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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
f3d2eac9-6e37-4d58-9f57-38de713d9f42 | ddg-da-data-distribution-generation-for | 2201.04038 | null | https://arxiv.org/abs/2201.04038v2 | https://arxiv.org/pdf/2201.04038v2.pdf | DDG-DA: Data Distribution Generation for Predictable Concept Drift Adaptation | In many real-world scenarios, we often deal with streaming data that is sequentially collected over time. Due to the non-stationary nature of the environment, the streaming data distribution may change in unpredictable ways, which is known as concept drift. To handle concept drift, previous methods first detect when/where the concept drift happens and then adapt models to fit the distribution of the latest data. However, there are still many cases that some underlying factors of environment evolution are predictable, making it possible to model the future concept drift trend of the streaming data, while such cases are not fully explored in previous work. In this paper, we propose a novel method DDG-DA, that can effectively forecast the evolution of data distribution and improve the performance of models. Specifically, we first train a predictor to estimate the future data distribution, then leverage it to generate training samples, and finally train models on the generated data. We conduct experiments on three real-world tasks (forecasting on stock price trend, electricity load and solar irradiance) and obtain significant improvement on multiple widely-used models. | ['Jiang Bian', 'Yingce Xia', 'Weiqing Liu', 'Xiao Yang', 'Wendi Li'] | 2022-01-11 | null | null | null | null | ['stock-trend-prediction', 'stock-market-prediction', 'stock-prediction'] | ['time-series', 'time-series', 'time-series'] | [ 1.36145324e-01 -6.04221761e-01 -6.05143793e-02 -5.72619498e-01
7.73812234e-02 -6.61529243e-01 3.58571768e-01 3.63066822e-01
-8.82828534e-02 7.00359762e-01 1.83605209e-01 -1.22624718e-01
6.99577332e-02 -9.23959315e-01 -7.23320782e-01 -8.21991682e-01
-2.31467411e-01 2.98094153e-01 2.61853635e-01 -3.25231194e-01
2.92670548e-01 1.45607784e-01 -1.74906158e+00 7.17369020e-02
1.06866360e+00 1.02864206e+00 2.11714178e-01 8.10855925e-01
-3.12726587e-01 7.07163572e-01 -1.04356480e+00 2.34064475e-01
1.41737968e-01 -3.39471310e-01 -6.28229231e-02 3.93631905e-02
-2.53725380e-01 -3.52584690e-01 -1.87659841e-02 9.91459668e-01
4.00570333e-01 -6.54567778e-02 5.46784461e-01 -1.24738085e+00
-2.71956593e-01 7.79120147e-01 -8.25545013e-01 7.22570539e-01
2.31390566e-01 -1.28308712e-02 4.82595831e-01 -2.89483905e-01
1.45855948e-01 1.15535605e+00 5.62481761e-01 4.44499403e-01
-8.14319253e-01 -8.25254798e-01 8.50183487e-01 7.19478205e-02
-1.05165029e+00 -1.28261805e-01 9.67457235e-01 -3.81751090e-01
6.26231611e-01 2.17529293e-02 8.29972029e-01 1.18325889e+00
5.32117605e-01 1.00234139e+00 7.92721987e-01 -4.76582944e-02
7.17586398e-01 -6.52691349e-02 -2.14634329e-01 -2.57087916e-01
2.72484422e-01 1.70415752e-02 -7.15884447e-01 -2.38529757e-01
1.17298909e-01 4.68045652e-01 -3.99717599e-01 1.09255970e-01
-8.49053979e-01 4.27490801e-01 2.75711060e-01 1.95091709e-01
-6.93069935e-01 -1.47050411e-01 3.40620369e-01 3.99101704e-01
8.99411380e-01 -9.84079391e-02 -7.42373526e-01 -4.06369388e-01
-9.13682818e-01 5.11279285e-01 9.66345191e-01 8.01438093e-01
4.87379223e-01 2.64306277e-01 7.55562484e-02 3.16898286e-01
2.53418773e-01 7.52016842e-01 1.07519960e+00 -3.22328925e-01
5.73261499e-01 5.02220631e-01 4.27468300e-01 -8.21681261e-01
-3.66051555e-01 -5.40459871e-01 -8.66002560e-01 -1.37592629e-01
1.89742193e-01 -5.68347037e-01 -8.58557224e-01 1.58033693e+00
5.54016054e-01 5.59000671e-01 1.36485353e-01 5.59121609e-01
-3.39630619e-02 8.30878973e-01 -1.87971681e-01 -6.30213022e-01
7.43555546e-01 -4.66444820e-01 -8.61504853e-01 -2.26690933e-01
3.48400384e-01 -4.40196395e-01 6.56942248e-01 6.10772371e-01
-7.99961567e-01 -4.92656738e-01 -1.11269593e+00 9.16085541e-01
-1.78434685e-01 -4.95122552e-01 3.87829393e-01 4.41028029e-01
-6.61892951e-01 5.85306585e-01 -1.25228179e+00 -1.77614674e-01
2.21930683e-01 -1.04217388e-01 6.87530935e-01 -1.10809889e-03
-1.29219055e+00 4.70677495e-01 7.58836508e-01 2.66287118e-01
-9.43706453e-01 -8.98741484e-01 -4.19295341e-01 8.95455703e-02
3.27323198e-01 -4.07966763e-01 1.49726427e+00 -1.04161441e+00
-1.38215840e+00 -9.12543610e-02 -3.11006904e-01 -8.10884535e-01
6.94217145e-01 -2.09092051e-01 -9.25014734e-01 -4.13389474e-01
-3.24874610e-01 -5.39324582e-02 9.84096766e-01 -1.06616974e+00
-1.21510828e+00 -4.26361024e-01 -2.17682704e-01 2.68686801e-01
-5.99913538e-01 -2.79376388e-01 1.96738586e-01 -5.56648314e-01
2.10517086e-02 -7.22372591e-01 -2.14994207e-01 -6.34704173e-01
4.48295772e-02 -2.99237251e-01 1.06115818e+00 -3.81576955e-01
1.61701560e+00 -2.29598260e+00 -3.93326610e-01 1.49410650e-01
-1.86197564e-01 1.99813381e-01 2.01254711e-01 6.59641623e-01
6.04465082e-02 -5.15771005e-03 -3.85564208e-01 -1.92283630e-01
-2.47856230e-01 5.69469810e-01 -9.39813972e-01 6.09910935e-02
2.73406580e-02 4.83976483e-01 -1.13483262e+00 2.80548990e-01
-2.15082526e-01 2.13403702e-01 -3.54012549e-02 2.11995095e-01
-8.93819749e-01 4.09463346e-01 -6.81971252e-01 4.88301873e-01
8.77035081e-01 -1.82798982e-01 6.00342900e-02 5.45198977e-01
-2.24997118e-01 1.48227155e-01 -1.33464336e+00 1.27340806e+00
-3.77959371e-01 5.57019532e-01 -3.66647571e-01 -1.06741428e+00
1.19085145e+00 3.05564731e-01 6.16536677e-01 -5.82363009e-01
-2.68971294e-01 2.28659213e-01 -1.34012522e-02 -5.84831536e-01
3.53443891e-01 -1.20694354e-01 2.94999659e-01 5.64660013e-01
-6.96863234e-01 6.58903793e-02 2.37549201e-01 -1.72115296e-01
1.06303644e+00 9.65216458e-02 1.13669388e-01 4.00977544e-02
3.99896204e-01 -1.17385894e-01 1.03049827e+00 5.34760237e-01
4.43898216e-02 3.46543878e-01 4.76620436e-01 -8.06360185e-01
-7.35218346e-01 -9.37285364e-01 1.13852639e-02 7.57872999e-01
9.09269750e-02 -2.64817208e-01 -4.81046498e-01 -4.74419266e-01
3.04853171e-01 1.06497288e+00 -4.85623896e-01 -2.43815526e-01
-3.38151723e-01 -1.45197439e+00 -4.51868847e-02 5.71163535e-01
4.01422232e-01 -9.97084022e-01 -9.39273834e-01 8.07426453e-01
1.20150493e-02 -8.52345228e-01 -2.15708852e-01 -4.78861406e-02
-1.06680405e+00 -9.38734949e-01 -2.99134642e-01 -2.02842698e-01
5.40824056e-01 3.31621230e-01 1.10495508e+00 6.36382680e-03
-9.99880508e-02 1.79036230e-01 -4.01248217e-01 -1.45054197e+00
-4.19410557e-01 4.91797626e-02 2.09193587e-01 2.78222233e-01
5.08130372e-01 -6.97471201e-01 -8.97536933e-01 1.45662814e-01
-1.31283176e+00 -4.02981937e-01 1.52679920e-01 4.99068081e-01
5.03606856e-01 1.01150882e+00 9.96489227e-01 -9.02294636e-01
6.95610046e-01 -8.86510730e-01 -8.01820397e-01 2.85670757e-01
-1.07046127e+00 -1.37464270e-01 7.02797830e-01 -6.64402008e-01
-1.19041371e+00 -1.46620780e-01 2.39928186e-01 -3.07150245e-01
-6.73072785e-02 9.65255976e-01 9.67378765e-02 6.83912277e-01
3.65123391e-01 5.25626659e-01 -1.99568614e-01 -4.83293742e-01
-1.90856934e-01 8.19544196e-01 4.16575819e-01 -4.09818530e-01
8.08726966e-01 6.77625418e-01 -1.57256529e-01 -6.99689031e-01
-1.04458463e+00 -4.20050234e-01 -3.56003940e-01 -1.02412067e-01
2.61298865e-01 -1.14054084e+00 -2.34635741e-01 9.57851529e-01
-8.51859987e-01 -3.13923895e-01 -2.57702112e-01 3.99548084e-01
-1.17900953e-01 -1.59440357e-02 -1.59571812e-01 -1.30153632e+00
-3.90596390e-01 -3.79980683e-01 6.47908330e-01 5.51809609e-01
5.30468188e-02 -1.35722077e+00 4.37532425e-01 -3.12281579e-01
7.03317165e-01 5.86042285e-01 7.36101985e-01 -7.48027265e-01
-4.63080853e-01 -4.91621912e-01 3.95691454e-01 4.03152972e-01
5.10565400e-01 2.75649369e-01 -8.95270050e-01 -5.77827692e-01
5.03182173e-01 3.39173496e-01 7.57607877e-01 3.79558802e-01
1.24780560e+00 -3.16278458e-01 -3.92852515e-01 5.22929728e-01
1.33158183e+00 5.05306900e-01 3.63012254e-01 3.91811609e-01
3.59967619e-01 5.03358006e-01 7.73394406e-01 1.01097977e+00
7.01859772e-01 2.92778850e-01 6.01245344e-01 2.75374591e-01
5.96530199e-01 -4.66735840e-01 6.05436683e-01 9.69888210e-01
2.29212120e-01 -4.81669873e-01 -9.37191248e-01 6.88696444e-01
-2.05165148e+00 -9.50594485e-01 7.48494714e-02 2.33581066e+00
7.30187476e-01 4.76337492e-01 2.67400056e-01 2.16036901e-01
5.54346502e-01 2.35936835e-01 -1.23771751e+00 -1.21090366e-02
-1.55997306e-01 -3.52825880e-01 5.14825702e-01 1.18323043e-01
-8.08392227e-01 3.31333876e-01 6.22021294e+00 9.84103605e-02
-1.56120062e+00 -2.52521962e-01 5.86966634e-01 -4.04484540e-01
-4.06989604e-01 -1.35135248e-01 -9.79233444e-01 1.09756875e+00
1.33102047e+00 -8.11988890e-01 3.43486220e-01 7.12500393e-01
5.08110166e-01 -2.64228415e-02 -1.02510035e+00 6.88238919e-01
6.39102608e-02 -8.55972409e-01 -1.85257211e-01 -7.47367293e-02
9.73001540e-01 3.38169128e-01 -6.36673048e-02 2.97402769e-01
3.26817691e-01 -5.72764337e-01 5.59433103e-01 7.94812083e-01
-3.26514542e-02 -8.14795732e-01 8.15928161e-01 9.34436083e-01
-1.07350361e+00 -4.53561991e-01 -4.23351228e-01 -2.36401170e-01
1.56199083e-01 1.08726835e+00 -8.35587084e-01 6.79542303e-01
9.63587105e-01 1.00049758e+00 -4.12971556e-01 1.25535417e+00
-2.55134225e-01 1.11188066e+00 -6.17667139e-01 -6.35064393e-02
-1.25596121e-01 -6.31400719e-02 5.37662029e-01 7.72836268e-01
7.47310698e-01 -1.31846219e-01 2.22301498e-01 4.25409526e-01
1.86245143e-01 -2.27115527e-01 -4.05099720e-01 7.50053376e-02
6.54585183e-01 8.27930450e-01 -5.34067690e-01 -3.59000802e-01
-4.30480599e-01 5.71001947e-01 -2.73673981e-01 6.09064102e-01
-6.69602573e-01 -1.05036050e-01 7.67279446e-01 1.51853099e-01
4.84190017e-01 -2.91657180e-01 -1.83008835e-01 -1.41993713e+00
5.11808455e-01 -6.65252447e-01 5.68168938e-01 -4.73355442e-01
-1.48592043e+00 3.56966794e-01 -1.42123457e-02 -1.41274273e+00
-7.41522133e-01 -1.50000706e-01 -1.21295106e+00 7.61582434e-01
-1.96973932e+00 -5.28083742e-01 -2.85414875e-01 4.04454768e-01
6.99328661e-01 -1.27999321e-01 5.21254957e-01 -1.22625999e-01
-7.07019448e-01 1.66948736e-02 8.26572835e-01 -2.32444197e-01
5.82262099e-01 -1.47359478e+00 8.17140937e-01 8.92593861e-01
-1.39786035e-01 2.23701119e-01 9.62254822e-01 -8.25494528e-01
-1.24179554e+00 -1.20480263e+00 4.94742155e-01 -3.18173796e-01
8.76074195e-01 -3.95702243e-01 -1.43159878e+00 5.40442765e-01
-5.41505106e-02 -3.62136625e-02 5.18474996e-01 -1.37093678e-01
-1.21858612e-01 -6.42358482e-01 -1.01881051e+00 1.96028814e-01
6.27723813e-01 4.46034111e-02 -4.23784703e-01 2.85017073e-01
7.80688107e-01 -5.03300607e-01 -7.64665365e-01 3.47023338e-01
3.43181252e-01 -7.28683591e-01 4.79511410e-01 -4.76746947e-01
1.06604621e-01 -4.79051620e-01 -2.62581017e-02 -1.87622321e+00
5.59178181e-02 -7.78856754e-01 -6.57762647e-01 1.50913131e+00
3.17909002e-01 -1.05242264e+00 6.90681875e-01 7.50076830e-01
2.11017638e-01 -5.09908259e-01 -8.13712001e-01 -8.70389998e-01
-2.67410409e-02 -2.48248175e-01 1.46253753e+00 7.87435889e-01
-2.86246032e-01 -8.03256482e-02 -2.27801442e-01 4.53746110e-01
5.38627148e-01 3.98462713e-01 6.77963138e-01 -1.33825457e+00
-2.42033646e-01 -2.74198711e-01 -2.68141419e-01 -8.85809600e-01
-9.68732908e-02 -2.12287605e-01 1.31139964e-01 -1.18258035e+00
-2.35078737e-01 -2.14795366e-01 -7.46687949e-01 4.87489365e-02
-5.59547544e-01 -6.53475165e-01 -9.59294438e-02 3.94071311e-01
-3.75722915e-01 8.38735342e-01 7.34315932e-01 -6.29260167e-02
-5.25648117e-01 8.14435184e-01 -5.47597826e-01 4.84215349e-01
1.03557277e+00 -4.98293906e-01 -8.11479867e-01 -4.55422670e-01
4.84483004e-01 -1.17692024e-01 -1.72884777e-01 -1.10189676e+00
2.96274304e-01 -5.34493566e-01 6.04396522e-01 -9.46568310e-01
-1.88118339e-01 -8.28171194e-01 1.63282961e-01 4.57293391e-01
-5.61477430e-02 4.27963793e-01 2.07084447e-01 9.68955338e-01
-4.87938374e-01 -1.11057989e-01 2.59763390e-01 -6.37007281e-02
-5.75468361e-01 6.48602366e-01 -2.53555715e-01 9.06395763e-02
1.02648985e+00 5.35889305e-02 -2.58270025e-01 -5.77743590e-01
-2.38771081e-01 9.16127384e-01 4.17797476e-01 7.78117120e-01
3.65954012e-01 -1.06905890e+00 -9.12152469e-01 3.15482289e-01
-2.19310485e-02 5.89718759e-01 2.62634009e-01 3.49730104e-01
-9.08073559e-02 2.65663471e-02 2.30423629e-01 -6.86001658e-01
-7.53824174e-01 7.66093850e-01 3.69826734e-01 -3.13302758e-03
-6.46220624e-01 3.82182360e-01 5.57995848e-02 2.45325081e-02
2.52532989e-01 -5.17407954e-01 -2.34378889e-01 3.84540051e-01
1.09335268e+00 2.04855636e-01 2.09823862e-01 2.72867233e-02
-1.71548933e-01 2.89561987e-01 -2.09874555e-01 2.10255310e-01
1.64912558e+00 -3.68850887e-01 -6.64927363e-02 1.23603570e+00
6.56446397e-01 -3.93179357e-01 -1.79299188e+00 -4.95102644e-01
4.11344111e-01 -4.66699541e-01 -1.03541002e-01 -6.79990113e-01
-1.02608395e+00 6.40255392e-01 7.54712403e-01 8.04286718e-01
1.46752274e+00 -5.73894501e-01 9.40261006e-01 1.27503052e-01
3.58163416e-01 -1.31582773e+00 -2.19988003e-01 3.86282116e-01
6.25552595e-01 -1.08631611e+00 -5.51607870e-02 1.34845033e-01
-6.64474070e-01 1.14790404e+00 5.77459931e-01 1.06683716e-01
1.04290271e+00 3.59242111e-01 2.87229747e-01 2.12740064e-01
-1.34582567e+00 1.65207967e-01 -2.32448757e-01 6.50009573e-01
1.86956793e-01 2.11797580e-01 1.50222167e-01 3.60859871e-01
-3.32700670e-01 3.52886319e-01 6.61895216e-01 1.13661408e+00
-4.11125183e-01 -1.03986692e+00 -3.41353297e-01 4.12340015e-01
-4.48510915e-01 2.87659615e-01 -5.42890690e-02 2.79275745e-01
-6.63877726e-02 9.43134725e-01 4.24833298e-01 -1.38938904e-01
5.74085653e-01 1.13135658e-01 -1.45981491e-01 -4.69072133e-01
-3.39870393e-01 1.60813946e-02 -4.12888139e-01 -1.23847261e-01
-3.28254640e-01 -1.26860571e+00 -1.23148930e+00 -2.28646860e-01
4.98259701e-02 6.34081215e-02 9.23304379e-01 1.08174086e+00
3.54802668e-01 6.29341006e-01 1.23650908e+00 -5.08759260e-01
-9.60759759e-01 -1.12104368e+00 -6.07297957e-01 4.25716996e-01
6.61579013e-01 -4.14857030e-01 -6.01820350e-01 1.14550494e-01] | [7.37805700302124, 2.9896442890167236] |
0ec3254e-1838-46c5-9b3a-9a6127330e00 | relational-deep-feature-learning-for | 2003.00697 | null | https://arxiv.org/abs/2003.00697v3 | https://arxiv.org/pdf/2003.00697v3.pdf | Relational Deep Feature Learning for Heterogeneous Face Recognition | Heterogeneous Face Recognition (HFR) is a task that matches faces across two different domains such as visible light (VIS), near-infrared (NIR), or the sketch domain. Due to the lack of databases, HFR methods usually exploit the pre-trained features on a large-scale visual database that contain general facial information. However, these pre-trained features cause performance degradation due to the texture discrepancy with the visual domain. With this motivation, we propose a graph-structured module called Relational Graph Module (RGM) that extracts global relational information in addition to general facial features. Because each identity's relational information between intra-facial parts is similar in any modality, the modeling relationship between features can help cross-domain matching. Through the RGM, relation propagation diminishes texture dependency without losing its advantages from the pre-trained features. Furthermore, the RGM captures global facial geometrics from locally correlated convolutional features to identify long-range relationships. In addition, we propose a Node Attention Unit (NAU) that performs node-wise recalibration to concentrate on the more informative nodes arising from relation-based propagation. Furthermore, we suggest a novel conditional-margin loss function (C-softmax) for the efficient projection learning of the embedding vector in HFR. The proposed method outperforms other state-of-the-art methods on five HFR databases. Furthermore, we demonstrate performance improvement on three backbones because our module can be plugged into any pre-trained face recognition backbone to overcome the limitations of a small HFR database. | ['Ig-Jae Kim', 'Taeoh Kim', 'Kyungjae Lee', 'Sangyoun Lee', 'MyeongAh Cho'] | 2020-03-02 | null | null | null | null | ['heterogeneous-face-recognition'] | ['computer-vision'] | [ 1.29533917e-01 -5.97146526e-02 -2.55362123e-01 -5.74065745e-01
-4.96371120e-01 -1.05609760e-01 5.38168669e-01 -4.21466172e-01
3.54894958e-02 1.90176010e-01 1.81260154e-01 9.97711271e-02
-1.24596924e-01 -1.04397845e+00 -6.90036535e-01 -5.98287106e-01
1.31517932e-01 1.03184037e-01 -3.40200998e-02 -4.25355852e-01
-1.36447355e-01 1.08664274e+00 -1.71238601e+00 5.78284442e-01
4.93900955e-01 1.40420318e+00 1.28556848e-01 6.16176128e-02
-1.99166939e-01 4.96073782e-01 -1.20850071e-01 -7.02401102e-01
5.46814680e-01 -2.60996133e-01 -3.03870469e-01 5.42588197e-02
8.91757905e-01 -2.30324790e-01 -9.27748919e-01 9.52396452e-01
6.18733168e-01 2.80030351e-02 5.24453044e-01 -1.41559672e+00
-8.14915061e-01 3.62992585e-02 -8.97499382e-01 -3.35420519e-01
5.09501040e-01 1.39757842e-01 9.22909498e-01 -1.31611490e+00
8.03306103e-01 1.54841256e+00 7.88686693e-01 7.01040387e-01
-1.07209432e+00 -1.05662823e+00 1.86448649e-01 3.54787588e-01
-1.75064564e+00 -6.84577465e-01 1.37758696e+00 -2.55243778e-01
8.59180093e-01 2.35366791e-01 5.08652627e-01 1.06451690e+00
-4.93887514e-02 3.86904180e-01 9.19309378e-01 -1.31082505e-01
-3.19551796e-01 -2.52027027e-02 -3.34364474e-01 1.04899800e+00
-2.39824876e-02 1.83015615e-01 -8.90063643e-01 -1.06616221e-01
8.82032990e-01 2.81236380e-01 -3.79469216e-01 -3.97917122e-01
-6.41178310e-01 5.99866033e-01 7.75659025e-01 1.08752936e-01
-2.92592555e-01 -4.06812690e-02 1.23045117e-01 3.50458443e-01
4.74927723e-01 -5.72386794e-02 -2.04797417e-01 6.11411750e-01
-8.39953005e-01 -2.99284339e-01 5.40585756e-01 8.62173319e-01
1.18885338e+00 -7.63331428e-02 -2.66401887e-01 1.25908625e+00
7.14171946e-01 5.39974511e-01 3.95866185e-02 -6.00107074e-01
6.34209633e-01 9.00845408e-01 -5.47705054e-01 -1.49696088e+00
-2.04039007e-01 -3.53987783e-01 -1.05115271e+00 1.66250661e-01
1.22472517e-01 2.74806142e-01 -9.62234139e-01 1.64351439e+00
4.41056520e-01 2.19361290e-01 -6.32301345e-02 1.07195139e+00
1.46649218e+00 3.54104549e-01 -1.10748000e-01 5.18874377e-02
1.25461257e+00 -7.89463580e-01 -4.46859390e-01 -1.08535111e-01
2.51986891e-01 -8.85158658e-01 7.30836391e-01 -9.41808820e-02
-6.66886151e-01 -7.22655356e-01 -8.94523025e-01 -2.62869805e-01
-3.96333992e-01 2.63143539e-01 6.11427069e-01 4.99445498e-01
-1.06107581e+00 5.39024174e-01 -3.70099753e-01 -3.84711593e-01
7.71461606e-01 5.61278224e-01 -9.73528385e-01 -6.84201896e-01
-1.03034258e+00 5.85338771e-01 -1.64284840e-01 5.67071795e-01
-6.46486998e-01 -7.83565700e-01 -1.03468847e+00 9.11540166e-02
3.12996507e-01 -4.98460650e-01 1.76772997e-01 -1.04840064e+00
-1.60972238e+00 1.08750296e+00 -3.10467303e-01 3.59196156e-01
3.76643360e-01 2.13327661e-01 -5.95457315e-01 3.66159290e-01
-2.31215075e-01 6.52356029e-01 9.94316280e-01 -1.20523119e+00
9.10612270e-02 -7.51219153e-01 1.75936781e-02 1.09468453e-01
-5.75068295e-01 1.43439636e-01 -9.20856893e-01 -7.40603566e-01
2.93519676e-01 -8.86544168e-01 1.33827761e-01 4.78585631e-01
-2.73180485e-01 -2.52273649e-01 9.35442746e-01 -7.45532990e-01
7.78151751e-01 -2.39423704e+00 1.95005052e-02 5.00985682e-01
2.21849963e-01 1.61517218e-01 -7.37265706e-01 3.53591591e-01
-3.72270674e-01 -1.99787915e-01 7.45153502e-02 -2.80821294e-01
-1.93397149e-01 3.14586870e-02 -6.55933172e-02 6.96823239e-01
4.84583765e-01 1.03261220e+00 -5.77803969e-01 -5.82601190e-01
2.25411937e-01 9.66686070e-01 -5.02066433e-01 2.32416347e-01
2.24483735e-03 3.86295706e-01 -3.32535118e-01 1.03231144e+00
1.18119800e+00 -1.62035644e-01 2.30770051e-01 -8.27292979e-01
1.84015214e-01 -5.86098135e-02 -9.65306640e-01 1.88302159e+00
-5.00402808e-01 5.03543735e-01 2.14358717e-01 -1.03136587e+00
1.35765922e+00 2.10154116e-01 5.88526547e-01 -8.22854280e-01
7.77278990e-02 7.45853111e-02 -1.93709925e-01 -3.72377485e-01
1.87990367e-02 1.27942547e-01 4.76470947e-01 6.34141341e-02
2.83171296e-01 4.41271573e-01 -3.92850041e-01 4.80985120e-02
1.04223657e+00 1.87146515e-01 -8.43389928e-02 -4.95400615e-02
7.60282218e-01 -5.10094464e-01 7.59054959e-01 2.44415790e-01
-2.78658909e-03 7.81585813e-01 2.86802560e-01 -4.69281822e-01
-5.61430156e-01 -9.03754532e-01 -2.17795908e-01 1.01480758e+00
3.26238960e-01 -6.23065948e-01 -1.63065687e-01 -8.47829640e-01
3.13501865e-01 -1.46783501e-01 -8.47638249e-01 -2.45436415e-01
-5.31237245e-01 -5.03546059e-01 4.10348207e-01 4.60584670e-01
7.79754937e-01 -8.77290368e-01 2.41410658e-01 -2.57795155e-01
7.37407953e-02 -1.24799955e+00 -5.40534914e-01 -3.36922199e-01
-5.21059871e-01 -9.99748111e-01 -6.28070116e-01 -7.54913330e-01
9.08328891e-01 5.11490226e-01 8.46625149e-01 2.74015754e-01
-5.36870301e-01 4.67253089e-01 -1.77580357e-01 1.12318978e-01
7.75559172e-02 -1.31624147e-01 -2.03950137e-01 6.36005342e-01
4.57930028e-01 -6.82719588e-01 -7.10205495e-01 5.43208361e-01
-4.74585831e-01 7.15864971e-02 6.49694741e-01 9.22615588e-01
5.54750085e-01 -2.12947354e-01 3.29519868e-01 -7.13539839e-01
5.86344227e-02 -4.16690171e-01 -4.78857607e-01 6.00134790e-01
-2.65859783e-01 -6.99857250e-02 5.91759264e-01 -4.90917861e-01
-1.16095114e+00 2.53979087e-01 2.01617870e-02 -9.66736615e-01
-3.41881961e-02 3.06612104e-01 -7.58237123e-01 -6.90330684e-01
3.99178654e-01 1.65641174e-01 9.35494304e-02 -5.19277871e-01
2.44405866e-01 5.63340247e-01 4.09371614e-01 -6.06003582e-01
1.10491395e+00 5.49530208e-01 4.40573931e-01 -9.31385219e-01
-5.58868587e-01 -4.22725230e-01 -5.17284036e-01 -3.89845133e-01
5.48640072e-01 -1.14357328e+00 -8.07007849e-01 4.18967545e-01
-1.12532616e+00 2.71492377e-02 6.14434332e-02 3.90693307e-01
-1.01076372e-01 2.87197709e-01 -5.83348572e-01 -6.82841063e-01
-2.42687896e-01 -9.97982562e-01 1.19092822e+00 3.36285681e-01
4.24183249e-01 -6.54909670e-01 -2.21024573e-01 5.05570590e-01
3.55933368e-01 2.04902917e-01 6.68288171e-01 -2.50084460e-01
-8.85304093e-01 -1.57707766e-01 -6.90106690e-01 2.93814212e-01
3.65704715e-01 1.65003732e-01 -1.28414857e+00 -4.20687735e-01
-4.49061036e-01 -3.51436794e-01 1.10537577e+00 -7.81023651e-02
1.29680598e+00 -8.88908654e-02 -4.48937744e-01 1.07542121e+00
1.43476617e+00 -1.34547383e-01 6.87236786e-01 -3.11732702e-02
1.20127761e+00 9.60389733e-01 3.83025676e-01 2.70274282e-01
3.41991991e-01 9.83200431e-01 3.78044873e-01 -2.85140157e-01
-5.38892090e-01 -4.73173827e-01 4.45897073e-01 5.27429402e-01
-3.31920713e-01 1.88482985e-01 -6.10036671e-01 1.89944476e-01
-1.75175822e+00 -9.46124554e-01 1.93404362e-01 2.25050712e+00
6.43682182e-01 -4.47676390e-01 -3.00489575e-01 -2.97532290e-01
8.14742684e-01 3.03622812e-01 -5.23778498e-01 3.32715409e-03
-4.20711786e-01 3.84041101e-01 3.51685196e-01 2.05569118e-01
-8.72712731e-01 1.03548837e+00 4.82106066e+00 9.80330765e-01
-1.33919251e+00 -1.78444553e-02 4.64919329e-01 1.52082136e-02
-4.64132547e-01 -1.24093674e-01 -8.27457905e-01 2.47342110e-01
3.19340885e-01 3.33527714e-01 6.50609791e-01 7.75405526e-01
-2.21146837e-01 3.88112873e-01 -1.16210687e+00 1.43152726e+00
4.03452128e-01 -1.19726014e+00 2.02618673e-01 1.66658819e-01
6.30676389e-01 -8.40218812e-02 1.80588782e-01 1.61469191e-01
-7.78514668e-02 -1.28964651e+00 2.04419911e-01 5.34870386e-01
1.24274242e+00 -8.28602910e-01 5.20136237e-01 -2.59223849e-01
-1.64475644e+00 7.37251490e-02 -6.87485933e-01 3.25388819e-01
-3.51162434e-01 5.12187600e-01 -6.01182163e-01 8.79990339e-01
7.15355456e-01 9.26682413e-01 -7.51993597e-01 6.85483813e-01
-1.85273603e-01 2.11357087e-01 -4.55954522e-01 4.68711644e-01
-2.19656348e-01 -3.30181897e-01 3.26979220e-01 9.22050714e-01
2.06443951e-01 5.80940619e-02 5.04725054e-02 8.39344978e-01
-5.18848658e-01 2.33641952e-01 -7.91973233e-01 1.30560264e-01
4.22380716e-01 1.59149015e+00 -4.16085124e-01 1.32096067e-01
-8.10014904e-01 1.13479471e+00 5.20928204e-01 5.28769970e-01
-6.78736329e-01 -2.49606088e-01 7.92386115e-01 2.71093398e-01
3.66175264e-01 7.36462995e-02 8.18032250e-02 -1.21410227e+00
2.70934999e-01 -7.59676099e-01 2.51912326e-01 -6.79735720e-01
-1.58586872e+00 7.82267153e-01 -3.39512080e-01 -1.09946573e+00
2.54549026e-01 -5.94044983e-01 -4.80686635e-01 1.15119159e+00
-1.88284075e+00 -1.68397045e+00 -6.94356561e-01 1.06169689e+00
1.27582336e-02 -5.61236262e-01 7.54011095e-01 6.45143449e-01
-8.16410124e-01 1.24799418e+00 -3.22360605e-01 5.55663228e-01
1.03813338e+00 -5.22401631e-01 7.28768036e-02 5.41373670e-01
1.94647253e-01 7.96801448e-01 -5.53450622e-02 -6.74716830e-01
-1.78297532e+00 -1.34426594e+00 6.46755278e-01 4.13962966e-03
3.46331745e-01 -5.65490842e-01 -1.03810072e+00 4.54019397e-01
-1.84329048e-01 8.87997866e-01 7.34971106e-01 2.76303798e-01
-1.16288579e+00 -6.94416523e-01 -1.17023945e+00 4.47028607e-01
1.48534620e+00 -1.11335707e+00 -9.12491605e-02 2.03158975e-01
4.27205980e-01 -2.25998029e-01 -1.08533156e+00 6.79852962e-01
8.37181449e-01 -9.16385055e-01 1.12128651e+00 -5.17672122e-01
4.43605065e-01 -3.13679636e-01 -3.59832823e-01 -9.17213261e-01
-3.82121921e-01 -6.12252653e-01 -4.39077653e-02 1.58107328e+00
1.16572171e-01 -7.23443747e-01 9.24046040e-01 5.58288991e-01
1.31393820e-01 -8.64544094e-01 -9.37251389e-01 -6.70047283e-01
-4.71401095e-01 -2.11140230e-01 7.55756199e-01 1.15683043e+00
-4.42171991e-01 3.08448613e-01 -4.00873691e-01 3.96952361e-01
7.71012664e-01 4.46898013e-01 6.98807657e-01 -1.29159701e+00
-1.92454785e-01 -2.23166570e-01 -7.39603460e-01 -8.19036603e-01
5.38009882e-01 -1.21438134e+00 -3.44044358e-01 -1.15227723e+00
3.22264254e-01 -7.02479064e-01 -4.41583633e-01 7.84477353e-01
-7.06702657e-03 5.95367312e-01 3.17018211e-01 1.43900216e-01
-4.02974367e-01 8.56495798e-01 1.42660010e+00 -2.89443105e-01
-1.76448636e-02 -3.26375276e-01 -4.68761146e-01 4.91671503e-01
4.96134937e-01 -2.83166707e-01 -3.01520079e-01 -3.87441486e-01
1.19754961e-02 8.57097134e-02 5.66397607e-01 -7.86913991e-01
4.22795385e-01 -1.60057545e-01 7.95967758e-01 -4.29988652e-01
6.01029754e-01 -1.00869167e+00 3.68637621e-01 -6.99657947e-02
-5.81992380e-02 -2.35071719e-01 7.92017058e-02 5.18954039e-01
-2.80430794e-01 3.07502359e-01 9.01318192e-01 1.38177052e-01
-6.01598620e-01 8.90380561e-01 3.20728809e-01 -2.31458098e-01
8.44931483e-01 -4.28949207e-01 -3.46178263e-01 -1.51835650e-01
-3.60235631e-01 2.96629034e-03 5.72829545e-01 7.24458754e-01
9.56158519e-01 -1.69050348e+00 -6.87218666e-01 6.90394521e-01
4.61711735e-01 -1.63121983e-01 5.59247315e-01 6.68266118e-01
-2.14193940e-01 1.33914858e-01 -5.09144127e-01 -5.01587570e-01
-1.49621344e+00 5.16410589e-01 3.75069410e-01 4.06414829e-02
-5.93028843e-01 9.37769532e-01 4.86551762e-01 -4.54668611e-01
2.19726518e-01 3.38196397e-01 -2.84690917e-01 4.52793725e-02
4.80913401e-01 1.21249005e-01 1.43755034e-01 -1.07461882e+00
-6.66356146e-01 1.05959511e+00 3.10558099e-02 2.98011780e-01
1.33128607e+00 1.52297810e-01 -4.14649785e-01 -1.49068415e-01
1.68505394e+00 1.35616392e-01 -1.20605505e+00 -5.32600105e-01
-5.09183526e-01 -7.43998528e-01 1.36697054e-01 -4.74352837e-01
-1.76056588e+00 8.69507134e-01 6.52256191e-01 -5.21986723e-01
1.38867104e+00 -1.78596769e-02 5.33340275e-01 2.40743652e-01
3.62470359e-01 -7.58980334e-01 1.45915821e-01 1.75403953e-01
1.16996729e+00 -1.32859921e+00 6.33596331e-02 -9.26591575e-01
-3.03486109e-01 1.26694429e+00 9.06469941e-01 3.48933600e-02
9.57410812e-01 9.69807431e-02 -9.37431455e-02 -3.22951466e-01
-5.77019393e-01 -2.96413988e-01 7.49285817e-01 8.16072702e-01
3.48733753e-01 -1.50363892e-01 1.23217337e-01 5.31058490e-01
8.89253765e-02 -6.85076639e-02 -2.59121656e-01 4.50412363e-01
8.77202228e-02 -1.20793784e+00 -3.26139599e-01 4.16639060e-01
-2.53123969e-01 -1.05736434e-01 -5.46722829e-01 6.04779243e-01
2.70433784e-01 8.05518150e-01 -3.24816518e-02 -7.59936035e-01
2.28850767e-01 -1.39534160e-01 7.79480994e-01 -4.75492358e-01
-4.17559177e-01 2.60338429e-02 -1.31421059e-01 -9.32020128e-01
-3.77541959e-01 -2.01924056e-01 -8.70236814e-01 -5.49376845e-01
-2.96900153e-01 -2.33094618e-01 4.94297087e-01 6.20545208e-01
6.73642039e-01 2.48997092e-01 8.92811239e-01 -7.54416406e-01
-3.56457792e-02 -8.05529892e-01 -6.50094926e-01 5.47338545e-01
1.39172539e-01 -9.04831052e-01 -1.30890355e-01 -1.98922440e-01] | [13.114914894104004, 0.5339385271072388] |
db530c71-ee7f-4275-9c20-508c6d1b67cd | one-class-one-click-quasi-scene-level-weakly | 2211.12657 | null | https://arxiv.org/abs/2211.12657v1 | https://arxiv.org/pdf/2211.12657v1.pdf | One Class One Click: Quasi Scene-level Weakly Supervised Point Cloud Semantic Segmentation with Active Learning | Reliance on vast annotations to achieve leading performance severely restricts the practicality of large-scale point cloud semantic segmentation. For the purpose of reducing data annotation costs, effective labeling schemes are developed and contribute to attaining competitive results under weak supervision strategy. Revisiting current weak label forms, we introduce One Class One Click (OCOC), a low cost yet informative quasi scene-level label, which encapsulates point-level and scene-level annotations. An active weakly supervised framework is proposed to leverage scarce labels by involving weak supervision from global and local perspectives. Contextual constraints are imposed by an auxiliary scene classification task, respectively based on global feature embedding and point-wise prediction aggregation, which restricts the model prediction merely to OCOC labels. Furthermore, we design a context-aware pseudo labeling strategy, which effectively supplement point-level supervisory signals. Finally, an active learning scheme with a uncertainty measure - temporal output discrepancy is integrated to examine informative samples and provides guidance on sub-clouds query, which is conducive to quickly attaining desirable OCOC annotations and reduces the labeling cost to an extremely low extent. Extensive experimental analysis using three LiDAR benchmarks collected from airborne, mobile and ground platforms demonstrates that our proposed method achieves very promising results though subject to scarce labels. It considerably outperforms genuine scene-level weakly supervised methods by up to 25\% in terms of average F1 score and achieves competitive results against full supervision schemes. On terrestrial LiDAR dataset - Semantics3D, using approximately 2\textpertenthousand{} of labels, our method achieves an average F1 score of 85.2\%, which increases by 11.58\% compared to the baseline model. | ['Jie Shao', 'Wei Yao', 'Puzuo Wang'] | 2022-11-23 | null | null | null | null | ['scene-classification'] | ['computer-vision'] | [ 5.24019003e-01 3.23642969e-01 -6.57799363e-01 -5.90232074e-01
-1.46301305e+00 -5.48645496e-01 5.41877568e-01 3.10798347e-01
-2.95968324e-01 5.97066462e-01 -3.22150469e-01 -1.67951643e-01
-2.35878468e-01 -7.88971305e-01 -8.56641650e-01 -7.65379190e-01
4.98752259e-02 4.92175281e-01 5.23477852e-01 1.42245442e-01
1.90962940e-01 3.97622257e-01 -1.73242247e+00 -6.83960365e-03
1.32255125e+00 1.32180023e+00 4.86997515e-01 1.27571464e-01
-3.79337281e-01 5.64215481e-01 -1.91953182e-01 -3.27632517e-01
4.62025344e-01 1.63974062e-01 -6.88065827e-01 4.80005443e-01
6.95856512e-01 -1.11662224e-01 2.51981378e-01 1.09124231e+00
3.03158134e-01 -3.60558517e-02 5.68611860e-01 -1.10567725e+00
-2.69405872e-01 2.27006599e-01 -6.26128078e-01 -1.37300059e-01
-1.39867067e-02 2.23547250e-01 1.40823114e+00 -1.16373014e+00
3.61914158e-01 8.96082461e-01 6.64617836e-01 2.71794230e-01
-1.36099398e+00 -6.20285690e-01 6.68264449e-01 -1.63728952e-01
-1.56039560e+00 -2.59682238e-01 9.71698642e-01 -4.75097418e-01
5.97230077e-01 3.81781220e-01 3.72673512e-01 6.64516568e-01
-4.73231763e-01 9.89282787e-01 1.08667469e+00 -1.92046300e-01
4.62713480e-01 4.02848154e-01 1.28589943e-01 6.33849263e-01
4.19690870e-02 -1.90335698e-02 -5.53341568e-01 -9.44495350e-02
4.45004046e-01 -7.58287236e-02 -1.59081027e-01 -6.38682008e-01
-1.05260396e+00 6.48326099e-01 5.67186415e-01 -3.14522684e-02
-2.15591177e-01 -6.47926703e-02 1.68654591e-01 -9.97037962e-02
8.28251541e-01 2.82228261e-01 -5.51873922e-01 1.60894528e-01
-9.79173183e-01 -4.95209098e-02 3.24053615e-01 1.41044831e+00
1.12612462e+00 1.37505278e-01 -1.63660437e-01 9.49612319e-01
5.02811909e-01 7.18259454e-01 -2.82962233e-01 -1.14857090e+00
6.37511611e-01 9.90039408e-01 1.86356872e-01 -6.83539391e-01
3.01342010e-02 -8.83910477e-01 -4.23604250e-01 9.04998109e-02
-2.17164997e-02 3.10601085e-01 -1.00281048e+00 1.39895570e+00
6.28382981e-01 2.92545229e-01 -1.66342601e-01 8.82783055e-01
6.70693278e-01 4.29284215e-01 2.96570659e-01 -3.43937218e-01
1.09768426e+00 -8.67683053e-01 -4.96245801e-01 -2.15473667e-01
8.53519917e-01 -5.63016474e-01 1.46452367e+00 3.16056997e-01
-8.13500702e-01 -7.41078317e-01 -1.06506729e+00 1.44646928e-01
-3.07553053e-01 3.14989656e-01 6.11901343e-01 6.03326619e-01
-9.11651075e-01 2.68741906e-01 -6.97950959e-01 -4.26959209e-02
8.56700540e-01 4.58315253e-01 -3.58617716e-02 -1.78892270e-01
-8.73187363e-01 4.39287156e-01 3.56489450e-01 1.54229715e-01
-9.69504476e-01 -9.28017259e-01 -9.20506716e-01 -2.00942948e-01
7.95074940e-01 -3.27076733e-01 9.68882799e-01 -5.68122685e-01
-1.26120567e+00 9.10533309e-01 -2.00743258e-01 -4.12594378e-01
6.06430948e-01 -3.68014127e-01 -3.14208567e-01 3.51380944e-01
6.35558367e-01 1.06879246e+00 7.09067464e-01 -1.64625812e+00
-9.64439452e-01 -4.19409275e-01 1.33820534e-01 4.68725532e-01
-3.57571274e-01 -3.97454798e-01 -7.18013763e-01 -3.79774183e-01
4.36679900e-01 -8.64146113e-01 -4.18811083e-01 2.46866614e-01
-3.33821058e-01 -4.66944903e-01 1.11542296e+00 -5.44525608e-02
8.75596762e-01 -2.18037677e+00 -1.48007095e-01 1.05931617e-01
2.17137083e-01 5.45094050e-02 9.67173874e-02 1.14185080e-01
1.59558713e-01 1.45349219e-01 -8.61382425e-01 -7.28470445e-01
3.41880992e-02 3.81662428e-01 -5.87331355e-01 3.58309627e-01
6.77619219e-01 7.92912245e-01 -1.09397554e+00 -8.25345635e-01
4.82567579e-01 1.67790219e-01 -5.41129172e-01 2.41528839e-01
-5.84512949e-01 6.31025374e-01 -6.96130216e-01 1.10534489e+00
8.01303983e-01 -3.92569512e-01 -8.03759024e-02 -7.21088126e-02
-2.97334850e-01 1.27686501e-01 -1.09699941e+00 1.91718912e+00
-5.74753344e-01 8.42357576e-02 1.01180367e-01 -1.13241720e+00
1.14583516e+00 9.96324420e-02 6.29842639e-01 -5.40526628e-01
-1.61419272e-01 3.90252590e-01 -6.58349812e-01 -2.18611851e-01
4.48844582e-01 3.34830619e-02 -2.46042579e-01 -7.67577812e-02
5.65399677e-02 -6.02169871e-01 -1.42132761e-02 3.09268802e-01
8.51808012e-01 5.92816770e-01 1.61270965e-02 -4.24032003e-01
6.55989289e-01 2.24379167e-01 8.34212482e-01 6.96242034e-01
-3.64351243e-01 5.83089590e-01 1.11698575e-01 -8.13280493e-02
-5.82785964e-01 -1.12926412e+00 -6.37606680e-01 1.02518737e+00
6.35383546e-01 -3.36141795e-01 -4.50761378e-01 -9.22970355e-01
6.52289018e-03 7.24503160e-01 -2.86891073e-01 1.87046707e-01
-4.81246084e-01 -5.82696140e-01 3.29975039e-01 6.07441843e-01
8.18205059e-01 -6.25965655e-01 -2.52349019e-01 6.30796254e-02
-1.74315304e-01 -1.45037627e+00 -4.22606990e-02 4.67986703e-01
-1.12167919e+00 -7.98894644e-01 -4.11505938e-01 -6.49960399e-01
6.43926084e-01 4.45528001e-01 9.58676815e-01 -1.38401955e-01
-7.89014548e-02 3.18776339e-01 -3.81323755e-01 -2.88953960e-01
2.51482666e-01 2.20526263e-01 -2.38755718e-02 9.69202295e-02
2.96586394e-01 -6.50826514e-01 -5.74687839e-01 5.18420577e-01
-7.26055801e-01 -1.04701303e-01 6.58326089e-01 6.73989415e-01
1.15551007e+00 -1.21636681e-01 6.12606943e-01 -1.00905621e+00
-2.39310101e-01 -4.89654690e-01 -8.28230500e-01 1.47594824e-01
-8.94840717e-01 -2.47738764e-01 5.03520548e-01 -5.24044447e-02
-1.18644500e+00 5.32846868e-01 -6.57433495e-02 -6.02198660e-01
-3.44199330e-01 2.42494699e-02 -5.19806921e-01 -1.61726430e-01
5.67795873e-01 2.34231561e-01 -4.74065602e-01 -4.14027929e-01
5.48488557e-01 7.39872873e-01 4.88179028e-01 -9.81438577e-01
1.00957811e+00 8.19557488e-01 -3.85457948e-02 -7.16953576e-01
-1.21228659e+00 -9.19408083e-01 -8.66333723e-01 -3.04059923e-01
7.00736046e-01 -1.35105824e+00 -4.74938542e-01 7.08463192e-02
-7.50025868e-01 -1.39373183e-01 -6.38899922e-01 4.25251156e-01
-5.85589409e-01 3.74606788e-01 -1.07503735e-01 -1.06882703e+00
-6.93654865e-02 -1.18237054e+00 1.78300858e+00 -4.28094789e-02
1.04946800e-01 -7.45050609e-01 -2.50705093e-01 8.17260921e-01
-2.88364012e-02 2.21272007e-01 6.22830451e-01 -5.49657583e-01
-1.25889266e+00 -1.06223278e-01 -3.60416770e-01 5.34802139e-01
5.42018265e-02 -2.17728019e-01 -1.35518265e+00 -8.26267973e-02
-6.70306804e-03 -5.46972275e-01 8.71308088e-01 5.80719002e-02
1.37477338e+00 1.28423780e-01 -5.03167808e-01 4.65122700e-01
1.46830547e+00 1.47822514e-01 2.28718460e-01 -4.25227918e-03
9.30918396e-01 8.31249297e-01 1.24634469e+00 3.31850708e-01
4.65845078e-01 6.06981337e-01 8.19019556e-01 -1.21765636e-01
-7.81323314e-02 -4.08678502e-01 9.97427553e-02 8.32951307e-01
3.92111950e-02 -1.37364060e-01 -9.97065663e-01 6.78667307e-01
-1.84024549e+00 -4.79437351e-01 -3.30169290e-01 2.19132757e+00
7.60684192e-01 5.06436229e-01 -1.53491050e-01 2.78620422e-01
6.72158122e-01 4.35677648e-01 -6.27491891e-01 2.56864280e-01
-1.05389804e-01 1.01605631e-01 5.93305647e-01 4.53984320e-01
-1.30852258e+00 1.15584159e+00 5.01875210e+00 1.32535911e+00
-9.63622868e-01 3.16374034e-01 6.14867270e-01 4.45766263e-02
-4.26406145e-01 2.00948820e-01 -1.03194106e+00 5.41251361e-01
5.60819328e-01 4.11361784e-01 -1.03510022e-01 1.35086596e+00
2.20012203e-01 -1.03061005e-01 -9.82223451e-01 8.46292734e-01
-2.02173308e-01 -1.25498390e+00 2.30996148e-03 3.18434358e-01
7.84681082e-01 1.65376976e-01 1.21875532e-01 3.05000842e-01
2.50532746e-01 -6.00417197e-01 8.34649026e-01 2.50462562e-01
1.08141887e+00 -5.94465554e-01 5.61451852e-01 6.69983625e-01
-1.60061049e+00 -1.34103924e-01 -2.86539525e-01 6.83518425e-02
2.37476274e-01 8.37615848e-01 -7.37716198e-01 9.65185344e-01
7.40963459e-01 9.58848536e-01 -3.89815420e-01 6.45213544e-01
-2.90325552e-01 8.41838896e-01 -5.12545466e-01 2.86924481e-01
4.23485041e-01 -3.02594841e-01 5.12375534e-01 1.00277460e+00
3.76398377e-02 2.71313399e-01 6.79098666e-01 8.48035038e-01
-6.78371415e-02 1.88992307e-01 -6.09816849e-01 2.41352186e-01
7.35612631e-01 1.30445302e+00 -8.81998658e-01 -3.28426093e-01
-2.69869775e-01 7.32582450e-01 1.62824288e-01 2.34641612e-01
-8.97890568e-01 -6.24677390e-02 2.90583998e-01 1.31968990e-01
2.55093962e-01 -1.55597806e-01 -6.31073713e-01 -1.06199610e+00
3.26711953e-01 -2.90759832e-01 2.35180736e-01 -5.70840240e-01
-1.09107351e+00 4.48190272e-01 -1.06287664e-02 -1.78284252e+00
2.02125669e-01 -3.86089802e-01 -4.80354339e-01 6.43959820e-01
-1.88627815e+00 -1.48185802e+00 -5.55935025e-01 4.02678400e-01
8.66467476e-01 1.26775444e-01 5.83614588e-01 4.84082609e-01
-5.19637704e-01 2.79449463e-01 -1.91550702e-01 -1.95777982e-01
4.70077455e-01 -1.35351384e+00 -2.23441888e-02 8.13240647e-01
2.89041370e-01 3.88973236e-01 3.01101327e-01 -6.73639655e-01
-9.94014204e-01 -1.47360420e+00 7.27341056e-01 -7.14514911e-01
5.37803471e-01 -5.66590428e-01 -8.36243570e-01 3.63535225e-01
-3.89413685e-01 4.41304088e-01 5.60776711e-01 -3.60266827e-02
-1.91064045e-01 -4.02563304e-01 -1.12599897e+00 2.64015943e-01
1.51214850e+00 -7.30377734e-01 -4.64189500e-01 6.01641417e-01
1.22091591e+00 -2.23494709e-01 -8.52299929e-01 1.09666944e+00
1.78569201e-02 -8.46566260e-01 1.03152919e+00 -4.68694419e-02
2.69758612e-01 -6.70821846e-01 -5.64346969e-01 -5.89864910e-01
-1.19250715e-01 -3.79100442e-01 -9.97723714e-02 1.57237053e+00
3.63631725e-01 -4.88217026e-01 1.10561097e+00 3.29801887e-01
-6.90186381e-01 -1.03563797e+00 -9.80266511e-01 -8.75019789e-01
-2.79618233e-01 -8.23885977e-01 4.98625606e-01 1.07896006e+00
-5.69601536e-01 2.83276260e-01 -3.15275677e-02 5.34114718e-01
9.22057390e-01 3.42987835e-01 7.02237904e-01 -1.40905428e+00
3.73513368e-03 -7.17641339e-02 -2.77433097e-01 -1.46316969e+00
2.33925179e-01 -8.83877456e-01 1.66069210e-01 -1.22915673e+00
-7.97514245e-02 -1.00008595e+00 -3.86050373e-01 4.43433523e-01
-1.14924781e-01 4.88807619e-01 8.13023187e-03 5.68033338e-01
-8.33474219e-01 8.25974762e-01 1.04675019e+00 -2.92442769e-01
-3.10472548e-02 1.62327230e-01 -5.40516675e-01 7.95713902e-01
4.66456801e-01 -4.06672359e-01 -7.73126125e-01 -2.78927803e-01
1.89276543e-02 -1.20084457e-01 3.59331012e-01 -9.47985113e-01
1.30940571e-01 -1.44016936e-01 3.96060124e-02 -1.02065754e+00
6.39328897e-01 -9.65266287e-01 -9.39286202e-02 1.11301385e-01
-2.48310283e-01 -6.10677719e-01 -1.27846867e-01 1.02633643e+00
-3.98219049e-01 -6.49805814e-02 7.28459120e-01 -3.77613567e-02
-9.55266178e-01 5.67627609e-01 3.25489700e-01 -2.37570610e-02
1.13493419e+00 -5.90536416e-01 6.79987371e-02 1.15409039e-01
-7.20451057e-01 5.87901354e-01 3.69934022e-01 2.19453275e-01
3.60494345e-01 -1.25253344e+00 -3.88150573e-01 1.03054784e-01
6.75656736e-01 7.04498172e-01 2.31800348e-01 9.01968598e-01
-2.38026276e-01 5.54933846e-01 4.75351363e-01 -1.33344412e+00
-9.85412240e-01 2.45677292e-01 4.53471132e-02 -1.33816898e-01
-4.84663963e-01 1.02563715e+00 4.58310306e-01 -7.55916357e-01
3.15468311e-01 -3.03429872e-01 -1.05956219e-01 2.68164247e-01
-1.14855185e-01 2.74820954e-01 2.92309880e-01 -6.98837936e-01
-4.63854760e-01 8.64978671e-01 1.25652269e-01 2.19480827e-01
1.18769586e+00 -2.42444992e-01 1.84458181e-01 5.62903404e-01
9.94153380e-01 8.28880966e-02 -1.69077682e+00 -5.05763113e-01
2.05836758e-01 -6.39178991e-01 1.75894365e-01 -7.49126554e-01
-8.54073882e-01 1.12929928e+00 6.44487143e-01 4.85518575e-03
1.04802382e+00 2.58264989e-01 5.38064897e-01 3.63401741e-01
8.26393783e-01 -1.03356671e+00 1.60027444e-01 1.30689636e-01
4.78351295e-01 -1.68239415e+00 4.68980856e-02 -1.02087498e+00
-6.47478819e-01 5.52260876e-01 8.17128479e-01 -3.27299326e-03
4.57337439e-01 1.03147604e-01 -1.44403884e-02 -3.00163418e-01
-4.93603140e-01 -5.18513083e-01 3.02458316e-01 3.43815118e-01
3.01266517e-02 1.16039731e-01 -1.77310823e-04 3.98903221e-01
1.14404753e-01 -2.30737343e-01 -1.90007538e-01 8.33861291e-01
-6.54604793e-01 -1.00129497e+00 -2.23212838e-01 3.39081168e-01
-1.48742884e-01 -7.53869712e-02 -1.87120259e-01 7.96747029e-01
6.16047442e-01 8.98415983e-01 4.18016985e-02 -3.40454012e-01
1.81506813e-01 -1.08698875e-01 1.89105254e-02 -9.32993948e-01
-2.53389716e-01 2.57740349e-01 1.79111704e-01 -6.60839379e-01
-7.66814530e-01 -5.88979125e-01 -1.29794300e+00 2.47516781e-01
-8.15290570e-01 2.45567843e-01 4.68763083e-01 9.18619156e-01
3.96153241e-01 2.92614937e-01 9.08498764e-01 -7.37420022e-01
-5.07448614e-01 -6.85655653e-01 -5.19002438e-01 7.65688792e-02
1.14098623e-01 -8.63011241e-01 -5.45963943e-01 1.60836756e-01] | [8.04211711883545, -3.074188232421875] |
da93b289-1668-4ecc-8f83-357ec768acaf | reactive-multi-stage-feature-fusion-for | 1908.05067 | null | https://arxiv.org/abs/1908.05067v1 | https://arxiv.org/pdf/1908.05067v1.pdf | Reactive Multi-Stage Feature Fusion for Multimodal Dialogue Modeling | Visual question answering and visual dialogue tasks have been increasingly studied in the multimodal field towards more practical real-world scenarios. A more challenging task, audio visual scene-aware dialogue (AVSD), is proposed to further advance the technologies that connect audio, vision, and language, which introduces temporal video information and dialogue interactions between a questioner and an answerer. This paper proposes an intuitive mechanism that fuses features and attention in multiple stages in order to well integrate multimodal features, and the results demonstrate its capability in the experiments. Also, we apply several state-of-the-art models in other tasks to the AVSD task, and further analyze their generalization across different tasks. | ['Yun-Nung Chen', 'Shang-Yu Su', 'Yu-Hsuan Deng', 'Hsiao-Hua Cheng', 'Tzu-Chuan Lin', 'Yi-Ting Yeh'] | 2019-08-14 | null | null | null | null | ['scene-aware-dialogue'] | ['computer-vision'] | [-2.23191485e-01 -3.84116113e-01 1.95462957e-01 -2.92284161e-01
-6.11787558e-01 -4.01483625e-01 8.36557806e-01 1.08638391e-01
-4.64019835e-01 4.21115696e-01 5.96623838e-01 -7.41994753e-02
2.99783528e-01 -2.35310346e-01 -1.47092909e-01 -4.70263332e-01
1.41834065e-01 1.66740134e-01 4.91348982e-01 -4.55116481e-01
2.54406184e-01 1.59638777e-01 -1.56126344e+00 9.77551103e-01
5.64691722e-01 1.02791870e+00 2.94177175e-01 1.08170521e+00
-5.97029746e-01 1.23630619e+00 -7.47755289e-01 -4.48106855e-01
-4.37071592e-01 -5.02005041e-01 -1.18260527e+00 2.07106277e-01
4.20504481e-01 -4.78905499e-01 -6.23287261e-01 7.84937978e-01
6.77674472e-01 3.87478590e-01 6.55383587e-01 -1.70035684e+00
-9.37307596e-01 2.02715144e-01 -7.19416201e-01 2.96461105e-01
1.15611386e+00 5.21786332e-01 7.91707397e-01 -9.82678831e-01
5.35710275e-01 1.81612325e+00 1.81418195e-01 6.11797690e-01
-7.04753935e-01 -5.65158784e-01 5.00801146e-01 9.73219037e-01
-9.66095865e-01 -4.24499094e-01 9.40723598e-01 -5.03468692e-01
7.92901397e-01 3.77392977e-01 9.13661242e-01 1.49805892e+00
-8.33301991e-02 1.39886832e+00 7.72550344e-01 -3.57439905e-01
-1.80007800e-01 1.95855707e-01 1.09097369e-01 5.32799542e-01
-7.07200229e-01 -8.48383009e-02 -8.94136310e-01 7.80442581e-02
5.80083191e-01 1.38368122e-02 -5.44446290e-01 -2.48910904e-01
-1.27601600e+00 8.46764922e-01 5.27559578e-01 4.51091558e-01
-1.37074932e-01 1.07572928e-01 7.55504966e-01 3.46043617e-01
2.91479588e-01 1.66419074e-01 6.81168661e-02 -1.54139683e-01
-4.62965131e-01 1.40907347e-01 5.04158020e-01 8.87919784e-01
5.04250348e-01 -4.88290749e-03 -8.65407169e-01 8.86371136e-01
7.42790997e-01 3.22974712e-01 3.38472128e-01 -1.17554915e+00
4.86101896e-01 6.84410930e-01 1.63783431e-02 -1.00076926e+00
-5.64713955e-01 2.32080713e-01 -6.88925266e-01 -1.71383228e-02
4.05570149e-01 -2.71699056e-02 -6.05444431e-01 1.44569826e+00
6.16470873e-01 -6.11641118e-03 1.33573890e-01 1.33349383e+00
1.78722310e+00 9.39743578e-01 3.25299472e-01 -2.16527626e-01
1.79406989e+00 -1.49828732e+00 -1.15899110e+00 -1.97130486e-01
2.79403090e-01 -8.40322793e-01 1.56335878e+00 2.37254843e-01
-1.01604295e+00 -9.22309637e-01 -7.94301331e-01 -5.22672951e-01
-3.10313702e-01 3.70868258e-02 5.33125699e-01 1.47286370e-01
-9.46778893e-01 -2.41543293e-01 -2.24246070e-01 -7.65055001e-01
1.85867533e-01 -1.42377064e-01 -4.85305071e-01 1.48573648e-02
-1.44932878e+00 8.82887781e-01 1.72884703e-01 5.23681104e-01
-1.13095474e+00 -1.59448788e-01 -9.47626114e-01 -5.54905087e-02
5.29975533e-01 -1.09108794e+00 1.41567123e+00 -9.51325357e-01
-1.66059792e+00 9.19543266e-01 -2.90182173e-01 -4.61347289e-02
3.81233007e-01 -3.37309241e-01 -4.50173408e-01 7.98300087e-01
-3.21530610e-01 8.54029119e-01 9.73123074e-01 -1.36470842e+00
-5.90127230e-01 -3.96339715e-01 4.23248976e-01 6.02395654e-01
-3.68390262e-01 1.91788390e-01 -7.97865808e-01 -2.48850361e-01
-2.34393850e-01 -4.03050512e-01 1.49421990e-01 2.93140292e-01
-1.59003928e-01 -5.18580258e-01 1.29287052e+00 -5.19482017e-01
9.99603927e-01 -2.14844370e+00 4.62413311e-01 -5.08447826e-01
3.89100939e-01 2.54055470e-01 -3.48860681e-01 7.55449593e-01
1.99590594e-01 -2.65387714e-01 1.69088557e-01 -3.89757812e-01
-1.13821462e-01 7.59320334e-02 -2.28184521e-01 3.05074304e-01
1.54051691e-01 1.22645342e+00 -8.10701430e-01 -8.71921241e-01
3.92812848e-01 4.91036117e-01 -2.91237831e-01 6.00012600e-01
-3.98099691e-01 7.64748156e-01 -4.11251515e-01 6.81559443e-01
4.25697714e-01 -3.55756342e-01 -7.49229193e-02 -5.60962796e-01
-2.83726286e-02 -3.42847854e-01 -9.93744314e-01 1.96744156e+00
-3.57312143e-01 1.20497155e+00 4.46083218e-01 -1.01181424e+00
7.50886500e-01 5.19623160e-01 1.58635795e-01 -1.06955647e+00
3.13550502e-01 -4.48154777e-01 -3.15195084e-01 -1.55095530e+00
6.08189344e-01 3.35911922e-02 -4.20657918e-02 4.55094203e-02
3.59978884e-01 -5.06775565e-02 3.10544278e-02 4.89680976e-01
5.23420095e-01 2.14992806e-01 3.48952293e-01 3.10635507e-01
1.02856302e+00 -1.89430397e-02 4.58595529e-02 4.77414131e-01
-6.64889574e-01 5.92433572e-01 5.43127179e-01 -4.48477380e-02
-5.72710812e-01 -1.03701878e+00 2.57470787e-01 1.51199007e+00
6.25903964e-01 -4.04603869e-01 -4.37826395e-01 -7.46374130e-01
5.05281463e-02 4.34776068e-01 -8.05714130e-01 -1.16145812e-01
-2.21757099e-01 -2.43887901e-02 4.84079778e-01 4.80145991e-01
7.29657710e-01 -1.22319889e+00 -6.19475484e-01 -8.12687501e-02
-7.85580456e-01 -1.32815063e+00 -4.99425083e-01 -5.47473431e-01
-3.60662609e-01 -1.11812139e+00 -1.12009525e+00 -8.40789914e-01
9.58585814e-02 5.36874235e-01 1.01583779e+00 -3.63055170e-02
-2.59348273e-01 1.17952585e+00 -6.43404603e-01 -2.77746022e-01
-1.27729282e-01 -3.21760416e-01 -4.33297783e-01 5.20176113e-01
2.09818721e-01 -8.57046396e-02 -7.84168065e-01 4.75712091e-01
-7.88796902e-01 6.49121106e-02 2.72578299e-01 8.21320534e-01
2.51563519e-01 -8.69082570e-01 5.87122500e-01 -3.26012641e-01
9.22690570e-01 -4.15505230e-01 -1.68047883e-02 6.23282731e-01
2.17326701e-01 -2.10646063e-01 2.39354283e-01 -7.15478539e-01
-1.26367569e+00 -2.90281653e-01 -2.67813325e-01 -7.06284225e-01
-3.87174517e-01 5.19492686e-01 -3.42229277e-01 3.78895439e-02
4.34596121e-01 1.84943736e-01 1.96567178e-01 -2.40719259e-01
8.02357852e-01 8.29585016e-01 8.71953189e-01 -2.67680109e-01
4.77586925e-01 5.54499388e-01 -2.90613979e-01 -1.08598936e+00
-4.65071142e-01 -7.77205706e-01 -4.41557169e-01 -9.60060656e-01
1.27492857e+00 -9.62495208e-01 -1.54275811e+00 4.66937959e-01
-1.54674470e+00 -1.45703346e-01 -2.30441280e-02 5.30292451e-01
-6.07157290e-01 7.85332680e-01 -4.76676047e-01 -1.02920330e+00
-2.45904505e-01 -1.12579346e+00 1.31484914e+00 6.14713967e-01
2.51929741e-02 -9.71009493e-01 -5.67276366e-02 7.40084529e-01
3.00838292e-01 6.37112930e-02 6.80096030e-01 -6.59809649e-01
-5.40027559e-01 1.99735150e-01 -5.70679367e-01 1.41692281e-01
-2.39005044e-01 8.60321149e-02 -1.12354958e+00 -1.27189636e-01
1.60046816e-01 -6.68339610e-01 1.01002860e+00 2.47566387e-01
9.86416936e-01 -3.57410796e-02 -2.60316491e-01 2.15601385e-01
8.69605601e-01 3.98724586e-01 5.62561035e-01 8.41420367e-02
7.22041726e-01 1.03763771e+00 8.90784204e-01 2.93491244e-01
8.30027282e-01 8.46004844e-01 8.13619375e-01 -4.46900785e-01
-2.41018549e-01 -7.71880746e-02 3.32556963e-01 8.30699265e-01
1.49668649e-01 -4.60150868e-01 -8.94997239e-01 4.43734556e-01
-2.22285652e+00 -1.03075945e+00 -4.12735105e-01 1.43520486e+00
5.04622281e-01 -4.98537600e-01 2.72443205e-01 -8.92544389e-02
6.76483095e-01 5.03380656e-01 -4.33104515e-01 -3.83522838e-01
-1.36293262e-01 -4.71454889e-01 -5.14259517e-01 3.68325800e-01
-1.20251846e+00 9.28637624e-01 6.50604486e+00 6.32635713e-01
-1.09835052e+00 1.86670303e-01 3.19452375e-01 5.58676086e-02
-2.35746577e-01 -1.45233929e-01 -3.14527482e-01 1.68943897e-01
4.80437458e-01 1.39517123e-02 1.79668128e-01 4.76367176e-01
1.65547088e-01 -3.69285256e-01 -1.21196830e+00 1.42238033e+00
5.09344876e-01 -1.27302349e+00 2.48566642e-01 -5.43144882e-01
1.06148414e-01 -5.57630897e-01 1.75095931e-01 3.74645114e-01
-2.34858692e-01 -1.02410173e+00 7.07988381e-01 9.27478433e-01
5.82268715e-01 -4.53514814e-01 6.73660278e-01 2.95422941e-01
-1.55645704e+00 -2.75511384e-01 -1.37090236e-01 -1.16328336e-01
6.44845009e-01 -2.11864367e-01 -6.15847588e-01 7.78574944e-01
1.03636479e+00 5.97365141e-01 -7.84939349e-01 1.15003598e+00
-1.24394156e-01 1.74473703e-01 2.33572364e-01 -3.13490480e-01
2.64902860e-01 1.28209263e-01 5.22975743e-01 1.26568222e+00
-1.71641201e-01 1.23902582e-01 1.87627345e-01 6.05721653e-01
3.68591487e-01 4.17303830e-01 -6.44954741e-01 5.88667579e-02
2.37653047e-01 1.30435085e+00 -2.70429432e-01 -2.01155379e-01
-7.68522978e-01 1.00901997e+00 2.14722693e-01 7.33824611e-01
-9.66618657e-01 -3.82990748e-01 4.81754124e-01 -3.24466318e-01
7.44840205e-02 -3.71304870e-01 2.85682976e-01 -1.13533306e+00
-9.26908553e-02 -9.30929661e-01 6.77694559e-01 -1.44669688e+00
-1.43123710e+00 6.01763070e-01 3.68630774e-02 -1.34418261e+00
-2.43230909e-01 -5.29225349e-01 -5.79687476e-01 6.74690604e-01
-1.57443714e+00 -1.40076876e+00 -8.99354398e-01 1.09091485e+00
8.90404463e-01 -2.17686191e-01 6.14059985e-01 3.52486998e-01
-4.57928598e-01 5.07589459e-01 -4.13864493e-01 1.41162813e-01
1.05245543e+00 -7.45229304e-01 -1.46945670e-01 4.59310740e-01
2.65781671e-01 9.04097110e-02 6.04801893e-01 -4.90335487e-02
-1.52220988e+00 -6.52905226e-01 5.68006217e-01 -2.91888297e-01
5.39738536e-01 -4.21853483e-01 -9.55816746e-01 3.97900760e-01
9.66352046e-01 -3.56955260e-01 6.15873516e-01 -2.85953954e-02
-4.18482631e-01 -2.99826153e-02 -7.87481427e-01 7.40352571e-01
7.81272471e-01 -9.71659780e-01 -8.22575688e-01 9.39946324e-02
9.45793808e-01 -4.27829415e-01 -6.98934197e-01 4.28599000e-01
6.35510385e-01 -1.09143162e+00 1.08891058e+00 -7.79188454e-01
3.47616553e-01 -4.15546745e-01 -2.97345936e-01 -9.83981788e-01
6.52951077e-02 -6.82690740e-01 -2.27955967e-01 1.47254956e+00
2.31899694e-02 -1.78188711e-01 2.61077732e-01 4.26996611e-02
-1.69666745e-02 -5.31292319e-01 -1.00013840e+00 -2.79961616e-01
-2.67846078e-01 -3.70326579e-01 2.11909801e-01 9.00775671e-01
3.48855495e-01 9.99932349e-01 -6.70101166e-01 -6.43856749e-02
9.35816206e-03 1.54905200e-01 1.11835456e+00 -1.14901662e+00
-2.08586574e-01 -5.02305984e-01 -4.29591447e-01 -1.61874783e+00
2.05507010e-01 -5.60950100e-01 -5.80740580e-03 -1.95139837e+00
2.77903676e-01 5.89169383e-01 8.38018656e-02 1.21664532e-01
-4.35522467e-01 1.56931296e-01 6.73965693e-01 4.91547622e-02
-1.23173237e+00 1.15317357e+00 1.70496821e+00 -4.84588623e-01
-1.68885902e-01 -2.12316424e-01 -5.24496794e-01 6.97762191e-01
2.77654499e-01 2.41669148e-01 -4.74673629e-01 -5.25256872e-01
-3.03692892e-02 6.43050611e-01 5.93527973e-01 -7.00412691e-01
5.16006827e-01 -5.16472422e-02 3.96650434e-01 -9.53412354e-01
8.43931675e-01 -7.88113713e-01 -3.64972264e-01 1.61541924e-01
-5.88241935e-01 1.85430527e-01 2.85655141e-01 7.55826950e-01
-7.75152206e-01 5.16309999e-02 5.81643701e-01 -1.36611173e-02
-1.15523601e+00 1.89968079e-01 -5.28583825e-01 2.63222903e-01
1.29181921e+00 -1.48347080e-01 -7.31608987e-01 -1.00237048e+00
-1.00781202e+00 9.67438638e-01 -1.22789800e-01 7.19328284e-01
1.07708979e+00 -1.41805279e+00 -7.25426257e-01 -1.94131687e-01
4.54945385e-01 -3.55578005e-01 8.88824046e-01 1.05043650e+00
-4.08528537e-01 4.96099055e-01 -2.45351106e-01 -1.08918452e+00
-1.71530259e+00 7.54047573e-01 3.65638733e-01 1.46830946e-01
-4.48388338e-01 9.68768597e-01 5.78608215e-01 -2.38253683e-01
7.29062617e-01 5.22093922e-02 -8.70570838e-01 6.73404753e-01
5.61848760e-01 3.42067391e-01 -5.16772270e-01 -7.53814578e-01
-3.20222825e-01 6.99718833e-01 1.39592677e-01 -2.26942107e-01
7.00775146e-01 -7.60852039e-01 8.12839493e-02 8.74787927e-01
1.24070096e+00 -3.98340225e-01 -1.24472225e+00 -4.46817398e-01
-3.13982636e-01 -3.88927728e-01 -3.07339251e-01 -6.57328486e-01
-9.92097735e-01 1.53065157e+00 7.29213595e-01 4.95530516e-01
1.20711267e+00 4.95776534e-01 5.47350883e-01 2.73687333e-01
-1.50457054e-01 -9.74045157e-01 9.24555361e-01 5.02917409e-01
1.42085624e+00 -1.35333979e+00 -2.36446276e-01 -3.56573790e-01
-1.37423241e+00 1.19941163e+00 9.22468483e-01 3.40575665e-01
4.80553895e-01 -1.70602113e-01 5.07197261e-01 -1.83522731e-01
-1.11173022e+00 -6.32684708e-01 3.69554698e-01 8.30321908e-01
2.84234077e-01 -4.02879030e-01 -2.08749294e-01 7.04885602e-01
3.84212106e-01 -1.62301928e-01 2.47960076e-01 5.74200451e-01
-3.38186443e-01 -7.48548687e-01 -3.91422987e-01 -8.73418450e-02
-1.81935966e-01 9.29358974e-02 -6.63582683e-01 1.00894558e+00
-2.07696050e-01 1.41391194e+00 1.56199530e-01 -6.21367216e-01
5.42643368e-01 1.68633834e-01 3.71307343e-01 -2.53362834e-01
-7.13065803e-01 2.68608779e-01 1.53859556e-01 -7.15415597e-01
-8.57810497e-01 -4.05961305e-01 -1.20493126e+00 1.53394351e-02
-2.24738583e-01 -1.73348486e-02 4.15751427e-01 1.00592542e+00
2.24383637e-01 7.51818419e-01 5.38290501e-01 -8.55258584e-01
-8.08147490e-02 -8.97534966e-01 -3.06110173e-01 3.46284807e-01
4.83644962e-01 -7.45665491e-01 -2.57807255e-01 -1.29698709e-01] | [10.801175117492676, 1.26587975025177] |
52b56c21-69b3-4870-a39d-02d1baeb2f3d | escort-ethereum-smart-contracts-vulnerability | 2103.12607 | null | https://arxiv.org/abs/2103.12607v1 | https://arxiv.org/pdf/2103.12607v1.pdf | ESCORT: Ethereum Smart COntRacTs Vulnerability Detection using Deep Neural Network and Transfer Learning | Ethereum smart contracts are automated decentralized applications on the blockchain that describe the terms of the agreement between buyers and sellers, reducing the need for trusted intermediaries and arbitration. However, the deployment of smart contracts introduces new attack vectors into the cryptocurrency systems. In particular, programming flaws in smart contracts can be and have already been exploited to gain enormous financial profits. It is thus an emerging yet crucial issue to detect vulnerabilities of different classes in contracts in an efficient manner. Existing machine learning-based vulnerability detection methods are limited and only inspect whether the smart contract is vulnerable, or train individual classifiers for each specific vulnerability, or demonstrate multi-class vulnerability detection without extensibility consideration. To overcome the scalability and generalization limitations of existing works, we propose ESCORT, the first Deep Neural Network (DNN)-based vulnerability detection framework for Ethereum smart contracts that support lightweight transfer learning on unseen security vulnerabilities, thus is extensible and generalizable. ESCORT leverages a multi-output NN architecture that consists of two parts: (i) A common feature extractor that learns the semantics of the input contract; (ii) Multiple branch structures where each branch learns a specific vulnerability type based on features obtained from the feature extractor. Experimental results show that ESCORT achieves an average F1-score of 95% on six vulnerability types and the detection time is 0.02 seconds per contract. When extended to new vulnerability types, ESCORT yields an average F1-score of 93%. To the best of our knowledge, ESCORT is the first framework that enables transfer learning on new vulnerability types with minimal modification of the DNN model architecture and re-training overhead. | ['Farinaz Koushanfar', 'Ahmad Reza Sadeghi', 'Alexandra Dmitrienko', 'Christoph Sendner', 'Hossein Fereidooni', 'Huili Chen', 'Oliver Lutz'] | 2021-03-23 | null | null | null | null | ['vulnerability-detection'] | ['miscellaneous'] | [-2.26948425e-01 1.81287527e-01 -4.02629912e-01 -9.56133828e-02
-9.31893945e-01 -1.28060222e+00 6.03819549e-01 -1.30214468e-01
-2.00973958e-01 3.89814317e-01 -2.76598427e-02 -1.18949163e+00
1.72418192e-01 -9.66363907e-01 -6.97867036e-01 -6.59586132e-01
-4.49460477e-01 5.63595593e-01 2.76610494e-01 -4.38362151e-01
2.81103939e-01 3.51574421e-01 -9.46727455e-01 4.61281210e-01
4.07406181e-01 1.10884392e+00 -6.41217768e-01 4.51006621e-01
3.33573014e-01 7.19676435e-01 -6.26165211e-01 -6.35885835e-01
7.79321253e-01 2.07541153e-01 -8.52042198e-01 -1.14741778e+00
1.75342411e-01 -8.59656274e-01 -4.77530241e-01 1.28951204e+00
2.22427189e-01 -7.56961763e-01 2.88658619e-01 -1.53055453e+00
-3.97904009e-01 1.49581683e+00 -3.04081380e-01 6.78361729e-02
-1.01355568e-01 5.93651235e-01 1.34686196e+00 -3.80870700e-01
7.29584277e-01 8.55648935e-01 7.47979939e-01 6.12888694e-01
-1.15946126e+00 -1.05420721e+00 -3.66576999e-01 1.30972281e-01
-7.16594696e-01 -2.92884540e-02 8.27322721e-01 -2.82549113e-01
1.37829816e+00 9.49860290e-02 3.84709060e-01 1.40245998e+00
6.71125293e-01 6.73277736e-01 1.24824715e+00 6.91268444e-02
4.42001790e-01 -5.62055781e-02 2.62631655e-01 1.62643477e-01
4.82401341e-01 7.79547811e-01 2.63579991e-02 -7.54362583e-01
2.05602795e-01 5.63960560e-02 8.45699236e-02 -3.26478869e-01
-1.12283349e+00 1.33159721e+00 3.07725996e-01 5.46398699e-01
-1.75907418e-01 5.46458185e-01 1.21864402e+00 1.15932035e+00
-1.08763598e-01 7.49177694e-01 -1.23091400e+00 -4.76610154e-01
-7.20533073e-01 4.09021109e-01 1.42166126e+00 6.01215482e-01
5.77686071e-01 3.30172107e-02 3.75492513e-01 -3.65779608e-01
1.74685672e-01 5.30467749e-01 1.97460130e-01 -8.73026550e-01
7.19449222e-01 8.62071633e-01 -1.83113441e-01 -4.85096902e-01
-4.08336163e-01 -4.58775669e-01 -4.08025026e-01 6.75012350e-01
5.17507315e-01 -4.17279691e-01 -6.61315322e-01 1.69323552e+00
2.91880071e-01 -9.49600860e-02 4.03547287e-01 5.03158748e-01
2.80788362e-01 3.88415515e-01 -1.87627673e-01 1.92250058e-01
1.39233100e+00 -6.02555394e-01 -2.44170472e-01 2.72986203e-01
1.09600055e+00 -5.48134089e-01 6.10988200e-01 5.56812048e-01
-6.34950161e-01 3.20936918e-01 -1.25956392e+00 3.45758349e-01
-7.95821607e-01 -7.73211896e-01 9.43103075e-01 1.10247815e+00
-7.07073629e-01 7.06957817e-01 -7.97241986e-01 3.32136214e-01
4.67027485e-01 7.57853925e-01 -2.94368356e-01 4.57988977e-01
-1.70766771e+00 9.81545091e-01 6.11889958e-01 -1.46154791e-01
-1.38429606e+00 -6.97020233e-01 -7.08758950e-01 2.62717456e-01
5.30604780e-01 -1.15659833e-01 1.40557778e+00 -1.06509507e+00
-1.38873553e+00 4.02484596e-01 9.09385741e-01 -7.45313168e-01
5.60501158e-01 6.44772425e-02 -5.18934608e-01 -4.82537076e-02
-1.82398900e-01 8.92639458e-02 6.50937259e-01 -8.26949000e-01
-6.61081076e-01 -1.37397990e-01 4.55333859e-01 -6.24917269e-01
-1.61018416e-01 3.44553858e-01 5.15902400e-01 -6.60909772e-01
-2.69605637e-01 -1.17810130e+00 4.71419841e-02 -6.23745918e-01
-1.77631184e-01 -3.69354784e-01 1.30590308e+00 -8.27527702e-01
9.22108352e-01 -1.92399585e+00 -2.75063485e-01 7.70587325e-01
1.88695386e-01 5.79697371e-01 -3.32042485e-01 6.66858315e-01
-1.24262422e-01 2.35660285e-01 -2.00271785e-01 5.03477693e-01
4.47999865e-01 2.95729507e-02 -7.72671700e-01 4.69275981e-01
-1.87542424e-01 1.33916938e+00 -8.26999068e-01 -1.43798525e-02
-8.03357959e-02 3.00323129e-01 -5.25759697e-01 -1.90676656e-02
-4.65845019e-01 3.11180115e-01 -5.68134069e-01 1.10118473e+00
7.12333202e-01 6.06145822e-02 5.77372074e-01 -1.37134390e-02
7.53005520e-02 6.02208197e-01 -9.12591398e-01 1.49056113e+00
-4.39393699e-01 3.15856248e-01 2.59687573e-01 -9.88067091e-01
7.32578099e-01 9.32972848e-01 5.11995792e-01 -7.17195094e-01
5.23703218e-01 8.09081078e-01 2.97380954e-01 -3.07402253e-01
2.78286748e-02 -2.30199378e-03 -7.20976889e-01 1.23565972e+00
7.27930367e-02 2.10554823e-01 -1.12392657e-01 2.46681143e-02
1.50219262e+00 2.08694324e-01 2.08850756e-01 -2.68255085e-01
3.27899218e-01 -1.61995426e-01 8.38858545e-01 7.45768845e-01
-4.44126338e-01 -1.97707638e-01 9.93885398e-01 -1.25593936e+00
-1.08936846e+00 -9.52055275e-01 -1.48861399e-02 9.43346441e-01
-3.38148385e-01 -2.85868764e-01 -7.87244856e-01 -1.48004031e+00
3.77768546e-01 5.36480010e-01 -5.69997132e-01 -1.15484215e-01
-7.87007809e-01 -5.02410948e-01 1.05036950e+00 6.56318247e-01
4.84354645e-01 -1.43146825e+00 -9.02890921e-01 2.29592130e-01
5.04668839e-02 -8.47290456e-01 -3.20899725e-01 3.74134362e-01
-8.15542996e-01 -1.52462471e+00 -9.06178951e-02 -5.70776045e-01
2.53845632e-01 -4.12369788e-01 8.78897846e-01 1.61390781e-01
-1.04961686e-01 -1.78661436e-01 -1.98595151e-01 -7.55159631e-02
-8.37220132e-01 4.04953182e-01 5.30503802e-02 -2.89246321e-01
6.90530062e-01 -9.22271729e-01 -5.67839682e-01 2.87774116e-01
-8.28187823e-01 -5.85166037e-01 7.76608229e-01 8.26103628e-01
-1.67737678e-01 -4.17611711e-02 8.49579930e-01 -1.00121439e+00
5.27382135e-01 -5.73610485e-01 -1.33960414e+00 3.41978282e-01
-1.19129634e+00 2.55512267e-01 7.40731239e-01 -3.18589330e-01
-7.22715497e-01 -9.84380394e-02 7.34295622e-02 -3.01445097e-01
1.55884698e-01 4.70912725e-01 -1.97240800e-01 -4.08368468e-01
5.71859062e-01 -4.84318100e-02 -6.68839291e-02 -2.25466281e-01
5.96189678e-01 7.36628950e-01 3.64269048e-01 -7.70157814e-01
1.26572633e+00 2.41594464e-01 5.73879369e-02 2.37927750e-01
4.26432537e-03 -7.11157173e-02 -2.90874064e-01 3.14067841e-01
3.88555646e-01 -5.65220535e-01 -1.52897012e+00 6.75364554e-01
-1.12818241e+00 -3.58109772e-01 3.28458138e-02 2.92127609e-01
-4.53945726e-01 4.74749058e-01 -1.11050928e+00 -5.16265273e-01
-1.02175653e+00 -1.46302700e+00 5.81189513e-01 -8.58997628e-02
-4.24857795e-01 -7.72637904e-01 3.85812581e-01 3.20339590e-01
7.33283043e-01 4.50170964e-01 1.10372353e+00 -1.28497779e+00
-7.89731026e-01 -4.96697843e-01 -6.58534393e-02 3.81664872e-01
1.29271289e-02 -2.71120071e-01 -1.04886663e+00 -7.69069731e-01
1.64169714e-01 -5.35428643e-01 6.78909898e-01 -1.49723262e-01
6.05839670e-01 -8.02804708e-01 -2.14693621e-01 7.61719465e-01
1.35161233e+00 4.41430688e-01 5.58142960e-01 9.20466900e-01
2.48771176e-01 4.69665527e-01 4.70712781e-02 9.37172621e-02
1.62386984e-01 3.93288404e-01 8.04539382e-01 1.46280706e-01
5.75384676e-01 -1.32846937e-01 6.41093552e-01 1.75180867e-01
-1.59440622e-01 4.74727333e-01 -1.09356689e+00 3.69892925e-01
-1.72636366e+00 -1.02292120e+00 3.17173898e-01 1.96650529e+00
9.27471220e-01 3.39824677e-01 1.33025751e-01 1.69840336e-01
2.68447906e-01 -1.13336928e-01 -5.49682200e-01 -9.34363663e-01
8.95911902e-02 5.00596941e-01 8.60605240e-01 4.05914843e-01
-1.32581472e+00 7.63700962e-01 5.64806652e+00 5.44348657e-01
-1.31934309e+00 2.86689401e-01 1.89091399e-01 -8.04125518e-02
-5.25762439e-01 5.45051157e-01 -4.34348851e-01 6.64633512e-01
1.05564964e+00 -2.29084846e-02 7.72471130e-01 1.04443765e+00
-3.71959716e-01 6.30732834e-01 -1.18817425e+00 4.49813664e-01
-3.91692489e-01 -1.42124057e+00 -3.63827944e-01 3.91136140e-01
7.16735423e-01 5.23000062e-01 1.90601885e-01 6.45768940e-01
8.52560103e-01 -8.95356655e-01 6.83556736e-01 -7.95641616e-02
6.42594099e-01 -9.60709393e-01 1.29015112e+00 1.20701455e-02
-8.41232479e-01 -7.07225502e-01 2.45795667e-01 2.54643172e-01
-1.47464007e-01 1.49255008e-01 -9.53452647e-01 6.06361091e-01
7.78125942e-01 2.17520356e-01 -1.62285075e-01 4.07207102e-01
-6.23314202e-01 7.77128696e-01 -3.41007560e-01 1.30652219e-01
5.63758850e-01 1.01839878e-01 5.62658966e-01 1.07492220e+00
2.46212602e-01 -3.10882449e-01 -7.15586182e-04 9.85224068e-01
-1.34348243e-01 -4.01536286e-01 -5.39211690e-01 -1.73860356e-01
4.85887170e-01 1.36192501e+00 -4.60093886e-01 -2.01208726e-01
-6.60253882e-01 3.96407366e-01 -1.91431232e-02 4.80922937e-01
-6.66257441e-01 -6.90020084e-01 5.96321464e-01 -2.24280134e-01
7.00614214e-01 1.12564333e-01 -4.09882218e-01 -1.26745570e+00
1.14786625e-01 -1.52959597e+00 7.46754825e-01 8.69448334e-02
-1.12082696e+00 5.22042811e-01 -2.67971516e-01 -1.16527820e+00
-5.29063642e-01 -6.41972125e-01 -8.51022661e-01 6.86567545e-01
-1.71141720e+00 -1.73251295e+00 5.51576376e-01 5.35578191e-01
-1.49523079e-01 -7.86264181e-01 1.12970006e+00 -3.34424712e-02
-1.72401309e-01 8.63013387e-01 2.76626557e-01 6.13185644e-01
3.25994223e-01 -1.16522837e+00 8.99093866e-01 7.46211410e-01
-1.03036560e-01 9.38860059e-01 5.18260360e-01 -7.75973737e-01
-1.68895161e+00 -7.62685299e-01 1.03100574e+00 -4.59948361e-01
1.29373467e+00 -6.08330131e-01 -7.39607155e-01 1.13566434e+00
3.77505302e-01 -4.82085869e-02 6.03780687e-01 2.27162048e-01
-1.33786404e+00 -3.44843894e-01 -1.36787176e+00 3.31527352e-01
4.29186910e-01 -8.78082454e-01 -8.12181532e-01 -7.76013061e-02
6.06850863e-01 -5.13363704e-02 -1.12274539e+00 4.55608159e-01
9.72420514e-01 -9.01881635e-01 6.45426750e-01 -6.44579351e-01
1.41229242e-01 -1.61943063e-01 -1.58239469e-01 -8.79681408e-01
1.51831344e-01 -1.36951530e+00 -6.96137965e-01 9.78877187e-01
5.31426489e-01 -1.26178694e+00 8.09020579e-01 5.72589815e-01
2.10414216e-01 -7.10770190e-01 -1.45395577e+00 -9.46139693e-01
6.39515281e-01 -1.91422492e-01 1.26486218e+00 1.28526330e+00
6.73847198e-01 -2.03994200e-01 -6.79234490e-02 7.23449290e-02
9.06605303e-01 6.58757627e-01 5.73315680e-01 -1.21977079e+00
-5.20663142e-01 -6.83076441e-01 -3.41122806e-01 -1.38087720e-01
3.17304790e-01 -1.07840765e+00 -5.38980663e-01 -6.61908805e-01
1.36887565e-01 -4.10525680e-01 -7.96684742e-01 1.13096893e+00
4.15449530e-01 -1.47134945e-01 1.18324250e-01 2.84679472e-01
-1.50810689e-01 3.90778109e-03 5.96073925e-01 -6.79299772e-01
-3.88718694e-02 2.32269958e-01 -6.62671030e-01 4.47961450e-01
9.44235981e-01 -7.20728755e-01 -8.43608528e-02 -1.07221395e-01
6.50386870e-01 3.32498066e-02 3.10490131e-01 -3.63410026e-01
1.70624197e-01 -1.60805285e-01 -2.87381172e-01 -4.22636032e-01
-3.47554862e-01 -9.84608173e-01 1.53623030e-01 1.05883563e+00
-2.66587492e-02 1.08859770e-01 -4.67989547e-03 2.12025061e-01
1.77439917e-02 -1.38071358e-01 3.87671053e-01 -1.80217028e-01
-5.00141859e-01 2.43495658e-01 -2.27167815e-01 -3.18068027e-01
1.14122415e+00 1.18398391e-01 -6.46378696e-01 -1.48502126e-01
-3.40127677e-01 4.38855618e-01 5.23588836e-01 6.47164404e-01
2.73770332e-01 -1.10573590e+00 -7.23556161e-01 4.10830230e-01
4.88292910e-02 -4.35474783e-01 -1.76709786e-01 6.79791212e-01
-5.29434860e-01 5.82349837e-01 -3.65433425e-01 -9.36196744e-02
-1.10618615e+00 7.59165466e-01 5.14273643e-01 -7.03611314e-01
-7.19198346e-01 2.99314797e-01 -1.19191468e-01 -7.54390061e-01
1.75174907e-01 -4.49856132e-01 1.14575267e-01 -7.35448301e-02
5.23413360e-01 3.01884472e-01 1.69821665e-01 -2.57572204e-01
-5.03700376e-01 4.31052953e-01 -4.27120090e-01 -1.61119431e-01
1.55997634e+00 7.96271265e-01 -4.79885519e-01 -2.84756541e-01
1.23003411e+00 -1.88071698e-01 -1.04904974e+00 -2.08953053e-01
5.22835910e-01 -1.53504625e-01 -1.06907181e-01 -1.26201725e+00
-1.37820518e+00 7.07826257e-01 5.31951666e-01 2.56153315e-01
8.54617298e-01 -4.11024913e-02 1.26653504e+00 7.72094905e-01
7.93901384e-01 -1.05738544e+00 -3.71667117e-01 6.58572793e-01
5.39269507e-01 -9.11403179e-01 -4.28992659e-01 3.19048733e-01
-2.46463954e-01 1.29017305e+00 1.69064432e-01 -1.99685350e-01
6.40402138e-01 8.06186795e-01 4.23393369e-01 -4.17278796e-01
-6.74116969e-01 5.78366399e-01 -3.89135987e-01 5.47936976e-01
-1.52332067e-01 3.46503109e-01 -3.68118674e-01 8.49482775e-01
-1.12405591e-01 1.24362454e-01 3.61686975e-01 1.01032722e+00
-1.62172213e-01 -1.64618278e+00 -2.66342133e-01 8.96615461e-02
-9.94714737e-01 -1.58192083e-01 -2.22667009e-01 7.15989292e-01
-1.25438208e-02 7.25519836e-01 -4.78598207e-01 -6.86988175e-01
1.23134322e-01 1.84055135e-01 1.26342803e-01 -8.54386911e-02
-1.34074199e+00 -2.99113035e-01 1.18437558e-01 -7.34663785e-01
1.23065561e-01 -6.67026699e-01 -1.25968456e+00 -5.84589303e-01
-3.78669739e-01 3.11342359e-01 6.77667916e-01 7.83724129e-01
3.01626146e-01 -2.37507597e-01 1.01841009e+00 -2.62707770e-01
-1.37228429e+00 -9.35126245e-01 -5.53715110e-01 4.13689911e-01
5.18142402e-01 -3.02513093e-01 -6.56925499e-01 -3.58184248e-01] | [6.816624641418457, 7.321304798126221] |
fcb7f4f6-3278-43b9-ab6a-c06ec9d3bc0b | is-neural-topic-modelling-better-than | null | null | https://openreview.net/forum?id=UBk6b94uH7 | https://openreview.net/pdf?id=UBk6b94uH7 | Is Neural Topic Modelling Better than Clustering? An Empirical Study on Clustering with Contextual Embeddings for Topics | Recent work incorporates pre-trained word embeddings such as BERT embeddings into Neural Topic Models (NTMs), generating highly coherent topics. However, with high-quality contextualized document representations, do we really need sophisticated neural models to obtain coherent and interpretable topics? In this paper, we conduct thorough experiments showing that directly clustering high-quality sentence embeddings with an appropriate word selecting method can generate more coherent and diverse topics than NTMs, achieving also higher efficiency and simplicity. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['topic-models'] | ['natural-language-processing'] | [-3.35898638e-01 3.90660822e-01 -4.33734298e-01 -5.67848802e-01
-8.56884301e-01 -1.25996217e-01 8.71446013e-01 1.86862871e-01
-4.04851079e-01 7.71306336e-01 9.31549251e-01 -3.01651359e-01
9.02688503e-02 -9.55953360e-01 -4.46060330e-01 -4.62824643e-01
-9.26251262e-02 5.66750228e-01 5.28435819e-02 -2.34903440e-01
2.36461580e-01 -8.38557929e-02 -1.37441337e+00 -1.73750129e-02
1.07241762e+00 4.98930693e-01 3.63723636e-01 7.29196727e-01
-6.94163501e-01 4.64568473e-02 -1.08294439e+00 -3.38593066e-01
-2.17169553e-01 -3.83692861e-01 -6.45285428e-01 3.24153937e-02
2.98875183e-01 -1.23767190e-01 -3.98242436e-02 6.87241137e-01
2.28598922e-01 5.87793648e-01 8.98945391e-01 -7.99753189e-01
-1.31944275e+00 1.04612398e+00 -2.61799693e-01 8.39749202e-02
-1.77057087e-02 -1.06312141e-01 1.53732455e+00 -9.95292902e-01
6.19164169e-01 1.44500673e+00 3.52101773e-01 7.01798141e-01
-1.34827971e+00 -4.45833027e-01 5.38638234e-01 4.72915955e-02
-1.06348181e+00 -1.64642707e-02 7.85376966e-01 6.86877454e-03
1.14422441e+00 6.23167828e-02 7.72723258e-01 1.52173769e+00
4.29595858e-01 8.48509192e-01 4.14770991e-01 -4.11304355e-01
5.45508146e-01 4.40525085e-01 5.47263563e-01 5.21620400e-02
7.98624277e-01 -4.04020220e-01 -4.53864574e-01 -4.15654004e-01
6.36427522e-01 1.87860966e-01 -1.75664961e-01 -1.75373882e-01
-1.06843936e+00 1.54424942e+00 5.27126968e-01 5.83699763e-01
-6.67741358e-01 3.46012324e-01 3.81775707e-01 -5.28833792e-02
1.05202472e+00 1.18527305e+00 -2.95679688e-01 -1.98708221e-01
-1.19123209e+00 3.16644460e-01 6.06830001e-01 9.28480923e-01
6.88763320e-01 4.45209891e-01 -5.46733439e-01 1.00726771e+00
4.19984519e-01 4.37504143e-01 9.92552280e-01 -5.84465802e-01
1.35375127e-01 2.71200448e-01 -4.49136570e-02 -8.96252215e-01
-1.72064453e-01 -6.16001308e-01 -5.69912136e-01 -3.72740239e-01
-3.91546965e-01 -4.64392811e-01 -1.28251886e+00 1.69542563e+00
-1.62838981e-01 1.65864170e-01 3.67684722e-01 6.61912560e-01
7.30943084e-01 1.35684562e+00 6.16441786e-01 -5.98569736e-02
1.49650526e+00 -1.08085752e+00 -1.12940741e+00 -6.57826066e-01
3.80048037e-01 -6.15234911e-01 1.28486776e+00 1.47033975e-01
-9.38442647e-01 -4.07915384e-01 -1.08832240e+00 -2.80195773e-01
-4.85824347e-01 8.33350867e-02 8.95489812e-01 5.91024756e-01
-1.17856252e+00 2.33411610e-01 -9.00221169e-01 -5.15621841e-01
5.28390825e-01 8.98586437e-02 -1.31831780e-01 -1.52802676e-01
-1.52764535e+00 1.01293719e+00 4.49780792e-01 -3.97887379e-01
-8.89535666e-01 -6.40074134e-01 -1.12398374e+00 4.01967108e-01
3.83481309e-02 -8.08350980e-01 1.20206630e+00 -2.48328686e-01
-1.39632726e+00 1.10974632e-01 -6.79313898e-01 -7.23667026e-01
-5.04643559e-01 -6.53759539e-01 -2.41519004e-01 1.32460818e-01
1.17713124e-01 1.20100331e+00 7.97674775e-01 -1.15343630e+00
-3.72576833e-01 -1.00424297e-01 -7.84571320e-02 3.48815143e-01
-1.08882034e+00 -1.22271627e-01 -2.20596313e-01 -7.69001186e-01
-7.55152181e-02 -5.96161842e-01 -7.24859476e-01 -2.17967138e-01
-4.31844443e-01 -9.87603784e-01 8.83523226e-01 -4.64418858e-01
1.21700811e+00 -1.91632187e+00 1.64363742e-01 -2.49460801e-01
3.14667225e-01 2.84806788e-01 -3.74259293e-01 3.93174350e-01
2.25206360e-01 4.23074663e-01 1.00280764e-02 -8.00791860e-01
1.63751572e-01 3.70604992e-01 -8.59675825e-01 -1.11053728e-01
6.37918234e-01 1.08020544e+00 -8.58561814e-01 -4.63404357e-01
1.62942320e-01 6.46452963e-01 -6.86835527e-01 4.65676069e-01
-6.07640862e-01 -5.92130542e-01 -5.65959036e-01 1.05175197e-01
2.76044756e-01 -3.10981631e-01 8.47235695e-02 3.40504616e-01
3.56954515e-01 6.87645733e-01 -5.24796665e-01 1.72449160e+00
-8.42625439e-01 1.19908774e+00 -4.70745414e-01 -9.60260749e-01
1.31692028e+00 5.31831443e-01 -1.73119374e-03 -3.15233290e-01
2.36639045e-02 -1.35046229e-01 -3.34933400e-01 -9.52860191e-02
1.39998972e+00 -4.32606339e-01 -2.10403398e-01 7.37792492e-01
4.28177923e-01 -3.63997459e-01 1.08034454e-01 5.51363707e-01
7.78150797e-01 -3.34399372e-01 6.50172085e-02 -3.98177773e-01
-3.62964660e-01 1.16380811e-01 1.80563107e-01 6.33571804e-01
1.20496005e-01 9.02458012e-01 4.80079323e-01 -1.65228575e-01
-1.40189719e+00 -1.31334651e+00 -1.43374056e-01 1.15935409e+00
-2.12883860e-01 -6.65677369e-01 -4.61844355e-01 -4.88562256e-01
-1.87371612e-01 1.34913874e+00 -6.94417536e-01 -3.57646465e-01
-3.38063538e-01 -7.98360705e-01 2.54874170e-01 7.48020053e-01
-1.94195315e-01 -1.15524793e+00 -3.15635115e-01 4.88243759e-01
-1.17236659e-01 -7.64716566e-01 -4.14045244e-01 4.80169624e-01
-1.29331458e+00 -2.24450007e-01 -1.15061736e+00 -7.66244352e-01
6.00282967e-01 4.72224385e-01 1.28357172e+00 -3.25881064e-01
2.10530702e-02 -1.57184601e-01 -4.97842073e-01 -4.91235763e-01
-9.32416618e-02 4.95900333e-01 4.35542362e-03 -6.16098464e-01
6.75544441e-01 -4.21719283e-01 -4.14954185e-01 -3.30041558e-01
-1.23097336e+00 -2.43212730e-01 5.87134421e-01 9.09951627e-01
5.10599539e-02 -7.41288736e-02 9.77845013e-01 -7.65328169e-01
1.43061745e+00 -6.74734652e-01 -2.17228644e-02 2.05402479e-01
-9.34555709e-01 3.41717660e-01 6.55744374e-01 -6.44893587e-01
-1.07527292e+00 -7.58475304e-01 -2.70251870e-01 -2.84159929e-01
-2.23087668e-01 6.20567977e-01 3.26706320e-01 1.11575401e+00
9.00801480e-01 2.66676903e-01 -1.10039577e-01 -3.94391984e-01
1.09267604e+00 7.21264541e-01 7.09782615e-02 -4.17506725e-01
6.31577194e-01 3.14484626e-01 -8.55882764e-01 -9.19973433e-01
-9.26775336e-01 -4.95104492e-01 -9.71612036e-02 3.34149152e-01
9.18649137e-01 -1.00720501e+00 4.36682612e-01 -2.60540903e-01
-1.49421513e+00 6.33311197e-02 -5.84935546e-01 7.13355482e-01
-2.72727579e-01 1.09599434e-01 -6.22455776e-01 -6.54325962e-01
-7.40980625e-01 -9.20719028e-01 1.15587580e+00 4.61771429e-01
-8.63548458e-01 -1.53328276e+00 3.46908301e-01 -7.02008978e-02
9.21185613e-01 -3.54357392e-01 1.08955932e+00 -1.05737901e+00
-3.72714967e-01 -2.82140702e-01 -3.81305330e-02 4.27136600e-01
3.16709638e-01 9.04892758e-02 -9.86885965e-01 -4.34425520e-03
-3.63953374e-02 -1.94601521e-01 1.03734529e+00 7.60705352e-01
8.25332940e-01 -4.07557964e-01 -4.26257402e-01 1.42034009e-01
1.32697129e+00 3.35139446e-02 6.63587689e-01 1.53482869e-01
3.57926607e-01 4.88217980e-01 3.47944915e-01 1.76972985e-01
4.07797039e-01 3.53579223e-01 -6.68409765e-02 -1.02283776e-01
-5.35464250e-02 -4.18105006e-01 4.83665198e-01 1.13205957e+00
4.88624692e-01 -6.11955345e-01 -6.25624180e-01 1.20820940e+00
-1.55872476e+00 -7.18409300e-01 1.47256464e-01 1.56327081e+00
1.21799672e+00 2.01853558e-01 -2.58083373e-01 -1.68715045e-01
7.59411633e-01 6.95827067e-01 -2.70349175e-01 -8.66851151e-01
4.62658666e-02 6.82734787e-01 -1.44127399e-01 4.15073752e-01
-8.29225481e-01 1.16735184e+00 7.58529568e+00 1.06582308e+00
-1.04617882e+00 2.38431275e-01 6.24262571e-01 -3.36114377e-01
-1.21108747e+00 -9.28310975e-02 -9.13765669e-01 2.78803051e-01
1.40658617e+00 -5.71562409e-01 -3.56444746e-01 1.39927220e+00
1.92968294e-01 9.91744474e-02 -7.64921248e-01 4.31878746e-01
3.75625283e-01 -1.68507564e+00 5.62447667e-01 -6.74003214e-02
1.07472634e+00 -1.46252334e-01 1.60243660e-01 5.88661134e-01
7.34116137e-01 -1.20591700e+00 2.51173764e-01 3.02575529e-01
2.80540437e-01 -7.31508017e-01 8.40170443e-01 7.94988032e-03
-6.71560407e-01 2.48294353e-01 -1.08386087e+00 1.19331621e-01
5.69385469e-01 1.16561437e+00 -1.05179501e+00 3.72322291e-01
2.86441088e-01 7.42492914e-01 -4.75474149e-01 9.33877468e-01
-4.97557819e-01 9.95545089e-01 -1.96081385e-01 -7.20961094e-01
6.38987184e-01 1.41435424e-02 3.50832015e-01 1.26356959e+00
4.14773405e-01 -2.78453231e-01 -2.12744877e-01 1.12878335e+00
-7.80707225e-02 1.75733224e-01 -6.10068619e-01 -4.93838072e-01
5.46088219e-01 1.25255871e+00 -5.56387007e-01 -7.56820440e-01
-1.16109267e-01 6.86780930e-01 3.17283183e-01 4.88809645e-01
-5.56682885e-01 -6.16971731e-01 1.00826359e+00 -1.82518065e-01
3.71175110e-01 -4.83197421e-01 -7.11623371e-01 -1.17552733e+00
-2.00435996e-01 -3.22751850e-01 3.86802666e-02 -8.84388506e-01
-1.51087010e+00 1.06746972e+00 8.74084160e-02 -6.41643763e-01
-5.47997892e-01 -2.32479498e-01 -1.17610550e+00 9.02246892e-01
-1.43162870e+00 -9.69164073e-01 1.74354658e-01 -1.24411665e-01
1.06948674e+00 -2.07441345e-01 1.00473428e+00 -2.89469838e-01
-4.19126242e-01 2.45268285e-01 1.30223215e-01 -2.72434384e-01
5.85367322e-01 -1.50103819e+00 8.42434704e-01 6.26843929e-01
5.14576614e-01 1.10977876e+00 9.63137686e-01 -5.75847089e-01
-7.54478991e-01 -1.22104955e+00 1.27746618e+00 -4.70129609e-01
5.54228783e-01 -6.36270642e-01 -1.11826289e+00 5.11236727e-01
9.18417215e-01 -7.77084172e-01 9.17774558e-01 8.39658558e-01
-3.43881220e-01 3.01085204e-01 -5.71503222e-01 8.14193428e-01
3.54307890e-01 -4.62749690e-01 -1.39451051e+00 4.16608721e-01
1.64708829e+00 2.62557924e-01 -6.36726916e-01 4.37923446e-02
1.10782355e-01 -2.82590091e-01 9.27228749e-01 -8.77422631e-01
7.61810064e-01 1.34813517e-01 3.71038765e-02 -1.87462378e+00
-2.29117557e-01 -2.13058025e-01 -1.36475712e-01 1.16764379e+00
7.89744496e-01 -4.62359846e-01 8.73734295e-01 5.02281249e-01
-2.24393308e-01 -9.71185386e-01 -7.17809737e-01 -7.44452238e-01
6.26974106e-01 -5.99474669e-01 7.07938612e-01 6.60627246e-01
2.78147638e-01 7.23476470e-01 -1.93290129e-01 -1.31780162e-01
2.43956923e-01 8.90994519e-02 3.67946476e-01 -1.12041712e+00
2.60666251e-01 -4.88722801e-01 -2.05631942e-01 -1.24463713e+00
3.31394285e-01 -5.95510125e-01 4.95663375e-01 -2.23474813e+00
2.14138716e-01 -3.31435621e-01 -1.06271103e-01 2.47053325e-01
-6.58968806e-01 1.11147659e-02 -4.25701812e-02 -9.02881101e-02
-7.51905441e-01 1.30696177e+00 8.82812321e-01 -8.08380023e-02
4.68040258e-03 -2.21295908e-01 -1.10393441e+00 2.39902169e-01
8.44853520e-01 -5.58231652e-01 -5.69667161e-01 -6.02960348e-01
8.25194418e-02 -4.30019289e-01 -4.91011553e-02 -7.16944158e-01
4.12629582e-02 -1.89813212e-01 2.33409747e-01 -5.54959118e-01
6.68572366e-01 -3.22678566e-01 -5.43279886e-01 1.13369226e-01
-8.52801502e-01 -2.89048813e-02 1.89177036e-01 6.48613214e-01
-4.84684885e-01 -5.18995821e-01 3.90762120e-01 -9.30104926e-02
-3.81198347e-01 1.88558042e-01 -9.34189498e-01 7.98785985e-02
7.76589632e-01 8.42104778e-02 -3.52089614e-01 -8.22078049e-01
-5.35505295e-01 1.20478541e-01 3.04197282e-01 8.85804892e-01
9.02987897e-01 -1.46721828e+00 -6.37655020e-01 8.34185537e-03
7.90771097e-02 2.45545834e-01 1.59697443e-01 -1.70411557e-01
5.12953214e-02 1.02302492e+00 1.61683768e-01 -4.13577348e-01
-5.87122142e-01 3.32934678e-01 -3.49188715e-01 -2.32747048e-01
-5.15671670e-01 9.21585143e-01 1.64561838e-01 -3.09816420e-01
9.27343369e-02 -4.81769890e-01 -3.98174018e-01 1.83028981e-01
5.69664657e-01 -1.31005183e-01 -2.24581823e-01 1.11806318e-02
9.42094158e-03 1.58763155e-01 -2.62786955e-01 -3.98005307e-01
1.44720185e+00 -4.29117866e-02 1.67344123e-01 3.65198225e-01
1.29642272e+00 -3.62005621e-01 -9.03609693e-01 -2.24342450e-01
3.25284481e-01 -1.99437931e-01 2.62536019e-01 -2.27193087e-01
-6.89155161e-01 1.28181398e+00 1.61410749e-01 5.72875082e-01
5.95510244e-01 3.90259713e-01 1.14456093e+00 4.63460505e-01
2.20612232e-02 -1.04487276e+00 5.07582128e-01 5.96454144e-01
8.08820307e-01 -9.04512763e-01 -4.27548252e-02 -1.62329003e-02
-9.24768329e-01 9.87984240e-01 6.44042432e-01 -5.12514889e-01
7.06649482e-01 -1.21113062e-01 4.56940010e-02 -2.80094266e-01
-1.46173835e+00 -8.47218633e-02 3.82353365e-01 6.90862298e-01
5.54541171e-01 1.19560681e-01 -4.37378138e-01 7.05127299e-01
-5.83884656e-01 -4.03318465e-01 7.88464427e-01 5.61137617e-01
-1.05477750e+00 -1.10384858e+00 -9.57546756e-02 6.92467451e-01
-2.15667188e-01 -4.46141452e-01 -7.76237473e-02 4.07372802e-01
-5.38264215e-01 9.92568791e-01 5.55519402e-01 -1.62084639e-01
-2.15905398e-01 4.00937617e-01 -3.19401324e-01 -1.38320243e+00
-2.01041847e-01 7.69592589e-04 2.16774955e-01 -9.38572809e-02
-1.45114884e-01 -3.54325145e-01 -1.10191059e+00 -1.46351904e-02
-7.17101872e-01 7.15208352e-01 9.49566662e-01 1.00977731e+00
6.25833511e-01 6.55275822e-01 4.98204768e-01 -4.47255552e-01
-6.16195381e-01 -1.58008707e+00 -4.08656001e-01 1.25418052e-01
5.15260883e-02 -3.83469820e-01 -5.58901310e-01 -1.29429072e-01] | [10.517633438110352, 7.101612567901611] |
f2abf253-0cdc-4f29-98b6-9e09c9064ea4 | a-methodology-for-search-space-reduction-in | 1809.07045 | null | http://arxiv.org/abs/1809.07045v2 | http://arxiv.org/pdf/1809.07045v2.pdf | A Methodology for Search Space Reduction in QoS Aware Semantic Web Service Composition | The semantic information regulates the expressiveness of a web service.
State-of-the-art approaches in web services research have used the semantics of
a web service for different purposes, mainly for service discovery,
composition, execution etc. In this paper, our main focus is on semantic driven
Quality of Service (QoS) aware service composition. Most of the contemporary
approaches on service composition have used the semantic information to combine
the services appropriately to generate the composition solution. However, in
this paper, our intention is to use the semantic information to expedite the
service composition algorithm. Here, we present a service composition framework
that uses semantic information of a web service to generate different clusters,
where the services are semantically related within a cluster. Our final aim is
to construct a composition solution using these clusters that can efficiently
scale to large service spaces, while ensuring solution quality. Experimental
results show the efficiency of our proposed method. | ['Soumi Chattopadhyay', 'Ansuman Banerjee'] | 2018-09-19 | null | null | null | null | ['service-composition'] | ['miscellaneous'] | [ 5.91741269e-03 -3.14710736e-01 1.61541283e-01 -5.98500371e-01
-3.36185336e-01 -5.44871509e-01 5.14943540e-01 -2.02144638e-01
3.52767818e-02 2.18137756e-01 4.21559900e-01 -1.31770417e-01
-4.91203606e-01 -1.10888755e+00 7.40475729e-02 -8.08860600e-01
1.66176885e-01 5.11462629e-01 7.18037367e-01 -4.64355648e-01
4.16275531e-01 5.53512394e-01 -2.25132895e+00 5.51429689e-01
9.37976062e-01 1.03093767e+00 2.77331293e-01 2.28981182e-01
-9.93741035e-01 1.83563009e-01 -3.34754527e-01 -2.35057727e-01
2.42307857e-01 -8.60541701e-01 -8.37610602e-01 5.36705017e-01
-7.56477594e-01 3.66070092e-01 5.21322966e-01 1.18553853e+00
2.92177677e-01 2.16856614e-01 2.29930431e-01 -1.76592815e+00
-2.60865569e-01 8.68881822e-01 9.67729762e-02 -3.15753251e-01
5.59885204e-01 -3.15517813e-01 1.10421431e+00 -4.60354239e-01
7.74191380e-01 1.11192572e+00 1.41799197e-01 8.58979940e-01
-8.23922992e-01 -4.13731515e-01 6.61018848e-01 7.57632434e-01
-1.48680103e+00 -3.68478119e-01 9.07816350e-01 4.77853976e-02
6.38982177e-01 7.37769067e-01 9.06026661e-01 2.91523159e-01
-3.31503898e-01 5.02035260e-01 1.14477468e+00 -4.93726045e-01
7.48480856e-01 1.23142958e-01 2.08077636e-02 8.60510990e-02
9.37178656e-02 -5.40006220e-01 -6.19675815e-02 -4.10500675e-01
6.93776459e-02 5.50995767e-01 -3.23525369e-01 -6.58322155e-01
-6.16232455e-01 7.07718909e-01 1.60878032e-01 9.61163223e-01
-3.89641643e-01 -2.67734766e-01 2.36518607e-01 3.98435742e-01
6.82389587e-02 1.30196482e-01 -5.01755655e-01 -3.39330822e-01
-4.07928705e-01 3.83880377e-01 1.43069983e+00 1.08223963e+00
6.25760257e-01 -1.15319639e-01 4.59430665e-01 8.54255736e-01
7.20781744e-01 1.23215772e-01 7.03645170e-01 -8.72365594e-01
-3.77866805e-01 1.21360219e+00 1.34145051e-01 -7.75305271e-01
-2.18872830e-01 -5.62003534e-03 -6.91055804e-02 2.89839000e-01
-3.88526125e-03 3.24280471e-01 -6.57026768e-01 1.38246822e+00
6.66242659e-01 2.13124499e-01 5.32477260e-01 1.23491192e+00
6.09057724e-01 4.01069134e-01 1.84587687e-01 -3.03945869e-01
1.32842314e+00 -8.95231903e-01 -7.65566826e-01 5.35017490e-01
4.45644587e-01 -7.68177927e-01 1.05583882e+00 2.34558105e-01
-7.66414404e-01 2.57867694e-01 -9.04169500e-01 6.29225016e-01
-5.95947683e-01 -8.23847175e-01 6.93650007e-01 5.87912858e-01
-1.04891169e+00 4.53687221e-01 -5.01795828e-01 -8.77058446e-01
-1.68450609e-01 3.58929992e-01 -2.29287878e-01 -1.92735985e-01
-9.46374834e-01 5.80741465e-01 4.84771192e-01 -3.67465228e-01
-7.93482542e-01 5.13785109e-02 -4.88753110e-01 3.59476954e-01
8.51112545e-01 -4.66699719e-01 1.22881138e+00 -1.56069744e+00
-1.50163090e+00 4.63016719e-01 -9.56531391e-02 -1.05779141e-01
6.96552321e-02 6.51699424e-01 -1.22569454e+00 -4.14469317e-02
-5.36730066e-02 -2.18239814e-01 2.67283082e-01 -1.78677905e+00
-1.53196096e+00 -5.56772411e-01 1.88968554e-01 3.09590846e-01
-3.60684752e-01 5.92822134e-01 -7.32061923e-01 5.55301197e-02
6.19742274e-01 -7.60557413e-01 -4.95837033e-01 -5.42889416e-01
1.31889641e-01 -3.79720509e-01 8.05210173e-01 -2.71865457e-01
1.37918341e+00 -1.87379205e+00 1.32817179e-01 7.67833233e-01
-6.39948323e-02 -1.07321791e-01 -3.24035101e-02 8.71584892e-01
4.32776153e-01 1.31160170e-01 -5.22450089e-01 2.44314209e-01
3.69379580e-01 7.34088600e-01 1.79203928e-01 9.32380557e-02
-4.04189885e-01 1.59357131e-01 -7.27201939e-01 -5.43335855e-01
2.01464747e-03 5.23362696e-01 -7.19464481e-01 3.55722904e-01
-4.44374502e-01 3.66220355e-01 -7.87163258e-01 8.43984067e-01
4.93075728e-01 -7.27422759e-02 6.97276831e-01 -2.13703816e-03
-2.29102373e-01 2.72469074e-01 -1.63788092e+00 1.80469286e+00
-5.11841595e-01 -5.92642546e-01 2.78061926e-01 -9.50855196e-01
1.03081107e+00 6.67708933e-01 1.18040180e+00 -5.89660287e-01
2.16226637e-01 8.80439162e-01 3.21686596e-01 -4.46779191e-01
1.77925844e-02 -1.81305557e-01 1.04857728e-01 5.74882448e-01
-3.03816378e-01 2.06722140e-01 4.72956151e-01 -1.13998860e-01
9.55950916e-01 4.15801495e-01 1.45647272e-01 -2.39394248e-01
1.10518932e+00 8.13760981e-02 6.86922133e-01 1.52268291e-01
-4.09544520e-02 3.58092427e-01 3.13959509e-01 -4.13331449e-01
-1.10027611e+00 -9.42094684e-01 1.77504420e-01 1.07280064e+00
4.00731683e-01 -6.97629988e-01 -9.61809456e-01 -6.66124940e-01
-2.15241939e-01 6.64565682e-01 7.03988224e-02 4.72270101e-02
-6.13273561e-01 -5.79865575e-01 2.50715222e-02 -2.37092972e-02
2.32156605e-01 -1.30705953e+00 -5.60565174e-01 6.16283059e-01
-1.31799445e-01 -1.05448222e+00 -1.01212054e-01 -4.10452008e-01
-6.09521806e-01 -1.08654296e+00 -1.55237421e-01 -6.45399928e-01
6.44281507e-01 5.11291981e-01 8.44654560e-01 5.38199782e-01
3.74735519e-02 2.63235480e-01 -9.85380232e-01 6.26833290e-02
-3.76920760e-01 6.08832389e-02 -8.69381130e-02 6.18535876e-01
5.91546059e-01 -1.09964991e+00 -5.95504940e-01 6.74156010e-01
-1.44527090e+00 4.94205877e-02 2.36198425e-01 1.66612610e-01
5.37721634e-01 5.46460152e-01 5.63520849e-01 -8.04335952e-01
7.12270379e-01 -5.51487207e-01 -5.27248085e-01 1.85337186e-01
-1.20414150e+00 3.79882790e-02 8.57389450e-01 -1.06074125e-01
-1.14942718e+00 -1.98917732e-01 -4.60207045e-01 2.55828351e-01
-2.40825385e-01 4.64797735e-01 -1.06035638e+00 -6.40286058e-02
-3.79004031e-02 3.83954525e-01 -8.25640112e-02 -9.74184036e-01
3.89310390e-01 9.89283741e-01 -4.10626456e-03 -7.38703310e-01
5.04632533e-01 7.75188148e-01 -2.76608523e-02 -2.61452701e-03
-1.16185367e-01 -8.06309640e-01 -9.26755220e-02 -1.27925545e-01
5.89377999e-01 -1.40664741e-01 -9.27235782e-01 -1.38869479e-01
-6.64065659e-01 3.26683372e-01 -1.07223645e-01 1.54877320e-01
-6.04234993e-01 6.70761764e-02 -9.31459367e-02 -9.41790402e-01
-3.02041650e-01 -1.44793987e+00 7.73154616e-01 2.52121091e-01
4.31378186e-02 -7.12599337e-01 -1.26016065e-01 8.41447264e-02
6.13705993e-01 3.52051482e-02 7.59284794e-01 -1.10253954e+00
-5.85316062e-01 -1.97132617e-01 -1.34821594e-01 -4.71782908e-02
4.10086453e-01 -1.04518995e-01 -4.57333028e-01 2.79004183e-02
8.93247426e-02 6.93915546e-01 -4.36800644e-02 -3.28462869e-01
1.08956611e+00 -5.85387826e-01 -3.06203693e-01 5.29040635e-01
1.93688095e+00 8.43556404e-01 6.98745906e-01 6.58747435e-01
5.83604217e-01 9.58721817e-01 6.80308402e-01 6.38459504e-01
4.50630903e-01 9.48260784e-01 5.69858015e-01 3.22429568e-01
6.64491653e-02 -9.94387642e-02 1.83895826e-01 1.15722573e+00
-2.88959980e-01 -1.85352579e-01 -9.46454227e-01 4.32047486e-01
-2.38152957e+00 -9.00809765e-01 -1.44648537e-01 1.91956568e+00
5.34678876e-01 -2.72480011e-01 1.99917361e-01 4.74492580e-01
7.25482285e-01 -3.64338368e-01 -1.77984864e-01 -6.20541334e-01
1.43540785e-01 -1.56564519e-01 2.74131179e-01 3.83481622e-01
-3.14839244e-01 1.00424862e+00 5.02965784e+00 6.44835353e-01
-9.32434440e-01 4.47905749e-01 -1.34364262e-01 1.29170150e-01
-9.87567425e-01 5.01668751e-01 -5.55028498e-01 5.68642259e-01
1.06879616e+00 -5.49693406e-01 9.55266237e-01 9.67809975e-01
5.37295163e-01 2.05799311e-01 -6.13348246e-01 6.57941222e-01
9.56841558e-02 -1.13709295e+00 2.07638204e-01 -3.69892567e-02
7.40139782e-01 -4.06325489e-01 -6.04143918e-01 -1.87036008e-01
4.07246858e-01 -5.95617533e-01 6.60059988e-01 5.23783565e-01
2.71219373e-01 -1.06581676e+00 6.71005130e-01 2.44738862e-01
-1.35179126e+00 -3.09482604e-01 4.16094176e-02 1.03618205e-01
3.71209860e-01 3.42882723e-01 -5.59805751e-01 8.74106288e-01
9.94291544e-01 1.77852631e-01 -1.18063405e-01 1.26944637e+00
-1.09817665e-02 1.37414098e-01 -6.44429252e-02 -3.42158854e-01
2.82407194e-01 -7.44252980e-01 5.21198332e-01 9.26022112e-01
5.22491693e-01 2.45642334e-01 5.78054607e-01 6.29166305e-01
2.92703003e-01 8.55128944e-01 -2.95000583e-01 -1.27605945e-01
5.17896295e-01 1.26291680e+00 -1.00709355e+00 -4.93399978e-01
-6.11846328e-01 9.14514065e-01 -1.81673914e-01 3.25063497e-01
-7.08745956e-01 -4.01536703e-01 1.21191740e+00 3.34248811e-01
1.25115559e-01 -6.87765703e-02 -1.13139771e-01 -1.10496879e+00
6.37869537e-02 -1.11567414e+00 4.79901403e-01 -5.32483757e-01
-1.12938702e+00 7.27738619e-01 -1.36492252e-01 -1.19605494e+00
-2.41389051e-01 -1.86270550e-01 -2.28681579e-01 7.61939406e-01
-1.37525439e+00 -1.22574615e+00 -3.58425260e-01 7.99511909e-01
4.20971006e-01 -2.41601884e-01 1.19591117e+00 5.66088557e-01
-1.60326008e-02 -1.38781995e-01 9.59130377e-02 -3.95460933e-01
1.78251147e-01 -9.24897015e-01 -2.06787735e-02 8.91348422e-01
-3.82018313e-02 5.61939418e-01 9.40164149e-01 -5.57340324e-01
-1.66692543e+00 -6.29461288e-01 1.13915384e+00 1.95634931e-01
7.80937135e-01 -6.01110235e-02 -7.02032745e-01 2.67538160e-01
-1.31810859e-01 -7.85834119e-02 9.30935860e-01 -1.94337785e-01
-3.53648007e-01 -5.85727751e-01 -1.68309486e+00 9.01046574e-01
1.41038215e+00 -1.29130855e-01 -5.06816022e-02 3.98762561e-02
9.34970975e-01 3.03926587e-01 -1.02088714e+00 3.52766812e-01
6.06104434e-01 -1.04880834e+00 6.07880473e-01 -7.53577530e-01
3.83941224e-03 -1.00774598e+00 -5.87496221e-01 -1.16604888e+00
-1.84140965e-01 -3.23378056e-01 7.11323321e-02 1.34879982e+00
2.11188868e-01 -8.56903553e-01 6.51793182e-01 6.41381681e-01
-3.68508548e-01 -3.47001642e-01 -7.69212663e-01 -7.04368532e-01
-6.75397217e-01 -5.12911856e-01 1.76847553e+00 8.59723926e-01
1.69302925e-01 -2.50953913e-01 1.12226546e-01 1.33747891e-01
4.89678353e-01 6.65808141e-01 3.86299640e-01 -1.42636096e+00
-1.64384529e-01 -7.42864907e-01 -5.84734857e-01 -1.37027964e-01
5.55210970e-02 -9.87926185e-01 8.37676004e-02 -2.02242446e+00
-4.74034017e-03 -6.46615088e-01 -5.24884224e-01 3.18784893e-01
3.07625115e-01 7.96144009e-02 3.91875386e-01 3.83697391e-01
-8.52944851e-01 -3.50313191e-03 1.12239933e+00 3.01524550e-01
-2.30279371e-01 1.19462736e-01 -1.10865128e+00 5.67273617e-01
9.24446523e-01 -5.28834522e-01 -5.42807579e-01 -9.62494761e-02
-1.05462242e-02 -6.76442012e-02 -2.74135172e-01 -8.90565813e-01
6.14371449e-02 -8.84106755e-01 -5.33883989e-01 1.39490530e-01
3.30985114e-02 -1.86885309e+00 1.00093663e+00 3.76665711e-01
-1.89452052e-01 -2.10436825e-02 -6.23813272e-01 1.57695889e-01
-3.78647566e-01 -4.96579021e-01 5.07214963e-01 -4.79761481e-01
-1.12733197e+00 4.63942885e-01 -3.95399570e-01 -4.22303408e-01
1.26760197e+00 -2.89926771e-02 2.96848178e-01 -2.42524281e-01
-8.77154946e-01 1.97738372e-02 1.00140476e+00 8.17960620e-01
4.69635516e-01 -1.46698332e+00 -5.06547391e-01 2.60095075e-02
6.06286585e-01 -5.97563922e-01 -1.34164542e-01 5.11249185e-01
-5.35025716e-01 1.99436873e-01 -2.92048514e-01 -2.60593742e-01
-1.25580466e+00 7.48555720e-01 2.49162480e-01 1.89317346e-01
-6.99581027e-01 1.16322823e-02 -3.55993152e-01 -4.76867169e-01
1.46080807e-01 2.75516242e-01 -5.27575135e-01 -2.06980303e-01
7.23600924e-01 2.35446930e-01 1.04688205e-01 -8.94536853e-01
-6.07101440e-01 5.08538365e-01 6.17618918e-01 -6.32229865e-01
1.29700291e+00 -3.86706412e-01 -6.50877059e-01 2.75739551e-01
1.03161442e+00 6.48285374e-02 -7.06227899e-01 -1.54437035e-01
6.81772530e-01 -7.99321175e-01 -2.77398378e-01 -7.63401568e-01
-1.25277221e+00 8.91957134e-02 6.99650526e-01 6.97820246e-01
1.45281649e+00 1.15319826e-01 8.75345767e-01 -1.84354529e-01
8.65891695e-01 -1.21239376e+00 -6.45768464e-01 1.34225175e-01
4.57050711e-01 -6.70796633e-01 -5.17196596e-01 -6.93863094e-01
-6.72765017e-01 1.01443720e+00 3.85250121e-01 1.19249910e-01
6.33671105e-01 4.77422893e-01 1.61124811e-01 -1.60934612e-01
-7.33834326e-01 -8.67653489e-01 1.53397648e-02 7.62533486e-01
3.96103501e-01 3.47685456e-01 -1.28946996e+00 8.22895169e-01
6.07385077e-02 -1.30023826e-02 3.38178754e-01 1.20774269e+00
-9.60572958e-01 -2.00383687e+00 -3.75109911e-01 8.41133073e-02
-5.05772948e-01 7.72820637e-02 -1.84486970e-01 1.88647136e-01
3.74439687e-01 1.38569474e+00 -1.04855254e-01 -2.00316295e-01
5.64952552e-01 3.90453577e-01 1.36227533e-01 -4.55842227e-01
-6.81868196e-01 2.69116938e-01 2.71683335e-01 -7.94124842e-01
-5.95734060e-01 -6.51087642e-01 -1.94909966e+00 -4.15357828e-01
-5.33530787e-02 8.12891006e-01 1.38420916e+00 8.97362530e-01
3.77640903e-01 6.79469168e-01 8.38670313e-01 -2.12990031e-01
-3.10253233e-01 -3.81210417e-01 -6.46040738e-01 1.02706814e+00
-4.05156732e-01 -5.35793483e-01 -4.03654337e-01 6.80788979e-02] | [8.611181259155273, 6.954575538635254] |
5c1a013a-0228-4187-a242-8b6243988707 | multi-goal-multi-agent-path-finding-via | 2009.05161 | null | https://arxiv.org/abs/2009.05161v1 | https://arxiv.org/pdf/2009.05161v1.pdf | Multi-Goal Multi-Agent Path Finding via Decoupled and Integrated Goal Vertex Ordering | We introduce multi-goal multi agent path finding (MAPF$^{MG}$) which generalizes the standard discrete multi-agent path finding (MAPF) problem. While the task in MAPF is to navigate agents in an undirected graph from their starting vertices to one individual goal vertex per agent, MAPF$^{MG}$ assigns each agent multiple goal vertices and the task is to visit each of them at least once. Solving MAPF$^{MG}$ not only requires finding collision free paths for individual agents but also determining the order of visiting agent's goal vertices so that common objectives like the sum-of-costs are optimized. We suggest two novel algorithms using different paradigms to address MAPF$^{MG}$: a heuristic search-based search algorithm called Hamiltonian-CBS (HCBS) and a compilation-based algorithm built using the SMT paradigm, called SMT-Hamiltonian-CBS (SMT-HCBS). Experimental comparison suggests limitations of compilation-based approach. | ['Pavel Surynek'] | 2020-09-10 | null | null | null | null | ['multi-agent-path-finding'] | ['playing-games'] | [ 8.70463476e-02 3.52750719e-01 4.54231128e-02 7.70383552e-02
-5.52029550e-01 -7.10846066e-01 5.24896204e-01 5.98440230e-01
-5.36834955e-01 1.18834901e+00 -3.96363765e-01 -4.66352582e-01
-1.03174162e+00 -1.47488201e+00 -4.71406341e-01 -5.30610561e-01
-7.94276595e-01 1.17906606e+00 7.30193496e-01 -5.87342203e-01
3.88000697e-01 4.37883675e-01 -1.16444325e+00 -3.04625571e-01
6.77777052e-01 3.93517792e-01 5.63297391e-01 8.77936125e-01
-2.27058187e-01 4.13518101e-01 -5.48434615e-01 -2.01607808e-01
6.13403440e-01 -5.00972688e-01 -1.17613161e+00 -1.29527211e-01
-3.78830105e-01 -8.74200389e-02 -7.96744451e-02 1.12587917e+00
4.20260690e-02 5.48450947e-01 6.38335884e-01 -2.17604780e+00
-1.55775681e-01 5.66606879e-01 -7.60226488e-01 7.68177658e-02
8.54387403e-01 1.03519119e-01 1.08156109e+00 2.14732457e-02
8.98861527e-01 1.25581169e+00 3.27352226e-01 3.23558450e-01
-8.26101005e-01 -1.17484339e-01 5.93808591e-01 3.97700101e-01
-1.34352958e+00 3.56123820e-02 2.38461018e-01 -2.30191916e-01
1.57586396e+00 4.40746307e-01 6.05862975e-01 -3.46346498e-02
5.83266616e-01 1.45708695e-01 9.85325873e-01 -6.37476683e-01
5.13754547e-01 -3.01499784e-01 1.48709968e-01 1.05182219e+00
3.50443333e-01 2.39176527e-01 -2.24803671e-01 -3.28603119e-01
6.70740128e-01 -2.09331915e-01 3.15560848e-02 -3.13386708e-01
-1.30819070e+00 1.05754435e+00 1.76374257e-01 1.16092764e-01
-6.62241936e-01 5.78802109e-01 2.82412171e-01 5.21202564e-01
-2.61427134e-01 4.02316839e-01 -3.37045044e-01 1.79136656e-02
-1.72998652e-01 7.93362975e-01 8.24168444e-01 1.24948144e+00
1.15138757e+00 -3.16580653e-01 2.11078480e-01 3.05020154e-01
5.48506439e-01 3.97117287e-01 -2.27874964e-01 -1.12339079e+00
5.83739340e-01 8.77494276e-01 3.80963176e-01 -1.08172154e+00
-7.79612780e-01 1.90081820e-02 -2.85856992e-01 8.32745433e-01
3.96069318e-01 -1.61172330e-01 -6.32208169e-01 1.91200018e+00
6.02070153e-01 -1.16521910e-01 2.42973983e-01 6.88198030e-01
4.51817721e-01 9.83174443e-01 -2.08231118e-02 -6.59213305e-01
1.27335477e+00 -1.50197363e+00 -3.12284648e-01 -3.87397110e-01
7.18691170e-01 -4.02714938e-01 5.85905612e-01 1.99996978e-01
-1.25112891e+00 2.91539103e-01 -9.80723739e-01 5.12958407e-01
-6.48114860e-01 -8.60751688e-01 7.20964968e-01 6.23725057e-01
-1.49436247e+00 4.12558138e-01 -6.58931315e-01 -3.99699420e-01
-2.30531439e-01 7.33966827e-01 -3.38708282e-01 -2.10836977e-01
-9.19603646e-01 8.37098837e-01 4.30701703e-01 -3.82521778e-01
-1.28555036e+00 -2.13713432e-03 -8.27222466e-01 -9.05997232e-02
9.95438814e-01 -9.02733743e-01 1.13406301e+00 -4.36943710e-01
-1.34023809e+00 4.39603806e-01 -2.40661010e-01 -3.21587861e-01
3.18413377e-01 7.22645223e-01 -3.07667077e-01 1.39515936e-01
7.35432565e-01 4.24661279e-01 3.01244557e-01 -1.48101044e+00
-1.33204734e+00 -3.49270523e-01 7.17669487e-01 5.34711719e-01
3.40110779e-01 1.26455367e-01 -3.58365119e-01 3.24478075e-02
2.28284135e-01 -1.04699278e+00 -8.62167954e-01 -6.08435690e-01
-5.98786414e-01 -4.99993742e-01 4.07598048e-01 -1.07311837e-01
1.28515482e+00 -1.38760567e+00 4.74414468e-01 7.85424352e-01
3.59190434e-01 -3.38870108e-01 -4.52073693e-01 1.23993027e+00
3.15484941e-01 1.95903368e-02 -1.70633882e-01 -7.31723532e-02
2.13706642e-01 4.15102333e-01 5.31413257e-01 3.96825343e-01
-5.52433670e-01 5.34808040e-01 -1.22370791e+00 -4.26058263e-01
5.13565913e-02 -2.73455560e-01 -6.55707061e-01 -2.76998550e-01
-6.84824526e-01 6.04236573e-02 -7.45280027e-01 5.93882203e-01
5.91459930e-01 -1.49326026e-01 4.22362328e-01 6.08512223e-01
-6.77772224e-01 -8.85810852e-02 -1.62634683e+00 1.43139970e+00
2.89686043e-02 -8.05981308e-02 2.07225546e-01 -6.74423575e-01
7.02097237e-01 3.69979106e-02 9.21544075e-01 -8.02898288e-01
2.88507361e-02 3.44269335e-01 9.02247056e-02 -2.28030145e-01
7.25073457e-01 -1.17879650e-02 -4.08680975e-01 1.00284183e+00
-5.13554513e-01 6.37667626e-02 8.85581851e-01 5.67902207e-01
1.58679450e+00 -7.86421001e-02 4.20245826e-01 -4.75947529e-01
6.65764868e-01 7.39941955e-01 5.99794507e-01 1.01555812e+00
-4.45344269e-01 -2.55602956e-01 4.32508051e-01 -5.25153458e-01
-6.89722061e-01 -1.02634501e+00 7.11670935e-01 1.16368568e+00
7.73007333e-01 -6.71894073e-01 -7.81223416e-01 -5.94125688e-01
-1.00152291e-01 6.58363044e-01 -4.54327136e-01 2.39711523e-01
-7.33888388e-01 -6.95566535e-01 -4.92864177e-02 -2.85668910e-01
3.91998142e-01 -1.18067503e+00 -1.17366910e+00 8.41059625e-01
-1.71984404e-01 -7.16821134e-01 -4.96144742e-01 1.08527571e-01
-1.57780752e-01 -1.40695214e+00 -2.26602390e-01 -8.97093713e-01
8.94719005e-01 6.10948324e-01 8.46193254e-01 2.95093268e-01
-1.85781881e-01 5.24368703e-01 -7.09453940e-01 -4.47363794e-01
-2.91406929e-01 -2.45795753e-02 -6.27034679e-02 -4.48391110e-01
3.98119055e-02 -5.61956525e-01 -5.10408640e-01 4.84020740e-01
-5.48318863e-01 -4.96109091e-02 2.39358172e-01 2.10427031e-01
8.49459887e-01 8.20479631e-01 6.11889482e-01 -6.12915397e-01
1.00191855e+00 -4.50307488e-01 -7.26716697e-01 3.62036735e-01
-6.65354609e-01 -2.65837878e-01 6.65865660e-01 1.01985969e-01
-3.14480931e-01 -1.56645849e-01 1.29504278e-01 1.43931732e-01
1.14052042e-01 7.21330762e-01 1.84048098e-02 -3.20587963e-01
3.54726285e-01 1.41791835e-01 2.49050893e-02 6.92080781e-02
1.64905012e-01 -1.11138925e-01 3.25491548e-01 -5.92046022e-01
4.98922020e-01 4.27403897e-01 5.94235778e-01 -2.76515365e-01
1.92086548e-01 -2.95991302e-01 -1.91335991e-01 -4.95124161e-01
7.38133192e-01 -2.46347532e-01 -1.39824939e+00 2.81134397e-01
-1.04181564e+00 -6.58058286e-01 -3.06129865e-02 1.13297306e-01
-8.72463763e-01 2.60517389e-01 -3.39300215e-01 -9.47137833e-01
-2.50591695e-01 -1.37277663e+00 6.26734197e-01 2.16385707e-01
-8.97149891e-02 -9.76434112e-01 4.67897862e-01 -7.21835857e-03
2.17376083e-01 6.95841432e-01 1.21155882e+00 -5.81397712e-01
-9.77334082e-01 -3.59590873e-02 -1.23358600e-01 -8.38193893e-01
4.47764210e-02 -2.74987072e-01 3.40337068e-01 -6.85265005e-01
-4.80448931e-01 3.39514762e-01 1.93938110e-02 7.17250228e-01
9.28009450e-02 -5.22010624e-01 -9.59789991e-01 -5.54758981e-02
2.10728812e+00 1.14574718e+00 5.76441228e-01 1.21890581e+00
6.04129583e-02 6.20197654e-01 7.64013588e-01 4.96562749e-01
1.06523311e+00 8.38981569e-01 8.67551565e-01 3.59003216e-01
2.07042664e-01 1.72704428e-01 2.77571946e-01 1.57648385e-01
-3.75120759e-01 -8.41338396e-01 -1.12796807e+00 5.48238635e-01
-2.23943663e+00 -8.19312930e-01 -4.60178792e-01 2.09540725e+00
-2.78250892e-02 1.44678295e-01 5.76554000e-01 7.31374919e-02
8.52146447e-01 -1.68280303e-01 -3.48794460e-01 -8.41157019e-01
1.91107824e-01 -8.71304721e-02 6.16534770e-01 1.16177344e+00
-6.73116386e-01 9.82876062e-01 5.81895542e+00 4.89516765e-01
-3.83183300e-01 2.51568437e-01 -3.98219563e-03 9.68777388e-02
-3.48716885e-01 4.61815894e-01 -7.17396080e-01 2.83676207e-01
7.10546196e-01 -6.03671193e-01 9.45168138e-01 5.49846292e-01
4.65990424e-01 -6.20231330e-01 -8.10909688e-01 5.96531808e-01
-2.50028104e-01 -1.23903823e+00 -6.75921291e-02 4.53118265e-01
6.44437492e-01 -2.33049482e-01 -3.92062932e-01 1.13043383e-01
9.65546846e-01 -7.44099677e-01 8.20017874e-01 -5.44564351e-02
3.24221104e-01 -1.10357213e+00 2.41078407e-01 4.27555978e-01
-1.77835822e+00 -2.24264652e-01 -1.04025602e-02 -3.44509304e-01
8.56251001e-01 -6.04274608e-02 -8.22237313e-01 1.06272447e+00
7.10921347e-01 -9.46575701e-02 1.32561222e-01 1.13052201e+00
1.94328558e-02 -6.01754010e-01 -2.98548549e-01 -3.16192269e-01
8.98447156e-01 -7.80405343e-01 1.16613412e+00 6.92470193e-01
4.39895570e-01 4.07500714e-01 9.26074982e-01 5.37224829e-01
6.45759463e-01 1.59214631e-01 -5.25527954e-01 1.12156384e-01
6.93731725e-01 1.03857446e+00 -1.39093220e+00 -7.89511204e-02
-1.26841396e-01 7.71928132e-01 2.14880437e-01 2.88507044e-01
-8.51107657e-01 -6.35675490e-01 7.74190664e-01 6.45145103e-02
-6.56840727e-02 -5.25393903e-01 6.49276748e-02 -1.48719415e-01
-2.88219780e-01 -7.36315072e-01 7.70872712e-01 -6.72494590e-01
-5.50395608e-01 8.67111623e-01 3.24201256e-01 -9.24664259e-01
-4.20263827e-01 -2.05307588e-01 -7.98608363e-01 7.54646301e-01
-1.40547681e+00 -1.02027309e+00 -4.45113741e-02 1.04200220e+00
4.39098060e-01 -3.51179421e-01 8.42715144e-01 5.89007698e-02
-5.24084687e-01 1.60413057e-01 -7.89841413e-02 -4.85252380e-01
-2.38749430e-01 -1.13426888e+00 1.98515311e-01 9.89144564e-01
-4.60781455e-01 4.35448229e-01 9.59170699e-01 -8.97734880e-01
-1.73327720e+00 -8.47431779e-01 8.63926530e-01 1.16790444e-01
4.78473663e-01 1.38477594e-01 -1.21543333e-01 9.16407943e-01
3.69892120e-01 -7.01416075e-01 3.53430301e-01 -2.01772615e-01
2.64062494e-01 2.10889444e-01 -1.60384417e+00 8.89435172e-01
1.15121984e+00 3.69784504e-01 3.95437982e-03 5.69203079e-01
7.95282960e-01 -3.72922271e-01 -5.91286659e-01 1.85404867e-01
1.19020447e-01 -9.30410802e-01 9.11470115e-01 -4.07932490e-01
-3.74974310e-03 -7.32023656e-01 -2.73570120e-01 -1.53772914e+00
-7.43545115e-01 -8.81705046e-01 5.31791985e-01 6.61251962e-01
7.03126609e-01 -1.12460291e+00 8.87889922e-01 4.32440102e-01
-5.65284014e-01 -7.38438964e-01 -1.25181401e+00 -1.06224978e+00
-1.12172335e-01 -1.35467395e-01 9.67177153e-01 8.74390006e-01
3.58960956e-01 1.48469627e-01 -9.60954279e-02 6.83128238e-01
9.06463027e-01 3.09463948e-01 7.33283460e-01 -9.62971747e-01
-4.32749748e-01 -5.75160503e-01 -1.43018454e-01 -4.12236869e-01
-8.45731795e-02 -8.70553970e-01 3.46630178e-02 -2.44950533e+00
-5.87609597e-03 -9.63749528e-01 9.95859727e-02 6.76021695e-01
4.77234781e-01 -2.72515804e-01 3.48910123e-01 -3.82897183e-02
-9.33340967e-01 -8.02747682e-02 1.49554896e+00 2.42594583e-03
-5.46464026e-01 1.79255236e-04 -6.06491804e-01 3.50561589e-01
7.91325450e-01 -5.84415793e-01 -7.29295850e-01 -3.41188252e-01
5.64188540e-01 7.41334915e-01 6.48208410e-02 -7.49874651e-01
6.58774674e-01 -1.01148427e+00 -7.62738287e-01 -4.50915039e-01
4.12686348e-01 -8.11377287e-01 8.38216484e-01 1.00617611e+00
1.40021652e-01 6.65815651e-01 2.63794195e-02 5.48579514e-01
1.37965217e-01 -6.15438104e-01 4.35229748e-01 -6.50755942e-01
-1.11590278e+00 2.74560809e-01 -6.81283951e-01 -2.66757220e-01
1.85855281e+00 -5.02942920e-01 -8.05973113e-01 -4.42111462e-01
-7.58986235e-01 7.94652641e-01 3.39276046e-01 6.89873472e-02
7.60830879e-01 -1.08072960e+00 -6.09139979e-01 -1.89241663e-01
-9.83909294e-02 1.06659375e-01 2.85368621e-01 6.06342733e-01
-8.76564264e-01 4.99273002e-01 -3.78238618e-01 1.16130579e-02
-1.23527014e+00 7.37051725e-01 3.73146474e-01 -6.98857725e-01
-6.21425211e-01 1.02132595e+00 -1.10340923e-01 -3.05139542e-01
-4.20185663e-02 2.45492920e-01 -1.58737078e-01 -2.85511374e-01
4.93502557e-01 8.42274427e-01 -1.96209043e-01 -6.43167496e-01
-5.80639243e-01 5.44623435e-01 -1.19865052e-01 -4.91116375e-01
1.34081852e+00 -5.19246519e-01 -6.00376964e-01 -3.87358546e-01
5.43274283e-01 -1.68785542e-01 -6.08674526e-01 1.37799367e-01
-3.92381027e-02 -5.28391242e-01 -3.13190043e-01 -7.92455673e-01
-7.89056838e-01 -1.98795646e-01 1.74921840e-01 7.37042725e-01
9.89146829e-01 1.83011591e-02 6.32583618e-01 2.62697637e-01
1.42680371e+00 -1.03333008e+00 7.86645263e-02 7.06745803e-01
6.68403208e-01 -6.15621984e-01 -9.79470089e-02 -8.75725210e-01
-4.64774787e-01 8.37205052e-01 7.38754630e-01 -2.03327075e-01
3.58542114e-01 2.26057604e-01 -3.17079425e-01 -6.65386915e-01
-6.42882407e-01 -4.52983648e-01 -6.18280530e-01 8.87558281e-01
-6.16270602e-01 2.60023475e-01 -6.10649407e-01 2.75625020e-01
-3.62699568e-01 -2.47416973e-01 8.13406289e-01 1.43882108e+00
-9.51738954e-01 -1.54360414e+00 -5.66908181e-01 3.39457035e-01
1.08849712e-01 2.21849546e-01 -2.80801356e-01 9.30215001e-01
2.13939041e-01 1.38850534e+00 -5.11453953e-03 -3.30907345e-01
5.32949328e-01 -4.16675448e-01 6.02474630e-01 -6.25455439e-01
-5.95661819e-01 -3.82907428e-02 5.44680834e-01 -5.09511352e-01
-4.08631772e-01 -7.98280597e-01 -1.95226729e+00 -7.99549222e-01
-4.56596464e-02 5.34228086e-01 5.53074002e-01 9.38420057e-01
2.06913128e-02 6.40923798e-01 6.14015937e-01 -6.52234972e-01
6.70379847e-02 -1.74369514e-01 -7.39858985e-01 -7.99275935e-02
-5.36161885e-02 -7.86518633e-01 -1.27205700e-01 -5.50830245e-01] | [4.9755024909973145, 1.7805142402648926] |
48acd86c-4e72-4e5c-bed7-6c0f8c59cbc1 | study-of-proximal-normalized-subband-adaptive | 2108.10219 | null | https://arxiv.org/abs/2108.10219v1 | https://arxiv.org/pdf/2108.10219v1.pdf | Study of Proximal Normalized Subband Adaptive Algorithm for Acoustic Echo Cancellation | In this paper, we propose a novel normalized subband adaptive filter algorithm suited for sparse scenarios, which combines the proportionate and sparsity-aware mechanisms. The proposed algorithm is derived based on the proximal forward-backward splitting and the soft-thresholding methods. We analyze the mean and mean square behaviors of the algorithm, which is supported by simulations. In addition, an adaptive approach for the choice of the thresholding parameter in the proximal step is also proposed based on the minimization of the mean square deviation. Simulations in the contexts of system identification and acoustic echo cancellation verify the superiority of the proposed algorithm over its counterparts. | ['Qiangming Cai', 'Lu Lu', 'Zongsheng Zheng', 'Rodrigo C. de Lamare', 'Yi Yu', 'Gang Guo'] | 2021-08-14 | null | null | null | null | ['acoustic-echo-cancellation', 'acoustic-echo-cancellation'] | ['medical', 'speech'] | [ 5.71409941e-01 -8.78384039e-02 2.87980467e-01 -4.03544128e-01
-7.07957208e-01 -2.90260524e-01 2.99170315e-01 -6.17388412e-02
-4.76900309e-01 5.77925622e-01 3.32585782e-01 -1.80097193e-01
-6.78209245e-01 -2.13720828e-01 -2.59662986e-01 -1.16985464e+00
-6.07358292e-02 -1.92723379e-01 -2.90424824e-02 -4.27889898e-02
2.65075415e-01 3.66897196e-01 -1.33660412e+00 -1.54158548e-01
1.02912652e+00 1.33573735e+00 1.92072734e-01 6.51568115e-01
2.66308457e-01 2.62486875e-01 -2.96947658e-01 -3.73959728e-02
5.79049110e-01 -6.70738399e-01 9.05512050e-02 1.65454447e-01
2.12397307e-01 -1.91980615e-01 4.26313989e-02 1.10850799e+00
8.11885059e-01 4.07209814e-01 5.02668679e-01 -5.86417139e-01
1.31334439e-01 6.93282366e-01 -5.81485331e-01 5.27261019e-01
2.61894941e-01 -1.98273420e-01 7.78994083e-01 -1.26463389e+00
7.97330961e-02 9.35745001e-01 9.15665090e-01 3.29885341e-04
-8.70274901e-01 -6.99312031e-01 -1.74604543e-02 9.88060832e-02
-1.68830407e+00 -9.63842571e-01 8.37270916e-01 -2.89304912e-01
3.03157508e-01 2.01564118e-01 5.66860139e-01 3.42649072e-01
-9.54985321e-02 3.27168256e-01 1.26190829e+00 -9.41671968e-01
3.72245669e-01 -6.80774078e-02 2.60942847e-01 5.80365300e-01
6.34013295e-01 3.00186723e-01 -5.89999676e-01 -5.65454066e-01
5.60037196e-01 -4.67600316e-01 -6.06615424e-01 -2.58896947e-01
-8.67375612e-01 6.26341403e-01 -6.32426590e-02 5.95162570e-01
-6.18544042e-01 -8.25978257e-03 1.84655905e-01 1.59928888e-01
6.32741630e-01 3.26592624e-02 1.15437277e-01 8.01818743e-02
-1.59328043e+00 1.60658315e-01 9.12784815e-01 6.20540857e-01
3.29002231e-01 6.08048618e-01 -2.49130815e-01 7.02710152e-01
7.59371400e-01 7.25752056e-01 1.42727554e-01 -8.83663356e-01
2.18393460e-01 -9.64120477e-02 2.87297338e-01 -1.02876067e+00
-3.59434336e-01 -1.05816531e+00 -8.78378451e-01 -1.13836892e-01
4.32290256e-01 -5.36120534e-01 -6.67655051e-01 1.63454378e+00
6.46951020e-01 4.73634064e-01 1.07260589e-02 1.04631770e+00
4.14483875e-01 5.52399158e-01 -2.18406484e-01 -8.48890364e-01
8.53249788e-01 -4.45778757e-01 -1.30299282e+00 1.66769940e-02
9.99616534e-02 -8.50061893e-01 3.14010322e-01 5.62313795e-01
-1.35798144e+00 -5.37050605e-01 -1.05193889e+00 6.67116761e-01
2.96440393e-01 2.81768173e-01 3.39724906e-02 1.12023997e+00
-9.86768901e-01 4.20612574e-01 -5.46838343e-01 -1.87330171e-01
-1.32762060e-01 2.64475346e-01 1.15950644e-01 3.67748827e-01
-9.51292098e-01 4.48342592e-01 2.34773949e-01 6.54259920e-01
-2.37975225e-01 -5.48689723e-01 -5.97078681e-01 2.49937102e-01
1.06866293e-01 -5.40628672e-01 1.18383431e+00 -1.14402282e+00
-1.69889903e+00 1.41444385e-01 -3.12610626e-01 -6.05216384e-01
5.71494043e-01 -2.45620534e-01 -5.28349042e-01 4.26380426e-01
-1.93110108e-01 -3.58380765e-01 1.43764317e+00 -1.06906247e+00
-7.18309999e-01 -1.88633427e-01 -5.21754026e-01 2.86068380e-01
-3.70394289e-01 -2.61027992e-01 -4.91908863e-02 -9.17510986e-01
6.22354925e-01 -5.96192539e-01 -4.83926564e-01 -2.10808307e-01
-9.47895646e-02 4.50432211e-01 4.86791164e-01 -7.89030612e-01
1.70181978e+00 -2.63770342e+00 -1.13604024e-01 1.00957060e+00
-1.48186073e-01 2.31736422e-01 1.82381138e-01 7.02477813e-01
-2.23281626e-02 -5.73034763e-01 -5.25134504e-01 -4.70612608e-02
-1.28117070e-01 -2.17803642e-01 -3.26371610e-01 9.41728473e-01
-2.22078890e-01 -1.81078210e-01 -5.64646542e-01 -4.01059330e-01
2.03000903e-01 4.07713950e-01 -5.80091178e-01 1.04584426e-01
7.21465111e-01 3.40386569e-01 -3.10652971e-01 5.20313323e-01
9.66788411e-01 2.22246081e-01 5.42531252e-01 -6.57182992e-01
-5.39486825e-01 1.78777874e-01 -1.90933025e+00 1.11406362e+00
-2.89653212e-01 3.19053054e-01 9.40755546e-01 -1.35303462e+00
9.99924004e-01 6.40693843e-01 5.24984121e-01 -4.34605896e-01
3.82911056e-01 5.48001826e-01 2.12687537e-01 -3.45624536e-01
4.45536643e-01 -3.43748033e-01 3.21136534e-01 1.82132602e-01
-7.81155825e-02 2.75210470e-01 3.47479075e-01 2.37852428e-02
6.74858570e-01 -1.83755040e-01 8.72322321e-01 -9.74225640e-01
9.98503208e-01 -4.43499446e-01 5.24674773e-01 9.51922178e-01
-2.80758619e-01 2.80476689e-01 -2.15598911e-01 1.32217646e-01
-6.95289135e-01 -7.99765408e-01 -3.50728333e-01 8.38671029e-01
1.72639757e-01 -3.66092212e-02 -7.43888557e-01 -3.57015021e-02
-8.25960711e-02 6.13997400e-01 -3.07892710e-01 3.94040607e-02
-6.99966252e-01 -7.51124024e-01 3.49823266e-01 1.89921230e-01
3.87771577e-01 -2.57021546e-01 -6.75139368e-01 4.55628574e-01
-1.71423644e-01 -1.10505497e+00 -4.01335061e-01 2.98682928e-01
-8.63613009e-01 -7.86822975e-01 -7.20887959e-01 -7.43251979e-01
6.19567096e-01 4.47416425e-01 3.11359465e-01 -1.00598782e-01
2.97461629e-01 7.86844015e-01 -5.19414663e-01 -5.94858117e-02
-2.86127001e-01 -3.20011348e-01 3.34971309e-01 8.42275500e-01
-2.46617958e-01 -7.60973096e-01 -6.74026132e-01 1.82675123e-01
-6.07365608e-01 -3.36491555e-01 7.96919405e-01 8.05711329e-01
4.59307760e-01 4.25403938e-02 6.99764013e-01 -6.22758031e-01
9.62307513e-01 -4.41355616e-01 -6.70788944e-01 6.44411817e-02
-7.92240679e-01 -8.49869549e-02 5.96037328e-01 -4.55673575e-01
-1.40868437e+00 2.19157293e-01 -3.80356133e-01 -1.58901334e-01
1.99070036e-01 5.04613459e-01 5.66255040e-02 -5.68847239e-01
3.48821282e-01 5.30771494e-01 -7.57960323e-03 -5.76352477e-01
4.10457969e-01 7.86534429e-01 6.25433445e-01 -3.20522904e-01
9.70014095e-01 5.05640447e-01 6.78373799e-02 -1.31077611e+00
-3.72442067e-01 -7.18269587e-01 -2.85558999e-01 -4.00062710e-01
2.99521804e-01 -7.31391013e-01 -6.17688179e-01 3.59354079e-01
-8.00540864e-01 1.40339568e-01 -3.04456323e-01 9.93912637e-01
-4.93486494e-01 7.51269102e-01 -2.91024685e-01 -1.45721483e+00
-6.70286655e-01 -6.47993565e-01 5.14796793e-01 5.52402474e-02
-1.45755321e-01 -7.97875941e-01 7.72806928e-02 -1.74624220e-01
5.98912358e-01 -1.82761252e-02 3.47263575e-01 -7.04513192e-01
-1.85624555e-01 -1.39170349e-01 -4.33619693e-02 3.41282338e-01
-2.98962872e-02 -1.99121982e-01 -7.35397339e-01 -5.20475090e-01
4.08774436e-01 3.54543775e-01 6.84121907e-01 9.43500519e-01
4.72647816e-01 -4.85173821e-01 -1.91384614e-01 6.77990139e-01
1.59167910e+00 3.08474362e-01 2.56278068e-01 8.40698332e-02
6.35647327e-02 3.23217034e-01 6.52887702e-01 1.05693138e+00
-2.21987516e-01 5.98245025e-01 1.63908020e-01 -5.81073156e-03
-7.12591261e-02 2.36018538e-01 1.86167955e-01 1.22775269e+00
2.87740994e-02 -1.71656832e-01 -3.74371886e-01 4.99629200e-01
-1.66256118e+00 -9.64088023e-01 -2.83745378e-01 2.30312967e+00
4.94954914e-01 3.08637675e-02 -9.83039588e-02 7.22870409e-01
7.60801971e-01 2.04137206e-01 -1.08778454e-01 -5.82122743e-01
-1.38455853e-01 3.80144566e-01 6.88574970e-01 7.69242704e-01
-9.04197454e-01 2.36370549e-01 6.89689445e+00 9.55038607e-01
-9.91504550e-01 1.40898585e-01 6.20253664e-03 1.90690428e-01
-1.19067822e-02 -3.96889299e-02 -8.09554696e-01 4.40104246e-01
8.45539808e-01 -4.17214870e-01 2.31575266e-01 5.08505881e-01
7.01627493e-01 -3.29849005e-01 -3.57132673e-01 8.97442639e-01
1.05389312e-01 -8.73950243e-01 -2.34825343e-01 -1.77966163e-01
4.71706599e-01 -5.46701074e-01 -2.96342955e-03 -2.53885955e-01
-4.16316271e-01 -1.78342730e-01 1.10421014e+00 5.93420804e-01
3.20956439e-01 -5.95914245e-01 5.25768220e-01 3.60731989e-01
-1.42520165e+00 -4.73605603e-01 3.02389599e-02 -3.24112564e-01
4.24286991e-01 1.14797103e+00 -6.88219190e-01 6.69425189e-01
5.10881126e-01 2.59506643e-01 1.78606287e-01 1.51193702e+00
-9.82639715e-02 1.09947157e+00 -7.08395541e-01 -3.65721732e-02
1.23595595e-01 -5.71583331e-01 9.72298145e-01 1.59766734e+00
7.96240091e-01 4.51974392e-01 2.51786970e-02 4.44826335e-01
4.40340519e-01 4.52557743e-01 -2.31354818e-01 3.49115282e-01
7.28928626e-01 1.09065449e+00 -5.10520160e-01 -2.13266000e-01
-4.60071921e-01 3.62141907e-01 -3.19810897e-01 5.32471716e-01
-5.28344750e-01 -5.61227679e-01 1.99868172e-01 1.65891096e-01
8.02023768e-01 -4.11286771e-01 -3.77051681e-01 -6.34647787e-01
6.10047393e-02 -9.15793121e-01 4.56552148e-01 -1.18551478e-01
-8.58520985e-01 4.75564033e-01 1.65439889e-01 -1.50127852e+00
-3.04699421e-01 -2.06188247e-01 -5.88663697e-01 7.73525238e-01
-1.48591888e+00 -3.91931325e-01 4.23143394e-02 5.92119455e-01
2.77506351e-01 -2.93990467e-02 3.30308497e-01 5.77680111e-01
-4.29076642e-01 6.57334268e-01 5.67249537e-01 -2.93188989e-01
3.31014693e-01 -6.34443939e-01 -2.60143578e-01 1.31309140e+00
-1.72656178e-01 7.73423135e-01 1.24551630e+00 -3.83941531e-01
-1.12782121e+00 -5.97203612e-01 8.31603408e-01 7.39262462e-01
5.51249743e-01 -1.72346503e-01 -7.92776287e-01 1.13797143e-01
2.26730853e-01 -3.18240047e-01 7.81402588e-01 -2.11346924e-01
9.17937160e-02 -4.73313093e-01 -1.20548213e+00 3.06785285e-01
6.15536213e-01 -1.04746915e-01 -4.37491924e-01 1.67200923e-01
-3.10111362e-02 -4.19588178e-01 -7.94539034e-01 6.13958657e-01
7.13132620e-01 -1.07932997e+00 8.79911482e-01 1.82378426e-01
-3.62705141e-01 -4.67849851e-01 -5.45193315e-01 -1.03698075e+00
-6.09361231e-01 -9.26627517e-01 -2.55949706e-01 1.11097538e+00
3.52861583e-01 -6.84099555e-01 5.40452838e-01 -1.63630187e-03
-1.39594883e-01 -5.65997124e-01 -1.35219502e+00 -9.60813344e-01
-6.69701338e-01 -3.86427850e-01 -8.88348892e-02 3.88788879e-01
-4.13891971e-02 2.76758131e-02 -6.51255488e-01 5.46534419e-01
9.97967720e-01 1.36610910e-01 3.21018666e-01 -8.11077297e-01
-6.30914748e-01 -2.89640963e-01 -1.88365713e-01 -1.29015434e+00
-3.01298380e-01 -3.57870787e-01 2.99803883e-01 -1.19499135e+00
-2.95748830e-01 -3.16076964e-01 -4.64754701e-01 -1.96200013e-01
-1.22917883e-01 1.80679232e-01 1.30447432e-01 -6.76964642e-03
-2.06255943e-01 4.76496518e-01 7.60239542e-01 2.29290292e-01
-3.22232604e-01 5.04960477e-01 -5.16499817e-01 7.37420142e-01
6.40531480e-01 -4.68680561e-01 -2.80383080e-01 -8.13439935e-02
-1.18503021e-03 2.10442021e-01 -2.53229844e-03 -1.30074286e+00
3.65142077e-01 1.45853356e-01 5.05647063e-02 -5.20149648e-01
5.08112669e-01 -1.10873508e+00 1.53873518e-01 7.50160515e-01
-2.61473060e-01 -2.98328906e-01 1.03093069e-02 8.55353951e-01
-1.98504180e-01 -3.77844006e-01 1.14451790e+00 3.25361460e-01
-4.33988214e-01 -2.52018899e-01 -7.58300781e-01 -1.83745369e-01
6.86534107e-01 -4.63074654e-01 3.06630194e-01 -7.73755372e-01
-8.30960929e-01 -1.16736695e-01 -3.70859861e-01 -3.16843063e-01
5.97621024e-01 -9.92007136e-01 -8.89813900e-01 3.61912340e-01
-4.04034168e-01 -9.16538060e-01 3.77270907e-01 1.45009470e+00
-2.00227648e-01 3.83156300e-01 1.34758741e-01 -4.92923081e-01
-1.29333866e+00 1.62625238e-01 4.20339674e-01 -5.26788943e-02
-4.01561052e-01 6.04220867e-01 1.69463456e-02 1.66035473e-01
3.60605538e-01 -1.01850070e-01 -1.39804408e-01 -1.19318388e-01
5.85967243e-01 7.25515127e-01 1.76233664e-01 -7.40560830e-01
-3.45005512e-01 8.18126678e-01 4.64359790e-01 -3.48298937e-01
9.67384517e-01 -5.72566688e-01 -1.52772516e-01 2.63884872e-01
8.47945571e-01 6.04868174e-01 -9.68458831e-01 -3.17540973e-01
-1.29870042e-01 -6.11117780e-01 2.65864879e-01 -4.54406798e-01
-1.00034606e+00 5.03313899e-01 9.35846150e-01 2.85664737e-01
1.57972169e+00 -6.51577115e-01 5.91783822e-01 3.17063987e-01
2.29109243e-01 -9.97761488e-01 -5.00138164e-01 2.61410594e-01
5.77251911e-01 -6.79791689e-01 3.01698416e-01 -7.02718496e-01
2.37610098e-02 1.15004885e+00 1.31230503e-02 -1.86703727e-01
9.20106590e-01 5.36122859e-01 1.78583398e-01 3.50585252e-01
-3.90747845e-01 -3.34902465e-01 2.83528179e-01 4.41297114e-01
4.62198615e-01 -8.70765746e-02 -1.38495529e+00 6.14812553e-01
1.07367106e-01 -2.52904352e-02 4.01071817e-01 9.52920377e-01
-8.92076612e-01 -8.29044044e-01 -8.47390473e-01 3.29354763e-01
-5.98165035e-01 -2.33941719e-01 5.05383462e-02 5.23147285e-01
-2.04471350e-02 1.40791869e+00 -2.57067412e-01 -1.57293692e-01
4.36493695e-01 -9.92648900e-02 4.62437600e-01 1.99695788e-02
-6.19534135e-01 6.15319371e-01 1.74580872e-01 -1.70925453e-01
-6.69816196e-01 -7.43909955e-01 -9.91105199e-01 -1.56317576e-04
-6.67770624e-01 5.87652981e-01 6.18287504e-01 9.51846838e-01
2.02465281e-01 3.27725142e-01 8.54157090e-01 -7.09253013e-01
-9.35885966e-01 -8.30891609e-01 -8.37892473e-01 5.45109548e-02
4.34912205e-01 -4.05280948e-01 -7.13567257e-01 -3.46877836e-02] | [15.091806411743164, 5.730691432952881] |
033dfa78-fa8f-49eb-9b04-11f6853c8be8 | a-survey-on-organoid-image-analysis-platforms | 2301.02341 | null | https://arxiv.org/abs/2301.02341v1 | https://arxiv.org/pdf/2301.02341v1.pdf | A survey on Organoid Image Analysis Platforms | An in-vitro cell culture system is used for biological discoveries and hypothesis-driven research on a particular cell type to understand mechanistic or test pharmaceutical drugs. Conventional in-vitro cultures have been applied to primary cells and immortalised cell lines plated on 2D surfaces. However, they are unreliable in complex physiological environments and can not always predict in-vivo behaviour correctly. Organoids are multicellular spheroids of a primary donor or stem cells that are replaced in vitro cell culture systems and are widely used in biological, biomedical and translational studies. Native heterogeneity, microanatomy, and functionality of an organ or diseased tissue can be represented by three-dimensional in-vitro tissue models such as organoids. Organoids are essential in in-vitro models for drug discovery and personalised drug screening. Many imaging artefacts such as organoid occlusion, overlap, out-of-focus spheroids and considerable heterogeneity in size cause difficulty in conventional image processing. Despite the power of organoid models for biology, their size and shape have mostly not been considered. Drug responses depend on dynamic changes in individual organoid morphology, number and size, which means differences in organoid shape and size, movement through focal planes, and live-cell staining with limited options cause challenges for drug response and growth analysis. This study primarily introduces the importance of the role of the organoid culture system in different disciplines of medical science and various scopes of utilising organoids. Then studies the challenges of operating organoids, followed by reviewing image analysis systems or platforms applied to organoids to address organoid utilising challenges. | ['Azadeh Nazemi', 'Alireza Ranjbaran'] | 2023-01-06 | null | null | null | null | ['culture'] | ['speech'] | [-1.94072127e-01 -1.59154847e-01 -9.48107019e-02 7.70419836e-01
-2.15936661e-01 -7.92587638e-01 5.01531243e-01 6.63442492e-01
-2.78632134e-01 7.96216309e-01 8.61494392e-02 -3.06683183e-01
3.69467378e-01 -3.72196585e-01 -4.15879369e-01 -9.06792164e-01
-3.16105746e-02 5.65623701e-01 4.73237157e-01 1.83588918e-02
-2.55909786e-02 1.06016445e+00 -8.35770965e-01 4.49751923e-03
2.39956006e-01 4.58723664e-01 2.36173466e-01 8.19184601e-01
-6.47335723e-02 1.47141472e-01 -4.76178765e-01 3.29732388e-01
8.45787153e-02 -4.70251322e-01 -2.24354863e-01 3.08285058e-01
-2.86066949e-01 3.04121226e-02 3.33096683e-01 3.78890395e-01
6.99937522e-01 -5.19081891e-01 9.01905715e-01 -8.88398468e-01
-3.58462930e-01 -3.86467874e-01 -1.76871404e-01 1.57193184e-01
6.87409699e-01 1.73308596e-01 -2.10592374e-01 -9.70422804e-01
1.06086576e+00 1.07675052e+00 4.64927882e-01 6.71387792e-01
-1.58461010e+00 -2.18515635e-01 -3.00897866e-01 -3.31963509e-01
-1.26947248e+00 -4.43603992e-01 -1.20216802e-01 -8.75739753e-01
7.92157114e-01 5.58751941e-01 8.79371643e-01 6.76907003e-01
1.18259716e+00 5.31692728e-02 1.26505637e+00 -3.53866518e-01
4.85710800e-01 6.40822232e-01 -3.99785519e-01 2.51745850e-01
5.38123310e-01 -2.56040618e-02 -1.00278601e-01 -2.04094246e-01
1.32036853e+00 2.15459734e-01 -5.70559502e-01 -4.28241342e-01
-1.26937687e+00 3.60495180e-01 -1.30007535e-01 8.40234399e-01
-2.51681685e-01 -1.84548140e-01 4.48605984e-01 7.98019469e-02
-7.76622593e-02 2.98111200e-01 -4.69724298e-01 -1.44017758e-02
-5.82290590e-01 7.55044492e-03 9.04953122e-01 3.54754776e-01
-9.76283327e-02 -1.37602165e-01 7.20802695e-02 5.97473979e-01
2.91967750e-01 3.52457076e-01 1.00354135e+00 -5.79846799e-01
-5.67199409e-01 9.97995079e-01 -3.48221697e-02 -5.67226648e-01
-6.49629891e-01 -1.38651848e-01 -7.15690136e-01 5.37279069e-01
6.69320583e-01 2.37948149e-01 -8.67490053e-01 1.06433249e+00
8.39105189e-01 2.63399258e-02 2.11497411e-01 7.57602036e-01
7.89971411e-01 5.69197059e-01 3.08640718e-01 -4.85566288e-01
1.66490316e+00 -1.85641930e-01 -4.35433954e-01 3.94662261e-01
1.04532552e+00 -8.13713193e-01 1.03771496e+00 3.00337493e-01
-9.92903173e-01 1.83014661e-01 -8.23821425e-01 4.04388189e-01
-6.55212760e-01 -2.02840805e-01 2.37209544e-01 7.43733883e-01
-9.76116776e-01 3.05612952e-01 -8.58141005e-01 -9.48141277e-01
5.23068190e-01 5.65685391e-01 -8.00621927e-01 -9.62358341e-02
-4.41351086e-01 1.08664083e+00 -1.63540840e-01 -1.81965753e-01
-7.57550597e-01 -9.28832710e-01 -6.90140545e-01 -3.57274003e-02
-2.46759847e-01 -9.38355267e-01 6.78880274e-01 -1.65325835e-01
-1.81985331e+00 1.13171232e+00 -2.04499543e-01 -3.18928808e-01
3.15529108e-01 4.66915756e-01 -4.16126639e-01 2.46115267e-01
-2.88918037e-02 3.32097739e-01 -1.21032886e-01 -1.31010807e+00
-2.07359150e-01 -5.33611298e-01 -4.39196467e-01 1.64146036e-01
2.80566305e-01 3.53439778e-01 -7.48868361e-02 -3.27986866e-01
1.38779119e-01 -9.94690299e-01 -3.85071427e-01 2.12408796e-01
-7.07046315e-02 3.03069353e-01 1.23564863e+00 -3.08369815e-01
8.91501427e-01 -1.87070346e+00 -2.32277513e-01 1.74737629e-02
3.37839067e-01 4.90848482e-01 1.09258287e-01 1.03398228e+00
1.76325098e-01 5.47138751e-01 3.13226014e-01 4.73295718e-01
-3.59040618e-01 -3.20081264e-01 4.14550871e-01 8.53528917e-01
-2.55667359e-01 9.70831752e-01 -4.47983235e-01 -7.52485454e-01
3.90567958e-01 8.69113505e-01 2.30376869e-02 -3.15945536e-01
1.08296126e-01 9.36586022e-01 -3.70159656e-01 8.78477097e-01
4.33369011e-01 -1.42799675e-01 1.39819950e-01 9.31866243e-02
-2.50287652e-01 -1.53839722e-01 -6.53058350e-01 1.05675590e+00
-2.50376612e-01 5.15856862e-01 4.71292973e-01 -4.71620917e-01
6.75524712e-01 7.97282338e-01 7.02278137e-01 -7.42200613e-01
2.46439889e-01 4.74446863e-01 2.44670838e-01 -3.08332175e-01
-2.70911187e-01 -6.29219651e-01 2.99225420e-01 1.52257740e-01
-3.57167423e-01 -2.06432119e-01 4.11845297e-01 5.64494878e-02
7.27515936e-01 -2.48254970e-01 4.60438728e-01 -5.37929177e-01
7.18620896e-01 2.60953810e-02 1.47255078e-01 -5.34899272e-02
-3.25606227e-01 7.62384236e-01 7.53213882e-01 -4.80029672e-01
-1.11642385e+00 -8.02338243e-01 -8.46639454e-01 1.72654882e-01
2.73348004e-01 -5.01205958e-02 -5.27984738e-01 2.31222603e-02
9.53891054e-02 1.60834119e-01 -6.85398042e-01 1.04814887e-01
-1.24275640e-01 -7.36635864e-01 4.76455927e-01 -7.24349394e-02
-3.58321778e-02 -4.75716680e-01 -6.74536169e-01 6.61667585e-01
6.10707521e-01 -1.05053794e+00 4.85487208e-02 -3.08159411e-01
-8.41789544e-01 -1.35915816e+00 -1.34333456e+00 -8.19737971e-01
8.58080924e-01 8.50575790e-02 7.12610066e-01 2.19815895e-01
-5.28438270e-01 5.42368054e-01 -1.53247550e-01 -9.26348448e-01
-7.31046736e-01 -6.00971282e-01 3.89391392e-01 -1.91455051e-01
1.14528045e-01 -5.53132534e-01 -8.84425402e-01 8.20245862e-01
-8.64450753e-01 -1.67604521e-01 4.90202159e-01 6.82902634e-01
1.06786287e+00 -4.20708895e-01 8.15456331e-01 -9.62491274e-01
7.04494059e-01 -3.76329720e-01 -3.88946205e-01 2.51577824e-01
-3.17772508e-01 -6.22395992e-01 6.71718895e-01 -4.35938507e-01
-6.47027493e-01 -2.50383824e-01 2.48506203e-01 -1.25981897e-01
-4.94697750e-01 6.36684775e-01 5.00423312e-02 -2.54933417e-01
9.05951738e-01 1.17535539e-01 7.22890019e-01 1.17873810e-01
-6.08400583e-01 6.14172995e-01 -1.73665360e-02 -3.04130971e-01
1.65907413e-01 8.65382552e-01 5.25003016e-01 -1.02817047e+00
8.13251734e-02 -6.03223503e-01 -4.08007711e-01 -1.70626685e-01
4.81042027e-01 -8.21815670e-01 -6.98844492e-01 4.04918402e-01
-7.62375891e-01 -8.46514106e-01 -3.58250767e-01 9.09235597e-01
-2.42830828e-01 8.96250606e-02 -6.39763415e-01 -7.74213076e-01
-1.43947423e-01 -1.34234929e+00 9.79191899e-01 4.06383187e-01
-4.72364545e-01 -1.31636202e+00 4.47933733e-01 2.59226970e-02
5.40483952e-01 1.10048699e+00 9.44918931e-01 -3.40707928e-01
-3.27713251e-01 -6.79740250e-01 2.63545275e-01 -5.48274875e-01
2.06864372e-01 3.23020607e-01 -7.87962735e-01 -1.28194436e-01
-6.59930706e-02 -9.37905908e-02 2.86173522e-01 9.81840134e-01
3.56012613e-01 -1.74975112e-01 -1.01094162e+00 3.82157385e-01
1.59558129e+00 6.93130493e-01 9.75710034e-01 3.23641092e-01
1.87483609e-01 6.48855865e-01 2.99291700e-01 2.22246021e-01
-2.01054916e-01 5.43438494e-01 8.82495046e-02 -5.01067877e-01
-1.11323096e-01 2.40007252e-01 1.00552991e-01 3.86679292e-01
-1.11051857e-01 -3.06625903e-01 -8.18685949e-01 5.73082328e-01
-9.58871245e-01 -5.91165364e-01 -5.22003055e-01 2.49732971e+00
6.48184776e-01 -2.19902903e-01 4.94104236e-01 -2.40058497e-01
6.11817658e-01 -9.17355418e-01 -4.69119281e-01 -6.06257379e-01
-3.85441065e-01 -9.62276533e-02 5.85375965e-01 3.68613362e-01
-3.32058340e-01 3.80274862e-01 7.28735113e+00 4.96036291e-01
-1.70099592e+00 -2.90790796e-01 8.13755155e-01 -2.36357793e-01
-1.68371454e-01 3.31781685e-01 -7.67922223e-01 5.30795455e-01
7.36082137e-01 -4.81878877e-01 -2.32081771e-01 1.20197572e-01
6.90457702e-01 -7.58224666e-01 -1.29641247e+00 6.66562438e-01
-3.35816860e-01 -1.70392120e+00 -1.84684455e-01 7.88656831e-01
5.39784610e-01 -2.34765470e-01 -9.84165445e-02 -1.85476959e-01
-1.55122966e-01 -1.31776726e+00 2.12348536e-01 2.96532482e-01
1.16599727e+00 -1.77754834e-01 9.39198375e-01 4.06739771e-01
-7.76216805e-01 2.89069831e-01 -3.66156995e-01 6.77576568e-03
3.08726877e-01 4.72662032e-01 -1.36865211e+00 7.78683424e-02
3.91764015e-01 1.30634019e-02 -4.29021418e-01 1.41719592e+00
6.68801427e-01 4.04215425e-01 -6.46590173e-01 -2.28558749e-01
-1.40945867e-01 -4.17927653e-01 4.38678294e-01 8.42817128e-01
2.84559399e-01 1.82502970e-01 -2.16039330e-01 5.72486877e-01
6.52945995e-01 5.56049228e-01 -8.15807045e-01 -2.20033661e-01
4.34236825e-01 1.19056952e+00 -1.30156302e+00 -1.86023429e-01
-5.03047347e-01 3.64844441e-01 -3.82025599e-01 5.45372725e-01
-2.15077013e-01 5.62494844e-02 3.23913664e-01 8.98587883e-01
-7.88752884e-02 -8.60136598e-02 -3.37156713e-01 -7.62622893e-01
-4.46257770e-01 -6.71764433e-01 2.31732368e-01 -4.68728960e-01
-6.17930233e-01 2.36532211e-01 9.15306434e-02 -1.28120673e+00
1.65057793e-01 -7.52399623e-01 -9.66887534e-01 6.67696953e-01
-6.01021469e-01 -1.23522890e+00 -3.07227168e-02 -1.95711572e-03
2.34453991e-01 -5.92496134e-02 1.12129307e+00 -6.67531341e-02
-6.73069656e-01 4.20304500e-02 5.33068895e-01 -4.82199162e-01
6.17380381e-01 -9.26341653e-01 -4.15027708e-01 2.91014612e-01
-3.22362661e-01 4.90065128e-01 8.23598564e-01 -7.18585074e-01
-1.28650963e+00 -6.01305366e-01 5.24500370e-01 -4.78329062e-01
2.26087645e-01 -3.43387991e-01 -8.16120982e-01 3.04521739e-01
1.00747168e-01 1.64581984e-01 1.24952877e+00 -1.70510590e-01
3.58466536e-01 2.36478925e-01 -1.33350158e+00 9.54141021e-01
1.64648056e-01 8.79811570e-02 3.64095777e-01 5.61885476e-01
-1.96414962e-02 -7.38187551e-01 -1.26244962e+00 1.19002536e-01
1.04256332e+00 -8.63875210e-01 6.55593455e-01 -4.87838298e-01
4.92613837e-02 -5.09295881e-01 5.21705329e-01 -1.01499403e+00
-2.41089210e-01 -5.09828091e-01 4.08543438e-01 9.04819608e-01
6.22661531e-01 -9.91030157e-01 8.89849663e-01 9.34715509e-01
-8.41212794e-02 -1.25601852e+00 -1.00975883e+00 -9.41010177e-01
5.36291182e-01 5.43801069e-01 3.31680506e-01 7.19572365e-01
6.45506442e-01 2.48450249e-01 6.17609382e-01 -5.21305203e-01
3.98745202e-02 -2.83380866e-01 1.15415001e+00 -1.43864787e+00
3.21406312e-02 -3.37603927e-01 -8.75003636e-01 -1.38030663e-01
-4.04396564e-01 -4.66414303e-01 -6.08850062e-01 -1.74898016e+00
-4.71050814e-02 -4.03175771e-01 3.59398693e-01 -1.69820458e-01
3.05632383e-01 2.21979730e-02 -1.19622432e-01 4.33031827e-01
6.76060393e-02 -3.72012332e-02 1.73479414e+00 7.34406784e-02
-5.39144933e-01 -3.86113346e-01 -5.98775804e-01 4.36056346e-01
7.77205050e-01 -3.68235439e-01 -5.42520106e-01 2.58442819e-01
-1.52050601e-02 1.52539760e-01 4.42820072e-01 -9.92687821e-01
2.30252862e-01 -3.27888310e-01 5.77744365e-01 -2.11260721e-01
1.31545618e-01 -8.61335158e-01 1.00687945e+00 7.77153969e-01
2.01066226e-01 -1.62165463e-01 2.24156305e-01 5.01053572e-01
-2.38215283e-01 -2.02233326e-02 1.07018888e+00 -5.58134854e-01
8.49072561e-02 -2.47742031e-02 -1.08359289e+00 -4.11802799e-01
1.80071425e+00 -1.40132475e+00 -5.21018982e-01 -1.54825643e-01
-1.04061103e+00 7.83191621e-02 1.42719328e+00 -5.28393984e-01
3.34895074e-01 -1.13144028e+00 -7.19326556e-01 -1.19719870e-01
-3.09950504e-02 3.45737219e-01 5.93425274e-01 1.68253326e+00
-1.37582600e+00 8.44864309e-01 -3.84265035e-01 -9.36159730e-01
-1.27720547e+00 5.57134211e-01 6.83556437e-01 -5.62735140e-01
-5.69363713e-01 4.98544872e-01 9.00204360e-01 6.06570020e-03
-5.08990884e-01 -7.62882084e-02 -2.34650835e-01 -3.41149449e-01
3.21727872e-01 1.33526251e-01 1.10830121e-01 -7.91279376e-01
-3.17545712e-01 6.31523848e-01 -1.72944084e-01 7.18620664e-04
9.47539806e-01 -2.93515205e-01 6.00332990e-02 6.27037466e-01
1.01024354e+00 -6.70703361e-03 -9.27331686e-01 4.48195100e-01
-5.39976597e-01 -3.27232003e-01 -1.49578869e-01 -6.73785269e-01
-6.37603641e-01 5.19081116e-01 3.76365691e-01 3.48916680e-01
8.71003568e-01 1.89354375e-01 5.51626801e-01 -1.73056483e-01
2.48126850e-01 -1.11654043e+00 -4.21528906e-01 5.01089469e-02
1.27587008e+00 -7.08947897e-01 3.14449906e-01 -9.11357582e-01
-4.24409121e-01 1.00877059e+00 4.42895055e-01 -3.97164710e-02
7.76313901e-01 6.01235867e-01 4.55281436e-01 -5.13121426e-01
-9.80183601e-01 1.89130351e-01 -9.64113325e-02 1.00995886e+00
8.59066725e-01 -2.21009597e-01 -9.58648860e-01 4.64803785e-01
2.60227453e-02 2.34496206e-01 9.26613688e-01 1.00220716e+00
-2.87520766e-01 -1.11512792e+00 -7.51067221e-01 3.87890249e-01
-8.67466092e-01 4.51481193e-01 -7.46282041e-01 1.06372893e+00
-2.51147866e-01 4.40762162e-01 -1.40772434e-02 1.79247722e-01
2.54895478e-01 -4.60114749e-03 5.48571229e-01 -6.06967747e-01
-5.43514132e-01 6.52147532e-01 -3.49944942e-02 -8.77582356e-02
-4.40772027e-01 -7.89499164e-01 -1.37824237e+00 1.99784599e-02
-5.51810741e-01 4.90260199e-02 7.93338120e-01 8.85844231e-01
6.10685289e-01 1.18198261e-01 2.72512257e-01 -8.51213992e-01
4.34449196e-01 -7.46729732e-01 -1.02778900e+00 -2.36916766e-02
1.27912432e-01 -5.32275915e-01 -4.60836887e-01 7.05723107e-01] | [13.860261917114258, -3.088029623031616] |
a1cf74f8-11f6-4ca2-8d0b-54c7c3450c1b | open-world-semi-supervised-learning-1 | 2102.03526 | null | https://arxiv.org/abs/2102.03526v3 | https://arxiv.org/pdf/2102.03526v3.pdf | Open-World Semi-Supervised Learning | A fundamental limitation of applying semi-supervised learning in real-world settings is the assumption that unlabeled test data contains only classes previously encountered in the labeled training data. However, this assumption rarely holds for data in-the-wild, where instances belonging to novel classes may appear at testing time. Here, we introduce a novel open-world semi-supervised learning setting that formalizes the notion that novel classes may appear in the unlabeled test data. In this novel setting, the goal is to solve the class distribution mismatch between labeled and unlabeled data, where at the test time every input instance either needs to be classified into one of the existing classes or a new unseen class needs to be initialized. To tackle this challenging problem, we propose ORCA, an end-to-end deep learning approach that introduces uncertainty adaptive margin mechanism to circumvent the bias towards seen classes caused by learning discriminative features for seen classes faster than for the novel classes. In this way, ORCA reduces the gap between intra-class variance of seen with respect to novel classes. Experiments on image classification datasets and a single-cell annotation dataset demonstrate that ORCA consistently outperforms alternative baselines, achieving 25% improvement on seen and 96% improvement on novel classes of the ImageNet dataset. | ['Jure Leskovec', 'Maria Brbic', 'Kaidi Cao'] | 2021-02-06 | open-world-semi-supervised-learning | https://openreview.net/forum?id=6VhmvP7XZue | https://openreview.net/pdf?id=6VhmvP7XZue | iclr-2022-4 | ['open-world-semi-supervised-learning'] | ['computer-vision'] | [ 6.29183769e-01 4.09501046e-01 -2.99888998e-01 -6.41630888e-01
-8.73536110e-01 -7.05383480e-01 3.52518022e-01 4.22582120e-01
-6.29527152e-01 9.53064919e-01 -2.94658780e-01 -2.21883491e-01
1.14672624e-01 -6.43901706e-01 -8.91414225e-01 -1.03489411e+00
9.21573043e-02 9.25905287e-01 2.19617262e-01 3.87315065e-01
-1.22188151e-01 3.63486141e-01 -1.51579154e+00 4.78306741e-01
6.28567219e-01 1.09690177e+00 -3.90595682e-02 4.09763306e-01
5.16530946e-02 4.88288969e-01 -6.53287828e-01 -2.63482798e-02
5.01421630e-01 -5.38706064e-01 -7.16713488e-01 5.62022030e-01
8.79463673e-01 -1.09855719e-01 -5.58136553e-02 9.75139260e-01
5.33777118e-01 4.81848139e-03 7.25137293e-01 -1.28611946e+00
-4.28052247e-01 5.31452656e-01 -5.91841936e-01 3.15767407e-01
-1.15830377e-01 2.35480487e-01 8.61567080e-01 -9.28898275e-01
6.76687002e-01 6.77221775e-01 4.52886790e-01 6.74149811e-01
-1.48840272e+00 -5.96449792e-01 4.02189642e-01 2.17798725e-01
-1.35233283e+00 -2.36221328e-01 4.71699744e-01 -3.83845180e-01
5.43041348e-01 7.04504400e-02 2.90541083e-01 1.13487017e+00
-3.10298279e-02 9.47840273e-01 1.14727879e+00 -4.30834830e-01
6.09151304e-01 3.95523459e-01 4.12641972e-01 5.08467197e-01
2.75296509e-01 1.08571358e-01 -3.72356892e-01 -3.02503314e-02
2.13977516e-01 1.04370065e-01 -2.71425843e-01 -7.03119755e-01
-1.10321975e+00 7.51809895e-01 3.86597216e-01 1.83559462e-01
-5.91646135e-03 -3.11082840e-01 4.67039049e-01 3.12418580e-01
5.92263579e-01 2.25002065e-01 -9.46817577e-01 1.02213643e-01
-9.44100797e-01 1.10962251e-02 5.46682537e-01 7.99714625e-01
8.62038136e-01 -1.94261104e-01 -2.74514675e-01 7.43654907e-01
1.04357302e-01 4.19971794e-01 8.25172842e-01 -4.64535147e-01
3.79446775e-01 8.45980942e-01 -2.01609552e-01 -1.66958138e-01
-3.02045673e-01 -9.39724386e-01 -6.35688841e-01 2.51255423e-01
7.19844282e-01 -1.07700549e-01 -1.30511689e+00 1.76441813e+00
5.68419397e-01 4.89786595e-01 2.97947079e-01 7.59226441e-01
8.07582378e-01 3.77191693e-01 -3.15506250e-01 -4.04984117e-01
8.60815406e-01 -1.11511457e+00 -4.49900895e-01 -3.04054886e-01
8.56953144e-01 -4.46034044e-01 1.13447809e+00 4.66707766e-01
-7.08001018e-01 -5.35726726e-01 -1.10798812e+00 2.40451306e-01
-5.06812692e-01 -1.66809205e-02 2.19646111e-01 5.44547975e-01
-6.62743390e-01 4.06450927e-01 -6.27435446e-01 -2.45390683e-01
7.80554652e-01 3.94694746e-01 -4.88259107e-01 -3.27107042e-01
-6.18840814e-01 3.73410612e-01 6.42540753e-01 3.44142318e-02
-1.13402998e+00 -6.87626004e-01 -9.17521000e-01 1.00417465e-01
7.61221588e-01 -1.57532007e-01 1.20900023e+00 -1.46856523e+00
-8.85194242e-01 1.12939239e+00 -1.48575410e-01 -4.32516545e-01
6.50282562e-01 2.32956652e-02 -3.97150964e-01 -1.20786794e-01
1.36407822e-01 5.78743756e-01 7.37934768e-01 -1.37026453e+00
-7.49202430e-01 -5.87762594e-01 -1.26677275e-01 4.73306030e-02
-2.78634667e-01 -7.11875260e-01 -3.61710817e-01 -5.56622744e-01
2.45938480e-01 -1.09495568e+00 -1.23290718e-01 -1.46516766e-02
-6.43061757e-01 -2.29007453e-01 1.23144650e+00 5.13727330e-02
7.66696155e-01 -2.42235351e+00 -2.08526433e-01 3.29232216e-01
2.33100981e-01 4.10903275e-01 -2.13695869e-01 5.03202062e-03
-3.97052228e-01 -1.60585284e-01 -4.59798843e-01 -2.99855471e-01
-2.12754488e-01 3.99724692e-01 -4.47309673e-01 5.11749446e-01
2.80091286e-01 7.41236091e-01 -9.33737934e-01 -3.16379607e-01
-2.77898703e-02 1.59227490e-01 -3.97008330e-01 1.02167986e-01
-2.86435992e-01 6.67715788e-01 -1.58303827e-01 6.72363818e-01
7.72480607e-01 -4.71644044e-01 7.98821971e-02 1.22402810e-01
3.58200729e-01 1.99866332e-02 -1.17388403e+00 1.50630689e+00
-2.80566543e-01 5.77295065e-01 -4.73976433e-01 -1.23267162e+00
6.45273626e-01 1.89120993e-01 2.60546535e-01 -6.44423306e-01
7.62842819e-02 2.91758031e-01 3.01800609e-01 -2.70139903e-01
-1.48314729e-01 -1.78569272e-01 1.61139280e-01 4.18114692e-01
1.57757834e-01 1.64354444e-01 2.79784203e-01 3.20015885e-02
1.10618520e+00 -3.08852699e-02 1.51580632e-01 -1.87042609e-01
3.60433757e-01 -2.57672463e-02 9.30340528e-01 8.52264524e-01
-2.82931417e-01 8.84256780e-01 5.15009403e-01 -6.13109708e-01
-8.14810812e-01 -1.31660771e+00 -5.68751097e-01 9.22206283e-01
7.99824670e-03 1.38074890e-01 -6.36948407e-01 -1.43514872e+00
7.36075118e-02 5.18535018e-01 -9.26718056e-01 -2.73328990e-01
-2.60489613e-01 -7.65952408e-01 1.60333604e-01 6.00380480e-01
3.09853911e-01 -9.72244978e-01 -5.68739474e-01 9.43930894e-02
8.66554901e-02 -1.15346861e+00 -4.30183560e-01 7.52982318e-01
-9.00055647e-01 -1.34043467e+00 -5.42945266e-01 -1.00451899e+00
1.24448490e+00 6.34468794e-02 1.00802565e+00 3.27644404e-03
-5.26063740e-01 1.36732176e-01 -4.14105564e-01 -4.42103297e-01
-2.43795410e-01 2.62723677e-02 8.22181068e-03 3.71690392e-01
4.77221191e-01 -3.07589799e-01 -5.24781585e-01 4.82534736e-01
-1.02903461e+00 -1.45680323e-01 5.22987962e-01 1.25915897e+00
9.85972106e-01 2.11059287e-01 7.93966770e-01 -1.36157644e+00
-2.46721521e-01 -5.61283052e-01 -4.48267162e-01 2.56812364e-01
-6.87396109e-01 7.62524977e-02 7.31521666e-01 -7.16564476e-01
-6.92696512e-01 3.02803695e-01 1.09506056e-01 -3.13654184e-01
-3.50532740e-01 6.02164567e-01 -3.50084096e-01 2.01296076e-01
7.22274840e-01 2.65458077e-01 -9.35751498e-02 -2.89325356e-01
1.26761710e-02 6.56082034e-01 4.87353146e-01 -4.17721570e-01
8.39389503e-01 6.71697497e-01 -1.64889693e-01 -5.41210473e-01
-1.28399193e+00 -5.39610684e-01 -8.89140725e-01 2.32620537e-03
5.59698999e-01 -1.00345194e+00 -2.72218108e-01 4.35350031e-01
-5.49163997e-01 -7.72684872e-01 -8.90324056e-01 4.04603779e-01
-3.08358639e-01 2.62907892e-01 -3.18326592e-01 -4.31239009e-01
-3.22868563e-02 -1.37234485e+00 1.18324769e+00 3.37654293e-01
-4.90108393e-02 -1.02580488e+00 -3.16363759e-02 3.65412325e-01
2.13662963e-02 3.05340469e-01 9.59901869e-01 -1.30116391e+00
-3.04989398e-01 -5.26601136e-01 -6.30867705e-02 5.50695539e-01
2.34220862e-01 -2.79739410e-01 -1.22867918e+00 -7.73723722e-01
-1.43285856e-01 -8.16984653e-01 1.02605569e+00 1.45116877e-02
1.38469946e+00 -4.35514376e-02 -4.16219026e-01 4.51903552e-01
1.28637016e+00 2.64414996e-01 3.38116765e-01 5.25814705e-02
5.41411400e-01 5.29979706e-01 6.89594150e-01 4.17898655e-01
-3.70807387e-03 5.53358555e-01 4.01450515e-01 -3.26420039e-01
2.01409571e-02 -1.11718729e-01 1.44744292e-01 2.26079747e-01
7.20453799e-01 -5.30246556e-01 -8.94413292e-01 6.18407547e-01
-1.79010153e+00 -5.56477249e-01 -7.95953572e-02 2.59771228e+00
9.00642812e-01 4.45312917e-01 -1.56669617e-01 3.88910234e-01
6.85457826e-01 -2.80411750e-01 -1.01379669e+00 3.79228368e-02
-3.67577583e-01 3.35690558e-01 3.60610455e-01 2.82324255e-01
-1.28650582e+00 6.08477890e-01 5.36450768e+00 7.77350426e-01
-1.32450485e+00 6.02263920e-02 1.15037310e+00 -2.27920592e-01
3.79380807e-02 -9.10206959e-02 -9.51535404e-01 2.81206965e-01
7.48376369e-01 1.08237527e-01 -7.37020448e-02 7.86666453e-01
-2.44121635e-04 -2.93614179e-01 -1.65511668e+00 8.44662845e-01
2.78889447e-01 -9.73740041e-01 7.96673074e-03 1.22751236e-01
1.00227678e+00 2.27621272e-01 3.17208290e-01 4.58568841e-01
2.20585108e-01 -9.15434659e-01 6.17150545e-01 -2.57852655e-02
8.70936036e-01 -7.28183687e-01 9.17873800e-01 5.60886562e-01
-6.93780839e-01 -1.36864051e-01 -2.40570396e-01 2.30334885e-02
-3.48510116e-01 8.58856618e-01 -1.14390230e+00 1.66361570e-01
5.31601191e-01 7.00792074e-01 -6.92711234e-01 1.19924355e+00
-2.74045974e-01 8.43030572e-01 -2.61922777e-01 4.49223995e-01
3.62423271e-01 2.07238644e-01 2.01089978e-01 9.63544965e-01
1.15031570e-01 -2.84222603e-01 5.49337864e-01 5.92427015e-01
-2.79977143e-01 4.99459207e-02 -5.90033352e-01 1.17379457e-01
1.27872571e-01 1.11494172e+00 -1.05785692e+00 -4.40641731e-01
-4.60019201e-01 1.10175538e+00 3.34714949e-01 5.02549827e-01
-8.19930851e-01 -1.65116787e-01 1.34998858e-01 2.05041170e-01
4.63032514e-01 3.49214286e-01 -2.06305400e-01 -1.14448917e+00
2.37948641e-01 -8.52905750e-01 7.99652696e-01 -3.60134840e-01
-1.47225356e+00 5.03411174e-01 -2.61036426e-01 -1.45444703e+00
-1.09544776e-01 -7.50132501e-01 -5.02896786e-01 6.35311484e-01
-1.47336936e+00 -1.02995431e+00 -2.47219309e-01 4.64179844e-01
8.31413746e-01 -1.80524096e-01 7.99607038e-01 2.85853088e-01
-5.37035704e-01 9.51472700e-01 5.14111102e-01 2.65690744e-01
1.00696015e+00 -1.36505556e+00 7.03842714e-02 7.95102000e-01
4.83709335e-01 1.73608974e-01 4.82511252e-01 -4.29878324e-01
-9.22576010e-01 -1.39336550e+00 8.30877066e-01 -4.34597105e-01
3.53811234e-01 -6.08118653e-01 -1.05551553e+00 9.37170029e-01
-1.40214682e-01 8.73848677e-01 1.09033108e+00 1.00486964e-01
-6.42360866e-01 -1.56238750e-01 -1.21714842e+00 2.79731840e-01
7.64456809e-01 -5.33901513e-01 -2.72588372e-01 6.08173966e-01
4.92924899e-01 -4.50407237e-01 -4.80996609e-01 7.06693053e-01
2.46063411e-01 -6.76359892e-01 5.06246328e-01 -8.06433201e-01
3.25797141e-01 -5.06739676e-01 -1.52435571e-01 -1.38215065e+00
1.02260448e-01 -2.28884965e-01 -6.47850484e-02 1.01146245e+00
8.14548016e-01 -5.96931219e-01 1.10729218e+00 3.59090239e-01
-1.84559003e-01 -9.59219873e-01 -1.09633982e+00 -8.83459210e-01
2.37402111e-01 -2.93258190e-01 1.83154285e-01 1.21689069e+00
-1.19522795e-01 3.45092565e-01 5.03248684e-02 1.81468755e-01
5.21163523e-01 2.22239405e-01 7.48230457e-01 -1.14488769e+00
-5.97972393e-01 -8.10892582e-02 -7.05354631e-01 -8.78332496e-01
2.25651473e-01 -1.05531907e+00 2.61492461e-01 -1.07225549e+00
6.08286917e-01 -7.11159527e-01 -6.32674277e-01 7.09589601e-01
-4.07975107e-01 5.54431021e-01 -7.41678998e-02 5.07982075e-03
-8.58084500e-01 3.44215542e-01 1.06691837e+00 -3.48348528e-01
-1.59211934e-01 1.24507524e-01 -5.12637794e-01 7.35975802e-01
5.40432632e-01 -5.66002905e-01 -6.23304069e-01 -4.11828399e-01
-1.18585818e-01 -2.45970085e-01 2.70008504e-01 -1.05368710e+00
1.32773146e-01 1.40571386e-01 5.58094501e-01 -7.40808249e-01
-6.31546602e-02 -9.96816456e-01 -1.47381514e-01 4.32508260e-01
-6.17992043e-01 -3.66502762e-01 2.96652228e-01 7.41328716e-01
-1.75199494e-01 -3.17798376e-01 9.86177623e-01 3.09361190e-01
-6.62000120e-01 4.29054767e-01 -1.31580397e-01 4.11080241e-01
1.39149010e+00 -3.62915248e-01 -4.25364792e-01 3.97365578e-02
-1.14143348e+00 3.46521795e-01 4.25361574e-01 3.00818384e-01
5.20289123e-01 -1.10499799e+00 -7.11249113e-01 3.70528311e-01
7.07741082e-01 4.87927556e-01 2.20939860e-01 7.05202818e-01
-2.49861062e-01 6.36477172e-02 2.06234977e-01 -9.26073670e-01
-1.17167985e+00 7.02016294e-01 3.28497708e-01 -4.13611382e-01
-3.81349355e-01 1.07697988e+00 6.67391598e-01 -6.91987038e-01
4.99856800e-01 -2.47818843e-01 4.41042967e-02 -5.57529479e-02
5.32097459e-01 -1.67379677e-02 3.71708959e-01 -5.60961902e-01
-1.66402444e-01 5.94408438e-02 -6.29164159e-01 2.19992295e-01
1.13070297e+00 7.58547336e-02 2.60660857e-01 1.00047600e+00
1.42658901e+00 -1.30858317e-01 -1.47662389e+00 -6.15038753e-01
1.33623153e-01 -2.83755988e-01 -1.20695554e-01 -1.20887375e+00
-1.22868919e+00 8.89026046e-01 1.12082505e+00 -2.00618953e-01
1.10149372e+00 2.18127519e-01 6.06368363e-01 5.35528183e-01
3.16578180e-01 -1.01743186e+00 1.55713633e-01 4.35068905e-01
3.86286944e-01 -1.63844061e+00 -2.01303571e-01 -4.60107833e-01
-5.01881063e-01 9.82265055e-01 9.56073284e-01 1.29408017e-01
6.18473589e-01 2.56327301e-01 3.23171496e-01 -4.24116151e-03
-8.39915872e-01 -1.45672277e-01 1.90231934e-01 4.86051500e-01
3.61307412e-01 -2.24285331e-02 1.34421829e-02 2.62619466e-01
1.27931356e-01 -1.95153043e-01 4.42343205e-01 1.15850127e+00
-2.53311217e-01 -1.05017495e+00 -1.28200024e-01 6.49975717e-01
-4.42075878e-01 -1.36802465e-01 -4.09212798e-01 7.21603453e-01
5.54047942e-01 7.20679760e-01 3.75009894e-01 -1.51935488e-01
2.74449408e-01 2.01659545e-01 3.67098957e-01 -1.11716413e+00
-3.30734074e-01 1.69301286e-01 -2.71058470e-01 -2.42846891e-01
-2.82840848e-01 -6.53469145e-01 -1.58113194e+00 4.02559340e-01
-5.85221469e-01 1.82464018e-01 2.57323891e-01 1.16746676e+00
1.68084800e-01 5.55012941e-01 7.93078184e-01 -3.81995112e-01
-5.76700032e-01 -7.49437749e-01 -6.90049112e-01 6.46203756e-01
5.03214896e-01 -5.91278076e-01 -4.74248320e-01 3.01421970e-01] | [9.541582107543945, 3.356410026550293] |
935135f4-5ed7-467d-9990-1140bd63990f | symbolic-regression-via-control-variable | 2306.08057 | null | https://arxiv.org/abs/2306.08057v1 | https://arxiv.org/pdf/2306.08057v1.pdf | Symbolic Regression via Control Variable Genetic Programming | Learning symbolic expressions directly from experiment data is a vital step in AI-driven scientific discovery. Nevertheless, state-of-the-art approaches are limited to learning simple expressions. Regressing expressions involving many independent variables still remain out of reach. Motivated by the control variable experiments widely utilized in science, we propose Control Variable Genetic Programming (CVGP) for symbolic regression over many independent variables. CVGP expedites symbolic expression discovery via customized experiment design, rather than learning from a fixed dataset collected a priori. CVGP starts by fitting simple expressions involving a small set of independent variables using genetic programming, under controlled experiments where other variables are held as constants. It then extends expressions learned in previous generations by adding new independent variables, using new control variable experiments in which these variables are allowed to vary. Theoretically, we show CVGP as an incremental building approach can yield an exponential reduction in the search space when learning a class of expressions. Experimentally, CVGP outperforms several baselines in learning symbolic expressions involving multiple independent variables. | ['Yexiang Xue', 'Nan Jiang'] | 2023-05-25 | null | null | null | null | ['symbolic-regression'] | ['knowledge-base'] | [ 4.10300374e-01 1.09470934e-01 -5.19258022e-01 -5.06565332e-01
-7.10294545e-01 -7.07694471e-01 2.00816333e-01 2.71622449e-01
-3.24680686e-01 1.34682703e+00 -5.89774191e-01 -4.58403975e-01
-1.58565804e-01 -8.88730943e-01 -1.14923429e+00 -6.24548972e-01
-5.25706112e-01 9.45602596e-01 -3.20934914e-02 -2.44845331e-01
1.22570902e-01 6.58643365e-01 -1.58268821e+00 -2.19823033e-01
6.45373762e-01 6.47745013e-01 -3.85557353e-01 6.51371539e-01
-2.74811417e-01 6.86948597e-01 -7.38786995e-01 -2.16211811e-01
9.95728448e-02 -5.42827308e-01 -7.43214130e-01 -4.74600792e-01
-2.64954772e-02 1.59467965e-01 4.16690707e-01 9.11465347e-01
2.15927035e-01 2.84910262e-01 4.68605429e-01 -1.49237335e+00
-5.29323459e-01 1.10053885e+00 -6.54890716e-01 -2.78937608e-01
2.82109261e-01 2.68720120e-01 1.16903210e+00 -3.55910122e-01
7.64988482e-01 1.56006026e+00 6.25503778e-01 6.01465762e-01
-1.94062269e+00 -1.05023766e+00 3.16495985e-01 -1.41056761e-01
-1.49809253e+00 -8.77900720e-02 6.47671103e-01 -4.19218242e-01
1.04488969e+00 3.74525636e-01 6.98049188e-01 1.01983547e+00
-3.67346928e-02 5.31089008e-01 9.30921078e-01 -6.99429274e-01
5.57475924e-01 1.96807936e-01 3.14556509e-01 9.02327240e-01
2.95920193e-01 4.19551015e-01 -4.00817722e-01 -4.23318803e-01
6.34564936e-01 -4.14634705e-01 -1.56306475e-02 -4.48103040e-01
-9.51512754e-01 1.05413222e+00 1.28531680e-01 1.69889748e-01
-1.44152537e-01 6.68481708e-01 4.41610813e-01 6.51438117e-01
1.41947418e-01 1.26109958e+00 -1.22333026e+00 -2.54486352e-01
-7.35643208e-01 7.57644296e-01 8.96300375e-01 1.05629015e+00
1.02199817e+00 1.83201000e-01 -1.88166276e-02 5.61173260e-01
-2.12364029e-02 4.55345571e-01 2.38388211e-01 -8.37929666e-01
5.26545346e-02 6.88914955e-01 -1.01557091e-01 -7.84847438e-01
-4.36312824e-01 -1.35954857e-01 -4.05346960e-01 4.18246776e-01
6.76616073e-01 -6.44505978e-01 -1.04102933e+00 2.06750226e+00
3.03911567e-01 1.96035981e-01 -8.62752367e-03 4.65013415e-01
4.49253619e-01 6.48963153e-01 2.65139014e-01 -4.50262725e-01
8.52085829e-01 -5.82827508e-01 -4.16802019e-01 -1.15369678e-01
1.05353868e+00 -5.09192981e-02 9.48577881e-01 7.84343541e-01
-9.19996619e-01 -3.10756505e-01 -9.22152162e-01 1.64595053e-01
-6.15571737e-01 -3.08886647e-01 1.41858387e+00 8.65425587e-01
-7.06389189e-01 7.12672949e-01 -8.41181397e-01 2.17253327e-01
4.51320618e-01 1.01624882e+00 -3.29106331e-01 1.30654037e-01
-1.26137304e+00 6.31465137e-01 5.01743436e-01 -2.92515922e-02
-8.67488861e-01 -1.04576385e+00 -9.68972206e-01 -5.00265658e-02
8.94248784e-01 -5.77210486e-01 1.15517020e+00 -1.08374715e+00
-1.67007375e+00 5.94875336e-01 -3.71894389e-01 -2.68108338e-01
3.42434585e-01 -1.06780387e-01 -1.96558908e-01 -4.09384519e-01
-1.77072436e-01 5.81526339e-01 7.11016715e-01 -1.29650092e+00
-7.72673637e-02 -3.73408765e-01 4.02918831e-02 -5.30723453e-01
9.68326852e-02 2.67869025e-01 -3.59481633e-01 -3.32013518e-01
-2.37851426e-01 -1.20637155e+00 -6.50917470e-01 -2.34528203e-02
-4.25992161e-01 -1.70939013e-01 5.98820150e-01 -2.71653354e-01
1.24078953e+00 -1.68293285e+00 6.26328111e-01 8.06606054e-01
2.51904637e-01 9.19747129e-02 -6.63225725e-02 -2.37565935e-01
-4.87308979e-01 4.35083836e-01 -4.66295540e-01 2.08520159e-01
4.09235731e-02 2.21413940e-01 -1.30840272e-01 9.46712792e-02
6.10569835e-01 1.18900323e+00 -1.07597077e+00 -6.02740228e-01
-2.16069490e-01 9.99554396e-02 -8.63393009e-01 7.48361722e-02
-1.10711634e+00 4.93926913e-01 -8.81527305e-01 9.08416986e-01
3.41031015e-01 -2.97135919e-01 3.67530495e-01 4.05265808e-01
1.47864744e-01 -2.49099672e-01 -1.13542473e+00 1.55311131e+00
-2.41623446e-01 4.39114660e-01 -2.45965287e-01 -1.24728262e+00
1.29734373e+00 1.97501983e-02 4.80364710e-01 -2.43428662e-01
4.71644133e-01 2.09751442e-01 2.01347277e-01 -5.69068253e-01
-1.97007675e-02 -4.31832552e-01 -3.46883982e-01 1.75649464e-01
2.54641119e-02 -5.87560356e-01 4.45753962e-01 -1.85526371e-01
1.20256817e+00 5.37170827e-01 5.12325823e-01 1.45381615e-01
6.61278307e-01 2.49623597e-01 8.12960386e-01 7.31669307e-01
2.07803950e-01 1.66897625e-02 1.21911132e+00 -5.48303068e-01
-8.85767341e-01 -6.77199543e-01 -2.62162715e-01 1.17309535e+00
-3.38197380e-01 -4.91632015e-01 -4.69413787e-01 -6.16308331e-01
3.64057273e-01 9.56216872e-01 -9.32303905e-01 -1.46535277e-01
-7.43457794e-01 -9.32808638e-01 5.70429862e-01 5.10715783e-01
-1.90639477e-02 -9.07176971e-01 -4.30803508e-01 2.46998042e-01
4.02692139e-01 -5.64447641e-01 3.51245075e-01 8.90317738e-01
-8.36157143e-01 -8.50358605e-01 -4.54176724e-01 -5.26505053e-01
8.45102549e-01 -6.49412096e-01 1.25393164e+00 1.93708524e-01
-5.91278911e-01 -2.35163733e-01 -2.92257071e-02 -8.26493561e-01
-5.36763430e-01 9.75415111e-02 -1.34487182e-01 -6.11209750e-01
5.40108681e-01 -7.47615278e-01 4.35551822e-01 1.34112224e-01
-5.80690205e-01 -2.80875772e-01 3.72361124e-01 1.15172589e+00
8.62135708e-01 1.66827813e-01 4.36076194e-01 -1.26346803e+00
3.82460862e-01 -5.90233088e-01 -1.20622337e+00 5.53284943e-01
-5.72542608e-01 7.73781300e-01 7.37422526e-01 -8.59537065e-01
-6.22924507e-01 3.08423489e-01 2.49524191e-01 -6.03867769e-01
-1.50840163e-01 7.61009276e-01 -3.19876850e-01 -3.11518431e-01
8.56985331e-01 -2.29632691e-01 2.71822494e-02 -1.19913019e-01
2.76755542e-01 5.02601080e-02 2.20249996e-01 -1.37234998e+00
6.89097822e-01 -2.70176530e-01 7.51155257e-01 -4.98171300e-01
-4.13797796e-01 2.51294315e-01 -7.50866234e-01 6.83064982e-02
5.86607277e-01 -5.14934242e-01 -1.04300606e+00 7.58237317e-02
-9.05157983e-01 -6.24134064e-01 -2.02552751e-01 2.15144157e-01
-5.36067665e-01 -1.00478843e-01 5.82354739e-02 -8.03765953e-01
1.02566242e-01 -1.21813262e+00 9.40387130e-01 5.29684015e-02
-8.17785919e-01 -1.02189684e+00 1.56633958e-01 -1.05316751e-01
-6.07356895e-03 7.81831563e-01 1.31972802e+00 -8.48855019e-01
-4.53560144e-01 -4.44162160e-01 2.55023718e-01 -1.17101736e-01
-4.82797474e-02 4.75989342e-01 -6.58337235e-01 -1.82904452e-02
-5.16547263e-01 -7.57132530e-01 4.42798018e-01 3.19333434e-01
1.56509221e+00 -1.45214215e-01 -8.68430734e-01 7.58678555e-01
1.19545090e+00 4.89239484e-01 4.52177346e-01 3.59916091e-01
5.40662050e-01 4.30887669e-01 5.72524786e-01 3.37906778e-01
-1.66727990e-01 4.95749116e-01 1.21272132e-01 -8.89378786e-02
6.23712361e-01 -8.36313218e-02 7.89729506e-02 -2.05637529e-01
-2.49846697e-01 -1.27432212e-01 -1.20117188e+00 2.47165769e-01
-1.63701546e+00 -8.22177351e-01 -7.30325654e-02 2.14620328e+00
1.45211554e+00 2.98603475e-01 2.15384230e-01 1.70798376e-01
4.89331067e-01 -5.09683251e-01 -1.03227973e+00 -5.27685702e-01
-2.14548200e-01 9.15280044e-01 5.79536557e-01 5.66589773e-01
-8.72106075e-01 1.13487327e+00 6.95117950e+00 5.34324288e-01
-1.43653631e+00 -5.27217090e-01 7.89922416e-01 -3.78408998e-01
-4.79486465e-01 3.15054834e-01 -9.38260257e-01 2.42231056e-01
1.03142917e+00 -5.55084705e-01 5.83425641e-01 9.86234903e-01
-1.00775629e-01 1.92381709e-03 -1.50883436e+00 6.84380889e-01
-2.36477256e-01 -1.14775991e+00 -1.20454200e-01 -2.00985104e-01
8.42239738e-01 -5.59954047e-01 -7.25152418e-02 6.85675561e-01
9.28212643e-01 -1.50080192e+00 3.80212188e-01 4.33933944e-01
7.60150492e-01 -9.32291567e-01 4.95034128e-01 1.20983727e-01
-6.98019683e-01 -1.85684979e-01 -1.32318079e-01 -1.21558057e-02
-1.19688012e-01 4.40849483e-01 -8.86347532e-01 2.46526703e-01
3.72242928e-01 4.74217504e-01 -5.23239136e-01 5.61618209e-01
-4.71877515e-01 8.61066222e-01 -6.05714440e-01 -4.97839034e-01
1.16805792e-01 -1.37380004e-01 3.66689235e-01 1.01502991e+00
8.25840235e-02 2.52210528e-01 5.77205680e-02 1.50322092e+00
1.54148474e-01 4.74646986e-02 -5.25594711e-01 -3.91608804e-01
4.48890805e-01 8.17157149e-01 -5.44657826e-01 -2.94306576e-01
-8.10485631e-02 4.02635545e-01 4.49302107e-01 5.30856073e-01
-9.32981133e-01 -6.59015954e-01 5.88273287e-01 -1.54976979e-01
4.03631568e-01 -1.40225649e-01 -5.13836563e-01 -8.90026152e-01
-1.69395015e-01 -1.18804312e+00 2.23330110e-01 -5.33949137e-01
-6.92590117e-01 2.48345867e-01 4.52756673e-01 -5.70496798e-01
-7.34973788e-01 -7.51898408e-01 -4.08529311e-01 9.95202124e-01
-1.17382181e+00 -8.95738900e-01 -2.52789464e-02 3.39946538e-01
8.18062760e-03 -3.10556471e-01 1.23450899e+00 -2.16692492e-01
-8.72682393e-01 9.60187793e-01 -4.77548242e-02 -1.70451403e-02
4.44388390e-01 -1.09029996e+00 2.43872240e-01 4.14739639e-01
-1.79197639e-01 1.15302932e+00 9.29328978e-01 -7.59631097e-01
-1.63263178e+00 -7.42503524e-01 7.26638436e-01 -2.43028462e-01
8.97959113e-01 -4.95926052e-01 -1.06873643e+00 8.79605889e-01
-5.22861958e-01 1.29850656e-01 7.70070314e-01 5.60360670e-01
-4.89460081e-01 5.29743433e-02 -1.08945119e+00 6.15281641e-01
7.98981786e-01 -9.19754207e-02 -2.15907529e-01 1.24728620e-01
8.59968722e-01 -5.09409189e-01 -9.78402138e-01 3.25235516e-01
6.41281128e-01 -1.34171247e-01 9.68702018e-01 -1.32073045e+00
6.43512666e-01 -3.29820305e-01 -3.89704108e-02 -1.16148496e+00
-1.85397148e-01 -1.04452157e+00 -2.05439925e-01 1.04267001e+00
8.51241827e-01 -5.25420904e-01 9.66544569e-01 1.15883696e+00
3.69422615e-01 -8.49550366e-01 -5.11482298e-01 -9.37253296e-01
4.86840785e-01 -4.44961309e-01 9.22588766e-01 1.08736539e+00
1.37957230e-01 4.03632581e-01 -1.79688215e-01 -1.25566032e-02
4.25646693e-01 2.22078189e-01 1.15736985e+00 -1.47448397e+00
-7.43106186e-01 -5.21888852e-01 -6.08320177e-01 -5.57234406e-01
7.48369992e-01 -1.00928211e+00 2.39321757e-02 -6.96151793e-01
-1.26421871e-02 -7.10759342e-01 -8.12068135e-02 1.02289689e+00
-4.21892136e-01 -2.73517638e-01 -1.62332401e-01 -4.22384024e-01
-1.89361691e-01 3.63028198e-01 1.02867365e+00 -3.02304804e-01
-5.04547417e-01 -1.62153110e-01 -7.85164952e-01 5.18937111e-01
7.51998246e-01 -6.67964220e-01 -3.37857962e-01 4.95331585e-02
5.46287119e-01 1.86991140e-01 1.05677754e-01 -6.65966332e-01
-5.65955192e-02 -7.96943486e-01 5.10819912e-01 -2.03994989e-01
1.32229760e-01 -7.22588837e-01 4.66451645e-01 2.77024180e-01
-6.90396965e-01 -1.71918020e-01 7.10157335e-01 1.59072742e-01
1.72492806e-02 -4.30918962e-01 7.01128423e-01 -2.21834645e-01
-5.71491122e-01 3.44318524e-03 -1.35574698e-01 1.79699045e-02
1.28077936e+00 -1.66168481e-01 2.48318955e-01 8.63388181e-03
-9.70101178e-01 5.16269386e-01 2.13303670e-01 5.47009632e-02
3.20506424e-01 -9.01183605e-01 -5.51686883e-01 1.48717299e-01
8.71873498e-02 4.00539190e-01 -4.54315275e-01 3.59194160e-01
-3.07854384e-01 2.27215886e-01 -1.60320401e-01 -4.18238640e-01
-1.58181238e+00 9.32810664e-01 2.71549106e-01 -4.16989028e-01
-7.93093294e-02 1.29451835e+00 -2.07337379e-01 -6.41310394e-01
2.51115888e-01 -4.54180330e-01 1.05836123e-01 -1.20890945e-01
3.14401180e-01 2.57059366e-01 -2.61656553e-01 6.49849549e-02
-1.77848399e-01 6.72334254e-01 -1.94765702e-01 -3.38230468e-02
1.69447517e+00 6.24748111e-01 -2.21280307e-01 7.08734989e-01
1.22439754e+00 -7.53083080e-02 -8.83048832e-01 -1.93507001e-01
2.80260801e-01 -3.21073562e-01 3.58320326e-02 -1.04342473e+00
-9.68139708e-01 5.80706835e-01 2.79862825e-02 -3.21934730e-01
1.14971936e+00 -5.71932681e-02 1.07235722e-01 1.01580989e+00
4.25296396e-01 -7.16359615e-01 -1.04942575e-01 4.75627422e-01
7.65324950e-01 -1.24517000e+00 4.42065835e-01 -3.86118174e-01
-4.37316447e-01 1.33103037e+00 6.94857597e-01 4.39692214e-02
4.44101900e-01 8.06335449e-01 -2.44835421e-01 -3.01967561e-01
-8.79179358e-01 1.39725000e-01 2.43831053e-02 5.91781318e-01
6.40323460e-01 -8.17205757e-03 -1.33641809e-01 6.31169677e-01
-5.05719721e-01 4.63095456e-01 1.74637824e-01 1.20449722e+00
-1.16562657e-01 -1.52673030e+00 -4.67277139e-01 5.00801206e-01
-4.06068772e-01 -1.31593674e-01 -6.95041597e-01 1.15554011e+00
2.05598608e-01 4.63523954e-01 -1.53457865e-01 -1.36901438e-01
2.57750183e-01 2.86312699e-01 7.17075229e-01 -5.11561871e-01
-6.06236279e-01 -1.84012368e-01 1.24681950e-01 -3.26355129e-01
-4.66916375e-02 -9.92068350e-01 -1.61854708e+00 -2.19216704e-01
-5.18994510e-01 2.54571199e-01 6.74390614e-01 8.14645529e-01
8.33816752e-02 6.44735396e-01 4.62058038e-01 -5.90840995e-01
-4.80523586e-01 -4.92580861e-01 -3.47643167e-01 3.03421449e-02
4.06019926e-01 -8.79250228e-01 -2.82022297e-01 3.08623016e-01] | [8.51333236694336, 6.973209857940674] |
2f1df4f1-2cb8-44b9-8151-47038f7e363a | dias-a-domain-independent-alife-based-problem | 2203.06855 | null | https://arxiv.org/abs/2203.06855v2 | https://arxiv.org/pdf/2203.06855v2.pdf | DIAS: A Domain-Independent Alife-Based Problem-Solving System | A domain-independent problem-solving system based on principles of Artificial Life is introduced. In this system, DIAS, the input and output dimensions of the domain are laid out in a spatial medium. A population of actors, each seeing only part of this medium, solves problems collectively in it. The process is independent of the domain and can be implemented through different kinds of actors. Through a set of experiments on various problem domains, DIAS is shown able to solve problems with different dimensionality and complexity, to require no hyperparameter tuning for new problems, and to exhibit lifelong learning, i.e. adapt rapidly to run-time changes in the problem domain, and do it better than a standard non-collective approach. DIAS therefore demonstrates a role for Alife in building scalable, general, and adaptive problem-solving systems. | ['Risto Miikkulainen', 'Hormoz Shahrzad', 'Babak Hodjat'] | 2022-03-14 | null | null | null | null | ['artificial-life'] | ['miscellaneous'] | [-3.29347670e-01 3.41397703e-01 2.98771650e-01 2.24890172e-01
-2.89978832e-01 -8.12896848e-01 6.52629077e-01 8.76437202e-02
-2.33375296e-01 9.65628564e-01 1.15674123e-01 2.93931216e-02
-8.35596263e-01 -1.06680846e+00 -2.48624712e-01 -8.49254191e-01
-2.77614623e-01 1.18532908e+00 1.53771684e-01 -3.18087935e-01
4.01095092e-01 5.12352824e-01 -1.63315189e+00 2.51627982e-01
8.10910344e-01 4.31106895e-01 1.53643787e-01 8.29030216e-01
-2.66212016e-01 9.64051425e-01 -1.04610300e+00 5.33453584e-01
2.74188995e-01 -4.83189881e-01 -1.29518163e+00 4.94754106e-01
-3.13423425e-02 4.18280274e-01 -3.92553397e-02 3.57595980e-01
5.44494271e-01 4.27076578e-01 5.46724260e-01 -1.14592946e+00
-7.57319689e-01 5.42581558e-01 -1.50852218e-01 2.09871545e-01
6.61125362e-01 5.03681064e-01 5.05352318e-01 -7.06702411e-01
1.01216292e+00 1.39874768e+00 5.50922275e-01 5.59891880e-01
-1.66938031e+00 -3.56216937e-01 2.35630170e-01 -1.98176563e-01
-1.21008193e+00 -3.78570616e-01 5.00706851e-01 -4.88393664e-01
1.17060387e+00 2.44397447e-01 1.10731363e+00 6.82296753e-01
2.44760290e-01 2.86998510e-01 1.32691383e+00 -4.47973460e-01
8.43377531e-01 2.77404428e-01 -6.74184412e-02 2.70967841e-01
1.59154817e-01 -7.34039024e-02 -5.99781513e-01 -2.87890643e-01
9.92891490e-01 -4.40579146e-01 1.66140143e-02 -6.17360651e-01
-1.33488333e+00 4.80295569e-01 2.85937399e-01 1.01183414e+00
-4.75977987e-01 2.62447208e-01 -7.36078843e-02 9.00650561e-01
5.31796813e-01 1.28545153e+00 -4.71049219e-01 -4.03717399e-01
-4.59446311e-01 4.62291777e-01 1.32573259e+00 5.89173496e-01
8.32353532e-01 -1.71109170e-01 1.83748409e-01 6.06688738e-01
1.71426892e-01 3.72333169e-01 3.58509481e-01 -1.67080045e+00
2.08790660e-01 9.84241009e-01 2.43986771e-01 -1.00428891e+00
-8.81364107e-01 -5.82325041e-01 -5.43952823e-01 5.51245451e-01
7.01889575e-01 -4.75380391e-01 -6.07391834e-01 1.53553796e+00
6.95605755e-01 5.62321953e-02 2.70154923e-01 7.68668175e-01
6.93970382e-01 8.87574434e-01 9.82096121e-02 -2.29466930e-01
8.62470031e-01 -8.10252368e-01 -3.87568563e-01 -1.59898713e-01
1.03736889e+00 -4.21519965e-01 1.02400875e+00 6.96721792e-01
-1.56986153e+00 -2.90223539e-01 -4.97726947e-01 9.31401998e-02
-5.47882676e-01 -6.15576982e-01 5.55202425e-01 5.12536407e-01
-1.43542433e+00 3.30868363e-01 -4.14666474e-01 -8.28384101e-01
2.30687514e-01 4.74172294e-01 -2.54383504e-01 5.88736124e-02
-9.32651460e-01 9.58590150e-01 4.60934937e-01 -4.45760280e-01
-8.66376936e-01 -9.11057353e-01 -2.42013052e-01 1.21781141e-01
1.98929816e-01 -1.19387925e+00 9.91505623e-01 -1.55206394e+00
-1.56102455e+00 1.04940486e+00 2.38500118e-01 3.14368755e-02
4.77037728e-01 2.18760341e-01 -2.48799339e-01 5.66538945e-02
-2.02511549e-02 3.81145418e-01 5.86308599e-01 -1.72493315e+00
-4.89822835e-01 -3.96052629e-01 4.40935433e-01 5.53779006e-01
-6.25233591e-01 -6.89724237e-02 -1.48166254e-01 -3.49700391e-01
-1.66409798e-02 -8.49201500e-01 -5.97926199e-01 -7.06354752e-02
2.95868486e-01 -4.57263619e-01 7.54704714e-01 -1.21651307e-01
1.36927235e+00 -2.17080069e+00 8.62405598e-01 4.69493538e-01
6.30290568e-01 8.13319981e-02 -3.45976025e-01 9.80657637e-01
7.86022842e-02 2.89847404e-01 -2.49530882e-01 -1.10992007e-01
-3.16926232e-03 4.50834513e-01 2.17454106e-01 3.45584363e-01
-5.04647121e-02 6.70238078e-01 -1.19055998e+00 -4.09145951e-01
5.14142141e-02 1.85245484e-01 -6.44804180e-01 1.80285424e-01
-7.66449094e-01 7.54681468e-01 -6.52715266e-01 3.83251548e-01
3.91447872e-01 -7.82089055e-01 5.31783342e-01 9.05038297e-01
-2.38406032e-01 -2.51266092e-01 -1.53758240e+00 1.82536137e+00
-6.45365357e-01 5.65875292e-01 5.47059953e-01 -9.95821476e-01
8.22705746e-01 5.00539839e-01 6.86538994e-01 -7.45366991e-01
3.87015752e-02 1.46938384e-01 -3.74525860e-02 -7.52661169e-01
1.89974681e-01 1.69739842e-01 -1.37198627e-01 1.32353389e+00
-5.25318272e-02 -3.52531672e-01 4.72231060e-01 3.07614982e-01
1.47465181e+00 -2.27451831e-01 1.57153562e-01 -6.42268538e-01
5.40174842e-01 3.44383091e-01 3.21412802e-01 1.01211715e+00
1.26863897e-01 2.19141528e-01 6.36656225e-01 -8.24918807e-01
-1.12511504e+00 -8.58489633e-01 -1.39317244e-01 1.50615430e+00
2.97137350e-01 -3.71034086e-01 -7.60984421e-01 -2.43017271e-01
1.24793202e-01 4.51647371e-01 -7.69737780e-01 1.61583528e-01
-6.53735280e-01 -5.71680248e-01 -3.00686453e-02 -2.26836223e-02
6.15878463e-01 -1.40123451e+00 -7.41696715e-01 4.80644792e-01
2.86716104e-01 -4.59917963e-01 2.49341026e-01 -1.29723132e-01
-1.00224268e+00 -1.02752268e+00 -7.87043214e-01 -9.30431247e-01
8.95936549e-01 1.34937450e-01 1.49905050e+00 4.25991893e-01
-3.10041577e-01 1.13036501e+00 -2.53408551e-01 -1.62173420e-01
-5.71642101e-01 4.10094261e-01 -8.53115767e-02 -1.57109067e-01
7.15235481e-03 -1.02796412e+00 -5.02538085e-01 3.20890993e-01
-8.81280899e-01 -1.63565770e-01 2.74894714e-01 6.03231132e-01
2.08103299e-01 3.20932508e-01 5.59354961e-01 -1.08405530e+00
1.25191510e+00 -1.07557213e+00 -4.50047493e-01 4.79354680e-01
-6.30344510e-01 -1.73588976e-01 6.08630538e-01 -6.00876689e-01
-1.02355611e+00 -1.30952805e-01 8.07845175e-01 6.37567788e-02
-4.38327730e-01 5.79374611e-01 5.27316853e-02 -2.37445459e-01
1.33526433e+00 -6.62535727e-02 1.59844026e-01 -3.57025802e-01
5.16065121e-01 4.20583069e-01 -1.51216432e-01 -9.99717057e-01
7.98204243e-01 2.51081884e-01 -1.42005965e-01 -8.66978288e-01
-2.49519572e-01 -2.38073871e-01 -6.09914243e-01 -5.48998296e-01
5.37446976e-01 -6.87305868e-01 -5.82444608e-01 7.51339734e-01
-1.02741337e+00 -9.36988711e-01 -7.91224182e-01 -1.54856101e-01
-4.53885376e-01 -1.58807710e-01 -2.41900548e-01 -6.61855817e-01
2.80358464e-01 -5.73668778e-01 5.50031900e-01 4.78208780e-01
-5.63473642e-01 -1.64753544e+00 5.63078701e-01 3.94801438e-01
9.56230283e-01 3.70723575e-01 8.47924829e-01 -5.23550034e-01
-7.32818127e-01 1.15334466e-01 2.13946011e-02 -2.87779659e-01
9.54255387e-02 9.06722993e-02 -6.96731269e-01 -4.24067348e-01
-9.92635265e-02 -3.61075580e-01 2.56230950e-01 7.37643689e-02
8.49284887e-01 -5.27952433e-01 -3.60361665e-01 2.81437933e-01
1.23238909e+00 5.81340715e-02 4.90050614e-01 5.98177254e-01
2.79920995e-01 8.29179525e-01 2.14554474e-01 8.70359838e-01
5.92441440e-01 3.00519019e-01 1.72820657e-01 -3.17202240e-01
-1.51937038e-01 3.89461517e-01 4.28849533e-02 5.76696098e-01
-3.28398287e-01 -3.78099233e-01 -1.60196495e+00 6.97073102e-01
-2.06981349e+00 -9.81519938e-01 -2.82486584e-02 1.56503165e+00
7.83451378e-01 -2.61779785e-01 2.90543556e-01 -5.30427285e-02
4.15871501e-01 -1.37576452e-02 -7.91004241e-01 -4.54229474e-01
-1.38002247e-01 2.71488041e-01 -1.75422877e-02 5.46757042e-01
-7.15940654e-01 9.33848619e-01 7.97427273e+00 5.16609848e-01
-8.19723666e-01 9.61014107e-02 5.19231915e-01 -3.14001322e-01
-4.58059520e-01 -3.34896520e-03 -8.88397358e-03 2.59823024e-01
9.20649230e-01 -6.42113328e-01 9.46527421e-01 5.50109982e-01
1.82239845e-01 -5.02407372e-01 -1.05088758e+00 7.19204247e-01
-1.99880049e-01 -1.53584290e+00 -2.63922304e-01 1.25932187e-01
1.24736202e+00 -2.75709629e-01 6.34343848e-02 1.38717934e-01
1.00024140e+00 -1.26141036e+00 3.34549159e-01 8.26134384e-01
3.18724126e-01 -6.10278368e-01 3.53154242e-01 7.40965307e-01
-7.86017239e-01 -7.43264794e-01 1.12950556e-01 -6.62152350e-01
-8.31324458e-02 3.89494389e-01 -2.94776917e-01 1.15867779e-01
8.92008841e-01 6.50175452e-01 -4.29564267e-01 1.08962953e+00
8.14584643e-02 3.22773159e-01 -3.83309782e-01 -1.66252658e-01
1.06351711e-01 -1.10594057e-01 6.78609252e-01 9.03598666e-01
2.50260442e-01 6.42144918e-01 3.77212673e-01 8.53280246e-01
2.55523741e-01 1.03300177e-01 -7.06932127e-01 -1.54144630e-01
6.00260437e-01 1.02174878e+00 -7.51326501e-01 -2.80154824e-01
8.73256698e-02 4.43846822e-01 2.50656456e-01 7.70576298e-01
-2.78312445e-01 -1.49977714e-01 4.95540470e-01 3.88391435e-01
-2.75445934e-02 -3.49628896e-01 -6.36541784e-01 -9.40409124e-01
-3.01369637e-01 -1.34149837e+00 3.87506515e-01 -6.32851601e-01
-1.35418940e+00 3.86416435e-01 5.20339645e-02 -8.13796103e-01
1.44141377e-03 -1.04548164e-01 -5.85843027e-01 7.13871717e-01
-9.72354591e-01 -9.77526188e-01 -5.64548373e-01 9.10888314e-01
1.93576843e-01 -5.18765390e-01 1.04701900e+00 -1.91252783e-01
-5.47280073e-01 -7.32371733e-02 4.86868531e-01 -3.70141238e-01
4.15759593e-01 -1.25128651e+00 1.21398970e-01 2.64842659e-01
-1.16303205e-01 5.74265718e-01 7.65934885e-01 -4.17209178e-01
-1.57295406e+00 -5.40251255e-01 8.12702060e-01 -7.68301725e-01
7.80396879e-01 -4.09906507e-01 -9.16163802e-01 2.95968890e-01
5.53161800e-01 -1.48715302e-01 6.77607417e-01 4.86869484e-01
-1.23198517e-02 -1.86403498e-01 -1.38699687e+00 4.33420658e-01
1.20197034e+00 -2.62711883e-01 -6.19058728e-01 8.40171516e-01
4.04344618e-01 -3.70089829e-01 -9.92089331e-01 -3.14108163e-01
9.99575034e-02 -1.11397755e+00 8.82540703e-01 -6.79915488e-01
3.01429898e-01 -1.85257390e-01 2.96399444e-01 -1.41276491e+00
-8.53785336e-01 -9.14246738e-01 -1.99259475e-01 1.12449574e+00
3.64517987e-01 -1.12531137e+00 6.86488688e-01 1.03389227e+00
2.92143196e-01 -3.51217568e-01 -1.05188811e+00 -6.76594257e-01
5.43393612e-01 6.76993234e-03 7.96094298e-01 1.19562852e+00
3.01437229e-01 1.25680998e-01 9.13000032e-02 3.01863719e-02
5.00907481e-01 2.00007007e-01 1.08713782e+00 -1.60625756e+00
-4.89041418e-01 -5.15370488e-01 8.40398148e-02 -6.61653042e-01
-3.80953178e-02 -6.46798730e-01 -3.76058638e-01 -1.82676160e+00
-2.41215080e-01 -1.06603396e+00 -2.24042684e-01 4.55867916e-01
3.46168429e-01 -8.30305070e-02 2.54398793e-01 6.27710760e-01
-1.12325525e+00 1.30101234e-01 1.40794194e+00 3.26161459e-02
-6.66579962e-01 -4.29331869e-01 -7.00594246e-01 6.04111910e-01
9.23133492e-01 -5.52771389e-01 -5.66395462e-01 -5.33612192e-01
7.48361588e-01 2.12073818e-01 4.39689718e-02 -1.35553563e+00
5.51805794e-01 -7.54840553e-01 2.42771074e-01 3.01925421e-01
8.05079192e-02 -1.07076240e+00 7.64269292e-01 4.44841415e-01
-3.56458366e-01 -3.98082472e-03 1.38117611e-01 4.03869867e-01
1.09572947e-01 -3.45764123e-02 5.28396964e-01 -6.87113404e-01
-6.31752849e-01 -1.04960777e-01 -8.49665999e-01 4.59799886e-01
1.82475853e+00 -4.40217495e-01 -5.10624290e-01 -2.16414168e-01
-1.08996153e+00 6.95465326e-01 9.73922372e-01 1.39369801e-01
2.32093588e-01 -1.01839018e+00 -7.46045589e-01 1.71999454e-01
-8.86777118e-02 8.32746699e-02 3.38666558e-01 4.75927591e-01
-7.80557454e-01 -4.74194214e-02 -2.92703688e-01 -5.06047606e-01
-9.96372461e-01 4.32105452e-01 6.24546647e-01 -4.39202279e-01
-4.98647928e-01 8.46641243e-01 4.74257395e-02 -5.59027314e-01
3.93740267e-01 1.95942953e-01 -2.21450165e-01 3.05543840e-01
4.36246216e-01 3.30751717e-01 -5.86684942e-01 -1.37500629e-01
-2.07367375e-01 8.44878793e-01 3.36550772e-01 -1.36103466e-01
1.77092695e+00 -2.03712091e-01 -6.18591607e-01 6.82814002e-01
3.72227669e-01 -3.28722745e-01 -1.17386615e+00 -4.21683520e-01
-5.04902303e-02 -5.18662870e-01 -5.90992197e-02 -1.11382556e+00
-9.83744979e-01 4.85684693e-01 4.05719072e-01 1.05562079e+00
1.07587290e+00 1.43492624e-01 2.11905852e-01 5.62464595e-01
5.31876087e-01 -1.53936350e+00 6.08407855e-01 7.22785115e-01
1.16154110e+00 -8.37848961e-01 8.08674768e-02 1.04938872e-01
-4.70528066e-01 1.33421230e+00 6.14042521e-01 -2.92717546e-01
8.34259033e-01 3.88232350e-01 2.21149623e-01 -6.12886548e-01
-1.31850195e+00 3.38738784e-02 -2.35968813e-01 9.79829550e-01
9.96664092e-02 -2.69040942e-01 -3.41453731e-01 -8.28015506e-02
8.66728425e-02 2.44071126e-01 4.39866126e-01 1.16689932e+00
-7.04062700e-01 -1.31106830e+00 -7.40303755e-01 -3.01949922e-02
1.68527588e-01 4.59853560e-01 -7.30999589e-01 8.12583387e-01
5.03000915e-01 9.86935556e-01 1.84936196e-01 9.91052985e-02
2.27622300e-01 -1.42090902e-01 4.09764290e-01 -7.18321264e-01
-1.36934710e+00 -4.16667700e-01 1.66425452e-01 -6.02437198e-01
-6.27500772e-01 -7.54708827e-01 -1.33365965e+00 -7.99461842e-01
1.76017538e-01 3.57831001e-01 2.84177423e-01 6.52466357e-01
7.69881129e-01 4.15436059e-01 7.54302263e-01 -8.09068978e-01
-1.76401347e-01 -5.26625454e-01 -6.55265868e-01 4.61144000e-01
1.31723166e-01 -3.60934377e-01 -5.18997073e-01 5.02080582e-02] | [4.111660957336426, 1.6595423221588135] |
988cecbe-44fd-4ce5-a0bb-f0d8768135f2 | deep-residual-network-based-food-recognition | 2005.04292 | null | https://arxiv.org/abs/2005.04292v2 | https://arxiv.org/pdf/2005.04292v2.pdf | Deep Residual Network based food recognition for enhanced Augmented Reality application | Deep neural network based learning approaches is widely utilized for image classification or object detection based problems with remarkable outcomes. Realtime Object state estimation of objects can be used to track and estimate the features that the object of the current frame possesses without causing any significant delay and misclassification. A system that can detect the features of such objects in the present state from camera images can be used to enhance the application of Augmented Reality for improving user experience and delivering information in a much perceptual way. The focus behind this paper is to determine the most suitable model to create a low-latency assistance AR to aid users by providing them nutritional information about the food that they consume in order to promote healthier life choices. Hence the dataset has been collected and acquired in such a manner, and we conduct various tests in order to identify the most suitable DNN in terms of performance and complexity and establish a system that renders such information realtime to the user. | ['Sainath G', 'Siddarth S', 'Vignesh S'] | 2020-05-08 | null | null | null | null | ['food-recognition'] | ['computer-vision'] | [ 7.52537251e-02 -3.13416839e-01 -2.74163157e-01 -3.71433228e-01
9.19916183e-02 -3.01667005e-01 1.98038995e-01 3.06457818e-01
-3.60377818e-01 3.42976749e-01 -1.63697332e-01 7.24709928e-02
-6.67831972e-02 -9.15439367e-01 -4.74636167e-01 -5.56769729e-01
-1.00041099e-01 2.32017726e-01 9.77984816e-02 -2.57643580e-01
1.74346715e-01 7.74405420e-01 -1.97662151e+00 2.35673234e-01
4.26560938e-01 1.06811619e+00 8.70640814e-01 7.29072213e-01
2.33476218e-02 8.12460005e-01 -2.89828867e-01 1.36436522e-01
2.87140906e-01 -1.79412454e-01 -5.44326678e-02 1.61487192e-01
2.74548560e-01 -9.71989155e-01 -4.94880050e-01 1.26407373e+00
6.46983445e-01 2.33192399e-01 3.49235922e-01 -1.00797784e+00
-5.68511963e-01 1.59790188e-01 -1.75177157e-01 6.56624794e-01
3.53207707e-01 8.64817277e-02 5.51829860e-02 -5.82384586e-01
3.15560579e-01 1.11637974e+00 4.10502970e-01 6.25230134e-01
-4.76708800e-01 -6.44610941e-01 2.51931399e-02 7.04217136e-01
-1.21176100e+00 -5.06872535e-01 8.85655999e-01 -1.04064509e-01
7.26657450e-01 3.98946553e-01 9.77978230e-01 8.83254111e-01
2.32271478e-01 8.40069175e-01 7.28415787e-01 -4.51926857e-01
3.54966134e-01 3.53469670e-01 3.40341806e-01 1.02056766e+00
4.43504155e-01 8.29446092e-02 -4.42083269e-01 3.72834980e-01
1.06063271e+00 5.81876993e-01 -1.52085662e-01 -9.29322839e-02
-1.15634084e+00 6.80614352e-01 5.73024213e-01 4.78105873e-01
-9.43625152e-01 2.47995958e-01 2.63954312e-01 1.10584415e-01
1.96227878e-01 8.55954140e-02 -8.00537944e-01 -1.01970704e-02
-4.94103551e-01 -1.12114593e-01 4.40895259e-01 7.66305506e-01
3.86884362e-01 4.12878275e-01 2.36525208e-01 4.88533407e-01
4.20200855e-01 6.99190199e-01 7.52048731e-01 -9.20789719e-01
-6.28105849e-02 7.22387135e-01 1.95390180e-01 -1.23205686e+00
-6.46174371e-01 -3.51072609e-01 -5.98288238e-01 4.84742135e-01
5.24187349e-02 -1.27986923e-01 -8.61772597e-01 1.32733357e+00
5.11938274e-01 2.96275526e-01 1.50563896e-01 1.06251013e+00
1.07786179e+00 1.03266561e+00 8.22879225e-02 -1.39986619e-01
1.77343535e+00 -4.75447148e-01 -9.49881375e-01 5.72233833e-02
5.56218743e-01 -6.64384007e-01 7.78516710e-01 7.26035714e-01
-7.75736272e-01 -9.03683066e-01 -1.27609646e+00 1.95208177e-01
-6.00976884e-01 6.27556980e-01 8.09928536e-01 7.44741678e-01
-9.83656824e-01 5.12854695e-01 -1.03660083e+00 -5.45824945e-01
1.91745266e-01 6.67915642e-01 -1.16821609e-01 -5.49772196e-02
-8.33861291e-01 1.14767301e+00 4.57374185e-01 4.58130330e-01
-1.06438386e+00 -6.47187293e-01 -7.66258061e-01 -4.96120080e-02
2.47475043e-01 -6.27150655e-01 1.11873114e+00 -1.18501031e+00
-1.32564080e+00 3.56446058e-01 1.06500067e-01 -5.00680089e-01
2.81107336e-01 -2.83325404e-01 -6.19221509e-01 3.12015653e-01
-1.37645721e-01 7.19048560e-01 7.72694767e-01 -9.19121861e-01
-9.90639091e-01 -8.35898578e-01 1.67320058e-01 5.75127482e-01
-3.93769443e-01 -1.52821047e-02 -1.37129202e-01 -2.22595572e-01
4.45005834e-01 -8.44985306e-01 -1.35511309e-01 5.10918021e-01
-7.18697757e-02 -1.62107140e-01 1.34574676e+00 -7.89894581e-01
5.47469974e-01 -1.89651525e+00 -4.27945316e-01 -2.85813510e-01
-4.48815571e-03 6.14465296e-01 3.45532261e-02 1.52848341e-04
8.54459479e-02 -4.42061633e-01 4.78163093e-01 1.76080853e-01
-5.31597495e-01 1.58472404e-01 1.31242231e-01 5.82080603e-01
3.55171002e-02 3.83667499e-01 -7.46227920e-01 -3.26888829e-01
8.33373189e-01 9.67681408e-01 -3.17478806e-01 2.18344525e-01
-2.30417728e-01 3.30739766e-01 -7.29739368e-01 7.42027283e-01
4.06177759e-01 8.69623795e-02 -1.50433928e-01 -8.69654179e-01
-1.91295773e-01 -7.34933019e-02 -1.39366663e+00 1.42745805e+00
-5.93316674e-01 7.66132534e-01 -1.29718497e-01 -1.02074862e+00
9.07218456e-01 5.62150717e-01 7.09357798e-01 -1.08814764e+00
4.61202294e-01 -1.92848742e-01 -7.21558779e-02 -1.10572350e+00
6.94286942e-01 3.68595362e-01 3.61729860e-01 8.55641812e-02
-2.68506289e-01 6.25754297e-01 2.74708476e-02 -1.31160274e-01
8.76373887e-01 2.51273602e-01 2.69876778e-01 -1.00780196e-01
5.47765732e-01 8.98396745e-02 1.73385784e-01 4.78041023e-01
-4.54285741e-01 1.32245412e-02 -5.57009399e-01 -9.17617142e-01
-9.47203457e-01 -8.64255965e-01 -1.04318885e-03 8.29693258e-01
4.45587665e-01 4.91916835e-01 -3.83144468e-01 -3.33788425e-01
-4.19470876e-01 6.93971992e-01 -2.53142923e-01 -2.48889655e-01
-5.90068638e-01 -7.04832494e-01 6.41919486e-03 4.48520422e-01
8.11286747e-01 -1.32827878e+00 -1.33279920e+00 3.84813428e-01
3.11961509e-02 -8.89222324e-01 -4.75056060e-02 2.28582919e-01
-1.06838262e+00 -1.12544000e+00 -6.66354299e-01 -1.00107419e+00
7.69680023e-01 5.71523964e-01 6.87664926e-01 1.54056832e-01
-4.53140914e-01 3.59185427e-01 -4.03706580e-01 -6.59699500e-01
-3.51359338e-01 -3.80431294e-01 3.33635628e-01 -2.49506563e-01
4.92386580e-01 -1.25739187e-01 -1.00868857e+00 -2.64476091e-02
-8.31266642e-01 -1.01753678e-02 3.71066481e-01 1.06077410e-01
3.76653492e-01 4.24001127e-01 7.42350221e-01 -4.78880763e-01
2.72417843e-01 -3.11531156e-01 -7.41190612e-01 1.23297404e-02
-5.59467614e-01 -6.14472739e-02 4.92234111e-01 -6.69785678e-01
-9.94442463e-01 3.74023438e-01 -8.78490135e-02 -1.23448901e-01
-3.58447850e-01 2.38297880e-01 -1.42124057e-01 -1.85381137e-02
2.66904861e-01 3.09001684e-01 -3.97798195e-02 -5.66442132e-01
2.80402511e-01 7.07414269e-01 6.64405167e-01 9.32780951e-02
2.05006883e-01 4.24733549e-01 6.03774749e-02 -1.04926074e+00
-4.67226565e-01 -6.08833969e-01 -3.70708138e-01 -6.47116780e-01
8.70790422e-01 -1.11659515e+00 -9.87284601e-01 2.90887386e-01
-1.17449534e+00 7.10845664e-02 3.34487483e-02 8.60010743e-01
-2.57897943e-01 9.94553044e-02 -5.81491888e-01 -1.03315604e+00
-6.16403103e-01 -1.05769420e+00 6.78357065e-01 5.98309875e-01
1.72749981e-01 -7.06159532e-01 -2.65001595e-01 2.99693972e-01
4.00728405e-01 4.55129854e-02 7.57915676e-01 -5.47270298e-01
-7.37239361e-01 -2.06990942e-01 -3.31346482e-01 2.26970091e-01
2.05422938e-01 -9.90507528e-02 -7.67470837e-01 -8.44391361e-02
2.73206681e-01 2.73093849e-01 2.86229849e-01 7.45813608e-01
8.99267972e-01 -2.10990906e-01 -3.20952743e-01 3.65250170e-01
1.82818258e+00 1.01975977e+00 7.07360387e-01 4.12233353e-01
6.90244794e-01 2.56510645e-01 8.36886883e-01 3.04775745e-01
2.81143427e-01 7.22972870e-01 8.09841514e-01 -2.61989206e-01
-2.50488281e-01 2.27199435e-01 4.69687670e-01 6.77937567e-01
1.58920616e-01 -3.69382441e-01 -2.52214581e-01 4.41983610e-01
-1.69945526e+00 -8.56222749e-01 -3.04760098e-01 2.01842928e+00
1.86081693e-01 -1.49142951e-01 1.04568794e-01 3.82064611e-01
7.91762173e-01 -4.12051290e-01 -6.25481129e-01 -2.77919739e-01
3.99997801e-01 4.07959335e-03 5.46960890e-01 3.52969877e-02
-9.12910640e-01 6.11253262e-01 5.96372509e+00 2.81416953e-01
-1.39954638e+00 1.02572523e-01 4.99571174e-01 -2.08738334e-02
3.90554935e-01 -5.15334189e-01 -6.91244304e-01 5.27300477e-01
1.06796432e+00 2.79285848e-01 3.08546156e-01 1.01924431e+00
7.20787466e-01 -3.72102857e-01 -1.11175191e+00 1.08838642e+00
1.32585093e-01 -1.14963436e+00 -1.22966040e-02 -1.22494929e-01
4.01634604e-01 8.89363978e-03 1.15079820e-01 5.05789323e-03
-1.14173144e-01 -4.50105637e-01 5.89237154e-01 7.21976578e-01
1.42840266e-01 -9.26735163e-01 6.76033437e-01 5.05325556e-01
-1.16441333e+00 -2.22748712e-01 -5.88184297e-01 -2.53153354e-01
-1.32932970e-02 3.26604396e-01 -1.10815012e+00 1.34610087e-01
5.63556194e-01 2.46414766e-01 -3.44513714e-01 1.28910327e+00
3.52392524e-01 2.87804842e-01 -3.32381457e-01 -4.07384336e-01
8.81661102e-02 -2.62374371e-01 5.27863383e-01 6.55984044e-01
7.75826335e-01 3.41572881e-01 1.16762854e-01 6.07926250e-01
1.40910238e-01 1.27154678e-01 -7.03888297e-01 1.19338825e-01
2.71196455e-01 1.32915628e+00 -1.09111118e+00 -5.43350279e-01
-1.85619712e-01 1.04805326e+00 -2.66436160e-01 -1.86388101e-02
-8.47396612e-01 1.13903955e-02 3.22896570e-01 2.51124412e-01
2.72134274e-01 -3.21152240e-01 9.39831734e-02 -8.36243331e-01
-1.38217926e-01 -6.82440817e-01 2.09247068e-01 -1.27529299e+00
-5.56423545e-01 7.04183578e-01 -1.79268718e-01 -1.10206616e+00
-1.81535572e-01 -7.65717566e-01 -2.28924870e-01 3.25869679e-01
-1.41719019e+00 -1.12140095e+00 -6.39281869e-01 5.51253796e-01
8.52994144e-01 -2.93390244e-01 7.43754327e-01 7.93272257e-01
-6.66966081e-01 -8.37840047e-03 4.93774526e-02 -1.73916325e-01
1.05509035e-01 -8.63283098e-01 -2.54102319e-01 9.98960018e-01
3.19988251e-01 1.50718942e-01 8.99738967e-01 -7.47517645e-01
-1.80391812e+00 -1.10312760e+00 4.46852237e-01 -3.65742922e-01
-3.25821415e-02 1.48693293e-01 -5.37653625e-01 3.97084206e-01
-6.90051094e-02 1.24294385e-02 1.81572154e-01 -5.16765177e-01
4.08792734e-01 -4.86852020e-01 -1.56340826e+00 5.12044549e-01
6.26050055e-01 -2.47764707e-01 -2.81229675e-01 3.95435631e-01
5.98582745e-01 -2.81163514e-01 -5.19273043e-01 2.69884646e-01
6.39754772e-01 -9.50290680e-01 1.07493150e+00 -3.60507965e-01
2.13598292e-02 -4.33624774e-01 -1.77305982e-01 -8.80629480e-01
-1.68836012e-01 -5.67307882e-02 -2.21807063e-01 9.94866073e-01
-3.70242707e-02 -3.53520423e-01 1.14633048e+00 4.83809769e-01
-8.68292972e-02 -4.71091807e-01 -6.36436820e-01 -3.56205761e-01
-9.54487264e-01 -2.49378324e-01 4.79321390e-01 6.92708492e-01
-5.14322698e-01 7.29670981e-03 -4.40895379e-01 7.02516496e-01
4.96380031e-01 -3.49857718e-01 4.15823966e-01 -1.11375582e+00
3.01423490e-01 2.35450998e-01 -8.05831850e-01 -7.58953094e-01
-3.72112989e-01 -4.76210415e-01 1.27248377e-01 -1.52289188e+00
1.92279726e-01 -5.34558594e-01 -4.31668490e-01 1.57200426e-01
4.82548364e-02 4.94667709e-01 9.96082276e-02 1.57138053e-02
-5.08613944e-01 1.70882434e-01 1.09925640e+00 -2.37579867e-02
-3.49657059e-01 3.10356349e-01 -5.37713587e-01 8.90847266e-01
6.84975147e-01 -5.23813426e-01 -6.41579330e-01 -3.70257884e-01
-3.38724181e-02 2.17069581e-01 5.31456828e-01 -1.49927235e+00
2.40981966e-01 1.61863521e-01 9.28720057e-01 -1.07495070e+00
6.26604378e-01 -1.53842533e+00 3.79515767e-01 9.73298848e-01
-1.71570554e-01 4.01636958e-01 7.82556981e-02 5.78594267e-01
3.04577380e-01 -5.80792785e-01 5.70468664e-01 -3.39960247e-01
-1.30169737e+00 3.84374231e-01 -5.48383236e-01 -8.95286977e-01
1.18453562e+00 -3.26075166e-01 3.12596783e-02 -4.52975661e-01
-5.66590309e-01 -2.89724797e-01 2.76926868e-02 4.56428438e-01
7.41468668e-01 -1.16156912e+00 -2.29542270e-01 3.00907582e-01
-1.32555634e-01 -3.33958328e-01 2.58102983e-01 2.87315667e-01
-9.65498149e-01 4.21086967e-01 -6.95236504e-01 -6.59615874e-01
-1.43849552e+00 8.06012332e-01 2.54458934e-01 1.35623395e-01
-7.56067872e-01 2.96387553e-01 -2.86581755e-01 2.04740852e-01
4.04005677e-01 -5.33604383e-01 -5.75159967e-01 -1.31578326e-01
8.66551697e-01 6.25561774e-01 9.76549238e-02 -5.81582010e-01
-2.01347247e-01 2.50818521e-01 1.49048064e-02 2.59430647e-01
1.42650056e+00 -5.18025160e-01 3.23750973e-01 2.51542389e-01
1.05962002e+00 -5.94811261e-01 -1.35520637e+00 -4.84958254e-02
-2.38042399e-01 -3.60965818e-01 8.17636549e-01 -8.50603938e-01
-1.45018566e+00 5.43553054e-01 1.82664490e+00 2.94157535e-01
1.15513408e+00 -4.34733063e-01 9.44999218e-01 6.33344829e-01
4.48258370e-01 -8.78860950e-01 1.82037890e-01 -1.04455128e-01
4.10771400e-01 -1.38947582e+00 9.44051817e-02 -1.62694469e-01
-3.39779109e-01 1.38358736e+00 6.07302606e-01 -2.60393858e-01
6.13132238e-01 4.16634679e-01 2.81589419e-01 -4.89862144e-01
-3.26539963e-01 -2.71640748e-01 7.88535476e-02 1.03104985e+00
1.54974535e-01 2.58095115e-02 1.80785563e-02 2.04635501e-01
1.30323291e-01 4.54390228e-01 5.22715628e-01 8.88214290e-01
-7.98399329e-01 -7.05072165e-01 -4.04191673e-01 4.65677083e-01
-5.63390911e-01 2.59378046e-01 4.30757403e-01 7.35918283e-01
4.51053858e-01 8.38607192e-01 2.20511228e-01 -1.47199675e-01
2.07716957e-01 -3.42642397e-01 6.04672015e-01 -2.70134687e-01
-3.63804936e-01 -9.69290212e-02 8.48089680e-02 -3.11959177e-01
-5.48644841e-01 -3.41114700e-01 -1.34101033e+00 -1.96585774e-01
-5.32565832e-01 -2.55253404e-01 1.45079672e+00 1.01717150e+00
1.43240035e-01 6.89131737e-01 5.13354897e-01 -9.72303748e-01
-4.03514445e-01 -7.38369346e-01 -3.64561617e-01 1.90848723e-01
3.53886932e-01 -5.53564727e-01 2.13386014e-01 2.92903394e-01] | [8.331857681274414, -0.9234289526939392] |
39ea574f-33f6-40c0-9f24-209de88c437f | remips-physically-consistent-3d | null | null | http://proceedings.neurips.cc/paper/2021/hash/a1a2c3fed88e9b3ba5bc3625c074a04e-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/a1a2c3fed88e9b3ba5bc3625c074a04e-Paper.pdf | REMIPS: Physically Consistent 3D Reconstruction of Multiple Interacting People under Weak Supervision | The three-dimensional reconstruction of multiple interacting humans given a monocular image is crucial for the general task of scene understanding, as capturing the subtleties of interaction is often the very reason for taking a picture. Current 3D human reconstruction methods either treat each person independently, ignoring most of the context, or reconstruct people jointly, but cannot recover interactions correctly when people are in close proximity. In this work, we introduce \textbf{REMIPS}, a model for 3D \underline{Re}construction of \underline{M}ultiple \underline{I}nteracting \underline{P}eople under Weak \underline{S}upervision. \textbf{REMIPS} can reconstruct a variable number of people directly from monocular images. At the core of our methodology stands a novel transformer network that combines unordered person tokens (one for each detected human) with positional-encoded tokens from image features patches. We introduce a novel unified model for self- and interpenetration-collisions based on a mesh approximation computed by applying decimation operators. We rely on self-supervised losses for flexibility and generalisation in-the-wild and incorporate self-contact and interaction-contact losses directly into the learning process. With \textbf{REMIPS}, we report state-of-the-art quantitative results on common benchmarks even in cases where no 3D supervision is used. Additionally, qualitative visual results show that our reconstructions are plausible in terms of pose and shape and coherent for challenging images, collected in-the-wild, where people are often interacting. | ['Cristian Sminchisescu', 'Vlad Olaru', 'Eduard Bazavan', 'Teodor Szente', 'Mihai Zanfir', 'Mihai Fieraru'] | 2021-12-01 | null | https://openreview.net/forum?id=-AV3AKwgiG | https://openreview.net/pdf?id=-AV3AKwgiG | neurips-2021-12 | ['3d-human-reconstruction'] | ['computer-vision'] | [ 3.16356048e-02 1.33307293e-01 5.41828811e-01 -4.65944231e-01
-2.77235568e-01 -2.81245679e-01 8.59776020e-01 5.08135296e-02
-5.86764097e-01 7.33302951e-01 2.14105114e-01 5.38556218e-01
-2.50448704e-01 -6.27726734e-01 -1.20186055e+00 -5.54584086e-01
-1.19654633e-01 1.30159461e+00 1.78287908e-01 -2.95054287e-01
-3.23805660e-01 6.22754395e-01 -2.04268193e+00 3.87608409e-01
4.74107414e-01 9.44498003e-01 8.57017785e-02 8.24085474e-01
3.93168628e-01 5.96470296e-01 -5.53146303e-01 -6.58408105e-01
5.64523458e-01 -2.79117376e-01 -5.36835670e-01 4.38731104e-01
9.96681273e-01 -3.99598479e-01 -5.09920835e-01 5.32586992e-01
5.52782536e-01 1.48028240e-01 6.73595846e-01 -1.27748585e+00
-6.87584057e-02 6.90635145e-02 -5.14104128e-01 1.12129645e-02
9.57599103e-01 3.80853981e-01 8.60860229e-01 -1.01612353e+00
9.13974285e-01 1.69620001e+00 8.88007879e-01 3.98834705e-01
-1.52596068e+00 -3.99763405e-01 2.36769125e-01 1.60955954e-02
-1.42032492e+00 -4.28280622e-01 5.43244421e-01 -5.55853188e-01
9.29864645e-01 5.63077867e-01 1.00356209e+00 1.41682935e+00
-5.12591936e-02 9.42858100e-01 9.16682839e-01 -3.35831672e-01
-2.02818155e-01 2.01217398e-01 2.78781243e-02 6.68746710e-01
2.74348080e-01 3.98958892e-01 -9.08449888e-01 -9.52737182e-02
1.11004961e+00 2.65024155e-01 -1.45780191e-01 -7.89463878e-01
-1.23730731e+00 4.64500636e-01 4.89477277e-01 -4.58757691e-02
-3.76259416e-01 1.28685340e-01 1.77480742e-01 1.18330747e-01
4.76644725e-01 6.27303720e-02 -9.04198065e-02 1.27362818e-01
-8.22415709e-01 1.06914723e+00 6.87573731e-01 9.86977637e-01
7.87365139e-01 -3.79282653e-01 -1.26606956e-01 7.74893045e-01
-2.21598402e-01 6.46264434e-01 -3.78602475e-01 -1.26371360e+00
4.61972028e-01 5.54182827e-01 5.68183601e-01 -9.87732232e-01
-5.33120871e-01 -3.15077633e-01 -1.10108733e+00 4.33081716e-01
5.23481846e-01 -2.75963489e-02 -4.76569086e-01 1.63638961e+00
6.18083179e-01 1.17657401e-01 -3.42571557e-01 1.20071352e+00
7.55230248e-01 2.53706932e-01 -2.72781223e-01 9.79449153e-02
1.26875305e+00 -8.87943983e-01 -2.80326962e-01 -2.99895465e-01
5.88018708e-02 -3.49296212e-01 8.39114964e-01 7.03941107e-01
-1.36487138e+00 -8.11284661e-01 -5.87340713e-01 -2.65726596e-01
-1.45787567e-01 5.80403507e-02 2.83811808e-01 2.73240745e-01
-1.02941656e+00 5.91135859e-01 -5.11834145e-01 -5.09697258e-01
3.50902140e-01 4.82003719e-01 -9.49930370e-01 -2.38355339e-01
-8.33577693e-01 8.62792611e-01 -1.27783760e-01 3.37466776e-01
-8.61980796e-01 -5.39924026e-01 -1.01362991e+00 -2.89322972e-01
5.38472593e-01 -1.21913624e+00 7.73138642e-01 -8.85249078e-01
-9.27818954e-01 1.20969069e+00 -5.21320105e-02 -6.10547125e-01
1.27296364e+00 -7.50951409e-01 -1.05322309e-01 2.27856040e-01
2.18580067e-01 9.12858665e-01 7.97459722e-01 -1.73046613e+00
-5.84613264e-01 -5.65450490e-01 1.93615690e-01 5.36900759e-01
3.16046119e-01 1.12233423e-02 -4.36598390e-01 -5.97649336e-01
-1.92643955e-01 -8.00356805e-01 -5.24355285e-02 5.64394772e-01
-5.81061900e-01 -7.70209879e-02 7.66486943e-01 -7.67882466e-01
5.10344982e-01 -1.94780993e+00 8.69047403e-01 1.74255505e-01
4.68057662e-01 1.12985432e-01 1.03598781e-01 5.05532563e-01
1.01228409e-01 -4.38591301e-01 -3.31515968e-01 -1.21640170e+00
1.64273873e-01 4.11175668e-01 -3.95535976e-02 7.21954107e-01
-1.01502985e-01 8.88420105e-01 -6.46855891e-01 -4.49678451e-01
7.12809861e-01 8.31820846e-01 -7.43489742e-01 6.48703694e-01
-2.29282573e-01 7.05040812e-01 4.03007157e-02 4.19245303e-01
6.41237140e-01 -5.47600947e-02 -1.58439979e-01 -1.87615037e-01
-7.07947984e-02 -1.44692972e-01 -1.53037500e+00 1.60909593e+00
-2.51554400e-01 2.59476304e-01 6.62130773e-01 -8.22676837e-01
5.53790331e-01 8.40484574e-02 5.48917532e-01 -3.72878522e-01
1.40596777e-01 -1.58347115e-01 -5.43005943e-01 -3.54626954e-01
4.40222472e-01 -1.23085462e-01 -2.14301571e-01 2.82562792e-01
3.07464157e-03 -7.12506920e-02 -3.92407924e-02 1.93252236e-01
9.40680683e-01 3.49063724e-01 2.09356677e-02 -1.41457813e-02
4.25045401e-01 -3.65052581e-01 1.93424672e-01 8.11588347e-01
6.62568882e-02 1.05501974e+00 2.70967096e-01 -7.62185395e-01
-1.11078596e+00 -1.36196661e+00 7.03566195e-03 8.87479007e-01
3.68448585e-01 -2.87644386e-01 -7.15122581e-01 -5.31369388e-01
2.12285563e-01 3.86543572e-01 -8.89614761e-01 2.45927662e-01
-6.63562059e-01 -2.94382513e-01 5.72192788e-01 2.51488686e-01
5.01594543e-01 -1.22089434e+00 -1.14176357e+00 -1.32830683e-02
-5.39397359e-01 -1.28539681e+00 -3.40352416e-01 -1.54286101e-01
-1.82138577e-01 -1.14648104e+00 -9.21357036e-01 -3.32450122e-01
6.56528711e-01 2.25833505e-01 1.38040352e+00 1.76043302e-01
-6.38578117e-01 7.94605911e-01 -1.59465194e-01 -1.00379735e-01
2.13562232e-02 -3.78271252e-01 3.29718530e-01 2.60011762e-01
1.38789624e-01 -7.47708142e-01 -6.49979293e-01 5.02488136e-01
-4.51368719e-01 3.06864113e-01 3.27898175e-01 8.01253140e-01
5.81937790e-01 -3.85285877e-02 -3.15825760e-01 -8.03879380e-01
5.40888309e-02 -1.16625026e-01 -2.93343157e-01 -2.04799175e-02
2.71376371e-01 -2.25901827e-01 4.44464028e-01 -3.16836089e-01
-9.13155973e-01 1.03074558e-01 1.75079763e-01 -9.08816516e-01
-5.74332416e-01 -2.83774465e-01 -3.02167296e-01 -5.34887426e-03
7.35733747e-01 7.33359680e-02 6.46082759e-02 -5.05301237e-01
1.23723291e-01 1.94789827e-01 8.31264675e-01 -7.19207108e-01
6.94096327e-01 8.85291457e-01 1.43382266e-01 -1.09959769e+00
-8.56146991e-01 -3.74850988e-01 -8.99053514e-01 -5.76768517e-01
9.05396879e-01 -1.08510733e+00 -1.20411301e+00 6.67738020e-01
-1.18730795e+00 -4.78635252e-01 -5.35919666e-01 1.05683707e-01
-6.75830185e-01 4.69734222e-01 -3.69561523e-01 -1.04454279e+00
6.48859367e-02 -8.43634903e-01 1.64878809e+00 -2.82972604e-01
-4.12422121e-01 -7.65922070e-01 -3.98000469e-03 8.14126432e-01
-2.47042641e-01 6.06923342e-01 2.20220134e-01 -1.78237677e-01
-6.92496836e-01 -2.04460770e-01 -2.12508991e-01 3.06862861e-01
-2.69038230e-01 -5.57254434e-01 -8.93073440e-01 -2.31062084e-01
-1.89280778e-01 -3.79459709e-01 7.68290877e-01 3.60752046e-01
8.73077393e-01 -4.45894003e-01 -4.04260427e-01 5.48814476e-01
9.27869439e-01 -8.32337081e-01 6.66437685e-01 -2.44129494e-01
1.07318532e+00 1.18891108e+00 4.19495881e-01 7.96303451e-01
7.73567915e-01 1.11944604e+00 5.47577560e-01 -1.08218849e-01
-2.42474824e-01 -3.21606904e-01 3.41806710e-01 -1.13986820e-01
-4.93164062e-01 -2.84310222e-01 -7.41025031e-01 4.81485397e-01
-1.94936812e+00 -1.26755679e+00 -2.59410441e-01 2.32992721e+00
3.74844640e-01 8.80950913e-02 5.71336627e-01 8.74938816e-02
6.48371816e-01 -3.38696614e-02 -4.05893624e-01 3.12108696e-01
-3.89001757e-01 9.21962932e-02 2.67272294e-01 8.26526940e-01
-1.06035888e+00 6.18429303e-01 5.20350361e+00 4.70800072e-01
-4.85224724e-01 6.36648983e-02 4.73086983e-01 -5.71825027e-01
7.50667825e-02 -2.73434103e-01 -9.52719271e-01 3.60872298e-01
3.10101867e-01 5.70053220e-01 6.31960928e-01 3.06415290e-01
1.59466773e-01 -4.77585346e-01 -1.51584971e+00 1.32416296e+00
4.53675896e-01 -1.08922660e+00 3.98407094e-02 2.99581915e-01
4.38937336e-01 -3.14087003e-01 -2.61883903e-02 -4.17914847e-03
3.68401349e-01 -1.29634678e+00 1.21902120e+00 9.85414743e-01
7.80200362e-01 -6.57697380e-01 5.62068403e-01 7.80521274e-01
-1.33355904e+00 4.20546196e-02 -5.97662143e-02 -3.59101087e-01
5.02141774e-01 6.49225354e-01 -3.22730303e-01 7.75924623e-01
1.11504424e+00 8.27677369e-01 -4.31005925e-01 6.54817045e-01
2.23640129e-02 -2.71230906e-01 -7.27275372e-01 3.96267772e-01
-2.01370031e-01 -1.85700923e-01 8.52795899e-01 1.09900177e+00
1.51755080e-01 3.28108132e-01 4.39634562e-01 8.72779906e-01
1.74589649e-01 -3.50654215e-01 -5.45902073e-01 7.17186987e-01
1.11554906e-01 8.43236327e-01 -4.46642578e-01 -1.88750938e-01
-5.82202822e-02 1.39948821e+00 6.07090056e-01 1.93960920e-01
-7.31954575e-01 2.34793365e-01 7.07809329e-01 9.52200830e-01
4.54115421e-01 -1.71922073e-01 -9.30094421e-02 -1.25588596e+00
3.68444741e-01 -7.40148485e-01 2.51874745e-01 -9.34810281e-01
-1.48927975e+00 4.33257878e-01 5.69442987e-01 -9.66760874e-01
-2.40082413e-01 -4.85351294e-01 -1.18848331e-01 7.61901557e-01
-8.06344151e-01 -1.55030823e+00 -6.20687842e-01 8.38458121e-01
4.59843695e-01 2.27408223e-02 6.20047152e-01 1.71182394e-01
-2.87633747e-01 3.54676604e-01 -4.21467513e-01 -1.52834520e-01
5.85165381e-01 -1.16286337e+00 2.73639053e-01 5.95867157e-01
2.06484228e-01 4.16421294e-01 1.09858787e+00 -7.11051524e-01
-1.09340858e+00 -9.23956335e-01 8.97836626e-01 -8.60237479e-01
-1.24444507e-01 -1.04658008e+00 -5.89860141e-01 8.29868317e-01
-1.30201995e-01 2.07595363e-01 2.05681652e-01 1.23901829e-01
-2.77275383e-01 -1.16336541e-02 -1.49703419e+00 6.15799487e-01
1.63951671e+00 -4.39707786e-01 -5.52284241e-01 5.24165511e-01
2.65985191e-01 -5.58382154e-01 -5.20877242e-01 1.50560766e-01
6.72729731e-01 -1.79343164e+00 1.54970419e+00 8.85282904e-02
2.95665711e-01 -1.33946359e-01 -2.14024603e-01 -8.35785210e-01
-2.19755005e-02 -7.03405082e-01 -3.09817463e-01 8.86402607e-01
-3.93556178e-01 -4.07839924e-01 9.46758330e-01 5.80456972e-01
2.18809947e-01 -5.79026401e-01 -1.14655566e+00 -6.28094256e-01
-2.13388696e-01 -2.94980466e-01 3.17971140e-01 4.77629304e-01
-4.67335075e-01 2.09615275e-01 -9.01945531e-01 2.34669536e-01
1.25819564e+00 -3.15763593e-01 1.39661276e+00 -1.56665099e+00
-7.71248698e-01 -2.83392873e-02 -4.42290783e-01 -1.26277602e+00
2.64837116e-01 -4.42814469e-01 5.92654683e-02 -1.40835083e+00
2.72962660e-01 -3.37805510e-01 4.08744931e-01 2.24849150e-01
1.36890963e-01 5.03917873e-01 2.80508429e-01 2.31328249e-01
-7.63327837e-01 7.31390595e-01 1.07382619e+00 1.22784533e-01
9.85023975e-02 3.58046666e-02 -2.53955692e-01 8.16027820e-01
1.67937223e-02 -1.31521925e-01 -1.51530653e-01 -4.21533287e-01
1.78942546e-01 1.00672655e-01 1.43450236e+00 -1.33106160e+00
2.89338678e-01 5.59730008e-02 6.27256453e-01 -7.68457532e-01
1.23855972e+00 -9.31682348e-01 7.84068406e-01 2.30495796e-01
-1.01128496e-01 -7.63634667e-02 -1.04333527e-01 6.15035772e-01
1.59393758e-01 2.13552266e-01 6.21334255e-01 -4.51729923e-01
-2.73160040e-01 3.60323727e-01 -7.02668121e-03 1.55738324e-01
9.39598382e-01 -5.52142501e-01 2.71628559e-01 -6.24555290e-01
-9.06450331e-01 2.61691093e-01 7.81252146e-01 2.64690250e-01
7.12880611e-01 -1.14026189e+00 -8.35138738e-01 3.73571992e-01
-2.70172115e-02 4.33627337e-01 6.63800895e-01 6.08383775e-01
-5.13403356e-01 1.34993151e-01 -2.28664115e-01 -9.10077870e-01
-1.51813447e+00 3.87223810e-01 5.95184565e-01 -3.50038856e-01
-9.34109569e-01 1.02769339e+00 6.79658055e-01 -7.64382958e-01
4.53303695e-01 9.75361243e-02 1.51679516e-01 -1.43606201e-01
3.95372570e-01 6.36558652e-01 -1.97833404e-01 -1.25327587e+00
-2.21851245e-01 7.54765749e-01 1.04429625e-01 -4.07632947e-01
1.40923297e+00 -3.27681571e-01 -2.38910280e-02 5.32857597e-01
7.49556303e-01 -6.06305152e-02 -1.80328071e+00 -2.89033115e-01
-7.99098730e-01 -6.59173548e-01 -5.27925611e-01 -8.84865463e-01
-7.04622686e-01 8.77622545e-01 3.51611137e-01 -2.14484870e-01
6.63320363e-01 3.12980920e-01 6.34266138e-01 3.53398770e-01
9.35411930e-01 -8.47925425e-01 2.71300197e-01 2.95638949e-01
1.26359224e+00 -9.41788256e-01 2.78953701e-01 -7.48386562e-01
-6.64948046e-01 4.79908109e-01 5.63143671e-01 -5.25843084e-01
5.41084945e-01 1.38148457e-01 -3.48295540e-01 -4.49043065e-01
-4.56273079e-01 -4.06520590e-02 1.33848548e-01 8.24472725e-01
-1.93806082e-01 1.01755798e-01 3.30857784e-01 5.89732707e-01
-5.05621493e-01 -2.41134152e-01 1.17419325e-01 7.43430793e-01
-1.01336181e-01 -9.55031753e-01 -6.55579269e-01 3.18475634e-01
5.93229383e-02 2.89943486e-01 -7.11996734e-01 8.24781537e-01
7.89909184e-01 7.61217833e-01 2.20975608e-01 -2.32297346e-01
7.03480840e-01 -3.33245009e-01 1.04342282e+00 -5.44972360e-01
-9.61811841e-01 -5.11455014e-02 2.21184030e-01 -7.45003402e-01
-6.01252139e-01 -1.16654217e+00 -8.75346720e-01 -6.69131339e-01
1.76213965e-01 -2.24980995e-01 1.42471539e-02 1.08752418e+00
2.33160123e-01 2.16755822e-01 2.53083467e-01 -1.67405069e+00
-1.68665394e-01 -6.08332574e-01 -5.87297499e-01 7.59786129e-01
5.17005026e-01 -1.04604483e+00 -3.33857387e-01 3.43099982e-02] | [6.9468488693237305, -1.1860027313232422] |
1c0621ef-7d70-445f-a804-be69fb7233b2 | a-semi-supervised-multi-task-learning | 2106.07381 | null | https://arxiv.org/abs/2106.07381v1 | https://arxiv.org/pdf/2106.07381v1.pdf | A Semi-supervised Multi-task Learning Approach to Classify Customer Contact Intents | In the area of customer support, understanding customers' intents is a crucial step. Machine learning plays a vital role in this type of intent classification. In reality, it is typical to collect confirmation from customer support representatives (CSRs) regarding the intent prediction, though it can unnecessarily incur prohibitive cost to ask CSRs to assign existing or new intents to the mis-classified cases. Apart from the confirmed cases with and without intent labels, there can be a number of cases with no human curation. This data composition (Positives + Unlabeled + multiclass Negatives) creates unique challenges for model development. In response to that, we propose a semi-supervised multi-task learning paradigm. In this manuscript, we share our experience in building text-based intent classification models for a customer support service on an E-commerce website. We improve the performance significantly by evolving the model from multiclass classification to semi-supervised multi-task learning by leveraging the negative cases, domain- and task-adaptively pretrained ALBERT on customer contact texts, and a number of un-curated data with no labels. In the evaluation, the final model boosts the average AUC ROC by almost 20 points compared to the baseline finetuned multiclass classification ALBERT model. | ['Amir Biagi', 'Matthew C. Spencer', 'Li Dong'] | 2021-06-10 | null | https://aclanthology.org/2021.ecnlp-1.7 | https://aclanthology.org/2021.ecnlp-1.7.pdf | acl-ecnlp-2021-8 | ['semi-supervised-text-classification-1'] | ['natural-language-processing'] | [ 3.33391458e-01 -8.65317136e-02 -5.90579331e-01 -9.75366831e-01
-1.03471494e+00 -6.59650624e-01 4.53591466e-01 5.24248302e-01
-4.03174669e-01 5.42421639e-01 6.27477989e-02 -6.71160102e-01
-6.82532862e-02 -5.78163087e-01 -3.50982666e-01 -3.36513847e-01
2.00902075e-01 1.14072824e+00 -2.38557197e-02 -1.96666509e-01
3.90355200e-01 6.20175190e-02 -1.14405334e+00 9.55810845e-01
1.08633041e+00 1.29303193e+00 -1.42312229e-01 4.07572031e-01
-2.75594771e-01 6.88495517e-01 -5.34942508e-01 -6.94953263e-01
1.17999896e-01 -5.80555908e-02 -7.93974459e-01 1.70374438e-01
2.74890572e-01 -4.06891972e-01 4.57575649e-01 5.32524586e-01
8.75503495e-02 -1.20013937e-01 9.26615179e-01 -1.69212782e+00
-8.01681161e-01 4.72857684e-01 -5.83465815e-01 5.53693585e-02
3.60726595e-01 -1.70096681e-01 1.43606448e+00 -8.82842898e-01
2.85404414e-01 8.39411080e-01 8.18122387e-01 1.44641384e-01
-1.20677722e+00 -6.74194753e-01 2.99764544e-01 3.49891126e-01
-1.10301268e+00 -2.49704793e-01 7.99920142e-01 -4.14062321e-01
9.94629383e-01 3.21430504e-01 2.42890626e-01 1.29885614e+00
4.82856929e-02 1.05892849e+00 1.04169321e+00 -3.76987427e-01
3.22599590e-01 8.73724997e-01 7.69732952e-01 2.35375360e-01
1.90310284e-01 -3.11623573e-01 -2.15339333e-01 -5.23049176e-01
1.07137978e-01 3.96676093e-01 1.83459148e-01 -2.74544746e-01
-7.47793674e-01 1.18763089e+00 2.12315813e-01 3.45348001e-01
-4.04429793e-01 -5.79053283e-01 5.88351488e-01 4.48400766e-01
7.08052933e-01 2.35957861e-01 -8.49122465e-01 4.36837412e-03
-7.98882544e-01 -8.72237608e-02 1.22084308e+00 9.63928342e-01
7.80605435e-01 -4.10610050e-01 7.95182362e-02 1.11453295e+00
2.25016460e-01 2.05767438e-01 6.80428445e-01 -3.32205683e-01
4.52377796e-01 1.06768799e+00 1.78625852e-01 -8.05495620e-01
-4.84922498e-01 -7.53309369e-01 -6.83216214e-01 -3.09052408e-01
3.21980745e-01 -9.79616493e-03 -6.51690125e-01 1.19341409e+00
1.42803192e-01 -2.11475864e-01 1.30162863e-02 8.65231216e-01
5.34422278e-01 4.86790985e-01 2.41836429e-01 -2.52885193e-01
1.50816810e+00 -1.05360878e+00 -6.61430538e-01 -5.41942835e-01
1.07530093e+00 -7.55849302e-01 1.16614544e+00 6.70064330e-01
-3.67957652e-01 -3.80500615e-01 -8.36553752e-01 1.47792697e-01
-4.31440890e-01 4.46408778e-01 9.46615994e-01 7.31861413e-01
-5.94389319e-01 1.42584026e-01 -2.66674399e-01 -4.72516060e-01
2.70650595e-01 3.09959859e-01 -2.78039664e-01 -3.90256912e-01
-1.23998058e+00 7.68714726e-01 3.39811295e-01 -7.43146762e-02
-3.36554170e-01 -6.10501230e-01 -6.44769371e-01 1.85166914e-02
5.41201115e-01 -3.49005222e-01 1.35085440e+00 -1.10850966e+00
-7.69786775e-01 9.02187943e-01 -2.78622061e-01 -3.66823971e-01
5.69779813e-01 -1.58429816e-02 -7.23311901e-01 -3.21072012e-01
4.53682244e-01 1.55714288e-01 7.70801723e-01 -1.31697059e+00
-1.05642664e+00 -5.47409058e-01 1.72990514e-03 2.68915053e-02
-7.26990938e-01 6.34423196e-02 -1.87780559e-01 -4.17049140e-01
1.31890953e-01 -9.27904427e-01 2.07222886e-02 -4.69540387e-01
-3.44906002e-01 -6.86970651e-01 9.16274786e-01 -4.75037694e-01
1.15598667e+00 -2.15421605e+00 -6.50910497e-01 1.79871753e-01
1.99985340e-01 2.06642270e-01 -3.04462109e-02 3.46514046e-01
-8.48830342e-02 1.58618569e-01 9.05514359e-02 -5.07585168e-01
1.05220526e-02 2.06579089e-01 -4.00795639e-01 5.47923371e-02
2.37108871e-01 8.48672926e-01 -7.55979896e-01 -5.30019760e-01
2.69303992e-02 1.28888100e-01 -6.31257415e-01 8.62200111e-02
-1.79287151e-01 1.77004710e-01 -5.69701672e-01 9.17914689e-01
5.96187770e-01 -7.62623727e-01 4.44865674e-01 -2.53512740e-01
1.70281813e-01 3.99764806e-01 -1.08461070e+00 1.10945213e+00
-7.39323914e-01 3.39311883e-02 -6.83319662e-03 -1.37463188e+00
9.28474009e-01 1.91616118e-01 5.37533164e-01 -6.21139050e-01
-9.26623940e-02 3.74896348e-01 -1.58635899e-01 -3.10358644e-01
4.24016684e-01 -4.99025851e-01 -3.94044280e-01 9.67515409e-01
-1.97270233e-02 2.70848155e-01 8.92993435e-02 1.82071149e-01
9.95662630e-01 -2.54620165e-01 5.32710552e-01 7.60491639e-02
5.05200386e-01 4.08515543e-01 6.54571116e-01 7.03182817e-01
-4.23355758e-01 3.53588670e-01 5.00009239e-01 -6.96552575e-01
-7.75184512e-01 -6.96551204e-01 -3.30083877e-01 1.78315735e+00
-2.02421114e-01 -2.70473868e-01 -1.11854203e-01 -1.17655361e+00
2.80219972e-01 1.03120625e+00 -5.47174394e-01 -1.86579943e-01
-4.45524335e-01 -8.94293368e-01 2.57158905e-01 5.63571393e-01
2.69810706e-01 -8.15186799e-01 -4.16317508e-02 3.86914819e-01
-3.85670006e-01 -1.34432769e+00 -5.28980434e-01 6.22409403e-01
-6.92903519e-01 -1.10348701e+00 -3.34057122e-01 -8.83694172e-01
5.35618126e-01 3.57002169e-01 1.03440642e+00 -7.14954659e-02
-6.84568435e-02 1.44639298e-01 -6.25769675e-01 -5.66713214e-01
-6.42487407e-01 2.51439720e-01 1.40070185e-01 4.31567043e-01
9.41715896e-01 -3.53765398e-01 -4.17344391e-01 7.38597274e-01
-8.04858446e-01 -1.13532752e-01 7.60409534e-01 8.48134935e-01
2.04198942e-01 -7.48947412e-02 1.21336281e+00 -1.25069571e+00
6.16535485e-01 -7.12900221e-01 -9.72374752e-02 4.58352029e-01
-1.11865878e+00 -3.30296546e-01 8.16847682e-01 -6.76616132e-01
-9.41835582e-01 2.95591969e-02 -9.62870475e-03 -5.29970340e-02
-3.04683983e-01 6.24996066e-01 1.48826137e-01 1.86527342e-01
7.40069687e-01 1.99140772e-01 -8.51081982e-02 -4.12070215e-01
1.71939045e-01 1.33791828e+00 1.10559573e-03 -2.77503282e-01
4.22913313e-01 2.42571190e-01 -5.44411242e-01 -5.18695593e-01
-1.22678912e+00 -1.05036640e+00 -7.03107595e-01 -1.21259153e-01
3.89904886e-01 -8.37988198e-01 -7.52280593e-01 1.39888883e-01
-9.19328690e-01 -1.02562875e-01 1.98697336e-02 3.36060673e-01
-2.52365679e-01 3.38494480e-01 -6.76603079e-01 -9.27632630e-01
-3.87110204e-01 -9.28382576e-01 1.02035511e+00 -2.20767334e-01
-6.19438648e-01 -1.16751480e+00 -2.40955144e-01 1.09955204e+00
5.55810750e-01 -2.95229763e-01 1.29090607e+00 -1.66462743e+00
-1.61174715e-01 -8.02654505e-01 -3.26600671e-01 3.44230831e-01
3.25788200e-01 -4.05890882e-01 -9.73036826e-01 -1.70838222e-01
1.42820522e-01 -8.00066948e-01 7.76630104e-01 1.34668618e-01
8.67540300e-01 -3.34839076e-01 -4.27486628e-01 6.18756451e-02
1.07437110e+00 3.63262087e-01 1.35428220e-01 4.15413916e-01
4.29885417e-01 7.57582307e-01 9.53185439e-01 5.72246611e-01
7.46659398e-01 6.62023008e-01 1.79360613e-01 -5.15112169e-02
4.92667198e-01 -1.31296128e-01 2.25349993e-01 4.16228890e-01
2.93713182e-01 -2.64491171e-01 -1.00339568e+00 2.79725045e-01
-1.83198822e+00 -6.47474229e-01 -2.72079468e-01 2.03405213e+00
7.90097415e-01 3.05052638e-01 2.37626389e-01 2.63199836e-01
7.09170640e-01 -1.77352726e-01 -6.61148071e-01 -3.63055617e-01
7.04853190e-03 -3.11271191e-01 1.97762653e-01 2.37637281e-01
-1.34332192e+00 7.84570217e-01 5.86140394e+00 7.01475441e-01
-1.06617785e+00 2.18090013e-01 9.84637737e-01 -2.35382356e-02
-4.00525719e-01 -7.55002797e-02 -9.90271807e-01 3.47027630e-01
1.05751157e+00 -5.07541327e-03 1.13252677e-01 1.29942024e+00
8.60929266e-02 -8.49017054e-02 -1.30264366e+00 7.44029224e-01
4.37782854e-01 -9.65116143e-01 6.28268495e-02 1.58921644e-01
4.86157656e-01 -2.06632599e-01 -7.51065463e-02 9.75824654e-01
2.89300263e-01 -6.83639884e-01 4.25347745e-01 -3.46350037e-02
7.11760521e-01 -4.50044036e-01 1.00847614e+00 8.33403349e-01
-8.70171309e-01 -6.39897525e-01 -4.42132503e-02 2.66205035e-02
3.63038599e-01 7.42048562e-01 -1.25295293e+00 3.47441375e-01
4.87353891e-01 5.79217851e-01 -4.17514652e-01 5.86752295e-01
2.40151674e-01 6.85298026e-01 -1.30195647e-01 -2.21601754e-01
1.88352570e-01 -3.60644668e-01 -3.55750360e-02 1.11361432e+00
4.12920862e-02 1.16519481e-01 4.67101276e-01 4.63307053e-01
-4.98706214e-02 3.99189234e-01 -4.73235130e-01 -1.44448534e-01
-2.00171806e-02 1.53874505e+00 -7.17060626e-01 -4.55463141e-01
-6.48715436e-01 1.08008313e+00 3.32356334e-01 2.72684783e-01
-7.81443775e-01 -2.92263702e-02 2.26939186e-01 2.98890471e-01
8.97805542e-02 1.15926370e-01 -5.98203123e-01 -9.28001583e-01
8.20128545e-02 -8.46643150e-01 7.15543568e-01 -4.68430340e-01
-1.68015492e+00 4.19305772e-01 -3.06197077e-01 -1.17512071e+00
-3.51664037e-01 -6.54560626e-01 -5.34527242e-01 6.45271361e-01
-1.55695009e+00 -1.40357721e+00 -2.74229288e-01 2.44613111e-01
6.62882090e-01 -2.34545290e-01 7.46849597e-01 6.04484260e-01
-4.82699215e-01 7.93073773e-01 -1.47038132e-01 5.47351576e-02
1.17404413e+00 -1.02921569e+00 8.48522186e-02 1.87696531e-01
-9.39213187e-02 7.10952401e-01 4.48888302e-01 -6.94594681e-01
-1.30410194e+00 -1.21643448e+00 1.38393712e+00 -5.33061028e-01
8.93443048e-01 -4.35626924e-01 -1.03958273e+00 1.01771104e+00
-2.46354461e-01 -2.68710166e-01 1.43095839e+00 6.54507935e-01
-4.59792376e-01 -1.68849185e-01 -1.22699714e+00 3.17002952e-01
8.04292619e-01 -4.40851301e-01 -6.06214523e-01 8.19856346e-01
6.61905468e-01 1.15654744e-01 -7.49024570e-01 5.16100466e-01
5.48561394e-01 -8.02931845e-01 7.34260976e-01 -9.73328829e-01
3.22139263e-01 2.07818612e-01 -2.16873750e-01 -1.03037655e+00
-2.98752546e-01 -5.05724102e-02 1.59744859e-01 1.17079031e+00
8.21700215e-01 -9.40038979e-01 8.56362760e-01 8.92447293e-01
-2.28689000e-01 -8.30188990e-01 -7.87068963e-01 -6.05852008e-01
-1.38711542e-01 -8.05190563e-01 3.21810782e-01 1.16274178e+00
2.77196556e-01 7.11683691e-01 -2.67644316e-01 -7.12890029e-02
4.20402467e-01 4.92628366e-01 7.25346863e-01 -1.53433037e+00
-2.42103145e-01 -1.59861356e-01 3.41730267e-02 -1.01576984e+00
1.89806923e-01 -1.08004105e+00 -2.11849838e-01 -1.47729349e+00
4.16647911e-01 -9.41822469e-01 -2.82749623e-01 7.67181635e-01
-1.05654031e-01 1.31645560e-01 -1.52014941e-01 5.51284015e-01
-9.64458406e-01 2.01387331e-01 7.16675758e-01 -7.16851726e-02
-3.73857319e-01 5.63591838e-01 -1.13430262e+00 7.05012143e-01
8.01465631e-01 -4.74405795e-01 -2.56662548e-01 9.35164317e-02
3.40088785e-01 4.92899828e-02 2.23522127e-01 -3.63834113e-01
1.94642454e-01 -2.47839138e-01 2.30099171e-01 -6.99569821e-01
2.85489351e-01 -9.31450427e-01 -1.92839041e-01 3.80136788e-01
-5.69511592e-01 -1.89822406e-01 -1.27932698e-01 6.57362580e-01
-1.80406645e-01 -3.69388580e-01 4.09399211e-01 3.51599348e-03
-6.72920704e-01 9.64962542e-02 -2.77102172e-01 -3.53575945e-02
1.12842989e+00 -2.61152595e-01 -4.63707298e-01 -5.25961161e-01
-7.56180048e-01 3.82963270e-01 1.95868239e-01 5.29625535e-01
4.60732937e-01 -1.05552208e+00 -6.52579784e-01 4.34011817e-01
7.01414227e-01 -4.54935819e-01 2.93204844e-01 9.71450686e-01
7.19900355e-02 8.59316945e-01 2.33636230e-01 -6.35556221e-01
-1.10943198e+00 6.23459697e-01 -6.19921647e-02 -5.71891785e-01
-5.95773272e-02 5.36376953e-01 1.08748131e-01 -8.08785617e-01
2.11015329e-01 -2.15027854e-01 -3.66157174e-01 4.71594989e-01
3.80375683e-01 2.36409321e-01 3.59402984e-01 -4.06194776e-01
-4.46450800e-01 1.94484651e-01 -7.15636611e-01 1.46113768e-01
1.27105558e+00 7.42965788e-02 1.13855705e-01 5.95096111e-01
1.30479038e+00 -1.36717662e-01 -6.82243764e-01 -4.20230329e-01
2.99675643e-01 -2.25967318e-01 -1.95442140e-01 -1.12294209e+00
-6.03584528e-01 6.28519177e-01 3.32786381e-01 5.48295021e-01
9.53857243e-01 3.24133933e-01 7.74098396e-01 4.88385826e-01
5.99015355e-01 -1.16497827e+00 4.02622193e-01 4.78484750e-01
8.39234948e-01 -2.01145959e+00 -2.19279885e-01 -5.20557642e-01
-1.25565565e+00 9.39374566e-01 5.66198528e-01 2.87172735e-01
6.38280332e-01 1.31430954e-01 2.14426115e-01 -5.51348701e-02
-9.44026530e-01 1.09235406e-01 2.53101617e-01 4.85224396e-01
5.80665171e-01 1.08149759e-01 -2.49560043e-01 1.04554021e+00
6.53782338e-02 4.65101115e-02 2.66605049e-01 9.10976827e-01
-5.28349161e-01 -1.09147370e+00 -1.51775539e-01 1.05987287e+00
-3.78242552e-01 -2.55930036e-01 -3.63066345e-01 5.58752120e-01
-1.38027146e-01 1.24944365e+00 2.06311867e-02 -5.09500325e-01
4.40476775e-01 4.71030056e-01 -2.75634110e-01 -8.15124691e-01
-7.37464964e-01 2.83939708e-02 3.61409813e-01 -1.72153071e-01
-3.37222889e-02 -7.36578643e-01 -1.21555018e+00 -1.21369898e-01
-4.84056145e-01 2.10330129e-01 8.00507188e-01 1.04269135e+00
4.36788291e-01 2.02304095e-01 9.56028998e-01 -3.41658950e-01
-9.83204663e-01 -1.10366118e+00 -5.08763969e-01 5.48568964e-01
2.11270764e-01 -4.66704845e-01 -7.49389172e-01 -2.62031257e-02] | [10.042876243591309, 6.113528728485107] |
2635152e-55b9-4023-9589-2a3bdefe0859 | censored-sampling-of-diffusion-models-using-3 | 2307.02770 | null | https://arxiv.org/abs/2307.02770v1 | https://arxiv.org/pdf/2307.02770v1.pdf | Censored Sampling of Diffusion Models Using 3 Minutes of Human Feedback | Diffusion models have recently shown remarkable success in high-quality image generation. Sometimes, however, a pre-trained diffusion model exhibits partial misalignment in the sense that the model can generate good images, but it sometimes outputs undesirable images. If so, we simply need to prevent the generation of the bad images, and we call this task censoring. In this work, we present censored generation with a pre-trained diffusion model using a reward model trained on minimal human feedback. We show that censoring can be accomplished with extreme human feedback efficiency and that labels generated with a mere few minutes of human feedback are sufficient. Code available at: https://github.com/tetrzim/diffusion-human-feedback. | ['Ernest K. Ryu', 'Albert No', 'Jaewoong Cho', 'Keon Lee', 'Kibeom Myoung', 'Taeho Yoon'] | 2023-07-06 | null | null | null | null | ['image-generation'] | ['computer-vision'] | [ 2.43360072e-01 6.04594588e-01 -1.67676006e-02 -2.30566934e-01
-8.26486707e-01 -7.91470230e-01 7.73017645e-01 -2.92008072e-01
-1.84724078e-01 9.16562676e-01 2.03185216e-01 -2.87367672e-01
2.51398623e-01 -4.90849942e-01 -6.83129311e-01 -6.38526738e-01
2.99319327e-01 4.01050359e-01 -5.77611476e-02 -3.39956023e-02
2.10000947e-01 2.23540485e-01 -1.20616901e+00 7.03715608e-02
1.04261971e+00 5.21867692e-01 1.96040809e-01 1.28764582e+00
2.15615779e-01 9.43416417e-01 -9.54854846e-01 -5.68235695e-01
3.39950204e-01 -9.15548205e-01 -6.32360280e-01 3.91130626e-01
3.56748164e-01 -6.60284638e-01 -2.87751049e-01 1.05050147e+00
5.97798467e-01 -8.03464875e-02 8.26049507e-01 -1.33138156e+00
-1.33154666e+00 4.58345234e-01 -5.57618082e-01 4.63565998e-02
3.45227510e-01 5.32144070e-01 7.84732342e-01 -9.07801867e-01
8.82022858e-01 9.66651022e-01 2.67564863e-01 1.20411289e+00
-1.36421728e+00 -5.78830481e-01 -1.11704744e-01 -2.90403157e-01
-1.18196630e+00 -6.47907853e-01 5.09304166e-01 -7.17137456e-01
5.79092443e-01 2.94454485e-01 5.17623365e-01 1.37268233e+00
2.18896925e-01 8.10876250e-01 1.11075950e+00 -4.13365930e-01
2.11460903e-01 2.80755728e-01 -2.80388176e-01 7.01113701e-01
1.21776596e-01 4.83443558e-01 -2.69671589e-01 -2.10612655e-01
1.01878858e+00 -3.54497992e-02 -4.27808553e-01 -2.56983817e-01
-1.13541532e+00 9.87117171e-01 4.69517231e-01 1.70082644e-01
-3.40384722e-01 3.91664267e-01 -5.48990443e-02 5.20565033e-01
6.82807386e-01 5.87228537e-01 -1.23636030e-01 -1.20716291e-02
-1.06654739e+00 1.65495187e-01 5.74703336e-01 1.00861049e+00
6.28130972e-01 3.25796306e-01 -4.11299974e-01 8.18455637e-01
-5.31264357e-02 5.60641527e-01 6.13918841e-01 -1.47448897e+00
-9.18225199e-02 2.64742170e-02 6.62309825e-01 -5.87099910e-01
6.41472405e-03 -2.91604877e-01 -8.94116282e-01 8.29659760e-01
5.14697075e-01 -6.10649288e-01 -1.31330395e+00 1.72170258e+00
7.54329981e-03 -2.08145365e-01 -2.29649451e-02 9.62294638e-01
2.18162104e-01 5.56971431e-01 -1.88818663e-01 -2.55986124e-01
6.80652201e-01 -1.29447865e+00 -6.89893842e-01 -2.07005262e-01
4.29608405e-01 -8.38981807e-01 1.13165963e+00 6.64671540e-01
-1.40748811e+00 -3.79487604e-01 -1.00961840e+00 1.82495296e-01
2.59140197e-05 2.23880544e-01 3.78687978e-01 6.48875594e-01
-1.40504384e+00 8.25163543e-01 -5.42682350e-01 -3.05323541e-01
4.97534931e-01 2.76180476e-01 -8.29246193e-02 -2.03213155e-01
-8.97359133e-01 7.14071155e-01 -2.37225443e-01 -1.54950857e-01
-1.43743587e+00 -3.33117396e-01 -4.66711879e-01 -1.87062442e-01
1.94438159e-01 -9.07153010e-01 1.70119238e+00 -1.46604300e+00
-1.46219218e+00 7.45600641e-01 -5.20791784e-02 -3.51350784e-01
1.22856963e+00 -3.99426967e-01 -1.97278753e-01 1.07860260e-01
2.42586643e-01 1.11585820e+00 1.32839203e+00 -1.53947747e+00
-4.38355297e-01 -4.80114035e-02 -8.15860182e-02 4.95947935e-02
-2.59755909e-01 -7.75638297e-02 -1.60202205e-01 -7.62730539e-01
-4.07945156e-01 -1.17993164e+00 -4.27992076e-01 2.28205949e-01
-4.42903668e-01 -8.23503882e-02 5.00065207e-01 -5.35255492e-01
1.00275922e+00 -2.18476772e+00 -6.33091033e-02 -1.46917075e-01
4.99196142e-01 1.69469610e-01 -1.96262047e-01 3.76017362e-01
-8.01040034e-04 5.96125782e-01 -3.99172187e-01 -5.00681341e-01
-8.41734782e-02 6.97718048e-03 -4.49407578e-01 2.81161249e-01
1.76873282e-01 9.49399412e-01 -1.13248301e+00 -4.25471783e-01
-8.60238075e-02 3.97455573e-01 -4.15686846e-01 3.53157014e-01
-3.26100856e-01 5.91580570e-01 -1.21099435e-01 3.74793291e-01
4.94368494e-01 -6.94391370e-01 -1.47492707e-01 6.43026352e-01
1.67769328e-01 -3.37324113e-01 -7.34534740e-01 1.53267479e+00
-2.66477346e-01 8.06110442e-01 1.49708046e-02 -5.35668284e-02
7.04959929e-01 5.36991119e-01 1.97773397e-01 -3.91139865e-01
-4.39096149e-03 4.77023751e-01 -4.39064056e-02 -1.44525766e-01
5.07124960e-01 -2.41140872e-01 2.36295968e-01 8.34078848e-01
-6.14291057e-02 -4.31515992e-01 2.09924638e-01 5.75164914e-01
1.18914020e+00 3.24028246e-02 -6.98904768e-02 2.49704123e-01
-7.40041509e-02 1.68225795e-01 3.88632715e-01 1.04963446e+00
-3.49349469e-01 1.22086251e+00 4.07825917e-01 -8.52564648e-02
-1.49753058e+00 -1.08979011e+00 2.72902340e-01 9.14309919e-01
-5.46455272e-02 -1.85818732e-01 -9.67420936e-01 -9.76749301e-01
-1.54987529e-01 7.75303245e-01 -6.94537699e-01 -1.24321036e-01
-2.08399832e-01 -4.69089270e-01 5.97676277e-01 2.30644256e-01
1.37967691e-01 -1.31347561e+00 -4.52215910e-01 8.99403021e-02
7.77484402e-02 -4.94347483e-01 -8.63311946e-01 7.88580924e-02
-8.00730109e-01 -8.19249272e-01 -1.60749912e+00 -5.64796627e-01
1.08722353e+00 3.56912255e-01 1.15357852e+00 4.35170323e-01
-4.62637953e-02 1.85312331e-01 -2.95674175e-01 -1.94875598e-01
-7.49681532e-01 -1.05508968e-01 -1.02324404e-01 -1.92065552e-01
-1.36951387e-01 -3.49545211e-01 -9.37859476e-01 3.86666983e-01
-1.05650234e+00 1.55013934e-01 5.85389912e-01 1.13586807e+00
3.57461035e-01 -2.03721195e-01 8.03218007e-01 -1.08271611e+00
1.02543485e+00 -4.50157613e-01 -5.20996749e-01 9.69202816e-02
-9.84587312e-01 -4.19883803e-02 7.06913173e-01 -5.81841469e-01
-1.06332362e+00 9.76510346e-02 1.05204880e-01 -7.03475416e-01
-9.12357271e-02 -2.34270059e-02 2.97819674e-01 1.15228489e-01
1.13524902e+00 6.37705624e-02 -8.09093192e-02 -2.52807498e-01
5.98550677e-01 6.67104542e-01 3.28112006e-01 -2.63613611e-01
6.28465950e-01 4.49398994e-01 -4.81766611e-01 -2.94831306e-01
-7.68274069e-01 8.51238593e-02 -2.81505108e-01 -3.74587744e-01
6.09337687e-01 -7.04459786e-01 -1.50373459e-01 6.11130059e-01
-1.16089547e+00 -9.34121132e-01 -6.89942539e-01 6.74394816e-02
-8.21294665e-01 2.16245383e-01 -8.02306056e-01 -9.81942594e-01
-2.30032980e-01 -9.45323348e-01 7.98528969e-01 2.90410906e-01
-4.65687901e-01 -8.77473772e-01 1.91235870e-01 1.90450415e-01
4.22430217e-01 7.87132904e-02 4.80346471e-01 -3.08666050e-01
-6.69691861e-01 -2.18345597e-01 -9.24504399e-02 6.24845207e-01
2.27124929e-01 1.92418009e-01 -9.29847956e-01 -3.08354050e-01
-1.11173362e-01 -6.21979058e-01 9.50758040e-01 4.28821117e-01
7.76929617e-01 -3.74058664e-01 -1.89628571e-01 3.80479127e-01
1.23400772e+00 4.53987494e-02 5.40425718e-01 -2.71698795e-02
5.20000815e-01 5.33095121e-01 4.64980155e-01 4.32032377e-01
1.81533431e-03 1.71770334e-01 2.82460511e-01 -2.59145916e-01
-5.17750144e-01 -6.05297506e-01 3.52248281e-01 5.16761959e-01
1.30843580e-01 -7.37564385e-01 -7.92307377e-01 7.88291454e-01
-1.88825738e+00 -8.54400039e-01 -1.70110062e-01 2.17363715e+00
9.23647046e-01 2.47137755e-01 6.41299710e-02 1.42228855e-02
8.93871427e-01 1.65514559e-01 -7.89124072e-01 -4.99911815e-01
-2.33669460e-01 -2.56416738e-01 5.17505765e-01 8.65776122e-01
-7.01302230e-01 9.07055318e-01 7.40341520e+00 8.49259913e-01
-9.09322321e-01 4.19238001e-01 9.52551126e-01 -3.01779419e-01
-7.02241719e-01 1.12347648e-01 -3.38798583e-01 5.50475359e-01
9.15413380e-01 -3.76217425e-01 4.83567864e-01 7.99619138e-01
4.08550739e-01 -2.51319468e-01 -8.38587701e-01 7.36986220e-01
1.17480839e-02 -1.07255149e+00 1.27696345e-04 3.38094532e-01
1.28152549e+00 -7.51557872e-02 3.58399719e-01 -6.44777343e-02
1.03595543e+00 -1.18473375e+00 8.57226074e-01 6.53718650e-01
9.99637902e-01 -5.86304307e-01 3.86886597e-01 6.52332604e-01
-1.74088210e-01 4.86606779e-03 -3.36775184e-01 2.52653053e-03
4.07499492e-01 9.05906320e-01 -8.09789538e-01 3.10364757e-02
1.90894902e-01 4.99918878e-01 -5.85562408e-01 1.09048784e+00
-6.96335733e-01 6.76753283e-01 9.85138211e-03 1.08931839e-01
2.88137905e-02 -9.23517421e-02 5.38442671e-01 9.13576245e-01
6.14018559e-01 5.88117056e-02 -1.19493879e-01 9.91175711e-01
-2.37878665e-01 -1.49996489e-01 -8.45634222e-01 -4.57451530e-02
1.81876272e-01 1.07289004e+00 -6.66229904e-01 -6.33325756e-01
-2.09729653e-02 1.71231771e+00 3.13481539e-01 6.30878210e-01
-7.52543092e-01 -2.32722193e-01 2.45459646e-01 1.65161282e-01
1.57878593e-01 -1.48886144e-01 -5.04740596e-01 -1.17776012e+00
-1.88919842e-01 -6.59522235e-01 1.43316820e-01 -1.37575614e+00
-1.23598874e+00 7.67440736e-01 -5.27013063e-01 -1.03401506e+00
-6.55505538e-01 -1.05094962e-01 -5.06237268e-01 9.18922365e-01
-1.18673825e+00 -7.94529855e-01 -1.54553503e-01 4.35658395e-01
6.52990341e-01 6.82768449e-02 6.13205492e-01 4.11443338e-02
-2.45916173e-01 5.13709247e-01 1.88874617e-01 -1.78775355e-01
9.83038366e-01 -1.55424976e+00 5.78366041e-01 7.69155741e-01
2.61297494e-01 3.13246489e-01 1.03566313e+00 -9.41166818e-01
-6.38373435e-01 -1.02900887e+00 1.02729130e+00 -6.30033672e-01
3.42026681e-01 -1.65460140e-01 -6.92984343e-01 6.09746933e-01
6.10213697e-01 -7.05351755e-02 4.54135269e-01 -3.58908534e-01
-1.74279034e-01 4.16481167e-01 -1.15035748e+00 7.83047438e-01
9.75173414e-01 -3.95870388e-01 -9.59996879e-02 5.39903462e-01
6.69122219e-01 -1.50711134e-01 -4.23831135e-01 -1.06481947e-01
4.25810575e-01 -1.08100069e+00 4.89215523e-01 -3.54341894e-01
7.42504001e-01 -2.27447376e-01 2.77501166e-01 -1.66313839e+00
-4.89521623e-01 -1.06293929e+00 -1.52224183e-01 1.10233891e+00
6.74356878e-01 -3.20368826e-01 9.92265224e-01 6.89453125e-01
1.81682765e-01 -5.11542499e-01 -5.14043629e-01 -1.03163815e+00
2.34192029e-01 -6.25763610e-02 3.44166338e-01 6.55963540e-01
-7.69225508e-02 3.68611395e-01 -8.57498288e-01 -2.26141959e-01
6.27626479e-01 -1.46445170e-01 7.01468945e-01 -9.05850589e-01
-3.46548378e-01 -3.65116894e-01 1.73560098e-01 -1.34037566e+00
-1.63050994e-01 -5.56302130e-01 3.04510921e-01 -1.59606242e+00
1.87127754e-01 -3.32318544e-01 -1.18837230e-01 3.85066152e-01
-3.14328283e-01 5.41249394e-01 2.99752414e-01 6.40824974e-01
-5.98306537e-01 5.75016618e-01 1.64135361e+00 -2.20802352e-02
-1.08247384e-01 4.18885648e-02 -8.25722337e-01 4.40835536e-01
1.06013715e+00 -8.00966501e-01 -4.44437325e-01 -6.24015987e-01
3.09799463e-01 2.83563286e-01 4.80344951e-01 -8.29196095e-01
2.15554923e-01 -1.97387233e-01 4.97228235e-01 -1.09050430e-01
2.43212298e-01 -4.94741350e-01 3.26903313e-01 5.25493324e-01
-6.56825006e-01 -7.93444272e-03 -3.09031636e-01 6.26309931e-01
-5.93416542e-02 -5.61299145e-01 7.79827237e-01 -3.60728323e-01
-2.11764827e-01 2.55913645e-01 -5.72266579e-01 1.93459869e-04
9.43672657e-01 -1.75170854e-01 -3.33494186e-01 -1.05204356e+00
-1.01013565e+00 1.48571342e-01 1.11078048e+00 3.01255435e-01
6.92601740e-01 -1.39972234e+00 -7.84106314e-01 3.49137150e-02
-5.36204390e-02 -2.59024769e-01 1.87633529e-01 5.17316282e-01
-2.83078790e-01 3.04248668e-02 2.66566277e-02 -3.31328243e-01
-1.00563228e+00 6.71703696e-01 3.34665835e-01 -1.85015485e-01
-4.12669629e-01 9.32705641e-01 3.50477576e-01 -6.84210360e-02
-4.52286266e-02 1.52727872e-01 2.70533502e-01 -1.06664672e-01
4.63858277e-01 2.37056345e-01 -2.02170178e-01 -3.40017378e-01
9.50410068e-02 1.72441527e-01 -1.25602752e-01 -8.33242774e-01
9.66694355e-01 -4.01802212e-01 3.92903447e-01 3.44821841e-01
8.43060791e-01 2.26201355e-01 -1.92182171e+00 1.08584881e-01
-3.10691684e-01 -6.97432458e-01 -1.02154046e-01 -1.14188504e+00
-1.14685082e+00 9.70135391e-01 4.40102905e-01 5.87834895e-01
1.01249921e+00 -2.18297392e-02 7.64084041e-01 -1.99223846e-01
2.59079993e-01 -1.06385338e+00 3.52227807e-01 2.69585043e-01
1.15235436e+00 -1.24534738e+00 -2.78500497e-01 -8.83976594e-02
-1.04084134e+00 6.24200284e-01 5.39599180e-01 -4.22616422e-01
4.31344301e-01 1.44123390e-01 4.10969198e-01 -8.52183551e-02
-1.11696303e+00 -1.11505032e-01 -1.33825704e-01 7.48919189e-01
4.28768069e-01 8.96170288e-02 -2.30140731e-01 2.72629112e-01
-2.69953519e-01 4.91972744e-01 8.97188425e-01 8.84585619e-01
-5.37096739e-01 -1.06352532e+00 -3.88450891e-01 4.56069857e-01
-5.98971963e-01 -2.62811750e-01 -7.58148789e-01 2.34550625e-01
-6.12427555e-02 1.27409613e+00 -1.24413572e-01 -2.98276156e-01
2.60678325e-02 3.41279730e-02 3.12931180e-01 -6.38023615e-01
-6.71929061e-01 2.34410122e-01 1.08668888e-02 -3.19712549e-01
-8.24142769e-02 -5.20686030e-01 -9.32895124e-01 -4.76628393e-01
-4.24404114e-01 1.68735176e-01 4.03136343e-01 2.89795071e-01
5.47392666e-01 3.39775592e-01 8.61302733e-01 -7.36661077e-01
-5.89416087e-01 -9.92414415e-01 -6.83236182e-01 5.02173662e-01
7.23619640e-01 -2.23440677e-01 -6.94576263e-01 4.48938757e-01] | [11.460121154785156, -0.22809697687625885] |
641855ed-a7ff-4cb3-bd74-89bf2728d85b | prototypical-cross-attention-networks-for | 2106.11958 | null | https://arxiv.org/abs/2106.11958v2 | https://arxiv.org/pdf/2106.11958v2.pdf | Prototypical Cross-Attention Networks for Multiple Object Tracking and Segmentation | Multiple object tracking and segmentation requires detecting, tracking, and segmenting objects belonging to a set of given classes. Most approaches only exploit the temporal dimension to address the association problem, while relying on single frame predictions for the segmentation mask itself. We propose Prototypical Cross-Attention Network (PCAN), capable of leveraging rich spatio-temporal information for online multiple object tracking and segmentation. PCAN first distills a space-time memory into a set of prototypes and then employs cross-attention to retrieve rich information from the past frames. To segment each object, PCAN adopts a prototypical appearance module to learn a set of contrastive foreground and background prototypes, which are then propagated over time. Extensive experiments demonstrate that PCAN outperforms current video instance tracking and segmentation competition winners on both Youtube-VIS and BDD100K datasets, and shows efficacy to both one-stage and two-stage segmentation frameworks. Code and video resources are available at http://vis.xyz/pub/pcan. | ['Fisher Yu', 'Chi-Keung Tang', 'Yu-Wing Tai', 'Martin Danelljan', 'Xia Li', 'Lei Ke'] | 2021-06-22 | null | http://proceedings.neurips.cc/paper/2021/hash/093f65e080a295f8076b1c5722a46aa2-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/093f65e080a295f8076b1c5722a46aa2-Paper.pdf | neurips-2021-12 | ['video-instance-segmentation', 'multiple-object-track-and-segmentation', 'multi-object-tracking-and-segmentation'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 2.01346025e-01 -3.07133555e-01 -6.21944010e-01 -2.02076137e-01
-7.37331390e-01 -4.87586051e-01 4.33638036e-01 -6.90951049e-02
-4.19328690e-01 3.39402497e-01 -2.09355295e-01 -4.10223641e-02
9.55482721e-02 -3.44439179e-01 -8.85677040e-01 -4.39460695e-01
-9.43770409e-02 5.47648847e-01 8.15219879e-01 3.96889359e-01
8.87985714e-03 4.27937835e-01 -1.57827735e+00 4.05957997e-01
4.26953673e-01 1.19786859e+00 3.97968233e-01 9.11086917e-01
-1.72959313e-01 7.63628304e-01 -3.79952073e-01 -3.03621888e-01
3.81474137e-01 -1.82993233e-01 -8.15431952e-01 2.51491249e-01
8.87398005e-01 -5.27497947e-01 -3.84699970e-01 9.89130557e-01
1.63797379e-01 2.35386550e-01 2.34336004e-01 -1.37044251e+00
-6.40444517e-01 4.87486899e-01 -8.41459334e-01 7.96692491e-01
1.54409006e-01 4.53584522e-01 9.44937766e-01 -8.19917977e-01
7.03624845e-01 1.21990705e+00 6.64391100e-01 8.89354110e-01
-1.26373231e+00 -8.30311239e-01 7.48034120e-01 3.33726406e-01
-1.22027659e+00 -4.67557907e-01 5.66534877e-01 -5.28826892e-01
8.88604283e-01 2.01623589e-01 1.04485643e+00 1.11110878e+00
-5.41414656e-02 1.53420663e+00 6.35809839e-01 -6.72686398e-02
-6.31802231e-02 -1.50774211e-01 4.57089245e-01 7.18372464e-01
9.60617736e-02 2.41253704e-01 -5.07647216e-01 1.38214931e-01
8.10864210e-01 3.27575266e-01 -1.45746559e-01 -3.91082257e-01
-1.26604199e+00 4.78711188e-01 4.30335194e-01 3.62167865e-01
-4.53621209e-01 4.58593100e-01 2.91299224e-01 -1.31982103e-01
3.04851025e-01 8.68787915e-02 -6.17059410e-01 5.06301224e-03
-1.29775155e+00 1.10349283e-01 3.78245085e-01 1.03292441e+00
7.78332472e-01 1.68763921e-02 -5.31921148e-01 4.05695319e-01
5.07444143e-01 5.54016471e-01 4.94700283e-01 -1.25086343e+00
2.49297693e-01 5.24894476e-01 1.67892456e-01 -7.33320773e-01
-2.17124805e-01 -1.24423817e-01 -3.06411207e-01 1.94223955e-01
4.93690759e-01 -1.31074578e-01 -1.20017767e+00 1.65936422e+00
7.01408803e-01 9.24285948e-01 -2.03795612e-01 1.00332582e+00
7.83722639e-01 5.41381598e-01 5.13295591e-01 -1.58925459e-01
1.22188914e+00 -1.35267138e+00 -5.37744880e-01 -4.04382050e-01
2.17320040e-01 -4.82216507e-01 6.66672349e-01 8.30506161e-02
-1.37638414e+00 -1.00619042e+00 -7.29258537e-01 -4.59224917e-02
-3.33644927e-01 3.48078869e-02 4.51005310e-01 3.67655963e-01
-1.02446485e+00 6.52110040e-01 -1.15876245e+00 -3.48801345e-01
8.03833604e-01 4.90909129e-01 1.21661173e-02 1.46018080e-02
-7.57805169e-01 4.75389749e-01 6.44106627e-01 2.66698807e-01
-1.15466547e+00 -8.02113295e-01 -6.51733756e-01 -1.14625320e-01
4.54242855e-01 -6.84589505e-01 1.35427237e+00 -1.32953537e+00
-1.19215465e+00 7.95388341e-01 -2.37253994e-01 -7.56809294e-01
6.17976725e-01 -5.75733125e-01 -5.14678240e-01 3.68217140e-01
1.99052662e-01 1.28947508e+00 1.09232950e+00 -1.21383059e+00
-1.13444448e+00 -1.63915947e-01 1.19732380e-01 1.87593430e-01
-1.34509638e-01 2.59638965e-01 -1.20546067e+00 -4.08424437e-01
-1.19813345e-02 -9.24190283e-01 -3.17182332e-01 1.98925093e-01
-5.76026559e-01 -2.06311852e-01 1.53352273e+00 -6.96583688e-01
1.24669123e+00 -2.01630735e+00 2.72213936e-01 -9.71385390e-02
3.46425086e-01 5.81498682e-01 -3.14937741e-01 -3.70118439e-01
5.66036068e-02 -7.26271123e-02 1.75859988e-01 -6.52400255e-01
-1.66804776e-01 2.20713317e-01 -8.55409801e-02 5.40061653e-01
3.36213529e-01 1.28759742e+00 -9.45868850e-01 -8.43159795e-01
5.51596761e-01 5.26906133e-01 -3.30402225e-01 1.31858364e-01
-7.58374095e-01 5.42102158e-01 -4.13957655e-01 8.82200301e-01
3.66109163e-01 -6.54149413e-01 1.32564353e-02 -3.62565845e-01
-2.38870695e-01 -2.58245915e-01 -1.06377935e+00 1.65954232e+00
1.21703506e-01 7.58249164e-01 -5.63005246e-02 -6.15193546e-01
3.39157581e-01 1.45480603e-01 9.04259980e-01 -7.24030614e-01
2.49026224e-01 -1.53550074e-01 -1.40247077e-01 -4.39823806e-01
5.47093391e-01 4.08504039e-01 1.93637371e-01 2.56302863e-01
2.07148015e-01 6.81840658e-01 5.04420161e-01 2.16050163e-01
1.00990748e+00 4.96746749e-01 -7.08761588e-02 -3.02145183e-02
4.52094078e-01 9.25920755e-02 9.26074564e-01 8.35903823e-01
-7.08857596e-01 4.90216851e-01 8.45173933e-03 -5.68974793e-01
-9.60520506e-01 -1.00792527e+00 8.98872688e-03 1.33736336e+00
5.35524249e-01 -1.28306717e-01 -5.76121926e-01 -7.70205498e-01
1.54676586e-01 5.25859356e-01 -8.01500440e-01 1.42110363e-01
-8.20467293e-01 -4.15975094e-01 3.03227156e-01 7.75441527e-01
3.55479091e-01 -1.29926658e+00 -1.08600545e+00 3.14896286e-01
-1.61010981e-01 -1.11930823e+00 -7.40992546e-01 -9.29346457e-02
-8.48266542e-01 -1.25257373e+00 -7.63279498e-01 -6.76877916e-01
3.42410982e-01 4.39519525e-01 1.07225931e+00 2.48518765e-01
-4.39005643e-01 8.74402285e-01 -1.43457487e-01 -1.93886131e-01
-2.18666300e-01 3.79162580e-02 -8.05946067e-02 2.76398718e-01
4.36311305e-01 -2.46026695e-01 -7.27476537e-01 2.40027457e-01
-5.56166887e-01 3.42719227e-01 6.23211563e-01 2.90565610e-01
8.55875075e-01 -4.69657272e-01 2.33148992e-01 -6.27308786e-01
-3.08183014e-01 -5.07096469e-01 -8.30537140e-01 5.09449720e-01
-3.34073871e-01 -4.09892678e-01 5.88221923e-02 -7.80373216e-01
-1.00976431e+00 4.58927691e-01 1.87394321e-01 -1.30281889e+00
-4.37773645e-01 8.80342815e-03 3.40189897e-02 2.94511281e-02
1.36268079e-01 2.04884350e-01 -1.72831401e-01 -4.96025890e-01
5.07590413e-01 3.29150222e-02 7.21157968e-01 -3.12746733e-01
6.00089908e-01 5.38510621e-01 -4.43393677e-01 -4.66176391e-01
-1.11777568e+00 -8.99287760e-01 -9.05351758e-01 -7.57779956e-01
1.40351295e+00 -1.02498746e+00 -7.05953479e-01 3.45626712e-01
-1.12596905e+00 -6.57692373e-01 -4.38498050e-01 2.90823817e-01
-4.96789724e-01 2.20080614e-01 -8.01793635e-01 -6.98603272e-01
-2.98816890e-01 -9.81269181e-01 1.24121416e+00 6.21579528e-01
-8.65783542e-02 -9.24809396e-01 -8.26942250e-02 3.48670989e-01
1.79920226e-01 1.82233334e-01 2.45115682e-01 -7.77758300e-01
-1.11096191e+00 -1.48064226e-01 -3.54825824e-01 -4.66784611e-02
-6.99792057e-02 3.49665999e-01 -7.54158676e-01 -4.83327031e-01
-4.89070445e-01 -1.41340673e-01 9.28696573e-01 7.87722588e-01
1.17981410e+00 -2.06281841e-01 -8.67287874e-01 5.77409029e-01
1.49142504e+00 3.84030491e-01 2.02797994e-01 1.73088089e-01
1.07808793e+00 1.48681030e-01 5.95668197e-01 3.36540341e-01
3.70372266e-01 6.91867769e-01 4.49750215e-01 -1.83491856e-01
-4.58783597e-01 -9.33187827e-03 3.26052606e-01 5.22272348e-01
-9.41863563e-03 -1.81477681e-01 -8.91587794e-01 7.26716161e-01
-2.06041431e+00 -1.39168286e+00 1.24013197e-04 1.85677218e+00
5.54873109e-01 3.03228199e-01 5.11921883e-01 -4.73608375e-01
1.09592545e+00 1.50686651e-01 -1.03373194e+00 2.71308392e-01
-1.14243455e-01 -9.36112255e-02 4.77094054e-01 3.39304239e-01
-1.56126213e+00 1.05553102e+00 6.15729475e+00 6.26807630e-01
-1.11494398e+00 2.40702853e-01 6.64244890e-01 -4.18874264e-01
2.03389570e-01 -2.40130424e-02 -1.22549260e+00 5.61481833e-01
1.03535676e+00 -5.68261892e-02 1.52900487e-01 8.64857316e-01
5.78155145e-02 -7.08230212e-02 -1.08126569e+00 7.35313416e-01
-1.70298100e-01 -1.60835350e+00 -9.37702283e-02 -1.28489375e-01
7.93198347e-01 5.13191164e-01 1.28828779e-01 3.83737296e-01
4.12700355e-01 -5.20174921e-01 1.17440474e+00 7.21498787e-01
3.06029469e-01 -2.88224339e-01 1.95328742e-01 1.00447215e-01
-1.66879570e+00 -2.95966268e-01 -2.22332459e-02 3.50356698e-01
3.03454340e-01 2.90092453e-02 -3.99039388e-01 3.60092968e-01
1.09716618e+00 1.14906895e+00 -8.20495665e-01 1.45916986e+00
2.28236601e-01 5.24313688e-01 -4.09420460e-01 3.68639886e-01
4.14258003e-01 -1.16665132e-01 5.18104255e-01 1.31614745e+00
9.09446180e-02 3.50959413e-02 6.91383243e-01 8.15196097e-01
8.98958743e-02 -2.62021661e-01 -7.98808187e-02 7.56846815e-02
4.74065185e-01 1.31108332e+00 -1.17309415e+00 -8.20017695e-01
-8.26546967e-01 9.54140306e-01 1.63970172e-01 5.54176629e-01
-1.34661627e+00 2.65681088e-01 7.19236255e-01 -1.62402600e-01
9.29706991e-01 -1.13852739e-01 -8.44226405e-03 -1.07880092e+00
-2.31352091e-01 -5.71132660e-01 7.39863038e-01 -7.60723412e-01
-1.04667628e+00 5.20747900e-01 5.69326477e-03 -1.21771455e+00
1.23218402e-01 -3.55572909e-01 -5.20016670e-01 4.71511573e-01
-1.48978186e+00 -1.42849708e+00 -3.61981988e-01 6.95637465e-01
9.63727355e-01 1.41216859e-01 3.70625496e-01 5.39437830e-01
-1.07585609e+00 3.22981626e-01 -5.32903485e-02 4.56931263e-01
3.80981147e-01 -1.08913732e+00 4.62197810e-01 1.06384635e+00
3.88215005e-01 2.04056084e-01 5.56581259e-01 -8.80238593e-01
-1.15998149e+00 -1.62345839e+00 3.11301917e-01 -6.21079087e-01
7.00933695e-01 -9.34240296e-02 -9.94228244e-01 1.12912691e+00
3.67446750e-01 5.28509378e-01 4.88704473e-01 -2.65319705e-01
-2.80980408e-01 1.67180046e-01 -7.99519002e-01 5.71355760e-01
1.10549057e+00 -2.72612542e-01 -3.28318655e-01 3.15859616e-01
8.96720886e-01 -6.35116577e-01 -8.50240588e-01 3.30561787e-01
6.44668460e-01 -8.57771039e-01 1.17587399e+00 -7.23977745e-01
2.89246067e-02 -5.28075457e-01 -1.16685539e-01 -5.03550291e-01
-4.94380653e-01 -5.26901841e-01 -7.25700796e-01 1.16307235e+00
3.07918757e-01 -1.22324917e-02 1.06378281e+00 7.76879549e-01
-1.77790388e-01 -6.72954142e-01 -8.46034586e-01 -8.83672118e-01
-2.39697307e-01 -5.08513331e-01 3.65638584e-01 6.83433533e-01
-7.04845667e-01 1.28723562e-01 -2.88800120e-01 4.75627720e-01
9.30160284e-01 5.14777482e-01 5.61108530e-01 -1.10272896e+00
-3.29705358e-01 -6.91758811e-01 -3.12593967e-01 -1.28895378e+00
2.01836929e-01 -6.63096845e-01 1.71766683e-01 -1.34208310e+00
2.51310080e-01 -3.94492149e-01 -6.28760993e-01 6.19771600e-01
-3.11740249e-01 4.05617177e-01 5.96224725e-01 4.08356309e-01
-1.56129265e+00 2.57394582e-01 1.12251854e+00 -3.46927196e-01
-3.26338708e-01 1.74457222e-01 -2.16770545e-01 6.92065775e-01
7.01414347e-01 -4.26253527e-01 -1.38243899e-01 -4.90608722e-01
-5.19317210e-01 1.64786890e-01 7.05413222e-01 -1.42307496e+00
4.97606754e-01 -3.14541698e-01 6.78741395e-01 -1.03354502e+00
5.60536385e-01 -8.43055844e-01 4.26094890e-01 6.02237940e-01
-3.22314620e-01 2.05386639e-01 5.09315491e-01 9.52148080e-01
1.64225139e-02 1.29783496e-01 8.95398676e-01 -1.91331163e-01
-1.48636580e+00 8.44325125e-01 -2.95796365e-01 -2.36550197e-02
1.38090134e+00 -5.31314790e-01 -1.20374404e-01 2.22276852e-01
-1.07872498e+00 5.71910620e-01 4.40218866e-01 6.18346810e-01
5.32659411e-01 -1.32477999e+00 -3.18175048e-01 3.14392266e-03
-5.03723025e-02 -1.80560842e-01 5.79514205e-01 9.99831259e-01
-1.25617892e-01 3.24680328e-01 -3.01016152e-01 -1.10002625e+00
-1.49827373e+00 9.72413421e-01 4.66757566e-01 1.08291442e-03
-8.01246941e-01 1.02097571e+00 1.87912762e-01 2.43147925e-01
4.46788996e-01 -3.59053463e-01 -1.19749069e-01 9.05353501e-02
5.88909030e-01 3.24958384e-01 -4.49166715e-01 -1.00664151e+00
-3.59400123e-01 6.10404253e-01 -2.79821754e-01 1.75759241e-01
1.01924837e+00 -3.58294934e-01 2.01665744e-01 5.25005043e-01
1.11835682e+00 -6.15230262e-01 -2.00654101e+00 -4.85257655e-01
2.76311040e-01 -4.89791721e-01 -1.21729627e-01 -6.16732180e-01
-1.42920804e+00 5.67372680e-01 8.38717043e-01 1.68608293e-01
1.15166032e+00 1.36820346e-01 9.98934805e-01 4.68009338e-02
1.49034172e-01 -9.88891363e-01 2.45114639e-01 2.99105793e-01
4.01819885e-01 -1.30860698e+00 -2.05868244e-01 -9.09934938e-02
-5.76141477e-01 9.21317756e-01 1.14866555e+00 -7.26899356e-02
7.23169684e-01 2.52499968e-01 1.56783611e-01 -2.18028724e-01
-8.26724291e-01 -5.82711101e-01 4.88948494e-01 4.29703891e-01
8.72108042e-02 -1.33491307e-01 3.55447918e-01 2.43718371e-01
5.57973802e-01 -5.49306534e-02 6.10231347e-02 8.33840609e-01
-4.85382289e-01 -8.85228992e-01 -3.16912383e-01 5.25038362e-01
-4.57022607e-01 2.21700162e-01 1.31962262e-02 9.35009897e-01
4.41068530e-01 6.31421745e-01 3.86584491e-01 -2.29336485e-01
9.09074023e-02 2.88969912e-02 4.69783366e-01 -4.34624612e-01
-6.52332544e-01 3.21261108e-01 -2.16210127e-01 -9.60046172e-01
-9.20468748e-01 -1.17875040e+00 -1.20362914e+00 -9.66444165e-02
-4.62834358e-01 -1.56787485e-01 2.39956856e-01 9.12047982e-01
4.90907222e-01 7.04064727e-01 -9.41116654e-04 -1.34382880e+00
1.28652304e-01 -4.85962540e-01 -2.20192894e-01 5.60558558e-01
5.60676992e-01 -7.50680923e-01 2.61885494e-01 3.53210092e-01] | [9.007563591003418, -0.21997593343257904] |
d7c40fe8-e99f-4569-ba03-b57194716d96 | learning-attention-as-disentangler-for | 2303.15111 | null | https://arxiv.org/abs/2303.15111v1 | https://arxiv.org/pdf/2303.15111v1.pdf | Learning Attention as Disentangler for Compositional Zero-shot Learning | Compositional zero-shot learning (CZSL) aims at learning visual concepts (i.e., attributes and objects) from seen compositions and combining concept knowledge into unseen compositions. The key to CZSL is learning the disentanglement of the attribute-object composition. To this end, we propose to exploit cross-attentions as compositional disentanglers to learn disentangled concept embeddings. For example, if we want to recognize an unseen composition "yellow flower", we can learn the attribute concept "yellow" and object concept "flower" from different yellow objects and different flowers respectively. To further constrain the disentanglers to learn the concept of interest, we employ a regularization at the attention level. Specifically, we adapt the earth mover's distance (EMD) as a feature similarity metric in the cross-attention module. Moreover, benefiting from concept disentanglement, we improve the inference process and tune the prediction score by combining multiple concept probabilities. Comprehensive experiments on three CZSL benchmark datasets demonstrate that our method significantly outperforms previous works in both closed- and open-world settings, establishing a new state-of-the-art. | ['Kwan-Yee K. Wong', 'Kai Han', 'Shaozhe Hao'] | 2023-03-27 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Hao_Learning_Attention_As_Disentangler_for_Compositional_Zero-Shot_Learning_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Hao_Learning_Attention_As_Disentangler_for_Compositional_Zero-Shot_Learning_CVPR_2023_paper.pdf | cvpr-2023-1 | ['compositional-zero-shot-learning'] | ['computer-vision'] | [ 2.02739537e-01 -1.18205078e-01 -1.84843391e-01 -3.34299088e-01
-4.23166543e-01 -7.56185472e-01 7.88990200e-01 1.48957744e-01
-2.02507719e-01 3.75054330e-01 3.48239928e-01 9.32112783e-02
-1.31616876e-01 -8.08507919e-01 -7.88159430e-01 -1.04769790e+00
1.30661935e-01 4.22505409e-01 -2.45438889e-01 6.39769658e-02
5.43335592e-03 3.27565819e-01 -1.75266206e+00 4.05901462e-01
8.75616848e-01 1.05841064e+00 2.50499696e-01 3.99794966e-01
-2.18417346e-01 6.65928066e-01 -3.48981589e-01 -5.87738156e-01
2.35335365e-01 -1.97031394e-01 -5.65549076e-01 3.00477147e-01
6.18256927e-01 -9.01195556e-02 -9.22799334e-02 1.15014708e+00
2.00920641e-01 2.16673940e-01 8.94882619e-01 -1.59945381e+00
-1.49616098e+00 6.73485041e-01 -6.37014329e-01 -1.25524819e-01
8.38668197e-02 3.93103689e-01 1.80578911e+00 -1.15573084e+00
4.60583657e-01 1.38455260e+00 5.51044792e-02 7.03948140e-01
-1.75389314e+00 -9.90516782e-01 5.93219936e-01 5.23802102e-01
-1.47723758e+00 -1.86687559e-01 8.84955704e-01 -6.97705209e-01
4.61796910e-01 3.88755858e-01 7.36651659e-01 1.24971569e+00
-3.54892135e-01 1.05421174e+00 7.73563385e-01 -2.86567599e-01
2.95347214e-01 1.48385823e-01 -1.98893696e-01 6.61901772e-01
3.25120628e-01 2.53393184e-02 -5.98860919e-01 4.59824093e-02
4.58772868e-01 3.32215130e-01 -2.67621189e-01 -9.45967555e-01
-1.44040883e+00 8.68835986e-01 8.41518641e-01 -1.46760792e-02
1.95948463e-02 8.41504931e-02 1.93317860e-01 1.06710317e-02
3.63851607e-01 6.75351977e-01 -3.29869866e-01 4.67990637e-01
-2.97180623e-01 1.13279760e-01 4.81899381e-01 1.08930755e+00
8.91860008e-01 -1.09456047e-01 -4.34887737e-01 5.56279302e-01
4.56863493e-01 3.84126872e-01 2.76419163e-01 -5.40844798e-01
4.12694961e-01 8.98009419e-01 4.85007875e-02 -6.95213914e-01
8.01616907e-02 -2.82103688e-01 -8.64756048e-01 3.71956736e-01
2.06038505e-01 1.65173374e-02 -1.08605540e+00 2.03745270e+00
3.19263786e-01 5.99550486e-01 1.57966480e-01 1.16895533e+00
7.88958490e-01 4.86790240e-01 4.19968665e-01 2.05877960e-01
1.73867548e+00 -1.12296844e+00 -5.26951134e-01 -1.83427453e-01
2.25518569e-01 -3.92532468e-01 1.29432404e+00 6.97207078e-02
-6.09593332e-01 -6.08400285e-01 -1.28015316e+00 -3.44997197e-01
-5.65989554e-01 2.94381499e-01 6.76590145e-01 2.02942207e-01
-1.93450257e-01 5.47778130e-01 -8.01168323e-01 3.67742628e-02
6.74207509e-01 1.40783906e-01 -5.41042864e-01 -1.75183907e-01
-1.03781867e+00 6.24722421e-01 4.37964201e-01 -9.84471012e-03
-1.22875333e+00 -1.12219131e+00 -1.16010189e+00 5.13463557e-01
7.43197620e-01 -8.13849330e-01 7.21459568e-01 -1.18345058e+00
-1.15662491e+00 8.28817964e-01 7.44693875e-02 -1.01733625e-01
1.36103764e-01 -2.37681121e-01 -4.58795726e-01 4.42124419e-02
4.44544069e-02 7.10015595e-01 1.05019128e+00 -1.50641215e+00
-5.41550756e-01 -4.26410347e-01 6.80531114e-02 3.46221656e-01
-3.33744407e-01 -3.52622896e-01 -1.00275673e-01 -7.01878488e-01
5.89276999e-02 -7.67934442e-01 2.59550035e-01 6.41909540e-01
-5.51282108e-01 -2.33360529e-01 7.34409094e-01 -2.45270699e-01
6.40955687e-01 -2.44216776e+00 7.06942737e-01 -1.39929578e-01
6.52817130e-01 4.65299701e-03 -3.65736902e-01 1.58578381e-01
-2.90094227e-01 4.08540405e-02 -3.18058521e-01 -4.62852746e-01
2.64207065e-01 3.09334993e-01 -3.53626519e-01 3.92985582e-01
8.64590704e-01 1.01624024e+00 -1.18143427e+00 -1.72586516e-01
1.98789209e-01 5.73419809e-01 -4.20342773e-01 4.32802618e-01
-5.70975244e-01 4.14866298e-01 -3.16057593e-01 5.43322563e-01
5.54931641e-01 -4.13673282e-01 2.36619338e-01 -4.53196764e-01
1.44333929e-01 7.29448944e-02 -1.01532733e+00 1.90333974e+00
-4.11148399e-01 6.71808898e-01 -3.01696837e-01 -8.47434044e-01
8.45245242e-01 2.41079003e-01 1.17470011e-01 -3.63261312e-01
7.42859915e-02 -1.59020662e-01 1.37342975e-01 -5.65371931e-01
1.33314610e-01 -3.93624008e-01 1.42108932e-01 3.15689385e-01
4.13175523e-01 2.61579186e-01 -1.06357701e-01 3.04367781e-01
5.33692122e-01 2.92673916e-01 3.86012286e-01 -4.64859515e-01
4.36560929e-01 -4.99232918e-01 6.28716648e-01 1.24595106e-01
5.09893745e-02 6.52759254e-01 7.36780524e-01 -3.60605210e-01
-9.95513916e-01 -1.63211131e+00 1.35678217e-01 1.24479640e+00
5.18023610e-01 -3.77517700e-01 -2.52829164e-01 -8.12763929e-01
3.82847488e-01 8.25103045e-01 -1.19027340e+00 -4.79455978e-01
-1.69284001e-01 -6.06667817e-01 1.16430774e-01 7.71707535e-01
4.65865657e-02 -7.88036764e-01 -4.28583086e-01 -2.43159249e-01
-4.81215455e-02 -9.70018923e-01 -6.42583728e-01 1.89749077e-01
-3.15017790e-01 -1.18176734e+00 -5.35905600e-01 -6.86376512e-01
8.22311342e-01 2.18009040e-01 1.15994143e+00 -1.52396798e-01
-6.36730373e-01 2.90644020e-02 -2.95085073e-01 -2.57889181e-01
1.41269088e-01 -9.82353166e-02 -4.28101234e-02 6.67414308e-01
6.39657915e-01 -6.91451848e-01 -7.27566838e-01 -7.85207283e-03
-8.27472866e-01 3.02744776e-01 6.23712003e-01 9.50998724e-01
5.15866101e-01 -3.49022031e-01 4.41927314e-01 -8.63065600e-01
2.33878449e-01 -7.26568103e-01 -6.24158204e-01 6.56726003e-01
-5.08955300e-01 6.07619703e-01 5.85215509e-01 -8.58200848e-01
-1.04373574e+00 -2.02787966e-02 4.26550388e-01 -8.71654630e-01
-2.22977281e-01 1.47269042e-02 -9.72515643e-01 4.69704539e-01
3.21770906e-01 1.26799658e-01 -1.92641139e-01 -5.85510373e-01
8.71593416e-01 1.03778027e-01 4.55817640e-01 -7.19052672e-01
8.32187116e-01 7.29254007e-01 -2.83670891e-02 -4.85184044e-01
-1.03613436e+00 -3.68428648e-01 -7.04427063e-01 1.90391600e-01
1.27588248e+00 -1.14599812e+00 -1.11163735e+00 8.36203992e-02
-9.77656126e-01 1.24593705e-01 -3.92776072e-01 3.46946567e-01
-3.28443468e-01 3.26742306e-02 -1.77131012e-01 -6.63901865e-01
-1.10050514e-01 -1.01640522e+00 1.16887999e+00 2.59685189e-01
-1.19905197e-03 -8.81660640e-01 -1.67529639e-02 7.13315904e-02
-4.90318686e-02 4.40544277e-01 1.39178288e+00 -6.69337213e-01
-8.85680437e-01 1.38008922e-01 -5.94642580e-01 1.93405002e-01
2.94839710e-01 9.05537680e-02 -1.10428917e+00 -2.26500541e-01
-2.85248339e-01 -3.86246711e-01 1.05798483e+00 -1.73514441e-01
1.22664809e+00 -3.95832837e-01 -2.35695556e-01 7.29506195e-01
1.53704476e+00 -1.36853144e-01 3.90705943e-01 -1.56230330e-01
1.22402155e+00 7.16784060e-01 4.19272572e-01 4.28718090e-01
4.18044597e-01 5.25298774e-01 6.09532714e-01 1.84225872e-01
-1.94786534e-01 -7.27307618e-01 1.91779360e-02 5.66783488e-01
7.65484422e-02 -2.17908755e-01 -4.99348789e-01 4.36652988e-01
-1.81460536e+00 -8.41415703e-01 2.48289794e-01 2.12563348e+00
7.68864632e-01 -1.86797991e-01 -1.22221038e-01 -2.61388928e-01
7.17013538e-01 2.70633101e-01 -8.67226839e-01 1.94019750e-01
-6.04484081e-02 4.18434925e-02 2.48960238e-02 3.96810651e-01
-1.01144111e+00 9.26807523e-01 4.24572086e+00 5.97865939e-01
-1.00578606e+00 1.47014767e-01 4.04879004e-01 -4.44995075e-01
-7.30955064e-01 2.41737843e-01 -6.59154534e-01 5.36757112e-01
1.28723592e-01 -1.73066363e-01 4.41850990e-01 7.51263738e-01
-4.67716962e-01 2.27903649e-01 -1.57322919e+00 8.71695161e-01
3.53147119e-01 -1.15231466e+00 4.17850465e-01 -1.38545021e-01
6.80886745e-01 -3.83497924e-01 3.33593071e-01 4.02575493e-01
4.12377059e-01 -1.12590683e+00 7.65850186e-01 6.97299600e-01
8.62020254e-01 -6.50505424e-01 1.68345347e-01 3.19184363e-02
-1.51316035e+00 -1.42289236e-01 -4.96737361e-01 7.02980673e-03
-9.30769518e-02 2.39505395e-01 -3.73263299e-01 5.83266377e-01
4.47946519e-01 8.87051582e-01 -4.73609775e-01 7.57553756e-01
-6.27223432e-01 3.36404890e-02 1.98596716e-01 -4.73927706e-02
4.16764617e-02 -3.76944005e-01 5.23627698e-01 7.70401180e-01
5.46142682e-02 1.07990749e-01 -4.87537077e-03 1.55391014e+00
-3.89825284e-01 -2.01274544e-01 -2.88481504e-01 -2.68463016e-01
4.25446540e-01 1.36694467e+00 -3.34955752e-01 -4.45688903e-01
-4.22727525e-01 1.20329320e+00 7.04566121e-01 6.09153271e-01
-7.63298750e-01 -4.27249193e-01 1.38190603e+00 -2.41337731e-01
6.82068110e-01 -3.79159711e-02 -1.97340384e-01 -1.61306250e+00
-6.49438202e-02 -5.80021560e-01 2.28937119e-01 -8.62340868e-01
-1.64278734e+00 5.62298894e-01 -1.16676658e-01 -1.30626595e+00
2.23081067e-01 -7.88321376e-01 -7.19388306e-01 9.21821475e-01
-1.41090524e+00 -1.57208383e+00 -4.23222363e-01 3.46803248e-01
5.71509719e-01 -1.19189613e-01 9.55964148e-01 1.33884221e-01
-5.92150331e-01 5.86859345e-01 1.22943752e-01 1.28014430e-01
7.26765871e-01 -1.52960372e+00 3.86559844e-01 6.72299147e-01
6.17578328e-01 6.81995749e-01 6.02585196e-01 -4.13217902e-01
-1.44008100e+00 -1.07374167e+00 5.04838765e-01 -6.58802867e-01
8.67340267e-01 -1.11168361e+00 -1.02450621e+00 6.62685215e-01
1.80046707e-01 4.35195535e-01 9.73243892e-01 2.61915922e-01
-1.22163534e+00 -2.32549012e-01 -7.39801347e-01 8.21604073e-01
9.75217164e-01 -8.79194081e-01 -6.19023681e-01 -2.84611359e-02
1.18856370e+00 -5.31942758e-04 -7.07897365e-01 2.74725676e-01
8.67142916e-01 -6.17050052e-01 1.07731497e+00 -1.09187067e+00
7.51209378e-01 -3.79629970e-01 -2.53013492e-01 -1.49052370e+00
-4.66815323e-01 -6.37355298e-02 -4.90081280e-01 1.31523740e+00
4.80697423e-01 -2.96888351e-01 5.81871510e-01 7.02642679e-01
1.76902294e-01 -7.39232659e-01 -6.31646574e-01 -6.92419589e-01
7.11553395e-02 -1.39044449e-01 9.42100346e-01 1.07668591e+00
-1.13213949e-01 7.56920457e-01 -4.74329740e-01 4.13527161e-01
8.06242645e-01 5.99200845e-01 3.60501379e-01 -1.35636473e+00
-5.41003466e-01 -3.89469862e-01 -5.10599375e-01 -8.19557607e-01
4.05305713e-01 -9.55488145e-01 1.51143074e-02 -1.21780360e+00
5.42974651e-01 -3.11884552e-01 -5.32419741e-01 4.42024052e-01
-6.45345926e-01 9.82623845e-02 4.81874168e-01 -8.96970481e-02
-5.50492048e-01 1.09118116e+00 1.34834802e+00 -5.78646839e-01
2.41611257e-01 -2.74014652e-01 -9.56187010e-01 4.77733284e-01
5.02433240e-01 -3.88673335e-01 -5.52296758e-01 -6.52147591e-01
3.41613770e-01 -4.60473478e-01 6.67766869e-01 -5.64031422e-01
6.07191287e-02 -1.81062192e-01 4.99087185e-01 -2.25076914e-01
6.08784974e-01 -1.05678713e+00 -9.10624564e-02 2.34611481e-01
-5.79532981e-01 -2.51606494e-01 -2.92508826e-02 1.01696718e+00
2.59521957e-02 -1.49929345e-01 6.73550069e-01 -3.64128686e-02
-7.32022345e-01 5.78715920e-01 1.57828256e-01 1.53063551e-01
1.20104396e+00 1.50143653e-01 -5.10046601e-01 2.00235024e-01
-7.71199167e-01 3.51280987e-01 4.95133758e-01 8.10727179e-01
7.31536090e-01 -1.73185706e+00 -5.97240984e-01 6.01538002e-01
8.33153009e-01 7.46403411e-02 3.47468674e-01 5.74091315e-01
-3.33582461e-02 -8.03489536e-02 -3.84992212e-01 -3.83054644e-01
-1.22773397e+00 1.12347209e+00 4.61639911e-02 2.80074030e-01
-5.60189009e-01 1.12885034e+00 9.16389346e-01 -3.53185445e-01
3.30538183e-01 -2.50829041e-01 -7.37684742e-02 2.04022586e-01
6.44205630e-01 1.13255039e-01 -4.50294644e-01 -4.79057342e-01
-2.95282632e-01 4.21152890e-01 -2.35913843e-01 4.67998581e-03
1.14468968e+00 -7.10989162e-02 -5.39095141e-02 8.46108496e-01
1.51116586e+00 -8.78277346e-02 -1.67985606e+00 -4.86026704e-01
6.36823401e-02 -7.64279425e-01 -2.77841538e-01 -6.79748654e-01
-1.11453569e+00 1.33405280e+00 6.35436296e-01 -1.10031068e-02
9.63018596e-01 3.75502795e-01 2.82726318e-01 1.78279161e-01
-1.67539462e-01 -5.55410326e-01 3.30331892e-01 2.99819931e-02
9.18268144e-01 -1.47331560e+00 -1.98461220e-01 -3.80296439e-01
-9.55602527e-01 9.14392591e-01 8.64644170e-01 -2.51942873e-01
5.93422055e-01 3.61304767e-02 -1.89034611e-01 -2.11811900e-01
-1.05501819e+00 -4.03800219e-01 8.14698219e-01 4.91338968e-01
3.08789521e-01 5.03312409e-01 2.25423679e-01 6.84007525e-01
1.89370245e-01 -5.19555390e-01 4.18670923e-02 3.95022541e-01
-2.70465344e-01 -9.37870622e-01 4.71994057e-02 2.31777206e-01
2.11400148e-02 -3.21124464e-01 -4.40963924e-01 4.16753709e-01
4.24927503e-01 5.32704055e-01 4.21270877e-01 -3.10488701e-01
8.50391015e-02 6.93481341e-02 5.72065532e-01 -8.52334499e-01
-1.42743573e-01 -2.84355581e-01 -3.64937931e-01 -2.70104378e-01
-2.96261221e-01 -4.44420666e-01 -1.01720178e+00 -5.48247397e-02
-3.68844986e-01 -4.10773940e-02 4.67405081e-01 9.55619633e-01
3.59544307e-01 6.95404589e-01 5.78518808e-01 -5.49954593e-01
-4.28330630e-01 -8.40295613e-01 -6.83804035e-01 7.33419657e-01
5.17870784e-01 -9.73352134e-01 -2.66016096e-01 1.32216513e-01] | [10.219938278198242, 2.2641961574554443] |
1df5a644-2f16-4dc8-8f0a-5f3ca0bea1b8 | entity-based-de-noising-modeling-for | null | null | https://aclanthology.org/2022.sigdial-1.40 | https://aclanthology.org/2022.sigdial-1.40.pdf | Entity-based De-noising Modeling for Controllable Dialogue Summarization | Although fine-tuning pre-trained backbones produces fluent and grammatically-correct text in various language generation tasks, factual consistency in abstractive summarization remains challenging. This challenge is especially thorny for dialogue summarization, where neural models often make inaccurate associations between personal named entities and their respective actions. To tackle this type of hallucination, we present an entity-based de-noising model via text perturbation on reference summaries. We then apply this proposed approach in beam search validation, conditional training augmentation, and inference post-editing. Experimental results on the SAMSum corpus show that state-of-the-art models equipped with our proposed method achieve generation quality improvement in both automatic evaluation and human assessment. | ['Nancy Chen', 'Zhengyuan Liu'] | null | null | null | null | sigdial-acl-2022-9 | ['abstractive-text-summarization'] | ['natural-language-processing'] | [ 4.47817892e-01 8.16040874e-01 -3.60171981e-02 -2.57101178e-01
-9.91425216e-01 -3.43288124e-01 9.37788546e-01 4.44813579e-01
-4.11172628e-01 1.63842368e+00 1.12568247e+00 -7.20134676e-02
1.67153612e-01 -6.47308350e-01 -5.19751906e-01 -1.51619196e-01
1.64401457e-01 8.00714552e-01 -1.82145238e-01 -5.56430697e-01
5.05927503e-01 -4.94504459e-02 -9.24472332e-01 6.64338112e-01
1.69216776e+00 3.96029115e-01 -7.76897892e-02 1.09501469e+00
-3.16819362e-02 1.25102830e+00 -1.49186468e+00 -9.93534088e-01
-4.73111749e-01 -9.50468063e-01 -1.25446963e+00 -6.64469898e-02
7.07824051e-01 -4.01520997e-01 -1.95789427e-01 9.52120662e-01
1.08527553e+00 5.36689281e-01 1.00552583e+00 -6.85552537e-01
-8.52306902e-01 1.25715911e+00 -2.11512163e-01 3.20278108e-01
8.50588381e-01 1.23679720e-01 1.05262709e+00 -5.21030605e-01
6.87608600e-01 1.30492198e+00 6.04974687e-01 1.01235163e+00
-1.21114159e+00 -3.08964968e-01 -1.94788333e-02 1.14798211e-01
-8.19804251e-01 -8.52129996e-01 4.64959770e-01 -6.37641456e-03
1.37828827e+00 5.41083395e-01 3.69922519e-01 1.65171552e+00
2.73376793e-01 9.89839435e-01 6.47697449e-01 -2.59753615e-01
1.20716147e-01 1.23459376e-01 2.28974968e-01 3.85949552e-01
3.71209294e-01 -4.44437027e-01 -8.51183116e-01 -2.25907326e-01
1.26065359e-01 -8.66345525e-01 -6.59791708e-01 6.45505786e-01
-1.41484106e+00 6.78207099e-01 7.06810653e-02 1.68335065e-01
-9.11096096e-01 -1.40334982e-02 6.57492936e-01 9.06526595e-02
6.86614454e-01 1.12880802e+00 -2.50546098e-01 -3.95294040e-01
-1.40746582e+00 6.42437220e-01 1.26841414e+00 9.42994654e-01
-7.59046897e-02 4.65326667e-01 -1.24153006e+00 9.74592745e-01
-1.39160126e-01 4.16937470e-01 8.09829056e-01 -9.35982525e-01
8.45948339e-01 3.01156044e-01 2.21112266e-01 -9.21452284e-01
-4.51654315e-01 -8.71101856e-01 -1.51414585e+00 -4.08523649e-01
1.05207441e-02 -6.24485910e-01 -4.78880167e-01 1.87042367e+00
1.21915407e-01 2.08484046e-02 6.54529750e-01 4.06334758e-01
1.28877044e+00 8.13816786e-01 2.31464565e-01 -8.12835991e-01
1.07882392e+00 -1.21923602e+00 -1.29928601e+00 -2.24554867e-01
5.01299322e-01 -4.49991643e-01 7.74814606e-01 2.72420257e-01
-1.73696041e+00 -5.05872428e-01 -9.19519484e-01 -2.10413173e-01
-6.06772071e-03 2.36919165e-01 2.32957929e-01 3.37406874e-01
-1.02848911e+00 8.51217568e-01 -5.89157939e-01 -3.51813495e-01
3.85310560e-01 -1.39170550e-02 -1.95867926e-01 6.25423908e-01
-1.58812547e+00 1.37871349e+00 8.05967450e-01 5.26596121e-02
-5.78741491e-01 -6.56901300e-01 -8.43460798e-01 3.04936439e-01
1.10438719e-01 -1.44806480e+00 1.68912101e+00 -2.99033433e-01
-1.80185544e+00 5.83925009e-01 -2.42131650e-01 -1.10878801e+00
5.91300130e-01 -6.45840645e-01 -4.81222749e-01 2.77421661e-02
1.45938307e-01 5.25757551e-01 5.81577897e-01 -1.05475378e+00
-3.99612784e-01 -5.78298569e-02 -2.51871735e-01 7.84940302e-01
-3.26416343e-01 -1.22040533e-01 2.64829755e-01 -7.74750888e-01
-3.99211228e-01 -3.69395852e-01 -2.42171139e-01 -1.03124821e+00
-1.34592617e+00 -6.03805304e-01 2.05890894e-01 -1.16643190e+00
1.90777421e+00 -1.35039592e+00 4.50367570e-01 -4.67525929e-01
1.08854465e-01 5.72688580e-01 3.86184379e-02 7.77468622e-01
1.57926992e-01 2.32200623e-01 -4.43405032e-01 -6.43727183e-01
1.81273684e-01 -2.00822100e-01 -7.34896183e-01 -5.80788068e-02
3.57932836e-01 1.02983367e+00 -1.18906713e+00 -6.01144850e-01
-1.20870955e-01 1.17356032e-01 -4.26418364e-01 5.76854408e-01
-4.41043317e-01 5.73046625e-01 -2.16052726e-01 1.89257309e-01
2.16234550e-01 -1.43919796e-01 6.26116479e-03 -4.92241606e-02
9.99978110e-02 7.96214521e-01 -7.58199573e-01 1.89758384e+00
-2.97968745e-01 4.38468039e-01 -4.18708205e-01 -7.08895743e-01
7.59980679e-01 5.97925365e-01 -4.10077602e-01 -4.34076041e-01
5.88799231e-02 1.74513068e-02 -1.07268944e-01 -5.60929060e-01
1.43590808e+00 -1.27374694e-01 -1.55694604e-01 6.10158980e-01
3.52296889e-01 -3.63292783e-01 5.66885769e-01 7.11809397e-01
9.66049731e-01 3.58177982e-02 7.87192404e-01 5.27338013e-02
6.37403429e-01 -8.60183239e-02 3.58576596e-01 1.08311892e+00
1.12460785e-01 6.39694333e-01 5.82614779e-01 2.34139696e-01
-8.31836164e-01 -7.36872256e-01 6.74447417e-02 1.09456456e+00
-4.16638702e-01 -6.15675092e-01 -1.23548877e+00 -6.92355037e-01
-4.22542900e-01 1.89710510e+00 -5.38472533e-01 -5.03323555e-01
-5.48279583e-01 -8.58671308e-01 1.11306870e+00 4.63780671e-01
7.30044007e-01 -1.48913956e+00 -5.35579681e-01 5.78666031e-01
-7.13864982e-01 -9.25846517e-01 -3.36884618e-01 -2.34472737e-01
-8.80869567e-01 -6.46712959e-01 -7.71988869e-01 -4.64992434e-01
3.92970383e-01 -5.97825170e-01 1.50102782e+00 -9.06661451e-02
3.86671096e-01 1.12273537e-01 -3.14996034e-01 -5.26565313e-01
-1.11721718e+00 5.48431635e-01 1.00122094e-01 -3.08865219e-01
-1.52422622e-01 -6.50349140e-01 -4.22105551e-01 -4.52062309e-01
-6.35311484e-01 3.80109966e-01 5.83863139e-01 1.07788789e+00
7.54188448e-02 -4.25620914e-01 1.20760441e+00 -1.35383487e+00
1.65151620e+00 -4.80877042e-01 2.50474572e-01 5.98647237e-01
-5.06869614e-01 1.80673003e-01 7.28720725e-01 -1.83826491e-01
-1.62997663e+00 -7.59420335e-01 -3.35729510e-01 2.29673326e-01
-1.53232038e-01 6.46108508e-01 1.79672204e-02 9.12166417e-01
1.17665100e+00 6.62515163e-01 -3.56922090e-01 -2.42866799e-01
5.96695483e-01 8.76233220e-01 1.19466758e+00 -4.73382086e-01
4.18069601e-01 -1.92442685e-01 -2.95753777e-01 -8.68955791e-01
-1.30867863e+00 -3.88346128e-02 -3.99790496e-01 -3.51931266e-02
9.05703008e-01 -8.22991312e-01 -4.66571242e-01 3.76021087e-01
-1.81195474e+00 -1.75076604e-01 -4.93984044e-01 6.54174015e-02
-7.08056331e-01 5.61026633e-01 -6.74725533e-01 -9.27629650e-01
-1.35911345e+00 -3.34847599e-01 9.70772386e-01 4.86653388e-01
-9.71085131e-01 -1.11244071e+00 5.98674834e-01 4.41246748e-01
5.43484569e-01 3.78418803e-01 8.10630858e-01 -1.32088149e+00
6.14832900e-02 -2.21850455e-01 2.04578802e-01 4.44632590e-01
-5.00520431e-02 -1.57041162e-01 -8.04501593e-01 -2.67680716e-02
-3.83070577e-03 -6.59966290e-01 8.79609525e-01 2.29926229e-01
8.52790952e-01 -9.91321027e-01 -6.47397945e-03 1.62090138e-01
5.83971679e-01 -2.74336129e-01 6.84480011e-01 -1.05433449e-01
5.36950648e-01 6.35989606e-01 5.58248162e-01 7.41439581e-01
5.18017888e-01 4.34732199e-01 1.70261655e-02 2.60010779e-01
-3.16015124e-01 -5.75459778e-01 4.26603794e-01 1.07081413e+00
-2.66597420e-01 -9.40621614e-01 -5.59695840e-01 5.47843039e-01
-2.02045751e+00 -1.53913033e+00 -9.55388248e-02 1.89380455e+00
1.62478411e+00 3.70241478e-02 -2.49435976e-01 -1.57987654e-01
7.69159853e-01 3.96947980e-01 -4.61818665e-01 -6.63678944e-01
-4.99816895e-01 2.49998838e-01 -1.71113595e-01 5.84638298e-01
-8.88707459e-01 1.06454110e+00 6.03162193e+00 8.46150279e-01
-6.30199432e-01 1.51941240e-01 5.89365721e-01 -1.57833528e-02
-2.48861849e-01 -3.86539578e-01 -8.01788568e-01 3.44214827e-01
1.32048619e+00 -7.33853221e-01 4.18301597e-02 4.42091793e-01
3.17892879e-01 -1.13211609e-01 -1.23077977e+00 6.18571877e-01
5.12767792e-01 -1.49242902e+00 4.70908105e-01 -5.39063871e-01
1.11398506e+00 -3.81032705e-01 -2.41443530e-01 6.74304068e-01
5.20862758e-01 -1.13882589e+00 6.12079382e-01 8.42034698e-01
6.43213153e-01 -5.97088158e-01 7.03000844e-01 7.66421795e-01
-2.52372950e-01 3.51969272e-01 -3.73678178e-01 -1.97400004e-02
6.23008072e-01 3.22653413e-01 -1.34720480e+00 7.96031654e-01
-1.13659941e-01 5.57507157e-01 -6.15440845e-01 8.57863724e-01
-8.25983763e-01 7.45136321e-01 1.62978455e-01 -3.06758016e-01
-9.11672562e-02 1.39014825e-01 1.05556643e+00 1.63015866e+00
1.74493060e-01 2.01838374e-01 -2.52699375e-01 9.64733481e-01
-5.78907073e-01 1.80776507e-01 -4.97391909e-01 -1.35958865e-01
5.66205025e-01 1.11542797e+00 -4.64764133e-04 -9.25986350e-01
3.23962152e-01 1.38956332e+00 6.75940812e-01 3.84580195e-01
-6.35709643e-01 -4.08910930e-01 6.82201982e-02 -2.86065906e-01
-8.20255131e-02 2.40516320e-01 -3.48917931e-01 -1.25270808e+00
-1.69831768e-01 -1.02064824e+00 3.95333380e-01 -8.17603707e-01
-1.05405319e+00 8.59503806e-01 -4.62842965e-03 -8.25665176e-01
-1.00948691e+00 1.86337590e-01 -1.09617054e+00 7.33961523e-01
-1.28374755e+00 -1.00675893e+00 -8.00428353e-03 2.42154464e-01
9.24714804e-01 -2.82478541e-01 1.27145875e+00 -1.75370559e-01
-7.15529978e-01 8.40166271e-01 -7.64898881e-02 -8.71264860e-02
8.94094527e-01 -1.69073820e+00 4.63761687e-01 9.62590694e-01
-1.18054608e-02 7.93540895e-01 1.23412287e+00 -9.93018746e-01
-7.16838717e-01 -1.41015160e+00 1.37271881e+00 -5.56310177e-01
3.52089673e-01 5.05298935e-02 -1.00291538e+00 7.38374293e-01
1.12508130e+00 -8.85747433e-01 7.64238536e-01 1.60058692e-01
4.51998971e-02 4.83157665e-01 -1.02646863e+00 8.00633371e-01
1.05606902e+00 -2.25339636e-01 -1.34829843e+00 9.09201920e-01
8.48171771e-01 -7.89455533e-01 -7.14722872e-01 3.16346467e-01
2.65132505e-02 -9.13664937e-01 5.90427399e-01 -1.28879178e+00
9.61582363e-01 -4.24763560e-02 2.24236161e-01 -2.00818181e+00
-1.32397205e-01 -1.13959956e+00 -6.96371973e-01 1.50280535e+00
7.39788055e-01 -2.67544657e-01 3.25016350e-01 5.85706174e-01
-6.21474266e-01 -5.66080034e-01 -6.86779499e-01 -4.20397252e-01
1.97683483e-01 4.37073074e-02 3.37562025e-01 9.11619008e-01
4.50081229e-01 1.23394907e+00 -7.16693938e-01 -1.11808687e-01
4.62207019e-01 -9.30044800e-02 8.36550355e-01 -9.42324519e-01
-2.20114037e-01 -5.36615372e-01 1.13742724e-01 -9.43063676e-01
5.79596162e-01 -8.29123378e-01 8.48168731e-02 -2.04436636e+00
3.41092587e-01 4.23531592e-01 2.93004453e-01 2.41082922e-01
-7.81031728e-01 -1.88852429e-01 -8.90557766e-02 -8.70614201e-02
-1.04895151e+00 1.14163947e+00 1.16641617e+00 -7.07223192e-02
-2.31612653e-01 2.07966015e-01 -9.40468311e-01 5.26315331e-01
6.94861472e-01 -2.41511092e-01 -3.17228287e-01 -3.42610627e-01
3.68715793e-01 5.20791352e-01 1.97486326e-01 -1.04562426e+00
6.63672566e-01 1.36032611e-01 2.92072892e-01 -7.24579215e-01
2.48358265e-01 1.91240624e-01 -1.05216295e-01 3.90712231e-01
-1.04553437e+00 -1.28369518e-02 1.33636639e-01 5.88925719e-01
-3.45897794e-01 -4.47833568e-01 4.14593518e-01 -2.17313886e-01
-7.80234337e-02 -1.64300352e-01 -3.52662027e-01 7.79558003e-01
5.57527959e-01 7.30887055e-02 -6.38961077e-01 -8.55114877e-01
-7.74112344e-01 4.10616159e-01 -1.01300813e-02 1.47629127e-01
5.87149441e-01 -1.04078305e+00 -1.59262300e+00 -4.12124515e-01
-2.19602436e-01 2.46110380e-01 5.30009925e-01 7.66198575e-01
-1.92484125e-01 6.51977241e-01 -3.68643962e-02 -6.96696565e-02
-1.16739559e+00 -2.81690173e-02 2.00646058e-01 -7.99211502e-01
-3.61393541e-01 1.02055562e+00 -1.18520148e-01 -4.39816535e-01
2.75006384e-01 -9.73755717e-02 -6.94113493e-01 2.99421012e-01
8.11285436e-01 6.99601054e-01 2.00888798e-01 -4.31785285e-01
1.84907973e-01 -3.36717665e-01 -3.98012847e-01 -3.00845265e-01
1.01028979e+00 -9.14854705e-02 -1.95780784e-01 4.63174820e-01
5.19353747e-01 8.44619200e-02 -6.31724477e-01 -1.17283896e-01
-5.70920631e-02 2.46738195e-01 -1.62999883e-01 -1.34141314e+00
-2.21721381e-01 7.50127673e-01 -3.07203054e-01 3.16937000e-01
8.45913827e-01 -3.27765942e-01 9.03447032e-01 9.15052652e-01
-2.14680791e-01 -1.10994303e+00 -3.24919564e-03 9.61906672e-01
1.47588873e+00 -9.78394389e-01 1.38483733e-01 -2.23752875e-02
-1.16667938e+00 9.22581851e-01 8.32931817e-01 9.00017992e-02
-2.15368658e-01 -4.72317934e-01 -3.13862741e-01 -4.99950945e-02
-1.30655539e+00 2.45629966e-01 4.53114480e-01 4.26508576e-01
7.77324677e-01 3.31241004e-02 -7.07212329e-01 9.50396299e-01
-8.70006263e-01 -3.51168036e-01 9.16556299e-01 5.07922232e-01
-4.12871301e-01 -6.47769570e-01 -5.85690923e-02 6.06804490e-01
-6.65441751e-01 -5.63752949e-01 -6.47280276e-01 3.80316585e-01
-6.24775410e-01 1.06231284e+00 -8.07042196e-02 -1.06070921e-01
5.98232269e-01 4.25654709e-01 4.64312524e-01 -9.34217334e-01
-1.07771385e+00 -4.11965102e-01 9.09940541e-01 -2.20462605e-01
-3.38279814e-01 -6.91208124e-01 -1.23010802e+00 -2.25261688e-01
-4.44242656e-01 4.41534519e-01 3.68063509e-01 8.75864685e-01
5.74065864e-01 9.90836501e-01 2.96749890e-01 -5.77658236e-01
-1.28605616e+00 -1.93078399e+00 -2.14050904e-01 5.49895585e-01
1.80941269e-01 -8.32194909e-02 -2.49760568e-01 6.75110668e-02] | [12.27811336517334, 9.197392463684082] |
44cd8547-f569-46b0-a544-13962f958288 | on-the-futility-of-learning-complex-frame | 1702.00178 | null | http://arxiv.org/abs/1702.00178v2 | http://arxiv.org/pdf/1702.00178v2.pdf | On the Futility of Learning Complex Frame-Level Language Models for Chord Recognition | Chord recognition systems use temporal models to post-process frame-wise
chord preditions from acoustic models. Traditionally, first-order models such
as Hidden Markov Models were used for this task, with recent works suggesting
to apply Recurrent Neural Networks instead. Due to their ability to learn
longer-term dependencies, these models are supposed to learn and to apply
musical knowledge, instead of just smoothing the output of the acoustic model.
In this paper, we argue that learning complex temporal models at the level of
audio frames is futile on principle, and that non-Markovian models do not
perform better than their first-order counterparts. We support our argument
through three experiments on the McGill Billboard dataset. The first two show
1) that when learning complex temporal models at the frame level, improvements
in chord sequence modelling are marginal; and 2) that these improvements do not
translate when applied within a full chord recognition system. The third, still
rather preliminary experiment gives first indications that the use of complex
sequential models for chord prediction at higher temporal levels might be more
promising. | ['Filip Korzeniowski', 'Gerhard Widmer'] | 2017-02-01 | null | null | null | null | ['chord-recognition'] | ['audio'] | [ 3.11042398e-01 2.40212709e-01 8.87446702e-02 -2.01202422e-01
-6.00635171e-01 -6.55463457e-01 6.37893796e-01 1.60384670e-01
-7.12864697e-01 4.43881691e-01 4.32422698e-01 -5.35973907e-01
-1.02974154e-01 -4.72033262e-01 -3.87714088e-01 -5.39945841e-01
-3.86768043e-01 2.45866179e-01 7.83697724e-01 -3.94106060e-01
4.18495685e-01 3.79579306e-01 -1.60813844e+00 7.04551816e-01
1.25417978e-01 6.86747134e-01 3.40666354e-01 1.22514915e+00
-6.06484041e-02 1.25640261e+00 -6.04659319e-01 -7.10477829e-02
2.15211473e-02 -8.93906057e-01 -1.03728175e+00 -3.66507918e-01
5.42133003e-02 -1.13114752e-01 -1.26013443e-01 5.45814753e-01
2.52952009e-01 3.46922278e-01 4.85941112e-01 -5.63515842e-01
-8.22768211e-02 1.02578902e+00 1.32593393e-01 1.84861526e-01
3.35679680e-01 -9.90955904e-02 1.46904123e+00 -5.47698677e-01
4.48537946e-01 1.15348911e+00 1.14784801e+00 4.87847090e-01
-1.33272088e+00 -4.06814635e-01 3.05029452e-01 4.73086089e-01
-9.95039880e-01 -5.29952765e-01 7.78281212e-01 -4.44262803e-01
1.23768961e+00 3.86470467e-01 9.11975503e-01 9.05928314e-01
1.38116241e-01 8.08393240e-01 1.16314733e+00 -6.68030381e-01
2.93810014e-02 -1.94488361e-01 -3.33264135e-02 3.18294764e-01
-4.90211338e-01 4.19953555e-01 -8.48231614e-01 -1.99433237e-01
7.97322571e-01 -5.35645962e-01 -1.26160532e-01 2.39733532e-02
-1.10067928e+00 7.73602784e-01 1.08692348e-01 7.76723206e-01
-3.88086319e-01 2.59269089e-01 7.02594399e-01 4.57888871e-01
1.45177156e-01 6.14640713e-01 -6.51980042e-01 -7.82408059e-01
-1.38303387e+00 3.02205414e-01 8.21392119e-01 3.17103922e-01
2.76517540e-01 2.11760864e-01 1.12270176e-01 8.36692214e-01
1.89368442e-01 -2.02971190e-01 6.84321523e-01 -1.17000771e+00
-1.75658688e-01 -1.37327075e-01 -1.13842569e-01 -7.09402144e-01
-4.12203401e-01 -1.87677264e-01 -3.62332404e-01 3.76355022e-01
8.33999336e-01 6.23047948e-02 -8.42554390e-01 1.84828389e+00
-2.54924357e-01 3.05008799e-01 9.48709249e-02 7.14843631e-01
2.76094913e-01 8.00600231e-01 1.04723126e-02 -6.01996660e-01
1.07110190e+00 -6.72379494e-01 -6.93783641e-01 -2.30402052e-02
8.28543007e-01 -1.12907207e+00 1.13152826e+00 9.46438909e-01
-1.19554842e+00 -8.03610325e-01 -1.08562982e+00 -3.80874276e-02
-6.53716996e-02 -2.14359909e-01 5.65205872e-01 4.64964926e-01
-9.69786763e-01 1.16470468e+00 -1.10740995e+00 -2.96274692e-01
-3.85815352e-01 2.25386009e-01 1.55132279e-01 5.57051182e-01
-1.31582284e+00 9.34823155e-01 5.25665700e-01 1.70542732e-01
-7.31855214e-01 -4.10195976e-01 -5.14356494e-01 2.55158748e-02
2.38937587e-01 -1.73884794e-01 1.60728228e+00 -1.03339338e+00
-1.91360331e+00 5.28148174e-01 -1.97645292e-01 -6.93088591e-01
3.70922506e-01 -2.51742929e-01 -6.08518533e-02 7.32573196e-02
-5.12854815e-01 5.87760091e-01 6.21115685e-01 -1.04629397e+00
-6.98222101e-01 6.14600480e-02 1.54233769e-01 8.52507278e-02
-3.56554869e-03 1.72525838e-01 -1.32211046e-02 -8.86662900e-01
1.59858942e-01 -1.37939286e+00 -4.24013495e-01 -5.30409098e-01
-3.79614346e-02 -4.20285791e-01 5.51070809e-01 -8.82581472e-01
1.47323239e+00 -2.12897038e+00 2.56849110e-01 1.15544230e-01
-5.09328365e-01 2.82511353e-01 -2.23861169e-02 6.19114935e-01
-3.14149618e-01 -7.80403763e-02 -2.41110176e-01 -2.70897120e-01
-8.78093615e-02 3.86438012e-01 -5.06978512e-01 5.76175116e-02
-4.34439965e-02 6.74261391e-01 -6.97148621e-01 -3.64628345e-01
1.00718148e-01 6.21243179e-01 -6.78635955e-01 -6.63968995e-02
-3.81442219e-01 5.54128706e-01 -7.71706132e-03 3.65901589e-02
-1.52767584e-01 2.97172070e-01 4.88201499e-01 9.97638330e-02
-4.11723614e-01 9.84985709e-01 -1.23955119e+00 1.91618025e+00
-3.90825659e-01 7.60173440e-01 -1.80382892e-01 -1.09363866e+00
7.21558869e-01 7.83607721e-01 3.23682964e-01 -6.24760628e-01
-1.88710585e-01 3.36062580e-01 5.93931973e-01 -2.24624425e-01
5.79296529e-01 -8.02512825e-01 -1.15265831e-01 4.22995150e-01
1.06247611e-01 -3.25787455e-01 8.23921636e-02 -1.89728826e-01
1.02971113e+00 6.88571215e-01 2.35168333e-03 -1.00887030e-01
3.91205579e-01 -1.14941604e-01 7.25233138e-01 7.16119826e-01
6.58075139e-02 6.25643432e-01 4.91131693e-01 -4.55878109e-01
-9.69603360e-01 -7.69541025e-01 3.80101288e-03 1.42344725e+00
-5.23256540e-01 -8.87221038e-01 -5.79126835e-01 -4.35396701e-01
-4.78976041e-01 6.62268758e-01 -3.84865463e-01 2.17563491e-02
-1.05212605e+00 -4.48076725e-01 7.73038447e-01 6.21224463e-01
-1.68006897e-01 -1.44989765e+00 -9.56304133e-01 6.40407562e-01
-1.39515728e-01 -7.98868775e-01 -1.04572564e-01 6.96861744e-01
-1.22552633e+00 -8.22689831e-01 -6.75417602e-01 -7.53678858e-01
-1.31910339e-01 -2.58113801e-01 9.64531064e-01 3.39440316e-01
-4.38500233e-02 1.79285511e-01 -6.04331017e-01 -5.03938377e-01
-6.56695962e-01 2.02963486e-01 6.84779212e-02 -3.25433403e-01
2.02663094e-01 -9.30276692e-01 -2.25644454e-01 1.53353363e-01
-8.73117149e-01 -3.99816073e-02 5.02689064e-01 9.30580795e-01
3.42535049e-01 1.81953632e-03 6.02311909e-01 -8.54435444e-01
5.65338731e-01 2.76544094e-01 -3.60857040e-01 -3.88636068e-02
-4.42064434e-01 2.10829109e-01 5.97243845e-01 -6.27834380e-01
-8.94141793e-01 2.56393850e-01 -5.37467122e-01 -1.90897152e-01
-2.66051590e-01 7.72426248e-01 2.75350899e-01 3.70283872e-01
6.80806875e-01 2.45691404e-01 -1.10718548e-01 -6.62950635e-01
2.40006715e-01 3.30983013e-01 5.23629785e-01 -6.60298049e-01
6.03372335e-01 1.80337206e-01 1.57637373e-02 -1.09766364e+00
-7.66854465e-01 -3.96981627e-01 -1.03159237e+00 -2.87993103e-01
8.31704557e-01 -5.66936731e-01 -6.31776273e-01 2.49540329e-01
-1.18267846e+00 -7.93994665e-01 -4.64776158e-01 7.28162348e-01
-1.11801922e+00 3.68501127e-01 -1.13422453e+00 -1.17716455e+00
1.93917796e-01 -9.24672961e-01 5.74149251e-01 -5.23432381e-02
-8.55830908e-01 -9.95158076e-01 3.25784445e-01 9.91305932e-02
2.87484318e-01 2.81284750e-02 1.16196799e+00 -5.52471042e-01
-2.40483180e-01 -8.22470561e-02 5.07075429e-01 4.94455159e-01
1.36504266e-02 2.00600684e-01 -1.17603540e+00 -1.76407546e-01
1.91342890e-01 -2.97075480e-01 9.33887839e-01 3.08884680e-01
8.09658289e-01 -8.32508281e-02 3.71913314e-02 2.10003167e-01
1.00708246e+00 4.69391078e-01 6.77715778e-01 2.86362946e-01
4.15109664e-01 9.17131841e-01 7.22021580e-01 2.25424454e-01
1.34390667e-01 9.99825120e-01 2.23351195e-02 1.89055994e-01
-2.85986871e-01 -3.35928351e-01 7.22421885e-01 1.35480380e+00
-5.67150176e-01 2.55936146e-01 -9.78768229e-01 6.60970867e-01
-1.87474060e+00 -1.27173209e+00 -2.85573781e-01 2.25678325e+00
1.02556908e+00 7.24510908e-01 6.89164519e-01 6.78793132e-01
3.00823957e-01 1.33786649e-01 -3.11734751e-02 -8.19151700e-01
-7.80650973e-02 4.66224134e-01 2.31708307e-02 6.48797691e-01
-7.85818994e-01 9.68829393e-01 6.97458744e+00 7.33587325e-01
-1.17054629e+00 -7.11874478e-03 2.59114087e-01 -1.34377822e-01
-1.92414194e-01 6.06410563e-01 -5.25351405e-01 2.09487885e-01
1.50708175e+00 2.73223519e-01 3.02577347e-01 3.89486015e-01
3.48091424e-01 -1.57025129e-01 -1.28698957e+00 5.52581489e-01
-2.80306280e-01 -1.20763969e+00 -8.99425894e-02 7.95211494e-02
2.71178722e-01 -2.01994702e-01 -1.01500690e-01 5.21997929e-01
1.85175445e-02 -1.21320593e+00 9.70472515e-01 7.05061436e-01
1.98919311e-01 -8.78545225e-01 4.91791576e-01 6.23499870e-01
-1.25711465e+00 3.14985216e-02 -2.53137976e-01 -5.52322209e-01
4.95525122e-01 5.77601865e-02 -7.70209551e-01 4.13442641e-01
5.70116758e-01 5.31371236e-01 -4.90132451e-01 9.82621312e-01
-4.02368397e-01 1.17734289e+00 -4.40344274e-01 1.29936993e-01
3.97925079e-01 3.29361274e-03 5.49975574e-01 1.54628563e+00
1.76435217e-01 9.87659469e-02 3.95330898e-02 2.87613720e-01
8.18267345e-01 1.17689036e-01 -3.43047589e-01 -1.88235164e-01
9.33820233e-02 7.69243360e-01 -9.22065735e-01 -2.37574205e-01
-5.03652871e-01 8.10537994e-01 9.53941867e-02 1.26257539e-01
-6.86596155e-01 -2.57872909e-01 3.31117272e-01 1.87926203e-01
5.95260739e-01 -5.70176184e-01 -3.15764338e-01 -8.03445935e-01
-2.54635960e-01 -9.13403869e-01 4.99991328e-01 -6.08786702e-01
-7.35079706e-01 6.70661628e-01 3.78525071e-02 -1.04458928e+00
-8.34439397e-01 -5.04212797e-01 -5.91056228e-01 7.89689541e-01
-1.13589478e+00 -8.93687367e-01 6.04542732e-01 3.53058577e-01
7.22446561e-01 1.90646797e-01 1.27838504e+00 1.98777411e-02
1.47952974e-01 2.68914759e-01 -3.45734715e-01 1.84737980e-01
6.06618345e-01 -1.36683571e+00 3.07657242e-01 7.58842468e-01
1.00471246e+00 6.50788486e-01 1.16135728e+00 -4.21750873e-01
-7.09755182e-01 -2.18091100e-01 1.26387000e+00 -5.01230359e-01
6.04492784e-01 -3.76515597e-01 -1.23869932e+00 5.98176360e-01
3.17993104e-01 -4.37775344e-01 8.26695859e-01 6.19971395e-01
-2.94960767e-01 2.79474974e-01 -2.92628497e-01 5.19090295e-01
7.66863286e-01 -8.93032491e-01 -1.06113493e+00 -9.89642963e-02
6.05737805e-01 -1.43380940e-01 -9.90612686e-01 6.59106553e-01
9.99296069e-01 -1.23982215e+00 7.45316982e-01 -5.58120608e-01
1.97646081e-01 -2.88537413e-01 -2.21491739e-01 -1.07407093e+00
-2.64573246e-01 -7.29203403e-01 -9.60219204e-02 1.17936790e+00
6.64328396e-01 3.67374495e-02 1.00493133e+00 1.96730882e-01
-2.28244752e-01 -5.44894218e-01 -9.97555554e-01 -9.42381799e-01
3.68166655e-01 -9.34824109e-01 -4.00219336e-02 9.43230510e-01
2.85026610e-01 4.83568251e-01 -6.05801880e-01 -6.54841065e-02
1.41040251e-01 9.08431932e-02 5.51532507e-01 -1.36220849e+00
-9.28507447e-01 -5.89117467e-01 -3.97957921e-01 -1.19023907e+00
1.68675736e-01 -6.02315068e-01 3.75654638e-01 -1.08272779e+00
-2.71134943e-01 -4.74593043e-01 -6.88183308e-01 4.55595404e-01
-1.21518038e-02 2.42276683e-01 5.42580783e-01 2.21432433e-01
-2.18194440e-01 2.81740725e-01 8.25470805e-01 2.54028767e-01
-3.40706021e-01 3.14685404e-01 -1.16136692e-01 1.04285765e+00
7.68765152e-01 -6.10083938e-01 -4.64474946e-01 -1.23677298e-01
2.97930390e-01 4.05665636e-01 1.74693421e-01 -1.25170147e+00
3.44748527e-01 -1.18607007e-01 2.20721170e-01 -6.06878459e-01
7.29303598e-01 -4.27086174e-01 1.59312055e-01 6.39397740e-01
-6.03947997e-01 9.26257223e-02 4.55909222e-01 4.25975770e-01
-4.87472475e-01 -4.95994210e-01 7.16615796e-01 -4.25384581e-01
-7.42693782e-01 -3.22032124e-01 -9.56235647e-01 -3.33352238e-01
3.90027463e-01 -4.09090042e-01 5.25386035e-01 -6.09336019e-01
-1.43546915e+00 -3.98760378e-01 2.50129849e-01 4.79330033e-01
2.95141518e-01 -1.09035909e+00 -5.83157599e-01 -4.72579384e-03
-1.11411415e-01 -2.95711279e-01 1.24414518e-01 9.33946431e-01
-2.42797509e-01 5.81418455e-01 -5.68991574e-03 -6.67312622e-01
-1.57654905e+00 3.62036645e-01 2.55902588e-01 -3.87799680e-01
-6.86586142e-01 1.05450320e+00 1.32646918e-01 -3.34760427e-01
4.42344427e-01 -4.22324628e-01 -2.10738927e-01 3.37178826e-01
3.46319556e-01 -1.05043545e-01 8.74907989e-03 -7.02348411e-01
-3.01571131e-01 4.60719764e-01 -9.83500928e-02 -7.48310030e-01
1.45613050e+00 -1.16644632e-02 1.46450669e-01 1.39080560e+00
6.75403118e-01 1.55361205e-01 -1.30762565e+00 -2.30599549e-02
7.19367683e-01 -7.63553157e-02 -4.11374122e-02 -7.21336424e-01
-3.33159328e-01 1.14784920e+00 1.59911096e-01 4.75281268e-01
1.17916965e+00 -2.14585498e-01 5.96943319e-01 2.60126114e-01
2.66575217e-01 -1.22322655e+00 2.34763813e-03 9.83897507e-01
7.62509465e-01 -4.98397678e-01 -1.84122950e-01 2.35433895e-02
-5.48805892e-01 1.47467101e+00 1.92334294e-01 -2.59473056e-01
5.54861367e-01 2.77037740e-01 2.22894266e-01 1.69446498e-01
-1.26054215e+00 -3.43229115e-01 1.69089898e-01 3.88269454e-01
1.09389389e+00 -2.29401037e-01 -3.40263665e-01 5.26477516e-01
-6.79293752e-01 -6.46086782e-02 6.56253457e-01 1.02878845e+00
-4.81092334e-01 -1.66900003e+00 -3.70614946e-01 2.44251173e-03
-8.47929120e-01 -1.40959740e-01 -4.69380587e-01 7.47224331e-01
1.78476304e-01 9.25499797e-01 -6.88160136e-02 -4.67504382e-01
3.25683117e-01 6.80656314e-01 7.54449785e-01 -7.61707067e-01
-7.73373902e-01 6.34278595e-01 3.22487682e-01 -4.48106974e-01
-6.95769131e-01 -9.74615932e-01 -1.23312891e+00 1.44516826e-01
-3.31733286e-01 6.05249226e-01 5.78145444e-01 1.08754289e+00
-2.37094045e-01 6.63792312e-01 2.19582587e-01 -1.06491065e+00
-4.38447654e-01 -1.01355374e+00 -4.77799475e-01 1.71196446e-01
3.49473745e-01 -1.81156740e-01 -2.59664774e-01 3.63485277e-01] | [15.904216766357422, 5.341151714324951] |
4a574455-bdd0-4841-98d5-5be5632decdd | sign-language-translation-with-transformers | 2004.00588 | null | https://arxiv.org/abs/2004.00588v2 | https://arxiv.org/pdf/2004.00588v2.pdf | Better Sign Language Translation with STMC-Transformer | Sign Language Translation (SLT) first uses a Sign Language Recognition (SLR) system to extract sign language glosses from videos. Then, a translation system generates spoken language translations from the sign language glosses. This paper focuses on the translation system and introduces the STMC-Transformer which improves on the current state-of-the-art by over 5 and 7 BLEU respectively on gloss-to-text and video-to-text translation of the PHOENIX-Weather 2014T dataset. On the ASLG-PC12 corpus, we report an increase of over 16 BLEU. We also demonstrate the problem in current methods that rely on gloss supervision. The video-to-text translation of our STMC-Transformer outperforms translation of GT glosses. This contradicts previous claims that GT gloss translation acts as an upper bound for SLT performance and reveals that glosses are an inefficient representation of sign language. For future SLT research, we therefore suggest an end-to-end training of the recognition and translation models, or using a different sign language annotation scheme. | ['Jesse Read', 'Kayo Yin'] | 2020-04-01 | sign-language-translation-with-transformers-1 | https://aclanthology.org/2020.coling-main.525 | https://aclanthology.org/2020.coling-main.525.pdf | coling-2020-8 | ['sign-language-translation'] | ['computer-vision'] | [ 5.81779242e-01 -9.52489022e-03 -2.09363848e-01 -6.34600282e-01
-1.42711008e+00 -6.72023892e-01 8.25494349e-01 -8.91707838e-01
-5.50138831e-01 5.27230918e-01 6.73236966e-01 -3.98989558e-01
3.21243972e-01 -1.17782712e-01 -6.39413893e-01 -5.91916442e-01
4.16001976e-01 9.10632551e-01 2.50284672e-01 -2.01545119e-01
1.40523696e-02 -8.32174905e-03 -1.44064081e+00 6.95855796e-01
5.51280022e-01 7.97854245e-01 -7.12302029e-02 1.01146030e+00
-2.34628156e-01 8.03767800e-01 -7.83644736e-01 -4.40930486e-01
4.55142796e-01 -1.04270303e+00 -6.08259737e-01 -4.24971730e-02
1.01456523e+00 -7.96379924e-01 -2.89535284e-01 5.37797868e-01
7.16545343e-01 -4.01706457e-01 6.63673460e-01 -1.13923109e+00
-5.35200179e-01 5.40046394e-01 5.58088273e-02 -3.61695349e-01
5.58152795e-01 3.56583655e-01 9.67263997e-01 -8.73836994e-01
1.21592486e+00 1.15833306e+00 5.01080513e-01 9.46649373e-01
-6.39990568e-01 -6.35262430e-01 -6.25994131e-02 -5.71029037e-02
-1.16092229e+00 -6.01541519e-01 -1.42327726e-01 -2.05076486e-01
1.35841191e+00 4.33093220e-01 7.82742262e-01 1.42328942e+00
-5.22286966e-02 1.22182333e+00 1.40001583e+00 -6.87591970e-01
-4.40072045e-02 -3.74173135e-01 -1.55235991e-01 8.42960060e-01
1.64476529e-01 1.64375633e-01 -1.02592838e+00 1.09611951e-01
6.48440719e-01 -5.97928464e-01 -3.55747610e-01 1.06954940e-01
-1.44227743e+00 3.80175531e-01 -2.53774524e-01 1.24376208e-01
-2.33560264e-01 4.35126334e-01 5.84371507e-01 8.32585633e-01
1.73497185e-01 -1.09953312e-02 -4.35375243e-01 -8.22382569e-01
-1.18001926e+00 -7.06606358e-02 9.87712324e-01 1.28537619e+00
3.99450846e-02 1.84352487e-01 -3.02624166e-01 6.15955591e-01
5.52530646e-01 1.25776160e+00 6.77619159e-01 -9.14720178e-01
7.18007088e-01 3.11784506e-01 2.63561867e-02 8.68907124e-02
7.66268969e-02 1.83113486e-01 -1.03756050e-02 3.90821487e-01
7.65466094e-01 -2.22281501e-01 -1.76969826e+00 1.38662064e+00
-1.61287114e-01 -2.47808263e-01 2.21303463e-01 1.36741149e+00
7.88325667e-01 4.22204107e-01 -6.50882125e-02 5.86246960e-02
1.20283604e+00 -1.06278932e+00 -8.01155567e-01 -2.01922357e-01
8.64544094e-01 -1.24770558e+00 1.28725696e+00 4.42708433e-01
-8.70612562e-01 9.85616744e-02 -7.92450368e-01 -1.45257697e-01
-2.35378936e-01 6.25923991e-01 2.07466841e-01 6.97493553e-01
-1.08120549e+00 1.11714326e-01 -1.12595618e+00 -1.10987473e+00
-1.26247376e-01 2.06978887e-01 -3.40344220e-01 -3.60641062e-01
-6.75894558e-01 1.28735578e+00 -1.68401077e-01 1.99779272e-01
-7.50916362e-01 -7.42816329e-02 -6.38942540e-01 -6.13745749e-01
1.32743955e-01 -5.79610407e-01 1.57294822e+00 -1.07244599e+00
-2.14832449e+00 1.19534051e+00 -4.19703633e-01 -2.68666834e-01
9.82308328e-01 -4.30648029e-01 -5.41205406e-01 2.97817618e-01
-7.93502629e-02 5.60613930e-01 9.33986962e-01 -9.38419402e-01
-8.10453653e-01 4.21440527e-02 -3.65299672e-01 3.75885457e-01
4.02269989e-01 6.15997851e-01 -6.30174816e-01 -5.41002095e-01
1.45570830e-01 -1.18928933e+00 4.02760953e-01 2.21401602e-01
-7.06389695e-02 -2.34803017e-02 8.66536558e-01 -1.11715066e+00
1.06501377e+00 -1.68801343e+00 1.99402750e-01 2.47615278e-01
-2.21056059e-01 5.61054468e-01 -4.65542465e-01 6.95078492e-01
3.61584127e-01 -1.02876276e-01 -2.48892412e-01 -6.27387524e-01
3.50229412e-01 8.39241743e-01 -4.70299155e-01 3.87065232e-01
-5.15844440e-03 1.06278861e+00 -6.82929039e-01 -5.48522472e-01
2.21604675e-01 4.92767066e-01 -2.46646360e-01 2.80279815e-02
-3.01814795e-01 3.44058156e-01 -1.38675123e-01 1.01528394e+00
8.60173106e-02 1.86034739e-01 3.21770757e-01 1.42303323e-02
-3.54969621e-01 6.51289165e-01 -4.49990958e-01 1.78144777e+00
-3.54355991e-01 1.14056587e+00 3.57701629e-02 -5.10061920e-01
7.86688030e-01 6.23305976e-01 2.37991229e-01 -7.51446605e-01
2.53616393e-01 1.06918716e+00 -9.03834775e-02 -8.18241239e-01
2.37662703e-01 -2.16640294e-01 -8.17151815e-02 7.56681085e-01
1.29731700e-01 -3.02779794e-01 3.39096338e-01 -1.68263689e-01
1.09382141e+00 7.39667535e-01 -1.69553921e-01 1.29681125e-01
1.38588041e-01 2.77019501e-01 1.63430169e-01 5.51404178e-01
-3.94610018e-01 7.83909202e-01 1.00860307e-02 -2.86472589e-01
-1.10897970e+00 -9.63969707e-01 2.79683232e-01 1.06914175e+00
-3.51514697e-01 -4.85185862e-01 -8.76067281e-01 -7.40391612e-01
-3.93718243e-01 9.26183105e-01 -1.72832996e-01 1.99870259e-01
-1.04833972e+00 -2.91303605e-01 1.33032608e+00 5.11506796e-01
5.02355218e-01 -1.02303576e+00 -9.51570868e-01 -1.19084110e-02
-7.75982559e-01 -1.43941152e+00 -9.38017309e-01 -2.22003236e-01
-6.38608873e-01 -8.91097307e-01 -1.15775597e+00 -9.66758251e-01
5.30403078e-01 -1.56031460e-01 8.42788279e-01 -4.00168970e-02
-4.74360511e-02 7.53808677e-01 -8.28349233e-01 -4.16428804e-01
-8.06110203e-01 -1.03223242e-01 4.07977924e-02 -1.69825763e-01
8.85931730e-01 -2.04978108e-01 -3.23419988e-01 6.48879051e-01
-7.35696256e-01 2.04716042e-01 8.47826898e-01 6.78108156e-01
3.98475856e-01 -9.57147419e-01 -9.81166884e-02 8.94155800e-02
5.47732055e-01 4.79987830e-01 -5.31973898e-01 7.09363580e-01
-6.92667544e-01 3.60142142e-01 1.32491693e-01 -5.67649782e-01
-8.91853273e-01 8.96968320e-02 -1.20917067e-01 -3.92392486e-01
6.87228590e-02 4.41212296e-01 2.06612542e-01 -7.29259402e-02
4.94244039e-01 6.53160393e-01 1.86971039e-01 -3.46471876e-01
3.09006482e-01 1.08483696e+00 4.70751911e-01 -5.17996430e-01
6.28477871e-01 3.95530015e-01 -2.78985649e-01 -8.87377143e-01
-4.27324325e-01 -3.81238192e-01 -5.38202763e-01 -5.65863788e-01
9.29271460e-01 -7.59646475e-01 -3.51145655e-01 5.58200538e-01
-1.08510125e+00 -8.73364091e-01 -2.78055638e-01 8.35685611e-01
-1.03439784e+00 4.59554344e-01 -6.86266482e-01 -8.34538579e-01
-6.05872154e-01 -1.14392662e+00 1.68539429e+00 -2.65782058e-01
-4.82960075e-01 -4.52344477e-01 2.43296415e-01 6.67321861e-01
5.27197897e-01 5.45563176e-03 2.80600160e-01 -1.64234236e-01
-6.09967351e-01 -3.40376586e-01 -2.96632648e-01 5.21425486e-01
9.54260156e-02 6.32198378e-02 -7.40678251e-01 -2.79339045e-01
-5.64663947e-01 -5.18614888e-01 8.52961063e-01 2.07026958e-01
-1.12616375e-01 -4.06478643e-01 4.87677604e-02 5.02497017e-01
1.18481469e+00 3.36119682e-02 9.00793791e-01 1.91671208e-01
4.07069802e-01 3.91218513e-01 6.76375031e-01 5.89294620e-02
4.92644846e-01 7.95798719e-01 -1.95466876e-01 1.27732396e-01
-8.96016479e-01 -5.64827740e-01 1.12215602e+00 1.18042958e+00
-6.39635682e-01 -3.73720676e-01 -8.96671236e-01 5.64016223e-01
-1.90461326e+00 -8.65714252e-01 -4.22368273e-02 2.11378074e+00
7.80550003e-01 -2.48097047e-01 1.83453336e-01 -4.91286293e-02
1.68590024e-01 -1.40179351e-01 -2.43636459e-01 -5.55247009e-01
-2.26544440e-01 2.38528565e-01 9.51740742e-01 7.12874413e-01
-6.23129189e-01 1.46951854e+00 7.09365177e+00 5.49204648e-01
-1.44602370e+00 7.20860511e-02 -3.88366967e-01 -2.37840816e-01
-3.39582041e-02 1.09240217e-02 -8.07873905e-01 2.03841463e-01
1.15038490e+00 2.01902568e-01 7.77022839e-01 4.13327485e-01
3.86399239e-01 -8.99569839e-02 -9.91301775e-01 9.89152670e-01
5.61906815e-01 -7.69861877e-01 3.20614070e-01 2.21717164e-01
5.27814806e-01 6.58969045e-01 -2.30590045e-01 3.09029639e-01
4.00932550e-01 -7.22538352e-01 1.15322483e+00 5.58968306e-01
1.50190628e+00 7.30592161e-02 7.36067832e-01 -1.19352890e-02
-1.18903065e+00 3.89039963e-01 5.09175239e-03 1.37892529e-01
6.56579673e-01 -2.57356733e-01 -1.07334268e+00 3.95793587e-01
4.61001933e-01 3.92789125e-01 -2.31187910e-01 8.24308276e-01
-6.40117407e-01 1.12346458e+00 -8.58963430e-01 -2.91240513e-01
4.81329620e-01 -2.20271185e-01 6.08660102e-01 1.56254244e+00
7.18191564e-01 -7.09960163e-02 4.08684798e-02 3.13878179e-01
2.45679229e-01 1.11307390e-01 -5.43087542e-01 -4.49144214e-01
-6.23469166e-02 3.83424342e-01 -4.66407448e-01 -7.10613132e-01
-4.55404013e-01 1.63653111e+00 -3.22569191e-01 6.46953881e-01
-7.77047098e-01 -1.85377136e-01 6.57133877e-01 -9.41287205e-02
4.37689692e-01 -5.39582014e-01 -9.84709859e-02 -1.44918203e+00
3.46084118e-01 -1.14027250e+00 1.31338567e-01 -1.04473865e+00
-8.82085383e-01 4.79854465e-01 -2.96473410e-02 -1.70513356e+00
-6.50305212e-01 -9.09906626e-01 -1.23302855e-01 6.94234610e-01
-1.55244958e+00 -1.69507790e+00 -3.35350573e-01 6.61651909e-01
6.97308302e-01 -6.46724105e-02 1.01803577e+00 3.21267217e-01
4.45389785e-02 8.27721953e-01 9.38998759e-02 4.18543339e-01
1.09886563e+00 -8.87043595e-01 4.11080390e-01 1.01060319e+00
3.15054119e-01 4.23045456e-01 7.91982532e-01 -7.79793918e-01
-1.37634051e+00 -8.15711677e-01 1.55094123e+00 -9.19318497e-01
6.64535820e-01 -9.92125422e-02 -7.16622621e-02 1.07846022e+00
4.27744091e-01 -4.72739160e-01 3.57457936e-01 -4.96847451e-01
-6.59554005e-01 2.27258667e-01 -8.97152424e-01 9.08320010e-01
1.42238474e+00 -8.34136844e-01 -7.98266828e-01 3.96812707e-01
2.79281199e-01 -5.04503191e-01 -6.88686252e-01 1.44936547e-01
1.44304144e+00 -5.65413356e-01 4.76930201e-01 -5.30424297e-01
2.38052160e-01 -4.24744338e-01 -6.21426642e-01 -1.14441669e+00
5.54796517e-01 -9.06230569e-01 2.01198205e-01 7.17584848e-01
6.03271425e-01 -6.30699098e-01 7.04148591e-01 6.79956257e-01
-4.11919415e-01 -4.14975882e-02 -1.21302342e+00 -1.16382790e+00
-6.37487918e-02 -5.33957303e-01 1.81105345e-01 6.13698542e-01
1.00510277e-01 2.87589341e-01 -6.63014352e-01 -2.13010728e-01
6.43734694e-01 -2.87239486e-03 1.01066709e+00 -6.75651073e-01
-1.65539220e-01 -5.75169206e-01 -3.25985074e-01 -1.38136661e+00
-2.86375850e-01 -7.73460865e-01 5.55453718e-01 -1.86764371e+00
-2.88724959e-01 2.71337450e-01 3.56484473e-01 9.03031766e-01
4.45366949e-01 4.47435856e-01 4.80090082e-01 4.61420417e-01
-4.60679293e-01 3.70825112e-01 1.42085958e+00 -2.11859122e-01
1.81139330e-03 -1.82606727e-01 2.79606599e-02 4.65442121e-01
6.89508140e-01 -3.33247811e-01 2.21628807e-02 -9.31901157e-01
-1.49487227e-01 -2.15184152e-01 9.97264758e-02 -7.84893632e-01
9.19630602e-02 1.66505072e-02 -3.98824990e-01 -5.16969621e-01
4.78223175e-01 -8.35395336e-01 -8.34928602e-02 4.77222413e-01
-2.21304268e-01 -4.51054275e-02 1.10705784e-02 1.29974216e-01
-2.48936743e-01 4.45708781e-02 4.91124004e-01 -2.46702284e-01
-6.09862626e-01 -2.26830810e-01 -1.06149590e+00 -1.52840197e-01
3.90027881e-01 -5.44079721e-01 -5.05265236e-01 -7.90690124e-01
-5.81176937e-01 5.61801158e-02 4.56295550e-01 4.42815840e-01
6.85589492e-01 -1.22107744e+00 -9.48705435e-01 2.83878505e-01
4.11209852e-01 -5.05916297e-01 -5.00332773e-01 1.13642108e+00
-9.49156880e-01 7.63905466e-01 -2.04880551e-01 -5.74088097e-01
-1.72685421e+00 -4.15209562e-01 3.36562097e-01 -1.07183121e-01
-7.97237277e-01 6.11209989e-01 -4.79593575e-01 -6.29125893e-01
4.27271724e-01 -9.84457016e-01 5.57835162e-01 -4.10333753e-01
3.78149897e-01 1.31826371e-01 -7.06839189e-02 -9.04471993e-01
-2.91685462e-01 9.63207960e-01 3.03394347e-01 -9.61949348e-01
9.83324945e-01 -7.41011649e-02 2.14775860e-01 3.02674532e-01
9.47649062e-01 -4.64665852e-02 -8.65388274e-01 -9.90438648e-03
1.28943816e-01 -3.24727356e-01 -1.88177228e-01 -1.44716132e+00
-6.44211650e-01 7.41807580e-01 6.87218904e-01 -6.68341696e-01
1.11477172e+00 -9.71831903e-02 1.10541749e+00 7.64912724e-01
6.35585070e-01 -1.34650135e+00 -2.28157446e-01 7.93348193e-01
1.19777548e+00 -1.14129257e+00 -3.42822492e-01 -3.55481915e-02
-8.23718131e-01 1.10483325e+00 1.12399533e-01 1.18849032e-01
2.32570261e-01 4.12469506e-01 9.62929487e-01 -1.50450751e-01
-6.03490889e-01 -5.83186924e-01 3.64341557e-01 5.36127508e-01
4.68028218e-01 2.56277949e-01 -7.67482817e-01 -1.21033967e-01
-3.94768447e-01 5.62028050e-01 3.13224584e-01 1.30538142e+00
-3.56928498e-01 -1.38457501e+00 -4.00004536e-01 2.16330364e-01
-6.83474913e-02 -2.52935886e-01 -1.04255271e+00 8.85068715e-01
-1.09025799e-01 1.00186265e+00 -3.19914222e-01 -4.41423684e-01
5.85727453e-01 7.79811740e-01 9.00121152e-01 -2.01398328e-01
-4.73730505e-01 4.06465560e-01 7.49264836e-01 -7.36087203e-01
-8.38321209e-01 -9.13575351e-01 -1.25539851e+00 -1.18237942e-01
-1.27690598e-01 -3.41257930e-01 8.97236228e-01 1.22006786e+00
9.15246755e-02 5.68266842e-04 -2.38667607e-01 -6.26313210e-01
-4.99845862e-01 -1.33298099e+00 -1.91772223e-01 3.05118531e-01
1.89378619e-01 -2.34954119e-01 -3.99848819e-01 4.42425817e-01] | [9.204514503479004, -6.532102584838867] |
fc71d53b-01fc-4fb2-93aa-4b7f0a266ab7 | learning-deep-visual-object-models-from-noisy | 1702.08513 | null | http://arxiv.org/abs/1702.08513v1 | http://arxiv.org/pdf/1702.08513v1.pdf | Learning Deep Visual Object Models From Noisy Web Data: How to Make it Work | Deep networks thrive when trained on large scale data collections. This has
given ImageNet a central role in the development of deep architectures for
visual object classification. However, ImageNet was created during a specific
period in time, and as such it is prone to aging, as well as dataset bias
issues. Moving beyond fixed training datasets will lead to more robust visual
systems, especially when deployed on robots in new environments which must
train on the objects they encounter there. To make this possible, it is
important to break free from the need for manual annotators. Recent work has
begun to investigate how to use the massive amount of images available on the
Web in place of manual image annotations. We contribute to this research thread
with two findings: (1) a study correlating a given level of noisily labels to
the expected drop in accuracy, for two deep architectures, on two different
types of noise, that clearly identifies GoogLeNet as a suitable architecture
for learning from Web data; (2) a recipe for the creation of Web datasets with
minimal noise and maximum visual variability, based on a visual and natural
language processing concept expansion strategy. By combining these two results,
we obtain a method for learning powerful deep object models automatically from
the Web. We confirm the effectiveness of our approach through object
categorization experiments using our Web-derived version of ImageNet on a
popular robot vision benchmark database, and on a lifelong object discovery
task on a mobile robot. | ['Barbara Caputo', 'Jay Young', 'Nizar Massouh', 'Tatiana Tommasi', 'Nick Hawes', 'Francesca Babiloni'] | 2017-02-28 | learning-deep-visual-object-models-from-noisy-1 | null | null | ieee-xplore-2017-12 | ['object-categorization'] | ['computer-vision'] | [-1.11985654e-01 -1.48139428e-02 4.95508201e-02 -3.30644846e-01
-2.76803255e-01 -7.60384440e-01 7.24937797e-01 3.37265342e-01
-8.20959270e-01 4.82884705e-01 -1.57367066e-01 -1.93159655e-01
-2.12090045e-01 -6.89402342e-01 -9.05336142e-01 -3.91789794e-01
-2.02132434e-01 6.71901047e-01 4.20875788e-01 -3.08942020e-01
2.33356223e-01 4.36827123e-01 -1.97385800e+00 1.85507864e-01
5.09287775e-01 1.29521608e+00 5.54791093e-01 4.86683667e-01
-2.08220974e-01 7.84896612e-01 -4.63468343e-01 -3.78151536e-01
5.43968916e-01 8.46085511e-03 -8.67211819e-01 2.69481957e-01
7.67038167e-01 -2.26701468e-01 -6.09381646e-02 1.14524829e+00
2.86003083e-01 -1.41807608e-02 5.99135756e-01 -1.48949957e+00
-8.17832470e-01 4.22283500e-01 -2.37325013e-01 3.30158591e-01
-1.44654773e-02 6.11947954e-01 8.76503706e-01 -7.78920889e-01
8.66347373e-01 1.23930073e+00 8.26829910e-01 3.90384197e-01
-1.36242163e+00 -5.23119330e-01 5.81765361e-02 2.85603791e-01
-1.21376228e+00 -3.73124331e-01 4.35704231e-01 -7.02380359e-01
9.90261197e-01 -3.01364660e-01 7.61033475e-01 1.45347703e+00
4.31941412e-02 4.48643029e-01 1.11018527e+00 -4.98984367e-01
4.23553228e-01 4.92162466e-01 2.06559420e-01 5.34614563e-01
6.80559158e-01 1.61145374e-01 -2.04801604e-01 1.78095117e-01
4.97710973e-01 -4.27555256e-02 -1.45885766e-01 -9.07938719e-01
-1.00288892e+00 7.91810989e-01 6.41487181e-01 5.20309687e-01
-4.13548023e-01 1.71304196e-01 6.78064764e-01 5.40934086e-01
1.96097568e-01 6.06424749e-01 -6.24621749e-01 -1.91960663e-01
-5.19430518e-01 2.05678135e-01 9.29516435e-01 9.17672038e-01
8.62871468e-01 -8.22787918e-03 3.69543642e-01 8.91181052e-01
2.68717349e-01 4.33523744e-01 8.26784849e-01 -9.92307663e-01
1.63674146e-01 6.61826193e-01 -7.37737119e-02 -9.42674041e-01
-5.37472069e-01 -2.28957817e-01 -4.86961603e-01 8.07710528e-01
6.02395773e-01 1.51745304e-01 -1.15647125e+00 1.70310831e+00
1.12933271e-01 -4.73789573e-01 1.54360279e-01 8.46890688e-01
6.26882493e-01 2.89173812e-01 3.61766219e-01 2.98313707e-01
1.42025173e+00 -8.07175159e-01 -2.44130388e-01 -4.67284113e-01
5.43611646e-01 -3.34596366e-01 1.39642167e+00 6.11450016e-01
-5.48020899e-01 -6.80498183e-01 -1.30846179e+00 -3.67718749e-02
-8.98508251e-01 -1.90191299e-01 6.94555163e-01 5.72748125e-01
-1.21767044e+00 4.21428680e-01 -7.05416679e-01 -9.92106616e-01
6.66314244e-01 2.96578050e-01 -6.68684125e-01 -2.16678381e-01
-8.12668443e-01 1.26749539e+00 7.07120597e-01 -2.06450328e-01
-1.13608539e+00 -2.46714100e-01 -7.84250855e-01 1.01217106e-02
4.01462346e-01 -4.41379756e-01 1.25582469e+00 -1.68519568e+00
-9.80135262e-01 1.04819500e+00 3.94868433e-01 -7.69164741e-01
4.58030403e-01 -9.34013873e-02 -1.25955954e-01 8.12770501e-02
2.27055669e-01 1.10368478e+00 7.73660839e-01 -1.61079144e+00
-7.06808925e-01 -4.66069728e-01 2.54159182e-01 -3.82420421e-02
-6.28390431e-01 -2.67881244e-01 -3.71512085e-01 -1.73775300e-01
-2.41603702e-01 -1.03528631e+00 -1.32612184e-01 1.25872448e-01
1.50348529e-01 -3.71106893e-01 9.10113275e-01 -4.52951729e-01
3.72984231e-01 -2.19075227e+00 -1.45469472e-01 7.36042112e-02
3.77191722e-01 3.95469755e-01 -2.25446820e-01 2.32841894e-01
-2.28378419e-02 2.77015597e-01 -7.29525313e-02 -4.28208672e-02
-8.14854130e-02 3.41567695e-01 -1.16315834e-01 1.81236148e-01
2.35906720e-01 9.27581787e-01 -9.70606625e-01 -4.75323588e-01
1.68962583e-01 2.08601445e-01 -3.74658495e-01 -1.81138180e-02
-2.59598881e-01 -6.28393218e-02 1.10654540e-01 6.49482608e-01
4.12389368e-01 -3.62526864e-01 2.38344774e-01 -5.75231910e-02
-8.06562304e-02 -1.17856516e-02 -8.55462968e-01 1.68147814e+00
-4.18307990e-01 9.83503401e-01 1.65470690e-01 -1.15117264e+00
9.24790204e-01 1.54897824e-01 4.85864818e-01 -1.00775301e+00
2.71247894e-01 2.70560026e-01 1.45420596e-01 -7.66429484e-01
4.05452430e-01 -1.16976067e-01 1.53574273e-01 2.13930935e-01
5.89515209e-01 -3.83312814e-02 5.26857495e-01 2.23059416e-01
1.25742280e+00 -5.60314022e-03 1.38751209e-01 -3.90226990e-01
-1.32065313e-02 4.76366460e-01 1.41480729e-01 8.30928087e-01
-4.41303074e-01 4.66159582e-01 2.93640614e-01 -7.35412538e-01
-1.34326291e+00 -9.56279159e-01 -2.37426236e-01 1.05270445e+00
1.01760194e-01 -2.55644679e-01 -5.23164272e-01 -8.23918402e-01
6.07414246e-02 4.47039425e-01 -6.56220376e-01 -1.64091572e-01
-2.01535687e-01 -5.75763941e-01 4.60821986e-01 4.43711132e-01
5.65710008e-01 -1.30611908e+00 -9.67936039e-01 2.43741676e-01
6.96708113e-02 -1.19328356e+00 2.43621245e-01 6.54815018e-01
-7.08170772e-01 -1.30854607e+00 -5.44258833e-01 -9.74035800e-01
5.28668165e-01 4.68510598e-01 1.36853254e+00 1.23271957e-01
-4.02996629e-01 8.56391191e-01 -4.89693046e-01 -7.26524711e-01
-5.42640030e-01 1.89352706e-01 2.15616837e-01 -3.48240554e-01
6.73355103e-01 -6.64385021e-01 -3.44572276e-01 2.35259116e-01
-1.14869010e+00 -2.48404935e-01 9.18272257e-01 8.31429362e-01
8.27251971e-02 1.06834792e-01 6.26576364e-01 -4.39300239e-01
6.23336852e-01 -7.87783384e-01 -7.38798440e-01 3.80079225e-02
-9.51117814e-01 9.09732506e-02 4.90683436e-01 -6.88941002e-01
-6.49213672e-01 1.62641734e-01 4.37635966e-02 -3.70981812e-01
-3.96316260e-01 4.28730637e-01 2.90779565e-02 -1.75377473e-01
1.27585649e+00 1.09845446e-02 2.62349576e-01 -3.30826581e-01
4.53327745e-01 6.96389437e-01 5.20601690e-01 -4.53223825e-01
6.54921353e-01 5.10554254e-01 -3.27692628e-01 -1.12065160e+00
-4.24416393e-01 -5.28559864e-01 -7.52418697e-01 -1.89321518e-01
8.72573912e-01 -9.61704433e-01 -5.84074497e-01 4.77152854e-01
-9.52535570e-01 -6.52357876e-01 -3.27782035e-01 3.37603480e-01
-3.95301044e-01 3.50036621e-01 -1.76036313e-01 -5.84648430e-01
-9.41062495e-02 -9.89081264e-01 7.17324317e-01 1.79642886e-01
-1.35180175e-01 -8.60421538e-01 1.25601143e-01 1.44897312e-01
4.74999547e-01 1.09346941e-01 8.30657959e-01 -7.93434620e-01
-6.33267462e-01 -1.78757578e-01 -5.13292134e-01 8.66201937e-01
-6.30883574e-02 -9.27960426e-02 -9.31777239e-01 -3.41203481e-01
3.68391387e-02 -9.06476498e-01 8.82297754e-01 -6.86469674e-02
7.85941601e-01 -2.49102846e-01 -2.53971785e-01 3.03001046e-01
1.75842571e+00 1.23324804e-01 5.14020979e-01 8.90261114e-01
5.89350522e-01 5.95230579e-01 2.69603938e-01 8.62255841e-02
5.04902303e-01 6.47701621e-01 8.03053617e-01 4.90178838e-02
-2.60611504e-01 -2.75703091e-02 3.11712503e-01 3.02414060e-01
1.24703877e-01 -1.48821056e-01 -1.21679926e+00 8.44832718e-01
-1.77565742e+00 -7.54865825e-01 1.14391804e-01 2.19614959e+00
6.21097684e-01 4.04378861e-01 2.65881747e-01 -4.86223167e-03
3.20236504e-01 -3.21538419e-01 -4.85794127e-01 -2.28175208e-01
-8.70602056e-02 -4.61604446e-02 6.40710831e-01 2.00340468e-02
-1.12655044e+00 8.33038390e-01 6.18743753e+00 4.18326676e-01
-1.31126833e+00 1.70654252e-01 4.35971469e-01 4.90610376e-02
1.76840812e-01 -1.23690307e-01 -6.87782168e-01 5.05444944e-01
1.11296475e+00 3.60046811e-02 4.14121807e-01 1.46743631e+00
-1.63395733e-01 -2.34544635e-01 -1.22401798e+00 1.11248350e+00
1.99905798e-01 -1.14786828e+00 4.97945957e-03 1.37441173e-01
4.57219154e-01 5.90500116e-01 -2.89555322e-02 5.36671758e-01
5.60836613e-01 -1.01855338e+00 9.80373383e-01 1.25469893e-01
3.79669011e-01 -3.22859436e-01 6.82782650e-01 2.84520030e-01
-7.83901453e-01 -4.59038764e-01 -4.68242288e-01 -7.81384930e-02
-2.72778898e-01 2.57057577e-01 -1.18385375e+00 -2.75896844e-02
1.24124980e+00 5.29273748e-01 -1.09335995e+00 1.27552366e+00
5.28451167e-02 4.25187171e-01 -4.04023707e-01 -8.07020590e-02
3.02803278e-01 1.91652447e-01 3.26225966e-01 1.11785531e+00
9.27679166e-02 -3.62740368e-01 1.18701503e-01 6.94921553e-01
-1.06762461e-01 -3.90721820e-02 -1.04251420e+00 -3.12838368e-02
3.94084007e-01 1.49666154e+00 -1.03250265e+00 -3.33871931e-01
-5.10422349e-01 7.72216141e-01 6.15128279e-01 2.41495967e-01
-5.63121557e-01 -2.47039795e-01 5.76665699e-01 1.70414224e-01
3.57080519e-01 -6.05049610e-01 -3.63618910e-01 -9.26975131e-01
1.31047904e-01 -1.15121627e+00 1.95955724e-01 -8.94834757e-01
-1.34740770e+00 6.27592504e-01 1.63630024e-02 -1.00336015e+00
-1.72616333e-01 -1.04697955e+00 -8.94387290e-02 3.55849743e-01
-1.34249282e+00 -1.03518224e+00 -5.84305167e-01 5.22526383e-01
6.13159180e-01 -1.27724245e-01 6.95356965e-01 3.44858468e-01
-8.66060331e-03 3.23045433e-01 9.06756148e-02 1.74630716e-01
6.41975164e-01 -1.25945413e+00 3.77166122e-01 6.41027689e-01
5.80876768e-01 4.23493266e-01 6.13367975e-01 -4.30522829e-01
-1.31310797e+00 -8.36130977e-01 4.05610055e-01 -1.01332235e+00
6.84803903e-01 -5.50018370e-01 -9.75355208e-01 6.96952105e-01
2.20430478e-01 -2.78861701e-01 2.26588950e-01 2.81774074e-01
-5.70618272e-01 -3.58886570e-01 -1.01152837e+00 4.78272438e-01
9.76781070e-01 -3.50165814e-01 -7.21958399e-01 3.74991983e-01
7.52258420e-01 6.02584984e-03 -5.57347059e-01 2.31032163e-01
5.92431903e-01 -1.00753462e+00 9.56456065e-01 -4.68723118e-01
3.61066759e-01 -1.72146380e-01 -1.93561599e-01 -1.30707181e+00
-3.14850539e-01 -3.16211432e-02 3.34681302e-01 1.20534015e+00
4.66667563e-01 -5.28345764e-01 6.44267023e-01 6.11263394e-01
-2.17720032e-01 -3.10882151e-01 -8.36085439e-01 -1.09704804e+00
-3.49060930e-02 -6.39835596e-01 1.08201265e-01 8.23640168e-01
-3.07059854e-01 4.50482756e-01 -3.10988575e-02 -7.53850341e-02
4.57487643e-01 -5.23777962e-01 1.04047012e+00 -1.63395369e+00
-2.22738124e-02 -5.23740053e-01 -8.38931739e-01 -4.84755367e-01
2.54860613e-02 -8.28315914e-01 2.58435100e-01 -1.46294475e+00
2.57565826e-01 -4.63186771e-01 -2.20979109e-01 7.76977897e-01
2.50317156e-01 4.78235483e-01 3.11408579e-01 3.81475478e-01
-8.21885347e-01 2.38580033e-01 6.66336894e-01 -3.65877569e-01
-4.09552492e-02 -4.28923398e-01 -7.97972381e-01 7.97661185e-01
7.33407915e-01 -4.60449338e-01 -4.42865878e-01 -4.87155676e-01
4.21563923e-01 -8.84760678e-01 6.08956754e-01 -1.49028814e+00
-4.95294817e-02 8.49228576e-02 4.66258228e-01 -1.11645311e-01
2.10452601e-01 -1.28704703e+00 -3.73632126e-02 3.84131312e-01
-1.41117826e-01 1.53084630e-02 4.42566723e-01 6.70139074e-01
-2.91600414e-02 -3.24096233e-01 8.55926812e-01 -5.07648647e-01
-1.42202079e+00 9.00725275e-02 -4.39943373e-01 2.69785933e-02
9.91816878e-01 -3.15691739e-01 -4.55892712e-01 -2.29959488e-01
-4.92843479e-01 1.79531768e-01 8.32107365e-01 1.02349222e+00
3.41391176e-01 -1.16180170e+00 -3.93046796e-01 1.26957163e-01
4.84818727e-01 -6.44421577e-02 -1.36151031e-01 5.66574991e-01
-6.29525661e-01 3.03749114e-01 -6.21047497e-01 -8.55141521e-01
-1.03859198e+00 9.88369763e-01 2.44934693e-01 -4.33673970e-02
-6.64680123e-01 6.10510349e-01 1.21125683e-01 -2.88036466e-01
3.73849243e-01 -4.36224014e-01 -9.57020819e-02 1.67317346e-01
3.58174145e-01 6.70495331e-02 2.26429895e-01 -4.92365867e-01
-2.19266161e-01 1.77905053e-01 -1.21751606e-01 -1.59636050e-01
1.39110994e+00 -2.55430698e-01 -2.23388895e-02 6.39592767e-01
1.08317316e+00 -4.30208713e-01 -1.46844780e+00 -9.11626741e-02
4.51172888e-01 -3.12496930e-01 -1.52276814e-01 -1.00593936e+00
-7.73630083e-01 7.17168808e-01 1.12503767e+00 6.05429769e-01
7.55767643e-01 2.13914499e-01 4.00442481e-01 7.76568949e-01
5.71573198e-01 -1.20932078e+00 5.59882879e-01 6.78944290e-01
8.55572462e-01 -1.73834920e+00 -1.94504470e-01 3.09443295e-01
-5.46738029e-01 1.07683420e+00 7.97520518e-01 -6.64073601e-02
5.09292960e-01 5.81263490e-02 2.65371740e-01 -5.50813854e-01
-6.64595068e-01 -4.82248962e-01 1.39965452e-02 1.05928409e+00
6.15409389e-02 -2.36674219e-01 -6.94323182e-02 3.78931075e-01
-1.58999875e-01 1.28916949e-01 5.56551814e-01 9.71541166e-01
-8.01783144e-01 -9.31739271e-01 -3.14153820e-01 3.38915974e-01
-2.15188444e-01 -7.75582194e-02 -5.29488206e-01 1.13418329e+00
4.21091795e-01 8.74872327e-01 1.20954253e-01 -3.96374613e-01
2.67424941e-01 3.48532885e-01 3.32050711e-01 -6.01351798e-01
-3.57538670e-01 -5.70896983e-01 4.34462950e-02 -5.11512935e-01
-3.27275008e-01 -4.37385529e-01 -9.77078855e-01 -1.01522366e-02
-1.86684087e-01 -7.33562186e-02 1.15939295e+00 7.58971930e-01
3.87029201e-01 2.70546079e-01 1.55948088e-01 -1.24004996e+00
-5.46486735e-01 -1.05361164e+00 -4.48378503e-01 7.77573586e-01
1.76297083e-01 -7.82153666e-01 -4.30914968e-01 2.68104672e-01] | [9.632906913757324, 2.3769278526306152] |
08486e62-d03c-4423-9539-411d8671a231 | sgva-clip-semantic-guided-visual-adapting-of | 2211.16191 | null | https://arxiv.org/abs/2211.16191v2 | https://arxiv.org/pdf/2211.16191v2.pdf | SgVA-CLIP: Semantic-guided Visual Adapting of Vision-Language Models for Few-shot Image Classification | Although significant progress has been made in few-shot learning, most of existing few-shot image classification methods require supervised pre-training on a large amount of samples of base classes, which limits their generalization ability in real world application. Recently, large-scale Vision-Language Pre-trained models (VLPs) have been gaining increasing attention in few-shot learning because they can provide a new paradigm for transferable visual representation learning with easily available text on the Web. However, the VLPs may neglect detailed visual information that is difficult to describe by language sentences, but important for learning an effective classifier to distinguish different images. To address the above problem, we propose a new framework, named Semantic-guided Visual Adapting (SgVA), which can effectively extend vision-language pre-trained models to produce discriminative adapted visual features by comprehensively using an implicit knowledge distillation, a vision-specific contrastive loss, and a cross-modal contrastive loss. The implicit knowledge distillation is designed to transfer the fine-grained cross-modal knowledge to guide the updating of the vision adapter. State-of-the-art results on 13 datasets demonstrate that the adapted visual features can well complement the cross-modal features to improve few-shot image classification. | ['Changsheng Xu', 'YaoWei Wang', 'Linhui Xiao', 'Xiaoshan Yang', 'Fang Peng'] | 2022-11-28 | null | null | null | null | ['few-shot-image-classification'] | ['computer-vision'] | [ 2.84196645e-01 -4.39824194e-01 -4.61812586e-01 -4.58256811e-01
-6.10778272e-01 -7.92406797e-02 7.63638914e-01 -5.74619286e-02
-4.56773132e-01 5.70287764e-01 2.58269757e-01 2.01840788e-01
6.23713769e-02 -9.77925599e-01 -6.37292266e-01 -6.90244198e-01
4.54728991e-01 6.44035786e-02 6.51354313e-01 -3.41373146e-01
1.00815415e-01 4.53325100e-02 -1.96302783e+00 3.69515389e-01
8.91152978e-01 1.19533408e+00 5.83817959e-01 3.30218583e-01
-6.47008181e-01 1.07850122e+00 -2.74793684e-01 -2.09590316e-01
-7.27057159e-02 -6.53904140e-01 -6.09582722e-01 1.19268499e-01
4.01696712e-01 -4.06497359e-01 -5.41103184e-01 1.30330956e+00
5.12934923e-01 4.98925209e-01 8.25934649e-01 -1.25554061e+00
-1.29537880e+00 2.60103583e-01 -4.17739153e-01 3.26682091e-01
1.40522793e-01 4.73073572e-01 9.27958906e-01 -1.00533378e+00
6.43796325e-01 1.33383238e+00 4.04141605e-01 8.90087187e-01
-9.35375452e-01 -6.45429850e-01 2.05824569e-01 9.50630367e-01
-1.29175985e+00 -3.24570984e-01 9.25571859e-01 -5.33031523e-01
8.91246855e-01 -6.15683608e-02 7.76014864e-01 1.20235741e+00
-3.91974598e-02 8.50265086e-01 1.10254157e+00 -5.36921382e-01
2.74361223e-01 3.01466823e-01 2.42029324e-01 9.18293834e-01
8.90409425e-02 9.62111354e-02 -5.66078961e-01 3.41854692e-01
4.72611308e-01 6.18913114e-01 -3.82768452e-01 -7.19642401e-01
-7.67225802e-01 1.07034886e+00 8.51684391e-01 2.84149766e-01
-8.32671225e-02 -4.81123589e-02 6.43851697e-01 2.43778259e-01
3.40573043e-01 1.27183422e-01 -1.98834136e-01 8.51460770e-02
-6.51459217e-01 -2.33897761e-01 4.14066762e-01 1.00282562e+00
1.27090597e+00 2.16519848e-01 -7.57723272e-01 1.17939961e+00
1.75224558e-01 6.59300685e-01 9.45357025e-01 -6.10821366e-01
2.83844531e-01 7.69116640e-01 -4.36950684e-01 -7.28350282e-01
9.03748274e-02 -1.42109796e-01 -1.03134155e+00 1.31751984e-01
-4.01595607e-02 2.26566806e-01 -1.35601914e+00 1.61595893e+00
9.71583202e-02 3.74829888e-01 7.15346709e-02 8.59341741e-01
1.13987708e+00 7.46941686e-01 2.27873117e-01 -3.43172550e-01
1.35037887e+00 -1.20994902e+00 -6.98992550e-01 -4.01594013e-01
3.74017090e-01 -4.44750011e-01 1.45299256e+00 -1.29067108e-01
-5.41471362e-01 -7.74595618e-01 -1.26489353e+00 -2.81472355e-01
-7.69987822e-01 -4.59112495e-01 5.53855240e-01 4.18620050e-01
-6.01681232e-01 3.93880665e-01 -4.84256506e-01 -4.78453457e-01
1.05233443e+00 -1.53859690e-01 -2.78988570e-01 -8.29959393e-01
-1.36517859e+00 8.64113510e-01 6.19210422e-01 -3.13584834e-01
-1.19082189e+00 -6.63175344e-01 -1.32163155e+00 2.08819494e-01
6.09713078e-01 -7.83409774e-01 1.00122941e+00 -9.29117620e-01
-1.46568084e+00 8.80992055e-01 -6.33843467e-02 -2.90703863e-01
3.76920164e-01 5.13203628e-02 -5.21334052e-01 3.26252788e-01
2.25981101e-01 6.73365116e-01 1.16488266e+00 -1.16901457e+00
-5.10458410e-01 -3.39066565e-01 7.53665119e-02 1.93791881e-01
-8.07382703e-01 -1.91760346e-01 -6.30949080e-01 -6.00358427e-01
-3.53782386e-01 -5.35736084e-01 -9.61123407e-02 3.10780346e-01
-1.73892379e-02 -3.73400271e-01 9.52841222e-01 -4.11443323e-01
9.42754269e-01 -2.10221076e+00 3.81905138e-02 -3.04437608e-01
3.50235850e-02 6.79522574e-01 -5.30641019e-01 2.55778939e-01
2.51393795e-01 -2.86692709e-01 -2.74601728e-01 -1.63886070e-01
-4.04494628e-02 4.20608342e-01 -3.61141652e-01 1.28794298e-01
2.23594487e-01 1.25405586e+00 -1.20494819e+00 -6.98214769e-01
6.13065362e-01 5.47361612e-01 -3.38871539e-01 3.31169784e-01
-2.34581232e-01 2.34899428e-02 -3.51937324e-01 6.05043471e-01
4.65277046e-01 -3.77659798e-01 -3.41149539e-01 -2.01966852e-01
1.70235693e-01 -3.56791317e-01 -7.25458860e-01 1.96600091e+00
-5.29100716e-01 5.72057068e-01 -4.41276997e-01 -1.33711934e+00
9.08718348e-01 -1.21140979e-01 1.48287937e-01 -1.00930357e+00
2.55240023e-01 -2.11842462e-01 -1.84297025e-01 -6.97760344e-01
-1.99420061e-02 -4.65548337e-01 5.29717244e-02 1.79640785e-01
5.23560286e-01 -1.81497976e-01 2.52886027e-01 2.80467063e-01
8.48391354e-01 -8.73261392e-02 5.34240961e-01 1.68615252e-01
6.10873580e-01 -9.29800197e-02 7.06090689e-01 7.42104292e-01
-5.36045134e-01 6.55205309e-01 -1.08654201e-01 -3.10329974e-01
-7.69282579e-01 -1.07821929e+00 -6.29026666e-02 1.35972786e+00
4.01276350e-01 -3.76279831e-01 -5.37032902e-01 -8.87148142e-01
-4.00713757e-02 7.49749660e-01 -8.42939198e-01 -8.60641420e-01
-4.23176102e-02 -3.68968934e-01 5.01223579e-02 6.54124320e-01
8.38547766e-01 -1.24421704e+00 -6.12641633e-01 -1.21404463e-02
1.34770088e-02 -1.12454224e+00 -5.69706917e-01 1.16724305e-01
-5.73449969e-01 -1.16538846e+00 -1.06064200e+00 -9.41793740e-01
6.43718898e-01 8.69914412e-01 7.81018436e-01 -5.55781871e-02
-7.93442845e-01 6.62240267e-01 -6.84702933e-01 -6.20325804e-01
-8.13436136e-02 -3.43046904e-01 9.02349781e-03 1.38804197e-01
8.32119644e-01 -4.73495573e-01 -5.33324659e-01 4.85639162e-02
-9.60771739e-01 4.48299386e-02 6.88957334e-01 1.35490954e+00
6.68786764e-01 -2.99450476e-02 4.81898814e-01 -7.62898803e-01
3.26626301e-01 -3.84220213e-01 -1.47528142e-01 5.48319042e-01
-6.04748189e-01 2.23122418e-01 8.38321388e-01 -6.64873064e-01
-1.11869597e+00 -2.12501705e-01 1.06342584e-01 -1.03952777e+00
-1.25360206e-01 4.69723552e-01 -3.13492507e-01 -2.94222742e-01
6.03079915e-01 7.91892231e-01 5.39446287e-02 -3.20124954e-01
7.81051993e-01 6.57547772e-01 3.59167725e-01 -2.27566555e-01
8.61228287e-01 4.37231094e-01 -1.46489426e-01 -1.02973831e+00
-1.29514849e+00 -7.28316128e-01 -6.91207826e-01 -1.40125658e-02
9.22628999e-01 -1.03682196e+00 -7.37123713e-02 4.93965775e-01
-9.17733848e-01 -1.65932998e-01 -5.76882541e-01 4.00060266e-01
-6.47409141e-01 5.33360302e-01 -3.01945627e-01 -5.46883345e-01
-3.02079707e-01 -9.63969171e-01 7.45058954e-01 4.54768389e-01
2.32199103e-01 -8.97972047e-01 9.50281918e-02 3.58486325e-01
4.41629678e-01 -2.01352015e-01 1.12040484e+00 -5.02757967e-01
-4.43247318e-01 -1.23904563e-01 -6.04558945e-01 7.16062605e-01
2.10117891e-01 -4.49068218e-01 -1.09201992e+00 -3.44569266e-01
-9.46816057e-02 -9.28266466e-01 1.44725144e+00 1.77411199e-01
1.28197455e+00 -9.52953994e-02 -2.73191690e-01 6.69083059e-01
1.64945209e+00 2.79977825e-03 5.56540430e-01 1.06362484e-01
9.32860374e-01 2.76590079e-01 6.31426632e-01 3.87462795e-01
3.77142638e-01 5.88768363e-01 2.98814923e-01 2.92901456e-01
-6.29985154e-01 -4.89766479e-01 3.23791116e-01 9.14041221e-01
-1.61897182e-01 3.70243639e-01 -6.87938571e-01 5.10545492e-01
-1.97441697e+00 -1.27444029e+00 6.67362392e-01 1.95539832e+00
8.78526747e-01 2.25440748e-02 -1.57032743e-01 -1.86293483e-01
7.80230701e-01 5.03238678e-01 -9.36833978e-01 -1.99704561e-02
-3.67526077e-02 6.24192618e-02 1.41326472e-01 4.99954149e-02
-1.08185220e+00 1.11799324e+00 5.58120632e+00 1.20660031e+00
-1.20621979e+00 4.53476667e-01 1.93966299e-01 -6.27999827e-02
-2.88131446e-01 -6.87606558e-02 -8.21539640e-01 4.68843937e-01
5.84433496e-01 -4.69801784e-01 3.26477110e-01 1.20806575e+00
-2.60465324e-01 1.50971532e-01 -9.05232728e-01 1.54634905e+00
6.28739536e-01 -1.50332153e+00 6.35561824e-01 -1.89794064e-01
8.35332036e-01 8.87725055e-02 -9.08621773e-03 1.01470852e+00
2.03513116e-01 -8.53690624e-01 3.17855358e-01 7.24197924e-01
1.07330084e+00 -6.27091587e-01 6.41769111e-01 3.97052705e-01
-1.32046735e+00 -4.01557297e-01 -9.90249753e-01 -1.17571868e-01
7.07081705e-02 2.34507486e-01 -4.32235897e-01 4.43786412e-01
7.32807457e-01 1.11132240e+00 -7.34739482e-01 1.09262836e+00
-2.11872295e-01 3.09205413e-01 2.70074636e-01 -2.58264810e-01
3.58710110e-01 5.34589635e-03 3.31526846e-01 9.63522971e-01
2.05013499e-01 2.11848930e-01 4.11661386e-01 7.87570477e-01
-2.05373317e-01 1.92101300e-01 -7.70282209e-01 -2.44726047e-01
2.69567281e-01 1.13228965e+00 -4.83126283e-01 -5.83597600e-01
-9.05674458e-01 1.15595222e+00 6.52384996e-01 4.69500482e-01
-5.53554952e-01 -6.98739767e-01 6.05020106e-01 -1.10351652e-01
5.82949460e-01 7.31920600e-02 1.65299982e-01 -1.59589911e+00
-2.03288540e-01 -5.70039630e-01 4.88000602e-01 -7.82931685e-01
-1.65117931e+00 3.29500616e-01 -1.21227138e-01 -1.40072894e+00
-1.44221142e-01 -6.93665266e-01 -7.00758159e-01 6.72097743e-01
-1.83811820e+00 -1.47878444e+00 -6.51266158e-01 1.05205250e+00
9.96560872e-01 -6.26318216e-01 8.80816758e-01 8.74398649e-02
-4.89957720e-01 6.78379834e-01 1.65701061e-01 2.32414365e-01
8.54176342e-01 -9.69748557e-01 -2.06400588e-01 7.92417347e-01
3.30060810e-01 4.38728899e-01 4.15377200e-01 -3.40960830e-01
-1.40299523e+00 -1.34388089e+00 3.01311374e-01 -2.65693605e-01
5.91524839e-01 -2.28240877e-01 -1.12005603e+00 4.43315834e-01
2.63081402e-01 6.26680374e-01 7.41536379e-01 4.28509489e-02
-8.70807946e-01 -3.40690762e-01 -8.54013026e-01 4.68414485e-01
1.21948242e+00 -9.10554826e-01 -1.03479028e+00 2.53691554e-01
9.39527392e-01 2.03836307e-01 -3.81955862e-01 2.92155236e-01
3.77991319e-01 -7.41098464e-01 1.05716050e+00 -9.09857452e-01
5.21171808e-01 -1.35570884e-01 -4.18911964e-01 -1.45695674e+00
-5.65652549e-01 -7.32198358e-02 -3.54090124e-01 1.27536476e+00
-1.92955554e-01 -3.18366200e-01 4.52783793e-01 3.21680397e-01
-1.49037808e-01 -5.98308027e-01 -7.15596259e-01 -1.04020488e+00
4.44435067e-02 -3.23787540e-01 2.04947978e-01 1.05224705e+00
-6.00140207e-02 7.95041800e-01 -5.65832138e-01 -3.33398044e-01
8.30745816e-01 2.67714202e-01 6.30783856e-01 -1.37333429e+00
-1.59652755e-01 -4.78230715e-01 -8.25394094e-01 -8.81919324e-01
4.84015495e-01 -1.13326156e+00 1.49220169e-01 -1.80519187e+00
8.00378025e-01 7.91618153e-02 -7.76334941e-01 6.47059798e-01
-3.25485915e-01 2.74643123e-01 3.80997062e-01 1.45883724e-01
-1.08690071e+00 1.24941623e+00 1.49402142e+00 -7.05954373e-01
6.66398257e-02 -3.52213472e-01 -7.19576359e-01 7.22253919e-01
2.72719741e-01 -3.36133003e-01 -7.74315000e-01 -1.17841415e-01
-3.33819270e-01 -3.63636762e-01 4.85460818e-01 -1.12686646e+00
3.58552516e-01 -5.36355913e-01 5.89250028e-01 -3.25000674e-01
4.50729311e-01 -6.57278001e-01 -5.78972578e-01 4.63818371e-01
-2.86468863e-01 -7.25357473e-01 -2.65674293e-02 9.72988367e-01
-4.33889955e-01 -2.98683703e-01 1.09392238e+00 -3.73244822e-01
-1.80040216e+00 8.50614190e-01 1.13397665e-01 3.65180224e-01
1.27075028e+00 -2.11726427e-01 -4.54350203e-01 -1.60753220e-01
-5.52491546e-01 2.37366855e-01 3.87078077e-01 7.93791294e-01
9.99243736e-01 -1.62229955e+00 -4.33662981e-01 2.83521593e-01
9.62218404e-01 -3.08864206e-01 8.01184654e-01 4.12071377e-01
-3.98702808e-02 2.89892077e-01 -6.68604374e-01 -5.24124801e-01
-1.16923749e+00 1.13559747e+00 9.36762914e-02 7.65830949e-02
-8.97706032e-01 8.83085847e-01 4.44978714e-01 -2.23933354e-01
1.41428128e-01 2.28346717e-02 -3.51637125e-01 3.00678104e-01
1.11980355e+00 2.14217946e-01 -3.42695177e-01 -5.46220005e-01
-3.06224912e-01 7.96678960e-01 -3.04484159e-01 2.24140123e-01
1.33027327e+00 -3.17640126e-01 1.46952599e-01 8.47576380e-01
1.41338384e+00 -6.10726595e-01 -1.44092882e+00 -8.82391334e-01
-3.77229989e-01 -6.10687494e-01 2.38936186e-01 -6.55519128e-01
-9.38632309e-01 1.46205294e+00 7.74592161e-01 -5.60222492e-02
1.01884544e+00 8.39440003e-02 8.86373162e-01 7.04685092e-01
5.03613830e-01 -1.25159168e+00 5.48964918e-01 6.86814606e-01
7.39435077e-01 -1.80899417e+00 -1.83741555e-01 -2.24132970e-01
-8.32864821e-01 9.98297453e-01 8.36671174e-01 -1.74016580e-02
8.10102284e-01 -4.43211496e-01 -2.40698457e-02 1.02504129e-02
-6.44292235e-01 -6.85757995e-01 5.91666281e-01 1.04029655e+00
-8.60869065e-02 -1.45680517e-01 5.25790863e-02 8.13892305e-01
3.26943666e-01 2.40592480e-01 2.10507512e-01 7.48713911e-01
-8.94458830e-01 -6.86792612e-01 1.27877086e-01 5.79626083e-01
2.08659872e-01 -2.56156117e-01 -7.35700727e-02 5.53842366e-01
7.27137551e-02 6.86740994e-01 -2.61321925e-02 -3.58323246e-01
2.71059781e-01 2.48409942e-01 7.51991034e-01 -1.04281366e+00
2.90248424e-01 -3.46161127e-01 -3.93736452e-01 -4.95483994e-01
-4.21141744e-01 -1.73507631e-01 -1.06185806e+00 9.46617126e-03
-5.14757335e-02 -1.15804888e-01 1.74244002e-01 1.31558692e+00
2.67130464e-01 5.93610287e-01 5.84543526e-01 -8.11653078e-01
-7.59919584e-01 -8.85309219e-01 -8.20810020e-01 8.48844886e-01
3.21346074e-01 -1.10916972e+00 -3.26039970e-01 7.46219233e-02] | [10.023812294006348, 2.571859121322632] |
9bb06567-425a-4633-90dd-831897f97fba | crisisltlsum-a-benchmark-for-local-crisis | 2210.14190 | null | https://arxiv.org/abs/2210.14190v1 | https://arxiv.org/pdf/2210.14190v1.pdf | CrisisLTLSum: A Benchmark for Local Crisis Event Timeline Extraction and Summarization | Social media has increasingly played a key role in emergency response: first responders can use public posts to better react to ongoing crisis events and deploy the necessary resources where they are most needed. Timeline extraction and abstractive summarization are critical technical tasks to leverage large numbers of social media posts about events. Unfortunately, there are few datasets for benchmarking technical approaches for those tasks. This paper presents CrisisLTLSum, the largest dataset of local crisis event timelines available to date. CrisisLTLSum contains 1,000 crisis event timelines across four domains: wildfires, local fires, traffic, and storms. We built CrisisLTLSum using a semi-automated cluster-then-refine approach to collect data from the public Twitter stream. Our initial experiments indicate a significant gap between the performance of strong baselines compared to the human performance on both tasks. Our dataset, code, and models are publicly available. | ['Alejandro Jaimes', 'Joel Tetreault', 'Shihao Ran', 'Ke Zhang', 'Bashar Alhafni', 'Hossein Rajaby Faghihi'] | 2022-10-25 | null | null | null | null | ['abstractive-text-summarization'] | ['natural-language-processing'] | [-2.12666709e-02 1.99054480e-02 5.41763194e-02 -4.30899054e-01
-1.47698188e+00 -8.59514117e-01 8.49259853e-01 1.31639671e+00
-8.56187463e-01 7.43670344e-01 1.11609149e+00 -2.86109388e-01
1.42676802e-02 -6.73636854e-01 -1.13317356e-01 -3.18888634e-01
-5.04917800e-01 7.17361391e-01 1.17375471e-01 -6.40121877e-01
5.34699202e-01 5.61845839e-01 -1.00395823e+00 8.10587108e-01
2.90972114e-01 7.69784987e-01 -2.30164319e-01 9.28364456e-01
-9.58337560e-02 1.16821754e+00 -8.54983628e-01 3.45824748e-01
3.11523434e-02 -2.25932226e-02 -9.31379914e-01 -3.39367956e-01
-1.17649406e-01 -2.97212243e-01 -3.56061250e-01 4.14084613e-01
1.03552818e+00 2.36554876e-01 5.47654212e-01 -1.27825701e+00
4.78518844e-01 1.02084661e+00 -5.45637727e-01 1.11333466e+00
8.44616771e-01 1.79178953e-01 8.33182156e-01 -7.95333207e-01
7.54981518e-01 1.07608390e+00 1.00568020e+00 -1.30622208e-01
-1.02187121e+00 -6.98210955e-01 1.22136891e-01 -8.26190487e-02
-1.23057485e+00 -7.86183119e-01 1.95962951e-01 -7.64134169e-01
1.60892248e+00 3.83992612e-01 1.82124808e-01 1.08045006e+00
-5.88632859e-02 6.25431657e-01 8.46777022e-01 -5.39602479e-04
1.79677844e-01 -5.53704500e-01 3.29913765e-01 -3.42675224e-02
-2.55740955e-02 -3.16793501e-01 -8.39236617e-01 -6.07873976e-01
-5.33167161e-02 2.90341407e-01 -4.06583518e-01 9.19690371e-01
-1.41873300e+00 7.10042417e-01 3.62478375e-01 3.18276733e-01
-8.68456185e-01 -2.93781757e-01 8.70320082e-01 2.27840275e-01
1.05170679e+00 7.56895006e-01 -3.67266357e-01 -4.28105384e-01
-1.26464152e+00 6.52216256e-01 9.20455098e-01 6.23105109e-01
4.37819898e-01 -3.26820791e-01 -1.12293102e-01 4.47567463e-01
-3.51331890e-01 8.09514284e-01 1.80121500e-03 -6.75478935e-01
1.35980773e+00 5.51943123e-01 3.74676108e-01 -1.53154790e+00
-1.28126848e+00 1.31139919e-01 -9.16852653e-01 -3.24696839e-01
3.24404985e-01 -7.51318693e-01 -4.67045814e-01 1.21307361e+00
2.81132817e-01 1.52866855e-01 -2.87700593e-02 5.99212229e-01
1.00147069e+00 9.95590210e-01 1.88101098e-01 -4.60566252e-01
1.12780845e+00 -1.22249790e-01 -8.84524524e-01 -4.05952424e-01
5.78552604e-01 -7.78225601e-01 5.93599916e-01 3.16547871e-01
-9.71979737e-01 1.38336554e-01 -5.98593235e-01 1.87040374e-01
-7.06239760e-01 -5.69915295e-01 3.39526892e-01 -2.08399504e-01
-7.41680264e-01 5.79853594e-01 -7.87323654e-01 -7.66949654e-01
3.14156532e-01 -2.70809233e-01 -5.66633463e-01 2.43911222e-01
-1.24880195e+00 9.50748742e-01 6.04020953e-01 1.79328583e-02
-7.35768080e-01 -6.88857734e-01 -7.20785022e-01 -1.42115533e-01
2.98283637e-01 4.22055721e-02 1.56056404e+00 1.74382910e-01
-4.33301270e-01 6.67221010e-01 1.31362572e-01 -6.33196950e-01
4.48493093e-01 -4.15599674e-01 -3.63643885e-01 4.97816294e-01
5.27750969e-01 2.19314203e-01 3.02219629e-01 -7.38812089e-01
-1.17981172e+00 -1.72148839e-01 -2.72539407e-01 1.52757034e-01
-1.52557239e-01 7.37150371e-01 2.86292523e-01 -5.83603084e-01
3.72347869e-02 -4.24234658e-01 -3.75941008e-01 -8.56087029e-01
-6.44206583e-01 -1.43701613e-01 8.03130686e-01 -9.84247148e-01
1.76193142e+00 -2.00567865e+00 -9.53323543e-02 1.14103474e-01
2.73064703e-01 8.25094339e-03 4.13917862e-02 1.59089291e+00
-2.03179896e-01 5.66563189e-01 -3.88501316e-01 -3.66269261e-01
-1.02858797e-01 -1.01749681e-01 -1.18009019e+00 4.36016321e-01
7.20920553e-03 4.87529308e-01 -1.21240115e+00 -4.32550550e-01
-1.83659360e-01 1.59951914e-02 -3.51887047e-01 5.23716867e-01
-8.28315318e-02 5.24484098e-01 -3.68616164e-01 5.05130708e-01
2.74725676e-01 7.65759312e-03 -2.03524753e-01 -8.90265703e-02
-6.97795630e-01 5.23963749e-01 -7.12946713e-01 1.22983539e+00
-5.06715104e-02 8.99999201e-01 2.22436011e-01 -6.19692326e-01
6.19525909e-01 6.77191556e-01 1.00556290e+00 -3.33285302e-01
1.10855743e-01 -6.28776802e-03 -7.55875051e-01 -7.05045819e-01
8.27424109e-01 9.78772864e-02 -8.35634291e-01 1.04166234e+00
-3.31740290e-01 -4.30569917e-01 5.68775177e-01 6.89493060e-01
1.58850455e+00 -5.88920057e-01 5.13183177e-01 4.12485488e-02
-3.44966650e-01 6.29869699e-01 3.67751837e-01 8.59647691e-01
-6.93044588e-02 8.23179364e-01 5.71904540e-01 -7.71781206e-01
-9.74423528e-01 -6.46458089e-01 1.79755479e-01 1.17528629e+00
-5.64040005e-01 -8.42155635e-01 -5.96535683e-01 -3.63860041e-01
-3.22657943e-01 8.76395226e-01 -3.30877483e-01 5.01584530e-01
-7.88986444e-01 -1.09806776e+00 1.13395584e+00 3.82037938e-01
3.96363705e-01 -1.28139603e+00 -1.11870623e+00 5.02244592e-01
-1.00947821e+00 -1.19242823e+00 -4.45718557e-01 1.63564503e-01
-4.60181206e-01 -1.47758198e+00 -1.41676411e-01 -1.63749874e-01
2.70876408e-01 3.03850442e-01 1.04356468e+00 -2.60475934e-01
-3.17728579e-01 4.17935461e-01 -7.23407090e-01 -6.81479931e-01
-4.53667156e-02 3.82238120e-01 6.46138936e-02 -3.59761000e-01
2.00142056e-01 -7.54509091e-01 -4.87154543e-01 -9.21298787e-02
-1.13030291e+00 -1.74239837e-02 -1.28826916e-01 2.63909131e-01
7.87193980e-03 6.94970638e-02 8.39837134e-01 -7.16671228e-01
1.20985162e+00 -1.14664733e+00 -1.61978185e-01 -2.37667978e-01
-1.48727849e-01 -5.20213246e-01 3.13873261e-01 1.41412929e-01
-7.16805935e-01 -1.87379152e-01 -8.64540935e-02 3.24626029e-01
-4.21241641e-01 1.23949301e+00 6.86347604e-01 9.17761981e-01
1.09060121e+00 -1.53903782e-01 -4.52707082e-01 -7.19533563e-01
1.40112475e-01 1.14226079e+00 9.52394485e-01 -2.16279194e-01
5.69384933e-01 7.02472150e-01 -5.09225786e-01 -9.66637671e-01
-1.10016787e+00 -1.09824514e+00 -4.63659883e-01 -2.30578244e-01
7.29607165e-01 -9.98202860e-01 -6.81609094e-01 6.03968084e-01
-1.59398127e+00 -4.21848148e-01 -3.32321018e-01 2.13292345e-01
-1.48025841e-01 2.02870339e-01 -7.02751219e-01 -7.97580361e-01
-8.56307745e-01 -2.70081848e-01 1.07546425e+00 -1.92302153e-01
-6.38389885e-01 -7.80275404e-01 5.81079662e-01 -5.26069850e-02
6.41003609e-01 1.18381417e+00 3.85934502e-01 -1.32048535e+00
4.17064801e-02 -7.34481156e-01 -2.88132817e-01 -4.47410792e-01
1.95980087e-01 -4.23791260e-02 -9.24340487e-01 -3.27379286e-01
-1.73697338e-01 -4.31329757e-01 1.11526489e+00 1.48145184e-01
5.43605328e-01 -9.67435956e-01 -4.02659148e-01 2.74921179e-01
8.01568210e-01 -7.60530084e-02 4.26860988e-01 6.48504734e-01
3.72451693e-01 6.69744492e-01 6.55572951e-01 1.28062332e+00
8.08396935e-01 1.29944965e-01 2.34637365e-01 -1.23494938e-01
4.43782449e-01 -1.09828681e-01 4.28904921e-01 8.01124513e-01
4.74669673e-02 -5.66222966e-01 -1.76238549e+00 8.12073350e-01
-2.09982324e+00 -1.54194820e+00 -1.52965754e-01 1.86586273e+00
8.50633502e-01 5.64194657e-02 2.35736743e-01 3.87156457e-01
5.72538137e-01 6.24092221e-01 -1.00909643e-01 -7.02048391e-02
-2.54839689e-01 3.26303430e-02 5.35707772e-01 6.45537436e-01
-1.57179356e+00 8.10693026e-01 7.05887270e+00 4.68432695e-01
-9.53600228e-01 2.23111331e-01 3.77398223e-01 -3.56936634e-01
1.02117158e-01 1.77799106e-01 -7.79696763e-01 1.82343438e-01
1.24755836e+00 -3.48503828e-01 1.98708296e-01 3.80581498e-01
9.31049407e-01 -4.02613491e-01 -8.89626682e-01 7.57317126e-01
-6.95825443e-02 -1.44393206e+00 -3.93819213e-01 -3.11230749e-01
4.24419224e-01 6.05984807e-01 -5.87305486e-01 2.01999143e-01
5.37159622e-01 -7.88003623e-01 6.07440829e-01 7.93462992e-01
6.35884821e-01 -6.12421215e-01 7.97891676e-01 8.10743153e-01
-1.11776340e+00 -5.44855651e-03 2.44537994e-01 -4.04945463e-01
9.14825082e-01 6.84142590e-01 -1.35877550e+00 2.92990208e-01
9.20558512e-01 9.65909243e-01 -4.86647248e-01 1.02507591e+00
-1.57671377e-01 7.22714126e-01 -7.19607353e-01 4.15495664e-01
3.35246533e-01 5.24921060e-01 9.63945150e-01 1.85597050e+00
9.69291255e-02 6.48968697e-01 7.63019323e-01 -4.31272155e-03
2.16578349e-01 3.84957716e-02 -9.21545088e-01 -3.08100194e-01
6.25308812e-01 1.24075282e+00 -8.18806231e-01 -4.72125739e-01
1.37677893e-01 3.54240388e-01 1.09294295e-01 3.38098168e-01
-6.60109639e-01 -7.86286712e-01 3.55705619e-01 4.72797692e-01
-2.59624660e-01 -4.86427277e-01 -2.30362471e-02 -1.15097010e+00
-3.82862121e-01 -8.81564260e-01 1.02151346e+00 -7.45110750e-01
-1.39811945e+00 1.02697790e+00 6.65704608e-01 -1.07563484e+00
-7.51883745e-01 2.67092615e-01 -8.29189301e-01 4.84565109e-01
-1.19132197e+00 -8.87315750e-01 -3.71376902e-01 5.81300735e-01
5.79602063e-01 8.13012272e-02 8.62070858e-01 2.96239585e-01
-5.00892341e-01 -1.64768755e-01 -3.30535442e-01 4.15451884e-01
8.25169921e-01 -9.57497001e-01 8.71975601e-01 9.89895225e-01
-3.75356317e-01 3.86872411e-01 9.28082168e-01 -1.05430830e+00
-8.12268615e-01 -1.17386854e+00 1.43002355e+00 -6.30973101e-01
9.21883583e-01 -2.73620397e-01 -7.62009501e-01 8.01425993e-01
6.72925949e-01 -7.00085044e-01 6.47927523e-01 -1.55417556e-02
-2.66381800e-01 1.33986577e-01 -9.20437455e-01 4.30782825e-01
6.69193566e-01 -5.33128202e-01 -9.03507531e-01 9.56005037e-01
6.66871428e-01 -3.94024253e-01 -7.36942410e-01 2.30859295e-01
7.93760084e-03 -7.61117697e-01 5.54709613e-01 -7.21670270e-01
2.80804992e-01 -2.48722628e-01 -1.46215051e-01 -1.56964576e+00
5.61394654e-02 -1.13594985e+00 3.41957986e-01 1.25424910e+00
4.93933469e-01 -6.71416640e-01 -6.21958971e-02 6.88594699e-01
-3.03464770e-01 -1.52333556e-02 -8.11417520e-01 -2.45534286e-01
-3.07267517e-01 -7.25106955e-01 5.91458440e-01 1.11310387e+00
6.80010259e-01 3.97340119e-01 -3.35641682e-01 2.74849564e-01
4.23808604e-01 -1.85868628e-02 9.24918532e-01 -1.27729118e+00
3.98387343e-01 -3.40765715e-01 4.28670086e-02 -3.33178818e-01
-3.02382976e-01 -7.26767540e-01 2.36444131e-01 -1.80051339e+00
1.73788387e-02 -2.48778060e-01 -1.95740219e-02 9.56664383e-01
1.36498690e-01 -5.03138378e-02 7.66384304e-02 3.94585431e-01
-9.26298320e-01 9.75799337e-02 2.46825054e-01 -1.35750368e-01
-4.62536693e-01 -1.96875930e-01 -5.88937044e-01 8.99893522e-01
1.11097574e+00 -9.67010498e-01 1.19918093e-01 -4.78880733e-01
7.19118178e-01 2.64995873e-01 4.13668863e-02 -1.08275902e+00
8.25797558e-01 -4.92718190e-01 -1.53837025e-01 -1.07164371e+00
4.46833260e-02 -3.26102018e-01 1.77088469e-01 1.27397701e-01
-4.78705168e-01 7.08562553e-01 4.52112198e-01 4.20707583e-01
-4.67540354e-01 3.47979397e-01 2.78616279e-01 -1.94843560e-01
-4.51146871e-01 2.33615324e-01 -9.72713351e-01 7.40059197e-01
6.80306911e-01 3.60081285e-01 -7.88709581e-01 -8.54979873e-01
-7.81200409e-01 5.88615000e-01 -2.31248066e-02 3.35709602e-01
5.92133045e-01 -7.70362318e-01 -1.25504827e+00 -3.01561922e-01
1.48337066e-01 2.53086895e-01 7.86071643e-02 9.51600254e-01
-7.67055213e-01 1.45838782e-01 2.74174334e-03 -2.97251642e-01
-8.33222270e-01 2.84658700e-01 -1.36896759e-01 -4.34014320e-01
-7.21827686e-01 4.70619082e-01 -5.54824829e-01 -3.56893510e-01
3.14296514e-01 -5.55601358e-01 -4.57627922e-01 8.70010436e-01
1.38912117e+00 6.07584715e-01 1.73246935e-01 -7.10280120e-01
-4.57994610e-01 2.68492103e-02 1.52201846e-01 -6.17824495e-01
1.96855199e+00 -1.11339651e-01 -2.19819829e-01 5.00934780e-01
7.64197767e-01 -3.50932851e-02 -8.35032701e-01 -6.78918809e-02
5.01879334e-01 -1.38837248e-01 2.59602629e-02 -6.48543715e-01
-4.57487434e-01 7.78680742e-01 -6.37163147e-02 7.16892242e-01
1.14561772e+00 -1.22025877e-01 1.03949034e+00 7.10932076e-01
3.13189149e-01 -1.24418890e+00 -6.27991557e-02 1.00750577e+00
1.46660721e+00 -1.15663326e+00 2.88350374e-01 1.45854056e-01
-7.45598197e-01 9.78914976e-01 -3.94518226e-02 -3.57273668e-02
8.63288820e-01 5.18370092e-01 1.26249999e-01 -6.24786735e-01
-1.06304407e+00 -2.90548414e-01 -3.36892515e-01 3.38600606e-01
1.81636423e-01 1.11395955e-01 1.04907323e-02 7.03074753e-01
-1.85025603e-01 -3.32421482e-01 6.02539122e-01 1.32879508e+00
-9.61894751e-01 -4.49636191e-01 -6.55171752e-01 3.26944649e-01
-7.29210436e-01 -4.19908255e-01 -6.58552110e-01 6.43700182e-01
-3.66713583e-01 1.58537579e+00 1.09755985e-01 -5.02487719e-01
6.52820349e-01 -1.06258214e-01 -4.44150150e-01 -7.51483083e-01
-1.18175590e+00 8.71645734e-02 4.99291241e-01 -6.35010362e-01
-2.52842486e-01 -9.88847256e-01 -1.47342217e+00 -6.36108279e-01
2.01264009e-01 4.49033797e-01 6.73934281e-01 7.72743165e-01
5.34346819e-01 1.11026734e-01 7.18362153e-01 -1.39719832e+00
-1.91746578e-01 -1.33780444e+00 -2.33131349e-01 5.36874115e-01
4.61544722e-01 -1.31904647e-01 -5.32762408e-01 3.22882421e-02] | [8.720222473144531, 9.46078872680664] |
7a2686c6-68a5-440c-b950-f99a2ab9f96d | adaptive-and-scalable-android-malware | 1606.07150 | null | http://arxiv.org/abs/1606.07150v2 | http://arxiv.org/pdf/1606.07150v2.pdf | Adaptive and Scalable Android Malware Detection through Online Learning | It is well-known that malware constantly evolves so as to evade detection and
this causes the entire malware population to be non-stationary. Contrary to
this fact, prior works on machine learning based Android malware detection have
assumed that the distribution of the observed malware characteristics (i.e.,
features) do not change over time. In this work, we address the problem of
malware population drift and propose a novel online machine learning based
framework, named DroidOL to handle it and effectively detect malware. In order
to perform accurate detection, security-sensitive behaviors are captured from
apps in the form of inter-procedural control-flow sub-graph features using a
state-of-the-art graph kernel. In order to perform scalable detection and to
adapt to the drift and evolution in malware population, an online
passive-aggressive classifier is used.
In a large-scale comparative analysis with more than 87,000 apps, DroidOL
achieves 84.29% accuracy outperforming two state-of-the-art malware techniques
by more than 20% in their typical batch learning setting and more than 3% when
they are continuously re-trained. Our experimental findings strongly indicate
that online learning based approaches are highly suitable for real-world
malware detection. | ['Annamalai Narayanan', 'Liu Jinliang', 'Liu Yang', 'Lihui Chen'] | 2016-06-23 | null | null | null | null | ['android-malware-detection'] | ['miscellaneous'] | [ 7.02631176e-02 -4.71477300e-01 -6.26527727e-01 4.96953987e-02
-2.01456413e-01 -7.21737564e-01 6.45806968e-01 3.28779280e-01
-1.71795383e-01 3.88504744e-01 -6.42817795e-01 -5.88710904e-01
1.67754024e-01 -6.65479302e-01 -7.50375450e-01 -4.77286130e-01
-7.34795451e-01 2.26853579e-01 7.29721785e-01 -1.71831742e-01
2.72273868e-01 4.49313134e-01 -1.65322709e+00 2.81252921e-01
8.26025605e-01 8.16072166e-01 -2.89470494e-01 8.47595215e-01
-3.12398523e-02 5.88653207e-01 -1.01302063e+00 -4.83304977e-01
1.60267740e-01 -2.16058418e-01 -3.47629666e-01 -9.90856215e-02
2.39916652e-01 -3.83160383e-01 -1.92917958e-01 1.23856628e+00
3.21551077e-02 -4.32039261e-01 6.70556605e-01 -1.43392384e+00
-2.45753333e-01 4.40068305e-01 -8.42168152e-01 5.85023224e-01
3.93493712e-01 3.09805036e-01 5.93069732e-01 -1.44835308e-01
5.77167153e-01 1.03194273e+00 8.38824749e-01 6.29398048e-01
-1.20131373e+00 -7.82372296e-01 3.32910240e-01 2.73790151e-01
-1.20763826e+00 -2.41014749e-01 1.02465808e+00 -7.14170039e-01
8.31801891e-01 3.13377023e-01 1.06511128e+00 1.62531292e+00
8.14165533e-01 5.04574060e-01 1.00334394e+00 2.73250858e-03
5.13406813e-01 2.81827003e-01 2.34933287e-01 7.41760492e-01
6.97711051e-01 1.49506494e-01 -2.28125557e-01 -8.14440250e-01
5.92044331e-02 4.50378239e-01 5.09545915e-02 -3.87145191e-01
-5.60898662e-01 9.03321147e-01 1.95133492e-01 4.06572551e-01
-9.89261940e-02 -1.34417996e-01 9.72983122e-01 2.34311253e-01
6.81755900e-01 2.14536637e-01 -7.47896791e-01 -6.80799067e-01
-8.06338549e-01 -2.07668059e-02 9.32643831e-01 3.69348824e-01
7.21875310e-01 2.63225228e-01 1.78807840e-01 4.83854383e-01
2.35267937e-01 3.94253641e-01 1.00484490e+00 -3.17760825e-01
8.17180276e-02 9.76430058e-01 -4.28311735e-01 -1.31640697e+00
-5.86041510e-02 -4.90284234e-01 -5.97516596e-01 1.21003382e-01
2.77110815e-01 -6.78136572e-02 -6.50347531e-01 1.47407317e+00
5.59258521e-01 6.58940971e-01 -2.35513195e-01 -1.31350890e-01
1.41497627e-01 6.09489024e-01 -3.29556465e-02 -6.47665679e-01
1.25055575e+00 -4.50084001e-01 -6.10900760e-01 -1.41766265e-01
7.46560752e-01 -2.17643708e-01 1.24678969e+00 4.55328166e-01
-3.18971455e-01 -2.07993329e-01 -1.18515038e+00 9.70890403e-01
-7.15552270e-01 -3.56225938e-01 5.11844158e-01 1.20235598e+00
-7.81469584e-01 4.11662638e-01 -1.33373177e+00 -3.20636570e-01
7.77289569e-01 4.86757576e-01 1.05805069e-01 3.54909956e-01
-6.64184332e-01 1.66088507e-01 4.11965340e-01 -3.87175649e-01
-1.32417977e+00 -7.80301988e-01 -5.05731761e-01 -2.71258742e-01
6.80613339e-01 2.99376771e-02 1.13856375e+00 -1.05002618e+00
-1.59687829e+00 6.97848022e-01 -1.21641427e-01 -6.17817640e-01
5.37044227e-01 -1.45345017e-01 -6.38133168e-01 -1.76026255e-01
-2.61153936e-01 -2.51769245e-01 1.41061878e+00 -1.33340323e+00
-4.82467443e-01 -7.06602037e-01 -8.56416747e-02 -6.17725253e-01
-1.06619489e+00 -9.10624042e-02 -2.39048049e-01 -4.03821260e-01
-6.46871269e-01 -1.18747818e+00 2.80499011e-01 -6.10598028e-01
-3.23497593e-01 -1.01851292e-01 1.88059151e+00 -7.06222773e-01
1.66321397e+00 -2.11962938e+00 -2.47213066e-01 1.17988281e-01
4.35363591e-01 7.75756180e-01 1.85881764e-01 1.38935760e-01
2.02983767e-01 2.09254637e-01 -4.72306162e-01 -1.02132596e-01
-5.20583272e-01 1.45593494e-01 -3.42860162e-01 7.03482091e-01
9.01577845e-02 7.67290354e-01 -9.73704159e-01 -3.13460141e-01
1.70735437e-02 5.87709188e-01 -5.74020922e-01 1.96740672e-01
-5.00448108e-01 4.11790609e-01 -3.21377933e-01 9.07520711e-01
5.34617543e-01 -2.62346566e-01 4.18968350e-01 3.61177683e-01
2.59185433e-01 -1.88863128e-01 -5.27420461e-01 8.61423612e-01
-4.22068238e-01 7.59436131e-01 6.00156263e-02 -9.23567712e-01
7.79578567e-01 -7.38995150e-02 7.05609798e-01 -3.97477239e-01
4.35844362e-01 1.30675763e-01 3.93687636e-01 -4.45802987e-01
1.75688118e-01 6.97330117e-01 1.51456311e-01 4.29804742e-01
-2.02378184e-01 2.48854727e-01 1.04606494e-01 9.43550654e-03
1.61959231e+00 -4.25521612e-01 5.73451102e-01 -1.49707511e-01
8.69084477e-01 -3.25009115e-02 4.34354663e-01 5.41112840e-01
-5.06222904e-01 -3.93610924e-01 7.96031058e-01 -3.77976149e-01
-5.56426108e-01 -8.05548728e-01 1.00259557e-01 1.27631283e+00
-9.77141187e-02 -7.06956208e-01 -1.26983249e+00 -1.34432042e+00
1.39863640e-01 3.34963650e-01 -8.00054967e-01 -6.00480855e-01
-7.45158851e-01 -9.93926644e-01 5.71144581e-01 6.89379033e-03
4.63927627e-01 -7.99691737e-01 -8.08670342e-01 2.08532855e-01
4.40306664e-01 -1.06008351e+00 -3.31788093e-01 -9.73853469e-02
-1.06768453e+00 -1.48615372e+00 -2.44345397e-01 -3.49615514e-01
5.67343652e-01 1.71373144e-01 8.17670643e-01 3.78083140e-01
-5.29649436e-01 5.47981679e-01 -4.10633326e-01 -5.28653800e-01
-9.87814724e-01 4.70870733e-01 1.41721070e-01 3.15451413e-01
4.24094468e-01 -5.84773540e-01 -3.21432412e-01 2.50542521e-01
-8.75275671e-01 -9.22409773e-01 3.93451959e-01 6.76104128e-01
2.97014654e-01 5.68488836e-01 3.57201785e-01 -1.17655671e+00
9.12836373e-01 -8.01698267e-01 -9.29866076e-01 1.56830698e-01
-9.51866210e-01 -2.43965030e-01 1.10466766e+00 -1.35887802e+00
-6.47460997e-01 2.76693627e-02 1.61617428e-01 -5.19667327e-01
5.08159734e-02 2.59640105e-02 -3.24971884e-01 -2.91103065e-01
8.03942561e-01 4.14386362e-01 1.79321557e-01 -8.27252194e-02
1.41604505e-02 8.78734112e-01 2.09051356e-01 -1.18835919e-01
1.01214635e+00 5.44185758e-01 -1.05373099e-01 -1.13961422e+00
-2.14367837e-01 -3.06634098e-01 -5.11366367e-01 -2.70686865e-01
5.95994174e-01 -6.35335207e-01 -9.41272020e-01 8.11192155e-01
-8.38333309e-01 -5.84133148e-01 8.73074904e-02 -1.02178112e-01
-1.84767082e-01 5.09143054e-01 -5.11045873e-01 -1.19278181e+00
-4.25549716e-01 -1.19824505e+00 1.03257275e+00 1.18166059e-01
-1.72109038e-01 -1.17241335e+00 2.95149982e-01 -8.35505649e-02
4.66693521e-01 3.75810266e-01 7.88413584e-01 -1.00445461e+00
-1.72659129e-01 -5.69324970e-01 2.51298606e-01 2.74996221e-01
5.74644744e-01 6.32594585e-01 -9.14930284e-01 -7.82489955e-01
3.69417131e-01 3.12285841e-01 5.32569468e-01 3.97018641e-02
1.33249986e+00 -7.37937093e-01 -6.93329334e-01 4.60064292e-01
1.23378539e+00 5.88901401e-01 2.86692888e-01 2.58581281e-01
7.74720490e-01 4.00578052e-01 4.00689662e-01 5.02010703e-01
1.89398617e-01 7.12647736e-01 8.31651032e-01 4.55634683e-01
3.19475114e-01 -3.53881329e-01 1.00077915e+00 7.55747914e-01
2.73583591e-01 -9.91146788e-02 -1.11170530e+00 2.46721447e-01
-1.55724359e+00 -9.24059808e-01 6.90439790e-02 2.38304687e+00
7.17574239e-01 6.03162944e-01 7.93411791e-01 2.04057947e-01
7.10379839e-01 2.66630292e-01 -7.58659482e-01 -5.56270480e-01
3.66699308e-01 -8.96509066e-02 6.48770511e-01 3.27191025e-01
-1.13482523e+00 9.40964222e-01 6.17741823e+00 9.86610651e-01
-1.65670896e+00 4.83857810e-01 6.69006169e-01 1.74716562e-01
2.28193864e-01 -1.14522465e-01 -7.93498158e-01 9.58337963e-01
1.35335779e+00 -9.83610228e-02 6.86058581e-01 1.20804107e+00
-8.11397284e-03 2.72765905e-02 -7.24134028e-01 1.13539100e+00
2.87210047e-01 -9.84373569e-01 -2.34759077e-01 5.70836365e-01
7.03978121e-01 -4.27955873e-02 2.72926241e-01 4.72035319e-01
1.69129461e-01 -7.06901312e-01 1.32160336e-01 1.06396474e-01
6.57872379e-01 -7.09845126e-01 5.15904307e-01 6.23752058e-01
-1.36364746e+00 -7.20802307e-01 1.32681817e-01 9.17374119e-02
-2.92242825e-01 6.68572962e-01 -1.12751675e+00 9.43755060e-02
7.34722614e-01 9.12539780e-01 -1.15881371e+00 6.55751884e-01
7.19924122e-02 1.16789520e+00 -1.03645600e-01 -3.28259557e-01
-7.98033550e-02 1.21649019e-01 6.84816420e-01 1.12025821e+00
1.99965797e-02 -5.84844768e-01 1.62319735e-01 5.06964386e-01
4.29089926e-02 8.72496739e-02 -8.46508861e-01 -5.38645208e-01
3.88350219e-01 1.17277133e+00 -7.98176646e-01 -4.44565386e-01
-1.69270456e-01 8.39270353e-01 1.73847467e-01 5.94355687e-02
-9.49442565e-01 -4.41337340e-02 8.00241768e-01 5.08641064e-01
3.18592608e-01 -4.65221405e-01 6.03047200e-02 -1.13453698e+00
6.04791520e-03 -1.25275624e+00 2.43622571e-01 3.32069576e-01
-1.22900975e+00 7.12859750e-01 7.77590945e-02 -1.06848884e+00
-6.55336380e-01 -7.27070570e-01 -9.39585805e-01 -1.35143369e-01
-1.05179822e+00 -9.72639501e-01 -3.86871129e-01 6.10018671e-01
5.63536227e-01 -5.19831777e-01 4.89321917e-01 1.06487602e-01
-8.18630517e-01 7.64271080e-01 1.77758783e-01 -3.47662829e-02
3.19654077e-01 -1.01368845e+00 3.84671181e-01 7.98524022e-01
1.20544307e-01 6.18133187e-01 6.63152993e-01 -1.32779717e+00
-1.86812615e+00 -1.21508968e+00 -1.55430762e-02 -5.80361545e-01
1.09259307e+00 -8.30344021e-01 -1.20495296e+00 5.04762232e-01
-9.69037116e-02 2.42241979e-01 4.66769129e-01 -2.21110940e-01
-5.16444683e-01 -4.72148329e-01 -1.22096753e+00 4.87582594e-01
1.01247132e+00 -5.09521484e-01 3.90515625e-02 2.02627078e-01
5.78558326e-01 -7.79410452e-02 -5.22676885e-01 6.39165223e-01
5.61027288e-01 -9.99323249e-01 6.99171126e-01 -5.11908352e-01
-2.97999829e-01 -1.27076972e-02 1.02174699e-01 -8.40008616e-01
1.99045166e-01 -1.03906155e+00 -1.14046502e+00 1.10191345e+00
1.60528347e-01 -1.00616741e+00 9.52933013e-01 -2.47359484e-01
4.36417490e-01 -9.95015025e-01 -9.30599511e-01 -1.21976149e+00
-1.36028394e-01 -2.98447698e-01 5.18600464e-01 8.87473702e-01
-2.11067840e-01 1.10164694e-01 -2.64354080e-01 -1.55553585e-02
5.86293519e-01 -5.04884481e-01 1.04401445e+00 -1.38965654e+00
-4.17255819e-01 -5.71851790e-01 -7.38985121e-01 -4.39466178e-01
5.54466069e-01 -6.55129969e-01 -3.56066316e-01 -4.50780302e-01
2.31733575e-01 -7.66100511e-02 -1.70193519e-02 2.13006034e-01
-1.22795843e-01 4.04561237e-02 -1.59631357e-01 3.03084701e-01
-4.96674001e-01 3.44806343e-01 5.82814753e-01 -2.92423308e-01
-6.91051066e-01 3.83155137e-01 -3.39703172e-01 7.15077102e-01
9.98517573e-01 -4.78294939e-01 -5.23056448e-01 2.70886272e-01
-6.02055527e-02 -5.86561024e-01 2.62888938e-01 -1.03857744e+00
-6.65195957e-02 -4.38898802e-03 1.35747075e-01 -3.91637564e-01
1.50912941e-01 -9.33716655e-01 5.80253862e-02 1.11534142e+00
2.28903860e-01 4.45202291e-01 3.46137732e-01 9.03917313e-01
1.64441630e-01 1.11328520e-01 7.17375875e-01 1.08697124e-01
-4.50731903e-01 4.38802719e-01 -6.08637750e-01 6.17037714e-02
1.60480595e+00 -3.24693859e-01 -4.56505865e-01 -2.00895399e-01
-9.06371102e-02 -3.26347470e-01 7.94869184e-01 5.96539259e-01
4.06039834e-01 -8.51304710e-01 -2.53952920e-01 4.37739372e-01
1.52677819e-01 -7.31653988e-01 1.10102817e-02 8.28720808e-01
-3.44866544e-01 1.43892780e-01 -6.54025078e-02 -9.63139236e-01
-1.69157791e+00 8.78251731e-01 2.60291070e-01 -6.05656981e-01
-3.99729133e-01 2.70509154e-01 -4.12914783e-01 -1.80150405e-01
3.15909415e-01 5.08003943e-02 -6.05969205e-02 1.05935618e-01
6.68008626e-01 6.20721757e-01 1.61148161e-01 -7.11854577e-01
-6.54096067e-01 4.07456845e-01 -2.63368100e-01 3.36435139e-01
1.02489078e+00 3.37392926e-01 -3.06872606e-01 6.51481330e-01
1.43628705e+00 4.59716052e-01 -1.20095575e+00 2.08084255e-01
3.21052879e-01 -5.73724329e-01 -4.29444611e-01 -3.67433161e-01
-1.05069220e+00 5.34523547e-01 1.12556756e+00 8.35136294e-01
1.02008390e+00 -1.43366143e-01 7.57470608e-01 1.92706227e-01
3.98745269e-01 -7.51849294e-01 4.95927244e-01 4.72794741e-01
3.00201118e-01 -1.29975128e+00 -1.30249253e-02 -4.74661618e-01
-1.39431730e-01 9.53685522e-01 7.62332797e-01 -1.27865791e-01
8.51961195e-01 4.49564606e-01 -3.28190595e-01 -1.37067661e-01
-6.49101734e-01 3.28742117e-01 1.88512862e-01 8.65234375e-01
1.11757610e-02 2.16780022e-01 -8.51684734e-02 2.41061047e-01
-1.39820740e-01 -3.51121753e-01 3.57938021e-01 1.05605054e+00
-3.41233253e-01 -1.17334008e+00 -3.61506552e-01 6.99225247e-01
-5.46089888e-01 3.82856250e-01 -9.08707142e-01 8.05370569e-01
1.06203683e-01 9.64704573e-01 -6.63121119e-02 -9.44351554e-01
-3.38164419e-02 -1.45291314e-01 2.40709513e-01 -5.98803818e-01
-7.81952262e-01 -2.76291579e-01 -3.77836406e-01 -6.08900487e-01
4.91701476e-02 -5.83746135e-01 -9.15102541e-01 -3.58996272e-01
-3.61606538e-01 -1.39297187e-01 7.90902853e-01 7.39617527e-01
5.29582500e-01 4.92914051e-01 9.66477990e-01 -6.74188435e-01
-6.79068387e-01 -7.54013896e-01 -1.37578294e-01 3.45539093e-01
4.62413073e-01 -7.79721737e-01 -7.18649745e-01 -1.64813876e-01] | [14.418810844421387, 9.678032875061035] |
5efcfe75-40d4-4c3c-8228-67d84530d6bc | you-need-to-read-again-multi-granularity | 2205.12886 | null | https://arxiv.org/abs/2205.12886v2 | https://arxiv.org/pdf/2205.12886v2.pdf | You Need to Read Again: Multi-granularity Perception Network for Moment Retrieval in Videos | Moment retrieval in videos is a challenging task that aims to retrieve the most relevant video moment in an untrimmed video given a sentence description. Previous methods tend to perform self-modal learning and cross-modal interaction in a coarse manner, which neglect fine-grained clues contained in video content, query context, and their alignment. To this end, we propose a novel Multi-Granularity Perception Network (MGPN) that perceives intra-modality and inter-modality information at a multi-granularity level. Specifically, we formulate moment retrieval as a multi-choice reading comprehension task and integrate human reading strategies into our framework. A coarse-grained feature encoder and a co-attention mechanism are utilized to obtain a preliminary perception of intra-modality and inter-modality information. Then a fine-grained feature encoder and a conditioned interaction module are introduced to enhance the initial perception inspired by how humans address reading comprehension problems. Moreover, to alleviate the huge computation burden of some existing methods, we further design an efficient choice comparison module and reduce the hidden size with imperceptible quality loss. Extensive experiments on Charades-STA, TACoS, and ActivityNet Captions datasets demonstrate that our solution outperforms existing state-of-the-art methods. Codes are available at github.com/Huntersxsx/MGPN. | ['Xi Zhou', 'Qiong Liu', 'Jialin Gao', 'Xuan Wang', 'Xin Sun'] | 2022-05-25 | null | null | null | null | ['moment-retrieval'] | ['computer-vision'] | [ 2.06478447e-01 -4.06057209e-01 -2.56049305e-01 -3.66484612e-01
-1.03115165e+00 -4.42736596e-01 6.33985341e-01 1.01604737e-01
-6.04851186e-01 2.63786018e-01 6.48234367e-01 3.23270261e-02
-2.28457332e-01 -4.82505828e-01 -8.77397835e-01 -6.39918506e-01
2.99838781e-01 -1.03917599e-01 4.19948459e-01 -2.32964426e-01
4.81073111e-01 -2.50755727e-01 -1.73651516e+00 5.08948088e-01
8.99520695e-01 1.18997395e+00 5.52454174e-01 6.46191657e-01
2.76257992e-01 9.64309275e-01 -4.84718859e-01 -3.36365193e-01
-1.49936572e-01 -6.09687984e-01 -8.71619165e-01 3.17357451e-01
5.37225366e-01 -5.66458762e-01 -7.00675845e-01 1.29019618e+00
7.01988935e-01 4.40447807e-01 4.97070879e-01 -1.06683445e+00
-7.91970253e-01 5.43714345e-01 -6.49184823e-01 5.92633188e-01
9.26173866e-01 3.48698407e-01 8.81643832e-01 -7.59745359e-01
3.82218599e-01 1.27685392e+00 -6.56051636e-02 4.06426013e-01
-8.03755164e-01 -4.68512118e-01 4.57121432e-01 9.95866776e-01
-1.38676143e+00 -4.65288907e-01 9.83227134e-01 -2.11369768e-01
5.98437130e-01 3.87683034e-01 6.38995349e-01 1.46908486e+00
1.65848985e-01 1.21885478e+00 8.23119462e-01 -1.82581902e-01
3.87402587e-02 -2.33084068e-01 7.99922720e-02 7.19124079e-01
-3.80882800e-01 -1.60664350e-01 -9.60541368e-01 3.05861503e-01
7.44852960e-01 1.32526517e-01 -7.61917830e-01 -2.44227543e-01
-1.78712392e+00 5.79741240e-01 4.27895188e-01 3.94374818e-01
-4.63355660e-01 -1.09689094e-01 6.35523558e-01 4.78902638e-01
2.81875074e-01 1.10728689e-01 -3.29684079e-01 -2.80794382e-01
-9.41685140e-01 3.94956395e-02 4.68398035e-01 9.44246113e-01
5.75947523e-01 -3.87618899e-01 -5.47896802e-01 7.20482409e-01
1.87333018e-01 4.88730907e-01 7.53793001e-01 -1.02163112e+00
6.90063536e-01 2.83522934e-01 -9.63521830e-04 -1.34680152e+00
-1.57097638e-01 -2.69818127e-01 -9.11369085e-01 -4.11116064e-01
8.14442188e-02 2.64222056e-01 -7.01990247e-01 1.81557393e+00
3.81043226e-01 2.30469897e-01 1.79357734e-02 1.37999845e+00
9.67371464e-01 8.69311750e-01 5.62099777e-02 -4.72654253e-01
1.68590057e+00 -1.40376532e+00 -7.92724609e-01 -8.94913971e-02
1.65669501e-01 -6.34041548e-01 1.33780789e+00 3.78905326e-01
-1.26382291e+00 -9.73457694e-01 -1.00658488e+00 -4.27139461e-01
-2.14880407e-01 9.30517912e-02 2.58058369e-01 -1.73594326e-01
-8.17257643e-01 3.04509610e-01 -6.20834649e-01 -3.05565387e-01
2.26935923e-01 4.84829955e-02 -3.39273810e-01 -2.29714304e-01
-1.50230932e+00 7.00883508e-01 6.55276418e-01 1.24915488e-01
-9.46463168e-01 -2.83334970e-01 -9.95606482e-01 1.07307442e-01
8.05381179e-01 -9.14651573e-01 1.32213581e+00 -1.04977989e+00
-1.42282939e+00 6.68452084e-01 -2.28223160e-01 -1.86149657e-01
2.83839822e-01 -2.98740119e-01 -5.46035111e-01 9.07883644e-01
1.80011228e-01 8.21873724e-01 1.20974696e+00 -8.28336775e-01
-6.83366120e-01 -1.18596725e-01 5.12139916e-01 7.88121045e-01
-2.65132993e-01 5.92898810e-03 -9.49778020e-01 -7.86473989e-01
1.93866566e-01 -5.10297000e-01 1.33369282e-01 -1.25629604e-01
-2.29009479e-01 -4.48670238e-01 5.98917067e-01 -8.98903847e-01
1.26343346e+00 -2.25008512e+00 7.37085998e-01 -2.44956687e-01
3.87366146e-01 -3.55661996e-02 -3.87281537e-01 3.38946730e-01
1.09104410e-01 -2.06715211e-01 1.44008905e-01 -1.79720730e-01
1.55252218e-01 -1.02385700e-01 -9.27249193e-02 4.08847928e-01
1.08369635e-02 9.22633588e-01 -1.12482417e+00 -8.70518625e-01
3.80029500e-01 4.95443225e-01 -3.56870145e-01 4.20289010e-01
-3.12433988e-01 4.90988106e-01 -5.96994042e-01 8.76879454e-01
4.48938608e-01 -4.68795538e-01 -3.79644483e-01 -6.83182716e-01
6.32900223e-02 4.41683158e-02 -8.35698724e-01 2.42854691e+00
-3.11028123e-01 4.58589673e-01 -5.31863719e-02 -1.03515530e+00
4.77128804e-01 2.96671122e-01 2.81196296e-01 -1.14790583e+00
3.94441545e-01 -8.40512142e-02 -1.55360147e-01 -1.06040013e+00
5.35432994e-01 3.37086052e-01 -2.52723843e-01 2.13945165e-01
2.95912892e-01 2.96287537e-01 4.36913967e-01 5.24508715e-01
9.30751026e-01 2.62520343e-01 2.77889192e-01 8.76006931e-02
8.30981076e-01 -4.04299170e-01 3.86326730e-01 6.38621569e-01
-4.63756144e-01 6.81140006e-01 4.42647040e-01 -1.47219628e-01
-6.44191563e-01 -9.83789563e-01 2.77197242e-01 1.57179654e+00
7.57364810e-01 -5.54784477e-01 -7.16475725e-01 -4.74026084e-01
-5.71503162e-01 3.79569918e-01 -6.19571030e-01 -3.52559477e-01
-2.99134940e-01 -3.79980147e-01 1.94875360e-01 4.23555702e-01
1.06393468e+00 -1.19380975e+00 -6.28213584e-01 7.79462559e-03
-1.04007399e+00 -1.12112653e+00 -9.68007982e-01 -3.63568306e-01
-3.54506701e-01 -1.04199946e+00 -9.20904636e-01 -1.05085385e+00
4.82319683e-01 4.53661859e-01 1.09581256e+00 8.90598353e-03
4.98096347e-02 5.87975442e-01 -7.56175935e-01 1.71141833e-01
1.60290092e-01 1.31662451e-02 -5.38739003e-02 2.04387963e-01
5.72019637e-01 -5.03395200e-01 -9.64822829e-01 3.46833348e-01
-1.04937708e+00 3.88479918e-01 8.66573632e-01 9.43725407e-01
6.12978101e-01 7.17051700e-03 3.77687305e-01 -2.32312918e-01
7.85797238e-01 -5.62635183e-01 -1.97333038e-01 4.78011608e-01
-1.53476104e-01 -1.18239529e-01 4.12558883e-01 -7.06664205e-01
-1.09278762e+00 -6.02870174e-02 -2.04949435e-02 -7.98861086e-01
-2.88818866e-01 6.20964408e-01 -5.49930811e-01 2.26144001e-01
2.75283098e-01 8.18014324e-01 -2.34728068e-01 -2.96169490e-01
4.21003073e-01 6.48093164e-01 9.50702250e-01 -6.04993284e-01
5.82010686e-01 1.48607582e-01 -3.55367661e-01 -5.86122215e-01
-1.07873809e+00 -6.60405517e-01 -5.88871479e-01 -4.24563885e-01
1.16702485e+00 -1.23382783e+00 -1.09595215e+00 5.46033978e-01
-1.13670552e+00 1.23406155e-02 5.75546101e-02 4.83118623e-01
-7.24262595e-01 7.41148591e-01 -6.66535139e-01 -3.39945018e-01
-2.38365859e-01 -1.13961792e+00 1.31478739e+00 5.15067816e-01
7.49371573e-02 -5.93793809e-01 -1.83115840e-01 7.26418197e-01
1.19806245e-01 -6.18963921e-03 4.86202061e-01 -3.83913130e-01
-8.23315322e-01 7.10884482e-02 -4.18639570e-01 1.64009660e-01
-2.17405707e-01 -3.95609468e-01 -6.34985626e-01 -3.83060247e-01
2.29170784e-01 -6.72891796e-01 1.09078383e+00 1.99173182e-01
1.51964116e+00 -3.05788279e-01 -1.13009349e-01 5.48214138e-01
1.04333138e+00 3.08386739e-02 6.66785896e-01 4.22648400e-01
7.58193254e-01 5.71831167e-01 9.71244872e-01 4.06150937e-01
7.52397418e-01 3.91651452e-01 4.30028528e-01 1.67432919e-01
9.21088979e-02 -3.47381473e-01 3.40456873e-01 1.18437254e+00
-9.82959643e-02 -4.53610033e-01 -4.06224400e-01 5.53420186e-01
-2.13058543e+00 -1.36872292e+00 3.43572259e-01 1.85165179e+00
8.32383573e-01 1.60516173e-01 5.60504310e-02 1.47809535e-02
6.77039564e-01 5.66746593e-01 -7.01339483e-01 7.85814151e-02
-2.09891960e-01 -4.05955017e-01 -4.50030267e-02 2.57957518e-01
-1.21690798e+00 7.10938454e-01 4.40538692e+00 1.16869056e+00
-9.34429526e-01 1.29310891e-01 4.30471152e-01 -2.19471917e-01
-7.46497437e-02 -2.47771010e-01 -3.89189124e-01 7.66455591e-01
6.36336863e-01 -1.34533569e-01 4.85225260e-01 5.09729266e-01
2.17828780e-01 -3.40128183e-01 -1.16557467e+00 1.47703171e+00
6.57919705e-01 -1.07434309e+00 4.51722592e-02 -3.69486481e-01
3.12714487e-01 -1.30834252e-01 5.32071330e-02 4.69002753e-01
-5.12015343e-01 -6.87198400e-01 7.80160546e-01 9.37691212e-01
6.98476911e-01 -7.26476133e-01 5.47882974e-01 4.62836176e-01
-1.43762183e+00 -5.76195940e-02 -3.57663274e-01 1.62562281e-02
3.45312566e-01 1.71175018e-01 2.65360158e-02 6.84875131e-01
9.03597236e-01 1.01634657e+00 -8.26237619e-01 1.07720852e+00
-2.78328001e-01 2.81470180e-01 -7.65123144e-02 -1.31605685e-01
1.76645473e-01 1.45081148e-01 5.72981119e-01 1.09547567e+00
1.33138493e-01 4.31474537e-01 3.39088589e-01 3.99894357e-01
-1.40835315e-01 -2.34745312e-02 -5.91180325e-02 9.77978259e-02
4.48254347e-01 1.17369580e+00 -4.60024387e-01 -5.33354342e-01
-5.77356458e-01 1.39133692e+00 2.98511416e-01 5.01590192e-01
-1.04633403e+00 -6.68626964e-01 2.59309590e-01 -2.72021443e-01
3.47179860e-01 -5.86390495e-02 4.96868849e-01 -1.84270465e+00
2.92760432e-01 -1.18813014e+00 6.10569119e-01 -1.24271071e+00
-1.24063444e+00 6.43495500e-01 1.26711667e-01 -1.61016691e+00
-1.96749046e-01 -2.96798766e-01 -4.34989095e-01 5.67902744e-01
-1.47532237e+00 -1.15046096e+00 -5.97618103e-01 1.02283514e+00
1.06890726e+00 1.56265926e-02 4.35977221e-01 3.84484172e-01
-5.06719351e-01 5.49329758e-01 -1.62796825e-01 1.21430174e-01
8.86706412e-01 -9.90634859e-01 -1.41939193e-01 9.60237324e-01
-3.20478366e-03 5.84038317e-01 6.70111060e-01 -4.60482866e-01
-1.46276450e+00 -8.74612868e-01 7.37095892e-01 -2.62556404e-01
6.46156907e-01 -1.70327350e-01 -9.55427170e-01 5.01739502e-01
7.16450572e-01 -2.36367702e-01 4.67131168e-01 -1.44718260e-01
-4.14933324e-01 -1.54759362e-01 -6.57321334e-01 7.25170135e-01
1.09381068e+00 -9.17929292e-01 -1.05047798e+00 3.77605587e-01
1.02618194e+00 -5.95965564e-01 -8.62810791e-01 2.77809978e-01
5.43912411e-01 -9.37746227e-01 9.78939235e-01 -2.52678245e-01
6.70032859e-01 -4.21449810e-01 -2.75271237e-01 -1.21037257e+00
-3.55506361e-01 -6.36707962e-01 -3.63901109e-01 1.20319200e+00
-1.22850366e-01 -1.03286758e-01 2.31269777e-01 2.83800691e-01
-1.37757316e-01 -7.26011276e-01 -7.92413116e-01 -4.52960849e-01
-4.53739852e-01 -2.11272791e-01 3.96301150e-01 7.72704363e-01
4.01686490e-01 8.45527530e-01 -7.43888259e-01 2.49019548e-01
6.24768138e-01 3.22826982e-01 5.56304038e-01 -7.94756949e-01
-4.59304869e-01 -3.28042507e-01 -3.75557274e-01 -1.71070373e+00
1.60979480e-01 -4.37463731e-01 1.81521922e-01 -1.36270058e+00
5.86348355e-01 4.85139072e-01 -6.21077955e-01 2.11783096e-01
-2.76141435e-01 2.46915072e-01 3.12228262e-01 3.09524149e-01
-1.57506943e+00 8.45094502e-01 1.66350663e+00 -2.75757402e-01
3.19905914e-02 -3.28858972e-01 -6.26145422e-01 6.61334574e-01
5.28533399e-01 -5.38253859e-02 -5.95596731e-01 -6.67521656e-01
1.62791222e-01 5.34945309e-01 7.00684488e-01 -1.10563385e+00
5.55854440e-01 -1.25999525e-01 5.63556135e-01 -1.10895252e+00
5.33338010e-01 -6.34992599e-01 -3.23311180e-01 1.37608141e-01
-6.08740270e-01 1.61135063e-01 -1.36555970e-01 8.64246070e-01
-5.98882616e-01 5.61969131e-02 4.61130142e-01 -2.06070811e-01
-1.16507781e+00 4.63292271e-01 -2.77193099e-01 3.48997787e-02
9.06892121e-01 -1.38708306e-02 -5.67206562e-01 -5.82721353e-01
-8.26211810e-01 7.25055814e-01 3.24476779e-01 7.41668582e-01
8.89535189e-01 -1.36646128e+00 -5.68082571e-01 9.04355384e-03
3.30833942e-01 -1.22254863e-01 7.77273357e-01 9.96287465e-01
-1.40446901e-01 2.80264795e-01 -3.86880666e-01 -6.97525561e-01
-1.25203311e+00 8.76804411e-01 1.91420451e-01 1.10571079e-01
-4.06571001e-01 9.56263244e-01 3.11781585e-01 1.26253068e-01
4.48051155e-01 -2.05315828e-01 -5.25946021e-01 2.68207848e-01
8.02733123e-01 2.44437680e-01 -3.55635792e-01 -9.50824797e-01
-3.21257055e-01 7.37839460e-01 -2.01700881e-01 4.32575531e-02
9.29914773e-01 -8.73611987e-01 -1.13639303e-01 5.19695580e-01
1.32866740e+00 -3.30529928e-01 -1.35900462e+00 -4.14898694e-01
-5.23872972e-01 -4.37024236e-01 -5.56732081e-02 -7.05586493e-01
-8.46773863e-01 9.57942605e-01 5.26371181e-01 9.58805755e-02
1.73599327e+00 1.96832314e-01 9.64939415e-01 6.10687554e-01
1.07737370e-01 -1.12241244e+00 5.26899040e-01 3.42552811e-01
9.68770206e-01 -1.53537607e+00 -7.74045959e-02 -1.60901412e-01
-7.59404778e-01 9.52472329e-01 9.80622113e-01 5.95395919e-03
4.17183965e-01 -5.28343201e-01 -6.11925647e-02 -1.32164314e-01
-9.41078961e-01 -3.84162694e-01 4.96366113e-01 3.32997054e-01
1.47447243e-01 -2.60962397e-01 -4.35379654e-01 8.94031048e-01
3.69842462e-02 9.87569764e-02 2.99215794e-01 7.92184114e-01
-5.44560134e-01 -6.37163937e-01 -2.11595029e-01 2.45416597e-01
-3.11007202e-01 -1.80480972e-01 -1.72793698e-02 4.82479930e-01
9.84325483e-02 1.11262619e+00 4.03968953e-02 -5.96229255e-01
2.66612738e-01 -1.91755116e-01 4.30436850e-01 -2.14049295e-01
-2.65680194e-01 4.98879880e-01 -1.31084338e-01 -8.78947377e-01
-8.53339732e-01 -5.90995431e-01 -9.41415250e-01 -1.19929351e-01
-1.92669556e-01 1.63340598e-01 1.11489668e-01 1.03006661e+00
4.09334898e-01 7.03050911e-01 6.09905183e-01 -1.10342944e+00
-4.82851058e-01 -9.77421582e-01 -3.92999887e-01 5.24302304e-01
5.50186217e-01 -5.57445049e-01 -3.71605277e-01 3.63254607e-01] | [10.309380531311035, 0.9683672785758972] |
2299aa3e-daf7-4c2a-9c26-0e14134843b3 | multi-task-learning-with-high-order | 1903.12058 | null | https://arxiv.org/abs/1903.12058v2 | https://arxiv.org/pdf/1903.12058v2.pdf | Multi-Task Learning with High-Order Statistics for X-vector based Text-Independent Speaker Verification | The x-vector based deep neural network (DNN) embedding systems have demonstrated effectiveness for text-independent speaker verification. This paper presents a multi-task learning architecture for training the speaker embedding DNN with the primary task of classifying the target speakers, and the auxiliary task of reconstructing the first- and higher-order statistics of the original input utterance. The proposed training strategy aggregates both the supervised and unsupervised learning into one framework to make the speaker embeddings more discriminative and robust. Experiments are carried out using the NIST SRE16 evaluation dataset and the VOiCES dataset. The results demonstrate that our proposed method outperforms the original x-vector approach with very low additional complexity added. | ['Jun Du', 'LiRong Dai', 'Wu Guo', 'Lanhua You'] | 2019-03-28 | null | null | null | null | ['text-independent-speaker-verification'] | ['speech'] | [-2.70346999e-02 -1.25750586e-01 -2.29697358e-02 -8.21923435e-01
-8.67334604e-01 -2.05476359e-01 7.09768951e-01 -2.61181146e-01
-5.66892266e-01 2.51494437e-01 6.99161351e-01 -5.48711658e-01
2.44189650e-02 9.26550478e-03 -2.08975762e-01 -1.08488381e+00
2.73905806e-02 1.67903543e-01 -4.81179982e-01 1.11790158e-01
1.57987431e-01 5.81242263e-01 -1.62119746e+00 -2.03500345e-01
3.86630058e-01 1.02859795e+00 6.66902587e-02 8.83118331e-01
-1.33816199e-02 5.67063093e-01 -8.01465988e-01 -1.44066647e-01
1.11690916e-01 -1.20621979e-01 -4.81232554e-01 1.28545761e-01
3.90103668e-01 -5.03863096e-01 -6.58884704e-01 1.10938704e+00
1.03591740e+00 4.37575847e-01 7.17006266e-01 -1.23627305e+00
-9.02993202e-01 8.04638684e-01 -2.15648800e-01 4.58587497e-01
5.04360013e-02 -1.64815843e-01 1.08626509e+00 -1.20754528e+00
-5.36451489e-02 1.46395493e+00 4.93201643e-01 9.27862406e-01
-8.70208085e-01 -8.25860500e-01 -1.05629973e-01 5.20991862e-01
-1.54096353e+00 -9.21621323e-01 1.21622956e+00 -2.48235986e-01
1.04763973e+00 3.28917950e-01 -9.85757634e-02 1.15772080e+00
6.15706742e-02 8.40900898e-01 8.02425802e-01 -6.87163293e-01
2.34113678e-01 4.41239148e-01 6.83736324e-01 4.30541694e-01
-1.44911885e-01 3.91526788e-01 -7.36537695e-01 -2.89311439e-01
1.53298721e-01 -1.85374636e-02 -1.42985433e-01 8.00929442e-02
-9.22577500e-01 1.11057734e+00 1.75923426e-02 6.56649351e-01
-4.32260871e-01 3.25071402e-02 7.50705063e-01 3.70396674e-01
4.32790726e-01 -3.09207350e-01 -4.96808022e-01 -6.42358661e-02
-1.04901373e+00 -1.12488873e-01 5.73712349e-01 6.96179748e-01
3.75972509e-01 9.12347615e-01 5.48046734e-03 1.01562881e+00
9.56762075e-01 4.95473713e-01 1.10013425e+00 -3.74151766e-01
3.47067505e-01 2.08160385e-01 -3.26341391e-01 -6.67284727e-01
1.52413687e-02 -4.98629510e-01 -7.93399632e-01 1.63480297e-01
-1.58937573e-01 -2.76500106e-01 -1.13683081e+00 1.64015794e+00
2.10216865e-01 1.24232531e-01 5.75407565e-01 6.99753404e-01
1.29036880e+00 1.00941741e+00 9.94754881e-02 1.57270581e-02
1.40716946e+00 -9.26685214e-01 -1.34803450e+00 -8.28356072e-02
2.46754944e-01 -8.34140003e-01 8.60958397e-01 6.09573685e-02
-6.43699884e-01 -9.52584624e-01 -1.43062341e+00 -3.57432477e-02
-4.40228462e-01 6.09804153e-01 2.44821519e-01 1.08397007e+00
-1.28566623e+00 -3.57613564e-02 -6.43158793e-01 -7.57969022e-02
1.95157796e-01 6.69312298e-01 -4.31921333e-01 2.26925984e-01
-1.23043478e+00 7.00017214e-01 4.71302688e-01 3.93138826e-01
-1.18951035e+00 -4.95003670e-01 -1.19298923e+00 3.85084629e-01
-2.31586620e-01 1.32765770e-01 1.32232785e+00 -5.91853678e-01
-1.78005350e+00 4.99873608e-01 -5.31765282e-01 -3.52803528e-01
-4.22553234e-02 2.18484495e-02 -8.44530404e-01 -1.50095731e-01
-3.02837521e-01 4.81323928e-01 1.04456961e+00 -1.18051827e+00
-3.36767346e-01 -7.19626009e-01 -4.59780365e-01 9.98261049e-02
-7.56222367e-01 3.46097648e-01 3.92252207e-02 -5.78283787e-01
2.91058332e-01 -5.41161060e-01 2.02626407e-01 -3.36437315e-01
-6.06165648e-01 -6.98692918e-01 1.57098544e+00 -1.05704272e+00
1.03113055e+00 -2.64339948e+00 1.56226642e-02 6.82517216e-02
5.17177284e-02 3.06330115e-01 -1.89087793e-01 3.55810016e-01
-3.60282600e-01 -2.14801013e-01 7.00060651e-02 -9.12694693e-01
3.56225431e-01 1.95471942e-01 -4.27350670e-01 6.02612138e-01
-8.37786123e-03 6.45671427e-01 -2.71589488e-01 -4.83283758e-01
3.47553998e-01 1.09454536e+00 -1.23712428e-01 2.51145750e-01
3.95704299e-01 2.36822009e-01 6.08239546e-02 6.69581771e-01
7.24457145e-01 4.07562703e-01 -1.00882282e-03 -2.07999453e-01
1.63742900e-02 6.36554062e-01 -1.23089981e+00 1.43164957e+00
-3.52261543e-01 1.14708006e+00 3.69614393e-01 -9.52551782e-01
1.16883075e+00 8.82967591e-01 8.97665322e-02 -2.88672388e-01
3.54861259e-01 4.54465933e-02 1.93370938e-01 -3.65881562e-01
5.11057615e-01 -5.30258238e-01 5.23346849e-02 5.73216558e-01
5.97843230e-01 3.78362924e-01 -4.93352294e-01 6.40745163e-02
5.29935360e-01 -4.93513197e-01 1.42790779e-01 -2.70257711e-01
1.06773067e+00 -8.04475725e-01 4.71770793e-01 2.66367435e-01
-6.26265347e-01 4.80008572e-02 8.54168311e-02 -2.89558023e-01
-8.68773341e-01 -9.83715594e-01 -2.89606839e-01 1.37254882e+00
-3.30784351e-01 -3.21284048e-02 -6.99727833e-01 -6.53287590e-01
7.99247772e-02 1.08202755e+00 -7.17674434e-01 -1.38274610e-01
-3.89869690e-01 -3.58646363e-01 1.01540482e+00 6.75536871e-01
3.75204921e-01 -9.92962718e-01 6.59871921e-02 6.19185250e-03
4.82919514e-02 -9.15986180e-01 -8.20358634e-01 3.53437096e-01
-8.19533169e-01 -4.45593774e-01 -6.40954554e-01 -1.42090392e+00
5.90142190e-01 6.38500080e-02 3.49468619e-01 -3.94477695e-01
7.65745947e-03 3.99067819e-01 5.99289499e-02 -5.78635514e-01
-5.93319654e-01 -1.86186969e-01 7.32728899e-01 3.42555791e-01
1.03714883e+00 -4.00819123e-01 -1.31585017e-01 1.32431760e-01
-8.91790807e-01 -8.90658021e-01 4.29572850e-01 9.81124222e-01
1.75331563e-01 1.88874483e-01 8.65810692e-01 -6.67254999e-02
9.16688859e-01 7.50281438e-02 -3.44323218e-01 8.42130408e-02
-5.11134267e-01 1.62832603e-01 3.61174792e-01 -5.20587385e-01
-1.11923540e+00 -8.03819671e-02 -4.95562881e-01 -4.99759108e-01
-2.57947922e-01 4.57342356e-01 -7.05273449e-01 5.10346703e-02
3.33325386e-01 7.97300339e-01 2.10666627e-01 -7.33007729e-01
2.92108625e-01 1.48227131e+00 5.36780775e-01 -9.44344327e-02
7.81225622e-01 -8.26395527e-02 -5.30566335e-01 -1.27252650e+00
-3.27845275e-01 -7.44296372e-01 -7.66068041e-01 -2.25060508e-01
8.64530504e-01 -1.02022994e+00 -9.29302096e-01 5.20073950e-01
-1.34924841e+00 3.81596178e-01 -1.97982386e-01 8.70344460e-01
-9.66587886e-02 5.48598826e-01 -5.91252804e-01 -1.10563159e+00
-6.53395951e-01 -1.34937072e+00 1.02954507e+00 8.78136456e-02
1.00608580e-02 -1.19955349e+00 3.24298412e-01 4.15680498e-01
6.36322558e-01 -5.30294240e-01 9.65695977e-01 -1.52288699e+00
-1.31308451e-01 -5.17551720e-01 4.68307883e-02 1.08901751e+00
3.91396970e-01 -2.28787601e-01 -1.67253864e+00 -5.57414591e-01
6.88570619e-01 -1.03648402e-01 7.87864447e-01 3.92245620e-01
9.98167574e-01 -4.51027364e-01 3.28299799e-03 3.71648997e-01
1.35658717e+00 2.68331617e-01 4.34642732e-01 -4.34937850e-02
7.45094359e-01 5.30929446e-01 8.29710886e-02 1.49552360e-01
1.30352139e-01 5.24986923e-01 7.21211657e-02 1.73998743e-01
-1.35261059e-01 -1.11370325e-01 6.79987669e-01 1.52539992e+00
2.65056401e-01 -1.74302563e-01 -7.96633363e-01 6.53420627e-01
-1.25291312e+00 -1.17540324e+00 4.79949147e-01 1.96800888e+00
5.54947436e-01 -9.19750854e-02 -3.65141630e-02 7.37020910e-01
8.78671885e-01 5.82266510e-01 -4.87383932e-01 -7.09600747e-01
-7.64957368e-02 1.32638931e-01 1.28360122e-01 9.85533059e-01
-1.11281991e+00 5.11502981e-01 6.49477291e+00 5.80382943e-01
-1.49221146e+00 3.70233387e-01 2.75213897e-01 5.76103143e-02
-1.94234177e-01 -5.06387115e-01 -1.11367857e+00 2.31675535e-01
1.48278058e+00 -2.84625500e-01 6.44492358e-02 1.01320767e+00
1.79330036e-01 6.51484787e-01 -1.16235280e+00 1.28298545e+00
6.29155040e-01 -8.79057229e-01 1.08096689e-01 2.18495578e-01
2.92114168e-01 1.00182943e-01 3.54818583e-01 5.41921377e-01
-1.93962723e-01 -1.03712606e+00 6.11395955e-01 2.56098479e-01
7.41643965e-01 -8.37275505e-01 1.04979539e+00 1.67544842e-01
-1.09602487e+00 -1.85661107e-01 -3.40311915e-01 2.94530720e-01
7.65116438e-02 2.42064774e-01 -1.28564799e+00 2.74243057e-01
3.41516167e-01 5.54947197e-01 -2.08146125e-01 6.37437701e-01
-6.12608157e-02 8.67900670e-01 3.55860367e-02 -2.50733912e-01
2.41148651e-01 -3.38521376e-02 8.18464279e-01 1.34916210e+00
9.17230174e-02 -2.33188912e-01 -2.43232384e-01 6.02642536e-01
-2.97466516e-01 1.35440558e-01 -5.57841301e-01 -2.46764794e-01
4.71435547e-01 9.91138756e-01 -2.15111356e-02 -4.18977708e-01
-2.90292412e-01 9.06158745e-01 -2.50453260e-02 5.88020623e-01
-6.62312031e-01 -7.22513855e-01 1.04172480e+00 -3.76238942e-01
6.97006881e-01 -4.70384866e-01 -2.97556251e-01 -9.20803010e-01
1.24231480e-01 -8.62757087e-01 -2.39342283e-02 -1.84057161e-01
-1.26805985e+00 9.86488581e-01 -3.48121732e-01 -1.10249472e+00
-4.57662314e-01 -7.91659832e-01 -8.22487652e-01 1.32424831e+00
-1.65380728e+00 -1.13900685e+00 9.26740840e-02 5.39169729e-01
7.77220011e-01 -9.08167958e-01 1.24510145e+00 4.79958147e-01
-7.40147233e-01 1.03054237e+00 5.04099250e-01 4.56013232e-01
3.97193253e-01 -1.23504925e+00 2.00992316e-01 9.07758594e-01
2.06594065e-01 8.28035176e-01 6.40109181e-01 -1.88958332e-01
-1.43177915e+00 -9.47986007e-01 1.34351051e+00 -2.21691176e-01
3.73736978e-01 -6.59209192e-01 -8.31645846e-01 8.05499077e-01
4.84570295e-01 -3.03023279e-01 1.37468636e+00 1.09906740e-01
-4.71894860e-01 -3.89978260e-01 -1.25997353e+00 6.94734529e-02
1.25920132e-01 -1.28389430e+00 -1.01816618e+00 8.41686502e-02
7.24998474e-01 3.89234945e-02 -7.33131170e-01 5.42828813e-02
5.03197968e-01 -4.63659078e-01 9.52296734e-01 -6.19974375e-01
-8.65612328e-02 -2.95263916e-01 -7.81372905e-01 -1.10175550e+00
-2.12103456e-01 -4.46091771e-01 -2.20854864e-01 1.54039598e+00
4.79445606e-01 -7.05917835e-01 6.08949780e-01 3.90573621e-01
-1.52448699e-01 -4.09019768e-01 -1.48526967e+00 -7.75535643e-01
-3.41834396e-01 -5.46274900e-01 5.85551798e-01 8.51865828e-01
-1.51688576e-01 6.77158892e-01 -4.12950099e-01 6.05502248e-01
8.67672145e-01 -4.82086301e-01 5.18412113e-01 -1.19073951e+00
4.67949831e-05 -2.62958318e-01 -9.71019030e-01 -1.06080222e+00
6.43457353e-01 -9.54477906e-01 1.48163453e-01 -1.25220430e+00
2.18229722e-02 1.48021102e-01 -6.24444664e-01 4.07096416e-01
-1.08906902e-01 -6.28795922e-02 -1.21547855e-01 5.11601120e-02
-6.31045625e-02 1.07088065e+00 4.46725607e-01 -5.44756174e-01
-1.92180380e-01 2.36964896e-01 -6.88603640e-01 3.16171229e-01
7.77671933e-01 -4.34326559e-01 -4.86628056e-01 -3.75893861e-01
-9.22156930e-01 -2.47726589e-01 1.42267019e-01 -8.43873560e-01
3.52737427e-01 6.49979651e-01 4.12468642e-01 -1.06571496e+00
6.58187628e-01 -9.09448743e-01 -4.35889870e-01 4.88561153e-01
-5.57671785e-01 -5.99266682e-03 4.84818637e-01 5.51357090e-01
-6.62210703e-01 -3.27407539e-01 7.82245517e-01 5.86740732e-01
-5.41379750e-01 1.49190903e-01 -3.53590101e-01 -6.55133843e-01
7.28476703e-01 -1.82777300e-01 1.69618040e-01 -5.07637680e-01
-8.29256654e-01 -2.96695143e-01 -2.45183408e-01 7.32903361e-01
1.04901183e+00 -1.57130742e+00 -9.01604116e-01 7.48875499e-01
5.10550104e-02 -5.83537817e-01 1.75076976e-01 2.67762452e-01
-1.72454238e-01 9.60270226e-01 -1.01410426e-01 -6.69015884e-01
-1.67951584e+00 4.27220374e-01 3.72889608e-01 1.25410691e-01
-2.27785870e-01 1.01124728e+00 -4.30137403e-02 -8.97179306e-01
9.57409561e-01 -3.11260730e-01 -4.16208237e-01 6.23238496e-02
9.26776350e-01 4.08408672e-01 2.24745885e-01 -1.20139313e+00
-6.71263993e-01 1.92109033e-01 -4.91497785e-01 -5.06049156e-01
1.45025134e+00 -2.03475103e-01 -1.51640356e-01 6.96765184e-01
1.76168954e+00 3.92740816e-02 -6.28888726e-01 -5.25558472e-01
-1.37797166e-02 -1.60928354e-01 6.53674722e-01 -5.68309128e-01
-9.09444153e-01 1.17792177e+00 1.05470634e+00 6.64729699e-02
9.07290578e-01 -7.22485334e-02 7.20569432e-01 3.47729802e-01
-2.37106442e-01 -8.99797261e-01 -1.68870598e-01 3.61983299e-01
8.82973611e-01 -1.25479662e+00 -1.74818709e-01 2.17576861e-01
-5.59418499e-01 1.16496408e+00 2.99593210e-01 2.81229436e-01
1.12284899e+00 2.00058836e-02 6.11401677e-01 -1.15136176e-01
-6.19498491e-01 2.80789942e-01 6.18574202e-01 5.74972272e-01
5.56572199e-01 1.95617869e-01 3.05976868e-02 6.33458912e-01
-3.24053705e-01 -4.23150033e-01 8.05905014e-02 7.54421830e-01
-4.14306253e-01 -1.08493471e+00 -5.71395457e-01 2.20400557e-01
-5.65768659e-01 -7.73797557e-02 -3.31824660e-01 4.40982968e-01
-2.71276057e-01 1.10660934e+00 6.15850687e-02 -6.18040085e-01
2.44588166e-01 6.35751665e-01 7.45095909e-02 -3.66276801e-01
-3.77181977e-01 5.23778424e-02 -6.43926337e-02 -4.17120345e-02
-3.68211389e-01 -6.97919011e-01 -1.01740134e+00 -1.41028836e-01
-5.93732059e-01 3.53723794e-01 1.52380168e+00 1.00710225e+00
1.92816377e-01 5.16433954e-01 1.08550644e+00 -8.20676565e-01
-1.17562985e+00 -1.61123395e+00 -8.34641874e-01 -9.87618510e-03
1.02540207e+00 -3.13254088e-01 -8.85177016e-01 8.47796425e-02] | [14.3565673828125, 6.09227991104126] |
c7b8b28a-dd5c-4486-aa03-10eba41b76f9 | distributional-reinforcement-learning-with-6 | 2305.16877 | null | https://arxiv.org/abs/2305.16877v1 | https://arxiv.org/pdf/2305.16877v1.pdf | Distributional Reinforcement Learning with Dual Expectile-Quantile Regression | Successful applications of distributional reinforcement learning with quantile regression prompt a natural question: can we use other statistics to represent the distribution of returns? In particular, expectile regression is known to be more efficient than quantile regression for approximating distributions, especially on extreme values, and by providing a straightforward estimator of the mean it is a natural candidate for reinforcement learning. Prior work has answered this question positively in the case of expectiles, with the major caveat that expensive computations must be performed to ensure convergence. In this work, we propose a dual expectile-quantile approach which solves the shortcomings of previous work while leveraging the complementary properties of expectiles and quantiles. Our method outperforms both quantile-based and expectile-based baselines on the MuJoCo continuous control benchmark. | ['Maarten de Rijke', 'Paul Groth', 'Jean-Michel Renders', 'Romain Deffayet', 'Sami Jullien'] | 2023-05-26 | null | null | null | null | ['distributional-reinforcement-learning', 'continuous-control'] | ['methodology', 'playing-games'] | [-4.74739134e-01 -6.22484833e-02 -5.72539091e-01 -4.58786935e-01
-1.15233040e+00 -5.16741037e-01 6.72464669e-01 5.70096195e-01
-4.84475523e-01 1.17658865e+00 4.07837749e-01 -6.11650705e-01
-2.72215128e-01 -9.63546872e-01 -6.97664499e-01 -5.40000558e-01
-1.74203321e-01 5.85885704e-01 -1.06869705e-01 -3.10726672e-01
5.73290348e-01 3.12267154e-01 -1.30702603e+00 -1.57744750e-01
9.43007410e-01 1.18855906e+00 -4.18881118e-01 5.20749569e-01
-8.78654495e-02 9.01696563e-01 -6.25291228e-01 -4.73521560e-01
2.18016341e-01 -4.81037796e-01 -5.07853866e-01 -5.42687654e-01
3.25537831e-01 -6.01196110e-01 -1.77918207e-02 8.07841122e-01
4.03892070e-01 3.15865189e-01 1.10626018e+00 -1.29995155e+00
-3.97457838e-01 6.00487471e-01 -7.93581486e-01 2.31957883e-01
2.31021270e-01 1.76878184e-01 1.22408152e+00 -3.82505357e-01
5.53638600e-02 1.23880517e+00 7.28303552e-01 2.83114463e-01
-1.49265683e+00 -5.80002725e-01 -1.69767756e-02 -3.04959297e-01
-8.99179161e-01 -1.75177425e-01 3.76121223e-01 -4.44898248e-01
7.72407651e-01 -3.55147161e-02 4.39147353e-01 1.02449274e+00
5.14681816e-01 8.33245277e-01 1.41569924e+00 -2.60867834e-01
5.91950715e-01 1.10094808e-01 4.20640260e-02 1.88748926e-01
2.89238870e-01 6.74963713e-01 -2.94633597e-01 -4.93569046e-01
8.93876970e-01 5.66461086e-02 1.71012297e-01 -4.19941455e-01
-8.14968884e-01 1.32596588e+00 1.89536795e-01 -1.21550038e-01
-6.10614896e-01 7.22605646e-01 6.83160484e-01 5.08109808e-01
7.40013003e-01 4.09879088e-01 -4.96962309e-01 -5.70952356e-01
-1.01575387e+00 8.80507648e-01 8.61183584e-01 7.54642069e-01
7.00299382e-01 2.81660229e-01 -4.60619926e-01 5.21278441e-01
1.40390009e-01 6.14729583e-01 2.44650841e-01 -1.10270131e+00
4.69036132e-01 2.90974025e-02 4.89385098e-01 -4.74900424e-01
-2.67735451e-01 -3.83283347e-01 -4.19942439e-01 6.28945172e-01
1.00483441e+00 -6.00813568e-01 -5.36037743e-01 1.65130067e+00
2.25857005e-01 9.50786397e-02 -2.33076550e-02 5.33558726e-01
-3.07326410e-02 5.03754795e-01 4.25394773e-01 -7.53795281e-02
9.20423746e-01 -3.02034676e-01 -3.42053801e-01 1.31441131e-01
3.88794452e-01 -5.40511608e-01 1.31361043e+00 4.62432116e-01
-8.92036796e-01 -2.73837507e-01 -7.70869732e-01 3.75167251e-01
-2.01244727e-01 -4.12155092e-01 9.40784276e-01 1.03215766e+00
-6.85577095e-01 8.79347265e-01 -7.64915228e-01 1.74799562e-01
3.36085707e-01 9.59874541e-02 6.33619651e-02 1.99211091e-01
-9.57597494e-01 9.47777152e-01 3.51522416e-01 -5.98446786e-01
-8.44009817e-01 -1.02255499e+00 -7.23960161e-01 3.62880766e-01
4.84923035e-01 -4.27127898e-01 1.58694184e+00 -7.02767313e-01
-1.65790343e+00 6.34711832e-02 1.79612219e-01 -9.16055381e-01
6.88371122e-01 -3.20650518e-01 2.83876304e-02 -1.28086388e-01
-4.20536362e-02 3.17462355e-01 7.28990972e-01 -8.01813245e-01
-7.77230620e-01 -3.25552166e-01 1.26284570e-01 -1.11752160e-01
-9.35915858e-02 -1.34364843e-01 4.39475179e-01 -5.82673371e-01
-5.59590399e-01 -5.64453959e-01 -3.94869596e-01 -4.50545758e-01
-1.38092577e-01 -6.80417418e-01 2.34998643e-01 -3.55142444e-01
1.09716201e+00 -2.11711550e+00 -3.98092002e-01 4.10913765e-01
-3.93217504e-02 -2.75039673e-01 6.92015663e-02 6.89184785e-01
-3.15511711e-02 8.40749219e-02 -1.88490197e-01 -6.97548389e-02
4.92887914e-01 -4.56996635e-02 -8.25506866e-01 6.54347241e-01
1.40107438e-01 7.91435182e-01 -9.09756362e-01 -3.90262961e-01
8.88588205e-02 2.37256154e-01 -8.62075508e-01 1.96005002e-01
-5.45012057e-01 2.75608271e-01 -3.95018488e-01 5.20546854e-01
4.76996303e-01 1.70316219e-01 -1.69248711e-02 4.19542879e-01
-1.94686979e-01 3.57181221e-01 -1.04482305e+00 1.04447281e+00
-5.09356618e-01 2.96471566e-01 -2.84496337e-01 -1.17618334e+00
1.00106096e+00 1.73303589e-01 7.44990468e-01 -6.56263411e-01
1.14872791e-01 3.74047548e-01 -1.66299433e-01 -1.26781344e-01
5.27740240e-01 -6.98407471e-01 -3.69371891e-01 6.20256484e-01
-6.00897335e-02 -4.65554327e-01 2.90431052e-01 -7.42602125e-02
8.16815853e-01 5.77194929e-01 3.41910958e-01 -5.50113261e-01
-3.04354113e-02 -1.60425156e-02 4.41639870e-01 8.26634586e-01
-2.38108084e-01 3.74851227e-01 1.21213460e+00 -2.54694700e-01
-1.01971924e+00 -1.52642548e+00 -2.18331367e-01 1.22800279e+00
-2.97605246e-01 -9.23414603e-02 -6.23830676e-01 -8.53733361e-01
5.60344160e-01 1.14278698e+00 -6.80685401e-01 5.67620993e-02
-3.39868307e-01 -6.52570963e-01 5.98251343e-01 8.60458851e-01
8.05730969e-02 -8.84601235e-01 -8.27259660e-01 2.86293387e-01
2.36722216e-01 -4.69138056e-01 -3.83180052e-01 4.54131454e-01
-9.14789200e-01 -1.16435111e+00 -8.31306100e-01 -3.53366951e-03
7.19532743e-02 -2.82283008e-01 1.39012754e+00 -3.51777703e-01
1.03203595e-01 5.84175348e-01 -1.41917929e-01 -8.49205136e-01
-2.65224785e-01 6.77716509e-02 -1.61835134e-01 -2.79413372e-01
2.74662822e-01 -7.52599359e-01 -6.17427826e-01 -6.17935471e-02
-8.84480357e-01 -7.38019705e-01 3.18653971e-01 9.22899425e-01
5.01009881e-01 -1.08208761e-01 1.37467718e+00 -1.13123012e+00
1.09788203e+00 -8.48160684e-01 -9.76285100e-01 -1.02691010e-01
-7.56377578e-01 1.82592213e-01 1.00860870e+00 -2.29080305e-01
-9.95135307e-01 -1.38386115e-01 -2.28585660e-01 -1.62367336e-02
-1.29752293e-01 7.27459908e-01 4.14338142e-01 5.48076689e-01
6.69219911e-01 9.36672278e-03 2.49588281e-01 -3.56371343e-01
3.18489313e-01 3.54568481e-01 4.29787815e-01 -1.07585871e+00
6.36788666e-01 2.42343038e-01 1.98944896e-01 -6.06817007e-01
-8.76720965e-01 -2.82164305e-01 -1.81893453e-01 1.85043085e-02
6.03918016e-01 -7.73714840e-01 -1.12771738e+00 -9.70421061e-02
-6.59537554e-01 -6.16216063e-01 -6.96726024e-01 6.64699435e-01
-1.13245964e+00 1.35980532e-01 -4.61728841e-01 -1.15009272e+00
-1.68104708e-01 -7.18814373e-01 7.37124085e-01 2.51589954e-01
-3.08199465e-01 -1.19204414e+00 2.89946616e-01 -3.19629699e-01
7.65668094e-01 7.80019343e-01 1.17024100e+00 -7.02114999e-01
-1.01617455e-01 -2.76817828e-01 -1.28998756e-01 5.12245595e-01
-1.05743110e-01 -2.48472262e-02 -7.06933498e-01 -2.71035135e-01
-1.67133018e-01 -7.30985582e-01 9.83030379e-01 7.44650185e-01
1.02250195e+00 -1.13046154e-01 2.72183836e-01 2.75950760e-01
1.46226692e+00 -6.94122165e-02 4.53714669e-01 4.22055900e-01
-1.14064641e-01 4.05559421e-01 8.60581338e-01 9.05774891e-01
4.17843938e-01 1.48978785e-01 5.81896961e-01 1.56848535e-01
3.95042181e-01 -4.81280655e-01 4.58453774e-01 4.85316329e-02
-2.43483335e-02 8.91252384e-02 -6.94082141e-01 5.55858552e-01
-1.73862696e+00 -1.11390162e+00 2.11463068e-02 2.52253628e+00
8.69999766e-01 3.25878561e-01 8.12864721e-01 -5.22161322e-03
1.97861984e-01 5.25277965e-02 -6.12412870e-01 -6.37028277e-01
2.14896709e-01 6.68387532e-01 6.51364207e-01 4.39268589e-01
-1.00056720e+00 5.12528181e-01 7.38345671e+00 6.99393809e-01
-8.71110260e-01 -2.46228367e-01 8.54399860e-01 -1.69076517e-01
-4.17843223e-01 5.53785451e-02 -7.13698328e-01 4.39161927e-01
1.19418848e+00 -4.44413006e-01 3.68432164e-01 1.23841369e+00
1.03227518e-01 -4.88050729e-01 -1.20237601e+00 5.17206967e-01
-4.65116084e-01 -1.03594744e+00 -3.21020305e-01 1.28135696e-01
7.29033232e-01 1.20687462e-01 3.50236654e-01 8.23017657e-01
7.35307038e-01 -1.32277775e+00 6.48215353e-01 7.04177678e-01
5.50264835e-01 -1.44846880e+00 7.66551077e-01 4.56795096e-01
-6.77002430e-01 -1.41689122e-01 -6.53191328e-01 -1.52111739e-01
-1.49426192e-01 6.82752669e-01 -7.10027397e-01 3.35862607e-01
3.72821867e-01 4.01916504e-01 -3.40786189e-01 1.14610684e+00
-5.86802885e-02 1.05721283e+00 -4.35220331e-01 -7.42833465e-02
5.41457236e-01 -3.27295274e-01 1.04641333e-01 1.23889637e+00
4.48433965e-01 -3.19232613e-01 2.49066755e-01 8.82636905e-01
1.50590986e-01 2.28170782e-01 -6.17902577e-01 2.05160994e-02
4.04969901e-01 9.13581491e-01 -3.50684881e-01 -3.71915996e-01
-4.43782479e-01 1.55511200e-01 3.45627427e-01 4.17148530e-01
-8.96445811e-01 -4.76817369e-01 6.03766978e-01 2.24763393e-01
3.97699326e-01 -3.85281324e-01 -3.91073048e-01 -8.01954091e-01
-1.79483056e-01 -8.54536951e-01 5.94313085e-01 -1.71484768e-01
-1.39986980e+00 1.65483087e-01 2.88168699e-01 -9.46968615e-01
-8.59605730e-01 -6.19468331e-01 -7.30270505e-01 9.96191621e-01
-1.79298770e+00 -5.40001035e-01 2.48763129e-01 5.85014760e-01
2.56637633e-01 -1.61702365e-01 7.20579565e-01 -1.20212957e-02
-9.70416516e-02 4.71924514e-01 5.65115333e-01 -7.39912763e-02
8.17443371e-01 -1.86213303e+00 1.64362237e-01 1.51112363e-01
4.61464236e-03 4.86629963e-01 1.09367800e+00 -5.42120755e-01
-1.38899183e+00 -8.57643485e-01 3.22055638e-01 -4.26976025e-01
7.55155027e-01 5.97362332e-02 -6.85805798e-01 7.05039084e-01
2.77976841e-01 -1.39974505e-01 6.33319676e-01 3.91639560e-01
-4.33400959e-01 -1.69973582e-01 -1.05885422e+00 3.72134447e-01
1.25911325e-01 -3.78029555e-01 -4.00484711e-01 2.63261020e-01
3.82848680e-01 -2.16623217e-01 -1.18963838e+00 1.44705594e-01
5.99692285e-01 -1.30110717e+00 7.95098305e-01 -6.38454020e-01
7.30655909e-01 3.72467227e-02 -2.32402802e-01 -1.60940671e+00
1.41062543e-01 -6.82375908e-01 -9.97615978e-02 1.25738358e+00
2.93401450e-01 -8.67827952e-01 8.72069716e-01 4.60423291e-01
1.33130670e-01 -1.02577281e+00 -1.04551935e+00 -1.15537965e+00
7.95331001e-01 -5.64715624e-01 7.95658886e-01 5.07279277e-01
2.04576984e-01 1.36653781e-02 -5.30281544e-01 -2.84148812e-01
8.32653701e-01 2.59289145e-01 9.47563112e-01 -1.04814970e+00
-5.01149356e-01 -6.27292633e-01 -4.74618971e-02 -1.06900048e+00
2.59596020e-01 -7.43020594e-01 6.23724908e-02 -1.15945375e+00
-1.94113180e-02 -4.00527894e-01 -5.18220842e-01 2.09598079e-01
-1.00265406e-01 8.89210904e-04 8.79893303e-02 -3.11072975e-01
-4.31959927e-01 8.26458752e-01 9.40119386e-01 9.92401466e-02
-1.15091458e-01 4.83805895e-01 -8.48258734e-01 5.92161000e-01
1.03302372e+00 -4.47528362e-01 -3.34099591e-01 2.16549456e-01
3.12619030e-01 4.92454290e-01 2.34589994e-01 -7.19383180e-01
-1.32573778e-02 -6.60329700e-01 5.04478872e-01 -6.61046386e-01
4.74377573e-02 -5.20888805e-01 -5.49457312e-01 2.63373464e-01
-5.45238972e-01 3.77467901e-01 2.02066466e-01 6.99359596e-01
-3.17536473e-01 -3.17525595e-01 8.42116237e-01 -5.74130267e-02
-1.08702995e-01 2.26328403e-01 -5.80487967e-01 4.90086913e-01
8.48509192e-01 1.21692829e-01 -1.72956854e-01 -8.30049455e-01
-2.72548884e-01 3.90237033e-01 2.46363088e-01 2.57144850e-02
4.32612062e-01 -1.23766482e+00 -9.59921539e-01 -1.34098813e-01
-2.32848734e-01 -3.17467809e-01 -2.99569249e-01 8.24603558e-01
-2.79065251e-01 2.33476892e-01 -1.88131198e-01 -2.56308317e-01
-5.50726414e-01 5.34795403e-01 3.11203152e-01 -5.38298011e-01
-5.04287779e-01 3.37572932e-01 9.31832790e-02 -3.66468906e-01
2.82121569e-01 -4.30496216e-01 8.01307335e-02 2.62413695e-02
5.72370350e-01 6.76396668e-01 -2.66731501e-01 7.07251728e-02
-5.71877090e-03 1.93332627e-01 3.55377942e-02 -2.68390656e-01
1.65583670e+00 1.81023180e-01 1.89418986e-01 5.98066092e-01
8.19683075e-01 1.99543700e-01 -1.65409470e+00 7.89759085e-02
5.72845161e-01 -3.85285676e-01 -1.92129388e-01 -7.77794242e-01
-7.70933390e-01 1.02999115e+00 2.07624376e-01 5.64699531e-01
8.60441566e-01 -2.63413459e-01 6.15984917e-01 2.36569121e-01
2.41797224e-01 -1.35140848e+00 1.73596293e-02 7.23790944e-01
7.23408401e-01 -1.16286695e+00 2.25308627e-01 2.03054637e-01
-7.97427416e-01 1.13806939e+00 3.81086916e-01 -8.03298056e-01
6.41322434e-01 3.38153273e-01 -9.58494619e-02 8.81098583e-02
-8.10479939e-01 -1.13981403e-01 1.97758712e-02 5.35837412e-01
8.10760677e-01 2.19472036e-01 -3.50642830e-01 4.74990696e-01
-4.77627307e-01 -1.91254586e-01 5.41912675e-01 8.13394725e-01
-5.53235590e-01 -1.26455593e+00 -3.86240751e-01 6.89071953e-01
-8.39563787e-01 -9.09946784e-02 8.59970823e-02 1.12001026e+00
-4.27965134e-01 8.51335883e-01 1.32462293e-01 1.26761317e-01
4.15737718e-01 1.99854627e-01 4.78194743e-01 -4.69567865e-01
-6.97016358e-01 1.14532061e-01 -5.56562841e-02 -5.12355268e-01
2.73304828e-03 -8.69357228e-01 -1.23838758e+00 -3.77215028e-01
-1.57567874e-01 3.96892160e-01 6.09060109e-01 8.74709070e-01
-5.98597974e-02 3.01880419e-01 6.99759126e-01 -5.89479506e-01
-1.64719415e+00 -8.96007419e-01 -9.18685138e-01 2.98156410e-01
5.33249915e-01 -6.20307565e-01 -2.73591161e-01 -4.67343152e-01] | [4.099339962005615, 2.5989716053009033] |
9bf73746-3b6a-46f5-a55e-2caa14d0bf95 | style-normalization-and-restitution-for | 2005.11037 | null | https://arxiv.org/abs/2005.11037v1 | https://arxiv.org/pdf/2005.11037v1.pdf | Style Normalization and Restitution for Generalizable Person Re-identification | Existing fully-supervised person re-identification (ReID) methods usually suffer from poor generalization capability caused by domain gaps. The key to solving this problem lies in filtering out identity-irrelevant interference and learning domain-invariant person representations. In this paper, we aim to design a generalizable person ReID framework which trains a model on source domains yet is able to generalize/perform well on target domains. To achieve this goal, we propose a simple yet effective Style Normalization and Restitution (SNR) module. Specifically, we filter out style variations (e.g., illumination, color contrast) by Instance Normalization (IN). However, such a process inevitably removes discriminative information. We propose to distill identity-relevant feature from the removed information and restitute it to the network to ensure high discrimination. For better disentanglement, we enforce a dual causal loss constraint in SNR to encourage the separation of identity-relevant features and identity-irrelevant features. Extensive experiments demonstrate the strong generalization capability of our framework. Our models empowered by the SNR modules significantly outperform the state-of-the-art domain generalization approaches on multiple widely-used person ReID benchmarks, and also show superiority on unsupervised domain adaptation. | ['Wen-Jun Zeng', 'Xin Jin', 'Li Zhang', 'Zhibo Chen', 'Cuiling Lan'] | 2020-05-22 | style-normalization-and-restitution-for-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Jin_Style_Normalization_and_Restitution_for_Generalizable_Person_Re-Identification_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Jin_Style_Normalization_and_Restitution_for_Generalizable_Person_Re-Identification_CVPR_2020_paper.pdf | cvpr-2020-6 | ['generalizable-person-re-identification'] | ['computer-vision'] | [ 2.11301401e-01 -5.00037849e-01 6.24491647e-02 -6.43145382e-01
-2.31555298e-01 -6.78619742e-01 7.02705979e-01 -1.20237462e-01
-6.92866027e-01 7.36771107e-01 3.51720959e-01 1.97139516e-01
-2.59943217e-01 -6.94325924e-01 -4.48349208e-01 -8.43701720e-01
1.86874524e-01 3.32548112e-01 -1.40781283e-01 -2.86666393e-01
7.95912743e-02 5.57649791e-01 -1.47796237e+00 -6.87769800e-02
1.41308594e+00 5.99011064e-01 -7.83327501e-03 -8.24296102e-02
-1.15769610e-01 2.68741727e-01 -7.72715092e-01 -5.12734950e-01
4.38337415e-01 -4.83774215e-01 -4.91229326e-01 -8.59864280e-02
5.78927100e-01 -3.45473558e-01 -5.65120578e-01 1.24378490e+00
8.62772048e-01 4.08570051e-01 8.47432733e-01 -1.17629778e+00
-1.08752692e+00 3.49206835e-01 -5.91412067e-01 2.42009178e-01
4.02552515e-01 1.73249200e-01 4.45439011e-01 -8.45022082e-01
4.22887087e-01 1.39836490e+00 4.21521246e-01 9.71810281e-01
-1.52819681e+00 -1.19456446e+00 5.29896617e-01 3.28070372e-01
-1.63017237e+00 -5.41608512e-01 1.04250348e+00 -3.07695240e-01
3.66852969e-01 2.69432843e-01 2.13493615e-01 1.67922127e+00
-2.54442692e-01 7.03283191e-01 1.32392228e+00 -2.05122501e-01
1.59282058e-01 2.91928142e-01 3.59476358e-01 1.74513683e-01
4.20110732e-01 1.99354082e-01 -5.77903688e-01 7.06935301e-02
7.35937059e-01 8.94710496e-02 -3.50737035e-01 -4.46247160e-01
-1.10318494e+00 3.57421994e-01 5.89954495e-01 1.46841586e-01
-1.59507066e-01 -5.79382002e-01 4.54617858e-01 4.79286999e-01
2.94408083e-01 4.34539348e-01 -1.93183780e-01 2.92555362e-01
-6.20074749e-01 3.33326757e-01 4.12575454e-01 1.12584639e+00
6.97837353e-01 -1.07841395e-01 -7.21404493e-01 1.18451238e+00
-9.67544615e-02 4.92916644e-01 6.32466316e-01 -4.64273661e-01
5.45251191e-01 7.03960717e-01 3.98063362e-02 -7.55701184e-01
-4.73430812e-01 -8.78935874e-01 -1.22158039e+00 1.11109458e-01
5.42312384e-01 -4.32007946e-02 -9.58916605e-01 2.10192513e+00
2.12646589e-01 2.47479767e-01 1.37821838e-01 1.12658715e+00
7.75061965e-01 8.00842717e-02 5.69459438e-01 5.45273647e-02
1.47721291e+00 -6.13990426e-01 -5.07588089e-01 -3.75767946e-01
1.72026664e-01 -4.62878078e-01 1.02664852e+00 1.97429091e-01
-6.96523070e-01 -9.73487496e-01 -1.19358659e+00 -6.35186285e-02
-4.72176164e-01 2.56538004e-01 3.70680958e-01 8.26746523e-01
-6.50949597e-01 5.56432188e-01 -3.27349871e-01 -5.72538555e-01
3.95450443e-01 5.81946373e-01 -7.22909510e-01 -3.05212498e-01
-1.29383147e+00 7.92805254e-01 5.01558721e-01 8.24677795e-02
-4.68324721e-01 -6.39824688e-01 -8.02507520e-01 -3.36744189e-02
1.20547786e-01 -9.79529381e-01 7.37865388e-01 -1.13636768e+00
-1.54549396e+00 1.02600920e+00 -1.60557076e-01 -1.77351654e-01
7.55853117e-01 -2.55855799e-01 -6.98576629e-01 -1.97339356e-01
2.56299734e-01 5.28420925e-01 8.54090452e-01 -1.29280210e+00
-5.88136017e-01 -7.14031637e-01 9.67947766e-03 3.69951665e-01
-8.83344293e-01 1.20763794e-01 -4.11196262e-01 -9.38939214e-01
1.71291158e-01 -8.16731989e-01 2.01566368e-01 -2.02597499e-01
-3.85956824e-01 -3.12695742e-01 4.86789644e-01 -8.83127749e-01
8.71814907e-01 -2.18906713e+00 2.05416977e-01 2.96614975e-01
1.35989815e-01 3.57390672e-01 -4.37090337e-01 -2.28541512e-02
-1.74464032e-01 -1.92513898e-01 -6.22049626e-03 -5.18432200e-01
1.40120676e-02 1.44299418e-01 -1.53197214e-01 5.88681698e-01
2.08659694e-01 7.67709613e-01 -8.47487271e-01 -3.68337452e-01
2.20292732e-01 4.87033278e-01 -5.08172274e-01 2.19948038e-01
2.91183233e-01 1.04713380e+00 -5.55572629e-01 4.98624653e-01
1.09190786e+00 2.21978709e-01 -2.29100883e-02 -2.76126236e-01
5.20070903e-02 2.45507509e-01 -1.23866844e+00 1.78771365e+00
-2.16100886e-01 1.70875624e-01 -2.10394226e-02 -1.00239193e+00
1.23024261e+00 -2.97249313e-02 2.23059088e-01 -1.06249404e+00
1.49405375e-01 5.66508137e-02 -6.24483377e-02 -2.22737670e-01
2.87003696e-01 -3.70933861e-02 -9.60465446e-02 8.82110074e-02
1.28097951e-01 7.11554348e-01 -6.89251646e-02 -5.19923717e-02
6.31387830e-01 2.17192039e-01 8.71923193e-02 -4.93763357e-01
9.47556078e-01 -3.52025896e-01 9.34807658e-01 8.77227902e-01
-4.96933639e-01 5.77139437e-01 5.25774471e-02 -2.75252789e-01
-9.03943598e-01 -1.26379991e+00 -2.08291501e-01 1.27119923e+00
5.90305626e-01 -9.34423357e-02 -7.74671853e-01 -8.21528614e-01
1.73909292e-01 6.61423504e-01 -7.12536216e-01 -5.25423169e-01
-7.00840414e-01 -8.92356694e-01 6.77040994e-01 7.94281840e-01
9.96315718e-01 -6.34556890e-01 2.02768445e-01 1.03443108e-01
-2.34937415e-01 -1.00601017e+00 -6.15650594e-01 -1.89639270e-01
-4.89318609e-01 -7.46420562e-01 -1.13698936e+00 -8.97096097e-01
8.76030326e-01 5.51525593e-01 7.29166389e-01 -2.13768274e-01
-4.84575182e-02 1.99052438e-01 -2.58756995e-01 -2.57982850e-01
-7.40106031e-02 1.85881510e-01 5.17329216e-01 2.29095757e-01
6.68974698e-01 -7.44723976e-01 -7.79442012e-01 7.29031324e-01
-6.09997332e-01 -2.73461472e-02 6.45103216e-01 1.00535572e+00
3.13551039e-01 1.06407791e-01 8.05224478e-01 -4.91830468e-01
7.25690603e-01 -2.79302448e-01 -3.71692955e-01 3.79983991e-01
-5.52320540e-01 8.91904980e-02 7.43378937e-01 -6.94370747e-01
-1.45095325e+00 -5.68080284e-02 8.38325396e-02 -1.84946179e-01
-4.89986509e-01 -9.96365678e-03 -7.96413660e-01 -1.08888134e-01
8.45865905e-01 3.93811196e-01 -1.74865440e-01 -6.56547487e-01
3.30661684e-01 7.01911926e-01 9.64048922e-01 -9.66448426e-01
1.15982449e+00 5.41932583e-01 -2.10533842e-01 -6.01031244e-01
-6.28477871e-01 -5.73196411e-01 -7.81405210e-01 9.60059185e-03
6.49867475e-01 -1.22093546e+00 -6.71319902e-01 5.39289653e-01
-9.65286016e-01 9.44419652e-02 -3.51079889e-02 4.60384309e-01
-1.45243362e-01 5.49702823e-01 -3.07823569e-01 -6.86320841e-01
-3.86708051e-01 -7.35608459e-01 7.18658805e-01 6.34505987e-01
-1.90035120e-01 -5.67425370e-01 -9.13691446e-02 3.93812597e-01
5.24136782e-01 2.93041784e-02 6.32494926e-01 -8.63099694e-01
-1.24968663e-01 -1.65779233e-01 -6.57179654e-01 4.32281941e-01
2.38549218e-01 -6.71813846e-01 -1.28325343e+00 -5.35320580e-01
-2.03724086e-01 1.53964991e-02 1.10538840e+00 -6.18586550e-03
1.15982604e+00 -1.26021311e-01 -4.12120879e-01 9.25204694e-01
9.86491561e-01 -5.42679131e-02 5.66211641e-01 5.72679222e-01
8.68458629e-01 7.56282270e-01 2.66763180e-01 2.85945863e-01
4.42333907e-01 8.34357202e-01 -1.62192136e-01 -1.00516394e-01
-2.58864075e-01 -5.00198543e-01 3.50104839e-01 1.56058371e-01
-3.58795315e-01 -6.82022944e-02 -5.45248926e-01 3.84654939e-01
-1.82869279e+00 -9.12151098e-01 1.08481146e-01 2.45231318e+00
7.99079895e-01 8.39980990e-02 3.92721713e-01 4.95802425e-02
9.25007820e-01 -2.22293571e-01 -7.63550878e-01 1.60555661e-01
-4.89419997e-01 3.42700183e-02 5.03818870e-01 8.58124718e-02
-1.29867446e+00 9.29215729e-01 4.83962440e+00 8.17180514e-01
-8.39909315e-01 1.28344093e-02 3.61270279e-01 -2.18827184e-02
-9.46820304e-02 -3.75928789e-01 -8.86449516e-01 4.68730181e-01
3.82371873e-01 -3.29994857e-01 5.76259553e-01 7.84901202e-01
1.45471290e-01 3.94181430e-01 -1.24414337e+00 1.21814573e+00
1.74098685e-01 -5.27457356e-01 1.87716722e-01 -4.74559925e-02
6.13584697e-01 -4.77886170e-01 1.81923792e-01 6.42829776e-01
1.07309513e-01 -9.54343021e-01 5.98508239e-01 4.35834914e-01
7.71576762e-01 -8.43294621e-01 6.41244471e-01 2.39449099e-01
-1.10385561e+00 -3.02051097e-01 -6.46062136e-01 -4.18312326e-02
-1.92931309e-01 3.58621627e-01 -3.34763348e-01 7.83302784e-01
6.99953020e-01 6.75828397e-01 -7.25893497e-01 1.09222567e+00
-3.87854159e-01 9.06777307e-02 -1.36690110e-01 3.45006049e-01
-3.22067529e-01 -1.94944367e-01 5.48778474e-01 1.29111445e+00
1.58229321e-01 1.97539240e-01 2.34280571e-01 1.11293685e+00
-1.11164249e-01 1.50685713e-01 -3.24135840e-01 4.54738677e-01
6.13836229e-01 9.85634685e-01 -3.10835987e-01 -3.01916659e-01
-3.71168554e-01 1.59069943e+00 3.98361266e-01 7.38586426e-01
-6.82635188e-01 -3.03487927e-01 1.04178584e+00 -1.27822636e-02
-4.92601395e-02 -1.39463082e-01 -4.13160771e-01 -1.51729524e+00
2.30949923e-01 -8.75306904e-01 5.69548249e-01 -2.83084303e-01
-1.93838274e+00 4.58400935e-01 -4.49980535e-02 -1.40051711e+00
6.01517037e-02 -5.35564125e-01 -6.53269887e-01 1.38496172e+00
-1.76013136e+00 -1.52158380e+00 -7.23478615e-01 1.07560301e+00
2.75841385e-01 -3.32477748e-01 7.30811179e-01 6.07527018e-01
-9.28316951e-01 1.34148848e+00 1.04900233e-01 4.60028082e-01
1.22954369e+00 -1.16179240e+00 4.84997392e-01 9.52619076e-01
-3.08223903e-01 1.15534580e+00 6.22683764e-01 -6.43178582e-01
-1.24793077e+00 -1.17265117e+00 6.00823045e-01 -3.09521586e-01
1.98786154e-01 -5.52077830e-01 -1.12351418e+00 4.84090716e-01
-2.03656778e-01 -9.24065560e-02 5.73797524e-01 2.65734613e-01
-8.84637415e-01 -4.18948412e-01 -1.29352999e+00 6.72324240e-01
1.56316745e+00 -5.66920757e-01 -7.38993406e-01 -4.12674993e-03
2.45637804e-01 -8.48160535e-02 -6.07574999e-01 4.96658117e-01
7.27604091e-01 -7.95922697e-01 1.30118477e+00 -5.88067710e-01
-8.92362371e-02 -4.10770684e-01 9.75741744e-02 -1.34673023e+00
-7.99109280e-01 -5.07594585e-01 2.17036143e-01 1.81291521e+00
-7.79788271e-02 -9.06794548e-01 6.40559673e-01 8.20158541e-01
1.51192114e-01 1.11944012e-01 -8.67018938e-01 -1.26546133e+00
3.17705691e-01 7.21453652e-02 8.94305587e-01 1.16815686e+00
-2.02335529e-02 2.98706114e-01 -4.62701172e-01 4.50947672e-01
1.04337823e+00 -8.40432569e-02 9.45369422e-01 -1.56951106e+00
-5.87692894e-02 -6.10758722e-01 -4.71729428e-01 -1.11539245e+00
4.47261184e-01 -8.53159189e-01 -2.19061613e-01 -1.18288648e+00
5.05805254e-01 -5.50426185e-01 -7.69334078e-01 4.24464554e-01
-4.80512738e-01 9.57930163e-02 1.20804220e-01 3.39032352e-01
-3.72803748e-01 6.86638832e-01 1.08008599e+00 -3.82146865e-01
-3.98742586e-01 -4.59496789e-02 -1.08574665e+00 4.45182174e-01
9.17183399e-01 -1.58959836e-01 -3.59693557e-01 -5.12494445e-01
-3.88246596e-01 -5.35048842e-01 7.12625444e-01 -1.19601536e+00
2.42485240e-01 -6.51546866e-02 1.01157153e+00 -1.78008750e-01
1.70159236e-01 -5.30653536e-01 -8.72248933e-02 6.83470815e-02
-3.11479062e-01 -3.26749533e-01 2.17830479e-01 4.64630097e-01
-3.09396759e-02 2.16917366e-01 8.55208635e-01 7.78837427e-02
-9.25292194e-01 3.20718229e-01 2.18137458e-01 -9.11244601e-02
7.30059266e-01 -2.85025716e-01 -5.69011271e-01 -1.08595639e-01
-5.42719841e-01 3.69612128e-01 6.54810667e-01 8.47995043e-01
3.97486448e-01 -1.37010968e+00 -1.03919435e+00 6.40607297e-01
5.25320768e-01 -3.97511423e-01 5.50269186e-01 5.29318690e-01
2.48819873e-01 2.20106438e-01 -5.80076516e-01 -3.64601672e-01
-1.15763831e+00 7.83498108e-01 4.17651683e-01 7.77978227e-02
-6.64427698e-01 7.99254119e-01 7.39369988e-01 -4.60888594e-01
3.53951365e-01 2.14946553e-01 -4.46621656e-01 -8.00209343e-02
8.61729383e-01 4.28749442e-01 -7.25396797e-02 -7.83180535e-01
-6.23945415e-01 6.51608050e-01 -3.35274547e-01 1.33909658e-01
1.02481544e+00 -2.47671381e-01 1.95839494e-01 -2.59973079e-01
1.01784956e+00 -5.56191839e-02 -1.29981816e+00 -4.68024462e-01
-4.44369502e-02 -4.52655613e-01 -2.55551398e-01 -9.33652341e-01
-7.60652244e-01 6.14967942e-01 9.83742952e-01 -2.27062196e-01
1.27074981e+00 -1.39154658e-01 5.78234792e-01 3.04033458e-01
4.16017264e-01 -1.18519557e+00 -2.10769922e-01 3.76674831e-01
7.82077730e-01 -1.37003493e+00 -3.14387120e-02 -4.37291682e-01
-4.58796322e-01 8.24030638e-01 8.87347043e-01 -8.80917907e-02
1.75595209e-01 -2.34877661e-01 -8.31933506e-03 2.81737596e-01
6.10344224e-02 -3.52038264e-01 6.45422161e-01 1.04706252e+00
2.81560808e-01 1.81671411e-01 -4.03050780e-01 1.12542450e+00
-2.03780815e-01 -1.06905982e-01 -7.95534253e-02 5.09142399e-01
-1.84436813e-01 -1.28333521e+00 -6.22154713e-01 3.89652736e-02
-1.04040943e-01 1.30554974e-01 -4.36793983e-01 6.49981201e-01
3.87723833e-01 9.86467838e-01 -1.21452101e-01 -4.63495791e-01
7.61212230e-01 2.02121921e-02 5.34621358e-01 -2.86216974e-01
-5.15300512e-01 -1.24695785e-01 -4.70476300e-02 -3.76027972e-01
-2.23906308e-01 -6.43746912e-01 -8.38730156e-01 -4.13219154e-01
-5.95335029e-02 -1.57830179e-01 4.12850171e-01 8.88806760e-01
4.12010193e-01 4.58886832e-01 5.96455038e-01 -7.69476950e-01
-7.39764690e-01 -9.02294576e-01 -6.10447645e-01 9.98323858e-01
2.70379484e-01 -8.75777006e-01 -2.14931488e-01 -1.21926673e-01] | [14.737116813659668, 1.0265278816223145] |
2d8175a6-b17e-42e9-8c5b-0fe848fd3a62 | eventpoint-self-supervised-local-descriptor | 2109.00210 | null | https://arxiv.org/abs/2109.00210v3 | https://arxiv.org/pdf/2109.00210v3.pdf | EventPoint: Self-Supervised Interest Point Detection and Description for Event-based Camera | This paper proposes a self-supervised learned local detector and descriptor, called EventPoint, for event stream/camera tracking and registration. Event-based cameras have grown in popularity because of their biological inspiration and low power consumption. Despite this, applying local features directly to the event stream is difficult due to its peculiar data structure. We propose a new time-surface-like event stream representation method called Tencode. The event stream data processed by Tencode can obtain the pixel-level positioning of interest points while also simultaneously extracting descriptors through a neural network. Instead of using costly and unreliable manual annotation, our network leverages the prior knowledge of local feature extraction on color images and conducts self-supervised learning via homographic and spatio-temporal adaptation. To the best of our knowledge, our proposed method is the first research on event-based local features learning using a deep neural network. We provide comprehensive experiments of feature point detection and matching, and three public datasets are used for evaluation (i.e. DSEC, N-Caltech101, and HVGA ATIS Corner Dataset). The experimental findings demonstrate that our method outperforms SOTA in terms of feature point detection and description. | ['Songzhi Su', 'Song Li', 'Cheng Zhao', 'Li Sun', 'Ze Huang'] | 2021-09-01 | null | null | null | null | ['interest-point-detection'] | ['computer-vision'] | [ 1.03696361e-01 -7.33650923e-01 -2.82171398e-01 -3.76078397e-01
-9.54552650e-01 -3.69863451e-01 6.74185991e-01 3.04847986e-01
-6.39982998e-01 3.27200502e-01 -1.09349675e-01 5.51221907e-01
-4.16164333e-03 -6.84971631e-01 -8.67691755e-01 -7.15792954e-01
-2.90560931e-01 -2.18607802e-02 6.35103643e-01 2.04294488e-01
2.94550210e-01 9.13829327e-01 -1.54263461e+00 1.05373941e-01
2.53922880e-01 1.53206158e+00 -5.30428253e-03 5.77892900e-01
3.23762029e-01 8.07801187e-01 -5.06310523e-01 4.00757492e-02
3.96830380e-01 -3.42542320e-01 -1.81093991e-01 -1.19597189e-01
7.52128303e-01 -5.87674975e-01 -8.21123064e-01 7.08884537e-01
6.60614550e-01 4.22470160e-02 5.17410755e-01 -1.65709651e+00
-3.09952557e-01 4.73670252e-02 -6.26587272e-01 4.54742402e-01
4.42000896e-01 1.90760329e-01 9.05831218e-01 -9.96133387e-01
8.01911712e-01 7.76926994e-01 1.05611670e+00 1.75648451e-01
-8.88934493e-01 -1.11377668e+00 3.27495858e-02 4.64935362e-01
-1.85659325e+00 -6.19827271e-01 9.34514523e-01 -1.82864681e-01
9.62939620e-01 -2.68879175e-01 8.36658597e-01 1.18766308e+00
4.15575147e-01 8.56073141e-01 5.97716153e-01 -5.16743511e-02
3.69044095e-01 -4.83691871e-01 -2.10350513e-01 7.68203378e-01
5.75727485e-02 3.22816253e-01 -1.21077859e+00 -3.16413939e-01
1.00937772e+00 4.99384403e-01 -8.24626312e-02 -5.56077182e-01
-1.54343390e+00 5.23033679e-01 7.40829289e-01 1.31500047e-02
-4.68861371e-01 6.76633418e-01 5.67847133e-01 9.49085429e-02
2.84348041e-01 -1.65554255e-01 -3.79170775e-01 -3.71670067e-01
-1.03968716e+00 2.33683754e-02 5.13772130e-01 1.16025019e+00
7.81781733e-01 2.28915047e-02 3.64990868e-02 3.56273800e-01
2.33298570e-01 9.11501944e-01 5.22454500e-01 -7.76246130e-01
1.48811325e-01 6.17971480e-01 -4.16853204e-02 -1.35958755e+00
-5.20458698e-01 -5.71293384e-02 -6.42415285e-01 6.37284666e-02
4.03622419e-01 -6.56630918e-02 -6.41543627e-01 1.41069710e+00
2.81034142e-01 9.62771833e-01 -2.32325315e-01 9.34786856e-01
8.10560048e-01 6.70165718e-01 1.95729092e-01 -1.02369808e-01
1.21467173e+00 -5.06662548e-01 -5.00251889e-01 -2.86855357e-04
3.32496136e-01 -4.43680286e-01 5.12101829e-01 2.33730406e-01
-6.64443076e-01 -4.45118427e-01 -1.19737029e+00 -6.06404245e-02
-4.40240413e-01 1.45102784e-01 7.84382939e-01 8.50457177e-02
-5.69440663e-01 5.76402068e-01 -1.33620012e+00 -7.14394212e-01
7.36424148e-01 3.65732223e-01 -4.39725220e-01 1.84411839e-01
-9.11793232e-01 3.21474582e-01 1.06452271e-01 -8.65162164e-02
-8.95828068e-01 -6.47251487e-01 -9.39589381e-01 -7.54118264e-02
1.01933971e-01 -4.48959529e-01 1.13127208e+00 -7.34026074e-01
-1.28245735e+00 7.99004972e-01 -3.35543334e-01 -6.71228528e-01
2.40808129e-01 -1.92608178e-01 -4.13183898e-01 5.59475303e-01
2.71040916e-01 7.69374907e-01 9.14082229e-01 -8.04400504e-01
-1.07152498e+00 -3.66930276e-01 -2.05863148e-01 -7.19986334e-02
-4.35431421e-01 1.93245545e-01 -5.80254793e-01 -6.61617517e-01
2.86287457e-01 -7.47955799e-01 3.37736905e-02 7.14957356e-01
-6.42709658e-02 -3.14812303e-01 1.04930103e+00 -4.17824350e-02
9.30637062e-01 -2.25246811e+00 -4.51298505e-01 2.37907156e-01
2.25708574e-01 -1.08075291e-01 4.80744168e-02 3.54484051e-01
1.61977425e-01 -4.27416235e-01 1.68309100e-02 -3.26070905e-01
-1.67356029e-01 -3.39038335e-02 -4.71632451e-01 1.05671799e+00
2.11577639e-01 1.02404511e+00 -1.07690895e+00 -8.49082112e-01
5.07668316e-01 7.51741827e-01 -2.08409682e-01 -1.04511753e-01
1.91882834e-01 1.51383892e-01 -6.12156987e-01 1.05740201e+00
4.90564525e-01 -2.49224499e-01 -4.01955247e-01 -5.20594418e-01
-3.16099226e-01 -6.14552684e-02 -1.23784316e+00 2.15548587e+00
-2.93589383e-01 9.49886501e-01 -4.19200212e-01 -8.64390790e-01
1.00602102e+00 2.23378569e-01 1.02254510e+00 -8.44961703e-01
2.77815640e-01 1.11211583e-01 -6.16338134e-01 -2.28393197e-01
4.52949405e-01 3.37861270e-01 -2.27534607e-01 2.45892480e-01
2.71386147e-01 2.79821426e-01 -3.19482647e-02 2.11091265e-01
1.36175895e+00 5.01943588e-01 4.79594767e-01 8.35749432e-02
3.40380967e-01 5.78124411e-02 8.56311023e-01 8.25039804e-01
-5.70666254e-01 8.90061319e-01 6.25844225e-02 -6.29464090e-01
-7.23328412e-01 -1.14170229e+00 -4.00314450e-01 9.92397070e-01
5.49741805e-01 -6.90340817e-01 -3.34411353e-01 -5.48791826e-01
1.71068326e-01 -5.69059048e-04 -4.99076366e-01 -2.35010818e-01
-7.32525289e-01 -4.29620832e-01 8.77556324e-01 9.41012859e-01
5.72054565e-01 -1.01347160e+00 -1.21198308e+00 3.10592353e-01
7.65362754e-02 -1.32739043e+00 -6.84170365e-01 2.96072304e-01
-8.12699854e-01 -1.17050505e+00 -4.47928101e-01 -7.94484913e-01
5.62485814e-01 4.71045613e-01 8.35193217e-01 -2.35233650e-01
-5.90615034e-01 7.32188761e-01 -3.82403016e-01 -5.06221533e-01
5.10031998e-01 -1.44495025e-01 8.06061402e-02 2.35652700e-01
7.23289967e-01 -7.59429038e-01 -1.04483414e+00 4.20384377e-01
-6.95144892e-01 -3.80557835e-01 4.86770302e-01 6.55502677e-01
1.08882689e+00 -2.57173896e-01 4.26340759e-01 -3.98925900e-01
-2.39606723e-02 -4.45309043e-01 -8.71579111e-01 -4.59329486e-02
-1.25215545e-01 -2.27770478e-01 6.24981284e-01 -4.93292868e-01
-6.85322225e-01 7.17376828e-01 2.91221499e-01 -7.00449467e-01
-3.26702952e-01 1.80232331e-01 -1.19532617e-02 -4.43432152e-01
8.16561162e-01 3.57603163e-01 -3.52178305e-01 -1.15916699e-01
1.67005718e-01 4.70762044e-01 9.45774734e-01 -3.83125365e-01
9.97661650e-01 1.14812994e+00 1.31671593e-01 -8.25126946e-01
-5.70905268e-01 -8.45640421e-01 -8.21993172e-01 -3.52057248e-01
8.11209321e-01 -1.22933459e+00 -9.60422814e-01 6.78934693e-01
-1.21177697e+00 5.07375561e-02 -4.67424214e-01 5.42416334e-01
-7.12974250e-01 2.25115612e-01 -4.50971693e-01 -6.48102999e-01
-3.64991784e-01 -7.25372195e-01 1.61560726e+00 4.00782555e-01
-5.36437072e-02 -7.64518082e-01 1.64367929e-01 -4.02270734e-01
2.79010952e-01 7.97548950e-01 -2.15208635e-01 -4.27492112e-01
-9.49921727e-01 -7.32301712e-01 -3.58758539e-01 -2.01982468e-01
1.06312864e-01 3.37588750e-02 -1.18447244e+00 -1.96627095e-01
-1.60315469e-01 -3.23939502e-01 8.11832726e-01 4.71323490e-01
1.13501608e+00 2.59998381e-01 -7.58976042e-01 1.20545840e+00
1.50767100e+00 2.76117951e-01 3.23836058e-01 5.99430144e-01
7.19251215e-01 -7.91999325e-02 6.75209105e-01 9.88546073e-01
5.86269438e-01 6.69556379e-01 3.70744735e-01 -1.01080000e-01
-1.79989949e-01 -4.50770348e-01 4.73418742e-01 4.36001241e-01
2.08139103e-02 -5.42858839e-02 -9.53189909e-01 6.53476477e-01
-2.04369378e+00 -1.08522475e+00 1.48564726e-01 2.14182210e+00
5.21699965e-01 -1.74424183e-02 9.93757043e-03 -5.21545373e-02
6.88306749e-01 1.69855371e-01 -8.81930053e-01 3.99905920e-01
-3.80781800e-01 2.28214294e-01 8.65312874e-01 -1.88360229e-01
-1.61049497e+00 8.60278010e-01 5.42333984e+00 6.86018944e-01
-1.29187548e+00 -7.42763281e-02 2.06085250e-01 -1.81815952e-01
6.12253785e-01 -4.74488623e-02 -1.05421555e+00 4.70395446e-01
9.62284029e-01 -1.51015431e-01 1.72780633e-01 8.46407473e-01
1.48312792e-01 -1.47190481e-01 -1.37273991e+00 1.57764673e+00
2.27130979e-01 -1.44310880e+00 -2.85268337e-01 -1.01192199e-01
4.52661067e-01 4.22333986e-01 -2.37069398e-01 -4.70407829e-02
-2.90574301e-02 -5.43882549e-01 8.70454729e-01 6.51666105e-01
7.89394736e-01 -9.14683878e-01 5.45576215e-01 3.29741724e-02
-1.86385286e+00 -4.99956310e-02 -3.71590883e-01 1.38989300e-01
2.00118959e-01 4.06056136e-01 -4.36421841e-01 3.98731977e-01
1.04440105e+00 1.56599641e+00 -6.09714568e-01 1.39894819e+00
3.50400619e-02 6.06805742e-01 -8.42090547e-01 2.09398016e-01
5.18206060e-02 2.54734486e-01 4.76734847e-01 1.23730516e+00
4.83146101e-01 1.99869145e-02 4.08412099e-01 7.31062710e-01
-1.40033692e-01 -6.80853799e-02 -8.02114308e-01 2.61401325e-01
7.50704467e-01 1.41139269e+00 -9.53088760e-01 -1.85860872e-01
-6.74177647e-01 1.03025651e+00 9.76658836e-02 1.49552166e-01
-1.02357686e+00 -8.90125811e-01 6.19298577e-01 7.12956721e-03
4.73680049e-01 -4.56152529e-01 -6.55969465e-03 -1.25767171e+00
8.58211815e-02 -3.39388430e-01 5.08634746e-01 -6.55923903e-01
-1.48486197e+00 4.14526016e-01 3.37462761e-02 -1.79715657e+00
-8.07793811e-02 -3.98342907e-01 -7.43883133e-01 2.29990944e-01
-1.65874708e+00 -1.43434870e+00 -7.53234982e-01 1.04413366e+00
3.09532344e-01 -1.90230012e-01 4.72465396e-01 4.78491813e-01
-4.29584414e-01 6.99450314e-01 1.96333453e-01 6.04229331e-01
1.03038728e+00 -9.43056941e-01 4.46671307e-01 7.81562090e-01
2.60272473e-01 4.01840568e-01 2.48384655e-01 -5.61736465e-01
-1.89921057e+00 -1.24224877e+00 6.10446811e-01 -4.65608686e-01
9.14410830e-01 -5.42685568e-01 -5.78511715e-01 6.13172054e-01
-1.37866393e-01 7.87926793e-01 4.00218368e-01 -3.84432852e-01
-3.16585511e-01 -6.01136804e-01 -9.89888787e-01 3.08676124e-01
1.37091887e+00 -7.81310380e-01 -3.42308879e-01 4.15995240e-01
4.70315039e-01 -4.94733274e-01 -8.94295752e-01 1.82031363e-01
6.02235317e-01 -6.30237222e-01 1.18481338e+00 -1.63225865e-03
1.61230713e-02 -6.81912243e-01 -2.06209943e-01 -7.17365801e-01
-1.79402739e-01 -5.76448739e-01 -2.57665217e-01 1.33844399e+00
-3.14611882e-01 -5.02123475e-01 8.99457216e-01 2.80653954e-01
-1.14156576e-02 -4.25763100e-01 -1.24685287e+00 -9.96833503e-01
-4.62038994e-01 -5.71651936e-01 5.56539953e-01 7.63438702e-01
-1.53077036e-01 -8.66656601e-02 -2.73637384e-01 4.64607298e-01
9.43083405e-01 2.54379541e-01 8.14073682e-01 -1.17161918e+00
2.01192677e-01 -9.91153717e-02 -1.00743198e+00 -1.17866933e+00
9.62126628e-02 -8.20225537e-01 2.71269143e-01 -1.07740617e+00
1.32177882e-02 -2.82443136e-01 -5.00535369e-01 5.68062127e-01
2.60480255e-01 6.14821255e-01 5.29763438e-02 4.02547449e-01
-1.15196955e+00 6.47167325e-01 6.57258630e-01 -3.73592563e-02
-1.97500899e-01 -4.25029278e-01 -7.19700679e-02 8.69463980e-01
7.32253790e-01 -8.04015994e-01 -5.13107330e-02 -3.55483830e-01
3.44248153e-02 -6.12971559e-02 8.30903530e-01 -1.52668405e+00
9.27420199e-01 -2.77330969e-02 7.67282844e-01 -8.99813712e-01
3.88080955e-01 -1.09954298e+00 -8.29476714e-02 2.16366932e-01
-1.49333268e-01 2.29288697e-01 1.78323731e-01 9.44122374e-01
-3.09407651e-01 1.45377353e-01 6.33533120e-01 1.36998489e-01
-1.10586166e+00 7.17014015e-01 -3.21516097e-01 7.17489375e-03
1.20917380e+00 -5.47287524e-01 -2.80955166e-01 -2.57739544e-01
-1.09544635e-01 1.70079798e-01 5.17112255e-01 3.76799792e-01
9.05641317e-01 -1.65819454e+00 -4.44491148e-01 2.27509126e-01
5.96364021e-01 8.19111913e-02 4.40328494e-02 8.40819359e-01
-5.48209786e-01 2.48999044e-01 -2.87365466e-01 -1.07332289e+00
-9.13253546e-01 2.96440840e-01 2.60615587e-01 3.75008732e-01
-1.06909406e+00 5.62169194e-01 9.15590301e-02 1.02275945e-01
4.51301455e-01 -4.66296166e-01 3.00000012e-02 1.14825435e-01
5.69571495e-01 3.47117692e-01 -6.67815879e-02 -8.52535605e-01
-7.47961164e-01 9.71990705e-01 4.14947867e-02 1.96653396e-01
1.44579077e+00 -6.49765432e-02 1.97771445e-01 3.81223530e-01
1.40998960e+00 -2.67169297e-01 -1.59588444e+00 -5.40260851e-01
7.01332912e-02 -5.74740887e-01 2.72816181e-01 -2.20443204e-01
-1.26477969e+00 6.89857125e-01 9.58015144e-01 -3.40199202e-01
1.22060204e+00 -2.01044772e-02 1.21678472e+00 7.56166101e-01
5.23741126e-01 -1.06540704e+00 1.60491288e-01 4.48709220e-01
5.96148074e-01 -1.29972577e+00 1.11235909e-01 3.86168920e-02
-4.29383725e-01 1.23322761e+00 5.71120679e-01 -5.66624284e-01
7.34814286e-01 4.10866529e-01 -3.95566486e-02 -2.54054070e-01
-6.52357221e-01 -1.86992407e-01 6.69211149e-02 7.90805578e-01
3.04583535e-02 -3.95243645e-01 2.00753793e-01 4.17065531e-01
1.75719768e-01 2.56677598e-01 2.47178167e-01 1.35152376e+00
-3.91437262e-01 -6.52407289e-01 -3.81759882e-01 3.20576966e-01
-2.30703294e-01 1.14382975e-01 -1.32790595e-01 8.53514850e-01
2.63513505e-01 6.86384201e-01 3.75738949e-01 -2.73958951e-01
3.93766254e-01 -1.40779540e-01 2.48121098e-01 -1.40985891e-01
-5.10026097e-01 6.21301010e-02 -4.09434229e-01 -9.90098834e-01
-7.03726113e-01 -9.32038248e-01 -1.64817154e+00 -1.28378436e-01
-2.57802814e-01 -2.47798696e-01 6.91949666e-01 6.80025220e-01
5.75833499e-01 3.13047141e-01 6.51039064e-01 -1.31074345e+00
-6.50400817e-02 -3.89539927e-01 -4.43483412e-01 5.27803063e-01
4.84579533e-01 -8.28849196e-01 -2.50471681e-01 4.37663943e-01] | [8.40736198425293, -1.2040108442306519] |
67ed08c4-ea57-4b4f-9744-29fef3ed5417 | learning-to-reason-for-text-generation-from | 2104.08296 | null | https://arxiv.org/abs/2104.08296v1 | https://arxiv.org/pdf/2104.08296v1.pdf | Learning to Reason for Text Generation from Scientific Tables | In this paper, we introduce SciGen, a new challenge dataset for the task of reasoning-aware data-to-text generation consisting of tables from scientific articles and their corresponding descriptions. Describing scientific tables goes beyond the surface realization of the table content and requires reasoning over table values. The unique properties of SciGen are that (1) tables mostly contain numerical values, and (2) the corresponding descriptions require arithmetic reasoning. SciGen is therefore the first dataset that assesses the arithmetic reasoning capabilities of generation models on complex input structures, i.e., tables from scientific articles. We study the effectiveness of state-of-the-art data-to-text generation models on SciGen and evaluate the results using common metrics as well as human evaluation. Our results and analyses show that (a) while humans like to reason for describing scientific tables, the ability of state-of-the-art models is severely limited on this task, (b) while adding more training data improves the results, it is not the solution for reasoning-aware text generation, and (c) one of the main bottlenecks for this task is the lack of proper automatic evaluation metrics. The data, code, and annotations for human evaluation will be available at https://github.com/UKPLab/SciGen. SciGen opens new avenues for future research in reasoning-aware text generation and evaluation. | ['Iryna Gurevych', 'Dan Roth', 'Andreas Rücklé', 'Nafise Sadat Moosavi'] | 2021-04-16 | null | https://openreview.net/forum?id=_QCy6HJJ4wd | https://openreview.net/pdf?id=_QCy6HJJ4wd | null | ['data-to-text-generation', 'arithmetic-reasoning'] | ['natural-language-processing', 'reasoning'] | [-3.50803882e-02 4.45932925e-01 3.72595601e-02 -2.09087536e-01
-1.02318394e+00 -1.00859284e+00 9.16712821e-01 6.81515872e-01
3.85263152e-02 9.41157401e-01 4.64668691e-01 -4.80079323e-01
2.55953278e-02 -1.16852570e+00 -9.16075468e-01 9.08200964e-02
3.74497473e-01 8.88043821e-01 4.76298388e-03 -4.96812463e-01
4.02635485e-01 2.52327710e-01 -1.54629946e+00 7.27521658e-01
1.14354110e+00 8.90516043e-01 -3.23712938e-02 8.25035691e-01
-6.51157200e-01 9.54429328e-01 -9.46813226e-01 -8.69465172e-01
7.23431110e-02 -6.04058683e-01 -1.04183447e+00 -3.78294140e-01
5.99197090e-01 -1.59810185e-01 -3.61911729e-02 8.40879738e-01
4.33821738e-01 -3.15338373e-02 7.64583290e-01 -1.37380517e+00
-8.78966212e-01 1.21485734e+00 6.34087622e-02 -3.37517299e-02
8.22563112e-01 2.42256090e-01 1.10832727e+00 -8.39598060e-01
9.98174727e-01 1.34496164e+00 5.55201292e-01 5.93400359e-01
-1.10379887e+00 -3.70238721e-01 -5.22717908e-02 -4.21000086e-02
-1.21052027e+00 -4.74551827e-01 3.56955320e-01 -4.94049937e-01
1.05572319e+00 5.78142524e-01 5.79631507e-01 1.05714476e+00
1.00006340e-02 7.24914730e-01 8.90919209e-01 -3.54054779e-01
3.43341142e-01 2.14049324e-01 6.62649050e-02 6.10753179e-01
6.68550789e-01 -4.34027970e-01 -5.99591196e-01 -9.15834866e-03
5.74671865e-01 -5.64967096e-01 -9.17629600e-02 -2.17431430e-02
-1.65168440e+00 6.30600870e-01 1.88141748e-01 1.93129852e-01
-1.76306605e-01 8.09534788e-02 5.92192054e-01 1.16740629e-01
1.84999064e-01 1.21885574e+00 -5.11565626e-01 -4.56983984e-01
-7.96571672e-01 9.62191522e-01 1.34722447e+00 1.43426216e+00
2.51242846e-01 -1.25093505e-01 -6.43187106e-01 4.60553974e-01
-1.85941651e-01 5.48253000e-01 1.93058789e-01 -1.04608822e+00
8.65210056e-01 8.94720614e-01 3.01462114e-01 -8.96047235e-01
-3.73883635e-01 -2.60981858e-01 -6.49358809e-01 1.35087013e-01
8.46364558e-01 -2.33850315e-01 -5.91816366e-01 1.58912444e+00
5.69867417e-02 -6.03212118e-01 2.84985751e-01 6.08203411e-01
1.43288648e+00 8.41469884e-01 7.59227127e-02 3.07328906e-02
1.58477390e+00 -7.06451654e-01 -8.13619554e-01 -8.79883841e-02
7.95447886e-01 -9.23187912e-01 1.34009969e+00 4.34713781e-01
-1.61007428e+00 -5.32929480e-01 -9.25165951e-01 -4.81517315e-01
-9.84923244e-01 1.51356444e-01 8.49721909e-01 4.14305001e-01
-1.08068740e+00 4.82890069e-01 -5.18154860e-01 -3.77147138e-01
2.52404749e-01 -1.18386053e-01 -2.61314005e-01 -6.15015700e-02
-1.34341443e+00 9.68755722e-01 4.75615382e-01 -2.99972683e-01
-4.45240855e-01 -1.14391398e+00 -1.02462161e+00 2.93332309e-01
5.25399089e-01 -9.53177452e-01 1.37361491e+00 -1.41932189e-01
-1.03183222e+00 7.41357923e-01 -6.90580159e-02 -4.14011419e-01
8.15326333e-01 -8.47369432e-02 -1.45092845e-01 -1.92320310e-02
3.04265350e-01 8.23109686e-01 -1.20775113e-02 -1.33845413e+00
-3.74995768e-01 3.04048089e-03 3.28057140e-01 1.14810757e-01
-2.38165483e-02 -2.70171762e-01 -4.07528073e-01 -7.94841170e-01
-2.32784301e-01 -6.80287123e-01 2.91955844e-02 4.65547433e-03
-7.69091725e-01 -3.22894305e-01 3.22122633e-01 -5.97290993e-01
1.23250687e+00 -1.67911518e+00 -2.65392456e-02 -9.91590768e-02
1.33193001e-01 2.06972584e-02 1.38474703e-02 8.64487231e-01
2.34514754e-02 6.95852101e-01 -6.15608133e-02 -1.94825917e-01
3.53143662e-01 -2.03410357e-01 -5.89360058e-01 -3.86462301e-01
2.76417226e-01 1.19018543e+00 -9.81588781e-01 -6.09172583e-01
9.27372351e-02 1.61328718e-01 -5.14660358e-01 1.43744558e-01
-8.72883558e-01 1.08044758e-01 -2.76900083e-01 6.48406327e-01
5.26572347e-01 -3.88325304e-01 1.97189264e-02 -2.76018530e-01
-9.41404328e-02 7.98714578e-01 -1.26970732e+00 1.70423913e+00
-4.78034645e-01 4.75577384e-01 -4.72066462e-01 -3.49170297e-01
6.83231413e-01 2.32347861e-01 2.82139983e-02 -6.10682607e-01
-1.31469116e-01 2.18194500e-01 -8.81425589e-02 -3.45240057e-01
8.26231718e-01 1.59410015e-01 -4.59572703e-01 6.59447789e-01
-7.49542043e-02 -8.43620718e-01 1.16172040e+00 6.64899290e-01
1.10849714e+00 2.27403238e-01 3.50780070e-01 -2.67976612e-01
4.95003402e-01 4.46688592e-01 2.01635510e-01 9.03204918e-01
4.36522275e-01 6.68534696e-01 8.95431280e-01 -3.46154749e-01
-1.08524847e+00 -1.09079552e+00 -6.78962395e-02 8.83860707e-01
-1.54530317e-01 -1.00268912e+00 -8.10218513e-01 -5.07375240e-01
2.35505085e-02 1.34171212e+00 -7.18948066e-01 4.47702408e-02
-3.80032480e-01 -4.31405604e-01 6.87736392e-01 7.08340645e-01
2.47955263e-01 -1.15793729e+00 -4.56796020e-01 8.48575532e-02
-6.28911495e-01 -1.25042784e+00 -3.88511509e-01 -1.99580193e-02
-8.85034561e-01 -1.05455363e+00 -4.42681581e-01 -4.85233247e-01
7.72664011e-01 -1.41739875e-01 1.69606531e+00 1.79875910e-01
-1.16072908e-01 2.20186830e-01 -3.24432880e-01 -8.27906966e-01
-8.85478675e-01 2.15660468e-01 -4.20044899e-01 -8.62145543e-01
2.74717975e-02 -2.39029676e-01 -3.70704293e-01 7.67055228e-02
-9.33873653e-01 6.03685617e-01 5.62170029e-01 5.03882289e-01
5.79614758e-01 1.05619349e-01 3.74265581e-01 -1.17701995e+00
8.58180106e-01 -2.41919726e-01 -6.60774410e-01 3.65308613e-01
-4.97683465e-01 3.61573517e-01 9.80435133e-01 -6.74035922e-02
-1.00245881e+00 -4.32695508e-01 5.36846779e-02 1.89193249e-01
-6.98940232e-02 6.76325381e-01 -2.66547352e-01 5.26927352e-01
1.00598276e+00 1.45146951e-01 -2.23433211e-01 -3.30270052e-01
6.51896954e-01 2.15864986e-01 6.11520767e-01 -1.00872314e+00
8.12180519e-01 2.12404564e-01 1.53411552e-01 -3.52521122e-01
-8.38634133e-01 5.88966832e-02 -3.50916207e-01 2.82518025e-02
7.22835600e-01 -8.13795030e-01 -8.44983816e-01 1.13558866e-01
-1.40572667e+00 -4.73608464e-01 -5.14847219e-01 8.73558819e-02
-6.06120229e-01 1.24474891e-01 -7.91751027e-01 -6.57098055e-01
-6.24275446e-01 -1.03327775e+00 1.09300232e+00 8.22914019e-02
-8.13721895e-01 -1.01788950e+00 -2.59500355e-01 5.81646323e-01
4.46882606e-01 5.33710301e-01 1.45548546e+00 -7.24059999e-01
-7.76557088e-01 -2.69549131e-01 -2.44584844e-01 -9.74422786e-03
-7.21929967e-03 3.11743200e-01 -7.31555760e-01 3.26020628e-01
-4.17723149e-01 -5.90793908e-01 4.48785990e-01 6.56140000e-02
1.39977992e+00 -5.73577940e-01 -1.36258870e-01 3.26673597e-01
1.15355480e+00 1.00659050e-01 7.46554315e-01 2.30561614e-01
6.51340902e-01 6.07717335e-01 5.37393689e-01 5.43107927e-01
8.34799767e-01 4.93700981e-01 1.75858706e-01 6.28686547e-02
-3.41143548e-01 -5.33854544e-01 2.68190894e-02 6.03689790e-01
1.06860735e-01 -5.68489492e-01 -1.10667229e+00 5.13838112e-01
-1.67619812e+00 -9.92834806e-01 -3.61609608e-01 2.11018610e+00
1.31681621e+00 3.64817381e-01 -1.17137535e-02 2.72149622e-01
2.91102022e-01 -2.42811128e-01 -3.63847882e-01 -5.96623003e-01
-2.12255359e-01 3.25418413e-01 3.68727222e-02 3.76664698e-01
-8.14307868e-01 8.61152411e-01 6.46720934e+00 8.11507463e-01
-6.80728912e-01 -3.48245203e-01 6.98290348e-01 -6.43355399e-02
-6.90611601e-01 -1.83228031e-03 -8.76804709e-01 4.58425552e-01
8.16308022e-01 -5.51776052e-01 4.77590859e-01 8.07471275e-01
-1.73829608e-02 -6.00513145e-02 -1.54475272e+00 6.99248612e-01
4.29856544e-03 -1.61710584e+00 6.91429555e-01 -2.20197827e-01
5.91756284e-01 -6.63789332e-01 -1.64256305e-01 4.65333641e-01
5.23350537e-01 -1.31984913e+00 1.12704766e+00 6.09380960e-01
8.27105880e-01 -5.85453629e-01 7.03744531e-01 1.06468126e-01
-8.91333282e-01 3.49762619e-01 -2.51855820e-01 -1.14632837e-01
-4.13090326e-02 7.72555113e-01 -7.53890455e-01 6.25545561e-01
4.79776949e-01 5.06828427e-01 -1.01123393e+00 6.53619766e-01
-4.34479415e-01 2.87718266e-01 -1.64792106e-01 -3.49284589e-01
-1.58820212e-01 2.50409469e-02 2.70764530e-01 1.38866222e+00
4.94045854e-01 3.98789309e-02 -1.67615727e-01 1.41994965e+00
-4.67121929e-01 9.42114964e-02 -5.76223075e-01 -5.26399910e-01
5.20129859e-01 1.34731579e+00 -8.26967359e-01 -6.90383375e-01
-2.07967713e-01 5.45149326e-01 3.26298237e-01 1.35304332e-01
-7.04011202e-01 -6.41061604e-01 3.96905601e-01 2.87428617e-01
1.60819948e-01 -5.18035740e-02 -1.21681440e+00 -1.05457044e+00
2.48983219e-01 -1.10555816e+00 5.50426543e-01 -1.31054854e+00
-1.18228686e+00 4.95986283e-01 2.11215690e-01 -1.05154693e+00
-5.69447041e-01 -5.69335222e-01 -4.39530522e-01 8.63635957e-01
-9.72173154e-01 -1.05709839e+00 -5.78937769e-01 1.42776951e-01
4.22147959e-01 -3.17712314e-02 1.03327072e+00 -1.27928033e-01
-1.64301962e-01 6.00592256e-01 -3.17053646e-01 2.46793985e-01
6.20968223e-01 -1.65706980e+00 9.63537931e-01 7.39569426e-01
6.74545169e-02 8.87566507e-01 9.40330923e-01 -8.50394666e-01
-1.65886831e+00 -1.04804540e+00 1.21708345e+00 -9.17824745e-01
6.37165725e-01 -7.49292493e-01 -8.31345379e-01 5.37048340e-01
2.63743013e-01 -4.86657053e-01 5.84564328e-01 1.46337301e-01
-5.03368437e-01 4.49076183e-02 -1.01958406e+00 1.02642882e+00
1.12905836e+00 -2.02725500e-01 -5.67727566e-01 7.38932192e-01
7.76007414e-01 -9.59860146e-01 -1.03736424e+00 1.15161039e-01
4.40030545e-01 -7.63233483e-01 8.77284706e-01 -4.19861913e-01
1.28077006e+00 -3.79398763e-01 3.87001261e-02 -1.28774548e+00
-9.70781595e-02 -5.85854232e-01 -2.07132339e-01 1.34764504e+00
9.22498465e-01 -2.84474224e-01 5.95727921e-01 8.40704560e-01
-1.52570724e-01 -7.40963995e-01 -3.29998761e-01 -7.22263098e-01
3.36229265e-01 -5.16397834e-01 8.55357707e-01 8.42260361e-01
2.74447978e-01 4.71294135e-01 2.38670409e-01 -3.14730704e-01
3.32653224e-01 3.19830537e-01 9.69557643e-01 -9.91476119e-01
-1.92534655e-01 -6.54194176e-01 -7.20883459e-02 -5.43473125e-01
-6.56964332e-02 -1.15741348e+00 -8.27701017e-02 -2.38743806e+00
1.16552316e-01 -3.50938082e-01 3.87850195e-01 6.44398749e-01
-3.90372008e-01 -1.41492546e-01 5.28915346e-01 6.17868733e-03
-4.93108302e-01 2.68819571e-01 1.41216266e+00 -9.65949073e-02
4.89632450e-02 -2.71817297e-01 -1.27451241e+00 2.70593494e-01
8.24246824e-01 -2.13549659e-01 -6.01277709e-01 -3.25739354e-01
7.44439423e-01 2.65746266e-01 2.57841974e-01 -1.18115509e+00
2.17907980e-01 -4.04863805e-01 7.16602802e-01 -6.64101958e-01
1.50277615e-01 -2.64903069e-01 1.29441679e-01 3.30948532e-01
-6.19767070e-01 4.17847425e-01 6.23913527e-01 -7.84546137e-02
5.32819256e-02 -9.29084793e-02 4.09813464e-01 -4.66390252e-01
-2.43727893e-01 -1.23549208e-01 -4.28137749e-01 7.00151443e-01
5.61043084e-01 1.39629185e-01 -9.09557283e-01 -5.64468980e-01
-1.10482596e-01 2.58512378e-01 7.00225949e-01 4.38397616e-01
4.03347969e-01 -1.20807266e+00 -8.33431363e-01 -1.34841859e-01
3.34044933e-01 3.48263204e-01 -8.18845034e-02 3.15858275e-01
-8.72084498e-01 6.28573656e-01 -2.64845818e-01 -1.70895368e-01
-9.04373288e-01 5.43126941e-01 1.28165856e-01 -5.94643295e-01
-4.47573870e-01 6.31078362e-01 2.05513343e-01 -5.98300993e-01
8.98558944e-02 -7.91574597e-01 -1.11884013e-01 -3.28790732e-02
7.12954342e-01 3.69835109e-01 3.25678915e-01 -2.27659824e-03
-2.17534348e-01 1.03871576e-01 -1.60455570e-01 -3.30582559e-01
1.06804991e+00 4.17739391e-01 -2.06623986e-01 5.03840566e-01
5.98301232e-01 1.74189374e-01 -6.03224039e-01 2.10134163e-01
-1.59127310e-01 -1.80044100e-01 -4.20750886e-01 -1.47778809e+00
-6.49739623e-01 7.28460848e-01 -2.91189253e-01 2.74439424e-01
7.93897092e-01 -1.08982787e-01 7.65108705e-01 5.15867770e-01
1.59901023e-01 -9.80015397e-01 -1.36211223e-03 6.67533994e-01
1.35134709e+00 -1.05961418e+00 2.80121863e-01 -6.63386703e-01
-8.24703574e-01 1.22654128e+00 7.55931497e-01 3.51536512e-01
1.31562725e-01 5.25897622e-01 1.10678777e-01 -3.01384032e-01
-1.14862251e+00 8.25880468e-03 3.53147805e-01 5.31700790e-01
9.09069419e-01 -3.50619406e-02 -3.08187783e-01 7.96017706e-01
-1.10423148e+00 2.16371059e-01 8.18085909e-01 1.01193750e+00
7.39768893e-02 -9.75951612e-01 -4.20459360e-01 7.41578460e-01
-3.57872695e-01 -3.79701197e-01 -7.62983918e-01 8.93492818e-01
-1.24992453e-01 1.13098812e+00 -5.21211289e-02 -1.16653271e-01
5.70572734e-01 1.87824130e-01 5.05983472e-01 -7.41186142e-01
-7.34420717e-01 -6.38285160e-01 4.81179744e-01 -3.28770280e-01
3.13191921e-01 -5.53723872e-01 -1.49756861e+00 -7.66845167e-01
2.98293322e-01 2.97285825e-01 7.90529609e-01 6.47070467e-01
5.31349063e-01 7.96782792e-01 -1.03226900e-01 -4.35620576e-01
-4.63300735e-01 -1.02413893e+00 -1.46515161e-01 7.02348113e-01
-8.33997726e-02 -3.63088131e-01 -2.57100224e-01 2.26735771e-01] | [11.402009010314941, 8.76620101928711] |
21b7eaf5-739d-4d12-b479-a3325b59837c | distributional-reinforcement-learning-with-5 | null | null | https://openreview.net/forum?id=C8Ltz08PtBp | https://openreview.net/pdf?id=C8Ltz08PtBp | Distributional Reinforcement Learning with Monotonic Splines | Distributional Reinforcement Learning (RL) differs from traditional RL by estimating the distribution over returns to capture the intrinsic uncertainty of MDPs. One key challenge in distributional RL lies in how to parameterize the quantile function when minimizing the Wasserstein metric of temporal differences. Existing algorithms use a step functions or piecewise linear functions. In this paper, we propose to learn smooth continuous quantile functions represented by monotonic rational-quadratic splines, which also naturally solve the quantile crossing problem. Experiments in stochastic environments show that a dense estimation for quantile functions enhances distributional RL in terms of faster empirical convergence and higher rewards in most cases. | ['Pascal Poupart', 'Oliver Schulte', 'Haonan Duan', 'Guiliang Liu', 'Yudong Luo'] | 2021-09-29 | null | null | null | iclr-2022-4 | ['distributional-reinforcement-learning'] | ['methodology'] | [-6.48896039e-01 3.32509726e-02 -5.00760913e-01 -5.37882924e-01
-1.57789695e+00 -6.31430805e-01 2.87718147e-01 2.43163690e-01
-7.62641668e-01 1.32600260e+00 3.22559893e-01 -1.98704392e-01
-4.48360860e-01 -7.96478748e-01 -8.92267883e-01 -7.28448272e-01
-6.07844889e-01 6.61746681e-01 -1.23946264e-01 -1.04748242e-01
2.97596455e-01 3.43074888e-01 -1.03179467e+00 -9.41921026e-02
1.10074866e+00 1.06187034e+00 -2.93689519e-02 7.01965928e-01
-1.37525618e-01 7.75821388e-01 -7.67422080e-01 -3.70740026e-01
4.09088612e-01 -2.37988248e-01 -3.04767400e-01 -6.76644325e-01
5.78207001e-02 -7.41246998e-01 -1.25134602e-01 9.95564878e-01
5.76742351e-01 6.53690100e-01 9.53720450e-01 -1.32467473e+00
-7.45141029e-01 8.90152991e-01 -7.49165297e-01 1.26731753e-01
6.47549853e-02 1.97982341e-01 1.25553930e+00 -4.59656984e-01
8.93159807e-02 1.60261750e+00 6.33480608e-01 4.52219754e-01
-1.61964273e+00 -4.86720473e-01 1.98332950e-01 -1.97566614e-01
-1.09307683e+00 -3.32944207e-02 3.56641859e-01 -3.47209901e-01
5.79865515e-01 -2.01413870e-01 3.22405785e-01 1.02900541e+00
3.70516181e-01 9.51390028e-01 1.13980877e+00 3.63166593e-02
6.78055584e-01 -5.15889823e-02 -9.82525125e-02 2.22378030e-01
2.27864921e-01 5.47937572e-01 -1.74996495e-01 -5.02181113e-01
1.08919048e+00 1.77335978e-01 1.66830286e-01 -4.83798504e-01
-7.88051009e-01 1.34267366e+00 3.00144166e-01 -5.02436399e-01
-7.88393915e-01 9.54680502e-01 5.52946091e-01 3.81190598e-01
6.56399190e-01 3.75106931e-01 -4.03722644e-01 -6.11812651e-01
-8.13991487e-01 7.79100001e-01 7.59902179e-01 8.84938538e-01
6.47988319e-01 3.45945060e-01 -7.61924386e-01 5.79026341e-01
3.83594155e-01 9.38190043e-01 7.01976866e-02 -1.35717559e+00
7.59095013e-01 -1.55172065e-01 8.62208843e-01 -3.73815238e-01
-1.32203519e-01 -2.18817636e-01 -3.02604616e-01 5.03404379e-01
8.70669067e-01 -7.14809835e-01 -6.15988851e-01 1.86641097e+00
2.57991582e-01 2.48984188e-01 1.02960259e-01 7.93709099e-01
-3.67232263e-02 6.95187747e-01 3.03564489e-01 -1.91315815e-01
7.65595496e-01 -4.98899966e-01 -6.81817591e-01 2.66324669e-01
1.09250158e-01 -2.28530094e-01 1.56690788e+00 4.36055899e-01
-1.20331645e+00 -2.46380389e-01 -6.06752872e-01 1.92650780e-01
3.77978161e-02 -4.26161028e-02 4.32291418e-01 5.66888094e-01
-8.88419271e-01 9.47488725e-01 -8.42477143e-01 5.55310190e-01
5.73099673e-01 2.79079974e-01 4.10793245e-01 1.22664392e-01
-1.31531715e+00 7.72680044e-01 4.88433428e-02 -2.84516394e-01
-1.28102422e+00 -1.04468179e+00 -7.77121782e-01 2.33290687e-01
4.41176951e-01 -2.58983582e-01 1.65411890e+00 -6.28863275e-01
-1.98116481e+00 5.69493584e-02 3.03935528e-01 -8.75425577e-01
9.56871390e-01 -5.46788752e-01 1.29969120e-01 -8.40937812e-03
6.97641224e-02 4.56151783e-01 9.72366452e-01 -9.37567294e-01
-5.88849545e-01 -2.57665277e-01 1.08624376e-01 2.19549283e-01
9.34385359e-02 -4.60935503e-01 3.27410340e-01 -5.73434770e-01
-8.62187266e-01 -5.19294977e-01 -5.38915992e-01 -2.11304858e-01
-9.12022144e-02 -5.75502872e-01 1.88100800e-01 -5.34943342e-01
1.05958557e+00 -2.02337122e+00 -1.41889855e-01 3.14186424e-01
-5.60355186e-02 -4.31483477e-01 -2.59487122e-01 6.03615046e-01
4.24539924e-01 -3.03656757e-02 -1.29471660e-01 -2.43267655e-01
6.44785106e-01 3.05366069e-01 -8.99107158e-01 6.73485875e-01
1.44924179e-01 9.21140254e-01 -1.14972711e+00 -7.64824972e-02
-1.26606211e-01 3.30294043e-01 -6.16272688e-01 3.32314223e-01
-6.48690581e-01 3.61930132e-01 -6.62280381e-01 3.69609505e-01
6.28276110e-01 3.32995594e-01 -2.02984765e-01 4.66412067e-01
-1.78090081e-01 2.14332625e-01 -1.15009928e+00 1.44596565e+00
-5.11340737e-01 2.31689543e-01 1.21816173e-02 -9.31556463e-01
1.17151833e+00 9.52507555e-02 6.37799203e-01 -6.13313556e-01
-6.23219088e-03 1.73370674e-01 -3.98983538e-01 -3.05208594e-01
6.20489597e-01 -4.71600711e-01 -4.27498698e-01 4.93633151e-01
-2.65899897e-01 -3.40989470e-01 -1.61343981e-02 -1.41346201e-01
6.52815640e-01 5.32774210e-01 3.21356468e-02 -5.26230216e-01
-1.93709135e-01 -1.77573308e-01 5.64137757e-01 8.68437171e-01
-3.29297900e-01 1.42856821e-01 1.20606411e+00 -2.22796336e-01
-1.07635105e+00 -1.71976423e+00 -1.09789215e-01 1.37949991e+00
-2.57170442e-02 5.81362620e-02 -6.09062493e-01 -7.68015325e-01
7.89081693e-01 1.22851789e+00 -7.43447244e-01 -1.14699185e-01
-4.33616877e-01 -3.73175561e-01 6.23365998e-01 8.40181589e-01
-2.47303285e-02 -9.43138838e-01 -8.06129754e-01 3.77517551e-01
2.14586884e-01 -3.89052689e-01 -9.25781965e-01 3.00460309e-01
-9.78348374e-01 -9.20450687e-01 -1.02273393e+00 -1.33361906e-01
2.78544784e-01 -2.76380867e-01 1.10132849e+00 -8.84614289e-01
-2.05904506e-02 7.69316137e-01 7.09899068e-02 -6.97001815e-01
6.43640608e-02 -2.95056462e-01 8.56263414e-02 -3.43525231e-01
9.66291130e-02 -3.57555866e-01 -6.76517189e-01 9.07343104e-02
-7.47616589e-01 -9.11911666e-01 2.54452765e-01 9.37375844e-01
9.37017202e-01 -2.06977695e-01 1.25852764e+00 -7.91442394e-01
1.44050133e+00 -8.30345333e-01 -1.11642206e+00 2.13390246e-01
-6.49852335e-01 6.54339194e-01 7.52985299e-01 -8.09616566e-01
-1.10653782e+00 -2.20884383e-01 1.53515473e-01 -4.35014546e-01
1.81785896e-01 4.36619163e-01 3.32635701e-01 5.75816810e-01
6.53223097e-01 -9.38136503e-03 1.99991971e-01 -3.00107002e-01
7.39340305e-01 2.46274173e-01 4.05997455e-01 -1.44510567e+00
4.56636935e-01 2.54501045e-01 -2.36418284e-02 -3.42543036e-01
-9.19387698e-01 -1.75436169e-01 -1.15080826e-01 -2.07613230e-01
5.65275192e-01 -7.83279657e-01 -1.33316314e+00 -3.14339429e-01
-7.47999370e-01 -9.73524332e-01 -1.16176212e+00 7.44119406e-01
-1.30049193e+00 4.44240272e-02 -6.10912561e-01 -1.39822948e+00
-3.29721063e-01 -8.80990624e-01 1.06346953e+00 3.31956118e-01
2.00814173e-01 -1.22266757e+00 6.20511413e-01 -3.92826319e-01
6.08734131e-01 5.94263434e-01 8.10588002e-01 -3.42742950e-01
-2.07504839e-01 1.49370462e-01 -3.01606357e-02 4.00325686e-01
-2.06548199e-01 -9.34908986e-02 -5.91014147e-01 -3.84464622e-01
-1.82812870e-01 -6.87861443e-01 8.83870304e-01 1.13042808e+00
1.25768626e+00 -3.83793324e-01 2.42546856e-01 5.38908005e-01
1.33008921e+00 3.26601416e-01 4.02301788e-01 2.03088194e-01
1.14203364e-01 4.63004827e-01 1.04413331e+00 1.13205945e+00
3.01101714e-01 6.11247979e-02 5.37886560e-01 3.68960619e-01
6.88549459e-01 -7.47925699e-01 8.41322362e-01 -6.12494573e-02
1.10716008e-01 2.26182610e-01 -7.69560039e-01 4.37330633e-01
-2.11587286e+00 -1.08387840e+00 1.72285229e-01 2.49130130e+00
1.16757596e+00 6.69570863e-02 6.89712763e-01 -6.00012898e-01
6.65111303e-01 -2.45848745e-01 -1.19917357e+00 -6.56348348e-01
2.79921353e-01 3.33972067e-01 8.69723499e-01 7.01828361e-01
-7.06282079e-01 8.39368999e-01 7.44913149e+00 9.92999077e-01
-7.44158268e-01 -1.76482555e-03 7.50027418e-01 -3.89034063e-01
-5.09504616e-01 -2.00544700e-01 -7.99036682e-01 5.92554569e-01
1.00706828e+00 -4.73545283e-01 7.71340311e-01 1.22106946e+00
5.13632059e-01 -3.23422551e-02 -1.09027708e+00 6.48745835e-01
-7.61844933e-01 -9.95516956e-01 -5.27597547e-01 9.26426277e-02
8.79286587e-01 4.94645163e-02 6.15946293e-01 9.38082814e-01
1.21133363e+00 -1.35910618e+00 8.43502522e-01 1.05286705e+00
9.57383454e-01 -1.47145855e+00 4.78107929e-01 4.70726490e-01
-9.33428407e-01 -2.55824834e-01 -1.03384113e+00 -3.86261493e-02
3.26688662e-02 5.04351676e-01 -7.77799189e-01 -8.64787921e-02
6.16299093e-01 4.21208620e-01 2.29031742e-01 1.16079891e+00
-4.04516727e-01 7.45242298e-01 -3.31269264e-01 -2.83889204e-01
6.04545474e-01 -5.95629632e-01 2.44118229e-01 1.15299928e+00
3.86704445e-01 -2.42251426e-01 4.09525543e-01 1.25880837e+00
-2.32203025e-02 5.02314344e-02 -5.68966806e-01 8.00149143e-03
6.17808700e-01 9.35235023e-01 -1.50356799e-01 -8.75929222e-02
-1.20276147e-02 3.69234204e-01 5.77287912e-01 7.21483707e-01
-1.18117177e+00 -5.65482140e-01 8.71603489e-01 -1.43353358e-01
4.13662404e-01 -4.58177060e-01 -7.59072527e-02 -7.15293050e-01
-2.66004503e-01 -4.04705882e-01 6.57221258e-01 -2.34895095e-01
-1.77478719e+00 1.37892425e-01 1.97399870e-01 -8.69284928e-01
-5.84965646e-01 -3.97340536e-01 -7.77170777e-01 9.36038196e-01
-1.55682755e+00 -5.99750102e-01 3.03913832e-01 8.44450772e-01
3.87372226e-01 1.95000395e-02 4.46056813e-01 -1.86526746e-01
-2.04902112e-01 5.62264502e-01 9.27625537e-01 -1.89542308e-01
9.18535471e-01 -1.93261194e+00 1.66983932e-01 6.73296824e-02
-1.53977200e-01 4.19147640e-01 7.95728087e-01 -6.26417816e-01
-1.31643772e+00 -1.00518918e+00 -1.03260212e-01 -2.65059233e-01
9.19424236e-01 -1.67675599e-01 -6.63835526e-01 6.32153869e-01
6.18812181e-02 2.38490459e-02 5.26458025e-01 1.02811053e-01
-4.07716125e-01 -2.79696196e-01 -1.59878135e+00 3.23146403e-01
3.38509917e-01 -3.79860610e-01 -3.32665622e-01 2.40358889e-01
9.14042234e-01 -3.17371547e-01 -1.22801936e+00 -9.60504040e-02
4.31534439e-01 -6.62810385e-01 1.01851809e+00 -1.07140982e+00
1.91433966e-01 2.16689985e-02 -1.18154272e-01 -1.79761052e+00
-4.67644855e-02 -1.01141918e+00 -3.36932153e-01 1.09544837e+00
1.49431080e-01 -8.13207209e-01 6.53604805e-01 6.27991855e-01
2.21657813e-01 -9.74854648e-01 -1.06584036e+00 -1.16608286e+00
6.89786017e-01 -5.31385660e-01 7.29543924e-01 4.21933681e-01
6.63390607e-02 -2.50986218e-01 -4.68388110e-01 -1.25978291e-01
1.13522768e+00 1.71880454e-01 5.91563821e-01 -9.06407237e-01
-4.57978994e-01 -5.70543110e-01 2.10440591e-01 -1.25465631e+00
4.41977322e-01 -5.72779834e-01 4.72217023e-01 -1.21248281e+00
-4.78829145e-02 -5.31753302e-01 -5.56188285e-01 2.06759676e-01
-1.59657523e-01 -5.03227055e-01 -8.43732655e-02 -1.08192630e-01
-8.57215703e-01 1.20136082e+00 1.26254344e+00 3.84244248e-02
-4.75567192e-01 4.45456803e-01 -6.59129083e-01 4.72382665e-01
1.03603292e+00 -6.62179768e-01 -7.21071601e-01 -1.56590328e-01
2.32770130e-01 7.16567934e-01 1.35137945e-01 -2.97641724e-01
1.02880582e-01 -8.85773778e-01 3.16981286e-01 -6.66691780e-01
8.31122175e-02 -3.84164631e-01 -4.28920835e-01 4.10519123e-01
-8.24357867e-01 2.86293596e-01 2.15235591e-01 1.10750866e+00
-4.56555523e-02 -2.19827443e-01 8.56056392e-01 1.32524848e-01
-8.55948478e-02 5.53110719e-01 -3.97154093e-01 7.18402505e-01
1.03103650e+00 2.83872366e-01 -1.66935444e-01 -8.35647345e-01
-5.94083726e-01 6.69602573e-01 5.56329228e-02 -9.78848189e-02
8.46413136e-01 -1.46875441e+00 -7.45198905e-01 -3.50960910e-01
-3.96546662e-01 7.94770941e-02 -9.77114514e-02 7.00603902e-01
-1.71581328e-01 3.95417400e-02 -2.02710569e-01 -2.95440823e-01
-3.38142633e-01 5.01914680e-01 4.29283857e-01 -5.98801672e-01
-3.86377424e-01 7.83374012e-01 -2.91263796e-02 -3.87650251e-01
6.91635489e-01 -6.53380096e-01 1.07967764e-01 9.37629566e-02
5.18403292e-01 7.63108969e-01 -7.02894092e-01 2.64315993e-01
5.85325658e-02 1.85895205e-01 2.02644035e-01 -5.40960550e-01
1.62525165e+00 1.56150879e-02 3.17250729e-01 7.50805914e-01
9.67251003e-01 -3.54334735e-03 -2.38542104e+00 -2.32465699e-01
4.64947730e-01 -5.40550590e-01 -1.32197365e-01 -7.14516819e-01
-7.78149486e-01 8.04614484e-01 4.86938864e-01 2.91567087e-01
6.86176181e-01 -1.84201866e-01 7.08869100e-01 3.83800805e-01
3.51189047e-01 -1.39141190e+00 2.68419683e-01 6.46714509e-01
8.47474933e-01 -1.03085804e+00 -1.23845778e-01 5.59778571e-01
-1.20436847e+00 1.29485631e+00 3.24171454e-01 -8.33342910e-01
5.88160574e-01 5.31447887e-01 -2.12616503e-01 1.58313498e-01
-7.71999836e-01 -1.85606480e-01 1.11451887e-01 7.47979343e-01
3.05781901e-01 4.38575804e-01 -1.68402240e-01 8.67890358e-01
-2.11287990e-01 -3.26971143e-01 3.64594698e-01 5.94800293e-01
-7.57651687e-01 -1.07696736e+00 -3.35991979e-01 3.94338369e-01
-5.95983207e-01 8.29721466e-02 2.74329066e-01 7.23602057e-01
-4.91229713e-01 7.71193624e-01 1.55809194e-01 1.93985835e-01
3.97653937e-01 -7.64353052e-02 5.87748468e-01 -4.28528786e-01
-4.44059879e-01 1.04938298e-01 -3.39630216e-01 -6.66120768e-01
2.22468197e-01 -9.23161387e-01 -1.63268542e+00 -4.01008993e-01
3.22653130e-02 4.67901856e-01 7.14365363e-01 5.79626799e-01
1.27707226e-02 2.89581060e-01 8.32357109e-01 -6.26905322e-01
-1.97339511e+00 -6.92561865e-01 -1.12370050e+00 1.93554044e-01
5.93381226e-01 -7.75126159e-01 -2.78216034e-01 -5.18332541e-01] | [4.087531566619873, 2.607748508453369] |
a90fd1e7-abf8-436f-83d1-b020650880fc | interaction-relational-network-for-mutual | 1910.04963 | null | https://arxiv.org/abs/1910.04963v2 | https://arxiv.org/pdf/1910.04963v2.pdf | Interaction Relational Network for Mutual Action Recognition | Person-person mutual action recognition (also referred to as interaction recognition) is an important research branch of human activity analysis. Current solutions in the field -- mainly dominated by CNNs, GCNs and LSTMs -- often consist of complicated architectures and mechanisms to embed the relationships between the two persons on the architecture itself, to ensure the interaction patterns can be properly learned. Our main contribution with this work is by proposing a simpler yet very powerful architecture, named Interaction Relational Network, which utilizes minimal prior knowledge about the structure of the human body. We drive the network to identify by itself how to relate the body parts from the individuals interacting. In order to better represent the interaction, we define two different relationships, leading to specialized architectures and models for each. These multiple relationship models will then be fused into a single and special architecture, in order to leverage both streams of information for further enhancing the relational reasoning capability. Furthermore we define important structured pair-wise operations to extract meaningful extra information from each pair of joints -- distance and motion. Ultimately, with the coupling of an LSTM, our IRN is capable of paramount sequential relational reasoning. These important extensions we made to our network can also be valuable to other problems that require sophisticated relational reasoning. Our solution is able to achieve state-of-the-art performance on the traditional interaction recognition datasets SBU and UT, and also on the mutual actions from the large-scale dataset NTU RGB+D. Furthermore, it obtains competitive performance in the NTU RGB+D 120 dataset interactions subset. | ['Alex C. Kot', 'Mauricio Perez', 'Jun Liu'] | 2019-10-11 | null | null | null | null | ['human-interaction-recognition'] | ['computer-vision'] | [ 6.95608407e-02 2.36663088e-01 -2.26703450e-01 -3.69472176e-01
-2.04039410e-01 -1.14881054e-01 6.93365812e-01 -3.50222647e-01
-2.57911444e-01 4.67894822e-01 5.04377902e-01 2.37473413e-01
-3.21931213e-01 -8.83430183e-01 -7.23193109e-01 -6.15022123e-01
-2.19181359e-01 7.60870814e-01 6.03279583e-02 -6.44307256e-01
-3.39234918e-01 4.71024930e-01 -1.61599696e+00 4.85472620e-01
4.16080952e-01 1.04775214e+00 -2.48768643e-01 6.43924296e-01
1.90532997e-01 1.22819483e+00 -3.53931814e-01 -4.74296272e-01
1.80279166e-01 -5.31536162e-01 -1.00523734e+00 8.84326473e-02
3.66461277e-01 -5.38110018e-01 -6.97375596e-01 3.84638727e-01
5.83971918e-01 5.50668836e-01 4.60010827e-01 -1.19956839e+00
-4.17685091e-01 7.94256330e-01 -3.24494123e-01 -1.07691035e-01
8.57840121e-01 2.17292786e-01 1.02089047e+00 -5.43668628e-01
5.58113158e-01 1.51030397e+00 6.92822516e-01 7.32371807e-01
-9.13473487e-01 -2.98632681e-01 4.11352903e-01 5.14862955e-01
-1.07757568e+00 -2.75177449e-01 7.85396755e-01 -2.43152469e-01
1.22794008e+00 3.25112909e-01 1.11574054e+00 1.64785743e+00
-9.53051001e-02 1.30462480e+00 4.57651228e-01 -2.33293638e-01
-1.91356152e-01 -4.77679908e-01 2.88187474e-01 8.27078521e-01
-1.10253930e-01 -8.43144432e-02 -7.19032586e-01 3.14680755e-01
9.63608384e-01 3.78902137e-01 -3.29355210e-01 -3.67791325e-01
-1.48125255e+00 4.09740210e-01 7.42333055e-01 4.79687661e-01
-3.19231093e-01 5.65301061e-01 3.61019194e-01 2.06850961e-01
1.61259070e-01 1.18880026e-01 -4.07221824e-01 -4.04491633e-01
-3.16221923e-01 4.14128035e-01 9.78466511e-01 7.55338550e-01
4.78053838e-01 -2.37754136e-01 -3.70122850e-01 7.73074150e-01
3.68696362e-01 2.45874375e-01 2.91854769e-01 -9.67042804e-01
7.42131531e-01 9.87622797e-01 -3.13162729e-02 -1.03260291e+00
-6.16542459e-01 -4.37830418e-01 -9.80292499e-01 -3.35650630e-02
5.41860282e-01 -3.72302067e-03 -7.01018751e-01 1.79111576e+00
2.87517071e-01 1.27746880e-01 -4.62712348e-02 1.05224526e+00
8.97052348e-01 1.91557765e-01 -8.36535990e-02 3.66347790e-01
1.53150201e+00 -1.24466622e+00 -7.31006503e-01 -2.79742986e-01
8.44030917e-01 -6.46946579e-02 8.61262619e-01 3.18416297e-01
-1.15638161e+00 -9.07487214e-01 -1.03044689e+00 -3.95714462e-01
-4.61882949e-01 3.47339600e-01 9.60210741e-01 3.64104211e-01
-9.44589496e-01 7.24916220e-01 -9.88344431e-01 -6.75996125e-01
3.41991901e-01 6.52327716e-01 -6.73128009e-01 1.75709084e-01
-1.33608377e+00 1.07074201e+00 2.57797033e-01 7.54420280e-01
-6.35019243e-01 -1.67148560e-01 -9.90564764e-01 4.20802366e-03
7.67882526e-01 -1.19632041e+00 1.14192963e+00 -6.72242701e-01
-1.61306202e+00 7.81150222e-01 1.06486566e-01 -6.37728989e-01
8.51911426e-01 -7.59059131e-01 -1.69736639e-01 2.32069753e-02
-2.19661325e-01 6.69864357e-01 4.36542034e-01 -9.73329782e-01
-6.05145395e-01 -6.52024627e-01 2.92461157e-01 5.11193335e-01
-4.15861875e-01 -1.10419579e-01 -9.85139966e-01 -6.63968921e-01
1.00389250e-01 -1.13983107e+00 -1.36545554e-01 -2.26144027e-02
-5.63504517e-01 -4.74478096e-01 7.46351182e-01 -7.34311700e-01
9.10567164e-01 -1.89064384e+00 8.60310853e-01 9.76389349e-02
3.19493234e-01 1.44409940e-01 -1.83998682e-02 3.14571053e-01
-9.31603983e-02 -2.71061152e-01 4.15996872e-02 -7.07162797e-01
2.50639021e-01 5.60311854e-01 -2.86885109e-02 4.03518349e-01
7.81688839e-02 1.54603124e+00 -7.00031459e-01 -3.23460311e-01
4.57966536e-01 6.91913307e-01 -2.58422375e-01 3.89111757e-01
-1.20688289e-01 6.02319777e-01 -3.82597297e-01 7.14485049e-01
3.61394696e-02 -4.80268262e-02 3.28450322e-01 -3.71082783e-01
3.44857305e-01 7.40925744e-02 -1.15527308e+00 1.90010369e+00
-2.73189306e-01 4.20362264e-01 -1.49038076e-01 -1.11609519e+00
8.40574980e-01 3.13176423e-01 7.87904859e-01 -5.54016054e-01
5.19342184e-01 -2.39920154e-01 1.25054806e-01 -6.65456474e-01
1.66500211e-01 3.86170417e-01 -1.31292135e-01 4.37208325e-01
2.14790925e-01 4.55089867e-01 2.75592357e-01 -1.44184530e-01
1.17604363e+00 7.25109458e-01 1.47890121e-01 2.33286142e-01
6.62426412e-01 -3.91713083e-01 3.04784805e-01 6.75163269e-01
-2.09268868e-01 4.43701476e-01 3.10481131e-01 -7.89798558e-01
-5.71436584e-01 -9.38987136e-01 4.58986610e-01 1.18461800e+00
1.09751914e-02 -4.36723650e-01 -6.48651719e-01 -8.29174519e-01
-1.01380922e-01 1.95939243e-01 -9.47359502e-01 -3.76251757e-01
-1.00772119e+00 -5.12511075e-01 7.96630442e-01 1.14870274e+00
9.51946139e-01 -1.32822585e+00 -7.13202298e-01 1.63541228e-01
-5.21271706e-01 -1.34215486e+00 -1.65452927e-01 1.90217122e-01
-7.68538237e-01 -1.18809247e+00 -6.53012872e-01 -5.61568141e-01
3.01099181e-01 1.98813546e-02 1.08931005e+00 1.27186328e-01
-8.91195312e-02 5.21711111e-01 -4.47580576e-01 -3.40892114e-02
1.10292658e-01 1.49992183e-01 1.47388428e-01 2.91404486e-01
3.33761662e-01 -8.24127376e-01 -5.65467000e-01 5.25664985e-01
-5.14402568e-01 1.22828931e-02 7.04070449e-01 6.42945647e-01
2.00033486e-01 -5.33703044e-02 -1.83945879e-01 -4.75896209e-01
3.61631006e-01 -6.74960315e-02 2.05537215e-01 3.37735295e-01
1.31546900e-01 1.51700661e-01 4.38172668e-01 -5.96785069e-01
-1.18540394e+00 1.47124261e-01 -1.48424909e-01 -5.20150721e-01
-3.07558388e-01 4.10622090e-01 -5.12672603e-01 8.40337947e-03
5.15088618e-01 1.24945052e-01 -3.61531670e-03 -5.93942285e-01
5.74217379e-01 2.49940425e-01 8.67213309e-01 -7.22364962e-01
5.11268735e-01 6.83880508e-01 1.46876499e-01 -5.82744181e-01
-7.54657924e-01 -3.10526580e-01 -1.18150687e+00 -5.14027476e-01
1.18873012e+00 -8.23445737e-01 -1.40919054e+00 8.32260311e-01
-1.17774403e+00 -6.31827414e-01 -2.94028014e-01 4.10668641e-01
-6.25561714e-01 4.45272982e-01 -9.41025078e-01 -9.60167587e-01
-1.83841169e-01 -9.29413617e-01 1.22618961e+00 1.06198922e-01
-4.69818950e-01 -1.01342964e+00 3.53298895e-02 8.02769363e-01
-5.61962696e-03 4.54485685e-01 6.00229919e-01 -6.00901604e-01
-4.22212362e-01 -2.62886494e-01 -2.18582191e-02 2.71835774e-01
9.05257985e-02 -2.05846623e-01 -7.15400517e-01 -2.27428135e-03
-2.29444727e-02 -3.43677193e-01 1.01225138e+00 1.89706132e-01
9.58622634e-01 -5.63405603e-02 -4.81944114e-01 6.25505805e-01
6.35760665e-01 2.02806488e-01 8.82318854e-01 3.06698829e-01
1.34636748e+00 1.00064588e+00 5.05351901e-01 3.09469879e-01
6.47943556e-01 1.19686520e+00 3.88040483e-01 -2.51599669e-01
-1.82227537e-01 -9.73423719e-02 5.76301157e-01 5.54948986e-01
-9.18391883e-01 -2.38333851e-01 -8.44611466e-01 1.33687600e-01
-2.58452344e+00 -1.21926498e+00 -1.91039592e-01 1.80361748e+00
4.10198122e-01 4.40610386e-02 3.62170100e-01 2.72594661e-01
4.11048561e-01 1.86501995e-01 -3.96905512e-01 -2.73194555e-02
-9.73873436e-02 1.12361513e-01 1.31831616e-01 1.84105352e-01
-1.25962162e+00 8.98405015e-01 6.05498266e+00 4.36673492e-01
-7.04116702e-01 -1.05470523e-01 2.96024680e-01 -2.27132842e-01
4.10356849e-01 -2.61076212e-01 -8.51058304e-01 1.29691854e-01
7.52125204e-01 6.29209220e-01 4.00815070e-01 7.13511884e-01
2.11392432e-01 -5.26652150e-02 -1.61949456e+00 1.24908566e+00
2.36807749e-01 -9.33258832e-01 6.80329800e-02 2.10112408e-01
1.27782598e-01 -4.07874316e-01 -1.48662627e-01 5.10486722e-01
3.98294389e-01 -1.07846880e+00 7.34093666e-01 1.01368141e+00
2.25459635e-01 -6.59767926e-01 8.68875146e-01 3.89389485e-01
-1.54393542e+00 -3.07197899e-01 2.97417697e-02 -3.69352967e-01
3.36226106e-01 1.91476211e-01 -2.59128571e-01 9.84741807e-01
1.05740905e+00 1.18887341e+00 -6.75298572e-01 4.37937945e-01
-4.56606686e-01 -1.98449902e-02 -3.29373956e-01 5.89504354e-02
9.19866189e-02 -2.71254331e-01 3.14975351e-01 9.16313469e-01
3.40175815e-02 1.43638507e-01 2.81395793e-01 7.33886778e-01
1.47077397e-01 -3.72031420e-01 -5.81176639e-01 8.99902880e-02
-1.84073165e-01 1.08918464e+00 -5.26439667e-01 -4.38960582e-01
-3.69248509e-01 1.31201303e+00 5.80276012e-01 2.55897313e-01
-1.11967504e+00 5.01988269e-02 7.84963250e-01 -1.17581561e-01
1.92557007e-01 -4.19201106e-01 -7.45813921e-02 -1.28261149e+00
3.57297063e-01 -1.01194167e+00 6.29362822e-01 -6.92356229e-01
-1.15507317e+00 6.21218681e-01 2.57911146e-01 -1.14786339e+00
-5.47647715e-01 -8.10082495e-01 -4.76969749e-01 5.88036537e-01
-7.96458244e-01 -1.75347590e+00 -6.39092565e-01 9.14710999e-01
3.74723822e-01 -4.52009104e-02 7.08880186e-01 3.87648493e-01
-9.20550346e-01 4.93763864e-01 -6.72968745e-01 6.23839080e-01
4.38069314e-01 -1.16692376e+00 4.56801891e-01 7.17665732e-01
3.73078138e-01 8.41277659e-01 3.90917808e-01 -6.12134099e-01
-1.46401918e+00 -7.96589494e-01 7.50870228e-01 -8.40820909e-01
5.38049042e-01 -6.19106889e-01 -5.81868470e-01 1.12420785e+00
-5.58506660e-02 -4.39752303e-02 6.79964244e-01 4.82363492e-01
-2.80210584e-01 -7.98724964e-02 -6.75662339e-01 9.50814366e-01
1.93535483e+00 -5.85478067e-01 -7.92983353e-01 2.51885355e-01
4.77932811e-01 -4.93486315e-01 -1.01857615e+00 5.37032664e-01
8.98542941e-01 -1.22224724e+00 1.38084912e+00 -8.32759440e-01
3.84473592e-01 -2.30493739e-01 -1.07736073e-01 -8.23535264e-01
-4.26752329e-01 -4.57287669e-01 -7.17164755e-01 1.08641505e+00
8.79672393e-02 -5.55658937e-01 1.03164947e+00 7.79637933e-01
-3.11855897e-02 -9.22890186e-01 -6.82212949e-01 -6.90449238e-01
-5.98300338e-01 -6.86425745e-01 5.27207196e-01 7.43003786e-01
3.09765581e-02 5.18732905e-01 -8.01037848e-01 -1.89850822e-01
4.68471825e-01 -8.32369477e-02 1.21728265e+00 -1.22098076e+00
-6.12375379e-01 -5.12585640e-01 -6.15440309e-01 -1.41764534e+00
2.42895156e-01 -5.37987411e-01 -5.56291081e-02 -1.65560555e+00
6.63895831e-02 1.33542120e-01 -1.67733252e-01 9.35540199e-01
-6.20686449e-02 3.53053927e-01 2.92018712e-01 2.43268728e-01
-7.25535274e-01 7.87351489e-01 1.29928863e+00 -2.98640013e-01
-3.87357086e-01 1.41033277e-01 -3.89934450e-01 8.61529648e-01
3.61108035e-01 8.44829008e-02 -4.36540365e-01 -3.57644886e-01
6.62096813e-02 7.29397610e-02 9.18172956e-01 -1.47981322e+00
3.38762283e-01 1.98731735e-01 6.75675690e-01 -6.96118057e-01
8.17488313e-01 -8.66127610e-01 3.23803216e-01 5.25928080e-01
-3.06014717e-01 -9.38107148e-02 -2.22332388e-01 4.49560463e-01
-2.17883408e-01 4.15633231e-01 2.22235978e-01 -3.25952858e-01
-8.84082675e-01 4.34428781e-01 -2.41088912e-01 -4.12131965e-01
1.06111932e+00 -4.75296110e-01 -1.46352157e-01 -6.20901763e-01
-1.26997983e+00 3.70298713e-01 -2.85051335e-02 7.85313785e-01
3.94723415e-01 -1.53606546e+00 -3.55801105e-01 -2.02782210e-02
-1.37234479e-02 8.37320611e-02 4.85596478e-01 1.12422502e+00
-2.66225398e-01 5.67922473e-01 -4.95085001e-01 -6.04362547e-01
-1.31434083e+00 3.36418360e-01 6.08103573e-01 -5.61492503e-01
-7.93474615e-01 8.58887613e-01 4.64554764e-02 -5.62502861e-01
4.16216731e-01 -5.21254778e-01 -4.60949242e-01 1.79070309e-01
3.81539524e-01 5.29879332e-01 -3.92513089e-02 -9.01574969e-01
-4.51026380e-01 6.35709465e-01 1.25671983e-01 -1.03896387e-01
1.30237699e+00 4.09004800e-02 -1.63184926e-01 4.27382946e-01
1.03627646e+00 -5.67143738e-01 -1.39290988e+00 -2.58894503e-01
-1.35539770e-02 -2.97128826e-01 -4.28186566e-01 -6.09765768e-01
-1.26875710e+00 8.13144684e-01 3.17984998e-01 1.21000580e-01
9.19921458e-01 1.42411843e-01 8.62936556e-01 7.75522888e-01
4.65977341e-01 -1.05343866e+00 4.14779961e-01 5.57946801e-01
1.00502837e+00 -1.05198920e+00 -5.32573052e-02 -4.11507547e-01
-5.75528145e-01 1.13299274e+00 9.72998202e-01 -5.75114526e-02
3.80878121e-01 1.08494133e-01 -8.41130763e-02 -4.48536515e-01
-5.52189291e-01 -5.69408894e-01 5.99336207e-01 6.28974915e-01
3.67047101e-01 -3.45063768e-02 -8.10590945e-03 7.24268019e-01
-2.66519755e-01 -5.10729104e-02 -2.43740499e-01 8.67425680e-01
2.45375596e-02 -1.34965158e+00 -3.16636950e-01 2.70563513e-01
1.02762133e-02 4.17718828e-01 -7.94898152e-01 1.04744911e+00
3.89360100e-01 7.56072640e-01 -6.93553314e-03 -7.94572532e-01
7.76098132e-01 5.45508368e-03 7.66898036e-01 -4.02961344e-01
-8.38068128e-01 -1.04323387e-01 2.96978980e-01 -1.21536064e+00
-7.13031888e-01 -6.76071584e-01 -1.30499637e+00 -4.02464896e-01
8.39001313e-02 -3.39669853e-01 8.24823976e-02 1.41312039e+00
1.21609151e-01 7.95923650e-01 -6.31490797e-02 -1.35001004e+00
-2.01842278e-01 -1.01557779e+00 -4.40031260e-01 6.36613309e-01
3.94913219e-02 -9.78177786e-01 -1.47367353e-02 -6.54621124e-02] | [7.961227893829346, 0.41291823983192444] |
455d8eea-73a9-4433-946c-4eab0a3500d1 | nowj-at-coliee-2023-multi-task-and-ensemble | 2306.04903 | null | https://arxiv.org/abs/2306.04903v1 | https://arxiv.org/pdf/2306.04903v1.pdf | NOWJ at COLIEE 2023 -- Multi-Task and Ensemble Approaches in Legal Information Processing | This paper presents the NOWJ team's approach to the COLIEE 2023 Competition, which focuses on advancing legal information processing techniques and applying them to real-world legal scenarios. Our team tackles the four tasks in the competition, which involve legal case retrieval, legal case entailment, statute law retrieval, and legal textual entailment. We employ state-of-the-art machine learning models and innovative approaches, such as BERT, Longformer, BM25-ranking algorithm, and multi-task learning models. Although our team did not achieve state-of-the-art results, our findings provide valuable insights and pave the way for future improvements in legal information processing. | ['Ha-Thanh Nguyen', 'Thai-Binh Nguyen', 'Hoang-Trung Nguyen', 'Tan-Minh Nguyen', 'Hai-Long Nguyen', 'Thi-Hai-Yen Vuong'] | 2023-06-08 | null | null | null | null | ['multi-task-learning', 'natural-language-inference'] | ['methodology', 'natural-language-processing'] | [-8.63826647e-02 -4.04289253e-02 -8.45545888e-01 -3.07439655e-01
-1.73623049e+00 -7.28406727e-01 7.33615756e-01 2.26063102e-01
-7.90117085e-01 9.10402596e-01 8.57268333e-01 -1.06072998e+00
-8.40142250e-01 -4.52772260e-01 -3.77532303e-01 2.55678713e-01
2.88406690e-03 8.78024936e-01 1.42302647e-01 -4.95170385e-01
8.03918540e-01 3.38191599e-01 -8.44671488e-01 1.05661023e+00
1.28112268e+00 5.19828498e-01 -7.65239000e-01 6.05963528e-01
-3.37094396e-01 1.61890948e+00 -6.64592505e-01 -1.42758119e+00
4.66627002e-01 3.78121704e-01 -1.31943142e+00 -9.77104604e-01
7.72908807e-01 -7.14030445e-01 -6.21136606e-01 6.24741256e-01
6.90901518e-01 -1.87124088e-01 1.02787519e+00 -8.49402249e-01
-7.65631735e-01 6.32098913e-01 -5.68584025e-01 7.69016802e-01
9.09644127e-01 -9.77222472e-02 1.45352566e+00 -2.63081789e-01
1.22573936e+00 1.45974123e+00 6.63848102e-01 4.25417930e-01
-7.90820777e-01 -3.44244450e-01 -5.67328855e-02 8.79985034e-01
-6.85094535e-01 -6.29526198e-01 4.81573313e-01 -9.35765505e-01
1.39925694e+00 -6.65783510e-02 2.51899451e-01 8.79876554e-01
8.01880658e-01 1.26215625e+00 1.04223228e+00 -4.65175956e-01
-2.74324231e-02 -4.44015384e-01 6.34641528e-01 1.71094075e-01
4.88110781e-01 1.59002960e-01 2.46644532e-03 -7.52411842e-01
1.28569424e-01 -1.18649058e-01 5.59002399e-01 3.26466531e-01
-4.10665303e-01 9.26456690e-01 -9.57234055e-02 6.91690207e-01
-4.55498010e-01 -1.90742705e-02 8.23961675e-01 7.39178002e-01
4.67985421e-01 3.13585222e-01 -5.50750494e-01 -5.16358554e-01
-1.08462608e+00 1.06625247e+00 7.81614780e-01 6.07626796e-01
1.38374254e-01 -7.83371091e-01 -9.27676916e-01 9.04821455e-01
3.60131890e-01 3.62212360e-01 -4.53864038e-02 -1.23802567e+00
1.35890472e+00 6.53406739e-01 6.22764602e-02 -4.99525338e-01
-2.57114947e-01 -2.36298665e-02 -6.17934912e-02 9.90121141e-02
4.96313900e-01 -2.58501381e-01 -1.02984810e+00 8.50960851e-01
-2.72622973e-01 -2.49452651e-01 1.48821637e-01 2.25309983e-01
6.69581175e-01 2.48740792e-01 3.93499911e-01 -2.56120740e-03
1.16046083e+00 -6.03685021e-01 -7.21486330e-01 -1.79646388e-01
8.76558840e-01 -1.08586526e+00 5.88862002e-01 2.64274359e-01
-1.43079197e+00 -1.53333947e-01 -4.33849066e-01 -5.55505276e-01
-3.89785618e-01 -1.38037175e-01 9.69341815e-01 7.07969189e-01
-5.83981037e-01 3.53372663e-01 -8.31491426e-02 -2.97887158e-02
9.44469988e-01 -3.23126167e-01 -1.20077245e-01 -4.60838050e-01
-1.71617949e+00 1.57459879e+00 -7.19188079e-02 -9.74419191e-02
-3.80436271e-01 -8.60238373e-01 -7.20309615e-01 -1.06254466e-01
7.07672417e-01 -4.36147630e-01 1.32358766e+00 5.36322594e-01
-7.67343163e-01 1.27003431e+00 -2.36878797e-01 -6.83169246e-01
8.70635986e-01 -4.54074174e-01 -8.60551536e-01 -1.04453363e-01
7.81440914e-01 -1.25998542e-01 1.91932946e-01 -6.16701424e-01
-8.31582427e-01 -3.52978945e-01 4.49805051e-01 -2.93026745e-01
3.03617090e-01 8.83487821e-01 -3.22481930e-01 -2.97407866e-01
-5.08132935e-01 -6.31622851e-01 -4.61929113e-01 -6.07453346e-01
-2.67354250e-01 -7.44397283e-01 4.47909310e-02 -8.37793052e-01
1.62780714e+00 -1.76385844e+00 -2.71860898e-01 4.00288552e-01
7.02144057e-02 6.68709636e-01 -1.70109689e-01 9.86193836e-01
1.81993753e-01 4.74443287e-01 3.02753448e-01 -5.41194715e-02
4.72235769e-01 2.56333411e-01 -4.20437127e-01 2.16271609e-01
-1.32662505e-01 1.15190148e+00 -6.40518844e-01 -9.99325097e-01
1.98677912e-01 -2.25659996e-01 -5.30807734e-01 -4.21555340e-01
-1.20455713e-03 -3.79441452e-04 -8.04637372e-01 7.58874714e-01
5.53213596e-01 2.77644068e-01 -5.47180399e-02 1.50116548e-01
-9.53576863e-02 8.55003476e-01 -6.76041126e-01 1.73202848e+00
-4.62605506e-01 2.36037359e-01 -1.28328592e-01 -7.36315906e-01
2.52169251e-01 2.94928044e-01 5.19601345e-01 -1.33324242e+00
-2.06550695e-02 3.06042522e-01 2.88957477e-01 -1.15939450e+00
2.25433081e-01 1.69469994e-02 -4.36585516e-01 6.33695662e-01
-5.37390172e-01 -1.16359480e-02 1.08050382e+00 5.40239275e-01
1.34114444e+00 -1.06119692e-01 4.61455107e-01 4.30873595e-02
5.71930885e-01 2.14921571e-02 5.23113549e-01 1.13999999e+00
-3.89509678e-01 -1.48616925e-01 8.41324389e-01 -7.88296938e-01
-7.13133156e-01 -1.09164059e+00 -7.51831651e-01 1.12648404e+00
-5.55131555e-01 -3.16023976e-01 -3.50160956e-01 -9.18337584e-01
6.68595076e-01 7.67618835e-01 -6.50960982e-01 1.13792405e-01
-6.54578269e-01 -3.66701394e-01 8.72969210e-01 4.05605912e-01
1.41170859e-01 -1.13397920e+00 -3.48104358e-01 4.74069595e-01
-2.41624013e-01 -1.22154939e+00 -3.02658021e-01 -5.56697726e-01
-5.74591517e-01 -1.59712434e+00 -4.87995028e-01 -2.19534576e-01
-1.84365790e-02 -4.45474476e-01 9.89241242e-01 1.74368754e-01
-6.75941110e-01 6.80301785e-02 -3.69324327e-01 -6.67625189e-01
-5.42492807e-01 1.78774163e-01 -4.36877728e-01 -6.36399567e-01
8.33648980e-01 -3.59835446e-01 -2.43308857e-01 -3.36403012e-01
-9.08733070e-01 -7.65416861e-01 8.87739003e-01 4.82193798e-01
-7.84260482e-02 -4.53159243e-01 8.30380559e-01 -1.47461379e+00
1.66195297e+00 -3.35982203e-01 -5.08053780e-01 6.96686447e-01
-1.05786836e+00 4.85397205e-02 2.13883325e-01 4.25506145e-01
-1.33730531e+00 -8.17852974e-01 -4.98343319e-01 2.86154509e-01
-1.98754981e-01 9.85257387e-01 3.55537534e-01 1.21128045e-01
7.82059133e-01 -2.68716961e-01 -4.37818259e-01 -4.51332211e-01
8.71822476e-01 1.02905512e+00 3.92036766e-01 -8.04256558e-01
7.48342514e-01 3.62855464e-01 -1.16396524e-01 -1.55418189e-02
-1.14549029e+00 -8.20758939e-01 -8.48870516e-01 -3.38210940e-01
7.93934345e-01 -5.64483285e-01 -7.69848108e-01 1.72974780e-01
-1.35012734e+00 1.51753187e-01 6.52739406e-02 3.70146930e-01
-2.28596389e-01 7.10597038e-01 -1.21489942e+00 -1.02697110e+00
-5.73899388e-01 -8.37962687e-01 9.63672936e-01 -3.09698433e-01
-3.51277828e-01 -9.81918037e-01 5.30037344e-01 1.72810757e+00
4.16904569e-01 2.18995884e-01 1.40419853e+00 -8.33069384e-01
-1.66364610e-01 -8.72785091e-01 -3.35650593e-01 9.46940705e-02
-3.16716224e-01 1.55371085e-01 -6.40637219e-01 7.82381594e-02
-5.01521587e-01 -5.45349240e-01 1.22002304e+00 6.20387018e-01
5.98165810e-01 -8.04758370e-02 -4.00311738e-01 -1.25733137e-01
1.15094841e+00 6.47125781e-01 1.08901417e+00 7.34060645e-01
1.38594702e-01 5.31517267e-01 7.14913130e-01 3.99138063e-01
7.16134846e-01 5.82872748e-01 -1.72794104e-01 1.49823338e-01
7.42667988e-02 -8.95076618e-02 -1.28570035e-01 2.10410863e-01
-6.44649267e-01 9.71715078e-02 -1.28009927e+00 6.40053093e-01
-2.24170423e+00 -1.58671129e+00 -2.22371876e-01 1.55410528e+00
6.26125753e-01 6.45641208e-01 1.91193238e-01 7.74687901e-02
2.89756984e-01 2.91398138e-01 -3.36905867e-01 -1.00204933e+00
3.25950272e-02 3.83240551e-01 3.75841945e-01 4.39053863e-01
-1.20972371e+00 1.14040327e+00 7.46174669e+00 1.16336966e+00
-2.95145541e-01 1.74799904e-01 5.85938692e-01 -2.76170731e-01
-5.21956980e-01 1.44541457e-01 -5.74862421e-01 4.70434070e-01
6.72384977e-01 -6.01150751e-01 3.26662838e-01 8.02490890e-01
2.60597169e-01 -9.80604514e-02 -8.34115803e-01 6.12149179e-01
1.62764117e-01 -1.66486752e+00 6.69178143e-02 4.79258299e-01
8.08326602e-01 3.28083158e-01 -1.48102522e-01 1.03649700e+00
4.82913524e-01 -9.21880007e-01 4.68261838e-01 7.19915569e-01
5.12601554e-01 -4.61547375e-01 1.13694477e+00 2.54279315e-01
-7.42834210e-01 -6.16270363e-01 -4.26134378e-01 -3.37618858e-01
7.32877791e-01 6.29000962e-01 -2.44917646e-01 9.98824656e-01
3.90784174e-01 7.59565532e-01 -4.46487576e-01 1.17727470e+00
-2.51148343e-01 2.42688701e-01 2.72745579e-01 -1.28881913e-02
4.90704358e-01 -2.23221049e-01 3.45578641e-01 1.24053478e+00
-2.79362768e-01 2.15380371e-01 2.47984752e-01 4.17848706e-01
-2.61226714e-01 3.15288752e-01 -8.54927957e-01 3.69467258e-01
4.64225560e-01 7.25128055e-01 9.12192613e-02 -6.46075904e-01
-6.43652856e-01 5.89779429e-02 3.55111271e-01 3.19418699e-01
-7.12287962e-01 -3.29604298e-01 3.20930451e-01 1.67981803e-01
1.61242813e-01 2.84678668e-01 -4.32057023e-01 -1.05687368e+00
4.03645754e-01 -8.05199862e-01 9.13253725e-01 -4.19153750e-01
-1.75759149e+00 2.25979865e-01 3.91482323e-01 -9.64138985e-01
-5.39251268e-01 -5.88535607e-01 -9.26375449e-01 5.10333896e-01
-1.87750816e+00 -1.11415625e+00 7.45397806e-01 5.07880092e-01
2.37828285e-01 -6.53235853e-01 4.72221375e-01 1.03413129e+00
-3.04219335e-01 5.28511405e-01 1.19296961e-01 3.55404377e-01
1.01390839e+00 -9.94266212e-01 4.72921640e-01 6.13745034e-01
3.59568410e-02 9.35541868e-01 1.63348123e-01 -9.44482505e-01
-1.03211355e+00 -6.06900215e-01 1.43653607e+00 -6.99546516e-01
1.15537214e+00 2.55432338e-01 -2.87520170e-01 7.01020479e-01
2.23075405e-01 -6.81621909e-01 8.34624827e-01 8.13289940e-01
-5.71482122e-01 -2.68491238e-01 -1.16431713e+00 4.72230971e-01
1.18982804e+00 -1.01654398e+00 -1.30110419e+00 8.73166025e-01
-1.17182679e-01 -3.86858463e-01 -1.05037522e+00 3.72518212e-01
9.57572937e-01 -7.61619270e-01 1.07178104e+00 -1.34675646e+00
6.99804664e-01 3.58657122e-01 2.98498601e-01 -6.96874797e-01
-4.11398858e-01 -4.05515969e-01 3.68789524e-01 9.80936766e-01
1.22769213e+00 -6.49272680e-01 5.15220106e-01 1.21854556e+00
-1.92438751e-01 -9.70748007e-01 -1.50235951e+00 -9.11838770e-01
6.30604625e-01 -9.97204900e-01 4.85167921e-01 9.23626602e-01
5.28323710e-01 3.42259884e-01 -4.54644799e-01 -5.42956829e-01
5.40091932e-01 3.78995717e-01 6.93307817e-01 -1.35210943e+00
3.13584395e-02 -6.52937412e-01 -5.00943422e-01 -6.22871757e-01
5.42726755e-01 -9.59313333e-01 -7.65874743e-01 -2.27796340e+00
4.79049474e-01 6.92888945e-02 -2.49807924e-01 6.32234275e-01
-2.74432093e-01 -4.91655022e-01 3.13719779e-01 2.83350289e-01
-9.68640983e-01 -1.27089784e-01 1.19847059e+00 -5.22010803e-01
2.18442842e-01 4.27035242e-01 -9.47559118e-01 4.76100773e-01
5.03218710e-01 -5.82893789e-01 -1.98709264e-01 -6.92017496e-01
4.53361660e-01 2.19769582e-01 -7.54266828e-02 -5.33545494e-01
6.28190815e-01 -2.18208402e-01 3.64918113e-02 -7.47340679e-01
8.77968669e-02 -4.42107677e-01 -4.92581189e-01 5.13393164e-01
-5.20528495e-01 -7.36701414e-02 6.89822435e-02 3.77616644e-01
-3.91668439e-01 -5.88999927e-01 8.68140906e-02 -2.41901338e-01
-4.31675613e-01 2.26141423e-01 -7.16379762e-01 4.99983817e-01
9.52795386e-01 -2.39462554e-02 -1.05166543e+00 -4.36342001e-01
-6.78715527e-01 8.08516562e-01 -2.88098186e-01 3.92679781e-01
4.37965482e-01 -9.50272918e-01 -1.48444116e+00 -4.34941590e-01
2.33656809e-01 -7.23476827e-01 3.71201873e-01 7.53384233e-01
-3.36802214e-01 1.16548705e+00 6.62050769e-02 1.01979874e-01
-1.28822362e+00 3.23361695e-01 -7.58969188e-02 -1.16465628e+00
-2.52295434e-01 3.14380914e-01 -8.89247596e-01 -3.49769741e-01
1.47338316e-01 -3.70600462e-01 -7.39085615e-01 2.31299296e-01
5.39631784e-01 6.68119311e-01 1.61217719e-01 -2.01521680e-01
-4.83651519e-01 8.58465493e-01 -6.98758841e-01 -1.60159916e-01
1.44252574e+00 5.53042471e-01 -4.17768627e-01 3.51424702e-02
8.82675052e-01 1.78277507e-01 1.48714827e-02 -4.55117412e-02
7.06176281e-01 -6.41716897e-01 -1.22424617e-01 -1.28996980e+00
-6.65310740e-01 5.01127422e-01 7.33438209e-02 1.93554804e-01
5.74586928e-01 2.48459414e-01 1.03537500e+00 8.98928642e-01
5.36544383e-01 -1.57030809e+00 -6.20654225e-01 7.45587051e-01
9.52605963e-01 -1.18709540e+00 3.48901033e-01 -2.49249130e-01
-6.55594885e-01 9.04271781e-01 1.29651621e-01 -6.33528605e-02
6.83349311e-01 2.06369504e-01 4.52687740e-01 -6.30245864e-01
-9.90217030e-01 -3.62072051e-01 5.95383883e-01 1.79477304e-01
5.43015480e-01 -1.15380943e-01 -1.16968584e+00 3.29510450e-01
-4.39412408e-02 3.98973137e-01 1.84570447e-01 9.56259787e-01
-1.19265199e-01 -1.63662636e+00 -1.44131467e-01 1.22705793e+00
-1.10261083e+00 -5.93160152e-01 -7.42637038e-01 9.91186976e-01
1.50793448e-01 1.27824092e+00 -5.23426712e-01 -5.81854545e-02
9.48737204e-01 2.14527518e-01 3.89182985e-01 -5.68422198e-01
-8.48690748e-01 -4.17809002e-02 1.00110757e+00 -6.07477248e-01
-3.90931845e-01 -9.19427931e-01 -9.22671497e-01 -3.96611124e-01
-5.56116067e-02 3.92325044e-01 3.75875950e-01 1.46671021e+00
2.66184658e-01 3.87117416e-01 6.12147935e-02 1.10395946e-01
-6.98336601e-01 -1.25549006e+00 -3.90481114e-01 6.92982674e-01
3.56379300e-02 -3.80847037e-01 -4.17639799e-02 -3.12401533e-01] | [9.916411399841309, 9.220111846923828] |
48339829-3251-41dd-8575-0aa2578b3b62 | learning-document-embeddings-with-cnns | null | null | https://openreview.net/forum?id=ryHM_fbA- | https://openreview.net/pdf?id=ryHM_fbA- | Learning Document Embeddings With CNNs | This paper proposes a new model for document embedding. Existing approaches either require complex inference or use recurrent neural networks that are difficult to parallelize. We take a different route and use recent advances in language modeling to develop a convolutional neural network embedding model. This allows us to train deeper architectures that are fully parallelizable. Stacking layers together increases the receptive filed allowing each successive layer to model increasingly longer range semantic dependences within the document. Empirically we demonstrate superior results on two publicly available benchmarks. Full code will be released with the final version of this paper. | ['Maksims Volkovs', 'Shunan Zhao', 'Chundi Lui'] | 2018-01-01 | null | null | null | iclr-2018-1 | ['document-embedding'] | ['methodology'] | [-1.46677494e-01 -2.88073737e-02 -4.76483047e-01 -4.69657958e-01
-4.29201871e-01 -5.64404607e-01 7.34624326e-01 1.53458729e-01
-5.80514014e-01 4.41484421e-01 4.40012664e-01 -6.34417117e-01
3.20650846e-01 -8.76317680e-01 -5.54222286e-01 -1.17755435e-01
-6.32263273e-02 1.53788850e-01 2.31955022e-01 -3.12391907e-01
3.76704425e-01 5.79479396e-01 -1.12609720e+00 6.68249071e-01
1.38011724e-01 4.80580747e-01 1.63131401e-01 1.05477583e+00
-5.01473367e-01 8.36980820e-01 -5.26218772e-01 -2.60098994e-01
9.24056992e-02 -2.19038632e-02 -7.83451915e-01 -2.31380448e-01
2.89526403e-01 -5.12127876e-01 -7.60809779e-01 7.05398738e-01
3.58966976e-01 2.01505691e-01 4.27231163e-01 -7.88957119e-01
-1.01114917e+00 8.41634154e-01 -1.73434630e-01 3.75791281e-01
-3.09972502e-02 -2.89965302e-01 1.22878230e+00 -1.10694385e+00
4.69698280e-01 1.33647180e+00 8.37975800e-01 5.69884002e-01
-1.22802341e+00 -5.57912767e-01 4.34724033e-01 2.96804570e-02
-1.10370505e+00 -4.79209989e-01 8.34354103e-01 -1.14581846e-01
1.82850897e+00 1.74506694e-01 5.28960824e-01 1.19555068e+00
2.78935671e-01 8.80948961e-01 7.98689425e-01 -5.91302693e-01
-1.71908528e-01 2.22449943e-01 6.97628856e-01 8.63173366e-01
2.95512408e-01 -1.18002132e-01 -2.45011657e-01 -1.02638401e-01
7.41493881e-01 4.23806995e-01 1.23929821e-01 -8.30341205e-02
-9.79362905e-01 1.22023368e+00 5.62718928e-01 6.33292317e-01
-6.44310936e-02 4.82281655e-01 6.46416545e-01 5.32477319e-01
4.97765034e-01 4.60742801e-01 -4.50355262e-01 -5.15365936e-02
-9.87789989e-01 1.72508791e-01 8.90706718e-01 7.55805433e-01
5.04621744e-01 9.01085660e-02 -3.50962617e-02 1.07914662e+00
2.69630700e-01 4.74533774e-02 7.84616888e-01 -7.71203399e-01
3.57467502e-01 4.39292699e-01 -2.14447692e-01 -1.00105965e+00
-4.57881629e-01 -4.63863790e-01 -6.50332987e-01 9.23510566e-02
-4.97945957e-02 -6.58812597e-02 -9.41876113e-01 1.25683212e+00
-3.09516281e-01 6.37167459e-03 1.88940436e-01 5.66908300e-01
7.36084044e-01 1.00772142e+00 -1.48105556e-02 4.24637526e-01
1.13598943e+00 -1.53548968e+00 -7.23966897e-01 -3.39495987e-01
9.13949311e-01 -7.59992898e-01 1.15795350e+00 2.45319948e-01
-1.17072117e+00 -6.06611669e-01 -1.39809012e+00 -6.04266047e-01
-9.55029190e-01 2.08047360e-01 1.06990993e+00 5.49597025e-01
-1.43918145e+00 5.07489979e-01 -1.16229320e+00 -3.20541799e-01
2.91879326e-01 5.05204678e-01 -1.75425157e-01 7.75685683e-02
-1.29048121e+00 9.30837750e-01 5.48470676e-01 1.47652939e-01
-4.35058802e-01 -4.25979704e-01 -1.07573295e+00 1.81229323e-01
-1.87788993e-01 -5.76957166e-01 1.32663548e+00 -8.99021685e-01
-1.55829978e+00 6.88655734e-01 -3.35895538e-01 -6.55294716e-01
7.72099346e-02 -2.12707207e-01 -3.21945518e-01 -1.54856563e-01
-2.87871987e-01 7.88148344e-01 6.12122655e-01 -9.43888068e-01
-1.60088390e-01 -2.03193814e-01 3.28224033e-01 -7.77321905e-02
-9.26554620e-01 4.03062820e-01 -5.45892179e-01 -9.35122311e-01
-3.76153253e-02 -1.01171434e+00 -3.28922451e-01 -1.35611566e-02
-3.14225942e-01 -3.84343714e-01 8.36889803e-01 -5.41329801e-01
1.47663713e+00 -2.19315934e+00 1.85875550e-01 2.53053270e-02
2.82417238e-01 3.94003242e-01 -3.98117334e-01 6.46367729e-01
-1.26754686e-01 3.35935175e-01 -7.86563531e-02 -7.19255328e-01
1.55834779e-01 5.26492596e-01 -5.70324123e-01 6.94278553e-02
1.55722141e-01 1.23252141e+00 -4.76365298e-01 -1.94617763e-01
3.88923362e-02 7.41913319e-01 -6.83805883e-01 6.10455647e-02
-1.53133750e-01 -4.13720757e-01 -2.10025638e-01 2.47829989e-01
4.93158430e-01 -4.96256769e-01 1.29235789e-01 -2.03312133e-02
-1.21768877e-01 8.65910232e-01 -8.25308681e-01 1.82072043e+00
-7.23360837e-01 1.00104034e+00 -2.53099740e-01 -1.06132722e+00
9.05526459e-01 1.87628672e-01 -5.51739149e-02 -4.88293320e-01
-1.00835614e-01 1.72047138e-01 -9.84064862e-02 -2.11295053e-01
8.91620219e-01 -6.19403794e-02 2.48095561e-02 6.93007112e-01
2.01439962e-01 3.45667824e-02 1.31356359e-01 2.42723256e-01
1.13449121e+00 -1.79339364e-01 -1.65900157e-03 -1.30696490e-01
3.76178294e-01 -1.77795008e-01 -2.24896166e-02 7.45811999e-01
9.91208106e-02 4.42696959e-01 4.49762970e-01 -8.40845525e-01
-1.25656772e+00 -9.96408582e-01 -1.34232953e-01 1.29623735e+00
-5.02469063e-01 -8.50935638e-01 -3.26433122e-01 -6.11854374e-01
5.80886193e-02 6.04834557e-01 -7.69974887e-01 -9.34331492e-02
-8.17120671e-01 -5.94086647e-01 8.83308053e-01 1.06319582e+00
1.52574092e-01 -1.13602340e+00 -2.20515981e-01 2.32500449e-01
2.64416546e-01 -8.65457952e-01 -1.74543872e-01 5.77865064e-01
-1.07167220e+00 -4.21051353e-01 -7.30995715e-01 -1.04018128e+00
6.22365177e-01 7.77008981e-02 1.06336641e+00 4.96119559e-01
-1.41335055e-01 6.33721128e-02 -9.71417055e-02 -3.95072937e-01
-3.19349766e-01 6.11076951e-01 -1.85160398e-01 -6.47996426e-01
6.80959582e-01 -3.61162931e-01 -3.53170186e-01 -3.10277224e-01
-1.06233943e+00 -5.61991669e-02 4.37070996e-01 9.97480929e-01
2.79001713e-01 -1.12497211e-01 3.53134632e-01 -9.80036318e-01
1.03066969e+00 -4.22434479e-01 -4.28657055e-01 1.38121814e-01
-6.85594141e-01 3.66084397e-01 6.87858701e-01 -3.65000784e-01
-6.19428337e-01 -2.07442090e-01 -4.74010438e-01 -2.49994919e-01
-5.74333407e-02 8.25485110e-01 5.88886142e-01 1.14198096e-01
3.25576156e-01 1.29011557e-01 -9.10143107e-02 -5.93549848e-01
3.22351396e-01 7.13652968e-01 2.26488829e-01 -4.23935711e-01
5.74676812e-01 3.79366398e-01 -4.27456290e-01 -8.37983727e-01
-7.24712908e-01 -2.52443075e-01 -5.92067301e-01 4.01876539e-01
6.59884453e-01 -8.85989189e-01 -2.76148528e-01 1.21496491e-01
-1.54485786e+00 -4.61623609e-01 3.56334671e-02 5.79836965e-01
-8.83803368e-02 1.95359498e-01 -1.22049975e+00 -3.39514285e-01
-4.04759943e-01 -1.06807888e+00 9.70212340e-01 -6.44819960e-02
-3.93489838e-01 -1.57788277e+00 4.33423668e-01 -5.51447608e-02
7.08154678e-01 -4.81585711e-01 8.43123794e-01 -7.23557174e-01
-2.86952227e-01 -3.64211291e-01 -3.73794883e-01 5.18124700e-01
3.22935656e-02 2.01595098e-01 -9.62040067e-01 -4.55287129e-01
-2.48916626e-01 -3.62657875e-01 1.33750045e+00 2.39803791e-01
1.47397745e+00 -2.31068596e-01 -4.14259821e-01 7.33687103e-01
1.34508586e+00 -1.85172111e-01 5.81252396e-01 6.88984454e-01
7.37483323e-01 3.76337260e-01 -1.10416792e-01 1.46556139e-01
4.83496964e-01 6.44697368e-01 -1.35174647e-01 -2.26527810e-01
-2.06307799e-01 -2.93600976e-01 5.79050183e-01 1.39748406e+00
1.37949675e-01 -3.14205438e-01 -1.11740994e+00 5.60906172e-01
-1.69104600e+00 -8.62885892e-01 1.39091119e-01 1.54410386e+00
8.31435502e-01 3.35971147e-01 -1.66707888e-01 5.95442839e-02
1.53716117e-01 4.14650351e-01 -2.43425325e-01 -1.15907943e+00
-5.67148328e-02 6.12021029e-01 4.99380648e-01 8.62893522e-01
-9.93368208e-01 1.25349820e+00 7.88722992e+00 3.14834386e-01
-1.38112557e+00 6.94635659e-02 4.00517702e-01 -1.80167869e-01
-7.22840846e-01 7.67405238e-03 -1.03439260e+00 2.51700372e-01
1.25888932e+00 6.04447126e-02 2.95889378e-01 6.71812534e-01
-1.23709425e-01 3.92034739e-01 -1.04820132e+00 8.18324566e-01
2.45836094e-01 -1.57495284e+00 2.28327870e-01 -3.18072066e-02
5.87834358e-01 5.38549960e-01 2.69130111e-01 4.90410060e-01
4.19973612e-01 -1.35640585e+00 1.97969213e-01 5.39555252e-01
5.34517229e-01 -8.62766743e-01 6.21439934e-01 1.28170654e-01
-1.08644974e+00 -1.13074489e-01 -7.05964267e-01 -4.49827403e-01
-6.19369075e-02 2.66219348e-01 -7.05832481e-01 2.57973045e-01
6.06766999e-01 9.65195358e-01 -8.47788811e-01 4.98580396e-01
-2.16991290e-01 5.82216918e-01 -2.32673243e-01 -3.12901348e-01
5.47138393e-01 2.00277224e-01 1.96383268e-01 1.72164750e+00
1.76611081e-01 -3.84588093e-01 1.49656668e-01 7.65861213e-01
-2.93042243e-01 5.31413555e-02 -8.87287676e-01 -4.96388733e-01
1.86386138e-01 1.10476661e+00 -7.04632282e-01 -5.28037548e-01
-7.40488112e-01 1.13094175e+00 7.66067803e-01 5.41777909e-01
-6.70046866e-01 -6.68289363e-01 6.15184367e-01 -1.72938824e-01
6.55283272e-01 -7.44528890e-01 -3.80328387e-01 -1.46159041e+00
-6.08058237e-02 -7.14702129e-01 2.34930351e-01 -7.16147423e-01
-1.19487417e+00 7.92358100e-01 -1.34891704e-01 -5.86176336e-01
-4.16633695e-01 -1.03740895e+00 -5.47161460e-01 8.72117400e-01
-1.67747223e+00 -1.04536796e+00 2.34702498e-01 3.30055028e-01
6.19421184e-01 -3.07688564e-01 1.42715251e+00 3.25094014e-01
-5.48080504e-01 8.34541023e-01 1.88951954e-01 5.30853689e-01
4.96568352e-01 -1.10230243e+00 8.61501038e-01 7.02689290e-01
5.08887172e-01 1.11412573e+00 3.75178069e-01 -3.36135238e-01
-1.27359271e+00 -9.20519412e-01 1.41304648e+00 -4.55929548e-01
1.06723976e+00 -6.75267100e-01 -1.03863502e+00 1.18674588e+00
5.63013136e-01 6.64909258e-02 8.88892770e-01 4.62187201e-01
-7.75313437e-01 1.17922843e-01 -5.11340797e-01 6.37918770e-01
6.30663455e-01 -1.11301422e+00 -7.42642820e-01 1.62205204e-01
9.44770813e-01 -1.30195007e-01 -7.43704140e-01 2.10836545e-01
6.53011501e-01 -5.11526167e-01 9.76567447e-01 -9.03825283e-01
4.34229165e-01 -2.24960111e-02 -9.94696245e-02 -1.04376376e+00
-3.84108394e-01 -1.98980063e-01 -4.51912105e-01 8.68307412e-01
7.34077454e-01 -7.98735321e-01 7.87564456e-01 3.51589739e-01
-6.76390529e-02 -9.35798466e-01 -3.71288955e-01 -7.82142997e-01
5.55878758e-01 -5.36140144e-01 4.06953365e-01 1.09898984e+00
1.66870669e-01 4.40656036e-01 -1.41382933e-01 -7.26319402e-02
1.94753096e-01 4.43118028e-02 4.00298178e-01 -1.01451957e+00
-4.26410347e-01 -6.09747827e-01 -4.48690325e-01 -1.33683372e+00
6.05456531e-01 -1.33491516e+00 -2.68133551e-01 -1.72295630e+00
2.04571649e-01 -3.67002696e-01 -7.10283935e-01 7.91655540e-01
1.23856068e-01 3.79122019e-01 9.53608751e-02 1.70509890e-01
-4.49170649e-01 3.42153281e-01 8.18044543e-01 -4.12657797e-01
-5.03513850e-02 -3.20157737e-01 -6.13818467e-01 6.18395925e-01
1.32874775e+00 -4.79170173e-01 -4.62956905e-01 -1.08227837e+00
4.32131708e-01 -4.19984281e-01 2.97247231e-01 -7.97511637e-01
2.89887488e-01 4.59135383e-01 7.02707589e-01 -5.27288795e-01
4.11179960e-01 -6.39707148e-01 -3.58490765e-01 5.10029078e-01
-8.14164281e-01 5.71589649e-01 6.32055342e-01 2.57607937e-01
-3.53161633e-01 -3.39699745e-01 3.80495399e-01 -7.97514319e-02
-5.37609458e-01 1.92708567e-01 -5.83071053e-01 -4.51492637e-01
5.76918304e-01 -3.82668409e-03 -1.59776181e-01 -3.41279954e-01
-6.87888265e-01 -9.15219113e-02 5.13659120e-01 8.34205747e-01
7.84807622e-01 -1.29705942e+00 -5.82770169e-01 4.16576475e-01
-1.09305754e-01 -5.41338146e-01 -1.78254530e-01 4.54021871e-01
-6.39583647e-01 9.07154858e-01 -1.52470991e-01 -2.59658992e-01
-1.25707507e+00 6.00600362e-01 3.27242404e-01 -5.12384593e-01
-7.35337377e-01 9.35283780e-01 -1.30290121e-01 -7.61333168e-01
3.42817754e-01 -4.94613796e-01 -2.46597439e-01 -1.84235930e-01
7.32809067e-01 3.54975462e-02 3.96578126e-02 -3.94850910e-01
-4.00627196e-01 3.29314440e-01 -6.53877914e-01 -2.37485036e-01
1.30980027e+00 1.95302084e-01 -3.89539748e-01 7.92673767e-01
1.75591695e+00 -8.01642090e-02 -8.60850871e-01 -2.57115871e-01
1.87513709e-01 -1.01326756e-01 3.85157257e-01 -4.83066499e-01
-7.95049846e-01 1.09660590e+00 4.18900907e-01 3.03137243e-01
8.35434794e-01 -8.16391110e-02 9.38408434e-01 8.35533500e-01
-1.88747674e-01 -1.18080986e+00 3.39732058e-02 1.13055134e+00
6.82130337e-01 -1.05421281e+00 5.11355959e-02 7.37797618e-02
-1.88098177e-01 1.58944774e+00 4.57556695e-01 -5.74545264e-01
9.36205983e-01 7.50194848e-01 1.32492676e-01 -1.91189393e-01
-1.08655202e+00 7.63592049e-02 3.21918458e-01 1.64921522e-01
1.21625829e+00 -4.76240404e-02 -4.08024251e-01 3.19115222e-01
-2.53595859e-01 -7.38819242e-02 2.82738268e-01 1.07743466e+00
-2.40843549e-01 -1.67010427e+00 4.89069298e-02 3.14513117e-01
-5.41047752e-01 -4.73951280e-01 -3.63265276e-01 6.25906110e-01
-3.27883929e-01 6.01049244e-01 5.05499363e-01 -2.92424440e-01
5.26045635e-02 2.99882025e-01 4.50437099e-01 -8.09615016e-01
-5.61850607e-01 -1.10544972e-01 1.87695236e-03 -5.57456136e-01
-3.15690935e-01 -3.23240995e-01 -1.50076652e+00 -5.25795817e-01
-8.09424669e-02 -2.76998840e-02 9.05333817e-01 6.72079682e-01
5.10385633e-01 6.74592435e-01 2.54851729e-01 -7.01719582e-01
-4.85824198e-01 -9.67983425e-01 -9.98647735e-02 -9.40499753e-02
5.08469880e-01 -2.76476592e-01 -1.18842959e-01 -1.30936548e-01] | [10.73326587677002, 8.17473030090332] |
dc232366-10a6-4063-b424-39dcb3f4219e | simulating-counterfactuals | 2306.15328 | null | https://arxiv.org/abs/2306.15328v1 | https://arxiv.org/pdf/2306.15328v1.pdf | Simulating counterfactuals | Counterfactual inference considers a hypothetical intervention in a parallel world that shares some evidence with the factual world. If the evidence specifies a conditional distribution on a manifold, counterfactuals may be analytically intractable. We present an algorithm for simulating values from a counterfactual distribution where conditions can be set on both discrete and continuous variables. We show that the proposed algorithm can be presented as a particle filter leading to asymptotically valid inference. The algorithm is applied to fairness analysis in credit scoring. | ['Matti Vihola', 'Santtu Tikka', 'Juha Karvanen'] | 2023-06-27 | null | null | null | null | ['fairness', 'fairness', 'counterfactual-inference'] | ['computer-vision', 'miscellaneous', 'miscellaneous'] | [-5.21032922e-02 3.38108689e-01 -3.93105447e-01 -2.74576306e-01
-4.78787839e-01 -4.29208010e-01 9.26813543e-01 -1.89177692e-01
-6.72355831e-01 1.59927213e+00 1.48385167e-01 -7.62954950e-01
-3.01645160e-01 -1.03998160e+00 -8.08775246e-01 -5.77617228e-01
-4.38658565e-01 7.39293694e-01 -4.19472277e-01 1.28218397e-01
6.21473014e-01 4.26875442e-01 -1.38471329e+00 -1.48189425e-01
1.00345755e+00 4.39614356e-01 -4.91741486e-02 8.76505792e-01
3.31767052e-02 5.99413574e-01 -9.74055827e-01 -7.37130463e-01
6.18879914e-01 -4.87344474e-01 -6.78948879e-01 -2.00020462e-01
2.28961706e-01 -4.92240399e-01 -1.48522839e-01 1.63235533e+00
4.74827975e-01 4.15711492e-01 7.83746660e-01 -1.29036391e+00
-4.59878653e-01 6.64924622e-01 -3.48994046e-01 5.20662479e-02
5.55327713e-01 2.94266373e-01 8.69203687e-01 -2.67568946e-01
4.09301460e-01 1.50358522e+00 4.80016708e-01 3.28464925e-01
-1.47147477e+00 -7.20175624e-01 6.90396056e-02 2.63725400e-01
-9.41326857e-01 -5.51639087e-02 6.84721053e-01 -1.46469444e-01
5.99530220e-01 3.64392102e-01 1.12272501e+00 1.05656385e+00
8.36083472e-01 3.83771539e-01 1.68550360e+00 -5.51127434e-01
1.02438962e+00 -5.87408729e-02 8.13987255e-02 3.38423193e-01
8.40952814e-01 1.12352204e+00 -4.36322808e-01 -7.77190149e-01
5.59756696e-01 -2.00508386e-01 -2.85624087e-01 -3.47463906e-01
-1.05341744e+00 1.24014723e+00 -8.07628259e-02 -2.77090281e-01
-9.00675535e-01 3.33992928e-01 2.13481009e-01 4.92838025e-01
2.75967658e-01 7.33215094e-01 -5.70858121e-02 -1.47175089e-01
-9.87271190e-01 9.48290586e-01 1.37182939e+00 4.01679516e-01
1.95096478e-01 7.31480960e-03 -4.77666616e-01 4.97860610e-02
3.74146909e-01 1.06266344e+00 3.38062674e-01 -1.56494045e+00
3.90385151e-01 -3.58916223e-01 1.13512969e+00 -4.24667627e-01
-2.11441249e-01 -2.51901358e-01 -3.91098648e-01 8.56609523e-01
8.50334167e-01 -6.08740985e-01 -4.95019048e-01 1.61175275e+00
3.56068134e-01 6.82508171e-01 4.15374897e-02 8.83429527e-01
-2.07721069e-01 5.01495659e-01 1.35486171e-01 -7.81499803e-01
1.07369435e+00 5.37011474e-02 -1.08546281e+00 5.67038022e-02
-7.07689300e-02 -6.49543822e-01 8.98904681e-01 4.91106927e-01
-1.22477102e+00 -7.78729990e-02 -1.19875431e+00 7.23863482e-01
-2.77836947e-03 -6.84291542e-01 5.76008797e-01 1.22573149e+00
-8.87301207e-01 8.97141933e-01 -5.17884374e-01 1.53637841e-01
5.03880560e-01 1.69623196e-01 2.89140493e-01 2.93799073e-01
-1.46805179e+00 1.40621948e+00 4.85810578e-01 -3.82636413e-02
-1.09025133e+00 -6.10643983e-01 -4.40155506e-01 2.14155823e-01
3.02559763e-01 -1.00060272e+00 1.23110628e+00 -7.06230164e-01
-1.98374951e+00 3.45900893e-01 3.92352156e-02 -1.14535534e+00
1.08530951e+00 2.19218396e-02 -5.11683404e-01 5.39900661e-02
2.43283257e-01 -1.06881529e-01 9.14842725e-01 -1.07443345e+00
-4.38763916e-01 -4.01509255e-01 2.57714659e-01 5.12477219e-01
4.67050463e-01 -1.49910614e-01 8.10554028e-01 -3.61378223e-01
-2.05115214e-01 -8.31195235e-01 -4.25103009e-01 -1.21879049e-01
-5.44723570e-01 6.66404590e-02 1.47261888e-01 -2.58992761e-01
8.36454093e-01 -1.39339864e+00 -6.99677467e-01 4.32371885e-01
-2.71540761e-01 -1.86425984e-01 3.27964634e-01 1.42104700e-01
-9.08289999e-02 2.78905556e-02 -1.99790686e-01 2.75549680e-01
5.18816471e-01 1.60237812e-02 -6.17012084e-01 1.15019763e+00
-3.78645211e-01 7.86700189e-01 -1.15059948e+00 -4.41807657e-01
3.97602588e-01 -2.06037715e-01 -4.70691383e-01 2.54615486e-01
-1.89638957e-01 -1.73591807e-01 -4.11808550e-01 -5.24073541e-02
1.01187873e+00 2.04979748e-01 5.82847774e-01 5.26756346e-01
-2.41806075e-01 3.74002159e-01 -1.42878664e+00 1.13478029e+00
-3.18911850e-01 2.94887900e-01 -2.97134191e-01 -1.14610243e+00
6.02526188e-01 4.93119061e-01 1.06221708e-02 -3.85296792e-01
4.36045438e-01 -8.78145248e-02 2.22414210e-01 -2.43099153e-01
4.17975664e-01 -1.42456055e+00 -1.07654437e-01 9.77786779e-01
-1.20554864e-01 -4.35192108e-01 -1.43939599e-01 -8.87269303e-02
7.32492387e-01 -2.59025469e-02 9.25258875e-01 -7.64752448e-01
1.43868580e-01 -6.49601072e-02 4.88536835e-01 1.53517199e+00
-6.55993342e-01 1.71772853e-01 6.05766594e-01 -3.21297020e-01
-6.96379185e-01 -1.77172685e+00 -5.01116455e-01 2.36156151e-01
2.23982409e-01 5.94541371e-01 -6.52661085e-01 -7.31496096e-01
4.21273530e-01 1.60582745e+00 -7.91023672e-01 -1.63715556e-01
-1.74284607e-01 -1.07727313e+00 2.64328986e-01 7.63449296e-02
6.51025951e-01 -8.55742216e-01 -8.66800606e-01 2.54213870e-01
1.95939206e-02 -3.61518592e-01 -1.92540213e-01 -2.27273032e-01
-9.14464831e-01 -1.23189807e+00 -4.92550671e-01 2.21415594e-01
2.34768108e-01 -1.23050399e-01 1.20917726e+00 -4.64143157e-01
2.51164675e-01 1.52890980e-01 4.19362813e-01 -8.91816616e-01
-4.70313162e-01 -1.04184198e+00 3.49496126e-01 -5.11147827e-02
5.17464936e-01 -4.02223319e-01 -7.47851908e-01 -7.57720396e-02
-2.75007993e-01 -2.27690831e-01 -9.99496803e-02 1.20320308e+00
2.46952444e-01 2.99009293e-01 7.70080507e-01 -1.11154592e+00
1.09991741e+00 -6.63509130e-01 -1.07981122e+00 3.61806840e-01
-6.68059707e-01 1.67180657e-01 7.23953962e-01 -3.28097105e-01
-1.89181828e+00 -2.81813890e-01 3.55509192e-01 5.25149330e-02
-2.10361332e-01 2.78723657e-01 -2.97282249e-01 2.81176150e-01
7.37111568e-01 -1.22528099e-01 4.49649766e-02 -1.41248718e-01
8.37605834e-01 3.87792319e-01 6.17268562e-01 -1.06291914e+00
5.55546582e-01 9.06601906e-01 3.72199655e-01 -3.42396706e-01
-7.49940515e-01 2.31084466e-01 -3.29518378e-01 -3.41438383e-01
5.00236094e-01 -4.29013193e-01 -1.20372474e+00 -8.41308571e-03
-1.06468916e+00 -1.72840968e-01 -8.37677479e-01 1.09276700e+00
-1.19708157e+00 2.66872376e-01 -1.20971270e-01 -1.47366142e+00
1.76621065e-01 -6.25985801e-01 2.43502483e-01 3.35159391e-01
-3.72186542e-01 -1.41680646e+00 5.18808901e-01 -1.01842396e-01
4.78822663e-02 3.61745864e-01 4.22216922e-01 -5.40390491e-01
-5.52517653e-01 -2.79483825e-01 1.93106011e-01 -2.46395871e-01
-2.16955513e-01 -5.58069497e-02 -9.96421218e-01 -2.75397986e-01
6.58567607e-01 5.97494468e-02 4.02167559e-01 1.00219309e+00
7.56197393e-01 -7.35263228e-01 -1.75765947e-01 3.25397879e-01
1.40672517e+00 4.89673615e-01 4.96384591e-01 4.75660533e-01
-3.75067800e-01 6.30104125e-01 9.88324702e-01 7.78476954e-01
9.90671739e-02 2.90436119e-01 3.40187699e-01 3.87299836e-01
5.92143655e-01 -3.65259290e-01 2.96806127e-01 -2.97997326e-01
-6.18950315e-02 -6.34802431e-02 -7.00291693e-01 4.51836675e-01
-1.78999269e+00 -1.82145441e+00 6.53595999e-02 2.57293534e+00
6.78548694e-01 3.21584105e-01 1.68440178e-01 -4.60366458e-02
1.21269429e+00 -9.25676227e-02 -7.18019426e-01 -8.08470905e-01
-2.86016520e-02 4.49811727e-01 7.57432640e-01 8.62249434e-01
-7.16730177e-01 2.26837456e-01 8.13791561e+00 5.55778682e-01
-7.19482422e-01 2.68656075e-01 5.74265122e-01 -4.55143005e-01
-5.46679437e-01 4.01904434e-01 -2.92637140e-01 9.15788770e-01
1.15944660e+00 -1.36859882e+00 1.94587857e-01 4.75673318e-01
5.60715497e-01 -5.81432045e-01 -1.06949854e+00 5.39059162e-01
-4.66893166e-01 -1.35694921e+00 -4.52846661e-02 1.54981598e-01
9.14198875e-01 -3.07411075e-01 2.43453428e-01 8.28982666e-02
1.35449255e+00 -1.01804376e+00 8.58385563e-01 7.17469215e-01
5.03506541e-01 -1.01958346e+00 9.20498669e-01 6.37062848e-01
-2.57423490e-01 -1.22354001e-01 -5.36917806e-01 -6.57914519e-01
4.72260177e-01 5.77344954e-01 -8.49568248e-01 5.26645839e-01
-3.15138847e-02 1.50493756e-01 4.34306622e-01 1.26831138e+00
-3.96310210e-01 6.76661432e-01 -2.24727511e-01 -3.36767465e-01
8.71676058e-02 -6.07939065e-01 7.44687915e-01 8.78172278e-01
4.10168797e-01 2.41786525e-01 -4.32468913e-02 1.22717571e+00
2.74040818e-01 -1.78838730e-01 -8.21285963e-01 6.55359209e-01
7.62305200e-01 3.48096043e-01 -6.92301273e-01 -5.57813287e-01
-3.13525230e-01 5.43856144e-01 1.15667507e-01 6.40517354e-01
-9.51815724e-01 -6.84978887e-02 7.73836851e-01 -2.41217822e-01
-1.74271315e-01 4.98006612e-01 -7.75320113e-01 -1.24322855e+00
-3.01509351e-01 -6.38715446e-01 7.65628874e-01 -5.34788966e-01
-1.45456636e+00 -1.95899472e-01 3.19643140e-01 -1.11927426e+00
-3.13512564e-01 -5.05930781e-01 -1.10934889e+00 1.14816797e+00
-1.23398352e+00 -1.08018756e-01 6.14053130e-01 5.10432124e-01
-1.58568338e-01 -2.85280347e-01 7.90078282e-01 -5.26533723e-01
-2.93697655e-01 1.54766977e-01 3.81936938e-01 -3.98546338e-01
2.48410985e-01 -1.74772859e+00 3.17005545e-01 1.04363286e+00
1.00842625e-01 7.92852819e-01 1.40900075e+00 -7.32765377e-01
-9.81558561e-01 -8.02174926e-01 5.00481129e-01 -3.40178281e-01
8.74541640e-01 8.22073817e-02 -5.54978907e-01 6.48918331e-01
3.46784949e-01 -1.80148140e-01 5.05039930e-01 2.89456218e-01
9.51255113e-02 2.23966807e-01 -1.67440045e+00 8.90573382e-01
6.82345510e-01 -5.50921679e-01 -1.33590579e+00 3.76500040e-01
1.33439869e-01 -3.35346431e-01 -5.62968135e-01 -2.34274492e-01
7.34622717e-01 -1.11203289e+00 1.01693177e+00 -1.00162148e+00
7.81697556e-02 -1.12771735e-01 1.69861019e-02 -1.69322085e+00
-2.02981099e-01 -1.14259541e+00 -1.52708143e-01 5.45649290e-01
-5.95772564e-02 -1.20067859e+00 6.96736991e-01 9.33273852e-01
6.43318415e-01 -1.74490869e-01 -1.64225769e+00 -1.05136895e+00
6.26313865e-01 -6.65051043e-01 1.14258552e+00 8.15606713e-01
6.59903407e-01 -1.14799753e-01 -2.11474434e-01 2.59191602e-01
1.43802321e+00 7.11541474e-01 3.75640154e-01 -1.10394347e+00
-3.86231005e-01 -4.31018978e-01 -2.49856517e-01 -6.77401304e-01
7.10314870e-01 -7.20052540e-01 1.10648647e-02 -9.62794542e-01
4.54180509e-01 -3.34525526e-01 -3.22129965e-01 -3.94463509e-01
-4.54115182e-01 -3.51799995e-01 2.58953154e-01 -3.24017882e-01
-2.91246265e-01 7.15971649e-01 1.21814179e+00 1.45461515e-01
2.68036872e-01 2.88555622e-01 -6.09832168e-01 8.78941357e-01
1.02247739e+00 -4.86613363e-01 -3.14398527e-01 4.87054169e-01
3.43805589e-02 8.24681580e-01 6.08862817e-01 -6.18141592e-01
-7.18430011e-03 -7.75376856e-01 2.27823019e-01 -5.40671527e-01
1.34221524e-01 -6.01396382e-01 5.40604889e-01 9.50007200e-01
-4.42271709e-01 -2.87705570e-01 -5.03415614e-02 9.35504317e-01
9.99551341e-02 -6.26695335e-01 9.67065871e-01 -2.77092785e-01
-1.50179207e-01 -8.78335685e-02 -5.12871861e-01 2.61917353e-01
1.13304913e+00 -3.34670022e-02 -6.24405146e-01 -7.00882554e-01
-7.95228779e-01 2.86329627e-01 6.08974278e-01 -3.05021703e-01
6.44936442e-01 -1.36847270e+00 -9.18162525e-01 -1.51680574e-01
-6.47378206e-01 -8.46826494e-01 2.10292608e-01 2.92099476e-01
-5.10446072e-01 5.82780004e-01 -4.69075263e-01 -6.30205050e-02
-7.02426434e-01 6.43263698e-01 6.80577159e-01 -2.91244090e-01
-3.39048982e-01 3.32502156e-01 1.10561013e-01 -1.55479610e-01
-3.53655994e-01 4.62601706e-02 -3.92443836e-02 -2.00193852e-01
7.58820057e-01 8.70640635e-01 -5.34739435e-01 -2.44246736e-01
5.00515923e-02 -1.30448490e-01 2.66850680e-01 -7.30372369e-01
1.03224874e+00 -3.51978153e-01 1.12179354e-01 5.71953833e-01
7.48116493e-01 -1.32667780e-01 -1.50209844e+00 1.69906884e-01
1.00117333e-01 -1.14115489e+00 -1.84597466e-02 -7.88791955e-01
-4.92890179e-01 5.93953669e-01 5.47848880e-01 5.53914309e-01
5.91657460e-01 -5.19398451e-01 4.65876609e-02 2.69419730e-01
8.43751788e-01 -1.38956702e+00 -6.51580274e-01 3.57344970e-02
8.33046913e-01 -1.14953268e+00 5.24750412e-01 -1.35724954e-02
-5.33101201e-01 8.36562634e-01 1.22823622e-02 -6.38931513e-01
1.03514194e+00 3.44033718e-01 -5.72525449e-02 -1.30922962e-02
-8.73679519e-01 2.29555126e-02 1.98170729e-02 8.69054437e-01
2.52830803e-01 7.86761403e-01 -8.79742205e-01 6.20642543e-01
-7.46632814e-01 3.06634218e-01 1.07136595e+00 4.40335244e-01
-3.77735764e-01 -5.79214990e-01 -8.64126146e-01 7.42982209e-01
-7.46652782e-01 -6.76011071e-02 2.76422292e-01 8.42118859e-01
-2.19974667e-01 1.00934517e+00 2.10170448e-01 5.03121972e-01
2.43675664e-01 1.48770377e-01 6.18290722e-01 -3.14053327e-01
-1.91523626e-01 -3.23860556e-01 1.57040954e-01 -8.44273388e-01
-3.50430340e-01 -1.15613699e+00 -1.03401721e+00 -7.70511091e-01
-3.72784317e-01 5.32513142e-01 2.12405339e-01 8.97483408e-01
-2.34511733e-01 1.58520326e-01 6.02657139e-01 -5.22539377e-01
-1.40304661e+00 -7.01386273e-01 -1.23133683e+00 2.82926351e-01
3.98262262e-01 -9.12593424e-01 -6.27973080e-01 -2.61167228e-01] | [8.326698303222656, 5.511471748352051] |
87ca8ab7-abc2-47ab-b517-c9768f972eb2 | a-segmental-framework-for-fully-unsupervised | 1606.06950 | null | http://arxiv.org/abs/1606.06950v2 | http://arxiv.org/pdf/1606.06950v2.pdf | A segmental framework for fully-unsupervised large-vocabulary speech recognition | Zero-resource speech technology is a growing research area that aims to
develop methods for speech processing in the absence of transcriptions,
lexicons, or language modelling text. Early term discovery systems focused on
identifying isolated recurring patterns in a corpus, while more recent
full-coverage systems attempt to completely segment and cluster the audio into
word-like units---effectively performing unsupervised speech recognition. This
article presents the first attempt we are aware of to apply such a system to
large-vocabulary multi-speaker data. Our system uses a Bayesian modelling
framework with segmental word representations: each word segment is represented
as a fixed-dimensional acoustic embedding obtained by mapping the sequence of
feature frames to a single embedding vector. We compare our system on English
and Xitsonga datasets to state-of-the-art baselines, using a variety of
measures including word error rate (obtained by mapping the unsupervised output
to ground truth transcriptions). Very high word error rates are reported---in
the order of 70--80% for speaker-dependent and 80--95% for speaker-independent
systems---highlighting the difficulty of this task. Nevertheless, in terms of
cluster quality and word segmentation metrics, we show that by imposing a
consistent top-down segmentation while also using bottom-up knowledge from
detected syllable boundaries, both single-speaker and multi-speaker versions of
our system outperform a purely bottom-up single-speaker syllable-based
approach. We also show that the discovered clusters can be made less speaker-
and gender-specific by using an unsupervised autoencoder-like feature extractor
to learn better frame-level features (prior to embedding). Our system's
discovered clusters are still less pure than those of unsupervised term
discovery systems, but provide far greater coverage. | ['Sharon Goldwater', 'Aren Jansen', 'Herman Kamper'] | 2016-06-22 | null | null | null | null | ['unsupervised-speech-recognition'] | ['speech'] | [ 2.96477079e-01 2.62359619e-01 -2.41289705e-01 -5.85999668e-01
-1.39363408e+00 -5.07096052e-01 5.56846797e-01 1.19939856e-01
-5.86846113e-01 2.93798447e-01 4.03192967e-01 -3.61972570e-01
1.66856796e-01 -3.18139315e-01 -3.87219191e-01 -6.86034203e-01
-1.64519459e-01 6.73060238e-01 1.54278234e-01 -2.21672490e-01
9.87584293e-02 -2.83861160e-02 -1.76996732e+00 1.82478085e-01
3.37439090e-01 5.22688687e-01 2.95062989e-01 8.32938612e-01
-2.43789673e-01 1.80288300e-01 -7.52476037e-01 -1.31259441e-01
-4.41614874e-02 -4.94905382e-01 -8.22654486e-01 3.00698131e-01
2.05847323e-01 2.19093874e-01 -8.49379823e-02 9.75106597e-01
6.20259166e-01 1.48344308e-01 5.05732536e-01 -7.08556771e-01
-4.08345699e-01 1.02254367e+00 -6.42380714e-02 3.36904645e-01
4.17929977e-01 -1.66333839e-01 1.31184459e+00 -9.64447796e-01
4.25617695e-01 1.27213562e+00 5.29855371e-01 5.94169080e-01
-1.39951074e+00 -5.92592061e-01 1.50536537e-01 2.12408835e-03
-1.86102092e+00 -1.15385664e+00 6.91181183e-01 -5.56647599e-01
1.53293216e+00 5.46201885e-01 3.54639113e-01 1.17327213e+00
-2.64957100e-01 5.72747469e-01 7.72872269e-01 -9.87748802e-01
3.56704503e-01 2.59583592e-01 2.13692069e-01 5.63908994e-01
-2.22555786e-01 -5.16761579e-02 -6.88487589e-01 -4.43222634e-02
4.58128005e-01 -6.77428246e-01 -1.89272940e-01 9.09339935e-02
-1.22504723e+00 9.94863987e-01 -2.67915010e-01 9.04204786e-01
-2.43211806e-01 -1.11099906e-01 4.70440626e-01 1.68356150e-01
6.82211220e-01 3.31940562e-01 -5.05747020e-01 -4.62084979e-01
-1.50705791e+00 -1.04599662e-01 9.65580583e-01 9.78786111e-01
7.87268639e-01 5.04685819e-01 -7.17939883e-02 1.20628762e+00
5.64820945e-01 2.96212614e-01 1.10496759e+00 -6.52984083e-01
2.35447928e-01 2.96162651e-03 -3.81163687e-01 -6.29629493e-01
-2.18017787e-01 -3.44812244e-01 -3.09388429e-01 -3.30017120e-01
2.20827311e-01 -4.67038304e-02 -1.26328182e+00 1.77069879e+00
2.05492496e-01 1.61132783e-01 9.12527367e-02 5.01540601e-01
6.01776898e-01 9.68088984e-01 -1.79855283e-02 -5.72282076e-01
1.57976377e+00 -7.93082535e-01 -8.56240034e-01 -3.93343449e-01
4.37823176e-01 -7.74572372e-01 9.85197425e-01 3.96488965e-01
-1.07861722e+00 -5.62933207e-01 -1.04658175e+00 1.59515858e-01
-5.09603858e-01 1.11901246e-01 2.31016293e-01 1.30259407e+00
-1.16955006e+00 2.60133713e-01 -8.51192713e-01 -5.68901956e-01
-6.94288015e-02 4.41790789e-01 -2.99710691e-01 3.45076442e-01
-1.15435290e+00 6.74643219e-01 4.22983348e-01 -2.30196133e-01
-8.70860338e-01 -5.87765992e-01 -1.03322768e+00 5.84718809e-02
3.26895744e-01 -8.95852372e-02 1.45594740e+00 -7.76309550e-01
-1.82398856e+00 8.47677886e-01 -8.26097608e-01 -4.41257954e-01
-3.99142981e-01 1.45601884e-01 -7.59125590e-01 1.60469994e-01
1.14215448e-01 5.17549753e-01 9.78126585e-01 -1.18722320e+00
-6.87367797e-01 -2.77702212e-01 -5.79246163e-01 1.79368004e-01
-5.42814314e-01 5.44261575e-01 -7.13496804e-01 -7.48308718e-01
2.49009579e-01 -9.49657083e-01 -1.21757388e-01 -9.38694775e-01
-4.47265118e-01 -4.79749590e-01 6.58926368e-01 -9.05089080e-01
1.64586890e+00 -2.33742785e+00 1.15431130e-01 3.12678844e-01
-1.30072072e-01 1.70336276e-01 7.05785304e-02 3.30160230e-01
-2.36481056e-01 3.59398365e-01 -3.53380293e-01 -8.18514168e-01
2.30858281e-01 4.78001505e-01 -2.38829806e-01 2.84319282e-01
2.17861235e-01 5.77943146e-01 -6.75112009e-01 -5.42785883e-01
3.00567508e-01 5.93779385e-01 -3.97065222e-01 -6.99890479e-02
2.73447800e-02 1.85127810e-01 2.41755888e-01 5.56437671e-01
2.89105084e-02 4.17708576e-01 2.36885801e-01 2.85590380e-01
-3.55610371e-01 9.05919790e-01 -1.28074992e+00 1.68241501e+00
-5.38144350e-01 7.75282562e-01 2.44439349e-01 -1.29778409e+00
8.99549425e-01 7.91897416e-01 2.96204060e-01 -3.69939446e-01
-8.04070476e-03 4.00204360e-01 2.22202912e-01 -2.59308159e-01
6.41333640e-01 -4.10790682e-01 -3.62385839e-01 2.02665254e-01
6.51589155e-01 -2.20015079e-01 2.21040294e-01 -1.56265005e-01
9.03904438e-01 -4.20063645e-01 1.77457005e-01 -3.12135547e-01
4.11614954e-01 -9.61237326e-02 5.12187779e-01 6.89206958e-01
-1.65334284e-01 8.79207015e-01 1.29588261e-01 1.32388264e-01
-9.27002907e-01 -9.91628468e-01 -3.82587016e-01 1.58484018e+00
-4.38562393e-01 -7.05170929e-01 -9.95303452e-01 -2.14493901e-01
-3.26022625e-01 9.66597080e-01 -3.03978533e-01 9.19323936e-02
-5.76599121e-01 -7.24200010e-01 1.03844094e+00 3.27594101e-01
-1.97207302e-01 -1.13123894e+00 -9.63247493e-02 5.85508466e-01
-2.39556342e-01 -1.08318365e+00 -5.55930793e-01 6.89260066e-01
-5.56022644e-01 -3.18392009e-01 -6.98952317e-01 -1.09337556e+00
2.85751462e-01 -1.74253523e-01 9.05814469e-01 -4.09215569e-01
-2.85630047e-01 4.26426679e-01 -4.63461041e-01 -4.53599721e-01
-6.65176332e-01 2.51984715e-01 4.27431673e-01 1.35252535e-01
6.83691800e-01 -3.51836443e-01 -6.36627339e-03 1.84353381e-01
-7.07420230e-01 -5.01546741e-01 5.01577735e-01 6.51811182e-01
5.30367255e-01 1.61810428e-01 6.83397770e-01 -5.12503505e-01
6.33130491e-01 -2.90556520e-01 -2.24905193e-01 -9.99075919e-03
-5.64895391e-01 1.32937189e-02 2.23979801e-01 -4.91800904e-01
-8.17869246e-01 2.26744667e-01 -4.79574144e-01 -3.04748923e-01
-6.11059546e-01 6.22215509e-01 -2.75856227e-01 3.51326376e-01
6.00427210e-01 5.80549955e-01 -8.06325451e-02 -5.68198383e-01
4.36958045e-01 1.27928174e+00 7.42601097e-01 -4.45411056e-01
5.89730382e-01 1.06230497e-01 -8.40793729e-01 -1.64839673e+00
-4.07736748e-01 -1.03039002e+00 -7.33642936e-01 -3.86010297e-02
1.02177405e+00 -1.00505114e+00 -8.59957188e-02 1.78188339e-01
-1.13017941e+00 -9.57353786e-02 -4.59443957e-01 6.89310193e-01
-3.99603486e-01 4.30161536e-01 -5.85894048e-01 -1.16821694e+00
-1.32636920e-01 -1.20500052e+00 1.33073759e+00 -1.01266406e-01
-6.24968410e-01 -9.15497065e-01 2.90640473e-01 5.30017972e-01
3.16804767e-01 -5.28711736e-01 8.11271966e-01 -1.22575867e+00
1.37965083e-01 -1.91048399e-01 4.65009451e-01 5.90569437e-01
1.42020866e-01 6.63099997e-03 -1.42822444e+00 -2.02738240e-01
2.21707486e-02 1.68485954e-01 7.24354148e-01 5.16853571e-01
7.61021972e-01 -3.12556654e-01 -3.08489919e-01 1.80693343e-01
9.17704463e-01 4.63624030e-01 4.50694919e-01 2.71335989e-03
6.10428989e-01 7.18214154e-01 1.35739848e-01 1.70964167e-01
3.16177696e-01 7.61783183e-01 -1.82079107e-01 1.46274626e-01
-1.65659025e-01 -7.17477798e-02 7.05483437e-01 1.68574870e+00
8.10613334e-02 -1.70214057e-01 -1.14705849e+00 1.18340993e+00
-1.54814410e+00 -1.00042939e+00 2.44515222e-02 2.39618731e+00
1.08186007e+00 2.66879290e-01 4.27148074e-01 4.75568861e-01
9.66029048e-01 9.47169289e-02 -3.50825489e-02 -5.94416380e-01
-6.88241376e-03 5.43595374e-01 2.60369450e-01 7.57448316e-01
-1.14770269e+00 1.24201751e+00 7.07546949e+00 1.08415926e+00
-1.11293411e+00 2.67022580e-01 3.39966834e-01 -1.40255883e-01
-1.45140111e-01 -2.75655137e-03 -1.00398588e+00 3.54222000e-01
1.79562020e+00 -7.82601386e-02 4.10937935e-01 8.00263643e-01
1.69901863e-01 4.80493605e-02 -1.16240585e+00 9.93376553e-01
3.19597512e-01 -1.10134852e+00 -2.45290071e-01 2.93033868e-01
4.14584070e-01 2.32495785e-01 -8.34149197e-02 6.05275631e-01
1.18815057e-01 -1.14710224e+00 1.04748642e+00 -3.80821675e-02
7.68407106e-01 -8.01205337e-01 5.35944641e-01 4.03906792e-01
-1.30064821e+00 1.78839356e-01 -2.22568452e-01 5.50445169e-02
2.47203544e-01 5.11218488e-01 -1.27480590e+00 4.79262292e-01
6.61922395e-01 2.04036906e-01 -4.03780550e-01 7.39790082e-01
7.75629580e-02 1.30390584e+00 -5.51894128e-01 -9.17739123e-02
2.06065848e-01 1.00905605e-01 7.56290615e-01 1.89108944e+00
2.61435091e-01 -1.18380904e-01 2.43530631e-01 4.96764213e-01
1.54512927e-01 3.66721541e-01 -4.94779885e-01 -2.83852249e-01
6.17247701e-01 1.02770519e+00 -8.00534904e-01 -4.16366488e-01
-3.95549297e-01 8.70672882e-01 -5.03816968e-03 2.87798911e-01
-4.51606929e-01 -6.21986628e-01 7.98824728e-01 2.42376812e-02
7.04996824e-01 -4.73682255e-01 -3.25461924e-01 -9.68642533e-01
-1.09820083e-01 -8.80638599e-01 7.99133331e-02 -3.84177089e-01
-1.11343634e+00 8.26491177e-01 1.27746299e-01 -8.10650706e-01
-8.09056520e-01 -5.00069559e-01 -5.16185582e-01 9.51542020e-01
-1.01625764e+00 -8.26862514e-01 5.60575366e-01 3.32662135e-01
1.02421749e+00 -5.17532885e-01 1.33622599e+00 3.10965478e-01
-6.53674185e-01 7.33642936e-01 1.32814586e-01 3.43216330e-01
5.87482274e-01 -1.30684006e+00 4.36979234e-01 9.08997118e-01
8.79904568e-01 7.95753360e-01 7.20983505e-01 -4.71434534e-01
-1.06860518e+00 -8.10501695e-01 1.37435448e+00 -5.71593225e-01
8.12998176e-01 -8.75724733e-01 -1.02222109e+00 6.34422958e-01
3.66711438e-01 -2.76404530e-01 1.08180964e+00 6.48652256e-01
-2.55596906e-01 8.34069252e-02 -7.15803802e-01 3.94443333e-01
8.05026174e-01 -9.99046981e-01 -1.02160394e+00 1.32614776e-01
1.01447260e+00 -2.52700108e-03 -7.65772641e-01 7.50672147e-02
4.02380228e-01 -6.13245249e-01 8.01387966e-01 -3.89950812e-01
-2.34953687e-01 -2.81333685e-01 -5.73420286e-01 -1.32748485e+00
-2.73342997e-01 -8.87103021e-01 2.29760349e-01 1.58787966e+00
8.36078942e-01 -3.05767626e-01 5.72998762e-01 1.74109563e-01
-4.98530716e-01 -3.33339423e-01 -1.51421988e+00 -9.13782895e-01
9.51291099e-02 -1.08401573e+00 3.25738013e-01 9.60174918e-01
3.61290962e-01 6.27691150e-01 -3.19129601e-02 4.06325608e-01
3.47054690e-01 -4.54984963e-01 4.63530272e-01 -1.12094104e+00
-3.67651820e-01 -6.07895553e-01 -5.19121647e-01 -8.47181857e-01
3.63387883e-01 -8.98470402e-01 6.98119462e-01 -1.14747310e+00
-2.85832316e-01 -1.11381836e-01 -3.06972951e-01 5.66354573e-01
9.61109251e-02 2.48706877e-01 -6.25038967e-02 1.41445264e-01
-3.83387864e-01 3.24064106e-01 2.05066070e-01 -1.37697592e-01
-4.48971868e-01 -1.21085741e-01 -6.39061034e-01 6.88584149e-01
7.01371193e-01 -6.52543247e-01 -2.24042878e-01 -2.22699195e-01
-2.85833389e-01 -1.37435302e-01 -1.05544962e-01 -1.03222346e+00
3.94555807e-01 1.40327364e-01 -8.19564760e-02 -4.56249982e-01
4.54999059e-01 -3.50882828e-01 2.11577807e-02 6.68027550e-02
-2.72385120e-01 -2.25573614e-01 3.37565601e-01 4.27587509e-01
-4.84582305e-01 -5.60722470e-01 4.83397633e-01 -1.03340380e-01
-8.02567124e-01 -2.55691648e-01 -1.03266799e+00 -1.16313867e-01
7.10784078e-01 -3.32806438e-01 2.21855804e-01 -4.04821426e-01
-9.35901642e-01 -2.40336582e-01 -1.80653157e-03 5.95096350e-01
3.62633556e-01 -1.02990544e+00 -8.53300810e-01 5.52014947e-01
2.02140257e-01 -1.46545008e-01 -1.81782350e-01 5.76776087e-01
-1.04689233e-01 7.03675270e-01 4.99620408e-01 -8.70861709e-01
-1.29139650e+00 2.56596178e-01 8.60539451e-02 -2.03154329e-03
-2.96713203e-01 1.13783073e+00 -5.00615267e-03 -6.76301003e-01
5.04781127e-01 -3.58927459e-01 -5.85440211e-02 3.63380820e-01
4.08314824e-01 2.99327783e-02 4.20159549e-01 -1.13031530e+00
-6.74139977e-01 3.76804382e-01 -3.83348987e-02 -8.27251911e-01
1.28528249e+00 -3.57997626e-01 1.79028660e-01 1.03269100e+00
1.17634654e+00 3.46657872e-01 -6.36108160e-01 -9.58192572e-02
3.28693658e-01 9.67317075e-02 3.62214237e-01 -4.86170411e-01
-4.81908739e-01 8.45587850e-01 4.99802798e-01 5.53770781e-01
7.21565783e-01 3.97065282e-01 5.78859568e-01 2.87801147e-01
1.68745458e-01 -1.44196582e+00 -2.29350343e-01 6.94704175e-01
6.02267921e-01 -1.19030488e+00 -3.78132939e-01 -3.19427013e-01
-6.21813357e-01 8.96808743e-01 -6.70887763e-03 1.77216649e-01
7.45021522e-01 2.99792200e-01 2.89563358e-01 -4.74663526e-02
-6.95166409e-01 -5.89606941e-01 4.69395638e-01 6.47320747e-01
6.42740667e-01 1.46038532e-01 3.98721109e-04 8.32461655e-01
-6.99905157e-01 -6.70152485e-01 3.14203858e-01 7.55862594e-01
-7.44582355e-01 -1.22561717e+00 -6.86974227e-01 2.09487006e-01
-5.70472538e-01 -3.49445701e-01 -3.11125785e-01 5.97946405e-01
3.40003073e-02 1.41489983e+00 3.61692578e-01 -5.26011229e-01
1.75661951e-01 8.82886648e-01 1.17250197e-01 -1.28752244e+00
-5.07479429e-01 8.35796654e-01 3.83303851e-01 -2.76426941e-01
-4.01547641e-01 -1.06428349e+00 -1.21293116e+00 2.75311768e-01
-5.93806922e-01 3.72486949e-01 1.15085280e+00 1.01459980e+00
1.10086843e-01 6.39866292e-01 5.12236178e-01 -1.04716444e+00
-3.34662914e-01 -1.27499747e+00 -6.94490731e-01 -4.32023266e-03
3.50935489e-01 -3.28706920e-01 -6.48259282e-01 4.22272891e-01] | [14.472715377807617, 6.653144836425781] |
1aae76bb-2554-4102-a6f4-78967c55a0a9 | dnn-filter-for-bias-reduction-in-distribution | 2211.04047 | null | https://arxiv.org/abs/2211.04047v3 | https://arxiv.org/pdf/2211.04047v3.pdf | DNN Filter for Bias Reduction in Distribution-to-Distribution Scan Matching | Distribution-to-distribution (D2D) point cloud registration techniques such as the Normal Distributions Transform (NDT) can align point clouds sampled from unstructured scenes and provide accurate bounds of their own solution error covariance -- an important feature for safety-of-life navigation tasks. D2D methods rely on the assumption of a static scene and are therefore susceptible to bias from range-shadowing, self-occlusion, moving objects, and distortion artifacts as the recording device moves between frames. Deep Learning-based approaches can achieve higher accuracy in dynamic scenes by relaxing these constraints, however, DNNs produce uninterpretable solutions which can be problematic from a safety perspective. In this paper, we propose a method of down-sampling LIDAR point clouds to exclude voxels that violate the assumption of a static scene and introduce error to the D2D scan matching process. Our approach uses a solution consistency filter -- identifying and suppressing voxels where D2D contributions disagree with local estimates from a PointNet-based registration network. Our results show that this technique provides significant benefits in registration accuracy, and is particularly useful in scenes containing dense foliage. | ['Jason Rife', 'Matthew McDermott'] | 2022-11-08 | null | null | null | null | ['point-cloud-registration'] | ['computer-vision'] | [ 3.22092801e-01 -1.77827835e-01 7.85476491e-02 -4.32332039e-01
-6.41705692e-01 -5.66827416e-01 6.05072379e-01 2.01462418e-01
-3.12756836e-01 6.79363132e-01 -1.27600044e-01 -5.76432832e-02
-2.66870856e-01 -1.09577453e+00 -7.75715590e-01 -5.21586120e-01
-8.83691981e-02 1.06618309e+00 4.95558113e-01 4.70427833e-02
2.05594286e-01 1.20030093e+00 -1.61040521e+00 -1.45704076e-01
8.30910146e-01 8.90651643e-01 1.17566593e-01 1.58563718e-01
-3.33414882e-01 2.19312549e-01 -6.04094923e-01 3.37259322e-02
8.29500675e-01 1.82280332e-01 -2.49380425e-01 -2.69175142e-01
1.04628372e+00 -4.73305285e-01 -1.15589060e-01 1.04788029e+00
5.05110145e-01 1.89051643e-01 6.29181743e-01 -1.35913694e+00
-1.27134606e-01 -1.12194590e-01 -7.95145988e-01 -2.52378970e-01
2.75867492e-01 1.91130564e-01 3.71110797e-01 -7.88735390e-01
8.03579807e-01 1.15759516e+00 1.38750744e+00 3.48895222e-01
-1.69062293e+00 -7.69729435e-01 1.68726310e-01 -2.00464994e-01
-1.79324102e+00 -4.27030295e-01 7.87777960e-01 -6.48760736e-01
1.02219105e+00 4.48582798e-01 7.69849002e-01 7.64974773e-01
3.94956768e-01 5.48047479e-03 8.66456926e-01 -4.61449996e-02
3.99516582e-01 -2.24211887e-01 -2.54409790e-01 1.83796421e-01
4.75832343e-01 3.56376380e-01 -5.38243592e-01 -3.82493317e-01
9.57385778e-01 -9.29340422e-02 -3.70844364e-01 -1.18733835e+00
-1.17892194e+00 6.04104400e-01 7.34150112e-01 -7.73217576e-03
-2.79903054e-01 2.65470684e-01 1.36267141e-01 1.83154881e-01
7.47064888e-01 4.01138127e-01 -4.63601410e-01 4.45721857e-02
-1.20642591e+00 5.61548114e-01 5.44012845e-01 1.07863176e+00
8.91718328e-01 -2.27599666e-02 2.02507213e-01 4.28527743e-01
3.80541056e-01 9.26726341e-01 -1.89548358e-01 -1.13936687e+00
3.50753516e-01 2.06268296e-01 1.71844959e-01 -1.21466458e+00
-5.35103679e-01 -3.56649846e-01 -9.39884067e-01 7.48366535e-01
3.47452968e-01 2.62618393e-01 -8.60988796e-01 1.45422244e+00
6.88457906e-01 3.29244137e-01 -1.90189525e-01 1.01012743e+00
6.45067275e-01 2.05776021e-01 -1.27748817e-01 -5.08162864e-02
6.14895880e-01 -5.09654451e-03 -5.08560658e-01 -1.08873501e-01
4.36165333e-01 -7.25268185e-01 8.23610902e-01 1.03813998e-01
-1.02828372e+00 -3.47254127e-01 -1.00134027e+00 2.18574107e-02
-3.38078737e-02 -6.33776844e-01 3.10631603e-01 4.60369498e-01
-1.16245461e+00 8.48816156e-01 -9.44925547e-01 -3.66056234e-01
5.63987076e-01 5.95685065e-01 -2.84454972e-01 -1.11644380e-01
-5.46582341e-01 1.20533788e+00 -1.04483560e-01 1.00359835e-01
-5.64950407e-01 -1.32905436e+00 -7.29221404e-01 -2.34187439e-01
7.65565485e-02 -7.72896826e-01 1.02943313e+00 -5.50262153e-01
-9.95270789e-01 9.03743446e-01 -2.91515112e-01 -4.12582517e-01
7.46843278e-01 -2.80268520e-01 -4.68733832e-02 -1.91953018e-01
3.16467345e-01 7.85003543e-01 7.25868404e-01 -1.52178371e+00
-2.90000856e-01 -7.51666129e-01 -3.23402762e-01 4.12955850e-01
4.52309370e-01 -5.23894012e-01 -1.70568153e-01 -2.10081726e-01
8.20385396e-01 -1.07707322e+00 -4.41770196e-01 7.10550606e-01
-2.23399982e-01 3.73435408e-01 1.00741100e+00 -3.01493078e-01
4.52090621e-01 -2.12860894e+00 -2.11289972e-01 6.08407795e-01
3.61541539e-01 -1.02593437e-01 -3.65232229e-02 1.13753140e-01
2.00535599e-02 2.16388684e-02 -3.80888373e-01 -4.28530663e-01
-1.71387315e-01 2.34064013e-01 -3.66432905e-01 9.46377754e-01
-6.37951866e-02 4.28101808e-01 -7.26271629e-01 -2.63811350e-01
6.86064363e-01 7.32826948e-01 -3.61189663e-01 -1.25643238e-01
-3.06791365e-01 7.26455510e-01 -2.42663115e-01 4.64248925e-01
1.35607862e+00 3.29925060e-01 -2.65211642e-01 -4.90942448e-01
-4.16120857e-01 3.57017875e-01 -1.35667086e+00 1.89251125e+00
-5.26891768e-01 8.18275273e-01 3.03625882e-01 -4.35293972e-01
1.04689443e+00 -1.64138228e-01 9.43795919e-01 -6.57726228e-01
-6.64170384e-02 2.65108377e-01 -7.95381367e-02 2.71887910e-02
5.35505474e-01 -1.97734967e-01 3.39066535e-01 6.89968020e-02
-4.33135182e-01 -9.78154600e-01 -5.76837540e-01 -8.73166770e-02
1.05976903e+00 2.27416471e-01 -5.51219620e-02 -5.08134902e-01
1.14875443e-01 2.35591561e-01 5.87940753e-01 6.56804204e-01
7.77278244e-02 9.13014293e-01 -5.55684865e-02 -6.10420346e-01
-1.16907477e+00 -1.34642732e+00 -6.43929541e-01 -9.77850556e-02
4.81468379e-01 -1.21307075e-01 -3.83265883e-01 -2.25539461e-01
5.58256447e-01 9.04138982e-01 -1.42445534e-01 -8.68712515e-02
-6.31296933e-01 -4.75180060e-01 3.20131660e-01 5.13186455e-01
4.25862402e-01 -2.84415156e-01 -8.10059369e-01 2.36166611e-01
-6.41164109e-02 -1.05882108e+00 -1.12033479e-01 2.53649741e-01
-1.15749383e+00 -9.58181202e-01 -4.69829857e-01 -2.87867367e-01
7.38217175e-01 4.12416190e-01 1.31986558e+00 -1.49248512e-02
-8.72378498e-02 4.15232927e-01 1.89842850e-01 -5.80657244e-01
-2.91186064e-01 -1.37787774e-01 2.76957333e-01 -5.73303938e-01
5.80391288e-01 -1.00037217e+00 -4.39326644e-01 5.01110494e-01
-3.87296736e-01 -3.42522226e-02 -1.71887025e-01 2.57913977e-01
1.09195256e+00 4.08995040e-02 -2.00202793e-01 -5.02642810e-01
1.95251748e-01 -1.34550646e-01 -1.29251885e+00 -2.73386955e-01
-5.13421059e-01 -2.25477628e-02 1.90132812e-01 -3.01409215e-01
-6.83853388e-01 3.92370731e-01 2.19089109e-02 -8.51401508e-01
-1.99518412e-01 9.65949073e-02 -3.64551157e-01 -5.70484579e-01
8.85022700e-01 -1.29291221e-01 6.21931069e-02 -3.78165603e-01
2.32513741e-01 2.31902316e-01 6.65934801e-01 -7.17187524e-01
1.24463522e+00 1.03902686e+00 6.81956232e-01 -1.07814980e+00
-5.48142910e-01 -4.52283144e-01 -1.01687205e+00 -4.09282148e-01
7.13817835e-01 -9.76947844e-01 -4.98889297e-01 3.62310290e-01
-1.41650891e+00 -3.76656860e-01 -6.41875565e-01 5.99797547e-01
-4.81010914e-01 3.04246783e-01 6.75319061e-02 -7.33389020e-01
-5.81704080e-02 -1.12538898e+00 1.27162278e+00 -1.29885622e-03
-3.49223256e-01 -7.41947472e-01 3.08263361e-01 -1.21563807e-01
5.68180263e-01 6.56458437e-01 5.94253182e-01 -1.10722408e-01
-9.72890317e-01 -2.08662963e-03 -9.16557312e-02 -4.51806262e-02
1.50102660e-01 1.03416622e-01 -1.19931686e+00 -2.81253994e-01
3.62827331e-01 2.63469160e-01 3.86363328e-01 7.42978811e-01
1.06193066e+00 1.30967796e-01 -5.77000499e-01 1.02448440e+00
1.64846814e+00 8.32369402e-02 6.65922403e-01 4.15261507e-01
8.62995327e-01 6.65066421e-01 6.24797940e-01 1.54669330e-01
4.03568029e-01 9.22736764e-01 9.12539661e-01 -2.04953104e-01
-2.06809461e-01 -1.01528503e-01 -1.15813971e-01 3.39758694e-01
-9.19021219e-02 -1.31563306e-01 -1.17010844e+00 3.88313949e-01
-1.60682750e+00 -9.16488767e-01 -8.51121187e-01 2.71195960e+00
3.94089818e-01 1.99805737e-01 -2.90601701e-01 -1.04530163e-01
5.66293359e-01 8.72800276e-02 -8.16485226e-01 -4.29154150e-02
-1.06407695e-01 2.75868267e-01 1.06456590e+00 6.90076947e-01
-8.58029842e-01 7.67792523e-01 6.17733431e+00 3.91188323e-01
-1.12081981e+00 1.47686437e-01 -3.65127437e-02 -2.60913193e-01
-5.85422039e-01 5.19251451e-02 -7.35763669e-01 1.99520737e-01
6.31081641e-01 1.29912436e-01 1.42765909e-01 9.07804012e-01
3.55523109e-01 -2.87679523e-01 -1.26460147e+00 1.24085271e+00
-1.33833503e-02 -1.46822619e+00 -2.64067113e-01 3.79678071e-01
5.45391023e-01 6.41755164e-01 -1.47253633e-01 -3.22191983e-01
3.46821070e-01 -8.41677129e-01 8.98682296e-01 4.09152478e-01
9.63699460e-01 -5.60010612e-01 4.48949099e-01 4.15663779e-01
-1.33008194e+00 4.96559858e-01 -7.07034588e-01 3.72823365e-02
4.41516340e-01 1.09457135e+00 -7.55110383e-01 4.38388169e-01
7.87077248e-01 5.43167770e-01 -1.82401702e-01 1.34340131e+00
9.14749131e-02 5.15518263e-02 -8.85572612e-01 3.75765830e-01
-1.55657932e-01 -3.98905575e-01 8.63322139e-01 6.08003616e-01
5.14849365e-01 -1.45129990e-02 5.01929335e-02 1.20662534e+00
2.84877002e-01 -3.95721227e-01 -1.07994449e+00 7.59145856e-01
7.02574372e-01 8.15962970e-01 -8.79933894e-01 3.87056656e-02
-2.60988414e-01 7.66713679e-01 -6.27908111e-02 1.22314878e-01
-7.83094466e-01 1.07383154e-01 1.23048389e+00 8.31863225e-01
1.05828211e-01 -7.13608027e-01 -6.85084999e-01 -9.91449893e-01
1.83588669e-01 -5.10926425e-01 -2.16164798e-01 -8.82171094e-01
-1.23311818e+00 4.12068009e-01 3.67645621e-01 -1.56316864e+00
1.62159987e-02 -4.42663610e-01 -4.38903421e-01 9.46405232e-01
-1.43152964e+00 -9.55846250e-01 -6.92628324e-01 6.43809319e-01
2.84804136e-01 2.29366630e-01 7.19481945e-01 4.80682373e-01
1.58947855e-01 1.61432400e-01 2.55552441e-01 -3.88050497e-01
6.16304755e-01 -1.01240051e+00 6.74893439e-01 7.35416830e-01
1.28269330e-01 4.23320949e-01 7.88769484e-01 -1.10669768e+00
-1.26763082e+00 -1.07286465e+00 5.37451923e-01 -8.37128520e-01
1.98434219e-01 -4.57329094e-01 -9.77739751e-01 5.20500124e-01
-4.50305700e-01 1.50831014e-01 3.15283597e-01 4.20327634e-02
-5.78152537e-02 -2.74014115e-01 -1.49452376e+00 4.26107913e-01
1.32397783e+00 -5.75410187e-01 -4.10289675e-01 4.60183322e-01
6.18489027e-01 -1.00635266e+00 -7.11796105e-01 5.53284824e-01
5.30372024e-01 -1.03595304e+00 1.33296907e+00 6.33333325e-02
-1.77031279e-01 -6.83464170e-01 -4.23364222e-01 -1.22405720e+00
-1.90667421e-01 -4.38820213e-01 1.23077668e-01 1.10234702e+00
-4.48139608e-02 -5.90794027e-01 8.53997409e-01 9.65771854e-01
-4.04197216e-01 -1.41250193e-01 -1.39839494e+00 -1.14417124e+00
-2.84118447e-02 -7.59036541e-01 9.62121606e-01 1.14975262e+00
-7.56009519e-01 -4.84982394e-02 1.03038289e-01 7.48231053e-01
1.09980941e+00 -9.91107374e-02 1.14111304e+00 -1.78707278e+00
3.55809033e-01 -1.92790687e-01 -5.41635871e-01 -1.02730644e+00
1.83904126e-01 -7.50901699e-01 2.96063155e-01 -1.57863772e+00
-4.77418989e-01 -1.07506335e+00 4.04296160e-01 5.75814545e-02
6.37095451e-01 1.04942173e-01 1.04950249e-01 6.14530087e-01
1.41688779e-01 5.43702722e-01 9.71242726e-01 -1.14951558e-01
-3.37769359e-01 1.10245515e-02 2.49782298e-02 1.00537658e+00
6.31695151e-01 -8.70273113e-01 -4.61730927e-01 -1.05616307e+00
1.20200835e-01 -2.72522360e-01 6.15226388e-01 -1.32153201e+00
2.44010985e-01 -2.53017366e-01 5.15988946e-01 -1.19820499e+00
6.55268610e-01 -1.49839878e+00 8.75644445e-01 5.17428875e-01
1.67874321e-01 2.33176365e-01 4.59964722e-01 4.68534797e-01
2.21032977e-01 -8.41419250e-02 7.97900617e-01 -1.47770837e-01
-3.07742655e-01 4.83305335e-01 -1.99967593e-01 -2.61645228e-01
8.76731396e-01 -7.01612294e-01 -1.87991947e-01 -2.53174961e-01
-3.44971359e-01 8.57156068e-02 1.19327092e+00 1.41974047e-01
6.46301150e-01 -1.41165769e+00 -6.04560792e-01 5.06236374e-01
-3.94176245e-02 8.70490491e-01 -4.59003933e-02 8.29824746e-01
-9.45101142e-01 1.07895717e-01 -8.87585357e-02 -1.30014825e+00
-1.26868701e+00 1.37316450e-01 6.39347255e-01 2.52845675e-01
-9.98245656e-01 8.03111613e-01 1.34033009e-01 -7.59350061e-01
1.79658636e-01 -7.08995581e-01 4.23539668e-01 -7.93962255e-02
8.29950571e-02 2.92134613e-01 4.15462881e-01 -9.00747538e-01
-6.73903644e-01 9.15646017e-01 3.25728774e-01 -6.52484372e-02
1.30114532e+00 -6.91728368e-02 -7.11165369e-02 3.17969084e-01
8.41515958e-01 1.77357346e-01 -1.28168893e+00 -4.36531715e-02
-1.17423490e-01 -9.82360959e-01 4.30641115e-01 -3.82451266e-01
-1.06212425e+00 7.68766642e-01 9.71104503e-01 1.31943719e-02
7.00789750e-01 -1.31318077e-01 6.45518482e-01 2.50464469e-01
6.44081831e-01 -8.22896481e-01 -5.51436901e-01 4.36252296e-01
8.71918917e-01 -9.89898264e-01 3.83836895e-01 -5.06142855e-01
-3.59880589e-02 1.03768098e+00 6.89172208e-01 -1.39982909e-01
7.87633419e-01 6.19136691e-01 5.54596931e-02 -4.28056538e-01
-1.07210979e-03 5.67592718e-02 1.47811159e-01 1.29801989e+00
-5.08397864e-03 -2.00225443e-01 1.84917182e-01 -2.09640533e-01
-4.26420450e-01 -1.81723058e-01 2.77923197e-01 9.36797023e-01
-2.00326771e-01 -1.00822234e+00 -8.49509358e-01 4.89655197e-01
4.50502813e-01 -4.91097895e-03 -3.10247689e-01 9.87612247e-01
2.74625808e-01 4.69402373e-01 6.44950032e-01 -2.37231582e-01
5.82192719e-01 -3.12959045e-01 6.95757866e-01 -5.94560146e-01
-4.10366446e-01 2.08099172e-01 -4.66553159e-02 -9.76506650e-01
-7.18051016e-01 -9.61371005e-01 -1.15754735e+00 -5.43492794e-01
-5.45352042e-01 -2.55564362e-01 9.32510614e-01 6.39788628e-01
4.99444872e-01 1.92762181e-01 3.34585845e-01 -1.43261290e+00
-3.39685380e-01 -4.08493102e-01 -3.97093534e-01 8.50363672e-02
3.76712829e-01 -8.47493351e-01 -4.39270407e-01 -3.22573990e-01] | [7.821380138397217, -2.8385369777679443] |
6c14d1c7-8941-42d9-a9b4-9eaeb30d3e90 | generating-text-through-adversarial-training | 1808.08703 | null | https://arxiv.org/abs/1808.08703v3 | https://arxiv.org/pdf/1808.08703v3.pdf | Generating Text through Adversarial Training using Skip-Thought Vectors | GANs have been shown to perform exceedingly well on tasks pertaining to image generation and style transfer. In the field of language modelling, word embeddings such as GLoVe and word2vec are state-of-the-art methods for applying neural network models on textual data. Attempts have been made to utilize GANs with word embeddings for text generation. This study presents an approach to text generation using Skip-Thought sentence embeddings with GANs based on gradient penalty functions and f-measures. The proposed architecture aims to reproduce writing style in the generated text by modelling the way of expression at a sentence level across all the works of an author. Extensive experiments were run in different embedding settings on a variety of tasks including conditional text generation and language generation. The model outperforms baseline text generation networks across several automated evaluation metrics like BLEU-n, METEOR and ROUGE. Further, wide applicability and effectiveness in real life tasks are demonstrated through human judgement scores. | ['Afroz Ahamad'] | 2018-08-27 | generating-text-through-adversarial-training-1 | https://aclanthology.org/N19-3008 | https://aclanthology.org/N19-3008.pdf | naacl-2019-6 | ['conditional-text-generation'] | ['natural-language-processing'] | [ 4.10924405e-01 3.47049683e-01 4.21121903e-02 -2.81972855e-01
-7.27761209e-01 -3.40112716e-01 1.32616675e+00 -4.43281800e-01
-3.61073315e-01 1.16264582e+00 6.29303038e-01 -1.70478076e-01
5.84150553e-01 -8.30380321e-01 -4.59758669e-01 -3.93361181e-01
6.10630512e-01 3.77236217e-01 -6.77396297e-01 -4.19023067e-01
4.10166204e-01 3.15394342e-01 -1.14867926e+00 3.47903013e-01
7.57089615e-01 5.38802385e-01 2.30765969e-01 1.13536537e+00
-3.47063720e-01 6.73449934e-01 -1.11721754e+00 -9.42452431e-01
5.09860404e-02 -9.38347161e-01 -7.41026044e-01 5.43258823e-02
5.11617303e-01 -2.48871997e-01 -3.97842646e-01 7.03966498e-01
9.53738570e-01 1.76567882e-01 1.14728677e+00 -1.00798047e+00
-1.60095990e+00 5.61076105e-01 -2.86264867e-01 -1.68800279e-01
5.74562848e-01 3.79313737e-01 9.29851174e-01 -1.03745627e+00
7.67943025e-01 1.36416852e+00 3.85098666e-01 1.16897130e+00
-1.35073984e+00 -5.27741253e-01 -3.15593570e-01 -2.20020175e-01
-9.65295792e-01 -2.08117709e-01 6.76831186e-01 -4.23082530e-01
1.14887679e+00 2.19008639e-01 4.21518981e-01 1.78221083e+00
6.33236766e-01 7.39992976e-01 1.28043258e+00 -7.90980995e-01
1.21782199e-02 4.11840498e-01 -5.31964660e-01 4.46290106e-01
1.09304346e-01 -3.36400829e-02 -8.45765769e-01 2.52937764e-01
9.06567872e-01 -5.36950052e-01 -9.17191952e-02 3.06077331e-01
-1.50292099e+00 1.25510931e+00 2.26357892e-01 4.57375884e-01
-4.07170475e-01 5.11212945e-01 5.16299784e-01 2.73132116e-01
9.17226553e-01 9.64098155e-01 -3.59986871e-02 -2.49091983e-01
-1.10015285e+00 4.22779530e-01 6.56430840e-01 1.20318425e+00
2.96627373e-01 6.73537791e-01 -1.04624474e+00 9.44036245e-01
1.99274004e-01 6.40887558e-01 1.05752325e+00 -4.80565518e-01
5.82581222e-01 3.36859941e-01 1.03576437e-01 -8.41002643e-01
2.71192461e-01 -1.55080497e-01 -1.13307285e+00 2.31138334e-01
-1.01699881e-01 -5.04158795e-01 -1.23551798e+00 1.66058064e+00
-2.15714127e-01 -1.68818340e-01 2.73685485e-01 5.10058463e-01
9.54460144e-01 9.20035839e-01 2.26747915e-01 4.84425202e-02
1.18430817e+00 -1.06250989e+00 -1.10873377e+00 -2.82254457e-01
4.35955822e-01 -1.17427027e+00 1.40638590e+00 1.46446273e-01
-1.30187798e+00 -1.03326881e+00 -8.67831647e-01 -7.81194121e-02
-5.53871453e-01 4.72191542e-01 5.01829445e-01 8.78479302e-01
-1.45231760e+00 4.58640575e-01 -3.17068130e-01 -4.76492643e-01
4.75317210e-01 8.85790288e-02 -2.72482425e-01 1.95144370e-01
-1.28784120e+00 1.06204927e+00 3.26453298e-01 2.82259192e-02
-8.64746094e-01 -4.93551999e-01 -9.12566066e-01 -3.37161347e-02
-4.18969125e-01 -1.27592981e+00 1.18621039e+00 -1.04827213e+00
-1.89689910e+00 9.77807999e-01 -1.46194696e-01 -6.78771436e-01
7.57485390e-01 -4.39364076e-01 -4.84171331e-01 -1.01814561e-01
1.52641743e-01 1.32473302e+00 9.74132001e-01 -1.08720636e+00
-1.29875198e-01 1.12362109e-01 -2.53991872e-01 4.09139812e-01
-7.56272972e-01 -3.46946828e-02 1.67942733e-01 -1.24006104e+00
-7.84275115e-01 -7.12519944e-01 -2.43277237e-01 -2.87228912e-01
-5.44363856e-01 -3.83976340e-01 5.91851711e-01 -8.36955011e-01
1.12115490e+00 -1.56043220e+00 1.53919250e-01 -2.85168707e-01
-1.25561118e-01 3.45519364e-01 -3.77660543e-01 7.17481136e-01
7.84095600e-02 5.06405711e-01 -7.96557218e-03 -5.58908224e-01
2.33157992e-01 -6.53542057e-02 -5.68684399e-01 -2.22659871e-01
5.06951749e-01 1.39013290e+00 -6.57872975e-01 -4.17825848e-01
3.06231201e-01 7.48945355e-01 -3.23863626e-01 6.36959791e-01
-3.55880499e-01 3.46140921e-01 -1.99774250e-01 -2.27795448e-02
1.58335820e-01 3.99980396e-02 -2.61386395e-01 1.35105938e-01
2.10831717e-01 2.57579148e-01 -5.74792385e-01 1.90302765e+00
-1.04813170e+00 9.73057091e-01 -6.16703928e-01 -3.90733391e-01
1.19311881e+00 7.43543625e-01 -2.90920168e-01 -4.64542717e-01
1.42877638e-01 -7.22192004e-02 -2.10977688e-01 -3.69068980e-01
1.05733526e+00 -2.75289267e-01 -1.52642354e-01 7.04280794e-01
5.54094672e-01 -6.35915041e-01 3.91722590e-01 1.92600787e-01
8.03974748e-01 4.33981091e-01 1.77349284e-01 -2.28297994e-01
5.08660913e-01 -2.82172501e-01 -4.28872198e-01 7.71705151e-01
3.38801414e-01 8.26561272e-01 2.16700345e-01 -3.30649942e-01
-1.52350950e+00 -9.47643101e-01 4.08040404e-01 1.07020009e+00
-5.35221040e-01 -1.86394989e-01 -1.17667747e+00 -5.56359231e-01
-4.46150690e-01 1.37047982e+00 -9.15143609e-01 -2.39806786e-01
-3.41346592e-01 -5.42606711e-01 7.83257186e-01 6.85330987e-01
4.39003527e-01 -1.87725139e+00 -3.45434695e-01 3.10402453e-01
-1.66907310e-01 -1.00514829e+00 -7.28508174e-01 -2.68345565e-01
-8.10613871e-01 -2.78298408e-01 -1.29414845e+00 -8.40855300e-01
7.39139497e-01 -3.03892434e-01 1.42344713e+00 -1.71956450e-01
-4.00482416e-01 4.16677058e-01 -4.13162559e-01 -6.75423741e-01
-9.02500451e-01 3.66424024e-01 -2.24253327e-01 -3.05495746e-02
2.69117624e-01 -2.34838009e-01 -5.72433174e-01 -1.30515933e-01
-1.11915839e+00 2.83023506e-01 6.70654953e-01 1.18132448e+00
1.84982345e-01 -5.76696277e-01 8.86233747e-01 -8.29245806e-01
1.63990927e+00 1.90334748e-02 -2.29296088e-02 3.36084545e-01
-7.83853173e-01 1.60002172e-01 6.94136441e-01 -4.66145635e-01
-1.36476266e+00 -3.90309215e-01 -1.05271660e-01 -4.71091978e-02
-4.92715299e-01 1.46193936e-01 -3.09333229e-03 3.97207707e-01
8.01279426e-01 5.75801432e-01 -6.87000006e-02 6.20476231e-02
8.03544760e-01 8.56325746e-01 3.82956266e-01 -3.77185822e-01
8.64356279e-01 -6.21806048e-02 -1.85301125e-01 -8.91382217e-01
-7.46918857e-01 7.15674385e-02 -6.05237603e-01 -2.77801603e-01
1.28503203e+00 -8.84197533e-01 -6.41217083e-02 4.62110668e-01
-1.48977792e+00 -3.79314184e-01 -5.85396707e-01 2.11252406e-01
-7.61534214e-01 7.33578503e-02 -6.06752098e-01 -8.72305036e-01
-1.06805050e+00 -8.98535788e-01 1.22705233e+00 2.94793904e-01
-7.37492204e-01 -1.50254357e+00 1.95595592e-01 3.60172868e-01
6.83236063e-01 3.09552580e-01 7.50331342e-01 -4.16298062e-01
-9.31150913e-02 -3.07201236e-01 5.17899469e-02 6.85512364e-01
2.59942621e-01 -1.30127400e-01 -1.06935167e+00 -1.66745663e-01
-2.03684911e-01 -4.93124992e-01 7.08773792e-01 2.85543919e-01
1.00303805e+00 -4.51663703e-01 1.09835356e-01 3.56036127e-01
1.24589705e+00 8.62629339e-02 1.05277395e+00 2.15936795e-01
7.44834840e-01 4.19433504e-01 2.06209019e-01 2.12199077e-01
-1.01225369e-01 4.22498405e-01 4.47778702e-02 -2.29101568e-01
-5.20765901e-01 -5.57518244e-01 5.36311567e-01 8.07895839e-01
-8.73603448e-02 -9.31842625e-01 -4.99598682e-01 7.46895611e-01
-1.38344920e+00 -1.16080785e+00 -6.12697490e-02 1.80621374e+00
1.04919589e+00 9.96726155e-02 -1.26117513e-01 -1.35503128e-01
7.47230470e-01 2.91622877e-01 -1.69261292e-01 -1.16378057e+00
-2.57259220e-01 7.13676393e-01 2.17373878e-01 3.75861526e-01
-8.06992888e-01 1.27019632e+00 6.61977148e+00 9.12522078e-01
-1.02626324e+00 1.18276678e-01 9.89891350e-01 5.79903871e-02
-4.97754216e-01 -4.20841932e-01 -8.45065296e-01 4.37370867e-01
1.07368553e+00 -3.06786776e-01 3.45606714e-01 6.26134217e-01
2.38775775e-01 2.22694054e-01 -1.08442175e+00 7.49744236e-01
5.84816277e-01 -1.35559821e+00 7.12543786e-01 -4.21977788e-02
1.36945832e+00 -5.06381989e-01 4.78300154e-01 3.30126613e-01
4.90729600e-01 -1.67927396e+00 5.34992158e-01 6.90711081e-01
1.32565069e+00 -7.36767828e-01 9.05622959e-01 -7.00772693e-03
-6.23513162e-01 4.52492088e-01 -3.54509950e-01 -4.49495651e-02
4.12143141e-01 3.40469033e-01 -1.31851745e+00 4.36065912e-01
1.15387954e-01 4.12685603e-01 -6.29345238e-01 3.66578102e-01
-6.83550000e-01 6.75575018e-01 3.48771453e-01 -4.72004205e-01
3.36431265e-01 -1.10224240e-01 1.96940437e-01 1.58222282e+00
6.66496277e-01 -5.37182689e-01 -4.45449680e-01 1.15038562e+00
-4.25150424e-01 3.71669888e-01 -1.00407779e+00 -5.33948481e-01
4.27683536e-03 1.39057517e+00 -4.51259971e-01 -5.38217425e-01
-1.12802185e-01 1.48662865e+00 3.07447985e-02 3.83124828e-01
-7.00929165e-01 -6.24350727e-01 3.97164226e-01 9.84456241e-02
3.17919761e-01 -1.25655398e-01 -6.32530868e-01 -8.76100957e-01
-1.06618918e-01 -8.54182839e-01 -2.31354907e-01 -1.25587940e+00
-1.54481137e+00 1.07112002e+00 -1.51870891e-01 -9.60126162e-01
-7.44330049e-01 -7.29382813e-01 -9.58135724e-01 1.43910968e+00
-1.06239057e+00 -1.57409728e+00 -2.89613783e-01 3.12714159e-01
1.11685932e+00 -8.95778298e-01 1.21637058e+00 -2.46461868e-01
-8.70366767e-02 8.75823915e-01 -6.41786754e-02 1.63298652e-01
8.41323376e-01 -1.42205954e+00 1.02628148e+00 8.40852261e-01
4.41157937e-01 6.74001098e-01 7.25892246e-01 -7.47763395e-01
-8.34386706e-01 -1.35160518e+00 1.24983156e+00 -4.01673675e-01
4.14905071e-01 -5.79208851e-01 -3.58168274e-01 5.68360209e-01
1.25230181e+00 -6.43318176e-01 7.55535543e-01 -3.57771426e-01
-1.22358084e-01 2.79693693e-01 -1.03666985e+00 9.83363092e-01
7.92977750e-01 -5.30796468e-01 -5.40898383e-01 5.81755638e-01
6.34429514e-01 -1.80101171e-01 -8.25653791e-01 1.12578494e-03
3.72954041e-01 -6.54933333e-01 6.79126859e-01 -7.85282671e-01
1.27350140e+00 1.17771342e-01 1.30123109e-01 -1.86063516e+00
-2.23167717e-01 -9.45864975e-01 2.01623235e-02 1.64439178e+00
5.85210681e-01 -3.38473260e-01 5.91373801e-01 2.64466584e-01
2.20847204e-02 -5.83991706e-01 -5.64979911e-01 -5.57627559e-01
4.21509832e-01 -9.86409560e-02 6.23578191e-01 6.00530684e-01
-4.20583218e-01 9.44073260e-01 -7.47556746e-01 -6.18787348e-01
2.69301116e-01 -3.65262449e-01 9.73028541e-01 -8.21853638e-01
-1.63892329e-01 -5.57235301e-01 -3.42203170e-01 -7.63793647e-01
4.83540177e-01 -1.00558054e+00 4.86926325e-02 -1.89854240e+00
1.01711638e-01 2.52666920e-01 1.41065687e-01 1.04468428e-01
-4.11434382e-01 5.48760116e-01 3.20685565e-01 -2.10477814e-01
-9.67139974e-02 7.92451084e-01 1.56977713e+00 -2.34610647e-01
1.73798695e-01 -2.39430353e-01 -8.40463281e-01 2.38456801e-01
1.12286460e+00 -2.24414498e-01 -5.46611249e-01 -6.46328330e-01
1.50156438e-01 -1.33412227e-01 2.53803551e-01 -9.81277168e-01
-2.52370507e-01 2.36802511e-02 8.17500770e-01 -2.52513617e-01
4.06269848e-01 -3.74762952e-01 3.76415849e-02 2.82142431e-01
-1.04859006e+00 5.45637727e-01 5.44396751e-02 3.86586964e-01
-2.04807281e-01 -4.76193726e-01 6.06789410e-01 -3.16554338e-01
-2.35548586e-01 1.57183502e-02 -4.94574219e-01 2.50840932e-01
1.07321644e+00 -3.09826195e-01 -3.30174528e-02 -9.04126048e-01
-4.61326599e-01 -3.01765203e-01 3.01597476e-01 7.52968729e-01
8.28103304e-01 -1.63307774e+00 -1.10836196e+00 1.72606573e-01
1.19658336e-02 -3.83825362e-01 -1.57466866e-02 9.89226699e-02
-5.64141035e-01 6.61621392e-01 -4.26670611e-01 -1.71650007e-01
-1.16647303e+00 2.35381514e-01 1.57071531e-01 -7.55928755e-01
-1.49235576e-01 9.80431497e-01 1.83559880e-01 -3.63082677e-01
-8.01431015e-02 5.24896802e-03 -1.71384156e-01 -1.36070326e-01
5.64817905e-01 2.37166315e-01 -6.63084239e-02 -5.35334229e-01
3.51207376e-01 2.97638953e-01 -1.91949546e-01 -5.75438321e-01
1.13434899e+00 1.54141277e-01 8.74725953e-02 4.63025540e-01
9.62588429e-01 -2.98881024e-01 -9.60788071e-01 2.32309282e-01
-3.24335515e-01 -2.63459295e-01 -1.83601305e-01 -1.08034611e+00
-8.07803035e-01 1.21131492e+00 5.75069308e-01 1.90540716e-01
8.26154470e-01 -4.19838846e-01 8.80951405e-01 1.04793802e-01
-1.29208425e-02 -1.31436491e+00 4.81066942e-01 5.26276708e-01
1.29527581e+00 -1.09884679e+00 -1.98757291e-01 1.57359004e-01
-1.06962419e+00 1.12504053e+00 7.90066063e-01 -2.75416076e-01
7.33753368e-02 7.19712600e-02 2.41030663e-01 2.05477744e-01
-8.16503108e-01 2.31327057e-01 5.74753463e-01 8.52955401e-01
1.21400273e+00 2.40539432e-01 -6.20653749e-01 6.55631050e-02
-7.62075245e-01 4.80106240e-03 6.63505852e-01 5.45817018e-01
5.99291697e-02 -1.29010665e+00 -2.13381752e-01 5.76872647e-01
-7.71072090e-01 -4.39709485e-01 -6.74790084e-01 6.71121299e-01
-2.18689486e-01 9.16458607e-01 -6.21511182e-03 -3.37999523e-01
8.08805227e-02 4.96473938e-01 5.49745381e-01 -8.18299294e-01
-9.65337873e-01 -2.21978843e-01 2.07126379e-01 4.87346165e-02
-4.03197199e-01 -3.75795960e-01 -6.78762734e-01 -1.57991245e-01
-2.38493651e-01 9.01569873e-02 8.62130284e-01 6.65576100e-01
3.80614102e-01 8.38083744e-01 4.41885620e-01 -8.44296575e-01
-7.19899476e-01 -1.73428953e+00 -3.36613894e-01 7.58747756e-01
-2.99577922e-01 -7.42793307e-02 -1.53137252e-01 5.21305144e-01] | [11.879911422729492, 9.352286338806152] |
4be019e0-a213-4668-96f3-6a13bf4d6215 | lnl-k-learning-with-noisy-labels-and-noise | 2306.11911 | null | https://arxiv.org/abs/2306.11911v1 | https://arxiv.org/pdf/2306.11911v1.pdf | LNL+K: Learning with Noisy Labels and Noise Source Distribution Knowledge | Learning with noisy labels (LNL) is challenging as the model tends to memorize noisy labels, which can lead to overfitting. Many LNL methods detect clean samples by maximizing the similarity between samples in each category, which does not make any assumptions about likely noise sources. However, we often have some knowledge about the potential source(s) of noisy labels. For example, an image mislabeled as a cheetah is more likely a leopard than a hippopotamus due to their visual similarity. Thus, we introduce a new task called Learning with Noisy Labels and noise source distribution Knowledge (LNL+K), which assumes we have some knowledge about likely source(s) of label noise that we can take advantage of. By making this presumption, methods are better equipped to distinguish hard negatives between categories from label noise. In addition, this enables us to explore datasets where the noise may represent the majority of samples, a setting that breaks a critical premise of most methods developed for the LNL task. We explore several baseline LNL+K approaches that integrate noise source knowledge into state-of-the-art LNL methods across three diverse datasets and three types of noise, where we report a 5-15% boost in performance compared with the unadapted methods. Critically, we find that LNL methods do not generalize well in every setting, highlighting the importance of directly exploring our LNL+K task. | ['Bryan A. Plummer', 'Siqi Wang'] | 2023-06-20 | null | null | null | null | ['learning-with-noisy-labels', 'learning-with-noisy-labels'] | ['computer-vision', 'natural-language-processing'] | [ 2.60596901e-01 -6.12706989e-02 -5.15772635e-03 -3.26190859e-01
-1.18398356e+00 -8.92895699e-01 5.59476912e-01 2.71656573e-01
-4.20769811e-01 6.50533974e-01 9.93027166e-02 -7.68825784e-02
6.51924983e-02 -6.52625203e-01 -9.34568405e-01 -9.96270061e-01
2.32958764e-01 1.11741200e-01 5.07094339e-03 2.08948985e-01
7.54166469e-02 2.70677328e-01 -1.55470538e+00 2.86687881e-01
4.24346209e-01 1.05271804e+00 1.57158405e-01 4.16318148e-01
-3.00484180e-01 1.02380705e+00 -8.59272778e-01 -4.26168352e-01
3.97459358e-01 -4.62820560e-01 -7.41954803e-01 8.10145810e-02
7.93000877e-01 5.50993681e-02 -1.32781595e-01 1.47228348e+00
4.64264542e-01 1.24319188e-01 7.56593287e-01 -1.25153565e+00
-5.87061465e-01 7.28230715e-01 -6.69474244e-01 1.95647955e-01
1.12957329e-01 3.43114257e-01 1.04807103e+00 -8.18349957e-01
5.65090537e-01 1.49720502e+00 1.04616630e+00 6.43575966e-01
-1.57559776e+00 -7.41556048e-01 3.59377950e-01 1.70705810e-01
-1.48468876e+00 -3.04540008e-01 6.30177736e-01 -4.97974664e-01
3.43189448e-01 2.65922606e-01 1.41266644e-01 1.50885713e+00
-2.44901359e-01 9.20559704e-01 1.58756208e+00 -4.37690616e-01
4.38537031e-01 2.69834697e-01 3.53362799e-01 2.97283888e-01
4.17924643e-01 9.89902243e-02 -6.07068777e-01 -2.86252171e-01
2.87693381e-01 -5.42205349e-02 -3.25373918e-01 -1.08528055e-01
-1.08165336e+00 6.84367418e-01 5.51956415e-01 3.45024839e-02
-7.59645179e-02 2.47341350e-01 4.26160127e-01 2.62360871e-01
4.68499571e-01 5.81881344e-01 -3.62748832e-01 -1.39196301e-02
-1.01524210e+00 2.31614992e-01 7.80288875e-01 8.25941980e-01
9.78576005e-01 -1.86919779e-01 -1.74320713e-01 8.50919724e-01
7.37599209e-02 7.11366832e-01 2.55693555e-01 -1.15252233e+00
2.10089117e-01 4.17014986e-01 2.54427731e-01 -8.33209455e-01
-4.47936535e-01 -5.25699019e-01 -8.91082883e-01 2.73389727e-01
6.57099903e-01 -1.39812395e-01 -1.20365834e+00 2.00841331e+00
2.58155257e-01 2.34651625e-01 -4.79247905e-02 8.04307401e-01
6.59434855e-01 4.91098940e-01 1.72299340e-01 -1.07268378e-01
1.23931265e+00 -6.66738212e-01 -6.56723678e-01 -5.12071371e-01
5.91085434e-01 -7.10431457e-01 1.17672110e+00 5.99155962e-01
-5.10733247e-01 -5.28165758e-01 -7.51791477e-01 4.01607789e-02
-4.97226894e-01 -8.59568194e-02 5.00111282e-01 7.16776729e-01
-7.83813536e-01 5.72390735e-01 -5.85710645e-01 -2.67489552e-01
6.58092022e-01 -8.40957612e-02 -1.85298324e-01 -5.51044822e-01
-9.69576299e-01 8.62283707e-01 1.91403911e-01 1.25262305e-01
-1.04738367e+00 -6.00177050e-01 -8.98409188e-01 -9.50229689e-02
7.14990497e-01 -2.91809112e-01 1.36414039e+00 -9.82988000e-01
-5.60524940e-01 9.50436413e-01 -4.10269916e-01 -3.40218186e-01
6.13133073e-01 -9.43352580e-02 -2.57695824e-01 -9.44070071e-02
3.09615493e-01 6.51136577e-01 8.30530107e-01 -1.85674965e+00
-6.81933761e-01 -2.68521219e-01 -6.19488070e-03 -1.29142860e-02
2.87477747e-02 -2.22733751e-01 -1.03216410e-01 -7.08223522e-01
1.55841589e-01 -8.30038369e-01 -1.52621880e-01 2.27283970e-01
-6.34524941e-01 -3.28559995e-01 6.73074424e-01 -2.83987552e-01
9.96800900e-01 -2.28603435e+00 -3.58307779e-01 3.34793210e-01
4.01289910e-01 3.47939968e-01 -3.04469794e-01 1.72012955e-01
1.81227904e-02 3.83035123e-01 -1.80863321e-01 -5.80930650e-01
1.50197402e-01 4.19619203e-01 -3.92577767e-01 5.07029951e-01
1.41103491e-01 7.70976782e-01 -1.24278080e+00 -2.55771458e-01
-1.19253201e-02 3.46507639e-01 1.30115047e-01 4.27207202e-02
-3.02289039e-01 2.56957918e-01 -5.03864028e-02 7.45676875e-01
8.55196476e-01 -3.55445564e-01 -1.28451455e-03 -2.06444904e-01
2.09041759e-01 2.70123541e-01 -1.48232162e+00 1.15395474e+00
-3.09283495e-01 6.19392335e-01 1.83758780e-01 -7.30311036e-01
6.85003817e-01 2.01555848e-01 8.03208873e-02 -5.44592381e-01
4.12313864e-02 2.50258595e-01 -1.99065834e-01 -4.97424036e-01
3.93108390e-02 -4.71719891e-01 2.61131991e-02 4.99122083e-01
3.36555056e-02 -1.39758870e-01 1.82205930e-01 2.30364695e-01
1.24399543e+00 -2.83849567e-01 2.04733044e-01 -7.91685656e-02
-7.66959935e-02 -4.84790727e-02 8.44478071e-01 1.56392491e+00
-3.40549827e-01 8.41156185e-01 5.90298831e-01 -3.29358041e-01
-9.42039251e-01 -1.17948568e+00 -2.28181288e-01 8.96676719e-01
2.13196769e-01 -2.97342837e-01 -6.51059985e-01 -9.76228356e-01
-9.83392540e-03 6.20557725e-01 -7.88548708e-01 -1.27332136e-01
-1.27712786e-01 -9.15413260e-01 7.17134893e-01 4.87280309e-01
2.92748511e-01 -8.69157076e-01 -2.77671725e-01 -7.69882090e-03
-3.11322629e-01 -1.04282117e+00 -3.75139117e-01 6.82775199e-01
-4.33717340e-01 -1.28566158e+00 -3.69728118e-01 -6.53791368e-01
7.78347373e-01 4.70748842e-01 1.40191734e+00 2.10336640e-01
-1.96479306e-01 3.93235713e-01 -4.63909149e-01 -5.85076451e-01
-4.56303447e-01 -1.96409941e-01 1.07399277e-01 -3.61372121e-02
7.76761055e-01 -3.85078311e-01 -4.24231023e-01 2.60574639e-01
-1.04819047e+00 -3.77432108e-01 5.47732472e-01 7.97443748e-01
7.77202487e-01 4.22477394e-01 5.81136405e-01 -1.24795735e+00
3.79636735e-01 -8.05050015e-01 -4.77980256e-01 2.35693201e-01
-3.84488076e-01 1.24431558e-01 6.62791371e-01 -7.19667912e-01
-8.25577438e-01 1.63380817e-01 -9.11577493e-02 -2.71321237e-01
-5.40120661e-01 2.37049118e-01 -2.84488529e-01 4.91081923e-02
9.59994078e-01 -7.59694818e-03 -2.72781849e-01 -7.08289921e-01
3.87890399e-01 6.67671144e-01 5.82862198e-01 -6.29674196e-01
8.10423017e-01 6.59413755e-01 9.65444185e-03 -6.95409596e-01
-1.55887735e+00 -6.81248844e-01 -3.48522127e-01 -5.12716360e-02
4.95563895e-01 -9.73936856e-01 -4.68531251e-01 5.59913218e-01
-9.14328873e-01 -4.28918540e-01 -4.74030286e-01 2.46914521e-01
-2.34806255e-01 4.51364905e-01 -5.57221115e-01 -1.00317192e+00
1.84968516e-01 -1.15092909e+00 8.97356808e-01 1.65296152e-01
-2.07300931e-01 -8.81048262e-01 -2.37975150e-01 2.69237727e-01
3.21415067e-01 2.58245200e-01 9.47352588e-01 -8.51804972e-01
-4.79419202e-01 -9.91322771e-02 -4.04129654e-01 7.80935526e-01
1.68680623e-01 -3.56388152e-01 -1.32947481e+00 -2.35480979e-01
1.89562812e-01 -7.59259105e-01 1.23621273e+00 2.36776024e-01
1.17454243e+00 -2.53444225e-01 -1.83491573e-01 2.98158199e-01
1.48237455e+00 -2.37964854e-01 4.14086670e-01 1.66878745e-01
8.26193035e-01 5.98062754e-01 4.26411152e-01 2.11907074e-01
3.34322006e-01 3.56637269e-01 5.41004062e-01 -1.62028253e-01
-4.66783136e-01 -3.59949529e-01 2.86585450e-01 3.73348951e-01
5.30861974e-01 -3.19524765e-01 -8.26436698e-01 8.00297499e-01
-1.67578232e+00 -8.15813959e-01 -2.91515023e-01 2.21220565e+00
1.12339330e+00 1.43597394e-01 -4.81731184e-02 3.00146163e-01
8.20205271e-01 -1.48205217e-02 -8.65898132e-01 1.52972296e-01
-3.27130824e-01 -4.15415503e-02 6.30881369e-01 4.65488285e-01
-1.32700622e+00 5.73310018e-01 6.55043602e+00 1.27429974e+00
-8.83253396e-01 1.52393013e-01 9.53159809e-01 -1.59821004e-01
-3.07390392e-01 -1.10707551e-01 -9.61495519e-01 7.02039838e-01
6.05804384e-01 2.45658085e-01 4.93407935e-01 8.39403391e-01
6.90930784e-02 -5.73875487e-01 -1.28350687e+00 1.13257718e+00
1.38376758e-01 -8.13719511e-01 -2.41750717e-01 -4.03901450e-02
1.06024551e+00 1.00600280e-01 7.20841810e-02 2.86369503e-01
7.38613367e-01 -1.26185763e+00 8.72470975e-01 5.37475586e-01
5.69926143e-01 -5.57652295e-01 9.83214259e-01 6.11403584e-01
-7.86565006e-01 -1.30874798e-01 -5.09349585e-01 -1.45454213e-01
-5.84201030e-02 1.29693890e+00 -6.13690317e-01 4.60721888e-02
8.97746384e-01 4.16390598e-01 -8.99776518e-01 1.27028382e+00
-6.42279029e-01 9.56572533e-01 -6.32405221e-01 2.11927906e-01
2.16568470e-01 2.14715779e-01 3.65809262e-01 1.25718701e+00
2.54322052e-01 -1.26207143e-01 3.70598346e-01 9.79470670e-01
-4.65480626e-01 -2.94919521e-01 -6.02605462e-01 2.49391586e-01
6.76015198e-01 1.14116192e+00 -9.60539401e-01 -3.41728330e-01
-5.17933190e-01 7.06995487e-01 3.78819019e-01 4.10665959e-01
-4.27610397e-01 -7.39483982e-02 6.20509684e-01 2.58167479e-02
1.99933782e-01 1.21824808e-01 -5.92756629e-01 -9.82694626e-01
1.68418467e-01 -1.04715014e+00 2.92132080e-01 -7.14189053e-01
-2.03016520e+00 2.83079624e-01 -2.11075544e-01 -1.02152109e+00
1.48558483e-01 -5.49416959e-01 -2.04456583e-01 7.95598090e-01
-1.62785482e+00 -1.02948761e+00 -3.55837107e-01 2.27990136e-01
4.78977859e-01 3.69775146e-01 5.08961201e-01 1.31050467e-01
-3.19933742e-01 5.31628907e-01 4.04702306e-01 1.38053492e-01
1.05383122e+00 -1.58677888e+00 2.80021727e-01 9.90453959e-01
2.85423607e-01 4.58829582e-01 6.85435951e-01 -7.51915395e-01
-8.51588726e-01 -1.38606453e+00 9.07923043e-01 -8.24472427e-01
6.00696266e-01 -6.32693887e-01 -1.20739055e+00 6.02755725e-01
-2.85499394e-01 3.85115296e-01 6.31863117e-01 1.44498214e-01
-7.86184072e-01 -7.29587078e-02 -1.30276608e+00 3.87543827e-01
1.06490505e+00 -6.06253922e-01 -3.18667740e-01 4.43279684e-01
5.13033748e-01 -2.05087066e-01 -4.04144973e-01 1.72639012e-01
1.32775173e-01 -9.03439641e-01 8.94336283e-01 -2.03173831e-01
2.06347346e-01 -4.99471128e-01 -2.39656955e-01 -1.51738536e+00
-2.13963464e-01 -2.22294867e-01 1.28718421e-01 1.57088232e+00
3.80563259e-01 -2.55186349e-01 5.51307678e-01 7.25766420e-01
1.40879288e-01 -4.62265909e-01 -8.63377750e-01 -1.02337348e+00
1.06307231e-01 -7.83822417e-01 4.40044910e-01 1.02749395e+00
-4.52060193e-01 1.17624797e-01 -5.89841604e-01 2.79588878e-01
8.37151408e-01 -2.56674588e-01 4.31026816e-01 -1.29270613e+00
-1.54631987e-01 -2.33218431e-01 -2.50597388e-01 -9.80814099e-01
2.65006691e-01 -9.20922160e-01 5.91778338e-01 -1.40663862e+00
4.77582633e-01 -6.81315720e-01 -3.88538480e-01 8.24009955e-01
-5.31945705e-01 6.96211755e-01 2.39854425e-01 2.28152975e-01
-9.52389002e-01 1.40204459e-01 9.49370682e-01 -2.14440227e-01
2.43326828e-01 -3.53771262e-02 -1.00497448e+00 9.25412714e-01
6.61281884e-01 -9.04855013e-01 -2.55103707e-01 -3.79404962e-01
4.66926187e-01 -6.72888696e-01 5.63721776e-01 -9.11989212e-01
2.24376559e-01 -1.64828658e-01 4.33808833e-01 -3.02060485e-01
1.27296135e-01 -9.35567439e-01 -2.28591077e-02 7.14287385e-02
-4.61970747e-01 -3.60051811e-01 -4.48110551e-02 8.68414819e-01
-1.05432160e-01 -5.97566664e-01 9.47286606e-01 -5.01904070e-01
-7.39550531e-01 -1.21384874e-01 -4.78358567e-01 4.67743337e-01
5.83818793e-01 7.68606961e-02 -6.12254977e-01 -4.72787648e-01
-6.72026932e-01 2.67066061e-01 6.36260927e-01 1.90855518e-01
2.94410706e-01 -1.23418367e+00 -5.14365911e-01 8.70026127e-02
2.37402618e-01 1.66018009e-01 1.12711959e-01 5.92215776e-01
-6.49960188e-04 -7.88522363e-02 3.53271782e-01 -5.04314780e-01
-1.13542259e+00 6.72202408e-01 2.62222826e-01 -2.24692643e-01
-3.70371729e-01 9.67481911e-01 2.98991293e-01 -5.44710040e-01
6.66028023e-01 -1.72821775e-01 1.37567654e-01 1.11951351e-01
7.57804394e-01 4.76995975e-01 2.12916523e-01 -5.07278860e-01
-2.79970914e-01 4.78878379e-01 3.12492661e-02 2.06445307e-01
1.12876105e+00 -2.76165724e-01 -6.41240552e-02 7.54730761e-01
1.15036118e+00 8.74823332e-03 -1.48168921e+00 -7.01110542e-01
2.01875672e-01 -4.62486565e-01 -2.27005621e-05 -1.22404742e+00
-9.31114197e-01 8.44558060e-01 6.53459907e-01 2.27871403e-01
1.09620225e+00 2.76733816e-01 5.95259607e-01 5.40615618e-01
3.82065862e-01 -1.11709094e+00 1.77385867e-01 4.89979237e-01
4.04796690e-01 -1.73969913e+00 -2.09168375e-01 -4.60530281e-01
-6.04209542e-01 8.04733992e-01 5.07847726e-01 5.14273345e-02
5.62589049e-01 4.16479170e-01 5.98492801e-01 -2.30742723e-01
-5.21669567e-01 -6.07120872e-01 2.68676076e-02 8.50407362e-01
-3.15367430e-02 8.35781358e-03 -2.67625526e-02 7.05939949e-01
-2.93755196e-02 -3.09581548e-01 6.47072792e-01 8.31092238e-01
-5.89173257e-01 -1.11160755e+00 -7.40593433e-01 6.23816729e-01
-5.32891929e-01 -9.43944454e-02 -4.39420402e-01 5.08544147e-01
7.44005382e-01 1.31334305e+00 -1.60853356e-01 -1.95086971e-01
3.25078785e-01 3.03819329e-01 2.50756294e-01 -7.82360017e-01
-4.45303798e-01 8.04518461e-02 -1.71512485e-01 -4.27433014e-01
-4.32166934e-01 -4.68938917e-01 -9.60871875e-01 -2.82018811e-01
-3.80237639e-01 6.04978278e-02 5.37387967e-01 9.87738132e-01
1.20965600e-01 2.00365990e-01 3.97344321e-01 -7.16580331e-01
-6.47828937e-01 -9.54241455e-01 -8.46718490e-01 9.03681278e-01
6.43857241e-01 -7.15996921e-01 -8.70629728e-01 1.44154981e-01] | [9.400724411010742, 3.927699327468872] |
b2aed967-48cd-451e-b292-6788b2137309 | a-hierarchical-spatio-temporal-graph | 2112.04294 | null | https://arxiv.org/abs/2112.04294v2 | https://arxiv.org/pdf/2112.04294v2.pdf | A Hierarchical Spatio-Temporal Graph Convolutional Neural Network for Anomaly Detection in Videos | Deep learning models have been widely used for anomaly detection in surveillance videos. Typical models are equipped with the capability to reconstruct normal videos and evaluate the reconstruction errors on anomalous videos to indicate the extent of abnormalities. However, existing approaches suffer from two disadvantages. Firstly, they can only encode the movements of each identity independently, without considering the interactions among identities which may also indicate anomalies. Secondly, they leverage inflexible models whose structures are fixed under different scenes, this configuration disables the understanding of scenes. In this paper, we propose a Hierarchical Spatio-Temporal Graph Convolutional Neural Network (HSTGCNN) to address these problems, the HSTGCNN is composed of multiple branches that correspond to different levels of graph representations. High-level graph representations encode the trajectories of people and the interactions among multiple identities while low-level graph representations encode the local body postures of each person. Furthermore, we propose to weightedly combine multiple branches that are better at different scenes. An improvement over single-level graph representations is achieved in this way. An understanding of scenes is achieved and serves anomaly detection. High-level graph representations are assigned higher weights to encode moving speed and directions of people in low-resolution videos while low-level graph representations are assigned higher weights to encode human skeletons in high-resolution videos. Experimental results show that the proposed HSTGCNN significantly outperforms current state-of-the-art models on four benchmark datasets (UCSD Pedestrian, ShanghaiTech, CUHK Avenue and IITB-Corridor) by using much less learnable parameters. | ['Zifeng Qiu', 'Yafeng Hao', 'Hongguang Li', 'Wenrui Ding', 'Yalong Jiang', 'Xianlin Zeng'] | 2021-12-08 | null | null | null | null | ['anomaly-detection-in-surveillance-videos', 'anomaly-detection-in-surveillance-videos'] | ['computer-vision', 'methodology'] | [-1.62315011e-01 -1.04517214e-01 9.75891724e-02 -4.91121002e-02
2.74387777e-01 -1.32542118e-01 5.08098125e-01 4.28611748e-02
-1.54764250e-01 1.70937553e-01 4.09236252e-01 -1.19150672e-02
8.85033831e-02 -8.30267429e-01 -7.58784890e-01 -6.11478448e-01
-5.33168316e-01 1.39506295e-01 6.31288052e-01 -3.79665524e-01
-9.48037654e-02 5.35004675e-01 -1.38719821e+00 4.25900370e-01
6.21625960e-01 9.58272755e-01 -4.28653598e-01 8.03088367e-01
6.25824928e-02 9.48049963e-01 -3.98780107e-01 -3.91164452e-01
4.01436538e-01 -4.91925478e-01 -4.54676002e-01 4.07255083e-01
7.03575552e-01 -6.52928233e-01 -1.00184202e+00 1.22445369e+00
1.25294656e-01 1.01409853e-01 6.15327477e-01 -1.44246590e+00
-5.98851085e-01 8.75170529e-02 -8.75013411e-01 8.25195134e-01
4.64120567e-01 3.33377808e-01 7.51328290e-01 -3.62211734e-01
3.83760333e-01 1.50982904e+00 8.60969126e-01 4.88541633e-01
-6.21301889e-01 -4.30167139e-01 8.34554791e-01 6.67639256e-01
-1.09835625e+00 -1.06617203e-02 6.65731132e-01 -5.88576019e-01
9.53693151e-01 6.98767826e-02 1.17093360e+00 1.48589313e+00
4.42739785e-01 7.41216719e-01 4.15297210e-01 1.52517006e-01
-1.73935264e-01 -6.97169304e-01 1.95737362e-01 1.25752223e+00
6.61244333e-01 5.59216514e-02 -3.64256948e-01 -1.17368080e-01
8.81997287e-01 5.28444111e-01 -4.95411783e-01 -4.40783650e-01
-1.18298924e+00 6.11831367e-01 5.38945675e-01 2.77618289e-01
-5.56744993e-01 1.92200005e-01 6.61539972e-01 2.03760713e-01
2.91102201e-01 -1.36717096e-01 -2.39511773e-01 1.70272067e-01
-5.18082619e-01 2.51855582e-01 2.96484232e-01 7.70597696e-01
4.68160540e-01 5.69586098e-01 -4.76962447e-01 4.16059613e-01
4.71390456e-01 3.10353339e-01 6.09266162e-01 -5.80151975e-01
5.38910985e-01 9.30433989e-01 -2.26254370e-02 -1.87194908e+00
-7.28631258e-01 -3.85891795e-01 -1.16253746e+00 1.24688700e-01
4.43253040e-01 -1.79297790e-01 -1.35734510e+00 1.52943051e+00
2.78457165e-01 8.80635142e-01 -6.31300434e-02 9.36751246e-01
9.61455405e-01 7.62474179e-01 8.81081745e-02 3.18712480e-02
1.32147741e+00 -1.03946006e+00 -8.55683565e-01 -3.29488963e-01
4.80086118e-01 -3.10748637e-01 6.04748249e-01 3.63303542e-01
-8.43964398e-01 -8.48077059e-01 -9.19655979e-01 2.43914858e-01
-4.99031901e-01 -2.36060787e-02 5.85399389e-01 4.58774894e-01
-1.14951420e+00 4.08300757e-01 -1.01150477e+00 -5.46587288e-01
4.59418803e-01 1.80207357e-01 -5.72370589e-01 -8.71932879e-02
-1.14260101e+00 5.40914178e-01 4.58659619e-01 4.25451875e-01
-1.00702548e+00 -2.40348130e-01 -1.29238701e+00 -1.28290756e-02
3.90793443e-01 -7.23315120e-01 5.50606966e-01 -1.00009000e+00
-9.72972512e-01 7.05767810e-01 1.63144916e-01 -5.05736172e-01
6.11824691e-01 -2.45699212e-01 -7.94301391e-01 2.66358137e-01
2.17652377e-02 3.51416171e-01 9.52472389e-01 -8.92979145e-01
-7.31040061e-01 -5.18682599e-01 1.27582252e-01 2.05093116e-01
-2.26736903e-01 -1.48430452e-01 -7.85585463e-01 -1.04226923e+00
3.28420907e-01 -7.55035281e-01 -2.84969568e-01 1.32344246e-01
-4.28849816e-01 -1.09507874e-01 1.20311642e+00 -9.96710718e-01
1.28813088e+00 -2.10242510e+00 4.22435313e-01 4.70944382e-02
3.22447091e-01 5.59382617e-01 -1.86118275e-01 7.56163970e-02
-7.00139850e-02 -5.91755658e-02 -1.21959612e-01 -9.44570154e-02
-4.07318622e-01 5.59322357e-01 9.93836001e-02 8.13439846e-01
2.24877790e-01 7.65973330e-01 -9.12323415e-01 -3.98706436e-01
3.27983648e-01 4.15347546e-01 -6.81955218e-01 2.72609890e-01
-2.00593118e-02 5.23330927e-01 -6.14571154e-01 7.48730302e-01
5.69542587e-01 -8.76890346e-02 -1.01038836e-01 -3.94688517e-01
2.80169398e-01 -4.78689790e-01 -1.24859202e+00 1.52453065e+00
3.86067539e-01 4.01427060e-01 4.93162349e-02 -1.34177399e+00
5.56038260e-01 3.19875419e-01 7.00627148e-01 -5.31586170e-01
1.88234270e-01 -3.33589762e-01 1.37069017e-01 -9.14220333e-01
1.84543610e-01 5.01773596e-01 1.28887877e-01 -7.30613694e-02
1.32303074e-01 6.04519427e-01 3.04097265e-01 2.54824698e-01
1.13011801e+00 1.17287338e-01 2.41797134e-01 6.30699247e-02
8.30702484e-01 -2.97312587e-01 9.23997641e-01 7.08994269e-01
-6.17114902e-01 7.61599123e-01 7.57822394e-01 -1.12645614e+00
-7.83237815e-01 -1.07698512e+00 3.85243207e-01 8.01113665e-01
3.42823058e-01 -4.44123000e-01 -6.15047455e-01 -1.05439687e+00
-1.24788746e-01 2.66858786e-01 -9.45369661e-01 -3.67718458e-01
-8.12419713e-01 -9.42737579e-01 5.50849080e-01 6.50542736e-01
8.28598499e-01 -1.00262213e+00 -5.90179920e-01 5.43900905e-03
-4.30231035e-01 -1.43863690e+00 -4.88287747e-01 -6.43439889e-01
-8.87453496e-01 -1.60232627e+00 -5.51361740e-01 -5.29576838e-01
9.45690155e-01 3.31526875e-01 9.86848116e-01 5.00115454e-01
-5.13388753e-01 7.77252853e-01 -5.53045094e-01 -3.96867871e-01
-4.11416851e-02 -4.06328589e-01 1.91919386e-01 6.82805419e-01
2.66214788e-01 -4.47060853e-01 -7.95889497e-01 2.95188099e-01
-9.27417040e-01 -2.21845105e-01 5.01758695e-01 7.94920802e-01
5.61991870e-01 1.68591350e-01 1.09752215e-01 -6.78568363e-01
2.80640155e-01 -6.80455208e-01 -4.38137949e-01 1.85749814e-01
-7.48492032e-02 -2.52506107e-01 7.16503918e-01 -2.70115107e-01
-7.05201030e-01 -1.28813609e-01 -1.10304169e-01 -8.55375350e-01
-3.83039474e-01 7.91487247e-02 -6.83015287e-02 1.84466112e-02
3.91283512e-01 2.96863675e-01 -1.01997647e-02 -1.58523336e-01
3.30724083e-02 -9.05412585e-02 6.80134475e-01 -3.01272154e-01
8.05569232e-01 5.28861582e-01 1.35846600e-01 -9.13938284e-01
-7.47325659e-01 -5.28725803e-01 -7.57941186e-01 -6.03997111e-01
1.36134577e+00 -8.60208869e-01 -3.30937147e-01 8.04258108e-01
-9.19795334e-01 -1.10617064e-01 9.40743182e-03 5.58011472e-01
-4.40688789e-01 9.04388428e-01 -6.10113323e-01 -5.79841554e-01
-1.96297452e-01 -1.03933930e+00 1.04495609e+00 2.79118299e-01
1.12263262e-01 -1.14500034e+00 -2.18957603e-01 2.33175099e-01
1.23313077e-01 9.96631205e-01 7.63175428e-01 -5.66017807e-01
-5.28854251e-01 -3.58231008e-01 -1.41502157e-01 2.19325989e-01
2.69275159e-01 6.25676885e-02 -5.70771456e-01 -6.31876349e-01
-2.55310923e-01 3.29669006e-02 8.67543221e-01 6.94270015e-01
1.55428624e+00 -6.70335889e-01 -4.65163559e-01 9.50466871e-01
1.09614921e+00 2.59237327e-02 8.40330839e-01 3.64230156e-01
1.35836291e+00 4.49699849e-01 3.93185467e-01 3.30519378e-01
2.55203694e-01 6.97785616e-01 9.00251627e-01 -2.07725838e-01
-2.40917280e-01 -7.93438777e-02 6.37208760e-01 4.75845397e-01
-7.20492542e-01 -3.39239657e-01 -7.95862973e-01 5.35269916e-01
-2.27978015e+00 -1.23578382e+00 -4.59538072e-01 1.92448354e+00
-1.13058552e-01 2.34423786e-01 4.38677251e-01 2.70930515e-03
6.74524128e-01 5.49665391e-01 -4.80959177e-01 -1.01600297e-01
-1.45249829e-01 -3.78284991e-01 3.68448317e-01 1.67881772e-01
-1.35792220e+00 7.65788794e-01 5.70659208e+00 4.15580541e-01
-8.61191273e-01 -9.42853466e-02 5.51913798e-01 3.35746370e-02
2.63867676e-01 -4.73273903e-01 -4.81934994e-01 6.91931307e-01
5.49360514e-01 2.21779451e-01 1.83518931e-01 6.21686935e-01
1.18311211e-01 3.49753737e-01 -8.40774238e-01 1.03315008e+00
2.85789400e-01 -1.02855766e+00 5.51059663e-01 -1.15442082e-01
5.62375844e-01 3.63905579e-02 -6.59027845e-02 3.13057154e-01
1.45434096e-01 -9.80646610e-01 4.79943246e-01 6.94758534e-01
3.10163885e-01 -7.42159128e-01 9.83444035e-01 1.19135439e-01
-1.54140437e+00 -2.94726282e-01 -5.42060196e-01 4.19303291e-02
1.48321897e-01 -4.21341974e-03 -4.38144177e-01 9.05495822e-01
1.15439415e+00 1.17580462e+00 -9.05347109e-01 1.03020656e+00
-1.50496855e-01 4.57341194e-01 -1.67141203e-02 5.98623395e-01
6.05836511e-01 -3.14586759e-01 8.30780685e-01 1.29675853e+00
4.77952302e-01 2.49353155e-01 5.98926306e-01 4.66119230e-01
2.85726756e-01 -1.30733490e-01 -9.16900158e-01 2.86749184e-01
-2.65325963e-01 1.07133496e+00 -5.98208785e-01 -3.32609177e-01
-8.20073605e-01 1.18873715e+00 1.98772073e-01 5.85707366e-01
-1.09046996e+00 9.47465971e-02 9.13594604e-01 2.84106284e-01
3.61465722e-01 -3.01963478e-01 4.46481824e-01 -1.43514097e+00
6.40293658e-02 -9.43691254e-01 1.06866062e+00 -5.11024296e-01
-1.34215629e+00 5.98904192e-01 2.03626871e-01 -1.36823642e+00
-2.25315437e-01 -9.77133095e-01 -8.82539988e-01 3.62577409e-01
-1.38628805e+00 -1.28247595e+00 -7.63670981e-01 1.21206069e+00
8.63800049e-01 -4.72049594e-01 4.80422825e-01 2.66426921e-01
-9.09895778e-01 5.43006957e-01 -3.90915692e-01 8.24527800e-01
1.93814829e-01 -1.15197098e+00 5.36824882e-01 1.42898917e+00
1.23380601e-01 1.62937194e-01 5.52412271e-01 -8.93516421e-01
-1.16142333e+00 -1.41350555e+00 2.37524003e-01 -3.05718184e-01
5.18997014e-01 -5.18743042e-03 -1.16507006e+00 9.96647358e-01
2.21536332e-03 6.64296508e-01 4.44196075e-01 -1.52828321e-01
-3.57704908e-01 -1.08956292e-01 -1.12415791e+00 6.42554402e-01
1.35500550e+00 -9.78203937e-02 -4.70120281e-01 4.57620561e-01
3.62235993e-01 -5.49146116e-01 -4.99824673e-01 6.75636888e-01
3.02572131e-01 -1.37473142e+00 1.21338344e+00 -1.06216311e+00
1.66106746e-01 -5.40840268e-01 2.48405002e-02 -1.21274674e+00
-6.92737997e-01 -2.33507127e-01 -6.68589652e-01 6.93489909e-01
-7.75968581e-02 -6.19153619e-01 7.16250956e-01 2.15966538e-01
-3.34342539e-01 -8.53663921e-01 -7.45735049e-01 -7.09524870e-01
-5.39138675e-01 -1.60047829e-01 3.83285731e-01 1.03468919e+00
-4.09934670e-01 -3.27613726e-02 -7.77848125e-01 8.24111760e-01
7.37919271e-01 -3.42689037e-01 6.71129823e-01 -1.19852448e+00
-1.39215499e-01 -5.05345345e-01 -1.28623402e+00 -7.96525419e-01
2.76985411e-02 -5.40151000e-01 -4.10628021e-01 -1.63488424e+00
2.37975456e-03 4.53890972e-02 -4.50340509e-01 4.62431222e-01
-5.07559180e-01 2.63851136e-01 -3.60646024e-02 -1.51909478e-02
-7.11708486e-01 4.37758148e-01 1.19390833e+00 -2.61887074e-01
4.65267226e-02 9.52388793e-02 -1.65863693e-01 1.16777408e+00
7.25186706e-01 -8.15053284e-02 -5.07833421e-01 -6.44206762e-01
-1.61182135e-01 -1.43529400e-01 7.26628900e-01 -1.43960440e+00
2.09539190e-01 -5.86558282e-02 8.80677044e-01 -5.37756979e-01
1.92547396e-01 -9.78193998e-01 6.86194450e-02 7.32590914e-01
7.08276629e-02 5.69673538e-01 3.04773629e-01 1.24143958e+00
-3.01664561e-01 1.61457703e-01 7.52841413e-01 -4.62334812e-01
-1.13140774e+00 9.49342787e-01 -4.87053633e-01 3.10207630e-04
1.29762948e+00 -4.82178450e-01 -7.89565369e-02 -4.94325757e-01
-8.20185959e-01 3.19346875e-01 2.25860015e-01 8.31972539e-01
8.66264403e-01 -1.42834282e+00 -7.57734776e-01 4.74814117e-01
1.43885538e-01 4.59197052e-02 5.85836887e-01 7.63660014e-01
-8.51810932e-01 1.86232835e-01 -7.30325282e-01 -7.93267548e-01
-1.31376410e+00 7.06586540e-01 7.20393062e-01 -3.02424878e-01
-1.07512343e+00 7.51053631e-01 5.21447241e-01 -4.12757955e-02
3.42593431e-01 -4.52840239e-01 -8.03912461e-01 2.08505243e-02
5.70634663e-01 5.00857592e-01 -9.04091299e-02 -1.09378541e+00
-3.16343248e-01 8.90227020e-01 -1.52512446e-01 6.21244609e-01
1.13126826e+00 -4.83123213e-02 -4.01821658e-02 6.53734878e-02
7.76758552e-01 -2.92041600e-01 -1.45104265e+00 -3.80017906e-02
-3.01832736e-01 -5.71430445e-01 -1.23967469e-01 -2.07718581e-01
-1.62661862e+00 7.64059722e-01 7.68311441e-01 2.91229904e-01
1.23911881e+00 -3.70433837e-01 8.07351053e-01 1.92464858e-01
2.24562317e-01 -8.71321380e-01 5.09058237e-01 2.93229610e-01
8.03024173e-01 -1.34261453e+00 7.06167519e-03 -3.67334396e-01
-6.47165596e-01 1.31121552e+00 1.14253092e+00 -1.85202122e-01
4.90089267e-01 -3.33244860e-01 -2.60787215e-02 -6.16061091e-01
-2.48473704e-01 -2.20998198e-01 7.41805136e-01 8.09766591e-01
1.59742087e-01 -1.20602675e-01 9.56241190e-02 3.50948900e-01
2.53275871e-01 -5.16232133e-01 4.49360698e-01 6.27057195e-01
-3.49074990e-01 -6.38871610e-01 -4.34603840e-01 5.59829295e-01
-6.40988410e-01 3.38084191e-01 -2.93939412e-01 8.98425281e-01
4.79723692e-01 9.38983381e-01 2.92288601e-01 -3.59003127e-01
6.33588970e-01 -1.68737307e-01 2.66058266e-01 -4.68286633e-01
-2.77551740e-01 -2.15899333e-01 -1.48465419e-02 -1.12001610e+00
-4.12791550e-01 -5.88151097e-01 -1.03703415e+00 -2.11842149e-01
2.38042772e-01 -1.46297619e-01 -1.44164339e-01 8.20861399e-01
3.48916411e-01 8.56865764e-01 1.91097617e-01 -9.05693591e-01
4.55737673e-02 -7.29669273e-01 -4.34122741e-01 8.64325583e-01
6.97711647e-01 -8.79775405e-01 -2.19807133e-01 5.99651448e-02] | [7.859288215637207, 1.5446853637695312] |
715ece65-0744-499c-a4e5-bff8b1262c5c | learning-formation-energy-of-inorganic | 1901.06016 | null | http://arxiv.org/abs/1901.06016v2 | http://arxiv.org/pdf/1901.06016v2.pdf | Learning formation energy of inorganic compounds using matrix variate deep Gaussian process | Future advancement of engineering applications is dependent on design of
novel materials with desired properties. Enormous size of known chemical space
necessitates use of automated high throughput screening to search the desired
material. The high throughput screening uses quantum chemistry calculations to
predict material properties, however, computational complexity of these
calculations often imposes prohibitively high cost on the search for desired
material. This critical bottleneck is resolved by using deep machine learning
to emulate the quantum computations. However, the deep learning algorithms
require a large training dataset to ensure an acceptable generalization, which
is often unavailable a-priory. In this paper, we propose a deep Gaussian
process based approach to develop an emulator for quantum calculations. We
further propose a novel molecular descriptor that enables implementation of the
proposed approach. As demonstrated in this paper, the proposed approach can be
implemented using a small dataset. We demonstrate efficacy of our approach for
prediction of formation energy of inorganic molecules. | ['Saket Mishra', 'Piyush Tagade'] | 2018-12-22 | null | null | null | null | ['formation-energy'] | ['miscellaneous'] | [ 2.37902135e-01 -3.68088573e-01 2.17452630e-01 -1.55419111e-01
-6.87172592e-01 -4.41716164e-01 5.10831833e-01 5.27015686e-01
-4.50356632e-01 1.23294449e+00 -4.32388633e-01 -3.73866796e-01
-7.72382393e-02 -1.24738514e+00 -7.59842575e-01 -9.41612840e-01
1.90984622e-01 4.61630851e-01 2.27534413e-01 -1.76739037e-01
8.07382107e-01 8.31375122e-01 -1.38252497e+00 -4.89813425e-02
9.82445121e-01 8.51105332e-01 3.48795742e-01 5.41388571e-01
1.16128638e-01 3.13055694e-01 -2.48820841e-01 -2.56109059e-01
3.91587883e-01 -3.49960297e-01 -6.70432985e-01 -3.15514535e-01
-4.93149157e-04 -2.62641340e-01 4.49758768e-02 9.67087805e-01
8.66664112e-01 5.42504549e-01 1.26566303e+00 -6.49693727e-01
-5.70471406e-01 1.49701029e-01 -2.31362417e-01 -7.34846294e-02
4.26804006e-01 2.92690575e-01 7.73238480e-01 -9.21411216e-01
4.37997848e-01 7.77692437e-01 5.98992527e-01 3.34886670e-01
-1.28244185e+00 -8.83352160e-01 -6.68344080e-01 1.43719539e-01
-1.52106988e+00 -3.75699759e-01 8.17729414e-01 -2.50674814e-01
1.36842668e+00 -6.12673722e-02 5.09577513e-01 7.08563864e-01
4.33858693e-01 8.59794915e-02 1.28221130e+00 -4.07796085e-01
7.12554455e-01 4.10185128e-01 -2.16522422e-02 6.96824789e-01
4.28640723e-01 3.48914057e-01 -4.05688018e-01 -2.29785413e-01
3.53133351e-01 -1.28685489e-01 3.02941293e-01 -1.62034884e-01
-5.99163294e-01 8.89417052e-01 4.39532816e-01 2.47707986e-03
-5.79849899e-01 1.36988461e-01 2.72664785e-01 6.73121139e-02
-1.09082088e-01 7.63132453e-01 -3.94708246e-01 -1.70961961e-01
-9.26863253e-01 3.48265052e-01 7.60915756e-01 6.47028804e-01
8.75272214e-01 1.26209900e-01 3.07540059e-01 3.82458538e-01
4.31332856e-01 2.44787619e-01 2.10962519e-01 -5.03891468e-01
5.13818078e-02 5.18299997e-01 2.43469864e-01 -6.67835176e-01
-4.90687430e-01 -3.34841698e-01 -7.26841271e-01 4.37385738e-01
1.81631431e-01 -2.51911342e-01 -9.68275964e-01 1.18531752e+00
6.39597535e-01 -1.85617283e-01 3.58519763e-01 4.28269804e-01
5.69441259e-01 1.02070260e+00 4.73954916e-01 -3.32021832e-01
9.13276434e-01 -5.33592880e-01 -3.44002455e-01 6.53151572e-01
6.18243694e-01 -1.10472155e+00 6.98891401e-01 6.89116418e-01
-9.43032086e-01 -4.98934150e-01 -1.28467929e+00 1.05480924e-01
-7.14261651e-01 1.69339448e-01 1.12009156e+00 8.72319162e-01
-5.97208619e-01 1.29207742e+00 -8.31545711e-01 -2.21387446e-01
4.56920862e-01 1.16318738e+00 -4.41213012e-01 3.29610854e-01
-9.90322471e-01 8.49937916e-01 8.16024125e-01 9.42291021e-02
-1.08343744e+00 -4.10061836e-01 -3.15475434e-01 -3.19181047e-02
1.23199392e-02 -5.83995938e-01 8.94139051e-01 -4.66677248e-01
-2.00943327e+00 3.86123866e-01 -4.43506278e-02 -4.62652802e-01
2.12062210e-01 -9.54617094e-03 -2.32047990e-01 -3.38059524e-03
-1.98210239e-01 4.47250992e-01 5.85536301e-01 -7.43834555e-01
-2.05233946e-01 -3.63236308e-01 -6.81500807e-02 2.39658654e-01
-1.90984070e-01 -4.45240662e-02 3.56463134e-01 -1.90412737e-02
6.48703426e-03 -8.60685170e-01 -6.15527391e-01 -4.49625880e-01
-5.41667223e-01 -2.13448644e-01 3.95951003e-01 -1.68745860e-01
9.51742470e-01 -1.63043118e+00 -1.86920926e-01 4.56839353e-01
7.52389356e-02 2.69166946e-01 3.82600814e-01 9.54115748e-01
2.64090523e-02 5.91206923e-02 -4.75462303e-02 3.44270706e-01
-1.54402331e-01 -3.89816970e-01 6.93601221e-02 5.91299176e-01
3.60839605e-01 4.57223207e-01 -5.20226717e-01 -4.14025664e-01
4.20570970e-01 6.56298459e-01 -8.64433289e-01 1.54869169e-01
-2.31635019e-01 5.13794363e-01 -9.34117138e-01 6.65936887e-01
9.25187528e-01 -2.18486786e-01 -1.94824800e-01 -6.11133933e-01
-4.81915236e-01 3.51795435e-01 -1.22292149e+00 1.41760123e+00
-1.66449174e-01 -2.91913971e-02 -5.03595054e-01 -1.26476395e+00
1.08766246e+00 2.77919233e-01 4.43090767e-01 -4.36380446e-01
5.62765837e-01 4.45185065e-01 5.02122700e-01 -2.79998243e-01
3.21234316e-01 -6.07583344e-01 3.33824426e-01 1.96056366e-01
7.29872659e-02 -2.40646377e-01 2.48192847e-01 -2.20586210e-01
8.53930473e-01 4.58996445e-01 4.49500203e-01 -2.84987301e-01
8.50173891e-01 3.40869933e-01 1.26799285e-01 6.11846268e-01
-1.95404723e-01 4.21221890e-02 -4.18592319e-02 -5.46883643e-01
-1.41301501e+00 -9.90317404e-01 -5.25810897e-01 6.78466737e-01
1.10605851e-01 -1.43000662e-01 -7.53576279e-01 -1.26242757e-01
-1.92776024e-01 3.59040141e-01 -3.33189666e-01 -2.45104820e-01
-3.76389802e-01 -1.21396053e+00 2.34932363e-01 1.80797204e-01
4.19807434e-01 -1.09485543e+00 -4.17308211e-01 6.38867497e-01
8.12362313e-01 -7.71664143e-01 3.78886372e-01 5.50782859e-01
-9.80848372e-01 -9.05725241e-01 -3.24405134e-01 -6.76565886e-01
5.94580412e-01 -2.51463622e-01 5.17287433e-01 -8.95495564e-02
-7.76581407e-01 -3.95128191e-01 -1.92424402e-01 -5.88028193e-01
-6.49169326e-01 7.41673857e-02 2.57112175e-01 -3.44658315e-01
7.07751572e-01 -8.18570673e-01 -1.10804594e+00 -2.74671912e-01
-4.71054077e-01 4.97757420e-02 8.96008492e-01 7.03208089e-01
6.47437394e-01 4.36858237e-01 8.25985312e-01 -6.57123625e-01
8.02542865e-01 -3.08386117e-01 -9.19997692e-01 1.27311543e-01
-5.66713393e-01 6.28259242e-01 1.00993299e+00 -2.99327940e-01
-1.03688991e+00 5.14710844e-01 -5.30031383e-01 3.94524872e-01
-3.34021509e-01 6.11505687e-01 -5.38549274e-02 -5.71665943e-01
5.89326978e-01 3.08620244e-01 -3.16188008e-01 -4.41871643e-01
1.95258092e-02 7.83287525e-01 6.92110360e-02 -8.84635448e-01
7.71997988e-01 2.21681193e-01 9.02450204e-01 -9.97173250e-01
-3.67369503e-01 -1.62846461e-01 -6.39576018e-01 1.24645568e-01
9.97093439e-01 -6.77085817e-01 -1.19057345e+00 9.38985348e-02
-1.03682840e+00 2.87213504e-01 2.86820501e-01 7.16677308e-01
-4.35654849e-01 3.67076427e-01 -3.47256392e-01 -1.01057780e+00
-8.72918069e-01 -1.56012285e+00 7.85737574e-01 3.19271952e-01
-1.84853628e-01 -7.14356303e-01 2.02528194e-01 2.72334605e-01
3.36438358e-01 5.08473992e-01 9.77624893e-01 -6.71924829e-01
-9.55438316e-01 -6.38594270e-01 -2.08926231e-01 2.00245872e-01
2.48538002e-01 1.64700270e-01 -9.15628433e-01 -7.37302229e-02
-1.17739476e-01 -4.46653843e-01 5.03663540e-01 4.60464060e-01
1.18034530e+00 2.91860968e-01 -2.11858273e-01 4.55624074e-01
1.81330585e+00 6.55708790e-01 6.09351218e-01 3.32620144e-01
6.64114237e-01 1.78600535e-01 7.29410470e-01 6.53871894e-01
-1.80190384e-01 2.82580405e-01 2.47314498e-01 1.95614904e-01
3.05111617e-01 -1.90410599e-01 4.37026583e-02 6.64469063e-01
-5.08635998e-01 1.24410037e-02 -9.23098326e-01 1.11248747e-01
-1.29486465e+00 -8.70807946e-01 -2.10284844e-01 2.46240997e+00
1.13361430e+00 1.80087775e-01 -4.40022610e-02 1.88798279e-01
3.54212075e-01 -6.66334152e-01 -6.33530021e-01 -7.79683888e-01
2.64358312e-01 1.04340160e+00 6.17894650e-01 5.22149265e-01
-1.07500398e+00 1.13351178e+00 6.53779745e+00 1.05826592e+00
-1.36938405e+00 -1.85384348e-01 4.57625836e-01 2.17717648e-01
1.00767940e-01 3.27953666e-01 -1.13489044e+00 2.03780785e-01
1.25006485e+00 -3.40065956e-01 2.96352297e-01 8.09810102e-01
4.16192949e-01 -2.97170490e-01 -1.18842101e+00 9.87451315e-01
-6.70605838e-01 -1.62484372e+00 1.60097420e-01 1.74112976e-01
7.77618766e-01 -2.12795258e-01 6.21438883e-02 2.52898298e-02
-6.51337206e-02 -1.20383275e+00 7.33918846e-02 2.73760438e-01
4.50909555e-01 -1.03106189e+00 4.18204308e-01 2.23534808e-01
-8.78067195e-01 1.11046284e-01 -6.72622740e-01 -3.00770342e-01
-1.31365165e-01 2.63705522e-01 -1.15406859e+00 3.40323418e-01
6.68110102e-02 2.26668343e-01 -2.66947895e-01 1.17175746e+00
4.42608595e-01 3.06928307e-01 -5.33019602e-01 -6.83963537e-01
4.16033566e-01 -6.63588583e-01 1.13343798e-01 9.43457484e-01
5.04151285e-01 -3.24128158e-02 1.54229224e-01 9.66082335e-01
-1.94194317e-02 6.29806757e-01 -6.89313352e-01 -5.25502086e-01
3.77553672e-01 1.08444452e+00 -8.39003742e-01 -2.91161418e-01
-3.45445752e-01 7.54478872e-01 2.08554745e-01 1.01128072e-01
-7.60084867e-01 -5.32085776e-01 2.29788795e-01 2.27579266e-01
1.77725255e-01 -2.74251759e-01 -7.13120997e-02 -6.81318104e-01
-4.30551291e-01 -6.94939077e-01 -1.88122168e-01 -4.08865541e-01
-9.73260343e-01 4.36690450e-01 -1.51149958e-01 -9.62575376e-01
1.11050494e-01 -1.08537900e+00 -6.35598063e-01 1.02092075e+00
-1.26780653e+00 -8.94280732e-01 2.52908878e-02 5.16083956e-01
2.35062733e-01 -3.89231086e-01 1.01316512e+00 3.72993231e-01
-5.87490261e-01 3.28784049e-01 4.55656022e-01 -4.51739013e-01
6.31345987e-01 -1.25608301e+00 5.86278876e-03 4.23171073e-01
-2.91506916e-01 9.63325441e-01 1.12197781e+00 -7.73651302e-01
-1.84683144e+00 -7.71195292e-01 2.69643068e-01 -2.53179073e-01
7.16576695e-01 -2.04488531e-01 -5.33231974e-01 -5.96663915e-02
1.42644510e-01 -2.75681555e-01 1.08185029e+00 -4.32850271e-02
2.99783826e-01 -8.31196085e-03 -1.30886638e+00 5.08919001e-01
4.47599083e-01 -4.82578307e-01 -2.45935246e-01 6.59118414e-01
3.41704488e-01 -2.34263595e-02 -1.18051255e+00 3.89355093e-01
7.25677490e-01 -8.77936423e-01 9.91845787e-01 -7.10488141e-01
3.75857174e-01 -3.38997692e-01 -2.47979969e-01 -6.80047333e-01
-7.09832087e-02 -7.13609278e-01 1.46756589e-03 8.23680222e-01
6.80146813e-01 -2.81598866e-01 1.23378766e+00 9.91619408e-01
9.63720828e-02 -8.25007915e-01 -6.76166892e-01 -6.04209006e-01
4.25004512e-01 -1.95790172e-01 4.77166474e-01 5.71864307e-01
-1.49114236e-01 6.17941678e-01 -3.79318565e-01 1.91563219e-01
7.43112266e-01 3.21094953e-02 5.80384076e-01 -1.29825377e+00
-4.11284298e-01 -1.46992821e-02 -5.71273744e-01 -4.83927727e-01
-1.67947516e-01 -6.85010910e-01 -3.00620526e-01 -1.29952621e+00
4.29940373e-01 -3.44759166e-01 -4.79884565e-01 -1.32026285e-01
1.22128099e-01 4.47491929e-02 -3.70638996e-01 -1.25094488e-01
-3.03194642e-01 7.14064360e-01 1.28516722e+00 8.11676532e-02
-4.30847257e-01 5.38652651e-02 -4.21430200e-01 5.97662985e-01
1.18454385e+00 -5.67656696e-01 -4.58867282e-01 2.42638558e-01
4.85120893e-01 -1.86429664e-01 3.89564373e-02 -1.30838478e+00
2.25067049e-01 -2.77047783e-01 7.00389624e-01 -8.68460834e-01
5.50837755e-01 -7.21998930e-01 2.44848818e-01 6.40464306e-01
-5.70993908e-02 -2.83902049e-01 1.20295703e-01 4.33654457e-01
-1.75716374e-02 -5.95823526e-01 9.17624533e-01 -3.22153836e-01
-4.01421338e-01 5.06189764e-01 -3.56814802e-01 -6.41115606e-01
1.12548780e+00 -5.66211760e-01 2.25151360e-01 1.86713830e-01
-8.02494407e-01 -3.27939868e-01 4.47885782e-01 -3.20960045e-01
6.08996272e-01 -1.09168088e+00 -1.76096261e-01 -3.72508075e-04
-1.57825992e-01 -3.55432957e-01 1.03577852e-01 5.78478992e-01
-1.20738065e+00 5.58882296e-01 -4.59425867e-01 -1.17105581e-01
-1.15570509e+00 6.77866042e-01 2.88401395e-01 -1.30004779e-01
-1.55822590e-01 7.16022372e-01 -7.14060664e-02 -2.11173132e-01
-8.00282806e-02 -1.15412809e-01 -9.62599143e-02 -3.23626548e-01
2.52292126e-01 2.87568659e-01 5.31305596e-02 -4.69037443e-01
-3.80092323e-01 6.89195633e-01 -3.13358337e-01 1.46161422e-01
1.62728179e+00 1.67556152e-01 -1.51963830e-01 -8.50519314e-02
1.35963774e+00 -1.47877321e-01 -9.94805574e-01 3.01636070e-01
-9.40200873e-03 -1.36139438e-01 3.97471875e-01 -5.67437410e-01
-2.53117979e-01 1.07914948e+00 9.58224118e-01 -2.26808697e-01
9.12529349e-01 -5.66300333e-01 7.35514700e-01 1.08723295e+00
3.42975438e-01 -1.36546659e+00 -2.28016868e-01 1.87989071e-01
4.41627175e-01 -1.70773470e+00 4.00443435e-01 -3.19333494e-01
-1.24382220e-01 1.52168691e+00 5.00626743e-01 -3.44096184e-01
8.97221684e-01 3.62693854e-02 -5.48217595e-01 -4.89499927e-01
-3.87638122e-01 -1.50194436e-01 1.48203835e-01 4.92622823e-01
8.66190314e-01 -6.53507710e-02 -6.83387995e-01 2.27366641e-01
-1.26708820e-01 1.15567110e-01 4.63592291e-01 1.16667616e+00
-6.94466114e-01 -1.53772223e+00 -1.98246405e-01 3.43043804e-01
-7.88774788e-01 -4.42464054e-01 -3.90563279e-01 6.17355227e-01
1.62154540e-01 7.54926920e-01 -5.49641311e-01 -2.46033177e-01
-8.42648298e-02 1.70287341e-02 7.49151468e-01 -5.45994520e-01
-3.44911605e-01 2.58571189e-02 1.06230810e-01 -1.67987093e-01
-3.77254367e-01 -3.10179502e-01 -1.39835846e+00 -4.08546686e-01
-6.30116165e-01 4.01803643e-01 9.68564391e-01 9.90719914e-01
2.72610456e-01 1.48776606e-01 5.44849098e-01 -7.94067204e-01
-6.24256909e-01 -7.16340363e-01 -6.08564675e-01 2.26999730e-01
1.75538694e-03 -8.14310789e-01 1.15893632e-02 -4.66937274e-02] | [5.190478324890137, 5.394010543823242] |
acf087d4-c030-40b7-ab33-bab7a3bebad8 | lea-improving-sentence-similarity-robustness | 2307.02912 | null | https://arxiv.org/abs/2307.02912v1 | https://arxiv.org/pdf/2307.02912v1.pdf | LEA: Improving Sentence Similarity Robustness to Typos Using Lexical Attention Bias | Textual noise, such as typos or abbreviations, is a well-known issue that penalizes vanilla Transformers for most downstream tasks. We show that this is also the case for sentence similarity, a fundamental task in multiple domains, e.g. matching, retrieval or paraphrasing. Sentence similarity can be approached using cross-encoders, where the two sentences are concatenated in the input allowing the model to exploit the inter-relations between them. Previous works addressing the noise issue mainly rely on data augmentation strategies, showing improved robustness when dealing with corrupted samples that are similar to the ones used for training. However, all these methods still suffer from the token distribution shift induced by typos. In this work, we propose to tackle textual noise by equipping cross-encoders with a novel LExical-aware Attention module (LEA) that incorporates lexical similarities between words in both sentences. By using raw text similarities, our approach avoids the tokenization shift problem obtaining improved robustness. We demonstrate that the attention bias introduced by LEA helps cross-encoders to tackle complex scenarios with textual noise, specially in domains with short-text descriptions and limited context. Experiments using three popular Transformer encoders in five e-commerce datasets for product matching show that LEA consistently boosts performance under the presence of noise, while remaining competitive on the original (clean) splits. We also evaluate our approach in two datasets for textual entailment and paraphrasing showing that LEA is robust to typos in domains with longer sentences and more natural context. Additionally, we thoroughly analyze several design choices in our approach, providing insights about the impact of the decisions made and fostering future research in cross-encoders dealing with typos. | ['David Jiménez', 'Diego Ortego', 'Emilio Almazán', 'Mario Almagro'] | 2023-07-06 | null | null | null | null | ['natural-language-inference'] | ['natural-language-processing'] | [ 4.21407133e-01 -2.41219342e-01 -7.07494169e-02 -3.20655704e-01
-9.37213123e-01 -7.28785336e-01 5.72824359e-01 5.10541081e-01
-7.20517874e-01 4.39179361e-01 4.77443993e-01 -4.45196599e-01
-1.95122540e-01 -8.20482552e-01 -1.00029385e+00 -3.47060770e-01
4.78979468e-01 3.54072779e-01 -8.69494956e-03 -5.21030426e-01
1.05642028e-01 1.20074585e-01 -1.59152007e+00 6.48799300e-01
1.04024982e+00 9.63051200e-01 3.68549228e-01 3.29782218e-01
-3.05043250e-01 6.08553946e-01 -6.27400398e-01 -9.69031930e-01
3.69924247e-01 -3.16578895e-01 -8.82345498e-01 -4.64873947e-02
7.12797582e-01 7.73439696e-03 -2.19139948e-01 1.05719578e+00
6.73103392e-01 1.99696440e-02 4.81020272e-01 -8.26107979e-01
-7.11337090e-01 1.08220303e+00 -3.93158048e-01 3.36485922e-01
5.38861573e-01 2.36593008e-01 1.41810203e+00 -8.80906463e-01
6.86972618e-01 1.38551509e+00 1.14023042e+00 3.03587854e-01
-1.56777179e+00 -4.88495529e-01 5.15348427e-02 3.22506338e-01
-1.07292438e+00 -4.19903785e-01 5.79217792e-01 -3.10978055e-01
1.23936093e+00 1.65900692e-01 7.64422657e-05 1.56561279e+00
5.81338033e-02 7.90060103e-01 8.40290129e-01 -5.73779881e-01
2.17175726e-02 4.14044589e-01 2.10029155e-01 8.16096142e-02
3.21645766e-01 -1.48269370e-01 -3.42187077e-01 -7.07608610e-02
1.06710546e-01 -1.58839032e-01 -2.62483865e-01 -2.66247630e-01
-1.09053206e+00 1.02489364e+00 2.27389798e-01 7.30002046e-01
-3.81438017e-01 -1.71837881e-01 9.66219068e-01 7.79978096e-01
4.54069883e-01 8.00657034e-01 -5.79820991e-01 -5.96201792e-02
-9.77598667e-01 2.87138194e-01 7.84386754e-01 1.02311242e+00
5.17301321e-01 -1.55022115e-01 -4.52576667e-01 1.13367867e+00
-1.96296915e-01 3.13982159e-01 9.97535884e-01 -5.23080528e-01
9.24426794e-01 5.01148283e-01 -1.10162981e-01 -9.37039733e-01
-2.33553395e-01 -6.84299529e-01 -8.35614204e-01 -2.86155879e-01
5.09021163e-01 -9.96575356e-02 -6.67509973e-01 2.08492470e+00
-1.08027257e-01 -1.29842475e-01 2.08026707e-01 5.92191637e-01
7.25439429e-01 2.74672151e-01 1.65175855e-01 9.71571580e-02
1.55299604e+00 -8.71446133e-01 -7.62228847e-01 -4.67267603e-01
9.88970816e-01 -1.03710365e+00 1.40098727e+00 1.15995020e-01
-1.10454977e+00 -7.08242118e-01 -1.00127840e+00 -3.22032481e-01
-6.33121729e-01 1.14578225e-01 1.09282136e-01 5.97524047e-01
-6.83852553e-01 1.18978202e+00 -2.68205106e-01 -5.99998891e-01
2.82280356e-01 1.94678038e-01 -4.01508689e-01 -1.70160875e-01
-1.61968088e+00 1.27726018e+00 2.71931469e-01 -8.91229734e-02
-1.16607010e-01 -9.86631870e-01 -1.00286317e+00 3.81578565e-01
4.04165149e-01 -8.76556039e-01 1.31590652e+00 -1.14557791e+00
-1.10267770e+00 7.12987661e-01 -1.40419528e-01 -8.18391860e-01
7.03601599e-01 -3.27241242e-01 -4.07419771e-01 -3.36389065e-01
4.15842175e-01 2.49921262e-01 6.28128409e-01 -7.21768320e-01
-2.85034120e-01 -3.26570690e-01 9.15936008e-02 1.28056034e-01
-4.75016743e-01 2.23544203e-02 -4.38801758e-02 -9.60460722e-01
-3.10235083e-01 -6.92441046e-01 -1.19964004e-01 -4.35261935e-01
-3.75343978e-01 -2.24531665e-01 3.98175389e-01 -6.59610927e-01
1.37542713e+00 -2.17339420e+00 1.71056435e-01 8.96373540e-02
-3.37157190e-01 4.85181928e-01 -5.56868553e-01 7.93358088e-01
-4.00817037e-01 2.61423945e-01 -2.73920596e-01 -6.76480472e-01
2.07775071e-01 2.63773501e-01 -4.10789937e-01 9.17818397e-02
5.57980299e-01 9.76880908e-01 -8.04513276e-01 -2.42276728e-01
-5.28481789e-02 2.85690844e-01 -5.21750808e-01 2.04677023e-02
-2.85166800e-02 4.51172478e-02 4.37943973e-02 2.27176845e-01
7.02591956e-01 3.14028794e-03 2.29625031e-01 -3.15336972e-01
1.28397122e-01 8.89595389e-01 -1.22761035e+00 1.76451468e+00
-9.77307975e-01 3.87473792e-01 6.04240857e-02 -1.09975803e+00
7.77179658e-01 2.24065602e-01 1.01989970e-01 -1.10593235e+00
1.33259654e-01 6.08489394e-01 1.33388117e-01 -5.57366908e-01
5.29213011e-01 -4.64898884e-01 -1.73203707e-01 3.77680480e-01
2.56849915e-01 -9.24610347e-02 5.78990042e-01 7.37268552e-02
1.21332359e+00 -9.58702415e-02 2.24483341e-01 -2.70899266e-01
6.76494002e-01 -2.41779372e-01 4.18808281e-01 9.91694272e-01
1.15493178e-01 7.36355782e-01 5.17090976e-01 -1.62751138e-01
-1.22754753e+00 -6.26479805e-01 -2.80715644e-01 9.58930910e-01
-1.62189528e-01 -6.18519068e-01 -7.03358471e-01 -7.88131177e-01
3.27766865e-01 7.50305235e-01 -6.31169319e-01 -4.15496379e-01
-7.45167851e-01 -6.34167612e-01 7.59806156e-01 6.35186374e-01
1.95052132e-01 -9.16596532e-01 -3.79748851e-01 5.21159649e-01
-4.34594661e-01 -1.41094255e+00 -4.97462243e-01 5.97393930e-01
-7.69568324e-01 -9.15750623e-01 -6.46202803e-01 -6.17446661e-01
8.69479626e-02 2.29501858e-01 1.45555830e+00 -7.85630941e-02
-9.45925787e-02 1.36169270e-01 -5.55861473e-01 -2.35250920e-01
-7.55052805e-01 4.65727687e-01 1.32166338e-03 -4.13431786e-02
6.73709810e-01 -6.34654999e-01 -1.46917462e-01 2.14272544e-01
-1.19918871e+00 -2.89605260e-01 6.44911766e-01 1.32673740e+00
1.03356175e-01 -6.77564293e-02 7.34793365e-01 -1.14478004e+00
8.67524087e-01 -7.12876081e-01 -2.41907924e-01 2.03263730e-01
-4.71284747e-01 5.41506767e-01 9.92953897e-01 -5.40651083e-01
-8.59190047e-01 -1.79829329e-01 -4.30444598e-01 -2.30215296e-01
-2.96555251e-01 5.29647768e-01 -3.87028545e-01 3.03125739e-01
7.08465338e-01 1.69449151e-02 6.24818634e-03 -8.33360016e-01
3.65294039e-01 8.19607019e-01 3.23364049e-01 -5.05807817e-01
6.57278001e-01 4.49882336e-02 -2.48700038e-01 -5.41266441e-01
-7.93731689e-01 -5.83398163e-01 -5.77438354e-01 4.67337102e-01
3.64928812e-01 -8.32968771e-01 -4.38440979e-01 2.63527542e-01
-1.30546761e+00 1.79173276e-02 -5.58442950e-01 3.35463226e-01
-4.16955560e-01 6.23243928e-01 -6.02770627e-01 -3.70078832e-01
-3.51031095e-01 -1.21741271e+00 1.08348632e+00 -3.06122959e-01
-5.95647216e-01 -9.74885881e-01 -3.81223783e-02 3.66554916e-01
6.72608852e-01 -1.21631414e-01 1.25043035e+00 -1.17800832e+00
-1.20229989e-01 -8.34513009e-02 -1.47716358e-01 6.40226603e-01
1.70816824e-01 -4.31505173e-01 -1.08166075e+00 -2.30758384e-01
3.99630181e-02 -2.77076244e-01 1.16798246e+00 -6.24875352e-02
8.87133956e-01 -3.25075179e-01 2.58305408e-02 3.47651869e-01
1.35588133e+00 -1.68786377e-01 5.94344020e-01 6.00258291e-01
4.14188564e-01 8.58478010e-01 5.44687212e-01 2.95556456e-01
5.35224527e-02 9.90436316e-01 1.94190934e-01 -1.65520310e-01
-2.77629226e-01 -2.74350673e-01 3.95051688e-01 6.55633986e-01
5.04209876e-01 -2.90435225e-01 -5.92860281e-01 7.67554462e-01
-1.77179527e+00 -1.01202393e+00 -1.98008776e-01 2.20110917e+00
1.14764762e+00 2.04162464e-01 1.14251114e-01 5.37247121e-01
5.63691318e-01 2.31183693e-02 -3.37230355e-01 -7.46571541e-01
-5.05169868e-01 5.14610946e-01 5.25332272e-01 4.44910616e-01
-1.15026021e+00 7.06291914e-01 5.10095501e+00 1.07978833e+00
-9.25588489e-01 1.69652894e-01 3.46127450e-01 -1.00224465e-01
-4.97471392e-01 -1.59574598e-01 -7.15694487e-01 7.62668848e-01
9.88327265e-01 -1.50060520e-01 1.86948970e-01 6.75558329e-01
1.50229484e-01 1.38904192e-02 -1.43276346e+00 6.69133365e-01
3.52803152e-03 -1.02433515e+00 1.34419292e-01 -2.88096100e-01
5.60201705e-01 4.38993834e-02 1.40397966e-01 4.61416215e-01
5.11896014e-02 -9.66942012e-01 7.73609400e-01 -4.44317348e-02
5.60356319e-01 -6.69651568e-01 1.18117332e+00 2.07484707e-01
-9.55843389e-01 -2.01518252e-01 -4.29198027e-01 5.70505001e-02
1.86521597e-02 7.74100721e-01 -7.12053716e-01 9.14537132e-01
5.57355762e-01 7.89725363e-01 -6.93920493e-01 9.30336952e-01
8.18385631e-02 3.33584070e-01 -3.53125274e-01 1.31279588e-01
3.58278483e-01 -4.57789302e-02 5.53029239e-01 1.79282677e+00
3.50047588e-01 -6.26538217e-01 -2.48939022e-01 9.18124437e-01
-1.80479154e-01 4.02376205e-01 -7.98037410e-01 1.17248535e-01
3.31381023e-01 8.92657876e-01 -2.84666955e-01 -2.93524355e-01
-6.39137030e-01 1.21298206e+00 4.01320964e-01 2.38477349e-01
-7.38304496e-01 -5.96650958e-01 7.59853184e-01 2.34974042e-01
6.66952968e-01 2.58484006e-01 -4.55502480e-01 -1.23253191e+00
4.35712993e-01 -1.27529263e+00 4.06947166e-01 -4.27034378e-01
-1.64243722e+00 7.63909280e-01 -1.54411778e-01 -1.22167087e+00
-2.77106673e-01 -6.79650843e-01 -3.98782849e-01 1.01987994e+00
-1.76189506e+00 -1.04218280e+00 1.21897146e-01 3.77126396e-01
7.59714782e-01 6.27420545e-02 6.76155686e-01 7.89665043e-01
-4.80541438e-01 9.53968585e-01 2.75489151e-01 7.93525353e-02
1.05730152e+00 -1.26506758e+00 6.11927032e-01 8.82848263e-01
2.52272308e-01 8.77950430e-01 7.67938554e-01 -5.29298782e-01
-1.08003044e+00 -1.13930190e+00 1.49691844e+00 -3.49755079e-01
8.70274186e-01 -5.78934789e-01 -1.23337758e+00 4.25879270e-01
3.93403620e-01 -3.67967308e-01 5.23183703e-01 2.65419066e-01
-7.26666749e-01 -1.13731913e-01 -9.01452541e-01 4.66807693e-01
1.04427040e+00 -7.88001060e-01 -8.13383341e-01 2.22334266e-01
7.39364326e-01 -2.33135626e-01 -6.40624046e-01 4.02888268e-01
3.36696714e-01 -9.07327175e-01 8.64827693e-01 -7.93633759e-01
6.05798841e-01 6.85667098e-02 -1.65186033e-01 -1.53695631e+00
-2.91493922e-01 -4.50587958e-01 3.78643483e-01 1.62077081e+00
5.88466048e-01 -4.27281618e-01 4.05730724e-01 2.65430063e-01
-5.72279952e-02 -5.27883112e-01 -9.41138983e-01 -1.12203228e+00
3.97110462e-01 -5.21148324e-01 6.54738426e-01 9.11268532e-01
1.10769778e-01 5.78221858e-01 -2.41337538e-01 -2.51950830e-01
7.00204968e-02 -4.73695062e-02 4.56925899e-01 -1.06869352e+00
-4.05063272e-01 -6.32331967e-01 -2.68306613e-01 -9.79133427e-01
4.29618567e-01 -1.09514070e+00 -3.77409868e-02 -1.03017545e+00
1.24865174e-01 -2.95745552e-01 -3.15111168e-02 3.40298086e-01
-3.20117474e-01 1.85936302e-01 3.31477553e-01 -1.95756644e-01
-2.99494177e-01 4.84294146e-01 7.40938604e-01 -2.66834885e-01
2.59361994e-02 5.79330474e-02 -7.80774474e-01 4.86947000e-01
6.70517027e-01 -5.26109934e-01 -1.32675216e-01 -6.76254153e-01
2.97461212e-01 -3.88396680e-01 3.16325575e-01 -8.39080274e-01
1.69553339e-01 3.96426708e-01 -7.73883164e-02 -3.08271974e-01
1.08253546e-01 -1.08655179e+00 -7.71713480e-02 3.57644290e-01
-7.05746233e-01 4.28176433e-01 4.00388241e-01 4.68630821e-01
-4.42399144e-01 -7.12244391e-01 7.45771170e-01 -1.74252167e-01
-2.66881227e-01 -3.11303198e-01 -5.11993170e-01 3.98069501e-01
5.01935661e-01 -1.09439671e-01 -5.24737872e-02 -3.11701804e-01
-5.56607902e-01 1.17229871e-01 3.55195880e-01 6.59717143e-01
1.47003487e-01 -1.30930412e+00 -8.44547808e-01 3.92418206e-01
2.69868463e-01 -3.11232090e-01 1.33145779e-01 8.83773446e-01
8.94298181e-02 5.55135667e-01 2.62152739e-02 -2.82011569e-01
-1.34232199e+00 7.00657129e-01 4.22125787e-01 -7.09371150e-01
-2.60516584e-01 6.52566314e-01 1.38066038e-02 -4.51755166e-01
4.22510356e-01 -7.60437608e-01 -1.20899670e-01 4.95185643e-01
3.13953847e-01 2.61512399e-01 8.06257546e-01 -6.46475196e-01
-2.81207919e-01 6.24844015e-01 -3.10980856e-01 1.18412174e-01
1.34843814e+00 -3.04830760e-01 4.55266982e-02 2.52616853e-01
1.51740348e+00 4.89111952e-02 -5.64485669e-01 -7.36234605e-01
5.02767086e-01 -3.17359209e-01 -1.57808840e-01 -8.08532178e-01
-8.53084981e-01 1.00919056e+00 2.73296326e-01 2.01132104e-01
1.02526808e+00 -2.02719882e-01 1.04711151e+00 3.61474752e-01
7.09547428e-03 -1.03192294e+00 -3.09017375e-02 7.26830006e-01
8.41811419e-01 -1.30200899e+00 -3.42583776e-01 -3.14877778e-01
-6.48433685e-01 1.13588393e+00 3.99283111e-01 2.45090090e-02
2.05805928e-01 3.52045357e-01 -2.72298362e-02 1.47990093e-01
-7.38626957e-01 -5.06151497e-01 2.23253027e-01 4.40801054e-01
5.64505994e-01 -3.74336362e-01 -5.39095998e-01 8.56639266e-01
-1.83434278e-01 -2.03478262e-01 4.84120697e-01 7.49558389e-01
3.37608978e-02 -1.61750829e+00 -2.19046488e-01 3.34729970e-01
-8.17483485e-01 -6.41518652e-01 -4.42205399e-01 6.45627022e-01
3.08065832e-01 9.52892542e-01 1.04589872e-01 -1.93430439e-01
7.87704229e-01 3.84186685e-01 2.54191369e-01 -6.70428157e-01
-1.27894199e+00 -1.33440152e-01 3.40295672e-01 -5.57731271e-01
-2.51952052e-01 -7.75623262e-01 -5.86522043e-01 -1.94308221e-01
-4.76146162e-01 1.65056556e-01 5.18605649e-01 8.58091831e-01
5.48577130e-01 5.27225018e-01 4.86238003e-01 -3.98309588e-01
-1.06542063e+00 -1.17930365e+00 -3.98702085e-01 1.18497026e+00
4.67455298e-01 -5.18743455e-01 -5.00579238e-01 -1.75660089e-01] | [11.034833908081055, 8.927603721618652] |
49f837b3-f773-40c9-bfa7-c774294fc254 | mdg-metaphorical-sentence-detection-and | null | null | https://openreview.net/forum?id=mPLmX6RVDT2 | https://openreview.net/pdf?id=mPLmX6RVDT2 | MDG: Metaphorical Sentence Detection and Generation with Masked Metaphor Modeling | This study tackles literal to metaphorical sentence generation, presenting a framework that can potentially lead to the production of an infinite number of new metaphors. To achieve this goal, we propose a complete workflow that tackles metaphorical sentence classification and metaphor reconstruction. Unlike similar research works regarding metaphor generation, our approach does not require any custom
or closed-source model, hence with this work we introduce a complete literal to metaphorical open-source model. The obtained results show that a good ratio of originally literal sentences, coming from different data sources and topics, are turned to metaphorical. Human evaluation shows that our constructed metaphors are considered more fluent, creative and metaphorical than figurative statements created by a real person. Furthermore, by using our artificial data to increase the training size of a metaphorical sentence classification dataset, we register an improvement of 3% over the baseline. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['sentence-classification'] | ['natural-language-processing'] | [ 5.61583191e-02 4.70146000e-01 4.15524900e-01 -2.34162241e-01
-2.35843077e-01 -7.73211181e-01 1.41204786e+00 4.58496690e-01
-2.51786560e-01 8.70460093e-01 5.95614851e-01 -3.15499723e-01
1.04681879e-01 -1.05216980e+00 -2.67619103e-01 -3.56008351e-01
2.42324919e-01 7.98866510e-01 -2.43000627e-01 -9.65695500e-01
5.74643314e-01 2.00085968e-01 -1.37528741e+00 6.85432017e-01
6.27225697e-01 5.62539756e-01 4.47420388e-01 3.99270475e-01
-7.71481693e-01 4.80438739e-01 -1.22405970e+00 -4.89144087e-01
-2.66774725e-02 -4.28440899e-01 -1.19641590e+00 -4.81821388e-01
2.01447353e-01 2.23417908e-01 6.47088826e-01 7.06338763e-01
4.53710675e-01 3.39566350e-01 7.53134906e-01 -5.31873226e-01
-1.01325703e+00 9.65608835e-01 1.98134109e-02 -3.01531255e-01
9.86234069e-01 -6.31888881e-02 7.17078149e-01 -8.19245756e-01
9.00355458e-01 1.76025629e+00 5.08239210e-01 5.95567107e-01
-1.53908134e+00 -2.33085588e-01 -1.31216526e-01 2.83530429e-02
-9.13876653e-01 -6.18761219e-02 1.13487792e+00 -4.21747923e-01
1.11283362e+00 7.86707699e-01 1.02887595e+00 2.00246024e+00
-1.13109201e-01 1.34560317e-01 1.59647536e+00 -9.94138062e-01
2.60458767e-01 6.91900849e-01 -8.18069354e-02 1.33347258e-01
-1.93776146e-01 -1.05394237e-01 -2.67706156e-01 -3.69962156e-01
4.34165537e-01 -2.80208886e-01 -2.39172131e-01 1.20963618e-01
-1.08372962e+00 1.08139670e+00 5.74391842e-01 1.11504436e+00
-2.65637487e-01 -1.31333947e-01 7.28984177e-01 4.25381303e-01
5.39342582e-01 1.13230538e+00 8.42217579e-02 -3.41361523e-01
-7.03290820e-01 4.05802965e-01 1.04721081e+00 7.13180542e-01
3.67884517e-01 -1.97794691e-01 -1.10742547e-01 1.06822491e+00
-1.12568233e-02 5.05408585e-01 9.56054449e-01 -4.69823480e-01
-2.91452184e-02 9.97760713e-01 1.71806470e-01 -1.34284580e+00
-1.90263882e-01 -1.21623963e-01 -6.25403881e-01 -8.68803412e-02
-7.42132515e-02 3.15483689e-01 -3.02033752e-01 1.50949454e+00
1.07369930e-01 -2.38546178e-01 2.56008238e-01 9.02769983e-01
1.08658588e+00 5.46549320e-01 1.69784740e-01 -2.20000729e-01
1.73886263e+00 -9.36646342e-01 -5.90102613e-01 -3.55554432e-01
3.06736261e-01 -1.18178141e+00 1.99665594e+00 3.38830024e-01
-9.24984634e-01 -5.75611413e-01 -9.01860297e-01 -3.74326974e-01
-1.15472972e+00 -6.50427714e-02 5.59109151e-01 4.89386410e-01
-9.63212788e-01 3.14570814e-01 1.33186001e-02 -8.45761240e-01
-3.81807014e-02 -4.12045032e-01 -2.49180138e-01 4.80377913e-01
-1.44314456e+00 1.40362048e+00 7.49656439e-01 -2.93532163e-01
-4.42061990e-01 -5.06567299e-01 -4.77453053e-01 5.85355014e-02
1.39081866e-01 -1.11330831e+00 8.95327806e-01 -9.70471680e-01
-1.29048359e+00 1.20566583e+00 -3.94170247e-02 -8.13712478e-01
6.64817870e-01 -2.53646940e-01 -1.39024913e-01 1.45122975e-01
2.11369336e-01 8.97036254e-01 8.73973191e-01 -1.54657853e+00
1.89023837e-01 5.89541681e-02 5.99803567e-01 -9.72150266e-03
-9.67897058e-01 5.75544313e-02 4.03299361e-01 -1.05680478e+00
-3.82598877e-01 -5.41315436e-01 7.38667920e-02 -4.63298827e-01
-3.09796661e-01 -2.87807465e-01 5.40607572e-01 -5.61126888e-01
1.07590377e+00 -1.91392708e+00 3.25772375e-01 -2.00168356e-01
7.51015767e-02 1.42907829e-03 1.11890502e-01 9.86318350e-01
-5.30043654e-02 3.16369146e-01 -9.70867053e-02 -5.34506202e-01
2.11820543e-01 3.15629721e-01 -1.02584302e+00 -5.82792103e-01
-4.62508574e-02 8.80292058e-01 -1.13957369e+00 -5.73576391e-01
4.64133948e-01 4.52914804e-01 -1.60041422e-01 8.28738660e-02
-4.61717844e-01 4.20801818e-01 -1.40950620e-01 3.11045259e-01
5.88564217e-01 2.84873486e-01 2.20563263e-01 9.65958536e-02
-2.81789955e-02 3.43332320e-01 -1.97420955e-01 2.10129166e+00
-1.32998824e+00 7.32389450e-01 -4.81592149e-01 -8.74955952e-01
1.49370241e+00 2.07133994e-01 -6.79600611e-02 -7.47643471e-01
2.30106592e-01 2.59544283e-01 -4.66581881e-01 -5.59821844e-01
8.70638728e-01 -7.91451216e-01 -2.37645566e-01 4.49473262e-01
-1.90558106e-01 -4.92414147e-01 4.64968421e-02 4.60260548e-02
8.05062950e-01 2.34110370e-01 4.19503152e-01 -5.76814234e-01
6.79257751e-01 2.58642107e-01 -5.64986050e-01 6.07148111e-01
4.52236563e-01 2.63880283e-01 5.32731652e-01 -7.81124830e-01
-7.98279226e-01 -1.28103662e+00 -2.55902916e-01 9.06646550e-01
-1.97766945e-02 -5.29407024e-01 -8.64713371e-01 -2.26797163e-01
-3.52510005e-01 1.41452384e+00 -7.44855642e-01 -9.34567526e-02
-4.65976208e-01 -3.37060064e-01 8.15131128e-01 -1.05101287e-01
4.89886373e-01 -1.50666261e+00 -1.48036897e+00 2.10774794e-01
-6.83713436e-01 -9.94293571e-01 3.35941434e-01 -1.14813335e-01
-8.11265826e-01 -7.66553700e-01 -4.64183241e-01 -5.41253865e-01
4.37919408e-01 1.75950885e-01 1.68868327e+00 1.38582617e-01
-4.97661710e-01 3.08118425e-02 -7.89910674e-01 -6.15981460e-01
-1.02603865e+00 -7.42569864e-02 -3.41112643e-01 -5.79199135e-01
3.54344815e-01 -9.03293729e-01 -4.66079116e-01 -2.78934449e-01
-9.66069400e-01 4.37835336e-01 2.62032330e-01 9.47216868e-01
-1.56918630e-01 -3.53349835e-01 4.24646109e-01 -6.89331770e-01
1.60791087e+00 -5.13838530e-01 6.60944507e-02 3.11782718e-01
-6.16573572e-01 1.01365137e-03 1.00794876e+00 -6.85586154e-01
-1.15216112e+00 -5.91368914e-01 -1.27311185e-01 -4.24744904e-01
-3.33586693e-01 4.04409766e-01 4.32129085e-01 2.75376260e-01
1.20285857e+00 1.97336540e-01 4.98934127e-02 -6.93114758e-01
8.16525400e-01 5.76391757e-01 2.92025477e-01 -9.93292332e-01
5.41236877e-01 6.01473451e-01 -3.14509869e-02 -7.07500696e-01
-3.57020736e-01 8.73334855e-02 -4.16202158e-01 -1.96243346e-01
5.27185857e-01 -3.95166039e-01 -5.50855815e-01 -3.16697448e-01
-1.74504924e+00 4.07691821e-02 -6.26156092e-01 -2.54393041e-01
-4.98529553e-01 3.14287215e-01 -1.10364668e-01 -1.00986731e+00
-8.59776020e-01 -6.57029152e-01 1.29654288e+00 -7.31840432e-02
-9.90424693e-01 -1.54902816e+00 1.45516708e-01 2.83553690e-01
7.39764094e-01 8.54371667e-01 1.37223554e+00 -6.39712334e-01
1.60010621e-01 -1.65679336e-01 2.96296459e-02 1.09790601e-01
2.67154813e-01 -2.03343183e-01 -1.05415118e+00 1.74931392e-01
5.07923067e-01 -4.43146050e-01 7.01275110e-01 -3.39866608e-01
5.82720101e-01 -6.24273121e-01 -1.08742639e-01 -1.88364554e-02
1.46153617e+00 -3.02246884e-02 5.87899685e-01 6.09430611e-01
2.63023153e-02 9.15765464e-01 8.13013136e-01 4.80421484e-01
2.64420658e-01 8.59810829e-01 3.21313441e-01 -8.08825418e-02
-4.76204194e-02 -4.32039827e-01 3.28617603e-01 4.27615315e-01
-1.22056253e-01 2.12529395e-02 -1.01288164e+00 5.18825889e-01
-1.56656551e+00 -1.07344949e+00 -2.18054757e-01 1.90558147e+00
7.91160226e-01 2.14858070e-01 6.85099363e-02 -6.36574998e-02
8.65799487e-02 3.04026186e-01 1.28354341e-01 -1.36820686e+00
-3.19131851e-01 6.76680505e-01 -6.07576430e-01 6.85657084e-01
-5.99334002e-01 1.14738786e+00 5.77253723e+00 9.17478502e-01
-1.26290405e+00 3.14562261e-01 2.17179552e-01 -2.12411299e-01
-7.87533045e-01 2.90149301e-01 8.93442407e-02 3.91773701e-01
9.44375098e-01 -3.78556877e-01 5.52429974e-01 8.10649931e-01
4.40854013e-01 -3.10062677e-01 -9.92150664e-01 8.60713303e-01
2.35288039e-01 -1.38489532e+00 6.43972039e-01 -1.55618116e-01
1.21080816e-01 -6.13323748e-01 -1.38360798e-01 3.00399899e-01
-9.30812061e-02 -1.27474749e+00 7.05955982e-01 6.72724962e-01
4.35889333e-01 -3.96645963e-01 6.70156181e-01 4.77570683e-01
-7.48130560e-01 -7.46902004e-02 -4.84883159e-01 -6.11894131e-01
2.37530649e-01 4.94882256e-01 -1.10869491e+00 7.73254037e-01
5.22718549e-01 3.15055281e-01 -9.24280405e-01 1.54146776e-01
-2.54390001e-01 1.53599858e-01 -1.07010528e-01 -5.56441307e-01
3.66179682e-02 -2.90350229e-01 8.75567913e-01 1.43895650e+00
2.48569161e-01 -4.30444747e-01 -1.34226933e-01 1.51533389e+00
4.35929507e-01 6.61369085e-01 -1.02207208e+00 -1.19091369e-01
4.95008975e-01 1.41790342e+00 -7.63802111e-01 -4.38066930e-01
1.96646318e-01 1.18291807e+00 2.72805214e-01 -5.98502979e-02
-9.66072917e-01 -4.79243957e-02 2.54670650e-01 5.59862331e-02
-5.22590816e-01 1.65277287e-01 -6.57351553e-01 -9.81319487e-01
2.30997711e-01 -6.06726229e-01 4.39706296e-02 -1.36409140e+00
-1.46364176e+00 1.01422632e+00 5.88891208e-01 -1.03835368e+00
-5.44887304e-01 -6.14203393e-01 -9.60841596e-01 9.51391518e-01
-9.83758211e-01 -1.54078960e+00 -4.75751728e-01 1.63139015e-01
7.42143869e-01 -1.69032216e-01 1.76049292e+00 -1.07855067e-01
3.48745853e-01 6.81425035e-02 -4.50050354e-01 -3.82495612e-01
6.31948352e-01 -1.33434355e+00 4.76147473e-01 3.49561036e-01
3.05350959e-01 1.27199304e+00 1.33744097e+00 -3.56586844e-01
-9.08647954e-01 -3.60446572e-01 1.10571146e+00 -6.10962152e-01
9.83305693e-01 -5.86726844e-01 -6.11641526e-01 8.95751566e-02
9.42370117e-01 -8.56758833e-01 7.37960994e-01 3.33641231e-01
-5.59749961e-01 4.40231919e-01 -1.24985981e+00 1.01073790e+00
1.19237566e+00 -4.23663110e-01 -1.41340053e+00 7.05114782e-01
8.82959247e-01 -1.91414669e-01 -6.45571530e-01 -9.27294642e-02
4.46342587e-01 -1.31788588e+00 1.17870975e+00 -7.04029143e-01
9.71621573e-01 2.75299430e-01 3.52795795e-02 -1.46105576e+00
3.83556247e-01 -4.45423841e-01 9.72287878e-02 1.31483114e+00
2.35689785e-02 -8.41964185e-01 1.71943859e-03 2.67683953e-01
-8.19761679e-02 -6.76017284e-01 -1.12530589e+00 -7.01835692e-01
5.54050088e-01 -2.35273346e-01 6.96540475e-01 1.06909251e+00
4.52081442e-01 8.10991585e-01 -3.37308832e-02 -9.06449199e-01
2.74695605e-02 6.13383710e-01 7.88219750e-01 -1.33833599e+00
-3.48636210e-01 -9.06354070e-01 -3.10430497e-01 -4.82866317e-01
4.40468788e-01 -9.49262142e-01 -4.00986582e-01 -1.56270778e+00
6.96262121e-02 -9.87417698e-02 1.44718513e-01 3.94041926e-01
2.01926500e-01 8.54330659e-02 5.05291343e-01 5.20729870e-02
8.27378333e-02 5.14757752e-01 1.14655066e+00 -1.18743211e-01
-1.05016366e-01 -6.55223787e-01 -9.92098153e-01 7.64012635e-01
8.14820468e-01 -3.03981125e-01 -6.35821044e-01 -2.23177657e-01
4.47395593e-01 -2.61142910e-01 9.68123674e-01 -6.99630201e-01
-5.49213946e-01 -2.43866399e-01 1.17941871e-02 -1.09182997e-02
6.96987212e-01 -8.35178733e-01 2.43752122e-01 5.61339617e-01
-2.96047688e-01 1.59206718e-01 3.30284238e-01 1.51568204e-01
-1.79933682e-01 -5.39670348e-01 4.66093868e-01 -5.81127226e-01
-2.46545941e-01 -1.08382714e+00 -2.66602457e-01 1.05487414e-01
9.57886577e-01 -4.10450250e-01 -4.91566777e-01 -2.97371566e-01
-5.91187537e-01 -1.87971979e-01 6.75821006e-01 9.27824318e-01
6.21222019e-01 -1.19903421e+00 -6.04965389e-01 -2.25041211e-01
1.46529272e-01 -2.62462139e-01 6.82217181e-02 4.61085469e-01
-6.26352370e-01 5.11845708e-01 -4.52887803e-01 -3.11328322e-01
-1.15205979e+00 7.50719547e-01 1.71604171e-01 -2.68319041e-01
-8.25549781e-01 6.41591549e-01 -9.90066528e-02 -4.39942062e-01
-4.19351429e-01 -3.52344334e-01 -3.56430620e-01 4.25248474e-01
2.15632960e-01 2.26964325e-01 -2.31423512e-01 -5.36173463e-01
-2.43488312e-01 2.42068604e-01 3.15552711e-01 -4.67225671e-01
1.10601842e+00 6.35320768e-02 -6.82148516e-01 8.50633085e-01
9.11508262e-01 2.15674073e-01 -1.06366508e-01 4.96389151e-01
2.11759821e-01 -6.76079333e-01 -6.28366947e-01 -1.07859528e+00
3.68233025e-02 9.92540181e-01 4.15315866e-01 8.72644901e-01
1.13148189e+00 7.93397501e-02 5.73032498e-01 4.45382565e-01
6.26934350e-01 -8.10530424e-01 2.61025339e-01 4.58184928e-01
1.72882557e+00 -8.72395039e-01 -1.61408871e-01 -2.75497615e-01
-5.67350388e-01 1.46631551e+00 4.40218776e-01 -1.32608786e-01
4.29828949e-02 1.78916170e-03 1.51659638e-01 -3.44047725e-01
-7.95257747e-01 2.13859990e-01 3.19858156e-02 5.02501249e-01
6.94932401e-01 3.08952451e-01 -1.17043090e+00 7.50510752e-01
-1.07058644e+00 9.90410894e-03 5.69620788e-01 4.79911387e-01
-1.96704566e-01 -1.35177815e+00 -5.30895412e-01 -6.53337240e-02
2.19386324e-01 -4.10061270e-01 -1.15946651e+00 7.28562951e-01
1.98517829e-01 9.89076257e-01 -5.14345616e-02 -2.31167480e-01
2.65089661e-01 5.22715211e-01 5.30831993e-01 -6.26868784e-01
-1.05833054e+00 -2.83884764e-01 3.38818997e-01 -1.11059554e-01
-5.75718760e-01 -2.90736169e-01 -1.22195590e+00 -2.95549124e-01
8.57561678e-02 3.58859360e-01 7.66537964e-01 9.07929599e-01
4.08285767e-01 5.38768351e-01 2.78879046e-01 -8.10134113e-01
-5.07544577e-01 -1.43789637e+00 1.85602188e-01 7.24749088e-01
-2.74707705e-01 -8.11729610e-01 -2.26790100e-01 1.12958197e-02] | [11.118101119995117, 9.13170337677002] |
7cde111b-eaee-4fc5-959a-ffd3572a4d99 | towards-semi-supervised-deep-facial | 2203.12341 | null | https://arxiv.org/abs/2203.12341v2 | https://arxiv.org/pdf/2203.12341v2.pdf | Towards Semi-Supervised Deep Facial Expression Recognition with An Adaptive Confidence Margin | Only parts of unlabeled data are selected to train models for most semi-supervised learning methods, whose confidence scores are usually higher than the pre-defined threshold (i.e., the confidence margin). We argue that the recognition performance should be further improved by making full use of all unlabeled data. In this paper, we learn an Adaptive Confidence Margin (Ada-CM) to fully leverage all unlabeled data for semi-supervised deep facial expression recognition. All unlabeled samples are partitioned into two subsets by comparing their confidence scores with the adaptively learned confidence margin at each training epoch: (1) subset I including samples whose confidence scores are no lower than the margin; (2) subset II including samples whose confidence scores are lower than the margin. For samples in subset I, we constrain their predictions to match pseudo labels. Meanwhile, samples in subset II participate in the feature-level contrastive objective to learn effective facial expression features. We extensively evaluate Ada-CM on four challenging datasets, showing that our method achieves state-of-the-art performance, especially surpassing fully-supervised baselines in a semi-supervised manner. Ablation study further proves the effectiveness of our method. The source code is available at https://github.com/hangyu94/Ada-CM. | ['Xinbo Gao', 'Xiaoyu Wang', 'Xi Yang', 'Nannan Wang', 'Hangyu Li'] | 2022-03-23 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Li_Towards_Semi-Supervised_Deep_Facial_Expression_Recognition_With_an_Adaptive_Confidence_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Li_Towards_Semi-Supervised_Deep_Facial_Expression_Recognition_With_an_Adaptive_Confidence_CVPR_2022_paper.pdf | cvpr-2022-1 | ['facial-expression-recognition'] | ['computer-vision'] | [ 2.06451297e-01 1.93620682e-01 -7.26352394e-01 -9.08987582e-01
-9.92386520e-01 -3.58597696e-01 2.38784060e-01 -1.13120459e-01
-4.81589913e-01 7.30450273e-01 -7.92930499e-02 2.37677902e-01
3.58351856e-01 -4.13327068e-01 -4.59416896e-01 -9.43525434e-01
1.33177340e-01 2.55276978e-01 -3.01588565e-01 6.12357594e-02
-2.42989555e-01 3.94710541e-01 -1.59610653e+00 2.85386175e-01
9.45182800e-01 1.52657330e+00 -2.85701483e-01 2.61040367e-02
3.10171512e-03 6.75614774e-01 -3.59733850e-01 -2.66708791e-01
2.72872329e-01 -4.96484339e-01 -4.75140214e-01 3.79999161e-01
3.94162595e-01 -4.56321269e-01 9.29080471e-02 1.29016745e+00
2.69141406e-01 -1.21014891e-02 7.77151644e-01 -1.60131848e+00
-3.54200155e-01 2.99002916e-01 -7.75457442e-01 -2.90446669e-01
-3.54404747e-02 -1.26406029e-01 8.81568074e-01 -1.17678022e+00
5.48777938e-01 7.66450584e-01 3.55846107e-01 9.03554738e-01
-1.26175845e+00 -1.14884722e+00 2.85411090e-01 -6.58722222e-02
-1.67784822e+00 -7.39243150e-01 8.76339734e-01 -3.18575174e-01
3.55810404e-01 1.09956367e-02 3.18129689e-01 1.03398931e+00
-2.63582438e-01 1.05081463e+00 1.32181334e+00 -4.38268721e-01
5.57561100e-01 4.41642076e-01 1.80783704e-01 8.50345433e-01
-1.76188573e-01 9.54193808e-03 -6.39711142e-01 -2.72134542e-01
4.98227656e-01 -2.49746572e-02 -1.52890071e-01 -3.86689812e-01
-6.34465456e-01 1.04054046e+00 1.80592149e-01 1.91774994e-01
-3.19295019e-01 -3.13889682e-01 4.19753551e-01 2.64594108e-01
6.57741725e-01 -1.07372336e-01 -5.32194972e-01 5.31999879e-02
-1.03014338e+00 -2.61293054e-01 5.17408729e-01 9.32962835e-01
1.16093898e+00 -5.03432797e-03 -1.21738754e-01 1.18571460e+00
2.36890331e-01 4.12257224e-01 4.62958723e-01 -1.02037406e+00
6.97394907e-02 7.74329126e-01 3.77267040e-02 -5.34961224e-01
-1.38570905e-01 -2.80563027e-01 -9.11659777e-01 3.58063400e-01
5.09359896e-01 -5.19950390e-01 -8.64913166e-01 2.08537912e+00
4.24290985e-01 2.79536456e-01 5.95607888e-03 1.04468083e+00
6.83830917e-01 4.28324908e-01 1.82075724e-01 -6.27404094e-01
1.14547908e+00 -1.01119089e+00 -5.63730657e-01 -1.21565752e-01
7.93275237e-01 -4.26856101e-01 1.09138930e+00 4.68072146e-01
-6.20692670e-01 -4.87267733e-01 -1.09989440e+00 2.45024979e-01
1.59038715e-02 7.60133266e-01 7.41726458e-01 7.40623295e-01
-8.43799651e-01 4.74649549e-01 -8.00204217e-01 -2.90765930e-02
7.61955738e-01 5.10863006e-01 -6.80168867e-01 7.57640377e-02
-9.99649704e-01 2.78963149e-01 -6.96756458e-03 3.00897509e-01
-7.47066319e-01 -4.53273386e-01 -1.06237447e+00 -8.14107955e-02
1.99163422e-01 -7.04763979e-02 1.25454962e+00 -1.62406623e+00
-1.55470586e+00 1.33807814e+00 -4.72856015e-01 -1.46594316e-01
5.46251237e-01 -1.87403768e-01 -3.42739403e-01 2.05033138e-01
1.02425553e-01 8.67625952e-01 9.32803929e-01 -1.16369116e+00
-4.24443752e-01 -5.94402909e-01 -2.79223472e-01 -1.67893320e-02
-6.94553435e-01 1.89502910e-02 -5.66283703e-01 -3.65677655e-01
1.37321055e-01 -1.09289432e+00 -1.23714142e-01 4.76592094e-01
-3.57729644e-01 -3.09798896e-01 8.01904976e-01 -2.25863233e-01
1.07203269e+00 -2.25169873e+00 -2.76186138e-01 2.97390878e-01
1.80165127e-01 4.70632195e-01 -2.87180036e-01 -1.06264353e-01
-1.74131587e-01 -1.05859645e-01 -2.46716052e-01 -5.68603337e-01
-1.49037004e-01 1.39225662e-01 -1.65827155e-01 5.04781902e-01
4.66533840e-01 6.83058858e-01 -7.95308232e-01 -4.46962774e-01
-5.05747609e-02 3.04718614e-01 -3.48959029e-01 3.20795387e-01
-9.77697819e-02 4.63705868e-01 -6.89816892e-01 9.27208185e-01
9.50308621e-01 -3.48916769e-01 3.14422190e-01 -3.21456343e-01
2.26803109e-01 -2.68297791e-01 -1.11841357e+00 1.34826851e+00
-2.30126217e-01 4.98239964e-01 1.18437149e-01 -9.54002976e-01
1.21804821e+00 2.70146847e-01 5.72277427e-01 -4.56336588e-01
1.95435524e-01 2.61075556e-01 -2.13105187e-01 -1.62862986e-01
-7.48796314e-02 -2.78611749e-01 1.58098876e-01 3.83934528e-01
1.90815076e-01 2.15956002e-01 -3.62927429e-02 4.99354079e-02
6.74432635e-01 2.47601300e-01 3.03112924e-01 -2.28696182e-01
6.08167708e-01 -4.40008461e-01 1.02730834e+00 3.81575078e-01
-6.74501061e-01 6.52487576e-01 8.28182161e-01 -1.94728076e-01
-7.50406444e-01 -9.23652053e-01 -3.83987844e-01 1.03304172e+00
-4.43951935e-02 -3.82396668e-01 -1.01129103e+00 -1.15242481e+00
8.19172524e-03 3.26445132e-01 -1.05359209e+00 -2.04745233e-01
-1.12787083e-01 -6.39148653e-01 4.91863042e-01 6.75638080e-01
5.82605183e-01 -9.00488734e-01 -2.01884687e-01 -2.64626205e-01
-3.45603889e-03 -1.02702940e+00 -5.96481144e-01 3.60690951e-01
-6.43528521e-01 -9.91617620e-01 -6.43561780e-01 -6.96215272e-01
1.16990149e+00 -1.93289537e-02 7.24747360e-01 1.11990228e-01
2.42610369e-02 3.35418060e-02 -4.87404972e-01 -3.39124590e-01
-2.30686635e-01 -1.95799559e-01 1.51952401e-01 5.05116880e-01
6.52101278e-01 -3.57799649e-01 -4.73068595e-01 6.36886358e-01
-7.02654958e-01 2.58449204e-02 4.58909482e-01 1.09850550e+00
9.03882027e-01 -4.00650442e-01 7.93114483e-01 -8.91662657e-01
9.91583318e-02 -4.32999313e-01 -4.42979366e-01 1.64955899e-01
-5.63758254e-01 6.95649488e-03 5.35869837e-01 -7.43463218e-01
-8.49832535e-01 4.02127624e-01 -1.89100534e-01 -7.31576085e-01
-2.44536951e-01 4.12736207e-01 -4.72296417e-01 -3.84867890e-04
3.68755966e-01 -2.60212291e-02 2.73394018e-01 -1.31349802e-01
1.39221428e-02 8.89758289e-01 2.05032155e-01 -5.81244290e-01
5.28997183e-01 4.76387471e-01 -1.58248723e-01 -6.45366848e-01
-1.27721977e+00 -2.58349538e-01 -6.87347293e-01 -3.83615732e-01
4.76027220e-01 -1.08716702e+00 -5.05449235e-01 6.06008470e-01
-5.47403753e-01 -6.30797982e-01 -1.56650335e-01 5.58327973e-01
-5.11350572e-01 2.16879010e-01 -6.14064872e-01 -1.01372433e+00
-4.05933708e-01 -9.57584321e-01 1.25473428e+00 4.41241860e-01
-3.04578274e-01 -7.96209633e-01 -8.92194733e-02 4.30301487e-01
2.37344727e-02 4.55049902e-01 4.78014857e-01 -8.98705721e-01
1.73869193e-01 -3.25801909e-01 -1.97546780e-01 8.36697102e-01
3.73198658e-01 1.68053254e-01 -1.32274437e+00 -2.70577610e-01
-2.10359886e-01 -1.05221987e+00 9.69556808e-01 1.36036351e-01
1.39201450e+00 -9.71449241e-02 -2.36070648e-01 6.17419243e-01
9.85517263e-01 6.17963411e-02 4.83991504e-01 -3.84642333e-02
2.60685325e-01 5.71463525e-01 9.28885639e-01 6.78742468e-01
1.04155630e-01 6.52614951e-01 2.92328447e-01 -1.51315719e-01
2.06244513e-01 -4.06395853e-01 4.90905404e-01 3.63091141e-01
1.59888551e-01 -5.25399633e-02 -6.12716258e-01 2.43908167e-01
-1.90152812e+00 -7.00460494e-01 1.56196326e-01 2.18480086e+00
1.14803195e+00 1.08236849e-01 1.15912795e-01 8.26771632e-02
6.78233385e-01 1.36744827e-01 -9.39936042e-01 -1.22264035e-01
-2.60286033e-01 1.48256794e-01 6.78031147e-02 4.32608545e-01
-1.31336033e+00 1.00544953e+00 5.29325771e+00 1.01438212e+00
-1.24568653e+00 7.91435502e-03 1.24893880e+00 -2.19448045e-01
-8.48792791e-02 -1.16856910e-01 -8.47795486e-01 5.96906841e-01
7.88729489e-01 1.22174382e-01 -9.14451759e-03 1.18426955e+00
2.62746632e-01 -2.09683731e-01 -1.11449683e+00 9.73426461e-01
1.53517783e-01 -9.29364920e-01 -1.70091137e-01 3.98651548e-02
8.80899727e-01 -1.33012354e-01 1.09790437e-01 4.33862478e-01
1.24769017e-01 -9.25044358e-01 5.93579113e-01 3.13326716e-01
1.20448256e+00 -7.31028378e-01 8.49555850e-01 3.78051400e-01
-9.78862464e-01 1.37118161e-01 -2.81980127e-01 6.75884187e-02
-1.60729557e-01 8.74154627e-01 -5.94051301e-01 1.97987735e-01
3.89467299e-01 8.12711418e-01 -3.09620649e-01 4.89050925e-01
-5.11919618e-01 8.65182817e-01 -3.53206992e-01 -2.71475948e-02
1.58884048e-01 -4.73440289e-01 -4.26229052e-02 9.75961804e-01
1.42846063e-01 1.93771586e-01 1.61140710e-01 7.07193553e-01
-2.89962351e-01 2.44006872e-01 -1.83055729e-01 -9.67317596e-02
4.97850209e-01 1.44244778e+00 -4.80809718e-01 -3.45912993e-01
-4.22780186e-01 1.04732132e+00 6.10363841e-01 3.76127571e-01
-7.98026025e-01 -2.22704917e-01 7.43510485e-01 -2.80837566e-02
1.41203448e-01 1.83633626e-01 -1.78414345e-01 -1.12558651e+00
8.52516815e-02 -7.46322453e-01 4.54314828e-01 -6.78103387e-01
-1.38687909e+00 7.36994624e-01 -2.25514546e-01 -1.37424016e+00
-2.81236917e-01 -7.04303384e-01 -5.59304178e-01 6.77875102e-01
-1.35608661e+00 -1.12253964e+00 -4.99847174e-01 5.26071489e-01
2.45871097e-01 -2.32773691e-01 1.08568990e+00 1.73107371e-01
-8.88193905e-01 1.11846411e+00 1.07879743e-01 5.35870314e-01
9.27524030e-01 -8.34478974e-01 -4.93942320e-01 5.20351827e-01
1.78870991e-01 5.45278907e-01 3.23958397e-01 -3.05501252e-01
-9.05212164e-01 -1.00866866e+00 6.10136271e-01 -4.79266159e-02
5.09606361e-01 -5.64050734e-01 -9.95940089e-01 6.04441404e-01
-2.90807724e-01 5.68245649e-01 1.24299109e+00 6.88444301e-02
-6.56981170e-01 -3.46568644e-01 -1.36006546e+00 4.72817421e-01
8.05312097e-01 -5.22765994e-01 1.80217810e-02 2.26698190e-01
1.83602601e-01 -2.30805203e-01 -8.10715497e-01 7.59453535e-01
8.79123509e-01 -9.74236846e-01 2.39501655e-01 -6.24485672e-01
4.60264653e-01 -1.82671338e-01 -2.95822948e-01 -1.13648772e+00
5.32879382e-02 -4.79578614e-01 -2.71717161e-01 1.29190981e+00
5.00172019e-01 -5.77514291e-01 1.30217874e+00 7.94284642e-01
3.10390204e-01 -1.26029599e+00 -9.78837907e-01 -6.29673958e-01
4.76145111e-02 -2.98527718e-01 3.08468878e-01 1.06383848e+00
1.39247522e-01 1.37443632e-01 -3.33300233e-01 -9.26079508e-03
5.19628763e-01 3.47660571e-01 6.70222580e-01 -9.71643448e-01
-1.64172694e-01 -2.65275508e-01 -2.92651743e-01 -1.03042662e+00
7.22208321e-01 -6.52478337e-01 8.46002623e-03 -9.18757737e-01
6.01925671e-01 -5.28135002e-01 -5.18794596e-01 1.02061021e+00
-3.44134957e-01 5.27527809e-01 -1.19208366e-01 2.68915862e-01
-6.29865408e-01 8.57377231e-01 1.01081800e+00 -6.29041642e-02
-1.44881502e-01 8.10056254e-02 -7.64719367e-01 8.01992416e-01
9.98526633e-01 -3.30910265e-01 -3.87437612e-01 1.66905113e-02
-2.93324649e-01 3.72261251e-03 8.12201481e-03 -6.06546342e-01
-7.14609995e-02 -3.71166825e-01 5.61241746e-01 -2.54346758e-01
3.21661830e-01 -5.06093681e-01 -2.56255835e-01 1.09208584e-01
-5.71142435e-01 -6.67247295e-01 2.17757031e-01 2.76764899e-01
-3.80074680e-01 -1.81464344e-01 1.03867424e+00 4.39445049e-01
-6.57990277e-01 5.43615341e-01 -1.99391127e-01 -1.08205922e-01
1.04539847e+00 -2.19709426e-01 -9.64491745e-04 -5.60292661e-01
-9.11241472e-01 3.07302654e-01 4.96501833e-01 2.86233515e-01
5.53151906e-01 -1.43111980e+00 -6.50479913e-01 4.34496194e-01
5.79117596e-01 -2.30157554e-01 1.75719097e-01 8.04660201e-01
1.28813580e-01 1.26754373e-01 -1.26728281e-01 -5.69109797e-01
-1.41765690e+00 2.25359663e-01 4.54828262e-01 -5.27099296e-02
-3.61898355e-02 9.82803166e-01 3.60704750e-01 -4.92731333e-01
1.90367714e-01 4.31427509e-02 -3.45204361e-02 -1.22157168e-02
7.58518636e-01 -7.36515075e-02 -8.26288387e-02 -7.77341783e-01
-4.27727997e-01 4.95284557e-01 -1.72462955e-01 -1.17576899e-04
1.20068622e+00 1.35807276e-01 -1.73846968e-02 3.28539312e-01
1.54940748e+00 1.18644886e-01 -1.70500588e+00 -3.82555485e-01
-2.32782245e-01 -3.62294793e-01 4.79750484e-02 -8.89259100e-01
-1.30120385e+00 7.00870454e-01 7.10274041e-01 -5.77911258e-01
1.44145930e+00 1.25490367e-01 3.24307352e-01 2.29861468e-01
3.58612180e-01 -1.08116555e+00 2.23031908e-01 1.42537162e-01
6.79669499e-01 -1.63357139e+00 -4.55626361e-02 -5.38398504e-01
-9.82836485e-01 9.77110267e-01 9.53371704e-01 -4.78731357e-02
7.16450989e-01 5.36300361e-01 2.75691211e-01 1.45948991e-01
-8.15120280e-01 -1.23664737e-01 2.59303570e-01 4.48877722e-01
5.69412053e-01 9.30647105e-02 -1.67664945e-01 1.02433550e+00
8.87695178e-02 2.71706522e-01 3.79003622e-02 8.28894973e-01
-3.94517452e-01 -1.08713138e+00 -7.04461709e-02 5.48996747e-01
-2.71704078e-01 2.02657729e-01 -4.74284530e-01 4.86440420e-01
2.05694318e-01 1.06422293e+00 1.03129461e-01 -4.37868863e-01
3.63108367e-02 2.05570415e-01 4.09407139e-01 -5.35496712e-01
-6.87722340e-02 1.65981442e-01 1.78440839e-01 -5.22198200e-01
-4.32841599e-01 -7.88676262e-01 -1.43227386e+00 1.19748838e-01
-5.26115656e-01 2.90843070e-01 3.85475129e-01 8.00560474e-01
3.21476728e-01 -1.94336295e-01 1.05134618e+00 -6.23274267e-01
-6.82963967e-01 -1.03359020e+00 -7.64744878e-01 4.48301703e-01
2.88262486e-01 -7.00661957e-01 -4.52308953e-01 1.02016255e-01] | [13.564806938171387, 1.637524127960205] |
5ab9fefd-9041-4d19-9677-4acaade4b398 | advances-and-challenges-in-meta-learning-a | 2307.04722 | null | https://arxiv.org/abs/2307.04722v1 | https://arxiv.org/pdf/2307.04722v1.pdf | Advances and Challenges in Meta-Learning: A Technical Review | Meta-learning empowers learning systems with the ability to acquire knowledge from multiple tasks, enabling faster adaptation and generalization to new tasks. This review provides a comprehensive technical overview of meta-learning, emphasizing its importance in real-world applications where data may be scarce or expensive to obtain. The paper covers the state-of-the-art meta-learning approaches and explores the relationship between meta-learning and multi-task learning, transfer learning, domain adaptation and generalization, self-supervised learning, personalized federated learning, and continual learning. By highlighting the synergies between these topics and the field of meta-learning, the paper demonstrates how advancements in one area can benefit the field as a whole, while avoiding unnecessary duplication of efforts. Additionally, the paper delves into advanced meta-learning topics such as learning from complex multi-modal task distributions, unsupervised meta-learning, learning to efficiently adapt to data distribution shifts, and continual meta-learning. Lastly, the paper highlights open problems and challenges for future research in the field. By synthesizing the latest research developments, this paper provides a thorough understanding of meta-learning and its potential impact on various machine learning applications. We believe that this technical overview will contribute to the advancement of meta-learning and its practical implications in addressing real-world problems. | ['KC Santosh', 'Thorsteinn Rögnvaldsson', 'Joaquin Vanschoren', 'Mohamed-Rafik Bouguelia', 'Anna Vettoruzzo'] | 2023-07-10 | null | null | null | null | ['self-supervised-learning', 'meta-learning', 'domain-adaptation', 'continual-learning', 'federated-learning', 'personalized-federated-learning', 'multi-task-learning', 'transfer-learning'] | ['computer-vision', 'methodology', 'methodology', 'methodology', 'methodology', 'methodology', 'methodology', 'miscellaneous'] | [ 1.29535735e-01 -9.66302231e-02 -4.56866592e-01 -2.60404348e-01
-9.58073854e-01 -3.52510065e-01 3.65287811e-01 3.54799628e-01
-3.61081272e-01 9.13866103e-01 -1.75606892e-01 -1.51944542e-02
-4.92856234e-01 -6.21538758e-01 -5.90653598e-01 -8.42223108e-01
-3.34797710e-01 4.80589539e-01 -2.57668287e-01 -2.82207102e-01
2.66354233e-01 1.96149707e-01 -2.06728411e+00 7.56850839e-01
1.09100592e+00 7.05354691e-01 3.71980399e-01 6.24845088e-01
-4.36102927e-01 5.93561172e-01 -1.04044628e+00 -2.39860237e-01
3.35167460e-02 -1.16684467e-01 -1.04424930e+00 1.22299440e-01
5.42355955e-01 3.01310778e-01 2.70367786e-02 8.14585209e-01
6.80501938e-01 5.41646421e-01 7.77100861e-01 -1.47227085e+00
-4.54650402e-01 3.84102851e-01 -5.68007290e-01 4.57604975e-01
-8.92272219e-02 -1.12797335e-01 5.88561952e-01 -9.24188137e-01
2.44624391e-01 1.07589936e+00 9.11749184e-01 7.43766427e-01
-7.72990048e-01 -4.47929949e-01 4.23122406e-01 5.01516998e-01
-7.35573590e-01 -1.80667847e-01 7.71602333e-01 -4.69776720e-01
1.01680136e+00 1.36669099e-01 4.01457608e-01 1.15739548e+00
4.49178755e-01 1.22460973e+00 1.16764808e+00 -8.27544689e-01
3.54364455e-01 2.79252559e-01 -9.41064861e-03 4.34993804e-01
3.20041459e-03 1.49867997e-01 -7.49640465e-01 -2.75177628e-01
1.78490102e-01 3.35487425e-01 3.32881153e-01 -4.13033068e-01
-1.27002537e+00 5.89179397e-01 9.71438661e-02 6.22116327e-01
-5.20829499e-01 -3.00751384e-02 9.03334618e-01 5.72264850e-01
7.82541573e-01 4.99385029e-01 -9.15815771e-01 -2.36517325e-01
-9.82667625e-01 2.78710753e-01 6.61145627e-01 1.02547014e+00
1.00376916e+00 4.33361530e-01 3.36985243e-03 1.22616673e+00
-1.57649100e-01 4.59462017e-01 9.99144256e-01 -8.59110534e-01
5.06605864e-01 5.01368940e-01 -1.03517704e-01 -3.55968386e-01
-6.40936852e-01 -9.17944193e-01 -8.37181091e-01 1.04223713e-01
-7.35620558e-02 -4.09938186e-01 -6.33718073e-01 1.44963121e+00
4.41209525e-01 1.32548288e-01 1.77128971e-01 2.08485946e-01
6.56780005e-01 5.72919548e-01 3.48192453e-01 -3.87006432e-01
1.20967436e+00 -1.39700246e+00 -6.24115229e-01 -3.33605438e-01
6.60172462e-01 -7.52761126e-01 8.74606490e-01 4.60270971e-01
-1.05513775e+00 -9.02439713e-01 -7.72115171e-01 5.52591741e-01
-8.38923931e-01 -2.14800894e-01 8.89894724e-01 6.10195518e-01
-7.02875316e-01 4.73488152e-01 -8.01479042e-01 -6.40918851e-01
4.48841423e-01 2.93748349e-01 -6.92813620e-02 -8.32150399e-04
-1.16222525e+00 9.30492640e-01 6.85908914e-01 -2.47210190e-01
-9.66841042e-01 -1.13866472e+00 -5.79026580e-01 -2.30614007e-01
3.77272874e-01 -9.59604561e-01 1.55961514e+00 -1.22013652e+00
-1.33072639e+00 6.80555642e-01 -1.13276720e-01 -2.90835351e-01
5.29722631e-01 -1.27507582e-01 -8.44569921e-01 -1.87615618e-01
4.40077856e-02 2.62969911e-01 9.75558639e-01 -1.20627570e+00
-1.44481361e+00 -4.72306788e-01 -2.39996836e-01 5.02740324e-01
-8.64262700e-01 -3.67436439e-01 2.03673299e-02 -7.83322275e-01
-4.35020030e-01 -6.25126362e-01 -2.19018906e-01 -8.31135809e-01
-4.73914261e-04 -5.28367460e-01 1.33892429e+00 6.45925477e-02
1.11176836e+00 -2.07145905e+00 4.69758511e-02 -1.79313302e-01
1.23518713e-01 4.09953028e-01 -4.35315102e-01 1.02637565e+00
-1.74394194e-02 -1.08921356e-01 6.81769326e-02 -4.26305681e-01
-2.58347660e-01 1.36700869e-01 -3.04518968e-01 1.38700038e-01
-2.93622732e-01 1.17637360e+00 -1.26739109e+00 -2.58476377e-01
3.16334009e-01 2.81628817e-01 -1.75878704e-01 2.87873685e-01
-2.65314221e-01 5.05978286e-01 -4.33355033e-01 9.82874215e-01
4.52438414e-01 -3.78625363e-01 2.36025721e-01 -1.44041091e-01
-6.61142990e-02 -2.21533537e-01 -9.73416388e-01 1.88686311e+00
-1.08685100e+00 5.19482136e-01 -2.67218146e-02 -1.45508611e+00
8.83943498e-01 3.33419800e-01 1.01682508e+00 -6.87821388e-01
-2.06920177e-01 4.29381162e-01 -3.89278322e-01 -6.87821686e-01
6.53624117e-01 -1.14112779e-01 -5.66900633e-02 7.85658062e-01
5.16874373e-01 1.93608522e-01 1.89225867e-01 -2.27364302e-01
7.43225574e-01 -8.00867081e-02 4.18884039e-01 -2.28170544e-01
5.94520092e-01 2.11149082e-01 2.23706529e-01 9.92400408e-01
-3.37608278e-01 -7.00810924e-02 -4.25826162e-01 -9.34834898e-01
-8.53771269e-01 -9.41179454e-01 1.39668640e-02 2.03096080e+00
-1.22786924e-01 -1.93695962e-01 -5.35337746e-01 -9.99421537e-01
3.78266484e-01 4.77169961e-01 -9.36112463e-01 -3.52424711e-01
-4.41226691e-01 -1.13732076e+00 2.31462300e-01 5.58464468e-01
7.24574447e-01 -1.15603721e+00 -6.40581965e-01 4.22925055e-01
-3.10488462e-01 -6.98935866e-01 -5.08891605e-02 3.94837081e-01
-1.52420974e+00 -1.03124607e+00 -8.37289870e-01 -1.00072277e+00
5.26841164e-01 6.81565166e-01 1.23644960e+00 -1.80693999e-01
-3.76101494e-01 1.02446115e+00 -4.32992399e-01 -6.73110366e-01
-5.25979936e-01 7.67705441e-01 3.34330469e-01 -1.65470377e-01
5.06633282e-01 -5.62080145e-01 -4.32415009e-01 3.08175743e-01
-9.47500288e-01 -3.17195773e-01 7.27538466e-01 1.29448402e+00
5.83154142e-01 4.23901528e-01 1.28565967e+00 -1.45524848e+00
8.27135205e-01 -8.53188574e-01 -2.33008727e-01 6.89312994e-01
-1.07750988e+00 -2.46075973e-01 6.58746421e-01 -6.70950115e-01
-1.30853462e+00 -2.23672166e-01 2.52880037e-01 -4.57066238e-01
-1.64385885e-01 5.61748266e-01 7.85839632e-02 -1.43331036e-01
1.00712657e+00 4.59482014e-01 9.58622470e-02 -5.73203683e-01
5.91632783e-01 9.93242323e-01 4.89310414e-01 -8.63948107e-01
4.10327435e-01 4.53759402e-01 -7.87416622e-02 -8.58662665e-01
-1.09687436e+00 -6.81251407e-01 -9.12314296e-01 -3.51271093e-01
2.19912767e-01 -8.48440945e-01 -4.99915391e-01 5.93499720e-01
-7.30173051e-01 -6.96507156e-01 -6.52359962e-01 1.98285744e-01
-9.60968137e-01 1.32683173e-01 -2.26387426e-01 -7.61340857e-01
-5.58869541e-01 -8.69491279e-01 7.68165827e-01 3.70131642e-01
2.63192505e-02 -1.90206516e+00 2.60001570e-01 4.05587971e-01
7.15407610e-01 1.48817644e-01 1.00100636e+00 -8.90786767e-01
-1.67628288e-01 -2.05757350e-01 5.58176696e-01 2.91823149e-01
6.56882644e-01 -2.64359146e-01 -1.26738930e+00 -7.54684865e-01
-1.56073987e-01 -6.46001399e-01 7.40895271e-01 3.97796571e-01
1.27238655e+00 -1.71999872e-01 -5.84033608e-01 6.52011156e-01
1.37932134e+00 1.23712830e-01 4.17335555e-02 9.22931075e-01
3.73031735e-01 7.28804350e-01 1.03310513e+00 6.34163797e-01
5.61640620e-01 3.06305468e-01 3.98797333e-01 3.21026221e-02
-2.32080802e-01 -4.14350256e-02 1.67019039e-01 1.00359094e+00
-9.10383016e-02 -1.31898820e-02 -9.51010108e-01 5.94591737e-01
-2.06877804e+00 -1.13585496e+00 3.11917275e-01 2.17497897e+00
6.42436266e-01 -3.91432881e-01 4.21709001e-01 3.12668160e-02
7.85679400e-01 4.88999039e-02 -8.65985513e-01 -4.35958624e-01
-1.23204753e-01 -4.93180826e-02 2.02105179e-01 2.65773416e-01
-1.40580833e+00 9.60347414e-01 7.76059437e+00 9.02124763e-01
-1.26310122e+00 5.05422533e-01 3.87189388e-01 -1.29320875e-01
-1.85699631e-02 -4.35176969e-01 -7.98210025e-01 3.54764014e-01
1.02974129e+00 -3.69553000e-01 4.84038711e-01 1.22024441e+00
-2.48056099e-01 1.48512632e-01 -1.00095344e+00 9.96314406e-01
1.84346676e-01 -1.44226229e+00 1.07823171e-01 -1.14940502e-01
1.31573367e+00 4.40238893e-01 5.17526805e-01 7.65987039e-01
1.98716775e-01 -5.82127571e-01 3.94128978e-01 3.43385726e-01
7.11836040e-01 -9.63839591e-01 6.83441997e-01 5.76473236e-01
-1.11064506e+00 -6.55843496e-01 -4.68293995e-01 2.10860193e-01
-1.71665922e-01 7.44398654e-01 -7.71079242e-01 9.85409915e-01
7.83469141e-01 8.17612350e-01 -6.42882288e-01 1.10638714e+00
3.15875709e-01 3.52738917e-01 2.61899441e-01 7.10055456e-02
3.38938162e-02 2.61799008e-01 5.92019558e-01 1.55666673e+00
2.68631220e-01 -3.79553050e-01 4.92214680e-01 2.11470157e-01
1.11074254e-01 1.18838564e-01 -5.35864830e-01 6.42284602e-02
7.56587148e-01 1.31160057e+00 -2.08685502e-01 -4.36400861e-01
-4.60707873e-01 8.23911965e-01 4.08027023e-01 4.83140498e-01
-2.31495082e-01 -5.56294978e-01 5.34942150e-01 -3.18169892e-01
-1.21201158e-01 -9.01296083e-03 -3.61940265e-01 -1.10121536e+00
-4.05613691e-01 -1.12247813e+00 1.06067884e+00 -2.01880485e-01
-1.71196175e+00 6.72882199e-01 2.43442044e-01 -1.43906987e+00
-6.59552038e-01 -5.43545127e-01 -5.29290259e-01 6.85910523e-01
-1.93088722e+00 -1.36214471e+00 -5.30173182e-01 7.49297738e-01
1.00908542e+00 -1.02111435e+00 9.60490286e-01 2.97699720e-01
-4.66044962e-01 9.15445685e-01 6.22117698e-01 -3.66354108e-01
1.05589736e+00 -1.29714870e+00 2.35536560e-01 2.38860995e-01
8.41043964e-02 5.17393947e-01 4.84363556e-01 -4.45969135e-01
-1.61259031e+00 -1.46461177e+00 4.99808341e-01 -5.00400305e-01
4.80399668e-01 -2.99492721e-02 -8.90870869e-01 6.56244993e-01
2.40085259e-01 6.86551034e-02 1.17600012e+00 6.08937562e-01
-3.13806266e-01 -4.74284351e-01 -1.23861349e+00 2.08148867e-01
6.85398459e-01 -3.75950277e-01 -5.10593653e-01 6.42575443e-01
3.14640462e-01 -4.02280062e-01 -1.14781380e+00 4.12882417e-01
5.54873586e-01 -8.12155843e-01 9.10651147e-01 -9.80805695e-01
6.28865957e-02 1.79895982e-01 -9.02606621e-02 -1.91079438e+00
-6.15248442e-01 -5.88007867e-01 -7.18365431e-01 1.13956261e+00
1.42273799e-01 -8.00367177e-01 9.02664542e-01 6.26628026e-02
-3.93992841e-01 -8.56225789e-01 -9.15038943e-01 -1.17346025e+00
5.31964540e-01 -4.18000072e-01 8.21758807e-01 1.25829327e+00
2.61404663e-01 -5.05093532e-03 -3.10395092e-01 -1.00719251e-01
8.81201565e-01 2.52363175e-01 8.85481536e-01 -1.32949257e+00
-2.18501240e-01 -4.88912523e-01 -1.03290603e-01 -6.93773389e-01
4.25573707e-01 -1.11126578e+00 -2.98222691e-01 -1.42778826e+00
1.80095330e-01 -6.05284035e-01 -8.38214576e-01 6.50413334e-01
-2.80077517e-01 3.78764570e-02 1.74050421e-01 5.18116593e-01
-9.20326233e-01 4.87860829e-01 1.29099810e+00 -4.44932938e-01
-4.49213773e-01 5.19503415e-01 -7.62675345e-01 5.69273412e-01
9.77673709e-01 -4.85100418e-01 -5.77727556e-01 -5.73080540e-01
3.82701419e-02 -2.99183071e-01 -1.07271887e-01 -9.00486946e-01
4.95102167e-01 -4.91263866e-01 4.52296913e-01 -4.60076839e-01
1.50495350e-01 -9.15437281e-01 -3.41597021e-01 3.92951220e-01
-2.00821638e-01 3.59184742e-01 5.53184271e-01 6.92243040e-01
-3.46336544e-01 -3.25297713e-01 7.87311792e-01 -4.67771351e-01
-1.20285738e+00 4.20856833e-01 -3.89495492e-01 2.69227356e-01
1.28290057e+00 -1.85344145e-01 -4.70853925e-01 -1.07350439e-01
-8.52230787e-01 4.91843253e-01 7.37056881e-02 7.66535699e-01
5.20325005e-01 -1.33704960e+00 -7.03043401e-01 1.65044099e-01
4.34177428e-01 -2.60302514e-01 8.03352356e-01 6.23776436e-01
1.98680237e-01 3.17500293e-01 -3.87641907e-01 -8.85650039e-01
-1.06735086e+00 6.37767076e-01 5.52766979e-01 -2.67338097e-01
-3.24688375e-01 5.69834709e-01 -2.34678194e-01 -8.72372925e-01
5.02040267e-01 3.18710327e-01 -1.59524739e-01 2.08584771e-01
6.71355426e-01 9.24352586e-01 4.45605814e-01 -9.67592373e-02
-2.92883813e-01 5.18993139e-01 -3.47040415e-01 8.81188512e-02
1.26698864e+00 -2.79476464e-01 -4.58341055e-02 9.55393732e-01
1.05713379e+00 -4.02229816e-01 -1.20838332e+00 -4.76600826e-01
1.96451992e-01 -3.58294606e-01 -9.85925943e-02 -1.19403195e+00
-9.45743263e-01 1.03335249e+00 8.65363896e-01 1.66977182e-01
1.25654042e+00 -4.79150042e-02 5.51146746e-01 7.23055065e-01
6.98796690e-01 -1.54516602e+00 7.13896513e-01 7.59793580e-01
5.54767728e-01 -1.48387647e+00 4.35463414e-02 3.06745172e-01
-6.98025107e-01 1.36212218e+00 8.86503875e-01 3.23614478e-01
9.50568557e-01 8.82580429e-02 2.40105361e-01 -9.24603519e-05
-1.08723569e+00 6.98308796e-02 4.04180020e-01 1.08246171e+00
4.23340589e-01 3.12446430e-02 3.84289354e-01 2.61620134e-01
-6.91556260e-02 -4.49218303e-02 1.96015731e-01 1.20934629e+00
-7.59704173e-01 -1.60605109e+00 -3.64671648e-01 4.63505715e-01
-3.19084108e-01 2.36468136e-01 -1.20918266e-02 7.46393561e-01
2.44953007e-01 9.93367612e-01 1.83496237e-01 -3.87073785e-01
4.17997837e-01 2.93598741e-01 6.42715454e-01 -9.34353292e-01
-8.63827348e-01 -2.34078839e-01 -3.96637022e-01 -1.35881737e-01
-6.18280470e-01 -6.55032277e-01 -8.72267067e-01 -2.63306946e-01
-1.30899340e-01 4.12860453e-01 8.22764516e-01 9.42307472e-01
5.83634973e-01 7.43227184e-01 8.66408587e-01 -7.78886557e-01
-7.89325833e-01 -1.01829422e+00 -5.17073333e-01 5.02993912e-02
5.29574573e-01 -6.77636921e-01 -7.96370134e-02 1.13640189e-01] | [9.922472953796387, 3.163855791091919] |
2083c277-cc5f-488b-bd50-3b1382a35bb2 | genetic-algorithm-for-program-synthesis | 2211.11937 | null | https://arxiv.org/abs/2211.11937v2 | https://arxiv.org/pdf/2211.11937v2.pdf | Genetic Algorithm for Program Synthesis | A deductive program synthesis tool takes a specification as input and derives a program that satisfies the specification. The drawback of this approach is that search spaces for such correct programs tend to be enormous, making it difficult to derive correct programs within a realistic timeout. To speed up such program derivation, we improve the search strategy of a deductive program synthesis tool, SuSLik, using evolutionary computation. Our cross-validation shows that the improvement brought by evolutionary computation generalises to unforeseen problems. | ['Yutaka Nagashima'] | 2022-11-22 | null | null | null | null | ['program-synthesis'] | ['computer-code'] | [ 4.66424674e-01 3.40304136e-01 -8.79513994e-02 -3.18216205e-01
-2.82750100e-01 -5.78229129e-01 4.32661086e-01 5.72838113e-02
8.74085817e-03 9.15927708e-01 -3.77866060e-01 -7.86418557e-01
-5.66043817e-02 -1.22556627e+00 -5.45808434e-01 -1.44796118e-01
2.48379171e-01 2.36707509e-01 5.72312951e-01 -3.82271469e-01
5.66474080e-01 3.55605960e-01 -2.10326958e+00 1.41778871e-01
1.15841901e+00 3.43928516e-01 1.30673289e-01 9.44612682e-01
-3.08338165e-01 6.11594439e-01 -7.97930002e-01 -4.54106897e-01
1.45062238e-01 -7.09835827e-01 -8.81008446e-01 -3.60789567e-01
-1.61387101e-01 -4.24971767e-02 1.67836443e-01 1.38092363e+00
1.43362075e-01 -2.95704782e-01 3.20509911e-01 -1.64006972e+00
-4.24895197e-01 7.39798725e-01 -1.54005870e-01 -3.36736113e-01
5.86003244e-01 2.82458723e-01 9.34906781e-01 -3.37334454e-01
8.08352828e-01 1.16741419e+00 8.19329858e-01 6.92021787e-01
-1.63400424e+00 -4.47437793e-01 -2.66824037e-01 -3.07389706e-01
-1.47231197e+00 -3.43989521e-01 6.69674933e-01 -3.90074283e-01
1.40575445e+00 8.91452610e-01 1.06253052e+00 4.57349777e-01
4.88965154e-01 5.85478067e-01 1.17626965e+00 -7.64811158e-01
6.68787718e-01 5.35170257e-01 -5.13785630e-02 5.96650481e-01
6.04278803e-01 7.84580469e-01 5.79322726e-02 -5.11256456e-01
2.70777673e-01 -3.94799501e-01 -1.64063513e-01 -5.86873114e-01
-7.16196120e-01 9.47770178e-01 8.22528033e-04 5.16004384e-01
-1.00659542e-01 3.44367862e-01 4.89489317e-01 8.46246183e-01
-5.01810983e-02 1.16486859e+00 -6.85591459e-01 -3.33679885e-01
-1.18793273e+00 5.38577557e-01 1.22358143e+00 9.98470485e-01
6.05585158e-01 3.69742483e-01 1.84487879e-01 1.39031082e-01
1.59609795e-01 4.50005502e-01 3.01088870e-01 -8.60699892e-01
2.08932441e-02 9.79766667e-01 -6.61649927e-02 -9.88828897e-01
4.57205763e-03 -2.04963565e-01 -1.37955859e-01 9.08309340e-01
1.23730622e-01 -4.15420234e-01 -3.08326989e-01 1.28702927e+00
2.81767786e-01 -6.28116131e-01 1.80863336e-01 4.33139831e-01
3.31384063e-01 6.93347454e-01 -3.28064561e-01 -6.57959759e-01
6.19927466e-01 -8.58879387e-01 -3.85848045e-01 -6.71789870e-02
9.28413868e-01 -4.92906541e-01 8.37551951e-01 5.70082605e-01
-9.67225671e-01 -2.24905670e-01 -1.30984151e+00 5.91203690e-01
-2.22417414e-01 -1.95340380e-01 9.51667309e-01 1.19008791e+00
-1.05133271e+00 7.66641378e-01 -5.28588712e-01 -2.50824749e-01
1.09382108e-01 7.41410851e-01 2.09344514e-02 1.97045282e-01
-7.31210172e-01 1.00459409e+00 9.69092488e-01 -2.17011198e-01
-3.32225382e-01 -8.31743777e-01 -7.24730372e-01 8.34315792e-02
5.79716265e-01 -7.89566875e-01 1.38708174e+00 -1.42656982e+00
-1.91140652e+00 6.66929960e-01 5.63409217e-02 -5.40201962e-01
6.04090273e-01 6.22237802e-01 -5.36122322e-01 -4.79339182e-01
-4.50827330e-01 4.15146947e-01 7.11478114e-01 -1.12376094e+00
-5.78413963e-01 2.55589150e-02 2.68269181e-01 -4.76995468e-01
2.34620590e-02 3.63557726e-01 3.08699697e-01 -3.46178979e-01
-5.30154705e-01 -1.19754422e+00 -5.39340436e-01 -1.55270249e-01
-2.45087415e-01 -8.78187642e-02 6.54921293e-01 -3.18140388e-01
1.65316164e+00 -2.02395773e+00 3.24719816e-01 5.50968826e-01
-4.95918877e-02 4.27083284e-01 5.14127910e-02 5.54785907e-01
-3.17113638e-01 3.01651120e-01 -4.54609424e-01 6.55175447e-01
3.91535193e-01 2.49313027e-01 -3.04481924e-01 7.53413737e-02
3.27673495e-01 9.14361417e-01 -9.54821646e-01 -5.09462237e-01
4.20659743e-02 -1.30040690e-01 -1.07351315e+00 1.16897851e-01
-8.30895841e-01 -2.13582382e-01 -5.57869315e-01 4.76190031e-01
4.35503870e-01 6.89776689e-02 4.06267583e-01 3.92565161e-01
-4.53017503e-01 1.18474998e-01 -1.10609019e+00 1.29087651e+00
-7.72653759e-01 6.51171088e-01 -3.58837664e-01 -8.70423496e-01
1.30237150e+00 2.63729066e-01 -4.17402945e-02 -3.49758834e-01
2.37404525e-01 5.69401324e-01 3.63724440e-01 -5.08402288e-01
4.06298131e-01 -2.20549971e-01 -2.36029238e-01 6.89985633e-01
-3.53079081e-01 -1.12524426e+00 4.19094205e-01 -3.12164009e-01
1.11714029e+00 3.92803907e-01 7.66479731e-01 -4.52514738e-01
7.83031225e-01 6.04802608e-01 4.84509975e-01 5.13698041e-01
2.29432717e-01 2.78791606e-01 1.02512419e+00 -5.93238056e-01
-1.30632997e+00 -4.85570371e-01 1.46410912e-01 5.44334233e-01
-4.20134097e-01 -6.94667459e-01 -7.94810712e-01 -8.15084398e-01
-5.52972034e-02 1.38870513e+00 -3.62197161e-01 -4.60153341e-01
-6.93975508e-01 -5.17305732e-01 4.90967453e-01 1.17796667e-01
8.16526711e-02 -1.14341891e+00 -1.38583648e+00 4.73494083e-01
3.30309212e-01 -1.70196652e-01 -2.93298047e-02 2.17823029e-01
-1.00585127e+00 -9.22303975e-01 -2.11744428e-01 -7.78718650e-01
8.53398740e-01 -4.13304418e-01 1.29102314e+00 5.89204788e-01
-3.96660417e-01 -1.87153205e-01 -2.06893638e-01 -4.23087627e-01
-1.33385861e+00 1.76390889e-03 -3.05343121e-01 -8.03632200e-01
1.55555263e-01 -4.42851067e-01 3.49792331e-01 -7.11969845e-03
-8.40484381e-01 8.58674422e-02 4.32681441e-01 1.04567766e+00
2.60334909e-01 9.06552136e-01 4.40275997e-01 -9.14568126e-01
8.14010799e-01 -5.25004156e-02 -1.39427686e+00 5.92573345e-01
-1.17487681e+00 6.33072972e-01 8.95171046e-01 -3.50984544e-01
-9.11154389e-01 4.94425058e-01 -1.69950023e-01 7.81024098e-02
5.22009134e-02 5.67814112e-01 -1.72544003e-01 -3.16792458e-01
9.12136436e-01 2.52704114e-01 9.90134254e-02 5.07541336e-02
1.36554748e-01 5.38779974e-01 2.37362742e-01 -6.61354005e-01
9.29380357e-01 -3.23988110e-01 1.38099074e-01 -5.40660083e-01
2.66592167e-02 3.72408330e-01 -2.46357262e-01 7.22368732e-02
1.58024043e-01 -1.90642655e-01 -5.96699834e-01 -2.61833910e-02
-1.30784643e+00 -2.90136755e-01 -5.19228697e-01 -7.92838261e-02
-8.28733921e-01 6.21208996e-02 3.37355494e-01 -1.01480782e+00
-3.56032461e-01 -1.31235838e+00 5.87285697e-01 -9.46862251e-03
-1.01279581e+00 -6.90532982e-01 4.71649528e-01 -3.50408167e-01
4.89517629e-01 3.29921752e-01 1.27390933e+00 -5.55895925e-01
-4.35048610e-01 -4.03292686e-01 2.09245935e-01 2.13604987e-01
-6.50977194e-02 9.60017622e-01 -4.98117715e-01 4.40807790e-02
-1.90006234e-02 -4.44348343e-02 -9.99404639e-02 3.49920690e-02
8.45978022e-01 -6.56282187e-01 -4.58013445e-01 3.45460713e-01
1.62002683e+00 5.82842052e-01 7.57926166e-01 4.77822721e-01
-1.91460878e-01 6.63616896e-01 7.32342601e-01 2.93630302e-01
-2.08499223e-01 7.44007289e-01 2.68568069e-01 3.64033043e-01
7.48318359e-02 -2.31414318e-01 2.21189201e-01 3.32008600e-01
1.21014975e-01 1.15566641e-01 -1.14991164e+00 7.29983687e-01
-1.77656412e+00 -9.57164705e-01 -2.04802319e-01 2.07493973e+00
1.27062523e+00 2.84677118e-01 3.22833478e-01 5.82484484e-01
5.15970767e-01 -4.44372267e-01 -9.78700444e-02 -1.20774388e+00
2.85675615e-01 4.87217635e-01 2.76643991e-01 5.24958551e-01
-4.25357938e-01 7.12181866e-01 7.02385998e+00 5.91354668e-01
-1.32519817e+00 -4.09406662e-01 -5.60312234e-02 6.43262127e-03
-8.35675180e-01 4.29464847e-01 -3.41971308e-01 4.30539697e-01
1.38675570e+00 -1.11756086e+00 6.40632272e-01 1.23216069e+00
-5.81586175e-02 -3.02823663e-01 -1.31004000e+00 4.79006588e-01
-2.89853930e-01 -1.45579743e+00 -3.42922747e-01 -1.69660449e-01
8.26722205e-01 -5.72289824e-01 -1.78869203e-01 4.48770821e-01
5.38349807e-01 -1.12620437e+00 8.51783812e-01 2.93533862e-01
4.12592381e-01 -1.32478547e+00 7.11135566e-01 4.93498445e-01
-5.34125805e-01 -2.15124622e-01 -1.59746587e-01 -2.95744330e-01
-2.00448081e-01 5.06866992e-01 -1.12365162e+00 2.79756278e-01
3.16168547e-01 7.75125250e-03 -6.60289526e-01 1.32745838e+00
-4.14404452e-01 2.50264376e-01 -1.71707287e-01 -7.70216525e-01
-6.31873496e-03 -1.14202194e-01 5.45669198e-01 1.13368559e+00
6.33130848e-01 -3.03016882e-02 -4.67056811e-01 1.45704103e+00
5.71228206e-01 6.63454160e-02 -1.05184245e+00 -4.87678647e-01
3.89747620e-01 7.68644154e-01 -6.19431734e-01 -3.03357959e-01
-8.79386142e-02 6.01385474e-01 -7.67051429e-02 -8.66833031e-02
-8.94076526e-01 -9.69998956e-01 1.81087956e-01 -9.87462252e-02
3.60800862e-01 -6.05295449e-02 -7.77079165e-01 -7.63589501e-01
6.71230908e-03 -1.43200922e+00 -1.53823942e-01 -7.12012649e-01
-3.33810985e-01 8.17771018e-01 1.45325184e-01 -9.73402619e-01
-9.61562634e-01 -3.03352654e-01 -6.35832727e-01 8.05484474e-01
-8.09099376e-01 -7.52074063e-01 2.35369727e-01 -7.35222027e-02
2.74003029e-01 -1.83146596e-01 9.25993025e-01 -5.57310209e-02
-3.51704568e-01 6.19435787e-01 -4.78484511e-01 -6.73868537e-01
1.76352151e-02 -1.12646937e+00 4.67999220e-01 1.05851269e+00
-2.97140867e-01 7.27714062e-01 1.34663868e+00 -5.76096058e-01
-1.91823518e+00 -8.67640078e-01 1.34048641e+00 -2.63559818e-01
9.74737346e-01 -6.93854839e-02 -7.13331223e-01 3.61640573e-01
2.26278454e-01 -2.49939039e-01 1.49469674e-01 -1.29384249e-01
-2.45289505e-01 3.16454400e-03 -1.34134281e+00 8.49375308e-01
7.77986288e-01 -3.47972482e-01 -6.80172145e-01 1.55584216e-01
8.47837090e-01 -2.71444321e-01 -8.90259087e-01 2.67519385e-01
6.44591928e-01 -1.03958631e+00 6.93149865e-01 -4.04437929e-01
6.89823449e-01 -5.92364430e-01 1.35199249e-01 -1.34238827e+00
1.58564702e-01 -9.92804885e-01 2.18990277e-02 1.06658888e+00
7.43826747e-01 -7.91642010e-01 7.13841259e-01 1.01775599e+00
1.21808434e-02 -5.23268461e-01 -5.59049129e-01 -1.20412123e+00
2.54785061e-01 -4.29528356e-01 1.14237928e+00 7.17039049e-01
5.94943643e-01 1.94951668e-01 2.96294503e-02 -2.00356439e-01
2.39010036e-01 7.77683973e-01 8.00608993e-01 -1.15094042e+00
-9.43698883e-01 -8.14671695e-01 -2.12582663e-01 -1.01167090e-01
4.42384273e-01 -7.84459829e-01 1.86277956e-01 -9.56345081e-01
-1.75451934e-01 -6.39376044e-01 4.79146093e-01 5.35369813e-01
2.03102529e-01 -2.32437864e-01 1.58674166e-01 -2.88700938e-01
-7.18811378e-02 2.50365615e-01 9.49162066e-01 -2.54115790e-01
-4.84261692e-01 2.67732382e-01 -6.64380789e-01 6.19401991e-01
7.35110343e-01 -8.11760128e-01 -4.04701322e-01 -1.12686045e-01
9.62803960e-01 4.86609817e-01 1.22437820e-01 -9.39418614e-01
1.10582327e-02 -6.06022418e-01 -4.96376455e-01 -2.11305767e-01
-3.86046976e-01 -1.28763211e+00 1.16880655e+00 1.12764561e+00
-3.79427612e-01 2.30295330e-01 3.31818938e-01 1.54310223e-02
-3.66114050e-01 -1.09860730e+00 7.31261015e-01 3.51946466e-02
-7.46463478e-01 -3.13401639e-01 -6.08349621e-01 -3.35299373e-01
1.32342803e+00 -4.69355226e-01 -3.44818272e-02 7.98363686e-02
-2.79311389e-01 -8.57293308e-02 1.24691451e+00 1.56124890e-01
4.26865697e-01 -1.05508292e+00 -3.73955339e-01 3.07812095e-01
2.14283481e-01 -4.00344968e-01 -4.42576110e-01 6.58339024e-01
-9.61541593e-01 4.47016209e-01 -4.08757091e-01 -3.15428108e-01
-1.55466223e+00 9.63927984e-01 3.84462476e-01 -7.53064603e-02
-4.56310630e-01 5.07428408e-01 -4.64127719e-01 -7.58356452e-01
-2.99383372e-01 -4.00292039e-01 2.04509780e-01 -5.36845922e-01
3.83329242e-01 3.30822676e-01 1.81049496e-01 1.58310413e-01
-5.05123913e-01 3.02180856e-01 5.54437697e-01 -1.26700535e-01
1.51194417e+00 5.30215204e-01 -5.94686568e-01 3.69564630e-02
1.02083826e+00 1.32981762e-01 -6.58666670e-01 3.30751598e-01
2.52949029e-01 -5.51877916e-01 8.70809108e-02 -7.78911829e-01
-7.45470107e-01 4.44240808e-01 1.42804220e-01 5.12452841e-01
1.37657285e+00 -6.31161034e-01 4.65060890e-01 7.02192664e-01
6.93547308e-01 -7.87574828e-01 -4.87958372e-01 4.39609379e-01
9.23346281e-01 -7.88264632e-01 1.64545625e-01 -4.91197437e-01
-3.52292717e-01 1.58524418e+00 3.77057642e-01 -2.32326731e-01
2.25469261e-01 9.23482120e-01 -5.57129145e-01 5.90736233e-02
-9.55110550e-01 2.67683268e-01 -2.76836306e-02 7.60984063e-01
2.80265838e-01 5.99232800e-02 -7.19871521e-01 1.59214273e-01
-6.83928847e-01 5.93418598e-01 1.04096615e+00 1.34193146e+00
-3.26712310e-01 -1.52090490e+00 -4.81237024e-01 5.28953448e-02
-1.46821916e-01 -6.62718639e-02 -6.38888299e-01 1.08645594e+00
9.61085856e-02 8.84513557e-01 -2.71072537e-01 -4.18404847e-01
1.61567152e-01 -1.30890369e-01 9.56091166e-01 -7.25070059e-01
-8.91461015e-01 -3.65407646e-01 6.81285679e-01 -3.76838446e-01
-2.21306846e-01 -5.73645651e-01 -1.24191046e+00 -5.09986162e-01
-4.54431385e-01 3.67221147e-01 7.09762335e-01 6.68987572e-01
1.63524717e-01 5.28923035e-01 5.44336438e-01 -2.42113188e-01
-7.24405229e-01 2.72426195e-02 -2.42836118e-01 -2.94168115e-01
1.54257789e-01 -1.74178660e-01 -8.87126550e-02 1.15569346e-01] | [8.068137168884277, 7.3107590675354] |
826c7f9a-3b26-48fa-9e88-ff3118033fd2 | compound-figure-separation-of-biomedical | 2107.08650 | null | https://arxiv.org/abs/2107.08650v1 | https://arxiv.org/pdf/2107.08650v1.pdf | Compound Figure Separation of Biomedical Images with Side Loss | Unsupervised learning algorithms (e.g., self-supervised learning, auto-encoder, contrastive learning) allow deep learning models to learn effective image representations from large-scale unlabeled data. In medical image analysis, even unannotated data can be difficult to obtain for individual labs. Fortunately, national-level efforts have been made to provide efficient access to obtain biomedical image data from previous scientific publications. For instance, NIH has launched the Open-i search engine that provides a large-scale image database with free access. However, the images in scientific publications consist of a considerable amount of compound figures with subplots. To extract and curate individual subplots, many different compound figure separation approaches have been developed, especially with the recent advances in deep learning. However, previous approaches typically required resource extensive bounding box annotation to train detection models. In this paper, we propose a simple compound figure separation (SimCFS) framework that uses weak classification annotations from individual images. Our technical contribution is three-fold: (1) we introduce a new side loss that is designed for compound figure separation; (2) we introduce an intra-class image augmentation method to simulate hard cases; (3) the proposed framework enables an efficient deployment to new classes of images, without requiring resource extensive bounding box annotations. From the results, the SimCFS achieved a new state-of-the-art performance on the ImageCLEF 2016 Compound Figure Separation Database. The source code of SimCFS is made publicly available at https://github.com/hrlblab/ImageSeperation. | ['Yuankai Huo', 'Haichun Yang', 'Catie Chang', 'Bennett A. Landman', 'Agnes B. Fogo', 'Mengyang Zhao', 'Shunxing Bao', 'Aadarsh Jha', 'Jiachen Xu', 'Yuanhan Tian', 'Ruining Deng', 'Quan Liu', 'Chang Qu', 'Tianyuan Yao'] | 2021-07-19 | null | null | null | null | ['image-augmentation'] | ['computer-vision'] | [ 2.36307174e-01 1.26888767e-01 -3.46578002e-01 -4.00064975e-01
-1.12915361e+00 -5.29481947e-01 2.08991319e-01 3.46468627e-01
-3.61349314e-01 6.73677683e-01 -2.77912080e-01 -4.69017535e-01
3.12263995e-01 -6.13574862e-01 -8.89018118e-01 -5.77120125e-01
7.40318969e-02 1.83455050e-01 1.11644395e-01 5.49925528e-02
1.28293745e-02 6.46230936e-01 -1.38458955e+00 4.67529476e-01
1.06891835e+00 1.03450537e+00 4.47503597e-01 6.67929292e-01
-3.61784697e-01 6.06558740e-01 -7.46975064e-01 -2.63019502e-01
3.44804257e-01 -3.18476707e-01 -7.53203094e-01 1.95635691e-01
6.58758819e-01 -4.33271348e-01 -2.06093967e-01 1.04928672e+00
6.29221022e-01 -3.95614445e-01 5.23992181e-01 -1.25818789e+00
-5.90885043e-01 2.26253271e-01 -7.64473736e-01 4.39284772e-01
4.63162102e-02 1.79375842e-01 7.86928117e-01 -8.72776091e-01
7.01659977e-01 8.48979056e-01 5.10328591e-01 5.34106970e-01
-1.01761591e+00 -1.08852899e+00 -1.57154411e-01 1.23975329e-01
-1.41901326e+00 -2.18572736e-01 5.49094915e-01 -4.12659347e-01
6.38450086e-01 2.88306713e-01 4.87325400e-01 1.03097332e+00
1.13289207e-01 1.05629349e+00 1.20761085e+00 -4.22262013e-01
2.09831614e-02 1.71164498e-01 2.49384329e-01 1.01628900e+00
3.96077186e-01 -2.83743531e-01 -4.82662134e-02 6.55271998e-03
9.88899350e-01 3.19647491e-01 -2.60677159e-01 -3.62646937e-01
-1.08310509e+00 5.80098271e-01 6.55715585e-01 3.95922422e-01
2.08500270e-02 -3.35966498e-01 4.40694183e-01 6.67251572e-02
2.86154091e-01 3.35803986e-01 -4.25987095e-01 2.20570788e-01
-9.46916938e-01 3.18744108e-02 3.70753348e-01 1.12527144e+00
7.16776371e-01 -1.61601037e-01 -1.02281161e-01 8.42463911e-01
1.42457202e-01 4.56546217e-01 7.12689996e-01 -5.26802957e-01
4.13043559e-01 8.68891776e-01 -1.53139919e-01 -8.42731595e-01
-5.44347525e-01 -4.98859912e-01 -1.14272225e+00 1.47769496e-01
4.98332143e-01 1.21422991e-01 -1.32789576e+00 1.43560970e+00
2.84030110e-01 7.43111670e-02 -8.16795826e-02 7.21704185e-01
1.19784486e+00 3.66316080e-01 1.25600070e-01 7.94032775e-03
1.62525558e+00 -1.04875565e+00 -8.46597373e-01 2.68586390e-02
8.80467355e-01 -7.15395033e-01 1.22068584e+00 4.27453697e-01
-7.88577795e-01 -5.21388650e-01 -1.36418998e+00 -2.17521667e-01
-7.86461830e-01 5.48025548e-01 7.48336017e-01 6.76744342e-01
-7.23728120e-01 2.54716098e-01 -9.49059069e-01 -2.62851447e-01
1.18738043e+00 3.04942310e-01 -7.26107717e-01 -1.92448601e-01
-8.31256807e-01 6.40912712e-01 2.50792414e-01 6.79363497e-03
-9.70053196e-01 -8.34179342e-01 -8.01692724e-01 3.80249135e-02
3.95201892e-01 -4.87354994e-01 9.31416154e-01 -1.09059954e+00
-9.46019590e-01 1.30135906e+00 9.77856889e-02 -4.03881133e-01
3.46035719e-01 -7.55641162e-02 -3.81533742e-01 3.94440264e-01
2.45681927e-01 7.07131684e-01 6.19243979e-01 -1.14252722e+00
-5.77491403e-01 -4.46388572e-01 4.16398384e-02 -1.36387289e-01
-5.30834079e-01 5.16276881e-02 -5.89411139e-01 -7.68080056e-01
-1.17718041e-01 -7.04847991e-01 -1.59524158e-01 2.69942015e-01
-6.64611340e-01 -1.90296188e-01 8.02702427e-01 -6.26936436e-01
1.05458260e+00 -2.20540428e+00 -2.73668885e-01 -7.42533505e-02
6.57382548e-01 5.16410887e-01 -9.92264599e-02 -4.74113896e-02
-4.45476681e-01 3.85541916e-01 -3.14993531e-01 -1.82708323e-01
-3.45902771e-01 -5.11678867e-02 7.28745833e-02 4.86004293e-01
2.59974986e-01 8.44716489e-01 -7.53909528e-01 -8.34632397e-01
1.92357935e-02 2.62439460e-01 -3.33973199e-01 2.47648314e-01
1.09761223e-01 3.78345221e-01 -3.81881207e-01 1.00992608e+00
8.84704947e-01 -8.26258183e-01 1.68409973e-01 -4.13517624e-01
7.12930858e-02 -2.34038889e-01 -8.45169842e-01 1.82419348e+00
-1.09764427e-01 4.10111040e-01 4.08550017e-02 -1.17546463e+00
5.97971916e-01 2.12335944e-01 5.27002513e-01 -6.05448306e-01
1.66625887e-01 1.70254886e-01 -7.01610669e-02 -5.64881861e-01
1.87923815e-02 7.48136789e-02 6.22278750e-02 1.01982117e-01
2.79958904e-01 3.19543123e-01 4.25307751e-01 5.09699643e-01
1.28285611e+00 4.77311388e-02 2.63885349e-01 -4.45820615e-02
5.61885357e-01 9.13704634e-02 7.91363001e-01 5.16855359e-01
-4.33539391e-01 8.19005311e-01 5.14103830e-01 -5.40946662e-01
-1.05322003e+00 -1.11580586e+00 -2.89636999e-01 6.58186376e-01
2.11456139e-02 -4.50419575e-01 -8.22726011e-01 -1.12027490e+00
-1.67427793e-01 1.87590256e-01 -5.07356524e-01 9.81400013e-02
-4.33192462e-01 -9.43736494e-01 7.25834012e-01 5.85905731e-01
5.51386595e-01 -8.42119217e-01 -4.37154502e-01 -2.74141908e-01
-9.56603810e-02 -1.19803286e+00 -5.13941348e-01 1.40282497e-01
-8.40197325e-01 -1.40963686e+00 -1.04026234e+00 -1.02785552e+00
1.28073740e+00 2.93792635e-01 8.59325826e-01 2.19241396e-01
-1.07393086e+00 1.77062228e-01 -3.18250388e-01 -6.21301413e-01
-8.88864473e-02 2.41615381e-02 -1.66482717e-01 -8.37648511e-02
2.88895458e-01 -2.52124429e-01 -8.09405863e-01 2.49910325e-01
-1.16098642e+00 2.37260893e-01 1.02943897e+00 9.73684788e-01
9.64707196e-01 -1.28426418e-01 6.57851160e-01 -1.23951399e+00
2.99745023e-01 -3.70006830e-01 -6.24188006e-01 3.07800144e-01
-6.73044741e-01 -1.16159230e-01 6.34976149e-01 -2.16997713e-01
-9.04167712e-01 1.65089920e-01 -1.21489279e-01 -5.75799942e-01
-3.88862222e-01 3.57052356e-01 -2.68323392e-01 1.26955817e-02
5.16429305e-01 1.78888202e-01 1.97789326e-01 -5.93442857e-01
1.96313381e-01 8.72807980e-01 5.39722025e-01 -3.01020384e-01
6.87804759e-01 4.42452222e-01 -2.23549530e-01 -6.43928230e-01
-1.01713276e+00 -6.62331641e-01 -5.92589319e-01 -1.58118438e-02
9.81070220e-01 -1.01132977e+00 -4.97476190e-01 4.67105031e-01
-8.58805716e-01 -2.93742865e-01 1.16244696e-01 3.65879625e-01
-2.26020157e-01 6.22686148e-01 -7.30063379e-01 -4.86607432e-01
-6.42925620e-01 -1.21565127e+00 1.03407848e+00 5.23499370e-01
-1.78187601e-02 -5.67372084e-01 -1.33634686e-01 7.74469256e-01
1.70280501e-01 3.36451203e-01 9.23705280e-01 -8.26626062e-01
-6.99569166e-01 -3.62429023e-01 -5.49887836e-01 4.83970314e-01
2.60630727e-01 -4.26496975e-02 -8.84542882e-01 -3.59587908e-01
-3.65523010e-01 -5.33588231e-01 8.06533456e-01 2.02319607e-01
1.69887102e+00 -1.78077221e-01 -5.70125163e-01 7.71495223e-01
1.30354571e+00 2.12107152e-01 8.55498254e-01 4.18923467e-01
7.77011991e-01 2.70442486e-01 5.84657907e-01 1.53184980e-01
1.27088651e-01 4.34641063e-01 2.85366565e-01 -6.78654969e-01
-2.49538749e-01 -3.42354625e-02 -7.90632516e-02 6.93390727e-01
-6.13657981e-02 1.02754487e-02 -8.28581274e-01 4.70776111e-01
-1.49664927e+00 -7.79648602e-01 -4.63016294e-02 2.13916993e+00
9.26849365e-01 1.93775088e-01 4.64350693e-02 -6.90150186e-02
6.46886766e-01 -2.38325849e-01 -6.62972450e-01 1.14312232e-01
-1.32821441e-01 3.38317215e-01 5.97137392e-01 -8.62705186e-02
-1.38406754e+00 7.23139584e-01 5.86394835e+00 1.07266712e+00
-9.91243124e-01 1.69823736e-01 1.05767798e+00 2.20626909e-02
1.71127990e-01 -4.09020305e-01 -8.39457214e-01 4.90388930e-01
6.66444123e-01 2.89579225e-03 -2.17619747e-01 1.06705213e+00
-8.22377577e-02 1.12687699e-01 -8.66251826e-01 1.22983611e+00
8.82344842e-02 -1.56532514e+00 1.16450630e-01 1.96064822e-02
5.23417413e-01 -9.12484452e-02 1.37581497e-01 3.50743294e-01
-1.23177797e-01 -1.09282804e+00 7.67517462e-02 2.20356971e-01
1.20302176e+00 -6.42604709e-01 8.76767278e-01 1.36554971e-01
-9.40769613e-01 1.32508174e-01 -3.64721477e-01 4.37076300e-01
-2.81706572e-01 6.33962154e-01 -7.74697483e-01 6.82688057e-01
8.77130389e-01 5.70364237e-01 -8.15213442e-01 1.21951807e+00
-1.54756352e-01 5.42299211e-01 -5.34175113e-02 1.38946012e-01
3.75698805e-02 -2.17372552e-03 1.73258975e-01 1.21301389e+00
8.34849402e-02 -8.98452476e-04 1.86388865e-01 8.41406524e-01
-4.34964508e-01 3.09836328e-01 -2.81236678e-01 -2.50795543e-01
2.17358559e-01 1.74218881e+00 -1.07050025e+00 -4.38056201e-01
-7.11628258e-01 9.23491478e-01 4.10551578e-01 2.57162303e-01
-9.40184593e-01 -6.31882370e-01 3.39842260e-01 2.33882055e-01
3.43982875e-01 1.75789803e-01 -1.31528214e-01 -1.24683845e+00
-1.48188025e-01 -1.11180270e+00 6.13517404e-01 -6.47286832e-01
-1.33723652e+00 6.78991914e-01 -2.22763553e-01 -1.30069041e+00
3.68884802e-01 -8.74835253e-01 -3.66376162e-01 6.11046791e-01
-1.48916161e+00 -1.10769200e+00 -4.54668432e-01 5.94853818e-01
5.38232327e-01 -4.92351681e-01 1.07207596e+00 6.55852914e-01
-9.44460571e-01 9.21410143e-01 1.59378707e-01 6.76014960e-01
9.36421871e-01 -1.30389476e+00 2.43513733e-02 6.99503779e-01
1.64401799e-01 7.47647226e-01 2.67968208e-01 -5.16499579e-01
-1.36037493e+00 -1.08624649e+00 4.35533494e-01 -3.19694161e-01
5.22674561e-01 -5.12848914e-01 -8.53918076e-01 6.79266989e-01
1.41360760e-01 5.15426636e-01 1.08138561e+00 -8.53317901e-02
-5.40598154e-01 -3.04487288e-01 -1.26596498e+00 4.26643461e-01
6.06511414e-01 -1.84351563e-01 -3.34788620e-01 6.58485949e-01
4.06189919e-01 -5.18882096e-01 -9.05084968e-01 4.50749069e-01
5.72567105e-01 -7.61332750e-01 9.46337461e-01 -5.67598224e-01
5.42423844e-01 -3.62460852e-01 2.49663532e-01 -7.08711028e-01
-1.63676664e-01 -3.96169722e-01 -1.22702077e-01 1.06582654e+00
5.15233994e-01 -5.31852603e-01 1.04109621e+00 4.35551763e-01
-3.03349614e-01 -1.18901908e+00 -5.95686436e-01 -6.53356969e-01
2.16949619e-02 8.57101455e-02 2.53675282e-01 9.82061684e-01
2.13797297e-02 3.16683829e-01 -1.55357450e-01 9.50825140e-02
7.18000710e-01 1.94261461e-01 8.51597369e-01 -9.93576050e-01
-3.15518618e-01 -3.08927268e-01 -5.62104404e-01 -8.80064011e-01
-1.63599268e-01 -1.04608119e+00 -2.28253141e-01 -1.55337691e+00
6.65030897e-01 -4.05180007e-01 -6.22090161e-01 7.79471695e-01
-3.23907495e-01 4.21286315e-01 4.13074493e-02 2.84602374e-01
-8.43271911e-01 3.19605097e-02 1.42903471e+00 -3.62476736e-01
1.49089456e-01 -1.85635865e-01 -8.10397089e-01 6.76152408e-01
9.68869984e-01 -3.79778057e-01 -2.77385563e-01 -1.17350094e-01
-1.40341952e-01 -2.32911900e-01 2.46062428e-01 -9.57949519e-01
9.54530016e-02 2.26538062e-01 8.03377330e-01 -8.46349001e-01
-9.58999619e-02 -5.43299615e-01 -1.47884518e-01 4.42356855e-01
-2.59633362e-01 -2.37634167e-01 4.38752174e-01 6.53256774e-01
-2.22721294e-01 -9.70048532e-02 9.16228712e-01 -3.22458684e-01
-5.71361184e-01 5.06890833e-01 -2.62977183e-01 -1.13799982e-01
1.12819600e+00 -7.87532851e-02 -4.58056986e-01 -3.87604311e-02
-6.49975359e-01 2.25998774e-01 3.22873771e-01 2.39444613e-01
6.07181668e-01 -1.18983793e+00 -5.36533713e-01 1.32677391e-01
3.58803600e-01 2.34737709e-01 4.83650774e-01 1.00354230e+00
-8.44473004e-01 4.09991741e-01 -3.36658239e-01 -4.97438937e-01
-1.64005888e+00 8.57792139e-01 2.99964193e-02 -5.81698060e-01
-7.70794213e-01 8.49782050e-01 6.52126729e-01 -4.28246498e-01
3.06912690e-01 -2.48465121e-01 -1.49672151e-01 -2.03012779e-01
8.72708559e-01 8.63406137e-02 1.03860021e-01 -4.13357496e-01
-3.04661870e-01 1.40995026e-01 -6.84175491e-01 6.47198319e-01
1.25527740e+00 1.73587158e-01 2.98294704e-02 1.64848849e-01
1.42900693e+00 -9.12887752e-02 -9.05669630e-01 6.00111559e-02
-1.76027104e-01 -2.67797858e-01 9.75419953e-02 -9.30334687e-01
-1.24105406e+00 8.06478322e-01 9.39529598e-01 -1.12982199e-01
1.18508852e+00 -3.97181064e-02 8.54698062e-01 2.37126291e-01
2.59393424e-01 -1.00276613e+00 1.93394348e-01 1.94859284e-03
5.49235225e-01 -1.53552413e+00 2.10269123e-01 -6.11088157e-01
-3.65450263e-01 1.07267118e+00 7.27523029e-01 5.23982979e-02
4.30474788e-01 5.26126027e-01 1.35978341e-01 -3.76968235e-01
-3.09535056e-01 -1.11820824e-01 2.31679186e-01 4.26846266e-01
6.70266688e-01 -1.13897815e-01 -3.28323781e-01 9.16858435e-01
1.80957973e-01 9.74820778e-02 3.68125528e-01 9.71217573e-01
-2.71391958e-01 -1.17250192e+00 -4.01873231e-01 7.58590877e-01
-8.63918185e-01 -1.86002925e-01 -2.93942124e-01 8.62754285e-01
1.61106363e-01 5.52865505e-01 -1.48582458e-03 -1.17102303e-02
1.45224676e-01 -1.60379633e-02 5.16882062e-01 -7.12064743e-01
-2.23496005e-01 2.19608873e-01 -2.14493006e-01 -5.28050125e-01
-3.03726345e-01 -2.88673609e-01 -1.46913052e+00 1.57540739e-01
-3.22825253e-01 3.69867422e-02 4.99993443e-01 7.14374423e-01
5.07986903e-01 7.73202479e-01 3.46221864e-01 -5.10664642e-01
-1.57148525e-01 -1.00095975e+00 -6.60627663e-01 5.77612817e-01
9.35434327e-02 -6.41335309e-01 -2.35760868e-01 3.11549634e-01] | [15.041478157043457, -2.5126662254333496] |
c8fcfbaf-ba0e-400c-bf9c-53aaa6d99797 | applying-deep-reinforcement-learning-to-the | 2211.14939 | null | https://arxiv.org/abs/2211.14939v2 | https://arxiv.org/pdf/2211.14939v2.pdf | Applying Deep Reinforcement Learning to the HP Model for Protein Structure Prediction | A central problem in computational biophysics is protein structure prediction, i.e., finding the optimal folding of a given amino acid sequence. This problem has been studied in a classical abstract model, the HP model, where the protein is modeled as a sequence of H (hydrophobic) and P (polar) amino acids on a lattice. The objective is to find conformations maximizing H-H contacts. It is known that even in this reduced setting, the problem is intractable (NP-hard). In this work, we apply deep reinforcement learning (DRL) to the two-dimensional HP model. We can obtain the conformations of best known energies for benchmark HP sequences with lengths from 20 to 50. Our DRL is based on a deep Q-network (DQN). We find that a DQN based on long short-term memory (LSTM) architecture greatly enhances the RL learning ability and significantly improves the search process. DRL can sample the state space efficiently, without the need of manual heuristics. Experimentally we show that it can find multiple distinct best-known solutions per trial. This study demonstrates the effectiveness of deep reinforcement learning in the HP model for protein folding. | ['Liang Dai', 'Roland H. C. Yap', 'Fujia Tian', 'Adithya Murali', 'Olafs Vandans', 'Houjing Huang', 'Kaiyuan Yang'] | 2022-11-27 | null | null | null | null | ['protein-folding'] | ['natural-language-processing'] | [ 6.99398145e-02 9.17660743e-02 -1.23740308e-01 -1.63430229e-01
-7.95128345e-01 -5.31420767e-01 -1.31691590e-01 1.21907674e-01
-6.67901695e-01 1.29534507e+00 -1.87833294e-01 -7.51675010e-01
1.74488306e-01 -6.71200514e-01 -1.46927679e+00 -1.09570456e+00
-4.65094686e-01 7.41860509e-01 -7.05113038e-02 -2.69693613e-01
4.39099222e-01 6.20467126e-01 -1.10697746e+00 1.54714435e-01
9.50505018e-01 6.44479334e-01 7.25966811e-01 7.59037912e-01
1.38940617e-01 4.55995381e-01 -3.92054081e-01 -1.53345824e-03
1.69196859e-01 -7.14790285e-01 -1.23389339e+00 -3.96907955e-01
1.93828553e-01 -8.95382091e-02 2.39584781e-02 8.66825402e-01
7.48432696e-01 1.34924620e-01 4.16140169e-01 -5.41376650e-01
-6.17492139e-01 1.50849536e-01 -1.53549045e-01 1.31807327e-01
3.78964722e-01 4.98394102e-01 1.26618016e+00 -6.26213908e-01
8.89291227e-01 1.13224685e+00 5.44485569e-01 6.99684322e-01
-1.67278242e+00 -2.26428270e-01 -1.07356302e-01 6.09799445e-01
-9.76587772e-01 7.34002441e-02 4.21074152e-01 -1.03053883e-01
1.86406279e+00 -3.15853119e-01 1.03795135e+00 1.12328720e+00
9.30470943e-01 6.98378086e-01 1.08884621e+00 -5.08225977e-01
6.68466628e-01 -7.43021965e-01 1.40851557e-01 8.00414085e-01
-1.71417698e-01 4.94806617e-01 -4.68687117e-01 -6.12403512e-01
3.68503004e-01 8.10882822e-02 -1.12404369e-01 -6.40365362e-01
-9.61853266e-01 1.02047610e+00 5.58933854e-01 -3.64564247e-02
-6.54820263e-01 1.16649404e-01 1.78494558e-01 5.45909524e-01
-2.39822090e-01 1.05769837e+00 -9.34421718e-01 -1.90492913e-01
-5.39094806e-01 7.45832562e-01 1.09922433e+00 4.10808802e-01
8.69395733e-01 -2.87640661e-01 3.25291604e-01 4.66814101e-01
1.51801094e-01 5.30788720e-01 2.37917170e-01 -1.12617815e+00
1.56937316e-01 3.21084797e-01 5.09626567e-01 -3.69056582e-01
-7.38561928e-01 -2.33524982e-02 -6.50500238e-01 3.55378419e-01
5.12854755e-01 -4.18443710e-01 -8.16360593e-01 1.97528851e+00
1.63676530e-01 -6.23011254e-02 4.18509722e-01 7.71538734e-01
2.55446523e-01 1.18048215e+00 2.18997717e-01 -9.19230819e-01
1.04736567e+00 -8.30025554e-01 -5.88945687e-01 -2.75949389e-02
5.44008434e-01 -4.36004400e-01 8.52407753e-01 6.80041015e-01
-1.26101756e+00 -4.30361569e-01 -1.25633597e+00 -1.41228199e-01
-3.14364403e-01 -3.32478017e-01 4.30166334e-01 1.82991251e-01
-1.13903475e+00 1.25383282e+00 -8.33443642e-01 -3.35482687e-01
1.72970761e-02 8.90236199e-01 -2.34187171e-01 1.57970250e-01
-1.58209562e+00 1.18960190e+00 6.72581494e-01 3.97191942e-02
-9.89220738e-01 -3.52508068e-01 -4.33137447e-01 -6.83237016e-02
3.27709794e-01 -5.83658397e-01 1.48165476e+00 -1.17370188e+00
-1.69165182e+00 5.29226184e-01 -4.85408098e-01 -9.17882502e-01
-1.16388360e-02 -2.96562407e-02 1.17416747e-01 1.32136643e-01
-4.80032533e-01 9.01011467e-01 4.28239346e-01 -8.57558489e-01
-7.76695460e-02 -4.93394792e-01 -5.58104105e-02 3.19801152e-01
3.94490868e-01 -1.04888976e-01 3.70728761e-01 -1.26122400e-01
-5.12426021e-03 -1.27962744e+00 -7.19124019e-01 -4.07579511e-01
-6.42339736e-02 -4.46831107e-01 2.34902456e-01 -7.61899948e-01
9.29945350e-01 -1.40673995e+00 6.66958094e-01 3.89659464e-01
1.15214519e-01 3.70417655e-01 -1.94509491e-01 9.35190141e-01
-1.23960719e-01 -1.58172101e-01 -2.19731420e-01 4.40846950e-01
-1.60979822e-01 5.81059933e-01 -2.90366024e-01 3.69973332e-01
9.71856266e-02 1.29429042e+00 -7.49810636e-01 -1.42100766e-01
-2.34056845e-01 2.04889923e-01 -7.06349432e-01 3.91073525e-01
-9.65081751e-01 3.09894979e-01 -3.10932070e-01 3.58740628e-01
6.49524093e-01 -6.94500327e-01 1.09193814e+00 -4.68163611e-03
-1.75567850e-01 4.02565271e-01 -6.34122849e-01 1.58193839e+00
-8.66714939e-02 1.93629548e-01 -3.74585897e-01 -1.06664205e+00
9.84571278e-01 1.92476645e-01 5.67220807e-01 -9.61234272e-01
-1.39063939e-01 2.86203533e-01 6.57649100e-01 -4.22152430e-01
3.92746963e-02 -2.91966528e-01 2.06469014e-01 6.29079759e-01
2.84272972e-02 4.24137354e-01 2.24887282e-01 -2.05337390e-01
1.16036451e+00 5.02298653e-01 5.49252331e-01 -3.58769000e-01
2.61399657e-01 9.75211635e-02 8.04681063e-01 6.65824056e-01
-1.51600301e-01 1.04163259e-01 8.44083130e-01 -9.83589530e-01
-1.46355236e+00 -8.64804924e-01 1.69496108e-02 1.20159400e+00
-5.95535571e-03 -3.46102089e-01 -1.16442859e+00 -4.06313121e-01
2.67632101e-02 2.55867302e-01 -4.55837488e-01 -2.20261201e-01
-1.14742267e+00 -1.00972247e+00 2.72254702e-02 3.31714362e-01
3.53426710e-02 -1.89395785e+00 -1.02874637e+00 8.54618311e-01
-4.91144881e-02 -4.25356895e-01 -3.94284189e-01 9.03717399e-01
-1.14494824e+00 -1.08453345e+00 -6.64497137e-01 -1.07934928e+00
3.60755861e-01 -2.29617044e-01 1.33218646e+00 -2.83740591e-02
-2.98353046e-01 -4.40994203e-01 -9.78379622e-02 1.04423240e-02
-5.25061786e-01 2.96865374e-01 1.66207775e-01 -5.96035480e-01
6.93419337e-01 -6.28294110e-01 -9.02637362e-01 2.20526338e-01
-6.01334333e-01 5.29448241e-02 6.13710940e-01 1.37703729e+00
9.72920001e-01 -5.93073845e-01 7.36775577e-01 -6.24250412e-01
6.87046468e-01 -2.56916910e-01 -7.77172029e-01 5.92334807e-01
-8.43943000e-01 8.47524822e-01 1.11211538e+00 -3.46751362e-01
-5.20028770e-01 4.60176259e-01 -4.74835247e-01 -5.88335469e-02
-2.68037375e-02 3.78302515e-01 -1.41970858e-01 -3.48519236e-01
5.96949518e-01 5.96149743e-01 2.42305532e-01 -4.30863649e-01
1.43105581e-01 2.55462319e-01 7.37268850e-02 -8.90533090e-01
-7.88546540e-03 -2.16297701e-01 2.49194816e-01 -6.76397502e-01
-5.55877388e-01 2.42796391e-02 -5.31661749e-01 3.33778024e-01
7.86551058e-01 -4.48380023e-01 -1.63305974e+00 1.53105810e-01
-1.31807911e+00 -6.72958851e-01 8.59830976e-02 2.13122055e-01
-1.23609352e+00 7.16525316e-01 -8.50290000e-01 -4.95939314e-01
-5.13498008e-01 -1.42437780e+00 8.24379027e-01 1.61020786e-01
-1.91265628e-01 -6.61568165e-01 6.66702151e-01 4.28027920e-02
4.23460826e-02 2.05712944e-01 1.46256447e+00 -4.21995848e-01
-6.30649209e-01 4.71932113e-01 2.52734333e-01 9.21374485e-02
-3.62317652e-01 -1.41072765e-01 -4.66852963e-01 -6.09599829e-01
-2.32047454e-01 -8.28083277e-01 1.00212467e+00 7.52527773e-01
1.17791569e+00 -4.97836322e-01 -1.69759303e-01 5.29530823e-01
1.43463814e+00 7.33550072e-01 6.53638303e-01 4.00003642e-01
1.79703578e-01 4.06193078e-01 5.80325365e-01 3.36187690e-01
6.63667172e-02 5.99018753e-01 5.51217496e-01 -9.21781808e-02
5.75643480e-01 -2.94459432e-01 4.97460812e-01 5.21743238e-01
-2.26201534e-01 -3.27262163e-01 -1.07558799e+00 -7.56411180e-02
-1.99640608e+00 -1.06683993e+00 5.06458461e-01 1.92062187e+00
1.34347069e+00 -2.60225628e-02 3.55168253e-01 -3.09727520e-01
4.04945314e-01 -1.03616834e-01 -1.43714499e+00 -9.09439206e-01
-2.26202771e-01 5.87737679e-01 4.76416588e-01 7.84488738e-01
-9.27629590e-01 1.08632874e+00 7.08822727e+00 4.23362553e-01
-9.98239517e-01 -3.72326851e-01 7.94781208e-01 -1.89864468e-02
1.08052783e-01 -9.45520867e-03 -8.35794270e-01 3.99807572e-01
1.51591933e+00 -8.10007751e-03 8.58448505e-01 6.80836499e-01
3.97598147e-01 -2.40489785e-02 -1.10435677e+00 7.27120876e-01
-4.28157657e-01 -1.65753508e+00 -5.35106026e-02 3.01180035e-01
7.75547206e-01 1.55741736e-01 8.62638466e-03 2.26774514e-01
4.00877327e-01 -1.36992419e+00 1.05328307e-01 5.60627699e-01
5.51171184e-01 -1.29869807e+00 5.96214533e-01 5.81751287e-01
-5.88644862e-01 -1.64441943e-01 -8.89757574e-01 -7.69904479e-02
-4.52836826e-02 1.86485723e-02 -8.27155113e-01 -4.38391678e-02
3.93177569e-01 6.10958517e-01 2.92226262e-02 6.63106263e-01
7.78556839e-02 5.14936507e-01 -1.78427950e-01 -6.14438355e-01
4.56866682e-01 -7.22506762e-01 2.24478100e-03 9.45286989e-01
-1.30676910e-01 4.99416888e-01 4.91487533e-01 7.66829848e-01
-2.95839578e-01 1.37015209e-01 -4.11459059e-01 -7.46543631e-02
1.21921398e-01 6.94180608e-01 -4.27037328e-01 -1.29442468e-01
-2.49576047e-02 7.78858006e-01 8.68345141e-01 5.35312116e-01
-6.16255105e-01 -3.32109839e-01 7.60253847e-01 -1.36772767e-01
5.85173070e-01 -2.67169267e-01 3.64837468e-01 -8.16834211e-01
-9.42237750e-02 -1.45650661e+00 1.53346822e-01 -3.83808523e-01
-1.19651997e+00 4.88556087e-01 -6.52613282e-01 -4.34281230e-01
-4.34514493e-01 -9.16513026e-01 -9.38203782e-02 9.92163062e-01
-1.51246500e+00 -4.12470847e-01 5.77488244e-01 1.50858954e-01
4.44628358e-01 -6.57212064e-02 1.07657325e+00 -3.68712276e-01
-4.47638929e-01 5.05842686e-01 9.23386931e-01 -4.74413306e-01
4.68479902e-01 -1.61584711e+00 7.51045763e-01 2.67782807e-01
-1.74910247e-01 8.53905737e-01 9.87195849e-01 -6.93819225e-01
-1.98397791e+00 -7.54479170e-01 9.44544971e-01 -7.69863129e-02
2.53551751e-01 -2.95520425e-01 -1.15403795e+00 2.92728335e-01
3.37112308e-01 -2.66865134e-01 6.78672433e-01 -3.94761637e-02
-1.64302401e-02 5.74940480e-02 -1.05305922e+00 4.70066488e-01
9.67813611e-01 -3.56988728e-01 -5.59826791e-01 6.79180622e-01
7.44673550e-01 -3.39507550e-01 -9.19653237e-01 2.04648063e-01
5.97778082e-01 -9.46695745e-01 1.05806708e+00 -1.37569392e+00
4.04117942e-01 -1.67710975e-01 -8.05381611e-02 -1.24173760e+00
-4.50040162e-01 -9.09717739e-01 -5.70115805e-01 -2.09651306e-01
5.88470340e-01 -4.87725526e-01 9.77818191e-01 3.15017790e-01
1.03503972e-01 -1.51427865e+00 -9.37946200e-01 -7.97659755e-01
6.71589375e-01 4.28613365e-01 5.14560282e-01 3.70249391e-01
2.96906114e-01 5.07405400e-01 -4.98183638e-01 -1.16774485e-01
5.24143457e-01 7.49630034e-01 2.78676957e-01 -1.01025403e+00
-7.01173604e-01 2.28644628e-02 2.41468102e-01 -1.48512328e+00
5.94558179e-01 -8.44915450e-01 2.25955963e-01 -1.35414112e+00
4.42941934e-01 -3.27385999e-02 -3.40039253e-01 5.09142995e-01
5.66226877e-02 -4.42753822e-01 -2.87520066e-02 3.20724100e-02
-8.08049262e-01 7.57158160e-01 1.36043000e+00 -3.12232971e-02
-2.24779725e-01 -1.04130276e-01 -2.51992881e-01 2.99654186e-01
1.16718233e+00 -4.92250532e-01 -9.32269245e-02 9.96927693e-02
7.68137455e-01 5.38752615e-01 -2.52466574e-02 -7.01968908e-01
-4.70314436e-02 -4.24988627e-01 4.24648553e-01 -6.65428638e-01
1.79199189e-01 -4.22498763e-01 -3.58745083e-02 1.11211038e+00
-6.26186490e-01 5.68750083e-01 2.71349438e-02 6.23517454e-01
1.27317727e-01 -1.23701960e-01 1.00768483e+00 -3.79100531e-01
-2.87354529e-01 3.30582470e-01 -5.64427316e-01 6.53878450e-02
9.17846322e-01 -1.45569388e-02 -6.76278695e-02 -9.65544730e-02
-8.82677972e-01 2.71277964e-01 6.40272737e-01 -1.52876109e-01
7.00566113e-01 -8.26456606e-01 -3.19818705e-01 1.73038036e-01
-2.56577432e-01 -4.08727050e-01 1.66450050e-02 3.93064946e-01
-9.81813490e-01 7.80100882e-01 -4.77198690e-01 -3.50996107e-01
-1.21875632e+00 8.98904979e-01 7.52468467e-01 -5.46660721e-01
-4.99556482e-01 4.03190345e-01 -1.68794140e-01 -4.39255089e-01
1.89403310e-01 -3.67975205e-01 2.69044209e-02 -2.92461962e-01
3.86859030e-01 2.31426090e-01 6.87829256e-02 -1.09787613e-01
-2.11994261e-01 5.56558251e-01 -4.95962679e-01 2.45918825e-01
1.49657857e+00 4.65058982e-01 -2.09860370e-01 9.27873179e-02
1.18670917e+00 -8.09427321e-01 -1.47774935e+00 -7.11902678e-02
3.91177833e-01 3.17894310e-01 -3.96282613e-01 -9.45506275e-01
-3.06904823e-01 9.97898579e-01 8.34557056e-01 -1.28078356e-01
8.27960670e-01 -2.87911415e-01 1.24267185e+00 1.50137711e+00
5.79753816e-01 -1.16588771e+00 4.37143184e-02 9.62347865e-01
6.54986620e-01 -1.15861964e+00 -1.85108662e-01 6.05247021e-01
-5.58800220e-01 1.50836408e+00 4.46796268e-01 -3.35578471e-01
1.90062463e-01 1.68199316e-01 -3.44095938e-02 -9.53396037e-02
-1.34120536e+00 1.33147612e-02 -1.94507256e-01 3.20579141e-01
5.15022159e-01 1.24346964e-01 -5.41409254e-01 8.14357698e-02
-4.09615710e-02 -6.45810142e-02 3.11001062e-01 9.34831560e-01
-9.18071687e-01 -1.58680618e+00 -1.01868957e-01 1.36149496e-01
-4.18094963e-01 -1.10149205e-01 -6.01628959e-01 1.94433287e-01
-2.84210015e-02 5.65701902e-01 -3.58710915e-01 7.46491402e-02
1.52266528e-02 3.55585068e-01 9.19692338e-01 -1.76114559e-01
-8.31369162e-01 -1.34114757e-01 -2.49226287e-01 -7.96682894e-01
-2.09332064e-01 -5.68223417e-01 -1.64442813e+00 -3.62607688e-01
-7.21210241e-02 5.44380903e-01 4.02225554e-01 7.52656162e-01
7.44596303e-01 1.02171883e-01 6.83634162e-01 -6.83725655e-01
-1.23078597e+00 -6.67390525e-01 -5.56966603e-01 4.11871821e-02
6.92287207e-01 -4.30802971e-01 1.04130276e-01 -1.27112210e-01] | [4.7280402183532715, 5.582693576812744] |
11d6832a-36b9-40d1-ab9a-c0bc6fcc26f7 | pay-attention-to-relations-multi-embeddings | 2203.01903 | null | https://arxiv.org/abs/2203.01903v1 | https://arxiv.org/pdf/2203.01903v1.pdf | Pay Attention to Relations: Multi-embeddings for Attributed Multiplex Networks | Graph Convolutional Neural Networks (GCNs) have become effective machine learning algorithms for many downstream network mining tasks such as node classification, link prediction, and community detection. However, most GCN methods have been developed for homogenous networks and are limited to a single embedding for each node. Complex systems, often represented by heterogeneous, multiplex networks present a more difficult challenge for GCN models and require that such techniques capture the diverse contexts and assorted interactions that occur between nodes. In this work, we propose RAHMeN, a novel unified relation-aware embedding framework for attributed heterogeneous multiplex networks. Our model incorporates node attributes, motif-based features, relation-based GCN approaches, and relational self-attention to learn embeddings of nodes with respect to the various relations in a heterogeneous, multiplex network. In contrast to prior work, RAHMeN is a more expressive embedding framework that embraces the multi-faceted nature of nodes in such networks, producing a set of multi-embeddings that capture the varied and diverse contexts of nodes. We evaluate our model on four real-world datasets from Amazon, Twitter, YouTube, and Tissue PPIs in both transductive and inductive settings. Our results show that RAHMeN consistently outperforms comparable state-of-the-art network embedding models, and an analysis of RAHMeN's relational self-attention demonstrates that our model discovers interpretable connections between relations present in heterogeneous, multiplex networks. | ['Siddharth Krishnan', 'Michael Ridenhour', 'Joshua Melton'] | 2022-03-03 | null | null | null | null | ['network-embedding'] | ['methodology'] | [ 5.52265085e-02 3.59420121e-01 -7.87762225e-01 -6.55272603e-02
9.76551771e-02 -7.30084896e-01 9.58500803e-01 5.82682908e-01
-1.65774748e-02 4.86881196e-01 5.51121473e-01 -4.01877731e-01
-7.06847429e-01 -1.32545161e+00 -5.02422035e-01 -5.22071183e-01
-7.54055142e-01 7.88477838e-01 2.15859473e-01 -3.85397553e-01
-3.76187950e-01 4.57585812e-01 -7.55780697e-01 1.94178328e-01
3.57047200e-01 5.45752048e-01 -3.20990264e-01 6.99014962e-01
-1.25218526e-01 7.23651409e-01 -1.66830346e-01 -7.49627769e-01
-4.40304317e-02 -1.90888822e-01 -8.52027893e-01 -4.88939136e-01
9.83845219e-02 1.45752296e-01 -1.26394701e+00 6.80138946e-01
3.70023131e-01 -2.46481106e-01 8.44153285e-01 -1.73411095e+00
-8.51408601e-01 1.07623208e+00 -2.25919649e-01 5.46786785e-01
9.11058784e-02 9.16799679e-02 1.95281792e+00 -4.74363834e-01
1.21399415e+00 1.42088127e+00 1.06014383e+00 3.15677792e-01
-1.92179728e+00 -3.83744568e-01 1.76157489e-01 4.27470431e-02
-1.26928008e+00 -4.38206606e-02 6.64441049e-01 -3.93354088e-01
1.16026092e+00 1.62972555e-01 1.06341314e+00 1.35244346e+00
3.07249427e-01 3.55388105e-01 4.95700449e-01 -4.35446447e-04
-2.72671610e-01 -2.41741374e-01 2.91283011e-01 7.44583368e-01
4.61801261e-01 -4.03632522e-02 -4.70766187e-01 -5.75746715e-01
7.73793638e-01 4.89122242e-01 -1.28474757e-01 -2.96598554e-01
-1.37397289e+00 9.67140794e-01 9.74780917e-01 5.00642180e-01
-3.36643606e-01 5.51093340e-01 6.36987746e-01 4.49199915e-01
6.68843567e-01 4.82002765e-01 -5.21734595e-01 1.45524055e-01
-5.07288456e-01 8.84137824e-02 1.11600459e+00 8.64828289e-01
7.60777533e-01 -5.78555644e-01 -1.16228715e-01 6.83031201e-01
4.55803901e-01 -1.38561577e-01 1.23836733e-01 -5.16241074e-01
2.87230521e-01 1.23534882e+00 -7.01651812e-01 -1.60771203e+00
-5.87107420e-01 -5.26442170e-01 -1.08881068e+00 -5.34198642e-01
6.04233071e-02 1.75381795e-01 -4.91959900e-01 1.87752581e+00
4.54555035e-01 4.18198347e-01 -1.90357521e-01 4.41886127e-01
1.23592138e+00 4.15155262e-01 2.01404080e-01 2.88103491e-01
1.20893180e+00 -7.85002112e-01 -6.19117081e-01 7.82775134e-02
8.51664305e-01 -6.39687553e-02 5.79069853e-01 -2.72468090e-01
-6.53224289e-01 1.24031305e-01 -7.64414251e-01 -3.10420722e-01
-8.86677623e-01 -5.76793909e-01 1.16877449e+00 1.34233087e-01
-1.19052923e+00 8.16250265e-01 -6.34096384e-01 -7.70920575e-01
7.67944038e-01 5.57958841e-01 -8.38501751e-01 -1.30427316e-01
-1.53829777e+00 6.45819783e-01 3.06951195e-01 6.43855259e-02
-7.25693107e-01 -9.88145053e-01 -8.83201301e-01 4.00830984e-01
3.81770343e-01 -8.67874205e-01 3.38804334e-01 -4.58981544e-01
-8.19751561e-01 8.64402652e-01 1.29320383e-01 -3.77563149e-01
-4.96019535e-02 4.40825701e-01 -6.13996029e-01 2.29200616e-01
5.29309139e-02 5.09721279e-01 2.89328322e-02 -9.08461213e-01
-4.06595729e-02 -1.79100558e-01 3.29519391e-01 -1.23280324e-01
-5.98275602e-01 -1.24425188e-01 -4.69396979e-01 -4.45049942e-01
9.54889432e-02 -9.18070674e-01 -1.33641779e-01 3.71017784e-01
-8.37580383e-01 -4.45441455e-01 8.65203440e-01 6.80729076e-02
1.39758110e+00 -1.85631919e+00 5.82140565e-01 6.86752558e-01
1.28891468e+00 1.03889987e-01 -4.99305367e-01 1.04068899e+00
-3.00573796e-01 6.50111794e-01 3.40449475e-02 -2.25639313e-01
1.26502112e-01 5.43055713e-01 1.92010105e-01 4.69678581e-01
5.79584181e-01 1.48856175e+00 -1.15959966e+00 -5.87166071e-01
-1.02770887e-01 6.97763085e-01 -5.74146152e-01 -3.70688513e-02
-2.81781167e-01 -1.15531147e-01 -4.02156055e-01 7.80158758e-01
1.51163071e-01 -8.60826254e-01 8.93304408e-01 -4.86629546e-01
4.30001736e-01 2.20225394e-01 -7.25990891e-01 1.18668866e+00
-1.70848116e-01 9.88309085e-01 2.98207738e-02 -1.14399636e+00
7.27334797e-01 2.72707224e-01 6.65157378e-01 -1.54590055e-01
7.14233369e-02 1.62659679e-02 5.13538659e-01 -4.16240871e-01
-1.01683643e-02 8.11409056e-02 9.63476077e-02 5.82843542e-01
5.44685841e-01 3.95000219e-01 2.80217379e-01 9.69371259e-01
1.92771316e+00 -4.18231517e-01 3.11505765e-01 -3.17597210e-01
4.91846632e-03 -1.93766281e-01 4.54039574e-01 4.92966980e-01
-2.07319468e-01 2.35722110e-01 1.51146877e+00 -4.86899287e-01
-9.63400424e-01 -1.06778193e+00 -1.84871927e-01 1.04399371e+00
-1.12414442e-01 -9.60508943e-01 -2.26296745e-02 -6.82650924e-01
5.31544805e-01 -4.11204189e-01 -1.13450086e+00 -3.60982567e-01
-3.68873209e-01 -1.10163510e+00 9.41259563e-01 3.69117707e-01
-1.08362287e-01 -7.83325970e-01 6.38988078e-01 3.93600792e-01
-3.46262194e-02 -1.19292378e+00 -4.11120713e-01 3.14422518e-01
-8.69462073e-01 -1.63952744e+00 -6.35945844e-03 -8.25018287e-01
5.02526820e-01 -8.82857442e-02 1.69639778e+00 5.56874454e-01
-5.79876781e-01 3.95964831e-01 -2.17381313e-01 6.38476387e-02
-3.30186158e-01 4.87401485e-01 1.58267692e-01 -6.24507628e-02
3.76840442e-01 -1.12437546e+00 -6.39115572e-01 1.79732233e-01
-1.05171752e+00 -3.73781443e-01 6.46680951e-01 1.16261756e+00
5.22635520e-01 -1.03014046e-02 7.70830095e-01 -1.44032049e+00
8.28995526e-01 -1.18801785e+00 3.27553265e-02 1.97708428e-01
-7.97027647e-01 -4.11816239e-02 4.63178188e-01 -5.97091734e-01
-1.28320843e-01 -4.59154516e-01 1.73674777e-01 -2.38228679e-01
3.71159554e-01 1.00345457e+00 -1.64670736e-01 -2.51735985e-01
6.05632842e-01 -3.52277607e-02 3.55435699e-01 -9.02508348e-02
6.08559608e-01 2.46135026e-01 1.45911053e-01 -4.78741497e-01
7.93434858e-01 5.22396088e-01 5.83367467e-01 -8.61131847e-01
-3.30847442e-01 -4.65688020e-01 -6.74137115e-01 3.68908010e-02
5.55439711e-01 -7.56806612e-01 -1.08634043e+00 8.99772793e-02
-9.17510211e-01 -2.39693940e-01 -5.04364856e-02 1.65586844e-01
-7.01609701e-02 2.55183190e-01 -1.16081655e+00 -3.53147298e-01
-2.78626502e-01 -7.77773678e-01 7.53415585e-01 -2.56635368e-01
-4.89362746e-01 -1.77284634e+00 4.66007054e-01 4.62406985e-02
4.38453585e-01 6.88492119e-01 1.47096980e+00 -9.14515257e-01
-6.88543618e-01 -2.42225185e-01 -6.12824440e-01 -2.80957311e-01
2.50469059e-01 4.56530273e-01 -5.82618594e-01 -1.38059715e-02
-1.34885359e+00 -2.57718980e-01 1.08211124e+00 9.29524750e-02
8.66774917e-01 -5.21778882e-01 -1.03862762e+00 7.59106338e-01
1.28630710e+00 -5.16149580e-01 4.62024361e-01 9.79562998e-02
1.06629515e+00 6.25784636e-01 -1.54347062e-01 2.03157160e-02
6.14948690e-01 4.48277920e-01 7.97634125e-01 -2.18133003e-01
1.34066064e-02 -5.10504723e-01 7.53885740e-03 7.86944985e-01
-9.68150795e-02 -6.29307628e-01 -8.86277437e-01 7.03139663e-01
-2.02808213e+00 -1.00917840e+00 -4.56049323e-01 1.34484673e+00
9.62720931e-01 1.50819257e-01 2.50288904e-01 -9.17690620e-02
6.71347022e-01 3.84249330e-01 -5.78323543e-01 -1.39192730e-01
-5.01256347e-01 1.79246813e-01 1.66050807e-01 3.85687560e-01
-9.62059021e-01 8.36500049e-01 6.35005283e+00 5.29306710e-01
-7.76269555e-01 3.39803621e-02 5.32789052e-01 -1.36597052e-01
-7.48510420e-01 4.17275503e-02 -3.87342900e-01 4.09903377e-01
1.13703477e+00 -1.41708404e-01 3.97574723e-01 3.56947035e-01
-1.33270860e-01 5.08835137e-01 -1.47712171e+00 6.66122615e-01
-4.07479793e-01 -2.03205872e+00 8.58299807e-02 5.44521570e-01
6.44547939e-01 4.48425233e-01 -8.99640098e-02 1.97362393e-01
7.17010736e-01 -1.46449614e+00 -7.14806095e-02 5.24215162e-01
7.36936152e-01 -4.25546259e-01 7.73822427e-01 -3.74808721e-02
-1.40087450e+00 3.11327409e-02 -2.11582124e-01 7.47249797e-02
-1.91552684e-01 8.87951553e-01 -9.60911036e-01 6.38622105e-01
6.02818668e-01 1.35813916e+00 -3.81145924e-01 6.22031271e-01
9.30241682e-03 6.61281049e-01 -3.80711138e-01 -2.56581366e-01
3.24661553e-01 -2.09491789e-01 5.26296318e-01 1.40590751e+00
-2.16184229e-01 -4.03708853e-02 -2.29360145e-02 1.18814766e+00
-6.43882453e-01 -1.25404075e-01 -9.90849674e-01 -8.18778157e-01
6.78360105e-01 1.64612734e+00 -7.05725610e-01 -3.21709961e-02
-4.13851291e-01 6.82515383e-01 8.63541186e-01 4.12324548e-01
-5.47886550e-01 -3.51875871e-01 1.11380255e+00 3.75839829e-01
5.31463884e-02 -1.48075178e-01 1.71696469e-01 -8.93015444e-01
-3.62872750e-01 -4.83951062e-01 5.58998823e-01 -1.74755037e-01
-1.99685335e+00 5.76415360e-01 -2.34812260e-01 -7.38062620e-01
5.68759590e-02 -8.04708719e-01 -8.82010996e-01 6.68206871e-01
-1.36957383e+00 -1.56918550e+00 -4.19958942e-02 4.51944023e-01
-3.05265099e-01 -8.37887377e-02 9.84559834e-01 5.23009181e-01
-8.25668931e-01 5.53581178e-01 1.17122166e-01 5.68496346e-01
4.42007869e-01 -1.36998987e+00 4.56117749e-01 2.17560381e-01
3.23631167e-01 1.08127224e+00 3.47164661e-01 -7.15699315e-01
-1.57373309e+00 -1.27456486e+00 1.12698185e+00 -7.04415917e-01
1.27138996e+00 -8.02559853e-01 -8.12571526e-01 1.08245480e+00
-9.46310014e-02 7.42442906e-01 1.20840192e+00 8.02139401e-01
-7.31439412e-01 -1.63227975e-01 -8.89001846e-01 5.94286561e-01
1.64802909e+00 -8.71034026e-01 5.74031286e-02 5.17110527e-01
9.74236429e-01 1.95516840e-01 -1.60477340e+00 2.47108430e-01
6.80907667e-01 -4.98417854e-01 1.22736835e+00 -1.20551121e+00
6.65927708e-01 -8.35087802e-03 6.48894459e-02 -1.45312464e+00
-7.57163703e-01 -5.46613753e-01 -5.29581606e-01 1.25978112e+00
7.75722802e-01 -9.44261849e-01 7.89176941e-01 2.13795930e-01
8.00986215e-02 -1.24070525e+00 -8.91107798e-01 -4.09607440e-01
-1.56825539e-02 -5.22046164e-03 6.54562116e-01 1.47108793e+00
3.59876871e-01 5.90986609e-01 -2.63887309e-02 4.06106040e-02
5.54925323e-01 -1.74514875e-02 5.22895455e-01 -1.80424762e+00
-3.17239493e-01 -7.48742759e-01 -1.04255176e+00 -5.09584665e-01
5.33976316e-01 -1.42546713e+00 -6.09930396e-01 -1.72237170e+00
5.14538229e-01 -8.64625692e-01 -3.95638496e-01 5.28482199e-01
-8.89578834e-02 4.11427438e-01 -2.51337290e-01 2.07623586e-01
-7.13434577e-01 5.04146218e-01 1.04348385e+00 -5.50699115e-01
5.67606725e-02 -3.35399508e-01 -8.81518483e-01 3.77525836e-01
4.75893319e-01 -4.06230450e-01 -4.05965507e-01 -1.87388688e-01
8.90407741e-01 -2.13843137e-01 5.22009552e-01 -2.33152583e-01
3.04074466e-01 1.09554477e-01 2.68116951e-01 -1.13023080e-01
4.10208672e-01 -7.96407402e-01 3.65971386e-01 3.86549056e-01
-4.79386240e-01 -8.22294652e-02 -1.31804138e-01 1.15921450e+00
1.42901270e-02 4.59565401e-01 2.01948792e-01 -3.47574316e-02
-3.42682570e-01 9.36727762e-01 -8.66008475e-02 1.28229320e-01
8.43155682e-01 -5.23286639e-03 -8.58111322e-01 -3.71819079e-01
-9.24417615e-01 4.71937478e-01 1.95949405e-01 3.71448070e-01
5.53330243e-01 -1.50176358e+00 -6.73243165e-01 -2.77547538e-03
3.36013883e-01 -8.10873508e-02 -3.38356360e-03 1.00971019e+00
-3.35117877e-01 1.75927192e-01 1.41648827e-02 -5.01075625e-01
-1.18331861e+00 4.60497826e-01 3.88961852e-01 -6.31966889e-01
-4.51379061e-01 9.30911958e-01 5.67234643e-02 -9.14254308e-01
-1.69769719e-01 -3.13510716e-01 -2.59634614e-01 2.51464248e-01
-1.41796112e-01 4.21377242e-01 -2.19985887e-01 -5.61923265e-01
-5.43810308e-01 1.90576553e-01 -8.99821967e-02 4.62277919e-01
1.55322564e+00 1.41728610e-01 -8.70757520e-01 5.79645813e-01
1.45819676e+00 -1.96383581e-01 -6.32351339e-01 -5.81199884e-01
7.84425065e-02 -4.34118770e-02 -1.72908902e-01 -5.57086408e-01
-1.16348875e+00 6.55764878e-01 -2.11620852e-01 6.06756806e-01
5.61511993e-01 6.44515276e-01 4.77842838e-01 3.32091212e-01
3.51532176e-02 -5.21305025e-01 1.09051242e-01 4.68292356e-01
5.49242795e-01 -1.14329278e+00 5.71965128e-02 -7.47414529e-01
-6.41357899e-02 1.06206119e+00 6.03515983e-01 -1.42694443e-01
1.12006128e+00 1.50961295e-01 -5.03289044e-01 -7.56509185e-01
-1.23574829e+00 -1.53275087e-01 2.63277113e-01 6.78344905e-01
5.69792747e-01 2.46216282e-01 -3.19043025e-02 5.62340736e-01
1.13266982e-01 -4.32834923e-01 3.19403946e-01 5.08038282e-01
2.67178528e-02 -1.27520955e+00 3.65417480e-01 1.00510263e+00
-4.04993951e-01 -5.18252730e-01 -8.30228031e-01 9.35051501e-01
5.45814820e-02 6.30874813e-01 1.91546097e-01 -6.17435396e-01
-3.05511001e-02 -1.52030304e-01 1.99086756e-01 -8.06942344e-01
-7.04757214e-01 -3.50733519e-01 3.69932413e-01 -5.06581187e-01
-3.36078763e-01 -3.65010351e-01 -1.16109943e+00 -7.75755882e-01
-3.86276692e-01 -8.85420814e-02 1.31028131e-01 5.99974811e-01
7.89818406e-01 7.57012963e-01 4.65223730e-01 -4.81742889e-01
-7.44317323e-02 -9.74418581e-01 -8.90789807e-01 5.13562143e-01
4.26543504e-01 -6.67741299e-01 -4.55084115e-01 -5.36889255e-01] | [7.076934337615967, 6.1961798667907715] |
27b608c7-d6db-40b1-bae2-95e1638a5044 | raven-s-progressive-matrices-completion-with | 2103.12045 | null | https://arxiv.org/abs/2103.12045v2 | https://arxiv.org/pdf/2103.12045v2.pdf | Raven's Progressive Matrices Completion with Latent Gaussian Process Priors | Abstract reasoning ability is fundamental to human intelligence. It enables humans to uncover relations among abstract concepts and further deduce implicit rules from the relations. As a well-known abstract visual reasoning task, Raven's Progressive Matrices (RPM) are widely used in human IQ tests. Although extensive research has been conducted on RPM solvers with machine intelligence, few studies have considered further advancing the standard answer-selection (classification) problem to a more challenging answer-painting (generating) problem, which can verify whether the model has indeed understood the implicit rules. In this paper we aim to solve the latter one by proposing a deep latent variable model, in which multiple Gaussian processes are employed as priors of latent variables to separately learn underlying abstract concepts from RPMs; thus the proposed model is interpretable in terms of concept-specific latent variables. The latent Gaussian process also provides an effective way of extrapolation for answer painting based on the learned concept-changing rules. We evaluate the proposed model on RPM-like datasets with multiple continuously-changing visual concepts. Experimental results demonstrate that our model requires only few training samples to paint high-quality answers, generate novel RPM panels, and achieve interpretability through concept-specific latent variables. | ['xiangyang xue', 'Bin Li', 'Fan Shi'] | 2021-03-22 | null | null | null | null | ['answer-selection'] | ['natural-language-processing'] | [ 2.43543819e-01 2.23180488e-01 1.03557728e-01 -3.45106155e-01
-4.55754846e-02 -4.05694932e-01 6.86808109e-01 -3.43641527e-02
-4.16833311e-02 7.16721654e-01 -7.59736896e-02 -3.94065738e-01
-5.37390590e-01 -9.74633217e-01 -5.53080916e-01 -5.47273993e-01
2.61321276e-01 1.08837306e+00 6.47630543e-02 -6.62591830e-02
6.32861257e-01 2.39712343e-01 -1.75927246e+00 4.11782533e-01
1.56262159e+00 7.06829906e-01 2.82371014e-01 6.06183887e-01
-6.11667275e-01 1.09042394e+00 -9.26601052e-01 -9.10992742e-01
-4.66338322e-02 -4.64582980e-01 -8.43553782e-01 -5.34109175e-02
1.92679480e-01 -2.04773903e-01 1.51974380e-01 1.21640766e+00
-4.84565087e-02 3.17751378e-01 1.05221617e+00 -1.32994509e+00
-1.49672520e+00 4.71739084e-01 -7.03087986e-01 9.85039324e-02
4.19512480e-01 3.75831127e-01 1.05748034e+00 -8.03486347e-01
2.71660715e-01 1.73992300e+00 2.91517079e-01 4.84064430e-01
-1.54929829e+00 -7.98741341e-01 3.94591719e-01 8.51843715e-01
-1.06471705e+00 1.88384414e-01 9.89792943e-01 -6.25164747e-01
5.64104855e-01 4.32012081e-01 8.92434895e-01 1.08578646e+00
1.46236718e-01 9.55562711e-01 1.54230738e+00 -3.22667718e-01
3.66509467e-01 1.63799569e-01 4.33737278e-01 8.19032133e-01
3.15551698e-01 -1.94540605e-01 -7.01241732e-01 1.06718197e-01
1.07576370e+00 9.67587009e-02 -3.48378867e-01 -4.08593386e-01
-1.16022182e+00 9.36130524e-01 4.39284295e-01 -5.29017486e-02
-3.40388775e-01 3.49444151e-02 6.96116388e-02 4.06469941e-01
2.05875441e-01 5.19126892e-01 -1.37696132e-01 -4.65689078e-02
-7.10438371e-01 3.81215274e-01 5.12477398e-01 8.64755154e-01
6.93020642e-01 9.82961059e-02 -6.02915287e-01 6.92376852e-01
5.51367640e-01 5.29037952e-01 4.99918848e-01 -1.07068682e+00
6.19077563e-01 9.30256605e-01 2.78491318e-01 -1.36229253e+00
-1.90795451e-01 -6.47684693e-01 -8.32783163e-01 5.65995216e-01
3.95246565e-01 4.04216528e-01 -8.42284799e-01 1.61843276e+00
2.10363656e-01 7.86944404e-02 -3.77202965e-02 1.12811434e+00
7.77603388e-01 6.91378176e-01 2.66120821e-01 -1.75846055e-01
1.62223840e+00 -1.06017399e+00 -1.06631947e+00 -1.59692347e-01
5.11532687e-02 -2.54016191e-01 1.73420322e+00 8.73433709e-01
-1.09265745e+00 -1.04589057e+00 -1.24270713e+00 -4.14415717e-01
-2.72004664e-01 8.21631849e-02 8.39480042e-01 6.66632056e-01
-7.54096687e-01 3.35502148e-01 -4.77746218e-01 1.48484066e-01
6.09188974e-01 1.77163363e-01 -2.60452718e-01 -1.41892686e-01
-1.23275959e+00 8.60174179e-01 4.88357902e-01 2.04351723e-01
-5.80479562e-01 -5.99203825e-01 -5.30061185e-01 3.99163604e-01
5.10559738e-01 -1.29345679e+00 9.95462120e-01 -1.03747010e+00
-1.68449426e+00 8.66379797e-01 -1.77361637e-01 -2.61742830e-01
5.63342690e-01 -4.21456248e-01 -2.45534241e-01 4.71769780e-01
4.31041904e-02 5.54753304e-01 1.34390509e+00 -1.37028861e+00
-2.40482181e-01 -4.34992164e-01 2.77677804e-01 3.28544170e-01
-2.29005292e-01 -2.92064369e-01 -2.27593318e-01 -7.16047049e-01
2.67163962e-01 -4.41985458e-01 2.90262938e-01 1.95278570e-01
-2.20563143e-01 -6.44750953e-01 5.22970498e-01 -8.60074401e-01
1.02466941e+00 -1.78581595e+00 6.03953719e-01 4.84687202e-02
6.42282367e-01 8.00441951e-02 -8.21000338e-03 -5.37074059e-02
-1.60062462e-02 1.63571015e-02 -1.35216326e-01 -2.82156438e-01
4.65172142e-01 2.44415313e-01 -7.81466901e-01 -6.57186564e-03
2.12691024e-01 1.22502673e+00 -9.72799242e-01 -4.50486302e-01
5.05597368e-02 1.38829350e-01 -5.93659222e-01 4.09558177e-01
-5.69078863e-01 5.00989199e-01 -3.23044389e-01 2.62948424e-01
7.44877458e-01 -3.99395585e-01 8.67857262e-02 1.11529134e-01
4.53522474e-01 1.45907942e-02 -1.13162172e+00 1.54970443e+00
-2.21788332e-01 5.85551560e-01 -4.15289611e-01 -9.79258537e-01
1.34219337e+00 4.41059237e-03 -5.38901985e-01 -7.35917330e-01
-2.25634038e-01 -1.61991015e-01 8.03872496e-02 -8.28239322e-01
2.61551499e-01 -5.12786090e-01 1.78523660e-01 2.57692546e-01
-4.31919098e-02 -2.80535489e-01 2.10523233e-01 1.69144437e-01
6.99834645e-01 5.37524521e-01 1.49758145e-01 -2.10051522e-01
7.21665621e-01 -2.45332807e-01 5.16876101e-01 8.03121507e-01
8.01073760e-02 3.75015497e-01 7.99534559e-01 -5.40254176e-01
-6.59739256e-01 -1.67010581e+00 1.52805075e-01 6.99406087e-01
2.95426488e-01 -2.04959258e-01 -5.57261884e-01 -5.37885308e-01
-1.42153865e-03 1.31783235e+00 -7.69342005e-01 -2.84992963e-01
-3.69137704e-01 -4.68413621e-01 1.55058965e-01 5.82587004e-01
8.16604197e-01 -1.26348293e+00 -6.03193700e-01 -1.51670754e-01
-2.67456383e-01 -5.89830101e-01 2.09876090e-01 -2.70544499e-01
-9.46946263e-01 -1.10071325e+00 -5.44592500e-01 -4.29320961e-01
8.23098838e-01 -7.14751929e-02 1.28366709e+00 1.54110894e-01
-1.32154524e-01 4.06812310e-01 -1.85710639e-01 -3.97377610e-01
-3.11280526e-02 -3.70935589e-01 -9.32776704e-02 1.01199977e-01
7.05107391e-01 -7.49733210e-01 -6.77348137e-01 5.58774471e-02
-9.63155150e-01 6.62073135e-01 6.25017107e-01 9.10530567e-01
3.88659209e-01 2.25265726e-01 5.15501857e-01 -9.24559653e-01
9.90445197e-01 -4.83752251e-01 -4.93690848e-01 4.07328129e-01
-5.33246517e-01 3.28995556e-01 5.99729180e-01 -5.24537385e-01
-1.54946816e+00 -6.69719398e-01 2.18470588e-01 -5.60383379e-01
-2.71109790e-01 4.89698887e-01 -3.32520097e-01 4.83167291e-01
5.74398577e-01 3.83150786e-01 -1.31309286e-01 -4.14737374e-01
7.53593206e-01 1.96427763e-01 5.23717523e-01 -1.14091396e+00
1.04521549e+00 3.96867573e-01 1.87032610e-01 -3.38635772e-01
-9.39409435e-01 3.10955048e-02 -4.84851182e-01 -1.34154588e-01
1.17609727e+00 -6.19254529e-01 -1.15006793e+00 2.43984774e-01
-1.26850808e+00 -2.23905131e-01 -2.36161858e-01 3.12744379e-01
-7.72417784e-01 4.95790660e-01 -7.46712863e-01 -9.52528834e-01
-4.42429818e-02 -1.07900834e+00 8.00503492e-01 5.58647096e-01
-5.20988584e-01 -1.11568403e+00 3.86962015e-03 8.76832426e-01
4.84236106e-02 1.51348382e-01 1.60707581e+00 -2.51478374e-01
-1.03726530e+00 2.12995544e-01 -3.55253726e-01 3.21308225e-02
-2.25074664e-01 -1.67143568e-01 -9.09519315e-01 2.83554327e-02
4.74019438e-01 -4.61247683e-01 1.00170958e+00 -6.01160107e-04
1.57185793e+00 -2.98382968e-01 2.67751794e-02 4.80896115e-01
1.11930716e+00 2.20851034e-01 1.01042092e+00 3.55855435e-01
6.69885159e-01 8.21192801e-01 5.79002857e-01 7.80519843e-02
4.09213573e-01 5.39981186e-01 2.79147625e-01 9.95853171e-02
6.70006126e-02 -5.02708197e-01 1.61127046e-01 8.66351187e-01
-3.83459955e-01 2.35690206e-01 -8.60578001e-01 1.71991915e-01
-1.85670006e+00 -1.10535634e+00 -3.94068182e-01 1.85817528e+00
1.08960438e+00 3.19153219e-01 -2.97508091e-01 3.63465309e-01
5.54185867e-01 -9.77134407e-02 -5.47534347e-01 -4.20935482e-01
3.97273339e-02 5.11620820e-01 -5.65410435e-01 5.51687062e-01
-7.02723861e-01 7.14450300e-01 5.25936842e+00 9.91676271e-01
-4.95681316e-01 1.88354868e-02 6.33409262e-01 1.67891793e-02
-7.46305585e-01 5.06728068e-02 -4.58082587e-01 3.41799736e-01
3.57643217e-01 -8.71786699e-02 4.44924533e-01 6.85984492e-01
-8.97574611e-03 -1.29116639e-01 -1.26234257e+00 1.31397486e+00
2.15935364e-01 -1.03110290e+00 6.74968779e-01 -2.47273967e-01
8.07410419e-01 -1.32751000e+00 5.08933306e-01 8.28346610e-01
2.72925019e-01 -1.39962006e+00 8.68007541e-01 1.32841122e+00
6.51886761e-01 -7.22621322e-01 4.03147578e-01 4.67591077e-01
-8.65203738e-01 -2.35289663e-01 -5.52060187e-01 -6.34042621e-01
-2.19757229e-01 5.76317668e-01 -3.92855108e-01 5.36543608e-01
5.86756289e-01 4.36725348e-01 -8.49673390e-01 7.66576350e-01
-1.08462071e+00 4.83422130e-01 3.22964102e-01 -9.34884474e-02
-2.95051157e-01 -5.94136834e-01 4.29082721e-01 5.21089196e-01
1.83902770e-01 2.60353982e-01 -1.92149401e-01 1.81117368e+00
1.45348370e-01 -1.08464882e-02 -1.84207335e-01 2.05980632e-02
3.44339639e-01 1.10006535e+00 -9.76543069e-01 -4.95546520e-01
-1.74378291e-01 1.18595648e+00 7.04729974e-01 8.35826159e-01
-8.54890704e-01 -2.83619851e-01 4.40602183e-01 -6.87472224e-02
2.70620376e-01 -3.35949540e-01 -7.29142547e-01 -1.36847198e+00
1.22212186e-01 -9.56362367e-01 3.13089848e-01 -1.48835194e+00
-1.56704295e+00 1.20950803e-01 4.87936176e-02 -8.50767255e-01
-4.59450446e-02 -9.29007947e-01 -6.17485464e-01 1.35708499e+00
-1.34156704e+00 -9.60148573e-01 -5.20166636e-01 5.80076754e-01
6.47230864e-01 -1.62016183e-01 8.54174078e-01 -3.19347978e-01
-3.49664003e-01 3.69862199e-01 -2.47981668e-01 -2.55468469e-02
4.28003460e-01 -1.66566765e+00 2.08422497e-01 6.81033373e-01
3.29275042e-01 1.19201899e+00 8.22755575e-01 -8.48466814e-01
-1.06495678e+00 -4.07891929e-01 3.96217048e-01 -7.77210236e-01
5.08293450e-01 -4.25426513e-01 -1.26611471e+00 5.84724247e-01
1.89784735e-01 -5.67179143e-01 7.89637148e-01 3.45114946e-01
-5.86896479e-01 -1.11939773e-01 -8.68972838e-01 8.96851242e-01
9.68412220e-01 -5.89387715e-01 -1.34217751e+00 2.03213878e-02
8.08851719e-01 -1.35342732e-01 -5.01081347e-01 2.28331193e-01
2.84419656e-01 -1.27537513e+00 9.67754543e-01 -6.98605120e-01
8.97677243e-01 -5.66610515e-01 3.61066818e-01 -1.33435822e+00
-6.33492887e-01 -2.22635716e-01 -6.76828921e-01 1.11722875e+00
4.29201825e-03 -6.91514313e-01 7.82683194e-01 6.72237933e-01
6.50603995e-02 -7.99655139e-01 -6.46798134e-01 -5.91595888e-01
9.35412027e-05 -3.21666867e-01 7.45352983e-01 1.04515529e+00
-1.05326504e-01 5.43489456e-01 -1.00273758e-01 4.48903292e-02
7.95993447e-01 4.71952379e-01 8.65603268e-01 -1.64182150e+00
-7.30549693e-01 -6.70445442e-01 -3.90031368e-01 -1.32500803e+00
1.77680030e-01 -8.02248120e-01 -4.08499658e-01 -1.79719043e+00
3.89544785e-01 -2.21930414e-01 -1.94340691e-01 1.47317186e-01
-9.40707147e-01 -7.06390738e-02 3.37721020e-01 2.54449606e-01
-6.62445426e-01 9.35452163e-01 1.62641084e+00 -2.66329110e-01
1.43046841e-01 -1.09569140e-01 -7.69447029e-01 7.30402648e-01
7.34419703e-01 -1.05040856e-01 -1.11191821e+00 -3.27293187e-01
7.98100948e-01 -3.13633005e-03 6.25527501e-01 -9.37225223e-01
1.11748399e-02 -2.86700726e-01 9.34800923e-01 -5.80019057e-01
4.64713693e-01 -7.14081049e-01 8.27774405e-02 3.97734642e-01
-2.74505764e-01 1.59266368e-01 1.51945755e-01 7.95262873e-01
-1.25616044e-01 -3.90932381e-01 3.40487778e-01 -2.24878237e-01
-7.96414852e-01 -8.95644277e-02 -1.52699858e-01 1.98682249e-01
9.67805147e-01 -4.13254112e-01 -5.88359416e-01 -3.41384888e-01
-8.75152826e-01 2.38628015e-01 1.40245795e-01 4.11189705e-01
9.96071935e-01 -1.48957443e+00 -5.96923828e-01 1.34942740e-01
-9.43893008e-03 1.21786110e-01 5.41742504e-01 5.70379853e-01
-7.78999448e-01 3.22620571e-01 -4.74776268e-01 -6.91448510e-01
-9.86045063e-01 1.03331685e+00 1.26282573e-01 -4.39048767e-01
-6.36310220e-01 1.03312898e+00 6.98822379e-01 -3.51134419e-01
1.87131956e-01 -2.99382508e-01 -5.94987810e-01 -4.91028513e-05
6.29200280e-01 2.92204320e-01 -4.56517756e-01 -3.04357987e-02
2.67105252e-01 4.47711110e-01 -1.20367706e-02 -1.21450350e-01
1.04681551e+00 -3.11811306e-02 -3.44249606e-01 9.21607554e-01
4.17915881e-01 -8.56796205e-02 -1.41439497e+00 -9.04573724e-02
8.29314291e-02 -7.06427336e-01 -5.31243443e-01 -8.19955647e-01
-5.11903942e-01 1.44516110e+00 3.12326699e-01 1.56644210e-02
1.11340320e+00 -2.83295363e-01 2.27516696e-01 3.67624104e-01
3.97518754e-01 -8.80708694e-01 7.15508878e-01 2.37043679e-01
1.23231745e+00 -8.84653568e-01 -1.98042579e-02 -5.28250694e-01
-6.03807211e-01 1.31205785e+00 8.99657845e-01 -6.18869513e-02
1.02418467e-01 -3.50325644e-01 -1.09662861e-01 -3.86676013e-01
-8.09232712e-01 7.07294270e-02 4.94018435e-01 8.27328801e-01
1.41724810e-01 2.12637588e-01 -1.64027363e-01 1.07745206e+00
-5.15048444e-01 -3.56814638e-02 1.73500985e-01 3.89928669e-01
-4.71439481e-01 -7.78894186e-01 -7.49982715e-01 4.62458193e-01
1.33240089e-01 -1.13154404e-01 -1.27308428e-01 4.97556210e-01
3.05827111e-01 6.59470737e-01 3.04277036e-02 2.86051049e-03
2.00337946e-01 4.66445684e-01 8.22934985e-01 -6.65864825e-01
-1.49755329e-01 -6.33186638e-01 -2.00124457e-01 -3.04558665e-01
-2.60370016e-01 -4.55747247e-01 -1.29945850e+00 -2.78145760e-01
1.16233364e-01 2.67621279e-01 2.17722535e-01 1.14518976e+00
-6.08327724e-02 8.87020826e-01 -2.19953448e-01 -5.00423968e-01
-7.23870814e-01 -8.73813152e-01 -5.50059438e-01 8.83269787e-01
-1.76418871e-01 -1.14523315e+00 -2.39366740e-01 1.81175426e-01] | [10.625925064086914, 2.004633665084839] |
aaa00da2-9aaa-41c4-b6df-0a4dcd19e9ee | deep-multi-view-subspace-clustering-with | 2305.06939 | null | https://arxiv.org/abs/2305.06939v1 | https://arxiv.org/pdf/2305.06939v1.pdf | Deep Multi-View Subspace Clustering with Anchor Graph | Deep multi-view subspace clustering (DMVSC) has recently attracted increasing attention due to its promising performance. However, existing DMVSC methods still have two issues: (1) they mainly focus on using autoencoders to nonlinearly embed the data, while the embedding may be suboptimal for clustering because the clustering objective is rarely considered in autoencoders, and (2) existing methods typically have a quadratic or even cubic complexity, which makes it challenging to deal with large-scale data. To address these issues, in this paper we propose a novel deep multi-view subspace clustering method with anchor graph (DMCAG). To be specific, DMCAG firstly learns the embedded features for each view independently, which are used to obtain the subspace representations. To significantly reduce the complexity, we construct an anchor graph with small size for each view. Then, spectral clustering is performed on an integrated anchor graph to obtain pseudo-labels. To overcome the negative impact caused by suboptimal embedded features, we use pseudo-labels to refine the embedding process to make it more suitable for the clustering task. Pseudo-labels and embedded features are updated alternately. Furthermore, we design a strategy to keep the consistency of the labels based on contrastive learning to enhance the clustering performance. Empirical studies on real-world datasets show that our method achieves superior clustering performance over other state-of-the-art methods. | ['Lifang He', 'Xiaorong Pu', 'Jingyu Pu', 'Yazhou Ren', 'Chenhang Cui'] | 2023-05-11 | null | null | null | null | ['multi-view-subspace-clustering'] | ['computer-vision'] | [-3.41180950e-01 -4.28473264e-01 6.74980953e-02 -1.98797598e-01
-4.97023493e-01 -5.07112503e-01 3.93541157e-01 -1.74114943e-01
-1.75379783e-01 7.64182806e-02 3.53325337e-01 2.15036750e-01
-2.51812458e-01 -5.29982984e-01 -3.94024104e-01 -1.13684869e+00
1.29578978e-01 2.56173790e-01 2.95666791e-02 1.67502448e-01
3.99346687e-02 1.77245960e-01 -1.44466102e+00 1.01808421e-01
1.06066239e+00 7.05864608e-01 3.19729656e-01 -2.21025012e-02
-1.34338781e-01 2.96344876e-01 -3.11589807e-01 -1.08427443e-01
2.81828523e-01 -5.17446756e-01 -2.97484696e-01 5.09703815e-01
-5.17094247e-02 -1.87903330e-01 -3.90262783e-01 1.22970104e+00
5.08237123e-01 1.90822199e-01 7.00517058e-01 -1.30773723e+00
-6.16804063e-01 4.90219980e-01 -7.96814203e-01 -2.04947636e-01
-5.14239185e-02 -5.79432286e-02 1.01843143e+00 -9.97112513e-01
4.20673221e-01 1.16327095e+00 4.69614506e-01 4.72851783e-01
-1.32959390e+00 -7.24704742e-01 3.87977660e-01 4.49972183e-01
-1.73887694e+00 -2.92177349e-01 1.26116025e+00 -5.14657795e-01
3.06046844e-01 4.67940532e-02 7.49737084e-01 8.67996097e-01
-2.49290630e-01 7.72835076e-01 1.02328444e+00 -1.41984776e-01
3.63289595e-01 1.64305389e-01 -6.60586059e-02 5.91657698e-01
-1.62156355e-02 -2.91009307e-01 -5.18680252e-02 -1.30186304e-01
6.89482987e-01 4.74295467e-01 -4.35817420e-01 -7.68285751e-01
-1.35242677e+00 1.09956992e+00 5.99460781e-01 5.02974927e-01
-3.60570610e-01 -2.25340381e-01 4.58185405e-01 -5.34813888e-02
3.29833508e-01 1.01207420e-01 -1.66538894e-01 2.11716101e-01
-7.60475159e-01 -1.19638622e-01 3.86554062e-01 7.81406283e-01
9.41840053e-01 -1.86832398e-02 2.43838280e-01 1.14323413e+00
4.28353071e-01 1.94146857e-01 6.78307891e-01 -1.01480043e+00
5.11197507e-01 1.02433097e+00 -1.53121918e-01 -1.40340471e+00
-4.38871056e-01 -4.81406361e-01 -1.35723031e+00 -1.06398590e-01
6.31884858e-02 5.79464920e-02 -6.32525802e-01 1.61367595e+00
4.24796969e-01 2.99091280e-01 1.72130719e-01 1.20033443e+00
6.95699096e-01 8.09574544e-01 -2.30514601e-01 -3.24811012e-01
1.18178952e+00 -1.05889785e+00 -7.04053581e-01 1.93202645e-02
6.21424198e-01 -5.64466178e-01 1.05914879e+00 3.10463876e-01
-5.39228082e-01 -6.45813406e-01 -1.17088616e+00 2.33937725e-01
-2.48966485e-01 4.43054855e-01 3.23117405e-01 3.28632563e-01
-8.65101159e-01 3.01741481e-01 -1.00572884e+00 -1.95297316e-01
1.73492357e-01 3.52806628e-01 -4.35915500e-01 -2.71435738e-01
-9.95251000e-01 2.76126713e-01 6.29331112e-01 1.09105147e-01
-5.37068307e-01 -2.24679038e-01 -8.92897964e-01 1.19046293e-01
4.75687355e-01 -3.58767152e-01 4.89215374e-01 -8.37956667e-01
-1.33112621e+00 4.42373753e-01 -1.78886965e-01 1.30660415e-01
1.94900036e-01 5.30010164e-02 -5.57154477e-01 2.69055128e-01
2.40601480e-01 6.39572084e-01 9.04391348e-01 -1.57988000e+00
-4.50131267e-01 -5.46232641e-01 -1.13563471e-01 4.61126119e-01
-8.98011148e-01 -3.71201605e-01 -1.06211030e+00 -7.59686172e-01
6.75518274e-01 -1.06087410e+00 -3.16039085e-01 -3.82307112e-01
-3.75999451e-01 -2.91179538e-01 1.20373738e+00 -5.67156911e-01
1.47197306e+00 -2.60682702e+00 6.65711761e-01 1.99226439e-01
3.81104946e-01 8.42210650e-02 -1.19678594e-01 3.22923869e-01
-1.11194275e-01 -3.28032412e-02 -3.41478199e-01 -3.63767087e-01
-1.71101242e-01 2.40402877e-01 1.52943283e-01 5.83546221e-01
-1.29951403e-01 5.34922779e-01 -7.06286073e-01 -7.35459805e-01
4.28146958e-01 4.79738891e-01 -6.54104769e-01 2.31575668e-01
1.43619075e-01 5.18985927e-01 -4.76658642e-01 3.68421108e-01
8.01464736e-01 -6.27048075e-01 2.22837880e-01 -5.31075776e-01
-6.51797056e-02 -4.27813560e-01 -1.59907687e+00 1.66720760e+00
-2.70257592e-01 1.10943861e-01 2.95968354e-01 -1.31855881e+00
7.60869086e-01 2.86502630e-01 7.69264758e-01 -1.26714826e-01
1.68488577e-01 1.80183381e-01 -1.65552106e-02 -2.73480296e-01
4.63592857e-02 -1.66983068e-01 1.17992811e-01 3.37893128e-01
-3.77876423e-02 2.90847927e-01 5.72375171e-02 3.02741617e-01
7.34077096e-01 -1.97259605e-01 1.90940619e-01 -1.70194700e-01
9.65427518e-01 -1.61670223e-01 9.39858377e-01 -8.18475932e-02
-2.14179441e-01 7.33034372e-01 3.06453109e-01 -3.16508800e-01
-8.52902412e-01 -8.78071725e-01 1.57175362e-02 6.44964159e-01
4.34768915e-01 -6.21494532e-01 -9.62098062e-01 -7.95469403e-01
-2.02183172e-01 3.44047695e-01 -4.61584687e-01 -2.30879158e-01
-5.44683635e-01 -8.26673985e-01 6.40279502e-02 6.53443456e-01
6.69943511e-01 -6.77995801e-01 -1.50255367e-01 2.88952202e-01
-3.91706467e-01 -1.06829095e+00 -7.34882236e-01 -3.51544209e-02
-9.03876603e-01 -1.06999922e+00 -8.09216976e-01 -9.33110893e-01
9.70886230e-01 7.14518964e-01 3.78994316e-01 1.49995208e-01
1.66426435e-01 2.45468244e-01 -5.86305141e-01 2.22690448e-01
-1.80149257e-01 1.61688998e-01 3.25630397e-01 3.90967667e-01
4.99088079e-01 -7.15995669e-01 -7.74429858e-01 5.01788318e-01
-1.17050076e+00 2.35578269e-01 6.07874334e-01 1.01606333e+00
6.99685872e-01 5.09563029e-01 4.95606899e-01 -6.50596440e-01
4.17545527e-01 -4.46291775e-01 -6.12575233e-01 1.41570553e-01
-7.60606408e-01 -6.37189597e-02 1.17038977e+00 -4.45487171e-01
-7.53919423e-01 4.05516773e-01 1.10237233e-01 -1.07651591e+00
-1.12292878e-01 6.60067976e-01 -5.33841372e-01 1.25773907e-01
1.66395172e-01 4.41348970e-01 1.87357947e-01 -4.97120887e-01
4.52137917e-01 7.90475070e-01 2.34015271e-01 -2.10398540e-01
1.01285076e+00 5.77709854e-01 -1.24958545e-01 -6.56507671e-01
-6.23605311e-01 -6.89526498e-01 -8.31383824e-01 -1.67541638e-01
1.04933155e+00 -1.15900612e+00 -5.25662363e-01 3.76573175e-01
-8.14014554e-01 7.63967410e-02 3.47704798e-01 7.51224935e-01
-2.90187627e-01 7.75984704e-01 -6.56794310e-01 -4.83959526e-01
-1.25554189e-01 -1.35154378e+00 8.72180402e-01 2.57990688e-01
2.43126273e-01 -1.01891983e+00 -1.24016546e-01 3.67651314e-01
6.63963854e-02 1.00589655e-01 1.04268646e+00 -5.07664204e-01
-5.01335680e-01 -4.50256541e-02 -2.72376806e-01 4.47117299e-01
3.57378691e-01 -1.63337648e-01 -7.13701725e-01 -6.21528506e-01
1.16433091e-01 -8.48952234e-02 7.36492217e-01 3.41678143e-01
1.47088253e+00 -2.33661264e-01 -4.53335047e-01 7.35252678e-01
1.43817794e+00 3.12543601e-01 2.53814846e-01 3.20259064e-01
1.06026697e+00 5.98676980e-01 4.95340466e-01 4.90449637e-01
5.45116723e-01 6.33734465e-01 3.06966871e-01 1.58037189e-02
2.48873815e-01 -2.63512582e-01 2.71836758e-01 1.71863830e+00
3.38349156e-02 4.23859619e-03 -8.00079703e-01 5.35855532e-01
-2.02949858e+00 -7.17334449e-01 2.50002872e-02 2.04277778e+00
3.90526295e-01 -1.12984329e-01 1.44244328e-01 2.42384508e-01
1.01620626e+00 3.35994691e-01 -6.72823012e-01 2.64564872e-01
-6.53192252e-02 -5.23099542e-01 8.17255601e-02 7.26560578e-02
-1.33536482e+00 8.08445871e-01 4.95035696e+00 9.68127012e-01
-1.14890611e+00 1.60290778e-01 5.09801507e-01 4.81403619e-02
-2.99864501e-01 6.09054975e-02 -5.32738805e-01 6.74678683e-01
4.46650952e-01 1.43525943e-01 7.01006413e-01 9.03578401e-01
1.03087373e-01 3.03071886e-01 -8.69278729e-01 1.39935076e+00
2.47878924e-01 -1.01764762e+00 1.03483796e-01 1.92545265e-01
8.48211706e-01 -2.01522961e-01 -6.80185389e-03 3.70699197e-01
1.04905859e-01 -6.08387291e-01 3.20008248e-01 1.92606509e-01
6.51944637e-01 -9.57174122e-01 6.60270810e-01 5.64908624e-01
-1.45841324e+00 -2.36060008e-01 -5.46713173e-01 2.82246768e-01
2.26808954e-02 5.93110740e-01 -2.56766677e-01 7.57913411e-01
7.08199561e-01 1.09928954e+00 -6.74099207e-01 7.15029538e-01
4.18867636e-03 5.12304008e-01 -3.02428067e-01 8.78259391e-02
4.10566598e-01 -8.29307258e-01 3.80783886e-01 7.45725274e-01
4.45290267e-01 2.15921804e-01 4.97938633e-01 7.93130159e-01
3.22241932e-02 3.57693195e-01 -5.21738470e-01 -2.71843169e-02
6.80792868e-01 1.49060380e+00 -8.58652055e-01 -2.98525453e-01
-7.38792717e-01 1.20936489e+00 3.68554920e-01 5.74250460e-01
-7.65476227e-01 -2.50633299e-01 2.98502415e-01 -9.28748250e-02
4.36381370e-01 -3.25083494e-01 -2.96202116e-02 -1.54215026e+00
1.77079201e-01 -9.24180031e-01 4.77447301e-01 -4.11273986e-01
-1.33411634e+00 4.89488602e-01 -9.88917202e-02 -1.69106710e+00
-7.70365000e-02 -3.26185077e-01 -5.30577302e-01 5.20472527e-01
-1.04486644e+00 -1.09671915e+00 -5.87855160e-01 7.19194114e-01
4.71145689e-01 -3.89297336e-01 6.75155878e-01 5.28685331e-01
-9.51426268e-01 4.69581127e-01 6.06228948e-01 2.17632607e-01
6.89451277e-01 -1.18351161e+00 -1.20617241e-01 7.32231677e-01
1.43864229e-01 7.19976485e-01 2.55774319e-01 -4.53578949e-01
-1.50079715e+00 -1.28227103e+00 1.39273182e-01 -5.63974716e-02
4.96608287e-01 -4.11782712e-01 -1.12378752e+00 5.03279567e-01
3.59290056e-02 2.77655032e-02 9.00498271e-01 1.08929120e-01
-3.61275077e-01 -2.70595342e-01 -7.69738853e-01 6.88505352e-01
8.70176435e-01 -5.40026128e-01 -3.53390127e-01 1.55699089e-01
7.55936205e-01 -1.39831733e-02 -9.99585629e-01 5.25785625e-01
2.84658521e-01 -1.00434136e+00 8.12302411e-01 -3.97553407e-02
2.99208015e-01 -7.99707711e-01 -3.14092427e-01 -1.58071947e+00
-6.80942833e-01 -1.04054123e-01 -1.62996650e-01 1.46709299e+00
-1.83984917e-02 -6.32236302e-01 8.51622701e-01 2.56130457e-01
-1.39712870e-01 -8.72849226e-01 -8.34979117e-01 -7.45893538e-01
-1.35140091e-01 6.17317185e-02 6.04556203e-01 1.29488111e+00
-1.02281034e-01 5.73340833e-01 -3.30946952e-01 2.72503465e-01
7.10511565e-01 4.22381818e-01 6.92610323e-01 -1.27701414e+00
-2.74681568e-01 -3.19714934e-01 -3.68996710e-01 -9.33411062e-01
2.51614600e-01 -9.09566581e-01 -1.23076655e-01 -1.45079517e+00
4.98413026e-01 -4.74700391e-01 -4.16921407e-01 1.94696128e-01
-4.14225429e-01 1.05099581e-01 1.79784909e-01 6.49860799e-01
-7.06976295e-01 1.03326213e+00 1.14714396e+00 -5.32000400e-02
-3.78804058e-01 -2.73892373e-01 -6.90056145e-01 7.30072081e-01
6.54543281e-01 -1.71935439e-01 -5.17860174e-01 -3.72038335e-01
-1.76374912e-01 4.42543775e-02 1.27951935e-01 -1.06668961e+00
3.51109624e-01 2.71281358e-02 4.90941018e-01 -8.20636630e-01
3.09592426e-01 -9.78770912e-01 2.71229267e-01 3.92825067e-01
6.16199411e-02 3.26905549e-02 -1.98314324e-01 8.77571464e-01
-5.53888798e-01 -7.18434602e-02 7.88746119e-01 -7.82676265e-02
-6.18041694e-01 4.37533081e-01 -2.10499153e-01 -2.35342085e-01
1.10988319e+00 -9.02814493e-02 6.01665489e-02 -4.60204810e-01
-6.07361019e-01 5.71571946e-01 8.13550174e-01 3.96396905e-01
6.35915458e-01 -1.74790013e+00 -4.22328919e-01 2.52856433e-01
2.86045104e-01 2.78200448e-01 5.28018355e-01 9.12342727e-01
-3.40533197e-01 3.03130925e-01 2.37169079e-02 -7.62980819e-01
-1.14100075e+00 1.10485017e+00 -5.83438724e-02 -7.23520070e-02
-7.43568242e-01 3.66228163e-01 5.45979977e-01 -5.90564132e-01
2.24366784e-01 2.41385162e-01 -4.60769862e-01 1.73439965e-01
3.26927781e-01 3.47832471e-01 -2.01019153e-01 -8.09025109e-01
-3.55479479e-01 9.18857157e-01 -1.48142233e-01 6.11936487e-02
1.34252834e+00 -3.09286445e-01 -2.39026293e-01 4.58593577e-01
1.56594026e+00 -1.13648966e-01 -1.23218930e+00 -2.97472954e-01
-1.46762446e-01 -3.30618232e-01 1.72739416e-01 -1.13240376e-01
-1.36915064e+00 9.34976220e-01 6.53711498e-01 9.36402082e-02
1.40630448e+00 -5.62822493e-03 8.65223229e-01 2.60676354e-01
2.02001959e-01 -1.19305193e+00 2.39851922e-01 1.37239233e-01
6.08609855e-01 -1.20164204e+00 2.82824840e-02 -4.71980512e-01
-7.45579422e-01 9.94897544e-01 7.54191816e-01 -1.16613075e-01
7.43858039e-01 -2.42688403e-01 1.45531073e-01 -2.67202497e-01
-3.41223508e-01 -2.88661141e-02 2.49651387e-01 2.52261937e-01
9.68783274e-02 7.76529014e-02 -3.25115979e-01 6.73008323e-01
1.36124507e-01 -4.69040066e-01 2.43864357e-01 5.41529059e-01
-2.11731583e-01 -1.12312210e+00 -4.90601361e-01 4.36412901e-01
-2.11934090e-01 2.54509181e-01 -3.02734524e-01 6.36380076e-01
1.69155061e-01 9.77568686e-01 -1.09887384e-01 -8.51326287e-01
1.13025077e-01 -5.35504706e-02 2.81276666e-02 -4.99568731e-01
-1.11335836e-01 7.14869022e-01 -5.34285665e-01 -4.33280349e-01
-6.46320641e-01 -7.41889179e-01 -1.26626348e+00 -9.77084264e-02
-5.49996078e-01 4.29243177e-01 3.89723420e-01 7.85706460e-01
5.13183355e-01 3.65364373e-01 1.06210470e+00 -7.35223293e-01
-4.11177784e-01 -9.01204884e-01 -8.57313097e-01 6.91481352e-01
-2.64674108e-02 -9.20205474e-01 -5.39530516e-01 -3.13373730e-02] | [8.397245407104492, 4.183085918426514] |
7aaaf7e2-db4c-4175-aea0-bb9c21cf8a6a | verify-and-edit-a-knowledge-enhanced-chain-of | 2305.03268 | null | https://arxiv.org/abs/2305.03268v1 | https://arxiv.org/pdf/2305.03268v1.pdf | Verify-and-Edit: A Knowledge-Enhanced Chain-of-Thought Framework | As large language models (LLMs) have become the norm in NLP, demonstrating good performance in generation and reasoning tasks, one of its most fatal disadvantages is the lack of factual correctness. Generating unfactual texts not only leads to lower performances but also degrades the trust and validity of their applications. Chain-of-Thought (CoT) prompting improves trust and model performance on complex reasoning tasks by generating interpretable reasoning chains, but still suffers from factuality concerns in knowledge-intensive tasks. In this paper, we propose the Verify-and-Edit framework for CoT prompting, which seeks to increase prediction factuality by post-editing reasoning chains according to external knowledge. Building on top of GPT-3, our framework lead to accuracy improvements in multiple open-domain question-answering tasks. | ['Lidong Bing', 'Chengwei Qin', 'Shafiq Joty', 'Xingxuan Li', 'Ruochen Zhao'] | 2023-05-05 | null | null | null | null | ['open-domain-question-answering'] | ['natural-language-processing'] | [ 1.90639514e-02 8.57781589e-01 4.18285616e-02 -4.32012469e-01
-7.50551462e-01 -4.36752141e-01 7.90101051e-01 3.68145257e-01
-1.15410857e-01 9.39368248e-01 2.76314795e-01 -6.91295326e-01
-1.73810542e-01 -9.14840817e-01 -6.95704877e-01 2.41706565e-01
6.12199724e-01 7.30373681e-01 1.86376452e-01 -4.31463867e-01
1.56986058e-01 -4.84523214e-02 -1.20545983e+00 1.04038751e+00
1.43442249e+00 8.06883276e-01 -6.46777079e-02 4.22430754e-01
-3.09308022e-01 1.84719658e+00 -4.94704843e-01 -1.40554738e+00
-3.36252339e-02 -1.79939196e-01 -1.26845908e+00 -4.01173830e-01
3.86237293e-01 -2.83185184e-01 2.25125670e-01 8.16407084e-01
1.87295943e-01 -8.64581466e-02 4.24462795e-01 -1.32662725e+00
-8.93211484e-01 1.04286516e+00 1.23587310e-01 -1.12444341e-01
7.81759202e-01 2.52618819e-01 1.40718901e+00 -7.07262933e-01
8.18701148e-01 1.47154129e+00 1.00188398e+00 5.16341627e-01
-1.21152329e+00 -3.20971221e-01 1.28546461e-01 5.52462757e-01
-9.07012403e-01 -3.34538549e-01 5.09086847e-01 -4.34210956e-01
1.54338133e+00 5.10957778e-01 5.11967301e-01 1.16390896e+00
3.97001922e-01 7.60424495e-01 1.42456925e+00 -6.41966820e-01
2.46820718e-01 6.47706091e-01 2.23882452e-01 5.91935337e-01
3.61971170e-01 -4.00748290e-02 -4.70556885e-01 -3.12765121e-01
2.02790007e-01 -3.42935860e-01 -1.48157150e-01 5.70783429e-02
-1.10823941e+00 9.48006213e-01 2.47466505e-01 2.44487390e-01
-5.78756213e-01 -7.70158321e-03 2.45985463e-01 5.71616590e-01
6.17460310e-01 9.66538310e-01 -8.60640287e-01 -3.76112610e-01
-7.57576585e-01 8.86844337e-01 1.36553776e+00 6.89412236e-01
4.48737919e-01 -2.80723304e-01 -5.23512721e-01 7.87016153e-01
3.96199077e-01 4.89656478e-01 5.21993041e-01 -1.15997684e+00
7.11950898e-01 1.00030208e+00 3.23074013e-01 -1.29303145e+00
-3.50941688e-01 -5.29937446e-01 -3.74550790e-01 -5.69206625e-02
5.18193483e-01 -1.30930901e-01 -2.57623315e-01 1.61190856e+00
3.70571584e-01 -3.58319491e-01 2.93185323e-01 7.05960155e-01
6.72362208e-01 4.33202654e-01 4.08578217e-01 -1.12627506e-01
1.33683014e+00 -1.09838593e+00 -7.85451055e-01 -5.27530253e-01
9.31159079e-01 -6.03681445e-01 1.22515357e+00 5.11334896e-01
-9.57790852e-01 -3.85341465e-01 -6.17606223e-01 -3.07636261e-01
-3.10298055e-01 -6.60658702e-02 6.28435075e-01 4.90903258e-01
-8.02665293e-01 5.33290684e-01 -4.79428798e-01 -6.23607710e-02
2.66237020e-01 -1.79336861e-01 -2.22892717e-01 -1.80271968e-01
-1.57314825e+00 1.50885701e+00 4.63383913e-01 -1.12374209e-01
-4.42114711e-01 -1.08203018e+00 -7.36515045e-01 1.34961933e-01
6.19983673e-01 -9.38785434e-01 1.66656089e+00 -8.11386228e-01
-1.42246890e+00 5.47114074e-01 3.76799740e-02 -8.76429319e-01
1.01008165e+00 -5.21634459e-01 -3.84067804e-01 -1.15976542e-01
2.00596154e-01 5.18502176e-01 6.42309129e-01 -1.06976128e+00
-4.33759063e-01 -2.12712422e-01 3.87903661e-01 1.05802894e-01
-1.17491774e-01 3.45824733e-02 4.11409885e-01 -3.29355508e-01
-8.22516978e-02 -6.99747086e-01 -2.90104061e-01 -2.96514720e-01
-2.46002570e-01 -4.77297366e-01 4.24344510e-01 -9.73895192e-01
1.39915001e+00 -1.53800404e+00 -2.09063724e-01 -4.63856049e-02
4.23292220e-01 4.52517271e-01 9.61645395e-02 6.30625725e-01
4.34705243e-02 3.22686106e-01 2.01004371e-01 1.18950248e-01
3.24319512e-01 3.09068829e-01 -7.96085775e-01 -4.24889296e-01
3.16010296e-01 1.37978351e+00 -9.53357697e-01 -7.05483019e-01
1.10254504e-01 1.06080763e-01 -7.52070010e-01 2.19749629e-01
-1.04132938e+00 -2.60171015e-02 -3.78469408e-01 1.96490243e-01
3.71651024e-01 -5.42916656e-01 2.88463861e-01 1.35195971e-01
2.97549218e-01 9.23931599e-01 -9.44082856e-01 1.32281685e+00
-6.25339329e-01 3.25901896e-01 -4.81231511e-01 -5.05270004e-01
7.75937438e-01 5.30381262e-01 -2.02998236e-01 -7.93988883e-01
-2.40538642e-01 2.46916473e-01 1.52707547e-01 -1.05665898e+00
5.03102720e-01 -5.33024013e-01 1.32202297e-01 7.46197999e-01
-1.06961131e-01 -5.14633238e-01 1.90486059e-01 6.57732308e-01
1.07643974e+00 2.33963758e-01 5.20300746e-01 -1.52763501e-01
6.12232804e-01 4.47280139e-01 5.20051062e-01 6.39323354e-01
7.73513019e-02 4.66139019e-02 7.40517735e-01 -5.01116455e-01
-9.43433523e-01 -6.13488913e-01 4.16021720e-02 8.17965448e-01
-5.77447653e-01 -6.74127102e-01 -7.72409678e-01 -8.77301753e-01
1.91497579e-01 1.71299016e+00 -4.03459489e-01 -1.04205787e-01
-2.08304048e-01 -3.17141503e-01 7.92445421e-01 4.18972224e-01
3.78839761e-01 -1.19194460e+00 -8.00518811e-01 5.22714317e-01
-7.55834520e-01 -1.30353785e+00 2.02103764e-01 -1.88585415e-01
-8.18087280e-01 -1.09060073e+00 4.34844941e-02 -6.42034486e-02
4.64112848e-01 -1.07498214e-01 1.53315449e+00 3.21319312e-01
2.11026549e-01 2.10902110e-01 -5.30834198e-01 -6.56554818e-01
-1.13766801e+00 -4.50469181e-02 -2.11432561e-01 -3.18145812e-01
5.43467104e-01 -4.40737367e-01 -1.22527085e-01 2.48628125e-01
-8.09819102e-01 4.64513242e-01 3.67852837e-01 7.69965291e-01
4.72027212e-02 -1.10906810e-01 6.93227112e-01 -1.34633636e+00
1.13028681e+00 -2.56807685e-01 -3.00045341e-01 7.17592657e-01
-1.24733794e+00 2.43252859e-01 8.55064332e-01 -9.68394503e-02
-1.60556829e+00 -7.63630092e-01 -6.25873059e-02 -7.45132705e-03
-8.43744129e-02 8.09463859e-01 1.25847071e-01 2.17604339e-01
1.22255290e+00 -1.61934629e-01 3.63738798e-02 -1.87126130e-01
6.04576349e-01 5.74341476e-01 1.33573562e-01 -1.03400636e+00
7.64657378e-01 6.02347404e-02 -4.24409896e-01 -2.38049597e-01
-1.40719795e+00 2.28756089e-02 -3.39603901e-01 -2.41388500e-01
4.05300766e-01 -8.05556715e-01 -8.49124312e-01 9.37966108e-02
-1.33033729e+00 -4.68085468e-01 -3.34195882e-01 7.59699866e-02
-4.10700202e-01 4.88640815e-01 -6.91253483e-01 -1.08164728e+00
-6.24417961e-01 -7.62793183e-01 5.55285454e-01 2.30604243e-02
-8.73998284e-01 -1.09515285e+00 2.98014618e-02 1.16773081e+00
5.47792912e-01 1.26560098e-02 1.32659805e+00 -1.00092256e+00
-5.22184908e-01 -3.59070957e-01 -1.05989479e-01 5.45950890e-01
-4.77480382e-01 -4.74409498e-02 -1.02384531e+00 3.95352542e-01
2.04629242e-01 -7.44483292e-01 1.84897572e-01 -1.16658248e-01
5.88397980e-01 -7.95347452e-01 -1.01667508e-01 -1.29031345e-01
1.04548252e+00 -1.47445884e-03 6.30765021e-01 3.70919138e-01
3.13645244e-01 1.07018471e+00 8.58730257e-01 3.56971890e-01
7.30333149e-01 4.71489698e-01 -8.84211287e-02 3.34355980e-01
-1.79554448e-02 -8.69518280e-01 3.18899930e-01 5.44992089e-01
-8.37271512e-02 -3.73587795e-02 -1.13216662e+00 4.82031345e-01
-2.10408449e+00 -1.14289474e+00 -7.10080147e-01 1.67393017e+00
1.24283886e+00 1.12986296e-01 -3.74271870e-01 1.02286451e-01
1.77042990e-03 -7.37552345e-02 -2.26550251e-01 -6.77119672e-01
-4.63124481e-04 1.16837472e-01 -2.21970320e-01 7.04823434e-01
-2.00264826e-01 1.02345622e+00 5.71374702e+00 7.58602917e-01
-5.65952301e-01 3.85164618e-01 5.32345951e-01 1.87410846e-01
-8.96160543e-01 4.70532358e-01 -5.89287758e-01 1.95229679e-01
9.61149156e-01 -2.87359059e-01 3.49994063e-01 1.09478712e+00
2.70149499e-01 -2.38370627e-01 -1.40345788e+00 4.23943728e-01
4.23119627e-02 -1.30023956e+00 1.41273379e-01 -2.44961500e-01
6.39434934e-01 -3.10826242e-01 -3.88061255e-01 7.15896726e-01
4.49792773e-01 -1.00034666e+00 9.85484660e-01 4.47440237e-01
1.14147902e-01 -3.94202501e-01 9.07176256e-01 1.01775396e+00
-3.78027737e-01 -2.22963318e-01 -3.46102804e-01 -5.95912695e-01
2.64470637e-01 9.22103167e-01 -1.23507214e+00 5.74923575e-01
3.57614458e-01 1.92551255e-01 -7.39446580e-01 3.98873985e-01
-8.34120214e-01 7.38394141e-01 -1.06942162e-01 -3.63532722e-01
-8.86953175e-02 -2.88836285e-02 2.15385482e-01 1.05516648e+00
-1.47856653e-01 2.03880221e-01 -8.38796347e-02 1.34727645e+00
1.23309512e-02 1.81904584e-01 -4.57790673e-01 -1.85205370e-01
4.21926975e-01 1.23583555e+00 -1.09132573e-01 -7.13271976e-01
-2.62775391e-01 5.69141746e-01 7.11118758e-01 1.60358846e-01
-7.51108229e-01 6.30572438e-02 1.82586387e-01 3.63796890e-01
-6.73342049e-02 1.29874095e-01 -5.73942423e-01 -1.17645752e+00
3.75127256e-01 -1.41944492e+00 3.63165140e-01 -1.17184186e+00
-1.34970796e+00 4.29432869e-01 2.55641211e-02 -6.21763647e-01
-5.47476709e-01 -5.54907978e-01 -3.35268527e-01 8.45147908e-01
-1.47316396e+00 -1.36484158e+00 -1.67160779e-01 2.38138422e-01
3.16343933e-01 2.09707081e-01 9.20457602e-01 7.05093741e-02
-1.05087534e-01 4.15947646e-01 -6.77313924e-01 -5.77279702e-02
6.77589953e-01 -1.16512835e+00 5.16286612e-01 6.72727108e-01
2.90470183e-01 1.00977635e+00 9.38030660e-01 -9.36074734e-01
-1.19040227e+00 -8.66772830e-01 1.71861959e+00 -1.14035404e+00
9.67701018e-01 -1.93581432e-01 -1.18620253e+00 9.50348258e-01
1.82292119e-01 -5.53324044e-01 6.09858453e-01 3.87027830e-01
-7.17174411e-01 1.39823742e-02 -1.11481774e+00 6.42081797e-01
8.48984420e-01 -6.77225113e-01 -1.35143244e+00 7.49715149e-01
8.99807811e-01 -4.99219865e-01 -9.10352647e-01 1.31362319e-01
4.09833610e-01 -1.13062263e+00 6.78192198e-01 -9.20461416e-01
9.51523304e-01 -2.36696526e-01 6.52707415e-03 -1.19837022e+00
-2.03382015e-01 -5.53936124e-01 -7.67854750e-02 1.22883356e+00
9.82647121e-01 -8.21630478e-01 4.29917872e-01 1.40887153e+00
7.76975453e-02 -6.05229855e-01 -6.24461770e-01 -5.79955518e-01
5.04057854e-02 -9.61499214e-01 5.81435144e-01 9.91610467e-01
6.55760884e-01 7.90234029e-01 -1.60257861e-01 -1.15568237e-02
3.80565166e-01 1.79666460e-01 8.77069116e-01 -1.30416405e+00
-6.16116941e-01 -1.37239974e-02 1.96388617e-01 -6.89028919e-01
2.52099484e-01 -1.05913115e+00 -6.31750524e-02 -1.76757336e+00
1.51818201e-01 -3.62460762e-01 2.77539700e-01 7.87888229e-01
-4.89904791e-01 -4.06699449e-01 3.37376982e-01 3.84574644e-02
-6.53962135e-01 2.97017455e-01 1.23625863e+00 1.52179703e-01
7.14214370e-02 -1.45719334e-01 -7.89466798e-01 6.56002223e-01
5.63507795e-01 -5.38385212e-01 -5.29343307e-01 -3.48062217e-01
1.04396605e+00 2.22167090e-01 6.57805741e-01 -8.36512148e-01
2.40184441e-01 -2.88109660e-01 3.47804241e-02 -1.70934126e-01
9.76170599e-02 -7.72016048e-01 4.20956224e-01 6.20435417e-01
-7.57131100e-01 1.63869783e-01 1.68243170e-01 4.50067699e-01
-1.61988527e-01 -4.13554937e-01 3.46041650e-01 -2.93361813e-01
-3.92021537e-01 -4.01593566e-01 -2.79961735e-01 2.45874539e-01
8.23745072e-01 4.06449102e-02 -6.58508301e-01 -4.06999886e-01
-3.09092581e-01 3.03809613e-01 2.28588611e-01 4.29558665e-01
4.11207706e-01 -9.27292049e-01 -8.18453908e-01 2.49653552e-02
3.11054379e-01 -7.11459890e-02 2.31055722e-01 9.89895046e-01
-4.70577598e-01 8.83089185e-01 9.59131401e-03 -3.73696059e-01
-1.04741800e+00 5.65927565e-01 2.91339606e-01 -8.29886675e-01
-4.29401368e-01 8.71934175e-01 -3.23341727e-01 -7.77683079e-01
-1.62451744e-01 -5.78604877e-01 -1.74671426e-01 -2.10375562e-01
4.73532110e-01 3.32710892e-01 3.40666980e-01 5.54460920e-02
-2.23877117e-01 -1.69857517e-01 -4.17303920e-01 -3.42665851e-01
1.05670059e+00 2.49986663e-01 -3.86660457e-01 2.90950477e-01
3.91449541e-01 2.81348266e-02 -7.74975836e-01 -2.33275026e-01
5.07579327e-01 -5.03441155e-01 -1.81144953e-01 -1.61099672e+00
-2.46108472e-01 7.93520093e-01 -2.83847839e-01 4.68440652e-01
4.26134408e-01 -1.95246980e-01 7.34598994e-01 7.47030854e-01
6.28097236e-01 -1.21516597e+00 -8.74646306e-02 5.56108773e-01
1.08636403e+00 -1.22607458e+00 -2.67101526e-02 -3.21175396e-01
-1.08837163e+00 9.56343949e-01 9.16319668e-01 4.20273572e-01
1.62206471e-01 1.60567239e-01 3.59005749e-01 -3.20885390e-01
-1.53133690e+00 3.81448418e-01 1.63403794e-01 5.38035691e-01
7.69894063e-01 2.07554713e-01 -5.33617914e-01 9.77513909e-01
-5.64629436e-01 4.44651246e-01 3.96416396e-01 6.69816017e-01
-3.45079005e-01 -1.09408271e+00 -3.62920225e-01 5.52230239e-01
-3.52659285e-01 -3.84563744e-01 -7.07306743e-01 6.67484462e-01
7.10373297e-02 1.30924761e+00 -5.26659906e-01 -2.74962664e-01
4.12768751e-01 5.51945090e-01 3.21032017e-01 -5.84565222e-01
-1.06690192e+00 -6.73159122e-01 8.21112514e-01 -5.36499262e-01
-1.60517558e-01 -4.50959533e-01 -1.23377001e+00 -5.93108296e-01
-4.59941894e-01 5.43361068e-01 4.99648273e-01 1.20007360e+00
6.64362669e-01 3.32689077e-01 -6.57126904e-02 3.06236476e-01
-1.29396117e+00 -1.29963696e+00 -3.57731730e-02 7.58600891e-01
-1.67508960e-01 -3.46683562e-01 -1.96836159e-01 4.51104119e-02] | [10.075693130493164, 7.676489353179932] |
90ab6010-5a3b-4a5b-8ae8-ef0e63bb250f | generative-logic-with-time-beyond-logical | 2301.08509 | null | https://arxiv.org/abs/2301.08509v2 | https://arxiv.org/pdf/2301.08509v2.pdf | Generative Logic with Time: Beyond Logical Consistency and Statistical Possibility | This paper gives a simple theory of inference to logically reason symbolic knowledge fully from data over time. We take a Bayesian approach to model how data causes symbolic knowledge. Probabilistic reasoning with symbolic knowledge is modelled as a process of going the causality forwards and backwards. The forward and backward processes correspond to an interpretation and inverse interpretation of formal logic, respectively. The theory is applied to a localisation problem to show a robot with broken or noisy sensors can efficiently solve the problem in a fully data-driven fashion. | ['Hiroyuki Kido'] | 2023-01-20 | null | null | null | null | ['formal-logic'] | ['reasoning'] | [ 3.37770313e-01 1.01087594e+00 -3.11427563e-02 -6.54755533e-01
-2.00828165e-01 -5.06938577e-01 1.07636869e+00 -3.55814360e-02
-3.06295574e-01 8.40704679e-01 9.21808705e-02 -6.80689216e-01
-6.05730534e-01 -9.81671453e-01 -9.96461689e-01 -6.29417896e-01
-4.36987072e-01 9.34387565e-01 6.19495571e-01 -5.65800183e-02
2.02494085e-01 4.51441437e-01 -1.19977236e+00 9.65276510e-02
-2.70344112e-02 5.76757252e-01 1.91451430e-01 9.18165624e-01
-2.48016924e-01 1.56035483e+00 -1.81372002e-01 1.43595785e-01
-7.55526200e-02 -2.57203281e-01 -1.51112032e+00 -3.61894220e-01
-6.64974034e-01 -1.61882177e-01 -3.25304568e-01 1.16476011e+00
-4.92254436e-01 -9.74068344e-02 5.82084477e-01 -1.60371506e+00
-4.75881457e-01 1.09635091e+00 2.63148657e-04 -2.37779990e-01
1.01353347e+00 -2.04587802e-01 6.27397358e-01 -4.61640149e-01
6.96705222e-01 2.07674599e+00 6.24170959e-01 3.40067476e-01
-1.46088684e+00 -1.38767660e-01 1.33371919e-01 4.93012041e-01
-1.48709071e+00 -2.52079546e-01 5.32472074e-01 -4.60928440e-01
1.01642609e+00 6.50107637e-02 5.45919299e-01 9.02470827e-01
4.97527331e-01 7.84548163e-01 1.40267909e+00 -7.78889000e-01
7.56087899e-01 -1.81689724e-01 4.73320931e-01 4.84724820e-01
4.17264491e-01 8.62963796e-01 -9.19465482e-01 -4.76163000e-01
5.70989251e-01 -2.32701540e-01 1.01354703e-01 -3.71031225e-01
-1.26270330e+00 6.75059438e-01 1.23356894e-01 6.68270886e-02
-4.62661207e-01 9.15566862e-01 2.67475814e-01 2.26064771e-01
-4.18912023e-01 6.42436892e-02 -8.10079634e-01 -1.91995278e-01
-3.77034962e-01 4.34064448e-01 1.41021371e+00 1.00618708e+00
5.86961329e-01 -2.48541147e-01 6.49862707e-01 -2.09665343e-01
1.31374872e+00 1.17189324e+00 2.62322705e-02 -1.52454162e+00
-1.97941467e-01 1.66802421e-01 5.48477173e-01 -8.82298946e-01
-3.06651205e-01 6.61062121e-01 -2.67091632e-01 5.76284587e-01
4.58977073e-01 1.35235190e-01 -1.04561806e+00 1.63334787e+00
2.08634317e-01 1.17431872e-01 7.06975460e-01 5.86676955e-01
4.26627338e-01 6.30713224e-01 1.34823784e-01 -2.14403361e-01
1.52271318e+00 1.51155636e-01 -1.18697929e+00 -1.86436370e-01
4.09911364e-01 -1.27117723e-01 1.12142466e-01 9.54301357e-01
-1.00020242e+00 -9.82654002e-03 -1.11460340e+00 -4.93590683e-02
-4.02906090e-01 -8.94487441e-01 8.10072899e-01 3.34588885e-01
-1.08547282e+00 4.82417732e-01 -1.55004823e+00 -3.25178921e-01
-2.97396481e-02 3.93444955e-01 -3.47095668e-01 -2.15143651e-01
-1.55497515e+00 1.20504785e+00 1.07492423e+00 5.32001019e-01
-1.28435504e+00 1.60653278e-01 -9.81778443e-01 -5.17689586e-01
6.07116759e-01 -7.05165863e-01 1.58501375e+00 -5.01619995e-01
-1.70875120e+00 5.89373589e-01 -4.60238248e-01 -7.84443855e-01
1.48513839e-01 -3.64113480e-01 -4.57006484e-01 8.20215642e-02
1.55644193e-01 3.35639805e-01 4.64625865e-01 -1.44880772e+00
-5.80334663e-01 -5.61143756e-01 1.57368794e-01 -3.78170401e-01
9.45787668e-01 1.69867426e-01 2.31151492e-03 2.18829870e-01
8.68230879e-01 -9.45817113e-01 -3.51888537e-01 -4.35791284e-01
-6.01687908e-01 -2.19266146e-01 5.51557779e-01 -2.57884830e-01
6.64469838e-01 -1.73081934e+00 1.71831414e-01 6.93746686e-01
-2.80054696e-02 -5.98439872e-01 3.28183681e-01 5.29039741e-01
-1.00154974e-01 2.44843870e-01 -2.02822849e-01 3.22821796e-01
4.26061451e-01 1.19868064e+00 -1.02579689e+00 5.59893310e-01
2.01271340e-01 7.78871953e-01 -1.35011446e+00 -5.97270250e-01
1.70585036e-01 2.26671860e-01 -9.75023359e-02 9.25212633e-03
-5.74498355e-01 2.19689235e-01 -7.74595976e-01 3.75956178e-01
5.08240044e-01 1.99089170e-01 8.30174387e-01 4.84974891e-01
-2.10232154e-01 5.01895607e-01 -1.52718282e+00 1.60448825e+00
-3.42271551e-02 4.41404462e-01 1.32502243e-01 -7.16798663e-01
7.94071555e-01 7.87375629e-01 -2.65080392e-01 -2.27845654e-01
3.24727446e-01 1.72806755e-01 -2.01139688e-01 -7.34653175e-01
3.11856896e-01 -8.25001001e-01 -6.37092292e-01 3.35024923e-01
-2.46937558e-01 -5.20339429e-01 -2.28179976e-01 3.28913897e-01
1.44228232e+00 8.45576346e-01 6.10204279e-01 -2.38604039e-01
3.28572929e-01 5.18905401e-01 6.54581428e-01 1.01542020e+00
2.69113109e-02 -2.19199792e-01 6.00436389e-01 -5.12024820e-01
-3.69789541e-01 -1.60125673e+00 7.09365830e-02 8.79667163e-01
4.55493957e-01 -2.94408917e-01 -3.16792667e-01 -1.55272394e-01
9.80733428e-03 1.36489177e+00 -5.69040358e-01 -1.79345593e-01
-1.69182390e-01 -3.30532521e-01 8.28376591e-01 6.22227371e-01
1.63786083e-01 -1.18161738e+00 -1.09551787e+00 2.08871260e-01
-1.05858721e-01 -8.82605016e-01 7.71220982e-01 7.47173846e-01
-9.82097983e-01 -1.03270209e+00 5.08347392e-01 -9.12476704e-02
4.95556414e-01 -2.35551313e-01 7.43327916e-01 -1.89819559e-01
1.37482837e-01 5.34674525e-01 -3.25256586e-01 -6.57386541e-01
-7.40354776e-01 -8.44696641e-01 2.94286489e-01 -5.62734723e-01
5.77722788e-01 -6.48142934e-01 3.40837300e-01 1.83807462e-01
-8.42222869e-01 -1.85735539e-01 3.75579596e-01 5.78359127e-01
3.40585232e-01 6.58305466e-01 1.30541679e-02 -6.00305855e-01
3.96716624e-01 -6.88501120e-01 -7.62159407e-01 1.68396682e-01
-3.51929009e-01 6.90120339e-01 1.21004798e-01 -6.20659478e-02
-1.31724250e+00 2.97880679e-01 5.80107331e-01 -8.68195668e-02
-6.69994175e-01 8.89917135e-01 -5.09654403e-01 4.62679297e-01
5.98893642e-01 3.76195423e-02 -1.48132473e-01 -2.32387394e-01
7.18902767e-01 2.68660277e-01 8.92123401e-01 -1.22452605e+00
5.63112676e-01 9.36121047e-01 7.09841251e-01 -3.02497298e-01
-5.34443557e-01 -9.90408137e-02 -9.73193824e-01 -1.21158093e-01
7.73673475e-01 -6.31155550e-01 -1.29119039e+00 1.37200266e-01
-1.53258288e+00 -4.42473739e-01 -4.63695228e-01 4.89246398e-01
-1.11328948e+00 3.22872847e-01 -2.01270685e-01 -1.62976193e+00
6.07485592e-01 -8.19660544e-01 8.94902349e-01 -2.46672198e-01
-7.52127707e-01 -1.11771107e+00 3.32888097e-01 -1.86996952e-01
-3.50051045e-01 4.25637275e-01 4.99455839e-01 -7.39926517e-01
-8.06390643e-01 -2.48572782e-01 -6.75759912e-02 -1.46352455e-01
-3.23721468e-01 1.95156619e-01 -9.61583555e-01 4.03103918e-01
1.16361976e-01 -1.19039200e-01 3.67080361e-01 1.20689318e-01
3.98922890e-01 -3.45654279e-01 -8.47561538e-01 -1.95223644e-01
1.38821876e+00 3.92855257e-01 5.78940809e-01 4.23535556e-01
1.57643452e-01 8.02902818e-01 8.48870516e-01 1.97215885e-01
5.81270874e-01 3.74913603e-01 4.90326524e-01 7.57022083e-01
3.36731255e-01 -4.27965879e-01 5.59500158e-01 2.72181034e-01
-4.15269613e-01 5.91606796e-02 -1.56376529e+00 7.48779476e-01
-2.25578189e+00 -1.38123620e+00 -6.66519761e-01 1.85199630e+00
1.08707809e+00 3.36437464e-01 -4.75893497e-01 3.76191318e-01
5.23544490e-01 -4.07235056e-01 -7.55245462e-02 -7.09438264e-01
1.07869521e-01 1.20654292e-02 6.92121446e-01 1.03580928e+00
-5.91172338e-01 9.22530353e-01 8.29438686e+00 3.73538077e-01
-3.02488476e-01 1.44596741e-01 -4.87729996e-01 2.95843333e-01
-4.26955491e-01 8.99437904e-01 -4.90497440e-01 5.07912077e-02
1.50492609e+00 7.34093338e-02 6.53205037e-01 6.92659736e-01
2.01388747e-01 -9.20702159e-01 -1.55470943e+00 5.41613460e-01
-4.91535276e-01 -1.02266037e+00 -1.50118366e-01 5.33562750e-02
2.82106936e-01 5.22180013e-02 -6.01898015e-01 -2.06222050e-02
1.65349209e+00 -1.09006131e+00 1.53304827e+00 1.30509317e+00
1.72159478e-01 -4.97598529e-01 8.36973011e-01 6.48748219e-01
-7.43900955e-01 -2.54812300e-01 8.90402794e-02 -7.00349331e-01
7.51634359e-01 8.33491921e-01 -1.26652920e+00 6.98342621e-01
7.38147080e-01 2.29751393e-01 1.27999961e-01 4.24122691e-01
-1.05919993e+00 5.53515375e-01 -7.60960519e-01 1.06590223e-02
-3.71617973e-02 1.09381497e-01 7.03352213e-01 1.12219930e+00
4.85834554e-02 4.22202915e-01 3.00428066e-02 1.04648292e+00
8.16798806e-01 -1.14188695e+00 -8.01264465e-01 1.79825798e-01
5.90949357e-01 2.72196859e-01 -6.72077000e-01 -4.17110622e-01
1.02305897e-01 4.18918312e-01 -2.76820064e-01 4.20058519e-01
-4.63749230e-01 -2.87179649e-01 1.55394286e-01 -3.11952889e-01
4.07033078e-02 -5.16009748e-01 -4.68998402e-01 -4.96192396e-01
-3.55812520e-01 -1.86795890e-01 3.93715203e-01 -1.46266103e+00
-9.48131621e-01 -3.31818592e-03 9.69376147e-01 -5.40886581e-01
-6.34646237e-01 -6.77251637e-01 -3.06737453e-01 8.44327807e-01
-1.10223103e+00 -9.22177196e-01 2.07034230e-01 4.09733981e-01
-3.15761179e-01 3.59749317e-01 1.13248634e+00 -7.66477466e-01
1.40217587e-01 -6.49657726e-01 -2.73349792e-01 -1.28347222e-02
1.47668481e-01 -1.54899549e+00 1.91980675e-01 8.88234913e-01
-2.61729181e-01 1.10626590e+00 1.43526340e+00 -9.40328419e-01
-1.67955971e+00 -5.42837501e-01 1.08524477e+00 -1.06457007e+00
1.07014358e+00 -6.95886686e-02 -6.69174492e-01 1.34928429e+00
2.07990650e-02 -1.37296796e-01 4.42247778e-01 2.97661066e-01
-5.34434974e-01 1.64221764e-01 -1.31630719e+00 4.20674354e-01
9.98130620e-01 -5.56451619e-01 -1.80720425e+00 -3.54339294e-02
6.87277973e-01 -2.49769658e-01 -7.98270106e-01 3.96391064e-01
7.37779796e-01 -6.44980192e-01 7.50442445e-01 -4.09010679e-01
5.96027374e-02 -1.06554985e+00 -6.02357030e-01 -7.02033818e-01
-4.30403128e-02 -8.74803364e-01 -2.38566980e-01 8.26448798e-01
1.51364848e-01 -9.88695025e-01 2.80267149e-01 7.85076380e-01
2.62753218e-01 4.99985926e-02 -1.35573924e+00 -8.97245347e-01
-1.93897739e-01 -1.29926550e+00 6.84423685e-01 5.64835966e-01
6.66660488e-01 7.51511082e-02 1.97011814e-01 7.44396985e-01
8.28423917e-01 -1.10330908e-02 4.21088785e-01 -1.55586028e+00
-6.89487383e-02 -6.60742968e-02 -8.09869826e-01 -8.18126619e-01
3.54366034e-01 -6.51517212e-01 9.76855397e-01 -1.59560561e+00
-1.37485387e-02 -3.19545895e-01 4.90413010e-02 8.20920467e-01
6.30030513e-01 -4.27954853e-01 -1.41354024e-01 1.61868811e-01
-7.39223003e-01 2.15623260e-01 6.08393371e-01 7.72942901e-02
-1.90564860e-02 -3.58174205e-01 -4.28795397e-01 1.25171375e+00
4.35643226e-01 -8.88884246e-01 -5.09750128e-01 -1.05991334e-01
1.16505826e+00 4.48352575e-01 9.57479060e-01 -5.33501387e-01
6.08588338e-01 -7.84597218e-01 -8.18162188e-02 -7.28973866e-01
3.12733114e-01 -1.09159780e+00 7.06440032e-01 6.07684970e-01
-3.02862704e-01 -3.86996388e-01 1.17168985e-01 9.60408211e-01
-1.35995582e-01 -4.21079755e-01 1.63572550e-01 -1.37701422e-01
-1.07321644e+00 -5.97886384e-01 -9.16936398e-01 -3.97948682e-01
8.95543873e-01 -1.48920432e-01 -2.27809381e-02 -3.02493662e-01
-1.52731311e+00 3.33741426e-01 3.20962578e-01 2.90583838e-02
7.19496310e-01 -1.22575927e+00 -1.46151572e-01 -7.95222521e-02
-1.19703852e-01 4.55463856e-01 -2.57706553e-01 1.01378763e+00
-5.08090436e-01 8.12156320e-01 1.65712804e-01 -7.89580405e-01
-7.71035969e-01 7.40559459e-01 2.10217834e-01 1.04635671e-01
-4.72536802e-01 7.42331982e-01 -1.78461835e-01 -5.79780221e-01
9.19745266e-02 -8.60527635e-01 2.41939962e-01 -5.22771060e-01
7.04423368e-01 5.91246367e-01 -5.58383048e-01 -5.01976371e-01
-8.35383892e-01 4.33708578e-01 6.63011014e-01 -9.79667962e-01
9.52375233e-01 -4.77515340e-01 -7.91361690e-01 1.26689363e+00
3.00051510e-01 -3.30518305e-01 -8.87817860e-01 -2.39204362e-01
6.65071011e-01 -8.13232511e-02 -1.14238262e-02 -1.10629070e+00
4.36139330e-02 6.23201251e-01 4.51862812e-02 5.52469671e-01
5.61922789e-01 6.93705022e-01 -1.17133648e-04 7.85247982e-01
1.04179502e+00 -9.90787566e-01 -5.91369212e-01 8.36690366e-01
1.05371761e+00 -6.84501827e-01 1.72202840e-01 -4.54211593e-01
-2.73138553e-01 1.29075229e+00 -1.98019013e-01 -3.65047678e-02
9.00285304e-01 7.83751011e-01 -2.23167658e-01 -6.19085491e-01
-1.00440204e+00 -5.99122234e-02 -3.82561594e-01 9.42429483e-01
-2.70048410e-01 5.86880147e-01 1.28169730e-01 8.05368602e-01
-3.83676559e-01 8.12701821e-01 6.80578470e-01 1.43643904e+00
-7.70333707e-01 -7.62844443e-01 -1.26419938e+00 -3.86954874e-01
-1.24631323e-01 3.19724441e-01 -4.55619812e-01 9.69884396e-01
2.13005498e-01 1.47672284e+00 -3.47787254e-02 -2.75311142e-01
2.31866047e-01 3.80905539e-01 7.28839636e-01 -5.86231172e-01
4.11744028e-01 -2.88083345e-01 3.09734344e-01 -8.62496793e-01
-9.29653227e-01 -1.11746192e+00 -2.32660151e+00 -3.04347992e-01
-6.77121878e-02 1.82411879e-01 8.07537735e-01 1.44644856e+00
-2.36922145e-01 5.62377393e-01 1.08932748e-01 -6.46977127e-01
-5.89842856e-01 -7.27483928e-01 -7.03272581e-01 -4.36502010e-01
4.94515657e-01 -8.57341826e-01 -6.00254357e-01 4.45560366e-01] | [8.60002613067627, 6.586549282073975] |
8439f68b-06f5-40b2-9596-91b7c0e1e07e | signet-semantic-instance-aided-unsupervised | 1812.05642 | null | http://arxiv.org/abs/1812.05642v2 | http://arxiv.org/pdf/1812.05642v2.pdf | SIGNet: Semantic Instance Aided Unsupervised 3D Geometry Perception | Unsupervised learning for geometric perception (depth, optical flow, etc.) is
of great interest to autonomous systems. Recent works on unsupervised learning
have made considerable progress on perceiving geometry; however, they usually
ignore the coherence of objects and perform poorly under scenarios with dark
and noisy environments. In contrast, supervised learning algorithms, which are
robust, require large labeled geometric dataset. This paper introduces SIGNet,
a novel framework that provides robust geometry perception without requiring
geometrically informative labels. Specifically, SIGNet integrates semantic
information to make depth and flow predictions consistent with objects and
robust to low lighting conditions. SIGNet is shown to improve upon the
state-of-the-art unsupervised learning for depth prediction by 30% (in squared
relative error). In particular, SIGNet improves the dynamic object class
performance by 39% in depth prediction and 29% in flow prediction. Our code
will be made available at https://github.com/mengyuest/SIGNet | ['Aman Raj', 'Samuel Sunarjo', 'Dinesh Bharadia', 'Rui Guo', 'Yue Meng', 'Yongxi Lu', 'Tara Javidi', 'Gaurav Bansal'] | 2018-12-13 | signet-semantic-instance-aided-unsupervised-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Meng_SIGNet_Semantic_Instance_Aided_Unsupervised_3D_Geometry_Perception_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Meng_SIGNet_Semantic_Instance_Aided_Unsupervised_3D_Geometry_Perception_CVPR_2019_paper.pdf | cvpr-2019-6 | ['3d-geometry-perception'] | ['computer-vision'] | [ 2.88926885e-02 1.24645106e-01 -1.75044402e-01 -6.13624752e-01
-2.16878101e-01 -5.22160828e-01 6.00138903e-01 1.25355840e-01
-2.45395780e-01 5.59837639e-01 1.66877329e-01 7.20616207e-02
-2.57247686e-02 -9.03937876e-01 -4.57365364e-01 -7.35353112e-01
-2.18390003e-02 4.13057059e-01 4.27805007e-01 2.25882113e-01
6.20840192e-01 4.78792012e-01 -1.79401445e+00 2.67908294e-02
8.09306502e-01 1.12580919e+00 1.15064025e-01 8.91608655e-01
-1.61584392e-01 1.08304584e+00 -3.63431185e-01 7.88722709e-02
3.38428020e-01 -3.21161509e-01 -7.12364912e-01 2.89581120e-01
1.03546381e+00 -4.61234003e-01 -4.96193588e-01 8.66721213e-01
4.13101196e-01 3.10725272e-01 8.78563225e-01 -1.35142255e+00
-5.34147620e-01 2.25091755e-01 -2.80943871e-01 -6.97580352e-02
3.24287921e-01 2.25291535e-01 9.97851312e-01 -9.69777048e-01
6.61980569e-01 1.20621538e+00 3.27426642e-01 6.03915274e-01
-9.13655698e-01 -4.14890736e-01 1.93159372e-01 3.99686754e-01
-1.12797749e+00 -6.24620318e-01 7.73869455e-01 -4.78113472e-01
9.40424442e-01 2.72696670e-02 5.53137660e-01 7.82413900e-01
-8.74486417e-02 8.86860549e-01 1.06995308e+00 -3.22490752e-01
5.26732922e-01 -1.03206731e-01 -1.17694445e-01 9.34518874e-01
2.74162650e-01 4.45260823e-01 -7.55701780e-01 5.47146499e-01
9.35882866e-01 -1.20628543e-01 -4.00426202e-02 -5.65962732e-01
-1.26638031e+00 6.05198503e-01 7.73786485e-01 -1.38832048e-01
7.34685957e-02 3.89657199e-01 -2.05644313e-02 1.50671214e-01
4.92644101e-01 4.32373255e-01 -3.65679234e-01 -3.14689994e-01
-5.62276423e-01 -5.76590709e-02 8.44283760e-01 1.05472267e+00
8.84756148e-01 2.03455925e-01 3.38525832e-01 7.17983782e-01
5.61132431e-01 8.46365571e-01 1.81427419e-01 -1.34258819e+00
2.86189377e-01 5.87562263e-01 2.48832721e-02 -1.09174073e+00
-6.00280702e-01 -1.82252437e-01 -6.61918283e-01 6.01132512e-01
7.91173518e-01 -8.94004852e-02 -1.04520762e+00 1.27184999e+00
2.50874609e-01 3.56730700e-01 1.29953951e-01 1.12807143e+00
1.02252257e+00 4.71746057e-01 -4.16857414e-02 1.52101070e-01
7.17894852e-01 -1.15692568e+00 -3.61669242e-01 -3.73125404e-01
7.20624268e-01 -8.56468916e-01 8.86257231e-01 6.39169216e-01
-9.36270058e-01 -7.02650785e-01 -9.73738611e-01 -1.91333324e-01
-4.41165835e-01 -3.35027352e-02 8.44983518e-01 7.69803703e-01
-9.43841875e-01 5.12585580e-01 -9.45764303e-01 -5.32920659e-01
6.70094907e-01 3.12963098e-01 -2.51040161e-01 -2.20026121e-01
-4.93610561e-01 7.21514404e-01 3.20942909e-01 1.76870171e-02
-8.48536074e-01 -5.19659340e-01 -1.14914274e+00 -2.36950755e-01
4.06008840e-01 -6.78847075e-01 1.24878395e+00 -8.98793995e-01
-1.79986572e+00 8.56749833e-01 -3.37739289e-01 -4.37866181e-01
7.76525080e-01 -4.98508632e-01 -5.99533767e-02 3.75875711e-01
-9.32361633e-02 1.13910210e+00 7.26426959e-01 -1.37391770e+00
-6.94062948e-01 -3.18502873e-01 -1.48364715e-02 3.97731930e-01
-1.08057268e-01 -3.87865007e-01 -1.97531104e-01 -3.46242189e-01
5.28661311e-01 -9.13331985e-01 -2.24551454e-01 5.82670212e-01
-2.91668504e-01 -1.68109834e-01 1.08588183e+00 -5.86616956e-02
4.69033569e-01 -1.87907302e+00 -1.29985198e-01 1.11186258e-01
3.34674925e-01 1.50642037e-01 1.48028016e-01 7.89023563e-02
3.91253769e-01 -8.16039219e-02 -2.76803702e-01 -4.66578960e-01
4.04702425e-02 4.40392166e-01 -5.89351840e-02 6.51243091e-01
1.59357369e-01 9.27669346e-01 -1.17896807e+00 -6.61706865e-01
9.31848943e-01 3.82564992e-01 -6.13642037e-01 1.36376753e-01
-3.06652158e-01 9.17139292e-01 -2.95555264e-01 8.46117437e-01
6.49561644e-01 -3.80600691e-01 -2.64541116e-02 7.73844123e-02
-1.15625076e-01 3.15439582e-01 -1.26157331e+00 1.72333825e+00
-2.90352643e-01 1.01004493e+00 -4.44850981e-01 -9.54401672e-01
1.22592473e+00 -8.74140412e-02 6.79880321e-01 -7.05760181e-01
1.53540581e-01 2.10415319e-01 -5.26723117e-02 -6.25110030e-01
3.97787005e-01 4.98656332e-02 2.25420833e-01 2.98892170e-01
-5.26735839e-03 -4.60363388e-01 4.16280292e-02 4.01684381e-02
7.59432852e-01 4.32190031e-01 1.25896841e-01 -2.43515056e-02
4.44101512e-01 -1.29712239e-01 5.40604949e-01 5.52196980e-01
-4.97699142e-01 7.69940495e-01 1.86480749e-02 -5.13062119e-01
-8.56027842e-01 -1.36665702e+00 -7.21636340e-02 9.04051661e-01
7.22443581e-01 -1.94004357e-01 -3.65508139e-01 -4.90284264e-01
9.18464884e-02 5.50723791e-01 -3.86791915e-01 5.98917939e-02
-5.32860994e-01 -3.11314166e-01 2.74944335e-01 8.77168238e-01
8.16027522e-01 -1.20227087e+00 -8.62853825e-01 -9.58444029e-02
-4.90194187e-02 -1.50516105e+00 -9.16798413e-02 -1.67178631e-01
-1.11772275e+00 -1.23690522e+00 -5.01764178e-01 -8.46265197e-01
9.11942840e-01 3.47878039e-01 1.04110205e+00 1.59698665e-01
-4.52069908e-01 5.52609146e-01 -5.42856932e-01 -5.01039088e-01
8.68093520e-02 -4.79298458e-02 1.27805889e-01 7.20994025e-02
2.84101784e-01 -5.77607453e-01 -8.64438236e-01 6.48069143e-01
-5.51879108e-01 1.21787786e-01 3.13584417e-01 4.39524561e-01
6.08937263e-01 -1.90427601e-01 2.95494080e-01 -7.18751073e-01
-1.54141545e-01 -1.06788531e-01 -7.62056768e-01 -1.95206672e-01
-6.18115723e-01 1.10893715e-02 4.42258209e-01 -1.52619973e-01
-1.13292706e+00 2.29930609e-01 2.09374726e-01 -3.04956168e-01
-6.16113544e-01 -4.98094223e-02 -6.73609525e-02 -3.44120294e-01
6.98412120e-01 1.12857170e-01 -9.18815285e-02 -3.24518889e-01
6.33515239e-01 4.39833641e-01 5.61465204e-01 -5.63568592e-01
8.54278028e-01 9.33989406e-01 3.14774275e-01 -9.83383179e-01
-8.22949350e-01 -7.27359712e-01 -1.27833974e+00 -5.31741440e-01
8.40400159e-01 -6.09041750e-01 -7.29368091e-01 6.59468353e-01
-7.41510630e-01 -8.07534695e-01 -2.64389545e-01 8.61406624e-01
-9.61084783e-01 4.22418296e-01 -3.49050194e-01 -9.62257504e-01
5.59252910e-02 -7.44571030e-01 9.82787311e-01 5.05951464e-01
-7.95243457e-02 -1.32157004e+00 -1.58931300e-01 4.40695673e-01
2.20161796e-01 3.26918840e-01 2.86799371e-01 -6.95907250e-02
-1.02768326e+00 7.63243139e-02 -5.03227115e-01 3.33736092e-01
3.71762156e-01 1.61702007e-01 -1.28645539e+00 -5.60701415e-02
-4.59206671e-01 -4.95934874e-01 1.00104368e+00 5.45991540e-01
1.16978276e+00 -4.10482194e-03 -1.56433523e-01 8.27270508e-01
1.50111151e+00 1.77359357e-01 6.89287484e-01 2.14924127e-01
1.00695705e+00 7.93426812e-01 6.98772967e-01 5.05816400e-01
5.09640872e-01 4.27736402e-01 6.98727965e-01 -1.15422755e-01
-4.64401722e-01 -3.49535346e-01 2.29362115e-01 6.60560966e-01
-4.32446450e-01 -2.63756633e-01 -1.04327154e+00 5.14872849e-01
-1.90044153e+00 -9.33192313e-01 -2.85080701e-01 2.15225506e+00
1.90592602e-01 3.11966360e-01 1.48887619e-01 3.73221755e-01
3.26963484e-01 1.27189443e-01 -6.30219579e-01 -2.14247823e-01
-3.04828852e-01 4.57409434e-02 6.22032523e-01 6.91427052e-01
-1.27389991e+00 1.32693398e+00 6.07401037e+00 2.58330405e-01
-1.13285387e+00 -2.25939497e-01 4.20193017e-01 1.31618738e-01
1.00866526e-01 1.11137897e-01 -7.03752577e-01 2.50057966e-01
5.21132350e-01 6.22033179e-02 -2.39623268e-03 8.95127356e-01
5.28487325e-01 -5.06383300e-01 -1.20116317e+00 1.24738812e+00
1.64340049e-01 -1.12907195e+00 -1.82351530e-01 -9.34632216e-03
1.07619596e+00 1.20645493e-01 1.15748517e-01 -1.51607469e-01
2.97776550e-01 -1.14687049e+00 5.42009115e-01 7.93723404e-01
5.45179665e-01 -4.17465538e-01 6.15345180e-01 2.26065293e-01
-1.30208087e+00 4.54652635e-03 -5.74778497e-01 -4.12629575e-01
3.19710195e-01 4.90036279e-01 -7.83493280e-01 2.53615171e-01
6.65554285e-01 1.42340803e+00 -7.29274511e-01 1.54471529e+00
-5.96299648e-01 6.14401758e-01 -4.67909992e-01 -1.33374989e-01
2.64024109e-01 -2.27957353e-01 4.50955123e-01 1.09383440e+00
3.54149938e-02 1.20091870e-01 5.94583988e-01 6.09159112e-01
1.12469852e-01 1.28422335e-01 -6.52895808e-01 3.25528532e-01
3.06213140e-01 9.57351625e-01 -9.81511414e-01 -4.95185643e-01
-3.24720085e-01 9.16532874e-01 9.80261788e-02 3.46156359e-01
-6.31662250e-01 -2.57443994e-01 5.38685560e-01 1.62474096e-01
2.02296600e-01 -7.07394421e-01 -8.21343303e-01 -1.21189713e+00
-5.39474040e-02 -6.08235151e-02 3.12846005e-01 -1.04621947e+00
-1.18848622e+00 1.44323051e-01 -1.08266778e-01 -1.59052515e+00
-1.36277139e-01 -1.03615904e+00 -4.37358290e-01 2.23122284e-01
-1.79665279e+00 -1.00226033e+00 -9.23352182e-01 6.14121377e-01
7.45213568e-01 -7.00619295e-02 7.49197960e-01 1.77542120e-01
-2.58554161e-01 3.64692420e-01 6.90525547e-02 4.09728855e-01
8.38830948e-01 -1.49577439e+00 1.69481352e-01 7.28036761e-01
4.94869322e-01 3.43835577e-02 4.09080386e-01 -3.87298256e-01
-1.13033271e+00 -1.09695876e+00 7.62436390e-01 -6.16548598e-01
4.92615283e-01 -2.91864693e-01 -5.89277983e-01 5.46814203e-01
-1.67729467e-01 4.36921507e-01 3.95193100e-01 -1.09266408e-01
-5.71335793e-01 -1.07818261e-01 -1.13398468e+00 6.05878294e-01
1.48806405e+00 -4.68638480e-01 -3.65194768e-01 4.46052521e-01
3.04557592e-01 -3.78052920e-01 -7.69930661e-01 3.84274989e-01
6.83097720e-01 -1.26379430e+00 9.85104203e-01 -4.65369225e-01
4.03166950e-01 -4.20809776e-01 -2.66065836e-01 -9.48533654e-01
-1.03687294e-01 -5.09166300e-01 -3.48157376e-01 7.50518262e-01
3.39339048e-01 -7.44464099e-01 1.35578430e+00 3.47286075e-01
-1.93408161e-01 -4.29745704e-01 -5.68956375e-01 -9.58414257e-01
1.84835456e-02 -7.62919247e-01 4.62943427e-02 8.54308903e-01
-2.96611656e-02 1.34910271e-01 -2.52117753e-01 1.91267356e-01
9.88117099e-01 1.97531164e-01 9.37871397e-01 -1.38611305e+00
-1.04382209e-01 -6.33224070e-01 -1.01082230e+00 -1.50992739e+00
1.00397673e-02 -7.43719399e-01 2.36118212e-01 -1.82518578e+00
-3.63040626e-01 -6.76394522e-01 -6.48368374e-02 5.17177224e-01
8.69865119e-02 6.17264450e-01 3.11203301e-01 3.41354340e-01
-8.54534745e-01 5.36622584e-01 1.46194065e+00 -1.85184166e-01
-2.49060795e-01 2.35441886e-02 -1.80688217e-01 1.05565155e+00
1.35809016e+00 -1.13174438e-01 -5.90150237e-01 -6.37001991e-01
-2.24055916e-01 -3.46195549e-01 5.53586662e-01 -1.41101289e+00
3.91038328e-01 -3.34806204e-01 5.38971126e-01 -5.70128202e-01
4.66115236e-01 -6.15701735e-01 -4.60704267e-01 4.75871980e-01
-2.31236711e-01 -3.05385381e-01 7.92722106e-02 5.22301674e-01
-9.89663079e-02 -6.59919903e-03 8.85250211e-01 -2.27473021e-01
-1.33183026e+00 4.00851607e-01 -2.22657815e-01 1.58339798e-01
8.94789994e-01 -7.34377801e-01 -3.42550576e-01 -6.76369905e-01
-8.80575657e-01 1.12567998e-01 4.70809400e-01 4.70566064e-01
9.96377826e-01 -1.16345954e+00 -7.25617230e-01 2.48670727e-01
3.14806670e-01 3.10261041e-01 3.80887724e-02 6.77510917e-01
-9.39820051e-01 5.22505164e-01 -3.03059101e-01 -9.62709785e-01
-1.17752147e+00 1.22054905e-01 3.16814274e-01 4.43574458e-01
-8.03720176e-01 1.07235670e+00 3.04083556e-01 -3.68415684e-01
4.04980540e-01 -3.86049986e-01 -2.71120798e-02 -2.03754008e-01
3.55013728e-01 6.03321433e-01 -2.33675361e-01 -7.60539114e-01
-3.40815544e-01 1.18621445e+00 3.41384560e-01 1.21958880e-02
1.04673624e+00 -2.53579140e-01 2.19440639e-01 5.26242733e-01
1.09857595e+00 -8.03011805e-02 -1.74851465e+00 -3.15612584e-01
9.05861259e-02 -7.66177595e-01 6.07071705e-02 -7.55861819e-01
-1.13400221e+00 1.08001947e+00 6.33895636e-01 -8.23004693e-02
9.61949170e-01 5.26749454e-02 5.68670034e-01 5.10680974e-01
6.09263420e-01 -1.04468060e+00 4.89925683e-01 7.08366513e-01
4.01117176e-01 -1.71678400e+00 -1.13904895e-02 -7.78850675e-01
-4.89228547e-01 1.26903319e+00 8.83095443e-01 -3.70197952e-01
8.44004929e-01 4.89002764e-02 4.94545877e-01 -1.84557915e-01
-4.76923376e-01 -5.37692845e-01 3.95819217e-01 1.01334393e+00
3.12639713e-01 1.29498735e-01 1.56658366e-01 -5.65657377e-01
-5.25057733e-01 -1.81603149e-01 4.68635678e-01 9.21217859e-01
-7.81266749e-01 -8.27740550e-01 -2.07333937e-01 1.69205770e-01
1.53854027e-01 1.24431709e-02 -4.62145925e-01 6.47856653e-01
2.78417289e-01 1.10292494e+00 3.20594311e-01 -2.39297017e-01
1.52488798e-01 -1.97965845e-01 8.01363170e-01 -6.49885774e-01
2.05798686e-01 2.54135784e-02 6.84837624e-02 -6.77151084e-01
-9.07906473e-01 -5.91425359e-01 -1.67341805e+00 -1.78331345e-01
-2.49689501e-02 -2.64402956e-01 5.69261849e-01 9.51677203e-01
1.11321853e-02 1.66140512e-01 6.62389278e-01 -9.48205352e-01
8.65240991e-02 -7.83303738e-01 -4.01132345e-01 4.13737565e-01
2.91418254e-01 -9.06438768e-01 -4.58668023e-01 4.58567262e-01] | [8.62179946899414, -1.975090503692627] |
f7abcafc-1be9-44e2-ae44-85152b2c8d5e | privacy-preserving-household-load-forecasting | 2206.15192 | null | https://arxiv.org/abs/2206.15192v1 | https://arxiv.org/pdf/2206.15192v1.pdf | Privacy-preserving household load forecasting based on non-intrusive load monitoring: A federated deep learning approach | Load forecasting is very essential in the analysis and grid planning of power systems. For this reason, we first propose a household load forecasting method based on federated deep learning and non-intrusive load monitoring (NILM). For all we know, this is the first research on federated learning (FL) in household load forecasting based on NILM. In this method, the integrated power is decomposed into individual device power by non-intrusive load monitoring, and the power of individual appliances is predicted separately using a federated deep learning model. Finally, the predicted power values of individual appliances are aggregated to form the total power prediction. Specifically, by separately predicting the electrical equipment to obtain the predicted power, it avoids the error caused by the strong time dependence in the power signal of a single device. And in the federated deep learning prediction model, the household owners with the power data share the parameters of the local model instead of the local power data, guaranteeing the privacy of the household user data. The case results demonstrate that the proposed approach provides a better prediction effect than the traditional methodology that directly predicts the aggregated signal as a whole. In addition, experiments in various federated learning environments are designed and implemented to validate the validity of this methodology. | ['Jianhong Pan', 'Jian Wang', 'Jingru Feng', 'Xinxin Zhou'] | 2022-06-30 | null | null | null | null | ['non-intrusive-load-monitoring', 'load-forecasting', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'miscellaneous', 'time-series'] | [-5.77858210e-01 -6.07370026e-02 -1.92280427e-01 -3.59512448e-01
-1.15432613e-01 -3.09586704e-01 2.66616285e-01 -3.21643353e-01
2.66924232e-01 8.07044864e-01 8.51236805e-02 -7.38602057e-02
-1.19531803e-01 -1.14864337e+00 -3.95148367e-01 -1.27243912e+00
-3.12878162e-01 1.18319847e-01 -6.68808937e-01 1.67273656e-01
-3.73588085e-01 4.21651632e-01 -1.27132380e+00 -4.35367189e-02
9.47674155e-01 1.51257372e+00 -4.60603423e-02 1.36485305e-02
4.24375804e-03 1.13651788e+00 -9.20179904e-01 -2.48018667e-01
2.53888845e-01 -3.57688129e-01 -5.91768146e-01 -1.11983195e-01
-3.68660480e-01 -1.01570439e+00 -3.78416508e-01 1.08648562e+00
7.81232178e-01 7.01615140e-02 3.54849666e-01 -1.79852033e+00
-5.30419469e-01 9.29968238e-01 -4.79167342e-01 -6.28415942e-02
2.13005424e-01 3.56892310e-02 8.14052701e-01 1.26526266e-01
-1.21874101e-01 7.26533771e-01 6.23012185e-01 1.11015171e-01
-1.16291618e+00 -1.18303406e+00 1.79115131e-01 4.78439540e-01
-1.24637353e+00 -1.04634106e-01 1.14375353e+00 -3.43670905e-01
9.22835767e-01 3.38413745e-01 7.24820137e-01 8.87980819e-01
4.42435086e-01 1.01037192e+00 8.71654749e-01 -1.25142589e-01
5.44121206e-01 2.21106336e-01 2.30672117e-02 2.67766088e-01
1.55722350e-01 2.48422891e-01 -2.33072490e-02 -4.57620502e-01
4.89032678e-02 5.93738973e-01 -3.04140478e-01 -3.05946410e-01
-7.30664313e-01 6.92508638e-01 4.13169205e-01 5.31184137e-01
-5.45463204e-01 -2.06364021e-01 7.77217805e-01 3.84381264e-01
8.06140602e-01 -2.25701600e-01 -6.55080080e-01 1.92033976e-01
-1.05025280e+00 -2.05864683e-01 1.24757433e+00 8.55536222e-01
7.26218462e-01 5.94511986e-01 2.63178609e-02 3.18627089e-01
1.81279585e-01 6.20722473e-01 7.66643763e-01 -7.59009838e-01
2.09024042e-01 5.91447175e-01 1.43850356e-01 -1.04449546e+00
-7.45762408e-01 -5.42173505e-01 -1.50754046e+00 1.08625457e-01
-1.09859081e-02 -1.02638996e+00 2.65466571e-01 1.56984234e+00
2.87402838e-01 4.29868340e-01 1.46718398e-01 4.63385820e-01
4.75765496e-01 1.16531277e+00 6.27939627e-02 -7.09615648e-01
1.03907561e+00 -7.45940924e-01 -1.16690230e+00 4.42817837e-01
6.10471308e-01 -1.79199830e-01 2.85570353e-01 5.19751668e-01
-4.65877265e-01 -4.57639784e-01 -1.11568046e+00 4.23312068e-01
-5.13733625e-01 2.36522928e-01 5.10649264e-01 6.44948065e-01
-5.47581434e-01 7.72900999e-01 -6.77747965e-01 -1.53622687e-01
5.96834958e-01 5.32773197e-01 -1.10625856e-01 1.93256736e-01
-1.31041682e+00 6.10437334e-01 5.95310390e-01 -1.05164228e-02
-5.57455957e-01 -7.92609572e-01 -6.91855669e-01 4.22750622e-01
1.66704245e-02 -3.85663718e-01 1.23270142e+00 -1.04916716e+00
-1.42035949e+00 1.67882144e-02 2.27532625e-01 -6.92126811e-01
4.30504680e-01 -1.10051833e-01 -9.47173238e-01 -2.72753030e-01
-1.23797715e-01 -3.49110067e-01 8.32059085e-01 -1.06316125e+00
-1.15860856e+00 -5.29984891e-01 -2.96000302e-01 -1.85427159e-01
-6.35387957e-01 -4.12579268e-01 5.07135987e-01 -3.90399992e-01
-5.71634054e-01 -3.67489487e-01 1.19124755e-01 -5.96452057e-01
-3.88480216e-01 -5.48031986e-01 1.48384726e+00 -1.29309618e+00
1.27499831e+00 -2.42771435e+00 -4.52095538e-01 3.38896543e-01
1.64156571e-01 1.44905120e-01 2.51756221e-01 5.57013512e-01
-2.58939534e-01 -2.86108375e-01 2.46776715e-01 -3.52646083e-01
4.76934731e-01 2.71348655e-01 -4.95198190e-01 7.80358553e-01
-5.57972610e-01 1.07084262e+00 -8.00831020e-01 -2.54172444e-01
7.11860180e-01 5.03049195e-01 3.64336185e-02 4.07317221e-01
-1.32145926e-01 4.88238782e-01 -5.15325725e-01 4.68227297e-01
7.85375357e-01 -1.47637144e-01 4.15430665e-01 -5.97513735e-01
-1.83254983e-02 -1.62669346e-01 -1.14666986e+00 1.22858858e+00
-8.02539647e-01 2.55559951e-01 1.93104476e-01 -1.32244134e+00
1.04640591e+00 8.43409836e-01 1.08266819e+00 -7.35309303e-01
5.71436524e-01 3.61151323e-02 -2.44445100e-01 -3.91017526e-01
-1.17289819e-01 2.30109856e-01 -7.70436972e-02 8.29731882e-01
3.14349562e-01 5.56778550e-01 -2.81437218e-01 -3.10195804e-01
8.00593913e-01 -7.26862848e-02 5.04444778e-01 -1.76293224e-01
4.55531299e-01 -6.82264149e-01 6.77617133e-01 3.18561852e-01
-2.10461691e-01 -4.60981518e-01 1.48002386e-01 -8.26551557e-01
-7.90781140e-01 -6.19864225e-01 -2.01898180e-02 1.09790301e+00
-5.29263541e-02 -8.35686252e-02 -8.00778091e-01 -8.00923228e-01
1.32452101e-01 1.14235365e+00 -3.68505925e-01 -2.73820192e-01
-3.14020932e-01 -9.46189463e-01 2.51608014e-01 5.07159114e-01
8.58425975e-01 -1.20933402e+00 -5.62527120e-01 4.98557061e-01
-1.34685785e-01 -6.39947772e-01 -3.79093587e-01 5.15509665e-01
-2.85733551e-01 -9.08134341e-01 -3.89497578e-01 -8.35422873e-01
3.35828394e-01 -9.18994769e-02 9.00245070e-01 -2.67307609e-01
6.29825443e-02 2.36824617e-01 -1.53301701e-01 -4.63548839e-01
-1.86951607e-01 1.91481665e-01 2.17256397e-01 5.24578393e-01
7.35919416e-01 -9.99253809e-01 -4.17219460e-01 5.04250787e-02
-3.79898041e-01 -2.92239666e-01 1.20404661e-01 8.26579750e-01
2.40775108e-01 1.16398203e+00 1.00419950e+00 -5.78658640e-01
7.07044840e-01 -9.08184767e-01 -1.07040799e+00 2.34316662e-01
-8.17402303e-01 -5.14500380e-01 1.33334136e+00 -5.79443991e-01
-1.22650170e+00 1.58512369e-01 1.31430060e-01 -3.72324735e-01
-3.36184859e-01 5.11619188e-02 -7.33904898e-01 3.86452675e-02
-5.63565753e-02 5.20048022e-01 -5.24341315e-02 -7.07914889e-01
2.15330645e-01 1.01942110e+00 4.70215052e-01 -2.37378091e-01
6.38808489e-01 2.35657617e-01 4.01491001e-02 -5.50164044e-01
-5.69825649e-01 -5.25503084e-02 -4.55072373e-01 -9.54290405e-02
7.44992912e-01 -1.06181109e+00 -1.46247900e+00 8.22768152e-01
-1.11696434e+00 -9.18291211e-02 -6.94056809e-01 5.04928172e-01
-4.82603222e-01 2.35968083e-01 -7.07876563e-01 -1.05139768e+00
-9.20632958e-01 -8.90309870e-01 7.27337122e-01 2.27254018e-01
-1.63973004e-01 -1.24679196e+00 -3.36978100e-02 1.97836850e-02
5.04285634e-01 3.48760277e-01 1.15531898e+00 -1.15174627e+00
-2.62746185e-01 -5.79232097e-01 1.35877758e-01 6.58067167e-01
6.67950213e-01 -4.87616569e-01 -1.10916018e+00 -6.82385027e-01
6.29339635e-01 8.77962336e-02 -2.07578152e-01 3.34613740e-01
1.33031487e+00 -9.74673808e-01 -2.14203820e-01 8.02430868e-01
1.50632155e+00 6.07435107e-01 2.64816731e-01 8.46066047e-03
6.73193038e-01 4.13868785e-01 -1.08275190e-02 9.02218342e-01
4.96899664e-01 2.37366468e-01 5.87726712e-01 -9.12993327e-02
4.32859629e-01 -3.75258416e-01 3.40075463e-01 9.89018619e-01
2.28645608e-01 -3.13037992e-01 -2.20524684e-01 3.46899301e-01
-2.00678372e+00 -1.20239973e+00 3.35125774e-01 1.91993403e+00
3.17827225e-01 -4.11896169e-01 1.47532895e-01 5.33901155e-01
7.25750864e-01 2.09162965e-01 -1.02427220e+00 -2.05088019e-01
-7.74439424e-02 2.43731644e-02 5.89353502e-01 2.46743977e-01
-1.20841169e+00 -3.82412896e-02 5.70651579e+00 9.71026242e-01
-1.10558093e+00 4.64521289e-01 6.72571957e-01 -3.77195552e-02
2.24193051e-01 -4.85926747e-01 -4.93957967e-01 1.03134954e+00
8.45506728e-01 -7.15833962e-01 8.92302811e-01 1.22925997e+00
3.70698392e-01 3.11822861e-01 -1.24977922e+00 1.20343137e+00
-2.75861531e-01 -9.10209417e-01 -3.14790994e-01 4.17400777e-01
7.65564322e-01 1.32527277e-01 -3.09755594e-01 2.60787815e-01
5.19954383e-01 -7.57186532e-01 3.72785509e-01 6.77328825e-01
3.02878112e-01 -1.16635275e+00 1.22138560e+00 8.36593151e-01
-1.37239635e+00 -7.96882510e-01 -2.84660794e-02 -2.66075075e-01
-7.60319084e-03 7.32299209e-01 -1.69355065e-01 1.00095558e+00
9.06100094e-01 8.17043722e-01 -4.87577431e-02 5.54777980e-01
2.64704786e-02 7.41746843e-01 -4.28125173e-01 1.97273865e-01
-3.23099881e-01 -3.95162225e-01 9.02561992e-02 6.91562951e-01
4.29561913e-01 -6.41359910e-02 5.00949025e-01 6.82827294e-01
-2.08215833e-01 2.89049387e-01 -6.69948220e-01 4.14029211e-01
5.91325879e-01 1.34987116e+00 2.74513252e-02 -2.06820175e-01
-7.58446038e-01 8.58538508e-01 1.99739099e-03 3.69272202e-01
-5.58268964e-01 -3.77092421e-01 5.08583307e-01 -1.93189487e-01
3.40136498e-01 3.22616607e-01 -8.82682353e-02 -1.17136014e+00
-4.70063038e-04 -6.94868207e-01 4.97431368e-01 -5.35353124e-01
-2.04039812e+00 2.93343633e-01 -1.93810627e-01 -1.13500619e+00
-6.95409358e-01 1.59524810e-02 -1.05022776e+00 1.04265416e+00
-1.39178526e+00 -1.31719542e+00 -1.48008212e-01 1.00326872e+00
1.05330393e-01 -5.26741862e-01 1.17375028e+00 4.89683896e-01
-7.74420798e-01 5.94100714e-01 7.37102747e-01 3.59643340e-01
-7.12440461e-02 -1.03548181e+00 8.19920599e-02 7.71893084e-01
-2.33946517e-01 -2.07570478e-01 2.45982334e-01 -5.75859070e-01
-1.25989783e+00 -1.42075503e+00 5.86291730e-01 1.55937746e-01
5.30054748e-01 -3.86872798e-01 -8.11228573e-01 9.21431303e-01
6.32086515e-01 1.65168643e-01 9.05154765e-01 -1.93038017e-01
4.13388088e-02 -8.04809690e-01 -1.80113876e+00 -5.30839898e-02
2.44161963e-01 -4.38606858e-01 -3.84981513e-01 4.33369875e-01
6.56104684e-01 3.11748803e-01 -1.42366946e+00 2.30936289e-01
3.32268238e-01 -7.73225486e-01 4.73431408e-01 -1.49272829e-01
-3.19242030e-01 -2.77003407e-01 -4.43091691e-01 -1.59330928e+00
-7.31009424e-01 -6.74461961e-01 -9.95560646e-01 1.81381094e+00
-3.71312290e-01 -1.20659566e+00 6.37196839e-01 6.20143354e-01
3.82844448e-01 -4.77983326e-01 -1.11059880e+00 -7.02746749e-01
-6.10213839e-02 -1.65375397e-01 1.84147394e+00 1.17467785e+00
2.17876345e-01 1.66562617e-01 -7.73774028e-01 5.13938129e-01
9.18847680e-01 3.89852434e-01 4.98675615e-01 -1.30396485e+00
-2.09022909e-01 -1.73310965e-01 -2.80944228e-01 -4.45335686e-01
6.06180251e-01 -6.92246497e-01 -4.93114978e-01 -1.26157784e+00
-1.75195873e-01 -1.60703018e-01 -6.54492259e-01 8.36270571e-01
5.19517839e-01 -2.16071084e-01 3.47557962e-01 1.66259587e-01
-2.44928747e-01 7.48733103e-01 6.40472472e-01 -4.90267098e-01
-1.19949944e-01 3.66623789e-01 -5.28429806e-01 7.54040301e-01
1.24546349e+00 -3.41244154e-02 -5.34486175e-01 -1.83264703e-01
-4.52592641e-01 1.31666988e-01 2.33586997e-01 -1.01332200e+00
2.41234913e-01 1.73529722e-02 7.73744822e-01 -8.55147362e-01
-3.93907689e-02 -1.76108181e+00 7.09711611e-01 7.01822221e-01
1.96276039e-01 -3.11806023e-01 -1.90567542e-02 2.46482491e-01
5.86345531e-02 8.87226090e-02 4.80883896e-01 3.42175364e-02
-4.35179085e-01 5.19933164e-01 -1.94318384e-01 -5.81624627e-01
1.25646758e+00 3.17375422e-01 -2.79990852e-01 -5.59138000e-01
-6.32953882e-01 6.63807213e-01 2.91852113e-02 3.35742950e-01
-1.04291774e-01 -1.58964407e+00 -3.72350305e-01 5.29469430e-01
-3.69992763e-01 -1.16023384e-01 3.44160706e-01 3.84299189e-01
2.56728947e-01 4.46803182e-01 6.08041622e-02 -3.19340318e-01
-7.15971172e-01 1.10401523e+00 7.06532300e-01 -3.13242286e-01
-7.03272283e-01 -2.18061775e-01 9.85789225e-02 -4.85871494e-01
5.55669487e-01 -9.86143872e-02 -2.24221334e-01 3.91277045e-01
7.00677633e-01 9.67291951e-01 2.91779041e-01 -6.44483507e-01
-1.02368973e-01 2.89054513e-01 2.43900090e-01 7.73207188e-01
1.60177708e+00 -4.27034289e-01 -5.07820487e-01 4.96410280e-01
1.47970486e+00 -2.98667341e-01 -1.04900873e+00 -2.48796493e-01
-4.93958414e-01 -4.55161035e-02 3.99472713e-01 -9.75529432e-01
-1.80889773e+00 5.52038610e-01 1.08106101e+00 7.10882127e-01
1.50680947e+00 -3.68649334e-01 1.09435451e+00 1.82555392e-01
7.94998646e-01 -9.48123276e-01 -8.82563472e-01 -1.02238595e-01
2.78805077e-01 -7.79350996e-01 -7.69057795e-02 2.21146747e-01
-3.43147144e-02 9.08696175e-01 4.42387223e-01 5.05456924e-02
1.22228587e+00 5.86693645e-01 -2.14031667e-01 2.11539671e-01
-5.15892148e-01 3.56439680e-01 -2.21032798e-01 9.48512733e-01
-1.95914805e-01 4.71737504e-01 1.14196502e-01 1.32777596e+00
-2.86188632e-01 4.84168887e-01 8.84132981e-02 7.43174374e-01
7.11961687e-02 -7.74951577e-01 -5.09428620e-01 7.98062861e-01
-5.65707982e-01 3.10205907e-01 3.61408859e-01 2.28820622e-01
5.28315902e-01 1.16891873e+00 2.29549989e-01 -5.58048546e-01
2.60137439e-01 1.96599767e-01 -3.01258708e-03 7.89595693e-02
-5.65495014e-01 -8.30589309e-02 -4.94818807e-01 -4.85734820e-01
-2.02840924e-01 -3.98945272e-01 -1.09579754e+00 -8.02088320e-01
-3.50031883e-01 2.35253394e-01 4.85283047e-01 1.04903400e+00
3.59216541e-01 4.87697572e-01 1.53893435e+00 -1.09157598e+00
-7.91409075e-01 -7.89892614e-01 -1.44488192e+00 2.97770530e-01
5.69511414e-01 -1.84854209e-01 -6.16258144e-01 -2.08685309e-01] | [5.893728256225586, 2.742708683013916] |
49f55f83-96da-42ee-926c-31e8a3a1f789 | knowledge-distillation-with-error-correcting | 2204.00649 | null | https://arxiv.org/abs/2204.00649v1 | https://arxiv.org/pdf/2204.00649v1.pdf | Knowledge distillation with error-correcting transfer learning for wind power prediction | Wind power prediction, especially for turbines, is vital for the operation, controllability, and economy of electricity companies. Hybrid methodologies combining advanced data science with weather forecasting have been incrementally applied to the predictions. Nevertheless, individually modeling massive turbines from scratch and downscaling weather forecasts to turbine size are neither easy nor economical. Aiming at it, this paper proposes a novel framework with mathematical underpinnings for turbine power prediction. This framework is the first time to incorporate knowledge distillation into energy forecasting, enabling accurate and economical constructions of turbine models by learning knowledge from the well-established park model. Besides, park-scale weather forecasts non-explicitly are mapped to turbines by transfer learning of predicted power errors, achieving model correction for better performance. The proposed framework is deployed on five turbines featuring various terrains in an Arctic wind park, the results are evaluated against the competitors of ablation investigation. The major findings reveal that the proposed framework, developed on favorable knowledge distillation and transfer learning parameters tuning, yields performance boosts from 3.3 % to 23.9 % over its competitors. This advantage also exists in terms of wind energy physics and computing efficiency, which are verified by the prediction quality rate and calculation time. | ['Hao Chen'] | 2022-04-01 | null | null | null | null | ['weather-forecasting'] | ['miscellaneous'] | [-3.58441740e-01 -9.88697112e-02 2.31513567e-02 8.70386884e-02
-1.73159033e-01 -5.66387177e-01 5.11152923e-01 -8.31262916e-02
-9.72329751e-02 1.09558904e+00 -8.53524953e-02 -4.00830418e-01
-3.89251560e-01 -1.10209942e+00 -3.22876960e-01 -7.54288852e-01
-5.66555917e-01 3.30232233e-01 -1.50971964e-01 -6.95535600e-01
5.87134659e-02 5.81734240e-01 -1.71800041e+00 -3.82530510e-01
1.36310661e+00 7.95505822e-01 3.80431145e-01 7.28658557e-01
-1.71427038e-02 3.60290706e-01 -2.92429268e-01 -1.41044647e-01
4.06909645e-01 7.02859983e-02 -4.45749283e-01 -5.34404874e-01
-6.63999617e-02 -3.14901561e-01 4.16602880e-01 6.02084517e-01
7.63268948e-01 2.43216351e-01 6.27940536e-01 -8.01032543e-01
-2.79452324e-01 4.23510790e-01 -3.11555833e-01 3.14782113e-01
-2.87287712e-01 1.64822921e-01 9.11682129e-01 -7.50772059e-01
1.48159623e-01 7.29996502e-01 1.08998597e+00 -1.57373939e-02
-8.25896740e-01 -5.11003733e-01 -2.48454407e-01 1.25509471e-01
-1.30948210e+00 -2.55539566e-01 5.35868764e-01 -7.24616647e-01
1.43432069e+00 1.82494655e-01 1.20146942e+00 3.25327218e-01
3.81625831e-01 8.06766897e-02 1.10057402e+00 -4.94046032e-01
4.93337661e-01 2.50351280e-01 -3.97341043e-01 3.34565014e-01
5.59904933e-01 7.88726330e-01 -4.76013988e-01 3.68379742e-01
3.77074361e-01 -4.77840751e-01 -2.14205995e-01 -2.28140429e-01
-5.32473981e-01 8.41768324e-01 6.13961577e-01 4.26422894e-01
-5.98269224e-01 -2.37158433e-01 4.36182201e-01 1.46805212e-01
8.13803017e-01 5.63637376e-01 -9.95212078e-01 -4.50696051e-01
-1.28142476e+00 2.61096835e-01 6.99959815e-01 5.96343756e-01
6.01686776e-01 1.00570893e+00 4.58996594e-01 3.18617970e-01
1.28771514e-01 1.03438246e+00 7.00916290e-01 -3.41554999e-01
3.20200950e-01 3.58518451e-01 5.03415942e-01 -8.74034524e-01
-8.20807815e-01 -1.05047131e+00 -1.06782067e+00 4.42768961e-01
-2.23636761e-01 -7.56001949e-01 -5.78506052e-01 1.24789882e+00
3.67979884e-01 4.36825842e-01 3.27147156e-01 7.19174623e-01
4.39142287e-01 9.65112567e-01 2.50903487e-01 -3.42116863e-01
1.07461989e+00 -9.17151392e-01 -7.57053137e-01 -9.38087031e-02
9.50971603e-01 -5.63928962e-01 5.81340551e-01 2.43085146e-01
-8.45948994e-01 -9.92181063e-01 -1.22769356e+00 3.35558057e-01
-1.01253784e+00 5.11160314e-01 6.38612270e-01 9.42401469e-01
-1.03828764e+00 9.91305351e-01 -8.21270108e-01 -3.47706318e-01
1.37031540e-01 1.82023585e-01 1.93023775e-02 7.15440869e-01
-1.58535135e+00 1.50585008e+00 8.44604969e-01 6.00517213e-01
-5.64715087e-01 -1.25757849e+00 -5.53199828e-01 4.18236703e-01
-1.70961753e-01 -9.92399275e-01 1.07795286e+00 -5.58878243e-01
-1.84594619e+00 -1.35889485e-01 1.81462802e-02 -9.86523569e-01
3.45091701e-01 -6.03623867e-01 -7.63580203e-01 -1.71285048e-01
-1.28186271e-01 1.13748886e-01 1.00998330e+00 -1.06731379e+00
-1.13286304e+00 -3.72403897e-02 -2.47521698e-01 4.07271802e-01
-6.13193393e-01 -6.22631729e-01 6.90897167e-01 -4.57509369e-01
-4.19853956e-01 -6.29737914e-01 -2.20013961e-01 -6.79955065e-01
3.85545254e-01 1.58925951e-02 8.49043727e-01 -9.51349854e-01
1.25147879e+00 -1.77472746e+00 4.65875641e-02 9.82951298e-02
-1.49745807e-01 7.57487535e-01 2.72719383e-01 7.04845488e-01
-1.44210264e-01 3.35526973e-01 5.91583103e-02 -1.79212447e-02
4.13625836e-02 5.08544028e-01 -6.15822911e-01 1.10568583e-01
4.15864855e-01 8.88171613e-01 -8.19553077e-01 3.93318199e-02
8.04726303e-01 5.00451326e-01 -3.55450660e-01 3.87900323e-02
2.35132053e-02 4.52319056e-01 -4.91189688e-01 2.24543139e-01
8.25890720e-01 2.45010376e-01 -1.35222618e-02 -2.78206527e-01
-7.99276948e-01 -1.17954098e-01 -1.13786888e+00 1.35659397e+00
-1.26593840e+00 1.81223288e-01 2.25742668e-01 -1.22492909e+00
1.06953657e+00 3.99091780e-01 4.29139555e-01 -4.99751449e-01
-1.35388687e-01 5.05890489e-01 5.43868029e-03 -5.48829138e-01
6.98070288e-01 -5.46733975e-01 4.16327208e-01 -1.77576303e-01
1.52900204e-01 -7.40225255e-01 2.56212950e-01 -3.49848628e-01
2.18980134e-01 6.43763423e-01 3.82525772e-01 -5.46621919e-01
6.74643159e-01 1.64094046e-01 5.23356795e-01 2.56014854e-01
2.85199076e-01 -1.43743962e-01 -2.16152608e-01 -7.93314040e-01
-1.09146273e+00 -7.74318993e-01 -3.41010600e-01 7.22382963e-01
-2.32650384e-01 -4.50922251e-01 -4.61187214e-01 -1.88598379e-01
2.03466326e-01 8.29415679e-01 -3.35821569e-01 -3.21983796e-04
-1.84018821e-01 -8.91919017e-01 2.82892466e-01 5.48749864e-01
7.08974063e-01 -5.66021740e-01 -9.32750821e-01 2.33076781e-01
2.33627707e-01 -8.63181531e-01 5.90783954e-01 4.43144888e-01
-1.29110336e+00 -5.83038092e-01 -4.02049065e-01 -3.08394670e-01
1.50492519e-01 -7.84631371e-02 1.07816684e+00 4.19691578e-02
-1.38296545e-01 -4.35874276e-02 -7.09392130e-01 -7.97286391e-01
-1.98336601e-01 5.45446396e-01 4.71465856e-01 -4.03685689e-01
-2.10324943e-01 -7.94653952e-01 -4.71821368e-01 3.19642350e-02
-5.41052222e-01 1.22135311e-01 8.43155146e-01 1.03646863e+00
1.45230442e-01 9.25829709e-01 8.74009132e-01 -5.84261477e-01
3.57232064e-01 -6.33540571e-01 -9.19382215e-01 4.64366861e-02
-1.23986495e+00 4.56477813e-02 1.17257655e+00 3.30069125e-01
-1.51052260e+00 2.92425398e-02 -1.97307453e-01 -7.96951726e-02
-2.12370485e-01 7.89723039e-01 2.25106984e-01 -1.93439320e-01
5.51039696e-01 2.90420026e-01 -1.33973673e-01 -6.27617061e-01
5.72526217e-01 4.73503768e-01 3.37745041e-01 -3.46631914e-01
1.45422888e+00 2.20671490e-01 3.52139384e-01 -1.19230902e+00
-4.69827384e-01 -3.28171968e-01 -9.34036911e-01 -2.98274100e-01
6.16616249e-01 -1.20912218e+00 -2.41417721e-01 4.38211560e-01
-6.99366808e-01 -2.23484769e-01 -4.61013854e-01 6.80162370e-01
-3.67240995e-01 1.42631978e-01 2.36621536e-02 -1.03732061e+00
-7.60800838e-01 -7.32445419e-01 6.24407589e-01 4.96214390e-01
2.36792624e-01 -1.48778319e+00 7.07518980e-02 -3.21208611e-02
1.15810823e+00 2.23843232e-01 7.92586327e-01 -2.35776186e-01
-9.80938375e-02 -2.46099353e-01 1.64436191e-01 5.32278121e-01
1.87974796e-01 2.71717995e-01 -1.07560110e+00 -2.91418195e-01
-1.19158626e-01 -1.69612676e-01 4.87013936e-01 3.93718362e-01
4.41530645e-01 -9.04875472e-02 -2.87267745e-01 6.18473589e-01
1.67268455e+00 5.52300401e-02 2.49885246e-01 4.45265412e-01
3.17249447e-01 3.30797583e-01 7.85874844e-01 7.70935059e-01
1.98004901e-01 4.41007376e-01 6.47671938e-01 -3.27825367e-01
2.58385055e-02 -2.51001209e-01 2.57761657e-01 1.12368691e+00
-8.05410147e-01 1.19959682e-01 -8.92551482e-01 6.43968523e-01
-1.54773307e+00 -1.00564694e+00 -1.97853342e-01 2.20573187e+00
2.77482361e-01 2.35597745e-01 -2.72540331e-01 1.46867990e-01
3.78059983e-01 1.89713985e-01 -2.01958969e-01 -7.93783486e-01
9.12811756e-02 7.55995154e-01 8.09981704e-01 5.92368424e-01
-1.16257477e+00 9.42553759e-01 5.88435841e+00 9.99978721e-01
-1.34029865e+00 1.56089067e-01 7.47560784e-02 1.44436732e-01
-5.88311739e-02 1.09088197e-01 -8.30633402e-01 3.00161690e-01
1.44755590e+00 -5.87692082e-01 4.56027478e-01 9.71573055e-01
8.49259257e-01 -1.90604970e-01 -3.75084907e-01 4.69283551e-01
-5.86803675e-01 -1.34468687e+00 9.97864157e-02 -3.22749093e-02
8.52435350e-01 1.17094874e-01 -1.36065170e-01 6.62195683e-01
1.21760681e-01 -9.25322831e-01 5.41680515e-01 7.36003876e-01
4.16286021e-01 -8.01318347e-01 1.38035214e+00 4.61972356e-01
-1.75165868e+00 -3.67753953e-01 -4.83157724e-01 -8.16710293e-01
3.35385591e-01 9.20537949e-01 -1.06876767e+00 1.73912740e+00
1.02548456e+00 9.39467549e-01 -3.66512060e-01 8.57819259e-01
-1.36981681e-01 8.15706968e-01 -5.65081239e-01 8.39754567e-02
1.63538873e-01 -4.94187325e-01 2.56602973e-01 8.91127586e-01
9.49908614e-01 -1.27310932e-01 1.32107306e-02 4.39242542e-01
4.78062034e-01 3.75891238e-01 -8.14589918e-01 9.89014730e-02
6.29110754e-01 1.76867855e+00 -1.79865599e-01 -2.87474185e-01
-2.80069351e-01 2.68745720e-01 -6.38801828e-02 1.65412694e-01
-8.58540714e-01 -3.92538458e-01 6.84037983e-01 7.64518138e-03
5.78262389e-01 -2.83143759e-01 -4.58598763e-01 -6.79917991e-01
-4.91691902e-02 -8.59665871e-02 3.03423442e-02 -9.40106571e-01
-1.15947270e+00 5.19404113e-01 9.62972715e-02 -1.56931710e+00
-4.96514052e-01 -9.43733335e-01 -8.52188766e-01 1.35436475e+00
-2.19662857e+00 -1.42707884e+00 -2.75476843e-01 -3.61328907e-02
4.65322822e-01 -3.20290267e-01 1.25143218e+00 1.18820891e-01
-5.35174191e-01 -6.43527210e-02 6.53427124e-01 -4.80692446e-01
2.18941405e-01 -1.53182554e+00 4.24229801e-01 8.06068122e-01
-2.23147362e-01 1.90920204e-01 7.44800866e-01 -5.50498486e-01
-1.42364275e+00 -1.20918202e+00 7.93767452e-01 -3.38698179e-02
9.87649500e-01 1.53628290e-01 -7.34663725e-01 1.33194193e-01
5.29245496e-01 1.19038433e-01 4.20151860e-01 2.41579801e-01
2.37025425e-01 -6.15555108e-01 -9.49268937e-01 2.40465358e-01
5.92240810e-01 -1.01747073e-01 -6.37176037e-01 2.78162748e-01
2.12294906e-01 -3.90169531e-01 -1.59916890e+00 8.25047195e-01
6.01722360e-01 -9.11402225e-01 7.72230744e-01 -5.66223919e-01
4.05735999e-01 -4.36118066e-01 2.98037827e-01 -2.01065922e+00
-4.86101478e-01 -5.60271859e-01 -2.61131495e-01 1.31986880e+00
6.15979552e-01 -7.24321008e-01 5.44815063e-01 1.58712983e-01
-3.78280699e-01 -6.23410106e-01 -1.25190246e+00 -8.28799129e-01
6.21004581e-01 -2.87904143e-01 7.19405293e-01 8.76368046e-01
-4.68974672e-02 2.77696073e-01 -3.70493501e-01 7.75335073e-01
1.04824081e-01 1.72877446e-01 7.57244229e-01 -1.57487249e+00
-4.47734147e-02 -4.08651054e-01 -2.52616733e-01 -4.54943717e-01
-1.69848232e-03 -6.76783383e-01 -1.41271695e-01 -1.48924279e+00
-7.43334472e-01 -3.92167121e-01 -3.76747213e-02 4.62422967e-01
-1.27406642e-01 -6.46990463e-02 5.47778010e-02 6.34671673e-02
4.17386085e-01 9.39694524e-01 1.07197380e+00 8.24300796e-02
-1.90543503e-01 1.52086228e-01 -5.26389852e-02 7.37340569e-01
1.35848665e+00 7.90433139e-02 -6.06367588e-01 -3.96616340e-01
4.40622389e-01 -7.82622248e-02 8.78732204e-02 -1.55400991e+00
7.32172467e-03 8.82437974e-02 4.58795279e-01 -5.26075065e-01
7.50506893e-02 -1.03300643e+00 4.95254666e-01 7.34525681e-01
4.91714895e-01 1.40144348e-01 7.20887184e-01 4.01290655e-01
-2.98254430e-01 -2.81518131e-01 4.17045623e-01 3.36931497e-02
-1.20082855e+00 -1.34551123e-01 -4.67920542e-01 -3.54739487e-01
1.07536602e+00 -1.43436387e-01 -4.03813154e-01 1.94807295e-02
-8.51244926e-01 3.94993693e-01 -1.13251917e-02 3.26042712e-01
1.77822500e-01 -8.69554758e-01 -7.93252289e-01 3.54751587e-01
-2.70709157e-01 6.59615695e-02 6.19204223e-01 8.90869915e-01
-6.97923124e-01 6.66110754e-01 -3.01478475e-01 -4.09650087e-01
-7.94463098e-01 3.46140057e-01 6.15705669e-01 -6.08144283e-01
-5.16102314e-01 3.89045537e-01 -5.06940603e-01 -6.26884222e-01
-6.56338215e-01 -6.16326511e-01 -4.12017822e-01 2.25777179e-01
9.95073393e-02 4.74291563e-01 7.08165824e-01 -4.57988143e-01
-7.06284344e-02 9.22525227e-01 4.73230481e-01 2.63735175e-01
1.39651322e+00 -3.22806239e-01 8.08325931e-02 3.09877276e-01
4.35136259e-01 1.12509176e-01 -1.08794117e+00 2.89238185e-01
-4.21685837e-02 -2.12946743e-01 6.92966998e-01 -1.18332887e+00
-1.21808505e+00 9.61532652e-01 9.32207525e-01 4.60079640e-01
1.23740959e+00 -8.66119027e-01 5.64530849e-01 3.87262225e-01
4.87448007e-01 -1.16904676e+00 -7.25886464e-01 5.68033874e-01
7.52572954e-01 -9.81601119e-01 1.81309387e-01 -1.56402305e-01
-5.27721107e-01 1.28500056e+00 5.45820177e-01 6.34798408e-03
9.25255537e-01 4.80525523e-01 5.48198633e-02 1.12694822e-01
-6.38218105e-01 -4.33139771e-01 -1.12886496e-01 8.25266957e-01
2.62904584e-01 2.97557265e-01 -3.92887443e-01 8.08322728e-01
-5.71841896e-01 3.64556611e-01 2.63213128e-01 7.97144294e-01
-6.01188123e-01 -9.58401322e-01 -3.42678219e-01 4.62994277e-01
-9.70643833e-02 -4.15920138e-01 4.72270548e-01 9.19707179e-01
5.03041387e-01 8.46069038e-01 -6.60624057e-02 -4.41551387e-01
4.32194412e-01 2.10703194e-01 -6.02875762e-02 -3.16934019e-01
-6.46491230e-01 -2.37556309e-01 1.99992403e-01 2.78194547e-02
-5.53014398e-01 -3.56730849e-01 -1.02210414e+00 -3.37845147e-01
-5.36590934e-01 8.52130473e-01 8.67603779e-01 1.00565612e+00
5.10850608e-01 8.95194650e-01 9.94183242e-01 -1.30579269e+00
-8.62370193e-01 -1.25351679e+00 -9.12094355e-01 -2.70737976e-01
4.00628783e-02 -1.07578862e+00 -5.47899246e-01 1.21361099e-01] | [6.311424732208252, 2.880448818206787] |
fc2ae55f-d9d0-414f-a27e-86fe7d5dc13f | garmentnets-category-level-pose-estimation | 2104.05177 | null | https://arxiv.org/abs/2104.05177v2 | https://arxiv.org/pdf/2104.05177v2.pdf | GarmentNets: Category-Level Pose Estimation for Garments via Canonical Space Shape Completion | This paper tackles the task of category-level pose estimation for garments. With a near infinite degree of freedom, a garment's full configuration (i.e., poses) is often described by the per-vertex 3D locations of its entire 3D surface. However, garments are also commonly subject to extreme cases of self-occlusion, especially when folded or crumpled, making it challenging to perceive their full 3D surface. To address these challenges, we propose GarmentNets, where the key idea is to formulate the deformable object pose estimation problem as a shape completion task in the canonical space. This canonical space is defined across garments instances within a category, therefore, specifies the shared category-level pose. By mapping the observed partial surface to the canonical space and completing it in this space, the output representation describes the garment's full configuration using a complete 3D mesh with the per-vertex canonical coordinate label. To properly handle the thin 3D structure presented on garments, we proposed a novel 3D shape representation using the generalized winding number field. Experiments demonstrate that GarmentNets is able to generalize to unseen garment instances and achieve significantly better performance compared to alternative approaches. | ['Shuran Song', 'Cheng Chi'] | 2021-04-12 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Chi_GarmentNets_Category-Level_Pose_Estimation_for_Garments_via_Canonical_Space_Shape_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Chi_GarmentNets_Category-Level_Pose_Estimation_for_Garments_via_Canonical_Space_Shape_ICCV_2021_paper.pdf | iccv-2021-1 | ['3d-shape-representation'] | ['computer-vision'] | [ 1.61333784e-01 8.54954496e-02 6.06820136e-02 -2.07125813e-01
-6.20265007e-01 -8.13482702e-01 3.44185024e-01 -2.47931123e-01
2.32930481e-01 2.80986667e-01 -1.80016905e-01 1.78496048e-01
-1.83960915e-01 -7.83996224e-01 -9.78338122e-01 -6.74526334e-01
-3.59465368e-02 1.06306612e+00 -1.10906258e-01 -1.12818740e-01
8.68961588e-03 8.84070694e-01 -1.42147160e+00 -2.54451372e-02
5.97373724e-01 9.38667953e-01 1.69988886e-01 3.48902285e-01
9.13093165e-02 -5.05025387e-01 -5.09714544e-01 -2.25193441e-01
4.37464148e-01 1.54704466e-01 -5.43938637e-01 8.39345753e-01
8.01991284e-01 -5.55717759e-02 -2.01713637e-01 9.75421131e-01
1.16083138e-01 2.23245189e-01 8.69541883e-01 -1.09000206e+00
-5.15801549e-01 2.27911565e-02 -2.73584366e-01 -5.81585407e-01
5.51537573e-01 -1.46202832e-01 8.70459378e-01 -1.26625943e+00
8.60678732e-01 1.56514549e+00 7.04380810e-01 3.80859464e-01
-1.38413513e+00 -2.30946034e-01 3.90356839e-01 -2.54012167e-01
-1.60908425e+00 4.01726067e-02 1.14746237e+00 -4.48450476e-01
5.81630468e-01 5.31803429e-01 7.76553273e-01 8.53634000e-01
2.27995455e-01 8.08602452e-01 8.49952817e-01 -2.62385249e-01
3.03861082e-01 -4.86446470e-01 2.82366350e-02 6.17717206e-01
2.22645521e-01 -4.50387269e-01 -4.46264334e-02 -3.49181324e-01
1.20894516e+00 4.19353515e-01 -1.08305566e-01 -1.04930711e+00
-1.30853319e+00 5.08094430e-01 4.37794238e-01 7.60180801e-02
-3.94513220e-01 1.29349947e-01 -9.14919972e-02 -7.82625973e-02
6.38736248e-01 3.40884596e-01 -5.23200333e-01 1.30625248e-01
-3.36134642e-01 8.42366159e-01 9.02290702e-01 1.37852454e+00
9.12380099e-01 -1.56833172e-01 2.75679469e-01 6.94082379e-01
4.06337142e-01 6.56639874e-01 -4.78489518e-01 -7.40009844e-01
5.85914552e-01 1.02421498e+00 1.24636225e-01 -1.06606150e+00
-5.36114097e-01 -6.50604427e-01 -8.89127672e-01 1.31087899e-01
2.88269371e-01 3.59343618e-01 -1.26362824e+00 1.51036334e+00
7.00166166e-01 6.01169430e-02 -2.05225199e-01 1.10759759e+00
7.22963870e-01 4.02186155e-01 -2.26839378e-01 1.19011246e-01
1.59525037e+00 -5.19927859e-01 -2.75071472e-01 -1.49545640e-01
5.54228760e-02 -7.38606036e-01 8.25759709e-01 3.34383130e-01
-1.14417005e+00 -4.69840884e-01 -1.01079690e+00 5.27797118e-02
-7.25817010e-02 1.77591801e-01 4.77014452e-01 3.76242876e-01
-6.35494053e-01 6.45695865e-01 -1.09771216e+00 -2.38745108e-01
2.05113456e-01 4.66787785e-01 -5.71736157e-01 -2.93624818e-01
-6.65859222e-01 7.08681405e-01 1.69300497e-01 2.86360621e-01
-7.67589152e-01 -6.81673765e-01 -1.00479555e+00 -2.94965774e-01
6.45881832e-01 -8.43141973e-01 9.41591501e-01 -2.16320291e-01
-8.87276053e-01 8.14133227e-01 -1.54946387e-01 2.93547243e-01
6.29128039e-01 -1.51066214e-01 -1.21075563e-01 -4.43060398e-02
6.82921633e-02 2.20675562e-02 1.13054144e+00 -1.72494137e+00
1.32011414e-01 -9.85336125e-01 2.72969216e-01 3.06038290e-01
2.10682958e-01 -6.45113766e-01 -6.81773305e-01 -8.26950014e-01
1.22993743e+00 -1.28075111e+00 -2.62828588e-01 1.44834131e-01
-4.77150023e-01 -2.95658886e-01 1.07711041e+00 -6.79651678e-01
7.35286236e-01 -2.07018709e+00 6.51332617e-01 4.69607383e-01
3.49484116e-01 -6.80593327e-02 6.67895423e-03 4.53109860e-01
2.03733761e-02 3.42532396e-02 -5.76654017e-01 -6.16162658e-01
2.19735563e-01 6.09380841e-01 7.27247670e-02 7.55853474e-01
3.20864350e-01 8.47862959e-01 -8.52850854e-01 -1.63266554e-01
3.11953485e-01 6.93098187e-01 -4.71214533e-01 -3.99460942e-02
-3.66839886e-01 6.20813906e-01 -9.27839696e-01 1.17357981e+00
1.01826191e+00 -7.80769885e-02 5.66418953e-02 -6.23654366e-01
5.92892841e-02 -1.50579020e-01 -1.68330228e+00 1.99986470e+00
-4.54083294e-01 -1.38697982e-01 4.31604743e-01 -7.79887617e-01
1.29550350e+00 2.49508709e-01 5.23972511e-01 -3.42228226e-02
1.83903740e-03 2.23784909e-01 -4.59460557e-01 -2.98674256e-01
5.16507864e-01 -1.39702961e-01 -5.04011214e-01 2.39167914e-01
-1.88247904e-01 -5.53089142e-01 -4.05373931e-01 -1.99986294e-01
7.64420450e-01 3.69112611e-01 4.95549850e-02 -2.96651572e-01
3.33172590e-01 -1.68747872e-01 4.43659514e-01 2.70582139e-01
3.95240158e-01 8.80677164e-01 2.81897426e-01 -6.88719094e-01
-1.24429679e+00 -1.55085051e+00 -4.65483487e-01 4.20759052e-01
4.45319057e-01 -4.25812513e-01 -6.99783444e-01 -5.95430434e-01
5.02696097e-01 3.29582006e-01 -6.25811219e-01 -1.83637266e-03
-9.02399480e-01 -4.11029488e-01 -8.66623893e-02 4.66118336e-01
2.99930483e-01 -6.97386980e-01 -5.29364645e-01 3.03885788e-01
-2.92121112e-01 -9.98630643e-01 -6.80448353e-01 -1.25530541e-01
-8.99588466e-01 -1.17569208e+00 -7.79158771e-01 -8.63460183e-01
1.03615022e+00 9.04702246e-02 9.78402734e-01 3.26768830e-02
-5.70872068e-01 6.61033511e-01 -3.33138019e-01 -1.57511570e-02
-2.57591635e-01 -9.21456292e-02 2.90331960e-01 2.97180474e-01
-6.85566217e-02 -7.14160621e-01 -5.77946961e-01 6.48959279e-01
-8.79668534e-01 1.35789895e-02 2.71222562e-01 4.90391940e-01
1.14799142e+00 5.84301315e-02 2.59943634e-01 -3.89134109e-01
3.64927924e-03 -2.31968224e-01 -2.48430863e-01 1.13308661e-01
9.37648416e-02 2.61445772e-02 2.73745954e-01 -6.45499706e-01
-5.95659018e-01 2.60613948e-01 -1.28930241e-01 -7.41009593e-01
-2.49930114e-01 2.57538944e-01 -5.28770924e-01 -1.37836695e-01
1.34376004e-01 2.70005837e-02 2.35761497e-02 -1.13403487e+00
1.94998249e-01 3.96845043e-01 5.00147700e-01 -1.02156854e+00
1.02404618e+00 5.43851316e-01 4.79875594e-01 -9.25396800e-01
-6.30429327e-01 -3.83975536e-01 -1.06416261e+00 -4.09590095e-01
9.99547899e-01 -7.42313862e-01 -9.24010634e-01 5.10289907e-01
-1.23986709e+00 8.80612582e-02 -2.30009764e-01 3.33883792e-01
-7.62800038e-01 3.20556343e-01 -1.97799623e-01 -8.19312692e-01
-6.21320754e-02 -1.26211560e+00 1.66412699e+00 -2.96484143e-01
-2.72364348e-01 -8.80372822e-01 -2.09555998e-01 2.88954228e-01
8.11866075e-02 1.11961579e+00 1.06159580e+00 -1.37028202e-01
-8.26980531e-01 -3.99873704e-01 2.24471420e-01 2.85573691e-01
1.66129708e-01 -3.34548086e-01 -6.25571132e-01 -4.63963568e-01
4.52122502e-02 6.35689422e-02 5.16308248e-01 2.03937396e-01
1.08969593e+00 -1.55001104e-01 -3.90411466e-01 6.18041098e-01
1.45616627e+00 -1.56496987e-01 2.48233289e-01 -1.01679657e-02
9.35705483e-01 6.05989277e-01 8.30575764e-01 4.74265426e-01
4.41039771e-01 1.01376271e+00 9.47477996e-01 1.43290818e-01
-1.46049991e-01 -4.49856579e-01 2.91884020e-02 7.67045379e-01
-5.04731476e-01 4.16253097e-02 -8.28609288e-01 3.97960842e-01
-1.65407014e+00 -4.04724061e-01 -2.70148516e-01 2.28243184e+00
2.78598905e-01 -8.58362988e-02 -5.73673770e-02 1.07709296e-01
7.97610044e-01 1.51336655e-01 -8.53578806e-01 2.25013625e-02
6.71020523e-03 1.92420200e-01 3.39751989e-01 3.51319224e-01
-9.81651187e-01 5.10917425e-01 5.86704063e+00 6.86495543e-01
-6.39447212e-01 -1.33681878e-01 6.07269406e-02 2.80405641e-01
-4.18352365e-01 -2.49014944e-02 -8.38089645e-01 1.80092379e-01
2.15876177e-02 1.32812425e-01 3.30782861e-01 8.13453257e-01
-2.02165574e-01 2.75932610e-01 -1.24259543e+00 9.95700061e-01
2.70591408e-01 -1.00364208e+00 2.03401491e-01 2.75644004e-01
6.89381957e-01 -3.74607980e-01 3.59735563e-02 7.18067884e-02
-3.34662437e-01 -9.22839165e-01 1.06418383e+00 5.98566234e-01
1.31391609e+00 -7.07303524e-01 1.99323446e-01 5.34006357e-01
-1.67910933e+00 2.19518363e-01 -3.36455077e-01 -4.79247235e-02
3.60668063e-01 2.73560345e-01 -7.36052573e-01 8.66305530e-01
3.74147385e-01 6.92965508e-01 -5.55215329e-02 9.00588930e-01
1.76600397e-01 1.14064962e-01 -4.17283386e-01 1.83118045e-01
2.02872958e-02 -4.88352001e-01 1.06807947e+00 6.73223674e-01
5.73626518e-01 2.51259297e-01 6.57224357e-01 9.71363902e-01
-8.49886090e-02 -2.14287475e-01 -3.84725243e-01 2.51360387e-01
2.00615823e-01 1.19815814e+00 -8.32405567e-01 -5.52852638e-02
-4.20429111e-02 7.56987095e-01 -2.06427738e-01 3.49147171e-01
-4.30652380e-01 3.21482718e-02 1.03311419e+00 3.86638284e-01
3.17914486e-01 -5.63205838e-01 -2.22010568e-01 -1.17380726e+00
4.79528904e-01 -4.48032826e-01 -5.20704910e-02 -7.39945233e-01
-1.22330487e+00 4.37702477e-01 4.24857736e-01 -1.39558566e+00
1.25865266e-01 -7.20759392e-01 -2.59298712e-01 8.62920403e-01
-9.34321880e-01 -1.57144499e+00 -3.90076637e-01 5.14967501e-01
6.63323700e-01 3.60923290e-01 1.07437420e+00 1.43120512e-01
-1.54895876e-02 2.81976342e-01 -1.49338260e-01 -1.92841500e-01
7.55226910e-02 -1.21269500e+00 7.62370169e-01 2.65659750e-01
-8.72119889e-02 4.66965467e-01 7.00469971e-01 -9.45317864e-01
-2.11800861e+00 -1.29153895e+00 5.18627584e-01 -7.32144833e-01
2.92811900e-01 -8.10932696e-01 -1.00574505e+00 8.82411242e-01
-9.04426396e-01 1.23163015e-02 5.80420531e-02 1.51784741e-03
-3.14853460e-01 1.77913234e-01 -1.40980601e+00 3.89713019e-01
1.57580769e+00 -3.17071199e-01 -8.62377346e-01 4.47452247e-01
6.94826424e-01 -1.02557087e+00 -1.30355203e+00 6.26438618e-01
6.00482225e-01 -2.06706524e-01 1.35322988e+00 -3.36726427e-01
3.75187024e-02 -4.81379330e-01 -4.66492385e-01 -1.22701359e+00
-2.76182234e-01 -5.57844043e-01 -1.01928495e-01 8.46411407e-01
-2.20727980e-01 -4.97970939e-01 1.00195456e+00 4.82117057e-01
-5.80100119e-01 -1.00479758e+00 -1.28412652e+00 -7.29213238e-01
-1.08371340e-01 -6.57626510e-01 1.04752445e+00 7.21577823e-01
-5.79819918e-01 -1.39772847e-01 -2.70622700e-01 6.06878340e-01
1.05620944e+00 4.89704311e-01 6.97860122e-01 -1.46292448e+00
-8.70071948e-02 1.29638866e-01 -9.49774563e-01 -1.28232598e+00
1.04851015e-01 -1.15087509e+00 -1.88177060e-02 -1.61782086e+00
1.34956568e-01 -7.75286555e-01 3.21289212e-01 3.05979133e-01
3.00428540e-01 3.97980034e-01 2.26593748e-01 4.80943732e-02
-2.20460281e-01 4.61311311e-01 1.84169078e+00 -1.51696742e-01
-1.73054710e-01 2.63777405e-01 -2.58210331e-01 8.80423069e-01
3.63962620e-01 -1.30243912e-01 1.67946927e-02 -5.31508982e-01
1.27666876e-01 2.78682590e-01 6.98404312e-01 -7.86368370e-01
-1.61250338e-01 -2.65816301e-01 3.19146633e-01 -1.28442693e+00
8.49269688e-01 -1.14695597e+00 7.60494888e-01 2.27601379e-01
1.40958428e-01 -4.13643606e-02 -3.41620110e-02 6.27773106e-01
4.82370369e-02 6.47997707e-02 4.87032861e-01 -2.92109936e-01
-3.84228736e-01 7.75867105e-01 2.59610027e-01 -2.76115090e-01
1.10726178e+00 -5.44482350e-01 2.32286990e-01 1.02545083e-01
-1.34642959e+00 2.25672960e-01 1.07130015e+00 6.40854776e-01
9.64364290e-01 -1.61562610e+00 -7.65935481e-01 7.17283428e-01
2.19650105e-01 8.78577948e-01 5.12147367e-01 4.50713873e-01
-3.48143488e-01 2.41546214e-01 -2.75876429e-02 -1.11912906e+00
-1.20092249e+00 2.70248979e-01 2.42841825e-01 9.81034264e-02
-1.02076828e+00 5.61692953e-01 3.94931346e-01 -6.38832152e-01
8.09146166e-02 -5.63811302e-01 8.64464492e-02 -9.88251343e-02
1.94955692e-01 4.61163074e-01 4.15355444e-01 -1.06230557e+00
-4.03860658e-01 1.31575286e+00 2.78828949e-01 2.50284821e-01
1.54123843e+00 -1.18778907e-02 -2.19654441e-01 4.92305279e-01
1.38765311e+00 -5.19827567e-02 -1.46309209e+00 -1.73538297e-01
-4.74362195e-01 -7.47855067e-01 -5.05097866e-01 -4.07127470e-01
-7.95424104e-01 4.48058665e-01 2.31686085e-01 -1.80888660e-02
7.00176001e-01 6.87879264e-01 7.31838465e-01 2.07931995e-01
9.34107184e-01 -5.65048814e-01 4.29284684e-02 4.38111722e-01
1.41471934e+00 -6.65561020e-01 -4.64805700e-02 -8.28841388e-01
-1.84467271e-01 1.33019233e+00 1.66749895e-01 -5.96244872e-01
6.47326648e-01 3.35626937e-02 -3.58873457e-01 -4.73615736e-01
-2.35207111e-01 2.07697913e-01 7.45972335e-01 6.68804228e-01
-1.16564676e-01 5.80981791e-01 -2.99691828e-03 5.41000664e-01
-3.91231567e-01 -5.37613273e-01 1.79115146e-01 9.91293073e-01
-3.00082833e-01 -1.14272666e+00 -9.11980927e-01 3.96090209e-01
-1.13308601e-01 6.36055827e-01 -7.63959289e-02 8.37300539e-01
3.27040046e-01 5.19581914e-01 3.60935986e-01 -2.98092157e-01
9.26822662e-01 6.28399253e-02 8.27160180e-01 -9.02779937e-01
-6.40475303e-02 1.31557241e-01 -1.14629574e-01 -5.18040895e-01
-3.87071490e-01 -8.57253730e-01 -1.24615920e+00 -4.57272008e-02
-1.77828372e-01 -1.68316185e-01 1.09627104e+00 9.15177226e-01
3.34144868e-02 2.62139678e-01 5.56296110e-01 -1.45470679e+00
-7.51269996e-01 -6.62880123e-01 -9.14168656e-01 7.19042659e-01
5.46121180e-01 -1.24342501e+00 -3.53775620e-01 -2.44297348e-02] | [8.308576583862305, -3.187530994415283] |
96eb5937-0aed-44c9-9a6b-a18c3385d37c | deep-reinforcement-learning-for-cost | 2302.10261 | null | https://arxiv.org/abs/2302.10261v2 | https://arxiv.org/pdf/2302.10261v2.pdf | Deep Reinforcement Learning for Cost-Effective Medical Diagnosis | Dynamic diagnosis is desirable when medical tests are costly or time-consuming. In this work, we use reinforcement learning (RL) to find a dynamic policy that selects lab test panels sequentially based on previous observations, ensuring accurate testing at a low cost. Clinical diagnostic data are often highly imbalanced; therefore, we aim to maximize the $F_1$ score instead of the error rate. However, optimizing the non-concave $F_1$ score is not a classic RL problem, thus invalidates standard RL methods. To remedy this issue, we develop a reward shaping approach, leveraging properties of the $F_1$ score and duality of policy optimization, to provably find the set of all Pareto-optimal policies for budget-constrained $F_1$ score maximization. To handle the combinatorially complex state space, we propose a Semi-Model-based Deep Diagnosis Policy Optimization (SM-DDPO) framework that is compatible with end-to-end training and online learning. SM-DDPO is tested on diverse clinical tasks: ferritin abnormality detection, sepsis mortality prediction, and acute kidney injury diagnosis. Experiments with real-world data validate that SM-DDPO trains efficiently and identifies all Pareto-front solutions. Across all tasks, SM-DDPO is able to achieve state-of-the-art diagnosis accuracy (in some cases higher than conventional methods) with up to $85\%$ reduction in testing cost. The code is available at [https://github.com/Zheng321/Deep-Reinforcement-Learning-for-Cost-Effective-Medical-Diagnosis]. | ['Mengdi Wang', 'Yuan Luo', 'Kaixuan Huang', 'Joseph Kim', 'Yikuan Li', 'Zheng Yu'] | 2023-02-20 | null | null | null | null | ['mortality-prediction', 'medical-diagnosis'] | ['medical', 'medical'] | [-4.50560451e-02 5.62738106e-02 -2.96161085e-01 -3.09932441e-01
-9.26558733e-01 -2.78161675e-01 -2.76678771e-01 4.07347381e-01
-2.51622885e-01 9.59435046e-01 -3.16090524e-01 -6.46521926e-01
-6.49554431e-01 -6.72982216e-01 -7.10311770e-01 -6.95743084e-01
-4.00037706e-01 1.05257773e+00 -3.57131511e-01 3.74302417e-01
-1.84343979e-02 2.13085011e-01 -1.23225296e+00 2.55752020e-02
1.26977491e+00 1.02731085e+00 1.57504782e-01 7.05253899e-01
2.25824729e-01 7.60235548e-01 -4.27387893e-01 -2.75214642e-01
3.03385139e-01 -7.16098249e-01 -5.71972549e-01 1.01509467e-01
-3.94618094e-01 -3.81743819e-01 1.65693179e-01 8.28037977e-01
8.10346067e-01 2.06116810e-02 3.15721691e-01 -1.25179887e+00
-1.70217752e-01 5.05541861e-01 -4.74998623e-01 1.21315911e-01
1.91524670e-01 6.20427787e-01 8.79995048e-01 -1.99312285e-01
3.74787450e-01 1.01676190e+00 4.24093753e-01 7.44445801e-01
-1.20433497e+00 -4.00019288e-01 2.49159649e-01 8.75246711e-03
-1.09159601e+00 8.08488131e-02 2.84036487e-01 -4.19509470e-01
8.60978186e-01 2.50428081e-01 8.97157073e-01 1.02037120e+00
4.72094983e-01 1.00600755e+00 1.22516239e+00 -1.45592779e-01
4.20595318e-01 -1.89308375e-01 -2.12452024e-01 9.70690012e-01
2.88326859e-01 2.39791870e-01 -2.70775050e-01 -3.38187814e-01
7.28480101e-01 3.39470237e-01 -2.04473317e-01 -3.89053702e-01
-1.14856994e+00 7.95330167e-01 -6.52856054e-03 -9.67239812e-02
-7.90314198e-01 1.74014121e-01 3.32033485e-01 4.49650377e-01
2.76578546e-01 7.15989590e-01 -7.39566863e-01 -3.61481547e-01
-6.47686899e-01 3.67575467e-01 7.74531484e-01 3.45849365e-01
2.08085179e-01 5.99130131e-02 -5.61058760e-01 6.52036548e-01
1.35561436e-01 7.49862373e-01 2.58165240e-01 -1.04784369e+00
2.65454680e-01 6.58043027e-01 5.24918318e-01 -3.91287357e-01
-7.34195232e-01 -9.40377593e-01 -7.95853198e-01 3.41082998e-02
3.89639705e-01 -6.81544721e-01 -7.62907147e-01 1.85311556e+00
5.35169721e-01 1.95924059e-01 -5.19678779e-02 1.01818955e+00
-5.93584925e-02 3.40520442e-01 9.95148718e-02 -7.08296537e-01
1.30305851e+00 -5.22824526e-01 -4.96680856e-01 -9.88385752e-02
6.46734297e-01 -2.56811559e-01 1.03610265e+00 4.93840277e-01
-1.25388825e+00 6.18084483e-02 -6.70584738e-01 8.01342309e-01
1.79995492e-01 6.04679696e-02 4.74169105e-01 4.64989066e-01
-7.53818750e-01 7.17708468e-01 -1.14721489e+00 -5.39287068e-02
5.85242212e-01 6.33359253e-01 2.55126119e-01 -2.68880367e-01
-1.07196999e+00 6.92549765e-01 2.01332048e-01 3.15685682e-02
-1.11534643e+00 -1.09368801e+00 -5.97491264e-01 1.62864968e-01
8.51729035e-01 -1.09134674e+00 1.45372713e+00 -7.82522619e-01
-1.64432657e+00 5.97795188e-01 7.31641576e-02 -5.75784147e-01
8.82057488e-01 -1.50649309e-01 -1.02016985e-01 1.87919989e-01
-4.26176824e-02 4.34210718e-01 6.40187681e-01 -7.82592654e-01
-7.66940594e-01 -3.23340267e-01 -8.18755478e-02 2.11738661e-01
-1.48008943e-01 -1.34246126e-01 3.17696780e-02 -3.69113266e-01
-4.31267142e-01 -8.66845667e-01 -4.04610574e-01 -4.03800964e-01
-3.39730293e-01 -3.69854659e-01 2.48705462e-01 -6.56410396e-01
1.16166008e+00 -1.74294937e+00 1.26576379e-01 1.58902124e-01
8.53935480e-02 3.18038255e-01 -1.00816183e-01 2.05326855e-01
1.52006447e-01 -5.29578589e-02 -2.92860448e-01 -1.26894861e-01
1.45946324e-01 1.84071422e-01 5.30122854e-02 4.86498624e-01
5.66667497e-01 1.07779098e+00 -1.19431198e+00 -2.75899351e-01
1.85621142e-01 2.19037503e-01 -7.75277734e-01 5.21155119e-01
-6.92408144e-01 6.43864393e-01 -7.48595834e-01 9.62119401e-01
3.34129065e-01 -8.06562126e-01 5.79035342e-01 4.75735635e-01
2.52737224e-01 -8.76615718e-02 -1.07820833e+00 1.23581696e+00
-3.67325097e-01 -1.27851576e-01 1.25809561e-03 -1.34414017e+00
6.51195586e-01 2.83482492e-01 1.17377961e+00 -7.88048923e-01
2.93412268e-01 3.00678909e-01 1.21425092e-01 -6.28240764e-01
-3.94828528e-01 -4.33721185e-01 1.36387842e-02 5.83798826e-01
-2.71798819e-01 2.21407518e-01 2.80214548e-01 -2.46162534e-01
1.47305536e+00 1.33319035e-01 2.42359772e-01 -3.28210413e-01
2.99577981e-01 5.07408828e-02 9.43336308e-01 7.48006940e-01
-2.73003548e-01 6.97443336e-02 1.10894537e+00 -3.32090288e-01
-7.11490512e-01 -1.01530659e+00 -5.77209108e-02 7.98459113e-01
-9.48867947e-02 3.56770992e-01 -6.09837592e-01 -7.70399809e-01
4.75701600e-01 6.92534864e-01 -5.58421075e-01 -2.47509181e-01
-5.84007919e-01 -9.94699419e-01 1.57423437e-01 3.18443716e-01
1.00691915e-01 -1.09372807e+00 -1.18899429e+00 5.95467985e-01
-5.94869889e-02 -7.52583623e-01 -3.44175190e-01 2.35240892e-01
-8.81826460e-01 -1.30201232e+00 -8.90725374e-01 -2.56365418e-01
7.96869814e-01 -3.18791032e-01 1.16845298e+00 8.76488835e-02
-4.79148299e-01 5.80287933e-01 -2.21883893e-01 -5.07009208e-01
-3.29053879e-01 -1.24435276e-01 -5.90634113e-03 7.87659362e-02
1.80229455e-01 -1.45154417e-01 -1.09279561e+00 2.29942650e-01
-6.86615825e-01 -5.85709326e-02 5.86161256e-01 1.21519899e+00
8.65366757e-01 -2.30406508e-01 1.06605792e+00 -8.74590218e-01
7.96386898e-01 -4.83953714e-01 -9.28532600e-01 4.27463800e-01
-1.09661961e+00 2.12357491e-01 7.51117587e-01 -5.39708912e-01
-5.94804049e-01 -5.26448712e-03 6.32389486e-02 -8.01474810e-01
2.26623446e-01 5.77430010e-01 3.33228409e-01 5.19763768e-01
3.00242603e-01 2.88537472e-01 3.39071751e-01 -1.41380534e-01
-3.53696227e-01 3.23311836e-01 8.11118633e-02 -7.87909448e-01
9.63910446e-02 1.17636539e-01 2.62972981e-01 8.25695868e-04
-9.09446716e-01 -3.01201940e-01 2.28457123e-01 -3.53774428e-01
4.86554742e-01 -7.66884089e-01 -1.51351643e+00 2.81217903e-01
-5.36206663e-01 -8.24783862e-01 -5.09428263e-01 5.94001889e-01
-8.51576269e-01 -1.55972391e-02 -5.07849038e-01 -9.71401513e-01
-6.88950896e-01 -1.30091333e+00 8.75875890e-01 1.98907062e-01
-6.85181245e-02 -9.40350413e-01 1.70131475e-01 2.63969392e-01
3.42753857e-01 5.98907471e-01 1.05890524e+00 -5.18656671e-01
-4.31958735e-01 4.40914705e-02 8.65703523e-02 2.87771553e-01
3.70833129e-02 -1.62080094e-01 -3.89303952e-01 -7.12389708e-01
-9.70785692e-02 -5.69031477e-01 5.40956020e-01 7.85879016e-01
1.48886228e+00 -4.04991001e-01 -1.85763359e-01 6.22275591e-01
1.49575603e+00 5.57257950e-01 1.81200281e-01 2.75275767e-01
2.01448396e-01 5.16855359e-01 9.06579912e-01 1.13967741e+00
3.91336948e-01 3.41975808e-01 5.85288346e-01 -1.47608435e-02
4.19068187e-01 -6.73068762e-02 2.93555051e-01 2.32547894e-01
1.53963596e-01 -3.36279392e-01 -1.16045618e+00 4.66554463e-01
-1.94968188e+00 -5.85255444e-01 3.73535663e-01 2.32734561e+00
9.67154324e-01 1.64131075e-02 3.81528050e-01 -1.76961452e-01
5.19845068e-01 -4.84203696e-01 -1.27053285e+00 -4.85503644e-01
2.37779215e-01 3.22700560e-01 3.99982929e-01 3.16286355e-01
-7.03956604e-01 4.27146018e-01 5.44454288e+00 5.11240840e-01
-1.39201152e+00 -2.64587868e-02 1.16366637e+00 -4.87896889e-01
-1.32803991e-01 -4.16103631e-01 -4.40977663e-01 7.05067396e-01
1.10406852e+00 -1.94716305e-01 7.03528404e-01 7.06568301e-01
6.52948558e-01 -7.36047626e-02 -1.19044840e+00 8.66803885e-01
-4.79992568e-01 -1.17284203e+00 -5.16924739e-01 1.19758388e-02
8.29495192e-01 6.08198866e-02 1.31526649e-01 4.71270591e-01
5.43477237e-01 -9.68870938e-01 4.06057179e-01 5.28563738e-01
9.78448987e-01 -1.04501617e+00 8.17282438e-01 3.77194345e-01
-6.16605043e-01 -5.00773847e-01 -5.19002713e-02 2.33184531e-01
6.12240881e-02 9.28207934e-01 -1.06437492e+00 4.26133513e-01
5.52685618e-01 4.57194537e-01 3.98995169e-02 9.41044867e-01
-1.33947521e-01 5.33598483e-01 -1.38389155e-01 -3.39982569e-01
1.69906586e-01 -1.49949208e-01 3.93435121e-01 1.10906637e+00
4.32961851e-01 1.24483526e-01 4.49647844e-01 7.93134570e-01
-7.87321255e-02 1.62674814e-01 -2.41658211e-01 -3.06621403e-01
3.26580375e-01 9.85136449e-01 -7.66412020e-01 -2.60918170e-01
7.23219737e-02 5.83747923e-01 2.26614624e-01 1.60786226e-01
-1.04316771e+00 -3.86180822e-03 6.82768285e-01 9.38090216e-03
2.42581293e-01 2.74481893e-01 -2.78380752e-01 -1.05469394e+00
9.15160850e-02 -1.21272540e+00 8.13622832e-01 -2.11893693e-01
-1.28700101e+00 1.85187772e-01 -2.10743561e-01 -1.20148742e+00
-7.20243156e-01 -5.07788360e-01 -2.42444411e-01 6.15286946e-01
-1.76181567e+00 -4.55047667e-01 5.61110117e-02 4.82448071e-01
3.67878437e-01 7.96956792e-02 6.30815566e-01 2.24823505e-01
-9.03395414e-01 6.74270570e-01 3.10679078e-01 -2.53310323e-01
5.48091233e-01 -1.24485540e+00 -2.69723564e-01 5.97115576e-01
-7.69170463e-01 2.60848582e-01 5.69128573e-01 -5.83878517e-01
-1.77698755e+00 -1.02658546e+00 3.89877170e-01 -5.20633310e-02
5.82824171e-01 -9.92643647e-03 -7.31802285e-01 2.10740298e-01
-1.52016297e-01 6.57650596e-03 4.61638331e-01 -1.06105819e-01
2.19758481e-01 -4.27654862e-01 -1.45308483e+00 4.83514518e-01
8.98169518e-01 -1.75255239e-02 2.78748069e-02 6.60013437e-01
6.53765678e-01 -5.42556584e-01 -1.12038326e+00 7.21177340e-01
4.26441014e-01 -6.81635082e-01 8.32786739e-01 -7.85549283e-01
5.22754371e-01 1.52111471e-01 2.58834392e-01 -1.37191463e+00
-1.28489152e-01 -8.08223069e-01 -5.18839836e-01 5.63233256e-01
3.41691822e-01 -1.02860022e+00 6.62572980e-01 4.57796335e-01
-1.88362168e-03 -1.58493674e+00 -1.08975160e+00 -8.40001523e-01
-8.63000285e-03 -6.37450591e-02 5.45415938e-01 9.46258843e-01
-1.04415752e-02 -1.02784567e-01 -2.73119241e-01 1.14890538e-01
7.79386222e-01 4.24564451e-01 2.33818933e-01 -8.94547045e-01
-9.83624816e-01 -6.88227654e-01 1.51569664e-01 -5.63591361e-01
8.37662816e-02 -6.06716990e-01 1.08400226e-01 -1.52251947e+00
2.69511640e-01 -6.09527409e-01 -7.24622846e-01 8.26236069e-01
-4.36878711e-01 -5.02160072e-01 -6.36274042e-03 -1.61286607e-01
-7.11185813e-01 3.86848807e-01 1.48182571e+00 -8.67855027e-02
-4.12125111e-01 1.30720630e-01 -7.33772635e-01 2.33413041e-01
1.06038105e+00 -6.39613271e-01 -4.62743521e-01 -2.58490771e-01
2.46200189e-01 8.89839888e-01 1.65984094e-01 -7.01829553e-01
-1.61424130e-01 -7.75431573e-01 2.63552725e-01 -3.95217568e-01
-8.35204795e-02 -5.69057345e-01 1.57940939e-01 1.37795091e+00
-2.86894351e-01 2.62162179e-01 1.01753667e-01 4.71848845e-01
2.33051583e-01 1.56631805e-02 9.26767588e-01 -1.38658807e-01
-1.65537655e-01 4.16516095e-01 -2.00154439e-01 3.20275784e-01
1.27664161e+00 2.74949610e-01 -2.47738436e-01 -7.80228451e-02
-4.69505072e-01 9.32032526e-01 1.52459443e-01 1.67858154e-01
5.76761484e-01 -7.65167236e-01 -1.01735950e+00 5.83676547e-02
-1.56332508e-01 7.62147084e-02 4.41028178e-01 1.16284883e+00
-5.68932593e-01 4.56640005e-01 -2.43199244e-02 -7.61774778e-01
-8.79174292e-01 6.32687688e-01 7.12734401e-01 -9.81057525e-01
-4.74184543e-01 6.29369676e-01 -1.41047522e-01 -3.84179085e-01
1.95633441e-01 -3.28326702e-01 3.29825103e-01 -1.39219567e-01
3.68027717e-01 5.90656042e-01 1.33243904e-01 5.36453247e-01
-4.70059991e-01 1.56788155e-01 1.29462615e-01 2.26578098e-02
1.62052453e+00 2.64176428e-01 -5.82000352e-02 1.01499952e-01
8.15710783e-01 -6.13433182e-01 -1.62640440e+00 -1.05575167e-01
1.75260007e-02 -3.24219972e-01 -1.25530064e-01 -1.27918327e+00
-1.18169200e+00 6.49380684e-01 7.62962639e-01 4.99805026e-02
1.40203035e+00 -2.33797938e-01 5.94739616e-01 3.54606479e-01
3.90327871e-01 -1.12757170e+00 2.36328125e-01 2.06311762e-01
5.63113451e-01 -1.22108531e+00 -1.43174931e-01 2.62154311e-01
-8.64684761e-01 8.37737501e-01 6.70083582e-01 -9.00461525e-02
4.34630245e-01 3.59447181e-01 -3.77310696e-03 -1.12931207e-01
-1.23012936e+00 -6.15614690e-02 -2.50888228e-01 2.29097024e-01
1.59378871e-01 5.57534635e-01 -5.00938892e-01 4.86587584e-01
3.88006866e-01 3.09033871e-01 3.53023887e-01 1.12737846e+00
-1.91693693e-01 -1.17877936e+00 -4.31065172e-01 8.45947623e-01
-6.11909866e-01 2.32896015e-01 1.63769960e-01 4.81607616e-01
-6.27567098e-02 8.05330575e-01 4.38529290e-02 -6.27157837e-02
2.70804048e-01 2.24697039e-01 5.62766850e-01 -3.38656574e-01
-5.61668515e-01 2.79536366e-01 -1.92386717e-01 -6.77520871e-01
-2.18885750e-01 -7.85075009e-01 -1.59416163e+00 -3.80444318e-01
1.22746654e-01 5.48885800e-02 5.67602277e-01 9.33631778e-01
7.04368889e-01 9.12687540e-01 9.81062829e-01 -1.44625157e-01
-1.24527740e+00 -5.76263607e-01 -5.43573320e-01 2.49403715e-01
4.20934170e-01 -6.61673665e-01 -9.18971077e-02 -4.44757968e-01] | [4.030922889709473, 2.716982364654541] |
3a22876f-af18-4224-a6f0-3d1e144925e0 | self-paced-deep-regression-forests-with | 2004.01459 | null | https://arxiv.org/abs/2004.01459v4 | https://arxiv.org/pdf/2004.01459v4.pdf | Self-Paced Deep Regression Forests with Consideration on Underrepresented Examples | Deep discriminative models (e.g. deep regression forests, deep neural decision forests) have achieved remarkable success recently to solve problems such as facial age estimation and head pose estimation. Most existing methods pursue robust and unbiased solutions either through learning discriminative features, or reweighting samples. We argue what is more desirable is learning gradually to discriminate like our human beings, and hence we resort to self-paced learning (SPL). Then, a natural question arises: can self-paced regime lead deep discriminative models to achieve more robust and less biased solutions? To this end, this paper proposes a new deep discriminative model--self-paced deep regression forests with consideration on underrepresented examples (SPUDRFs). It tackles the fundamental ranking and selecting problem in SPL from a new perspective: fairness. This paradigm is fundamental and could be easily combined with a variety of deep discriminative models (DDMs). Extensive experiments on two computer vision tasks, i.e., facial age estimation and head pose estimation, demonstrate the efficacy of SPUDRFs, where state-of-the-art performances are achieved. | ['Lili Pan', 'Zenglin Xu', 'Yazhou Ren', 'Shijie Ai'] | 2020-04-03 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/7424_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123750273.pdf | eccv-2020-8 | ['head-pose-estimation'] | ['computer-vision'] | [-1.85527384e-01 1.22848161e-01 -3.20729554e-01 -9.08146679e-01
-4.15751368e-01 -3.25492099e-02 7.02424228e-01 -3.64825279e-01
-6.76971555e-01 1.01704752e+00 3.67479682e-01 7.20537901e-02
-1.11614577e-01 -6.68064833e-01 -2.53088713e-01 -1.10071504e+00
5.51536642e-02 6.20499730e-01 -2.24850535e-01 3.11083975e-04
2.16627210e-01 5.67230880e-01 -1.77644777e+00 -4.38814491e-01
1.12719142e+00 1.09233356e+00 -2.80368209e-01 1.78146347e-01
-1.04876133e-02 4.45863664e-01 -1.35911047e-01 -8.74930680e-01
7.92081207e-02 -4.07973342e-02 -7.34292567e-01 -2.38496527e-01
6.24517024e-01 -5.84261835e-01 -3.38749401e-02 8.72655511e-01
1.09843600e+00 6.88699707e-02 1.07467568e+00 -1.55147874e+00
-6.03833139e-01 4.27508980e-01 -9.87359226e-01 1.37828231e-01
1.66277334e-01 5.57430834e-02 8.61524224e-01 -1.06170118e+00
2.90184408e-01 1.49119914e+00 8.79753649e-01 1.06143522e+00
-1.15936017e+00 -1.12104487e+00 3.30824614e-01 5.42770267e-01
-1.43511939e+00 -8.01876962e-01 8.63207459e-01 -3.65557611e-01
1.51984259e-01 2.05300748e-01 6.51255846e-01 1.63619697e+00
9.82941762e-02 1.10751474e+00 1.69318807e+00 -2.67784029e-01
2.38382012e-01 6.50944933e-02 8.78485218e-02 7.22933054e-01
3.42113942e-01 2.57329673e-01 -8.86328995e-01 -1.52368948e-01
5.06516278e-01 -1.57499194e-01 1.61806941e-01 -5.52604079e-01
-4.19954270e-01 9.87776220e-01 2.49576539e-01 1.64879531e-01
-3.32841039e-01 -1.10719770e-01 3.64774734e-01 1.14630066e-01
8.41596663e-01 -7.50191286e-02 -3.94590199e-01 -3.03356890e-02
-1.10165703e+00 5.69552481e-01 4.83786196e-01 5.94024658e-01
6.93228424e-01 1.73418745e-01 -4.52903211e-01 1.02643883e+00
5.19198596e-01 4.99659091e-01 6.80427909e-01 -6.96898222e-01
-5.94211221e-02 1.58111572e-01 -1.93774804e-01 -7.61188567e-01
-5.84210038e-01 -6.40634596e-01 -1.05090010e+00 4.05007869e-01
6.60058200e-01 -2.03369513e-01 -9.15819764e-01 2.14317775e+00
7.01154292e-01 -1.04832418e-01 -3.98901016e-01 7.70192862e-01
8.01253736e-01 8.52936357e-02 6.41859710e-01 -5.74476004e-01
1.27993977e+00 -6.80630863e-01 -5.89958251e-01 -2.62826115e-01
1.90736353e-01 -5.22822142e-01 8.27326059e-01 5.65844357e-01
-8.93047929e-01 -5.14367640e-01 -5.34133315e-01 -5.15689142e-02
-1.44960269e-01 7.42195770e-02 8.36170554e-01 1.01796389e+00
-1.24448621e+00 6.31644249e-01 -4.82579887e-01 -4.34229404e-01
8.03972065e-01 5.90554535e-01 -3.76222640e-01 4.39863130e-02
-1.13239086e+00 1.02727878e+00 -2.46446088e-01 1.54983625e-01
-9.01190758e-01 -4.44960803e-01 -6.77878916e-01 -3.20704848e-01
1.03485473e-01 -1.14055192e+00 1.41706729e+00 -1.09229386e+00
-1.55009937e+00 1.35745442e+00 -4.05700028e-01 -4.86768067e-01
9.70546365e-01 -5.73053062e-01 -1.12947941e-01 -3.68268043e-01
1.77428529e-01 7.85322964e-01 1.27258670e+00 -9.91669655e-01
-5.95359623e-01 -1.18067718e+00 -3.56921017e-01 1.27720535e-01
-6.15610600e-01 1.28270999e-01 1.26799554e-01 -5.31814456e-01
7.18483478e-02 -9.80602324e-01 -1.81384638e-01 2.05733925e-01
-3.31170201e-01 -9.21074510e-01 5.36318839e-01 -6.83622658e-01
1.29889131e+00 -1.77792943e+00 9.89893675e-02 -1.50375171e-02
5.88914394e-01 1.53362334e-01 3.66225354e-02 -1.57039896e-01
-7.34206736e-02 -3.05989653e-01 6.01474270e-02 -6.06436253e-01
1.16155356e-01 9.10689607e-02 5.66297285e-02 7.25258410e-01
2.54541803e-02 7.98698545e-01 -6.49595141e-01 -9.06591296e-01
-2.21383780e-01 4.62819397e-01 -5.67006826e-01 3.55974764e-01
2.59971112e-01 5.98307669e-01 -4.44474280e-01 9.39605236e-01
1.01384938e+00 3.47195148e-01 2.47281487e-03 -1.86952323e-01
-2.17881590e-01 -6.48304075e-03 -8.09711576e-01 1.29785001e+00
-4.68174636e-01 3.46203715e-01 -3.15688811e-02 -9.74939704e-01
1.23342502e+00 -1.09201647e-01 3.95495653e-01 -9.02132869e-01
2.06193343e-01 2.23736107e-01 -1.36982411e-01 -5.76276422e-01
3.52028698e-01 -4.90594953e-01 8.35739598e-02 9.08777565e-02
2.52833247e-01 2.48892739e-01 -2.45225370e-01 -3.05336744e-01
6.38203382e-01 3.19633096e-01 6.27720594e-01 -4.24648762e-01
5.64122856e-01 -7.06583381e-01 8.53025377e-01 5.47182560e-01
-7.44640112e-01 3.33664477e-01 4.17102367e-01 -7.12806523e-01
-7.05867052e-01 -1.14549899e+00 -3.90988648e-01 1.78859615e+00
-2.52471060e-01 -8.83278117e-05 -9.05520797e-01 -9.65101719e-01
2.35445291e-01 4.83848035e-01 -1.11935115e+00 -2.48954773e-01
-5.13432980e-01 -1.03116012e+00 3.88044536e-01 7.58932710e-01
5.67047238e-01 -8.36443841e-01 -3.26344103e-01 -4.67098132e-02
3.24243270e-02 -6.05991304e-01 -8.23099092e-02 2.41328865e-01
-1.02261198e+00 -7.41293907e-01 -1.40180993e+00 -6.69666886e-01
6.16464734e-01 1.35519141e-02 1.27887464e+00 -3.44285697e-01
-1.38521597e-01 2.61101544e-01 -8.76451358e-02 -5.75537205e-01
1.30058885e-01 3.76707166e-01 5.48638940e-01 3.25445086e-01
6.91606402e-01 -1.06437016e+00 -9.81402874e-01 3.60050827e-01
-2.40032569e-01 -2.12992355e-01 8.66830647e-01 9.94037986e-01
3.07718039e-01 -4.75851983e-01 7.89076865e-01 -1.06585467e+00
5.40650964e-01 -3.68181020e-01 -2.13043675e-01 1.95372120e-01
-1.03805804e+00 2.55764455e-01 2.81057864e-01 -3.72485816e-01
-1.23351431e+00 1.04320541e-01 -5.33974469e-01 -1.66039824e-01
-3.05343956e-01 -1.09501052e-04 -4.97604072e-01 -8.25119689e-02
7.10144818e-01 2.36421078e-01 1.61822602e-01 -6.01654053e-01
3.38550687e-01 8.81642520e-01 3.70125055e-01 -7.83622384e-01
8.67972374e-01 3.31520796e-01 1.05843961e-01 -5.66636026e-01
-9.73410130e-01 -2.33359843e-01 -6.74610794e-01 -4.80124235e-01
6.80275738e-01 -1.01263380e+00 -8.24870110e-01 8.49263906e-01
-8.51346552e-01 -1.10001579e-01 -9.81838331e-02 2.21299186e-01
-5.68289936e-01 2.05395579e-01 -2.99423426e-01 -1.30863643e+00
-6.71692431e-01 -6.20269418e-01 1.18961108e+00 5.94051480e-01
-4.44640279e-01 -7.03521848e-01 2.71088749e-01 2.51402438e-01
5.31471908e-01 1.47619918e-01 7.22122967e-01 -5.81566989e-01
-3.61403041e-02 -3.70340943e-02 -2.30401590e-01 1.63874611e-01
-1.02869801e-01 -9.00145169e-05 -1.58137774e+00 -2.94098556e-01
-1.53526843e-01 -5.71073592e-01 1.06688726e+00 6.39572918e-01
1.41843343e+00 -3.74639630e-01 -2.45056704e-01 7.20046282e-01
1.06688917e+00 -1.43775165e-01 6.81354642e-01 3.70358407e-01
5.80940485e-01 9.19014037e-01 6.43444121e-01 7.31295764e-01
8.00804734e-01 7.41463125e-01 4.54751343e-01 -6.86226189e-02
4.54062037e-02 -3.93561304e-01 2.16383889e-01 6.12474620e-01
-7.07793295e-01 3.12552810e-01 -6.77039862e-01 3.15266848e-01
-1.74993575e+00 -8.48752856e-01 1.43389806e-01 2.07093310e+00
7.57486999e-01 -2.20165625e-02 8.09414148e-01 3.17228347e-01
6.60145879e-01 3.09284806e-01 -7.35293448e-01 -4.81037498e-01
-2.75362618e-02 3.07430267e-01 1.85662180e-01 -2.22784728e-02
-1.26143467e+00 8.11183453e-01 5.75656223e+00 9.93345439e-01
-1.26374257e+00 3.10879022e-01 1.22944748e+00 6.74136505e-02
-2.39700302e-01 -4.66576099e-01 -1.11076534e+00 4.55004483e-01
7.26142108e-01 -2.21240669e-01 -5.41615300e-03 1.31961191e+00
1.42131373e-01 -1.04823411e-01 -1.02184796e+00 1.33546150e+00
1.53717655e-03 -6.08532667e-01 -1.53834358e-01 1.81285769e-01
5.13295770e-01 -3.32470119e-01 5.94883919e-01 6.49852276e-01
3.11556101e-01 -1.06193113e+00 7.96501696e-01 6.83103859e-01
9.81232345e-01 -7.84373224e-01 6.43796027e-01 2.00373709e-01
-8.41981530e-01 -2.85764933e-01 -4.33065712e-01 -3.89948264e-02
-2.47934237e-01 8.03224862e-01 -5.87620199e-01 2.84148037e-01
8.76497090e-01 6.23749912e-01 -7.12058067e-01 1.15259981e+00
-3.53035271e-01 4.41120952e-01 -2.05824170e-02 -3.95487636e-01
-1.86668426e-01 8.66914541e-02 1.86397880e-01 9.47735012e-01
2.66984910e-01 -1.32737413e-01 -3.78926367e-01 4.60280478e-01
6.32529184e-02 3.83702248e-01 -5.36541164e-01 4.66629684e-01
2.84855366e-01 1.43947816e+00 -3.32916290e-01 7.79004171e-02
-5.66257797e-02 8.13209295e-01 6.14371181e-01 9.73540023e-02
-6.91665530e-01 1.63984492e-01 8.25004160e-01 3.01546872e-01
-8.86286551e-04 -5.06477728e-02 -4.45986837e-01 -1.09337771e+00
-2.71942079e-01 -8.23775113e-01 4.99142170e-01 -3.88763487e-01
-1.65500820e+00 6.23641789e-01 -1.56365886e-01 -1.12237704e+00
-3.90860915e-01 -5.45259297e-01 -5.57598889e-01 6.96078241e-01
-1.70686758e+00 -1.38885581e+00 -5.32505155e-01 7.30807662e-01
4.01724845e-01 -1.95710152e-01 8.73513281e-01 3.90536636e-01
-7.14231551e-01 1.17673647e+00 -1.60444491e-02 -1.15218997e-01
9.81018007e-01 -1.26342237e+00 1.42387584e-01 4.37755138e-01
-3.78854759e-03 4.82269794e-01 7.81766474e-01 -1.83496416e-01
-1.08193862e+00 -8.70996952e-01 1.02816176e+00 -2.38245308e-01
2.25061461e-01 -4.23775822e-01 -6.18022859e-01 1.54268831e-01
-1.13093473e-01 1.10414378e-01 8.27879131e-01 6.86468422e-01
-4.76405889e-01 -7.00311303e-01 -1.35066068e+00 4.98420894e-01
1.52540720e+00 -2.74285942e-01 -2.97784030e-01 3.88638563e-02
7.42392838e-02 -1.42331824e-01 -4.50220555e-01 6.19962871e-01
1.14480543e+00 -1.34770930e+00 8.43442440e-01 -7.67753184e-01
5.21639585e-01 2.30279818e-01 4.85214451e-03 -1.23897529e+00
-4.85797584e-01 -5.40665209e-01 -3.81610304e-01 1.46138668e+00
7.20045865e-02 -6.26238823e-01 1.21978223e+00 7.18702734e-01
3.75759900e-01 -1.08596885e+00 -1.26936662e+00 -7.25251555e-01
2.59166062e-01 -1.48427337e-01 7.05800891e-01 7.86345124e-01
-5.79685926e-01 3.99655133e-01 -6.83384836e-01 -1.88635916e-01
8.99231613e-01 7.99044222e-02 9.06855285e-01 -1.74428058e+00
1.25489190e-01 -4.91050810e-01 -5.47083914e-01 -9.46025372e-01
4.58296537e-01 -3.94597769e-01 3.44022475e-02 -1.14549494e+00
4.37777340e-01 -4.69296426e-01 -3.58002692e-01 4.11144972e-01
-3.51066798e-01 1.72884226e-01 -6.14400804e-02 9.55701433e-03
-6.07418656e-01 9.59888518e-01 1.06623137e+00 -8.90088230e-02
1.20988786e-01 6.14146769e-01 -1.12623191e+00 8.79832625e-01
6.80902600e-01 -3.15398693e-01 -1.50869712e-02 -1.01960316e-01
-2.20343675e-02 -2.07226768e-01 2.27707341e-01 -8.93636465e-01
9.20509994e-02 -1.66651011e-01 8.20421338e-01 -3.47438484e-01
1.91703096e-01 -4.13485885e-01 -2.50870019e-01 4.48019266e-01
-1.65329829e-01 1.48353921e-02 -1.63930893e-01 4.36286747e-01
-7.33909532e-02 -1.07595824e-01 1.03605247e+00 -2.77402624e-02
-8.48963737e-01 6.40641987e-01 -6.61660135e-02 9.20674205e-02
8.11536014e-01 -2.82529473e-01 -6.88066036e-02 -4.80435610e-01
-8.94821167e-01 -7.35895410e-02 1.58607885e-01 3.38143915e-01
5.06209910e-01 -1.44380319e+00 -8.74648571e-01 2.25215942e-01
1.51133642e-01 -2.60929435e-01 2.18553901e-01 9.82137203e-01
7.03936815e-02 1.60177067e-01 -4.97908354e-01 -5.74567854e-01
-1.48303020e+00 1.85872838e-01 2.47869000e-01 -2.95057744e-01
-6.95076287e-02 1.30244005e+00 4.03854311e-01 -4.32944894e-01
4.16487545e-01 9.06046927e-02 -5.18691778e-01 6.49811387e-01
4.68095213e-01 5.75030684e-01 -1.20691815e-02 -5.80758750e-01
-6.48453176e-01 6.67220294e-01 -2.66668916e-01 2.31627166e-01
1.64603436e+00 -1.87123075e-01 1.96875166e-02 1.35318175e-01
1.04340947e+00 -5.68343960e-02 -1.36092877e+00 -2.39582315e-01
9.20288861e-02 -6.17987633e-01 -1.44247210e-03 -6.34187520e-01
-1.21824038e+00 9.84046996e-01 1.05618870e+00 -1.23709671e-01
1.27825344e+00 -5.60446009e-02 6.19719923e-01 3.90654653e-02
7.01931179e-01 -1.12043798e+00 1.28691033e-01 2.71594912e-01
8.59596372e-01 -1.46112943e+00 1.83408573e-01 -3.74633893e-02
-5.45421064e-01 1.06586766e+00 9.39074576e-01 7.66795799e-02
6.07183635e-01 7.91901909e-03 1.34872794e-02 2.27939606e-01
-5.41228890e-01 -4.43673015e-01 3.22608739e-01 8.33628178e-01
5.88445067e-01 2.32908279e-01 -6.20406508e-01 8.48149061e-01
-3.84510815e-01 1.68913022e-01 -1.51225075e-01 4.83073324e-01
-3.81877363e-01 -1.23342669e+00 -3.21074843e-01 5.50544500e-01
-4.01580095e-01 1.63392186e-01 -4.37816054e-01 4.90239114e-01
3.70319575e-01 4.81846511e-01 -1.17670409e-01 -4.79228377e-01
1.55430809e-01 1.46782279e-01 7.46697783e-01 -2.16029152e-01
-2.72468567e-01 -1.14902645e-01 5.44559658e-02 -4.49201524e-01
-6.21250093e-01 -9.11514223e-01 -4.03534830e-01 -5.70901811e-01
-3.10751408e-01 -3.39630954e-02 5.69441974e-01 8.43583941e-01
3.27793881e-02 -3.03126639e-03 1.02490520e+00 -9.51974511e-01
-8.93140435e-01 -9.97903407e-01 -8.02169263e-01 9.96836647e-02
2.10052356e-01 -1.06453168e+00 -2.61935622e-01 -3.91539395e-01] | [13.606466293334961, 0.9083948135375977] |
32272b25-209b-419e-a386-25b01da3ec05 | superior-performance-with-diversified | null | null | https://openreview.net/forum?id=tvwNdOKhuF5 | https://openreview.net/pdf?id=tvwNdOKhuF5 | Superior Performance with Diversified Strategic Control in FPS Games Using General Reinforcement Learning | This paper offers an overall solution for first-person shooter (FPS) games to achieve superior performance using general reinforcement learning (RL). We introduce an agent in ViZDoom that can surpass previous top agents ranked in the open ViZDoom AI Competitions by a large margin. The proposed framework consists of a number of generally applicable techniques, including hindsight experience replay (HER) based navigation, hindsight proximal policy optimization (HPPO), rule-guided policy search (RGPS), prioritized fictitious self-play (PFSP), and diversified strategic control (DSC). The proposed agent outperforms existing agents by taking advantage of diversified and human-like strategies, instead of larger neural networks, more accurate frag skills, or hand-craft tricks, etc. We provide comprehensive analysis and experiments to elaborate the effect of each component in affecting the agent performance, and demonstrate that the proposed and adopted techniques are important to achieve superior performance in general end-to-end FPS games. The proposed methods can contribute to other games and real-world tasks which also require spatial navigation and diversified behaviors. | ['Lei Han', 'Zhengyou Zhang', 'Zhuobin Zheng', 'Peng Sun', 'Chun Yuan', 'Jiawei Xu', 'Shuxing Li'] | 2021-09-29 | null | null | null | null | ['fps-games'] | ['playing-games'] | [-3.84041518e-01 -1.05519732e-02 -6.82633668e-02 1.36767671e-01
-3.55558068e-01 -5.66165149e-01 6.66339695e-01 -3.96136850e-01
-8.31636906e-01 1.34297431e+00 2.52034158e-01 -1.76337346e-01
-9.28198755e-01 -4.09803838e-01 -4.63339150e-01 -8.28339815e-01
-3.87795299e-01 7.76075244e-01 4.72577780e-01 -1.26467061e+00
8.75396013e-01 3.63298595e-01 -1.55098093e+00 -1.90390259e-01
1.17579019e+00 8.09257090e-01 6.22912228e-01 8.66931379e-01
3.04162294e-01 1.51697278e+00 -5.91981411e-01 -3.74586172e-02
5.69887757e-01 -4.61017132e-01 -7.94979334e-01 -4.51944917e-01
-2.55445868e-01 -6.07670724e-01 -2.22264767e-01 7.68352211e-01
9.54746008e-01 8.48524868e-01 6.54553056e-01 -1.37707448e+00
-2.13836357e-01 7.12201059e-01 -6.52568579e-01 5.21628559e-01
5.27138889e-01 6.11328661e-01 7.18944252e-01 -4.48368713e-02
4.92802858e-01 1.25747955e+00 6.08380079e-01 6.47727489e-01
-3.12406242e-01 -5.82760036e-01 1.60813883e-01 4.41961139e-01
-7.29721129e-01 -1.86022226e-04 6.14128053e-01 -1.05964839e-01
1.05619705e+00 1.43549681e-01 9.50645149e-01 1.28148198e+00
4.78172392e-01 8.73858035e-01 1.40857744e+00 -2.21118778e-01
5.01226723e-01 -9.35952291e-02 -3.07909757e-01 6.90227628e-01
1.62111327e-01 9.27669764e-01 -4.61435437e-01 -6.09218627e-02
9.53753591e-01 -3.73526514e-01 -3.58938538e-02 -5.39603829e-01
-1.11968315e+00 9.03492510e-01 6.13440573e-01 9.12986845e-02
-9.39323366e-01 3.52526605e-01 4.18251246e-01 5.02707124e-01
-5.25156677e-01 1.23055422e+00 -5.71170926e-01 -6.52662396e-01
-5.73809326e-01 1.06585920e+00 7.61884212e-01 7.26981223e-01
2.44574115e-01 5.55322230e-01 -3.91487479e-01 5.97325861e-01
6.82254136e-02 3.09775174e-01 8.04654717e-01 -1.49283743e+00
3.36691350e-01 2.37734437e-01 5.78300416e-01 -8.43562782e-01
-9.03114140e-01 -7.71532118e-01 -2.69886494e-01 7.52313733e-01
2.27179542e-01 -6.60945594e-01 -7.46727765e-01 1.74869335e+00
2.23379627e-01 -3.28900740e-02 3.09127808e-01 1.14460576e+00
6.29210174e-01 8.10980260e-01 7.21095875e-02 8.77381712e-02
9.55509126e-01 -1.51312602e+00 -5.08801699e-01 -1.69590279e-01
3.71460944e-01 -1.53489724e-01 9.58460450e-01 9.30413604e-01
-1.12267935e+00 -4.75038946e-01 -1.11842740e+00 5.19902408e-01
-3.20556104e-01 -3.01654279e-01 8.61914814e-01 4.80467707e-01
-1.02735436e+00 7.91699767e-01 -4.67570603e-01 -5.71413457e-01
1.19228229e-01 5.42746603e-01 -2.96732467e-02 5.02380610e-01
-1.33879828e+00 1.23257720e+00 7.78691769e-01 -1.51557475e-01
-1.57817364e+00 -3.17705631e-01 -3.85285437e-01 1.37676718e-03
9.71309543e-01 -6.88449085e-01 1.45538807e+00 -7.56659150e-01
-2.21461511e+00 2.96514064e-01 8.18553686e-01 -7.45645761e-01
6.75899506e-01 -3.07745934e-01 -4.86413538e-02 1.77104428e-01
9.39497054e-02 6.25450969e-01 5.91575384e-01 -1.17954957e+00
-1.13636363e+00 -2.87554383e-01 4.20528859e-01 1.12201488e+00
1.24890573e-01 -3.49934399e-02 1.85853437e-01 -3.17849338e-01
-3.49538267e-01 -9.09989119e-01 -7.09099293e-01 -7.78917670e-01
-1.60472412e-02 -3.42667609e-01 3.99335057e-01 -5.71069419e-01
1.00439036e+00 -1.63786995e+00 5.74019790e-01 3.49194072e-02
3.76769677e-02 4.10751134e-01 -1.61098279e-02 8.52225602e-01
5.31488180e-01 -5.86240292e-01 3.01327586e-01 4.34393287e-01
1.98066533e-01 2.73611784e-01 6.03931397e-02 3.23281772e-02
-8.32952797e-01 9.97583270e-01 -1.11748922e+00 -2.41403192e-01
1.22186989e-01 -3.45987231e-01 -5.96129179e-01 2.55844772e-01
-3.95843178e-01 7.18887389e-01 -7.17828810e-01 3.91041160e-01
1.93016097e-01 4.68905061e-01 -1.98860437e-01 5.65939069e-01
-3.59042764e-01 -8.71300995e-02 -1.31258309e+00 1.60796225e+00
-2.41581813e-01 1.12217898e-02 1.24119952e-01 -1.01387584e+00
1.07999039e+00 1.51265234e-01 3.74725610e-01 -8.86254668e-01
4.23143059e-01 4.01972890e-01 2.91364521e-01 -6.97860062e-01
6.75059378e-01 1.34908259e-01 -6.06117733e-02 1.66091189e-01
2.26707920e-01 -6.47447482e-02 2.29447499e-01 5.01579754e-02
9.60836768e-01 5.57979703e-01 6.72570646e-01 -5.52755773e-01
6.54757917e-01 4.94677335e-01 5.21941960e-01 1.36120224e+00
-7.17769086e-01 -1.44834623e-01 6.31154239e-01 -4.00624186e-01
-9.62038696e-01 -8.22265625e-01 6.34160042e-01 1.57521331e+00
5.45526266e-01 1.71145320e-01 -6.33866847e-01 -4.31260437e-01
-4.55321446e-02 9.53667402e-01 -6.78979576e-01 -2.21671358e-01
-6.25875115e-01 -5.12987494e-01 5.96237421e-01 5.20088136e-01
1.01254690e+00 -1.77378416e+00 -1.22878671e+00 3.44692349e-01
1.07799396e-01 -5.08511186e-01 -2.38803655e-01 3.65310401e-01
-5.98423660e-01 -8.79390061e-01 -9.23222721e-01 -7.14276731e-01
-1.17856950e-01 2.38815248e-01 6.59523726e-01 -1.80101454e-01
2.70701021e-01 3.54519784e-01 -7.32682943e-01 -4.11758095e-01
-3.99706140e-02 1.23731740e-01 3.52327496e-01 -4.55587685e-01
1.27914801e-01 -7.00943649e-01 -6.98529601e-01 4.07095164e-01
-3.50506812e-01 -7.90950209e-02 7.44297743e-01 1.13567746e+00
-8.43707770e-02 1.82711884e-01 7.93069363e-01 -6.47863865e-01
1.46053839e+00 -4.86580312e-01 -8.33209574e-01 1.71192378e-01
-7.67062366e-01 8.23641345e-02 9.40296769e-01 -4.11254555e-01
-1.27569282e+00 -4.03345466e-01 -9.02903974e-02 2.71433359e-03
-1.44631520e-01 3.37562561e-01 1.20907508e-01 -3.37621838e-01
1.13905990e+00 7.01117456e-01 1.24527723e-01 -9.50317085e-02
3.89241993e-01 6.28832400e-01 5.26166141e-01 -8.33960354e-01
3.34074080e-01 1.00153424e-01 9.04694200e-02 -4.55138683e-01
-3.03977847e-01 -3.97490174e-01 -1.82531744e-01 -3.86117905e-01
8.30665052e-01 -7.06888974e-01 -1.37445211e+00 6.19146943e-01
-5.69914818e-01 -4.45111156e-01 -2.21677110e-01 6.51738882e-01
-1.16782570e+00 1.11014545e-01 -5.66580892e-01 -1.09728873e+00
-3.75922203e-01 -1.06126666e+00 3.98913831e-01 8.98662210e-01
1.39784425e-01 -6.60541654e-01 3.20620537e-01 5.29966772e-01
7.10083783e-01 2.22288325e-01 5.74522078e-01 -8.05589557e-01
-3.30980629e-01 1.97558552e-01 3.52811277e-01 -8.88972133e-02
-4.37677294e-01 -4.98740971e-01 -1.67819500e-01 -3.79265010e-01
-9.49131846e-02 -7.71173120e-01 6.07210100e-01 5.95179617e-01
3.83376867e-01 -3.26465368e-01 -1.05768710e-01 6.44968987e-01
1.41806114e+00 1.17533553e+00 6.11282170e-01 1.13362050e+00
2.86804855e-01 5.00737429e-01 8.93459678e-01 7.91061223e-01
4.64365959e-01 5.93246281e-01 8.69245708e-01 3.68871719e-01
4.25774843e-01 -5.50202131e-01 4.23080981e-01 1.00418113e-01
-8.26952040e-01 -1.44166887e-01 -6.00621998e-01 2.98783064e-01
-2.20371389e+00 -1.32310665e+00 2.36927778e-01 1.97324455e+00
2.32451305e-01 2.37320423e-01 8.62316072e-01 -1.05914377e-01
4.47013110e-01 5.30737787e-02 -8.63539279e-01 -7.42467105e-01
-1.07540391e-01 6.80937171e-02 6.63657725e-01 5.31285286e-01
-7.94801116e-01 1.37810695e+00 6.12751579e+00 1.06666315e+00
-7.89158285e-01 -1.57335345e-02 1.23168729e-01 -1.28250480e-01
1.82516336e-01 -7.62719065e-02 -7.13423014e-01 2.92378217e-01
4.12052095e-01 -2.09259614e-01 1.05851400e+00 1.26269186e+00
3.75071228e-01 -4.41562921e-01 -3.40675622e-01 9.38792109e-01
-1.01323545e-01 -1.07809401e+00 -2.16040909e-01 -6.47551864e-02
9.06485140e-01 -1.04626857e-01 1.45955205e-01 1.05258262e+00
9.04014468e-01 -1.12386060e+00 7.17013001e-01 4.37057763e-01
-6.09559976e-02 -1.08439183e+00 8.96746635e-01 8.55447888e-01
-6.84322119e-01 -9.69174206e-01 -4.87642407e-01 -6.01536274e-01
-6.59211874e-02 -6.63335502e-01 -6.88794136e-01 8.23058963e-01
6.41599298e-01 1.82144150e-01 -1.92716256e-01 1.15126657e+00
-2.77526855e-01 1.75127223e-01 -1.53056055e-01 -1.00171888e+00
1.15400600e+00 -3.56682330e-01 8.53697240e-01 3.42445672e-01
8.07515159e-02 3.97278637e-01 2.31382534e-01 4.29536104e-01
8.80849481e-01 1.96222812e-01 -4.28641737e-01 1.70607954e-01
3.38470072e-01 1.04905403e+00 -5.43751776e-01 -9.51320380e-02
1.28019629e-02 9.02324200e-01 3.24413717e-01 3.82306218e-01
-8.22615623e-01 -3.93482566e-01 3.57008398e-01 4.64262851e-02
3.04191947e-01 -9.92055982e-02 7.77293369e-02 -5.96011996e-01
-3.67022634e-01 -1.27618384e+00 5.45423448e-01 -8.90399516e-01
-8.80237281e-01 7.90625215e-01 1.36350483e-01 -1.07479703e+00
-5.98488569e-01 -5.91255009e-01 -5.92236817e-01 5.83095729e-01
-1.21049976e+00 -9.60961163e-01 -2.91619062e-01 8.37358594e-01
5.87498665e-01 -9.73850548e-01 4.08875614e-01 -1.44749120e-01
-3.87883484e-01 4.18517083e-01 1.95784047e-01 -3.57120574e-01
1.80465579e-01 -1.32385647e+00 -9.58905593e-02 4.66804862e-01
-3.59973401e-01 1.75438419e-01 1.00604820e+00 -5.53259075e-01
-1.13317108e+00 -3.11827719e-01 -1.45305604e-01 -8.63302946e-02
4.18517202e-01 1.12165235e-01 3.26571539e-02 4.71732855e-01
4.29873616e-01 -7.77233779e-01 -2.14483533e-02 1.97447270e-01
3.37507248e-01 -5.91501482e-02 -1.08175147e+00 1.07005000e+00
1.11175835e+00 3.88373971e-01 -8.57496262e-01 9.45341662e-02
3.85051399e-01 -6.38487339e-01 -3.95037234e-02 2.43704200e-01
5.84503591e-01 -1.38937950e+00 8.84925127e-01 -1.16280746e+00
3.95517528e-01 -2.09693879e-01 -2.02570278e-02 -1.76588988e+00
-8.24754298e-01 -1.09718275e+00 7.38337860e-02 4.20228601e-01
1.61133170e-01 -6.16437018e-01 1.00606322e+00 1.10495411e-01
-2.34191760e-01 -9.96093094e-01 -8.08503270e-01 -1.08870387e+00
1.66463479e-01 2.56603390e-01 5.76719463e-01 5.92829347e-01
3.45788091e-01 3.34141195e-01 -1.13428557e+00 -4.76050079e-02
4.31437641e-01 -2.21451700e-01 1.03349650e+00 -9.28852320e-01
-7.10375607e-01 -5.88673651e-01 -3.25829178e-01 -1.25085163e+00
-2.75967211e-01 -3.61838579e-01 9.38588083e-02 -1.58720100e+00
-1.90216675e-01 -4.09796029e-01 -3.48583102e-01 5.36789596e-02
-1.46734923e-01 -5.07775068e-01 3.57381672e-01 1.68827027e-02
-1.00334299e+00 8.06642056e-01 1.74979091e+00 3.07185084e-01
-6.33983254e-01 2.56778568e-01 -9.95103836e-01 7.24347413e-01
1.09746850e+00 -2.38464683e-01 -8.40573490e-01 -1.00625277e-01
1.44992039e-01 7.99835563e-01 2.11855173e-01 -1.33407855e+00
5.10705590e-01 -6.76877260e-01 1.89718530e-01 -2.44477585e-01
4.05975610e-01 -6.23985946e-01 -1.11034691e-01 1.04498315e+00
-2.25322142e-01 3.22954118e-01 -3.21432902e-03 4.97903615e-01
4.28703204e-02 -4.77126420e-01 7.85871387e-01 -8.02728415e-01
-1.28251588e+00 -1.14887510e-03 -5.94345272e-01 1.28490299e-01
1.56982577e+00 -6.14124060e-01 -3.27812493e-01 -9.77544606e-01
-5.54175556e-01 7.89310575e-01 -1.16263427e-01 2.08321735e-01
3.48366946e-01 -9.73285913e-01 -7.86699295e-01 -1.95833981e-01
-2.69658238e-01 -5.26723087e-01 5.61350644e-01 7.40358531e-01
-1.13524652e+00 6.54234529e-01 -1.15445638e+00 -6.52922392e-02
-7.47063339e-01 5.91667891e-01 7.01491296e-01 -7.25570858e-01
-5.31887949e-01 1.09037030e+00 -6.88989758e-02 -8.08996022e-01
4.41981316e-01 1.68780372e-01 -6.71410322e-01 -3.79755467e-01
3.46813321e-01 8.05971920e-01 -5.20289242e-01 -2.48111308e-01
-7.34579340e-02 3.21406633e-01 -1.27015933e-01 -5.95410228e-01
1.29656863e+00 5.19408770e-02 4.48140085e-01 -7.39743859e-02
3.54171097e-02 -3.53461355e-01 -1.56158233e+00 7.24586993e-02
-2.59844393e-01 -2.77650386e-01 -1.16026215e-01 -1.38106155e+00
-6.65937960e-01 2.85557479e-01 4.34365153e-01 3.97637524e-02
1.13320112e+00 -5.50481081e-01 7.05420434e-01 8.15222681e-01
1.00057232e+00 -1.50759852e+00 3.49024653e-01 6.55279934e-01
9.71529663e-01 -9.86202657e-01 -1.71378106e-01 2.09147364e-01
-1.46281052e+00 8.13727677e-01 1.20082104e+00 -6.55820906e-01
1.96136102e-01 -7.50615522e-02 1.38354450e-01 -1.30242422e-01
-9.31469262e-01 -3.62491816e-01 -3.44892293e-01 1.00991917e+00
-5.47783077e-01 -1.27020665e-02 -6.69398367e-01 7.48322606e-01
-6.73898041e-01 5.21457903e-02 4.54103738e-01 9.36567307e-01
-1.04294169e+00 -5.95740378e-01 -5.65288067e-01 4.70881730e-01
-1.76434770e-01 2.56827027e-01 -2.20864326e-01 1.00708449e+00
1.23225018e-01 8.96699667e-01 -5.30155301e-01 -3.84881943e-01
3.81433398e-01 -4.55789238e-01 7.38297999e-01 1.60129517e-02
-8.61690700e-01 -2.16628864e-01 9.58156362e-02 -6.22778594e-01
-9.64942127e-02 -5.69638729e-01 -1.05953586e+00 -5.19474387e-01
4.89079207e-02 4.25535470e-01 2.72684723e-01 9.78853106e-01
3.34645025e-02 5.55586040e-01 4.30244982e-01 -6.95324838e-01
-1.11794567e+00 -1.16331291e+00 -8.71039569e-01 5.24476990e-02
2.62472872e-02 -1.20498407e+00 -2.16419935e-01 -8.24131846e-01] | [3.666414976119995, 1.6032196283340454] |
90d06537-f822-4bb4-afda-6f9511790614 | a-neural-approach-to-automated-essay-scoring | null | null | https://aclanthology.org/D16-1193 | https://aclanthology.org/D16-1193.pdf | A Neural Approach to Automated Essay Scoring | null | ['Hwee Tou Ng', 'Kaveh Taghipour'] | 2016-11-01 | null | null | null | emnlp-2016-11 | ['automated-essay-scoring'] | ['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] | [-7.42611026763916, 3.758726119995117] |
67f42db2-4cda-460e-b1cf-96016e3b2723 | improving-dependency-parsers-with-supertags | null | null | https://aclanthology.org/E14-4030 | https://aclanthology.org/E14-4030.pdf | Improving Dependency Parsers with Supertags | null | ['Yuji Matsumoto', 'Kevin Duh', 'Hiroki Ouchi'] | 2014-04-01 | null | null | null | eacl-2014-4 | ['transition-based-dependency-parsing'] | ['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] | [-7.354365348815918, 3.66866135597229] |
e7351d1e-7c15-4e48-ba33-1daea16910b2 | learning-to-generate-language-supervised-and | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Zhang_Learning_To_Generate_Language-Supervised_and_Open-Vocabulary_Scene_Graph_Using_Pre-Trained_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Zhang_Learning_To_Generate_Language-Supervised_and_Open-Vocabulary_Scene_Graph_Using_Pre-Trained_CVPR_2023_paper.pdf | Learning To Generate Language-Supervised and Open-Vocabulary Scene Graph Using Pre-Trained Visual-Semantic Space | Scene graph generation (SGG) aims to abstract an image into a graph structure, by representing objects as graph nodes and their relations as labeled edges. However, two knotty obstacles limit the practicability of current SGG methods in real-world scenarios: 1) training SGG models requires time-consuming ground-truth annotations, and 2) the closed-set object categories make the SGG models limited in their ability to recognize novel objects outside of training corpora. To address these issues, we novelly exploit a powerful pre-trained visual-semantic space (VSS) to trigger language-supervised and open-vocabulary SGG in a simple yet effective manner. Specifically, cheap scene graph supervision data can be easily obtained by parsing image language descriptions into semantic graphs. Next, the noun phrases on such semantic graphs are directly grounded over image regions through region-word alignment in the pre-trained VSS. In this way, we enable open-vocabulary object detection by performing object category name grounding with a text prompt in this VSS. On the basis of visually-grounded objects, the relation representations are naturally built for relation recognition, pursuing open-vocabulary SGG. We validate our proposed approach with extensive experiments on the Visual Genome benchmark across various SGG scenarios (i.e., supervised / language-supervised, closed-set / open-vocabulary). Consistent superior performances are achieved compared with existing methods, demonstrating the potential of exploiting pre-trained VSS for SGG in more practical scenarios. | ['Chang-Wen Chen', 'Tao Mei', 'Rui Huang', 'Ting Yao', 'Yingwei Pan', 'Yong Zhang'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['open-vocabulary-object-detection', 'scene-graph-generation', 'word-alignment'] | ['computer-vision', 'computer-vision', 'natural-language-processing'] | [ 5.72647214e-01 6.32882655e-01 -2.74752468e-01 -3.33327621e-01
-5.07246494e-01 -6.49043739e-01 8.33151996e-01 2.98778594e-01
-4.61930335e-02 2.18389317e-01 -2.89908964e-02 -3.17205817e-01
1.28177688e-01 -1.02124429e+00 -9.44694340e-01 -4.64559942e-01
6.88482597e-02 3.17210525e-01 3.94574910e-01 -1.58878639e-01
1.25860617e-01 4.46566075e-01 -1.59463322e+00 1.60548761e-01
8.75005841e-01 9.01471794e-01 6.23174071e-01 3.15410465e-01
-5.18832028e-01 6.17995203e-01 -4.21029955e-01 -5.61289728e-01
3.14143270e-01 -3.61308008e-01 -8.54421914e-01 6.12173557e-01
4.55395460e-01 7.16531053e-02 -2.43459046e-01 1.33210313e+00
-9.05907247e-03 1.15596332e-01 5.39921284e-01 -1.32484066e+00
-8.95859659e-01 4.63319123e-01 -4.57091689e-01 -1.12024769e-01
5.99596322e-01 3.73382717e-01 1.32249808e+00 -9.79213297e-01
1.02540731e+00 1.35579622e+00 2.24121854e-01 4.59639311e-01
-1.15881193e+00 -3.55959624e-01 6.18070960e-01 8.81365389e-02
-1.51616955e+00 -2.79191315e-01 8.52251470e-01 -5.62721729e-01
1.08668578e+00 2.60337323e-01 6.49667263e-01 9.40075338e-01
-1.26850232e-01 7.01622605e-01 8.37818325e-01 -5.52194834e-01
2.06271365e-01 8.88653845e-03 9.88466740e-02 9.67352211e-01
3.95200878e-01 -8.71432796e-02 -5.33110261e-01 1.23653270e-01
7.62494028e-01 -3.26603167e-02 -2.77862906e-01 -8.09431791e-01
-1.32157171e+00 6.97147608e-01 9.10334408e-01 1.07497871e-01
-3.58772188e-01 7.23909512e-02 4.35304046e-01 -5.75917512e-02
4.67344522e-01 3.63564432e-01 -7.08212852e-02 5.57506323e-01
-5.29900312e-01 8.52070004e-03 4.29307729e-01 1.69594049e+00
9.53848064e-01 -5.98082095e-02 -3.03010464e-01 6.47661209e-01
4.88159895e-01 6.20604873e-01 1.22369930e-01 -4.08633947e-01
6.94181561e-01 9.86474276e-01 -3.09943050e-01 -1.20759058e+00
-1.88277766e-01 -3.43324274e-01 -7.62245655e-01 -1.93955377e-01
3.56386125e-01 5.18335462e-01 -1.12320423e+00 1.56017208e+00
5.19880891e-01 1.36368051e-01 2.80441195e-01 8.58831704e-01
1.26327944e+00 5.09689748e-01 4.02024835e-01 -6.10598624e-02
1.74212742e+00 -9.84531403e-01 -5.65204561e-01 -6.68839395e-01
6.94374561e-01 -3.83968294e-01 1.48614621e+00 3.30091268e-02
-5.69487035e-01 -5.85410297e-01 -8.90742242e-01 -1.55448541e-01
-5.74352443e-01 2.86544152e-02 7.27887511e-01 2.54264086e-01
-8.75368416e-01 5.96149042e-02 -5.53129375e-01 -7.42891908e-01
6.85521483e-01 5.71987741e-02 -5.00643075e-01 -2.89978266e-01
-8.93125057e-01 4.27662492e-01 9.47896659e-01 1.06007926e-01
-1.06396902e+00 -3.00922692e-01 -1.33804739e+00 4.22875434e-02
8.30018759e-01 -7.31699646e-01 8.77931774e-01 -8.81194890e-01
-8.83496523e-01 1.36453331e+00 -4.50855009e-02 -2.89931059e-01
1.66260794e-01 8.27502459e-02 -3.07711810e-01 3.57259601e-01
4.00193244e-01 7.59138525e-01 9.25981998e-01 -1.44126260e+00
-5.12631416e-01 -4.72343445e-01 3.75050247e-01 3.42188776e-01
-1.26807690e-01 -4.17901017e-02 -7.86706030e-01 -6.97008491e-01
3.92925173e-01 -8.01550746e-01 -2.77729124e-01 6.34131441e-03
-7.34149456e-01 -3.46454293e-01 5.86591721e-01 -5.84578872e-01
9.70624447e-01 -2.27237582e+00 5.27798757e-02 2.64304310e-01
4.20601487e-01 1.68787181e-01 -2.64819533e-01 3.81285906e-01
-2.02266946e-02 -4.56032194e-02 -1.84358999e-01 -1.53902080e-02
-1.24455400e-01 3.99727762e-01 -4.00776267e-01 1.57246545e-01
4.02446270e-01 1.34295654e+00 -1.30914700e+00 -6.31057322e-01
1.97113901e-01 -1.17914349e-01 -4.93211836e-01 3.98512155e-01
-6.84925377e-01 3.75141531e-01 -6.86128378e-01 7.34071136e-01
4.63161767e-01 -5.93452156e-01 3.24400812e-01 -3.47171724e-01
2.14933068e-01 2.96252996e-01 -1.02706552e+00 1.84831321e+00
-3.49119842e-01 3.76249254e-01 -3.13677579e-01 -1.22863626e+00
1.10367048e+00 -5.88348284e-02 -9.63813663e-02 -6.61827385e-01
-9.83116683e-03 5.11148237e-02 -2.50777334e-01 -5.44709563e-01
2.65160322e-01 3.45834717e-02 -2.12619975e-01 2.30861064e-02
3.18897396e-01 -2.36805290e-01 2.32734531e-01 6.08512282e-01
9.03648555e-01 2.59168446e-01 6.61598623e-01 -2.68242449e-01
5.76479375e-01 2.54578501e-01 3.84925902e-01 7.22625971e-01
-6.36407360e-02 5.92935681e-01 3.74266714e-01 -2.52006292e-01
-7.40356743e-01 -1.23373592e+00 2.32672498e-01 1.03559804e+00
6.45023882e-01 -6.55928791e-01 -8.01659226e-01 -1.03223050e+00
-1.33281901e-01 7.27846980e-01 -4.43153679e-01 -2.48191878e-01
-3.73922527e-01 -2.94071376e-01 3.65265459e-01 5.71062386e-01
4.60760117e-01 -1.24573398e+00 -5.13124406e-01 4.05299962e-02
-1.78669557e-01 -1.63315988e+00 -3.36571097e-01 5.08749038e-02
-6.28503919e-01 -1.12406802e+00 -4.09831464e-01 -1.14981937e+00
1.17943192e+00 6.22619390e-01 1.17274630e+00 1.76683635e-01
-3.66219759e-01 5.74258029e-01 -5.38059771e-01 -3.01811129e-01
-5.14985085e-01 -1.75923377e-01 -2.40144327e-01 2.36433759e-01
3.07218701e-01 -2.05766290e-01 -5.13433576e-01 1.92160711e-01
-8.42689514e-01 6.02376819e-01 5.73033154e-01 6.62363231e-01
9.34220672e-01 -1.09834895e-01 4.34472829e-01 -1.09411001e+00
9.69887823e-02 -2.42022291e-01 -9.11783814e-01 6.33456051e-01
-4.54363078e-01 6.84361085e-02 5.46070457e-01 -3.03289175e-01
-1.01691091e+00 9.59496424e-02 1.17388450e-01 -4.19528872e-01
-2.84265876e-01 5.38638353e-01 -4.71906871e-01 5.68032824e-02
6.71837568e-01 4.29177850e-01 -1.44880578e-01 -1.59786254e-01
9.47642565e-01 4.37248439e-01 7.48850644e-01 -5.20090342e-01
1.16936839e+00 5.09408295e-01 -4.56027091e-02 -7.39958167e-01
-1.01968420e+00 -7.69564509e-01 -7.30490148e-01 -1.65499389e-01
1.14561045e+00 -1.04022741e+00 -2.82843530e-01 1.98222682e-01
-1.07276905e+00 -2.87234694e-01 -3.40294451e-01 1.39776587e-01
-5.55735111e-01 4.56202358e-01 -2.43144408e-01 -5.70411503e-01
-2.67570853e-01 -9.44790363e-01 1.48590457e+00 1.82611749e-01
2.12067477e-02 -9.46070194e-01 -4.93250698e-01 4.88633394e-01
-1.50631234e-01 2.71020681e-01 1.08282149e+00 -7.23425746e-01
-1.04815066e+00 -7.86725059e-02 -6.78121090e-01 1.80023044e-01
2.74351031e-01 -3.25280845e-01 -9.53467906e-01 -2.41273835e-01
-4.00492013e-01 -3.31885338e-01 5.88176966e-01 -5.38532110e-03
1.09075308e+00 -2.60333687e-01 -6.50710821e-01 6.33075714e-01
1.49807632e+00 6.08842000e-02 3.97957861e-01 2.11953163e-01
1.11694074e+00 6.30131423e-01 9.66157198e-01 1.61710382e-01
4.32594001e-01 6.31358624e-01 5.28458595e-01 -4.18844640e-01
-4.60322827e-01 -7.30521560e-01 1.06402650e-01 5.95198035e-01
1.01811595e-01 -3.83732378e-01 -1.04077911e+00 6.25251591e-01
-1.74303257e+00 -5.80104947e-01 -3.21573019e-01 1.99482000e+00
6.40111268e-01 1.62507579e-01 -2.08081797e-01 -1.93133205e-01
7.37145007e-01 2.18681186e-01 -4.06293899e-01 6.62406832e-02
-2.48155743e-01 -4.93499041e-02 4.70075548e-01 2.15111628e-01
-1.13594198e+00 1.43100023e+00 4.80382252e+00 8.31478417e-01
-8.27859104e-01 1.43869787e-01 4.85110939e-01 4.07138616e-01
-4.60683435e-01 3.05249929e-01 -8.82706285e-01 -3.72880779e-04
3.32485855e-01 -1.93985537e-01 4.23208147e-01 9.46797431e-01
-7.91168809e-02 5.21096066e-02 -1.21081018e+00 1.23701942e+00
2.57938683e-01 -1.43243694e+00 5.07531524e-01 -3.52355354e-02
6.49377406e-01 -6.44480959e-02 -3.08294803e-01 3.59977514e-01
2.93022722e-01 -7.70540237e-01 9.57469821e-01 4.30812389e-02
1.10180342e+00 -1.98781610e-01 4.12452310e-01 3.18137079e-01
-1.54460502e+00 1.03056923e-01 -4.09468055e-01 6.58025369e-02
2.04714984e-01 3.75740945e-01 -1.13012874e+00 9.62094247e-01
4.73388672e-01 7.10146844e-01 -8.57736111e-01 6.92054510e-01
-7.09357440e-01 5.76043785e-01 -9.76496711e-02 6.46724924e-02
3.88398290e-01 -3.73410374e-01 5.06056547e-01 1.05473423e+00
3.11140567e-02 2.00529829e-01 4.86640543e-01 1.10471368e+00
-2.99118757e-02 2.32118651e-01 -1.06161559e+00 -2.31141567e-01
3.17059457e-01 1.30100298e+00 -1.27858698e+00 -5.45711696e-01
-7.66008735e-01 9.77737844e-01 4.92499918e-01 6.00225925e-01
-8.09706211e-01 -2.32192799e-01 3.43075663e-01 2.56777257e-01
2.95025080e-01 -1.16027057e-01 -1.75538853e-01 -1.24889648e+00
1.92746669e-01 -7.96841562e-01 6.11891866e-01 -9.70339954e-01
-1.19684982e+00 6.13705456e-01 2.93571591e-01 -1.07445562e+00
-6.90438747e-02 -6.66409850e-01 -1.58030570e-01 4.70504850e-01
-1.37790775e+00 -1.60825264e+00 -6.77122355e-01 7.33966827e-01
6.91856086e-01 -1.17557095e-02 7.19250798e-01 1.49749452e-02
-3.75224173e-01 4.42616105e-01 -4.43958372e-01 3.36787015e-01
2.40559533e-01 -1.15867233e+00 8.22633266e-01 1.18926096e+00
7.68943310e-01 6.02728307e-01 4.44975108e-01 -9.40942645e-01
-1.34624445e+00 -1.57274711e+00 7.74516582e-01 -4.36657161e-01
6.62632048e-01 -1.05086982e+00 -9.58069026e-01 8.54990721e-01
-1.37401506e-01 2.96299845e-01 2.74727583e-01 5.88217415e-02
-6.38575315e-01 1.56704322e-01 -8.00931036e-01 8.39644611e-01
1.80056024e+00 -7.64193773e-01 -6.63521528e-01 7.30178654e-01
1.02388012e+00 -4.99133259e-01 -5.49500108e-01 4.82369572e-01
-4.23531793e-02 -4.45273697e-01 1.10271454e+00 -6.77559614e-01
2.30246291e-01 -5.74110627e-01 -2.64603943e-01 -9.88162518e-01
-1.91995502e-01 -4.90058035e-01 8.66761953e-02 1.32473361e+00
2.01543018e-01 -6.59743845e-01 5.16520023e-01 2.38123134e-01
-3.19585830e-01 -4.95661348e-01 -6.03316128e-01 -1.02658129e+00
-4.79598284e-01 -4.79051292e-01 5.63452244e-01 9.85527813e-01
-9.86846536e-02 5.74955404e-01 1.37723625e-01 4.12228972e-01
6.22182310e-01 4.25335228e-01 9.43901479e-01 -1.11607587e+00
-2.77058214e-01 -2.00605541e-01 -8.36108029e-01 -1.19400930e+00
2.49283344e-01 -1.35029447e+00 2.19906300e-01 -1.83743286e+00
3.46252501e-01 -6.04977131e-01 -1.63377356e-02 7.35431075e-01
-4.20459121e-01 2.43125945e-01 3.02114636e-01 5.66971302e-02
-8.81990552e-01 4.80327964e-01 1.31644917e+00 -3.75112325e-01
2.78091338e-02 -4.42476362e-01 -6.74999595e-01 8.33158851e-01
4.58718359e-01 -3.44227433e-01 -8.99468064e-01 -3.57931852e-01
1.07685857e-01 -1.65674090e-01 7.06186712e-01 -6.42874062e-01
4.87843826e-02 -3.50384802e-01 -8.89455676e-02 -3.71292830e-01
-1.65335461e-02 -6.04274452e-01 1.45765334e-01 1.78997248e-01
-1.87286690e-01 -4.36829358e-01 5.02559319e-02 9.07244921e-01
-2.14250609e-01 -9.28698778e-02 5.34999430e-01 -2.04626828e-01
-1.32634699e+00 2.81946182e-01 -1.31396726e-02 3.07982355e-01
1.22751665e+00 -4.93686169e-01 -2.99671859e-01 -1.29543036e-01
-7.28268623e-01 1.30929023e-01 6.14661813e-01 6.40224636e-01
7.53140211e-01 -1.22578335e+00 -3.19026440e-01 4.51747537e-01
1.03428364e+00 4.75990236e-01 8.54889229e-02 4.89054918e-01
-4.09731627e-01 3.59577149e-01 9.99903455e-02 -9.82539952e-01
-1.25720882e+00 1.04122603e+00 6.56236261e-02 -1.26915753e-01
-1.02763796e+00 1.01718092e+00 1.15531409e+00 -4.33182359e-01
1.40902773e-01 -6.07023239e-01 -9.36174989e-02 -2.79620677e-01
1.79073408e-01 -3.33391815e-01 -7.32289487e-03 -7.73265123e-01
-4.09575820e-01 5.69768727e-01 9.26009491e-02 2.40944669e-01
9.61424649e-01 -2.35511824e-01 1.28328260e-02 2.61341810e-01
9.63864565e-01 -1.31503001e-01 -1.04849780e+00 -5.55553615e-01
3.41779858e-01 -4.59848344e-01 -2.12496713e-01 -4.79888022e-01
-9.00491416e-01 6.36817336e-01 2.91813970e-01 6.43691570e-02
1.18791294e+00 6.14121258e-01 4.71946627e-01 4.31206912e-01
8.03913057e-01 -6.80154383e-01 2.80842692e-01 1.55667856e-01
9.15749013e-01 -1.27011931e+00 -1.28422439e-01 -1.25202334e+00
-7.04019725e-01 7.62003660e-01 6.43694103e-01 7.90818706e-02
2.69865781e-01 -2.57219970e-01 -1.22904005e-02 -7.67737448e-01
-4.45414543e-01 -5.88680446e-01 6.49383724e-01 8.18365455e-01
2.59727389e-02 2.09694162e-01 -7.56748812e-03 5.04241645e-01
-7.07389861e-02 -2.64917731e-01 2.88640112e-01 7.50764787e-01
-2.95133948e-01 -7.92304397e-01 -1.01756170e-01 3.46626043e-01
4.68081944e-02 -2.35256627e-01 -3.81157964e-01 8.83983314e-01
5.48845157e-02 8.68243277e-01 -6.67636655e-03 -9.63633042e-03
4.38372761e-01 -4.90707792e-02 4.45909262e-01 -1.22008133e+00
-1.35848954e-01 1.23291668e-02 2.55530447e-01 -7.79318750e-01
-3.52491319e-01 -4.35796738e-01 -1.48781526e+00 4.69440669e-01
-5.04697978e-01 -1.40450448e-01 3.34717810e-01 9.88667727e-01
4.48020548e-01 5.43356419e-01 2.55597502e-01 -4.56844687e-01
-2.31555283e-01 -5.94632268e-01 -4.55781490e-01 1.00820005e+00
-1.73772462e-02 -8.78740013e-01 -6.64563105e-02 4.34345812e-01] | [10.372176170349121, 1.6105009317398071] |
fddc475b-8485-45a9-9c44-16a8614523ea | pso-convolutional-neural-networks-with | 2205.10456 | null | https://arxiv.org/abs/2205.10456v2 | https://arxiv.org/pdf/2205.10456v2.pdf | PSO-Convolutional Neural Networks with Heterogeneous Learning Rate | Convolutional Neural Networks (ConvNets or CNNs) have been candidly deployed in the scope of computer vision and related fields. Nevertheless, the dynamics of training of these neural networks lie still elusive: it is hard and computationally expensive to train them. A myriad of architectures and training strategies have been proposed to overcome this challenge and address several problems in image processing such as speech, image and action recognition as well as object detection. In this article, we propose a novel Particle Swarm Optimization (PSO) based training for ConvNets. In such framework, the vector of weights of each ConvNet is typically cast as the position of a particle in phase space whereby PSO collaborative dynamics intertwines with Stochastic Gradient Descent (SGD) in order to boost training performance and generalization. Our approach goes as follows: i) [regular phase] each ConvNet is trained independently via SGD; ii) [collaborative phase] ConvNets share among themselves their current vector of weights (or particle-position) along with their gradient estimates of the Loss function. Distinct step sizes are coined by distinct ConvNets. By properly blending ConvNets with large (possibly random) step-sizes along with more conservative ones, we propose an algorithm with competitive performance with respect to other PSO-based approaches on Cifar-10 (accuracy of 98.31%). These accuracy levels are obtained by resorting to only four ConvNets -- such results are expected to scale with the number of collaborative ConvNets accordingly. We make our source codes available for download https://github.com/leonlha/PSO-ConvNet-Dynamics. | ['Bernardete Ribeiro', 'Augusto Santos', 'Nguyen Huu Phong'] | 2022-05-20 | null | null | null | null | ['pso-convnets-dynamics-1', 'pso-convnets-dynamics-2'] | ['computer-vision', 'computer-vision'] | [ 5.74794300e-02 -1.89101592e-01 1.47495747e-01 -1.47041798e-01
-8.63175169e-02 -3.26355755e-01 6.79996133e-01 1.51288047e-01
-1.05074561e+00 6.93967223e-01 -4.94882137e-01 -1.27643257e-01
-3.58630657e-01 -7.33288407e-01 -7.56401896e-01 -1.11530435e+00
-1.77669004e-01 3.44647467e-01 4.14018244e-01 -2.70286560e-01
1.87652305e-01 6.49668753e-01 -1.54752648e+00 -1.16196558e-01
7.75851846e-01 1.21339214e+00 5.19372880e-01 8.14282835e-01
1.21202499e-01 6.37511909e-01 -7.24372029e-01 -3.06592852e-01
2.89495856e-01 -9.66514796e-02 -4.21071678e-01 -2.31308024e-02
2.40157709e-01 2.51524687e-01 -4.50327963e-01 1.27280581e+00
5.31112134e-01 3.03580344e-01 2.64610231e-01 -1.05785465e+00
-4.24310476e-01 5.18654704e-01 -3.07781637e-01 6.03716850e-01
-4.45974737e-01 1.59997076e-01 8.18118572e-01 -8.20537984e-01
3.65389973e-01 8.00185382e-01 9.49529946e-01 7.33334780e-01
-1.10045362e+00 -5.04079580e-01 2.50083894e-01 1.84988335e-01
-1.26211238e+00 -2.47806489e-01 8.17756236e-01 -1.97269097e-01
9.53265071e-01 1.32090673e-01 1.01475179e+00 1.12338543e+00
1.69779494e-01 7.17432439e-01 7.03150690e-01 -3.29622030e-01
5.43932021e-01 2.32765079e-01 1.35252506e-01 6.87368810e-01
3.83218169e-01 2.58544963e-02 -4.01910752e-01 8.77268165e-02
6.87544286e-01 2.78757848e-02 -1.68060794e-01 -2.57841617e-01
-1.01877987e+00 9.83475327e-01 7.34396219e-01 6.65337861e-01
-6.52015328e-01 4.96407747e-01 1.96406350e-01 2.35402480e-01
3.52373511e-01 4.75672841e-01 -5.89416385e-01 -1.13294236e-01
-9.28838491e-01 2.83489853e-01 8.02027583e-01 4.63808298e-01
6.66899681e-01 2.77567953e-01 -4.08032537e-03 9.47432756e-01
2.33928904e-01 3.13956171e-01 6.24848604e-01 -7.08670199e-01
4.34201032e-01 5.01971960e-01 6.67449683e-02 -1.26635373e+00
-6.24032319e-01 -1.04450655e+00 -1.02924275e+00 3.58058900e-01
4.75164294e-01 -4.89998668e-01 -7.75460541e-01 1.63313639e+00
2.96524346e-01 4.49074745e-01 7.18659312e-02 5.83029449e-01
5.89734316e-01 5.71328104e-01 7.68732056e-02 -5.78817464e-02
1.18335605e+00 -1.11980772e+00 -2.51500219e-01 -3.36705357e-01
2.19929054e-01 -5.07321835e-01 5.89857936e-01 4.32603925e-01
-1.08019662e+00 -7.94544697e-01 -1.09549797e+00 4.89712000e-01
-4.53654081e-01 6.12812638e-01 5.59740007e-01 8.95484149e-01
-1.39221120e+00 1.08906114e+00 -1.28709602e+00 -4.88241874e-02
9.84060109e-01 8.42514634e-01 7.66259134e-02 5.20445228e-01
-8.00703526e-01 6.63816333e-01 4.93524700e-01 4.22235310e-01
-6.43717229e-01 -6.73708677e-01 -4.43007112e-01 1.29642382e-01
-3.34461257e-02 -5.64189851e-01 9.06290472e-01 -1.13883221e+00
-1.58428991e+00 5.24436653e-01 4.98242229e-02 -1.01822877e+00
6.67953253e-01 -1.37433335e-01 -5.59885204e-02 -1.19197229e-03
-2.21122161e-01 8.29018652e-01 1.04427445e+00 -1.02698648e+00
-8.55000317e-01 -1.20007582e-01 8.40555951e-02 1.20031707e-01
-7.83549607e-01 -7.50099570e-02 -3.33737791e-01 -5.53083181e-01
1.66170403e-01 -1.00879955e+00 -5.10080516e-01 1.99871920e-02
-1.96748957e-01 -3.45962375e-01 8.03465128e-01 -4.70921695e-02
1.09365654e+00 -1.97641313e+00 1.21262394e-01 -2.43595093e-02
3.03464353e-01 8.66745710e-01 -1.37347534e-01 1.27309322e-01
2.80795954e-02 2.33071838e-02 -1.52955472e-01 -7.67117143e-01
3.61284427e-02 2.60592490e-01 -1.40053213e-01 6.01562142e-01
3.92214477e-01 9.99087811e-01 -8.99317384e-01 -4.85050827e-02
3.94378722e-01 8.87995481e-01 -6.24559879e-01 -2.93067455e-01
-1.11827187e-01 4.13571984e-01 -6.34268463e-01 2.72758067e-01
7.79426754e-01 -4.31839764e-01 8.15347582e-02 -2.65622854e-01
-3.00828189e-01 -7.41851702e-02 -1.11543739e+00 1.34711838e+00
-3.48980904e-01 7.97789991e-01 8.63173977e-03 -1.36075437e+00
9.51379299e-01 5.75613528e-02 5.76468170e-01 -4.08722818e-01
3.65098476e-01 1.12957232e-01 2.21905515e-01 -1.22802027e-01
4.25037056e-01 2.66116440e-01 4.36990380e-01 8.91923755e-02
2.95407802e-01 1.61117792e-01 3.53611767e-01 -2.84273177e-01
1.16383410e+00 -1.13495432e-01 -2.13087499e-02 -2.48180985e-01
7.68341780e-01 6.92135841e-02 3.95219892e-01 1.02053583e+00
-2.56518811e-01 5.05920231e-01 2.67590731e-01 -7.51591086e-01
-1.01884222e+00 -7.60772645e-01 -3.45978677e-01 9.63084042e-01
5.57543151e-02 -1.12787679e-01 -7.83300400e-01 -4.85655904e-01
-2.32632160e-01 3.50893140e-01 -7.43036985e-01 -5.37306480e-02
-7.88005233e-01 -1.26528037e+00 6.18639231e-01 3.54365259e-01
7.37815201e-01 -1.30279410e+00 -1.05966198e+00 5.71574152e-01
5.15848279e-01 -1.05321443e+00 6.06363229e-02 5.37028849e-01
-9.07773852e-01 -1.03645861e+00 -8.50203753e-01 -9.29035485e-01
6.94858193e-01 -3.28124664e-03 9.44715679e-01 3.71980757e-01
-2.68562436e-01 1.72121525e-01 -3.50731671e-01 -4.12238926e-01
-2.83013374e-01 2.05323562e-01 7.62221143e-02 1.87972754e-01
3.04645374e-02 -8.28305244e-01 -8.70181739e-01 2.10807517e-01
-7.24423647e-01 -1.71766609e-01 4.97565746e-01 9.80843246e-01
5.98709643e-01 2.89061487e-01 3.31577450e-01 -6.53383613e-01
6.13407016e-01 -2.99249887e-01 -8.73663545e-01 3.47511172e-02
-5.03172159e-01 -2.15928420e-01 8.95867705e-01 -7.97298253e-01
-6.66270554e-01 1.28282726e-01 -4.52418536e-01 -5.95190465e-01
-1.89337909e-01 2.38759369e-01 4.48872387e-01 -5.89816391e-01
6.19829059e-01 3.94250572e-01 -1.71692356e-01 -2.58899480e-01
-6.55525923e-02 1.36519670e-01 3.48205566e-01 -2.09844008e-01
8.58166635e-01 5.74693680e-01 -5.49177267e-02 -7.61427581e-01
-9.01336670e-01 -1.77926019e-01 -3.81818533e-01 -2.01452360e-01
8.57149899e-01 -6.31123245e-01 -9.93524730e-01 8.53905857e-01
-1.05862284e+00 -5.58440447e-01 -4.38531846e-01 5.99732459e-01
-2.06893578e-01 -1.70178954e-02 -5.33148110e-01 -7.49016583e-01
-4.76449043e-01 -1.18556273e+00 6.25343680e-01 7.80174553e-01
2.99448669e-01 -1.40573645e+00 -1.42515991e-02 2.88028456e-03
7.77400434e-01 1.91549495e-01 3.36288989e-01 -7.50204086e-01
-2.86155343e-01 -2.22982302e-01 -1.80430889e-01 6.89518213e-01
-4.73779440e-02 -5.47002535e-04 -9.22985852e-01 -4.28083152e-01
3.00025672e-01 -1.99578375e-01 8.42210948e-01 8.13839614e-01
1.16842699e+00 -1.73703954e-01 -3.57704043e-01 7.65944123e-01
1.59349895e+00 2.35952899e-01 3.87580365e-01 5.53379416e-01
5.79038024e-01 1.88086003e-01 1.26049981e-01 6.02093756e-01
1.41975150e-01 5.38278639e-01 5.94891608e-01 -3.27114798e-02
-1.56787977e-01 1.25559703e-01 1.01964816e-01 6.75948143e-01
-2.28110880e-01 -3.91184151e-01 -9.06765580e-01 3.84447128e-01
-1.82355833e+00 -1.02749848e+00 -1.01834387e-01 1.78126144e+00
4.90503699e-01 4.38984603e-01 -1.57041281e-01 1.90711603e-01
6.58531070e-01 3.06646138e-01 -6.62222326e-01 2.77998950e-03
-2.94622928e-01 4.50892299e-01 6.04067087e-01 3.31339002e-01
-1.28902447e+00 8.67757618e-01 5.10208797e+00 8.29625726e-01
-1.49249327e+00 2.13497892e-01 7.45068312e-01 -6.49279132e-02
2.49782950e-01 -2.26987258e-01 -1.17982817e+00 6.27319038e-01
7.79308379e-01 1.18118957e-01 4.61125255e-01 8.21069300e-01
1.12440787e-01 -1.71872839e-01 -6.44695401e-01 1.08229697e+00
-3.41288894e-02 -1.77567816e+00 -2.50108570e-01 -1.37713021e-02
1.03991890e+00 6.72466040e-01 3.24723065e-01 2.31790930e-01
3.42098951e-01 -9.29590464e-01 7.22441673e-01 3.94828588e-01
1.08924888e-01 -6.73598289e-01 7.01112807e-01 4.48254347e-01
-1.17254210e+00 -4.30676758e-01 -6.56012237e-01 5.32175694e-03
-4.04521860e-02 6.08989596e-01 -5.51850677e-01 3.10987532e-01
9.28183496e-01 7.83053756e-01 -5.84330738e-01 1.30689847e+00
2.25492902e-02 7.30610371e-01 -4.89052117e-01 -5.04029274e-01
6.72300100e-01 -3.19032788e-01 6.56472623e-01 1.01990485e+00
3.32859814e-01 -3.42395723e-01 4.94720079e-02 8.13987494e-01
-1.23742297e-01 -1.39226124e-01 -6.90922141e-02 1.24713048e-01
5.01963377e-01 1.45501733e+00 -1.22017884e+00 -2.81472445e-01
-1.13040760e-01 7.12225378e-01 4.44656789e-01 2.85470873e-01
-9.81471837e-01 -2.93770611e-01 5.84048748e-01 -2.92014569e-01
8.21408749e-01 -2.83049554e-01 -2.73507416e-01 -1.00069773e+00
-3.63581590e-02 -4.35571492e-01 2.71812230e-02 -3.60575676e-01
-1.07992709e+00 1.13189816e+00 -3.49187016e-01 -1.10696721e+00
4.15212363e-02 -8.17545593e-01 -7.29131579e-01 5.93821943e-01
-1.70551729e+00 -7.20900118e-01 -3.20486695e-01 5.69594085e-01
6.46427631e-01 -4.01311904e-01 5.91822743e-01 3.86549711e-01
-6.68713510e-01 4.58143979e-01 2.96824098e-01 1.90941975e-01
-5.21899983e-02 -1.13390946e+00 4.85149264e-01 6.54399097e-01
4.41267818e-01 4.55106497e-01 8.11853468e-01 -1.71066359e-01
-1.26303446e+00 -1.16209340e+00 6.72947109e-01 -1.85581520e-01
9.93164241e-01 -3.24168563e-01 -6.75997853e-01 1.72736526e-01
8.83766264e-02 4.59215850e-01 1.28830552e-01 -2.14817569e-01
2.69636571e-01 -3.73334259e-01 -9.07745004e-01 4.64814633e-01
9.79858935e-01 1.25363823e-02 -2.56850183e-01 6.35621130e-01
3.75430971e-01 -6.06732488e-01 -5.91341317e-01 2.38931388e-01
2.49782667e-01 -1.27102578e+00 1.04561186e+00 -2.91764736e-01
-1.35243451e-03 -2.59622782e-01 7.37600625e-02 -1.28571951e+00
-1.03979453e-01 -5.88901401e-01 -4.54322957e-02 8.06215227e-01
4.09509391e-01 -1.04971170e+00 1.22562826e+00 1.31199390e-01
-2.13505819e-01 -1.07552707e+00 -1.11494398e+00 -7.25974023e-01
2.52131373e-02 -4.85096663e-01 1.89696968e-01 5.47854245e-01
-7.53652811e-01 8.31390964e-04 -7.65516981e-02 3.40811968e-01
6.62908375e-01 -2.75105566e-01 3.63991559e-01 -1.36781609e+00
-3.38241279e-01 -8.31074357e-01 -5.36287725e-01 -1.04638648e+00
6.73315004e-02 -6.87041104e-01 -1.28762037e-01 -1.26425231e+00
-2.52935439e-01 -9.40466046e-01 -4.34687376e-01 5.18593609e-01
1.75662607e-01 4.55579817e-01 3.19689870e-01 2.21516103e-01
-6.47443891e-01 4.85836595e-01 9.86615837e-01 -7.87905976e-02
-3.78738880e-01 4.25343394e-01 -2.71959156e-01 8.16176772e-01
1.09783053e+00 -7.71758258e-01 -3.93193036e-01 -5.98026335e-01
2.61000156e-01 -2.54679173e-01 7.14299798e-01 -1.67225432e+00
7.89461851e-01 2.63369352e-01 5.13533294e-01 -4.39875543e-01
3.94389182e-01 -6.43914580e-01 2.12051392e-01 8.22353721e-01
-1.73535094e-01 1.85805097e-01 2.09294841e-01 5.51445484e-01
-2.37143844e-01 -6.03415608e-01 8.95198762e-01 -1.13046505e-01
-4.70892727e-01 5.42780936e-01 -3.00275534e-01 1.30589092e-02
1.00979924e+00 -2.98663884e-01 -1.39398620e-01 7.10788518e-02
-8.27930391e-01 3.47924642e-02 1.34043574e-01 3.29017699e-01
5.33749759e-01 -9.22673821e-01 -5.52215755e-01 1.69316888e-01
-5.92201591e-01 2.35715602e-03 2.98051029e-01 1.05810571e+00
-5.59910476e-01 2.65508473e-01 -2.40764141e-01 -8.61045301e-01
-1.14187717e+00 6.06741905e-02 7.75101662e-01 -3.92491221e-01
-5.71813166e-01 1.22755003e+00 -2.21318394e-01 -2.68981874e-01
4.00786281e-01 -1.78310990e-01 -4.44375724e-01 1.19559407e-01
4.77143735e-01 1.53842032e-01 2.58185446e-01 -4.75126654e-01
-3.44410121e-01 5.29132545e-01 -1.02485746e-01 2.46100217e-01
1.83740568e+00 6.76681325e-02 4.09783646e-02 3.80313024e-02
1.24471438e+00 -2.70287693e-01 -1.69947517e+00 -4.43131998e-02
5.81352077e-02 -5.83567917e-02 2.48056009e-01 -4.95106786e-01
-1.51344037e+00 6.73961043e-01 8.61527741e-01 4.49758232e-01
1.04650426e+00 -9.07340497e-02 6.25108659e-01 5.50288916e-01
1.79073572e-01 -9.89821136e-01 2.34836861e-01 7.78068960e-01
5.83874464e-01 -1.09444380e+00 -1.11178830e-01 1.93876401e-01
-3.64470810e-01 1.29722452e+00 3.12549710e-01 -7.09594905e-01
8.85517538e-01 4.00052577e-01 3.88648398e-02 -2.48964205e-01
-8.20824146e-01 -1.89168960e-01 1.57834008e-01 3.79107982e-01
3.16963345e-02 -2.66459465e-01 -2.32726768e-01 3.37205499e-01
-2.29000852e-01 -6.56720847e-02 1.31415114e-01 8.89265358e-01
-4.18521762e-01 -1.03223968e+00 -6.35642186e-02 3.52089375e-01
-4.92627352e-01 -1.97665185e-01 2.15540320e-01 6.88820124e-01
2.89705068e-01 8.62075210e-01 1.68853834e-01 -4.11190331e-01
1.29336894e-01 -5.33609688e-01 3.48938465e-01 -2.92837769e-01
-9.53695178e-01 -1.71284348e-01 -2.26527020e-01 -4.20001745e-01
-6.62821651e-01 -6.99387968e-01 -1.19830608e+00 -1.73829824e-01
-3.01579267e-01 2.04772741e-01 8.67832899e-01 8.69515657e-01
1.85198173e-01 6.16695225e-01 6.75616860e-01 -1.33699417e+00
-6.58156991e-01 -6.91492260e-01 -2.76306301e-01 -1.96590903e-03
2.78881699e-01 -6.05554760e-01 -5.29017210e-01 -8.35642889e-02] | [8.426104545593262, 3.1744253635406494] |
2bc41dda-a1b3-4f23-b20a-c5ce6731f244 | dt2i-dense-text-to-image-generation-from | 2204.02035 | null | https://arxiv.org/abs/2204.02035v1 | https://arxiv.org/pdf/2204.02035v1.pdf | DT2I: Dense Text-to-Image Generation from Region Descriptions | Despite astonishing progress, generating realistic images of complex scenes remains a challenging problem. Recently, layout-to-image synthesis approaches have attracted much interest by conditioning the generator on a list of bounding boxes and corresponding class labels. However, previous approaches are very restrictive because the set of labels is fixed a priori. Meanwhile, text-to-image synthesis methods have substantially improved and provide a flexible way for conditional image generation. In this work, we introduce dense text-to-image (DT2I) synthesis as a new task to pave the way toward more intuitive image generation. Furthermore, we propose DTC-GAN, a novel method to generate images from semantically rich region descriptions, and a multi-modal region feature matching loss to encourage semantic image-text matching. Our results demonstrate the capability of our approach to generate plausible images of complex scenes using region captions. | ['Andreas Dengel', 'Jörn Hees', 'Prateek Bansal', 'Stanislav Frolov'] | 2022-04-05 | null | null | null | null | ['conditional-image-generation'] | ['computer-vision'] | [ 6.77046120e-01 2.38627940e-01 -1.21188775e-01 -4.69633371e-01
-9.83741045e-01 -7.33266950e-01 8.93088698e-01 -3.41011345e-01
1.79147393e-01 9.16148126e-01 2.40192801e-01 -2.94636171e-02
2.52988249e-01 -9.74869490e-01 -9.59765494e-01 -5.36518574e-01
5.34922361e-01 4.45139587e-01 1.13908164e-01 -2.37422898e-01
1.31246924e-01 5.78284204e-01 -1.43890858e+00 3.87587994e-01
9.85285401e-01 7.91992545e-01 2.89648980e-01 4.07283902e-01
-2.39858612e-01 5.45077145e-01 -6.12222552e-01 -4.50854778e-01
2.91746825e-01 -8.50192010e-01 -6.55285716e-01 3.53911728e-01
5.58557034e-01 -2.05886558e-01 -1.58719346e-01 8.43028188e-01
4.20823514e-01 8.89234245e-02 8.21738064e-01 -1.41218805e+00
-7.18720257e-01 4.98840809e-01 -6.22819424e-01 -5.33327997e-01
4.17793781e-01 1.67379469e-01 9.43763733e-01 -8.21304321e-01
8.13472807e-01 1.26734090e+00 2.59308070e-01 8.10653746e-01
-1.42093837e+00 -6.57980919e-01 2.60189086e-01 -2.03417107e-01
-1.35322213e+00 -3.79803985e-01 8.62497747e-01 -4.96465176e-01
4.27444279e-01 3.64829838e-01 6.62774682e-01 1.12453687e+00
-2.13083938e-01 9.18450415e-01 1.11280835e+00 -5.96530378e-01
2.36532271e-01 3.30781341e-01 -6.09077454e-01 7.23979652e-01
8.87350738e-02 3.21448371e-02 -4.14949566e-01 -4.83887903e-02
1.26218796e+00 6.88731950e-03 -3.29151869e-01 -5.05069375e-01
-1.34326899e+00 8.29815626e-01 5.35703599e-01 5.35632372e-02
-2.47450992e-01 4.25259978e-01 8.90503377e-02 -3.02716166e-01
4.82297927e-01 5.84511578e-01 2.16252394e-02 2.39235103e-01
-1.02580953e+00 5.68443477e-01 3.53996187e-01 1.18854046e+00
8.03718626e-01 2.32904196e-01 -5.36653757e-01 8.79597247e-01
-6.76782196e-03 6.78477705e-01 3.91925164e-02 -1.06883621e+00
4.59318876e-01 3.92237812e-01 2.91222513e-01 -8.87378931e-01
4.20394242e-02 -3.53082299e-01 -8.97247493e-01 1.60109103e-01
3.43408227e-01 -8.81543010e-03 -1.17212307e+00 1.81607366e+00
2.68866807e-01 -9.22169015e-02 -5.49089462e-02 9.34834063e-01
3.67883831e-01 9.18558359e-01 -5.53127155e-02 2.09700763e-01
9.92136836e-01 -1.14845812e+00 -7.38934398e-01 -1.68731555e-01
2.61454880e-01 -6.77737951e-01 1.26014471e+00 1.96794063e-01
-1.25772417e+00 -4.91250873e-01 -6.41267598e-01 1.07404506e-02
-2.82654345e-01 1.88303664e-01 6.65737629e-01 5.29764354e-01
-9.91458178e-01 1.52670950e-01 -4.11591977e-01 -2.29846060e-01
5.74096262e-01 1.27116725e-01 -2.56694883e-01 -2.85536051e-01
-9.12087679e-01 6.62754297e-01 5.27307689e-01 5.99468406e-03
-9.98273015e-01 -5.81085145e-01 -1.00671422e+00 -9.50327050e-03
3.93929571e-01 -9.38074529e-01 1.01405287e+00 -1.10546589e+00
-1.50738764e+00 8.82430494e-01 -2.45659471e-01 -2.22222999e-01
5.90148151e-01 1.20096616e-01 -7.12738112e-02 2.05236867e-01
2.48821735e-01 1.36724019e+00 9.35659349e-01 -1.75774705e+00
-4.45453048e-01 -6.30670832e-03 6.25698864e-02 6.99040145e-02
-9.85976756e-02 -1.14860095e-01 -5.39409399e-01 -1.00224733e+00
-7.86737502e-02 -8.83564830e-01 -5.31021953e-01 2.46731922e-01
-7.33596265e-01 5.55900037e-02 5.64541996e-01 -3.89620632e-01
8.72402430e-01 -1.98123002e+00 7.96648487e-02 2.39982396e-01
4.15995643e-02 7.29123428e-02 -3.47591490e-01 4.54994082e-01
-7.80073460e-04 3.86006743e-01 -4.44105983e-01 -3.78147155e-01
1.29085392e-01 6.11209124e-02 -6.18188143e-01 1.25146732e-01
3.83134127e-01 1.08702981e+00 -8.56556594e-01 -6.49965107e-01
3.12208146e-01 5.35892427e-01 -7.63686955e-01 3.06507289e-01
-6.70250118e-01 8.00850511e-01 -6.12138212e-01 5.06330431e-01
7.09697008e-01 -3.99407387e-01 -8.19821730e-02 -1.16351694e-01
1.34973712e-02 -1.92478582e-01 -7.20577478e-01 1.79161429e+00
-5.89996397e-01 4.52911556e-01 -1.01880141e-01 -7.87876248e-01
9.35252309e-01 1.59058392e-01 4.86482650e-01 -5.23746252e-01
2.38288362e-02 1.10373013e-01 -3.33089292e-01 -1.39458880e-01
5.90952516e-01 -3.88961881e-01 -2.35861555e-01 3.53627771e-01
-5.47748566e-01 -7.20067322e-01 1.42935589e-01 2.61956930e-01
5.18405378e-01 4.26938355e-01 -2.31463043e-03 6.24537421e-03
4.09126103e-01 1.02578647e-01 5.29767394e-01 5.88040650e-01
1.93973839e-01 1.35747027e+00 4.01459098e-01 -2.64389873e-01
-1.31857860e+00 -1.29447103e+00 -7.82838184e-03 4.60434914e-01
2.30704919e-01 -2.12657839e-01 -1.00714493e+00 -6.84921026e-01
-2.39524543e-01 8.47659826e-01 -5.64308345e-01 2.70247888e-02
-6.00938916e-01 -3.28198850e-01 3.69707942e-01 5.91005206e-01
5.03751516e-01 -1.11094379e+00 -2.49456421e-01 9.47362483e-02
-5.30466378e-01 -1.28645861e+00 -8.08915019e-01 -3.21840107e-01
-7.02885151e-01 -6.05997562e-01 -1.24569201e+00 -9.41207886e-01
1.22778285e+00 2.63590664e-01 1.21343398e+00 -2.69254372e-02
-4.87311900e-01 2.10272193e-01 -2.70364195e-01 -2.66097069e-01
-5.12497187e-01 -7.60847479e-02 -3.75716805e-01 1.04829915e-01
-4.80568528e-01 -4.08839315e-01 -8.16300154e-01 5.07087469e-01
-1.21410084e+00 7.13708937e-01 5.33147991e-01 8.91533077e-01
8.31136286e-01 -2.30106395e-02 6.02178395e-01 -8.74300420e-01
2.79614478e-01 -1.88025415e-01 -8.36170316e-01 3.29557359e-01
-2.54560560e-01 2.00178429e-01 8.56126606e-01 -2.10168630e-01
-1.35713780e+00 2.85489529e-01 -6.73285797e-02 -5.74736357e-01
-2.24808365e-01 1.19078539e-01 -4.81106281e-01 4.65505049e-02
4.32752043e-01 5.02850294e-01 -6.61654100e-02 -1.38247550e-01
9.57139969e-01 3.30936372e-01 5.99299073e-01 -9.71304417e-01
1.02504492e+00 7.30982065e-01 -4.92702760e-02 -5.29494107e-01
-6.83933496e-01 -7.17471242e-02 -5.00954390e-01 -2.15799436e-01
1.02408779e+00 -9.96021152e-01 -3.80003244e-01 2.41543740e-01
-1.28647375e+00 -5.26162505e-01 -3.50721627e-01 3.71261761e-02
-7.29229093e-01 2.81803846e-01 -2.04529360e-01 -7.96606421e-01
-1.79256350e-01 -1.28043354e+00 1.53079534e+00 1.33814260e-01
-1.27694696e-01 -8.36814106e-01 -1.79744229e-01 5.52642524e-01
4.41537827e-01 6.25921190e-01 8.44649255e-01 1.73705816e-01
-1.22810483e+00 1.41074201e-02 -5.60847580e-01 1.76831722e-01
1.62488922e-01 -8.09329897e-02 -8.82758141e-01 -1.86842710e-01
-6.61516786e-01 -3.31624150e-01 7.32537925e-01 3.76129955e-01
1.63082814e+00 -3.75727713e-01 -6.12278819e-01 7.32050538e-01
1.45532787e+00 1.51968777e-01 8.09944987e-01 4.80062440e-02
9.33346868e-01 6.94578648e-01 5.47830760e-01 4.61786419e-01
2.90801376e-01 9.86853004e-01 2.96459019e-01 -5.82684457e-01
-4.42433745e-01 -7.16231287e-01 4.29948606e-02 2.33345166e-01
2.25808308e-01 -6.35513842e-01 -6.64010406e-01 6.96310282e-01
-1.66490448e+00 -9.32467699e-01 -3.20100822e-02 2.09549832e+00
8.63308668e-01 -9.19800028e-02 1.10505112e-01 -2.70333469e-01
8.72150540e-01 1.29450172e-01 -4.00073797e-01 -1.83309838e-01
-1.98157430e-01 2.07715839e-01 3.48447263e-01 4.46100473e-01
-7.93402910e-01 1.25446641e+00 6.56729221e+00 1.12863445e+00
-1.05191028e+00 -2.25112736e-01 1.00409532e+00 7.48467520e-02
-9.74866092e-01 1.28729567e-01 -7.19298482e-01 5.27081013e-01
3.53571653e-01 2.52578128e-02 2.91983902e-01 7.30394840e-01
3.72523695e-01 -1.42928660e-01 -9.53820705e-01 1.04854620e+00
2.02027231e-01 -1.63060927e+00 5.64060152e-01 6.83819950e-02
1.22614026e+00 -5.75516224e-01 2.43493974e-01 -9.78919119e-02
4.14426744e-01 -1.17827535e+00 9.22836483e-01 4.53954548e-01
1.24778521e+00 -7.33806670e-01 6.47970811e-02 1.09855302e-01
-1.16884661e+00 1.95574403e-01 -2.44418740e-01 2.76813567e-01
4.92636681e-01 6.71740770e-01 -8.13597798e-01 5.38032591e-01
3.47459257e-01 4.79881257e-01 -4.58293647e-01 8.65570903e-01
-1.72445595e-01 3.89285833e-01 -1.44582346e-01 6.95474148e-02
2.45585904e-01 -3.32648367e-01 2.29847357e-01 1.06954849e+00
3.90395522e-01 -2.60472558e-02 1.91768318e-01 1.53337145e+00
-1.05110943e-01 5.64197376e-02 -1.00409913e+00 2.91581172e-03
4.16664392e-01 1.10105085e+00 -1.03108692e+00 -3.60502005e-01
-1.20393097e-01 1.29613543e+00 2.21106797e-01 5.17371535e-01
-1.04123938e+00 -3.98653299e-01 4.99740183e-01 3.06371242e-01
2.22169578e-01 -1.42263755e-01 -4.57122922e-01 -1.27365160e+00
9.96981114e-02 -7.32421160e-01 -6.69310242e-02 -1.07722032e+00
-1.09084368e+00 5.00639737e-01 7.41406903e-02 -1.28452575e+00
-3.22782397e-01 -2.88477361e-01 -6.13771558e-01 9.08257842e-01
-1.54506111e+00 -1.52843034e+00 -3.73721510e-01 5.44265985e-01
5.98569870e-01 1.43331140e-02 5.42817414e-01 1.67170405e-01
-4.04338807e-01 7.04459488e-01 8.28459039e-02 1.92565441e-01
6.42485559e-01 -1.05497551e+00 5.74889600e-01 8.17249835e-01
2.15482134e-02 5.64290822e-01 5.24994254e-01 -5.90120316e-01
-1.13581586e+00 -1.38209021e+00 5.04654109e-01 -3.90109599e-01
9.37096998e-02 -7.27555037e-01 -4.78374213e-01 6.18492484e-01
3.33979696e-01 -1.79964170e-01 3.05973709e-01 -4.74404246e-01
-2.65362471e-01 -1.21538870e-01 -1.04344952e+00 1.02658486e+00
1.13353145e+00 -3.62292051e-01 -3.07665560e-02 4.11283761e-01
7.86364257e-01 -3.21777850e-01 -6.12782955e-01 4.15333360e-01
3.61230731e-01 -8.56647193e-01 1.17866278e+00 -7.44669065e-02
7.01591730e-01 -5.84330082e-01 -2.17621755e-02 -1.22974658e+00
-1.15050137e-01 -8.22150052e-01 6.10077918e-01 1.44160569e+00
4.90722865e-01 -4.14509296e-01 1.06924939e+00 5.54233909e-01
-2.95646161e-01 -5.73900819e-01 -4.58944261e-01 -8.10387254e-01
1.42701894e-01 -1.99931026e-01 7.29638934e-01 7.00468838e-01
-2.85771906e-01 2.43055791e-01 -6.15618467e-01 -1.36107191e-01
6.38950050e-01 4.82900739e-01 1.02150953e+00 -8.19767296e-01
-2.39803374e-01 -5.88850200e-01 3.91025506e-02 -1.31856585e+00
2.62502730e-01 -9.02279258e-01 3.82591754e-01 -1.76440227e+00
2.98124045e-01 -7.92865038e-01 3.12794119e-01 3.93890798e-01
-2.11942911e-01 4.94689375e-01 2.09594324e-01 9.10968482e-02
-4.28688049e-01 8.46000671e-01 1.81096041e+00 -2.08586439e-01
1.00219429e-01 -1.55617967e-01 -6.79364502e-01 3.46586585e-01
7.38847792e-01 -2.17763573e-01 -7.25806594e-01 -3.76951635e-01
1.85425356e-01 2.31952578e-01 4.79385465e-01 -8.38325262e-01
-1.97588935e-01 -3.79032016e-01 4.78453726e-01 -7.66067863e-01
6.12104177e-01 -7.23460853e-01 3.81761879e-01 1.80144370e-01
-4.85951126e-01 -1.26920328e-01 9.65391248e-02 6.08076155e-01
-3.45216155e-01 6.24426939e-02 8.97747934e-01 -2.87990600e-01
-4.75338161e-01 4.66831595e-01 -2.11862326e-01 1.34245351e-01
1.05611038e+00 -4.07993972e-01 -1.87298164e-01 -6.02144480e-01
-3.86594534e-01 2.54149586e-01 8.16976726e-01 3.57079327e-01
8.89509380e-01 -1.47475231e+00 -7.96582162e-01 2.26698861e-01
3.45942646e-01 2.91460633e-01 1.52257577e-01 2.89005041e-01
-5.54967284e-01 5.56897044e-01 -1.55374795e-01 -6.63511872e-01
-9.38713908e-01 5.77776134e-01 1.51771128e-01 -1.32930398e-01
-7.29406178e-01 7.39122868e-01 1.00660813e+00 -5.18041551e-01
8.99332762e-02 -2.91655868e-01 2.01931253e-01 -3.99607271e-01
1.76760703e-01 -8.80518705e-02 -3.20037335e-01 -5.74537098e-01
1.29308160e-02 8.24563026e-01 8.46722052e-02 -4.09028083e-01
1.03988254e+00 -1.71671838e-01 5.17200120e-02 -2.11118851e-02
1.01979041e+00 1.10108085e-01 -1.40398657e+00 1.76472738e-02
-3.39514434e-01 -7.70485699e-01 -2.93014795e-01 -8.56575847e-01
-1.12763941e+00 7.72546291e-01 1.82301387e-01 -1.56496793e-01
1.12720144e+00 6.23630472e-02 9.67888594e-01 -8.39398727e-02
4.87055928e-01 -8.32480788e-01 4.66318965e-01 1.20513014e-01
1.05304277e+00 -1.19792962e+00 -2.52734452e-01 -7.54873037e-01
-7.38110900e-01 9.11881030e-01 7.98360229e-01 9.96608734e-02
2.31327578e-01 1.80585429e-01 9.92082339e-03 -5.16870357e-02
-4.84647810e-01 -2.59991229e-01 3.26908231e-01 9.26405668e-01
4.94111061e-01 5.14104813e-02 -9.59444940e-02 1.23112187e-01
-1.86838940e-01 -3.08381766e-02 5.09747028e-01 5.17465234e-01
-1.78410545e-01 -1.29231429e+00 -5.50144732e-01 1.57889411e-01
-2.64530450e-01 -2.15414628e-01 -3.32004279e-01 7.31772602e-01
4.30353731e-02 8.33244920e-01 -4.27600835e-03 -5.91162145e-02
1.77940562e-01 -2.36655504e-01 6.15753770e-01 -7.12895155e-01
-1.77766711e-01 1.86103523e-01 -3.96138728e-02 -4.18329209e-01
-1.20810285e-01 -4.89459515e-01 -1.24214077e+00 3.76269594e-02
-3.78359586e-01 1.51469797e-01 6.80349529e-01 6.15125179e-01
4.48715717e-01 4.65155363e-01 7.63379753e-01 -9.99504685e-01
6.94242213e-03 -5.09769917e-01 -4.43915337e-01 6.24538124e-01
1.26928911e-01 -5.88357925e-01 -2.00434551e-02 3.50926250e-01] | [11.273294448852539, -0.25225284695625305] |
fbba435b-8ad1-452f-856f-921743c9a274 | community-detection-in-networks-a-user-guide | 1608.00163 | null | http://arxiv.org/abs/1608.00163v2 | http://arxiv.org/pdf/1608.00163v2.pdf | Community detection in networks: A user guide | Community detection in networks is one of the most popular topics of modern
network science. Communities, or clusters, are usually groups of vertices
having higher probability of being connected to each other than to members of
other groups, though other patterns are possible. Identifying communities is an
ill-defined problem. There are no universal protocols on the fundamental
ingredients, like the definition of community itself, nor on other crucial
issues, like the validation of algorithms and the comparison of their
performances. This has generated a number of confusions and misconceptions,
which undermine the progress in the field. We offer a guided tour through the
main aspects of the problem. We also point out strengths and weaknesses of
popular methods, and give directions to their use. | ['Darko Hric', 'Santo Fortunato'] | 2016-07-30 | null | null | null | null | ['misconceptions'] | ['miscellaneous'] | [ 2.18737368e-02 -2.79706493e-02 -1.02614544e-01 1.15639457e-04
2.70413846e-01 -8.20123255e-01 7.36565173e-01 5.52090406e-01
-1.25957668e-01 7.28599429e-01 -5.32218926e-02 -4.14114565e-01
-4.03328121e-01 -1.05260980e+00 -6.64747879e-02 -8.69798005e-01
-8.27559173e-01 7.87976980e-01 7.40688145e-01 -2.63111204e-01
4.30462122e-01 3.91015142e-01 -1.22976243e+00 9.99679342e-02
5.11996746e-01 3.17381412e-01 -3.54467034e-02 5.79064488e-01
-3.44425738e-01 4.73514497e-01 -6.94160283e-01 -3.85486454e-01
-1.27325580e-01 -5.37025332e-01 -9.75557923e-01 3.30303997e-01
-2.10953698e-01 4.16988224e-01 -1.89765066e-01 1.29727995e+00
5.56041040e-02 -3.48347843e-01 6.97986722e-01 -1.66130626e+00
-2.19189152e-01 8.58081400e-01 -8.03484619e-01 2.12241456e-01
4.45901573e-01 -2.47036263e-01 9.82404113e-01 -4.19413298e-01
8.45347822e-01 1.18121302e+00 7.51542687e-01 3.39589804e-01
-1.31804597e+00 -5.28810561e-01 6.35004640e-02 3.17823559e-01
-1.73592854e+00 -2.39908114e-01 4.11530465e-01 -7.70482838e-01
3.41424018e-01 5.02669454e-01 7.76086986e-01 7.53877699e-01
-1.44218326e-01 1.96720749e-01 1.11979032e+00 -4.81952518e-01
2.61681378e-01 4.79309142e-01 1.71398893e-01 3.65942240e-01
9.96111393e-01 -3.70182097e-01 2.35612933e-02 -6.24606907e-01
8.97765279e-01 1.58550795e-02 -4.17603940e-01 -8.74925792e-01
-1.29898500e+00 1.11697578e+00 2.87205398e-01 1.01217592e+00
6.96063042e-02 -1.67948961e-01 3.33719343e-01 4.03301179e-01
1.27621546e-01 2.74528652e-01 -2.21925303e-02 1.70030847e-01
-8.89569759e-01 1.19657747e-01 1.09264922e+00 7.76672184e-01
7.55713582e-01 -6.22388363e-01 5.35885870e-01 3.54367375e-01
4.69912440e-01 1.31870182e-02 -1.30046412e-01 -6.83282197e-01
1.10864125e-01 8.96575391e-01 -8.94993320e-02 -1.45241404e+00
-2.72127628e-01 -3.14665794e-01 -1.29609144e+00 2.04921395e-01
6.84938967e-01 -8.62155929e-02 -2.58001298e-01 1.36914039e+00
2.03010008e-01 6.28589769e-04 -1.92791656e-01 6.38862431e-01
6.90861344e-01 4.41847146e-01 -2.64474273e-01 -4.77708966e-01
1.24528193e+00 -6.38327539e-01 -5.57170808e-01 1.77724555e-01
4.11332250e-01 -8.19690883e-01 2.34203100e-01 1.18685544e-01
-7.90589929e-01 -6.10297360e-02 -7.08116174e-01 6.68730736e-01
-4.03569818e-01 -5.83236575e-01 9.60672140e-01 9.00004089e-01
-1.34648216e+00 6.40451193e-01 -5.25736630e-01 -1.03343570e+00
9.30255726e-02 4.58480477e-01 -4.87673670e-01 -8.49956553e-03
-8.39861453e-01 6.56349599e-01 5.38549900e-01 1.42542943e-01
-3.56695861e-01 1.14196599e-01 -3.99185181e-01 -3.36402208e-02
6.70773089e-01 -4.68992293e-01 4.18890834e-01 -9.58272457e-01
-6.28996372e-01 1.04795170e+00 -6.46569133e-02 -3.14966798e-01
4.77284342e-01 7.51701474e-01 -7.59337425e-01 2.01872081e-01
2.29640290e-01 1.97430462e-01 1.75744772e-01 -1.39981210e+00
-5.66399872e-01 -1.35581885e-02 1.52514558e-02 -9.49997008e-02
-1.43730149e-01 3.89178097e-01 -3.46383840e-01 -2.91130036e-01
4.80216563e-01 -9.90685344e-01 -7.16830552e-01 -4.69799228e-02
-6.20516658e-01 -5.23958385e-01 6.62208974e-01 -2.69533694e-02
1.47994149e+00 -1.98648632e+00 3.48853506e-02 8.70475948e-01
9.60880160e-01 1.30345970e-01 1.27734154e-01 1.01357102e+00
-2.22582072e-01 7.45139122e-01 -2.00532153e-01 2.67798394e-01
-2.21845627e-01 1.32430747e-01 6.64446801e-02 8.62178087e-01
3.89602818e-02 3.59349132e-01 -1.23632860e+00 -6.30738735e-01
7.32455850e-02 2.90073097e-01 -1.73507839e-01 -3.20500404e-01
2.53257394e-01 3.42014521e-01 -4.98291105e-01 5.08659184e-01
5.59327841e-01 -8.74057472e-01 8.72704566e-01 3.46648395e-01
-4.27748978e-01 2.76765049e-01 -1.69422543e+00 7.76553690e-01
6.09464526e-01 8.76949847e-01 2.95525879e-01 -1.14077568e+00
7.77826369e-01 5.98946333e-01 6.41669214e-01 3.02910894e-01
6.73725456e-02 3.64150077e-01 6.10703230e-01 -2.99712241e-01
2.10506544e-01 -6.00368008e-02 2.61688203e-01 8.38157952e-01
-2.60365933e-01 2.72389084e-01 7.33463109e-01 6.24765694e-01
1.37381554e+00 -8.06431234e-01 5.61704457e-01 -5.68956792e-01
4.47588056e-01 3.61438654e-02 3.16303164e-01 8.18775058e-01
-3.76097709e-01 4.84107316e-01 9.81229842e-01 -3.26736927e-01
-9.67573643e-01 -8.62122357e-01 -6.82877079e-02 4.89503205e-01
4.22015458e-01 -6.79350853e-01 -5.78126907e-01 -5.67991316e-01
6.72132568e-03 -3.04726899e-01 -7.78018773e-01 5.37247121e-01
-3.19023848e-01 -7.51615345e-01 1.29567146e-01 4.02699411e-02
9.64601152e-03 -9.18201268e-01 -1.08301550e-01 1.57528341e-01
-4.20327783e-01 -9.32655931e-01 -1.91352770e-01 1.13563575e-01
-1.11266220e+00 -1.74687564e+00 -6.04740441e-01 -8.73484612e-01
1.09184921e+00 7.85061002e-01 1.30953789e+00 9.29531634e-01
4.32040617e-02 1.04826011e-01 -4.69853878e-01 -2.75590289e-02
-6.91064179e-01 8.98905843e-02 1.93706676e-01 -6.63428828e-02
4.89699960e-01 -8.22819591e-01 -3.72162908e-01 6.73786402e-01
-5.64347625e-01 -3.54034632e-01 4.86252397e-01 3.48700821e-01
6.12063408e-02 5.28489888e-01 3.51087123e-01 -1.05585301e+00
7.79905140e-01 -7.70068526e-01 -3.50566357e-01 2.45083213e-01
-5.75535417e-01 -2.51836360e-01 6.50245026e-02 -2.65877098e-01
-1.35081992e-01 -2.53207535e-01 2.18050987e-01 7.17530400e-02
-1.97283328e-01 5.75423181e-01 -5.19004762e-02 -2.95068473e-01
6.36063337e-01 -5.13463616e-02 1.49211630e-01 -3.87001395e-01
5.39081991e-02 5.80834389e-01 9.40660387e-02 -3.29558074e-01
9.61792827e-01 8.02663147e-01 -5.39935157e-02 -1.25053287e+00
-1.48092285e-01 -1.01961446e+00 -8.04987609e-01 -4.84337568e-01
3.82287920e-01 -5.61525106e-01 -7.15292871e-01 2.10567057e-01
-1.27274203e+00 9.61439535e-02 7.78495520e-02 2.12543130e-01
-6.90258443e-02 9.05780673e-01 -6.11869991e-01 -9.34027374e-01
9.04284716e-02 -8.99365425e-01 1.50473654e-01 1.23473607e-01
-5.72185695e-01 -1.31741512e+00 2.92444378e-01 1.50891453e-01
2.95442313e-01 4.92692024e-01 7.18807757e-01 -8.07810664e-01
-4.77598667e-01 -2.21473545e-01 -3.10667902e-01 -1.39401957e-01
2.73522943e-01 6.66833580e-01 -5.10640323e-01 -6.21691167e-01
-2.94938684e-01 2.77228504e-01 6.43227637e-01 3.69896024e-01
6.63978457e-01 -3.64844978e-01 -1.07355452e+00 -7.90792108e-02
1.56847906e+00 1.12067737e-01 8.16812038e-01 3.27550113e-01
3.31528246e-01 9.32619750e-01 -1.45646995e-02 2.14945257e-01
6.33442774e-02 3.41696650e-01 5.27990162e-01 -1.98672354e-01
2.34973520e-01 1.24448799e-01 -1.48060350e-02 1.08134186e+00
-6.67573392e-01 -3.71031553e-01 -1.11239421e+00 6.60493016e-01
-1.74245906e+00 -1.45715678e+00 -1.10359442e+00 2.14495039e+00
5.89070678e-01 3.20340008e-01 6.00624382e-01 4.76855576e-01
1.57018626e+00 -7.14280456e-02 1.38195291e-01 7.24658370e-04
-2.79853553e-01 -1.34393707e-01 2.44707927e-01 3.76578778e-01
-1.03556883e+00 5.45234144e-01 7.63163948e+00 6.00919604e-01
-9.38763976e-01 -1.51353270e-01 5.34255207e-01 6.09738231e-01
-2.29715452e-01 4.48144257e-01 -4.61948127e-01 3.01120669e-01
4.86763269e-01 -4.40628469e-01 2.60930091e-01 6.18950963e-01
2.55851954e-01 -3.43858629e-01 -9.93193746e-01 5.34384370e-01
-1.76467434e-01 -1.43366051e+00 -3.32974195e-01 5.64466834e-01
8.59872460e-01 8.62877592e-02 -6.02025032e-01 -1.91823110e-01
5.34683883e-01 -1.01840520e+00 3.84082228e-01 1.08148217e-01
5.10636091e-01 -4.41313654e-01 7.19530880e-01 2.54529595e-01
-1.41083729e+00 1.80846274e-01 -3.69651735e-01 -4.90310580e-01
1.20813988e-01 7.93616414e-01 -9.16841865e-01 4.13954198e-01
7.51949728e-01 6.01098418e-01 -5.68086863e-01 1.63864362e+00
-1.26066744e-01 6.04757786e-01 -4.59689885e-01 -4.14351106e-01
1.86297178e-01 -5.95091939e-01 7.13952661e-01 1.17984521e+00
4.49036472e-02 -6.63424283e-02 2.61879284e-02 7.83052504e-01
1.91839188e-01 1.21341601e-01 -8.06462348e-01 -3.31318647e-01
6.61545813e-01 1.38537049e+00 -1.62373948e+00 -2.72626817e-01
-5.10149658e-01 4.42530870e-01 -1.28046162e-02 1.95442960e-01
-4.39278066e-01 -3.69671136e-01 6.06534362e-01 6.97279572e-01
1.90549552e-01 -4.24014926e-01 -1.04778223e-01 -1.03249502e+00
-2.15183794e-01 -9.77825582e-01 4.80767339e-01 -2.05490872e-01
-1.53500402e+00 4.85413939e-01 -1.04935788e-01 -1.17790949e+00
7.43865520e-02 -3.77180964e-01 -1.01684535e+00 6.59575164e-01
-9.73735511e-01 -5.93997478e-01 -2.01781064e-01 5.26313901e-01
-1.20505773e-01 3.03575937e-02 7.82296658e-01 2.36672923e-01
-3.63634408e-01 -1.39394710e-02 4.02497530e-01 5.91827095e-01
4.61243659e-01 -1.11776769e+00 3.26893747e-01 8.00192893e-01
2.22257841e-02 8.94088566e-01 9.84139800e-01 -7.01486409e-01
-6.52102530e-01 -6.21080875e-01 1.43616962e+00 -4.25023377e-01
1.16740680e+00 -5.88025749e-01 -8.94483387e-01 4.49169576e-01
3.87379766e-01 -3.09691668e-01 8.51275206e-01 5.82009137e-01
-1.80549443e-01 3.67141932e-01 -9.50376868e-01 7.42533267e-01
1.15472555e+00 -2.63442367e-01 -1.94182739e-01 4.05035317e-01
1.67310774e-01 5.57195783e-01 -7.00638413e-01 1.38267189e-01
5.22684515e-01 -1.40248811e+00 7.96273112e-01 -3.95608962e-01
5.09521700e-02 -5.47138572e-01 1.04169391e-01 -9.29302514e-01
-6.89932346e-01 -6.73485100e-01 2.84821898e-01 1.23487711e+00
4.01474267e-01 -7.95411825e-01 1.11404860e+00 -1.13604194e-03
4.05528665e-01 -4.80568409e-01 -7.80440569e-01 -8.28619003e-01
6.11435287e-02 9.39384699e-02 3.53014499e-01 1.68938267e+00
6.26821697e-01 5.65927625e-01 5.96356764e-02 9.10730213e-02
9.79859531e-01 1.03648730e-01 7.49254704e-01 -2.08447313e+00
8.40939507e-02 -8.92416358e-01 -7.07328498e-01 -5.34826636e-01
-4.11022365e-01 -5.93743086e-01 -3.00606489e-01 -1.80927610e+00
5.92838645e-01 -8.17578614e-01 7.79652223e-02 1.34638578e-01
1.61732599e-01 3.50959629e-01 -2.46773660e-02 5.90791285e-01
-8.10691297e-01 -2.84305006e-01 1.06353700e+00 -4.77412567e-02
-1.55364245e-01 1.55141309e-01 -7.30572701e-01 7.81828582e-01
9.07723308e-01 -6.54353499e-01 -1.85033798e-01 2.82993704e-01
5.87627411e-01 -1.45927314e-02 2.01551154e-01 -1.01016116e+00
4.88036692e-01 -3.65026534e-01 -1.61494166e-02 -5.05661726e-01
-2.24875838e-01 -7.53243208e-01 7.69503891e-01 7.53887296e-01
8.41152221e-02 2.86419213e-01 -3.51535052e-01 7.38384366e-01
-1.46510959e-01 -4.59857106e-01 6.30496800e-01 -4.60473865e-01
-3.76818836e-01 1.28639534e-01 -6.61574304e-01 -1.18483752e-01
1.26971769e+00 -6.91093743e-01 -2.50267208e-01 -4.84957159e-01
-7.87832856e-01 3.76430988e-01 9.47970212e-01 2.29698181e-01
2.11160928e-01 -1.08097291e+00 -8.99715126e-01 -1.42167225e-01
1.41851529e-01 -2.33278945e-01 -1.62914410e-01 1.17048478e+00
-7.55978942e-01 2.42311791e-01 -1.74608722e-01 -7.21414208e-01
-1.74210787e+00 6.75655246e-01 4.78689760e-01 -2.95545369e-01
-3.13646406e-01 4.15188074e-01 5.23498245e-02 -2.05271006e-01
8.49292725e-02 1.32731587e-01 -5.68629086e-01 1.70433313e-01
4.68938053e-01 5.14079094e-01 -2.03585267e-01 -6.52323127e-01
-6.80575728e-01 2.92585403e-01 2.36179326e-02 1.53899506e-01
1.35084558e+00 -2.62719840e-01 -9.70068216e-01 5.30422807e-01
6.84290528e-01 2.37781778e-01 -3.56225193e-01 -4.11103033e-02
4.30472016e-01 -3.76029313e-01 -4.42301929e-01 -2.81768501e-01
-9.63695347e-01 6.18773222e-01 8.77260119e-02 1.29193985e+00
5.69496691e-01 4.21607912e-01 -1.28852218e-01 1.24269560e-01
6.48477912e-01 -7.96237230e-01 2.02788711e-02 4.34361845e-01
5.20600915e-01 -1.08176684e+00 1.54276997e-01 -1.08209610e+00
-9.97426957e-02 1.20769620e+00 1.41787544e-01 -2.36837283e-01
8.42721939e-01 4.53609638e-02 -1.32111162e-01 -4.86657262e-01
-6.72564507e-01 -4.12105501e-01 -1.54811725e-01 9.49634433e-01
6.83817267e-01 1.34193286e-01 -7.92194188e-01 1.75603896e-01
2.38640606e-01 -4.01120484e-01 9.71086025e-01 8.08358490e-01
-7.21145213e-01 -1.34729040e+00 -7.69008934e-01 4.87346590e-01
-2.03234866e-01 7.69892111e-02 -8.96105051e-01 9.79274631e-01
3.71062875e-01 1.33370101e+00 1.16428323e-01 -3.89287084e-01
-8.76137614e-02 -2.85589963e-01 3.36823195e-01 -7.00664878e-01
-5.45539260e-01 1.05252862e-01 2.58640140e-01 4.46859226e-02
-7.77761281e-01 -7.81791508e-01 -7.90152907e-01 -1.21469986e+00
-8.74156833e-01 8.10249984e-01 3.15341800e-01 7.06623077e-01
-3.08326315e-02 3.42117473e-02 7.06218779e-01 -5.49453914e-01
7.68711194e-02 -8.13630819e-01 -1.12804842e+00 2.05591276e-01
6.41436428e-02 -4.72281367e-01 -7.19497085e-01 -5.96605316e-02] | [6.952313423156738, 5.298880577087402] |
1edbd283-2fc5-4f2f-8eaf-d1075bcfafe1 | mixed-attention-transformer-for | 2109.02789 | null | https://arxiv.org/abs/2109.02789v2 | https://arxiv.org/pdf/2109.02789v2.pdf | Mixed Attention Transformer for Leveraging Word-Level Knowledge to Neural Cross-Lingual Information Retrieval | Pretrained contextualized representations offer great success for many downstream tasks, including document ranking. The multilingual versions of such pretrained representations provide a possibility of jointly learning many languages with the same model. Although it is expected to gain big with such joint training, in the case of cross lingual information retrieval (CLIR), the models under a multilingual setting are not achieving the same level of performance as those under a monolingual setting. We hypothesize that the performance drop is due to the translation gap between query and documents. In the monolingual retrieval task, because of the same lexical inputs, it is easier for model to identify the query terms that occurred in documents. However, in the multilingual pretrained models that the words in different languages are projected into the same hyperspace, the model tends to translate query terms into related terms, i.e., terms that appear in a similar context, in addition to or sometimes rather than synonyms in the target language. This property is creating difficulties for the model to connect terms that cooccur in both query and document. To address this issue, we propose a novel Mixed Attention Transformer (MAT) that incorporates external word level knowledge, such as a dictionary or translation table. We design a sandwich like architecture to embed MAT into the recent transformer based deep neural models. By encoding the translation knowledge into an attention matrix, the model with MAT is able to focus on the mutually translated words in the input sequence. Experimental results demonstrate the effectiveness of the external knowledge and the significant improvement of MAT embedded neural reranking model on CLIR task. | ['James Allan', 'Razieh Rahimi', 'Sheikh Muhammad Sarwar', 'Hamed Bonab', 'Zhiqi Huang'] | 2021-09-07 | null | null | null | null | ['cross-lingual-information-retrieval'] | ['natural-language-processing'] | [-8.52297097e-02 -2.74008602e-01 -3.29373121e-01 -2.34161347e-01
-8.55371535e-01 -6.40238047e-01 7.47235239e-01 -7.82340392e-02
-5.63263714e-01 5.10145783e-01 5.82203329e-01 -4.38030660e-01
-8.97801667e-02 -7.77183890e-01 -9.53006446e-01 -5.23033619e-01
3.47034514e-01 7.36183882e-01 -1.29549444e-01 -6.01407886e-01
-9.67924893e-02 4.79158126e-02 -1.13024950e+00 5.71058631e-01
9.77620542e-01 5.02701223e-01 7.02021003e-01 -9.33096111e-02
-5.37891626e-01 5.29181778e-01 -5.45348346e-01 -6.28985167e-01
1.85337603e-01 -3.60330343e-01 -8.78373742e-01 -4.87044871e-01
5.80815971e-01 -2.62872964e-01 -4.67455715e-01 1.10586405e+00
3.65088135e-01 8.27407837e-02 6.17452502e-01 -7.05843985e-01
-1.26678061e+00 1.23922205e+00 -4.57863837e-01 1.33733794e-01
2.01257810e-01 -4.85923946e-01 1.48746860e+00 -1.25859666e+00
6.75689340e-01 1.44149125e+00 3.67367268e-01 3.13300163e-01
-1.02653003e+00 -6.47544086e-01 5.14440775e-01 3.73152047e-01
-1.53175437e+00 -2.14987770e-01 5.38082898e-01 -2.07670882e-01
1.07776821e+00 3.36417794e-01 2.96394855e-01 1.22293520e+00
2.06425101e-01 7.61898875e-01 8.29769135e-01 -5.51693797e-01
-5.24623752e-01 4.64915127e-01 2.00689718e-01 3.20547372e-01
4.03916463e-02 -8.54114443e-02 -4.19139326e-01 6.55326173e-02
4.84395087e-01 2.26094455e-01 -4.74701554e-01 -1.58281595e-01
-1.37927854e+00 8.29559803e-01 8.39508712e-01 8.29188704e-01
-2.57501036e-01 -3.55449691e-02 3.49738181e-01 4.63556468e-01
3.75691622e-01 6.47847593e-01 -6.11808836e-01 4.69309449e-01
-6.45193994e-01 -1.74789637e-01 4.44697946e-01 1.01012385e+00
8.40292633e-01 -2.45452791e-01 -3.46111983e-01 1.12127042e+00
4.00640219e-01 7.29051530e-01 9.85344648e-01 -2.03558296e-01
6.84979081e-01 5.92518985e-01 -2.30312809e-01 -9.11375821e-01
-5.80808707e-02 -8.96825254e-01 -8.23690116e-01 -6.10088170e-01
3.00725680e-02 1.72233418e-01 -9.07734334e-01 2.13467216e+00
-1.67993844e-01 -1.24278769e-01 4.23222423e-01 9.92618740e-01
7.47806728e-01 9.23831940e-01 -1.12054259e-01 -1.17372863e-01
1.37867975e+00 -1.23617530e+00 -8.35427523e-01 -4.47645903e-01
7.64902830e-01 -1.04087865e+00 1.37520742e+00 -1.19077906e-01
-7.38718808e-01 -7.07389653e-01 -9.51024234e-01 -3.81807446e-01
-7.25959778e-01 1.14042468e-01 4.72171992e-01 1.15951061e-01
-9.97303903e-01 3.93365800e-01 -5.09112120e-01 -5.89032292e-01
-3.62650782e-01 -2.04462651e-02 -2.59141535e-01 -3.47483128e-01
-1.93677223e+00 1.22820401e+00 3.99045378e-01 2.80543566e-01
-7.85869300e-01 -5.17767787e-01 -7.31987715e-01 3.17245811e-01
2.01927632e-01 -6.16366029e-01 9.05137658e-01 -1.13105714e+00
-1.04338074e+00 8.07576418e-01 -3.39915544e-01 -2.53617167e-01
2.29687586e-01 -5.41777134e-01 -5.41236639e-01 -3.28058213e-01
4.52255428e-01 5.39236009e-01 4.33872133e-01 -1.08565640e+00
-4.41477805e-01 -4.43640292e-01 1.34717315e-01 6.59231007e-01
-5.92829823e-01 1.53533250e-01 -8.29793394e-01 -7.18200803e-01
1.76990535e-02 -9.32822943e-01 1.19017646e-01 -6.28362417e-01
-3.98297995e-01 -3.55419874e-01 4.94382381e-01 -7.35037327e-01
1.31413507e+00 -2.04187465e+00 3.81111503e-01 1.57224491e-01
-2.55133420e-01 2.29473203e-01 -3.97885948e-01 8.03056479e-01
-1.82759896e-01 3.59342128e-01 1.82342172e-01 -8.52155089e-02
-1.70948133e-01 2.21087560e-01 -8.71090472e-01 -1.57296401e-03
1.26518890e-01 1.11177683e+00 -9.78428543e-01 -1.85003698e-01
-3.36133629e-01 5.22506177e-01 -3.55153292e-01 5.09380959e-02
-4.01933677e-02 2.58063585e-01 -4.00668979e-01 3.54577154e-01
3.46213549e-01 -2.23980099e-01 5.08833170e-01 -3.67537588e-01
1.79744717e-02 8.42703581e-01 -7.63687372e-01 1.85757184e+00
-7.79859662e-01 5.49851060e-01 -3.37203473e-01 -8.26745093e-01
6.90054595e-01 5.43855488e-01 1.71975493e-01 -1.15122843e+00
-1.87314510e-01 4.88606602e-01 7.84766600e-02 -1.86420217e-01
6.20857358e-01 -1.03570491e-01 -1.32314593e-01 4.03845102e-01
6.28479868e-02 2.00117692e-01 1.61447704e-01 3.14138323e-01
7.06684768e-01 1.01773277e-01 1.21505260e-02 -3.23536813e-01
5.80240488e-01 -6.33809790e-02 5.04003346e-01 7.54885137e-01
4.16733831e-01 2.77945876e-01 6.76849484e-02 -3.22917163e-01
-8.37633073e-01 -8.97515297e-01 -3.11788708e-01 1.40930581e+00
3.70854288e-01 -5.73499918e-01 -3.48342508e-01 -5.07678092e-01
-1.57946125e-01 6.78526282e-01 -4.16913182e-01 -4.70063508e-01
-7.62545228e-01 -5.65290391e-01 5.56799412e-01 4.13717896e-01
3.57590020e-01 -9.93305624e-01 9.35833529e-02 2.50308216e-01
-5.40314674e-01 -1.02891386e+00 -8.19446981e-01 3.68755102e-01
-7.54031658e-01 -5.84075332e-01 -8.08224559e-01 -1.07236624e+00
4.42890525e-01 5.11322558e-01 1.30855012e+00 2.24459946e-01
2.36721769e-01 -2.90734484e-03 -4.47288305e-01 -4.26694527e-02
-2.39990622e-01 3.11867893e-01 1.75086975e-01 -3.64165902e-02
4.99157995e-01 -2.08617523e-01 -4.38311547e-01 3.25633705e-01
-1.03209460e+00 2.97311251e-03 6.85446739e-01 1.17895758e+00
5.63463151e-01 -1.16060309e-01 4.82753158e-01 -7.75552213e-01
8.05136561e-01 -6.22781873e-01 -3.77363384e-01 7.83738017e-01
-7.49103427e-01 4.39191759e-01 6.21170044e-01 -6.57465160e-01
-8.26623619e-01 -5.16721368e-01 -3.47423367e-02 -4.98266250e-01
3.14743668e-01 1.18414652e+00 -1.41074196e-01 4.66311455e-01
4.81333852e-01 3.51980805e-01 -4.54902232e-01 -8.01955700e-01
5.13886511e-01 5.69352269e-01 9.55851153e-02 -6.97876751e-01
6.73101902e-01 2.70387474e-02 -5.74458420e-01 -2.90795803e-01
-9.61002827e-01 -3.83393526e-01 -3.79773945e-01 1.89119443e-01
9.23101842e-01 -1.14017081e+00 1.20234499e-02 1.19133808e-01
-1.37074363e+00 2.62044761e-02 6.77071838e-03 6.04525685e-01
1.43313035e-01 2.30989173e-01 -6.98125005e-01 -3.49909037e-01
-3.43003511e-01 -1.40878129e+00 1.11191821e+00 -1.91809330e-02
-2.01539621e-02 -1.10373425e+00 2.63906837e-01 2.49085054e-01
6.09762728e-01 -5.66671133e-01 1.52826345e+00 -9.52222466e-01
-4.51926410e-01 3.28554097e-03 -1.63771152e-01 3.37614447e-01
3.77598107e-01 -1.73688337e-01 -8.61932337e-01 -4.77654040e-01
-1.55492172e-01 -2.94032663e-01 1.01158094e+00 -4.94887214e-03
7.47218013e-01 -3.60365659e-01 -4.48771805e-01 4.57899570e-01
1.39131165e+00 2.07230046e-01 5.23680985e-01 4.55015838e-01
8.37080717e-01 6.80379927e-01 2.87124813e-01 -5.18307745e-01
4.82789367e-01 1.02594137e+00 9.06167254e-02 -2.81072378e-01
-2.70382166e-01 -5.26696205e-01 7.08762407e-01 1.45695972e+00
2.73569703e-01 -2.84100860e-01 -1.00835454e+00 6.51916385e-01
-1.73585725e+00 -8.17152381e-01 9.92329493e-02 2.38893008e+00
1.15939021e+00 -1.81572095e-01 -5.80155730e-01 -4.85004187e-01
7.96748698e-01 -4.26040813e-02 -3.90273511e-01 -3.67081225e-01
-3.81316453e-01 1.42357305e-01 2.09732413e-01 7.58598685e-01
-7.81800985e-01 1.36890066e+00 5.44290829e+00 8.92600298e-01
-1.44191730e+00 2.31982484e-01 4.14653867e-01 5.62691279e-02
-7.57431149e-01 1.60801724e-01 -8.69383812e-01 3.15077960e-01
7.97315717e-01 -2.54927754e-01 5.49383283e-01 5.48129499e-01
-2.19944999e-01 4.23100621e-01 -1.31718361e+00 7.50253022e-01
2.31508926e-01 -8.88881505e-01 8.20177436e-01 7.26063550e-02
7.68815637e-01 4.00752515e-01 2.14699909e-01 8.38198602e-01
2.31662765e-01 -1.05554187e+00 6.84944987e-01 3.45670760e-01
7.23922789e-01 -5.34489870e-01 8.48235428e-01 3.44871014e-01
-1.08197296e+00 1.14822038e-01 -6.59626842e-01 2.67291605e-01
-4.34282422e-02 4.12477314e-01 -7.63466358e-01 8.33239436e-01
6.76531851e-01 6.71673357e-01 -5.99318922e-01 5.82704544e-01
-3.09854358e-01 3.15104663e-01 3.68054020e-06 1.61406875e-01
4.46307540e-01 -4.08850968e-01 5.54410934e-01 1.12152326e+00
5.24676979e-01 -5.37501514e-01 2.07716167e-01 7.57260501e-01
-3.66284609e-01 5.85825145e-01 -8.49375188e-01 -2.50287920e-01
3.69350970e-01 1.03931081e+00 -1.36333853e-01 -4.65147674e-01
-6.59624815e-01 1.19544446e+00 6.96777284e-01 7.87539005e-01
-5.77012479e-01 -1.63100481e-01 7.32756615e-01 -1.00143857e-01
4.90806587e-02 1.31294560e-02 1.87541097e-01 -1.51920676e+00
2.66529202e-01 -1.11202884e+00 4.07120556e-01 -7.49150395e-01
-1.42159200e+00 1.11901748e+00 -1.37804881e-01 -1.14323139e+00
-3.33953202e-01 -5.57592511e-01 -2.23380297e-01 1.22172678e+00
-1.86384451e+00 -1.20998836e+00 2.40972400e-01 8.07429314e-01
5.83492577e-01 -2.88609743e-01 9.81304765e-01 6.89375222e-01
-5.08666515e-01 7.80534148e-01 3.89800847e-01 3.43044847e-01
1.19392884e+00 -1.05695963e+00 1.40473634e-01 8.07121813e-01
7.78126121e-01 1.35490024e+00 2.28652045e-01 -6.35956407e-01
-1.38805068e+00 -1.08103073e+00 1.40689552e+00 -4.69542027e-01
8.16291273e-01 -4.17776167e-01 -1.10094798e+00 9.23614621e-01
6.06475115e-01 -4.20710266e-01 6.73124075e-01 5.33114374e-01
-7.34887183e-01 -2.11280793e-01 -3.45971793e-01 8.69921267e-01
9.56888616e-01 -1.01723146e+00 -9.36380208e-01 5.74957013e-01
1.10255039e+00 -1.56876951e-01 -6.49284661e-01 5.37481427e-01
4.15109843e-01 -4.10442859e-01 9.54921603e-01 -9.59540367e-01
5.03207624e-01 -2.30527535e-01 -3.48306865e-01 -1.62696600e+00
-4.88763034e-01 -8.71818885e-02 1.52606085e-01 1.32818139e+00
8.88243973e-01 -7.13073850e-01 -1.51896253e-01 1.44284889e-01
-1.05013832e-01 -6.03268147e-01 -9.22796488e-01 -9.00087237e-01
6.20366573e-01 -1.34597331e-01 7.23902762e-01 1.21303892e+00
-3.37653682e-02 9.77716744e-01 -3.16449553e-01 1.60953939e-01
-7.56600813e-04 3.85723472e-01 2.53696084e-01 -1.04618657e+00
-3.61331016e-01 -5.80089569e-01 9.25300177e-03 -1.31984031e+00
4.70483869e-01 -1.57244706e+00 2.62088384e-02 -1.54663157e+00
5.07347345e-01 -5.69260657e-01 -8.47562194e-01 5.11329532e-01
-6.10397518e-01 7.07869977e-02 1.83942005e-01 5.10325491e-01
-3.95142078e-01 5.95502555e-01 1.10275400e+00 -4.76413727e-01
1.77585743e-02 -4.51098680e-01 -8.06002319e-01 3.53451908e-01
4.82307792e-01 -5.06392121e-01 -4.46986288e-01 -1.40315211e+00
7.17751384e-01 -7.57679790e-02 -1.26315095e-02 -5.50853014e-01
1.51034191e-01 5.32246754e-02 1.85927153e-01 -3.58512700e-01
1.64522380e-01 -1.06365836e+00 7.13959485e-02 3.27038616e-01
-6.48976982e-01 4.54460114e-01 1.12691082e-01 4.14402694e-01
-6.29756570e-01 -1.56849444e-01 2.79108942e-01 -1.96029395e-01
-5.52492797e-01 1.89045459e-01 -2.29999363e-01 2.42365338e-02
4.01886225e-01 2.94566631e-01 -3.91921937e-01 -3.75652194e-01
-4.27112103e-01 3.64501148e-01 2.71818787e-01 1.07052886e+00
3.85981858e-01 -1.60171044e+00 -8.84314060e-01 1.70186773e-01
4.26568598e-01 -3.80471915e-01 -1.42655045e-01 6.47999465e-01
5.02540581e-02 8.98893714e-01 6.25445768e-02 -4.42745209e-01
-8.23948979e-01 6.96523905e-01 5.03770351e-01 -7.64569819e-01
-2.92351365e-01 7.56995738e-01 7.55746841e-01 -8.48647237e-01
2.48851076e-01 -2.53619194e-01 -3.20184678e-01 1.02058552e-01
4.30978239e-01 -3.41838092e-01 3.83134872e-01 -7.61006951e-01
-4.95608062e-01 7.04596996e-01 -5.68580151e-01 -2.12178245e-01
1.05420506e+00 -2.06386670e-01 -4.20614690e-01 5.93186021e-01
1.37365294e+00 2.62830734e-01 -3.04236382e-01 -8.36479723e-01
1.82969376e-01 -2.48656362e-01 6.13399372e-02 -8.43868256e-01
-1.16196620e+00 9.80739176e-01 5.57086706e-01 -2.76538078e-02
8.80813360e-01 1.39747456e-01 7.51164794e-01 6.26523674e-01
4.64678884e-01 -8.87838840e-01 -7.06927702e-02 9.69244599e-01
1.17210412e+00 -1.19140935e+00 -4.69541162e-01 -7.50169205e-03
-6.40422642e-01 9.36605930e-01 7.20499337e-01 1.62202463e-01
4.25465614e-01 -1.20584995e-01 3.56209397e-01 -1.94424495e-01
-8.29314113e-01 -2.40079895e-01 6.95166707e-01 1.65628403e-01
9.26825762e-01 -8.96111224e-03 -5.34916341e-01 4.22177076e-01
-1.65734887e-01 -6.01658165e-01 -3.01514752e-02 5.31136155e-01
-8.41371417e-02 -1.44151151e+00 -1.88780472e-01 1.56551301e-01
-4.92658764e-01 -8.77621233e-01 -5.62629104e-01 4.23651725e-01
-1.00506812e-01 8.23337734e-01 1.11745849e-01 -2.88531870e-01
2.38338411e-01 4.90387350e-01 5.80716208e-02 -7.65914559e-01
-7.37776816e-01 2.33935893e-01 -7.07311481e-02 -4.46620494e-01
-1.49987996e-01 -1.81049123e-01 -9.70391273e-01 1.50949970e-01
-4.52642471e-01 4.05620903e-01 6.15077496e-01 1.01542807e+00
4.35122699e-01 6.38199747e-01 5.80244780e-01 -2.57003605e-01
-5.37768245e-01 -1.24112082e+00 -3.11385423e-01 6.31068885e-01
2.22982064e-01 -5.54309309e-01 -1.77333802e-01 -2.75909334e-01] | [11.369359016418457, 9.839285850524902] |
4f84ee31-4e67-47fe-9d7d-442903ae859e | perceptually-optimized-deep-high-dynamic | 2109.00180 | null | https://arxiv.org/abs/2109.00180v3 | https://arxiv.org/pdf/2109.00180v3.pdf | Perceptually Optimized Deep High-Dynamic-Range Image Tone Mapping | We describe a deep high-dynamic-range (HDR) image tone mapping operator that is computationally efficient and perceptually optimized. We first decompose an HDR image into a normalized Laplacian pyramid, and use two deep neural networks (DNNs) to estimate the Laplacian pyramid of the desired tone-mapped image from the normalized representation. We then end-to-end optimize the entire method over a database of HDR images by minimizing the normalized Laplacian pyramid distance (NLPD), a recently proposed perceptual metric. Qualitative and quantitative experiments demonstrate that our method produces images with better visual quality, and runs the fastest among existing local tone mapping algorithms. | ['Kede Ma', 'Yuming Fang', 'Jiebin Yan', 'Chenyang Le'] | 2021-09-01 | null | null | null | null | ['tone-mapping'] | ['computer-vision'] | [ 3.76692474e-01 -2.27967188e-01 -2.41516560e-01 -3.43056947e-01
-9.96471226e-01 -3.58632386e-01 1.53056577e-01 -5.90416253e-01
-2.52870858e-01 4.26371127e-01 3.44260871e-01 9.93836764e-03
1.13797598e-01 -8.83375227e-01 -7.25386262e-01 -3.41066748e-01
4.41664830e-02 5.01243286e-02 3.39237690e-01 -3.83745939e-01
2.79176742e-01 7.04423070e-01 -1.47579443e+00 2.46071309e-01
8.35667312e-01 1.20865834e+00 3.19231123e-01 9.84919012e-01
4.75070089e-01 1.22076368e+00 -5.13383031e-01 -3.25236797e-01
6.60522103e-01 -5.48789263e-01 -8.76730680e-01 2.16446698e-01
7.81619906e-01 -6.84121132e-01 -8.26910019e-01 1.36266410e+00
8.40870917e-01 2.63094127e-01 3.01539332e-01 -7.17237592e-01
-1.71756709e+00 4.24232543e-01 -1.04660869e+00 3.65188479e-01
2.19517022e-01 1.08369775e-01 9.73934591e-01 -1.12643290e+00
8.18211675e-01 1.34029925e+00 4.86919165e-01 9.48588029e-02
-1.63907003e+00 -5.28774440e-01 -5.70538223e-01 2.59586602e-01
-1.83099222e+00 -3.67772818e-01 9.51224744e-01 -2.53774822e-02
9.43769693e-01 1.87066153e-01 4.98910636e-01 4.78984505e-01
4.05017287e-01 3.73996466e-01 1.42549598e+00 -3.97494435e-01
-5.45850322e-02 -9.88133773e-02 -6.79716051e-01 6.86411679e-01
-3.87523562e-01 4.96798873e-01 -6.77656472e-01 4.57108378e-01
1.44729495e+00 -3.57377440e-01 -3.19516361e-01 -1.95935801e-01
-1.24301720e+00 7.61420012e-01 7.71160066e-01 2.52872825e-01
-4.24889982e-01 3.80147457e-01 1.09741874e-02 5.56856811e-01
4.55830038e-01 3.20256740e-01 1.73058778e-01 -5.18096425e-03
-1.02215290e+00 -1.51559085e-01 2.95638055e-01 6.91502750e-01
1.02619529e+00 2.52763808e-01 -2.21130759e-01 1.37981939e+00
-3.80124971e-02 7.21349657e-01 2.40535274e-01 -1.53068602e+00
-4.61309738e-02 4.44921777e-02 -3.76265543e-03 -1.22934425e+00
-4.21250202e-02 -1.43897310e-01 -1.04910290e+00 6.79558516e-01
-1.52305901e-01 1.82321325e-01 -1.09628880e+00 1.49871886e+00
-1.29370287e-01 -9.09407809e-02 -1.56540856e-01 1.35005021e+00
6.17035091e-01 1.02043486e+00 9.03679356e-02 9.03562829e-02
9.82800901e-01 -9.51136470e-01 -7.55038321e-01 -1.71886340e-01
-2.24261716e-01 -1.06481051e+00 1.61543620e+00 2.31214464e-01
-1.50803828e+00 -1.02549601e+00 -1.41753256e+00 -9.55588877e-01
-2.04340339e-01 1.54373929e-01 1.71132833e-01 4.71507847e-01
-1.68003988e+00 5.59494793e-01 -3.79176170e-01 -5.84213026e-02
3.40653777e-01 1.59302145e-01 -1.63558632e-01 -1.06716610e-01
-1.38782179e+00 9.38046038e-01 3.21852714e-01 -7.93781430e-02
-7.87528515e-01 -9.12178040e-01 -8.56537104e-01 8.52481425e-02
-1.10386208e-01 -4.62999612e-01 1.06898224e+00 -1.07039142e+00
-1.88810837e+00 1.26787651e+00 7.41957575e-02 -3.96800458e-01
3.23116243e-01 8.40782449e-02 -7.80239165e-01 4.29306448e-01
7.66023323e-02 1.21616495e+00 8.82816613e-01 -1.43318605e+00
-6.19248688e-01 1.30597427e-01 8.52900266e-04 4.77510601e-01
-1.32520735e-01 7.26817697e-02 -7.54360259e-01 -1.12524796e+00
-1.89345479e-02 -6.95393264e-01 6.85510784e-02 4.60446745e-01
-4.50033039e-01 3.03100109e-01 8.08240771e-01 -1.03399050e+00
1.27171004e+00 -2.32943916e+00 2.00042389e-02 3.65340114e-01
2.95146793e-01 -1.16258241e-01 -5.37052095e-01 5.00138327e-02
-2.37929478e-01 -1.33439630e-01 -2.86606431e-01 1.85401421e-02
5.11086695e-02 -1.82733640e-01 -5.21759868e-01 4.98718619e-01
-8.12987145e-03 1.02361071e+00 -6.08973384e-01 -5.39180875e-01
5.45798182e-01 7.90144920e-01 -4.83659685e-01 1.39293537e-01
6.58591762e-02 -2.04697661e-02 3.02486598e-01 5.91148794e-01
8.41042757e-01 -1.96592972e-01 -5.79817481e-02 -8.14371884e-01
-2.18707994e-01 -2.40622506e-01 -9.45765138e-01 1.81492710e+00
-6.75902903e-01 1.25139320e+00 -2.16268748e-01 -3.68041366e-01
1.25657177e+00 -2.02332452e-01 4.19058859e-01 -1.50666404e+00
9.88446996e-02 2.27340207e-01 -3.33084524e-01 -1.38944775e-01
8.97042751e-01 -2.25527942e-01 -1.74166746e-02 5.31638801e-01
-1.87539831e-02 -4.34572190e-01 -4.95364470e-03 -1.76931825e-02
6.68318391e-01 -1.34142905e-01 2.75761664e-01 -3.73367578e-01
3.49588424e-01 -1.77659377e-01 -1.23845125e-02 6.84096098e-01
-4.12852079e-01 1.09545815e+00 2.16093227e-01 -5.23943484e-01
-1.60463798e+00 -1.74117756e+00 -1.09184839e-01 1.51913869e+00
6.22981131e-01 2.43450701e-01 -5.91509938e-01 6.01849630e-02
-3.36789161e-01 5.57912946e-01 -7.27930188e-01 -3.61087918e-01
-5.24896383e-01 -2.56181955e-01 5.08242011e-01 4.77652460e-01
1.03335059e+00 -1.29194844e+00 -4.25449520e-01 4.00737412e-02
-4.41743225e-01 -9.43537056e-01 -1.17141032e+00 1.40166640e-01
-2.15065449e-01 -5.21909595e-01 -9.23158705e-01 -1.41456604e+00
3.58217567e-01 3.46212476e-01 1.40967143e+00 -1.38063803e-01
-5.79500437e-01 9.15858895e-02 -5.82179613e-02 3.63026351e-01
-5.78677833e-01 -2.43022114e-01 -1.31760523e-01 -1.71981081e-02
1.86842933e-01 -5.17012954e-01 -9.16890681e-01 4.45488185e-01
-1.15068197e+00 5.03089316e-02 7.72935987e-01 5.75247169e-01
1.26183796e+00 3.73363674e-01 3.23403001e-01 -4.56644684e-01
9.00838077e-01 2.95123637e-01 -6.44195914e-01 4.42673445e-01
-2.34621719e-01 -5.62038086e-02 5.67596138e-01 -5.07537603e-01
-1.07607782e+00 7.31567815e-02 -2.23648980e-01 -6.49294078e-01
1.15293123e-01 8.08783919e-02 6.04413562e-02 -5.95272481e-01
1.05076015e+00 4.35770154e-01 -3.15957814e-01 3.50902565e-02
9.38147306e-01 5.53343058e-01 1.26203144e+00 -2.43960887e-01
1.04882932e+00 5.87091982e-01 -2.67532349e-01 -7.50681460e-01
-8.00303757e-01 4.26950306e-02 -5.82438827e-01 -2.93414444e-01
1.18264043e+00 -1.01749218e+00 -6.49830580e-01 3.80836308e-01
-7.29866385e-01 -8.30120027e-01 -5.16275167e-01 1.53375417e-01
-8.85371566e-01 1.06844053e-01 -9.02112961e-01 -1.81174338e-01
-3.13077331e-01 -1.05271113e+00 1.13448763e+00 2.84715921e-01
9.66730192e-02 -8.25038493e-01 3.51208568e-01 -4.38315719e-02
8.05974185e-01 1.11278869e-01 8.78648520e-01 4.91628379e-01
-6.43220603e-01 1.39590174e-01 -9.60497260e-01 3.83364975e-01
2.49923497e-01 -8.15784782e-02 -9.84420240e-01 -3.23252797e-01
-2.88162410e-01 -5.82145751e-01 8.12335312e-01 7.19813049e-01
1.54811728e+00 -2.11400568e-01 3.36817741e-01 9.88233685e-01
1.79382038e+00 1.85442239e-01 1.16911077e+00 3.51475507e-01
6.84750378e-01 1.21066228e-01 4.77677107e-01 1.46755561e-01
1.87733024e-01 7.37673759e-01 -4.22759876e-02 -9.20125842e-01
-7.38952994e-01 -3.75535339e-01 3.13052684e-01 5.10771215e-01
1.41327962e-01 9.60376561e-02 -4.69660610e-01 3.12472314e-01
-1.14591670e+00 -1.02955818e+00 7.32059121e-01 1.85041797e+00
1.21501589e+00 8.37068707e-02 1.29417375e-01 -1.28332406e-01
9.70436990e-01 4.69895780e-01 -8.32985103e-01 -6.74044192e-01
-4.82098162e-01 4.84778374e-01 6.68615282e-01 8.01018059e-01
-1.14277267e+00 1.19424701e+00 7.73656368e+00 1.04127896e+00
-1.26381469e+00 -7.28798062e-02 1.15122473e+00 -9.04377252e-02
-2.50107765e-01 -5.69818020e-01 -3.79362330e-02 1.02179199e-01
7.17522860e-01 -4.34133261e-01 1.27300262e+00 7.58560359e-01
2.33174890e-01 7.49536753e-02 -7.75539637e-01 1.30321145e+00
1.99077100e-01 -1.35863316e+00 1.42208591e-01 -4.46697213e-02
1.08147609e+00 -1.32480403e-03 8.80095482e-01 2.96667684e-02
5.32317936e-01 -1.08685720e+00 8.90430272e-01 4.84954894e-01
1.66921043e+00 -1.14258170e+00 2.00211674e-01 -5.57715893e-01
-1.50923669e+00 -2.11527944e-02 -8.46478164e-01 6.00570560e-01
4.20678630e-02 4.16810542e-01 -2.93775290e-01 -9.85144675e-02
1.09659362e+00 4.97813642e-01 -7.34021485e-01 6.89265072e-01
2.70453710e-02 -3.55286077e-02 2.64544301e-02 5.73295593e-01
-1.33908531e-02 -2.37665370e-01 1.41553655e-01 1.08790660e+00
1.79990098e-01 2.05541983e-01 -7.94264302e-03 1.26234770e+00
-6.91834450e-01 -1.87584385e-01 -5.05005658e-01 4.80331518e-02
6.47194982e-01 1.34768701e+00 -9.32511330e-01 -2.20882803e-01
-2.42605597e-01 1.63192320e+00 1.93145573e-01 5.23538649e-01
-1.04879582e+00 -9.92813826e-01 6.52405143e-01 -1.37828127e-01
3.63574833e-01 -3.97005081e-02 -3.95953238e-01 -6.30069196e-01
-1.98034152e-01 -1.00166798e+00 2.20746808e-02 -1.44706237e+00
-1.35707390e+00 8.41657221e-01 -4.28124845e-01 -1.41781616e+00
1.23215869e-01 -4.43162173e-01 -2.98811883e-01 8.71197343e-01
-1.56111801e+00 -9.46557045e-01 -4.33721572e-01 9.65524733e-01
4.30864602e-01 9.26363990e-02 4.45385993e-01 5.77573478e-01
-1.10833406e-01 8.18189085e-01 7.03016296e-02 2.59628743e-01
9.42131162e-01 -1.36190820e+00 6.31506205e-01 9.76868272e-01
9.93008446e-03 3.91416878e-01 5.80306768e-01 -3.74823362e-01
-1.11936462e+00 -1.41020179e+00 4.12725419e-01 4.58334908e-02
6.77383304e-01 -3.32716078e-01 -8.74626517e-01 5.25449574e-01
5.53465784e-01 4.09449749e-02 3.75300348e-01 -4.66789395e-01
-5.63652754e-01 -3.56511265e-01 -1.36437464e+00 8.82054806e-01
1.03907382e+00 -1.26144028e+00 -2.65804410e-01 1.17595494e-01
1.12381661e+00 -5.09454012e-01 -1.21135831e+00 1.21625930e-01
5.44348121e-01 -1.15872848e+00 1.44314826e+00 2.54257739e-01
6.32020831e-01 -5.12688875e-01 -3.97732764e-01 -1.42373776e+00
-8.22583914e-01 -4.38119799e-01 2.61473328e-01 7.79334068e-01
4.23835039e-01 -2.92947382e-01 4.45734799e-01 2.18726292e-01
1.10115092e-02 -4.25294846e-01 -7.45629787e-01 -6.03367746e-01
-4.55216579e-02 -5.77474274e-02 4.75003928e-01 8.34986985e-01
-3.99336457e-01 6.92106560e-02 -7.04754531e-01 1.81966901e-01
7.73683965e-01 3.08274657e-01 2.56434232e-01 -6.40704334e-01
-1.92021057e-01 -6.50845349e-01 -3.85739475e-01 -1.05040324e+00
-3.67813036e-02 -6.01661444e-01 4.50203866e-01 -1.46619332e+00
2.51854688e-01 -1.17487013e-02 -3.73401999e-01 3.29742849e-01
8.84839706e-03 1.33278036e+00 1.13763250e-01 1.01707339e-01
-5.80986261e-01 5.90646505e-01 1.63181341e+00 -2.85590440e-01
-3.80981565e-01 -7.48896658e-01 -7.54992604e-01 3.67016971e-01
6.85169160e-01 -7.36346468e-02 -4.04686600e-01 -5.68767011e-01
7.41302371e-02 9.26313996e-02 2.72411883e-01 -1.05432212e+00
1.57556146e-01 -1.75208986e-01 9.38785613e-01 -5.95559120e-01
3.11478496e-01 -6.80976868e-01 6.35719970e-02 2.29669124e-01
-8.16031516e-01 1.91160992e-01 1.26078561e-01 2.48414725e-01
-3.28180820e-01 5.98263025e-01 1.62196016e+00 2.38033995e-01
-1.07927430e+00 5.32131076e-01 -2.06911638e-01 6.24683453e-03
7.97481835e-01 -3.42991233e-01 -5.33562481e-01 -5.94359040e-01
-6.40525997e-01 -3.89410496e-01 7.43923485e-01 5.96744955e-01
9.40761387e-01 -1.85344267e+00 -5.69924712e-01 2.89899319e-01
3.07829562e-03 -6.27475202e-01 5.55504143e-01 4.18051988e-01
-1.13558209e+00 2.48112939e-02 -8.09036016e-01 -3.81363720e-01
-7.73143232e-01 7.56592155e-01 7.26306260e-01 1.38738185e-01
-7.40718663e-01 6.87521994e-01 4.06781465e-01 -2.51103967e-01
2.13655353e-01 -9.63532180e-02 1.08086079e-01 -4.92838413e-01
6.75065160e-01 4.29779410e-01 -1.42030984e-01 -8.64586651e-01
-1.02386340e-01 1.03896153e+00 5.01160026e-02 -4.65388298e-01
1.10117137e+00 -6.57407761e-01 -2.32070684e-01 1.29828498e-01
1.78871679e+00 -6.43914193e-02 -1.45795274e+00 -3.08339745e-01
-5.89841187e-01 -7.84991324e-01 3.56673211e-01 -8.98582280e-01
-1.41154099e+00 6.34585500e-01 1.29220784e+00 2.67420501e-01
1.84566641e+00 -6.40590787e-02 9.40684974e-01 2.63747454e-01
5.41198887e-02 -1.30810499e+00 5.66739023e-01 3.00866812e-01
1.02916789e+00 -1.06634331e+00 -1.02130376e-01 -1.30363882e-01
-8.55474830e-01 9.98526573e-01 5.34049511e-01 -4.84332919e-01
7.34450400e-01 3.14519256e-01 5.98011315e-01 6.27002819e-03
-3.12815458e-01 -4.01963621e-01 4.44101840e-01 8.28800917e-01
1.87489673e-01 -5.84089458e-02 3.60295981e-01 -2.94706196e-01
-4.71697181e-01 9.54850987e-02 4.72299129e-01 4.31652933e-01
-7.84378052e-01 -4.80065048e-01 -2.60062575e-01 1.29305631e-01
-2.88492709e-01 -2.76783288e-01 -3.28671008e-01 5.20495296e-01
-1.92141198e-02 8.02557170e-01 3.30772847e-01 -6.43019676e-01
3.91490340e-01 -6.33875728e-01 5.07982492e-01 -5.48714399e-02
-6.80076778e-02 1.80342540e-01 -4.71038282e-01 -9.30978596e-01
-3.08726132e-01 3.17044646e-01 -1.04433060e+00 -7.57551134e-01
2.09056675e-01 -5.56618690e-01 3.10147762e-01 3.37761879e-01
2.48250574e-01 5.56585491e-01 1.05049825e+00 -1.13687205e+00
-1.63059533e-02 -5.53836524e-01 -9.59148705e-01 4.92069721e-01
5.28935611e-01 -2.50458628e-01 -2.57166058e-01 4.40809071e-01] | [10.980438232421875, -2.199376344680786] |
3d623f04-0932-4414-956f-5936f1a3b183 | improving-generalizability-of-graph-anomaly-1 | 2306.10534 | null | https://arxiv.org/abs/2306.10534v1 | https://arxiv.org/pdf/2306.10534v1.pdf | Improving Generalizability of Graph Anomaly Detection Models via Data Augmentation | Graph anomaly detection (GAD) is a vital task since even a few anomalies can pose huge threats to benign users. Recent semi-supervised GAD methods, which can effectively leverage the available labels as prior knowledge, have achieved superior performances than unsupervised methods. In practice, people usually need to identify anomalies on new (sub)graphs to secure their business, but they may lack labels to train an effective detection model. One natural idea is to directly adopt a trained GAD model to the new (sub)graph for testing. However, we find that existing semi-supervised GAD methods suffer from poor generalization issue, i.e., well-trained models could not perform well on an unseen area (i.e., not accessible in training) of the same graph. It may cause great troubles. In this paper, we base on the phenomenon and propose a general and novel research problem of generalized graph anomaly detection that aims to effectively identify anomalies on both the training-domain graph and unseen testing graph to eliminate potential dangers. Nevertheless, it is a challenging task since only limited labels are available, and the normal background may differ between training and testing data. Accordingly, we propose a data augmentation method named \textit{AugAN} (\uline{Aug}mentation for \uline{A}nomaly and \uline{N}ormal distributions) to enrich training data and boost the generalizability of GAD models. Experiments verify the effectiveness of our method in improving model generalizability. | ['Long-Kai Huang', 'Fu-Lai Chung', 'Huachi Zhou', 'Ninghao Liu', 'Xiao Huang', 'Shuang Zhou'] | 2023-06-18 | null | null | null | null | ['graph-anomaly-detection', 'anomaly-detection'] | ['graphs', 'methodology'] | [ 2.01437756e-01 1.12497933e-01 -1.94702744e-01 -2.00957075e-01
-2.56201237e-01 -6.07878506e-01 4.51969951e-01 4.60455149e-01
1.33862086e-02 5.58336973e-01 -4.02757883e-01 -5.70949733e-01
-1.47159351e-02 -1.07256877e+00 -6.39440656e-01 -6.37265384e-01
-2.03841925e-03 2.66016096e-01 4.68311101e-01 -2.68530399e-01
-2.05646157e-02 4.75274026e-01 -1.12517929e+00 -1.40272379e-01
1.35388052e+00 9.58749175e-01 -3.35658878e-01 2.72886515e-01
-2.66258400e-02 5.36544204e-01 -6.99263453e-01 -5.23756862e-01
3.50799501e-01 -5.87007642e-01 -6.11392498e-01 2.88211286e-01
3.70144010e-01 -2.76565790e-01 -6.29880726e-01 1.58351386e+00
2.48455703e-01 7.37409815e-02 6.51992381e-01 -1.57854235e+00
-7.17532516e-01 2.78851628e-01 -7.91874647e-01 5.59636295e-01
1.66705251e-01 1.03540323e-03 9.53269482e-01 -5.87852299e-01
1.54541060e-01 8.60175788e-01 5.40114343e-01 6.50865436e-01
-9.53450024e-01 -8.24616194e-01 5.68225980e-01 5.98025583e-02
-1.43694806e+00 -1.71564132e-01 1.12856996e+00 -1.78600535e-01
1.72186002e-01 5.12796938e-01 3.88891459e-01 1.26852560e+00
-1.24674596e-01 8.18597317e-01 9.66536164e-01 -2.10468501e-01
1.16779953e-01 2.47383267e-01 3.83027464e-01 8.36468220e-01
7.16090739e-01 -6.19720668e-02 -2.21692532e-01 -3.33072871e-01
4.93689895e-01 2.66359776e-01 -2.18447566e-01 -2.03474820e-01
-6.40830159e-01 7.49761581e-01 5.74170470e-01 4.02081013e-01
-2.59025007e-01 -3.53210360e-01 3.34689200e-01 2.43244216e-01
5.32750309e-01 2.52477586e-01 -1.32300183e-01 1.64224893e-01
-4.73862857e-01 -2.60727018e-01 6.07371092e-01 9.02906299e-01
7.68056571e-01 2.93785006e-01 9.59863663e-02 6.89807773e-01
2.76355445e-01 3.49369794e-01 4.42984492e-01 5.96262775e-02
3.46013665e-01 1.00418615e+00 -3.56962204e-01 -1.55576885e+00
-3.01477879e-01 -7.90600657e-01 -1.07915306e+00 -1.84462503e-01
3.83837342e-01 -1.96433917e-01 -1.06605744e+00 1.49965012e+00
4.55374271e-01 4.85880613e-01 -2.07294270e-01 5.84931493e-01
7.81604469e-01 5.49354553e-01 1.30722225e-01 -2.75421232e-01
9.65186179e-01 -8.33825588e-01 -6.34451687e-01 -5.54415524e-01
9.39921856e-01 -3.04750085e-01 1.17067564e+00 5.85776448e-01
-4.35293257e-01 -4.01414484e-01 -1.13795710e+00 5.47018528e-01
-4.39680904e-01 -1.94071680e-01 6.86218262e-01 9.74650860e-01
-7.20157921e-01 3.47147435e-01 -7.17470467e-01 -3.25531751e-01
5.72412968e-01 3.13655615e-01 -3.33123565e-01 -4.27661389e-01
-1.23811579e+00 3.73437881e-01 5.80544889e-01 2.31208950e-01
-1.02119470e+00 5.51398669e-04 -1.03435993e+00 -2.66562670e-01
8.83973956e-01 1.19453482e-01 7.01463819e-01 -9.49742794e-01
-7.18300760e-01 7.07874596e-01 1.51351944e-01 -4.00376469e-01
1.96590140e-01 -5.94262592e-02 -1.09584987e+00 1.76407043e-02
1.83409467e-01 -1.68983694e-02 1.08858514e+00 -1.24865437e+00
-5.85928798e-01 -5.95786631e-01 5.42640053e-02 8.41455162e-02
-1.00236154e+00 -2.09005430e-01 -5.81777215e-01 -7.95546532e-01
3.91359627e-01 -7.90229678e-01 -1.85272515e-01 -5.24038196e-01
-7.99992740e-01 -2.57544935e-01 1.28056669e+00 -8.07161808e-01
1.72149050e+00 -2.32363534e+00 -3.68084401e-01 7.85466969e-01
5.66420972e-01 6.93399251e-01 3.41082662e-02 1.07608818e-01
-2.24699304e-01 3.24176908e-01 -3.66421372e-01 1.09403409e-01
-3.86224151e-01 4.91603047e-01 -2.92188436e-01 5.31909049e-01
2.14454293e-01 6.64195955e-01 -9.32953894e-01 -3.82547617e-01
-1.33408576e-01 -5.34063019e-02 -3.75047088e-01 2.06156209e-01
-1.28281847e-01 5.57008445e-01 -8.52815628e-01 1.00553989e+00
7.43619263e-01 -3.05206448e-01 3.31910588e-02 1.69918016e-01
4.22620326e-01 4.64928858e-02 -1.25654888e+00 1.01898038e+00
2.88514644e-02 1.51630700e-01 -1.81750044e-01 -1.35813582e+00
1.19060171e+00 3.34337689e-02 3.31931025e-01 -6.32272840e-01
3.31902807e-03 2.77376235e-01 1.92127019e-01 -3.33632201e-01
2.04593539e-02 1.52321413e-01 -2.38765441e-02 4.14417028e-01
-1.76756084e-01 4.11229700e-01 1.21339776e-01 3.75045985e-01
1.31684208e+00 -3.58495295e-01 2.87893474e-01 -4.88810763e-02
7.86974669e-01 -1.87234282e-01 7.41890073e-01 7.77020574e-01
-3.82453173e-01 3.61273557e-01 8.03804576e-01 -3.30261379e-01
-5.42031109e-01 -9.83771086e-01 -9.64173675e-02 1.05863726e+00
3.57619047e-01 -6.24638498e-01 -6.43179715e-01 -1.53652084e+00
-1.16443060e-01 6.26806855e-01 -5.41576087e-01 -7.68294752e-01
-3.83415312e-01 -1.03156006e+00 6.61748946e-01 4.85975355e-01
7.24002242e-01 -9.79484737e-01 2.19353929e-01 -4.73295674e-02
-6.18233420e-02 -1.08816624e+00 -4.02355105e-01 -1.36201203e-01
-7.19823539e-01 -1.23482990e+00 -2.02531278e-01 -7.30600178e-01
1.25590920e+00 3.80766422e-01 9.22674656e-01 7.01996744e-01
-5.78402989e-02 3.08852971e-01 -4.92964178e-01 -6.23737156e-01
-4.73537534e-01 8.17459673e-02 3.69452864e-01 5.19090950e-01
5.63276291e-01 -5.64111590e-01 -4.55308020e-01 5.95998645e-01
-1.10830975e+00 -5.35860181e-01 5.68773448e-01 6.47924721e-01
5.68792939e-01 6.98522747e-01 8.36364746e-01 -1.36369061e+00
7.14209259e-01 -6.27431035e-01 -5.01428902e-01 1.24384478e-01
-8.28060269e-01 -1.81339622e-01 7.61221230e-01 -4.46681589e-01
-6.86683416e-01 -1.63793728e-01 -2.00962707e-01 -4.42677110e-01
-2.56887496e-01 6.53752506e-01 -6.67650700e-01 -2.81575590e-01
7.79373348e-01 2.64986783e-01 -9.24594775e-02 -3.60482603e-01
-1.00217529e-01 5.31015098e-01 4.59340841e-01 -4.77448106e-01
1.34262240e+00 3.19938600e-01 7.94791505e-02 -8.39549839e-01
-9.36325788e-01 -5.82507312e-01 -4.62419063e-01 -1.57977909e-01
4.91585493e-01 -6.42497361e-01 -2.33704060e-01 6.38747036e-01
-5.62319994e-01 -3.74913216e-03 -7.08276778e-02 2.22019851e-01
2.69922972e-01 8.48136008e-01 -2.40419507e-01 -8.42066586e-01
-4.47229408e-02 -7.39481568e-01 6.91229641e-01 3.65688384e-01
1.10831961e-01 -1.16358161e+00 -1.69455498e-01 1.95284367e-01
1.03362858e-01 4.17693108e-01 8.38101029e-01 -1.39972579e+00
-3.50420684e-01 -7.42271364e-01 -2.64455140e-01 6.94900095e-01
4.73075479e-01 -1.50567159e-01 -9.38094676e-01 -5.34115911e-01
2.26875052e-01 -1.74898669e-01 6.63100421e-01 -1.94305092e-01
1.54542315e+00 -5.31222701e-01 -3.96855444e-01 2.97489196e-01
9.05250430e-01 2.15169653e-01 6.16690695e-01 1.48254171e-01
1.14837372e+00 3.05063188e-01 6.98262334e-01 2.35945672e-01
9.96598601e-02 3.05281729e-01 6.56377017e-01 -1.96206957e-01
2.60115415e-01 -4.37693357e-01 4.54794019e-01 7.13401496e-01
2.47018831e-03 -3.96212727e-01 -1.04673028e+00 3.96581799e-01
-1.66802251e+00 -5.02482176e-01 -2.94967771e-01 2.47813559e+00
6.10132158e-01 6.67369068e-01 2.83083767e-01 4.25193548e-01
8.86868119e-01 1.06449150e-01 -6.09699070e-01 -1.47511661e-02
-5.03645688e-02 1.65287092e-01 2.37701818e-01 1.44405663e-01
-1.32760990e+00 8.03301871e-01 4.89141989e+00 1.02009118e+00
-9.54926550e-01 1.38413748e-02 8.11452210e-01 4.75862116e-01
-2.09985361e-01 9.14025903e-02 -8.33896458e-01 5.53237140e-01
7.73572505e-01 -1.08857378e-01 1.09679334e-01 1.00252175e+00
-2.14678958e-01 1.67089045e-01 -9.18647945e-01 8.22154701e-01
4.60697472e-01 -6.44171059e-01 1.86466143e-01 2.99458265e-01
6.33876622e-01 -3.29092383e-01 2.05006331e-01 6.06829464e-01
6.44615516e-02 -9.45246220e-01 -1.32332712e-01 2.03962639e-01
7.07762897e-01 -8.14244092e-01 8.30864727e-01 7.65918016e-01
-1.12860906e+00 -7.23913684e-02 -3.01662773e-01 2.31998600e-02
-3.19996774e-01 6.37029409e-01 -7.24354267e-01 7.58938611e-01
5.65433681e-01 7.96472073e-01 -1.12695634e+00 9.76498842e-01
-5.27466059e-01 9.53721821e-01 -2.83945799e-01 1.71993911e-01
2.43075803e-01 -3.00632596e-01 7.91334867e-01 6.91559017e-01
2.19319150e-01 1.59332454e-01 6.15214586e-01 3.90380293e-01
-2.15492353e-01 2.89046615e-01 -9.43551302e-01 -2.67085791e-01
1.56783968e-01 1.25219703e+00 -8.60557079e-01 -9.78278890e-02
-5.92825770e-01 8.75235200e-01 3.00156683e-01 3.94787222e-01
-8.30522537e-01 -4.27415788e-01 1.08792253e-01 4.02978212e-01
-1.43119961e-01 -8.77595916e-02 -2.77310666e-02 -1.31323206e+00
2.39863142e-01 -1.03405046e+00 1.00631464e+00 -2.94250965e-01
-1.57831895e+00 6.05345666e-01 -1.31824851e-01 -1.42603910e+00
-1.86872892e-02 -6.27459764e-01 -8.98579419e-01 5.20328701e-01
-1.13096583e+00 -1.06948030e+00 -5.03994346e-01 8.98274958e-01
1.36984900e-01 -4.12958235e-01 6.29788220e-01 4.64202166e-01
-1.06778646e+00 9.95914876e-01 -2.43956119e-01 5.35697460e-01
6.04957283e-01 -1.20678413e+00 2.08113670e-01 1.51364481e+00
4.02265310e-01 5.81029713e-01 4.02248561e-01 -1.04044318e+00
-1.10974407e+00 -1.39724565e+00 3.69658023e-01 -5.82010269e-01
6.88782215e-01 -4.26344931e-01 -1.31933248e+00 8.59713495e-01
-3.97738606e-01 3.86414766e-01 6.17709935e-01 2.95118093e-01
-2.43187502e-01 -1.57087967e-01 -1.20203447e+00 7.18705714e-01
1.15129280e+00 -4.58328456e-01 -2.81915456e-01 5.76093197e-01
6.11310840e-01 -2.56729990e-01 -4.41481292e-01 6.91058815e-01
-3.10108989e-01 -8.00908267e-01 7.03818798e-01 -8.23252022e-01
-1.01511553e-01 -3.39009047e-01 1.47782862e-01 -1.07470989e+00
-1.00587517e-01 -7.48831272e-01 -5.87035656e-01 1.41744590e+00
3.31992149e-01 -1.02230239e+00 9.52997923e-01 5.57658017e-01
-1.98631331e-01 -5.86225331e-01 -7.13733792e-01 -8.85618508e-01
-2.76275486e-01 -5.02029955e-01 5.39491832e-01 1.31547117e+00
-1.26454711e-01 4.24697936e-01 -3.90714139e-01 6.48042381e-01
6.06760442e-01 -2.70369142e-01 8.73971164e-01 -1.52494395e+00
-1.34892911e-01 -1.63183913e-01 -7.34203041e-01 -8.48638058e-01
9.77697521e-02 -9.86176670e-01 -3.02910209e-01 -1.12977159e+00
-2.26812046e-02 -6.92252338e-01 -6.87701404e-01 8.18228900e-01
-5.27402163e-01 3.55159998e-01 -3.78764451e-01 2.28860959e-01
-6.77963436e-01 4.91033167e-01 1.13348556e+00 -8.80737454e-02
-3.62546712e-01 5.04274368e-01 -9.91352975e-01 9.64072168e-01
8.22587967e-01 -4.19152319e-01 -6.54024124e-01 5.57522289e-02
2.33394444e-01 -4.76588905e-01 4.22598600e-01 -1.10520935e+00
1.26711622e-01 -3.74655263e-03 3.51657927e-01 -3.01710218e-01
-1.95600480e-01 -7.84526229e-01 -3.33607972e-01 3.57487708e-01
1.60607770e-01 5.84885962e-02 1.83530152e-01 1.05615020e+00
-2.48128086e-01 -3.27734321e-01 5.46363592e-01 9.55321267e-02
-8.86521399e-01 9.07864809e-01 -8.63760412e-02 1.63792968e-01
1.20344079e+00 -1.60488576e-01 -3.70094359e-01 -4.97256428e-01
-8.11612844e-01 4.66179281e-01 1.91983819e-01 4.48792219e-01
6.91762030e-01 -1.47404444e+00 -5.69571495e-01 5.94817817e-01
3.97222579e-01 1.98972255e-01 2.49699563e-01 6.78067267e-01
-2.94114321e-01 -9.93046090e-02 -3.53347361e-02 -5.25891304e-01
-1.07994914e+00 8.97346973e-01 2.22094908e-01 -3.14546078e-01
-5.55245578e-01 7.82583117e-01 4.10826802e-01 -4.17056918e-01
2.25827023e-01 2.62191534e-01 -3.62656504e-01 -1.52638465e-01
3.66280556e-01 1.94993630e-01 1.62361920e-01 -7.08836973e-01
-3.73848587e-01 5.41720651e-02 -5.58279932e-01 4.96061713e-01
8.35648060e-01 4.17943820e-02 -2.34865516e-01 1.83858320e-01
7.98546970e-01 3.47474843e-01 -8.39100480e-01 -5.31203210e-01
1.22205079e-01 -5.92494309e-01 -1.56537622e-01 -4.53936666e-01
-1.18156207e+00 7.93039322e-01 5.17391622e-01 7.63463199e-01
1.27051663e+00 -2.45239004e-03 7.80953884e-01 4.79008734e-01
2.58733958e-01 -9.38631833e-01 4.64574218e-01 2.94259548e-01
5.15111446e-01 -1.37715757e+00 3.12160607e-02 -8.11515033e-01
-6.59653246e-01 7.83258617e-01 1.18914187e+00 -2.17051934e-02
7.97515869e-01 -3.50368738e-01 -9.93974954e-02 -4.75333273e-01
-6.32578433e-02 -2.03921497e-01 5.56406140e-01 6.84791267e-01
2.45879684e-02 -4.83986251e-02 -1.28217399e-01 7.44293153e-01
4.84287813e-02 -7.00354397e-01 4.87848043e-01 9.55810726e-01
-3.52349669e-01 -1.14228404e+00 -2.90961653e-01 8.79862964e-01
-5.05784512e-01 3.70926298e-02 -6.46343231e-01 7.86050498e-01
1.84578374e-01 1.01504230e+00 -2.60996222e-01 -6.37887359e-01
4.60286707e-01 4.61015888e-02 -7.06413388e-02 -9.23532367e-01
-1.34445086e-01 -2.05448177e-02 4.86046597e-02 -2.75283664e-01
-7.09492539e-04 -4.83622164e-01 -1.17013514e+00 -2.28726491e-01
-5.42533696e-01 3.43922675e-01 -3.98104824e-02 1.10799718e+00
1.64016619e-01 4.48966891e-01 9.33732092e-01 3.11258342e-02
-4.69071835e-01 -8.14903557e-01 -8.84285867e-01 5.88023782e-01
1.70907184e-01 -6.94228053e-01 -5.33292353e-01 -3.33929658e-01] | [6.6421403884887695, 5.773899555206299] |
3fdf0276-aa46-44f4-8e5e-dda23434fe8f | learning-dialogue-representations-from | 2205.13568 | null | https://arxiv.org/abs/2205.13568v2 | https://arxiv.org/pdf/2205.13568v2.pdf | Learning Dialogue Representations from Consecutive Utterances | Learning high-quality dialogue representations is essential for solving a variety of dialogue-oriented tasks, especially considering that dialogue systems often suffer from data scarcity. In this paper, we introduce Dialogue Sentence Embedding (DSE), a self-supervised contrastive learning method that learns effective dialogue representations suitable for a wide range of dialogue tasks. DSE learns from dialogues by taking consecutive utterances of the same dialogue as positive pairs for contrastive learning. Despite its simplicity, DSE achieves significantly better representation capability than other dialogue representation and universal sentence representation models. We evaluate DSE on five downstream dialogue tasks that examine dialogue representation at different semantic granularities. Experiments in few-shot and zero-shot settings show that DSE outperforms baselines by a large margin. For example, it achieves 13% average performance improvement over the strongest unsupervised baseline in 1-shot intent classification on 6 datasets. We also provide analyses on the benefits and limitations of our model. | ['Bing Xiang', 'Andrew O. Arnold', 'Xiaofei Ma', 'Nicholas Dingwall', 'Wei Xiao', 'Dejiao Zhang', 'Zhihan Zhou'] | 2022-05-26 | null | https://aclanthology.org/2022.naacl-main.55 | https://aclanthology.org/2022.naacl-main.55.pdf | naacl-2022-7 | ['dialogue-understanding', 'conversational-response-selection', 'dialogue-act-classification', 'goal-oriented-dialogue-systems', 'dialogue-management'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 2.83753276e-01 4.69964981e-01 -3.98005694e-01 -5.82283437e-01
-8.47436011e-01 -2.56161720e-01 1.28016889e+00 3.53650331e-01
-5.29738843e-01 7.91220427e-01 1.12892497e+00 -1.67413116e-01
1.40923038e-01 -6.76180363e-01 2.32075527e-01 -3.64266962e-01
3.43159260e-03 5.93924403e-01 1.54597268e-01 -1.05346787e+00
4.97537434e-01 -1.01168472e-02 -1.08763540e+00 5.11757255e-01
7.64218271e-01 6.16799831e-01 -8.37503746e-02 9.76430237e-01
-4.88988936e-01 1.14382327e+00 -1.08181679e+00 -4.79409754e-01
-2.32425228e-01 -7.73676157e-01 -1.32890689e+00 1.44529864e-02
2.14762747e-01 -7.34758496e-01 -6.49042010e-01 5.37451923e-01
7.58409441e-01 5.47757983e-01 9.28914845e-01 -1.10383439e+00
-7.07173645e-01 4.68117744e-01 -2.65013486e-01 4.66136307e-01
7.65293002e-01 1.77913681e-01 1.53140712e+00 -1.05322063e+00
5.66166699e-01 1.58509886e+00 5.90523839e-01 1.09180450e+00
-1.43881714e+00 -1.41338035e-01 -1.66126326e-01 -8.31310898e-02
-4.04774874e-01 -6.97141290e-01 6.77131832e-01 -2.33561724e-01
1.46450388e+00 4.34976578e-01 1.84174106e-01 1.37172544e+00
4.67006899e-02 1.06494725e+00 9.17659521e-01 -5.64616859e-01
1.87644154e-01 1.83932751e-01 7.25593388e-01 6.01108670e-01
-3.46659571e-01 -4.37816471e-01 -7.75367141e-01 -5.34915566e-01
3.26863319e-01 -6.26606941e-02 -2.80827224e-01 -1.49684966e-01
-9.46944296e-01 1.55882597e+00 2.28235513e-01 3.80971640e-01
-1.08879909e-01 -1.51348978e-01 9.13484454e-01 7.67099023e-01
8.27146828e-01 9.46640074e-01 -2.60139883e-01 -5.67406297e-01
-1.74936652e-01 3.24550837e-01 1.22898340e+00 5.65328419e-01
4.11370605e-01 7.51571730e-02 -7.75793254e-01 1.50619686e+00
5.02248853e-02 -8.51882771e-02 7.74369359e-01 -8.74141216e-01
6.25905037e-01 5.58044434e-01 -5.20944558e-02 -7.64280796e-01
-4.42920387e-01 3.05166841e-01 -6.93384409e-01 -1.53930023e-01
3.16994905e-01 -4.14888501e-01 -2.44692564e-01 1.56254196e+00
2.79473048e-02 -3.99399608e-01 6.04186714e-01 5.41284919e-01
1.36968768e+00 9.98584569e-01 2.31602058e-01 -2.30025977e-01
1.35244799e+00 -1.35507309e+00 -9.84727263e-01 -3.52215797e-01
1.03036189e+00 -4.37783778e-01 1.45214331e+00 -1.30113989e-01
-8.59403431e-01 -3.27032387e-01 -8.86179686e-01 -4.01132256e-01
-1.94872379e-01 -3.05848747e-01 7.69756794e-01 6.29560530e-01
-9.16347921e-01 5.36429763e-01 -3.03265154e-01 -6.55780911e-01
2.23822817e-01 -7.07951784e-02 -3.05106401e-01 1.48263276e-01
-1.51537895e+00 1.26839411e+00 8.42557997e-02 -4.07466829e-01
-5.98362982e-01 -6.55647337e-01 -1.26459026e+00 2.22629026e-01
2.22914353e-01 -2.74324417e-01 1.69597387e+00 -3.76269400e-01
-1.94235444e+00 8.57786298e-01 -6.32265955e-02 -7.59777486e-01
2.09714368e-01 -2.97052205e-01 -1.49563596e-01 1.69383243e-01
4.54553366e-02 5.49597263e-01 4.33222294e-01 -7.20019221e-01
-3.85738075e-01 -1.29212752e-01 5.34013152e-01 7.24885643e-01
-7.54010499e-01 -2.41783895e-02 4.20237035e-02 -4.16449696e-01
-4.42637116e-01 -5.75882018e-01 -2.84583837e-01 -5.58832772e-02
-1.83976233e-01 -9.23787117e-01 8.53205919e-01 -5.85628152e-01
1.14111817e+00 -2.01448989e+00 2.91573942e-01 -6.84710801e-01
3.69514108e-01 5.29176354e-01 -2.30836496e-01 9.73325372e-01
3.35391134e-01 -5.95090948e-02 -3.37367803e-01 -4.66914088e-01
2.42067248e-01 2.51188785e-01 -3.58181298e-01 1.10678941e-01
5.27703166e-01 9.86792028e-01 -1.23377728e+00 -2.74518520e-01
3.72170895e-01 1.82125583e-01 -5.04137635e-01 6.89794302e-01
-2.86133468e-01 2.49136016e-01 -2.68074483e-01 6.50889874e-02
2.61658318e-02 -1.97708890e-01 3.65174860e-01 1.20788522e-01
3.31272840e-01 9.97918963e-01 -3.77453238e-01 1.84065175e+00
-7.74586856e-01 1.09053802e+00 -1.66445702e-01 -9.96136844e-01
1.08108211e+00 5.35923898e-01 7.77515769e-02 -8.30452085e-01
4.64517594e-04 -1.37722909e-01 3.29483524e-02 -4.85281974e-01
1.10630274e+00 -4.45933759e-01 -5.54297566e-01 8.00869823e-01
4.78592128e-01 -2.94296473e-01 1.48799941e-01 6.76872253e-01
1.24993491e+00 -4.69001889e-01 7.47732997e-01 -1.77614316e-01
4.37214047e-01 -1.62586004e-01 1.60581499e-01 7.28469610e-01
-4.98617858e-01 3.26125801e-01 9.51529503e-01 -2.31841758e-01
-8.08639467e-01 -8.64019156e-01 -9.55019519e-02 1.64854550e+00
-4.10103574e-02 -6.09962881e-01 -3.60231549e-01 -1.01376724e+00
4.31831293e-02 9.93824780e-01 -6.14565670e-01 -4.23346490e-01
-3.45158726e-01 -6.56263590e-01 6.14078820e-01 5.23294687e-01
5.36824167e-01 -1.22953558e+00 -4.01851743e-01 1.78037375e-01
-1.97238788e-01 -8.39882195e-01 -3.91433686e-01 1.41534105e-01
-6.55240715e-01 -8.86075616e-01 -8.28558147e-01 -7.56734312e-01
1.74985901e-01 5.26798904e-01 1.31821859e+00 -5.77302091e-02
-2.82416940e-01 6.44762397e-01 -5.92299402e-01 -2.63751559e-02
-8.91716361e-01 1.22254387e-01 -9.95576233e-02 -4.78822827e-01
7.13678241e-01 -1.10687859e-01 -2.53671587e-01 -7.07988366e-02
-5.37979305e-01 -1.11875087e-01 3.60150300e-02 1.59194720e+00
-4.51790869e-01 -5.62663972e-01 1.28855383e+00 -1.32512987e+00
1.76183248e+00 -6.84950888e-01 1.11453988e-01 2.22087234e-01
-6.13835514e-01 1.20908245e-01 4.51394945e-01 -7.76265338e-02
-1.37190819e+00 -7.64214396e-01 -3.74806523e-01 1.81179225e-01
-8.70236829e-02 4.21140492e-01 1.67253241e-01 4.89736378e-01
9.45957899e-01 1.28029719e-01 4.75558072e-01 -3.24564457e-01
7.26707518e-01 1.06304741e+00 1.43884748e-01 -3.95911962e-01
5.14485613e-02 -4.71606217e-02 -6.00607812e-01 -1.34752882e+00
-7.52277195e-01 -7.15195596e-01 -4.10643846e-01 -1.01553097e-01
8.17404747e-01 -6.68436229e-01 -5.76473653e-01 2.05355868e-01
-1.27045834e+00 -4.53200400e-01 -4.96088892e-01 2.30872348e-01
-5.18066704e-01 6.16497159e-01 -1.14456546e+00 -9.82802927e-01
-6.10852182e-01 -7.45994508e-01 8.99931312e-01 1.85425147e-01
-7.70192325e-01 -1.37749100e+00 4.55007881e-01 5.17221272e-01
4.70293313e-01 -1.41333461e-01 9.22274530e-01 -1.37966907e+00
3.12187910e-01 -3.13642174e-02 -1.82510898e-01 5.20003200e-01
3.99084032e-01 -4.90954041e-01 -1.23425782e+00 -3.53066832e-01
1.13493197e-01 -1.14559734e+00 9.95907485e-01 -1.32484302e-01
5.41553020e-01 -5.22601545e-01 1.38024926e-01 -1.91977620e-01
8.41425657e-01 1.35756269e-01 4.73560363e-01 1.26897901e-01
3.19211781e-01 1.05346870e+00 7.35068321e-01 7.18509257e-01
4.20523018e-01 5.93374968e-01 4.20905128e-02 -9.71851684e-03
-7.59453923e-02 -9.11014900e-02 4.68533903e-01 1.00052381e+00
4.24468309e-01 -3.98223877e-01 -7.62083411e-01 5.50067246e-01
-1.86118603e+00 -1.04525173e+00 1.75271511e-01 1.98336017e+00
1.19452274e+00 7.11636245e-02 4.16399837e-01 -1.56210646e-01
5.22889137e-01 8.28203738e-01 -4.35340673e-01 -9.51205611e-01
8.72107893e-02 2.56074727e-01 -2.71421611e-01 6.44461632e-01
-1.11463165e+00 9.85214770e-01 6.39239359e+00 5.92285991e-01
-6.32260859e-01 2.70917624e-01 5.66338122e-01 1.48488328e-01
-1.96171895e-01 -3.00595403e-01 -6.04054928e-01 3.22435439e-01
1.18112338e+00 -4.38822955e-01 2.64041908e-02 8.22775722e-01
-2.72753835e-01 2.61379369e-02 -1.24819052e+00 7.44906545e-01
4.40319717e-01 -1.46747971e+00 -3.77418809e-02 -2.17332602e-01
5.98945141e-01 -1.42074004e-01 -1.65898934e-01 9.78154600e-01
5.18185675e-01 -1.05053902e+00 -3.45911652e-01 9.99109596e-02
6.28092170e-01 -6.11391008e-01 8.73005927e-01 3.01354796e-01
-6.53930962e-01 3.78525779e-02 -4.59777027e-01 -3.74184132e-01
1.87554091e-01 -3.73334363e-02 -1.32964122e+00 3.16122472e-01
7.05672950e-02 8.18329155e-01 -3.12540174e-01 3.68588716e-01
-2.14833096e-01 5.42014539e-01 2.96742648e-01 -5.03227532e-01
4.74709839e-01 -1.29262939e-01 3.97505432e-01 1.57864666e+00
-2.92403281e-01 2.95280486e-01 1.72721684e-01 5.41431248e-01
-3.80967408e-01 2.53644556e-01 -1.06269383e+00 -4.51789618e-01
5.77534676e-01 1.18131530e+00 -1.92663297e-01 -3.86388153e-01
-6.21217012e-01 1.17569792e+00 6.19707704e-01 2.15922017e-02
-3.13482016e-01 -8.15627337e-01 8.79646242e-01 -5.11023879e-01
-3.94863188e-02 -4.42884490e-02 1.39998659e-01 -1.13261080e+00
-3.10020000e-01 -8.78234386e-01 4.91732091e-01 -2.41652414e-01
-1.74364603e+00 5.74436426e-01 -1.57432541e-01 -9.85751510e-01
-9.71395969e-01 -5.79436898e-01 -9.21483576e-01 6.95460081e-01
-1.30689025e+00 -7.21265137e-01 -1.81986543e-03 2.33319908e-01
1.37880707e+00 -5.25721371e-01 1.46406865e+00 -1.24455385e-01
-4.13271546e-01 6.34282053e-01 2.18766749e-01 4.22529906e-01
8.99248481e-01 -1.51592529e+00 6.60272181e-01 1.18857786e-01
3.52850161e-03 5.50034046e-01 6.88187599e-01 -3.32556009e-01
-1.08598900e+00 -6.82418883e-01 1.06116605e+00 -4.84481484e-01
7.85049498e-01 -4.79893595e-01 -1.15805042e+00 5.83209276e-01
8.27812910e-01 -4.71338868e-01 1.20307088e+00 6.16659820e-01
-4.16247100e-01 3.91344815e-01 -1.14082074e+00 7.39503741e-01
6.07042611e-01 -1.02630556e+00 -1.39342380e+00 6.15916252e-01
9.22577143e-01 -1.67437971e-01 -1.12487435e+00 6.67511374e-02
3.94598246e-01 -1.02011240e+00 9.81773674e-01 -1.07978666e+00
8.10403168e-01 7.22679555e-01 -1.46349818e-01 -1.59433973e+00
5.01888767e-02 -7.42825210e-01 -3.27998608e-01 1.28676450e+00
2.66889125e-01 -6.54932261e-01 5.86818933e-01 7.43701458e-01
-1.54138386e-01 -8.79762053e-01 -8.62349510e-01 -6.94933355e-01
3.20883811e-01 2.87659615e-01 1.88314393e-01 1.21326566e+00
8.70024323e-01 1.43124473e+00 -6.55966938e-01 -6.78587794e-01
2.08286449e-01 -4.50894833e-02 1.00192368e+00 -1.17182434e+00
-2.49605983e-01 -5.09606719e-01 -3.36178571e-01 -1.44606674e+00
6.28250659e-01 -8.16783309e-01 2.06300408e-01 -1.74602270e+00
2.74920225e-01 -1.90733150e-02 7.86758512e-02 1.98580429e-01
-5.67186117e-01 -5.83811700e-02 1.66091025e-01 1.89671444e-03
-7.96214759e-01 1.23194253e+00 8.40983570e-01 -4.04238552e-01
-2.68311888e-01 -1.34538775e-02 -6.74936950e-01 5.31808913e-01
8.99633586e-01 -1.09875645e-03 -7.22837269e-01 -7.20033571e-02
-4.47759271e-01 3.73658001e-01 -1.26335129e-01 -3.99224699e-01
-2.43442971e-02 5.97042218e-02 -2.75802433e-01 -2.59088248e-01
8.45271051e-01 -6.39522225e-02 -1.01114905e+00 4.61566031e-01
-1.19868410e+00 -8.82318318e-02 5.63594624e-02 7.54778504e-01
-3.16082656e-01 -7.43969977e-01 6.71722949e-01 -1.27160102e-01
-8.63806665e-01 -2.67827064e-01 -6.61852658e-01 6.46402359e-01
6.54827535e-01 8.06579292e-02 -7.94317186e-01 -8.11791420e-01
-4.58076924e-01 3.91731977e-01 1.36413738e-01 6.19053006e-01
7.80639708e-01 -1.19641697e+00 -7.98969150e-01 -1.33430387e-03
5.51980555e-01 -4.32557940e-01 2.27094457e-01 4.74940836e-01
-2.10737020e-01 5.44316709e-01 -1.85873985e-01 -2.90732443e-01
-1.48837245e+00 1.31618753e-01 4.41417955e-02 -5.05854964e-01
-7.94829249e-01 9.24958110e-01 1.78857565e-01 -9.59051073e-01
4.35023218e-01 1.07586257e-01 -6.18903637e-01 3.26092869e-01
8.40552986e-01 5.66767573e-01 -2.45523930e-01 -5.50990880e-01
8.50760266e-02 -2.96867251e-01 -4.96758908e-01 -2.07516551e-01
1.30393171e+00 -1.10945493e-01 -1.04296990e-01 1.05165851e+00
1.39940274e+00 -3.18741113e-01 -1.09131873e+00 -5.38555503e-01
1.70764342e-01 -5.23371279e-01 -2.12367684e-01 -5.80475569e-01
-3.02473187e-01 1.14624941e+00 2.96178311e-01 6.27226949e-01
4.04984623e-01 8.35390389e-02 8.43178093e-01 8.92059028e-01
1.30417809e-01 -1.07444823e+00 7.63113022e-01 9.23959374e-01
1.04216588e+00 -1.60515296e+00 5.38528804e-03 -1.37384415e-01
-1.31086123e+00 1.15922332e+00 7.16192663e-01 -6.08844347e-02
1.70802310e-01 -2.10480139e-01 6.15597062e-04 -4.24682707e-01
-1.08038640e+00 -1.44925952e-01 1.20616823e-01 4.99983937e-01
8.87263536e-01 -1.15664706e-01 -5.10101855e-01 4.15946126e-01
6.96664723e-03 -5.79591870e-01 6.41431451e-01 1.02965391e+00
-5.83088458e-01 -1.10143292e+00 3.35413158e-01 7.19133496e-01
-2.13155568e-01 -2.20244497e-01 -7.91801989e-01 7.48444319e-01
-1.05625033e+00 1.29902411e+00 1.88534454e-01 -4.02616233e-01
2.58541316e-01 4.69001561e-01 2.37649724e-01 -1.25550854e+00
-9.24369037e-01 -3.76533180e-01 8.40668201e-01 -2.42531568e-01
-3.76889825e-01 -4.02769655e-01 -1.16665208e+00 -3.11398357e-01
-4.59633291e-01 4.68871832e-01 2.43670747e-01 8.49624097e-01
3.38355988e-01 6.05803490e-01 8.74551594e-01 -6.45894706e-01
-1.24107671e+00 -1.44717133e+00 -5.91101408e-01 6.29386246e-01
3.78872812e-01 -6.42073274e-01 -4.92697209e-01 -5.02443612e-01] | [12.703225135803223, 7.877967357635498] |
a4a36d7d-bc0b-4ea6-8969-770c3bd8b98c | detecting-overfitting-of-deep-generative | 1901.03396 | null | http://arxiv.org/abs/1901.03396v1 | http://arxiv.org/pdf/1901.03396v1.pdf | Detecting Overfitting of Deep Generative Networks via Latent Recovery | State of the art deep generative networks are capable of producing images
with such incredible realism that they can be suspected of memorizing training
images. It is why it is not uncommon to include visualizations of training set
nearest neighbors, to suggest generated images are not simply memorized. We
demonstrate this is not sufficient and motivates the need to study
memorization/overfitting of deep generators with more scrutiny. This paper
addresses this question by i) showing how simple losses are highly effective at
reconstructing images for deep generators ii) analyzing the statistics of
reconstruction errors when reconstructing training and validation images, which
is the standard way to analyze overfitting in machine learning. Using this
methodology, this paper shows that overfitting is not detectable in the pure
GAN models proposed in the literature, in contrast with those using hybrid
adversarial losses, which are amongst the most widely applied generative
methods. The paper also shows that standard GAN evaluation metrics fail to
capture memorization for some deep generators. Finally, the paper also shows
how off-the-shelf GAN generators can be successfully applied to face inpainting
and face super-resolution using the proposed reconstruction method, without
hybrid adversarial losses. | ['Julien Rabin', 'Loic Simon', 'Frederic Jurie', 'Ryan Webster'] | 2019-01-09 | detecting-overfitting-of-deep-generative-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Webster_Detecting_Overfitting_of_Deep_Generative_Networks_via_Latent_Recovery_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Webster_Detecting_Overfitting_of_Deep_Generative_Networks_via_Latent_Recovery_CVPR_2019_paper.pdf | cvpr-2019-6 | ['facial-inpainting'] | ['computer-vision'] | [ 3.29741329e-01 5.63803136e-01 5.14643729e-01 -7.19964579e-02
-5.42380393e-01 -4.04637516e-01 8.67457867e-01 -3.79049182e-01
-1.83117539e-01 1.18211663e+00 -5.03335744e-02 -5.41473851e-02
-5.13782492e-03 -1.03059030e+00 -9.40801561e-01 -8.26281190e-01
1.92110151e-01 4.91180748e-01 -3.16800654e-01 -3.73185158e-01
8.68441761e-02 6.40114486e-01 -1.80134702e+00 2.22266331e-01
7.32069075e-01 7.72807777e-01 -1.12224013e-01 7.29676962e-01
5.51255746e-03 9.67669189e-01 -1.19258010e+00 -9.95147347e-01
1.99440137e-01 -1.16587996e+00 -6.97399259e-01 -8.73975977e-02
8.18923771e-01 -5.09907782e-01 -8.49937424e-02 1.02893353e+00
6.67425156e-01 -1.76788941e-01 7.54858017e-01 -1.24529183e+00
-8.90508413e-01 4.38425511e-01 -2.30775476e-01 2.15022564e-01
4.33068365e-01 2.81093687e-01 3.39550287e-01 -7.19043493e-01
6.87723696e-01 1.35535264e+00 8.33139002e-01 8.19489419e-01
-1.31839585e+00 -7.05583394e-01 -5.66362739e-01 2.02913210e-02
-1.21866477e+00 -4.99927968e-01 7.90140986e-01 -2.82007009e-01
7.31389582e-01 5.31099498e-01 5.39476335e-01 1.52843595e+00
2.06962749e-01 2.18772724e-01 1.57713950e+00 -5.40366769e-01
1.32218063e-01 3.76296997e-01 -5.62350333e-01 6.06574297e-01
3.80210310e-01 3.84893686e-01 -4.70515370e-01 7.14039356e-02
8.85852695e-01 -3.26304227e-01 -3.63806039e-01 7.54191130e-02
-6.98520184e-01 1.01289225e+00 3.55678767e-01 5.98667085e-01
-3.80356997e-01 2.90462822e-01 3.70812714e-01 6.44033194e-01
6.55706584e-01 6.16730869e-01 3.01636130e-01 -1.22049913e-01
-1.38400769e+00 2.91659564e-01 5.81086099e-01 6.97397470e-01
5.37547410e-01 7.38237143e-01 1.29941413e-02 6.86032474e-01
-1.44248605e-01 4.10679817e-01 5.82734168e-01 -9.46178496e-01
3.35780755e-02 1.64535239e-01 -8.04031640e-02 -1.04079521e+00
8.96641836e-02 -5.03922343e-01 -1.16489828e+00 9.09529269e-01
3.76702249e-01 8.63648802e-02 -7.60963917e-01 1.85137594e+00
-9.39163640e-02 1.39427707e-01 1.71274245e-01 5.69846392e-01
9.59395766e-01 4.09040511e-01 -6.26191348e-02 -3.33420962e-01
9.75939035e-01 -5.04163921e-01 -8.47199619e-01 5.07773794e-02
8.46625417e-02 -8.85538697e-01 1.14840794e+00 5.98967373e-01
-1.61631846e+00 -8.97963047e-01 -1.24309075e+00 1.18301280e-01
-3.26010346e-01 -7.74010792e-02 4.43696499e-01 1.11616004e+00
-1.40411544e+00 1.03707111e+00 -5.86947322e-01 -3.64024878e-01
7.01379836e-01 2.05610022e-01 -4.99105334e-01 6.95350319e-02
-1.09255004e+00 1.28557575e+00 1.21439993e-01 1.25073725e-02
-1.01591253e+00 -6.46362960e-01 -6.47862971e-01 -1.77812930e-02
-2.04228938e-01 -6.77051246e-01 7.75300860e-01 -1.50043881e+00
-1.42572927e+00 1.07808280e+00 9.16042030e-02 -6.74899936e-01
1.05653453e+00 -1.39264345e-01 -5.19241989e-01 1.67858526e-01
-1.85520411e-01 7.57272780e-01 1.25183487e+00 -1.59406877e+00
1.82189718e-02 -1.79207683e-01 -8.85608867e-02 -2.59640217e-01
-1.38831288e-01 -3.19794565e-01 3.14790100e-01 -9.24816251e-01
-3.44472885e-01 -6.48174942e-01 1.94104001e-01 -1.58987746e-01
-3.97378862e-01 8.74866098e-02 8.36737335e-01 -8.52036476e-01
7.92985439e-01 -1.96849990e+00 -7.99142569e-02 1.30912930e-01
1.23769149e-01 4.79384422e-01 -1.75893813e-01 4.55348670e-01
-3.90246600e-01 2.99670398e-01 -3.09327275e-01 -6.49040759e-01
-4.62558679e-02 3.66039306e-01 -7.24552989e-01 4.64267641e-01
3.34702730e-01 1.03269017e+00 -5.39210320e-01 -3.11462253e-01
4.01801825e-01 1.05192733e+00 -3.30660284e-01 2.34001204e-01
4.11575064e-02 6.55399501e-01 2.16804862e-01 2.60085106e-01
9.56463933e-01 1.28721863e-01 -2.06489176e-01 -4.31475136e-03
2.50320524e-01 8.42086077e-02 -8.97311032e-01 1.34187376e+00
-4.58109587e-01 1.02512383e+00 -3.06900054e-01 -1.06677318e+00
1.04438400e+00 4.16067958e-01 -3.11170332e-02 -9.25704241e-01
-1.71821881e-02 2.32745588e-01 -1.59576014e-01 -1.85992822e-01
4.31842864e-01 -4.86529380e-01 4.13593709e-01 4.10035223e-01
2.85396039e-01 -3.75993550e-01 -1.76097825e-03 -4.87035774e-02
9.97226655e-01 1.10534683e-01 7.65933022e-02 -1.43666178e-01
3.88178855e-01 -3.58967394e-01 -1.60156898e-02 7.47331679e-01
3.46808195e-01 1.06912446e+00 5.82700849e-01 -4.03813094e-01
-1.37841415e+00 -1.14016759e+00 -4.71283942e-02 4.45272326e-01
-3.91818464e-01 -1.17327131e-01 -1.31761861e+00 -4.31233406e-01
-2.92209476e-01 8.86397779e-01 -1.08240521e+00 -3.34331602e-01
-4.49542910e-01 -5.91452837e-01 9.38475549e-01 3.26696068e-01
7.24523604e-01 -1.42566669e+00 -8.30334783e-01 -2.15669852e-02
6.80434285e-03 -7.40647435e-01 2.25660548e-01 4.08082716e-02
-9.90600944e-01 -8.79798591e-01 -1.05353975e+00 -4.92514700e-01
8.77097905e-01 -3.23746532e-01 1.55980217e+00 4.52828109e-01
-3.30105364e-01 3.78839850e-01 -3.19190443e-01 -3.56058121e-01
-1.24340856e+00 -2.03997806e-01 -1.15171589e-01 -2.37881273e-01
-1.32073537e-01 -9.58276093e-01 -7.16858447e-01 1.15382247e-01
-1.24934506e+00 -1.71440527e-01 6.74782932e-01 8.26484501e-01
3.54516566e-01 1.14646442e-01 6.89061403e-01 -1.21150291e+00
8.10730338e-01 -1.95369050e-01 -2.77430326e-01 1.57060847e-01
-7.58090377e-01 4.09115013e-03 8.95240903e-01 -3.77404749e-01
-1.03583908e+00 -3.46428603e-01 -4.92939025e-01 -5.08417785e-01
-3.57502431e-01 1.07232474e-01 -1.84573874e-01 -3.33604574e-01
1.06622279e+00 4.24522072e-01 2.54233688e-01 -3.40713561e-01
2.08159313e-01 6.94224313e-02 7.10953593e-01 -1.83162317e-01
1.00659859e+00 5.62332630e-01 2.45259285e-01 -8.03132117e-01
-3.30507815e-01 4.42519367e-01 -3.06511164e-01 -3.39451402e-01
5.89220285e-01 -5.78101039e-01 -5.71428001e-01 4.29119736e-01
-1.21989369e+00 -2.93659270e-01 -8.62625360e-01 6.28804928e-03
-7.99466252e-01 3.32382500e-01 -6.10874891e-01 -7.78391242e-01
-2.71918774e-01 -9.91308689e-01 7.71618426e-01 1.05777808e-01
-3.63150686e-01 -1.00545478e+00 3.63734625e-02 2.50366956e-01
8.01862121e-01 8.04061949e-01 7.83235133e-01 -2.27407277e-01
-3.01139504e-01 -1.92134038e-01 1.16565563e-01 9.30061817e-01
1.51203945e-01 9.61417407e-02 -1.34451735e+00 -4.83763278e-01
4.00671065e-01 -4.20903742e-01 7.73273706e-01 2.88804233e-01
1.15856004e+00 -5.29803276e-01 2.26522967e-01 6.42755032e-01
1.49331391e+00 3.81263010e-02 1.59625661e+00 1.42032653e-01
2.95127004e-01 6.78960264e-01 1.58043072e-01 2.07926750e-01
-4.19531316e-01 5.81221819e-01 7.04241931e-01 -3.10460150e-01
-4.37334687e-01 -4.56924915e-01 3.92358899e-01 3.40230048e-01
-5.03508627e-01 -3.42108876e-01 -3.73403043e-01 1.62464127e-01
-1.18266153e+00 -1.35695004e+00 -8.28333125e-02 2.25897646e+00
7.06384003e-01 1.50577530e-01 4.56754155e-02 5.88853300e-01
5.76036155e-01 6.35361299e-02 -2.03480914e-01 -7.64954627e-01
-5.28631449e-01 9.98950303e-01 3.21718633e-01 4.09251809e-01
-7.10067809e-01 6.08834505e-01 6.51909971e+00 9.85222161e-01
-1.13409579e+00 3.97339314e-01 1.03627658e+00 3.26498151e-02
-4.80138808e-01 -2.03425318e-01 -1.83836833e-01 7.06583560e-01
9.88421500e-01 1.17829405e-01 4.43383902e-01 6.85349226e-01
-1.98331475e-02 -1.84577182e-01 -1.08931792e+00 9.81841207e-01
4.15436119e-01 -1.46983778e+00 3.80393863e-01 9.01779756e-02
7.99173772e-01 -6.90242708e-01 6.56646192e-01 5.96648306e-02
1.68798938e-02 -1.93353045e+00 6.71328664e-01 7.94886529e-01
1.04276693e+00 -9.64836061e-01 7.74906993e-01 1.55223638e-01
-4.42596555e-01 3.66070330e-01 -5.21698534e-01 4.75675464e-02
6.07364550e-02 6.21326804e-01 -7.97697186e-01 5.21567941e-01
5.64388752e-01 1.81465894e-01 -8.64499748e-01 7.47943044e-01
-3.39167178e-01 7.80439556e-01 -1.09944150e-01 4.24164921e-01
-7.42512792e-02 -6.06238358e-02 4.71165657e-01 1.03083920e+00
6.95669293e-01 -2.97212541e-01 -9.58445191e-01 1.28476310e+00
-3.87295336e-02 -8.41533020e-02 -9.28702056e-01 1.12348147e-01
1.15677252e-01 9.26928878e-01 -7.39330947e-01 -1.73614696e-01
8.79075471e-03 1.38653994e+00 -1.52372820e-02 2.38600746e-01
-9.12811816e-01 -1.03407994e-01 2.98967987e-01 4.33585137e-01
1.06180057e-01 1.07269779e-01 -2.27715224e-01 -6.61954820e-01
6.33497760e-02 -1.02231801e+00 9.46921948e-03 -1.03168619e+00
-1.27159107e+00 1.06731129e+00 -2.13902211e-04 -1.12297738e+00
-7.02175379e-01 -3.39162439e-01 -8.42321754e-01 1.09389937e+00
-1.15907598e+00 -1.16495228e+00 -5.37141681e-01 6.36127591e-01
2.63482660e-01 -2.95052826e-01 1.03818500e+00 2.46256053e-01
-4.54380997e-02 9.08090055e-01 -1.30053937e-01 -2.09558934e-01
5.26393652e-01 -1.15929341e+00 4.17975575e-01 8.86668861e-01
3.70721817e-01 4.93210584e-01 1.20434535e+00 -3.92285436e-01
-8.45063150e-01 -9.44415212e-01 6.13938034e-01 -3.48300070e-01
-9.44584087e-02 -2.55327582e-01 -1.09231877e+00 4.11661714e-01
7.11201668e-01 -1.34324372e-01 4.96147782e-01 -5.28478563e-01
-2.54573256e-01 -3.95431370e-02 -1.61791730e+00 3.95143360e-01
8.67338419e-01 -4.17539775e-01 -4.79801238e-01 2.10002884e-01
2.63263762e-01 -2.53561437e-01 -6.57314003e-01 2.78186947e-01
2.75802970e-01 -1.61192667e+00 1.06601453e+00 -2.83380449e-01
9.77479339e-01 4.58154902e-02 2.21080184e-01 -1.40863550e+00
1.25840813e-01 -6.87799096e-01 1.12592258e-01 1.56006324e+00
1.76827341e-01 -6.92372262e-01 9.71947014e-01 2.39555150e-01
7.00311139e-02 -4.55104262e-01 -1.24749684e+00 -8.48220527e-01
3.59880179e-01 -1.66706070e-01 5.85650027e-01 1.00627232e+00
-6.90016627e-01 -1.09332711e-01 -6.04749143e-01 -3.32972467e-01
5.74675739e-01 -2.69547164e-01 9.42480564e-01 -9.91028607e-01
-3.48833054e-01 -5.02493680e-01 -7.89673626e-01 -2.76281178e-01
2.48946503e-01 -6.61317229e-01 -3.10369372e-01 -1.33132923e+00
-7.90664703e-02 -2.95377374e-01 6.34157211e-02 1.82450339e-01
1.23493679e-01 8.69515479e-01 4.76231463e-02 1.63550183e-01
1.20506123e-01 4.51434225e-01 1.40122473e+00 6.32877201e-02
3.28349113e-01 -6.86758831e-02 -6.63667679e-01 5.51803350e-01
8.08313906e-01 -3.51931155e-01 -6.23760581e-01 -1.25336677e-01
4.23642129e-01 -1.29826039e-01 9.66705799e-01 -1.39161420e+00
-2.15483010e-01 2.75348246e-01 8.53465319e-01 -1.43298790e-01
5.53361475e-01 -8.06193948e-01 8.57009709e-01 5.88427842e-01
-2.88104415e-01 2.40244582e-01 3.03683639e-01 2.72605032e-01
-3.97305369e-01 -3.95733476e-01 1.16307199e+00 -4.14834857e-01
-3.30632597e-01 -1.11655302e-01 -1.43298253e-01 6.04831129e-02
8.82679224e-01 -5.05879939e-01 -2.81699091e-01 -9.11104739e-01
-6.48425639e-01 -7.15337873e-01 7.03315854e-01 9.27004591e-02
8.23274434e-01 -1.32346380e+00 -8.60727668e-01 3.23269308e-01
-4.06485349e-01 -2.69921690e-01 4.73433495e-01 5.37905455e-01
-9.34805572e-01 -3.77032682e-02 -7.13137448e-01 -3.91738534e-01
-1.32658875e+00 5.87646127e-01 4.32104379e-01 -1.08970001e-01
-6.05006874e-01 9.53336716e-01 2.00121272e-02 1.51490077e-01
-1.28397476e-02 1.79256853e-02 4.55861166e-03 4.51268535e-03
3.70504200e-01 2.90323198e-01 3.18827122e-01 -6.44286513e-01
-6.48206547e-02 2.96239346e-01 1.84685349e-01 -1.80529401e-01
1.26100647e+00 2.14000329e-01 -2.40011856e-01 2.41311267e-01
1.04251313e+00 7.38124475e-02 -1.14923990e+00 4.70275462e-01
-5.54475009e-01 -5.88498890e-01 -2.96620488e-01 -8.65776122e-01
-1.37641180e+00 1.17619538e+00 8.31668496e-01 4.50818777e-01
1.42378378e+00 -2.26846904e-01 5.78999937e-01 -1.91022947e-01
3.24347079e-01 -8.28124106e-01 2.15918168e-01 2.15992182e-02
1.40082300e+00 -9.71029341e-01 5.45221660e-03 -1.75063565e-01
-5.07087231e-01 1.01417410e+00 4.20975626e-01 -5.05619407e-01
3.16044450e-01 4.05024111e-01 -5.84631264e-02 -7.71703422e-02
-4.46578294e-01 1.06278628e-01 5.21590300e-02 8.51073861e-01
4.42923695e-01 -1.35520712e-01 -4.21782941e-01 1.73588112e-01
-7.25578666e-01 -4.73615341e-02 7.14430809e-01 5.84348261e-01
1.90334339e-02 -1.25700796e+00 -5.59949160e-01 3.72995794e-01
-6.99684083e-01 -1.00189403e-01 -5.32333136e-01 1.13121223e+00
4.01813090e-01 7.12865651e-01 9.41369981e-02 -3.40040535e-01
1.00583620e-01 1.34568959e-01 9.02286172e-01 -2.21096352e-01
-8.44054520e-01 -4.56055105e-01 -1.93131432e-01 -3.90953332e-01
-5.60309291e-01 -2.87839502e-01 -6.54651761e-01 -7.24221289e-01
-1.61609173e-01 1.99774668e-01 6.79196000e-01 6.30386949e-01
1.90971896e-01 6.23661280e-01 4.28380281e-01 -7.97900081e-01
-4.27149445e-01 -1.10303247e+00 -4.93894488e-01 7.61490047e-01
2.80361652e-01 -5.87423682e-01 -6.87123239e-01 2.63110667e-01] | [12.050931930541992, -0.2565203607082367] |
9a403219-b619-4051-bc6e-61c72920b174 | group-wise-deep-co-saliency-detection | 1707.07381 | null | http://arxiv.org/abs/1707.07381v2 | http://arxiv.org/pdf/1707.07381v2.pdf | Group-wise Deep Co-saliency Detection | In this paper, we propose an end-to-end group-wise deep co-saliency detection
approach to address the co-salient object discovery problem based on the fully
convolutional network (FCN) with group input and group output. The proposed
approach captures the group-wise interaction information for group images by
learning a semantics-aware image representation based on a convolutional neural
network, which adaptively learns the group-wise features for co-saliency
detection. Furthermore, the proposed approach discovers the collaborative and
interactive relationships between group-wise feature representation and
single-image individual feature representation, and model this in a
collaborative learning framework. Finally, we set up a unified end-to-end deep
learning scheme to jointly optimize the process of group-wise feature
representation learning and the collaborative learning, leading to more
reliable and robust co-saliency detection results. Experimental results
demonstrate the effectiveness of our approach in comparison with the
state-of-the-art approaches. | ['Xi Li', 'Omar El Farouk Bourahla', 'Fei Wu', 'Shanshan Zhao', 'Lina Wei'] | 2017-07-24 | null | null | null | null | ['co-saliency-detection'] | ['computer-vision'] | [ 2.56828696e-01 -2.32300926e-02 -6.26445115e-02 -5.90179741e-01
-5.45726895e-01 9.70340520e-02 4.28077251e-01 2.88175821e-01
-3.24257493e-01 1.21745147e-01 3.66685838e-01 2.32135117e-01
-4.84110892e-01 -5.38440943e-01 -9.11510110e-01 -4.37131941e-01
-2.67310381e-01 -6.57625347e-02 5.37905037e-01 5.36331795e-02
4.18474764e-01 2.02932805e-01 -2.11180043e+00 5.26026130e-01
1.14401698e+00 1.26613593e+00 9.19389009e-01 4.57826704e-01
1.23440310e-01 1.02326953e+00 -1.03574023e-01 2.57882714e-01
-1.20341014e-02 -4.05816078e-01 -7.29852498e-01 4.56610113e-01
6.41328156e-01 -9.63103846e-02 5.51329404e-02 1.17753482e+00
4.89095628e-01 3.44485939e-01 3.27077389e-01 -1.30217206e+00
-8.85051012e-01 6.63325667e-01 -7.49629319e-01 4.37494159e-01
-2.46315505e-02 1.08488895e-01 1.28449237e+00 -1.44617212e+00
3.67570490e-01 1.36807299e+00 2.59712815e-01 2.62933597e-02
-8.71098220e-01 -7.52189457e-01 5.78785419e-01 6.20684624e-01
-1.38449979e+00 -2.93184340e-01 1.25979531e+00 -2.13126853e-01
6.96330905e-01 -5.39374873e-02 7.53885508e-01 3.32948267e-01
-1.18728817e-01 1.37259603e+00 6.83145881e-01 -6.08940542e-01
1.36860669e-01 -1.09779820e-01 -2.07052827e-02 1.04086328e+00
2.20803872e-01 -1.56017244e-01 -9.78315115e-01 -8.30024332e-02
9.05343413e-01 6.87000394e-01 4.06294279e-02 -8.65936220e-01
-1.50018215e+00 6.88238919e-01 1.04625201e+00 4.66612607e-01
-6.72166348e-01 3.71603996e-01 2.89343417e-01 -9.26205795e-03
5.75420499e-01 2.43597865e-01 -4.20638472e-01 5.44066310e-01
-1.07439721e+00 4.16387349e-01 1.73447028e-01 8.88113320e-01
1.17060566e+00 -2.59987682e-01 -5.31305134e-01 7.72358894e-01
5.95414698e-01 1.02707662e-01 5.84836721e-01 -6.28904283e-01
2.91351885e-01 9.43176687e-01 7.09006339e-02 -1.05933046e+00
-4.09900129e-01 -8.34670782e-01 -5.45748830e-01 1.34766281e-01
-7.43778273e-02 -7.73736164e-02 -6.88485205e-01 1.74820507e+00
4.46619213e-01 6.70124412e-01 5.02202101e-02 1.24951625e+00
9.11119640e-01 2.03712031e-01 2.68166214e-01 -3.77842283e-04
1.23568010e+00 -1.65805650e+00 -6.19730592e-01 -1.95773005e-01
5.02117872e-01 -7.41003156e-01 8.74720156e-01 -2.96339303e-01
-1.17622828e+00 -9.44177091e-01 -9.67319369e-01 -2.21933156e-01
-3.00984293e-01 5.57704329e-01 7.46863008e-01 -2.50682503e-01
-1.15355051e+00 4.99213576e-01 -6.24521971e-01 -2.69842267e-01
9.05881643e-01 5.43127716e-01 -6.62273774e-03 5.04715964e-02
-9.66711164e-01 4.61882144e-01 8.03423584e-01 3.42395231e-02
-1.08045506e+00 -9.53006625e-01 -8.87542009e-01 3.94851238e-01
6.12479031e-01 -9.39091206e-01 1.28455222e+00 -1.49854589e+00
-1.17121685e+00 7.28619814e-01 -3.44657868e-01 -5.58761895e-01
2.01605469e-01 -5.41182339e-01 -2.14077935e-01 2.43842244e-01
3.98114383e-01 7.51066864e-01 1.08204567e+00 -1.28917837e+00
-1.20896101e+00 -2.67377496e-01 1.45160668e-02 6.27267957e-01
-4.95579898e-01 3.80414069e-01 -3.93855065e-01 -5.90791881e-01
1.08153738e-01 -3.55385333e-01 -3.72388512e-01 2.73445487e-01
-2.00231776e-01 -5.02425730e-01 1.27631545e+00 -1.87228024e-01
1.09440017e+00 -2.27207851e+00 2.28001982e-01 1.65134890e-03
5.62547505e-01 3.21344376e-01 -1.58817634e-01 2.23256811e-01
-1.44697636e-01 -2.58327603e-01 -1.72509700e-01 -7.70812988e-01
-2.28844613e-01 -2.95337766e-01 -1.06926434e-01 4.08256084e-01
4.89777058e-01 1.27253842e+00 -1.36746871e+00 -6.41194820e-01
4.28800493e-01 3.84984165e-01 -4.50447172e-01 4.20463681e-01
-1.50486097e-01 2.28194565e-01 -5.68159401e-01 7.68152535e-01
5.65913260e-01 -5.27760744e-01 -6.45002276e-02 -3.03761423e-01
-3.28081548e-01 -1.92838803e-01 -9.52392578e-01 2.00777125e+00
-4.18715030e-01 3.50920022e-01 -1.05062924e-01 -1.05930388e+00
1.03542376e+00 -2.36472607e-01 4.33631778e-01 -5.01959205e-01
6.09241463e-02 3.10395002e-01 -1.87803544e-02 -3.45765024e-01
5.13985813e-01 3.12283814e-01 6.46908134e-02 4.33828026e-01
4.77189600e-01 5.26085913e-01 -2.23658994e-01 3.81429017e-01
4.79947329e-01 1.83494310e-04 3.55404556e-01 -4.47366178e-01
7.47439384e-01 -2.83713728e-01 3.84313703e-01 7.52279878e-01
-3.39517564e-01 5.34280598e-01 -1.32763758e-02 -3.81960809e-01
-6.73211515e-01 -6.64985955e-01 4.15969282e-01 1.52809536e+00
7.40292072e-01 -2.48969376e-01 -5.57132423e-01 -7.99761534e-01
1.66610196e-01 3.25915366e-01 -8.17571163e-01 -5.17411053e-01
-2.71180660e-01 -2.53538877e-01 -1.32462338e-01 6.47111237e-01
8.42597365e-01 -1.26998961e+00 -8.43306243e-01 2.52710164e-01
-3.81925702e-02 -9.70343292e-01 -6.90189362e-01 4.82713478e-03
-7.60956526e-01 -1.10599792e+00 -6.10073149e-01 -1.39434052e+00
8.14539731e-01 1.02467823e+00 8.38777781e-01 4.43954140e-01
-2.80249238e-01 3.24993879e-01 -3.95377278e-01 -5.87937176e-01
3.31714332e-01 -4.67136800e-02 -1.93548977e-01 7.12970018e-01
3.53925943e-01 -4.88344282e-01 -1.06021845e+00 2.72230417e-01
-7.96817660e-01 4.04057175e-01 7.59277999e-01 7.78205931e-01
8.06088269e-01 -5.65857701e-02 8.82001042e-01 -6.26284003e-01
3.82382363e-01 -5.70411742e-01 -4.43877935e-01 3.80892217e-01
-3.61748248e-01 -1.46258175e-01 4.00934190e-01 -2.12217793e-01
-1.05999267e+00 2.44766533e-01 5.52395403e-01 -1.00213170e+00
-3.27471830e-02 6.20201051e-01 -2.96597421e-01 -2.65650004e-01
2.86943883e-01 3.64550918e-01 -9.56765786e-02 -5.01897335e-01
6.92458212e-01 4.50715184e-01 8.16052794e-01 -1.45711854e-01
5.90021253e-01 5.74386418e-01 -3.16450268e-01 -3.57793212e-01
-1.29859567e+00 -9.46602225e-01 -9.13854539e-01 -4.13723201e-01
8.39033544e-01 -1.50097609e+00 -3.33987802e-01 6.75205946e-01
-1.00956881e+00 -4.17665653e-02 -3.23532432e-01 3.91615689e-01
-7.24201322e-01 2.26100504e-01 3.24631408e-02 -7.22998023e-01
-5.68985641e-01 -1.02896690e+00 1.43679607e+00 6.49466395e-01
1.81464747e-01 -1.04979241e+00 -3.27379286e-01 1.69066399e-01
3.87934655e-01 3.61416757e-01 4.25848722e-01 -7.89355397e-01
-8.60023141e-01 1.01925299e-01 -8.04489851e-01 6.93665594e-02
3.24008882e-01 -2.68506110e-01 -8.45893979e-01 -3.63498062e-01
-3.28508198e-01 -2.50468761e-01 1.18497515e+00 4.51060593e-01
1.60127878e+00 -1.57244280e-01 -4.49893892e-01 5.69042742e-01
1.61501420e+00 -2.35849515e-01 1.81815445e-01 2.56359071e-01
1.10612619e+00 5.11104167e-01 1.05356812e+00 5.70384800e-01
7.92305291e-01 5.64209700e-01 8.48396719e-01 -4.84047949e-01
-3.97563159e-01 -4.01808381e-01 1.33966550e-01 4.80698586e-01
1.64866388e-01 2.51999319e-01 -4.62775052e-01 1.09218693e+00
-2.56391358e+00 -9.19182122e-01 1.55105526e-02 1.85443819e+00
5.22849739e-01 -1.73237324e-01 2.45683014e-01 -1.22320488e-01
1.05943847e+00 3.80992681e-01 -6.84791625e-01 8.31408575e-02
-1.29620314e-01 -5.83995543e-02 3.39767367e-01 8.44729096e-02
-1.52682030e+00 1.31258678e+00 5.50475121e+00 8.21004808e-01
-1.19290662e+00 4.19128627e-01 4.07070965e-01 -1.39965579e-01
-3.12490344e-01 -6.97922930e-02 -5.48730195e-01 4.91363078e-01
3.86953056e-01 -5.20800292e-01 1.71090096e-01 9.43325996e-01
3.09195727e-01 -9.08405036e-02 -8.98811460e-01 9.98541176e-01
4.03423369e-01 -1.57890713e+00 2.37956066e-02 -1.62858829e-01
1.09209216e+00 8.11349601e-02 2.88710415e-01 4.40799557e-02
3.63569260e-01 -6.29696727e-01 1.14501667e+00 6.65689588e-01
4.09622997e-01 -9.82549191e-01 5.30935705e-01 2.62710929e-01
-1.51462948e+00 -5.20454228e-01 -4.29959118e-01 -1.12371355e-01
-9.03120730e-03 6.88035905e-01 -7.77005315e-01 7.46991932e-01
8.03320706e-01 1.40831590e+00 -8.14189315e-01 1.38840628e+00
-1.99336931e-01 3.58036011e-01 4.46269996e-02 -1.40707240e-01
5.13646543e-01 3.09838653e-01 4.75234658e-01 1.12464368e+00
1.37677193e-01 -9.90725830e-02 5.77959597e-01 1.07742918e+00
-7.14578554e-02 1.11226059e-01 -2.75710225e-01 1.57744706e-01
4.88877654e-01 1.40692484e+00 -8.32720757e-01 -5.08779824e-01
-3.30999494e-01 1.02627611e+00 7.91411400e-01 2.69427478e-01
-6.80805445e-01 -4.52280253e-01 7.15489566e-01 -2.40536928e-01
8.83152664e-01 -7.03189820e-02 -2.41623610e-01 -1.01306558e+00
2.92311497e-02 -2.93344676e-01 4.09440249e-01 -7.04261005e-01
-1.16717160e+00 2.11357355e-01 3.29856155e-03 -1.29494846e+00
-1.55985609e-01 -1.01117238e-01 -9.93611217e-01 8.80796075e-01
-2.09372211e+00 -1.72886717e+00 -3.90239805e-01 8.52344394e-01
7.81497836e-01 -2.20383272e-01 3.79304975e-01 8.78982246e-02
-3.31598610e-01 5.93372822e-01 -1.18856244e-01 -1.82743013e-01
4.96758550e-01 -1.10040414e+00 2.50116825e-01 1.06696105e+00
1.65081903e-01 4.37268853e-01 3.74879777e-01 -4.89194393e-01
-1.26854467e+00 -1.70319998e+00 9.34493303e-01 -3.07569206e-02
6.06271684e-01 -3.36982578e-01 -8.86699557e-01 3.57961595e-01
2.35406324e-01 5.51238239e-01 6.33986652e-01 -7.58591760e-03
-2.76698977e-01 -3.92515324e-02 -1.03056681e+00 3.96969527e-01
1.41461360e+00 -7.42428362e-01 -5.74007332e-01 3.61194074e-01
1.25189972e+00 -2.65157759e-01 -4.88733113e-01 3.65223020e-01
3.82500082e-01 -8.36325049e-01 1.02852619e+00 -3.89671534e-01
4.98472393e-01 -8.15732062e-01 8.44710879e-03 -1.25455832e+00
-8.89275730e-01 -3.78369480e-01 -3.43373537e-01 1.10628796e+00
2.59898193e-02 -3.22660595e-01 4.59994286e-01 3.81663665e-02
-4.07065213e-01 -9.74255323e-01 -5.64536512e-01 -3.65206242e-01
-5.29660463e-01 -3.78653780e-02 7.54374385e-01 7.95209646e-01
-2.65181482e-01 8.54908079e-02 -4.30663019e-01 4.75504458e-01
9.20993984e-01 7.41103113e-01 5.48114419e-01 -1.28965819e+00
-1.02018766e-01 -4.85271275e-01 -6.54967546e-01 -7.63803184e-01
3.09345871e-01 -1.11045814e+00 1.34442121e-01 -1.58447134e+00
5.47846437e-01 -2.77139008e-01 -1.06189859e+00 6.45790815e-01
-6.95031345e-01 1.95341304e-01 5.34728229e-01 2.39508271e-01
-1.52497840e+00 7.07633197e-01 1.26507628e+00 -1.37457877e-01
-2.74183929e-01 -3.34159613e-01 -1.20423400e+00 5.68183124e-01
5.20105958e-01 -1.92712232e-01 -5.69223762e-01 -4.46413606e-01
-3.24009150e-01 -5.51227689e-01 8.62573981e-01 -1.08793378e+00
6.01915300e-01 -1.29788399e-01 4.97031242e-01 -8.73458564e-01
-1.66031107e-01 -8.11364770e-01 -3.76926750e-01 2.35427737e-01
-7.20327973e-01 -4.18560654e-01 -2.04172917e-02 7.62790561e-01
-4.71580774e-01 1.39240667e-01 7.36635685e-01 -1.41727507e-01
-1.29424095e+00 5.28669655e-01 2.26832718e-01 -2.71746367e-01
1.19480741e+00 -9.31263044e-02 -6.34456053e-02 -1.82612598e-01
-6.20299518e-01 5.88950455e-01 3.19171786e-01 8.63164306e-01
9.31969762e-01 -1.47628260e+00 -6.37727559e-01 3.95678788e-01
5.01445651e-01 2.25523040e-01 5.44670641e-01 7.12295115e-01
6.42301291e-02 -4.32535494e-03 -3.22093576e-01 -8.70637596e-01
-1.29135609e+00 6.72538340e-01 1.45355538e-01 -1.30473912e-01
-4.49721694e-01 1.11630034e+00 4.95875061e-01 -1.48983315e-01
2.05782190e-01 -1.79927602e-01 -5.37359476e-01 -6.15347065e-02
8.39151084e-01 2.24390393e-03 -1.14436515e-01 -8.87289941e-01
-4.51163322e-01 5.39914012e-01 -4.29979175e-01 3.14145982e-01
1.44332027e+00 -4.09777164e-01 -1.55050710e-01 3.65357041e-01
1.40942895e+00 -5.82113326e-01 -1.47660112e+00 -7.97180653e-01
-7.43392706e-02 -7.39965618e-01 2.99562812e-01 -6.78746462e-01
-1.33087671e+00 6.29210174e-01 8.08173716e-01 -3.38321060e-01
1.29999340e+00 2.68395662e-01 6.41472399e-01 1.81794658e-01
3.89662564e-01 -1.01901817e+00 4.45229769e-01 3.48575801e-01
9.91199613e-01 -1.44429731e+00 -2.62378044e-02 -5.34257829e-01
-5.62683642e-01 7.86005616e-01 7.75562942e-01 -5.50288439e-01
8.86229217e-01 -4.29937065e-01 -2.85781354e-01 -4.41092432e-01
-8.63035202e-01 -8.13478827e-01 6.08875513e-01 4.78304625e-01
1.07448317e-01 1.90093279e-01 -2.16276556e-01 6.91192806e-01
2.26284370e-01 2.89149225e-01 -1.44756176e-02 1.02954865e+00
-8.17976236e-01 -6.49139464e-01 -1.27277240e-01 3.74537915e-01
-1.87784404e-01 -2.14542560e-02 -3.15730333e-01 1.63924485e-01
4.75672930e-01 8.92929554e-01 3.23482543e-01 -4.48231429e-01
1.44071996e-01 -2.36707553e-01 1.16762593e-01 -7.87207425e-01
-6.38582647e-01 8.64829719e-02 -4.30766076e-01 -7.20294237e-01
-8.69288683e-01 -7.41996646e-01 -1.42063141e+00 5.21164000e-01
-4.55693275e-01 -1.36975288e-01 5.44165373e-01 9.95538533e-01
8.51738393e-01 7.76797175e-01 1.08748746e+00 -1.33188164e+00
-1.16135404e-01 -7.52258360e-01 -5.67407191e-01 5.78850806e-01
6.30454123e-01 -8.39303374e-01 -1.18015870e-01 5.32945767e-02] | [9.826104164123535, -0.27957963943481445] |
697a4d55-f2bb-4945-b01e-a9ccd92f9377 | global-entity-disambiguation-with-bert | null | null | https://openreview.net/forum?id=STY-d6qQS9t | https://openreview.net/pdf?id=STY-d6qQS9t | Global Entity Disambiguation with BERT | We propose a global entity disambiguation (ED) model based on BERT. To capture global contextual information for ED, our model treats not only words but also entities as input tokens, and solves the task by sequentially resolving mentions to their referent entities and using resolved entities as inputs. We train the model using a large entity-annotated corpus obtained from Wikipedia. We achieve new state-of-the-art results on five standard ED datasets: AIDA-CoNLL, MSNBC, AQUAINT, ACE2004, and WNED-WIKI. | ['Anonymous'] | 2022-01-16 | null | null | null | acl-arr-january-2022-1 | ['entity-disambiguation'] | ['natural-language-processing'] | [-8.61033142e-01 4.39337701e-01 -4.49943155e-01 -2.54222900e-01
-7.98948765e-01 -8.12534392e-01 7.93870151e-01 6.83396041e-01
-1.34103346e+00 1.12926555e+00 5.47145486e-01 4.06878777e-02
6.26352951e-02 -7.95154512e-01 -4.48783785e-01 9.01983008e-02
-2.77209044e-01 9.10542488e-01 5.41400969e-01 -4.90868121e-01
-5.88958561e-02 1.10115983e-01 -1.18295038e+00 2.56284356e-01
1.39612794e+00 6.59034252e-01 2.90301204e-01 2.36125544e-01
-7.25494623e-01 9.10419047e-01 -9.45504367e-01 -9.72760558e-01
-1.60285622e-01 3.48687202e-01 -1.39647138e+00 -6.15700603e-01
3.26187402e-01 2.40915582e-01 -3.63857955e-01 1.14141393e+00
4.33488935e-01 3.99142236e-01 7.19690800e-01 -1.09636581e+00
-1.31030059e+00 1.25071931e+00 -2.58539528e-01 3.10566604e-01
3.61042708e-01 -4.73988533e-01 1.65905523e+00 -1.24212694e+00
1.42413890e+00 1.28473055e+00 4.42293644e-01 5.38969636e-01
-7.13118434e-01 -6.35633528e-01 4.41698849e-01 4.48953450e-01
-1.71567309e+00 -2.39580825e-01 -1.26234949e-01 -1.03223443e-01
1.89233816e+00 2.10684445e-03 9.80871469e-02 7.89973319e-01
-4.21681345e-01 8.58136654e-01 3.63011032e-01 -5.68876386e-01
4.52756556e-03 -1.37405947e-01 7.07154930e-01 5.98001540e-01
6.86402977e-01 -5.11338294e-01 -3.39508682e-01 -2.22591497e-02
1.21043012e-01 -5.13618052e-01 -2.67612576e-01 1.27251014e-01
-1.11980999e+00 5.18675387e-01 5.07788777e-01 6.31918669e-01
-6.00577652e-01 -7.06217885e-02 5.68549633e-01 1.19471058e-01
5.97000539e-01 1.17921412e+00 -9.94795918e-01 9.80640529e-04
-4.27244157e-01 4.82516766e-01 1.04893279e+00 1.22690284e+00
6.98352754e-01 -4.98531967e-01 -2.76067764e-01 1.16399908e+00
2.09877744e-01 6.06430590e-01 6.55084074e-01 -6.38921857e-01
9.93696928e-01 7.36040473e-01 6.18470848e-01 -9.09283459e-01
-6.25460923e-01 -2.34853417e-01 -4.54309672e-01 -6.33633792e-01
-7.08268136e-02 -5.08254051e-01 -1.07065809e+00 1.82284331e+00
3.92067760e-01 3.24917942e-01 4.96026039e-01 6.36363149e-01
1.73767674e+00 5.99084616e-01 8.65008056e-01 1.46710142e-01
1.47630191e+00 -1.05745447e+00 -1.11092472e+00 -6.39654517e-01
9.22978282e-01 -3.43928188e-01 4.99554843e-01 -1.96178004e-01
-7.33864903e-01 -1.85194373e-01 -7.08356082e-01 -5.14177144e-01
-1.24885011e+00 3.98382723e-01 4.97056186e-01 -1.71048731e-01
-7.96075523e-01 3.51295352e-01 -6.46003783e-01 -5.35533130e-01
-1.18353173e-01 9.43770334e-02 -6.51089549e-01 4.32454310e-02
-2.22475791e+00 1.35293365e+00 1.05380046e+00 1.02293007e-01
-2.87124962e-01 -7.48862565e-01 -1.45123768e+00 2.00300902e-01
5.13991714e-01 -5.73355973e-01 1.23758924e+00 6.78616241e-02
-6.45831287e-01 1.15358865e+00 -3.32016230e-01 -6.68512166e-01
-2.12934047e-01 -5.39414704e-01 -1.14689970e+00 -1.77781746e-01
8.34328830e-01 7.27900803e-01 -1.76799759e-01 -1.23621619e+00
-1.14163494e+00 -1.13323972e-01 3.39353770e-01 2.10994825e-01
-1.84641823e-01 3.27269614e-01 -7.24896610e-01 -6.34815574e-01
-3.51588011e-01 -6.25219464e-01 -2.55301863e-01 -8.66885424e-01
-8.11686099e-01 -8.75862002e-01 4.50678825e-01 -7.91836560e-01
1.74226606e+00 -1.81919491e+00 1.79385304e-01 -1.30813688e-01
4.42613840e-01 5.72700977e-01 -3.89415711e-01 4.70407963e-01
-7.95213953e-02 6.15930974e-01 1.07782677e-01 -6.77295208e-01
4.28674579e-01 4.04296339e-01 -2.20787749e-01 -2.50813842e-01
1.63441777e-01 1.09181368e+00 -1.36024678e+00 -6.92032576e-01
-1.98922470e-01 1.36047050e-01 -3.29765350e-01 1.60157159e-01
-7.05868080e-02 -4.01815563e-01 -6.52793229e-01 6.12000287e-01
3.86112750e-01 -1.76513866e-01 4.47971374e-01 -2.64086992e-01
-1.99883178e-01 7.92790711e-01 -1.39990556e+00 1.46680284e+00
-6.13299668e-01 4.40135866e-01 1.70458779e-01 -6.79546475e-01
6.37133420e-01 5.05015433e-01 2.10644484e-01 -7.62456357e-01
-1.14724882e-01 4.26402777e-01 -6.38616383e-01 -4.74491954e-01
1.41235316e+00 3.26652408e-01 -8.65399241e-01 -9.75607485e-02
7.75035620e-01 2.91313201e-01 7.43889391e-01 6.04739308e-01
1.11677074e+00 -1.48142129e-01 8.70967984e-01 -3.00448269e-01
3.50159913e-01 1.71362326e-01 8.58363450e-01 5.76729178e-01
4.16696742e-02 -1.14941448e-02 3.36488873e-01 -1.34184331e-01
-4.08626318e-01 -9.23864603e-01 -1.00725822e-01 1.10455346e+00
4.10240173e-01 -1.11004412e+00 -4.56425011e-01 -1.10553181e+00
2.71849811e-01 1.09842527e+00 -7.77669072e-01 1.59104913e-01
-6.23193502e-01 -5.30233562e-01 6.86881185e-01 5.84779263e-01
5.41670740e-01 -1.20108485e+00 -6.83991536e-02 4.91994172e-01
-6.07174575e-01 -1.68605375e+00 -5.43096840e-01 2.67926663e-01
3.34051102e-02 -1.14100742e+00 -3.29060555e-01 -1.08929467e+00
3.86234254e-01 -8.78986493e-02 1.73307657e+00 2.78535075e-02
-2.10198045e-01 1.26551643e-01 -4.82926607e-01 -4.74044800e-01
1.34903044e-01 6.12750769e-01 1.25054494e-01 -6.63428724e-01
7.55945444e-01 -4.48391028e-02 -1.65572420e-01 2.35980853e-01
-6.46295547e-01 -4.34830040e-01 1.34523273e-01 7.51314938e-01
6.88851655e-01 -1.04962952e-01 8.16776037e-01 -1.41539228e+00
6.07158899e-01 -8.84144247e-01 -3.74933422e-01 7.00021386e-01
-5.21055877e-01 2.16043234e-01 4.87685233e-01 -7.37116113e-02
-1.44600654e+00 -3.17016721e-01 -2.56932735e-01 5.61858714e-02
-4.07727584e-02 1.09297621e+00 -4.27341372e-01 5.51759660e-01
6.28738225e-01 -3.44283164e-01 -1.43846118e+00 -7.39739716e-01
9.59520638e-01 6.33684635e-01 1.02693665e+00 -1.04556119e+00
6.50976360e-01 -2.06987426e-01 -6.10162616e-01 -5.60816348e-01
-1.28335810e+00 -8.28164399e-01 -9.13421154e-01 2.83375591e-01
9.71468449e-01 -1.38771868e+00 -2.58809268e-01 2.93064207e-01
-1.51713109e+00 -5.34345284e-02 -2.82817304e-01 2.68795162e-01
3.80478382e-01 6.64683580e-02 -8.06460202e-01 -8.12091678e-02
-4.92707759e-01 -4.48225409e-01 1.00561798e+00 4.22230989e-01
-4.70590979e-01 -1.27847958e+00 3.77527684e-01 -9.84599218e-02
3.04790199e-01 2.66016811e-01 8.37368488e-01 -1.35625088e+00
-3.25890422e-01 -2.30480328e-01 -2.74156332e-01 -8.81492794e-02
2.39830047e-01 1.22049235e-01 -4.06928688e-01 2.42795855e-01
-1.03224993e+00 -3.57179552e-01 1.04359162e+00 -1.33985683e-01
7.25436687e-01 -5.67062974e-01 -8.70594501e-01 3.98079246e-01
1.43579876e+00 1.89257544e-02 4.36685175e-01 8.66463184e-01
7.02804685e-01 3.33525747e-01 8.57392073e-01 3.04145873e-01
1.19749832e+00 4.53548193e-01 1.82962209e-01 3.76387127e-02
-2.06582829e-01 -3.98315221e-01 -1.72616974e-01 9.13314700e-01
1.70724541e-02 -8.33691895e-01 -1.16718984e+00 1.39940739e+00
-1.90402234e+00 -1.05941689e+00 -1.54574573e-01 1.64804459e+00
1.35680139e+00 -1.75855532e-02 -3.49952966e-01 -8.63924503e-01
8.48933280e-01 2.38912687e-01 -3.37330133e-01 6.13955744e-02
-5.83479583e-01 4.64232206e-01 5.97527683e-01 6.63775027e-01
-1.38019276e+00 1.69666588e+00 6.21430922e+00 8.37064326e-01
-6.24082327e-01 1.26174688e-01 1.93766087e-01 1.19249113e-01
-2.53739804e-01 -8.01693723e-02 -1.43577671e+00 3.14887792e-01
8.40343058e-01 -8.03782880e-01 4.08008993e-02 9.83761847e-01
-6.39106274e-01 8.82875472e-02 -7.22438633e-01 6.41359568e-01
-1.28380388e-01 -1.29359722e+00 -6.72259107e-02 -3.30345958e-01
8.99981081e-01 4.74917412e-01 -3.93473417e-01 1.06877971e+00
1.18368506e+00 -8.13437343e-01 7.18090594e-01 3.07670921e-01
8.97018194e-01 -6.81736469e-01 1.08755994e+00 -4.84909688e-05
-1.68638039e+00 3.36236238e-01 -3.67256999e-01 4.72934961e-01
3.47990274e-01 5.60724258e-01 -7.22353339e-01 9.84861970e-01
8.06614637e-01 9.67621386e-01 -6.31516457e-01 8.41675043e-01
-9.41581905e-01 4.73300308e-01 -2.19654381e-01 -6.07640445e-02
3.10131878e-01 3.44611913e-01 6.42965376e-01 1.95103502e+00
1.85615897e-01 4.51930404e-01 2.20266968e-01 6.55870318e-01
-9.43786502e-01 2.18088925e-01 -1.75222382e-01 -1.54858857e-01
1.32233405e+00 1.48088324e+00 -1.97174922e-01 -6.80870414e-01
-5.01962304e-01 9.61992502e-01 1.24363029e+00 3.80300432e-01
-6.14009023e-01 -1.23086011e+00 9.73161221e-01 -5.23911059e-01
5.06288886e-01 -2.57895648e-01 2.65509486e-01 -1.52937174e+00
-1.90002277e-01 -3.42071116e-01 8.98781419e-01 -7.07438171e-01
-1.55265796e+00 8.55742872e-01 1.01643905e-01 -6.92944586e-01
-4.94173586e-01 -5.15441358e-01 -6.03687048e-01 8.93046081e-01
-1.95569217e+00 -1.01602292e+00 -1.46402270e-01 2.74165511e-01
1.77460641e-01 -3.76427025e-02 9.77330565e-01 7.02963769e-01
-6.91007376e-01 7.05546677e-01 5.35589308e-02 9.81903195e-01
1.20819795e+00 -1.76018178e+00 9.65776920e-01 9.34346735e-01
1.81341752e-01 9.11804676e-01 6.56158984e-01 -9.41676676e-01
-6.93288326e-01 -1.51886034e+00 1.87062895e+00 -7.51114607e-01
1.07064152e+00 -1.52975529e-01 -8.91753376e-01 1.10328102e+00
6.07755065e-01 2.63072044e-01 5.39527714e-01 7.58574128e-01
-5.73360085e-01 4.39899147e-01 -1.11417019e+00 3.72023106e-01
1.28479075e+00 -5.88078380e-01 -1.34725320e+00 9.85753238e-02
1.12817025e+00 -9.06033039e-01 -1.24366987e+00 1.59356296e-01
-9.02511030e-02 2.40249172e-01 1.09705055e+00 -1.07778430e+00
2.44542643e-01 -3.35747033e-01 -2.35384136e-01 -2.04057908e+00
-3.10424984e-01 -2.99639165e-01 -5.18740833e-01 1.91137218e+00
1.09251273e+00 -5.50504386e-01 6.23090230e-02 8.65420520e-01
-4.40725416e-01 -2.69249529e-01 -9.33716536e-01 -1.00743365e+00
7.68033117e-02 -5.30041754e-02 9.49562609e-01 1.47861433e+00
4.40458268e-01 5.98429024e-01 1.51707098e-01 4.67099011e-01
2.99888074e-01 6.05382994e-02 3.03666800e-01 -1.28701627e+00
1.93948999e-01 -1.51046023e-01 -3.56391072e-02 -9.33001041e-01
9.31836426e-01 -9.93847132e-01 1.80976868e-01 -2.06984401e+00
3.88496439e-03 -6.45631015e-01 -5.25305569e-01 1.08492374e+00
-8.51793528e-01 -1.69721514e-01 4.53222506e-02 -6.44975603e-02
-1.36405587e+00 5.06968379e-01 6.61752880e-01 -1.38784096e-01
-3.86815593e-02 -7.96975613e-01 -8.55284929e-01 4.74431634e-01
2.02300549e-01 -6.47246778e-01 2.17874393e-01 -9.62027490e-01
4.88117516e-01 -1.51654616e-01 -5.98240308e-02 -4.99283046e-01
6.40987217e-01 -3.15833569e-01 1.56570580e-02 -5.54139972e-01
3.93940210e-02 -4.60884213e-01 -2.75447190e-01 -1.31799057e-01
-4.80152816e-01 2.06069559e-01 3.01236302e-01 3.74974519e-01
-6.62163138e-01 -3.49966198e-01 1.55091301e-01 -1.68587118e-01
-1.52741206e+00 4.03196901e-01 -1.67863220e-01 1.02843797e+00
5.77792346e-01 7.22928941e-01 -1.00297678e+00 1.40563920e-01
-8.22191596e-01 1.06892228e+00 1.98678821e-01 7.39481926e-01
-1.74174935e-03 -1.52808285e+00 -7.98993826e-01 -6.32191479e-01
6.26529217e-01 2.78176337e-01 1.54545352e-01 4.27122712e-02
-9.86800715e-02 7.96299756e-01 9.96873602e-02 2.20418006e-01
-1.11730349e+00 5.11066258e-01 3.51000160e-01 -7.11873174e-01
-2.63404101e-01 9.46978211e-01 -1.72002345e-01 -9.04066443e-01
4.38500866e-02 -4.31659341e-01 -7.43268907e-01 3.31390947e-01
6.94781244e-01 1.65598899e-01 3.09999704e-01 -6.57264650e-01
-7.14195430e-01 1.42559037e-01 -2.81186372e-01 2.86772028e-02
1.47261679e+00 -2.79898822e-01 -2.71085858e-01 1.34715671e-02
1.05102921e+00 4.87932682e-01 -3.61985415e-01 -8.07837605e-01
7.36985803e-01 -3.65867205e-02 -3.51016261e-02 -1.15758109e+00
-9.78362620e-01 2.53935277e-01 -3.06604028e-01 1.70236096e-01
6.95458293e-01 3.92501026e-01 9.89661694e-01 8.92031133e-01
5.99700332e-01 -1.38234258e+00 -5.67659378e-01 1.20227730e+00
7.14886069e-01 -9.31315899e-01 -2.48330802e-01 -5.39913714e-01
-8.05722952e-01 6.98473215e-01 1.01253259e+00 -5.07501177e-02
6.61698699e-01 2.87156284e-01 -1.39936566e-01 -4.32970196e-01
-9.39157367e-01 -7.59993732e-01 6.12061858e-01 5.66783130e-01
5.93559325e-01 2.06408307e-01 -4.58027482e-01 1.41947448e+00
-1.33650407e-01 -3.30782741e-01 3.48052859e-01 8.17817092e-01
-5.10601878e-01 -1.16239357e+00 4.17426795e-01 7.19690099e-02
-6.09809697e-01 -6.64775014e-01 -5.94181895e-01 1.26114726e+00
2.22696945e-01 7.00492084e-01 3.70328352e-02 -1.00323185e-01
7.26690650e-01 9.37547535e-02 -1.46792799e-01 -1.04024494e+00
-6.34748757e-01 -6.94296122e-01 8.51618588e-01 -3.56001914e-01
-3.81173104e-01 -6.57327414e-01 -1.89206314e+00 -1.73756242e-01
-3.10012162e-01 7.47422576e-01 3.93823296e-01 1.06111479e+00
7.58828461e-01 6.94867611e-01 1.81183890e-01 -1.31079391e-01
9.83354747e-02 -9.86949861e-01 -4.77972239e-01 6.07592881e-01
1.80269718e-01 -8.01127255e-01 -2.10777536e-01 -2.45424151e-01] | [9.502453804016113, 8.966890335083008] |
b95b9c51-2d64-480f-a168-c18f07b29801 | optical-remote-sensing-image-understanding | 2204.09120 | null | https://arxiv.org/abs/2204.09120v1 | https://arxiv.org/pdf/2204.09120v1.pdf | Optical Remote Sensing Image Understanding with Weak Supervision: Concepts, Methods, and Perspectives | In recent years, supervised learning has been widely used in various tasks of optical remote sensing image understanding, including remote sensing image classification, pixel-wise segmentation, change detection, and object detection. The methods based on supervised learning need a large amount of high-quality training data and their performance highly depends on the quality of the labels. However, in practical remote sensing applications, it is often expensive and time-consuming to obtain large-scale data sets with high-quality labels, which leads to a lack of sufficient supervised information. In some cases, only coarse-grained labels can be obtained, resulting in the lack of exact supervision. In addition, the supervised information obtained manually may be wrong, resulting in a lack of accurate supervision. Therefore, remote sensing image understanding often faces the problems of incomplete, inexact, and inaccurate supervised information, which will affect the breadth and depth of remote sensing applications. In order to solve the above-mentioned problems, researchers have explored various tasks in remote sensing image understanding under weak supervision. This paper summarizes the research progress of weakly supervised learning in the field of remote sensing, including three typical weakly supervised paradigms: 1) Incomplete supervision, where only a subset of training data is labeled; 2) Inexact supervision, where only coarse-grained labels of training data are given; 3) Inaccurate supervision, where the labels given are not always true on the ground. | ['Antonio J Plaza', 'Jocelyn Chanussot', 'Jun Li', 'Weiying Xie', 'Pedram Ghamisi', 'Leyuan Fang', 'Jun Yue'] | 2022-04-18 | null | null | null | null | ['remote-sensing-image-classification'] | ['miscellaneous'] | [ 6.27971828e-01 -3.46630096e-01 -3.41305345e-01 -4.92787093e-01
-5.34971297e-01 -5.66647828e-01 2.87012815e-01 2.89182007e-01
-4.01405841e-01 9.54676270e-01 -2.79625833e-01 -3.73031169e-01
-1.98966518e-01 -1.06707478e+00 -4.87921327e-01 -1.02535248e+00
3.45517576e-01 2.60033935e-01 1.67455837e-01 -6.95599765e-02
-4.97967936e-02 3.29259604e-01 -1.67997181e+00 -7.32443482e-02
1.13908148e+00 9.15041804e-01 4.51790422e-01 2.57903248e-01
-3.23975205e-01 5.39350510e-01 -3.14890146e-01 4.22666669e-01
2.92471975e-01 -4.42221820e-01 -7.97497213e-01 5.29835820e-01
8.53139013e-02 -3.33371371e-01 2.27461055e-01 1.50292277e+00
1.61317527e-01 -1.86722472e-01 5.11178315e-01 -8.90888274e-01
-3.57215077e-01 4.45403934e-01 -8.52187753e-01 1.55563727e-02
-2.32097745e-01 -5.47398478e-02 1.04792726e+00 -7.47004509e-01
2.21845120e-01 8.37029397e-01 4.40400809e-01 2.86249697e-01
-9.42785323e-01 -6.67301714e-01 5.32619417e-01 1.15001395e-01
-1.43423820e+00 -1.29083246e-01 7.71196783e-01 -5.45951426e-01
2.81469971e-01 2.68214762e-01 5.83019197e-01 3.07881832e-01
-5.06882131e-01 6.37122214e-01 1.54619706e+00 -4.19858754e-01
3.12924594e-01 1.92675978e-01 2.34746069e-01 3.59602392e-01
1.71436027e-01 3.27047974e-01 -1.17417373e-01 5.09771556e-02
8.26383352e-01 5.10748327e-01 -4.04087663e-01 1.36781588e-01
-1.14777005e+00 7.21606493e-01 7.85862803e-01 5.74912250e-01
-3.46332312e-01 -3.83782208e-01 8.16576555e-02 3.07660073e-01
8.30497742e-01 2.88745970e-01 -5.84269702e-01 2.75356412e-01
-1.06440139e+00 -1.47403389e-01 3.62503827e-01 6.33247972e-01
1.44745111e+00 -3.71611677e-02 5.27086377e-01 8.43348920e-01
1.78253636e-01 7.90848255e-01 1.93735331e-01 -5.97533286e-01
4.48986053e-01 7.90412724e-01 4.27895367e-01 -9.99848366e-01
-3.59274626e-01 -4.67356831e-01 -1.13765192e+00 2.98445255e-01
4.52942282e-01 -1.93751737e-01 -1.18857384e+00 1.43582165e+00
3.97271067e-01 5.82551248e-02 1.28066202e-03 1.22015297e+00
8.31283629e-01 9.98395085e-01 1.67937372e-02 -4.46499527e-01
8.26510370e-01 -8.99473906e-01 -6.17920995e-01 -4.28442538e-01
4.76858169e-01 -5.90767324e-01 1.22892380e+00 1.40307084e-01
-3.06698203e-01 -6.49726748e-01 -9.85851049e-01 3.37083519e-01
-4.76848990e-01 2.37169698e-01 6.29110754e-01 3.54561448e-01
-5.79527318e-01 4.39601272e-01 -6.36920094e-01 -3.17398101e-01
4.14679050e-01 1.18883058e-01 -3.29368383e-01 -4.29493487e-01
-1.13274801e+00 5.84706783e-01 4.94217575e-01 7.32890069e-01
-8.47722352e-01 -4.24977481e-01 -6.82885051e-01 -1.03460439e-01
7.11657345e-01 -2.14905869e-02 9.17237520e-01 -1.32995689e+00
-9.62856293e-01 9.50096726e-01 -1.16874747e-01 2.45460376e-01
4.86322969e-01 -1.41104877e-01 -3.97657931e-01 -6.65947571e-02
3.35747123e-01 3.78629446e-01 8.01249206e-01 -1.50042355e+00
-8.26667368e-01 -7.61184335e-01 1.27106816e-01 3.68531078e-01
-8.38511735e-02 -2.27623343e-01 5.84542267e-02 -3.97697031e-01
7.06057131e-01 -7.44287372e-01 -3.23425174e-01 1.09828271e-01
-3.81270230e-01 -1.18987195e-01 1.14968204e+00 -5.46589911e-01
7.33458698e-01 -2.41670251e+00 -1.27713621e-01 1.13070369e-01
1.47264287e-01 3.77031356e-01 2.14098282e-02 7.86436871e-02
-4.69577499e-02 3.66362542e-01 -7.02122390e-01 3.25147420e-01
-4.96799886e-01 5.73695064e-01 -3.37132871e-01 4.24315065e-01
2.01142445e-01 5.78215361e-01 -1.17763078e+00 -4.93561119e-01
3.05012912e-01 1.21676438e-01 1.57267228e-01 4.20743257e-01
-4.58313853e-01 9.60162282e-01 -8.64466250e-01 8.98752689e-01
6.88519597e-01 -4.73445028e-01 -1.94455355e-01 -7.74055123e-02
-3.20032358e-01 -5.55043519e-02 -1.39127970e+00 1.30026221e+00
-4.35400844e-01 3.30870509e-01 3.40708733e-01 -1.30177295e+00
9.78137553e-01 2.76460111e-01 5.33274710e-01 -5.76168001e-01
-1.81094542e-01 5.27446151e-01 -3.71646285e-01 -6.80869520e-01
2.18778476e-01 -6.53230488e-01 2.24599361e-01 5.86782098e-01
-6.90445304e-01 -2.44523481e-01 -9.07994583e-02 -1.76021397e-01
5.00157833e-01 1.36094257e-01 2.67738491e-01 1.58173442e-02
4.86213654e-01 4.91542816e-01 8.20503414e-01 6.31310761e-01
-2.36460567e-01 6.47087574e-01 -7.15469345e-02 -5.95856547e-01
-8.83037567e-01 -6.33574009e-01 -4.00520861e-01 1.14269459e+00
4.69077051e-01 2.59264082e-01 -4.65989530e-01 -6.73196256e-01
-1.97769240e-01 7.12502301e-02 -3.71789962e-01 2.10114837e-01
-2.47055054e-01 -1.04675889e+00 2.58234948e-01 4.49254155e-01
1.04488182e+00 -1.17724502e+00 -3.11649203e-01 2.30749488e-01
-3.78920227e-01 -1.07348680e+00 -5.70816100e-02 1.74295858e-01
-1.16951334e+00 -1.16496396e+00 -7.80331790e-01 -6.17718756e-01
1.04518354e+00 8.43370736e-01 8.39201927e-01 5.13541758e-01
-1.27380004e-03 -3.57403368e-01 -6.02546573e-01 -4.30968583e-01
-1.48749217e-01 -1.09301209e-02 -1.46548480e-01 6.85744360e-02
4.11062479e-01 -4.18824315e-01 -3.61185730e-01 6.53880298e-01
-1.15043151e+00 -2.41246843e-03 7.35122740e-01 9.75549221e-01
8.68682444e-01 7.64783561e-01 7.24228442e-01 -1.21665919e+00
-1.96206257e-01 -2.75871694e-01 -7.43525267e-01 3.25160891e-01
-6.43154025e-01 -2.85307914e-01 6.62124991e-01 -2.33662367e-01
-1.13121903e+00 1.27842888e-01 -7.10626617e-02 1.88521445e-02
-5.40060043e-01 9.80358422e-01 -4.19257969e-01 -3.66047248e-02
7.23948717e-01 2.83074379e-01 -1.35674194e-01 -5.50856590e-01
9.39693376e-02 1.16137934e+00 2.66600192e-01 -4.10202980e-01
7.79830217e-01 6.88414454e-01 -1.86254382e-01 -9.41517115e-01
-1.53189397e+00 -8.43159318e-01 -7.79177368e-01 -1.37013584e-01
6.86931551e-01 -9.57793534e-01 2.65382320e-01 9.15532768e-01
-6.08116210e-01 -3.53467792e-01 -2.95222104e-01 6.12569034e-01
-4.45603244e-02 3.33608836e-01 -3.30227405e-01 -8.16127360e-01
-2.02583537e-01 -1.02568793e+00 1.05928290e+00 3.23227286e-01
3.51028740e-01 -8.89839351e-01 -2.92608380e-01 5.77214122e-01
2.97553003e-01 1.96106777e-01 9.06003177e-01 -1.35284260e-01
-6.11491442e-01 -8.46732706e-02 -6.65150046e-01 5.84585547e-01
6.99339926e-01 2.26477087e-02 -9.60098207e-01 -1.42156661e-01
1.49850035e-02 -6.84899271e-01 8.47605705e-01 4.49382544e-01
9.35181975e-01 -9.68936309e-02 -2.11576253e-01 4.41874504e-01
1.54490387e+00 -5.93544841e-02 2.53527492e-01 2.38838062e-01
9.82267022e-01 9.10790682e-01 1.09385514e+00 2.27134973e-01
2.44709000e-01 2.38077238e-01 5.23682296e-01 -5.73531747e-01
1.93355247e-01 -3.43188405e-01 -1.15677655e-01 5.98211527e-01
-2.42951766e-01 1.87771395e-01 -9.85449553e-01 6.55361831e-01
-1.89069104e+00 -7.83658743e-01 -3.43473285e-01 2.24122143e+00
1.05789614e+00 -1.11807227e-01 -8.80437866e-02 5.97523868e-01
9.73563313e-01 3.88404042e-01 -8.98216903e-01 4.70426023e-01
-1.55380622e-01 -2.63273299e-01 4.76850569e-01 4.87979561e-01
-1.26147258e+00 1.00880408e+00 5.75338650e+00 6.20774508e-01
-1.46216857e+00 1.96116850e-01 5.98478615e-01 4.24511254e-01
-3.16013545e-01 2.38426238e-01 -7.73868680e-01 3.69619340e-01
2.16276199e-01 3.42904747e-01 1.62366793e-01 8.08510780e-01
5.57692111e-01 -5.10174870e-01 -8.60594809e-01 9.10261393e-01
-2.66605526e-01 -8.80410075e-01 -4.49167825e-02 1.30516291e-02
1.10199690e+00 2.18058929e-01 -3.73423189e-01 -8.50416571e-02
2.60081023e-01 -1.09076488e+00 4.41516548e-01 2.05559313e-01
9.39777613e-01 -4.75662827e-01 9.21359122e-01 1.02259815e+00
-1.17825377e+00 2.92131416e-02 -5.86850941e-01 -3.91401500e-01
1.60636753e-01 1.34305751e+00 -2.82310784e-01 4.61875081e-01
7.69130945e-01 1.14493096e+00 -1.69924721e-01 8.15631628e-01
-6.47566378e-01 7.89898872e-01 -5.91858149e-01 2.18908370e-01
3.44434083e-01 -4.53010112e-01 1.79973826e-01 6.19037926e-01
5.95112592e-02 6.06335163e-01 8.22321773e-01 6.95369661e-01
1.47892758e-02 1.03024349e-01 -5.05593598e-01 -2.40031723e-02
3.82323354e-01 1.11810672e+00 -7.67042875e-01 -2.48859495e-01
-4.20588523e-01 5.21890759e-01 8.45034793e-03 5.56819141e-01
-2.82463372e-01 8.74948576e-02 1.65023908e-01 1.53977588e-01
-1.42990112e-01 -2.29740560e-01 -5.29390156e-01 -1.30944192e+00
1.41593561e-01 -7.56051362e-01 4.87103879e-01 -7.86207974e-01
-1.38682067e+00 3.44843537e-01 -1.39854133e-01 -1.30138791e+00
9.90151167e-02 -3.33051473e-01 -3.46158534e-01 9.41204250e-01
-2.18525362e+00 -1.12277615e+00 -8.12966466e-01 4.81374711e-01
4.03445542e-01 1.97647884e-01 6.29367709e-01 3.38999182e-01
-4.66146439e-01 -1.50161520e-01 1.87279537e-01 2.52772003e-01
5.71583331e-01 -1.07910824e+00 -3.20261568e-01 8.26920688e-01
1.14785582e-01 1.23998582e-01 5.03841519e-01 -4.96449441e-01
-8.05743992e-01 -1.33901298e+00 7.25185513e-01 1.49376303e-01
4.76248741e-01 1.93046212e-01 -1.39193022e+00 4.80931073e-01
-6.14561915e-01 4.60077256e-01 7.37547278e-01 1.20396733e-01
-2.89016157e-01 -4.75862652e-01 -1.12075377e+00 1.09401889e-01
7.61281490e-01 -7.01233685e-01 -4.78634536e-01 6.52226567e-01
2.54675537e-01 -1.81263834e-01 -6.69610143e-01 5.44843376e-01
1.50320768e-01 -7.60573089e-01 4.85882849e-01 -3.15069199e-01
4.16723281e-01 -8.09644759e-01 -4.23331261e-02 -1.13984072e+00
-1.58046082e-01 2.41252989e-01 6.12591267e-01 1.01068783e+00
5.34747005e-01 -5.83029628e-01 8.38713884e-01 4.17755753e-01
9.85268578e-02 -5.08779228e-01 -5.50864577e-01 -7.03795612e-01
9.89915896e-03 -2.91372359e-01 5.98820150e-01 1.42169321e+00
-3.99428725e-01 3.88855726e-01 -4.02843803e-01 5.99162042e-01
6.39230907e-01 7.66670108e-01 5.17505348e-01 -1.52877414e+00
-1.17371775e-01 -1.20369434e-01 -1.69075236e-01 -1.18440974e+00
5.21405041e-02 -4.19963181e-01 4.66782212e-01 -1.73372328e+00
3.75369370e-01 -1.09446430e+00 -1.74565986e-02 8.93489659e-01
-4.84488219e-01 3.52853715e-01 -2.35788807e-01 8.46572220e-01
-3.31780851e-01 4.77671027e-01 1.34110045e+00 -4.41440701e-01
-3.40184480e-01 2.80647427e-01 -5.69884717e-01 8.75648499e-01
7.86556065e-01 -5.28115392e-01 -4.51405853e-01 -6.70178175e-01
3.40260476e-01 -1.37591273e-01 4.03891087e-01 -6.05171621e-01
6.30314797e-02 -7.53128946e-01 7.32170716e-02 -6.05681896e-01
-1.51366159e-01 -1.08289731e+00 1.59100816e-02 2.77505189e-01
-4.59072702e-02 -5.75020254e-01 -3.75583589e-01 5.60912788e-01
-7.31652200e-01 -4.03288573e-01 9.35932100e-01 -5.40290475e-01
-1.11547267e+00 5.32867968e-01 1.11408383e-02 -8.97824317e-02
9.20478821e-01 -2.46784002e-01 -1.14656225e-01 -2.96923816e-01
-8.54267061e-01 3.63735646e-01 5.50880253e-01 1.36236683e-01
3.56408507e-01 -9.65942800e-01 -9.69181061e-01 1.85199261e-01
4.55407023e-01 7.43508995e-01 2.36198321e-01 5.46285212e-01
-3.50230247e-01 1.82879530e-02 1.34368360e-01 -9.03427482e-01
-1.05661309e+00 3.00687134e-01 4.45731431e-01 -1.37491584e-01
-4.73850131e-01 5.33066094e-01 7.56119937e-02 -6.13020539e-01
-5.41055249e-03 -1.05879478e-01 -3.57089967e-01 4.78437580e-02
5.80271959e-01 1.82561114e-01 7.48261139e-02 -7.01238394e-01
-1.94464788e-01 9.36208487e-01 1.65172234e-01 1.44086957e-01
1.31215203e+00 -3.30065459e-01 -4.04768258e-01 6.83147311e-01
7.82197893e-01 -2.82939762e-01 -1.35073209e+00 -7.67227173e-01
-1.71694174e-01 -6.03807867e-01 4.26285863e-01 -7.75452316e-01
-1.28419852e+00 1.21165454e+00 4.74031836e-01 2.97463834e-01
1.37449872e+00 6.56286031e-02 5.77326894e-01 6.49569154e-01
7.31622994e-01 -1.14189434e+00 -2.89956927e-01 4.33010876e-01
4.06289816e-01 -1.89002657e+00 1.88804880e-01 -6.91971123e-01
-3.84073108e-01 8.55736613e-01 5.16090810e-01 3.88921231e-01
9.19805646e-01 2.11740620e-02 6.67289555e-01 -1.48211792e-01
-3.60803008e-02 -5.50259829e-01 1.35313228e-01 4.83285218e-01
3.49913657e-01 2.34934255e-01 -2.82912731e-01 2.05722421e-01
1.32145524e-01 2.58698408e-02 2.39380866e-01 1.03760159e+00
-8.06154490e-01 -1.00595140e+00 -5.51165879e-01 5.45798302e-01
-3.53321105e-01 -3.75967398e-02 -3.28693062e-01 4.17056412e-01
3.40058416e-01 1.22392952e+00 -2.63594776e-01 -6.02330938e-02
1.50187671e-01 -4.28247422e-01 -7.10006207e-02 -1.01071393e+00
-1.12917982e-01 7.54320174e-02 -6.47542402e-02 -9.10405517e-02
-1.14478648e+00 -4.45022732e-01 -1.43950224e+00 -1.91968054e-01
-7.54218400e-01 3.92226189e-01 5.96418023e-01 1.51589453e+00
-1.15241811e-01 -4.16223193e-03 9.28943574e-01 -6.12954617e-01
-5.56420088e-01 -1.03187013e+00 -1.07197309e+00 4.53525007e-01
5.18678904e-01 -5.80520689e-01 -5.79278171e-01 1.89834699e-01] | [9.70629596710205, -1.3549489974975586] |
7922dc80-a507-45f7-80e0-72c40744d4dc | viscoelastic-constitutive-artificial-neural | 2303.12164 | null | https://arxiv.org/abs/2303.12164v1 | https://arxiv.org/pdf/2303.12164v1.pdf | Viscoelastic Constitutive Artificial Neural Networks (vCANNs) $-$ a framework for data-driven anisotropic nonlinear finite viscoelasticity | The constitutive behavior of polymeric materials is often modeled by finite linear viscoelastic (FLV) or quasi-linear viscoelastic (QLV) models. These popular models are simplifications that typically cannot accurately capture the nonlinear viscoelastic behavior of materials. For example, the success of attempts to capture strain rate-dependent behavior has been limited so far. To overcome this problem, we introduce viscoelastic Constitutive Artificial Neural Networks (vCANNs), a novel physics-informed machine learning framework for anisotropic nonlinear viscoelasticity at finite strains. vCANNs rely on the concept of generalized Maxwell models enhanced with nonlinear strain (rate)-dependent properties represented by neural networks. The flexibility of vCANNs enables them to automatically identify accurate and sparse constitutive models of a broad range of materials. To test vCANNs, we trained them on stress-strain data from Polyvinyl Butyral, the electro-active polymers VHB 4910 and 4905, and a biological tissue, the rectus abdominis muscle. Different loading conditions were considered, including relaxation tests, cyclic tension-compression tests, and blast loads. We demonstrate that vCANNs can learn to capture the behavior of all these materials accurately and computationally efficiently without human guidance. | ['Christian J. Cyron', 'Kevin Linka', 'Kian P. Abdolazizi'] | 2023-03-21 | null | null | null | null | ['physics-informed-machine-learning'] | ['graphs'] | [ 2.6737475e-01 -1.0445893e-01 -2.7918616e-01 4.2646300e-02
6.7631342e-02 -3.1133074e-01 -4.1206971e-02 5.8154162e-02
-3.5136321e-01 9.2304301e-01 -1.6401831e-02 -1.8346900e-01
-6.1424673e-01 -5.9273547e-01 -1.0025958e+00 -8.8603222e-01
-5.3274876e-01 6.0027045e-01 3.6253995e-01 -1.3576323e-01
1.4195972e-02 7.0161664e-01 -1.0385368e+00 6.5574400e-02
8.9707851e-01 1.1132566e+00 1.0584300e-01 8.2947582e-01
3.6160240e-01 7.1965730e-01 -8.0303252e-02 1.9816998e-03
-1.7134918e-02 2.4813433e-01 -7.9327381e-01 -3.4453660e-01
2.5367898e-01 -6.8332803e-01 -6.7509925e-01 6.2581605e-01
4.5070282e-01 2.0386708e-01 1.0789510e+00 -7.0531446e-01
-9.1710162e-01 8.7856120e-01 -4.5112786e-01 -7.7058405e-02
1.3886979e-01 -1.2819983e-01 6.5106505e-01 -5.2263528e-01
7.3953229e-01 9.9751526e-01 1.2470415e+00 8.2781214e-01
-1.4389145e+00 -1.5398425e-01 -3.0562896e-01 -1.6382691e-01
-5.6433833e-01 3.8570192e-02 1.3398279e+00 -6.2157977e-01
7.2370404e-01 2.7510160e-01 8.4287447e-01 1.4221891e+00
1.1421319e+00 6.3569331e-01 1.1923869e+00 -9.7152241e-02
3.1810224e-01 -2.5599417e-01 3.9894527e-01 4.7215265e-01
7.3038244e-01 3.5394752e-01 -2.2741473e-01 -4.5504689e-01
1.0720572e+00 -1.7563103e-01 -1.8882184e-01 -2.8307694e-01
-7.1936572e-01 4.2706889e-01 1.4564338e-01 4.4476274e-01
-3.5905999e-01 4.9488333e-01 6.7384779e-01 1.8544482e-01
6.0416013e-01 6.9513577e-01 -7.4502748e-01 -1.9659972e-01
-8.3996987e-01 4.4847235e-01 1.1881025e+00 4.6077332e-01
2.2562478e-01 7.1833605e-01 5.2547801e-01 8.6901593e-01
3.5358328e-01 5.1994675e-01 6.7818391e-01 -1.2315364e+00
-1.2495921e-01 1.0807035e-01 1.9286099e-01 -1.1405739e+00
-8.3006871e-01 -4.4617528e-01 -1.1150815e+00 2.6100550e-03
3.0114001e-01 -2.4624114e-01 -9.3254280e-01 1.3266323e+00
1.5004599e-01 -7.6106772e-02 2.5993595e-01 9.5424354e-01
9.9811268e-01 5.3363037e-01 2.5876191e-01 -3.4722975e-01
8.2057446e-01 -4.4885984e-01 -6.7627543e-01 -1.6767006e-01
5.6031811e-01 -4.7463378e-01 7.9551065e-01 4.3719804e-01
-1.5319023e+00 -5.2751178e-01 -1.1083263e+00 5.4809276e-02
-1.7795803e-01 -1.5219136e-01 8.7563699e-01 3.1820390e-01
-7.2174931e-01 1.5071433e+00 -1.3354360e+00 2.1816239e-01
4.3778617e-02 4.9280149e-01 -2.5423190e-01 6.0005873e-02
-1.2590532e+00 6.9244945e-01 2.5872967e-01 5.3478444e-01
-4.3773323e-01 -1.1429770e+00 -7.7835202e-01 -2.9055836e-02
-6.6923723e-03 -7.8559572e-01 9.3127650e-01 -6.1396372e-01
-1.7127864e+00 5.3088593e-01 3.9129904e-01 -3.5564956e-01
6.1856270e-01 -3.0621377e-01 -3.1828630e-01 3.4838441e-01
-6.5286583e-01 2.6403192e-02 6.0966933e-01 -1.8565243e+00
8.1629872e-01 1.8138875e-01 -3.3622447e-01 -5.1516587e-01
-4.1099179e-01 -2.1873137e-01 3.9458549e-01 -1.0282465e+00
4.8203221e-01 -1.2488364e+00 -2.1706960e-01 -8.9860551e-02
-5.3344208e-01 1.2614961e-01 8.2093579e-01 -1.0206556e+00
7.1363354e-01 -1.8267580e+00 3.8896596e-01 3.7747803e-01
3.4996343e-01 4.0785766e-01 -2.6691568e-01 7.0862317e-01
4.7131805e-04 -3.3821575e-02 -5.7553595e-01 3.5671756e-02
4.4169676e-02 5.6133634e-01 -1.8695356e-01 3.2303923e-01
7.3801011e-02 1.0018369e+00 -8.4655011e-01 -4.1253996e-01
-2.9328090e-01 6.3417304e-01 -4.8870718e-01 6.0417354e-02
-3.3515602e-01 2.6032126e-01 -3.8534963e-01 9.8258656e-01
9.2532784e-01 -9.0708822e-02 3.9270869e-01 -7.0457244e-01
1.7296563e-01 -4.5919061e-01 -7.0649832e-01 1.1314281e+00
-2.8876168e-01 5.1566106e-01 2.9752243e-01 -1.2719783e+00
9.9132270e-01 3.2858384e-01 1.1577252e+00 -5.4340476e-01
3.1170154e-01 7.7462620e-01 2.2657542e-01 -1.0034046e+00
4.9741849e-01 -4.0115666e-01 9.7886913e-02 3.5866880e-01
1.1614521e-02 -3.1491637e-01 9.6495904e-02 -1.2963031e-01
9.5825994e-01 6.6207457e-01 -2.6985970e-01 -4.0482873e-01
1.5697095e-01 -1.9206467e-01 5.6706518e-01 6.5812403e-01
9.9671043e-02 4.2925543e-01 4.3835109e-01 -7.4660641e-01
-1.2713022e+00 -1.1600301e+00 -3.9418021e-01 7.4469674e-01
1.7055041e-01 2.5884920e-01 -5.2519912e-01 3.0611759e-01
5.4512691e-01 1.3006513e-01 -7.1296507e-01 -4.3217564e-01
-1.1439302e+00 -8.1992096e-01 3.8890037e-01 9.0013576e-01
7.8380771e-02 -1.1148715e+00 -4.6884331e-01 6.2049782e-01
3.3474585e-01 -1.4318166e+00 -8.1580818e-02 2.7692497e-01
-1.4991447e+00 -8.7000668e-01 -7.9174489e-01 -6.4678508e-01
6.0955179e-01 -4.2088550e-01 9.6030396e-01 1.1922338e-01
-4.4858950e-01 5.2890402e-01 -1.9612451e-01 -2.5626054e-01
-1.1021680e+00 -6.2469300e-02 5.5374849e-01 -3.7719768e-01
-4.0940660e-01 -1.1740108e+00 -7.5191158e-01 2.7769747e-01
-1.0988333e+00 -2.2521707e-01 8.8971168e-01 6.4603746e-01
6.6658318e-01 -5.8237676e-02 5.3652227e-01 -9.7079825e-01
7.8992391e-01 -3.6918601e-01 -8.7498702e-02 -1.8464541e-01
-4.9348176e-01 -2.5076550e-02 8.8294506e-01 -1.3877180e+00
-7.5543511e-01 -3.2475802e-01 -1.9700499e-01 -6.0098332e-01
6.1976146e-02 1.1658958e+00 3.2398316e-01 -5.6562269e-01
4.7588518e-01 3.2247162e-01 8.9682192e-02 -4.4792360e-01
-3.8628656e-01 3.3642885e-01 7.8607482e-01 -1.3639208e+00
6.3926500e-01 1.3898893e-01 5.7266545e-01 -1.0857768e+00
-3.9704004e-01 1.7626031e-01 -4.9755052e-01 -5.2288187e-01
2.3640713e-01 -1.7051429e-01 -1.2031717e+00 8.6464143e-01
-9.7949713e-01 -7.6998991e-01 -3.8254279e-01 6.5760952e-01
-8.6608028e-01 7.2334486e-01 -1.4881047e+00 -8.2646316e-01
-6.5657616e-01 -9.3168688e-01 6.7799604e-01 1.6169229e-01
-3.4876198e-01 -1.4485826e+00 1.5336987e-01 2.1447164e-01
8.5873008e-01 1.0320965e+00 1.1673441e+00 -1.2395828e-01
-7.4874833e-02 -5.9221870e-01 4.6662409e-02 6.3365537e-01
-1.1491281e-01 5.0223637e-01 -7.5495738e-01 -4.7489753e-01
-1.4537612e-02 -3.6985654e-01 8.9657199e-01 1.0409119e+00
1.1501993e+00 -4.2063919e-01 -3.4735745e-01 5.2971882e-01
1.6469820e+00 3.2468945e-01 5.4400766e-01 -1.0363365e-01
7.8462243e-01 3.7189972e-01 -1.3747551e-01 3.8384902e-01
-1.6435403e-01 2.4054424e-01 3.5960349e-01 -4.1621223e-02
3.4665871e-02 1.4641698e-01 2.1843985e-01 1.4908851e+00
-9.3535858e-01 -1.6278930e-01 -6.6962206e-01 2.4789016e-01
-1.3272042e+00 -6.8695402e-01 -6.3514106e-02 1.6824203e+00
1.1270785e+00 4.9615094e-01 4.1087531e-02 3.2185301e-01
2.3071298e-01 -1.1027032e-01 -8.8656408e-01 -6.6952264e-01
-2.6540026e-01 2.0381761e-01 3.7450776e-01 3.9442268e-01
-8.8394922e-01 2.4414057e-01 7.1248498e+00 6.1867017e-02
-1.3982905e+00 -2.5060114e-01 3.1839442e-01 2.4329872e-01
-3.4748235e-01 -4.3142813e-01 -2.0118654e-01 7.1420109e-01
1.1188233e+00 1.3582054e-01 2.8377238e-01 6.5983325e-01
1.5378976e-01 -4.9022764e-02 -8.7383413e-01 3.7311998e-01
-3.3110419e-01 -1.4559218e+00 -2.6572782e-01 -2.8476916e-02
6.4235973e-01 -1.4734124e-01 1.1923218e-01 5.6720026e-02
-6.5190122e-02 -7.2966826e-01 5.4893434e-01 1.2859359e+00
6.1020792e-01 -2.3080456e-01 9.3392855e-01 1.0573059e-01
-8.4956038e-01 -1.7278582e-01 -6.8729305e-01 6.6967085e-02
3.0117446e-01 9.2608517e-01 -7.8568511e-02 5.2371579e-01
6.3104820e-01 5.4794246e-01 1.9306542e-01 8.8583791e-01
6.0810488e-01 9.5970243e-01 -4.1390204e-01 -1.9929306e-01
5.3133104e-02 -2.5068304e-01 7.0449358e-01 1.1602992e+00
2.2078097e-01 7.1690962e-02 2.6146695e-02 8.9073271e-01
-9.2219636e-03 -1.5658906e-01 -2.8854489e-01 -4.9290946e-01
1.7765844e-03 9.1881359e-01 -7.1706933e-01 1.0563045e-01
2.3452779e-02 2.7845219e-01 -3.1456355e-02 7.4532706e-01
-7.3629791e-01 -2.9235873e-01 2.1801631e-01 5.2109617e-01
-1.0989353e-01 -9.2070186e-01 -5.4120686e-02 -8.7914556e-01
2.6978668e-02 -3.1320316e-01 -7.4133150e-02 -9.7778761e-01
-1.6289085e+00 4.6188027e-01 2.9736590e-01 -1.1395718e+00
2.2851348e-01 -1.1100389e+00 -4.2112255e-01 6.3755041e-01
-1.3308594e+00 -1.2704172e+00 -3.5769445e-01 -1.7719580e-01
-5.7171836e-02 1.0069981e-01 6.5389824e-01 2.9709542e-01
-1.9344188e-01 2.4030049e-01 3.9278203e-01 1.6226010e-04
5.7579231e-01 -1.1129354e+00 -5.3225186e-02 4.6061341e-02
-9.1169888e-01 8.7508476e-01 1.3760567e+00 -1.0783278e+00
-1.7593011e+00 -5.9492034e-01 1.3892503e-01 2.6552385e-01
9.2841053e-01 -3.2831229e-02 -1.5914170e+00 3.2000202e-01
-2.5456330e-01 3.3310208e-01 8.1105977e-01 -2.5569066e-01
-4.7873072e-02 -6.2611975e-02 -8.5323000e-01 3.6382407e-01
8.9864635e-01 -1.6707318e-01 -2.8156304e-01 4.5149869e-01
6.0341239e-01 -7.0261049e-01 -1.7732502e+00 8.1726909e-01
1.1882907e+00 -5.5388737e-01 1.1947004e+00 -9.7278112e-01
1.0552408e+00 4.4400465e-01 6.5878578e-02 -1.0953331e+00
-4.3815988e-01 -7.1612173e-01 -9.3910623e-01 6.1682719e-01
-5.5971455e-02 -7.9038310e-01 8.8140857e-01 8.1020653e-01
-5.5493051e-01 -1.4031680e+00 -8.5715002e-01 -1.0868980e+00
5.2414280e-01 -1.4672090e-01 1.7676543e-01 8.5288560e-01
-2.0181158e-01 -4.9258468e-01 -4.3094584e-01 -2.3541036e-01
6.5516341e-01 4.5983039e-02 -3.5320256e-02 -1.6637921e+00
-3.1354788e-01 -3.0986962e-01 -3.4755883e-01 -6.1244053e-01
4.3240893e-01 -4.9368066e-01 -1.1519219e-01 -1.4475681e+00
1.7822459e-02 -5.7606339e-01 -2.6774520e-01 1.7092615e-01
1.6289461e-01 9.2306010e-02 -4.0935636e-01 5.0662291e-01
3.5899067e-01 4.8390856e-01 1.8936197e+00 -3.5509074e-01
-3.3234936e-01 1.1861478e-02 -1.7080779e-01 6.7984742e-01
7.4842960e-01 -2.1514983e-01 -1.9946286e-01 -2.7776703e-01
5.4618555e-01 3.9334878e-01 4.0931627e-01 -1.1572914e+00
1.0409765e-01 -3.6340153e-01 5.0409526e-01 -3.0543080e-01
3.9395237e-01 -8.7589347e-01 4.9807471e-01 7.2927880e-01
-3.9104876e-01 -7.8807026e-02 4.2836514e-01 5.1812893e-01
1.0498753e-02 -2.8974739e-01 8.6544901e-01 8.5986443e-02
-1.5148601e-01 3.6445072e-01 -6.3501912e-01 -3.1535041e-01
4.7408357e-01 -6.9949627e-01 -4.7075400e-01 -4.2044576e-02
-1.0688577e+00 -4.2459384e-01 3.7463593e-01 -1.4376585e-01
7.3887044e-01 -1.1119647e+00 -4.3031552e-01 1.4943181e-01
-6.4490026e-01 -1.1718655e-01 7.9244876e-01 1.0253896e+00
-1.0183002e+00 9.2023417e-02 -5.0146037e-01 -4.3537185e-01
-7.6636505e-01 5.1715827e-01 4.9130243e-01 -3.4240961e-01
-4.6983287e-01 5.9996432e-01 -4.8209432e-01 -3.3911860e-01
-3.6589256e-01 -6.5739942e-01 -1.0449143e-01 -4.1011849e-01
-3.2367989e-01 6.0199732e-01 -9.7823717e-02 -4.3701047e-01
-5.0565153e-03 8.9951885e-01 2.4182510e-01 4.7056922e-01
1.7090954e+00 3.8142478e-01 -2.9372022e-01 5.5273235e-01
1.1156039e+00 -1.3383500e-01 -1.3554542e+00 -1.6676147e-01
-3.4150168e-01 1.6087738e-01 -1.2815972e-01 -5.5780149e-01
-1.2474191e+00 5.2560151e-01 1.4145039e-01 6.0435855e-01
9.1150886e-01 -2.2151376e-01 1.3019011e+00 4.2959866e-01
-7.6418832e-02 -1.2963971e+00 3.9321303e-01 4.2035934e-01
1.0580878e+00 -7.1471119e-01 5.2911919e-01 -7.1679384e-01
-4.0126760e-02 1.5895412e+00 5.3890514e-01 -3.7695238e-01
9.2199409e-01 8.8552737e-01 -7.6899357e-02 8.1526637e-02
-3.5970974e-01 8.0791777e-01 4.5992866e-01 4.8585027e-01
6.1545002e-01 -2.1994209e-02 -3.3985060e-01 7.2285008e-01
4.5877956e-02 2.4741958e-01 7.0696771e-01 1.4345536e+00
-2.5806960e-01 -7.8316945e-01 -1.2779486e-01 7.3482764e-01
-4.4407687e-01 2.9811108e-01 4.1923601e-02 7.9785478e-01
-3.4303007e-01 2.0293760e-01 -1.8510196e-01 -3.9164338e-01
2.7721411e-01 -3.9268788e-02 7.7841103e-01 1.0018782e-01
-1.7199484e-01 2.1337405e-01 9.6427754e-02 -2.7214846e-01
-7.7119142e-01 -5.0100797e-01 -1.1542481e+00 -3.4400058e-01
-1.9207290e-01 8.2506843e-02 7.2409779e-01 8.4713715e-01
1.4753073e-01 6.5773940e-01 2.4457036e-01 -1.4028648e+00
-6.2601167e-01 -1.1109927e+00 -1.0359528e+00 5.6116241e-01
3.1590977e-01 -1.0180767e+00 -4.6457735e-01 3.2576650e-01] | [6.374929428100586, 3.426564931869507] |
74ab2158-f985-4c45-a44a-734029e7642e | hqdec-self-supervised-monocular-depth | 2305.18706 | null | https://arxiv.org/abs/2305.18706v1 | https://arxiv.org/pdf/2305.18706v1.pdf | HQDec: Self-Supervised Monocular Depth Estimation Based on a High-Quality Decoder | Decoders play significant roles in recovering scene depths. However, the decoders used in previous works ignore the propagation of multilevel lossless fine-grained information, cannot adaptively capture local and global information in parallel, and cannot perform sufficient global statistical analyses on the final output disparities. In addition, the process of mapping from a low-resolution feature space to a high-resolution feature space is a one-to-many problem that may have multiple solutions. Therefore, the quality of the recovered depth map is low. To this end, we propose a high-quality decoder (HQDec), with which multilevel near-lossless fine-grained information, obtained by the proposed adaptive axial-normalized position-embedded channel attention sampling module (AdaAxialNPCAS), can be adaptively incorporated into a low-resolution feature map with high-level semantics utilizing the proposed adaptive information exchange scheme. In the HQDec, we leverage the proposed adaptive refinement module (AdaRM) to model the local and global dependencies between pixels in parallel and utilize the proposed disparity attention module to model the distribution characteristics of disparity values from a global perspective. To recover fine-grained high-resolution features with maximal accuracy, we adaptively fuse the high-frequency information obtained by constraining the upsampled solution space utilizing the local and global dependencies between pixels into the high-resolution feature map generated from the nonlearning method. Extensive experiments demonstrate that each proposed component improves the quality of the depth estimation results over the baseline results, and the developed approach achieves state-of-the-art results on the KITTI and DDAD datasets. The code and models will be publicly available at \href{https://github.com/fwucas/HQDec}{HQDec}. | ['Jun Cheng', 'Fei Wang'] | 2023-05-30 | null | null | null | null | ['monocular-depth-estimation'] | ['computer-vision'] | [ 1.27593651e-01 -1.83194816e-01 -1.43552542e-01 -4.71708238e-01
-1.28156149e+00 -1.98842101e-02 3.19438607e-01 -8.88223797e-02
-4.21195626e-01 8.18800807e-01 4.03710246e-01 3.53296816e-01
-1.98247790e-01 -1.08741021e+00 -8.71238470e-01 -9.12635982e-01
2.52769440e-01 1.44984007e-01 5.32828212e-01 4.77750972e-03
4.79000688e-01 3.79285544e-01 -2.10411429e+00 4.01467294e-01
1.19707716e+00 1.31648171e+00 5.84763408e-01 6.41486526e-01
-8.41177106e-02 6.47801757e-01 -1.34076178e-01 2.74277432e-03
3.30215484e-01 -2.72481054e-01 -3.51587117e-01 -1.98960379e-01
7.40296781e-01 -8.76948655e-01 -4.86534983e-01 1.29697216e+00
6.91695809e-01 1.44634947e-01 5.18792510e-01 -7.00210094e-01
-3.20616663e-01 1.33868411e-01 -8.99126828e-01 3.01744610e-01
3.16818804e-01 2.06256166e-01 7.70878017e-01 -1.12616313e+00
5.65910459e-01 1.26701105e+00 3.66536468e-01 3.47765028e-01
-1.10004413e+00 -9.69704449e-01 2.62045562e-01 3.35868597e-01
-1.56224835e+00 -5.37518263e-01 7.79628277e-01 -4.38914835e-01
8.20633888e-01 -8.02121758e-02 6.36168301e-01 8.03862214e-01
3.55518728e-01 5.35262108e-01 9.89532411e-01 -2.04668254e-01
1.42560124e-01 -2.10276783e-01 1.89711928e-01 8.07585239e-01
2.49235690e-01 4.05514002e-01 -1.06018305e+00 9.90750790e-02
1.11105096e+00 -4.69413772e-02 -5.39720953e-01 -7.21808523e-02
-8.85420918e-01 6.52671397e-01 6.31478310e-01 -2.77063251e-02
-3.87720019e-01 2.37181336e-01 -6.92557320e-02 -1.34898916e-01
5.46307921e-01 -2.19536871e-02 -3.91162902e-01 -1.95586175e-01
-9.02363062e-01 1.92930803e-01 2.12553099e-01 9.51341808e-01
1.44307756e+00 -7.56307840e-02 -2.47663781e-01 7.54376054e-01
4.51933354e-01 6.15452290e-01 5.58418095e-01 -1.15964592e+00
5.55121422e-01 4.43522990e-01 -2.63362508e-02 -9.18754399e-01
-3.73855919e-01 -3.63718569e-01 -7.60329962e-01 4.46239740e-01
2.34574601e-01 9.89308059e-02 -9.61752236e-01 1.79873240e+00
4.17550236e-01 4.16271806e-01 7.94620588e-02 1.16674244e+00
8.93705487e-01 6.92716956e-01 -1.62053600e-01 -1.30518973e-01
1.15584612e+00 -9.23216343e-01 -6.32902265e-01 -2.90826201e-01
1.33992970e-01 -5.96191585e-01 9.67654943e-01 3.47818583e-01
-1.25085413e+00 -7.96671033e-01 -1.22320020e+00 -7.53825188e-01
-3.64183746e-02 -5.01342192e-02 4.43436086e-01 1.41160771e-01
-1.05978012e+00 4.30855930e-01 -1.00874043e+00 8.62511694e-02
5.37084043e-01 4.15061384e-01 -1.34765714e-01 -5.93440652e-01
-1.13868833e+00 4.50300425e-01 2.23492578e-01 1.90547079e-01
-8.35945189e-01 -7.96091974e-01 -1.09597516e+00 -1.05323650e-01
5.62730022e-02 -8.52192700e-01 8.45792055e-01 -6.69325471e-01
-1.33672202e+00 5.12528360e-01 -5.86452186e-01 -1.24269694e-01
3.91311854e-01 -3.62236887e-01 7.32434960e-03 4.10434186e-01
3.92382205e-01 9.81311202e-01 8.69716763e-01 -1.29366446e+00
-9.42276657e-01 -6.02092266e-01 -6.83885366e-02 6.05216742e-01
-2.27627367e-01 -5.77813089e-01 -7.99189866e-01 -5.25285900e-01
4.00115311e-01 -3.55614603e-01 -1.15477532e-01 2.94737905e-01
-1.52096897e-01 4.48874563e-01 3.95305246e-01 -5.75803936e-01
1.21631503e+00 -2.39478827e+00 1.94678292e-01 2.36606672e-02
2.75274754e-01 -1.27986670e-01 8.05347281e-06 -1.58204913e-01
4.01950568e-01 -2.54647136e-01 -4.21401322e-01 -5.94899952e-01
-2.44093657e-01 2.13740349e-01 -1.42181560e-01 5.52208304e-01
4.10733104e-04 6.64114237e-01 -9.83294129e-01 -4.93073106e-01
5.87452829e-01 9.02230144e-01 -8.33555698e-01 3.44969332e-01
1.48874931e-02 6.90507293e-01 -7.05721140e-01 6.31954610e-01
1.06458259e+00 -1.72594368e-01 -2.61652082e-01 -5.52527785e-01
-3.61557782e-01 1.79109395e-01 -1.13309693e+00 2.16870642e+00
-4.46058422e-01 4.63163018e-01 -1.56020019e-02 -3.52711856e-01
8.62125695e-01 -1.64633140e-01 3.85774583e-01 -1.13511825e+00
-1.34954769e-02 2.84323603e-01 -1.86607569e-01 -3.15431297e-01
7.72443593e-01 -1.38107464e-02 5.13576530e-02 2.44657937e-02
2.53687124e-03 -4.92059514e-02 -2.08894432e-01 -4.08644266e-02
9.14215684e-01 2.61640519e-01 2.16342300e-01 -1.09329864e-01
6.48776591e-01 -1.50943533e-01 7.91905701e-01 5.18749297e-01
-5.84184080e-02 1.11132085e+00 6.90806657e-02 -8.99588093e-02
-8.45531821e-01 -1.19542599e+00 -4.88741696e-01 7.01046109e-01
7.49079347e-01 -2.38641545e-01 -6.29813850e-01 -1.92422196e-01
1.92037225e-02 3.50120127e-01 -6.47295177e-01 -2.20589310e-01
-4.51801062e-01 -6.57014549e-01 3.03191692e-01 4.93082166e-01
9.09890890e-01 -8.17335188e-01 -6.91819549e-01 2.37118959e-01
-3.57828945e-01 -1.15389562e+00 -4.35832351e-01 1.91298589e-01
-9.34249043e-01 -8.87034714e-01 -6.24697626e-01 -5.58321118e-01
6.38356328e-01 2.57258087e-01 8.08362246e-01 -1.71972126e-01
-1.89058363e-01 8.36527571e-02 -2.27547660e-01 4.83694859e-02
2.51909256e-01 1.59964040e-02 -2.15858787e-01 1.29199624e-01
2.14272082e-01 -7.86671519e-01 -1.09256113e+00 1.50638297e-01
-9.62681770e-01 3.61004829e-01 5.81280649e-01 8.19587886e-01
1.07920170e+00 6.28427267e-02 4.87485051e-01 -7.42203116e-01
1.82338897e-02 -4.60942507e-01 -7.03607619e-01 -3.88943672e-01
-5.52276850e-01 2.30968714e-01 5.03324628e-01 -7.97328427e-02
-1.38717556e+00 1.76080447e-02 -4.83469009e-01 -5.43512702e-01
-1.62320822e-01 2.00628459e-01 -4.85750347e-01 -9.14575607e-02
3.79065663e-01 3.77322078e-01 -3.16999853e-01 -5.36426604e-01
1.91449121e-01 4.50175554e-01 6.86917782e-01 -6.41446173e-01
6.24300241e-01 8.48561347e-01 3.43203614e-03 -5.23693204e-01
-9.94449496e-01 -4.29783911e-01 -5.45791149e-01 -1.13990739e-01
1.04511738e+00 -1.36140776e+00 -4.16148692e-01 8.62122118e-01
-9.66050804e-01 -4.30404007e-01 -3.41147333e-01 5.70333719e-01
-5.46298504e-01 2.73833543e-01 -7.00099289e-01 -5.57406604e-01
-1.96569815e-01 -1.31527495e+00 1.38648260e+00 5.51942945e-01
2.44517311e-01 -8.30172777e-01 -1.36020139e-03 3.61115873e-01
2.66185015e-01 1.52208701e-01 7.89925396e-01 1.65031329e-01
-1.22957134e+00 2.90509492e-01 -5.43150902e-01 2.69884616e-01
9.31617022e-02 -3.40570718e-01 -1.34899414e+00 -2.21490070e-01
-1.72852036e-02 -2.88499534e-01 1.13923824e+00 8.32668006e-01
1.21006441e+00 6.25333712e-02 -1.85440168e-01 1.26658666e+00
1.77735913e+00 1.80149246e-02 8.99286509e-01 4.01812732e-01
9.79543984e-01 4.51934993e-01 9.26917255e-01 7.79557586e-01
8.64212334e-01 6.69886231e-01 6.49828017e-01 -5.70205450e-02
-6.39051139e-01 -4.23335910e-01 2.44196385e-01 7.52936244e-01
4.26144805e-03 -1.39571145e-01 -5.30370414e-01 4.72996473e-01
-1.78622782e+00 -7.15325236e-01 -6.64920732e-02 2.21393180e+00
8.47134352e-01 9.10230577e-02 -4.73487675e-01 -8.42676163e-02
6.10747457e-01 3.63717049e-01 -8.74642909e-01 3.54612916e-04
-3.41190040e-01 2.44413361e-01 6.13836884e-01 9.62812126e-01
-7.42250443e-01 9.03492212e-01 5.07028008e+00 9.63963628e-01
-1.19142032e+00 9.25868377e-02 6.61160171e-01 -2.32272446e-01
-5.90833366e-01 -1.29437745e-01 -1.15758371e+00 5.75476229e-01
6.41853869e-01 -3.93257998e-02 3.22373629e-01 5.23913383e-01
3.72424126e-01 -5.34864843e-01 -1.10100961e+00 1.24174142e+00
8.42166543e-02 -1.18630683e+00 5.32398261e-02 1.47919238e-01
9.12817597e-01 1.69153079e-01 3.29946309e-01 6.06124802e-03
-5.57154277e-03 -9.82682526e-01 7.81771779e-01 8.39151859e-01
1.20597446e+00 -8.77162874e-01 5.81054866e-01 2.53596157e-01
-1.43045723e+00 -9.96344835e-02 -5.02488315e-01 -1.17603652e-01
2.05763847e-01 8.38947535e-01 -1.00927979e-01 7.60022700e-01
9.37841177e-01 1.10761297e+00 -3.92476022e-01 9.52334940e-01
-1.70872435e-01 1.73252538e-01 -2.53917634e-01 5.99061489e-01
-3.11546922e-02 -4.69033159e-02 3.52937400e-01 8.66445482e-01
5.03532529e-01 3.36049914e-01 -1.01550363e-01 8.84625971e-01
1.47945043e-02 -1.80738643e-01 -2.19166279e-01 7.51362801e-01
5.31826675e-01 9.94658768e-01 -3.29840243e-01 -1.98846489e-01
-4.90864187e-01 8.04473281e-01 4.45818305e-01 5.18704176e-01
-7.15690613e-01 -2.34257206e-01 9.67673600e-01 3.27513993e-01
4.39813524e-01 2.54822313e-03 -5.79583347e-01 -1.34916925e+00
1.40298501e-01 -6.62182033e-01 2.93508261e-01 -8.89907718e-01
-1.13927662e+00 6.59287632e-01 -1.06156550e-01 -1.40512395e+00
-2.52188087e-01 -1.89874217e-01 -2.24226981e-01 1.19628417e+00
-2.22329450e+00 -8.61711442e-01 -8.90451312e-01 8.59880209e-01
7.71689177e-01 2.75810272e-01 4.78031754e-01 6.63956404e-01
-4.65124905e-01 6.37651384e-01 -8.35447684e-02 -1.72940105e-01
7.82213032e-01 -8.27929556e-01 -6.48454651e-02 9.41393435e-01
-3.66052508e-01 1.44724295e-01 3.52000952e-01 -6.18685246e-01
-1.24339390e+00 -1.24691188e+00 3.05069804e-01 -7.70734400e-02
1.99959353e-01 -3.01291585e-01 -1.14036787e+00 3.94174874e-01
-3.27496111e-01 3.37084472e-01 1.50502264e-01 -4.51208293e-01
-2.15150505e-01 -2.88241118e-01 -1.30781972e+00 2.44277969e-01
1.10623932e+00 -6.38520241e-01 -2.13865653e-01 -3.26269507e-01
8.72855961e-01 -7.68979728e-01 -8.06476235e-01 5.46949625e-01
5.87060153e-01 -1.41223216e+00 9.77046192e-01 4.51308608e-01
7.55132318e-01 -5.61205506e-01 -5.84102631e-01 -8.81541729e-01
-3.30464393e-01 -1.40777051e-01 -2.14089528e-01 1.15366888e+00
8.03063139e-02 -7.99911559e-01 7.81713724e-01 5.52747965e-01
-4.94866371e-01 -8.33571076e-01 -1.17504966e+00 -1.94375172e-01
-3.59612107e-02 -5.63952506e-01 6.47472501e-01 5.96448362e-01
-4.56090063e-01 1.27890855e-01 -2.40643203e-01 5.77864826e-01
1.00982559e+00 1.53864995e-01 5.90145469e-01 -8.94869447e-01
-4.03294742e-01 -3.20397973e-01 -5.80272019e-01 -1.63401818e+00
1.08290985e-01 -6.51424050e-01 2.43182331e-01 -1.60276616e+00
2.24187195e-01 -6.04477167e-01 -3.04333895e-01 1.47968292e-01
-2.76692510e-01 4.52161759e-01 -7.79685527e-02 3.00904125e-01
-3.44717145e-01 8.72714579e-01 1.50300848e+00 1.06035121e-01
-1.50724605e-01 -3.69040430e-01 -7.27070928e-01 7.29195535e-01
3.70762855e-01 -4.29260373e-01 -5.88357627e-01 -8.76290619e-01
4.14135978e-02 3.04812312e-01 3.98227602e-01 -1.19007027e+00
5.02759099e-01 -7.64967427e-02 6.74622297e-01 -7.33629167e-01
7.55490780e-01 -6.69750035e-01 -6.05064370e-02 1.13269158e-01
-9.47123617e-02 -2.79226273e-01 2.78837323e-01 5.90397000e-01
-5.48504710e-01 1.57508254e-01 1.01799977e+00 7.33054951e-02
-9.72618401e-01 7.10733712e-01 6.52141050e-02 1.32149979e-01
7.18746066e-01 -5.04368901e-01 -5.20539165e-01 -2.67253876e-01
-4.08531547e-01 2.02137277e-01 8.41409206e-01 1.37709782e-01
1.02453625e+00 -1.29176962e+00 -6.33518457e-01 6.30115688e-01
1.44668043e-01 7.72411704e-01 7.95822263e-01 6.98582053e-01
-4.93868023e-01 -2.13095732e-03 -3.26865047e-01 -8.10950518e-01
-7.60339200e-01 1.99651942e-01 4.08617586e-01 -1.23844840e-01
-9.18683887e-01 9.57883000e-01 7.96778083e-01 -6.98960200e-02
1.57967508e-01 -4.09456223e-01 -1.99922785e-01 -4.91237044e-02
6.86283648e-01 3.05270910e-01 -2.14575306e-02 -7.12573588e-01
-2.61080801e-01 1.21015465e+00 -5.73573670e-05 -2.31731534e-01
1.36746395e+00 -6.66035712e-01 5.10536544e-02 4.22781497e-01
1.42848957e+00 1.55278593e-01 -1.92241180e+00 -2.46032804e-01
-7.22949803e-01 -9.85660315e-01 4.73235667e-01 -4.96469587e-01
-1.26136553e+00 9.55253005e-01 7.63282955e-01 -6.09667182e-01
1.54452062e+00 -1.25170469e-01 9.34860468e-01 -2.88101673e-01
5.22998273e-01 -7.61324942e-01 6.79858252e-02 4.53002065e-01
5.91975391e-01 -1.26643467e+00 4.10576835e-02 -5.58347106e-01
-3.33039641e-01 9.28979039e-01 9.60735261e-01 -1.80744112e-01
6.59538865e-01 5.23995638e-01 2.65288893e-02 -8.11520070e-02
-7.73641288e-01 -2.20642745e-01 3.60242575e-01 6.66788280e-01
1.61329910e-01 -3.26509655e-01 3.95406894e-02 5.33209801e-01
-1.11493491e-01 -9.15415026e-03 4.99971092e-01 6.06339335e-01
-5.26019156e-01 -8.98028910e-01 -2.88355082e-01 4.04782236e-01
-1.23579182e-01 -4.35640812e-01 4.10028636e-01 4.68744606e-01
3.81830513e-01 7.91239142e-01 4.16220009e-01 -4.00003850e-01
2.72541940e-01 -4.56011444e-01 4.93320078e-01 -5.83226562e-01
-8.03840309e-02 2.50246018e-01 -1.36172056e-01 -9.93106902e-01
-3.00691694e-01 -5.78086674e-01 -1.62420881e+00 -2.57713795e-01
4.82952036e-02 -2.69652724e-01 3.82164836e-01 7.15302229e-01
4.21674550e-01 6.95201218e-01 6.18936300e-01 -1.15209401e+00
6.11168370e-02 -6.88229561e-01 -6.89188600e-01 1.64945751e-01
6.89946949e-01 -7.68565297e-01 -5.46368599e-01 -1.80004090e-01] | [9.382522583007812, -2.495950222015381] |
07158f45-aeab-4a2c-a7ba-82da6d820781 | geometric-dimensionality-reduction-for | 1709.01233 | null | https://arxiv.org/abs/1709.01233v9 | https://arxiv.org/pdf/1709.01233v9.pdf | Supervised Dimensionality Reduction for Big Data | To solve key biomedical problems, experimentalists now routinely measure millions or billions of features (dimensions) per sample, with the hope that data science techniques will be able to build accurate data-driven inferences. Because sample sizes are typically orders of magnitude smaller than the dimensionality of these data, valid inferences require finding a low-dimensional representation that preserves the discriminating information (e.g., whether the individual suffers from a particular disease). There is a lack of interpretable supervised dimensionality reduction methods that scale to millions of dimensions with strong statistical theoretical guarantees.We introduce an approach, XOX, to extending principal components analysis by incorporating class-conditional moment estimates into the low-dimensional projection. The simplest ver-sion, "Linear Optimal Low-rank" projection (LOL), incorporates the class-conditional means. We prove, and substantiate with both synthetic and real data benchmarks, that LOL and its generalizations in the XOX framework lead to improved data representations for subsequent classification, while maintaining computational efficiency and scalability. Using multiple brain imaging datasets consisting of >150 million features, and several genomics datasets with>500,000 features, LOL outperforms other scalable linear dimensionality reduction techniques in terms of accuracy, while only requiring a few minutes on a standard desktop computer. | ['Christopher Douville', 'Minh Tang', 'Joshua T. Vogelstein', 'Mauro Maggioni', 'Eric Bridgeford', 'Da Zheng', 'Randal Burns'] | 2017-09-05 | null | null | null | null | ['supervised-dimensionality-reduction'] | ['computer-vision'] | [ 3.18282843e-01 -7.77685270e-02 -2.22973093e-01 -6.39507473e-01
-7.54728973e-01 -4.87042785e-01 5.97602367e-01 2.80182064e-01
-3.29589933e-01 7.65314996e-01 5.04250586e-01 -1.74516961e-01
-6.38147950e-01 -6.94502413e-01 -4.34463501e-01 -7.26394415e-01
-2.97761798e-01 6.36445522e-01 -4.74758923e-01 3.19535047e-01
2.57834285e-01 5.82757890e-01 -1.50542974e+00 2.12788478e-01
5.98169565e-01 9.87171352e-01 -1.46150872e-01 4.66988504e-01
9.90028605e-02 -2.55919807e-02 -1.02697514e-01 -1.72771037e-01
3.72179031e-01 -3.11126769e-01 -5.62447369e-01 -8.17031711e-02
5.25760829e-01 -3.22443962e-01 -3.26814830e-01 1.04173648e+00
5.14368892e-01 -9.58748087e-02 8.97768855e-01 -1.17019904e+00
-5.85363328e-01 3.93214911e-01 -6.76527679e-01 2.84965485e-01
1.98659524e-01 -5.83361760e-02 1.01761639e+00 -8.58187199e-01
7.00968444e-01 1.39973283e+00 7.52044678e-01 5.57928145e-01
-1.86917150e+00 -4.48268622e-01 -2.35028222e-01 -1.55872002e-01
-1.19434893e+00 -6.31237209e-01 5.22285879e-01 -7.74381101e-01
7.61267900e-01 6.06445491e-01 4.98439461e-01 1.03562129e+00
5.60814738e-01 2.53990084e-01 1.27516592e+00 -1.14423253e-01
5.09712994e-01 -1.17693208e-01 6.91095948e-01 5.14925838e-01
7.88627446e-01 1.67553667e-02 -6.04122400e-01 -8.46197844e-01
4.04223979e-01 2.56211877e-01 -5.55327013e-02 -5.62772453e-01
-1.55450964e+00 9.38737452e-01 -1.62538379e-01 -2.06484608e-02
-5.55068016e-01 5.86418658e-02 5.85729003e-01 1.47754952e-01
4.37143624e-01 6.38270259e-01 -4.50931340e-01 -1.84448659e-01
-7.90646911e-01 3.12799662e-01 6.90901518e-01 5.25549889e-01
6.11757874e-01 -3.55178475e-01 8.52662232e-03 6.80886924e-01
-4.89746826e-03 6.65796399e-01 7.81250417e-01 -1.07564533e+00
2.49476567e-01 6.63871348e-01 -1.57736585e-01 -1.28255820e+00
-7.62327194e-01 -4.05485570e-01 -1.37728953e+00 -1.15510680e-01
2.19490334e-01 -1.68448254e-01 -5.22152424e-01 1.96416974e+00
4.90145832e-01 -2.86102388e-03 -1.12810489e-02 6.43716335e-01
2.95380682e-01 3.46884161e-01 -5.44963107e-02 -4.22037810e-01
1.67643321e+00 8.11666474e-02 -5.79494476e-01 -1.40660241e-01
8.19420040e-01 -2.83854336e-01 1.10207546e+00 5.58813930e-01
-5.72066307e-01 2.88761929e-02 -8.10940146e-01 -1.23505756e-01
-2.57386446e-01 4.23144475e-02 1.24771631e+00 7.32229173e-01
-7.05899894e-01 6.66847289e-01 -1.00447237e+00 -3.58603388e-01
8.50252509e-01 3.63916367e-01 -8.52064729e-01 -3.14273626e-01
-8.11678886e-01 4.35216039e-01 1.81819633e-01 -2.16701254e-01
-6.06931806e-01 -1.22907209e+00 -5.44738531e-01 1.57671258e-01
-1.86966565e-02 -7.80763030e-01 4.23614711e-01 -3.15589830e-02
-8.60910237e-01 6.58769369e-01 -3.33214015e-01 -4.50215995e-01
2.29836494e-01 -2.49310419e-01 -1.43363297e-01 3.07092577e-01
2.43421480e-01 5.61490595e-01 6.49292409e-01 -6.17543697e-01
-2.76266724e-01 -1.02525496e+00 -4.36196238e-01 -9.90334228e-02
-9.51274276e-01 -1.43461347e-01 -6.06135232e-03 -5.10527730e-01
6.34787977e-01 -1.09829748e+00 -4.03079718e-01 3.91031533e-01
-4.73658115e-01 -1.02169424e-01 6.19051576e-01 -5.92593670e-01
1.14235497e+00 -2.21313787e+00 3.04622948e-01 2.12093011e-01
7.90891349e-01 -6.71206340e-02 -1.39562950e-01 1.73760846e-01
-3.31814080e-01 3.89965363e-02 -4.05951589e-01 -1.13953643e-01
-4.82583307e-02 1.68948144e-01 -3.63689214e-01 7.85423696e-01
1.08851187e-01 7.97363162e-01 -7.83521652e-01 -2.34046742e-01
3.78056839e-02 5.47519445e-01 -8.53004515e-01 -2.51900196e-01
2.37090096e-01 1.35768801e-01 -5.49041390e-01 3.88153911e-01
5.51124394e-01 -5.41095078e-01 3.49290401e-01 -2.97094762e-01
1.51786804e-01 1.08089194e-01 -1.01020837e+00 1.62549758e+00
-5.72467223e-02 6.45938337e-01 -3.21568042e-01 -1.16908252e+00
7.65506506e-01 3.07268519e-02 8.67295682e-01 -3.84346515e-01
-3.99876684e-02 6.26100898e-02 -3.16347368e-02 -2.87914336e-01
-7.24290535e-02 -1.52417615e-01 -2.81323433e-01 6.30715728e-01
-1.60334438e-01 2.44802713e-01 2.14713231e-01 3.19385231e-01
1.53456604e+00 -5.30577302e-01 5.06965816e-01 -6.35176361e-01
8.12694132e-02 -7.55449608e-02 8.71260703e-01 5.00602603e-01
-7.50253797e-02 4.37504530e-01 1.16194594e+00 -6.51367426e-01
-1.07542276e+00 -1.19751418e+00 -7.51694977e-01 7.79659510e-01
-4.71255481e-01 -5.76768160e-01 -7.40388036e-01 -3.33011776e-01
4.61758286e-01 5.20286620e-01 -9.01013196e-01 -3.79994154e-01
-1.10971801e-01 -1.67315006e+00 5.07042944e-01 3.56960893e-01
4.94791679e-02 -3.35571915e-01 -5.88122070e-01 1.06238704e-02
1.49653181e-01 -9.40191627e-01 -1.09516397e-01 1.38316542e-01
-1.19850147e+00 -1.02079260e+00 -3.90467733e-01 -1.27940297e-01
9.72776115e-01 1.22291699e-01 7.94244409e-01 -4.80368286e-01
-9.05322969e-01 1.33129224e-01 -1.15080015e-03 -2.14198753e-01
-4.11170796e-02 -1.71110019e-01 7.14786589e-01 -1.09652929e-01
8.11875582e-01 -9.40670252e-01 -7.69336164e-01 -7.02382848e-02
-8.07015061e-01 -1.09685525e-01 8.73468399e-01 9.56236839e-01
8.20800841e-01 4.20534611e-02 8.42871785e-01 -1.08544517e+00
7.38924980e-01 -6.82081997e-01 -4.93144572e-01 1.01536177e-02
-9.54499722e-01 4.65798110e-01 6.40308440e-01 -4.39192325e-01
-5.02089977e-01 2.12526634e-01 1.34119302e-01 -3.33860993e-01
-9.28499475e-02 5.52462101e-01 -1.22501306e-01 1.33910537e-01
8.66593719e-01 3.43185097e-01 3.03997815e-01 -7.59173691e-01
4.67743248e-01 6.19986594e-01 3.27661783e-01 -5.75827360e-01
3.66414398e-01 7.76807666e-01 5.57586253e-01 -1.06819761e+00
-8.38806033e-01 -2.14466110e-01 -7.54854202e-01 3.50326747e-01
7.03603745e-01 -7.08206713e-01 -8.06984961e-01 1.03713520e-01
-5.70736289e-01 2.32473403e-01 -3.83870929e-01 7.89292872e-01
-6.17482483e-01 3.76902908e-01 -5.15061080e-01 -3.36800069e-01
-3.17420661e-01 -8.60083401e-01 1.10131717e+00 -3.52177233e-01
-4.24698204e-01 -6.07643902e-01 3.87138098e-01 2.66136557e-01
1.42143771e-01 4.11151230e-01 1.51790905e+00 -6.93325996e-01
-3.18839811e-02 -5.49386322e-01 -3.69542927e-01 2.94852465e-01
2.45023012e-01 -2.78900027e-01 -8.63421142e-01 -4.16450441e-01
1.31813377e-01 -1.70579076e-01 7.55968869e-01 5.31709433e-01
1.42629170e+00 -4.55082327e-01 -5.99142909e-01 7.09263921e-01
1.20499504e+00 -2.79892057e-01 4.40863460e-01 -3.28561962e-01
6.36560321e-01 6.50024891e-01 2.56677240e-01 7.30184555e-01
9.60318297e-02 4.56495285e-01 -1.45361081e-01 3.16261202e-01
4.67053354e-02 4.99280170e-03 1.81771480e-02 9.33464408e-01
1.88748747e-01 3.06982219e-01 -1.01033461e+00 3.15251499e-01
-1.52514267e+00 -8.88126075e-01 -1.40735647e-02 2.43005991e+00
8.25052738e-01 3.67072821e-02 2.68225297e-02 2.01635763e-01
4.01979685e-01 -1.19001046e-01 -1.04092360e+00 -6.40727207e-02
-1.24972984e-01 -5.44704571e-02 3.58848423e-01 2.25862443e-01
-1.06282353e+00 3.04821104e-01 7.07359457e+00 5.47990441e-01
-9.06117022e-01 -1.05098225e-01 9.52649355e-01 -4.84109849e-01
-3.11363131e-01 -1.26064718e-01 -7.57858813e-01 5.06883264e-01
1.21990979e+00 -5.87726712e-01 4.65637922e-01 9.00794804e-01
3.47962290e-01 3.82728055e-02 -1.38223743e+00 1.27677250e+00
5.15935719e-02 -1.40924692e+00 2.19620258e-01 6.88696742e-01
6.49772406e-01 1.17959708e-01 1.82468414e-01 -1.00766264e-01
1.08319826e-01 -9.87312794e-01 -5.32416031e-02 8.25989783e-01
9.00947094e-01 -7.39840508e-01 3.27952981e-01 4.14584666e-01
-6.29923224e-01 -2.23825425e-01 -8.73864591e-01 -7.43777957e-03
-1.92016870e-01 1.37059236e+00 -6.32636607e-01 1.02788381e-01
5.27786136e-01 7.54449069e-01 -6.63212061e-01 6.84148312e-01
4.11360443e-01 5.27969837e-01 -5.33971190e-01 1.10658862e-01
-1.67913094e-01 -3.68090689e-01 5.52629173e-01 9.71411347e-01
3.58812809e-01 2.89841652e-01 -1.04512431e-01 7.80025721e-01
-1.28344864e-01 9.65748355e-02 -6.65692031e-01 -5.83905458e-01
5.83423197e-01 1.27171683e+00 -7.51262128e-01 -3.30127746e-01
-3.20698649e-01 7.13879049e-01 3.03799361e-01 -1.10136457e-01
-1.43644109e-01 -4.03803617e-01 1.22451484e+00 8.70123506e-02
-1.09650962e-01 -3.30500990e-01 -4.44866002e-01 -1.22501469e+00
1.24132566e-01 -1.06058860e+00 3.36587459e-01 -2.96625078e-01
-1.75363731e+00 3.21331233e-01 -2.23138481e-02 -1.07976806e+00
-3.21712464e-01 -8.34474146e-01 6.36733770e-02 7.11215258e-01
-8.96352172e-01 -5.61763167e-01 7.23398104e-02 1.83542356e-01
5.21065481e-03 -2.42734000e-01 1.28054559e+00 3.85016948e-01
-7.19162107e-01 3.67581248e-01 6.81184471e-01 3.64009058e-04
5.24402797e-01 -1.05022597e+00 2.64028281e-01 4.12988067e-01
-5.93851097e-02 9.73284066e-01 6.83956265e-01 -6.59329891e-01
-1.95850968e+00 -9.54574108e-01 7.77939439e-01 -5.83530784e-01
6.98114038e-01 -6.27209187e-01 -8.67023945e-01 5.67881286e-01
-7.21096098e-01 3.92879635e-01 1.23677206e+00 5.34008026e-01
-4.89939362e-01 -1.91706166e-01 -1.30671275e+00 4.54839587e-01
1.13012278e+00 -5.69869459e-01 -5.84963143e-01 6.68891966e-01
4.25093919e-01 2.87489891e-01 -1.35226047e+00 2.06457704e-01
8.20490062e-01 -6.27462745e-01 1.08575022e+00 -1.20011437e+00
3.40600997e-01 4.44231071e-02 -4.46611404e-01 -1.05410111e+00
-4.08399850e-01 -4.00153577e-01 -2.49730647e-01 8.84858131e-01
1.81083933e-01 -7.67012954e-01 9.08243418e-01 1.09143162e+00
3.76842797e-01 -9.14755166e-01 -1.09588552e+00 -7.87678182e-01
3.11385155e-01 -4.38222259e-01 6.85125172e-01 1.13742578e+00
2.47047663e-01 3.58399838e-01 -3.10375869e-01 7.04062507e-02
1.04433453e+00 2.71580249e-01 6.71716928e-01 -1.82575142e+00
-1.84148386e-01 -2.82689035e-01 -1.05015159e+00 -5.98669887e-01
7.47019500e-02 -1.03119481e+00 -4.56056952e-01 -1.15473771e+00
7.30182588e-01 -4.74698216e-01 -2.54699141e-01 4.59618419e-01
-1.67697236e-01 1.95150942e-01 -2.66501963e-01 1.64859012e-01
-3.19423735e-01 4.43018556e-01 7.74879992e-01 5.48595637e-02
-6.01175688e-02 -3.28854918e-01 -1.12915897e+00 7.84819901e-01
5.37767112e-01 -7.10768819e-01 -5.25890648e-01 -2.81721473e-01
4.71996784e-01 -4.88762707e-02 2.25860864e-01 -9.99330401e-01
-1.61832366e-02 -9.90537778e-02 7.81757832e-01 -3.97905648e-01
4.61451501e-01 -4.24700230e-01 2.00596318e-01 5.11834621e-01
-6.79251194e-01 -8.29702914e-02 1.76559445e-02 7.10622132e-01
2.26367489e-01 1.24383800e-01 6.17105901e-01 1.43231228e-01
-1.40183210e-01 4.10881907e-01 -1.19499698e-01 9.59130526e-02
9.97228861e-01 1.09578945e-01 -3.81327063e-01 2.26173038e-03
-7.65769184e-01 -6.23612665e-03 2.88497210e-01 1.86186701e-01
7.05811799e-01 -1.38510871e+00 -8.75609815e-01 4.25655425e-01
1.33825481e-01 -4.31426823e-01 3.91599000e-01 8.22955608e-01
-1.68975592e-01 7.48583019e-01 -3.19709837e-01 -5.84808588e-01
-1.13501072e+00 7.48156667e-01 -9.86137986e-02 2.20795884e-03
-9.46920574e-01 4.34877276e-01 4.13172811e-01 -4.32918161e-01
-2.29293048e-01 -2.72391945e-01 1.29612818e-01 3.45830590e-01
1.01794934e+00 6.87827289e-01 1.85873695e-02 -4.38408315e-01
-4.78970051e-01 3.95813137e-01 -2.39050239e-01 1.48912277e-02
1.84444463e+00 -1.64590422e-02 -3.82552266e-01 4.65632617e-01
1.49214470e+00 -2.04650253e-01 -1.05164552e+00 -1.12677991e-01
1.16187856e-01 -6.60771012e-01 2.48680755e-01 -3.84803027e-01
-8.79851043e-01 1.00528026e+00 7.11178780e-01 4.13445123e-02
9.73057568e-01 1.09412052e-01 5.01572371e-01 9.08847749e-01
5.18781185e-01 -7.92000473e-01 -3.07199329e-01 7.65827578e-03
7.02394307e-01 -1.00691748e+00 4.79313523e-01 -2.27414221e-01
-4.04984325e-01 8.70963514e-01 4.78917807e-02 -2.74338186e-01
8.75636280e-01 3.15292567e-01 -4.04442489e-01 -5.14920175e-01
-1.10054410e+00 5.00543535e-01 2.60317951e-01 5.24208248e-01
3.03207159e-01 4.08371627e-01 -5.22541106e-01 7.25466371e-01
-3.95676911e-01 3.49637843e-03 3.09908122e-01 7.35013664e-01
-4.64597911e-01 -9.35943902e-01 -3.36949319e-01 1.47143281e+00
-4.14739877e-01 -9.78922099e-02 -2.75871485e-01 4.12777364e-01
-2.55761921e-01 5.20858884e-01 2.51324862e-01 -2.59076744e-01
1.08952597e-02 2.18874827e-01 2.71157801e-01 -5.95806777e-01
3.27170700e-01 -6.96684122e-02 -1.73929319e-01 -9.06883836e-01
-2.40755863e-02 -1.17900693e+00 -1.15785587e+00 -4.81440336e-01
3.69245708e-02 -3.74749862e-02 7.24066317e-01 8.20658863e-01
9.07043874e-01 2.36677766e-01 5.33234715e-01 -5.16384900e-01
-9.13223028e-01 -6.88129842e-01 -8.87745440e-01 4.94442523e-01
2.22827390e-01 -5.79797268e-01 -4.32760090e-01 2.39611529e-02] | [7.2881855964660645, 4.852302074432373] |
54eb000a-9f7c-45c3-a3c3-801aef003867 | classification-of-lung-nodules-in-ct-images | 1807.00094 | null | http://arxiv.org/abs/1807.00094v1 | http://arxiv.org/pdf/1807.00094v1.pdf | Classification of lung nodules in CT images based on Wasserstein distance in differential geometry | Lung nodules are commonly detected in screening for patients with a risk for
lung cancer. Though the status of large nodules can be easily diagnosed by fine
needle biopsy or bronchoscopy, small nodules are often difficult to classify on
computed tomography (CT). Recent works have shown that shape analysis of lung
nodules can be used to differentiate benign lesions from malignant ones, though
existing methods are limited in their sensitivity and specificity. In this work
we introduced a new 3D shape analysis within the framework of differential
geometry to calculate the Wasserstein distance between benign and malignant
lung nodules to derive an accurate classification scheme. The Wasserstein
distance between the nodules is calculated based on our new spherical optimal
mass transport, this new algorithm works directly on sphere by using spherical
metric, which is much more accurate and efficient than previous methods. In the
process of deformation, the area-distortion factor gives a probability measure
on the unit sphere, which forms the Wasserstein space. From known cases of
benign and malignant lung nodules, we can calculate a unique optimal mass
transport map between their correspondingly deformed Wasserstein spaces. This
transportation cost defines the Wasserstein distance between them and can be
used to classify new lung nodules into either the benign or malignant class. To
the best of our knowledge, this is the first work that utilizes Wasserstein
distance for lung nodule classification. The advantages of Wasserstein distance
are it is invariant under rigid motions and scalings, thus it intrinsically
measures shape distance even when the underlying shapes are of high complexity,
making it well suited to classify lung nodules as they have different sizes,
orientations, and appearances. | ['Xianfeng GU', 'Deruo Liu', 'Qianli Ma', 'Xiaoyin Xu', 'Chengfeng Wen', 'Min Zhang', 'Hai Chen', 'Jie He'] | 2018-06-30 | null | null | null | null | ['lung-nodule-classification'] | ['medical'] | [-1.34043023e-02 2.52657354e-01 -5.77341095e-02 -5.78212412e-03
-3.51916105e-01 -6.70508087e-01 4.99822706e-01 2.28696037e-02
-3.07248980e-01 2.74645329e-01 -2.87122846e-01 -5.24951875e-01
-2.85158277e-01 -9.71551597e-01 -2.11592510e-01 -8.76907885e-01
1.42978832e-01 9.65604961e-01 1.05697823e+00 7.80041441e-02
-1.57776177e-01 1.10833609e+00 -1.13670444e+00 -3.07431906e-01
7.21954584e-01 6.89514816e-01 2.98550189e-01 8.56323242e-01
-6.76127374e-02 1.56331137e-01 1.30855618e-03 -3.21573168e-01
3.76797348e-01 -7.77423799e-01 -1.07227468e+00 2.43544266e-01
7.62806833e-02 -2.11952910e-01 -1.26317367e-01 1.04595518e+00
2.25178823e-01 -1.31771892e-01 1.15015066e+00 -7.94077337e-01
-4.38329101e-01 9.13704559e-02 -2.17341810e-01 1.13087632e-01
3.05534601e-02 -3.17860752e-01 8.11873972e-01 -9.30961490e-01
4.50598449e-01 9.09296513e-01 7.64004767e-01 6.42769396e-01
-8.98228347e-01 -8.74157101e-02 -6.05889916e-01 -4.96848151e-02
-1.19299281e+00 1.54893875e-01 2.72141963e-01 -7.00854182e-01
2.73932815e-01 8.17336798e-01 8.96767974e-01 2.37390473e-01
4.84035105e-01 3.50145817e-01 9.03654277e-01 -2.78276414e-01
2.03995835e-02 3.44787389e-01 -2.50487059e-01 1.03931499e+00
8.06806862e-01 -4.93097156e-02 7.00156271e-01 -3.10386449e-01
1.13179016e+00 4.94903564e-01 -2.90455341e-01 -9.60755825e-01
-1.67730308e+00 5.96873164e-01 5.15892029e-01 1.04433167e+00
-2.46935412e-02 9.89669710e-02 -1.19795278e-01 -1.17728993e-01
1.10730432e-01 3.20598871e-01 -4.85201599e-03 1.39241949e-01
-6.32469714e-01 -8.01381916e-02 8.94373715e-01 5.66617608e-01
4.00562406e-01 -3.45822603e-01 -1.96861163e-01 5.26237190e-01
6.34841979e-01 8.81379426e-01 7.34585345e-01 -5.83092928e-01
-4.39849794e-02 7.05018997e-01 -9.88052338e-02 -9.05581355e-01
-3.65140170e-01 -8.07567686e-02 -8.14203918e-01 1.06402420e-01
6.94148481e-01 5.77278852e-01 -6.27394497e-01 1.08291185e+00
5.40510595e-01 1.20380111e-01 -3.73809412e-02 6.93055749e-01
6.26210988e-01 3.12372744e-02 -3.75842601e-01 -1.58284202e-01
1.38229501e+00 -6.93943262e-01 -3.77385050e-01 4.16779727e-01
9.51203942e-01 -6.99485838e-01 8.76234174e-01 -2.28276685e-01
-1.12468672e+00 -4.37981915e-03 -8.08012605e-01 2.62557089e-01
-1.34940609e-01 3.67339589e-02 6.99938685e-02 9.73707616e-01
-1.05481327e+00 8.37417364e-01 -1.42182255e+00 -7.31015742e-01
4.66355868e-02 3.67965490e-01 -2.11901948e-01 3.74681115e-01
-8.51448953e-01 1.10114658e+00 1.97023779e-01 -1.24999173e-01
-4.85550165e-01 -4.18927014e-01 -4.52357978e-01 -6.02348372e-02
2.04521686e-01 -6.26422584e-01 1.27888548e+00 -4.47693110e-01
-1.32221794e+00 1.02013040e+00 -1.66763999e-02 -2.32301816e-01
8.24017048e-01 6.64985657e-01 -2.80830592e-01 3.23946744e-01
-3.63329686e-02 2.60218054e-01 7.79597580e-01 -7.15947628e-01
-4.64270711e-01 -3.29624474e-01 -2.97279119e-01 7.87253082e-02
-3.01593393e-01 -1.48259252e-01 -1.66287482e-01 -4.04560775e-01
6.52550519e-01 -1.34930515e+00 -2.24795222e-01 2.53618479e-01
-2.70430833e-01 -3.37493271e-01 9.49517429e-01 -3.83865297e-01
1.03592396e+00 -1.96693349e+00 -7.14873746e-02 4.39768285e-01
5.44174075e-01 1.92385048e-01 2.83074170e-01 2.04880070e-02
4.85785790e-02 5.74392438e-01 -3.71303439e-01 5.37976265e-01
-3.57268564e-02 1.67626351e-01 3.20391357e-01 7.54740119e-01
7.12376311e-02 1.01023257e+00 -1.05650949e+00 -8.50919008e-01
1.37780696e-01 4.98999894e-01 -2.31008381e-01 -5.83500825e-02
3.42994869e-01 2.60057837e-01 -7.78505147e-01 4.60669726e-01
5.91237664e-01 -3.41967046e-01 5.58723416e-03 -1.67377502e-01
1.04580954e-01 -2.11179014e-02 -1.08107829e+00 6.67731881e-01
-1.75945833e-01 4.20750499e-01 -8.19565356e-02 -8.02316785e-01
9.31639433e-01 6.18975759e-01 8.77590895e-01 1.72105581e-01
1.40427321e-01 7.91223645e-01 3.43543649e-01 -5.02527952e-01
-1.21329352e-02 -6.46681070e-01 2.88828254e-01 5.69983900e-01
-4.39337552e-01 -7.16768205e-01 2.10071534e-01 -8.17185938e-02
1.17174804e+00 -3.58602911e-01 5.73255241e-01 -5.83291471e-01
8.61731052e-01 -2.35036820e-01 3.11993621e-02 1.81787640e-01
-1.98135808e-01 7.84372330e-01 4.41117942e-01 -2.19195902e-01
-1.13622129e+00 -1.47764266e+00 -7.68762231e-01 3.40111971e-01
2.50780225e-01 1.97494894e-01 -8.28924656e-01 -8.82103443e-01
1.31318539e-01 2.64575183e-01 -5.88196695e-01 -2.05168322e-01
-5.45396090e-01 -6.86777890e-01 4.30867851e-01 5.38460195e-01
4.33971733e-01 -6.20100856e-01 -6.26804709e-01 -7.91355595e-03
-1.19039096e-01 -7.63578653e-01 -7.96308398e-01 -1.83067665e-01
-1.33689451e+00 -1.50549042e+00 -1.23194945e+00 -9.22820151e-01
1.01579511e+00 5.45820713e-01 8.25966477e-01 2.61138856e-01
-5.43969512e-01 7.18573213e-01 -1.05874583e-01 -3.52849036e-01
-9.68825161e-01 -1.00587957e-01 1.29791107e-02 4.59350161e-02
1.22875459e-01 -8.89787748e-02 -5.61411440e-01 1.00156236e+00
-1.04093695e+00 -2.47415110e-01 7.14639187e-01 5.59062243e-01
9.78783667e-01 1.27048433e-01 2.03521661e-02 -6.78209841e-01
3.62519920e-01 -4.22608823e-01 -3.71038526e-01 2.39271566e-01
-5.86074412e-01 2.07978696e-01 4.09247369e-01 -4.49687570e-01
-6.44615233e-01 -1.35601595e-01 2.24715143e-01 -1.18814096e-01
8.90508518e-02 8.71405452e-02 1.65861815e-01 -4.70337987e-01
7.00456381e-01 8.45619291e-02 2.40120053e-01 -2.54085541e-01
-2.92075872e-02 5.68497121e-01 2.11484730e-01 7.77844340e-02
1.22588992e+00 8.81499469e-01 6.65550292e-01 -8.32463264e-01
-6.45284414e-01 -7.78023005e-01 -1.15161121e+00 -9.07016546e-02
1.07823610e+00 -6.18329383e-02 -3.62401485e-01 -4.69059311e-02
-6.13001585e-01 -2.67654900e-02 -6.65628374e-01 1.02879882e+00
-6.33570790e-01 7.23473608e-01 -4.22156185e-01 -5.04498780e-01
-1.06506713e-01 -9.25641716e-01 8.97741973e-01 -1.15614995e-01
-9.41251814e-02 -1.60526156e+00 2.53707677e-01 9.34588388e-02
5.18085182e-01 2.96362847e-01 8.94092679e-01 -8.36316228e-01
-5.21734536e-01 -6.61478162e-01 -7.24156797e-02 3.11265647e-01
4.59541053e-01 2.18442380e-01 -3.29567462e-01 -3.76119256e-01
5.21182716e-01 3.36741120e-01 6.09026968e-01 3.72928441e-01
9.77373779e-01 -1.31953180e-01 -7.20668256e-01 3.91318440e-01
1.18519878e+00 3.10734630e-01 4.37109172e-01 1.29954040e-01
6.38516963e-01 3.87083203e-01 5.88559031e-01 -8.83329511e-02
-9.55854356e-02 4.12629545e-01 5.98648429e-01 -5.47826849e-03
-2.17738509e-01 1.36894792e-01 2.42601082e-01 1.03740573e+00
-5.24969995e-01 1.07725382e-01 -1.10832071e+00 3.86504740e-01
-1.39185309e+00 -1.04969990e+00 -5.03569663e-01 2.56054950e+00
5.87013781e-01 3.08388900e-02 1.33435741e-01 1.36733651e-01
8.97794485e-01 -2.70717084e-01 -3.09026659e-01 -1.32880345e-01
2.58955956e-01 -1.06816879e-02 6.99922502e-01 4.50025767e-01
-9.23710704e-01 1.30082116e-01 6.62245846e+00 7.49608040e-01
-1.24230719e+00 1.27225026e-01 2.74055719e-01 2.60508806e-01
-4.29464519e-01 -1.00944713e-01 -6.17312372e-01 1.47407174e-01
5.75171173e-01 -5.07823825e-01 -1.31053373e-01 7.12056637e-01
3.92070003e-02 2.71957330e-02 -1.16996372e+00 6.28011644e-01
-4.02919874e-02 -8.80463421e-01 -1.03086140e-02 3.72030079e-01
6.10957563e-01 1.04570035e-02 1.58310458e-02 -2.41754219e-01
-2.01744542e-01 -1.04798687e+00 2.33235791e-01 5.82247913e-01
9.09953177e-01 -2.51867563e-01 1.01363099e+00 4.23516810e-01
-1.50740457e+00 4.82295781e-01 -5.65141082e-01 5.72322011e-01
-9.08986479e-02 4.80164289e-01 -1.57119274e+00 3.02280813e-01
3.94078523e-01 6.48709178e-01 -7.48418510e-01 1.24502015e+00
2.52582520e-01 4.20877606e-01 -4.02789265e-01 -5.29275775e-01
1.53295293e-01 -6.21308982e-01 8.11375499e-01 8.83275390e-01
9.99252439e-01 -1.18728191e-01 -1.15594256e-03 9.06459928e-01
2.21333697e-01 4.33231324e-01 -6.60235286e-01 -7.35843331e-02
1.11992963e-01 1.43004298e+00 -1.38944459e+00 -2.21172422e-01
-3.71572495e-01 7.22891390e-01 -3.49878401e-01 -2.23531857e-01
-9.65056479e-01 -2.88584620e-01 -1.54677480e-02 7.67400444e-01
1.69696912e-01 -5.68010285e-02 1.31960794e-01 -1.07204568e+00
4.65054847e-02 -6.95091635e-02 1.43207103e-01 -5.12419581e-01
-9.11106169e-01 6.63136482e-01 1.91131234e-01 -1.56542790e+00
-1.68276176e-01 -8.06927323e-01 -6.85675561e-01 4.77372050e-01
-9.79960322e-01 -7.29608774e-01 -4.58265841e-01 3.83266002e-01
-9.99673153e-04 3.23151471e-03 8.04097414e-01 -1.78955883e-01
2.13373695e-02 6.44962788e-02 6.11049712e-01 1.88039526e-01
5.44504702e-01 -1.67641616e+00 4.60568070e-03 4.31305230e-01
-3.24093513e-02 1.13661051e-01 3.13635916e-01 -6.53298497e-01
-1.07941163e+00 -1.24525666e+00 9.14490700e-01 -6.54603601e-01
8.74134004e-01 6.47233352e-02 -9.81553137e-01 5.03065884e-01
-3.97550881e-01 3.46927375e-01 5.88045418e-01 -7.47615099e-01
-4.05910611e-02 5.69728017e-02 -1.38277686e+00 5.79754829e-01
8.12002778e-01 -2.21065879e-01 -4.84450459e-01 8.11782420e-01
3.10924023e-01 -2.17408121e-01 -1.25946081e+00 6.25900090e-01
5.49890101e-01 -1.04598033e+00 9.52635825e-01 -1.67298779e-01
-2.92427033e-01 -3.38267893e-01 2.42300943e-01 -9.92597699e-01
-6.09809577e-01 -2.86453843e-01 4.11576092e-01 5.37243068e-01
5.45378745e-01 -9.53177214e-01 8.02429855e-01 3.40825766e-01
2.06883579e-01 -9.91460323e-01 -9.01669145e-01 -1.18095005e+00
3.23862761e-01 2.18792539e-02 4.18058664e-01 6.33211613e-01
-1.25360146e-01 -2.46191457e-01 4.65676159e-01 -2.74973035e-01
4.93493468e-01 1.00361730e-03 2.41199985e-01 -1.80311489e+00
-2.39201456e-01 -7.92455494e-01 -8.18136573e-01 -6.19483411e-01
-3.51183683e-01 -1.45504522e+00 -6.95821941e-02 -1.67621994e+00
4.52853262e-01 -5.43374777e-01 -8.67959037e-02 8.83273855e-02
1.50475189e-01 3.12767684e-01 1.60570594e-03 6.89959884e-01
-1.80227026e-01 4.03955914e-02 1.86818290e+00 6.41185939e-02
-5.07878326e-02 6.78117931e-01 -2.34866768e-01 1.08180642e+00
8.83654177e-01 -5.13760448e-01 -3.54656070e-01 9.48961750e-02
4.42789197e-02 4.26628347e-03 3.29005152e-01 -9.44980562e-01
-4.70210128e-02 -2.37756670e-01 4.64208750e-03 -6.16221666e-01
-7.06293285e-02 -1.20154607e+00 4.57137287e-01 1.19835687e+00
-6.59880266e-02 -2.12375358e-01 -3.04090112e-01 6.15512609e-01
-7.16735199e-02 -8.70510161e-01 1.25517797e+00 -3.26609135e-01
2.42330991e-02 5.89202762e-01 -6.00843728e-01 -3.54296044e-02
1.60350525e+00 -5.56655049e-01 3.10463101e-01 -2.19095841e-01
-9.03854787e-01 -4.09528255e-01 8.05975735e-01 -1.78644642e-01
5.40325761e-01 -1.70034719e+00 -6.36369467e-01 1.37399316e-01
-1.19342998e-01 2.28073806e-01 -3.27493012e-01 1.68748915e+00
-1.14494765e+00 6.40440166e-01 1.41047254e-01 -9.40897346e-01
-1.56164443e+00 5.21819174e-01 9.11509514e-01 -5.34390688e-01
-4.71215725e-01 5.14945447e-01 3.26798499e-01 -4.90334362e-01
-1.22452304e-01 -8.37038636e-01 -1.21698543e-01 -1.73963025e-01
1.48864418e-01 6.70659602e-01 1.84400171e-01 -7.94138610e-01
-3.43694240e-01 1.03547668e+00 2.86074966e-01 6.17248416e-02
7.95604587e-01 4.73787747e-02 -4.93360490e-01 5.02816021e-01
1.30089068e+00 2.31850132e-01 -6.34184420e-01 -2.24781651e-02
1.61571175e-01 -4.51239645e-01 -2.89204389e-01 -7.30846524e-02
-9.13798809e-01 9.53695655e-01 4.97524202e-01 1.00611567e+00
7.43175507e-01 5.38726687e-01 7.18460679e-01 4.39739406e-01
-3.18859778e-02 -7.46824145e-01 2.78581053e-01 4.34631974e-01
9.36191618e-01 -9.73150373e-01 -1.26851201e-01 -7.87979007e-01
-2.88362056e-01 1.47477138e+00 2.54532248e-01 -1.89785361e-01
1.03753161e+00 2.20510095e-01 -1.46604404e-01 -1.44868006e-03
-2.71231800e-01 -2.31840193e-01 7.14859664e-01 5.22478282e-01
6.53601348e-01 3.07462990e-01 -5.05978763e-01 6.52813017e-02
-2.12585524e-01 -2.56434470e-01 5.53181291e-01 8.53458285e-01
-1.03417253e+00 -1.01011348e+00 -6.47727787e-01 7.29681373e-01
-5.17184496e-01 2.83036053e-01 -5.10437310e-01 1.03140914e+00
-1.91844225e-01 4.02711958e-01 1.89751238e-01 1.01608559e-02
2.97664046e-01 -6.56537786e-02 7.48124361e-01 -7.56831169e-01
-1.45252168e-01 1.60070136e-01 -4.19132620e-01 -9.11584869e-02
-5.88050187e-01 -8.47183645e-01 -1.56117797e+00 -1.48879085e-02
-6.50252998e-01 4.29257065e-01 4.84633952e-01 6.25563622e-01
-1.71007112e-01 2.28550643e-01 7.95420945e-01 -5.69942772e-01
-9.97418582e-01 -7.85737753e-01 -1.02865922e+00 5.69262266e-01
9.29667726e-02 -5.43497324e-01 -7.91366994e-01 -2.88787484e-02] | [15.306270599365234, -2.1898996829986572] |
16170bb9-8f92-471a-b95e-569dea0d6790 | sensor-fusion-using-emg-and-vision-for-hand | 1910.11126 | null | https://arxiv.org/abs/1910.11126v1 | https://arxiv.org/pdf/1910.11126v1.pdf | Sensor fusion using EMG and vision for hand gesture classification in mobile applications | The discrimination of human gestures using wearable solutions is extremely important as a supporting technique for assisted living, healthcare of the elderly and neurorehabilitation. This paper presents a mobile electromyography (EMG) analysis framework to be an auxiliary component in physiotherapy sessions or as a feedback for neuroprosthesis calibration. We implemented a framework that allows the integration of multisensors, EMG and visual information, to perform sensor fusion and to improve the accuracy of hand gesture recognition tasks. In particular, we used an event-based camera adapted to run on the limited computational resources of mobile phones. We introduced a new publicly available dataset of sensor fusion for hand gesture recognition recorded from 10 subjects and used it to train the recognition models offline. We compare the online results of the hand gesture recognition using the fusion approach with the individual sensors with an improvement in the accuracy of 13% and 11%, for EMG and vision respectively, reaching 85%. | ['Elisa Donati', 'Melika Payvand', 'Lyes Khacef', 'Gemma Taverni', 'Enea Ceolini'] | 2019-10-19 | null | null | null | null | ['electromyography-emg'] | ['medical'] | [ 3.04853112e-01 -5.51282354e-02 1.38586042e-02 -1.47080645e-01
-8.06369126e-01 -2.29765698e-01 1.47696018e-01 -3.04326952e-01
-9.06671703e-01 7.02594280e-01 3.72654706e-01 1.22246280e-01
-3.74613851e-01 -3.37613612e-01 -4.80787545e-01 -7.76695728e-01
7.66309723e-02 6.35473669e-01 2.14063302e-01 9.36275348e-03
2.61002854e-02 8.14557016e-01 -2.07063580e+00 4.26210135e-01
5.24628937e-01 9.67508674e-01 3.94486040e-01 8.78924847e-01
2.25672275e-01 3.26038450e-02 -5.08298397e-01 2.53666312e-01
8.49675536e-02 -1.63346872e-01 -1.79899782e-01 -1.10907346e-01
2.31069997e-01 -4.83319193e-01 -8.39918107e-02 5.94388843e-01
1.18764603e+00 -4.15327884e-02 4.83843118e-01 -9.42573845e-01
3.42917711e-01 3.09871137e-01 -8.40475336e-02 -9.43224281e-02
7.24275172e-01 2.81575590e-01 1.90895692e-01 -5.46677709e-01
6.94822669e-01 6.68057621e-01 5.66536725e-01 1.07059050e+00
-8.28048944e-01 -5.03889203e-01 -4.19688582e-01 8.11258316e-01
-1.01317847e+00 -3.96182388e-01 4.45460200e-01 -4.97494608e-01
1.15833497e+00 7.48721838e-01 1.12164795e+00 1.65933859e+00
5.19625209e-02 5.25823474e-01 1.23398435e+00 -6.72466159e-01
4.48745608e-01 -1.44654095e-01 3.17177832e-01 1.74767867e-01
4.04331446e-01 2.36118585e-01 -8.02440107e-01 -4.17272337e-02
7.13171184e-01 3.04304987e-01 -4.21450287e-01 -6.08846284e-02
-1.22747552e+00 1.28690097e-02 2.02285990e-01 6.54680490e-01
-1.12290013e+00 2.21830562e-01 6.65460154e-02 2.19486907e-01
-4.27839495e-02 -1.85421839e-01 -4.05011505e-01 -8.66310954e-01
-8.01061928e-01 7.56132230e-02 1.07671642e+00 5.13726056e-01
-1.62634235e-02 -3.03677648e-01 -3.81556749e-01 4.48504776e-01
3.85483801e-01 7.24735379e-01 7.79404104e-01 -7.07631171e-01
4.31663930e-01 7.85228074e-01 9.93763804e-02 -3.46330374e-01
-8.53487611e-01 5.27376533e-02 -4.53661680e-01 6.41339362e-01
4.27318066e-01 -3.14894229e-01 -8.89197767e-01 1.28257608e+00
3.38262767e-01 1.33023098e-01 -2.36962184e-01 1.10070622e+00
5.52410007e-01 -1.62101820e-01 1.62956163e-01 -3.32625210e-01
1.38075495e+00 -3.16144377e-01 -8.71646404e-01 1.18477918e-01
4.39494699e-01 -4.97790992e-01 9.92328823e-01 8.61283600e-01
-7.06246614e-01 -2.62494832e-01 -1.06686628e+00 5.11158764e-01
-2.79061288e-01 4.67031270e-01 4.06356066e-01 9.58142281e-01
-8.29459071e-01 7.79634595e-01 -1.46482706e+00 -7.31168926e-01
2.60258108e-01 1.00429332e+00 -6.79461658e-01 4.01061207e-01
-5.43349147e-01 1.10790265e+00 2.78686404e-01 2.33955875e-01
-3.13709289e-01 -2.56774575e-01 -1.99744076e-01 -9.59405303e-02
-6.23554885e-02 -6.07972026e-01 8.30611348e-01 -5.97841203e-01
-1.88073635e+00 7.66890407e-01 1.21183544e-02 -3.88362795e-01
7.68579900e-01 -6.84861600e-01 -2.04056352e-01 1.97305486e-01
-6.41812921e-01 1.35970309e-01 8.39200020e-01 -6.75554693e-01
-3.45684290e-01 -1.10373294e+00 -5.37078679e-01 1.51201129e-01
-6.49899542e-01 -1.14556015e-01 -7.50196055e-02 -4.31703776e-01
1.79452688e-01 -9.55807447e-01 2.48204872e-01 -1.59414232e-01
1.02015428e-01 -3.33458222e-02 7.34122872e-01 -1.18617773e+00
8.85510623e-01 -1.84332311e+00 7.38887668e-01 5.45270324e-01
-1.50835991e-01 4.28749502e-01 2.77385652e-01 1.86654031e-01
7.40187764e-02 -6.18886292e-01 2.32137442e-02 -1.97782502e-01
-2.75295258e-01 4.13466185e-01 3.13504726e-01 3.39848429e-01
-3.45592439e-01 5.87167978e-01 -4.32553947e-01 -2.16402709e-01
7.25710452e-01 7.70857155e-01 -1.34436116e-01 4.42561179e-01
1.94308773e-01 9.06870484e-01 -2.41263807e-01 8.04103255e-01
3.37464735e-02 4.61715221e-01 3.29691768e-01 -3.50146294e-01
-3.87345590e-02 -7.49107823e-02 -1.51262903e+00 1.92113495e+00
-3.68946344e-01 4.11910266e-01 2.96726286e-01 -5.65213144e-01
7.92130113e-01 8.42222154e-01 1.03685427e+00 -5.07287085e-01
7.10364997e-01 5.58476686e-01 -7.16039687e-02 -1.11350739e+00
-2.13232726e-01 2.16049418e-01 6.11899912e-01 3.94367129e-01
2.99002618e-01 3.22558552e-01 -3.41234207e-01 -5.47817767e-01
1.28588963e+00 5.49761117e-01 2.16727287e-01 1.40629724e-01
5.60297251e-01 -1.41068429e-01 -1.45957515e-01 4.12705928e-01
-3.48283611e-02 4.82929736e-01 -1.97319984e-01 -6.01486377e-02
-6.09495521e-01 -9.52408135e-01 1.20533958e-01 7.11588621e-01
-3.78202647e-01 -1.61956251e-01 -1.00334084e+00 -2.79314548e-01
5.57289533e-02 2.05118611e-01 -3.39272022e-01 -3.68654765e-02
-6.68311954e-01 -4.97252941e-01 4.00357366e-01 9.22457695e-01
2.04051077e-01 -1.29772031e+00 -1.25307596e+00 2.75584906e-01
-5.05078509e-02 -9.29590344e-01 1.73879161e-01 4.29067612e-01
-1.14629209e+00 -1.23969305e+00 -1.05270875e+00 -4.76899177e-01
2.37534776e-01 -3.45943749e-01 9.72661376e-02 -1.41340986e-01
-3.70336860e-01 1.08440268e+00 -6.77510321e-01 -5.73089182e-01
-1.15666866e-01 1.43646002e-01 2.67203212e-01 -1.06337756e-01
4.98603076e-01 -1.05186665e+00 -4.89000857e-01 2.07591519e-01
-4.08130914e-01 -2.56600350e-01 6.97093308e-01 6.19380534e-01
4.54819351e-01 -1.15725660e+00 1.70784026e-01 -1.51015699e-01
8.23189497e-01 -4.23266254e-02 -3.85943055e-01 3.89683306e-01
-6.03124142e-01 1.05178421e-02 -6.37300834e-02 -7.66236484e-01
-7.69111514e-01 6.80802763e-01 -5.50998211e-01 -3.38391572e-01
-3.67796212e-01 1.61077708e-01 -2.37590551e-01 -5.28206408e-01
8.24974358e-01 -1.99045371e-02 3.57319415e-01 -8.76984596e-01
6.43325821e-02 1.44706881e+00 8.65047991e-01 -4.16476935e-01
-6.09011613e-02 2.72750229e-01 1.45464286e-01 -8.73298168e-01
3.53649497e-01 -7.77729571e-01 -9.23316896e-01 -8.39796185e-01
8.13635707e-01 -4.87913013e-01 -1.24230075e+00 7.35177577e-01
-1.23638511e+00 -3.45935762e-01 -1.79014012e-01 1.07159472e+00
-8.70447814e-01 2.74006158e-01 -6.94263801e-02 -1.28453517e+00
-8.15757632e-01 -9.45287526e-01 1.09882128e+00 2.60528117e-01
-4.22170222e-01 -3.48078519e-01 1.85867354e-01 3.28421354e-01
3.34053010e-01 5.37397504e-01 1.59069926e-01 -5.42823851e-01
-1.42527744e-01 -6.15811229e-01 3.14843863e-01 2.91710228e-01
1.34173065e-01 -4.57771927e-01 -1.12842035e+00 -1.20289631e-01
1.39325276e-01 -4.20428813e-02 4.29879665e-01 1.73720658e-01
7.22453237e-01 1.57643810e-01 -3.20938110e-01 2.48299137e-01
1.14278316e+00 2.84757823e-01 9.86940444e-01 3.56809974e-01
6.44463301e-01 4.60052967e-01 4.59231496e-01 5.36421895e-01
-1.40055090e-01 1.13939619e+00 4.82278824e-01 5.41709840e-01
-2.50582725e-01 3.92551512e-01 5.47568321e-01 6.69894636e-01
-1.15087676e+00 1.86189577e-01 -8.82747710e-01 2.24633738e-01
-2.07414246e+00 -7.40106106e-01 -1.79254666e-01 2.55568910e+00
4.08269048e-01 -1.90429851e-01 3.72305930e-01 7.42643476e-01
5.34039557e-01 -6.90686047e-01 -4.58800793e-01 -1.06738292e-01
1.64594799e-01 1.03334630e+00 7.65119612e-01 3.13722491e-01
-6.67284489e-01 4.06961381e-01 5.78870869e+00 1.54712915e-01
-1.32234132e+00 4.55645829e-01 -7.27847338e-01 -3.58399987e-01
6.16729558e-01 -5.64981222e-01 -5.43896496e-01 5.32179356e-01
1.12135875e+00 5.08107662e-01 5.72804451e-01 7.56400228e-01
4.12916213e-01 -3.25758189e-01 -1.07162464e+00 1.43595874e+00
2.30453372e-01 -8.47197175e-01 -3.90391797e-01 6.19980097e-02
-3.51937301e-02 2.26319581e-01 -6.47879183e-01 -2.70815402e-01
-7.40855515e-01 -7.21691370e-01 4.26465541e-01 1.16810465e+00
6.48827672e-01 -2.59034008e-01 8.79724026e-01 4.80735540e-01
-8.92220259e-01 -1.00651912e-01 4.62362617e-01 -2.48552203e-01
2.65205801e-01 -1.52508557e-01 -6.77999258e-01 4.00427878e-01
7.84634173e-01 3.74567866e-01 -3.06101561e-01 1.08521032e+00
-2.54491419e-01 3.05146396e-01 -8.10822248e-01 -4.15579975e-01
-4.81323481e-01 1.01067588e-01 9.09712195e-01 8.40227842e-01
6.40132666e-01 1.17666945e-01 -3.60701084e-01 2.57570028e-01
4.43326980e-01 7.57030025e-02 -2.31545806e-01 6.39432222e-02
1.34025782e-01 1.18920684e+00 -4.42035288e-01 -1.31036965e-02
-7.33374879e-02 1.23323202e+00 -2.82069087e-01 1.66325718e-01
-4.65506405e-01 -4.45593685e-01 3.87846470e-01 2.85241604e-01
9.11662653e-02 -4.95831579e-01 -5.28447032e-01 -1.05658305e+00
7.70220816e-01 -6.62102282e-01 2.31732860e-01 -8.50764036e-01
-6.27865314e-01 2.27212757e-01 -5.16854450e-02 -1.23305225e+00
-8.50061893e-01 -1.26496077e+00 -3.66993368e-01 9.60694730e-01
-6.10254467e-01 -1.02106369e+00 -9.58671391e-01 8.80663931e-01
1.09821290e-01 -1.98790401e-01 1.24346256e+00 4.33719724e-01
-2.14754730e-01 4.26832676e-01 -4.69117761e-02 -2.29894340e-01
5.90555668e-01 -9.60628748e-01 -3.96839887e-01 5.77364087e-01
7.48656467e-02 4.60735410e-01 6.30865633e-01 -7.11020350e-01
-1.67032051e+00 -2.41928995e-01 5.82841992e-01 -5.26327312e-01
2.76134878e-01 3.87730822e-02 -5.78467548e-01 4.45570022e-01
-4.71171178e-02 -3.42719167e-01 6.23856068e-01 -2.74988919e-01
2.37779990e-01 -2.32176557e-01 -1.38722038e+00 2.50239760e-01
1.29973245e+00 -4.22610730e-01 -9.04057622e-01 2.33129486e-01
-2.02067941e-01 -6.03614807e-01 -1.15017557e+00 4.20298994e-01
1.58829439e+00 -7.02223897e-01 7.60775805e-01 -4.18545157e-01
-3.69020760e-01 -4.13201982e-03 -3.17512155e-01 -1.02407312e+00
3.30076426e-01 -4.37122285e-01 -5.81381738e-01 8.50953102e-01
-4.08693142e-02 -6.92368448e-01 9.66191351e-01 6.88277960e-01
8.95552114e-02 -4.79289830e-01 -1.41474926e+00 -6.10551476e-01
-7.21378744e-01 -7.83475995e-01 2.16020927e-01 -5.30194677e-02
4.20168310e-01 -8.63375962e-02 -3.18211317e-01 -7.37521797e-02
5.28321981e-01 -5.38043261e-01 7.05448031e-01 -1.47278106e+00
-4.29344714e-01 -2.18864873e-01 -1.18141174e+00 -3.72111022e-01
-2.48854205e-01 -6.36702418e-01 -5.31005189e-02 -1.54187763e+00
-2.36655489e-01 2.13814214e-01 -5.02120912e-01 6.10711157e-01
4.14867043e-01 3.67375135e-01 1.21165842e-01 2.95964152e-01
1.02805905e-01 6.04369119e-03 7.38180995e-01 6.71915859e-02
-5.95117807e-01 3.77118856e-01 -2.39058770e-02 4.48878825e-01
6.11303806e-01 -4.66591030e-01 7.13255256e-02 -2.06026599e-01
-2.02688798e-02 7.84141347e-02 7.43312061e-01 -1.66810083e+00
4.99032617e-01 3.46560895e-01 5.15866995e-01 -5.91382682e-01
5.13639987e-01 -1.40586710e+00 6.09885693e-01 9.42647696e-01
5.79611473e-02 -2.63828456e-01 3.25999148e-02 2.73024946e-01
1.65219516e-01 2.44060867e-02 2.94577867e-01 1.08975813e-01
-5.64191461e-01 -2.30035752e-01 -4.47186261e-01 -8.14781785e-01
1.16560268e+00 -7.62839079e-01 1.28710717e-01 -1.16120897e-01
-1.45152521e+00 -1.70982867e-01 4.63638604e-02 4.95123148e-01
6.34528935e-01 -1.16101360e+00 -1.91227511e-01 3.69026035e-01
5.40692955e-02 -4.68575984e-01 8.10555518e-02 1.35782361e+00
-4.48423862e-01 3.11736315e-01 -1.10195649e+00 -6.66280210e-01
-1.81828558e+00 -1.45588472e-01 3.18813860e-01 2.43740469e-01
-5.83699048e-01 3.04421544e-01 -1.30073571e+00 -2.76973341e-02
9.16678846e-01 -5.38103640e-01 -5.54278374e-01 1.60748750e-01
6.39803767e-01 8.46820951e-01 6.97211444e-01 -4.52560425e-01
-7.03568101e-01 7.86788702e-01 5.65668225e-01 -6.00792408e-01
1.60147274e+00 1.35083586e-01 1.72773480e-01 6.39787495e-01
6.30645990e-01 -2.63905048e-01 -9.03129876e-01 2.75148511e-01
1.68659806e-01 -1.55521100e-02 4.76135351e-02 -1.21739650e+00
-7.31517255e-01 9.16118622e-01 1.70965087e+00 -3.47579211e-01
1.37849212e+00 -2.43605286e-01 5.27828872e-01 7.31851101e-01
8.67512286e-01 -1.10235810e+00 -3.78120750e-01 -2.70562991e-02
1.10057211e+00 -8.87280822e-01 -6.00391738e-02 -1.97235510e-01
-2.68490642e-01 1.50746882e+00 2.01662481e-01 -1.79466903e-01
7.20431387e-01 4.40154731e-01 1.45876314e-02 3.81670706e-02
1.39275659e-02 -5.77234745e-01 4.99274850e-01 8.71113181e-01
3.50620627e-01 2.63178587e-01 -1.17483377e+00 1.04807365e+00
2.70358384e-01 9.37704980e-01 -6.91735372e-02 1.23101008e+00
-2.82621652e-01 -1.32023191e+00 -5.65914214e-01 6.48584366e-01
-2.44252309e-01 4.54987586e-01 -4.97091740e-01 7.22605169e-01
3.60179633e-01 6.89509928e-01 -3.28553885e-01 -9.51390326e-01
7.55406082e-01 4.58971411e-01 1.16879880e+00 -4.10285711e-01
-1.06388927e+00 3.36156110e-03 1.93904430e-04 -9.74877596e-01
-7.64202774e-01 -8.99456799e-01 -1.29630816e+00 2.08242342e-01
-5.64581305e-02 -3.52058500e-01 1.44135153e+00 1.24779749e+00
3.26759875e-01 5.49345195e-01 2.43628602e-02 -1.28589690e+00
-6.44276261e-01 -1.48707068e+00 -5.34472466e-01 2.39268169e-01
-2.92881038e-02 -9.75756705e-01 -9.78662595e-02 7.04183504e-02] | [6.865433692932129, 0.2123619168996811] |
cc738096-b64f-4e15-99f1-b60c19d3db23 | discussion-paper-the-threat-of-real-time | 2306.02487 | null | https://arxiv.org/abs/2306.02487v1 | https://arxiv.org/pdf/2306.02487v1.pdf | Discussion Paper: The Threat of Real Time Deepfakes | Generative deep learning models are able to create realistic audio and video. This technology has been used to impersonate the faces and voices of individuals. These ``deepfakes'' are being used to spread misinformation, enable scams, perform fraud, and blackmail the innocent. The technology continues to advance and today attackers have the ability to generate deepfakes in real-time. This new capability poses a significant threat to society as attackers begin to exploit the technology in advances social engineering attacks. In this paper, we discuss the implications of this emerging threat, identify the challenges with preventing these attacks and suggest a better direction for researching stronger defences. | ['Yisroel Mirsky', 'Guy Frankovits'] | 2023-06-04 | null | null | null | null | ['misinformation'] | ['miscellaneous'] | [ 2.42906548e-02 1.43437311e-01 3.44142973e-01 1.89419866e-01
-4.80892062e-01 -8.83235991e-01 8.64473581e-01 -1.96666002e-01
-3.65380228e-01 7.51443207e-01 2.97033548e-01 -3.92669201e-01
2.44579285e-01 -9.27945316e-01 -3.09712350e-01 -5.35638571e-01
-2.69139290e-01 1.10797761e-02 -1.00906841e-01 -4.04575527e-01
2.49159798e-01 5.28386891e-01 -1.05537331e+00 3.49387914e-01
2.17494681e-01 2.72782832e-01 -6.87151253e-01 1.11729670e+00
4.62155461e-01 8.07357371e-01 -1.46590364e+00 -9.63223755e-01
2.52131999e-01 -5.04579544e-01 -3.64218622e-01 -4.74396408e-01
3.58291119e-01 -8.70423555e-01 -5.16050041e-01 1.10070491e+00
8.65409076e-01 -3.00584376e-01 2.42029384e-01 -1.49700701e+00
-5.68345368e-01 5.81464291e-01 -7.41713941e-01 3.60934258e-01
6.76997006e-01 4.56976295e-01 2.48436138e-01 -2.22634792e-01
2.45071083e-01 1.62293541e+00 1.00273228e+00 1.05818689e+00
-1.06073081e+00 -1.49576247e+00 -5.68074405e-01 -1.29812002e-01
-1.24734986e+00 -5.46374381e-01 6.80423141e-01 -4.26358998e-01
7.93532312e-01 4.10189092e-01 8.89979362e-01 1.83821857e+00
4.21569020e-01 3.44448745e-01 8.04066002e-01 -1.50363922e-01
7.38046318e-02 2.32755080e-01 -2.62417138e-01 4.56391335e-01
6.47052407e-01 7.61214614e-01 -5.47708094e-01 -8.51445973e-01
6.92436874e-01 -2.62212068e-01 1.02770239e-01 6.63615704e-01
-6.62706017e-01 1.29045963e+00 2.48308435e-01 4.73753899e-01
-3.40412110e-01 7.60023236e-01 5.13521075e-01 3.04517269e-01
2.50574291e-01 9.04378891e-01 1.49103642e-01 -7.24984765e-01
-7.56569207e-01 3.78238618e-01 9.93375480e-01 -1.38396909e-02
6.15463182e-02 5.74288309e-01 5.51484704e-01 4.76601809e-01
4.95919615e-01 6.44968867e-01 2.68773645e-01 -1.14900780e+00
1.74394742e-01 3.98758650e-02 -8.37887079e-02 -1.71738362e+00
-1.83277410e-02 -3.85397434e-01 -5.12214363e-01 3.79010350e-01
3.94120991e-01 -9.71318007e-01 -5.18488109e-01 1.38464451e+00
1.34010941e-01 5.70068598e-01 -1.30659774e-01 3.05432320e-01
4.39891964e-01 6.96987808e-01 1.18649900e-02 4.60488588e-01
1.08510208e+00 1.28498420e-01 -6.26924574e-01 -1.83513403e-01
2.73749590e-01 -7.72366285e-01 3.97060782e-01 7.02720642e-01
-8.53125691e-01 -1.39187157e-01 -1.20757580e+00 6.89989865e-01
-4.67920840e-01 -5.63754559e-01 7.20811725e-01 1.89093304e+00
-8.64736140e-01 4.43508089e-01 -7.67835557e-01 -1.52581036e-01
8.69379997e-01 5.07875323e-01 -1.27798721e-01 3.09392452e-01
-1.62499368e+00 7.74132788e-01 -6.98100775e-02 2.13196740e-01
-1.07486355e+00 -5.97701609e-01 -5.64682961e-01 -1.38597354e-01
-2.05445305e-01 -6.35816276e-01 9.82335627e-01 -6.32220924e-01
-1.35011959e+00 7.43716240e-01 7.44551182e-01 -6.91542029e-01
5.64811110e-01 -4.11483258e-01 -8.06520224e-01 1.84994429e-01
-1.44863024e-01 5.81148803e-01 1.18657494e+00 -9.20905530e-01
-3.66222769e-01 -1.59904569e-01 1.13135546e-01 -3.27020735e-01
-6.25063419e-01 6.05269551e-01 8.26708615e-01 -8.08304965e-01
-6.46977186e-01 -1.05254197e+00 -4.23175693e-02 -3.70952696e-01
-6.57302499e-01 2.25260198e-01 1.45927596e+00 -5.98856270e-01
9.40387249e-01 -2.20880389e+00 -4.16813314e-01 2.54030824e-01
5.09648979e-01 8.07605386e-01 1.63498551e-01 8.96570325e-01
1.74207121e-01 7.46704400e-01 3.84755850e-01 -6.05664961e-02
-2.48772800e-02 -1.84379682e-01 -5.86154342e-01 5.78139484e-01
-6.29924685e-02 8.41333687e-01 -7.37685919e-01 5.93527295e-02
2.13635140e-05 9.16771293e-01 -6.55479431e-01 -8.29124004e-02
2.82022834e-01 4.11707163e-01 -4.34268534e-01 4.83752549e-01
4.33272004e-01 2.53354341e-01 1.43085286e-01 4.27451253e-01
1.20221555e-01 2.99556762e-01 -7.70978928e-01 5.33334374e-01
-2.50763029e-01 1.12923789e+00 3.44870180e-01 -5.76670349e-01
8.03268790e-01 2.19569817e-01 1.98589861e-01 1.04960706e-02
7.78823435e-01 1.19649224e-01 2.90291965e-01 -3.13098073e-01
1.89364195e-01 -3.21665823e-01 -5.07197201e-01 8.87225568e-01
-3.62613887e-01 -2.02406451e-01 -4.87447709e-01 3.16546828e-01
1.32155037e+00 -5.06668985e-01 -1.44272953e-01 1.36003673e-01
2.40524456e-01 -3.58841121e-01 6.41515702e-02 8.50134373e-01
-3.49447787e-01 -2.25784183e-02 3.13228816e-01 -5.79832077e-01
-7.49817014e-01 -1.09367752e+00 4.93079931e-01 9.36812699e-01
-2.14561373e-01 -4.66811419e-01 -1.02208674e+00 -8.26205969e-01
2.32769787e-01 7.70001948e-01 -5.67781329e-01 -6.84616268e-01
-5.51872551e-01 -6.70779586e-01 1.80040395e+00 3.65127861e-01
7.68386900e-01 -8.54837954e-01 -8.59539330e-01 2.38532662e-01
7.72330724e-03 -9.85288084e-01 -2.17231646e-01 -5.67057729e-01
-3.14269572e-01 -8.77445936e-01 -5.78507662e-01 -2.64213592e-01
2.80733287e-01 1.40137270e-01 4.98449236e-01 4.20414567e-01
-6.08608544e-01 5.13190091e-01 -8.14952552e-02 -8.56744170e-01
-9.89400029e-01 -2.89254904e-01 2.31145337e-01 9.93917789e-03
5.40781975e-01 -8.63235354e-01 -3.03484917e-01 7.37365484e-02
-8.44564438e-01 -4.66791511e-01 1.46627977e-01 5.44177115e-01
-8.08703899e-01 6.06384873e-01 9.27461267e-01 -9.02064085e-01
1.28898489e+00 -5.74668825e-01 -2.04576060e-01 -3.68476361e-01
-1.04147054e-01 -5.08714616e-01 3.53733718e-01 -7.89815009e-01
-8.38490009e-01 -4.93769884e-01 -3.69140804e-01 -4.53222021e-02
-2.59377271e-01 2.98168302e-01 -9.21300426e-02 -5.78048408e-01
8.50900769e-01 -6.20862022e-02 6.80682734e-02 -3.24453205e-01
2.12473288e-01 7.86579907e-01 4.45897967e-01 -1.93143964e-01
1.33492029e+00 5.93335629e-01 -5.70585504e-02 -1.46655154e+00
-2.48170108e-01 2.14311883e-01 9.23945904e-02 -6.18415236e-01
7.49409974e-01 -5.55003643e-01 -1.15982258e+00 9.96248305e-01
-1.32581699e+00 2.54669338e-02 1.09524064e-01 1.56610876e-01
1.16697267e-01 5.33104539e-01 -1.01977527e+00 -9.82860625e-01
-2.37578660e-01 -6.72794700e-01 6.62660658e-01 3.01631451e-01
-8.67759228e-01 -9.75960553e-01 3.48657161e-01 7.05102146e-01
6.73246980e-01 6.82765603e-01 4.23611224e-01 -9.34746146e-01
-1.81062639e-01 -7.99630463e-01 -2.09698807e-02 5.43169558e-01
6.39283657e-02 2.14917332e-01 -1.04253507e+00 -3.33063662e-01
2.31687605e-01 -2.26094157e-01 3.52526963e-01 8.61537904e-02
5.96723020e-01 -6.97021782e-01 -3.93563092e-01 4.16763753e-01
6.13343477e-01 5.30613482e-01 8.58815491e-01 1.20949700e-01
6.94379508e-01 6.83596313e-01 -1.07558593e-01 4.87756163e-01
-6.26652502e-03 3.00727665e-01 2.43404865e-01 2.64343619e-01
-2.38612220e-02 -6.52493000e-01 5.89370549e-01 4.72064406e-01
-1.00763775e-01 -3.02547336e-01 -9.21168923e-01 -6.29659593e-02
-1.16064239e+00 -1.36655080e+00 -6.62248433e-02 1.90457594e+00
4.24665391e-01 4.89257812e-01 5.06049156e-01 2.80963063e-01
9.56774950e-01 2.09889431e-02 -1.60474449e-01 -8.74277890e-01
2.60985374e-01 4.12540317e-01 5.66346407e-01 7.03291476e-01
-1.07990587e+00 9.71618414e-01 7.22976923e+00 5.01153052e-01
-1.18753660e+00 1.46808103e-01 4.28089052e-01 -6.20050915e-02
-1.25677928e-01 -1.21465728e-01 -7.98982024e-01 8.20738614e-01
1.39408600e+00 -1.96640015e-01 4.06121582e-01 4.02088970e-01
1.44051865e-01 7.21154884e-02 -6.58283114e-01 8.42281282e-01
2.60158300e-01 -1.38302481e+00 -1.94167256e-01 7.71525204e-01
2.96337128e-01 -3.67747486e-01 6.06689692e-01 1.04585513e-01
8.80753100e-01 -1.33769739e+00 1.99708357e-01 6.41715527e-02
6.49369419e-01 -1.12799299e+00 6.22791767e-01 2.74342597e-01
-5.20346999e-01 -3.41657788e-01 -4.77241017e-02 -4.18677837e-01
3.51921588e-01 2.77650058e-01 -1.23514342e+00 -3.56287062e-01
4.28930640e-01 3.01122844e-01 -4.21021461e-01 7.85358787e-01
-4.03239995e-01 9.44105387e-01 -5.09696722e-01 -3.26304942e-01
2.75107652e-01 3.61132622e-01 9.55778837e-01 1.13023329e+00
2.40955070e-01 -1.31874219e-01 -3.25977653e-01 8.18339109e-01
9.05567259e-02 -7.01444626e-01 -1.01657569e+00 -8.64838302e-01
7.42581069e-01 8.08299780e-01 -7.51754105e-01 1.37419969e-01
2.52253532e-01 9.73235846e-01 -5.16912997e-01 1.68340690e-02
-9.75810409e-01 -8.21021259e-01 9.63940918e-01 4.53402251e-01
-2.30740219e-01 -5.25035501e-01 6.02200767e-03 -8.98312747e-01
-8.27412546e-01 -1.40775812e+00 2.96240658e-01 -3.97257537e-01
-1.23005879e+00 4.23602998e-01 -1.30484700e-01 -7.79701293e-01
-6.18307710e-01 -2.75884330e-01 -7.60228455e-01 5.61981559e-01
-5.22778034e-01 -1.16156757e+00 2.11810768e-01 3.79926771e-01
1.09290443e-01 -5.48078537e-01 7.23057091e-01 3.99698257e-01
-3.26234788e-01 7.45699704e-01 -3.84195596e-01 4.82214987e-01
3.83609682e-01 -6.33374155e-01 9.83594835e-01 8.58540297e-01
2.68104941e-01 8.49038064e-01 9.43992496e-01 -9.80093122e-01
-1.30669928e+00 -5.06005466e-01 5.01731336e-01 -6.95264041e-01
8.66955400e-01 -7.82524467e-01 -4.99280632e-01 6.44219875e-01
1.26252517e-01 -7.56919205e-01 1.21247613e+00 -2.33732685e-02
-4.90465462e-01 7.77826533e-02 -1.54060674e+00 7.12199569e-01
5.46829522e-01 -8.14304829e-01 -5.24753451e-01 7.83091858e-02
5.95642865e-01 1.60412893e-01 -2.19210565e-01 -2.06364021e-01
8.96045446e-01 -1.09146988e+00 1.20261121e+00 -6.58857226e-01
1.60065517e-01 2.35300168e-01 1.97342440e-01 -1.19480515e+00
-6.75204769e-02 -1.41160631e+00 -5.44113256e-02 1.30515432e+00
1.81760341e-01 -1.14690495e+00 1.16078651e+00 5.50229847e-01
5.65086722e-01 -7.10429698e-02 -1.02728951e+00 -7.36233175e-01
1.44658446e-01 -6.48723245e-01 6.62944794e-01 1.19024885e+00
2.25001395e-01 2.41731361e-01 -8.33628118e-01 2.97286212e-01
8.60730350e-01 -9.00842428e-01 8.74272227e-01 -1.19441581e+00
-4.84721750e-01 -4.92347658e-01 -9.15141761e-01 -4.44533318e-01
-5.03978953e-02 -4.02752876e-01 -4.83192712e-01 -6.93727493e-01
-2.97788560e-01 3.62765417e-02 1.41964704e-01 2.34876350e-01
3.19711804e-01 7.66947508e-01 5.14398396e-01 -1.14381537e-01
2.07422990e-02 9.49331094e-03 6.60603404e-01 -1.04825780e-01
-1.57010198e-01 2.58267969e-01 -1.21166456e+00 1.04122710e+00
1.06405067e+00 -6.67085588e-01 -3.51909876e-01 8.01693648e-03
6.49752975e-01 -8.27162638e-02 6.83987737e-01 -1.11283565e+00
7.68206641e-02 4.85001579e-02 6.38826549e-01 -2.74113510e-02
6.58102036e-01 -6.87347770e-01 2.66519606e-01 1.08111906e+00
-1.47256166e-01 7.73881078e-02 3.89237583e-01 5.77543378e-01
3.31525058e-01 -2.09563345e-01 9.14398968e-01 -9.41717438e-03
-1.35771483e-02 -2.39295185e-01 -1.15671575e+00 -4.16815989e-02
1.27083969e+00 -2.86167979e-01 -6.85326040e-01 -1.23016274e+00
-5.69887817e-01 -5.59344478e-02 2.83603549e-01 5.60419738e-01
9.61608231e-01 -1.11007094e+00 -7.43582785e-01 4.45652962e-01
-6.99989140e-01 -1.05410147e+00 2.19070166e-01 8.11997652e-02
-8.71190488e-01 1.48427917e-03 -3.25736880e-01 -3.18214446e-02
-1.61683214e+00 2.96592057e-01 3.68847221e-01 2.57605612e-01
-5.04756093e-01 1.43321824e+00 -1.35842443e-01 3.01534608e-02
6.52516931e-02 5.92777848e-01 -1.71670988e-01 -1.00650832e-01
1.02156723e+00 8.54623377e-01 -6.51768565e-01 -6.01502657e-01
-5.55110097e-01 -1.32716060e-01 -2.07734346e-01 -4.44219410e-01
1.19388807e+00 2.47897327e-01 -1.50242284e-01 -8.39743763e-02
1.08638418e+00 5.43897271e-01 -7.70730317e-01 6.35008574e-01
-1.68895632e-01 -8.14931214e-01 -3.64392553e-03 -8.28824341e-01
-9.50429499e-01 1.04169810e+00 2.85416842e-01 9.42855716e-01
5.56730986e-01 -2.03379616e-01 1.35167372e+00 1.14573494e-01
3.19376558e-01 -7.38216817e-01 4.66960192e-01 1.67480081e-01
5.29978096e-01 -3.59009504e-01 -2.28212327e-02 -3.05343300e-01
-4.76812273e-01 9.65395927e-01 3.60949963e-01 -3.64102304e-01
7.03297794e-01 7.34854102e-01 8.22130218e-02 -3.39898169e-01
-3.77210319e-01 6.46424651e-01 -2.66287774e-01 1.25636280e+00
1.43688187e-01 1.76422998e-01 1.54325925e-03 3.51567924e-01
-6.27209961e-01 -1.94290772e-01 1.03729939e+00 7.63841331e-01
-4.57941324e-01 -1.27993488e+00 -7.91910946e-01 2.12464049e-01
-9.94331479e-01 1.31898001e-01 -1.17985308e+00 5.71180344e-01
1.50538832e-01 1.41110754e+00 -3.72218221e-01 -9.94334579e-01
-1.35290205e-01 -1.78140163e-01 3.77269417e-01 -4.75834489e-01
-8.31661999e-01 -3.33524793e-02 4.13156658e-01 -2.28343174e-01
1.45171762e-01 -6.54890180e-01 -6.97865546e-01 -1.09795344e+00
-2.41957575e-01 2.08303377e-01 6.32891893e-01 5.57650685e-01
4.20991689e-01 9.57285687e-02 6.53578877e-01 -7.21846879e-01
-5.81451952e-01 -6.52753294e-01 -5.48006058e-01 -6.22293428e-02
2.26117000e-01 -5.19719422e-01 -6.02479160e-01 -4.28979129e-01] | [6.189533710479736, 7.748828887939453] |
0aa9987a-2d8b-43e6-939a-aa1886b47f87 | acvnet-attention-concatenation-volume-for | 2203.02146 | null | https://arxiv.org/abs/2203.02146v3 | https://arxiv.org/pdf/2203.02146v3.pdf | Attention Concatenation Volume for Accurate and Efficient Stereo Matching | Stereo matching is a fundamental building block for many vision and robotics applications. An informative and concise cost volume representation is vital for stereo matching of high accuracy and efficiency. In this paper, we present a novel cost volume construction method which generates attention weights from correlation clues to suppress redundant information and enhance matching-related information in the concatenation volume. To generate reliable attention weights, we propose multi-level adaptive patch matching to improve the distinctiveness of the matching cost at different disparities even for textureless regions. The proposed cost volume is named attention concatenation volume (ACV) which can be seamlessly embedded into most stereo matching networks, the resulting networks can use a more lightweight aggregation network and meanwhile achieve higher accuracy, e.g. using only 1/25 parameters of the aggregation network can achieve higher accuracy for GwcNet. Furthermore, we design a highly accurate network (ACVNet) based on our ACV, which achieves state-of-the-art performance on several benchmarks. | ['Xin Yang', 'Peng Guo', 'Junda Cheng', 'Gangwei Xu'] | 2022-03-04 | attention-concatenation-volume-for-accurate | http://openaccess.thecvf.com//content/CVPR2022/html/Xu_Attention_Concatenation_Volume_for_Accurate_and_Efficient_Stereo_Matching_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Xu_Attention_Concatenation_Volume_for_Accurate_and_Efficient_Stereo_Matching_CVPR_2022_paper.pdf | cvpr-2022-1 | ['stereo-depth-estimation', 'patch-matching', 'stereo-matching-1'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 2.47165024e-01 -1.96872503e-01 -1.56761277e-02 -2.46103749e-01
-4.58290368e-01 2.10678186e-02 2.99868762e-01 -7.31211603e-02
-3.35804522e-01 4.12657827e-01 3.47199112e-01 3.49634662e-02
-2.80013889e-01 -9.82946098e-01 -6.75971687e-01 -5.86975574e-01
3.86719942e-01 4.90421653e-02 6.12965345e-01 -3.93505931e-01
3.60618174e-01 4.14842546e-01 -1.80119264e+00 4.38902639e-02
1.32223105e+00 1.23409605e+00 5.69422901e-01 -4.31073382e-02
-1.34472817e-01 4.24814433e-01 -2.02877298e-01 -3.29136103e-01
6.65153563e-01 1.91051699e-02 -4.74090189e-01 -1.70534402e-01
8.88937473e-01 -4.32107061e-01 -7.12032676e-01 1.33442247e+00
6.51554167e-01 1.90985426e-01 3.26412618e-01 -1.24139392e+00
-6.12922430e-01 4.80761677e-01 -7.76994407e-01 1.99056268e-01
1.33793861e-01 2.15242684e-01 1.05509138e+00 -8.76975417e-01
5.22522330e-01 1.45748580e+00 7.22133279e-01 2.68744290e-01
-1.07191372e+00 -1.00669789e+00 3.17799747e-01 5.20024896e-01
-1.41376877e+00 -2.29711503e-01 1.06439614e+00 -2.13889375e-01
7.50070333e-01 1.99275970e-01 8.68299067e-01 6.15296960e-01
4.87434044e-02 7.53730297e-01 8.71256888e-01 -1.11864910e-01
-2.23911420e-01 -4.33022976e-01 -1.64029244e-02 6.96343422e-01
3.47674638e-01 2.77137876e-01 -4.23688948e-01 1.14047937e-01
1.27502537e+00 4.10095215e-01 -6.58438385e-01 -3.87258768e-01
-1.38707197e+00 7.75162220e-01 1.09159613e+00 2.12217063e-01
-3.39198709e-01 2.25288078e-01 2.76899844e-01 5.49804419e-02
2.79412270e-01 2.28008598e-01 -1.43978059e-01 2.09057033e-01
-5.59566736e-01 2.51846045e-01 7.30874985e-02 1.07207990e+00
1.07236743e+00 -6.05572052e-02 -6.09994590e-01 1.13186193e+00
1.77414998e-01 4.78492171e-01 3.41831386e-01 -1.16153526e+00
6.75378323e-01 9.53485608e-01 -1.41151577e-01 -1.39664948e+00
-4.64770705e-01 -6.60694122e-01 -1.17964709e+00 1.26206234e-01
-2.85843927e-02 2.01883465e-01 -8.08378041e-01 1.77315784e+00
3.29039097e-01 3.57652396e-01 -2.77370512e-01 1.19392979e+00
1.20639825e+00 6.06015563e-01 -1.62826791e-01 9.03949142e-02
1.25549996e+00 -1.18143821e+00 -4.11833167e-01 -3.24153543e-01
2.15619385e-01 -7.83912599e-01 1.03791773e+00 7.77438581e-02
-1.07124257e+00 -8.80788863e-01 -1.28693628e+00 -4.17977542e-01
4.41717403e-03 -7.12011084e-02 8.58204246e-01 1.29852772e-01
-8.51721883e-01 7.32295990e-01 -2.74357706e-01 -1.27650633e-01
5.99373460e-01 4.78461981e-01 -3.68941635e-01 -4.93361533e-01
-1.13436770e+00 6.63154185e-01 3.62282544e-01 1.03664979e-01
-2.88384169e-01 -7.66049922e-01 -1.08951783e+00 1.26449287e-01
1.37427062e-01 -8.84482682e-01 8.63124847e-01 -7.08663404e-01
-1.13340414e+00 6.81004763e-01 -2.84049690e-01 -1.46951437e-01
4.53065008e-01 8.25475529e-03 -2.06361175e-01 2.43240278e-02
2.67115533e-01 1.07132542e+00 6.75311565e-01 -1.02608633e+00
-7.63755441e-01 -3.89053047e-01 -3.20376009e-02 3.65841329e-01
-3.13227028e-01 -2.53927678e-01 -7.62983978e-01 -9.57200229e-01
5.69413185e-01 -5.57368994e-01 -3.05420011e-01 4.25866902e-01
-1.37093499e-01 -3.12146604e-01 8.46165240e-01 -6.56751633e-01
1.10638535e+00 -2.08258343e+00 1.80545896e-01 3.03703725e-01
5.17187774e-01 3.39736044e-01 -4.02397007e-01 -8.91473293e-02
5.54027744e-02 -2.42386699e-01 -3.06555778e-01 -2.32725948e-01
-1.37466967e-01 1.34276599e-01 -2.99084242e-02 4.16464806e-01
2.51169264e-01 1.08290112e+00 -8.44355345e-01 -5.23451149e-01
4.88090485e-01 5.19780874e-01 -8.19334090e-01 2.23104075e-01
1.52701855e-01 2.06834435e-01 -5.38220108e-01 6.51835740e-01
1.04515886e+00 -1.17218666e-01 -4.31405306e-01 -7.72505224e-01
-2.06533447e-01 -1.98494294e-03 -1.27136660e+00 1.88408065e+00
-4.42817986e-01 7.60397196e-01 9.63960662e-02 -8.88400555e-01
1.41777229e+00 -3.40800434e-01 4.14285958e-01 -1.16684890e+00
3.38012159e-01 2.38793671e-01 -4.35266122e-02 -2.18970016e-01
5.89523852e-01 3.36269021e-01 1.76222309e-01 1.12365410e-01
-7.86306262e-02 6.50252029e-02 -2.24904455e-02 -1.14433318e-01
7.54036307e-01 -1.30676195e-01 1.10339656e-01 -5.01403093e-01
8.95546019e-01 -3.02850783e-01 1.18062198e+00 5.05105913e-01
-2.22754568e-01 8.95548224e-01 1.22645341e-01 -6.86189055e-01
-1.23002768e+00 -7.32295573e-01 -2.29100689e-01 6.73516273e-01
8.70460689e-01 -1.72265604e-01 -5.13406754e-01 -2.49661595e-01
1.18192740e-01 3.46838087e-02 -3.97592217e-01 -3.17538530e-01
-8.09437215e-01 -3.56897801e-01 1.77273825e-01 6.96758986e-01
1.30160379e+00 -1.01641130e+00 -2.47988299e-01 2.50915557e-01
-3.69305670e-01 -1.02539933e+00 -9.04874682e-01 -4.06145275e-01
-6.67184293e-01 -1.03662050e+00 -8.61253142e-01 -1.02915275e+00
7.06776202e-01 7.73446083e-01 9.92792904e-01 4.60466534e-01
-3.04234028e-01 -2.24981517e-01 -1.33561626e-01 -2.38312602e-01
2.53956646e-01 9.23517197e-02 -2.94716209e-01 7.98132047e-02
1.97081223e-01 -7.49475241e-01 -9.76776719e-01 5.14872909e-01
-7.54897714e-01 4.68769550e-01 6.24880016e-01 9.72589970e-01
6.93548441e-01 -3.80988896e-01 2.64621317e-01 -3.83003503e-01
5.29195189e-01 -4.36633453e-02 -8.95517468e-01 8.34219903e-02
-4.85964864e-01 2.07314119e-01 5.94722390e-01 -3.78922373e-01
-7.88906753e-01 -8.79695639e-02 -3.38000298e-01 -6.65826619e-01
3.27469021e-01 1.03565894e-01 -2.22986430e-01 -7.40497231e-01
2.91421115e-01 2.31061429e-01 9.37294308e-03 -4.48640555e-01
2.06945375e-01 4.67361033e-01 7.72840023e-01 -5.50381839e-01
6.88434660e-01 3.94966096e-01 2.77202159e-01 -4.08275455e-01
-5.99095702e-01 -5.45759618e-01 -3.89532447e-01 -2.28242755e-01
5.90837240e-01 -9.92879927e-01 -9.41278160e-01 6.17837846e-01
-1.38164556e+00 -3.25403400e-02 -2.48465613e-02 4.47697997e-01
-3.44968468e-01 5.06002426e-01 -4.70421493e-01 -4.32740927e-01
-6.74986959e-01 -1.38061738e+00 1.12516689e+00 6.02494419e-01
2.39009216e-01 -6.40003622e-01 -1.76646858e-01 2.05573142e-01
5.36222994e-01 2.24400938e-01 7.30484486e-01 -7.12641701e-02
-9.89125669e-01 1.59459144e-01 -9.89544570e-01 1.50170743e-01
-4.05446328e-02 -1.31600752e-01 -8.97357523e-01 -2.13368282e-01
-2.21499756e-01 4.96676452e-02 1.20536971e+00 4.05518651e-01
1.45291674e+00 -1.52703792e-01 -2.75414973e-01 1.19589484e+00
1.60518217e+00 9.66982841e-02 8.77516329e-01 4.81324255e-01
9.29212868e-01 6.24831557e-01 5.93058109e-01 2.40656748e-01
4.44412231e-01 8.18473518e-01 4.98846918e-01 -3.62415463e-01
-3.62372011e-01 -3.28436375e-01 -4.21970822e-02 7.83198237e-01
-1.28004566e-01 1.83735147e-01 -6.26645982e-01 4.96769875e-01
-2.01078296e+00 -9.24567044e-01 -1.86016515e-01 2.31071973e+00
4.53873843e-01 6.66388497e-02 -1.67174086e-01 5.01643196e-02
1.05381489e+00 3.77120197e-01 -5.98271191e-01 -2.07770634e-02
-3.64877313e-01 2.14510694e-01 6.16083920e-01 6.00214720e-01
-9.99428451e-01 8.68308127e-01 5.90829372e+00 1.23148453e+00
-9.30556178e-01 -3.11340336e-02 5.35918593e-01 1.72784835e-01
-6.48308933e-01 -5.24845496e-02 -7.57182539e-01 6.40633583e-01
-1.73184305e-01 -8.43838155e-02 3.71151507e-01 7.79968619e-01
8.89995769e-02 1.03929393e-01 -7.51682162e-01 1.44017792e+00
2.07560677e-02 -1.68622828e+00 3.52988243e-01 -1.05261818e-01
8.59075963e-01 1.05533794e-01 -1.16392665e-01 -3.78777683e-02
3.59422341e-02 -8.39052320e-01 4.33348149e-01 4.26461458e-01
7.83933818e-01 -9.02822018e-01 9.82966542e-01 -1.62131190e-02
-1.66669786e+00 -1.97695270e-01 -8.40486765e-01 8.44811350e-02
9.11170542e-02 6.94889903e-01 9.61647928e-02 7.57204473e-01
7.45938063e-01 1.05259728e+00 -5.23115039e-01 1.56509101e+00
1.02594174e-01 -2.31119573e-01 -3.34987074e-01 -3.41113620e-02
3.32884550e-01 -4.87885296e-01 4.89161462e-01 9.28960443e-01
5.44478655e-01 1.00622602e-01 1.63806468e-01 9.67392743e-01
-1.34088114e-01 1.98394790e-01 -5.17282963e-01 6.27468526e-01
7.93361008e-01 1.22696555e+00 -4.20335591e-01 -2.58611411e-01
-4.85154986e-01 8.43649745e-01 5.42634785e-01 8.79634842e-02
-7.30930567e-01 -6.44229531e-01 9.61573780e-01 -1.71061948e-01
2.86941648e-01 -7.81116134e-04 -3.21202189e-01 -1.20280159e+00
3.33462387e-01 -6.38052166e-01 2.66282737e-01 -9.82003748e-01
-1.29596508e+00 6.80590391e-01 -3.61015886e-01 -1.48051202e+00
3.51046652e-01 -5.28407872e-01 -8.71109724e-01 1.07858205e+00
-1.75371325e+00 -1.12080622e+00 -1.17847645e+00 8.05136442e-01
4.88401383e-01 -8.01439732e-02 2.31186867e-01 6.38526618e-01
-5.64682961e-01 6.80798054e-01 -1.89482868e-01 1.08175008e-02
6.15654826e-01 -7.57356048e-01 5.57674110e-01 7.60000646e-01
-2.29104713e-01 3.83828491e-01 2.57136524e-01 -4.68245596e-01
-1.16710174e+00 -1.27031517e+00 6.52951300e-01 1.80334389e-01
2.17397094e-01 -1.40273079e-01 -1.00767946e+00 1.46164536e-01
1.04454495e-02 3.06818914e-02 1.89319847e-03 -2.27930561e-01
-5.53182065e-01 -6.40570045e-01 -1.21347976e+00 7.64955640e-01
1.66849875e+00 -4.07521397e-01 -5.30266404e-01 9.19843167e-02
1.15562713e+00 -4.92436409e-01 -7.20986903e-01 8.48690271e-01
5.93539238e-01 -1.37551951e+00 1.20034480e+00 -8.33275318e-02
4.86504108e-01 -5.64066231e-01 -1.48509473e-01 -9.68732893e-01
-8.06816280e-01 -4.84908968e-01 2.17837632e-01 1.23263848e+00
9.82128978e-02 -9.06495214e-01 8.40238452e-01 3.78064066e-01
-4.10717428e-01 -9.65824425e-01 -8.97818446e-01 -6.20868564e-01
-2.94584125e-01 -2.03183755e-01 9.93650734e-01 9.14588869e-01
-4.16348130e-01 5.41940480e-02 -2.58717835e-01 -2.94784512e-02
9.25999761e-01 3.30311745e-01 6.68412805e-01 -1.36813259e+00
6.49112761e-02 -7.96496987e-01 -7.89322734e-01 -1.51708925e+00
1.94591731e-02 -6.84368849e-01 -9.50909257e-02 -1.62146032e+00
3.40911388e-01 -6.65132642e-01 -2.50984341e-01 1.88382789e-01
-3.07295889e-01 4.27165806e-01 3.03903699e-01 2.75217354e-01
-2.34712169e-01 9.04368937e-01 1.62500358e+00 -2.83241689e-01
-5.95712028e-02 -3.46949965e-01 -6.52190804e-01 5.77588081e-01
6.64976776e-01 -1.82421564e-03 -3.16141695e-01 -6.46844447e-01
-4.68584821e-02 -2.24237144e-01 3.72063190e-01 -1.21390605e+00
4.87245440e-01 -1.32958636e-01 3.90343457e-01 -7.50308037e-01
3.49033445e-01 -8.72909784e-01 2.07713261e-01 4.73345846e-01
-7.44629046e-03 2.30350554e-01 1.35025367e-01 4.00288463e-01
-5.13280511e-01 -1.87804978e-02 7.95139551e-01 -1.95904393e-02
-1.11612678e+00 8.36983621e-01 4.31959122e-01 -4.53092129e-04
8.45992267e-01 -4.51948106e-01 -5.41033387e-01 -1.74818605e-01
7.36930147e-02 5.66151738e-01 6.39258444e-01 5.16378045e-01
8.42641592e-01 -1.90479970e+00 -6.89153850e-01 4.17826384e-01
2.41852760e-01 4.00081784e-01 5.58476508e-01 6.21280611e-01
-7.76460290e-01 3.74848038e-01 -6.70750439e-01 -7.50792861e-01
-1.13786924e+00 4.45267111e-01 3.54965597e-01 -2.20676586e-02
-7.29317009e-01 8.62618923e-01 5.05067468e-01 -2.64682561e-01
2.38038838e-01 -3.18111777e-01 -3.97288859e-01 -2.42820859e-01
5.62845886e-01 4.29604501e-01 -9.02346522e-02 -7.14099407e-01
-1.92732722e-01 1.28968096e+00 9.77630839e-02 3.07730168e-01
1.15706289e+00 -7.21753538e-02 -2.43715122e-01 -2.64147252e-01
1.39097667e+00 -2.15198576e-01 -1.51946366e+00 -4.50517833e-01
-5.14345109e-01 -9.57740963e-01 3.13385397e-01 -3.14499103e-02
-1.51360559e+00 7.60823131e-01 5.91665804e-01 -2.08607167e-01
1.26038074e+00 -1.91855997e-01 1.18645740e+00 2.17313796e-01
4.16242659e-01 -9.74351287e-01 1.32260928e-02 3.52857858e-01
1.11121881e+00 -1.30797338e+00 3.99424955e-02 -8.21975768e-01
-3.52470577e-01 1.05241764e+00 1.10258710e+00 -4.35747027e-01
4.65478033e-01 -1.14297912e-01 -2.19965205e-01 -1.51304081e-01
-3.60395938e-01 -4.22727168e-01 5.91477513e-01 7.66972125e-01
1.34908751e-01 -1.98273093e-01 -5.01836836e-01 3.74327868e-01
-9.07814577e-02 -3.81695390e-01 2.35348865e-02 5.46242595e-01
-6.35234416e-01 -9.67535734e-01 -2.31154487e-01 6.43564045e-01
8.26737098e-03 -3.61566991e-01 -1.48085147e-01 5.96124649e-01
3.78091425e-01 6.22591376e-01 3.94439459e-01 -6.61821187e-01
5.60739636e-01 -6.21977687e-01 4.00644183e-01 -1.46665215e-01
-5.18478036e-01 -6.48797825e-02 -6.23019561e-02 -9.18876231e-01
-6.41242325e-01 -2.73773611e-01 -8.93299520e-01 -5.91045737e-01
-3.15388739e-01 -2.40162462e-01 1.92206681e-01 7.47354090e-01
4.69985425e-01 3.82137001e-01 7.73450971e-01 -1.01841879e+00
-2.62582731e-02 -7.48188138e-01 -3.46076936e-01 5.52980125e-01
1.94845855e-01 -9.35649872e-01 -2.00342596e-01 -4.30618763e-01] | [8.853891372680664, -2.226198196411133] |
32fe6e03-04ce-4a76-a39b-b87dc982fd9b | video-object-detection-for-privacy-preserving | 2306.14620 | null | https://arxiv.org/abs/2306.14620v1 | https://arxiv.org/pdf/2306.14620v1.pdf | Video object detection for privacy-preserving patient monitoring in intensive care | Patient monitoring in intensive care units, although assisted by biosensors, needs continuous supervision of staff. To reduce the burden on staff members, IT infrastructures are built to record monitoring data and develop clinical decision support systems. These systems, however, are vulnerable to artifacts (e.g. muscle movement due to ongoing treatment), which are often indistinguishable from real and potentially dangerous signals. Video recordings could facilitate the reliable classification of biosignals using object detection (OD) methods to find sources of unwanted artifacts. Due to privacy restrictions, only blurred videos can be stored, which severely impairs the possibility to detect clinically relevant events such as interventions or changes in patient status with standard OD methods. Hence, new kinds of approaches are necessary that exploit every kind of available information due to the reduced information content of blurred footage and that are at the same time easily implementable within the IT infrastructure of a normal hospital. In this paper, we propose a new method for exploiting information in the temporal succession of video frames. To be efficiently implementable using off-the-shelf object detectors that comply with given hardware constraints, we repurpose the image color channels to account for temporal consistency, leading to an improved detection rate of the object classes. Our method outperforms a standard YOLOv5 baseline model by +1.7% mAP@.5 while also training over ten times faster on our proprietary dataset. We conclude that this approach has shown effectiveness in the preliminary experiments and holds potential for more general video OD in the future. | ['Thilo Stadelmann', 'Emanuela Keller', 'Lukas Tuggener', 'Shufan Huo', 'Marko Seric', 'Daniel Baumann', 'Jens Michael Boss', 'Raphael Emberger'] | 2023-06-26 | null | null | null | null | ['video-object-detection'] | ['computer-vision'] | [ 5.42202830e-01 -7.85099193e-02 6.77222908e-02 -2.48156227e-02
-2.48841465e-01 -4.70161110e-01 -2.03230102e-02 4.45493102e-01
-6.29483461e-01 7.54180789e-01 -2.94625163e-01 -1.18764155e-01
-5.01900539e-02 -4.68924403e-01 -5.24357855e-01 -8.02847862e-01
-3.59707534e-01 -9.90895480e-02 4.45639402e-01 3.44732940e-01
-9.75157470e-02 7.62321472e-01 -1.48154497e+00 6.29022658e-01
4.18358445e-01 1.18249655e+00 3.30066592e-01 7.96115458e-01
1.87730074e-01 7.65380442e-01 -8.15904081e-01 1.09678596e-01
2.84314841e-01 -5.85875273e-01 -2.18524650e-01 1.55356443e-02
3.51239860e-01 -6.30562007e-01 -1.64568990e-01 1.03225040e+00
7.79906094e-01 -2.11564839e-01 8.84131938e-02 -9.18677747e-01
1.45129025e-01 -5.87378591e-02 -3.01748604e-01 4.62189615e-01
3.72588247e-01 2.90162534e-01 1.70653224e-01 -5.05635083e-01
6.61167622e-01 8.28937352e-01 6.51647091e-01 6.46400213e-01
-1.52784729e+00 -6.26247942e-01 -9.92175192e-02 3.44311863e-01
-1.15160799e+00 -6.10918760e-01 5.81369698e-01 -6.31591499e-01
7.03948557e-01 7.22979426e-01 6.19330943e-01 1.26164150e+00
4.45086181e-01 3.80366743e-01 1.10235059e+00 -2.81921774e-01
3.23958158e-01 5.88825524e-01 -1.15283644e-02 5.13706803e-01
6.11091316e-01 -9.37888771e-02 -4.71703500e-01 -1.51814193e-01
8.63966525e-01 3.89175266e-01 -8.53546023e-01 -5.00698984e-01
-1.50316298e+00 3.03588420e-01 9.70006436e-02 5.52674592e-01
-5.01166344e-01 -5.13954274e-02 5.56617618e-01 2.87747920e-01
8.27615783e-02 5.62156081e-01 -2.86847234e-01 -3.02299410e-01
-8.68861437e-01 -2.40913019e-01 5.89394987e-01 9.20008898e-01
2.36923918e-01 -2.45976686e-01 -4.29943651e-01 4.06195730e-01
5.50130829e-02 3.02242756e-01 5.78857601e-01 -1.01196516e+00
3.09624881e-01 4.82962787e-01 4.07110244e-01 -9.25200522e-01
-7.47772694e-01 -3.55083972e-01 -8.87594640e-01 2.92063326e-01
5.54455280e-01 -1.83862373e-01 -6.69555426e-01 1.16815054e+00
2.91241765e-01 3.86164725e-01 -1.12059444e-01 9.58404839e-01
7.44635940e-01 2.91449636e-01 -8.06661546e-02 -6.85462832e-01
1.66211879e+00 -3.55412543e-01 -1.13606107e+00 1.72886923e-01
5.83408117e-01 -6.37690604e-01 7.32534409e-01 6.57496512e-01
-8.45275402e-01 -4.94690984e-01 -1.19310141e+00 4.22189891e-01
-1.69782996e-01 1.34683758e-01 3.16991985e-01 9.19962287e-01
-8.22647393e-01 8.51908326e-01 -1.26744199e+00 -4.37434524e-01
5.27514338e-01 4.81276512e-01 -2.87123561e-01 -3.94277610e-02
-8.70774686e-01 7.53941059e-01 1.69733003e-01 3.21414918e-01
-5.76432168e-01 -7.65495837e-01 -6.15999937e-01 -1.51161268e-01
1.55097142e-01 -4.05171633e-01 9.03169930e-01 -7.92209208e-01
-1.45025921e+00 7.32384562e-01 -6.17838874e-02 -5.29823661e-01
9.84876096e-01 -1.81326494e-01 -4.44892347e-01 6.57947361e-01
-2.26446986e-01 2.38350734e-01 8.74611795e-01 -7.80423939e-01
-4.51762766e-01 -3.95884126e-01 -3.40453461e-02 -2.08879471e-01
-5.05814493e-01 9.65020359e-02 -3.65573913e-01 -6.20379448e-01
-1.35977805e-01 -9.87330973e-01 -1.98574498e-01 3.24736476e-01
-1.29112080e-01 6.17108583e-01 8.40044975e-01 -6.35478914e-01
1.14913225e+00 -2.30871153e+00 -3.30469012e-01 -5.09988051e-03
2.33055472e-01 5.86775482e-01 9.95448977e-02 2.46079862e-01
-2.57604159e-02 -2.64129251e-01 -2.05217134e-02 -1.47013478e-02
-4.46745217e-01 -7.93004557e-02 6.98802099e-02 8.89917076e-01
2.61892825e-02 5.95862389e-01 -8.71787131e-01 -3.88546795e-01
5.47721863e-01 6.06318057e-01 -1.76009312e-01 3.67796659e-01
1.96500704e-01 9.14454520e-01 -3.02998722e-01 5.68059683e-01
6.30797267e-01 -2.16948867e-01 4.05500025e-01 -3.67377013e-01
-2.05786034e-01 9.41907689e-02 -1.24493742e+00 1.64235294e+00
-2.60514230e-01 7.81009436e-01 2.56050199e-01 -1.08464646e+00
7.96750426e-01 8.04091334e-01 9.17524278e-01 -6.60849750e-01
2.14371130e-01 2.12084562e-01 1.12659290e-01 -1.00819147e+00
-7.09089786e-02 -1.57355383e-01 5.59794962e-01 -1.80902332e-02
-2.37795889e-01 4.14915055e-01 8.77973065e-02 -2.82689184e-01
1.40016472e+00 1.34388387e-01 2.62575388e-01 -2.90414125e-01
4.46401536e-01 -6.90733492e-02 5.32440722e-01 8.24556172e-01
-5.74347436e-01 7.07250953e-01 2.71476686e-01 -6.04962230e-01
-6.49788678e-01 -7.71877706e-01 -4.84693140e-01 3.86115521e-01
1.97697192e-01 -1.59806207e-01 -6.71345890e-01 -4.44317937e-01
-7.40090162e-02 3.13394129e-01 -3.43737811e-01 -1.55463398e-01
-6.78226829e-01 -9.63124335e-01 4.19476926e-01 4.56275225e-01
2.33079702e-01 -9.34477031e-01 -1.62024891e+00 7.84571826e-01
-7.23705739e-02 -1.39261496e+00 -1.30214587e-01 2.10114419e-01
-1.20054495e+00 -1.26318097e+00 -1.02770019e+00 -4.36223119e-01
6.27701342e-01 4.48635519e-01 6.20831311e-01 -1.70860171e-01
-8.74015033e-01 5.04598439e-01 -2.20048830e-01 -6.21957898e-01
-3.48067105e-01 -6.80069923e-01 1.06988601e-01 3.05167943e-01
4.03518379e-01 -4.54622775e-01 -1.06851578e+00 3.57243985e-01
-7.70128787e-01 -8.80778730e-02 3.70704114e-01 5.29314160e-01
4.95912343e-01 -1.21602006e-01 3.78489196e-01 -6.73360229e-01
3.30260843e-01 -1.51509538e-01 -7.90946603e-01 -2.50996891e-02
-5.83558321e-01 -1.71190232e-01 6.23991132e-01 -5.53148985e-01
-8.03129375e-01 2.18994513e-01 3.05251271e-01 -6.17976129e-01
-3.72377545e-01 1.13298949e-02 1.64400294e-01 -8.11426863e-02
6.82289183e-01 1.35657499e-02 2.70988345e-01 -4.67474252e-01
-2.99605310e-01 8.26499641e-01 3.50603312e-01 9.01690647e-02
2.67246336e-01 8.47287059e-01 2.47203633e-01 -1.08804190e+00
-2.78211534e-01 -7.91552901e-01 -4.48058575e-01 -3.29935789e-01
9.36296880e-01 -1.09649193e+00 -8.50109160e-01 2.82044530e-01
-1.21442175e+00 -3.00692786e-02 -2.40850657e-01 1.01815569e+00
-3.28510284e-01 4.80386913e-01 -6.50941551e-01 -9.00128484e-01
-3.92357737e-01 -1.07972085e+00 7.93556690e-01 -3.48354168e-02
-3.96363944e-01 -5.87315321e-01 -9.03509408e-02 1.79755136e-01
3.90276909e-01 6.39987230e-01 3.79074961e-01 -3.00479263e-01
-7.34691501e-01 -4.34852362e-01 -2.85585225e-02 3.83505017e-01
6.44210875e-01 -1.51375398e-01 -1.25201714e+00 -5.63103676e-01
5.25898159e-01 3.03054333e-01 5.95874548e-01 6.38559103e-01
1.16686857e+00 -2.95637250e-01 -5.53340197e-01 4.81690556e-01
1.23626161e+00 5.43488324e-01 6.83827162e-01 4.59822804e-01
5.04184484e-01 6.49350822e-01 6.69713259e-01 6.12105548e-01
-2.73235172e-01 8.65066707e-01 4.52563524e-01 -2.25848094e-01
-9.67749655e-02 4.56834048e-01 4.58573341e-01 3.12352747e-01
-3.92011166e-01 3.61245088e-02 -4.10036474e-01 4.17964488e-01
-1.66122234e+00 -1.09337962e+00 -5.04432380e-01 2.65369892e+00
6.37228727e-01 1.81314182e-02 1.66916937e-01 2.57148713e-01
7.58960247e-01 -2.46838331e-01 -4.88645166e-01 -2.06787229e-01
2.08787829e-01 1.93960667e-01 6.88273132e-01 -9.58002210e-02
-1.11377048e+00 -7.47501478e-02 5.67106009e+00 -4.81838733e-02
-1.30748439e+00 1.16259202e-01 3.84764165e-01 -5.71965635e-01
4.69203711e-01 -5.48044443e-01 -5.41871250e-01 7.68612802e-01
1.16722131e+00 2.35123977e-01 1.56976700e-01 7.27406025e-01
8.64010394e-01 -1.08023219e-01 -1.28614259e+00 1.35121262e+00
-5.47494590e-02 -1.33382583e+00 -5.37263393e-01 3.58296894e-02
2.26493806e-01 -1.73338324e-01 -2.23050728e-01 -3.12091559e-01
-6.17866158e-01 -6.92737877e-01 5.50698340e-01 4.84690845e-01
7.75201142e-01 -2.99542397e-01 9.19759989e-01 2.49243170e-01
-8.75687897e-01 -6.45992979e-02 -2.68086970e-01 -1.18813980e-02
9.42036808e-02 8.62916350e-01 -9.96724606e-01 3.68685126e-01
8.93281758e-01 6.77928805e-01 -3.32462549e-01 1.34073484e+00
1.33442387e-01 3.43957633e-01 -2.34306604e-01 6.92685768e-02
-1.84994057e-01 8.59978958e-04 6.36645913e-01 1.42385328e+00
6.35722995e-01 1.45516125e-02 -1.93461895e-01 4.69331622e-01
2.52619326e-01 -5.51150218e-02 -6.13531470e-01 9.18787196e-02
8.29601213e-02 1.15196431e+00 -8.86125207e-01 -3.95376444e-01
-6.85212791e-01 1.04348564e+00 -3.69270056e-01 1.95445985e-01
-8.16389978e-01 -3.90217632e-01 6.36979342e-01 5.47422230e-01
3.72349083e-01 -5.60479462e-02 -1.19391210e-01 -1.14007306e+00
3.75474691e-01 -9.67341602e-01 4.07514334e-01 -4.18180674e-01
-7.25709200e-01 5.06361365e-01 -2.15392336e-01 -1.84004748e+00
-5.40898414e-03 -7.89774835e-01 -1.70970023e-01 6.11028314e-01
-1.57883883e+00 -2.57696062e-01 -6.66292250e-01 6.55976832e-01
3.54757577e-01 2.87279487e-01 1.01771402e+00 5.78191042e-01
-5.12343884e-01 2.67427564e-01 2.43505180e-01 -2.75110155e-02
8.78518820e-01 -9.48535740e-01 -2.45480746e-01 7.82704830e-01
3.19754030e-03 5.62040329e-01 7.98169315e-01 -3.74133587e-01
-1.43272293e+00 -1.05619621e+00 5.31019330e-01 -4.10180926e-01
5.13798356e-01 -4.25294161e-01 -8.90349448e-01 4.79149848e-01
3.49767646e-03 3.29596668e-01 7.62793481e-01 -5.04129231e-01
9.58081260e-02 -4.32504475e-01 -1.18739295e+00 3.91289800e-01
8.00186694e-01 -2.92583883e-01 -4.50555205e-01 3.68162304e-01
2.37806424e-01 -4.46864843e-01 -9.35045362e-01 4.56148535e-01
7.55861700e-01 -1.10791516e+00 9.05779541e-01 -3.83519143e-01
-1.58276230e-01 -5.26342750e-01 2.69958049e-01 -8.23289514e-01
-2.02316329e-01 -9.02810633e-01 -1.80766091e-01 8.23373735e-01
3.89061868e-02 -7.99886584e-01 6.26531959e-01 7.55201876e-01
5.77312037e-02 -3.43495339e-01 -1.09497559e+00 -1.04475713e+00
-8.15813899e-01 -3.62535030e-01 1.66929320e-01 8.50630879e-01
4.17338699e-01 -8.82711411e-02 -4.39399391e-01 2.45809942e-01
5.21535337e-01 1.62643805e-01 4.94117707e-01 -1.18100226e+00
-4.31712180e-01 -7.86561817e-02 -8.91716421e-01 -5.76201618e-01
-6.74889445e-01 -5.08462012e-01 -8.40579160e-03 -1.27532053e+00
2.63734255e-03 -2.13414550e-01 -7.21817493e-01 2.45834604e-01
-4.08228189e-02 4.76767808e-01 9.80278328e-02 2.34835312e-01
-3.52395028e-01 -3.50066312e-02 1.13199186e+00 -1.48098588e-01
-5.25880158e-01 1.27471969e-01 -2.76845694e-01 6.44105196e-01
7.73545444e-01 -7.26827562e-01 -2.45159715e-01 -6.96815774e-02
-2.72347987e-01 5.75459711e-02 5.27271867e-01 -1.35501409e+00
-8.11256468e-03 1.01423435e-01 4.81180459e-01 -2.35421896e-01
3.39503467e-01 -1.32135487e+00 5.63012421e-01 1.02611721e+00
-2.95462124e-02 -1.96791634e-01 3.79438043e-01 6.56528890e-01
-1.51454210e-01 -1.27380267e-01 9.48233962e-01 -2.80630440e-01
-2.94743359e-01 -4.23066355e-02 -6.82378173e-01 -3.92202973e-01
1.33948731e+00 -4.81165648e-01 -3.02416801e-01 -1.19085498e-02
-8.22911084e-01 -2.45455354e-01 4.18515444e-01 3.15091819e-01
4.24526721e-01 -9.48929012e-01 -4.08029437e-01 2.41592512e-01
1.53221309e-01 -3.35183203e-01 3.68838161e-01 1.45341742e+00
-7.03941405e-01 5.76787531e-01 -2.95636088e-01 -7.99100995e-01
-1.60096467e+00 8.18163693e-01 3.33604664e-01 3.63635011e-02
-1.08010113e+00 4.33105767e-01 2.21713275e-01 5.88583827e-01
3.37209314e-01 -7.74845183e-01 -2.61363864e-01 1.97615474e-01
1.05152380e+00 5.75572133e-01 5.59660852e-01 -2.08897188e-01
-7.28958964e-01 3.37664336e-01 8.59198272e-02 3.36400867e-01
1.10921073e+00 -1.07981712e-01 1.25637293e-01 3.48868370e-01
1.13815415e+00 -4.40611541e-02 -1.41352403e+00 1.60160542e-01
2.82838792e-02 -4.88129973e-01 -9.22575518e-02 -5.76660752e-01
-1.03568745e+00 9.54302251e-01 1.20031643e+00 3.80616248e-01
1.35685837e+00 -2.90871292e-01 4.32333022e-01 2.81188577e-01
3.93240988e-01 -9.14764702e-01 -1.27644941e-01 -4.18709457e-01
5.99554360e-01 -1.00934637e+00 4.45438251e-02 -5.25184214e-01
-3.70923817e-01 1.32729316e+00 5.50951771e-02 1.97391778e-01
3.70362729e-01 4.22406316e-01 2.91405588e-01 -8.02667961e-02
-4.08262581e-01 -8.26433524e-02 4.69869701e-03 7.05065608e-01
3.99977922e-01 3.98346521e-02 -4.34739411e-01 2.66322404e-01
5.26264906e-01 5.41423082e-01 7.36751139e-01 1.15661705e+00
-2.70264268e-01 -8.94284904e-01 -5.96262872e-01 4.10902739e-01
-7.59668291e-01 1.12205721e-01 1.65682703e-01 6.43730164e-01
2.13756070e-01 1.16724801e+00 -1.40315279e-01 3.96556370e-02
7.90574789e-01 4.36287634e-02 5.53957045e-01 -5.17911434e-01
-4.31532115e-01 2.73395687e-01 6.27419502e-02 -9.68667567e-01
-7.10447133e-01 -7.40598857e-01 -1.09751916e+00 2.88606048e-01
-1.19871818e-01 -1.84316412e-01 5.89896202e-01 5.83598137e-01
7.57740855e-01 8.03015411e-01 3.02065969e-01 -7.22761095e-01
-3.27426612e-01 -8.48750293e-01 -6.63990259e-01 4.89270031e-01
5.77620327e-01 -5.98365784e-01 -2.89860517e-01 5.80014944e-01] | [13.874136924743652, 2.835064649581909] |
8395dd55-d037-497c-a64f-ca725df9a3d2 | skeletor-skeletal-transformers-for-robust | 2104.11712 | null | https://arxiv.org/abs/2104.11712v1 | https://arxiv.org/pdf/2104.11712v1.pdf | Skeletor: Skeletal Transformers for Robust Body-Pose Estimation | Predicting 3D human pose from a single monoscopic video can be highly challenging due to factors such as low resolution, motion blur and occlusion, in addition to the fundamental ambiguity in estimating 3D from 2D. Approaches that directly regress the 3D pose from independent images can be particularly susceptible to these factors and result in jitter, noise and/or inconsistencies in skeletal estimation. Much of which can be overcome if the temporal evolution of the scene and skeleton are taken into account. However, rather than tracking body parts and trying to temporally smooth them, we propose a novel transformer based network that can learn a distribution over both pose and motion in an unsupervised fashion. We call our approach Skeletor. Skeletor overcomes inaccuracies in detection and corrects partial or entire skeleton corruption. Skeletor uses strong priors learn from on 25 million frames to correct skeleton sequences smoothly and consistently. Skeletor can achieve this as it implicitly learns the spatio-temporal context of human motion via a transformer based neural network. Extensive experiments show that Skeletor achieves improved performance on 3D human pose estimation and further provides benefits for downstream tasks such as sign language translation. | ['Richard Bowden', 'Necati Cihan Camgoz', 'Tao Jiang'] | 2021-04-23 | null | null | null | null | ['sign-language-translation'] | ['computer-vision'] | [ 1.59171879e-01 -3.11804354e-01 -4.62949127e-02 -3.28673571e-01
-7.76884019e-01 -4.57532883e-01 4.66563940e-01 -5.99273682e-01
-6.21708632e-01 6.32823050e-01 5.91343522e-01 3.74740720e-01
1.52351111e-01 -1.89653948e-01 -8.50336015e-01 -5.09540856e-01
5.75936362e-02 5.95057249e-01 5.15050173e-01 -3.96103673e-02
-2.70960033e-02 4.27388340e-01 -1.44236076e+00 -8.32436830e-02
4.44022566e-01 6.19594574e-01 -8.65964368e-02 9.58834171e-01
2.69773394e-01 6.47551119e-01 -5.93175530e-01 -2.39246443e-01
5.68275154e-01 -3.76526833e-01 -6.23532951e-01 3.43392134e-01
1.04834604e+00 -6.66743159e-01 -5.28130651e-01 7.81588852e-01
7.05785811e-01 2.34793738e-01 5.54239273e-01 -1.04847789e+00
-3.20569612e-02 1.04409441e-01 -8.93476784e-01 6.05601408e-02
4.58596855e-01 4.41081494e-01 5.69897592e-01 -7.67117620e-01
8.28146160e-01 1.38235044e+00 1.10474002e+00 5.53581059e-01
-1.17027199e+00 -5.72431982e-01 1.18328601e-01 -5.59496619e-02
-1.19125402e+00 -7.56434381e-01 5.79907060e-01 -5.20190954e-01
7.53658056e-01 2.21638903e-02 9.02251303e-01 1.18282413e+00
1.51895091e-01 9.02389228e-01 6.94457352e-01 -3.54424179e-01
-1.27076000e-01 -7.47213721e-01 -2.26076365e-01 7.81466246e-01
7.73588270e-02 1.54847473e-01 -8.08437228e-01 4.99414690e-02
1.22859132e+00 -5.58009110e-02 -3.78267020e-01 -6.74524009e-01
-1.45431328e+00 2.15399653e-01 2.35561714e-01 -1.68667793e-01
-3.80874902e-01 6.57036066e-01 2.84637064e-01 1.10539086e-01
3.12690407e-01 1.98709294e-01 -5.39184570e-01 -3.48252088e-01
-1.05211568e+00 4.21256304e-01 4.00738984e-01 7.96593428e-01
3.58455807e-01 2.08351329e-01 -3.18280719e-02 5.23192525e-01
2.34811813e-01 6.25292480e-01 3.74639839e-01 -1.38427079e+00
4.13336098e-01 7.93606117e-02 2.56110132e-01 -8.48352551e-01
-4.72642183e-01 -3.58150423e-01 -4.90204781e-01 4.70738649e-01
8.68178487e-01 -3.29053164e-01 -1.49301112e+00 1.94432151e+00
5.48509240e-01 2.69965053e-01 -4.52039242e-01 1.31746614e+00
5.29217541e-01 1.30251944e-01 -3.43429819e-02 1.55162588e-01
1.16811192e+00 -8.82893503e-01 -4.63714629e-01 -5.48011959e-01
8.72446001e-02 -8.40978265e-01 7.97962785e-01 3.93662363e-01
-1.34400988e+00 -3.79620165e-01 -8.20737123e-01 -2.33331412e-01
4.24742401e-01 -6.14855364e-02 4.21628743e-01 4.23059434e-01
-8.88261259e-01 5.82111299e-01 -1.42270565e+00 -4.03552264e-01
2.45383859e-01 4.32484150e-01 -6.12614870e-01 -6.36802390e-02
-7.85765469e-01 9.78782952e-01 1.05699711e-02 2.96006471e-01
-5.90959191e-01 -6.30575478e-01 -1.11283278e+00 -3.34474862e-01
5.13113916e-01 -1.25151098e+00 1.42923188e+00 -8.86521876e-01
-1.63706064e+00 8.25236559e-01 -2.96485752e-01 -3.52202684e-01
9.45166051e-01 -7.38961160e-01 1.61458209e-01 2.10845932e-01
2.72875912e-02 7.07840621e-01 1.29871416e+00 -8.50684881e-01
-4.42886055e-01 -4.03186858e-01 -2.47490391e-01 4.60993439e-01
4.43388075e-02 -6.57717660e-02 -9.39135373e-01 -7.99450457e-01
4.78250265e-01 -1.25610292e+00 -2.05631301e-01 5.13450086e-01
-1.78851902e-01 3.10030490e-01 6.58756256e-01 -9.89666522e-01
6.75677001e-01 -1.83176708e+00 5.25486529e-01 2.71870699e-02
1.04957469e-01 2.07108915e-01 -8.45287461e-03 -1.23534761e-01
9.18854550e-02 -3.83356720e-01 -1.39810666e-01 -3.47901255e-01
-2.57056922e-01 4.56166744e-01 1.54694393e-01 7.16431141e-01
1.62691832e-01 9.94189858e-01 -8.82854939e-01 -4.40541267e-01
3.16838145e-01 8.23982358e-01 -6.48180485e-01 -1.14484228e-01
-2.48116359e-01 9.15805697e-01 -1.97513506e-01 6.66354537e-01
2.91349083e-01 -2.52068341e-01 2.70016268e-02 -3.49072814e-01
1.57386720e-01 3.30856621e-01 -1.27470112e+00 2.10476303e+00
-2.31101662e-01 6.73733711e-01 3.46728027e-01 -8.24748039e-01
3.99170190e-01 4.70531195e-01 7.79104590e-01 -4.24371183e-01
1.39250249e-01 8.40225443e-02 -1.35504052e-01 -5.20012677e-01
3.39200705e-01 -3.87546062e-01 9.72677916e-02 4.60039079e-01
-7.31999502e-02 -1.70758590e-01 -1.12977266e-01 -9.17775705e-02
1.19804823e+00 8.08410764e-01 1.41070515e-01 2.61665434e-01
1.48213282e-02 5.10085374e-02 7.99250722e-01 4.46990579e-01
-4.56578225e-01 1.00683498e+00 4.65279333e-02 -5.82478404e-01
-1.20669055e+00 -1.32099402e+00 2.99212933e-01 7.32944727e-01
-6.32153600e-02 -2.23609537e-01 -4.69623744e-01 -5.36210656e-01
2.02974245e-01 1.02395499e-02 -3.13166410e-01 -4.82280478e-02
-1.13215995e+00 -4.99668717e-01 6.10735297e-01 8.38083446e-01
4.07356620e-01 -5.41350365e-01 -9.19604540e-01 1.78866193e-01
-5.08007050e-01 -1.33375967e+00 -9.59814966e-01 -7.79280290e-02
-9.90997553e-01 -9.95388746e-01 -1.06250763e+00 -5.65057755e-01
6.03203177e-01 3.13259602e-01 1.10331571e+00 -1.39011234e-01
-4.45170850e-01 4.88347858e-01 2.29336880e-02 -1.50854930e-01
-1.94568619e-01 -2.91363180e-01 2.73419619e-01 -2.40564734e-01
9.74153280e-02 -7.51129329e-01 -7.25642622e-01 5.28253019e-01
-5.02642810e-01 2.18128815e-01 3.62973571e-01 9.32365239e-01
4.82928455e-01 -1.35845035e-01 -5.31221814e-02 -4.50478166e-01
1.46086916e-01 8.64616707e-02 -5.78011155e-01 -7.93544427e-02
-2.01931804e-01 3.15861970e-01 6.18564487e-02 -5.89959800e-01
-1.11913013e+00 6.89786136e-01 -1.13788851e-01 -7.76334465e-01
-1.70924112e-01 1.76694825e-01 2.03010693e-01 -8.51253141e-03
7.51670063e-01 -2.57004984e-02 4.44825858e-01 -4.95187312e-01
3.15261841e-01 1.70153648e-01 1.12627935e+00 -5.29528618e-01
9.86591101e-01 6.63085401e-01 1.28906116e-01 -8.04897726e-01
-9.22071993e-01 -3.96632761e-01 -9.15572822e-01 -3.75732958e-01
8.77789974e-01 -1.20512736e+00 -5.51476419e-01 6.85627878e-01
-1.08871818e+00 -3.49423528e-01 -1.96073309e-01 8.23179305e-01
-7.39917696e-01 7.34080553e-01 -8.28717947e-01 -5.89247584e-01
-2.40462050e-01 -9.61896896e-01 1.24637139e+00 7.79922120e-03
-6.62791789e-01 -6.92590058e-01 7.50661641e-02 4.54164028e-01
1.39165998e-01 5.99390507e-01 3.70174110e-01 1.71784773e-01
-6.23503745e-01 -2.51961738e-01 2.34982111e-02 2.16528133e-01
3.33571315e-01 -1.32542983e-01 -7.86795437e-01 -4.21124637e-01
-2.36479431e-01 -3.37359518e-01 7.85897434e-01 8.36817920e-01
4.61882800e-01 -1.71673283e-01 -5.94204701e-02 7.64541447e-01
1.02235985e+00 -3.31530511e-01 4.07657385e-01 3.51065844e-01
9.97562647e-01 5.81970096e-01 5.37443578e-01 3.17947477e-01
3.51910949e-01 1.06385815e+00 2.00150713e-01 -1.66845527e-02
-6.20381415e-01 -3.26849908e-01 4.35936987e-01 6.13948286e-01
-5.20969093e-01 2.56915003e-01 -9.02226150e-01 3.95861208e-01
-1.89552259e+00 -1.09949410e+00 -1.22850398e-02 2.32660270e+00
9.38067973e-01 1.66792735e-01 4.24304098e-01 1.48684099e-01
6.61841393e-01 -2.17516571e-02 -7.28276610e-01 1.74724653e-01
-6.77290708e-02 2.26979569e-01 7.15087235e-01 6.27761662e-01
-1.04967654e+00 1.02950311e+00 6.87964058e+00 1.90329194e-01
-1.26640964e+00 -1.25365425e-02 2.77678552e-03 -5.94830513e-01
1.21299483e-01 -1.94422081e-01 -5.77546716e-01 2.04658017e-01
4.45257902e-01 2.57043183e-01 3.00683081e-01 4.81869161e-01
3.06560874e-01 -9.70655382e-02 -1.07151544e+00 1.13011634e+00
1.38796374e-01 -9.13312256e-01 -2.70313442e-01 4.43057623e-03
7.19116747e-01 3.18028301e-01 -1.79064080e-01 -1.85659155e-01
4.24502611e-01 -8.71245444e-01 1.05057991e+00 6.40718460e-01
7.92062283e-01 -5.57965994e-01 3.91054183e-01 4.33425754e-01
-1.12790370e+00 1.91317737e-01 -4.09167632e-02 -2.22335741e-01
5.22070110e-01 4.34135914e-01 -7.90083110e-01 1.70357242e-01
7.14697063e-01 8.16973984e-01 -2.24258482e-01 1.19211531e+00
-4.25602376e-01 4.11431581e-01 -7.95178592e-01 4.64990616e-01
-1.46052256e-01 2.94209123e-02 8.43755245e-01 9.36454892e-01
2.84220189e-01 9.60749164e-02 1.80352375e-01 5.04075050e-01
2.79719293e-01 -3.82526517e-01 -4.19327825e-01 1.90086603e-01
1.50293335e-01 7.70341039e-01 -5.78905761e-01 -2.81446844e-01
-3.09015781e-01 1.28040302e+00 1.72788575e-01 3.52811873e-01
-6.95757449e-01 3.73180658e-02 8.07886541e-01 3.08047563e-01
5.06539941e-01 -7.39037633e-01 -2.62046784e-01 -1.52413857e+00
3.44204456e-01 -9.54272509e-01 3.89751315e-01 -8.24424922e-01
-9.07341599e-01 2.13884383e-01 4.20207977e-02 -1.32934809e+00
-6.16691113e-01 -5.27029872e-01 -2.84122318e-01 6.23929381e-01
-1.19983459e+00 -1.09231627e+00 -3.18067938e-01 7.31856585e-01
6.38774693e-01 2.26121753e-01 4.87563819e-01 3.20323437e-01
-2.09882185e-01 4.55195606e-01 -4.25388128e-01 2.41153315e-01
1.18680060e+00 -1.14011788e+00 7.89685249e-01 9.56866860e-01
3.34132046e-01 5.10185838e-01 9.33092654e-01 -6.71559453e-01
-1.39757466e+00 -7.16511726e-01 8.91039252e-01 -1.00787544e+00
3.43287975e-01 -6.31758273e-02 -5.83701074e-01 8.65497172e-01
-4.15247470e-01 3.09097022e-01 1.71644270e-01 -4.99188416e-02
-3.94778669e-01 4.19983305e-02 -8.77021432e-01 6.74180210e-01
1.40976858e+00 -4.67789412e-01 -6.84535742e-01 2.10689351e-01
3.63490909e-01 -1.04722619e+00 -7.71662474e-01 3.69682193e-01
1.09505689e+00 -1.01560724e+00 1.28039634e+00 -5.40048242e-01
3.23069245e-01 -3.92264456e-01 1.37842074e-01 -1.06724668e+00
-1.48120105e-01 -7.36101151e-01 -2.77631074e-01 6.46635890e-01
3.41806263e-02 -2.28384778e-01 1.20247674e+00 7.92704344e-01
1.28773704e-01 -3.34737569e-01 -1.03779626e+00 -8.00486565e-01
-2.05279633e-01 -5.06007135e-01 2.42964312e-01 6.13645434e-01
-3.14111471e-01 3.29225242e-01 -8.80232334e-01 1.83015525e-01
9.62133944e-01 -6.87746480e-02 1.21437764e+00 -1.18034315e+00
-5.06916702e-01 -1.18896879e-01 -6.54725552e-01 -1.46166205e+00
-7.31940791e-02 -3.31598103e-01 4.33842152e-01 -1.19737065e+00
-3.57136056e-02 -1.16264589e-01 1.34253830e-01 5.14742732e-01
-8.76839682e-02 5.84338784e-01 1.88459292e-01 3.40793431e-01
-3.85879397e-01 2.95129806e-01 1.43264198e+00 7.03968033e-02
-1.68126151e-01 1.35867074e-01 -8.98005739e-02 1.09753156e+00
3.92825246e-01 -4.56078023e-01 -2.51580507e-01 -9.11679387e-01
9.61570814e-02 3.99060696e-01 8.27639580e-01 -1.09411180e+00
2.97740549e-01 -2.42689475e-02 7.87641943e-01 -6.41506970e-01
6.34156048e-01 -6.93786144e-01 3.82813424e-01 5.91593862e-01
-1.49057895e-01 1.47196755e-01 1.48416802e-01 6.64965570e-01
-4.97056358e-02 1.61308944e-01 8.51747155e-01 -4.36771572e-01
-6.06886327e-01 4.09568310e-01 -2.53927052e-01 1.54123083e-01
5.16541302e-01 -5.38368821e-01 1.88312605e-01 -7.33543456e-01
-8.69611859e-01 1.22196019e-01 7.32749462e-01 5.35942674e-01
4.87717420e-01 -1.32799864e+00 -5.54859459e-01 2.56273001e-01
-1.70379505e-01 2.10303336e-01 6.29013255e-02 1.03444242e+00
-6.11647487e-01 2.12091804e-01 -3.55276316e-01 -9.26247001e-01
-1.55663180e+00 1.54631641e-02 4.27051783e-01 3.01487651e-03
-1.12692821e+00 9.23378468e-01 -2.68785120e-03 -3.19604307e-01
3.52010101e-01 -2.24275336e-01 3.40059787e-01 -3.74842703e-01
2.99308687e-01 2.72154182e-01 -1.98639423e-01 -8.77987862e-01
-3.20443153e-01 1.15763116e+00 1.10108286e-01 -4.69800621e-01
1.25883067e+00 -3.44643444e-01 2.61130422e-01 1.38234675e-01
9.51767743e-01 1.28814369e-01 -1.87427938e+00 -4.83453751e-01
-1.95693165e-01 -6.15893722e-01 6.56185746e-02 -6.24254227e-01
-1.06606507e+00 6.97995663e-01 4.10861433e-01 -6.51756585e-01
9.11835194e-01 -2.20840126e-01 1.23004973e+00 2.12744981e-01
4.23138171e-01 -1.10011578e+00 1.58639535e-01 4.86584514e-01
6.62483573e-01 -1.18826389e+00 2.98475593e-01 -2.54776120e-01
-3.78780514e-01 1.08733964e+00 4.59463358e-01 -2.24214450e-01
3.70569259e-01 3.84766012e-01 3.31177533e-01 1.65286642e-02
-4.90794778e-01 -1.43923074e-01 5.24327397e-01 6.59519553e-01
4.12486196e-01 -3.09532017e-01 -4.00609598e-02 -4.62995283e-02
-3.24795336e-01 1.79900661e-01 1.72197253e-01 1.27474272e+00
-2.28453219e-01 -1.07547152e+00 -7.37270594e-01 3.10174795e-03
-4.12194759e-01 2.52101272e-01 -1.86372712e-01 7.80107737e-01
8.90980661e-02 4.49547291e-01 -8.54322221e-03 -2.84769535e-01
3.85249615e-01 7.18247294e-02 9.94204402e-01 -4.32469785e-01
-3.16915572e-01 5.45738697e-01 2.85202622e-01 -7.77276456e-01
-7.16474712e-01 -9.52769816e-01 -1.23974037e+00 -2.03574479e-01
-1.68974534e-01 -3.80169541e-01 5.32442749e-01 1.08604312e+00
1.92195415e-01 2.71898150e-01 -7.70101771e-02 -1.27942502e+00
-6.93677902e-01 -6.43755794e-01 -2.23349482e-01 5.76404929e-01
6.91819847e-01 -8.17447186e-01 -2.56378204e-01 3.94872993e-01] | [7.155967712402344, -0.776003897190094] |
16e64e2b-cdab-4897-a4e0-8d6f418555f2 | point2point-a-framework-for-efficient-deep | 2306.16306 | null | https://arxiv.org/abs/2306.16306v1 | https://arxiv.org/pdf/2306.16306v1.pdf | Point2Point : A Framework for Efficient Deep Learning on Hilbert sorted Point Clouds with applications in Spatio-Temporal Occupancy Prediction | The irregularity and permutation invariance of point cloud data pose challenges for effective learning. Conventional methods for addressing this issue involve converting raw point clouds to intermediate representations such as 3D voxel grids or range images. While such intermediate representations solve the problem of permutation invariance, they can result in significant loss of information. Approaches that do learn on raw point clouds either have trouble in resolving neighborhood relationships between points or are too complicated in their formulation. In this paper, we propose a novel approach to representing point clouds as a locality preserving 1D ordering induced by the Hilbert space-filling curve. We also introduce Point2Point, a neural architecture that can effectively learn on Hilbert-sorted point clouds. We show that Point2Point shows competitive performance on point cloud segmentation and generation tasks. Finally, we show the performance of Point2Point on Spatio-temporal Occupancy prediction from Point clouds. | ['Athrva Atul Pandhare'] | 2023-06-28 | null | null | null | null | ['point-cloud-segmentation'] | ['computer-vision'] | [ 2.09652677e-01 -7.72921368e-02 -2.11143315e-01 -4.98851389e-01
-1.11934972e+00 -6.55988395e-01 5.94395459e-01 3.81513059e-01
-9.44034606e-02 4.92245108e-01 -1.61600575e-01 -5.36830544e-01
-4.17422950e-01 -9.48667705e-01 -1.29127204e+00 -4.34636444e-01
-4.25458878e-01 1.04428220e+00 2.46920004e-01 1.29896536e-01
5.51937938e-01 1.16241550e+00 -1.76147711e+00 3.18649918e-01
7.49494553e-01 7.61715770e-01 1.96514145e-01 5.48544407e-01
-6.72888279e-01 9.59435701e-02 -4.20781553e-01 3.86720598e-01
6.56192005e-01 1.33272454e-01 -8.34662616e-01 8.89400318e-02
1.00359631e+00 -2.30673388e-01 -1.65913090e-01 8.03904891e-01
-4.67767939e-03 3.33775908e-01 7.46063411e-01 -1.57615674e+00
-5.15420258e-01 2.20119476e-01 -7.92607784e-01 2.01111451e-01
-8.97222906e-02 -1.42726656e-02 1.11545050e+00 -1.10079122e+00
3.96887541e-01 1.29899347e+00 8.40978146e-01 1.61631525e-01
-1.41813862e+00 -6.53545678e-01 2.33535543e-01 -6.03097607e-04
-1.59161973e+00 -1.45604417e-01 6.45779252e-01 -6.23727441e-01
1.27071714e+00 6.46443784e-01 8.43717396e-01 4.53546971e-01
-9.08304751e-03 6.58606589e-01 9.11767721e-01 -3.69851999e-02
3.87682945e-01 -4.90292341e-01 1.93890765e-01 4.53498721e-01
3.13469321e-01 8.85723457e-02 -3.32057238e-01 -5.23573160e-01
1.27918160e+00 3.96369159e-01 3.71697242e-03 -9.21706259e-01
-1.25765741e+00 7.49578178e-01 8.60557795e-01 1.30966380e-01
-3.20617378e-01 5.96651495e-01 8.59518722e-02 4.29754034e-02
3.92719567e-01 6.33060753e-01 -4.74009812e-01 1.79267198e-01
-1.10298920e+00 5.72623193e-01 3.98961216e-01 1.20312166e+00
1.10902786e+00 -2.16743425e-01 7.33382627e-02 4.78335559e-01
3.94258797e-01 5.28528869e-01 3.44374403e-03 -1.18414927e+00
4.52275604e-01 4.60188150e-01 1.21509485e-01 -1.00338113e+00
-4.33529019e-01 -2.30556339e-01 -8.17045391e-01 3.83677512e-01
1.38056427e-01 4.28860426e-01 -1.35602248e+00 1.27977872e+00
2.77385056e-01 7.56034613e-01 -2.00493753e-01 8.13185215e-01
5.06810129e-01 9.95006084e-01 4.17015105e-02 2.06154346e-01
9.19147611e-01 -6.64974034e-01 -7.08021373e-02 -1.25020519e-01
3.94376129e-01 -3.64506334e-01 9.24414992e-01 1.79783866e-01
-1.18270147e+00 -5.02340257e-01 -9.84055817e-01 -3.06621522e-01
-3.69728804e-01 -4.98728871e-01 7.99133301e-01 1.17249981e-01
-1.15881944e+00 9.31996822e-01 -1.23000395e+00 -1.12790413e-01
6.61932826e-01 6.85236931e-01 -2.22824752e-01 -1.33618545e-02
-4.98922497e-01 5.47313631e-01 3.26738745e-01 -6.75045624e-02
-4.59445417e-01 -1.27437246e+00 -7.86794126e-01 1.33513123e-01
-8.09505954e-02 -8.07998240e-01 1.15168464e+00 -3.00199121e-01
-7.51710832e-01 8.42301488e-01 -3.96594971e-01 -5.00229359e-01
2.50179470e-01 -1.66161478e-01 1.22439481e-01 -4.51427326e-02
4.53230321e-01 1.10237610e+00 7.73118794e-01 -1.57136941e+00
-1.00487399e+00 -6.13248825e-01 -2.51687378e-01 2.53999025e-01
3.18388015e-01 -6.57049000e-01 -5.40515184e-01 -3.19249779e-01
8.40813994e-01 -1.20538342e+00 -6.48904979e-01 1.93194337e-02
-4.35006052e-01 -4.38532114e-01 1.06405854e+00 -1.96845651e-01
5.37336588e-01 -2.03752923e+00 -1.14937663e-01 6.95756137e-01
2.92661458e-01 -2.41435040e-02 -2.89542317e-01 2.70971477e-01
-2.93755114e-01 4.94130552e-01 -3.93195808e-01 -2.69560188e-01
1.69869438e-01 5.10305882e-01 -7.63365448e-01 5.28962731e-01
4.18592334e-01 1.04971313e+00 -9.15309966e-01 -2.31437251e-01
5.22757173e-01 6.12375915e-01 -7.06076980e-01 -1.65189579e-01
-4.27794546e-01 4.46145236e-01 -4.71062154e-01 6.30364001e-01
1.03936052e+00 -4.94928122e-01 -2.30062127e-01 3.98970954e-03
-1.77931577e-01 3.35288346e-01 -1.04094148e+00 1.82903838e+00
-3.14714551e-01 5.76673865e-01 -1.72183007e-01 -6.07475877e-01
8.82587135e-01 -1.65875241e-01 1.05885208e+00 -5.52580535e-01
-4.86072153e-01 1.92686170e-01 -3.72287273e-01 1.84585139e-01
8.08581233e-01 -1.71798825e-01 -1.48388594e-01 3.71406287e-01
-3.68253678e-01 -7.09725797e-01 -2.84258991e-01 2.52968952e-04
1.13095772e+00 4.29633446e-02 -8.09264109e-02 -2.21477732e-01
-4.37353589e-02 4.98724699e-01 3.11709642e-01 9.70174491e-01
2.22612530e-01 9.33907807e-01 1.98457718e-01 -7.99584210e-01
-1.40104640e+00 -1.52488589e+00 -3.98617983e-01 6.78173840e-01
2.31739506e-01 -3.45849305e-01 -3.39294732e-01 -4.10704523e-01
4.82824266e-01 6.79901063e-01 -2.45787829e-01 1.90340266e-01
-7.92804062e-01 -3.52223337e-01 3.05670023e-01 7.05203176e-01
1.67953238e-01 -7.45045304e-01 -5.10514200e-01 1.54261276e-01
-1.95078328e-02 -9.98554468e-01 -2.05912650e-01 3.20168108e-01
-1.41926157e+00 -9.35447335e-01 -2.96084404e-01 -7.54106700e-01
8.85973811e-01 6.39327049e-01 1.48620915e+00 1.55091450e-01
-1.27433315e-01 3.81188840e-01 -4.61035222e-02 -5.11794209e-01
2.40409188e-02 3.02992672e-01 -3.34248245e-02 -5.16034603e-01
7.38390863e-01 -7.54900515e-01 -5.07949114e-01 1.42277271e-01
-8.48432958e-01 -6.73302412e-02 4.54165399e-01 6.18747294e-01
1.32640707e+00 2.06922516e-01 8.59469771e-02 -7.99757063e-01
2.77753174e-01 -3.98061812e-01 -8.36925387e-01 -2.50940979e-01
-1.67228892e-01 1.36879191e-01 2.42838055e-01 1.85000673e-01
-2.70133764e-01 6.65768445e-01 -1.69239849e-01 -1.09988439e+00
-4.10942167e-01 2.21001968e-01 1.49003014e-01 -2.02201381e-01
5.66611469e-01 9.04500335e-02 -1.17699109e-01 -4.17986125e-01
4.96578246e-01 1.88609600e-01 6.13199770e-01 -7.70404875e-01
1.17394400e+00 8.86601269e-01 2.26736128e-01 -9.09210205e-01
-6.38405144e-01 -7.16058731e-01 -1.00663900e+00 3.10258180e-01
8.53986204e-01 -9.57111061e-01 -5.02669096e-01 -8.36802647e-02
-1.30845523e+00 -1.84169918e-01 -6.82351291e-01 1.59437180e-01
-8.35131347e-01 2.09269911e-01 -3.08551997e-01 -5.54196000e-01
3.32291611e-02 -9.45038557e-01 1.69721401e+00 -1.92139596e-01
-2.38091484e-01 -8.51009130e-01 2.48847216e-01 8.12010020e-02
1.21951541e-02 4.74782914e-01 1.13231254e+00 -4.60009158e-01
-1.49665725e+00 -1.15384080e-01 -2.79626071e-01 -3.87725234e-02
9.85542387e-02 -8.66566822e-02 -8.05350304e-01 -4.24599737e-01
-1.61314666e-01 7.04666786e-03 8.73580217e-01 6.53040946e-01
1.80775487e+00 -1.42134219e-01 -6.41356111e-01 9.91528988e-01
1.48961771e+00 -7.09643215e-02 5.56632519e-01 2.60060668e-01
1.05736399e+00 4.58059281e-01 5.80014348e-01 2.64766455e-01
4.05037105e-01 3.72845441e-01 7.33581781e-01 -2.66981095e-01
1.75097212e-01 -6.24696255e-01 -3.60337764e-01 7.45443583e-01
-1.92111619e-02 4.35729995e-02 -1.47557950e+00 6.73284531e-01
-1.94650006e+00 -8.40646148e-01 -2.88909733e-01 2.20808816e+00
2.09195986e-01 -5.29430881e-02 -1.27228513e-01 -3.56168598e-02
6.48906469e-01 2.38140866e-01 -6.83448017e-01 -1.90443024e-01
2.04861979e-03 2.81292409e-01 9.49917316e-01 5.70501924e-01
-1.23232734e+00 9.25387502e-01 6.98499775e+00 4.89318848e-01
-9.61830497e-01 -1.06710032e-01 4.87728238e-01 -1.72082797e-01
-6.45848393e-01 -1.90257579e-02 -6.59059107e-01 2.66903818e-01
7.43692338e-01 1.09287627e-01 2.80257553e-01 8.64564061e-01
4.41570207e-03 2.05945075e-01 -1.44341290e+00 1.33826983e+00
-3.54580253e-01 -1.73618817e+00 3.57476771e-01 5.23855865e-01
8.55059564e-01 5.79604387e-01 2.68319637e-01 2.12631792e-01
4.70059067e-01 -1.44067037e+00 4.25636560e-01 3.94655555e-01
7.64423609e-01 -9.07075405e-01 2.20747009e-01 4.20113921e-01
-1.30241978e+00 1.92801178e-01 -8.85847092e-01 -8.86073411e-02
1.30966336e-01 3.50194305e-01 -1.23812222e+00 3.27898234e-01
7.98303962e-01 7.41394997e-01 -3.45972300e-01 1.32670200e+00
3.01251173e-01 2.13529155e-01 -7.56656885e-01 3.44494909e-01
4.17715043e-01 -3.35544139e-01 5.29680908e-01 9.39958394e-01
5.45236945e-01 2.33517811e-01 2.84992397e-01 1.08901799e+00
2.62737274e-02 -2.84469783e-01 -1.22474194e+00 8.67343917e-02
7.45283306e-01 8.11875343e-01 -8.84122372e-01 -2.07792476e-01
-2.66563565e-01 5.73601067e-01 4.09602255e-01 2.64427543e-01
-5.79095900e-01 -9.08829421e-02 1.22044420e+00 3.58106971e-01
5.33761501e-01 -7.86864638e-01 -6.70980453e-01 -7.64379323e-01
-5.18364646e-02 -4.64220673e-01 9.45400074e-02 -8.25058520e-01
-1.38645351e+00 3.12877208e-01 1.98506981e-01 -1.33008850e+00
-1.93245262e-01 -6.43617153e-01 -4.87440735e-01 8.10044885e-01
-1.55003166e+00 -9.18204844e-01 -2.86396891e-01 5.05363524e-01
5.25182784e-01 1.91189915e-01 6.38578832e-01 2.09615156e-01
2.93110404e-02 6.83065504e-02 2.03224093e-01 -9.00348574e-02
5.95438182e-02 -1.43102825e+00 1.22647405e+00 4.76521909e-01
5.27343094e-01 7.34284759e-01 4.48668659e-01 -7.26530671e-01
-1.64170086e+00 -1.46911359e+00 5.83118021e-01 -9.79084909e-01
3.61046106e-01 -5.05484641e-01 -1.28493154e+00 9.20701981e-01
-3.50575536e-01 2.46919855e-01 5.54348052e-01 2.42860556e-01
-3.33629072e-01 1.60133541e-01 -1.10171103e+00 5.29095173e-01
1.35242867e+00 -6.30983233e-01 -6.86343074e-01 6.13335133e-01
8.67543697e-01 -6.26539409e-01 -8.54051590e-01 6.09091222e-01
2.95809265e-02 -8.45327973e-01 1.41779399e+00 -7.96702921e-01
2.28709668e-01 -5.51569760e-01 -2.58227706e-01 -1.23373473e+00
-7.56556213e-01 -2.58916885e-01 1.22744508e-01 6.98199034e-01
3.13924015e-01 -4.17481095e-01 1.46431553e+00 6.00268900e-01
-2.49225199e-01 -5.91809511e-01 -1.07301855e+00 -7.46870279e-01
3.50863546e-01 -7.23860145e-01 1.13061345e+00 1.01621521e+00
-5.28686285e-01 1.64298207e-01 1.59272864e-01 6.93855286e-01
8.52239132e-01 4.40672964e-01 9.18952703e-01 -1.53844094e+00
1.73111275e-01 -4.16369557e-01 -6.75896764e-01 -1.31338799e+00
3.08180869e-01 -1.05054724e+00 2.19241709e-01 -1.85935462e+00
-2.14557841e-01 -1.08876431e+00 -2.70563275e-01 4.54152346e-01
2.21641153e-01 3.17924410e-01 2.27049857e-01 6.76361024e-01
-5.18684804e-01 3.72091711e-01 1.12261403e+00 -2.90298641e-01
-3.46352279e-01 -6.64520934e-02 -3.66720170e-01 6.23659611e-01
9.12146747e-01 -4.13286507e-01 -5.61960399e-01 -8.98397624e-01
2.35094458e-01 1.89207867e-02 6.67452693e-01 -1.14986336e+00
3.21376383e-01 -2.67216027e-01 5.99037766e-01 -1.38421631e+00
6.95033610e-01 -1.09049284e+00 3.23779434e-01 1.25312164e-01
-1.00252107e-01 4.60675240e-01 4.27942902e-01 6.73600197e-01
-9.27281007e-02 1.36223868e-01 5.59365213e-01 -3.66908371e-01
-8.10892582e-01 7.46738136e-01 2.66662240e-02 -1.77152559e-01
9.37859178e-01 -4.71427649e-01 -1.97457373e-01 -6.27427697e-02
-5.77447057e-01 4.16635633e-01 8.34820867e-01 5.22438109e-01
8.78812850e-01 -1.39665759e+00 -5.00012815e-01 5.06513953e-01
-6.77609444e-03 7.51669943e-01 1.29843056e-01 3.40650350e-01
-8.80213916e-01 7.28605688e-01 -3.93735379e-01 -1.35206985e+00
-9.25152302e-01 4.89001304e-01 3.61984283e-01 1.28472954e-01
-9.05851364e-01 8.28304946e-01 4.85893339e-01 -8.35669935e-01
1.88295245e-02 -7.54513085e-01 2.55354941e-01 -2.37941310e-01
1.90786734e-01 1.43197641e-01 2.28493094e-01 -6.90476179e-01
-4.01087612e-01 8.53569984e-01 5.98570108e-02 -8.12644735e-02
1.41450787e+00 2.49695137e-01 -3.25304449e-01 5.31979263e-01
1.21201408e+00 -1.90834552e-01 -1.49752319e+00 -1.16963893e-01
2.03243434e-01 -8.97785783e-01 -1.46809965e-02 -1.16896629e-01
-7.63139546e-01 8.73216629e-01 6.33088708e-01 4.32877451e-01
5.96913159e-01 2.73690760e-01 7.72213459e-01 5.80619574e-01
5.91949046e-01 -6.61640406e-01 -2.90182531e-01 7.68841684e-01
7.90141821e-01 -1.22463870e+00 7.81051442e-03 -4.65576261e-01
-1.02452956e-01 8.59945357e-01 5.85053980e-01 -5.63436985e-01
7.12947309e-01 1.69359475e-01 -9.78911147e-02 -5.20718694e-01
-6.74136817e-01 -2.11744025e-01 3.82602334e-01 8.52353156e-01
4.65156846e-02 1.26393467e-01 4.75713581e-01 -1.66519836e-01
-6.72382236e-01 -2.80794293e-01 1.48072451e-01 1.09998989e+00
-6.94494128e-01 -9.28224564e-01 -6.93969131e-01 6.93741560e-01
1.23902477e-01 -7.00028334e-03 -2.33362302e-01 8.28822494e-01
3.81070450e-02 3.56560051e-01 9.65724170e-01 -3.78882498e-01
4.35223550e-01 -5.76514415e-02 4.13937807e-01 -8.82142544e-01
-1.23603754e-01 -8.57109427e-02 -4.48833346e-01 -7.30045199e-01
-1.47246301e-01 -7.94891715e-01 -1.55589449e+00 -3.63524467e-01
1.91384763e-01 3.23625952e-01 9.18508410e-01 5.30788660e-01
6.88846648e-01 4.29341972e-01 4.05232757e-01 -1.48593915e+00
-3.00829709e-01 -3.14063042e-01 -5.13643026e-01 2.70016730e-01
7.46103406e-01 -4.39048231e-01 -1.60138130e-01 -3.09295803e-01] | [8.00989818572998, -3.4988503456115723] |
0c616553-8ce5-45b4-8440-0ef5d1cb56af | hyperspectral-image-analysis-with-subspace | 2304.09730 | null | https://arxiv.org/abs/2304.09730v1 | https://arxiv.org/pdf/2304.09730v1.pdf | Hyperspectral Image Analysis with Subspace Learning-based One-Class Classification | Hyperspectral image (HSI) classification is an important task in many applications, such as environmental monitoring, medical imaging, and land use/land cover (LULC) classification. Due to the significant amount of spectral information from recent HSI sensors, analyzing the acquired images is challenging using traditional Machine Learning (ML) methods. As the number of frequency bands increases, the required number of training samples increases exponentially to achieve a reasonable classification accuracy, also known as the curse of dimensionality. Therefore, separate band selection or dimensionality reduction techniques are often applied before performing any classification task over HSI data. In this study, we investigate recently proposed subspace learning methods for one-class classification (OCC). These methods map high-dimensional data to a lower-dimensional feature space that is optimized for one-class classification. In this way, there is no separate dimensionality reduction or feature selection procedure needed in the proposed classification framework. Moreover, one-class classifiers have the ability to learn a data description from the category of a single class only. Considering the imbalanced labels of the LULC classification problem and rich spectral information (high number of dimensions), the proposed classification approach is well-suited for HSI data. Overall, this is a pioneer study focusing on subspace learning-based one-class classification for HSI data. We analyze the performance of the proposed subspace learning one-class classifiers in the proposed pipeline. Our experiments validate that the proposed approach helps tackle the curse of dimensionality along with the imbalanced nature of HSI data. | ['Moncef Gabbouj', 'Turker Ince', 'Fahad Sohrab', 'Mete Ahishali', 'Sertac Kilickaya'] | 2023-04-19 | null | null | null | null | ['one-class-classification'] | ['miscellaneous'] | [ 7.97424734e-01 -3.81575167e-01 -3.03720742e-01 -2.59291530e-01
-7.26606190e-01 -5.44470251e-01 4.07900929e-01 2.16278851e-01
-2.72214890e-01 7.18364298e-01 -2.20629022e-01 -2.31367454e-01
-4.94542480e-01 -8.02369952e-01 -2.34175831e-01 -1.21615899e+00
1.25199603e-02 2.22160622e-01 -2.46663406e-01 4.69776914e-02
-3.17492485e-02 8.44361424e-01 -2.22874093e+00 2.26673007e-01
1.37391436e+00 1.22760856e+00 2.85844266e-01 3.76623720e-01
-1.43018737e-01 1.81676686e-01 -1.40364215e-01 3.67172837e-01
2.22895801e-01 -4.34023201e-01 -3.29387456e-01 3.99963289e-01
5.03501356e-01 -9.55938101e-02 6.11031987e-02 1.12043667e+00
3.68805587e-01 2.99037367e-01 8.79485786e-01 -1.24732471e+00
1.92251429e-01 6.73112050e-02 -6.06162548e-01 -9.25819948e-02
-1.28425047e-01 -1.58203200e-01 8.08696508e-01 -1.00768864e+00
3.72174084e-01 7.65093088e-01 3.41797888e-01 1.33532986e-01
-1.37193108e+00 -5.07245839e-01 2.10203193e-02 4.34715956e-01
-1.48837578e+00 -4.39004868e-01 1.16018200e+00 -6.80078149e-01
4.69440609e-01 7.22924769e-01 8.33226442e-01 6.30220056e-01
-1.61974028e-01 5.77361703e-01 1.31454182e+00 -4.15746927e-01
5.19199491e-01 3.22051108e-01 3.98446769e-01 1.27795860e-01
4.88698930e-01 -6.59760833e-02 -3.98182213e-01 -9.28847045e-02
-3.62651306e-03 2.12445125e-01 -2.76227176e-01 -8.15509617e-01
-9.51740921e-01 7.78986633e-01 4.33714569e-01 3.63965958e-01
-4.16464239e-01 -6.25154734e-01 3.14921051e-01 7.76299685e-02
5.67320704e-01 1.17353208e-01 -1.71143979e-01 2.80699611e-01
-1.30980456e+00 -4.37764488e-02 4.38141316e-01 5.75217128e-01
1.05081487e+00 7.23231360e-02 7.64527768e-02 8.63968253e-01
7.37755820e-02 7.90402830e-01 3.89230460e-01 -3.35232168e-01
4.13217515e-01 8.92683268e-01 9.23635066e-02 -1.11541104e+00
-7.30178297e-01 -1.00249362e+00 -1.29756248e+00 2.19298258e-01
2.79225171e-01 2.84106731e-01 -5.68974793e-01 1.37396467e+00
4.30659741e-01 -6.22055195e-02 4.14094716e-01 9.06557262e-01
4.90493774e-01 6.39279485e-01 1.11603238e-01 -6.88435495e-01
1.15331781e+00 -5.40250897e-01 -6.16970122e-01 -2.42134869e-01
7.09102511e-01 -4.29480523e-01 1.09441948e+00 4.88787353e-01
-1.37537926e-01 -4.42664921e-01 -1.20888019e+00 3.55787009e-01
-5.03553152e-01 7.71695554e-01 5.74972034e-01 7.95700729e-01
-3.48651737e-01 4.54161078e-01 -7.37985253e-01 -5.58133960e-01
5.73802114e-01 2.59932756e-01 -5.39579690e-01 -4.37151819e-01
-7.64712274e-01 4.44885254e-01 6.73795521e-01 4.69933540e-01
-3.18353534e-01 -5.94727993e-01 -7.18136489e-01 2.02209190e-01
1.91263735e-01 1.49045875e-02 2.49279484e-01 -7.94072092e-01
-9.98315990e-01 7.16809869e-01 -2.60247201e-01 -2.41400793e-01
2.97065824e-01 -7.62628391e-02 -4.79105234e-01 1.99183956e-01
-1.09294690e-01 4.58997712e-02 1.07690728e+00 -1.28288007e+00
-5.40766120e-01 -9.68515098e-01 -5.79796195e-01 3.54712129e-01
-9.65578616e-01 -4.37295347e-01 3.74177188e-01 -4.11851466e-01
8.59268725e-01 -1.02899992e+00 1.08650684e-01 -1.38839379e-01
-3.15420002e-01 6.82779327e-02 1.21495235e+00 -5.88336468e-01
1.19342446e+00 -2.36802268e+00 2.88429379e-01 3.69763404e-01
-4.43531200e-02 2.44003996e-01 -3.00755445e-03 2.50908226e-01
-3.74003172e-01 -1.79857910e-01 -6.56139910e-01 4.43037078e-02
-2.15940341e-01 -7.14228600e-02 -1.70307964e-01 8.56598198e-01
2.12163433e-01 3.84768784e-01 -7.40576625e-01 -3.51648152e-01
4.98887360e-01 4.13814396e-01 -2.23195240e-01 2.31977403e-01
9.40166786e-02 6.76664233e-01 -3.32721382e-01 7.99010634e-01
1.13853264e+00 1.40051199e-02 1.44311249e-01 -8.19963574e-01
-4.16104615e-01 -2.97761381e-01 -1.47020900e+00 1.58903301e+00
-3.60076517e-01 4.63096380e-01 -4.81120683e-03 -1.52956498e+00
1.00645983e+00 1.64710864e-01 6.97463155e-01 -3.58442962e-01
-1.70276850e-01 5.83863437e-01 2.37804763e-02 -5.18947065e-01
-6.02518134e-02 -2.59471685e-01 1.34659752e-01 1.06421068e-01
-9.14257690e-02 -7.65614286e-02 1.43266827e-01 -4.00227249e-01
5.89750886e-01 9.13925096e-03 4.34936017e-01 -4.40497577e-01
9.62309778e-01 2.54485577e-01 4.74113375e-01 3.34622055e-01
-1.65119842e-01 2.51323879e-01 1.56166047e-01 -4.30825561e-01
-9.05375242e-01 -4.86330748e-01 -6.69878840e-01 8.42455268e-01
-6.34587556e-03 5.51602878e-02 -4.84230280e-01 -4.26997006e-01
-6.62382469e-02 7.21931756e-01 -2.82121330e-01 -2.15008229e-01
-1.98655948e-01 -1.25860441e+00 1.89390898e-01 -1.70453712e-02
5.28539062e-01 -7.03910410e-01 -8.22600722e-01 3.74686420e-02
-1.66914001e-01 -9.95777905e-01 3.02852452e-01 4.16543692e-01
-1.18365717e+00 -1.16582894e+00 -6.67119443e-01 -5.12879848e-01
7.21134245e-01 5.63663304e-01 4.96109664e-01 -4.10122633e-01
-4.65187550e-01 1.52176917e-01 -4.20104891e-01 -3.00492376e-01
-2.53530830e-01 3.13211262e-01 1.62388295e-01 5.57993472e-01
4.84208822e-01 -6.22911632e-01 -6.42091811e-01 1.25640437e-01
-1.10699439e+00 1.59070760e-01 7.08958566e-01 9.64399397e-01
6.11610234e-01 7.11590886e-01 6.48763001e-01 -6.08085930e-01
1.12205848e-01 -4.08283442e-01 -7.63268292e-01 3.51146311e-01
-5.85834444e-01 -6.84871748e-02 8.00434768e-01 -1.80251628e-01
-1.02895343e+00 5.46419859e-01 1.80482700e-01 -1.57005787e-01
-6.69672906e-01 7.86595404e-01 -5.18921435e-01 -1.64245188e-01
7.87532866e-01 5.95944405e-01 3.13506909e-02 -5.75003624e-01
-2.09168363e-02 1.00916922e+00 2.47608244e-01 -1.03411660e-01
1.04535902e+00 5.15669644e-01 5.01431525e-01 -1.40080357e+00
-7.35188842e-01 -8.10270488e-01 -1.12170923e+00 -3.69297355e-01
5.19477367e-01 -1.03919137e+00 -3.56079429e-01 5.56287467e-01
-5.62892973e-01 1.90091565e-01 -1.73996404e-01 6.85923338e-01
-3.77350748e-01 4.59949851e-01 1.75443038e-01 -1.25047290e+00
-2.67207563e-01 -1.02548838e+00 9.43543196e-01 2.09830388e-01
1.88566774e-01 -5.66720665e-01 -2.63839781e-01 4.40337837e-01
3.09273511e-01 5.04847229e-01 1.15226722e+00 -5.64104915e-01
-4.34324801e-01 -5.29513419e-01 -2.02958882e-01 2.87135929e-01
4.10486549e-01 -4.04849678e-01 -1.15471995e+00 -5.71336269e-01
2.10576907e-01 -2.02145293e-01 9.45000887e-01 3.59365553e-01
1.20237505e+00 -1.28114119e-01 -2.10792616e-01 6.45076513e-01
1.77620280e+00 1.16997041e-01 2.70184040e-01 3.78313273e-01
7.84658372e-01 9.37090695e-01 1.02397168e+00 6.23730421e-01
-1.90907568e-01 5.83964646e-01 5.96460521e-01 -1.87754124e-01
1.41589999e-01 1.32690147e-01 5.91786839e-02 4.26869869e-01
6.37647808e-02 2.36144718e-02 -1.04460931e+00 3.04347038e-01
-1.75831914e+00 -8.32792282e-01 -2.76208133e-01 2.58809257e+00
3.24309468e-01 -3.51166993e-01 7.42267221e-02 9.31541920e-01
6.12762630e-01 1.52028784e-01 -9.61103082e-01 2.68636495e-01
-4.48297828e-01 -1.38333663e-01 4.54755068e-01 1.41482890e-01
-1.49918544e+00 5.80274045e-01 4.09305906e+00 7.77384758e-01
-1.38600707e+00 -2.16607809e-01 5.56098104e-01 1.54500350e-01
-1.23288482e-02 -1.74047366e-01 -6.80081427e-01 1.37790263e-01
7.41731763e-01 -3.70890722e-02 5.31126797e-01 7.40590036e-01
3.85770649e-01 -3.05554241e-01 -9.33130741e-01 1.28663886e+00
8.57407004e-02 -9.08186018e-01 5.87606467e-02 1.29709005e-01
5.75478733e-01 -2.37630710e-01 8.22006986e-02 -1.86894853e-02
-9.39082265e-01 -8.33355665e-01 3.25468063e-01 2.77516186e-01
1.02056515e+00 -8.71658385e-01 7.27659404e-01 7.05539346e-01
-1.16288149e+00 -5.69665849e-01 -4.36817288e-01 2.46641651e-01
-3.89499158e-01 9.33762372e-01 -4.88012016e-01 7.97347426e-01
5.45340478e-01 1.09820879e+00 -6.93604410e-01 1.17976117e+00
2.59130716e-01 5.04974067e-01 -3.37786824e-01 3.11449498e-01
-5.41332662e-02 -5.69003046e-01 6.86103761e-01 8.49424660e-01
7.24622428e-01 3.16874176e-01 1.92385897e-01 5.39904416e-01
2.71116376e-01 3.70883137e-01 -6.60226762e-01 -1.87492326e-01
1.75052270e-01 1.37942016e+00 -9.11466002e-01 -1.06793493e-01
-2.23511249e-01 7.21308649e-01 -1.93848014e-01 4.20469582e-01
-3.52274388e-01 -2.88187742e-01 4.67042625e-01 9.63725895e-03
1.12228796e-01 -2.71435946e-01 -4.20641750e-01 -1.30514121e+00
1.67991504e-01 -7.98537612e-01 4.21840847e-01 -3.20460707e-01
-9.44906414e-01 5.64664304e-01 2.59485897e-02 -1.84695923e+00
-2.89339498e-02 -6.05927408e-01 -1.22207388e-01 7.31056511e-01
-1.98380518e+00 -1.20897424e+00 -7.31517315e-01 4.00392324e-01
4.52092528e-01 -2.72499323e-01 1.06427944e+00 2.82898515e-01
-8.61696243e-01 2.16729313e-01 7.29613245e-01 -2.41791531e-01
5.75085819e-01 -9.33882535e-01 -7.05816746e-01 9.16522622e-01
-1.77315876e-01 2.05744371e-01 6.42995417e-01 -4.74458158e-01
-1.64086354e+00 -1.22089732e+00 4.09169495e-01 3.17048132e-01
4.07649130e-01 -3.62608641e-01 -9.75420177e-01 -7.15574771e-02
-5.32640398e-01 2.94237733e-01 1.11791635e+00 -1.18861720e-02
-1.80886567e-01 -7.18491375e-01 -1.10907149e+00 1.53782576e-01
6.08057618e-01 -5.48709571e-01 4.18170802e-02 6.11474872e-01
9.94257256e-02 1.09956443e-01 -9.26895797e-01 6.47527218e-01
6.77939713e-01 -9.85030591e-01 8.39270055e-01 -3.02763224e-01
1.66564003e-01 -6.09623969e-01 -4.30302739e-01 -1.23316431e+00
-2.27943197e-01 1.78148538e-01 -1.14388220e-01 9.92777109e-01
6.62859529e-02 -5.29618859e-01 9.10606265e-01 4.89810258e-01
1.48498798e-02 -3.47321928e-01 -1.04436219e+00 -8.17490816e-01
-2.22443730e-01 -1.72255769e-01 5.13027251e-01 9.91539478e-01
-1.85341090e-01 1.35819688e-01 -2.55251378e-01 5.95298052e-01
1.29492521e+00 7.03709781e-01 5.17099023e-01 -1.84988010e+00
5.87540567e-02 -2.50558585e-01 -5.63197553e-01 -1.92199081e-01
2.43470728e-01 -9.61511791e-01 -1.92674533e-01 -1.35758591e+00
2.58640140e-01 -5.89458585e-01 -3.42648298e-01 4.13669735e-01
-7.71893486e-02 2.58121908e-01 1.88074678e-01 5.43891013e-01
-4.36105840e-02 7.46459603e-01 6.63759708e-01 -4.64657485e-01
-5.39640248e-01 1.63439229e-01 -3.46476048e-01 3.68696272e-01
7.86469698e-01 -3.45503390e-01 -4.59458053e-01 -1.00176111e-02
9.98887792e-02 2.13021301e-02 4.43430036e-01 -1.39566529e+00
1.76713228e-01 -1.81844890e-01 5.78421891e-01 -8.87267888e-01
2.80067831e-01 -1.38229346e+00 3.67528558e-01 5.11951208e-01
-9.20086429e-02 -9.10067499e-01 2.56193608e-01 6.58902824e-01
-4.01314646e-01 -7.71440789e-02 9.39895093e-01 1.46861061e-01
-7.87919343e-01 1.33578151e-01 -2.89657354e-01 -6.96838439e-01
1.09385860e+00 -5.09693205e-01 -4.28853929e-02 -3.56534049e-02
-9.31573749e-01 -2.82518677e-02 3.82584810e-01 1.16406254e-01
4.82824296e-01 -1.19363308e+00 -8.00527513e-01 5.07435858e-01
6.42188787e-01 7.32233524e-02 6.80584729e-01 8.86015117e-01
-1.32558957e-01 5.49157560e-01 -3.30847770e-01 -9.88207638e-01
-1.30809283e+00 3.80110472e-01 2.99268872e-01 -6.03563152e-02
-4.05490994e-01 2.95223981e-01 7.60172680e-02 -3.58848184e-01
-1.22757554e-02 7.63374753e-03 -8.14944685e-01 6.94698751e-01
4.40021724e-01 6.27581298e-01 4.63432789e-01 -8.87957573e-01
-3.70629996e-01 7.41398871e-01 3.78800452e-01 2.80727684e-01
1.70526063e+00 -1.58970088e-01 -4.09116477e-01 6.73533738e-01
1.30139041e+00 -4.80719745e-01 -9.52647567e-01 -8.82988274e-02
1.50805652e-01 -4.92625326e-01 4.04046178e-01 -4.90235180e-01
-6.23847127e-01 1.04138649e+00 1.28216732e+00 1.19485959e-01
1.59444654e+00 -3.57214093e-01 1.64353848e-01 5.46563268e-01
3.11570376e-01 -1.11823702e+00 -4.76774722e-01 1.10687986e-01
8.28414083e-01 -1.56525624e+00 2.02760607e-01 -6.85155034e-01
-2.85909057e-01 1.37880552e+00 3.56216133e-01 3.61339957e-01
7.04056799e-01 -3.27599585e-01 -2.19645843e-01 4.75874804e-02
-9.28032026e-02 -3.68114412e-01 5.19839108e-01 5.95849454e-01
3.34731102e-01 2.22544223e-01 -2.83315778e-01 4.43746626e-01
1.54097341e-02 -1.77656889e-01 3.15883815e-01 7.05025136e-01
-5.40317833e-01 -9.84627426e-01 -7.73435473e-01 6.06645763e-01
1.77036300e-01 1.08646579e-01 -7.33655766e-02 4.29696769e-01
3.87419537e-02 1.08886278e+00 -2.27276593e-01 -1.28251791e-01
3.58548820e-01 3.86611193e-01 6.08594455e-02 -4.10159290e-01
7.37268552e-02 2.55582213e-01 -3.12142104e-01 -3.42080295e-01
-7.13629901e-01 -7.57837653e-01 -8.22795093e-01 3.39850307e-01
-4.29094434e-01 1.53018832e-01 1.03009236e+00 9.07146990e-01
2.50097662e-01 2.57153243e-01 1.06152523e+00 -1.01762044e+00
-5.42133391e-01 -7.04968333e-01 -1.07493269e+00 3.54125708e-01
5.06968796e-01 -6.12360179e-01 -6.00706220e-01 -3.51588242e-03] | [10.00133991241455, -1.8804876804351807] |
d5c9b365-d69b-424e-88ac-ccb89a26abe1 | bidirectional-feature-pyramid-network-with | null | null | http://openaccess.thecvf.com/content_ECCV_2018/html/Lei_Zhu_Bi-directional_Feature_Pyramid_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Lei_Zhu_Bi-directional_Feature_Pyramid_ECCV_2018_paper.pdf | Bidirectional Feature Pyramid Network with Recurrent Attention Residual Modules for Shadow Detection | This paper presents a network to detect shadows by exploring and combining global context in deep layers and local context in shallow layers of a deep convolutional neural network (CNN). There are two technical contributions in our network design. First, we formulate the recurrent attention residual (RAR) module to combine the contexts in two adjacent CNN layers and learn an attention map to select a residual and then refine the context features. Second, we develop a bidirectional feature pyramid network (BFPN) to aggregate shadow contexts spanned across different CNN layers by deploying two series of RAR modules in the network to iteratively combine and refine context features: one series to refine context features from deep to shallow layers, and another series from shallow to deep layers. Hence, we can better suppress false detections and enhance shadow details at the same time. We evaluate our network on two common shadow detection benchmark datasets: SBU and UCF. Experimental results show that our network outperforms the best existing method with 34.88% reduction on SBU and 34.57% reduction on UCF for the balance error rate. | ['Xiao-Wei Hu', 'Pheng-Ann Heng', 'Chi-Wing Fu', 'Zijun Deng', 'Lei Zhu', 'Xuemiao Xu', 'Jing Qin'] | 2018-09-01 | null | null | null | eccv-2018-9 | ['shadow-detection'] | ['computer-vision'] | [ 5.75835884e-01 1.73610337e-02 2.35267669e-01 -5.05198598e-01
-5.24072945e-01 2.15848032e-02 2.29333758e-01 -4.59542662e-01
-2.76082996e-02 6.56138480e-01 3.88718575e-01 -3.65540534e-01
3.80201608e-01 -7.10322618e-01 -7.21364260e-01 -7.66627550e-01
-5.74515127e-02 -4.85376596e-01 9.84450936e-01 -2.62838185e-01
1.48779750e-02 5.31106055e-01 -1.38469827e+00 5.23612857e-01
7.50813365e-01 1.37653971e+00 5.92860639e-01 8.76556695e-01
1.61527291e-01 1.29770923e+00 -8.98615658e-01 8.39940980e-02
-2.28844658e-02 -1.97068036e-01 -3.60886335e-01 -3.11695606e-01
3.88386339e-01 -6.11043394e-01 -7.03706503e-01 6.56376302e-01
7.22626567e-01 5.31077050e-02 1.44532114e-01 -1.06389797e+00
-9.36259031e-01 4.75545168e-01 -7.46428728e-01 4.28864151e-01
-1.48961255e-02 -9.56072286e-02 1.03171504e+00 -1.19517744e+00
3.10741663e-01 1.46554077e+00 9.03036535e-01 3.14016372e-01
-9.59244430e-01 -7.78726161e-01 7.23950148e-01 3.23104322e-01
-1.26141560e+00 -1.24621287e-01 7.53545821e-01 2.27134779e-01
8.49140167e-01 3.78856957e-01 5.77368081e-01 1.30584335e+00
3.52694601e-01 1.11587203e+00 9.47685719e-01 -2.13928353e-02
-8.05416256e-02 -2.73055702e-01 4.48136359e-01 1.03286004e+00
1.93962261e-01 4.01091650e-02 -6.47562027e-01 -7.67442212e-02
6.26339674e-01 2.28699803e-01 -3.86999249e-01 2.17351183e-01
-9.17168200e-01 6.86028123e-01 1.09476614e+00 9.25179422e-02
-2.44615316e-01 7.10310161e-01 9.79560539e-02 -1.99842110e-01
3.87549043e-01 8.59482959e-02 -3.85512888e-01 5.74106395e-01
-6.98797524e-01 -1.96060594e-02 3.98274660e-01 1.01398003e+00
9.73976314e-01 6.32116944e-02 -9.55760956e-01 8.45140874e-01
2.63850421e-01 8.73701513e-01 -8.61043856e-02 -8.28916907e-01
1.97723225e-01 3.97220761e-01 4.09256332e-02 -1.17903578e+00
-3.21933091e-01 -7.20274210e-01 -8.27650368e-01 1.36860207e-01
-4.46212023e-01 -3.07370871e-01 -1.29566503e+00 1.64307868e+00
-9.45824757e-03 8.34445238e-01 -9.25301388e-02 9.09326077e-01
1.16794574e+00 8.76566887e-01 -1.26009271e-01 9.96541530e-02
1.36934566e+00 -1.32584929e+00 -7.08062351e-01 -5.35012245e-01
7.65009969e-02 -8.59597921e-01 1.09896600e+00 2.20082719e-02
-6.19648337e-01 -8.76982093e-01 -1.45933628e+00 -5.34840167e-01
-3.29974294e-01 5.40218651e-01 6.65789127e-01 2.28416473e-01
-1.18930614e+00 3.44427854e-01 -5.19237876e-01 -5.16599715e-02
7.60493696e-01 2.79040128e-01 4.88674432e-01 -2.13370875e-01
-1.30411792e+00 3.84586990e-01 -5.20617217e-02 6.78334057e-01
-1.27062654e+00 -6.38823211e-01 -8.34885180e-01 4.28960145e-01
5.20852566e-01 -3.55928481e-01 1.16899955e+00 -7.89561689e-01
-1.17440331e+00 1.45643160e-01 -6.17418706e-01 -4.55627441e-01
1.67335287e-01 -6.90444529e-01 -3.16153228e-01 -2.96229422e-01
1.51213840e-01 4.64591116e-01 7.32015073e-01 -1.61609590e+00
-1.03079975e+00 5.01751602e-02 2.33714432e-01 4.99517545e-02
-1.28413752e-01 -2.48930052e-01 -9.07438755e-01 -7.79958546e-01
1.02439985e-01 -8.18877757e-01 -2.78466970e-01 -1.37503773e-01
-5.66382349e-01 -1.45657822e-01 1.75798774e+00 -7.16056585e-01
1.41259420e+00 -2.10381603e+00 -2.35365629e-01 2.53929198e-01
5.17227709e-01 4.11673963e-01 -2.54469007e-01 -1.49268299e-01
1.81460053e-01 -1.54386923e-01 -3.75736862e-01 -6.71732724e-01
-1.78695858e-01 5.03085494e-01 -8.22920561e-01 1.85602605e-01
4.66539949e-01 1.30589271e+00 -9.90823209e-01 -3.07154357e-01
1.44523934e-01 8.53872597e-01 -3.55386466e-01 3.21554363e-01
-1.39381245e-01 1.37160998e-04 -5.44271648e-01 8.69195163e-01
8.70089471e-01 -3.58648419e-01 1.99218869e-01 -4.41315740e-01
-2.18944252e-01 3.67052197e-01 -8.37501347e-01 1.29812634e+00
-8.03595603e-01 1.33958757e+00 -3.18930373e-02 -4.78851259e-01
1.02086675e+00 -2.33955264e-01 5.53160720e-02 -9.39548194e-01
3.20546418e-01 -9.77958087e-03 -2.04151630e-01 -2.46843204e-01
8.06787312e-01 4.92411882e-01 -8.84404853e-02 2.00196937e-01
-1.36640653e-01 5.32078505e-01 -5.09303927e-01 2.71664768e-01
1.41976905e+00 -3.40627208e-02 -2.24987507e-01 -4.26422060e-01
5.94632745e-01 -7.09462821e-01 9.03735936e-01 1.11109543e+00
-2.29925707e-01 7.39003062e-01 6.33520186e-01 -5.84193289e-01
-4.07244325e-01 -1.00594091e+00 1.34995595e-01 1.58954883e+00
6.52625382e-01 -4.88698512e-01 -3.62735450e-01 -8.64460766e-01
-2.55244933e-02 4.30058837e-01 -1.10687470e+00 -2.36078084e-01
-8.18432689e-01 -8.81414890e-01 5.42274415e-01 9.97852802e-01
9.68921244e-01 -1.28492522e+00 -8.29791665e-01 -1.84121057e-02
-1.51272908e-01 -1.20222306e+00 -4.47625786e-01 5.66787660e-01
-3.49094629e-01 -9.31435227e-01 -5.73160648e-01 -8.08961093e-01
5.42389929e-01 9.16553497e-01 1.13145900e+00 6.22452915e-01
-4.25405532e-01 -5.47044724e-03 -4.55491155e-01 -6.72455192e-01
3.77534449e-01 1.80910096e-01 -4.82273787e-01 1.61959499e-01
-2.56659137e-03 -5.06735504e-01 -9.05549645e-01 2.55807966e-01
-5.52218556e-01 1.70354784e-01 1.00242722e+00 8.64294648e-01
5.35127342e-01 -3.54326516e-01 2.80845225e-01 -1.08644629e+00
4.47586775e-01 -3.31032544e-01 -5.09846747e-01 4.48487461e-01
-3.21795076e-01 -8.66597239e-03 3.54897141e-01 2.02236697e-02
-1.37268269e+00 8.62030163e-02 1.12114605e-02 -4.87105727e-01
2.14933366e-01 -4.71549556e-02 -2.91836083e-01 -2.63988107e-01
4.19019073e-01 1.22579247e-01 -8.55907083e-01 -1.08614348e-01
2.29106188e-01 3.72060597e-01 7.04564750e-01 -3.30163568e-01
9.09328878e-01 6.83539629e-01 -1.57776922e-01 -8.26042235e-01
-1.33640981e+00 -3.12644988e-01 -4.49638635e-01 -2.52407432e-01
9.64025974e-01 -9.81200755e-01 -6.86230659e-01 4.62744474e-01
-1.18032467e+00 -8.75627935e-01 2.55123619e-02 -2.64779061e-01
1.40604183e-01 1.81074962e-01 -3.67861599e-01 -7.69132018e-01
-5.33463955e-01 -1.20188522e+00 1.66102087e+00 5.48028767e-01
3.64332557e-01 -6.63323402e-01 -2.00225696e-01 -1.90368965e-01
5.39850593e-01 3.86569142e-01 4.91514564e-01 8.79354402e-02
-9.75913942e-01 2.31426641e-01 -1.03112555e+00 3.10356736e-01
2.84680843e-01 -4.72651096e-03 -1.57634413e+00 -2.13770762e-01
-2.42403194e-01 -2.18880296e-01 1.68301797e+00 5.02666414e-01
1.75806069e+00 1.74741261e-03 -6.31744385e-01 8.61656785e-01
1.28383482e+00 8.79920945e-02 8.96695316e-01 5.58219198e-03
1.07511854e+00 1.62218153e-01 6.46888256e-01 2.49302924e-01
2.90965915e-01 3.16712171e-01 5.36299407e-01 -6.33071661e-01
-5.52292347e-01 1.42591223e-02 3.02372128e-01 2.95270652e-01
-1.69158936e-01 -1.98578581e-01 -6.70794368e-01 5.83418131e-01
-2.02791715e+00 -7.44194746e-01 -1.46301789e-02 1.66408384e+00
2.27432072e-01 3.64537686e-01 -4.24387425e-01 -2.06450135e-01
7.23324955e-01 8.00343513e-01 -4.81967121e-01 -7.25567117e-02
-3.17465156e-01 3.26015443e-01 5.84641874e-01 8.23431790e-01
-1.19400930e+00 1.26723385e+00 5.62823439e+00 6.79479897e-01
-1.11909699e+00 7.30686635e-02 8.49004507e-01 -5.51242381e-02
-3.94958764e-01 2.75853034e-02 -8.33347738e-01 3.88162345e-01
3.95750254e-01 6.62645400e-01 2.86051869e-01 9.27616298e-01
7.47939572e-02 -2.97068805e-01 -5.38710713e-01 7.45723486e-01
1.82014585e-01 -1.54424036e+00 -1.21114232e-01 -7.93413892e-02
1.03451431e+00 3.15633506e-01 1.56411514e-01 6.20155692e-01
4.26074266e-01 -1.09491765e+00 6.25422478e-01 4.89802271e-01
7.16829658e-01 -8.79571617e-01 9.37701523e-01 -2.54203409e-01
-1.91932106e+00 -4.06870604e-01 -5.90430379e-01 9.95298326e-02
-1.79631442e-01 6.53274596e-01 -7.90609598e-01 3.61723930e-01
1.15138543e+00 7.72757113e-01 -7.73178518e-01 5.87500572e-01
-4.57094878e-01 5.61188459e-01 -7.71534666e-02 -1.38859451e-01
4.14698899e-01 3.12267751e-01 2.50485599e-01 1.69173503e+00
1.41141608e-01 2.23558098e-01 1.90998942e-01 7.39295065e-01
-3.15814406e-01 -6.44535899e-01 -2.79164970e-01 6.95143998e-01
6.70996249e-01 1.54411352e+00 -8.41927826e-01 -4.20466363e-01
-4.43856508e-01 1.23789167e+00 3.78616333e-01 5.96430779e-01
-1.45672345e+00 -7.07998276e-01 8.13124955e-01 -2.45331079e-01
5.36772490e-01 -7.22819045e-02 -5.37339449e-01 -7.83427298e-01
-5.30329049e-02 -4.44529027e-01 1.60914227e-01 -9.17488933e-01
-7.23367929e-01 7.34886229e-01 -3.99999887e-01 -6.92039847e-01
7.26252258e-01 -5.58103561e-01 -1.08808792e+00 9.53232586e-01
-2.14418411e+00 -1.47988689e+00 -1.14270377e+00 6.67633176e-01
7.32070029e-01 1.75889015e-01 3.92430544e-01 3.19566637e-01
-7.99213052e-01 6.79753900e-01 -2.30874404e-01 2.91698396e-01
7.26261854e-01 -1.33312511e+00 8.40543985e-01 1.13794434e+00
-2.10548028e-01 5.82880795e-01 2.15982497e-01 -8.20515454e-01
-1.01108408e+00 -1.66266394e+00 6.95668161e-01 -3.06832910e-01
4.52788472e-01 -7.02173233e-01 -9.18475330e-01 5.97432792e-01
1.02637053e-01 4.32817489e-01 2.80497253e-01 2.29120627e-01
-4.74720269e-01 -2.94894725e-01 -5.17236471e-01 6.29463434e-01
1.35829639e+00 -7.06692398e-01 -1.64594382e-01 1.07561454e-01
1.20617914e+00 -6.93353117e-01 -2.29083464e-01 7.10859656e-01
6.50230169e-01 -1.26366889e+00 9.48289275e-01 -6.02166764e-02
3.84839118e-01 -4.70152497e-01 -4.87542838e-01 -7.98474193e-01
-6.40080392e-01 -7.24477291e-01 -4.90313381e-01 1.08588624e+00
5.11503220e-01 -4.20933902e-01 7.52110064e-01 -1.47421598e-01
-6.50671422e-01 -1.35388517e+00 -5.01504481e-01 -2.21439376e-01
-5.88021040e-01 -5.44446409e-01 7.07944810e-01 3.63857925e-01
-9.57461894e-01 5.32025754e-01 -6.41022503e-01 7.96971679e-01
3.90398920e-01 5.32943249e-01 7.39389658e-01 -9.97391820e-01
-1.54222161e-01 -2.07943529e-01 1.88941494e-01 -1.35311711e+00
5.43667451e-02 -3.64530891e-01 6.72819555e-01 -1.56947351e+00
4.66497719e-01 -4.79703933e-01 -6.97371244e-01 6.77683175e-01
-7.76479125e-01 4.86554086e-01 2.80045688e-01 -3.85367759e-02
-8.63061607e-01 7.84167111e-01 1.12792957e+00 -9.16668251e-02
-3.40801179e-01 8.55619982e-02 -8.17302525e-01 8.07326496e-01
5.67082584e-01 -2.56287754e-01 -3.97878408e-01 -7.12038159e-01
-6.77553266e-02 -4.35976863e-01 8.05270314e-01 -1.26521862e+00
3.40420902e-02 1.73594542e-02 9.85947549e-01 -1.02056980e+00
3.54329616e-01 -6.79679811e-01 -5.86013675e-01 5.77431977e-01
-1.63278818e-01 -2.51151085e-01 2.50136018e-01 8.01737010e-01
2.03017592e-01 4.85364854e-01 8.37530136e-01 2.58144706e-01
-1.08994436e+00 3.35035801e-01 -2.57794291e-01 -3.14431190e-01
7.48129666e-01 9.53658074e-02 -7.11463869e-01 -3.42449218e-01
-4.46658999e-01 5.15013278e-01 -9.24327001e-02 5.87132215e-01
9.01596129e-01 -1.29935813e+00 -3.57170373e-01 1.58379808e-01
-1.41260117e-01 2.17486754e-01 3.36923450e-01 6.31840706e-01
-3.87876809e-01 3.85811359e-01 -1.21497229e-01 -6.25777721e-01
-1.29400742e+00 6.45373240e-02 4.91896033e-01 -3.31874996e-01
-6.53714299e-01 1.41466081e+00 7.83691406e-01 -1.04966305e-01
7.01545775e-01 -7.23532677e-01 -1.33609504e-01 -8.80168602e-02
6.51539087e-01 3.58132124e-01 -1.67517185e-01 -4.71138597e-01
-4.19182211e-01 4.41797853e-01 2.31423993e-02 3.41565311e-01
1.16879904e+00 -1.31322443e-01 -7.34114125e-02 5.66961244e-02
1.20732009e+00 1.46196395e-01 -1.85124564e+00 -3.32888752e-01
-2.71466851e-01 -2.92014420e-01 2.13994712e-01 -8.23267758e-01
-1.31135941e+00 8.51055145e-01 7.96532989e-01 9.50079262e-02
1.51729596e+00 9.76963062e-03 1.09883451e+00 6.26334786e-01
-1.06934667e-01 -9.09739017e-01 5.70325553e-01 1.02682817e+00
1.03648233e+00 -1.21799588e+00 4.56603877e-02 -4.87098038e-01
-5.18806577e-01 8.49844217e-01 9.33898211e-01 -5.07349849e-01
7.61426210e-01 5.42044520e-01 8.33822694e-03 -3.17639947e-01
-6.35130405e-01 -6.06944442e-01 3.62223476e-01 5.50825179e-01
2.73447663e-01 -1.19493008e-02 2.16725677e-01 8.32498848e-01
1.48058787e-01 -4.11688447e-01 1.85231060e-01 9.22627687e-01
-8.17043960e-01 -5.94573736e-01 -4.04935747e-01 3.70591998e-01
-2.16948867e-01 -3.69974524e-01 -7.97161460e-01 6.70357823e-01
5.70245624e-01 8.74896348e-01 -4.38567698e-02 -8.95322323e-01
2.77177989e-01 -4.24352229e-01 -2.38287672e-02 -6.47391319e-01
-4.83408242e-01 1.53443649e-01 -6.54774010e-02 -1.01359808e+00
-2.19190687e-01 -2.61447668e-01 -1.25474679e+00 -2.80791782e-02
-3.83426815e-01 -1.79588884e-01 2.30744004e-01 9.30599570e-01
4.68763322e-01 1.48552454e+00 7.79383600e-01 -1.21766436e+00
3.31000358e-01 -7.83287227e-01 -1.43273205e-01 -3.93744886e-01
8.52196634e-01 -7.01515257e-01 1.62071064e-02 -2.10803434e-01] | [10.863597869873047, -4.128228664398193] |
f66e27f8-a932-452c-a9ff-7e5825d842e1 | tracking-deformable-parts-via-dynamic | 1311.0262 | null | http://arxiv.org/abs/1311.0262v1 | http://arxiv.org/pdf/1311.0262v1.pdf | Tracking Deformable Parts via Dynamic Conditional Random Fields | Despite the success of many advanced tracking methods in this area, tracking
targets with drastic variation of appearance such as deformation, view change
and partial occlusion in video sequences is still a challenge in practical
applications. In this letter, we take these serious tracking problems into
account simultaneously, proposing a dynamic graph based model to track object
and its deformable parts at multiple resolutions. The method introduces well
learned structural object detection models into object tracking applications as
prior knowledge to deal with deformation and view change. Meanwhile, it
explicitly formulates partial occlusion by integrating spatial potentials and
temporal potentials with an unparameterized occlusion handling mechanism in the
dynamic conditional random field framework. Empirical results demonstrate that
the method outperforms state-of-the-art trackers on different challenging video
sequences. | ['Zhixin Sun', 'Xu Cheng', 'Zhenyang Wu', 'Suofei Zhang'] | 2013-10-30 | null | null | null | null | ['occlusion-handling'] | ['computer-vision'] | [-1.39450356e-01 -5.69518745e-01 -1.97817251e-01 -7.14804679e-02
-3.71788800e-01 -7.31964052e-01 5.07777035e-01 -4.41901237e-01
-2.74376988e-01 6.51502490e-01 -2.87677497e-01 1.95173845e-01
-5.67547046e-02 -3.23814541e-01 -6.76098466e-01 -8.49438429e-01
-6.92427829e-02 3.76982063e-01 9.83963609e-01 1.45894572e-01
-1.13570407e-01 6.48007631e-01 -1.18827260e+00 -4.27817136e-01
9.43207800e-01 9.08935547e-01 3.44920397e-01 4.62213337e-01
9.99876633e-02 6.54995799e-01 -4.16489601e-01 -5.54821670e-01
3.76219153e-01 1.67665482e-01 -8.37813541e-02 5.55261254e-01
9.83108521e-01 -2.78021842e-01 -7.57931888e-01 1.46831691e+00
3.88206780e-01 3.51272732e-01 3.88810277e-01 -1.21593034e+00
-7.83595383e-01 1.02188803e-01 -9.88831162e-01 6.67286456e-01
1.35801747e-01 3.99693340e-01 4.50465083e-01 -8.67789924e-01
6.16398513e-01 1.41219544e+00 7.09014714e-01 5.54526865e-01
-1.13263977e+00 -7.12379456e-01 8.60103965e-01 2.50908852e-01
-1.23153710e+00 -1.56016901e-01 7.45131671e-01 -8.17690372e-01
4.81394261e-01 -2.39521787e-02 7.48832703e-01 9.20450032e-01
3.97406220e-01 8.28396022e-01 7.91673243e-01 1.96420830e-02
-1.95337057e-01 -2.77015179e-01 2.68645972e-01 7.24288940e-01
9.12446499e-01 4.23985898e-01 -2.95058787e-01 -2.30358958e-01
1.09309530e+00 2.20798746e-01 -5.17879426e-01 -9.16099727e-01
-1.12366939e+00 4.66434896e-01 3.56143177e-01 5.68127185e-02
-3.88594508e-01 2.98515916e-01 8.95113349e-02 -2.00943723e-01
5.57696402e-01 -2.45175675e-01 -3.44597787e-01 3.42215329e-01
-7.12423384e-01 3.30392212e-01 3.69344264e-01 1.27873015e+00
4.15746957e-01 6.17330849e-01 -5.82097769e-01 4.01684284e-01
6.96602404e-01 1.04594183e+00 -1.00207433e-01 -6.65675342e-01
3.76745343e-01 3.68875533e-01 5.50942421e-01 -1.17572546e+00
-2.29674250e-01 -9.38692451e-01 -6.60575032e-01 2.01310456e-01
4.33134854e-01 -1.05528302e-01 -1.01195562e+00 1.84024739e+00
9.78981733e-01 8.58012259e-01 -4.38753963e-01 1.04630387e+00
7.93961763e-01 3.06126684e-01 3.76463532e-01 -6.25673831e-01
1.36955619e+00 -1.00423563e+00 -1.20497859e+00 -2.65494525e-01
-4.23082113e-02 -8.84715855e-01 1.27850473e-01 2.17762202e-01
-1.12406385e+00 -8.12447488e-01 -8.21750462e-01 4.45042789e-01
2.37389073e-01 1.28024563e-01 7.05597639e-01 5.49705267e-01
-8.24829280e-01 2.09257960e-01 -1.26517737e+00 -2.92843938e-01
4.75488186e-01 4.92321014e-01 -9.29658860e-02 -1.46543011e-01
-9.09262180e-01 9.12136316e-01 2.81908810e-01 5.41770339e-01
-1.10834694e+00 -7.21169353e-01 -7.52765119e-01 -2.78631836e-01
6.93756223e-01 -9.66523111e-01 9.18088257e-01 -6.64650321e-01
-1.24476767e+00 6.04417741e-01 -1.93681359e-01 -1.93576247e-01
8.04926634e-01 -4.66596693e-01 -5.55315852e-01 -1.60430551e-01
-9.02420282e-02 1.18773893e-01 1.16195941e+00 -1.22594261e+00
-4.65470284e-01 -5.87952077e-01 -9.37393606e-02 3.28294009e-01
-3.67239006e-02 3.05006981e-01 -9.25652742e-01 -8.47740471e-01
1.38629079e-01 -1.03836107e+00 -2.88957238e-01 4.56954032e-01
-1.80534795e-01 -1.49994552e-01 1.26394713e+00 -6.45663261e-01
1.04096222e+00 -1.87565649e+00 4.30029958e-01 -2.57029951e-01
3.34362119e-01 5.31436324e-01 1.20896660e-03 -2.37933248e-01
2.81603485e-01 -4.83012229e-01 5.88768013e-02 -2.92132020e-01
-4.78003994e-02 1.44052297e-01 -1.64851502e-01 8.98076594e-01
2.69356854e-02 1.14708722e+00 -1.08962262e+00 -7.18750894e-01
4.22976077e-01 9.64304090e-01 -3.44161212e-01 2.63087571e-01
-2.76017874e-01 8.55627418e-01 -9.84207332e-01 9.12802100e-01
1.24177372e+00 -3.00230294e-01 -9.13130119e-02 -4.70749706e-01
-2.28895470e-01 -5.28757453e-01 -1.42076147e+00 1.51773524e+00
3.31216186e-01 3.32216799e-01 4.67693120e-01 -7.85543323e-01
5.99511385e-01 2.53555238e-01 7.51569450e-01 -1.45235002e-01
2.38750666e-01 -1.46489009e-01 2.00621374e-02 -4.78053093e-01
4.07949686e-01 -3.19741666e-02 4.22953963e-01 -2.45136052e-01
4.63693440e-02 3.59773308e-01 2.97061820e-02 7.59529173e-02
1.04036987e+00 5.48854887e-01 4.08228785e-02 -1.73892245e-01
6.56420946e-01 -2.70500541e-01 1.10544109e+00 6.54224992e-01
-6.29014492e-01 4.32963908e-01 -3.63883227e-01 -4.76334482e-01
-4.76105660e-01 -1.21069229e+00 -2.08507344e-01 6.52695596e-01
6.98989451e-01 -2.24863756e-02 -2.86764860e-01 -6.26095057e-01
1.71949178e-01 1.08472668e-02 -4.74449366e-01 5.74015304e-02
-8.50718021e-01 -9.09581304e-01 9.48822945e-02 5.79290867e-01
4.67352390e-01 -7.78429687e-01 -3.42084050e-01 3.62572014e-01
-1.33267641e-01 -1.50853395e+00 -7.80884445e-01 -3.37180972e-01
-1.03995681e+00 -1.25298810e+00 -8.62172365e-01 -6.00704432e-01
5.55339038e-01 6.62620008e-01 9.76311028e-01 -4.21643890e-02
-5.75326562e-01 7.58614898e-01 -6.06749021e-02 -1.90996736e-01
-4.58988501e-03 -4.99534965e-01 1.71420023e-01 2.80185729e-01
2.86292844e-02 -2.68630981e-01 -7.24328816e-01 6.58116341e-01
-7.79627860e-01 -2.53206581e-01 3.46637994e-01 5.71829140e-01
7.98416734e-01 9.27679092e-02 -1.49504185e-01 -4.41330552e-01
-1.75367817e-01 -2.20630854e-01 -1.28619361e+00 5.20374715e-01
-1.42428041e-01 -3.21696460e-01 5.10606058e-02 -1.01096928e+00
-1.31743431e+00 3.68756294e-01 3.95892650e-01 -1.06704485e+00
1.96963072e-01 6.13371916e-02 -4.61160779e-01 -8.01941216e-01
5.25626093e-02 3.03476423e-01 -8.19976330e-02 -4.93034959e-01
1.70597941e-01 -2.83476323e-01 5.87945163e-01 -5.76460183e-01
1.53820920e+00 7.93937981e-01 1.71751678e-01 -5.47923088e-01
-9.67587411e-01 -5.06526411e-01 -7.50194490e-01 -6.98483288e-01
1.01709104e+00 -1.10711288e+00 -8.53686631e-01 8.29146147e-01
-1.31325519e+00 -1.54654058e-02 -2.02903450e-02 8.39637041e-01
-2.77447224e-01 7.53647447e-01 -4.30160820e-01 -9.57997262e-01
-5.23094423e-02 -1.12060726e+00 9.98151243e-01 5.91253281e-01
7.10797429e-01 -1.25510848e+00 9.82378423e-02 1.49850473e-01
4.73380774e-01 4.55679715e-01 5.23119606e-03 -1.12899080e-01
-1.42980552e+00 -2.39711225e-01 -1.92552641e-01 1.27301961e-01
1.75533921e-01 8.71053264e-02 -5.23539186e-01 -6.75072134e-01
1.88693419e-01 1.93503618e-01 8.48810554e-01 9.48377490e-01
7.27034450e-01 6.89139143e-02 -8.00297320e-01 8.86014581e-01
1.41211748e+00 3.11840653e-01 3.62889558e-01 5.93231581e-02
1.18633795e+00 1.98221728e-01 9.11899209e-01 4.19551998e-01
2.30019584e-01 1.22331059e+00 6.41769528e-01 1.61297381e-01
-3.34772974e-01 9.59997177e-02 5.18378019e-01 5.54529548e-01
-2.79909670e-01 -2.94796079e-01 -5.97283542e-01 3.25936884e-01
-2.09899044e+00 -1.16009414e+00 -4.95612085e-01 2.17369175e+00
3.44919384e-01 4.86494079e-02 -6.54428229e-02 -6.61850512e-01
1.28229833e+00 2.72928268e-01 -8.18873465e-01 8.62839639e-01
-2.88587123e-01 -3.17578137e-01 6.08463109e-01 4.88899142e-01
-1.35925686e+00 1.00610042e+00 6.21643305e+00 6.86204553e-01
-9.57467854e-01 3.56190950e-01 -1.05942581e-02 8.01840499e-02
5.96638881e-02 1.00876048e-01 -1.46904075e+00 5.61185002e-01
1.20146915e-01 -4.54722419e-02 1.27614975e-01 5.69123447e-01
1.39182787e-02 1.61330536e-01 -6.23518050e-01 9.58448708e-01
1.15287244e-01 -1.00562704e+00 -1.23113044e-01 9.72445905e-02
7.16871023e-01 1.05971590e-01 2.51292259e-01 2.13557109e-01
4.00814503e-01 -4.16847616e-01 7.04670191e-01 7.89362371e-01
2.25484848e-01 -1.54070809e-01 3.13930303e-01 8.77271295e-02
-1.92656052e+00 1.73849761e-01 -5.10887146e-01 2.89278686e-01
6.89014614e-01 5.00293195e-01 7.78660402e-02 7.83179224e-01
6.13033533e-01 1.07202601e+00 -6.37898028e-01 1.71026528e+00
-2.42501236e-02 4.09705520e-01 -4.05019164e-01 5.37071884e-01
-7.98241943e-02 -2.26840258e-01 1.22140527e+00 8.49233091e-01
2.40273103e-01 3.29906315e-01 7.63292491e-01 7.32242763e-01
3.59653115e-01 -3.46834630e-01 -3.21642578e-01 2.92041689e-01
2.84583390e-01 1.41386533e+00 -7.15254545e-01 -1.09114252e-01
-5.90246201e-01 6.12398744e-01 9.99116749e-02 7.01795161e-01
-1.38241696e+00 4.29602921e-01 5.88265479e-01 2.39427030e-01
7.52532125e-01 -4.88544822e-01 4.46250111e-01 -1.76650202e+00
2.74545133e-01 -3.53753567e-01 3.87212873e-01 -6.20934963e-01
-1.39726150e+00 5.24664581e-01 7.15687275e-02 -1.61127925e+00
2.47533500e-01 -5.69998145e-01 -6.58508897e-01 7.39250183e-01
-1.69008958e+00 -1.53238606e+00 -5.13589978e-01 7.66127527e-01
4.79787141e-01 -6.04570881e-02 4.83237728e-02 6.70896232e-01
-7.02853441e-01 3.47052008e-01 -6.18578605e-02 1.94454163e-01
6.81139231e-01 -8.09830248e-01 1.84691116e-01 1.13473022e+00
-9.69319791e-02 5.82827747e-01 8.62919211e-01 -1.09969509e+00
-1.77690637e+00 -1.29573369e+00 2.32175931e-01 -5.71915269e-01
8.98427725e-01 -2.01109543e-01 -1.17923617e+00 8.58662009e-01
7.86426663e-02 8.49044800e-01 1.04487464e-01 -2.37152427e-01
-1.89474717e-01 8.87792557e-02 -7.47395515e-01 1.69551075e-01
1.43481183e+00 1.01292832e-02 -3.34927619e-01 4.06716347e-01
5.44440508e-01 -9.01099026e-01 -7.68167496e-01 8.31044793e-01
3.77358556e-01 -4.68830973e-01 1.36311233e+00 -4.63540643e-01
-6.05772495e-01 -8.38539243e-01 -1.35196419e-02 -8.47575545e-01
-6.70315146e-01 -8.38324487e-01 -7.30951905e-01 1.29180098e+00
-2.85537511e-01 -5.02734959e-01 7.28700638e-01 6.09520078e-01
-1.53044134e-01 -3.18908244e-01 -1.03066528e+00 -1.14710343e+00
-2.78501898e-01 -3.92801277e-02 -1.99071951e-02 8.53804946e-01
-9.06008184e-01 -4.44839858e-02 -7.33521402e-01 6.76466048e-01
1.47142136e+00 9.81327370e-02 8.03738892e-01 -1.59893918e+00
-4.61995453e-01 -1.23499855e-01 -6.72550559e-01 -1.32537913e+00
3.04321021e-01 -6.18192375e-01 5.43741360e-02 -1.37342811e+00
5.19266009e-01 -4.35040951e-01 -4.33066398e-01 -2.26970506e-03
-6.68083072e-01 1.30147427e-01 4.23647016e-01 3.87497604e-01
-1.09007561e+00 8.32165301e-01 1.56254208e+00 -3.60213757e-01
1.36029676e-01 2.96969712e-01 -1.79056138e-01 7.82960713e-01
1.76666975e-01 -6.77659035e-01 -2.26931512e-01 -6.43078327e-01
-2.46205896e-01 2.17333093e-01 6.92282736e-01 -1.05473149e+00
5.31159580e-01 -3.26219738e-01 5.54710984e-01 -9.74982679e-01
4.66703862e-01 -1.03598905e+00 5.96269608e-01 5.01172006e-01
3.15930992e-01 1.83736190e-01 4.68747497e-01 1.24794328e+00
-1.56822458e-01 1.99357808e-01 8.88301671e-01 1.68708682e-01
-8.00260842e-01 1.11305690e+00 4.31697350e-03 2.33073905e-01
1.23822176e+00 -3.29996198e-01 -1.87525779e-01 1.17508620e-02
-1.02639198e+00 4.57714379e-01 5.17576218e-01 6.57263994e-01
4.43555802e-01 -1.50407875e+00 -7.60926723e-01 -5.13158925e-02
-1.55400664e-01 -1.18448816e-01 7.24934936e-01 1.17741859e+00
-2.43637636e-02 2.26296440e-01 -1.03011854e-01 -1.01943278e+00
-1.40273643e+00 7.55236387e-01 5.37498593e-01 -4.57354546e-01
-7.55984843e-01 8.09158564e-01 7.36538291e-01 -2.75553521e-02
3.64045292e-01 5.21799456e-03 -2.71868020e-01 -1.23325974e-01
5.69877505e-01 2.73145825e-01 -3.90756130e-01 -1.05942988e+00
-5.76713324e-01 1.27242386e+00 -2.10885122e-01 3.58940989e-01
9.78971541e-01 -3.36066902e-01 2.47391179e-01 1.39244378e-01
6.55949056e-01 -5.13693206e-02 -1.90421855e+00 -6.41242385e-01
-1.71771526e-01 -9.26705003e-01 2.80312216e-03 -5.20522892e-01
-1.65258539e+00 6.19290173e-01 8.13051939e-01 -5.48058301e-02
7.99661756e-01 -8.67075697e-02 4.96134132e-01 6.39492199e-02
6.66202188e-01 -5.38236916e-01 1.87316760e-01 5.57854831e-01
6.22885287e-01 -1.36028767e+00 2.94244945e-01 -8.78314018e-01
-4.09252733e-01 7.20615208e-01 9.70669746e-01 -3.84987146e-01
8.45723271e-01 3.74204636e-01 -9.27463770e-02 -9.08267125e-02
-6.06106162e-01 -2.23198324e-01 6.72341943e-01 8.43408048e-01
1.91371739e-01 -2.51215875e-01 -2.94699490e-01 2.27535069e-01
7.81492591e-01 -6.10268116e-03 -1.45686464e-02 9.69243348e-01
-3.32916945e-01 -1.02946091e+00 -6.88975096e-01 1.90662537e-02
-5.62934995e-01 1.52699292e-01 9.09562781e-02 8.04699421e-01
7.01770931e-02 5.91707289e-01 -1.82007268e-01 5.27527481e-02
3.82176906e-01 -3.99070382e-01 8.99584949e-01 -4.22408611e-01
-5.49447477e-01 6.65532708e-01 -4.46042120e-01 -6.11554623e-01
-8.75911176e-01 -1.08085513e+00 -9.59855258e-01 3.73799503e-02
-8.66189957e-01 -5.02468348e-02 4.02004510e-01 8.27419043e-01
3.04974616e-01 8.05359244e-01 2.05507398e-01 -1.17873847e+00
-5.77832758e-01 -7.48460293e-01 -6.18925929e-01 3.53526652e-01
3.35533023e-01 -1.46904230e+00 -3.25005293e-01 1.48259714e-01] | [6.312592029571533, -2.1886115074157715] |
9fd100da-2e44-4bcd-a140-098e3e821123 | covid-19-diagnosis-from-x-ray-using-neural | 2105.14333 | null | https://arxiv.org/abs/2105.14333v1 | https://arxiv.org/pdf/2105.14333v1.pdf | Covid-19 diagnosis from x-ray using neural networks | Corona virus or COVID-19 is a pandemic illness, which has influenced more than million of causalities worldwide and infected a few large number of individuals .Innovative instrument empowering quick screening of the COVID-19 contamination with high precision can be critically useful to the medical care experts. The primary clinical device presently being used for the analysis of COVID-19 is the Reverse record polymerase chain response as known as RT-PCR, which is costly, less-delicate and requires specific clinical work force. X-Ray imaging is an effectively available apparatus that can be a great option in the COVID-19 conclusion. This exploration was taken to examine the utility of computerized reasoning in the quick and exact recognition of COVID-19 from chest X-Ray pictures. The point of this paper is to propose a procedure for programmed recognition of COVID-19 from advanced chest X-Ray images applying pre-prepared profound learning calculations while boosting the discovery exactness. The point is to give over-focused on clinical experts a second pair of eyes through a learning picture characterization models. We distinguish an appropriate Convolutional Neural Network-CNN model through beginning similar investigation of a few mainstream CNN models. | ['Mohammed Rhithick A', 'Dinesh J'] | 2021-05-29 | null | null | null | null | ['covid-19-detection'] | ['medical'] | [ 3.32975835e-01 -1.96497262e-01 2.16968715e-01 -5.85512072e-02
-1.49417281e-01 -6.63324893e-01 1.60267532e-01 3.08420688e-01
-6.00570917e-01 7.42070913e-01 -1.39781937e-01 -9.56160605e-01
-4.38209444e-01 -7.30784416e-01 -3.98944765e-01 -6.54445291e-01
-8.34162012e-02 1.05237639e+00 -2.59673297e-01 -3.02319318e-01
2.98996061e-01 9.79312003e-01 -1.02753079e+00 4.47235256e-01
6.23400390e-01 9.76663888e-01 3.58627170e-01 1.25022161e+00
-1.04890252e-02 7.84275234e-01 -3.03758144e-01 -1.13161616e-01
6.58664480e-02 -3.66221666e-01 -9.12131667e-01 -5.80052197e-01
-4.79264297e-02 -6.20330691e-01 4.22302127e-01 7.54850149e-01
8.25080633e-01 -4.10384774e-01 1.02664912e+00 -8.20685267e-01
-3.27047378e-01 1.03689685e-01 -1.32065758e-01 7.16321409e-01
4.00457025e-01 4.95861530e-01 2.27010757e-01 -6.70141459e-01
5.52936018e-01 8.08099210e-01 1.22396004e+00 4.67639863e-01
-4.13481325e-01 -4.69656259e-01 -7.36879349e-01 2.68534184e-01
-1.18267322e+00 4.85303938e-01 4.01478976e-01 -9.45310593e-01
1.21457398e+00 3.84616613e-01 9.26153004e-01 9.56206918e-01
8.04316878e-01 1.26754776e-01 1.12388527e+00 -2.51783192e-01
1.78957656e-02 2.93333143e-01 4.77109581e-01 1.05267978e+00
4.64931816e-01 1.86352298e-01 2.86479026e-01 -3.19511950e-01
7.67265558e-01 7.76094019e-01 -3.95877957e-01 2.02077061e-01
-9.88543391e-01 9.82352257e-01 3.16023946e-01 5.23131311e-01
-6.11353755e-01 -2.87211210e-01 6.98050141e-01 7.56676793e-02
-1.14143610e-01 8.09469521e-01 -7.55843341e-01 2.80628838e-02
-7.47030199e-01 1.56893983e-01 5.53754687e-01 3.84488612e-01
2.05554426e-01 -1.58733830e-01 -1.39665648e-01 6.21981025e-02
2.45846286e-01 9.30019855e-01 4.91185665e-01 -2.87516773e-01
6.91992640e-02 8.13662887e-01 7.38978162e-02 -9.47616518e-01
-9.10343289e-01 -4.38662440e-01 -1.35302961e+00 3.33892554e-01
2.70398378e-01 -3.18226606e-01 -1.02979839e+00 9.81067479e-01
3.19818944e-01 -1.58814788e-01 -6.87747374e-02 8.90765846e-01
9.25785840e-01 6.41511202e-01 1.63268134e-01 -2.80805200e-01
2.01829648e+00 -5.05874336e-01 -5.02220452e-01 4.89150107e-01
5.50115108e-01 -7.33936489e-01 6.95224762e-01 8.37398112e-01
-7.14897275e-01 -5.73847413e-01 -1.14501643e+00 3.06623191e-01
-7.59093702e-01 1.80705503e-01 5.53391814e-01 8.79790008e-01
-7.04176009e-01 6.57141626e-01 -4.89969283e-01 -6.26257360e-01
4.07731891e-01 5.35068154e-01 -2.02790067e-01 1.09397709e-01
-1.13941181e+00 1.21617985e+00 3.53213578e-01 2.26143152e-01
-9.16630864e-01 -8.34297478e-01 -2.66295135e-01 -1.07850470e-02
1.99841708e-02 -1.09603095e+00 1.03820562e+00 -6.53169692e-01
-1.12308717e+00 1.11616719e+00 3.02147567e-01 -4.80518281e-01
5.60461521e-01 -2.71400064e-01 -4.41622823e-01 6.49766684e-01
-2.37103611e-01 3.22957695e-01 8.43225479e-01 -8.11580360e-01
-4.34994310e-01 -2.75673658e-01 -3.83800358e-01 -3.37584376e-01
2.80402929e-01 3.27283114e-01 2.90519357e-01 -4.99587178e-01
-3.15656781e-01 -8.57272685e-01 -2.09903196e-01 -1.38501704e-01
-2.51912892e-01 -3.15258712e-01 7.78172970e-01 -7.30713904e-01
8.73699129e-01 -1.56725693e+00 -4.96087819e-01 2.56887078e-01
5.47142267e-01 1.08052790e+00 3.72440606e-01 5.82341075e-01
-4.09281254e-01 3.40467483e-01 -2.66657442e-01 5.60255468e-01
-3.47861171e-01 -1.66490465e-01 -3.60075653e-01 5.29112339e-01
4.13167238e-01 9.24017131e-01 -7.11929739e-01 -7.41021752e-01
1.84412301e-01 8.36288691e-01 -3.85132313e-01 5.35815239e-01
-2.14392468e-01 5.17734110e-01 -6.47925854e-01 7.55408645e-01
8.65183711e-01 -6.70674086e-01 4.21876833e-02 -3.77725422e-01
2.51307078e-02 -3.65017891e-01 -7.28980839e-01 8.56239498e-01
7.63299540e-02 5.08058667e-01 -2.27028728e-01 -8.69572163e-01
6.31006837e-01 6.25293374e-01 2.47006938e-01 -4.32969838e-01
6.48934603e-01 9.12760943e-02 -2.24479720e-01 -1.39466763e+00
3.60459834e-02 -6.17210209e-01 3.48694444e-01 6.58325493e-01
-3.30316275e-01 4.77219597e-02 -2.52538234e-01 -6.48540035e-02
9.34148788e-01 -2.21905336e-01 8.34857106e-01 -2.89017290e-01
8.06110203e-01 3.61891508e-01 7.17042759e-02 8.66849065e-01
-3.96716386e-01 6.78809941e-01 1.10419102e-01 -1.01836538e+00
-1.03853452e+00 -1.01552439e+00 -2.08855063e-01 5.67869663e-01
-3.55874360e-01 3.54017764e-01 -8.97447586e-01 -4.27049130e-01
-3.73273581e-01 2.48798355e-01 -7.88779795e-01 3.75102222e-01
-8.52867901e-01 -5.82379103e-01 9.32578325e-01 5.38112342e-01
4.93423969e-01 -1.32153726e+00 -1.32051039e+00 1.63012654e-01
1.39583811e-01 -6.37263894e-01 1.31735697e-01 2.56223738e-01
-9.95715678e-01 -1.48859060e+00 -9.41395164e-01 -9.38750982e-01
5.40182948e-01 -3.26231532e-02 9.77150977e-01 7.53450930e-01
-9.81314838e-01 3.62248868e-01 -2.54860729e-01 -8.76702189e-01
-6.61534548e-01 -3.21378862e-03 -1.50855677e-02 -4.00153071e-01
4.83825237e-01 -2.58438379e-01 -9.54557419e-01 -1.00056380e-01
-7.66141355e-01 2.16956642e-02 5.92583179e-01 6.15273297e-01
4.31229085e-01 -3.24041724e-01 4.24831688e-01 -8.99015725e-01
5.91201603e-01 -5.06775558e-01 -3.46065581e-01 3.22994798e-01
-8.26118648e-01 -1.51004881e-01 9.32582855e-01 -2.35370561e-01
-8.68224978e-01 -2.34872699e-01 -3.40318114e-01 -3.44335049e-01
-4.28204060e-01 2.00741306e-01 4.56793338e-01 3.47298980e-02
8.72983277e-01 1.78726450e-01 -4.00549285e-02 -2.66914070e-01
6.06272044e-03 7.27222621e-01 6.88929021e-01 -1.37750059e-01
4.95042741e-01 5.83047926e-01 2.92125732e-01 -9.99049604e-01
-5.26130974e-01 -6.74790382e-01 -4.98037755e-01 -4.90752548e-01
1.70776534e+00 -5.36044300e-01 -1.23015952e+00 5.19258916e-01
-1.52381921e+00 4.42765616e-02 2.34587595e-01 6.09782100e-01
-2.30982602e-01 4.83914226e-01 -7.16139436e-01 -7.09256351e-01
-1.12704253e+00 -1.07939434e+00 6.65381253e-01 2.29227334e-01
-2.50180274e-01 -9.25299346e-01 5.55857778e-01 3.07849765e-01
5.32396555e-01 5.92716098e-01 1.33684516e+00 -9.24964547e-01
-5.44521213e-01 -4.13819999e-01 -4.73332375e-01 3.12110931e-01
-1.88274711e-01 1.14196166e-01 -1.06369305e+00 -1.02195218e-01
5.12499154e-01 -2.57704079e-01 5.75615406e-01 3.68346363e-01
9.44165349e-01 -1.11345917e-01 -5.46324492e-01 7.68744051e-01
1.76483548e+00 6.97452188e-01 6.68379247e-01 1.29961133e-01
6.62591875e-01 2.85713404e-01 4.38801020e-01 3.63641322e-01
-2.35794216e-01 -1.36312246e-01 3.32227647e-01 -2.32917845e-01
2.72278458e-01 2.81079441e-01 -3.32554072e-01 6.39156342e-01
-6.40105009e-01 -3.08658510e-01 -1.34245718e+00 1.87528789e-01
-1.11584473e+00 -1.19230771e+00 -5.98339379e-01 1.47326744e+00
5.39530337e-01 -7.04661980e-02 5.85162006e-02 1.96537346e-01
7.55090535e-01 -5.32746375e-01 -1.87556803e-01 -5.97353399e-01
2.04300016e-01 5.80580831e-01 1.57173619e-01 1.23586990e-01
-8.57853472e-01 6.16599433e-02 6.92441082e+00 3.13498646e-01
-1.53054798e+00 1.68741882e-01 5.22350311e-01 3.39412987e-01
6.44137710e-02 -4.38728750e-01 -5.13735652e-01 3.70343655e-01
8.23844850e-01 3.42276603e-01 2.50203665e-02 9.40251827e-01
1.71090290e-01 5.91818709e-03 -1.04261899e+00 1.14694810e+00
2.08718076e-01 -1.84400165e+00 5.78873493e-02 -1.44734278e-01
3.89477730e-01 1.81638673e-01 -1.68782219e-01 2.99107004e-02
-2.42242500e-01 -1.21273065e+00 1.41214564e-01 7.52627432e-01
1.04933250e+00 -4.90850925e-01 1.09058487e+00 2.34047666e-01
-8.88028979e-01 -2.38562133e-02 -2.62301803e-01 -3.24281193e-02
1.15873523e-01 2.93270469e-01 -1.50207019e+00 1.81621283e-01
8.10419083e-01 -2.26331890e-01 -2.62702644e-01 8.91600311e-01
2.21107617e-01 6.69128656e-01 -1.13053910e-01 -5.30421317e-01
3.09285939e-01 -2.94382274e-02 3.03519905e-01 1.76187074e+00
2.69269168e-01 6.44985318e-01 -3.12299579e-01 6.43221140e-01
4.23886478e-01 3.27589005e-01 -7.29133010e-01 -7.92623311e-02
-8.01736861e-02 1.22758067e+00 -1.17277133e+00 -7.13375866e-01
1.38106234e-02 6.65620208e-01 -3.74945626e-02 -2.77160946e-02
-8.58759165e-01 -3.62603545e-01 7.24504516e-02 3.55679125e-01
5.60569167e-01 3.41848075e-01 -2.71453142e-01 -4.89724636e-01
-5.10089457e-01 -1.13713884e+00 5.45820832e-01 -1.18375659e+00
-1.24017155e+00 8.04708958e-01 2.36524623e-02 -1.11351752e+00
-2.02659979e-01 -1.21213770e+00 -8.82367134e-01 8.99755120e-01
-1.22326541e+00 -1.10338604e+00 -5.51665962e-01 5.88858783e-01
3.24354708e-01 -3.72483373e-01 1.10692251e+00 1.44554060e-02
-1.74180582e-01 1.07649248e-02 3.56568843e-02 6.80979639e-02
3.17333579e-01 -1.01859748e+00 -9.02531371e-02 3.76006007e-01
-8.98376465e-01 8.09617698e-01 6.80079520e-01 -8.82412970e-01
-1.12600720e+00 -8.89287829e-01 1.10456395e+00 -6.19865358e-01
9.42895114e-02 1.28659278e-01 -7.83621788e-01 4.30131406e-01
5.46178222e-01 -3.66565198e-01 8.06798577e-01 -6.18840158e-01
-1.33962974e-01 1.43941075e-01 -1.41898692e+00 2.88237184e-01
3.37039977e-01 -6.15009904e-01 -9.22976851e-01 6.23919547e-01
5.67808747e-01 -9.00317132e-02 -6.69378221e-01 2.78686315e-01
6.00727618e-01 -1.25108051e+00 1.01341891e+00 -9.37574506e-01
3.96111876e-01 -2.12577999e-01 2.28033945e-01 -5.14032602e-01
-2.67349958e-01 -7.59447873e-01 3.77450228e-01 4.04084772e-01
1.45856693e-01 -7.25298107e-01 6.41511023e-01 3.79180722e-02
4.76161540e-02 -9.12825048e-01 -5.59535444e-01 -3.45278889e-01
-1.03828780e-01 -4.73280907e-01 5.83903968e-01 1.05926144e+00
-1.70820653e-01 2.60670453e-01 -9.31844413e-02 2.22738504e-01
3.16548824e-01 -4.97162230e-02 5.17643332e-01 -1.34133255e+00
-3.61466646e-01 -2.52293766e-01 -2.53997713e-01 -3.01371515e-01
-7.61347115e-01 -6.56293273e-01 -1.94462404e-01 -1.51729453e+00
3.03633809e-01 -7.24478662e-02 -2.61984587e-01 4.22117114e-02
3.08279656e-02 1.86992422e-01 4.59946506e-02 1.15384504e-01
-1.34608924e-01 -4.57901865e-01 1.33497655e+00 -1.28554953e-02
1.26340061e-01 -6.20193370e-02 -4.37303603e-01 8.02534759e-01
1.01517987e+00 -7.83036768e-01 -3.91275316e-01 1.73317760e-01
9.60336745e-01 2.76636302e-01 4.42709506e-01 -1.03997052e+00
7.13512152e-02 -4.22287211e-02 6.44935548e-01 -1.24944723e+00
-1.76602900e-01 -9.12532926e-01 3.05191010e-01 1.15508091e+00
-2.26886105e-03 6.67503476e-01 3.04263998e-02 3.69452059e-01
2.51903117e-01 -5.18035471e-01 8.35223079e-01 -6.67740166e-01
-3.50473046e-01 2.79030483e-02 -9.12209332e-01 1.49566323e-01
1.09944665e+00 -4.58326310e-01 -3.60778183e-01 -1.65127907e-02
-4.48230118e-01 -2.75463790e-01 2.02986449e-01 -2.57769287e-01
6.53827727e-01 -6.55949056e-01 -4.75191206e-01 3.77387971e-01
-9.88155007e-02 -1.18825078e-01 6.21287346e-01 1.07739425e+00
-1.65411353e+00 6.26701057e-01 -5.67267835e-01 -6.76230967e-01
-1.47750819e+00 1.23586929e+00 5.79190612e-01 -5.34947217e-01
-6.34979904e-01 7.24077404e-01 1.29594907e-01 -1.04294099e-01
1.26483113e-01 -7.00140893e-01 -5.28367281e-01 -3.88056189e-02
7.63424933e-01 4.37942535e-01 9.05217156e-02 -2.87795454e-01
-5.75703740e-01 7.73944438e-01 -8.98447558e-02 3.64131659e-01
1.41869891e+00 2.52760649e-01 -3.00868452e-01 6.65321201e-02
1.24891388e+00 -3.83096874e-01 -5.27741969e-01 3.93990815e-01
-4.51232851e-01 1.64642811e-01 -2.09867984e-01 -1.09401786e+00
-5.23496687e-01 1.23128676e+00 1.20945477e+00 3.17998797e-01
1.05650175e+00 -2.51151383e-01 8.18346262e-01 7.51357317e-01
-9.87671390e-02 -7.82537341e-01 1.62389651e-01 3.15692991e-01
8.04862678e-01 -1.27784860e+00 1.51508883e-01 -4.66217883e-02
-5.46851873e-01 1.44714248e+00 1.67352781e-01 -1.68476880e-01
8.88251543e-01 3.17373514e-01 4.09420371e-01 -9.76331532e-01
-3.20268899e-01 2.40456715e-01 6.93029165e-02 1.02886605e+00
3.24934512e-01 2.28042439e-01 -2.44659364e-01 5.38656712e-01
1.06442466e-01 5.21162212e-01 2.57322133e-01 1.08966041e+00
-6.02475166e-01 -4.19746876e-01 -8.24337780e-01 5.23652315e-01
-7.51567900e-01 -2.84516782e-01 -1.20611288e-01 9.67141688e-01
6.50149703e-01 5.46287775e-01 -8.07259381e-02 -1.68250784e-01
9.56092551e-02 2.05022484e-01 4.20051187e-01 -1.50391743e-01
-9.94449258e-01 -3.24590504e-01 -1.84833482e-01 -1.72942832e-01
-4.18764710e-01 -2.48283103e-01 -1.33626807e+00 -4.12032276e-01
-6.56132028e-03 1.59783103e-02 7.58819401e-01 1.10329771e+00
1.05232932e-02 4.00213361e-01 2.25014895e-01 -5.51263273e-01
-5.76269031e-01 -7.58252561e-01 -4.20414001e-01 2.03155667e-01
6.83202684e-01 -1.78930044e-01 -1.83189005e-01 2.43280068e-01] | [15.581006050109863, -1.6898729801177979] |
52cc5946-26de-4fc6-992c-1c812cdbc2bc | improving-low-resource-cross-lingual-document | 1906.03492 | null | https://arxiv.org/abs/1906.03492v1 | https://arxiv.org/pdf/1906.03492v1.pdf | Improving Low-Resource Cross-lingual Document Retrieval by Reranking with Deep Bilingual Representations | In this paper, we propose to boost low-resource cross-lingual document retrieval performance with deep bilingual query-document representations. We match queries and documents in both source and target languages with four components, each of which is implemented as a term interaction-based deep neural network with cross-lingual word embeddings as input. By including query likelihood scores as extra features, our model effectively learns to rerank the retrieved documents by using a small number of relevance labels for low-resource language pairs. Due to the shared cross-lingual word embedding space, the model can also be directly applied to another language pair without any training label. Experimental results on the MATERIAL dataset show that our model outperforms the competitive translation-based baselines on English-Swahili, English-Tagalog, and English-Somali cross-lingual information retrieval tasks. | ['Neha Verma', 'Sungrok Shim', 'Rui Zhang', 'Garrett Bingham', 'Caitlin Westerfield', 'William Hu', 'Dragomir Radev', 'Alexander Fabbri'] | 2019-06-08 | improving-low-resource-cross-lingual-document-1 | https://aclanthology.org/P19-1306 | https://aclanthology.org/P19-1306.pdf | acl-2019-7 | ['cross-lingual-information-retrieval'] | ['natural-language-processing'] | [-3.18737984e-01 -4.59747225e-01 -7.40697443e-01 -3.58497530e-01
-1.78610158e+00 -8.14129293e-01 9.73609209e-01 1.92235366e-01
-1.01866257e+00 4.62723047e-01 5.91360390e-01 -4.25963193e-01
-1.67653143e-01 -5.82506955e-01 -6.03758454e-01 -3.15344930e-01
2.41253912e-01 1.16239417e+00 1.54009880e-02 -4.57190841e-01
7.40698874e-02 2.35909134e-01 -7.70268559e-01 5.34262002e-01
7.09757328e-01 6.15434527e-01 1.68378815e-01 3.72790277e-01
-2.94119179e-01 3.11454028e-01 -3.22590441e-01 -4.84114140e-01
1.78437069e-01 -7.54270777e-02 -9.14962947e-01 -7.40229487e-01
5.84228694e-01 -6.18343294e-01 -5.35302460e-01 6.39530480e-01
9.10686910e-01 8.01392086e-03 8.29774380e-01 -6.20587707e-01
-1.48059869e+00 7.87553668e-01 -4.97902632e-01 2.34618425e-01
2.10872918e-01 -4.30849731e-01 1.50455332e+00 -1.41030657e+00
8.25889230e-01 1.54510832e+00 3.67709666e-01 2.58461654e-01
-1.13286412e+00 -6.53793275e-01 -1.19438224e-01 2.29240507e-01
-1.66903794e+00 -4.15212095e-01 4.22661811e-01 1.33949947e-02
1.61708021e+00 4.93168011e-02 1.59101292e-01 1.09041917e+00
1.22585088e-01 9.81417537e-01 5.56152940e-01 -7.27407515e-01
-4.14604992e-01 2.81923085e-01 8.64231065e-02 5.07606089e-01
1.54796422e-01 -1.04566477e-01 -5.21762073e-01 -4.47651446e-01
3.19634825e-01 -1.38430961e-03 -9.98313949e-02 -1.65663317e-01
-1.18675625e+00 1.05740929e+00 6.22051835e-01 6.47947609e-01
-2.98349738e-01 2.92064637e-01 6.69472158e-01 5.70124209e-01
8.31578016e-01 5.98603606e-01 -7.46669173e-01 4.69546378e-01
-7.63573229e-01 2.37368822e-01 5.54149389e-01 1.02500236e+00
8.22491467e-01 -4.20382649e-01 -6.18801951e-01 1.42129660e+00
5.50074220e-01 1.08980262e+00 9.30224657e-01 -4.71816838e-01
5.28893769e-01 2.86710173e-01 1.54579366e-02 -8.47167790e-01
-8.60230997e-02 -5.84309876e-01 -2.98904806e-01 -7.94339836e-01
-2.09795892e-01 1.80758402e-01 -7.82135367e-01 1.62393904e+00
-1.47070724e-03 -6.76289558e-01 3.84080321e-01 1.00812614e+00
8.50533247e-01 9.40529048e-01 5.41701876e-02 1.15419939e-01
1.46815836e+00 -1.33971727e+00 -6.56143785e-01 -3.99402142e-01
9.56727862e-01 -1.35781026e+00 1.15335095e+00 -4.69071805e-01
-1.16493082e+00 -3.68776113e-01 -7.41624892e-01 -8.65731359e-01
-9.91368651e-01 3.66736531e-01 5.59348404e-01 7.24355429e-02
-1.41519344e+00 -1.49870455e-01 -5.64021230e-01 -7.11955428e-01
-2.32848987e-01 1.26186416e-01 -2.53522396e-01 -6.74080551e-01
-1.76569760e+00 1.10504377e+00 3.80382419e-01 -2.45135836e-02
-6.75707459e-01 -6.89502954e-01 -8.01725864e-01 2.56095290e-01
-1.37173146e-01 -5.53427756e-01 1.18499792e+00 -4.48604792e-01
-9.27396536e-01 1.25765634e+00 -3.16057831e-01 -1.43511310e-01
2.09725276e-01 -4.33714747e-01 -6.01698518e-01 2.65217423e-01
7.53448486e-01 9.79155064e-01 3.69430959e-01 -9.38893855e-01
-4.28405672e-01 -3.44608486e-01 -1.11286737e-01 6.52556956e-01
-5.46256244e-01 5.87986290e-01 -1.31628060e+00 -7.52342641e-01
9.72933844e-02 -8.54745090e-01 2.20695719e-01 -2.71990120e-01
-1.15405284e-02 -8.48855317e-01 5.09634852e-01 -8.34514201e-01
1.08875120e+00 -1.90953112e+00 1.79619957e-02 8.16840306e-02
-3.54349881e-01 1.10134192e-01 -7.63613880e-01 1.07547855e+00
3.82686883e-01 2.13463277e-01 3.35258663e-01 -2.62390524e-01
2.63135463e-01 1.72969252e-01 -6.43885672e-01 1.44470364e-01
3.41348886e-01 1.49168932e+00 -1.06930935e+00 -6.50521874e-01
-2.54420608e-01 7.25356519e-01 -4.52964336e-01 1.84305564e-01
-9.96984728e-03 -2.92615771e-01 -6.39073849e-01 6.98757589e-01
3.94682199e-01 -5.22126593e-02 4.13743138e-01 -1.76496208e-01
1.25736520e-01 1.00749159e+00 -1.60217479e-01 2.10904646e+00
-8.58817339e-01 5.71863115e-01 -1.42924368e-01 -5.72756946e-01
6.20418966e-01 3.68456900e-01 2.96511263e-01 -1.53521144e+00
-2.40312532e-01 7.41874099e-01 -3.99153054e-01 -4.23823416e-01
9.43391204e-01 2.08226845e-01 -4.63959634e-01 9.77448404e-01
2.22477138e-01 -4.73792292e-02 3.10207635e-01 3.45640212e-01
6.70121014e-01 1.97882235e-01 -3.31313252e-01 -3.98387253e-01
3.82808775e-01 4.61209603e-02 1.63814008e-01 8.77351880e-01
2.73400813e-01 2.91374028e-01 -1.20259397e-01 -3.59823197e-01
-9.47119951e-01 -1.16506493e+00 -2.29673982e-01 1.98306012e+00
-3.94435823e-02 -3.68963629e-01 -2.52357006e-01 -6.52958989e-01
2.71322876e-01 5.65356493e-01 -4.14571285e-01 -3.86493325e-01
-9.57078457e-01 -5.83201945e-01 8.78112376e-01 3.31083655e-01
1.64090231e-01 -1.09231532e+00 4.69599441e-02 2.69198149e-01
-4.25865144e-01 -1.02230620e+00 -1.14148998e+00 2.59124011e-01
-6.60125494e-01 -5.72671473e-01 -1.15951502e+00 -1.47208703e+00
4.05784726e-01 5.67160010e-01 1.52252185e+00 2.14055166e-01
-1.65808171e-01 4.91117775e-01 -3.17738235e-01 -5.54310791e-02
-8.66055489e-02 6.44830644e-01 1.37852475e-01 -4.94712383e-01
1.03946579e+00 1.27675608e-01 -7.58318186e-01 3.95978875e-02
-1.03833449e+00 -4.47086543e-01 6.80330694e-01 1.11937141e+00
6.43760502e-01 -5.90797007e-01 6.61920249e-01 -4.07492697e-01
1.09137177e+00 -5.14292955e-01 -5.24296939e-01 1.08170497e+00
-7.95973241e-01 3.78705412e-01 2.21809477e-01 -4.16892350e-01
-8.71360064e-01 -7.00395644e-01 1.36815533e-01 -1.23360209e-01
4.81822878e-01 1.00089478e+00 2.86454946e-01 3.68776470e-01
4.87388432e-01 3.19707394e-01 -5.80223978e-01 -8.42384160e-01
7.87402749e-01 9.91778255e-01 1.85256362e-01 -8.37936342e-01
5.25712013e-01 -1.22946100e-02 -7.30534554e-01 -3.93529683e-01
-8.11297596e-01 -7.67067134e-01 -5.28370857e-01 2.74043232e-01
8.42519343e-01 -1.23946190e+00 -2.14639649e-01 1.76474657e-02
-1.41331840e+00 -8.29536319e-02 4.38288376e-02 7.42623627e-01
7.00407997e-02 -2.35395934e-02 -1.08823252e+00 -3.92517090e-01
-7.83607662e-01 -1.14574361e+00 1.65800047e+00 -2.41267577e-01
2.87406542e-03 -1.25936556e+00 4.51722264e-01 3.07572067e-01
7.24360824e-01 -8.85771811e-01 1.34005189e+00 -9.13010240e-01
-4.19895291e-01 -4.27290708e-01 -5.98043740e-01 9.45769250e-02
1.98838696e-01 -3.69385034e-01 -6.89270794e-01 -7.68371046e-01
-4.25813287e-01 -9.73646283e-01 1.14888799e+00 4.13149558e-02
6.11572921e-01 -3.21252793e-01 -4.19301689e-01 3.35340202e-01
1.34234476e+00 -2.67140344e-02 1.97017659e-02 3.42377841e-01
5.29052734e-01 5.68389356e-01 3.95980865e-01 -3.29855949e-01
7.52305746e-01 8.44183445e-01 -3.50162059e-01 -2.17801422e-01
-3.49936724e-01 -4.82687026e-01 4.07158792e-01 1.49209845e+00
6.48796558e-01 -4.10393834e-01 -1.00913918e+00 9.20708835e-01
-1.66341734e+00 -6.69938624e-01 4.66847390e-01 2.03583384e+00
1.24834573e+00 -4.32115167e-01 -4.32109565e-01 -8.53987932e-01
3.22860837e-01 1.72386274e-01 -5.08897066e-01 -4.40420806e-01
-4.52681780e-01 3.77980530e-01 5.76307058e-01 8.45883369e-01
-9.53820884e-01 1.54292548e+00 6.50459528e+00 9.30468798e-01
-1.11094677e+00 2.97682434e-01 3.66164505e-01 -1.90779701e-01
-7.79058754e-01 -1.92291394e-01 -7.27588654e-01 -5.68303913e-02
8.27828228e-01 -3.03752154e-01 5.47952116e-01 4.36138153e-01
-4.03390765e-01 3.46360445e-01 -1.11734533e+00 7.50119150e-01
4.43317711e-01 -1.22964656e+00 4.98289406e-01 -1.48879558e-01
7.30780125e-01 8.10765982e-01 3.38583827e-01 7.91801870e-01
6.10325515e-01 -8.79326046e-01 3.46589684e-01 1.89579144e-01
9.24823165e-01 -6.11738622e-01 7.58676410e-01 -4.32196818e-02
-9.75791156e-01 3.34051132e-01 -6.26980424e-01 6.56371176e-01
-2.14668494e-02 -3.89284268e-02 -9.23802555e-01 6.67897224e-01
8.25972736e-01 6.21396244e-01 -6.87815428e-01 5.04236996e-01
-1.44945368e-01 3.45613807e-01 -2.22338438e-01 -1.26142740e-01
6.54331028e-01 -1.25682682e-01 1.40558884e-01 1.56030381e+00
4.34056252e-01 -4.72684979e-01 2.12457165e-01 7.05477595e-01
-5.63457727e-01 5.92921972e-01 -7.76396275e-01 -3.06775242e-01
6.05934739e-01 1.08841825e+00 -9.87419859e-02 -5.25841236e-01
-4.92441386e-01 1.32522392e+00 7.36710370e-01 8.34014177e-01
-1.34980246e-01 -5.57465792e-01 5.75463653e-01 -4.71587688e-01
1.07941963e-01 -1.28386140e-01 3.13055098e-01 -1.24890208e+00
1.84562311e-01 -8.17789853e-01 6.80525124e-01 -9.74778652e-01
-1.68587244e+00 8.71074915e-01 -7.56345764e-02 -9.37454164e-01
-7.92838871e-01 -4.25514966e-01 8.57216939e-02 1.50248110e+00
-2.00385833e+00 -1.66515756e+00 6.82339668e-01 4.75847423e-01
4.94093329e-01 -5.17210484e-01 1.31022871e+00 7.08975911e-01
-1.90550432e-01 9.68605399e-01 7.93134034e-01 4.83826071e-01
1.30808890e+00 -8.63494158e-01 3.93362433e-01 4.79839355e-01
4.79240149e-01 1.16467786e+00 5.58459200e-03 -5.88047981e-01
-1.54916787e+00 -9.79926884e-01 1.93653059e+00 -5.59863627e-01
9.82087612e-01 -4.97661322e-01 -9.19927716e-01 7.72864223e-01
9.19100046e-01 -1.83690295e-01 8.68935943e-01 4.50767875e-01
-9.53013241e-01 -1.47488847e-01 -6.82275116e-01 7.43892848e-01
6.91784501e-01 -1.41304743e+00 -8.33166242e-01 7.83634186e-01
1.23248863e+00 -3.54487896e-02 -8.46032262e-01 2.16480121e-01
7.76092350e-01 -6.07141182e-02 1.28560925e+00 -1.07440567e+00
1.21786803e-01 4.72769290e-02 -5.36901593e-01 -1.32189441e+00
-4.36952710e-01 -4.80921417e-02 2.31281951e-01 1.01089358e+00
6.48177505e-01 -6.44309402e-01 -7.09249899e-02 2.21262559e-01
-2.62521319e-02 -6.15627706e-01 -1.05798948e+00 -7.71075845e-01
7.75992215e-01 -1.95061833e-01 7.38240421e-01 1.03877079e+00
-2.45244578e-02 7.17583120e-01 -9.31994617e-02 -7.71997944e-02
3.01328450e-01 3.39087874e-01 2.26386666e-01 -8.33736420e-01
-1.89219981e-01 -6.11731410e-01 3.20860326e-01 -1.23078203e+00
7.50834107e-01 -1.67955410e+00 1.25484452e-01 -1.63635838e+00
4.15276051e-01 -5.68956792e-01 -8.07275712e-01 5.30996561e-01
-2.23838955e-01 5.12260675e-01 -2.02214152e-01 4.86586004e-01
-7.83899426e-01 6.70450091e-01 8.98995757e-01 -6.17606223e-01
9.10082310e-02 -6.66694224e-01 -5.30319870e-01 -3.10100950e-02
4.71352607e-01 -6.65955007e-01 -2.50453264e-01 -1.68360400e+00
5.40405989e-01 -8.59640464e-02 -3.80718149e-02 -6.59020320e-02
2.61252493e-01 9.01565552e-02 4.03613687e-01 -6.96121275e-01
3.58613849e-01 -6.90252423e-01 -5.12676120e-01 3.31606656e-01
-1.10755706e+00 5.74074745e-01 1.01981021e-01 2.08145231e-01
-3.17605287e-01 -3.68625782e-02 1.06248043e-01 8.17698799e-03
-2.92840809e-01 3.44763637e-01 -3.09513807e-01 2.50722706e-01
2.24636588e-03 6.20424509e-01 -5.71254313e-01 -3.64114493e-01
-4.85111214e-02 6.69634700e-01 3.22891027e-01 1.15212333e+00
4.65308070e-01 -1.90311134e+00 -1.03303719e+00 3.12756896e-01
8.68445814e-01 -5.93559325e-01 -1.06458843e-01 3.81889820e-01
-2.99178153e-01 1.43394732e+00 1.36187255e-01 -3.50713342e-01
-7.19215631e-01 5.13397932e-01 1.49538025e-01 -5.95155954e-01
-3.99416536e-02 8.96884680e-01 1.71128243e-01 -1.07335472e+00
2.90834337e-01 8.44117552e-02 -1.27436310e-01 1.10881105e-01
4.82790887e-01 -6.52574003e-02 3.20402056e-01 -9.43412840e-01
-5.35018921e-01 6.87856734e-01 -6.37825191e-01 -6.49329722e-01
1.13474202e+00 -3.60044569e-01 -5.72182000e-01 4.05030280e-01
2.02638817e+00 7.14102834e-02 -7.07472637e-02 -9.57722425e-01
4.35195208e-01 -1.45362899e-01 3.28566521e-01 -1.00537646e+00
-9.38111424e-01 8.98709476e-01 7.18804061e-01 -2.57273108e-01
8.82046640e-01 4.44582015e-01 9.46159780e-01 1.00660706e+00
3.37701738e-01 -1.40321124e+00 2.14964002e-02 8.40801179e-01
1.07097542e+00 -1.46445382e+00 -2.98272252e-01 5.29502630e-01
-2.91266739e-01 8.67255986e-01 3.70288253e-01 7.73936063e-02
5.44833899e-01 -3.77110451e-01 6.03413224e-01 -3.53787988e-01
-9.26915526e-01 -3.36148322e-01 9.42857265e-01 3.25370491e-01
1.03581393e+00 -2.52993498e-02 -6.47515774e-01 1.51985705e-01
-1.07568000e-02 -3.31014365e-01 -4.29064631e-01 8.15760612e-01
-1.06673487e-01 -1.52063739e+00 -5.33336699e-02 3.77194613e-01
-6.79503858e-01 -8.80485296e-01 -4.86385137e-01 5.71191490e-01
-5.76637447e-01 9.14991856e-01 3.19452852e-01 -1.10216439e-01
3.95384878e-02 3.66381347e-01 1.83713943e-01 -5.62250912e-01
-6.85468614e-01 4.56397563e-01 3.21446359e-02 -5.20705402e-01
-1.77003056e-01 -2.79155731e-01 -8.66854668e-01 7.39826187e-02
-2.47445121e-01 4.72220391e-01 7.74759173e-01 6.65086091e-01
6.82449937e-01 1.69137537e-01 5.45173943e-01 -4.44133401e-01
-6.24459505e-01 -1.32138014e+00 -2.15311959e-01 3.59954923e-01
2.96399862e-01 -1.72093108e-01 -1.05586089e-02 -4.84285325e-01] | [11.378799438476562, 9.810039520263672] |
cffb8cb1-2bed-4e0c-afb9-839436863dd8 | fine-grained-visual-classification-via-2 | null | null | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10042971 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10042971 | Fine-Grained Visual Classification via Internal Ensemble Learning Transformer | Recently, vision transformers (ViTs) have been investigated in fine-grained visual recognition (FGVC) and are now
considered state of the art. However, most ViT-based works ignore
the different learning performances of the heads in the multihead self-attention (MHSA) mechanism and its layers. To address
these issues, in this paper, we propose a novel internal ensemble
learning transformer (IELT) for FGVC. The proposed IELT
involves three main modules: multi-head voting (MHV) module,
cross-layer refinement (CLR) module, and dynamic selection (DS)
module. To solve the problem of the inconsistent performances of
multiple heads, we propose the MHV module, which considers
all of the heads in each layer as weak learners and votes for
tokens of discriminative regions as cross-layer feature based on
the attention maps and spatial relationships. To effectively mine
the cross-layer feature and suppress the noise, the CLR module
is proposed, where the refined feature is extracted and the assist
logits operation is developed for the final prediction. In addition,
a newly designed DS module adjusts the token selection number
at each layer by weighting their contributions of the refined
feature. In this way, the idea of ensemble learning is combined
with the ViT to improve fine-grained feature representation. The
experiments demonstrate that our method achieves competitive
results compared with the state of the art on five popular FGVC
datasets. Source code has been released and can be found at
https://github.com/mobulan/IELT. | ['Bin Luo', 'Bo Jiang', 'Jiahui Wang', 'Qin Xu'] | 2023-02-13 | null | null | null | ieee-transactions-on-multimedia-2023-2 | ['fine-grained-visual-recognition', 'fine-grained-image-classification'] | ['computer-vision', 'computer-vision'] | [-1.11182585e-01 -3.52590799e-01 3.21714655e-02 -2.71616548e-01
-4.29593623e-01 -9.06101540e-02 4.92971867e-01 -2.04338804e-01
-5.55747926e-01 6.35800660e-01 3.04066300e-01 5.82468770e-02
-9.80847180e-02 -7.02775896e-01 -5.07908106e-01 -1.12742698e+00
5.09159863e-01 -1.58170182e-02 6.50681734e-01 -3.14685591e-02
3.86772662e-01 4.20081705e-01 -1.85900855e+00 5.59954286e-01
1.17399526e+00 1.33292687e+00 6.50294006e-01 3.11319023e-01
-4.48610961e-01 9.11237478e-01 -4.36207741e-01 -1.39761716e-01
-1.42922655e-01 -3.00961882e-01 -4.88882869e-01 -9.81717557e-02
2.45771766e-01 -1.56106398e-01 -1.70900822e-01 9.80122626e-01
8.44287157e-01 7.06187412e-02 6.44417584e-01 -1.19781876e+00
-4.00937349e-01 5.38523734e-01 -8.01872611e-01 3.77378851e-01
-2.72160530e-01 5.94274774e-02 8.58949661e-01 -1.35232508e+00
9.77380723e-02 1.09422183e+00 4.65529561e-01 5.28729558e-01
-7.64773309e-01 -1.01577806e+00 5.50969660e-01 7.53113568e-01
-1.56189299e+00 -6.09083176e-01 8.47019255e-01 -3.15890908e-01
7.68481076e-01 1.00091264e-01 6.26968384e-01 7.54788756e-01
1.86787128e-01 1.06398189e+00 1.07171452e+00 -4.60188657e-01
2.93382760e-02 2.98323423e-01 2.09950030e-01 8.78695548e-01
5.54076880e-02 -9.92440730e-02 -4.95062202e-01 3.00202519e-01
6.45283222e-01 9.12029520e-02 -1.91315368e-01 -2.22093299e-01
-9.04129028e-01 6.93935275e-01 7.37797320e-01 6.50681496e-01
-5.20144463e-01 -5.15279267e-03 2.63714820e-01 -7.77880475e-02
4.75078046e-01 -2.66845375e-02 -4.20843959e-01 3.26338142e-01
-9.54632521e-01 -1.39086947e-01 3.23343217e-01 6.39219642e-01
8.59650433e-01 -1.17584858e-02 -7.28107750e-01 1.10114980e+00
5.36949337e-01 2.92933017e-01 6.09922588e-01 -5.65625072e-01
4.23617393e-01 9.99757767e-01 -2.69301295e-01 -8.92929435e-01
-2.17475116e-01 -8.84085655e-01 -9.83433962e-01 2.36306593e-01
-7.15360790e-02 -8.83717090e-02 -1.06311572e+00 1.58041179e+00
3.60459745e-01 3.59360367e-01 -2.37793505e-01 9.02563274e-01
1.24710071e+00 5.95538974e-01 3.31891060e-01 -2.21740484e-01
1.33434451e+00 -1.15504515e+00 -5.48715532e-01 -1.36792764e-01
3.35188150e-01 -6.48543954e-01 7.38590658e-01 2.42702872e-01
-9.19032991e-01 -9.26929593e-01 -9.08904076e-01 6.60986230e-02
-3.27907592e-01 5.39493322e-01 3.88470620e-01 2.70096213e-01
-9.18734848e-01 1.76666796e-01 -5.37073970e-01 -7.26194382e-02
7.46841669e-01 4.02825326e-01 -1.43929586e-01 2.81366520e-03
-1.13885403e+00 7.63664126e-01 2.82599449e-01 4.78188485e-01
-8.17287683e-01 -4.75131661e-01 -6.01665378e-01 1.50143534e-01
2.94543505e-01 -7.00702190e-01 9.58136559e-01 -1.22317934e+00
-1.20028567e+00 5.14018476e-01 -4.85338151e-01 1.03512779e-02
2.59652525e-01 -6.04561865e-02 -3.40797901e-01 -1.85166523e-01
1.39309675e-01 5.53981662e-01 1.03050721e+00 -1.26059568e+00
-1.01506734e+00 -6.34917974e-01 -2.38634571e-01 4.02475178e-01
-4.34395134e-01 -1.46711506e-02 -7.04379201e-01 -6.31708920e-01
2.57531721e-02 -4.89112169e-01 -7.98398033e-02 -3.66165280e-01
-2.49778360e-01 -6.13599181e-01 8.58398318e-01 -6.00895047e-01
1.62477291e+00 -2.26936388e+00 3.24697495e-01 1.65509194e-01
2.66553968e-01 6.73375130e-01 -1.33522481e-01 5.45479022e-02
9.82806385e-02 -5.45694008e-02 -5.72436117e-02 -4.16918874e-01
-1.73371375e-01 9.78081580e-03 4.82700765e-03 1.93306029e-01
2.34337717e-01 9.80572402e-01 -5.85402071e-01 -8.71075392e-01
5.70469320e-01 5.88353813e-01 -3.63888323e-01 1.89108327e-01
7.60420645e-03 4.58456606e-01 -8.28809142e-01 5.92553496e-01
8.90483737e-01 -1.88577116e-01 -2.38598380e-02 -7.09329963e-01
-5.66802621e-01 -1.18659258e-01 -1.33505535e+00 1.30520940e+00
-3.49745303e-01 1.61075965e-01 3.80846560e-02 -7.77719915e-01
9.94789422e-01 1.13323852e-01 1.45990029e-01 -9.29680169e-01
2.83739567e-01 1.82430819e-01 2.36013997e-02 -4.90946501e-01
3.89247924e-01 8.16361979e-02 9.42875221e-02 1.35269582e-01
2.83959836e-01 4.57547575e-01 -1.92213595e-01 3.61514576e-02
7.74481595e-01 2.32749358e-01 3.10509950e-01 -3.72192217e-03
1.09044886e+00 -2.86773860e-01 8.38236153e-01 5.64256430e-01
-3.17414641e-01 3.81379634e-01 -2.88054664e-02 -2.98887640e-01
-7.20077693e-01 -8.30915213e-01 6.63965046e-02 1.27211308e+00
3.61954033e-01 -2.59224147e-01 -7.46418834e-01 -8.41834724e-01
4.09537274e-03 5.39760053e-01 -7.34724581e-01 -2.86288530e-01
-4.61170435e-01 -6.04604185e-01 4.27957177e-01 6.85060561e-01
1.00272858e+00 -1.46094322e+00 -4.90900010e-01 2.15805903e-01
-1.47737160e-01 -6.69436276e-01 -4.76973951e-01 1.74841806e-01
-5.58921278e-01 -7.88163245e-01 -7.70079494e-01 -8.43251109e-01
5.31574249e-01 4.55542058e-01 6.84425890e-01 2.48780653e-01
-7.13147074e-02 -4.31785872e-03 -4.84631509e-01 -3.18446040e-01
1.78317115e-01 2.28606358e-01 -1.52460188e-01 4.86514509e-01
4.63990271e-01 -3.49002868e-01 -7.87327290e-01 3.49436551e-01
-4.98258144e-01 1.93906352e-01 8.31266403e-01 9.68742013e-01
7.33267844e-01 9.90755558e-02 6.33598983e-01 -6.50178552e-01
3.99845660e-01 -3.75697762e-01 -3.90791446e-01 4.40959007e-01
-5.64035654e-01 1.37289196e-01 7.18333304e-01 -2.52029538e-01
-1.43568957e+00 -2.54819579e-02 -2.01334104e-01 -6.69077933e-01
-1.92433044e-01 2.90245056e-01 -4.99841034e-01 -2.55261540e-01
9.83173177e-02 4.71580356e-01 -4.18435782e-01 -6.43431962e-01
9.52547044e-02 8.39676499e-01 3.09985310e-01 -2.28449062e-01
5.04692316e-01 1.24484681e-01 -4.33148354e-01 -5.39849401e-01
-8.68031442e-01 -4.28070605e-01 -5.99184752e-01 -4.81031984e-01
9.25297976e-01 -9.72447157e-01 -5.91945946e-01 8.20384920e-01
-9.78755057e-01 -1.33081734e-01 -1.30812243e-01 5.04609406e-01
-5.15525751e-02 8.22374076e-02 -3.45174819e-01 -9.56809878e-01
-5.17584145e-01 -1.22431755e+00 1.06937921e+00 8.32648575e-01
2.48365238e-01 -6.48834109e-01 -1.50122777e-01 3.79225671e-01
5.17273843e-01 -2.14425310e-01 6.67773664e-01 -3.75261337e-01
-6.28208280e-01 3.46524388e-01 -5.48244655e-01 4.45683151e-01
5.96448742e-02 -3.91285028e-03 -1.26818788e+00 -1.73148677e-01
-1.97118521e-01 -2.05284238e-01 1.49689507e+00 6.34779751e-01
1.37166798e+00 -4.59122099e-02 -5.50928533e-01 7.39490926e-01
1.46804810e+00 3.43980134e-01 6.19443297e-01 3.63438815e-01
9.67996836e-01 4.66150939e-01 6.33548796e-01 5.85521877e-01
6.54387534e-01 6.00229681e-01 4.23418164e-01 -2.11210176e-01
-3.33102584e-01 -1.68538541e-01 4.21977073e-01 9.28091884e-01
-3.46941262e-01 -1.77606925e-01 -5.47475815e-01 4.69151318e-01
-1.93737721e+00 -1.10193932e+00 -2.18846519e-02 1.99780905e+00
5.50841868e-01 -2.44251061e-02 -1.02043174e-01 1.58929899e-01
9.52492118e-01 3.92569095e-01 -6.93658054e-01 -3.40717852e-01
-2.76671529e-01 2.97341347e-01 3.24280143e-01 2.49300361e-01
-1.20656562e+00 1.06674659e+00 4.03855944e+00 1.21100771e+00
-1.20240653e+00 3.31294149e-01 4.94935095e-01 6.73847869e-02
-2.83315241e-01 -1.80274799e-01 -1.05064404e+00 6.24177754e-01
2.25246772e-01 2.88893759e-01 3.11077505e-01 4.50586379e-01
1.36751145e-01 -1.30692229e-01 -4.49105412e-01 9.12453353e-01
2.09482431e-01 -1.12762725e+00 1.62795812e-01 -1.27955258e-01
4.96627867e-01 1.27290890e-01 9.73427445e-02 4.17364866e-01
3.08657717e-02 -8.47820222e-01 7.71642566e-01 9.12792563e-01
7.76983202e-01 -8.69901359e-01 9.06983197e-01 4.41068143e-01
-1.68987000e+00 -3.91303837e-01 -4.11656529e-01 3.31838280e-01
-1.40763536e-01 5.60346425e-01 -3.66268903e-01 8.25363040e-01
9.91831064e-01 7.29707956e-01 -7.72743821e-01 1.18269312e+00
-2.96125442e-01 6.26526356e-01 -7.81129971e-02 3.65715809e-02
2.04479650e-01 9.03742611e-02 5.14215529e-01 1.19432843e+00
2.91628629e-01 7.90528506e-02 5.68140000e-02 6.40771031e-01
4.68868427e-02 1.51806697e-01 -4.45857979e-02 4.18276519e-01
5.96532106e-01 1.51356733e+00 -5.68648160e-01 -2.86315352e-01
-3.57964128e-01 9.63368773e-01 6.38712525e-01 2.60829002e-01
-8.14026654e-01 -4.32253301e-01 4.06027585e-01 -1.31210804e-01
8.17033231e-01 2.62079924e-01 -2.05844879e-01 -1.20666993e+00
1.82856582e-02 -7.21868873e-01 5.59754848e-01 -7.49911845e-01
-1.25001860e+00 8.01338315e-01 -3.38089764e-01 -1.17009258e+00
2.42386147e-01 -2.40570694e-01 -7.94461191e-01 1.03686118e+00
-1.74922514e+00 -1.30953264e+00 -5.58631539e-01 9.50006485e-01
5.99113464e-01 -3.17443192e-01 6.12094641e-01 4.37511057e-01
-9.27399576e-01 7.52841830e-01 -2.07067221e-01 -2.26975791e-02
6.57102406e-01 -9.81000125e-01 -5.31102009e-02 8.12014222e-01
-1.41245410e-01 3.65497828e-01 1.30875677e-01 -6.36245549e-01
-9.61204231e-01 -1.36525738e+00 9.08885717e-01 9.90680382e-02
1.86168581e-01 -2.00591564e-01 -1.01223898e+00 3.90972704e-01
1.22933879e-01 1.24346495e-01 3.35212529e-01 -6.87699169e-02
-2.76324987e-01 -4.82437134e-01 -9.44865763e-01 2.89489061e-01
9.47927713e-01 -3.42744082e-01 -5.61774313e-01 -3.45526397e-01
4.01450992e-01 -2.99453400e-02 -5.67890406e-01 5.87005973e-01
7.12947726e-01 -1.03157675e+00 7.00008631e-01 -1.30737022e-01
1.84663340e-01 -7.76874423e-01 -1.14703603e-01 -1.32344508e+00
-9.08726573e-01 1.75511375e-01 -1.13261186e-01 1.57466662e+00
2.81648725e-01 -5.74131548e-01 5.21293819e-01 5.81191815e-02
-3.14511001e-01 -9.66740549e-01 -9.23769712e-01 -1.89313680e-01
-3.35089236e-01 -2.80919284e-01 6.79244995e-01 6.67798638e-01
-4.41661507e-01 5.51813900e-01 -3.24669361e-01 5.63417859e-02
6.29197598e-01 2.72036642e-01 4.31899071e-01 -1.25374222e+00
-1.28761932e-01 -4.69969153e-01 -1.82703808e-01 -8.86252761e-01
1.04373932e-01 -9.65167165e-01 -2.29187682e-02 -1.71070230e+00
4.82566357e-01 -6.32587492e-01 -8.39789569e-01 5.62915564e-01
-4.36161101e-01 2.55633920e-01 4.15105253e-01 1.92463979e-01
-9.28938091e-01 8.18308175e-01 1.39379752e+00 -1.88304886e-01
-1.86029702e-01 5.90679422e-02 -8.24233651e-01 7.90365160e-01
8.20975125e-01 -3.08184266e-01 -2.82303363e-01 -4.03227448e-01
-3.68171453e-01 -2.49107987e-01 4.43826109e-01 -1.05051911e+00
6.39833987e-01 8.54633898e-02 1.01994812e+00 -8.40123057e-01
2.10225984e-01 -7.28271663e-01 6.91488087e-02 4.10802633e-01
-1.55388415e-01 -1.48757964e-01 1.45769387e-01 2.44158924e-01
-2.35379383e-01 2.72519253e-02 8.26639414e-01 -1.65213138e-01
-1.15200746e+00 5.01316428e-01 -1.81771159e-01 -1.00045152e-01
1.01933241e+00 -3.63632202e-01 -3.73485625e-01 -6.30442286e-04
-5.00482976e-01 5.24504185e-01 1.52222350e-01 3.64841729e-01
8.95539224e-01 -1.29423952e+00 -8.98629963e-01 5.12127101e-01
1.74234673e-01 -8.37969333e-02 8.15954268e-01 9.40693021e-01
1.11832224e-01 2.48692572e-01 -3.05797040e-01 -4.96816367e-01
-1.34256351e+00 3.18430126e-01 4.70133781e-01 -4.21171397e-01
-2.61188805e-01 1.00705600e+00 5.70341945e-01 -2.67222613e-01
2.28002518e-01 1.09640338e-01 -9.29903030e-01 3.79237950e-01
5.70782065e-01 3.56260777e-01 1.23999175e-02 -9.12144363e-01
-6.98093891e-01 8.53115439e-01 -2.95875490e-01 2.08759859e-01
1.27081525e+00 -3.11602980e-01 -1.49594918e-01 3.09322715e-01
9.38932121e-01 5.25706671e-02 -1.25182164e+00 -4.54440832e-01
-3.15033793e-01 -2.20209241e-01 3.00477594e-01 -8.47693026e-01
-1.37039685e+00 8.11531961e-01 7.74765611e-01 -2.70657539e-01
1.58008063e+00 7.18585253e-02 6.98297799e-01 -1.96860656e-01
3.02353054e-01 -9.99579787e-01 -1.71708018e-01 6.44880176e-01
7.92525411e-01 -1.01593363e+00 -1.50656015e-01 -1.59427628e-01
-7.97854245e-01 7.32414603e-01 1.05947673e+00 -1.34366632e-01
7.46950328e-01 3.07296574e-01 -4.87253927e-02 -5.58829401e-03
-9.42032516e-01 -4.67567742e-01 4.31630671e-01 3.32029045e-01
3.81764919e-01 1.35656074e-01 -4.62421089e-01 9.18124020e-01
1.55685514e-01 1.29756689e-01 -1.82421729e-01 6.50188804e-01
-6.35358989e-01 -9.65725064e-01 -2.09202930e-01 6.39954269e-01
-3.46701235e-01 -2.75401682e-01 -2.32728690e-01 4.31935936e-01
8.54160190e-01 1.04117954e+00 1.05939649e-01 -8.95818174e-01
3.30263078e-01 -7.51557723e-02 5.16866744e-01 -3.95187825e-01
-1.00196218e+00 1.39993891e-01 -1.24324366e-01 -6.03329360e-01
-6.82725668e-01 -5.21315277e-01 -1.14544249e+00 -6.23378642e-02
-5.52623391e-01 3.03801030e-01 2.59171933e-01 1.03689599e+00
4.09684569e-01 8.40270340e-01 8.19478273e-01 -9.03365195e-01
-2.16663092e-01 -1.04773581e+00 -5.85012913e-01 -2.44722161e-02
3.71505439e-01 -9.51459944e-01 -2.15167537e-01 -3.17850202e-01] | [9.641074180603027, 1.9649802446365356] |
bff85f59-7c1a-4c5c-8c27-1a6a537ba0d9 | doc2graph-a-task-agnostic-document | 2208.11168 | null | https://arxiv.org/abs/2208.11168v1 | https://arxiv.org/pdf/2208.11168v1.pdf | Doc2Graph: a Task Agnostic Document Understanding Framework based on Graph Neural Networks | Geometric Deep Learning has recently attracted significant interest in a wide range of machine learning fields, including document analysis. The application of Graph Neural Networks (GNNs) has become crucial in various document-related tasks since they can unravel important structural patterns, fundamental in key information extraction processes. Previous works in the literature propose task-driven models and do not take into account the full power of graphs. We propose Doc2Graph, a task-agnostic document understanding framework based on a GNN model, to solve different tasks given different types of documents. We evaluated our approach on two challenging datasets for key information extraction in form understanding, invoice layout analysis and table detection. Our code is freely accessible on https://github.com/andreagemelli/doc2graph. | ['Simone Marinai', 'Josep Lladós', 'Enrico Civitelli', 'Sanket Biswas', 'Andrea Gemelli'] | 2022-08-23 | null | null | null | null | ['document-layout-analysis', 'table-detection', 'semantic-entity-labeling', 'key-information-extraction'] | ['computer-vision', 'miscellaneous', 'natural-language-processing', 'natural-language-processing'] | [ 1.38751611e-01 7.95909092e-02 -2.40284085e-01 -1.63386986e-01
-1.90294832e-01 -9.82522726e-01 7.72908866e-01 6.76869631e-01
-1.41900796e-02 2.18299791e-01 2.76386976e-01 -7.34791338e-01
-5.93553424e-01 -1.24391079e+00 -6.78315341e-01 -1.49695054e-01
6.26227930e-02 7.13911176e-01 -1.35664195e-02 -3.30653965e-01
6.00841641e-01 7.17561662e-01 -9.73269045e-01 6.64726734e-01
8.23892355e-01 8.36773813e-01 1.51658431e-01 6.00605965e-01
-6.04206085e-01 8.23937654e-01 -5.62657952e-01 -6.60150170e-01
1.54430330e-01 -3.39585096e-02 -1.10411346e+00 -5.54631576e-02
7.11593151e-01 -4.14043926e-02 -5.52533329e-01 1.09756148e+00
1.91025496e-01 5.78298382e-02 6.19095325e-01 -1.12094367e+00
-9.22836185e-01 1.21365249e+00 -6.16677821e-01 1.59373745e-01
2.27260441e-01 -1.12427078e-01 1.47724545e+00 -6.76498771e-01
7.96806395e-01 1.44632173e+00 4.98933911e-01 2.19608665e-01
-1.13517940e+00 -3.75830382e-01 3.77001524e-01 5.49840868e-01
-1.19071531e+00 1.46240508e-02 1.00931585e+00 -3.84727597e-01
8.92370760e-01 2.05040902e-01 5.92642486e-01 9.28593993e-01
1.27062678e-01 1.24106228e+00 6.39832675e-01 -4.79428977e-01
-3.16591375e-02 -2.42013440e-01 5.96081197e-01 7.60991454e-01
6.73963547e-01 -4.86766785e-01 -4.54615712e-01 2.17814356e-01
6.87898815e-01 3.15746628e-02 -2.63962269e-01 -4.99869019e-01
-7.91244984e-01 9.12105203e-01 7.75190353e-01 5.12713790e-01
-2.07189649e-01 3.47490996e-01 6.16475105e-01 2.73223579e-01
2.59742022e-01 7.66329765e-01 -4.87260133e-01 4.79052681e-03
-7.61955261e-01 3.95244420e-01 9.34428155e-01 1.03829896e+00
6.15601361e-01 -4.10475172e-02 -2.03447133e-01 7.70558000e-01
2.59954184e-01 1.38159499e-01 5.50325625e-02 -3.97102594e-01
7.50143170e-01 1.16170931e+00 -4.75232422e-01 -1.47518349e+00
-5.75709641e-01 -6.60086155e-01 -9.56220925e-01 -3.36569369e-01
5.31012177e-01 2.45449960e-01 -9.37094748e-01 1.19996250e+00
-4.31187227e-02 -3.93729985e-01 -3.52265626e-01 5.87629974e-01
1.11786342e+00 4.96235400e-01 -1.24560662e-01 6.02373719e-01
1.47217274e+00 -8.87415051e-01 -6.49178386e-01 -2.41022006e-01
7.87934184e-01 -7.39546955e-01 1.06578708e+00 8.52431238e-01
-8.27601850e-01 -4.58784550e-01 -1.12762618e+00 -4.97722179e-01
-9.58385408e-01 1.78516820e-01 8.87107313e-01 6.07408583e-01
-1.10313261e+00 6.94135368e-01 -5.79956591e-01 -5.42603791e-01
8.76479149e-01 3.77060354e-01 -1.53615341e-01 -4.40882564e-01
-8.85416925e-01 5.11798441e-01 6.31756425e-01 3.05334121e-01
-5.48709810e-01 -5.20962000e-01 -8.28513861e-01 4.23609197e-01
8.58220279e-01 -8.18376422e-01 1.11448753e+00 -4.03585494e-01
-8.45630884e-01 7.25513518e-01 7.69627988e-02 -4.40549433e-01
4.29756463e-01 -5.92569888e-01 -2.62889653e-01 5.90566806e-02
-2.06969008e-01 2.34783411e-01 5.55372775e-01 -9.76610005e-01
-1.03978261e-01 -7.09305882e-01 2.37370297e-01 -9.41580907e-02
-4.77707237e-01 -2.12402195e-01 -7.16953635e-01 -7.20528901e-01
1.20764621e-01 -5.97338259e-01 -3.77077125e-02 -1.75746456e-01
-1.09769070e+00 -4.16203648e-01 8.87475610e-01 -6.48895621e-01
1.28870046e+00 -1.64692664e+00 2.62022257e-01 4.10261035e-01
6.58015966e-01 3.87916148e-01 -3.20094764e-01 8.48307908e-01
7.13809505e-02 4.46214050e-01 -1.18980363e-01 -1.42322451e-01
1.72387511e-01 -1.07091978e-01 -3.47712040e-01 2.83130351e-03
1.05793759e-01 1.37303185e+00 -6.39307559e-01 -1.76746339e-01
2.48042181e-01 2.93142796e-01 -3.12813610e-01 -9.55758765e-02
-7.57987857e-01 -1.05571590e-01 -6.12888038e-01 5.07719338e-01
9.01179731e-01 -6.26518071e-01 6.10679865e-01 -3.41579527e-01
2.10488334e-01 3.51628572e-01 -1.22623003e+00 1.84123397e+00
-3.38215500e-01 8.76459360e-01 -8.03514346e-02 -1.19880354e+00
9.15332794e-01 -2.03384548e-01 1.94607168e-01 -7.38246262e-01
2.72315085e-01 -1.43425511e-02 6.20918050e-02 -2.98062354e-01
8.36605847e-01 5.17817140e-01 -2.72141155e-02 5.63079655e-01
1.29123434e-01 -2.19999343e-01 6.50134206e-01 8.46492112e-01
1.22593391e+00 -1.07661113e-01 1.07252665e-01 -3.65090966e-01
5.59545219e-01 -1.09849170e-01 -2.31050700e-03 7.68896043e-01
4.57980812e-01 5.32436788e-01 9.73110437e-01 -6.43805206e-01
-9.05917525e-01 -7.78815627e-01 1.52063757e-01 9.08539414e-01
-4.89116758e-02 -1.07202864e+00 -7.46515512e-01 -7.30083287e-01
2.78120995e-01 5.18961847e-01 -6.51092172e-01 2.08132230e-02
-7.86113381e-01 -5.59637606e-01 4.71698344e-01 7.04567373e-01
2.56137103e-01 -1.25688517e+00 3.99979763e-02 2.32785285e-01
3.91595438e-02 -1.29068732e+00 -2.45311171e-01 1.11008093e-01
-9.24008846e-01 -1.67467904e+00 -4.06153053e-01 -6.33612812e-01
7.12497950e-01 2.85186738e-01 1.35408258e+00 4.53400612e-01
-4.89650548e-01 5.61263442e-01 -3.34081978e-01 -5.37294924e-01
-4.18068558e-01 8.05125713e-01 -4.99778926e-01 -1.53419927e-01
5.63745618e-01 -4.50009376e-01 -3.54611605e-01 -3.79599854e-02
-1.13497627e+00 1.32484153e-01 4.25497264e-01 3.99461269e-01
2.78387845e-01 7.96681494e-02 2.25977093e-01 -1.36786509e+00
1.11918712e+00 -2.23399222e-01 -8.85418236e-01 3.83421034e-01
-5.79665124e-01 3.79790604e-01 8.03321123e-01 9.95372757e-02
-6.02590322e-01 -5.31013370e-01 1.41213499e-02 -5.44576161e-02
-2.25205019e-01 8.09138238e-01 -5.42734981e-01 1.52914792e-01
3.86308193e-01 1.07127316e-01 -4.07497585e-01 -7.25803494e-01
5.49298584e-01 1.78524822e-01 3.23140979e-01 -7.33410656e-01
9.60113168e-01 3.94010663e-01 3.76207322e-01 -8.26184928e-01
-8.24521661e-01 -5.05715251e-01 -9.35694635e-01 8.55208188e-02
6.72307193e-01 -3.50652725e-01 -7.85037756e-01 6.01283610e-01
-1.41891909e+00 -4.59778309e-01 -1.53851067e-03 -8.70345756e-02
-1.61837533e-01 5.22610903e-01 -5.88542461e-01 -4.19270009e-01
-6.61745727e-01 -8.42923760e-01 1.06488574e+00 2.31910557e-01
-1.87408045e-01 -1.43613064e+00 -2.83662193e-02 3.82204294e-01
2.33514771e-01 2.04816803e-01 1.44834578e+00 -7.80216932e-01
-1.15134180e+00 -1.32714450e-01 -6.78551733e-01 3.16702984e-02
1.73711866e-01 2.33342294e-02 -7.20633447e-01 -7.15261921e-02
-7.42908955e-01 -1.16799138e-01 1.28402042e+00 3.65271419e-01
1.59147298e+00 -3.57705116e-01 -4.07769442e-01 7.63864517e-01
1.57983541e+00 -1.93465110e-02 6.42429113e-01 5.26972473e-01
1.34974647e+00 6.52880549e-01 2.27217916e-02 3.08413029e-01
4.01939958e-01 5.60710669e-01 7.07392335e-01 -1.43899862e-02
-1.80918962e-01 -4.14839596e-01 -1.68275863e-01 6.15556061e-01
-4.27515171e-02 -9.79205012e-01 -1.18303311e+00 6.45916939e-01
-1.84301829e+00 -6.28659010e-01 -5.76445520e-01 1.51861954e+00
2.85771728e-01 3.29394221e-01 -1.36221081e-01 3.40229452e-01
5.36074102e-01 4.24718916e-01 -4.04855967e-01 -5.45996785e-01
-2.18576401e-01 5.00016510e-01 3.02786499e-01 2.02360824e-01
-1.11805546e+00 1.23270500e+00 4.80019331e+00 7.92008340e-01
-8.19558084e-01 -5.04772305e-01 5.36019921e-01 2.92983353e-01
-5.40249348e-01 -4.45613898e-02 -8.11328292e-01 1.08708087e-02
5.02256095e-01 -1.55429468e-01 3.76541704e-01 7.94979572e-01
7.32406462e-03 -2.59590615e-02 -1.22417367e+00 8.36502373e-01
9.01960135e-02 -1.76899517e+00 6.21377766e-01 2.47344211e-01
5.79240739e-01 -8.01345780e-02 1.15168802e-01 -4.20102105e-02
4.31629002e-01 -1.27900493e+00 4.91459996e-01 4.56918746e-01
4.59457248e-01 -8.12583983e-01 5.80630183e-01 2.52227839e-02
-1.35809076e+00 6.23340830e-02 -2.68159956e-01 1.10282876e-01
8.14666599e-02 6.61854506e-01 -1.01091766e+00 8.82460892e-01
6.04974687e-01 1.01798046e+00 -1.00265336e+00 1.03964162e+00
-6.66500211e-01 5.93975961e-01 -7.60594830e-02 -2.28972018e-01
3.60890657e-01 -3.76893610e-01 4.75154459e-01 1.18441761e+00
4.23643142e-02 -2.64018595e-01 -1.72898844e-02 1.10810721e+00
-5.17394364e-01 3.74599874e-01 -6.41885340e-01 -4.98055786e-01
1.91387743e-01 1.48237109e+00 -1.24772823e+00 -1.84204593e-01
-1.62478805e-01 6.98797464e-01 2.82165170e-01 2.71216691e-01
-2.63388366e-01 -6.14479721e-01 2.54052222e-01 3.74756038e-01
5.00425160e-01 -6.16809845e-01 -5.24665833e-01 -1.03961086e+00
1.92680553e-01 -8.71943116e-01 6.81216180e-01 -7.20206618e-01
-1.04510951e+00 5.78385890e-01 -7.85930827e-02 -6.25207543e-01
-2.04820320e-01 -1.19945717e+00 -7.87734628e-01 5.93324244e-01
-1.40201056e+00 -1.46324062e+00 -5.65018356e-01 4.51555252e-01
6.56676948e-01 -1.46738291e-01 6.52030706e-01 1.89455420e-01
-6.37793601e-01 5.95189393e-01 1.25764355e-01 5.89206517e-01
3.29226643e-01 -1.58259118e+00 9.87967730e-01 9.04278696e-01
6.54407859e-01 8.44197273e-01 3.73758614e-01 -5.54857075e-01
-1.63138163e+00 -1.02465177e+00 7.42884219e-01 -6.16803348e-01
7.27992177e-01 -8.55087161e-01 -1.10229492e+00 6.91796541e-01
3.63285065e-01 -3.95057499e-01 4.10485059e-01 4.67299938e-01
-4.87937212e-01 -2.42321822e-03 -5.17636716e-01 5.79170465e-01
1.22611880e+00 -4.79153365e-01 -1.46687180e-01 4.33575839e-01
6.01082861e-01 -1.59244493e-01 -5.46739042e-01 1.40987575e-01
4.74161148e-01 -7.64525354e-01 9.14715290e-01 -7.19008267e-01
6.16058826e-01 1.25412438e-02 7.47536123e-02 -1.24074507e+00
-2.73642451e-01 -6.09115124e-01 -5.29127896e-01 1.03053832e+00
4.14940029e-01 -4.46493387e-01 1.10262680e+00 1.28972948e-01
-1.19431421e-01 -7.05516636e-01 -2.66645163e-01 -6.92172408e-01
1.15010135e-01 -7.10703313e-01 8.64533961e-01 9.56655383e-01
-1.96581513e-01 5.91300488e-01 -6.53538629e-02 -9.25802141e-02
4.85242158e-01 2.98198313e-01 1.05739772e+00 -1.67080939e+00
-1.65752381e-01 -8.33372891e-01 -3.48544598e-01 -1.09477139e+00
4.87795696e-02 -1.39698517e+00 -6.83966517e-01 -2.24225831e+00
1.54709041e-01 -1.91646650e-01 -1.26111969e-01 4.80122626e-01
-1.31616360e-02 -7.12625757e-02 3.97494197e-01 -1.01466484e-01
-6.31777942e-01 4.62454200e-01 1.23625422e+00 -6.59994483e-01
-5.98259345e-02 2.90896464e-02 -1.00215757e+00 6.71854913e-01
1.02440643e+00 -2.86154151e-01 -4.43258345e-01 -8.61372292e-01
9.19440031e-01 -4.76747751e-01 2.76463538e-01 -7.67681062e-01
2.57620156e-01 1.27624437e-01 4.27278996e-01 -9.24472213e-01
-5.58658550e-03 -6.48970604e-01 -3.30836833e-01 3.18598658e-01
-2.70002604e-01 2.54897177e-01 6.03904307e-01 4.84629095e-01
-8.65324661e-02 -2.97282726e-01 1.53189734e-01 -2.55416542e-01
-8.85113835e-01 5.01625478e-01 -1.79961473e-01 1.18281156e-01
3.52771014e-01 -1.53009087e-01 -7.27904260e-01 -4.45134312e-01
-5.88156462e-01 2.34300986e-01 2.83831537e-01 7.10382223e-01
8.24166834e-01 -1.05419433e+00 -6.01437151e-01 7.61701241e-02
2.31839314e-01 2.12540805e-01 1.71423808e-01 3.07604074e-01
-7.98008204e-01 9.01220500e-01 -8.78583789e-02 -4.46599364e-01
-1.15993404e+00 6.86801434e-01 1.12381428e-01 -8.59683871e-01
-5.48651636e-01 7.66379297e-01 3.21729183e-01 -5.85508525e-01
9.62555632e-02 -5.68980515e-01 -3.83623213e-01 1.35733962e-01
4.25781578e-01 5.20680666e-01 3.93097878e-01 -1.87358886e-01
-1.88709214e-01 6.68494999e-01 -5.93811393e-01 2.52150059e-01
1.51460457e+00 2.95913726e-01 -3.37003410e-01 5.29344231e-02
1.09166443e+00 -2.36050040e-02 -5.92962801e-01 -4.14106816e-01
5.77541828e-01 -1.93229884e-01 1.58272348e-02 -7.79487967e-01
-1.07303405e+00 1.17840385e+00 7.23422766e-02 4.86851841e-01
1.03198874e+00 5.48048131e-02 7.50748158e-01 6.81075931e-01
1.38954490e-01 -1.03295350e+00 4.79266286e-01 7.53203928e-01
1.09515405e+00 -1.08895814e+00 3.11923504e-01 -6.73282385e-01
-9.26569626e-02 1.51012838e+00 6.36062920e-01 -8.40525925e-02
6.07183397e-01 8.00061449e-02 -2.95505643e-01 -7.71787465e-01
-5.41305423e-01 -1.86810330e-01 7.90290475e-01 5.78485787e-01
4.70283180e-01 2.74615409e-03 -1.86310709e-02 5.51057637e-01
-3.76420856e-01 -2.31964022e-01 6.42799795e-01 8.54087889e-01
-1.46357447e-01 -1.41803169e+00 -1.91365659e-01 6.35762155e-01
-3.61015677e-01 -4.03905928e-01 -9.78791714e-01 8.92714441e-01
-2.97558039e-01 6.38990760e-01 1.14553489e-01 -2.23272979e-01
2.98363239e-01 -1.36387497e-01 6.21390045e-01 -7.49301434e-01
-4.15206999e-01 -1.59655467e-01 1.31073430e-01 -5.09104788e-01
-1.75912142e-01 -4.39778775e-01 -1.06646800e+00 -5.06309569e-01
-1.32713348e-01 -8.63460265e-03 6.46691740e-01 6.89276814e-01
4.88718927e-01 8.12099874e-01 9.24888328e-02 -3.12662393e-01
-9.44592878e-02 -9.89799917e-01 -5.19399107e-01 2.63172507e-01
6.12175278e-02 -2.16330037e-01 1.48083657e-01 -1.48022950e-01] | [11.663901329040527, 2.811992645263672] |
fe0ad60c-e0a0-4474-bf8b-156e78479e06 | tensor-networks-for-high-order-polynomial | 2204.07743 | null | https://arxiv.org/abs/2204.07743v1 | https://arxiv.org/pdf/2204.07743v1.pdf | Tensor-networks for High-order Polynomial Approximation: A Many-body Physics Perspective | We analyze the problem of high-order polynomial approximation from a many-body physics perspective, and demonstrate the descriptive power of entanglement entropy in capturing model capacity and task complexity. Instantiated with a high-order nonlinear dynamics modeling problem, tensor-network models are investigated and exhibit promising modeling advantages. This novel perspective establish a connection between quantum information and functional approximation, which worth further exploration in future research. | ['Tong Yang'] | 2022-04-16 | null | null | null | null | ['tensor-networks'] | ['methodology'] | [-1.23546757e-01 8.93888324e-02 -1.85665622e-01 -2.45031845e-02
-1.14736699e-01 -3.53108585e-01 7.23709464e-01 -2.20092565e-01
-2.59723276e-01 5.41384876e-01 1.74829409e-01 -1.57476872e-01
-8.03632557e-01 -5.07709563e-01 -4.17006791e-01 -7.34806836e-01
-7.70547569e-01 4.58189875e-01 -2.26164192e-01 -4.70815688e-01
1.94141552e-01 6.17689312e-01 -8.38868856e-01 -6.87641576e-02
9.28985596e-01 8.34949434e-01 -2.94590056e-01 6.49553180e-01
5.46751320e-01 9.47434247e-01 5.18042669e-02 -5.11500299e-01
5.81347227e-01 -2.63673425e-01 -9.61667120e-01 -3.26576293e-01
2.45728180e-01 -1.91629335e-01 -1.41161156e+00 1.29183936e+00
1.37893930e-01 5.01793027e-01 5.64534664e-01 -1.07841587e+00
-8.90254438e-01 5.44178426e-01 4.31838363e-01 6.70075238e-01
8.11145082e-02 6.14866734e-01 1.47216606e+00 -5.88582098e-01
5.72769642e-01 1.19736028e+00 7.43332446e-01 5.94194829e-01
-1.60756528e+00 -3.76593173e-01 -3.76659125e-01 2.43286371e-01
-1.54120159e+00 -3.76105636e-01 8.13046217e-01 -5.64590991e-01
9.37531054e-01 3.95174474e-01 8.74831617e-01 9.26307559e-01
1.11991274e+00 3.87794375e-01 1.57340395e+00 -1.78041384e-01
1.97044954e-01 -2.59951890e-01 8.10963392e-01 1.00592983e+00
3.27267349e-01 8.92434597e-01 -9.13259447e-01 -4.55234379e-01
9.19324815e-01 -2.41163094e-02 -2.83658445e-01 -3.32584888e-01
-1.23016059e+00 7.66868830e-01 6.84612393e-01 3.88438612e-01
-5.72904348e-01 8.55710268e-01 1.28184453e-01 5.25239527e-01
2.14201421e-01 1.02762318e+00 -6.84302673e-02 -4.43331480e-01
-6.92785501e-01 5.93818367e-01 9.65688050e-01 8.70143950e-01
7.90742099e-01 3.64238094e-03 -1.51422262e-01 -2.69934028e-01
3.01751256e-01 4.55389529e-01 -9.36020985e-02 -1.29965639e+00
-7.62549117e-02 2.46304318e-01 1.52518094e-01 -9.01884675e-01
-8.27253938e-01 -8.29717159e-01 -1.07155943e+00 -3.53597045e-01
2.96826482e-01 3.24057221e-01 -1.72339678e-01 1.58497787e+00
4.04912084e-02 8.30683261e-02 2.31200363e-02 1.24353588e+00
2.24162295e-01 4.37936574e-01 -2.08684847e-01 -3.22023779e-01
1.35849833e+00 -8.03756714e-01 -9.36242640e-01 2.60397196e-01
7.39036202e-01 -1.11664586e-01 8.27627122e-01 3.37864965e-01
-1.37093651e+00 -2.60113269e-01 -8.06722403e-01 -2.78572112e-01
-1.57726109e-02 -4.32250232e-01 1.40923679e+00 7.05546319e-01
-1.26702952e+00 1.36790800e+00 -1.25782633e+00 -8.45651031e-02
1.15968771e-01 5.37945926e-01 -1.58515856e-01 -1.74989676e-04
-1.34741628e+00 1.16476905e+00 3.06224644e-01 4.05392200e-01
-9.79896665e-01 -8.55961084e-01 -4.53998625e-01 1.85112774e-01
1.54432386e-01 -1.28193510e+00 1.04722703e+00 -1.59197412e-02
-1.54612684e+00 1.73051596e-01 3.33670080e-02 -6.16510868e-01
8.49205405e-02 1.66313216e-01 -5.50814390e-01 3.91089112e-01
-5.09949446e-01 -1.52596664e-02 3.99425507e-01 -5.77217460e-01
4.41291898e-01 -4.20678496e-01 5.82264721e-01 4.27857451e-02
-3.54829520e-01 -3.09526056e-01 3.04377466e-01 -1.41066805e-01
3.47909182e-01 -1.30459654e+00 -5.61310351e-01 -3.87051582e-01
-4.12457883e-01 -1.25014856e-01 5.87315019e-03 -3.03930342e-01
1.46894646e+00 -1.70116770e+00 7.89042234e-01 4.34161514e-01
9.06910002e-01 -2.26383820e-01 -1.60028368e-01 1.05887485e+00
1.15912281e-01 3.69080007e-01 1.48884550e-01 -1.75349876e-01
5.81914604e-01 1.72041565e-01 -1.56291798e-01 7.71715343e-01
8.78820047e-02 1.51856852e+00 -8.06657970e-01 -2.32025146e-01
-4.70305122e-02 3.47163945e-01 -9.11264777e-01 -2.20411330e-01
1.87125981e-01 8.72942686e-01 -8.37569416e-01 4.00386572e-01
4.38162714e-01 -8.53107572e-01 6.46951348e-02 -5.35903752e-01
-1.28301471e-01 4.69671458e-01 -6.01995409e-01 1.81155193e+00
-2.68118858e-01 5.70531607e-01 3.77648212e-02 -8.74421835e-01
1.01513028e-01 1.67226583e-01 5.70157766e-01 -6.50599897e-01
1.78660199e-01 -2.22045332e-02 8.64336491e-01 -6.71856463e-01
9.58633959e-01 -5.67523181e-01 -3.08471680e-01 3.10801625e-01
2.35110432e-01 -1.04307473e-01 2.19087347e-01 4.32808310e-01
1.33702898e+00 -2.02853635e-01 1.26959890e-01 -8.45668197e-01
3.29700559e-01 6.37915134e-02 3.49575691e-02 1.14287686e+00
-4.75049615e-01 -3.24347019e-01 4.35022235e-01 -4.01051283e-01
-1.43812311e+00 -8.88486981e-01 -7.71699667e-01 6.88868761e-01
1.15974545e-01 -9.88401234e-01 -5.96330881e-01 8.78936499e-02
1.39953285e-01 4.43465620e-01 -7.31428742e-01 -5.08292317e-01
-3.43869328e-01 -7.44316339e-01 6.32823408e-01 2.82004736e-02
3.13903868e-01 -2.63404280e-01 -1.12396732e-01 -9.20795649e-02
-8.52070898e-02 -1.13474882e+00 -3.32407624e-01 -8.30476210e-02
-1.06446397e+00 -7.75731802e-01 -2.17381373e-01 2.10116748e-02
3.50219101e-01 -4.53612581e-02 9.57722127e-01 2.07511291e-01
-2.54052281e-01 7.83810854e-01 -5.16456664e-02 4.29149687e-01
-6.54194415e-01 2.27009784e-02 6.75093830e-01 -2.40193635e-01
8.14480260e-02 -8.12876940e-01 -6.49778366e-01 6.48764297e-02
-8.73490036e-01 1.68138295e-01 6.63114190e-01 9.82337654e-01
2.24144071e-01 2.53410757e-01 -1.59830913e-01 -4.07025099e-01
8.77723575e-01 -4.45056975e-01 -6.27385736e-01 4.95916009e-02
-6.83377326e-01 5.44155002e-01 8.49823117e-01 -2.37465993e-01
-5.88838577e-01 -5.64421356e-01 2.21891031e-01 -3.99740189e-01
4.05050576e-01 9.18488204e-01 4.65012342e-01 -9.15102363e-01
6.32803142e-01 4.72033948e-01 1.16606511e-01 -1.73744082e-01
4.33824748e-01 3.10040750e-02 1.92241639e-01 -9.72469211e-01
1.20964801e+00 4.87369299e-01 9.80067790e-01 -7.74496555e-01
-7.32103229e-01 -2.41012439e-01 -8.03329825e-01 -1.51315093e-01
6.80337608e-01 -7.61688054e-01 -1.68211591e+00 3.32777724e-02
-1.24653924e+00 -1.28964782e-01 -4.47771043e-01 9.42742646e-01
-7.42473960e-01 5.45710623e-01 -1.16451240e+00 -1.32046366e+00
-1.14329932e-02 -1.07219768e+00 7.08215714e-01 3.46239731e-02
2.01955393e-01 -1.31871939e+00 2.83322036e-01 2.76045829e-01
6.45699799e-01 -5.52026890e-02 9.10504818e-01 -4.44744915e-01
-1.25318885e+00 -3.19860965e-01 1.10799726e-02 6.60716593e-02
-9.46476758e-01 -4.10277843e-01 -5.93117297e-01 -5.61483800e-01
2.83017814e-01 -3.10965300e-01 7.51644909e-01 2.14692429e-01
9.78101671e-01 -4.87986475e-01 -1.84677884e-01 6.16292298e-01
1.40222979e+00 -4.99132127e-01 4.05229926e-01 -4.48361754e-01
6.82700336e-01 3.98324460e-01 -1.64724663e-01 3.80748093e-01
5.58361590e-01 7.24573970e-01 2.87805617e-01 5.40665329e-01
1.24012865e-01 -3.36087883e-01 2.40535080e-01 1.78152204e+00
-7.59184659e-01 2.19913930e-01 -8.79333973e-01 -2.03682184e-01
-1.66493022e+00 -1.20775497e+00 -4.54190522e-01 1.96421385e+00
4.71268445e-01 -2.01420724e-01 1.97306007e-01 -3.04176331e-01
2.86739528e-01 -1.05107269e-04 -5.98047554e-01 -2.30196804e-01
-1.42360047e-01 3.20739418e-01 5.92009127e-01 5.12636364e-01
-5.99586606e-01 7.92835355e-01 8.13809299e+00 8.95088077e-01
-6.18030488e-01 6.48236096e-01 3.78980599e-02 2.89166011e-02
-5.37053227e-01 5.81525981e-01 -5.12518585e-01 4.10494387e-01
1.33225107e+00 -3.99076074e-01 1.34972489e+00 2.62395948e-01
1.55825973e-01 1.71615690e-01 -1.37682414e+00 9.44381595e-01
-3.41759622e-01 -1.54042304e+00 -7.75213167e-02 9.14160430e-01
7.08932400e-01 2.63978869e-01 2.40915269e-01 3.38675529e-01
1.08840028e-02 -1.11116362e+00 5.29203653e-01 1.19232213e+00
3.45414639e-01 -3.57252091e-01 3.49335998e-01 5.69070160e-01
-9.29899216e-01 -2.95638621e-01 -5.30777752e-01 -8.13859165e-01
2.46546000e-01 5.03056824e-01 -2.08587758e-02 7.99982309e-01
1.12954728e-01 5.95022142e-01 -3.51253867e-01 1.00086498e+00
2.75670081e-01 5.16469181e-01 -3.14692855e-01 -3.54920000e-01
1.12458214e-01 -6.22331619e-01 9.03647840e-01 7.08510220e-01
4.37524058e-02 4.50748026e-01 1.80841535e-01 1.67836320e+00
1.95209488e-01 -2.48599440e-01 -5.93392432e-01 -8.07651460e-01
1.92172617e-01 1.32900488e+00 -3.47793818e-01 1.66631620e-02
-1.55551776e-01 8.06365609e-01 4.52672899e-01 4.81821507e-01
-8.89868915e-01 4.68251072e-02 6.26304448e-01 -1.23528019e-01
-1.94626167e-01 -1.09331942e+00 -1.59887061e-01 -1.79316354e+00
-1.95088629e-02 -4.28676188e-01 -2.91166127e-01 -3.55483919e-01
-1.57543850e+00 2.35673040e-01 4.68215123e-02 -1.03693402e+00
-2.83545982e-02 -9.14120436e-01 -3.25935423e-01 7.27495074e-01
-1.25148141e+00 -1.00564504e+00 1.61097765e-01 5.73180199e-01
-5.58119714e-01 1.41358986e-01 1.09767890e+00 2.09454805e-01
-6.86742783e-01 4.36563104e-01 4.46030796e-01 -2.92525053e-01
-4.24445271e-01 -1.16315210e+00 4.21324223e-01 6.67040825e-01
1.28910094e-01 1.30459261e+00 8.94177079e-01 -4.37836260e-01
-2.61781144e+00 -4.48614717e-01 7.23105133e-01 -8.69609058e-01
1.32447791e+00 -6.86085522e-01 -4.49739426e-01 5.16052544e-01
-5.63637167e-02 3.13007891e-01 8.63577962e-01 5.07205606e-01
-2.80310065e-01 4.59163606e-01 -9.68955636e-01 4.44145143e-01
1.50823510e+00 -1.21063626e+00 -1.60797477e-01 8.20049167e-01
6.38195455e-01 -3.94397974e-01 -1.59135115e+00 -1.63523816e-02
5.05534172e-01 -6.71273649e-01 8.48707974e-01 -1.06416082e+00
4.73216116e-01 1.06982619e-01 -2.01658815e-01 -1.12651384e+00
-8.76201570e-01 -1.26754832e+00 -7.67150283e-01 2.21235588e-01
1.14580981e-01 -7.70213962e-01 3.64335567e-01 1.04929733e+00
-2.55071908e-01 -6.90223217e-01 -1.29538202e+00 -1.20987296e+00
6.19241953e-01 -4.34386820e-01 1.90746114e-01 8.84500742e-01
8.39238942e-01 3.69452924e-01 -6.11080050e-01 1.97267160e-01
7.33605385e-01 -8.19968358e-02 4.03211325e-01 -1.40927911e+00
-7.54057884e-01 -5.57171404e-01 -8.53915453e-01 -1.42251074e+00
4.39420253e-01 -1.50157583e+00 -5.07355511e-01 -6.51412487e-01
6.45528495e-01 -2.87448943e-01 -3.30149382e-01 -2.71011442e-01
3.61835659e-01 2.51554668e-01 5.04990101e-01 7.82438695e-01
-9.99751151e-01 8.90971065e-01 1.68464947e+00 4.34410609e-02
3.13230783e-01 -6.42895252e-02 -3.26449275e-01 2.38973305e-01
5.07010460e-01 -3.21479291e-01 -6.77466467e-02 -1.54161856e-01
7.92202890e-01 5.81200659e-01 9.59497213e-01 -1.08547330e+00
6.01568222e-01 -2.79714525e-01 -2.56607272e-02 -3.40671018e-02
8.18601727e-01 -6.25096917e-01 1.11442745e-01 8.90979052e-01
-5.46228170e-01 2.20484018e-01 -1.51195928e-01 8.43468070e-01
2.79199630e-01 -1.08689778e-01 3.35718066e-01 -1.26222298e-01
-2.60129631e-01 6.48984253e-01 -3.42818767e-01 4.58886884e-02
5.17713428e-01 1.73870087e-01 -6.39171600e-01 -3.07210118e-01
-1.05593026e+00 5.72979599e-02 5.09138823e-01 -1.53786108e-01
1.09322734e-01 -1.58207154e+00 -4.10332948e-01 -7.01032430e-02
2.96221133e-02 -8.79798412e-01 8.05898309e-01 1.66512048e+00
-2.04879060e-01 9.09267545e-01 -1.48781687e-01 -4.70987022e-01
-1.17152095e-01 6.44176304e-01 5.29360294e-01 -5.32739103e-01
-5.37365735e-01 3.50481182e-01 7.48622715e-02 -3.44955891e-01
-5.02597332e-01 -5.11744261e-01 5.12211561e-01 -7.61125743e-01
2.21186832e-01 5.78033805e-01 -2.50152916e-01 -7.81111538e-01
-2.10444868e-01 2.21357390e-01 2.82159269e-01 -2.18372583e-01
9.01085675e-01 -1.31516993e-01 -5.37044883e-01 6.01356447e-01
1.22095370e+00 -2.42383465e-01 -8.84549439e-01 -3.16995651e-01
-5.80229051e-02 -2.53065467e-01 1.83206409e-01 -6.52400076e-01
-4.44393396e-01 9.71310258e-01 3.88060004e-01 7.87569880e-01
4.91366208e-01 1.57528237e-01 6.49997294e-01 9.92332578e-01
7.93724716e-01 -7.75776029e-01 2.35965289e-02 8.45037460e-01
7.01547980e-01 -9.49049234e-01 -9.55240950e-02 -2.69305199e-01
-7.53248408e-02 1.20981705e+00 2.23623633e-01 -5.82949340e-01
9.09318924e-01 -6.66783452e-02 -1.08542609e+00 -7.12716758e-01
-8.24076355e-01 -1.97847649e-01 6.21370852e-01 3.25470179e-01
4.45674658e-01 4.39283967e-01 -6.30707562e-01 5.65523684e-01
-6.23234928e-01 -4.93468456e-02 7.15381563e-01 3.89483064e-01
-5.65161593e-02 -9.53398168e-01 1.97770432e-01 4.38526571e-01
-3.48179460e-01 -3.87226671e-01 -1.23087741e-01 6.92741513e-01
-3.89366210e-01 7.09663570e-01 -2.93475259e-02 -6.89990819e-01
-9.68554690e-02 1.53511211e-01 9.34815943e-01 -3.19254428e-01
-5.88035464e-01 -5.01051664e-01 2.35630944e-02 -8.58950198e-01
-4.71211463e-01 -5.11249602e-01 -9.63026166e-01 -1.00177288e+00
-6.80046499e-01 2.74627656e-01 6.79182708e-01 1.07776535e+00
7.60060251e-01 4.69059616e-01 4.37389225e-01 -9.92410898e-01
-1.31822944e+00 -1.03780770e+00 -8.47648978e-01 9.54894423e-02
4.34445620e-01 -7.37353146e-01 -5.19792259e-01 -6.46010816e-01] | [5.701115608215332, 4.987032890319824] |
2dc23d08-7501-4d9d-9ea4-237beaec4481 | learning-to-forecast-vegetation-greenness-at | 2210.13648 | null | https://arxiv.org/abs/2210.13648v2 | https://arxiv.org/pdf/2210.13648v2.pdf | Learning to forecast vegetation greenness at fine resolution over Africa with ConvLSTMs | Forecasting the state of vegetation in response to climate and weather events is a major challenge. Its implementation will prove crucial in predicting crop yield, forest damage, or more generally the impact on ecosystems services relevant for socio-economic functioning, which if absent can lead to humanitarian disasters. Vegetation status depends on weather and environmental conditions that modulate complex ecological processes taking place at several timescales. Interactions between vegetation and different environmental drivers express responses at instantaneous but also time-lagged effects, often showing an emerging spatial context at landscape and regional scales. We formulate the land surface forecasting task as a strongly guided video prediction task where the objective is to forecast the vegetation developing at very fine resolution using topography and weather variables to guide the prediction. We use a Convolutional LSTM (ConvLSTM) architecture to address this task and predict changes in the vegetation state in Africa using Sentinel-2 satellite NDVI, having ERA5 weather reanalysis, SMAP satellite measurements, and topography (DEM of SRTMv4.1) as variables to guide the prediction. Ours results highlight how ConvLSTM models can not only forecast the seasonal evolution of NDVI at high resolution, but also the differential impacts of weather anomalies over the baselines. The model is able to predict different vegetation types, even those with very high NDVI variability during target length, which is promising to support anticipatory actions in the context of drought-related disasters. | ['Markus Reichstein', 'Nuno Carvalhais', 'Jeran Poehls', 'Lazaro Alonso', 'Vitus Benson', 'Christian Requena-Mesa', 'Claire Robin'] | 2022-10-24 | null | null | null | null | ['video-prediction'] | ['computer-vision'] | [ 4.03399289e-01 -3.69218767e-01 -4.03663665e-02 -7.86605105e-02
1.55878350e-01 -7.66685069e-01 7.45170474e-01 2.34311596e-01
-3.05097818e-01 9.67369676e-01 3.50562155e-01 -8.22444975e-01
-1.95428446e-01 -1.43903291e+00 -4.99354064e-01 -8.92610669e-01
-8.10729861e-01 -7.17074424e-02 8.19660649e-02 -1.01685798e+00
-2.28070483e-01 9.48273897e-01 -1.71640956e+00 2.81588972e-01
6.12516463e-01 6.64921939e-01 8.07838440e-01 9.31175828e-01
-8.01161081e-02 4.40328777e-01 -2.57427692e-01 4.31114048e-01
9.52960700e-02 1.17571820e-02 -5.79719663e-01 8.96700770e-02
-9.10197035e-04 -4.43718016e-01 1.31908730e-02 8.19170892e-01
3.59955102e-01 -1.33559555e-01 4.23653036e-01 -6.23711824e-01
-3.02531630e-01 3.28956932e-01 -7.29896009e-01 2.34085083e-01
3.11145075e-02 3.10524464e-01 6.67173803e-01 -6.09824598e-01
6.58811748e-01 1.30600941e+00 7.24011540e-01 -2.66272962e-01
-1.49143720e+00 -3.23896348e-01 4.91469413e-01 1.88030466e-01
-1.04243243e+00 -2.90375322e-01 3.30484420e-01 -9.91270781e-01
8.71147633e-01 4.23480898e-01 7.60979950e-01 1.07017553e+00
7.46517718e-01 2.16422871e-01 1.03500223e+00 -1.51981175e-01
2.96191294e-02 -4.93995577e-01 -3.66730303e-01 2.19073534e-01
3.07888463e-02 6.52916908e-01 -1.24826558e-01 1.70721307e-01
5.32301903e-01 3.43821794e-01 -2.92453736e-01 3.13779145e-01
-1.40546942e+00 6.48425937e-01 9.99027312e-01 5.33813894e-01
-1.32824254e+00 9.87671018e-02 4.25471127e-01 4.53768283e-01
9.07177389e-01 2.21281528e-01 -9.96670783e-01 3.04790378e-01
-1.10675025e+00 3.58736694e-01 2.45481357e-01 9.26265270e-02
6.61219597e-01 4.04654950e-01 3.86475883e-02 7.03531742e-01
1.67891264e-01 1.24581254e+00 -1.23405203e-01 -7.38759637e-01
5.29236346e-03 4.13862199e-01 3.82644624e-01 -1.42149818e+00
-7.81624615e-01 -3.91763031e-01 -1.38001978e+00 5.91878057e-01
1.21451095e-01 -4.48988259e-01 -9.26870823e-01 1.69839549e+00
1.49927258e-01 -1.01866268e-01 -6.52736723e-02 8.67096722e-01
3.78847778e-01 1.32769084e+00 5.35662651e-01 -3.92834783e-01
1.19938529e+00 4.51408550e-02 -8.19577694e-01 -3.74257326e-01
7.17583537e-01 -3.78250450e-01 5.58978796e-01 -1.87261164e-01
-3.76381427e-01 -6.15253389e-01 -6.05724752e-01 7.34709918e-01
-9.41332042e-01 2.10749239e-01 3.93710077e-01 -1.61241248e-01
-1.39763892e+00 7.20772266e-01 -7.17508614e-01 -7.35283732e-01
2.01587766e-01 5.30940928e-02 -3.45533758e-01 4.07475263e-01
-1.58096623e+00 1.15379238e+00 4.34357405e-01 1.07627022e+00
-5.88909328e-01 -5.60767055e-01 -8.27020884e-01 2.81047106e-01
-3.38165984e-02 -2.34803617e-01 4.36793089e-01 -1.14887965e+00
-9.95055735e-01 8.61423910e-01 -5.16133457e-02 -6.57614648e-01
3.08621883e-01 1.77604377e-01 -6.33463264e-01 -4.03809220e-01
3.81592005e-01 8.26551557e-01 5.18409371e-01 -1.05716360e+00
-1.14748693e+00 -6.91093206e-01 -1.42242059e-01 3.02862018e-01
1.47864083e-02 6.30739257e-02 6.84643745e-01 -6.50795639e-01
-5.67782968e-02 -9.88896549e-01 -6.58140302e-01 2.29959972e-02
6.05210699e-02 3.55572253e-01 9.94775116e-01 -1.00597215e+00
9.15859282e-01 -2.07003903e+00 3.11933309e-01 -1.72781318e-01
-9.18716714e-02 6.08849108e-01 -4.66203660e-01 6.82456076e-01
6.44551069e-02 3.71533006e-01 -6.90347850e-01 4.78526413e-01
-3.67009938e-01 4.92674917e-01 -8.71474504e-01 2.78090715e-01
3.77404898e-01 7.18048215e-01 -7.91797578e-01 1.98287755e-01
6.05574131e-01 5.93651533e-01 2.32034072e-01 8.65083858e-02
-4.62165654e-01 6.28714919e-01 -2.15842962e-01 5.46264350e-01
1.02899671e+00 3.84430468e-01 3.06521952e-01 2.28402287e-01
-1.03560221e+00 -1.53522626e-01 -7.11457610e-01 8.43569219e-01
-3.17593306e-01 1.06874561e+00 4.66933489e-01 -1.03837299e+00
1.11275268e+00 4.43871468e-01 3.71519387e-01 -8.03080380e-01
-3.00214678e-01 2.99443424e-01 8.47256109e-02 -5.43615699e-01
4.95808750e-01 -1.79497302e-01 2.09422544e-01 2.02352908e-02
-4.94215131e-01 1.12907976e-01 -2.04621926e-01 -5.07755995e-01
5.61601698e-01 3.33479702e-01 2.28933334e-01 -5.05042851e-01
4.52675372e-01 3.21033150e-01 4.63723987e-01 4.10689294e-01
-2.10610643e-01 2.39501998e-04 6.36523247e-01 -1.07995498e+00
-8.77166629e-01 -6.58577859e-01 -4.16372448e-01 1.31185520e+00
-3.32778215e-01 4.31763440e-01 1.10568374e-01 1.29228653e-02
2.29698852e-01 7.94975102e-01 -8.56486738e-01 8.77472013e-02
-2.70418882e-01 -1.32250464e+00 4.24398214e-01 2.53183097e-01
5.74480176e-01 -1.62874508e+00 -1.05483472e+00 5.89589059e-01
-1.77299574e-01 -9.21485305e-01 6.89433515e-01 3.26212108e-01
-9.39175129e-01 -6.53295934e-01 -8.24186623e-01 -3.41327816e-01
1.08467050e-01 3.51922035e-01 9.95594800e-01 -2.24832609e-01
-2.26497967e-02 5.53610362e-02 -2.05320939e-01 -4.17807251e-01
-3.71314585e-01 2.41243228e-01 -1.20572839e-02 1.81981281e-03
-4.21635062e-02 -9.23319459e-01 -5.66843867e-01 2.52577633e-01
-9.57404375e-01 3.98012727e-01 4.06345487e-01 6.09112501e-01
2.68802911e-01 -3.27737778e-02 6.81466818e-01 -4.33842599e-01
1.12014979e-01 -8.17500114e-01 -7.67103374e-01 1.56532064e-01
-3.10401201e-01 -3.30881327e-01 3.79821509e-01 4.85074818e-02
-1.15025163e+00 2.31147736e-01 -3.19755524e-02 3.18913996e-01
-6.98083043e-01 1.44640803e+00 1.63848862e-01 2.34170765e-01
5.00148833e-01 3.74383777e-01 -2.55108148e-01 -2.91051924e-01
1.22635186e-01 6.25543058e-01 5.53952217e-01 2.91495882e-02
6.17825449e-01 6.08278632e-01 2.83110917e-01 -1.51034713e+00
-5.94794333e-01 -1.62141293e-01 -1.03539264e+00 -5.96501648e-01
6.66916072e-01 -1.10599506e+00 -5.10242879e-01 8.41928303e-01
-1.25450945e+00 -6.68788910e-01 -3.69878411e-02 2.49666899e-01
-1.26751989e-01 -3.00180376e-01 -3.55582349e-02 -9.36972797e-01
-4.58615780e-01 -6.03148580e-01 6.64034367e-01 5.44710606e-02
4.96815443e-02 -1.25106394e+00 5.30565143e-01 -5.27348578e-01
9.98348594e-01 1.17171967e+00 7.54218459e-01 3.02594483e-01
-2.79388785e-01 -1.16604818e-02 -5.04094422e-01 1.55132180e-02
6.82256997e-01 5.24525106e-01 -8.94985318e-01 -1.26393020e-01
-3.69172484e-01 1.83903620e-01 1.18273306e+00 1.07908285e+00
6.19409561e-01 -3.13287318e-01 -2.56548882e-01 8.53995919e-01
1.47324836e+00 2.20886171e-01 5.80302775e-01 5.04198611e-01
5.39610684e-01 1.00386786e+00 6.22251868e-01 5.39264321e-01
3.56933594e-01 5.71479261e-01 1.00874984e+00 -7.89130628e-01
2.21565887e-01 3.27642351e-01 6.98959053e-01 4.79955710e-02
-3.05915803e-01 1.21969786e-02 -1.42455339e+00 8.44478965e-01
-1.88301957e+00 -1.59945595e+00 -7.50801861e-01 2.19602323e+00
3.19402099e-01 -2.14912772e-01 -1.10659353e-01 -2.19716027e-01
6.54760897e-01 1.01438296e+00 -6.32842541e-01 -5.51570117e-01
-7.57371247e-01 -3.95832397e-02 1.06975102e+00 5.61287999e-01
-1.23465693e+00 1.24606657e+00 6.02784348e+00 -1.45117074e-01
-1.88438308e+00 -1.22720443e-01 4.30129796e-01 2.21838072e-01
-2.72462130e-01 9.73807424e-02 -4.32294130e-01 1.53370753e-01
1.24694264e+00 -2.72317324e-02 2.95287669e-01 1.96072370e-01
1.26363897e+00 -3.80936921e-01 -4.01465781e-02 -2.34947260e-03
-9.35133994e-01 -1.45227766e+00 -2.32387349e-04 2.86678582e-01
6.63299501e-01 6.38716042e-01 4.62313518e-02 1.17891967e-01
4.72807437e-01 -7.70004928e-01 3.40421915e-01 7.00989008e-01
9.95192647e-01 -4.23733592e-01 6.42230213e-01 3.86508614e-01
-1.56290269e+00 -1.35043845e-01 -5.08056283e-01 -5.88517487e-01
1.46921992e-01 6.56143546e-01 -5.82564831e-01 4.77779925e-01
7.88994610e-01 1.34501290e+00 -1.96675211e-01 5.03985643e-01
-9.73260477e-02 9.47437942e-01 -3.53063524e-01 6.28500521e-01
6.83669925e-01 -2.09869385e-01 8.17172885e-01 1.14656949e+00
6.03382647e-01 3.73037070e-01 9.27806795e-02 6.03161395e-01
4.55660760e-01 -8.37412837e-04 -1.09980774e+00 6.51908517e-02
1.25695840e-01 1.10897040e+00 -3.32935393e-01 -2.99374610e-02
1.95063874e-02 8.01861703e-01 -1.51079595e-01 6.90439463e-01
-3.49482179e-01 1.26416355e-01 1.16098332e+00 2.08860502e-01
2.07886934e-01 -5.34583032e-01 3.31421709e-03 -1.04880416e+00
-2.45268047e-01 -5.18599689e-01 1.83068290e-01 -8.80585253e-01
-4.52500731e-01 5.10602713e-01 -2.10219964e-01 -1.12997186e+00
-2.77543664e-01 -5.19766450e-01 -1.00287580e+00 1.34694600e+00
-2.24473763e+00 -1.39852142e+00 -2.98380852e-01 1.02459103e-01
3.59549552e-01 2.18702376e-01 1.32494617e+00 -2.72204727e-01
-5.41996598e-01 -7.10919499e-01 6.68737054e-01 -3.45075905e-01
4.82762158e-01 -1.03782821e+00 8.28443587e-01 7.89852083e-01
-5.63605428e-01 -3.24333757e-01 7.26814508e-01 -8.59554172e-01
-8.50661516e-01 -1.58730209e+00 1.23392558e+00 4.24931973e-01
7.53989637e-01 2.10032612e-01 -1.00989640e+00 6.63088918e-01
-7.36743510e-02 5.64751960e-02 3.44310015e-01 1.65302828e-01
5.57092726e-02 -4.27722126e-01 -9.42387760e-01 5.30201197e-01
5.89731693e-01 -4.15196896e-01 7.28527084e-02 4.11753505e-01
4.84053433e-01 1.89140961e-01 -7.02752650e-01 7.27184296e-01
9.31583703e-01 -9.49090123e-01 6.79231048e-01 -7.07730591e-01
6.39302135e-01 -2.45098427e-01 -6.64856970e-01 -1.68343043e+00
-8.93960714e-01 -1.41097337e-01 3.57072920e-01 9.27350819e-01
1.90291852e-01 -6.53438866e-01 3.05259749e-02 2.34081849e-01
3.49415064e-01 -1.03965059e-01 -7.85062373e-01 -5.44527888e-01
2.28137448e-01 -1.92007869e-01 5.90595484e-01 9.97255802e-01
-6.42658710e-01 -9.98363048e-02 -5.27691364e-01 8.48803401e-01
1.27956867e-01 2.52566010e-01 5.18648326e-01 -1.71069396e+00
5.98399043e-01 -5.63391924e-01 -3.39595944e-01 -8.43857378e-02
9.85852480e-02 -3.18900257e-01 -2.72730082e-01 -1.40304744e+00
-3.27030510e-01 -2.74264634e-01 -3.17359477e-01 8.37120473e-01
-1.62490115e-01 -1.75873578e-01 1.16302274e-01 2.35126838e-01
6.84127510e-01 9.20601010e-01 7.72980928e-01 -2.65877992e-01
-5.73654652e-01 1.66760042e-01 -2.60775872e-02 5.58805108e-01
8.10698092e-01 -3.62788022e-01 1.70552716e-01 -6.29191995e-01
4.77950931e-01 4.61851150e-01 7.29742706e-01 -1.05716109e+00
-5.25195241e-01 -9.12887216e-01 3.94993424e-01 -9.65764046e-01
-1.05955161e-01 -9.40088272e-01 5.19280374e-01 7.54663110e-01
-2.88607121e-01 9.16512012e-02 8.10021698e-01 3.09026927e-01
-8.42930228e-02 3.61900121e-01 4.61468309e-01 3.07262335e-02
-9.86591041e-01 6.47986114e-01 -1.22832584e+00 -7.27834880e-01
6.34425700e-01 6.66723028e-02 -2.24119753e-01 -4.46193635e-01
-8.27170193e-01 5.82521141e-01 4.23451394e-01 5.83506644e-01
1.23183459e-01 -1.05132997e+00 -1.42609394e+00 2.83827543e-01
1.25507079e-02 -4.91666228e-01 4.99787450e-01 6.23228014e-01
-5.83083212e-01 4.33995694e-01 -8.11258376e-01 -6.73241794e-01
-1.06543994e+00 1.90838173e-01 6.91654503e-01 -3.74027640e-01
-4.23823833e-01 5.18418610e-01 1.58610478e-01 -6.95142448e-01
-4.32443410e-01 -3.32363456e-01 -7.36042142e-01 7.93265641e-01
4.95752990e-01 -2.71211211e-02 -2.26945132e-01 -1.02629662e+00
-2.79619575e-01 3.28092009e-01 4.79418039e-01 -2.24681616e-01
1.85819161e+00 -5.62511146e-01 -2.55473644e-01 9.17524695e-01
6.32333219e-01 -5.74993551e-01 -1.44627726e+00 -4.19124335e-01
1.70110449e-01 -2.48968169e-01 7.76614249e-01 -9.65009153e-01
-1.23634720e+00 1.09402955e+00 9.36736524e-01 4.54045385e-01
1.29205918e+00 -6.52050853e-01 5.44255555e-01 3.44129860e-01
-1.29620852e-02 -6.09739184e-01 -1.10272801e+00 6.49413526e-01
1.35029495e+00 -1.35692787e+00 -8.04517716e-02 3.12637925e-01
-2.94924557e-01 1.30735171e+00 1.06504105e-01 2.70550609e-01
8.31031322e-01 1.23149399e-02 1.94468573e-01 -1.20415330e-01
-8.51748526e-01 -7.86540270e-01 -6.30887225e-03 6.21458888e-01
4.39430445e-01 9.43692565e-01 5.41829802e-02 -8.67553428e-02
2.10275650e-01 1.36290103e-01 4.31418419e-01 5.81702948e-01
-7.05259502e-01 -7.52001286e-01 -6.99558198e-01 4.83945608e-01
-4.12531830e-02 -1.90477267e-01 -2.69432068e-01 3.77107650e-01
4.84057702e-02 6.25473619e-01 1.93030581e-01 -2.78363734e-01
2.16870561e-01 -9.22927782e-02 -1.95241794e-01 -2.86208957e-01
-4.94435340e-01 -1.12175889e-01 -1.30800620e-01 -4.99158710e-01
-6.00798726e-01 -1.11975741e+00 -6.69485986e-01 -8.43175352e-01
-7.58154988e-02 -4.97784346e-01 8.51631165e-01 1.13282776e+00
5.39689779e-01 4.38470572e-01 1.04277718e+00 -1.56011939e+00
1.82099938e-02 -1.14775014e+00 -7.83462465e-01 -2.64851123e-01
8.77100110e-01 -4.66117829e-01 -1.77724123e-01 -1.07130043e-01] | [9.509114265441895, -1.5604009628295898] |
f6d08d7d-28a7-4d88-aefc-f6caba42d212 | explainable-verbal-deception-detection-using | 2210.03080 | null | https://arxiv.org/abs/2210.03080v1 | https://arxiv.org/pdf/2210.03080v1.pdf | Explainable Verbal Deception Detection using Transformers | People are regularly confronted with potentially deceptive statements (e.g., fake news, misleading product reviews, or lies about activities). Only few works on automated text-based deception detection have exploited the potential of deep learning approaches. A critique of deep-learning methods is their lack of interpretability, preventing us from understanding the underlying (linguistic) mechanisms involved in deception. However, recent advancements have made it possible to explain some aspects of such models. This paper proposes and evaluates six deep-learning models, including combinations of BERT (and RoBERTa), MultiHead Attention, co-attentions, and transformers. To understand how the models reach their decisions, we then examine the model's predictions with LIME. We then zoom in on vocabulary uniqueness and the correlation of LIWC categories with the outcome class (truthful vs deceptive). The findings suggest that our transformer-based models can enhance automated deception detection performances (+2.11% in accuracy) and show significant differences pertinent to the usage of LIWC features in truthful and deceptive statements. | ['Bennett Kleinberg', 'Felix Soldner', 'Loukas Ilias'] | 2022-10-06 | null | null | null | null | ['deception-detection'] | ['miscellaneous'] | [-1.00792602e-01 -2.26579271e-02 -2.49929875e-01 -5.86423218e-01
-4.79345709e-01 -5.53903043e-01 9.40528452e-01 1.82261780e-01
-2.27023840e-01 6.28899992e-01 4.59407002e-01 -6.48586690e-01
4.46914621e-02 -3.83349687e-01 -3.67694050e-01 -3.73313636e-01
2.27125168e-01 1.91876069e-01 -4.20589000e-01 -3.61729681e-01
1.01356769e+00 4.64447439e-01 -1.18825841e+00 6.36169851e-01
1.05468166e+00 1.07682490e+00 -7.71884561e-01 5.34683824e-01
-2.02160254e-01 1.41266167e+00 -1.28392875e+00 -1.28591704e+00
-1.78403482e-01 -5.77731013e-01 -8.02892685e-01 -1.74967974e-01
8.76203835e-01 -8.86800349e-01 -4.67297226e-01 1.16055739e+00
2.10281938e-01 -1.71257094e-01 8.45044017e-01 -1.33915639e+00
-1.53225112e+00 4.34199572e-01 -3.67611557e-01 8.44240606e-01
4.19934452e-01 4.60188121e-01 9.43545222e-01 -7.42362857e-01
9.45413485e-02 1.43921363e+00 8.80623698e-01 7.02367246e-01
-8.76189590e-01 -7.89058864e-01 1.02322018e-02 6.34604156e-01
-1.11459148e+00 -8.30140412e-01 8.31074417e-01 -4.94166404e-01
9.72184956e-01 3.43339711e-01 6.46615028e-01 1.82962334e+00
6.31181300e-01 9.13799167e-01 1.39480674e+00 -2.53318071e-01
-4.32193466e-02 4.64191347e-01 5.22146642e-01 5.88412285e-01
6.67107284e-01 3.36955160e-01 -8.18761885e-01 -4.68477160e-01
2.47189045e-01 -1.97030261e-01 -2.68536001e-01 1.24597333e-01
-6.88666642e-01 1.37103522e+00 3.86534214e-01 5.40378630e-01
-9.69313532e-02 1.09428473e-01 6.24860823e-01 5.48853695e-01
8.71869028e-01 9.01515007e-01 -1.88421860e-01 -2.13055491e-01
-1.14265609e+00 2.68290341e-01 9.36201692e-01 4.01704103e-01
2.89273132e-02 4.51763570e-01 -2.13821277e-01 5.73525727e-01
2.57771522e-01 3.36035877e-01 8.27284276e-01 -4.31263924e-01
2.95233458e-01 2.47799486e-01 3.58443886e-01 -1.89943421e+00
-3.15083236e-01 -5.60139596e-01 -4.33338910e-01 8.31103176e-02
4.53454584e-01 -7.67043307e-02 -6.30251527e-01 1.04654694e+00
-3.56191218e-01 -3.23175728e-01 -1.94535941e-01 1.15388370e+00
6.82639301e-01 3.24676961e-01 1.87809914e-01 6.64813444e-02
1.23293293e+00 -8.90313983e-01 -9.70480919e-01 -6.40534520e-01
8.37095737e-01 -4.39415663e-01 1.05707526e+00 5.59702277e-01
-9.77724075e-01 -1.66361541e-01 -1.11896491e+00 -5.11268198e-01
-6.60025895e-01 -1.58530504e-01 7.32775509e-01 1.19267499e+00
-7.25404739e-01 7.30216086e-01 -3.77688885e-01 -1.36449784e-01
8.71333241e-01 -6.27747998e-02 7.09371502e-03 1.39610529e-01
-1.58137798e+00 1.70768511e+00 3.05002015e-02 2.60485619e-01
-1.02377880e+00 -3.53492051e-01 -7.35625446e-01 1.07560813e-01
8.28600004e-02 -5.32350838e-01 1.09131205e+00 -1.84163880e+00
-1.28322661e+00 1.03795516e+00 -9.75487083e-02 -7.34014809e-01
7.02918172e-01 -4.86362308e-01 -8.66443992e-01 1.46788523e-01
-4.16755788e-02 1.57112569e-01 1.42174661e+00 -1.00017452e+00
-2.22275704e-01 -5.75341284e-01 9.49352309e-02 9.87130404e-02
-4.10016626e-01 2.56464094e-01 8.67614806e-01 -5.35853982e-01
-4.47829455e-01 -2.35743627e-01 6.46259308e-01 2.29413398e-02
-3.76857549e-01 -3.11388284e-01 8.08145344e-01 -9.60657775e-01
1.31764972e+00 -2.14644814e+00 -4.03118849e-01 -1.61737755e-01
7.15835273e-01 6.64305687e-01 1.10305138e-01 4.70170021e-01
8.47081318e-02 8.19255412e-01 1.20124258e-01 -4.45731878e-01
2.21968606e-01 -8.20319727e-02 -3.67198199e-01 8.64509046e-01
1.93669692e-01 1.36995089e+00 -8.73043060e-01 -2.20606670e-01
1.43542439e-01 4.12344724e-01 -1.49490386e-01 -2.21643895e-01
1.53211668e-01 -4.75164829e-03 -3.12337309e-01 6.68639243e-01
4.99046057e-01 -1.84927478e-01 -1.46484852e-01 -3.08282766e-02
-7.11833909e-02 8.77423704e-01 2.07679980e-02 7.29092181e-01
-1.97624713e-01 1.43512070e+00 3.19556482e-02 -9.35156524e-01
8.19905281e-01 8.51882175e-02 -5.11533618e-01 -6.11224771e-01
5.47947824e-01 3.47945511e-01 2.97976285e-01 -5.95572054e-01
5.55543602e-01 -7.09924400e-01 3.11123896e-02 4.25242633e-01
-1.40750006e-01 -3.43659595e-02 -5.27444959e-01 2.18750477e-01
6.58828855e-01 -4.91768032e-01 3.98594320e-01 -4.12432015e-01
4.44009691e-01 1.67724445e-01 -2.29412746e-02 1.04154158e+00
-7.45085597e-01 2.73332179e-01 7.89323449e-01 -7.55355716e-01
-9.14498746e-01 -4.65740532e-01 -7.78557137e-02 9.19818580e-01
6.69850549e-03 -7.45556951e-02 -5.67263961e-01 -8.94724250e-01
3.62592787e-01 1.74220681e+00 -1.02378988e+00 -1.00737321e+00
-1.34885773e-01 -5.87306082e-01 1.12582231e+00 2.29504362e-01
6.07953966e-01 -8.11201394e-01 -8.07952523e-01 -2.48416767e-01
-2.81856716e-01 -6.38942301e-01 -4.02157545e-01 -1.89622954e-01
-7.05547273e-01 -8.73965323e-01 -1.93255797e-01 -2.68084317e-01
1.98673770e-01 3.31500590e-01 1.13954759e+00 4.63946968e-01
7.90283158e-02 2.64487058e-01 -3.76604050e-01 -5.03426433e-01
-6.26426995e-01 -2.29556933e-01 4.29594107e-02 9.96645540e-02
9.55987394e-01 -9.06898156e-02 -4.35496569e-01 -4.78737149e-03
-7.47783303e-01 -3.53334904e-01 3.93359214e-01 1.02247393e+00
-5.45418203e-01 -2.48757958e-01 7.55900502e-01 -6.88462555e-01
1.37293923e+00 -6.94944978e-01 -8.14988166e-02 -1.97397862e-02
-8.69097829e-01 -1.90024734e-01 4.88984317e-01 -5.58775008e-01
-9.73306537e-01 -9.11200523e-01 -3.67932357e-02 -4.00165319e-01
-1.26115724e-01 5.57732105e-01 2.67734110e-01 -2.98149973e-01
8.17878306e-01 3.82221043e-01 2.57665783e-01 -2.26730376e-01
1.55601770e-01 9.52199757e-01 6.92243204e-02 -1.94353815e-02
5.74308097e-01 2.76881248e-01 -7.18340337e-01 -8.65209579e-01
-1.32699645e+00 1.15505615e-02 -3.81268322e-01 -1.48365125e-01
5.12088597e-01 -5.03530383e-01 -9.44094002e-01 6.21485889e-01
-1.20024109e+00 1.09547945e-02 3.01176701e-02 1.99802786e-01
-2.50737399e-01 5.80777586e-01 -8.14181328e-01 -1.09988499e+00
-4.09682065e-01 -6.29481852e-01 7.85530388e-01 -1.82211921e-01
-7.07414567e-01 -1.33830833e+00 -3.39969307e-01 9.47441518e-01
7.06050992e-01 8.36165100e-02 1.02934551e+00 -1.41359556e+00
1.07391521e-01 -5.27508795e-01 -3.09406132e-01 5.67032635e-01
-3.80541347e-02 -2.64614910e-01 -9.61340308e-01 -7.28030577e-02
4.96529609e-01 -8.02860260e-01 7.83206940e-01 1.75996512e-01
1.10248816e+00 -9.97829139e-01 -1.62726983e-01 2.51024872e-01
9.68722999e-01 -7.54028186e-02 6.15276039e-01 2.55313605e-01
4.83946055e-01 5.17639697e-01 2.00854197e-01 3.19022119e-01
3.04373503e-01 4.71170962e-01 6.77316785e-01 3.50780964e-01
8.36902112e-02 -4.76237804e-01 5.98716438e-01 2.33079940e-01
3.29180300e-01 -5.81333637e-01 -9.29832458e-01 6.18246138e-01
-1.34448946e+00 -1.17914176e+00 -5.10672033e-01 1.87100017e+00
5.60896993e-01 2.74369031e-01 -1.93720963e-02 6.95729181e-02
5.47284365e-01 4.76902813e-01 -6.61696494e-01 -1.18254483e+00
-2.85660774e-01 -3.18503916e-01 3.09022635e-01 7.51332462e-01
-8.51780057e-01 1.07098484e+00 7.34455776e+00 7.03110158e-01
-1.09637678e+00 1.83370516e-01 9.39861953e-01 -4.00624067e-01
-4.17340487e-01 -5.53768694e-01 -4.32145119e-01 6.36983097e-01
1.22944093e+00 -2.84660727e-01 3.13031256e-01 6.99692786e-01
3.09419930e-01 -6.89177886e-02 -1.20884955e+00 8.76410365e-01
1.01018929e+00 -1.07213330e+00 3.71482819e-01 5.26927374e-02
4.18486834e-01 -2.90054590e-01 4.80447143e-01 4.38196629e-01
2.29457378e-01 -1.54844582e+00 1.11397994e+00 3.96094769e-01
3.46465141e-01 -6.16113484e-01 1.09876907e+00 6.77501678e-01
3.66880447e-01 -2.13528499e-01 -3.04105550e-01 -6.53691590e-01
-5.07856300e-03 3.96229774e-01 -8.01914811e-01 -8.21603183e-03
5.14055371e-01 6.94503129e-01 -6.80540740e-01 3.83855015e-01
-3.99200469e-01 9.24653113e-01 2.29054615e-01 -5.89647651e-01
5.59871852e-01 1.63853765e-01 6.62503362e-01 1.17477190e+00
-4.72505987e-02 4.31258157e-02 -4.37143117e-01 1.29644012e+00
-3.13588269e-02 -2.39977151e-01 -6.62337303e-01 -6.14882648e-01
2.19518423e-01 9.53038990e-01 -1.33739591e-01 -4.42344755e-01
-3.51096630e-01 1.38740730e+00 4.10249174e-01 1.72495440e-01
-9.08329487e-01 -1.79934844e-01 7.58854926e-01 2.20655844e-01
-2.27647033e-02 -1.18960842e-01 -7.34139860e-01 -1.32117486e+00
-1.93446323e-01 -1.21868300e+00 6.86447248e-02 -9.35139656e-01
-1.79290199e+00 2.13910371e-01 -2.80474484e-01 -4.15668637e-01
-7.79211447e-02 -7.57065237e-01 -4.66035277e-01 8.73046339e-01
-1.38626122e+00 -1.16386867e+00 -5.31853810e-02 4.56300914e-01
5.87220311e-01 1.38303917e-02 6.79336131e-01 -6.21281331e-03
-4.29282814e-01 7.46370196e-01 2.60983445e-02 3.21521968e-01
6.18202806e-01 -1.00419223e+00 4.33645308e-01 4.64096129e-01
-5.42163625e-02 6.61108017e-01 8.24557900e-01 -6.37844503e-01
-9.46361423e-01 -5.02923667e-01 1.37795484e+00 -1.10566795e+00
1.15109503e+00 -4.03027475e-01 -1.00746226e+00 8.32383037e-01
2.19680041e-01 -2.72383422e-01 8.98289263e-01 1.81021795e-01
-7.08206296e-01 3.33956093e-01 -1.41688085e+00 5.05807579e-01
9.07953262e-01 -6.19102001e-01 -1.09088337e+00 3.23552758e-01
2.24656671e-01 5.83089888e-03 -3.35853398e-01 -1.87760323e-01
7.74879992e-01 -1.30529308e+00 6.52233720e-01 -1.34595120e+00
9.03247714e-01 4.47237074e-01 2.91725934e-01 -1.51729584e+00
-6.71538472e-01 -2.53297657e-01 -2.92554796e-01 6.69931889e-01
3.14578354e-01 -9.65934157e-01 4.28183824e-01 7.87450671e-01
-1.35918006e-01 -5.20493567e-01 -9.61555183e-01 -6.37482345e-01
6.47499621e-01 -1.85579091e-01 2.64592260e-01 1.43196285e+00
3.92360389e-01 4.42184120e-01 -7.28343308e-01 7.57146580e-03
3.00895661e-01 -1.00782119e-01 2.17068106e-01 -1.18587887e+00
1.27532616e-01 -7.55208969e-01 -5.79441071e-01 -1.04017568e+00
3.60185713e-01 -6.10167503e-01 -3.64376694e-01 -1.21744835e+00
1.81634158e-01 2.32102156e-01 6.92841932e-02 5.13345063e-01
-2.01307163e-01 2.92709380e-01 2.39912480e-01 3.82137448e-01
-5.06542087e-01 5.40572047e-01 1.06418717e+00 -3.11472416e-01
3.21241796e-01 -1.44860178e-01 -1.23698676e+00 8.30091178e-01
1.13300061e+00 -3.21646899e-01 -8.33681673e-02 -6.80054009e-01
3.66306245e-01 -2.09586248e-01 9.08647239e-01 -3.48752975e-01
-2.40635332e-02 -1.41376471e-02 6.20720506e-01 -9.22916979e-02
3.56107712e-01 -5.53880632e-01 -5.43231070e-01 6.08770907e-01
-6.71230733e-01 -7.29895681e-02 2.07590774e-01 5.08316338e-01
-6.97807893e-02 -4.68911678e-01 8.44847500e-01 -3.01709145e-01
-2.62829006e-01 -3.41563404e-01 -8.94558787e-01 8.94349664e-02
6.82888687e-01 -4.91548002e-01 -7.33832657e-01 -1.06150115e+00
-6.63327098e-01 -9.78781004e-03 3.14480513e-01 7.03635216e-01
7.13453710e-01 -1.05607498e+00 -8.42401087e-01 -1.16673494e-02
1.30924791e-01 -1.01874697e+00 -1.40973264e-02 1.00611937e+00
-3.42649162e-01 8.19965899e-01 -2.41941184e-01 1.37210727e-01
-1.10894990e+00 7.07116425e-01 7.53047168e-01 2.26603150e-02
-1.80806428e-01 8.37674320e-01 -2.69816555e-02 4.83577512e-02
-1.20206408e-01 4.42156978e-02 -1.15671672e-01 8.63557383e-02
7.40617156e-01 7.94809341e-01 3.12325489e-02 -8.75809014e-01
-4.34490383e-01 -2.63082862e-01 -4.62991506e-01 1.11174293e-01
7.92248070e-01 -2.40302697e-01 -5.70724197e-02 6.81838632e-01
1.12852299e+00 8.79126936e-02 -6.64775014e-01 2.07811594e-01
3.89459245e-02 -7.99748719e-01 4.48917925e-01 -1.49228060e+00
-7.38357186e-01 1.12187421e+00 1.23213097e-01 6.06525004e-01
4.96570051e-01 -1.19994991e-01 5.75978577e-01 6.28969818e-02
2.12289885e-01 -7.96824336e-01 2.87584007e-01 3.15523773e-01
1.23426139e+00 -1.39888930e+00 2.98917443e-02 -7.35753775e-02
-8.24141800e-01 1.09853148e+00 5.86876035e-01 -1.36495978e-01
2.65920788e-01 -2.76156843e-01 1.04926908e-02 -5.76315999e-01
-6.16120040e-01 4.08761740e-01 3.76241982e-01 6.41067088e-01
5.16208470e-01 4.00695764e-02 -7.00903893e-01 6.34112835e-01
-5.22019923e-01 -2.70004541e-01 7.19829082e-01 4.12577301e-01
-5.62848032e-01 -3.55245531e-01 -3.70866776e-01 8.42909932e-01
-8.22793305e-01 -5.45600057e-01 -1.42142904e+00 8.41262162e-01
-9.87660885e-02 1.35682404e+00 3.81291695e-02 -3.84056240e-01
-1.24246478e-02 3.83659303e-01 3.12979072e-01 -4.36569005e-01
-9.79145467e-01 -5.33392072e-01 4.71117258e-01 -4.79501009e-01
-7.30437860e-02 -7.23111689e-01 -3.84002000e-01 -9.49733853e-01
-7.43485749e-01 3.75368893e-02 3.69064748e-01 1.15719259e+00
4.88302380e-01 1.02400735e-01 1.06678300e-01 -3.83167595e-01
-1.13758433e+00 -1.22911131e+00 -5.25331020e-01 4.84593391e-01
7.91143894e-01 -5.29076636e-01 -1.12792957e+00 -3.74810755e-01] | [8.09903335571289, 10.295369148254395] |
b5ffdf49-f01e-4d74-84f1-f2d3883a410f | mantra-net-manipulation-tracing-network-for | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Wu_ManTra-Net_Manipulation_Tracing_Network_for_Detection_and_Localization_of_Image_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Wu_ManTra-Net_Manipulation_Tracing_Network_for_Detection_and_Localization_of_Image_CVPR_2019_paper.pdf | ManTra-Net: Manipulation Tracing Network for Detection and Localization of Image Forgeries With Anomalous Features | To fight against real-life image forgery, which commonly involves different types and combined manipulations, we propose a unified deep neural architecture called ManTra-Net. Unlike many existing solutions, ManTra-Net is an end-to-end network that performs both detection and localization without extra preprocessing and postprocessing. \manifold is a fully convolutional network and handles images of arbitrary sizes and many known forgery types such splicing, copy-move, removal, enhancement, and even unknown types. This paper has three salient contributions. We design a simple yet effective self-supervised learning task to learn robust image manipulation traces from classifying 385 image manipulation types. Further, we formulate the forgery localization problem as a local anomaly detection problem, design a Z-score feature to capture local anomaly, and propose a novel long short-term memory solution to assess local anomalies. Finally, we carefully conduct ablation experiments to systematically optimize the proposed network design. Our extensive experimental results demonstrate the generalizability, robustness and superiority of ManTra-Net, not only in single types of manipulations/forgeries, but also in their complicated combinations.
| [' Premkumar Natarajan', ' Wael AbdAlmageed', 'Yue Wu'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['image-forensics', 'fake-image-detection'] | ['computer-vision', 'computer-vision'] | [ 2.80995429e-01 -7.17891574e-01 1.51277825e-01 -2.91317344e-01
-6.64881229e-01 -5.19280195e-01 3.20607305e-01 -9.71943438e-02
-2.40329862e-01 2.43165955e-01 -5.38431183e-02 -3.92813981e-01
-2.26058848e-02 -6.17101490e-01 -8.99093688e-01 -5.91835797e-01
-3.10940355e-01 -4.16463077e-01 1.12398207e-01 -2.63926268e-01
6.01825655e-01 7.23020971e-01 -1.33510172e+00 5.25106549e-01
6.69912815e-01 1.15855026e+00 -2.14878604e-01 8.36103916e-01
3.82200122e-01 1.08501470e+00 -8.57671618e-01 -6.76680923e-01
5.51267862e-01 -3.27775270e-01 -6.26085818e-01 5.03539026e-01
7.21591473e-01 -7.95858622e-01 -8.71841788e-01 1.34927297e+00
7.80912161e-01 7.80732483e-02 2.96984673e-01 -1.47387862e+00
-1.06134057e+00 2.87295043e-01 -9.07044411e-01 4.76057589e-01
1.69318005e-01 8.44518661e-01 5.61790287e-01 -1.11324906e+00
3.73360366e-01 1.26523280e+00 1.15613973e+00 2.83458024e-01
-6.73118472e-01 -7.89323032e-01 4.94330227e-02 4.41403151e-01
-1.32270432e+00 -4.74902272e-01 7.99448490e-01 -1.01368457e-01
8.56109202e-01 2.12129161e-01 -9.90965664e-02 1.37533641e+00
3.61743093e-01 9.22660470e-01 8.67344677e-01 -3.01439494e-01
-1.21491976e-01 -4.77442294e-01 1.87826723e-01 1.03576458e+00
2.74404287e-01 2.51711398e-01 -5.08297920e-01 -2.23819926e-01
7.31778979e-01 2.78921127e-01 -2.59614587e-01 -2.03104392e-02
-1.15284395e+00 5.99576950e-01 3.91585082e-01 1.69503272e-01
-2.74528831e-01 4.18937504e-01 6.90224886e-01 6.62686288e-01
1.05808631e-01 5.01043677e-01 -4.04061854e-01 3.14809307e-02
-8.70365620e-01 1.69413149e-01 2.42833421e-01 1.01646698e+00
7.11587727e-01 3.11188906e-01 -3.41634125e-01 7.92937696e-01
-1.13313816e-01 1.75187051e-01 5.39393842e-01 -6.62793219e-01
5.22427559e-01 5.45849741e-01 -1.63411155e-01 -1.42523909e+00
-3.66126239e-01 -5.71146369e-01 -1.25618541e+00 2.05096960e-01
2.44491741e-01 -1.68594494e-01 -1.05712855e+00 1.41262102e+00
-4.31686044e-02 5.00892460e-01 -2.78793991e-01 9.40633953e-01
5.10374427e-01 2.06861466e-01 -1.53717577e-01 -4.62944880e-02
1.12798107e+00 -1.09758890e+00 -8.17710280e-01 -2.45809361e-01
7.53632784e-01 -7.94462621e-01 1.10978639e+00 5.55386424e-01
-1.03484344e+00 -5.36586523e-01 -1.20207000e+00 -1.76259696e-01
-5.66631794e-01 2.17646956e-01 5.96894383e-01 7.90449917e-01
-1.03071976e+00 9.92544293e-01 -5.51381052e-01 4.70546819e-02
7.54673302e-01 1.57730147e-01 -5.07814765e-01 -2.28222758e-01
-9.83455539e-01 6.61536992e-01 3.66650909e-01 4.79957283e-01
-1.18715739e+00 -6.91585064e-01 -8.22646499e-01 9.04842094e-02
6.07825816e-01 -4.12337840e-01 1.01826310e+00 -9.04986680e-01
-1.04317832e+00 9.46378171e-01 6.88301921e-02 -4.47213918e-01
5.29085994e-01 -2.28236869e-01 -8.86182189e-01 -5.74350581e-02
5.18388040e-02 8.34557414e-03 1.20190120e+00 -1.17404592e+00
-4.50012654e-01 -5.93599200e-01 -2.59594887e-01 -3.58308882e-01
-4.56070125e-01 5.66751480e-01 -5.26811361e-01 -1.22245061e+00
-4.10525352e-02 -4.48801666e-01 -1.99949041e-01 1.89665183e-01
-6.01201177e-01 1.05067462e-01 1.08824813e+00 -8.49284470e-01
1.33838999e+00 -2.60775018e+00 -4.48539287e-01 2.29985729e-01
4.17456657e-01 5.97567141e-01 -6.16491497e-01 2.77691483e-01
-4.45502520e-01 1.33101732e-01 -5.80432832e-01 -3.86215478e-01
-3.46695445e-02 -1.80847831e-02 -6.20384276e-01 9.09870148e-01
5.22362649e-01 1.13114500e+00 -8.36515725e-01 -1.49488047e-01
1.98978841e-01 1.01797514e-01 -3.71845841e-01 3.00117522e-01
-3.61770615e-02 -1.51146412e-01 -1.57594949e-01 1.17268610e+00
1.05906403e+00 -1.20808683e-01 -3.81210178e-01 -3.29660654e-01
1.54599547e-01 -2.98307806e-01 -1.12686276e+00 1.90215337e+00
-2.23964781e-01 5.47761500e-01 2.83610165e-01 -1.00055540e+00
7.88221657e-01 -1.07544452e-01 2.36893550e-01 -6.57091737e-01
3.37957501e-01 2.09268808e-01 -2.08828300e-01 -7.58457780e-01
7.41498232e-01 2.49053016e-01 -1.80511594e-01 6.04536295e-01
2.69529670e-01 4.60993856e-01 -1.11026019e-01 2.11244568e-01
1.72586465e+00 -2.32643545e-01 7.11599737e-02 -2.87872702e-02
3.83446962e-01 -3.81202549e-01 6.04385316e-01 1.16024852e+00
-5.89662015e-01 8.36096048e-01 4.30525452e-01 -6.75495744e-01
-7.18925595e-01 -9.85056281e-01 1.94461226e-01 1.14557970e+00
3.52651954e-01 -3.91673863e-01 -5.79609513e-01 -1.05110812e+00
2.72704884e-02 5.43839097e-01 -5.31334519e-01 -7.14036822e-01
-6.40143573e-01 -1.20917153e+00 1.16086507e+00 4.84911025e-01
1.07947099e+00 -1.22358930e+00 -2.78933167e-01 -3.89210917e-02
-1.63309738e-01 -1.13096440e+00 -8.14452171e-01 -4.35194857e-02
-5.81162512e-01 -1.31316185e+00 -3.42940271e-01 -8.33424091e-01
5.74021161e-01 6.39506638e-01 1.03514254e+00 6.49318933e-01
-8.75615776e-01 2.45186672e-01 -3.40727121e-01 -1.20870911e-01
-3.80926579e-01 -4.12808299e-01 -1.91062801e-02 2.52532572e-01
2.22467273e-01 -5.57163298e-01 -6.41178250e-01 4.83779609e-01
-1.18751931e+00 -6.90978825e-01 8.64136577e-01 9.52733219e-01
2.47369260e-01 5.93972981e-01 4.72402602e-01 -6.81556404e-01
8.22416067e-01 -5.42523146e-01 -4.42551345e-01 4.68804687e-01
-4.32649255e-01 -2.14700013e-01 6.08059347e-01 -3.95463288e-01
-9.61666107e-01 -1.25840709e-01 -2.26092666e-01 -9.33382690e-01
-3.87608528e-01 1.88549384e-01 -4.68019545e-01 -4.30388808e-01
8.96565437e-01 5.32552302e-01 -2.36261748e-02 -3.99579138e-01
4.87595499e-01 5.52873552e-01 1.24528861e+00 -3.73148292e-01
1.14972210e+00 4.15343910e-01 -1.79546848e-01 -8.24624538e-01
-5.76537549e-01 -4.52577323e-01 -6.02406859e-01 -1.36411622e-01
4.21222985e-01 -8.51970434e-01 -7.21373737e-01 1.38518310e+00
-1.20507908e+00 -2.80111402e-01 1.51491119e-02 -2.16001689e-01
-3.66117001e-01 1.04981172e+00 -8.44253600e-01 -5.91748297e-01
-5.55824816e-01 -1.02595723e+00 1.12265098e+00 -1.50892884e-01
2.55698770e-01 -7.95340300e-01 -5.10405183e-01 2.97624320e-01
4.49886888e-01 5.26116014e-01 6.26012146e-01 -4.81334358e-01
-6.49336457e-01 -4.89911586e-01 -5.60518265e-01 5.76151609e-01
1.45479053e-01 -5.54340743e-02 -9.46147740e-01 -5.77203155e-01
1.61701292e-01 -5.08047283e-01 1.12850893e+00 -1.06865847e-02
2.03032589e+00 -3.29799265e-01 -3.30795161e-02 1.02438867e+00
1.36275089e+00 2.44161068e-03 1.04473484e+00 4.13984835e-01
8.18776369e-01 6.44169003e-02 4.50572163e-01 4.74318027e-01
-1.10084780e-01 3.28566551e-01 8.55877161e-01 -2.24661782e-01
-1.05515033e-01 -1.22399673e-01 6.24021411e-01 3.32134724e-01
3.22484821e-01 -3.93328041e-01 -5.61930835e-01 6.41860664e-01
-1.75233805e+00 -1.09629500e+00 -1.01191409e-01 2.01687646e+00
4.57032144e-01 3.14995423e-02 7.06448779e-03 2.93001503e-01
9.54899073e-01 2.43950278e-01 -5.61510146e-01 -1.81640968e-01
-5.05725205e-01 2.59496450e-01 6.75488889e-01 -4.41341102e-03
-1.45507240e+00 1.02046478e+00 6.80086708e+00 1.17364657e+00
-9.33881700e-01 1.03632234e-01 6.96685314e-01 -9.66716856e-02
2.12520137e-01 -3.44261736e-01 -2.23288119e-01 6.15388095e-01
6.08706176e-01 4.30886239e-01 5.42577267e-01 7.09007561e-01
1.49035707e-01 1.97628170e-01 -8.13581288e-01 1.16591799e+00
5.00139475e-01 -1.45708442e+00 1.56114185e-02 -2.09954426e-01
4.97191608e-01 -1.54791728e-01 2.98780054e-01 2.84961611e-01
3.11332375e-01 -1.05533290e+00 5.92412531e-01 2.36249000e-01
8.14755023e-01 -9.01712120e-01 7.49344945e-01 2.66962528e-01
-7.88863122e-01 -4.08532917e-01 -3.92213464e-01 4.91644032e-02
3.55706736e-02 6.44576132e-01 -2.17811853e-01 6.42860234e-01
8.86636138e-01 8.23865235e-01 -9.03835356e-01 1.14526093e+00
-2.15692241e-02 6.73496485e-01 9.47908387e-02 6.02368295e-01
2.84501135e-01 1.40499175e-01 5.54234862e-01 1.29475522e+00
3.25330257e-01 -8.99739042e-02 4.69665192e-02 9.63993609e-01
-4.50463504e-01 -2.52778560e-01 -5.67169130e-01 1.48670956e-01
4.02818948e-01 1.22168028e+00 -6.75969303e-01 -1.93020571e-02
-3.51368546e-01 1.58066082e+00 3.24332744e-01 2.80346125e-01
-9.45347965e-01 -8.14141870e-01 7.71312952e-01 -1.59275442e-01
2.86658168e-01 -7.22064525e-02 -2.66201794e-01 -1.38056791e+00
4.22775000e-01 -1.20137691e+00 6.57401621e-01 -6.97527528e-01
-1.68827236e+00 3.11478317e-01 -4.18703675e-01 -1.11061025e+00
3.05322468e-01 -8.16185534e-01 -1.13468742e+00 4.37886477e-01
-1.47179496e+00 -1.19512451e+00 -4.75744843e-01 9.03823912e-01
5.51353872e-01 -5.22426844e-01 5.05221844e-01 5.62671244e-01
-1.11977565e+00 1.24593318e+00 -3.45291421e-02 7.15187907e-01
8.47024798e-01 -9.91168439e-01 7.35376239e-01 1.56080401e+00
-4.33974713e-02 6.44808471e-01 4.31440771e-01 -6.31655216e-01
-1.61541259e+00 -1.65920031e+00 3.26560944e-01 -3.44332099e-01
7.93953001e-01 -3.81921649e-01 -1.17611611e+00 6.87395155e-01
1.38338476e-01 5.70362687e-01 3.23124588e-01 -2.27588564e-01
-7.60492206e-01 3.96076217e-02 -1.44072938e+00 5.20504713e-01
1.40426433e+00 -6.76827073e-01 -2.97588885e-01 4.34325665e-01
7.11507678e-01 -4.81843740e-01 -6.44837618e-01 4.46138024e-01
8.32295939e-02 -1.26601899e+00 1.14677298e+00 -7.45486498e-01
5.66453934e-01 -2.18888938e-01 2.26324727e-03 -1.10049152e+00
-5.21162808e-01 -1.16190732e+00 -3.04239720e-01 1.15064192e+00
7.86369387e-03 -4.92623061e-01 7.36810207e-01 2.05086708e-01
-5.30271828e-01 -5.39933801e-01 -8.26171339e-01 -9.65707302e-01
-1.46454088e-02 -6.45693243e-01 7.73677289e-01 1.21919119e+00
-3.09539050e-01 -3.64297211e-01 -8.76521945e-01 5.43687224e-01
7.27381349e-01 -5.87782264e-02 8.26443613e-01 -5.36311150e-01
-2.63139993e-01 -4.94490266e-01 -7.82495141e-01 -8.75664592e-01
1.89287111e-01 -7.10181952e-01 2.09338754e-01 -7.49858379e-01
7.27168173e-02 -2.39796251e-01 -4.02043402e-01 6.78710461e-01
-4.98359710e-01 3.60884428e-01 -1.03253938e-01 7.76378885e-02
-7.55265117e-01 5.74439585e-01 9.73166883e-01 -2.70935267e-01
2.11234555e-01 -3.48475948e-02 -7.77916789e-01 6.07498169e-01
7.57558405e-01 -2.04218015e-01 -9.31854248e-02 -6.31448567e-01
-3.50754410e-02 -9.89995301e-02 8.17245543e-01 -1.17662954e+00
8.48067924e-02 -1.57928355e-02 5.09182215e-01 -5.70225179e-01
-4.35315706e-02 -6.99627161e-01 -3.33129227e-01 2.79313415e-01
-2.15110675e-01 4.54564691e-01 3.32876801e-01 8.33081543e-01
-1.72579318e-01 -4.96603623e-02 9.06388462e-01 -1.01686515e-01
-1.14557064e+00 5.88927925e-01 -3.33964646e-01 -3.39050777e-02
1.05276883e+00 -2.61625707e-01 -6.29908741e-01 -2.86831111e-01
-4.41618919e-01 2.03607425e-01 3.81158978e-01 7.45641232e-01
1.03615618e+00 -1.38771605e+00 -6.89994276e-01 4.54069495e-01
2.99654216e-01 -1.90803602e-01 5.37108600e-01 5.47751725e-01
-5.87113738e-01 -2.53554255e-01 -3.08706880e-01 -2.55577117e-01
-1.14878190e+00 1.06511569e+00 5.08725107e-01 -8.73163193e-02
-6.68484986e-01 8.54974389e-01 -1.66042343e-01 -4.88221526e-01
4.36216116e-01 -1.01270348e-01 3.73476624e-01 -5.26648164e-01
6.89122021e-01 6.38989210e-01 3.31827313e-01 -5.44218481e-01
-2.33934790e-01 1.77392170e-01 -2.93310106e-01 6.40263319e-01
1.16699088e+00 -9.46379229e-02 -3.83498341e-01 -2.90530473e-01
1.30593944e+00 -3.80299062e-01 -1.06270909e+00 -4.17345166e-01
1.33640915e-01 -7.98171937e-01 2.14880258e-01 -9.25285757e-01
-1.56984186e+00 8.34751904e-01 6.76181257e-01 -1.05379045e-01
1.51836050e+00 -4.20401335e-01 1.12747812e+00 5.25169134e-01
8.20824206e-02 -1.02041543e+00 4.27398860e-01 4.42048937e-01
9.88225341e-01 -1.28157210e+00 -1.27686588e-02 -3.07976604e-01
-4.42194313e-01 1.08003259e+00 9.19313371e-01 -1.51799619e-01
5.46649694e-01 3.40347618e-01 1.25491410e-01 -4.24982429e-01
-4.60333377e-01 1.42343953e-01 -4.93128374e-02 5.56749284e-01
-5.15770651e-02 -2.96411037e-01 2.99012244e-01 5.80646038e-01
8.77417848e-02 -1.31343096e-01 6.31112218e-01 1.07152689e+00
-1.53195798e-01 -7.71077633e-01 -5.68693042e-01 6.27604306e-01
-5.94353020e-01 -5.75288124e-02 -4.89042073e-01 5.97932696e-01
3.34942013e-01 1.01932478e+00 5.54153584e-02 -7.03220844e-01
3.75114352e-01 -2.09632829e-01 1.86921671e-01 -1.11397319e-01
-8.80920231e-01 -3.17983180e-01 -2.85691738e-01 -9.94015455e-01
5.10916486e-02 -5.09866655e-01 -6.73483372e-01 -5.16193032e-01
-5.19835234e-01 -4.98706013e-01 2.99357235e-01 8.61154497e-01
6.31715000e-01 4.57777798e-01 8.51786554e-01 -7.28560507e-01
-9.78755176e-01 -9.55927551e-01 -6.51189685e-01 7.92917728e-01
6.84349954e-01 -5.66201687e-01 -5.38240910e-01 -1.09105498e-01] | [12.333465576171875, 0.9579513669013977] |
b1738197-f29c-40ba-a0cf-d334b0c72852 | a-comprehensive-survey-on-deep-music | 2011.06801 | null | https://arxiv.org/abs/2011.06801v1 | https://arxiv.org/pdf/2011.06801v1.pdf | A Comprehensive Survey on Deep Music Generation: Multi-level Representations, Algorithms, Evaluations, and Future Directions | The utilization of deep learning techniques in generating various contents (such as image, text, etc.) has become a trend. Especially music, the topic of this paper, has attracted widespread attention of countless researchers.The whole process of producing music can be divided into three stages, corresponding to the three levels of music generation: score generation produces scores, performance generation adds performance characteristics to the scores, and audio generation converts scores with performance characteristics into audio by assigning timbre or generates music in audio format directly. Previous surveys have explored the network models employed in the field of automatic music generation. However, the development history, the model evolution, as well as the pros and cons of same music generation task have not been clearly illustrated. This paper attempts to provide an overview of various composition tasks under different music generation levels, covering most of the currently popular music generation tasks using deep learning. In addition, we summarize the datasets suitable for diverse tasks, discuss the music representations, the evaluation methods as well as the challenges under different levels, and finally point out several future directions. | ['Xinyu Yang', 'Jing Luo', 'Shulei Ji'] | 2020-11-13 | null | null | null | null | ['audio-generation'] | ['audio'] | [ 4.15330917e-01 -3.06371957e-01 8.28815624e-02 1.76691964e-01
-7.50439644e-01 -7.29696393e-01 5.49778879e-01 -4.24385428e-01
1.98084936e-01 7.27603972e-01 5.55352509e-01 3.89355749e-01
-4.65603262e-01 -8.53909910e-01 -3.08286965e-01 -8.99604440e-01
6.83063418e-02 3.82217228e-01 -3.96335632e-01 -3.56836200e-01
4.92975116e-01 3.07294458e-01 -1.80978656e+00 6.12684667e-01
5.08216560e-01 8.64117801e-01 4.78045121e-02 8.89372706e-01
-2.65802085e-01 7.79153287e-01 -1.50648630e+00 -5.70793390e-01
1.00188069e-01 -9.98605192e-01 -7.66712844e-01 -9.36448649e-02
4.29319352e-01 -1.96030408e-01 -2.10440099e-01 1.00857127e+00
1.01989162e+00 2.70073682e-01 8.77944827e-01 -1.33649039e+00
-9.33806360e-01 1.58057034e+00 -2.38387913e-01 -2.59633064e-01
4.62418646e-01 -1.14763111e-01 1.22778440e+00 -7.36006975e-01
4.97274816e-01 9.78722095e-01 7.74272323e-01 7.58049965e-01
-1.10102510e+00 -9.36857224e-01 -4.59646434e-01 1.60864934e-01
-1.36217606e+00 -2.21666455e-01 1.17773604e+00 -5.05360663e-01
6.59186900e-01 4.20001239e-01 1.10391116e+00 1.20672870e+00
-1.44478410e-01 1.15732539e+00 7.70141304e-01 -5.26648998e-01
-6.59829229e-02 -2.83969671e-01 -4.50893342e-01 5.95625937e-02
-2.06549957e-01 1.88501239e-01 -1.05435884e+00 -8.35981965e-02
1.04945862e+00 -5.11222541e-01 -1.38936073e-01 -3.55064757e-02
-1.55241525e+00 7.25818753e-01 5.15539706e-01 6.62887871e-01
-3.98668051e-01 5.69560528e-01 6.12375855e-01 3.46871585e-01
9.63928029e-02 8.86816859e-01 -1.21498749e-01 -2.41375640e-01
-1.41985261e+00 8.71769369e-01 5.65185666e-01 8.75989676e-01
1.46014169e-01 1.00168359e+00 -5.13509750e-01 1.11943030e+00
-1.03798702e-01 2.54599065e-01 8.47882807e-01 -1.04891694e+00
2.96280146e-01 2.01731026e-01 -8.53826702e-02 -8.58004928e-01
-1.51764363e-01 -7.15562820e-01 -1.15007007e+00 2.78091431e-01
2.11013854e-01 -3.71929914e-01 -5.98849118e-01 1.79724407e+00
-2.83730209e-01 2.12346837e-01 1.76962130e-02 8.14894676e-01
1.41425407e+00 9.19041693e-01 -1.57087713e-01 -1.05341621e-01
1.06955218e+00 -1.01969934e+00 -9.06228423e-01 3.91171187e-01
-4.12545726e-02 -1.29180396e+00 8.81675124e-01 7.84133613e-01
-1.60839236e+00 -1.20569277e+00 -1.16121984e+00 7.55297616e-02
-2.11495146e-01 4.76099461e-01 8.23129117e-01 5.12351751e-01
-1.32399368e+00 1.15463316e+00 -1.30334303e-01 5.03690541e-02
2.05549911e-01 3.44906241e-01 1.32768244e-01 6.43576205e-01
-1.43828690e+00 4.46970016e-01 6.39568985e-01 -1.56052917e-01
-7.82199383e-01 -6.57061517e-01 -5.34445584e-01 5.18568605e-02
-2.54036486e-01 -1.06943464e+00 1.64485157e+00 -1.23167694e+00
-1.93386209e+00 7.37632811e-01 2.73847193e-01 -3.71116132e-01
3.63615513e-01 -3.77822518e-01 -6.21342182e-01 -1.70584843e-01
1.95553396e-02 1.05562317e+00 8.80827010e-01 -9.39579606e-01
-6.26793742e-01 1.25197470e-01 4.42826934e-02 4.52917576e-01
-2.67665654e-01 1.79911956e-01 -1.59394026e-01 -1.42327237e+00
-2.00025856e-01 -9.00631011e-01 9.74768847e-02 -4.41991925e-01
-4.78898436e-01 -4.93853807e-01 2.24696621e-01 -3.99952501e-01
1.66324937e+00 -2.20518875e+00 6.21401906e-01 -1.71757266e-02
-1.75871462e-01 1.16575971e-01 -2.24870786e-01 6.69087052e-01
-4.33610290e-01 6.85478002e-02 -9.07503814e-02 -7.29791149e-02
2.77280003e-01 -3.43540192e-01 -6.66910946e-01 -3.75320584e-01
5.32341525e-02 1.02895963e+00 -7.87919104e-01 -2.50905603e-01
6.97163194e-02 4.97716427e-01 -4.18425441e-01 1.97691679e-01
-9.20952111e-02 4.70895141e-01 -2.91137725e-01 4.70290720e-01
2.01703578e-01 9.65562612e-02 -6.94568008e-02 -2.98747897e-01
-2.84428567e-01 6.26577795e-01 -1.39597487e+00 2.12251306e+00
-3.60767752e-01 6.48956895e-01 -3.11000586e-01 -5.40078878e-01
1.21986055e+00 8.33307147e-01 6.00705862e-01 -2.79144704e-01
6.30186349e-02 4.55956757e-01 2.68582374e-01 -2.19537035e-01
9.48926270e-01 -4.38927829e-01 -3.67646247e-01 7.28783369e-01
2.24362776e-01 -6.10956371e-01 5.28519332e-01 -2.01451138e-01
6.58942580e-01 4.82900649e-01 4.11021650e-01 2.17561707e-01
4.02201623e-01 -2.35101301e-02 1.87348485e-01 5.66802382e-01
1.43063918e-01 8.65176737e-01 1.55764788e-01 -3.63838404e-01
-1.09246612e+00 -1.27612281e+00 1.01824068e-01 1.33425510e+00
-2.56760240e-01 -4.99168575e-01 -8.83662403e-01 -2.42682099e-02
-2.18675449e-01 5.81496119e-01 -4.77949291e-01 -1.44759327e-01
-7.15396345e-01 -7.92262018e-01 1.08362913e+00 6.60078526e-01
4.95555192e-01 -2.09960651e+00 -1.99108124e-01 4.53633100e-01
-5.00439227e-01 -1.32051751e-01 -3.54155779e-01 -1.05454706e-01
-9.57800984e-01 -7.00036407e-01 -1.25480890e+00 -1.02211916e+00
-1.59046352e-01 -2.89874747e-02 1.42154896e+00 -1.49306491e-01
-1.90405816e-01 -6.54960796e-02 -4.75368172e-01 -7.20555544e-01
-5.94001114e-01 3.22185993e-01 9.70284920e-03 -1.21861957e-01
9.03685912e-02 -8.30669105e-01 -5.37019074e-01 -9.05160829e-02
-9.30499375e-01 2.95806736e-01 6.83103740e-01 6.23170495e-01
5.54682195e-01 1.48982510e-01 1.14676106e+00 -6.28722310e-01
1.26000631e+00 -1.44540042e-01 1.26814455e-01 -1.78898677e-01
-5.12535945e-02 -2.13574067e-01 5.11551023e-01 -6.13961995e-01
-9.77567375e-01 -3.96879502e-02 -4.09832031e-01 -1.53268814e-01
-2.07739815e-01 4.72467303e-01 1.38344150e-02 4.46958274e-01
9.21039701e-01 3.18846077e-01 -4.37540859e-01 -6.21618271e-01
5.63774168e-01 6.54731631e-01 6.93540752e-01 -7.45162487e-01
8.75209212e-01 8.66805986e-02 -2.78633267e-01 -5.11348128e-01
-8.54461372e-01 -1.71566352e-01 -6.39144480e-01 -5.44087946e-01
6.79053128e-01 -7.12720931e-01 -2.94993371e-01 6.84775829e-01
-1.22165191e+00 -1.91493705e-01 -7.86459386e-01 3.92748028e-01
-9.95678365e-01 -1.07578926e-01 -1.01017964e+00 -6.51306450e-01
-8.92389476e-01 -1.05018079e+00 1.06529570e+00 4.46460545e-01
-8.61957252e-01 -6.33480310e-01 3.12782586e-01 3.09009224e-01
4.81965721e-01 4.19012368e-01 9.52116907e-01 -1.20999627e-01
-3.28585982e-01 -1.62686601e-01 1.16920896e-01 4.13660288e-01
1.99339494e-01 7.00197145e-02 -1.19772983e+00 9.51106027e-02
-3.86374831e-01 -6.20608330e-01 8.33948493e-01 5.43738902e-01
1.45466077e+00 -2.93416940e-02 2.79164582e-01 5.38357198e-01
8.99239361e-01 4.14481252e-01 7.15441406e-01 2.84997016e-01
5.53168952e-01 4.76885974e-01 3.06070060e-01 4.28121269e-01
-3.22585136e-01 7.71814108e-01 1.08367547e-01 1.15353344e-02
-7.43200004e-01 -4.29120004e-01 5.28632820e-01 1.44525814e+00
-7.77729094e-01 -2.06941113e-01 -2.89382428e-01 3.85551035e-01
-1.61026263e+00 -1.44315207e+00 -1.41036272e-01 1.93946314e+00
1.12329125e+00 -1.68015823e-01 4.81417835e-01 9.36830938e-01
8.77411962e-01 2.79151887e-01 -5.61477900e-01 -5.05128384e-01
-2.43339688e-01 8.65926862e-01 -3.59204769e-01 -1.59690648e-01
-1.05303979e+00 1.01034594e+00 7.25473356e+00 1.12184191e+00
-1.08751559e+00 -1.66061401e-01 1.82906285e-01 -2.95730710e-01
-8.82738605e-02 -4.08793867e-01 -3.57312471e-01 2.57895261e-01
7.25505233e-01 -3.94902229e-01 7.14392424e-01 8.23783219e-01
1.31304294e-01 7.24317372e-01 -1.02056646e+00 1.29141688e+00
2.47375414e-01 -1.42633367e+00 7.19179392e-01 -1.79561511e-01
1.23990357e+00 -3.52937937e-01 3.80050898e-01 5.50663710e-01
2.02822909e-01 -1.09683371e+00 1.14484370e+00 6.26379251e-01
1.08867466e+00 -1.03729582e+00 5.58931351e-01 -1.28357843e-01
-1.43720663e+00 3.42159793e-02 -1.60691559e-01 -3.31577539e-01
3.23645324e-01 2.99169838e-01 -5.79154491e-01 6.88393116e-01
5.88786900e-01 8.72537792e-01 -3.02269727e-01 1.40240133e+00
-4.33251470e-01 6.53152347e-01 3.00805360e-01 -1.54832080e-01
4.47050780e-02 -9.04513896e-02 4.73763078e-01 1.13187659e+00
8.90690863e-01 -4.05984849e-01 -1.44899562e-01 1.08058059e+00
-3.09499472e-01 4.28057581e-01 -3.16011399e-01 -3.08329076e-01
2.87207991e-01 1.08128345e+00 -4.09151822e-01 -4.99202371e-01
1.58964038e-01 8.58785331e-01 -8.89049396e-02 2.44437099e-01
-7.99750924e-01 -7.81942844e-01 5.63405335e-01 5.63129522e-02
-1.80954590e-01 8.76562148e-02 -4.59594548e-01 -8.19261909e-01
-3.79462272e-01 -1.17046726e+00 2.50962347e-01 -1.45136070e+00
-1.24149442e+00 6.16814017e-01 -1.87136218e-01 -1.73767698e+00
-5.40773749e-01 -4.05488104e-01 -8.58414412e-01 8.95525038e-01
-8.90436709e-01 -9.75192726e-01 -2.62519419e-01 4.56518233e-01
9.27267611e-01 -5.49281657e-01 1.19907486e+00 4.50092554e-01
-9.51303020e-02 5.19681513e-01 -6.65935352e-02 2.66132712e-01
7.99205720e-01 -1.23185086e+00 4.92089719e-01 1.55138314e-01
7.15315044e-01 4.98509765e-01 4.26241010e-01 -3.64717275e-01
-6.05941653e-01 -9.77698922e-01 9.51157987e-01 -1.03035234e-01
5.28385043e-01 7.62860477e-02 -5.74253678e-01 2.21305519e-01
6.41748190e-01 -1.06986606e+00 1.05310774e+00 1.03614971e-01
-1.71783358e-01 -1.35799035e-01 -6.04876697e-01 9.55947101e-01
1.22531474e+00 -2.51292080e-01 -7.27836668e-01 8.63957331e-02
5.63143849e-01 -3.75827044e-01 -7.13282049e-01 3.87069672e-01
7.74576843e-01 -9.97313857e-01 9.80280340e-01 -5.43372273e-01
8.04770350e-01 -4.78872716e-01 5.95338307e-02 -1.54146516e+00
-7.90860415e-01 -7.38417864e-01 -1.51870653e-01 1.50770986e+00
4.22323674e-01 2.60976791e-01 6.81002617e-01 -3.45354944e-01
-4.36402410e-01 -3.99340361e-01 -4.30848241e-01 -3.77912730e-01
7.32866153e-02 -6.65371776e-01 1.11308825e+00 8.71622264e-01
-8.83532390e-02 6.01699948e-01 -6.35019898e-01 -5.73130250e-01
4.12466735e-01 5.45155168e-01 9.38861907e-01 -1.45029485e+00
-6.02234542e-01 -1.28049016e+00 -2.94470191e-01 -7.77998149e-01
-4.92491797e-02 -1.28344190e+00 -1.63582578e-01 -1.75007725e+00
2.15549260e-01 -2.10113540e-01 -6.29384995e-01 3.58878732e-01
1.48881786e-02 8.26584697e-01 5.12312889e-01 3.66340578e-01
-2.21372575e-01 5.18598020e-01 1.66270185e+00 -3.35839033e-01
-4.15153563e-01 3.83349419e-01 -8.98997366e-01 7.99569368e-01
1.35891068e+00 -2.69967526e-01 -5.26084423e-01 -3.99024248e-01
6.77009106e-01 1.06441393e-01 -9.01662633e-02 -1.38897586e+00
-1.42483398e-01 -3.78366411e-02 7.16102242e-01 -8.01235616e-01
5.49591184e-01 -1.03069901e-01 6.07361078e-01 3.85894001e-01
-8.19620550e-01 1.87656373e-01 9.97308716e-02 5.15293591e-02
-6.43054008e-01 -5.75968325e-01 7.11883664e-01 -3.10681641e-01
-5.40966809e-01 5.43792956e-02 -3.33852798e-01 -1.56931922e-01
5.08976817e-01 -2.99141347e-01 2.23246992e-01 -5.58880806e-01
-1.02853918e+00 -4.51142460e-01 -4.50657830e-02 8.27792823e-01
6.43439472e-01 -2.00750780e+00 -1.02119470e+00 3.68549004e-02
4.53019813e-02 -2.84886450e-01 4.34735328e-01 2.48969764e-01
-5.61780393e-01 2.19483301e-01 -4.50937420e-01 -2.11202890e-01
-1.26805961e+00 2.23317653e-01 1.55796394e-01 -3.50154728e-01
-4.06684399e-01 8.99768770e-01 2.20389128e-01 -3.89672428e-01
3.32183331e-01 2.20783446e-02 -7.25775182e-01 2.95650721e-01
7.27549016e-01 5.29102743e-01 -1.16030417e-01 -6.07506156e-01
3.80897671e-01 7.11650670e-01 3.67212802e-01 -4.02070910e-01
1.15771055e+00 3.96044523e-01 -1.84451729e-01 7.66640782e-01
5.55827141e-01 -7.79129788e-02 -6.93671405e-01 1.02788545e-01
-1.84810653e-01 -9.63200033e-02 -2.71365047e-01 -1.16256714e+00
-1.22070038e+00 9.86512661e-01 3.47143978e-01 2.82470405e-01
1.22248650e+00 -1.80819169e-01 1.00202167e+00 7.68955052e-02
4.23969656e-01 -1.28676641e+00 5.30534446e-01 5.90503216e-01
1.48053575e+00 -3.79295766e-01 -3.05568010e-01 7.40787089e-02
-6.17465556e-01 1.09250867e+00 4.18716431e-01 -3.18453580e-01
2.57616550e-01 6.04698174e-02 1.41728133e-01 1.38966545e-01
-5.17943978e-01 -1.76636592e-01 5.85115671e-01 6.99173391e-01
1.16344392e+00 3.43172371e-01 -3.81780386e-01 7.66470492e-01
-1.39062035e+00 9.29963291e-02 4.82334882e-01 4.03334677e-01
-4.06541169e-01 -1.24225819e+00 -6.30201697e-01 3.89503956e-01
-6.16642416e-01 -2.01290339e-01 -8.97697389e-01 4.56293851e-01
6.21947408e-01 9.42455947e-01 -2.20825821e-02 -5.35280049e-01
3.94595265e-01 3.54829222e-01 6.33027434e-01 -7.24208176e-01
-1.01361346e+00 3.13397229e-01 -2.22098734e-03 7.81459268e-03
-6.60549402e-01 -5.13420343e-01 -1.11524630e+00 -3.17082137e-01
-1.75196320e-01 2.98991591e-01 6.54234529e-01 3.66154104e-01
1.13625646e-01 1.15569615e+00 3.73309791e-01 -1.36619329e+00
-3.57169896e-01 -1.42754126e+00 -8.77611935e-01 6.07051790e-01
-1.93560988e-01 -4.18810636e-01 -1.26808360e-01 3.97652596e-01] | [16.062641143798828, 5.538880348205566] |
3e38a948-ab0c-4ba3-850d-3c3eb99b1a91 | plan2scene-converting-floorplans-to-3d-scenes | 2106.05375 | null | https://arxiv.org/abs/2106.05375v1 | https://arxiv.org/pdf/2106.05375v1.pdf | Plan2Scene: Converting Floorplans to 3D Scenes | We address the task of converting a floorplan and a set of associated photos of a residence into a textured 3D mesh model, a task which we call Plan2Scene. Our system 1) lifts a floorplan image to a 3D mesh model; 2) synthesizes surface textures based on the input photos; and 3) infers textures for unobserved surfaces using a graph neural network architecture. To train and evaluate our system we create indoor surface texture datasets, and augment a dataset of floorplans and photos from prior work with rectified surface crops and additional annotations. Our approach handles the challenge of producing tileable textures for dominant surfaces such as floors, walls, and ceilings from a sparse set of unaligned photos that only partially cover the residence. Qualitative and quantitative evaluations show that our system produces realistic 3D interior models, outperforming baseline approaches on a suite of texture quality metrics and as measured by a holistic user study. | ['Manolis Savva', 'Angel X. Chang', 'Yasutaka Furukawa', 'Qirui Wu', 'Madhawa Vidanapathirana'] | 2021-06-09 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Vidanapathirana_Plan2Scene_Converting_Floorplans_to_3D_Scenes_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Vidanapathirana_Plan2Scene_Converting_Floorplans_to_3D_Scenes_CVPR_2021_paper.pdf | cvpr-2021-1 | ['plan2scene'] | ['computer-vision'] | [ 1.05460846e+00 6.93105519e-01 3.55670422e-01 -5.91718674e-01
-8.42821121e-01 -6.66855752e-01 4.57231760e-01 9.43402424e-02
6.38051748e-01 4.54442799e-01 4.35382456e-01 -2.23589346e-01
2.30205595e-01 -1.41434312e+00 -1.16104507e+00 -1.29496872e-01
-4.74088043e-02 6.94170296e-01 6.67350367e-02 -2.53112465e-01
-8.39892104e-02 3.52198750e-01 -1.87866271e+00 7.32650340e-01
8.32136095e-01 1.17576849e+00 4.04318929e-01 8.59680593e-01
-6.51529878e-02 3.37191105e-01 -9.04712006e-02 -5.61097413e-02
4.37120974e-01 -3.80258895e-02 -7.60360420e-01 8.62344205e-01
1.17237198e+00 -2.98331082e-01 1.66745707e-01 6.87528074e-01
7.53267109e-02 6.19514734e-02 7.68908560e-01 -8.63161385e-01
-8.37410271e-01 1.33596763e-01 -5.78681350e-01 -6.97861910e-01
6.51721716e-01 1.43377408e-01 8.75688612e-01 -7.38769352e-01
7.78015435e-01 1.47282708e+00 7.99866140e-01 1.77635979e-02
-1.72418809e+00 -1.95189863e-01 5.99922478e-01 -7.26233900e-01
-1.56159651e+00 -4.50412363e-01 5.66281974e-01 -5.27704477e-01
7.66159296e-01 5.62652290e-01 8.22258413e-01 1.11747897e+00
1.45000622e-01 4.67240095e-01 1.22001934e+00 -2.92111099e-01
4.58642989e-01 -1.47183657e-01 -3.62373978e-01 1.11904585e+00
1.09820604e-01 -2.18655914e-01 -3.23355168e-01 -8.08181763e-02
9.81853008e-01 -1.95952505e-01 -2.66357243e-01 -5.70817411e-01
-1.03628767e+00 9.13596749e-02 5.45543671e-01 -4.85002041e-01
-4.86276805e-01 1.86445013e-01 -3.91100049e-01 3.78517397e-02
9.49526846e-01 1.84380963e-01 -4.04933542e-01 3.16199243e-01
-7.22406030e-01 3.52316439e-01 8.87234092e-01 1.44647014e+00
1.06499410e+00 4.63463180e-02 1.58409439e-02 7.05798209e-01
4.00729716e-01 1.08790970e+00 -6.25958562e-01 -1.34550679e+00
6.73057735e-01 8.43221247e-01 3.95125240e-01 -9.69175279e-01
-9.86434892e-02 2.71861970e-01 -7.52432406e-01 2.80266702e-01
1.02293961e-01 8.87895003e-02 -1.52141321e+00 1.38003111e+00
4.81631041e-01 -2.97823101e-01 -3.35651219e-01 6.41828060e-01
8.84721041e-01 5.52641451e-01 -7.17649832e-02 6.14921629e-01
1.15354550e+00 -8.24741006e-01 -3.46959829e-01 -6.05065703e-01
3.63406725e-02 -6.88595712e-01 1.59623635e+00 3.22741479e-01
-1.05213416e+00 -4.64633495e-01 -9.95550036e-01 -2.45313019e-01
-3.94398183e-01 3.41431260e-01 5.90692401e-01 7.48654962e-01
-1.45338869e+00 5.44389725e-01 -9.85051274e-01 -6.16164267e-01
5.13603210e-01 2.14799061e-01 -3.56512040e-01 -3.60968947e-01
-1.91744328e-01 4.38867807e-01 -1.24205261e-01 3.02047972e-02
-9.78699386e-01 -6.73811018e-01 -1.22232878e+00 2.10726243e-02
3.03905547e-01 -9.11935151e-01 1.27101934e+00 -9.47926879e-01
-1.46562076e+00 8.40884268e-01 -2.38720447e-01 3.95666480e-01
3.36909354e-01 -1.68581754e-01 -1.88172355e-01 -2.99310744e-01
4.57687944e-01 7.24525154e-01 6.68360174e-01 -1.95578766e+00
-5.35762310e-01 -3.30902338e-01 1.42212540e-01 4.85469848e-01
3.03962022e-01 -9.16743636e-01 -6.12149417e-01 -3.99230868e-01
5.84920526e-01 -9.23793197e-01 -2.95675993e-01 3.12841475e-01
-9.15763736e-01 5.85991681e-01 6.29968882e-01 -8.96879494e-01
5.83793700e-01 -2.13282394e+00 -5.57416305e-02 7.48289108e-01
-5.80187403e-02 -7.55819142e-01 -2.76653498e-01 4.40257341e-01
1.27787888e-01 3.21690410e-01 -4.29351181e-01 -8.80903780e-01
2.14068875e-01 3.36788088e-01 -4.96727705e-01 1.17441386e-01
1.54041484e-01 5.04358888e-01 -7.80590177e-01 -2.28355810e-01
4.71675307e-01 6.51356101e-01 -7.48883665e-01 1.87955782e-01
-7.21999466e-01 4.62426692e-01 -2.72158980e-01 9.79909718e-01
8.34947646e-01 -5.40543795e-01 6.39011264e-01 -1.38291731e-01
-3.31488252e-02 3.71158570e-01 -1.41125166e+00 1.94077373e+00
-7.89586365e-01 3.88093501e-01 2.30410755e-01 7.23777637e-02
9.52339351e-01 4.92389947e-02 1.23074315e-01 -5.59284747e-01
-2.90751070e-01 1.08623169e-01 -9.87875640e-01 -3.09490591e-01
8.85529280e-01 3.13638866e-01 -1.99725881e-01 3.95710856e-01
-3.99651319e-01 -8.50655198e-01 -3.63545388e-01 -7.65593722e-02
1.19402742e+00 7.53701508e-01 -9.22442600e-02 -4.58633035e-01
-9.48350132e-02 1.40356228e-01 1.58305958e-01 6.97204292e-01
6.11001313e-01 1.16887069e+00 1.65121332e-01 -8.29104245e-01
-1.38422930e+00 -1.61358881e+00 2.29053244e-01 9.63722587e-01
2.83653527e-01 -3.36992025e-01 -9.47983265e-01 -4.39281672e-01
1.94366753e-01 5.47318876e-01 -9.44407701e-01 5.60726166e-01
-3.10806423e-01 -4.07651186e-01 2.43318863e-02 3.91531885e-01
6.75662637e-01 -7.70316422e-01 -2.67602801e-01 -6.12260737e-02
-4.43484068e-01 -1.12373006e+00 -5.19346535e-01 5.77687025e-02
-5.75173378e-01 -1.19012463e+00 -2.20196068e-01 -8.94461572e-01
1.26871204e+00 3.50476593e-01 1.62752485e+00 2.89955050e-01
-1.57543257e-01 5.35280704e-01 -1.23689160e-01 -3.42760205e-01
-3.71137828e-01 1.18463978e-01 -2.11324230e-01 1.67692900e-01
-2.53591895e-01 -7.50906527e-01 -4.81569558e-01 3.28119934e-01
-8.40595961e-01 8.62396479e-01 1.69396788e-01 1.55918479e-01
1.16826940e+00 3.14474143e-02 -2.82172412e-01 -9.83228445e-01
1.48254022e-01 -2.57638633e-01 -8.59129727e-01 3.20866674e-01
-2.62978196e-01 1.14748469e-02 3.57307673e-01 2.26520002e-01
-1.46140933e+00 4.53687787e-01 1.67515531e-01 5.25599867e-02
-5.05603075e-01 5.97196035e-02 -4.91070807e-01 2.02522296e-02
7.58279145e-01 -1.39450863e-01 -4.67685610e-01 -5.46491921e-01
3.22451472e-01 4.97702301e-01 5.43458879e-01 -9.51328278e-01
9.08678770e-01 9.73457277e-01 -5.04179709e-02 -1.08059609e+00
-1.01783395e+00 1.76011220e-01 -6.52471185e-01 -2.44157463e-01
9.07230437e-01 -1.07803595e+00 -5.61178207e-01 2.76227802e-01
-9.72113192e-01 -1.29736972e+00 -2.75964856e-01 -4.93482687e-02
-7.00531006e-01 -1.62347838e-01 -3.70444685e-01 -8.64145517e-01
3.04976553e-02 -9.80653882e-01 2.02066851e+00 -6.91286474e-02
-3.90377432e-01 -7.63081789e-01 -2.12588981e-02 4.05579001e-01
3.40645909e-01 7.47724175e-01 1.12285507e+00 6.67614579e-01
-1.31040907e+00 2.82978684e-01 -3.36503059e-01 -1.02652721e-01
4.31648463e-01 1.69490874e-01 -1.35635507e+00 -2.76500527e-02
-8.50424647e-01 -2.09212035e-01 5.76798201e-01 5.01143634e-01
1.23981094e+00 -3.46742690e-01 -3.52053523e-01 9.17239785e-01
1.47385359e+00 -1.68474406e-01 6.36905968e-01 2.53972948e-01
8.78371298e-01 7.47130871e-01 1.83881044e-01 2.33253241e-01
1.02014828e+00 4.52010393e-01 6.69549763e-01 -4.12390709e-01
-3.31968129e-01 -7.58244991e-01 2.29193345e-02 3.39193881e-01
-2.72755742e-01 -6.08776212e-01 -9.56130087e-01 6.20316863e-01
-1.56541920e+00 -4.40617234e-01 -2.44278029e-01 2.12277985e+00
3.56087863e-01 -5.61321322e-05 -4.04405951e-01 -2.44756594e-01
4.43740278e-01 3.05881798e-01 -3.73102605e-01 -4.31619167e-01
-2.02456966e-01 3.16360772e-01 7.07635581e-01 9.46885526e-01
-1.22434449e+00 1.00395572e+00 6.81930351e+00 2.29667068e-01
-6.74250782e-01 -2.72019535e-01 1.01959348e+00 7.69654140e-02
-9.13774312e-01 -2.06521899e-02 -3.88089597e-01 -3.97338904e-02
4.57764149e-01 6.48099899e-01 1.01436949e+00 9.37339365e-01
2.95048743e-01 -3.34081620e-01 -1.11294687e+00 6.95779145e-01
-9.83123854e-02 -1.53922224e+00 6.44934326e-02 3.25015634e-01
1.36593258e+00 7.40213469e-02 1.00518785e-01 6.58619925e-02
1.17320085e+00 -1.32440686e+00 1.04929662e+00 6.37453079e-01
1.42223918e+00 -1.65462121e-01 2.01481119e-01 3.58184651e-02
-1.57982004e+00 4.55489784e-01 1.01521328e-01 -2.28697211e-01
9.66008529e-02 4.89559442e-01 -9.60780621e-01 5.86919069e-01
1.04349613e+00 5.04944623e-01 -5.55347204e-01 6.03948236e-01
-3.62869710e-01 3.51373255e-01 -6.48339152e-01 4.29780543e-01
-1.41385838e-01 -2.49833852e-01 -7.04620704e-02 8.08752716e-01
5.24490893e-01 6.21281751e-02 5.15114129e-01 1.36200333e+00
-2.09479570e-01 -2.09675491e-01 -9.56702888e-01 3.03570986e-01
6.76299334e-01 1.15639436e+00 -1.00926661e+00 -3.26021612e-01
-1.03129230e-01 1.13548207e+00 2.48260543e-01 6.22076094e-01
-4.18582588e-01 -1.74542740e-02 6.37916923e-01 5.44205129e-01
2.62067556e-01 -2.49038577e-01 -7.98278332e-01 -9.50042307e-01
2.55039781e-01 -6.68995142e-01 -2.77410209e-01 -1.38979578e+00
-1.09714413e+00 4.98385489e-01 -1.66724995e-01 -9.23958361e-01
2.38994822e-01 -2.99486935e-01 -3.70650440e-01 9.40011322e-01
-1.27583265e+00 -1.70561886e+00 -1.10455036e+00 1.15452096e-01
4.80189025e-01 5.36249399e-01 1.23951840e+00 2.67819129e-02
-1.81593955e-01 -6.78233281e-02 -1.58813149e-01 -1.59401700e-01
2.46011928e-01 -1.45301056e+00 1.30933988e+00 6.56276345e-01
-1.17358468e-01 8.37283731e-02 5.29458582e-01 -1.04603505e+00
-1.30736125e+00 -1.66376555e+00 7.22543359e-01 -7.67122865e-01
-9.21391770e-02 -8.62836123e-01 -5.22101283e-01 1.05734217e+00
-1.53284641e-02 -1.94497973e-01 3.07115525e-01 1.39755264e-01
-2.75806338e-01 8.05248767e-02 -1.15386248e+00 9.02955830e-01
1.58613741e+00 -4.66608912e-01 -6.24637604e-02 4.19054180e-01
5.90002000e-01 -1.02244163e+00 -8.94027770e-01 3.34469497e-01
6.69582725e-01 -9.97734904e-01 9.18217301e-01 4.05478887e-02
5.92245638e-01 -4.85326022e-01 -7.94029772e-01 -1.49395406e+00
-3.37855667e-01 -4.76339877e-01 3.96950096e-01 7.13143289e-01
5.31238496e-01 -1.39999449e-01 1.08097827e+00 8.43007624e-01
-4.16735590e-01 -4.17254120e-01 -4.47150439e-01 -4.12390441e-01
-5.11079788e-01 -4.97535199e-01 1.22317290e+00 6.89031124e-01
-6.89475119e-01 1.80917121e-02 -1.83274001e-01 7.20910609e-01
7.42758095e-01 4.98798430e-01 1.22627354e+00 -1.25092959e+00
-1.35804519e-01 -3.37668099e-02 9.98669788e-02 -1.02195072e+00
-1.79761633e-01 -7.26235032e-01 4.17535424e-01 -2.16738915e+00
-5.66991903e-02 -7.83303559e-01 6.34907365e-01 7.44070172e-01
1.19017370e-01 5.34489274e-01 -2.71101326e-01 -1.74539164e-01
-4.01508480e-01 5.24088740e-01 1.29854536e+00 -3.16347063e-01
-3.91361773e-01 -3.34084183e-01 -6.93902850e-01 9.19067442e-01
5.30691504e-01 -6.73517883e-02 -4.94642496e-01 -1.00208056e+00
4.97211397e-01 -2.06898749e-02 5.53289711e-01 -9.52426016e-01
-5.08681476e-01 -5.00317752e-01 8.48475814e-01 -5.24954438e-01
7.31930614e-01 -8.67257297e-01 7.26277590e-01 -1.37219906e-01
-1.75789163e-01 5.80595545e-02 2.94811934e-01 6.27691805e-01
4.32816267e-01 7.15884447e-01 2.61492401e-01 -3.76897544e-01
-5.57033300e-01 4.01863307e-01 -3.41184467e-01 -1.03673086e-01
4.78442609e-01 -4.59795296e-01 -3.34022969e-01 -4.91777718e-01
-7.68790305e-01 8.40237066e-02 1.31722009e+00 5.63800871e-01
5.46409309e-01 -1.50564206e+00 -3.85292321e-01 7.08445549e-01
1.60737276e-01 7.27815151e-01 2.94781446e-01 6.11460134e-02
-9.88597512e-01 -4.59215865e-02 -4.29355837e-02 -6.95446968e-01
-1.00080371e+00 -1.50813445e-01 4.62538362e-01 -7.28122145e-02
-9.54377174e-01 6.26832902e-01 6.87160373e-01 -1.15603638e+00
8.17412362e-02 -1.11248946e+00 3.64254773e-01 -6.53825879e-01
1.88584626e-02 3.03694271e-02 3.85265619e-01 -5.43421626e-01
-1.81795105e-01 6.68291807e-01 6.88176513e-01 -2.65483975e-01
1.53411996e+00 -2.75347620e-01 4.47391160e-02 3.89094144e-01
7.05967069e-01 2.55867809e-01 -1.71380901e+00 -4.81618866e-02
-3.13887686e-01 -6.95517898e-01 -5.69765642e-02 -9.82333481e-01
-6.78873479e-01 4.26947057e-01 4.65936065e-01 2.81778667e-02
9.20479715e-01 -2.37619206e-01 5.53723097e-01 4.20434594e-01
8.87553811e-01 -1.13848841e+00 -4.37571630e-02 4.77708876e-01
9.64095414e-01 -1.19136655e+00 -1.86633575e-03 -8.68400931e-01
-4.52290654e-01 8.36358428e-01 5.58413386e-01 -2.21869633e-01
6.40616119e-01 5.32849252e-01 8.31959024e-02 -5.39384902e-01
-6.39778912e-01 -1.75646663e-01 3.84760261e-01 8.12378228e-01
1.90237552e-01 5.55054247e-01 9.21176314e-01 2.13585526e-01
-7.32756793e-01 -4.33374575e-04 4.64151680e-01 1.04436851e+00
-4.15240705e-01 -8.53783309e-01 -6.81459248e-01 5.60757756e-01
3.20930690e-01 -1.35466933e-01 -4.50531602e-01 7.71745265e-01
2.99753189e-01 8.75326931e-01 2.50606209e-01 -5.72592556e-01
6.18528128e-01 -1.93603918e-01 5.13867974e-01 -1.04828215e+00
-3.01939756e-01 4.81064618e-02 5.80610931e-01 -7.91492701e-01
-1.87860355e-01 -4.97841954e-01 -8.08136284e-01 -3.78590792e-01
1.17237963e-01 -3.92400831e-01 8.03176582e-01 6.77572608e-01
6.15964651e-01 7.95191765e-01 4.20990258e-01 -1.46305001e+00
3.50312889e-01 -6.90225065e-01 -7.33528078e-01 4.89632249e-01
3.37075770e-01 -6.13812208e-01 3.62159908e-02 4.88881767e-01] | [9.265108108520508, -3.0721046924591064] |
b4d93157-d194-43fa-9576-1a3ff3f37608 | guard-graph-universal-adversarial-defense | 2204.09803 | null | https://arxiv.org/abs/2204.09803v3 | https://arxiv.org/pdf/2204.09803v3.pdf | GUARD: Graph Universal Adversarial Defense | Graph convolutional networks (GCNs) have shown to be vulnerable to small adversarial perturbations, which becomes a severe threat and largely limits their applications in security-critical scenarios. To mitigate such a threat, considerable research efforts have been devoted to increasing the robustness of GCNs against adversarial attacks. However, current approaches for defense are typically designed for the whole graph and consider the global performance, posing challenges in protecting important local nodes from stronger adversarial targeted attacks. In this work, we present a simple yet effective method, named Graph Universal Adversarial Defense (GUARD). Unlike previous works, GUARD protects each individual node from attacks with a universal defensive patch, which is generated once and can be applied to any node (node-agnostic) in a graph. Extensive experiments on four benchmark datasets demonstrate that our method significantly improves robustness for several established GCNs against multiple adversarial attacks and outperforms state-of-the-art defense methods by large margins. Our code is publicly available at https://github.com/EdisonLeeeee/GUARD. | ['Jiawang Dan', 'Weiqiang Wang', 'Zibin Zheng', 'Changhua Meng', 'Liang Chen', 'Ruofan Wu', 'Jie Liao', 'Jintang Li'] | 2022-04-20 | null | null | null | null | ['adversarial-defense'] | ['adversarial'] | [ 1.83076859e-01 3.31813693e-01 -2.01200560e-01 8.33968073e-02
-5.70006073e-01 -1.32362092e+00 6.10902071e-01 8.46394077e-02
-1.72162743e-03 4.98279363e-01 -7.91374967e-03 -8.03358555e-01
5.45148551e-02 -1.09848630e+00 -7.13725388e-01 -8.10862362e-01
-4.51658458e-01 -1.62910387e-01 5.98100603e-01 -6.43938363e-01
-1.27979256e-02 9.24317122e-01 -4.92588282e-01 -2.10637197e-01
6.50915921e-01 5.35924673e-01 -7.31426775e-01 8.65260959e-01
5.15461326e-01 5.87872803e-01 -8.67988706e-01 -8.16316664e-01
7.77443349e-01 -3.62891436e-01 -7.10620940e-01 -3.92630428e-01
6.02891207e-01 -3.70806307e-01 -1.08352792e+00 1.45089924e+00
6.74033344e-01 -7.94802085e-02 1.75163448e-01 -1.64634657e+00
-7.26803243e-01 9.68248427e-01 -5.60344636e-01 2.70611495e-01
5.60867377e-02 3.48580956e-01 8.66457880e-01 -1.48125887e-01
3.33189160e-01 1.29655516e+00 6.10815525e-01 9.80905294e-01
-1.05847824e+00 -1.01365244e+00 5.46805382e-01 -1.75756335e-01
-1.18107927e+00 -2.82013685e-01 8.86929452e-01 1.10585459e-01
6.15279615e-01 5.40825605e-01 4.28336374e-02 1.36838257e+00
4.01762784e-01 3.28932732e-01 6.88498914e-01 -1.71114430e-02
1.18515998e-01 -2.99159020e-01 -1.36681646e-01 5.15229702e-01
5.44347465e-01 5.05396128e-01 2.09834635e-01 -5.90851367e-01
5.69554627e-01 -9.79243685e-03 -5.25400877e-01 -1.58992007e-01
-5.57260513e-01 9.66297567e-01 1.12560332e+00 1.53354570e-01
-5.72532266e-02 5.14847100e-01 6.09209478e-01 4.37765300e-01
3.53783935e-01 2.97497511e-01 -3.31637323e-01 4.99743134e-01
-2.30706617e-01 1.77693233e-01 1.03862250e+00 6.40021563e-01
4.22511548e-01 3.85158122e-01 6.91815913e-02 3.53880286e-01
8.26842487e-02 4.39947128e-01 -2.24728525e-01 -6.22916818e-01
3.73371452e-01 4.95261341e-01 -5.90892375e-01 -1.50275958e+00
-2.43387297e-01 -6.72564447e-01 -1.16009414e+00 4.18964237e-01
1.73062712e-01 -5.14524341e-01 -1.02588296e+00 2.02994490e+00
5.90274930e-01 4.52386618e-01 1.20520271e-01 5.93139112e-01
7.56287277e-01 4.35494542e-01 1.86695099e-01 2.12328881e-01
8.19894493e-01 -7.65262842e-01 -1.58594310e-01 -5.69804966e-01
4.17133719e-01 -3.72605145e-01 6.55246317e-01 1.29367888e-01
-6.98380172e-01 -1.23910056e-02 -1.16098309e+00 5.30974329e-01
-6.09029531e-01 -7.93837786e-01 6.40356481e-01 1.28373241e+00
-1.18367374e+00 6.99597895e-01 -7.23058164e-01 -2.84716368e-01
5.54583371e-01 5.83792150e-01 -4.88505781e-01 -1.13389835e-01
-1.56081247e+00 5.64415872e-01 3.42912138e-01 2.26283204e-02
-1.54166889e+00 -6.30548000e-01 -8.18441808e-01 1.06507003e-01
7.21873820e-01 -2.67859399e-01 8.69611979e-01 -6.41261220e-01
-1.03258264e+00 5.40784001e-01 6.67256057e-01 -7.20906854e-01
5.18908441e-01 5.14365127e-03 -8.07657003e-01 3.18091571e-01
-1.83191165e-01 2.20580027e-01 1.00187600e+00 -1.38298583e+00
-1.70253608e-02 -2.85989046e-01 8.96766961e-01 -1.71956137e-01
-8.00920129e-01 3.95473480e-01 -3.92621070e-01 -9.21749532e-01
-2.10234478e-01 -1.22785437e+00 -6.67810559e-01 -2.66269930e-02
-8.80264282e-01 1.38817728e-01 1.14062893e+00 -4.27390337e-01
1.21647108e+00 -2.12025905e+00 7.46075585e-02 4.26313609e-01
6.40339315e-01 7.96078146e-01 -3.30594301e-01 7.35795617e-01
-2.32736841e-01 6.95402920e-01 -2.20615968e-01 1.61262646e-01
-4.49743792e-02 2.37600029e-01 -6.29310489e-01 8.33108485e-01
8.33387077e-02 8.80533874e-01 -9.04972792e-01 2.24149995e-03
-2.77525075e-02 5.91702938e-01 -5.33737957e-01 1.08277358e-01
-6.73974752e-02 3.03902298e-01 -6.56127632e-01 9.10659075e-01
1.04560220e+00 -1.12988420e-01 3.43779564e-01 1.27396569e-01
5.71270287e-01 -1.10052377e-01 -8.44833493e-01 8.78986239e-01
2.48988587e-02 2.86489636e-01 4.20913190e-01 -7.59885430e-01
7.41754770e-01 1.50320724e-01 1.60566643e-01 -2.19251990e-01
3.89865756e-01 2.39304341e-02 2.16191754e-01 3.01914155e-01
4.71215546e-02 1.97232708e-01 -4.69613105e-01 4.21109885e-01
-1.43434837e-01 1.05591737e-01 1.38296619e-01 8.64508152e-01
1.69538176e+00 -5.65167606e-01 1.21751502e-01 -3.25214446e-01
7.44939923e-01 -4.37800735e-01 7.11764276e-01 9.77017164e-01
-5.78268170e-01 3.05726618e-01 8.73254716e-01 -4.38927442e-01
-5.63287199e-01 -1.26653361e+00 3.21721315e-01 1.04705060e+00
3.75816286e-01 -6.35421097e-01 -9.21015680e-01 -1.42018151e+00
1.84040233e-01 3.61146241e-01 -8.73473048e-01 -7.67946482e-01
-5.80385447e-01 -7.16170847e-01 1.16690230e+00 5.26519775e-01
6.18080676e-01 -8.15664411e-01 1.47644266e-01 -2.75488216e-02
2.68825024e-01 -1.19400799e+00 -8.02726090e-01 -1.08653337e-01
-4.46236491e-01 -1.39903736e+00 -2.88927466e-01 -3.39321166e-01
8.62629235e-01 5.59172690e-01 9.50464904e-01 7.45932281e-01
-4.36153620e-01 3.23917896e-01 -4.89547670e-01 -4.24866110e-01
-7.74466038e-01 2.02960327e-01 2.70398319e-01 -1.33755282e-01
-2.08695039e-01 -7.79366970e-01 -5.35534739e-01 4.93221551e-01
-1.24928570e+00 -8.09221327e-01 3.61956745e-01 5.04402459e-01
6.66742697e-02 4.22239989e-01 6.24570072e-01 -1.17669761e+00
8.93369973e-01 -6.72780156e-01 -6.86085522e-01 1.24873154e-01
-4.42448437e-01 -3.15543711e-01 1.14121079e+00 -5.79983652e-01
-4.65580195e-01 -3.24778974e-01 -2.76079327e-01 -4.42588389e-01
-5.55933677e-02 2.38487840e-01 -4.98067260e-01 -1.06788850e+00
1.06910551e+00 -6.77175075e-02 -1.77929685e-01 -3.59372832e-02
3.42892647e-01 4.25365288e-03 5.29369473e-01 -5.45428336e-01
1.78490555e+00 3.33270282e-01 3.13310623e-01 -5.22425592e-01
-7.21770108e-01 1.07761189e-01 -1.61817521e-01 -3.17221195e-01
3.81767780e-01 -6.26309335e-01 -6.88823164e-01 7.59917498e-01
-7.29675233e-01 -3.85224372e-01 2.69090116e-01 -2.99575359e-01
1.41902104e-01 6.89096510e-01 -7.95226634e-01 -4.88791794e-01
-6.00895166e-01 -1.01202369e+00 4.07699734e-01 3.60604405e-01
3.24231088e-01 -1.24004352e+00 1.77117251e-02 1.86595783e-01
7.35077441e-01 1.02004993e+00 5.41193664e-01 -9.20636117e-01
-6.42197192e-01 -7.94558346e-01 -1.71442732e-01 5.81149042e-01
9.92954671e-02 1.65105224e-01 -7.50697374e-01 -9.91087615e-01
-2.79592067e-01 -4.70067590e-01 7.19547927e-01 -1.04617395e-01
1.24990237e+00 -7.39526391e-01 -3.71629089e-01 1.01463079e+00
1.56584156e+00 -1.02481231e-01 7.37762213e-01 2.61488676e-01
9.21738923e-01 1.59873679e-01 2.10721269e-01 1.89485729e-01
-7.02334493e-02 1.72691479e-01 1.33851862e+00 -2.80401200e-01
1.21748224e-01 -2.92351305e-01 4.95717287e-01 -2.06582434e-02
1.78025626e-02 -8.78130078e-01 -9.40539598e-01 3.63780409e-01
-1.58487236e+00 -8.44292939e-01 1.12033553e-01 1.86919999e+00
4.49511737e-01 5.19653082e-01 3.07243317e-02 3.42651308e-02
8.13353240e-01 7.12865949e-01 -7.42972493e-01 -4.28208739e-01
-2.22107798e-01 1.80644214e-01 9.18226540e-01 4.32520121e-01
-1.27742088e+00 1.24277008e+00 6.47922087e+00 8.34491491e-01
-9.74261940e-01 -5.20510711e-02 5.50588191e-01 9.94589403e-02
-1.04857937e-01 2.12386519e-01 -6.51321769e-01 2.40546390e-01
9.74834800e-01 -5.18743575e-01 5.13098300e-01 9.11834836e-01
-3.25432628e-01 5.83546221e-01 -4.81492430e-01 6.18358999e-02
-2.47780997e-02 -1.30908990e+00 1.18184604e-01 4.42026645e-01
7.56200552e-01 1.50438055e-01 2.59449720e-01 2.81523943e-01
1.02491283e+00 -1.08369827e+00 1.13986902e-01 -2.69169211e-01
6.96913958e-01 -1.27014208e+00 5.41703701e-01 1.90761447e-01
-1.24941683e+00 -1.59153149e-01 -4.35167819e-01 3.22799742e-01
-1.68190017e-01 3.07584316e-01 -5.41952968e-01 7.00392008e-01
6.95812225e-01 3.41596186e-01 -7.95625389e-01 5.97069204e-01
-6.32310569e-01 8.86434436e-01 -2.17877313e-01 2.65560687e-01
5.11061072e-01 2.97560304e-01 8.98139954e-01 9.40217197e-01
-9.01797339e-02 2.80892365e-02 4.60111290e-01 5.36440730e-01
-4.76039857e-01 -2.05559626e-01 -8.42641473e-01 -2.50119328e-01
6.04560316e-01 1.46035111e+00 -6.83995724e-01 1.37434095e-01
-1.76286206e-01 9.53248322e-01 3.52973312e-01 2.92113751e-01
-1.15689301e+00 -6.99151456e-01 1.06974959e+00 -3.81119959e-02
9.15745869e-02 -1.79434657e-01 2.02064380e-01 -9.96185839e-01
-1.04933888e-01 -1.31041932e+00 7.69882500e-01 -5.41232973e-02
-1.42847133e+00 8.85106683e-01 -1.54615089e-01 -1.12076950e+00
1.64687276e-01 -5.79687178e-01 -1.22103548e+00 7.17241526e-01
-1.11144030e+00 -1.44196951e+00 -2.93301404e-01 9.92067397e-01
-1.15236916e-01 -3.45737964e-01 7.68596768e-01 1.62440076e-01
-8.49673986e-01 1.26968062e+00 -2.24456489e-01 6.30604804e-01
6.05427206e-01 -1.03588116e+00 1.03242075e+00 1.74829912e+00
-1.61637664e-01 7.82506943e-01 7.38112926e-01 -8.80160570e-01
-1.48842990e+00 -1.33387685e+00 1.39828905e-01 -3.12926143e-01
1.15832460e+00 -6.54134989e-01 -9.93658721e-01 8.02836180e-01
2.22801968e-01 5.53596616e-01 6.65840447e-01 -2.24890605e-01
-9.72102880e-01 -4.75855451e-03 -1.56420195e+00 9.58780289e-01
1.22576201e+00 -5.10878861e-01 2.01133028e-01 5.31905293e-01
1.18854547e+00 -5.16749501e-01 -7.56773114e-01 5.90608299e-01
1.01148762e-01 -9.50117528e-01 1.16155410e+00 -8.50400031e-01
3.78385112e-02 -3.60045105e-01 4.03119773e-02 -1.38313496e+00
-4.83196974e-01 -1.11739933e+00 -3.30168456e-01 1.15428782e+00
3.23899031e-01 -1.14919448e+00 8.03358316e-01 4.46660489e-01
-7.22982295e-05 -6.32269740e-01 -7.40438402e-01 -9.93417501e-01
3.90277714e-01 -8.42088386e-02 7.77556896e-01 1.04153585e+00
-2.82986969e-01 -1.86649695e-01 -5.23387909e-01 9.74948347e-01
9.30771708e-01 -4.09247339e-01 9.74785268e-01 -9.36837852e-01
-3.67540777e-01 -4.39573020e-01 -7.56149173e-01 -4.03082103e-01
3.91288161e-01 -8.78622890e-01 -2.50433296e-01 -1.06502581e+00
-1.59764260e-01 -1.97533533e-01 -4.27813202e-01 8.22079837e-01
-4.38355803e-01 5.69261193e-01 2.91741520e-01 -2.16105387e-01
-4.01455551e-01 1.47671148e-01 9.08843756e-01 -2.90705055e-01
2.20546752e-01 2.14397639e-01 -1.40547478e+00 6.04009151e-01
1.30132818e+00 -5.53702772e-01 -5.62429905e-01 -2.07414746e-01
2.50414927e-02 -1.70296162e-01 5.48659980e-01 -1.02895713e+00
1.88590348e-01 -3.24771047e-01 5.83955012e-02 -4.29292247e-02
-7.11288378e-02 -6.89608395e-01 1.48119003e-01 7.80533373e-01
-1.23757809e-01 1.18014038e-01 4.31604534e-01 8.62729549e-01
6.88219517e-02 1.63115084e-01 1.11014497e+00 -5.85569721e-03
-4.57908690e-01 1.05783033e+00 5.42544350e-02 2.51158953e-01
1.30046308e+00 1.94256887e-01 -1.00481999e+00 -6.26778364e-01
-5.27994871e-01 4.46031511e-01 6.40626669e-01 5.67787528e-01
6.42231047e-01 -1.26020813e+00 -9.25245285e-01 1.29632890e-01
1.30200803e-01 -3.80196601e-01 4.36222792e-01 2.97156960e-01
-5.74787617e-01 -6.98850974e-02 -2.46837437e-01 6.53559640e-02
-1.48016560e+00 1.00475597e+00 6.61127329e-01 -3.89107287e-01
-5.87122083e-01 1.20462024e+00 2.82757282e-01 -4.25770938e-01
3.10536265e-01 5.44483304e-01 2.27673855e-02 -6.58438921e-01
6.05053961e-01 2.20064580e-01 -1.65752724e-01 -7.88447976e-01
-5.77356458e-01 2.02248409e-01 -4.51915115e-01 4.69922125e-01
1.08897388e+00 3.30093950e-01 -3.27499956e-01 -6.64379537e-01
1.15495062e+00 3.80201161e-01 -1.14543831e+00 -1.61089629e-01
-3.98933172e-01 -7.11648822e-01 -8.84958580e-02 -6.96956933e-01
-1.70059967e+00 4.78169203e-01 1.53028756e-01 7.84239054e-01
1.37921214e+00 -9.70335528e-02 8.32463920e-01 2.24153548e-01
4.67628211e-01 -4.50867683e-01 1.07158214e-01 5.16601443e-01
8.69158506e-01 -9.61957514e-01 1.27183180e-02 -8.13280225e-01
-3.43256623e-01 8.76060545e-01 1.01906669e+00 -6.57047629e-01
8.53036642e-01 4.11944002e-01 7.75470883e-02 -5.35164997e-02
-6.01667047e-01 2.29321688e-01 2.27818385e-01 7.94935942e-01
-1.35029778e-02 1.91930071e-01 -1.90023016e-02 2.67531335e-01
2.98552290e-02 -9.24566329e-01 7.56103933e-01 1.01622081e+00
-2.80459225e-01 -1.54594338e+00 -3.56834620e-01 1.04026422e-01
-8.50166619e-01 -1.26536384e-01 -1.02763486e+00 7.91246891e-01
-2.59111971e-01 1.03989744e+00 -7.34920800e-01 -8.86063039e-01
2.05145627e-01 -6.27922714e-01 3.98151800e-02 -4.84843940e-01
-9.69701886e-01 -3.19025278e-01 5.73714823e-02 -8.60504806e-01
2.19057113e-01 -1.86333433e-01 -1.02191818e+00 -9.12731588e-01
-2.70171374e-01 9.01201069e-02 2.88866937e-01 3.23582113e-01
3.79643381e-01 5.99966705e-01 1.05311191e+00 -6.65058911e-01
-8.38779330e-01 -4.74862635e-01 -5.07578194e-01 2.88366437e-01
4.15883601e-01 -3.74822885e-01 -9.72888947e-01 -7.20182180e-01] | [6.108344078063965, 7.296267986297607] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.