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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
020f9d3c-907f-4f99-9102-e348f9e2360d | divide-and-contrast-source-free-domain | 2211.06612 | null | https://arxiv.org/abs/2211.06612v1 | https://arxiv.org/pdf/2211.06612v1.pdf | Divide and Contrast: Source-free Domain Adaptation via Adaptive Contrastive Learning | We investigate a practical domain adaptation task, called source-free domain adaptation (SFUDA), where the source-pretrained model is adapted to the target domain without access to the source data. Existing techniques mainly leverage self-supervised pseudo labeling to achieve class-wise global alignment [1] or rely on local structure extraction that encourages feature consistency among neighborhoods [2]. While impressive progress has been made, both lines of methods have their own drawbacks - the "global" approach is sensitive to noisy labels while the "local" counterpart suffers from source bias. In this paper, we present Divide and Contrast (DaC), a new paradigm for SFUDA that strives to connect the good ends of both worlds while bypassing their limitations. Based on the prediction confidence of the source model, DaC divides the target data into source-like and target-specific samples, where either group of samples is treated with tailored goals under an adaptive contrastive learning framework. Specifically, the source-like samples are utilized for learning global class clustering thanks to their relatively clean labels. The more noisy target-specific data are harnessed at the instance level for learning the intrinsic local structures. We further align the source-like domain with the target-specific samples using a memory bank-based Maximum Mean Discrepancy (MMD) loss to reduce the distribution mismatch. Extensive experiments on VisDA, Office-Home, and the more challenging DomainNet have verified the superior performance of DaC over current state-of-the-art approaches. The code is available at https://github.com/ZyeZhang/DaC.git. | ['Guanbin Li', 'Liang Lin', 'Siyuan Li', 'Zhen Li', 'Hui Cheng', 'Weikai Chen', 'Ziyi Zhang'] | 2022-11-12 | null | null | null | null | ['source-free-domain-adaptation'] | ['computer-vision'] | [ 8.77503157e-02 -1.59256980e-02 -4.68394697e-01 -6.41183317e-01
-1.01800001e+00 -4.65415359e-01 5.76659679e-01 1.16548622e-02
-2.20987394e-01 7.58914411e-01 5.96695952e-02 8.65425244e-02
-9.28851366e-02 -6.35865867e-01 -5.60282946e-01 -9.09120798e-01
2.88474977e-01 6.90953255e-01 3.05911571e-01 -2.72805482e-01
4.97908406e-02 2.31435180e-01 -1.37262893e+00 4.75605369e-01
1.04668736e+00 1.15904129e+00 2.94983298e-01 9.46697444e-02
-2.10217461e-01 7.05193818e-01 -5.21262825e-01 -1.60072163e-01
4.45967585e-01 -5.29791892e-01 -8.07870150e-01 5.77039309e-02
6.17980659e-01 2.26347260e-02 -1.26593098e-01 1.16245127e+00
6.44692183e-01 1.15329251e-01 6.71098173e-01 -1.25041747e+00
-7.88475573e-01 4.22714472e-01 -5.75903058e-01 1.91815704e-01
-8.53619576e-02 8.31705481e-02 1.05331743e+00 -1.05428123e+00
7.47966230e-01 1.03296220e+00 6.33373678e-01 6.72043025e-01
-1.51058853e+00 -8.52754176e-01 3.98215085e-01 3.61316681e-01
-1.46338391e+00 -5.56006074e-01 1.14755535e+00 -4.79785651e-01
7.75663555e-01 -1.32499248e-01 1.81750417e-01 1.41784596e+00
-2.11369589e-01 7.85862446e-01 1.15537214e+00 -4.66756433e-01
5.06556153e-01 4.13102746e-01 3.43977273e-01 3.56366664e-01
4.84571084e-02 1.23059422e-01 -5.84725916e-01 -1.61001220e-01
4.97007877e-01 -1.09313533e-01 -2.62173772e-01 -9.16592062e-01
-1.04378819e+00 6.69631064e-01 6.40047491e-01 4.65638280e-01
-2.23218530e-01 -4.30084169e-01 3.53606224e-01 3.43697518e-01
5.62576830e-01 2.95561969e-01 -6.78801775e-01 1.63561866e-01
-9.28840339e-01 8.59960988e-02 5.93552172e-01 1.09055054e+00
1.06765926e+00 -3.79101224e-02 -7.48347566e-02 1.24273813e+00
2.26330817e-01 3.55605185e-01 7.53736138e-01 -7.09982216e-01
6.35382354e-01 7.11530447e-01 -6.64496720e-02 -8.92129242e-01
-3.45931768e-01 -7.29453504e-01 -8.28608334e-01 2.94492364e-01
7.20826566e-01 7.44308010e-02 -9.49514866e-01 2.10912514e+00
5.05325854e-01 3.49075735e-01 6.62073791e-02 8.49918306e-01
6.81858301e-01 4.63002384e-01 1.95987821e-02 5.31428866e-02
9.63214219e-01 -1.21789718e+00 -3.83900285e-01 -5.45339465e-01
6.30974948e-01 -5.66980779e-01 1.31388354e+00 3.05460095e-01
-6.80761874e-01 -7.53305435e-01 -1.16259825e+00 1.02356402e-02
-5.50181150e-01 1.60747781e-01 1.95142906e-02 5.44941545e-01
-7.79731870e-01 5.30626237e-01 -7.66981423e-01 -4.86200154e-01
7.26175904e-01 2.49262288e-01 -3.42546433e-01 -2.30826348e-01
-1.06019950e+00 7.67159462e-01 3.70580673e-01 -3.80783767e-01
-8.93701255e-01 -9.08108771e-01 -7.09006310e-01 -7.89187923e-02
3.35885823e-01 -5.16621828e-01 1.04809117e+00 -1.28110194e+00
-1.44247067e+00 1.06535935e+00 -1.69919506e-01 -3.97672683e-01
4.67361838e-01 -4.22275811e-02 -5.44340611e-01 -7.91520532e-03
3.23593140e-01 6.61897898e-01 1.00334561e+00 -1.40351284e+00
-7.04896033e-01 -4.48388398e-01 -2.90110081e-01 2.51698017e-01
-5.76142311e-01 -3.24383229e-01 -2.95542687e-01 -8.43758047e-01
6.95024505e-02 -7.11638629e-01 -9.10696536e-02 -5.86847998e-02
-3.10062796e-01 -1.56816661e-01 9.88768101e-01 -4.31378126e-01
1.24113619e+00 -2.27663326e+00 8.38720873e-02 3.63054246e-01
1.88103512e-01 3.88683558e-01 -4.55140740e-01 1.65720508e-01
-3.71715486e-01 -2.50696659e-01 -5.14337361e-01 -3.98870826e-01
-9.64412540e-02 1.84945866e-01 -3.86509269e-01 4.46209878e-01
2.76593626e-01 5.99093080e-01 -1.02261353e+00 -3.32565844e-01
8.75869393e-02 3.53002548e-01 -5.78209043e-01 1.60576195e-01
-1.82102188e-01 6.36564314e-01 -3.26892048e-01 6.31938338e-01
8.04476678e-01 -3.16529989e-01 3.34112078e-01 -3.78487229e-01
3.10850680e-01 5.67662537e-01 -1.28241634e+00 1.91349638e+00
-3.90122682e-01 2.47664794e-01 1.94688782e-01 -1.21253550e+00
1.15604842e+00 7.21429139e-02 3.91653687e-01 -8.85752678e-01
-8.75903070e-02 4.53985155e-01 -1.21063091e-01 -1.74766675e-01
2.64469981e-01 -6.26806393e-02 -2.12133806e-02 2.18506932e-01
2.64541358e-01 1.61264986e-01 4.56910059e-02 6.92328364e-02
9.74070728e-01 4.46468055e-01 4.79274869e-01 -4.43914503e-01
5.74754238e-01 1.72381341e-01 8.47288728e-01 4.73370522e-01
-4.55171078e-01 8.80630136e-01 1.80896625e-01 -2.96354890e-01
-9.19409990e-01 -1.26099622e+00 -2.56515205e-01 1.15566027e+00
3.19648504e-01 -3.27336460e-01 -8.43641996e-01 -1.17347836e+00
-5.11199795e-02 7.14252949e-01 -4.65843439e-01 -3.37574631e-01
-6.75268233e-01 -6.79916918e-01 3.68973911e-01 5.66947520e-01
6.51713014e-01 -8.39170694e-01 -1.83617502e-01 2.41026804e-01
-3.02120507e-01 -9.55272734e-01 -5.23227036e-01 2.91918010e-01
-8.24942291e-01 -9.32384729e-01 -6.81047499e-01 -8.94521236e-01
7.77796805e-01 2.76463449e-01 1.30352020e+00 -3.35259438e-01
1.37426779e-01 2.29592815e-01 -2.37071246e-01 -4.91788089e-02
-3.35846424e-01 4.05656636e-01 2.23988533e-01 1.87883973e-01
8.05577397e-01 -9.35013831e-01 -4.65136111e-01 6.20897353e-01
-7.10398376e-01 -3.13899517e-01 5.46506286e-01 1.07163036e+00
8.78190279e-01 -8.71127099e-02 8.98254216e-01 -1.09693015e+00
4.39817280e-01 -8.12613904e-01 -3.83159369e-01 1.62010729e-01
-7.78096795e-01 -1.34070083e-01 8.22228253e-01 -6.35313749e-01
-1.09577942e+00 1.78392649e-01 5.68893999e-02 -6.05781019e-01
-5.35118222e-01 2.95845509e-01 -6.31185353e-01 1.10370308e-01
1.22235751e+00 1.82806373e-01 -8.99259280e-03 -6.67497575e-01
3.40262145e-01 7.86497891e-01 6.37802780e-01 -8.66508782e-01
8.14833105e-01 4.85179693e-01 -4.83511835e-01 -6.45484686e-01
-9.70540822e-01 -5.78671217e-01 -9.31480527e-01 1.12142876e-01
4.55146462e-01 -1.07287383e+00 1.32656187e-01 4.34131652e-01
-8.05086732e-01 -6.18199825e-01 -6.22784019e-01 1.98199973e-01
-5.00348806e-01 2.82139748e-01 -2.21308157e-01 -3.89951229e-01
-1.39487103e-01 -1.01867414e+00 8.98668706e-01 1.46449745e-01
-3.69295716e-01 -1.11878717e+00 1.54621899e-01 2.87006825e-01
3.82684708e-01 1.55772150e-01 8.89769316e-01 -1.08590364e+00
-3.58959854e-01 -4.80168164e-02 -1.74541190e-01 7.06138432e-01
4.33550864e-01 -4.11616504e-01 -1.15707493e+00 -3.61237556e-01
1.70031548e-01 -3.45847428e-01 7.41592884e-01 1.87135965e-01
9.86535072e-01 -2.26155147e-01 -4.15697992e-01 6.77713335e-01
1.43142080e+00 1.71769578e-02 4.23500061e-01 5.55856884e-01
7.31205881e-01 5.86579025e-01 7.95864522e-01 3.24213475e-01
6.00231349e-01 8.27727914e-01 2.68231124e-01 -8.81684273e-02
-5.29678285e-01 -2.45942548e-01 5.31606972e-01 6.91075027e-01
3.57762575e-01 2.87031243e-03 -1.08266270e+00 8.34273815e-01
-1.97621870e+00 -8.84151697e-01 6.01162063e-03 2.27732730e+00
1.10380697e+00 1.26270413e-01 2.83260345e-01 -8.08903426e-02
8.18740666e-01 1.25076786e-01 -7.73807287e-01 6.79289699e-02
-2.50952750e-01 1.65162012e-01 2.35354468e-01 2.98449039e-01
-1.26132619e+00 9.37077224e-01 5.09807205e+00 1.23649836e+00
-1.15546441e+00 2.99265474e-01 6.73335552e-01 -2.30790630e-01
-8.04833174e-02 -7.97134489e-02 -1.01556456e+00 6.71510160e-01
9.55070555e-01 2.62941383e-02 2.84601659e-01 1.22511041e+00
-5.19995317e-02 1.02607645e-01 -1.29409826e+00 9.97815847e-01
-2.41305423e-03 -1.12145233e+00 -1.03235632e-01 -2.83091553e-02
8.95901084e-01 2.95753658e-01 2.22137123e-01 3.95941138e-01
2.85017699e-01 -7.08357573e-01 7.61691630e-01 2.42560983e-01
8.15016747e-01 -5.85701287e-01 6.34096503e-01 4.80834156e-01
-1.12498105e+00 -1.24696366e-01 -4.20965075e-01 9.22818258e-02
-1.39356047e-01 7.97804296e-01 -6.31954849e-01 5.78881025e-01
9.42471266e-01 1.05265439e+00 -6.42697036e-01 8.07464123e-01
-1.67698532e-01 6.91585302e-01 -4.05101717e-01 5.54866314e-01
1.25724271e-01 -3.94295335e-01 6.86218441e-01 1.15318286e+00
2.07252085e-01 -2.83631027e-01 3.02270770e-01 1.01081562e+00
-1.60227373e-01 1.22672051e-01 -5.28450608e-01 3.04007798e-01
8.12781394e-01 1.21630836e+00 -4.08311009e-01 -4.32602584e-01
-3.89594316e-01 9.69505548e-01 6.25162303e-01 3.95162523e-01
-7.23439097e-01 -3.71435344e-01 8.19796681e-01 1.78169325e-01
4.16774631e-01 1.14378165e-02 -7.70457268e-01 -1.28697920e+00
2.51361787e-01 -1.19281459e+00 6.13453686e-01 -4.01219517e-01
-1.79459167e+00 6.81653798e-01 -1.74293835e-02 -1.57559669e+00
-1.57892689e-01 -4.15496022e-01 -4.93086725e-01 9.14212644e-01
-1.73114538e+00 -1.19722569e+00 -3.54522467e-01 8.45643222e-01
5.86955547e-01 -3.58035237e-01 8.36478233e-01 5.14991999e-01
-7.19713509e-01 9.36914265e-01 3.72281611e-01 1.58482611e-01
1.22058594e+00 -1.18110013e+00 1.41943902e-01 8.74229550e-01
9.73409489e-02 5.60313523e-01 4.17113423e-01 -5.16194224e-01
-6.05099976e-01 -1.38725221e+00 8.46588135e-01 -6.48460388e-01
5.92037976e-01 -4.75260019e-01 -1.23483729e+00 6.31291389e-01
8.80319923e-02 2.53481954e-01 7.08935201e-01 7.82388449e-02
-8.81923556e-01 -5.26991963e-01 -1.46923411e+00 4.92876709e-01
1.13162827e+00 -4.91718084e-01 -6.51031852e-01 3.17544192e-01
4.87143755e-01 -2.75081277e-01 -9.00594592e-01 3.32737416e-01
1.27705187e-01 -1.17708480e+00 1.03821802e+00 -3.63725007e-01
3.27006280e-01 -5.48168778e-01 -3.58559877e-01 -1.43055284e+00
-4.92252201e-01 -2.86920786e-01 -9.41473693e-02 1.66202331e+00
3.44840735e-01 -8.07779610e-01 8.54531586e-01 3.78952175e-01
-2.48587951e-01 -6.13409758e-01 -1.06315792e+00 -1.14199007e+00
3.90537053e-01 -3.13744783e-01 6.26534998e-01 1.28100061e+00
-5.95168248e-02 5.34561276e-01 -2.08191976e-01 2.62931406e-01
6.34879827e-01 1.13003269e-01 7.88459241e-01 -1.15692437e+00
-3.11058313e-01 -4.67324764e-01 -1.52926251e-01 -1.09658241e+00
1.95987806e-01 -1.12425816e+00 3.01405881e-02 -1.16613293e+00
4.99052033e-02 -7.28239179e-01 -4.90214050e-01 7.67615139e-01
-1.14301801e-01 2.92563260e-01 9.39647183e-02 5.36678255e-01
-5.89561164e-01 6.14250124e-01 8.58518898e-01 -2.45481566e-01
-3.71796548e-01 -8.34615529e-02 -8.22320104e-01 6.49231732e-01
9.32556152e-01 -5.60920417e-01 -5.49476087e-01 -3.14675748e-01
-2.73734778e-01 -5.92660725e-01 2.84428567e-01 -1.28305233e+00
1.77609131e-01 -4.93864007e-02 2.20336214e-01 -2.60895282e-01
1.75811350e-01 -9.55367088e-01 -1.06066301e-01 7.58289173e-02
-3.54689389e-01 -2.99309820e-01 9.12309662e-02 6.72845781e-01
-4.27700102e-01 -1.28450006e-01 1.37895954e+00 1.05269313e-01
-9.21347797e-01 2.36059368e-01 5.98460473e-02 3.54715884e-01
1.02882791e+00 -2.52694219e-01 -4.53386635e-01 -1.47262380e-01
-7.16237187e-01 1.67119637e-01 6.01997495e-01 4.17232156e-01
4.13198471e-01 -1.49118674e+00 -5.90975642e-01 3.67642820e-01
4.82143193e-01 2.64422208e-01 1.53776169e-01 8.87106299e-01
-2.69065448e-03 4.82761525e-02 -1.99984163e-01 -7.64754772e-01
-9.82101560e-01 6.30220294e-01 5.03345311e-01 -4.33213681e-01
-4.96704280e-01 9.19196010e-01 4.60161835e-01 -9.22083557e-01
3.16000193e-01 -7.40197971e-02 -5.64826913e-02 1.52481973e-01
4.84818548e-01 3.55754703e-01 2.69080728e-01 -6.54786646e-01
-4.59612787e-01 5.49897671e-01 -3.16503763e-01 2.12234408e-01
1.38476539e+00 -3.01197261e-01 5.83253466e-02 2.99611270e-01
1.15083933e+00 -1.04894415e-01 -1.55192983e+00 -7.82747328e-01
3.15208167e-01 -4.11629200e-01 -1.40126869e-01 -9.83191729e-01
-1.14324844e+00 8.88603508e-01 7.46358573e-01 -1.75955877e-01
1.30366659e+00 -3.59247029e-02 7.54521310e-01 8.87077227e-02
3.94348323e-01 -1.29840505e+00 2.10831672e-01 5.47308683e-01
7.62775660e-01 -1.34729564e+00 -2.99921781e-01 -3.40887696e-01
-7.41578758e-01 8.41237426e-01 8.29854608e-01 -2.97403336e-01
6.24443471e-01 2.11849347e-01 2.54646868e-01 5.22461757e-02
-5.88213563e-01 -2.03684896e-01 2.63570130e-01 1.12494624e+00
8.33077580e-02 -1.82956770e-01 1.63935184e-01 8.47375691e-01
1.19629297e-02 -1.29104361e-01 1.56983703e-01 7.04312265e-01
-2.89629579e-01 -1.40397453e+00 -4.22769427e-01 1.45193398e-01
-1.14268474e-01 -1.54678151e-01 -4.07982767e-01 7.67211914e-01
3.53271663e-01 8.71091604e-01 4.90779094e-02 -4.00976926e-01
4.51205671e-01 3.74885291e-01 2.08593830e-01 -7.50379503e-01
-4.81213570e-01 4.03858200e-02 -2.84112245e-02 -6.53338671e-01
-4.60770309e-01 -6.82091475e-01 -1.23615909e+00 -5.16975746e-02
-5.68054877e-02 -1.12554263e-02 3.16373497e-01 8.24393690e-01
7.93328583e-01 3.65515739e-01 5.75840890e-01 -8.59193027e-01
-7.45212793e-01 -8.21223736e-01 -5.02939999e-01 6.85862660e-01
3.16688776e-01 -6.78490758e-01 -3.40016991e-01 1.36726677e-01] | [10.276878356933594, 2.999518871307373] |
c8dfed0a-7832-4b22-baf2-101c24b699cf | impactcite-an-xlnet-based-method-for-citation | 2005.06611 | null | https://arxiv.org/abs/2005.06611v1 | https://arxiv.org/pdf/2005.06611v1.pdf | ImpactCite: An XLNet-based method for Citation Impact Analysis | Citations play a vital role in understanding the impact of scientific literature. Generally, citations are analyzed quantitatively whereas qualitative analysis of citations can reveal deeper insights into the impact of a scientific artifact in the community. Therefore, citation impact analysis (which includes sentiment and intent classification) enables us to quantify the quality of the citations which can eventually assist us in the estimation of ranking and impact. The contribution of this paper is two-fold. First, we benchmark the well-known language models like BERT and ALBERT along with several popular networks for both tasks of sentiment and intent classification. Second, we provide ImpactCite, which is XLNet-based method for citation impact analysis. All evaluations are performed on a set of publicly available citation analysis datasets. Evaluation results reveal that ImpactCite achieves a new state-of-the-art performance for both citation intent and sentiment classification by outperforming the existing approaches by 3.44% and 1.33% in F1-score. Therefore, we emphasize ImpactCite (XLNet-based solution) for both tasks to better understand the impact of a citation. Additional efforts have been performed to come up with CSC-Clean corpus, which is a clean and reliable dataset for citation sentiment classification. | ['Sheraz Ahmed', 'Dominique Mercier', 'Vikas Rajashekar', 'Syed Tahseen Raza Rizvi', 'Andreas Dengel'] | 2020-05-05 | null | null | null | null | ['citation-intent-classification'] | ['natural-language-processing'] | [-3.68255675e-01 -4.08196121e-01 -6.38589203e-01 6.41788691e-02
-7.27101505e-01 -7.38649368e-01 1.07013750e+00 5.42075455e-01
-4.31499571e-01 7.93447793e-01 4.63410795e-01 -6.15171254e-01
-2.05991626e-01 -7.62273967e-01 -4.72172320e-01 -2.66710252e-01
6.85593486e-02 1.95749864e-01 3.84001881e-02 -8.69330242e-02
1.07848358e+00 4.13116693e-01 -1.31566918e+00 -4.77979407e-02
1.17662799e+00 9.11901176e-01 1.28898621e-01 5.05666256e-01
-6.02378547e-01 9.75904703e-01 -9.98979688e-01 -7.39566505e-01
-4.20380741e-01 -1.92505330e-01 -8.03838134e-01 -8.65675390e-01
1.59725085e-01 3.97767335e-01 -9.01562050e-02 1.16079688e+00
3.19215536e-01 -2.90282339e-01 9.60992217e-01 -1.08888006e+00
-7.74022222e-01 9.54379380e-01 -7.18312383e-01 6.94187462e-01
5.22818565e-01 -6.12759292e-02 1.46259999e+00 -7.96323717e-01
8.11855674e-01 1.17381656e+00 4.45984870e-01 -1.84708580e-01
-6.60693347e-01 -9.74802256e-01 1.77328549e-02 4.20169204e-01
-9.23229039e-01 -6.71223775e-02 9.24632907e-01 -6.78696752e-01
5.57994604e-01 3.75082403e-01 7.50753105e-01 1.21622443e+00
4.99675661e-01 6.78571522e-01 1.47019207e+00 -3.39110970e-01
2.49585763e-01 -4.71759103e-02 6.68530405e-01 4.44559485e-01
6.86322868e-01 -4.30169731e-01 -7.81662524e-01 -2.78084636e-01
1.33684099e-01 2.76518390e-02 -3.73012513e-01 7.23140836e-01
-1.40085101e+00 6.45748734e-01 5.99957049e-01 7.30121911e-01
-5.35274625e-01 1.55448541e-01 2.88073868e-01 1.67098820e-01
8.17978442e-01 9.03717101e-01 -2.02182397e-01 -5.68486392e-01
-8.89671803e-01 1.34260833e-01 1.14240313e+00 6.64360881e-01
3.22602063e-01 -2.92543173e-01 -5.82840145e-01 5.05983412e-01
3.03817868e-01 6.47866905e-01 4.14425761e-01 -8.85161340e-01
3.70219767e-01 9.46087122e-01 -8.86320919e-02 -1.39688325e+00
-1.09661385e-01 -1.14661014e+00 -7.34461844e-01 -1.11818828e-01
2.11621076e-02 -8.67921785e-02 -5.47850311e-01 1.30022180e+00
-2.70275712e-01 3.03571373e-01 -1.73210815e-01 6.98035479e-01
1.32795155e+00 7.02788711e-01 2.25249141e-01 -1.18077159e-01
1.43681645e+00 -9.14287388e-01 -8.53842258e-01 7.22755417e-02
5.82767487e-01 -1.17904913e+00 1.05245757e+00 2.94896483e-01
-8.02612066e-01 -1.37474775e-01 -1.06829786e+00 2.93852568e-01
-7.04086125e-01 -1.41331963e-02 8.09511960e-01 2.42382556e-01
-9.15644467e-01 6.25252068e-01 -2.69266158e-01 -1.75182477e-01
6.86891139e-01 -1.40084914e-04 -1.00794986e-01 -4.52560447e-02
-1.27863443e+00 9.77131069e-01 -1.82128027e-01 -3.88706595e-01
-3.49296540e-01 -1.04078686e+00 -2.44867116e-01 1.86990410e-01
1.19713508e-01 -7.14094341e-01 1.05203629e+00 -5.80051780e-01
-1.10452628e+00 6.55998051e-01 -3.97345483e-01 -9.09463316e-02
2.82744795e-01 -1.50353923e-01 -4.43660796e-01 -1.07769877e-01
5.47927797e-01 -3.73583436e-02 1.25605874e-02 -1.14572191e+00
-6.83374226e-01 -3.31952065e-01 1.62191987e-01 -1.24817267e-01
-8.60837996e-01 3.25077206e-01 -8.38160157e-01 -7.80148327e-01
-4.25413579e-01 -7.43590295e-01 8.55735466e-02 -5.22297621e-01
-5.94713032e-01 -7.43877709e-01 5.39818645e-01 -4.58191663e-01
1.52487099e+00 -1.62869024e+00 5.16912267e-02 2.18360677e-01
5.82771361e-01 1.16416372e-01 -2.24327564e-01 5.15763640e-01
1.76536337e-01 7.22452819e-01 1.83175445e-01 -4.46973741e-01
-1.68935031e-01 -3.63999456e-01 -1.73879638e-01 2.87394971e-01
2.25268193e-02 1.14552903e+00 -1.00215209e+00 -5.04749894e-01
-3.62148702e-01 3.16993475e-01 -1.22281663e-01 8.79693776e-02
-4.95624878e-02 3.93369675e-01 -9.66157913e-01 1.00756550e+00
3.98931563e-01 -6.26771510e-01 -2.43602529e-01 -8.91406313e-02
-5.20323932e-01 5.34670293e-01 -4.58159864e-01 1.33562541e+00
-4.90468651e-01 9.89511192e-01 -4.16882157e-01 -6.58735752e-01
7.81994581e-01 5.21862842e-02 3.46390814e-01 -9.58028853e-01
4.59168404e-01 5.21371961e-01 1.19818775e-02 -3.04296136e-01
4.30844456e-01 1.69950694e-01 -1.31690055e-01 5.67366362e-01
-1.82636932e-01 1.80904225e-01 6.74487054e-01 6.01700306e-01
1.46529651e+00 -3.70317042e-01 -4.48788665e-02 -4.99372393e-01
5.20965576e-01 -2.45351139e-02 2.20206261e-01 8.15012217e-01
3.43327783e-02 3.03259403e-01 8.16596746e-01 -1.21411623e-03
-6.75105751e-01 -5.37710607e-01 -6.63137063e-02 8.35051775e-01
1.90377638e-01 -4.45250988e-01 -5.31834483e-01 -5.28439999e-01
1.47670671e-01 7.16199040e-01 -7.14521289e-01 9.01325792e-02
-2.97375202e-01 -8.84806573e-01 4.83274400e-01 4.79937404e-01
3.99019420e-01 -1.06765556e+00 1.28865167e-01 -5.98557759e-03
-3.74775946e-01 -9.12427902e-01 -1.47404030e-01 -9.31075811e-02
-7.47867763e-01 -1.27986920e+00 -9.42610383e-01 -4.08139318e-01
5.15533507e-01 1.66752368e-01 1.37026000e+00 3.77307564e-01
6.43206853e-03 1.70045063e-01 -4.27977085e-01 -7.83867717e-01
-4.60355058e-02 5.43299258e-01 -1.98862568e-01 -3.33267838e-01
5.07012010e-01 -4.23528194e-01 -7.00132072e-01 -2.44110078e-01
-5.57599187e-01 -3.47904176e-01 9.45350707e-01 4.44336146e-01
2.28047282e-01 -2.30350077e-01 8.69258761e-01 -1.10019910e+00
1.39320087e+00 -9.02588248e-01 -3.52484584e-01 4.97950353e-02
-1.21674573e+00 -1.26652688e-01 4.50341374e-01 -1.11549489e-01
-9.05003309e-01 -1.13757062e+00 -9.52916592e-02 -7.76573867e-02
3.11080426e-01 1.25444984e+00 4.34180737e-01 -1.36338979e-01
3.04472208e-01 -1.81869462e-01 -3.67896080e-01 -6.41320109e-01
1.95425555e-01 1.00185001e+00 3.30779314e-01 -6.29923582e-01
6.60482943e-01 2.15508834e-01 2.23936528e-01 -3.69311631e-01
-1.24124706e+00 -8.28252256e-01 -2.70141661e-01 -4.63631541e-01
4.38972086e-01 -9.08970475e-01 -9.79278326e-01 3.15101713e-01
-1.53097522e+00 4.86871809e-01 2.78797776e-01 5.13580620e-01
3.19178998e-01 2.53040969e-01 -6.33881152e-01 -6.46743596e-01
-7.65227258e-01 -1.08466220e+00 7.91185975e-01 4.09978777e-01
-4.37220633e-01 -1.23529780e+00 3.16876382e-01 6.45448029e-01
6.49481416e-01 2.71511585e-01 9.36911583e-01 -6.03680432e-01
-5.47166765e-01 -4.24972206e-01 -6.42186761e-01 2.14482825e-02
-7.31852129e-02 2.60625809e-01 -8.03440988e-01 1.03654377e-02
-3.50857824e-01 1.20591149e-01 1.10052001e+00 3.41889232e-01
1.16230428e+00 -2.81544358e-01 -6.38125002e-01 2.28299379e-01
1.37864125e+00 8.57784078e-02 5.52896023e-01 9.48955953e-01
7.28804588e-01 5.33479393e-01 5.51810503e-01 3.34683985e-01
5.82381010e-01 2.53554225e-01 4.09749448e-01 -2.35666949e-02
-1.96615100e-01 4.96582314e-02 2.30460260e-02 1.45010912e+00
-6.28365338e-01 -5.76004982e-01 -1.17441142e+00 7.17598259e-01
-1.66166282e+00 -7.24405885e-01 -7.01803505e-01 1.64647341e+00
9.35371995e-01 4.71166641e-01 -1.19492471e-01 4.33936948e-03
6.64132297e-01 2.02843934e-01 -2.20383704e-01 -2.52185136e-01
-3.34779739e-01 2.20226705e-01 4.82910812e-01 2.07498044e-01
-6.91455066e-01 6.96564436e-01 6.03193712e+00 9.25492108e-01
-1.24139452e+00 1.75736994e-01 7.15112567e-01 1.30807415e-01
-5.99562049e-01 1.48536637e-01 -5.78489840e-01 8.59939992e-01
9.76208568e-01 -6.30617321e-01 -1.13798074e-01 6.50506258e-01
2.04431027e-01 -5.81267811e-02 -7.58102238e-01 7.26139247e-01
3.44661921e-02 -1.70526814e+00 1.87004492e-01 9.86835137e-02
9.23561394e-01 3.65589589e-01 4.52717543e-02 3.21561366e-01
1.69145033e-01 -9.04994726e-01 6.52325213e-01 8.64363372e-01
7.08595932e-01 -7.68681347e-01 1.25873268e+00 2.25467995e-01
-7.80662119e-01 1.18285239e-01 -1.77937523e-01 -4.11764979e-01
-2.41432544e-02 1.22999525e+00 -4.11391526e-01 6.45001173e-01
9.55588877e-01 1.30040276e+00 -6.38719797e-01 1.17947173e+00
-2.87892133e-01 1.04812944e+00 1.72338814e-01 -7.16026664e-01
2.14128405e-01 -2.02395394e-01 7.79443026e-01 1.50487065e+00
4.61598545e-01 -8.40057526e-03 -2.09967121e-01 9.82023895e-01
-8.61292899e-01 2.61203080e-01 -4.98160243e-01 -5.67344069e-01
6.89782619e-01 1.47370899e+00 -6.22726381e-01 -4.15917128e-01
-3.66652250e-01 3.83835703e-01 3.02668750e-01 2.38712385e-01
-5.28459191e-01 -6.87143028e-01 4.86099094e-01 1.29986778e-01
2.74985880e-02 1.24740534e-01 -7.39479363e-01 -1.10892844e+00
-4.20968197e-02 -3.93677860e-01 -4.41318341e-02 -7.25365043e-01
-1.58672440e+00 6.63891375e-01 -4.35115844e-01 -1.13166928e+00
2.43429169e-01 -6.07842565e-01 -1.12621200e+00 1.05547559e+00
-2.12400842e+00 -9.05859530e-01 -5.71398795e-01 -1.47146717e-01
3.29210281e-01 -2.38458574e-01 5.27113795e-01 4.33963597e-01
-6.77187324e-01 2.24040598e-01 2.21679077e-01 -1.63367987e-02
8.06296170e-01 -1.26892877e+00 2.68098891e-01 5.98733783e-01
-1.47338033e-01 8.34802628e-01 6.18196309e-01 -8.45907867e-01
-1.53426182e+00 -1.05582809e+00 1.28127074e+00 -7.85291672e-01
1.16082513e+00 2.38852188e-01 -7.24862993e-01 1.91738561e-01
4.97246832e-01 -5.01946032e-01 8.61676812e-01 5.22144794e-01
-2.47560725e-01 1.03337429e-01 -6.04025304e-01 6.43989861e-01
1.00092793e+00 -3.19861293e-01 -6.94705486e-01 1.33698985e-01
8.02264214e-01 1.78181648e-01 -1.24595797e+00 4.42809999e-01
5.75538397e-01 -7.31596887e-01 7.26379395e-01 -3.88943762e-01
1.15116978e+00 -5.25274463e-02 2.50891209e-01 -1.40561819e+00
-5.79733193e-01 -2.71671861e-01 -1.80679008e-01 1.61496317e+00
6.47733510e-01 -6.96826756e-01 4.42607135e-01 1.12762064e-01
2.11080555e-02 -1.13003755e+00 -3.86792094e-01 -5.24854064e-01
1.60551399e-01 -4.48149085e-01 6.38984561e-01 1.27094221e+00
1.28442556e-01 6.53944254e-01 2.76776880e-01 -3.60039532e-01
7.45268047e-01 3.48737895e-01 4.22883987e-01 -1.91712821e+00
4.14529026e-01 -1.15734398e+00 -2.10764989e-01 -5.73861420e-01
4.48731482e-01 -9.71697390e-01 -4.76203650e-01 -1.92113531e+00
7.38535523e-01 -5.26167035e-01 -8.04942012e-01 2.41334647e-01
-5.81888855e-01 4.60859299e-01 4.66698855e-02 8.52358401e-01
-7.91329145e-01 3.79565418e-01 1.36607194e+00 -3.82487208e-01
2.78238893e-01 -1.61147669e-01 -1.30846798e+00 6.46335363e-01
6.34905994e-01 -6.39779150e-01 1.04622968e-01 -3.21191519e-01
7.44156778e-01 -1.64648697e-01 6.86466768e-02 -6.94718003e-01
5.80681264e-01 -2.42199957e-01 1.51704237e-01 -7.25429833e-01
-1.54508740e-01 -1.77823424e-01 -2.64817446e-01 2.88580835e-01
-4.82795328e-01 2.16853753e-01 -2.68523414e-02 5.79571426e-01
-4.46635693e-01 -3.21905077e-01 1.85523957e-01 -5.71323745e-02
-4.02643919e-01 1.15206204e-01 -1.08604178e-01 1.78194597e-01
6.51076078e-01 2.30018646e-01 -8.48620236e-01 -2.92563617e-01
1.43292919e-01 1.83981910e-01 3.65834475e-01 7.00743675e-01
3.63843679e-01 -1.06986642e+00 -9.84783709e-01 -5.19177377e-01
2.96857774e-01 -3.30830842e-01 -3.29079688e-01 1.05811763e+00
-4.60253596e-01 8.29384148e-01 7.13212416e-02 -2.32412234e-01
-1.12080681e+00 -9.05027264e-04 -3.08342576e-01 -4.65396196e-01
-5.61982989e-02 8.23233664e-01 -4.53398861e-02 -1.06198855e-01
9.68001857e-02 -1.10400215e-01 -9.10482049e-01 3.34814608e-01
4.15205300e-01 7.18336344e-01 1.33035600e-01 -4.71762598e-01
-4.55627322e-01 6.40253425e-01 -3.80738005e-02 1.06354348e-01
1.62794030e+00 1.19702995e-01 -8.67802680e-01 6.74476087e-01
1.20210326e+00 4.34976935e-01 -7.99043030e-02 -1.45623190e-02
3.33827615e-01 -2.84304798e-01 5.31375766e-01 -1.29647577e+00
-1.00400198e+00 6.70937359e-01 1.02127222e-02 2.93124825e-01
7.70138800e-01 1.60287276e-01 7.75300443e-01 2.13229433e-01
1.80009335e-01 -7.76249588e-01 2.24832475e-01 5.87146699e-01
9.73904014e-01 -1.17859006e+00 1.45321816e-01 -2.80142248e-01
-3.03900778e-01 1.00067759e+00 4.00286704e-01 5.88703007e-02
6.25925779e-01 2.07075238e-01 1.15839383e-02 -7.00612545e-01
-8.16933811e-01 -2.44128406e-02 8.25040996e-01 -4.39634100e-02
1.04528606e+00 1.31665528e-01 -8.67551267e-01 7.84066856e-01
-2.88290113e-01 -8.77972469e-02 5.80921233e-01 8.05635512e-01
-2.98933953e-01 -1.08701539e+00 -3.67738545e-01 9.00790870e-01
-1.05226433e+00 -3.90234530e-01 -8.19242120e-01 3.96130115e-01
-2.66788095e-01 1.00547397e+00 -1.77313954e-01 -3.99818569e-01
3.50023597e-01 -3.58233392e-01 1.67943798e-02 -4.86045122e-01
-7.30500400e-01 -1.55753002e-01 1.16939314e-01 -2.52623379e-01
-5.61836302e-01 -5.08578300e-01 -1.08943033e+00 -6.23110771e-01
-5.84988356e-01 5.28116941e-01 1.15964055e+00 9.57114041e-01
6.18251503e-01 1.18125343e+00 5.39156258e-01 -4.95239079e-01
-9.65536013e-02 -1.37732732e+00 -4.28227395e-01 1.73517585e-01
7.87399244e-03 -7.42320955e-01 -7.90227950e-01 -1.68409735e-01] | [9.63268756866455, 8.259973526000977] |
a9324587-c59d-4afa-b4a4-348fb1cfe22d | positional-contrastive-learning-for | 2106.09157 | null | https://arxiv.org/abs/2106.09157v3 | https://arxiv.org/pdf/2106.09157v3.pdf | Positional Contrastive Learning for Volumetric Medical Image Segmentation | The success of deep learning heavily depends on the availability of large labeled training sets. However, it is hard to get large labeled datasets in medical image domain because of the strict privacy concern and costly labeling efforts. Contrastive learning, an unsupervised learning technique, has been proved powerful in learning image-level representations from unlabeled data. The learned encoder can then be transferred or fine-tuned to improve the performance of downstream tasks with limited labels. A critical step in contrastive learning is the generation of contrastive data pairs, which is relatively simple for natural image classification but quite challenging for medical image segmentation due to the existence of the same tissue or organ across the dataset. As a result, when applied to medical image segmentation, most state-of-the-art contrastive learning frameworks inevitably introduce a lot of false-negative pairs and result in degraded segmentation quality. To address this issue, we propose a novel positional contrastive learning (PCL) framework to generate contrastive data pairs by leveraging the position information in volumetric medical images. Experimental results on CT and MRI datasets demonstrate that the proposed PCL method can substantially improve the segmentation performance compared to existing methods in both semi-supervised setting and transfer learning setting. | ['Yiyu Shi', 'Jingtong Hu', 'Jian Zhuang', 'Meiping Huang', 'Haiyun Yuan', 'Xiaowei Xu', 'Xinrong Hu', 'Yawen Wu', 'Dewen Zeng'] | 2021-06-16 | null | null | null | null | ['volumetric-medical-image-segmentation'] | ['medical'] | [ 6.37480319e-01 2.04553887e-01 -2.73510247e-01 -6.39777243e-01
-1.03673565e+00 -5.12925327e-01 2.89366782e-01 2.01430738e-01
-5.31278014e-01 9.54356194e-01 -1.30955741e-01 -2.19569281e-01
3.76294777e-02 -5.92724621e-01 -7.00859010e-01 -9.72976804e-01
-1.74719449e-02 4.79332089e-01 2.75273234e-01 1.94782764e-01
-2.61544455e-02 3.07683378e-01 -9.11489487e-01 1.72597691e-01
1.00962245e+00 1.05457854e+00 3.32173258e-01 1.00179590e-01
-1.54712200e-01 9.81740057e-01 -3.06220472e-01 -3.00985515e-01
5.09040534e-01 -5.08297324e-01 -9.17426050e-01 3.72940838e-01
3.03844154e-01 -3.22158635e-01 -2.57079490e-02 1.31176543e+00
5.86727679e-01 -3.54193896e-02 7.75329709e-01 -1.09031951e+00
-4.77375597e-01 6.28667772e-01 -8.01565230e-01 3.79522830e-01
-2.96956897e-01 2.36552842e-02 7.33550310e-01 -8.08404744e-01
4.82865453e-01 6.77867353e-01 3.35147858e-01 5.73599517e-01
-1.03892720e+00 -8.63499045e-01 -1.35927454e-01 -1.60154685e-01
-1.19437802e+00 -2.68708527e-01 9.36181784e-01 -5.19535303e-01
-9.93101895e-02 2.15734467e-01 4.34049487e-01 7.93471932e-01
1.40060663e-01 8.58565688e-01 1.47704113e+00 -1.17601648e-01
2.35865116e-01 1.91213399e-01 -8.66562054e-02 9.10761714e-01
5.40334769e-02 2.58401960e-01 -1.02049671e-01 2.47472282e-02
9.07877207e-01 2.61471421e-01 -3.14803243e-01 -5.97679138e-01
-1.23756993e+00 7.31062829e-01 1.01587570e+00 1.44212157e-01
-2.79398352e-01 -2.42729321e-01 6.20644987e-01 2.58781947e-02
4.80927140e-01 3.62175047e-01 -4.59365666e-01 4.52679455e-01
-9.12126601e-01 -1.06814526e-01 2.88615257e-01 8.30781817e-01
6.23199701e-01 -1.05528034e-01 -3.18625063e-01 7.39488065e-01
1.99523956e-01 1.74469441e-01 7.18961000e-01 -4.11821187e-01
5.19677460e-01 6.43350661e-01 -1.26828656e-01 -7.14326143e-01
-4.57695335e-01 -8.68654490e-01 -1.15833402e+00 1.27354383e-01
3.50871116e-01 -3.54099661e-01 -1.22954488e+00 1.55411541e+00
5.34165502e-01 3.59096020e-01 -9.60242450e-02 1.26007318e+00
9.50236499e-01 3.65640819e-01 2.93289632e-01 -5.10106623e-01
1.02461386e+00 -1.04371572e+00 -6.43905878e-01 -2.81067461e-01
7.15508759e-01 -5.53819358e-01 9.51743066e-01 6.71910420e-02
-8.88158441e-01 -4.58234698e-01 -1.20471346e+00 1.26891404e-01
-6.09122626e-02 2.67431960e-02 7.14707673e-01 5.30996621e-01
-5.41447818e-01 3.82322699e-01 -8.72645319e-01 2.75601566e-01
1.22328901e+00 5.43970346e-01 -3.27489197e-01 -3.67849350e-01
-1.06555665e+00 4.87833649e-01 3.76488954e-01 2.52142400e-01
-1.01451850e+00 -8.28679979e-01 -7.59576261e-01 -1.47587523e-01
5.31375527e-01 -3.79603893e-01 1.00184405e+00 -1.16152501e+00
-1.24010062e+00 1.11392438e+00 4.06380028e-01 -3.42322946e-01
1.02082062e+00 1.41567960e-01 -3.90861779e-02 2.72542328e-01
3.93776119e-01 9.10262525e-01 8.58594894e-01 -1.23368430e+00
-5.29527009e-01 -5.51587164e-01 -2.63896763e-01 3.20609987e-01
-5.00154607e-02 -3.81371945e-01 -6.53706044e-02 -6.98568285e-01
2.01743737e-01 -1.02140939e+00 -5.70267975e-01 3.23812068e-01
-5.59076428e-01 8.93844590e-02 7.56649137e-01 -4.10831898e-01
5.05371988e-01 -2.19508839e+00 -1.49822816e-01 8.43586773e-02
3.59809935e-01 4.79467601e-01 1.84193775e-01 -3.23959529e-01
-1.17913119e-01 5.45124151e-02 -5.75620770e-01 4.90586385e-02
-5.34253120e-01 1.04800254e-01 7.32634217e-02 6.86168849e-01
3.53388518e-01 1.00870740e+00 -1.24506807e+00 -1.02632916e+00
3.10394734e-01 2.26622269e-01 -3.17349315e-01 3.16608489e-01
-7.03536943e-02 1.32425749e+00 -7.15370536e-01 6.30252957e-01
7.73120344e-01 -5.33860326e-01 5.53137138e-02 -3.64777565e-01
2.58415908e-01 -3.80700678e-02 -6.48116291e-01 1.72623622e+00
-1.76821515e-01 2.32719496e-01 -1.18224397e-01 -1.49266028e+00
7.42098510e-01 4.88702565e-01 7.54701257e-01 -7.11982369e-01
4.68788981e-01 2.63704449e-01 2.04804212e-01 -7.63923705e-01
-2.11473778e-01 -5.57934761e-01 -5.35070822e-02 2.19941646e-01
2.54882034e-02 -1.39816061e-01 -1.43028200e-01 -4.53573186e-03
8.45905483e-01 -1.37067363e-01 1.40674025e-01 -3.59091371e-01
6.16870165e-01 -1.26673607e-02 8.53607714e-01 3.57401729e-01
-6.46401942e-01 7.46266782e-01 2.52107143e-01 -3.20947081e-01
-8.49453866e-01 -1.08573985e+00 -3.72166663e-01 8.23747098e-01
2.18323141e-01 2.92063087e-01 -7.67649531e-01 -1.10658193e+00
-3.46457958e-01 2.48113684e-02 -6.97467923e-01 -1.85332537e-01
-5.04281223e-01 -9.61471796e-01 3.76413703e-01 4.95917708e-01
9.11685646e-01 -1.01040232e+00 -5.97819090e-01 1.27486840e-01
-1.14026196e-01 -1.23580778e+00 -5.75668752e-01 3.65550160e-01
-1.00170827e+00 -1.14207613e+00 -9.80505228e-01 -1.30221748e+00
1.13347280e+00 2.41727099e-01 8.15965414e-01 1.19218141e-01
-4.63260740e-01 -2.44815037e-01 -3.09116304e-01 -3.72961104e-01
-3.41923445e-01 8.52514207e-02 -3.58570427e-01 2.00297996e-01
-6.61852881e-02 -4.04955089e-01 -1.05397856e+00 3.02805513e-01
-9.61295426e-01 9.45614278e-02 7.88813710e-01 1.19318151e+00
9.75998044e-01 9.85377058e-02 9.02750134e-01 -1.54928005e+00
3.20770383e-01 -4.58456308e-01 -5.04755914e-01 2.65230715e-01
-5.02553225e-01 -5.46953583e-04 6.47143781e-01 -4.13114667e-01
-9.88747478e-01 4.68806088e-01 1.11397626e-02 -4.17807132e-01
4.40183282e-02 5.84546030e-01 3.49028893e-02 -1.92356169e-01
6.17954016e-01 2.37913832e-01 1.91971824e-01 -9.32517499e-02
1.73159182e-01 5.69840670e-01 5.91052353e-01 -3.39015573e-01
4.93556142e-01 5.00909925e-01 2.33984664e-01 -3.53013158e-01
-1.04431283e+00 -3.92041028e-01 -6.89610183e-01 -2.26450250e-01
9.48615134e-01 -8.11637223e-01 -2.08175719e-01 4.44490492e-01
-5.88860691e-01 -3.77993554e-01 -1.23563625e-01 5.89866638e-01
-4.35949802e-01 3.14529747e-01 -5.98569512e-01 -3.63416314e-01
-5.96271932e-01 -1.57445955e+00 8.39089215e-01 3.78295600e-01
2.86583513e-01 -8.42408061e-01 -2.41483852e-01 5.48976839e-01
1.43698052e-01 4.52141553e-01 9.57950950e-01 -5.52385628e-01
-5.08487940e-01 -2.32343495e-01 -3.85427415e-01 4.84953433e-01
6.21379018e-01 -5.60235202e-01 -7.50430346e-01 -3.81140202e-01
1.82471558e-01 -6.94605649e-01 6.31618917e-01 5.36496997e-01
1.49588573e+00 -1.54280365e-01 -4.22040492e-01 5.39984107e-01
1.38821542e+00 1.41525775e-01 3.66731286e-01 -1.46135703e-01
9.03531253e-01 6.34059906e-01 8.72709692e-01 7.27175698e-02
7.69263431e-02 2.57713497e-01 4.26576793e-01 -4.37780648e-01
-1.10378154e-01 -1.83300510e-01 -3.82710814e-01 8.64780426e-01
1.98052078e-01 6.56882674e-02 -8.42899680e-01 5.18708229e-01
-1.65725601e+00 -5.13996780e-01 -3.77563015e-02 2.17767739e+00
1.11379874e+00 1.29764214e-01 -9.76759940e-02 1.34555474e-01
8.60766292e-01 -1.82318464e-01 -7.94417441e-01 1.74629971e-01
3.32029402e-01 3.84305507e-01 7.40461349e-01 1.71212256e-01
-1.39395070e+00 7.64184713e-01 5.59250879e+00 7.11818695e-01
-1.60530651e+00 4.62250233e-01 1.26299226e+00 1.65348619e-01
7.79865235e-02 -4.17156339e-01 -1.67548180e-01 7.03791976e-01
4.00053859e-01 2.32359633e-01 -3.31719704e-02 7.33003616e-01
1.46759897e-01 -2.66746562e-02 -1.10150456e+00 1.06933796e+00
-1.64429858e-01 -1.30407953e+00 -1.25907376e-01 -7.46104494e-02
9.34652686e-01 4.09998484e-02 3.30643028e-01 2.12074310e-01
8.98071155e-02 -1.05701566e+00 2.54524410e-01 -6.37319963e-03
8.79639268e-01 -6.81359053e-01 8.09729755e-01 4.20618683e-01
-7.93359995e-01 9.76709127e-02 -4.17876571e-01 2.96775728e-01
7.73991197e-02 5.83127320e-01 -1.05439889e+00 3.44005764e-01
6.01816297e-01 5.85694015e-01 -4.30266708e-01 1.18624341e+00
-2.48188108e-01 5.85225046e-01 4.39794436e-02 2.02191412e-01
4.73365515e-01 -1.81125507e-01 8.44659135e-02 8.18482637e-01
-7.44268447e-02 1.78299874e-01 4.75008130e-01 7.42327690e-01
-4.64893311e-01 3.80209804e-01 -4.26904976e-01 1.17593497e-01
3.46701711e-01 1.51377034e+00 -1.02121985e+00 -3.24006855e-01
-3.18505973e-01 8.70300949e-01 4.26604837e-01 9.55001563e-02
-8.41135681e-01 1.48553997e-01 -2.52047211e-01 2.27854595e-01
1.60170212e-01 1.94729835e-01 -2.68555671e-01 -9.52322066e-01
-7.70522654e-02 -7.63910830e-01 3.90677422e-01 -4.62019026e-01
-1.58874285e+00 6.68902159e-01 -1.78355455e-01 -1.50014102e+00
-3.67711000e-05 -3.17946166e-01 -4.62534457e-01 4.98677462e-01
-1.56208909e+00 -1.16024458e+00 -4.89817053e-01 7.20813096e-01
4.83285129e-01 -3.56080905e-02 6.05287611e-01 6.49701118e-01
-5.18681347e-01 7.76667774e-01 -1.27250394e-02 4.92795348e-01
8.06856394e-01 -1.20494974e+00 -3.07755142e-01 5.35866678e-01
-9.74754170e-02 2.65822023e-01 2.90494323e-01 -5.33544302e-01
-8.49907041e-01 -1.46567428e+00 1.19268782e-01 3.31427045e-02
2.13515863e-01 -2.54799575e-01 -8.88558924e-01 6.45220578e-01
-7.00272098e-02 9.13136363e-01 8.93104911e-01 -4.04151738e-01
7.27413744e-02 -2.06621781e-01 -1.57068205e+00 3.71751428e-01
7.53342927e-01 -3.38181764e-01 -4.24119443e-01 6.45185590e-01
5.52394748e-01 -5.79616368e-01 -9.82888162e-01 6.55223668e-01
2.08761826e-01 -7.02286839e-01 8.46673846e-01 -5.45311272e-01
5.56127548e-01 -1.28083304e-01 1.60649255e-01 -1.20380712e+00
-7.33986050e-02 -3.09540004e-01 4.58232075e-01 1.13617241e+00
4.91581976e-01 -4.70514297e-01 1.01583326e+00 6.96212411e-01
-2.13969380e-01 -8.95666718e-01 -9.11720634e-01 -4.72007126e-01
3.79395992e-01 1.88948974e-01 3.10331553e-01 1.25651896e+00
-7.99585655e-02 5.35749793e-01 -3.66394460e-01 1.01853654e-01
8.97226870e-01 1.28320843e-01 3.36351186e-01 -9.70587909e-01
-2.15920910e-01 -4.08911221e-02 -5.39484560e-01 -7.78362513e-01
6.02493845e-02 -1.12973583e+00 4.15373683e-01 -1.18926239e+00
3.95716995e-01 -1.08968890e+00 -6.20179951e-01 3.11165690e-01
-5.07301271e-01 6.32240117e-01 -1.83859527e-01 4.14433867e-01
-4.51693863e-01 3.92756522e-01 1.90256310e+00 -4.34736311e-01
7.61990100e-02 4.82783876e-02 -5.10627389e-01 5.09473443e-01
8.27826679e-01 -7.48671830e-01 -8.21712255e-01 -3.32714021e-01
-1.60647660e-01 2.87316799e-01 2.71392882e-01 -7.93953657e-01
9.58456025e-02 1.21540390e-02 6.25975251e-01 -5.16706049e-01
-3.54052335e-02 -8.06846380e-01 -1.67933181e-01 5.91625571e-01
-5.72220504e-01 -3.93482327e-01 -1.89038306e-01 5.59294760e-01
-2.96213180e-01 -3.04707021e-01 1.08205557e+00 -5.24566233e-01
-4.40797627e-01 8.23615849e-01 2.20968723e-02 2.89359540e-01
1.19692016e+00 2.12473348e-02 2.58811396e-02 -1.74520779e-02
-8.09849679e-01 3.01512837e-01 3.83537672e-02 1.14468418e-01
5.29211164e-01 -1.23396111e+00 -6.54480219e-01 1.23930760e-01
5.58892339e-02 6.26619101e-01 2.72545040e-01 1.05439913e+00
-5.46178937e-01 1.21745907e-01 -4.63969231e-01 -8.20911288e-01
-9.16457355e-01 7.53452063e-01 3.69057447e-01 -3.26624244e-01
-5.91606855e-01 9.02882516e-01 6.56835318e-01 -5.16122043e-01
1.01310186e-01 -2.90303290e-01 -7.24831745e-02 -2.24327952e-01
3.32299709e-01 -2.80743912e-02 1.14592724e-01 -5.37695467e-01
-3.57470334e-01 2.39979118e-01 -5.65457940e-01 2.22952589e-01
1.21769094e+00 -4.39779684e-02 5.92209958e-02 1.06189124e-01
1.47631383e+00 -3.97505730e-01 -1.47563255e+00 -2.88087666e-01
-2.21372217e-01 -4.40137476e-01 3.59991074e-01 -7.63790667e-01
-1.61937845e+00 9.57540154e-01 1.04004252e+00 -1.19194493e-01
9.25796151e-01 -8.26824531e-02 9.40971375e-01 -6.42799214e-02
4.52042252e-01 -9.01672125e-01 2.85240829e-01 -3.16137165e-01
4.54642862e-01 -1.86005926e+00 6.79644495e-02 -7.82131076e-01
-7.06171572e-01 7.34302759e-01 6.82971537e-01 -2.91158974e-01
6.94929659e-01 2.28110150e-01 3.08084041e-01 -3.05319399e-01
-5.84559292e-02 6.80079758e-02 1.14445619e-01 5.19839525e-01
4.90345359e-01 1.63259730e-01 -4.30381954e-01 4.65091407e-01
1.43751845e-01 2.17318296e-01 2.65621215e-01 1.14091730e+00
-6.95987567e-02 -1.04473817e+00 3.20975631e-02 7.69673169e-01
-8.70816648e-01 6.46497831e-02 -1.64732113e-01 6.38553500e-01
1.98903427e-01 6.81832433e-01 -1.56137543e-02 -2.53241174e-02
-1.02856942e-01 -3.72652888e-01 7.05589831e-01 -8.04534733e-01
-5.61239600e-01 1.09885067e-01 -3.72428954e-01 -1.38724297e-01
-5.91270506e-01 -5.17370582e-01 -1.53833938e+00 2.17078283e-01
-5.02660632e-01 1.06425181e-01 2.70298600e-01 1.09965098e+00
9.62064788e-03 5.51382601e-01 9.07798767e-01 -5.09793818e-01
-5.56765378e-01 -9.05058026e-01 -4.92821425e-01 6.45332694e-01
3.15295011e-01 -7.50074625e-01 5.40494360e-02 1.12157427e-01] | [14.738402366638184, -2.1292803287506104] |
96f5c54d-0674-419f-9e9e-44049d3c8f1f | effects-of-word-embeddings-on-neural-network | 1805.05237 | null | http://arxiv.org/abs/1805.05237v2 | http://arxiv.org/pdf/1805.05237v2.pdf | Effects of Word Embeddings on Neural Network-based Pitch Accent Detection | Pitch accent detection often makes use of both acoustic and lexical features
based on the fact that pitch accents tend to correlate with certain words. In
this paper, we extend a pitch accent detector that involves a convolutional
neural network to include word embeddings, which are state-of-the-art vector
representations of words. We examine the effect these features have on
within-corpus and cross-corpus experiments on three English datasets. The
results show that while word embeddings can improve the performance in
corpus-dependent experiments, they also have the potential to make
generalization to unseen data more challenging. | ['Sabrina Stehwien', 'Antje Schweitzer', 'Ngoc Thang Vu'] | 2018-05-14 | null | null | null | null | ['cross-corpus'] | ['computer-vision'] | [-4.45950657e-01 -2.89807111e-01 -2.19137266e-01 -5.01548111e-01
-5.57054102e-01 -7.71926105e-01 4.25695509e-01 4.16710436e-01
-1.02571917e+00 4.40672636e-01 5.10787904e-01 -2.83040494e-01
2.59113610e-01 -8.07589948e-01 -3.84928375e-01 -2.45263904e-01
-2.58701712e-01 1.65892392e-01 1.72781751e-01 -5.41949928e-01
5.71498163e-02 4.68612194e-01 -1.29907823e+00 -1.69918746e-01
2.00093180e-01 6.04113936e-01 2.63897534e-02 5.49830139e-01
-2.87649900e-01 2.04301462e-01 -9.22196329e-01 -4.97859955e-01
-4.01742384e-02 3.47105674e-02 -6.69341981e-01 -4.03090298e-01
5.13600290e-01 6.77474067e-02 -2.26109579e-01 9.93390858e-01
6.67466521e-01 4.81763214e-01 4.58303005e-01 -7.05877960e-01
-7.93067217e-01 7.30591238e-01 -2.18522012e-01 7.93497860e-01
2.25664303e-01 -3.94236706e-02 1.46156478e+00 -9.38629210e-01
4.50329155e-01 1.28972435e+00 8.48660827e-01 2.11530298e-01
-1.15149891e+00 -9.23639357e-01 1.09975785e-01 -3.21009420e-02
-1.36427712e+00 -3.12811971e-01 9.40478861e-01 -2.05342323e-01
1.31408131e+00 2.07672059e-03 4.77690578e-01 1.16032183e+00
3.16252172e-01 4.05312896e-01 8.95653486e-01 -5.62510669e-01
4.38295715e-02 3.40721011e-02 3.15189987e-01 3.38532269e-01
8.62879902e-02 4.38098252e-01 -6.35032356e-01 -1.89754903e-01
5.60837030e-01 -5.42663991e-01 -1.70719475e-01 9.77907404e-02
-1.01740503e+00 1.33913875e+00 3.19048226e-01 6.37719989e-01
-3.59619409e-01 1.57131091e-01 7.41336942e-01 1.54195249e-01
4.40697819e-01 9.50748861e-01 -8.12093616e-01 -3.69113892e-01
-7.41511762e-01 4.92416769e-01 7.51719773e-01 2.42401168e-01
6.09506130e-01 2.77719468e-01 2.50431448e-01 1.03807890e+00
1.53874412e-01 3.13818961e-01 1.01893210e+00 -3.96133304e-01
1.49668962e-01 -4.94803861e-02 -1.20725162e-01 -8.20830524e-01
-7.79359758e-01 -2.56039828e-01 1.15424526e-04 -1.69214364e-02
3.61936271e-01 -5.29051185e-01 -9.43828464e-01 2.12151241e+00
1.64271072e-01 -3.61414962e-02 1.28855437e-01 7.60583401e-01
9.32267845e-01 6.21338785e-01 5.87414980e-01 -9.20002833e-02
1.63168383e+00 -5.85690498e-01 -1.10365415e+00 -4.63548005e-01
5.27770519e-01 -1.15150893e+00 1.34205723e+00 1.93821117e-01
-9.11983430e-01 -8.33362043e-01 -1.26813209e+00 -2.97560599e-02
-8.46926808e-01 -4.54865932e-01 5.54225504e-01 1.00534940e+00
-6.66914463e-01 3.73326778e-01 -6.35661662e-01 -3.60326886e-01
-1.79580823e-01 2.56178170e-01 -3.31210196e-01 4.20452327e-01
-1.78494084e+00 1.24007154e+00 7.50695527e-01 -3.86983961e-01
8.95451661e-03 -7.50153005e-01 -1.08159137e+00 1.07870527e-01
-2.44090185e-02 7.67423809e-02 1.55574596e+00 -6.95637882e-01
-1.38279033e+00 5.48494697e-01 1.98628291e-01 -3.12658012e-01
-2.94077873e-01 -3.71447474e-01 -9.52602446e-01 -4.39333245e-02
-1.01567730e-01 9.53828096e-01 4.37130749e-01 -5.35594642e-01
-6.10215485e-01 -5.21230549e-02 -1.47264823e-01 1.50548264e-01
-5.23199141e-01 2.28826150e-01 -8.71967226e-02 -9.66693044e-01
-2.08489448e-01 -1.06769121e+00 -1.74484700e-02 -6.02206945e-01
-6.79858327e-02 -6.99538648e-01 8.06545913e-01 -6.02110445e-01
1.10910738e+00 -2.39224696e+00 -2.31304526e-01 5.51923513e-02
-2.88290799e-01 3.64659488e-01 -1.74334690e-01 4.18477416e-01
-1.08717255e-01 2.64863968e-01 2.85893947e-01 -1.30400479e-01
9.77889299e-02 3.64043981e-01 -2.80268937e-01 3.14682394e-01
5.59614480e-01 7.90321946e-01 -6.83056533e-01 -3.14831465e-01
3.03853810e-01 7.63009369e-01 -5.13181031e-01 1.60779133e-02
-7.30899945e-02 -5.91647960e-02 9.78422761e-02 2.05597281e-01
4.07069176e-01 4.32812989e-01 2.56764263e-01 -7.55129755e-02
-2.47868076e-01 9.72438037e-01 -1.07836962e+00 1.41280746e+00
-6.63752079e-01 8.38533878e-01 -1.27137139e-01 -6.46104991e-01
1.01196241e+00 5.23813128e-01 -4.26124297e-02 -5.86290658e-01
1.68472052e-01 1.74313739e-01 5.43540537e-01 -1.35048807e-01
9.68836665e-01 -6.44733250e-01 -4.14103180e-01 1.97044045e-01
3.29293460e-01 -1.49212703e-01 2.09416360e-01 -2.15835303e-01
7.66377628e-01 -3.71431500e-01 3.71700525e-01 -5.40853143e-01
7.38320425e-02 -1.95436850e-01 6.86436713e-01 3.95044953e-01
-3.28456253e-01 3.92556608e-01 1.98381707e-01 -2.78122753e-01
-9.23386514e-01 -1.25796890e+00 -5.99302948e-01 1.41016579e+00
-1.89132914e-01 -6.18548393e-01 -4.04867828e-01 -5.14852583e-01
7.28429481e-02 9.86283541e-01 -6.59068108e-01 -8.97380784e-02
-6.02893412e-01 -7.55885184e-01 7.10069835e-01 9.99801397e-01
-5.39537333e-02 -1.38688898e+00 -5.99742293e-01 6.00243151e-01
1.10297114e-01 -1.26825523e+00 -6.54696763e-01 7.13628590e-01
-3.67572099e-01 -7.08739698e-01 -3.67872149e-01 -9.41290677e-01
-1.44529164e-01 -2.01379851e-01 1.32212198e+00 -1.74676225e-01
-1.56481341e-01 4.13530588e-01 -7.88798153e-01 -9.28129435e-01
-4.72190499e-01 3.78201216e-01 3.59424531e-01 -4.20177728e-01
1.09498215e+00 -3.81030738e-01 -1.42296195e-01 3.71599719e-02
-9.50557292e-01 -8.11802685e-01 3.78869206e-01 9.27824080e-01
4.88206804e-01 4.62396778e-02 9.49997783e-01 -6.64066434e-01
1.06307518e+00 -2.35335797e-01 -4.57147747e-01 -3.19563091e-01
-4.99503613e-01 7.17981309e-02 4.04474646e-01 -7.57948279e-01
-6.17340982e-01 -1.00841381e-01 -5.41939080e-01 -2.61564195e-01
-2.68153012e-01 9.10364270e-01 3.31964381e-02 1.76353782e-01
5.90202272e-01 -3.15001458e-01 -2.31946141e-01 -5.73817670e-01
5.12645841e-01 6.67873263e-01 4.55630839e-01 -2.53369480e-01
5.58503866e-01 -3.00092667e-01 -3.54268163e-01 -1.20344031e+00
-7.60948777e-01 -5.09280384e-01 -4.35427219e-01 2.23675922e-01
1.28064013e+00 -8.17402661e-01 -2.27018550e-01 4.04989123e-02
-1.06894839e+00 -1.12561911e-01 -3.51812899e-01 9.04522240e-01
-2.22124502e-01 1.77695438e-01 -7.49292552e-01 -4.71136302e-01
-7.23889023e-02 -1.20822144e+00 6.87696815e-01 2.88110405e-01
-9.48713064e-01 -1.31739485e+00 5.22231519e-01 -1.58608511e-01
3.98973197e-01 -9.07889307e-02 9.57294822e-01 -1.18831348e+00
2.83528388e-01 -1.64714441e-01 3.19477111e-01 5.52734852e-01
3.02372605e-01 4.18814458e-02 -1.26255393e+00 -1.86896563e-01
-1.76184207e-01 -4.02349412e-01 7.97601700e-01 4.48293895e-01
6.52773023e-01 1.73212364e-01 1.02338299e-01 3.14742893e-01
1.24517798e+00 2.65134454e-01 4.64236975e-01 4.58993256e-01
5.36939681e-01 4.81881112e-01 4.69395787e-01 3.89452018e-02
1.35982381e-02 7.22389162e-01 8.19783565e-03 -1.00913279e-01
-3.62849794e-02 -2.08012134e-01 3.28761846e-01 1.18248010e+00
5.16421795e-01 -1.31103009e-01 -1.00062084e+00 8.93330038e-01
-1.16448152e+00 -7.20859289e-01 9.67644155e-02 1.97645247e+00
1.01477826e+00 2.94721156e-01 3.77396792e-01 3.00249517e-01
6.33718789e-01 4.81304795e-01 7.35495314e-02 -1.21066523e+00
-1.54164210e-01 9.34865594e-01 5.12079716e-01 6.33025527e-01
-1.38738000e+00 1.20136702e+00 7.29895449e+00 5.83391249e-01
-1.26433241e+00 9.48874503e-02 2.55697340e-01 1.26609236e-01
-1.08695649e-01 -4.96945769e-01 -9.67306793e-01 1.84739262e-01
1.40080321e+00 -1.05553426e-01 1.39816314e-01 7.63998806e-01
-2.71107793e-01 3.53736803e-02 -1.08120906e+00 6.38203561e-01
7.63377361e-03 -9.42626297e-01 -3.42913121e-01 4.15401720e-02
4.29084808e-01 3.04117382e-01 2.53233463e-01 5.41306674e-01
2.90580362e-01 -1.20469439e+00 5.05993962e-01 -3.41665864e-01
5.74767530e-01 -1.18886125e+00 9.81403887e-01 -1.56058058e-01
-1.15449202e+00 2.13626802e-01 -5.17829955e-01 -2.32166126e-01
2.60270685e-01 2.94078529e-01 -1.15055335e+00 -3.75801474e-02
7.55737782e-01 1.10188678e-01 -4.04642910e-01 7.26476312e-01
-3.45374286e-01 9.45229709e-01 -3.87502491e-01 -3.11756760e-01
4.75308418e-01 1.58063859e-01 3.62275988e-01 1.60195875e+00
2.17608623e-02 -1.21915974e-01 3.00956517e-01 5.79771578e-01
-1.77451558e-02 4.78103489e-01 -6.29427731e-01 -4.10982966e-01
6.72277749e-01 1.18439949e+00 -6.97774291e-01 -7.19736814e-02
-8.62686634e-01 5.54985881e-01 5.05162954e-01 1.59673125e-01
-7.03155041e-01 -7.76539683e-01 1.27960563e+00 -6.41458556e-02
7.86595941e-01 -5.15403092e-01 -1.83532044e-01 -5.49780011e-01
-2.42757291e-01 -6.57395065e-01 2.44546250e-01 -4.12110955e-01
-1.50906682e+00 5.68014741e-01 -4.30155955e-02 -5.91903269e-01
-4.36698675e-01 -8.99111092e-01 -6.53179586e-01 1.20146883e+00
-1.53833461e+00 -7.40133941e-01 2.91366518e-01 1.86604813e-01
6.14399612e-01 -1.14505924e-01 1.27632868e+00 1.67791247e-01
-3.15018773e-01 7.66670346e-01 -1.13320366e-01 6.39452159e-01
9.83042657e-01 -1.42231596e+00 9.65669155e-01 5.17527521e-01
7.92521000e-01 5.97000897e-01 8.06691706e-01 -5.45414269e-01
-8.60288382e-01 -8.79428566e-01 1.05776989e+00 -5.27885556e-01
9.22356784e-01 -4.40390885e-01 -1.12779117e+00 8.13296735e-01
5.06307960e-01 6.35422617e-02 1.18711185e+00 1.00526488e+00
-5.25414944e-01 2.64225423e-01 -7.72157073e-01 4.86110061e-01
4.84743237e-01 -7.84742177e-01 -1.13955045e+00 2.40331572e-02
9.98924494e-01 -3.40682268e-01 -9.81491923e-01 3.29646587e-01
4.62116599e-01 -5.55652320e-01 8.67562830e-01 -7.21780181e-01
1.81405693e-02 -1.27446447e-02 -3.59461546e-01 -1.89324999e+00
-5.30299306e-01 -2.65522152e-01 3.29776406e-01 1.41089296e+00
8.22039723e-01 -4.69517648e-01 4.83038366e-01 3.01960409e-01
-1.30631670e-01 -3.00028473e-01 -1.08785355e+00 -1.09102714e+00
6.54450357e-01 -6.01547897e-01 5.44758677e-01 1.08849680e+00
3.44752133e-01 7.97041655e-01 -7.85883144e-03 2.03354299e-01
-1.18211709e-01 -3.57837826e-01 2.32165709e-01 -1.34696281e+00
-2.13429511e-01 -4.28417832e-01 -8.79302442e-01 -5.01276016e-01
5.27927160e-01 -7.53832579e-01 1.48014739e-01 -8.66042674e-01
-4.61713076e-01 -3.57527316e-01 -6.94631875e-01 3.47387820e-01
-4.92397904e-01 4.01561946e-01 3.15405101e-01 -3.92375380e-01
-6.81098923e-02 4.50595587e-01 6.33142591e-01 6.58916384e-02
-4.05670971e-01 -3.09579909e-01 -4.49553072e-01 6.35732353e-01
1.09327269e+00 -4.71283704e-01 -3.13347578e-01 -2.99566597e-01
1.41216749e-02 -3.33994985e-01 -2.24297106e-01 -9.72876668e-01
-2.15864748e-01 4.93364222e-02 4.73312110e-01 -2.94761956e-01
6.02293730e-01 -4.23001349e-01 -3.43017220e-01 1.45983607e-01
-4.13435161e-01 7.80666709e-01 8.64627004e-01 3.69820833e-01
-4.60663557e-01 -3.40685844e-01 8.10387254e-01 6.83531016e-02
-8.10467303e-01 -2.06685603e-01 -5.28295875e-01 3.30788106e-01
5.23402572e-01 9.96563137e-02 -3.14481184e-02 -2.93531895e-01
-4.75687474e-01 -2.81672686e-01 1.95512295e-01 8.21604073e-01
1.63140446e-01 -1.36245298e+00 -5.83978236e-01 3.86913866e-01
8.33819360e-02 -5.61848164e-01 -2.90628999e-01 3.28835368e-01
-2.90570438e-01 4.74102974e-01 -1.88353688e-01 -3.85906935e-01
-1.23774183e+00 7.41205096e-01 2.45926067e-01 4.08730060e-02
-4.87414598e-01 1.03278506e+00 2.64184535e-01 -5.41050613e-01
2.59982854e-01 -4.14443970e-01 -4.17512625e-01 2.91329861e-01
4.56741393e-01 -1.16932370e-01 2.44632527e-01 -9.02868629e-01
-4.84645247e-01 2.79891253e-01 -4.53321606e-01 -5.28241038e-01
1.10915482e+00 2.30887264e-01 4.83365387e-01 8.82598639e-01
1.22186589e+00 5.06663740e-01 -7.30041623e-01 -1.13924794e-01
2.52404600e-01 -9.25756469e-02 5.66652060e-01 -7.73341656e-01
-7.85013199e-01 8.80663216e-01 8.49808037e-01 4.04491067e-01
6.83677733e-01 -7.92494044e-02 9.31657970e-01 2.27607206e-01
-2.44104460e-01 -1.43898726e+00 -2.14746714e-01 7.50516713e-01
4.33414668e-01 -1.11671185e+00 -2.55591363e-01 -2.95876592e-01
-6.95792615e-01 1.14854157e+00 5.14330626e-01 -2.38888413e-01
9.97446835e-01 4.57445860e-01 5.62860966e-01 -9.43946689e-02
-5.58584511e-01 -3.35250407e-01 1.26419812e-01 6.19823098e-01
1.04759395e+00 3.37004811e-01 -5.76076865e-01 6.03006423e-01
-7.75037110e-01 -4.96533543e-01 4.76481497e-01 8.16916466e-01
-4.57710534e-01 -1.51598680e+00 -3.90902370e-01 1.66552410e-01
-8.85186493e-01 -3.76408070e-01 -4.53002810e-01 1.20395446e+00
-1.42440675e-02 9.86179650e-01 5.33388793e-01 -4.86297756e-01
4.33013260e-01 6.37996137e-01 3.15384686e-01 -7.32118547e-01
-7.89388657e-01 2.06742138e-01 4.57062483e-01 -1.07159123e-01
-2.94448137e-01 -7.91891873e-01 -1.23493946e+00 -1.41908452e-01
-6.13426685e-01 4.08001751e-01 8.41269851e-01 8.85590792e-01
-5.85899176e-03 6.77188993e-01 3.38842779e-01 -6.31203115e-01
-5.74477732e-01 -1.25382388e+00 -9.01183546e-01 4.17683125e-01
3.27022254e-01 -7.70083725e-01 -3.60465080e-01 -2.58754522e-01] | [10.738890647888184, 9.500494956970215] |
d1cdccad-5662-42da-aab2-b52643814e49 | limits-of-an-ai-program-for-solving-college | 2208.06906 | null | https://arxiv.org/abs/2208.06906v1 | https://arxiv.org/pdf/2208.06906v1.pdf | Limits of an AI program for solving college math problems | Drori et al. (2022) report that "A neural network solves, explains, and generates university math problems by program synthesis and few-shot learning at human level ... [It] automatically answers 81\% of university-level mathematics problems." The system they describe is indeed impressive; however, the above description is very much overstated. The work of solving the problems is done, not by a neural network, but by the symbolic algebra package Sympy. Problems of various formats are excluded from consideration. The so-called "explanations" are just rewordings of lines of code. Answers are marked as correct that are not in the form specified in the problem. Most seriously, it seems that in many cases the system uses the correct answer given in the test corpus to guide its path to solving the problem. | ['Ernest Davis'] | 2022-08-14 | null | null | null | null | ['program-synthesis'] | ['computer-code'] | [ 1.39116682e-02 4.94282156e-01 -4.85010408e-02 -2.71112144e-01
-8.11389029e-01 -8.13120186e-01 4.61345077e-01 6.90915063e-02
9.00235102e-02 1.15372968e+00 2.37122804e-01 -9.61240172e-01
-3.33178878e-01 -1.05414116e+00 -6.49028122e-01 -2.45758951e-01
5.00434041e-01 5.23061335e-01 9.98492315e-02 -6.15076721e-01
9.08084452e-01 3.61691087e-01 -1.75853598e+00 3.05996567e-01
1.22168803e+00 5.54229096e-02 1.27978504e-01 9.47735190e-01
-6.84823692e-01 1.39095414e+00 -6.96056545e-01 -6.38567090e-01
7.46449009e-02 -7.78471291e-01 -1.25134027e+00 -3.31510380e-02
5.42535186e-01 -2.57038176e-01 -4.63929206e-01 1.37287128e+00
-1.25830108e-02 4.15914446e-01 6.68812752e-01 -1.19638264e+00
-7.14978814e-01 9.48577464e-01 -1.99573077e-02 4.09004420e-01
5.91693819e-01 3.97069335e-01 1.01640594e+00 -8.42081726e-01
7.07689881e-01 8.66813421e-01 7.83452272e-01 5.74445426e-01
-1.38531566e+00 -5.61390758e-01 -3.32061499e-01 3.36975932e-01
-1.29042327e+00 -3.28064173e-01 5.25364518e-01 -8.71059597e-01
1.41958201e+00 4.93776381e-01 7.59508431e-01 5.96369684e-01
4.54792887e-01 3.74142349e-01 7.01822937e-01 -5.74036837e-01
4.07463729e-01 2.79305786e-01 3.68841738e-01 8.91410947e-01
2.94370651e-01 -1.50236413e-02 -1.18564762e-01 -3.06892961e-01
7.91206121e-01 -2.02670589e-01 -1.64516345e-01 -1.06116861e-01
-9.58463788e-01 1.19244790e+00 2.68364232e-02 6.99273407e-01
-1.43269897e-01 1.87422320e-01 3.73279899e-01 4.26478058e-01
2.76204105e-02 1.13225174e+00 -4.23611104e-01 -3.04779559e-01
-1.09589100e+00 6.95178628e-01 1.18768716e+00 1.17484391e+00
6.18404031e-01 5.70230246e-01 1.97040185e-01 5.29804707e-01
-8.68629292e-02 -1.99650824e-02 6.69079900e-01 -1.68936658e+00
3.95243853e-01 6.01170480e-01 1.76836923e-01 -1.05527210e+00
-1.28490105e-01 -1.00604028e-01 -3.35132092e-01 4.90187347e-01
6.83093488e-01 -4.58503932e-01 -5.12626231e-01 1.39216650e+00
-2.85884321e-01 1.26014471e-01 1.89002529e-01 5.70290744e-01
1.03488743e+00 1.11319852e+00 2.20982760e-01 -3.48089308e-01
7.46845365e-01 -9.47387755e-01 -7.60686874e-01 -1.46864325e-01
7.79745281e-01 -5.93529701e-01 8.94361556e-01 5.23387253e-01
-1.59012532e+00 -7.81165779e-01 -1.11159849e+00 -4.07772465e-03
-3.31298411e-01 -1.34188890e-01 7.88447618e-01 4.28034365e-01
-1.06424820e+00 9.94761705e-01 -2.51623601e-01 -7.86165744e-02
-4.10461202e-02 2.16538101e-01 3.60896848e-02 -1.69436224e-02
-9.72591639e-01 9.94338751e-01 5.40176809e-01 -3.52412611e-01
-6.00519300e-01 -8.37803960e-01 -1.11524522e+00 5.53601980e-01
4.25770015e-01 -5.09389162e-01 1.63468492e+00 -1.06043887e+00
-1.28687453e+00 8.15241396e-01 -2.67079234e-01 -2.74120241e-01
4.16984767e-01 1.16301574e-01 -3.93225074e-01 7.65696988e-02
3.93254161e-01 3.27860773e-01 3.19279164e-01 -9.74553347e-01
-7.04729855e-01 1.05774060e-01 3.34155381e-01 -2.24754110e-01
2.23207906e-01 2.30905056e-01 8.90554413e-02 -5.76168120e-01
1.32383704e-01 -6.03308797e-01 -5.10946274e-01 -5.45202315e-01
-1.57919601e-01 -6.98341310e-01 1.93292946e-01 -6.98670328e-01
1.40350950e+00 -1.90461063e+00 9.94370654e-02 2.44497553e-01
3.32828462e-01 3.04691851e-01 7.07231089e-02 3.27311009e-01
-6.78616166e-01 5.40958464e-01 -1.53206110e-01 5.33264279e-01
4.03244674e-01 -1.90291852e-02 -6.80679560e-01 8.99351016e-02
2.11005494e-01 8.64308119e-01 -9.50348794e-01 -4.07655716e-01
2.50776559e-01 -1.24481939e-01 -6.86058700e-01 1.52238905e-01
-6.01913512e-01 -6.00866135e-03 -3.85232478e-01 2.65823841e-01
1.79716691e-01 -3.07072073e-01 2.84294784e-01 5.70003629e-01
-2.97313482e-01 2.61715353e-01 -1.29754269e+00 1.54913509e+00
-2.14178964e-01 9.44284081e-01 -4.77239877e-01 -1.16331005e+00
1.01079559e+00 6.82579875e-01 -2.00066008e-02 -2.30762631e-01
9.06436071e-02 3.67589861e-01 3.06165218e-01 -1.03272533e+00
5.88660598e-01 -2.34687477e-01 -2.28983343e-01 5.27652562e-01
5.44153303e-02 -5.67552149e-01 6.27244592e-01 3.71163815e-01
1.20172632e+00 3.45369101e-01 5.45415938e-01 -2.38976777e-01
6.56496644e-01 8.10795188e-01 4.76147383e-01 1.07621920e+00
-7.08910152e-02 5.06867349e-01 9.22765553e-01 -7.67771959e-01
-1.38767874e+00 -8.73351336e-01 1.57375187e-01 9.97816682e-01
-4.42770302e-01 -5.59639633e-01 -8.47347498e-01 -2.50604779e-01
-2.27260143e-01 1.45324695e+00 -2.72617877e-01 8.90205279e-02
-5.86683512e-01 -7.66207650e-03 3.79561752e-01 4.33654875e-01
1.97509363e-01 -1.38721633e+00 -4.48321730e-01 6.30470872e-01
-1.11626014e-01 -5.81448019e-01 7.40547404e-02 2.52192885e-01
-8.15507650e-01 -1.10861123e+00 -5.89518964e-01 -9.39667463e-01
5.46411097e-01 -2.17356589e-02 1.15185213e+00 4.10415024e-01
-3.53488505e-01 1.81439847e-01 -3.74475941e-02 -2.51984835e-01
-7.09888816e-01 -2.19831079e-01 -2.08152816e-01 -9.21610713e-01
6.66873336e-01 -6.52241468e-01 2.91357577e-01 -3.87258440e-01
-4.54871595e-01 -1.26265317e-01 1.08720779e-01 8.44568253e-01
1.52921677e-01 4.98760372e-01 5.82457840e-01 -1.28087819e+00
8.49959373e-01 -5.75045466e-01 -7.25597620e-01 3.15884978e-01
-4.03424770e-01 1.15543596e-01 9.01491225e-01 -2.85537809e-01
-1.05381703e+00 3.82169299e-02 -1.49478361e-01 -2.05312312e-01
-7.08393455e-01 5.24986506e-01 5.46990521e-03 7.97394216e-02
1.05920053e+00 2.60736883e-01 -3.17711323e-01 -1.47953369e-02
1.39492169e-01 2.35916853e-01 6.94027960e-01 -8.19192290e-01
8.83440495e-01 -6.05849981e-01 -1.57488838e-01 -6.24436498e-01
-7.75336027e-01 -3.18452455e-02 -2.66345888e-01 -1.37520552e-01
6.80233955e-01 -6.26865387e-01 -5.94391942e-01 -2.00260490e-01
-1.56308424e+00 -3.15263420e-01 -4.63599861e-01 2.76985794e-01
-9.62335706e-01 1.36393681e-01 -7.42425501e-01 -8.19406509e-01
7.47268200e-02 -1.19048703e+00 9.41726491e-02 5.31046927e-01
-1.06319618e+00 -8.34878504e-01 9.98062664e-04 3.04490089e-01
2.24849463e-01 2.14874581e-01 1.51107836e+00 -1.07483077e+00
-5.00684083e-01 -2.79389024e-01 -1.53251469e-01 1.29826859e-01
-3.92784655e-01 3.38954598e-01 -6.87204719e-01 3.81016821e-01
2.85413653e-01 -2.72756428e-01 1.85370445e-01 3.25142264e-01
1.26394510e+00 -5.89889348e-01 5.33090718e-02 2.10445464e-01
1.62386858e+00 4.69631076e-01 6.03177190e-01 2.33182311e-01
3.23978812e-01 6.57971084e-01 2.04198539e-01 2.37358034e-01
1.03845119e-01 9.20327902e-02 -1.21175379e-01 5.93024611e-01
1.94845140e-01 -1.76152319e-01 2.98015535e-01 7.08866239e-01
-6.32190108e-02 5.81951104e-02 -1.37103355e+00 6.21005654e-01
-1.99539804e+00 -1.55066240e+00 -7.17153370e-01 1.64729369e+00
8.85537684e-01 4.22827303e-01 -6.36325851e-02 1.61233589e-01
8.09724510e-01 -2.10046783e-01 -2.57773250e-01 -1.02005148e+00
3.60988617e-01 7.61115015e-01 1.09024886e-02 8.50152373e-01
-5.74676573e-01 9.73420441e-01 7.55702543e+00 8.05602670e-01
-5.18821836e-01 -2.12614134e-01 3.45938176e-01 1.88129619e-01
-4.68250483e-01 1.68967202e-01 -6.30253017e-01 3.17996413e-01
1.37790966e+00 -8.87301326e-01 6.37468755e-01 1.27022934e+00
1.33085415e-01 -1.20721146e-01 -1.24546528e+00 5.81327856e-01
5.25496677e-02 -1.88235950e+00 7.44342282e-02 -2.86093831e-01
9.12920654e-01 -6.19333267e-01 -2.80531555e-01 8.74139071e-01
7.97541916e-01 -1.34699154e+00 6.74114943e-01 6.95432305e-01
3.45736444e-01 -1.23451436e+00 7.09474802e-01 8.14785242e-01
-8.95306170e-01 -2.68447906e-01 -5.53772509e-01 -7.52072275e-01
-2.90026009e-01 7.77240470e-02 -7.37153769e-01 2.12642491e-01
2.58117497e-01 3.22521448e-01 -6.42026782e-01 1.21093786e+00
-5.23506820e-01 6.13598764e-01 2.35885620e-01 -4.37252104e-01
3.58185649e-01 -3.19943912e-02 4.15117085e-01 1.12040722e+00
5.30147314e-01 8.75887334e-01 2.15612262e-01 1.48502421e+00
3.39827180e-01 7.15700909e-02 -9.83434737e-01 -4.08895582e-01
5.07599771e-01 9.55555320e-01 -6.31358087e-01 -9.14930820e-01
-2.42024273e-01 4.48194712e-01 9.71750095e-02 3.78472269e-01
-6.42715216e-01 -1.15796542e+00 3.10604423e-01 5.36146900e-03
1.62793338e-01 -5.19008972e-02 -7.24060893e-01 -1.06594229e+00
-5.28297067e-01 -1.20268500e+00 2.47575138e-02 -1.23428392e+00
-1.10649264e+00 3.02746654e-01 5.55650257e-02 -7.51173615e-01
-8.54877949e-01 -6.37257695e-01 -1.18995833e+00 1.16301334e+00
-7.09548354e-01 -7.03134164e-02 1.89962629e-02 3.28653574e-01
8.57392251e-01 -4.69155461e-01 1.02409911e+00 -1.69808057e-03
-3.89260352e-01 1.31094739e-01 -2.04925850e-01 2.20619574e-01
8.39633718e-02 -1.33794081e+00 2.33909786e-01 8.05820107e-01
-7.63216317e-02 9.61670160e-01 1.29335570e+00 -6.55186474e-01
-1.13870537e+00 -6.95148110e-01 1.56196618e+00 -4.43573922e-01
9.73949850e-01 3.38465124e-01 -1.17531037e+00 7.47742653e-01
4.11621839e-01 -5.84068060e-01 7.80206442e-01 -1.81678236e-01
-1.65825449e-02 4.24520075e-01 -1.00589395e+00 6.87813938e-01
5.38964689e-01 -5.31335533e-01 -1.38475287e+00 6.36343062e-01
7.84170210e-01 -3.95916760e-01 -5.85963130e-01 -2.49138147e-01
1.08372726e-01 -9.36413884e-01 7.71074414e-01 -1.09982598e+00
1.26948941e+00 -1.12890787e-01 -2.09445860e-02 -1.21738935e+00
-5.39571941e-01 -6.97314262e-01 8.97588134e-02 1.05111718e+00
4.45220053e-01 -1.51160568e-01 8.83222461e-01 1.08864832e+00
-3.29683214e-01 -3.68157238e-01 -5.14555395e-01 -4.67399985e-01
4.35095906e-01 -5.59357464e-01 4.94019985e-01 1.29728329e+00
7.10798025e-01 4.45890605e-01 2.41686832e-02 -2.14374304e-01
5.98303080e-01 3.01673770e-01 6.51204348e-01 -1.37057030e+00
-3.95437419e-01 -7.06385016e-01 4.16130871e-02 -4.63664800e-01
5.80483735e-01 -1.05533957e+00 2.48220786e-01 -1.52089798e+00
-2.16349158e-02 1.45491352e-02 3.01612854e-01 3.88016641e-01
1.44763321e-01 -2.00585216e-01 1.12493485e-01 -1.00021020e-01
-2.98071355e-01 -1.59385696e-01 1.11738265e+00 -1.02617472e-01
-1.06405914e-01 -1.39809608e-01 -9.19471085e-01 1.30258691e+00
8.59039307e-01 -4.27068979e-01 -3.24556142e-01 -1.55263051e-01
3.57009917e-01 8.59499931e-01 3.76759320e-01 -1.15938032e+00
6.21613681e-01 -6.55452311e-01 5.61295807e-01 -2.96036601e-01
-8.28830227e-02 -6.91646814e-01 3.73736441e-01 7.14611113e-01
-8.16784084e-01 3.06921810e-01 3.06902856e-01 1.10044502e-01
2.32164077e-02 -1.26703715e+00 6.41198933e-01 -8.40319812e-01
-1.03206599e+00 -3.36487353e-01 -9.35203850e-01 1.35682732e-01
1.04901254e+00 -3.59116673e-01 -1.64604738e-01 -5.23298383e-01
-9.96829569e-01 1.90766733e-02 1.01254299e-01 -6.21212870e-02
6.55258119e-01 -1.32108128e+00 -4.86378282e-01 1.08873341e-02
-1.88084647e-01 -1.80976048e-01 1.00914717e-01 3.15885842e-01
-7.45139241e-01 6.22899771e-01 -4.08707380e-01 2.76256297e-02
-8.90542924e-01 5.72907984e-01 2.67693132e-01 -4.70549054e-02
-7.57347524e-01 7.75070131e-01 -3.44899684e-01 -5.23866951e-01
3.59029502e-01 -8.43235850e-02 -3.99330437e-01 -8.23414102e-02
7.84102857e-01 1.97163582e-01 -3.55746746e-01 -7.35101104e-02
1.85974360e-01 1.22696862e-01 9.78459939e-02 -6.81406409e-02
1.67014837e+00 3.06929708e-01 -2.56695181e-01 7.19986141e-01
7.99497008e-01 -1.40076801e-01 -5.19226968e-01 -2.35262960e-01
2.33878598e-01 -3.43864232e-01 -3.63897055e-01 -8.38059068e-01
-5.04040718e-01 1.02706242e+00 -1.24060333e-01 4.00840104e-01
5.01690090e-01 -3.91258687e-01 5.44947088e-01 9.51966584e-01
2.81435639e-01 -1.17451811e+00 4.63700108e-02 9.77728128e-01
7.23716319e-01 -9.58384097e-01 4.21498604e-02 -1.18014038e-01
-5.54232299e-01 1.69872284e+00 9.73358333e-01 -6.21542037e-01
1.20970868e-01 2.47607052e-01 -4.18126255e-01 -2.38516554e-01
-8.95602942e-01 2.45528758e-01 -5.00319786e-02 3.25091481e-01
6.27375662e-01 -2.08779991e-01 -4.31362659e-01 9.16060627e-01
-7.29502201e-01 3.81266266e-01 1.11995554e+00 9.34585810e-01
-1.18120325e+00 -6.65066540e-01 -6.73322439e-01 3.46086860e-01
-2.81552285e-01 -1.57825485e-01 -3.37840647e-01 8.46209049e-01
2.10798830e-01 8.26705873e-01 -2.64791753e-02 -2.84573764e-01
2.41716847e-01 7.08649814e-01 3.82317245e-01 -1.09683311e+00
-8.26306701e-01 -3.01072001e-01 1.46336764e-01 -2.16279507e-01
2.00590745e-01 -5.07965147e-01 -1.50849569e+00 -1.00592208e+00
1.64361611e-01 5.83780587e-01 2.00893447e-01 1.20545769e+00
-2.72335619e-01 7.60185003e-01 9.04370323e-02 -6.43361330e-01
-9.19889390e-01 -6.03941381e-01 -4.67148274e-01 -1.11781128e-01
6.62150383e-02 -2.29345664e-01 -3.99020284e-01 2.10390612e-01] | [9.245269775390625, 7.18405294418335] |
9741e9c3-3dc1-4fa9-a252-78868c4ebdd1 | double-check-your-state-before-trusting-it | 2206.07989 | null | https://arxiv.org/abs/2206.07989v2 | https://arxiv.org/pdf/2206.07989v2.pdf | Double Check Your State Before Trusting It: Confidence-Aware Bidirectional Offline Model-Based Imagination | The learned policy of model-free offline reinforcement learning (RL) methods is often constrained to stay within the support of datasets to avoid possible dangerous out-of-distribution actions or states, making it challenging to handle out-of-support region. Model-based RL methods offer a richer dataset and benefit generalization by generating imaginary trajectories with either trained forward or reverse dynamics model. However, the imagined transitions may be inaccurate, thus downgrading the performance of the underlying offline RL method. In this paper, we propose to augment the offline dataset by using trained bidirectional dynamics models and rollout policies with double check. We introduce conservatism by trusting samples that the forward model and backward model agree on. Our method, confidence-aware bidirectional offline model-based imagination, generates reliable samples and can be combined with any model-free offline RL method. Experimental results on the D4RL benchmarks demonstrate that our method significantly boosts the performance of existing model-free offline RL algorithms and achieves competitive or better scores against baseline methods. | ['Zongqing Lu', 'Xiu Li', 'Jiafei Lyu'] | 2022-06-16 | null | null | null | null | ['d4rl'] | ['robots'] | [-3.44521135e-01 4.07978445e-01 -9.39625084e-01 -2.18642280e-01
-8.80532742e-01 -9.18376088e-01 8.06017518e-01 -2.66981542e-01
-4.10347462e-01 1.16497862e+00 2.84249276e-01 -6.92572832e-01
1.21967278e-01 -5.36445558e-01 -1.11506879e+00 -6.83112919e-01
-2.10598156e-01 7.79046834e-01 -1.21280039e-02 -3.14426750e-01
1.78598259e-02 2.86993176e-01 -1.17049873e+00 1.88809708e-01
8.55472863e-01 8.87683570e-01 -5.78950346e-02 5.29431760e-01
1.43177152e-01 1.45716524e+00 -5.00226915e-01 -1.00515045e-01
6.44336641e-01 -5.95926940e-01 -5.61171532e-01 -2.02220708e-01
-4.79455143e-02 -8.89770567e-01 -7.02411592e-01 8.74996066e-01
2.66974628e-01 3.69095296e-01 3.89127582e-01 -1.43011951e+00
-5.49219131e-01 8.05040836e-01 -3.01428229e-01 -4.91593108e-02
2.77073622e-01 8.10872555e-01 7.55585670e-01 -4.56008106e-01
7.21447349e-01 1.29854512e+00 4.67137724e-01 8.45557332e-01
-1.33723426e+00 -7.76333213e-01 6.74538136e-01 2.04831392e-01
-8.88151526e-01 -5.17150104e-01 6.78660095e-01 -1.86012164e-01
9.71794665e-01 -9.73947495e-02 1.03310466e+00 1.79942548e+00
4.30919714e-02 1.19250786e+00 1.64317727e+00 -5.94180338e-02
7.12018609e-01 4.14339155e-02 -3.68332088e-01 8.42331886e-01
-7.47825429e-02 9.38575864e-01 -5.76856673e-01 -2.31376350e-01
9.31212723e-01 -1.73929513e-01 -2.73802076e-02 -6.71330392e-01
-1.24424505e+00 9.51440036e-01 4.78818744e-01 -3.91646445e-01
-2.48523548e-01 4.09749389e-01 4.11839098e-01 4.98495191e-01
3.27780575e-01 6.34544194e-01 -4.74853873e-01 -6.77204072e-01
-8.78083289e-01 7.50411034e-01 7.65959203e-01 1.11617517e+00
4.27808762e-01 2.51092851e-01 -4.51111287e-01 2.78213054e-01
6.25628605e-02 4.09619600e-01 5.64605296e-01 -1.40165949e+00
5.42717516e-01 3.90285522e-01 7.66400516e-01 -4.45299536e-01
-2.49682933e-01 -3.82903636e-01 -4.03127253e-01 3.83977830e-01
6.99867070e-01 -2.73077756e-01 -7.91760147e-01 1.85555518e+00
4.61560905e-01 3.64350885e-01 1.86189324e-01 8.46316993e-01
-5.55823967e-02 6.66067243e-01 -1.85884267e-01 -3.56315821e-01
4.64133143e-01 -1.42308509e+00 -6.83857560e-01 -2.58677125e-01
7.55540311e-01 -2.77709588e-02 1.44901502e+00 7.04588175e-01
-1.04955816e+00 -2.80428350e-01 -9.74663913e-01 1.95144713e-01
1.10178202e-01 1.37723222e-01 7.31701672e-01 2.85842389e-01
-9.21838641e-01 1.05143142e+00 -1.32245374e+00 2.53762960e-01
5.88839531e-01 2.03774869e-01 -1.53484493e-01 8.23648460e-03
-1.12050140e+00 1.08493745e+00 3.90828967e-01 -1.36161242e-02
-1.67642975e+00 -7.55546391e-01 -7.49605060e-01 -3.79371226e-01
6.86504543e-01 -3.63779485e-01 1.68989551e+00 -1.10126364e+00
-2.05541539e+00 2.23915279e-01 4.18632627e-02 -1.01816535e+00
1.21398509e+00 -2.62379646e-01 -1.73970997e-01 -2.36278330e-03
-1.49686590e-01 6.72877073e-01 1.00515330e+00 -1.29270267e+00
-4.20623153e-01 -3.29575385e-03 2.55709767e-01 3.27318370e-01
1.02958754e-01 -5.91954410e-01 -1.46065488e-01 -4.35696065e-01
-5.08140922e-01 -1.21158576e+00 -4.39538926e-01 1.39402999e-02
-2.68708348e-01 -7.14571774e-02 6.60916269e-01 -6.35944486e-01
1.16640460e+00 -1.82044137e+00 -1.73670929e-02 7.45331198e-02
-7.35917017e-02 1.76989555e-01 -2.66994387e-01 5.72025836e-01
3.64150643e-01 -5.89405447e-02 -1.57322027e-02 -4.23533559e-01
1.86096802e-01 5.97884595e-01 -9.73965049e-01 6.45530760e-01
-7.43192509e-02 1.01157987e+00 -1.41633737e+00 -2.92307198e-01
1.93564191e-01 -6.67003244e-02 -7.52201378e-01 2.91861862e-01
-7.94668436e-01 8.64952266e-01 -3.75485361e-01 4.55777854e-01
3.46047878e-01 -7.20247626e-02 5.42097986e-01 2.13182643e-01
1.67013705e-02 5.17833650e-01 -8.17018986e-01 1.84857821e+00
-6.38141334e-01 3.66015226e-01 -2.66760260e-01 -7.62947142e-01
8.34875286e-01 9.35997367e-02 2.70476729e-01 -1.02295375e+00
-1.92894831e-01 7.13397041e-02 1.58631038e-02 -2.23080397e-01
4.63833511e-01 -1.32184684e-01 -3.66627611e-02 6.66852117e-01
2.28525568e-02 -2.42831722e-01 9.41392325e-04 1.97955951e-01
1.01756978e+00 1.05375254e+00 1.49689719e-01 -1.16087779e-01
-7.57689923e-02 1.17053144e-01 7.20529437e-01 1.00437713e+00
-5.80452859e-01 8.99535269e-02 7.35591412e-01 -3.89310420e-01
-1.04038978e+00 -1.19070387e+00 3.61242950e-01 8.85549366e-01
8.70041624e-02 -3.21697950e-01 -5.20911157e-01 -1.37265086e+00
2.90054411e-01 1.23381114e+00 -7.46387839e-01 -4.90432739e-01
-6.62454665e-01 -1.81048587e-01 7.27078259e-01 7.44199991e-01
3.79475951e-01 -8.93989205e-01 -6.05494738e-01 4.24173355e-01
-2.61622727e-01 -9.20215130e-01 -4.81294870e-01 1.08882166e-01
-1.03733075e+00 -9.95724440e-01 -4.44185853e-01 -2.46732444e-01
5.87698877e-01 -7.50024021e-02 9.64626074e-01 -2.40601510e-01
2.94104099e-01 3.28684092e-01 -2.93414742e-01 -3.43470156e-01
-7.58394122e-01 -2.20327169e-01 4.56907481e-01 -3.01632434e-01
-1.31670497e-02 -4.04013306e-01 -7.07723737e-01 2.73625880e-01
-3.46323282e-01 1.67812839e-01 1.15281940e-01 1.08186293e+00
5.20664215e-01 -3.45084250e-01 9.30044711e-01 -6.25806451e-01
7.38797367e-01 -5.06502688e-01 -9.24720645e-01 1.90525055e-01
-1.01024091e+00 5.46392500e-01 1.05638325e+00 -8.50025177e-01
-1.06216335e+00 -2.81066317e-02 1.60983205e-01 -9.00073349e-01
2.32262492e-01 1.20569348e-01 1.76273137e-01 1.77600652e-01
7.60148942e-01 4.63092625e-01 4.08094347e-01 -2.90391773e-01
6.85839832e-01 2.40097582e-01 3.03950131e-01 -9.51942623e-01
6.83910191e-01 5.36271036e-01 -1.76699653e-01 -1.19548500e-01
-9.63863373e-01 1.09158978e-01 -1.51748046e-01 -4.64320898e-01
1.79108202e-01 -1.12768197e+00 -7.72016585e-01 2.90440112e-01
-6.31058276e-01 -1.32971978e+00 -7.14679658e-01 4.55854326e-01
-1.25241339e+00 1.32359847e-01 -6.70785010e-01 -1.13933444e+00
5.71720116e-02 -1.20950520e+00 7.29493678e-01 1.51475534e-01
-9.08717662e-02 -1.08411300e+00 3.23698521e-01 2.24467423e-02
3.94813806e-01 2.44968325e-01 6.48054183e-01 -5.96545458e-01
-4.55458701e-01 1.03882797e-01 3.78927767e-01 3.96287203e-01
-3.51805910e-02 -1.76399678e-01 -9.67156470e-01 -5.65568268e-01
-1.39962316e-01 -1.18181562e+00 7.60890484e-01 2.15921313e-01
1.36986136e+00 -9.09342349e-01 -1.70743495e-01 4.13317770e-01
1.09269619e+00 9.49120894e-02 4.56102997e-01 3.06405544e-01
2.54966915e-01 5.17427921e-01 1.21081913e+00 8.69469464e-01
5.24736881e-01 4.91757601e-01 5.64624608e-01 2.56151974e-01
2.00524226e-01 -1.29177415e+00 1.04349852e+00 4.27692801e-01
8.59431699e-02 -9.79671404e-02 -5.50165236e-01 4.86155003e-01
-2.17329502e+00 -1.08960009e+00 3.98911446e-01 2.29742050e+00
1.37047577e+00 1.93873569e-01 5.65571308e-01 -2.49622434e-01
1.16448119e-01 1.95274681e-01 -1.08035636e+00 -3.73103052e-01
1.55680507e-01 1.66513816e-01 6.45559072e-01 6.93264306e-01
-6.75680459e-01 1.16306961e+00 6.40815926e+00 1.02346623e+00
-1.01477623e+00 1.41081020e-01 7.07510710e-01 -6.52957916e-01
-3.97311091e-01 2.68975914e-01 -8.22053790e-01 5.10978341e-01
1.10602689e+00 -1.34321541e-01 1.08035421e+00 1.04321170e+00
5.01416802e-01 -1.68531746e-01 -1.38995862e+00 7.34196007e-01
-3.31993550e-01 -1.47787333e+00 -1.75496891e-01 1.31414279e-01
9.81540263e-01 7.02890828e-02 1.85391814e-01 1.04067099e+00
1.03061450e+00 -1.12837732e+00 1.18997717e+00 6.67608142e-01
7.98337936e-01 -9.09190953e-01 1.22581474e-01 9.40956354e-01
-7.42673218e-01 -3.01589549e-01 -2.19381765e-01 -1.44339785e-01
7.11536109e-02 9.14402083e-02 -9.02334332e-01 2.93507099e-01
3.15665126e-01 8.26892853e-01 -8.51120725e-02 4.82698530e-01
-4.59585696e-01 9.71340537e-01 -3.48739266e-01 -2.34912664e-01
6.11634076e-01 -3.32337379e-01 4.28513706e-01 7.42255032e-01
-9.26703215e-03 -1.91678554e-01 5.07771194e-01 1.05982637e+00
-2.78414059e-02 -3.25704306e-01 -8.98568392e-01 -4.00242716e-01
4.91539419e-01 8.46210122e-01 -2.20860466e-01 -4.69629109e-01
6.19752407e-02 9.13773954e-01 7.02461004e-01 4.71205711e-01
-1.16612375e+00 2.16362208e-01 6.01384640e-01 -6.19176179e-02
1.41301662e-01 -4.97429639e-01 4.60790470e-02 -1.34328675e+00
-1.61067799e-01 -1.28686845e+00 2.20333979e-01 -4.97955263e-01
-1.11853361e+00 7.33578578e-02 1.62269875e-01 -1.35652637e+00
-7.74113297e-01 -4.45537150e-01 -3.60088766e-01 4.21755284e-01
-1.45131254e+00 -1.00035870e+00 1.23117752e-01 5.69508791e-01
7.12678611e-01 -3.54624130e-02 6.43403709e-01 -3.00934851e-01
-3.09372157e-01 7.33296275e-01 3.54244202e-01 -1.38420835e-01
7.29889512e-01 -1.30734646e+00 2.99694628e-01 5.17048657e-01
1.49445785e-02 4.73011822e-01 6.32452607e-01 -6.86420798e-01
-1.62958586e+00 -1.16755462e+00 1.75911888e-01 -6.03396773e-01
8.66087317e-01 -4.96473461e-01 -5.31459510e-01 1.05038726e+00
-1.13903545e-02 2.45913118e-01 2.48079091e-01 -1.11086912e-01
-3.54853123e-01 7.07161129e-02 -1.25063205e+00 9.75362718e-01
1.29506516e+00 -4.61413622e-01 -3.48700494e-01 4.27779973e-01
8.21117043e-01 -7.39602447e-01 -7.23168612e-01 8.06906894e-02
6.02302909e-01 -8.65745664e-01 9.02130544e-01 -1.18698549e+00
3.40069652e-01 -1.13949709e-01 8.53463169e-03 -1.61395097e+00
1.06351390e-01 -1.14362407e+00 -9.67724979e-01 7.84593821e-01
3.59070122e-01 -8.36292505e-01 6.35942519e-01 4.48422134e-01
-4.83231843e-02 -1.09576225e+00 -8.39200079e-01 -1.28727877e+00
3.81730676e-01 -4.62369323e-01 6.98778689e-01 7.21957445e-01
2.40060627e-01 -1.55307561e-01 -7.00295568e-01 -1.86001047e-01
5.81641495e-01 2.43930578e-01 9.49965537e-01 -4.27155733e-01
-7.96201229e-01 -3.69175524e-01 4.31297570e-01 -1.51373363e+00
5.53034604e-01 -8.85207593e-01 1.64506733e-01 -1.18449485e+00
-1.51546195e-01 -8.59536290e-01 -1.03392281e-01 6.75388575e-01
1.90452039e-01 -2.98386186e-01 2.49538511e-01 9.10185203e-02
-6.81026518e-01 1.23397338e+00 1.57385457e+00 8.11543390e-02
-4.14886892e-01 -4.39288002e-03 -4.81896490e-01 6.68649793e-01
9.48902249e-01 -4.70860124e-01 -9.45284247e-01 5.48410788e-02
3.28874201e-01 4.34443891e-01 4.63186294e-01 -7.02236772e-01
-8.54924917e-02 -6.34381115e-01 2.72242486e-01 -3.69519353e-01
2.82431424e-01 -4.30456877e-01 -2.53626466e-01 8.09421420e-01
-7.79107332e-01 2.32120212e-02 2.55871087e-01 1.05820239e+00
2.29703635e-01 8.53781030e-02 7.22580612e-01 -2.53388435e-01
-4.48839277e-01 3.61259341e-01 -2.83556521e-01 5.07481098e-01
1.04817903e+00 8.43473524e-02 -3.95040959e-01 -6.19488835e-01
-6.00657582e-01 5.73800504e-01 6.52901888e-01 4.19708759e-01
5.86015582e-01 -1.32865548e+00 -2.79104948e-01 1.77641973e-01
-2.58929171e-02 -1.67473480e-01 7.81157091e-02 9.72886384e-01
-2.24250391e-01 2.08255395e-01 -6.43498451e-02 -3.50587308e-01
-4.23168451e-01 7.61022151e-01 5.88952720e-01 -6.38701737e-01
-8.50694895e-01 4.93932098e-01 -1.09790325e-01 -8.23534667e-01
3.71409684e-01 -6.60506666e-01 1.93808556e-01 -2.99015015e-01
3.78118545e-01 4.87138242e-01 -3.11087370e-01 7.97495544e-02
-8.96534175e-02 -1.42182037e-01 -6.58494309e-02 -5.95315993e-01
1.08982670e+00 1.64959446e-01 6.22297704e-01 7.12322414e-01
6.97917581e-01 -7.40002170e-02 -2.27934408e+00 9.89324227e-02
-2.36488670e-01 -4.28786278e-01 -8.06261674e-02 -1.18588305e+00
-6.03165507e-01 7.15363264e-01 3.46896619e-01 -1.89307481e-01
6.64540887e-01 -3.30964565e-01 7.61711240e-01 5.33138812e-01
1.01996922e+00 -1.50822103e+00 3.29692930e-01 5.16483724e-01
9.09284115e-01 -1.41809154e+00 -1.25784338e-01 4.67454344e-01
-1.16265059e+00 9.05610740e-01 8.70164812e-01 -6.02002859e-01
2.72427350e-01 2.50172108e-01 -2.63453305e-01 3.66843998e-01
-1.37324488e+00 6.16111942e-02 -7.16716945e-02 7.34916747e-01
-2.08329275e-01 2.09705338e-01 -9.69796106e-02 6.63810492e-01
-9.12486240e-02 3.06527317e-01 5.03738046e-01 1.08219099e+00
-2.33864382e-01 -1.17440653e+00 2.30929498e-02 3.29595149e-01
-3.00567271e-03 1.98803008e-01 -2.30113611e-01 1.07534313e+00
-4.43303287e-01 8.72164130e-01 -9.05817869e-05 -3.16626221e-01
-6.67174906e-02 4.68739122e-02 9.72304642e-01 -1.69187963e-01
-6.01097047e-01 -1.97440431e-01 3.80906671e-01 -1.23900914e+00
1.13221310e-01 -5.83187699e-01 -1.51166916e+00 -3.95267367e-01
-8.88746977e-02 -1.90350488e-02 4.20647740e-01 9.86102879e-01
5.94193339e-01 7.77379647e-02 8.15092683e-01 -8.71919870e-01
-1.65647852e+00 -8.47639084e-01 -2.90576726e-01 1.46576151e-01
5.41268766e-01 -9.43604112e-01 -4.48774666e-01 -3.25775295e-01] | [4.052041053771973, 2.055196523666382] |
d69f6f00-2c12-443a-b215-5635f290e26e | recognizing-arguing-subjectivity-and-argument | null | null | https://aclanthology.org/W12-3810 | https://aclanthology.org/W12-3810.pdf | Recognizing Arguing Subjectivity and Argument Tags | null | ['Rebecca Hwa', 'Janyce Wiebe', 'er', 'Alex Conrad'] | 2012-07-01 | recognizing-arguing-subjectivity-and-argument-1 | https://aclanthology.org/W12-3810 | https://aclanthology.org/W12-3810.pdf | ws-2012-7 | ['subjectivity-analysis'] | ['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] | [-7.384544372558594, 3.661895275115967] |
26026692-b2ee-48b9-a25e-a892c1d9a2f7 | few-shot-relational-reasoning-via-connection | 2210.06722 | null | https://arxiv.org/abs/2210.06722v1 | https://arxiv.org/pdf/2210.06722v1.pdf | Few-shot Relational Reasoning via Connection Subgraph Pretraining | Few-shot knowledge graph (KG) completion task aims to perform inductive reasoning over the KG: given only a few support triplets of a new relation $\bowtie$ (e.g., (chop,$\bowtie$,kitchen), (read,$\bowtie$,library), the goal is to predict the query triplets of the same unseen relation $\bowtie$, e.g., (sleep,$\bowtie$,?). Current approaches cast the problem in a meta-learning framework, where the model needs to be first jointly trained over many training few-shot tasks, each being defined by its own relation, so that learning/prediction on the target few-shot task can be effective. However, in real-world KGs, curating many training tasks is a challenging ad hoc process. Here we propose Connection Subgraph Reasoner (CSR), which can make predictions for the target few-shot task directly without the need for pre-training on the human curated set of training tasks. The key to CSR is that we explicitly model a shared connection subgraph between support and query triplets, as inspired by the principle of eliminative induction. To adapt to specific KG, we design a corresponding self-supervised pretraining scheme with the objective of reconstructing automatically sampled connection subgraphs. Our pretrained model can then be directly applied to target few-shot tasks on without the need for training few-shot tasks. Extensive experiments on real KGs, including NELL, FB15K-237, and ConceptNet, demonstrate the effectiveness of our framework: we show that even a learning-free implementation of CSR can already perform competitively to existing methods on target few-shot tasks; with pretraining, CSR can achieve significant gains of up to 52% on the more challenging inductive few-shot tasks where the entities are also unseen during (pre)training. | ['Jure Leskovec', 'Hongyu Ren', 'Qian Huang'] | 2022-10-13 | null | null | null | null | ['relational-reasoning'] | ['natural-language-processing'] | [ 9.28663462e-02 8.08894336e-01 -3.88240010e-01 -3.89347345e-01
-7.51251936e-01 -1.18839391e-01 3.23897183e-01 3.34246188e-01
-1.15056396e-01 7.04266369e-01 -7.29864389e-02 -3.58192593e-01
-2.74254024e-01 -1.23288763e+00 -1.24637926e+00 -3.51986885e-01
-2.07454383e-01 8.78353059e-01 3.88978302e-01 -4.64603096e-01
-4.53003168e-01 -6.78866869e-03 -1.64582980e+00 4.32579577e-01
7.49514222e-01 9.55610693e-01 1.87404469e-01 4.36240077e-01
-2.51314282e-01 1.13698196e+00 -2.74190336e-01 -7.36059606e-01
7.35456422e-02 -2.55903721e-01 -1.36645937e+00 -1.48786038e-01
3.01157624e-01 1.95898786e-02 -2.27319092e-01 9.06856775e-01
4.97580655e-02 4.69827354e-01 4.95817453e-01 -1.32209074e+00
-4.59471613e-01 1.19871247e+00 -3.45650971e-01 1.16141364e-01
3.42341661e-01 -8.52167085e-02 1.51783621e+00 -1.01466155e+00
9.08455670e-01 1.06884730e+00 5.87277293e-01 3.63205761e-01
-1.42251885e+00 -7.51939774e-01 3.15631241e-01 5.05265355e-01
-1.62553990e+00 -3.43359172e-01 5.55606425e-01 -1.07914992e-01
1.33244908e+00 8.56238380e-02 5.23910820e-01 8.48397970e-01
-1.87842533e-01 8.07279348e-01 5.26931822e-01 -5.62540233e-01
1.80846214e-01 2.40931228e-01 5.12659371e-01 1.08941233e+00
3.39834213e-01 -2.52717316e-01 -5.75894535e-01 -1.42010465e-01
9.46928188e-02 1.17258854e-01 -2.45426744e-01 -7.87480235e-01
-8.39860857e-01 7.56088734e-01 7.77834177e-01 3.68931592e-01
-3.90003924e-03 2.45601937e-01 3.40159982e-01 5.20392478e-01
4.85454053e-01 4.01089936e-01 -8.60027313e-01 4.05375779e-01
-6.74828053e-01 1.97978958e-01 1.20644617e+00 1.40031362e+00
1.53066313e+00 -3.23380768e-01 -5.85275330e-02 6.18944943e-01
1.95521772e-01 1.72745213e-01 2.21782893e-01 -4.46557909e-01
6.51021779e-01 8.78214180e-01 -1.72340646e-01 -6.11161709e-01
-4.24322158e-01 -4.64641303e-01 -7.14297831e-01 -4.60702449e-01
1.36006817e-01 -1.13722801e-01 -1.20383680e+00 1.77178538e+00
4.83562320e-01 5.04323304e-01 3.04942608e-01 4.81736809e-01
1.02588582e+00 4.89576966e-01 2.86400825e-01 -2.62755930e-01
1.30136299e+00 -9.58505571e-01 -2.26221800e-01 -6.15788758e-01
1.28018177e+00 -3.80254574e-02 1.08857059e+00 7.83178359e-02
-7.14337766e-01 -4.82812017e-01 -1.07928324e+00 -1.44177735e-01
-7.97776937e-01 -2.34079316e-01 9.65653837e-01 2.08755106e-01
-7.89341867e-01 6.34924591e-01 -5.96791804e-01 -6.52866721e-01
3.52981538e-01 4.36531484e-01 -4.41823781e-01 -6.15735233e-01
-1.55774045e+00 8.91962469e-01 1.00164282e+00 -2.37513036e-01
-1.01541066e+00 -1.02106726e+00 -1.27734101e+00 4.29120004e-01
1.14876306e+00 -8.65748882e-01 1.15047526e+00 -5.37391186e-01
-8.86602044e-01 1.08240509e+00 -7.68696144e-02 -7.33085871e-01
-9.28422958e-02 -1.00363024e-01 -4.87699479e-01 -3.04528349e-03
3.16658080e-01 5.08088350e-01 7.14336038e-01 -1.16899157e+00
-7.91428089e-01 -3.73921365e-01 4.85220343e-01 2.33278155e-01
-1.78539708e-01 -2.74151444e-01 -5.47399700e-01 -2.19162330e-01
1.45311117e-01 -7.24629879e-01 6.31135851e-02 -7.11511970e-01
-6.58248305e-01 -5.51745594e-01 5.78260243e-01 -1.32321388e-01
1.07367361e+00 -2.06580496e+00 1.28661171e-01 3.25541139e-01
3.85192245e-01 2.53418833e-01 4.76475023e-02 3.53743881e-01
-2.89840907e-01 8.04964304e-02 -1.96061999e-01 -1.71668410e-01
4.34065945e-02 5.76723516e-01 -4.78455067e-01 7.71195963e-02
9.40970331e-02 1.13499188e+00 -1.14789426e+00 -5.22629082e-01
-4.47377050e-03 -1.59367591e-01 -5.42290688e-01 2.78316736e-01
-7.01153398e-01 -2.25646704e-01 -3.53265047e-01 7.16388702e-01
2.38489792e-01 -6.50942564e-01 4.68221039e-01 -3.25297058e-01
5.33203304e-01 3.49562615e-01 -1.12320983e+00 1.87318730e+00
-6.02427721e-01 2.31524631e-02 -2.27123752e-01 -1.35790694e+00
7.18351781e-01 9.10626575e-02 8.41573104e-02 -3.59404117e-01
-5.03579937e-02 1.38998836e-01 -3.85618582e-03 -3.27225208e-01
2.00184658e-01 -5.70066988e-01 -3.06359142e-01 4.13817495e-01
7.34106004e-01 -1.18378654e-01 4.31019664e-01 7.32446432e-01
1.45124388e+00 8.62646773e-02 5.48846066e-01 -1.43089697e-01
2.79438227e-01 3.16511869e-01 6.33347034e-01 8.87326300e-01
3.20732743e-01 6.74668029e-02 5.08588493e-01 -4.65210170e-01
-5.43263555e-01 -1.05858850e+00 1.25663891e-01 1.57012880e+00
3.46221775e-01 -8.31120312e-01 -4.44966823e-01 -1.15812051e+00
1.37854517e-01 1.03207886e+00 -6.53524220e-01 -4.82338905e-01
-2.91557997e-01 -5.52427948e-01 4.49144274e-01 5.49118876e-01
4.49621290e-01 -1.11764967e+00 -2.74193853e-01 3.01654607e-01
-1.53691769e-01 -1.21901345e+00 -6.92635179e-02 6.49444878e-01
-6.91401899e-01 -1.38932526e+00 -2.41002627e-02 -8.56371105e-01
7.07520247e-01 2.20358133e-01 1.53420699e+00 3.85542631e-01
-4.56975341e-01 3.34680080e-01 -3.40907604e-01 -2.73042530e-01
-7.11726844e-02 4.16278601e-01 -1.21912830e-01 -1.46860197e-01
7.41248071e-01 -7.48978853e-01 -2.35417068e-01 2.18848810e-01
-6.78732812e-01 1.36340633e-01 5.25264680e-01 9.32525039e-01
6.75175011e-01 3.11313599e-01 6.40945256e-01 -1.66741216e+00
1.46984428e-01 -8.25314283e-01 -2.57095814e-01 6.69437170e-01
-8.40382874e-01 2.62338787e-01 6.93557680e-01 -2.52998114e-01
-1.02443409e+00 -1.07507937e-01 1.83977708e-01 -6.28199697e-01
-8.00243914e-02 9.65260923e-01 -3.80605519e-01 1.64138466e-01
9.26527679e-01 2.90788785e-02 -4.43867683e-01 -5.04429102e-01
6.99754894e-01 1.13596104e-01 6.48715973e-01 -7.94196963e-01
9.57904518e-01 3.03966135e-01 -7.15496913e-02 -4.79456663e-01
-1.48462725e+00 -7.44604349e-01 -5.23240983e-01 2.53181756e-01
5.29516637e-01 -1.16666865e+00 -6.92521632e-01 -1.34438789e-02
-8.26316953e-01 -6.44933164e-01 -5.77814937e-01 1.52566329e-01
-4.31828767e-01 -3.35819498e-02 -4.47403073e-01 -3.94804627e-01
-5.00060976e-01 -5.52804470e-01 8.18885624e-01 -5.29935136e-02
-1.30014643e-01 -1.01216722e+00 -3.24224867e-02 3.06451827e-01
-5.28499708e-02 1.56180505e-02 1.54158747e+00 -1.03092277e+00
-6.57621264e-01 -1.27873421e-01 -2.55216300e-01 1.12361759e-01
-8.02479386e-02 -5.20485818e-01 -8.98790538e-01 -2.18289986e-01
-2.73022056e-01 -8.31998348e-01 1.14385974e+00 -2.15532780e-01
1.08104241e+00 -3.84207189e-01 -8.41572464e-01 6.70915067e-01
1.54586756e+00 -3.15480828e-01 5.14963388e-01 -1.18141342e-02
9.07418132e-01 3.30785811e-01 7.73116767e-01 6.48979023e-02
7.70978391e-01 4.27119881e-01 3.47392768e-01 1.20573223e-01
-3.61325033e-02 -6.14620626e-01 2.97827460e-03 3.56792033e-01
-1.27713412e-01 -2.75731198e-02 -1.08664632e+00 7.15618432e-01
-2.03293848e+00 -9.25684392e-01 3.04360598e-01 2.08845782e+00
9.88730550e-01 3.80898476e-01 -2.47312590e-01 1.19615737e-02
5.16475976e-01 4.51014601e-02 -6.61818266e-01 -1.91578522e-01
1.82538792e-01 6.31205499e-01 3.60530972e-01 4.09037143e-01
-1.02233171e+00 1.43164098e+00 5.12308359e+00 8.82304013e-01
-6.07101500e-01 2.35874251e-01 4.48583782e-01 1.39169898e-02
-3.44277918e-01 4.47937250e-01 -1.17543006e+00 -2.66587287e-02
8.66133928e-01 -4.51350719e-01 5.18966913e-01 1.17753911e+00
-6.80161655e-01 -9.84968022e-02 -1.60201752e+00 7.69736230e-01
2.68607765e-01 -1.27666795e+00 6.27091900e-02 -2.75267482e-01
6.78936303e-01 1.05651505e-01 -3.93441945e-01 1.29240048e+00
7.78496921e-01 -9.16648209e-01 3.65051270e-01 1.74044460e-01
8.53375375e-01 -7.03055918e-01 6.36279821e-01 7.20805168e-01
-1.44439375e+00 -1.38927862e-01 -5.67258596e-01 -8.80727693e-02
-1.09570898e-01 7.86195815e-01 -1.22528827e+00 9.73177314e-01
7.52601087e-01 7.95332372e-01 -4.34100211e-01 4.66329634e-01
-6.83134675e-01 3.97859931e-01 -3.03078681e-01 1.31671861e-01
3.17576975e-02 2.06528813e-01 2.22372815e-01 9.69631970e-01
1.64500892e-01 5.25464237e-01 3.89781773e-01 8.37586164e-01
-5.45178533e-01 1.29097477e-01 -8.93833280e-01 6.41165450e-02
4.53827918e-01 1.42916703e+00 -4.61210221e-01 -7.27175713e-01
-4.73325312e-01 8.00490081e-01 1.11519194e+00 4.82935429e-01
-7.83929169e-01 -3.90680939e-01 4.29054081e-01 3.32171649e-01
6.75123394e-01 4.36875582e-01 1.57007560e-01 -1.37213647e+00
-1.27458468e-01 -6.69603348e-01 9.75814402e-01 -7.61114478e-01
-1.31693947e+00 3.41151685e-01 2.06133917e-01 -5.99874020e-01
-4.15934980e-01 -3.57871264e-01 -6.85578108e-01 6.44111633e-01
-1.38833022e+00 -1.41213310e+00 -2.76571929e-01 9.94172990e-01
2.89384842e-01 1.72773027e-03 9.67214704e-01 8.88180435e-02
-4.26770359e-01 5.79060972e-01 -5.74353993e-01 1.55892834e-01
5.90337873e-01 -1.40748310e+00 2.93343663e-01 6.92780733e-01
4.20653313e-01 7.62601614e-01 4.69961166e-01 -7.69647419e-01
-1.44525754e+00 -1.50542247e+00 1.05148709e+00 -4.36102837e-01
8.16885471e-01 -5.15193164e-01 -1.00850940e+00 1.29442656e+00
-2.54365057e-01 4.56464380e-01 7.20727563e-01 1.03289318e+00
-6.41002536e-01 -3.31492543e-01 -1.03267384e+00 4.02602941e-01
1.53891087e+00 -6.03818715e-01 -1.06087101e+00 5.38952410e-01
1.00912309e+00 -4.80900317e-01 -8.75682592e-01 6.17774308e-01
1.16107520e-02 -5.71781695e-01 9.99573588e-01 -1.00668883e+00
2.72197574e-01 -1.38198495e-01 -1.20175943e-01 -1.38436401e+00
-2.72410572e-01 -4.38126743e-01 -7.80998588e-01 1.09102631e+00
6.58034086e-01 -5.38091838e-01 7.36270189e-01 7.62358069e-01
-2.69434869e-01 -8.02784860e-01 -6.91303909e-01 -7.72356391e-01
-3.25649649e-01 -5.28869390e-01 7.19475031e-01 1.03912055e+00
4.74857330e-01 9.77946281e-01 -1.68359175e-01 3.33035797e-01
6.86915100e-01 5.99117577e-01 8.97399187e-01 -1.36966157e+00
-6.67996228e-01 2.53210485e-01 -3.38133603e-01 -7.79062748e-01
4.79599029e-01 -1.39222968e+00 1.86290830e-01 -1.57328427e+00
5.03120422e-01 -7.07364082e-01 -4.12619710e-01 1.23067713e+00
-3.47138077e-01 -3.41958962e-02 -4.19337004e-02 1.81005988e-02
-1.00150394e+00 3.31588894e-01 9.27087843e-01 -4.27436262e-01
-1.54458046e-01 -8.61759782e-02 -9.80214059e-01 7.99998105e-01
3.67799163e-01 -5.44634104e-01 -7.93277442e-01 -1.70222968e-01
4.94627744e-01 -1.41361514e-02 3.71243238e-01 -8.60294700e-01
5.53639889e-01 1.25917047e-01 7.06210732e-02 -3.33566070e-01
2.11970121e-01 -6.34777129e-01 1.78252161e-01 2.77193367e-01
-1.63411617e-01 -5.92875838e-01 3.16677578e-02 8.93002272e-01
-1.66111022e-01 -2.03644112e-01 4.46450412e-01 -3.47717524e-01
-1.14632249e+00 5.91833115e-01 4.18879837e-01 4.76020873e-01
1.11207116e+00 9.98568758e-02 -3.57732356e-01 6.74616359e-03
-1.18775785e+00 5.27128100e-01 1.61555842e-01 1.16950192e-01
5.63464403e-01 -1.02848041e+00 -3.80961955e-01 6.93420991e-02
6.24988198e-01 5.98925591e-01 1.95308805e-01 6.75327301e-01
1.57210138e-02 2.41196021e-01 3.16017002e-01 -3.35575521e-01
-9.42942977e-01 1.05080295e+00 2.11417750e-01 -7.09964216e-01
-7.84402788e-01 1.09567845e+00 3.58723581e-01 -4.15599018e-01
2.88706511e-01 -4.02631611e-01 -3.14470604e-02 -6.38713241e-02
3.73682201e-01 -9.78153758e-03 3.08511347e-01 -7.90772140e-02
-4.03661430e-01 1.09416716e-01 -5.42349458e-01 4.38467324e-01
1.39431381e+00 1.74315110e-01 -1.20357387e-01 4.93129969e-01
1.06129372e+00 -4.48507220e-01 -6.49711132e-01 -6.85859978e-01
2.44039282e-01 -2.03544959e-01 -2.56913811e-01 -7.78890610e-01
-1.00988376e+00 5.13224721e-01 -1.20494872e-01 6.10090569e-02
9.30492282e-01 6.38378739e-01 7.28742540e-01 9.74015713e-01
8.88407230e-01 -9.49771464e-01 -1.02030136e-01 4.62249458e-01
5.32594502e-01 -1.14269006e+00 -3.04918806e-03 -8.05942059e-01
-4.49276835e-01 5.69378078e-01 8.02129745e-01 -5.91454543e-02
7.67185211e-01 2.11774856e-02 -5.98422885e-01 -7.83743560e-01
-1.14597011e+00 -5.73405743e-01 2.16805995e-01 4.26600218e-01
1.31251216e-01 1.47136316e-01 1.55421451e-01 9.63247836e-01
-2.54905950e-02 2.69567966e-02 6.92319497e-02 8.23512495e-01
-5.18255055e-01 -1.00174940e+00 2.65510648e-01 8.82792175e-01
-1.09608494e-01 -5.26148498e-01 -1.52754083e-01 9.41868961e-01
3.46828699e-01 6.61088407e-01 -1.97286978e-01 -4.09865528e-01
4.14156228e-01 5.19466877e-01 4.23817337e-01 -1.39714813e+00
-4.81713921e-01 -3.54255885e-01 4.53580856e-01 -6.18676782e-01
-1.55590981e-01 -2.68688947e-01 -1.55484152e+00 -2.57525086e-01
-4.55516934e-01 4.43642497e-01 -1.69009413e-03 1.37057996e+00
2.81699896e-01 4.25578684e-01 3.32044810e-01 -2.43696257e-01
-3.93301427e-01 -9.73295391e-01 -8.72688413e-01 7.46205270e-01
-1.47596791e-01 -8.53685677e-01 -4.14629489e-01 1.16716169e-01] | [8.95909595489502, 7.953689098358154] |
e8ba4492-8667-4053-a023-4046397371f5 | enhanced-single-shot-detector-for-small | 2205.05927 | null | https://arxiv.org/abs/2205.05927v1 | https://arxiv.org/pdf/2205.05927v1.pdf | Enhanced Single-shot Detector for Small Object Detection in Remote Sensing Images | Small-object detection is a challenging problem. In the last few years, the convolution neural networks methods have been achieved considerable progress. However, the current detectors struggle with effective features extraction for small-scale objects. To address this challenge, we propose image pyramid single-shot detector (IPSSD). In IPSSD, single-shot detector is adopted combined with an image pyramid network to extract semantically strong features for generating candidate regions. The proposed network can enhance the small-scale features from a feature pyramid network. We evaluated the performance of the proposed model on two public datasets and the results show the superior performance of our model compared to the other state-of-the-art object detectors. | ['Jie Yang', 'Jocelyn Chanussot', 'Eric Granger', 'Masoumeh Zareapoor', 'Pourya Shamsolmoali'] | 2022-05-12 | null | null | null | null | ['small-object-detection'] | ['computer-vision'] | [ 9.76434574e-02 -3.18870485e-01 7.83946291e-02 -2.46909931e-01
-6.37653589e-01 8.06065872e-02 6.01615489e-01 -1.49064034e-01
-5.87289155e-01 4.63042438e-01 1.74970210e-01 3.59737128e-01
8.95587653e-02 -1.01070344e+00 -5.87018073e-01 -5.81553459e-01
7.30177313e-02 -1.57490164e-01 1.26624084e+00 -1.35195494e-01
2.76559681e-01 6.66107416e-01 -1.77099860e+00 4.54480231e-01
7.72892416e-01 1.23344457e+00 5.96256495e-01 3.98117065e-01
-2.38470346e-01 7.02711165e-01 -4.63446707e-01 -8.30744430e-02
5.02349734e-01 -4.72361445e-02 -3.68484467e-01 5.91299264e-03
3.23536634e-01 -5.35348177e-01 -6.72719955e-01 1.22779298e+00
7.51916766e-01 2.65857987e-02 5.44787943e-01 -1.16535974e+00
-8.81703973e-01 3.60537857e-01 -8.15069556e-01 6.72507465e-01
2.69235913e-02 2.26433590e-01 6.57093585e-01 -1.44249535e+00
5.65424979e-01 1.46681392e+00 6.19175375e-01 2.63001859e-01
-7.17257440e-01 -8.05099070e-01 -3.54549214e-02 4.02332008e-01
-1.69012439e+00 -2.84836262e-01 6.06895447e-01 -2.14370444e-01
1.07378089e+00 -1.86131924e-01 5.07629097e-01 6.32924438e-01
-1.19354688e-02 1.11024475e+00 9.45553482e-01 -2.88316846e-01
2.59374641e-02 2.43100166e-01 2.10537225e-01 8.42018545e-01
4.40351874e-01 2.86578327e-01 -3.44316006e-01 1.54668093e-01
7.38131523e-01 4.47712421e-01 1.11467309e-01 -4.65719141e-02
-9.56240714e-01 7.90098548e-01 1.24302411e+00 6.02051735e-01
-4.85689878e-01 -2.44585164e-02 2.09343195e-01 -9.61523280e-02
3.10518116e-01 -3.06116547e-02 -1.37950271e-01 3.08064312e-01
-9.45371807e-01 2.84241170e-01 3.60512704e-01 9.15135980e-01
6.78090751e-01 1.49890184e-01 -8.08157146e-01 6.84349597e-01
1.78223148e-01 3.64554614e-01 3.82606566e-01 -2.05466971e-01
3.81604940e-01 1.08317721e+00 7.95537531e-02 -1.07683873e+00
-3.84696722e-01 -4.62044626e-01 -9.14955676e-01 2.61696607e-01
2.13764176e-01 -1.05497301e-01 -1.37508285e+00 1.21913588e+00
3.25660557e-01 4.63799447e-01 1.37574941e-01 9.86188948e-01
1.32252634e+00 1.03516173e+00 2.66176373e-01 1.28580883e-01
1.49870455e+00 -1.12904787e+00 -4.03940856e-01 -3.04256022e-01
2.09648103e-01 -8.60668421e-01 6.83832824e-01 -5.26834056e-02
-8.92353773e-01 -1.13974857e+00 -1.17871416e+00 1.33083805e-01
-5.46547413e-01 4.87753063e-01 6.10130191e-01 7.00078428e-01
-7.52779543e-01 4.18835461e-01 -7.63488114e-01 -5.70180357e-01
1.02901566e+00 4.54292566e-01 -3.24540913e-01 -2.56876141e-01
-1.03921318e+00 7.84882724e-01 1.02657139e+00 3.56519222e-02
-8.95159483e-01 -5.43416142e-01 -5.51430106e-01 4.48563576e-01
4.56535310e-01 -4.14808750e-01 1.07115591e+00 -7.83150792e-01
-9.09912348e-01 6.38784468e-01 5.12324013e-02 -6.03534877e-01
1.97987840e-01 -9.12668034e-02 -4.64796126e-01 2.30901167e-01
2.30567567e-02 9.86940444e-01 8.51045728e-01 -8.80258560e-01
-1.14711916e+00 -2.08132938e-01 -7.60426372e-02 1.78428888e-01
-6.64453328e-01 7.05425680e-01 -5.48927486e-01 -4.32971507e-01
-2.43807007e-02 -4.60062891e-01 -3.62186968e-01 1.31956175e-01
-3.74478102e-01 -5.14340818e-01 1.29829383e+00 -3.09789151e-01
1.09804821e+00 -2.39069700e+00 -3.48275810e-01 -1.94612309e-01
1.16646260e-01 7.03316212e-01 -1.16736099e-01 1.42515704e-01
2.16560379e-01 -1.60662487e-01 -9.11797658e-02 -4.86125425e-02
-1.32060930e-01 -1.48951709e-01 -1.31154940e-01 2.02747807e-01
7.06007481e-01 1.06201315e+00 -6.24470413e-01 -9.16813314e-01
2.22806215e-01 3.65587384e-01 -2.96267062e-01 2.32772067e-01
-2.68078670e-02 -7.90829808e-02 -7.61398971e-01 9.62154090e-01
1.01441371e+00 -3.14146638e-01 -5.12724340e-01 -2.86861122e-01
-3.65940571e-01 -3.83876532e-01 -1.54611170e+00 1.41355968e+00
1.68154195e-01 4.43894744e-01 -1.95252687e-01 -7.44251609e-01
1.02349746e+00 1.06797971e-01 1.83518887e-01 -6.17583513e-01
2.88976669e-01 4.17282492e-01 2.18892649e-01 -4.16761249e-01
4.79093820e-01 -1.52998984e-01 1.40865296e-01 -1.90445408e-02
2.70598769e-01 8.08628201e-02 2.49118477e-01 3.06633711e-01
1.11172199e+00 -7.35104680e-02 4.53826636e-01 -1.94275856e-01
5.25882721e-01 8.45437050e-02 5.97188830e-01 8.69208992e-01
-4.58976626e-01 7.10219383e-01 -5.18424287e-02 -5.36913872e-01
-9.35877323e-01 -1.13613605e+00 -1.61290169e-01 1.04802144e+00
4.24767613e-01 -1.08867951e-01 -6.54003561e-01 -7.37951517e-01
1.65252149e-01 3.07781816e-01 -4.44056034e-01 -2.72373974e-01
-5.55287302e-01 -1.06546307e+00 6.10583246e-01 9.46789026e-01
1.12865770e+00 -1.36580181e+00 -8.04432154e-01 2.53785372e-01
3.04596990e-01 -1.26181209e+00 -3.89037818e-01 1.14564814e-01
-6.51840806e-01 -7.72513866e-01 -9.17575061e-01 -1.25665355e+00
6.79687500e-01 7.45905519e-01 8.73687625e-01 1.11604087e-01
-9.09805477e-01 -1.96916416e-01 -3.88040811e-01 -7.75502324e-01
-7.29697421e-02 1.17301811e-02 -1.69234529e-01 1.73102379e-01
5.99560797e-01 -1.95772022e-01 -8.93883288e-01 2.56998599e-01
-1.07842207e+00 -1.56053472e-02 1.19496393e+00 6.22873247e-01
4.95579541e-01 2.12007746e-01 8.46195519e-01 -4.27100360e-01
5.25484145e-01 -4.11908388e-01 -7.78511286e-01 1.87515229e-01
-3.30340564e-01 1.63976979e-02 5.68316460e-01 -3.68860364e-01
-1.40914297e+00 3.30259055e-01 -1.15029141e-01 -3.25619251e-01
-2.52684265e-01 1.91750288e-01 -9.73391160e-03 -3.31741780e-01
7.48725235e-01 4.62060988e-01 -4.08100069e-01 -6.06772780e-01
2.29326770e-01 9.69458461e-01 5.32671809e-01 -1.50438964e-01
8.89912248e-01 5.30321956e-01 -6.69041276e-02 -7.57129967e-01
-9.88309145e-01 -7.58696854e-01 -6.80622101e-01 -8.01887140e-02
8.84015322e-01 -1.00946987e+00 -2.65090197e-01 6.23984993e-01
-1.22727334e+00 3.27882171e-01 -2.97190756e-01 3.80879164e-01
-1.19813956e-01 9.28250179e-02 -5.22126853e-01 -8.25201452e-01
-7.51371861e-01 -1.05441129e+00 1.15926552e+00 8.37057054e-01
6.15732312e-01 -2.80093729e-01 -2.80697644e-01 -2.13831082e-01
6.93712711e-01 7.01885596e-02 4.55941767e-01 -7.20847428e-01
-9.38674510e-01 -5.30599833e-01 -1.02048552e+00 3.41565043e-01
-1.58796217e-02 3.57279703e-02 -1.05700874e+00 -1.62979648e-01
-1.79185256e-01 -4.33022469e-01 1.33945966e+00 4.04518902e-01
1.16629636e+00 1.43517897e-01 -6.69297576e-01 4.90139991e-01
1.68589592e+00 1.26009181e-01 7.41589844e-01 1.20886654e-01
4.03296649e-01 -2.73484364e-02 7.30574846e-01 5.74847817e-01
-9.70353261e-02 4.26036030e-01 2.77238876e-01 -3.27510774e-01
-3.53945762e-01 -2.93123245e-01 1.11211717e-01 2.56294072e-01
-1.42070696e-01 -2.19317693e-02 -8.36909831e-01 6.22130513e-01
-1.69770622e+00 -1.00389671e+00 -1.27609745e-01 1.71028638e+00
4.13055569e-01 5.49850881e-01 7.90889710e-02 -1.92887574e-01
9.05612826e-01 2.45088935e-01 -4.55405205e-01 1.05251707e-01
-4.19864655e-01 3.51745129e-01 5.67770183e-01 -2.35780060e-01
-1.51942766e+00 1.19084990e+00 6.44374609e+00 1.15593398e+00
-9.86847997e-01 2.27279276e-01 4.14786309e-01 1.29033715e-01
4.96085405e-01 -5.92088029e-02 -1.40984511e+00 3.16982269e-01
5.89066207e-01 -2.93739378e-01 -2.21733719e-01 1.15348613e+00
-6.65498152e-02 -8.81257206e-02 -6.53851509e-01 9.01065946e-01
1.01450972e-01 -1.45584285e+00 2.73826476e-02 -2.38240376e-01
8.71368408e-01 2.80881792e-01 -2.92141852e-03 6.11330092e-01
1.08821236e-01 -9.23473239e-01 4.59757239e-01 4.38528985e-01
5.51226497e-01 -7.69887745e-01 8.86725724e-01 5.27343035e-01
-1.70632517e+00 -4.95105356e-01 -9.95208204e-01 -7.27766454e-02
2.51624659e-02 5.16142726e-01 -8.53453338e-01 4.10562992e-01
8.18994462e-01 5.24568856e-01 -8.52003038e-01 1.69343162e+00
4.18412499e-02 5.69518268e-01 -4.76691067e-01 -3.48962247e-01
6.05861902e-01 7.48071671e-02 3.95031840e-01 1.34641373e+00
4.66698796e-01 1.11041874e-01 4.46302861e-01 1.04398215e+00
-3.52958739e-01 2.78556615e-01 -6.02999866e-01 -1.03804573e-01
3.62140834e-01 1.60322845e+00 -1.05545366e+00 -7.27106452e-01
-6.79635465e-01 8.17568421e-01 2.45147318e-01 -3.43147330e-02
-7.76495993e-01 -7.34130979e-01 2.41614535e-01 8.75621475e-03
6.61080718e-01 5.06352447e-02 4.38938960e-02 -8.38553846e-01
5.00424244e-02 -4.42810863e-01 5.65925479e-01 -7.04607189e-01
-1.23622572e+00 6.25037491e-01 2.69307066e-02 -1.23957121e+00
3.45839471e-01 -6.63606822e-01 -9.07525420e-01 7.51568496e-01
-1.62223637e+00 -1.37961638e+00 -5.98061144e-01 6.37533724e-01
6.92860842e-01 -3.47510040e-01 5.19389451e-01 4.28332061e-01
-4.62594509e-01 2.89208353e-01 2.02377945e-01 3.35540086e-01
4.35142368e-01 -8.40639532e-01 6.12310767e-01 1.12109089e+00
6.45253360e-02 3.72608960e-01 2.95430243e-01 -7.24880338e-01
-1.22129440e+00 -1.37573922e+00 3.90195429e-01 1.58262894e-01
4.67226088e-01 -2.25123405e-01 -8.40920568e-01 4.08598095e-01
2.71004826e-01 7.29202032e-01 3.19522321e-02 -4.03121233e-01
-2.50229269e-01 -3.03523362e-01 -1.38412821e+00 3.16596538e-01
1.08640802e+00 -3.10869664e-02 -7.35307336e-01 3.32435161e-01
8.09082210e-01 -7.85686001e-02 -3.37560713e-01 5.54055095e-01
4.48736548e-01 -9.20114994e-01 1.11248910e+00 -5.19656897e-01
1.98431864e-01 -4.36813444e-01 -1.78540230e-01 -1.00552428e+00
-6.69599175e-01 -1.20767660e-01 1.12607945e-02 1.05984497e+00
1.63730130e-01 -4.37948495e-01 8.89367402e-01 1.51979223e-01
-1.73913524e-01 -6.22077107e-01 -7.40637422e-01 -9.39899981e-01
-2.90348321e-01 4.14250717e-02 5.45360029e-01 1.83197215e-01
-4.14947331e-01 3.11374545e-01 -1.76102176e-01 1.18907630e-01
6.67816937e-01 2.93461949e-01 5.37667751e-01 -1.35882020e+00
-1.49407357e-01 -3.81300330e-01 -8.83080304e-01 -8.88861835e-01
-2.97171026e-01 -8.66524696e-01 2.43810162e-01 -1.78200531e+00
6.82687461e-01 -1.61442742e-01 -6.84254587e-01 3.17582548e-01
-6.10362113e-01 4.30128902e-01 3.99260342e-01 -9.38371941e-02
-8.87874901e-01 6.17182791e-01 1.22620928e+00 -2.82832474e-01
-2.15271890e-01 -3.06193531e-02 -6.06730461e-01 7.95099378e-01
8.40770781e-01 -6.26052737e-01 -1.34900168e-01 -1.07021458e-01
-3.76130611e-01 -4.59681809e-01 3.60417545e-01 -1.60790610e+00
4.35237378e-01 -3.94707322e-02 8.73359263e-01 -9.66120720e-01
2.96501487e-01 -5.45776904e-01 -2.05578446e-01 9.06209588e-01
1.21450640e-01 -2.91309208e-01 2.68940628e-01 6.29553139e-01
-3.83931965e-01 -2.98096836e-01 1.19999766e+00 -3.36119235e-01
-1.39084256e+00 4.42405373e-01 -3.27086858e-02 -1.23465702e-01
1.34991932e+00 -4.06383544e-01 -2.83735484e-01 2.94538617e-01
-3.25787872e-01 8.60997885e-02 -6.65798644e-03 5.36422074e-01
1.16828787e+00 -1.32872355e+00 -9.68682647e-01 1.85788929e-01
3.89593422e-01 -4.00253385e-03 3.61798137e-01 6.65845156e-01
-5.08686960e-01 4.24592346e-01 -3.96855950e-01 -4.84231561e-01
-1.34900713e+00 8.29438984e-01 2.34075218e-01 -1.90456569e-01
-9.00148690e-01 1.07802546e+00 4.29256797e-01 -2.21425872e-02
1.93874285e-01 -3.19062203e-01 -3.52413505e-01 -1.76433161e-01
8.35092604e-01 2.89988041e-01 -1.32495403e-01 -5.13740540e-01
-3.65221828e-01 4.16693717e-01 -2.03628033e-01 4.09868121e-01
1.57429242e+00 2.05282435e-01 9.20385271e-02 4.48949262e-02
9.37556565e-01 -4.00880426e-01 -1.15418637e+00 -5.94392002e-01
-2.19890587e-02 -6.21547043e-01 2.33616263e-01 -4.91787165e-01
-1.07999027e+00 8.76083374e-01 1.02702951e+00 7.04659298e-02
1.15349352e+00 7.52602071e-02 9.33214009e-01 4.74673301e-01
2.96286196e-01 -1.14170289e+00 3.64438295e-01 4.49204296e-01
8.38574052e-01 -1.59533787e+00 -2.34227511e-03 -4.90362644e-01
-4.82626110e-01 1.10907304e+00 9.49158609e-01 -5.74257135e-01
8.80589604e-01 3.19317162e-01 -3.88266712e-01 -1.91501781e-01
-4.73574936e-01 -6.80606186e-01 4.78017747e-01 4.01148617e-01
1.15345038e-01 -2.05465639e-03 -2.05611438e-01 6.62907839e-01
4.03967142e-01 3.71451527e-01 1.78036004e-01 1.09443855e+00
-1.15450346e+00 -6.82103634e-01 -6.32466018e-01 8.10140908e-01
-4.29482967e-01 -1.79269001e-01 -3.14326733e-01 6.86347663e-01
3.83245856e-01 7.88857341e-01 3.90305929e-02 -2.30985641e-01
4.34890270e-01 -7.05019832e-02 3.76943648e-01 -6.97910786e-01
-5.16410887e-01 -3.72780785e-02 -3.07883680e-01 -3.41547221e-01
-3.77878636e-01 -3.06688756e-01 -1.40093148e+00 1.86419353e-01
-5.22909045e-01 -1.99512333e-01 4.73612875e-01 7.23841727e-01
3.80470723e-01 5.92216909e-01 5.43937504e-01 -9.17814434e-01
-8.19193602e-01 -1.26203799e+00 -7.52459586e-01 4.68673706e-01
-1.04620494e-01 -8.87605190e-01 -5.61123826e-02 -7.79575333e-02] | [8.820202827453613, -0.4182262718677521] |
c3fb8049-9cac-4e44-8d65-794fce6f5540 | adaptive-warm-start-mcts-in-alphazero-like | 2105.06136 | null | https://arxiv.org/abs/2105.06136v1 | https://arxiv.org/pdf/2105.06136v1.pdf | Adaptive Warm-Start MCTS in AlphaZero-like Deep Reinforcement Learning | AlphaZero has achieved impressive performance in deep reinforcement learning by utilizing an architecture that combines search and training of a neural network in self-play. Many researchers are looking for ways to reproduce and improve results for other games/tasks. However, the architecture is designed to learn from scratch, tabula rasa, accepting a cold-start problem in self-play. Recently, a warm-start enhancement method for Monte Carlo Tree Search was proposed to improve the self-play starting phase. It employs a fixed parameter $I^\prime$ to control the warm-start length. Improved performance was reported in small board games. In this paper we present results with an adaptive switch method. Experiments show that our approach works better than the fixed $I^\prime$, especially for "deep," tactical, games (Othello and Connect Four). We conjecture that the adaptive value for $I^\prime$ is also influenced by the size of the game, and that on average $I^\prime$ will increase with game size. We conclude that AlphaZero-like deep reinforcement learning benefits from adaptive rollout based warm-start, as Rapid Action Value Estimate did for rollout-based reinforcement learning 15 years ago. | ['Aske Plaat', 'Mike Preuss', 'Hui Wang'] | 2021-05-13 | null | null | null | null | ['board-games'] | ['playing-games'] | [-4.45527643e-01 -4.35173884e-02 -8.92599300e-02 -1.20346770e-01
-5.02140880e-01 -4.35909539e-01 3.52898657e-01 -1.59926578e-01
-9.94772315e-01 1.07508957e+00 -1.57787830e-01 -4.96350467e-01
-3.01672727e-01 -1.13692689e+00 -7.26896584e-01 -7.46632576e-01
-4.28404957e-01 5.86700737e-01 4.78076726e-01 -9.31117177e-01
3.52009654e-01 1.59600645e-01 -1.55430889e+00 1.70571715e-01
7.25624144e-01 7.32383192e-01 2.85242826e-01 7.94504046e-01
4.94463835e-03 1.00013602e+00 -7.99938023e-01 -4.77881670e-01
5.81229627e-01 -8.34302127e-01 -9.14345324e-01 -5.18121600e-01
8.91144481e-03 -3.23991567e-01 -2.25733727e-01 9.45601702e-01
6.67937934e-01 4.09255177e-01 1.80485711e-01 -1.09680653e+00
2.54337668e-01 9.83837664e-01 -4.28048521e-01 5.11632562e-01
6.19531162e-02 6.92210317e-01 9.58446681e-01 -1.13201417e-01
4.97955024e-01 8.89562488e-01 8.14426541e-01 6.66363537e-01
-9.96910870e-01 -1.00855637e+00 -9.82743353e-02 3.37812394e-01
-1.12387931e+00 1.43188955e-02 6.17742121e-01 -4.62641716e-02
1.35855877e+00 1.81791168e-02 1.21165061e+00 1.04810786e+00
5.18136621e-01 5.75571656e-01 1.32712662e+00 -5.27948499e-01
7.77325273e-01 -2.61654556e-01 -3.29760671e-01 7.68205762e-01
1.53080486e-02 8.40636551e-01 -5.25098801e-01 -1.70393568e-02
9.84024763e-01 -5.59526742e-01 3.07825685e-01 -4.24447298e-01
-5.89524090e-01 1.21484029e+00 6.44757926e-01 3.03140521e-01
-4.34334040e-01 7.17086375e-01 6.07834041e-01 6.03824079e-01
-2.12895349e-02 1.10472047e+00 -5.60869992e-01 -7.81681180e-01
-1.03343356e+00 6.88118279e-01 8.47490609e-01 4.15350050e-01
7.81701446e-01 4.48938161e-01 9.10271779e-02 7.61162937e-01
-1.31963015e-01 1.39094934e-01 7.72913575e-01 -1.29011571e+00
3.25664461e-01 1.10880658e-01 6.10200912e-02 -4.80706930e-01
-7.56347060e-01 -7.17248321e-01 -6.37457490e-01 8.84956658e-01
6.58014059e-01 -6.61276758e-01 -7.76093841e-01 1.89315689e+00
1.28331080e-01 -5.69413714e-02 -1.30328462e-02 7.46161580e-01
3.46735895e-01 3.95581126e-01 1.59345374e-01 -7.08971592e-03
1.11515522e+00 -9.67961609e-01 -2.58737922e-01 -4.27338302e-01
7.42576778e-01 -2.75160253e-01 1.14828682e+00 7.86228538e-01
-1.28840792e+00 -6.06409609e-01 -1.18726778e+00 6.92581773e-01
-1.92446083e-01 -3.56694281e-01 1.03543627e+00 1.04619408e+00
-1.15274978e+00 1.03806198e+00 -7.66680777e-01 -1.23755358e-01
2.12779328e-01 5.98185599e-01 -3.80287543e-02 4.35575068e-01
-1.47235155e+00 1.04720294e+00 8.82959843e-01 -3.11999261e-01
-1.08968151e+00 -2.62304634e-01 -4.40375060e-01 1.90622494e-01
7.32782722e-01 -6.08122706e-01 1.54645526e+00 -1.09658194e+00
-2.09633422e+00 5.56884766e-01 4.74149525e-01 -1.02421165e+00
4.41981196e-01 4.61923890e-02 8.40794221e-02 -1.48580655e-01
-1.13889106e-01 7.87437260e-01 6.58384860e-01 -9.03565586e-01
-7.18424916e-01 -2.39052325e-01 5.91921031e-01 4.94219095e-01
1.27550974e-01 -1.03901289e-01 -6.49292618e-02 -4.60227907e-01
-1.88067615e-01 -9.84013975e-01 -8.11837912e-01 -7.08837867e-01
3.59576166e-01 -2.54084617e-01 -1.78055763e-02 -9.86883976e-03
1.22073889e+00 -1.69886398e+00 -2.13484928e-01 3.52958620e-01
2.08242133e-01 2.94007212e-01 -2.06191033e-01 5.01885295e-01
-1.47685662e-01 -1.15656257e-01 1.71601996e-01 3.36421132e-01
-4.01155911e-02 4.00384158e-01 5.93448952e-02 3.84743288e-02
-2.83955127e-01 9.29377198e-01 -1.00333190e+00 -1.77157417e-01
-1.57664791e-02 -2.62010992e-01 -1.13790894e+00 3.76954675e-02
-3.74302417e-01 4.46131453e-02 -1.98029429e-01 1.07645676e-01
4.03924614e-01 -8.45350400e-02 3.27958941e-01 4.32743013e-01
-2.09803864e-01 4.30297107e-01 -1.20438135e+00 1.73411381e+00
-5.50383985e-01 3.35631877e-01 -2.52014808e-02 -1.03018594e+00
9.99784887e-01 7.35093281e-02 4.35358524e-01 -1.18728805e+00
4.70292360e-01 1.06892854e-01 6.23010278e-01 -1.48647055e-01
6.29256845e-01 -4.09234166e-01 -7.30427280e-02 3.45598847e-01
1.43847272e-01 -4.92731333e-01 5.83274543e-01 1.72346607e-02
1.48899984e+00 3.80160838e-01 2.32124165e-01 -4.11202878e-01
1.50511876e-01 2.71814704e-01 5.44877052e-01 1.32573044e+00
-3.03257644e-01 1.53264999e-01 7.43357658e-01 -7.32146740e-01
-1.10762382e+00 -9.36830997e-01 2.22043350e-01 1.59779704e+00
-1.26004750e-02 -6.63398504e-01 -1.00455999e+00 -6.48833573e-01
-2.55330145e-01 7.02981174e-01 -7.76100039e-01 -3.98515552e-01
-7.89888203e-01 -9.19992030e-01 6.81536853e-01 6.08279645e-01
7.88916886e-01 -1.56172311e+00 -1.19187784e+00 6.93096578e-01
1.30733088e-01 -4.91937697e-01 2.98499260e-02 1.01399672e+00
-9.82409418e-01 -9.13818717e-01 -5.75559437e-01 -4.32128370e-01
6.74905553e-02 -2.26282567e-01 1.38033664e+00 1.32666171e-01
-1.13378882e-01 -1.36308193e-01 -5.22753358e-01 -3.63561332e-01
-5.12239873e-01 4.00682330e-01 1.55706834e-02 -9.77255106e-01
1.78495914e-01 -7.18145669e-01 -7.75487065e-01 2.44860291e-01
-7.59679079e-01 4.87039723e-02 5.54380119e-01 1.26838529e+00
2.80645899e-02 2.50837535e-01 5.33605516e-01 -7.31375754e-01
9.44314182e-01 -2.16339290e-01 -7.61357844e-01 -2.10044712e-01
-8.64318132e-01 5.96198142e-01 8.34347546e-01 -4.14546162e-01
-6.82952821e-01 -1.85545191e-01 -6.95435286e-01 -2.34063983e-01
1.18239209e-01 4.46081668e-01 3.32146913e-01 -1.08542360e-01
1.22954428e+00 1.67934015e-01 3.66478264e-02 -1.27032906e-01
2.38530427e-01 7.99817443e-02 1.50887147e-01 -7.27889538e-01
4.35337782e-01 -7.10959639e-03 -1.24947689e-01 -2.22835049e-01
-3.92833382e-01 8.27685744e-02 -7.25859404e-02 -3.05113643e-01
5.04517257e-01 -7.33523607e-01 -1.23648763e+00 5.14549911e-01
-6.04142606e-01 -1.04633832e+00 -6.60107136e-01 4.31010187e-01
-9.89594877e-01 3.04408502e-02 -7.05025911e-01 -5.29107690e-01
-3.32813561e-01 -1.12311625e+00 3.26482117e-01 5.20038903e-01
-2.08260879e-01 -6.72584355e-01 5.33499002e-01 2.06720605e-01
5.75371385e-01 -1.36938542e-01 6.44520998e-01 -5.30334115e-01
-1.75553471e-01 1.73919782e-01 1.42301008e-01 2.85253912e-01
-2.71739244e-01 -3.91055882e-01 -5.17288864e-01 -3.82551670e-01
-1.23858459e-01 -7.20129609e-01 7.91719854e-01 4.54172164e-01
9.62505639e-01 -2.75925159e-01 1.76027209e-01 5.50012112e-01
1.38354623e+00 6.92736626e-01 9.36085582e-01 1.06675899e+00
8.47201571e-02 7.74971545e-02 6.84533656e-01 8.01320434e-01
1.05306059e-01 6.11617684e-01 8.40259492e-01 6.98346272e-02
1.46809906e-01 -2.16350257e-01 3.69722486e-01 3.21176380e-01
-3.67948800e-01 -4.65380289e-02 -1.05848706e+00 3.59793335e-01
-1.69143963e+00 -1.17598736e+00 2.69953668e-01 2.03609848e+00
1.17788672e+00 8.44904602e-01 6.05845749e-01 9.98840854e-02
2.25915581e-01 1.01474673e-01 -5.64832032e-01 -1.01109695e+00
8.82811621e-02 1.07052875e+00 7.58996904e-01 4.90656972e-01
-6.36282682e-01 1.50308919e+00 6.90513229e+00 1.27516913e+00
-1.07316732e+00 1.85411461e-02 6.01920903e-01 -3.50516081e-01
6.70581236e-02 1.20256282e-01 -6.42444968e-01 4.92725223e-01
1.17084980e+00 -8.33020881e-02 6.40346467e-01 1.03785706e+00
2.05276310e-01 -6.27967596e-01 -4.89658058e-01 6.38025343e-01
-3.74220759e-01 -1.49481153e+00 -4.00513947e-01 1.12890229e-01
8.31915140e-01 2.14878112e-01 -4.93452922e-02 8.81808639e-01
1.09894311e+00 -9.62872505e-01 6.33937001e-01 -5.10091484e-02
5.13508976e-01 -1.27762568e+00 6.84484839e-01 5.42494118e-01
-9.14519608e-01 -3.15865546e-01 -3.20630729e-01 -5.42849243e-01
-3.90166849e-01 1.30094722e-01 -8.85488927e-01 2.05759823e-01
7.71158099e-01 -7.12517723e-02 -5.01386881e-01 1.10666239e+00
-2.99984902e-01 6.82968020e-01 -3.69678855e-01 -5.92536986e-01
8.60278845e-01 -8.45425427e-02 1.41779348e-01 9.59823489e-01
1.15683161e-01 1.71017304e-01 2.60606166e-02 5.27027667e-01
1.79377928e-01 1.61040612e-02 -3.83822680e-01 2.46077180e-01
1.23281777e-01 8.74453843e-01 -8.22211087e-01 -1.78580657e-01
1.98933691e-01 7.24682391e-01 3.51088345e-01 2.71600261e-02
-1.09703279e+00 -5.31234205e-01 5.71946979e-01 1.95168078e-01
4.97785151e-01 -3.53672862e-01 -3.29392254e-01 -8.13178003e-01
-4.84398991e-01 -1.26041210e+00 3.19777131e-01 -6.66865885e-01
-6.43885136e-01 7.34905005e-01 3.27849500e-02 -9.91333008e-01
-7.99982488e-01 -5.93585312e-01 -7.62463748e-01 5.76052964e-01
-8.92348826e-01 -3.31455350e-01 -3.01296660e-03 6.06361210e-01
5.78128040e-01 -4.23644453e-01 6.46801472e-01 -2.40887493e-01
-3.13766509e-01 8.06187868e-01 3.86519551e-01 1.87606476e-02
3.06023329e-01 -1.37528276e+00 4.65650827e-01 4.46469307e-01
1.64533183e-01 3.46918017e-01 9.82121944e-01 -4.62120026e-01
-1.02590847e+00 -3.14691097e-01 1.85353085e-01 5.89730218e-02
6.79290056e-01 -1.45613253e-01 -5.18528819e-01 1.68737218e-01
4.62536663e-01 -3.53817433e-01 4.31940734e-01 4.75327790e-01
-5.29669411e-02 -3.32691550e-01 -1.11977708e+00 7.79631913e-01
9.95099366e-01 -6.52200729e-02 -5.19685149e-01 -1.99091490e-02
2.24313453e-01 -6.43120706e-01 -5.19960344e-01 2.19068468e-01
5.62214315e-01 -1.42690420e+00 7.18218148e-01 -5.58067262e-01
4.96482670e-01 1.42485797e-01 3.02807927e-01 -1.69727135e+00
-4.20564353e-01 -8.16139162e-01 1.63109839e-01 3.46572131e-01
3.69264960e-01 -4.75386709e-01 1.58753526e+00 1.40513316e-01
-9.29899663e-02 -8.95814359e-01 -1.19525290e+00 -9.83917594e-01
5.47079265e-01 -4.74057823e-01 3.74454737e-01 5.06637752e-01
2.67809391e-01 3.60469252e-01 -3.80740732e-01 -4.81438875e-01
4.15246218e-01 -2.46325716e-01 7.18144178e-01 -9.22715425e-01
-8.60243201e-01 -6.67846918e-01 -1.68728307e-01 -9.69966412e-01
-1.69887871e-01 -5.37970006e-01 1.96364582e-01 -1.14852691e+00
-1.32168800e-01 -6.80931687e-01 -3.85923564e-01 6.63466930e-01
-2.12269444e-02 1.52913049e-01 3.89483690e-01 -1.47191465e-01
-7.02312827e-01 5.64423561e-01 1.29844069e+00 8.53477344e-02
-4.61261719e-01 5.13094589e-02 -6.02912307e-01 7.15514064e-01
1.25964117e+00 -5.83424091e-01 -4.16456163e-01 -9.18337256e-02
7.78410971e-01 4.00864720e-01 9.21255276e-02 -1.56348872e+00
2.42074564e-01 -4.83458668e-01 2.30145410e-01 -1.95086554e-01
3.08766842e-01 -3.91240954e-01 3.00090946e-02 9.78426814e-01
-4.00717646e-01 4.19436485e-01 4.67954993e-01 2.05349088e-01
5.98499514e-02 -6.59569204e-01 9.05760288e-01 -6.10509932e-01
-7.23252177e-01 -7.90056586e-02 -7.95775890e-01 3.41584593e-01
9.83271897e-01 -4.27394480e-01 -6.00048043e-02 -5.11528552e-01
-8.37060213e-01 2.67795414e-01 2.32889622e-01 1.88509285e-01
3.01921397e-01 -1.02711093e+00 -5.41882575e-01 1.81681737e-01
-2.31215149e-01 -3.96452606e-01 2.08446756e-01 4.05797362e-01
-9.20082271e-01 1.17830627e-01 -7.45586097e-01 -2.27352664e-01
-1.05380666e+00 2.65290588e-01 7.24682570e-01 -9.35859025e-01
-3.22702765e-01 1.06918240e+00 -2.95883477e-01 -1.52422667e-01
2.98459738e-01 -1.96276709e-01 8.69962499e-02 -5.63864075e-02
2.66622752e-01 2.59432316e-01 1.82864904e-01 2.69813150e-01
-1.24596514e-01 7.37194270e-02 -3.19726229e-01 -5.22310555e-01
1.26172578e+00 3.50479633e-01 3.15770686e-01 4.49844636e-02
6.03619874e-01 -3.38290185e-01 -1.46325076e+00 4.14493345e-02
-1.28201351e-01 -2.07562432e-01 5.06850481e-02 -1.07307124e+00
-1.08318114e+00 6.45342052e-01 7.77685761e-01 2.22666457e-01
1.06998098e+00 -3.86332452e-01 6.06011808e-01 6.93526268e-01
9.24811661e-01 -1.45078158e+00 3.90957206e-01 1.08440983e+00
5.13230324e-01 -8.90414417e-01 -7.08015263e-02 5.06985128e-01
-7.80001938e-01 1.05133915e+00 1.10011029e+00 -5.18543601e-01
3.25117469e-01 4.21784550e-01 -7.73463324e-02 -1.21541508e-01
-1.05879414e+00 -4.53727245e-01 -5.16477168e-01 4.78424937e-01
1.77109808e-01 -1.88238770e-02 -7.20879018e-01 3.94165784e-01
-7.99135387e-01 2.09122986e-01 5.27734280e-01 1.02448487e+00
-8.36318374e-01 -1.46408951e+00 -2.58418262e-01 4.08143848e-01
-4.06795353e-01 -2.83515811e-01 -5.38484417e-02 1.05770576e+00
3.82902414e-01 7.15562999e-01 4.90498096e-02 -6.48394406e-01
1.34622812e-01 -2.17547305e-02 8.56788039e-01 -2.69591004e-01
-1.41534233e+00 -1.47266552e-01 1.70568913e-01 -6.74678683e-01
2.80895028e-02 -3.10147941e-01 -1.33010304e+00 -7.51919568e-01
-2.15541780e-01 6.18499517e-01 5.22651613e-01 7.91635513e-01
-2.62808818e-02 5.70924580e-01 5.65215707e-01 -7.17606008e-01
-8.52771938e-01 -7.31394410e-01 -6.74487114e-01 -5.68534434e-02
-1.85165912e-01 -7.66072392e-01 -1.52090237e-01 -6.97240353e-01] | [3.5645768642425537, 1.4894155263900757] |
533200ac-70c5-43df-8d15-2da2c6397100 | e-2-go-motion-motion-augmented-event-stream | 2112.03596 | null | https://arxiv.org/abs/2112.03596v3 | https://arxiv.org/pdf/2112.03596v3.pdf | E$^2$(GO)MOTION: Motion Augmented Event Stream for Egocentric Action Recognition | Event cameras are novel bio-inspired sensors, which asynchronously capture pixel-level intensity changes in the form of "events". Due to their sensing mechanism, event cameras have little to no motion blur, a very high temporal resolution and require significantly less power and memory than traditional frame-based cameras. These characteristics make them a perfect fit to several real-world applications such as egocentric action recognition on wearable devices, where fast camera motion and limited power challenge traditional vision sensors. However, the ever-growing field of event-based vision has, to date, overlooked the potential of event cameras in such applications. In this paper, we show that event data is a very valuable modality for egocentric action recognition. To do so, we introduce N-EPIC-Kitchens, the first event-based camera extension of the large-scale EPIC-Kitchens dataset. In this context, we propose two strategies: (i) directly processing event-camera data with traditional video-processing architectures (E$^2$(GO)) and (ii) using event-data to distill optical flow information (E$^2$(GO)MO). On our proposed benchmark, we show that event data provides a comparable performance to RGB and optical flow, yet without any additional flow computation at deploy time, and an improved performance of up to 4% with respect to RGB only information. | ['Barbara Caputo', 'Matteo Matteucci', 'Emanuele Gusso', 'Marco Cannici', 'Gabriele Goletto', 'Mirco Planamente', 'Chiara Plizzari'] | 2021-12-07 | null | null | null | null | ['event-based-vision'] | ['computer-vision'] | [ 3.99435103e-01 -5.17009497e-01 -5.94637841e-02 -9.07009467e-02
-1.14200026e-01 -4.20750231e-01 4.21730757e-01 -8.63562375e-02
-7.62172818e-01 6.27728462e-01 2.41496593e-01 3.76594253e-02
6.71852157e-02 -7.22880363e-01 -5.23906291e-01 -7.36607075e-01
-8.33661482e-03 -3.10976446e-01 5.67019939e-01 1.15512282e-01
1.93477020e-01 4.80671495e-01 -1.74863410e+00 3.26171219e-01
3.63388866e-01 1.22955394e+00 5.41687831e-02 1.01815867e+00
1.48222253e-01 1.21939182e+00 -3.92419457e-01 -2.68156886e-01
1.86260417e-01 -6.34928763e-01 -4.99217659e-01 6.74411133e-02
6.20983481e-01 -6.49973691e-01 -8.35843563e-01 8.08612645e-01
5.27311921e-01 3.41048747e-01 2.22594738e-02 -1.32796013e+00
-2.22982541e-01 1.69489291e-02 -5.70932567e-01 7.14677334e-01
7.51318932e-01 6.48127794e-01 5.96295357e-01 -7.45712638e-01
7.93928504e-01 9.37183261e-01 4.45160568e-01 6.35830700e-01
-1.04521859e+00 -3.27439964e-01 3.93219516e-02 7.38304138e-01
-1.11051571e+00 -6.76919222e-01 7.96741664e-01 -2.47258112e-01
1.24835265e+00 2.20719546e-01 1.06941080e+00 1.19183016e+00
3.85110974e-01 7.84786046e-01 9.94637012e-01 -7.50592500e-02
5.62416911e-01 -4.44304466e-01 -1.67533636e-01 6.12294734e-01
1.21597573e-01 1.79846227e-01 -1.08146131e+00 2.21750587e-01
1.01783347e+00 5.23273289e-01 -6.65564418e-01 -1.26180157e-01
-1.73623502e+00 4.10813123e-01 4.13509309e-01 1.50250569e-01
-5.73193669e-01 9.08095717e-01 4.84771848e-01 1.75976709e-01
9.90767851e-02 -2.08149757e-02 -1.76495180e-01 -7.74374247e-01
-8.63693297e-01 -1.04637049e-01 6.14061892e-01 7.35903800e-01
6.26909316e-01 1.22775882e-01 -6.09423891e-02 7.83421025e-02
8.67897645e-02 6.89970195e-01 6.36476398e-01 -1.27147937e+00
3.73575091e-01 5.42556584e-01 1.89496562e-01 -1.05301976e+00
-4.79779243e-01 1.70215398e-01 -9.14865673e-01 3.25262338e-01
6.19413853e-01 -1.33561239e-01 -4.73872751e-01 1.49773681e+00
2.12209627e-01 6.50361478e-01 3.04303449e-02 1.10257578e+00
6.94210529e-01 5.72699070e-01 -7.07171932e-02 -5.80836535e-01
1.47949445e+00 -6.73445404e-01 -7.64513433e-01 -1.92625195e-01
2.87762970e-01 -6.55824721e-01 9.23085570e-01 6.13565922e-01
-1.22326803e+00 -5.01107037e-01 -8.93312752e-01 -1.27781287e-01
-2.48370647e-01 -4.67336401e-02 1.00992954e+00 6.74090385e-01
-1.07786441e+00 5.34028411e-01 -1.26298714e+00 -6.48965597e-01
3.99120361e-01 3.14580590e-01 -5.54162323e-01 -1.12272836e-01
-7.76409030e-01 4.70612437e-01 2.51207501e-01 9.55084711e-03
-7.84839332e-01 -5.91798723e-01 -8.09200406e-01 -1.02234460e-01
2.98478991e-01 -7.30537593e-01 9.27324593e-01 -9.00571287e-01
-1.76322830e+00 7.38724113e-01 -3.77283126e-01 -7.09544539e-01
4.42532659e-01 -3.20648193e-01 -4.04127747e-01 7.80832648e-01
-2.76126176e-01 7.53735006e-01 8.69003117e-01 -4.23024207e-01
-7.69070983e-01 -3.59721094e-01 3.38952184e-01 1.36008456e-01
-4.40927833e-01 1.64752051e-01 -5.07057846e-01 -5.01522183e-01
-3.57401595e-02 -8.91533494e-01 -1.02152795e-01 5.44058442e-01
8.38285759e-02 -1.04971062e-02 9.62861300e-01 -7.60025457e-02
1.07929540e+00 -2.17502642e+00 -6.35909513e-02 -3.69298756e-01
3.53569061e-01 3.42504919e-01 5.57526760e-02 2.17908904e-01
5.09272553e-02 -4.49791253e-01 -7.04489648e-02 -1.59237087e-01
-5.33892334e-01 2.96372503e-01 -2.49187648e-01 6.90477312e-01
5.05139045e-02 1.01413798e+00 -1.24741781e+00 -5.22953868e-01
9.15329695e-01 5.52677572e-01 -5.08555293e-01 1.65032782e-02
8.44811555e-03 6.23202562e-01 -2.34291673e-01 6.67452157e-01
3.55839610e-01 -2.71773219e-01 -2.60743890e-02 -4.48444903e-01
-3.30539316e-01 -1.03912055e-01 -1.37942207e+00 2.06103539e+00
-2.93326169e-01 1.04524648e+00 -2.42104679e-01 -7.31706202e-01
6.15329027e-01 3.40445697e-01 1.01079261e+00 -8.96900952e-01
2.19087392e-01 -5.19848138e-04 -1.91824540e-01 -5.53232014e-01
5.77349365e-01 1.95881933e-01 1.08682312e-01 3.70810479e-01
4.39679511e-02 1.32454544e-01 4.18736547e-01 9.07521322e-02
1.66566908e+00 2.34814033e-01 3.35238874e-01 1.74203292e-01
5.54429650e-01 1.08026462e-02 6.68881655e-01 6.06479406e-01
-6.78060532e-01 7.21558750e-01 1.59140781e-01 -7.07727313e-01
-6.75531864e-01 -1.17580795e+00 8.10968354e-02 6.24089360e-01
5.53140819e-01 -5.04633009e-01 -5.79628468e-01 -3.25484872e-01
-2.67932087e-01 2.18200281e-01 -2.98999190e-01 -1.15920134e-01
-6.82562172e-01 -7.31488049e-01 6.19815826e-01 8.05187523e-01
9.26046431e-01 -1.04183817e+00 -1.56478322e+00 3.85625184e-01
-2.39821896e-01 -1.60992324e+00 -4.44284171e-01 -5.42786568e-02
-1.18154919e+00 -1.29940975e+00 -5.88991642e-01 -2.84727722e-01
3.74759853e-01 6.52690649e-01 9.68108833e-01 -3.42934370e-01
-5.62165499e-01 8.99412632e-01 -3.90993357e-01 -1.34633198e-01
3.15582395e-01 -5.24142206e-01 1.53077841e-01 4.26556677e-01
5.29105127e-01 -7.73574769e-01 -1.31280744e+00 3.48938137e-01
-9.73897040e-01 7.32803196e-02 3.72519672e-01 5.51667511e-01
6.96418524e-01 -1.66639760e-01 1.19863585e-01 -3.13282251e-01
5.78138679e-02 -1.76920503e-01 -5.24030030e-01 -9.05568004e-02
-2.20348492e-01 -4.13843304e-01 8.04771185e-01 -5.42826116e-01
-9.99122620e-01 4.97204661e-01 2.38968164e-01 -7.68585563e-01
-1.96351737e-01 1.52219027e-01 2.04404414e-01 -1.01236239e-01
6.89068079e-01 3.12211931e-01 -1.14039108e-01 -1.68435633e-01
4.58777010e-01 4.03819442e-01 1.02718949e+00 -1.55151725e-01
3.00505608e-01 1.32469058e+00 1.80611879e-01 -9.61112857e-01
-3.53146970e-01 -7.63660610e-01 -5.64610004e-01 -5.31493902e-01
1.11966467e+00 -1.07054138e+00 -1.20002639e+00 7.52914250e-01
-1.17316878e+00 -3.10674906e-01 -6.84460223e-01 8.84904802e-01
-8.04878771e-01 4.94078189e-01 -6.93271816e-01 -7.64096558e-01
-2.52980679e-01 -8.73733997e-01 1.00681686e+00 6.05155826e-01
-5.19775003e-02 -8.32363725e-01 9.12779495e-02 2.83765674e-01
4.12653476e-01 5.74030161e-01 -5.90398386e-02 2.08292231e-01
-8.48762512e-01 -9.66212377e-02 -3.38013828e-01 1.01669148e-01
1.72969386e-01 -4.77963015e-02 -1.06034482e+00 -1.83736041e-01
1.38670146e-01 -1.06801197e-01 7.89538682e-01 4.62574244e-01
1.12495863e+00 1.76581159e-01 -1.18954368e-01 8.95413756e-01
1.53990531e+00 3.91945302e-01 1.12981570e+00 1.10341519e-01
7.01859117e-01 8.84380266e-02 4.55425262e-01 8.27189684e-01
3.41204911e-01 7.25705326e-01 5.13342261e-01 3.04930657e-02
-4.26202387e-01 -5.00536524e-02 7.20659435e-01 6.08793616e-01
-5.52199841e-01 -3.29582989e-01 -5.68105698e-01 5.90341985e-01
-2.05050540e+00 -1.34209347e+00 -4.74323839e-01 2.24036026e+00
5.74886620e-01 -2.68033333e-02 8.73566046e-02 2.69980848e-01
6.32115543e-01 4.02248800e-01 -6.08663201e-01 -2.31612384e-01
-2.44245932e-01 4.06742543e-01 6.18815422e-01 -6.32199645e-02
-1.13429308e+00 7.21833348e-01 6.23450279e+00 3.76538724e-01
-1.29774702e+00 1.84634671e-01 2.25799814e-01 -6.01843119e-01
2.87500620e-01 5.30831181e-02 -5.38788140e-01 7.47524619e-01
1.05320227e+00 -1.57274008e-02 4.18548405e-01 5.65611780e-01
5.78065574e-01 -6.44462049e-01 -1.21848941e+00 1.68226588e+00
1.31469622e-01 -1.40934920e+00 -2.29808196e-01 5.39484993e-02
5.76220870e-01 1.12073153e-01 -3.56918722e-01 -3.15779924e-01
-3.89287509e-02 -6.55549943e-01 5.83668351e-01 6.64486766e-01
9.07594502e-01 -4.04441714e-01 5.13966382e-01 -4.16251607e-02
-1.46394515e+00 -5.45395575e-02 -3.93568695e-01 -4.50413883e-01
5.49953580e-01 8.33936691e-01 -7.11884648e-02 3.09518963e-01
9.40328658e-01 1.16695857e+00 -3.44307780e-01 1.05847681e+00
-1.04434274e-01 4.99094069e-01 -4.85497743e-01 7.72742927e-02
3.31535153e-02 -1.10109918e-01 6.09332204e-01 1.13212371e+00
3.22820932e-01 3.48482937e-01 5.44276787e-04 5.98854899e-01
1.55597487e-02 -3.80110472e-01 -7.13813841e-01 1.10060811e-01
1.51857480e-01 1.21783352e+00 -8.44843566e-01 -3.80038410e-01
-8.61492395e-01 1.59399676e+00 -1.43676206e-01 2.73273796e-01
-9.51321781e-01 -5.24787724e-01 8.75408173e-01 -1.11197144e-01
3.79170299e-01 -5.63779891e-01 2.99821626e-02 -1.51432288e+00
1.47276461e-01 -3.79616201e-01 4.74113435e-01 -9.93399560e-01
-8.47211659e-01 2.17413291e-01 -2.28721976e-01 -1.54273200e+00
-1.81197479e-01 -7.56876826e-01 -6.06381655e-01 2.45122790e-01
-1.55578291e+00 -7.65302598e-01 -9.47385013e-01 1.33841121e+00
5.21919191e-01 1.27110139e-01 5.29690683e-01 4.28944618e-01
-6.47830904e-01 2.17694104e-01 2.93242410e-02 2.56566703e-01
7.09351838e-01 -1.08373177e+00 1.82689458e-01 1.18560421e+00
3.69806349e-01 2.78007567e-01 3.26179117e-01 -3.32157016e-01
-2.15506148e+00 -1.04142320e+00 5.34933567e-01 -5.21007121e-01
5.85451603e-01 -9.79496911e-02 -3.77089262e-01 5.29109120e-01
6.73432201e-02 7.86829352e-01 4.25063968e-01 -6.30625129e-01
-3.12641114e-02 -5.39291143e-01 -1.04923964e+00 4.98638690e-01
1.36713243e+00 -5.86997330e-01 -4.11869556e-01 6.61723539e-02
3.87252539e-01 -4.03281331e-01 -9.59225714e-01 5.72093390e-02
5.91353834e-01 -1.41627610e+00 1.06894994e+00 -1.69987172e-01
3.33838999e-01 -6.77767813e-01 -1.14894152e-01 -7.79439867e-01
-4.66750897e-02 -9.01231647e-01 -5.84817290e-01 9.58019376e-01
-4.73500878e-01 -6.47839069e-01 9.60278928e-01 4.97524410e-01
6.85040131e-02 -3.37498277e-01 -1.22868121e+00 -7.95170665e-01
-7.66006112e-01 -7.36176729e-01 8.88078883e-02 6.53927863e-01
4.95405644e-02 -7.44985491e-02 -3.42380255e-01 -1.62759110e-01
6.85819983e-01 1.47469968e-01 7.27428079e-01 -9.28517461e-01
-3.04824114e-01 -2.58938670e-01 -1.09154642e+00 -1.28586400e+00
-3.95353109e-01 -3.77851725e-01 -6.44995049e-02 -1.22249043e+00
9.78596583e-02 7.59032667e-02 -3.49374354e-01 3.02253574e-01
1.07485829e-02 8.37787569e-01 3.02142888e-01 2.16047823e-01
-9.53127027e-01 3.58803511e-01 1.01666844e+00 4.22934815e-02
-1.87431470e-01 -4.38357294e-01 -2.61213362e-01 8.45429957e-01
5.01814544e-01 -1.49015069e-01 -4.08786774e-01 -4.84439701e-01
2.11247131e-01 2.69321769e-01 8.88088763e-01 -1.47376466e+00
6.65187120e-01 -2.05063105e-01 4.48246032e-01 -4.00667816e-01
4.55479264e-01 -7.87556529e-01 2.07509905e-01 6.36674464e-01
1.81633696e-01 2.09632561e-01 4.22185548e-02 8.28587651e-01
-3.65913242e-01 1.96363866e-01 6.33133531e-01 -2.76544780e-01
-1.28885591e+00 3.35991681e-01 -6.96668267e-01 6.97061419e-02
1.18524778e+00 -6.42436862e-01 -5.58834076e-01 -2.67749518e-01
-3.86976719e-01 -1.22276403e-01 5.87053657e-01 2.28840351e-01
7.91958332e-01 -1.26218200e+00 -3.45517516e-01 1.40389621e-01
1.50624543e-01 -9.56045911e-02 3.83875549e-01 1.18585312e+00
-7.76762187e-01 3.18413079e-01 -5.27550042e-01 -1.00122178e+00
-1.07398367e+00 4.93393153e-01 1.64837986e-01 5.20862117e-02
-9.14027214e-01 5.94942391e-01 7.98166096e-02 5.11260331e-01
4.80726585e-02 -4.74952340e-01 2.95733251e-02 6.58985525e-02
8.66770625e-01 8.12059462e-01 -6.69223517e-02 -4.80936319e-01
-6.08965099e-01 7.51219988e-01 4.24967080e-01 -9.30404812e-02
1.09207785e+00 -3.79385948e-01 7.45865554e-02 3.83493066e-01
9.99527872e-01 -2.35854730e-01 -1.63270390e+00 -4.05051671e-02
-2.54206717e-01 -7.20295370e-01 2.86320508e-01 -3.65005881e-01
-1.42852366e+00 9.70300436e-01 8.88926625e-01 5.95893860e-02
1.71798897e+00 -2.27617025e-01 1.13450098e+00 2.71901220e-01
6.64880693e-01 -1.05364668e+00 3.26083362e-01 2.62039691e-01
3.72561663e-01 -1.20744467e+00 3.83986421e-02 -3.35241288e-01
-4.17442203e-01 1.27271545e+00 4.89932537e-01 -2.51997411e-01
3.06211978e-01 2.99743295e-01 -1.86888888e-01 -1.24371953e-01
-7.84714937e-01 -5.41136920e-01 -1.46913603e-01 6.96489215e-01
2.66875148e-01 -1.95369110e-01 -2.98741311e-01 6.83704317e-02
3.19571137e-01 6.37720227e-01 6.72606885e-01 1.08192670e+00
-1.23393565e-01 -7.77627766e-01 -4.22271609e-01 2.34044701e-01
-3.83207113e-01 1.07263073e-01 -7.10184723e-02 5.22790253e-01
1.72172055e-01 1.10461318e+00 4.15424258e-01 -3.37663293e-01
4.06696916e-01 -1.97949022e-01 6.97196066e-01 -8.25836286e-02
-5.77392578e-01 -1.26231670e-01 -1.62036419e-01 -1.36772048e+00
-9.44698572e-01 -7.58309722e-01 -1.22985268e+00 -6.10065579e-01
9.28289741e-02 -5.36341071e-01 5.55957913e-01 6.98891878e-01
5.97862780e-01 4.24632132e-01 4.55170661e-01 -9.28766012e-01
1.48857519e-01 -3.92717838e-01 -4.77663815e-01 6.99880302e-01
2.08371893e-01 -5.18591344e-01 -2.64000297e-01 5.85172474e-01] | [8.631546974182129, -1.2885894775390625] |
8a12e111-9ba2-427a-bd3a-b0f83ed399be | cross-modal-learning-by-hallucinating-missing | null | null | https://doi.org/10.1016/B978-0-12-817358-9.00018-4 | https://www.researchgate.net/publication/334901581_Cross-modal_Learning_by_Hallucinating_Missing_Modalities_in_RGB-D_Vision | Cross-modal Learning by Hallucinating Missing Modalities in RGB-D Vision | Diverse input data modalities can provide complementary cues for several tasks, usually leading to more robust algorithms and better performance. However, while a (training) dataset could be accurately designed to include a variety of sensory inputs, it is often the case that not all modalities are available in real life (testing) scenarios, when the model is to be deployed. This raises the challenge of how to learn robust representations leveraging multimodal data in the training stage, while considering limitations at test time, such as noisy or missing modalities. This chapter presents a new approach for multimodal video action recognition, developed within the unified frameworks of distillation and privileged information, named generalized distillation. We consider the particular case of learning representations from depth and RGB videos, while relying on RGB data only at test time. Our approach consists in training a hallucination network that learns to distill depth features through multiplicative connections of spatiotemporal representations, leveraging soft labels and hard labels, and the euclidean distance between feature maps. We report state-of-the-art or comparable results on video action recognition on the largest multimodal dataset available for this task, the NTU RGB+D, as well as on the UWA3DII and Northwestern-UCLA. | ['Nuno C. Garcia', 'Vittorio Murino', 'Pietro Morerio'] | 2019-01-01 | null | null | null | multimodal-scene-understanding-algorithms | ['multimodal-activity-recognition'] | ['computer-vision'] | [ 6.34682536e-01 -1.39151454e-01 -1.81760728e-01 -4.03695434e-01
-9.74193335e-01 -5.00900686e-01 8.49894166e-01 -1.98540539e-01
-5.81688106e-01 8.20735633e-01 3.85957360e-01 1.10945717e-01
-1.53628170e-01 -4.08589691e-01 -6.91831410e-01 -1.00389755e+00
-1.49524391e-01 2.73238093e-01 4.33064270e-04 -3.36102657e-02
1.74581647e-01 5.45697570e-01 -1.89129341e+00 7.90213048e-01
3.39043945e-01 1.36906338e+00 8.72096568e-02 9.06724572e-01
3.30235630e-01 1.08781314e+00 -3.16359192e-01 -4.58124019e-02
3.90351146e-01 -3.77843022e-01 -7.41995871e-01 2.97211975e-01
7.07929313e-01 -5.65234959e-01 -8.04796517e-01 6.76088989e-01
6.18911326e-01 3.82326931e-01 5.98926187e-01 -1.46656430e+00
-6.98234200e-01 2.42893338e-01 -2.06125140e-01 2.41384953e-01
7.45361686e-01 4.05795246e-01 7.57494867e-01 -6.73053741e-01
7.44144440e-01 1.14262533e+00 2.04241902e-01 6.90122843e-01
-1.20458198e+00 -2.72112310e-01 1.36283562e-01 5.40324926e-01
-1.09983313e+00 -6.28098547e-01 6.43412173e-01 -3.73707086e-01
1.07226312e+00 1.74878061e-01 4.31019932e-01 1.87165332e+00
-1.99078232e-01 1.24531090e+00 1.22305691e+00 -3.12466651e-01
3.51226300e-01 3.00705116e-02 -1.79764107e-01 4.47251350e-01
-1.53115734e-01 1.71562627e-01 -9.65209723e-01 9.61121768e-02
7.62147009e-01 2.25046754e-01 -5.29620409e-01 -6.19065762e-01
-1.54916775e+00 5.69894493e-01 5.69528997e-01 3.63780111e-01
-3.41978341e-01 2.19871864e-01 3.97032917e-01 5.07956266e-01
8.34205449e-02 3.32991332e-01 -3.34688812e-01 -5.32130957e-01
-9.61183012e-01 3.45268287e-02 4.84527975e-01 7.35560894e-01
6.18486106e-01 -2.14183945e-02 -1.08843222e-01 6.01640105e-01
3.11191738e-01 6.27953827e-01 7.22560704e-01 -1.17597306e+00
6.75603390e-01 4.71717119e-01 -5.61194457e-02 -7.84451663e-01
-4.77062166e-01 1.18383512e-01 -7.39248991e-01 4.57123667e-01
6.64588988e-01 2.85660997e-02 -1.19738376e+00 1.91274118e+00
4.77355160e-03 2.64816523e-01 3.51421654e-01 1.26116645e+00
7.79295206e-01 4.21920955e-01 -5.93684278e-02 1.06171273e-01
8.43351603e-01 -6.37412846e-01 -4.65495616e-01 -2.60805726e-01
5.91209173e-01 -4.06956673e-01 8.43720675e-01 7.21039176e-01
-8.74710321e-01 -5.19983709e-01 -1.06468177e+00 -2.53711462e-01
-8.17929029e-01 1.57996550e-01 5.75397551e-01 4.02734071e-01
-1.00480700e+00 5.99151194e-01 -9.61742222e-01 -7.10751057e-01
4.15070683e-01 4.24819767e-01 -1.06202185e+00 -5.84369659e-01
-9.72190619e-01 1.02207911e+00 3.49597514e-01 3.60653669e-01
-1.24873233e+00 -2.30991870e-01 -1.02274656e+00 -2.61499286e-01
4.09165353e-01 -3.94793421e-01 9.11554277e-01 -1.06978214e+00
-1.33717334e+00 7.33294189e-01 2.11360574e-01 -2.19183609e-01
6.73891127e-01 -2.21879527e-01 -3.57165426e-01 4.02181596e-01
-1.28377214e-01 9.43296254e-01 8.64887118e-01 -1.17569649e+00
-4.00422901e-01 -5.56066990e-01 3.83360773e-01 3.55810761e-01
-1.82001695e-01 -3.15707028e-01 -2.86509722e-01 -4.28802222e-01
2.06827447e-01 -9.47143316e-01 3.73183750e-02 2.36918256e-01
-3.89665902e-01 2.09210604e-01 1.05684590e+00 -6.51535332e-01
5.47339141e-01 -2.30108809e+00 7.85815597e-01 1.30986869e-01
-1.36508003e-01 -1.47400079e-02 -3.38269353e-01 4.22588527e-01
-3.02821189e-01 -2.35237062e-01 -3.27231944e-01 -4.78511631e-01
1.57442704e-01 5.58352709e-01 -2.89793938e-01 6.70639515e-01
3.58399987e-01 8.28729808e-01 -8.89743865e-01 -1.90634876e-01
5.61032355e-01 6.19841278e-01 -2.85729021e-01 2.47855797e-01
-1.14723697e-01 7.77819812e-01 -1.80234283e-01 1.11760044e+00
2.69984633e-01 -2.99086049e-02 -8.60425085e-02 -2.34986499e-01
1.37319148e-01 -6.56470880e-02 -1.31008518e+00 2.18663359e+00
-3.11616898e-01 8.44504297e-01 -7.99990296e-02 -1.09830332e+00
5.70728481e-01 4.54445273e-01 6.61469221e-01 -8.86812150e-01
1.25159010e-01 3.30763012e-01 -2.08146274e-01 -8.32633615e-01
4.39789742e-01 1.20950420e-03 -7.31410682e-02 4.43346709e-01
3.88983130e-01 1.67690236e-02 1.82205141e-01 2.89735030e-02
1.31329417e+00 5.07853866e-01 -3.06189712e-02 4.02983934e-01
2.35135883e-01 -2.22470433e-01 1.10687979e-01 6.06903434e-01
-3.45782995e-01 9.89818454e-01 6.30811632e-01 -1.98668882e-01
-8.88927579e-01 -1.05059087e+00 -1.43111810e-01 1.08091307e+00
-3.13340127e-02 -4.49303351e-02 -2.44728476e-01 -7.91357517e-01
-5.31779565e-02 4.22481477e-01 -1.02388978e+00 -3.29291672e-01
-2.66711652e-01 -4.28354084e-01 6.76564515e-01 6.46380365e-01
5.71172774e-01 -1.13418865e+00 -8.76291633e-01 -1.55685946e-01
-1.40932515e-01 -1.28003776e+00 8.16024914e-02 4.98445243e-01
-8.13283801e-01 -1.23942029e+00 -9.17769134e-01 -2.22515926e-01
4.96649504e-01 1.53200299e-01 6.20706260e-01 -3.67709011e-01
-3.08657974e-01 8.65258098e-01 -5.70166528e-01 1.80739030e-01
4.88987826e-02 -1.59727499e-01 6.00489937e-02 3.17530930e-01
2.70039797e-01 -4.82779831e-01 -3.98282677e-01 1.72574565e-01
-1.34006608e+00 -6.71509430e-02 6.87327325e-01 9.03847039e-01
4.10996258e-01 -3.88845772e-01 1.72537342e-01 -3.61424863e-01
2.15380237e-01 -6.68898404e-01 -5.85679300e-02 3.45887452e-01
1.08808808e-01 6.86681643e-02 3.45003545e-01 -6.34434879e-01
-7.99451292e-01 2.29221493e-01 -4.48503561e-04 -8.46966684e-01
-5.82969785e-01 4.44673508e-01 -3.61621469e-01 -2.35933304e-01
7.23564863e-01 1.67411521e-01 -1.74753610e-02 -3.50421548e-01
4.80398923e-01 6.96099579e-01 6.57444000e-01 -4.85344499e-01
3.40320289e-01 6.27687871e-01 9.24483165e-02 -8.78580272e-01
-4.37654912e-01 -2.75422126e-01 -1.06835282e+00 -4.27973300e-01
9.48155522e-01 -8.46398592e-01 -4.10802633e-01 5.54534912e-01
-7.67275393e-01 -4.78199124e-01 -3.94868970e-01 7.83303022e-01
-1.00680315e+00 2.28596479e-01 -2.89345771e-01 -5.81737220e-01
4.66776818e-01 -1.33235371e+00 1.25108242e+00 8.95980224e-02
-5.39319925e-02 -8.67177904e-01 5.34076132e-02 4.52006876e-01
4.07998145e-01 7.55416632e-01 6.86967313e-01 -6.19159281e-01
-6.92351639e-01 -3.66851121e-01 -1.62848473e-01 4.97929603e-01
-5.57371881e-03 -1.33865237e-01 -1.34240270e+00 -2.48739213e-01
-2.77924806e-01 -1.02356946e+00 1.18236232e+00 1.58553813e-02
9.06463623e-01 5.67937829e-02 -2.98247766e-02 5.55791795e-01
1.21385360e+00 -9.81886014e-02 7.94269204e-01 4.02319252e-01
6.07514024e-01 5.80453932e-01 4.37690049e-01 5.39053440e-01
2.87862718e-01 7.35080004e-01 8.06775749e-01 1.15954719e-01
-7.39835426e-02 -1.00041546e-01 5.91599524e-01 1.76704422e-01
-2.69153029e-01 -2.16205850e-01 -9.31823671e-01 4.10845906e-01
-2.08510733e+00 -1.02751184e+00 3.63914073e-01 2.28165460e+00
5.59751928e-01 -1.52627349e-01 1.21775255e-01 4.01992559e-01
2.26468876e-01 2.36909866e-01 -7.15373099e-01 -6.20768405e-02
-5.57795584e-01 -2.65922695e-02 2.78045118e-01 1.60814568e-01
-1.28817677e+00 4.79252607e-01 6.01835394e+00 2.02308938e-01
-1.43218160e+00 1.10302247e-01 4.08234864e-01 -6.94234908e-01
1.60711557e-02 -2.93721706e-01 -2.72132427e-01 3.76259983e-01
1.10480988e+00 4.59043920e-01 4.98804927e-01 6.71442747e-01
-6.38687611e-02 -5.09142518e-01 -1.50659716e+00 1.30720913e+00
5.46791196e-01 -9.46877718e-01 -1.82774454e-01 -2.20522680e-03
5.75431824e-01 2.88073212e-01 3.06802571e-01 3.79470974e-01
-1.04486234e-01 -1.22568011e+00 7.46759593e-01 9.49626803e-01
7.65265822e-01 -4.01430070e-01 6.86390400e-01 1.65190265e-01
-6.91462755e-01 -3.21721971e-01 -1.74464434e-01 1.05244830e-01
8.79992843e-02 1.28152948e-02 -3.96967262e-01 5.20965219e-01
7.34400094e-01 1.14950144e+00 -8.03826988e-01 1.09830868e+00
-1.91883549e-01 -5.15541341e-03 -3.44519854e-01 2.79879659e-01
3.63813013e-01 1.82846501e-01 3.46389204e-01 9.87007141e-01
4.81797308e-01 2.82400884e-02 1.39123976e-01 4.44045782e-01
4.59317714e-02 -3.36950183e-01 -9.94302154e-01 -1.51555270e-01
-3.18335854e-02 1.03562808e+00 -3.85700285e-01 -8.66206512e-02
-5.45553565e-01 1.19240987e+00 2.60295600e-01 7.36707091e-01
-6.58208132e-01 -9.79785323e-02 6.77690923e-01 -8.75427574e-02
1.97639257e-01 -6.14155591e-01 -4.25147964e-03 -1.36829007e+00
1.04825512e-01 -9.23768342e-01 5.99750340e-01 -1.11462867e+00
-1.20421398e+00 4.36963350e-01 2.41969839e-01 -1.40245056e+00
-5.52599847e-01 -1.14041805e+00 -2.05560967e-01 6.06287777e-01
-1.44482636e+00 -1.27848518e+00 -4.67491895e-01 1.02968252e+00
2.67772734e-01 -2.98838168e-01 8.62755179e-01 4.46789324e-01
-5.57734311e-01 3.57952893e-01 1.71752602e-01 -1.08754169e-02
9.03443754e-01 -1.07000959e+00 -5.02028286e-01 6.56425774e-01
3.23602319e-01 2.95943260e-01 5.00674784e-01 -1.78184569e-01
-1.79729688e+00 -7.44819939e-01 3.25829953e-01 -5.47723591e-01
6.15697563e-01 -3.52913648e-01 -7.70591795e-01 8.13632488e-01
1.59854800e-01 3.96521986e-01 7.42338657e-01 -1.18052378e-01
-4.83098358e-01 1.12154610e-01 -1.11626184e+00 4.30024743e-01
9.19498503e-01 -9.79787171e-01 -5.21426558e-01 3.21078271e-01
1.19857930e-01 -5.04582345e-01 -8.19947720e-01 3.51646662e-01
6.58010483e-01 -1.12439001e+00 8.51455867e-01 -7.66956449e-01
5.49749911e-01 -4.78311539e-01 -8.33450198e-01 -1.25077105e+00
1.32308021e-01 -2.49148026e-01 -4.73635286e-01 9.00121212e-01
3.67363781e-01 -2.96559364e-01 6.85406029e-01 8.44322443e-01
8.63322318e-02 -6.25348747e-01 -1.41920590e+00 -6.30339682e-01
-1.82176709e-01 -7.45943367e-01 2.21273094e-01 8.30830634e-01
4.08400148e-02 -2.09613219e-01 -6.19742990e-01 1.08724721e-01
2.74145067e-01 -3.45346555e-02 6.85014784e-01 -7.24255323e-01
-3.03918421e-01 -3.32429379e-01 -1.06651390e+00 -9.04591203e-01
1.70428216e-01 -7.66362250e-01 9.78208557e-02 -1.43256593e+00
5.99167822e-03 -9.61379334e-02 -4.34726655e-01 8.32316458e-01
3.52325946e-01 5.87910175e-01 3.47112566e-01 1.59612313e-01
-7.96403110e-01 8.08655500e-01 1.08338785e+00 -3.59522730e-01
7.96306357e-02 -2.99123019e-01 -4.05486643e-01 4.69620287e-01
4.90586847e-01 -1.17044367e-01 -4.66921270e-01 -6.01682484e-01
1.24351280e-02 1.12268031e-01 7.81078517e-01 -1.21923125e+00
2.80727893e-01 -1.63961425e-01 7.20567286e-01 -3.28611940e-01
9.53597188e-01 -9.73657072e-01 -9.02338978e-03 -2.05915850e-02
-5.32269120e-01 -1.69715881e-01 1.31637171e-01 6.11413538e-01
-4.31982130e-01 -1.13871634e-01 5.32227874e-01 -1.38535082e-01
-1.21003473e+00 2.79779047e-01 -1.81089729e-01 -1.45690620e-01
1.19239295e+00 -4.74807084e-01 -5.61068594e-01 -3.96870762e-01
-1.18883836e+00 7.90517330e-02 6.43055618e-01 6.97613537e-01
7.88672686e-01 -1.50120246e+00 -4.97152627e-01 5.25632322e-01
4.95019317e-01 -3.62453520e-01 4.50950384e-01 1.15534759e+00
-1.86650261e-01 3.92536700e-01 -7.13477254e-01 -7.06374943e-01
-1.00160563e+00 4.28209662e-01 3.54408622e-01 2.03068748e-01
-5.03159702e-01 7.05262303e-01 -1.91446334e-01 -3.26680392e-01
5.28723598e-01 -2.69398272e-01 -6.86084107e-02 2.99164325e-01
6.63439512e-01 2.46112734e-01 8.28084126e-02 -8.65383506e-01
-3.64745051e-01 4.67515588e-01 2.62258589e-01 -4.43903774e-01
1.24978471e+00 -1.44622028e-01 1.83442697e-01 8.51684391e-01
1.40671933e+00 -6.84911907e-01 -1.58095145e+00 -3.42617959e-01
-1.10819325e-01 -7.43649781e-01 -3.95598523e-02 -8.71235132e-01
-1.05698252e+00 1.23385835e+00 9.52868938e-01 1.00105762e-01
1.28482771e+00 7.48749375e-02 3.36192966e-01 6.68161571e-01
4.26357925e-01 -9.77988124e-01 4.04500037e-01 4.29732770e-01
1.10343790e+00 -1.66679895e+00 -1.44096956e-01 2.29081601e-01
-9.35537755e-01 1.29170501e+00 4.31436300e-01 4.17940244e-02
4.22035575e-01 -1.51024787e-02 9.71478373e-02 9.88044124e-03
-7.54545867e-01 -4.60678458e-01 3.31642330e-01 7.61442363e-01
2.86776096e-01 -1.80529520e-01 4.58297849e-01 1.79371238e-01
2.55544007e-01 -3.92470602e-03 4.26311165e-01 1.32035744e+00
-2.13229254e-01 -9.71517265e-01 -3.37624699e-01 3.50566715e-01
-1.36811852e-01 2.68670976e-01 -5.76246321e-01 8.80447984e-01
2.72478670e-01 8.85473728e-01 -5.14772944e-02 -5.79166770e-01
3.10178399e-01 2.04748183e-01 7.64839292e-01 -4.63986278e-01
-1.19857743e-01 -2.53450394e-01 9.34519917e-02 -9.90662992e-01
-8.54423940e-01 -9.50051010e-01 -1.01664793e+00 -7.77548999e-02
1.21152155e-01 -4.19614494e-01 7.49810994e-01 1.05243039e+00
2.01429948e-01 3.84172946e-01 3.60601515e-01 -1.44314885e+00
-2.96594322e-01 -9.21056926e-01 -7.34000027e-01 5.60412288e-01
6.33770227e-01 -1.05801952e+00 -3.01359594e-01 1.16838649e-01] | [8.149580955505371, 0.6654144525527954] |
aec5de6d-06ad-4bc3-9770-f1a9fbe3e2ea | semi-supervised-deep-large-baseline | 2212.02763 | null | https://arxiv.org/abs/2212.02763v1 | https://arxiv.org/pdf/2212.02763v1.pdf | Semi-supervised Deep Large-baseline Homography Estimation with Progressive Equivalence Constraint | Homography estimation is erroneous in the case of large-baseline due to the low image overlay and limited receptive field. To address it, we propose a progressive estimation strategy by converting large-baseline homography into multiple intermediate ones, cumulatively multiplying these intermediate items can reconstruct the initial homography. Meanwhile, a semi-supervised homography identity loss, which consists of two components: a supervised objective and an unsupervised objective, is introduced. The first supervised loss is acting to optimize intermediate homographies, while the second unsupervised one helps to estimate a large-baseline homography without photometric losses. To validate our method, we propose a large-scale dataset that covers regular and challenging scenes. Experiments show that our method achieves state-of-the-art performance in large-baseline scenes while keeping competitive performance in small-baseline scenes. Code and dataset are available at https://github.com/megvii-research/LBHomo. | ['Shuaicheng Liu', 'Songchen Han', 'Yuhang Lu', 'Haipeng Li', 'Hai Jiang'] | 2022-12-06 | null | null | null | null | ['homography-estimation'] | ['computer-vision'] | [ 2.03899905e-01 2.84316782e-02 4.12045792e-02 -3.35642457e-01
-9.24173057e-01 -2.68307686e-01 4.71831352e-01 -3.67149204e-01
-3.15267473e-01 5.36749482e-01 2.80020922e-01 3.14953566e-01
1.75930977e-01 -6.53412640e-01 -9.95778739e-01 -7.43668735e-01
2.68579990e-01 2.44387999e-01 4.11467761e-01 -4.20830362e-02
3.55752319e-01 2.44991943e-01 -1.29474533e+00 2.30002757e-02
1.09598923e+00 8.99399579e-01 4.55191046e-01 2.53160030e-01
4.10073280e-01 6.20960653e-01 -1.47309676e-02 -5.34146845e-01
8.19178760e-01 -5.15997350e-01 -6.61185503e-01 3.34888428e-01
9.93842602e-01 -6.59245491e-01 -5.65095007e-01 1.42845392e+00
4.27151889e-01 1.54676720e-01 3.93578291e-01 -1.06151497e+00
-4.59282875e-01 1.83892131e-01 -8.28066826e-01 -3.06661159e-01
2.80673385e-01 3.94304067e-01 1.18499231e+00 -9.00476933e-01
7.74444580e-01 1.35893285e+00 5.91422558e-01 -4.31789495e-02
-1.46489489e+00 -6.97580516e-01 -4.93227132e-02 3.26781571e-02
-1.39964533e+00 -5.25271177e-01 9.01179731e-01 -4.23387617e-01
6.90860271e-01 -1.52460650e-01 5.73456585e-01 7.84460068e-01
1.58406034e-01 6.93431139e-01 1.38573360e+00 -3.20838839e-01
-1.98672667e-01 -1.00241683e-01 -1.32008240e-01 7.93773353e-01
1.45788789e-01 3.27222496e-01 -3.92656595e-01 1.41850755e-01
1.05419624e+00 6.85253665e-02 -5.65547526e-01 -6.57719135e-01
-1.26550841e+00 5.74529111e-01 7.38539219e-01 -7.29399696e-02
-3.80520225e-01 3.13766114e-02 -1.09955907e-01 2.57286668e-01
2.79490113e-01 4.75024700e-01 -2.10752171e-02 1.01327062e-01
-9.35546219e-01 1.73318431e-01 5.77434599e-01 1.12984383e+00
1.30503738e+00 -2.30090097e-01 1.21446349e-01 9.72436011e-01
2.01951027e-01 6.27195835e-01 2.40524575e-01 -1.22281575e+00
6.39712572e-01 5.05880415e-01 2.41323113e-02 -1.20326078e+00
-5.08952886e-02 -3.16315740e-01 -9.64427710e-01 1.46732599e-01
4.12633985e-01 1.29819825e-01 -8.49441350e-01 1.68639755e+00
3.00676227e-01 2.80271620e-02 -2.74191469e-01 1.09394610e+00
4.92012024e-01 5.84057570e-01 -4.89121199e-01 -8.72730017e-02
1.15130186e+00 -1.36988151e+00 -6.21747553e-01 -3.94621372e-01
1.19457483e-01 -1.12648904e+00 1.10064721e+00 3.76942545e-01
-1.27995098e+00 -4.79345173e-01 -1.18076289e+00 -7.03777492e-01
-2.29566339e-02 3.41183811e-01 4.02475744e-01 1.67463109e-01
-9.67674732e-01 5.68913877e-01 -7.48426139e-01 -2.57603526e-01
2.37616181e-01 2.56652862e-01 -3.11442614e-01 -1.63211584e-01
-9.70014751e-01 6.22033238e-01 3.49982381e-01 -1.28673539e-01
-8.10436070e-01 -5.66470087e-01 -7.62958825e-01 -3.37504111e-02
5.09669483e-01 -8.25579226e-01 9.03668582e-01 -6.50091648e-01
-1.52215540e+00 1.14355981e+00 4.14434262e-02 -3.84258240e-01
9.44814026e-01 -2.94701070e-01 -8.23060274e-02 3.17132622e-01
2.06829965e-01 9.08493340e-01 9.36759055e-01 -1.48357928e+00
-5.28092802e-01 -4.66824293e-01 -3.41248289e-02 4.69012886e-01
-1.60239607e-01 -3.49079877e-01 -9.17008698e-01 -6.04994118e-01
5.15869081e-01 -1.04749060e+00 -1.00226633e-01 7.42506161e-02
-5.48189104e-01 4.22479779e-01 3.38106304e-01 -8.65558326e-01
1.04482615e+00 -2.18328834e+00 2.95964420e-01 1.01833619e-01
2.89344162e-01 -2.62856986e-02 -3.81187089e-02 3.89673382e-01
8.89947191e-02 -3.50751758e-01 -5.10249257e-01 -5.62674582e-01
-1.86068296e-01 -1.77833080e-01 -4.64259714e-01 5.44756234e-01
-8.25167596e-02 7.40621865e-01 -8.15532386e-01 -2.98327297e-01
3.33295316e-01 4.63446140e-01 -6.10103190e-01 2.83947706e-01
-8.01519782e-04 5.76313913e-01 -2.79103249e-01 4.05931562e-01
1.05921376e+00 -3.23444366e-01 5.45593388e-02 -6.38593674e-01
-3.79340529e-01 1.19793981e-01 -1.20401025e+00 2.01485944e+00
-3.35615277e-01 4.81556982e-01 1.07728839e-01 -5.57330906e-01
9.68231380e-01 -2.95110136e-01 3.90082747e-01 -5.82506716e-01
3.00808161e-01 4.09941047e-01 -1.65968850e-01 -1.46982506e-01
4.60761577e-01 1.75788760e-01 3.33356053e-01 6.41708523e-02
1.32799551e-01 -5.29056609e-01 2.67211527e-01 2.81378359e-01
7.21836030e-01 2.60530829e-01 2.50946999e-01 -2.82868415e-01
6.46285176e-01 -2.37748384e-01 5.78025401e-01 4.05802637e-01
-1.42593086e-01 1.06444204e+00 5.53160846e-01 -2.64780194e-01
-1.14512885e+00 -1.14652216e+00 -9.48917046e-02 4.20783639e-01
7.46865273e-01 -4.31519657e-01 -6.37725174e-01 -3.61675143e-01
-8.95982515e-03 2.98100471e-01 -4.00293440e-01 -9.82199237e-02
-4.70373303e-01 -5.47244191e-01 1.99898675e-01 2.32295737e-01
1.09991586e+00 -7.04389870e-01 -3.76482755e-01 -7.99860358e-02
-4.33914095e-01 -1.24418068e+00 -8.84655952e-01 -2.12979674e-01
-9.28460538e-01 -1.08061326e+00 -8.31397235e-01 -7.49164760e-01
9.62781727e-01 5.60111821e-01 8.90888751e-01 -6.90016225e-02
-1.88144878e-01 -4.22509126e-02 -1.69548634e-02 8.42046738e-02
-1.75532605e-02 -7.71617666e-02 -1.39437750e-01 2.29217187e-01
3.55430134e-02 -9.01066542e-01 -8.84400606e-01 6.25228643e-01
-8.38912189e-01 3.47979218e-01 8.98078620e-01 9.32194054e-01
9.86651003e-01 -1.97758436e-01 -1.67999759e-01 -7.82892764e-01
1.51494041e-01 -7.97827914e-03 -1.05917037e+00 2.28506625e-01
-7.40138352e-01 7.20322505e-02 6.94990933e-01 -3.11946154e-01
-1.17931128e+00 3.26261789e-01 -2.28249803e-02 -6.62757099e-01
1.07958756e-01 1.53947966e-02 -3.65754366e-01 -3.80833387e-01
4.53992516e-01 3.63047272e-01 6.85652345e-03 -6.16907060e-01
3.62894744e-01 2.57748485e-01 8.48771274e-01 -4.73192930e-01
1.17852366e+00 9.08026397e-01 2.69561168e-02 -7.30883420e-01
-7.61499703e-01 -6.29729211e-01 -8.07823122e-01 -1.29490048e-01
8.56742203e-01 -1.17446780e+00 -4.31653261e-01 8.03900421e-01
-1.02423859e+00 -4.02807236e-01 -1.48131892e-01 7.30827153e-01
-7.24951744e-01 5.80917537e-01 -5.95051825e-01 -3.67330581e-01
-2.13201880e-01 -1.25502038e+00 1.20918965e+00 1.45753518e-01
2.57596225e-01 -6.86857045e-01 1.99282110e-01 5.94945729e-01
2.75382083e-02 3.57404016e-02 4.55706149e-01 -4.65962104e-02
-1.26430929e+00 -3.48083046e-03 -5.24810612e-01 4.82636899e-01
-4.23485227e-02 -1.91700280e-01 -1.00548482e+00 -4.88514334e-01
2.05489233e-01 -5.75046718e-01 1.06040800e+00 3.61078411e-01
9.12372410e-01 -1.46246389e-01 -6.70781732e-03 1.36594093e+00
1.64882660e+00 9.04426798e-02 9.48798954e-01 3.87564033e-01
1.04350770e+00 5.84771395e-01 6.70486987e-01 3.08575869e-01
4.72042561e-01 7.98346221e-01 2.91825980e-01 -2.02520967e-01
-4.06263262e-01 -5.72057307e-01 2.66882092e-01 9.32024777e-01
-1.17056616e-01 4.98726070e-02 -7.14542210e-01 3.62612754e-01
-1.76981699e+00 -7.50275552e-01 -6.12082221e-02 2.53510642e+00
7.71227181e-01 -6.94409991e-03 2.26134229e-02 -2.60398835e-01
6.22622073e-01 5.31455874e-01 -7.39055634e-01 2.98956960e-01
-3.90879393e-01 -4.98694569e-01 6.95044696e-01 7.77162373e-01
-1.01559234e+00 1.15041506e+00 5.37370157e+00 8.92443240e-01
-1.14457178e+00 5.49277849e-02 6.16393626e-01 4.41230228e-03
-2.59975821e-01 3.94089013e-01 -7.06461728e-01 4.22939509e-01
1.13067366e-01 -1.12162150e-01 7.53952086e-01 6.11209095e-01
-5.34020364e-02 -1.92861080e-01 -8.62346649e-01 1.22606242e+00
2.76330322e-01 -1.17605996e+00 7.09151551e-02 2.61994869e-01
1.03411198e+00 2.05397725e-01 4.98874001e-02 -1.04394756e-01
1.61991075e-01 -4.97656912e-01 6.04766607e-01 4.34924930e-01
9.09275770e-01 -4.78148073e-01 4.24974889e-01 1.91542014e-01
-1.21674740e+00 2.42115542e-01 -4.87926871e-01 1.26874268e-01
2.43216872e-01 5.63680172e-01 -1.81166232e-01 7.70775259e-01
7.27614880e-01 9.71464157e-01 -6.83893681e-01 1.18063939e+00
-4.61864024e-01 3.56830657e-02 -4.09162909e-01 5.91137648e-01
1.00668155e-01 -8.89625549e-01 6.02229476e-01 8.12432230e-01
2.19164893e-01 1.63343266e-01 3.49021763e-01 9.31465805e-01
-2.71046758e-01 1.44718885e-01 -5.81796944e-01 3.14362526e-01
4.17922109e-01 1.32396662e+00 -6.38451636e-01 -2.07986519e-01
-4.27863240e-01 1.24201882e+00 3.82376343e-01 4.62928146e-01
-5.92143953e-01 -4.47775275e-01 4.46152687e-01 9.45006013e-02
4.28713709e-02 -2.52696157e-01 -3.41778517e-01 -1.55807316e+00
2.69933552e-01 -8.27860355e-01 1.47363022e-01 -8.60029459e-01
-1.14443111e+00 5.86929977e-01 4.91496138e-02 -1.47018945e+00
-2.42737960e-02 -4.25344318e-01 -5.97756147e-01 7.60727704e-01
-1.47764933e+00 -1.27595389e+00 -6.94288731e-01 4.99512166e-01
4.63365585e-01 -1.35769956e-02 1.99340224e-01 4.86428112e-01
-4.91614699e-01 5.62176406e-01 2.79039711e-01 1.94688123e-02
1.10719049e+00 -1.13283789e+00 3.44826609e-01 1.18707943e+00
-1.11566387e-01 5.96469343e-01 5.28771281e-01 -6.59539402e-01
-1.32548463e+00 -7.99705446e-01 7.91073143e-01 -1.62042663e-01
7.84872472e-01 -4.49956924e-01 -9.27363336e-01 7.51971960e-01
1.12717152e-01 -1.79282114e-01 5.72830848e-02 -3.40645939e-01
-5.19227564e-01 -3.08683038e-01 -9.71311569e-01 6.58186913e-01
1.29587984e+00 -5.88536978e-01 -3.33547264e-01 3.28410774e-01
6.46584213e-01 -6.12679243e-01 -8.57164383e-01 3.91148299e-01
7.44804621e-01 -1.31613457e+00 1.06603551e+00 2.01194167e-01
7.77173221e-01 -4.75429058e-01 -1.56757429e-01 -1.23727500e+00
-3.37686658e-01 -9.75768626e-01 7.91867152e-02 1.12299526e+00
1.53243124e-01 -8.21318924e-01 5.48792124e-01 3.08725864e-01
-1.37608364e-01 -7.04461634e-01 -5.26150048e-01 -1.06050336e+00
-1.54733285e-01 1.40688717e-01 3.93733501e-01 9.75410938e-01
-3.98607045e-01 4.13926393e-01 -7.31840611e-01 2.12890014e-01
1.08465803e+00 4.04459149e-01 1.03919804e+00 -7.97940075e-01
-3.67571741e-01 -4.85567331e-01 -4.30316001e-01 -1.57542026e+00
1.63407028e-02 -7.52517223e-01 1.89101949e-01 -1.14572656e+00
3.48994583e-01 -1.69920623e-01 8.68719891e-02 2.05770567e-01
-1.59777805e-01 4.25494641e-01 2.86556333e-01 6.94139004e-01
-3.99675518e-01 8.60648334e-01 1.53495789e+00 3.80881242e-02
-2.98832923e-01 -2.49971315e-01 -4.23345238e-01 7.78999031e-01
5.73200107e-01 -2.58603126e-01 -5.82110465e-01 -6.68920755e-01
-1.11024618e-01 -8.15807953e-02 2.74205118e-01 -1.22249150e+00
4.46019769e-01 -6.81957826e-02 2.66323060e-01 -5.97274244e-01
5.27981341e-01 -6.40640795e-01 1.94104776e-01 3.08160186e-01
-2.92187244e-01 -8.92308354e-02 -1.85870886e-01 4.35200542e-01
-5.03971338e-01 7.93977529e-02 1.09979594e+00 1.06847733e-01
-5.92690170e-01 6.97064579e-01 4.62942868e-01 1.51376739e-01
6.75533235e-01 -2.05961421e-01 -5.05207539e-01 -4.36075568e-01
-2.83532023e-01 4.03994471e-01 1.04632652e+00 3.47943455e-01
6.01063550e-01 -1.25868738e+00 -5.93282163e-01 4.21009064e-01
1.93329886e-01 9.57396030e-02 3.05945516e-01 1.05064046e+00
-8.60606194e-01 1.60537168e-01 -4.91938978e-01 -5.80137432e-01
-1.12952948e+00 3.87522727e-01 2.24766329e-01 -2.13859692e-01
-7.04773903e-01 7.41296291e-01 8.32780361e-01 -3.88042718e-01
3.44262958e-01 -1.74085736e-01 1.84595183e-01 -1.40588090e-01
3.82949501e-01 4.95740086e-01 -1.40742227e-01 -9.40323055e-01
-7.99614117e-02 1.03644896e+00 -1.23224109e-01 -3.12565684e-01
1.09190428e+00 -4.81977940e-01 -1.58669636e-01 2.78124958e-01
1.67883146e+00 3.15863639e-02 -1.65470791e+00 -2.69219667e-01
-4.07832950e-01 -9.48115349e-01 8.29981938e-02 -4.50334042e-01
-1.16350746e+00 9.58681524e-01 5.78386784e-01 -3.35710138e-01
1.40670145e+00 -2.54040033e-01 9.89940166e-01 5.23184121e-01
3.61637622e-01 -9.29293811e-01 2.08514228e-01 6.49706185e-01
1.14106119e+00 -1.41725457e+00 2.69194514e-01 -8.11810732e-01
-6.64478242e-01 8.58182728e-01 7.57126510e-01 -3.30822617e-01
4.20540959e-01 -1.66131869e-01 1.35930166e-01 -1.74998328e-01
-4.67405587e-01 -3.16279531e-01 4.22891825e-01 3.65387470e-01
1.79424807e-01 -2.35000208e-01 -2.91241437e-01 -2.94115003e-02
-3.55150461e-01 -1.34848356e-01 3.22975844e-01 5.23767531e-01
-2.35861063e-01 -1.14487612e+00 -3.18381488e-01 2.16238573e-01
-1.86821803e-01 -1.96390435e-01 -4.65411901e-01 6.84407890e-01
-5.44216447e-02 6.23486459e-01 3.75083312e-02 -4.96905655e-01
3.96265864e-01 -4.43852276e-01 5.63451767e-01 -3.38436991e-01
-2.60538459e-01 5.52982330e-01 -1.10440731e-01 -9.02192116e-01
-2.63191253e-01 -4.00835752e-01 -9.31693375e-01 -3.03981811e-01
-9.44163203e-02 -2.76211828e-01 4.12343264e-01 5.00733674e-01
2.42958888e-01 4.06364091e-02 9.01496649e-01 -1.11322832e+00
-5.50019026e-01 -7.17859805e-01 -6.34350777e-01 8.15541565e-01
3.66180092e-01 -6.23085678e-01 -6.32799089e-01 3.08252424e-01] | [8.742971420288086, -2.295504331588745] |
31070792-ba56-4a62-98bc-b289b06c395a | electrophysiological-indicators-of-gesture | 1811.05058 | null | http://arxiv.org/abs/1811.05058v1 | http://arxiv.org/pdf/1811.05058v1.pdf | Electrophysiological indicators of gesture perception | Background: While there has been abundant research concerning neurological
responses to gesture generation, the time course of gesture processing is not
well understood. Specifically, it is not clear if or how particular
characteristics within the kinematic execution of gestures capture attention
and aid in the classification of gestures with communicative intent. If indeed
key features of gestures with perceptual saliency exist, such features could
help form the basis of a compact representation of the gestures in memory.
Methods: This study used a set of available gesture videos as stimuli. The
timing for salient features of performed gestures was determined by isolating
inflection points in the hands' motion trajectories. Participants passively
viewed the gesture videos while continuous EEG data was collected. We focused
on mu oscillations (10 Hz) and used linear regression to test for associations
between the timing of mu oscillations and inflection points in motion
trajectories. Results: Peaks in the EEG signals at central and occipital
electrodes were used to isolate the salient events within each gesture. EEG
power oscillations were detected 343 and 400ms on average after inflection
points at occipital and central electrodes, respectively. A regression model
showed that inflection points in the motion trajectories strongly predicted
subsequent mu oscillations (R^2=0.961, p<.01). Conclusion: The results suggest
that coordinated activity in the visual and motor cortices are highly
correlated with key motion components within gesture trajectories. These points
may be associated with neural signatures used to encode gestures in memory for
later identification and even recognition. | [] | 2018-11-13 | null | null | null | null | ['gesture-generation'] | ['robots'] | [ 1.74011871e-01 -5.52710295e-01 -3.54604930e-01 -6.93397671e-02
-4.85687882e-01 -5.94624341e-01 6.60441458e-01 1.38570622e-01
-5.58274269e-01 3.39233965e-01 7.82457471e-01 1.95816889e-01
-4.40883994e-01 4.17942501e-04 -4.53480870e-01 -7.50017881e-01
-8.03515315e-01 -2.47798905e-01 6.60064146e-02 1.06483154e-01
8.37304175e-01 6.72772586e-01 -1.38771641e+00 6.49997234e-01
3.44870985e-01 8.60007048e-01 3.45289439e-01 7.53997922e-01
2.39256009e-01 2.55435944e-01 -7.90884256e-01 1.83511913e-01
-1.66927665e-01 -1.04633498e+00 -5.26866972e-01 -2.21029967e-01
-1.50474578e-01 -4.26650345e-01 -3.34895670e-01 7.64314890e-01
6.93308413e-01 3.00078839e-01 7.76248813e-01 -1.18346107e+00
-2.50425190e-01 6.15462303e-01 -5.62879443e-01 1.01719761e+00
6.08945608e-01 3.45477015e-01 7.00948656e-01 -9.08219874e-01
9.06529546e-01 9.04995382e-01 4.29058135e-01 5.89950323e-01
-1.02947569e+00 -8.38385761e-01 7.08541870e-02 6.69256926e-01
-1.36703134e+00 -4.62350726e-01 7.71387935e-01 -5.84658325e-01
1.34402430e+00 5.50421953e-01 1.15585709e+00 1.58213222e+00
4.96796072e-01 5.59120595e-01 9.93141830e-01 -2.18010128e-01
-1.51275378e-02 -2.60318339e-01 5.53855240e-01 9.17304978e-02
-1.14523053e-01 2.72009343e-01 -1.27132010e+00 -1.83066323e-01
9.93505418e-01 -2.86750272e-02 -6.97110116e-01 5.95769644e-01
-1.53883052e+00 4.40474182e-01 4.30473685e-01 9.32482898e-01
-8.00248623e-01 2.53537804e-01 2.95329273e-01 7.58247003e-02
1.62023067e-01 3.05289537e-01 1.91924833e-02 -8.84379983e-01
-1.03762865e+00 -1.10834047e-01 6.41522780e-02 5.12217820e-01
7.66241085e-03 -1.16783455e-01 -3.28041792e-01 2.58792996e-01
1.53484002e-01 4.53583062e-01 8.26849341e-01 -5.78477859e-01
2.64990896e-01 4.79156494e-01 -2.71603316e-01 -1.00474548e+00
-1.00971234e+00 2.15821072e-01 -1.84022263e-01 4.20389101e-02
6.22821689e-01 -1.85778007e-01 -5.42927682e-01 1.69660640e+00
-1.37604952e-01 2.16067821e-01 -4.56951529e-01 1.36547863e+00
5.73572993e-01 4.32572454e-01 4.45130646e-01 -5.77777803e-01
1.47004092e+00 1.47469148e-01 -7.11883962e-01 3.46226729e-02
2.02511728e-01 -4.02245015e-01 1.01013219e+00 1.10704660e-01
-9.74862516e-01 -4.74274725e-01 -7.86569118e-01 3.53236765e-01
1.18635625e-01 -5.58511503e-02 5.42159379e-01 4.19354111e-01
-5.66034496e-01 7.46962249e-01 -1.48338521e+00 -5.52595496e-01
2.66915470e-01 6.08331084e-01 -2.67589360e-01 7.99474001e-01
-8.81254137e-01 1.14517879e+00 5.36203235e-02 4.82670248e-01
-6.37808621e-01 -3.59988362e-01 -1.32336244e-01 -1.71782389e-01
-3.03681970e-01 -1.98889002e-01 8.57581019e-01 -1.16933155e+00
-1.23206496e+00 6.17234468e-01 -4.59029078e-01 2.12301999e-01
-2.37229485e-02 -1.48192853e-01 -5.14890373e-01 7.26210713e-01
-1.84171811e-01 5.09634256e-01 9.33590114e-01 -7.02154577e-01
-3.16132545e-01 -8.81288707e-01 -4.84569341e-01 4.19352442e-01
-9.10382867e-02 6.90911949e-01 1.01173036e-02 -8.46133053e-01
4.13972437e-01 -8.26528549e-01 4.24284071e-01 -3.94793451e-01
-3.17443572e-02 -2.87765265e-01 2.41969198e-01 -9.23753560e-01
1.39307153e+00 -2.10523486e+00 2.33292535e-01 5.36414623e-01
1.54702261e-01 -4.49599713e-01 -2.31275856e-01 4.73194867e-01
-1.24756426e-01 6.20056540e-02 4.48507443e-02 7.69831181e-01
9.23062966e-04 -4.24972683e-01 -4.18325633e-01 1.02458107e+00
8.39979127e-02 1.40088248e+00 -8.61317635e-01 4.19603512e-02
2.84938551e-02 5.87175250e-01 -3.53465587e-01 1.07714318e-01
1.13130398e-01 9.81524885e-01 -2.36254364e-01 6.92937076e-01
4.74083349e-02 3.07862431e-01 1.05358161e-01 -2.76241183e-01
-3.13341916e-01 5.81318915e-01 -7.12229788e-01 1.62904215e+00
2.87532955e-01 1.63962209e+00 -1.18738130e-01 -7.01798975e-01
5.56398153e-01 6.74319625e-01 6.04592443e-01 -1.03686440e+00
6.62248731e-01 2.61672497e-01 9.85645592e-01 -9.76515353e-01
1.08943544e-02 -1.27601281e-01 2.41424125e-02 5.35324156e-01
-5.98087907e-02 1.88168332e-01 -2.74520610e-02 -2.66268313e-01
1.07183981e+00 1.78449646e-01 4.92905825e-02 -1.16903014e-01
-2.37473562e-01 4.18050475e-02 -7.45475143e-02 4.64093417e-01
-1.08188257e-01 4.93667960e-01 5.20822942e-01 -6.55112863e-02
-2.87981182e-01 -1.28906643e+00 -2.60405004e-01 1.04464531e+00
2.84784913e-01 -3.19813251e-01 -6.50638878e-01 5.10652829e-03
-3.55184019e-01 7.31691539e-01 -6.68785214e-01 -3.99664104e-01
-8.33029211e-01 -8.14418972e-01 4.46192801e-01 8.00714850e-01
4.22271714e-02 -1.75891447e+00 -1.64045453e+00 9.84068438e-02
-3.18774313e-01 -9.16390717e-01 -3.82019043e-01 3.18746895e-01
-1.07299376e+00 -1.20364082e+00 -5.73429227e-01 -5.21154344e-01
6.12385154e-01 -7.50334039e-02 3.73939335e-01 -2.05131993e-01
-5.47279596e-01 7.64335752e-01 -4.64878410e-01 -3.59627724e-01
2.04136103e-01 -3.63589376e-01 4.41783220e-02 -1.73245594e-01
7.83058941e-01 -7.72703648e-01 -6.18078291e-01 9.97579396e-02
-6.76860332e-01 -2.36129403e-01 4.54340518e-01 4.13227558e-01
2.61057615e-01 -4.24970329e-01 4.17350292e-01 3.88015211e-01
1.00860655e+00 -5.66585839e-01 2.86576711e-02 -8.88760313e-02
-1.39708128e-02 -2.34956384e-01 8.01422969e-02 -1.07936096e+00
-5.86333632e-01 -2.70075649e-01 3.03496122e-01 -2.34299019e-01
-4.21222866e-01 3.17232549e-01 1.62890136e-01 6.47761524e-02
7.17779875e-01 4.09135610e-01 -2.29340687e-01 9.63188335e-02
-1.51151761e-01 5.77537298e-01 5.99279106e-01 -3.29598337e-01
-1.02761418e-01 3.46946299e-01 -2.49783739e-01 -1.31780016e+00
2.65110105e-01 -4.01567221e-01 -9.45679307e-01 -7.52271533e-01
1.08558416e+00 -5.69076955e-01 -9.49310839e-01 2.63902515e-01
-1.25363410e+00 -5.93502820e-01 2.78706044e-01 1.31006646e+00
-6.43818796e-01 8.16328675e-02 -3.91848445e-01 -9.83994424e-01
-4.07098621e-01 -9.66931581e-01 7.80364454e-01 2.35867828e-01
-1.21390331e+00 -3.32704455e-01 1.07455634e-01 -3.58581066e-01
-7.49937370e-02 4.02547807e-01 7.58087277e-01 -6.36367679e-01
-1.75132573e-01 -2.30408445e-01 2.87739784e-01 -6.22307062e-01
1.21038236e-01 4.14072126e-02 -1.00842047e+00 2.65609264e-01
3.28553498e-01 1.22490367e-02 4.00310397e-01 8.58470321e-01
7.93620586e-01 -4.32314388e-02 -5.06520748e-01 3.39187056e-01
8.27882290e-01 8.34309638e-01 5.79227149e-01 -1.28870860e-01
2.58902401e-01 8.74457896e-01 1.08689882e-01 1.97995201e-01
-2.87611246e-01 7.16270506e-01 2.59205610e-01 5.42866111e-01
1.08298779e-01 1.27936780e-01 8.17557871e-01 5.22287130e-01
-7.15813279e-01 2.49466121e-01 -9.75645542e-01 6.53144062e-01
-1.52334917e+00 -1.12529755e+00 -7.98300326e-01 2.37042284e+00
3.98571581e-01 -6.33327989e-03 9.43871215e-02 1.05261549e-01
5.47449470e-01 -1.63030177e-01 -6.43323481e-01 -3.13111246e-01
1.42954168e-05 5.08626938e-01 3.29979211e-02 1.27208740e-01
-6.02793396e-01 5.93006611e-01 6.01149702e+00 1.45809472e-01
-1.30438268e+00 2.13568568e-01 -6.86802119e-02 -9.74668980e-01
2.35048950e-01 -5.07706225e-01 -2.63193578e-01 4.60634381e-01
1.40041292e+00 7.44111612e-02 7.57133067e-01 9.81571153e-02
7.61259615e-01 -5.44925094e-01 -1.29246473e+00 1.08948815e+00
4.91224043e-02 -9.11021292e-01 -4.32089716e-01 2.35844702e-02
1.56143978e-01 1.13515705e-01 2.83661559e-02 -2.84295231e-01
-6.80572093e-01 -1.15526676e+00 8.98189664e-01 7.47395754e-01
8.17386508e-01 -5.67840576e-01 1.94208071e-01 4.84639198e-01
-1.30397761e+00 2.17586104e-02 -7.34267533e-02 -3.92431915e-01
2.59469450e-01 -3.70666027e-01 -7.14708388e-01 -2.46106371e-01
6.31801903e-01 4.66621488e-01 -2.33525649e-01 7.67407537e-01
-3.92374218e-01 7.61956990e-01 -3.11875075e-01 -5.60983837e-01
1.77784696e-01 2.39829063e-01 8.78304780e-01 1.18262398e+00
5.89494109e-01 6.86311543e-01 -7.17128634e-01 1.04331112e+00
5.76928258e-01 4.37800772e-03 -3.99464697e-01 -5.19801438e-01
5.47373712e-01 9.33993995e-01 -1.40373325e+00 -4.70193103e-02
-2.59921253e-01 1.13440633e+00 -7.99182281e-02 5.76124907e-01
-7.87760615e-01 -2.89800614e-01 6.77478731e-01 -1.84775442e-01
-3.68352026e-01 -5.70692778e-01 -5.96820235e-01 -9.01455343e-01
3.02732825e-01 -4.25347030e-01 6.64535835e-02 -8.25859845e-01
-7.38762617e-01 5.33132553e-01 2.71104574e-01 -9.79766548e-01
-5.00456214e-01 -4.90635693e-01 -8.59336555e-01 1.05984342e+00
-6.50978863e-01 -4.95901167e-01 -4.03741181e-01 1.05426550e+00
5.76679587e-01 5.38860381e-01 1.10326552e+00 -1.48966476e-01
-3.05279076e-01 1.87238112e-01 -2.09507719e-01 4.89841960e-02
4.78207558e-01 -7.39292264e-01 1.20492630e-01 6.37098730e-01
6.41452909e-01 1.00423872e+00 3.86982232e-01 -1.00270391e+00
-1.36530018e+00 -1.89573124e-01 7.09123492e-01 -6.12509847e-01
7.31934130e-01 -3.82398039e-01 -7.05774367e-01 4.60588247e-01
6.89626411e-02 -5.03979743e-01 8.87860000e-01 -4.06224549e-01
2.03789622e-01 6.34980261e-01 -5.88290155e-01 7.97913849e-01
1.25499272e+00 -7.81920493e-01 -1.14113915e+00 1.65038966e-02
-3.10620338e-01 -1.01987809e-01 -6.98571265e-01 2.15572730e-01
1.29781532e+00 -8.81606460e-01 8.37733984e-01 -7.31671929e-01
2.08463222e-01 2.51110971e-01 -3.05788685e-02 -1.22863066e+00
-2.41589621e-01 -4.29200172e-01 1.23740118e-02 7.03760386e-01
1.29480541e-01 -3.19820762e-01 3.84998947e-01 5.77641606e-01
-7.37620220e-02 -3.99280548e-01 -1.09466696e+00 -5.04788518e-01
-2.23080188e-01 -9.81548786e-01 3.38310301e-02 4.29720819e-01
1.02568257e+00 2.03827769e-03 2.32823253e-01 2.38461465e-01
-2.65262146e-02 5.48200235e-02 -6.24525081e-03 -9.21639979e-01
-1.24218185e-02 -1.15539217e+00 -3.97394180e-01 -5.38891733e-01
3.68420705e-02 -1.13909328e+00 1.56523913e-01 -1.50948882e+00
3.06097299e-01 1.97096944e-01 -3.16613019e-01 3.25016409e-01
-1.69315919e-01 2.59033114e-01 2.55637020e-01 4.81380463e-01
3.94863158e-01 2.86846310e-01 8.36292028e-01 9.56301689e-02
-9.85805809e-01 -7.84488581e-03 -1.66308627e-01 5.91910124e-01
8.77486944e-01 -3.24403554e-01 -3.40247810e-01 -1.07217096e-01
-1.80413499e-01 2.85004377e-01 5.12645364e-01 -9.04074252e-01
1.04686171e-01 7.90511258e-03 8.15492511e-01 -6.41901553e-01
5.21306932e-01 -7.54132152e-01 2.36980215e-01 7.00503707e-01
-2.69785166e-01 1.75730154e-01 3.51595253e-01 1.53049052e-01
2.20843896e-01 -7.13144839e-02 3.14169407e-01 -5.82836044e-04
-7.52524018e-01 -2.91464299e-01 -9.60213900e-01 -1.75522864e-01
1.10056853e+00 -7.08459675e-01 -8.75712186e-02 -1.81693226e-01
-1.00911057e+00 -3.79475802e-01 -1.04061343e-01 6.11602783e-01
7.93044686e-01 -1.40898681e+00 -3.81630421e-01 4.12707865e-01
-2.28027657e-01 -8.16264153e-01 2.78210878e-01 1.66616488e+00
-6.71316683e-02 6.66772842e-01 -6.99810803e-01 -8.06813359e-01
-1.30388701e+00 1.84396684e-01 4.30259332e-02 6.16195619e-01
-9.40787017e-01 1.16149259e+00 -6.67827055e-02 7.86396503e-01
3.46793562e-01 -6.11793041e-01 -4.54252481e-01 8.53583515e-01
9.70902383e-01 6.52147710e-01 -2.31003225e-01 -9.93312955e-01
-7.38784790e-01 6.62367582e-01 7.74666190e-01 -7.15489924e-01
1.21554732e+00 -9.28526744e-02 -1.92645654e-01 1.16875184e+00
9.80096459e-01 -6.80775195e-02 -1.37681580e+00 6.34987533e-01
2.22481579e-01 -1.85062826e-01 -2.11499587e-01 -9.21430767e-01
-8.35108876e-01 1.23413980e+00 9.46260810e-01 -1.95691139e-01
1.14330471e+00 1.94524646e-01 4.07368988e-01 -1.45185620e-01
4.64388728e-01 -1.11475694e+00 4.01950860e-03 -1.71881039e-02
1.14751291e+00 -4.47563082e-01 -3.43558967e-01 9.27730054e-02
-6.83581948e-01 1.48475838e+00 1.92849502e-01 -4.13103193e-01
5.73201656e-01 1.27312586e-01 -2.55619437e-01 -4.03963238e-01
-3.90310913e-01 -2.05591574e-01 7.21376300e-01 5.49500287e-01
6.93914056e-01 3.15530479e-01 -9.67608929e-01 1.16931820e+00
-3.80658001e-01 -1.65380940e-01 2.26103395e-01 9.32814181e-01
-4.90741938e-01 -3.14626396e-01 -7.32075214e-01 6.77927911e-01
-3.10788125e-01 -8.17695260e-02 -6.56865537e-01 1.01498258e+00
5.49661145e-02 1.02333450e+00 5.53389728e-01 -6.94497108e-01
2.87284732e-01 7.98366725e-01 1.08880782e+00 -4.67149645e-01
-8.75819623e-01 5.42202830e-01 -9.57772322e-03 -8.97881150e-01
-5.53884208e-01 -1.20669675e+00 -2.06916428e+00 2.83575654e-01
-1.00698106e-01 -1.73735067e-01 6.73334539e-01 1.09657168e+00
-1.47375464e-02 5.68712890e-01 -1.14081070e-01 -1.17715895e+00
-2.03421656e-02 -1.38243115e+00 -5.79255581e-01 1.50080666e-01
2.60488749e-01 -8.37729871e-01 -1.99291959e-01 3.45423937e-01] | [13.016499519348145, 3.429910659790039] |
be2d553d-434a-4fc9-b2d6-2043bb04903e | actgan-flexible-and-efficient-one-shot-face | 2003.13840 | null | https://arxiv.org/abs/2003.13840v1 | https://arxiv.org/pdf/2003.13840v1.pdf | ActGAN: Flexible and Efficient One-shot Face Reenactment | This paper introduces ActGAN - a novel end-to-end generative adversarial network (GAN) for one-shot face reenactment. Given two images, the goal is to transfer the facial expression of the source actor onto a target person in a photo-realistic fashion. While existing methods require target identity to be predefined, we address this problem by introducing a "many-to-many" approach, which allows arbitrary persons both for source and target without additional retraining. To this end, we employ the Feature Pyramid Network (FPN) as a core generator building block - the first application of FPN in face reenactment, producing finer results. We also introduce a solution to preserve a person's identity between synthesized and target person by adopting the state-of-the-art approach in deep face recognition domain. The architecture readily supports reenactment in different scenarios: "many-to-many", "one-to-one", "one-to-another" in terms of expression accuracy, identity preservation, and overall image quality. We demonstrate that ActGAN achieves competitive performance against recent works concerning visual quality. | ['Volodymyr Budzan', 'Marian Petruk', 'Ivan Kosarevych', 'Orest Kupyn', 'Mykola Maksymenko', 'Markian Kostiv'] | 2020-03-30 | null | null | null | null | ['face-reenactment'] | ['computer-vision'] | [ 4.04542834e-01 3.70422959e-01 3.47205251e-01 -5.87671280e-01
-6.04293644e-01 -5.39319158e-01 6.48486853e-01 -8.36204886e-01
-7.24471780e-03 7.69410193e-01 1.01282157e-01 2.96763718e-01
2.92250097e-01 -9.39604163e-01 -8.88616800e-01 -8.29263628e-01
2.69742697e-01 1.68829337e-01 -4.32572901e-01 -5.19764423e-01
-3.22329611e-01 7.35596657e-01 -1.35670984e+00 4.08619583e-01
4.28242654e-01 8.77877891e-01 -3.91520917e-01 6.08428299e-01
2.16233909e-01 7.78792202e-01 -7.48124540e-01 -1.23957586e+00
6.98565245e-01 -9.03846920e-01 -6.49776757e-01 2.93471307e-01
7.11323202e-01 -4.70075309e-01 -4.70322400e-01 1.22527659e+00
7.72334456e-01 -5.90122156e-02 4.95951802e-01 -1.65397382e+00
-1.04263699e+00 2.82036781e-01 -7.51103282e-01 -5.25535107e-01
4.71042335e-01 4.41879719e-01 4.27126616e-01 -8.09258759e-01
7.27011621e-01 1.58374953e+00 6.61011517e-01 1.11876345e+00
-1.30326545e+00 -1.06470466e+00 -1.21198267e-01 -1.62594438e-01
-1.40640879e+00 -8.51427376e-01 8.34847152e-01 -2.93868095e-01
2.27275819e-01 2.98678786e-01 5.65553606e-01 1.39449787e+00
7.43338764e-02 3.66842836e-01 1.23627484e+00 -3.69048953e-01
-7.25728050e-02 1.58893898e-01 -5.31179607e-01 7.10814297e-01
-1.34323314e-01 3.21682036e-01 -5.53278804e-01 -9.03577954e-02
9.64153171e-01 4.05198224e-02 -3.90256107e-01 -2.19681472e-01
-7.05470383e-01 6.78446531e-01 5.01532495e-01 1.48977786e-01
-4.79738533e-01 2.85355061e-01 2.41901264e-01 6.73544645e-01
4.04440314e-01 2.64466166e-01 1.10714801e-01 2.13481277e-01
-1.11497521e+00 3.24156672e-01 6.23756766e-01 9.32641447e-01
8.07513297e-01 4.48814869e-01 -4.46070582e-01 8.33190739e-01
9.87682194e-02 3.84897172e-01 3.04698199e-01 -1.00599134e+00
4.66548167e-02 3.70504707e-01 5.86357107e-03 -9.41173911e-01
3.56606990e-01 -5.00199795e-01 -1.11915040e+00 9.07524586e-01
5.95969940e-03 -3.30061704e-01 -1.07417202e+00 2.18107009e+00
3.52031887e-01 3.76684725e-01 3.48719805e-01 8.72298419e-01
9.14677501e-01 5.74828446e-01 -2.08152514e-02 -2.18974039e-01
1.34984338e+00 -1.06817400e+00 -6.48550332e-01 -5.68507984e-02
-6.35994002e-02 -7.61838794e-01 1.01616764e+00 1.67426571e-01
-1.33854628e+00 -6.78493083e-01 -9.69401836e-01 4.09138389e-02
-1.47491068e-01 1.04396284e-01 2.52755612e-01 1.01712537e+00
-1.42061412e+00 5.95032573e-01 -2.13398844e-01 -3.13282311e-01
6.92227900e-01 5.97202420e-01 -1.02232230e+00 -1.17113315e-01
-1.01411498e+00 6.49570346e-01 8.60751346e-02 9.24143717e-02
-1.26782250e+00 -8.22901309e-01 -7.40278900e-01 1.49468139e-01
9.71080512e-02 -1.10112154e+00 1.06743503e+00 -1.85538411e+00
-1.84343290e+00 1.32834208e+00 -1.48752749e-01 -3.87812585e-01
9.43411291e-01 4.87299599e-02 -5.71702898e-01 1.75467193e-01
1.71856452e-02 8.66696596e-01 1.30263174e+00 -1.63740337e+00
-2.29301304e-01 -4.53520983e-01 1.47845551e-01 1.31988928e-01
-1.07460864e-01 2.66242087e-01 -4.28415149e-01 -8.58737707e-01
-3.23315978e-01 -7.91187108e-01 1.59459069e-01 4.27312672e-01
-4.45800573e-01 1.40522838e-01 1.23681605e+00 -6.59532964e-01
5.34292459e-01 -2.30135322e+00 2.38670483e-02 3.90904918e-02
2.76033640e-01 5.20169318e-01 -4.35356587e-01 3.42017382e-01
-4.72710311e-01 -7.06615373e-02 -1.66790605e-01 -8.51265728e-01
-4.52942699e-02 9.40196589e-02 -2.30329752e-01 5.12390316e-01
3.14613104e-01 1.01711702e+00 -6.43916249e-01 -3.07903796e-01
-4.18792805e-03 8.56586039e-01 -5.97463131e-01 6.30755961e-01
1.21456683e-01 6.37440681e-01 6.80090860e-02 5.94737053e-01
1.04374373e+00 2.08131105e-01 -4.01880965e-02 -3.41061592e-01
2.70832300e-01 -5.65264106e-01 -1.00733876e+00 1.44497287e+00
-3.86021644e-01 4.79275614e-01 3.53687078e-01 -5.21679640e-01
1.17058742e+00 4.89282608e-01 2.48314917e-01 -4.26260173e-01
2.02796057e-01 3.27922106e-02 -9.36678872e-02 -2.12169424e-01
2.55508512e-01 -5.96251667e-01 1.27889469e-01 1.82455853e-01
1.73319578e-01 9.37252417e-02 -9.38480794e-02 2.69038286e-02
7.98688710e-01 1.34184748e-01 5.64996935e-02 -1.91319481e-01
5.79158425e-01 -5.42570651e-01 6.51496410e-01 4.26210999e-01
-9.53090712e-02 9.44396734e-01 6.30307317e-01 -3.89830798e-01
-1.16058493e+00 -1.22412086e+00 4.45957512e-01 8.91164601e-01
-4.13618125e-02 -2.09713820e-02 -1.18291426e+00 -6.16736174e-01
-1.95601791e-01 6.47227108e-01 -1.06151736e+00 -3.86100799e-01
-4.08922404e-01 -2.99118102e-01 1.06029892e+00 2.11237371e-01
1.08393455e+00 -1.14226615e+00 -3.98542106e-01 6.73961043e-02
-1.28927715e-02 -1.06760466e+00 -7.83283770e-01 -5.21285176e-01
-1.70201510e-01 -9.62782085e-01 -1.02348804e+00 -9.04643774e-01
9.45009530e-01 -7.48039261e-02 9.06356990e-01 7.35018402e-02
-1.76728949e-01 4.20975655e-01 -1.74264625e-01 -2.93454111e-01
-7.76045918e-01 -5.13130844e-01 1.28165275e-01 8.18682075e-01
-1.17005333e-01 -7.12680519e-01 -8.77326369e-01 2.46881172e-01
-9.38166857e-01 6.74165413e-02 4.46148783e-01 7.31408715e-01
2.96153724e-01 -2.65350267e-02 6.14473403e-01 -9.39561844e-01
6.81459963e-01 -3.42920586e-03 -2.06848830e-01 3.57006639e-01
-3.04979980e-01 -2.69944936e-01 9.55225527e-01 -6.15594566e-01
-1.17995739e+00 1.31508261e-01 -3.05004746e-01 -9.84392166e-01
-2.71350257e-02 -2.70123243e-01 -6.87334359e-01 -4.36754137e-01
5.59783041e-01 5.13680816e-01 2.95170248e-01 -1.92210555e-01
5.57229161e-01 4.84832466e-01 1.09530187e+00 -4.48557675e-01
1.12038660e+00 6.44878268e-01 2.52814963e-02 -5.73351800e-01
-3.61066520e-01 2.37609252e-01 -4.91750419e-01 -2.34137237e-01
8.01132500e-01 -1.01709080e+00 -9.70471382e-01 1.01885378e+00
-1.26730406e+00 -1.52993098e-01 -4.81205791e-01 -1.31150752e-01
-5.55471420e-01 8.03804472e-02 -3.96108747e-01 -5.64438164e-01
-7.80424833e-01 -1.07906413e+00 1.08910275e+00 5.12507141e-01
3.07934191e-02 -7.31000304e-01 -5.60365431e-02 1.85627326e-01
5.79443932e-01 8.95933032e-01 5.10519981e-01 -2.73493856e-01
-2.40595967e-01 -2.10130259e-01 -2.45725036e-01 6.34584367e-01
2.85818279e-01 8.13263729e-02 -1.03094828e+00 -7.88509130e-01
2.21915767e-01 -3.01592529e-01 5.84972084e-01 -1.26190349e-01
8.14346373e-01 -7.76770413e-01 -5.84008396e-02 8.72863472e-01
1.42415071e+00 1.71496257e-01 1.18448198e+00 -1.02734059e-01
6.80997431e-01 6.61276937e-01 8.30646455e-02 2.93601692e-01
1.88415989e-01 6.39799595e-01 5.10526597e-01 -6.40222311e-01
-5.76952577e-01 -5.25672555e-01 4.53456372e-01 1.12904636e-02
1.31311209e-03 -2.48722315e-01 -3.49948406e-01 4.90009934e-01
-1.45707047e+00 -1.20647776e+00 3.57372135e-01 1.96572065e+00
7.10392416e-01 -2.50888050e-01 1.12290598e-01 -9.89087820e-02
1.03784895e+00 1.96900681e-01 -4.33225274e-01 -4.29237664e-01
-2.79229075e-01 4.24993694e-01 1.19247682e-01 3.78107607e-01
-7.22755790e-01 1.12675428e+00 5.61777020e+00 7.66239405e-01
-1.48714006e+00 2.42914885e-01 9.26943123e-01 -1.33816004e-01
-6.23232983e-02 -3.32667008e-02 -5.96949458e-01 4.72968370e-01
4.89436299e-01 -3.01123768e-01 4.99350190e-01 7.45953023e-01
9.67285186e-02 3.72631311e-01 -1.11130452e+00 1.28303921e+00
4.41104323e-01 -1.30343425e+00 4.33188528e-01 -8.97422582e-02
8.47425461e-01 -7.42123067e-01 3.67524415e-01 1.54479563e-01
2.91144252e-01 -1.39044654e+00 8.00084949e-01 5.58859289e-01
1.43510222e+00 -1.01778531e+00 4.45996404e-01 -5.50822876e-02
-1.15626979e+00 1.23510078e-01 -3.22136551e-01 3.01466227e-01
1.66601822e-01 2.77304530e-01 -6.16622984e-01 6.51497245e-01
5.34435511e-01 2.84920543e-01 -4.02457178e-01 5.40246725e-01
-2.19551772e-01 1.23506606e-01 2.27959733e-02 4.34898138e-01
1.11446396e-01 -1.59770131e-01 5.03761947e-01 1.05717087e+00
5.18728793e-01 2.53495872e-01 -4.79246341e-02 1.04223573e+00
-6.37870193e-01 -1.07699342e-01 -7.65277743e-01 2.29385078e-01
3.01064789e-01 1.24855614e+00 -3.42255473e-01 -2.90882200e-01
5.72624020e-02 1.54717970e+00 1.67084590e-01 3.29143196e-01
-9.11085427e-01 -4.95461911e-01 7.98182011e-01 3.04489464e-01
2.11796388e-01 2.53806561e-01 2.00615510e-01 -8.63590121e-01
1.60005447e-02 -1.34300590e+00 -4.09322511e-03 -9.40640211e-01
-1.35020232e+00 1.08470452e+00 -2.52212256e-01 -9.56079245e-01
-2.18801171e-01 -3.15933615e-01 -9.44593966e-01 1.05713737e+00
-1.08571935e+00 -1.89122963e+00 -5.31288624e-01 1.11458862e+00
4.40231264e-01 -4.72838134e-01 7.88990557e-01 2.51327425e-01
-2.94302732e-01 1.05614126e+00 -2.32314095e-01 4.91329491e-01
6.80550277e-01 -7.13617742e-01 4.84145463e-01 1.04330015e+00
-6.51255921e-02 4.10324305e-01 7.13520706e-01 -4.11066443e-01
-1.24888563e+00 -1.40385628e+00 5.61494827e-01 -7.14508221e-02
9.62341763e-03 -5.10223806e-01 -6.68998897e-01 7.03617334e-01
5.66283643e-01 1.97822571e-01 4.71414000e-01 -4.91432309e-01
-5.11429131e-01 -3.63320082e-01 -1.71945095e+00 8.90819848e-01
1.07894039e+00 -6.32198930e-01 -1.41147301e-01 -1.84437297e-02
4.98886108e-01 -4.05223817e-01 -7.73242712e-01 4.36259598e-01
7.35959768e-01 -1.27770865e+00 9.26076472e-01 -5.83982110e-01
5.66467941e-01 -4.81952608e-01 -1.68308228e-01 -1.29032803e+00
-2.74461210e-01 -1.18810165e+00 2.92157590e-01 1.59651005e+00
4.17307997e-03 -7.11258531e-01 8.64472330e-01 4.54263300e-01
1.47837281e-01 -5.72827697e-01 -1.06808770e+00 -7.78195798e-01
1.37063891e-01 2.11172566e-01 1.04139328e+00 1.10324001e+00
-5.91963112e-01 2.95470208e-01 -1.07624865e+00 2.25248873e-01
6.55384481e-01 -2.85109375e-02 1.15493715e+00 -7.85006285e-01
-4.27289128e-01 -1.90880150e-01 -6.14590526e-01 -6.85855567e-01
3.78962278e-01 -7.91938841e-01 -1.99778616e-01 -1.13214314e+00
1.88633204e-01 3.36089246e-02 4.33132537e-02 6.63843155e-01
1.13771316e-02 7.73540497e-01 5.62891245e-01 2.88970638e-02
-4.97444496e-02 7.09416211e-01 1.43136406e+00 -1.41138837e-01
1.23742156e-01 -8.59600976e-02 -9.17518258e-01 3.89258534e-01
7.48960853e-01 -4.40178543e-01 -5.41409612e-01 -2.78433710e-01
-3.54253054e-01 1.53152272e-01 7.46316195e-01 -1.04416358e+00
1.80104762e-01 -9.50258505e-03 6.29405737e-01 1.42815188e-02
7.34389365e-01 -8.32378626e-01 8.09311092e-01 4.76792157e-01
-1.31071448e-01 2.29362622e-01 1.68635771e-01 2.51720011e-01
-2.30821356e-01 -1.62108559e-02 1.43581772e+00 -1.98340490e-01
-4.64271337e-01 5.19211888e-01 1.97770759e-01 -1.48949608e-01
1.37476301e+00 -3.62505436e-01 -3.93629432e-01 -7.57751882e-01
-7.42743611e-01 -2.12785736e-01 7.14600623e-01 5.05884588e-01
7.86284626e-01 -1.62429750e+00 -1.25517166e+00 6.07540309e-01
-1.30044863e-01 -3.41230929e-01 2.75199026e-01 3.54880601e-01
-5.72068036e-01 -3.27633470e-01 -7.38894224e-01 -2.80919313e-01
-1.59866631e+00 4.95417744e-01 8.13033760e-01 -1.17228553e-01
-4.73514229e-01 1.03510284e+00 6.21314049e-01 -2.84907103e-01
-3.40157859e-02 5.89728236e-01 3.25916320e-01 -2.42331371e-01
5.00737071e-01 1.18740849e-01 -2.14379564e-01 -1.05229104e+00
-1.30392089e-01 6.27783000e-01 -1.30362630e-01 -1.99063271e-01
1.16369128e+00 -6.35643080e-02 -2.67451644e-01 -1.41269699e-01
1.34675395e+00 3.99368480e-02 -1.41856086e+00 -6.08873256e-02
-7.46679068e-01 -8.00711989e-01 -3.55684698e-01 -8.34929883e-01
-1.51841223e+00 7.14344740e-01 8.93289626e-01 -1.79262489e-01
1.40956020e+00 -3.17045420e-01 7.97670960e-01 -1.53140560e-01
3.45482618e-01 -5.79056919e-01 4.23412710e-01 7.94137195e-02
1.28878200e+00 -9.93146598e-01 -3.06129605e-01 -3.45592469e-01
-7.71457970e-01 8.86991441e-01 6.65409207e-01 -2.08022818e-01
4.33945328e-01 1.86613977e-01 2.46409148e-01 -1.20959379e-01
-4.87643421e-01 1.12349607e-01 2.38069907e-01 8.51174474e-01
9.87989604e-02 -5.55332601e-02 3.36246043e-01 3.18608940e-01
-4.69501913e-01 1.22934788e-01 3.75391006e-01 6.36800408e-01
8.91304687e-02 -1.29392922e+00 -4.59127545e-01 6.05392531e-02
-5.76471746e-01 -1.35219854e-03 -5.61253309e-01 8.34307969e-01
3.39306325e-01 9.29992616e-01 -4.22804579e-02 -3.78842443e-01
4.63977873e-01 -8.69241059e-02 6.61012471e-01 -3.94144475e-01
-9.45967674e-01 -1.43612474e-01 -1.70735851e-01 -3.62410665e-01
-3.59300166e-01 -2.24832460e-01 -8.30779910e-01 -8.51806402e-01
8.14733431e-02 -1.52630746e-01 3.56347948e-01 5.59782505e-01
5.31008959e-01 3.83310318e-01 1.00675869e+00 -7.45311916e-01
-3.63839537e-01 -7.14828074e-01 -6.57986283e-01 7.56711662e-01
3.38802576e-01 -2.74548441e-01 -9.43269581e-02 2.89781958e-01] | [12.705150604248047, -0.08667301386594772] |
090d43cd-e346-4a22-afa5-7bc4b095f1c1 | benchmark-data-to-study-the-influence-of-pre | 2306.12150 | null | https://arxiv.org/abs/2306.12150v1 | https://arxiv.org/pdf/2306.12150v1.pdf | Benchmark data to study the influence of pre-training on explanation performance in MR image classification | Convolutional Neural Networks (CNNs) are frequently and successfully used in medical prediction tasks. They are often used in combination with transfer learning, leading to improved performance when training data for the task are scarce. The resulting models are highly complex and typically do not provide any insight into their predictive mechanisms, motivating the field of 'explainable' artificial intelligence (XAI). However, previous studies have rarely quantitatively evaluated the 'explanation performance' of XAI methods against ground-truth data, and transfer learning and its influence on objective measures of explanation performance has not been investigated. Here, we propose a benchmark dataset that allows for quantifying explanation performance in a realistic magnetic resonance imaging (MRI) classification task. We employ this benchmark to understand the influence of transfer learning on the quality of explanations. Experimental results show that popular XAI methods applied to the same underlying model differ vastly in performance, even when considering only correctly classified examples. We further observe that explanation performance strongly depends on the task used for pre-training and the number of CNN layers pre-trained. These results hold after correcting for a substantial correlation between explanation and classification performance. | ['Stefan Haufe', 'Kerstin Ritter', 'Fabian Eitel', 'Céline Budding', 'Benedict Clark', 'Rick Wilming', 'Marta Oliveira'] | 2023-06-21 | null | null | null | null | ['explainable-artificial-intelligence', 'transfer-learning'] | ['computer-vision', 'miscellaneous'] | [ 5.45302689e-01 5.23820281e-01 -2.91382402e-01 -6.05880678e-01
-4.58571374e-01 -1.00540057e-01 7.92247593e-01 5.10865748e-01
-3.86645496e-01 8.19404244e-01 3.06580186e-01 -3.48255932e-01
-5.32294631e-01 -5.60114324e-01 -8.87402713e-01 -6.86632931e-01
-1.69706643e-01 6.19972050e-01 -3.88339683e-02 -1.71279162e-01
1.26685873e-01 1.95086583e-01 -1.49111104e+00 5.99888206e-01
1.02135241e+00 8.99231791e-01 -5.49470559e-02 4.98941004e-01
3.42753809e-03 7.76512742e-01 -6.20462894e-01 -4.47801769e-01
-1.51203290e-01 -6.40439332e-01 -1.07597721e+00 -4.85345758e-02
1.72827482e-01 1.27195880e-01 -1.00325055e-01 8.04170668e-01
2.06417114e-01 -2.00089514e-01 9.21991348e-01 -1.25147331e+00
-6.78718925e-01 8.26534867e-01 8.75596777e-02 3.69856358e-01
1.27262026e-01 1.26290470e-01 1.09172130e+00 -5.82573354e-01
6.94916129e-01 8.07920992e-01 9.46910560e-01 6.77116871e-01
-1.34331954e+00 -5.30538857e-01 -3.37824404e-01 5.81236422e-01
-8.76073778e-01 -1.02949180e-01 5.33085167e-01 -3.78998041e-01
7.01180518e-01 3.13907832e-01 4.35398698e-01 1.33128905e+00
4.47200954e-01 6.51011825e-01 1.34346128e+00 -4.96035874e-01
2.86994874e-01 3.17044526e-01 5.32862544e-01 5.23994625e-01
5.16722620e-01 3.31772178e-01 -5.62702775e-01 1.06256701e-01
5.64316630e-01 -4.92333174e-02 -5.82282305e-01 -3.75317544e-01
-1.36032295e+00 1.11741698e+00 9.84666586e-01 7.34463096e-01
-5.34185588e-01 1.60306677e-01 5.50209165e-01 4.99221802e-01
4.72785741e-01 1.02556181e+00 -8.50693285e-01 -3.22201550e-02
-7.23124564e-01 4.15119499e-01 5.12241066e-01 3.37197423e-01
7.20969975e-01 -6.38516396e-02 -2.37995386e-01 5.92508256e-01
-2.03952417e-01 -1.64877120e-02 6.74977779e-01 -6.15356863e-01
2.22120658e-01 8.50851536e-01 -1.75828055e-01 -7.93286979e-01
-7.42941797e-01 -9.10723388e-01 -1.08081627e+00 3.39482069e-01
7.12549686e-01 9.90273654e-02 -7.46841550e-01 1.65492916e+00
-6.59880042e-02 1.76056549e-02 1.84416175e-01 1.09735489e+00
9.16065574e-01 1.00602277e-01 1.61272645e-01 5.96672818e-02
1.28080261e+00 -9.19900656e-01 -5.30371785e-01 -2.68536031e-01
8.13340068e-01 -5.24793386e-01 1.21181464e+00 5.69058895e-01
-9.57491994e-01 -5.55862725e-01 -1.01879263e+00 1.38147205e-01
-5.51173031e-01 1.91510245e-01 9.42330122e-01 4.41534787e-01
-8.04756045e-01 1.19658852e+00 -6.28553152e-01 -1.89314932e-01
6.12205327e-01 6.20793581e-01 -5.54811239e-01 -5.92681132e-02
-1.06816328e+00 1.23458481e+00 4.13663149e-01 -7.78979016e-03
-5.19990325e-01 -9.97407973e-01 -5.79835057e-01 4.54166055e-01
1.04495153e-01 -8.08891594e-01 1.14154625e+00 -1.41736674e+00
-1.04152310e+00 7.30606377e-01 3.05078030e-01 -8.18645120e-01
6.19931340e-01 -1.47929072e-01 -2.11742371e-01 -1.82466358e-01
1.36237815e-01 7.84533203e-01 5.74658513e-01 -1.30778491e+00
-7.52668381e-02 -3.69737595e-01 1.79118186e-01 -3.20000172e-01
-8.84954259e-02 -4.18509752e-01 1.26034915e-01 -5.79587936e-01
1.03003256e-01 -1.09359527e+00 -2.19566479e-01 -2.91935951e-01
-5.71761370e-01 -1.62166938e-01 3.24608862e-01 -3.47963572e-01
8.13213050e-01 -2.01022840e+00 1.02778099e-01 1.70451641e-01
3.10877264e-01 3.73471498e-01 -8.69165808e-02 1.38803497e-01
-6.32523417e-01 2.30690360e-01 -4.27086622e-01 -2.96128482e-01
-6.89810561e-03 4.20891732e-01 -5.45842834e-02 2.76750416e-01
4.30376649e-01 1.06174862e+00 -9.26087618e-01 -2.69721448e-01
2.93944627e-01 7.15734959e-01 -7.12340653e-01 2.25880444e-01
-2.89562732e-01 1.05245435e+00 -3.23778391e-01 1.53486848e-01
1.66417077e-01 -7.11122453e-01 5.29277362e-02 -1.74688414e-01
2.34427229e-01 3.98897707e-01 -5.16523123e-01 1.54701650e+00
-6.75762296e-01 9.38060939e-01 -7.40575492e-01 -1.24424243e+00
7.02881217e-01 4.35404718e-01 4.41777885e-01 -7.90230393e-01
3.95984888e-01 4.49287742e-01 6.55534327e-01 -3.50319654e-01
1.52702570e-01 -3.48882109e-01 2.93067932e-01 3.49887252e-01
1.13434158e-01 1.62570015e-01 -1.07086569e-01 -2.00017914e-01
1.20227861e+00 -1.85577512e-01 4.29683447e-01 -4.90759969e-01
4.37679112e-01 3.12027305e-01 3.13556314e-01 8.45291197e-01
1.03032872e-01 9.78669047e-01 6.72274411e-01 -1.04470265e+00
-1.04902434e+00 -6.14890516e-01 -4.94259983e-01 6.71415627e-01
8.29230249e-03 -1.15829922e-01 -8.42625022e-01 -7.32465088e-01
-1.86537892e-01 8.45235050e-01 -1.35950053e+00 -4.86035317e-01
-3.16880614e-01 -8.63534153e-01 2.23333463e-01 5.91328502e-01
3.97076756e-01 -1.51428246e+00 -9.59263802e-01 1.29269317e-01
-1.53164923e-01 -9.63308632e-01 2.19827235e-01 4.68904018e-01
-1.17181540e+00 -1.30479503e+00 -6.02740407e-01 -1.27326548e-01
6.47439003e-01 1.57776129e-04 1.64711142e+00 9.16580379e-01
-2.41446227e-01 1.10414237e-01 -4.83621478e-01 -6.03757381e-01
-6.41811252e-01 5.33103228e-01 -1.52783930e-01 -1.66084126e-01
1.84947386e-01 -5.00545681e-01 -8.13904464e-01 2.72266209e-01
-1.00881171e+00 2.88735598e-01 9.25637305e-01 1.13807476e+00
2.46815503e-01 -3.36852551e-01 6.25057638e-01 -1.27311742e+00
5.09370387e-01 -4.77335423e-01 1.80919282e-02 9.55317169e-02
-1.06368136e+00 5.47075450e-01 5.36640882e-01 -3.73397589e-01
-7.33699560e-01 -4.02554274e-01 -1.36533216e-01 -2.02425376e-01
-3.79943103e-01 7.08868504e-01 2.19942585e-01 -7.07978979e-02
9.62632418e-01 -1.49729684e-01 5.68479858e-02 -3.73222679e-01
4.64669615e-02 1.14909910e-01 5.90021074e-01 -2.33107775e-01
4.45984066e-01 2.79493570e-01 1.17217124e-01 -4.87852514e-01
-1.20132172e+00 -1.18155554e-01 -5.84703028e-01 -1.17704660e-01
1.03712654e+00 -3.16681176e-01 -6.51267111e-01 -2.18065873e-01
-1.30516958e+00 -4.55371886e-01 -2.95015574e-01 7.40971029e-01
-6.49696827e-01 -8.23258236e-02 -2.33101547e-01 -2.45172247e-01
-2.75162548e-01 -1.41491699e+00 9.06491220e-01 -1.86560641e-03
-6.31447732e-01 -1.18609488e+00 8.55321437e-02 4.05198008e-01
7.61626661e-01 4.31293994e-01 1.41455793e+00 -9.58725631e-01
-5.41818976e-01 -1.81031033e-01 -4.73394513e-01 1.93485469e-01
-1.13482907e-01 -3.75215650e-01 -1.06319594e+00 9.70989391e-02
-2.85252929e-01 -1.90323517e-01 9.39063132e-01 5.33342481e-01
1.51270223e+00 -2.15814248e-01 -2.12831378e-01 2.62770563e-01
1.30193031e+00 -3.45137119e-01 8.60054374e-01 7.38825440e-01
4.88699377e-01 7.53204763e-01 3.20553929e-01 1.06027164e-01
-8.39778334e-02 8.33476663e-01 7.78819203e-01 -3.42019111e-01
-1.08965680e-01 2.12077752e-01 -1.66176200e-01 5.00612736e-01
-5.23293495e-01 -2.61235684e-02 -1.13123453e+00 4.40037221e-01
-1.84414136e+00 -7.75186121e-01 -5.63139975e-01 2.11053467e+00
5.58330595e-01 2.60741174e-01 -1.04748614e-01 5.53386092e-01
2.46191442e-01 -3.45806271e-01 -3.05100977e-01 -4.56213146e-01
-1.28912814e-02 3.27854127e-01 3.25150400e-01 3.77583057e-01
-7.45571554e-01 3.98098022e-01 6.08384848e+00 4.11773592e-01
-1.25386560e+00 1.70816779e-01 9.31771040e-01 3.16762537e-01
-4.98048425e-01 -2.77476698e-01 -5.06822765e-02 2.67158121e-01
1.10160959e+00 1.10223755e-01 1.49877131e-01 9.06098545e-01
5.37060238e-02 1.88857451e-01 -1.48841989e+00 8.17134023e-01
-8.07625726e-02 -1.54991031e+00 -9.01684631e-03 9.84911621e-02
6.76519334e-01 -4.97254655e-02 2.01628417e-01 4.10487086e-01
-2.89651304e-01 -1.44026411e+00 5.55997789e-01 4.35038537e-01
3.16460997e-01 -5.69219828e-01 1.28731763e+00 1.65763691e-01
-4.54121351e-01 -4.31748778e-02 -3.87641490e-01 -6.68388531e-02
-2.72249967e-01 6.21644795e-01 -1.14393318e+00 4.09328163e-01
7.80330002e-01 3.92866492e-01 -7.75806725e-01 9.08858597e-01
-3.89461040e-01 8.58205497e-01 1.28440276e-01 9.77587607e-03
4.47015464e-01 3.89460959e-02 2.55843312e-01 1.16028023e+00
1.68707311e-01 -4.21757177e-02 -4.73800302e-01 1.13183522e+00
3.41833010e-02 1.32480159e-01 -5.46983898e-01 1.72019556e-01
-2.47537434e-01 1.05417967e+00 -9.53221619e-01 -3.59732032e-01
-2.55342722e-01 8.65002930e-01 3.51783007e-01 1.14274509e-02
-1.05816197e+00 3.40247959e-01 5.96208274e-01 3.68698597e-01
6.78464249e-02 1.89234614e-01 -7.51708448e-01 -9.61406946e-01
-1.70470148e-01 -8.84103715e-01 1.87434509e-01 -9.40976799e-01
-1.27226770e+00 1.04258645e+00 7.50858784e-02 -9.72795188e-01
-3.54504138e-01 -8.49188447e-01 -7.66939402e-01 5.66468418e-01
-1.62359047e+00 -9.72015560e-01 -7.01257288e-01 3.88392955e-01
4.71965015e-01 -6.30409345e-02 1.00218582e+00 2.47529551e-01
-2.51413375e-01 4.15280610e-01 -6.80448636e-02 -1.24728523e-01
3.81209105e-01 -1.53150022e+00 1.21592447e-01 2.47154608e-01
2.93736726e-01 5.79696715e-01 1.23043096e+00 -2.13225380e-01
-8.04918826e-01 -9.06527221e-01 9.95732546e-01 -6.31170630e-01
3.07894140e-01 2.29306389e-02 -1.44148588e+00 5.23897290e-01
3.02926928e-01 1.41131103e-01 7.88453639e-01 6.07280850e-01
-2.84777761e-01 1.71326444e-01 -8.91878128e-01 3.72518301e-01
8.97056401e-01 -2.59252906e-01 -6.00686967e-01 3.91714931e-01
5.03245652e-01 -1.88305497e-01 -9.05295730e-01 6.20833278e-01
5.67078829e-01 -1.44490492e+00 1.05205071e+00 -8.43830407e-01
1.20924616e+00 2.26091042e-01 1.96951255e-01 -1.42787969e+00
-2.31142133e-01 1.64056227e-01 3.20486099e-01 7.49063253e-01
8.10102940e-01 -5.56362212e-01 9.63407278e-01 7.17310727e-01
-2.16292217e-01 -1.12454128e+00 -8.78135860e-01 -6.92720830e-01
1.90663457e-01 -5.40619075e-01 7.67920971e-01 1.15234613e+00
-1.22484550e-01 4.15771902e-01 -2.03506261e-01 1.59112692e-01
3.68132949e-01 9.51384455e-02 7.13040769e-01 -1.50348365e+00
-3.50002736e-01 -6.02840066e-01 -6.18742049e-01 -1.03868574e-01
2.87670821e-01 -1.07803059e+00 -8.69360566e-02 -1.30817223e+00
3.15596938e-01 -5.07652164e-01 -3.73601317e-01 5.41557014e-01
-3.80525172e-01 3.35284084e-01 1.03686370e-01 4.36972082e-01
-4.59330738e-01 4.95349735e-01 1.03175271e+00 -6.71708509e-02
1.12546710e-02 -5.98102361e-02 -5.53459346e-01 7.52540529e-01
9.58670437e-01 -6.88893855e-01 -3.64731312e-01 -5.35970271e-01
2.51094490e-01 1.31857337e-03 7.58285642e-01 -1.17739952e+00
-2.84882873e-01 2.27977112e-01 4.41964358e-01 -7.92568773e-02
-1.72942430e-02 -8.54681015e-01 3.02572459e-01 6.87478304e-01
-7.07793117e-01 2.64967620e-01 1.73871130e-01 4.30247575e-01
-3.14998418e-01 -4.10405517e-01 6.69476271e-01 -4.46186252e-02
-3.75775307e-01 2.55831927e-01 -2.40442112e-01 -2.54798651e-01
5.40149033e-01 -3.23259652e-01 -1.91077348e-02 -3.56649548e-01
-9.28015530e-01 -3.35315287e-01 3.09293300e-01 4.03925270e-01
4.06556100e-01 -1.19985962e+00 -6.89759374e-01 -9.42952652e-03
3.51594180e-01 -2.13790417e-01 1.65790260e-01 9.74353135e-01
-5.06097078e-01 6.20521367e-01 -5.39400518e-01 -7.91753292e-01
-1.08293438e+00 5.36917150e-01 4.48983252e-01 -5.61615705e-01
-7.40206540e-01 4.73729104e-01 3.98467004e-01 -5.61282218e-01
-2.26080790e-02 -5.56439340e-01 -2.33803004e-01 -3.81479055e-01
3.63718569e-01 1.80328533e-01 2.21660346e-01 -2.40320116e-01
-2.45831072e-01 1.28977105e-01 5.43631017e-02 2.41342634e-01
1.71606648e+00 3.39302242e-01 1.35796130e-01 4.48866099e-01
1.16865635e+00 -6.44553304e-01 -7.70791173e-01 -8.53643045e-02
2.83262581e-01 -2.96499461e-01 -8.39055479e-02 -9.22947049e-01
-1.20757329e+00 1.23446524e+00 6.86951578e-01 2.66681582e-01
9.02547121e-01 9.19863209e-02 2.66280472e-01 1.45159557e-01
1.73068419e-01 -5.16935647e-01 3.65237534e-01 1.08237632e-01
1.23474658e+00 -1.58501530e+00 -6.09172583e-02 -4.78595823e-01
-6.02236450e-01 1.20693994e+00 3.76156956e-01 -1.88411370e-01
4.81853217e-01 -2.18664438e-01 2.55551767e-02 -6.46824002e-01
-5.44344246e-01 -2.03454290e-02 5.60699165e-01 4.64482039e-01
7.51087487e-01 1.06675394e-01 -2.29930505e-01 8.39595914e-01
-5.17551601e-01 5.67322522e-02 4.42992389e-01 2.65588492e-01
-2.48402134e-02 -1.02120078e+00 -2.69256294e-01 5.90057194e-01
-6.27944291e-01 -9.05064642e-02 -4.32637721e-01 1.19425416e+00
2.74814367e-01 5.21562874e-01 -4.18939069e-02 -1.81714728e-01
1.82642341e-01 1.23090766e-01 4.35498476e-01 -5.17071426e-01
-9.00798976e-01 -6.71923995e-01 -4.42611687e-02 -6.08757436e-01
-4.98409718e-01 -5.79619646e-01 -1.21646392e+00 -3.86297673e-01
-4.59395826e-01 3.19311261e-01 7.47063279e-01 1.38875854e+00
1.85541064e-01 9.65271533e-01 1.19354412e-01 -8.02202761e-01
-3.18198889e-01 -1.03452599e+00 -2.29194522e-01 9.47769344e-01
2.85644859e-01 -9.38602388e-01 -3.14351588e-01 -6.53354526e-02] | [8.771265029907227, 5.678853511810303] |
66fd7349-8039-4603-b3dc-b2ef1b32c4db | detail-preserved-point-cloud-completion-via | 2007.02374 | null | https://arxiv.org/abs/2007.02374v1 | https://arxiv.org/pdf/2007.02374v1.pdf | Detail Preserved Point Cloud Completion via Separated Feature Aggregation | Point cloud shape completion is a challenging problem in 3D vision and robotics. Existing learning-based frameworks leverage encoder-decoder architectures to recover the complete shape from a highly encoded global feature vector. Though the global feature can approximately represent the overall shape of 3D objects, it would lead to the loss of shape details during the completion process. In this work, instead of using a global feature to recover the whole complete surface, we explore the functionality of multi-level features and aggregate different features to represent the known part and the missing part separately. We propose two different feature aggregation strategies, named global \& local feature aggregation(GLFA) and residual feature aggregation(RFA), to express the two kinds of features and reconstruct coordinates from their combination. In addition, we also design a refinement component to prevent the generated point cloud from non-uniform distribution and outliers. Extensive experiments have been conducted on the ShapeNet dataset. Qualitative and quantitative evaluations demonstrate that our proposed network outperforms current state-of-the art methods especially on detail preservation. | ['Chunxia Xiao', 'Wenxiao Zhang', 'Qingan Yan'] | 2020-07-05 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/5105_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123700511.pdf | eccv-2020-8 | ['point-cloud-completion'] | ['computer-vision'] | [-6.16160855e-02 3.04774530e-02 3.12850177e-01 -4.27971572e-01
-6.87768042e-01 -4.62762356e-01 6.26471400e-01 1.52775943e-01
3.90665755e-02 3.26453239e-01 1.62185043e-01 2.97706634e-01
-4.49893251e-02 -7.36340046e-01 -8.69660795e-01 -5.66949964e-01
8.99252072e-02 5.27561307e-01 2.84030080e-01 8.28663558e-02
3.72398496e-01 1.06033909e+00 -1.63518190e+00 1.50546834e-01
8.04266989e-01 1.28366041e+00 3.61081570e-01 8.28424543e-02
-3.77351373e-01 2.26244316e-01 -3.78690988e-01 -1.63332134e-01
7.10130334e-01 2.99233407e-01 -2.86026120e-01 4.21658516e-01
6.69510961e-01 -5.05278945e-01 -2.61993229e-01 1.19389904e+00
5.34354568e-01 -1.54633150e-01 6.37888789e-01 -1.11649907e+00
-6.79700196e-01 -4.67184708e-02 -7.53659785e-01 -6.17754042e-01
4.15398568e-01 1.59201875e-01 7.35673845e-01 -1.40090597e+00
8.03018510e-01 1.34217656e+00 8.32207322e-01 3.09921145e-01
-9.80601192e-01 -5.81276357e-01 2.41446123e-01 -5.96506037e-02
-1.54876792e+00 -4.81560081e-01 1.25218248e+00 -5.23016512e-01
8.65569472e-01 9.40751806e-02 6.87799633e-01 4.03895110e-01
1.95863843e-01 7.67299891e-01 7.22909153e-01 -1.35098398e-01
1.68891922e-01 -2.66137272e-01 -1.42779857e-01 8.95633519e-01
3.86468560e-01 7.69369304e-02 -4.66887593e-01 -2.51045108e-01
1.17892528e+00 4.16197717e-01 -2.89271593e-01 -1.00034523e+00
-1.25096512e+00 4.79069382e-01 7.07336485e-01 -5.77756986e-02
-6.67411923e-01 1.33543730e-01 4.39353427e-03 2.19303295e-01
4.95837569e-01 1.50198922e-01 -6.08599305e-01 1.89620867e-01
-7.03386605e-01 4.99725550e-01 6.86485171e-01 1.46264803e+00
1.37700033e+00 3.84767354e-02 -2.05061715e-02 7.68136144e-01
4.77332562e-01 5.54717660e-01 -2.80049425e-02 -1.01287770e+00
5.44302583e-01 1.15723979e+00 2.33018965e-01 -1.16000879e+00
-3.60656708e-01 -4.04521585e-01 -8.52337003e-01 6.28552735e-01
1.66080862e-01 1.14285015e-01 -1.14946878e+00 1.44940341e+00
5.35893738e-01 3.86656076e-01 -2.21192464e-01 9.63660121e-01
1.12617564e+00 4.25681591e-01 -6.11803353e-01 1.57840431e-01
9.26726103e-01 -7.65732586e-01 -4.64565843e-01 -1.06954142e-01
1.20151155e-01 -8.04766297e-01 6.41714096e-01 2.04281017e-01
-1.20807230e+00 -5.93413889e-01 -1.10041296e+00 -4.42632675e-01
-1.82168275e-01 2.71890819e-01 5.64704537e-01 1.92823391e-02
-9.67074752e-01 6.24101877e-01 -9.05779362e-01 6.98073651e-04
7.06246138e-01 4.12340999e-01 -6.46755993e-01 -3.61829758e-01
-2.98136413e-01 6.55992270e-01 -1.55014426e-01 2.01104730e-01
-7.32944906e-01 -9.50514495e-01 -9.22819614e-01 3.15231085e-02
1.47589937e-01 -9.67534125e-01 1.00682676e+00 -3.67538005e-01
-1.22008848e+00 6.59291387e-01 -3.77988309e-01 3.27170007e-02
3.73523474e-01 -3.31214100e-01 3.49112332e-01 -4.40844856e-02
1.36648297e-01 7.25498497e-01 9.44629312e-01 -1.72730350e+00
-5.39499998e-01 -7.18980491e-01 -1.25596374e-01 3.16824794e-01
2.47991443e-01 -4.73822713e-01 -5.72968066e-01 -4.97425914e-01
8.94205749e-01 -6.72968030e-01 -2.29115441e-01 7.45233715e-01
-2.85855293e-01 -1.81621686e-01 1.02211535e+00 -5.73302567e-01
4.60947782e-01 -2.15036464e+00 1.31358877e-01 2.49580860e-01
3.52942109e-01 3.65451090e-02 -3.33977282e-01 2.41033688e-01
1.98731631e-01 -1.77987471e-01 -3.29507440e-01 -7.40553498e-01
-5.54628596e-02 3.32886815e-01 -2.84256995e-01 5.88697314e-01
4.94426996e-01 9.71694887e-01 -8.25284660e-01 -9.32617933e-02
4.16685611e-01 8.20011437e-01 -7.70494521e-01 5.30087948e-02
-2.35405400e-01 4.52781141e-01 -6.95887089e-01 1.13982546e+00
1.27389371e+00 7.79615715e-03 -4.29717869e-01 -3.97111773e-01
-1.78310618e-01 1.26216024e-01 -1.45099127e+00 2.22710872e+00
-1.99932128e-01 1.22140504e-01 3.66460562e-01 -6.19795501e-01
1.37342334e+00 6.23981878e-02 8.29923332e-01 -3.21707040e-01
7.42132366e-02 2.95628130e-01 -2.75717348e-01 -2.02752545e-01
4.66842592e-01 1.05510101e-01 2.71313608e-01 9.74778533e-02
-4.25362363e-02 -4.61499751e-01 -6.00700796e-01 -2.37461358e-01
1.15690076e+00 5.15542746e-01 1.14296421e-01 -4.78748558e-03
5.54011345e-01 -9.07278657e-02 8.06600332e-01 3.36317331e-01
-9.49441567e-02 1.13411605e+00 2.15078056e-01 -5.74015617e-01
-1.11207998e+00 -1.07221496e+00 -3.62521410e-02 3.10180783e-01
4.56024110e-01 -3.26721072e-01 -3.15734357e-01 -6.39921665e-01
4.18100089e-01 3.87678564e-01 -2.58630216e-01 -2.66157109e-02
-6.06747031e-01 -9.15187150e-02 7.37049058e-02 5.16902268e-01
4.24610496e-01 -8.82956088e-01 -6.29312754e-01 1.30806431e-01
5.63186929e-02 -9.84512866e-01 -5.08472741e-01 2.54769158e-02
-1.02238095e+00 -1.11329091e+00 -5.14543355e-01 -7.95557141e-01
9.93970454e-01 6.47587359e-01 7.93202817e-01 1.61008090e-01
-2.16710582e-01 3.08223456e-01 -3.37120891e-01 -3.80246878e-01
5.60236797e-02 -2.35602438e-01 -6.36923909e-02 1.97614636e-02
2.91970789e-01 -9.20576155e-01 -5.85575461e-01 5.81828840e-02
-6.46080434e-01 6.37255460e-02 7.27870107e-01 6.75156415e-01
1.10527933e+00 -1.74323544e-01 2.53136814e-01 -4.71602410e-01
2.53051966e-01 -2.66740859e-01 -5.26604414e-01 4.22342308e-03
-2.84455359e-01 1.48148835e-01 3.92478317e-01 -2.62202293e-01
-8.15183938e-01 6.29237056e-01 -1.50915980e-01 -1.18768239e+00
-1.42349258e-01 3.33243132e-01 -4.44311231e-01 -2.80720443e-01
2.46646687e-01 3.54637355e-01 2.03317732e-01 -9.60214257e-01
3.30564439e-01 4.02942747e-01 4.47251290e-01 -4.93130833e-01
1.06275916e+00 7.04702735e-01 2.31671616e-01 -8.30136836e-01
-5.78874707e-01 -5.86583257e-01 -8.88375103e-01 -1.88178625e-02
5.22296906e-01 -1.07570672e+00 -6.22021914e-01 5.58619022e-01
-1.56295574e+00 2.50471830e-01 -5.22056699e-01 1.74967676e-01
-7.65360534e-01 3.72209549e-01 -9.06947404e-02 -6.20635986e-01
-4.21132714e-01 -1.19515657e+00 1.53090465e+00 1.30079566e-02
3.17117423e-01 -5.24822235e-01 2.10779980e-02 -1.09134391e-01
2.59760618e-01 6.29850090e-01 7.16907680e-01 -2.68773198e-01
-1.03977656e+00 -3.65073383e-01 -3.57965648e-01 1.91286400e-01
4.32573318e-01 -5.25657013e-02 -9.08826530e-01 -3.82364035e-01
1.84538573e-01 -5.35371192e-02 8.09434354e-01 3.14173639e-01
9.99218106e-01 -2.21459523e-01 -3.14135969e-01 9.37031865e-01
1.53648448e+00 -1.00948647e-01 5.49564958e-01 2.69746874e-02
1.04330885e+00 4.70784724e-01 5.27693033e-01 6.44556165e-01
5.77772677e-01 5.45824528e-01 9.49637115e-01 2.45985404e-01
-4.75066513e-01 -5.42248726e-01 1.75126657e-01 8.26475322e-01
-1.91468582e-01 2.90180743e-01 -8.27749133e-01 5.03649533e-01
-1.81139493e+00 -6.60175502e-01 -1.78809628e-01 2.15571594e+00
4.95176762e-01 -8.11648592e-02 -3.66051376e-01 4.73219231e-02
6.67498052e-01 1.07681565e-01 -6.74320042e-01 -5.99249415e-02
-1.59787029e-01 1.12451598e-01 4.08242911e-01 5.48410714e-01
-9.25305724e-01 8.69266272e-01 5.76476431e+00 5.93595207e-01
-1.08179271e+00 -1.61564931e-01 3.17320861e-02 2.92680953e-02
-5.89632928e-01 1.32326394e-01 -7.07784712e-01 7.95857161e-02
2.24376563e-02 1.54711166e-02 2.54564047e-01 9.34230685e-01
-4.32371311e-02 1.66036502e-01 -1.03792655e+00 1.31938148e+00
2.40277126e-02 -1.34011769e+00 3.49042475e-01 1.36333346e-01
7.54800618e-01 3.48679870e-01 -2.65143961e-01 -3.93682159e-02
9.04329680e-03 -7.12746859e-01 9.71729636e-01 1.02850187e+00
8.71689141e-01 -7.49476433e-01 5.08735061e-01 5.75476825e-01
-1.40992260e+00 9.18952823e-02 -8.40596497e-01 2.46756114e-02
6.11652285e-02 8.27103317e-01 -5.92750013e-01 8.74895215e-01
6.62758291e-01 1.06794214e+00 -5.76879203e-01 1.38138294e+00
-4.82220091e-02 -1.15046307e-01 -6.22070611e-01 4.49842244e-01
4.81415801e-02 -2.65546858e-01 9.76097703e-01 5.88282287e-01
5.09289145e-01 2.37136222e-02 2.88770616e-01 1.13834393e+00
-1.08246282e-02 -8.17489475e-02 -7.55679309e-01 3.55970025e-01
6.81559443e-01 1.27295887e+00 -3.86092335e-01 -1.08932950e-01
-6.25685036e-01 8.28942537e-01 4.67774689e-01 1.32895693e-01
-3.02544832e-01 -2.46766701e-01 8.87492716e-01 3.36686939e-01
6.03692353e-01 -6.14074469e-01 -7.44391799e-01 -1.22106338e+00
4.55714464e-01 -3.75451207e-01 -2.65525252e-01 -7.99005747e-01
-1.22233438e+00 3.76100957e-01 -2.54426718e-01 -1.71211314e+00
3.31235081e-02 -4.18799520e-01 -5.31133533e-01 8.65343869e-01
-1.79332387e+00 -1.38806307e+00 -6.51283801e-01 6.31161332e-01
5.83280087e-01 -1.27319530e-01 5.70372939e-01 1.33880243e-01
2.67760009e-02 3.18047076e-01 -1.09574780e-01 -1.17923930e-01
4.29295510e-01 -9.76427376e-01 3.82621437e-01 5.90110540e-01
-6.28698692e-02 5.47621369e-01 3.89394432e-01 -9.35020030e-01
-1.82900655e+00 -1.11916578e+00 7.41199553e-01 -3.53844523e-01
6.08419888e-02 -2.97017485e-01 -9.93491769e-01 6.17482007e-01
-3.08629632e-01 3.71294439e-01 -2.86695398e-02 -3.70887607e-01
-3.62163037e-01 -1.35469466e-01 -1.40942562e+00 3.41270119e-01
1.29726672e+00 -2.19292566e-01 -6.78199351e-01 -1.48668766e-01
8.91793132e-01 -6.22498155e-01 -8.56208742e-01 7.57815659e-01
5.85828960e-01 -1.04686606e+00 1.06888509e+00 -1.07019104e-01
4.89724040e-01 -7.40534067e-01 -4.94564772e-01 -1.14364100e+00
-5.14993608e-01 -5.40019691e-01 -4.29724067e-01 1.23859882e+00
-6.01703487e-02 -4.31744218e-01 9.58983839e-01 5.63950181e-01
-7.11254299e-01 -9.08697665e-01 -1.13671184e+00 -5.51018655e-01
-2.03168374e-02 -3.43204409e-01 1.09931815e+00 8.58703494e-01
-5.32475471e-01 -4.61755757e-04 -1.88843399e-01 4.06721354e-01
7.30715513e-01 4.21746701e-01 9.76546586e-01 -1.53431034e+00
2.76394665e-01 -3.37051123e-01 -7.15522408e-01 -1.28252304e+00
1.52161205e-02 -9.52091753e-01 1.46715477e-01 -1.76853609e+00
-4.78908680e-02 -6.54181302e-01 5.15861809e-03 6.83212698e-01
1.76798254e-01 9.43846926e-02 2.02103630e-01 3.15771431e-01
-2.60677874e-01 9.85509217e-01 1.64240193e+00 -6.52810931e-02
-3.30678225e-01 4.88492399e-02 -6.45205140e-01 8.37851346e-01
5.32624483e-01 -4.08920079e-01 -2.00235501e-01 -7.81980991e-01
4.32985462e-02 -8.87534246e-02 6.35830700e-01 -1.00329280e+00
4.33322370e-01 -2.12766901e-01 6.19525135e-01 -1.33566093e+00
7.17473447e-01 -1.26390576e+00 1.79105580e-01 1.90262586e-01
1.93351775e-01 1.60697028e-02 6.09429665e-02 6.82496548e-01
-2.06657380e-01 -2.17086878e-02 6.02327347e-01 -2.05157042e-01
-4.92659807e-01 7.90649295e-01 3.30177695e-01 -4.21164632e-01
9.24272656e-01 -5.37058473e-01 -1.19294785e-01 -8.91942680e-02
-5.34945309e-01 2.43478477e-01 8.61902356e-01 6.28449678e-01
1.29508722e+00 -1.64462757e+00 -7.76930809e-01 7.57983506e-01
1.76755548e-01 8.44244719e-01 1.55931324e-01 5.59154928e-01
-5.79056501e-01 1.85497150e-01 -2.52063125e-01 -8.34203303e-01
-9.44972277e-01 4.72168595e-01 2.15450034e-01 1.82166815e-01
-1.05872548e+00 5.87042689e-01 2.16122642e-01 -7.91968882e-01
2.77235478e-01 -5.05697370e-01 -1.12103568e-02 -1.37818277e-01
3.08891565e-01 3.75000685e-01 1.68147951e-01 -7.93553114e-01
-4.53071475e-01 1.21137083e+00 2.84717139e-02 2.94662178e-01
1.71383679e+00 -1.67057961e-01 -2.57934183e-01 2.32261479e-01
1.15019310e+00 1.61808938e-01 -1.70908892e+00 -4.91010964e-01
-1.60222143e-01 -7.57852077e-01 9.65538174e-02 -4.93587226e-01
-1.29701149e+00 7.59944081e-01 3.59753489e-01 -1.19932733e-01
9.51964021e-01 -1.48596661e-02 7.11711407e-01 3.54246289e-01
6.17172301e-01 -7.12863624e-01 -9.76308361e-02 6.82852685e-01
1.35175908e+00 -1.01834154e+00 2.94369131e-01 -7.35212088e-01
-3.46138388e-01 1.15615642e+00 5.72318435e-01 -7.80844092e-01
9.90034699e-01 2.72508591e-01 -1.86289504e-01 -4.20658827e-01
-7.01825619e-01 -2.66648918e-01 4.79616016e-01 7.55191684e-01
-7.96261802e-02 -1.73729002e-01 -1.93218142e-02 6.53807938e-01
-2.80046433e-01 -1.00201003e-01 3.24602395e-01 1.01278615e+00
-6.53895319e-01 -1.04635406e+00 -3.62864822e-01 4.62784529e-01
1.94594450e-03 2.81676650e-01 -4.80410516e-01 6.24405682e-01
3.00302416e-01 4.95515108e-01 1.38273388e-01 -4.96589452e-01
6.14996076e-01 -6.52030557e-02 5.47134399e-01 -7.46835053e-01
-3.11399102e-01 3.10684592e-01 -3.10025811e-01 -7.74872839e-01
-3.67012709e-01 -8.19395959e-01 -1.40260887e+00 -1.49684399e-01
-3.02973121e-01 -5.08519173e-01 9.39410925e-01 6.90652311e-01
6.76152289e-01 3.35144341e-01 7.50712037e-01 -1.48399782e+00
-5.97058833e-01 -8.60957921e-01 -5.26955068e-01 3.14002872e-01
6.38496518e-01 -9.75716531e-01 -3.37680876e-01 -1.60576344e-01] | [8.375473022460938, -3.5162010192871094] |
b1246912-f00a-4fc8-b0ea-7381cfafa661 | qvmix-and-qvmix-max-extending-the-deep | 2012.12062 | null | https://arxiv.org/abs/2012.12062v1 | https://arxiv.org/pdf/2012.12062v1.pdf | QVMix and QVMix-Max: Extending the Deep Quality-Value Family of Algorithms to Cooperative Multi-Agent Reinforcement Learning | This paper introduces four new algorithms that can be used for tackling multi-agent reinforcement learning (MARL) problems occurring in cooperative settings. All algorithms are based on the Deep Quality-Value (DQV) family of algorithms, a set of techniques that have proven to be successful when dealing with single-agent reinforcement learning problems (SARL). The key idea of DQV algorithms is to jointly learn an approximation of the state-value function $V$, alongside an approximation of the state-action value function $Q$. We follow this principle and generalise these algorithms by introducing two fully decentralised MARL algorithms (IQV and IQV-Max) and two algorithms that are based on the centralised training with decentralised execution training paradigm (QVMix and QVMix-Max). We compare our algorithms with state-of-the-art MARL techniques on the popular StarCraft Multi-Agent Challenge (SMAC) environment. We show competitive results when QVMix and QVMix-Max are compared to well-known MARL techniques such as QMIX and MAVEN and show that QVMix can even outperform them on some of the tested environments, being the algorithm which performs best overall. We hypothesise that this is due to the fact that QVMix suffers less from the overestimation bias of the $Q$ function. | ['Matthia Sabatelli', 'Jonathan Pisane', 'Gilles Louppe', 'Pierre Geurts', 'Damien Ernst', 'Pascal Leroy'] | 2020-12-22 | null | null | null | null | ['smac-1', 'smac'] | ['playing-games', 'playing-games'] | [-4.89214778e-01 1.91846535e-01 -3.01977426e-01 1.08342633e-01
-7.65151203e-01 -5.03544986e-01 1.00790894e+00 4.35491264e-01
-8.85567546e-01 1.27079129e+00 2.72998493e-02 -3.15295994e-01
-6.83720052e-01 -7.02240765e-01 -7.96778381e-01 -6.95139349e-01
-6.95636749e-01 8.93466890e-01 4.16296005e-01 -9.33938205e-01
2.69879818e-01 1.08408727e-01 -1.56337869e+00 -6.81636576e-03
5.47117949e-01 8.02014351e-01 2.61673313e-02 9.25393522e-01
1.86784342e-01 1.31430805e+00 -9.03565526e-01 -2.57394254e-01
4.22387481e-01 -5.19919455e-01 -1.16801512e+00 -3.17467898e-01
-4.63330671e-02 -3.10541600e-01 -8.94606039e-02 6.35729373e-01
6.07988179e-01 5.02733827e-01 4.89256233e-01 -1.63141680e+00
2.41508670e-02 8.15833747e-01 -5.02701461e-01 1.78505898e-01
3.87223423e-01 5.88956773e-01 1.14904654e+00 -1.06901392e-01
8.18456233e-01 1.27453494e+00 6.06568754e-01 5.63600063e-01
-1.23489332e+00 -8.40867683e-02 6.41101301e-02 4.12662745e-01
-9.87313271e-01 -1.88852891e-01 5.11079967e-01 -2.72550672e-01
1.44445229e+00 -4.99035791e-02 6.72309577e-01 9.79302049e-01
4.49712455e-01 1.09857118e+00 1.44199979e+00 -4.24516231e-01
8.92406642e-01 -1.97078779e-01 -4.45595086e-01 8.02929401e-01
-6.12307265e-02 7.92240798e-01 -4.55365449e-01 -3.49282414e-01
7.28534639e-01 -4.22315240e-01 1.55803248e-01 -8.50026846e-01
-1.43875647e+00 1.18102312e+00 4.41979438e-01 2.89954513e-01
-7.64358222e-01 7.35667050e-01 9.11175072e-01 8.74459147e-01
3.53678204e-02 6.46955073e-01 -7.26035118e-01 -5.52781880e-01
-6.87561810e-01 9.94191766e-01 8.75423670e-01 4.31991965e-01
8.87837827e-01 3.59699219e-01 -1.02584265e-01 5.54779649e-01
4.00245160e-01 2.49024153e-01 7.25580812e-01 -1.44341624e+00
1.50639370e-01 3.01600367e-01 4.37340021e-01 -2.69228190e-01
-5.97805321e-01 -2.61520773e-01 -4.59896058e-01 9.26189423e-01
2.85980582e-01 -4.16333854e-01 -3.30117285e-01 1.71891308e+00
5.47354639e-01 1.19943246e-01 7.19760120e-01 8.10920715e-01
3.59991312e-01 6.11182272e-01 6.76366761e-02 -4.32079136e-01
1.02665412e+00 -1.13675344e+00 -4.63069439e-01 -1.98095992e-01
9.12525296e-01 -3.04745018e-01 8.40655565e-01 5.71579516e-01
-1.15724826e+00 -5.12711108e-01 -1.18963265e+00 7.44507015e-01
-3.79830062e-01 -5.28869867e-01 8.07093978e-01 4.67022032e-01
-1.36378002e+00 9.42165792e-01 -8.75804782e-01 -1.05111599e-01
1.57803804e-01 4.72532541e-01 -2.52096891e-01 3.19648504e-01
-1.20051360e+00 1.30425942e+00 6.83871508e-01 -3.43125492e-01
-1.61083782e+00 -3.55300069e-01 -6.77720904e-01 -2.89743990e-01
6.66174591e-01 -4.83863294e-01 1.87801087e+00 -1.12992513e+00
-2.17142344e+00 5.34505546e-01 6.39003575e-01 -9.53594506e-01
6.06261551e-01 5.06988578e-02 3.78721021e-02 2.17548996e-01
9.60038304e-02 5.58939099e-01 8.25644791e-01 -1.35368621e+00
-8.71631622e-01 2.32335902e-03 4.70327854e-01 2.94785529e-01
3.87794673e-01 -2.83759743e-01 4.47393924e-01 -2.60936558e-01
-8.72170150e-01 -6.37700379e-01 -6.52520061e-01 -4.54036534e-01
3.03503960e-01 -7.81682193e-01 6.23643339e-01 -3.04440171e-01
8.37490320e-01 -1.63893318e+00 7.11529970e-01 8.49392265e-02
1.69148892e-01 4.10739481e-01 -5.57515442e-01 1.11878192e+00
2.39583760e-01 -3.14829320e-01 -1.64819837e-01 -1.96150526e-01
5.87469280e-01 8.02568853e-01 3.51167992e-02 5.86933374e-01
-1.48025602e-02 9.78007555e-01 -1.24135971e+00 -4.82891262e-01
2.96557635e-01 -3.22320946e-02 -5.90797305e-01 4.17540789e-01
-8.69619489e-01 1.69887632e-01 -2.93900251e-01 4.08531159e-01
2.57873863e-01 2.05444857e-01 4.45969075e-01 4.20571834e-01
-1.90990955e-01 -4.96159121e-02 -1.29859149e+00 1.82471108e+00
-4.50319856e-01 1.97759032e-01 3.16765130e-01 -1.50494051e+00
8.63316476e-01 5.41755438e-01 8.12311649e-01 -9.79293048e-01
1.56575054e-01 2.19662771e-01 5.27922623e-02 -4.15719092e-01
3.91827941e-01 -3.97873998e-01 -2.12175578e-01 7.56060481e-01
4.87605244e-01 -4.09441590e-01 5.23710847e-01 2.99671680e-01
1.33700073e+00 6.01822913e-01 7.18418360e-01 -3.79662931e-01
5.54475546e-01 6.34047166e-02 4.41358984e-01 1.00623047e+00
-7.03341782e-01 -4.91490752e-01 8.43688726e-01 -6.08198285e-01
-1.07087147e+00 -1.10982478e+00 5.20981312e-01 1.40951848e+00
-9.53135937e-02 -5.26666224e-01 -5.65277100e-01 -1.10645103e+00
1.12350971e-01 5.71803510e-01 -7.29593515e-01 -1.80774592e-02
-5.55694818e-01 -5.46789527e-01 4.23537284e-01 2.53384411e-01
4.08367068e-01 -1.62374127e+00 -1.38390481e+00 5.68803191e-01
2.37505957e-01 -5.83826184e-01 1.56260077e-02 4.24101233e-01
-5.76150537e-01 -1.11648750e+00 -4.62301672e-01 -4.77477521e-01
-1.25222251e-01 -3.76483738e-01 1.42663383e+00 7.83157945e-02
-1.22648552e-01 8.37884486e-01 -8.47292542e-01 -5.02201557e-01
-8.72762620e-01 1.14555568e-01 1.59258842e-01 -2.70554990e-01
-5.70672415e-02 -5.65016329e-01 -4.59782660e-01 1.16829453e-02
-9.52800274e-01 -2.44140044e-01 6.51250839e-01 1.11160851e+00
2.91955948e-01 2.14973073e-02 8.32533717e-01 -5.22597909e-01
9.02978241e-01 -4.08239901e-01 -9.42606449e-01 1.21978140e-02
-7.54832089e-01 5.19734442e-01 8.24544668e-01 -3.91085118e-01
-4.34653968e-01 -5.49833849e-02 -4.03631419e-01 -3.22122127e-01
-1.29425332e-01 5.35105824e-01 3.51062268e-01 -6.04569279e-02
7.80899823e-01 3.09294432e-01 4.85736609e-01 -7.63663873e-02
5.64856648e-01 3.62074077e-01 2.44715959e-01 -8.37478280e-01
4.30426300e-01 1.16431616e-01 2.27078855e-01 -5.66248477e-01
-4.02566701e-01 -3.36603463e-01 -2.10169241e-01 -4.12004173e-01
5.76159835e-01 -5.89760959e-01 -1.41906393e+00 2.63603985e-01
-7.46368229e-01 -1.13529098e+00 -8.67070913e-01 2.95603365e-01
-1.37502420e+00 2.71465719e-01 -6.17999315e-01 -9.93817449e-01
-3.59428465e-01 -1.19513130e+00 9.23655570e-01 4.70129140e-02
2.25195646e-01 -1.22579360e+00 9.53528106e-01 -2.20158361e-02
6.04009151e-01 5.39834082e-01 6.90894604e-01 -6.29178584e-01
-1.52216792e-01 1.09101817e-01 4.08451259e-01 2.63281673e-01
-4.24867034e-01 -2.58785099e-01 -5.61954737e-01 -9.40965593e-01
-3.76610398e-01 -9.46954966e-01 4.11967367e-01 3.42430651e-01
2.67528176e-01 -4.38223958e-01 1.14978597e-01 -1.97333153e-02
1.78558135e+00 2.36327186e-01 4.79961425e-01 9.77779031e-01
-1.78642962e-02 2.82158464e-01 7.50971973e-01 8.54266584e-01
6.60057604e-01 8.86650264e-01 1.18058884e+00 1.46886975e-01
3.77623647e-01 1.77932263e-03 7.59408832e-01 4.17144507e-01
-2.51341552e-01 2.34818868e-02 -7.78113186e-01 5.65775692e-01
-2.49296999e+00 -1.13455284e+00 2.32195675e-01 1.99124348e+00
7.14740515e-01 1.89941097e-02 9.16693032e-01 8.30916241e-02
8.23440105e-02 2.93692350e-01 -4.97214556e-01 -9.30076540e-01
9.72894803e-02 6.86424077e-01 2.85558134e-01 6.34067893e-01
-1.05843699e+00 9.67288971e-01 6.64511776e+00 7.06690311e-01
-6.06728852e-01 4.33832020e-01 1.06650600e-02 -2.84163523e-02
1.59201309e-01 5.80717903e-03 -3.34292620e-01 2.38133609e-01
1.30306613e+00 -1.71283558e-02 9.07358944e-01 9.59643900e-01
1.45397440e-01 -3.47559363e-01 -9.95742261e-01 7.56437123e-01
-3.18605781e-01 -1.41711795e+00 -2.91861326e-01 1.63637847e-02
7.02094018e-01 3.96717489e-01 -2.56495148e-01 1.09822464e+00
1.06969833e+00 -9.74661052e-01 7.95360684e-01 2.18286619e-01
1.61607668e-01 -1.12749434e+00 9.19973671e-01 5.80158353e-01
-9.52538252e-01 -3.33058238e-01 -3.56301606e-01 -3.38079154e-01
-1.34853706e-01 -9.90180001e-02 -6.32403433e-01 8.52281511e-01
6.26273513e-01 5.78918695e-01 -2.99003810e-01 7.94404089e-01
-3.06856692e-01 3.18265229e-01 -1.18920550e-01 -3.77546400e-01
9.99161959e-01 -8.92742053e-02 4.23633605e-01 9.15673614e-01
-2.28146732e-01 -3.15068245e-01 5.15139163e-01 4.67928767e-01
3.06064278e-01 8.04437473e-02 -4.32910025e-01 5.49525656e-02
1.36200309e-01 1.26627195e+00 -3.66975725e-01 -3.43222082e-01
-2.62196094e-01 9.09652472e-01 6.88821554e-01 -5.95140122e-02
-7.19421506e-01 -2.95941621e-01 8.48555088e-01 -3.33144963e-01
6.73375189e-01 -2.60082543e-01 7.71980762e-01 -8.67032707e-01
-3.53897423e-01 -1.34364712e+00 6.51596844e-01 -4.43492055e-01
-1.10815716e+00 6.24856055e-01 6.09865263e-02 -1.10356069e+00
-9.75975394e-01 -5.27560234e-01 -5.24382889e-01 3.65442514e-01
-1.64956105e+00 -1.01236558e+00 1.98171929e-01 7.53415465e-01
5.35699844e-01 -6.35242820e-01 9.89833057e-01 -2.01928005e-01
-2.56079823e-01 3.45554978e-01 4.63270456e-01 -2.59144694e-01
3.62540185e-01 -1.70174468e+00 7.99461901e-02 4.29457068e-01
4.78245355e-02 -1.96039468e-01 9.49355543e-01 -1.64127752e-01
-1.72671437e+00 -8.23491931e-01 1.54116243e-01 -4.40962642e-01
8.55154574e-01 -2.10967958e-02 -4.54749405e-01 6.53387368e-01
7.98144221e-01 2.26928480e-02 2.26402029e-01 -6.90477043e-02
-3.38635370e-02 -5.35780452e-02 -1.14821517e+00 2.60072500e-01
4.22745645e-01 -1.79999441e-01 -6.90492034e-01 3.56267363e-01
4.83479679e-01 -2.26764098e-01 -1.09948933e+00 1.22782387e-01
2.70712912e-01 -1.26315379e+00 8.36144626e-01 -8.64278555e-01
3.73844266e-01 -2.51489371e-01 -2.74707615e-01 -2.04692364e+00
-2.24512160e-01 -1.03167510e+00 -5.20187795e-01 7.05726027e-01
1.06650805e-02 -6.58183992e-01 6.81185305e-01 -3.55401307e-01
8.83205980e-03 -8.39644372e-01 -1.37239814e+00 -1.05245245e+00
4.77491081e-01 -5.32520562e-02 4.67821240e-01 7.89610922e-01
3.21045697e-01 4.14565682e-01 -5.13299823e-01 -1.27353415e-01
7.61747897e-01 5.14227711e-02 9.93827999e-01 -8.38183403e-01
-9.93276596e-01 -6.80830836e-01 -4.75224048e-01 -5.25766253e-01
3.72080475e-01 -7.32252598e-01 7.82815516e-02 -1.56121397e+00
-2.22883880e-01 -1.47893399e-01 -3.99725407e-01 5.56430578e-01
1.21114030e-01 -1.06783524e-01 4.49092120e-01 -8.09007138e-03
-1.17930210e+00 7.49533355e-01 1.08520734e+00 -7.35203698e-02
-3.38043422e-01 -5.50114959e-02 -2.11938888e-01 3.45851868e-01
8.71210337e-01 -2.53929526e-01 -3.58646899e-01 -8.99711922e-02
4.05801147e-01 5.41616857e-01 3.91299427e-01 -9.73728299e-01
3.64019051e-02 -5.32036185e-01 -1.81354299e-01 -6.06944673e-02
1.17446586e-01 -5.80845773e-01 -1.91765174e-01 1.05785632e+00
-3.44606757e-01 4.84420896e-01 1.09929301e-01 3.88359427e-01
-1.64314225e-01 -3.73459071e-01 9.67729747e-01 -4.56367612e-01
-1.05431855e+00 5.25381118e-02 -7.98806369e-01 2.35040575e-01
1.39897418e+00 1.85420454e-01 -3.47854048e-01 -4.78353143e-01
-4.98080850e-01 6.20081484e-01 2.09211960e-01 2.18229949e-01
5.07633984e-01 -1.17801166e+00 -9.57335770e-01 7.34275021e-03
1.22430570e-01 -3.33037555e-01 4.49796617e-02 8.25034320e-01
-6.33600295e-01 2.20161632e-01 -5.28656363e-01 -2.83265591e-01
-9.17728961e-01 8.06748211e-01 8.27414632e-01 -8.35494637e-01
-5.80377162e-01 4.86687213e-01 -4.63249147e-01 -7.71828651e-01
1.74927965e-01 8.29950273e-02 -8.57236050e-03 -1.20444946e-01
4.31366891e-01 4.90411878e-01 -1.30587548e-01 -3.57864141e-01
-3.00970435e-01 1.37871206e-01 -8.37971643e-02 -5.25400937e-01
1.56161153e+00 2.55456597e-01 1.71633989e-01 2.61827856e-01
7.40280569e-01 -6.50508642e-01 -1.49578130e+00 -2.37626255e-01
1.16194554e-01 6.83400258e-02 1.47193789e-01 -9.53342378e-01
-8.14819574e-01 4.57905382e-01 7.58173645e-01 4.08883750e-01
8.42231452e-01 -1.76275969e-02 3.61070156e-01 4.53674018e-01
8.37267578e-01 -1.48860514e+00 3.68151605e-01 6.48248434e-01
7.85391808e-01 -1.13880432e+00 1.35516077e-01 7.18868136e-01
-9.38386679e-01 9.78734076e-01 6.10798478e-01 -6.05044365e-01
2.40866840e-01 2.37278670e-01 1.12499170e-01 -3.23781103e-01
-1.26702321e+00 -5.61759830e-01 -4.61988389e-01 8.38757634e-01
-2.53042355e-02 1.85324311e-01 -4.14041460e-01 1.06515579e-01
-5.16761579e-02 -1.60110667e-01 5.78150690e-01 1.47653270e+00
-5.04106820e-01 -1.39319146e+00 -3.11424375e-01 1.96503744e-01
-1.02772713e-01 4.83945221e-01 -2.19941363e-01 1.16020799e+00
-5.13455719e-02 8.36688757e-01 -1.46952972e-01 -2.31906280e-01
5.15320361e-01 -8.42200369e-02 7.77031839e-01 -1.28820643e-01
-1.23129451e+00 -9.81323496e-02 5.02245724e-02 -9.32233393e-01
-6.94676876e-01 -5.08642614e-01 -1.23660588e+00 -4.52154696e-01
4.29572165e-02 6.28221512e-01 6.53418243e-01 1.02858639e+00
1.63286887e-02 6.65263653e-01 6.72448575e-01 -1.18481648e+00
-1.24103987e+00 -8.20463419e-01 -7.86175072e-01 2.70014077e-01
4.78129327e-01 -7.62118340e-01 -3.28489579e-02 -4.99345303e-01] | [3.849879264831543, 1.8436626195907593] |
a82e3ba2-721d-42e8-accf-e430b2309837 | cpgnet-cascade-point-grid-fusion-network-for | 2204.09914 | null | https://arxiv.org/abs/2204.09914v3 | https://arxiv.org/pdf/2204.09914v3.pdf | CPGNet: Cascade Point-Grid Fusion Network for Real-Time LiDAR Semantic Segmentation | LiDAR semantic segmentation essential for advanced autonomous driving is required to be accurate, fast, and easy-deployed on mobile platforms. Previous point-based or sparse voxel-based methods are far away from real-time applications since time-consuming neighbor searching or sparse 3D convolution are employed. Recent 2D projection-based methods, including range view and multi-view fusion, can run in real time, but suffer from lower accuracy due to information loss during the 2D projection. Besides, to improve the performance, previous methods usually adopt test time augmentation (TTA), which further slows down the inference process. To achieve a better speed-accuracy trade-off, we propose Cascade Point-Grid Fusion Network (CPGNet), which ensures both effectiveness and efficiency mainly by the following two techniques: 1) the novel Point-Grid (PG) fusion block extracts semantic features mainly on the 2D projected grid for efficiency, while summarizes both 2D and 3D features on 3D point for minimal information loss; 2) the proposed transformation consistency loss narrows the gap between the single-time model inference and TTA. The experiments on the SemanticKITTI and nuScenes benchmarks demonstrate that the CPGNet without ensemble models or TTA is comparable with the state-of-the-art RPVNet, while it runs 4.7 times faster. | ['Zhenhua Wang', 'Hongyu Pan', 'Gang Zhang', 'Xiaoyan Li'] | 2022-04-21 | null | null | null | null | ['robust-3d-semantic-segmentation', 'lidar-semantic-segmentation'] | ['computer-vision', 'computer-vision'] | [-7.42047355e-02 -2.89809048e-01 -1.98867097e-01 -5.84906816e-01
-6.42068386e-01 -1.20979100e-01 5.25485694e-01 -7.94150010e-02
-5.64231813e-01 4.78753954e-01 -3.52855921e-01 -4.04038876e-01
-2.87267387e-01 -1.26883483e+00 -8.24356019e-01 -6.95681334e-01
4.10508722e-01 8.29527020e-01 8.45144272e-01 -1.54383540e-01
2.12812424e-01 5.83311200e-01 -2.05904055e+00 -3.01470608e-01
1.32465589e+00 1.42916667e+00 3.04093122e-01 9.94315520e-02
-3.38881195e-01 2.01810360e-01 -1.54630661e-01 -2.66500652e-01
4.03422266e-01 3.27415317e-01 -4.65207361e-02 -3.05212736e-01
6.11566186e-01 -4.04740602e-01 -4.12400007e-01 1.18818319e+00
6.21644676e-01 2.64367461e-01 3.57501566e-01 -1.52988267e+00
2.35384051e-02 1.04398638e-01 -9.32533920e-01 -1.54550299e-01
-1.63438424e-01 2.14632079e-01 5.29354393e-01 -1.18438745e+00
3.00196707e-01 1.28828263e+00 8.40401590e-01 1.12267487e-01
-8.85941505e-01 -1.08721030e+00 2.76279539e-01 5.82481384e-01
-1.56459785e+00 -2.74002433e-01 8.16729307e-01 -1.29998982e-01
9.40201402e-01 3.24330628e-01 9.22986448e-01 8.11773062e-01
3.49535108e-01 5.46272457e-01 1.12340641e+00 2.21981421e-01
1.61310866e-01 -2.22681724e-02 7.99823031e-02 7.83032119e-01
4.85639244e-01 3.59083056e-01 -5.62448442e-01 4.69628237e-02
6.41709864e-01 2.03940764e-01 -9.84069034e-02 -7.36060500e-01
-1.06663382e+00 8.45673978e-01 6.25569463e-01 -1.20110452e-01
-2.68408805e-01 2.14419201e-01 4.87748712e-01 -6.18876470e-03
6.13873363e-01 -2.66760588e-01 -3.97546172e-01 -1.76518530e-01
-1.05334997e+00 4.67065215e-01 3.99305969e-01 1.05154121e+00
1.14043403e+00 -3.58049236e-02 3.94099541e-02 8.25986207e-01
4.75320905e-01 8.59506667e-01 2.32136980e-01 -1.00895238e+00
7.48644352e-01 7.74317384e-01 -2.11116105e-01 -1.01993608e+00
-4.85660553e-01 -4.15737331e-01 -9.52235162e-01 4.95729566e-01
-1.71998162e-02 2.30001658e-01 -1.19035256e+00 1.28658867e+00
7.48981059e-01 4.76359278e-01 6.83542117e-02 8.77235115e-01
9.69333768e-01 5.50356030e-01 -1.96119085e-01 -9.42705348e-02
1.32233095e+00 -1.04566550e+00 -5.57019353e-01 -4.62107092e-01
5.60596645e-01 -6.49498880e-01 9.57762122e-01 3.11442882e-01
-8.49196970e-01 -8.18386555e-01 -1.24296570e+00 -3.50799203e-01
-3.99937779e-01 -3.63014340e-02 7.00033605e-01 5.37996709e-01
-7.14164853e-01 3.50260973e-01 -9.60173130e-01 -1.09267905e-01
6.59864962e-01 4.00500566e-01 -3.32755178e-01 -4.80853856e-01
-9.28404033e-01 8.88531983e-01 2.32993603e-01 3.33346367e-01
-6.15519047e-01 -1.09855592e+00 -1.02457595e+00 -8.92653465e-02
6.38127625e-01 -9.21266675e-01 9.40649450e-01 1.12872594e-03
-1.42520487e+00 4.70822781e-01 -4.60933089e-01 -4.50680763e-01
7.46241391e-01 -3.28836232e-01 -2.57333428e-01 -1.05308101e-01
3.65372777e-01 9.32127714e-01 4.77210641e-01 -1.29162872e+00
-1.02240443e+00 -8.32566798e-01 -2.01377958e-01 5.28228283e-01
-6.78227516e-03 -6.86746240e-01 -7.95933485e-01 -1.38696074e-01
7.30839312e-01 -8.32198858e-01 -4.30258125e-01 2.47390494e-01
-2.37057075e-01 -2.45221481e-01 1.46684229e+00 -4.35631067e-01
7.58344471e-01 -2.15459323e+00 -2.74782687e-01 3.47272187e-01
2.11918995e-01 2.99918592e-01 2.40570828e-01 -5.83063774e-02
3.86765212e-01 -1.85826838e-01 -4.40069467e-01 -5.32538593e-01
-6.73505068e-02 6.09006524e-01 -2.07856074e-01 5.09978354e-01
-9.17510130e-03 8.56522918e-01 -7.40432620e-01 -7.10285604e-01
7.85024762e-01 6.24540567e-01 -5.20859957e-01 -2.10290134e-01
-1.22748800e-01 3.66034955e-01 -6.53044760e-01 7.12141335e-01
1.29270732e+00 1.51242524e-01 -3.95613462e-01 -4.33946282e-01
-3.85934234e-01 3.66870075e-01 -1.29630899e+00 1.95700192e+00
-6.41722739e-01 2.49981493e-01 1.20361941e-02 -8.12089682e-01
1.08352900e+00 -7.04240501e-02 4.40979958e-01 -9.89897907e-01
1.19807450e-02 3.51816803e-01 -2.12303460e-01 -2.22185403e-01
5.31457365e-01 1.07622728e-01 6.61848560e-02 -2.08302334e-01
-4.06368554e-01 -5.06722808e-01 -9.13186520e-02 -3.57519984e-02
7.48691440e-01 3.71353298e-01 -3.62044456e-03 -1.62873179e-01
5.85462272e-01 2.16040209e-01 9.79959428e-01 4.02078480e-01
-3.48792449e-02 4.74607855e-01 5.08808419e-02 -3.79554719e-01
-6.76374316e-01 -9.40764606e-01 -3.78093809e-01 2.29665786e-01
8.42807651e-01 -2.40571529e-01 -4.13152158e-01 -6.54144824e-01
3.17981809e-01 9.04199004e-01 -1.95656970e-01 -1.40465140e-01
-6.00283563e-01 -6.31342649e-01 3.28473926e-01 7.18163311e-01
1.13686073e+00 -3.96810293e-01 -8.17153096e-01 1.83611587e-01
-1.71363473e-01 -1.31866181e+00 -2.10498914e-01 5.85863777e-02
-1.13044202e+00 -8.49434793e-01 -2.17568785e-01 -3.98875713e-01
5.40258884e-01 7.18336821e-01 7.95966804e-01 -7.12023750e-02
8.06437954e-02 -8.43052119e-02 -8.81237611e-02 -5.48079908e-01
2.57943928e-01 -7.67425820e-02 8.88528451e-02 -2.22822592e-01
4.84466285e-01 -9.29355204e-01 -7.16831863e-01 5.73207200e-01
-4.77987885e-01 3.65121216e-01 5.99241972e-01 7.12597728e-01
1.20356143e+00 2.79715896e-01 1.04480043e-01 -6.04839444e-01
-1.37650341e-01 -2.47090042e-01 -9.44544077e-01 -1.57523245e-01
-9.37008083e-01 -1.84805289e-01 3.25656831e-01 -8.05262029e-02
-1.08571863e+00 2.47704029e-01 -4.26718861e-01 -8.11721742e-01
-3.45687009e-02 3.08816195e-01 -4.87296313e-01 -4.11292642e-01
2.41483420e-01 3.24333698e-01 6.92598820e-02 -3.57493490e-01
3.79512310e-01 4.86645252e-01 5.61156154e-01 -3.71697605e-01
1.00475645e+00 9.34524238e-01 3.77780616e-01 -6.76855206e-01
-6.16304815e-01 -6.07263327e-01 -5.06959379e-01 -2.29240969e-01
8.13713074e-01 -1.22209167e+00 -7.48721838e-01 4.54248726e-01
-9.94961381e-01 9.24091320e-03 -1.87908605e-01 7.54020691e-01
-5.56228817e-01 3.35759580e-01 -6.80264011e-02 -7.65077233e-01
-3.87718081e-01 -1.41446662e+00 1.31457520e+00 3.09374720e-01
3.93055111e-01 -4.49802279e-01 -2.52148569e-01 5.70858955e-01
2.84999013e-01 2.42545158e-01 6.56253099e-01 -1.52324975e-01
-8.94924939e-01 -5.95868342e-02 -5.53774118e-01 2.28819683e-01
-2.04954460e-01 -3.29191945e-02 -1.11403263e+00 -1.01637118e-01
1.39522955e-01 5.57840383e-03 9.70729291e-01 4.46101397e-01
1.24525785e+00 1.90899625e-01 -7.07546711e-01 1.04252243e+00
1.42132580e+00 2.16123134e-01 5.94951034e-01 3.22349906e-01
9.49186087e-01 4.53614712e-01 1.11946487e+00 5.81335537e-02
8.71550560e-01 7.95269728e-01 8.19334567e-01 -1.56255037e-01
-1.48273483e-01 -3.21988314e-01 7.53194466e-02 9.61430728e-01
-8.88531208e-02 4.10817824e-02 -8.29196334e-01 4.41850483e-01
-2.07045436e+00 -6.07335269e-01 -4.43892866e-01 2.33303666e+00
2.49960512e-01 5.50418973e-01 -2.30618685e-01 2.67141759e-01
4.97214735e-01 2.03208685e-01 -8.96487117e-01 -4.24093083e-02
2.71134148e-03 1.52596444e-01 1.00389910e+00 3.88471514e-01
-8.91808629e-01 8.86686325e-01 4.55431223e+00 1.25800300e+00
-1.02044010e+00 3.79666895e-01 4.59340304e-01 -2.07577720e-01
-3.26890707e-01 5.63965179e-02 -1.22297525e+00 3.14815968e-01
6.49233103e-01 9.51702446e-02 1.33084729e-01 1.13160801e+00
1.74993217e-01 -3.25858533e-01 -7.39102185e-01 1.19362450e+00
-7.81602785e-02 -1.32430935e+00 -8.75911191e-02 2.72902995e-01
5.24842978e-01 6.14398539e-01 -1.21354297e-01 4.15862352e-01
1.49175644e-01 -6.62832618e-01 8.39011490e-01 3.33715767e-01
8.41441333e-01 -1.07354593e+00 7.79669285e-01 5.92954218e-01
-1.56756127e+00 5.53991906e-02 -6.09888911e-01 2.77590044e-02
3.75608176e-01 1.04858065e+00 -4.91047949e-01 1.03965938e+00
9.79828417e-01 7.51174510e-01 -4.20103371e-01 1.02975476e+00
-1.49132848e-01 1.94319755e-01 -8.01081061e-01 7.55912587e-02
3.96154612e-01 -5.09207904e-01 7.13026464e-01 6.31318688e-01
6.01308167e-01 -7.90631697e-02 3.23076189e-01 7.91349590e-01
1.34867743e-01 -1.52221560e-01 -6.37009919e-01 6.27184391e-01
7.25204647e-01 1.36514783e+00 -8.95325243e-01 -3.82038265e-01
-5.53878486e-01 5.47910571e-01 8.24627187e-03 9.88115370e-02
-1.07145607e+00 -3.51085484e-01 8.62030268e-01 3.08195174e-01
4.86276358e-01 -4.60812181e-01 -6.74724400e-01 -8.01626444e-01
3.38739932e-01 -3.96795183e-01 2.13171486e-02 -6.78531408e-01
-9.69981670e-01 4.74836856e-01 1.35248035e-01 -1.39021397e+00
1.79790154e-01 -3.99412662e-01 -3.17530960e-01 8.50720108e-01
-1.87749505e+00 -1.23529792e+00 -7.66994715e-01 5.56368530e-01
6.46400630e-01 2.21734777e-01 4.06963140e-01 5.92945874e-01
-5.00054896e-01 3.36598426e-01 -1.35784432e-01 -4.30735767e-01
4.16397482e-01 -8.74845684e-01 7.08850145e-01 8.32941651e-01
3.79528813e-02 1.14056334e-01 3.38173211e-01 -7.55812347e-01
-1.54113054e+00 -1.40951741e+00 5.98465323e-01 -4.27464247e-01
1.55261233e-01 -2.61164844e-01 -8.67646575e-01 3.67732167e-01
-3.71856004e-01 2.71232963e-01 1.17467940e-02 -7.38229752e-02
-1.80478141e-01 -5.84774673e-01 -1.17909241e+00 4.56460923e-01
1.38803875e+00 -2.07922876e-01 -3.53536934e-01 2.18142033e-01
9.75460827e-01 -8.95899475e-01 -7.04473794e-01 1.04097974e+00
4.78987128e-01 -1.13566077e+00 1.12758696e+00 3.05946141e-01
6.38273507e-02 -7.52989054e-01 -3.49621087e-01 -9.47398186e-01
-9.29170921e-02 -1.50104314e-01 -1.39660150e-01 1.04594195e+00
1.58361316e-01 -1.02828300e+00 8.92699301e-01 2.50360548e-01
-6.60176098e-01 -1.03558636e+00 -1.46981859e+00 -8.51871312e-01
-3.01869482e-01 -9.37790215e-01 9.94553804e-01 5.67880929e-01
-8.01812053e-01 2.19933957e-01 9.87904295e-02 5.07444322e-01
7.19301879e-01 2.38616332e-01 1.04780293e+00 -1.53227758e+00
2.27209002e-01 -3.60187888e-01 -6.80701852e-01 -1.25303638e+00
-1.28359050e-01 -6.99278891e-01 1.64705455e-01 -1.72783399e+00
-1.39079332e-01 -8.30178201e-01 -8.99893120e-02 2.79555589e-01
-4.95203175e-02 3.31436217e-01 -4.93302271e-02 1.64607406e-01
-3.02748948e-01 8.79252374e-01 1.37422860e+00 -1.28011346e-01
-1.41130611e-01 7.45125413e-02 -4.35882449e-01 9.02505815e-01
4.68241632e-01 -5.56173146e-01 -6.84831023e-01 -6.04980350e-01
1.56659603e-01 -9.56085697e-02 4.79061425e-01 -1.41698432e+00
4.77121264e-01 -8.50599930e-02 2.75706351e-01 -1.53163397e+00
8.50000083e-01 -9.84321773e-01 2.60182530e-01 4.97180820e-01
6.32682741e-01 7.59134516e-02 2.66324788e-01 7.72094786e-01
-2.47278824e-01 1.68025672e-01 7.64424264e-01 6.00686893e-02
-8.44762981e-01 7.95388043e-01 1.91583067e-01 -3.41010243e-01
1.14452779e+00 -6.61405504e-01 -2.01655343e-01 -7.03391386e-03
-9.97208953e-02 5.84146678e-01 4.78571415e-01 3.79941851e-01
7.51127243e-01 -1.43354988e+00 -4.88887906e-01 4.12904024e-01
9.02040377e-02 9.48196232e-01 5.56564033e-01 1.18407536e+00
-5.69096386e-01 3.46562445e-01 -4.85473946e-02 -1.19490564e+00
-1.11734688e+00 2.59603411e-01 7.99564347e-02 -1.59102678e-01
-1.03227782e+00 9.24152672e-01 3.17514151e-01 -7.58736849e-01
8.58292803e-02 -5.13314188e-01 -6.51090145e-02 -5.64538613e-02
2.88261384e-01 5.51939726e-01 4.86455292e-01 -6.60917580e-01
-4.42622602e-01 1.05772305e+00 1.68910995e-02 1.48755731e-04
1.13536000e+00 -1.34230331e-01 3.11148469e-03 3.45313370e-01
9.66004968e-01 -1.44102409e-01 -1.38571310e+00 -1.41054928e-01
-5.04035294e-01 -7.41958797e-01 6.44482851e-01 -3.57870370e-01
-1.28835511e+00 1.02849782e+00 7.65615523e-01 -1.57853663e-01
1.06451619e+00 -4.03639764e-01 1.28260422e+00 2.99639016e-01
6.76973701e-01 -1.00214767e+00 -4.92359817e-01 4.72127944e-01
5.77912390e-01 -1.17824090e+00 3.18114102e-01 -8.85930002e-01
-3.47024381e-01 8.13771665e-01 8.74443173e-01 -4.33820002e-02
7.74231315e-01 2.78790087e-01 -7.25662857e-02 -2.72721916e-01
-4.36156660e-01 -1.11662783e-01 2.66085446e-01 5.63717246e-01
-3.74046624e-01 1.41350040e-02 -2.26263806e-01 4.49747950e-01
-4.73344028e-01 -4.17741574e-02 -7.58467019e-02 7.81029463e-01
-3.99691969e-01 -8.29145253e-01 -3.46406728e-01 6.67279482e-01
1.70178056e-01 6.63976744e-02 2.36891195e-01 9.54869866e-01
5.12646139e-01 7.75944412e-01 3.57307225e-01 -6.64853692e-01
6.85200870e-01 -1.99585631e-01 2.95362353e-01 -3.21541578e-01
-4.91640903e-02 7.61015713e-02 -5.96361905e-02 -1.06085873e+00
-3.90017450e-01 -6.71435356e-01 -1.40202856e+00 -4.88120168e-01
-6.43775344e-01 -1.84577089e-02 1.06431293e+00 9.68540728e-01
5.75298250e-01 5.78671873e-01 5.20160258e-01 -1.02107227e+00
-3.57854277e-01 -8.22653472e-01 -3.14852476e-01 -2.51430064e-01
-4.61769029e-02 -1.15023792e+00 -2.89265126e-01 -5.74793279e-01] | [8.094676971435547, -2.625796318054199] |
cb4c7602-deb8-4dd9-b9c3-68240b6137e1 | collagan-collaborative-gan-for-missing-image-1 | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Lee_CollaGAN_Collaborative_GAN_for_Missing_Image_Data_Imputation_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Lee_CollaGAN_Collaborative_GAN_for_Missing_Image_Data_Imputation_CVPR_2019_paper.pdf | CollaGAN: Collaborative GAN for Missing Image Data Imputation | In many applications requiring multiple inputs to obtain a desired output, if any of the input data is missing, it often introduces large amounts of bias. Although many techniques have been developed for imputing missing data, the image imputation is still difficult due to complicated nature of natural images. To address this problem, here we proposed a novel framework for missing image data imputation, called Collaborative Generative Adversarial Network (CollaGAN). CollaGAN convert the image imputation problem to a multi-domain images-to-image translation task so that a single generator and discriminator network can successfully estimate the missing data using the remaining clean data set. We demonstrate that CollaGAN produces the images with a higher visual quality compared to the existing competing approaches in various image imputation tasks.
| [' Jong Chul Ye', ' Won-Jin Moon', ' Junyoung Kim', 'Dongwook Lee'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['image-imputation'] | ['computer-vision'] | [ 7.59297013e-01 1.78758036e-02 7.44424239e-02 -5.01604974e-01
-1.01139915e+00 -5.15194833e-01 3.53202105e-01 -5.89779556e-01
-3.62273008e-02 1.23614955e+00 1.69822648e-01 6.13637315e-03
2.37554461e-01 -7.17076004e-01 -1.17686439e+00 -8.63710225e-01
6.86947465e-01 2.67929763e-01 -5.45399547e-01 1.32573381e-01
7.97987580e-02 1.58259869e-01 -1.45058811e+00 5.50558150e-01
1.28737652e+00 5.93347907e-01 2.19289362e-01 6.89106703e-01
-2.10938260e-01 1.04329252e+00 -9.02932584e-01 -5.89242518e-01
4.84612107e-01 -8.73118222e-01 -3.48675191e-01 8.13869312e-02
4.28875655e-01 -4.27314848e-01 -1.93136767e-01 1.16692007e+00
6.10700190e-01 -6.71866536e-02 9.29827750e-01 -1.81863904e+00
-1.28539860e+00 4.69094425e-01 -8.60213041e-01 -4.09432590e-01
1.93233028e-01 1.51118696e-01 1.29849672e-01 -9.39120770e-01
7.44576752e-01 1.08944607e+00 6.68618202e-01 6.24124110e-01
-1.59065938e+00 -1.05537951e+00 -3.13314408e-01 -2.20793560e-01
-1.27603686e+00 -5.48956692e-01 8.74021113e-01 -4.15364683e-01
2.62061507e-01 1.78141192e-01 4.42544967e-02 1.60506892e+00
1.72529340e-01 4.93106961e-01 1.49417615e+00 -2.79213190e-01
-6.29522502e-02 2.59741813e-01 -4.00301486e-01 1.89516753e-01
1.88826144e-01 1.64177403e-01 -4.45126861e-01 -1.09797120e-01
8.57207835e-01 2.73801535e-01 -6.43473119e-02 -9.44155380e-02
-1.41868532e+00 6.79735780e-01 3.79775256e-01 -2.01779440e-01
-3.58509988e-01 8.75623748e-02 1.10854708e-01 5.12964606e-01
3.80067497e-01 9.00403485e-02 -1.45144552e-01 2.95935214e-01
-1.00103915e+00 3.20766985e-01 4.45784867e-01 1.10611033e+00
7.68363476e-01 4.18077767e-01 -3.49142581e-01 8.29850197e-01
6.02515507e-03 8.18583190e-01 3.83004457e-01 -1.12091446e+00
8.30495954e-01 3.75357658e-01 3.95222783e-01 -9.78413641e-01
1.60027742e-01 -3.16586286e-01 -1.74197733e+00 7.09201515e-01
5.27846575e-01 -4.28741246e-01 -1.16470742e+00 1.98877716e+00
1.61836103e-01 2.76668608e-01 3.77551049e-01 8.39367986e-01
8.61468077e-01 7.78914571e-01 4.53794003e-02 -1.41782463e-01
8.41896534e-01 -7.97419846e-01 -9.05199051e-01 -3.19684982e-01
-2.27382705e-01 -1.00717425e+00 9.28345144e-01 3.24968636e-01
-1.06422246e+00 -9.32721496e-01 -9.34523642e-01 -2.46289313e-01
-1.41569853e-01 1.88688293e-01 3.11060160e-01 4.26157922e-01
-7.82711923e-01 2.08100677e-01 -2.86066681e-01 2.01766014e-01
6.29253030e-01 3.95223767e-01 -8.00374806e-01 -2.97114402e-01
-9.49101448e-01 6.54361725e-01 2.87790269e-01 2.24642083e-01
-9.62053716e-01 -6.81710780e-01 -8.50060880e-01 -1.64299130e-01
1.10205807e-01 -1.13649821e+00 9.48775053e-01 -1.61314964e+00
-1.16651356e+00 6.71256363e-01 -2.96090454e-01 -2.23054931e-01
8.36687028e-01 2.47713998e-02 -3.53899121e-01 -5.01582980e-01
2.20940098e-01 8.39994252e-01 1.39206195e+00 -1.74376142e+00
-1.41711354e-01 -3.52137774e-01 -4.41724151e-01 9.93560329e-02
1.96600407e-01 -1.76054850e-01 -8.86789635e-02 -1.14492381e+00
1.29218176e-01 -7.73472548e-01 -1.96508974e-01 -1.14053287e-01
-6.72172785e-01 4.90243465e-01 8.84854555e-01 -8.58409464e-01
5.39896667e-01 -2.09655237e+00 4.11957711e-01 -3.03128790e-02
2.63213366e-01 -4.55646627e-02 -4.05105233e-01 4.01514620e-01
-1.80013195e-01 2.02537715e-01 -5.49889266e-01 -6.48902059e-01
-8.49725306e-02 3.71655256e-01 -6.18320107e-01 2.45247111e-01
1.98999673e-01 1.05887079e+00 -5.70739210e-01 -4.40297604e-01
2.14941770e-01 7.80554950e-01 -3.33850116e-01 6.43675923e-01
-6.35443926e-02 1.11412871e+00 -9.55783352e-02 6.77152455e-01
1.21617496e+00 -8.56556296e-02 -1.92434177e-01 -1.35130197e-01
3.71891409e-01 -6.00846171e-01 -1.19437957e+00 1.72451556e+00
-4.53054816e-01 5.40593863e-01 2.05142591e-02 -7.67535865e-01
1.12373388e+00 3.63459229e-01 3.18702400e-01 -6.91818178e-01
1.30385598e-02 -4.19102274e-02 -1.17529228e-01 -4.89770055e-01
3.86286676e-01 -3.47272247e-01 -2.64433235e-01 4.13673103e-01
1.17109880e-01 -1.53370023e-01 -3.09787095e-01 1.50862366e-01
8.42669189e-01 5.23083746e-01 1.12204682e-02 3.08141887e-01
2.29019538e-01 1.09865695e-01 9.73693430e-01 1.00724292e+00
9.13682356e-02 1.41706479e+00 2.90124774e-01 -3.40932697e-01
-1.67835343e+00 -1.27283049e+00 1.07216716e-01 5.93025386e-01
4.94069569e-02 4.10047650e-01 -9.77917194e-01 -5.23822188e-01
1.11811630e-01 4.61357266e-01 -6.73095584e-01 -1.95879247e-02
-3.12167436e-01 -7.90499151e-01 7.23284006e-01 5.13155997e-01
6.99557543e-01 -1.39047205e+00 1.88955024e-01 7.49095604e-02
-5.21115959e-01 -1.09783614e+00 -6.58861160e-01 -6.53871596e-02
-7.55698442e-01 -8.31797600e-01 -8.86251688e-01 -8.17977071e-01
1.15047085e+00 1.99790239e-01 1.14282346e+00 -9.12336931e-02
-1.61056593e-01 -3.32781911e-01 -2.24450469e-01 -5.45463026e-01
-7.67625332e-01 -2.07511261e-01 -9.51148421e-02 2.22156301e-01
1.51766777e-01 -5.93059182e-01 -5.13419569e-01 1.58624962e-01
-1.26624966e+00 6.43569827e-01 8.42847347e-01 1.22606111e+00
6.60369098e-01 1.10593326e-01 1.00809181e+00 -1.25388324e+00
6.85597718e-01 -7.54469991e-01 -4.24655020e-01 1.23648576e-01
-3.45495701e-01 1.58152267e-01 9.67279911e-01 -5.62352836e-01
-1.29089034e+00 4.74603921e-01 -1.77030265e-01 -5.67942977e-01
-4.51807231e-01 2.55389571e-01 -5.02187371e-01 6.94416687e-02
5.84000230e-01 4.93461519e-01 2.74928123e-01 -4.51470733e-01
3.58227253e-01 7.89481044e-01 1.04746187e+00 -4.21399474e-01
1.09211767e+00 2.50267804e-01 -1.04114294e-01 -2.40710154e-01
-7.13280678e-01 1.07398912e-01 -7.47223258e-01 -2.89219674e-02
8.19397569e-01 -1.23847866e+00 -4.82371658e-01 7.72322714e-01
-1.36166549e+00 -1.13536306e-01 -2.89105177e-01 2.09009334e-01
-5.47858357e-01 1.54196620e-01 -2.88145959e-01 -7.26657271e-01
-4.45001453e-01 -1.28594744e+00 8.38734686e-01 1.99003935e-01
6.02990128e-02 -5.41211367e-01 -2.06771772e-02 5.95851898e-01
4.27036405e-01 9.91373241e-01 8.38822246e-01 7.48430984e-03
-6.78640425e-01 -2.36729145e-01 -3.68639499e-01 5.99107444e-01
1.51869401e-01 -1.42734319e-01 -9.49146986e-01 -1.55206770e-01
1.38602078e-01 -3.47885311e-01 7.49335408e-01 3.67386371e-01
1.33764482e+00 -5.01600981e-01 1.42073603e-02 7.40548134e-01
1.75622594e+00 -3.90148489e-03 1.06893814e+00 -1.25672168e-03
8.12689543e-01 3.07886332e-01 3.55192274e-01 3.47977191e-01
2.54042298e-01 5.18768549e-01 5.00014722e-01 -5.30444741e-01
-3.06923002e-01 -6.45126462e-01 3.22325557e-01 6.44643188e-01
-3.49604376e-02 -4.34120178e-01 -4.77791578e-01 5.45843363e-01
-1.95652115e+00 -1.14235187e+00 -3.23011816e-01 2.11945653e+00
9.77329075e-01 -2.99659520e-01 -1.28282830e-01 -1.42072309e-02
8.61636758e-01 -1.49652854e-01 -8.80509257e-01 -3.05018395e-01
-5.41922688e-01 1.88144207e-01 5.45283377e-01 2.75785983e-01
-7.92480826e-01 5.27724802e-01 6.80292177e+00 6.23079538e-01
-5.82841694e-01 3.15647364e-01 8.43515813e-01 2.78139770e-01
-3.83726299e-01 -2.14844104e-02 -4.32215720e-01 8.93937588e-01
6.22892857e-01 5.50456010e-02 5.84138513e-01 6.11674726e-01
2.83411682e-01 -8.57587606e-02 -8.54022503e-01 1.19771993e+00
2.00922221e-01 -1.15685606e+00 2.15626955e-01 9.24676582e-02
1.16831636e+00 -3.48043352e-01 1.33864239e-01 6.72065690e-02
6.14308834e-01 -1.52842760e+00 5.40228665e-01 9.73255873e-01
1.07297766e+00 -8.28585565e-01 9.08976972e-01 7.15955973e-01
-4.70686853e-01 1.06083147e-01 -6.11319423e-01 -1.03085980e-01
7.48123080e-02 8.27762723e-01 -4.29670483e-01 5.92731476e-01
3.88401449e-01 3.80905807e-01 -5.02026260e-01 9.23652709e-01
-2.71696866e-01 4.74905670e-01 7.38404766e-02 6.59229636e-01
-3.49744767e-01 -4.90030944e-01 2.87435263e-01 7.56738603e-01
6.88601315e-01 -1.89857855e-01 1.66260805e-02 1.24080372e+00
-6.33439302e-01 -1.47525236e-01 -1.00725818e+00 3.12670231e-01
4.47422117e-01 1.11427402e+00 -2.13916168e-01 -2.72634745e-01
-4.04429644e-01 1.52241158e+00 1.70524195e-01 5.68276942e-01
-1.04615521e+00 -3.56863707e-01 5.37595332e-01 -1.59265801e-01
-1.44729823e-01 -1.24677263e-01 -7.96677768e-01 -1.36724818e+00
1.07565433e-01 -1.17951941e+00 7.15305284e-02 -1.34710860e+00
-1.48808396e+00 6.56126678e-01 -3.70792270e-01 -1.33413994e+00
-4.09978211e-01 -4.77344505e-02 -5.18658221e-01 1.31511617e+00
-1.34690380e+00 -1.62639213e+00 -7.60533392e-01 7.79112279e-01
4.32989925e-01 -4.48813826e-01 9.83370841e-01 2.63776749e-01
-3.80640447e-01 6.01842284e-01 5.70101321e-01 2.61007547e-01
1.02266085e+00 -1.21040857e+00 4.33600634e-01 1.00959098e+00
1.98280647e-01 3.12781721e-01 7.08296895e-01 -7.25991130e-01
-1.50592017e+00 -1.32174206e+00 5.86569965e-01 -6.10835552e-01
-6.98629543e-02 -5.14828026e-01 -1.00372744e+00 8.52329254e-01
6.54288292e-01 2.58972466e-01 6.39019370e-01 -3.35378379e-01
-3.94455194e-01 -1.48033902e-01 -1.58440208e+00 3.91352206e-01
8.32779050e-01 -2.76591420e-01 -4.13810402e-01 6.96049333e-02
6.96440160e-01 -6.18019581e-01 -8.09619308e-01 3.16892654e-01
4.29256409e-01 -9.16576385e-01 1.18141544e+00 -4.67340648e-01
9.16266620e-01 -6.56818449e-01 -9.53549892e-02 -1.42273498e+00
-2.84509391e-01 -3.94707739e-01 3.39160897e-02 1.63407230e+00
2.90268153e-01 -3.20516974e-01 7.22571671e-01 6.66227639e-01
2.43236914e-01 1.73884071e-02 -7.20747471e-01 -4.99776572e-01
2.23514155e-01 -1.75137937e-01 1.01336086e+00 8.82035315e-01
-7.99326837e-01 2.46378258e-01 -1.39307594e+00 1.87840670e-01
1.17340839e+00 3.41563612e-01 1.24696946e+00 -9.51881647e-01
-3.08632791e-01 3.11935425e-01 -2.07779869e-01 -4.93958235e-01
2.41300896e-01 -9.56505001e-01 1.48690715e-01 -1.48124909e+00
6.32255435e-01 -4.26835209e-01 -2.24061728e-01 6.65869832e-01
-2.79048651e-01 8.38944614e-01 2.00324252e-01 3.64036649e-01
-9.98919904e-02 4.15979713e-01 1.51089942e+00 -3.28215152e-01
-1.63701642e-02 -2.47631818e-02 -1.01872981e+00 3.99450809e-01
8.98752391e-01 -8.44766915e-01 -3.89674753e-01 -7.45007038e-01
7.20195174e-02 4.70852643e-01 7.75588632e-01 -9.49764013e-01
1.47975549e-01 -4.22641546e-01 1.16735411e+00 -5.61277628e-01
2.04796076e-01 -9.60506618e-01 1.09500527e+00 4.94917482e-02
-2.74459630e-01 6.98020507e-04 2.36787945e-02 4.39405829e-01
-4.66764450e-01 -1.11806847e-01 7.88810492e-01 -2.01526821e-01
-4.86737579e-01 2.87178308e-01 3.85236293e-02 -5.27717285e-02
9.54938352e-01 -2.01025933e-01 -3.50781262e-01 -5.12885809e-01
-7.54039049e-01 -8.01391006e-02 6.63350940e-01 5.28792262e-01
8.41132998e-01 -1.62948811e+00 -1.13101256e+00 5.12748539e-01
-6.75262064e-02 1.08758941e-01 4.64740783e-01 4.59332436e-01
-9.70574394e-02 -3.49797577e-01 -5.65507770e-01 -3.80033404e-01
-1.19579613e+00 6.92854524e-01 4.59324345e-02 2.10403167e-02
-6.68255866e-01 2.66523749e-01 1.91077620e-01 -6.65565014e-01
-1.49422169e-01 1.24013938e-01 2.34377101e-01 -3.31779778e-01
4.80138600e-01 1.06689475e-01 -9.62950438e-02 -7.20185518e-01
1.53395355e-01 3.44322950e-01 2.44467169e-01 1.20228494e-03
1.38375473e+00 -2.46861890e-01 -1.92618206e-01 1.67526469e-01
1.03345311e+00 -1.03879357e-02 -1.46072876e+00 -6.95188120e-02
-6.74477994e-01 -8.25443983e-01 -1.45521730e-01 -1.13084137e+00
-1.16325748e+00 8.31126511e-01 6.62708938e-01 -1.47603348e-01
1.37951350e+00 -5.70684612e-01 7.71625459e-01 -6.17650375e-02
3.91710103e-01 -7.39436507e-01 -2.31433690e-01 6.25039190e-02
1.14518642e+00 -1.80199301e+00 -1.34367988e-01 -1.78635597e-01
-9.85331535e-01 7.84772098e-01 5.80088377e-01 -1.68396756e-01
2.23641708e-01 7.41463363e-01 2.11902350e-01 3.90436947e-01
-4.15854126e-01 1.23956755e-01 6.35662302e-02 1.08987808e+00
1.63897783e-01 -1.26370313e-02 -4.16253880e-03 7.70231307e-01
-1.03713229e-01 4.78520870e-01 6.39155030e-01 5.42949438e-01
1.43924896e-02 -1.57704246e+00 -8.34051788e-01 4.74630743e-01
-6.07543766e-01 -1.49582863e-01 -2.65603125e-01 4.39514011e-01
5.14681458e-01 9.03706014e-01 -1.36286125e-01 -3.52935374e-01
2.09515348e-01 -8.44159350e-03 3.37761641e-01 -3.52648109e-01
-4.75424051e-01 -1.11543402e-01 -4.49752599e-01 -2.00989828e-01
-5.77781677e-01 -5.17537117e-01 -9.70163941e-01 -4.17543679e-01
5.81189711e-03 -9.98640880e-02 7.10629463e-01 6.17833316e-01
6.58351779e-01 5.20116389e-01 7.12070704e-01 -8.03644717e-01
-2.80286074e-01 -1.01250172e+00 -5.52772582e-01 1.07921147e+00
4.15502369e-01 -3.74817997e-01 -1.58632115e-01 6.14587486e-01] | [11.70372200012207, -0.41153597831726074] |
40c46f0f-d979-4142-bb18-580cb438c33f | dalaj-a-dataset-for-linguistic-acceptability-1 | null | null | https://aclanthology.org/2021.nlp4call-1.3 | https://aclanthology.org/2021.nlp4call-1.3.pdf | DaLAJ – a dataset for linguistic acceptability judgments for Swedish | null | ['Julia Klezl', 'Yousuf Ali Mohammed', 'Elena Volodina'] | null | null | null | null | nlp4call-nodalida-2021-5 | ['linguistic-acceptability'] | ['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] | [-7.513768196105957, 3.5689454078674316] |
327cb444-e124-42e2-98b3-ac11394ac59b | dont-miss-the-labels-label-semantic-augmented | null | null | https://aclanthology.org/2021.findings-acl.245 | https://aclanthology.org/2021.findings-acl.245.pdf | Don’t Miss the Labels: Label-semantic Augmented Meta-Learner for Few-Shot Text Classification | null | ['Wei zhang', 'YuHao Lin', 'Lingqiao Liu', 'Qiaoyang Luo'] | null | null | null | null | findings-acl-2021-8 | ['few-shot-text-classification'] | ['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] | [-7.348437786102295, 3.6334152221679688] |
4e80d4a5-31e4-447f-8a33-969dc4706dbb | mobilebrick-building-lego-for-3d | 2303.01932 | null | https://arxiv.org/abs/2303.01932v2 | https://arxiv.org/pdf/2303.01932v2.pdf | MobileBrick: Building LEGO for 3D Reconstruction on Mobile Devices | High-quality 3D ground-truth shapes are critical for 3D object reconstruction evaluation. However, it is difficult to create a replica of an object in reality, and even 3D reconstructions generated by 3D scanners have artefacts that cause biases in evaluation. To address this issue, we introduce a novel multi-view RGBD dataset captured using a mobile device, which includes highly precise 3D ground-truth annotations for 153 object models featuring a diverse set of 3D structures. We obtain precise 3D ground-truth shape without relying on high-end 3D scanners by utilising LEGO models with known geometry as the 3D structures for image capture. The distinct data modality offered by high-resolution RGB images and low-resolution depth maps captured on a mobile device, when combined with precise 3D geometry annotations, presents a unique opportunity for future research on high-fidelity 3D reconstruction. Furthermore, we evaluate a range of 3D reconstruction algorithms on the proposed dataset. Project page: http://code.active.vision/MobileBrick/ | ['Victor Adrian Prisacariu', 'Philip H. S. Torr', 'Robert Castle', 'Jia-Wang Bian', 'Kejie Li'] | 2023-03-03 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Li_MobileBrick_Building_LEGO_for_3D_Reconstruction_on_Mobile_Devices_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Li_MobileBrick_Building_LEGO_for_3D_Reconstruction_on_Mobile_Devices_CVPR_2023_paper.pdf | cvpr-2023-1 | ['3d-object-reconstruction', 'object-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 2.05687404e-01 8.42754450e-03 2.32026309e-01 -4.28165436e-01
-1.00449133e+00 -7.69508362e-01 4.31961447e-01 -2.54385859e-01
-4.07750867e-02 2.73822337e-01 5.80412485e-02 -3.11827697e-02
-1.32677341e-02 -8.61283958e-01 -9.03044641e-01 -2.89009273e-01
1.67427897e-01 9.57366407e-01 3.88653070e-01 -1.46967769e-01
2.58150309e-01 9.64173079e-01 -1.92698193e+00 2.48804018e-02
3.25023681e-01 1.25156891e+00 4.08686757e-01 5.94101369e-01
-7.59944022e-02 -1.02066167e-01 -1.31135195e-01 -3.07173103e-01
6.87198997e-01 -2.19709471e-01 -3.49881560e-01 5.15584946e-01
6.78392112e-01 -6.55252934e-01 8.03802609e-02 6.77510440e-01
7.09117889e-01 -4.16731894e-01 4.33270603e-01 -9.80080426e-01
-2.21571922e-01 -2.70926297e-01 -4.25287008e-01 -3.46996307e-01
1.12692010e+00 3.24487120e-01 5.58523118e-01 -1.06730330e+00
9.08430398e-01 1.00362265e+00 1.08353901e+00 4.61378306e-01
-1.18393409e+00 -3.35241050e-01 -4.39561903e-01 -2.69090742e-01
-1.45226359e+00 -5.54731309e-01 1.04234183e+00 -5.49640536e-01
6.94487870e-01 3.34538788e-01 1.02860320e+00 1.25510347e+00
1.25921041e-01 2.78393090e-01 1.47162688e+00 -3.63757253e-01
2.60159761e-01 -4.96073347e-03 -3.98391366e-01 7.45754361e-01
3.38846147e-01 3.19071025e-01 -7.99610019e-01 -1.87169641e-01
1.28814089e+00 1.46358430e-01 -3.54802579e-01 -1.06139350e+00
-1.46607745e+00 1.98579952e-01 4.04644340e-01 -1.22046344e-01
-6.15339994e-01 2.62056202e-01 -1.23440988e-01 -5.57822287e-02
4.05816257e-01 1.59214944e-01 -5.03392100e-01 -4.13421303e-01
-7.07412720e-01 3.24099027e-02 4.02869403e-01 1.26490021e+00
1.01134050e+00 -1.71258181e-01 4.96764153e-01 3.75477344e-01
7.97210991e-01 9.40093994e-01 2.15052679e-01 -1.37183857e+00
4.08751607e-01 7.72389710e-01 2.44801059e-01 -8.81493986e-01
-3.66746277e-01 -2.49619842e-01 -4.43716198e-01 5.51880300e-01
3.95465642e-01 5.08698225e-01 -8.25482070e-01 9.44665909e-01
7.82510400e-01 -9.60194543e-02 -1.05072677e-01 1.29545867e+00
8.92496049e-01 -1.47544384e-01 -7.11248815e-01 1.08208507e-01
1.11942327e+00 -2.59300083e-01 -3.57735604e-01 -1.63675621e-01
4.02283490e-01 -7.31289744e-01 1.37006915e+00 4.39381808e-01
-1.25590312e+00 -2.97739357e-01 -1.03091681e+00 -2.33428493e-01
3.97435650e-02 -2.73780763e-01 4.99518842e-01 8.00352752e-01
-9.05469000e-01 3.86582553e-01 -8.92163575e-01 -3.95830989e-01
5.13461530e-01 5.32855727e-02 -8.19246650e-01 -3.51861805e-01
-5.99896908e-01 8.81054938e-01 8.01381283e-03 1.42597511e-01
-8.26910317e-01 -6.87873721e-01 -8.10696661e-01 -7.26554632e-01
2.92744219e-01 -9.85385418e-01 1.31134939e+00 -4.28667426e-01
-1.42548060e+00 1.61971712e+00 1.34100690e-01 1.25535473e-01
8.26911509e-01 -8.23156759e-02 -1.11703190e-03 2.89149046e-01
6.20802157e-02 1.94839224e-01 6.13061130e-01 -1.93018115e+00
-9.09198225e-02 -1.04640973e+00 1.14811555e-01 2.99860626e-01
4.15900171e-01 -4.78844672e-01 -3.91870081e-01 -7.25593939e-02
1.07956493e+00 -7.55311251e-01 -1.48057804e-01 6.12335205e-01
-3.14299613e-01 7.49345362e-01 5.56254327e-01 -5.22316277e-01
3.76194715e-01 -1.91385269e+00 -2.24973056e-02 2.09412009e-01
8.14860240e-02 -2.54055023e-01 2.56047457e-01 2.60937065e-01
3.70818108e-01 1.60392627e-01 -3.64347190e-01 -5.59128582e-01
4.76753488e-02 3.73190135e-01 2.04879139e-02 7.40763903e-01
-2.73571849e-01 8.35757017e-01 -8.64862859e-01 -4.88023072e-01
4.71463650e-01 7.65647173e-01 -2.96780974e-01 1.94317952e-01
-2.03707442e-01 7.16769874e-01 -4.80620712e-01 1.28882170e+00
8.44324768e-01 -1.95592493e-01 -8.23900625e-02 -4.45619851e-01
-7.82069117e-02 2.58865327e-01 -1.35284853e+00 2.52144194e+00
-4.56272125e-01 1.71206057e-01 2.78298020e-01 -3.41945112e-01
1.01882637e+00 2.88020670e-01 5.85513234e-01 -8.98119330e-01
9.80658233e-02 5.65297186e-01 -6.11311316e-01 -4.52626497e-01
5.96642554e-01 -4.76936370e-01 -3.05390488e-02 4.65383679e-01
-1.49530694e-01 -1.01784122e+00 -7.87951350e-01 -2.01800629e-01
8.15308332e-01 7.70628810e-01 1.84414551e-01 1.38570875e-01
9.32342932e-02 2.43107319e-01 1.23649530e-01 5.38236320e-01
7.81146437e-02 1.34024894e+00 -1.08783282e-01 -4.06847060e-01
-1.42253900e+00 -1.50029862e+00 -4.58517075e-01 -5.16310073e-02
3.80564958e-01 -1.52030244e-01 -3.31127018e-01 -2.47620940e-01
1.47054181e-01 2.45813683e-01 -5.36849618e-01 1.55036956e-01
-2.96815276e-01 -2.97403276e-01 3.88532311e-01 4.01895285e-01
5.27210176e-01 -4.98606354e-01 -1.30211473e+00 -3.76745760e-02
-1.09926738e-01 -1.23834908e+00 -5.36333099e-02 -1.23648504e-02
-1.49559617e+00 -1.27608693e+00 -5.83823740e-01 -1.85910031e-01
6.50482476e-01 3.99373651e-01 1.46518397e+00 3.08890697e-02
-1.00874141e-01 9.72460985e-01 -4.78845239e-01 -2.89900303e-01
-2.50512540e-01 -3.47211152e-01 9.12019908e-02 -2.26789087e-01
-9.97100696e-02 -8.57965946e-01 -6.62775278e-01 7.91301668e-01
-7.11183786e-01 3.15506756e-01 3.12959999e-01 2.62690008e-01
1.43873143e+00 -2.04363614e-01 -1.34092033e-01 -4.57024187e-01
1.09179029e-02 -1.57074392e-01 -6.99595571e-01 -7.58079067e-02
-2.91205645e-01 -3.30090404e-01 -7.90863037e-02 -1.79188341e-01
-8.25919330e-01 3.24250400e-01 -2.37922177e-01 -6.34761035e-01
-3.87970299e-01 2.36531124e-01 -3.32030058e-01 -8.90633762e-02
9.74201500e-01 2.65098155e-01 3.26492071e-01 -7.41437078e-01
2.45511353e-01 6.36917770e-01 5.53960741e-01 -3.97199124e-01
8.10960889e-01 1.05234325e+00 2.35376924e-01 -8.81862760e-01
-6.47148788e-01 -3.52714479e-01 -1.04608738e+00 -4.92174715e-01
5.11215210e-01 -1.16131628e+00 -6.13130689e-01 5.31193495e-01
-9.55892086e-01 -4.45808083e-01 -5.12976229e-01 5.79268992e-01
-1.07188857e+00 2.89730906e-01 -3.43773700e-02 -9.35187757e-01
-1.02469459e-01 -1.12272692e+00 1.74554646e+00 -4.73306254e-02
-1.53686941e-01 -6.62528753e-01 4.08559181e-02 8.12536299e-01
1.82224810e-02 9.49188113e-01 2.40845844e-01 4.96051937e-01
-9.95216489e-01 -3.96802187e-01 1.71491563e-01 2.52814754e-03
2.12626949e-01 -1.50614142e-01 -1.17152536e+00 6.95445687e-02
2.03555092e-01 -3.66924137e-01 -2.04448123e-03 2.67188698e-01
7.15668142e-01 3.18935275e-01 -6.29101321e-02 8.07568848e-01
1.68153775e+00 -1.77746668e-01 7.97467709e-01 5.07592261e-01
8.98441434e-01 4.61967140e-01 7.78646231e-01 5.24022520e-01
6.14066064e-01 1.07905757e+00 9.69602942e-01 1.51585951e-01
-3.88171762e-01 -4.99907613e-01 -1.54787898e-02 7.49132991e-01
-4.20083016e-01 1.30681098e-01 -1.22238636e+00 2.55494922e-01
-1.24232090e+00 -5.60277462e-01 -8.20979238e-01 2.47446752e+00
5.45777142e-01 2.08829734e-02 1.58492818e-01 4.79883254e-01
2.49947086e-01 -1.65958300e-01 -7.23960876e-01 1.19492337e-01
-2.85399705e-01 7.89936185e-02 5.74893594e-01 5.11473775e-01
-3.26962084e-01 4.24065143e-01 6.13304567e+00 1.82549670e-01
-9.50462043e-01 2.04722524e-01 -5.89872189e-02 -1.29797712e-01
-7.16751814e-01 5.27140759e-02 -5.74168921e-01 2.53794611e-01
6.74412787e-01 2.67549485e-01 2.89963067e-01 8.54928493e-01
3.20282966e-01 -6.40979409e-01 -8.54439676e-01 1.39541852e+00
8.46326873e-02 -1.25566697e+00 -2.51996964e-01 4.53438640e-01
6.60642982e-01 3.42631906e-01 -2.77659386e-01 -4.90366369e-01
-1.87530145e-01 -7.74820685e-01 1.30588222e+00 9.02422071e-01
1.22034514e+00 -4.25435305e-01 5.21708608e-01 6.58024490e-01
-9.49521780e-01 4.92496341e-01 -1.95234492e-01 -1.30827323e-01
4.79792088e-01 7.71717608e-01 -9.34678674e-01 8.01320553e-01
9.07859266e-01 5.94365418e-01 -6.23649597e-01 1.03312707e+00
-1.87190413e-01 4.91905995e-02 -6.18701398e-01 2.82857269e-01
-3.55272561e-01 -2.63957530e-01 4.52223390e-01 4.58915949e-01
6.07864916e-01 1.16631664e-01 -2.85871476e-01 8.11620295e-01
1.40860245e-01 -3.40375125e-01 -1.02395558e+00 2.43914306e-01
3.52124423e-01 9.05175924e-01 -9.62958157e-01 1.14960246e-01
-1.36725038e-01 1.04332733e+00 -1.04519561e-01 7.37628434e-03
-6.35403693e-01 2.20149606e-01 5.29550314e-01 7.32165277e-01
5.47378464e-03 -7.98608661e-01 -8.07603598e-01 -1.21811819e+00
4.88116384e-01 -4.26739633e-01 -1.60824776e-01 -1.58656096e+00
-1.12194121e+00 4.51873183e-01 6.33988157e-02 -1.60214484e+00
-2.05746144e-01 -4.88435030e-01 1.29795000e-01 8.21546197e-01
-1.25777686e+00 -1.14714384e+00 -8.70617270e-01 6.29685462e-01
1.81123197e-01 4.08117563e-01 1.10060072e+00 1.58537090e-01
1.96879953e-01 1.04972878e-02 -1.26924992e-01 -4.45950955e-01
2.14881539e-01 -9.27322924e-01 4.52446222e-01 3.55687112e-01
3.21814656e-01 1.84336379e-01 5.93041599e-01 -7.29396105e-01
-2.07129574e+00 -6.56861782e-01 3.93170953e-01 -1.19838822e+00
-9.85578150e-02 -4.72056359e-01 -6.55676961e-01 6.97760046e-01
-6.78701401e-01 3.47386688e-01 4.91279572e-01 -2.72038788e-01
-1.62475377e-01 1.09392822e-01 -1.71459365e+00 1.18799940e-01
1.76527333e+00 -8.73591661e-01 -5.29890001e-01 8.90867114e-02
4.14054900e-01 -1.14192891e+00 -1.07731164e+00 4.07605618e-01
1.05621159e+00 -1.47355938e+00 1.34652865e+00 1.58873856e-01
3.79213065e-01 -6.18650138e-01 -8.11303735e-01 -9.59370911e-01
3.00478250e-01 -2.49464497e-01 -2.64849484e-01 8.26577067e-01
9.08234194e-02 -4.74091500e-01 1.05844879e+00 6.85240567e-01
-2.66768873e-01 -6.25208795e-01 -1.13681734e+00 -8.88261318e-01
-4.50710505e-01 -1.10259485e+00 7.93582380e-01 7.56870270e-01
-6.39343977e-01 -1.79704681e-01 -1.38087021e-02 3.50549281e-01
9.54772651e-01 3.97707492e-01 1.19240928e+00 -1.32283020e+00
-4.84770574e-02 6.98416755e-02 -8.00828934e-01 -1.07152832e+00
-3.38619500e-01 -8.91989827e-01 -1.07286997e-01 -1.70267856e+00
-2.84458995e-01 -8.11985731e-01 6.13117218e-01 8.73669889e-03
6.77311063e-01 9.48962748e-01 -1.56259120e-01 4.15969044e-01
-3.01532060e-01 6.29240870e-01 1.45279765e+00 4.32704568e-01
-2.55268365e-02 -4.53690328e-02 -3.76651615e-01 9.18931186e-01
5.75240850e-01 -4.68705952e-01 -3.05125386e-01 -7.91905582e-01
4.62539256e-01 2.73246169e-01 8.00704122e-01 -9.74516690e-01
-9.77815688e-02 -5.46400957e-02 5.93330801e-01 -1.00213182e+00
9.77520823e-01 -1.38080156e+00 1.01220942e+00 2.87025362e-01
3.35760266e-01 -2.08310410e-02 2.67221108e-02 4.15843040e-01
3.11611503e-01 -5.18555827e-02 6.25029385e-01 -6.09516621e-01
-5.48724055e-01 4.39019412e-01 1.19113915e-01 -9.97730196e-02
8.31331730e-01 -1.19552159e+00 3.03919166e-01 -1.21172808e-01
-6.24730170e-01 -3.31339478e-01 1.58504081e+00 1.92814127e-01
1.12527287e+00 -1.68182766e+00 -3.28117788e-01 5.88559747e-01
3.97221029e-01 7.78492868e-01 2.06383243e-01 5.84546983e-01
-9.40715909e-01 2.14057602e-02 -3.60849589e-01 -1.22415841e+00
-9.66673791e-01 1.33491466e-02 5.73455334e-01 4.97315139e-01
-9.18572485e-01 4.72322315e-01 -4.53648567e-01 -9.04986680e-01
-1.76879138e-01 -5.35238504e-01 5.22487342e-01 -1.57901302e-01
2.42971331e-01 3.03072333e-01 5.36036015e-01 -9.61063266e-01
-5.10020018e-01 1.16547656e+00 9.01746273e-01 -4.24783856e-01
1.50332057e+00 -4.97949421e-01 2.58279562e-01 7.94941962e-01
9.04042542e-01 3.27391773e-02 -1.47820365e+00 -1.12936310e-01
-2.33329415e-01 -1.06513369e+00 1.89493388e-01 -7.55485415e-01
-8.51940155e-01 7.38224268e-01 7.48700619e-01 -1.02016270e-01
9.50649142e-01 3.27514708e-01 4.90713626e-01 8.01119506e-02
1.38842177e+00 -6.13386452e-01 -1.95551571e-03 1.89579144e-01
1.18685246e+00 -1.34051681e+00 2.17827410e-01 -4.26061392e-01
-2.43841663e-01 9.37672496e-01 2.71926016e-01 -2.16673817e-02
4.87658471e-01 1.26485556e-01 1.58223540e-01 -6.94816887e-01
-1.94440205e-02 -1.82166100e-01 9.84084010e-02 1.08237505e+00
1.45378426e-01 5.30767255e-03 2.16648862e-01 1.94828063e-01
-4.28103238e-01 8.58353004e-02 6.46136045e-01 1.19831514e+00
-7.06280768e-02 -9.33427930e-01 -8.32269728e-01 2.03340396e-01
-5.08736782e-02 3.60418290e-01 -3.62444639e-01 7.54757464e-01
6.59523681e-02 6.53424323e-01 1.08073182e-01 -4.84488457e-01
8.98696899e-01 -1.24218687e-01 1.11324751e+00 -6.24022305e-01
-1.55434534e-01 -1.96939602e-01 5.81614077e-02 -1.02440286e+00
-7.16100633e-01 -7.64108777e-01 -1.10646999e+00 -2.48219460e-01
-2.00268492e-01 -4.62195635e-01 1.39287233e+00 5.31622946e-01
4.22035724e-01 -5.14627136e-02 5.57536602e-01 -1.53114390e+00
-1.26780435e-01 -6.70421898e-01 -7.81065643e-01 4.11918491e-01
3.32514077e-01 -9.54386771e-01 -3.78824502e-01 1.54910505e-01] | [8.494458198547363, -2.8441519737243652] |
fb036535-c278-469e-8703-5b527e2f867a | a-neural-based-program-decompiler | 1906.12029 | null | https://arxiv.org/abs/1906.12029v1 | https://arxiv.org/pdf/1906.12029v1.pdf | A Neural-based Program Decompiler | Reverse engineering of binary executables is a critical problem in the computer security domain. On the one hand, malicious parties may recover interpretable source codes from the software products to gain commercial advantages. On the other hand, binary decompilation can be leveraged for code vulnerability analysis and malware detection. However, efficient binary decompilation is challenging. Conventional decompilers have the following major limitations: (i) they are only applicable to specific source-target language pair, hence incurs undesired development cost for new language tasks; (ii) their output high-level code cannot effectively preserve the correct functionality of the input binary; (iii) their output program does not capture the semantics of the input and the reversed program is hard to interpret. To address the above problems, we propose Coda, the first end-to-end neural-based framework for code decompilation. Coda decomposes the decompilation task into two key phases: First, Coda employs an instruction type-aware encoder and a tree decoder for generating an abstract syntax tree (AST) with attention feeding during the code sketch generation stage. Second, Coda then updates the code sketch using an iterative error correction machine guided by an ensembled neural error predictor. By finding a good approximate candidate and then fixing it towards perfect, Coda achieves superior performance compared to baseline approaches. We assess Coda's performance with extensive experiments on various benchmarks. Evaluation results show that Coda achieves an average of 82\% program recovery accuracy on unseen binary samples, where the state-of-the-art decompilers yield 0\% accuracy. Furthermore, Coda outperforms the sequence-to-sequence model with attention by a margin of 70\% program accuracy. | ['Yuandong Tian', 'Haolan Liu', 'Huili Chen', 'Farinaz Koushanfar', 'Xinyun Chen', 'Jishen Zhao', 'Cheng Fu'] | 2019-06-28 | null | null | null | null | ['computer-security'] | ['miscellaneous'] | [ 4.64709640e-01 -1.40018553e-01 -7.67703950e-01 1.78823676e-02
-7.25173414e-01 -7.41250813e-01 2.40232214e-01 1.09976642e-01
6.59243986e-02 2.10985988e-01 -1.26451284e-01 -1.22303557e+00
5.12428939e-01 -7.07948327e-01 -9.57225978e-01 -1.56891868e-01
1.87808663e-01 7.52522498e-02 2.94937998e-01 -1.56172901e-01
6.35479569e-01 2.45506138e-01 -1.27201736e+00 4.09292400e-01
1.03156090e+00 8.54814589e-01 1.11562602e-01 9.04922247e-01
-1.20334275e-01 8.64933014e-01 -6.80071592e-01 -7.45503485e-01
2.76977301e-01 -2.51410931e-01 -8.39765191e-01 -2.00687274e-01
1.67339861e-01 -6.77071333e-01 -4.86190438e-01 1.57645679e+00
-1.34820685e-01 -6.30779922e-01 5.25055766e-01 -1.29453969e+00
-8.97809029e-01 8.29000831e-01 -7.80822217e-01 2.26412520e-01
1.57608196e-01 5.34733713e-01 9.71034825e-01 -7.26908326e-01
3.50521773e-01 1.06103027e+00 5.88797390e-01 6.71420515e-01
-1.28087044e+00 -7.23651767e-01 -2.31732745e-02 1.03130728e-01
-1.13593197e+00 -3.14572781e-01 7.35540569e-01 -5.17859519e-01
1.38735962e+00 2.94933885e-01 1.94896951e-01 1.11986804e+00
5.38078368e-01 6.82616293e-01 8.02194059e-01 -1.67548478e-01
2.15059623e-01 1.26735941e-01 5.30076623e-01 9.62423384e-01
4.43604976e-01 3.92118663e-01 4.76494879e-02 -4.21721280e-01
5.23266755e-02 2.15333939e-01 -3.61163408e-01 4.16201428e-02
-8.26951265e-01 8.18985164e-01 4.75789666e-01 1.23061806e-01
3.42509784e-02 2.88582116e-01 8.55185091e-01 4.82740432e-01
1.04292817e-02 5.99077404e-01 -4.06706214e-01 -3.79671931e-01
-1.05291474e+00 1.33423790e-01 8.69521856e-01 8.22887063e-01
5.70806503e-01 4.47957903e-01 8.55313763e-02 2.94033498e-01
4.18063939e-01 4.58894730e-01 7.37760901e-01 -4.49556917e-01
8.93834710e-01 8.60228300e-01 -4.13698494e-01 -1.12532914e+00
1.72484487e-01 -3.35671484e-01 -7.78221488e-01 4.90236640e-01
1.44540370e-01 1.30515531e-01 -8.42529893e-01 1.40689135e+00
-1.49705827e-01 4.83435392e-02 5.72280437e-02 4.46870804e-01
4.36596274e-01 7.00243413e-01 -2.34660760e-01 -8.08827728e-02
1.40793467e+00 -1.05953836e+00 -3.50926578e-01 -6.70972764e-01
6.69732809e-01 -4.93165582e-01 1.10061657e+00 3.17632884e-01
-7.32164383e-01 -4.75260228e-01 -1.55215943e+00 7.07028657e-02
-1.77071527e-01 1.42206565e-01 3.55482072e-01 9.53208208e-01
-8.83115411e-01 5.08822203e-01 -9.60164130e-01 2.89938897e-01
5.18211365e-01 3.68825555e-01 -1.84339777e-01 1.78486537e-02
-7.00190246e-01 3.66898358e-01 5.83811760e-01 -7.60380551e-02
-1.31268156e+00 -7.05038846e-01 -1.10068464e+00 3.16867352e-01
3.91457677e-01 -3.33879739e-01 1.24649155e+00 -1.08142710e+00
-1.36430061e+00 6.02898061e-01 -3.03567648e-01 -7.00055361e-01
2.43565336e-01 -1.24439456e-01 -5.04039645e-01 -1.47480622e-01
1.06481593e-02 9.50178783e-03 1.26312494e+00 -1.16805625e+00
-4.78201509e-01 -2.37084955e-01 7.65861571e-02 -5.06052017e-01
-4.93833363e-01 1.60836920e-01 -2.62678593e-01 -8.05557966e-01
-3.38671595e-01 -9.85717773e-01 -2.00281627e-02 -2.81782895e-01
-7.17565596e-01 1.67170390e-01 1.19426465e+00 -1.07551241e+00
1.86234975e+00 -2.35113788e+00 1.87185034e-01 1.16804622e-01
5.19048631e-01 7.41483510e-01 -7.98817053e-02 9.29937363e-02
-4.80227649e-01 4.99225259e-01 -6.58160329e-01 -1.71937525e-01
7.16766436e-03 -1.76174402e-01 -1.02425385e+00 3.24315041e-01
2.69125342e-01 1.19382048e+00 -7.85993934e-01 -5.43491803e-02
-2.05347151e-01 2.61039380e-02 -8.51545334e-01 3.45325202e-01
-5.07532299e-01 -1.14765786e-01 -2.77289808e-01 8.86056304e-01
6.29073858e-01 -4.15236443e-01 1.97288409e-01 -5.88840693e-02
1.08256556e-01 4.64271456e-01 -5.43552816e-01 1.12093186e+00
-5.79533458e-01 8.91731501e-01 -1.95111051e-01 -8.28357697e-01
8.17408800e-01 1.78625174e-02 -2.48577446e-01 -3.79601598e-01
1.77111402e-01 4.19258118e-01 2.30630055e-01 -4.39777821e-01
6.49117887e-01 9.52175111e-02 -3.44181895e-01 8.00388753e-01
-3.89832973e-01 -2.87203304e-02 -1.47880062e-01 2.34552145e-01
1.45900488e+00 -4.27858680e-02 5.05693197e-01 6.58188313e-02
6.39493227e-01 1.00554273e-01 4.95139420e-01 5.92438936e-01
-1.46371141e-01 2.53274769e-01 9.27080750e-01 -4.09934700e-01
-1.08025026e+00 -7.41626918e-01 2.25982860e-01 7.54281223e-01
6.37708139e-03 -6.36650503e-01 -1.09092033e+00 -1.21501148e+00
-1.21512957e-01 1.04996979e+00 -5.05107045e-01 -6.39403224e-01
-9.43803072e-01 -6.60006464e-01 1.00222254e+00 7.38842785e-01
5.38359046e-01 -8.71203423e-01 -5.78463256e-01 2.61152610e-02
-1.69367537e-01 -8.77816916e-01 -8.96211267e-01 1.57062605e-01
-8.17921877e-01 -1.19321561e+00 -1.82899818e-01 -7.18403101e-01
8.63915026e-01 2.25560039e-01 8.92504454e-01 5.81084251e-01
-1.28823102e-01 -3.18640977e-01 -2.49807417e-01 3.43029201e-02
-1.12400830e+00 1.65934622e-01 -1.47160158e-01 -2.45152339e-01
4.39759344e-01 -4.15157497e-01 -1.65132597e-01 1.11162141e-01
-1.13059759e+00 -8.28206241e-02 6.18003428e-01 1.08071029e+00
3.10277045e-01 3.89475733e-01 2.08749190e-01 -9.15283740e-01
5.41829288e-01 -5.07649601e-01 -9.66713309e-01 1.79966360e-01
-7.64576495e-01 3.88003081e-01 1.25528312e+00 -4.92774785e-01
-7.94896364e-01 -1.25547796e-01 -3.64383101e-01 -6.48299873e-01
1.89685166e-01 4.88006413e-01 -4.14796054e-01 4.33258028e-05
8.10754240e-01 5.47721744e-01 1.13801889e-01 -2.23444164e-01
6.31017163e-02 9.04899061e-01 7.56724298e-01 -5.23736835e-01
1.25052094e+00 1.00691728e-01 -2.02194497e-01 -4.14768726e-01
-3.52989078e-01 1.54764116e-01 -3.99249762e-01 3.19576621e-01
7.03699172e-01 -5.69001555e-01 -6.98167980e-01 7.46546328e-01
-1.35092580e+00 -3.90928715e-01 1.08429648e-01 -1.78033873e-01
-2.36055940e-01 9.19281542e-01 -8.49996209e-01 -4.70410705e-01
-5.63387036e-01 -1.96367800e+00 1.00547743e+00 -1.28390249e-02
-3.14250380e-01 -6.75860167e-01 -2.48494521e-01 3.63860339e-01
2.85006106e-01 8.56620148e-02 1.64996290e+00 -7.40149856e-01
-8.20517480e-01 -3.73876393e-01 -2.71407485e-01 5.88641346e-01
7.00088590e-02 8.39297548e-02 -7.65488267e-01 -5.21369994e-01
1.74403161e-01 -2.20940128e-01 6.91239595e-01 -2.68425971e-01
1.40554929e+00 -7.18528032e-01 -4.17613298e-01 8.98205698e-01
1.35535812e+00 4.18772250e-01 7.12869227e-01 2.29781285e-01
8.09429705e-01 1.13997675e-01 3.77599895e-01 1.53499156e-01
1.58793196e-01 5.80378711e-01 8.50001931e-01 4.68159378e-01
-1.29961342e-01 -6.24872088e-01 9.09756064e-01 8.85949850e-01
4.76193458e-01 -1.46060526e-01 -1.09521854e+00 4.37753558e-01
-1.40658438e+00 -7.66632080e-01 -8.85653943e-02 2.31992006e+00
8.37545991e-01 4.95002091e-01 -2.46833880e-02 3.57097894e-01
6.39964759e-01 1.77501112e-01 -7.65758514e-01 -7.88844049e-01
3.62634033e-01 2.55097121e-01 7.24162519e-01 4.52759653e-01
-8.94322991e-01 8.52630079e-01 5.54916286e+00 9.87750649e-01
-1.32158422e+00 1.10791996e-01 6.54446721e-01 3.47819924e-01
-5.53798378e-01 3.55365723e-01 -9.88219082e-01 7.76760697e-01
1.23215723e+00 -3.05026233e-01 8.29817712e-01 1.26279020e+00
-4.56756413e-01 3.75254035e-01 -1.17845833e+00 7.26944029e-01
2.31854022e-01 -1.25518847e+00 -7.63623714e-02 1.87407404e-01
4.75204259e-01 -8.26885924e-02 2.88263202e-01 5.97658277e-01
3.82410824e-01 -1.13846862e+00 9.37172472e-01 -2.18897983e-02
1.16714346e+00 -8.04270566e-01 6.36395216e-01 4.60475564e-01
-1.23574126e+00 -4.86019164e-01 -1.92487091e-01 1.87332198e-01
-7.17675760e-02 2.94036865e-01 -9.27754462e-01 3.02397639e-01
4.14478123e-01 5.98923385e-01 -8.43243837e-01 5.67130089e-01
-4.96214747e-01 8.64868164e-01 1.54137626e-01 -8.16141963e-02
1.61077157e-01 2.05845207e-01 7.28970885e-01 1.17898607e+00
3.93251598e-01 -2.58814156e-01 3.14310454e-02 1.20262480e+00
-3.44958454e-01 -4.35683161e-01 -6.12502456e-01 -5.33891261e-01
4.58537102e-01 9.03428435e-01 -6.37317836e-01 -2.65297681e-01
-4.00749981e-01 1.05927801e+00 3.32361668e-01 2.15448990e-01
-1.13321650e+00 -6.99915826e-01 7.65488267e-01 6.88836798e-02
4.66449142e-01 -1.46732286e-01 -5.92457712e-01 -1.35676813e+00
1.37076065e-01 -1.56064296e+00 2.22415492e-01 -4.79477972e-01
-7.32310116e-01 9.58984137e-01 -2.22618327e-01 -1.16521633e+00
-4.54736263e-01 -8.04218233e-01 -6.69192672e-01 8.12352419e-01
-1.28479040e+00 -9.49134648e-01 -7.45744482e-02 -3.26937772e-02
7.83432007e-01 -4.86996561e-01 6.58653975e-01 1.47779241e-01
-8.58388901e-01 1.10199046e+00 -4.67767827e-02 4.45728689e-01
1.28032506e-01 -1.06574070e+00 1.17638385e+00 1.47519267e+00
-2.28736058e-01 9.29086924e-01 4.92457032e-01 -1.06848645e+00
-1.87446821e+00 -1.30956614e+00 7.41395175e-01 -5.23211896e-01
1.01814234e+00 -4.73377794e-01 -1.14624214e+00 8.82844508e-01
-2.81775575e-02 -6.29456714e-02 4.58368391e-01 -4.86934304e-01
-9.90904748e-01 1.14328079e-01 -9.73351836e-01 6.97379410e-01
7.18694925e-01 -8.79548132e-01 -4.94826525e-01 9.83203650e-02
1.11141849e+00 -4.79202718e-01 -5.22333682e-01 1.33849263e-01
4.44409370e-01 -9.11179006e-01 8.46627653e-01 -6.75337255e-01
1.08259475e+00 -4.90771979e-01 -2.73077399e-01 -9.85142648e-01
-2.05073357e-01 -7.31358469e-01 -6.86492503e-01 1.14250183e+00
4.49109733e-01 -6.86365545e-01 6.46028399e-01 4.07247812e-01
-1.87500536e-01 -8.37139904e-01 -7.04372168e-01 -8.08690488e-01
1.33794174e-01 -6.55496359e-01 8.84188235e-01 6.57340467e-01
1.08491806e-02 2.53259748e-01 -3.56755376e-01 2.85118371e-01
5.50346136e-01 3.70875627e-01 7.79901683e-01 -5.71689367e-01
-9.89835322e-01 -6.66313171e-01 -2.88502097e-01 -1.22269809e+00
6.80608630e-01 -1.19435203e+00 3.68191563e-02 -6.96386576e-01
3.69908065e-01 -2.57092029e-01 6.31274357e-02 6.61766291e-01
-3.85232091e-01 7.63500854e-02 2.16881841e-01 2.70484716e-01
-4.27100882e-02 2.57709950e-01 5.41327059e-01 -5.69104016e-01
5.40554374e-02 5.55618517e-02 -9.79396641e-01 6.89528048e-01
6.69197857e-01 -6.93963051e-01 -3.48303914e-01 -4.63988602e-01
3.18377018e-01 1.99739113e-01 4.58125204e-01 -8.22581053e-01
1.85208935e-02 -1.85055673e-01 -2.01967582e-01 -4.05689925e-01
-8.78146738e-02 -7.63309240e-01 1.31680235e-01 1.03393757e+00
-1.69716567e-01 3.85622889e-01 2.68372744e-01 6.29416406e-01
-1.08517520e-01 -7.70440280e-01 8.64977896e-01 7.40627050e-02
-6.64214134e-01 2.32733727e-01 -3.49868655e-01 8.28703642e-02
1.00474012e+00 -2.42169946e-01 -5.64284027e-01 2.98065059e-02
3.61440033e-02 -2.28873059e-01 8.95875514e-01 6.02439642e-01
6.86774015e-01 -1.01258171e+00 -3.90361607e-01 6.05790913e-01
1.30272567e-01 -3.13727945e-01 -9.03622359e-02 5.73357105e-01
-7.57155478e-01 4.27979290e-01 3.22312191e-02 -3.06174129e-01
-1.53561163e+00 1.12854576e+00 2.82547563e-01 -5.65060735e-01
-4.57452178e-01 6.06166124e-01 2.94716865e-01 -3.17323923e-01
-7.65250102e-02 -4.37504411e-01 1.98401228e-01 -6.00495398e-01
8.61066818e-01 2.05364317e-01 1.05470635e-01 -6.73826277e-01
-3.58202249e-01 1.84059396e-01 -4.37308282e-01 3.74448955e-01
1.14174306e+00 3.33809495e-01 -4.33227807e-01 -1.13469623e-01
1.48877096e+00 2.33176529e-01 -9.84065950e-01 -7.66726211e-02
2.95213815e-02 -6.44996941e-01 2.35010739e-02 -6.57185972e-01
-1.07936442e+00 1.10471547e+00 1.96709216e-01 8.32587257e-02
1.26923776e+00 -3.32171321e-01 1.22999573e+00 2.91118264e-01
3.52264643e-01 -3.87898177e-01 2.58296013e-01 6.60147429e-01
5.27897656e-01 -1.01642215e+00 -1.53025702e-01 -3.89379948e-01
-4.04799491e-01 1.29926741e+00 7.38260984e-01 5.81335695e-03
3.39408129e-01 7.58622706e-01 -4.10211921e-01 1.11455873e-01
-6.74322546e-01 4.35945302e-01 2.33741894e-01 4.72231418e-01
4.69416156e-02 5.13593927e-02 2.43666936e-02 9.06240046e-01
-2.44408652e-01 -1.60833597e-01 7.37531900e-01 1.00037622e+00
-2.32203379e-01 -1.24049723e+00 -4.61400986e-01 5.83197176e-01
-6.94889128e-01 -4.53005344e-01 -2.78176218e-01 4.46427375e-01
-1.83494523e-01 7.50386178e-01 -3.41521829e-01 -9.25880492e-01
1.87226420e-03 -3.55387270e-03 2.27755792e-02 -7.05571532e-01
-8.26961339e-01 -3.92232716e-01 -1.60100147e-01 -4.96564895e-01
5.76801538e-01 -4.50574368e-01 -1.17197835e+00 -5.83858073e-01
-2.10855782e-01 -1.52888864e-01 5.05748332e-01 7.51099527e-01
6.29304767e-01 7.14286745e-01 7.07201958e-01 -5.59540689e-01
-1.03262579e+00 -5.34757793e-01 -1.26374653e-02 2.41377443e-01
5.89007437e-01 -2.67000794e-01 -4.54489529e-01 2.98639596e-01] | [7.110142707824707, 7.816238880157471] |
6da2c0f5-1f61-40c5-b352-dda6462c2480 | joint-dimensionality-reduction-for-separable | 2101.05500 | null | https://arxiv.org/abs/2101.05500v1 | https://arxiv.org/pdf/2101.05500v1.pdf | Joint Dimensionality Reduction for Separable Embedding Estimation | Low-dimensional embeddings for data from disparate sources play critical roles in multi-modal machine learning, multimedia information retrieval, and bioinformatics. In this paper, we propose a supervised dimensionality reduction method that learns linear embeddings jointly for two feature vectors representing data of different modalities or data from distinct types of entities. We also propose an efficient feature selection method that complements, and can be applied prior to, our joint dimensionality reduction method. Assuming that there exist true linear embeddings for these features, our analysis of the error in the learned linear embeddings provides theoretical guarantees that the dimensionality reduction method accurately estimates the true embeddings when certain technical conditions are satisfied and the number of samples is sufficiently large. The derived sample complexity results are echoed by numerical experiments. We apply the proposed dimensionality reduction method to gene-disease association, and predict unknown associations using kernel regression on the dimension-reduced feature vectors. Our approach compares favorably against other dimensionality reduction methods, and against a state-of-the-art method of bilinear regression for predicting gene-disease associations. | ['Yoram Bresler', 'Hao Cheng', 'Bihan Wen', 'Yanjun Li'] | 2021-01-14 | null | null | null | null | ['supervised-dimensionality-reduction'] | ['computer-vision'] | [-7.37168267e-02 -1.33778289e-01 -3.70477855e-01 -2.63696790e-01
-9.72836196e-01 -4.90569264e-01 4.71508473e-01 4.63216364e-01
-5.25550961e-01 5.42008460e-01 3.80298048e-01 -1.32951856e-01
-5.97196579e-01 -6.23635054e-01 -3.39807004e-01 -7.90580392e-01
-4.48804647e-01 5.03810346e-01 -6.77705407e-02 1.56504288e-01
-1.98124051e-02 4.98618573e-01 -1.51409650e+00 -1.41694427e-01
6.00053489e-01 7.51110494e-01 -2.90271640e-01 8.24205399e-01
-8.55790004e-02 9.28443894e-02 -7.78212771e-02 -1.88099608e-01
5.10741957e-02 -1.04191639e-01 -5.38200736e-01 -9.99803990e-02
3.89025062e-01 -1.81992024e-01 -6.30828977e-01 8.68406415e-01
7.29125917e-01 2.42562760e-02 9.90151405e-01 -1.42372274e+00
-1.02394247e+00 4.58661802e-02 -5.42764604e-01 1.98274255e-01
3.05754215e-01 -5.67900538e-01 1.10334408e+00 -1.41102850e+00
8.69225204e-01 1.20761406e+00 7.16549397e-01 6.21289611e-01
-1.31706285e+00 -2.71942019e-01 -1.01778142e-01 4.89522964e-02
-1.62316155e+00 -3.94614130e-01 5.84993362e-01 -6.67171240e-01
5.93258202e-01 3.18624109e-01 4.19809997e-01 7.62617886e-01
8.83182958e-02 4.11592573e-01 5.51816225e-01 -6.25575185e-01
4.12475228e-01 3.86783868e-01 2.84796089e-01 9.07671154e-01
5.80065668e-01 -9.03691202e-02 -5.36422610e-01 -8.87536824e-01
2.00550601e-01 4.82363880e-01 -1.71512976e-01 -8.50215793e-01
-1.32952070e+00 1.05004442e+00 -5.98924570e-02 3.96550387e-01
-3.39197338e-01 1.40558988e-01 4.45190072e-01 2.86175400e-01
6.38034046e-01 2.01401889e-01 -5.75882494e-01 4.59115021e-02
-5.38863242e-01 1.86354518e-01 7.28987157e-01 8.75895202e-01
5.79212427e-01 -3.33038062e-01 2.00528741e-01 9.63429570e-01
3.82710218e-01 4.18386221e-01 7.10734785e-01 -5.59477150e-01
2.08803818e-01 6.03064597e-01 1.45317078e-01 -1.17962003e+00
-5.86352944e-01 1.35891780e-01 -7.55387068e-01 -2.46620774e-01
3.55183870e-01 -6.03374466e-02 -5.06408036e-01 1.74900711e+00
7.61337340e-01 2.98142850e-01 3.02033812e-01 8.19405556e-01
6.68866634e-01 4.92556244e-01 -2.04194989e-02 -1.96815893e-01
1.64836955e+00 -3.64608556e-01 -7.90282607e-01 3.90283078e-01
1.14286029e+00 -4.11869496e-01 7.08155930e-01 2.57832140e-01
-6.55704677e-01 -1.28537804e-01 -8.83032203e-01 -1.86727598e-01
-7.01455891e-01 2.58839071e-01 8.65001798e-01 8.02622736e-01
-7.93780863e-01 3.31387132e-01 -6.11442089e-01 -6.30678058e-01
4.21837717e-01 4.70929265e-01 -8.38823497e-01 -4.84461457e-01
-1.19088864e+00 5.90268016e-01 1.11678876e-01 -1.58596143e-01
-4.45483804e-01 -8.85502636e-01 -9.51723099e-01 -1.58704296e-01
-9.62516069e-02 -5.67985117e-01 5.03300667e-01 -3.33180636e-01
-9.20495570e-01 9.62031543e-01 -3.61134201e-01 -1.37366325e-01
6.69572502e-02 -3.10879350e-01 -8.45163882e-01 3.15263718e-01
8.15334842e-02 3.62769872e-01 8.19504082e-01 -7.79250681e-01
-5.35417676e-01 -6.86916590e-01 -3.48594218e-01 -3.90711203e-02
-1.02212107e+00 -1.55123949e-01 -3.61869007e-01 -4.96118039e-01
1.11652300e-01 -9.89648104e-01 -3.00620884e-01 5.58924377e-01
-2.75952965e-01 -3.34850729e-01 8.65431428e-01 -4.24743593e-01
1.21844184e+00 -2.46266627e+00 5.05662441e-01 3.68163317e-01
5.23489952e-01 -6.31055310e-02 -4.21512425e-01 4.63350534e-01
-3.87269169e-01 2.37422347e-01 -3.36130075e-02 -1.54696509e-01
1.30630257e-02 1.11769952e-01 -1.07698292e-01 9.92450595e-01
4.18826878e-01 5.62315524e-01 -8.70702505e-01 -4.56179559e-01
-5.38528748e-02 8.11885774e-01 -6.34436131e-01 1.07083812e-01
2.43727207e-01 -1.54874235e-01 -6.42391086e-01 7.86065936e-01
5.20415723e-01 -3.11714172e-01 4.51872826e-01 -4.81281877e-01
1.73520580e-01 -2.57317305e-01 -1.21919405e+00 1.73893821e+00
-2.61000067e-01 5.98386168e-01 -2.40236610e-01 -1.16074538e+00
7.41751671e-01 4.22482163e-01 8.94271731e-01 -1.76682115e-01
-6.12313636e-02 2.08194658e-01 -2.02586263e-01 -7.38953352e-01
2.55253792e-01 -1.42535359e-01 -2.47564822e-01 3.31079870e-01
2.67875195e-01 4.36026931e-01 1.07316777e-01 2.74855793e-01
1.23661053e+00 -4.23026413e-01 4.10411805e-01 -2.14862123e-01
4.63635653e-01 -1.04710720e-01 5.69680870e-01 3.76571059e-01
-1.75624594e-01 3.90731305e-01 7.42193878e-01 -4.51921254e-01
-1.14734387e+00 -9.65220273e-01 -5.07524312e-01 9.97168541e-01
-2.60066111e-02 -5.29706359e-01 -3.30411524e-01 -7.71242738e-01
4.56960201e-01 1.21359684e-01 -8.99910331e-01 -2.40356475e-01
-3.01625896e-02 -1.08456075e+00 4.91596520e-01 3.34345907e-01
-3.99377912e-01 -3.67760882e-02 1.94375347e-02 1.95741877e-02
3.43191266e-01 -9.20870602e-01 -2.74190664e-01 1.51025414e-01
-8.72801423e-01 -1.39013994e+00 -8.00511360e-01 -8.73582125e-01
9.34110940e-01 2.33389303e-01 5.80659151e-01 -1.25557959e-01
-7.18615949e-01 8.10931206e-01 -2.92115211e-01 -9.83543023e-02
-2.12444410e-01 -8.43684375e-02 7.47675896e-01 9.41291898e-02
8.89746964e-01 -3.78261864e-01 -4.86746311e-01 1.52026013e-01
-1.22120261e+00 -5.30111492e-01 5.14190137e-01 1.09878314e+00
7.17120886e-01 1.32346287e-01 8.13673198e-01 -9.05172527e-01
6.23720586e-01 -9.20392811e-01 -4.14369315e-01 2.00012282e-01
-6.41363204e-01 4.01712388e-01 4.12705272e-01 -6.96432292e-01
-4.02309746e-01 2.35270560e-01 2.51237541e-01 -4.99413431e-01
-2.16890693e-01 5.88044047e-01 -2.35478267e-01 -1.24184027e-01
5.53653002e-01 1.61524210e-02 2.33752087e-01 -5.85018694e-01
7.63359070e-01 8.43136072e-01 3.53186019e-02 -3.18972141e-01
8.70752454e-01 6.72589600e-01 2.62241691e-01 -1.11816561e+00
-3.57093006e-01 -5.82495809e-01 -6.92362964e-01 3.34872723e-01
6.69665456e-01 -9.74180520e-01 -5.89698017e-01 -3.90243763e-03
-9.88459468e-01 5.29506445e-01 -1.35581508e-01 8.18909228e-01
-4.03204173e-01 4.18904305e-01 -4.15801346e-01 -7.17608690e-01
2.13161390e-02 -7.02363968e-01 1.02752221e+00 -1.75757214e-01
-2.97298938e-01 -1.26692581e+00 5.73097467e-01 -5.80448397e-02
1.32328749e-01 2.00038612e-01 1.36304688e+00 -1.02358341e+00
-1.81254558e-02 -7.71197915e-01 -2.28851542e-01 3.13059427e-02
5.17492890e-01 -2.47421004e-02 -8.36332381e-01 -1.46581382e-01
-4.17633206e-01 -1.61596924e-01 8.14819455e-01 3.71960402e-01
1.12178624e+00 -2.73036838e-01 -7.04161286e-01 5.00837922e-01
1.43449533e+00 -1.95983812e-01 2.47509480e-01 4.78443764e-02
7.55853891e-01 7.28047192e-01 5.17905235e-01 7.31706202e-01
2.67594278e-01 6.30872250e-01 3.45403627e-02 -2.08455101e-02
3.13089132e-01 -5.61919361e-02 1.86002672e-01 9.78697240e-01
3.88908565e-01 -1.07351691e-01 -7.30992377e-01 8.65556717e-01
-1.80577922e+00 -7.38974333e-01 -1.39368385e-01 2.33032131e+00
7.87214398e-01 -4.27144617e-01 3.23926717e-01 1.10139988e-01
7.07619071e-01 -1.25024274e-01 -5.30178249e-01 -3.42457831e-01
-1.63654581e-01 -2.90445052e-02 4.52030510e-01 3.98275763e-01
-1.17552662e+00 4.32348937e-01 7.06126022e+00 5.48160732e-01
-7.22064853e-01 1.49134204e-01 2.63060629e-01 -2.76517153e-01
-4.95223105e-01 -3.05927217e-01 -7.98012018e-01 3.28694671e-01
1.16356945e+00 -3.00182521e-01 1.49102375e-01 8.44622433e-01
5.49442880e-02 1.59904122e-01 -1.36351359e+00 1.11966729e+00
2.31693521e-01 -1.38301337e+00 4.36297990e-02 2.49886870e-01
5.09975672e-01 -2.09512398e-01 3.74849081e-01 -3.83460559e-02
6.88535022e-03 -8.51031542e-01 -2.29370028e-01 5.95344186e-01
1.01983476e+00 -7.22621858e-01 5.59875846e-01 -4.44040038e-02
-1.11174023e+00 -1.48438960e-01 -5.31356514e-01 3.20640266e-01
-8.35971385e-02 1.05856872e+00 -7.79756248e-01 3.54230374e-01
3.08871865e-01 8.73799324e-01 -4.29513067e-01 7.74715066e-01
3.33676666e-01 3.45023543e-01 -4.49486375e-01 -5.75488247e-02
-1.67691216e-01 3.10319904e-02 4.36831504e-01 1.31410086e+00
4.44115639e-01 6.35504052e-02 -1.20850071e-01 5.18994689e-01
-1.62080690e-01 4.58531857e-01 -1.03046215e+00 -5.17976582e-01
6.90761149e-01 1.16933596e+00 -2.38632575e-01 -1.86962977e-01
-7.83451736e-01 1.08613968e+00 4.35345709e-01 4.15586829e-01
-4.34205115e-01 -6.97070658e-01 1.23457253e+00 -5.33405170e-02
3.65074307e-01 -2.76018053e-01 2.42506191e-01 -1.23719811e+00
2.06362978e-02 -4.91114080e-01 6.59012556e-01 -1.30259648e-01
-1.66881835e+00 2.89364457e-01 -1.30986929e-01 -1.40522528e+00
-2.82295972e-01 -9.33468997e-01 -9.49821323e-02 6.41100466e-01
-1.53131723e+00 -7.89880157e-01 2.42956698e-01 4.36991125e-01
-5.05670346e-02 -5.91756582e-01 1.35093546e+00 5.93729913e-01
-6.99519098e-01 6.48017347e-01 7.36247838e-01 1.48608103e-01
7.06987917e-01 -1.11929023e+00 -7.13252649e-02 4.10247594e-01
8.54304135e-02 5.93079209e-01 4.79978919e-01 -4.66591597e-01
-1.99205613e+00 -1.12675965e+00 9.74974573e-01 -3.87977034e-01
7.94675827e-01 -4.45255429e-01 -8.80793154e-01 3.45787197e-01
-3.87401611e-01 6.08728945e-01 1.54121411e+00 3.58719081e-01
-5.74041426e-01 -1.28378615e-01 -1.27534831e+00 3.10786039e-01
7.08702922e-01 -6.51379049e-01 -3.64740998e-01 5.66232264e-01
6.75204217e-01 6.92413598e-02 -1.39390910e+00 -1.81799922e-02
8.23846340e-01 -2.86784619e-01 1.11235559e+00 -1.44373679e+00
2.96156973e-01 -2.42122188e-01 -6.78644598e-01 -1.15219307e+00
-3.65812570e-01 -3.24790299e-01 -4.63509262e-01 1.08692527e+00
2.14857355e-01 -6.48219585e-01 6.88102067e-01 7.25474060e-01
5.03232777e-01 -7.98267841e-01 -1.16580129e+00 -8.68544698e-01
1.02991119e-01 -4.11161989e-01 3.63224000e-01 1.24158454e+00
3.59637290e-01 1.50609866e-01 -4.38471824e-01 4.69719887e-01
7.83870637e-01 -1.12862483e-01 6.12666428e-01 -1.49313903e+00
-1.66536316e-01 1.40948240e-02 -1.17675269e+00 -6.50077879e-01
3.79023433e-01 -8.83481443e-01 -4.48156983e-01 -1.18965042e+00
2.56448776e-01 -5.44058323e-01 -5.14303923e-01 3.32476616e-01
-8.91901925e-02 1.02746099e-01 -3.78702968e-01 5.97390421e-02
-5.14556468e-01 6.81306541e-01 5.29706955e-01 -1.19129024e-01
-2.05926538e-01 -4.12386686e-01 -7.41076946e-01 5.36133230e-01
4.72984612e-01 -6.79147363e-01 -3.26162249e-01 -1.39531910e-01
3.56491894e-01 -1.70008719e-01 2.43781790e-01 -5.30070722e-01
1.03045069e-01 -1.12623358e-02 5.05150199e-01 -2.48810366e-01
4.00948763e-01 -9.73989367e-01 -3.35057150e-03 3.88932765e-01
-5.54786503e-01 -8.53157192e-02 1.37966037e-01 1.02465737e+00
-9.17189345e-02 -1.00135691e-01 4.94981498e-01 4.18841839e-01
-6.06836319e-01 4.07760620e-01 -3.09213728e-01 1.23244978e-01
1.07494855e+00 -4.37415689e-02 -1.00422427e-01 -1.29750922e-01
-9.49397206e-01 2.79329680e-02 2.15957940e-01 5.84765673e-01
1.01763558e+00 -1.95180833e+00 -7.83127546e-01 3.37725103e-01
5.62549889e-01 -6.38660014e-01 9.87028629e-02 9.30552065e-01
-5.11697046e-02 3.70107085e-01 1.83763519e-01 -4.78558213e-01
-1.41978621e+00 8.26170444e-01 4.92570661e-02 1.37141421e-01
-3.76675367e-01 6.72966301e-01 8.45370740e-02 -5.26194990e-01
1.70674965e-01 -1.23028189e-01 -8.03393945e-02 3.33975852e-01
7.66730428e-01 4.49254006e-01 -5.90184741e-02 -5.52141666e-01
-7.03338325e-01 7.13851810e-01 -3.02389264e-01 -1.09447911e-01
1.51134074e+00 -1.47670820e-01 -2.11844146e-02 6.33926094e-01
1.86190140e+00 3.78940143e-02 -5.51488578e-01 -5.80310643e-01
3.13882008e-02 -5.96539855e-01 7.53401220e-02 -2.55754530e-01
-8.40026975e-01 7.36972392e-01 9.95392382e-01 1.58926547e-01
8.98156643e-01 3.29389155e-01 4.60250229e-01 5.45831561e-01
9.45239365e-02 -9.20121789e-01 -2.32852697e-01 -1.18423440e-03
4.86190587e-01 -1.30974102e+00 2.54706442e-01 -3.59181464e-01
-4.12672848e-01 1.23248971e+00 7.40965381e-02 -1.22858696e-01
1.02334177e+00 1.31958947e-01 -2.90883064e-01 -3.11200976e-01
-1.01674306e+00 9.49509721e-03 5.56801200e-01 8.74620855e-01
6.53479874e-01 9.79413539e-02 -3.89674067e-01 6.41634285e-01
4.10476893e-01 7.24231033e-03 2.66396463e-01 8.41453612e-01
-3.83929431e-01 -1.19422662e+00 -3.91968966e-01 7.44297683e-01
-4.40582633e-01 -1.17937885e-01 -3.21612984e-01 7.25162387e-01
-8.54769126e-02 5.79492331e-01 1.24474518e-01 -3.28100145e-01
2.34951124e-01 2.25037098e-01 3.62911403e-01 -4.82179314e-01
2.98263043e-01 -8.48178752e-03 -3.60589549e-02 -5.28478324e-01
-3.91600817e-01 -8.35394919e-01 -1.26311016e+00 -1.60985440e-01
-3.87980998e-01 1.66026101e-01 6.82986379e-01 7.75784254e-01
8.36885452e-01 1.63770884e-01 8.61741066e-01 -4.43099320e-01
-6.11586988e-01 -4.69870925e-01 -9.16159749e-01 5.35465062e-01
6.12733364e-01 -8.60848725e-01 -4.89837945e-01 4.70023719e-04] | [7.3635149002075195, 4.797533988952637] |
ef8b8879-0f73-442d-bfbf-fd7116b0741b | cater-a-diagnostic-dataset-for-compositional | 1910.04744 | null | https://arxiv.org/abs/1910.04744v2 | https://arxiv.org/pdf/1910.04744v2.pdf | CATER: A diagnostic dataset for Compositional Actions and TEmporal Reasoning | Computer vision has undergone a dramatic revolution in performance, driven in large part through deep features trained on large-scale supervised datasets. However, much of these improvements have focused on static image analysis; video understanding has seen rather modest improvements. Even though new datasets and spatiotemporal models have been proposed, simple frame-by-frame classification methods often still remain competitive. We posit that current video datasets are plagued with implicit biases over scene and object structure that can dwarf variations in temporal structure. In this work, we build a video dataset with fully observable and controllable object and scene bias, and which truly requires spatiotemporal understanding in order to be solved. Our dataset, named CATER, is rendered synthetically using a library of standard 3D objects, and tests the ability to recognize compositions of object movements that require long-term reasoning. In addition to being a challenging dataset, CATER also provides a plethora of diagnostic tools to analyze modern spatiotemporal video architectures by being completely observable and controllable. Using CATER, we provide insights into some of the most recent state of the art deep video architectures. | ['Deva Ramanan', 'Rohit Girdhar'] | 2019-10-10 | null | null | null | null | ['video-object-tracking'] | ['computer-vision'] | [ 1.50445446e-01 -1.14479497e-01 -3.85916919e-01 -4.27903652e-01
-4.15709764e-01 -8.40205133e-01 1.08444166e+00 -4.17855591e-01
-6.46342412e-02 5.30227244e-01 3.28049630e-01 -4.27067913e-02
7.16320947e-02 -4.16575164e-01 -9.93510604e-01 -7.13923752e-01
-3.70998085e-01 4.91297662e-01 3.98966193e-01 -3.00942779e-01
1.38998777e-01 4.52665746e-01 -1.76399612e+00 5.19176126e-01
2.35150620e-01 1.03270996e+00 -1.23160705e-02 9.76486444e-01
2.86917597e-01 1.51581335e+00 -4.11091000e-01 1.20307934e-02
2.84377933e-01 -4.17093664e-01 -8.81933033e-01 4.79787052e-01
9.77427363e-01 -7.16528535e-01 -9.13582742e-01 5.87932289e-01
1.68402225e-01 2.45850235e-02 4.73175853e-01 -1.59150195e+00
-8.30002487e-01 3.27695936e-01 -2.21413150e-01 5.78490376e-01
5.14257610e-01 6.66237950e-01 9.36383367e-01 -7.00471401e-01
9.77381349e-01 1.31578600e+00 5.49463689e-01 7.46110439e-01
-1.40522671e+00 -4.76481318e-01 4.70200300e-01 3.89749289e-01
-9.02305305e-01 -7.16691136e-01 8.12849045e-01 -7.68925428e-01
1.08632016e+00 1.39689684e-01 1.02807975e+00 1.75012743e+00
8.84290598e-03 1.10538363e+00 9.38459516e-01 5.06982673e-03
2.46094182e-01 -2.43611321e-01 -8.36329088e-02 7.85536587e-01
-3.62169407e-02 3.96343827e-01 -7.64961421e-01 3.00950438e-01
1.04352534e+00 -2.25919904e-03 -3.09331715e-01 -9.12597656e-01
-1.74453950e+00 5.12356579e-01 3.59029740e-01 9.25642326e-02
-5.11628166e-02 5.62679648e-01 3.60521615e-01 2.99511820e-01
1.00897543e-01 5.44051886e-01 -4.63937402e-01 -4.74918485e-01
-8.57207656e-01 7.22680092e-01 5.88875711e-01 1.18607402e+00
5.49421608e-01 4.15007472e-01 -2.28546157e-01 1.81008935e-01
-1.50691688e-01 6.34699702e-01 4.07448262e-01 -1.50080848e+00
2.14797765e-01 4.30573016e-01 2.46893540e-01 -1.25653052e+00
-3.20488155e-01 -1.95217818e-01 -5.86510122e-01 4.05060768e-01
5.88916600e-01 3.69694978e-01 -9.82857287e-01 2.01720095e+00
5.31518087e-02 5.29071569e-01 -1.18510120e-01 1.10681546e+00
6.02766871e-01 5.55527449e-01 -8.03634301e-02 1.17768019e-01
8.27017546e-01 -1.05128086e+00 -4.21654880e-01 -2.22456545e-01
5.16616225e-01 -2.56811976e-01 1.13136637e+00 3.72455120e-01
-1.33105779e+00 -6.10471249e-01 -9.67206597e-01 -2.92015761e-01
-2.52822459e-01 -4.63671595e-01 9.73313987e-01 2.70477057e-01
-1.22551942e+00 5.70717931e-01 -1.10099018e+00 -5.83220780e-01
6.76755846e-01 1.44339427e-01 -6.33389711e-01 -1.51536077e-01
-7.66716242e-01 7.70783544e-01 1.13920316e-01 -8.80400017e-02
-1.87762249e+00 -8.75931859e-01 -9.10699487e-01 -1.88412115e-01
4.53901023e-01 -7.14412630e-01 1.40564740e+00 -1.18218982e+00
-1.17438507e+00 1.16647375e+00 -1.49124205e-01 -6.75821126e-01
8.03380013e-01 -2.41567105e-01 -3.43577296e-01 4.92492974e-01
2.74299413e-01 9.72564101e-01 1.05379164e+00 -1.15882719e+00
-5.56648433e-01 -2.84569174e-01 5.30438781e-01 2.68140081e-02
-2.02106565e-01 5.11698984e-03 -4.31959063e-01 -7.34497488e-01
-1.19072445e-01 -1.12115335e+00 -1.94765497e-02 3.49603266e-01
-1.50618657e-01 9.18902010e-02 1.16945314e+00 -2.86194533e-01
7.59258986e-01 -1.94930482e+00 4.50038791e-01 -4.81455892e-01
3.87584805e-01 -2.96826154e-04 -2.23410279e-01 3.25599790e-01
-1.49136305e-01 1.01012655e-01 -1.53435707e-01 -1.74890384e-01
-6.03217892e-02 2.98418939e-01 -6.55179560e-01 5.89773655e-01
5.75076044e-01 1.15263689e+00 -1.22511125e+00 -4.26157147e-01
4.56301033e-01 2.97002643e-01 -8.69330406e-01 1.95728645e-01
-7.21294105e-01 7.75740325e-01 -3.39149296e-01 9.03841615e-01
9.08786803e-02 -5.15347004e-01 -1.26170158e-01 -1.31407574e-01
2.18101591e-02 5.25205582e-03 -7.11592138e-01 2.07925773e+00
-3.18005867e-02 1.14606261e+00 -8.87630358e-02 -1.23556280e+00
3.94954890e-01 2.74084508e-01 8.31478238e-01 -7.42036998e-01
2.60210317e-03 -8.09483975e-02 -5.24677970e-02 -9.19745326e-01
5.31625807e-01 2.44596288e-01 7.08278716e-02 2.47449532e-01
5.32942042e-02 -3.64997864e-01 3.73988301e-01 4.72219586e-01
1.49031317e+00 5.41810632e-01 -1.50037259e-01 -2.83971399e-01
1.25709802e-01 5.68418324e-01 4.67345625e-01 8.69720876e-01
-6.49144769e-01 8.97130072e-01 5.42623401e-01 -9.82705057e-01
-1.35528445e+00 -1.39214051e+00 -6.21979460e-02 8.25177729e-01
3.45730722e-01 -1.75170869e-01 -6.01155877e-01 -4.68859732e-01
-1.88584551e-02 4.42923546e-01 -8.95602286e-01 -1.79401562e-01
-7.28526413e-01 -2.94205904e-01 5.77725112e-01 7.51920223e-01
4.63474572e-01 -1.05583215e+00 -9.21045184e-01 1.84392318e-01
-1.24675736e-01 -1.54372287e+00 -2.45403171e-01 -9.98729020e-02
-8.53239238e-01 -1.35277295e+00 -5.19550145e-01 -6.96000755e-01
4.17457938e-01 6.27784073e-01 1.68116868e+00 8.22671428e-02
-4.36106414e-01 8.65337312e-01 -2.59537548e-01 -2.49828964e-01
-2.41166532e-01 -1.17597014e-01 3.14337164e-01 8.11518729e-03
3.52390617e-01 -6.23983145e-01 -7.64395475e-01 5.76965928e-01
-9.78726268e-01 1.92066118e-01 2.81191498e-01 8.39849651e-01
3.50291014e-01 -1.92164406e-01 3.20727408e-01 -6.24955297e-01
5.79953976e-02 -5.47549903e-01 -4.84133273e-01 3.42320800e-02
-9.84773487e-02 -4.16089557e-02 5.12448132e-01 -6.47212982e-01
-9.52716053e-01 -5.16464487e-02 3.64498347e-01 -9.70528841e-01
-3.35707307e-01 7.59524181e-02 3.44642550e-02 9.43246409e-02
8.26623261e-01 1.76883116e-01 -2.82870959e-02 -2.20927526e-03
5.23825169e-01 -3.44956592e-02 8.59497726e-01 -8.96327078e-01
7.68163919e-01 1.11291444e+00 -3.05228978e-02 -7.67030776e-01
-1.03148687e+00 -3.97488140e-02 -8.41591299e-01 -5.19165218e-01
1.12101400e+00 -1.11807728e+00 -6.75075352e-01 5.93575060e-01
-8.84994566e-01 -9.02894557e-01 -4.82573181e-01 2.58629262e-01
-9.98589218e-01 -6.56377450e-02 -6.75067067e-01 -4.34157223e-01
5.17698467e-01 -1.23160899e+00 1.29639280e+00 -1.28453940e-01
-4.20389175e-01 -8.64347398e-01 -7.69215971e-02 3.18813980e-01
3.57557476e-01 6.09123051e-01 7.97996342e-01 -6.47680312e-02
-1.35049260e+00 -5.13679609e-02 -2.04787955e-01 9.63731557e-02
6.50328025e-03 2.66526967e-01 -1.04885101e+00 -3.49635094e-01
-4.47056582e-03 -7.87694037e-01 6.33550465e-01 5.10553598e-01
1.42948580e+00 -3.74426097e-02 -2.37103775e-01 8.96604836e-01
1.15466487e+00 1.98178947e-01 5.18595278e-01 5.02101719e-01
7.29357183e-01 4.84446049e-01 5.42327046e-01 1.65348440e-01
3.51363003e-01 6.66457236e-01 7.07509875e-01 1.89602450e-01
-2.53625244e-01 -3.31948370e-01 3.74576420e-01 5.71950376e-01
-1.68097988e-01 -3.67761195e-01 -1.09874475e+00 6.15814865e-01
-1.90816414e+00 -1.43933451e+00 5.00538051e-02 1.67739487e+00
4.60001111e-01 3.91726613e-01 1.86104298e-01 9.83960330e-02
2.45875075e-01 5.05518317e-01 -8.91994715e-01 1.49016291e-01
-5.81511796e-01 -2.47177243e-01 1.97081476e-01 9.94505584e-02
-1.21858311e+00 9.85805929e-01 7.18339109e+00 1.77967384e-01
-1.29120946e+00 -2.42045507e-01 6.76916599e-01 -5.59726357e-01
-2.82923520e-01 4.26756218e-02 -4.71993357e-01 3.79443526e-01
6.68535113e-01 -3.06203812e-02 5.70697725e-01 9.22971904e-01
2.61304796e-01 1.97990499e-02 -1.85877347e+00 1.15066946e+00
2.07972184e-01 -1.87721860e+00 2.48649314e-01 1.85863915e-04
9.50886309e-01 1.15429021e-01 3.24145854e-01 2.80835479e-01
3.58059406e-01 -1.19376922e+00 1.30533600e+00 5.20915270e-01
7.49493599e-01 -1.34448126e-01 1.49782076e-02 7.90084749e-02
-9.19124246e-01 -3.11562896e-01 -1.67962968e-01 -4.47346389e-01
1.61254734e-01 1.32158056e-01 -3.57038200e-01 -9.32455137e-02
1.09242415e+00 1.26836276e+00 -4.78216916e-01 6.32466018e-01
1.59037467e-02 4.55154747e-01 -9.02002454e-02 1.73170164e-01
3.94141972e-01 3.70575562e-02 6.27137363e-01 8.66231501e-01
3.45769920e-03 1.50507182e-01 2.03510657e-01 8.03759694e-01
-1.90077990e-01 -6.50162876e-01 -1.15542853e+00 -2.63370603e-01
2.50316411e-01 8.12927663e-01 -5.69984317e-01 -5.38236082e-01
-5.40042818e-01 8.04100454e-01 4.24364775e-01 4.91679072e-01
-1.10356045e+00 4.01048601e-01 1.14054585e+00 1.98404297e-01
1.33295581e-01 -6.16730571e-01 -3.56812589e-02 -1.61157525e+00
1.00412965e-01 -1.03413939e+00 1.07923396e-01 -1.21140671e+00
-1.03402019e+00 3.66576850e-01 1.77375972e-01 -1.28726780e+00
-4.46919560e-01 -7.97781467e-01 -3.55678320e-01 1.17403783e-01
-1.18340695e+00 -1.11430335e+00 -5.98158777e-01 8.60343099e-01
1.07402325e+00 -1.82937235e-01 6.03702009e-01 3.07119936e-01
-4.52362984e-01 1.89479813e-01 -1.25756085e-01 2.59676009e-01
5.23493409e-01 -1.11641240e+00 6.45770490e-01 7.65237272e-01
5.04008293e-01 4.23816353e-01 8.80429804e-01 -2.70582676e-01
-1.83122158e+00 -9.85701740e-01 2.28595182e-01 -1.28368425e+00
8.78381610e-01 -6.57872677e-01 -6.72371686e-01 1.20635247e+00
1.03277475e-01 3.81327301e-01 7.91872367e-02 -8.62260535e-02
-6.64753854e-01 -1.04026638e-01 -7.64149785e-01 9.38422620e-01
1.75613260e+00 -6.92738116e-01 -5.52132547e-01 3.88671011e-01
8.06843698e-01 -5.49819946e-01 -6.88322008e-01 5.24738848e-01
7.88834631e-01 -1.34434044e+00 1.13458252e+00 -1.27669871e+00
8.65122736e-01 -3.35364729e-01 -3.40572417e-01 -1.05588222e+00
-3.23491395e-01 -5.26752293e-01 -4.67474073e-01 8.87397945e-01
1.06373921e-01 -1.08411200e-01 1.22142446e+00 8.44556868e-01
-1.92536145e-01 -6.37853563e-01 -4.84082103e-01 -9.64803219e-01
-9.06751305e-02 -6.14108860e-01 4.59975272e-01 1.12426186e+00
-3.17633212e-01 1.62281588e-01 -5.13099372e-01 -3.96159813e-02
7.01772988e-01 2.05039084e-01 1.11362827e+00 -8.46873283e-01
-2.91192174e-01 -5.28925478e-01 -8.87072682e-01 -1.25574708e+00
1.41729951e-01 -5.24627686e-01 -1.14335969e-01 -1.09446800e+00
2.23137647e-01 -2.38123387e-01 -1.74493268e-01 2.15431526e-01
3.21903646e-01 4.01890755e-01 1.23268232e-01 2.78019965e-01
-7.89778829e-01 5.10354638e-01 1.24837363e+00 -4.51417059e-01
1.44041330e-01 -4.01787370e-01 -3.44796032e-01 9.20777619e-01
5.55607617e-01 -9.49177071e-02 -7.39468098e-01 -1.08082914e+00
9.85557884e-02 9.76179987e-02 8.26818764e-01 -1.19799948e+00
1.32047266e-01 -3.80440831e-01 5.91302037e-01 -4.68936503e-01
7.03039467e-01 -8.03969681e-01 1.13687344e-01 2.00765952e-01
-5.77074051e-01 3.81298542e-01 7.68647492e-02 8.25359166e-01
-3.79209518e-01 5.53224087e-01 7.81609416e-01 -4.78315413e-01
-1.28232491e+00 6.50478184e-01 -3.18586886e-01 4.49409336e-01
1.25262380e+00 -3.63164067e-01 -4.67787325e-01 -4.31730837e-01
-7.33031929e-01 2.89480865e-01 8.77376318e-01 8.32648993e-01
4.89452213e-01 -1.26407754e+00 -5.21619201e-01 1.85891584e-01
3.90528947e-01 7.68340528e-02 4.03446734e-01 6.22973084e-01
-7.16121793e-01 4.39279556e-01 -3.91728789e-01 -1.22512949e+00
-9.22102451e-01 9.04368699e-01 3.48175406e-01 3.00314039e-01
-8.14804554e-01 8.94656003e-01 5.01753747e-01 -6.15178719e-02
4.25900191e-01 -5.00383854e-01 3.04660290e-01 -2.97122478e-01
4.20144588e-01 1.18470959e-01 -2.73897409e-01 -6.65775716e-01
-2.85294861e-01 4.74023014e-01 1.08756892e-01 -3.77984196e-02
1.36120665e+00 -1.09905228e-01 2.22681090e-01 5.54278791e-01
1.16161335e+00 -4.35052276e-01 -1.96161067e+00 1.89758129e-02
-5.37618771e-02 -7.33154893e-01 -2.51397610e-01 -5.89188635e-01
-1.07624137e+00 9.85626936e-01 3.89115572e-01 2.86054969e-01
9.60002482e-01 1.28373876e-01 6.27602398e-01 5.74603975e-01
6.21677399e-01 -7.53997386e-01 5.59583008e-01 4.83292043e-01
8.62279654e-01 -1.61381686e+00 -2.01294497e-01 -6.07901365e-02
-5.29278755e-01 9.10427451e-01 8.65320325e-01 -3.41914058e-01
5.39414287e-01 2.41136849e-01 4.32078652e-02 -3.23940963e-01
-1.06952214e+00 2.04233322e-02 -9.16631818e-02 6.61305964e-01
1.61899000e-01 -3.24621797e-01 5.64835310e-01 -1.72368243e-01
-1.53228417e-01 1.30526006e-01 5.14560461e-01 1.03078282e+00
-1.11407526e-01 -7.20616698e-01 -2.61233579e-02 3.16270471e-01
-3.15732509e-01 2.34180108e-01 -1.80721357e-01 1.04607582e+00
1.60326459e-03 7.76053786e-01 2.74666429e-01 -2.98964113e-01
1.91720933e-01 -2.75376618e-01 7.68704236e-01 -3.67969841e-01
-7.77039900e-02 -4.74279851e-01 9.75863412e-02 -1.11451483e+00
-6.89890921e-01 -8.89365971e-01 -9.09085989e-01 -3.03138047e-01
3.79507869e-01 -3.07933390e-01 4.23942953e-01 7.75821209e-01
2.50490129e-01 5.04431367e-01 3.61827463e-01 -1.42229676e+00
-2.32602820e-01 -4.28361028e-01 -3.04201573e-01 9.18330073e-01
8.03339124e-01 -8.67660165e-01 -3.07759911e-01 6.33694828e-01] | [8.680935859680176, 0.47692692279815674] |
97aea8c4-f223-41d1-91eb-0ba057f26dfe | learning-free-form-deformations-for-3d-object | 1803.10932 | null | http://arxiv.org/abs/1803.10932v1 | http://arxiv.org/pdf/1803.10932v1.pdf | Learning Free-Form Deformations for 3D Object Reconstruction | Representing 3D shape in deep learning frameworks in an accurate, efficient
and compact manner still remains an open challenge. Most existing work
addresses this issue by employing voxel-based representations. While these
approaches benefit greatly from advances in computer vision by generalizing 2D
convolutions to the 3D setting, they also have several considerable drawbacks.
The computational complexity of voxel-encodings grows cubically with the
resolution thus limiting such representations to low-resolution 3D
reconstruction. In an attempt to solve this problem, point cloud
representations have been proposed. Although point clouds are more efficient
than voxel representations as they only cover surfaces rather than volumes,
they do not encode detailed geometric information about relationships between
points. In this paper we propose a method to learn free-form deformations (FFD)
for the task of 3D reconstruction from a single image. By learning to deform
points sampled from a high-quality mesh, our trained model can be used to
produce arbitrarily dense point clouds or meshes with fine-grained geometry. We
evaluate our proposed framework on both synthetic and real-world data and
achieve state-of-the-art results on point-cloud and volumetric metrics.
Additionally, we qualitatively demonstrate its applicability to label
transferring for 3D semantic segmentation. | ['Anders Eriksson', 'Frederic Maire', 'Dominic Jack', 'Clinton Fookes', 'Sridha Sridharan', 'Sareh Shirazi', 'Jhony K. Pontes'] | 2018-03-29 | null | null | null | null | ['3d-object-reconstruction'] | ['computer-vision'] | [ 3.00025702e-01 1.00759029e-01 8.72522146e-02 -5.01475871e-01
-9.29895103e-01 -4.35336739e-01 6.32351637e-01 3.80467415e-01
-2.02230364e-01 3.88857633e-01 -4.46063936e-01 -1.37568846e-01
1.85110688e-01 -1.32646346e+00 -1.15800083e+00 -3.93495291e-01
-7.39298835e-02 9.47192013e-01 4.07614350e-01 -2.29394864e-02
2.50689864e-01 1.18771482e+00 -1.60961664e+00 1.76308811e-01
9.60890472e-01 1.17644024e+00 7.78981596e-02 2.23374739e-01
-7.08745301e-01 -4.79843840e-02 -3.08402807e-01 -1.57907143e-01
3.86729836e-01 7.80427158e-02 -9.21751797e-01 3.43458802e-01
7.34089375e-01 -3.59118551e-01 5.28515466e-02 9.44042563e-01
2.00685933e-01 4.85804640e-02 1.06650257e+00 -9.24704015e-01
-5.28809965e-01 -5.65980636e-02 -7.20431566e-01 -2.10406452e-01
6.90134615e-02 -9.05631259e-02 7.57847965e-01 -1.07676399e+00
5.96268177e-01 1.45899642e+00 8.68020952e-01 4.53254014e-01
-1.37330556e+00 -4.85445410e-01 6.65872321e-02 -3.78962725e-01
-1.29320920e+00 -7.16567039e-03 9.47631598e-01 -6.59141481e-01
1.03992522e+00 3.22348922e-01 8.02572072e-01 5.74805081e-01
7.81391039e-02 5.23422301e-01 1.16968668e+00 -2.33982697e-01
3.10708135e-01 -1.20409384e-01 -9.61454883e-02 7.82142997e-01
2.58902669e-01 -1.04051508e-01 -1.18515290e-01 -3.71739537e-01
1.49351037e+00 2.09805608e-01 -2.03398153e-01 -7.22664058e-01
-9.51499701e-01 1.10459638e+00 8.80915165e-01 2.03561440e-01
-4.87281501e-01 6.15261674e-01 2.31804207e-01 -4.32473123e-02
1.05289900e+00 2.03668863e-01 -2.84756869e-01 2.75849462e-01
-1.13572609e+00 5.37359178e-01 4.38638628e-01 8.02865505e-01
1.03406405e+00 2.84161139e-02 1.26370713e-01 7.56164372e-01
5.02512693e-01 3.85429204e-01 2.82438342e-02 -1.16232443e+00
2.51262814e-01 7.45685220e-01 1.19488001e-01 -9.58597004e-01
-3.06660295e-01 -4.58564728e-01 -7.65951097e-01 6.01387262e-01
1.44644320e-01 5.26098728e-01 -1.31659079e+00 1.11516356e+00
5.44452012e-01 3.88991088e-01 -2.61215955e-01 9.37876105e-01
9.53975916e-01 6.28762305e-01 7.99337476e-02 1.58212125e-01
1.06396747e+00 -6.29973590e-01 -1.77566692e-01 4.58277799e-02
4.54834342e-01 -5.08829653e-01 9.08662200e-01 1.90801278e-01
-1.40399528e+00 -3.73073906e-01 -8.69975984e-01 -3.55535746e-01
-2.07967937e-01 -3.66729766e-01 7.15095818e-01 5.66064000e-01
-1.20520139e+00 9.12169635e-01 -1.03378642e+00 -2.72400618e-01
1.05067754e+00 3.82337332e-01 -1.47308096e-01 -1.12462126e-01
-8.09202969e-01 6.78573072e-01 3.12454719e-02 -2.43541315e-01
-7.38746285e-01 -9.93812025e-01 -9.66428101e-01 -2.38209888e-02
-6.32710457e-02 -9.51753557e-01 1.07424271e+00 -4.73220825e-01
-1.30705464e+00 1.34041858e+00 -5.47693018e-03 -5.02780199e-01
4.56918001e-01 -1.96454301e-02 3.40426028e-01 2.58907348e-01
9.13095996e-02 7.70375907e-01 7.85855353e-01 -1.66648436e+00
-2.10742965e-01 -5.98664045e-01 2.24701807e-01 1.56044587e-01
9.23453122e-02 -4.10637468e-01 -2.81580567e-01 -5.75797796e-01
5.80928504e-01 -6.74657345e-01 -4.60944027e-01 6.41212940e-01
-1.30488738e-01 -2.67444551e-01 9.45071995e-01 -3.42260242e-01
2.77664244e-01 -1.91679716e+00 1.75662667e-01 1.05418772e-01
3.10038298e-01 1.81648478e-01 1.15667433e-01 1.71252370e-01
2.92118102e-01 4.26999927e-01 -8.49321544e-01 -6.07966900e-01
1.05787404e-02 4.45506662e-01 -2.01723829e-01 6.82797372e-01
5.31854570e-01 9.96790230e-01 -7.47529328e-01 -5.77642739e-01
6.32711947e-01 1.07043505e+00 -7.74416864e-01 -1.37139810e-03
-5.05470932e-01 6.53464854e-01 -6.82969391e-01 7.09036946e-01
1.04898429e+00 -3.83817822e-01 -4.52089876e-01 -3.03275939e-02
-1.09601982e-01 2.15804592e-01 -9.37812150e-01 2.09480882e+00
-5.88780344e-01 2.26699576e-01 3.11217427e-01 -1.22592938e+00
1.03970027e+00 2.48740777e-01 7.88679600e-01 -4.75984544e-01
1.42388538e-01 3.20838958e-01 -6.69268250e-01 -4.69342284e-02
3.33913088e-01 -6.16641462e-01 1.01184554e-03 2.81563789e-01
-8.42252150e-02 -1.05674207e+00 -3.44453484e-01 -1.41236216e-01
7.82116234e-01 3.94057482e-01 3.05103231e-02 -1.26534015e-01
3.72964472e-01 1.98236719e-01 2.92226583e-01 3.69462579e-01
2.27887824e-01 9.66011763e-01 1.20265745e-01 -4.63847667e-01
-1.13036859e+00 -1.28670037e+00 -6.40332699e-01 3.16314459e-01
2.41086036e-01 -1.11386832e-02 -7.82639265e-01 -4.25021946e-01
3.18805575e-01 5.25007546e-01 -4.60057259e-01 2.43307143e-01
-7.57271111e-01 -4.97522920e-01 2.61764169e-01 5.57172656e-01
4.40226793e-01 -8.43611300e-01 -9.17948008e-01 3.59748930e-01
6.12685941e-02 -1.15605950e+00 -3.27657796e-02 2.19895281e-02
-1.51564288e+00 -1.05250728e+00 -8.44606042e-01 -7.38090515e-01
8.44338179e-01 2.51104355e-01 1.39371741e+00 1.99384943e-01
-3.22187781e-01 5.23260832e-01 -1.67505160e-01 -4.22945857e-01
-3.65797341e-01 -2.02801183e-01 -4.04998124e-01 -1.62841275e-01
1.99178699e-02 -7.86906719e-01 -6.07693076e-01 6.13909736e-02
-1.04630876e+00 1.24579199e-01 4.02092963e-01 5.64153969e-01
1.40071249e+00 -4.05498557e-02 3.08755755e-01 -1.06308579e+00
2.76055694e-01 -4.13455158e-01 -6.00269377e-01 -2.23510176e-01
-3.00187558e-01 4.23694449e-03 4.17271912e-01 -3.55008282e-02
-7.60616601e-01 2.29608789e-01 -5.10706484e-01 -9.37147141e-01
-3.75536531e-01 2.62158066e-01 2.54907191e-01 -4.22775328e-01
4.33044732e-01 2.14338243e-01 -4.93212156e-02 -7.36951649e-01
4.82640535e-01 3.24268550e-01 2.93969065e-01 -8.24935198e-01
8.91423047e-01 1.04015613e+00 3.01399082e-01 -9.23079729e-01
-5.50250530e-01 -3.74722391e-01 -9.39903259e-01 -7.79826641e-02
8.90056908e-01 -8.34074438e-01 -4.23783630e-01 2.54188538e-01
-1.43230689e+00 -2.44405001e-01 -5.51082850e-01 9.66022685e-02
-8.90584171e-01 3.80046844e-01 -5.48171282e-01 -6.99015975e-01
-4.07758892e-01 -1.31116283e+00 1.63491952e+00 -1.49865702e-01
-2.21078489e-02 -9.72908497e-01 -5.36339432e-02 1.96888521e-01
3.57733190e-01 9.61982250e-01 1.21062768e+00 1.57344177e-01
-8.61264944e-01 -1.52916372e-01 -3.40658098e-01 2.58677036e-01
2.56387219e-02 -2.40486577e-01 -1.04914534e+00 -1.24611929e-01
1.01689890e-01 -2.77663499e-01 8.14324558e-01 5.97936094e-01
1.66175616e+00 6.11098148e-02 -4.32157427e-01 8.95667553e-01
1.68196034e+00 -2.09136009e-01 5.22307038e-01 9.54155028e-02
8.92417431e-01 5.40774405e-01 3.77190679e-01 2.83117414e-01
3.29998702e-01 6.91125214e-01 8.98645341e-01 -2.13752255e-01
-4.44488227e-01 -1.59937948e-01 -2.82722980e-01 7.18786776e-01
-3.00823271e-01 3.96519862e-02 -8.52204204e-01 4.97651458e-01
-1.45000815e+00 -5.15649498e-01 -5.19318223e-01 2.05686474e+00
6.77915096e-01 5.47090210e-02 -5.64709231e-02 3.44691128e-01
4.06901091e-01 1.39920980e-01 -5.76027393e-01 -5.51935911e-01
1.46505296e-01 6.99978292e-01 4.55318391e-01 4.82494414e-01
-1.05524850e+00 9.11407113e-01 5.81100845e+00 6.59069657e-01
-1.26464939e+00 3.08141708e-01 6.22086048e-01 -2.73365472e-02
-7.10077584e-01 -2.74351060e-01 -4.81298953e-01 2.65233696e-01
6.14050984e-01 1.95280716e-01 1.01961888e-01 8.38323534e-01
1.14156716e-01 1.28925547e-01 -1.12199199e+00 1.21920228e+00
-1.85171187e-01 -1.68081605e+00 3.36002350e-01 2.55573332e-01
8.19447160e-01 2.01606780e-01 1.16286725e-01 1.51305227e-02
5.52597046e-02 -1.43442130e+00 8.88827443e-01 4.32511330e-01
1.16083968e+00 -7.40117252e-01 3.46025169e-01 4.76633698e-01
-1.27189469e+00 5.42967498e-01 -6.59329295e-01 -3.77573147e-02
1.81125641e-01 8.59656751e-01 -7.17314184e-01 6.29062891e-01
7.74004161e-01 7.01191306e-01 -1.75875708e-01 1.06654620e+00
1.54173210e-01 2.93792605e-01 -4.96086001e-01 2.64663160e-01
4.89545614e-01 -2.61920691e-01 3.93365800e-01 9.88683581e-01
3.88031662e-01 2.01374188e-01 2.09487319e-01 1.47153962e+00
-2.57295042e-01 5.92132173e-02 -7.55382538e-01 1.35868013e-01
3.48846853e-01 1.00315213e+00 -1.03553545e+00 -1.93076387e-01
-3.82014126e-01 9.10066664e-01 5.33302486e-01 2.23784626e-01
-6.25994980e-01 -1.78995095e-02 7.58960545e-01 6.40408158e-01
3.79548103e-01 -6.56159937e-01 -7.06931233e-01 -7.40248024e-01
1.10409431e-01 -1.33064955e-01 -1.45743623e-01 -5.76074183e-01
-1.35636580e+00 6.37827218e-01 8.45663324e-02 -1.09693813e+00
4.26271930e-03 -5.67633390e-01 -3.81564379e-01 1.08660650e+00
-1.92716146e+00 -1.04591691e+00 -3.88366938e-01 4.96310622e-01
6.27917707e-01 4.72015768e-01 9.64984536e-01 2.25669608e-01
1.96662620e-01 3.88901569e-02 -7.08746165e-02 -6.25986233e-02
-2.22614426e-02 -1.21094930e+00 7.87433982e-01 2.48227179e-01
8.71838927e-02 2.39380464e-01 4.04756576e-01 -5.82824051e-01
-1.38808370e+00 -1.37834644e+00 4.11053151e-01 -4.46656883e-01
1.29259648e-02 -4.46343154e-01 -1.23369670e+00 4.40441430e-01
-4.12688196e-01 6.14006341e-01 3.09626371e-01 -3.45419168e-01
-2.28673473e-01 4.20996100e-01 -1.65103126e+00 1.47433951e-01
1.25967550e+00 -3.64848912e-01 -5.12088001e-01 3.52951914e-01
7.28216112e-01 -7.06520081e-01 -1.11452651e+00 5.15470505e-01
5.78333139e-02 -1.04933238e+00 1.36645019e+00 -2.16636047e-01
4.96795356e-01 -2.08167851e-01 -1.71220466e-01 -1.19653165e+00
-2.07119003e-01 -1.45675212e-01 -1.44494757e-01 7.59870887e-01
-1.34291276e-01 -5.29870391e-01 1.09625936e+00 3.57456356e-01
-5.57184279e-01 -1.22757113e+00 -1.11052108e+00 -7.47729897e-01
6.42721117e-01 -6.23641372e-01 8.72042596e-01 9.05782640e-01
-7.81345725e-01 -8.16565901e-02 2.31100366e-01 2.03044385e-01
1.02319992e+00 4.58439887e-01 4.40734982e-01 -1.74573636e+00
1.30224541e-01 -5.24976730e-01 -5.77835441e-01 -1.22820938e+00
2.40571111e-01 -1.32667220e+00 -5.29273711e-02 -2.04674578e+00
-1.73751682e-01 -1.15789962e+00 2.55570620e-01 1.37685284e-01
3.18169862e-01 6.45310819e-01 -4.74940985e-02 3.63888532e-01
-4.04723063e-02 6.65829241e-01 1.57790434e+00 -2.58089006e-01
2.01569758e-02 -1.67408451e-01 -2.47791842e-01 8.28572869e-01
6.70153320e-01 -4.42086637e-01 -3.11359972e-01 -8.59115362e-01
-1.24269836e-01 1.63469449e-01 8.05811465e-01 -8.90220404e-01
-2.27745920e-01 1.20193487e-04 5.29132724e-01 -9.51683819e-01
8.05930972e-01 -8.93207192e-01 6.37931079e-02 2.97408551e-01
-3.17095742e-02 -2.79556513e-01 2.89096415e-01 6.48074925e-01
-5.67450337e-02 -1.77249849e-01 1.07205582e+00 -5.40489674e-01
-4.15428251e-01 7.75836587e-01 6.19265251e-02 -2.61864699e-02
1.01070797e+00 -5.39491177e-01 1.89675644e-01 1.14683889e-01
-5.56846976e-01 -1.34305552e-01 9.84041631e-01 1.38228998e-01
1.06469798e+00 -1.42584848e+00 -6.64397657e-01 2.15073243e-01
-1.37632370e-01 9.99340653e-01 1.79618746e-01 3.17294478e-01
-9.98365700e-01 5.03665745e-01 -1.82716817e-01 -1.09109330e+00
-8.01619709e-01 2.85355717e-01 5.31276047e-01 4.73641587e-04
-1.13703263e+00 1.02644527e+00 4.34662193e-01 -6.67875767e-01
1.35561064e-01 -7.64166951e-01 1.36509044e-02 -4.00581300e-01
5.94698451e-02 1.23610094e-01 3.75511855e-01 -7.79271722e-01
-3.48549753e-01 1.05575562e+00 2.73983508e-01 1.62812144e-01
1.49901950e+00 2.16586620e-01 -6.79478198e-02 5.23978651e-01
1.32503200e+00 -3.99713427e-01 -1.36312401e+00 -2.42402419e-01
-2.83783704e-01 -7.55804539e-01 3.56369585e-01 -2.88870484e-01
-1.26444829e+00 1.29659152e+00 5.31745970e-01 1.79278061e-01
5.62519968e-01 4.42021489e-01 1.16558814e+00 -9.14193094e-02
7.64757216e-01 -5.74605286e-01 -6.62106369e-03 3.50924164e-01
1.02541816e+00 -1.22626829e+00 1.06062934e-01 -9.56620693e-01
3.64781320e-02 1.04847467e+00 2.77952552e-01 -6.98527277e-01
7.81547785e-01 1.24510013e-01 -3.61019880e-01 -6.63796067e-01
-1.92622066e-01 -5.31276204e-02 3.11725706e-01 8.21021140e-01
5.24748147e-01 2.01301172e-01 -8.45791847e-02 2.03891098e-01
-1.65045545e-01 -6.66562021e-02 2.93421537e-01 9.44762707e-01
-4.67291683e-01 -9.97863770e-01 -5.72251379e-01 6.46018624e-01
-1.60038322e-01 2.19240397e-01 -1.59032077e-01 6.63889229e-01
1.36123851e-01 3.85099769e-01 5.06527066e-01 1.99226532e-02
3.93602073e-01 -1.59045428e-01 9.78012443e-01 -1.04068553e+00
-2.75658965e-01 3.96502623e-03 -3.41474742e-01 -6.31813765e-01
-5.34048557e-01 -7.60721028e-01 -1.80264235e+00 -1.59353897e-01
5.11016212e-02 -1.54215023e-01 9.81602967e-01 7.69123077e-01
2.15963572e-01 5.26583314e-01 4.06701684e-01 -1.56708634e+00
-4.20887798e-01 -5.11078775e-01 -6.64529860e-01 4.64729726e-01
1.53045923e-01 -1.05862963e+00 -1.49830729e-01 -2.15448931e-01] | [8.524945259094238, -3.6043901443481445] |
46c30d80-dabb-470d-ace7-71d1712a25db | deep-relighting-networks-for-image-light | 2008.08298 | null | https://arxiv.org/abs/2008.08298v2 | https://arxiv.org/pdf/2008.08298v2.pdf | Deep Relighting Networks for Image Light Source Manipulation | Manipulating the light source of given images is an interesting task and useful in various applications, including photography and cinematography. Existing methods usually require additional information like the geometric structure of the scene, which may not be available for most images. In this paper, we formulate the single image relighting task and propose a novel Deep Relighting Network (DRN) with three parts: 1) scene reconversion, which aims to reveal the primary scene structure through a deep auto-encoder network, 2) shadow prior estimation, to predict light effect from the new light direction through adversarial learning, and 3) re-renderer, to combine the primary structure with the reconstructed shadow view to form the required estimation under the target light source. Experimental results show that the proposed method outperforms other possible methods, both qualitatively and quantitatively. Specifically, the proposed DRN has achieved the best PSNR in the "AIM2020 - Any to one relighting challenge" of the 2020 ECCV conference. | ['Daniel P. K. Lun', 'Chu-Tak Li', 'Zhi-Song Liu', 'Wan-Chi Siu', 'Li-Wen Wang'] | 2020-08-19 | null | null | null | null | ['image-relighting'] | ['computer-vision'] | [ 4.90610749e-01 -4.76118401e-02 4.53126222e-01 -1.61943018e-01
-2.21435696e-01 -2.57979453e-01 5.56154847e-01 -5.54111540e-01
-6.01846389e-02 8.51463318e-01 3.18246067e-01 -2.51145393e-01
4.25390214e-01 -8.04970920e-01 -1.00753546e+00 -8.19278657e-01
5.18769741e-01 -1.08389676e-01 1.76383644e-01 -2.89240539e-01
2.69564569e-01 4.83592808e-01 -1.26895046e+00 4.24536355e-02
8.91693592e-01 1.08419228e+00 3.80566955e-01 5.73508024e-01
1.59767300e-01 1.19496822e+00 -4.58802253e-01 -4.65812147e-01
4.37901229e-01 -5.22226036e-01 -5.68033099e-01 3.13753158e-01
4.29316729e-01 -1.05886614e+00 -8.38704348e-01 1.08004951e+00
5.04906356e-01 1.63623855e-01 3.06849301e-01 -1.11455858e+00
-6.21782124e-01 1.42703250e-01 -7.99463332e-01 7.40678236e-02
3.76472026e-01 3.57318401e-01 3.63869935e-01 -7.19911754e-01
6.96115613e-01 1.33254850e+00 5.28556764e-01 2.90333837e-01
-1.26345158e+00 -7.21237659e-01 -2.41766408e-01 4.60510015e-01
-1.24978280e+00 -4.51260298e-01 1.37128103e+00 -2.20205396e-01
2.52580523e-01 1.88608825e-01 5.71197987e-01 1.11968040e+00
3.96367937e-01 6.21805012e-01 1.64873183e+00 -3.03611726e-01
4.31009345e-02 6.23511896e-02 -4.73789692e-01 6.39719903e-01
-1.25292405e-01 4.71940219e-01 -4.23523515e-01 3.35284382e-01
1.03111219e+00 -3.44522228e-03 -7.56169558e-01 -3.05412233e-01
-1.07243276e+00 3.99841577e-01 7.76251018e-01 -2.25255061e-02
-3.44374597e-01 2.53506184e-01 1.29982427e-01 2.59030491e-01
4.47499633e-01 2.06282437e-01 -8.12518038e-03 3.10624063e-01
-7.52979994e-01 1.28456056e-01 5.64330816e-01 8.51510823e-01
9.81160700e-01 4.58907962e-01 -3.98210101e-02 6.30068719e-01
1.89840868e-01 5.97713411e-01 1.22754321e-01 -1.12125278e+00
5.07991135e-01 2.15780914e-01 5.41520298e-01 -1.29017687e+00
-1.68391615e-01 -3.31221014e-01 -1.26676083e+00 4.97820228e-01
4.13490385e-02 -2.14692265e-01 -8.07173789e-01 1.46559501e+00
5.18938780e-01 4.90141600e-01 7.08492026e-02 1.09903502e+00
1.13837135e+00 9.98674154e-01 -4.28262681e-01 -3.96033198e-01
9.79856491e-01 -9.57780063e-01 -9.49317276e-01 -2.61489898e-01
-3.44424620e-02 -1.11997437e+00 8.96396399e-01 4.72039014e-01
-1.12835634e+00 -7.17399359e-01 -1.18415499e+00 -4.30473417e-01
1.11629702e-02 4.18516286e-02 5.17063081e-01 3.98254246e-01
-1.06182051e+00 3.79391044e-01 -3.21237594e-01 1.87255889e-02
3.99017870e-01 1.57749150e-02 -1.30621821e-01 -5.11633694e-01
-1.18736875e+00 7.64521956e-01 3.47145408e-01 3.96624714e-01
-1.26735091e+00 -8.19689512e-01 -7.48455465e-01 6.24403469e-02
4.58800822e-01 -7.01227486e-01 8.41089845e-01 -1.05748761e+00
-1.72140038e+00 6.70073569e-01 7.13835210e-02 -2.94041544e-01
7.76069343e-01 -1.91369846e-01 -3.49833488e-01 1.73550829e-01
-1.94419593e-01 4.97117758e-01 1.04511666e+00 -1.88602495e+00
-3.03065419e-01 -1.75072432e-01 3.35697889e-01 3.34082723e-01
6.26409277e-02 -2.66028047e-01 -5.08672178e-01 -8.25480342e-01
-3.85316871e-02 -7.06248760e-01 -8.47482979e-02 2.04224080e-01
-6.40328765e-01 4.07378614e-01 1.18278444e+00 -1.14549518e+00
8.33123326e-01 -2.09538412e+00 1.18722178e-01 -1.03235833e-01
1.93236932e-01 2.96169460e-01 -1.05696291e-01 5.31473517e-01
-2.00562224e-01 -3.33208740e-01 -2.11360097e-01 -4.04615492e-01
-2.56889164e-01 1.85663730e-01 -6.79990709e-01 6.76661193e-01
-1.80150434e-01 9.25527573e-01 -8.29182029e-01 -4.66155678e-01
7.25755453e-01 7.33150661e-01 -3.69104505e-01 6.27707839e-01
-1.53881416e-01 8.57074380e-01 -1.25953525e-01 3.47193569e-01
1.24503648e+00 8.37093517e-02 -3.88777703e-02 -6.38901591e-01
-1.87524736e-01 -8.62712488e-02 -1.17048490e+00 1.71686757e+00
-6.76246285e-01 9.17352200e-01 2.77478337e-01 -6.99375391e-01
8.52139890e-01 -8.21429864e-03 4.80909824e-01 -9.14164186e-01
2.52159357e-01 -7.25296736e-02 -2.90390581e-01 -7.40513265e-01
5.50377309e-01 -1.62931606e-01 2.58803844e-01 1.48783639e-01
-3.28011811e-01 -5.07641733e-01 -2.10569963e-01 9.38577279e-02
7.58755088e-01 2.37670317e-01 4.92806099e-02 2.97048204e-02
6.17943585e-01 -5.66097915e-01 7.38859475e-01 3.18244100e-01
1.21555686e-01 7.69537389e-01 4.10458565e-01 -6.67007446e-01
-1.13387311e+00 -9.31405485e-01 1.02007519e-02 3.34071159e-01
7.60185480e-01 1.65672034e-01 -7.08075523e-01 -3.20175707e-01
-3.51005286e-01 9.50887859e-01 -5.45245469e-01 -1.84175253e-01
-5.38935661e-01 -2.83822924e-01 1.50405198e-01 2.51553040e-02
1.17899418e+00 -1.12123466e+00 -5.32387018e-01 5.59975253e-03
-6.27774358e-01 -1.35534251e+00 -5.36404610e-01 -1.50459275e-01
-4.63544250e-01 -1.03031409e+00 -8.18545282e-01 -6.29591405e-01
4.84251797e-01 6.16954625e-01 8.81759584e-01 1.17857039e-01
-2.81945884e-01 1.00792415e-01 -3.37637335e-01 -2.33207271e-01
-5.25031269e-01 -5.46430469e-01 -2.94006079e-01 5.83218336e-01
-4.71690327e-01 -7.56123543e-01 -1.06158769e+00 2.87132293e-01
-1.26835001e+00 5.53906262e-01 7.46907592e-01 9.11844611e-01
3.03999931e-01 5.59844494e-01 1.36284664e-01 -8.53426576e-01
1.48498654e-01 -8.45761225e-02 -6.50086880e-01 2.21195057e-01
-4.62033182e-01 -3.31943333e-01 9.22455430e-01 -1.61267459e-01
-1.62505519e+00 -1.84888896e-02 -3.17472398e-01 -7.20660806e-01
1.84134543e-02 4.44938578e-02 -6.20673716e-01 -4.11383897e-01
3.42793614e-01 7.16921031e-01 -2.39508525e-01 -4.00138021e-01
2.93670744e-01 3.34511548e-01 8.43966246e-01 -2.52633691e-01
1.19492221e+00 8.59528959e-01 3.53692144e-01 -7.19946921e-01
-8.58551919e-01 -1.59362853e-01 -4.79655355e-01 -4.41197395e-01
8.92727256e-01 -8.96784008e-01 -7.64653742e-01 7.94964075e-01
-1.52041626e+00 -4.73721564e-01 -1.60032406e-01 2.19118819e-01
-6.47798717e-01 7.65307367e-01 -4.99367356e-01 -4.36249733e-01
-2.91901320e-01 -1.32491219e+00 1.12788451e+00 3.58990967e-01
5.32748044e-01 -8.70821297e-01 2.15521101e-02 6.78773344e-01
2.61054188e-01 5.72486818e-01 7.91977286e-01 4.78748143e-01
-1.04992223e+00 8.86601955e-02 -5.24566472e-01 6.64374709e-01
1.87950253e-01 -2.51914769e-01 -1.12582409e+00 -4.17291611e-01
2.60032237e-01 -3.40600729e-01 8.42785060e-01 3.29706997e-01
1.46870553e+00 -4.01932508e-01 -1.86227225e-02 1.14735985e+00
1.81653464e+00 2.39726007e-01 1.32373178e+00 7.20538348e-02
1.03083849e+00 4.55250114e-01 5.27809799e-01 4.61375862e-01
2.60095626e-01 7.80256391e-01 9.88320231e-01 -5.28129399e-01
-5.70241809e-01 -4.59609270e-01 2.86496937e-01 5.58912158e-01
-6.57441393e-02 -5.65958679e-01 -5.12666583e-01 1.51532650e-01
-1.47700346e+00 -9.15935993e-01 -1.85770601e-01 2.04642391e+00
6.05798960e-01 -8.45300555e-02 -4.67317969e-01 5.27037643e-02
6.04794323e-01 5.96003413e-01 -7.49200583e-01 -2.55830169e-01
-2.50859410e-01 8.65196250e-03 4.70321953e-01 6.79605484e-01
-8.38918984e-01 9.26451206e-01 5.36149263e+00 9.67277169e-01
-1.16065681e+00 -2.54705525e-03 7.55771279e-01 3.57556850e-01
-4.88599002e-01 2.17505828e-01 -2.33733490e-01 6.58272922e-01
2.61506706e-01 5.24974093e-02 8.25787067e-01 3.52710992e-01
5.01823425e-01 -3.45260441e-01 -7.94759929e-01 1.12908208e+00
3.46583456e-01 -1.21164358e+00 1.90659776e-01 5.79987690e-02
9.77381349e-01 -3.21778029e-01 1.46605656e-01 -3.27775441e-02
1.38607472e-01 -8.00652742e-01 5.15506804e-01 7.93684781e-01
9.80739355e-01 -8.88250172e-01 5.36183476e-01 2.53853709e-01
-9.51165974e-01 7.12745031e-03 -4.28599179e-01 3.07766289e-01
3.44516903e-01 6.99383140e-01 -5.28535545e-01 8.21092308e-01
6.75169468e-01 8.64673138e-01 -3.69051427e-01 9.08373833e-01
-3.93366933e-01 5.07882893e-01 1.20899141e-01 5.05107224e-01
-2.94954721e-02 -4.72164690e-01 8.28772545e-01 7.65057623e-01
2.73836941e-01 2.58556366e-01 2.69924905e-02 9.15713906e-01
-3.38064641e-01 -1.54147804e-01 -6.27367139e-01 5.71529686e-01
1.15186483e-01 1.26588249e+00 -6.05848610e-01 -2.17314810e-01
-3.16832066e-01 1.51924908e+00 -3.63952629e-02 6.30989432e-01
-9.53119695e-01 -3.78132135e-01 2.97928035e-01 2.21003756e-01
1.87764421e-01 -3.06900609e-02 -1.72347650e-02 -1.29406917e+00
-1.19186868e-03 -9.13086236e-01 -2.45420799e-01 -1.63720942e+00
-9.12272811e-01 5.85371196e-01 -2.27490336e-01 -1.43205070e+00
1.16027907e-01 -3.45123678e-01 -6.40101075e-01 8.51892173e-01
-1.88625979e+00 -1.47415268e+00 -7.47566283e-01 7.78981209e-01
7.49430001e-01 5.38077317e-02 2.36398041e-01 4.31094021e-01
-4.85522062e-01 4.05508369e-01 4.09810662e-01 -4.30863276e-02
8.52119148e-01 -9.55764949e-01 1.11651637e-01 1.17298925e+00
-2.11666718e-01 -3.25372703e-02 8.74842227e-01 -5.93548775e-01
-1.55742455e+00 -1.28213537e+00 3.79117578e-01 5.13910018e-02
2.71736324e-01 -2.20400289e-01 -7.11761951e-01 5.44292808e-01
6.85557365e-01 -1.24739772e-02 4.65760343e-02 -5.63745916e-01
-2.19884306e-01 -5.59045315e-01 -1.14305484e+00 6.46396995e-01
7.39243507e-01 -3.60940129e-01 -1.29005060e-01 3.73882115e-01
8.76335800e-01 -7.23364949e-01 -6.56756461e-01 3.38319004e-01
5.09496868e-01 -1.34637725e+00 1.14069772e+00 1.44049991e-02
1.07975328e+00 -4.79455501e-01 -3.01703095e-01 -1.46913481e+00
-1.64551780e-01 -8.46049488e-01 -6.57408237e-02 1.46454608e+00
-4.07347888e-01 -4.92556244e-01 5.39002657e-01 2.89500982e-01
-2.99978942e-01 -6.01656497e-01 -7.30700254e-01 -4.50210243e-01
-2.62414545e-01 -2.10832164e-01 7.42967308e-01 1.02630663e+00
-7.98586547e-01 2.82474697e-01 -1.13413250e+00 3.93281072e-01
9.55019176e-01 3.46344233e-01 1.10346150e+00 -6.34288967e-01
-4.54913646e-01 8.01435672e-03 -2.57166445e-01 -1.40591109e+00
9.36684310e-02 -4.87462938e-01 1.81474030e-01 -1.53076410e+00
2.71866024e-01 -1.71148360e-01 1.78823806e-02 3.28492448e-02
-1.74926147e-01 3.43142152e-01 2.18556583e-01 8.49387869e-02
-2.90362924e-01 9.58871067e-01 1.76224935e+00 -2.83923745e-01
-1.10385725e-02 -1.32164538e-01 -5.33379376e-01 6.84454262e-01
4.58145976e-01 -1.59657583e-01 -6.39245212e-01 -7.70953894e-01
1.62414819e-01 5.89820504e-01 7.80794859e-01 -1.02191472e+00
1.92089006e-02 -3.13172907e-01 5.71692228e-01 -7.49285281e-01
6.47014499e-01 -1.14422059e+00 3.67685795e-01 4.09160525e-01
-1.47603348e-01 -1.78049818e-01 1.27990559e-01 7.36610949e-01
-1.70216680e-01 -5.95538840e-02 1.09990060e+00 5.30518629e-02
-7.07069933e-01 3.73807788e-01 -8.23783409e-03 -6.14465848e-02
1.01986051e+00 -1.27509996e-01 -4.13237542e-01 -7.09957480e-01
-2.32001811e-01 4.00679335e-02 5.61424017e-01 2.60828346e-01
8.71090651e-01 -1.34116018e+00 -7.81292737e-01 2.15396240e-01
-1.56479523e-01 1.07372910e-01 7.04848945e-01 5.34657121e-01
-9.34758425e-01 1.22318435e-02 -3.94538432e-01 -3.57932448e-01
-1.16718733e+00 9.12834167e-01 3.54860961e-01 -2.53133774e-01
-7.77757227e-01 6.01557732e-01 7.47889221e-01 -1.14365458e-01
5.73729910e-02 -6.45458326e-02 -4.46392521e-02 -3.40301812e-01
3.93740952e-01 3.95801067e-01 -1.70834526e-01 -7.09516466e-01
1.50585040e-01 5.98791361e-01 4.74970490e-02 9.75599885e-02
1.27790213e+00 -6.22520685e-01 -1.99082196e-01 1.39700860e-01
1.30612791e+00 -1.13479696e-01 -1.55640721e+00 -3.29464078e-01
-9.56136405e-01 -9.35417533e-01 3.82505447e-01 -7.30577230e-01
-1.46884215e+00 1.09728110e+00 6.67740345e-01 -8.69776309e-02
1.63289237e+00 -5.90097666e-01 1.21796668e+00 1.80414155e-01
3.46291482e-01 -7.99664915e-01 2.83387840e-01 1.09721489e-01
1.18002057e+00 -1.17664385e+00 1.99209139e-01 -4.86810029e-01
-5.38865268e-01 9.17962074e-01 6.10013127e-01 -2.15633661e-01
7.22121894e-01 5.17456681e-02 -3.31094712e-02 -3.70637607e-03
-4.45521593e-01 1.72888950e-01 2.04979956e-01 5.25002420e-01
-8.71929154e-02 -2.36260965e-01 -1.78421196e-02 5.21400385e-02
-7.05747455e-02 3.04029882e-02 7.88506508e-01 3.69356215e-01
-1.43928081e-01 -7.89701104e-01 -5.12433827e-01 1.59325853e-01
-2.40961939e-01 -2.04686075e-01 -3.23533773e-01 6.72660172e-01
3.29840630e-01 8.99675429e-01 -1.82783782e-01 -3.95363688e-01
2.24153355e-01 -7.27747023e-01 5.05012274e-01 -1.31078407e-01
-1.49112850e-01 6.66098669e-02 -8.76977742e-02 -6.96870983e-01
-5.46940088e-01 -3.48206997e-01 -6.61326945e-01 -6.35363281e-01
-9.43263397e-02 -3.36120486e-01 5.67022264e-01 8.21692109e-01
5.37718870e-02 6.52403355e-01 1.22087741e+00 -1.13812399e+00
-1.34047568e-01 -7.44350493e-01 -6.97953701e-01 4.84377682e-01
6.53460503e-01 -6.01927698e-01 -2.80020565e-01 3.69719476e-01] | [10.567606925964355, -2.385394334793091] |
dc88410d-30eb-456c-9faf-f885126481f4 | formality-style-transfer-with-hybrid-textual | 1903.06353 | null | http://arxiv.org/abs/1903.06353v1 | http://arxiv.org/pdf/1903.06353v1.pdf | Formality Style Transfer with Hybrid Textual Annotations | Formality style transformation is the task of modifying the formality of a
given sentence without changing its content. Its challenge is the lack of
large-scale sentence-aligned parallel data. In this paper, we propose an
omnivorous model that takes parallel data and formality-classified data jointly
to alleviate the data sparsity issue. We empirically demonstrate the
effectiveness of our approach by achieving the state-of-art performance on a
recently proposed benchmark dataset of formality transfer. Furthermore, our
model can be readily adapted to other unsupervised text style transfer tasks
like unsupervised sentiment transfer and achieve competitive results on three
widely recognized benchmarks. | ['Furu Wei', 'Tao Ge', 'Ruochen Xu'] | 2019-03-15 | null | null | null | null | ['formality-style-transfer'] | ['natural-language-processing'] | [ 6.57257497e-01 2.33276337e-01 -1.87894508e-01 -6.71268344e-01
-1.13351929e+00 -6.89171314e-01 6.34438396e-01 7.15193525e-03
-6.85618937e-01 9.34808314e-01 5.46026170e-01 -2.49994159e-01
3.12990457e-01 -4.01098311e-01 -8.42322171e-01 -3.50418657e-01
4.93691176e-01 5.31265497e-01 1.19992964e-01 -6.54595733e-01
3.37445349e-01 -2.59606410e-02 -9.69278634e-01 4.89714712e-01
1.11335385e+00 4.43234116e-01 1.21123865e-01 4.64033067e-01
-1.37128472e-01 5.78740895e-01 -5.84848642e-01 -7.22747803e-01
1.61276445e-01 -4.05945688e-01 -1.24429071e+00 3.40236276e-02
8.01324189e-01 -6.53256327e-02 -1.38640821e-01 1.04487705e+00
3.58652472e-01 1.13017866e-02 7.42599845e-01 -1.06580555e+00
-1.00603592e+00 7.66953707e-01 -5.15093505e-01 3.47247928e-01
2.10594341e-01 -9.34735686e-02 1.43142211e+00 -9.64494526e-01
6.84851944e-01 1.22833550e+00 5.37472665e-01 7.99569547e-01
-1.39372599e+00 -5.73730588e-01 3.80597174e-01 6.74200803e-02
-1.02356458e+00 -3.44936997e-01 9.09293413e-01 -3.26684445e-01
1.02478480e+00 4.16893438e-02 3.31197619e-01 1.07794356e+00
2.70350486e-01 8.47739816e-01 1.26516080e+00 -5.47598183e-01
-1.44368365e-01 3.31353559e-03 1.58351585e-01 6.27720594e-01
-6.22212850e-02 -3.55603397e-01 -4.71597970e-01 5.23959212e-02
2.03759849e-01 -1.62038118e-01 -2.33379275e-01 -2.09221050e-01
-1.44340086e+00 9.02650297e-01 3.62206489e-01 3.32313180e-01
1.30429015e-01 1.75748020e-01 8.50053132e-01 6.39397383e-01
7.70577013e-01 8.25576842e-01 -9.23465848e-01 -2.06312478e-01
-7.50212789e-01 1.86351523e-01 6.98470831e-01 1.06300330e+00
6.90243900e-01 -6.51775002e-02 -3.36720794e-01 8.60816240e-01
-6.68942854e-02 6.01302564e-01 4.75694686e-01 -7.16402769e-01
9.40971255e-01 5.25709569e-01 -1.45808384e-02 -6.56743467e-01
-2.72402853e-01 -3.43022466e-01 -8.68254423e-01 -3.88594955e-01
1.98211357e-01 -1.79414779e-01 -5.22363067e-01 1.89833450e+00
1.93941921e-01 -2.26171806e-01 2.43229464e-01 5.65977931e-01
9.02297735e-01 7.56028354e-01 7.19481781e-02 3.90094407e-02
1.23875785e+00 -1.38270533e+00 -6.66439652e-01 -3.34818929e-01
9.02028978e-01 -8.51198256e-01 1.75819862e+00 2.47146174e-01
-9.94274974e-01 -4.84650612e-01 -1.00240409e+00 -4.19624090e-01
-2.70242840e-01 -2.57521849e-02 5.67200959e-01 4.21415746e-01
-9.84555304e-01 5.02573133e-01 -6.63994253e-01 -4.18452770e-01
3.03031176e-01 2.85174519e-01 -4.41452205e-01 -8.02771673e-02
-1.40335500e+00 9.12258387e-01 3.82577717e-01 -9.26030725e-02
-4.64175433e-01 -1.01367962e+00 -1.03111565e+00 -3.82337645e-02
2.93926477e-01 -8.88076305e-01 1.36480856e+00 -1.21966255e+00
-1.66855764e+00 9.96236861e-01 -2.45517582e-01 -3.56765419e-01
3.61186922e-01 -3.67800564e-01 -3.11318249e-01 -7.89949521e-02
2.73402661e-01 9.47819591e-01 8.25768530e-01 -9.94094491e-01
-6.21305466e-01 -8.77076685e-02 3.86157751e-01 3.18364054e-01
-9.99559939e-01 7.76432157e-02 -3.08226138e-01 -1.01756120e+00
-3.01694185e-01 -9.57192838e-01 -2.59159893e-01 -3.12009871e-01
-2.64774740e-01 -4.16520029e-01 5.20915687e-01 -5.42715132e-01
1.14042413e+00 -2.00116587e+00 5.85458994e-01 -3.21913868e-01
1.47815764e-01 2.48435020e-01 -5.36581576e-01 6.31737173e-01
1.79819297e-02 1.75774261e-01 -4.35477227e-01 -4.10475284e-01
1.83911249e-01 2.04087436e-01 -5.83143294e-01 1.65625423e-01
4.09728914e-01 1.19527853e+00 -9.34968591e-01 -3.67494464e-01
-2.17982978e-01 1.93886727e-01 -9.17765379e-01 1.47762105e-01
-2.14472666e-01 4.84842330e-01 -2.83096373e-01 8.01199675e-03
5.22550464e-01 -2.95531660e-01 1.69039696e-01 -2.86950946e-01
1.48534521e-01 7.74536848e-01 -5.55766225e-01 2.01686668e+00
-7.49032199e-01 6.54117167e-01 -2.81829834e-01 -9.46069777e-01
8.68940711e-01 2.34012544e-01 2.02464014e-01 -6.73541546e-01
-4.40890566e-02 1.56168446e-01 -9.07079224e-03 -2.54116267e-01
8.47598433e-01 -4.42641228e-01 -5.48550308e-01 6.91468894e-01
2.10644513e-01 -4.47112888e-01 4.64949280e-01 3.01916540e-01
9.96306777e-01 1.25532135e-01 2.51864612e-01 -7.64858484e-01
7.59404302e-01 -5.50056808e-02 5.93109727e-01 4.22817528e-01
-7.61981606e-02 4.34620500e-01 5.37234604e-01 -3.86321157e-01
-1.07873392e+00 -1.08917356e+00 -1.47733204e-02 1.32511580e+00
-1.03496537e-01 -7.69342363e-01 -7.98712850e-01 -1.26080990e+00
-1.94443554e-01 5.80139160e-01 -8.10706973e-01 -1.83035240e-01
-8.33885193e-01 -4.62013811e-01 6.18019938e-01 4.83978868e-01
4.18003112e-01 -1.17838323e+00 1.65288541e-02 5.63266575e-02
-5.79401851e-01 -1.20743799e+00 -1.09644902e+00 -2.95619778e-02
-8.46486509e-01 -7.09357023e-01 -5.51985204e-01 -1.11930501e+00
9.26793456e-01 3.37780684e-01 1.40686381e+00 2.34835176e-03
2.42002472e-01 -8.92886147e-03 -3.95029813e-01 -3.94979268e-01
-5.66794276e-01 7.28900433e-01 2.88281459e-02 -1.57256722e-01
2.50927538e-01 -4.39894527e-01 -3.25325668e-01 2.44169250e-01
-1.15476716e+00 2.63361275e-01 4.07836765e-01 1.07409894e+00
4.92894202e-01 -2.24967986e-01 9.41181064e-01 -1.50209987e+00
6.73522234e-01 -6.40673563e-02 -3.66309494e-01 2.70472705e-01
-6.53971910e-01 2.21305415e-01 1.05149317e+00 -3.18134993e-01
-1.06528842e+00 -1.79665819e-01 -1.65620863e-01 1.12442516e-01
4.94992882e-02 6.37036741e-01 -2.45076641e-01 2.28666455e-01
5.58941722e-01 2.56180912e-01 -2.29996413e-01 -3.61069024e-01
5.80077112e-01 6.22438610e-01 5.07826030e-01 -8.58967900e-01
1.14865232e+00 4.56643224e-01 -2.34752566e-01 -6.05686188e-01
-1.51308560e+00 -3.17543328e-01 -9.21232879e-01 2.27591991e-01
5.81150293e-01 -9.73453104e-01 -1.18688129e-01 3.16322893e-01
-1.36905849e+00 -4.70335007e-01 -3.70059729e-01 6.98016658e-02
-4.95754004e-01 5.89697599e-01 -6.72098458e-01 1.51066840e-01
-7.44051039e-01 -8.70958269e-01 1.06045878e+00 -9.70490128e-02
-4.94175136e-01 -1.41980863e+00 4.20983642e-01 6.67237461e-01
3.42138708e-01 -1.87101454e-01 9.55131769e-01 -4.50350374e-01
-1.27127066e-01 -2.55826451e-02 -1.34503111e-01 6.57243669e-01
5.13373077e-01 -3.72256041e-02 -5.74847996e-01 -5.18804967e-01
-2.36345768e-01 -7.35357702e-01 8.80221963e-01 -1.65876642e-01
1.07267368e+00 -5.06053865e-01 8.54391232e-02 5.80679059e-01
1.13317502e+00 -4.24393266e-01 4.96363074e-01 3.57483566e-01
9.05691922e-01 6.80087566e-01 7.93171346e-01 1.19107530e-01
7.90991962e-01 6.11437917e-01 -6.78363144e-02 -3.62337142e-01
-2.26942897e-01 -4.07324642e-01 5.74974179e-01 1.44244146e+00
3.85213614e-01 -2.11240917e-01 -8.32892478e-01 7.15145826e-01
-1.81125331e+00 -9.01567757e-01 -8.54715034e-02 1.59659064e+00
1.31055987e+00 3.40259671e-01 -5.29102944e-02 -5.64775839e-02
6.38009131e-01 2.24853799e-01 -2.48770878e-01 -7.39495933e-01
-8.96856710e-02 3.26960027e-01 2.37436920e-01 6.46975279e-01
-1.24852276e+00 1.35431468e+00 6.69947100e+00 8.06394279e-01
-1.05955017e+00 3.02881241e-01 4.15863603e-01 -1.87579505e-02
-8.32991064e-01 -8.19722787e-02 -9.68873322e-01 2.93549150e-01
7.51249015e-01 -3.24603796e-01 4.24810499e-01 4.93381500e-01
1.22486718e-01 3.32367569e-01 -1.37287974e+00 6.43153906e-01
4.66440201e-01 -1.28932047e+00 2.97561198e-01 -2.85305887e-01
1.21577108e+00 1.20410062e-01 2.39309698e-01 5.94928980e-01
3.01539153e-01 -8.38324904e-01 7.80951619e-01 6.00998178e-02
1.04242039e+00 -6.49054348e-01 6.31942153e-01 3.01981598e-01
-1.02772987e+00 2.51926661e-01 -2.50900298e-01 -3.75039756e-01
1.89142540e-01 5.05409956e-01 -7.20059037e-01 6.45226359e-01
5.62870145e-01 1.12556338e+00 -6.82237864e-01 2.77674288e-01
-5.69370151e-01 6.34081483e-01 -1.19498901e-01 -1.91998184e-01
4.12915677e-01 4.90737334e-03 3.28565180e-01 1.43450367e+00
-1.20466746e-01 -2.09300905e-01 2.08036497e-01 7.01481700e-01
-4.75999564e-01 3.56654435e-01 -6.43382132e-01 -1.72131490e-02
1.75641999e-01 1.38303614e+00 -4.50399727e-01 -3.16228747e-01
-5.93154430e-01 1.04106700e+00 9.38858330e-01 2.20297158e-01
-7.68332362e-01 -4.50501978e-01 5.11319399e-01 -1.76865831e-01
4.69596475e-01 -2.07344070e-01 -4.80223745e-01 -1.53772426e+00
3.46693277e-01 -1.12648857e+00 5.01086116e-01 -5.62887132e-01
-1.73294389e+00 6.78652346e-01 -1.10390559e-01 -1.27949345e+00
3.68270352e-02 -5.69985986e-01 -7.16354430e-01 7.84617484e-01
-1.88548613e+00 -1.26710141e+00 -5.70144095e-02 5.40586531e-01
7.24307001e-01 -2.70256460e-01 9.28092360e-01 2.23471463e-01
-5.05700946e-01 9.56585646e-01 1.42416105e-01 2.28103325e-01
1.21271896e+00 -1.38311923e+00 7.79733479e-01 9.93675470e-01
2.29145542e-01 6.90219522e-01 6.54195189e-01 -5.26497722e-01
-1.14357960e+00 -1.50970864e+00 1.39158571e+00 -7.22565413e-01
9.66405153e-01 -7.41657734e-01 -7.86402643e-01 8.10812116e-01
4.91131097e-01 -1.60901532e-01 8.81038964e-01 4.28409278e-01
-6.89076722e-01 -2.72534639e-01 -8.53077173e-01 8.96013737e-01
1.14816022e+00 -5.20951331e-01 -1.11068869e+00 4.21273887e-01
9.24936950e-01 -4.71636176e-01 -9.71849859e-01 5.35604239e-01
1.96715191e-01 -4.70768124e-01 5.16188204e-01 -9.65168059e-01
1.01560533e+00 -7.53495768e-02 -1.06787987e-01 -1.61855090e+00
-4.61086929e-01 -7.35864520e-01 1.79047078e-01 1.51089919e+00
7.38577664e-01 -6.19843006e-01 6.99516177e-01 3.19219321e-01
-1.96541727e-01 -6.10456526e-01 -9.11025286e-01 -8.02259445e-01
5.48063815e-01 -3.36147659e-02 5.04346788e-01 1.11306465e+00
2.26834044e-01 1.10621965e+00 -3.11680228e-01 -2.44565934e-01
4.12692755e-01 3.65395397e-01 1.02859163e+00 -8.08094501e-01
-4.74089295e-01 -3.97191435e-01 2.44498346e-03 -1.26832926e+00
8.06918740e-01 -1.46913838e+00 5.25011383e-02 -1.52441776e+00
5.39033473e-01 -2.69776076e-01 -2.72196382e-01 6.18554652e-01
-4.74858195e-01 5.29199064e-01 2.48344272e-01 5.86980842e-02
-9.38700140e-01 1.03302622e+00 1.58345628e+00 -4.60404217e-01
-1.64820701e-02 -3.85960728e-01 -9.85133827e-01 5.94432771e-01
8.87189925e-01 -6.26378179e-01 -5.31788766e-01 -9.37109649e-01
3.73152405e-01 -3.68266791e-01 -2.76928898e-02 -4.70495760e-01
-2.25980833e-01 -1.60603121e-01 -9.43799838e-02 -3.58101577e-01
-4.72449549e-02 -5.73706567e-01 -4.66580033e-01 4.41942096e-01
-6.63575590e-01 3.09304416e-01 2.50775218e-01 3.29495519e-01
-3.25667500e-01 -1.16278619e-01 7.49145150e-01 2.24974141e-01
-4.01866198e-01 3.24652553e-01 -2.16169268e-01 7.18084157e-01
7.04956710e-01 2.28176966e-01 -4.94695991e-01 -1.87325031e-01
-1.94460437e-01 3.30338895e-01 5.85555315e-01 7.84271181e-01
3.47817868e-01 -1.48033249e+00 -1.04680967e+00 1.07181780e-01
5.05135238e-01 -6.90654144e-02 1.12920158e-01 6.76351607e-01
-3.93352240e-01 5.80479443e-01 -2.31491521e-01 -5.84467351e-01
-1.09164548e+00 5.55969357e-01 9.44276303e-02 -6.65915847e-01
-6.48054063e-01 6.21750057e-01 5.63191116e-01 -9.79766667e-01
-1.43107280e-01 -3.15472782e-01 -1.09941036e-01 -3.27799767e-01
4.57459033e-01 -7.31417444e-03 2.15230390e-01 -5.55992305e-01
-2.88758546e-01 5.72429359e-01 -5.63154459e-01 -1.69293568e-01
1.35394728e+00 -2.01941937e-01 -4.39571887e-01 6.32140398e-01
1.37813842e+00 2.36802064e-02 -1.24775398e+00 -6.10416532e-01
-2.74165832e-02 -2.62569189e-01 -3.48647833e-01 -6.66698158e-01
-8.58558774e-01 8.15050125e-01 -1.23704523e-01 -1.47822842e-01
1.01624417e+00 1.04317054e-01 1.01545799e+00 7.65253007e-01
9.79530141e-02 -1.20041633e+00 2.90044278e-01 1.08218861e+00
1.13537967e+00 -1.29075575e+00 -1.08370371e-01 -5.88674903e-01
-7.39374280e-01 8.43873799e-01 7.42964208e-01 -2.85818517e-01
4.78200555e-01 1.66187659e-01 2.24973097e-01 2.20970977e-02
-9.65768635e-01 7.65917748e-02 4.95342940e-01 4.35644090e-01
7.56581843e-01 1.39970761e-02 -6.15806520e-01 6.26089275e-01
-5.21488667e-01 -2.48537928e-01 5.75928211e-01 7.25491405e-01
-5.13651855e-02 -1.42642093e+00 2.61581223e-02 3.17233801e-01
-6.07973993e-01 -4.51802701e-01 -4.53233302e-01 5.30799568e-01
-3.46260875e-01 8.48688185e-01 -6.26331419e-02 -1.22461446e-01
5.38548768e-01 3.93871553e-02 7.78024614e-01 -1.08857465e+00
-7.61654973e-01 -6.99569359e-02 2.13748842e-01 -2.91694999e-01
-5.15385687e-01 -7.29321897e-01 -1.25602114e+00 -3.78315210e-01
1.15941979e-01 2.36014336e-01 3.96631509e-01 1.18551707e+00
3.15351903e-01 6.64354384e-01 8.47630858e-01 -7.27811873e-01
-7.96724856e-01 -1.08962297e+00 -2.52278388e-01 9.10006881e-01
2.25712836e-01 -3.15699369e-01 -1.50495932e-01 3.04386020e-01] | [11.572406768798828, 9.518424034118652] |
9c8bdb26-592c-4422-a1f4-cb2c6bde707b | paragraph-based-transformer-pre-training-for | 2205.01228 | null | https://arxiv.org/abs/2205.01228v2 | https://arxiv.org/pdf/2205.01228v2.pdf | Paragraph-based Transformer Pre-training for Multi-Sentence Inference | Inference tasks such as answer sentence selection (AS2) or fact verification are typically solved by fine-tuning transformer-based models as individual sentence-pair classifiers. Recent studies show that these tasks benefit from modeling dependencies across multiple candidate sentences jointly. In this paper, we first show that popular pre-trained transformers perform poorly when used for fine-tuning on multi-candidate inference tasks. We then propose a new pre-training objective that models the paragraph-level semantics across multiple input sentences. Our evaluation on three AS2 and one fact verification datasets demonstrates the superiority of our pre-training technique over the traditional ones for transformers used as joint models for multi-candidate inference tasks, as well as when used as cross-encoders for sentence-pair formulations of these tasks. Our code and pre-trained models are released at https://github.com/amazon-research/wqa-multi-sentence-inference . | ['Alessandro Moschitti', 'Luca Soldaini', 'Siddhant Garg', 'Luca Di Liello'] | 2022-05-02 | null | https://aclanthology.org/2022.naacl-main.181 | https://aclanthology.org/2022.naacl-main.181.pdf | naacl-2022-7 | ['answer-selection'] | ['natural-language-processing'] | [ 1.67874470e-01 8.35465044e-02 -2.44430497e-01 -8.10804963e-01
-1.76637220e+00 -6.09193027e-01 6.54703915e-01 3.93244892e-01
-3.08478385e-01 9.43125367e-01 3.17940742e-01 -6.08445704e-01
1.42153148e-02 -7.79888093e-01 -1.00404191e+00 -6.49238378e-02
3.07583272e-01 6.99819386e-01 4.32588696e-01 -3.29989731e-01
3.03018600e-01 -1.66545317e-01 -1.48094571e+00 9.41262007e-01
9.41333592e-01 9.09389079e-01 2.46504322e-01 1.00238502e+00
-2.47556463e-01 1.07601476e+00 -6.75217569e-01 -1.36637580e+00
-1.74005136e-01 -1.46945328e-01 -1.09831023e+00 -3.37738514e-01
1.04897785e+00 -2.90097177e-01 8.23202878e-02 8.17347765e-01
5.00241518e-01 -3.49363200e-02 4.36540514e-01 -9.56395745e-01
-6.38079762e-01 1.23072505e+00 -1.04772195e-01 2.90535241e-01
5.74543357e-01 -4.97696772e-02 1.79486930e+00 -1.05780649e+00
4.98614520e-01 1.39150286e+00 8.05141866e-01 5.50063014e-01
-1.22455430e+00 -4.73966748e-01 5.51461168e-02 6.19697690e-01
-1.11820781e+00 -7.15732992e-01 5.15601039e-01 -2.05499858e-01
1.41838074e+00 5.26230276e-01 -3.04386187e-02 1.29119456e+00
4.10874456e-01 1.29930556e+00 7.26669788e-01 -3.89977664e-01
-1.83363795e-01 2.23800674e-01 4.46568400e-01 8.65473747e-01
1.79527458e-02 -4.94291633e-01 -8.81741703e-01 -2.44372189e-01
9.56516489e-02 -5.15418708e-01 -9.52239558e-02 2.04067305e-01
-1.11328936e+00 9.80028689e-01 -3.01328972e-02 3.58941495e-01
-1.20803572e-01 2.80702651e-01 7.55354226e-01 6.28646255e-01
6.83472753e-01 6.39662206e-01 -1.14932537e+00 -3.09465617e-01
-1.07585204e+00 4.64237213e-01 7.86026180e-01 9.79864895e-01
5.17926633e-01 -4.05148774e-01 -9.61222231e-01 9.44706917e-01
5.14856540e-02 4.26145017e-01 3.78034621e-01 -8.05814862e-01
1.04226911e+00 3.24027538e-01 1.58615578e-02 -4.99093145e-01
-9.09809247e-02 -5.86375892e-01 -5.44199944e-01 -5.22809029e-01
4.71268147e-01 -2.18505226e-02 -5.71498632e-01 1.81988561e+00
2.81230032e-01 3.38437818e-02 -8.76319557e-02 4.09326315e-01
1.05004060e+00 3.49080384e-01 6.27883002e-02 -4.43160534e-04
1.62151241e+00 -1.24447215e+00 -8.92658293e-01 -1.99431151e-01
9.01862621e-01 -9.95367229e-01 1.34492159e+00 3.62831771e-01
-1.50594020e+00 -7.53011823e-01 -9.50645506e-01 -6.45711958e-01
-3.00017208e-01 5.15591145e-01 5.13504088e-01 5.37332892e-01
-1.09621501e+00 6.00509465e-01 -7.33379066e-01 -4.19322215e-02
3.57469767e-01 2.42355466e-01 -8.26415196e-02 6.93173781e-02
-1.58379948e+00 1.27245903e+00 3.82351838e-02 -8.48085340e-03
-8.26628089e-01 -1.04286933e+00 -8.87618780e-01 2.89592534e-01
3.26594234e-01 -1.13362491e+00 1.80580759e+00 -5.12194633e-01
-1.57991242e+00 1.05639565e+00 -7.29162991e-01 -5.74426472e-01
3.83050799e-01 -5.12737930e-01 -2.73642600e-01 5.20585515e-02
4.22697604e-01 2.34629929e-01 8.62514198e-01 -7.45332122e-01
-5.44328809e-01 -2.00584263e-01 4.81473774e-01 -1.21665202e-01
-2.23478422e-01 2.75847733e-01 -2.78937399e-01 -4.39093411e-01
-3.25806648e-01 -3.97501230e-01 1.01495720e-01 -4.62589741e-01
-9.18546259e-01 -6.98754013e-01 3.49126667e-01 -7.33627379e-01
1.31018448e+00 -1.60307944e+00 2.47814864e-01 -2.78751969e-01
3.33781280e-02 2.54340559e-01 -4.13196772e-01 5.70417583e-01
-7.34468400e-02 2.68167585e-01 -1.69960767e-01 -1.06462204e+00
3.67402911e-01 -9.32586566e-02 -4.28927243e-01 -1.09306134e-01
6.00685716e-01 1.07146227e+00 -8.82731438e-01 -6.25302136e-01
-2.30890289e-02 1.40730277e-01 -3.99851441e-01 3.40984911e-01
-5.63025653e-01 3.28251570e-01 -4.02107477e-01 4.71278787e-01
6.97228432e-01 -4.52455163e-01 2.49117300e-01 -2.95199543e-01
3.08880001e-01 1.28791821e+00 -7.30649590e-01 1.92514205e+00
-9.17040706e-01 3.83504987e-01 -6.04810715e-02 -9.81347620e-01
6.69785500e-01 4.78888482e-01 -6.01437353e-02 -6.39155030e-01
-4.50036116e-03 4.70712364e-01 -1.95208043e-01 -7.80335128e-01
6.36982262e-01 -2.65008241e-01 -1.59120977e-01 4.90385294e-01
5.38839579e-01 -4.47745711e-01 6.95993721e-01 5.50545812e-01
1.21943176e+00 9.28076506e-02 6.43624142e-02 -9.41494554e-02
1.05730832e+00 -2.77261823e-01 3.38929296e-01 8.43003452e-01
2.70678520e-01 5.48164785e-01 6.92006528e-01 -1.25796080e-01
-7.97074258e-01 -1.09731662e+00 -3.48733366e-01 1.16760635e+00
-2.53738612e-01 -9.14084256e-01 -6.20042920e-01 -1.08013546e+00
1.13757700e-01 1.30314052e+00 -5.77161372e-01 -5.54233007e-02
-5.98939180e-01 -3.69169384e-01 7.12391615e-01 6.05894029e-01
2.27381155e-01 -6.97025359e-01 -1.73818156e-01 2.35290304e-01
-7.07593560e-01 -1.35388589e+00 -3.95564824e-01 2.07448065e-01
-8.29066634e-01 -1.08215404e+00 -2.18228832e-01 -5.31566501e-01
2.48437867e-01 -2.95948386e-01 1.87374330e+00 3.07440013e-01
1.06719911e-01 2.06778631e-01 -3.69173199e-01 -2.23730206e-01
-5.19025385e-01 4.19688374e-01 -2.00249821e-01 -1.09382890e-01
5.29474437e-01 -5.48643172e-01 -1.57841533e-01 -3.92362811e-02
-5.52350402e-01 8.29361975e-02 4.80357230e-01 1.04654956e+00
2.60512620e-01 -3.86814922e-01 7.41705358e-01 -1.14906657e+00
9.32346344e-01 -3.81899178e-01 -3.46307188e-01 9.64634359e-01
-2.03077391e-01 3.33411068e-01 6.47444725e-01 -7.47894123e-02
-1.29570758e+00 -5.04003525e-01 -5.45542717e-01 -1.44211918e-01
5.44297174e-02 7.14294434e-01 -7.83373863e-02 4.72818971e-01
4.22234803e-01 6.73035532e-02 -2.39193037e-01 -4.55463558e-01
3.90801132e-01 5.78677475e-01 1.55822486e-01 -9.46260393e-01
7.74740517e-01 -9.33752880e-02 -1.39829665e-01 -2.29579270e-01
-1.59319627e+00 -3.36802751e-01 -6.32761419e-01 2.17771590e-01
6.79653585e-01 -8.83492649e-01 -7.58913875e-01 1.46046460e-01
-1.57629335e+00 -3.41724724e-01 -1.32801458e-01 2.48766601e-01
-3.71482372e-01 3.65808278e-01 -8.92945707e-01 -7.45098352e-01
-7.69284487e-01 -1.14946401e+00 1.45867860e+00 -1.49089664e-01
-3.92778873e-01 -1.31761312e+00 1.19762972e-01 1.02679622e+00
4.12678897e-01 -5.18977284e-01 9.54813540e-01 -7.18621433e-01
-5.46850979e-01 -2.79454887e-01 -4.34646420e-02 4.99753475e-01
-1.04704332e-02 2.20856145e-01 -1.08822298e+00 4.14900966e-02
9.50875804e-02 -8.18643034e-01 1.12523627e+00 2.87379056e-01
1.36096048e+00 -1.92740709e-01 -5.81952818e-02 2.51940817e-01
1.16267586e+00 -6.34388268e-01 4.20961231e-01 -1.25998147e-02
4.03014392e-01 4.75630611e-01 7.05307424e-01 2.78566301e-01
7.99636841e-01 8.01216841e-01 1.08554475e-01 3.10539097e-01
-1.66739568e-01 -2.89090633e-01 5.39016247e-01 1.02353144e+00
2.75578439e-01 -3.69731426e-01 -5.50432265e-01 6.54223442e-01
-1.97782457e+00 -1.14004886e+00 -4.98042822e-01 2.13449454e+00
1.38627577e+00 2.29301050e-01 -1.32477865e-01 9.06391516e-02
4.29638654e-01 -3.32919206e-03 -1.82250634e-01 -6.12944007e-01
-3.21120977e-01 8.61470342e-01 -6.05696402e-02 9.82724130e-01
-1.17483866e+00 8.72316062e-01 5.45586920e+00 1.27516949e+00
-7.48882949e-01 6.21905446e-01 6.76313579e-01 -8.02761391e-02
-8.00991893e-01 2.94264369e-02 -1.16473830e+00 3.33317727e-01
1.07393026e+00 -3.01853605e-02 -4.64340001e-02 4.10147399e-01
8.81478116e-02 -2.11160213e-01 -1.51921475e+00 4.86328572e-01
1.22867018e-01 -1.58514905e+00 -2.68462058e-02 -6.97181761e-01
6.50479078e-01 4.13789647e-03 3.03026978e-02 6.85325027e-01
2.59846717e-01 -8.35884929e-01 8.86075795e-01 3.41529757e-01
7.16319740e-01 -2.40937129e-01 9.97454703e-01 2.06755966e-01
-1.20335114e+00 4.58650589e-02 -2.37728447e-01 -1.12147696e-01
4.50819343e-01 1.05915201e+00 -7.63888478e-01 9.31568861e-01
4.26960886e-01 7.62590587e-01 -8.73989522e-01 6.44779801e-01
-8.33512425e-01 8.11919093e-01 -2.36372069e-01 -2.04340324e-01
3.11884116e-02 1.92429990e-01 3.18679899e-01 1.38396680e+00
1.49705440e-01 -3.96777958e-01 -2.54686683e-01 9.94986773e-01
-3.30377638e-01 -1.45893283e-02 -2.25081995e-01 1.58608258e-01
5.53887844e-01 1.30071795e+00 -7.21857771e-02 -8.17810774e-01
-4.72543806e-01 8.79352510e-01 8.69628906e-01 1.78128764e-01
-1.12427473e+00 -3.41130614e-01 4.85765547e-01 -8.51693526e-02
5.20037234e-01 4.17527929e-02 -6.47511482e-01 -1.75170600e+00
5.03941238e-01 -9.07983005e-01 4.62672383e-01 -7.46730566e-01
-1.62979901e+00 4.61019605e-01 -2.73919012e-02 -9.24099624e-01
-3.88761044e-01 -5.59057593e-01 -8.83320332e-01 9.27128136e-01
-1.75449944e+00 -1.51225281e+00 2.42647380e-01 5.75964928e-01
7.42269814e-01 1.31944954e-01 9.70406532e-01 5.57430267e-01
-4.97205138e-01 1.05910742e+00 -2.26751223e-01 1.65182725e-02
6.93755627e-01 -1.53873980e+00 4.26242173e-01 9.89433885e-01
2.78636783e-01 8.77144516e-01 6.25511289e-01 -5.02907693e-01
-1.30276024e+00 -8.70819032e-01 1.94324529e+00 -9.17965829e-01
9.65324581e-01 -6.02343023e-01 -8.24032903e-01 8.80648613e-01
5.92643917e-01 -3.03606063e-01 6.65989339e-01 8.44572663e-01
-5.72356224e-01 -2.48168647e-01 -1.01442349e+00 3.39862049e-01
7.74042070e-01 -9.98718560e-01 -7.23859072e-01 9.23454225e-01
9.42811012e-01 -5.54988027e-01 -1.04086745e+00 4.19823796e-01
3.18194181e-01 -9.18353617e-01 1.01205504e+00 -7.87308037e-01
1.06279898e+00 4.40149615e-03 -1.77109450e-01 -1.13620853e+00
-1.71692610e-01 -5.70763767e-01 -2.97871470e-01 1.47135210e+00
1.00328946e+00 -4.49191719e-01 5.39549232e-01 4.82960552e-01
-3.09981078e-01 -9.37138498e-01 -1.01276720e+00 -5.72290540e-01
2.95978874e-01 -7.09174931e-01 6.00364268e-01 6.69835567e-01
1.77118361e-01 9.35574472e-01 -1.61624730e-01 4.77421051e-03
5.99059701e-01 3.04472059e-01 6.44501328e-01 -6.52629972e-01
-5.97543895e-01 -4.86190379e-01 7.93975368e-02 -1.09729123e+00
4.62681949e-01 -1.01164031e+00 -1.91206881e-03 -1.47864389e+00
3.20935458e-01 -2.95045227e-01 -3.01048486e-03 4.93774295e-01
-7.79341340e-01 -7.24952444e-02 3.32345217e-02 -3.73836845e-01
-7.26072907e-01 7.62341738e-01 1.17581367e+00 -4.04476255e-01
5.04034102e-01 4.47063357e-01 -4.73473161e-01 1.85464516e-01
9.32285249e-01 -4.39774364e-01 -3.85032117e-01 -9.23395216e-01
5.38979888e-01 1.51796982e-01 2.82031894e-01 -6.18903160e-01
1.94980949e-01 2.23334581e-01 -5.59996553e-02 -7.54030585e-01
5.89740574e-01 -2.98944861e-01 -3.55327696e-01 2.09201485e-01
-7.66383410e-01 1.15492791e-01 1.88475847e-01 1.51164323e-01
-4.63175118e-01 -6.25727534e-01 3.16934645e-01 -2.28889033e-01
-1.77017584e-01 -1.89522523e-02 -2.38495260e-01 4.17642295e-01
4.44957435e-01 4.23775524e-01 -4.52270508e-01 -4.27368492e-01
-6.51781261e-01 4.75649416e-01 -2.01471001e-01 3.64552975e-01
7.24559963e-01 -1.01390803e+00 -1.20819211e+00 -7.85616785e-02
1.06188804e-01 -7.38424808e-02 3.76275808e-01 1.25280821e+00
-1.31364539e-01 6.80132329e-01 3.42686951e-01 -5.04373491e-01
-1.44541025e+00 2.72108734e-01 2.78952569e-01 -9.52706695e-01
-1.41396290e-02 1.52243924e+00 -2.41264731e-01 -9.29689288e-01
1.02035441e-01 -6.12397611e-01 6.57989532e-02 -6.96301386e-02
3.77596825e-01 2.28589401e-02 4.89699274e-01 -6.69706389e-02
-4.53414828e-01 3.91941726e-01 -1.96386546e-01 -3.37467603e-02
1.19981706e+00 -8.48120302e-02 -4.48971391e-01 5.18466234e-01
1.24816000e+00 7.59244189e-02 -4.59966123e-01 -4.45346415e-01
1.71824336e-01 -2.79410332e-01 -1.87086999e-01 -1.03740335e+00
-9.44311440e-01 1.20943618e+00 -2.92119503e-01 1.23740822e-01
9.33451772e-01 8.65551531e-02 9.01402116e-01 5.17943382e-01
1.85349122e-01 -8.30093205e-01 8.23912844e-02 7.94942021e-01
1.05098188e+00 -1.35971117e+00 -5.33429906e-02 -5.28797984e-01
-6.41193449e-01 1.25486720e+00 7.09557056e-01 7.05985129e-02
5.57123005e-01 3.52801621e-01 -1.82474017e-01 -2.63739794e-01
-1.53669059e+00 -1.30187258e-01 4.00346726e-01 7.94486627e-02
1.04573381e+00 2.83427034e-02 -3.41940910e-01 6.77323520e-01
-4.44347948e-01 1.27468899e-01 3.33740383e-01 6.69609964e-01
1.69982985e-02 -1.51853585e+00 3.10847908e-02 7.05376267e-01
-6.07766151e-01 -6.33095026e-01 -3.08511198e-01 2.62531638e-01
-3.14549319e-02 1.28884113e+00 -6.58650622e-02 -5.75526536e-01
1.34356633e-01 2.79123902e-01 9.09765661e-01 -7.90442288e-01
-1.27572143e+00 -5.84480882e-01 9.76204753e-01 -4.99706745e-01
-4.56147999e-01 -7.00171590e-01 -8.52776408e-01 -4.53302026e-01
-5.22980213e-01 3.70609969e-01 4.34725225e-01 1.21527958e+00
3.77303928e-01 6.82032466e-01 3.51972133e-01 -3.13739300e-01
-1.03700066e+00 -1.41636336e+00 -8.53656381e-02 4.30981725e-01
3.35327148e-01 -4.15304065e-01 -3.31297398e-01 -2.66557448e-02] | [11.091841697692871, 8.431422233581543] |
c8b05bf1-8ff8-4c2c-abb2-48d1cb00b703 | differentiable-frequency-based | 2209.09194 | null | https://arxiv.org/abs/2209.09194v2 | https://arxiv.org/pdf/2209.09194v2.pdf | Differentiable Frequency-based Disentanglement for Aerial Video Action Recognition | We present a learning algorithm for human activity recognition in videos. Our approach is designed for UAV videos, which are mainly acquired from obliquely placed dynamic cameras that contain a human actor along with background motion. Typically, the human actors occupy less than one-tenth of the spatial resolution. Our approach simultaneously harnesses the benefits of frequency domain representations, a classical analysis tool in signal processing, and data driven neural networks. We build a differentiable static-dynamic frequency mask prior to model the salient static and dynamic pixels in the video, crucial for the underlying task of action recognition. We use this differentiable mask prior to enable the neural network to intrinsically learn disentangled feature representations via an identity loss function. Our formulation empowers the network to inherently compute disentangled salient features within its layers. Further, we propose a cost-function encapsulating temporal relevance and spatial content to sample the most important frame within uniformly spaced video segments. We conduct extensive experiments on the UAV Human dataset and the NEC Drone dataset and demonstrate relative improvements of 5.72% - 13.00% over the state-of-the-art and 14.28% - 38.05% over the corresponding baseline model. | ['Dinesh Manocha', 'Ming Lin', 'Divya Kothandaraman'] | 2022-09-15 | null | null | null | null | ['activity-recognition-in-videos'] | ['computer-vision'] | [ 6.40719235e-01 -1.58950850e-01 -1.93725362e-01 -1.62069857e-01
-5.29153764e-01 -4.65099514e-01 5.70214987e-01 -4.77687240e-01
-5.24020433e-01 5.02439678e-01 2.55655617e-01 3.80211949e-01
-1.95723951e-01 -3.34432483e-01 -8.60377491e-01 -9.04078066e-01
-6.05508685e-01 -4.26862895e-01 -1.08093150e-01 1.65699378e-01
-2.00371947e-02 3.88313502e-01 -1.57852876e+00 3.80808711e-01
3.75704199e-01 1.24223053e+00 7.39784539e-02 9.26215231e-01
8.72801542e-01 1.23541331e+00 -4.57482904e-01 1.44531384e-01
5.79553902e-01 -2.78798789e-01 -4.90068108e-01 3.72184962e-01
6.22083485e-01 -6.65514588e-01 -9.72039223e-01 8.65446031e-01
2.09125504e-01 2.50740379e-01 5.87078214e-01 -1.26021445e+00
-3.53281796e-01 2.48054981e-01 -9.30010676e-01 6.93813205e-01
2.86190510e-01 3.42444509e-01 1.02366054e+00 -9.51668441e-01
4.23083574e-01 8.40366900e-01 2.93224066e-01 4.86208647e-01
-1.12744153e+00 -7.12111473e-01 2.94049650e-01 2.29057014e-01
-1.23611498e+00 -6.31853878e-01 9.29914832e-01 -7.50113308e-01
8.52838635e-01 1.21165447e-01 7.58323431e-01 1.33240509e+00
1.69663295e-01 9.10884798e-01 6.00146711e-01 -1.46312401e-01
-1.64692327e-01 -4.73589450e-01 -6.81056678e-02 8.39083731e-01
2.72341788e-01 4.59702164e-01 -9.39577341e-01 6.52985349e-02
1.12796795e+00 3.06777775e-01 -6.61505818e-01 -5.98583102e-01
-1.65562403e+00 5.93877614e-01 3.28071415e-01 -1.16569042e-01
-5.48248529e-01 4.64002788e-01 2.98410118e-01 2.09868029e-02
3.43735516e-01 4.55207825e-01 -3.16929758e-01 -2.29423612e-01
-1.04579031e+00 2.91211039e-01 4.09809798e-01 8.02750170e-01
4.06337053e-01 4.49675500e-01 -2.74773449e-01 3.12330097e-01
3.56075019e-01 4.76626009e-01 2.42878929e-01 -1.30675972e+00
5.66884756e-01 2.72895515e-01 2.51057833e-01 -1.13872361e+00
-8.60807225e-02 -5.54807484e-01 -8.47390354e-01 1.46776140e-01
3.00448149e-01 -3.91722649e-01 -6.59409523e-01 1.79452896e+00
1.75802201e-01 8.03456128e-01 2.81937085e-02 1.23506594e+00
3.59429747e-01 7.26948440e-01 -4.42892238e-02 -2.36323327e-01
1.24430013e+00 -9.63443518e-01 -5.64609170e-01 -3.82140368e-01
1.19501986e-01 -3.59303743e-01 9.14055288e-01 5.46478271e-01
-1.11112368e+00 -6.03985608e-01 -1.30276108e+00 8.60392749e-02
2.33842805e-01 5.87576926e-01 4.39666927e-01 1.81481525e-01
-5.08638918e-01 6.16637170e-01 -1.09716654e+00 6.19287938e-02
6.34330630e-01 3.52806538e-01 -3.85849297e-01 9.20958072e-02
-9.52947915e-01 3.49780083e-01 2.85577655e-01 1.82208508e-01
-1.55221128e+00 -8.80150080e-01 -9.27410245e-01 1.30889848e-01
6.19021714e-01 -7.24000573e-01 1.13612437e+00 -1.16136718e+00
-1.44383216e+00 6.92664206e-01 4.30958252e-03 -6.67704523e-01
3.86241287e-01 -7.48134255e-01 -3.60785276e-01 7.16343343e-01
-2.50793174e-02 4.88349795e-01 1.31805849e+00 -8.70775878e-01
-7.03732669e-01 -2.40469933e-01 2.67820537e-01 2.79988259e-01
-3.02024007e-01 3.96431945e-02 -2.69898981e-01 -1.18240905e+00
-2.56107688e-01 -9.44561005e-01 1.36587629e-02 3.08917969e-01
-8.91851410e-02 3.72345716e-01 1.05607903e+00 -7.97454953e-01
1.10057652e+00 -2.39018679e+00 5.82688034e-01 -1.38537496e-01
5.52000821e-01 7.00440779e-02 -9.16682705e-02 3.75823192e-02
-2.38669008e-01 -4.52043325e-01 -1.37547582e-01 -2.01938316e-01
-1.69277087e-01 -6.14728890e-02 -5.57063878e-01 8.72777700e-01
5.00054717e-01 7.29638338e-01 -9.35962558e-01 -5.81846246e-03
2.45116323e-01 6.42829776e-01 -6.91009641e-01 4.42294627e-01
-7.06435964e-02 5.92444301e-01 -3.78656328e-01 6.21869028e-01
2.55877912e-01 -3.54495436e-01 7.36274496e-02 -3.76662374e-01
-6.23122603e-02 6.20956644e-02 -9.47120547e-01 1.93678784e+00
-9.61722806e-02 9.46529269e-01 2.08221972e-01 -1.04777312e+00
5.42294860e-01 3.46753120e-01 7.48051584e-01 -2.95120597e-01
2.75323540e-01 -2.09731638e-01 2.14595310e-02 -5.66555262e-01
4.42889303e-01 1.93574995e-01 -1.20155655e-01 2.58246809e-01
2.80574232e-01 5.08583426e-01 4.46561649e-02 7.92984385e-03
1.19523537e+00 3.63325089e-01 3.37304562e-01 -2.50447094e-01
4.81965929e-01 -3.09446365e-01 7.49682009e-01 3.20783913e-01
-4.79742467e-01 5.53106070e-01 6.55965865e-01 -6.20998383e-01
-7.75070190e-01 -1.01546741e+00 3.55533987e-01 1.17758584e+00
2.16170654e-01 -2.44508892e-01 -8.02231431e-01 -6.20920837e-01
-2.45900854e-01 2.76244551e-01 -7.56224573e-01 -3.64169896e-01
-7.41052508e-01 -4.60214585e-01 5.22210777e-01 6.67295516e-01
6.95365906e-01 -7.39003956e-01 -1.27778363e+00 -1.15438215e-01
-1.46309763e-01 -1.37521827e+00 -7.41582930e-01 -2.43493868e-03
-6.63254380e-01 -1.18617165e+00 -7.19848514e-01 -5.31549037e-01
5.05983353e-01 5.32062829e-01 8.47672760e-01 -4.06961948e-01
-5.06653011e-01 5.44814229e-01 -2.29563266e-01 -5.35334945e-02
3.30827385e-01 -2.41386414e-01 2.75188863e-01 5.56938410e-01
2.18338698e-01 -6.54017568e-01 -8.33339572e-01 2.22047821e-01
-7.93743908e-01 1.56040266e-01 5.28726399e-01 9.86890435e-01
4.99135315e-01 -8.09917971e-03 -2.38471646e-02 -2.99821585e-01
-2.11089849e-04 -5.45843065e-01 -6.63645506e-01 -1.37478132e-02
6.19457103e-02 1.71322422e-03 5.22147596e-01 -6.99574888e-01
-8.01758945e-01 3.48074138e-01 6.13984108e-01 -1.13099039e+00
-1.53352752e-01 3.14276814e-01 -3.55779558e-01 -6.71576113e-02
4.69468176e-01 2.04613358e-01 -4.62784953e-02 -3.10954750e-02
2.79693484e-01 2.61445016e-01 8.44764352e-01 -3.80052507e-01
7.35919952e-01 6.29252732e-01 -6.78742379e-02 -1.04707503e+00
-7.70701110e-01 -3.51537019e-01 -6.04366899e-01 -4.23772305e-01
1.02483535e+00 -1.43503273e+00 -7.36599326e-01 4.00926143e-01
-1.02820909e+00 -1.95509344e-01 -3.87814015e-01 8.13613057e-01
-7.56547689e-01 2.85555243e-01 -4.09588367e-01 -6.62910223e-01
-1.17042221e-01 -1.04474485e+00 1.25987554e+00 1.78966239e-01
-8.87794346e-02 -5.66165924e-01 -1.19391367e-01 3.52675408e-01
-2.78091226e-02 7.91929066e-01 4.20317352e-01 -1.69447735e-01
-8.96603525e-01 -2.68746074e-02 2.36642938e-02 3.36609215e-01
-8.53285473e-03 -1.20815933e-02 -1.15741324e+00 -5.57936311e-01
1.07972860e-01 -3.15173000e-01 9.67422366e-01 5.66206872e-01
1.24636972e+00 -4.51250941e-01 -2.07865313e-01 9.90291953e-01
1.03447878e+00 2.62696594e-01 4.15164948e-01 -1.98392309e-02
7.99871981e-01 5.84220171e-01 6.28036737e-01 8.31060410e-01
8.51787105e-02 6.51936352e-01 6.31642759e-01 1.68061897e-01
1.01191297e-01 -2.76012391e-01 7.61438668e-01 4.04272377e-01
-2.63317704e-01 -2.59095877e-01 -6.01116180e-01 4.32695955e-01
-1.90958786e+00 -1.26518071e+00 3.49774867e-01 2.14295292e+00
3.68715823e-01 -4.15964983e-02 2.97294915e-01 1.99605376e-01
5.41028619e-01 6.78147197e-01 -7.87854791e-01 3.01736712e-01
-1.61644761e-02 -1.62631646e-02 4.58536893e-01 1.91081688e-01
-1.75614703e+00 5.31492651e-01 5.62054634e+00 3.95696491e-01
-1.07205808e+00 -2.01380953e-01 4.44641888e-01 -8.25801492e-01
3.53070378e-01 -3.24392378e-01 -5.19438148e-01 4.63911235e-01
9.48964417e-01 -1.87904388e-01 6.34347081e-01 7.00255454e-01
3.83232027e-01 2.90395599e-02 -1.42589176e+00 1.26489639e+00
1.71382293e-01 -1.21103311e+00 -7.16087408e-03 2.87357032e-01
7.22129226e-01 -1.72147900e-01 4.88833874e-01 -6.14276752e-02
-9.33645889e-02 -1.21181321e+00 9.74393725e-01 6.50887668e-01
8.59382749e-01 -7.61810482e-01 2.08446234e-01 3.21015805e-01
-1.35903275e+00 -2.86901653e-01 -9.89303738e-02 -2.46193722e-01
9.11281556e-02 2.27648169e-01 -2.56640434e-01 3.90613496e-01
7.64849722e-01 1.31458175e+00 -1.18257061e-01 6.92649305e-01
-1.70375764e-01 5.31996429e-01 -5.45875691e-02 3.58277410e-01
3.60610694e-01 -3.44368592e-02 7.98031092e-01 1.00812280e+00
2.42401913e-01 3.71608585e-01 2.42563531e-01 8.07403564e-01
-1.00515537e-01 -5.66178381e-01 -7.02688813e-01 -3.67487818e-01
1.63490430e-01 1.01854432e+00 -2.61118710e-01 -1.87146008e-01
-6.47056401e-01 1.14193785e+00 1.46140531e-01 5.48961639e-01
-1.12893093e+00 -4.34026778e-01 1.15234017e+00 -1.54416682e-02
6.08409524e-01 -4.65221614e-01 1.79634705e-01 -1.40493083e+00
2.59116441e-01 -1.09374976e+00 2.64362216e-01 -5.45045257e-01
-9.11275923e-01 5.48310220e-01 1.10159956e-01 -1.72476614e+00
-3.94898534e-01 -7.35925913e-01 -4.04609174e-01 6.81475103e-01
-1.28154981e+00 -1.06554675e+00 -5.24552047e-01 6.71785235e-01
7.86997914e-01 -4.62366045e-01 5.77720582e-01 2.94863582e-01
-8.09700131e-01 3.37197155e-01 -1.38677120e-01 4.24894780e-01
3.40624988e-01 -9.00095940e-01 2.35714629e-01 1.24203897e+00
3.70288163e-01 5.29992342e-01 5.38699567e-01 -3.69064629e-01
-1.57755053e+00 -1.25905013e+00 2.41760954e-01 -4.28921044e-01
6.43666089e-01 -4.16858047e-01 -5.72928667e-01 7.62810707e-01
2.31182799e-01 6.16726577e-01 7.21740603e-01 -4.83251452e-01
-3.05980504e-01 -5.91132268e-02 -7.69641757e-01 4.58501488e-01
1.09008300e+00 -6.26116633e-01 -6.65968597e-01 -1.65200431e-03
6.83415234e-01 -4.31065381e-01 -6.91501021e-01 4.93745804e-01
8.73439670e-01 -9.62884665e-01 1.09406412e+00 -7.86505699e-01
5.83255768e-01 -5.25661409e-01 -3.69626373e-01 -1.01276720e+00
-4.58568692e-01 -9.98684764e-01 -7.75666416e-01 5.97433209e-01
2.27651987e-02 -1.67876527e-01 8.31196964e-01 3.78240228e-01
9.27382186e-02 -7.32828438e-01 -8.25584173e-01 -6.28903210e-01
-4.68325526e-01 -2.56348699e-01 1.09660171e-01 8.13964963e-01
-1.38226405e-01 1.26262709e-01 -8.23662579e-01 6.01393759e-01
8.81350398e-01 -2.01546941e-02 5.81962883e-01 -9.34828460e-01
-5.39637625e-01 -1.38275430e-01 -4.65804756e-01 -1.41919160e+00
2.20070049e-01 -4.19658482e-01 -1.00055136e-01 -8.29964280e-01
1.89705908e-01 4.13068295e-01 -5.12432158e-01 1.90892264e-01
4.83385229e-04 1.26329243e-01 1.86254278e-01 1.54385790e-01
-5.09225428e-01 7.35308588e-01 1.03874075e+00 -2.01520890e-01
-1.61115822e-04 -2.07892835e-01 -4.64482486e-01 9.58374798e-01
6.34103954e-01 -3.46243709e-01 -7.31669545e-01 -7.37608731e-01
-2.35016122e-01 1.58376649e-01 9.12001193e-01 -1.09693587e+00
1.25593096e-01 -2.23281384e-01 7.32186258e-01 -3.22162449e-01
6.20126426e-01 -9.46717680e-01 1.10287316e-01 4.40621287e-01
-5.10397673e-01 4.54103425e-02 1.93577826e-01 9.86054242e-01
-2.93852150e-01 3.30688328e-01 7.53226459e-01 2.45477725e-02
-8.80909324e-01 5.32293141e-01 -2.60817558e-01 1.61464408e-01
1.40222931e+00 -1.97170362e-01 -1.19755067e-01 -1.74862325e-01
-6.03471100e-01 3.12599801e-02 1.39941767e-01 4.50512499e-01
9.11409020e-01 -1.25068796e+00 -5.12990177e-01 6.17494583e-01
-6.65132999e-02 -2.03020975e-01 3.53234410e-01 8.99523675e-01
-4.74537075e-01 4.91408259e-01 -4.14308965e-01 -5.74626327e-01
-1.20666230e+00 4.74801421e-01 4.63967502e-01 4.72424403e-02
-7.02289045e-01 9.33212280e-01 5.12839437e-01 3.33772629e-01
4.19696867e-01 -6.50978506e-01 -1.75661847e-01 -7.71934614e-02
6.88972354e-01 5.66886544e-01 -5.24855316e-01 -8.23471069e-01
-3.98881167e-01 5.82767546e-01 5.91964312e-02 -1.27491027e-01
1.25787973e+00 9.15715322e-02 3.73666167e-01 2.66571581e-01
1.31556165e+00 -1.90813154e-01 -2.14769840e+00 -2.60034382e-01
-2.94538707e-01 -7.07467794e-01 2.30153412e-01 -3.98413658e-01
-1.21842921e+00 9.49413717e-01 5.91411829e-01 -1.28444090e-01
1.33414793e+00 -3.28486770e-01 5.60762763e-01 3.21622103e-01
3.01745325e-01 -7.37863004e-01 3.43156248e-01 2.66459823e-01
9.15835023e-01 -1.08509958e+00 -3.70651484e-04 -2.83050239e-01
-7.56953657e-01 1.02756226e+00 5.18346369e-01 -5.35351217e-01
4.45606738e-01 2.63701290e-01 -3.24283570e-01 -2.91682631e-01
-1.00246060e+00 5.34720570e-02 5.85851133e-01 4.21933115e-01
1.68847427e-01 -3.32844667e-02 3.25587064e-01 6.43577158e-01
1.97038263e-01 5.14430217e-02 2.71968514e-01 1.03663754e+00
-2.93816507e-01 -3.76507431e-01 -7.37222508e-02 2.49221429e-01
-5.28623939e-01 7.84588009e-02 -1.37579307e-01 7.13024199e-01
1.25235662e-01 7.53979623e-01 1.94382384e-01 -5.41836023e-01
4.13436562e-01 -1.69026077e-01 5.01249552e-01 -4.25866902e-01
-9.14728642e-02 2.62147903e-01 -1.01117991e-01 -1.10118032e+00
-7.48551548e-01 -9.18268502e-01 -9.70010519e-01 5.98508939e-02
1.23997971e-01 -1.32249430e-01 2.21906826e-02 6.95527792e-01
4.34213191e-01 8.03270459e-01 7.44324744e-01 -1.15310645e+00
-5.37643135e-01 -6.53982401e-01 -6.54228985e-01 7.70031661e-02
9.07328427e-01 -1.01286328e+00 -3.31430554e-01 6.29416883e-01] | [8.482495307922363, 0.509787917137146] |
e3051d75-491e-4764-88c3-6cbc416b1258 | deep-keyphrase-generation | 1704.06879 | null | https://arxiv.org/abs/1704.06879v3 | https://arxiv.org/pdf/1704.06879v3.pdf | Deep Keyphrase Generation | Keyphrase provides highly-condensed information that can be effectively used for understanding, organizing and retrieving text content. Though previous studies have provided many workable solutions for automated keyphrase extraction, they commonly divided the to-be-summarized content into multiple text chunks, then ranked and selected the most meaningful ones. These approaches could neither identify keyphrases that do not appear in the text, nor capture the real semantic meaning behind the text. We propose a generative model for keyphrase prediction with an encoder-decoder framework, which can effectively overcome the above drawbacks. We name it as deep keyphrase generation since it attempts to capture the deep semantic meaning of the content with a deep learning method. Empirical analysis on six datasets demonstrates that our proposed model not only achieves a significant performance boost on extracting keyphrases that appear in the source text, but also can generate absent keyphrases based on the semantic meaning of the text. Code and dataset are available at https://github.com/memray/OpenNMT-kpg-release. | ['Peter Brusilovsky', 'Sanqiang Zhao', 'Rui Meng', 'Daqing He', 'Yu Chi', 'Shuguang Han'] | 2017-04-23 | deep-keyphrase-generation-1 | https://aclanthology.org/P17-1054 | https://aclanthology.org/P17-1054.pdf | acl-2017-7 | ['keyphrase-generation'] | ['natural-language-processing'] | [-5.85406609e-02 1.11130498e-01 -4.44203556e-01 2.00808957e-01
-9.51025546e-01 -6.89989090e-01 9.31936741e-01 7.32168019e-01
-2.87479669e-01 6.07242763e-01 1.07070553e+00 -2.50631962e-02
9.76721719e-02 -9.30300236e-01 -6.89494789e-01 -4.29276615e-01
2.76667476e-01 2.52729714e-01 2.10495412e-01 -3.23600799e-01
5.71622431e-01 -1.59016758e-01 -1.74322724e+00 6.58048332e-01
7.63515711e-01 9.49033082e-01 5.56209207e-01 4.86702234e-01
-6.76264405e-01 1.25758612e+00 -6.96727455e-01 -3.10381025e-01
-4.32587974e-02 -5.87331474e-01 -9.78700042e-01 -8.70817602e-02
2.43769214e-01 -5.24160445e-01 -7.78830409e-01 9.80894327e-01
2.35218465e-01 -2.86915511e-01 5.99198699e-01 -1.14150965e+00
-5.89469314e-01 1.35113680e+00 -2.73975313e-01 1.01055987e-02
5.41671395e-01 -6.05517507e-01 1.55500138e+00 -9.44825649e-01
6.09463036e-01 1.09826946e+00 2.95663238e-01 2.18135864e-01
-6.82348549e-01 -6.63354337e-01 -7.10555092e-02 2.82264709e-01
-1.48789382e+00 -3.81866723e-01 8.81481588e-01 -1.81057453e-01
7.13960409e-01 2.55089253e-01 9.24072742e-01 1.02354050e+00
3.51245493e-01 1.42139745e+00 7.19198823e-01 -3.31416488e-01
-3.25827152e-02 2.87953368e-03 2.43823528e-01 5.92688501e-01
3.63924742e-01 -5.53033948e-01 -8.11049521e-01 -2.96655387e-01
2.57979333e-01 1.61781907e-01 -2.46052250e-01 1.31940514e-01
-1.49578142e+00 7.37577975e-01 1.45677477e-01 3.75103801e-01
-6.98420227e-01 8.52414966e-02 5.31813979e-01 3.13444406e-01
6.09540105e-01 6.12933874e-01 -4.68799740e-01 -2.28917271e-01
-1.16166687e+00 5.89585304e-01 1.01132011e+00 9.29647923e-01
8.09249580e-01 -3.95858377e-01 -4.36716229e-01 5.52600265e-01
1.38443694e-01 4.49593544e-01 6.23455584e-01 -5.33123195e-01
5.95161498e-01 9.42460060e-01 1.96373969e-01 -1.20447695e+00
-3.01050544e-01 -2.22618818e-01 -6.28787816e-01 -7.19591558e-01
-1.83293208e-01 -3.09662461e-01 -8.96351993e-01 1.25942969e+00
2.05488905e-01 -3.65976393e-02 6.16516545e-02 4.78643268e-01
1.34538066e+00 1.08753109e+00 -1.34024560e-01 -3.84436697e-02
1.43360162e+00 -7.81526387e-01 -9.40100312e-01 -2.07275838e-01
4.62170482e-01 -1.07002604e+00 9.23948705e-01 2.18551815e-01
-8.73362780e-01 -3.10432225e-01 -1.05598354e+00 -5.13341010e-01
-5.18754065e-01 3.66547734e-01 4.84545380e-01 3.07508763e-02
-8.62717628e-01 3.93282324e-01 -5.83878815e-01 -2.33118281e-01
3.37227792e-01 -2.41893664e-01 -9.42727327e-02 6.70946017e-02
-1.40873682e+00 4.15696681e-01 9.59032953e-01 -4.96568531e-01
-8.40220392e-01 -7.59136558e-01 -7.82948792e-01 2.61308253e-01
8.93511951e-01 -6.96973503e-01 1.21557403e+00 -5.55664837e-01
-1.05492795e+00 5.79697847e-01 -4.49119419e-01 -5.59692264e-01
3.96232381e-02 -6.17254317e-01 -2.10966676e-01 3.86242211e-01
4.50242609e-01 6.87618792e-01 1.01828110e+00 -1.14104688e+00
-1.01377845e+00 -1.13801612e-02 8.51215124e-02 3.28301817e-01
-7.53618300e-01 1.03023477e-01 -6.06724381e-01 -1.17687380e+00
1.49562463e-01 -6.88893139e-01 2.77517825e-01 -5.59258699e-01
-8.76068175e-01 -3.78725320e-01 7.06086934e-01 -9.07247663e-01
1.81792951e+00 -1.96155143e+00 1.00804009e-01 -3.24151143e-02
5.12428105e-01 -7.37645999e-02 1.26777384e-02 1.28301251e+00
4.03925717e-01 2.99238384e-01 -4.39395532e-02 2.11011767e-02
1.45680383e-01 -1.27910540e-01 -1.18134379e+00 -9.48988348e-02
-2.64784485e-01 1.12805736e+00 -9.53975439e-01 -5.96892536e-01
-1.20355502e-01 3.00739676e-01 -2.11128265e-01 3.71487468e-01
-6.52941585e-01 -2.19101712e-01 -8.24025393e-01 4.93276387e-01
2.52193451e-01 -5.13296545e-01 -4.21734899e-02 -4.22816187e-01
7.63848349e-02 8.17060113e-01 -9.98599768e-01 1.45666921e+00
-3.09242666e-01 7.66813517e-01 -2.37632960e-01 -6.02716386e-01
7.67182589e-01 4.19481397e-01 8.83184671e-01 -5.54698408e-01
9.64212790e-02 1.22657649e-01 -6.10786796e-01 -1.53228879e-01
1.26270866e+00 1.50503173e-01 -3.18771303e-01 8.29822958e-01
5.30964956e-02 -3.47675920e-01 3.61785263e-01 7.91726470e-01
9.76425529e-01 -1.43809959e-01 4.63305354e-01 -1.66062400e-01
3.35652322e-01 2.07366541e-01 1.55683279e-01 8.06164861e-01
4.99829322e-01 5.33280432e-01 6.88880146e-01 -2.95360118e-01
-1.12054074e+00 -5.48120558e-01 2.12411687e-01 1.09708631e+00
3.24947625e-01 -1.40500259e+00 -6.28213048e-01 -7.05965817e-01
1.01351500e-01 8.21346998e-01 -6.15451217e-01 -1.96972311e-01
-2.87138462e-01 -5.09484351e-01 5.51376522e-01 1.96196601e-01
3.52563739e-01 -1.00484133e+00 -3.48419458e-01 2.63638079e-01
-7.67245829e-01 -1.24670303e+00 -4.67210025e-01 4.09907289e-02
-3.31880718e-01 -1.05060458e+00 -7.53748834e-01 -6.41631126e-01
5.60451150e-01 6.51503682e-01 1.15747797e+00 1.48642570e-01
3.24044973e-02 4.70427573e-01 -9.50559020e-01 -6.25442386e-01
-4.75852579e-01 4.86109585e-01 -2.83205897e-01 1.29209487e-02
5.56114733e-01 -2.66005874e-01 -5.19277573e-01 -2.27420807e-01
-1.22150946e+00 4.87690330e-01 5.05361199e-01 6.86533630e-01
6.19349599e-01 5.45014441e-01 4.10875410e-01 -8.52028966e-01
1.06493115e+00 -7.33593643e-01 -2.02229962e-01 2.60988623e-01
-9.16900039e-01 3.09261560e-01 7.33400762e-01 -6.50409907e-02
-8.22486937e-01 -3.91348511e-01 -6.90287650e-02 -8.86818618e-02
1.29577681e-01 8.14843595e-01 8.85427147e-02 6.59407318e-01
2.61605203e-01 8.73149931e-01 -5.30562997e-01 -7.58209348e-01
6.13654613e-01 7.42572606e-01 2.56132990e-01 -6.65887833e-01
7.96167970e-01 5.24330854e-01 -2.87296951e-01 -8.24724257e-01
-1.37382936e+00 -6.51632726e-01 -4.37947303e-01 2.94628018e-03
6.04826510e-01 -1.35784197e+00 -4.39584292e-02 5.22647619e-01
-1.11396670e+00 5.79151995e-02 -3.16317528e-01 9.18363184e-02
-2.21214041e-01 5.87185979e-01 -5.94469666e-01 -3.55634868e-01
-9.13888633e-01 -6.35563076e-01 1.33993983e+00 1.17122859e-01
-4.05230761e-01 -7.37162113e-01 -8.48600864e-02 1.56039089e-01
8.33323747e-02 -3.16729322e-02 1.03449488e+00 -8.93747985e-01
-5.45165241e-01 -4.16929632e-01 -2.27603197e-01 8.24231058e-02
4.82179940e-01 6.17613876e-03 -6.56758904e-01 -1.99724451e-01
-1.95911855e-01 -4.70976621e-01 1.31765473e+00 1.49844855e-01
1.24138272e+00 -9.64613855e-01 -2.87154973e-01 4.30007786e-01
1.20536005e+00 -6.93384483e-02 5.03052771e-01 4.31658775e-01
1.07406771e+00 5.57157457e-01 5.35806715e-01 7.91884124e-01
9.79272246e-01 3.90515924e-01 2.14333057e-01 2.50278234e-01
-1.46942362e-01 -8.82989585e-01 4.80773449e-01 1.25683594e+00
4.63917166e-01 -4.22216922e-01 -8.67229044e-01 7.06353068e-01
-1.88973761e+00 -9.89636123e-01 6.08853847e-02 1.58924043e+00
1.26696086e+00 2.37384111e-01 -9.35976766e-03 2.09801823e-01
3.84550005e-01 6.32601500e-01 -1.33277789e-01 1.88028157e-01
-1.00431934e-01 7.81735778e-02 4.13584769e-01 3.23923826e-01
-9.92749214e-01 1.23357189e+00 5.46997356e+00 1.03004205e+00
-9.91640508e-01 -1.20106421e-01 3.47647667e-01 6.95692748e-02
-7.29852259e-01 1.80418938e-01 -1.18879402e+00 5.64599395e-01
7.30945170e-01 -8.47095668e-01 1.48376212e-01 8.49619806e-01
-8.20068642e-03 -1.80293113e-01 -9.36187029e-01 9.02586102e-01
1.66607574e-01 -1.59897923e+00 6.61024213e-01 -1.36546239e-01
7.64362752e-01 -1.32277519e-01 -1.54236048e-01 1.16604589e-01
2.45407104e-01 -7.12101698e-01 9.97026443e-01 4.99507099e-01
4.33592916e-01 -7.26801157e-01 6.21505380e-01 3.51007253e-01
-1.37336528e+00 3.33099850e-02 -4.64924902e-01 1.52672395e-01
-2.76038885e-01 9.36901271e-01 -9.51081574e-01 5.10591209e-01
5.90545297e-01 1.01286912e+00 -7.92833984e-01 7.32616127e-01
-5.90618908e-01 7.38679767e-01 -1.63197041e-01 -4.33728337e-01
3.30960304e-01 1.99066162e-01 6.18510485e-01 1.36189020e+00
3.87041301e-01 -3.65251452e-02 3.46312195e-01 8.27159584e-01
-3.35503668e-01 4.75764573e-01 -3.68774891e-01 -7.61930168e-01
8.44374895e-01 1.26571953e+00 -1.00381792e+00 -7.01650560e-01
-3.28047872e-01 8.88194680e-01 -1.37749791e-01 2.31192380e-01
-4.05868948e-01 -7.30999947e-01 5.21362841e-01 1.49449393e-01
3.72732460e-01 -3.08160245e-01 -2.20321283e-01 -1.61929572e+00
2.14332759e-01 -1.06296968e+00 4.81348425e-01 -8.63427222e-01
-8.85804474e-01 4.53543335e-01 1.77857757e-01 -1.13264942e+00
-4.03063655e-01 -1.72349080e-01 -3.28510880e-01 6.41801596e-01
-1.30142021e+00 -1.35486197e+00 -2.79525340e-01 4.39973652e-01
7.87736237e-01 -1.73823223e-01 6.68506742e-01 -1.31503776e-01
-1.59012660e-01 2.40972832e-01 2.30271399e-01 4.82871503e-01
6.16846561e-01 -1.41594994e+00 7.69991219e-01 9.93548930e-01
6.03818715e-01 6.92974150e-01 7.28813767e-01 -8.44469905e-01
-1.59110200e+00 -9.90143776e-01 1.19972932e+00 -4.59028989e-01
8.38264287e-01 -5.15497148e-01 -9.15613234e-01 4.72074986e-01
5.40291250e-01 -7.32744575e-01 6.79631829e-01 -5.75619675e-02
-4.23304439e-01 -6.00634217e-02 -3.11312497e-01 8.65915239e-01
5.54335833e-01 -8.86978328e-01 -1.03326488e+00 2.29012743e-01
1.26382780e+00 -3.55436027e-01 -5.77904105e-01 1.24323405e-01
4.47319537e-01 -6.19300067e-01 8.50883842e-01 -2.75190115e-01
8.64325702e-01 -1.30270824e-01 -9.51929167e-02 -1.38080120e+00
-1.74866214e-01 -8.67270231e-01 -8.12451601e-01 1.26839852e+00
3.66802394e-01 -2.05190077e-01 5.18681884e-01 1.19708233e-01
1.14281997e-01 -7.50653625e-01 -3.78829002e-01 -2.42007181e-01
-7.82640576e-02 -4.55651015e-01 8.73308957e-01 9.02215421e-01
1.97405830e-01 6.53828263e-01 -5.68140090e-01 -1.78885311e-01
4.65632260e-01 5.00084281e-01 7.69922137e-01 -1.13144517e+00
3.28419246e-02 -5.59258759e-01 6.94194138e-02 -1.21690392e+00
6.68796748e-02 -9.77488756e-01 -1.98623568e-01 -2.03536057e+00
5.26724577e-01 -5.15511297e-02 -3.10752511e-01 7.47119486e-01
-4.30578202e-01 -2.39525408e-01 1.68695748e-01 5.05412877e-01
-7.58243620e-01 6.95209205e-01 1.23973894e+00 -2.74794281e-01
-5.03572337e-02 -1.52028501e-01 -1.29493809e+00 4.35960740e-01
7.38714516e-01 -6.37122095e-01 -5.60860455e-01 -2.07706004e-01
9.22939599e-01 -3.88107114e-02 3.25114012e-01 -8.62410545e-01
3.37516159e-01 -3.33703369e-01 2.79986173e-01 -9.55645978e-01
8.62284601e-02 -6.33616984e-01 -2.39868343e-01 2.37569481e-01
-6.51135147e-01 5.56019917e-02 1.64179489e-01 2.96405703e-01
-5.20412326e-01 -2.57183611e-01 1.21026598e-01 -2.83506155e-01
-8.06897938e-01 3.53295535e-01 -4.69041288e-01 3.71673971e-01
3.68859887e-01 3.09854567e-01 -3.01777869e-01 -7.99704731e-01
-8.38043690e-02 1.34796545e-01 2.65897393e-01 7.87826002e-01
8.16259444e-01 -1.29673135e+00 -9.53243315e-01 1.01981908e-02
3.67864192e-01 -6.18604794e-02 -1.37488350e-01 2.74315655e-01
-3.31103146e-01 7.61898816e-01 7.16527328e-02 8.29994306e-02
-1.06526625e+00 5.38110316e-01 -1.97582424e-01 -3.17963183e-01
-9.16750729e-01 6.05716765e-01 1.33428887e-01 3.83778289e-02
1.60489604e-01 -5.63851595e-01 -3.19182456e-01 6.45156026e-01
8.97934377e-01 1.09122500e-01 -3.01766228e-02 -5.13157964e-01
9.98518839e-02 2.96874255e-01 -4.13677931e-01 -7.94364214e-02
1.30032706e+00 -4.19741780e-01 -4.53922510e-01 5.34932196e-01
1.32324684e+00 2.00000495e-01 -7.88684905e-01 -7.13292420e-01
2.67596096e-01 -1.82680279e-01 4.64180261e-01 -6.60178781e-01
-6.96749508e-01 6.20339453e-01 -1.38677791e-01 5.43142080e-01
1.19242871e+00 2.17825562e-01 1.40440941e+00 6.86396539e-01
1.37012452e-01 -1.31520152e+00 2.72633493e-01 8.89574766e-01
8.86768401e-01 -1.11558855e+00 3.66607279e-01 -2.12270409e-01
-6.30798876e-01 1.28751850e+00 3.13169062e-01 2.20965892e-01
7.62373030e-01 2.04343021e-01 -3.86072993e-02 -4.68702704e-01
-9.05216634e-01 -4.08447623e-01 7.73205042e-01 3.76631580e-02
4.54492033e-01 1.60035025e-02 -2.97719300e-01 8.69308650e-01
-6.23998463e-01 -2.10835531e-01 5.86548030e-01 9.09158409e-01
-8.58430028e-01 -9.84717786e-01 -7.56588876e-02 7.14065731e-01
-8.70970726e-01 -4.47540224e-01 -8.93503964e-01 3.49585623e-01
-1.56785250e-01 7.56277263e-01 -1.41209990e-01 -6.86938584e-01
-1.20433025e-01 2.78258413e-01 1.52030746e-02 -7.46539116e-01
-3.81054372e-01 3.36391062e-01 2.39606015e-02 -3.29186440e-01
-2.05901757e-01 -3.82248670e-01 -1.22043693e+00 -4.07289356e-01
6.31196722e-02 5.03494382e-01 4.21454191e-01 9.25176263e-01
5.71173012e-01 4.81521964e-01 7.63335884e-01 -3.69045287e-01
-2.48915717e-01 -1.01071513e+00 -4.12608176e-01 3.39069605e-01
3.84181380e-01 -7.84761384e-02 -1.05829515e-01 3.42365652e-01] | [12.29585075378418, 8.9010009765625] |
32106577-59dc-4a23-8639-e863b7f30302 | improving-zero-shot-multilingual-neural-1 | 2305.07310 | null | https://arxiv.org/abs/2305.07310v1 | https://arxiv.org/pdf/2305.07310v1.pdf | Improving Zero-shot Multilingual Neural Machine Translation by Leveraging Cross-lingual Consistency Regularization | The multilingual neural machine translation (NMT) model has a promising capability of zero-shot translation, where it could directly translate between language pairs unseen during training. For good transfer performance from supervised directions to zero-shot directions, the multilingual NMT model is expected to learn universal representations across different languages. This paper introduces a cross-lingual consistency regularization, CrossConST, to bridge the representation gap among different languages and boost zero-shot translation performance. The theoretical analysis shows that CrossConST implicitly maximizes the probability distribution for zero-shot translation, and the experimental results on both low-resource and high-resource benchmarks show that CrossConST consistently improves the translation performance. The experimental analysis also proves that CrossConST could close the sentence representation gap and better align the representation space. Given the universality and simplicity of CrossConST, we believe it can serve as a strong baseline for future multilingual NMT research. | ['Haifeng Wang', 'Hua Wu', 'Zhongjun He', 'Liwen Zhang', 'Pengzhi Gao'] | 2023-05-12 | null | null | null | null | ['nmt'] | ['computer-code'] | [-8.86773765e-02 -1.24006368e-01 -7.23243773e-01 -3.78612310e-01
-1.37032926e+00 -5.72444022e-01 8.25063765e-01 -3.15041542e-01
-2.58184463e-01 8.87634993e-01 4.01649326e-01 -7.87720501e-01
4.24733520e-01 -5.78495443e-01 -1.06261075e+00 -4.77316350e-01
3.24795246e-01 7.11804271e-01 -4.36773092e-01 -5.21422267e-01
-3.55487138e-01 -1.17773153e-01 -8.86647582e-01 3.80140215e-01
1.22343481e+00 2.69947588e-01 2.69184500e-01 3.20264250e-02
-3.09048861e-01 3.62808585e-01 -3.31965685e-01 -7.95503795e-01
5.11871874e-01 -7.27841079e-01 -7.66775131e-01 -3.99700284e-01
6.38496697e-01 -2.47891657e-02 -2.44330361e-01 1.14620435e+00
8.05511713e-01 6.08698577e-02 6.69597805e-01 -9.81003344e-01
-1.35443723e+00 9.88803923e-01 -6.15397274e-01 1.84615001e-01
3.81705910e-02 -1.99401174e-02 1.32403088e+00 -1.44625199e+00
9.06084239e-01 1.36570871e+00 5.72899282e-01 7.52830327e-01
-1.23610377e+00 -6.68338239e-01 -1.20751830e-02 1.05465273e-03
-1.41138792e+00 -4.79811698e-01 2.62795657e-01 -2.44085088e-01
1.36233711e+00 2.24713415e-01 3.29986781e-01 1.62618780e+00
6.04531705e-01 9.54217434e-01 1.10471725e+00 -7.25982189e-01
-6.91279694e-02 3.04318190e-01 -3.99711579e-02 5.14596105e-01
1.84271887e-01 2.31177643e-01 -6.07045650e-01 9.62137729e-02
6.77712560e-01 -3.24434370e-01 -2.40498513e-01 -4.11921054e-01
-1.67965055e+00 8.36251676e-01 3.98179859e-01 6.44247055e-01
9.88016650e-02 -1.38273779e-02 6.62289500e-01 8.04405212e-01
7.75279939e-01 5.34737825e-01 -4.07493114e-01 -1.11332983e-01
-8.75724256e-01 -3.01849604e-01 4.75948602e-01 1.38402009e+00
8.00017178e-01 3.81395161e-01 -4.82389063e-01 1.25871706e+00
-1.73910871e-01 9.09003675e-01 9.15672719e-01 -4.34307009e-01
1.00196052e+00 3.08620691e-01 -1.99503526e-01 -4.52586889e-01
1.27716467e-01 -6.40667677e-01 -8.28107238e-01 -4.27037001e-01
1.08775713e-01 -3.76577288e-01 -6.36874974e-01 2.02105737e+00
-1.99575424e-01 -1.48137912e-01 3.75883251e-01 9.44904685e-01
5.49085021e-01 1.00148833e+00 -1.95540354e-01 -3.18043590e-01
1.09651399e+00 -1.20715070e+00 -9.13700283e-01 -6.16852880e-01
1.20247686e+00 -1.12430120e+00 1.59186268e+00 -3.84590536e-01
-8.55843961e-01 -4.41467434e-01 -1.13253689e+00 -3.68766874e-01
-4.12282854e-01 2.26921886e-01 6.80477083e-01 4.31099862e-01
-9.75966036e-01 4.53165501e-01 -7.96658456e-01 -6.47518575e-01
9.97451991e-02 5.22546358e-02 -5.04892349e-01 -3.94305408e-01
-1.69699037e+00 1.35636067e+00 1.61682829e-01 -1.02930851e-01
-6.15964055e-01 -5.26997864e-01 -1.00028813e+00 -1.19175717e-01
8.58943630e-03 -8.15697610e-01 1.14378393e+00 -1.11595535e+00
-1.46825838e+00 9.06940520e-01 -3.63226712e-01 -4.67018008e-01
5.03175735e-01 -3.16542059e-01 -3.42964977e-01 -5.39205432e-01
5.02713084e-01 7.65019000e-01 4.18163389e-01 -8.90451372e-01
-1.93805680e-01 -8.71313810e-02 -4.18597251e-01 5.12130737e-01
-5.97258866e-01 1.20236389e-01 -5.18602967e-01 -8.60107481e-01
7.85338879e-02 -1.12987924e+00 -4.03846195e-03 -4.06557977e-01
-3.50895017e-01 -2.66729891e-01 4.61278766e-01 -6.30256593e-01
9.87371683e-01 -1.88838327e+00 4.82332855e-01 -3.63432944e-01
-5.81131756e-01 1.25571817e-01 -5.49535036e-01 6.50960565e-01
-4.79202457e-02 7.13019865e-03 -2.58052230e-01 -5.21615148e-01
1.32823065e-02 5.65175414e-01 -5.39964497e-01 4.50614274e-01
2.91141361e-01 1.49224842e+00 -9.49508846e-01 -3.25462013e-01
-7.60711282e-02 5.30150890e-01 -2.85659850e-01 1.82995088e-02
-1.01900741e-01 3.46637458e-01 -1.21486764e-02 5.68632901e-01
3.76547962e-01 -2.27307588e-01 4.23247635e-01 7.92637244e-02
1.83745530e-02 5.76867104e-01 -3.50722015e-01 2.33229017e+00
-7.37718046e-01 7.70523131e-01 -4.65116322e-01 -6.89198196e-01
1.09290469e+00 4.66149986e-01 2.57457197e-01 -1.02292645e+00
2.40967244e-01 6.79911137e-01 7.53399730e-02 -2.32825309e-01
6.84900045e-01 -3.74054283e-01 -3.33694547e-01 7.40124166e-01
4.11175370e-01 -9.72179994e-02 6.04832806e-02 1.09290071e-01
4.12789851e-01 3.83931607e-01 2.83357233e-01 -6.53207719e-01
6.95668533e-02 1.24460503e-01 6.83266580e-01 5.23209393e-01
-1.91756003e-02 4.87518877e-01 -7.40499645e-02 -4.89918917e-01
-1.26058972e+00 -1.16818058e+00 -1.08063377e-01 1.26444900e+00
1.21146433e-01 -2.64921099e-01 -7.22691357e-01 -7.63334751e-01
-2.11077437e-01 9.47566330e-01 -3.18120003e-01 -3.86762291e-01
-8.01050842e-01 -9.10058677e-01 5.93352497e-01 4.43879783e-01
1.84605867e-01 -6.31941617e-01 2.25836590e-01 1.37607902e-01
-8.04464877e-01 -9.88879800e-01 -9.36884284e-01 3.93269628e-01
-9.36438560e-01 -3.60100925e-01 -1.02435780e+00 -1.25011265e+00
5.80496550e-01 6.36430383e-01 1.35821307e+00 -4.17428523e-01
1.73278645e-01 -1.99149355e-01 -3.19021672e-01 -1.53788522e-01
-5.94172060e-01 4.74237889e-01 4.70420182e-01 -2.32614353e-01
8.01863492e-01 -5.07813692e-01 -3.15272547e-02 3.38298917e-01
-4.67022657e-01 6.19658940e-02 5.19416273e-01 1.11855876e+00
6.05518997e-01 -7.76497841e-01 6.84141934e-01 -6.02607727e-01
9.06020343e-01 -5.90272903e-01 -3.20872456e-01 6.79878950e-01
-7.86600232e-01 3.23353618e-01 6.04949057e-01 -5.36079228e-01
-7.49735951e-01 -4.58891124e-01 2.19028354e-01 -5.45951426e-01
5.18715084e-01 4.19790983e-01 -1.90174282e-01 2.71354854e-01
8.35284889e-01 4.89275843e-01 -7.33738318e-02 -4.67215955e-01
7.09988892e-01 7.50231624e-01 3.01267415e-01 -8.83113921e-01
6.49638236e-01 -3.38413902e-02 -4.37889755e-01 -4.01583314e-01
-7.97512412e-01 -1.70036152e-01 -6.84610546e-01 1.66157067e-01
6.83880270e-01 -1.38015091e+00 3.38271141e-01 -1.94735397e-02
-1.34745228e+00 -2.45498091e-01 -2.86754280e-01 8.03451061e-01
-6.00913227e-01 1.77185178e-01 -8.36141229e-01 -3.79303068e-01
-6.32795691e-01 -1.34721661e+00 1.16333640e+00 -2.54328966e-01
-3.45775753e-01 -1.25979662e+00 3.07479918e-01 2.42611006e-01
3.90799284e-01 -3.87829989e-01 1.23479962e+00 -6.12826169e-01
-5.58930397e-01 7.08972802e-03 -8.09221938e-02 4.52050447e-01
2.86960691e-01 -4.48910683e-01 -7.81952083e-01 -7.11973310e-01
-2.05187555e-02 -5.16769350e-01 7.87631214e-01 1.95828170e-01
2.79284924e-01 -2.79758722e-01 -8.24856088e-02 7.17962801e-01
1.41047299e+00 -2.69869685e-01 5.29146492e-01 2.69486308e-01
6.70229673e-01 3.79010528e-01 6.41026080e-01 -2.76356459e-01
4.35658872e-01 8.55809808e-01 -3.10281827e-03 -2.55592495e-01
-3.37541401e-01 -5.08567274e-01 8.53103817e-01 1.95013142e+00
7.97290877e-02 -3.23050320e-01 -8.48363936e-01 6.60001755e-01
-1.89884543e+00 -7.61070967e-01 7.89011940e-02 2.25786614e+00
1.13092542e+00 -2.01725200e-01 -3.57862979e-01 -5.61389983e-01
8.69799972e-01 1.40140876e-01 -3.30069542e-01 -8.16603839e-01
-6.33972943e-01 9.19850767e-02 4.93824273e-01 5.98444700e-01
-6.27991915e-01 1.45189905e+00 6.87634516e+00 1.05428565e+00
-1.22149277e+00 6.43510520e-01 5.11324167e-01 -2.08981365e-01
-4.39382046e-01 -8.77761766e-02 -9.14792538e-01 3.31817836e-01
1.04498839e+00 -5.68742692e-01 5.43870628e-01 7.07459569e-01
3.41882706e-02 6.24677896e-01 -1.39148533e+00 9.12298441e-01
3.16371024e-01 -1.30995655e+00 4.10927534e-01 1.72880590e-01
1.25604427e+00 7.44732916e-01 2.21043319e-01 8.08945119e-01
5.07425666e-01 -8.76392066e-01 4.70465004e-01 1.16863415e-01
1.17811346e+00 -7.69654930e-01 7.98009932e-01 5.37864506e-01
-9.67151463e-01 3.69486928e-01 -8.40727866e-01 1.74390636e-02
1.93317533e-01 2.76628822e-01 -8.83511245e-01 8.50940406e-01
1.22973472e-01 1.00759983e+00 -2.66904771e-01 5.11767566e-01
-3.07151735e-01 3.58875245e-01 5.31852618e-02 1.70611348e-02
3.84628534e-01 -5.21183908e-01 5.12038291e-01 1.31803346e+00
7.77868032e-01 -7.23068237e-01 2.78215498e-01 8.02339256e-01
-4.39830124e-01 5.60074508e-01 -8.96228194e-01 -2.66737819e-01
5.52273035e-01 6.79067671e-01 -6.02727756e-02 -4.98262197e-01
-6.21763229e-01 1.46311593e+00 6.47478044e-01 5.52464604e-01
-7.84040928e-01 -8.42087269e-02 1.02320576e+00 -1.66775689e-01
-1.87799279e-02 -3.04864973e-01 -4.19529527e-01 -1.83963120e+00
1.82061687e-01 -1.10466433e+00 -3.63825485e-02 -5.80884397e-01
-1.65809298e+00 9.60923612e-01 -2.94252783e-01 -1.65992558e+00
-5.14479935e-01 -5.41452527e-01 -3.49100083e-01 1.22543669e+00
-1.50534248e+00 -1.37330115e+00 7.30115533e-01 5.17540276e-01
1.01203203e+00 -5.91201246e-01 1.24727976e+00 4.08321768e-01
-7.07902610e-01 1.13643992e+00 5.65611899e-01 2.17020139e-01
1.08340168e+00 -8.13149810e-01 8.47090781e-01 1.00402200e+00
5.50707698e-01 1.14819241e+00 4.68575895e-01 -7.08037674e-01
-1.63741422e+00 -1.39036226e+00 1.52425218e+00 -6.20798826e-01
7.97924638e-01 -4.60677445e-01 -9.01565850e-01 8.90000522e-01
6.99671865e-01 -7.30784461e-02 8.66715610e-01 4.15544361e-01
-8.88595045e-01 2.34016050e-02 -6.21095359e-01 8.16662669e-01
9.39558029e-01 -1.02230501e+00 -7.99325109e-01 6.95223510e-01
1.01846182e+00 -2.03262776e-01 -8.59786332e-01 2.80352026e-01
4.57665622e-01 -3.64507735e-01 7.87116468e-01 -8.70622039e-01
5.05709589e-01 1.10128179e-01 -5.16716540e-01 -1.76572454e+00
-2.77002931e-01 -6.18405044e-01 1.71012599e-02 1.07558429e+00
8.83279085e-01 -6.58911347e-01 2.41802752e-01 9.09341499e-02
-2.44289592e-01 -7.28744984e-01 -1.29209948e+00 -1.49761415e+00
1.04600787e+00 -1.29410028e-01 6.49595141e-01 1.47408032e+00
3.36468339e-01 9.65863705e-01 -9.13499236e-01 -1.89723805e-01
4.79182303e-01 1.93525493e-01 5.69336474e-01 -7.42348492e-01
-4.40943599e-01 -3.25501442e-01 -1.08053833e-01 -1.31493783e+00
4.39405710e-01 -1.72760892e+00 6.14069067e-02 -1.46105087e+00
4.28455472e-01 -1.13231942e-01 -4.48897272e-01 4.32178617e-01
-3.58739525e-01 2.76495695e-01 8.43384862e-02 5.94851017e-01
-2.25808039e-01 7.50322580e-01 1.52048242e+00 -3.90931219e-01
5.43894097e-02 -3.95663202e-01 -6.49065256e-01 9.68913659e-02
8.12143922e-01 -5.73997796e-01 -4.70309287e-01 -1.30408132e+00
-9.04422440e-03 -9.14691761e-02 -4.58207637e-01 -5.61426401e-01
-3.36209908e-02 -1.99263364e-01 3.28319962e-03 -2.71388650e-01
3.65394592e-01 -5.74705243e-01 -8.41982588e-02 2.21829951e-01
-4.37287420e-01 3.14453512e-01 8.07098597e-02 3.10814112e-01
-2.59471625e-01 1.24175828e-02 7.21370876e-01 -2.55649090e-01
-3.29539597e-01 2.60116845e-01 -1.05873577e-01 3.06813627e-01
6.58915043e-01 7.06356950e-03 -4.76694077e-01 -1.09967411e-01
-2.53785223e-01 1.72001347e-01 6.38591647e-01 1.01341391e+00
2.42849648e-01 -1.94965184e+00 -1.05634665e+00 4.31189775e-01
4.53265309e-01 -6.12962008e-01 -2.08803192e-01 8.16780329e-01
9.77707002e-03 5.97097158e-01 -1.85992777e-01 -7.71137297e-01
-8.74251842e-01 4.51140165e-01 3.62909853e-01 -2.21775904e-01
-6.27996981e-01 9.34826314e-01 1.26908496e-01 -9.80433106e-01
-2.02629361e-02 -1.14883736e-01 4.84314919e-01 -8.37079510e-02
5.37818253e-01 -3.81006524e-02 3.01416188e-01 -9.21095669e-01
-1.94524825e-01 4.80984330e-01 -3.81612927e-01 -3.70467484e-01
1.07176399e+00 -3.08990151e-01 -2.54937589e-01 8.69994044e-01
1.33660853e+00 -1.34312406e-01 -8.11797798e-01 -6.96726263e-01
5.22802770e-02 -4.54827815e-01 -1.65290996e-01 -7.58011103e-01
-7.71775246e-01 1.19076133e+00 2.94492006e-01 -4.72166896e-01
5.53590894e-01 -6.01093508e-02 1.09874833e+00 5.53585589e-01
8.12548697e-01 -1.10514402e+00 -1.44401312e-01 1.02657259e+00
8.55873168e-01 -1.38810003e+00 -3.64445567e-01 -1.20890327e-01
-8.09349239e-01 9.05771852e-01 5.21776795e-01 5.42072840e-02
-1.65825617e-02 2.04525262e-01 4.46236938e-01 3.11376482e-01
-1.03871059e+00 -5.87511668e-03 3.36868852e-01 5.79265118e-01
9.94719684e-01 3.55465859e-01 -5.34136176e-01 3.73351485e-01
-3.50947559e-01 -2.65153289e-01 1.94798782e-01 6.40443087e-01
-3.61141890e-01 -1.55274320e+00 -1.22229844e-01 -6.63106441e-02
-1.18687861e-01 -5.91955662e-01 -5.96854985e-01 6.25785649e-01
-3.58518921e-02 8.40274394e-01 -4.32520732e-02 -3.91376734e-01
1.76130012e-01 5.57655871e-01 5.92230380e-01 -7.56337225e-01
-6.18830442e-01 2.11585596e-01 1.04520969e-01 -3.30792397e-01
9.91416425e-02 -4.35997665e-01 -8.61323714e-01 -2.73311228e-01
-3.03982049e-01 2.24058956e-01 7.36095965e-01 8.97364557e-01
5.10748744e-01 3.72261852e-01 6.10635936e-01 -4.22719866e-01
-9.63921309e-01 -1.31126547e+00 -1.37679517e-01 3.39159846e-01
6.62563294e-02 -3.62986237e-01 -1.54391438e-01 -1.54358774e-01] | [11.548465728759766, 10.190595626831055] |
2a08ef4d-9c53-4e9a-9824-b5f0e5aa79de | understanding-and-stabilizing-gans-training | 1909.13188 | null | https://arxiv.org/abs/1909.13188v4 | https://arxiv.org/pdf/1909.13188v4.pdf | Understanding and Stabilizing GANs' Training Dynamics with Control Theory | Generative adversarial networks (GANs) are effective in generating realistic images but the training is often unstable. There are existing efforts that model the training dynamics of GANs in the parameter space but the analysis cannot directly motivate practically effective stabilizing methods. To this end, we present a conceptually novel perspective from control theory to directly model the dynamics of GANs in the function space and provide simple yet effective methods to stabilize GANs' training. We first analyze the training dynamic of a prototypical Dirac GAN and adopt the widely-used closed-loop control (CLC) to improve its stability. We then extend CLC to stabilize the training dynamic of normal GANs, where CLC is implemented as a squared $L2$ regularizer on the output of the discriminator. Empirical results show that our method can effectively stabilize the training and obtain state-of-the-art performance on data generation tasks. | ['Kun Xu', 'Jun Zhu', 'Chongxuan Li', 'Bo Zhang'] | 2019-09-29 | null | https://openreview.net/forum?id=BJe7h34YDS | https://openreview.net/pdf?id=BJe7h34YDS | null | ['l2-regularization'] | ['methodology'] | [ 1.58186153e-01 3.73228282e-01 1.14091121e-01 7.77070001e-02
-7.27395952e-01 -7.28303850e-01 8.06750834e-01 -7.92284667e-01
8.75981674e-02 9.95336592e-01 -7.33121410e-02 -3.35876703e-01
2.45111600e-01 -7.26402223e-01 -9.78625834e-01 -1.11971676e+00
3.05214852e-01 2.45539367e-01 -2.40630150e-01 -4.79643196e-01
-1.80230126e-01 2.27756858e-01 -1.07257032e+00 -1.04852632e-01
1.18265665e+00 7.46405423e-01 -3.74910496e-02 8.54365468e-01
3.38653684e-01 8.42387319e-01 -7.96225786e-01 -4.27384019e-01
4.47034568e-01 -1.20551586e+00 -5.61651289e-01 3.77996708e-03
2.12784231e-01 -5.68229966e-02 -2.27024511e-01 1.06691587e+00
6.35354817e-01 -2.47034226e-02 8.07557046e-01 -1.33741796e+00
-7.92569518e-01 5.11811733e-01 -3.97646964e-01 -1.76803172e-01
-2.91564822e-01 4.78661031e-01 6.17183506e-01 -3.07981908e-01
5.31900346e-01 1.26173842e+00 6.77861869e-01 1.33505678e+00
-1.46450138e+00 -5.72903335e-01 2.22204164e-01 -3.43129069e-01
-1.10738742e+00 -4.08746064e-01 9.24375415e-01 -4.68095452e-01
4.63770062e-01 4.13108855e-01 6.92518830e-01 1.43401933e+00
4.33520675e-01 5.50717235e-01 1.11559820e+00 -5.61461985e-01
3.88975084e-01 -6.72650263e-02 -3.88720959e-01 8.95995259e-01
1.47190522e-02 2.56061792e-01 -2.16831461e-01 5.16550541e-02
1.13883853e+00 -5.60461223e-01 -3.55414301e-01 -3.63170177e-01
-8.12253118e-01 9.58413720e-01 6.68746173e-01 1.99061245e-01
-2.29423583e-01 8.34521532e-01 1.40855983e-01 4.74112004e-01
5.90259671e-01 7.09815204e-01 -1.15997558e-02 -1.68469533e-01
-6.63683295e-01 3.65662694e-01 4.46309060e-01 9.14076447e-01
4.65605348e-01 7.92408109e-01 -3.34582359e-01 7.53571451e-01
2.84546345e-01 5.41762292e-01 5.96955240e-01 -1.30612600e+00
4.08877254e-01 2.51198381e-01 5.40902950e-02 -4.89439994e-01
4.57820520e-02 -5.09446084e-01 -1.24057186e+00 4.69477683e-01
4.32995826e-01 -6.42972410e-01 -1.08552980e+00 2.22151232e+00
1.73899889e-01 2.18854398e-01 -9.89300907e-02 7.01877892e-01
2.23153341e-03 8.38479400e-01 -2.51837879e-01 -4.92102206e-01
6.35010660e-01 -9.24743295e-01 -8.32827985e-01 -6.66267648e-02
2.92187810e-01 -4.10678208e-01 1.45505524e+00 2.19036117e-01
-1.58334196e+00 -5.57407260e-01 -1.02242053e+00 1.43367320e-01
-1.92643464e-01 2.34533951e-01 3.86519760e-01 7.72808611e-01
-1.44686306e+00 7.71158576e-01 -1.27089155e+00 5.11963367e-02
4.81870651e-01 3.31394643e-01 1.75989375e-01 3.08208257e-01
-1.04079950e+00 8.60069573e-01 6.01947308e-02 1.86075851e-01
-1.17387664e+00 -9.17083383e-01 -6.28954768e-01 -1.54543117e-01
1.56221688e-01 -1.21973217e+00 1.40151823e+00 -1.23406541e+00
-2.30697227e+00 5.10783136e-01 -1.48845054e-02 -5.40306211e-01
8.74030709e-01 -7.48370662e-02 -1.21571586e-01 -3.43459904e-01
-1.64010242e-01 8.15397918e-01 1.28704333e+00 -1.48898029e+00
-4.08843644e-02 9.78555083e-02 -2.02670693e-01 1.13204673e-01
-1.99439317e-01 -3.71028543e-01 -1.43175244e-01 -1.03395319e+00
-2.85682172e-01 -1.28581750e+00 -2.59443432e-01 -2.70805031e-01
-6.02676153e-01 -1.14356782e-02 7.71667004e-01 -4.04760808e-01
1.30006015e+00 -1.76774621e+00 4.31891859e-01 3.86550315e-02
-1.45147359e-02 4.47752327e-01 -1.91983119e-01 2.81182498e-01
-7.39059895e-02 2.80147731e-01 -3.97424161e-01 -5.74979782e-01
1.24638461e-01 4.22852218e-01 -8.12667608e-01 4.30932641e-01
3.17181438e-01 1.34723973e+00 -7.40649939e-01 2.02828646e-02
3.16722803e-02 6.79178894e-01 -7.82809556e-01 4.36047882e-01
-5.70184410e-01 7.64842868e-01 -3.09498638e-01 2.23884791e-01
3.66077513e-01 -1.26707152e-01 1.51496619e-01 5.52218594e-02
1.72999583e-03 3.23107988e-02 -6.84322238e-01 1.47783589e+00
-4.79730070e-01 7.14644611e-01 1.47973582e-01 -9.94514108e-01
6.71014726e-01 1.60209447e-01 2.15798452e-01 -3.20203573e-01
1.85421631e-01 1.30581468e-01 1.86250702e-01 -3.58331832e-03
4.15762328e-02 -4.85629439e-01 -2.59281881e-02 2.58958369e-01
-2.16311193e-03 -7.24180460e-01 3.91067117e-02 7.99636617e-02
9.86170888e-01 1.36940733e-01 -2.47189701e-02 -5.66800535e-01
5.05064309e-01 -3.24085027e-01 4.06348974e-01 7.15184629e-01
1.03217706e-01 6.29574835e-01 6.64288998e-01 -1.64579809e-01
-1.29870033e+00 -1.16605091e+00 2.00456053e-01 7.79100418e-01
-2.63943285e-01 -2.77761281e-01 -1.27788448e+00 -6.17610812e-01
-3.14922631e-02 8.85302424e-01 -8.87477338e-01 -4.05306965e-01
-6.86565399e-01 -7.52524972e-01 7.00592697e-01 5.43033481e-01
6.09733641e-01 -8.70399177e-01 -2.54069477e-01 2.65866742e-02
9.97315068e-03 -6.34560645e-01 -8.93939912e-01 8.14709291e-02
-6.79141045e-01 -7.27305949e-01 -7.97486603e-01 -6.56449795e-01
8.93847167e-01 -4.63714689e-01 1.08789515e+00 -1.23888738e-01
-2.87226904e-02 3.92629832e-01 2.17904717e-01 -5.71084917e-01
-1.04484773e+00 2.27821052e-01 1.99339062e-01 -6.91961721e-02
-7.00631559e-01 -6.86305523e-01 -5.10903537e-01 3.22158992e-01
-8.76053274e-01 2.71554917e-01 9.27104875e-02 1.04762673e+00
6.02191925e-01 3.10257971e-02 8.70287955e-01 -8.76704454e-01
9.38941658e-01 -1.19649783e-01 -9.74282324e-01 1.67726949e-01
-9.35506284e-01 3.81112903e-01 1.11870754e+00 -5.48829734e-01
-1.15299547e+00 3.45615707e-02 -2.22539574e-01 -7.40548551e-01
3.47182304e-01 -7.55264610e-02 -1.76501513e-01 -2.03510344e-01
7.83683240e-01 1.51222140e-01 3.79419714e-01 -1.21931836e-01
6.76446855e-01 2.35864893e-01 7.56281435e-01 -8.16991448e-01
1.11621296e+00 2.92518914e-01 3.00135851e-01 -4.78839427e-01
-7.86240518e-01 4.59189117e-01 -1.76323354e-01 -2.98397094e-01
7.03412592e-01 -7.83361077e-01 -9.03061330e-01 6.90606713e-01
-9.97439623e-01 -1.07978559e+00 -7.00725377e-01 -1.16493218e-01
-1.13209987e+00 -1.97738603e-01 -8.42367768e-01 -8.57695103e-01
-5.00749111e-01 -1.06512368e+00 9.66039956e-01 2.58630335e-01
1.10740773e-01 -1.24585307e+00 3.09312284e-01 -2.60329302e-02
7.21139014e-01 6.75908983e-01 9.07149911e-01 1.57194376e-01
-5.14501929e-01 2.40571678e-01 4.60316509e-01 7.97397971e-01
9.73519906e-02 3.09551477e-01 -8.49145532e-01 -5.59207261e-01
3.36881846e-01 -2.65469939e-01 7.51890242e-01 6.34367466e-01
1.27261841e+00 -6.86027646e-01 -1.66937992e-01 9.08758640e-01
1.45144331e+00 3.32045883e-01 9.64946449e-01 -1.51676297e-01
7.93152750e-01 1.25861749e-01 1.59361176e-02 1.42369688e-01
-3.10903764e-03 5.91682673e-01 4.20209587e-01 -5.34483651e-03
-2.14482829e-01 -5.83843648e-01 8.05937707e-01 9.95703936e-01
-2.79235333e-01 -4.23371732e-01 -6.31516695e-01 1.28402829e-01
-1.81316340e+00 -9.55118954e-01 1.58732131e-01 2.07531285e+00
1.14294600e+00 -3.40974494e-03 8.49626139e-02 -7.34056234e-02
7.66555965e-01 2.24652421e-03 -7.16662288e-01 -4.89755690e-01
-9.43851173e-02 4.24893558e-01 4.78304267e-01 7.58882046e-01
-8.68993044e-01 8.90459478e-01 7.60508299e+00 8.68624985e-01
-1.41018009e+00 -4.66648154e-02 9.94705617e-01 -3.13746601e-01
-3.63393009e-01 -3.19625527e-01 -6.15838945e-01 8.01840544e-01
1.09712625e+00 -3.39105785e-01 8.88437510e-01 8.70065928e-01
3.33821237e-01 2.39334360e-01 -1.07047594e+00 8.84312451e-01
-1.86393067e-01 -1.64664531e+00 3.53737101e-02 2.74529725e-01
1.33934462e+00 -3.41621101e-01 5.09476364e-01 4.19804394e-01
5.61925471e-01 -1.19172740e+00 6.80352211e-01 6.57580316e-01
1.10475171e+00 -8.60441625e-01 3.15488547e-01 4.49377835e-01
-7.31700480e-01 -1.46544173e-01 -3.02918196e-01 8.43298212e-02
2.34179720e-01 3.57882142e-01 -6.17946208e-01 1.93703949e-01
7.83887506e-02 5.10593534e-01 -5.07369101e-01 4.17082936e-01
-3.58999521e-01 8.63648713e-01 -2.04387024e-01 1.38651326e-01
2.11042762e-01 -6.16281629e-01 4.97207403e-01 7.78431058e-01
5.51636398e-01 2.28006952e-02 -2.79223949e-01 1.23849869e+00
-3.52155179e-01 -4.20973480e-01 -7.45681524e-01 -1.30364984e-01
1.90336257e-01 9.23795283e-01 -3.43539268e-01 -7.56926611e-02
2.74839222e-01 9.71418798e-01 3.47083181e-01 4.33171600e-01
-1.29281652e+00 -6.82656933e-03 8.85283947e-01 2.38121018e-01
1.10230923e-01 -2.52490699e-01 -3.51344287e-01 -1.18339825e+00
-2.31413245e-02 -1.00852847e+00 1.13199554e-01 -7.68973529e-01
-1.33218908e+00 5.38179100e-01 -2.52182931e-01 -9.89978135e-01
-8.39685798e-01 -4.65750486e-01 -9.88391578e-01 8.84467483e-01
-1.10299253e+00 -1.04787421e+00 -2.02378988e-01 6.98861659e-01
3.04975003e-01 -2.10265592e-01 6.39447331e-01 -4.60259952e-02
-8.49659085e-01 9.21576321e-01 4.48206991e-01 -1.32206053e-01
3.85057211e-01 -1.50144529e+00 6.28972828e-01 1.09380603e+00
-1.01636827e-01 6.78762078e-01 6.63544774e-01 -5.78925192e-01
-1.54959452e+00 -1.18938160e+00 1.71316504e-01 -6.26667321e-01
6.01664424e-01 -5.70318937e-01 -6.73369586e-01 6.98766589e-01
4.40246135e-01 -2.89495923e-02 1.64618656e-01 -4.43557829e-01
1.11445338e-02 -2.44138226e-01 -1.13463819e+00 9.39026713e-01
1.11837971e+00 -3.57114583e-01 -5.66506991e-03 2.54344046e-01
7.23253787e-01 -6.56533182e-01 -5.90383053e-01 3.76482844e-01
2.35022247e-01 -7.91859806e-01 8.52585614e-01 -7.71472931e-01
5.19528925e-01 -3.03554177e-01 1.26179174e-01 -2.04529881e+00
-1.47878185e-01 -1.32566512e+00 -3.80871594e-01 1.22111380e+00
9.07924250e-02 -7.43887663e-01 7.75480986e-01 4.75333452e-01
-3.25916320e-01 -8.86251330e-01 -1.09960008e+00 -9.03326511e-01
7.42717981e-01 -1.39287055e-01 2.36565620e-01 5.94181001e-01
-1.48803279e-01 4.54846203e-01 -4.83273894e-01 -1.61228597e-01
4.56841022e-01 -2.34865651e-01 7.53606319e-01 -7.43778646e-01
-3.21603537e-01 -6.21538520e-01 -5.05176792e-03 -1.00909388e+00
3.07639003e-01 -7.06853151e-01 3.27147514e-01 -1.05711079e+00
-3.01583797e-01 -5.12491703e-01 1.16373904e-01 1.59292057e-01
-2.37662792e-01 5.42155728e-02 2.88064390e-01 8.98154527e-02
-9.22743157e-02 8.46722782e-01 1.54138172e+00 1.44806243e-02
-2.23216474e-01 1.67177156e-01 -7.94097066e-01 4.70412195e-01
9.41769779e-01 -9.47385728e-02 -9.64357138e-01 -4.21610177e-01
9.92873162e-02 -1.30814478e-01 3.16852868e-01 -1.01461387e+00
2.08460856e-02 -2.78299391e-01 2.22889259e-01 5.51572815e-02
2.25074291e-01 -5.22454858e-01 4.64083523e-01 5.73015749e-01
-5.07528961e-01 1.54673085e-01 4.32940163e-02 4.42839235e-01
2.54328530e-02 8.08915719e-02 1.35028958e+00 1.82005558e-02
2.62032717e-01 3.32291722e-01 -2.78686345e-01 4.58575696e-01
9.26689446e-01 4.71234649e-01 -5.48343360e-01 -6.73134327e-01
-5.28911889e-01 3.74678057e-04 7.01678813e-01 6.68703541e-02
1.57591298e-01 -1.70251000e+00 -4.83869344e-01 4.62542564e-01
-6.09777331e-01 1.96667343e-01 -4.67174873e-02 3.28623444e-01
-5.43438256e-01 2.85824627e-01 -7.93289319e-02 -4.11516160e-01
-6.56161249e-01 5.80084741e-01 9.13190663e-01 -3.41552228e-01
-2.83776462e-01 8.31310928e-01 3.19654316e-01 -6.48068115e-02
9.80458483e-02 -4.72948879e-01 4.16342616e-01 -2.44124159e-01
2.17821211e-01 3.18389505e-01 -1.75038218e-01 -1.31131455e-01
-9.66310725e-02 4.40091819e-01 4.33063060e-01 -3.00237715e-01
1.18183494e+00 -2.12674960e-02 -5.61539903e-02 3.41770351e-01
1.03259397e+00 -1.03117920e-01 -1.77951765e+00 3.52416754e-01
-6.29640520e-01 -1.80592015e-01 -1.78416014e-01 -6.44159794e-01
-1.39850926e+00 8.28536391e-01 5.27415872e-01 5.56834340e-01
1.22363627e+00 -2.46426731e-01 6.44321740e-01 2.92189140e-02
1.74434990e-01 -1.26363540e+00 3.38223934e-01 5.02884269e-01
1.19709241e+00 -8.03467870e-01 -4.45563674e-01 -3.22421551e-01
-7.40797698e-01 8.22338820e-01 8.74531925e-01 -4.21279788e-01
4.90832984e-01 6.12962961e-01 -3.60242650e-02 2.35798404e-01
-1.00300717e+00 1.90051675e-01 1.95631102e-01 6.70858204e-01
3.61247748e-01 -3.27651426e-02 -2.46228099e-01 3.87084126e-01
-4.04425830e-01 -2.05230296e-01 5.90604424e-01 4.81258124e-01
-2.18272349e-03 -1.25194097e+00 -2.26787403e-01 2.01136306e-01
-3.02115142e-01 -7.07520824e-03 -2.90489614e-01 7.35636234e-01
8.37148875e-02 7.10080445e-01 6.83067515e-02 -2.20627666e-01
1.72282070e-01 4.05168593e-01 7.14827180e-01 -2.75377333e-01
-6.38289094e-01 1.50446966e-02 -2.90953070e-01 -4.33958709e-01
-1.49850816e-01 -6.10695899e-01 -7.76488602e-01 -5.32993138e-01
-2.46267021e-01 7.43115544e-02 4.67883348e-01 7.37780988e-01
4.21896398e-01 8.03450644e-01 9.59080935e-01 -7.17265964e-01
-1.02058351e+00 -7.96451926e-01 -3.50497484e-01 3.71005356e-01
3.31600547e-01 -4.93797094e-01 -3.50667953e-01 4.13747162e-01] | [11.579325675964355, 0.010694744996726513] |
d8bd095d-8350-4c75-889c-611cde3ad87c | distribution-aware-graph-representation | 2205.06576 | null | https://arxiv.org/abs/2205.06576v1 | https://arxiv.org/pdf/2205.06576v1.pdf | Distribution-Aware Graph Representation Learning for Transient Stability Assessment of Power System | The real-time transient stability assessment (TSA) plays a critical role in the secure operation of the power system. Although the classic numerical integration method, \textit{i.e.} time-domain simulation (TDS), has been widely used in industry practice, it is inevitably trapped in a high computational complexity due to the high latitude sophistication of the power system. In this work, a data-driven power system estimation method is proposed to quickly predict the stability of the power system before TDS reaches the end of simulating time windows, which can reduce the average simulation time of stability assessment without loss of accuracy. As the topology of the power system is in the form of graph structure, graph neural network based representation learning is naturally suitable for learning the status of the power system. Motivated by observing the distribution information of crucial active power and reactive power on the power system's bus nodes, we thus propose a distribution-aware learning~(DAL) module to explore an informative graph representation vector for describing the status of a power system. Then, TSA is re-defined as a binary classification task, and the stability of the system is determined directly from the resulting graph representation without numerical integration. Finally, we apply our method to the online TSA task. The case studies on the IEEE 39-bus system and Polish 2383-bus system demonstrate the effectiveness of our proposed method. | ['Mingli Song', 'Zunlei Feng', 'Jie Song', 'Quan Zhang', 'Rong Yan', 'Na Yu', 'Shunyu Liu', 'KaiXuan Chen'] | 2022-05-12 | null | null | null | null | ['numerical-integration'] | ['miscellaneous'] | [-2.22976610e-01 -9.76427495e-02 -1.33895800e-01 6.54002354e-02
-2.34353095e-01 -5.96447170e-01 9.46244895e-02 4.08449113e-01
1.79372326e-01 7.96911180e-01 -5.87795496e-01 -8.00594628e-01
-6.69822097e-01 -8.92156422e-01 -9.33697000e-02 -1.00943112e+00
-5.03510594e-01 3.56098533e-01 -3.63182127e-02 -6.27485037e-01
-6.47367761e-02 7.44465530e-01 -9.99640524e-01 -5.01974285e-01
1.18674386e+00 1.10551345e+00 -1.17346570e-01 1.70642927e-01
2.03049734e-01 6.34026229e-01 -9.73663747e-01 2.45954290e-01
-2.17535533e-02 -4.66886908e-01 -6.25854194e-01 1.91847116e-01
-7.44179666e-01 9.71359108e-03 -6.84307575e-01 1.47191381e+00
4.35423523e-01 3.03847700e-01 7.26850629e-01 -1.78137624e+00
1.49709359e-01 7.01909184e-01 -5.98267555e-01 3.82739425e-01
2.66757309e-01 -4.32375148e-02 7.29214728e-01 -4.73324209e-01
2.45206341e-01 8.20262611e-01 3.65065336e-01 -1.40121028e-01
-1.17723858e+00 -6.19414568e-01 1.81241304e-01 6.40223145e-01
-1.73019075e+00 1.71590403e-01 1.40818655e+00 -4.10670400e-01
6.10593975e-01 2.90314764e-01 9.77842033e-01 2.20859230e-01
6.17114186e-01 5.59637904e-01 9.22551632e-01 -3.11322331e-01
4.42397535e-01 -6.74618557e-02 3.31711054e-01 5.76450646e-01
1.64039969e-01 4.40701433e-02 1.46528244e-01 -6.24329858e-02
1.83969915e-01 -2.16513336e-01 -5.31260133e-01 -4.11481529e-01
-7.10284710e-01 7.70838916e-01 7.41930604e-01 5.61442256e-01
-1.59483347e-02 -2.39446089e-01 6.71713233e-01 5.90081453e-01
5.34177184e-01 3.07472259e-01 -1.07487693e-01 1.69903100e-01
-8.49842966e-01 -4.27770056e-02 7.89995909e-01 5.88237464e-01
5.51427543e-01 9.57021236e-01 3.21555585e-01 3.06263894e-01
9.48235467e-02 3.43603611e-01 4.58034724e-01 -2.99938887e-01
3.70165437e-01 1.04575539e+00 2.67237518e-02 -1.23640919e+00
-7.73640811e-01 -6.54258490e-01 -1.25959122e+00 7.50439346e-01
3.95806491e-01 -6.11938953e-01 -2.99775749e-01 1.58272314e+00
3.87242824e-01 8.38138536e-02 3.22612897e-02 6.40706718e-01
2.12872282e-01 1.22180545e+00 -2.62418747e-01 -8.88662338e-01
8.85947526e-01 -3.69315624e-01 -9.35677469e-01 2.46443450e-01
5.97734630e-01 -2.85822123e-01 4.96919781e-01 3.81338298e-01
-8.41966987e-01 -3.89273405e-01 -1.60334432e+00 8.49095225e-01
-5.45207739e-01 2.54946828e-01 2.20296353e-01 4.66673493e-01
-8.79662812e-01 7.79419959e-01 -7.63225019e-01 -9.50503275e-02
-6.85221255e-02 3.92228693e-01 -1.04465999e-01 4.61252540e-01
-1.65783751e+00 1.15491402e+00 8.07360530e-01 7.16861725e-01
-5.67167461e-01 -3.95597965e-01 -8.26980352e-01 2.81255394e-01
6.87094986e-01 -2.64025331e-01 7.13571370e-01 -8.67740631e-01
-1.38519728e+00 -7.53936693e-02 3.57867122e-01 -1.45088986e-01
4.04059380e-01 5.58960378e-01 -8.63047898e-01 2.48803377e-01
-2.33435065e-01 -6.19313478e-01 1.00560188e+00 -9.48214054e-01
-2.65056133e-01 -2.28970721e-01 -1.05267152e-01 2.19759360e-01
-3.98854822e-01 -2.51508951e-01 2.87803262e-01 -5.98127544e-01
1.64165422e-01 -6.18867159e-01 -5.43519616e-01 -2.57063955e-01
-3.91582072e-01 -4.60046917e-01 9.39415395e-01 -9.34529245e-01
1.69624400e+00 -2.06390619e+00 4.49739039e-01 8.90294194e-01
7.76662156e-02 4.01509315e-01 3.84803265e-01 8.62860978e-01
-5.02471268e-01 -2.91082501e-01 -3.18911612e-01 3.80946130e-01
-5.34657761e-02 -5.83870243e-03 -1.94884598e-01 7.32310593e-01
8.15730989e-02 4.80381340e-01 -7.98070312e-01 -3.47300619e-01
4.38714325e-01 1.17342025e-01 5.05645312e-02 1.52633861e-01
-3.38189416e-02 4.08208132e-01 -7.13424146e-01 2.44370431e-01
5.18473923e-01 -3.21349680e-01 5.17114580e-01 -3.01605880e-01
-8.94964114e-03 -3.64807934e-01 -1.49230063e+00 1.23558998e+00
-4.80416983e-01 4.43267345e-01 2.56865650e-01 -1.50030482e+00
9.29049492e-01 4.16881830e-01 7.52347767e-01 -5.71534753e-01
5.41437328e-01 2.54426952e-02 2.79041737e-01 -2.59625018e-01
7.00838224e-04 7.17873499e-02 -2.50454754e-01 3.69921654e-01
-1.24974445e-01 -3.85542393e-01 4.42159712e-01 3.23258549e-01
7.36581981e-01 -2.79749215e-01 7.15371907e-01 -7.33021438e-01
1.07566619e+00 3.64094712e-02 5.83660424e-01 -2.14760020e-01
3.01455874e-02 -2.81248629e-01 1.14340651e+00 -2.50252247e-01
-7.33039737e-01 -9.13196206e-01 -2.13709712e-01 3.28990042e-01
2.74579912e-01 -2.68807024e-01 -5.74514568e-01 -5.31348288e-01
-5.86507320e-02 8.40023160e-01 -3.58508229e-01 -6.91516459e-01
-5.54049015e-01 -1.11169934e+00 1.13558285e-01 2.72651345e-01
2.61714280e-01 -6.85484111e-01 -1.71056643e-01 4.74501580e-01
2.39691347e-01 -6.43679142e-01 -1.37734339e-01 3.12873900e-01
-7.49194741e-01 -1.27829421e+00 -5.11455059e-01 -9.27149951e-01
8.68584871e-01 -3.82763773e-01 6.97539687e-01 2.03506738e-01
-2.17318028e-01 -4.14279588e-02 -1.03946045e-01 2.77421802e-01
-5.83005965e-01 3.02015850e-03 2.09960341e-01 9.53234881e-02
-5.32386005e-01 -7.11795807e-01 -3.31492394e-01 2.99087256e-01
-7.20225751e-01 -1.42081276e-01 2.62245625e-01 1.04109526e+00
1.81190550e-01 1.22237182e+00 1.18904722e+00 -4.71671402e-01
9.00122225e-01 -6.11021698e-01 -1.29085517e+00 4.27831322e-01
-9.73642647e-01 -8.71557295e-02 1.32303250e+00 -2.90175736e-01
-6.91867948e-01 -1.21552087e-01 5.19057438e-02 -3.39378297e-01
3.30554694e-01 1.11380684e+00 -5.10181427e-01 -3.13090950e-01
1.82037085e-01 4.39703703e-01 3.11699748e-01 -1.48237795e-01
4.73766066e-02 3.83706391e-01 3.81814688e-01 -3.68935019e-01
1.18640614e+00 -2.23927289e-01 7.18509436e-01 -7.64949441e-01
-1.83340922e-01 -1.83586046e-01 -5.10353327e-01 -4.72534180e-01
2.85436153e-01 -6.43580556e-01 -1.43517780e+00 6.91197574e-01
-7.42921710e-01 -6.22879853e-03 -5.79543598e-02 2.43203133e-01
-2.32616201e-01 7.06369698e-01 -5.88626981e-01 -9.38509166e-01
-3.27837318e-01 -1.08344972e+00 2.14143321e-01 7.08634406e-02
1.49736673e-01 -1.36513829e+00 -5.21006919e-02 -3.74900520e-01
5.10554798e-02 6.70750618e-01 1.38130331e+00 -5.56292057e-01
-1.89743161e-01 -5.83260715e-01 9.44834761e-03 4.73631620e-01
5.92262670e-02 1.47374600e-01 -3.61078143e-01 -9.76193428e-01
3.43664229e-01 2.69153621e-02 -4.48710322e-02 2.63336301e-01
9.06694591e-01 -3.20120960e-01 -2.93698221e-01 3.49478543e-01
1.72123599e+00 6.17268741e-01 2.67087221e-01 -7.23847374e-02
5.37967205e-01 4.31583524e-01 6.20473564e-01 5.95287561e-01
1.95046887e-01 5.18628955e-01 4.76540387e-01 -5.07445149e-02
4.63396192e-01 6.98051378e-02 5.26097655e-01 1.18365407e+00
1.84178710e-01 -4.21006858e-01 -8.74056220e-01 1.61144584e-01
-1.77780247e+00 -5.86109698e-01 -1.55854188e-02 2.09323740e+00
4.10158366e-01 2.90452242e-01 1.53739035e-01 9.76993859e-01
9.30053711e-01 2.79462904e-01 -6.06541455e-01 -2.84036696e-01
-7.17970878e-02 -1.00340940e-01 9.47706327e-02 6.89900517e-01
-7.68765390e-01 1.11778811e-01 4.94537354e+00 1.34384298e+00
-1.33608401e+00 -4.00902271e-01 7.37590909e-01 5.61569154e-01
6.33606166e-02 -9.33516398e-02 -3.20891559e-01 6.76988006e-01
9.81601179e-01 -1.12179625e+00 5.46889484e-01 7.07677543e-01
4.39277500e-01 -2.45562032e-01 -7.78891921e-01 8.07352662e-01
-8.75906497e-02 -8.71743441e-01 -8.21784586e-02 -1.47873357e-01
5.44535577e-01 -5.70375204e-01 -5.85296489e-02 1.62919089e-01
-9.44285542e-02 -6.82904184e-01 6.71892583e-01 2.41713881e-01
7.78973162e-01 -1.23710954e+00 8.89815211e-01 5.15368342e-01
-1.49999785e+00 -4.57814902e-01 -2.15822533e-02 -9.47712734e-02
3.81705284e-01 7.59925008e-01 -7.27514923e-01 1.26611221e+00
3.28559071e-01 9.42031622e-01 -6.98089302e-01 1.02531755e+00
-2.26286426e-01 5.03386557e-01 -2.94944435e-01 -4.59980875e-01
1.69784337e-01 -6.44887865e-01 6.28807366e-01 4.38416868e-01
2.49317855e-01 2.30665922e-01 5.00153542e-01 6.34255409e-01
3.50208819e-01 1.26485742e-04 -4.54848856e-01 9.51105282e-02
3.27303410e-01 1.33804488e+00 -8.92519712e-01 -2.97837168e-01
-1.73080675e-02 2.58598953e-01 1.06574573e-01 6.99977636e-01
-7.27632225e-01 -7.88883746e-01 6.43050745e-02 -5.42555004e-02
-2.24244550e-01 -2.79059261e-01 -2.98358798e-01 -1.16109526e+00
4.46000583e-02 -6.45750642e-01 6.53625727e-01 -6.83074832e-01
-1.21524930e+00 8.78054917e-01 1.66495100e-01 -1.78074574e+00
-8.05968583e-01 -5.41206062e-01 -9.40971792e-01 1.01030886e+00
-1.27563608e+00 -7.46559441e-01 1.37910798e-01 8.12765419e-01
2.67387062e-01 -2.42477968e-01 5.35114706e-01 3.29470545e-01
-1.08345413e+00 3.26157808e-01 4.60008681e-01 2.24263459e-01
-2.51818836e-01 -1.31404257e+00 -7.98661634e-02 1.07100391e+00
-3.89106542e-01 5.70670888e-02 8.42318416e-01 -5.06440103e-01
-1.50039327e+00 -6.25952899e-01 4.03130382e-01 2.49436781e-01
1.26311886e+00 -3.48322392e-01 -9.98194873e-01 2.81170428e-01
3.51149619e-01 1.64179094e-02 1.09475352e-01 -2.90285617e-01
4.28330213e-01 -4.78588969e-01 -1.01210237e+00 5.84077418e-01
1.94410384e-01 -3.76679450e-01 -5.36974788e-01 2.89150268e-01
3.41010809e-01 -1.39604121e-01 -1.19630325e+00 5.12781680e-01
-9.42584872e-02 -3.19211066e-01 8.56722772e-01 -3.85010242e-01
-2.75857896e-01 -7.05192327e-01 3.41976970e-01 -1.75795376e+00
-2.80931085e-01 -8.01317811e-01 -2.83874780e-01 1.02485657e+00
2.21679837e-01 -9.64557111e-01 5.02534568e-01 8.23670551e-02
-2.94009373e-02 -7.65004277e-01 -1.31450331e+00 -7.72668302e-01
1.59287810e-01 1.95915606e-02 6.16889000e-01 9.89003956e-01
7.23389626e-01 4.66597646e-01 -1.77917480e-02 6.36239409e-01
7.39190459e-01 4.36246336e-01 7.19871223e-02 -1.04057956e+00
-1.17387243e-01 -4.78798896e-01 -5.18516839e-01 -2.97018796e-01
5.36368370e-01 -7.98990428e-01 -3.22343826e-01 -1.23883688e+00
-5.05486727e-01 -3.22480381e-01 -5.11769533e-01 1.45194054e-01
-7.99341798e-02 -3.50450486e-01 2.21437961e-02 -7.56272599e-02
-2.36361966e-01 8.58078420e-01 1.22986007e+00 -3.13978642e-01
6.23805597e-02 4.62509483e-01 2.06990600e-01 4.39788997e-01
7.03164756e-01 4.20653224e-02 -8.01791072e-01 2.93360800e-01
3.04588318e-01 1.11655748e+00 6.28574938e-02 -1.18978262e+00
4.05704111e-01 -1.80225540e-02 2.10084766e-01 -8.35615814e-01
1.19051650e-01 -1.36003363e+00 3.07691395e-01 1.04997933e+00
7.59629235e-02 4.13525790e-01 1.82590142e-01 4.76989537e-01
-5.15569210e-01 -2.08041474e-01 6.95989728e-01 3.87017876e-01
-7.26308107e-01 2.71568418e-01 -7.21828878e-01 -1.76607653e-01
1.01363325e+00 1.47049516e-01 -2.78229058e-01 -4.35226649e-01
-9.92835879e-01 6.51562631e-01 4.92062308e-02 6.07055239e-02
3.66339266e-01 -1.28443170e+00 -4.85643536e-01 3.81045282e-01
-2.20683962e-01 -2.95862734e-01 4.38127667e-01 7.08082497e-01
-5.11747122e-01 1.02733493e-01 -1.21846318e-01 -5.86130977e-01
-7.36852944e-01 7.73672044e-01 5.57133019e-01 -3.58611465e-01
-4.80388910e-01 -1.12209827e-01 -4.21539843e-01 -6.12329738e-03
-1.10237233e-01 -2.90802270e-01 -6.64271414e-01 3.76853347e-01
1.27062425e-01 7.49944329e-01 2.74027348e-01 -4.81226236e-01
-3.02629024e-01 4.91476715e-01 4.24095303e-01 1.49676993e-01
1.10060859e+00 -2.43146017e-01 -4.45861161e-01 5.28017581e-01
1.17584682e+00 -4.96046126e-01 -9.21636224e-01 6.65559620e-02
-3.59365270e-02 -1.05955832e-01 1.77685663e-01 -5.99292755e-01
-1.44273198e+00 6.49515510e-01 4.33818609e-01 7.69446909e-01
1.35606706e+00 -5.94806492e-01 3.59851122e-01 3.99598479e-01
5.66026092e-01 -1.00139785e+00 -2.32723728e-01 2.73675203e-01
7.05904365e-01 -7.74394214e-01 2.39938900e-01 -5.07040739e-01
-2.31786326e-01 1.42234540e+00 3.89788091e-01 -2.86730975e-01
1.13597596e+00 3.76914501e-01 -1.75272584e-01 -1.94768086e-02
-6.34786308e-01 3.74142230e-01 3.00404668e-01 2.66839862e-01
-2.07234249e-01 3.17690931e-02 -4.17272061e-01 5.93019009e-01
-1.62993699e-01 -2.62934089e-01 8.10850441e-01 8.60758483e-01
-1.75699636e-01 -6.45094514e-01 -3.74090314e-01 2.97903389e-01
-1.25048772e-01 4.36170429e-01 2.84900159e-01 8.81462216e-01
-3.31815571e-01 9.86793280e-01 -2.01369122e-01 -2.91083246e-01
4.12161589e-01 -1.08257808e-01 1.22951575e-01 -1.54839680e-01
-4.47013885e-01 2.82735340e-02 -1.39668614e-01 -3.03414136e-01
7.41115510e-02 -5.14022768e-01 -1.35843742e+00 -4.03919816e-01
-5.35846174e-01 7.75226831e-01 3.88832718e-01 8.84244919e-01
-2.35508367e-01 7.70673573e-01 1.37445688e+00 -6.75495386e-01
-7.60604799e-01 -6.11556172e-01 -1.21727228e+00 -1.18803650e-01
2.81336933e-01 -7.11383224e-01 -6.87283635e-01 -5.30084491e-01] | [5.9242353439331055, 2.6051368713378906] |
a4472a72-b8d8-4f4c-8184-a922ba0bfb5d | improving-electron-micrograph-signal-to-noise | 1807.11234 | null | http://arxiv.org/abs/1807.11234v2 | http://arxiv.org/pdf/1807.11234v2.pdf | Improving Electron Micrograph Signal-to-Noise with an Atrous Convolutional Encoder-Decoder | We present an atrous convolutional encoder-decoder trained to denoise
512$\times$512 crops from electron micrographs. It consists of a modified
Xception backbone, atrous convoltional spatial pyramid pooling module and a
multi-stage decoder. Our neural network was trained end-to-end to remove
Poisson noise applied to low-dose ($\ll$ 300 counts ppx) micrographs created
from a new dataset of 17267 2048$\times$2048 high-dose ($>$ 2500 counts ppx)
micrographs and then fine-tuned for ordinary doses (200-2500 counts ppx). Its
performance is benchmarked against bilateral, non-local means, total variation,
wavelet, Wiener and other restoration methods with their default parameters.
Our network outperforms their best mean squared error and structural similarity
index performances by 24.6% and 9.6% for low doses and by 43.7% and 5.5% for
ordinary doses. In both cases, our network's mean squared error has the lowest
variance. Source code and links to our new high-quality dataset and trained
network have been made publicly available at
https://github.com/Jeffrey-Ede/Electron-Micrograph-Denoiser | ['Jeffrey M. Ede'] | 2018-07-30 | null | null | null | null | ['2048'] | ['playing-games'] | [ 3.45953405e-01 -2.59396322e-02 6.05716228e-01 -3.72555912e-01
-1.18926620e+00 -2.00645819e-01 3.81860226e-01 7.42328689e-02
-1.03031671e+00 1.02261603e+00 4.60130572e-02 -1.02326311e-02
-1.33640483e-01 -8.34252000e-01 -8.81256282e-01 -1.02446258e+00
-1.55561287e-02 2.78660685e-01 3.59228313e-01 -3.77745293e-02
4.03439432e-01 5.39165914e-01 -1.12643504e+00 4.27249312e-01
1.65318087e-01 1.14459944e+00 3.99857491e-01 9.71447110e-01
5.54415107e-01 4.54274118e-01 -4.48815942e-01 -3.98973882e-01
1.75065890e-01 -2.85795212e-01 -6.99779153e-01 -2.93958545e-01
3.23656946e-01 -3.73416990e-01 -4.89685953e-01 9.80248213e-01
9.64552999e-01 -3.65598463e-02 7.80562043e-01 -5.12682319e-01
-9.86843348e-01 4.12652791e-01 -8.76827478e-01 8.12771916e-01
-9.86894667e-02 3.82788152e-01 4.72275078e-01 -9.30607796e-01
8.57891023e-01 1.07286537e+00 1.06960940e+00 6.50728226e-01
-1.65478814e+00 -5.38549185e-01 -6.96601272e-01 1.33637831e-01
-1.04452538e+00 -6.07894003e-01 3.02895695e-01 -2.84736246e-01
1.39049160e+00 1.05572432e-01 5.93006648e-02 9.49615836e-01
1.15362358e+00 -9.11192968e-03 1.30003786e+00 -1.19419821e-01
2.36293659e-01 -5.85339785e-01 1.14469901e-02 4.09417480e-01
-3.47067751e-02 -9.63418279e-03 -2.06111446e-01 -9.38650444e-02
9.56894577e-01 -2.82907821e-02 -1.79633811e-01 5.21802425e-01
-1.30018628e+00 4.52979296e-01 7.34057069e-01 3.62700939e-01
-6.72449768e-01 5.86470246e-01 3.62559915e-01 1.46506682e-01
5.54217994e-01 3.36849302e-01 -2.78387278e-01 2.19707307e-03
-7.79892027e-01 2.27830186e-01 2.01840743e-01 5.94864428e-01
6.85031414e-01 5.22404015e-02 2.30107084e-02 9.69568133e-01
1.19712204e-01 4.79279965e-01 4.52066302e-01 -1.23796606e+00
-1.21790625e-01 2.13489667e-01 -1.32344306e-01 -7.59147286e-01
-5.87674975e-01 -1.42059207e-01 -1.27865958e+00 5.87458611e-01
2.93622792e-01 -1.62945002e-01 -1.20004487e+00 1.52072287e+00
-1.28978565e-01 -7.49606118e-02 -2.72413582e-01 8.33656609e-01
9.20760453e-01 6.74315274e-01 6.60985196e-03 -2.24772498e-01
1.56638336e+00 -3.92062545e-01 -7.13322759e-01 -2.06470609e-01
-1.00556381e-01 -1.05192971e+00 7.76348472e-01 2.12749943e-01
-1.57270896e+00 -4.14827615e-01 -1.01845098e+00 -3.95438224e-01
-2.55694628e-01 -2.01803632e-02 -6.87459037e-02 -7.82145187e-02
-1.41248035e+00 9.99084055e-01 -1.16097510e+00 -3.29727143e-01
8.59807014e-01 6.42008722e-01 -4.38678443e-01 -4.13660705e-02
-7.05822468e-01 7.18244672e-01 2.10047774e-02 -6.58055767e-02
-1.05901170e+00 -9.36211109e-01 -5.34696460e-01 -9.84321162e-02
-1.74779609e-01 -6.65736794e-01 1.14646924e+00 -2.89529651e-01
-1.26276636e+00 9.80102956e-01 -1.46477550e-01 -5.67120910e-01
2.81345904e-01 1.23935156e-01 -2.22062215e-01 4.31300879e-01
3.95824909e-01 6.85159504e-01 4.14868712e-01 -1.24384785e+00
-1.57139972e-01 -3.46321374e-01 -3.62976819e-01 -1.89175382e-01
2.32395336e-01 3.18748862e-01 -2.23906130e-01 -5.92016280e-01
1.24778643e-01 -5.56519866e-01 -2.56492227e-01 5.25529236e-02
-4.21090424e-01 3.81444335e-01 7.00231552e-01 -9.15720046e-01
6.20668888e-01 -2.08785367e+00 -8.85973573e-02 6.33947691e-03
5.00292361e-01 -6.97217882e-02 -2.39222683e-02 5.54282010e-01
-3.95201951e-01 -2.04069223e-02 -5.99119902e-01 -6.37420654e-01
-1.53692201e-01 -6.19782992e-02 1.18398167e-01 7.79244959e-01
8.24519470e-02 8.88156354e-01 -7.04606295e-01 -9.77174640e-02
3.96462709e-01 9.68211830e-01 -3.31756413e-01 -4.63034026e-02
2.30529457e-01 2.84496218e-01 2.45639235e-01 6.34446084e-01
8.26079309e-01 -6.22386158e-01 -1.12177230e-01 -5.06915390e-01
-2.21059859e-01 -1.84533119e-01 -6.77310646e-01 1.48722005e+00
-2.52940953e-01 7.42262363e-01 5.09203434e-01 -5.73292196e-01
7.23723769e-01 1.44531637e-01 4.82659310e-01 -7.90763557e-01
4.26304221e-01 3.07198256e-01 -3.71928476e-02 -1.42547518e-01
3.85729134e-01 -7.43514657e-01 1.31620824e-01 3.25523168e-01
1.92230135e-01 -1.70511678e-01 1.72985241e-01 2.05240488e-01
1.79334950e+00 -4.23592776e-01 1.27512395e-01 -6.29662216e-01
3.42430115e-01 -2.50463128e-01 4.57434386e-01 5.47920525e-01
-2.07716405e-01 1.13752258e+00 4.83099133e-01 -3.83954108e-01
-1.56658959e+00 -1.24845374e+00 -7.47119367e-01 7.38098919e-01
-7.35804960e-02 -1.04542643e-01 -8.17200541e-01 -4.61857282e-02
-1.71604440e-01 4.40549552e-01 -8.98780882e-01 8.04607943e-02
-5.80484331e-01 -1.32124674e+00 5.93597293e-01 5.20920753e-01
6.63009822e-01 -1.36910808e+00 -5.89123726e-01 2.67904431e-01
9.52661037e-02 -9.29847002e-01 -3.43697369e-01 5.80347419e-01
-5.55299520e-01 -1.00775373e+00 -7.32595026e-01 -7.47163296e-01
8.92075598e-01 -2.25572869e-01 1.07268775e+00 9.34951305e-02
-5.39008379e-01 6.68933392e-02 -5.53921424e-02 -1.34564161e-01
-4.21893746e-01 -3.30710113e-01 1.31495968e-01 -4.51638252e-01
1.88170955e-01 -9.01422739e-01 -1.14103174e+00 3.05508226e-01
-1.20572329e+00 -1.78434804e-01 5.90330660e-01 1.04731512e+00
1.21575332e+00 7.80461356e-02 4.95346308e-01 -4.72917199e-01
6.31075561e-01 -3.47008198e-01 -4.32412207e-01 -3.26079935e-01
-5.04678607e-01 -3.18914503e-01 9.97295320e-01 -2.10663658e-02
-9.23711061e-01 -2.93710321e-01 -6.12776935e-01 -3.67749214e-01
-3.07052404e-01 2.96474304e-02 1.48524091e-01 -1.85313836e-01
8.58898938e-01 9.34923417e-04 3.17368470e-02 -4.36070174e-01
-1.39737390e-02 4.48825806e-01 1.06643736e+00 -2.89630920e-01
4.09671605e-01 9.40430343e-01 5.63770067e-03 -6.61976576e-01
-1.31347924e-01 -1.03461415e-01 -2.80928433e-01 -1.27938390e-01
1.34953725e+00 -5.50222456e-01 -8.67263734e-01 8.86533499e-01
-1.04299390e+00 -5.10854781e-01 -3.54156464e-01 3.77533138e-01
-6.72214210e-01 3.76970142e-01 -1.46169066e+00 -1.83290601e-01
-7.92829812e-01 -1.36718225e+00 1.07425654e+00 2.26160958e-01
-1.41189218e-01 -8.14600348e-01 -4.68596295e-02 9.66844782e-02
7.68707812e-01 5.48260868e-01 8.78424644e-01 -2.29401946e-01
-2.50323981e-01 -1.94993094e-01 -6.27083540e-01 6.18831635e-01
1.32070243e-01 -8.41929242e-02 -1.00680077e+00 -5.08764625e-01
3.76809806e-01 -4.22652900e-01 1.01808321e+00 9.43302155e-01
1.25945103e+00 -4.50079627e-02 -1.71626151e-01 7.60223866e-01
1.78581119e+00 3.36305797e-01 1.15454078e+00 4.54919934e-01
3.92592013e-01 8.25284347e-02 -2.46635944e-01 4.31728542e-01
-3.92607637e-02 3.87356341e-01 6.19504750e-01 -7.89460391e-02
-4.29785997e-01 4.99226719e-01 8.52801874e-02 7.63191640e-01
-2.95715153e-01 -2.80891627e-01 -1.00135899e+00 5.29637396e-01
-1.26172960e+00 -1.10721159e+00 -2.91389853e-01 1.81104803e+00
1.00666213e+00 2.22275838e-01 -3.13412756e-01 -1.10229077e-02
7.67574370e-01 1.33722559e-01 -6.41729176e-01 -3.65790814e-01
-3.27135265e-01 5.81878364e-01 7.10073054e-01 6.05882764e-01
-8.90908957e-01 3.23405683e-01 6.20468283e+00 9.82358932e-01
-8.89572680e-01 1.71343163e-01 1.02597272e+00 -2.73962915e-01
-1.07278928e-01 -4.37451035e-01 -5.50212979e-01 8.17542017e-01
1.10361516e+00 -9.12149921e-02 4.58924741e-01 3.71086955e-01
1.12377211e-01 -3.60936970e-01 -8.56088400e-01 9.54830527e-01
-1.29618570e-01 -1.66034079e+00 -5.43401957e-01 3.37802470e-01
7.05345988e-01 7.01693237e-01 -1.36442073e-02 -1.56646982e-01
3.49516988e-01 -1.16621602e+00 4.62279916e-01 8.16756427e-01
1.16455531e+00 -8.48506331e-01 1.02764750e+00 -3.58197056e-02
-8.93770218e-01 2.70288020e-01 -6.42996609e-01 2.30664283e-01
5.02318978e-01 9.44912910e-01 -4.27793324e-01 3.42456520e-01
1.23613167e+00 5.21830976e-01 -3.82921278e-01 5.86236715e-01
3.18190187e-01 2.84435034e-01 -4.69317436e-01 3.87757242e-01
8.54567438e-02 -3.15180458e-02 3.58641386e-01 1.30788934e+00
3.61815542e-01 2.75136799e-01 -5.91736972e-01 9.11999166e-01
-4.16654170e-01 -5.18743873e-01 -2.35954553e-01 3.46940309e-01
3.31512451e-01 1.36624646e+00 -1.25297499e+00 -1.78004816e-01
-1.87929004e-01 9.10112023e-01 6.61200657e-02 3.07724893e-01
-8.57326210e-01 -7.25839913e-01 5.09846807e-01 2.99233049e-01
3.92818362e-01 8.35955236e-03 -4.69619095e-01 -6.64494157e-01
-2.72060901e-01 -5.94794393e-01 5.22627793e-02 -1.12070143e+00
-1.83236861e+00 7.11919785e-01 -2.55444292e-02 -6.95042372e-01
1.03903309e-01 -8.19230139e-01 -7.07919180e-01 9.45627153e-01
-1.08157432e+00 -6.85202539e-01 -1.96068019e-01 3.45294416e-01
2.75760531e-01 2.90425513e-02 7.79266655e-01 5.09920001e-01
-3.98894370e-01 3.04148048e-01 6.21310830e-01 1.69821910e-03
6.34845495e-01 -1.33239961e+00 6.00594223e-01 7.66829550e-01
-7.64405251e-01 6.06561184e-01 8.36217403e-01 -6.93187237e-01
-1.10338187e+00 -1.01367545e+00 7.36042440e-01 -4.03948665e-01
7.13077068e-01 -1.40832275e-01 -1.22124243e+00 6.64070189e-01
7.09514320e-01 2.78490186e-01 5.18447578e-01 -6.96874142e-01
1.67912886e-01 -7.37398639e-02 -1.68196738e+00 4.13588971e-01
6.73637867e-01 -4.73670930e-01 -3.45804602e-01 2.71077901e-01
2.78573841e-01 -5.66228747e-01 -1.40578175e+00 4.34907794e-01
4.08488393e-01 -1.20271230e+00 1.12662363e+00 2.56608546e-01
9.21420693e-01 -3.00092995e-01 -2.43161812e-01 -1.12889779e+00
-5.71370065e-01 -2.18514323e-01 2.96035528e-01 8.70666146e-01
1.90132082e-01 -7.16368794e-01 6.89880311e-01 2.27095366e-01
-5.99036455e-01 -1.02981222e+00 -1.34469521e+00 -4.38640445e-01
2.85658598e-01 -3.02240819e-01 2.58981168e-01 5.39036036e-01
-3.71745944e-01 8.53227749e-02 -1.42058730e-02 -8.46539810e-02
9.50733006e-01 -4.27673578e-01 9.68971029e-02 -6.27761483e-01
-1.87168449e-01 -4.29986507e-01 -5.40023625e-01 -6.23027503e-01
-1.14715055e-01 -7.75115848e-01 2.45072320e-01 -1.64136255e+00
4.51725215e-01 1.45921588e-01 -2.96806961e-01 5.47833443e-01
1.08135030e-01 9.90153551e-01 -1.67871386e-01 1.67034835e-01
-2.95246750e-01 4.83835250e-01 1.17835760e+00 2.69243419e-02
4.22926456e-01 -5.66230953e-01 -4.79997724e-01 7.12754488e-01
9.49519217e-01 -5.60853124e-01 9.12620313e-03 -5.65453053e-01
1.50744483e-01 -2.68050693e-02 7.62687564e-01 -1.31481850e+00
1.52053058e-01 2.90509880e-01 8.49308550e-01 -6.41289353e-01
5.01014054e-01 -4.48758423e-01 4.66228604e-01 4.43719923e-01
-5.81477210e-02 4.77394700e-01 4.25154507e-01 3.43503773e-01
9.69121158e-02 -2.02409521e-01 1.33360958e+00 -3.11100036e-01
-3.58262986e-01 1.61351964e-01 -5.18168807e-01 -1.15741692e-01
8.94301414e-01 -2.45762900e-01 -8.40134144e-01 -2.59477705e-01
-8.16268146e-01 -1.05958097e-01 9.18926656e-01 -2.38550201e-01
8.76817822e-01 -1.25801027e+00 -7.27990270e-01 8.12459588e-02
-4.03503895e-01 1.08499102e-01 7.23003089e-01 1.02161407e+00
-9.35538709e-01 -5.44336922e-02 -6.31768882e-01 -4.88010466e-01
-1.02828419e+00 1.14071682e-01 4.87285078e-01 -1.01852179e-01
-6.24280095e-01 1.12740123e+00 1.71048567e-01 -2.03232303e-01
-3.99082392e-01 -3.55938613e-01 2.57373750e-01 -4.10906732e-01
6.66568339e-01 7.55353928e-01 3.33458215e-01 -6.57389343e-01
-3.91788572e-01 6.02182388e-01 -2.12957799e-01 -7.95184448e-02
1.85712969e+00 -1.34648666e-01 -5.16829848e-01 1.98109016e-01
1.49098074e+00 -2.00952291e-01 -1.47985458e+00 1.09210789e-01
-4.13675338e-01 -3.12824957e-02 1.05281398e-01 -8.20709825e-01
-1.20329118e+00 5.88667691e-01 8.07880104e-01 2.45708659e-01
1.21077418e+00 2.18805626e-01 9.05579925e-01 3.65634598e-02
1.12994611e-01 -8.57265890e-01 3.84983309e-02 6.05210841e-01
1.00455630e+00 -9.96737957e-01 4.00929898e-01 6.83941245e-02
-1.69209197e-01 1.05737340e+00 4.91852552e-01 -6.81530654e-01
6.89969897e-01 8.35509241e-01 6.28488734e-02 -6.80592179e-01
-7.90277719e-01 3.40499759e-01 -9.66418386e-02 5.18663585e-01
6.05148613e-01 -2.89999574e-01 -3.45252663e-01 3.41947198e-01
-2.74654150e-01 1.27379093e-02 6.30836666e-01 9.66770768e-01
-3.17842901e-01 -7.80747831e-01 -2.93659866e-01 8.94372821e-01
-9.98044550e-01 -1.34246305e-01 2.25763083e-01 6.43376827e-01
2.11218297e-01 6.69658720e-01 2.97454447e-01 -2.42075264e-01
3.11546922e-01 -2.46474713e-01 4.04413432e-01 -3.45669985e-01
-6.62017167e-01 3.04332078e-01 -1.83686912e-01 -6.51898324e-01
-4.70940411e-01 -5.65186501e-01 -1.61274087e+00 -9.09122765e-01
9.31314006e-02 -2.51722366e-01 5.51305830e-01 5.30055285e-01
2.87174582e-01 9.34072435e-01 1.72622561e-01 -1.31145561e+00
-1.83399141e-01 -1.09758639e+00 -7.21092582e-01 3.44430774e-01
3.40227842e-01 -2.74757624e-01 -6.81953192e-01 3.56048346e-01] | [13.124985694885254, -2.6573493480682373] |
8380a505-a11b-453f-9905-92dc3db19ddd | goal-conditioned-predictive-coding-as-an | 2307.03406 | null | https://arxiv.org/abs/2307.03406v1 | https://arxiv.org/pdf/2307.03406v1.pdf | Goal-Conditioned Predictive Coding as an Implicit Planner for Offline Reinforcement Learning | Recent work has demonstrated the effectiveness of formulating decision making as a supervised learning problem on offline-collected trajectories. However, the benefits of performing sequence modeling on trajectory data is not yet clear. In this work we investigate if sequence modeling has the capability to condense trajectories into useful representations that can contribute to policy learning. To achieve this, we adopt a two-stage framework that first summarizes trajectories with sequence modeling techniques, and then employs these representations to learn a policy along with a desired goal. This design allows many existing supervised offline RL methods to be considered as specific instances of our framework. Within this framework, we introduce Goal-Conditioned Predicitve Coding (GCPC), an approach that brings powerful trajectory representations and leads to performant policies. We conduct extensive empirical evaluations on AntMaze, FrankaKitchen and Locomotion environments, and observe that sequence modeling has a significant impact on some decision making tasks. In addition, we demonstrate that GCPC learns a goal-conditioned latent representation about the future, which serves as an "implicit planner", and enables competitive performance on all three benchmarks. | ['Chen Sun', 'Shijie Wang', 'Ce Zhang', 'Zilai Zeng'] | 2023-07-07 | null | null | null | null | ['offline-rl', 'decision-making'] | ['playing-games', 'reasoning'] | [ 9.94202420e-02 9.35770795e-02 -9.13572848e-01 -2.79034883e-01
-6.13773048e-01 -6.78911030e-01 1.02510190e+00 -1.65774468e-02
-1.73045278e-01 8.47898602e-01 7.28302300e-01 -6.78474665e-01
-4.64731865e-02 -6.87547386e-01 -7.81450927e-01 -7.09181964e-01
-3.64154965e-01 4.62594628e-01 -2.59693041e-02 -2.98894763e-01
3.45006943e-01 5.15707076e-01 -1.45178950e+00 4.11487699e-01
8.56096804e-01 5.24361968e-01 3.17582190e-01 6.30021930e-01
3.68388928e-02 1.30650580e+00 -2.18904942e-01 3.16617712e-02
2.95069933e-01 -3.79913479e-01 -9.65705276e-01 -1.66509375e-02
-2.01927528e-01 -6.34982288e-01 -5.03881335e-01 5.49320221e-01
1.28047436e-01 6.62668347e-01 9.56811011e-01 -1.22437572e+00
-4.94517207e-01 9.46704865e-01 -1.67478710e-01 3.57266068e-02
3.51443321e-01 4.93555963e-01 1.05490398e+00 -4.35267389e-01
7.11442888e-01 1.37370312e+00 3.29347670e-01 6.20467365e-01
-1.25715566e+00 -4.71534491e-01 6.33225024e-01 2.04125687e-01
-8.70117664e-01 -5.56918800e-01 6.61283970e-01 -4.80010629e-01
1.19506252e+00 1.74042806e-01 5.72384953e-01 1.46604633e+00
6.59157187e-02 1.43443084e+00 8.76000524e-01 -3.29626858e-01
5.17086267e-01 -1.23778321e-01 9.81267244e-02 5.44115424e-01
-7.78450221e-02 6.69340312e-01 -3.54798853e-01 -1.06658064e-01
6.76402211e-01 1.64929435e-01 -2.50727147e-01 -5.47546804e-01
-1.24270368e+00 8.76215458e-01 4.09350008e-01 9.17613357e-02
-3.78724962e-01 6.67019367e-01 3.71306777e-01 3.27945113e-01
4.29342508e-01 5.00376165e-01 -1.86787158e-01 -7.09316611e-01
-8.07889760e-01 6.41851723e-01 8.28894198e-01 1.13582551e+00
6.66260183e-01 2.46282980e-01 -4.94106323e-01 3.66303235e-01
1.56636655e-01 3.63895059e-01 5.19386828e-01 -1.38029623e+00
5.59768379e-01 3.87866676e-01 5.08592606e-01 -8.14094663e-01
-3.02544087e-01 -2.30801046e-01 -2.17962548e-01 2.57995784e-01
1.84370443e-01 -1.82357207e-01 -9.37285841e-01 1.85364962e+00
4.25471105e-02 4.17888165e-01 3.35915327e-01 9.34554040e-01
6.70969933e-02 8.54647279e-01 2.97662914e-01 -1.63954854e-01
5.40975630e-01 -1.43867326e+00 -4.97773588e-01 -1.76011294e-01
1.06409240e+00 -5.74846640e-02 1.15262938e+00 3.80087107e-01
-9.65172410e-01 -4.46839631e-01 -8.96186233e-01 -7.66922394e-03
-1.63662866e-01 2.31900483e-01 9.48206544e-01 4.03291166e-01
-1.07056630e+00 9.32934880e-01 -1.38503575e+00 -3.89245689e-01
3.72169942e-01 2.50337303e-01 3.07866689e-02 -2.96908468e-02
-6.91691220e-01 1.02180552e+00 5.89907110e-01 -3.26375604e-01
-1.55779922e+00 -4.81008917e-01 -8.32068384e-01 8.22594315e-02
6.00782871e-01 -6.23996079e-01 1.55100799e+00 -7.76703775e-01
-1.80770802e+00 2.51294583e-01 -4.29060489e-01 -8.50027025e-01
5.77055514e-01 -3.57011199e-01 -1.48248792e-01 8.42873603e-02
4.08868231e-02 7.93830156e-01 7.77747631e-01 -1.19936371e+00
-8.85241091e-01 -1.12652965e-01 4.41759974e-01 3.99507821e-01
-2.00047627e-01 -2.14918539e-01 -3.04954797e-01 -6.55499995e-01
-2.96373457e-01 -1.07558370e+00 -6.57347620e-01 -3.50319505e-01
-2.64011145e-01 -4.24875677e-01 5.55256248e-01 -5.20851970e-01
1.35651124e+00 -1.86687541e+00 4.75254864e-01 2.12720744e-02
6.00598902e-02 2.44046316e-01 -1.74324706e-01 8.88380647e-01
2.06838429e-01 2.38070756e-01 -2.41347790e-01 -5.25935113e-01
1.78816363e-01 4.95398521e-01 -9.16608274e-01 3.85316700e-01
9.05194208e-02 1.18272078e+00 -1.25369549e+00 1.83092356e-02
3.50578040e-01 -8.86873081e-02 -7.34431267e-01 2.03819230e-01
-7.63836980e-01 7.89440155e-01 -7.70281494e-01 5.49044788e-01
-3.67529020e-02 -1.74988866e-01 3.72827947e-01 5.34558177e-01
-3.06772292e-01 4.83047277e-01 -5.66456318e-01 1.98987186e+00
-2.70232737e-01 5.33633113e-01 -3.84945959e-01 -1.04787529e+00
7.53814757e-01 1.30400211e-01 6.95762277e-01 -6.04817033e-01
-9.25559253e-02 4.80135418e-02 -2.08196476e-01 -4.49766606e-01
7.75129199e-01 9.96899009e-02 -1.89539753e-02 7.19755173e-01
-9.37382430e-02 1.74934104e-01 2.55890340e-01 2.29367673e-01
1.07921374e+00 7.91158676e-01 2.61450499e-01 -4.69526947e-02
8.09379891e-02 4.80383962e-01 4.66245681e-01 8.97929490e-01
-1.38434678e-01 1.92607239e-01 4.93297905e-01 -4.00065660e-01
-1.04694629e+00 -8.40389073e-01 4.01059389e-01 1.30957198e+00
3.59628089e-02 -6.01143539e-01 -3.44461858e-01 -7.04572558e-01
2.18912706e-01 1.29983151e+00 -6.63082302e-01 -3.06530833e-01
-8.33186090e-01 -3.62283409e-01 5.81543684e-01 9.07369912e-01
6.61402792e-02 -1.07985258e+00 -8.54645073e-01 3.13158959e-01
-1.45473361e-01 -8.15687656e-01 -2.73943603e-01 1.47895202e-01
-1.04098272e+00 -9.33056533e-01 -6.47126913e-01 -4.55425084e-01
3.89993727e-01 3.68735403e-01 9.02938068e-01 -1.21901780e-01
2.67064959e-01 5.99972427e-01 -6.90096140e-01 -2.23806202e-01
-3.92318100e-01 3.29047978e-01 1.52572662e-01 -2.61825234e-01
9.10319537e-02 -6.85019076e-01 -3.71914893e-01 6.43224493e-02
-5.29035389e-01 3.05698365e-01 4.37628984e-01 7.48942971e-01
3.49041462e-01 -2.85656780e-01 6.30515218e-01 -7.06565619e-01
9.55942035e-01 -8.39853823e-01 -4.57425803e-01 1.95313260e-01
-7.23355830e-01 3.30400437e-01 7.36288607e-01 -4.46787894e-01
-1.01695108e+00 4.25309390e-02 -4.87012900e-02 -7.00168550e-01
-1.74544051e-01 6.34270966e-01 2.25317150e-01 3.32180858e-01
6.49711668e-01 4.35105473e-01 8.99080485e-02 -5.08268058e-01
9.05122876e-01 4.07835960e-01 3.84003848e-01 -1.07261050e+00
5.05501449e-01 6.26181483e-01 -1.63013026e-01 -5.47132611e-01
-5.90576172e-01 -5.43323040e-01 -4.55448180e-01 -8.92419890e-02
6.47851467e-01 -8.46751928e-01 -8.69933665e-01 -1.99290998e-02
-7.97533572e-01 -1.07692492e+00 -4.51543182e-01 5.25484681e-01
-1.34568799e+00 2.52793014e-01 -5.09183288e-01 -1.20895922e+00
6.18564058e-03 -1.29429615e+00 8.82630587e-01 -2.37392392e-02
-3.03003162e-01 -1.06548226e+00 2.99197167e-01 -1.04869194e-01
2.37537861e-01 3.76047224e-01 9.45224464e-01 -6.72987163e-01
-7.67148197e-01 2.24506736e-01 1.45205706e-01 -1.25359406e-03
-8.57301429e-02 -3.09709638e-01 -7.53764927e-01 -4.66812491e-01
-1.68385834e-01 -5.53534508e-01 1.11456335e+00 3.31742495e-01
1.20732045e+00 -4.60955769e-01 -6.43233240e-01 7.51002192e-01
1.17040849e+00 4.81529623e-01 4.51140314e-01 3.44157428e-01
4.97425497e-01 6.01108134e-01 1.02650774e+00 5.60785651e-01
5.85167706e-01 8.29082072e-01 4.31171477e-01 3.63997370e-01
1.31126791e-02 -9.20012474e-01 7.06312358e-01 5.38839877e-01
-1.93957433e-01 -3.57237279e-01 -9.93110776e-01 6.42792463e-01
-2.47058487e+00 -1.10386980e+00 1.48982346e-01 1.75456357e+00
6.00298166e-01 -2.04225600e-01 4.64177877e-01 -6.09272458e-02
8.72469172e-02 2.68980592e-01 -8.95406723e-01 -2.97463059e-01
3.70347559e-01 1.47461900e-02 5.65167665e-01 4.25285727e-01
-1.02713203e+00 1.35260189e+00 7.13433981e+00 7.44204402e-01
-1.09478295e+00 1.19684287e-03 4.51478332e-01 -9.26409364e-02
-5.01477838e-01 1.71364501e-01 -7.21152604e-01 4.07629967e-01
1.12510633e+00 -4.06258911e-01 8.17514360e-01 1.01977861e+00
5.64229727e-01 1.01202615e-01 -1.37500381e+00 7.01698124e-01
-3.15369785e-01 -1.50600648e+00 1.54387042e-01 2.44136289e-01
7.45819628e-01 6.99339360e-02 2.83559144e-01 8.47722888e-01
9.90116477e-01 -1.28063166e+00 1.02950156e+00 5.77441037e-01
7.01333880e-01 -9.04210091e-01 1.37489244e-01 9.08388376e-01
-1.03782415e+00 -6.16483808e-01 -4.45459217e-01 -4.33884919e-01
2.41017699e-01 -2.42059454e-01 -1.19282031e+00 8.01735044e-01
2.24005371e-01 1.30091166e+00 -1.65928945e-01 9.08752859e-01
-4.48792189e-01 9.69965875e-01 -8.05840045e-02 -3.05186287e-02
7.29179621e-01 -3.51282895e-01 6.06187344e-01 9.83404636e-01
2.97599047e-01 1.22840196e-01 6.54229045e-01 1.05737424e+00
2.31022999e-01 -9.61870104e-02 -9.18731689e-01 -6.25611067e-01
3.53432119e-01 6.13521159e-01 -5.31151116e-01 -3.04296792e-01
-9.33318362e-02 8.49397063e-01 6.23273492e-01 7.15465963e-01
-7.99374938e-01 2.96543270e-01 8.24829996e-01 -1.83524355e-01
2.34627098e-01 -6.93637490e-01 -3.48921835e-01 -1.30383599e+00
-3.45232695e-01 -9.28504109e-01 1.28415853e-01 -5.40322840e-01
-6.86200500e-01 1.17234074e-01 2.13896185e-01 -1.28525400e+00
-8.20986032e-01 -3.97211611e-01 -6.59325123e-01 6.31486833e-01
-1.60274565e+00 -1.26524687e+00 1.12347618e-01 4.91934925e-01
7.68185973e-01 -2.02606723e-01 7.77156234e-01 -7.67511129e-02
-5.63619077e-01 3.17858428e-01 2.93788433e-01 -6.52628541e-02
2.52415091e-01 -1.31768107e+00 5.76774776e-01 7.85933912e-01
2.50145555e-01 8.33048940e-01 6.13673568e-01 -7.63063431e-01
-1.52047658e+00 -1.30728936e+00 4.75548923e-01 -7.27379322e-01
5.56318820e-01 -1.26821280e-01 -5.75716376e-01 1.19893539e+00
-4.73759770e-02 -4.77022260e-01 5.05391538e-01 2.44129732e-01
-1.82436649e-02 3.87484103e-01 -6.59197628e-01 9.61388648e-01
1.43543673e+00 -3.77268463e-01 -6.58636093e-01 3.01633179e-01
1.07270133e+00 -4.17472988e-01 -5.54033935e-01 1.77657112e-01
5.12761533e-01 -6.97228074e-01 9.84859824e-01 -1.22878754e+00
4.63038027e-01 -4.65246290e-02 -2.13960409e-01 -1.44638050e+00
-3.01586002e-01 -8.45719993e-01 -6.61143005e-01 9.11356807e-01
2.65547842e-01 -4.51211631e-01 7.76097357e-01 5.40408432e-01
-5.34746528e-01 -9.26495969e-01 -3.49375099e-01 -1.02567577e+00
4.22423631e-01 -6.66301191e-01 8.76973867e-01 7.37858653e-01
3.27603996e-01 -1.29160672e-01 -7.61714697e-01 -4.04789448e-02
3.12962383e-01 3.99185658e-01 9.83447850e-01 -7.20590472e-01
-3.09048206e-01 -5.51733553e-01 1.28927734e-02 -1.72454190e+00
5.80950439e-01 -1.23890638e+00 4.54487167e-02 -1.65842211e+00
-8.52091238e-02 -6.69432104e-01 -2.77288079e-01 4.83133525e-01
9.33441818e-02 -6.25546575e-01 4.59041208e-01 4.90344822e-01
-7.08049178e-01 9.32697952e-01 1.26569712e+00 -4.90489714e-02
-4.58130926e-01 2.15197206e-01 -7.11678922e-01 5.75613916e-01
9.80999708e-01 -2.95902431e-01 -9.26286221e-01 -4.75133836e-01
-5.22218756e-02 3.83056015e-01 1.59678712e-01 -9.89772320e-01
2.48015895e-01 -7.23139346e-01 -4.72052619e-02 -6.15690768e-01
5.06004035e-01 -5.48376203e-01 -1.37354687e-01 5.02562284e-01
-8.22152317e-01 -2.86186878e-02 2.54061855e-02 1.05705976e+00
-3.05851791e-02 -2.54628975e-02 2.18945161e-01 -1.79649279e-01
-1.12772369e+00 3.68443698e-01 -6.22763515e-01 -7.94491172e-02
1.19617665e+00 -1.16840884e-01 -6.24160804e-02 -6.77856803e-01
-7.77644277e-01 6.58352077e-01 6.55208766e-01 5.02942443e-01
6.35352075e-01 -1.29209530e+00 -4.25209731e-01 6.89668506e-02
3.76809463e-02 -3.04326117e-01 -2.24808708e-01 7.35104442e-01
-2.87893593e-01 8.40155542e-01 -2.05679119e-01 -5.41523099e-01
-6.14171624e-01 7.98452854e-01 1.24867834e-01 -5.15580058e-01
-8.93188119e-01 5.10480583e-01 -5.54269813e-02 -4.78339583e-01
4.88243759e-01 -7.77209520e-01 -4.37163502e-01 -1.98684931e-01
3.56600165e-01 5.03224492e-01 -4.92853463e-01 -3.60978544e-01
5.70641980e-02 1.55404201e-02 6.54680207e-02 -4.68178660e-01
1.22883022e+00 -1.54645801e-01 4.07395750e-01 5.61845064e-01
8.19544435e-01 -3.05142343e-01 -1.76520526e+00 -9.37550887e-02
4.00768906e-01 -4.60107565e-01 -1.22285016e-01 -7.16777205e-01
-5.88696837e-01 9.45356846e-01 8.19794685e-02 -7.21645132e-02
7.92872965e-01 -2.99739867e-01 8.27854395e-01 6.62473023e-01
9.38729525e-01 -9.17638540e-01 8.92241672e-02 7.31699109e-01
7.79624164e-01 -1.01368594e+00 -3.97884458e-01 -4.20752652e-02
-9.41912293e-01 1.01625967e+00 3.89036626e-01 -2.92952597e-01
3.50991309e-01 -3.68303736e-03 -3.87275249e-01 1.16723694e-01
-1.19763625e+00 -5.38361967e-01 1.47023335e-01 7.86184430e-01
1.12937838e-01 3.49623382e-01 -1.67348668e-01 5.44674575e-01
-1.71358377e-01 3.68205130e-01 3.28665286e-01 1.10167968e+00
-6.65081203e-01 -1.26971543e+00 8.89754593e-02 3.71585071e-01
-4.19644937e-02 2.43362740e-01 -2.25188479e-01 5.88748455e-01
-2.58224636e-01 1.09070051e+00 -1.81486934e-01 -6.84859872e-01
-7.00874254e-03 4.72857319e-02 3.44915479e-01 -7.66827762e-01
-4.96121824e-01 -1.90189809e-01 2.86747009e-01 -1.04914653e+00
-3.02517980e-01 -6.69105291e-01 -1.38691843e+00 -3.86635989e-01
2.52776265e-01 1.68665200e-01 3.69836479e-01 1.08476448e+00
4.11860198e-01 5.20859957e-01 4.32138890e-01 -9.18973327e-01
-9.82583582e-01 -7.58034408e-01 -2.40922764e-01 3.28416228e-01
4.24093843e-01 -8.91723812e-01 1.60054117e-02 2.55782413e-03] | [4.148338317871094, 1.729852557182312] |
1aaaae61-a21a-4c91-bacc-b31470594349 | annotation-cost-efficient-active-learning-for | 2306.11605 | null | https://arxiv.org/abs/2306.11605v2 | https://arxiv.org/pdf/2306.11605v2.pdf | Annotation Cost Efficient Active Learning for Content Based Image Retrieval | Deep metric learning (DML) based methods have been found very effective for content-based image retrieval (CBIR) in remote sensing (RS). For accurately learning the model parameters of deep neural networks, most of the DML methods require a high number of annotated training images, which can be costly to gather. To address this problem, in this paper we present an annotation cost efficient active learning (AL) method (denoted as ANNEAL). The proposed method aims to iteratively enrich the training set by annotating the most informative image pairs as similar or dissimilar, while accurately modelling a deep metric space. This is achieved by two consecutive steps. In the first step the pairwise image similarity is modelled based on the available training set. Then, in the second step the most uncertain and diverse (i.e., informative) image pairs are selected to be annotated. Unlike the existing AL methods for CBIR, at each AL iteration of ANNEAL a human expert is asked to annotate the most informative image pairs as similar/dissimilar. This significantly reduces the annotation cost compared to annotating images with land-use/land cover class labels. Experimental results show the effectiveness of our method. The code of ANNEAL is publicly available at https://git.tu-berlin.de/rsim/ANNEAL. | ['Begüm Demir', 'Lars Möllenbrok', 'Gencer Sumbul', 'Genc Hoxha', 'Julia Henkel'] | 2023-06-20 | null | null | null | null | ['metric-learning', 'content-based-image-retrieval', 'active-learning', 'metric-learning', 'active-learning'] | ['computer-vision', 'computer-vision', 'methodology', 'methodology', 'natural-language-processing'] | [ 2.81444132e-01 -6.18985444e-02 8.01521167e-02 -6.63887620e-01
-1.24073946e+00 -5.99486053e-01 4.47201341e-01 5.08583486e-01
-7.06943393e-01 6.27182186e-01 -2.61942387e-01 -1.30701348e-01
-4.65109348e-01 -9.90555048e-01 -4.21359628e-01 -9.95733798e-01
-5.13198897e-02 7.83964157e-01 -9.35145095e-02 3.35904360e-02
1.33152381e-01 6.65057778e-01 -1.62424695e+00 1.75167620e-02
1.03358972e+00 1.27449203e+00 6.10285401e-01 4.33983773e-01
-2.29326814e-01 7.84518778e-01 -1.90712437e-01 -2.97682405e-01
5.74238539e-01 -3.61873329e-01 -1.03127646e+00 -6.20013699e-02
3.03355157e-01 -5.65069973e-01 9.39034671e-02 1.18777740e+00
7.22446740e-01 4.82627243e-01 6.64091349e-01 -1.16367877e+00
-3.14261794e-01 4.12551969e-01 -4.64197218e-01 -2.67475396e-02
-1.80279776e-01 -3.73933852e-01 1.27135158e+00 -1.41134107e+00
3.93535733e-01 9.76213694e-01 5.06706238e-01 2.11381137e-01
-1.04587841e+00 -7.14774191e-01 7.55321160e-02 4.58444744e-01
-1.80768704e+00 -4.71061945e-01 9.20569599e-01 -4.82612163e-01
2.17620999e-01 3.78054500e-01 5.55632889e-01 3.38836461e-01
-5.30978024e-01 1.03289568e+00 8.02301764e-01 -4.56098825e-01
6.48003280e-01 6.20628484e-02 8.62644762e-02 4.17519420e-01
-3.75144631e-02 -1.09870724e-01 -1.65653229e-02 -2.54540682e-01
4.83887970e-01 1.81057572e-01 -2.33867377e-01 -4.29342866e-01
-9.65763271e-01 1.07579935e+00 9.04054463e-01 3.50241333e-01
-6.88259721e-01 4.89403866e-02 2.44531602e-01 1.90032363e-01
9.35573518e-01 4.30771083e-01 -1.67109728e-01 3.71263146e-01
-1.23034394e+00 4.20729443e-02 5.48168600e-01 5.79571545e-01
1.31694400e+00 -6.06570661e-01 -6.52646199e-02 1.22451687e+00
6.03710532e-01 4.03451294e-01 1.09458655e-01 -9.50423062e-01
2.90756792e-01 7.18706727e-01 3.64953876e-01 -1.23795009e+00
-1.88313752e-01 -2.91057765e-01 -1.00588787e+00 1.30487695e-01
-8.18055030e-03 -7.60011794e-03 -8.11758757e-01 1.25620830e+00
3.63129377e-01 -9.51434765e-03 -8.36323127e-02 1.01282513e+00
9.31209624e-01 8.35928738e-01 2.12601572e-02 -6.35144413e-02
8.79889250e-01 -7.33251095e-01 -4.53049004e-01 -2.57007152e-01
5.24587512e-01 -7.37882435e-01 8.21955323e-01 3.15077119e-02
-7.43616760e-01 -5.01339912e-01 -8.22638392e-01 1.54047543e-02
-3.96292090e-01 4.42298323e-01 4.40326750e-01 2.38518506e-01
-1.25100803e+00 2.85460591e-01 -8.72855484e-01 -1.73212215e-01
7.38224089e-01 2.55617023e-01 -4.05558199e-01 -2.32061073e-01
-1.34031153e+00 6.51856661e-01 6.90451562e-01 6.23226225e-01
-9.67823982e-01 -5.85589170e-01 -8.78159165e-01 1.02275781e-01
1.85630873e-01 -8.67763069e-03 1.21896815e+00 -1.42219377e+00
-1.11116600e+00 1.05464149e+00 6.08531907e-02 -3.45534086e-01
6.40726209e-01 -1.33203298e-01 -8.16717464e-03 3.01306814e-01
9.47556794e-02 1.09396124e+00 5.12856305e-01 -1.50106680e+00
-6.05995595e-01 -3.86850268e-01 2.46624112e-01 5.23118734e-01
-4.48775262e-01 5.05519398e-02 -5.09646177e-01 -4.51142162e-01
3.82043213e-01 -8.71691465e-01 -3.55406135e-01 6.54984415e-01
-1.37637118e-02 -2.00792313e-01 7.92884171e-01 -6.58606231e-01
9.86533880e-01 -2.08247805e+00 7.64094517e-02 4.99764174e-01
2.95224041e-01 5.85057437e-01 -3.13042641e-01 5.27349651e-01
1.15320966e-01 3.86617482e-02 -7.46279836e-01 -4.30763215e-01
-4.55704927e-02 3.76574159e-01 -7.89387152e-02 4.62091506e-01
2.69061804e-01 7.88494766e-01 -1.19134641e+00 -6.92635536e-01
3.68840694e-01 4.57523912e-01 -7.56510869e-02 3.94007266e-01
-1.10212699e-01 6.05331421e-01 -3.99719298e-01 6.89611733e-01
9.77515578e-01 -2.50469834e-01 1.03359252e-01 -1.96776316e-01
-1.74064204e-01 -3.60977612e-02 -1.14744222e+00 1.48208594e+00
-6.39931083e-01 8.78511906e-01 2.56307609e-02 -1.47839952e+00
1.23119593e+00 2.66090900e-01 7.19194591e-01 -8.04055631e-01
-7.44092613e-02 4.67448235e-01 -2.39129648e-01 -4.15314823e-01
4.56252754e-01 2.19998330e-01 2.14821577e-01 4.08979833e-01
-2.13110134e-01 -4.66138981e-02 2.16541603e-01 2.90653948e-02
6.51037455e-01 4.20325920e-02 4.12648559e-01 -1.46898702e-01
6.80509329e-01 5.39133698e-02 5.56089818e-01 8.03701282e-01
-3.47625583e-01 5.98585129e-01 -1.61214620e-02 -6.43064022e-01
-1.07833099e+00 -7.62666523e-01 -2.94698000e-01 1.00422692e+00
2.58519232e-01 -5.13620451e-02 -6.34065688e-01 -5.19106627e-01
-2.71678001e-01 3.36242229e-01 -5.63770771e-01 1.05181284e-01
-4.45610106e-01 -8.74572396e-01 3.66291732e-01 3.84211570e-01
1.08650374e+00 -1.00273001e+00 -5.02810419e-01 1.95179015e-01
-5.53305566e-01 -5.91742456e-01 5.41213378e-02 2.00521350e-02
-5.43806195e-01 -9.59087431e-01 -9.92824733e-01 -7.82010734e-01
8.79021347e-01 4.65897888e-01 1.07741845e+00 2.75656670e-01
-2.69868791e-01 3.99230756e-02 -7.17583954e-01 -3.00943524e-01
-3.36184651e-01 2.32679266e-02 -4.75097775e-01 2.87805468e-01
4.92756277e-01 -2.09028587e-01 -9.61393416e-01 5.63406467e-01
-1.17962909e+00 -6.56062588e-02 7.51118183e-01 7.32723415e-01
9.02604640e-01 2.30615065e-01 3.37159127e-01 -7.76218235e-01
1.94540337e-01 -5.77496231e-01 -9.45414782e-01 4.60932881e-01
-4.25657004e-01 -2.29200080e-01 2.71655113e-01 -1.98249623e-01
-8.87458205e-01 4.59594220e-01 -2.70600945e-01 -1.04604132e-01
-1.31812647e-01 1.07823157e+00 -1.84224933e-01 -1.48989797e-01
6.33094132e-01 1.49927676e-01 -6.90533221e-02 -4.09470469e-01
1.06077112e-01 1.17314374e+00 2.15980873e-01 -3.17255557e-01
7.78758824e-01 4.78016973e-01 -1.63232863e-01 -7.84373105e-01
-1.10038304e+00 -8.55837405e-01 -8.79532993e-01 -4.16377604e-01
6.53930008e-01 -1.02503765e+00 -1.16426952e-01 6.41892612e-01
-8.67151797e-01 -4.80854958e-01 -1.97797164e-01 6.29705667e-01
-3.10399920e-01 4.34107810e-01 -4.36642990e-02 -9.39193904e-01
-5.46038687e-01 -1.21273398e+00 1.14108980e+00 1.03331342e-01
-3.21579054e-02 -9.96238112e-01 2.07101569e-01 3.45707685e-01
5.98186672e-01 2.12733537e-01 6.17788494e-01 -7.01140881e-01
-4.14749175e-01 -3.14341635e-01 -4.95341748e-01 6.02543592e-01
2.48816878e-01 -6.32356331e-02 -9.12473083e-01 -3.32839787e-01
-2.63587028e-01 -3.80939692e-01 7.39222765e-01 2.82341778e-01
1.23959899e+00 -2.92854935e-01 -1.55985698e-01 3.92875284e-01
1.55362642e+00 2.56900251e-01 7.42463350e-01 3.84481758e-01
5.29585779e-01 5.49523056e-01 1.04329479e+00 4.10830915e-01
3.30744803e-01 6.65743113e-01 6.18453503e-01 -4.28200573e-01
1.72088474e-01 1.76928237e-01 -8.33708327e-03 4.54056561e-01
1.02871563e-03 -4.28835124e-01 -1.28816748e+00 6.44374251e-01
-1.96807766e+00 -8.98424864e-01 -9.92456637e-03 2.47587967e+00
8.71526003e-01 -3.55939448e-01 -2.15540051e-01 3.40770364e-01
8.85331392e-01 8.84861276e-02 -6.14821792e-01 1.98016346e-01
-4.59231809e-03 1.24074996e-01 4.50275302e-01 6.54496133e-01
-1.37302005e+00 7.80189157e-01 4.43386126e+00 8.73645902e-01
-9.97506082e-01 1.78065225e-01 7.63385117e-01 1.41984344e-01
-2.26047248e-01 8.39012414e-02 -5.18758535e-01 3.59656900e-01
5.06122231e-01 1.48891151e-01 1.81633845e-01 7.15046644e-01
3.78013939e-01 -2.76182741e-01 -8.78558576e-01 1.03716290e+00
-7.56946132e-02 -1.12420082e+00 1.50410190e-01 -5.44979610e-02
8.90069544e-01 2.67363757e-01 -1.67105466e-01 5.91969900e-02
2.60145545e-01 -8.01864803e-01 5.98912716e-01 6.58273518e-01
7.38250315e-01 -7.63409972e-01 1.15168560e+00 4.00712818e-01
-1.26064277e+00 -2.24555537e-01 -6.70878470e-01 2.71184921e-01
-1.36618227e-01 8.12307656e-01 -5.69905579e-01 5.75678229e-01
8.43459845e-01 8.13034058e-01 -5.64632773e-01 1.25002468e+00
-1.80279583e-01 5.22835135e-01 -3.66694301e-01 9.06357989e-02
5.24445355e-01 -4.16820079e-01 3.62058789e-01 8.81657720e-01
3.51613611e-01 5.94900735e-02 2.74492413e-01 7.46158957e-01
-1.42545298e-01 1.73571855e-01 -3.84557247e-01 -7.56550878e-02
7.32502699e-01 1.35347009e+00 -6.32371008e-01 -2.65520126e-01
-3.65267918e-02 1.03367770e+00 2.84707487e-01 2.43160710e-01
-5.26134789e-01 -4.36934203e-01 2.21145436e-01 -1.13015331e-01
9.81725827e-02 -1.86880007e-01 1.55883953e-01 -7.81822145e-01
8.49635527e-02 -4.73422378e-01 4.15176272e-01 -8.81212354e-01
-1.25027096e+00 6.57707512e-01 1.20928183e-01 -1.51821125e+00
-7.78333545e-02 -2.16285363e-01 -3.69393736e-01 9.70823169e-01
-1.83686697e+00 -1.09767246e+00 -8.92754436e-01 3.95112336e-01
4.64996517e-01 4.03684638e-02 8.19371939e-01 6.44649625e-01
-2.97343254e-01 3.78666073e-01 6.33378923e-01 4.46904570e-01
3.85795176e-01 -1.20391941e+00 1.41854987e-01 7.64073610e-01
1.21423662e-01 2.72509605e-01 3.16123635e-01 -2.82781661e-01
-7.04875767e-01 -1.38640201e+00 9.27779257e-01 1.52865797e-01
4.72312540e-01 -2.37248614e-01 -8.41311514e-01 3.77864093e-01
-2.18405098e-01 2.32528627e-01 8.58168960e-01 -4.64376092e-01
-1.25940770e-01 -4.26508218e-01 -1.21035957e+00 2.07121432e-01
6.47874892e-01 -5.60534596e-01 -1.02151655e-01 5.89663625e-01
2.59699047e-01 -3.32830250e-02 -7.69047856e-01 4.09492940e-01
3.90764534e-01 -7.50231326e-01 9.23105240e-01 -4.63989414e-02
3.77576947e-01 -4.18363541e-01 -4.59213406e-01 -1.23247313e+00
-3.13949883e-01 -6.99205622e-02 3.68499219e-01 1.16800702e+00
3.06469828e-01 -5.89098990e-01 4.46149886e-01 4.86925393e-01
-2.28614286e-02 -5.22117019e-01 -8.49637985e-01 -6.18250549e-01
-6.82686046e-02 -1.63683534e-01 6.87106729e-01 1.06421614e+00
-5.83666861e-01 -1.91770047e-02 -1.21775687e-01 2.96817392e-01
6.09796524e-01 2.27924347e-01 5.48539340e-01 -1.62799537e+00
1.13188118e-01 -3.11361253e-01 -4.56979096e-01 -8.33788335e-01
1.35717258e-01 -9.05885637e-01 4.08064634e-01 -1.88212740e+00
2.23237216e-01 -9.92766738e-01 -4.09099400e-01 6.62020326e-01
-2.66649723e-01 5.73349893e-01 -9.69008654e-02 6.82533085e-01
-6.27328277e-01 5.39581120e-01 7.34610975e-01 -3.64136249e-01
-9.83356908e-02 2.13917509e-01 -2.83186555e-01 4.63763237e-01
1.04031253e+00 -5.70281088e-01 -4.82634485e-01 -5.90916336e-01
4.87475336e-01 -2.20535055e-01 6.58637881e-01 -8.16144586e-01
2.31817618e-01 -4.84056473e-02 3.08367103e-01 -8.81139934e-01
2.20102474e-01 -1.11323726e+00 2.53941745e-01 4.13731217e-01
-6.63987458e-01 -2.31882975e-01 -1.03922367e-01 4.59159940e-01
-4.79563862e-01 -6.08502567e-01 9.56437409e-01 -3.03769112e-01
-8.07373822e-01 7.13211238e-01 -2.69703627e-01 -2.35024914e-01
7.91401625e-01 -1.60261527e-01 1.86638944e-02 -3.99675578e-01
-6.10299408e-01 2.78839618e-01 2.92954504e-01 1.31057933e-01
5.64196706e-01 -1.32444334e+00 -1.01216722e+00 -9.42629427e-02
5.29902875e-01 4.35363144e-01 2.45903105e-01 6.89052403e-01
-9.05833244e-01 1.13900676e-01 4.67588659e-04 -7.73917556e-01
-1.22295570e+00 1.16526619e-01 5.61535239e-01 -1.41776219e-01
-3.64864796e-01 8.38965356e-01 -2.36820020e-02 -7.12713659e-01
2.02079967e-01 1.72062293e-01 -4.78262275e-01 4.47656959e-01
5.54499567e-01 2.74324059e-01 4.20891762e-01 -7.87095428e-01
-3.25559527e-01 4.70622122e-01 -6.42248243e-02 -1.34766564e-01
1.55400407e+00 -2.90172249e-01 -4.97410744e-01 3.14357311e-01
1.55495226e+00 -5.26601911e-01 -1.15792942e+00 -7.10417151e-01
1.57519743e-01 -6.17467582e-01 5.58912158e-01 -5.99031270e-01
-1.28369892e+00 8.36994469e-01 1.11293590e+00 1.57235727e-01
1.24757910e+00 -1.15126230e-01 5.09735227e-01 6.91174328e-01
1.86413780e-01 -1.17763507e+00 -1.61356255e-01 2.55934447e-01
9.68623698e-01 -1.67303538e+00 7.58983269e-02 7.66346883e-03
-3.93243939e-01 9.33792710e-01 2.87856102e-01 1.19915038e-01
8.12918961e-01 -2.03261822e-01 2.88582146e-01 -3.81769508e-01
-1.18477784e-01 -3.78114611e-01 3.47282350e-01 3.74137491e-01
3.40720206e-01 1.09981403e-01 -3.64128739e-01 -2.41527930e-01
1.95867494e-01 -3.97230349e-02 7.07612559e-02 9.79295850e-01
-5.26458740e-01 -1.04918921e+00 -3.15729111e-01 3.77989501e-01
-8.06413963e-02 -2.53393263e-01 -5.78132987e-01 4.92607981e-01
2.60526426e-02 1.07388031e+00 4.21339691e-01 -9.47925225e-02
-1.79244623e-01 -2.75889963e-01 7.00082257e-02 -5.22511005e-01
-3.46478254e-01 -1.33264422e-01 -1.84047014e-01 -2.95830727e-01
-8.48438382e-01 -4.99882042e-01 -1.00089526e+00 3.42057832e-02
-4.80960101e-01 3.68625402e-01 7.69678354e-01 8.71344864e-01
1.93558350e-01 -1.97884396e-01 1.12058651e+00 -9.69175875e-01
-3.77867669e-01 -9.88571763e-01 -3.82137120e-01 3.27563554e-01
2.72057027e-01 -5.84975541e-01 -4.85551119e-01 -9.00966153e-02] | [9.75486946105957, -1.3047770261764526] |
f63283b4-d504-4a1b-a67d-6add306fee14 | wizmap-scalable-interactive-visualization-for | 2306.09328 | null | https://arxiv.org/abs/2306.09328v1 | https://arxiv.org/pdf/2306.09328v1.pdf | WizMap: Scalable Interactive Visualization for Exploring Large Machine Learning Embeddings | Machine learning models often learn latent embedding representations that capture the domain semantics of their training data. These embedding representations are valuable for interpreting trained models, building new models, and analyzing new datasets. However, interpreting and using embeddings can be challenging due to their opaqueness, high dimensionality, and the large size of modern datasets. To tackle these challenges, we present WizMap, an interactive visualization tool to help researchers and practitioners easily explore large embeddings. With a novel multi-resolution embedding summarization method and a familiar map-like interaction design, WizMap enables users to navigate and interpret embedding spaces with ease. Leveraging modern web technologies such as WebGL and Web Workers, WizMap scales to millions of embedding points directly in users' web browsers and computational notebooks without the need for dedicated backend servers. WizMap is open-source and available at the following public demo link: https://poloclub.github.io/wizmap. | ['Duen Horng Chau', 'Fred Hohman', 'Zijie J. Wang'] | 2023-06-15 | null | null | null | null | ['navigate'] | ['reasoning'] | [-6.33646548e-01 1.18516333e-01 -3.60867232e-01 -1.14489421e-01
-5.09229958e-01 -8.19607258e-01 6.37551963e-01 4.49015111e-01
-1.79113448e-01 1.58518940e-01 7.01506257e-01 -7.36401320e-01
-8.55640415e-03 -9.46808636e-01 -2.21905485e-01 -2.36209571e-01
-1.87014788e-01 2.70749837e-01 6.30877316e-02 4.47478779e-02
1.09927684e-01 2.40864530e-01 -1.64146817e+00 2.93089598e-01
7.98716724e-01 4.91737396e-01 3.17535698e-01 6.37904048e-01
-3.61794591e-01 2.47299120e-01 -1.79373339e-01 -4.19365764e-01
1.30454734e-01 1.16122276e-01 -5.75683117e-01 -4.99030083e-01
2.41193473e-01 -4.94015276e-01 -5.64783633e-01 7.43527770e-01
3.71852607e-01 4.03879993e-02 5.18230915e-01 -1.49714482e+00
-1.20522237e+00 2.90561765e-01 -1.74008936e-01 1.83648258e-01
4.10485774e-01 3.23233426e-01 1.19343543e+00 -1.20966923e+00
6.09395087e-01 1.01335835e+00 5.46774447e-01 4.29612845e-01
-1.36739016e+00 -6.45174205e-01 6.91256374e-02 3.32609981e-01
-1.24489558e+00 -3.50095510e-01 6.91774607e-01 -7.00759172e-01
9.43815827e-01 4.01807219e-01 6.05668306e-01 1.33722258e+00
-2.57025331e-01 5.87815642e-01 7.10756540e-01 -3.40247035e-01
2.94811189e-01 4.81441349e-01 3.62587422e-01 8.16318512e-01
3.63379836e-01 -3.10923308e-01 -7.25592196e-01 -3.93737495e-01
8.80447447e-01 4.36160654e-01 -1.61891177e-01 -6.12300396e-01
-1.35244358e+00 8.33942592e-01 5.98328114e-01 1.16635412e-01
-2.99147159e-01 2.26764411e-01 3.11208367e-01 1.85302809e-01
6.88841045e-01 4.07407522e-01 -3.28980416e-01 -5.43197989e-01
-3.45017791e-01 2.51862973e-01 4.62257087e-01 8.44481945e-01
7.93872058e-01 -2.64902979e-01 5.20684063e-01 8.66158307e-01
5.49211979e-01 3.33358675e-01 5.97995698e-01 -8.51581514e-01
6.65833414e-01 9.31889057e-01 1.75876006e-01 -1.16982174e+00
-1.48578957e-01 3.78044024e-02 -5.06247163e-01 3.82468253e-01
3.41700107e-01 1.01542640e-02 -4.26672816e-01 1.33905423e+00
4.83303607e-01 -8.37790519e-02 -1.45544097e-01 6.77920282e-01
8.50594163e-01 7.28687584e-01 1.47478029e-01 6.46097720e-01
1.59701371e+00 -8.14022839e-01 -5.95375657e-01 -4.62092251e-01
1.06431615e+00 -5.13632059e-01 1.80208921e+00 1.23168968e-01
-6.86638355e-01 -2.68376589e-01 -1.15106761e+00 -6.86373472e-01
-9.92161989e-01 9.44228619e-02 6.91262186e-01 1.80278853e-01
-8.75819385e-01 6.05781913e-01 -1.22028148e+00 -5.72500229e-01
6.34387910e-01 -6.07247800e-02 -7.65502691e-01 -1.01608030e-01
-7.88886249e-01 7.92289197e-01 2.82310307e-01 -1.63643286e-01
-1.40423015e-01 -1.00071871e+00 -9.93045986e-01 2.30685577e-01
1.01526365e-01 -5.18242240e-01 1.09437752e+00 -3.06014985e-01
-1.02205062e+00 6.20428026e-01 -1.93595484e-01 5.42996414e-02
3.86910319e-01 -7.67829418e-01 -3.01764816e-01 5.92728611e-03
-9.76428092e-02 2.93439180e-01 4.31634814e-01 -9.64598417e-01
-6.46781698e-02 -5.26588500e-01 1.46020919e-01 -1.63318275e-03
-9.61000383e-01 -1.99862942e-01 -4.16786224e-01 -5.60907185e-01
1.89089924e-01 -5.28254569e-01 -5.60373589e-02 7.19293952e-01
-2.09264368e-01 -1.83237329e-01 1.37344670e+00 -8.57653677e-01
1.49238420e+00 -2.49970436e+00 1.07610628e-01 -9.66798216e-02
6.24802351e-01 2.98917949e-01 -8.56387541e-02 8.60001743e-01
-2.05759659e-01 5.12333691e-01 -5.30323759e-03 -4.71083641e-01
4.40629274e-01 1.43187702e-01 -2.99720675e-01 2.15460554e-01
3.89470011e-02 1.00044560e+00 -1.07818651e+00 -1.90683097e-01
5.43647528e-01 7.59151459e-01 -4.42335993e-01 1.98592111e-01
-2.36087397e-01 4.45896424e-02 -3.39112461e-01 4.58907396e-01
5.96196651e-01 -5.98306358e-01 4.65956539e-01 -1.91729050e-02
-2.22613752e-01 5.82158089e-01 -1.18213582e+00 1.85731220e+00
-6.71343327e-01 1.18278599e+00 -2.10162088e-01 -5.83818495e-01
8.27834487e-01 1.32132173e-01 -3.09283612e-03 -4.05630708e-01
-2.95638412e-01 1.38671339e-01 -7.32936561e-01 -6.00235283e-01
6.71053946e-01 3.69458288e-01 -3.72618139e-02 1.05913532e+00
-7.33486051e-03 2.75425851e-01 -3.13876905e-02 5.64377606e-01
9.64956343e-01 8.95714387e-02 3.48114938e-01 -6.58195242e-02
-2.58216262e-01 -9.59015340e-02 3.36056709e-01 1.98231995e-01
3.13076228e-01 4.71506119e-01 6.96607947e-01 -8.53689551e-01
-1.26497650e+00 -1.45337892e+00 -1.70237526e-01 1.07924259e+00
-2.14693204e-01 -1.06025529e+00 -3.22279125e-01 -4.58956212e-01
2.89924771e-01 7.71614373e-01 -5.76557577e-01 9.94874164e-02
-1.69670403e-01 -3.55088115e-01 2.36147553e-01 8.11774492e-01
1.55936360e-01 -7.40200579e-01 -6.17774963e-01 -8.06088671e-02
-1.52462080e-01 -7.40283191e-01 -1.88626140e-01 -1.44970745e-01
-9.94533360e-01 -1.00860488e+00 -4.74791497e-01 -6.00644469e-01
8.43794882e-01 3.95700544e-01 1.03075361e+00 1.76233724e-01
-5.06818414e-01 2.59636879e-01 -3.97738218e-01 -3.50337893e-01
5.32242768e-02 1.92459196e-01 1.06392711e-01 -5.51518381e-01
7.53200471e-01 -1.02882588e+00 -7.49281824e-01 8.87559354e-02
-1.03068757e+00 5.04320323e-01 3.07918079e-02 4.86690372e-01
1.81393951e-01 -2.75767058e-01 2.86318898e-01 -8.30649197e-01
8.59201372e-01 -9.30481017e-01 -5.56544125e-01 1.87467083e-01
-7.69447803e-01 1.04256302e-01 3.96293551e-01 -3.37844849e-01
-5.59354842e-01 -4.38600391e-01 2.72412580e-02 -2.86416650e-01
-2.07755446e-01 5.53842664e-01 -1.54934257e-01 5.10283828e-01
6.74841285e-01 -1.14005841e-01 1.18109986e-01 -1.03840804e+00
8.07213604e-01 1.11122227e+00 3.20267498e-01 -2.66349465e-01
8.72315884e-01 2.53677607e-01 -7.10520327e-01 -7.89515555e-01
-2.88995504e-01 -2.16234609e-01 -8.17402363e-01 4.31819633e-02
6.06104553e-01 -8.92858207e-01 -4.54306573e-01 1.21782497e-01
-9.00266588e-01 -5.09426057e-01 -2.21895829e-01 3.84828866e-01
-5.60116097e-02 2.65781045e-01 -2.88939416e-01 -3.88662368e-01
-2.37610802e-01 -7.49122918e-01 7.34524369e-01 1.90691486e-01
-6.05836987e-01 -1.32584333e+00 2.57778019e-01 1.63187563e-01
4.21042830e-01 3.93779665e-01 1.21379828e+00 -3.55765641e-01
-6.40149355e-01 -5.55390120e-01 -4.20927882e-01 1.04398973e-01
3.31702471e-01 2.55914837e-01 -1.11963809e+00 -5.94450571e-02
-6.61031365e-01 -1.88883096e-01 4.18381274e-01 -2.37820849e-01
1.39674163e+00 -5.38357019e-01 -4.10748512e-01 7.49317110e-01
1.38518095e+00 -3.79558980e-01 4.20865715e-01 6.02796376e-01
6.99697495e-01 4.39194590e-01 2.32511178e-01 5.95532477e-01
5.84222078e-01 6.05794013e-01 3.41024429e-01 -2.07567159e-02
1.70589328e-01 -6.35732412e-01 1.49965063e-01 9.56130505e-01
1.12187624e-01 -7.65296146e-02 -1.37200713e+00 6.40296161e-01
-1.93839383e+00 -8.74975443e-01 -1.37754241e-02 2.30896282e+00
4.72811878e-01 -6.89098909e-02 -7.44995177e-02 -5.32117002e-02
3.01752478e-01 4.19313818e-01 -5.20341814e-01 -4.81694728e-01
3.11533928e-01 6.16383553e-03 2.15628475e-01 4.71299976e-01
-7.16569960e-01 6.61543489e-01 5.82497215e+00 4.28660661e-01
-1.15158176e+00 3.40275794e-01 1.28601491e-01 -4.74208713e-01
-8.02648902e-01 3.35676782e-02 -3.42771977e-01 4.55635190e-01
9.82071221e-01 -4.93370712e-01 4.19235826e-01 1.05843449e+00
2.33830601e-01 1.06005393e-01 -1.30886137e+00 1.18646288e+00
-3.18615735e-01 -1.78966951e+00 -3.26665789e-02 4.23163801e-01
1.58647299e-01 3.67286533e-01 3.90361071e-01 1.00289851e-01
4.18592185e-01 -1.01122808e+00 3.93786162e-01 3.27231824e-01
1.04733610e+00 -6.26463771e-01 3.57339293e-01 1.19057015e-01
-1.09331715e+00 -1.11143589e-01 -4.10106570e-01 -3.02569270e-01
1.41229138e-01 4.34795678e-01 -6.91886127e-01 2.48309404e-01
6.88389957e-01 8.29169333e-01 -7.46974170e-01 7.10399926e-01
-3.30550015e-01 4.35808450e-01 -4.72481370e-01 -3.92247428e-04
-1.61510967e-02 -1.11654565e-01 3.54869604e-01 1.17077565e+00
4.57825065e-01 -9.21584517e-02 -2.15902984e-01 9.12356734e-01
-1.16010562e-01 -5.53086633e-03 -9.60476995e-01 -6.88076198e-01
8.38145077e-01 1.36394382e+00 -4.27197188e-01 -2.10919946e-01
-4.65392083e-01 1.04142773e+00 5.94028354e-01 5.28513014e-01
-6.98127747e-01 -6.85620546e-01 1.35817027e+00 5.66963255e-01
-4.01599426e-03 -9.29332018e-01 -4.14413720e-01 -1.44230890e+00
3.76611322e-01 -6.28218710e-01 2.24409491e-01 -1.10133839e+00
-1.05144382e+00 5.68156958e-01 1.39162824e-01 -1.17643595e+00
-3.14226538e-01 -7.05660701e-01 -7.05990911e-01 8.08835506e-01
-1.08181560e+00 -1.08802736e+00 -5.52167416e-01 3.39911997e-01
3.36196899e-01 -6.37284061e-03 1.32241023e+00 2.06572682e-01
-6.23223901e-01 4.33344275e-01 6.45304203e-01 3.23666841e-01
4.81580198e-01 -1.30881453e+00 1.01401150e+00 4.29843217e-01
5.43631613e-01 1.04123139e+00 5.22920132e-01 -3.80609125e-01
-1.40692413e+00 -7.77655184e-01 9.64226961e-01 -8.68259728e-01
9.05048370e-01 -9.97530222e-01 -1.18846905e+00 1.11333084e+00
2.11171746e-01 2.17305169e-01 1.34207797e+00 7.21111894e-01
-8.27488542e-01 1.26796648e-01 -7.90556729e-01 8.91406000e-01
9.25159693e-01 -1.00765002e+00 -6.05103076e-01 4.92454976e-01
6.80566013e-01 -2.02893525e-01 -8.77544105e-01 -3.12393248e-01
7.70688653e-01 -7.76632667e-01 9.81873453e-01 -6.74920440e-01
6.04312897e-01 -1.54630110e-01 -3.83099504e-02 -1.26019633e+00
-2.99910933e-01 -4.26891714e-01 -4.62524652e-01 1.02523112e+00
4.90387797e-01 -9.87838745e-01 5.91513693e-01 1.15897715e+00
3.64874661e-01 -7.72490978e-01 -6.23389781e-01 -5.90295434e-01
3.11080590e-02 -6.53421640e-01 9.50867414e-01 1.06805778e+00
6.08022094e-01 1.09348193e-01 -1.02217823e-01 3.51689458e-01
5.85086048e-01 3.52122635e-02 9.94412780e-01 -1.28493810e+00
-2.19747141e-01 -2.12271899e-01 -5.64070880e-01 -8.95866752e-01
-2.61952639e-01 -9.24814582e-01 -7.44266987e-01 -1.84403384e+00
-9.52250138e-03 -6.28328860e-01 -1.26168728e-01 7.99498320e-01
-3.01201511e-02 1.96482390e-01 3.47020417e-01 4.21276093e-01
-3.69729012e-01 4.71786350e-01 5.21323264e-01 1.11381151e-01
-2.09721968e-01 -5.33863008e-01 -8.83704245e-01 7.37078488e-01
9.75316286e-01 -5.21969318e-01 -4.91329700e-01 -1.22045934e+00
4.33254331e-01 -6.30744219e-01 6.62116230e-01 -8.94990623e-01
1.86487213e-02 5.92022426e-02 4.20679122e-01 -4.10962760e-01
4.91622686e-01 -6.96423113e-01 2.33364850e-01 9.40382406e-02
-2.94996619e-01 5.32678127e-01 4.48260993e-01 5.47412336e-01
-5.37436157e-02 -6.54385313e-02 2.12906793e-01 -1.76584739e-02
-6.62314892e-01 1.40381932e-01 -3.82375121e-01 -1.08132526e-01
1.06756985e+00 -1.82841256e-01 -6.01090550e-01 -3.14574420e-01
-7.57103443e-01 1.57747895e-01 9.27688003e-01 7.81547785e-01
6.88089073e-01 -1.45202875e+00 -2.15177685e-01 5.22083223e-01
4.53679353e-01 1.41886666e-01 2.42650867e-01 2.37130016e-01
-7.89336979e-01 2.02486157e-01 -2.52066106e-01 -2.20389187e-01
-1.34443116e+00 3.12781513e-01 -1.57076135e-01 1.28019154e-01
-1.30332768e+00 7.14888453e-01 5.49390204e-02 -7.26346076e-01
5.65173291e-02 -2.96731174e-01 1.03977546e-01 4.17505950e-02
1.10406506e+00 3.80514234e-01 -1.96123272e-01 -1.01327084e-01
-2.74746656e-01 2.64226139e-01 -7.39486143e-02 -1.15695439e-01
1.66864872e+00 -1.75195694e-01 1.63588747e-01 7.36744761e-01
1.36054075e+00 -5.31676114e-02 -1.27955139e+00 -3.45422149e-01
3.62993963e-02 -8.61872256e-01 1.14023149e-01 -4.64698970e-01
-6.63618624e-01 1.30110347e+00 4.97189432e-01 1.16489336e-01
6.52029395e-01 1.90572619e-01 7.81515360e-01 3.46004725e-01
1.48119017e-01 -8.34681392e-01 1.49272591e-01 2.08322182e-01
8.17661583e-01 -1.21937633e+00 1.48696274e-01 1.92805764e-03
-5.16112387e-01 1.31061494e+00 4.98201281e-01 4.67620529e-02
8.18228006e-01 1.85866639e-01 2.67982870e-01 -3.06041509e-01
-1.09795427e+00 1.92731053e-01 -1.60534047e-02 6.72778606e-01
5.17014444e-01 1.35730535e-01 8.57763514e-02 4.64399308e-01
-2.86784321e-01 6.29790276e-02 3.80981892e-01 9.34026957e-01
-2.69522130e-01 -1.31721318e+00 -9.40363407e-02 4.60326374e-01
-2.62428578e-02 -1.85076714e-01 -7.48637989e-02 6.88130856e-01
-2.71548361e-01 4.75156099e-01 4.58485425e-01 -5.09734213e-01
1.19116247e-01 2.44136289e-01 -1.52691185e-01 -7.17805266e-01
-3.53735983e-02 -6.27660871e-01 -8.96968096e-02 -6.85920119e-01
2.95725584e-01 -4.81821626e-01 -1.07130432e+00 -7.30500340e-01
1.42565638e-01 9.89539027e-02 1.06845975e+00 3.72517943e-01
1.01375818e+00 5.81834018e-02 1.86969459e-01 -7.50804663e-01
-2.76966333e-01 -1.04138362e+00 -3.98123622e-01 3.42133790e-01
2.22163662e-01 -6.27556741e-01 -3.82797956e-01 8.72434825e-02] | [10.39069652557373, 8.387252807617188] |
8c4f24a4-be15-4fcd-9b51-9778c82a8db2 | managing-large-dataset-gaps-in-urban-air | 2212.10273 | null | https://arxiv.org/abs/2212.10273v1 | https://arxiv.org/pdf/2212.10273v1.pdf | Managing Large Dataset Gaps in Urban Air Quality Prediction: DCU-Insight-AQ at MediaEval 2022 | Calculating an Air Quality Index (AQI) typically uses data streams from air quality sensors deployed at fixed locations and the calculation is a real time process. If one or a number of sensors are broken or offline, then the real time AQI value cannot be computed. Estimating AQI values for some point in the future is a predictive process and uses historical AQI values to train and build models. In this work we focus on gap filling in air quality data where the task is to predict the AQI at 1, 5 and 7 days into the future. The scenario is where one or a number of air, weather and traffic sensors are offline and explores prediction accuracy under such situations. The work is part of the MediaEval'2022 Urban Air: Urban Life and Air Pollution task submitted by the DCU-Insight-AQ team and uses multimodal and crossmodal data consisting of AQI, weather and CCTV traffic images for air pollution prediction. | ['Alan F. Smeaton', 'Mark Roantree', 'Elke Eichlemann', 'Adam Stapleton', 'Phuc H. Le-Khac', 'Dinh Viet Cuong'] | 2022-12-19 | null | null | null | null | ['air-pollution-prediction'] | ['miscellaneous'] | [ 2.42300227e-01 -4.32720751e-01 1.18828088e-01 -4.39305782e-01
-8.63892972e-01 -6.14449918e-01 3.48459154e-01 6.90787971e-01
-2.07575083e-01 8.76472950e-01 2.47636795e-01 -7.55717099e-01
-6.98034286e-01 -1.43720531e+00 -3.46705109e-01 -4.60393190e-01
2.66612381e-01 8.04628074e-01 9.72510651e-02 1.19456671e-01
-3.77238423e-01 5.63274860e-01 -1.67358208e+00 3.32142651e-01
7.65887082e-01 1.09075201e+00 -1.00079656e-01 1.34944665e+00
-2.79191703e-01 3.34427543e-02 -5.26051342e-01 2.01213345e-01
6.09925628e-01 -7.37605467e-02 -8.05840075e-01 -1.14166848e-01
4.01133329e-01 -8.04822445e-02 2.19631732e-01 7.02620566e-01
3.89932960e-01 4.34335709e-01 3.83760303e-01 -1.54684353e+00
2.98169047e-01 -9.49210525e-02 1.23834163e-01 5.42265594e-01
3.99888068e-01 6.60983324e-01 9.98663723e-01 -2.63887018e-01
-5.72386086e-02 1.00197661e+00 8.33621323e-01 1.87823214e-02
-1.05421352e+00 -6.10511839e-01 2.22352430e-01 1.95167243e-01
-1.11463642e+00 -6.22535348e-02 3.56777422e-02 -5.51886261e-01
1.20473719e+00 8.25599968e-01 1.06021070e+00 4.68336970e-01
6.74004927e-02 2.84593821e-01 1.21672404e+00 -6.45813718e-02
3.90344471e-01 -2.32611597e-02 3.08672786e-01 2.86034141e-02
1.74690083e-01 6.82790577e-01 -5.69408946e-02 -2.65125573e-01
-3.27292413e-01 5.98507464e-01 6.44766241e-02 7.34710753e-01
-1.08454382e+00 6.04741514e-01 5.22976458e-01 1.69213340e-01
-7.34247148e-01 -1.02867901e-01 -8.21305066e-02 6.89456403e-01
7.12273538e-01 4.36626673e-01 -8.29116821e-01 -2.62325704e-01
-1.16462886e+00 5.28452814e-01 4.62178946e-01 5.82498252e-01
1.07780159e+00 -3.20770949e-01 -2.11828277e-01 5.63687623e-01
7.13532269e-01 1.32024992e+00 -1.60434484e-01 -9.60609138e-01
6.85339153e-01 5.95650196e-01 6.97421849e-01 -7.02557743e-01
-6.53570116e-01 1.16681552e-03 -4.06722546e-01 2.63151854e-01
4.89639282e-01 -4.73099619e-01 -1.06078255e+00 8.53680313e-01
6.77091897e-01 2.33789250e-01 -1.58363760e-01 7.14142799e-01
8.42267692e-01 1.40272808e+00 3.67174238e-01 -2.41640806e-01
1.01996934e+00 -5.25892317e-01 -7.77139485e-01 -2.43809018e-02
4.74243492e-01 -5.75290263e-01 6.16369069e-01 5.31651556e-01
-9.27320302e-01 -8.03653240e-01 -4.65022594e-01 6.44866347e-01
-1.40579498e+00 -4.34292197e-01 8.51708278e-02 8.21785510e-01
-7.96537399e-01 5.89361727e-01 -4.69422489e-01 -3.51274014e-01
1.68332681e-01 2.69072026e-01 -6.84263855e-02 -3.72855753e-01
-1.60436785e+00 7.98222244e-01 3.06896567e-01 2.78407007e-01
-9.97712314e-01 -1.12602639e+00 -6.83798075e-01 -2.22713411e-01
-8.43878761e-02 -6.65504515e-01 1.18428361e+00 -4.10712361e-01
-8.22151363e-01 3.53149265e-01 -1.52484819e-01 -3.08500201e-01
3.13639730e-01 -2.34552875e-01 -1.27955389e+00 -2.94963449e-01
2.38743126e-01 4.67328876e-01 6.77762032e-01 -9.30137336e-01
-1.28077829e+00 -6.29555166e-01 2.01488197e-01 8.12446252e-02
2.06392601e-01 6.96258172e-02 3.48566681e-01 1.08864643e-01
-3.12129229e-01 -9.73471642e-01 -4.71251577e-01 -1.32830009e-01
-1.35400951e-01 -5.35555840e-01 7.95118511e-01 -7.99706221e-01
1.58178234e+00 -1.69476712e+00 -6.96932554e-01 7.07115173e-01
-2.14940712e-01 7.26115167e-01 -1.43047139e-01 5.10532796e-01
-4.17218328e-01 1.32882759e-01 -2.91111946e-01 -1.14600815e-01
6.79261163e-02 6.76489353e-01 -5.56568764e-02 5.10116398e-01
1.42093435e-01 5.96539676e-01 -1.10356677e+00 -1.09265089e-01
6.88452780e-01 1.54369488e-01 -5.66447787e-02 4.78007078e-01
-6.56477511e-01 8.10397804e-01 -3.36352676e-01 4.35408115e-01
7.90888488e-01 3.13084334e-01 -5.62837958e-01 5.53014278e-02
-8.25040638e-01 4.36744213e-01 -1.42366993e+00 1.26926386e+00
-7.27545917e-01 4.07492727e-01 -6.29789010e-02 -7.10926354e-01
5.79471767e-01 8.41010571e-01 5.86594403e-01 -1.12675321e+00
-1.89292803e-01 1.54513493e-01 -3.61816972e-01 -7.94157267e-01
2.17531666e-01 -2.49306127e-01 2.47834086e-01 5.77811301e-01
-6.79429114e-01 -3.86857659e-01 3.47642809e-01 -3.71489197e-01
1.10175109e+00 -1.49078414e-01 -2.96350837e-01 1.90197434e-02
5.56881368e-01 3.27264577e-01 4.33642119e-01 4.87063080e-01
-4.35276985e-01 4.22989368e-01 -1.73974633e-01 -7.20504761e-01
-1.28424549e+00 -1.14583063e+00 -4.56878185e-01 8.86510372e-01
-4.23835874e-01 -1.64928988e-01 -3.23249668e-01 -5.76035738e-01
-3.22512793e-03 1.00733376e+00 -3.04671496e-01 1.74674019e-01
-3.77622515e-01 -7.73337364e-01 -3.94369103e-02 2.31562093e-01
3.91719937e-01 -7.10501194e-01 -4.84385967e-01 4.52056289e-01
-2.94550151e-01 -7.57583082e-01 -1.89100113e-02 -2.44098783e-01
-8.32463443e-01 -1.14335287e+00 -4.43926901e-01 1.94891691e-01
3.71154636e-01 2.62973219e-01 1.28132308e+00 -4.84187864e-02
-1.96580634e-01 7.05229223e-01 -2.29079977e-01 -1.15855753e+00
-3.50318700e-01 -1.46782860e-01 -1.18210785e-01 3.69274546e-03
5.36869228e-01 -2.86730498e-01 -8.88834774e-01 4.53733087e-01
-1.05009985e+00 -3.90401483e-01 -1.91926524e-01 1.59486458e-01
8.80398929e-01 8.13378334e-01 4.76370454e-01 -5.84527552e-01
2.78086692e-01 -8.06335866e-01 -8.68779838e-01 -1.07221501e-02
-8.69168699e-01 -4.10865545e-01 4.43163633e-01 2.96714365e-01
-1.05281615e+00 1.27385661e-01 -5.74899375e-01 4.10167128e-02
-9.33930695e-01 1.92464754e-01 -1.40506580e-01 7.51150787e-01
5.19337237e-01 -2.91291267e-01 -4.60501045e-01 -8.65412891e-01
1.74573854e-01 1.01204073e+00 4.53582853e-02 -2.86117405e-01
8.92587841e-01 4.68147516e-01 1.20671712e-01 -9.79201138e-01
-1.30412424e+00 -1.23692334e+00 -7.03882396e-01 -7.25090325e-01
1.13009107e+00 -1.11890554e+00 -5.89972258e-01 3.45169574e-01
-1.04145098e+00 -2.67379999e-01 -5.47047317e-01 4.23739165e-01
9.50934589e-02 -2.49425098e-01 1.86521471e-01 -1.29657567e+00
-4.18888569e-01 -9.03790474e-01 1.18449771e+00 2.49090865e-01
-2.18915209e-01 -1.22115707e+00 6.80129230e-01 8.08451355e-01
6.45311415e-01 4.10683662e-01 5.95154941e-01 -1.98516965e-01
-3.60382080e-01 -3.76341671e-01 -3.44539016e-01 5.61155379e-01
1.73968241e-01 1.24003641e-01 -1.21923852e+00 -3.20531160e-01
-5.53187430e-01 2.49517843e-01 5.56743324e-01 3.13837707e-01
9.86182928e-01 -4.87800330e-01 -1.79344535e-01 -1.26956776e-01
1.54533732e+00 3.78284067e-01 7.38363326e-01 1.23035252e-01
2.71053314e-01 7.45327234e-01 8.86275113e-01 4.65362579e-01
4.20944542e-01 3.68602186e-01 9.77162778e-01 -2.19925210e-01
-8.51555094e-02 1.35091081e-01 8.05817768e-02 4.05192018e-01
-1.01783872e-01 -3.09317231e-01 -1.25461853e+00 1.14274704e+00
-1.50805032e+00 -1.07740176e+00 -9.81218338e-01 2.48998284e+00
5.66303551e-01 -3.45544629e-02 4.51340526e-01 6.12991631e-01
3.91711235e-01 -5.23168361e-03 -2.12167993e-01 -6.48564339e-01
4.53275591e-01 4.67872769e-01 6.59009159e-01 7.36694992e-01
-9.87584651e-01 1.78407028e-01 6.67320490e+00 1.77120134e-01
-6.17384970e-01 4.54692572e-01 5.74969351e-01 -1.71271324e-01
-3.49288911e-01 -8.72554258e-02 -9.21828449e-01 7.21504450e-01
2.05694938e+00 5.95466435e-01 4.89615113e-01 1.57912388e-01
9.33012664e-01 -4.84349877e-01 -8.14053357e-01 5.01502693e-01
-7.03043282e-01 -9.15350914e-01 -4.37205881e-01 1.83461621e-01
9.65254009e-01 5.72498918e-01 -8.74555707e-02 -9.88734365e-02
3.85979801e-01 -1.19782174e+00 2.80192137e-01 1.32814324e+00
5.56742013e-01 -9.10101175e-01 7.90202439e-01 3.75876158e-01
-1.24073088e+00 -4.17759925e-01 -2.10413203e-01 -1.68173507e-01
2.67882317e-01 1.12555456e+00 -6.61885321e-01 5.72955370e-01
1.03845263e+00 6.46591246e-01 -5.15781939e-01 1.51452720e+00
-1.67490795e-01 1.03470862e+00 -7.06702650e-01 1.86141148e-01
5.42101145e-01 -3.63056511e-01 6.35094106e-01 1.18012202e+00
5.47501564e-01 3.08020264e-01 1.73173156e-02 7.16053009e-01
6.01348758e-01 -2.71579504e-01 -9.21825111e-01 1.47198305e-01
1.42886236e-01 9.74175394e-01 8.32952484e-02 -3.57189775e-01
-6.25134289e-01 2.51388460e-01 -6.00572288e-01 5.06399691e-01
-7.79527962e-01 -2.02759817e-01 1.15335333e+00 5.24621189e-01
7.34266266e-02 -2.48835519e-01 -1.57378718e-01 -3.34554821e-01
-3.10138047e-01 -3.86448979e-01 5.68866670e-01 -1.08488131e+00
-1.20032549e+00 -3.75429317e-02 1.48744419e-01 -1.30775249e+00
5.93403988e-02 -6.54041946e-01 -9.80736732e-01 1.08145130e+00
-1.96425152e+00 -8.51315796e-01 -4.03526634e-01 5.66920459e-01
2.94147253e-01 3.67116541e-01 8.88329029e-01 7.21753299e-01
-5.71929991e-01 -3.39126945e-01 3.28814596e-01 -2.27475241e-01
1.67513192e-01 -1.27627850e+00 2.37602428e-01 5.41839302e-01
-2.04868659e-01 -7.72257224e-02 8.87655735e-01 -6.12558544e-01
-1.05132043e+00 -1.81931198e+00 1.52062738e+00 -8.01860631e-01
4.69557256e-01 1.76452622e-01 -6.78054571e-01 2.73253053e-01
3.05785060e-01 -5.23205288e-02 9.78757620e-01 -2.27894530e-01
2.47550249e-01 -6.51476085e-01 -1.41592467e+00 -1.88732207e-01
5.90409160e-01 -4.70549732e-01 -4.23473924e-01 9.52965200e-01
6.44413710e-01 -3.09584346e-02 -1.42686522e+00 1.73674107e-01
2.83041090e-01 -7.85714447e-01 1.01724613e+00 -9.04123783e-01
-1.62875485e-02 -7.29731381e-01 -2.57548749e-01 -1.55720425e+00
4.79288306e-03 -1.23835176e-01 2.40582466e-01 1.08514893e+00
8.62539351e-01 -6.34330809e-01 3.49727243e-01 1.15996969e+00
-4.20951582e-02 -6.83576241e-02 -1.25087881e+00 -5.82773387e-01
1.85773540e-02 -1.28336823e+00 1.48949277e+00 5.57929039e-01
-6.69180334e-01 4.22513299e-02 -1.45356119e-01 8.41254413e-01
4.38414425e-01 1.35730475e-01 5.25887430e-01 -1.62022865e+00
2.59062320e-01 5.28224139e-03 -1.70166254e-01 -4.69241947e-01
-3.10035497e-01 -5.27994037e-01 1.37883663e-01 -2.18142104e+00
-4.77556080e-01 -5.95363736e-01 -4.77042079e-01 5.74608028e-01
2.43834987e-01 1.82204396e-02 -7.91292414e-02 -2.24724188e-01
-5.01168787e-01 2.14431927e-01 1.14461720e+00 -5.88834524e-01
-2.21065521e-01 8.27552736e-01 1.18700065e-01 3.07316691e-01
9.89067554e-01 -6.52754605e-01 -2.76226342e-01 -4.34203625e-01
8.58419120e-01 9.27320197e-02 4.91797537e-01 -1.31591260e+00
4.21320312e-02 -5.36009848e-01 4.34791952e-01 -1.29776871e+00
2.68097907e-01 -1.84052908e+00 8.26648057e-01 4.22125846e-01
-7.45371059e-02 3.61570746e-01 2.41344810e-01 6.83825195e-01
-7.01506212e-02 -2.55926937e-01 7.81834781e-01 -3.90312523e-01
-2.94646472e-01 8.49753857e-01 -8.31390440e-01 -2.16159046e-01
8.89801979e-01 -1.42878190e-01 -1.90878049e-01 -1.06762595e-01
-1.11165798e+00 7.59966195e-01 -1.95583794e-02 2.51441121e-01
4.83404636e-01 -1.51221383e+00 -7.77649343e-01 3.57670695e-01
1.70164183e-01 2.74279833e-01 3.27824444e-01 7.17568934e-01
-2.80395418e-01 7.09849358e-01 2.83630937e-01 -6.78022563e-01
-1.31171930e+00 3.53605777e-01 7.69637883e-01 -3.39273483e-01
-3.22228283e-01 2.81465858e-01 -4.01716471e-01 -6.92140758e-01
-1.06025143e-02 -5.61323643e-01 -3.62850100e-01 3.45407337e-01
8.35194290e-01 9.46580112e-01 4.76260513e-01 -8.11067820e-01
-5.03635406e-02 4.86386031e-01 7.53705978e-01 -1.50108980e-02
1.51846349e+00 -4.87319142e-01 -1.43550172e-01 7.20410168e-01
1.27253759e+00 -6.25887692e-01 -8.88821781e-01 -2.09372714e-01
-1.80190280e-01 -6.94856465e-01 3.37606162e-01 -1.10879648e+00
-1.10536754e+00 1.10650003e+00 1.31189823e+00 6.29048586e-01
1.12774658e+00 -2.25376353e-01 1.38611794e+00 1.65041849e-01
-1.52230605e-01 -1.58340013e+00 -4.10433918e-01 5.13626277e-01
7.45065987e-01 -1.39584780e+00 -1.18175291e-01 1.72164261e-01
-6.99751899e-02 7.53554821e-01 2.63009042e-01 4.73285824e-01
1.28262317e+00 -2.02430129e-01 7.76226148e-02 -5.74667454e-01
-6.68925464e-01 -6.20962083e-01 4.97765303e-01 8.56911182e-01
1.50339589e-01 5.95906794e-01 7.88807496e-02 -1.12993874e-01
-1.32241711e-01 -8.68564025e-02 -1.25197724e-01 6.11191809e-01
-7.86853969e-01 -9.95461881e-01 -7.09432900e-01 8.02991092e-01
-3.16122383e-01 4.27998900e-02 8.82937610e-02 1.97014287e-01
8.74726951e-01 1.66788101e+00 2.69757181e-01 8.61020088e-02
9.47163045e-01 2.28652105e-01 -2.23700047e-01 -3.86760145e-01
-7.40041971e-01 -5.58824360e-01 5.29040515e-01 -7.68658638e-01
-8.44560325e-01 -9.45233047e-01 -1.13433087e+00 -4.23891991e-01
-9.38177034e-02 1.44499525e-01 1.06739938e+00 1.05127132e+00
1.86782628e-02 6.62075341e-01 9.60929692e-01 -4.19047832e-01
1.16300583e-01 -7.07868338e-01 -6.80229485e-01 4.11368579e-01
7.86752999e-01 -2.32682943e-01 -3.81576300e-01 -1.92035720e-01] | [6.2215962409973145, 2.5078747272491455] |
1621e0ad-27be-47f8-8624-d80997f7866c | can-wifi-estimate-person-pose | 1904.00277 | null | http://arxiv.org/abs/1904.00277v2 | http://arxiv.org/pdf/1904.00277v2.pdf | Can WiFi Estimate Person Pose? | WiFi human sensing has achieved great progress in indoor localization,
activity classification, etc. Retracing the development of these work, we have
a natural question: can WiFi devices work like cameras for vision applications?
In this paper We try to answer this question by exploring the ability of WiFi
on estimating single person pose. We use a 3-antenna WiFi sender and a
3-antenna receiver to generate WiFi data. Meanwhile, we use a synchronized
camera to capture person videos for corresponding keypoint annotations. We
further propose a fully convolutional network (FCN), termed WiSPPN, to estimate
single person pose from the collected data and annotations. Evaluation on over
80k images (16 sites and 8 persons) replies aforesaid question with a positive
answer. Codes have been made publicly available at
https://github.com/geekfeiw/WiSPPN. | ['Stanislav Panev', 'Fei Wang', 'Ziyi Dai', 'Dong Huang', 'Jinsong Han'] | 2019-03-30 | null | null | null | null | ['rf-based-pose-estimation'] | ['computer-vision'] | [ 8.47598836e-02 -1.52912214e-01 5.18835224e-02 -5.13655484e-01
-7.53588200e-01 -7.45557487e-01 5.52424252e-01 -7.45077729e-01
-3.53983819e-01 9.31003928e-01 6.23895943e-01 2.30185036e-03
-4.81711812e-02 -6.51329160e-01 -9.30283308e-01 -5.31645894e-01
-7.09176343e-03 -3.25749163e-03 -6.73320070e-02 3.47269237e-01
-1.35156959e-01 1.79810703e-01 -1.19904435e+00 2.70513028e-01
2.70273119e-01 1.04716456e+00 -1.47873327e-01 9.83971238e-01
3.89510840e-01 5.60692728e-01 -5.74430466e-01 -4.64215904e-01
4.40537453e-01 1.99726149e-02 -6.84330106e-01 -1.47911876e-01
7.39565909e-01 -5.60968399e-01 -8.67536247e-01 6.72538757e-01
7.47576177e-01 2.34634802e-02 1.31779283e-01 -1.57070780e+00
-3.05678964e-01 4.45615202e-01 -4.46381003e-01 1.78419635e-01
1.36055934e+00 8.12808871e-02 4.00929987e-01 -8.15682054e-01
1.95086569e-01 9.35419917e-01 1.32271922e+00 5.84035397e-01
-5.15343904e-01 -8.94063950e-01 -1.42428711e-01 2.08001301e-01
-1.87245500e+00 -6.28263533e-01 3.66384089e-01 -2.15080053e-01
7.61862516e-01 3.43440235e-01 8.61771822e-01 2.01235986e+00
-8.29894915e-02 7.41551101e-01 8.87022376e-01 -2.39574865e-01
-9.87677202e-02 -9.92523581e-02 1.78786423e-02 4.91438955e-01
5.11730075e-01 1.37660965e-01 -8.94959688e-01 -2.36625090e-01
9.63774920e-01 4.54969943e-01 -8.32736254e-01 2.03571320e-01
-1.51473010e+00 3.00209224e-01 4.87441748e-01 3.43917847e-01
-3.54447484e-01 6.75389349e-01 -2.15387061e-01 1.04299048e-02
-7.46024586e-03 1.32303312e-01 -2.89328068e-01 -5.67746341e-01
-8.14501822e-01 3.57717246e-01 8.15678716e-01 1.21107233e+00
5.56202292e-01 -4.27365810e-01 -1.92586690e-01 4.58033323e-01
6.35185242e-01 9.22599852e-01 2.46928722e-01 -1.22704875e+00
5.64023256e-01 2.10891739e-01 6.38762176e-01 -8.63465548e-01
-4.88373071e-01 -3.31413001e-01 -7.27831721e-01 -3.25722009e-01
7.81983256e-01 -9.48725998e-01 -5.99065781e-01 1.58535337e+00
2.26596385e-01 8.20409119e-01 -3.19364101e-01 1.11577559e+00
6.44179165e-01 4.94974375e-01 -1.41032681e-01 3.16637635e-01
1.48107636e+00 -9.89995956e-01 -5.08345842e-01 -2.09105700e-01
1.87601969e-01 -6.02656305e-01 7.91008353e-01 6.13515079e-01
-7.18118727e-01 -8.38846445e-01 -7.87120163e-01 2.10931867e-01
-1.40076324e-01 2.80979186e-01 7.62746632e-01 1.11273563e+00
-1.10186923e+00 1.57280460e-01 -9.38892305e-01 -8.01055908e-01
2.31501162e-01 6.30031407e-01 -5.67991078e-01 -3.17540079e-01
-1.23362184e+00 2.44566217e-01 -1.47722438e-01 4.39237177e-01
-8.16185534e-01 -3.34919721e-01 -5.14103770e-01 -1.02086850e-01
3.15072209e-01 -1.07431304e+00 1.42184496e+00 -8.92165780e-01
-1.32827139e+00 3.86354595e-01 -3.32727253e-01 -3.72861117e-01
5.09606600e-01 -7.97271192e-01 -8.79972577e-01 -6.56831041e-02
2.11240962e-01 4.90972966e-01 5.33772230e-01 -9.76606131e-01
-7.43074596e-01 -4.82151747e-01 2.63246715e-01 -1.58795267e-01
-3.60840112e-01 -1.11025319e-01 -5.94861507e-01 -5.92271507e-01
-1.48713320e-01 -1.22284436e+00 -4.91427630e-02 -8.09184611e-02
-6.44369483e-01 -7.18698725e-02 4.06628728e-01 -5.52148998e-01
1.01285970e+00 -1.92420721e+00 -4.53826636e-01 4.14520353e-01
1.80538818e-01 8.89510512e-02 9.00532454e-02 4.41844732e-01
2.36333609e-01 -2.13450104e-01 2.37024292e-01 -4.98831153e-01
5.63999787e-02 -8.08353722e-02 -3.22804898e-02 6.12683713e-01
-5.07109165e-01 9.36464489e-01 -8.23439419e-01 -1.50325865e-01
1.56795636e-01 1.07123041e+00 -5.22619843e-01 2.01611936e-01
6.18082643e-01 6.05702579e-01 -5.07330298e-01 7.42109835e-01
7.67659724e-01 -3.84422183e-01 1.35360718e-01 -2.59652138e-01
-8.92161503e-02 -3.00394129e-02 -1.67149985e+00 2.03574252e+00
-2.99209327e-01 6.75904334e-01 8.05115700e-02 -4.80408221e-01
5.57015836e-01 5.62670410e-01 4.49344933e-01 -1.95400670e-01
2.31240869e-01 -5.11285253e-02 -5.56802392e-01 -6.83211088e-01
3.16775471e-01 3.81148875e-01 -2.95828849e-01 2.97134817e-01
1.83993161e-01 9.28460181e-01 -1.49885684e-01 1.08171590e-01
1.70056510e+00 3.77358764e-01 5.91145046e-02 -6.48141578e-02
5.95511079e-01 -3.28865618e-01 3.35914105e-01 1.18809223e+00
-3.36006641e-01 8.22754264e-01 -3.28870118e-01 -5.76085031e-01
-3.51164967e-01 -1.51115227e+00 2.13220850e-01 9.44075704e-01
2.57891417e-01 -8.96496654e-01 -8.76734376e-01 -6.28293097e-01
1.02816857e-02 -3.42755243e-02 -5.79127133e-01 2.70762593e-01
-6.60179377e-01 -4.16646034e-01 1.16202998e+00 7.80919433e-01
9.37101722e-01 -7.22099900e-01 -5.37348211e-01 3.29747535e-02
-8.22972655e-01 -1.28421175e+00 -6.68883741e-01 -2.03960314e-01
-1.49472594e-01 -1.26260936e+00 -1.02415979e+00 -4.56116587e-01
4.22077835e-01 7.44544208e-01 9.20820415e-01 3.90328765e-02
2.47599818e-02 1.07898569e+00 -3.12392980e-01 -3.29987437e-01
6.07017636e-01 2.12703958e-01 4.60747629e-01 2.80288368e-01
7.45736897e-01 -7.35386133e-01 -1.06739450e+00 5.93685091e-01
-2.62282223e-01 -4.15941089e-01 4.65092391e-01 3.98694091e-02
5.33755645e-02 -2.10654542e-01 1.70217022e-01 -4.66430575e-01
3.12239200e-01 -6.05935276e-01 -3.12515169e-01 1.11386746e-01
-1.33698523e-01 -4.36620742e-01 3.20460916e-01 -1.92817003e-01
-1.04781473e+00 4.77684438e-01 -4.87852037e-01 -2.30643794e-01
-7.45207965e-01 -4.16254885e-02 -3.21287274e-01 -1.65305898e-01
7.12180555e-01 4.52817306e-02 -5.88049114e-01 -5.95933735e-01
2.02955008e-01 1.10645485e+00 9.08001184e-01 -5.23737729e-01
8.32266331e-01 8.03195119e-01 -4.75276828e-01 -8.55373621e-01
-7.60819197e-01 -8.31237614e-01 -4.68030006e-01 -4.92835402e-01
8.76621962e-01 -1.40522945e+00 -1.22060442e+00 5.88725209e-01
-1.14779305e+00 -1.96627483e-01 2.81024426e-01 6.56686604e-01
-2.27505088e-01 4.08592016e-01 -4.97421265e-01 -8.40886831e-01
-2.54554331e-01 -7.54225135e-01 1.26095057e+00 5.48609555e-01
-3.57273132e-01 -6.87208652e-01 2.85824448e-01 7.95334697e-01
5.53649068e-01 1.54273018e-01 -5.67970037e-01 -5.24093397e-02
-7.15546250e-01 -4.60680872e-01 -4.02631238e-03 -2.64430583e-01
1.21603362e-01 -5.35874844e-01 -1.41173136e+00 -2.98127502e-01
-2.92376667e-01 -4.17395160e-02 6.33956969e-01 6.30516887e-01
1.13507700e+00 -3.25402796e-01 -8.18877339e-01 9.92554426e-01
1.28679717e+00 -2.11618900e-01 8.74211371e-01 4.01596934e-01
7.91511476e-01 1.46456718e-01 6.79708347e-02 6.36694193e-01
6.27772391e-01 9.73444462e-01 1.82944611e-01 2.19586700e-01
-1.44154951e-01 -5.05259573e-01 6.06288612e-01 1.13878422e-03
-8.19920719e-01 -7.48390436e-01 -9.30797458e-01 1.02876127e-01
-1.85121942e+00 -1.11598754e+00 -4.32844341e-01 2.00120401e+00
9.69514847e-02 2.28852127e-02 4.18791234e-01 5.96590899e-02
6.99502587e-01 -2.67697051e-02 2.64003202e-02 4.62923467e-01
1.95815653e-01 1.09849358e-02 7.29378521e-01 4.65194404e-01
-1.31534040e+00 4.99679208e-01 5.63869286e+00 3.08020115e-01
-9.44522202e-01 2.74631768e-01 1.65029138e-01 -2.32354239e-01
-8.09896458e-03 -2.89698303e-01 -1.07295263e+00 7.83135772e-01
1.04686332e+00 5.21647453e-01 5.51554322e-01 7.77218878e-01
3.62311214e-01 -1.71767935e-01 -9.59246337e-01 1.51864982e+00
1.05674841e-01 -1.16675591e+00 -5.29464960e-01 3.10112655e-01
4.86851126e-01 2.01492757e-01 -1.28754646e-01 2.21307695e-01
5.03055044e-02 -1.02815235e+00 5.76947212e-01 9.13688719e-01
7.12868571e-01 -3.98558915e-01 8.34517419e-01 2.78318375e-01
-1.64701343e+00 -1.99519470e-01 -1.03985667e-01 -4.57500845e-01
3.78010511e-01 4.65541065e-01 -6.15984678e-01 5.04231095e-01
1.28537333e+00 5.35525918e-01 -6.70237958e-01 1.34215558e+00
-3.08270663e-01 7.67172456e-01 -6.12368941e-01 1.28436267e-01
-3.04427948e-02 2.11099967e-01 2.21071482e-01 1.40476668e+00
8.08962047e-01 9.39932615e-02 8.38467106e-02 4.10999715e-01
-1.77952752e-01 -4.42756683e-01 -6.71223223e-01 6.16004407e-01
6.58556879e-01 1.32910573e+00 -7.11693645e-01 -5.34116998e-02
-6.49769962e-01 1.27316630e+00 -9.92209613e-02 4.95758355e-01
-1.08946502e+00 -2.51979858e-01 7.16813505e-01 4.63565916e-01
1.21247344e-01 -3.48531902e-01 1.40952826e-01 -1.37823713e+00
1.07839309e-01 -5.00543118e-01 2.48096853e-01 -9.66750085e-01
-1.12638223e+00 2.86455780e-01 -2.05158904e-01 -1.29612219e+00
-2.42005557e-01 -5.74772239e-01 -5.39862216e-01 5.49524486e-01
-9.97312546e-01 -1.32763159e+00 -1.12099588e+00 1.22062457e+00
1.81724742e-01 1.26073360e-01 7.11392164e-01 8.49318504e-01
-4.37431604e-01 8.48167479e-01 -1.31574869e-01 6.21957004e-01
8.95378888e-01 -1.11461580e+00 4.99030560e-01 8.65685940e-01
4.43152487e-01 9.34351623e-01 5.68105400e-01 -5.09379327e-01
-1.49925148e+00 -1.00923777e+00 1.04163408e+00 -1.06004596e+00
1.89848300e-02 -6.18085980e-01 -6.80357665e-02 1.10850811e+00
2.33569115e-01 2.69952893e-01 9.14639056e-01 1.25260144e-01
-1.66028157e-01 -1.95604280e-01 -1.07061374e+00 2.85377413e-01
1.58976710e+00 -3.53835851e-01 -4.42169875e-01 2.50708282e-01
2.22135559e-01 -2.90119290e-01 -6.32941663e-01 4.44329530e-02
1.15835154e+00 -1.03608346e+00 1.27977490e+00 1.02761321e-01
-1.36865914e-01 -4.83748794e-01 -3.21447104e-01 -8.04495454e-01
-4.16788608e-01 -6.38378978e-01 -3.33596855e-01 1.09426439e+00
7.93779492e-02 -5.00716031e-01 1.23084641e+00 5.72987616e-01
5.82863018e-02 -3.26250255e-01 -1.01610887e+00 -7.01797724e-01
-6.59908831e-01 -8.00851643e-01 7.13273227e-01 6.80090308e-01
-9.29041430e-02 1.64520800e-01 -8.50647748e-01 5.48906326e-01
6.42281175e-01 -3.97986680e-01 1.08940625e+00 -1.13902998e+00
-5.08915782e-01 2.40066811e-01 -4.19330418e-01 -1.54189622e+00
-3.65025371e-01 -3.72368455e-01 -4.09013443e-02 -1.48438978e+00
-8.85177171e-04 -1.99647069e-01 -2.16126263e-01 4.88501012e-01
1.53949931e-01 8.68629873e-01 7.21687227e-02 1.05794147e-01
-9.51956272e-01 1.61082260e-02 5.19953251e-01 -4.34743389e-02
1.31965056e-01 5.04236698e-01 -7.20383406e-01 8.26857328e-01
8.98297071e-01 -1.62058800e-01 -1.49551302e-01 -6.75023854e-01
3.74477595e-01 -2.70394422e-02 1.04808867e+00 -1.79650986e+00
4.62382406e-01 2.59347469e-01 1.14046264e+00 -2.66805798e-01
6.38579667e-01 -1.13690233e+00 5.47732592e-01 3.69259089e-01
3.59515361e-02 1.90854102e-01 -1.75517365e-01 6.81651533e-01
2.12015197e-01 5.34282736e-02 1.30714299e-02 -2.75756031e-01
-7.34833479e-01 3.75565052e-01 -4.92408484e-01 -2.57647812e-01
7.67196894e-01 -4.67926323e-01 -3.11394036e-01 -8.00789475e-01
-7.50639260e-01 5.71779311e-02 2.90181249e-01 5.49149930e-01
5.20829320e-01 -1.53028274e+00 -3.39720339e-01 3.10097605e-01
-8.07750784e-03 -4.87392217e-01 3.15853387e-01 7.63384283e-01
-3.56173813e-01 6.81542218e-01 -4.61155325e-02 -5.32681227e-01
-1.05670452e+00 8.22348520e-02 4.98120129e-01 1.59322783e-01
-4.17872578e-01 1.04133320e+00 -2.21485570e-01 -3.95597309e-01
4.67574269e-01 -1.70247748e-01 -8.37406814e-02 -3.16680670e-01
8.55854392e-01 7.12403774e-01 -1.16235226e-01 -6.35962725e-01
-8.30701351e-01 7.43766010e-01 5.24508715e-01 -1.89792380e-01
1.12301755e+00 -4.92068261e-01 5.38164973e-01 -1.00885518e-01
9.56361949e-01 4.19708848e-01 -1.43766046e+00 2.81473547e-02
-1.25586867e-01 -6.34802282e-01 -2.44749948e-01 -8.45082462e-01
-9.74532366e-01 5.42365015e-01 1.12847328e+00 1.10069245e-01
7.68677771e-01 2.26279870e-01 9.23183143e-01 5.74204385e-01
8.43965411e-01 -7.11063862e-01 -1.32434785e-01 3.07488054e-01
4.61010933e-01 -1.16454935e+00 -2.28953287e-01 -3.12900335e-01
-1.14587486e-01 1.09695208e+00 6.46442473e-01 -1.14872426e-01
7.57001042e-01 3.50911379e-01 1.73943624e-01 -1.04480304e-01
-1.14659019e-01 -1.65868059e-01 -1.02389231e-02 1.04342294e+00
5.08884728e-01 2.88804732e-02 3.01073700e-01 9.50308979e-01
-5.12318313e-01 5.95693529e-01 3.52289379e-01 9.16681170e-01
-2.87419826e-01 -9.34540510e-01 -7.87250578e-01 3.46048743e-01
-5.33167899e-01 2.37551227e-01 -2.29689911e-01 5.32470644e-01
6.81711197e-01 1.33487236e+00 -2.23801091e-01 -6.84072196e-01
3.20279837e-01 -1.28430650e-01 5.59757113e-01 -2.20715806e-01
-5.76279998e-01 -1.02159113e-01 1.22019134e-01 -9.02628720e-01
-7.35235155e-01 -6.36917472e-01 -6.74151242e-01 -4.71095830e-01
-7.00474009e-02 2.98720658e-01 4.28199351e-01 9.17440116e-01
3.67698491e-01 2.58859068e-01 2.38886565e-01 -1.04410601e+00
-6.69999495e-02 -9.07957673e-01 -4.38669950e-01 1.14854552e-01
4.41195577e-01 -5.37372768e-01 -2.49017701e-01 7.68223554e-02] | [6.786060810089111, 0.5594683289527893] |
43d3bfcb-5bfd-4903-ba1d-b3abe0488bfe | fedmt-federated-learning-with-mixed-type | 2210.02042 | null | https://arxiv.org/abs/2210.02042v2 | https://arxiv.org/pdf/2210.02042v2.pdf | FedMT: Federated Learning with Mixed-type Labels | In federated learning (FL), classifiers (e.g., deep networks) are trained on datasets from multiple centers without exchanging data across them, and thus improves sample efficiency. In the classical setting of FL, the same labeling criterion is usually employed across all centers being involved in training. This constraint greatly limits the applicability of FL. For example, standards used for disease diagnosis are more likely to be different across clinical centers, which mismatches the classical FL setting. In this paper, we consider an important yet under-explored setting of FL, namely FL with mixed-type labels where different labeling criteria can be employed by various centers, leading to inter-center label space differences and challenging existing FL methods designed for the classical setting. To effectively and efficiently train models with mixed-type labels, we propose a theory-guided and model-agnostic approach that can make use of the underlying correspondence between those label spaces and can be easily combined with various FL methods such as FedAvg. We present convergence analysis based on over-parameterized ReLU networks. We show that the proposed method can achieve linear convergence in label projection, and demonstrate the impact of the parameters of our new setting on the convergence rate. The proposed method is evaluated and the theoretical findings are validated on benchmark and medical datasets. | ['Xiaoxiao Li', 'Gang Niu', 'Aline Talhouk', 'Qiong Zhang'] | 2022-10-05 | null | null | null | null | ['type'] | ['speech'] | [ 1.92338571e-01 -1.04323486e-02 -5.31595707e-01 -4.87315834e-01
-7.14392602e-01 -6.59038723e-01 1.96141869e-01 1.86113119e-01
-5.00018239e-01 7.94643879e-01 -6.90657422e-02 -3.37235838e-01
-3.77451897e-01 -6.89300299e-01 -6.69700682e-01 -1.03459167e+00
1.47275999e-01 5.22487104e-01 -3.35368603e-01 2.37817034e-01
-3.68254662e-01 5.70439398e-01 -1.12513804e+00 1.70429781e-01
8.74084175e-01 1.02173674e+00 -6.86811507e-02 1.38916135e-01
-2.53416389e-01 7.82950878e-01 -3.66400123e-01 -3.90451998e-01
5.97698092e-01 -4.52751249e-01 -8.78908932e-01 3.15786004e-01
4.26108867e-01 -7.47924671e-02 -8.95949975e-02 1.23267424e+00
4.30676103e-01 1.51469603e-01 5.18669903e-01 -1.21248007e+00
-5.55076003e-01 8.41888547e-01 -4.95505661e-01 -3.82969916e-01
-2.58199871e-01 -1.65315717e-01 9.95144546e-01 -6.47486925e-01
6.48934543e-01 1.09063339e+00 8.18203628e-01 7.60203779e-01
-1.27542293e+00 -6.96749508e-01 3.01324427e-01 2.10090913e-02
-1.37855017e+00 -2.00015843e-01 7.16388822e-01 -4.89568293e-01
4.02790047e-02 3.52653414e-01 3.01118523e-01 9.03260112e-01
-6.29717782e-02 7.12855637e-01 1.31164765e+00 -5.04506230e-01
4.00897175e-01 3.89446139e-01 3.22242916e-01 6.60264373e-01
3.88749301e-01 -1.51300773e-01 -3.20534140e-01 -5.41907549e-01
4.74336118e-01 5.44029236e-01 -5.48919797e-01 -8.35152984e-01
-1.39093220e+00 1.04037189e+00 5.53782761e-01 1.90930724e-01
-2.02941492e-01 -1.44696191e-01 5.86717725e-01 3.38313460e-01
5.21538317e-01 2.97347218e-01 -4.86803412e-01 4.69799936e-01
-8.84646058e-01 -9.64833051e-02 6.77256107e-01 9.84534919e-01
7.98429370e-01 -4.05434906e-01 -2.59256423e-01 8.98311913e-01
6.77331239e-02 1.77196503e-01 5.17877400e-01 -1.13741970e+00
4.38013792e-01 7.56467879e-01 3.79385054e-02 -7.78421283e-01
-5.36304593e-01 -5.81509173e-01 -1.21522439e+00 -1.15321875e-01
5.79886675e-01 -3.51237297e-01 -5.49381673e-01 2.14693213e+00
7.12910116e-01 3.13421458e-01 4.53252904e-02 7.13063061e-01
3.96071255e-01 1.98001023e-02 -1.17906205e-01 -4.10809934e-01
1.10166526e+00 -1.18645632e+00 -7.27434099e-01 1.96578383e-01
1.30650604e+00 -4.24570739e-01 8.41160119e-01 2.78762370e-01
-7.70774603e-01 -3.42477672e-02 -7.75412440e-01 1.97947770e-01
-2.40443543e-01 2.05538005e-01 6.46002710e-01 7.07454920e-01
-1.06491137e+00 5.90896785e-01 -6.40358567e-01 -4.64127541e-01
7.23189652e-01 4.23673600e-01 -4.70363349e-01 -4.85689342e-01
-1.00900400e+00 4.30849135e-01 3.28798324e-01 2.01403543e-01
-6.74412906e-01 -8.00542891e-01 -4.79716122e-01 -1.42726591e-02
5.69397748e-01 -7.54971445e-01 1.22691071e+00 -9.17255044e-01
-1.23065138e+00 7.35130847e-01 1.00123674e-01 -3.36665273e-01
8.29624236e-01 1.96622834e-01 -2.70548224e-01 9.25123617e-02
-5.93997352e-03 3.92612517e-01 4.81356949e-01 -1.12469077e+00
-7.77124822e-01 -4.67682034e-01 2.58315504e-01 1.15347035e-01
-7.98431933e-01 -1.21247739e-01 -1.27801076e-01 -5.06238520e-01
3.76135707e-02 -1.14877617e+00 -3.91930073e-01 2.15322047e-01
-4.86675262e-01 -2.03128532e-01 6.98060393e-01 -1.40910670e-01
1.04550219e+00 -2.17709661e+00 7.65239866e-03 2.11377174e-01
4.32604969e-01 1.23707242e-01 -1.01607978e-01 2.79387534e-01
2.31173113e-02 5.66651300e-02 -3.92252088e-01 -5.44459224e-01
-3.97654921e-02 2.76888758e-01 -8.13655779e-02 9.37641084e-01
-2.30303824e-01 5.93728900e-01 -9.48641479e-01 -6.68340623e-01
-1.04950584e-01 3.92588198e-01 -6.02582037e-01 5.75428233e-02
-6.63328543e-02 7.16230571e-01 -6.14592791e-01 5.93832552e-01
7.11154759e-01 -7.27302134e-01 6.49294376e-01 -1.77674279e-01
1.41202778e-01 -1.17614955e-01 -1.24269676e+00 1.79136658e+00
-6.64633036e-01 1.55384749e-01 2.48141170e-01 -1.21197009e+00
7.85680592e-01 4.98344451e-01 7.67537951e-01 -1.25821441e-01
2.41330713e-01 3.42644721e-01 -9.96005088e-02 -1.70423314e-01
-1.91286709e-02 -1.89332798e-01 5.40090203e-02 8.63029420e-01
-7.21810609e-02 5.65165222e-01 -3.22593264e-02 -2.16707047e-02
1.04777431e+00 -4.17599469e-01 2.70252198e-01 -4.28724080e-01
5.85312068e-01 -2.30285004e-01 9.34646666e-01 9.24923837e-01
-4.42518443e-01 3.37077081e-01 2.47949526e-01 -5.83198488e-01
-6.80575907e-01 -5.71797907e-01 -4.76786911e-01 1.11174572e+00
1.46840796e-01 -1.12341709e-01 -7.97398269e-01 -1.21532774e+00
8.35992098e-02 3.89461577e-01 -8.21923554e-01 -1.45933196e-01
-5.06715953e-01 -8.88160825e-01 5.06490827e-01 2.70077318e-01
2.75861442e-01 -7.78616905e-01 -5.81208110e-01 1.48100182e-01
-3.10910381e-02 -8.24453831e-01 -5.26826620e-01 2.84973323e-01
-9.69913125e-01 -1.18674803e+00 -7.96914041e-01 -7.40034044e-01
9.70605910e-01 3.68234128e-01 9.01344657e-01 -6.00997079e-03
-1.05307236e-01 3.40642333e-01 -2.51524687e-01 -1.61855578e-01
-4.01306480e-01 2.85788745e-01 1.85340017e-01 6.91124856e-01
1.17969215e-01 -3.64420652e-01 -8.00305545e-01 4.60425586e-01
-1.15656900e+00 1.21523242e-03 4.32648987e-01 1.24210668e+00
6.54561162e-01 -1.41146138e-01 7.83620775e-01 -1.57049084e+00
3.21563959e-01 -6.89555466e-01 -4.86876816e-01 7.61109471e-01
-8.46246064e-01 1.05807744e-01 1.00587845e+00 -6.81531072e-01
-9.44030166e-01 1.18500479e-01 3.52455109e-01 -7.22250104e-01
2.50812992e-02 4.37472820e-01 -2.35817254e-01 -3.34268361e-01
4.81236994e-01 -2.20699996e-01 4.46772464e-02 -6.20679319e-01
4.86567199e-01 8.14175189e-01 1.87901884e-01 -7.73360014e-01
4.39074725e-01 8.27118933e-01 1.15103476e-01 -1.26269072e-01
-1.00423229e+00 -3.57459724e-01 -4.74454910e-01 -5.89780845e-02
5.35950065e-01 -7.45519876e-01 -5.33705235e-01 3.49569976e-01
-9.57438529e-01 -2.53068894e-01 -5.58189452e-01 5.91789424e-01
-4.54332292e-01 2.33660907e-01 -7.23440707e-01 -5.94835222e-01
-4.07301724e-01 -1.26699543e+00 7.11767614e-01 1.23780712e-01
1.74060047e-01 -1.38685787e+00 1.22899629e-01 2.59344280e-01
4.20018375e-01 2.43564442e-01 1.08573794e+00 -9.61098075e-01
-3.56748253e-01 -1.17532797e-01 -2.38265723e-01 3.71563166e-01
4.65711296e-01 -4.02278692e-01 -9.53907907e-01 -7.43956447e-01
1.54412538e-01 -4.43457007e-01 6.13661647e-01 2.79101253e-01
1.43558908e+00 -4.67638701e-01 -5.56491733e-01 7.88699329e-01
1.43211961e+00 -2.46196389e-02 -1.15947783e-01 2.52948910e-01
8.95267427e-01 8.83364618e-01 5.07772923e-01 5.49807608e-01
2.51836628e-01 6.10471487e-01 2.25877866e-01 -3.11901033e-01
7.50919804e-02 5.25653921e-02 -5.19940145e-02 1.10237825e+00
4.13133949e-01 -1.37440175e-01 -7.88331032e-01 4.97896165e-01
-2.02265882e+00 -3.60739350e-01 2.18086004e-01 2.46184015e+00
8.88978243e-01 -4.98384088e-01 -8.18453506e-02 -1.75002694e-01
1.08195078e+00 -1.02685302e-01 -9.52624083e-01 -1.78619683e-01
-1.68443516e-01 -1.60680320e-02 7.87951112e-01 9.01895463e-02
-1.00593615e+00 3.59649718e-01 6.05364990e+00 8.80868077e-01
-1.24012017e+00 5.62179863e-01 9.73136544e-01 -3.82984102e-01
-1.73972875e-01 -9.31606665e-02 -6.21831834e-01 4.13865536e-01
8.62021983e-01 -3.40902030e-01 3.33237052e-01 8.43768001e-01
-9.10975635e-02 4.50468689e-01 -1.44073462e+00 1.04670453e+00
-1.19947754e-01 -1.40940344e+00 -1.62146360e-01 2.51839697e-01
1.07488048e+00 1.25413850e-01 8.00586268e-02 1.26929387e-01
5.82635462e-01 -7.23962486e-01 3.87244403e-01 3.15233171e-01
1.08929563e+00 -8.06065321e-01 7.02844918e-01 3.58803600e-01
-9.65139270e-01 -4.54603165e-01 -4.41729337e-01 3.94608885e-01
2.55580079e-02 7.92251348e-01 -5.54145098e-01 9.10398602e-01
6.13522887e-01 5.93324840e-01 -2.96808004e-01 1.01275861e+00
2.07623988e-01 5.67475557e-01 -3.11258316e-01 2.29515076e-01
1.57783598e-01 -3.60132039e-01 1.25146031e-01 9.86187518e-01
4.64384913e-01 -2.89135069e-01 3.29040855e-01 5.98586023e-01
-5.76277316e-01 4.71159339e-01 -6.85080051e-01 2.79077947e-01
7.15288222e-01 1.49250805e+00 -6.13609433e-01 -2.71995753e-01
-5.91025889e-01 5.73518753e-01 6.00540340e-01 3.68762463e-01
-7.28951454e-01 -3.04080211e-02 5.12768090e-01 -1.20365508e-01
8.96208286e-02 3.98887455e-01 -3.41595002e-02 -1.13697863e+00
2.61428207e-02 -8.94061387e-01 8.23174655e-01 5.10026002e-03
-1.76255536e+00 6.69776142e-01 -1.66524589e-01 -1.36829484e+00
2.73969416e-02 -3.03615689e-01 -3.01533759e-01 7.24494100e-01
-1.57011151e+00 -1.11141098e+00 -1.10074855e-01 8.56939495e-01
1.20660454e-01 -1.21719807e-01 7.56218731e-01 6.67677581e-01
-9.40523028e-01 1.07127988e+00 7.54436255e-01 2.36102622e-02
1.00390005e+00 -1.02296793e+00 -3.46018285e-01 6.05174601e-01
4.40925434e-02 6.97630048e-01 8.17513168e-02 -3.36854279e-01
-1.26702225e+00 -1.74088597e+00 6.74810588e-01 -4.09971401e-02
6.11429691e-01 -4.44556236e-01 -9.02464151e-01 7.79067338e-01
-1.54969499e-01 5.42763114e-01 1.01428950e+00 2.68870950e-01
-5.18755376e-01 -3.71263087e-01 -1.44365942e+00 5.78767657e-01
1.06644881e+00 -4.12943512e-01 5.77739812e-02 8.54925454e-01
7.45239496e-01 -5.82557879e-02 -9.66038167e-01 5.30119061e-01
2.75616735e-01 -6.93750679e-01 7.24711597e-01 -8.17053616e-01
-2.35356092e-02 -1.27323896e-01 -2.91562051e-01 -1.23209381e+00
-1.76288396e-01 -6.15340829e-01 -1.02381498e-01 1.12616789e+00
2.76392847e-01 -1.10062110e+00 9.11231101e-01 8.05651367e-01
8.55816081e-02 -1.01132250e+00 -1.21680713e+00 -9.28550601e-01
3.42430919e-01 -2.14082580e-02 9.50223088e-01 1.46168613e+00
-1.07686520e-01 -2.94334278e-03 -4.38216269e-01 1.85744502e-02
7.88389623e-01 3.33555609e-01 3.89456660e-01 -1.38646841e+00
-4.99857306e-01 -3.70499969e-01 -2.91537270e-02 -7.41183221e-01
3.98342222e-01 -1.25095439e+00 3.04780137e-02 -1.08244920e+00
4.40725178e-01 -1.25085187e+00 -7.24554956e-01 6.57133341e-01
-1.00573786e-01 1.69023350e-01 2.63905048e-01 5.81886768e-01
-7.30872929e-01 4.66842443e-01 1.24836862e+00 -1.37812674e-01
-6.06030365e-03 -3.13809849e-02 -7.66634345e-01 5.52086413e-01
7.61872649e-01 -7.50928044e-01 -5.18418610e-01 -2.75685430e-01
2.86423117e-02 -2.15990581e-02 3.00533548e-02 -8.29067945e-01
3.57407600e-01 -1.34445995e-01 -1.93306878e-02 -1.09550297e-01
-2.43431870e-02 -1.04387236e+00 3.63073051e-01 3.72078329e-01
-6.97931468e-01 -2.57095005e-02 -2.99421638e-01 7.00070322e-01
-4.81975377e-02 -3.49436402e-01 7.85922110e-01 -2.62788050e-02
6.99008256e-02 6.47332728e-01 2.08793342e-01 1.08340429e-02
1.22956800e+00 1.17495723e-01 -4.47972745e-01 -8.95770118e-02
-5.50477326e-01 2.90125310e-01 5.98962128e-01 3.30424495e-02
8.08656663e-02 -1.51545560e+00 -6.53963327e-01 4.86301146e-02
1.95905939e-01 1.84200361e-01 4.92002755e-01 1.25375807e+00
-1.83274254e-01 4.09321517e-01 1.13255531e-01 -6.31574929e-01
-9.49740827e-01 8.76220942e-01 4.50569004e-01 -6.35233521e-01
-6.86743200e-01 6.21453345e-01 6.99352980e-01 -1.03806686e+00
3.15361351e-01 -5.33393808e-02 5.90234576e-03 2.18550097e-02
4.35355633e-01 3.84543449e-01 2.16336727e-01 -4.44528699e-01
-2.82395065e-01 3.83313149e-01 -3.47786725e-01 1.78304568e-01
1.19401026e+00 -8.51600915e-02 -2.49541879e-01 6.10425472e-01
1.42395520e+00 -1.98896572e-01 -1.19968605e+00 -7.10383713e-01
-7.48886615e-02 -3.32430869e-01 -7.23779062e-03 -6.04818642e-01
-1.63014984e+00 7.05878377e-01 7.20766366e-01 7.47288018e-02
1.23636210e+00 1.65849074e-03 7.80141532e-01 3.39135021e-01
6.65836751e-01 -1.08642793e+00 -3.24087441e-01 -1.23676090e-02
2.66049027e-01 -1.13090384e+00 -1.90117851e-01 -2.85548627e-01
-3.67546231e-01 8.63944650e-01 3.55522841e-01 1.58363104e-01
6.83148921e-01 5.65450452e-02 2.35718593e-01 -9.53664333e-02
-6.82995200e-01 3.28389525e-01 -9.39414278e-02 2.44844109e-01
2.15617508e-01 3.46932322e-01 -4.67489362e-01 4.99439597e-01
3.19248140e-01 6.18327670e-02 4.57922518e-01 9.99454677e-01
2.22557560e-01 -1.43288159e+00 -4.81635123e-01 6.90083146e-01
-5.44211745e-01 1.25808626e-01 -1.12080134e-01 5.24341702e-01
2.53997862e-01 8.84637952e-01 -1.16705768e-01 -1.60778448e-01
3.67656946e-02 1.44842416e-01 2.99765646e-01 -4.99855846e-01
-6.96983635e-01 -5.26901670e-02 -2.96768010e-01 -5.08556128e-01
-6.00953043e-01 -7.00845599e-01 -1.01820171e+00 -3.35162014e-01
-4.67799306e-01 2.93855220e-01 5.87870955e-01 7.38366365e-01
5.55490136e-01 3.26693207e-01 9.97306228e-01 -3.56702209e-01
-9.25424099e-01 -7.75427639e-01 -8.76647413e-01 5.88218987e-01
2.93308049e-01 -8.14246953e-01 -5.40835440e-01 -1.82209060e-01] | [6.039620399475098, 6.435494899749756] |
2eb67920-ac3f-42e7-8893-e677e4499024 | boosting-value-decomposition-via-unit-wise | 2305.07182 | null | https://arxiv.org/abs/2305.07182v1 | https://arxiv.org/pdf/2305.07182v1.pdf | Boosting Value Decomposition via Unit-Wise Attentive State Representation for Cooperative Multi-Agent Reinforcement Learning | In cooperative multi-agent reinforcement learning (MARL), the environmental stochasticity and uncertainties will increase exponentially when the number of agents increases, which puts hard pressure on how to come up with a compact latent representation from partial observation for boosting value decomposition. To tackle these issues, we propose a simple yet powerful method that alleviates partial observability and efficiently promotes coordination by introducing the UNit-wise attentive State Representation (UNSR). In UNSR, each agent learns a compact and disentangled unit-wise state representation outputted from transformer blocks, and produces its local action-value function. The proposed UNSR is used to boost the value decomposition with a multi-head attention mechanism for producing efficient credit assignment in the mixing network, providing an efficient reasoning path between the individual value function and joint value function. Experimental results demonstrate that our method achieves superior performance and data efficiency compared to solid baselines on the StarCraft II micromanagement challenge. Additional ablation experiments also help identify the key factors contributing to the performance of UNSR. | ['Chunlin Chen', 'Zhi Wang', 'Zichuan Liu', 'Yuanyang Zhu', 'Qingpeng Zhao'] | 2023-05-12 | null | null | null | null | ['multi-agent-reinforcement-learning', 'starcraft-ii', 'starcraft'] | ['methodology', 'playing-games', 'playing-games'] | [-4.04127508e-01 3.71562451e-01 -4.33081955e-01 -6.01541437e-02
-8.02701831e-01 -5.60439289e-01 7.74398804e-01 -1.53500691e-01
-4.88624781e-01 1.15266871e+00 5.71502686e-01 -5.42975925e-02
-4.83722240e-01 -7.45613337e-01 -6.63379729e-01 -1.17990243e+00
-3.08952153e-01 6.80686295e-01 -1.50678873e-01 -4.41905499e-01
-1.62078872e-01 1.51812330e-01 -9.84419823e-01 -6.87716249e-03
8.41300189e-01 8.60520124e-01 1.48407623e-01 5.88247359e-01
4.60993856e-01 1.82130373e+00 -7.88325310e-01 -2.00851917e-01
4.68713671e-01 -4.43368375e-01 -6.62651956e-01 -8.59695580e-03
-3.25613916e-01 -8.16597223e-01 -7.04823732e-01 8.63148570e-01
4.54096913e-01 4.94622797e-01 6.27714992e-01 -1.70400763e+00
-5.57786584e-01 1.41986060e+00 -8.48607957e-01 -5.85678816e-02
-1.81197956e-01 4.37228769e-01 1.43518186e+00 -2.23428503e-01
4.00256306e-01 1.27892888e+00 2.19004348e-01 5.66508412e-01
-1.31034446e+00 -7.44728446e-01 5.67979217e-01 1.77448004e-01
-8.68286490e-01 -3.25249523e-01 7.34805226e-01 4.91708964e-02
1.08601093e+00 -4.48409244e-02 8.35130811e-01 1.06914032e+00
1.02497540e-01 1.18562293e+00 9.81420040e-01 7.52226263e-02
6.39288783e-01 -1.46186888e-01 -2.32873470e-01 6.98663116e-01
4.48935598e-01 3.00320506e-01 -1.02400792e+00 -2.77495921e-01
1.04228950e+00 -9.75041315e-02 -5.28839789e-02 -7.80886531e-01
-1.37362754e+00 8.27935517e-01 7.21809685e-01 -1.20040685e-01
-6.36194408e-01 9.13883150e-01 2.68811792e-01 6.47339463e-01
2.02886999e-01 8.84933054e-01 -5.02727628e-01 -3.25454116e-01
-6.37118220e-01 3.69528860e-01 5.96397817e-01 8.36289346e-01
5.55743635e-01 3.33745360e-01 -1.29151374e-01 3.63973230e-01
5.52672267e-01 5.51721334e-01 2.82104284e-01 -1.54063547e+00
7.22584188e-01 6.27802432e-01 4.05866563e-01 -4.03256267e-01
-5.50031900e-01 -6.68049276e-01 -7.69289792e-01 5.29562473e-01
4.56721127e-01 -7.03004837e-01 -7.26433039e-01 2.26080036e+00
3.43975276e-01 3.04398127e-02 6.72426105e-01 1.07975161e+00
3.56344610e-01 7.17269599e-01 -1.55622393e-01 -8.29830915e-02
1.15103006e+00 -1.31605053e+00 -6.38824344e-01 -5.03718793e-01
4.60356146e-01 1.35968596e-01 5.71618915e-01 3.69511545e-01
-1.36069500e+00 3.90457874e-03 -1.29232550e+00 1.08010747e-01
1.98119879e-01 -1.90023407e-01 9.88739669e-01 1.42532408e-01
-9.23646808e-01 5.06175995e-01 -1.35683739e+00 2.57600248e-01
5.09518504e-01 5.29083788e-01 -3.73812258e-01 1.11954853e-01
-1.34384441e+00 1.09737122e+00 3.44473183e-01 1.51919788e-02
-1.67680979e+00 -7.31468499e-01 -9.19492006e-01 3.77622038e-01
7.15347171e-01 -7.91501582e-01 1.44752550e+00 -5.10966778e-01
-1.81695795e+00 -1.15564056e-01 5.29926479e-01 -7.16355145e-01
3.36180925e-01 -1.78936332e-01 3.18992108e-01 1.63497522e-01
3.27885896e-02 6.36756599e-01 8.38313639e-01 -1.13015246e+00
-6.17040455e-01 -1.95489824e-01 5.75829029e-01 7.37792730e-01
-4.66533959e-01 -4.20268148e-01 3.18025537e-02 -5.09733498e-01
-1.30943358e-01 -9.37442720e-01 -6.79535210e-01 -3.73614430e-01
-3.11235897e-02 -2.21313700e-01 2.09903449e-01 -5.56065500e-01
7.62079597e-01 -1.84194756e+00 1.06130016e+00 1.91072270e-01
5.28038263e-01 -2.87368983e-01 -4.09751952e-01 6.18052602e-01
9.51799303e-02 -3.69600207e-01 -1.92013517e-01 -4.54623163e-01
4.41538870e-01 4.42774028e-01 -6.10402703e-01 5.00576675e-01
6.49824515e-02 1.01763296e+00 -1.17923307e+00 -7.00564533e-02
-1.35136217e-01 2.16667593e-01 -9.02735829e-01 3.85843486e-01
-4.44844991e-01 3.73053402e-01 -6.76485062e-01 3.88939768e-01
4.12270308e-01 -2.41164342e-01 6.74170792e-01 1.99163007e-03
1.29991829e-01 2.25450963e-01 -1.24042737e+00 1.92771196e+00
-4.66812789e-01 2.23493978e-01 3.34074259e-01 -8.85560811e-01
5.87570786e-01 2.55981386e-01 5.80670118e-01 -7.27150738e-01
3.03727746e-01 -9.59633440e-02 1.13293149e-01 -1.99592598e-02
5.88219881e-01 -1.25414744e-01 -6.07667267e-01 9.18958127e-01
3.59963685e-01 -3.90710711e-01 1.91467762e-01 8.57314944e-01
1.33337665e+00 3.88166726e-01 1.64447859e-01 -2.63394535e-01
1.22563154e-01 -1.97484657e-01 9.08029020e-01 7.38483131e-01
-3.72652799e-01 8.84381831e-02 1.10284853e+00 -3.46692264e-01
-9.05589461e-01 -1.06861854e+00 6.38305247e-01 1.25035942e+00
3.65734637e-01 -4.88029629e-01 -5.54086149e-01 -6.84716046e-01
2.09175661e-01 7.16957867e-01 -7.18472600e-01 -4.59727705e-01
-5.20484328e-01 -9.49110210e-01 3.93697202e-01 8.34942758e-01
5.93099415e-01 -8.31248164e-01 -7.72681415e-01 3.50683630e-01
-2.44555160e-01 -7.42289782e-01 -3.60880375e-01 5.83538055e-01
-6.82364762e-01 -7.36846387e-01 -5.62223494e-01 -2.47077212e-01
6.59188211e-01 1.70467198e-01 1.00351405e+00 -1.42762989e-01
1.13740578e-01 3.78943592e-01 -2.50105143e-01 -1.28395602e-01
-3.27282012e-01 2.96738863e-01 4.10238862e-01 -1.65096849e-01
-3.47604215e-01 -7.67434597e-01 -5.58228314e-01 1.69083327e-01
-7.24205673e-01 1.79249614e-01 7.89267719e-01 1.07450712e+00
-4.35653701e-02 1.36416227e-01 8.31686616e-01 -3.44616711e-01
7.48700380e-01 -5.54453015e-01 -8.83979201e-01 2.82969624e-01
-6.60432160e-01 8.18837702e-01 6.10347807e-01 -3.47248733e-01
-1.39115417e+00 -1.15687825e-01 3.50814641e-01 -1.86832324e-01
6.23419106e-01 5.18599033e-01 -1.07939005e-01 2.29559973e-01
5.16098559e-01 1.79964378e-01 3.48476887e-01 -7.60753751e-02
8.96063566e-01 2.23801062e-01 3.55928868e-01 -9.95255411e-01
8.86959374e-01 4.36113834e-01 8.70178193e-02 -1.48053676e-01
-7.97735453e-01 1.58106521e-01 -8.00628215e-02 -2.28710532e-01
6.16831899e-01 -1.50834405e+00 -1.06906319e+00 5.78580379e-01
-7.94802248e-01 -7.02198148e-01 -6.61850929e-01 5.56826115e-01
-8.02172244e-01 -2.97353603e-03 -8.69926155e-01 -9.70912695e-01
-1.36116579e-01 -1.20799077e+00 6.63367927e-01 4.56307262e-01
1.35238379e-01 -8.14614952e-01 2.98646331e-01 6.33988142e-01
6.01055920e-01 -1.99328829e-03 6.95831716e-01 -2.40711689e-01
-9.20036376e-01 1.40642196e-01 1.20436303e-01 1.92691624e-01
-1.88671112e-01 -5.20663381e-01 -7.29732037e-01 -7.42620587e-01
-1.41215429e-01 -9.50949073e-01 9.64593709e-01 1.86978385e-01
4.96808320e-01 -3.99621755e-01 -5.20697460e-02 4.32136834e-01
1.03709269e+00 -6.53686225e-02 2.96815515e-01 3.24847490e-01
4.86225992e-01 4.26877737e-01 5.16613603e-01 9.52478409e-01
9.29144859e-01 3.37127060e-01 1.01108646e+00 1.74477682e-01
1.16438858e-01 -3.33531052e-01 8.71912301e-01 7.31162071e-01
-3.70885134e-01 -4.36914265e-02 -4.64959621e-01 4.25968647e-01
-2.42703676e+00 -1.03472018e+00 5.51784217e-01 1.85712624e+00
1.07966125e+00 3.51920631e-03 1.64542720e-01 -3.22410643e-01
1.97444439e-01 4.92508292e-01 -1.12878966e+00 3.25790681e-02
-4.21277322e-02 -2.71077871e-01 5.79512894e-01 5.96232772e-01
-7.03307390e-01 9.39197958e-01 6.19010401e+00 6.93170428e-01
-5.41059554e-01 2.40678340e-01 4.35939163e-01 -8.07922363e-01
-4.18331832e-01 -2.66848970e-03 -4.96425897e-01 1.78636566e-01
8.44254792e-01 -2.99466729e-01 1.01207101e+00 8.52305591e-01
-5.46760745e-02 -1.63788229e-01 -1.01775002e+00 6.79920018e-01
-1.88659728e-01 -1.33781362e+00 -3.75036985e-01 2.39821658e-01
1.00679123e+00 2.50295609e-01 2.29480028e-01 6.84659660e-01
1.47034538e+00 -6.67992055e-01 1.16211796e+00 2.79594243e-01
3.67375016e-01 -1.04514945e+00 4.66411054e-01 4.24457580e-01
-1.11795819e+00 -7.18283653e-01 -4.39388543e-01 -3.92045707e-01
7.83868879e-02 1.03731222e-01 -5.76409042e-01 5.97335637e-01
3.54718536e-01 5.42707324e-01 -1.42840058e-01 3.36582631e-01
-7.33102679e-01 4.24395293e-01 -2.32261986e-01 -3.45598273e-02
4.20459569e-01 -2.63962001e-01 5.26580453e-01 3.74736100e-01
-2.73405295e-02 1.40487894e-01 3.71279716e-02 9.84669685e-01
-3.63837540e-01 -5.95634162e-01 -1.55683875e-01 -3.92650276e-01
4.84578967e-01 1.56401920e+00 -6.46579325e-01 -2.31724262e-01
-1.98133066e-01 7.45495796e-01 8.92123401e-01 4.53067154e-01
-1.08928502e+00 -9.23856273e-02 1.02385747e+00 -5.04686654e-01
2.02017441e-01 -4.19174701e-01 -6.12152964e-02 -1.58912790e+00
-3.17587107e-01 -1.04860902e+00 5.20439684e-01 -6.50910318e-01
-1.18636239e+00 4.13834661e-01 2.67766807e-02 -9.69992518e-01
-5.57643712e-01 -2.53844082e-01 -4.33228344e-01 4.93251950e-01
-1.52549028e+00 -1.24678278e+00 4.98662628e-02 5.76227903e-01
2.59891242e-01 -6.25229180e-01 6.45237088e-01 -1.23807989e-01
-8.45922828e-01 4.10393983e-01 9.65601876e-02 -9.20581073e-02
5.03115892e-01 -1.40450466e+00 3.17901015e-01 9.36951876e-01
-2.31640656e-02 5.50080955e-01 6.98159277e-01 -4.33132768e-01
-1.83262658e+00 -8.06052566e-01 -2.03660782e-02 -5.67968011e-01
1.06390071e+00 -5.45055985e-01 -3.39294612e-01 7.87473977e-01
5.59444010e-01 -1.09519556e-01 5.33689022e-01 1.74346343e-01
-4.95465875e-01 -2.53354162e-01 -1.01960719e+00 8.08638453e-01
8.49323809e-01 -3.25987250e-01 -7.49502718e-01 4.85176630e-02
1.04696298e+00 -3.30004930e-01 -7.75036156e-01 -3.46449986e-02
4.28432137e-01 -6.84004128e-01 8.81056726e-01 -7.98030555e-01
5.98306358e-01 -3.36623132e-01 -1.46515295e-01 -1.77964008e+00
-6.03878617e-01 -1.08927870e+00 -6.40498281e-01 9.62979317e-01
4.27529752e-01 -7.18696237e-01 8.78795385e-01 9.60925162e-01
2.84915995e-02 -6.61929607e-01 -1.01861989e+00 -6.32825732e-01
1.06302388e-01 -2.22583190e-02 7.65536726e-01 8.32364500e-01
4.20547843e-01 5.27329922e-01 -8.87201071e-01 3.05333376e-01
1.08272064e+00 2.49602973e-01 6.27811372e-01 -7.89049566e-01
-6.89594686e-01 -2.40124494e-01 6.72154278e-02 -1.06115305e+00
2.38369182e-01 -8.44503522e-01 -5.10908067e-02 -1.62328053e+00
4.98115391e-01 -4.41453338e-01 -6.37661278e-01 7.71006823e-01
-7.32486695e-02 -7.06462935e-02 7.51436651e-01 1.96426198e-01
-1.09036601e+00 1.11657047e+00 1.31851387e+00 -2.57600874e-01
-1.76220715e-01 -3.21310848e-01 -1.24725282e+00 5.15175939e-01
7.32771516e-01 -4.47088271e-01 -6.98833764e-01 -6.62669420e-01
8.10909748e-01 3.79282922e-01 1.91897258e-01 -7.78440356e-01
5.21681488e-01 -5.72907567e-01 1.53987154e-01 -2.60838598e-01
5.02353609e-01 -8.07155669e-01 -4.58662678e-03 7.19146430e-01
-6.03498340e-01 -8.49883854e-02 -1.37695581e-01 6.27099156e-01
1.37201652e-01 -8.72403532e-02 6.03150904e-01 -1.39118090e-01
-4.65762407e-01 2.64433861e-01 -5.11601210e-01 4.41057496e-02
1.04528284e+00 5.34678698e-01 -6.01806879e-01 -7.03196228e-01
-5.19014120e-01 1.06663024e+00 2.78944135e-01 3.32549691e-01
4.13862258e-01 -1.48664856e+00 -7.20975399e-01 2.17275307e-01
-2.68403441e-01 1.51912823e-01 4.90682542e-01 7.03154624e-01
-1.96350351e-01 2.36636344e-02 -4.98806655e-01 8.79131164e-03
-5.74149728e-01 4.06734854e-01 4.66536641e-01 -8.04041684e-01
-4.03476238e-01 1.01547039e+00 1.73057213e-01 -4.53211367e-01
2.55761713e-01 -2.77178943e-01 -3.05586513e-02 3.32553864e-01
5.78622520e-01 6.32342398e-01 -4.37300980e-01 -9.45787802e-02
-2.83003926e-01 -4.43033762e-02 -2.22969487e-01 -6.38324618e-01
1.87396443e+00 -1.65309355e-01 -1.25863448e-01 6.00824729e-02
5.27268350e-01 -2.92652607e-01 -2.01926064e+00 -4.17818606e-01
-3.79624128e-01 -6.88915551e-02 3.80890548e-01 -9.79840457e-01
-1.33456767e+00 5.10649920e-01 5.33650955e-03 1.76189855e-01
8.75595212e-01 8.67069960e-02 5.91369689e-01 7.61883557e-01
8.08127403e-01 -1.39505649e+00 5.34395039e-01 5.72313488e-01
9.89601910e-01 -1.04969239e+00 1.41419917e-01 3.64168376e-01
-1.17181718e+00 7.42200732e-01 7.82874405e-01 -2.18382522e-01
1.84329882e-01 7.03105211e-01 -1.37720302e-01 -6.97033778e-02
-1.38911390e+00 -1.41329914e-01 -3.86251748e-01 6.12195790e-01
-2.70636618e-01 1.55455858e-01 2.19636887e-01 1.10109162e+00
-5.78123182e-02 -3.60625774e-01 8.53919923e-01 1.00410771e+00
-4.70632255e-01 -1.13854063e+00 -2.52762496e-01 1.49940312e-01
-1.45417556e-01 7.50627294e-02 -5.86892441e-02 4.86748457e-01
-3.81265044e-01 9.74142015e-01 -7.59841353e-02 -3.22561622e-01
8.47210214e-02 -3.21844697e-01 6.34595037e-01 -3.28420788e-01
-6.43043518e-01 5.99944917e-03 2.39770785e-02 -8.64304185e-01
-2.06757516e-01 -5.97372174e-01 -1.55838716e+00 -2.74750113e-01
-1.92639530e-01 3.67517024e-01 4.07132477e-01 6.74635828e-01
4.61856812e-01 9.22022879e-01 7.71654963e-01 -9.60675776e-01
-1.42223477e+00 -7.79451072e-01 -7.38320649e-01 1.67227462e-01
5.15994132e-01 -9.03215110e-01 -4.32392925e-01 -3.75391096e-01] | [3.7099640369415283, 2.044111490249634] |
a26136fc-500a-4a64-8872-b7ed87737c0c | predicting-optimal-value-functions-by | 1909.05004 | null | https://arxiv.org/abs/1909.05004v4 | https://arxiv.org/pdf/1909.05004v4.pdf | Predicting optimal value functions by interpolating reward functions in scalarized multi-objective reinforcement learning | A common approach for defining a reward function for Multi-objective Reinforcement Learning (MORL) problems is the weighted sum of the multiple objectives. The weights are then treated as design parameters dependent on the expertise (and preference) of the person performing the learning, with the typical result that a new solution is required for any change in these settings. This paper investigates the relationship between the reward function and the optimal value function for MORL; specifically addressing the question of how to approximate the optimal value function well beyond the set of weights for which the optimization problem was actually solved, thereby avoiding the need to recompute for any particular choice. We prove that the value function transforms smoothly given a transformation of weights of the reward function (and thus a smooth interpolation in the policy space). A Gaussian process is used to obtain a smooth interpolation over the reward function weights of the optimal value function for three well-known examples: GridWorld, Objectworld and Pendulum. The results show that the interpolation can provide very robust values for sample states and action space in discrete and continuous domain problems. Significant advantages arise from utilizing this interpolation technique in the domain of autonomous vehicles: easy, instant adaptation of user preferences while driving and true randomization of obstacle vehicle behavior preferences during training. | ['Arpan Kusari', 'Jonathan P. How'] | 2019-09-11 | null | null | null | null | ['multi-objective-reinforcement-learning'] | ['methodology'] | [-1.52330905e-01 3.24855931e-02 -3.81561816e-01 -2.81179845e-01
-8.19089234e-01 -5.55835545e-01 4.34316009e-01 1.60578355e-01
-1.04815531e+00 1.33861244e+00 -3.88111353e-01 -2.27272987e-01
-5.48846304e-01 -6.15685701e-01 -7.52805948e-01 -9.27640975e-01
-1.42103553e-01 6.15546644e-01 1.74252972e-01 -5.45535982e-01
4.73845452e-01 5.05611837e-01 -1.81135571e+00 -4.02588278e-01
1.20338082e+00 7.44270504e-01 3.89633179e-01 8.06960285e-01
1.75133720e-01 4.31362718e-01 -8.31597209e-01 5.22036888e-02
4.35078979e-01 -2.18144327e-01 -7.90573001e-01 7.36848265e-02
-1.09053567e-01 -8.35567638e-02 1.93727031e-01 1.08483648e+00
5.09117842e-01 7.21237242e-01 8.60451698e-01 -1.58574104e+00
-5.57355642e-01 1.64308906e-01 -3.52303833e-01 3.68713471e-03
2.51978099e-01 2.67086923e-01 7.87247658e-01 -3.67976338e-01
4.47913468e-01 1.14055073e+00 3.88167292e-01 6.90270126e-01
-1.21779263e+00 -1.76847816e-01 5.64931929e-02 3.77561688e-01
-1.29474998e+00 -1.73334792e-01 4.15037364e-01 -3.75365555e-01
6.95763350e-01 8.74171928e-02 4.80459839e-01 7.32389212e-01
3.98351938e-01 4.97651607e-01 1.21325946e+00 -4.48237091e-01
5.80385983e-01 6.53473377e-01 -3.12206209e-01 6.36664152e-01
3.33983898e-02 3.10488880e-01 1.53188646e-01 -1.98304400e-01
8.84541571e-01 -2.98074245e-01 -1.05548441e-01 -6.89278364e-01
-9.07439590e-01 9.97724891e-01 2.48098876e-02 1.33768007e-01
-7.34063804e-01 3.12774122e-01 2.39218011e-01 6.04049563e-01
5.38841374e-02 6.51619196e-01 -4.70141888e-01 -1.69823632e-01
-5.26532531e-01 8.11557174e-01 5.64788520e-01 7.53044128e-01
8.62259984e-01 3.46080780e-01 -3.40770304e-01 8.15554559e-01
1.51522174e-01 5.41994631e-01 5.51858664e-01 -1.29422069e+00
4.11457926e-01 1.57861978e-01 1.02368903e+00 -4.21531767e-01
-2.86096722e-01 -2.26960286e-01 -2.38116477e-02 1.11311710e+00
9.71567512e-01 -6.70333147e-01 -5.55789769e-01 1.82706916e+00
5.81816792e-01 -2.54614383e-01 2.08079293e-01 7.87376702e-01
-2.94603109e-02 6.15099311e-01 8.94657150e-02 -4.11871344e-01
1.00468957e+00 -6.90345764e-01 -6.93500698e-01 -1.57025471e-01
5.44154286e-01 -4.62713659e-01 1.27288091e+00 4.71376538e-01
-1.24280679e+00 -6.22979045e-01 -1.16040504e+00 5.67055464e-01
-4.22320485e-01 1.33577846e-02 1.19355142e-01 4.86777663e-01
-1.05046999e+00 1.06229913e+00 -3.62220973e-01 4.57639322e-02
-6.97284639e-02 5.97959876e-01 -8.57982337e-02 3.65708619e-01
-1.38677347e+00 1.65800452e+00 3.95567745e-01 -1.11314692e-01
-8.77497137e-01 -5.61985075e-01 -5.97630680e-01 -7.97364861e-02
4.93458897e-01 -2.96675831e-01 1.49742591e+00 -1.33854198e+00
-2.00312257e+00 3.62482786e-01 8.87633637e-02 -3.53022844e-01
8.27010334e-01 1.34275347e-01 -2.91581869e-01 -1.45279020e-01
1.90645859e-01 5.49535036e-01 1.19165647e+00 -1.21792948e+00
-9.31157351e-01 -4.54229079e-02 3.77800465e-01 5.25647342e-01
-1.88827999e-02 -2.40946516e-01 2.00908393e-01 -2.40496993e-01
-5.95132172e-01 -8.47576797e-01 -5.89277864e-01 -2.63147503e-01
2.72546560e-01 -5.53320289e-01 5.65359890e-01 -4.99240071e-01
1.03330624e+00 -1.93065381e+00 3.72477055e-01 3.77006799e-01
-3.03765595e-01 -5.11759054e-03 -3.64774019e-01 3.87883335e-01
-5.98602071e-02 -1.83836102e-01 -1.30102664e-01 1.90196201e-01
1.73919097e-01 3.57086897e-01 -9.53177661e-02 6.82928920e-01
1.26709849e-01 5.72409272e-01 -1.06371224e+00 -3.41643214e-01
1.03275269e-01 1.08205490e-01 -5.41755795e-01 2.73372173e-01
-3.86050552e-01 4.53013569e-01 -6.52527809e-01 3.63007858e-02
3.33532900e-01 2.48138696e-01 -2.08772104e-02 2.78720409e-01
-5.20207345e-01 -1.62684903e-01 -1.59945416e+00 1.13349473e+00
-7.65121579e-01 2.09828272e-01 1.48440629e-01 -1.23850715e+00
1.19246626e+00 3.62326384e-01 7.00604558e-01 -4.99581426e-01
2.78365403e-01 2.39742398e-01 8.39829445e-02 -6.41474068e-01
5.91169655e-01 -4.92773205e-01 -4.91124131e-02 2.85245597e-01
9.88058969e-02 -3.56040537e-01 3.13743234e-01 -4.53925759e-01
5.75630307e-01 3.63146216e-01 4.31093931e-01 -5.97851932e-01
6.90640271e-01 2.02657492e-03 5.09465337e-01 5.85213661e-01
-4.77389127e-01 8.96504964e-04 5.95207036e-01 -2.29946837e-01
-1.12070692e+00 -8.36698472e-01 -1.02010638e-01 1.03554428e+00
1.12002857e-01 4.32947129e-01 -6.75699770e-01 -5.99547088e-01
2.45849118e-01 1.13346279e+00 -6.05040610e-01 -1.39563933e-01
-5.61082780e-01 -5.07296503e-01 1.54485285e-01 2.48817682e-01
3.47609162e-01 -1.07327282e+00 -9.01176095e-01 4.66953814e-01
2.46378724e-02 -6.91120207e-01 -5.71384430e-01 3.79436374e-01
-7.30989218e-01 -9.85074520e-01 -8.61379325e-01 -6.27399147e-01
6.78165972e-01 -3.10673922e-01 6.81232452e-01 -2.90357739e-01
6.77077174e-02 7.18888164e-01 4.94932979e-02 -3.39001983e-01
-5.62015951e-01 -3.72591645e-01 3.14204752e-01 8.83564204e-02
-1.03056893e-01 -2.47521698e-01 -3.38921487e-01 4.71488833e-01
-6.85438037e-01 -4.51832354e-01 9.53070000e-02 9.55158949e-01
4.73255903e-01 2.16136068e-01 1.09673715e+00 -3.73277843e-01
1.09752309e+00 -4.44161981e-01 -1.06872523e+00 4.01768774e-01
-7.08203495e-01 5.49410284e-01 9.21242177e-01 -7.58885324e-01
-9.55019176e-01 -1.20389741e-02 1.86721273e-02 -2.81953931e-01
-5.60295247e-02 1.88790724e-01 7.90625624e-03 -3.09412271e-01
7.75907993e-01 -1.16805643e-01 3.53528768e-01 -5.75263761e-02
4.53931719e-01 5.67547023e-01 2.98141181e-01 -9.51635599e-01
5.75191081e-01 -8.92930478e-02 1.65290534e-01 -7.14720845e-01
-3.54753524e-01 -1.33626103e-01 -3.81045580e-01 -5.84563375e-01
7.32835352e-01 -3.98403049e-01 -1.15463722e+00 1.24679074e-01
-9.29010451e-01 -6.92734838e-01 -8.17869663e-01 4.94026095e-01
-1.26311243e+00 2.24939764e-01 5.87987974e-02 -1.22186029e+00
1.28243208e-01 -1.40935147e+00 4.56665933e-01 4.83448356e-01
-1.44088924e-01 -1.29670691e+00 8.09487700e-02 -1.07037984e-01
3.26617301e-01 4.13556784e-01 9.62758839e-01 -3.07056308e-01
-2.12692201e-01 -2.32283086e-01 4.30645287e-01 5.04865408e-01
3.58220190e-02 -1.23830689e-02 -5.88685930e-01 -4.42913651e-01
2.78917909e-01 -3.97429228e-01 1.81512803e-01 6.42508864e-01
7.54475653e-01 -4.12923783e-01 1.66788930e-03 1.59876287e-01
1.59306264e+00 7.69449055e-01 4.08007473e-01 7.93718278e-01
1.72945917e-01 6.08683407e-01 1.02444398e+00 6.11569583e-01
3.00994873e-01 7.47897983e-01 5.08756042e-01 2.86463648e-01
5.66816032e-01 1.30639849e-02 4.94307876e-01 -3.39785330e-02
-3.40577006e-01 2.19017789e-01 -6.55119359e-01 5.27752578e-01
-2.06128263e+00 -1.04953289e+00 2.58576542e-01 2.71476436e+00
8.54256392e-01 7.11545944e-02 4.99449164e-01 6.79517835e-02
7.79735208e-01 -1.14387527e-01 -8.62531185e-01 -1.05736673e+00
2.14757130e-01 1.99130192e-01 9.74307477e-01 1.04435360e+00
-8.86489391e-01 5.35552561e-01 7.18292952e+00 1.04771006e+00
-8.73060763e-01 -2.71767154e-02 3.88541877e-01 -5.24631143e-02
-2.65432656e-01 -7.99140781e-02 -8.18282068e-01 4.32593405e-01
9.70547915e-01 -5.69556355e-01 9.98835623e-01 8.53677154e-01
5.75618267e-01 -3.45053792e-01 -9.50263202e-01 6.27582848e-01
-5.08841038e-01 -9.16684449e-01 -4.76322889e-01 -1.16188325e-01
8.11164200e-01 -3.61131281e-01 1.67428404e-01 5.27537584e-01
5.01312494e-01 -9.74954903e-01 7.93331802e-01 6.43385947e-01
7.46805608e-01 -1.07765162e+00 3.75825256e-01 6.00473046e-01
-7.72436380e-01 -5.38069487e-01 -3.29721183e-01 -2.85267159e-02
-9.05015916e-02 -6.64270073e-02 -9.32813287e-01 3.35359186e-01
1.88744560e-01 1.44566134e-01 -6.04533590e-03 1.04233146e+00
-2.70790532e-02 2.27474608e-02 -3.04175496e-01 -5.79412580e-01
6.06685758e-01 -4.95075136e-01 6.31815612e-01 6.75042391e-01
5.32675982e-01 -2.63483196e-01 1.81382090e-01 7.89455473e-01
4.91534293e-01 1.74440026e-01 -5.42792559e-01 1.62795290e-01
2.74618536e-01 1.10709977e+00 -3.50881785e-01 -4.96909432e-02
-1.94463134e-01 5.67800224e-01 4.03072566e-01 6.93336487e-01
-1.07837152e+00 -6.44280910e-01 9.02803600e-01 -5.45586720e-02
4.04136896e-01 -2.63328373e-01 -1.90681085e-01 -4.41286057e-01
-9.58324522e-02 -7.94052601e-01 2.40503311e-01 -3.79729629e-01
-8.38880301e-01 3.33480418e-01 3.21064591e-01 -1.41191053e+00
-6.52718544e-01 -5.68467379e-01 -5.98309577e-01 1.19189847e+00
-1.36105108e+00 -2.88717061e-01 4.32796568e-01 6.25294745e-01
4.80438828e-01 -5.06374121e-01 7.00255215e-01 8.52737948e-02
-2.29958147e-01 4.15192753e-01 5.37416041e-01 -6.76800191e-01
4.12369817e-01 -1.45265174e+00 -2.20781565e-01 3.61587167e-01
-6.06824636e-01 7.09751919e-02 1.26675630e+00 -3.20353299e-01
-1.23101068e+00 -7.68926501e-01 5.08095503e-01 -2.81699598e-01
7.65661180e-01 2.55967081e-01 -6.42455041e-01 3.67556274e-01
1.33913979e-01 -2.29942814e-01 8.23436677e-03 -2.32935503e-01
4.55436110e-01 -2.16833562e-01 -1.54657447e+00 8.21952164e-01
2.55559057e-01 -8.43369886e-02 -3.75737309e-01 2.51360714e-01
3.47701252e-01 -4.51568544e-01 -8.35006356e-01 1.01752572e-01
4.11873132e-01 -6.96644008e-01 9.71319795e-01 -8.40589285e-01
1.09325603e-01 -3.36189270e-01 8.14574510e-02 -1.92443633e+00
-3.50560069e-01 -8.70171905e-01 6.94644004e-02 7.99606442e-01
3.78518671e-01 -8.47198367e-01 5.56095958e-01 7.41226733e-01
3.87567841e-02 -1.10231209e+00 -1.35599971e+00 -1.01309931e+00
3.21907401e-01 -9.65020433e-02 4.87000346e-01 5.98104000e-01
1.88384682e-01 6.44247010e-02 -3.85214984e-01 1.37853265e-01
7.07476556e-01 -1.46043956e-01 5.68768322e-01 -8.54552031e-01
-5.00552654e-01 -4.67098951e-01 -4.71982285e-02 -9.41244364e-01
1.97551623e-01 -6.07343018e-01 2.41880685e-01 -1.41770506e+00
-5.89209735e-01 -6.50403917e-01 -2.06839547e-01 2.48353079e-01
-7.29153082e-02 -6.20626688e-01 2.50427306e-01 -1.34507164e-01
-2.16361284e-01 6.40272439e-01 1.81569970e+00 8.48681405e-02
-6.05879843e-01 5.64809263e-01 -5.00603437e-01 6.71663702e-01
1.07812345e+00 -4.21002209e-01 -7.34640598e-01 3.79492752e-02
7.28086606e-02 5.88970840e-01 1.92969739e-01 -8.33160937e-01
9.47077852e-03 -8.33706796e-01 3.10260300e-02 -8.47316310e-02
3.74679089e-01 -9.54548895e-01 4.69784588e-02 4.02549148e-01
-5.32839000e-01 2.36004874e-01 1.91827074e-01 4.62076038e-01
4.00521308e-02 -9.81716454e-01 1.14932477e+00 -1.07398324e-01
-6.59629345e-01 7.75431320e-02 -6.83451712e-01 1.24882333e-01
1.36071384e+00 -2.60184854e-01 1.08141005e-01 -5.06109715e-01
-8.47952724e-01 6.97077513e-01 1.22660249e-01 2.47012854e-01
4.41180617e-01 -1.36594093e+00 -4.87792790e-01 1.99284572e-02
-2.99483180e-01 -3.58978182e-01 -1.51584834e-01 5.93529165e-01
-1.96793392e-01 9.79285091e-02 -5.73105276e-01 -1.83245257e-01
-9.41733062e-01 5.03142595e-01 8.57432306e-01 -2.80434430e-01
-2.30208829e-01 3.53187293e-01 -3.33242387e-01 -3.58186334e-01
3.21453094e-01 -4.00527477e-01 -4.92906779e-01 2.04711333e-01
3.02262276e-01 7.22356200e-01 -3.09109449e-01 -2.72014022e-01
3.23521458e-02 4.63551581e-01 2.41920233e-01 -5.35368919e-01
1.20398176e+00 -1.66526496e-01 3.17743689e-01 5.74697673e-01
8.50620210e-01 -4.37758327e-01 -1.66515291e+00 1.38923854e-01
2.07801480e-02 -4.35354590e-01 -7.99167454e-02 -7.14612365e-01
-6.22829735e-01 3.95629823e-01 6.85622931e-01 4.66761082e-01
7.91029334e-01 -5.66709161e-01 2.53010511e-01 4.64246899e-01
5.12466073e-01 -1.73506045e+00 -5.11228405e-02 6.27114415e-01
1.13098705e+00 -9.89717007e-01 -2.47708976e-01 3.25274706e-01
-1.08638144e+00 1.27068007e+00 5.18424213e-01 -5.14932275e-01
5.17062366e-01 1.30086824e-01 9.06858444e-02 2.48214602e-01
-4.64226991e-01 -2.08721802e-01 9.45108384e-02 7.95875669e-01
3.16316597e-02 1.86056957e-01 -6.75654948e-01 2.30453670e-01
1.01961158e-01 -1.10786012e-03 6.25604928e-01 8.02531898e-01
-7.65792787e-01 -1.27872479e+00 -7.05415487e-01 2.29537934e-01
-1.04841515e-01 4.56964344e-01 3.85165125e-01 8.16978395e-01
1.04957879e-01 8.25057626e-01 -1.19376101e-01 7.90320039e-02
6.48834884e-01 2.95863807e-01 7.21879303e-01 -2.96122849e-01
-4.77342576e-01 -1.27649829e-01 1.59396932e-01 -3.74933034e-01
-2.15677232e-01 -8.38003337e-01 -1.57206464e+00 -1.08940983e-02
-1.48893118e-01 5.55376589e-01 7.10143566e-01 9.98457313e-01
-2.07874745e-01 5.27098656e-01 1.00633705e+00 -1.01465011e+00
-1.35415256e+00 -5.78240216e-01 -7.57313728e-01 1.50548190e-01
5.63883364e-01 -9.89673793e-01 -3.91040236e-01 -3.30571562e-01] | [4.281239032745361, 2.186870574951172] |
c5ebe427-292e-4b84-9d62-463948bebcca | learning-uncertainty-with-artificial-neural-1 | 2206.06317 | null | https://arxiv.org/abs/2206.06317v1 | https://arxiv.org/pdf/2206.06317v1.pdf | Learning Uncertainty with Artificial Neural Networks for Improved Predictive Process Monitoring | The inability of artificial neural networks to assess the uncertainty of their predictions is an impediment to their widespread use. We distinguish two types of learnable uncertainty: model uncertainty due to a lack of training data and noise-induced observational uncertainty. Bayesian neural networks use solid mathematical foundations to learn the model uncertainties of their predictions. The observational uncertainty can be calculated by adding one layer to these networks and augmenting their loss functions. Our contribution is to apply these uncertainty concepts to predictive process monitoring tasks to train uncertainty-based models to predict the remaining time and outcomes. Our experiments show that uncertainty estimates allow more and less accurate predictions to be differentiated and confidence intervals to be constructed in both regression and classification tasks. These conclusions remain true even in early stages of running processes. Moreover, the deployed techniques are fast and produce more accurate predictions. The learned uncertainty could increase users' confidence in their process prediction systems, promote better cooperation between humans and these systems, and enable earlier implementations with smaller datasets. | ['Jochen De Weerdt', 'Hans Weytjens'] | 2022-06-13 | null | null | null | null | ['predictive-process-monitoring'] | ['time-series'] | [ 2.43568808e-01 4.48084354e-01 8.46141428e-02 -8.31385732e-01
-3.34487647e-01 -4.57331061e-01 7.91396797e-01 4.70311522e-01
-2.73117214e-01 9.25591350e-01 -2.96422895e-02 -5.60212016e-01
-5.98292887e-01 -1.00690103e+00 -6.29928470e-01 -2.44597182e-01
-2.27731764e-01 6.36264741e-01 1.45707756e-01 4.24857795e-01
3.92563015e-01 5.19608378e-01 -1.57871759e+00 8.40570107e-02
7.21818209e-01 1.37249851e+00 -1.09370813e-01 6.55250788e-01
-2.33902171e-01 1.03705537e+00 -6.43658280e-01 -2.35842332e-01
2.31672361e-01 1.98861375e-01 -3.60401303e-01 -5.66691041e-01
-3.40664876e-04 -4.09158736e-01 -1.12829402e-01 7.25986719e-01
1.49535701e-01 2.79470176e-01 8.82060170e-01 -1.36619818e+00
-3.32764596e-01 9.96271551e-01 -4.42890152e-02 4.87724841e-02
1.79174468e-01 3.32391769e-01 5.29291749e-01 -2.33130693e-01
-4.14872505e-02 1.30916083e+00 1.01659727e+00 3.44961196e-01
-1.27652764e+00 -7.07246900e-01 1.05253182e-01 -3.73090580e-02
-1.17398489e+00 -3.46450895e-01 2.93995619e-01 -7.46943474e-01
9.78291810e-01 2.73665428e-01 3.39636326e-01 1.02071202e+00
6.37458503e-01 3.34858358e-01 1.12568164e+00 -2.26698413e-01
7.76242495e-01 3.66521806e-01 -9.15232487e-03 2.62333959e-01
6.03518426e-01 6.30524755e-01 -5.62779665e-01 -1.18068643e-01
9.20237899e-01 2.23466352e-01 -1.93882138e-01 -2.23747626e-01
-1.03034270e+00 5.21103501e-01 2.69462883e-01 1.77943651e-02
-4.86204505e-01 5.25115252e-01 3.43056828e-01 2.42210239e-01
4.48300034e-01 9.13951337e-01 -6.83706462e-01 -5.86869836e-01
-8.76564324e-01 2.02736869e-01 1.32180166e+00 8.22112083e-01
4.61785525e-01 1.87216662e-02 -3.14276546e-01 3.95842880e-01
5.39153993e-01 3.56060147e-01 1.67029083e-01 -1.35315621e+00
1.52631968e-01 2.65554786e-01 5.46851158e-01 -8.00748706e-01
-4.38854873e-01 -2.38690421e-01 -6.50277019e-01 5.47513723e-01
6.66938782e-01 -3.99244964e-01 -1.00197601e+00 1.35795784e+00
-2.04726905e-01 2.89575875e-01 -8.64324048e-02 3.92226160e-01
3.00566349e-02 5.53421617e-01 3.80692780e-01 -4.94990405e-03
9.65634644e-01 -3.40457380e-01 -7.46714354e-01 -2.86521524e-01
1.94848418e-01 -3.65510553e-01 6.73563957e-01 6.45576775e-01
-8.99737597e-01 -6.15916431e-01 -1.07615054e+00 5.85048735e-01
-5.06358504e-01 -3.47986192e-01 9.39281881e-01 8.58923376e-01
-7.20975041e-01 1.50767517e+00 -1.25319922e+00 -8.89654532e-02
4.50910151e-01 3.81520212e-01 9.49646682e-02 1.27678230e-01
-1.24201167e+00 1.27208972e+00 7.83395410e-01 2.21951887e-01
-8.45211267e-01 -9.67569232e-01 -8.52674186e-01 2.86473691e-01
3.14373702e-01 -5.96467674e-01 1.64822042e+00 -6.56831861e-01
-1.54204190e+00 1.67901143e-02 3.85111064e-01 -7.30702758e-01
8.22632134e-01 -2.56755173e-01 -3.76736611e-01 -3.98487955e-01
-4.33617055e-01 4.31354821e-01 7.42779732e-01 -1.21775019e+00
-8.08452070e-01 -2.51224756e-01 -1.97374538e-01 -1.49521723e-01
1.47594199e-01 -2.29957685e-01 1.90933496e-01 -7.56726265e-02
7.54486695e-02 -6.46583259e-01 -4.39143270e-01 1.97350442e-01
-1.50826171e-01 -6.20298088e-02 6.08146608e-01 -6.16833448e-01
1.05180800e+00 -1.68600512e+00 -3.72305810e-01 5.00473619e-01
2.10745677e-01 -2.00172476e-02 3.07302564e-01 3.72167557e-01
1.22385308e-01 4.21714753e-01 -1.60453469e-01 -1.82110995e-01
4.10803765e-01 3.78454328e-01 -2.48571441e-01 1.70102455e-02
5.35050213e-01 4.92559314e-01 -8.57942522e-01 -1.49683371e-01
6.51411474e-01 4.14922059e-01 1.22005016e-01 3.67805183e-01
-4.37565356e-01 3.00540447e-01 -3.12927842e-01 4.35205817e-01
3.81600440e-01 -4.59559746e-02 9.03597772e-02 1.10769972e-01
-1.50291711e-01 3.00826102e-01 -1.28531790e+00 1.13586271e+00
-6.77616417e-01 7.49849141e-01 -2.05504209e-01 -5.07466316e-01
1.08575404e+00 1.53563708e-01 2.21239984e-01 -1.11088693e-01
2.85209239e-01 -5.13040982e-02 4.35094945e-02 -2.90127039e-01
4.07919824e-01 -2.23865360e-01 6.74424171e-02 4.33249563e-01
-7.66124502e-02 -6.35397732e-01 2.68407967e-02 -3.15703481e-01
1.08097088e+00 4.46049541e-01 2.20346719e-01 -2.50161916e-01
5.64765893e-02 -3.60836178e-01 4.98114526e-01 1.12625325e+00
-4.25998926e-01 2.92312890e-01 6.45363569e-01 -6.66113079e-01
-9.97705281e-01 -1.35323167e+00 -4.28181857e-01 9.17842090e-01
-2.61990458e-01 -1.32263824e-01 -3.51300031e-01 -5.43494225e-01
4.96930271e-01 1.33857119e+00 -6.77362740e-01 -3.93031359e-01
1.45186216e-01 -4.12147433e-01 4.48348701e-01 1.04644537e+00
1.45385116e-01 -8.15976381e-01 -7.18635142e-01 4.11138535e-01
3.49948585e-01 -6.12589121e-01 2.21863180e-01 7.10753441e-01
-1.02207386e+00 -9.41855907e-01 -2.03681931e-01 9.83910486e-02
2.37643450e-01 -5.91994941e-01 1.24211514e+00 -4.48436260e-01
-5.82380295e-02 5.12566209e-01 1.46517143e-01 -1.39550781e+00
-7.69979835e-01 -3.90647888e-01 2.93759108e-01 -7.57945776e-01
6.42900288e-01 -6.37862206e-01 -3.62609088e-01 4.08468515e-01
-6.61435366e-01 -3.13292980e-01 5.46083152e-01 4.98360693e-01
3.46165687e-01 4.06329006e-01 4.81293976e-01 -6.63584948e-01
1.08341479e+00 -4.40196782e-01 -8.90836000e-01 3.14730793e-01
-1.34482479e+00 5.02345860e-01 1.85623363e-01 -4.94291186e-01
-1.31714797e+00 7.00950669e-03 2.60067463e-01 -2.93727934e-01
-4.10193026e-01 6.58728600e-01 2.57288694e-01 2.82850146e-01
7.82898545e-01 -3.21504325e-01 8.77661407e-02 -1.26773685e-01
2.22139150e-01 6.39119446e-01 5.60993254e-01 -8.88151228e-01
4.26070273e-01 2.80909780e-02 1.17947288e-01 -4.38642412e-01
-6.81641638e-01 -1.79436445e-01 -4.83305871e-01 -4.05806482e-01
5.29597640e-01 -7.10038841e-01 -1.05048513e+00 1.48819029e-01
-9.64066327e-01 -3.07956934e-01 -6.75166488e-01 7.45729864e-01
-5.74852288e-01 1.46998212e-01 -6.73856616e-01 -1.56152749e+00
-4.08918738e-01 -8.85368407e-01 8.44425261e-01 4.54829574e-01
-6.28821611e-01 -1.09984410e+00 -6.07784875e-02 3.19607160e-03
7.46224463e-01 2.02152908e-01 7.01495290e-01 -9.35490251e-01
-4.85028476e-01 -5.73327184e-01 -3.07724297e-01 6.96807146e-01
5.41340653e-03 4.70083535e-01 -1.23019552e+00 1.07849725e-01
5.57653606e-02 -2.17136651e-01 7.12604284e-01 6.41623020e-01
1.34223533e+00 -3.12044378e-02 -3.80671650e-01 1.17001690e-01
1.15415907e+00 2.63845712e-01 6.65457428e-01 2.90979594e-01
2.24216402e-01 8.50172997e-01 3.84093851e-01 8.29319715e-01
1.01054706e-01 -8.44872221e-02 5.98254740e-01 5.95333099e-01
5.28813124e-01 -1.74361825e-01 2.98222393e-01 5.10487139e-01
-4.81347233e-01 -6.62777349e-02 -1.26863897e+00 7.64761344e-02
-1.95018470e+00 -1.06707561e+00 -4.28811572e-02 2.71405959e+00
8.51599693e-01 6.76140130e-01 -3.47287506e-01 2.75406018e-02
5.58794677e-01 -2.22857967e-01 -6.74381196e-01 -5.90618372e-01
5.79950154e-01 1.79996230e-02 8.42884779e-01 3.24117094e-01
-1.17388213e+00 3.18728983e-01 7.55662918e+00 4.49562997e-01
-7.85384893e-01 -3.49121958e-01 8.20242107e-01 -1.52237594e-01
-1.55714884e-01 3.31032090e-02 -6.04260147e-01 4.80747700e-01
1.80418599e+00 -2.32434794e-01 4.40818071e-01 1.09822321e+00
4.26097631e-01 -4.25997138e-01 -1.62859559e+00 6.89226508e-01
-5.88597298e-01 -1.27510941e+00 -3.25083762e-01 -5.98202413e-03
5.89910686e-01 9.22356024e-02 -1.55273959e-01 6.79593205e-01
1.08077157e+00 -1.42500222e+00 6.32280648e-01 1.25818741e+00
4.10522193e-01 -9.57648695e-01 1.17679536e+00 3.20572287e-01
-8.30413401e-01 -3.27726990e-01 -6.67785645e-01 -5.38483977e-01
-4.46146280e-02 9.71217155e-01 -1.30725348e+00 3.62287641e-01
7.07784832e-01 3.26156706e-01 -3.03860813e-01 1.14393735e+00
-1.68418780e-01 5.83645642e-01 -7.20832825e-01 -2.76376188e-01
-3.24940532e-02 -1.27504647e-01 1.27412722e-01 1.01158786e+00
4.80176628e-01 -4.77934152e-01 4.71672490e-02 1.38793635e+00
2.35729799e-01 -4.91806537e-01 -6.30691707e-01 -2.86423713e-01
7.30449378e-01 8.68086517e-01 -4.58316803e-01 -5.73961288e-02
-1.55316815e-01 4.16157126e-01 1.40459523e-01 2.05082789e-01
-6.56737745e-01 -3.72162253e-01 7.48113811e-01 -1.69512376e-01
-8.76187906e-02 -1.25438869e-01 -5.49410284e-01 -7.16695487e-01
-2.33084783e-01 -5.48406541e-01 2.44360849e-01 -9.26749587e-01
-1.70543623e+00 3.38984281e-01 2.51647383e-01 -8.84662688e-01
-8.00135255e-01 -1.01621342e+00 -6.42312527e-01 1.23563910e+00
-1.22061098e+00 -7.92737484e-01 -2.53247708e-01 -5.78440502e-02
1.76614106e-01 1.15719467e-01 1.00211680e+00 -3.61486286e-01
-4.79846179e-01 -1.14381537e-01 4.54461694e-01 -2.13094324e-01
7.58215845e-01 -1.69505954e+00 1.03467666e-01 5.14968276e-01
-3.57689530e-01 6.78682685e-01 1.09275413e+00 -7.64505804e-01
-9.54882145e-01 -1.01541984e+00 3.88445884e-01 -1.03637433e+00
9.88283753e-01 -9.78222396e-03 -8.79217982e-01 8.14631224e-01
-5.65727353e-02 -1.70237914e-01 7.87687957e-01 5.82709193e-01
-4.00483489e-01 -1.03143126e-01 -1.30677080e+00 2.86905259e-01
5.25564134e-01 -5.08199692e-01 -8.20005894e-01 -3.43471318e-02
7.21649170e-01 -2.63070256e-01 -1.43876863e+00 4.48078871e-01
9.55353618e-01 -8.93267035e-01 6.98055983e-01 -7.02098787e-01
5.16813874e-01 -5.13141006e-02 -1.33808777e-01 -1.42907155e+00
-8.82948935e-02 -3.60457212e-01 -3.64430338e-01 1.19980359e+00
6.53186202e-01 -6.62508130e-01 7.23918498e-01 1.71602261e+00
9.87116918e-02 -5.47061920e-01 -7.08022654e-01 -8.76360834e-01
2.60503858e-01 -1.15782285e+00 9.04912651e-01 8.36950660e-01
-1.35096088e-02 -1.32565856e-01 -1.40847027e-01 3.69387090e-01
6.29455626e-01 -2.99963921e-01 3.66024882e-01 -1.91014349e+00
-2.96341568e-01 -5.08861244e-01 -5.80792904e-01 -2.98174530e-01
-1.70557201e-01 -1.89521044e-01 4.01188523e-01 -1.48789942e+00
-1.36253536e-01 -5.76035798e-01 -5.79476416e-01 3.18331361e-01
1.25151873e-02 -5.01525462e-01 4.61088978e-02 3.89640182e-02
-2.80937552e-01 4.95351404e-01 7.52443910e-01 -9.82393175e-02
-2.31167346e-01 4.99659956e-01 -4.91109848e-01 9.90245879e-01
9.51923430e-01 -4.48677570e-01 -6.26467228e-01 -2.24310141e-02
3.78517151e-01 -5.64489737e-02 3.26669455e-01 -1.36963511e+00
2.23786995e-01 -4.26236391e-01 8.54689598e-01 -3.81138712e-01
2.01855391e-01 -1.18926179e+00 5.03361940e-01 4.03074563e-01
-6.46037340e-01 -2.71739870e-01 5.34550130e-01 1.01525939e+00
4.75641377e-02 -5.78292787e-01 3.75738651e-01 -3.05375040e-01
-5.42344868e-01 -2.84728426e-02 -5.26773334e-01 -5.10968328e-01
1.03128910e+00 -2.23810077e-01 -1.91288844e-01 -6.20564699e-01
-9.48642552e-01 3.93022776e-01 1.37706622e-01 2.73405701e-01
3.37554187e-01 -9.05551851e-01 -4.35932130e-01 -9.13399905e-02
1.88596353e-01 1.57473460e-01 4.94656302e-02 3.75030130e-01
-5.67487419e-01 4.20910150e-01 -1.69743940e-01 -5.96706688e-01
-6.12034142e-01 4.22035098e-01 5.99578619e-01 -4.41209078e-01
-8.76603723e-02 7.21698582e-01 -3.67902488e-01 -6.48843408e-01
7.09249973e-01 -8.61248016e-01 1.50179982e-01 -2.10960284e-01
7.61859477e-01 5.49988747e-01 -4.18133400e-02 3.54347706e-01
-3.36909257e-02 -2.94861197e-01 6.06829785e-02 -1.09979369e-01
1.29046690e+00 -1.16078556e-01 3.59011777e-02 1.01314199e+00
4.10261661e-01 -4.88058567e-01 -1.64860630e+00 -1.20130718e-01
4.33925629e-01 -3.97163689e-01 3.76982331e-01 -1.33206749e+00
-5.19059062e-01 9.91982281e-01 8.49849224e-01 4.68598157e-01
7.50340521e-01 -2.52726227e-01 7.98503961e-03 8.47830415e-01
3.61616611e-01 -1.29730594e+00 -4.54983532e-01 3.39906901e-01
8.27353418e-01 -1.44834840e+00 1.74878493e-01 -2.22731650e-01
-5.26868343e-01 1.41097808e+00 6.49271846e-01 4.03596669e-01
7.99513936e-01 5.26690006e-01 -5.69179840e-02 6.63733929e-02
-9.70419765e-01 1.20346457e-01 2.81147897e-01 1.01715100e+00
5.78663886e-01 3.11174601e-01 1.66883349e-01 8.86897206e-01
-1.94757730e-02 4.95434165e-01 5.34291327e-01 9.88438427e-01
-5.83401620e-01 -8.49108815e-01 -5.10477483e-01 1.04193723e+00
-3.27991247e-01 4.55120727e-02 -6.32135347e-02 7.04158247e-01
-1.15155689e-01 1.12584484e+00 2.70836741e-01 -4.28448737e-01
3.80058438e-01 4.06931818e-01 2.08855554e-01 -5.67407548e-01
-3.01463336e-01 -4.92866755e-01 4.18174088e-01 -6.26660705e-01
-8.20823312e-02 -6.94962978e-01 -1.00128627e+00 -5.19801915e-01
-3.82531732e-01 1.17936313e-01 1.08911383e+00 8.02017391e-01
3.68890166e-01 9.17955339e-01 1.31634220e-01 -9.20598090e-01
-1.42534387e+00 -1.26984692e+00 -6.86882794e-01 -3.99917476e-02
-2.20577165e-01 -6.49243534e-01 -4.11958039e-01 3.73305380e-02] | [7.367177963256836, 3.8318915367126465] |
4aaeb95e-afe0-4200-a5a1-be4e51c6e056 | evaluating-machine-translation-in-cross | null | null | https://aclanthology.org/2022.amta-research.25 | https://aclanthology.org/2022.amta-research.25.pdf | Evaluating Machine Translation in Cross-lingual E-Commerce Search | Multilingual query localization is integral to modern e-commerce. While machine translation is widely used to translate e-commerce queries, evaluation of query translation in the context of the down-stream search task is overlooked. This study proposes a search ranking-based evaluation framework with an edit-distance based search metric to evaluate machine translation impact on cross-lingual information retrieval for e-commerce search query translation, The framework demonstrate evaluation of machine translation for e-commerce search at scale and the proposed metric is strongly associated with traditional machine translation and traditional search relevance-based metrics. | ['Amita Misra', 'Liling Tan', 'Hang Zhang'] | null | null | null | null | amta-2022-9 | ['cross-lingual-information-retrieval'] | ['natural-language-processing'] | [-1.23391345e-01 -6.32559180e-01 -7.99946785e-01 -1.54394269e-01
-1.82236767e+00 -1.21229994e+00 7.35781550e-01 3.84158581e-01
-8.80029559e-01 4.03862268e-01 1.78275228e-01 -6.62636399e-01
-3.89278978e-01 -6.89463854e-01 -5.27184784e-01 6.31651357e-02
3.87573749e-01 9.92952943e-01 3.19109946e-01 -7.16959357e-01
4.64692026e-01 1.97238669e-01 -6.54054523e-01 4.34379637e-01
1.00815856e+00 9.75396514e-01 1.56520262e-01 4.27192092e-01
-2.60089070e-01 -7.33842701e-02 -3.20358276e-01 -1.10701573e+00
2.76678622e-01 -5.97660363e-01 -1.09768021e+00 -8.47422481e-01
4.53845978e-01 1.31144419e-01 -4.42916639e-02 1.24765646e+00
8.47343266e-01 -2.56031364e-01 4.32379633e-01 -1.04959261e+00
-1.22306359e+00 4.97250289e-01 2.78971624e-02 6.23500526e-01
6.02930248e-01 -3.38999301e-01 1.42273533e+00 -1.21326661e+00
1.12808704e+00 1.01850188e+00 5.36980987e-01 -1.57703727e-01
-8.79800200e-01 -2.90321708e-01 -4.62107092e-01 2.33614102e-01
-1.42430079e+00 -4.66548353e-02 5.15476584e-01 -8.04440007e-02
1.20964909e+00 5.78040361e-01 2.60220408e-01 8.55990887e-01
5.78056335e-01 5.17169774e-01 9.91654694e-01 -6.65525734e-01
-1.65797919e-01 5.77446163e-01 -7.00851828e-02 4.76630896e-01
1.76182881e-01 8.39117467e-02 -6.89336836e-01 -3.74026954e-01
5.43891490e-01 -7.60827959e-02 3.20692182e-01 -7.53601938e-02
-1.35642862e+00 1.01607525e+00 2.86126733e-01 4.17149693e-01
-6.06181085e-01 -1.87652752e-01 7.34712541e-01 1.04398417e+00
5.91030598e-01 7.62009263e-01 -8.66060495e-01 -2.31805161e-01
-6.67419374e-01 2.21657231e-01 9.88464713e-01 1.36467147e+00
5.32706678e-01 -7.46606588e-01 -5.37693381e-01 9.80018020e-01
3.31252664e-01 1.02989912e+00 5.59610128e-01 -6.87356174e-01
6.87495768e-01 7.25736260e-01 -1.07939355e-02 -7.31543541e-01
2.40094990e-01 -6.96312428e-01 -2.09263250e-01 -5.27916610e-01
1.22434366e-02 2.07112551e-01 -7.61107430e-02 1.17646706e+00
1.27410516e-01 -8.47169578e-01 9.00754184e-02 1.06778789e+00
4.52016026e-01 4.19267327e-01 9.79429334e-02 -4.86433655e-01
1.45072711e+00 -1.36974776e+00 -7.13832200e-01 1.06825493e-01
8.98190022e-01 -1.84118629e+00 1.28457427e+00 -1.39832273e-01
-9.97578204e-01 -2.87234038e-01 -6.06332600e-01 -3.94452572e-01
-8.72643530e-01 -1.23305716e-01 6.10139906e-01 6.77714288e-01
-1.02503145e+00 -3.55916508e-02 -1.92498982e-01 -9.10756648e-01
-4.39134538e-01 1.85295135e-01 5.13229780e-02 -5.22590280e-02
-1.58749068e+00 1.18171251e+00 -1.16812527e-01 -1.64736271e-01
-3.35037619e-01 -2.76343644e-01 -1.03945524e-01 -1.43110752e-01
2.82112241e-01 -8.31768811e-01 1.38735425e+00 -5.52332819e-01
-1.32806981e+00 7.96887100e-01 -2.55832255e-01 -3.86163369e-02
6.42718911e-01 -1.92368254e-01 -8.82785559e-01 1.33573487e-01
5.89300275e-01 2.10714906e-01 3.32074881e-01 -7.63921440e-01
-8.78543675e-01 -3.73339385e-01 -4.65344638e-02 5.20221293e-01
-4.52801958e-02 9.11164105e-01 -8.21756959e-01 -7.38955438e-01
3.06679904e-01 -7.92122722e-01 1.47283420e-01 -3.84833008e-01
1.23538047e-01 -4.15549308e-01 6.65162325e-01 -8.88509452e-01
1.53616726e+00 -1.56879079e+00 -2.99049854e-01 2.21034139e-01
-4.14236516e-01 -2.83913910e-02 -3.50868195e-01 1.02940440e+00
6.71379507e-01 6.23071671e-01 4.02728796e-01 3.28535408e-01
5.17065585e-01 -1.84121087e-01 -3.49242777e-01 -7.66284242e-02
-3.32080603e-01 1.67190218e+00 -8.04714024e-01 -9.72891927e-01
-4.09733385e-01 1.18313782e-01 -3.77952963e-01 -1.24462299e-01
-1.15821868e-01 -2.21868549e-02 -9.27118540e-01 1.24526787e+00
1.57679141e-01 -1.22106485e-01 9.25547332e-02 -2.14482024e-01
-1.17467962e-01 5.65074801e-01 -3.31143796e-01 1.98073757e+00
-6.60560668e-01 1.59116939e-01 9.63310972e-02 -2.52871096e-01
6.20616257e-01 4.00044471e-01 5.45150220e-01 -1.75842977e+00
-1.13706611e-01 8.55190396e-01 -4.22574431e-01 -5.74114203e-01
9.27028537e-01 3.77367854e-01 -2.64013350e-01 8.35939467e-01
-1.18529461e-01 -7.16525018e-02 1.81525186e-01 1.73197649e-02
8.14940333e-01 2.65567154e-01 1.48293674e-01 -3.77284020e-01
4.62422580e-01 6.65894091e-01 3.88909802e-02 6.89564705e-01
-1.85807824e-01 2.92295702e-02 -2.75376469e-01 -3.62831593e-01
-1.14552402e+00 -8.38427842e-01 -1.75352976e-01 1.86962354e+00
5.95474422e-01 -3.23987186e-01 -6.56658828e-01 -1.07285452e+00
-1.39801316e-02 4.84698832e-01 -7.82444999e-02 -9.61350873e-02
-6.17471039e-01 -6.03014231e-01 6.38447523e-01 -8.62969607e-02
3.79512280e-01 -8.76338780e-01 -2.28974760e-01 3.45569104e-01
-6.31658912e-01 -9.97652411e-01 -1.55122149e+00 1.23629577e-01
-1.20698071e+00 -6.69790030e-01 -6.91850364e-01 -1.37349439e+00
1.91572800e-01 4.60331947e-01 1.52770841e+00 -1.62293002e-01
-4.74284701e-02 6.28330946e-01 -7.43114293e-01 -5.41785061e-02
-5.48261285e-01 6.10324383e-01 -2.09766999e-01 -5.26154935e-01
1.09772992e+00 -1.15490831e-01 -9.50948834e-01 9.08775806e-01
-9.64105487e-01 -7.14282095e-01 1.03088522e+00 8.44025671e-01
8.31982315e-01 -3.33093464e-01 7.25050628e-01 -4.31722701e-01
1.60650790e+00 -4.38346446e-01 -4.63674963e-01 1.11774755e+00
-1.80284691e+00 1.43613994e-01 8.92409403e-03 -5.55280089e-01
-7.23332703e-01 -6.53732181e-01 2.04053059e-01 5.91573864e-02
4.62361336e-01 9.63470876e-01 2.92720824e-01 -3.06962132e-01
6.90911174e-01 4.45308506e-01 -4.51990485e-01 -8.32624912e-01
4.49242800e-01 7.59904385e-01 2.47529638e-03 -6.49530172e-01
2.99894094e-01 -2.69383222e-01 -2.85011292e-01 -1.58663332e-01
1.47862181e-01 -1.01916111e+00 -6.28515780e-01 -4.58658449e-02
7.66052604e-01 -5.90968609e-01 -5.39356172e-01 -2.62934238e-01
-1.32888913e+00 3.96369070e-01 1.21392883e-01 8.04016352e-01
-2.85613745e-01 1.52989849e-01 -8.52610886e-01 -4.55025226e-01
-8.69540215e-01 -1.14941990e+00 1.49800014e+00 -4.41786438e-01
-1.45277768e-01 -1.00637031e+00 5.79414964e-01 8.73085737e-01
7.40891814e-01 -7.29168355e-01 1.18631804e+00 -9.48917031e-01
-6.43101990e-01 -5.51984251e-01 -2.79480785e-01 7.16535971e-02
4.02189605e-02 -5.86468756e-01 -1.87235042e-01 -1.67759061e-01
-9.85555798e-02 -1.67308643e-01 2.55674839e-01 -1.70828924e-01
-3.99276689e-02 -6.78834617e-01 -1.85160950e-01 1.10505395e-01
1.72505069e+00 5.57846606e-01 2.72414029e-01 7.26954877e-01
1.06071271e-01 5.60581505e-01 9.62918043e-01 -2.11760402e-01
6.18274450e-01 1.31413102e+00 -2.32964322e-01 -1.19424500e-02
1.19369384e-02 -7.21382737e-01 4.01047528e-01 1.44859529e+00
1.14771321e-01 -5.78878112e-02 -7.66275048e-01 1.59335598e-01
-1.71318412e+00 -6.61471307e-01 1.54533938e-01 2.18125057e+00
1.00790954e+00 -1.57157406e-01 -3.65942679e-02 -7.15378046e-01
6.50193572e-01 -4.12819743e-01 -5.09817839e-01 -7.01528370e-01
-2.03689381e-01 2.99741983e-01 8.79855096e-01 7.12170899e-01
-6.44674957e-01 1.26405334e+00 7.31447554e+00 1.08784819e+00
-1.23536706e+00 8.34466696e-01 3.01672906e-01 2.66281009e-01
-6.84078574e-01 -1.02346456e-02 -7.22845614e-01 2.94005305e-01
7.20614135e-01 -5.77052355e-01 6.70389473e-01 9.04970407e-01
1.64501622e-01 2.52215296e-01 -1.14268386e+00 1.00223577e+00
1.00089230e-01 -9.25552130e-01 3.75687867e-01 2.79368222e-01
7.74898112e-01 2.07845807e-01 3.26253802e-01 6.17543042e-01
1.01197585e-01 -5.47860801e-01 7.17570186e-01 1.65119275e-01
9.04159307e-01 -3.28960031e-01 9.40786302e-01 2.45765746e-01
-1.08862114e+00 3.66001815e-01 -1.15677163e-01 5.21470845e-01
1.17537208e-01 6.26137480e-02 -9.23525035e-01 7.63049185e-01
4.99310702e-01 -1.24823578e-01 -6.06220543e-01 7.09810495e-01
2.91802317e-01 2.87719041e-01 -2.77629226e-01 -5.20610392e-01
4.26784426e-01 -7.20725417e-01 5.22161126e-01 1.37101817e+00
8.39273930e-01 -5.26041329e-01 7.95153677e-02 6.56164408e-01
-3.68123770e-01 1.06679881e+00 -5.38785398e-01 -3.14142734e-01
5.59190750e-01 1.03625119e+00 -6.13819361e-01 -2.03137144e-01
-4.87001687e-01 1.50012100e+00 -2.87496150e-01 6.05320990e-01
-6.31988049e-01 -2.59828299e-01 2.91210353e-01 -5.57242781e-02
1.16261262e-02 -1.73202634e-01 -1.49384245e-01 -1.03428507e+00
6.13858283e-01 -1.13870978e+00 3.88667434e-01 -5.07762790e-01
-1.31624413e+00 7.27367818e-01 -1.12400025e-01 -1.48915112e+00
-5.49503624e-01 -1.55828267e-01 5.16951755e-02 1.13004518e+00
-1.40545213e+00 -1.43767571e+00 4.86263484e-01 4.56055015e-01
6.40800595e-01 -3.73826146e-01 1.03573632e+00 9.51631844e-01
1.10417441e-01 8.58086228e-01 5.33453345e-01 -1.17366277e-01
1.08471835e+00 -8.28623891e-01 3.17483723e-01 6.04117334e-01
4.68325794e-01 1.19296074e+00 3.78776431e-01 -8.30604315e-01
-1.91181922e+00 -7.85912573e-01 1.88363063e+00 -7.48423576e-01
7.22149432e-01 -3.60389352e-02 -2.28392929e-01 2.00205758e-01
4.84414488e-01 -5.76782703e-01 6.11822784e-01 3.05257857e-01
-6.72339976e-01 -5.38865089e-01 -1.08719885e+00 5.91227233e-01
1.10889208e+00 -1.23828459e+00 -6.47658169e-01 5.57852268e-01
1.02653253e+00 2.19043434e-01 -1.16288507e+00 4.08479840e-01
9.27904069e-01 -1.98753938e-01 1.21663022e+00 -4.88859445e-01
-1.88840125e-02 -5.91322519e-02 -4.43810403e-01 -9.03381586e-01
-2.31797844e-01 -5.68671465e-01 2.99677312e-01 9.11883950e-01
1.11437213e+00 -7.49707162e-01 3.33973348e-01 4.07016575e-01
2.41177917e-01 -4.78595704e-01 -1.22269976e+00 -9.95501697e-01
3.05227846e-01 1.38955250e-01 6.73200369e-01 1.01225483e+00
1.30411536e-01 8.59655678e-01 5.30599505e-02 -1.91446483e-01
3.72011632e-01 3.38392973e-01 2.91548967e-01 -7.58012176e-01
-2.20508903e-01 -9.96487558e-01 4.59010489e-02 -1.05395985e+00
-2.85042644e-01 -1.18787253e+00 -2.75600702e-01 -1.46948552e+00
3.30778062e-01 -4.07808155e-01 -5.81299901e-01 -9.52280313e-02
4.44388390e-03 3.65687937e-01 -7.71531835e-02 9.31204200e-01
-1.00275505e+00 2.88908869e-01 1.46216667e+00 -4.38969135e-01
1.75418407e-02 2.40103137e-02 -6.01175725e-01 -2.40179211e-01
3.92556518e-01 -6.40954733e-01 -5.19188106e-01 -7.19183743e-01
8.97865772e-01 2.03951001e-01 -7.27212653e-02 -2.12786332e-01
6.50866687e-01 -2.49844357e-01 -3.89543205e-01 -3.32746089e-01
-2.97972679e-01 -1.07639515e+00 9.72910598e-02 3.24574620e-01
-8.93134475e-01 1.02552474e+00 -5.11459291e-01 2.29607120e-01
-8.37380171e-01 -2.15103149e-01 2.56838481e-04 -2.14996234e-01
-5.09629965e-01 2.13428363e-01 4.90006730e-02 2.16171429e-01
5.29275239e-01 -4.92001362e-02 -2.28808731e-01 -3.82369339e-01
-2.63579577e-01 -8.65333155e-02 2.96285719e-01 7.73654759e-01
2.49633789e-01 -1.52482998e+00 -8.07754934e-01 -2.17840001e-01
5.53622961e-01 -9.84127402e-01 -6.16400659e-01 8.79590631e-01
-5.40066898e-01 1.28337181e+00 4.48815711e-02 -3.13424051e-01
-1.15802622e+00 6.28536999e-01 -5.31280562e-02 -7.60627151e-01
2.93044448e-01 5.92533171e-01 -3.27810735e-01 -7.62107074e-01
1.82564050e-01 -1.76248074e-01 -2.16461439e-02 -1.00870803e-01
1.02798857e-01 4.45095330e-01 5.13365090e-01 -6.83868289e-01
-2.79425800e-01 6.97008729e-01 1.41446944e-02 -7.72291660e-01
5.84483802e-01 -6.38809085e-01 -5.97187102e-01 1.56128824e-01
1.59739637e+00 2.90253222e-01 1.91941231e-01 -7.73557603e-01
6.64084017e-01 -4.00289744e-01 6.11254349e-02 -1.56385684e+00
-5.26921570e-01 3.99020642e-01 9.47189212e-01 1.25278011e-01
1.14406896e+00 3.51599753e-02 9.81942773e-01 9.99971747e-01
1.09502304e+00 -1.64891267e+00 -1.31678477e-01 6.14086926e-01
9.78371918e-01 -1.51067519e+00 -3.46341670e-01 -2.18145519e-01
-5.07800817e-01 9.33951974e-01 1.43442690e-01 7.86014915e-01
5.45936048e-01 -8.92847180e-02 5.78630090e-01 -3.17200869e-01
-6.09560430e-01 -1.10842824e-01 6.49154902e-01 5.87393269e-02
6.84275925e-01 4.33048308e-02 -1.28532600e+00 2.56928027e-01
-3.97754535e-02 1.84641048e-01 -7.31935441e-01 9.06959414e-01
-2.31055245e-01 -1.71494460e+00 -1.71556950e-01 1.61882460e-01
-8.42156231e-01 -7.89198518e-01 -5.79387844e-01 3.24609250e-01
-6.36613369e-02 1.11866200e+00 -5.23223460e-01 -5.86474597e-01
3.51056576e-01 5.26933730e-01 2.46941701e-01 -2.90799379e-01
-1.12282670e+00 3.73745859e-01 2.68935263e-01 -5.55350006e-01
-5.12679219e-01 -3.91197741e-01 -4.04577523e-01 -1.61026433e-01
-5.28740346e-01 1.02535224e+00 1.31251991e+00 4.16086346e-01
6.63574100e-01 -2.52288580e-01 6.32097781e-01 2.63165593e-01
-8.89819145e-01 -1.23506415e+00 -2.93538958e-01 4.45266336e-01
-2.46868268e-01 -1.29917160e-01 -1.18674695e-01 6.70960397e-02] | [11.542150497436523, 9.997648239135742] |
b58e62fb-2061-4579-b268-0489d4701001 | metatroll-few-shot-detection-of-state | 2303.07354 | null | https://arxiv.org/abs/2303.07354v1 | https://arxiv.org/pdf/2303.07354v1.pdf | MetaTroll: Few-shot Detection of State-Sponsored Trolls with Transformer Adapters | State-sponsored trolls are the main actors of influence campaigns on social media and automatic troll detection is important to combat misinformation at scale. Existing troll detection models are developed based on training data for known campaigns (e.g.\ the influence campaign by Russia's Internet Research Agency on the 2016 US Election), and they fall short when dealing with {\em novel} campaigns with new targets. We propose MetaTroll, a text-based troll detection model based on the meta-learning framework that enables high portability and parameter-efficient adaptation to new campaigns using only a handful of labelled samples for few-shot transfer. We introduce \textit{campaign-specific} transformer adapters to MetaTroll to ``memorise'' campaign-specific knowledge so as to tackle catastrophic forgetting, where a model ``forgets'' how to detect trolls from older campaigns due to continual adaptation. Our experiments demonstrate that MetaTroll substantially outperforms baselines and state-of-the-art few-shot text classification models. Lastly, we explore simple approaches to extend MetaTroll to multilingual and multimodal detection. Source code for MetaTroll is available at: https://github.com/ltian678/metatroll-code.git. | ['Jey Han Lau', 'Xiuzhen Zhang', 'Lin Tian'] | 2023-03-13 | null | null | null | null | ['misinformation', 'few-shot-text-classification'] | ['miscellaneous', 'natural-language-processing'] | [ 0.31964913 -0.19871451 -0.5911734 -0.09994107 -1.184722 -0.54283154
1.212251 0.3043932 -0.7719142 0.69065994 0.5167142 -0.4719228
0.10542185 -0.79632306 -0.72003406 -0.32002202 0.12656482 0.6725086
0.24392481 -0.86082995 0.2090474 -0.03945249 -1.2206322 0.5367905
0.60013515 0.27150384 -0.06080003 1.1897312 -0.12154001 0.8713749
-0.89763063 -0.93236375 0.09212755 -0.39435515 -0.48798352 -0.50081205
0.7808009 -0.10781807 -0.70851874 0.9667735 0.8023458 -0.02113381
0.5869734 -0.8244054 -0.1105408 0.75232077 -0.47152227 0.93498915
0.281172 0.4322124 0.7183843 -0.69319624 1.1666329 1.4427263
1.1793983 0.7253751 -1.1636269 -1.1350225 0.13655055 0.2987069
-0.8244214 -0.7535666 0.50395757 -0.5221506 0.9376768 0.40915573
0.5904136 1.9830859 0.8548983 0.8643312 0.9980695 -0.2014733
-0.05473239 0.11001594 0.4386078 0.59401274 0.32101405 0.27364042
-0.7629986 -0.5861485 0.12940127 0.09091197 0.43527725 0.20482306
-1.0041491 1.3241343 0.14808834 0.4496976 -0.02295196 0.29000208
0.9680564 0.61619824 1.0101213 0.46460676 -0.3329461 -0.40910295
-0.8072785 0.26559195 0.61842626 0.72953975 0.62767285 0.02621431
-0.47137496 0.6132459 -0.192536 0.98378307 0.43047282 -0.34209034
0.65581834 0.44962925 -0.08792977 -0.8627136 -0.70025736 -0.9882008
-0.5640798 -0.26839375 0.16598801 -0.42325568 -1.209261 1.4268649
0.3690173 0.23530836 -0.18208559 0.24503815 0.86317873 0.48025894
0.35081503 -0.08384785 1.4707785 -0.8020569 -0.59577954 -0.90627545
0.8972628 -0.9172986 0.7736412 0.05967417 -0.3101254 -0.26777592
-0.80707806 0.3469938 -0.6805991 -0.30493176 0.80178744 0.86227846
-0.42361945 0.33683366 -0.569146 -0.8050092 0.3138651 -0.12827663
-0.07206875 -0.15379763 -1.6738838 1.21845 0.25800526 -0.13523683
-1.2960867 -0.68024355 -0.98354834 -0.6577643 0.5409417 -0.5573386
1.4506564 -0.66123164 -1.0419832 0.6769141 -0.04630795 -0.7145857
0.5688193 -0.30135533 -0.8707832 -0.20945181 0.41781878 0.19672582
1.2705101 -1.0151416 -0.880751 -0.01623158 -0.03103392 0.06846441
-0.20636794 0.1660445 -0.06339954 -0.6682161 -0.65476656 -0.95115864
-0.14411475 -0.9883939 -0.26197058 -0.09637871 0.6983184 -0.84993666
1.5629566 -1.7582656 0.15642789 -0.13164896 0.1837138 0.6194164
-0.4471302 1.0251844 0.32573995 0.09757484 0.07859123 -0.37042886
-0.10077658 0.10261498 -0.53670603 0.64755106 -0.23799479 1.0726392
-1.324578 -0.3042666 0.30466512 0.27067697 -0.3065853 -0.4702481
-0.292549 0.6416282 -0.4328861 0.7509726 0.31849056 0.1843679
0.1087539 0.07933809 -0.17917582 0.14405823 -0.23485352 1.6045904
-0.5956281 0.8281774 -0.04350347 -0.31741044 0.53979015 -0.05668923
0.04270079 -0.8443732 0.80660385 0.02301739 -0.14996871 -0.39251435
0.8488781 -0.31662038 -1.0221221 0.4428421 0.33734798 -0.06815004
0.556958 0.62714434 1.4738808 -0.37601537 0.27743435 0.30596092
0.2527836 0.42971197 0.5118839 1.5153344 -0.1942179 -0.24453318
0.12385688 -0.68713135 -1.135108 -0.6407144 -0.17009963 1.8338766
-0.22243875 -0.68341434 -0.41417745 -0.9685752 -0.05284219 1.3141356
-0.88855624 -0.25323907 -0.591467 -0.8815832 1.0133799 -0.05276843
0.4157766 -0.7266736 -0.4214697 0.3473814 -0.42232957 -0.7107719
-0.44626033 -0.0883477 -0.6557403 -1.2833576 -0.51805174 -0.25304845
0.24839844 0.229748 0.9618086 -0.03166376 -0.43758392 0.58768755
-0.4439509 -0.6482952 -0.86195534 0.5982497 0.27060404 -0.11835213
0.5026981 -0.27384007 -0.1530173 0.23885457 -0.8554715 -0.01634545
0.8248619 0.87074155 -0.1029772 -0.26798636 0.34607506 -1.2773372
0.9277284 -0.5939698 -0.24644858 0.19325614 -0.15964505 -0.16624919
0.09790765 -0.767716 -1.0418739 -0.33934715 -0.25774342 -0.08519616
0.02951499 0.44141957 0.60633266 -0.10914516 1.1265146 0.46010503
-0.20584571 -0.5026235 0.5603175 0.8467759 0.25034943 -0.09835966
1.2211255 0.5270972 -0.35817668 -0.9811747 -1.3723364 -0.72392726
-0.38671032 -0.64133495 0.60346526 -0.8909166 -0.62320554 0.5410601
-1.100671 -0.51539195 -0.16298677 0.43769613 -0.24084972 0.29974645
-0.72921383 -0.81424075 -0.6129809 -0.5871257 1.2082685 -0.15351321
-0.26543543 -1.1410422 0.7372202 0.5054785 0.62919414 0.30399436
0.6566911 -0.8459762 0.054334 -0.75797415 0.09239072 0.00874923
-0.28566968 -0.7845713 -0.66743815 -0.76430887 -0.1357959 -0.46759197
1.3515297 0.06083919 0.05176255 -0.60502726 -0.6498495 0.2183533
0.9557237 -0.27165246 0.47432125 0.58977485 0.3615563 0.18165456
0.771392 0.3764191 0.33831233 0.58366454 0.21785533 0.25109577
-0.43847692 -0.681708 0.73018706 1.0051827 -0.08044706 -0.11153045
-1.0484719 0.4393126 -1.9109365 -1.413394 -0.16626205 2.0609791
0.82579666 0.5033302 0.01261572 -0.31771767 0.8780343 0.42836463
-0.41311195 -0.45043796 0.01522669 -0.06911715 1.1202105 0.92808044
-1.4004542 1.3292025 6.10168 1.5888726 -0.9028115 1.0559942
0.01272052 -0.33615896 -0.02701983 0.09248589 -1.4848834 0.27503031
1.018113 -0.2531506 0.23358937 0.5167605 -0.19829597 -0.14021042
-0.39615023 0.64884555 0.685745 -1.4046898 0.09757096 0.07811047
0.7272344 0.49172714 0.13145825 1.1972442 0.69394505 -0.6599025
0.7403474 0.66205364 0.8084001 -0.67495435 0.8122672 0.7094949
-1.0027455 -0.34284776 -0.37800294 -0.08153962 0.27430755 0.45224607
-1.2073941 0.55251634 0.3206983 0.7766036 -1.1376388 0.91515046
-0.46981663 0.85209084 -0.17176858 -0.21523827 0.4087664 0.62511617
1.2694508 1.3359438 0.14144243 -0.420496 0.28876653 -0.17718615
0.06853216 0.02957432 -0.7680278 0.01152427 0.29440153 1.0430136
-0.39173082 -0.4274784 0.03800356 0.8498805 0.2994112 0.16187516
-1.0346022 -0.43237382 0.25003 0.24852562 0.02301414 -0.20801914
0.21845661 -1.2286363 -0.7035664 -0.9670709 0.77775437 -0.17238678
-1.3531499 0.7166805 0.27548817 -1.2500745 -0.21654022 -0.19518022
-0.5046904 0.06467929 -1.3323061 -1.7287732 0.17505346 0.38064888
1.085654 -0.35744005 0.63448864 0.38546365 -0.3693984 0.6169776
0.09449147 0.07999092 0.9054886 -0.53366727 0.54856366 0.5870819
-0.20497453 0.44130775 1.2245488 -1.4445667 -1.5436144 -1.5095798
1.3302234 -1.1372004 1.2506533 -0.8318672 -0.1747216 1.0571986
0.280849 -0.792397 0.6799157 0.56729585 -0.49196827 0.14215505
-0.88276577 0.6658633 1.138352 -0.39975476 -0.991817 0.96144295
0.4944693 -0.3286442 -0.4259987 0.23482911 0.52580976 -0.54551375
0.84908646 -0.7181931 0.0635389 0.12139744 0.22586888 -1.3690534
-0.3947863 -0.8237558 -0.03572747 0.88315284 0.49061903 -0.69274735
0.24305372 -0.27993417 -0.11036406 0.05197232 -1.6048641 -1.0060308
0.02634591 -0.860099 0.07738764 0.9413991 -0.05803391 0.7598357
-1.2122492 -0.13009246 1.013209 -0.3289091 1.2398565 -1.222843
-0.27798182 -0.1943578 -0.37204176 -0.61295855 0.1338913 -1.1716987
-0.1047831 -1.2894301 0.46898457 -0.22752708 0.09533743 0.7360883
0.10082636 0.64540875 0.3141728 0.29129365 -1.145162 0.52405035
1.0663834 -0.71653533 -0.3006154 0.20271938 -0.4234417 0.70817506
0.864485 -1.1108832 0.27226105 -0.31199288 0.71688265 -0.112791
0.3262512 -1.0997274 0.3887893 -0.00626479 0.21601933 -0.61084265
0.48378327 -0.29511088 0.27243912 0.9754644 -0.14803039 -0.26624936
0.4906491 1.0507693 0.3103122 -0.28197598 0.4151572 -0.24595146
-0.6389594 0.08022945 -1.0518768 0.02291052 0.8703269 0.2268821
-0.9853531 -0.52411425 -0.7479905 0.26664707 0.2635244 0.870281
0.2679091 -1.137477 -1.0431377 -0.17455566 0.4699457 -1.0929042
0.4864842 1.0737058 -0.11642481 0.57074875 0.07066205 -0.25765088
-1.4835205 0.4340967 0.08820851 -0.8302902 -0.53344613 0.575772
-0.26244974 -0.81532204 0.12332243 0.2503781 -0.16149314 0.37498215
0.8933976 0.4416762 -0.09936661 -0.9019587 -0.40846825 0.40806493
-0.40176088 -0.14211094 1.2370062 -0.13396876 0.25171685 0.6046545
0.61748236 0.40259957 -0.7750669 -0.43104616 -0.18596663 -0.46904874
0.07625308 -0.9951051 -0.36571267 0.62673026 0.65184337 -0.13151155
0.412669 0.11986402 0.90585726 0.6282316 0.5621035 -1.0659184
0.09203357 1.2031324 0.83229536 -1.070723 0.3548914 0.05590308
-0.6144719 0.83573955 0.07028896 0.18664348 0.3226064 -0.13031317
-0.08419171 -0.4648865 -0.8294187 -0.5134719 0.22596972 0.53862214
-0.05779997 0.08583566 -0.5730907 0.3670426 -0.15234458 -0.12803684
0.34455052 1.1780417 -0.8500208 -1.261012 -0.5022397 0.70654136
-0.24895175 -0.35049096 -0.6695056 0.72664386 0.2931353 1.0656664
-0.23832293 -0.77234983 0.33731103 0.21426323 0.42262802 -0.71611935
-1.1349953 -0.0282408 0.7358432 -0.13327265 -0.28026074 -0.62530875
-0.3704965 -0.7846486 -0.28522345 0.17205808 0.6801979 0.86831665
0.25463384 0.43959534 0.38950258 -1.0400454 -0.36918575 -1.5684693
-0.14141762 0.24197662 0.14377339 -0.5675488 -0.3177191 -0.41800952] | [8.468006134033203, 10.665050506591797] |
e830d661-4ad2-474d-b20a-fc5b868c4ec3 | generative-models-for-local-network-community | 1804.04469 | null | http://arxiv.org/abs/1804.04469v1 | http://arxiv.org/pdf/1804.04469v1.pdf | Generative models for local network community detection | Local network community detection aims to find a single community in a large
network, while inspecting only a small part of that network around a given seed
node. This is much cheaper than finding all communities in a network. Most
methods for local community detection are formulated as ad-hoc optimization
problems. In this work, we instead start from a generative model for networks
with community structure. By assuming that the network is uniform, we can
approximate the structure of unobserved parts of the network to obtain a method
for local community detection. We apply this local approximation technique to
two variants of the stochastic block model. To our knowledge, this results in
the first local community detection methods based on probabilistic models.
Interestingly, in the limit, one of the proposed approximations corresponds to
conductance, a popular metric in this field. Experiments on real and synthetic
datasets show comparable or improved results compared to state-of-the-art local
community detection algorithms. | ['Twan van Laarhoven'] | 2018-04-12 | null | null | null | null | ['local-community-detection'] | ['graphs'] | [ 1.70096919e-01 3.47292215e-01 -7.04868697e-03 2.23191962e-01
-2.62358308e-01 -8.73626530e-01 5.91154397e-01 8.15925062e-01
-1.62107721e-01 5.32984316e-01 -1.91859394e-01 -1.40279442e-01
-1.70750618e-01 -1.25079834e+00 -6.03563130e-01 -8.22795630e-01
-4.51234102e-01 9.31926429e-01 7.00285077e-01 2.38734446e-02
3.01006138e-01 5.04445851e-01 -8.77117753e-01 9.01514813e-02
6.92889214e-01 1.47436813e-01 1.10697493e-01 8.15069854e-01
-4.24345285e-02 6.61791146e-01 -5.16610146e-01 -2.45615318e-01
7.49186128e-02 -6.05098188e-01 -9.52019334e-01 3.66491199e-01
3.79826203e-02 1.69799805e-01 -4.11120206e-01 1.22418487e+00
3.59647959e-01 -4.60182309e-01 6.16551936e-01 -1.31983304e+00
-1.11928046e-01 1.19643939e+00 -8.58777702e-01 1.18492477e-01
4.94450599e-01 -2.73352027e-01 1.24493420e+00 -5.87285280e-01
9.94760394e-01 1.22676122e+00 8.93335640e-01 4.24370877e-02
-1.91103613e+00 -4.62188095e-01 2.68414199e-01 -1.62995577e-01
-1.86959052e+00 -7.96570927e-02 6.85864329e-01 -6.75005794e-01
3.69100034e-01 -4.41446565e-02 7.42954373e-01 7.34807193e-01
-4.28742357e-02 6.48240924e-01 1.05476439e+00 -5.21531105e-01
2.76333362e-01 1.68744847e-01 1.57888114e-01 7.23700523e-01
8.45957100e-01 -4.07850206e-01 -1.50933325e-01 -6.84385538e-01
7.27615356e-01 8.30762535e-02 -3.50208461e-01 -9.86398041e-01
-1.30652368e+00 1.08771312e+00 5.28415740e-01 6.37075245e-01
-2.03773916e-01 3.85956258e-01 2.00784594e-01 3.23198706e-01
4.10672814e-01 -5.14794625e-02 2.35454336e-01 3.97024751e-01
-1.32666516e+00 1.34071171e-01 1.35569358e+00 8.58526945e-01
9.26228523e-01 -5.44871330e-01 7.58110061e-02 2.40875840e-01
4.61829394e-01 8.93262476e-02 -2.75845557e-01 -6.33956790e-01
2.80983269e-01 7.79148757e-01 -2.20317636e-02 -1.27437592e+00
-2.26922110e-01 -7.99651682e-01 -1.24660659e+00 1.37038408e-02
6.54140055e-01 -8.95053446e-02 -4.13096577e-01 1.58831489e+00
3.52625877e-01 3.07610691e-01 -2.58028179e-01 4.22313124e-01
2.74110198e-01 5.81967175e-01 -5.87493300e-01 -4.67992008e-01
9.85025465e-01 -8.36792469e-01 -3.06691200e-01 6.46843091e-02
2.53215104e-01 -5.74472487e-01 1.35340050e-01 3.60657305e-01
-8.44255209e-01 -3.04997414e-02 -9.04691219e-01 6.74723387e-01
-5.66546507e-02 -2.20348507e-01 3.93735975e-01 8.85872066e-01
-1.33013344e+00 6.97701097e-01 -7.46685624e-01 -8.80642295e-01
1.40123352e-01 2.62046605e-01 -1.47118539e-01 -3.65318507e-01
-6.43607557e-01 2.98733592e-01 4.36934441e-01 6.28644302e-02
-1.24612713e+00 -9.65145081e-02 -4.72070217e-01 1.83019951e-01
6.52553856e-01 -5.93624234e-01 7.87884772e-01 -9.31490004e-01
-8.00849199e-01 8.19682956e-01 -3.21040660e-01 -6.39744580e-01
6.35506272e-01 5.20886004e-01 7.12417960e-02 3.44636530e-01
2.50673532e-01 2.46988118e-01 6.32994890e-01 -1.39176488e+00
-1.40022144e-01 5.40608317e-02 4.80487570e-02 -3.58280361e-01
-1.69525266e-01 2.43570402e-01 -4.17344093e-01 -4.07809168e-01
3.84744704e-01 -9.42062318e-01 -6.19260728e-01 1.40524998e-01
-7.65057266e-01 -2.54000187e-01 6.48546398e-01 -1.46814093e-01
1.31616724e+00 -1.65379739e+00 2.01230481e-01 1.01115108e+00
8.84362876e-01 -8.28758106e-02 -1.10492572e-01 1.05377626e+00
-9.91238654e-02 6.02998257e-01 -4.34492528e-01 -4.92777407e-01
-2.24742830e-01 5.10534011e-02 4.65927944e-02 1.07619548e+00
-3.70660312e-02 4.89036024e-01 -1.15357292e+00 -6.26136065e-01
-2.38058880e-01 2.70105362e-01 -7.00927556e-01 -2.42957920e-01
1.27272710e-01 1.32967561e-01 -3.97288978e-01 4.21314806e-01
7.93892026e-01 -7.52043605e-01 7.87093699e-01 5.02649605e-01
-1.04261070e-01 1.37263788e-02 -1.81603670e+00 1.31611943e+00
6.84954002e-02 6.37575746e-01 6.07489944e-01 -1.44134521e+00
7.93324590e-01 5.86496413e-01 6.07701421e-01 3.48016441e-01
7.33444169e-02 1.99681342e-01 2.11690068e-01 2.43168458e-01
-5.35493866e-02 1.00962698e-01 1.39379099e-01 8.99654329e-01
-1.38932317e-02 2.64221162e-01 5.90508223e-01 7.79688656e-01
1.61346531e+00 -5.93249619e-01 5.77093661e-01 -6.40020192e-01
3.70287418e-01 -1.25659034e-01 3.07970107e-01 1.43261683e+00
-2.00001717e-01 7.70672262e-01 1.05252254e+00 1.74316429e-02
-1.18224657e+00 -1.09090662e+00 8.13419819e-02 3.42984587e-01
2.37973556e-01 -8.35439384e-01 -1.04495537e+00 -3.14536035e-01
-1.09795131e-01 -2.70884961e-01 -6.26928806e-01 1.89585835e-01
-4.51227635e-01 -8.18020523e-01 3.25373113e-01 1.56680092e-01
1.27021506e-01 -6.73632085e-01 -2.78690774e-02 5.91819823e-01
-4.89788324e-01 -9.17165041e-01 -4.82052952e-01 9.06265974e-02
-1.15068710e+00 -1.28041923e+00 -7.22401381e-01 -9.01722372e-01
8.76930773e-01 4.46533978e-01 1.23375988e+00 6.80552661e-01
-3.60135883e-01 1.38068974e-01 -3.34986687e-01 2.16256380e-01
-7.28335023e-01 3.31385255e-01 -2.29982864e-02 2.86071330e-01
-2.05499306e-02 -7.50414133e-01 -4.64343756e-01 2.69200385e-01
-6.22132480e-01 -2.21172273e-01 5.27515233e-01 6.57983661e-01
3.73480141e-01 5.91943622e-01 7.71068260e-02 -8.96759629e-01
6.95624709e-01 -8.19751620e-01 -6.96782947e-01 3.06600809e-01
-6.66545093e-01 1.85926691e-01 3.86650920e-01 -4.99817967e-01
-5.08610189e-01 2.85656482e-01 5.02045512e-01 -1.77903622e-01
-2.98256874e-02 8.29876363e-01 4.96319383e-02 -1.73326567e-01
5.66427588e-01 1.49468049e-01 -1.26660550e-02 -5.23826957e-01
1.29895225e-01 3.84322464e-01 6.00254051e-02 -5.29613554e-01
1.21379697e+00 9.27175641e-01 1.71911865e-01 -8.48615348e-01
-1.45146117e-01 -1.08789885e+00 -9.96381104e-01 -2.16675460e-01
2.93482840e-01 -9.63342190e-01 -7.89165616e-01 3.00412595e-01
-1.26779687e+00 -1.39217362e-01 -1.10280633e-01 1.04319744e-01
-1.97309107e-01 7.50828564e-01 -6.72122359e-01 -1.28841376e+00
9.16989073e-02 -8.02205980e-01 7.75520682e-01 -1.77363649e-01
-2.25501712e-02 -1.07088125e+00 6.01128817e-01 -1.89085886e-01
2.33065471e-01 4.73092496e-01 3.96311998e-01 -4.83909637e-01
-1.01503980e+00 -5.33663511e-01 -3.20039898e-01 -6.91619441e-02
7.28417840e-03 4.24345434e-01 -6.95703030e-01 -6.29725873e-01
-3.82377923e-01 2.04024568e-01 1.05929005e+00 5.36964774e-01
5.96764088e-01 -4.21884269e-01 -9.12752807e-01 1.76923305e-01
1.76452374e+00 -2.90032476e-01 6.27538204e-01 6.67998418e-02
5.37707269e-01 5.18523574e-01 1.13558620e-01 5.80198407e-01
6.89433664e-02 4.75816011e-01 4.73912001e-01 -1.38069108e-01
1.40662625e-01 -2.86119819e-01 2.84863144e-01 6.68162107e-01
-2.34493628e-01 -5.17793953e-01 -1.02789032e+00 1.06098199e+00
-2.02532578e+00 -1.36418366e+00 -7.14123607e-01 2.26411915e+00
8.36657405e-01 3.84013116e-01 6.29961431e-01 1.53970435e-01
1.50444055e+00 -7.65968934e-02 7.01228902e-02 -8.08979757e-03
-2.30782807e-01 3.28420848e-02 5.50161421e-01 6.36870503e-01
-1.09714925e+00 5.25950968e-01 6.71909428e+00 8.52807641e-01
-6.27046347e-01 1.90008447e-01 5.96448421e-01 2.21329063e-01
-2.67303258e-01 6.84068978e-01 -7.33287930e-01 2.14729741e-01
7.26935387e-01 -3.07971448e-01 5.20916641e-01 8.38148952e-01
1.76523760e-01 -2.57547554e-02 -9.33739424e-01 3.98266107e-01
-2.12765887e-01 -1.25306678e+00 -2.63591200e-01 6.48199797e-01
1.09361851e+00 1.36421159e-01 -6.12257659e-01 -3.61377954e-01
5.43116212e-01 -9.40870106e-01 6.63803697e-01 5.66411793e-01
3.74924242e-01 -7.19161570e-01 6.18119121e-01 6.36456013e-01
-1.74827540e+00 -5.46039129e-03 -3.83681148e-01 -2.50526965e-02
2.03870282e-01 1.05785525e+00 -9.41339612e-01 3.24091852e-01
5.88657856e-01 6.84032619e-01 -6.24870121e-01 1.69132674e+00
-1.22633446e-02 8.67125869e-01 -8.49765182e-01 -1.67752445e-01
1.93362564e-01 -3.02175254e-01 9.80850160e-01 1.29233968e+00
2.21838832e-01 -3.85607839e-01 4.12463456e-01 1.19257569e+00
-2.44866312e-01 -1.50972623e-02 -7.32683480e-01 -3.23727757e-01
3.83126289e-01 1.30550337e+00 -1.40661800e+00 -2.37801343e-01
-2.96641409e-01 8.71859968e-01 3.58393192e-01 2.87144780e-01
-4.54437673e-01 -4.63193238e-01 2.35141322e-01 6.43113852e-01
5.74128568e-01 -3.38183135e-01 5.61570823e-02 -1.23029065e+00
-5.45233600e-02 -7.12220252e-01 3.74480873e-01 -2.99673468e-01
-1.32232928e+00 3.17740232e-01 8.65616500e-02 -1.00722241e+00
-4.53080423e-02 -2.07946405e-01 -9.18663740e-01 7.55529463e-01
-1.05836844e+00 -9.96810019e-01 -2.35172525e-01 4.43510890e-01
-3.76323313e-02 2.79633194e-01 5.38394988e-01 1.09853037e-01
-3.07984352e-01 1.32635355e-01 6.22693300e-01 3.70883435e-01
4.26375598e-01 -1.56040931e+00 5.32040596e-01 1.35818660e+00
1.27330944e-01 1.05843091e+00 7.42484450e-01 -8.02650511e-01
-7.72319496e-01 -8.12058508e-01 1.17917633e+00 -8.27036202e-02
8.82409811e-01 -7.81154871e-01 -8.29313636e-01 5.30741155e-01
3.49896580e-01 5.42226322e-02 2.57419288e-01 1.78921714e-01
-2.08949283e-01 5.90373650e-02 -1.04804623e+00 4.13239837e-01
1.18884826e+00 -4.79446083e-01 1.53705612e-01 5.24127305e-01
3.26356560e-01 3.68189216e-01 -7.61111021e-01 -6.16316777e-03
4.27434146e-01 -8.60360146e-01 1.07048655e+00 -4.18446399e-02
-1.51025737e-02 -6.45454705e-01 2.28108525e-01 -1.03454173e+00
-5.07743120e-01 -1.08461463e+00 -6.07349277e-02 1.43887758e+00
2.23452672e-01 -5.26779234e-01 1.00966454e+00 -3.02061468e-01
5.52316487e-01 -3.10384542e-01 -1.00101733e+00 -7.57500172e-01
4.48585860e-02 9.67155993e-02 2.35459879e-01 9.32054818e-01
5.41608557e-02 2.90872574e-01 -1.15810849e-01 3.53232056e-01
1.31160510e+00 3.13760012e-01 6.76003277e-01 -1.92138767e+00
-5.32043219e-01 -6.33691072e-01 -4.85788345e-01 -8.77679467e-01
-9.65901315e-02 -7.69411922e-01 -1.67943481e-02 -1.72371233e+00
8.78140748e-01 -6.14124179e-01 7.13395029e-02 1.75370917e-01
7.71067590e-02 3.68563175e-01 -1.87143296e-01 3.71101677e-01
-6.78565919e-01 -7.77083188e-02 9.59548891e-01 -3.41835260e-01
-1.36796787e-01 4.13073808e-01 -7.50880718e-01 4.64985609e-01
6.98052347e-01 -1.02735782e+00 -2.20329657e-01 1.84723184e-01
4.29226369e-01 2.34163133e-03 5.67010522e-01 -9.68242347e-01
7.54783094e-01 -6.18790388e-02 -9.87608209e-02 -6.93260014e-01
1.09204911e-02 -6.49051070e-01 4.68116939e-01 8.37338746e-01
-2.76598409e-02 -6.96072504e-02 -2.87889749e-01 1.17816067e+00
-1.33919731e-01 -5.72768331e-01 8.15993369e-01 -3.45036209e-01
1.23808496e-02 1.30416542e-01 -8.99326146e-01 -1.20384678e-01
9.90108371e-01 -3.53488088e-01 -1.92170009e-01 -6.50448442e-01
-8.70605528e-01 1.25902608e-01 7.52377152e-01 -1.92937061e-01
3.55330974e-01 -9.13458526e-01 -1.02957582e+00 -3.06375235e-01
1.02153488e-01 -5.59552945e-02 -1.80052057e-01 1.10781765e+00
-8.38933647e-01 3.35667968e-01 2.91961819e-01 -8.35751772e-01
-1.56589258e+00 7.50871658e-01 4.51873034e-01 -6.56940341e-01
-3.18938762e-01 6.15645111e-01 1.68606997e-01 -3.05604219e-01
9.66544077e-02 -1.74152493e-01 2.19402313e-02 -3.02747991e-02
3.06213468e-01 4.15172428e-01 -2.01902106e-01 -6.18080378e-01
-5.76852441e-01 5.03035069e-01 4.16992530e-02 -1.50855303e-01
1.22953296e+00 -2.66239256e-01 -7.35465407e-01 2.52592742e-01
1.09984696e+00 3.90739650e-01 -6.62491322e-01 -3.31705272e-01
1.27068952e-01 -2.71207243e-01 -1.17764086e-01 -9.55167636e-02
-9.63533640e-01 6.44886553e-01 3.68674666e-01 7.61985242e-01
8.10016453e-01 3.63302946e-01 -8.66005644e-02 5.07149339e-01
5.70248306e-01 -6.57800317e-01 1.06428772e-01 1.98198870e-01
4.68739361e-01 -9.33567584e-01 2.90455222e-01 -8.06610942e-01
1.01489566e-01 1.08744264e+00 3.26466635e-02 -5.25402546e-01
8.87838304e-01 2.58278668e-01 -6.72725260e-01 -2.78351486e-01
-6.69872820e-01 -3.08425725e-01 -1.45285934e-01 6.92737877e-01
3.11464995e-01 -3.79813984e-02 -2.35366806e-01 1.00715883e-01
3.79281133e-01 -3.24791789e-01 9.88768637e-01 8.83992136e-01
-6.73130989e-01 -1.39761138e+00 -5.58517754e-01 1.93630740e-01
-3.69999081e-01 -3.37180585e-01 -8.35971475e-01 7.18770206e-01
-3.07322964e-02 1.20078981e+00 -4.06484902e-02 6.33543655e-02
-1.62715986e-01 -2.55271047e-01 4.43558246e-01 -9.07547057e-01
-5.50045371e-01 4.53353047e-01 -4.04707156e-02 -2.90235966e-01
-7.15985298e-01 -9.31659579e-01 -7.70806074e-01 -7.13923216e-01
-1.06039453e+00 4.54342335e-01 3.89510751e-01 5.75460672e-01
9.08512026e-02 1.86485499e-01 5.78504205e-01 -6.08870447e-01
-2.05756724e-01 -8.82980287e-01 -1.07258904e+00 -7.03293532e-02
3.02085459e-01 -4.33288753e-01 -6.58670723e-01 -3.97063047e-02] | [6.921823978424072, 5.208592891693115] |
77966e60-cb5c-49ad-ab62-3299420fd506 | a-generic-approach-for-enhancing-gans-by | 2112.03502 | null | https://arxiv.org/abs/2112.03502v1 | https://arxiv.org/pdf/2112.03502v1.pdf | A Generic Approach for Enhancing GANs by Regularized Latent Optimization | With the rapidly growing model complexity and data volume, training deep generative models (DGMs) for better performance has becoming an increasingly more important challenge. Previous research on this problem has mainly focused on improving DGMs by either introducing new objective functions or designing more expressive model architectures. However, such approaches often introduce significantly more computational and/or designing overhead. To resolve such issues, we introduce in this paper a generic framework called {\em generative-model inference} that is capable of enhancing pre-trained GANs effectively and seamlessly in a variety of application scenarios. Our basic idea is to efficiently infer the optimal latent distribution for the given requirements using Wasserstein gradient flow techniques, instead of re-training or fine-tuning pre-trained model parameters. Extensive experimental results on applications like image generation, image translation, text-to-image generation, image inpainting, and text-guided image editing suggest the effectiveness and superiority of our proposed framework. | ['Jinhui Xu', 'Changyou Chen', 'Chunyuan Li', 'Yufan Zhou'] | 2021-12-07 | null | null | null | null | ['text-guided-image-editing'] | ['computer-vision'] | [ 4.72145200e-01 7.29917735e-02 -4.89020832e-02 -3.91981870e-01
-7.24235237e-01 -2.65320808e-01 8.09492767e-01 -3.17613125e-01
-1.56502366e-01 8.09261143e-01 3.52511592e-02 -2.29409665e-01
1.06250562e-01 -7.67494798e-01 -6.62858844e-01 -7.69943953e-01
4.57236797e-01 5.82711399e-01 -2.71693170e-01 1.09279990e-01
3.07813078e-01 3.00032198e-01 -1.26769447e+00 -1.78289145e-01
1.32192159e+00 7.95016468e-01 6.10945761e-01 6.18788183e-01
-3.24755520e-01 7.76648343e-01 -5.97800434e-01 -5.91567099e-01
2.27379844e-01 -9.01861370e-01 -5.58006763e-01 4.32049900e-01
1.88599467e-01 -4.39270645e-01 -1.73763689e-02 1.13252831e+00
5.06084681e-01 2.56544769e-01 5.71733177e-01 -1.13597083e+00
-8.38762403e-01 4.20343608e-01 -7.16233194e-01 -1.88686326e-01
-3.05162854e-02 1.11077115e-01 6.55436575e-01 -7.07831740e-01
6.14538193e-01 1.21430242e+00 3.48932326e-01 7.65308797e-01
-1.40015697e+00 -5.54356873e-01 4.02692929e-02 7.76453540e-02
-1.34955215e+00 -5.08183718e-01 8.87360036e-01 -3.25186133e-01
4.37143058e-01 2.79260218e-01 3.55265945e-01 1.14312220e+00
-2.52902154e-02 8.76994848e-01 1.18390775e+00 -5.22424936e-01
2.52556652e-01 1.99966028e-01 -5.93671560e-01 6.45074666e-01
9.18325633e-02 -2.29348034e-01 -3.39415431e-01 -1.15889743e-01
1.22818172e+00 1.87550224e-02 -2.85193563e-01 -3.83809209e-01
-1.04975295e+00 8.92827392e-01 1.28204182e-01 1.50861025e-01
-5.02829254e-01 3.90001059e-01 1.56423196e-01 1.37435958e-01
6.98058605e-01 4.54817533e-01 -2.14044377e-01 -3.05256754e-01
-1.21626174e+00 2.71462560e-01 5.01794696e-01 1.05144334e+00
8.49121988e-01 3.27374578e-01 -6.91378117e-02 1.00504148e+00
3.35831553e-01 3.45879793e-01 4.61550802e-01 -1.01405942e+00
5.20953119e-01 3.17145050e-01 2.00286970e-01 -9.28174257e-01
1.41816929e-01 -4.74471658e-01 -1.04706216e+00 -2.03222986e-02
6.19864464e-03 -1.90737233e-01 -1.17496455e+00 1.81506634e+00
3.57498795e-01 2.12983221e-01 -9.93760377e-02 7.97531188e-01
2.09285751e-01 8.41321766e-01 5.41175418e-02 -2.53254145e-01
8.97309601e-01 -1.09109557e+00 -9.32915986e-01 -2.69996732e-01
3.65901291e-01 -1.00912321e+00 1.27189517e+00 1.47724405e-01
-1.23823237e+00 -7.02285230e-01 -8.31719697e-01 -1.05546579e-01
8.72496739e-02 2.20571175e-01 6.00571454e-01 5.61006308e-01
-1.11829650e+00 5.24085164e-01 -1.02836597e+00 -2.44200721e-01
4.83084142e-01 2.57133812e-01 -5.84152713e-02 -1.49761304e-01
-8.53079736e-01 6.80245459e-01 2.86271304e-01 2.00709358e-01
-1.04019213e+00 -7.10190833e-01 -8.20790768e-01 1.12078421e-01
3.02774727e-01 -9.75671589e-01 1.20590806e+00 -1.13112366e+00
-1.94988370e+00 5.46753347e-01 -2.58976489e-01 -3.72631133e-01
6.77328229e-01 -3.63098770e-01 -1.28288165e-01 -1.41711205e-01
-4.39507104e-02 8.71705234e-01 1.26857579e+00 -1.39464152e+00
-2.61036217e-01 -1.65129036e-01 -4.61545512e-02 2.76898742e-01
-5.16792417e-01 -1.00323692e-01 -5.02511263e-01 -9.78329003e-01
-1.37229383e-01 -8.46868038e-01 -4.35370713e-01 -1.72471210e-01
-4.89623070e-01 3.95872183e-02 9.63657141e-01 -7.59255111e-01
1.25842583e+00 -1.90236855e+00 4.31212068e-01 4.15015109e-02
1.19490638e-01 4.20086294e-01 -2.58408546e-01 5.55394292e-01
1.97797984e-01 1.75864398e-01 -4.53279108e-01 -7.05365956e-01
3.37392427e-02 3.87436301e-01 -3.76801670e-01 4.64903936e-02
2.87837058e-01 1.00267243e+00 -7.52356231e-01 -6.51484251e-01
4.29027021e-01 6.52104199e-01 -7.64131248e-01 4.70981151e-01
-4.61203516e-01 9.57108200e-01 -4.47355062e-01 3.86269450e-01
6.69611692e-01 -4.03106481e-01 2.45059595e-01 -2.48339195e-02
1.37660861e-01 5.71893305e-02 -1.06718075e+00 1.94998860e+00
-8.73377442e-01 6.14928246e-01 8.80413782e-03 -1.06288373e+00
9.91069913e-01 1.09207496e-01 3.26454729e-01 -6.75901294e-01
1.86783031e-01 6.35207891e-02 -1.72036171e-01 -4.16012883e-01
5.14168441e-01 -1.99830994e-01 2.60143042e-01 5.86620867e-01
-1.46590052e-02 -9.49779898e-02 1.90596476e-01 1.16199832e-02
5.85077524e-01 4.22428548e-01 5.71458414e-02 -8.17169249e-02
5.17931700e-01 -2.94466794e-01 5.45966208e-01 4.64445025e-01
3.06421548e-01 6.43907368e-01 3.06415439e-01 -2.70156711e-01
-1.24365485e+00 -9.16844785e-01 2.38559261e-01 9.02480423e-01
-3.21577564e-02 -4.36783403e-01 -1.12102199e+00 -4.68009353e-01
-4.56186414e-01 9.68125165e-01 -4.74282324e-01 -7.30760917e-02
-6.33377612e-01 -9.23513770e-01 3.18221301e-01 5.70327759e-01
7.59191453e-01 -9.69509900e-01 -5.77786863e-01 2.24515781e-01
-4.70431060e-01 -9.35452104e-01 -6.23440981e-01 -2.25193977e-01
-1.15676975e+00 -5.08235455e-01 -1.04364622e+00 -6.95487201e-01
1.08340323e+00 1.50978863e-01 1.12995970e+00 4.81596626e-02
-2.17417628e-01 -1.16652092e-02 -2.70066977e-01 -2.12818131e-01
-6.51529551e-01 1.88510001e-01 -3.19937468e-01 2.30836079e-01
-1.13443807e-01 -7.49755442e-01 -7.20242977e-01 3.60286295e-01
-1.29222453e+00 6.51196599e-01 7.49027729e-01 9.35237944e-01
7.45573699e-01 -2.34242026e-02 5.66976190e-01 -1.01102960e+00
7.43931174e-01 -2.08661422e-01 -6.90358400e-01 3.24848980e-01
-8.10964286e-01 3.60776961e-01 6.66462183e-01 -4.32503313e-01
-1.39192641e+00 -8.07280988e-02 -2.12432146e-01 -5.24676561e-01
2.46150084e-02 5.08798718e-01 -3.91953200e-01 1.68820232e-01
3.93181562e-01 4.84789222e-01 -4.39115390e-02 -5.57574272e-01
5.65614641e-01 5.45977235e-01 4.44439709e-01 -7.45410562e-01
8.51633132e-01 3.95216525e-01 -1.05509795e-01 -5.83797693e-01
-6.17040813e-01 4.85092141e-02 -3.65182370e-01 -1.65095031e-02
8.48813534e-01 -7.70751059e-01 -2.54795939e-01 4.42900568e-01
-1.10251915e+00 -5.53899646e-01 -2.22600088e-01 2.71176636e-01
-6.43482566e-01 4.18177873e-01 -4.00915921e-01 -6.79481566e-01
-4.89424944e-01 -1.32118487e+00 1.17305028e+00 2.09726840e-01
-7.29597583e-02 -1.18989682e+00 2.12506995e-01 4.05249804e-01
8.00356388e-01 2.87767828e-01 1.05979443e+00 1.10726729e-02
-8.33259940e-01 -1.21394530e-01 -1.98368967e-01 5.42944968e-01
4.37245756e-01 -1.82393398e-02 -7.74924099e-01 -3.35493147e-01
-6.30148947e-02 -2.30479941e-01 5.88142455e-01 4.51147020e-01
1.48278522e+00 -3.91486198e-01 -1.68615773e-01 8.56616735e-01
1.39545155e+00 2.04268306e-01 8.47056746e-01 1.32149696e-01
8.42294753e-01 4.06895667e-01 5.15831411e-01 5.08455753e-01
2.94988424e-01 7.02179134e-01 3.16720247e-01 -1.99012458e-01
-1.03280000e-01 -5.43776095e-01 3.08287829e-01 9.41132247e-01
-1.73992470e-01 -5.36310911e-01 -6.41160309e-01 4.80837196e-01
-1.89401042e+00 -7.76743412e-01 3.33764493e-01 2.05001187e+00
9.07962501e-01 -1.80198908e-01 -1.76720172e-01 -1.10376716e-01
7.70084023e-01 1.82904348e-01 -6.02440298e-01 -2.72396117e-01
2.21119866e-01 4.13921297e-01 2.53691316e-01 5.46873391e-01
-7.63448656e-01 1.01967931e+00 6.23737860e+00 1.05897605e+00
-1.25262129e+00 1.86899140e-01 8.41405272e-01 1.18403234e-01
-6.17036104e-01 1.04212716e-01 -5.25120854e-01 5.11061966e-01
7.30317771e-01 -1.24945439e-01 6.44703269e-01 7.98852384e-01
3.55157435e-01 -2.79793367e-02 -9.13469493e-01 1.11985898e+00
1.71186358e-01 -1.41163361e+00 4.14107084e-01 1.66746274e-01
1.08359933e+00 -4.65891719e-01 1.98675781e-01 1.58615232e-01
2.32475027e-01 -1.01742768e+00 6.55606091e-01 4.69512850e-01
8.22843134e-01 -7.65678227e-01 5.58146060e-01 4.33358103e-01
-8.82519126e-01 3.35310280e-01 -4.27534938e-01 5.33539429e-02
2.74342895e-01 6.10504568e-01 -8.62665474e-01 5.19030273e-01
3.01823914e-01 5.11003375e-01 -4.34720188e-01 7.76330590e-01
-4.05354917e-01 6.82874084e-01 -9.82933119e-02 2.16915876e-01
1.25245333e-01 -4.53310817e-01 2.68437743e-01 9.44859862e-01
6.11071944e-01 -2.87377089e-01 -5.51831722e-02 1.23047006e+00
-2.70068735e-01 -4.72917361e-03 -4.13709521e-01 -3.27456325e-01
2.32841656e-01 1.37749088e+00 -7.59534657e-01 -3.75365168e-01
-1.98205039e-01 1.32638788e+00 1.54786855e-01 4.74901199e-01
-1.15442550e+00 -3.03582460e-01 5.54830074e-01 1.51756540e-01
2.88542956e-01 -4.25815850e-01 -2.07756445e-01 -1.19774079e+00
-4.90017645e-02 -7.95899153e-01 -4.90312539e-02 -7.52387106e-01
-9.48125541e-01 7.07989275e-01 2.06833389e-02 -1.11788595e+00
-6.58718169e-01 -2.11287707e-01 -6.04649246e-01 9.07458246e-01
-1.57495761e+00 -1.26619983e+00 -4.89814878e-01 5.16303241e-01
7.74850965e-01 -6.09079842e-03 6.90869093e-01 4.40887839e-01
-6.29668891e-01 6.24699950e-01 2.45166793e-01 -1.83388337e-01
6.67837024e-01 -1.13381124e+00 4.91358727e-01 9.19023037e-01
1.58116043e-01 6.59931004e-01 7.20634282e-01 -4.96960312e-01
-1.23287737e+00 -1.23909974e+00 6.25935674e-01 -1.71458498e-01
3.06675822e-01 -4.31843132e-01 -6.70509636e-01 7.22470820e-01
5.32467186e-01 -3.32185090e-01 5.71986735e-01 -2.04521805e-01
-4.12267409e-02 -9.31495354e-02 -1.05862272e+00 7.27615893e-01
1.02340448e+00 -3.85627151e-01 -3.62768658e-02 2.49151140e-01
5.41869462e-01 -4.35217381e-01 -7.61925519e-01 3.73916298e-01
2.85738766e-01 -8.34907413e-01 7.60558128e-01 -4.03556973e-01
5.95892012e-01 -3.53657186e-01 -5.78062842e-03 -1.50419915e+00
-1.31827593e-01 -1.02062082e+00 -1.09015480e-01 1.38682997e+00
1.49305806e-01 -4.38628107e-01 8.16271007e-01 5.88061094e-01
-1.48801714e-01 -7.49190271e-01 -5.26087880e-01 -4.94572490e-01
-1.65819839e-01 -2.32592255e-01 7.61116087e-01 8.67318213e-01
-6.35484278e-01 2.45660394e-01 -8.89744461e-01 -7.39561021e-02
6.10001624e-01 2.90397048e-01 1.15272093e+00 -8.36361170e-01
-6.01138175e-01 -3.89211684e-01 -1.42319798e-01 -1.40971899e+00
2.90692765e-02 -6.58194244e-01 1.85645252e-01 -1.73267782e+00
2.71313131e-01 -4.24401253e-01 -4.53758650e-02 3.15357178e-01
-2.48708740e-01 1.10847473e-01 1.62925541e-01 2.18448266e-01
-4.47316676e-01 8.69751155e-01 1.59091222e+00 2.95247007e-02
-9.49290954e-03 -7.00764880e-02 -8.12009573e-01 6.04415596e-01
7.52665579e-01 -3.23686510e-01 -8.25523794e-01 -8.60534012e-01
1.71778888e-01 5.95419630e-02 3.63649607e-01 -8.47218931e-01
-7.57824183e-02 -3.96803081e-01 3.56321335e-01 -3.66211802e-01
2.87286758e-01 -7.36852348e-01 4.67426330e-01 2.65688300e-01
-3.48477423e-01 2.00341985e-01 1.26729161e-01 5.06768048e-01
-4.31040823e-01 -2.44580492e-01 8.35118532e-01 -4.70250882e-02
-4.51556593e-01 3.50643188e-01 -1.38940841e-01 -6.34065084e-03
9.67396915e-01 -9.42380205e-02 -3.46060358e-02 -6.88144088e-01
-5.85457385e-01 -1.74541119e-03 6.56960547e-01 4.55764681e-01
6.44660294e-01 -1.37882674e+00 -6.04758561e-01 3.31358224e-01
-1.22873157e-01 2.00148538e-01 3.66590738e-01 6.40903831e-01
-5.34983993e-01 2.63831854e-01 -1.98660240e-01 -5.22532046e-01
-9.77137446e-01 2.49896720e-01 1.84339315e-01 -6.19199038e-01
-3.86458218e-01 7.55847394e-01 4.92543578e-01 -2.38919586e-01
3.80989397e-03 -2.26751834e-01 3.11569810e-01 -3.47480267e-01
2.88616806e-01 3.77638638e-01 -1.59627169e-01 -3.86435628e-01
1.20570138e-01 4.39804792e-01 -1.73544705e-01 -1.66205928e-01
1.27624941e+00 -2.72867411e-01 -1.08774312e-01 7.67170489e-02
1.00910282e+00 -2.29244485e-01 -1.48128057e+00 -1.76501766e-01
-1.94626033e-01 -6.36696279e-01 1.39385223e-01 -6.84510410e-01
-1.27510285e+00 1.08727813e+00 5.40880561e-01 2.87274215e-02
1.32332838e+00 -1.94935083e-01 1.06938899e+00 1.50708944e-01
3.53923112e-01 -1.07027185e+00 3.19706112e-01 1.47465721e-01
8.52143943e-01 -9.94269013e-01 -2.04385325e-01 -1.99810684e-01
-6.49539828e-01 9.41549838e-01 6.43029392e-01 -9.24988613e-02
3.31423402e-01 2.16816947e-01 -4.63312045e-02 9.88957956e-02
-4.75718200e-01 1.13942862e-01 1.95225060e-01 5.27486384e-01
4.82209384e-01 -3.98959629e-02 -2.72407740e-01 1.85969129e-01
-9.11175907e-02 1.48351297e-01 2.07218051e-01 9.60274398e-01
-2.08971754e-01 -1.58103442e+00 -2.13811591e-01 3.58645767e-01
-4.02895093e-01 -1.40920326e-01 -9.56505388e-02 5.92144549e-01
-6.03192206e-03 7.37512410e-01 -5.74133992e-02 -2.16024548e-01
3.76902260e-02 7.08662421e-02 6.92407727e-01 -4.54073280e-01
-8.46826807e-02 4.09297913e-01 -2.64399379e-01 -3.80427748e-01
-6.37082458e-01 -5.69063842e-01 -1.01153970e+00 -3.04696321e-01
-4.31430250e-01 8.46686363e-02 7.53383934e-01 8.23196709e-01
6.51236594e-01 6.02297068e-01 6.13819599e-01 -8.89671504e-01
-6.30022943e-01 -1.16434932e+00 -2.83415556e-01 4.05257374e-01
8.05290192e-02 -5.53028464e-01 6.00285158e-02 4.47891027e-01] | [11.520604133605957, -0.4754030704498291] |
11a30642-512b-4d63-b5b3-5e008f0661b0 | 3d-semantic-scene-completion-from-a-single | 1905.06231 | null | https://arxiv.org/abs/1905.06231v1 | https://arxiv.org/pdf/1905.06231v1.pdf | 3D Semantic Scene Completion from a Single Depth Image using Adversarial Training | We address the task of 3D semantic scene completion, i.e. , given a single depth image, we predict the semantic labels and occupancy of voxels in a 3D grid representing the scene. In light of the recently introduced generative adversarial networks (GAN), our goal is to explore the potential of this model and the efficiency of various important design choices. Our results show that using conditional GANs outperforms the vanilla GAN setup. We evaluate these architecture designs on several datasets. Based on our experiments, we demonstrate that GANs are able to outperform the performance of a baseline 3D CNN in case of clean annotations, but they suffer from poorly aligned annotations. | ['Yueh-Tung Chen', 'Martin Garbade', 'Juergen Gall'] | 2019-05-15 | null | null | null | null | ['3d-semantic-scene-completion'] | ['computer-vision'] | [ 4.79900956e-01 6.56071067e-01 3.48867863e-01 -3.34126920e-01
-7.77157605e-01 -5.38171053e-01 9.76433873e-01 -4.76790577e-01
-3.51884216e-01 7.85554111e-01 4.13003653e-01 -1.31288305e-01
3.68391484e-01 -7.23961115e-01 -8.89896154e-01 -5.81036687e-01
1.31015047e-01 6.90115273e-01 2.33076513e-02 1.80328488e-02
4.29174602e-02 6.85237050e-01 -1.46651518e+00 1.05719082e-01
5.26039839e-01 1.12988722e+00 7.55535513e-02 7.31811285e-01
-6.35062680e-02 9.54456270e-01 -6.53734148e-01 -1.68534413e-01
4.72670913e-01 -3.56079161e-01 -8.49042416e-01 9.72702801e-02
6.41835749e-01 -5.59802949e-01 -4.13701832e-01 8.57425928e-01
4.59068507e-01 1.65726572e-01 9.21907604e-01 -1.13777006e+00
-3.99388790e-01 7.43389428e-02 -2.57345200e-01 -7.39613920e-02
3.66126090e-01 2.21993417e-01 7.95918465e-01 -8.86399209e-01
9.08110082e-01 1.23766303e+00 7.04365134e-01 9.06883657e-01
-1.32513154e+00 -3.67101103e-01 2.28286371e-01 -2.75748700e-01
-1.22992778e+00 -4.92137372e-01 8.00955534e-01 -6.15053952e-01
1.16350794e+00 -6.08536489e-02 5.98158658e-01 1.68301439e+00
1.24111913e-01 6.83762372e-01 1.33654726e+00 -3.35511148e-01
6.40020669e-01 -2.43904516e-01 -3.30626756e-01 5.48414648e-01
-1.80559009e-02 7.11974129e-02 -5.90586960e-01 1.24438845e-01
1.05080497e+00 -1.63674250e-01 -7.46063516e-02 -6.56920016e-01
-9.71474648e-01 1.00452387e+00 7.02054620e-01 7.20171183e-02
-3.58954936e-01 7.02908576e-01 5.57453632e-02 -8.11198875e-02
7.92769253e-01 5.12419879e-01 -2.06511304e-01 -2.15886366e-02
-8.81332994e-01 4.55218047e-01 5.53353548e-01 1.03061593e+00
7.15998232e-01 1.98530406e-01 -2.81594098e-01 4.57018346e-01
2.84995794e-01 2.85844922e-01 4.39745607e-03 -1.05234015e+00
3.30646038e-01 4.68807727e-01 3.29504237e-02 -3.99339437e-01
-3.81016999e-01 -2.60540754e-01 -9.42791045e-01 7.78468132e-01
3.87386471e-01 -9.15442221e-03 -1.70326281e+00 1.70465410e+00
3.20393831e-01 3.40102911e-01 4.23319452e-02 8.30279887e-01
7.25752950e-01 3.99717242e-01 3.01260531e-01 3.98461699e-01
8.93153787e-01 -8.90719950e-01 -5.45393169e-01 -5.43936133e-01
1.67141721e-01 -4.33841050e-01 9.79440868e-01 2.79052943e-01
-1.21346951e+00 -5.91136336e-01 -1.01792979e+00 -3.35259795e-01
-4.09302771e-01 -2.05325708e-01 8.68796706e-01 6.66500926e-01
-1.39637208e+00 5.00394106e-01 -1.16131663e+00 -2.68241853e-01
9.44336236e-01 1.47823051e-01 -4.56830174e-01 -2.80459702e-01
-7.40353286e-01 9.44562733e-01 1.74141094e-01 1.38879046e-01
-1.54929268e+00 -7.62154162e-01 -1.01920819e+00 -1.55939206e-01
1.10110443e-03 -1.07213974e+00 1.28184426e+00 -7.25564241e-01
-1.52158380e+00 1.24897480e+00 -3.32986079e-02 -5.49890041e-01
8.19018304e-01 -3.39563042e-01 3.14984232e-01 1.14770703e-01
7.28637949e-02 1.07297504e+00 6.95347428e-01 -1.54645205e+00
-3.20946127e-01 -4.32746708e-01 2.21073419e-01 2.24679559e-01
2.50656337e-01 -5.13878584e-01 -3.74722600e-01 -5.79201281e-01
8.89307410e-02 -9.50361490e-01 -6.95474386e-01 1.64240412e-02
-6.84052289e-01 2.54797012e-01 4.52969998e-01 -6.38363838e-01
1.46111280e-01 -2.07126069e+00 3.55612338e-01 1.41330600e-01
2.16222256e-01 -1.36001989e-01 3.65793742e-02 2.08943129e-01
2.60170475e-02 1.38900951e-01 -5.45797169e-01 -1.01578569e+00
1.27216488e-01 5.10999799e-01 -3.30763519e-01 3.68408799e-01
4.21555579e-01 1.23494363e+00 -7.08470643e-01 -2.12291270e-01
5.41244447e-01 8.24412584e-01 -8.13758790e-01 6.81253731e-01
-4.99743640e-01 1.09377849e+00 -4.60439950e-01 5.17404616e-01
6.92409515e-01 -2.50020564e-01 -7.26210102e-02 2.78505584e-04
2.10043088e-01 5.48356056e-01 -7.34695911e-01 2.41260815e+00
-5.07285535e-01 5.40312588e-01 4.65404205e-02 -8.61657381e-01
9.21365142e-01 1.21350817e-01 3.64504158e-01 -7.97162533e-01
8.99897888e-04 8.51228759e-02 -5.12218654e-01 1.37755508e-02
2.93648332e-01 -2.32919395e-01 -1.83794558e-01 2.57668793e-01
3.77319157e-01 -7.11125553e-01 -4.90338236e-01 1.28133833e-01
1.39827025e+00 4.98895705e-01 1.25926241e-01 -2.32467651e-01
5.58959432e-02 8.47835839e-02 1.04686692e-01 8.54084969e-01
1.36178419e-01 1.17131662e+00 5.10762572e-01 -5.41400254e-01
-1.24225187e+00 -1.33554208e+00 4.32340465e-02 5.35149515e-01
4.05582227e-02 -1.66334406e-01 -9.19681132e-01 -9.20841396e-01
-6.40411079e-02 7.75747478e-01 -1.03184223e+00 -9.68297571e-03
-5.55041552e-01 -5.84182620e-01 4.68005329e-01 6.55885518e-01
4.42146420e-01 -1.08111584e+00 -8.95172477e-01 2.77111307e-03
2.41468698e-01 -1.51224208e+00 1.49202317e-01 5.73867559e-01
-7.76237071e-01 -9.10102785e-01 -6.92049384e-01 -4.97906715e-01
7.07235932e-01 -7.88236633e-02 1.53422153e+00 -1.36801198e-01
-1.59491837e-01 6.87931776e-01 -3.13791752e-01 -4.09881175e-01
-4.51718986e-01 2.34046310e-01 -3.68959069e-01 -4.13151950e-01
-1.27424886e-02 -9.44052577e-01 -8.61021221e-01 -5.11786304e-02
-1.00353074e+00 4.05057698e-01 3.62989038e-01 8.53774905e-01
8.30643892e-01 -3.16394031e-01 1.97895318e-01 -1.13325644e+00
2.20732957e-01 -4.26737458e-01 -5.90857506e-01 -1.46973416e-01
-4.44361269e-01 1.18865624e-01 3.76474321e-01 7.22784325e-02
-1.00289810e+00 3.61811668e-01 -6.58256114e-01 -6.14067435e-01
-5.97745597e-01 -2.45951284e-02 -2.82393038e-01 -8.24071020e-02
5.18674791e-01 7.36830235e-02 -2.87249386e-01 -5.65698028e-01
5.24670422e-01 1.78018644e-01 7.68310368e-01 -5.69298029e-01
7.40720332e-01 7.51137257e-01 2.88851261e-01 -5.53452432e-01
-1.10589695e+00 -5.87742552e-02 -8.32202256e-01 -4.43795137e-02
1.39662659e+00 -1.19959402e+00 -2.04254061e-01 4.31748956e-01
-1.31574571e+00 -8.88911963e-01 -6.75020814e-01 6.93121552e-02
-1.03527021e+00 -8.11680108e-02 -4.53580379e-01 -7.22347915e-01
-1.41295284e-01 -1.13223040e+00 1.51472783e+00 8.14333633e-02
-2.48519078e-01 -1.22003818e+00 1.14176221e-01 3.56265694e-01
4.51202661e-01 9.69473064e-01 7.28668153e-01 -5.01510859e-01
-9.00693893e-01 2.34113373e-02 -1.51150733e-01 4.06215340e-01
-4.92911087e-03 -4.24448639e-01 -1.64219332e+00 -1.49857834e-01
6.29832223e-02 -4.55830902e-01 9.30386782e-01 5.12978137e-01
1.43991923e+00 1.31920353e-02 -1.39009401e-01 1.04756057e+00
1.56278050e+00 -1.67766623e-02 9.04279292e-01 2.15005264e-01
1.00452530e+00 4.18702126e-01 2.18069628e-01 2.97246397e-01
2.42691413e-01 5.58450103e-01 1.03101814e+00 -3.91509950e-01
-5.39776027e-01 -7.73464739e-01 -7.41931051e-02 2.65497148e-01
-1.02715708e-01 -4.53900754e-01 -9.41290021e-01 4.80168551e-01
-1.59815943e+00 -4.45655376e-01 2.77641386e-01 1.89827073e+00
4.09624636e-01 2.92271256e-01 -2.73283273e-01 -1.33737952e-01
9.19992700e-02 3.24743301e-01 -6.65786862e-01 -3.53156239e-01
-1.17969424e-01 9.66980100e-01 7.04638243e-01 5.36277711e-01
-9.62099195e-01 1.00018084e+00 7.71895599e+00 5.06622434e-01
-7.04476953e-01 3.75391543e-01 9.62125897e-01 -7.90240094e-02
-5.76372147e-01 -1.81825966e-01 -6.15515828e-01 2.94975579e-01
7.05682218e-01 4.14416850e-01 5.30927181e-01 9.10991311e-01
-1.55017287e-01 -1.62861854e-01 -1.44975901e+00 1.01207232e+00
1.35780618e-01 -1.43398547e+00 1.11402005e-01 1.63782462e-01
1.07055056e+00 3.56397152e-01 -2.56527169e-03 2.27812216e-01
8.50866675e-01 -1.68491507e+00 8.03294539e-01 4.88048702e-01
9.85147297e-01 -4.86477435e-01 6.38827562e-01 1.39713451e-01
-7.38772154e-01 3.38270038e-01 -1.93909168e-01 -8.81472453e-02
3.72033238e-01 5.51072180e-01 -8.98520827e-01 4.55786854e-01
6.85713530e-01 6.04724109e-01 -4.30811226e-01 7.48454988e-01
-6.51923716e-01 3.55101645e-01 -3.08804929e-01 4.95225132e-01
4.77979243e-01 -5.32164052e-02 2.93756396e-01 1.03550422e+00
4.24023628e-01 -5.18720560e-02 -9.88694131e-02 1.29798496e+00
-2.16280416e-01 -3.70266438e-01 -9.77150917e-01 4.43583220e-01
1.56358078e-01 8.96787047e-01 -7.83450723e-01 -1.32738888e-01
-2.08416179e-01 1.40727782e+00 5.57498634e-01 4.86527324e-01
-6.72655404e-01 3.23375553e-01 9.60391462e-01 7.61472061e-02
5.27327716e-01 -4.16927397e-01 -6.70208633e-01 -9.54078734e-01
-3.98394801e-02 -4.38450634e-01 2.24722102e-01 -1.05165839e+00
-1.25334179e+00 5.96964657e-01 -1.30459264e-01 -7.75026023e-01
-4.08938378e-01 -7.96812356e-01 -6.20888054e-01 9.55710292e-01
-1.61056447e+00 -1.37247920e+00 -7.30036199e-01 6.56439245e-01
5.07266998e-01 1.34002194e-01 1.04830289e+00 5.47428802e-02
-1.09450102e-01 2.83394843e-01 -1.53649360e-01 -9.50151961e-03
2.63841033e-01 -1.55170703e+00 1.08488429e+00 7.74762034e-01
3.26313555e-01 3.28499414e-02 6.74255133e-01 -5.23106873e-01
-1.22957349e+00 -1.11849725e+00 3.94006729e-01 -8.42445076e-01
1.05356090e-01 -8.37798834e-01 -5.31654418e-01 9.84040439e-01
2.22332686e-01 1.74852222e-01 3.24865937e-01 -1.46047056e-01
-2.20778212e-01 4.02838707e-01 -1.35121715e+00 4.66524214e-01
1.52166319e+00 -5.50522506e-01 -3.68067145e-01 2.77444720e-01
9.38814402e-01 -8.04115534e-01 -7.79700875e-01 5.23864508e-01
2.54547924e-01 -1.31935358e+00 1.17881691e+00 -4.56991076e-01
6.61113501e-01 -1.37046590e-01 -3.78707319e-01 -1.43110538e+00
-6.33667856e-02 -3.65091234e-01 2.08231155e-02 8.58374596e-01
3.69462557e-02 -3.66799146e-01 1.15525317e+00 7.93701053e-01
-4.26177979e-01 -4.97740299e-01 -1.28808820e+00 -5.01840711e-01
2.15819046e-01 -5.13263166e-01 6.09649539e-01 7.08797336e-01
-7.59306669e-01 2.86591500e-01 -4.17403728e-01 4.98712100e-02
6.69135273e-01 -1.05265550e-01 1.10177302e+00 -1.02466631e+00
-3.29067260e-01 -2.26552010e-01 -6.16865158e-01 -1.27345490e+00
3.15353066e-01 -7.75008142e-01 1.28838882e-01 -1.80938578e+00
-1.10971451e-01 -5.89362144e-01 -1.53225794e-01 5.12388527e-01
2.97011770e-02 6.63743556e-01 -2.45170519e-02 -1.18698590e-01
-5.09701192e-01 8.03283393e-01 1.28009915e+00 -1.92575440e-01
-1.46907065e-02 -2.07949579e-01 -4.72838044e-01 6.22300327e-01
6.42013967e-01 -3.78408641e-01 -4.03394282e-01 -6.62382424e-01
1.06471613e-01 -2.10316420e-01 7.65616417e-01 -1.11359954e+00
-1.40831679e-01 9.68309492e-02 6.67353213e-01 -4.73384500e-01
7.41556466e-01 -9.45832789e-01 3.92224759e-01 1.93864658e-01
-2.25527614e-01 -1.94404379e-01 3.32243919e-01 4.62024271e-01
-1.36187881e-01 7.65884146e-02 6.27414882e-01 -3.83601457e-01
-7.52005875e-01 3.92566025e-01 -2.32407041e-02 1.86174929e-01
9.60315704e-01 -2.49883801e-01 -7.84902722e-02 -3.53059441e-01
-6.97286010e-01 -6.75098300e-02 9.11404431e-01 3.06769192e-01
5.86547554e-01 -1.39923882e+00 -5.22731245e-01 3.80489260e-01
7.62180164e-02 8.24452579e-01 2.42070377e-01 2.34378248e-01
-6.50875270e-01 3.37295413e-01 -5.02184093e-01 -7.81507313e-01
-6.46867037e-01 3.61092925e-01 5.90801358e-01 -3.83704364e-01
-9.40894663e-01 9.40648973e-01 6.84995234e-01 -4.26682889e-01
2.62787044e-01 -4.06287879e-01 1.42857864e-01 -4.76577580e-01
1.14179537e-01 -1.24250457e-01 2.82556593e-01 -3.32643718e-01
-2.54894137e-01 4.94488746e-01 3.13949049e-01 -2.49287069e-01
1.56366146e+00 5.49427196e-02 2.24840760e-01 3.82046402e-01
1.04262972e+00 -1.84367031e-01 -1.89729047e+00 1.08646408e-01
-3.73900682e-01 -5.92278600e-01 -3.32798623e-02 -6.91224635e-01
-1.14363086e+00 8.42093229e-01 5.50708711e-01 8.41662101e-03
9.83367682e-01 2.60172129e-01 4.93502647e-01 -1.47411659e-01
4.69149947e-01 -5.45202911e-01 2.62872856e-02 4.47884738e-01
7.12899566e-01 -1.15931058e+00 -1.83103144e-01 -4.16226864e-01
-5.27992189e-01 7.40679085e-01 5.60409188e-01 -4.69301313e-01
4.73371983e-01 3.70627999e-01 -1.14091055e-03 -5.14567375e-01
-4.99210387e-01 -3.30381989e-01 1.97212994e-01 1.01728833e+00
2.04766735e-01 -7.65425935e-02 4.28384572e-01 1.72217831e-01
-5.00512123e-01 -1.57618001e-01 3.31221312e-01 6.84457421e-01
7.42416829e-02 -9.39273179e-01 -1.09536052e-01 1.63120970e-01
-3.60779583e-01 -5.44980206e-02 -4.87115204e-01 8.23012948e-01
1.41441628e-01 5.70191801e-01 1.90490976e-01 -1.97613090e-01
4.34860170e-01 1.68072537e-01 8.15019190e-01 -6.53550625e-01
-4.13104713e-01 -2.62861401e-01 -9.42408666e-03 -8.62524867e-01
-5.91177583e-01 -5.40127456e-01 -8.75523329e-01 -9.68947448e-03
2.24253029e-01 -1.80325672e-01 8.96399736e-01 8.55949283e-01
2.28038549e-01 6.98536754e-01 3.65535408e-01 -1.31922734e+00
-2.19690070e-01 -9.95191813e-01 -4.69835818e-01 6.08315229e-01
3.91409814e-01 -7.44409502e-01 -3.58447134e-01 1.14048265e-01] | [8.938161849975586, -3.0648856163024902] |
bf620573-4472-4bab-b554-bbea21396f08 | multi-armed-bandit-learning-for-tdma | 2302.05301 | null | https://arxiv.org/abs/2302.05301v1 | https://arxiv.org/pdf/2302.05301v1.pdf | Multi-armed Bandit Learning for TDMA Transmission Slot Scheduling and Defragmentation for Improved Bandwidth Usage | This paper proposes a Time Division Multiple Access (TDMA) MAC slot allocation protocol with efficient bandwidth usage in wireless sensor networks and Internet of Things (IoTs). The developed protocol has two primary components: a Multi-Armed Bandits (MAB)-based slot allocation mechanism for collision free transmission, and a Decentralized Defragmented Slot Backshift (DDSB) operation for improving bandwidth usage efficiency. The proposed framework is decentralized in that each node finds its transmission schedule independently without the control of any centralized arbitrator. The developed mechanism is suitable for networks with or without time synchronization, thus, making it suitable for low-complexity wireless transceivers for wireless sensor and IoT nodes. This framework is able to manage the trade-off between learning convergence time and bandwidth. In addition, it allows the nodes to adapt to topological changes while maintaining efficient bandwidth usage. The developed logic is tested for both fully-connected and arbitrary mesh networks with extensive simulation experiments. It is shown how the nodes can learn to select collision-free transmission slots using MAB. Moreover, the nodes learn to self-adjust their transmission schedules using a novel DDSB framework in order to reduce bandwidth usage. | ['Subir Biswas', 'Amit Kumar Bhuyan', 'Hrishikesh Dutta'] | 2023-01-14 | null | null | null | null | ['multi-armed-bandits'] | ['miscellaneous'] | [ 1.50521681e-01 7.14610994e-01 -6.20431781e-01 -3.47017199e-01
-2.01832615e-02 -2.29776919e-01 2.37130120e-01 6.59793839e-02
-1.03637779e+00 1.33649850e+00 -1.03110445e+00 -5.11523664e-01
-7.05322385e-01 -1.21827519e+00 -6.81503862e-02 -1.35143745e+00
-5.91330886e-01 8.91101420e-01 7.28935540e-01 -3.95704322e-02
-4.16347608e-02 6.57168150e-01 -1.40932965e+00 -8.77526700e-01
3.72656792e-01 1.30055702e+00 4.61252749e-01 1.96589604e-01
-5.18667996e-01 -1.20214626e-01 -6.67617798e-01 1.91837296e-01
1.37458920e-01 -2.12913379e-02 -7.19168365e-01 2.16709487e-02
-6.35631144e-01 -1.53427482e-01 2.80287921e-01 4.49068993e-01
4.56876516e-01 -6.72645569e-02 5.30074596e-01 -1.42939591e+00
8.22682977e-01 1.03401482e+00 -6.16727769e-01 1.82481229e-01
-2.36334547e-01 -3.35999250e-01 4.79577690e-01 1.66929290e-01
6.82967722e-01 9.55126643e-01 1.38227805e-01 6.48178220e-01
-1.21544707e+00 -1.04297042e+00 -3.47536393e-02 2.25065872e-01
-1.18204105e+00 -3.60435098e-01 7.01598346e-01 8.25045109e-02
3.79842788e-01 5.25561869e-01 7.05839038e-01 4.96920496e-01
3.61503989e-01 9.41797420e-02 7.59212196e-01 -7.16999233e-01
8.77941966e-01 2.08219126e-01 -2.77563035e-01 2.62862325e-01
7.26096451e-01 -1.80701241e-01 -1.23407124e-02 6.01217896e-03
8.16086352e-01 -4.17370856e-01 1.45524994e-01 -6.12713158e-01
-1.08253443e+00 7.90131867e-01 3.85760844e-01 9.45862234e-01
-7.00685382e-01 4.65559453e-01 2.91376889e-01 4.64938551e-01
1.65629625e-01 8.31584260e-03 -5.64949095e-01 2.52764404e-01
-6.57451868e-01 3.46700358e-03 1.02610469e+00 8.20973039e-01
9.75187361e-01 -7.98777938e-02 1.30270630e-01 3.89274389e-01
5.13678849e-01 9.15475190e-01 1.42465547e-01 -9.39301848e-01
5.33246323e-02 -5.40554225e-02 1.80988759e-01 -5.12286842e-01
-1.32396328e+00 -4.59044635e-01 -1.23217559e+00 4.09453392e-01
3.71119887e-01 -5.22103131e-01 -3.85105789e-01 1.92171931e+00
9.66100574e-01 -2.53907621e-01 2.11559951e-01 5.90260804e-01
4.28387858e-02 7.42995501e-01 3.62242162e-01 -1.00266612e+00
1.62184834e+00 -3.53800833e-01 -1.16332042e+00 -5.60268876e-04
4.85822380e-01 -5.30269384e-01 -8.71337652e-02 2.65654594e-01
-1.24161375e+00 -1.15271866e-01 -1.12689054e+00 8.56056094e-01
-3.37596774e-01 -2.16587365e-01 5.92751265e-01 1.02663088e+00
-8.03899884e-01 1.31747261e-01 -9.09450829e-01 -6.00108504e-01
9.23545510e-02 8.79099071e-01 6.74243420e-02 4.50849801e-01
-1.25815868e+00 5.22651613e-01 8.00966978e-01 -1.89657554e-01
-1.28224358e-01 -3.30377907e-01 -6.54058516e-01 -3.71578522e-02
4.77878124e-01 -7.49894500e-01 1.12077105e+00 -6.23702228e-01
-1.77187943e+00 2.76735783e-01 2.44778901e-01 -7.65745521e-01
4.47958648e-01 7.68931329e-01 -4.05196577e-01 5.13763726e-01
9.50543284e-02 6.36911035e-01 6.00110590e-01 -9.54964519e-01
-9.39441919e-01 -1.28186628e-01 -1.33767501e-01 -1.19298302e-01
-4.76093858e-01 -4.17721212e-01 1.54193246e-03 -2.60237634e-01
2.30753615e-01 -8.66922438e-01 -5.42569518e-01 3.65735441e-01
-1.66975304e-01 -4.55814630e-01 1.10409176e+00 4.31262016e-01
1.03899932e+00 -1.81160820e+00 4.62920740e-02 7.89913535e-01
-5.92586957e-02 1.32084906e-01 -8.05040747e-02 5.84359050e-01
5.59414089e-01 -1.30319387e-01 1.75860897e-01 -2.48625457e-01
-5.72278276e-02 6.81412876e-01 2.43402079e-01 5.01036763e-01
-2.77057052e-01 -8.86296704e-02 -5.85750878e-01 -8.23932946e-01
4.04416591e-01 2.02652201e-01 -7.93708935e-02 -1.83556512e-01
-4.16508079e-01 4.46511656e-01 -1.06116581e+00 1.13615885e-01
1.00359166e+00 -4.40266244e-02 6.28901362e-01 -3.12954515e-01
-7.98518121e-01 -1.32350385e-01 -1.54396081e+00 1.43060625e+00
-6.13087177e-01 1.06001440e-02 8.60890985e-01 -1.53570056e+00
9.72705662e-01 4.02444661e-01 1.21580279e+00 -1.10348499e+00
4.99764264e-01 5.12735784e-01 -3.01977694e-01 -2.55273968e-01
-7.59144351e-02 -1.97386190e-01 -8.50487500e-02 6.57326162e-01
-1.12094082e-01 7.46440217e-02 3.70660514e-01 -1.84906423e-01
9.33695316e-01 -5.44160962e-01 3.38075370e-01 -6.60966277e-01
8.14740539e-01 -1.81505591e-01 5.34404218e-01 7.76200116e-01
-1.07679656e-02 -9.04989660e-01 -9.61263571e-03 -4.32640672e-01
-6.56992435e-01 -1.04027843e+00 -1.43995464e-01 8.95378709e-01
7.21694946e-01 2.49336705e-01 -4.55195904e-01 -2.54971273e-02
-1.61245853e-01 4.85259235e-01 -2.41091937e-01 1.32832944e-01
-2.44638622e-01 -3.53076637e-01 2.43205145e-01 -4.58464593e-01
7.58229196e-01 -8.68870735e-01 -1.53023767e+00 9.04066741e-01
1.95119586e-02 -1.06795859e+00 1.76294252e-01 6.01331472e-01
-7.04825521e-01 -9.62463140e-01 -4.28300321e-01 -6.92980766e-01
2.55879700e-01 3.00710917e-01 5.10022759e-01 -6.24243580e-02
-3.58745009e-01 7.12630689e-01 -4.07648981e-01 -5.94653606e-01
-3.88338596e-01 6.76869333e-01 5.98054891e-03 -1.58537924e-01
5.76219559e-02 -7.83309162e-01 -7.92705953e-01 8.70469451e-01
-1.15154946e+00 1.31176308e-01 7.83174455e-01 4.21889007e-01
5.93836546e-01 3.74438763e-01 9.23416555e-01 -3.46759081e-01
7.03847334e-02 -5.55693388e-01 -1.18797386e+00 1.61791608e-01
-6.93367004e-01 3.11572880e-01 3.28880399e-01 -4.01516646e-01
-8.55035484e-01 1.49713963e-01 1.11354843e-01 4.87325042e-01
-8.03871453e-02 3.87461111e-02 1.58042625e-01 -5.32387555e-01
1.36437953e-01 -1.36778355e-01 4.35548484e-01 -3.77359122e-01
4.86489153e-03 6.22229040e-01 -8.86701569e-02 -4.11693484e-01
9.62117910e-01 7.54431844e-01 8.49644184e-01 -1.32350218e+00
-1.89365208e-01 -1.50307104e-01 -1.82768866e-01 -4.61736411e-01
7.88320959e-01 -5.77740014e-01 -1.22667742e+00 2.50886261e-01
-9.77888346e-01 -3.20167571e-01 -1.34716898e-01 6.75240815e-01
-4.68262762e-01 2.91717857e-01 -1.71432003e-01 -1.08812451e+00
-5.84452510e-01 -7.62271464e-01 3.23297560e-01 6.36118591e-01
-8.16244781e-02 -9.46517825e-01 -2.88951904e-01 -8.84352252e-02
1.06646299e+00 4.03699607e-01 5.78277171e-01 -4.18298185e-01
-6.15315378e-01 3.18106622e-01 -1.27931044e-01 -3.40774655e-01
2.36658067e-01 -2.46119767e-01 -1.66877225e-01 -6.86898887e-01
-2.48673916e-01 -2.06445038e-01 3.60868037e-01 6.22334421e-01
6.78133726e-01 -2.75262803e-01 -7.32076228e-01 2.14800108e-02
1.42756474e+00 3.78026813e-01 3.14384878e-01 7.12040842e-01
-3.49763125e-01 6.24642015e-01 9.35891569e-01 1.05475819e+00
3.33341092e-01 8.35803926e-01 1.17416167e+00 -1.38466761e-01
3.11549246e-01 6.54856563e-01 -2.18183864e-02 1.80380628e-01
2.81573951e-01 -4.94415224e-01 -3.61268580e-01 2.25953206e-01
-1.72158146e+00 -6.59476101e-01 1.19835235e-01 2.11903119e+00
5.59035063e-01 4.36974615e-01 2.68911093e-01 5.65757573e-01
7.92185485e-01 -4.91832569e-02 -5.01998663e-01 -4.85144436e-01
6.14744909e-02 5.07052779e-01 1.08844006e+00 4.93004233e-01
-8.84791732e-01 3.76731813e-01 4.20933485e+00 1.11430597e+00
-1.19606662e+00 1.03807844e-01 5.39706089e-02 2.77969390e-01
-6.37390539e-02 -2.20817417e-01 -3.17195624e-01 4.47862089e-01
8.60105097e-01 -2.88025826e-01 2.75429606e-01 4.72652107e-01
5.31738758e-01 -6.75022066e-01 -2.71705747e-01 7.15393782e-01
-8.17912281e-01 -1.30495405e+00 -3.29436481e-01 1.56389996e-01
1.77624434e-01 -1.40767679e-01 -3.28588307e-01 -4.04697865e-01
-1.53018534e-01 -2.61686891e-01 6.74136698e-01 2.92597115e-01
6.80150151e-01 -7.78523266e-01 7.03359902e-01 4.13263381e-01
-1.36016512e+00 -2.33341038e-01 -7.60909989e-02 -9.56950858e-02
4.07708675e-01 7.17580318e-01 -6.39710128e-01 7.25535989e-01
4.79556412e-01 -1.22242235e-01 3.20276201e-01 1.18674922e+00
4.31691825e-01 2.42183521e-01 -1.04168129e+00 -5.98275304e-01
2.85738856e-01 -5.77270724e-02 6.52216434e-01 7.80904531e-01
5.71563482e-01 1.18383050e-01 2.05540642e-01 4.75624859e-01
4.02516305e-01 -5.80358692e-03 -1.55238518e-02 3.94993961e-01
1.19163799e+00 1.25227785e+00 -1.18553543e+00 3.52199897e-02
-6.59456477e-02 3.45810860e-01 -4.76420403e-01 1.61615744e-01
-6.59998953e-01 -7.18013704e-01 6.32631719e-01 1.16101682e-01
6.26359224e-01 -2.51872301e-01 1.03323534e-01 -1.58293486e-01
-5.16754091e-01 8.93603340e-02 5.93976974e-01 -1.16534039e-01
-7.52123177e-01 2.55514890e-01 3.20306212e-01 -1.07750833e+00
-2.22184956e-01 -1.40395314e-01 -3.31962526e-01 2.19854563e-01
-1.76199603e+00 -8.44846308e-01 -1.93409294e-01 7.40221322e-01
3.07890207e-01 -7.20201246e-03 8.67655694e-01 4.68950570e-01
-2.74602830e-01 3.54214102e-01 1.08473964e-01 -5.24171054e-01
2.29504302e-01 -6.47185504e-01 -8.09544086e-01 2.90818185e-01
-6.70469403e-01 6.54386356e-02 1.22036469e+00 -4.48352508e-02
-1.41862822e+00 -6.41062677e-01 3.98443252e-01 1.18789196e+00
4.66950953e-01 -2.64727604e-02 -2.37247184e-01 -1.63102113e-02
1.53752834e-01 -3.11989278e-01 6.79819524e-01 -5.24088502e-01
4.30974036e-01 -8.44143391e-01 -1.62924516e+00 1.15818046e-01
3.57883543e-01 6.16966426e-01 2.81167328e-01 1.09539978e-01
4.41237926e-01 1.94058456e-02 -7.74051607e-01 4.01995808e-01
7.00449228e-01 -7.34288394e-01 6.89186692e-01 3.56581539e-01
-9.17373896e-01 -4.32266563e-01 1.63507268e-01 -9.47580636e-01
-1.31730571e-01 -9.31257963e-01 3.59731644e-01 1.26167929e+00
2.52459109e-01 -1.02976453e+00 6.73413217e-01 2.61786461e-01
4.16269362e-01 -2.55220801e-01 -1.92679167e+00 -7.92344809e-01
-2.84810960e-01 -1.92553774e-01 6.21890366e-01 4.19594407e-01
-5.14324456e-02 3.11567098e-01 -1.87819168e-01 3.53801876e-01
1.10581565e+00 -2.05363899e-01 7.84149706e-01 -1.72710967e+00
3.67386937e-02 -4.04439420e-01 -2.13277221e-01 -7.74684548e-01
4.55751969e-03 -2.20055461e-01 -9.68065783e-02 -1.28624570e+00
-6.77046955e-01 -1.19590068e+00 -2.27814004e-01 4.58348066e-01
1.06077075e+00 -1.38243530e-02 3.55517156e-02 -8.77982900e-02
-7.81134248e-01 1.94705069e-01 1.03032982e+00 -3.55671602e-03
-2.95274973e-01 6.18898749e-01 1.01891205e-01 1.87221691e-01
1.27902400e+00 -5.45454502e-01 -7.95843661e-01 2.03892719e-02
1.05010070e-01 6.98224306e-01 1.19386688e-01 -1.13215315e+00
3.55604202e-01 -4.46429521e-01 -2.38955259e-01 -4.73232836e-01
2.66228080e-01 -1.75958860e+00 6.34353876e-01 1.39726436e+00
-1.77714899e-01 -4.21997488e-01 1.52465403e-01 7.71040738e-01
4.56859708e-01 -9.95976031e-02 1.09915912e+00 3.59124273e-01
-6.02747798e-01 4.44498323e-02 -9.44025278e-01 -7.33739078e-01
1.64675057e+00 -2.16197655e-01 -1.00927681e-01 -3.40747595e-01
-8.46531987e-01 7.20898330e-01 -1.51221417e-02 8.57658535e-02
1.65508185e-02 -8.11912835e-01 -7.18543530e-02 -2.10032277e-02
-9.71409008e-02 -7.30829313e-02 4.52406406e-01 9.03606296e-01
-4.25051212e-01 4.53027070e-01 -5.33863485e-01 -5.15217781e-01
-1.15506947e+00 1.99115038e-01 3.32889557e-01 -3.58855873e-01
-1.26109570e-01 2.55941272e-01 -6.78174913e-01 1.71969473e-01
4.48714435e-01 -1.89380184e-01 -2.82480717e-01 1.73177958e-01
3.10107231e-01 5.69848359e-01 -1.91576377e-01 -1.78030759e-01
-4.96223778e-01 5.66610873e-01 3.59234989e-01 -2.08044752e-01
1.19737756e+00 -8.48023772e-01 -4.73322630e-01 7.96665847e-02
6.68398619e-01 -3.71717542e-01 -6.66551650e-01 -2.95267522e-01
2.12842524e-01 7.17976391e-02 3.23132396e-01 -4.30516243e-01
-1.20065486e+00 -1.83770433e-02 7.76988924e-01 8.61266553e-01
1.25975049e+00 -1.29483327e-01 5.86099982e-01 5.80798268e-01
1.07823539e+00 -1.21966922e+00 -1.18015796e-01 6.77850842e-02
-9.76557843e-03 -5.66080809e-01 -1.00887507e-01 -6.77091956e-01
2.01268673e-01 1.78254700e+00 4.11617696e-01 3.05904299e-01
1.00087225e+00 5.00964642e-01 -1.53209105e-01 -1.41075730e-01
-5.54435849e-01 -4.46598291e-01 -7.70010769e-01 7.17170954e-01
-2.55670786e-01 8.99668038e-02 -1.20382977e+00 1.74339801e-01
2.15470269e-01 5.13244048e-02 4.98162478e-01 1.04393089e+00
-1.21822619e+00 -1.57865131e+00 -3.60825568e-01 1.72302246e-01
-2.01488033e-01 8.56221795e-01 6.44991577e-01 1.25516915e+00
2.10644782e-01 1.21653068e+00 3.92076403e-01 9.42731202e-02
-8.53904709e-02 -3.25677127e-01 2.67911851e-01 -1.38689399e-01
1.68724343e-01 1.85926065e-01 1.24384560e-01 -5.53534091e-01
-9.56043839e-01 -3.00740957e-01 -1.60113525e+00 -2.03423202e-01
-4.25367832e-01 8.06274295e-01 1.19841695e+00 9.68214273e-01
1.22101232e-01 3.98378134e-01 1.07831073e+00 -7.50979602e-01
-2.38927871e-01 -5.28835952e-01 -9.75116074e-01 -5.02504826e-01
5.43318450e-01 -9.98554170e-01 -4.68633533e-01 -6.29222751e-01] | [5.958409309387207, 1.58686101436615] |
49f2a6d1-bd55-4a86-a8dd-448e5b46c2a0 | fvqa-2-0-introducing-adversarial-samples-into | 2303.10699 | null | https://arxiv.org/abs/2303.10699v1 | https://arxiv.org/pdf/2303.10699v1.pdf | FVQA 2.0: Introducing Adversarial Samples into Fact-based Visual Question Answering | The widely used Fact-based Visual Question Answering (FVQA) dataset contains visually-grounded questions that require information retrieval using common sense knowledge graphs to answer. It has been observed that the original dataset is highly imbalanced and concentrated on a small portion of its associated knowledge graph. We introduce FVQA 2.0 which contains adversarial variants of test questions to address this imbalance. We show that systems trained with the original FVQA train sets can be vulnerable to adversarial samples and we demonstrate an augmentation scheme to reduce this vulnerability without human annotations. | ['Bill Byrne', 'Zhilin Wang', 'Weizhe Lin'] | 2023-03-19 | null | null | null | null | ['common-sense-reasoning'] | ['reasoning'] | [-2.09389940e-01 6.74957335e-01 -9.76772383e-02 -1.58037752e-01
-9.56854522e-01 -1.17529583e+00 4.37405735e-01 2.29010180e-01
1.59358367e-01 6.81611240e-01 2.20520362e-01 -6.48887157e-01
1.86022390e-02 -1.04540741e+00 -1.00919187e+00 1.45799667e-02
8.27760920e-02 5.52917302e-01 6.70506239e-01 -3.99059981e-01
1.79336026e-01 4.46171641e-01 -1.06147480e+00 7.45946109e-01
8.09756458e-01 7.52507925e-01 -6.66860819e-01 9.85328197e-01
-3.74267310e-01 1.73059809e+00 -1.42881882e+00 -1.14463866e+00
3.25409085e-01 -3.64528447e-01 -1.48773372e+00 -1.64826348e-01
1.23537016e+00 -1.59620151e-01 -8.29071105e-01 1.02032578e+00
4.62033510e-01 6.47281716e-03 6.79251790e-01 -2.00291109e+00
-1.39056277e+00 -7.12103210e-04 -5.85352242e-01 5.15599906e-01
8.06222558e-01 3.37829322e-01 9.43283796e-01 -7.67081499e-01
1.11687064e+00 1.42194092e+00 5.76092780e-01 6.59454346e-01
-1.02034259e+00 -4.52564865e-01 -5.41487709e-02 8.14989686e-01
-1.33607256e+00 -8.26905742e-02 1.14084339e+00 -4.57332999e-01
8.23668718e-01 4.68582600e-01 5.64198136e-01 1.11454129e+00
1.01790018e-01 8.16346884e-01 1.10036838e+00 -3.76858592e-01
5.25876224e-01 5.25523312e-02 2.61498719e-01 9.87060845e-01
2.59930700e-01 -9.57199782e-02 -5.49372733e-01 -2.15285942e-01
4.57651854e-01 -4.41466242e-01 -4.01273757e-01 -6.91928685e-01
-8.36270630e-01 9.26826179e-01 1.23773789e+00 -1.28090814e-01
5.50547354e-02 2.83892542e-01 4.48992074e-01 6.44380808e-01
1.84486374e-01 6.15533173e-01 -2.14242607e-01 4.16509122e-01
-4.65723693e-01 2.70766020e-01 9.44459140e-01 9.52113271e-01
8.81159127e-01 1.73145473e-01 -6.89973593e-01 3.46760780e-01
1.45728216e-01 4.04005706e-01 8.71627480e-02 -1.05194914e+00
5.82856476e-01 1.16894448e+00 1.40841395e-01 -1.37120652e+00
-2.54742980e-01 -8.64537582e-02 -4.48582977e-01 6.91972613e-01
7.00277388e-01 1.72415003e-01 -1.44376338e+00 1.27087367e+00
5.79809844e-01 -2.24417731e-01 2.44783580e-01 9.49279785e-01
1.47894371e+00 3.77895862e-01 4.80039895e-01 5.81592441e-01
1.33085608e+00 -1.01943433e+00 -8.39257479e-01 -3.80315334e-01
2.77956784e-01 -3.91432047e-01 1.62831974e+00 1.35726305e-02
-8.23298931e-01 -2.45796904e-01 -1.30137491e+00 -5.17223537e-01
-9.51598406e-01 -3.88127804e-01 4.18805778e-01 9.24899042e-01
-1.13076293e+00 -7.75172859e-02 -2.09694564e-01 -1.95408911e-01
1.15025234e+00 -1.47381678e-01 -6.28580391e-01 -2.61177123e-01
-1.12919724e+00 1.09492314e+00 2.26761773e-01 -4.37961891e-02
-1.07926297e+00 -7.32880771e-01 -9.98610914e-01 -2.86306113e-01
5.50915360e-01 -8.05971622e-01 9.82625008e-01 -1.17038965e+00
-9.65551555e-01 1.24385607e+00 7.46317506e-02 -3.64328980e-01
5.74340284e-01 -6.25318512e-02 -5.06375194e-01 8.29169095e-01
-4.36292477e-02 5.83002567e-01 1.08473849e+00 -1.68349552e+00
2.33516186e-01 -6.21976256e-01 8.06286395e-01 -1.39041292e-02
5.93533274e-04 -3.12685281e-01 -3.57324153e-01 -8.36789966e-01
-3.26591432e-01 -4.41022843e-01 6.43062368e-02 5.69455087e-01
-7.52147734e-01 -1.25798747e-01 1.42518413e+00 -9.52720344e-01
7.76423514e-01 -1.69059670e+00 -2.59409964e-01 2.52407074e-01
7.42747247e-01 3.78945231e-01 -4.17998374e-01 3.64691645e-01
-9.41412672e-02 5.18603086e-01 -1.04247630e-01 3.24782014e-01
-3.62621881e-02 4.40397024e-01 -6.28889382e-01 4.32571262e-01
4.84914392e-01 1.66948020e+00 -9.89387870e-01 -7.09436595e-01
-6.81153759e-02 2.88606584e-01 -3.72995794e-01 3.21989983e-01
-6.57381415e-01 1.36473194e-01 -3.62004220e-01 1.12612760e+00
9.17799234e-01 -4.54697341e-01 4.62703444e-02 -4.55696970e-01
7.53779233e-01 -1.60405830e-01 -6.17980778e-01 1.40450704e+00
4.09245910e-03 1.08815527e+00 -1.54481232e-01 -7.78797925e-01
7.66679943e-01 1.38786554e-01 -2.80376405e-01 -1.07158947e+00
-9.97900516e-02 -3.80352825e-01 -2.91870207e-01 -7.76113629e-01
2.89199919e-01 -3.04914657e-02 -6.59024343e-02 1.91125512e-01
2.66672045e-01 -4.15270627e-01 4.99263965e-02 9.44302499e-01
1.43249869e+00 -6.91451952e-02 2.46567607e-01 -5.41911134e-03
3.00174743e-01 7.67763078e-01 2.88531691e-01 7.19414175e-01
-8.36687386e-01 9.06960249e-01 8.96035373e-01 -6.79181278e-01
-9.68305826e-01 -1.70830941e+00 3.45631897e-01 8.40507686e-01
3.51798356e-01 -2.25071892e-01 -6.53721035e-01 -1.43560553e+00
1.79437533e-01 7.66158581e-01 -1.20192993e+00 -2.83170432e-01
-2.05690399e-01 2.90914588e-02 9.96866465e-01 6.99900329e-01
5.58060169e-01 -1.32430506e+00 -5.41640699e-01 -3.84348989e-01
-1.67480391e-02 -1.03929913e+00 -1.44999966e-01 -3.22412670e-01
-4.40549463e-01 -2.01606607e+00 -6.17296576e-01 -6.83328271e-01
8.28733802e-01 5.40877618e-02 2.03360796e+00 4.27716345e-01
-3.42007458e-01 1.14503229e+00 -5.03169596e-01 -7.41762817e-01
-4.68730509e-01 -3.19467068e-01 -9.03885663e-01 -3.23183089e-01
3.78813803e-01 -1.86181486e-01 -8.90064716e-01 1.65042415e-01
-1.28380728e+00 -3.31283003e-01 -9.32512358e-02 5.62060595e-01
4.99067366e-01 -3.65013957e-01 7.95941830e-01 -9.76908028e-01
7.28123486e-01 -4.02985543e-01 -3.35059822e-01 8.68521273e-01
-2.30027288e-02 -4.71574627e-02 3.86273921e-01 -2.32854724e-01
-9.49012339e-01 -2.47057408e-01 9.17084068e-02 -6.20252311e-01
-1.13474041e-01 1.13916405e-01 -3.28060716e-01 -5.65352261e-01
1.19605196e+00 -4.32240009e-01 -3.22732627e-01 6.83119297e-02
9.23490584e-01 2.14901432e-01 7.88202822e-01 -2.45370835e-01
1.07403743e+00 7.78989315e-01 -8.64914954e-02 -3.79356414e-01
-1.12986517e+00 -1.39190301e-01 -3.37591439e-01 -5.43944478e-01
9.25609112e-01 -7.13293970e-01 -8.05309892e-01 1.07057635e-02
-1.27207220e+00 -4.46214229e-01 -7.03454852e-01 -5.57843924e-01
-4.36345607e-01 2.67629802e-01 -2.47839540e-01 -6.13823354e-01
-4.65789855e-01 -4.81591105e-01 8.04227948e-01 3.42283279e-01
-6.45697936e-02 -1.04512215e+00 4.19819444e-01 8.85527611e-01
3.28110546e-01 9.13361311e-01 1.07764602e+00 -6.84183478e-01
-6.09568536e-01 -1.64357439e-01 -5.55440009e-01 1.27744555e-01
-1.25936896e-01 1.29411072e-01 -1.02389967e+00 -2.32286841e-01
-2.97715664e-01 -9.10009086e-01 8.63760948e-01 -2.51490861e-01
1.05345905e+00 -4.56078261e-01 -1.57822967e-01 2.10976064e-01
1.66825926e+00 5.31446449e-02 1.13385534e+00 1.77461162e-01
1.03923321e+00 6.95100665e-01 1.72184378e-01 -3.69691670e-01
6.10023320e-01 2.36466900e-01 9.37983155e-01 -3.74909043e-01
-6.72690809e-01 -5.45679688e-01 4.64256592e-02 1.44996539e-01
3.57585698e-01 -5.61238170e-01 -1.24911106e+00 1.10699832e+00
-1.53911829e+00 -1.09277773e+00 -2.54605800e-01 1.62641025e+00
7.99568474e-01 -2.15261117e-01 5.94565868e-02 2.10131541e-01
5.48530459e-01 3.83359760e-01 -6.79748714e-01 -6.69194996e-01
-3.78863186e-01 3.79562378e-01 1.85382441e-01 5.10789633e-01
-1.03163242e+00 9.04346526e-01 7.52951717e+00 5.53926945e-01
-4.60465103e-01 2.92488158e-01 6.18999004e-01 1.16420746e-01
-7.61259735e-01 -2.28399262e-02 2.22297326e-01 4.71347868e-02
5.09973347e-01 -1.16276063e-01 4.69453894e-02 7.43477702e-01
-6.68840170e-01 -1.06070690e-01 -6.61754966e-01 7.05312133e-01
3.92419100e-01 -1.60300183e+00 6.05141461e-01 -6.48047864e-01
9.19740736e-01 -2.92907476e-01 7.57293329e-02 3.56507957e-01
7.48488545e-01 -1.45692790e+00 5.53287148e-01 5.85461020e-01
9.05431390e-01 -7.63477266e-01 5.74219048e-01 -2.69105941e-01
-1.05943573e+00 1.28607929e-01 -4.20315713e-01 1.98256820e-01
-8.47672224e-02 4.04898554e-01 -8.76055658e-01 6.34042978e-01
9.07541871e-01 4.93150242e-02 -1.25091863e+00 1.01675057e+00
-5.41053593e-01 6.81011319e-01 9.03741419e-02 2.30561376e-01
-1.25505432e-01 4.50703472e-01 5.17669320e-01 7.04818010e-01
-2.95887560e-01 -4.41917926e-02 -1.61020696e-01 8.99581611e-01
-5.31412601e-01 -6.67580068e-02 -9.52586174e-01 2.62828656e-02
8.48043263e-02 9.06188488e-01 -6.28965259e-01 -3.74179214e-01
-5.53512335e-01 1.20346856e+00 8.53461802e-01 8.41297507e-01
-5.63497722e-01 -4.73692447e-01 4.37985063e-01 1.65333748e-01
3.21988136e-01 1.65159792e-01 -1.95633352e-01 -1.12713110e+00
6.84574246e-02 -1.05601227e+00 9.29266036e-01 -1.38642013e+00
-1.70420229e+00 6.80064976e-01 -4.22681093e-01 -8.15176547e-01
1.37487203e-01 -7.67817557e-01 -6.16492033e-01 5.27845085e-01
-1.59071624e+00 -1.42859387e+00 -8.02010715e-01 1.15348983e+00
1.41608715e-01 4.15962946e-04 8.53390753e-01 -1.22618727e-01
-2.57257193e-01 7.30454504e-01 -4.61724371e-01 4.89478976e-01
6.86164379e-01 -1.70142043e+00 5.09558558e-01 6.75838172e-01
4.67707396e-01 5.09122051e-02 6.86760366e-01 -7.62252450e-01
-1.17701495e+00 -1.21934760e+00 3.75146419e-01 -1.35863066e+00
6.49851441e-01 -2.11279348e-01 -1.31061959e+00 9.32870328e-01
6.68988168e-01 5.63147902e-01 7.60027111e-01 -4.43723291e-01
-1.19623113e+00 1.52965650e-01 -1.57844603e+00 5.86424947e-01
9.04582262e-01 -1.07813108e+00 -1.21393323e+00 4.68814224e-01
8.75006914e-01 -5.48658669e-01 -6.84392869e-01 4.18665022e-01
2.12051287e-01 -1.10329247e+00 1.13175535e+00 -1.17627752e+00
2.98230201e-01 -7.15377808e-01 1.78008735e-01 -1.15301597e+00
1.58190340e-01 -3.66239309e-01 -5.88015735e-01 9.53001678e-01
2.40992695e-01 -3.44459951e-01 1.05706787e+00 4.37695771e-01
4.67478454e-01 -3.08973163e-01 -1.13248932e+00 -5.15468478e-01
2.48736635e-01 -2.18025446e-01 4.10427898e-01 1.14889944e+00
-1.31438717e-01 3.99855793e-01 -9.29366797e-02 3.35538656e-01
6.20178103e-01 1.85434252e-01 9.10118520e-01 -9.35747743e-01
1.90128028e-01 7.74869397e-02 -8.73293817e-01 -4.22388464e-01
-5.51828071e-02 -7.33874500e-01 -3.72350782e-01 -2.04983306e+00
9.64992717e-02 2.95721233e-01 -3.10042500e-01 6.13708794e-01
-5.32357633e-01 9.20354962e-01 1.97562993e-01 -8.80309120e-02
-9.55383599e-01 5.09243011e-01 1.58419657e+00 -7.72070169e-01
3.51011902e-01 -6.56717360e-01 -6.90979302e-01 5.93060553e-01
8.42504323e-01 -5.24275184e-01 -6.82824850e-01 -4.43795919e-01
8.41392100e-01 -2.06940085e-01 1.10675573e+00 -7.44990528e-01
2.02567324e-01 9.16279480e-02 7.00819135e-01 -6.13777518e-01
4.00196165e-02 -7.89468706e-01 -2.01846376e-01 2.99974799e-01
-2.20412984e-01 3.17915350e-01 6.89706981e-01 8.67959917e-01
-4.44292843e-01 5.86354844e-02 4.94268417e-01 -5.97558990e-02
-8.67789805e-01 3.49175185e-02 -1.04602940e-01 7.49817729e-01
1.23777664e+00 -6.80267736e-02 -1.23612916e+00 -8.75981510e-01
-6.53758109e-01 4.95746404e-01 5.98511994e-01 3.22927684e-01
9.86947954e-01 -1.45743716e+00 -6.39929473e-01 -3.24945927e-01
5.60552061e-01 -2.22285360e-01 6.08475983e-01 -2.91689355e-02
-1.09019995e+00 7.27205118e-03 -3.95122707e-01 -1.80502981e-01
-1.01822162e+00 1.12999749e+00 5.29949903e-01 -2.59288222e-01
-4.09411401e-01 9.90896106e-01 2.11595148e-01 -5.44070423e-01
6.87706321e-02 1.07841246e-01 -5.00428021e-01 -3.49778086e-02
6.15558445e-01 5.43237388e-01 -1.20594390e-02 -5.76347470e-01
-7.01806486e-01 4.09946412e-01 1.83474004e-01 9.52105671e-02
6.60609722e-01 2.89398786e-02 -1.76516008e-02 1.59125328e-02
1.15768075e+00 2.44402632e-01 -1.11760020e+00 -2.78713852e-02
-3.42000388e-02 -5.24830461e-01 -3.57483923e-01 -1.26419318e+00
-1.33599794e+00 9.61299002e-01 6.73969269e-01 5.96540570e-01
1.12753153e+00 2.97110647e-01 5.21635950e-01 3.29822332e-01
1.47603184e-01 -8.00862253e-01 6.17217004e-01 1.75552815e-01
1.55280268e+00 -1.42754018e+00 1.14047758e-01 -6.11620605e-01
-7.38798618e-01 8.66214693e-01 9.62421417e-01 -2.89794564e-01
4.21065658e-01 2.25600712e-02 8.62929165e-01 -8.98826897e-01
-5.99224925e-01 -2.16280296e-01 7.01997697e-01 1.34190810e+00
-1.82328731e-01 -8.95660892e-02 1.16316170e-01 2.86425501e-01
-1.65922701e-01 -2.71718770e-01 4.61994231e-01 9.03256059e-01
7.85798766e-03 -5.40934980e-01 -6.64891958e-01 2.90368378e-01
-5.34520149e-01 3.37658264e-02 -1.10546815e+00 1.00556588e+00
-2.15312257e-01 1.24036419e+00 8.31939727e-02 -2.24587977e-01
6.29523933e-01 1.00773826e-01 5.30682445e-01 -2.47907013e-01
-7.75923729e-01 -1.17088497e+00 1.41184986e-01 -9.53306317e-01
-5.37943959e-01 2.76868016e-01 -1.19649959e+00 -1.44195765e-01
1.62430964e-02 -1.86628345e-02 7.75504038e-02 6.48182213e-01
5.46451092e-01 6.20940149e-01 3.53999771e-02 -7.56922886e-02
-5.42250313e-02 -6.24049664e-01 -2.43704930e-01 9.80669558e-01
5.29721856e-01 -5.19929707e-01 -4.00334179e-01 1.28966585e-01] | [10.970978736877441, 1.947223424911499] |
fc08d343-f5fd-40f4-86ce-7034369845da | lifted-inference-with-linear-order-axiom | 2211.01164 | null | https://arxiv.org/abs/2211.01164v1 | https://arxiv.org/pdf/2211.01164v1.pdf | Lifted Inference with Linear Order Axiom | We consider the task of weighted first-order model counting (WFOMC) used for probabilistic inference in the area of statistical relational learning. Given a formula $\phi$, domain size $n$ and a pair of weight functions, what is the weighted sum of all models of $\phi$ over a domain of size $n$? It was shown that computing WFOMC of any logical sentence with at most two logical variables can be done in time polynomial in $n$. However, it was also shown that the task is $\texttt{#}P_1$-complete once we add the third variable, which inspired the search for extensions of the two-variable fragment that would still permit a running time polynomial in $n$. One of such extension is the two-variable fragment with counting quantifiers. In this paper, we prove that adding a linear order axiom (which forces one of the predicates in $\phi$ to introduce a linear ordering of the domain elements in each model of $\phi$) on top of the counting quantifiers still permits a computation time polynomial in the domain size. We present a new dynamic programming-based algorithm which can compute WFOMC with linear order in time polynomial in $n$, thus proving our primary claim. | ['Ondřej Kuželka', 'Jan Tóth'] | 2022-11-02 | null | null | null | null | ['relational-reasoning'] | ['natural-language-processing'] | [ 1.01908289e-01 5.93893409e-01 6.51001036e-02 -4.47947443e-01
-8.96083891e-01 -5.17625391e-01 3.57161015e-02 3.77621382e-01
-7.27960646e-01 6.67235196e-01 -6.32995129e-01 -9.24212337e-01
-3.31401616e-01 -1.59265530e+00 -8.49312961e-01 -6.28048182e-01
-5.87959230e-01 1.12403274e+00 6.86571956e-01 -2.79841442e-02
1.80799529e-01 2.51574367e-01 -1.69814658e+00 2.57179648e-01
3.68408322e-01 9.64895248e-01 6.95303530e-02 7.34981120e-01
-5.54638624e-01 7.14785516e-01 -1.81587949e-01 -7.13949919e-01
2.28599608e-01 -1.81429327e-01 -1.21359575e+00 -6.21809542e-01
4.50029373e-01 2.06210151e-01 -3.02872688e-01 1.55055356e+00
-1.32488132e-01 2.44259685e-01 3.24436784e-01 -1.46631670e+00
-6.87261671e-02 1.46017349e+00 -4.90016580e-01 1.37686282e-01
3.79581988e-01 -2.12610945e-01 1.56326735e+00 -1.44711629e-01
6.45361006e-01 1.54081583e+00 2.86048442e-01 5.71573377e-01
-1.63467407e+00 -8.38443696e-01 -1.04253106e-02 3.73849541e-01
-1.55336523e+00 2.15185389e-01 5.77358842e-01 -3.11527610e-01
1.40063691e+00 6.59676433e-01 1.96077526e-01 -6.44971281e-02
4.82931212e-02 2.33637363e-01 1.20056415e+00 -1.08733487e+00
3.46107036e-01 -8.56844038e-02 7.50534713e-01 1.31613433e+00
5.16729653e-01 -7.45130926e-02 -4.68448639e-01 -5.02622306e-01
9.64767560e-02 -3.85658860e-01 2.86283493e-01 8.93050507e-02
-6.68132067e-01 1.09215128e+00 -9.56307352e-02 3.08692127e-01
2.92140961e-01 7.27377415e-01 2.61131138e-01 3.46904933e-01
1.67268127e-01 2.75971860e-01 -8.60580385e-01 -2.70754122e-03
-6.50108039e-01 4.83604163e-01 1.07250094e+00 9.58859801e-01
9.18655634e-01 -4.03874934e-01 2.10781634e-01 1.52663976e-01
3.29580009e-01 9.44509506e-01 -2.51608163e-01 -1.35850656e+00
5.48774242e-01 7.12854922e-01 1.39008030e-01 -6.43349290e-01
-4.27075744e-01 3.44157070e-01 -4.08435374e-01 3.66245449e-01
8.32920909e-01 2.26768881e-01 -5.00720680e-01 2.37592411e+00
2.93786019e-01 -1.36763945e-01 -3.46829504e-01 2.12853819e-01
1.13170192e-01 6.46951318e-01 3.08773458e-01 -4.69267070e-01
1.73641372e+00 -6.55408502e-02 -4.56029564e-01 -1.40481055e-01
1.03266180e+00 -2.96719640e-01 8.04762900e-01 6.15135431e-01
-1.37374616e+00 3.61947878e-03 -1.02962923e+00 -2.65089095e-01
-2.85164416e-01 -6.27931893e-01 1.11573398e+00 6.52617037e-01
-8.55705440e-01 4.26344782e-01 -8.24685991e-01 3.50575149e-01
1.21530471e-03 5.85245073e-01 -4.16583121e-01 -4.93341416e-01
-1.49200761e+00 1.01638985e+00 3.78114223e-01 -2.32814535e-01
-4.87413079e-01 -4.65968847e-01 -1.05870056e+00 2.24100202e-01
8.21405709e-01 -3.12338978e-01 1.13535643e+00 -1.82724431e-01
-8.38212252e-01 1.03616607e+00 -8.26888263e-01 -7.58154213e-01
2.01869741e-01 4.85599190e-01 -2.45174140e-01 7.43520632e-02
6.74323812e-02 1.69380214e-02 8.50034058e-02 -8.20266187e-01
-9.19369817e-01 -8.48530650e-01 8.17670941e-01 -4.50425863e-01
4.76486772e-01 3.21695149e-01 -4.28974964e-02 3.96302193e-01
5.08440852e-01 -8.57837379e-01 -4.57734078e-01 -2.46652693e-01
-1.77979752e-01 -7.73492873e-01 2.04399228e-02 -1.04213089e-01
1.57016540e+00 -1.97228539e+00 5.68182096e-02 6.73380554e-01
6.38845682e-01 -2.25363389e-01 6.35461509e-02 1.68344423e-01
-6.57761768e-02 5.30114353e-01 -4.59403902e-01 -1.67854518e-01
5.33205152e-01 4.14402395e-01 -7.96681717e-02 3.47238004e-01
-7.94821158e-02 4.93789047e-01 -6.28506422e-01 -8.62794042e-01
-1.45532563e-01 -2.29888424e-01 -7.66845703e-01 -1.36897907e-01
-9.37579513e-01 -6.44000471e-01 -2.09852323e-01 1.22497760e-01
1.14100647e+00 -2.35643193e-01 6.07218981e-01 1.37008026e-01
-8.55393335e-02 4.63638246e-01 -1.74156392e+00 1.07456779e+00
-3.58781189e-01 1.11403503e-01 4.75340262e-02 -7.60929942e-01
5.34804821e-01 -2.37727095e-03 2.59313911e-01 -4.93471742e-01
2.06634417e-01 1.66613534e-01 2.50859857e-01 -3.80510747e-01
3.55937481e-01 -1.09772933e+00 -6.37938380e-01 6.47250772e-01
-7.96662178e-03 -2.14395389e-01 6.22878194e-01 4.69482064e-01
1.59205496e+00 -4.81011778e-01 -4.74056811e-04 -4.44266886e-01
7.93417573e-01 -1.28716575e-02 8.99800062e-01 9.19770241e-01
-8.09618384e-02 -2.15473488e-01 1.12881935e+00 -4.54632849e-01
-6.57574296e-01 -1.13273036e+00 -1.08099252e-03 1.38377309e+00
3.81717756e-02 -8.51159155e-01 -7.65675545e-01 -4.15764719e-01
-3.91045399e-02 9.75581408e-01 -6.71292126e-01 5.09674381e-03
-8.30077171e-01 -5.09525418e-01 5.22743940e-01 4.14973974e-01
3.38875204e-01 -8.78163636e-01 -7.55345821e-01 2.95292675e-01
-2.57385075e-01 -1.07191777e+00 -2.06828251e-01 8.44718456e-01
-7.60739625e-01 -1.55170810e+00 5.19043326e-01 -6.88841283e-01
6.00520492e-01 -2.44301543e-01 1.19062865e+00 2.83049971e-01
-1.38638899e-01 2.13480908e-02 1.45760551e-01 -5.26795030e-01
-2.40684316e-01 -2.16001242e-01 -1.60139233e-01 -6.68856144e-01
1.01162195e+00 -2.55581588e-01 8.84813368e-02 -2.23645449e-01
-9.77334857e-01 -2.39476085e-01 -1.78170919e-01 6.58216476e-01
7.78856218e-01 9.31039035e-01 -3.78776014e-01 -1.50076795e+00
1.81680322e-01 -1.85392573e-01 -1.39349759e+00 5.18850923e-01
-4.06476259e-01 6.91820621e-01 7.92205691e-01 -1.47354931e-01
-5.45049906e-01 7.43374377e-02 1.79506853e-01 2.17353299e-01
4.12815273e-01 1.07193232e+00 -5.40525317e-01 2.74661154e-01
5.08887470e-01 -6.45249113e-02 -1.03465058e-01 -1.03998534e-01
2.24483147e-01 -1.52509436e-01 3.25963676e-01 -1.23272336e+00
7.78395712e-01 3.65506977e-01 1.02091050e+00 -3.46839994e-01
-8.49454165e-01 -3.70679945e-02 -3.60603064e-01 3.09038311e-01
4.52569187e-01 -3.96175236e-01 -1.54431820e+00 1.25264779e-01
-1.14549410e+00 -5.64872384e-01 -3.66853893e-01 1.70547917e-01
-5.28013885e-01 4.44734603e-01 -4.17151511e-01 -1.53973401e+00
4.70697545e-02 -9.56534147e-01 5.02453685e-01 -2.44401589e-01
-1.49464011e-01 -6.67483568e-01 6.08486719e-02 1.63713694e-01
4.66671735e-02 9.37044621e-02 1.83414590e+00 -6.33248091e-01
-5.61124027e-01 -5.34025729e-01 -2.06624046e-01 2.26160228e-01
-7.02613354e-01 2.03940235e-02 -6.13215685e-01 1.99728981e-01
1.69621170e-01 -6.38756528e-02 5.58716118e-01 3.88548583e-01
1.24859178e+00 -3.28090608e-01 -2.55238473e-01 -1.94994956e-01
1.76478457e+00 1.04417488e-01 7.28056252e-01 6.26175776e-02
6.71385601e-02 5.52316666e-01 4.19026852e-01 2.99110830e-01
5.46479106e-01 3.95784825e-01 1.30345076e-01 6.64207876e-01
5.14388144e-01 -4.98028956e-02 -6.08114004e-02 4.00605917e-01
-2.68624127e-01 1.53554156e-01 -1.06775856e+00 3.55310649e-01
-1.53729188e+00 -1.17430151e+00 -4.14718509e-01 2.48080659e+00
1.26743698e+00 6.96736157e-01 -1.24246092e-03 5.63797951e-01
6.27740562e-01 -1.83154762e-01 -5.69031797e-02 -8.35186601e-01
1.91762149e-02 1.33544624e+00 8.56170058e-01 1.44794118e+00
-7.33290732e-01 7.91378558e-01 5.78591204e+00 8.59289348e-01
-4.57543254e-01 3.07755142e-01 2.48781562e-01 -1.19673856e-01
-6.57988787e-01 4.51792777e-01 -7.92511523e-01 4.71780717e-01
1.31025422e+00 -3.89399678e-01 8.34541798e-01 1.02848971e+00
-2.83350438e-01 -6.12647831e-01 -1.39947498e+00 6.35137260e-01
-2.93430001e-01 -9.03663933e-01 -4.57866073e-01 1.79793149e-01
2.60394990e-01 -4.50900227e-01 -2.12395996e-01 5.48167646e-01
8.79533410e-01 -1.04950237e+00 7.53287315e-01 2.48208925e-01
8.71069074e-01 -1.14322627e+00 8.50128174e-01 5.12392342e-01
-1.31543243e+00 -1.73534900e-01 -5.26027203e-01 -4.75474000e-01
1.67581528e-01 7.20742166e-01 -2.89003104e-01 1.80389062e-01
4.64390576e-01 -5.92990100e-01 -1.52937219e-01 4.98151422e-01
-1.87181562e-01 4.86249030e-01 -9.40173149e-01 -1.58576757e-01
-1.24973141e-01 -6.46989152e-04 3.45624909e-02 1.11605477e+00
1.40854359e-01 7.22898662e-01 -5.56728393e-02 8.08611512e-01
-2.03622237e-01 -4.97148007e-01 -2.52204359e-01 1.83201253e-01
7.10143447e-01 7.40895510e-01 -4.79201376e-01 -5.47160923e-01
-2.65458614e-01 8.52640197e-02 3.63105983e-01 -1.73407406e-01
-9.11521614e-01 -5.62817931e-01 3.55407357e-01 4.20796961e-01
2.85220832e-01 -3.39730620e-01 -5.10466933e-01 -8.27827930e-01
1.31900802e-01 -5.41933358e-01 8.28217745e-01 -3.29115659e-01
-1.16895998e+00 4.40415949e-01 3.71868700e-01 -1.97700486e-01
-9.38910991e-02 -1.06782377e+00 -2.62112111e-01 1.31353331e+00
-1.14614475e+00 -7.91848481e-01 4.23221707e-01 5.76133490e-01
-2.41566643e-01 4.32010412e-01 1.32992005e+00 -4.78858240e-02
-1.86281011e-01 6.73509777e-01 -5.61110616e-01 -1.17095947e-01
2.85740849e-02 -1.64132893e+00 -1.53275058e-01 9.94232774e-01
-2.17154827e-02 1.04210305e+00 1.14313328e+00 -3.92946959e-01
-1.66672766e+00 -6.86960161e-01 1.80425775e+00 -4.40554887e-01
7.70912528e-01 -4.26846743e-01 -7.24259496e-01 8.14418793e-01
-2.64179379e-01 2.23241314e-01 7.62577236e-01 6.24773383e-01
-1.02082944e+00 -3.07194859e-01 -1.55869889e+00 3.61368269e-01
9.44560111e-01 -8.32645118e-01 -6.91430986e-01 2.77952868e-02
7.53401279e-01 -2.62035131e-01 -8.28816354e-01 3.74767572e-01
6.41153395e-01 -4.40360248e-01 5.64942002e-01 -7.55415916e-01
1.87343702e-01 -6.26666605e-01 -7.61770189e-01 -2.57928163e-01
-3.28161240e-01 -3.68938953e-01 -3.70471150e-01 7.71237552e-01
7.28092074e-01 -8.18574071e-01 7.83392072e-01 1.48904824e+00
3.45471531e-01 -4.95563179e-01 -1.54032707e+00 -6.11355841e-01
5.78410685e-01 -1.22017133e+00 5.68687499e-01 4.11515474e-01
6.42275631e-01 2.85869241e-01 2.01303840e-01 2.49775156e-01
6.70304418e-01 1.89726010e-01 4.34060574e-01 -1.58913362e+00
-6.56326532e-01 -2.58739293e-01 -4.66601223e-01 -4.68248099e-01
4.20059830e-01 -1.10838461e+00 2.18068212e-01 -1.19274056e+00
5.70805073e-01 -7.38991737e-01 -3.33768427e-01 7.57343233e-01
1.62780300e-01 -2.76269466e-01 2.80949056e-01 -4.58207816e-01
-8.08544338e-01 -2.62861490e-01 7.84742653e-01 -3.36271673e-01
4.34473991e-01 -1.40479743e-01 -8.10179293e-01 8.15029681e-01
4.50299114e-01 -8.37229013e-01 -6.57231882e-02 -1.17837884e-01
9.08492386e-01 3.62602860e-01 1.01738900e-01 -7.25124061e-01
3.79133731e-01 -5.24556994e-01 -4.86728221e-01 -4.62146193e-01
1.43877819e-01 -9.33119833e-01 1.88176170e-01 6.35337293e-01
-5.32011569e-01 2.63823509e-01 2.43876636e-01 2.97938734e-01
9.86107886e-02 -8.25715482e-01 8.33677471e-01 -2.94195592e-01
-2.98401505e-01 -2.17676647e-02 -4.23763037e-01 2.53874749e-01
7.40998864e-01 4.30750221e-01 -3.44908595e-01 1.51875034e-01
-7.76585042e-01 7.11740106e-02 2.11041555e-01 -6.01392508e-01
3.30848157e-01 -9.21485245e-01 -3.00482094e-01 -3.22450437e-02
2.31192745e-02 3.67085606e-01 4.88182276e-01 4.94874686e-01
-5.18648922e-01 5.08421779e-01 3.27482045e-01 1.46238292e-02
-1.46016955e+00 8.43528867e-01 1.06905237e-01 -9.08543646e-01
7.96998106e-03 1.15871572e+00 -1.35689318e-01 -2.91390806e-01
1.53760657e-01 -7.21652746e-01 4.82291281e-01 -3.59727591e-01
8.81580949e-01 4.25086766e-01 -9.88895074e-02 -1.96224228e-01
-6.27579570e-01 3.30644101e-01 3.10012978e-02 -6.04395628e-01
1.16719365e+00 1.54600382e-01 -9.08560634e-01 4.63489622e-01
1.04116321e+00 1.82663962e-01 -4.57566112e-01 -4.51409876e-01
3.36523205e-01 -2.31542245e-01 -3.75723422e-01 -7.71304727e-01
-8.21799874e-01 7.57420897e-01 3.76894802e-01 5.30864239e-01
1.05043495e+00 4.96053010e-01 4.32927817e-01 7.23961890e-01
1.32592964e+00 -1.12500858e+00 -6.74574733e-01 8.86375248e-01
3.22636873e-01 -7.54744768e-01 8.30592215e-02 -4.69924808e-01
-1.10266551e-01 9.82911885e-01 5.03508806e-01 -1.36533782e-01
8.44920456e-01 7.53879905e-01 -6.44320190e-01 -3.04975927e-01
-1.00911796e+00 -2.32945308e-01 -3.61587942e-01 2.76529819e-01
3.27170312e-01 5.44871867e-01 -7.53596544e-01 1.14308274e+00
-5.09116054e-01 6.52764887e-02 3.34096074e-01 1.02045465e+00
-7.97699809e-01 -1.57452071e+00 -2.93425858e-01 5.53416669e-01
-7.82089710e-01 -2.84706652e-01 1.96872815e-01 7.69640028e-01
5.61009407e-01 1.01124477e+00 -1.52120903e-01 -2.05966443e-01
5.95661215e-02 5.29623508e-01 1.08221614e+00 -7.43513167e-01
-8.31187144e-02 -3.93139601e-01 1.79091185e-01 -3.82613152e-01
-2.01112941e-01 -6.72655761e-01 -1.72845995e+00 -9.66219842e-01
-4.49640542e-01 4.67035085e-01 4.31475550e-01 1.08824050e+00
-2.86427617e-01 -3.94164696e-02 2.30744764e-01 1.63407013e-01
-6.54402673e-01 -9.27564681e-01 -1.10142779e+00 -9.55310091e-02
-2.18594953e-01 -5.75299859e-01 -6.58297122e-01 -2.89682169e-02] | [8.62928581237793, 6.730592727661133] |
30a62a6c-355d-48e6-80ad-ed6da199f3ee | quantitative-analysis-of-automatic-image | 1701.01480 | null | http://arxiv.org/abs/1701.01480v1 | http://arxiv.org/pdf/1701.01480v1.pdf | Quantitative Analysis of Automatic Image Cropping Algorithms: A Dataset and Comparative Study | Automatic photo cropping is an important tool for improving visual quality of
digital photos without resorting to tedious manual selection. Traditionally,
photo cropping is accomplished by determining the best proposal window through
visual quality assessment or saliency detection. In essence, the performance of
an image cropper highly depends on the ability to correctly rank a number of
visually similar proposal windows. Despite the ranking nature of automatic
photo cropping, little attention has been paid to learning-to-rank algorithms
in tackling such a problem. In this work, we conduct an extensive study on
traditional approaches as well as ranking-based croppers trained on various
image features. In addition, a new dataset consisting of high quality cropping
and pairwise ranking annotations is presented to evaluate the performance of
various baselines. The experimental results on the new dataset provide useful
insights into the design of better photo cropping algorithms. | ['Bing-Yu Chen', 'Hwann-Tzong Chen', 'Yu-Chen Tsai', 'Kai-Han Chang', 'Tzu-Wei Huang', 'Yi-Ling Chen'] | 2017-01-05 | null | null | null | null | ['image-cropping'] | ['computer-vision'] | [ 5.72594106e-01 -2.59883732e-01 -3.06216598e-01 -3.43470246e-01
-1.26558948e+00 -6.55146182e-01 3.88260752e-01 2.75562525e-01
-2.20122010e-01 4.09814179e-01 3.05772781e-01 -1.59166716e-02
1.31371081e-01 -4.72391456e-01 -7.41532683e-01 -5.20234942e-01
7.92816207e-02 -1.75394982e-01 4.62930083e-01 2.29439717e-02
6.90602481e-01 1.17093593e-01 -1.77476776e+00 3.24776828e-01
1.09374988e+00 8.25577199e-01 6.12184346e-01 6.30503714e-01
3.35026145e-01 3.04088712e-01 -4.15465444e-01 -4.93768394e-01
4.09892678e-01 -3.78835022e-01 -4.12249982e-01 3.52517664e-01
1.11984313e+00 -4.24181432e-01 1.15740158e-01 1.24433625e+00
4.90577221e-01 2.34240368e-02 7.43276715e-01 -1.30433786e+00
-8.52018356e-01 4.35827196e-01 -8.80844355e-01 2.40176171e-01
4.55259353e-01 1.09282725e-01 1.42462277e+00 -1.31364274e+00
5.71457982e-01 1.05201459e+00 4.31704372e-01 1.26339018e-01
-1.18406892e+00 -5.85542977e-01 7.02931955e-02 3.11841577e-01
-1.29258549e+00 -4.57773447e-01 8.08788121e-01 -3.95511568e-01
3.06000769e-01 2.56403297e-01 4.75722224e-01 5.06480575e-01
-9.23789069e-02 9.06394124e-01 1.18497002e+00 -5.88043571e-01
1.06412977e-01 9.98732969e-02 3.15701030e-02 6.74188554e-01
4.13048387e-01 9.86739546e-02 -7.23574400e-01 -1.17240906e-01
5.82315743e-01 -1.43149689e-01 -2.77433604e-01 -6.94935799e-01
-1.20757198e+00 4.85450566e-01 8.47226202e-01 7.76420981e-02
-3.61688048e-01 1.68864042e-01 8.42572972e-02 -1.37299061e-01
3.25909048e-01 7.40770876e-01 -1.80425316e-01 2.25838706e-01
-1.41580832e+00 3.40648353e-01 2.09118724e-02 7.66148865e-01
1.02983725e+00 -1.75948590e-01 -5.07497966e-01 8.58891070e-01
1.06881052e-01 6.88419521e-01 1.95890322e-01 -1.24226201e+00
5.10922492e-01 6.35380328e-01 4.16757643e-01 -1.20385754e+00
-1.27845660e-01 -5.96905611e-02 -6.06750250e-01 4.79127377e-01
2.55810022e-01 1.51686624e-01 -1.07796121e+00 1.43530750e+00
8.85122567e-02 -1.62910834e-01 -4.05824453e-01 1.03438485e+00
6.00963116e-01 5.27392030e-01 1.65669203e-01 -1.34087026e-01
1.30728292e+00 -1.02954888e+00 -4.97521788e-01 -3.23951900e-01
1.16661541e-01 -1.12519801e+00 1.43577981e+00 3.43049049e-01
-1.15329158e+00 -6.96722507e-01 -1.25980222e+00 -2.97622979e-01
-2.80411780e-01 5.88974118e-01 3.78718823e-01 5.36305189e-01
-1.22564697e+00 4.80496436e-01 -3.75981808e-01 -4.13089991e-01
5.06240785e-01 -4.43121651e-03 -2.60839880e-01 -4.03257787e-01
-8.27216983e-01 8.03996086e-01 3.08778167e-01 -2.63478190e-01
-9.58192766e-01 -5.73445261e-01 -6.75946116e-01 8.84916782e-02
2.62732029e-01 -4.60317463e-01 1.26075244e+00 -1.13387871e+00
-9.17622268e-01 9.76809621e-01 -4.50147629e-01 -4.57437098e-01
4.71351624e-01 -4.20488596e-01 -8.64153504e-02 3.68593961e-01
4.13908750e-01 1.11248088e+00 1.11330581e+00 -1.56921732e+00
-9.15753543e-01 -1.38045236e-01 4.76724990e-02 5.02551496e-01
-4.36045617e-01 2.60040369e-02 -6.24325037e-01 -8.85365963e-01
5.98033033e-02 -1.00101209e+00 -1.78789705e-01 1.26095667e-01
-4.09749061e-01 -1.65309355e-01 5.36985695e-01 -6.80871606e-01
1.28924513e+00 -2.19674969e+00 -5.92123345e-02 1.43759012e-01
6.54562041e-02 2.86363602e-01 -3.61833096e-01 4.99366790e-01
-1.31786279e-02 3.44479531e-01 -2.21932396e-01 -1.51914701e-01
-8.43113065e-02 -2.96799630e-01 -3.15230638e-01 2.04793751e-01
4.27257597e-01 8.74344826e-01 -1.14234602e+00 -7.84283996e-01
3.65398109e-01 1.15524746e-01 -3.43706399e-01 1.91051438e-01
-2.07425103e-01 -1.11673124e-01 -2.03855336e-01 9.74262536e-01
7.14962065e-01 -4.02915120e-01 1.42264411e-01 -4.81275111e-01
-1.64227948e-01 3.92471580e-03 -8.46218765e-01 1.56580842e+00
-2.20657632e-01 8.42006385e-01 -3.95813376e-01 -3.73587400e-01
6.54372215e-01 -4.46687415e-02 3.71922970e-01 -8.07890117e-01
-2.39550829e-01 6.39630258e-02 -3.59251946e-01 -3.52628499e-01
1.11241806e+00 2.81009704e-01 -3.83057967e-02 4.02963609e-01
-4.03882921e-01 -2.62091100e-01 5.82212329e-01 3.83652002e-01
7.85727203e-01 2.67985016e-01 4.59894598e-01 -3.74723151e-02
2.27825120e-01 3.90973777e-01 4.44889963e-01 5.90930104e-01
-4.24383372e-01 1.15770054e+00 4.01604883e-02 -1.87289670e-01
-1.26046574e+00 -1.11557436e+00 1.76208057e-02 1.33953786e+00
6.69765770e-01 -4.91721272e-01 -8.26107502e-01 -6.69395626e-01
-9.55421850e-02 4.51087922e-01 -6.74994946e-01 1.85655449e-02
-3.90773565e-01 -7.24265993e-01 8.32609981e-02 4.63662505e-01
6.04248226e-01 -1.09912860e+00 -6.84517503e-01 -1.80176020e-01
-6.04105115e-01 -9.73309577e-01 -8.91709566e-01 -3.45497042e-01
-7.27793813e-01 -1.41868544e+00 -9.94263887e-01 -9.24389660e-01
8.06988180e-01 1.31494355e+00 1.30267823e+00 1.70164376e-01
-2.88460553e-01 1.87209800e-01 -6.50426209e-01 -3.01914334e-01
-2.44089663e-01 1.78983301e-01 -1.27454132e-01 -1.57705843e-01
2.34701619e-01 -1.62251383e-01 -9.97209728e-01 5.97006857e-01
-1.04595411e+00 2.20346019e-01 1.05959272e+00 5.81234634e-01
6.25529349e-01 1.33490533e-01 3.62317145e-01 -5.27456343e-01
5.89783609e-01 1.57075115e-02 -5.87289214e-01 6.39100969e-01
-6.63314521e-01 2.41988562e-02 8.33774656e-02 -3.18382621e-01
-8.92350376e-01 3.45962137e-01 5.12525976e-01 -3.28984171e-01
2.32079342e-01 2.96746939e-01 -6.29306883e-02 -1.98282212e-01
7.57935762e-01 7.34472834e-03 -4.76749957e-01 -3.24914724e-01
7.20718443e-01 3.76419663e-01 5.14495254e-01 -1.18643612e-01
1.11841536e+00 3.40490729e-01 -1.65706128e-01 -7.98085213e-01
-8.31336081e-01 -5.52760899e-01 -6.45356953e-01 -4.49755043e-01
8.45390797e-01 -9.65370238e-01 -1.95375413e-01 2.41353109e-01
-8.27449501e-01 -1.49989575e-01 2.93342695e-02 8.31888616e-03
-3.16033840e-01 7.11389780e-01 3.77421826e-02 -7.20800877e-01
-3.47229749e-01 -9.69909489e-01 1.35357904e+00 4.20067400e-01
-1.72892705e-01 -2.88714498e-01 -8.40332918e-03 5.38175464e-01
1.91961139e-01 1.13373347e-01 7.20476985e-01 2.23953635e-01
-8.99364889e-01 -3.28164130e-01 -7.57737517e-01 2.42776006e-01
8.44940096e-02 1.54442132e-01 -9.98713970e-01 -2.11699083e-01
-8.69796813e-01 -3.89401823e-01 1.15408015e+00 5.91564298e-01
1.02249932e+00 5.46981804e-02 -4.17030245e-01 1.84629515e-01
1.46638477e+00 3.32116000e-02 7.10619926e-01 3.37288260e-01
6.87552571e-01 5.90876281e-01 1.30415845e+00 2.26163283e-01
3.14415842e-01 6.73813760e-01 5.25283694e-01 -2.82260299e-01
-3.85255039e-01 -6.01730704e-01 3.51682335e-01 1.88394144e-01
4.93715219e-02 -2.46621251e-01 -8.05653095e-01 8.43669176e-01
-1.68292999e+00 -9.16096628e-01 7.64488801e-02 2.45479035e+00
8.14257264e-01 -4.19844985e-02 2.56852239e-01 3.45788002e-01
1.00827014e+00 2.63068527e-01 -4.39328402e-01 1.65447742e-01
-2.39424482e-01 -2.28539795e-01 5.34657419e-01 2.31831625e-01
-1.35695004e+00 1.05384195e+00 6.97675276e+00 7.78717577e-01
-9.14400101e-01 -1.91031992e-01 7.67000616e-01 5.88622279e-02
-7.54846483e-02 8.53190422e-02 -7.06918180e-01 4.03234243e-01
1.60737768e-01 -1.59527779e-01 3.28713328e-01 7.05605149e-01
4.32339728e-01 -7.07746863e-01 -8.94500732e-01 1.04501724e+00
2.58542657e-01 -1.12366104e+00 1.72127932e-01 -1.10942282e-01
1.09109855e+00 -1.33195147e-01 3.89286160e-01 -1.45695522e-01
3.55018467e-01 -8.82945716e-01 7.87415087e-01 3.64359707e-01
8.15933645e-01 -5.40035188e-01 4.15191352e-01 -1.70016393e-01
-1.22724938e+00 -1.22349016e-01 -3.10106575e-01 1.66565284e-01
1.51288271e-01 6.05512619e-01 -8.97978842e-01 2.86347121e-01
8.76379073e-01 6.79674923e-01 -1.13341427e+00 1.54057932e+00
-3.48578066e-01 4.22934353e-01 7.71895275e-02 8.75605568e-02
-2.73376033e-02 1.26282439e-01 3.98799121e-01 9.32339728e-01
4.24066782e-01 -1.07845068e-01 2.60786206e-01 4.18582290e-01
-2.07943022e-01 1.50703520e-01 -3.65547746e-01 -1.32780865e-01
5.76526582e-01 1.44614708e+00 -8.02338541e-01 -2.76027650e-01
-2.45376289e-01 9.12561357e-01 1.74049616e-01 3.23195159e-01
-6.62503898e-01 -2.31372282e-01 4.13475007e-01 3.11269850e-01
3.53778124e-01 -1.87488094e-01 -3.45581323e-01 -1.03128350e+00
1.88311413e-01 -8.19176733e-01 3.07523906e-01 -1.37167370e+00
-1.19111419e+00 4.24598575e-01 -7.64776692e-02 -1.68881762e+00
1.24548949e-01 -3.36555213e-01 -5.57293594e-01 6.25375390e-01
-1.78065741e+00 -1.21991372e+00 -7.10925877e-01 4.33414243e-02
7.15462327e-01 2.81373888e-01 4.20640141e-01 9.63432640e-02
-3.17072153e-01 4.31650817e-01 3.01831197e-02 -2.93209106e-02
1.22819602e+00 -1.33068240e+00 4.05351341e-01 1.20286739e+00
4.08150315e-01 2.74163157e-01 9.16289568e-01 -6.20435774e-01
-8.37304771e-01 -1.17763388e+00 8.86230290e-01 -4.67693567e-01
1.76384822e-01 3.74490395e-02 -8.11182857e-01 1.07700825e-01
4.56367791e-01 -2.74638921e-01 3.77150446e-01 -6.68297857e-02
-2.70683736e-01 -2.52980232e-01 -8.44387531e-01 6.72379196e-01
9.07943308e-01 -3.53224277e-01 -4.27413464e-01 2.33343109e-01
5.46558678e-01 -1.85636729e-01 -3.82150263e-01 3.95506978e-01
7.03347325e-01 -9.78405237e-01 1.22951412e+00 -4.69632223e-02
6.30268693e-01 -6.11708224e-01 -8.18381011e-02 -1.25257373e+00
-4.51819539e-01 -4.04147565e-01 4.23087806e-01 1.26646709e+00
2.99683928e-01 1.86639488e-01 7.64161050e-01 4.43146408e-01
2.11526886e-01 -3.04350704e-01 -3.07346016e-01 -4.80047137e-01
-3.77751201e-01 -1.83640227e-01 4.65685248e-01 7.16258347e-01
-1.99466616e-01 3.05491984e-01 -5.94008565e-01 3.54945391e-01
7.24633336e-01 4.14945841e-01 8.93151343e-01 -1.07892430e+00
1.37757674e-01 -5.20061791e-01 -3.05671662e-01 -6.40248477e-01
-3.52530181e-01 -5.01574039e-01 4.80370045e-01 -1.69769144e+00
5.80846667e-01 -2.69814789e-01 -4.18498129e-01 3.64376277e-01
-8.07479858e-01 8.98207247e-01 3.68540078e-01 3.65642428e-01
-1.00800550e+00 3.97432864e-01 1.09094381e+00 -2.47270972e-01
-1.90248415e-01 -1.82450414e-01 -7.47397006e-01 6.14829600e-01
8.79068017e-01 -3.89746368e-01 -4.28960443e-01 -4.07303214e-01
5.08839965e-01 -1.97459593e-01 5.38680851e-01 -1.02961159e+00
6.15876541e-02 -2.44242430e-01 5.54911077e-01 -9.41057026e-01
2.56388694e-01 -5.08851171e-01 -2.72217214e-01 3.64646241e-02
-4.67493594e-01 4.06030327e-01 -2.12903041e-02 7.07480073e-01
-1.79697648e-01 -1.69850186e-01 8.54632378e-01 -5.99300750e-02
-1.21333432e+00 7.93465525e-02 -1.44855231e-01 1.41005576e-01
1.00998187e+00 -1.55663669e-01 -2.64534533e-01 -4.22458529e-01
-4.11978662e-01 2.79203564e-01 9.78806853e-01 6.04472160e-01
7.90914595e-01 -1.33081126e+00 -5.67309320e-01 -1.69156238e-01
5.53295553e-01 -3.85033697e-01 2.12932825e-01 5.06936669e-01
-4.33459908e-01 2.42870480e-01 -2.99467772e-01 -5.07422149e-01
-1.81767869e+00 5.67346513e-01 -3.03409062e-02 -9.87586081e-02
-1.70843139e-01 6.64166808e-01 3.29198301e-01 3.08503151e-01
2.36172915e-01 -2.55085588e-01 -3.73661935e-01 1.76782921e-01
6.44103646e-01 4.21629846e-01 -5.53094260e-02 -6.59504771e-01
-3.01762164e-01 7.33631134e-01 -6.13033138e-02 -2.74238378e-01
9.80853856e-01 -5.01892149e-01 1.12606362e-01 1.72843024e-01
7.92734742e-01 -3.77682708e-02 -1.24978220e+00 -2.64426142e-01
1.75869420e-01 -9.80629861e-01 2.20624477e-01 -9.78684068e-01
-8.99435163e-01 8.20192099e-01 8.64154398e-01 1.49295807e-01
1.32715380e+00 -2.00974673e-01 5.31887054e-01 1.54251218e-01
3.09605926e-01 -1.16376746e+00 5.47469378e-01 -4.91821617e-02
9.60257888e-01 -1.80230498e+00 3.99712950e-01 -4.67982054e-01
-8.75140190e-01 7.28543282e-01 5.45895159e-01 -1.67677060e-01
2.68676281e-01 -3.05058628e-01 2.70923406e-01 8.37930106e-03
-4.75933760e-01 -6.39829814e-01 7.89289355e-01 6.60251021e-01
4.28548485e-01 -9.09083802e-03 -3.29596132e-01 4.50187400e-02
-1.53607912e-02 7.17781335e-02 5.11366427e-01 7.05940366e-01
-7.48111904e-01 -1.00855947e+00 -4.20597047e-01 3.31061214e-01
-3.32349151e-01 -2.96178937e-01 -7.27520406e-01 5.09967923e-01
5.54777980e-02 1.11323464e+00 -1.48602650e-01 -2.89142668e-01
1.31247848e-01 -1.72199622e-01 3.63743603e-01 -5.20377457e-01
-3.00046593e-01 7.02452287e-02 -2.19924953e-02 -5.05957544e-01
-6.47961974e-01 -8.11870813e-01 -7.59477794e-01 7.74133280e-02
-5.19958556e-01 -8.77448171e-02 5.44092357e-01 5.25330663e-01
3.55232090e-01 2.74545342e-01 5.99409401e-01 -1.15777338e+00
-1.96464762e-01 -8.80246222e-01 -4.05639946e-01 6.54207826e-01
2.47480944e-01 -6.94152415e-01 -8.79557803e-02 3.07972401e-01] | [11.252705574035645, -1.061044454574585] |
8959aade-cae4-425b-9b15-4538aa6a1438 | vnhsge-vietnamese-high-school-graduation | 2305.12199 | null | https://arxiv.org/abs/2305.12199v1 | https://arxiv.org/pdf/2305.12199v1.pdf | VNHSGE: VietNamese High School Graduation Examination Dataset for Large Language Models | The VNHSGE (VietNamese High School Graduation Examination) dataset, developed exclusively for evaluating large language models (LLMs), is introduced in this article. The dataset, which covers nine subjects, was generated from the Vietnamese National High School Graduation Examination and comparable tests. 300 literary essays have been included, and there are over 19,000 multiple-choice questions on a range of topics. The dataset assesses LLMs in multitasking situations such as question answering, text generation, reading comprehension, visual question answering, and more by including both textual data and accompanying images. Using ChatGPT and BingChat, we evaluated LLMs on the VNHSGE dataset and contrasted their performance with that of Vietnamese students to see how well they performed. The results show that ChatGPT and BingChat both perform at a human level in a number of areas, including literature, English, history, geography, and civics education. They still have space to grow, though, especially in the areas of mathematics, physics, chemistry, and biology. The VNHSGE dataset seeks to provide an adequate benchmark for assessing the abilities of LLMs with its wide-ranging coverage and variety of activities. We intend to promote future developments in the creation of LLMs by making this dataset available to the scientific community, especially in resolving LLMs' limits in disciplines involving mathematics and the natural sciences. | ['Nguyen Hong-Phuoc', 'Nguyen Thi-My-Thanh', 'Nguyen Van-Tien', 'Ngo Bac-Bien', 'Phan Xuan-Dung', 'Vo The-Duy', 'Le Ngoc-Bich', 'Dao Xuan-Quy'] | 2023-05-20 | null | null | null | null | ['reading-comprehension', 'question-rewriting'] | ['natural-language-processing', 'natural-language-processing'] | [-3.63778859e-01 1.24361344e-01 3.47656086e-02 1.58325970e-01
-1.10334456e+00 -8.36572409e-01 7.99732566e-01 5.89137554e-01
-6.14327133e-01 8.65414619e-01 2.53343791e-01 -8.54230165e-01
-3.05699557e-01 -8.11624706e-01 -4.94329184e-01 -2.08845004e-01
4.08599615e-01 4.17869121e-01 3.83093730e-02 -5.21941602e-01
7.28394389e-01 2.83300191e-01 -1.42222822e+00 2.22690925e-01
1.56245840e+00 2.64665455e-01 4.15766090e-01 7.83998668e-01
-4.79838461e-01 9.96914387e-01 -1.01368487e+00 -8.64528656e-01
-5.51079273e-01 -5.57938874e-01 -1.18105912e+00 -2.03492865e-01
9.04250920e-01 1.32162228e-01 -2.48242334e-01 7.34257579e-01
6.81532919e-01 4.64399904e-01 5.53465962e-01 -9.31949675e-01
-1.08861685e+00 5.54683626e-01 -1.48094445e-01 6.12192094e-01
8.61944079e-01 1.27801284e-01 1.01937175e+00 -1.22909307e+00
8.63572598e-01 1.34073794e+00 2.09966436e-01 2.22121730e-01
-7.73336470e-01 -5.89520097e-01 -1.34014785e-01 6.27170980e-01
-1.30617189e+00 -8.07892904e-02 4.62709635e-01 -6.88515961e-01
8.27020168e-01 1.84989855e-01 6.49263918e-01 1.16866732e+00
4.26364273e-01 7.57834971e-01 1.56535673e+00 -5.40558815e-01
3.68877910e-02 3.72011691e-01 1.74023122e-01 8.31548572e-01
4.26077843e-02 -3.50173235e-01 -9.03181732e-01 1.31116733e-01
6.29637599e-01 -6.24979317e-01 -1.87841132e-01 2.61812717e-01
-1.65930855e+00 8.11530530e-01 -2.16175988e-02 5.74367166e-01
4.41345833e-02 -3.46269757e-01 1.40289173e-01 6.17625177e-01
2.90266395e-01 7.38668382e-01 -2.72786319e-01 -4.84142125e-01
-7.99110413e-01 5.00623405e-01 1.07344234e+00 8.56462002e-01
3.27046692e-01 -3.04940566e-02 -5.00057161e-01 9.80791509e-01
3.43333222e-02 6.92214012e-01 6.54478014e-01 -8.39915574e-01
9.18287218e-01 6.71118736e-01 -4.31941569e-01 -1.11446869e+00
-2.21528858e-01 -4.49885875e-01 -7.12256074e-01 1.09006822e-01
6.53139234e-01 -1.05127573e-01 -5.79975247e-01 1.53700256e+00
3.76504436e-02 -2.75939465e-01 1.03177801e-01 6.14353240e-01
1.75273323e+00 9.57913160e-01 4.83794183e-01 3.72874849e-02
1.54178429e+00 -9.77647781e-01 -7.85447061e-01 -3.04731101e-01
8.27163994e-01 -1.00993335e+00 1.56637728e+00 6.06407225e-01
-1.52991199e+00 -8.00473809e-01 -8.99839222e-01 -5.96168756e-01
-7.97143877e-01 7.51495734e-02 1.06319815e-01 6.61003113e-01
-1.08568275e+00 3.26264650e-01 -1.01945519e-01 -2.88938195e-01
1.17300458e-01 -2.45933771e-01 -3.49972725e-01 -2.51798868e-01
-1.48714364e+00 1.34754944e+00 2.67412335e-01 -2.94028163e-01
-6.02199376e-01 -6.63808227e-01 -1.01130629e+00 2.16736153e-01
2.68600732e-01 -3.02876800e-01 1.22119224e+00 -1.28308833e-01
-1.32776761e+00 1.27050424e+00 4.62882034e-02 1.32325321e-01
5.95655799e-01 9.15099308e-02 -4.84892935e-01 2.06598148e-01
4.20475215e-01 4.45085496e-01 4.06799354e-02 -6.80379689e-01
-3.22614163e-01 -1.92517415e-01 1.26415744e-01 5.16570508e-01
-4.16524678e-01 1.27535895e-01 -3.25854242e-01 -8.59450102e-01
-7.25431144e-02 -6.20210588e-01 1.31920323e-01 -3.65115136e-01
-1.69955850e-01 -8.22625279e-01 7.07904160e-01 -1.07030773e+00
1.38943100e+00 -1.70625985e+00 9.13373530e-02 4.22821939e-02
4.92184967e-01 9.72297490e-02 -2.91564733e-01 6.30613208e-01
1.86026856e-01 5.92990637e-01 -3.30269672e-02 -1.96778834e-01
2.33716935e-01 -1.76442713e-02 -7.57121202e-03 2.39311725e-01
-1.01058543e-01 1.28724670e+00 -1.03799284e+00 -9.11463022e-01
-2.09966265e-02 2.59672720e-02 -1.81417000e-02 -7.75119141e-02
-1.47876456e-01 2.99259514e-01 -3.53698194e-01 6.32973611e-01
2.60646373e-01 -4.92151707e-01 -9.27266926e-02 8.87930810e-01
-3.95635694e-01 3.11747998e-01 -7.76930809e-01 1.61662292e+00
-4.72359478e-01 1.29493010e+00 -8.08998644e-02 -7.69903004e-01
9.37548339e-01 5.23116350e-01 -1.36945263e-01 -1.19119787e+00
1.30765177e-02 1.73718795e-01 1.92200840e-01 -6.74916387e-01
7.38031864e-01 1.06651552e-01 -3.01951766e-01 6.21539950e-01
1.84510648e-01 -5.11803210e-01 7.41524875e-01 6.56650603e-01
6.60367489e-01 2.48163994e-02 2.59596646e-01 -6.04823411e-01
7.28812695e-01 1.44747883e-01 -1.37845531e-01 8.59774947e-01
-9.46890041e-02 2.49847382e-01 5.75243771e-01 5.71689755e-02
-9.10535395e-01 -1.16069269e+00 -1.73979849e-01 1.59935319e+00
-1.20868087e-01 -5.36799133e-01 -7.51985669e-01 -2.51759887e-01
-4.26689923e-01 8.51933002e-01 -3.62514347e-01 2.64498055e-01
-5.09857118e-01 -4.39911842e-01 5.81083119e-01 1.92368552e-01
8.02998304e-01 -1.40355277e+00 -4.52294439e-01 2.23588329e-02
-5.90225816e-01 -1.31249833e+00 -1.78624734e-01 -2.97613680e-01
-4.28464979e-01 -9.72261488e-01 -1.05071330e+00 -1.23315978e+00
2.43819222e-01 1.37740700e-02 1.45458472e+00 3.66165310e-01
-2.29304120e-01 5.73962092e-01 -2.90748775e-01 -6.98832691e-01
-4.99465823e-01 4.56175566e-01 -4.64645922e-01 -8.76810074e-01
1.03800938e-01 -5.93326846e-03 -2.44791005e-02 1.16410464e-01
-8.35843384e-01 2.68801838e-01 5.71170211e-01 5.72991371e-01
1.57810241e-01 -2.27477565e-01 8.24905038e-01 -8.98951769e-01
1.16626275e+00 -5.21080732e-01 -4.11109149e-01 6.01488829e-01
-5.33177733e-01 -3.37230742e-01 5.08052945e-01 -4.18718964e-01
-1.04288876e+00 -1.06881285e+00 -3.05444777e-01 3.26427400e-01
-1.62039593e-01 9.73601222e-01 -5.07019162e-02 -2.49679327e-01
7.19363153e-01 2.65303582e-01 -3.09558600e-01 -3.14353794e-01
-3.92743293e-03 4.78871912e-01 7.40111053e-01 -9.25444603e-01
7.84528315e-01 -3.87448102e-01 -9.16710123e-03 -1.39525616e+00
-7.74071336e-01 -3.77234995e-01 -4.35227126e-01 -6.42105281e-01
8.52500141e-01 -8.31897080e-01 -1.09420514e+00 3.78127933e-01
-1.05383456e+00 -5.78917146e-01 6.50490895e-02 4.63810802e-01
-1.23361863e-01 3.94415766e-01 -7.73547292e-01 -4.84021068e-01
-4.76258248e-02 -1.11083364e+00 6.96856439e-01 5.58763266e-01
-2.77088344e-01 -1.45451319e+00 7.16542378e-02 9.34818685e-01
9.22609642e-02 1.19273327e-01 1.51720572e+00 -8.27143312e-01
-4.87094998e-01 2.52752453e-01 -1.03339419e-01 -6.98084943e-03
-6.60170257e-01 -3.74403447e-02 -6.48935556e-01 -1.56320959e-01
-1.90939680e-01 -8.28980803e-01 6.70576036e-01 4.13336232e-02
1.15182912e+00 -1.93381593e-01 4.59341109e-02 -6.31090701e-02
1.16725433e+00 9.81372297e-02 6.97887838e-01 6.43948317e-01
5.84780395e-01 8.32423151e-01 3.05701375e-01 -1.96313988e-02
8.60353172e-01 2.85471350e-01 -1.11825570e-01 -9.80530456e-02
-2.70633966e-01 -2.29994386e-01 4.41962451e-01 1.41182029e+00
-7.26519451e-02 -5.10911644e-01 -1.52640390e+00 6.47730768e-01
-1.37421370e+00 -9.27665770e-01 -6.51366889e-01 1.86592257e+00
7.45490134e-01 -4.24620435e-02 -7.16411695e-02 9.39772353e-02
4.30119067e-01 1.43101677e-01 -6.97422847e-02 -7.55595684e-01
-4.10243809e-01 6.26005828e-01 -3.85659076e-02 6.56912744e-01
-5.61854720e-01 1.07832396e+00 6.63837719e+00 1.15558088e+00
-7.74176478e-01 7.59135783e-02 7.06597209e-01 2.98467755e-01
-4.78952676e-01 -3.13538194e-01 -6.82332695e-01 2.15341523e-01
1.24283075e+00 -5.69798350e-01 5.67892902e-02 3.17217380e-01
2.10372537e-01 -3.79833430e-01 -5.34573972e-01 7.03341722e-01
3.72705102e-01 -1.41806507e+00 4.61243950e-02 5.38764633e-02
9.96589363e-01 -3.33042562e-01 2.76188999e-01 7.74473667e-01
1.68939009e-01 -1.46170962e+00 7.54830658e-01 3.70067239e-01
7.82204330e-01 -7.33540535e-01 6.41967535e-01 6.33824527e-01
-8.88216496e-01 1.93516895e-01 -9.98718739e-02 -4.69318837e-01
-2.79575199e-01 7.84853473e-02 -8.13697517e-01 4.16634977e-01
5.93656480e-01 2.61634231e-01 -1.20895207e+00 9.51860845e-01
-6.03397369e-01 7.38287866e-01 3.31487149e-01 -7.22667336e-01
3.98435742e-01 -2.99470901e-01 3.66531461e-01 1.24298501e+00
3.92511010e-01 3.91073912e-01 4.31838304e-01 7.35276163e-01
-2.56248325e-01 6.84977889e-01 -3.42992455e-01 -3.60393196e-01
3.76477689e-01 1.11182463e+00 -8.11564863e-01 -4.41903651e-01
-5.35775423e-01 6.40373468e-01 3.40838462e-01 5.72196543e-01
-5.36704898e-01 -7.04748929e-01 -2.67997049e-02 1.22743800e-01
-4.48710382e-01 -4.58949864e-01 -5.93605042e-01 -1.03556478e+00
-3.15214545e-01 -1.21118486e+00 5.08534551e-01 -1.04201996e+00
-1.03574121e+00 4.39054430e-01 1.09608278e-01 -5.88138402e-01
-1.82103068e-01 -8.47024739e-01 -6.57755196e-01 1.04984438e+00
-1.18234515e+00 -8.95286500e-01 -5.33609748e-01 6.20806813e-01
7.99859464e-01 -2.73002803e-01 6.29975677e-01 1.52663261e-01
-6.10861003e-01 5.89618683e-01 5.59618091e-03 3.29961091e-01
7.58789182e-01 -1.37701523e+00 2.45025039e-01 6.42888784e-01
2.81658441e-01 3.27748984e-01 4.23944980e-01 -6.33139014e-01
-9.60234523e-01 -6.31551921e-01 1.60200953e+00 -5.28685272e-01
8.21375132e-01 -2.39341706e-01 -9.28891003e-01 4.76882339e-01
6.61969960e-01 -8.30178082e-01 9.34207439e-01 -4.30767089e-02
5.36032654e-02 5.75387180e-01 -7.72293627e-01 7.74145246e-01
6.90126777e-01 -7.69305766e-01 -9.68964040e-01 7.25493193e-01
3.85910183e-01 -6.43819630e-01 -1.13591266e+00 -3.82393599e-02
3.53733301e-01 -6.77031934e-01 8.62572134e-01 -7.60483801e-01
1.13070309e+00 3.44984204e-01 3.01909685e-01 -1.29426455e+00
-2.72889525e-01 -4.91744787e-01 2.21702144e-01 1.20789397e+00
5.84375143e-01 -5.44203937e-01 6.80385411e-01 4.50756609e-01
-3.77747178e-01 -7.13455021e-01 -7.96019912e-01 -5.23349404e-01
8.63442779e-01 -3.56328309e-01 2.01027334e-01 1.22213542e+00
1.50759771e-01 6.51067615e-01 1.36709854e-01 -2.30372399e-01
3.90246719e-01 8.78508165e-02 6.10536635e-01 -1.51076436e+00
1.73285469e-01 -9.40403283e-01 2.38467883e-02 -6.50270343e-01
4.68489140e-01 -1.16038251e+00 -2.18120173e-01 -2.04189897e+00
1.28937721e-01 1.96872637e-01 3.19194019e-01 3.03883016e-01
-3.85547996e-01 2.64654428e-01 3.48715246e-01 -9.38968435e-02
-5.59206605e-01 1.89165398e-01 1.99563539e+00 4.65512183e-03
8.78561884e-02 -3.01690221e-01 -8.16360414e-01 7.17820466e-01
7.50332952e-01 -7.99568463e-03 -3.09370667e-01 -3.58924240e-01
3.14677358e-01 3.22096318e-01 4.02012229e-01 -8.70296061e-01
4.07204211e-01 -4.28841323e-01 8.65177035e-01 -4.77763265e-01
8.77539590e-02 -7.82743003e-03 -4.83615458e-01 3.73113960e-01
-5.88032782e-01 6.97451711e-01 3.03376943e-01 -1.40157536e-01
-4.30594414e-01 -3.61427486e-01 5.73875606e-01 -4.62804794e-01
-8.90883625e-01 2.76529323e-02 -9.44202662e-01 6.84249759e-01
9.46197093e-01 -3.01359892e-01 -7.75304377e-01 -7.07616687e-01
-7.38139749e-01 6.24804139e-01 -4.97933419e-04 5.24975419e-01
6.86620235e-01 -1.12854290e+00 -1.12551224e+00 -9.63430405e-02
7.89027810e-02 -1.99181631e-01 3.62718731e-01 6.73500657e-01
-7.61600256e-01 8.59432936e-01 -2.70225614e-01 -1.05561249e-01
-1.22141778e+00 6.20003566e-02 -2.35369682e-01 -6.62817657e-01
-2.84446180e-01 7.39745080e-01 -1.96741633e-02 -6.52351856e-01
1.82318464e-01 -4.78811227e-02 -6.64462566e-01 3.00191671e-01
6.54139936e-01 8.63707662e-01 5.77978268e-02 -6.90998733e-01
1.45864546e-01 4.12562728e-01 2.53054172e-01 -4.12888795e-01
8.54166389e-01 -1.00134701e-01 -9.12165567e-02 5.38745463e-01
1.00955439e+00 3.91143918e-01 -1.25724033e-01 8.91131461e-02
2.58631378e-01 -1.03650115e-01 -3.83103043e-02 -1.16394043e+00
-4.50751752e-01 1.05036616e+00 8.55655223e-02 2.98236817e-01
6.39905274e-01 3.95435579e-02 5.50268412e-01 5.57504475e-01
-7.94781595e-02 -1.23706853e+00 4.03450459e-01 1.03555262e+00
1.01907682e+00 -1.12556612e+00 -7.55404755e-02 -2.25242972e-01
-7.03250229e-01 1.17937434e+00 9.20203090e-01 3.92584980e-01
2.29228958e-01 -1.35801166e-01 9.27124321e-02 -2.64344692e-01
-6.83434665e-01 -1.15099609e-01 7.58049607e-01 3.64087373e-01
9.10743535e-01 -5.99840805e-02 -6.89939499e-01 3.86011392e-01
-9.71108377e-01 -3.42256516e-01 9.32960987e-01 7.47153819e-01
-5.18690288e-01 -1.05698049e+00 -7.14803934e-01 3.53692174e-01
-5.40534914e-01 -3.77416939e-01 -6.92990065e-01 1.20259202e+00
1.23940939e-02 1.06434512e+00 -6.05541803e-02 9.02007297e-02
2.45101959e-01 3.22274685e-01 5.64450979e-01 -6.61263525e-01
-8.58158231e-01 -2.90370882e-01 1.67873725e-01 7.86643848e-02
-7.46195167e-02 -4.79643583e-01 -1.07721210e+00 -7.24219620e-01
1.34646237e-01 5.25651038e-01 6.83198571e-01 1.16401863e+00
-9.50874910e-02 5.90382934e-01 -1.29086435e-01 -3.32793772e-01
-3.17781031e-01 -1.17775130e+00 -5.64990282e-01 1.15434669e-01
-7.55548291e-03 -3.63519430e-01 -1.39091849e-01 -2.45600864e-01] | [11.058526039123535, 8.662670135498047] |
1002634b-7d4b-4e0d-b5e4-33d95353410a | controlling-perceived-emotion-in-symbolic | 2208.05162 | null | https://arxiv.org/abs/2208.05162v4 | https://arxiv.org/pdf/2208.05162v4.pdf | Controlling Perceived Emotion in Symbolic Music Generation with Monte Carlo Tree Search | This paper presents a new approach for controlling emotion in symbolic music generation with Monte Carlo Tree Search. We use Monte Carlo Tree Search as a decoding mechanism to steer the probability distribution learned by a language model towards a given emotion. At every step of the decoding process, we use Predictor Upper Confidence for Trees (PUCT) to search for sequences that maximize the average values of emotion and quality as given by an emotion classifier and a discriminator, respectively. We use a language model as PUCT's policy and a combination of the emotion classifier and the discriminator as its value function. To decode the next token in a piece of music, we sample from the distribution of node visits created during the search. We evaluate the quality of the generated samples with respect to human-composed pieces using a set of objective metrics computed directly from the generated samples. We also perform a user study to evaluate how human subjects perceive the generated samples' quality and emotion. We compare PUCT against Stochastic Bi-Objective Beam Search (SBBS) and Conditional Sampling (CS). Results suggest that PUCT outperforms SBBS and CS in almost all metrics of music quality and emotion. | ['Levi H. S. Lelis', 'Jim Whitehead', 'Lili Mou', 'Lucas N. Ferreira'] | 2022-08-10 | null | null | null | null | ['music-generation', 'music-generation'] | ['audio', 'music'] | [ 2.29689389e-01 -7.80033097e-02 -8.14918354e-02 -4.21384454e-01
-1.08419323e+00 -6.30035162e-01 3.65578353e-01 4.39080670e-02
-2.84011424e-01 9.29233432e-01 2.76558161e-01 2.52288699e-01
-1.08516552e-01 -7.20496595e-01 -4.59456086e-01 -6.72535956e-01
-7.72692785e-02 7.01172948e-01 -6.15162365e-02 -4.95352037e-02
3.57199371e-01 1.87964246e-01 -1.76603806e+00 5.52049518e-01
8.50585163e-01 1.12783849e+00 2.67688960e-01 1.16016531e+00
5.43656647e-02 5.45773983e-01 -1.09964633e+00 -3.41996104e-01
1.55480597e-02 -1.03011012e+00 -4.07424808e-01 -2.41396919e-01
-2.02858493e-01 1.75135463e-01 4.84887391e-01 1.01737487e+00
9.50511992e-01 3.12082410e-01 8.92462850e-01 -1.11632097e+00
9.47878435e-02 1.01136541e+00 -1.46538869e-01 -2.71181732e-01
1.01472962e+00 6.68249801e-02 1.05642760e+00 -3.71556222e-01
6.86914623e-01 1.18922365e+00 5.57907403e-01 6.21718228e-01
-1.60421073e+00 -8.66554856e-01 -2.53279030e-01 -2.80096033e-03
-1.61095428e+00 -5.87159336e-01 7.03473389e-01 -3.77161652e-01
6.53167844e-01 5.49674213e-01 1.08492267e+00 1.06689787e+00
2.33438894e-01 7.64719307e-01 1.01950169e+00 -7.65908957e-01
9.62702274e-01 3.28090429e-01 -5.26073098e-01 6.14127398e-01
-3.52818787e-01 3.22149396e-01 -1.14092088e+00 -7.13455200e-01
5.22998571e-01 -9.79133129e-01 -2.28009284e-01 -1.30351052e-01
-9.84119177e-01 7.86747217e-01 7.40100890e-02 2.25275323e-01
-7.48464882e-01 4.92299676e-01 9.31220502e-02 9.34640542e-02
2.97789663e-01 7.97608733e-01 -2.29914084e-01 -8.31233442e-01
-1.34765601e+00 6.09810054e-01 1.33494186e+00 5.91855526e-01
3.04949760e-01 5.79466447e-02 -8.26923788e-01 8.97653997e-01
2.62395322e-01 4.92867649e-01 4.33898985e-01 -1.13681614e+00
-1.16009906e-01 -9.10231397e-02 4.74581987e-01 -7.29286373e-01
7.64962435e-02 -6.09808743e-01 -2.22990304e-01 5.04924238e-01
3.79077524e-01 -3.34017575e-01 -6.42144144e-01 1.87002265e+00
2.22290963e-01 3.29714529e-02 -2.32994363e-01 8.59524846e-01
2.00677335e-01 8.49180698e-01 1.39778286e-01 -6.25050843e-01
1.02767825e+00 -4.07470614e-01 -6.13431633e-01 1.11680582e-01
4.70845997e-01 -8.82450521e-01 1.26643217e+00 1.06474054e+00
-1.34225976e+00 -5.72835088e-01 -1.04953527e+00 7.09768772e-01
2.40777463e-01 2.32615754e-01 4.43621218e-01 1.09611380e+00
-8.08054507e-01 9.53032017e-01 -5.74354589e-01 8.14657286e-03
2.27523014e-01 1.52693331e-01 5.71560621e-01 3.20747465e-01
-1.00512409e+00 5.30295253e-01 2.88568646e-01 -3.46812248e-01
-1.29534781e+00 -3.22138637e-01 -3.13994586e-01 1.92498073e-01
1.76216699e-02 -5.41949749e-01 1.56567240e+00 -1.08112597e+00
-2.04455400e+00 6.39602244e-01 -2.55607039e-01 -3.69491249e-01
3.95068854e-01 -7.33985752e-02 -2.20130399e-01 -6.86871931e-02
3.64434868e-02 8.36340368e-01 9.51351285e-01 -1.48187971e+00
-6.30482018e-01 5.53735122e-02 -3.90531093e-01 4.02228057e-01
1.17526107e-01 -9.67807174e-02 -2.40891293e-01 -6.25244021e-01
-4.52947356e-02 -8.11246395e-01 -1.68329656e-01 -3.65641057e-01
-6.22877896e-01 -2.45094329e-01 -8.18848331e-03 -4.54721779e-01
1.67191851e+00 -2.08864307e+00 4.62284356e-01 6.58154428e-01
-4.40406173e-01 -2.76916772e-01 -8.59703571e-02 2.36931831e-01
1.45032197e-01 7.43459016e-02 -6.61478713e-02 -3.80818456e-01
2.64611036e-01 1.01700455e-01 -3.43234956e-01 5.42686172e-02
-4.20641214e-01 4.69919235e-01 -8.25359702e-01 -5.96433938e-01
-3.03019047e-01 2.42604494e-01 -8.59738708e-01 4.36627507e-01
-9.49669063e-01 3.02473724e-01 -4.70641494e-01 4.50915158e-01
1.75703287e-01 2.16141477e-01 1.74535066e-01 6.09296225e-02
1.19963266e-01 2.67730087e-01 -1.24132729e+00 1.77373731e+00
-5.34194767e-01 2.46552229e-01 1.18559990e-02 -2.39913136e-01
1.24234092e+00 4.55127835e-01 1.87776431e-01 -3.26688558e-01
2.28173569e-01 1.30066603e-01 -8.52120146e-02 -4.00779039e-01
5.34793854e-01 -6.46163225e-01 -4.17467117e-01 7.32714593e-01
-2.04689220e-01 -6.57931864e-01 9.65876281e-02 4.45723310e-02
9.63833988e-01 4.73424226e-01 2.81180263e-01 1.18714653e-01
1.60475716e-01 -1.03393115e-01 4.19155151e-01 1.08960080e+00
-4.00943756e-02 4.37989473e-01 7.49021292e-01 1.23288222e-01
-7.64946520e-01 -1.20344269e+00 2.26750806e-01 1.33463144e+00
-2.12320939e-01 -4.42038417e-01 -8.94587815e-01 -2.96265095e-01
-2.10909396e-01 1.48628509e+00 -4.77519929e-01 -2.45785758e-01
-4.61482294e-02 -5.08704066e-01 5.42370021e-01 -4.50167581e-02
-3.28099821e-03 -1.45851302e+00 -1.02144825e+00 4.52035457e-01
-4.57111537e-01 -5.27444065e-01 -4.03260887e-01 2.15004951e-01
-7.03929186e-01 -4.11328673e-01 -5.16543686e-01 -2.40316033e-01
1.17633931e-01 -7.47747958e-01 1.27302790e+00 -2.99899518e-01
-3.32504928e-01 1.84232950e-01 -6.83820546e-01 -3.70867282e-01
-7.94108927e-01 6.82172105e-02 -1.20660596e-01 5.96171021e-02
-1.05915286e-01 -7.91369736e-01 -4.47651356e-01 2.24688783e-01
-4.86578614e-01 6.00934103e-02 3.37667614e-01 6.34066224e-01
6.54647231e-01 3.00810695e-01 3.69726628e-01 -2.98613787e-01
1.26261127e+00 -2.45808810e-01 -4.35674906e-01 1.45925000e-01
-5.40816665e-01 2.02808335e-01 3.10220003e-01 -7.71755099e-01
-9.42559659e-01 9.66225713e-02 -7.13544935e-02 -2.99609125e-01
-1.02740571e-01 4.41028774e-01 -7.40955397e-03 4.11080748e-01
1.15213525e+00 1.72529489e-01 -2.65330046e-01 -1.98647439e-01
3.58733267e-01 7.85750389e-01 4.66255516e-01 -9.87391770e-01
7.97923878e-02 1.46160007e-03 -1.61459699e-01 -4.61005121e-01
-8.30133259e-01 8.46156180e-02 2.42093131e-02 -6.77099764e-01
8.80805254e-01 -4.01678115e-01 -1.14116907e+00 2.18559384e-01
-1.03876770e+00 -6.07387424e-01 -5.94711959e-01 6.27902091e-01
-1.14273608e+00 -1.55239671e-01 -2.06680819e-01 -1.71488941e+00
-3.39800537e-01 -8.49830985e-01 1.29292154e+00 2.70975024e-01
-9.72415268e-01 -3.85751009e-01 8.76535550e-02 3.24359676e-03
1.30966008e-01 4.71886158e-01 9.88188148e-01 -3.61573786e-01
-4.04377818e-01 -5.31661093e-01 5.40170133e-01 8.62802714e-02
-3.85954708e-01 -2.66719330e-02 -9.16200936e-01 -5.86695708e-02
-1.22103579e-01 -5.39782882e-01 4.05107886e-01 7.07543135e-01
1.02600896e+00 -5.59192486e-02 -2.78827459e-01 4.37145382e-01
1.18474960e+00 4.79431540e-01 6.73793137e-01 7.15464801e-02
-1.40791656e-02 3.62556607e-01 8.46457839e-01 1.10879970e+00
-2.24544212e-01 9.39326227e-01 1.87434629e-01 6.14507437e-01
1.41938180e-01 -7.12691426e-01 5.40669978e-01 3.56377393e-01
-7.24359080e-02 -5.49232841e-01 -5.00450015e-01 1.42473802e-01
-1.53122592e+00 -1.09847617e+00 3.69444549e-01 2.54295230e+00
1.19823933e+00 4.62854475e-01 4.65752751e-01 4.81451660e-01
5.27132928e-01 -1.32700741e-01 -4.75869626e-01 -7.92832017e-01
2.30150282e-01 6.74323618e-01 2.63121035e-02 6.75611556e-01
-4.62570935e-01 8.78544271e-01 6.72968769e+00 1.15350521e+00
-7.80414581e-01 -1.23520486e-01 5.73403418e-01 -5.20640612e-01
-3.49528313e-01 -6.11475334e-02 -5.80351055e-01 5.18664777e-01
9.54837143e-01 -1.56816930e-01 9.88976598e-01 6.45157516e-01
5.32019675e-01 -6.17919743e-01 -1.15832734e+00 9.27365720e-01
-5.46910204e-02 -9.73545253e-01 -9.94585454e-02 -9.21190232e-02
5.12021005e-01 -5.02728760e-01 -4.13915422e-03 3.15630794e-01
3.54872555e-01 -1.17788947e+00 1.30328918e+00 9.89520550e-01
8.63259614e-01 -9.67185616e-01 2.43106648e-01 6.09007418e-01
-8.14642489e-01 1.24868490e-02 -2.34674402e-02 -6.07309379e-02
3.98221701e-01 7.04656482e-01 -1.10620093e+00 2.90923994e-02
4.09544468e-01 -1.28979355e-01 6.56616092e-02 1.03415334e+00
-4.44822967e-01 1.03321564e+00 -3.51973057e-01 -7.72187173e-01
-1.51320651e-01 -9.23994556e-02 8.04384708e-01 1.06665421e+00
5.63250065e-01 -9.04615745e-02 1.59994513e-01 1.17652452e+00
3.88924003e-01 2.34172970e-01 2.53444314e-02 -1.08388267e-01
8.81186962e-01 8.56857002e-01 -7.94998586e-01 -4.71175373e-01
8.18328559e-01 1.17130172e+00 -1.52309492e-01 2.59135097e-01
-7.32316613e-01 -4.44132239e-01 4.33032840e-01 -1.40595526e-01
4.43498313e-01 -1.60903949e-02 -6.21957958e-01 -6.01789474e-01
-4.03970897e-01 -1.13231468e+00 2.07445174e-01 -1.30481613e+00
-7.07948744e-01 5.83286703e-01 3.25433351e-02 -1.02654397e+00
-7.39695787e-01 1.97711866e-02 -5.70694208e-01 1.22755492e+00
-4.83487576e-01 -3.27805936e-01 -1.13979451e-01 1.57183677e-01
4.55981195e-01 -1.96577609e-01 1.21518302e+00 -2.85324305e-01
-1.30173102e-01 6.22104645e-01 -2.57506102e-01 -5.00221133e-01
4.07748610e-01 -1.10999370e+00 -4.95456718e-02 1.68757141e-01
3.03568333e-01 1.49896845e-01 1.37770045e+00 -7.59679854e-01
-9.47689474e-01 -2.63770640e-01 7.47372866e-01 -1.04693711e-01
7.12215304e-02 -3.96089584e-01 -3.98715913e-01 -2.78589539e-02
-1.44083083e-01 -6.84361815e-01 7.44130969e-01 1.83758482e-01
2.89488081e-02 9.59166735e-02 -1.42463994e+00 5.77300131e-01
9.21425998e-01 -1.86458468e-01 -1.96157530e-01 3.64356004e-02
3.23448896e-01 -4.04439330e-01 -8.49450111e-01 1.96226463e-01
1.08348417e+00 -1.27917647e+00 5.74367285e-01 -2.12418884e-01
4.06539947e-01 -2.83558279e-01 -4.28146482e-01 -1.69066966e+00
-2.39040270e-01 -1.00931811e+00 1.36588002e-02 1.08144534e+00
5.87982953e-01 -9.03160200e-02 9.90917683e-01 1.52648598e-01
3.57159734e-01 -8.54725301e-01 -9.89070833e-01 -5.89258671e-01
-2.75658846e-01 -8.36681485e-01 6.33864522e-01 3.01494241e-01
1.48025900e-01 4.17899787e-01 -4.12956536e-01 -1.56422526e-01
6.81778133e-01 3.13952506e-01 6.74843609e-01 -9.22967494e-01
-1.01696301e+00 -5.41606009e-01 -2.18751937e-01 -8.46771896e-01
-1.28066212e-01 -9.40773129e-01 4.45107400e-01 -1.34292638e+00
-7.46945366e-02 -5.26560009e-01 6.79749474e-02 8.38312134e-02
6.87415153e-02 -2.44805999e-02 2.19663724e-01 -1.33231478e-02
-2.81082273e-01 7.06158996e-01 1.22827029e+00 3.51081073e-01
-5.60590208e-01 4.05088961e-01 -4.75567341e-01 5.90758502e-01
7.82440603e-01 -7.74016917e-01 -3.59406978e-01 3.09926361e-01
5.73608816e-01 7.65211403e-01 7.70835951e-02 -1.23308563e+00
-3.13905329e-02 -1.33652702e-01 5.31609654e-01 -6.40668750e-01
6.18123531e-01 -3.41201991e-01 2.91670114e-01 5.65341532e-01
-7.31598318e-01 -3.09569001e-01 -8.36379230e-02 5.49818993e-01
1.50700986e-01 -5.32399297e-01 7.23733783e-01 6.07945723e-03
1.18372189e-02 -3.27457249e-01 -6.08169615e-01 -5.97160943e-02
7.62462020e-01 -2.74048716e-01 5.06861150e-01 -1.02388942e+00
-1.16440535e+00 1.17585093e-01 2.01837882e-01 -1.58909876e-02
5.60487151e-01 -1.31113684e+00 -7.02160776e-01 1.89614847e-01
-5.06095961e-02 -4.76271808e-01 -2.23984942e-01 3.82790327e-01
-2.52731889e-01 -1.58110440e-01 9.16548446e-02 -5.05299389e-01
-1.33206534e+00 3.90760228e-02 5.65673769e-01 -2.64965147e-01
-4.89958189e-02 1.15131664e+00 -5.13497293e-01 -1.74256325e-01
5.75811684e-01 -2.41711468e-01 1.35762513e-01 2.37306505e-02
1.87267765e-01 4.08772558e-01 -3.69548738e-01 -1.23110302e-01
-8.69769827e-02 2.23181978e-01 6.28382325e-01 -1.17460775e+00
1.01660883e+00 2.27667913e-01 1.03345171e-01 7.95027375e-01
6.26576185e-01 2.15719923e-01 -1.08607149e+00 2.51726180e-01
1.24329086e-02 -5.03223658e-01 4.40222304e-03 -1.55981672e+00
-3.95104140e-01 3.60084176e-01 7.51435876e-01 3.85700673e-01
1.25242364e+00 1.09169513e-01 5.52602828e-01 -8.77162740e-02
5.10821283e-01 -1.17706871e+00 1.02631256e-01 1.54646933e-01
9.79627132e-01 -2.80443609e-01 -6.47439659e-02 -2.12886259e-01
-8.93706203e-01 8.61134648e-01 1.39876261e-01 1.70914549e-02
5.48634827e-01 5.01701415e-01 6.09377474e-02 3.13465893e-02
-1.09764528e+00 -1.92509845e-01 1.38793588e-01 5.00642061e-01
6.46009862e-01 5.79246938e-01 -4.17720884e-01 6.38730645e-01
-9.31098342e-01 4.94013608e-01 2.07168877e-01 6.80407166e-01
-8.14526796e-01 -1.18347430e+00 -6.25399947e-01 5.38152754e-01
-2.92329192e-01 1.96442641e-02 -4.40057963e-01 1.72632337e-01
3.12786937e-01 1.16727507e+00 -3.71697638e-03 -7.22410023e-01
3.32565635e-01 3.63609672e-01 8.86613727e-01 -5.09913862e-01
-5.26507318e-01 4.40091908e-01 6.43270135e-01 -3.71323019e-01
-8.10308531e-02 -9.04796541e-01 -1.18058360e+00 -5.12746070e-03
-3.37580144e-01 5.76762080e-01 1.04137361e+00 6.43381774e-01
4.97910194e-02 5.79761147e-01 8.20661306e-01 -1.02053630e+00
-7.42341280e-01 -1.12955093e+00 -8.72416615e-01 -1.27679901e-02
-2.61086047e-01 -3.88726562e-01 -3.26270431e-01 -8.27347562e-02] | [16.009960174560547, 5.549931049346924] |
0cdce88d-cc15-4877-ba1e-d76a22bf6425 | optimal-transport-for-change-detection-on | 2302.07025 | null | https://arxiv.org/abs/2302.07025v3 | https://arxiv.org/pdf/2302.07025v3.pdf | Optimal Transport for Change Detection on LiDAR Point Clouds | Unsupervised change detection between airborne LiDAR data points, taken at separate times over the same location, can be difficult due to unmatching spatial support and noise from the acquisition system. Most current approaches to detect changes in point clouds rely heavily on the computation of Digital Elevation Models (DEM) images and supervised methods. Obtaining a DEM leads to LiDAR informational loss due to pixelisation, and supervision requires large amounts of labelled data often unavailable in real-world scenarios. We propose an unsupervised approach based on the computation of the transport of 3D LiDAR points over two temporal supports. The method is based on unbalanced optimal transport and can be generalised to any change detection problem with LiDAR data. We apply our approach to publicly available datasets for monitoring urban sprawling in various noise and resolution configurations that mimic several sensors used in practice. Our method allows for unsupervised multi-class classification and outperforms the previous state-of-the-art unsupervised approaches by a significant margin. | ['Makoto Yamada', 'Peter Naylor', 'Marco Fiorucci'] | 2023-02-14 | null | null | null | null | ['change-detection'] | ['computer-vision'] | [ 7.99128234e-01 -5.49135506e-01 1.22360535e-01 -5.44406712e-01
-8.90427291e-01 -6.72121227e-01 8.00466537e-01 6.57076657e-01
-8.36370230e-01 7.80702710e-01 -3.45588773e-01 -3.87804091e-01
-4.27455127e-01 -1.40410411e+00 -5.33583760e-01 -5.57592332e-01
-2.30795771e-01 9.23821211e-01 8.55851114e-01 -1.50780007e-01
6.04063869e-02 1.00302815e+00 -1.91917074e+00 -2.09652647e-01
8.39466095e-01 9.01200771e-01 2.21930608e-01 7.58787692e-01
-2.05394849e-01 -8.64329487e-02 -1.29539356e-01 8.74811709e-02
6.05006635e-01 1.62629485e-01 -3.91423613e-01 3.06761682e-01
8.58132362e-01 -2.81771034e-01 7.88745210e-02 1.00233924e+00
4.52517927e-01 6.18596338e-02 7.99432814e-01 -1.08078480e+00
2.97205329e-01 1.03166588e-01 -5.60819685e-01 3.42986077e-01
-6.82392791e-02 -6.49078712e-02 8.80084634e-01 -9.41538990e-01
4.66165721e-01 1.05911946e+00 9.81950939e-01 -3.93674672e-01
-1.99406731e+00 -5.52761495e-01 5.92326820e-02 1.27055869e-01
-1.68209302e+00 -4.80136096e-01 5.73452294e-01 -7.56523967e-01
7.57167578e-01 4.49342519e-01 7.72326589e-01 2.06413314e-01
-3.09586585e-01 1.01503991e-01 1.14695907e+00 -3.53652716e-01
4.00427282e-01 -1.66115671e-01 -1.61527872e-01 2.57491350e-01
6.36876166e-01 4.53161925e-01 -3.19926918e-01 -2.42066041e-01
4.22321051e-01 2.32808679e-01 6.74973875e-02 -7.38069057e-01
-9.62044060e-01 7.03734875e-01 4.62785959e-01 1.34711802e-01
-3.81663442e-01 -3.64526771e-02 2.03073412e-01 4.54115421e-01
7.04296768e-01 -4.02576998e-02 -5.63226700e-01 -3.78457978e-02
-1.45368147e+00 3.36109877e-01 3.43384504e-01 8.86443853e-01
1.28549087e+00 -2.07431376e-01 6.02132618e-01 6.12283409e-01
3.13118607e-01 8.99798334e-01 -1.88849553e-01 -8.72910917e-01
5.36514938e-01 6.75862789e-01 2.32067659e-01 -9.11885321e-01
-1.98773384e-01 3.53405699e-02 -7.27344096e-01 6.31579816e-01
2.90608406e-01 1.49919376e-01 -9.19550240e-01 1.10319602e+00
5.74339747e-01 1.49540529e-01 -9.53846276e-02 3.30144495e-01
2.91914612e-01 3.20251822e-01 -5.60647510e-02 -3.54251802e-01
8.83998334e-01 1.64774299e-01 -3.25706959e-01 -2.31384158e-01
5.06555796e-01 -6.40757382e-01 6.85104668e-01 2.70028502e-01
-6.41271949e-01 -5.07188082e-01 -1.09673333e+00 2.02266186e-01
-7.89384365e-01 -1.64487273e-01 2.05231205e-01 7.16784418e-01
-9.65991139e-01 8.84434283e-01 -9.35984433e-01 -6.05340958e-01
5.78793645e-01 4.82065439e-01 -2.31998265e-01 -1.57116070e-01
-6.87270403e-01 7.14872003e-01 2.23187447e-01 3.91979992e-01
-4.03116792e-01 -6.27324879e-01 -8.12041342e-01 -3.97558630e-01
3.50704134e-01 -1.38509914e-01 9.49342608e-01 -4.13327247e-01
-9.48593616e-01 8.72311771e-01 -1.43502265e-01 -5.51360667e-01
7.51199782e-01 -7.71038756e-02 -3.73194784e-01 6.37138486e-02
3.82700115e-01 5.01152575e-01 8.03909004e-01 -1.22682238e+00
-1.01673412e+00 -6.68789625e-01 -4.33106005e-01 6.65245280e-02
2.04447404e-01 -1.47022888e-01 1.08489282e-01 -3.37395966e-01
6.52726471e-01 -9.70566213e-01 -4.85026032e-01 2.31863439e-01
-3.32545005e-02 3.25117409e-01 1.10418022e+00 -3.06402266e-01
9.07921195e-01 -2.10553551e+00 -4.05712306e-01 5.31975985e-01
-9.74419713e-02 2.74814904e-01 1.61312506e-01 5.42732537e-01
1.51724601e-02 2.46111333e-01 -1.03423059e+00 -2.14512944e-01
-1.91867709e-01 7.74240255e-01 -3.09058160e-01 6.81672335e-01
2.20921397e-01 4.15492058e-01 -8.98379028e-01 -5.90033531e-01
6.30965054e-01 1.42411113e-01 -1.96856588e-01 -2.27879807e-01
1.90789606e-02 3.77708972e-01 -2.37689644e-01 6.34225011e-01
1.15964663e+00 6.30002618e-01 5.46461828e-02 1.61976203e-01
-7.37186015e-01 4.49375421e-01 -1.72602677e+00 1.34992516e+00
-2.48170555e-01 7.65809953e-01 6.91335127e-02 -9.71008539e-01
1.09866774e+00 3.42102200e-02 7.05498099e-01 -3.50310236e-01
-3.47824425e-01 4.53704983e-01 -1.64266258e-01 -2.79606611e-01
6.11541331e-01 -2.61981755e-01 7.87896216e-02 1.99654877e-01
-4.71052140e-01 -8.11444163e-01 1.96876124e-01 -1.55769080e-01
1.17260599e+00 -3.15309279e-02 2.93100059e-01 -4.23383445e-01
4.18613255e-01 2.51787871e-01 4.18179512e-01 6.94285572e-01
-8.09523985e-02 4.26965743e-01 -8.27478096e-02 -3.60558480e-01
-1.11240113e+00 -1.35355568e+00 -6.51172280e-01 5.22279918e-01
-3.62600498e-02 -1.40479669e-01 -8.73914212e-02 -3.90389174e-01
5.01061499e-01 4.77817804e-01 -2.24156693e-01 3.74495476e-01
-4.57098097e-01 -7.92101800e-01 4.94948536e-01 5.00724375e-01
5.71140468e-01 -5.27854204e-01 -7.98366666e-01 4.31878626e-01
1.79331362e-01 -1.15744925e+00 2.94687778e-01 4.70270067e-01
-1.45099092e+00 -1.12517095e+00 5.40298335e-02 -3.52004021e-01
5.54420412e-01 3.93567652e-01 1.02155459e+00 -5.51349856e-02
-4.28357244e-01 3.67213666e-01 -8.20325539e-02 -6.71372056e-01
-1.52987555e-01 7.73266107e-02 1.26856968e-01 -5.54957576e-02
5.57508886e-01 -1.16969121e+00 -3.04055512e-01 4.10411775e-01
-9.44218397e-01 -5.25277734e-01 3.62848163e-01 3.99735957e-01
9.52900469e-01 6.61345720e-01 1.85728908e-01 -7.58226156e-01
2.52589229e-02 -2.97964811e-01 -1.12023592e+00 -2.72502005e-01
-8.30390215e-01 -5.52597344e-02 -7.72716701e-02 -5.61421216e-02
-8.11850429e-01 5.79141200e-01 1.65213615e-01 -5.03044240e-02
-7.69906640e-01 3.80181342e-01 -3.32454711e-01 -2.84276217e-01
7.07056999e-01 -5.60312942e-02 -2.15784743e-01 -6.06364429e-01
3.58906835e-01 8.72299254e-01 6.72302186e-01 -4.36121345e-01
1.33210540e+00 1.06367433e+00 4.57487404e-01 -1.37878156e+00
-1.60566404e-01 -1.03068984e+00 -1.50878298e+00 -1.99892744e-01
3.52878869e-01 -9.57204819e-01 -4.11437824e-02 4.59423661e-01
-9.34768856e-01 -2.84542859e-01 -7.14097619e-01 5.11316776e-01
-4.59559888e-01 3.94369096e-01 1.59309179e-01 -1.08645177e+00
1.68927431e-01 -7.25888073e-01 1.17882013e+00 -4.11783040e-01
1.24280741e-02 -7.51422286e-01 3.89276266e-01 6.91910135e-03
1.95156857e-01 5.58605671e-01 6.32576048e-01 -1.63604572e-01
-8.25415373e-01 -3.35751861e-01 -2.14302018e-01 2.57820278e-01
4.67263430e-01 4.08999741e-01 -9.18388426e-01 -1.41770989e-01
-8.48729461e-02 1.46378368e-01 1.03810680e+00 2.35940844e-01
6.27640247e-01 5.27525209e-02 -5.12344778e-01 2.69300103e-01
1.84942532e+00 -6.02984540e-02 7.14897037e-01 4.56726313e-01
3.99934888e-01 7.59822726e-01 8.09602201e-01 3.35365713e-01
2.84661859e-01 6.73968911e-01 6.49092436e-01 -5.35671003e-02
2.54468203e-01 -2.13051699e-02 3.44042927e-02 4.84061509e-01
-2.47853339e-01 3.30635428e-01 -1.26834893e+00 1.04572058e+00
-1.70158076e+00 -1.05940437e+00 -7.65531361e-01 2.66622901e+00
6.02955878e-01 3.08595508e-01 9.65984464e-02 6.75361574e-01
6.87104166e-01 1.17142126e-01 -3.06548417e-01 -1.48966670e-01
-1.80661511e-02 5.18448770e-01 1.20407414e+00 6.58161223e-01
-1.22058511e+00 6.31436884e-01 6.49037600e+00 3.26359510e-01
-8.10926676e-01 1.55667141e-01 -1.88043296e-01 -1.32227436e-01
-5.05352914e-02 2.78269231e-01 -8.18777502e-01 3.49210948e-01
9.82195318e-01 1.14782818e-01 -5.33105899e-03 6.01447940e-01
5.84906578e-01 -7.30989635e-01 -1.03207946e+00 8.13338816e-01
-4.72347915e-01 -9.26067829e-01 -2.80070990e-01 5.00455737e-01
7.10373342e-01 5.95269680e-01 -3.26745600e-01 -2.11436316e-01
5.60463130e-01 -6.16498530e-01 5.39975286e-01 5.80746472e-01
8.77179265e-01 -6.44656897e-01 4.52713549e-01 4.45272654e-01
-1.61418569e+00 4.97173741e-02 -4.00866151e-01 -3.91516477e-01
2.39661202e-01 1.07176912e+00 -9.23332572e-01 5.79366446e-01
7.92400122e-01 8.16409290e-01 -6.93067491e-01 1.32640827e+00
-2.41652697e-01 5.53602755e-01 -1.12359607e+00 4.32993799e-01
2.46944249e-01 -5.66763043e-01 7.60940611e-01 1.13371491e+00
5.32128334e-01 -7.63151236e-03 3.17307532e-01 5.01402259e-01
3.51563394e-01 -2.84877345e-02 -1.11299694e+00 4.49474841e-01
6.75065935e-01 8.88190389e-01 -5.97586751e-01 -2.19373032e-01
-2.98098356e-01 4.71949458e-01 -1.59179926e-01 7.30260387e-02
-1.80362076e-01 -2.22818747e-01 8.52891624e-01 6.48415685e-01
6.25006974e-01 -7.70452619e-01 -3.14719439e-01 -6.08200073e-01
2.26075023e-01 -2.59457886e-01 2.90707111e-01 -4.88094389e-01
-1.07254112e+00 4.01980914e-02 4.60712582e-01 -1.57963538e+00
-2.43336082e-01 -4.36296731e-01 -5.64823687e-01 8.22894394e-01
-1.84530234e+00 -9.65995014e-01 -5.07076263e-01 4.37031865e-01
3.56884956e-01 2.84477532e-01 7.22444713e-01 4.49272931e-01
-6.33017197e-02 -3.66206050e-01 4.17499274e-01 -8.53566602e-02
5.15747309e-01 -1.36812210e+00 6.57732964e-01 1.11549294e+00
3.28632087e-01 -5.48449438e-03 6.99674249e-01 -8.84648740e-01
-7.72990048e-01 -1.50805724e+00 1.05519187e+00 -4.58807170e-01
8.62267375e-01 -2.18497187e-01 -8.73902321e-01 5.77956975e-01
-3.68166566e-01 1.73360273e-01 5.40072680e-01 3.48589681e-02
-9.82213300e-03 -5.59851408e-01 -1.28025746e+00 2.40217105e-01
1.23896801e+00 -6.30407035e-01 -5.98827004e-01 1.61897436e-01
8.81928131e-02 -4.98351157e-02 -8.28115165e-01 6.19235218e-01
3.76844704e-01 -8.32936943e-01 9.03371751e-01 -3.15395892e-01
-1.38345584e-01 -7.22187638e-01 -5.71388841e-01 -1.19603384e+00
-1.24217086e-01 -2.68589497e-01 5.30135393e-01 1.24999225e+00
1.62573695e-01 -6.72441244e-01 6.67504013e-01 4.14096743e-01
2.23079443e-01 9.87095386e-02 -1.37509882e+00 -1.02327275e+00
-9.56648141e-02 -1.04050195e+00 6.72668993e-01 7.36150861e-01
-7.04201043e-01 1.25108808e-01 2.77586520e-01 8.18634272e-01
9.41899240e-01 5.65597378e-02 9.68625903e-01 -2.24062538e+00
3.29300016e-01 -2.63545573e-01 -1.00823033e+00 -7.00950325e-01
-9.50613096e-02 -7.16351032e-01 2.10472122e-01 -1.35640335e+00
-4.91541713e-01 -9.34332967e-01 2.56325036e-01 2.80258983e-01
3.88771653e-01 4.31258261e-01 -1.76601127e-01 4.43902045e-01
1.16384111e-01 4.09916580e-01 2.63028771e-01 -3.83740366e-01
-4.25670892e-01 2.54526496e-01 6.82493001e-02 1.00889552e+00
7.62123406e-01 -8.68958175e-01 -1.83006838e-01 -4.54585940e-01
2.78034717e-01 -5.50660312e-01 6.80335462e-01 -1.38571572e+00
2.52786934e-01 -3.59537870e-01 9.62552279e-02 -1.29678202e+00
3.92383754e-01 -1.52434695e+00 4.41381961e-01 2.66712368e-01
1.28663823e-01 -7.59991212e-03 1.47606567e-01 7.12845504e-01
-1.36258304e-01 -4.30323362e-01 9.70002830e-01 -2.49693975e-01
-6.89907908e-01 3.93033803e-01 -5.72418094e-01 -3.06657791e-01
8.35996985e-01 -4.14222181e-01 1.84192955e-01 -3.49935368e-02
-7.65244007e-01 1.88108742e-01 6.96698606e-01 -6.74644858e-02
4.62047219e-01 -1.34309661e+00 -7.42302895e-01 3.82518977e-01
3.36720407e-01 4.78911251e-01 -3.25184524e-01 7.13246286e-01
-6.00914061e-01 2.49695107e-01 1.32959578e-02 -1.12859571e+00
-1.52626741e+00 2.34972835e-01 3.56020987e-01 -4.57586013e-02
-5.46060324e-01 1.96969002e-01 -5.50528049e-01 -6.73498690e-01
-1.39217779e-01 -6.62780464e-01 1.98039338e-01 5.96893966e-01
2.09141463e-01 5.65038145e-01 5.72762549e-01 -9.29967999e-01
-3.57829422e-01 8.40009093e-01 3.32395524e-01 -3.95543545e-01
1.38776124e+00 -4.03458297e-01 -2.21524686e-02 8.56735468e-01
8.20268273e-01 5.04367575e-02 -1.07100630e+00 -6.37783408e-01
3.57534170e-01 -8.40253413e-01 2.73610234e-01 -2.26975456e-01
-7.04044640e-01 9.44759130e-01 1.03533471e+00 3.27784508e-01
9.57400024e-01 -2.31796190e-01 2.49793395e-01 8.20782542e-01
5.78945994e-01 -1.35651922e+00 -6.60191417e-01 3.10904741e-01
6.09780848e-01 -1.34886062e+00 4.74256933e-01 -6.30578101e-01
1.05456062e-01 9.03623402e-01 -8.08771774e-02 -3.56686302e-02
1.00071383e+00 2.95544893e-01 -1.14734858e-01 -2.06739411e-01
-2.23888412e-01 -9.43766534e-01 -1.63133591e-01 9.85713899e-01
-2.36109480e-01 2.85078734e-01 -8.24326947e-02 -3.42884362e-01
-2.21904248e-01 2.43323166e-02 4.07256424e-01 1.33744204e+00
-8.10111046e-01 -1.31278598e+00 -7.65482903e-01 6.76860332e-01
1.25351518e-01 -1.52758718e-01 -3.45205963e-01 9.94301617e-01
4.76559728e-01 9.20206308e-01 5.61960459e-01 -7.87553750e-03
6.65398717e-01 1.37225688e-01 2.70133555e-01 -7.11187482e-01
-9.10468921e-02 -9.35625168e-04 1.53982148e-01 -3.38763833e-01
-9.52164769e-01 -1.46988797e+00 -9.16189671e-01 -2.76910126e-01
-4.07462537e-01 -1.15153477e-01 8.22906017e-01 8.35917652e-01
1.21444076e-01 -3.62444366e-03 7.73908675e-01 -1.15187407e+00
-2.79758930e-01 -9.25297797e-01 -9.30239379e-01 7.57596269e-02
5.31457126e-01 -9.09674883e-01 -4.29056525e-01 2.67982543e-01] | [8.424668312072754, -2.5805282592773438] |
1e1118a3-3f29-4d9a-9a31-145ba0437e2d | evaluating-persuasion-strategies-and-deep | null | null | https://aclanthology.org/E17-2077 | https://aclanthology.org/E17-2077.pdf | Evaluating Persuasion Strategies and Deep Reinforcement Learning methods for Negotiation Dialogue agents | In this paper we present a comparative evaluation of various negotiation strategies within an online version of the game {``}Settlers of Catan{''}. The comparison is based on human subjects playing games against artificial game-playing agents ({`}bots{'}) which implement different negotiation dialogue strategies, using a chat dialogue interface to negotiate trades. Our results suggest that a negotiation strategy that uses persuasion, as well as a strategy that is trained from data using Deep Reinforcement Learning, both lead to an improved win rate against humans, compared to previous rule-based and supervised learning baseline dialogue negotiators. | ['Klaus-Peter Engelbrecht', "Heriberto Cuay{\\'a}huitl", 'Markus Guhe', 'Ioannis Efstathiou', 'Alex Lascarides', 'Mihai Dobre', 'Simon Keizer', 'Oliver Lemon'] | 2017-04-01 | null | null | null | eacl-2017-4 | ['persuasion-strategies'] | ['computer-vision'] | [-2.54962910e-02 8.64756763e-01 1.36325747e-01 -2.90513605e-01
-5.71471453e-01 -8.47177446e-01 9.91949797e-01 -6.49086088e-02
-8.81062925e-01 1.19967270e+00 2.77536124e-01 -5.92367828e-01
-1.32620811e-01 -1.13989425e+00 1.04612343e-01 -4.31817681e-01
8.53997990e-02 1.06843603e+00 3.45800579e-01 -1.47112489e+00
5.66463411e-01 -1.96443275e-01 -7.50733137e-01 3.56402487e-01
4.78090167e-01 5.48374474e-01 -2.12762848e-01 9.84470308e-01
-6.98558316e-02 1.44319391e+00 -1.12196112e+00 -1.12649548e+00
6.33650303e-01 -8.12516332e-01 -1.48301017e+00 -4.96338785e-01
-2.22639352e-01 -5.44023395e-01 -2.28414819e-01 8.57417226e-01
6.10912621e-01 3.35850626e-01 6.57709956e-01 -1.17862785e+00
-1.02900162e-01 1.20376766e+00 -1.20283313e-01 1.70919195e-01
6.07010126e-01 5.83930731e-01 1.17451310e+00 1.64989069e-01
9.92217422e-01 1.27312148e+00 8.41941774e-01 1.18929672e+00
-1.25995421e+00 -8.26254725e-01 -3.57633322e-01 7.58195594e-02
-5.93434989e-01 -3.04133505e-01 7.11967230e-01 -1.33921370e-01
1.26358700e+00 -8.35046023e-02 9.38401580e-01 1.52165639e+00
3.11238289e-01 6.50105536e-01 1.38836336e+00 -4.04996663e-01
4.36794817e-01 -9.21846926e-02 -2.13439658e-01 2.30628759e-01
-2.00652122e-01 7.34207630e-01 -4.40456480e-01 -4.86330301e-01
1.10863996e+00 -1.08068562e+00 4.17181730e-01 -3.05213511e-01
-8.19651365e-01 1.49196601e+00 2.60833979e-01 3.01744848e-01
-7.14374781e-01 2.11321697e-01 9.01317894e-01 9.45003450e-01
1.62611589e-01 1.25563633e+00 -5.58314264e-01 -9.85668719e-01
-1.77242726e-01 8.81117225e-01 1.53742003e+00 3.47292513e-01
2.66632915e-01 1.94070175e-01 7.83563126e-03 8.69234562e-01
2.57442623e-01 -9.92916152e-02 2.49402896e-01 -1.52606273e+00
3.39669496e-01 3.94815147e-01 3.46872717e-01 -6.83213890e-01
-6.43334448e-01 1.60668820e-01 -4.85495061e-01 9.07242894e-01
8.66515160e-01 -9.03443694e-01 -2.15107530e-01 1.66304183e+00
3.12273592e-01 -2.52258003e-01 6.85515821e-01 7.68754721e-01
8.29580069e-01 3.84672701e-01 4.28649545e-01 4.30498570e-02
1.47855330e+00 -8.94004345e-01 -4.37771648e-01 -1.38743639e-01
4.48694915e-01 -3.97518426e-01 8.23411942e-01 6.44822776e-01
-1.48957133e+00 -2.27093250e-01 -9.72620606e-01 1.73872724e-01
-1.51348382e-01 -9.41219389e-01 1.08423138e+00 1.11941290e+00
-1.17053366e+00 7.47269630e-01 -4.58497226e-01 -3.07227820e-01
1.21788703e-01 5.17108321e-01 -2.13803962e-01 8.17471981e-01
-1.85838163e+00 1.45089686e+00 5.94637096e-01 -3.21375877e-01
-5.77454507e-01 -2.23573923e-01 -8.37275803e-01 2.61990912e-02
5.14197290e-01 -4.88758892e-01 2.08469987e+00 -1.05131042e+00
-2.56758642e+00 1.04750371e+00 1.03542256e+00 -8.91134441e-01
8.61396134e-01 1.53522313e-01 3.93207937e-01 9.68276635e-02
2.15459839e-02 7.76991844e-01 1.32383510e-01 -1.09847343e+00
-9.11854446e-01 2.02765539e-01 1.06232631e+00 7.17844069e-01
5.68906069e-01 3.55065465e-01 4.04932082e-01 -4.86332953e-01
-5.24832428e-01 -8.07955503e-01 -5.01725316e-01 -6.84964001e-01
-1.33929832e-03 -7.75238156e-01 -5.35234846e-02 -4.70689982e-01
7.29072630e-01 -1.46642017e+00 2.25654513e-01 3.75394791e-01
3.02618265e-01 3.85550737e-01 -1.65732563e-01 8.91873360e-01
1.43575445e-01 1.37672395e-01 1.68189466e-01 2.38560304e-01
5.38132250e-01 4.17121112e-01 1.82909518e-01 -3.41969915e-02
-1.83168456e-01 1.13782752e+00 -1.23866808e+00 -2.66384840e-01
2.01968372e-01 -2.07462296e-01 -6.15320444e-01 3.12432915e-01
-5.22592247e-01 5.22814393e-01 -4.48964268e-01 1.20313996e-02
1.55559152e-01 1.83299914e-01 8.44527960e-01 5.69225430e-01
-1.29504472e-01 9.05941486e-01 -6.58864319e-01 1.58583844e+00
-2.24218592e-01 3.58280689e-01 3.56561124e-01 -9.41158116e-01
9.02618885e-01 6.98711157e-01 3.15842152e-01 -9.27811682e-01
8.04464757e-01 1.26164174e-02 8.22208405e-01 -3.66068557e-02
4.48785782e-01 -5.10769010e-01 -6.54886007e-01 1.21263957e+00
-1.81199405e-02 -8.14485312e-01 2.57157326e-01 8.59307349e-02
1.26479852e+00 5.19790351e-01 7.26407051e-01 -2.79284298e-01
4.14424479e-01 4.75063324e-01 4.90296841e-01 1.31426334e+00
-7.15707362e-01 -2.46569619e-01 1.06650150e+00 -7.59624302e-01
-1.22204804e+00 -8.16730380e-01 4.18402374e-01 1.54001522e+00
1.45806313e-01 -4.28351283e-01 -9.38819528e-01 -6.88798666e-01
-3.70432526e-01 8.36654782e-01 -6.99758291e-01 -2.12794989e-02
-8.66887867e-01 -4.51250702e-01 1.16816676e+00 2.70409584e-01
1.06789517e+00 -1.91684628e+00 -1.13222718e+00 7.11260736e-01
-3.94190162e-01 -6.40400708e-01 6.95251450e-02 1.71846718e-01
-1.29903153e-01 -1.21281147e+00 -1.45364776e-01 -7.54302144e-01
-5.37020028e-01 -2.89308131e-01 1.43349862e+00 4.15076137e-01
1.44541129e-01 3.04488301e-01 -6.78526223e-01 -7.41912961e-01
-1.19493330e+00 4.24467236e-01 -8.86822790e-02 -9.84690249e-01
6.40035033e-01 -7.23593950e-01 -5.84387362e-01 3.71112734e-01
-4.22883600e-01 4.14939940e-01 2.87359178e-01 1.15175402e+00
-7.19396174e-01 -2.30063558e-01 6.70858264e-01 -1.14354455e+00
1.66043019e+00 -3.08233172e-01 -4.97075647e-01 3.00592603e-03
-4.22872365e-01 4.44610370e-03 4.72273946e-01 -2.69976288e-01
-1.20826769e+00 -2.99528778e-01 -4.99194235e-01 5.98132133e-01
-2.03805611e-01 1.86132923e-01 1.28999054e-01 -8.26907530e-02
1.20939171e+00 -1.08815514e-01 4.10786837e-01 -1.22759249e-02
3.23188812e-01 6.91759586e-01 4.46880221e-01 -1.18002331e+00
6.57723606e-01 -1.94587752e-01 -6.00874901e-01 -3.94417465e-01
-1.02382734e-01 -9.15594622e-02 -4.64768559e-01 -3.60102147e-01
9.42730963e-01 -4.67072427e-01 -1.67679918e+00 6.74366236e-01
-1.11572814e+00 -9.87903833e-01 -2.17644900e-01 1.60835296e-01
-1.16859257e+00 1.22549459e-01 -1.22030210e+00 -8.25368881e-01
-3.91260028e-01 -8.89211833e-01 4.28306758e-01 6.35682702e-01
-8.84392738e-01 -9.93725181e-01 5.71454883e-01 6.02810383e-01
6.58458710e-01 2.08031952e-01 7.72410333e-01 -1.23699450e+00
1.79490700e-01 1.22704089e-01 6.32063369e-04 3.02137174e-02
1.19089760e-01 -3.46862972e-01 -5.97639441e-01 1.35424718e-01
-1.28128931e-01 -9.50383961e-01 -1.91244677e-01 1.15331978e-01
-1.30945578e-01 -4.87811208e-01 1.58345714e-01 -1.61482260e-01
7.07624435e-01 9.15643871e-01 8.91021729e-01 9.14769590e-01
-1.85179338e-01 9.23730612e-01 6.55973792e-01 4.71651614e-01
6.65359974e-01 6.93596303e-01 2.30956286e-01 1.69181183e-01
4.49149489e-01 -2.11129010e-01 3.00290227e-01 1.76404312e-01
-8.14575613e-01 -3.80300321e-02 -1.10553479e+00 1.00292683e-01
-2.11020780e+00 -1.42617178e+00 3.12671870e-01 1.37213957e+00
1.41196990e+00 6.14904583e-01 8.06256592e-01 7.93176442e-02
6.04398549e-01 2.71743208e-01 -2.68040389e-01 -1.34678400e+00
2.14284047e-01 1.03870475e+00 1.96064994e-01 7.66735315e-01
-1.01754475e+00 1.52439642e+00 6.98730373e+00 6.68722093e-01
-5.93434155e-01 -2.52451673e-02 4.29940939e-01 2.08981782e-01
9.57390442e-02 2.57746112e-02 9.32953656e-02 1.09443754e-01
1.00606287e+00 -5.64859092e-01 7.96568871e-01 6.95630848e-01
2.96610683e-01 -2.52034962e-01 -8.74505877e-01 3.54005277e-01
-3.71786803e-01 -1.53905582e+00 -3.64546806e-01 1.09920785e-01
2.59820700e-01 -3.71527523e-01 -3.68784696e-01 1.03396153e+00
1.61464119e+00 -1.14935231e+00 6.02311194e-01 -1.99111700e-01
1.80069298e-01 -8.17121387e-01 1.04344738e+00 5.65247059e-01
-7.40864575e-01 1.00021571e-01 -2.28847533e-01 -8.36009800e-01
-4.19448689e-03 -9.88425791e-01 -1.24872506e+00 5.00417411e-01
4.47692692e-01 -9.39492807e-02 1.96211249e-01 4.87343013e-01
-6.53944254e-01 7.20054567e-01 -8.70634466e-02 -6.43058479e-01
6.77254498e-01 -6.08922124e-01 3.60150486e-01 1.07190490e+00
-8.15765619e-01 1.01817131e+00 3.02193403e-01 5.89973986e-01
8.14225823e-02 1.85535744e-01 -4.02794570e-01 9.84257087e-02
4.54025030e-01 1.02632701e+00 -9.00803149e-01 -1.87969267e-01
-1.42973468e-01 8.41384172e-01 1.70693442e-01 -1.15119137e-01
-7.38814831e-01 -3.16816062e-01 9.15335357e-01 -3.44215065e-01
-7.25151002e-02 -7.02994168e-02 -2.29613006e-01 -6.72042727e-01
-7.11556733e-01 -1.46663272e+00 4.70828891e-01 -6.01668537e-01
-1.43825555e+00 5.69943845e-01 2.79040784e-02 -7.19849944e-01
-9.57642496e-01 -5.48919439e-01 -9.09326136e-01 6.95963562e-01
-6.91975653e-01 -9.09921110e-01 1.12706676e-01 3.01782370e-01
6.34837568e-01 -5.01910210e-01 1.16489875e+00 -4.72549319e-01
5.90345648e-04 5.48197091e-01 -3.77269208e-01 7.30267048e-01
3.88186604e-01 -1.54655170e+00 8.81917179e-01 -7.46053606e-02
-1.04807995e-01 4.41800505e-01 8.73549104e-01 -3.95908475e-01
-7.26307392e-01 1.09180555e-01 4.74615216e-01 -4.92629826e-01
9.15324628e-01 -3.49922895e-01 -5.06396532e-01 5.26973665e-01
1.07214570e+00 -1.02661443e+00 8.27893376e-01 2.64406353e-01
-1.30540326e-01 5.50022542e-01 -1.42266273e+00 8.79743576e-01
1.00764096e+00 -4.65470165e-01 -1.47326279e+00 3.05227879e-02
2.74055272e-01 -7.79500902e-01 -7.64407635e-01 6.69277981e-02
8.78536344e-01 -1.23555982e+00 8.93408895e-01 -1.28024828e+00
6.94969952e-01 3.11670303e-01 3.59506547e-01 -1.64007306e+00
-2.23365158e-01 -1.09980738e+00 6.78973138e-01 6.51275992e-01
2.79290617e-01 -6.67403996e-01 1.24057388e+00 1.00216329e+00
4.71956760e-01 -2.79768199e-01 -1.15628934e+00 -3.77296776e-01
8.97388458e-01 -2.06870258e-01 4.34918344e-01 1.09294963e+00
9.15924191e-01 8.14763248e-01 -4.75160003e-01 -4.99822855e-01
2.88690507e-01 -2.33353436e-01 9.37945783e-01 -1.27706337e+00
-4.95957166e-01 -8.87998641e-01 -4.46409941e-01 -8.38572264e-01
2.92415589e-01 -4.98728305e-01 2.67416656e-01 -1.46063530e+00
-1.87743455e-01 -5.01243472e-01 6.79283813e-02 7.70725667e-01
8.30713138e-02 1.08584419e-01 3.25001210e-01 -1.62103280e-01
-5.41093111e-01 2.84903347e-01 1.07917023e+00 -6.14021905e-02
-3.09604436e-01 2.89984316e-01 -9.22014654e-01 8.50709319e-01
1.05447364e+00 -2.31328130e-01 -3.48146707e-01 1.63135156e-01
4.73555237e-01 6.87276542e-01 -4.68036495e-02 -4.53198642e-01
4.94030416e-01 -6.33542717e-01 -1.38555244e-01 3.95100564e-01
1.95792764e-01 -1.69038370e-01 -1.28563821e-01 8.54759634e-01
-6.67057574e-01 2.20586941e-01 2.61583567e-01 1.07744768e-01
1.42986542e-02 -3.35414201e-01 7.19056964e-01 -6.73363388e-01
-6.45335734e-01 -4.07781661e-01 -1.34891760e+00 1.79150194e-01
1.01604950e+00 -4.24789906e-01 -4.41181630e-01 -1.31479752e+00
-7.88701892e-01 5.16305387e-01 3.39948446e-01 3.92563105e-01
8.83901492e-02 -8.68799269e-01 -1.10428011e+00 -4.25211161e-01
-2.54060328e-01 -2.95828074e-01 -1.74182177e-01 -2.74141096e-02
-1.27225256e+00 1.73527151e-01 -1.07302141e+00 2.06347302e-01
-1.34565890e+00 -1.63897619e-01 8.24943364e-01 -6.86723590e-01
-6.52753651e-01 9.41613138e-01 -1.50436372e-01 -1.01468968e+00
1.99939430e-01 1.83445096e-01 -5.37489891e-01 -8.28487352e-02
3.44696164e-01 1.80174127e-01 -3.47656429e-01 -2.26470500e-01
-8.35277513e-02 -2.34224379e-01 -8.38026926e-02 -8.18755805e-01
1.26926112e+00 1.61603704e-01 1.09469406e-01 1.08850077e-01
3.37678716e-02 -9.07805860e-02 -1.07852745e+00 -2.44118974e-01
4.95763838e-01 -3.29879582e-01 -7.30463862e-01 -1.36686647e+00
-3.71794343e-01 1.75703868e-01 5.66881821e-02 7.41319537e-01
5.13686180e-01 -4.12730217e-01 5.23042679e-01 7.65986621e-01
9.59456980e-01 -1.27821505e+00 1.38578355e-01 9.61314619e-01
7.97788858e-01 -1.05980682e+00 -2.01789886e-01 -1.40485868e-01
-1.25572598e+00 1.18185759e+00 8.42696548e-01 -4.79938000e-01
1.74477652e-01 3.74795794e-01 6.33919120e-01 -2.16177389e-01
-1.08635342e+00 -1.80793315e-01 -6.48286164e-01 8.71917844e-01
2.27913752e-01 7.88697079e-02 -8.61671984e-01 7.87725389e-01
-1.01086986e+00 -2.91502085e-02 8.48913610e-01 1.10860288e+00
-2.59710610e-01 -1.71460342e+00 -2.22103328e-01 7.07010999e-02
-3.81366938e-01 1.23163797e-01 -1.24335921e+00 1.33044100e+00
9.57358778e-02 1.23768485e+00 1.03476364e-02 -3.84811342e-01
5.07516623e-01 1.53002620e-01 7.16202259e-01 -6.70476556e-01
-1.79651403e+00 -3.89740884e-01 8.04062843e-01 -4.95041817e-01
-5.04368603e-01 -5.53278029e-01 -1.27877903e+00 -1.03258300e+00
-8.17993954e-02 6.53109908e-01 2.08614588e-01 9.76074517e-01
-4.52872902e-01 3.23798731e-02 4.36479598e-01 -7.17587292e-01
-8.92979205e-01 -1.19432449e+00 -4.85585481e-01 4.63527650e-01
-4.99986649e-01 -4.82440263e-01 -1.16572000e-01 -5.69334447e-01] | [3.6688332557678223, 1.5448708534240723] |
7e4020ff-57f2-46b5-b63e-34f7dc91ef94 | meta-learning-initializations-for-interactive | 2210.15371 | null | https://arxiv.org/abs/2210.15371v1 | https://arxiv.org/pdf/2210.15371v1.pdf | Meta-Learning Initializations for Interactive Medical Image Registration | We present a meta-learning framework for interactive medical image registration. Our proposed framework comprises three components: a learning-based medical image registration algorithm, a form of user interaction that refines registration at inference, and a meta-learning protocol that learns a rapidly adaptable network initialization. This paper describes a specific algorithm that implements the registration, interaction and meta-learning protocol for our exemplar clinical application: registration of magnetic resonance (MR) imaging to interactively acquired, sparsely-sampled transrectal ultrasound (TRUS) images. Our approach obtains comparable registration error (4.26 mm) to the best-performing non-interactive learning-based 3D-to-3D method (3.97 mm) while requiring only a fraction of the data, and occurring in real-time during acquisition. Applying sparsely sampled data to non-interactive methods yields higher registration errors (6.26 mm), demonstrating the effectiveness of interactive MR-TRUS registration, which may be applied intraoperatively given the real-time nature of the adaptation process. | ['Dean Barratt', 'Yipeng Hu', 'Zachary M. C. Baum'] | 2022-10-27 | null | null | null | null | ['medical-image-registration'] | ['medical'] | [ 5.40571094e-01 7.80318499e-01 -2.26198342e-02 -4.43491846e-01
-1.20579040e+00 -2.93826580e-01 4.25786823e-01 3.23111802e-01
-8.49699080e-01 3.87833208e-01 1.88152686e-01 -5.25301874e-01
-5.47304809e-01 -5.26037037e-01 -6.53228104e-01 -7.68640161e-01
-8.57646406e-01 9.13652599e-01 1.10889159e-01 -3.20512652e-01
1.19855307e-01 5.26903987e-01 -1.14991319e+00 2.91435719e-01
9.22583759e-01 4.68721271e-01 3.87082785e-01 8.91102850e-01
1.33425966e-01 5.23048222e-01 -2.26650611e-01 1.29116595e-01
3.71706247e-01 -3.84264350e-01 -1.13883221e+00 -4.23271567e-01
5.46937764e-01 -1.74340531e-01 -3.14748853e-01 8.47113311e-01
9.71029639e-01 4.74104732e-01 5.15030921e-01 -4.43152964e-01
2.45692298e-01 8.48980904e-01 -2.78088599e-01 2.18023315e-01
4.45244163e-01 -1.16716392e-01 6.18408993e-02 -4.27771151e-01
8.79932463e-01 4.44313169e-01 1.01033044e+00 7.09388673e-01
-1.12649906e+00 -4.27651495e-01 -5.21064281e-01 -1.37691945e-01
-1.11247206e+00 -3.09689850e-01 4.17914867e-01 -5.14615357e-01
7.91288853e-01 6.23629093e-01 7.65440106e-01 6.33360088e-01
8.92167509e-01 2.29790479e-01 1.41383553e+00 -4.76443797e-01
7.92836398e-02 -1.04638427e-01 3.84055786e-02 8.53348434e-01
-3.94441560e-02 4.50001240e-01 -1.64979815e-01 -3.97367239e-01
1.30157447e+00 -4.07856144e-02 -3.01759481e-01 -3.50040853e-01
-1.46475351e+00 3.20504487e-01 4.43288177e-01 7.20978737e-01
-7.87175953e-01 9.21516269e-02 6.02853000e-01 4.31088716e-01
2.83738941e-01 5.93463540e-01 -3.14405859e-01 -2.79650271e-01
-1.08471394e+00 -3.80284548e-01 9.09637451e-01 7.22915351e-01
4.65612084e-01 -8.22512135e-02 -1.03167586e-01 7.69510269e-01
1.60074696e-01 -1.30603155e-02 9.26517904e-01 -9.57812190e-01
-3.60648446e-02 1.43416539e-01 -1.42787650e-01 -6.71263814e-01
-8.72850716e-01 -4.54497278e-01 -1.04013336e+00 3.12927097e-01
1.47252172e-01 -3.75428855e-01 -1.05645204e+00 1.22987020e+00
4.75471318e-01 6.69529796e-01 6.03876486e-02 9.50748801e-01
1.17252743e+00 1.33612394e-01 1.25719439e-02 -5.95176041e-01
1.27757037e+00 -7.14425325e-01 -7.22857058e-01 1.03497468e-01
9.69859362e-01 -7.23674476e-01 8.19918215e-01 2.12596476e-01
-1.44976103e+00 -5.09704411e-01 -8.87947917e-01 4.50840324e-01
1.86211020e-01 -3.08164775e-01 7.50743985e-01 7.30007052e-01
-1.39238787e+00 8.73861492e-01 -1.33656275e+00 -7.15836808e-02
4.05470319e-02 1.11179101e+00 -6.14606321e-01 3.83338667e-02
-1.05175233e+00 1.34851646e+00 4.03640240e-01 1.60593912e-01
-8.35438430e-01 -1.18934500e+00 -9.28759873e-01 -5.94152868e-01
2.39618182e-01 -8.18611085e-01 1.03878701e+00 -8.52397799e-01
-1.74095476e+00 1.23459911e+00 2.65573785e-02 -3.56780320e-01
6.76360250e-01 -9.85919163e-02 -4.18639243e-01 1.66402072e-01
-4.98203486e-01 3.50291938e-01 4.88715649e-01 -1.47222495e+00
-1.19841076e-01 -3.29943299e-01 6.52532279e-02 5.42504311e-01
3.61230135e-01 -2.41566822e-01 -4.27111268e-01 -5.02391696e-01
6.57065928e-01 -1.31871474e+00 -8.23268235e-01 -3.98399740e-01
-5.79082966e-02 3.22591603e-01 2.54953027e-01 -7.53957093e-01
8.97687912e-01 -1.85915136e+00 9.43961143e-02 6.09974444e-01
3.16953123e-01 1.10773891e-01 -1.36281043e-01 1.57199666e-01
-5.58662593e-01 -2.51224935e-01 -3.18680674e-01 -9.09172893e-02
-7.47296989e-01 2.01742202e-01 4.99868840e-01 8.30063343e-01
-7.86076009e-01 8.90289366e-01 -1.26241076e+00 -6.21720433e-01
4.75598425e-01 5.49998879e-01 -5.21906137e-01 5.87856770e-01
3.90441179e-01 1.31733239e+00 -1.95871204e-01 3.18060279e-01
4.50235277e-01 1.27553791e-01 5.17189622e-01 -8.30908477e-01
-1.53335288e-01 -9.41208154e-02 -9.90202129e-01 2.43415952e+00
-9.52788711e-01 3.68890464e-01 2.18957394e-01 -1.01676381e+00
8.63537848e-01 6.60938799e-01 1.28212452e+00 -6.36381567e-01
2.48237327e-01 2.47090414e-01 3.10462892e-01 -9.27216589e-01
3.85774940e-01 -3.01096708e-01 3.20406854e-01 7.73086190e-01
3.20882618e-01 -3.09727520e-01 -3.45622331e-01 -7.39437863e-02
9.95305121e-01 1.23235643e-01 2.99024314e-01 -6.47099078e-01
5.03533900e-01 -3.63554917e-02 1.35172576e-01 1.06355333e+00
-6.32977560e-02 5.95423460e-01 -2.72706002e-01 -8.82131636e-01
-7.95370281e-01 -8.84210050e-01 -2.89968044e-01 8.47754598e-01
3.70063961e-01 1.86033518e-04 -7.03596115e-01 -5.82935572e-01
-5.17732620e-01 4.70337957e-01 -8.39966178e-01 -9.99058113e-02
-1.11665964e+00 -8.85425985e-01 1.40019551e-01 1.66692138e-01
-8.85145515e-02 -9.89650965e-01 -7.71065712e-01 3.16913217e-01
-1.28791705e-01 -5.86795211e-01 -5.11260927e-01 2.71098286e-01
-1.57094073e+00 -1.06092250e+00 -7.31675982e-01 -8.96023750e-01
1.11349261e+00 -3.25628728e-01 1.08904600e+00 2.09545195e-01
-4.45695579e-01 9.48560417e-01 -1.10305749e-01 3.63859092e-03
-9.66557622e-01 9.56282243e-02 4.23868373e-03 -2.56438911e-01
-3.99014801e-01 -7.66129017e-01 -8.34658563e-01 1.74244300e-01
-8.53680253e-01 2.00506940e-01 6.79013550e-01 9.93823647e-01
7.63250649e-01 -5.07834852e-01 2.44385839e-01 -1.53422892e+00
4.47631419e-01 -2.84593642e-01 -3.72811109e-01 3.83335590e-01
-6.50874972e-01 4.09499696e-03 6.98984191e-02 -6.49275064e-01
-1.12813890e+00 3.99574965e-01 -2.30488256e-01 -7.59910271e-02
-2.69685209e-01 7.30852485e-01 6.95522547e-01 -9.01705325e-01
1.06450653e+00 3.43704760e-01 5.09347498e-01 -5.49703725e-02
3.79624844e-01 4.37653244e-01 8.36970091e-01 -4.76824701e-01
5.73052704e-01 4.60656524e-01 8.55891705e-02 -5.60365498e-01
-3.26251894e-01 -4.33856905e-01 -1.12448502e+00 -5.63524365e-01
5.44409752e-01 -6.12084031e-01 -7.52087295e-01 1.57166034e-01
-8.20316792e-01 -6.76619411e-01 -4.43681091e-01 9.89073157e-01
-9.00931656e-01 2.68024772e-01 -7.00385153e-01 -4.54246581e-01
-8.28385711e-01 -1.50166655e+00 6.54504061e-01 -1.70668564e-03
-2.90981382e-01 -1.54801345e+00 3.75774056e-01 2.28499904e-01
8.37588429e-01 7.49703050e-01 7.40036130e-01 -7.36703992e-01
-2.98560411e-01 -3.59937340e-01 3.50375324e-01 -1.59522980e-01
4.77467358e-01 -6.30564570e-01 -8.01268041e-01 -6.05879962e-01
2.52777338e-01 -5.61622605e-02 2.71383226e-01 8.14027250e-01
1.10355711e+00 -8.43044668e-02 -3.05277944e-01 8.19623590e-01
1.40547216e+00 4.61802870e-01 6.60027444e-01 2.50542432e-01
5.33796310e-01 3.77822936e-01 5.75874448e-01 3.01562399e-01
3.66216242e-01 5.95300972e-01 4.19436902e-01 -6.12412453e-01
-7.58148879e-02 3.41706604e-01 -1.95971996e-01 1.20572269e+00
-6.84267044e-01 7.70750523e-01 -1.26514006e+00 2.06043184e-01
-1.62777698e+00 -7.80314147e-01 1.14097238e-01 2.64364123e+00
1.37518573e+00 -1.96833536e-01 -2.70216495e-01 -4.29805517e-01
5.55063248e-01 4.98986952e-02 -3.10553789e-01 -2.70379007e-01
5.92378497e-01 5.52345872e-01 5.76543510e-01 9.32418704e-01
-1.01262438e+00 4.54128712e-01 7.02732611e+00 4.16798562e-01
-1.44991159e+00 5.36028862e-01 5.62007785e-01 2.84095615e-01
-2.79339612e-01 -3.04325581e-01 5.63216694e-02 -1.38510302e-01
1.08339345e+00 -1.94113776e-01 5.15239179e-01 4.84175980e-01
1.59546956e-01 -2.77976155e-01 -1.26721096e+00 1.05451298e+00
2.41951048e-01 -1.67446613e+00 -2.61267871e-01 -2.86064923e-01
7.46445179e-01 2.49580130e-01 -2.44871750e-01 1.42684922e-01
4.33578081e-02 -1.15694988e+00 -3.98283713e-02 8.20625901e-01
1.24939680e+00 -6.17791533e-01 8.62425745e-01 3.00964385e-01
-8.36224973e-01 3.76392066e-01 7.99243525e-02 5.19638836e-01
5.72526865e-02 4.58847880e-01 -1.09830832e+00 8.28106940e-01
3.78284872e-01 3.62996787e-01 -2.11797297e-01 1.20954442e+00
4.32516187e-01 2.37543091e-01 -1.99410141e-01 5.23968220e-01
-7.00878352e-02 -1.39643326e-01 6.32717490e-01 1.30428255e+00
1.20818736e-02 4.90463763e-01 2.58424938e-01 2.88930148e-01
1.91007391e-01 1.32397458e-01 -4.38728571e-01 7.83531368e-01
2.81068653e-01 1.49735940e+00 -6.36340916e-01 -3.37320685e-01
-7.43018538e-02 9.34682071e-01 -9.53383669e-02 2.04480082e-01
-5.04366100e-01 -2.28683293e-01 -2.08766297e-01 2.24240094e-01
-4.70323652e-01 -2.33437657e-01 -3.17259818e-01 -8.41817200e-01
-4.93068993e-01 -8.69366288e-01 4.71190542e-01 -3.61402661e-01
-8.07526946e-01 1.08644390e+00 3.19913357e-01 -1.29864299e+00
-5.59592962e-01 -2.76549794e-02 -5.88211715e-01 8.95287454e-01
-1.11105108e+00 -9.99167264e-01 -6.94493890e-01 6.44358337e-01
2.53150821e-01 -1.15611898e-02 1.31926894e+00 3.41461688e-01
1.67216167e-01 6.08630419e-01 -1.25717759e-01 7.59980083e-02
8.00436378e-01 -1.28783441e+00 1.70544237e-02 8.03838670e-02
-1.46940395e-01 8.70726228e-01 5.41840076e-01 -5.95554471e-01
-1.73482847e+00 -8.79431367e-01 2.95938849e-01 -2.52922446e-01
2.39874586e-01 3.14870983e-01 -8.38464022e-01 7.37068295e-01
3.94984037e-01 4.93656814e-01 9.26156461e-01 -7.14461179e-03
4.15181398e-01 -1.71750352e-01 -1.66762006e+00 6.43267691e-01
8.69976044e-01 -3.52639794e-01 -7.17423856e-01 5.26612163e-01
3.59850079e-01 -1.54439187e+00 -1.75531352e+00 7.26875544e-01
6.88106656e-01 -6.99341536e-01 1.06971538e+00 -2.62212515e-01
-9.93983522e-02 3.23350914e-02 4.99355733e-01 -1.44631135e+00
-1.44188240e-01 -1.09212244e+00 1.56540066e-01 3.09306979e-01
3.35103899e-01 -8.33240509e-01 7.52404451e-01 7.07482994e-01
-5.38170159e-01 -8.27771485e-01 -1.24719322e+00 -2.54540890e-01
-9.27618518e-02 -3.60526204e-01 5.11425510e-02 1.48179436e+00
3.21007490e-01 -5.41161776e-01 -1.18883617e-01 3.56480688e-01
8.98610234e-01 -6.76583592e-03 3.17953706e-01 -1.06301856e+00
-3.33866596e-01 9.19463411e-02 -3.48980635e-01 -5.29027164e-01
2.72949003e-02 -1.12327969e+00 2.43211716e-01 -1.28767490e+00
1.51591986e-01 -1.06092358e+00 -4.65685576e-01 4.82356340e-01
-9.11446512e-02 2.64847845e-01 -2.57784128e-01 3.90004396e-01
-2.95346856e-01 -5.47989868e-02 1.57711089e+00 1.85101166e-01
-6.12416387e-01 1.98416874e-01 -1.27886340e-01 5.27963340e-01
7.26849973e-01 -5.53501666e-01 -3.42695028e-01 -3.18019569e-01
-1.61604509e-01 7.19789028e-01 1.16085947e-01 -9.02778208e-01
8.74695241e-01 6.56163916e-02 2.03568190e-01 -3.10712367e-01
1.15180258e-02 -9.36369836e-01 5.95163524e-01 9.96910930e-01
-4.34946835e-01 1.67686895e-01 3.80179107e-01 1.17767408e-01
-1.21152230e-01 -3.81858885e-01 1.00373423e+00 -4.90340233e-01
-3.87241811e-01 4.92605209e-01 -3.56979668e-01 -7.11531043e-02
8.36528063e-01 -4.13526207e-01 3.39931011e-01 -1.84605449e-01
-1.53027833e+00 -2.50821352e-01 6.67874888e-02 9.70146656e-02
7.35888958e-01 -9.69858527e-01 -9.13965642e-01 3.05540740e-01
-3.67133677e-01 2.03781113e-01 6.41667843e-01 1.46144128e+00
-9.60178733e-01 -7.68616721e-02 -3.31200361e-01 -1.01191437e+00
-1.20210528e+00 -4.56996262e-02 1.03430164e+00 -3.40495020e-01
-9.07868147e-01 6.01696849e-01 -3.09291184e-01 -8.56642485e-01
1.34727895e-01 -1.42860729e-02 -1.84714735e-01 -3.58819515e-01
2.42859051e-01 2.29160547e-01 4.67485160e-01 -4.05292273e-01
-2.88043290e-01 7.27347553e-01 -2.02292964e-01 -1.37992218e-01
1.38719666e+00 -6.20938698e-03 -2.30090827e-01 3.52533489e-01
8.26682985e-01 -1.70050964e-01 -8.56000781e-01 -4.53793079e-01
5.03777340e-02 -2.27141261e-01 4.92338359e-01 -9.35571790e-01
-1.31167185e+00 3.29094768e-01 1.27895188e+00 -4.03076231e-01
1.06530476e+00 -2.24549081e-02 4.71120626e-01 3.21875632e-01
6.95402741e-01 -7.46711969e-01 -1.72245353e-01 3.49930733e-01
9.31826591e-01 -1.25784028e+00 3.29961717e-01 -3.00786197e-01
-5.38299143e-01 1.28095806e+00 3.62102896e-01 -1.92029163e-01
9.35297489e-01 7.63872504e-01 5.70625901e-01 -3.10270131e-01
-2.55432218e-01 3.89025271e-01 6.57045007e-01 8.24429452e-01
8.43587399e-01 -2.30707638e-02 -3.85425985e-01 4.60619889e-02
5.73740304e-02 1.37375101e-01 4.24353361e-01 1.03438675e+00
-1.20488323e-01 -1.08183444e+00 -1.70605272e-01 4.83631551e-01
-5.25747597e-01 -2.85499036e-01 5.58682740e-01 6.74232781e-01
-1.29069850e-01 3.16314995e-01 5.49234375e-02 -1.28678530e-01
4.41891700e-01 -3.08967620e-01 8.66166949e-01 -7.92829633e-01
-1.41078460e+00 1.41060710e-01 1.92205030e-02 -9.45127547e-01
-7.82296896e-01 -5.16412437e-01 -1.48655868e+00 1.47533432e-01
-2.79014826e-01 2.82305807e-01 9.40983534e-01 8.90254676e-01
3.10765266e-01 6.60841882e-01 7.43999183e-01 -1.34267521e+00
-1.68936938e-01 -9.12113011e-01 -3.71740580e-01 3.07807207e-01
4.10896361e-01 -1.94300979e-01 -7.48057514e-02 1.69059515e-01] | [13.876609802246094, -2.6256189346313477] |
dcfce9d7-7687-4078-afa2-f9b05f46d45b | balancing-lexical-and-semantic-quality-in | 2305.09898 | null | https://arxiv.org/abs/2305.09898v1 | https://arxiv.org/pdf/2305.09898v1.pdf | Balancing Lexical and Semantic Quality in Abstractive Summarization | An important problem of the sequence-to-sequence neural models widely used in abstractive summarization is exposure bias. To alleviate this problem, re-ranking systems have been applied in recent years. Despite some performance improvements, this approach remains underexplored. Previous works have mostly specified the rank through the ROUGE score and aligned candidate summaries, but there can be quite a large gap between the lexical overlap metric and semantic similarity. In this paper, we propose a novel training method in which a re-ranker balances the lexical and semantic quality. We further newly define false positives in ranking and present a strategy to reduce their influence. Experiments on the CNN/DailyMail and XSum datasets show that our method can estimate the meaning of summaries without seriously degrading the lexical aspect. More specifically, it achieves an 89.67 BERTScore on the CNN/DailyMail dataset, reaching new state-of-the-art performance. Our code is publicly available at https://github.com/jeewoo1025/BalSum. | ['Yong Suk Choi', 'Jeewoo Sul'] | 2023-05-17 | null | null | null | null | ['abstractive-text-summarization', 'semantic-textual-similarity', 'semantic-similarity'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 1.69192016e-01 -1.67077661e-01 -4.66039181e-01 -5.72189629e-01
-1.02022862e+00 -4.84404355e-01 5.14935434e-01 2.82255501e-01
-6.76192462e-01 8.92853916e-01 7.95142949e-01 -6.31446987e-02
5.82207926e-02 -7.23208368e-01 -5.55002451e-01 -4.05449808e-01
3.41332138e-01 8.51268321e-02 2.78256088e-01 -1.56973436e-01
7.05521643e-01 -8.37971717e-02 -1.63614035e+00 6.77130401e-01
1.24586236e+00 7.90011168e-01 2.81919301e-01 5.85986733e-01
-3.62593293e-01 7.01801598e-01 -9.95506465e-01 -5.75568438e-01
-1.07467182e-01 -6.78593516e-01 -7.44959533e-01 -4.17149991e-01
6.45456314e-01 -3.08543772e-01 -4.07995284e-01 1.19976091e+00
9.16424632e-01 1.58125028e-01 5.56313276e-01 -9.41734493e-01
-7.06678987e-01 1.22451794e+00 -5.08609176e-01 3.68873745e-01
2.98022449e-01 -1.89037204e-01 1.40924418e+00 -8.11635792e-01
4.16729897e-01 1.05557859e+00 6.01444423e-01 5.27280390e-01
-8.63583267e-01 -7.60550261e-01 2.34356925e-01 4.68682140e-01
-1.19409192e+00 -4.45254117e-01 6.16517067e-01 -1.69885457e-01
1.01964927e+00 5.36536098e-01 5.21092296e-01 1.16621816e+00
1.58645555e-01 9.85224426e-01 8.02008867e-01 -2.27021128e-01
1.17893271e-01 -1.39305979e-01 4.74440366e-01 5.40293813e-01
4.37783480e-01 -3.25596601e-01 -7.68775225e-01 -2.08962247e-01
3.43181312e-01 -5.67195415e-02 -5.51388979e-01 1.42433152e-01
-1.15471458e+00 8.06077540e-01 3.39727759e-01 3.58652055e-01
-2.58732051e-01 2.11129323e-01 7.46087313e-01 1.87428266e-01
6.85521722e-01 6.86082721e-01 -3.89469683e-01 -3.77066106e-01
-1.20173728e+00 2.71735668e-01 7.65151501e-01 7.64058471e-01
3.95803005e-01 -2.38882136e-02 -4.64931101e-01 1.18734562e+00
-1.14287116e-01 2.77167767e-01 7.93628573e-01 -8.28061163e-01
4.90716398e-01 4.45254147e-01 -1.86131015e-01 -1.18289375e+00
-2.49181449e-01 -6.22689366e-01 -1.05623245e+00 -4.15443599e-01
8.89414847e-02 8.40998255e-03 -6.94325805e-01 1.62738323e+00
-1.82628736e-01 1.33548915e-01 -1.05732046e-01 8.68066072e-01
1.20723557e+00 7.40610838e-01 2.46720924e-03 -2.10540697e-01
1.23795772e+00 -1.04492640e+00 -8.93841505e-01 -1.00809410e-01
5.05932689e-01 -7.90732265e-01 1.26657867e+00 4.49068248e-01
-1.10547280e+00 -3.56812894e-01 -1.20984173e+00 -2.09127814e-02
-1.69950515e-01 3.72396737e-01 5.04236281e-01 4.11930948e-01
-1.05150092e+00 1.02387059e+00 -6.61092758e-01 -5.03048003e-01
2.99040347e-01 1.32419512e-01 -1.09956972e-02 2.11190805e-01
-1.42150259e+00 8.02489817e-01 7.47746646e-01 -1.69326678e-01
-5.32786846e-01 -5.10454834e-01 -4.34853405e-01 2.93750048e-01
4.68629807e-01 -7.83457220e-01 1.45639169e+00 -8.63716006e-01
-1.48108399e+00 6.55053675e-01 -1.50155246e-01 -5.52072763e-01
4.71274942e-01 -4.72477257e-01 -3.69525462e-01 -1.18319415e-01
9.12424028e-02 5.44491112e-01 3.24082881e-01 -1.06044185e+00
-8.19055438e-01 -1.03484489e-01 1.90239884e-02 3.87675196e-01
-6.34335220e-01 1.63216874e-01 -6.28087223e-01 -9.30810750e-01
-1.73258260e-02 -6.84404969e-01 -8.67765695e-02 -5.24665475e-01
-6.75860226e-01 -4.50773239e-01 1.02859847e-01 -6.95361793e-01
1.81390166e+00 -1.68786478e+00 -2.36343481e-02 -2.47430593e-01
2.53559023e-01 4.61344123e-01 -2.91605055e-01 6.73802257e-01
8.51979405e-02 3.60274255e-01 -3.89333606e-01 -2.71721542e-01
-1.03125818e-01 -1.79637060e-01 -4.81754333e-01 1.42291591e-01
1.10458195e-01 8.57249022e-01 -1.07960045e+00 -4.30532515e-01
-2.66365498e-01 2.69316137e-01 -4.97070014e-01 1.95306957e-01
-2.13837102e-01 -5.12574986e-02 -4.43567365e-01 4.70394015e-01
4.62095588e-01 -3.47012341e-01 1.33376747e-01 -1.30793259e-01
-1.18944407e-01 7.38230109e-01 -7.02470660e-01 1.82573581e+00
-1.67869255e-01 7.06778646e-01 -3.70371222e-01 -9.15006876e-01
8.85539770e-01 6.90404102e-02 2.41185591e-01 -6.83102965e-01
1.58249915e-01 3.82961661e-01 -1.40989001e-03 -4.61922020e-01
1.18162346e+00 4.96521220e-02 -2.17224192e-03 3.96915138e-01
-1.07949793e-01 9.00199413e-02 4.11034584e-01 3.43966663e-01
1.21936584e+00 8.11918005e-02 5.03198504e-01 -2.48430878e-01
3.40123892e-01 4.27312367e-02 6.16240859e-01 9.06819344e-01
-6.03464199e-03 9.70251560e-01 6.06878877e-01 -1.83391690e-01
-9.11757529e-01 -7.88992882e-01 1.48930117e-01 1.24755538e+00
6.33410290e-02 -7.92606711e-01 -8.15250576e-01 -7.65596747e-01
-1.32596180e-01 1.02639925e+00 -3.79816204e-01 -3.40130955e-01
-7.01936603e-01 -8.78552914e-01 9.01859164e-01 5.54271996e-01
5.16319394e-01 -1.11544979e+00 -5.21048546e-01 2.15202436e-01
-6.27554297e-01 -7.78003693e-01 -6.82497859e-01 -3.90396900e-02
-1.00409973e+00 -8.08154702e-01 -8.77738655e-01 -4.91728783e-01
3.43600750e-01 3.53369415e-01 1.36000681e+00 1.98952049e-01
1.81951657e-01 -1.16139866e-01 -5.31406879e-01 -3.96378279e-01
-3.35467726e-01 6.63550556e-01 -4.94493693e-02 -3.44447017e-01
5.05783141e-01 -4.32353556e-01 -6.74072146e-01 1.94249958e-01
-8.97899270e-01 1.22618750e-01 6.93657041e-01 8.26497436e-01
5.51185727e-01 -3.43138933e-01 1.14746141e+00 -8.59302819e-01
1.23414731e+00 -4.52120245e-01 -1.55773163e-01 2.41137296e-01
-8.55010867e-01 7.34802783e-02 7.38000512e-01 -3.90024036e-01
-8.34931433e-01 -3.92395377e-01 -2.40973264e-01 -1.69756755e-01
3.50768752e-02 8.16630781e-01 -7.93381110e-02 5.97349942e-01
6.35926247e-01 2.52744347e-01 -2.18228325e-01 -5.73806167e-01
3.22676659e-01 8.78920853e-01 4.38474834e-01 -3.57174277e-01
2.75178134e-01 2.17745706e-01 -5.37871778e-01 -7.74677098e-01
-1.14015794e+00 -5.35670161e-01 -2.32565850e-01 -1.25425681e-01
4.07491595e-01 -8.70231509e-01 -4.68627721e-01 3.51337790e-01
-1.44964373e+00 -9.18254536e-03 -1.32235199e-01 5.03774822e-01
-2.78412402e-01 6.35045230e-01 -5.57487726e-01 -4.85686183e-01
-9.26742315e-01 -8.94448400e-01 7.81523943e-01 1.93812191e-01
-5.61462641e-01 -6.50787652e-01 1.28086865e-01 2.40385517e-01
5.66092551e-01 -8.15677196e-02 8.01230490e-01 -1.05178571e+00
-2.66633511e-01 -1.42968476e-01 -3.53970557e-01 4.39897895e-01
4.21172306e-02 -1.16687603e-02 -8.70254517e-01 -1.47975951e-01
-1.12961501e-01 -1.68880641e-01 1.53946078e+00 5.14617801e-01
1.41564155e+00 -4.05293941e-01 -1.98005244e-01 2.90667504e-01
1.11595166e+00 1.51033178e-01 6.22408628e-01 3.75318766e-01
6.06533229e-01 5.67209423e-01 6.61666572e-01 4.69877392e-01
3.87247235e-01 5.21055520e-01 2.48759955e-01 1.90738305e-01
-1.70600966e-01 -3.03169698e-01 4.39337045e-01 1.38334846e+00
3.24233286e-02 -7.50008941e-01 -8.25788736e-01 5.14480114e-01
-2.12626266e+00 -1.05470669e+00 -1.21852232e-03 2.23203111e+00
1.07001674e+00 2.81515121e-01 1.61007494e-02 -2.59869713e-02
8.41188252e-01 4.15897459e-01 -5.03990233e-01 -4.59632307e-01
-3.10954005e-01 -1.98014468e-01 4.64819103e-01 3.82579923e-01
-9.27515984e-01 9.56530690e-01 5.76613188e+00 1.25027895e+00
-1.09294701e+00 -6.79133162e-02 7.17963099e-01 -3.19678277e-01
-5.33931136e-01 -2.45454758e-01 -8.98708045e-01 6.28695130e-01
9.14539039e-01 -4.10294414e-01 5.97950406e-02 6.77739441e-01
2.19972938e-01 -8.80040154e-02 -9.44511473e-01 8.56965780e-01
4.80462998e-01 -1.38818824e+00 3.79051238e-01 -1.45334229e-01
6.22509837e-01 2.26654887e-01 -1.54339939e-01 3.59075397e-01
9.40261707e-02 -9.32525337e-01 6.67832196e-01 6.41751409e-01
7.48123229e-01 -8.98087442e-01 9.32801902e-01 2.93835819e-01
-8.13230336e-01 1.67426273e-01 -5.08556306e-01 5.20615689e-02
1.06501952e-01 8.79937470e-01 -7.67929077e-01 6.01282120e-01
6.15052223e-01 8.60203147e-01 -5.52269995e-01 1.24876857e+00
-2.64126420e-01 8.12901139e-01 -2.63916310e-02 -5.23352802e-01
1.32798895e-01 -1.56077296e-01 8.14212143e-01 1.48983490e+00
4.98176277e-01 -3.00635938e-02 -2.57473383e-02 6.49406552e-01
-4.61899966e-01 3.28535289e-01 -4.08737808e-01 -1.47265524e-01
6.24619961e-01 1.06986010e+00 -6.09790802e-01 -5.65714478e-01
-1.70282945e-01 9.35116589e-01 4.22569096e-01 2.08818883e-01
-1.02810991e+00 -6.78232193e-01 5.82439065e-01 3.96216065e-02
1.81107044e-01 9.08758938e-02 -3.98751110e-01 -1.34515297e+00
1.83051109e-01 -9.60049868e-01 3.93060535e-01 -5.59568524e-01
-1.29280603e+00 7.86642849e-01 9.80555043e-02 -1.23740518e+00
-7.23127946e-02 -1.64808407e-01 -6.67927742e-01 5.27968228e-01
-1.45221353e+00 -6.62408352e-01 -2.78697044e-01 -1.72841996e-01
9.11920846e-01 -2.74756551e-01 6.18661821e-01 4.17896062e-01
-6.15064561e-01 6.47357941e-01 3.60569239e-01 2.72614714e-02
9.92671847e-01 -1.16540718e+00 5.80263853e-01 7.47404635e-01
-2.62397006e-02 7.90093660e-01 8.57832491e-01 -7.35546529e-01
-9.07569051e-01 -1.05874717e+00 1.13833189e+00 -1.87737763e-01
4.35040444e-01 -4.78478484e-02 -1.04399359e+00 2.91743338e-01
5.73452175e-01 -7.77671218e-01 7.72114277e-01 1.38588443e-01
-3.93521786e-01 -5.45767620e-02 -6.60187662e-01 8.34913552e-01
1.15986955e+00 -2.29449719e-01 -7.38917768e-01 9.78560224e-02
8.79783690e-01 -3.02912772e-01 -5.08724630e-01 5.60237885e-01
7.02676415e-01 -1.07728827e+00 6.76004291e-01 -3.98371518e-01
8.37617278e-01 -2.07831293e-01 -1.24866903e-01 -1.57470202e+00
-1.61134228e-01 -4.27951396e-01 -1.88124448e-01 1.38447464e+00
4.42123413e-01 -4.72122341e-01 7.15445876e-01 1.41587824e-01
-4.29694235e-01 -1.08271837e+00 -7.90981650e-01 -7.38295197e-01
1.34219497e-01 -2.60039449e-01 6.89832628e-01 8.89908791e-01
1.19695619e-01 6.12252772e-01 -4.34352487e-01 -4.14818436e-01
2.77227491e-01 7.64539838e-02 5.25388002e-01 -1.13819814e+00
-1.43001720e-01 -9.13046718e-01 -6.97607361e-03 -1.32036424e+00
2.16919631e-01 -1.07421756e+00 1.78157240e-01 -1.83237088e+00
5.77009797e-01 1.07645325e-01 -4.88607258e-01 3.71976018e-01
-4.62041616e-01 4.81987260e-02 2.10361984e-02 2.51238853e-01
-9.55759466e-01 7.81982422e-01 9.23188269e-01 -1.27740800e-01
-1.49496824e-01 -1.67593732e-01 -9.49510336e-01 6.75392449e-01
1.33632898e+00 -6.35946572e-01 -2.01610580e-01 -6.71958983e-01
3.92273307e-01 -2.33925268e-01 2.82722991e-03 -9.36923921e-01
2.69516736e-01 2.51656119e-02 1.12026505e-01 -1.00849414e+00
1.37620136e-01 -2.26831660e-01 -6.96624890e-02 4.84642684e-01
-8.79574537e-01 3.16726238e-01 3.03255636e-02 6.05979800e-01
-3.39024007e-01 -4.43257093e-01 4.92208809e-01 -1.59487128e-01
-5.09248316e-01 4.00141440e-03 -3.16078395e-01 1.39151424e-01
4.84598935e-01 -1.56190069e-02 -6.75245523e-01 -4.89510953e-01
-5.34378700e-02 2.81680554e-01 4.82688159e-01 5.48026443e-01
7.88252711e-01 -1.21310139e+00 -1.06138325e+00 -2.48326033e-01
1.92134693e-01 -3.14472139e-01 1.54787302e-01 7.03665614e-01
-3.44749331e-01 5.73620975e-01 -5.48163848e-03 -2.84906626e-01
-1.46823514e+00 7.11511001e-02 7.36650229e-02 -3.90320033e-01
-4.59429979e-01 7.71306217e-01 7.20365345e-03 -3.83620322e-01
4.17225987e-01 -3.21910828e-01 -5.48649132e-01 4.85142410e-01
6.72969162e-01 7.31706202e-01 2.04723641e-01 -4.25227702e-01
-2.00002953e-01 2.87227154e-01 -3.04913849e-01 -4.08822373e-02
1.22047329e+00 -8.95769820e-02 -2.84789085e-01 5.39976716e-01
1.13404489e+00 4.84870747e-03 -6.40226126e-01 -3.02782118e-01
2.56083906e-01 -3.78253877e-01 -7.35702813e-02 -7.71052182e-01
-9.61464643e-01 8.34014595e-01 2.32457042e-01 3.73634130e-01
1.04513490e+00 -2.12886065e-01 1.07258856e+00 5.76312244e-01
-4.13066894e-02 -1.22518468e+00 1.14410244e-01 8.85580420e-01
9.93450820e-01 -1.07138312e+00 2.59857237e-01 -1.37317836e-01
-8.16020429e-01 9.94018793e-01 6.01559341e-01 -9.14599374e-02
1.57153711e-01 -5.25712110e-02 -5.24371713e-02 -8.10496584e-02
-9.24884319e-01 -3.67000252e-02 4.32165802e-01 1.89801678e-02
9.44037259e-01 7.75851384e-02 -8.82455528e-01 8.27615619e-01
-4.21122611e-01 -1.02613486e-01 7.03544855e-01 6.91256583e-01
-6.63749158e-01 -9.37142909e-01 3.46706025e-02 8.86748254e-01
-8.65989983e-01 -4.73263323e-01 -5.96085846e-01 3.70636284e-01
-3.37056756e-01 1.01202035e+00 2.00633369e-02 -6.17024064e-01
4.48014110e-01 2.27057654e-02 9.71593186e-02 -6.21462286e-01
-8.21624994e-01 7.21891969e-02 4.55956459e-01 -4.94803697e-01
-3.67001206e-01 -5.50314903e-01 -1.14345837e+00 -3.25859427e-01
-5.74002266e-01 3.76642019e-01 6.18263841e-01 4.53661978e-01
5.35944462e-01 7.24431992e-01 4.64375347e-01 -6.55197620e-01
-8.29629600e-01 -1.23418260e+00 -2.83168435e-01 2.10649714e-01
1.87012643e-01 -2.78880179e-01 -4.05281067e-01 -1.97744295e-01] | [12.257293701171875, 9.292240142822266] |
270c83b7-0c57-4029-9ec9-f435f38d89e8 | can-characters-reveal-your-native-language-a | null | null | https://aclanthology.org/D14-1142 | https://aclanthology.org/D14-1142.pdf | Can characters reveal your native language? A language-independent approach to native language identification | null | ['Marius Popescu', 'Radu Tudor Ionescu', 'Aoife Cahill'] | 2014-10-01 | null | null | null | emnlp-2014-10 | ['native-language-identification'] | ['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] | [-7.373507022857666, 3.6945815086364746] |
0de3a08b-a065-43df-b1e1-544ab75b620e | lipreading-with-long-short-term-memory | 1601.08188 | null | http://arxiv.org/abs/1601.08188v1 | http://arxiv.org/pdf/1601.08188v1.pdf | Lipreading with Long Short-Term Memory | Lipreading, i.e. speech recognition from visual-only recordings of a
speaker's face, can be achieved with a processing pipeline based solely on
neural networks, yielding significantly better accuracy than conventional
methods. Feed-forward and recurrent neural network layers (namely Long
Short-Term Memory; LSTM) are stacked to form a single structure which is
trained by back-propagating error gradients through all the layers. The
performance of such a stacked network was experimentally evaluated and compared
to a standard Support Vector Machine classifier using conventional computer
vision features (Eigenlips and Histograms of Oriented Gradients). The
evaluation was performed on data from 19 speakers of the publicly available
GRID corpus. With 51 different words to classify, we report a best word
accuracy on held-out evaluation speakers of 79.6% using the end-to-end neural
network-based solution (11.6% improvement over the best feature-based solution
evaluated). | ['Jürgen Schmidhuber', 'Jan Koutník', 'Michael Wand'] | 2016-01-29 | null | null | null | null | ['lipreading'] | ['computer-vision'] | [ 3.21178198e-01 1.75433397e-01 -1.09215908e-01 -5.98642766e-01
-1.10759163e+00 -1.84833840e-01 4.39736813e-01 -2.42378995e-01
-5.60253203e-01 2.93878049e-01 4.19409841e-01 -3.73168141e-01
2.97865897e-01 -1.33886486e-01 -6.60933554e-01 -5.40356100e-01
5.04759327e-02 -3.55705223e-03 -1.21080957e-01 2.58485317e-01
3.68023425e-01 5.45732141e-01 -2.00155520e+00 7.13387311e-01
2.53750533e-01 1.36164749e+00 -4.14933786e-02 9.87897158e-01
-2.27244318e-01 6.20115578e-01 -6.48869872e-01 -4.63842541e-01
-2.11733818e-01 -3.20584700e-02 -6.07584894e-01 1.78095371e-01
9.92949367e-01 -1.93870381e-01 -5.68776056e-02 7.53788531e-01
9.90286529e-01 2.05363464e-02 3.68887186e-01 -8.90746236e-01
-7.78195739e-01 4.54861850e-01 -1.40365437e-01 -1.67927134e-03
4.65193719e-01 2.65050292e-01 9.33570266e-01 -1.53844428e+00
2.97778934e-01 1.34955096e+00 9.00957346e-01 9.18369412e-01
-1.23269522e+00 -5.26134968e-01 -6.03562370e-02 4.59287316e-01
-1.27870429e+00 -1.56188798e+00 6.13892198e-01 -2.98380643e-01
1.67878175e+00 2.66831696e-01 5.14008641e-01 1.24595737e+00
-6.04580566e-02 8.59042883e-01 1.06371355e+00 -8.26956332e-01
1.25995308e-01 5.24134636e-01 1.98112980e-01 9.09998596e-01
-1.01530246e-01 2.24605381e-01 -1.19785655e+00 1.48058617e-02
5.60710914e-02 -4.80453908e-01 -1.33395642e-01 -1.82453915e-01
-1.03348160e+00 6.23736858e-01 2.25326031e-01 1.68904334e-01
-3.84539962e-01 -7.63164386e-02 5.56113839e-01 1.97585329e-01
6.18376434e-01 -1.15152448e-01 -5.30138612e-01 -1.23790637e-01
-1.34798050e+00 -1.51601136e-01 9.10379112e-01 4.12421167e-01
4.17824239e-01 4.42622870e-01 -1.89615428e-01 1.36950910e+00
8.03447306e-01 4.52074111e-01 1.01466346e+00 -6.10913992e-01
4.78984684e-01 3.10014278e-01 -2.14665294e-01 -5.71233034e-01
-4.57726538e-01 -3.01204711e-01 -5.37955880e-01 7.30430722e-01
4.92476285e-01 -3.31626922e-01 -1.21792924e+00 1.47942543e+00
-1.07003525e-01 4.73978892e-02 3.78388762e-01 5.65950632e-01
1.36198580e+00 6.42941892e-01 6.81653842e-02 -1.51935384e-01
1.32467830e+00 -1.03043807e+00 -7.54509091e-01 -4.06222165e-01
3.29636812e-01 -8.14068139e-01 1.08510983e+00 3.80558163e-01
-1.33023727e+00 -5.85005760e-01 -1.21396995e+00 -1.52459204e-01
-7.00006068e-01 5.43282807e-01 2.63945460e-02 1.05915415e+00
-1.68277407e+00 2.82052726e-01 -3.39284986e-01 -2.70162642e-01
4.84895557e-01 5.44090867e-01 -5.91268241e-01 3.42073113e-01
-9.52169001e-01 8.91601861e-01 -2.08199155e-02 3.61463249e-01
-7.34880269e-01 -4.62756813e-01 -1.11174655e+00 1.47866994e-01
-1.63963437e-01 -3.00301790e-01 1.47873676e+00 -9.82007921e-01
-2.09204197e+00 1.13408458e+00 -8.64503443e-01 -5.97700715e-01
2.92920887e-01 -3.03120282e-03 -5.30407369e-01 1.23482570e-01
-4.42915231e-01 8.64878595e-01 1.27413237e+00 -9.20794010e-01
-3.92486900e-01 -4.78776008e-01 -7.65369117e-01 1.43398225e-01
-5.93102694e-01 5.13685346e-01 -9.43011232e-03 -3.61376673e-01
8.53594393e-02 -7.26711988e-01 3.65410089e-01 -2.02178247e-02
-4.79626745e-01 -6.19059682e-01 6.69659972e-01 -1.15359044e+00
8.52525115e-01 -2.15364289e+00 -3.36052477e-01 -3.25348228e-02
-1.08876847e-01 6.67142808e-01 -2.02365533e-01 2.03207303e-02
-3.33883584e-01 2.03592226e-01 -1.39371887e-01 -1.04876173e+00
2.24955633e-01 -2.33569562e-01 -1.55872166e-01 5.88473797e-01
2.81572759e-01 9.14357901e-01 -1.42328471e-01 -3.84161562e-01
3.65011692e-01 9.90065575e-01 -2.50812978e-01 1.62159622e-01
-1.53219076e-02 -2.78045416e-01 5.33305883e-01 7.53513932e-01
6.12439811e-01 7.01548755e-02 -1.74052030e-01 2.00816970e-02
-2.51173675e-01 6.58771694e-01 -6.84791088e-01 1.63903308e+00
-6.73081517e-01 1.32656157e+00 1.91439480e-01 -6.56916022e-01
1.01750338e+00 8.47141206e-01 -9.93424207e-02 -6.12250805e-01
-7.63160002e-04 2.35320672e-01 -1.59211203e-01 -6.73347056e-01
2.06319600e-01 -4.36696969e-02 3.62794429e-01 1.83309987e-01
4.46015477e-01 3.41225684e-01 -2.11824432e-01 -3.70578170e-01
5.12537122e-01 -8.15506205e-02 -1.57449573e-01 -4.91654724e-02
8.19558203e-01 -5.95801055e-01 1.66458175e-01 3.91815424e-01
-4.70903933e-01 7.49084294e-01 2.66288787e-01 -2.94479400e-01
-8.57569098e-01 -8.37461889e-01 -3.36670876e-01 1.47116101e+00
-8.86792004e-01 -3.53561878e-01 -1.08813620e+00 -4.23124522e-01
-7.63880387e-02 6.03907228e-01 -5.83702505e-01 6.53220937e-02
-4.23029780e-01 -5.36981225e-01 8.62351358e-01 3.79119903e-01
4.05401796e-01 -1.52788877e+00 -3.95250380e-01 1.25214130e-01
-1.65216699e-02 -1.16343939e+00 -2.93984294e-01 1.04963034e-01
-3.94040108e-01 -5.48978269e-01 -1.04528987e+00 -1.06611753e+00
1.30014926e-01 -1.38407990e-01 8.69279683e-01 -2.37772003e-01
-3.36843401e-01 1.25347152e-01 2.24345699e-01 -6.54312134e-01
-4.36859041e-01 3.07809785e-02 2.79248029e-01 4.15023834e-01
7.02225983e-01 -2.17430159e-01 -4.46947396e-01 1.11628518e-01
-2.15523332e-01 -1.50154561e-01 6.76974297e-01 8.55355263e-01
1.89938992e-01 -5.82973957e-01 7.42306471e-01 6.49849102e-02
5.97902179e-01 2.90810280e-02 -6.04915738e-01 2.41314977e-01
-5.88952124e-01 -2.23703310e-02 7.55453110e-02 -3.75812382e-01
-1.05604231e+00 1.66314885e-01 -6.68901205e-01 -3.27295005e-01
-5.57909429e-01 1.06406979e-01 -1.39094785e-01 -2.04057351e-01
4.42927569e-01 4.97413427e-01 4.32609707e-01 -7.35656679e-01
3.33572417e-01 1.43143058e+00 6.20068133e-01 2.29709402e-01
3.08148824e-02 2.57027578e-02 -4.42584693e-01 -1.43545580e+00
-3.56860638e-01 -4.45688188e-01 -5.64143300e-01 -2.75811762e-01
8.71990740e-01 -8.35145175e-01 -1.19908965e+00 1.02503169e+00
-1.30542791e+00 -2.87436038e-01 4.49837670e-02 6.18740857e-01
-5.64423084e-01 -8.44566897e-02 -4.81659025e-01 -1.34676933e+00
-5.96895039e-01 -1.11788976e+00 1.10300112e+00 2.07014769e-01
-2.60034561e-01 -9.60974038e-01 -2.11310089e-02 5.99923909e-01
5.81322789e-01 -3.58530939e-01 5.83259702e-01 -9.16824758e-01
8.44919980e-02 -1.64334878e-01 -1.92488357e-01 7.86096692e-01
-5.42062968e-02 2.11660177e-01 -2.03591108e+00 -1.84749678e-01
-5.63378856e-02 -4.33238477e-01 1.17122304e+00 6.79001331e-01
1.01804745e+00 -5.88072002e-01 -1.33121863e-01 5.07248223e-01
9.92216885e-01 -6.23013414e-02 5.49578369e-01 1.62073836e-01
4.88552272e-01 1.00687170e+00 -3.93796742e-01 8.52985084e-02
3.19534838e-01 6.27564371e-01 8.42307881e-02 -1.13718264e-01
-6.63477421e-01 -2.63290226e-01 8.75481606e-01 7.02527642e-01
4.11477059e-01 -3.69036049e-02 -1.01735067e+00 6.61342680e-01
-1.27116585e+00 -1.13215482e+00 2.50956059e-01 2.27925682e+00
6.43659472e-01 1.47373274e-01 4.25460368e-01 6.71359956e-01
9.51214790e-01 2.42621884e-01 -4.66018021e-01 -1.03841650e+00
-1.40633002e-01 1.48467988e-01 2.41681620e-01 7.69002676e-01
-1.06796408e+00 8.94441962e-01 7.55023050e+00 5.94746888e-01
-1.70391476e+00 6.07880577e-02 8.30815613e-01 -4.72884655e-01
1.62433118e-01 -8.22816968e-01 -1.20038581e+00 3.77371281e-01
1.81509829e+00 2.15492532e-01 2.76158184e-01 7.30775297e-01
3.33506435e-01 9.50431228e-02 -9.09878314e-01 1.36456144e+00
5.98335445e-01 -1.24383676e+00 -2.44147077e-01 9.42860469e-02
2.08585888e-01 4.26869452e-01 6.07909024e-01 2.51345128e-01
-1.67452291e-01 -1.40962553e+00 8.07762265e-01 4.40837264e-01
1.14036417e+00 -7.16027677e-01 4.33328241e-01 9.92302895e-02
-9.47124958e-01 -2.03585625e-01 -1.45073146e-01 1.03395604e-01
-9.66062099e-02 3.61378908e-01 -1.47655988e+00 -1.82134107e-01
7.64610589e-01 5.16896784e-01 -6.13427877e-01 1.05044413e+00
2.91770715e-02 8.18519652e-01 -2.87962914e-01 -3.59023958e-01
1.73082203e-01 4.94032264e-01 5.63368440e-01 1.60561478e+00
8.80450159e-02 -6.81562722e-01 -5.16094089e-01 5.47267795e-01
-2.99822360e-01 1.99169144e-01 -6.18499815e-01 7.16067106e-02
2.06617892e-01 1.24060762e+00 1.71275064e-03 -3.29564214e-01
-4.74464744e-01 7.29554355e-01 4.11939174e-01 4.78648961e-01
-3.51805240e-01 -6.31033480e-01 7.40833461e-01 1.11236060e-02
5.54591894e-01 5.75767308e-02 -4.28827137e-01 -7.66240001e-01
2.97517270e-01 -8.07102978e-01 5.64041287e-02 -8.49099815e-01
-1.05612326e+00 1.01515818e+00 -5.01085818e-01 -6.85522914e-01
-8.05775225e-01 -1.05284643e+00 -5.76855838e-01 1.28475654e+00
-1.68926811e+00 -1.18848503e+00 9.43859965e-02 5.25394261e-01
6.72246218e-01 -7.17814565e-01 1.20866251e+00 1.19785979e-01
-6.85559750e-01 1.07244599e+00 -1.89210810e-02 2.56689548e-01
6.47963345e-01 -1.03051817e+00 5.95762134e-01 6.18581593e-01
3.14605713e-01 3.80636364e-01 2.88037270e-01 -1.91340894e-01
-1.11814630e+00 -9.30535138e-01 1.57124114e+00 -1.20169237e-01
3.92032921e-01 -6.02506697e-01 -8.77734840e-01 5.52810252e-01
5.69044292e-01 1.78627744e-01 9.38649297e-01 1.68585718e-01
-5.93295276e-01 -2.31670275e-01 -1.31329536e+00 2.95853287e-01
6.94178343e-01 -1.00311315e+00 -7.13238895e-01 9.05888528e-02
4.41860795e-01 -1.12815164e-01 -4.24855113e-01 2.73411483e-01
1.09097159e+00 -1.09308732e+00 1.01986217e+00 -7.18931973e-01
-5.13747819e-02 1.68696195e-01 -2.11258665e-01 -1.19909441e+00
1.64253682e-01 -7.86805272e-01 -1.60131395e-01 1.20503044e+00
9.17208433e-01 -7.82015741e-01 7.85437584e-01 4.68108058e-01
-6.18068315e-02 -8.31648886e-01 -1.35406494e+00 -5.34095466e-01
1.52302235e-01 -6.64342523e-01 4.08917964e-01 3.40570778e-01
1.18376002e-01 4.18365389e-01 -1.71030059e-01 -4.58766781e-02
5.17845094e-01 -4.07653123e-01 2.80717492e-01 -1.12658155e+00
1.70160204e-01 -7.39203632e-01 -6.94296837e-01 -7.45791852e-01
7.12137818e-01 -9.25458312e-01 2.53334790e-01 -1.49699724e+00
-1.30743116e-01 -1.12839594e-01 -3.67831886e-01 7.71343768e-01
1.44835025e-01 6.23652101e-01 1.86756942e-02 -7.79823735e-02
-6.44513816e-02 2.75502354e-01 4.94743556e-01 -2.75589705e-01
-1.81656227e-01 1.91566676e-01 -4.14362937e-01 8.47423851e-01
9.65020657e-01 -2.35971995e-03 -1.79546152e-03 -3.41117173e-01
-2.65455037e-01 -4.37596515e-02 4.88291115e-01 -1.25653756e+00
4.76783216e-01 5.11217833e-01 6.87710285e-01 -6.24422312e-01
9.90747154e-01 -2.86592841e-01 -5.73218167e-01 2.87888199e-01
-7.52707243e-01 -1.09975986e-01 5.09234369e-01 -1.73581708e-02
-2.27786511e-01 -1.81067586e-01 9.00483787e-01 2.10272193e-01
-5.74982941e-01 -5.22540733e-02 -6.94765389e-01 -4.07668382e-01
5.14982879e-01 -2.67130613e-01 -2.40323424e-01 -3.91377568e-01
-1.03829277e+00 -4.75508183e-01 -9.14690569e-02 4.61716682e-01
9.60743487e-01 -1.22538292e+00 -1.02625990e+00 7.73033917e-01
-7.56162852e-02 -6.54139221e-01 2.55250502e-02 5.54964542e-01
-1.18692897e-01 8.08735907e-01 -5.97179122e-02 -6.75501227e-01
-1.92328799e+00 2.96668440e-01 7.71965802e-01 2.28968740e-01
-5.13625979e-01 1.24840546e+00 -4.27651912e-01 -4.58461076e-01
7.82512426e-01 -2.32539013e-01 -4.61399138e-01 3.33644152e-01
1.06356061e+00 3.81434321e-01 7.02545166e-01 -1.14999044e+00
-6.82705164e-01 4.71960127e-01 -3.68825952e-03 -7.21709371e-01
1.08147633e+00 -1.52345914e-02 6.18821308e-02 6.66604698e-01
1.62317085e+00 7.86686987e-02 -9.07325804e-01 -1.31303623e-01
-7.66642345e-03 -1.65217966e-01 4.42416728e-01 -1.07491505e+00
-7.83865392e-01 1.48278642e+00 1.11267126e+00 2.99830258e-01
7.04103351e-01 -2.05656707e-01 5.50597906e-01 4.84205514e-01
-3.57712477e-01 -9.98803616e-01 -1.55035198e-01 4.74460453e-01
1.15406418e+00 -1.43005466e+00 -5.52221060e-01 1.84825025e-02
-4.74346340e-01 1.27312922e+00 1.68528527e-01 2.66090244e-01
7.20816076e-01 3.86629999e-01 4.81334746e-01 2.46062949e-01
-9.75301325e-01 -2.24211052e-01 6.62899077e-01 6.67345941e-01
7.40189552e-01 7.48321274e-03 4.54114705e-01 2.81378359e-01
-4.45677191e-01 4.76495968e-03 2.77191177e-02 4.09645528e-01
-5.47324717e-01 -7.16470540e-01 -5.81702590e-01 3.62169117e-01
-6.78162634e-01 -3.44504744e-01 -3.91620040e-01 1.66970357e-01
-1.45501286e-01 1.45897543e+00 3.20232958e-01 -3.33140403e-01
2.02117696e-01 8.54809940e-01 2.82027423e-01 -3.41666937e-01
-6.71149254e-01 1.03562605e-02 3.77515256e-01 -3.93080443e-01
-1.43623292e-01 -8.24166059e-01 -9.20196116e-01 -2.21947566e-01
-1.37739077e-01 -1.51915506e-01 1.38068986e+00 7.94484675e-01
4.37684804e-01 3.72232825e-01 6.17842793e-01 -1.09555590e+00
-6.81279719e-01 -1.37140322e+00 -1.75168753e-01 -9.87876877e-02
9.95310307e-01 -3.80140513e-01 -6.43516660e-01 1.05791636e-01] | [14.350528717041016, 5.08786153793335] |
dcc8e4a8-ef86-498b-81aa-643fa13b7e54 | mixrts-toward-interpretable-multi-agent | 2209.07225 | null | https://arxiv.org/abs/2209.07225v2 | https://arxiv.org/pdf/2209.07225v2.pdf | MIXRTs: Toward Interpretable Multi-Agent Reinforcement Learning via Mixing Recurrent Soft Decision Trees | Multi-agent reinforcement learning (MARL) recently has achieved tremendous success in a wide range of fields. However, with a black-box neural network architecture, existing MARL methods make decisions in an opaque fashion that hinders humans from understanding the learned knowledge and how input observations influence decisions. Our solution is MIXing Recurrent soft decision Trees (MIXRTs), a novel interpretable architecture that can represent explicit decision processes via the root-to-leaf path of decision trees. We introduce a novel recurrent structure in soft decision trees to address partial observability, and estimate joint action values via linearly mixing outputs of recurrent trees based on local observations only. Theoretical analysis shows that MIXRTs guarantees the structural constraint with additivity and monotonicity in factorization. We evaluate MIXRTs on a range of challenging StarCraft II tasks. Experimental results show that our interpretable learning framework obtains competitive performance compared to widely investigated baselines, and delivers more straightforward explanations and domain knowledge of the decision processes. | ['Chunlin Chen', 'Yang Gao', 'Zhi Wang', 'Yuanyang Zhu', 'Zichuan Liu'] | 2022-09-15 | null | null | null | null | ['starcraft-ii', 'starcraft'] | ['playing-games', 'playing-games'] | [ 3.33030671e-01 6.54755235e-01 -5.09673715e-01 -3.84744167e-01
-4.74927247e-01 -7.55465686e-01 9.21637058e-01 5.76971844e-02
-5.44970185e-02 6.62738442e-01 5.01599073e-01 -5.55010200e-01
-3.30412477e-01 -4.35478956e-01 -6.41220808e-01 -6.70101285e-01
-7.81464428e-02 7.77589977e-01 -1.33033767e-01 -3.46609443e-01
-8.38833824e-02 -9.89216864e-02 -1.28781652e+00 7.46800482e-01
7.16250658e-01 9.14970100e-01 -1.84169918e-01 8.14604342e-01
1.43080354e-01 1.66279066e+00 -4.07046974e-01 -4.23367023e-01
4.25690740e-01 -3.71722996e-01 -8.78735244e-01 1.67369530e-01
1.16198309e-01 -5.89734614e-01 -9.55039635e-02 6.20083392e-01
-7.66572058e-02 -2.11533974e-03 7.48733580e-01 -1.41819668e+00
-7.86135316e-01 1.19409537e+00 -5.96497416e-01 1.79994151e-01
1.51520252e-01 3.48468304e-01 1.54044461e+00 -5.26589930e-01
3.46059144e-01 1.72962773e+00 5.47118843e-01 5.29248178e-01
-1.46964729e+00 -2.76626229e-01 9.58920538e-01 3.35566878e-01
-6.07489645e-01 -2.69944221e-01 5.03962159e-01 -4.21896845e-01
1.03993428e+00 8.13562796e-02 5.24867773e-01 1.54430723e+00
4.38485920e-01 1.33466113e+00 1.40931058e+00 -3.07568401e-01
6.99465990e-01 -3.40222418e-01 2.99931705e-01 9.54971135e-01
2.06447273e-01 3.82372081e-01 -1.05774438e+00 -1.65035531e-01
7.88304031e-01 8.90998766e-02 2.06347898e-01 -4.51353967e-01
-1.37999797e+00 9.96156693e-01 4.14446980e-01 -1.79086775e-01
-3.59675825e-01 6.31757855e-01 2.58876443e-01 6.55562043e-01
3.81399095e-01 5.50535500e-01 -7.57551491e-01 -7.97417462e-02
-3.15493107e-01 1.17730618e-01 8.60858440e-01 5.39116502e-01
6.24070585e-01 2.64065325e-01 -2.31135413e-01 4.00711447e-01
4.96531636e-01 5.67412972e-01 4.53804016e-01 -1.35530579e+00
4.73335773e-01 5.69097042e-01 1.06471650e-01 -6.55632675e-01
-5.17171323e-01 -5.25710464e-01 -7.51310229e-01 6.64309204e-01
6.17162764e-01 -5.24406552e-01 -9.75975454e-01 1.75080371e+00
2.56899238e-01 -1.62445549e-02 4.28673178e-01 1.01175702e+00
3.15258831e-01 5.17713606e-01 1.49211613e-02 1.42922495e-02
1.23681188e+00 -1.26171672e+00 -7.34554529e-01 -9.80455518e-01
4.61995393e-01 -2.99045622e-01 9.00552750e-01 7.43276596e-01
-7.05969810e-01 -2.87999511e-01 -1.10921144e+00 -2.43534446e-02
1.47548154e-01 2.92100042e-01 1.25281763e+00 3.98053825e-01
-8.92201781e-01 5.98206580e-01 -1.39372480e+00 7.77013376e-02
2.67744362e-01 4.80790913e-01 -1.22201815e-01 2.24746242e-01
-1.00824845e+00 8.93624842e-01 2.61725038e-01 3.47631842e-01
-1.33435261e+00 -3.83445323e-01 -9.17869449e-01 -8.28070268e-02
8.42512786e-01 -9.85049069e-01 1.92253339e+00 -1.28290832e+00
-1.87236190e+00 3.61492187e-01 -2.16255918e-01 -1.08644736e+00
6.03082955e-01 -7.38658607e-01 7.69916549e-02 6.08916804e-02
4.12076376e-02 5.33100128e-01 1.33520627e+00 -9.61728811e-01
-8.45429957e-01 -2.93713897e-01 4.38183963e-01 4.42616731e-01
2.15819389e-01 -4.14972305e-01 2.71860361e-01 -7.70149708e-01
9.44307595e-02 -1.04297531e+00 -6.81586027e-01 -1.49001181e-01
-3.01450968e-01 -3.56830239e-01 5.92694819e-01 -3.73213321e-01
9.65446532e-01 -1.77214551e+00 6.07561409e-01 -6.65920302e-02
5.91040194e-01 -4.76196446e-02 -1.99909344e-01 4.61235195e-01
3.95576172e-02 2.57318635e-02 -2.15800956e-01 -6.51450634e-01
4.08781499e-01 7.50195324e-01 -9.38362479e-01 1.82854608e-01
3.12591702e-01 1.03196752e+00 -1.07437563e+00 3.99582162e-02
3.49685967e-01 -3.03511135e-02 -5.90321422e-01 9.60715786e-02
-7.98052788e-01 3.37908238e-01 -7.27772951e-01 6.53782547e-01
9.97532606e-02 -4.94958639e-01 5.67667902e-01 2.99910158e-01
1.73861742e-01 6.47114933e-01 -1.24401867e+00 1.68632567e+00
-5.16450882e-01 5.90677381e-01 -1.29393965e-01 -8.68761778e-01
5.49480319e-01 2.24340603e-01 1.11253053e-01 -4.82817680e-01
-2.25938391e-02 1.81371234e-02 2.46972948e-01 -2.44518027e-01
1.74711838e-01 -1.04666054e-01 -2.93223917e-01 9.03275549e-01
1.35947734e-01 1.13073729e-01 4.42181304e-02 2.89492458e-01
1.29029489e+00 5.56092381e-01 6.27549827e-01 -8.86990502e-02
9.95919853e-02 -1.53098717e-01 9.31112349e-01 1.19515514e+00
-9.68825072e-02 2.49614626e-01 9.28611934e-01 -1.07197475e+00
-5.71689785e-01 -1.23606658e+00 4.64351296e-01 1.55969441e+00
-1.47525609e-01 -6.59466803e-01 -5.08538127e-01 -8.81975532e-01
1.07330300e-01 8.63048434e-01 -1.12000668e+00 -1.63329259e-01
-4.48483229e-01 -7.54597545e-01 3.52098197e-01 7.27492511e-01
4.56433535e-01 -9.39624429e-01 -8.41001630e-01 3.33460927e-01
-1.02798417e-01 -1.12629390e+00 -6.74567670e-02 8.31608117e-01
-9.18085873e-01 -1.00036693e+00 -7.48950765e-02 -3.00071031e-01
4.97879148e-01 1.86934859e-01 1.21132267e+00 -2.09685355e-01
2.78945696e-02 4.05930221e-01 -3.88035357e-01 -4.76918757e-01
-5.54984510e-01 1.62604958e-01 2.45178625e-01 1.51934907e-01
2.89307028e-01 -7.30719566e-01 -3.24879318e-01 3.65945041e-01
-6.41649485e-01 4.94791299e-01 1.00489616e+00 9.50705826e-01
2.61222750e-01 -2.99194872e-01 3.14995289e-01 -8.35136354e-01
6.23289764e-01 -3.34796429e-01 -5.89544773e-01 2.99875677e-01
-5.09773374e-01 1.00444877e+00 6.41379178e-01 -3.37934315e-01
-1.19394112e+00 -1.83404814e-02 3.80549788e-01 -3.18856359e-01
-2.93218613e-01 4.48997319e-01 2.67151445e-01 4.41340178e-01
7.29301751e-01 -5.89376427e-02 -2.14123148e-02 -2.09503621e-01
5.14612734e-01 2.20333382e-01 3.12401086e-01 -8.60752940e-01
6.43227637e-01 5.89432240e-01 -2.69837733e-02 -3.20341319e-01
-1.35466766e+00 1.93179511e-02 -5.51539779e-01 -3.33467871e-02
6.62708342e-01 -1.10957682e+00 -9.44297135e-01 3.55239719e-01
-1.09016764e+00 -7.48537719e-01 -4.35152858e-01 3.94319206e-01
-7.91496575e-01 -4.14053053e-02 -7.91219413e-01 -1.04905403e+00
-2.31690437e-01 -1.16354275e+00 1.19533205e+00 5.37171140e-02
-5.26841342e-01 -1.13574839e+00 9.06205028e-02 5.15112400e-01
-3.23604094e-03 1.56361833e-01 8.57147932e-01 -6.00990593e-01
-8.96498621e-01 4.29454058e-01 2.95495003e-01 1.29133806e-01
2.08731517e-01 -1.66337073e-01 -1.10393918e+00 -1.71833932e-01
-6.69061765e-02 -6.11984372e-01 1.30388474e+00 3.52283686e-01
7.02059746e-01 -5.54727614e-01 -1.42345935e-01 3.99374247e-01
8.69824469e-01 -8.91535208e-02 3.24154869e-02 3.55309963e-01
4.98878598e-01 4.54024494e-01 4.76174861e-01 6.35563612e-01
6.46934628e-01 3.60246569e-01 7.19228745e-01 4.41027805e-02
2.95817945e-02 -4.99592662e-01 9.82809782e-01 5.71158767e-01
-2.47206524e-01 -2.08410665e-01 -8.72075021e-01 3.02656800e-01
-2.54313231e+00 -8.47984970e-01 1.62675366e-01 1.90684986e+00
6.94271982e-01 3.99548769e-01 5.05249016e-02 3.53975408e-02
2.00821042e-01 3.32357109e-01 -1.02537704e+00 -7.37408102e-01
-1.24211229e-01 1.86683029e-01 4.07382339e-01 8.27936947e-01
-1.24295914e+00 9.89461958e-01 6.88168859e+00 4.73277330e-01
-7.50849247e-01 -2.61813253e-02 6.34248435e-01 -3.02924067e-01
-3.95450562e-01 2.27214426e-01 -7.85406590e-01 -4.35569137e-02
6.51294708e-01 1.33759394e-01 6.75712943e-01 7.59220958e-01
1.51156947e-01 -1.97228640e-01 -1.56704485e+00 6.93843186e-01
-2.09778622e-01 -1.31319988e+00 8.78371149e-02 -1.88190155e-02
6.99847817e-01 1.90512270e-01 4.15687978e-01 4.66215998e-01
1.55157661e+00 -1.08157516e+00 9.58359718e-01 1.96624130e-01
1.69125900e-01 -5.88772357e-01 3.51263285e-01 5.71376562e-01
-1.07622778e+00 -6.23038471e-01 -1.40557230e-01 -7.97295213e-01
-1.89985678e-01 3.97224724e-01 -1.08131981e+00 4.75050628e-01
4.64773118e-01 1.10110927e+00 -5.14384687e-01 2.14541718e-01
-1.08451128e+00 8.48928154e-01 -4.25389409e-01 -7.32462704e-02
6.27970457e-01 -2.00091586e-01 6.08506441e-01 8.58595073e-01
-3.40562314e-01 3.85277644e-02 3.48621726e-01 8.35399985e-01
1.33440331e-01 -4.95169312e-01 -5.63113034e-01 -8.35411549e-02
5.28567936e-04 1.08255970e+00 -6.73971415e-01 -8.37563127e-02
-2.33260840e-01 8.95119607e-01 6.73784733e-01 5.55468738e-01
-6.48016036e-01 3.37298810e-01 1.11318755e+00 -3.21835667e-01
5.22946060e-01 -4.12916839e-01 -4.56489980e-01 -1.47070920e+00
4.29895632e-02 -1.31549597e+00 6.18779778e-01 -6.33637428e-01
-1.18053460e+00 4.03707713e-01 -1.51552945e-01 -1.01504886e+00
-8.24864268e-01 -7.19228506e-01 -4.44551885e-01 3.82230818e-01
-1.39750016e+00 -1.28835428e+00 2.24277645e-01 2.35345185e-01
9.17647004e-01 -2.32979879e-01 1.02975941e+00 -5.89619815e-01
-6.09933853e-01 3.24981511e-01 9.79899168e-02 8.85615945e-02
4.44965214e-01 -1.66052675e+00 8.16168904e-01 8.58205676e-01
5.69628417e-01 6.13650560e-01 8.34439218e-01 -5.11531711e-01
-1.49758089e+00 -7.84917533e-01 3.61798972e-01 -7.46761262e-01
9.52228963e-01 -5.17391145e-01 -5.31489313e-01 1.27129853e+00
5.36409616e-01 -2.48217821e-01 8.09394181e-01 7.13339686e-01
-6.49873912e-01 -4.26594652e-02 -5.98238885e-01 7.83122122e-01
9.86746013e-01 -4.63333607e-01 -1.10631633e+00 2.47877821e-01
7.42959261e-01 -4.03707296e-01 -4.49890137e-01 2.25506544e-01
6.37275517e-01 -1.02853644e+00 7.43893087e-01 -9.72522855e-01
7.63709068e-01 -5.46489120e-01 -1.19091772e-01 -1.59646344e+00
-3.17588389e-01 -1.12540746e+00 -5.94881415e-01 6.28106952e-01
6.13004744e-01 -7.06710875e-01 6.53688133e-01 5.56793451e-01
-5.04967161e-02 -8.26736271e-01 -7.68100619e-01 -4.38776970e-01
-9.90239233e-02 -5.27934968e-01 3.54280680e-01 7.15582967e-01
1.72057524e-02 7.09521830e-01 -6.03287458e-01 3.71040881e-01
8.71831059e-01 4.66618568e-01 6.77353501e-01 -1.14076483e+00
-9.92637277e-01 -3.05544585e-01 3.31439590e-03 -1.39945531e+00
2.75790304e-01 -8.79235148e-01 -1.26321048e-01 -1.51901388e+00
-7.17656612e-02 -2.69405335e-01 -3.31644952e-01 1.03613472e+00
-2.28067085e-01 -2.79488176e-01 4.12788212e-01 8.49711969e-02
-1.06489599e+00 7.85005569e-01 1.10162652e+00 -2.60267973e-01
-1.29080638e-01 3.42043899e-02 -9.10760403e-01 1.33938372e+00
6.74893439e-01 -4.43423510e-01 -6.32800519e-01 -8.96342099e-01
7.33209848e-01 -1.08817909e-02 3.93129528e-01 -8.11658859e-01
2.45678321e-01 -4.39364910e-01 3.78900737e-01 -3.57410401e-01
3.62032980e-01 -5.34960687e-01 -1.91279694e-01 5.39803624e-01
-8.24400902e-01 6.65827841e-02 7.25252405e-02 9.23643291e-01
-4.49373759e-02 1.33919969e-01 2.70803928e-01 -1.10085018e-01
-8.24141443e-01 -1.81644447e-02 -8.92481923e-01 1.20950937e-01
8.45870137e-01 9.63529106e-03 -3.27134818e-01 -4.92680788e-01
-9.45669234e-01 7.14084387e-01 1.17782764e-01 6.23464167e-01
4.89767462e-01 -1.02396822e+00 -6.05855823e-01 1.95733413e-01
-9.15798470e-02 2.90397793e-01 -1.37218326e-01 4.20263529e-01
-1.20722733e-01 3.23618889e-01 -1.85472518e-01 -5.81425786e-01
-1.01369178e+00 2.15742007e-01 4.73620445e-01 -7.69225061e-01
-6.46948934e-01 8.14782143e-01 4.24281120e-01 -5.53537309e-01
2.08139852e-01 -9.13579047e-01 1.93257667e-02 4.74997051e-02
6.22904718e-01 1.48742959e-01 -1.93959847e-01 1.17728993e-01
-1.58295676e-01 3.21801573e-01 -3.72835040e-01 -3.00637424e-01
1.33602893e+00 -1.31380633e-02 1.84634715e-01 8.97938848e-01
2.22220644e-01 -4.57503587e-01 -1.75556505e+00 -5.26064515e-01
1.23114869e-01 -5.65788001e-02 6.87700184e-03 -1.05539358e+00
-6.68866098e-01 9.82285559e-01 4.47074771e-02 1.45701498e-01
8.44281971e-01 2.49793695e-04 4.58437860e-01 1.07551718e+00
4.59317565e-01 -1.08851552e+00 3.10919017e-01 8.57155800e-01
8.50951076e-01 -1.37217546e+00 5.32933837e-03 -1.75618380e-01
-1.06683362e+00 1.25498247e+00 5.53427517e-01 -2.04675660e-01
3.22886497e-01 4.34929669e-01 1.61614448e-01 6.53821006e-02
-1.71354592e+00 -2.35375032e-01 -8.25803820e-03 5.55332005e-01
5.40359430e-02 2.53780425e-01 4.52177405e-01 7.01869071e-01
-2.00474232e-01 -8.60365629e-02 4.88760710e-01 9.49550807e-01
-4.34672177e-01 -1.15913498e+00 -3.61809522e-01 3.54767919e-01
-2.93016940e-01 -1.52147770e-01 -5.79256952e-01 5.70783079e-01
-1.28127977e-01 1.09996486e+00 -1.45540349e-02 -3.45164895e-01
-6.20568693e-02 6.67149574e-02 4.94053334e-01 -5.03803790e-01
-7.36067712e-01 -1.08136751e-01 1.78680226e-01 -6.94292068e-01
-3.35557669e-01 -7.89450705e-01 -1.33075225e+00 -1.85665265e-01
1.42497823e-01 7.99608156e-02 2.96174824e-01 1.41134083e+00
4.24443692e-01 8.22436094e-01 4.51986551e-01 -6.28795862e-01
-1.10916901e+00 -8.28844547e-01 -2.61239499e-01 1.83720458e-02
8.76509011e-01 -8.48173499e-01 -3.55589360e-01 9.65670496e-02] | [4.14121150970459, 1.7261509895324707] |
9c1de94f-d5a2-4fec-ad75-fbdf8922fd85 | exploiting-robust-unsupervised-video-person | 2111.05170 | null | https://arxiv.org/abs/2111.05170v3 | https://arxiv.org/pdf/2111.05170v3.pdf | Exploiting Robust Unsupervised Video Person Re-identification | Unsupervised video person re-identification (reID) methods usually depend on global-level features. And many supervised reID methods employed local-level features and achieved significant performance improvements. However, applying local-level features to unsupervised methods may introduce an unstable performance. To improve the performance stability for unsupervised video reID, this paper introduces a general scheme fusing part models and unsupervised learning. In this scheme, the global-level feature is divided into equal local-level feature. A local-aware module is employed to explore the poentials of local-level feature for unsupervised learning. A global-aware module is proposed to overcome the disadvantages of local-level features. Features from these two modules are fused to form a robust feature representation for each input image. This feature representation has the advantages of local-level feature without suffering from its disadvantages. Comprehensive experiments are conducted on three benchmarks, including PRID2011, iLIDS-VID, and DukeMTMC-VideoReID, and the results demonstrate that the proposed approach achieves state-of-the-art performance. Extensive ablation studies demonstrate the effectiveness and robustness of proposed scheme, local-aware module and global-aware module. The code and generated features are available at https://github.com/deropty/uPMnet. | ['Xiujun Shu', 'Wei Gao', 'Ge Li', 'Xianghao Zang'] | 2021-11-09 | exploiting-robust-unsupervised-video-person-1 | https://arxiv.org/abs/2111.05170 | https://arxiv.org/pdf/2111.05170.pdf | null | ['unsupervised-person-re-identification'] | ['computer-vision'] | [-2.80389965e-01 -4.59651321e-01 -2.30235353e-01 -4.66482282e-01
-6.12867832e-01 -1.12133719e-01 6.51042402e-01 1.88038617e-01
-6.84495270e-01 5.91473818e-01 4.66910660e-01 4.13744360e-01
-8.95124972e-02 -6.36190116e-01 -3.96687478e-01 -8.24975491e-01
-1.25047028e-01 -9.84602720e-02 2.45421246e-01 7.30738267e-02
2.01691553e-01 2.32512876e-01 -1.90941715e+00 1.13345310e-01
8.89403224e-01 6.96435928e-01 -1.04869413e-03 4.76342767e-01
-9.29708965e-03 6.81202054e-01 -4.85704184e-01 -4.80152696e-01
4.10564482e-01 -3.78449440e-01 -6.41907096e-01 7.10687861e-02
3.61032784e-01 -4.19184774e-01 -6.30421460e-01 1.14906037e+00
6.55879200e-01 2.38915429e-01 7.51637220e-01 -1.44806862e+00
-4.68476176e-01 3.75702769e-01 -7.90790915e-01 4.84103501e-01
6.71842396e-01 1.35689303e-01 5.73241174e-01 -7.58163214e-01
3.88527304e-01 1.19446266e+00 6.65794015e-01 4.37780380e-01
-6.87591970e-01 -8.94642293e-01 2.76066780e-01 4.52912956e-01
-1.79825616e+00 -5.80793738e-01 7.18453228e-01 -4.43601340e-01
7.09840178e-01 1.53045297e-01 3.42576742e-01 8.98753643e-01
1.64492112e-02 9.97891843e-01 1.22626400e+00 -2.50883669e-01
-2.81583399e-01 2.80129641e-01 5.13807774e-01 7.91621089e-01
3.98200035e-01 1.27792880e-01 -5.76045156e-01 1.07759540e-03
5.43856680e-01 2.67594486e-01 -1.55704841e-01 -1.03194386e-01
-9.30295289e-01 4.51016575e-01 4.17601168e-01 2.30213180e-01
-8.33558198e-03 -3.09687376e-01 5.85258186e-01 2.36417860e-01
2.06510305e-01 -1.67100519e-01 -1.42666563e-01 -2.57084787e-01
-8.38422596e-01 2.00944752e-01 4.45110887e-01 1.15783441e+00
9.61825848e-01 -1.99627891e-01 -2.76175767e-01 9.80254233e-01
2.84146100e-01 1.74134925e-01 8.48642647e-01 -4.56822932e-01
3.81680548e-01 8.06801140e-01 -3.23603153e-02 -1.17189860e+00
-4.72132355e-01 -3.40109229e-01 -9.28078175e-01 -1.74844831e-01
1.47365764e-01 -1.29239321e-01 -9.74498212e-01 1.55986035e+00
3.51645917e-01 4.50169206e-01 5.63364103e-03 9.84567702e-01
1.28595388e+00 5.38733602e-01 2.77325511e-01 -2.70769447e-01
1.23388910e+00 -1.17752039e+00 -6.60421312e-01 1.45092234e-01
6.00867867e-01 -5.71800947e-01 5.27353108e-01 8.31075087e-02
-7.92274833e-01 -9.29505348e-01 -1.06501341e+00 9.77808833e-02
-6.57753170e-01 3.93071353e-01 3.77043992e-01 8.04207265e-01
-9.39410090e-01 2.49457464e-01 -4.07165051e-01 -7.87131846e-01
3.67840946e-01 5.58983803e-01 -7.84986734e-01 2.68444121e-02
-1.10841131e+00 4.54396665e-01 5.64424813e-01 1.90805897e-01
-5.89097619e-01 -2.98696995e-01 -9.27864075e-01 -1.41061649e-01
1.93215668e-01 -4.30833966e-01 9.72681761e-01 -7.76139796e-01
-1.31559241e+00 9.13956404e-01 -4.04538840e-01 -2.24523023e-01
6.54434860e-01 -1.05153546e-01 -6.74243927e-01 2.36502454e-01
2.80228049e-01 5.36726475e-01 8.12478244e-01 -1.09173024e+00
-8.94141912e-01 -2.43348822e-01 -1.33404210e-01 3.78644347e-01
-7.45912969e-01 2.14267656e-01 -8.57913494e-01 -6.82101130e-01
-1.08717546e-01 -6.97137475e-01 1.39276743e-01 -4.85575289e-01
-3.29193801e-01 -3.90392274e-01 9.57703829e-01 -6.02995276e-01
1.24901128e+00 -2.20608091e+00 -1.21769845e-01 1.89296082e-01
6.45812899e-02 3.82248104e-01 3.18905339e-02 3.74630719e-01
-1.08801305e-01 1.39275312e-01 -4.14597765e-02 -4.65566754e-01
-2.33463243e-01 -1.80817142e-01 4.24430698e-01 7.32493997e-01
2.78149210e-02 7.28527308e-01 -8.62526298e-01 -7.68070698e-01
3.54458302e-01 2.92188674e-01 -3.07611704e-01 2.74249315e-01
6.19219363e-01 5.17775416e-01 -4.64371085e-01 8.69447351e-01
8.40299308e-01 1.51800901e-01 -1.36039525e-01 -2.79156685e-01
-1.37059718e-01 -1.89215928e-01 -1.51530671e+00 1.45363224e+00
1.62627781e-03 2.59433508e-01 -1.90656915e-01 -9.87491310e-01
1.01339972e+00 1.90194443e-01 4.55823451e-01 -6.10807717e-01
3.19230109e-01 -9.70158055e-02 -3.03874999e-01 -6.57383382e-01
5.21263540e-01 3.05821508e-01 -1.12458870e-01 1.73402160e-01
2.86088914e-01 7.85529852e-01 2.71429241e-01 3.06643128e-01
9.84612823e-01 2.34311402e-01 2.40542024e-01 -3.55112582e-01
1.11064279e+00 -3.86493683e-01 8.08211029e-01 7.18084455e-01
-6.83992326e-01 6.35439396e-01 1.02417640e-01 -3.74399543e-01
-6.03146434e-01 -9.30872262e-01 -1.19035788e-01 1.00170100e+00
6.84253931e-01 -8.93903732e-01 -8.27185392e-01 -9.55742419e-01
-1.29880290e-02 4.76988778e-02 -6.41032577e-01 -8.65787044e-02
-4.15242225e-01 -8.47686052e-01 5.95255017e-01 5.46118975e-01
1.12735903e+00 -8.40307057e-01 -1.13656081e-01 -1.25721574e-01
-1.85143799e-01 -1.13766515e+00 -6.82628632e-01 -4.57570821e-01
-5.55248022e-01 -1.01609254e+00 -8.93051386e-01 -9.65301871e-01
9.49917316e-01 6.05253518e-01 5.30429542e-01 3.67793709e-01
-3.04152936e-01 5.52245855e-01 -6.16448164e-01 -1.57415226e-01
1.38638481e-01 2.16746330e-01 4.99441683e-01 4.82099056e-01
6.99178219e-01 -4.70705241e-01 -7.17308760e-01 6.12215459e-01
-6.74881876e-01 -8.63682553e-02 6.54203773e-01 8.29463124e-01
3.96271676e-01 2.77798653e-01 7.00450778e-01 -6.15785420e-01
4.68708754e-01 -5.03971934e-01 -2.07374722e-01 7.17938915e-02
-4.66445863e-01 -2.13440523e-01 5.83506048e-01 -3.86890173e-01
-1.25741971e+00 8.94361287e-02 7.32786730e-02 -2.02158943e-01
-4.15685207e-01 3.52225631e-01 -5.88356316e-01 -5.15948832e-02
4.30621207e-01 4.99329269e-01 -1.21227689e-01 -6.15628898e-01
-1.99312530e-02 1.06292427e+00 7.50404358e-01 -6.79467082e-01
1.07624114e+00 2.12781265e-01 -4.38712120e-01 -9.38296497e-01
-4.32927221e-01 -8.97776961e-01 -7.25914776e-01 -4.32028174e-01
7.57203221e-01 -1.35830963e+00 -7.19250083e-01 9.77945805e-01
-6.07639134e-01 2.58676112e-01 6.34004921e-02 4.71094012e-01
-2.56675124e-01 6.40659928e-01 -4.90347087e-01 -9.01875317e-01
-3.13649476e-01 -1.06250429e+00 7.89370954e-01 1.04325259e+00
7.93918893e-02 -6.05033338e-01 -1.33990198e-01 3.98537248e-01
1.62604287e-01 5.76680265e-02 1.20999686e-01 -8.00367892e-01
-3.62980783e-01 -5.93984365e-01 -4.36930090e-01 2.66178578e-01
4.08707976e-01 -1.06148615e-01 -1.15509570e+00 -5.91580331e-01
-3.71101588e-01 -2.89641678e-01 9.36112940e-01 4.97141294e-03
1.24518585e+00 -3.08888286e-01 -4.66820329e-01 7.64236629e-01
1.37816167e+00 -2.75617465e-02 6.04128122e-01 6.03740215e-01
8.74215126e-01 3.83931965e-01 7.39564300e-01 3.88576776e-01
8.08668554e-01 6.17769301e-01 -8.08965787e-02 1.53444991e-01
-9.46959853e-02 -4.28159326e-01 6.04759395e-01 8.67214739e-01
-3.98599476e-01 4.04214002e-02 -7.94227242e-01 6.80907786e-01
-2.04554963e+00 -1.21053767e+00 8.25356841e-02 2.19441819e+00
5.94082117e-01 1.87427089e-01 7.09031165e-01 1.26013845e-01
1.07245672e+00 2.15402454e-01 -3.19894910e-01 -2.12844908e-01
-1.57662228e-01 -2.92996943e-01 5.61468482e-01 1.72684118e-01
-1.54555488e+00 8.44665527e-01 5.45530891e+00 9.44103658e-01
-6.87153637e-01 2.22937837e-01 4.83798623e-01 1.39312744e-01
1.90168157e-01 -1.59436420e-01 -1.19419801e+00 8.31581354e-01
7.70631194e-01 -4.03269887e-01 2.29214635e-02 7.95334041e-01
1.19808592e-01 -2.53326535e-01 -8.97944331e-01 1.40632486e+00
3.13039660e-01 -8.26056778e-01 1.29101828e-01 -5.81145063e-02
6.60152495e-01 -2.38263011e-01 -7.36240074e-02 4.15990889e-01
-1.40433013e-01 -9.14528489e-01 4.27755952e-01 6.95023239e-01
7.56467760e-01 -1.04252386e+00 1.03212345e+00 3.22915912e-01
-1.67402363e+00 -2.12741539e-01 -4.42980886e-01 -1.06904797e-01
-1.60655662e-01 1.46436527e-01 -2.23033041e-01 1.14251840e+00
1.12058365e+00 1.21991503e+00 -9.06032026e-01 1.40174663e+00
-3.78335081e-02 4.53495145e-01 -2.88952976e-01 1.29646450e-01
-2.72556059e-02 -1.83592722e-01 5.15920579e-01 1.61443388e+00
1.31943654e-02 1.53735459e-01 3.67575705e-01 2.67471075e-01
-5.92242368e-02 3.13701034e-01 -4.51718509e-01 2.64104098e-01
3.41859132e-01 1.31462610e+00 -5.40196955e-01 -4.40842241e-01
-7.06540763e-01 1.13181579e+00 2.18696997e-01 3.24447721e-01
-8.19396853e-01 -6.03132904e-01 5.30070186e-01 1.13441110e-01
1.55205265e-01 -1.75296158e-01 1.47062451e-01 -1.35928726e+00
1.74160197e-01 -8.37374568e-01 7.66987860e-01 -2.74600953e-01
-1.44308221e+00 4.27442908e-01 3.23333055e-01 -1.51534832e+00
4.84740250e-02 -3.61410081e-01 -8.34020615e-01 7.20479906e-01
-1.48470247e+00 -1.41440642e+00 -8.26456606e-01 8.87748837e-01
6.30765617e-01 -5.48636615e-01 6.20057762e-01 4.04261976e-01
-1.15676892e+00 1.35872531e+00 1.50278419e-01 6.50718927e-01
1.00157356e+00 -9.78858471e-01 2.29618717e-02 1.12205422e+00
-2.40707010e-01 6.92271829e-01 4.32633549e-01 -6.61063850e-01
-1.27856815e+00 -1.06193340e+00 6.64755046e-01 -2.35264078e-01
2.53663689e-01 -2.46034727e-01 -7.50768185e-01 6.67734504e-01
3.26076224e-05 8.98879468e-02 6.83715045e-01 1.12428265e-02
-4.37530458e-01 -4.51620996e-01 -1.39270616e+00 4.71695930e-01
1.18379462e+00 -4.59278136e-01 -5.98569751e-01 5.71145043e-02
2.51377463e-01 -1.73200279e-01 -9.98862088e-01 3.78846705e-01
7.24105775e-01 -1.03770816e+00 8.41508687e-01 -2.68187881e-01
2.70468537e-02 -5.31194985e-01 9.12757739e-02 -8.11869204e-01
-4.80358481e-01 -6.23365760e-01 -1.61461309e-01 1.81429279e+00
5.60393147e-02 -7.80462027e-01 7.11654842e-01 5.17662942e-01
1.04606643e-01 -5.13206661e-01 -7.99023926e-01 -9.45371091e-01
-1.15022533e-01 4.48787250e-02 5.80606818e-01 9.28815901e-01
1.50671333e-01 8.07714090e-02 -4.87587690e-01 2.72815794e-01
8.17672074e-01 -3.12265396e-01 1.06541705e+00 -1.04113531e+00
5.89542538e-02 -3.31165075e-01 -1.08405268e+00 -9.32248414e-01
1.42169476e-01 -8.11429501e-01 -1.11379817e-01 -1.36331952e+00
7.88128078e-01 -3.58556390e-01 -7.98027635e-01 5.09136736e-01
-5.27230978e-01 4.39816356e-01 2.87101299e-01 4.06240225e-01
-8.63645911e-01 5.80982208e-01 7.28867173e-01 -3.08871835e-01
-3.38165462e-01 -2.38425881e-01 -6.16170645e-01 6.92444921e-01
9.78986144e-01 -3.39104801e-01 -3.08853596e-01 -1.42803639e-01
-4.79079276e-01 -6.27653778e-01 3.70744854e-01 -1.39167392e+00
5.85977018e-01 9.09309089e-02 8.51398230e-01 -5.86102486e-01
2.17627838e-01 -5.09037435e-01 -6.23623002e-03 2.06272230e-01
-3.64132784e-02 -1.30403675e-02 1.21438988e-01 5.87519169e-01
-3.50446433e-01 -1.23096190e-01 7.38805234e-01 -6.37787953e-02
-1.19114506e+00 4.76682454e-01 -1.42808512e-01 -2.80335456e-01
1.16435635e+00 -6.14562631e-01 -2.48642370e-01 -4.31790829e-01
-5.26819587e-01 5.12627125e-01 6.08136714e-01 7.42344737e-01
7.72247016e-01 -1.44243622e+00 -8.50153208e-01 3.47077370e-01
4.13213968e-01 -3.30194473e-01 4.95605230e-01 8.73071492e-01
-3.16706866e-01 1.02200136e-01 -3.72920990e-01 -6.49427176e-01
-1.45999801e+00 5.48230767e-01 2.82695979e-01 -7.41642043e-02
-5.52714109e-01 8.62862647e-01 1.46645114e-01 -3.51149350e-01
3.69887143e-01 3.60668153e-01 -4.82775360e-01 6.14670739e-02
8.43208730e-01 5.37852883e-01 -2.72046268e-01 -1.36265624e+00
-6.51192665e-01 9.19469535e-01 -4.55852181e-01 2.10943684e-01
1.06457496e+00 -3.78800422e-01 -1.98467281e-02 1.76332414e-01
1.32682669e+00 4.00756374e-02 -9.72933173e-01 -3.34184319e-01
-9.97189283e-02 -5.85730672e-01 -1.90860823e-01 -5.06857574e-01
-1.00381041e+00 4.16854203e-01 9.10340190e-01 -2.17883348e-01
1.32636499e+00 -2.25024506e-01 7.59106278e-01 1.01265773e-01
5.28588653e-01 -1.10804939e+00 -5.60465381e-02 1.21611848e-01
4.58916754e-01 -1.42479336e+00 2.97188491e-01 -3.87458831e-01
-5.39565384e-01 8.83775711e-01 1.02676773e+00 -1.27244368e-01
8.08556020e-01 -4.19914834e-02 -2.93049272e-02 1.48366153e-01
-7.76506215e-02 -4.06628907e-01 3.40157270e-01 5.45977592e-01
2.36526310e-01 -5.44066653e-02 -5.44476807e-01 8.72367084e-01
-1.03463620e-01 5.56967072e-02 3.13849628e-01 1.08244085e+00
-3.01797420e-01 -1.20103133e+00 -3.99352103e-01 4.79316771e-01
-4.77976590e-01 3.16562980e-01 -2.83158898e-01 7.89799809e-01
4.56806034e-01 1.21471393e+00 -5.59530221e-02 -7.96105444e-01
1.85863823e-01 -2.71793753e-02 3.09287310e-01 -2.93507308e-01
-4.55196381e-01 -3.36011387e-02 5.93263321e-02 -4.70140159e-01
-8.11969459e-01 -7.74089396e-01 -1.04345489e+00 -4.69003737e-01
-2.64695138e-01 2.12279186e-01 1.97697490e-01 8.63074124e-01
3.19570780e-01 6.54829741e-02 8.39461565e-01 -8.50120246e-01
-8.77513513e-02 -9.82956707e-01 -5.13566434e-01 5.96705556e-01
2.70799726e-01 -6.93991959e-01 -2.89550006e-01 1.46161571e-01] | [14.766327857971191, 1.0309139490127563] |
1843e0fa-81dc-4fa4-920a-dce7b166293b | suffering-toasters | 2306.17258 | null | https://arxiv.org/abs/2306.17258v2 | https://arxiv.org/pdf/2306.17258v2.pdf | Suffering Toasters -- A New Self-Awareness Test for AI | A widely accepted definition of intelligence in the context of Artificial Intelligence (AI) still eludes us. Due to our exceedingly rapid development of AI paradigms, architectures, and tools, the prospect of naturally arising AI consciousness seems more likely than ever. In this paper, we claim that all current intelligence tests are insufficient to point to the existence or lack of intelligence \textbf{as humans intuitively perceive it}. We draw from ideas in the philosophy of science, psychology, and other areas of research to provide a clearer definition of the problems of artificial intelligence, self-awareness, and agency. We furthermore propose a new heuristic approach to test for artificial self-awareness and outline a possible implementation. Finally, we discuss some of the questions that arise from this new heuristic, be they philosophical or implementation-oriented. | ['Ira Wolfson'] | 2023-06-29 | null | null | null | null | ['philosophy'] | ['miscellaneous'] | [ 3.33490163e-01 4.38917756e-01 2.30639011e-01 -2.42159784e-01
4.16549951e-01 -5.89626491e-01 1.01639175e+00 6.17888756e-02
-5.33087075e-01 6.14163935e-01 2.35385686e-01 -6.15842104e-01
-3.93612772e-01 -6.69906437e-01 -1.52935192e-01 -4.06137466e-01
2.56503731e-01 4.12761837e-01 3.09436605e-03 -4.33940679e-01
8.19826007e-01 5.31313121e-01 -1.59240162e+00 -4.07448977e-01
1.10747766e+00 6.48710072e-01 1.30709648e-01 3.33947599e-01
1.31204994e-02 1.04293585e+00 -6.58597529e-01 -5.93361616e-01
9.13176909e-02 -9.26575899e-01 -9.59062457e-01 2.25630775e-01
-5.06478772e-02 -2.23889217e-01 -1.34713575e-01 1.17642057e+00
7.12612295e-04 -1.08023144e-01 3.28995585e-01 -1.17319930e+00
-9.21156883e-01 4.67862189e-01 1.34898105e-03 3.10894132e-01
5.56789279e-01 6.12295628e-01 4.81812894e-01 -3.18968356e-01
4.19278771e-01 1.03074229e+00 1.54453978e-01 7.61898279e-01
-8.49222541e-01 -2.77216733e-01 4.28964943e-02 3.19642097e-01
-1.12125087e+00 -6.68934941e-01 6.40882611e-01 -3.35513979e-01
1.01818419e+00 1.99243695e-01 1.30899906e+00 7.80473292e-01
3.41202915e-01 3.23920161e-01 1.62153864e+00 -7.58463919e-01
4.04748172e-01 5.06485462e-01 4.01773334e-01 5.54515600e-01
9.42926884e-01 2.29201198e-01 -5.45844853e-01 1.22056268e-01
1.05890942e+00 -3.51154059e-01 -8.49617720e-02 -2.12462276e-01
-1.51968813e+00 5.49999237e-01 1.14849128e-01 8.65358770e-01
-5.57976782e-01 -1.09864846e-01 1.00921758e-01 2.70484865e-01
-8.78624842e-02 1.08975410e+00 -1.80050522e-01 -3.56519967e-01
-2.13636711e-01 6.55364990e-02 9.01732385e-01 3.35457563e-01
6.26156688e-01 3.26344222e-01 6.24517083e-01 1.89233392e-01
3.39836180e-01 3.70032609e-01 6.44244313e-01 -1.24388659e+00
-3.24460477e-01 8.57005656e-01 1.64989144e-01 -9.80960190e-01
-3.35838944e-01 -2.05825195e-01 -2.80387461e-01 5.05978942e-01
2.15876430e-01 -4.77161556e-02 -3.58119011e-01 1.62370884e+00
8.97579119e-02 -1.86318144e-01 5.52978814e-01 8.09511542e-01
4.09447759e-01 2.33273670e-01 5.26867248e-02 -5.53313613e-01
1.19474840e+00 -5.84934235e-01 -8.82990479e-01 -6.51725888e-01
3.40158701e-01 -2.19571829e-01 9.84703600e-01 5.96004605e-01
-1.25765991e+00 -1.71145499e-01 -1.29859674e+00 -8.62040836e-03
-5.05007386e-01 -5.37961900e-01 1.62105739e+00 1.02141845e+00
-1.16621590e+00 2.17316091e-01 -7.07287073e-01 -8.10859561e-01
-4.99363467e-02 3.84337455e-01 -8.21793973e-02 2.88835198e-01
-8.50452542e-01 1.43017876e+00 7.02712238e-01 6.23003468e-02
-2.67260373e-01 2.97159731e-01 -4.96227980e-01 -2.53343314e-01
2.72796243e-01 -9.72356141e-01 9.03163135e-01 -1.72840118e+00
-1.47770321e+00 1.18823910e+00 -1.31092519e-01 -5.41360795e-01
-2.38015447e-02 6.05269894e-02 -7.79331625e-01 3.11995864e-01
-7.86841139e-02 3.58000606e-01 2.45914549e-01 -1.24100685e+00
-5.10794342e-01 -7.20697761e-01 4.11704957e-01 4.38391596e-01
-2.11043522e-01 2.38342449e-01 3.85509968e-01 -3.32327411e-02
3.20658624e-01 -6.93902850e-01 -9.58683118e-02 -3.68604869e-01
3.01000059e-01 -5.17814815e-01 2.07466528e-01 -5.60987927e-02
8.53898883e-01 -2.18010616e+00 -2.25408562e-02 -1.16181485e-01
5.57101011e-01 1.62982538e-01 8.83983374e-02 5.23368597e-01
6.93432763e-02 1.38483509e-01 -2.16597784e-02 5.09508848e-01
2.91363418e-01 3.62416565e-01 -1.74777135e-01 3.47207397e-01
-8.21809620e-02 8.40323150e-01 -9.57591772e-01 -3.49476397e-01
3.67192686e-01 3.60481203e-01 -2.91434973e-01 -1.96837187e-02
3.17582749e-02 1.87553614e-01 -6.08630478e-01 5.51796079e-01
1.96379066e-01 -3.39150637e-01 3.76229405e-01 4.41563129e-01
-5.20132422e-01 6.05810285e-01 -6.52288318e-01 1.47095752e+00
-4.14620936e-02 6.66212857e-01 -1.76331535e-01 -9.70337689e-01
7.58354783e-01 5.59123218e-01 5.42039871e-02 -9.12950039e-01
3.92971516e-01 3.17636490e-01 8.89835835e-01 -6.16075277e-01
2.25429565e-01 -3.78588617e-01 2.06001073e-01 6.76266611e-01
-5.47706746e-02 -4.87643600e-01 1.31593034e-01 3.92206572e-02
1.16810858e+00 7.50977099e-02 7.00632870e-01 -5.11932254e-01
5.07010162e-01 1.53774232e-01 4.75904882e-01 6.35543704e-01
-5.02372324e-01 -1.24065116e-01 2.39477620e-01 -7.15590298e-01
-6.22832954e-01 -9.71526325e-01 -2.13429675e-01 7.35533237e-01
3.68875444e-01 -6.33214489e-02 -9.84201074e-01 -1.53438598e-01
-5.29240727e-01 1.37327290e+00 -3.19617033e-01 -1.54290527e-01
-2.15300411e-01 -8.02142024e-01 7.98106045e-02 1.03227369e-01
6.54569447e-01 -1.58023429e+00 -1.51435673e+00 -3.52065153e-02
1.23660669e-01 -1.01938641e+00 8.94803762e-01 1.03752613e-01
-9.80528176e-01 -6.53418481e-01 9.10718516e-02 -4.40083683e-01
6.54845774e-01 4.77825105e-01 9.99003053e-01 4.87152487e-01
7.24591315e-02 8.63442421e-01 -2.82736689e-01 -8.24451864e-01
-5.13610601e-01 -5.87357223e-01 2.91439444e-01 -4.47721332e-01
1.04358292e+00 -7.88145483e-01 -4.70556468e-01 -7.72495642e-02
-9.11143124e-01 1.93819031e-01 6.73917174e-01 2.04643250e-01
-1.78388298e-01 2.36204550e-01 7.67619133e-01 -5.05327404e-01
7.39657819e-01 -3.90382797e-01 -1.59544453e-01 2.14039952e-01
-7.79311359e-01 -2.33186215e-01 2.04346105e-01 -2.38761008e-01
-1.04611099e+00 -5.16679287e-01 1.78712994e-01 3.68408561e-01
-6.95812643e-01 3.61430645e-01 -2.47465774e-01 -1.86028212e-01
7.11575449e-01 4.56650615e-01 7.58350268e-02 1.13267921e-01
2.41158724e-01 5.66895723e-01 5.63995242e-01 -5.11706293e-01
5.79191327e-01 3.87126982e-01 -2.84598142e-01 -1.18377960e+00
-7.11249352e-01 1.44878909e-01 -3.44215751e-01 -3.34902853e-01
8.20051908e-01 -6.35564923e-01 -8.86731088e-01 9.06079262e-03
-1.06386781e+00 -1.05226889e-01 -4.40739930e-01 8.05594027e-01
-9.40574944e-01 3.59398663e-01 -1.94826469e-01 -1.17670250e+00
-6.74023926e-02 -8.99498284e-01 2.01581433e-01 5.30907929e-01
-5.81785321e-01 -1.09818482e+00 -8.60437900e-02 6.63035214e-01
4.02002990e-01 5.05056307e-02 8.51672232e-01 -8.27404737e-01
-5.71727335e-01 -1.82091057e-01 7.14284331e-02 9.58297253e-02
8.88967142e-02 -1.08213551e-01 -8.63551438e-01 1.06324859e-01
9.97399747e-01 -5.48488379e-01 2.12536722e-01 -8.20044279e-02
6.69551194e-01 -4.79395539e-01 1.66915357e-02 2.38102823e-01
1.52470315e+00 8.88928771e-01 9.25440669e-01 6.01688325e-01
-1.94688827e-01 9.73459423e-01 3.54591370e-01 3.46820325e-01
5.11416495e-01 5.85967042e-02 2.37528235e-01 2.47870490e-01
2.89933652e-01 1.02746040e-01 4.20359313e-01 1.00366056e+00
-5.23120046e-01 -1.35675132e-01 -1.13829589e+00 4.30161268e-01
-1.59576869e+00 -1.17878819e+00 8.26414227e-02 2.17775917e+00
4.69789475e-01 3.11756194e-01 1.33216798e-01 1.83186352e-01
4.75369692e-01 -1.10369258e-01 -4.70017850e-01 -8.46518517e-01
-2.16755271e-01 2.17334367e-02 -4.12326939e-02 3.16361994e-01
-3.87258917e-01 1.08787072e+00 7.80877781e+00 1.47340270e-02
-8.76639903e-01 -4.21051495e-02 4.77573544e-01 2.91752994e-01
-5.98429382e-01 2.55747199e-01 -1.90471977e-01 1.92180872e-01
9.01995063e-01 -5.61488628e-01 7.81912506e-01 4.71997976e-01
3.44095901e-02 -6.60619140e-01 -1.04916322e+00 5.73860884e-01
5.13048112e-01 -7.90186524e-01 7.99492300e-02 1.80939063e-01
2.82725692e-01 -1.78682193e-01 1.00217290e-01 6.56209737e-02
4.86584336e-01 -1.11472356e+00 6.75250471e-01 4.22007740e-01
1.58230454e-01 -4.52275574e-01 5.83241642e-01 5.61935365e-01
-2.28569195e-01 -1.12702258e-01 -4.65436041e-01 -1.13306773e+00
-2.17806578e-01 3.57309394e-02 -3.45175713e-01 1.15968458e-01
1.67103872e-01 -6.48519443e-03 -6.26469314e-01 8.56508493e-01
-2.29943722e-01 2.92943597e-01 -3.46488893e-01 -4.93838221e-01
1.93871707e-01 -5.49836934e-01 5.05369425e-01 5.13293326e-01
3.45611647e-02 7.06295729e-01 -3.32112610e-01 1.05200303e+00
7.67462254e-01 -3.16499993e-02 -9.01809037e-01 -5.19332886e-01
4.25674826e-01 1.12436438e+00 -1.12783468e+00 -3.25968236e-01
-7.55062997e-01 8.51665258e-01 -3.49332690e-02 2.42673188e-01
-5.39134026e-01 -8.12680200e-02 5.56210399e-01 -1.77483618e-01
-3.50994468e-01 -4.53403115e-01 -6.67465389e-01 -1.25448859e+00
-2.42939040e-01 -1.12980807e+00 -1.76550522e-01 -8.53027821e-01
-1.06733930e+00 4.88620043e-01 9.79606900e-03 -3.74293000e-01
-4.68036592e-01 -6.87077522e-01 -6.72150314e-01 5.04392266e-01
-7.15620637e-01 -9.22949016e-01 -1.34455711e-01 2.99206167e-01
2.56961584e-01 -2.36879617e-01 1.02150214e+00 -7.00790942e-01
-3.77642483e-01 -6.44453317e-02 -4.26677018e-01 -1.14172883e-01
1.18938752e-01 -1.00507700e+00 2.65171736e-01 8.03655148e-01
3.04316892e-03 1.04712522e+00 9.76014316e-01 -4.24349159e-01
-1.64497757e+00 2.21868083e-01 9.88070786e-01 -7.70681679e-01
9.67026412e-01 -9.52787045e-03 -6.17758453e-01 8.68839025e-01
7.86586761e-01 -7.24838376e-01 8.31479132e-01 1.55197307e-01
-5.32006323e-02 8.70705247e-02 -1.17907214e+00 9.96349990e-01
1.12633002e+00 -3.36676180e-01 -1.40205145e+00 1.51529163e-01
6.11062467e-01 4.61877823e-01 -6.90195560e-01 3.92998338e-01
5.60460746e-01 -1.52254772e+00 7.92922676e-01 -2.77707815e-01
1.59354016e-01 -2.02100232e-01 -1.04192443e-01 -8.60057414e-01
-5.80111623e-01 -7.67703950e-01 3.72238576e-01 1.01288474e+00
1.37091324e-01 -1.25425041e+00 4.83270854e-01 1.00285876e+00
2.34339461e-02 -2.43411735e-01 -7.72929549e-01 -4.60524946e-01
1.88639164e-01 -3.85027260e-01 5.14500558e-01 1.28733909e+00
1.00212085e+00 5.40079415e-01 4.16646391e-01 -1.57545984e-01
6.40442908e-01 -4.53163795e-02 6.46960735e-01 -1.34026706e+00
-1.75120980e-01 -7.74955153e-01 -7.38416135e-01 -5.35995603e-01
6.01913109e-02 -4.03418988e-01 -1.94027632e-01 -1.85461485e+00
2.70279169e-01 -1.19962893e-01 -3.03908497e-01 2.17627913e-01
6.41971678e-02 7.16989934e-02 3.69999200e-01 8.64578113e-02
-6.23799264e-01 1.65850893e-01 1.00390589e+00 5.82686901e-01
7.56256506e-02 -5.08831203e-01 -1.34323013e+00 1.34123063e+00
1.22527194e+00 1.49605736e-01 -5.01357853e-01 -3.09594452e-01
4.93306905e-01 2.94703357e-02 4.23484385e-01 -1.39951456e+00
2.57756740e-01 -7.03713953e-01 4.89849359e-01 -2.03249436e-02
3.82986605e-01 -9.70520139e-01 1.69927463e-01 5.29237807e-01
1.62635697e-03 3.23764831e-01 -1.29536642e-02 -7.77556673e-02
1.82164356e-01 -4.09814149e-01 6.46899164e-01 -6.49119854e-01
-8.48835468e-01 -4.95413095e-01 -9.04181659e-01 -1.20912567e-01
1.12598443e+00 -6.17697775e-01 -5.67933142e-01 -2.30181754e-01
-4.22179282e-01 -6.17409237e-02 1.04181576e+00 1.62371799e-01
5.56084454e-01 -5.79872966e-01 -3.01896423e-01 1.28769279e-01
-4.25116196e-02 -5.81224799e-01 -7.37825856e-02 6.61047935e-01
-7.01947451e-01 6.91777349e-01 -6.78411484e-01 7.57782683e-02
-6.49806440e-01 8.63224685e-01 2.69645363e-01 5.53773642e-01
-5.44135332e-01 3.25659543e-01 6.72676146e-01 9.43298489e-02
1.72565859e-02 3.12566608e-01 -1.95391417e-01 -4.90126252e-01
6.95960283e-01 2.84781933e-01 -6.26389265e-01 -6.67873144e-01
-3.50689828e-01 2.61916310e-01 2.46414870e-01 -6.61654532e-01
1.01733458e+00 -3.42578173e-01 -6.81777477e-01 9.17182982e-01
1.35138929e-01 -6.26009926e-02 -5.36959291e-01 2.82279015e-01
9.61500630e-02 -2.93413132e-01 -1.93897218e-01 -1.23166931e+00
-2.10228041e-01 8.52879703e-01 3.25919300e-01 5.96235573e-01
1.23707819e+00 1.45330116e-01 2.17095353e-02 5.78883827e-01
6.59776092e-01 -1.21547246e+00 -6.48067445e-02 4.01973039e-01
7.39708185e-01 -9.45044219e-01 1.74057260e-01 -7.39335865e-02
-7.15389669e-01 9.92269337e-01 8.58590722e-01 -2.18523711e-01
1.63221478e-01 8.97445232e-02 2.11819425e-01 -3.65273178e-01
-1.09061778e+00 -5.14127791e-01 -3.53239059e-01 8.96290481e-01
7.12677479e-01 1.21532299e-01 -9.32240486e-01 2.02637136e-01
-6.58118546e-01 6.63591325e-02 8.42982948e-01 7.69417763e-01
-9.96583164e-01 -8.11606288e-01 -4.74101305e-01 1.94937997e-02
-4.73910838e-01 1.80436969e-01 -1.09642482e+00 9.41271424e-01
4.09090340e-01 1.12210667e+00 2.37812623e-01 -1.53356984e-01
-3.04327130e-01 2.01875180e-01 8.40660214e-01 -4.33483750e-01
-5.78579605e-01 -6.89918101e-02 3.93217951e-02 -1.98981449e-01
-7.75035858e-01 -7.65652835e-01 -1.26896584e+00 -4.74153489e-01
-7.68578500e-02 3.82275343e-01 7.54545748e-01 1.00242305e+00
-4.39344645e-02 9.76518765e-02 8.71943161e-02 -2.68053800e-01
-1.68758556e-01 -8.34106445e-01 -5.10669470e-01 8.84742513e-02
-9.65055600e-02 -3.94757599e-01 -5.20190775e-01 -1.58158794e-01] | [9.005188941955566, 6.356170654296875] |
615a2f29-1ff0-4163-a836-35b67b002539 | fenrir-physics-enhanced-regression-for | 2202.01287 | null | https://arxiv.org/abs/2202.01287v2 | https://arxiv.org/pdf/2202.01287v2.pdf | Fenrir: Physics-Enhanced Regression for Initial Value Problems | We show how probabilistic numerics can be used to convert an initial value problem into a Gauss--Markov process parametrised by the dynamics of the initial value problem. Consequently, the often difficult problem of parameter estimation in ordinary differential equations is reduced to hyperparameter estimation in Gauss--Markov regression, which tends to be considerably easier. The method's relation and benefits in comparison to classical numerical integration and gradient matching approaches is elucidated. In particular, the method can, in contrast to gradient matching, handle partial observations, and has certain routes for escaping local optima not available to classical numerical integration. Experimental results demonstrate that the method is on par or moderately better than competing approaches. | ['Philipp Hennig', 'Nathanael Bosch', 'Filip Tronarp'] | 2022-02-02 | null | null | null | null | ['numerical-integration'] | ['miscellaneous'] | [-2.24376634e-01 -5.13592474e-02 -2.55150318e-01 -4.59349416e-02
-1.12296748e+00 -5.36479294e-01 7.75903106e-01 -1.49999633e-02
-7.37770438e-01 1.28873277e+00 -3.39877069e-01 -4.10685927e-01
-3.81243229e-01 -4.21486706e-01 -2.89075077e-01 -1.23995233e+00
-6.68220967e-02 8.56535673e-01 1.41152099e-01 -3.45229208e-01
5.47251582e-01 6.69062436e-01 -1.10522985e+00 -5.49736142e-01
7.77278781e-01 8.76697600e-01 -3.10889602e-01 8.69562566e-01
-1.83982208e-01 2.78907537e-01 -3.29326391e-01 -2.83205450e-01
7.12221116e-02 -4.79903519e-01 -8.13010037e-01 4.64318916e-02
-9.48563889e-02 2.36527979e-01 -6.26692101e-02 1.26086760e+00
3.74098033e-01 5.47949672e-01 1.21930742e+00 -1.12169027e+00
-2.43393198e-01 1.76970661e-01 -5.65877974e-01 3.36168528e-01
4.16068912e-01 1.54814318e-01 5.10192335e-01 -8.16839516e-01
5.43621004e-01 1.49135876e+00 1.02388358e+00 2.56409377e-01
-1.63739920e+00 -1.43602446e-01 1.18930429e-01 -2.47597411e-01
-1.63278425e+00 -2.45574072e-01 3.70589703e-01 -6.00537479e-01
7.15779841e-01 3.67373191e-02 5.78018546e-01 7.74493575e-01
4.93508607e-01 5.20196736e-01 1.31596196e+00 -4.88399059e-01
4.91434276e-01 1.95625827e-01 9.87922922e-02 8.13263178e-01
9.88773853e-02 4.55761403e-01 -3.07351589e-01 -6.75464928e-01
7.96046078e-01 -4.65242863e-01 -1.52427420e-01 -6.31323993e-01
-1.15617752e+00 1.10178542e+00 -6.43516928e-02 1.41310230e-01
-3.44494849e-01 3.54652524e-01 1.25500545e-01 3.02332461e-01
5.55724323e-01 6.72060788e-01 -4.31110531e-01 -3.85028273e-01
-1.03555226e+00 7.21520603e-01 1.14108014e+00 7.92153478e-01
6.36105180e-01 3.18057358e-01 1.89309707e-03 3.58834684e-01
4.84134793e-01 7.95972168e-01 3.05917889e-01 -1.34293735e+00
1.83897033e-01 -1.36629688e-02 7.74880350e-01 -5.65294683e-01
-5.37794173e-01 -3.26333523e-01 -8.04781914e-01 5.07297158e-01
7.18412101e-01 -4.62237954e-01 -5.95895410e-01 1.53087974e+00
5.12615740e-01 1.61628097e-01 1.33951947e-01 5.32911658e-01
-6.40279651e-02 9.26208854e-01 -1.25327379e-01 -8.50195706e-01
9.32839632e-01 -5.30985057e-01 -8.73633444e-01 2.70754606e-01
5.62200785e-01 -7.63901234e-01 6.17959499e-01 4.85932559e-01
-1.37417400e+00 -2.36659452e-01 -8.91046584e-01 4.56664920e-01
-2.30871782e-01 -3.80726039e-01 5.68847001e-01 7.62083650e-01
-1.21464872e+00 1.05721891e+00 -1.07646811e+00 -1.89266652e-02
-4.57467213e-02 5.32217085e-01 1.76931024e-01 5.45427918e-01
-1.11228669e+00 1.03893209e+00 3.99115652e-01 2.63092160e-01
-4.87412333e-01 -6.10056460e-01 -8.50487828e-01 -1.72307745e-01
3.22749466e-01 -6.69187665e-01 1.60268044e+00 -5.67745805e-01
-2.12352991e+00 3.95603865e-01 -3.21005553e-01 -4.12558913e-01
8.30078781e-01 -2.65977979e-02 -1.16060741e-01 8.31063241e-02
-2.13462532e-01 3.93842936e-01 9.14444327e-01 -1.11023653e+00
-3.82571846e-01 -1.26815006e-01 -4.36389357e-01 3.25941503e-01
3.22576940e-01 1.98245957e-01 -2.49610841e-01 -5.03445625e-01
1.67093948e-01 -1.09616911e+00 -8.49297881e-01 -7.26969317e-02
-2.38054797e-01 -2.70694047e-01 4.93882567e-01 -1.68978512e-01
1.33573675e+00 -1.59367764e+00 4.00209278e-01 5.75017691e-01
-1.27651125e-01 9.50032920e-02 2.75755942e-01 6.01834834e-01
6.25784509e-03 1.63192064e-01 -6.18454337e-01 -3.81427258e-01
3.81624103e-02 9.65105370e-02 -3.42972755e-01 9.80374634e-01
-1.83312651e-02 7.08804905e-01 -1.02881014e+00 -6.11709237e-01
2.27717996e-01 3.64348352e-01 -4.74788398e-01 -1.78508446e-01
-2.16932967e-01 4.67920452e-01 -6.86967731e-01 3.62399340e-01
5.25367022e-01 -1.88769326e-01 8.14890340e-02 4.18589205e-01
-3.15637499e-01 -1.66846424e-01 -1.78120971e+00 1.16338468e+00
-3.73784930e-01 4.78488594e-01 3.23940217e-01 -1.23028290e+00
7.68112779e-01 3.76764029e-01 4.62373972e-01 2.58857552e-02
1.99723333e-01 3.22025388e-01 -3.48704129e-01 -1.95928738e-01
5.45176387e-01 -5.87681532e-01 -1.17427826e-01 2.26337135e-01
-1.03132963e-01 -7.93980002e-01 1.90670252e-01 -1.28004819e-01
4.56580728e-01 4.46795404e-01 6.12465739e-01 -7.10239708e-01
5.79171360e-01 3.33802611e-01 4.81293380e-01 9.36451435e-01
-1.27989680e-01 4.01106954e-01 6.89525187e-01 -2.73221821e-01
-7.74015367e-01 -1.10721362e+00 -6.46043181e-01 5.23328662e-01
2.71184236e-01 -1.34323150e-01 -1.03008652e+00 -2.42688715e-01
1.19072869e-01 5.74097991e-01 -6.70809209e-01 4.21169400e-02
-5.93508184e-01 -1.30501282e+00 4.58534449e-01 2.20648095e-01
3.03498685e-01 -8.11765015e-01 -3.54744792e-01 6.24118388e-01
1.10170193e-01 -6.96866572e-01 -3.32205057e-01 2.60125697e-01
-1.30225158e+00 -7.26471186e-01 -1.03830600e+00 -3.33783686e-01
3.54378343e-01 -2.31696710e-01 1.02211595e+00 -2.88120329e-01
-1.61524206e-01 3.91520351e-01 2.76514947e-01 -9.69090536e-02
-6.95149958e-01 8.30595791e-02 3.58434916e-01 -1.70455530e-01
5.08335568e-02 -4.83594716e-01 -4.62418467e-01 5.79854488e-01
-6.69703186e-01 -6.65664613e-01 2.16517493e-01 1.03062129e+00
6.52744174e-01 8.76623392e-02 4.58212584e-01 -8.23381186e-01
7.42185056e-01 -4.69011992e-01 -1.20886493e+00 -3.45552899e-02
-9.89128232e-01 5.62078774e-01 4.77283537e-01 -6.01017594e-01
-1.29810154e+00 1.68611363e-01 6.12095259e-02 -2.18112782e-01
2.55619865e-02 5.33386886e-01 2.91960955e-01 -3.54091853e-01
7.70973623e-01 -1.69501588e-01 2.82088250e-01 -5.72073817e-01
2.96986490e-01 3.89192522e-01 5.09011447e-01 -8.32165360e-01
7.84029961e-01 4.78714645e-01 6.52843535e-01 -9.50818241e-01
-5.39018571e-01 -5.41241050e-01 -6.62841439e-01 -1.06466748e-01
8.96972537e-01 -4.73746747e-01 -8.25929165e-01 6.67461216e-01
-9.98129129e-01 -4.80021268e-01 -5.14791310e-01 6.59801841e-01
-1.08632827e+00 4.88604903e-01 -7.53471851e-01 -1.24304736e+00
4.11279976e-01 -1.15982890e+00 8.04399312e-01 4.31088448e-01
-2.04462394e-01 -1.56622922e+00 5.11330664e-01 -2.05829710e-01
3.78947258e-01 2.98474282e-01 7.19332755e-01 -3.83994341e-01
-3.48158896e-01 -4.27900553e-01 1.78650379e-01 7.68575743e-02
-3.00864518e-01 4.26187545e-01 -6.39898837e-01 -2.63920814e-01
3.52613926e-01 -9.66032371e-02 7.48929143e-01 8.96130979e-01
5.96074581e-01 -1.56964153e-01 -5.49198508e-01 5.85606635e-01
1.45470309e+00 2.04312012e-01 4.17120934e-01 3.49884629e-01
1.11835107e-01 5.16827166e-01 8.52739751e-01 5.55264831e-01
-1.91687606e-02 8.31345260e-01 1.28369657e-02 -8.42919052e-02
5.36744714e-01 -1.54833347e-02 3.23505014e-01 5.82113564e-01
-1.89395458e-01 -2.89174933e-02 -8.84818912e-01 2.34635428e-01
-2.03357768e+00 -8.88332844e-01 -4.72357482e-01 2.36840820e+00
1.04861450e+00 1.03881046e-01 3.49189043e-01 -9.82987508e-02
7.66333342e-01 -9.04867500e-02 -4.38404500e-01 -6.58814251e-01
-4.67016324e-02 -4.39045243e-02 7.60524273e-01 1.13461769e+00
-1.11842275e+00 8.52353513e-01 8.33456039e+00 1.03396213e+00
-6.06725156e-01 3.59706767e-03 5.48750937e-01 2.19994903e-01
-4.07770760e-02 1.87993586e-01 -1.08724713e+00 4.28473771e-01
1.08703697e+00 -2.24205688e-01 4.82727706e-01 7.53439844e-01
3.60080153e-01 -5.90306461e-01 -8.51082742e-01 6.80408120e-01
-2.71930635e-01 -1.07255089e+00 -2.78181702e-01 2.47688010e-01
1.07217586e+00 -3.20886254e-01 2.42951706e-01 2.95043498e-01
6.73872590e-01 -1.12109220e+00 4.64161873e-01 8.60813797e-01
3.59112144e-01 -9.77196634e-01 8.50712895e-01 4.89950418e-01
-1.06124806e+00 1.22674309e-01 -5.05012035e-01 -1.03637747e-01
5.10382950e-01 5.46469390e-01 -4.45533097e-01 4.30707455e-01
3.91295940e-01 1.69802308e-01 -7.08302110e-02 1.41171277e+00
9.07521620e-02 5.25227070e-01 -6.46668911e-01 -3.36331666e-01
6.26005113e-01 -9.46711361e-01 9.53813195e-01 1.31092656e+00
3.90733451e-01 1.40072197e-01 1.52597517e-01 7.46563852e-01
6.01195931e-01 2.13583738e-01 -6.79907680e-01 1.68518707e-01
1.66823745e-01 8.11678827e-01 -8.53995621e-01 -4.37893689e-01
2.44834945e-02 6.49306297e-01 1.03249095e-01 7.44286776e-01
-6.81694448e-01 -5.16728818e-01 5.51813364e-01 -1.59206107e-01
3.45948607e-01 -4.04680192e-01 -1.66633517e-01 -1.02103698e+00
-4.31074321e-01 -7.28794932e-01 5.19209743e-01 -5.75695634e-01
-9.89797413e-01 4.81762916e-01 5.95114291e-01 -9.03491020e-01
-9.22800124e-01 -8.61613154e-01 -4.54471022e-01 1.02402174e+00
-1.06864965e+00 -3.70379239e-01 4.30095911e-01 4.13230240e-01
3.67224514e-01 1.03110924e-01 6.33938432e-01 -4.44685549e-01
-5.25536776e-01 3.03023934e-01 8.23179722e-01 -4.71720159e-01
2.54013270e-01 -1.59190679e+00 2.59523571e-01 6.81621671e-01
-1.30866230e-01 5.52109420e-01 1.16714740e+00 -5.59548855e-01
-1.20623207e+00 -4.61783648e-01 6.09075785e-01 -4.86680955e-01
1.06024981e+00 2.20200270e-02 -1.03907382e+00 4.41909730e-01
3.19343731e-02 -2.27517039e-01 1.69946924e-01 1.93701148e-01
2.04623356e-01 3.84409130e-01 -1.09749329e+00 7.48466849e-01
4.62670416e-01 -1.92753211e-01 -5.57024777e-01 4.27907884e-01
1.85836911e-01 -4.52743709e-01 -8.80580485e-01 1.58915728e-01
2.70860046e-01 -5.51568270e-01 1.17230785e+00 -8.24353755e-01
-1.84533432e-01 -2.83154756e-01 9.43626985e-02 -1.37667763e+00
-4.53821532e-02 -1.62728834e+00 -3.38677727e-02 9.80665028e-01
4.93018925e-01 -8.45956028e-01 6.80456400e-01 6.46125674e-01
2.56488025e-01 -7.63935924e-01 -1.31284869e+00 -1.11386919e+00
7.12938368e-01 -3.93375665e-01 2.54704565e-01 6.01267397e-01
-6.91206614e-03 6.40482118e-04 -2.38370508e-01 1.95588484e-01
9.52372134e-01 5.83447665e-02 3.90796065e-01 -1.29575586e+00
-4.72041816e-01 -7.99713433e-01 -1.97047204e-01 -1.15169024e+00
3.58064830e-01 -5.66075981e-01 3.97848368e-01 -1.06672180e+00
-1.57527342e-01 -5.53939879e-01 -2.05052465e-01 -2.50260055e-01
-2.75165439e-01 4.64526452e-02 -2.85590321e-01 1.86337709e-01
-4.12893921e-01 6.21791720e-01 1.14702189e+00 3.09672534e-01
-6.26125038e-01 6.51452720e-01 -1.26141131e-01 1.13049698e+00
7.31254220e-01 -6.71468675e-01 -1.04182765e-01 4.38046485e-01
4.71908003e-01 6.96460009e-01 2.30533883e-01 -7.43660629e-01
3.54391396e-01 -3.94138694e-01 1.41262382e-01 -5.53076029e-01
4.43993807e-01 -2.30474263e-01 6.15702234e-02 4.36672479e-01
-2.87095726e-01 3.94857585e-01 1.97729185e-01 7.62792826e-01
-2.39682510e-01 -1.02638400e+00 1.07788324e+00 -1.83746919e-01
-1.65975124e-01 9.87469256e-02 -1.16796720e+00 4.81941611e-01
7.28421211e-01 -1.70737937e-01 2.10535973e-01 -6.42638206e-01
-1.00501680e+00 2.82788187e-01 6.30630732e-01 -4.34453577e-01
2.42603362e-01 -1.08308482e+00 -4.77402240e-01 -1.78967863e-01
-4.94004667e-01 -1.67940140e-01 1.07816666e-01 1.06458914e+00
-5.36512971e-01 3.59551907e-01 3.26657742e-01 -6.74393177e-01
-9.20945883e-01 7.52932549e-01 7.38571703e-01 -4.61932212e-01
-3.88346463e-01 7.11730421e-01 -4.42974940e-02 -5.01172602e-01
1.98970139e-01 -4.47042316e-01 -3.61474603e-02 1.23112872e-01
3.07882726e-01 7.57589042e-01 -2.91118741e-01 -4.23676461e-01
-9.34703797e-02 8.44397962e-01 2.04657465e-01 -8.11225653e-01
9.54923332e-01 -3.94677192e-01 -6.15942702e-02 8.37642908e-01
1.11036551e+00 -1.21461093e-01 -1.52706695e+00 -2.12532356e-01
2.47419611e-01 -1.47777334e-01 1.76602930e-01 -4.34973121e-01
-6.16999447e-01 8.47034156e-01 4.73130047e-01 3.75090986e-01
5.59528887e-01 -1.83462888e-01 2.37294018e-01 9.15090740e-01
2.20521972e-01 -1.28896952e+00 -2.92441607e-01 5.59117138e-01
7.25655138e-01 -1.11943364e+00 1.58746943e-01 -3.13920826e-01
-5.90343952e-01 1.32425213e+00 1.37564149e-02 -4.00685847e-01
9.36931014e-01 5.86253047e-01 3.37382257e-02 7.01855728e-03
-7.22088575e-01 -7.05411583e-02 2.06547111e-01 4.74887788e-01
-2.27447879e-03 -2.05182582e-01 -1.45323500e-01 5.45940548e-02
-3.39549109e-02 -1.09113559e-01 5.49530268e-01 8.30680609e-01
-4.49584693e-01 -1.06465220e+00 -6.93435311e-01 6.27940670e-02
-6.15091980e-01 -4.78305295e-02 1.38663605e-01 1.18433571e+00
-5.51758766e-01 7.93909252e-01 -7.40762502e-02 4.31459844e-01
5.68437576e-02 3.57709855e-01 4.81365472e-01 -2.06459075e-01
-5.10213673e-01 3.63156199e-01 1.29885167e-01 -6.64385855e-01
-3.33686650e-01 -1.22983563e+00 -1.04575443e+00 -1.87495545e-01
-4.56934929e-01 6.40816748e-01 6.34394109e-01 1.01371169e+00
-2.40245625e-01 -7.46507384e-03 6.58385694e-01 -1.14589262e+00
-1.30413675e+00 -6.61959231e-01 -7.36766636e-01 -3.32603455e-01
4.41689998e-01 -9.65720773e-01 -1.01374364e+00 -1.65730670e-01] | [6.626845836639404, 3.932269811630249] |
cc5c8506-fe6a-47ce-ac69-d57c4c36df8e | cnn-based-autoencoder-application-in-breast | null | null | https://ieeexplore.ieee.org/document/9502205 | https://ieeexplore.ieee.org/document/9502205 | CNN Based Autoencoder Application in Breast Cancer Image Retrieval | Content Based Medical Image Retrieval (CBMIR) is considered as a common technique to retrieve relevant images by comparing the features contained in the query image with the features contained in the image located in the database. Currently, the study related to CBMIR on breast cancer image however remains challenging due to inadequate research in such area. Previous study has a low performance and misinformation emphasizing the feature extraction process. Therefore, this study aims to utilize the CNN based Autoencoder method to minimize misinformation in the feature extraction process and to improve the performance result. The dataset used in this study is the BreakHis dataset. Overall, the results of image retrieval in breast cancer applying the CNN based Autoencoder method achieved higher performance compared to the method used in the previous study with an average precision of 0.9237 in the mainclass dataset category and 0.6825 in the subclass dataset category. | ['Fauzi Dwi Setiawan Sumadi', 'Lailatul Husniah', 'Trfebi Shina Sabrila', 'Kharisma Muzaki Ghufron', 'Agus Eko Minarno'] | 2021-08-04 | null | null | null | international-seminar-on-intelligent | ['medical-image-retrieval', 'medical-image-retrieval'] | ['computer-vision', 'medical'] | [-8.67924392e-02 -2.57266700e-01 -1.92845955e-01 -6.85965791e-02
-6.43194675e-01 1.86574087e-02 4.54894900e-01 5.83914220e-01
-7.78255939e-01 5.54655254e-01 3.71766001e-01 1.23063035e-01
-6.47560358e-01 -9.29821253e-01 -1.62727222e-01 -8.83789897e-01
2.85303473e-01 1.29610971e-01 6.34869114e-02 -2.61353731e-01
5.58124661e-01 4.77814406e-01 -1.72524524e+00 7.21728563e-01
3.32134813e-01 1.05555439e+00 4.53611404e-01 6.48536921e-01
-2.02637359e-01 9.59822416e-01 -9.19213712e-01 8.16807449e-02
1.31177651e-02 -3.60225290e-01 -8.89000654e-01 -1.28412724e-01
7.96994865e-02 -6.42390072e-01 -4.94597584e-01 9.87985253e-01
7.27184176e-01 2.42864266e-01 8.68756771e-01 -7.86774397e-01
-7.37930119e-01 3.50659132e-01 -2.44567141e-01 7.15502381e-01
-1.38485432e-02 -7.20453322e-01 3.22693586e-01 -7.69370496e-01
7.48315334e-01 9.06709969e-01 4.39781874e-01 2.22599521e-01
-4.61463511e-01 -7.30218291e-01 -6.21713340e-01 5.62797189e-01
-1.69027352e+00 -2.18611732e-02 6.25099599e-01 -3.94801795e-01
9.52902913e-01 2.70839423e-01 6.74823701e-01 1.67968959e-01
8.63701224e-01 3.88594031e-01 1.02513564e+00 -8.34110916e-01
-3.36376131e-02 4.77258176e-01 2.92961121e-01 6.27913475e-01
3.20828587e-01 1.69258639e-01 -3.68663669e-01 -7.00542331e-02
4.36575294e-01 4.15497780e-01 5.28575927e-02 4.39542502e-01
-7.62361526e-01 1.18004537e+00 6.21273935e-01 1.28079081e+00
-5.92022538e-01 -2.39436373e-01 5.75058401e-01 3.00643861e-01
3.58239204e-01 3.71753782e-01 -4.21072751e-01 2.52070159e-01
-1.04276264e+00 -3.12353261e-02 3.98320794e-01 6.41685963e-01
4.72966224e-01 -7.78064653e-02 -1.64948881e-01 8.01331937e-01
1.82626337e-01 4.04536515e-01 1.13152325e+00 -4.80836421e-01
-2.17883766e-01 7.71720946e-01 -5.22526920e-01 -1.67029989e+00
-3.03621620e-01 -4.18454677e-01 -7.87399292e-01 -1.77052002e-02
-7.12580979e-02 7.48606548e-02 -1.13250077e+00 9.45844412e-01
1.72474146e-01 -2.63733953e-01 5.58299124e-01 9.33938265e-01
1.58470547e+00 6.52249992e-01 3.05771440e-01 1.11375973e-01
1.70774913e+00 -6.52073741e-01 -1.16321528e+00 5.46792865e-01
7.22012103e-01 -1.21721303e+00 2.78007150e-01 1.95639685e-01
-5.41388869e-01 -5.51803589e-01 -1.28396702e+00 -2.61347666e-02
-9.48592603e-01 6.03811204e-01 5.51496804e-01 4.06001747e-01
-1.05356777e+00 4.32839632e-01 -6.05847538e-01 -7.06035376e-01
3.72357607e-01 4.18266803e-01 -6.97050750e-01 -1.94791049e-01
-1.03379679e+00 9.79707181e-01 8.73585761e-01 -1.33858845e-01
-7.55299449e-01 -5.44431508e-01 -5.77704787e-01 -5.69007173e-02
-8.95480625e-03 -2.74328947e-01 1.02124071e+00 -1.24001014e+00
-9.37810898e-01 8.31923902e-01 1.46491617e-01 -5.18606126e-01
-1.09819090e-02 3.96731980e-02 -5.27551174e-01 7.58369863e-01
1.20773420e-01 7.28260100e-01 1.66067868e-01 -9.52471793e-01
-8.09801280e-01 -4.13591951e-01 -1.13734864e-01 2.62149721e-01
-5.77370644e-01 8.84509161e-02 -4.20598805e-01 -7.65382946e-01
4.28942293e-01 -7.14462817e-01 8.46274644e-02 -1.41852692e-01
-1.61892906e-01 -2.18869150e-01 9.93029833e-01 -8.55738282e-01
1.09611475e+00 -2.17257404e+00 -5.60477614e-01 3.11243206e-01
-5.44564761e-02 4.32236403e-01 1.28679872e-01 5.56304932e-01
-2.46167585e-01 2.72037685e-01 2.73024321e-01 4.74016160e-01
-7.89374888e-01 1.04829527e-01 4.08822566e-01 4.64831591e-01
-2.35043969e-02 5.29280066e-01 -4.84618545e-01 -1.10479939e+00
4.32891339e-01 7.26761401e-01 -1.81650177e-01 1.50544047e-01
5.17982125e-01 -3.58256772e-02 -6.77928984e-01 8.24383438e-01
6.18955016e-01 -1.47715643e-01 -1.86045513e-01 -6.69673383e-01
5.92964962e-02 -4.98268604e-01 -7.72641957e-01 1.36141801e+00
-3.02028447e-01 1.05127048e+00 -3.71853679e-01 -9.81934726e-01
8.44324708e-01 7.33882964e-01 8.44203115e-01 -9.89450097e-01
6.71411335e-01 8.77907500e-02 1.74491957e-01 -1.06580639e+00
5.49270272e-01 -5.76971285e-02 5.88351846e-01 7.19579160e-02
4.88345847e-02 3.04595232e-01 7.19391480e-02 2.04229012e-01
7.22102821e-01 -4.81217772e-01 5.79410017e-01 -4.80096102e-01
5.68923593e-01 5.71681857e-01 1.37572303e-01 7.46345282e-01
-2.83525974e-01 4.37651247e-01 -1.45130098e-01 -4.34502721e-01
-9.17365253e-01 -4.44674581e-01 -5.99916637e-01 4.78830606e-01
4.15540747e-02 -2.96187345e-02 -6.27090394e-01 -2.19578862e-01
-2.11986974e-01 2.03957930e-01 -9.43028808e-01 -3.49283606e-01
-1.70121655e-01 -8.91553044e-01 5.64432442e-01 3.15457702e-01
1.01636732e+00 -1.13304400e+00 -7.41852522e-01 2.94261761e-02
-6.58605322e-02 -4.93850201e-01 2.80675087e-02 -1.92242451e-02
-1.17050886e+00 -1.34906626e+00 -8.43871474e-01 -1.13459933e+00
8.88127625e-01 3.86510313e-01 7.26972938e-01 6.69410706e-01
-8.24467838e-01 1.90241486e-01 -7.25657582e-01 -8.20857525e-01
-3.84996742e-01 1.47553682e-01 -6.92974389e-01 -5.22140443e-01
1.03475809e+00 2.82244682e-01 -8.31256509e-01 -2.16972411e-01
-1.35383987e+00 -3.52329880e-01 9.36476707e-01 1.03854799e+00
6.83498561e-01 7.83940613e-01 3.78882408e-01 -6.71903372e-01
6.58014953e-01 -4.88687754e-01 -3.15457195e-01 1.13191172e-01
-9.98218000e-01 -3.17860693e-01 1.52834440e-02 -2.49723583e-01
-8.45980525e-01 -2.68422723e-01 5.11987992e-02 2.47487333e-02
-2.50890106e-01 1.00361991e+00 3.84352267e-01 -2.42722824e-01
6.25219762e-01 2.48235032e-01 4.17715877e-01 -2.86848933e-01
-5.46576440e-01 1.00691509e+00 2.32944965e-01 1.56788304e-01
2.17918217e-01 4.41646874e-01 1.57122254e-01 -8.50057542e-01
-5.83823204e-01 -8.34616780e-01 -4.16098297e-01 -3.48137289e-01
1.18661785e+00 -1.06035566e+00 -5.70356071e-01 1.54923588e-01
-8.74858141e-01 4.71222430e-01 2.38155887e-01 1.05352592e+00
-3.24980193e-03 4.35250625e-02 -6.05316699e-01 -6.87331915e-01
-8.83009315e-01 -1.17453790e+00 7.05425501e-01 4.72416312e-01
-1.60314527e-03 -8.38024020e-01 -1.28946314e-02 1.85878396e-01
6.42152786e-01 2.45685622e-01 1.08524036e+00 -1.12574971e+00
-2.76291072e-01 -8.82006884e-01 -4.30843771e-01 2.77843982e-01
4.06042814e-01 -1.66679919e-02 -9.01835382e-01 -1.32966846e-01
1.53588891e-01 9.15533006e-02 8.25900137e-01 6.78287804e-01
1.18053913e+00 -3.40695083e-01 -4.74824637e-01 -4.37479327e-03
2.13082552e+00 8.40754807e-01 7.71126330e-01 7.89649248e-01
9.60283205e-02 5.36532164e-01 9.03618097e-01 1.25358254e-01
-1.50569677e-01 2.64676899e-01 1.81664839e-01 -1.85997903e-01
-2.61079699e-01 8.61393884e-02 -4.32966590e-01 7.60222554e-01
1.32523164e-01 -1.89338535e-01 -8.32407236e-01 7.00303018e-01
-1.56581831e+00 -8.72002959e-01 5.49192633e-03 1.77322590e+00
5.45360386e-01 -3.46198440e-01 -4.00431871e-01 4.42266047e-01
6.32760048e-01 -3.45557362e-01 5.72058000e-02 -2.69903958e-01
-4.88860421e-02 2.64711112e-01 5.27599871e-01 2.51165628e-01
-1.21825147e+00 5.99664927e-01 6.30969524e+00 9.56966162e-01
-1.32188654e+00 1.10743798e-01 6.99638724e-01 2.27841198e-01
3.22244674e-01 -2.53042221e-01 -5.27474582e-01 2.46623978e-01
1.04884136e+00 -1.13520525e-01 -1.11335777e-01 9.83036160e-01
1.85219839e-01 -7.44118094e-01 -4.93245661e-01 1.09407794e+00
6.15309894e-01 -1.37300277e+00 3.72248650e-01 6.44813925e-02
5.68097472e-01 -2.91523457e-01 -7.33003439e-03 5.47298566e-02
-3.77789229e-01 -1.17703938e+00 1.69772692e-02 9.81974900e-01
3.11950415e-01 -1.05808926e+00 1.73616505e+00 1.97215304e-02
-7.30325401e-01 -3.01829297e-02 -6.32905841e-01 2.18114913e-01
-7.33900607e-01 3.74747843e-01 -1.38019824e+00 5.58624923e-01
1.03723359e+00 4.52708006e-01 -8.25631618e-01 1.32813036e+00
5.30988634e-01 4.18037832e-01 8.41121748e-02 -2.86686689e-01
1.19413614e-01 1.71543956e-01 1.99392572e-01 1.12995148e+00
4.34852391e-01 1.83810160e-01 -4.77664918e-02 2.05355093e-01
1.10929295e-01 8.83644044e-01 -8.16708684e-01 -1.16214454e-01
2.68487185e-01 1.12702560e+00 -9.45750892e-01 -5.77733397e-01
-1.97882190e-01 6.08579576e-01 -4.68232393e-01 2.33203113e-01
-2.50125766e-01 -8.77403438e-01 -7.79898241e-02 -2.70321537e-02
1.45846203e-01 4.21274751e-01 -1.28523573e-01 -4.54331160e-01
-2.42972195e-01 -8.35843205e-01 7.56258070e-01 -9.25312877e-01
-9.64570642e-01 9.08600688e-01 2.81619489e-01 -1.28380454e+00
-4.36053127e-02 -5.93598723e-01 -1.02337182e-01 8.60082567e-01
-1.39786911e+00 -1.18839991e+00 -4.69381958e-01 5.00343323e-01
5.54283381e-01 -6.46309912e-01 1.21258783e+00 4.89835799e-01
-3.88040870e-01 4.15271878e-01 5.16118526e-01 5.73929667e-01
8.24435949e-01 -8.04594398e-01 -1.29998064e+00 5.19196868e-01
-2.73120910e-01 6.40069067e-01 5.71707487e-01 -8.57186317e-01
-1.06325674e+00 -8.36728632e-01 8.93440664e-01 7.82293230e-02
5.40361032e-02 6.86455190e-01 -6.33202195e-01 2.61084616e-01
6.91905439e-01 -2.97766060e-01 1.00204527e+00 -4.46942866e-01
1.48776129e-01 -2.71495700e-01 -1.60187352e+00 3.57960433e-01
-2.70421505e-01 -5.46227038e-01 -4.45694178e-01 3.38298202e-01
2.46455058e-01 -1.94018587e-01 -1.33874023e+00 4.00917470e-01
7.06338584e-01 -5.09081841e-01 8.32534194e-01 -5.24770081e-01
7.10017800e-01 -2.45638385e-01 -4.76399720e-01 -8.11852098e-01
-2.83139765e-01 6.07540250e-01 5.07601261e-01 1.10849345e+00
3.99215579e-01 -2.83304572e-01 8.33846867e-01 4.51518983e-01
8.06270540e-02 -6.91257000e-01 -6.97630525e-01 -3.33299935e-01
1.45233348e-01 1.14995167e-01 4.81671572e-01 7.01531291e-01
-2.57190973e-01 -3.70985359e-01 1.58687010e-01 -1.95679758e-02
2.69014776e-01 -3.95377278e-02 2.53004611e-01 -1.12308633e+00
3.04831862e-01 -1.38106078e-01 -8.15884888e-01 7.92861208e-02
-1.65027425e-01 -9.76445258e-01 2.40385085e-02 -1.69445133e+00
8.76881778e-01 -9.96394232e-02 -5.85743189e-01 3.57608676e-01
-4.13159505e-02 4.71743584e-01 9.99940261e-02 5.19517779e-01
-1.12248197e-01 6.47072643e-02 1.28712082e+00 -4.65880930e-01
-1.65489092e-02 -3.92477304e-01 -7.35901654e-01 5.39885938e-01
1.07047331e+00 -8.84808481e-01 -4.52922821e-01 -2.57903636e-01
-2.17870563e-01 2.72160601e-02 1.74663514e-01 -1.12633419e+00
4.85933959e-01 6.44749403e-02 1.05258667e+00 -1.17970109e+00
2.24971175e-01 -1.23911214e+00 2.54285812e-01 7.37518132e-01
-5.01220763e-01 3.44613284e-01 3.51400048e-01 4.47177023e-01
-6.90492928e-01 -7.63410270e-01 5.76375186e-01 -4.51058149e-01
-8.66568089e-01 1.24597840e-01 -7.71923602e-01 -6.03300095e-01
1.01499128e+00 -2.81754941e-01 -2.62251228e-01 -3.32124770e-01
-6.30565286e-01 -3.39805514e-01 5.60788484e-03 2.65385866e-01
8.56554329e-01 -1.44402218e+00 -4.97244239e-01 -2.58630663e-01
4.18738902e-01 -3.03423882e-01 4.41852450e-01 7.44558036e-01
-1.10439599e+00 8.75158668e-01 -4.52273905e-01 -3.55042815e-01
-1.72846556e+00 5.27602613e-01 4.61056024e-01 -1.20393604e-01
-4.88238990e-01 2.43153304e-01 -2.87077606e-01 2.21866265e-01
1.11618534e-01 1.34373233e-01 -1.18282592e+00 2.34995484e-01
8.48566473e-01 2.77611971e-01 5.29159367e-01 -9.41402793e-01
-3.24102253e-01 5.25352716e-01 -4.63456482e-01 -5.99870235e-02
1.26282263e+00 -3.58521678e-02 -3.77942413e-01 1.72957599e-01
1.72179592e+00 -4.84896272e-01 -1.07900992e-01 -6.99831843e-02
9.27366540e-02 -4.39899623e-01 7.51220167e-01 -9.33633029e-01
-1.28747380e+00 4.38934445e-01 1.68638051e+00 -1.90680087e-01
1.23747897e+00 -2.33710557e-01 4.15513337e-01 8.23575974e-01
-1.18281826e-01 -1.43387270e+00 7.59015530e-02 1.48561910e-01
8.92075300e-01 -1.58294320e+00 2.68770576e-01 -9.31909829e-02
-3.39839816e-01 1.50781298e+00 3.67510170e-01 -2.50461280e-01
1.53150785e+00 -2.67988374e-03 5.29860437e-01 -5.89322925e-01
-3.66993725e-01 -3.38414758e-02 4.12070423e-01 5.45871198e-01
8.21026087e-01 -1.68567300e-01 -9.59968567e-01 3.25556993e-01
-4.68167178e-02 4.38680828e-01 3.99867296e-01 1.37381089e+00
-4.49246526e-01 -7.46132135e-01 -5.51153243e-01 8.07943046e-01
-1.29905093e+00 -1.07997790e-01 -2.65589982e-01 1.10407412e+00
1.23225227e-02 1.46218598e+00 1.74566463e-01 -2.94652700e-01
2.45867986e-02 -1.16246559e-01 -6.24651974e-03 -3.05964112e-01
-7.19631076e-01 2.04180807e-01 -1.39405817e-01 -1.69381067e-01
-9.76362407e-01 -2.65793800e-01 -1.17451167e+00 -7.72245899e-02
-6.36834383e-01 3.24779391e-01 1.18070841e+00 8.91687274e-01
3.10311288e-01 7.87386417e-01 2.15869576e-01 -8.25942531e-02
-6.22317148e-03 -1.25575602e+00 -4.56278503e-01 3.52227986e-01
2.66338557e-01 -5.07391453e-01 1.05528245e-02 1.64936960e-01] | [14.36076545715332, -1.5549200773239136] |
3d1805c1-eb35-491a-ba8f-70cb3bfda4e1 | on-the-robustness-of-ensemble-based-machine | 2209.14013 | null | https://arxiv.org/abs/2209.14013v2 | https://arxiv.org/pdf/2209.14013v2.pdf | On the Robustness of Random Forest Against Untargeted Data Poisoning: An Ensemble-Based Approach | Machine learning is becoming ubiquitous. From finance to medicine, machine learning models are boosting decision-making processes and even outperforming humans in some tasks. This huge progress in terms of prediction quality does not however find a counterpart in the security of such models and corresponding predictions, where perturbations of fractions of the training set (poisoning) can seriously undermine the model accuracy. Research on poisoning attacks and defenses received increasing attention in the last decade, leading to several promising solutions aiming to increase the robustness of machine learning. Among them, ensemble-based defenses, where different models are trained on portions of the training set and their predictions are then aggregated, provide strong theoretical guarantees at the price of a linear overhead. Surprisingly, ensemble-based defenses, which do not pose any restrictions on the base model, have not been applied to increase the robustness of random forest models. The work in this paper aims to fill in this gap by designing and implementing a novel hash-based ensemble approach that protects random forest against untargeted, random poisoning attacks. An extensive experimental evaluation measures the performance of our approach against a variety of attacks, as well as its sustainability in terms of resource consumption and performance, and compares it with a traditional monolithic model based on random forest. A final discussion presents our main findings and compares our approach with existing poisoning defenses targeting random forests. | ['Chan Yeob Yeun', 'Ernesto Damiani', 'Nicola Bena', 'Alessandro Balestrucci', 'Claudio A. Ardagna', 'Marco Anisetti'] | 2022-09-28 | null | null | null | null | ['data-poisoning'] | ['adversarial'] | [ 3.46324921e-01 -2.17752885e-02 -1.02000520e-01 -1.61103718e-02
-5.07304192e-01 -7.33149946e-01 7.83685207e-01 4.57827240e-01
-5.02640784e-01 9.09014523e-01 -6.44190013e-02 -7.38061130e-01
5.01515940e-02 -1.22520435e+00 -5.89935005e-01 -1.08771789e+00
-2.75829315e-01 5.39289594e-01 5.16608715e-01 -3.90686393e-01
3.61183614e-01 6.38814569e-01 -1.32353497e+00 5.03394961e-01
7.12624311e-01 8.77279520e-01 -6.69074118e-01 6.25615180e-01
3.22003424e-01 8.53647709e-01 -7.17039824e-01 -9.26211417e-01
3.53176862e-01 -1.69882983e-01 -7.50672638e-01 -3.69412541e-01
5.89990057e-02 -3.62801820e-01 -1.90295637e-01 8.03308845e-01
6.81246400e-01 -4.88077223e-01 6.10688806e-01 -1.36956084e+00
-6.84009567e-02 7.99397767e-01 -4.84624594e-01 1.09275825e-01
1.95961252e-01 2.80905455e-01 9.03788209e-01 -3.42658550e-01
1.53142393e-01 1.04992187e+00 8.15509081e-01 7.13282645e-01
-1.31689858e+00 -1.07988167e+00 1.43928155e-02 4.08852160e-01
-1.09151411e+00 -2.40301698e-01 4.13038135e-01 -1.93361610e-01
6.59175754e-01 6.17816269e-01 4.01771873e-01 1.31714058e+00
4.49613959e-01 4.88569140e-01 1.63733017e+00 -4.73145157e-01
5.48284471e-01 3.98924589e-01 3.22023600e-01 4.82378960e-01
8.03150177e-01 5.43794155e-01 -6.01031125e-01 -8.36787999e-01
1.10343069e-01 -8.01734254e-03 -3.32126379e-01 -3.79464954e-01
-7.96479523e-01 1.28709126e+00 3.41439366e-01 9.68596786e-02
-3.47735852e-01 -7.07364306e-02 5.80760241e-01 2.74867833e-01
5.79970181e-01 3.14675450e-01 -5.05623579e-01 4.23738748e-01
-9.11697268e-01 3.68476152e-01 1.06869066e+00 1.40410468e-01
3.72779012e-01 -8.30509439e-02 -2.48547900e-03 -1.42615587e-02
5.69408238e-02 6.55225396e-01 3.14439595e-01 -3.69636029e-01
1.91869482e-01 3.74644995e-01 3.07216356e-03 -9.78366733e-01
-3.38251650e-01 -6.99433804e-01 -1.10917366e+00 5.07697403e-01
5.40908396e-01 -1.65460527e-01 -5.59007585e-01 1.63672197e+00
7.05927610e-01 2.40871117e-01 4.46940586e-02 4.50509131e-01
1.92999110e-01 2.89855778e-01 2.77984053e-01 -2.42667735e-01
1.36550808e+00 -5.82628489e-01 -1.55868083e-01 1.67366251e-01
6.51770592e-01 -4.93006945e-01 5.92060864e-01 7.47996509e-01
-5.33074915e-01 -1.05733909e-01 -1.06969202e+00 8.24087203e-01
-4.23413843e-01 -7.94434071e-01 5.14576674e-01 1.52971113e+00
-6.68969870e-01 6.66476965e-01 -7.38207459e-01 -8.60690027e-02
5.39266467e-01 5.54896832e-01 -2.67771274e-01 9.84249413e-02
-1.40239561e+00 1.15621281e+00 1.48840904e-01 -3.65288705e-01
-7.88751125e-01 -4.84346837e-01 -3.30781698e-01 -8.90975498e-05
2.07718149e-01 -9.30831015e-01 1.00457966e+00 -6.49443567e-01
-9.94729877e-01 4.97215003e-01 9.69508886e-02 -1.39100516e+00
7.19215512e-01 -2.61892855e-01 -1.93863899e-01 -7.50978515e-02
-3.07082653e-01 1.40639186e-01 8.88693750e-01 -1.16035306e+00
-6.44766212e-01 -7.17116952e-01 3.52000929e-02 -3.11221838e-01
-5.66274881e-01 2.78917074e-01 8.93722177e-01 -5.46818912e-01
-2.33974561e-01 -9.27016318e-01 -8.05804491e-01 -3.90029669e-01
-6.43989027e-01 6.12969249e-02 8.82299423e-01 -3.61959666e-01
1.30021870e+00 -1.43193793e+00 -2.71052867e-01 3.88800979e-01
1.51473880e-01 4.62791920e-01 2.08893225e-01 7.51117289e-01
-2.38232668e-02 3.90330195e-01 -4.51178312e-01 3.60279568e-02
-2.03063682e-01 4.50826064e-02 -1.02206933e+00 6.14336789e-01
-1.07921816e-01 4.84609932e-01 -4.63517994e-01 -2.97777653e-01
2.97588706e-02 4.44115520e-01 -5.90793610e-01 -1.18372031e-01
-1.72131762e-01 4.53821719e-01 -3.04891199e-01 5.47320187e-01
7.18936443e-01 -1.69642702e-01 3.47731829e-01 1.30442500e-01
7.44477585e-02 3.40918005e-01 -8.65662098e-01 5.07960439e-01
-9.87021253e-02 -6.55527636e-02 -1.75054386e-01 -1.07047009e+00
7.72529960e-01 3.42971742e-01 5.02253234e-01 -2.57382303e-01
1.80250227e-01 2.21661538e-01 1.41410321e-01 1.32967070e-01
1.28778696e-01 -3.93853784e-01 -2.71844745e-01 7.04071522e-01
-4.01210308e-01 2.20542714e-01 -2.25311816e-01 1.19801939e-01
1.46961868e+00 -2.06781492e-01 7.33816922e-01 -9.19291526e-02
7.25883961e-01 1.47064507e-01 2.88666815e-01 8.74682128e-01
-1.79469749e-01 9.51446667e-02 2.67282873e-01 -7.16729164e-01
-8.12832117e-01 -8.77514482e-01 -2.25950092e-01 1.02779353e+00
-4.24835011e-02 -5.30968487e-01 -1.07193041e+00 -1.34942222e+00
1.69868693e-01 7.43913651e-01 -6.34900033e-01 -2.06809998e-01
-5.63802421e-01 -1.42255557e+00 9.62969184e-01 1.88475475e-01
6.92322254e-01 -8.48469555e-01 -9.21677411e-01 2.13850692e-01
-1.69283107e-01 -7.51593232e-01 1.39051080e-01 5.44426441e-01
-1.01374567e+00 -1.23995590e+00 -1.51062787e-01 -8.32297653e-03
2.30933219e-01 2.96023488e-01 1.07640553e+00 2.72759199e-01
-7.86247402e-02 3.05635370e-02 -3.78981829e-01 -8.79926920e-01
-7.14359522e-01 1.99407741e-01 3.71266246e-01 6.80215955e-02
5.02402127e-01 -9.53196824e-01 -4.80535001e-01 3.97974074e-01
-8.95103335e-01 -2.96541929e-01 6.53236449e-01 9.00708973e-01
1.38406366e-01 1.75912872e-01 6.45726025e-01 -1.06632233e+00
2.67557979e-01 -6.66480720e-01 -3.37108135e-01 1.75603047e-01
-8.48463953e-01 1.90873921e-01 7.44342864e-01 -4.52671766e-01
-4.27235872e-01 1.30383551e-01 -4.23717618e-01 -5.16947880e-02
-1.95287302e-01 1.22632839e-01 -1.45866811e-01 -3.24516535e-01
1.05110347e+00 4.76741672e-01 3.15685495e-04 -5.00484526e-01
3.29194278e-01 7.48227417e-01 2.58166581e-01 -5.31914711e-01
1.20288897e+00 6.44268155e-01 3.80198240e-01 -6.27364397e-01
-7.59011090e-01 -2.23295659e-01 -5.08451402e-01 -9.00568962e-02
4.42936569e-01 -5.34756362e-01 -7.72534430e-01 5.82935095e-01
-1.04156923e+00 4.33360115e-02 -1.74015865e-01 8.23932961e-02
-3.73568147e-01 6.73481941e-01 -5.48263550e-01 -1.07978785e+00
-8.54644120e-01 -7.32885301e-01 6.19658709e-01 -1.62798539e-01
-2.34508589e-01 -6.96659625e-01 2.72976369e-01 4.26617175e-01
5.14107645e-01 5.53474844e-01 1.01486206e+00 -1.26649237e+00
-4.48973030e-01 -4.32332277e-01 1.88154817e-01 3.51031393e-01
-2.72305071e-01 -2.58604437e-01 -1.14432752e+00 -4.49489087e-01
3.33131850e-01 -2.39341393e-01 1.06538367e+00 -2.59372555e-02
9.94206369e-01 -8.41008008e-01 -4.15140212e-01 2.48914883e-01
1.24689829e+00 1.80515256e-02 7.20133126e-01 6.61078751e-01
1.97013348e-01 6.92695379e-01 4.41152364e-01 5.72200596e-01
9.75636579e-03 7.14502990e-01 9.47680533e-01 6.55349866e-02
3.35445583e-01 -1.97762102e-01 5.48671722e-01 2.48348460e-01
-2.73416460e-01 -1.31373316e-01 -8.08255136e-01 -1.52725950e-02
-1.55300248e+00 -1.44503307e+00 -2.53442973e-01 2.56366539e+00
7.55346060e-01 3.64937901e-01 5.98115742e-01 7.45805621e-01
6.93210423e-01 5.77159338e-02 -4.98185545e-01 -4.76850897e-01
-7.70288333e-02 6.28526270e-01 8.29077363e-01 2.14543283e-01
-1.43111873e+00 5.29367566e-01 6.53646088e+00 1.03704631e+00
-1.15938830e+00 2.87389874e-01 7.90292740e-01 1.03912480e-01
-3.56235206e-02 5.84330447e-02 -8.86951506e-01 5.35159051e-01
1.33476186e+00 -1.12986259e-01 1.52442589e-01 1.08201838e+00
-9.86935720e-02 2.12234661e-01 -7.24161625e-01 3.20240855e-01
-1.18838727e-01 -1.26472676e+00 1.62181601e-01 6.36418939e-01
3.19645613e-01 -7.80522823e-02 2.17919767e-01 3.06147128e-01
8.04962575e-01 -1.16479039e+00 5.88726997e-01 9.46140140e-02
3.06797355e-01 -1.09281671e+00 8.09554875e-01 9.45365667e-01
-7.04215467e-01 -3.67776126e-01 -2.51895905e-01 -2.32681528e-01
-8.75397101e-02 8.08399558e-01 -8.19542408e-01 7.20847607e-01
6.74680471e-01 -1.02493696e-01 -6.36765778e-01 8.81979823e-01
-1.64404243e-01 1.14497268e+00 -4.34790820e-01 -1.67040005e-01
7.69544467e-02 2.08601817e-01 5.94380856e-01 1.18129826e+00
5.89942932e-02 -1.26091046e-02 2.24399120e-01 2.13045508e-01
1.72823340e-01 2.61699021e-01 -7.88894534e-01 4.10197139e-01
4.96495962e-01 1.00028622e+00 -5.65529883e-01 -2.82361388e-01
-1.82188153e-01 6.21983886e-01 3.35933603e-02 -3.78163606e-01
-9.98975396e-01 1.26762718e-01 6.78665876e-01 4.02358621e-01
1.91036254e-01 9.93710943e-03 -5.69186211e-01 -1.02026677e+00
-3.23754430e-01 -1.50102258e+00 6.29760325e-01 1.65653229e-01
-1.42210722e+00 8.13031793e-01 -1.75794378e-01 -1.18618476e+00
-1.87611014e-01 -4.03486371e-01 -5.78042090e-01 5.79145491e-01
-1.10306644e+00 -1.30874193e+00 1.41303122e-01 6.19762421e-01
1.40610352e-01 -2.23130032e-01 1.26206303e+00 -1.39471337e-01
-4.24424738e-01 7.39323676e-01 5.42295873e-02 -1.36810899e-01
5.71372092e-01 -1.04459035e+00 3.93050194e-01 9.51131701e-01
2.71937698e-01 5.68358362e-01 9.55961108e-01 -6.20519936e-01
-9.29699957e-01 -1.08563566e+00 8.92456234e-01 -7.28838146e-01
4.39799100e-01 -3.29916120e-01 -7.93892682e-01 2.94652045e-01
4.29306328e-02 -1.63100094e-01 9.80771661e-01 1.30314633e-01
-9.68132496e-01 -9.70893353e-02 -1.54977369e+00 4.66972768e-01
7.68529356e-01 -1.96286187e-01 -3.00823480e-01 3.91925931e-01
4.70786780e-01 1.82144910e-01 -6.64170980e-01 6.89554453e-01
8.89089704e-01 -1.50532842e+00 1.14248323e+00 -9.82714593e-01
2.05089107e-01 -9.91627201e-02 -4.25518483e-01 -9.80198562e-01
-2.79814541e-01 -5.30438602e-01 -3.93378735e-01 9.96816099e-01
2.82562971e-01 -9.33033645e-01 1.11049163e+00 2.65802056e-01
5.31659007e-01 -9.74867404e-01 -1.01574099e+00 -8.55873644e-01
5.44433594e-01 -2.95519322e-01 7.79444218e-01 7.44721353e-01
-8.12590197e-02 3.15061331e-01 -7.53709912e-01 3.37273508e-01
1.09904993e+00 1.32111073e-01 1.04443491e+00 -1.41987157e+00
-6.49274826e-01 -2.36227036e-01 -6.46199942e-01 -3.78980398e-01
-1.36935174e-01 -6.94549143e-01 -5.04796565e-01 -8.24647903e-01
3.52704436e-01 -3.97838473e-01 -4.49245334e-01 6.33744657e-01
-2.36699298e-01 4.79768038e-01 3.88808995e-01 4.70757365e-01
-1.91019371e-01 1.19711481e-01 5.05699635e-01 -1.78295851e-01
1.76555336e-01 6.84405804e-01 -1.01111329e+00 7.25641727e-01
1.15910792e+00 -9.34736907e-01 -8.76741484e-02 4.75827038e-01
-1.00999609e-01 -6.41868263e-02 5.07898867e-01 -1.27462518e+00
1.06631577e-01 -1.05838455e-01 3.82092863e-01 -4.46644038e-01
8.81239548e-02 -8.54907513e-01 3.63617510e-01 1.38122654e+00
-2.45049536e-01 4.59583923e-02 -2.41493583e-01 8.07041109e-01
2.25751236e-01 -7.79418126e-02 1.09285712e+00 -5.30586541e-02
5.43569624e-02 2.22139731e-01 -4.97790515e-01 -4.23895448e-01
1.24413693e+00 -7.79495090e-02 -5.61373293e-01 -4.11929458e-01
-5.98319530e-01 -3.13184172e-01 5.37212491e-01 -1.12184703e-01
2.30573386e-01 -8.51263404e-01 -1.02334130e+00 9.98843312e-02
-2.17602611e-01 -5.77409744e-01 -3.39112245e-02 8.23789537e-01
-2.31032789e-01 4.32169139e-01 -2.13250026e-01 -3.27943295e-01
-1.81011593e+00 1.01013446e+00 2.91201770e-01 -1.06619108e+00
-3.61712635e-01 5.37178636e-01 7.05910623e-02 -2.67048210e-01
2.02775821e-01 3.89049411e-01 -7.96177387e-02 -6.07483722e-02
7.98801661e-01 4.40070331e-01 2.02806443e-01 -4.56339598e-01
-5.58973193e-01 9.48624238e-02 -2.62025625e-01 2.23820866e-03
1.12272537e+00 3.37439656e-01 -5.65726161e-02 -1.52615383e-01
5.86903751e-01 3.63317966e-01 -5.89320481e-01 -1.67856768e-01
2.73990184e-01 -3.23596209e-01 -2.74304569e-01 -1.02022231e+00
-8.32402170e-01 9.29782629e-01 5.88470817e-01 5.74179471e-01
1.15341854e+00 -4.15692151e-01 8.60069513e-01 3.90980154e-01
9.85960484e-01 -2.50077724e-01 -1.20086811e-01 3.20928246e-01
4.86214131e-01 -8.54064941e-01 1.80009335e-01 -4.89017725e-01
-5.43972969e-01 1.00509572e+00 2.85559148e-01 -2.91392118e-01
6.66654229e-01 5.73302209e-01 -1.31454006e-01 1.81274325e-01
-1.02684128e+00 -1.29413292e-01 -3.54300886e-01 9.55649555e-01
8.44330490e-02 1.48594007e-01 -5.25405645e-01 6.87198997e-01
-4.28495526e-01 -2.02339202e-01 2.50736922e-01 8.80109429e-01
-8.42403710e-01 -1.64365661e+00 -8.74587119e-01 4.27348524e-01
-9.52545643e-01 -1.13488980e-01 -6.47251368e-01 6.47146106e-01
1.94135845e-01 1.21941018e+00 -5.04977703e-01 -9.70930457e-01
1.52973533e-01 2.94130951e-01 2.54518896e-01 -2.18475044e-01
-1.21718085e+00 -4.41711813e-01 2.62311310e-01 -3.91528428e-01
-2.08768323e-01 -6.75241351e-01 -6.29833817e-01 -1.00939643e+00
-6.30222559e-01 5.19309461e-01 6.38354540e-01 6.73280716e-01
1.60856366e-01 1.34161813e-03 1.06993639e+00 -5.84576130e-01
-1.47893429e+00 -7.80017674e-01 -4.72510338e-01 8.64046440e-03
4.50781882e-02 -5.28851807e-01 -5.48711002e-01 -2.43009284e-01] | [5.813967227935791, 7.503772735595703] |
7002fa82-e8af-4cdb-b7d6-fafea64d0aed | answering-unanswered-questions-through | 2305.17393 | null | https://arxiv.org/abs/2305.17393v2 | https://arxiv.org/pdf/2305.17393v2.pdf | Answering Unanswered Questions through Semantic Reformulations in Spoken QA | Spoken Question Answering (QA) is a key feature of voice assistants, usually backed by multiple QA systems. Users ask questions via spontaneous speech which can contain disfluencies, errors, and informal syntax or phrasing. This is a major challenge in QA, causing unanswered questions or irrelevant answers, and leading to bad user experiences. We analyze failed QA requests to identify core challenges: lexical gaps, proposition types, complex syntactic structure, and high specificity. We propose a Semantic Question Reformulation (SURF) model offering three linguistically-grounded operations (repair, syntactic reshaping, generalization) to rewrite questions to facilitate answering. Offline evaluation on 1M unanswered questions from a leading voice assistant shows that SURF significantly improves answer rates: up to 24% of previously unanswered questions obtain relevant answers (75%). Live deployment shows positive impact for millions of customers with unanswered questions; explicit relevance feedback shows high user satisfaction. | ['Shervin Malmasi', 'Oleg Rokhlenko', 'Besnik Fetahu', 'Zhiyu Chen', 'Pedro Faustini'] | 2023-05-27 | null | null | null | null | ['specificity'] | ['natural-language-processing'] | [ 5.93837388e-02 7.29239285e-01 4.13483024e-01 -5.31723976e-01
-1.54436743e+00 -1.07578707e+00 1.81301907e-01 -5.07159419e-02
-2.03763664e-01 8.36038589e-01 9.87888336e-01 -7.72863328e-01
-1.71029434e-01 -4.60209221e-01 -2.04953477e-01 3.19715977e-01
6.33896530e-01 8.00535679e-01 7.58737743e-01 -1.23338807e+00
2.71021158e-01 4.58574593e-02 -1.30515003e+00 8.40088904e-01
1.26674783e+00 8.99281204e-01 4.27619368e-01 1.00726354e+00
-1.02730000e+00 1.36986017e+00 -1.14621735e+00 -6.08925879e-01
-4.03532296e-01 -4.09981281e-01 -1.90017879e+00 -2.24847794e-01
3.80013853e-01 -3.16978008e-01 2.45099664e-01 8.27177584e-01
4.24843401e-01 9.08176005e-02 -1.31197140e-01 -1.09369648e+00
-7.87706137e-01 5.66946805e-01 5.32535076e-01 3.35433602e-01
1.39083922e+00 2.86389858e-01 1.20338213e+00 -8.12781334e-01
3.94903213e-01 1.43129301e+00 6.97242498e-01 8.56460392e-01
-8.27505767e-01 1.06988259e-01 -1.39131978e-01 2.80347437e-01
-7.46955872e-01 -8.26562881e-01 2.05215767e-01 -7.74727985e-02
1.52742767e+00 9.80994463e-01 -9.71906185e-02 7.50320613e-01
-1.61466692e-02 3.92641455e-01 5.56128681e-01 -4.33857888e-01
3.60280395e-01 4.10433769e-01 6.25316203e-01 4.65993315e-01
-4.72191066e-01 -7.88194478e-01 -5.29034555e-01 -3.50157022e-01
-6.68615326e-02 -4.43219900e-01 -5.24074316e-01 6.01584435e-01
-6.44641101e-01 7.86503494e-01 -7.52068236e-02 2.32726976e-01
-4.84979749e-01 -3.99434417e-01 2.84319550e-01 6.75598681e-01
-3.13185722e-01 9.32578087e-01 -7.11284816e-01 -8.57878864e-01
-2.14571878e-01 5.36702275e-01 1.49363196e+00 1.03661585e+00
7.29845941e-01 -2.05747530e-01 -5.60110152e-01 1.21415269e+00
4.02632132e-02 7.69260406e-01 6.41247869e-01 -1.64301264e+00
5.19743145e-01 9.51586485e-01 7.45160758e-01 -7.98665226e-01
-4.46709961e-01 -1.98990181e-02 -1.68020409e-02 -6.94988072e-01
4.96557951e-01 -2.52103359e-01 -4.51727957e-01 1.45495093e+00
7.67392069e-02 -7.49685109e-01 2.19803169e-01 8.52765858e-01
1.30332088e+00 7.90544748e-01 -5.38942516e-02 -2.09485471e-01
1.76075923e+00 -1.15692961e+00 -1.28183460e+00 -5.13321459e-01
4.76942092e-01 -1.02286923e+00 2.02998686e+00 3.42434645e-01
-1.36855125e+00 -4.00202185e-01 -3.34753156e-01 -4.21967268e-01
-1.69688508e-01 -2.48411298e-01 1.50477335e-01 8.43232095e-01
-1.22217453e+00 5.14272973e-03 1.33808458e-03 -5.62963009e-01
-5.21080680e-02 2.60685682e-01 1.02695012e-02 -2.82202750e-01
-1.31858742e+00 8.60696733e-01 -4.50228900e-01 -9.50895920e-02
-1.24958903e-01 -7.41852105e-01 -6.65172100e-01 2.74401665e-01
7.31357396e-01 -6.92101657e-01 2.19410753e+00 -3.66793066e-01
-1.70578182e+00 3.94209534e-01 -7.22275496e-01 -3.55480671e-01
-9.12808999e-02 -5.73776722e-01 -7.90466130e-01 4.06668276e-01
4.46279138e-01 3.26746851e-01 5.46058238e-01 -8.97963524e-01
-6.79009974e-01 -2.78887123e-01 3.75844449e-01 3.02517653e-01
-1.70379460e-01 4.77473736e-01 -2.34676898e-02 4.38009910e-02
1.07137434e-01 -1.87771961e-01 1.55301020e-01 -6.99349225e-01
-1.41950652e-01 -6.38521671e-01 6.90373957e-01 -1.23746204e+00
1.83388889e+00 -1.78452051e+00 -2.68686682e-01 -2.83887565e-01
1.00657314e-01 3.98862928e-01 -3.27500492e-01 8.04983795e-01
4.02240306e-01 3.69842380e-01 -1.33768320e-01 7.72158056e-02
3.67190003e-01 5.76452434e-01 -9.88353670e-01 -7.35050380e-01
4.72211838e-01 1.31405115e+00 -8.18490565e-01 -1.83414936e-01
-6.12970032e-02 -1.41262069e-01 -6.58384979e-01 5.78859568e-01
-6.39663517e-01 4.94565181e-02 -2.60306597e-01 9.58497107e-01
3.17604750e-01 -2.49300048e-01 -8.22246373e-02 4.22396958e-02
7.11599514e-02 1.18767297e+00 -7.51007617e-01 1.29881728e+00
-5.01078188e-01 6.85126409e-02 4.32300568e-01 -2.86219925e-01
8.38580847e-01 2.50211149e-01 -1.47608265e-01 -1.13884509e+00
-2.35604197e-02 4.36527610e-01 -5.12525260e-01 -1.11704946e+00
8.28541934e-01 -1.21824570e-01 -3.37811440e-01 2.33410716e-01
5.00896163e-02 -5.80008566e-01 -9.99784246e-02 3.99314314e-01
1.63211286e+00 -4.76575345e-01 1.02011621e-01 -6.01794720e-02
8.05263698e-01 3.95446420e-01 1.45681217e-01 1.05956948e+00
-5.84662855e-01 6.21938229e-01 5.83390057e-01 -9.73141864e-02
-4.18755531e-01 -1.14923191e+00 3.89643788e-01 1.45397508e+00
-1.21781625e-01 -5.74563622e-01 -1.09186232e+00 -6.22864723e-01
-2.98550874e-01 1.22705960e+00 2.22369865e-01 -2.57244915e-01
-7.72160709e-01 2.79077291e-01 7.36068249e-01 2.93150187e-01
4.11506891e-01 -1.43757045e+00 -5.41802645e-01 4.66754049e-01
-9.85996664e-01 -1.29932380e+00 -7.25015342e-01 -1.09509677e-01
-5.33229589e-01 -7.95767009e-01 -4.36265707e-01 -7.78576434e-01
-1.10639535e-01 3.03679794e-01 1.34185886e+00 7.43709728e-02
1.43795744e-01 9.35202599e-01 -6.98877335e-01 -2.38385051e-02
-5.86033225e-01 4.13280018e-02 -1.60722181e-01 -3.76176566e-01
4.88784462e-01 -2.96880692e-01 -4.25069332e-01 6.44245267e-01
-7.72681653e-01 -7.11633623e-01 5.98432608e-02 7.45120585e-01
-8.02654400e-02 -6.03932440e-01 1.36539280e+00 -7.57623672e-01
1.50768602e+00 -3.02367657e-01 -1.23165607e-01 6.32453978e-01
-4.68331963e-01 -6.93854168e-02 7.14812100e-01 7.92235062e-02
-1.28146994e+00 -4.54143047e-01 -7.92144597e-01 4.38757390e-01
-4.32971030e-01 2.93351561e-01 -2.51489460e-01 1.72817245e-01
1.00863373e+00 1.07465424e-02 3.74213368e-01 -4.71562892e-01
5.22932947e-01 1.04763591e+00 8.10982406e-01 -4.49937820e-01
1.92545816e-01 -1.77142303e-02 -9.22299623e-01 -9.63823259e-01
-8.22819710e-01 -8.56113732e-01 1.45458356e-01 -1.57719895e-01
7.07570851e-01 -3.89416844e-01 -1.20423555e+00 -1.42793044e-01
-1.39655638e+00 -1.93594158e-01 -6.90200388e-01 -2.76158780e-01
-4.11502361e-01 7.58023798e-01 -6.77353382e-01 -1.06546605e+00
-5.54483414e-01 -9.08532917e-01 1.05677044e+00 4.38338310e-01
-1.07146549e+00 -5.73789835e-01 -1.87116578e-01 1.40749168e+00
9.11921561e-01 -8.37118626e-01 1.37026346e+00 -9.49153960e-01
-2.35642180e-01 1.08288899e-02 -1.61681101e-01 3.52664232e-01
3.14664304e-01 -6.19415820e-01 -8.55852664e-01 2.85008758e-01
3.16099554e-01 -5.63100219e-01 2.20275357e-01 2.47837473e-02
5.73982179e-01 -8.86695683e-01 4.04107779e-01 -4.63842005e-01
7.26657808e-01 3.58315915e-01 7.62199402e-01 4.47925813e-02
2.91223675e-02 1.04578829e+00 5.15421808e-01 4.96069670e-01
7.31633008e-01 4.61388588e-01 3.64957243e-01 7.23379612e-01
-7.91800320e-02 -3.36543798e-01 5.15730858e-01 8.22493494e-01
8.01876962e-01 -4.79114294e-01 -1.00687540e+00 5.30066431e-01
-1.64123118e+00 -7.07987309e-01 -3.99092674e-01 1.93098068e+00
9.28741157e-01 -1.64048970e-01 8.62775669e-02 -5.16122133e-02
3.60709518e-01 -1.85529411e-01 -3.51101547e-01 -8.78953636e-01
-1.16979249e-01 8.31147015e-01 -1.68516085e-01 1.19649863e+00
-2.49490738e-01 1.17762339e+00 6.19657660e+00 5.12679160e-01
-7.15456784e-01 1.67035043e-01 8.34914073e-02 2.13807583e-01
-9.54988003e-01 -7.41595775e-02 -7.84754455e-01 1.59134701e-01
1.09177125e+00 -2.16269210e-01 7.28824198e-01 7.35289395e-01
3.12908232e-01 -3.44696015e-01 -6.35046661e-01 8.16729724e-01
1.07178457e-01 -1.48687398e+00 1.97565988e-01 -6.63944662e-01
4.87622060e-02 -2.34282687e-01 -1.14452802e-01 7.77184725e-01
2.66823709e-01 -1.00524318e+00 5.01221359e-01 6.50614202e-01
4.55155045e-01 -5.24347007e-01 8.80751133e-01 5.24103165e-01
-5.56683719e-01 -4.45342779e-01 -1.36157379e-01 -3.17177892e-01
6.92683935e-01 2.06258729e-01 -1.09251130e+00 2.81139046e-01
9.22708452e-01 -3.47597450e-01 -5.78119099e-01 4.67930466e-01
-1.25246778e-01 8.44614983e-01 -4.62343544e-01 -4.54684973e-01
5.88581376e-02 3.85420322e-02 6.58657253e-01 7.54818499e-01
2.05288246e-01 5.25655091e-01 -2.04446539e-01 7.16982365e-01
1.55002428e-02 1.74458250e-01 -2.15510830e-01 -1.88111290e-01
6.26941502e-01 9.68223512e-01 -1.78807601e-01 -2.09649786e-01
-3.01348716e-01 1.33903205e+00 -5.57329366e-03 5.01783907e-01
-4.07102942e-01 -5.63981831e-01 5.25234699e-01 7.51462653e-02
3.65952402e-02 4.09036279e-02 -3.04528028e-01 -8.46265018e-01
5.38646877e-01 -1.36653435e+00 4.49922055e-01 -1.22920692e+00
-1.15331328e+00 7.98382044e-01 -6.30226910e-01 -2.96208233e-01
-6.57763422e-01 -3.34235430e-01 -3.73050928e-01 1.09980166e+00
-1.30548704e+00 -5.38421333e-01 -4.05865788e-01 4.47389007e-01
9.37602758e-01 -1.37475750e-03 1.18319511e+00 5.21804869e-01
-6.79087564e-02 5.60081184e-01 -5.90209484e-01 -4.27251011e-01
6.71703994e-01 -1.11607635e+00 3.07273537e-01 3.67396116e-01
-2.93929994e-01 8.49277794e-01 7.00013459e-01 -5.56977332e-01
-1.54665279e+00 -6.98252738e-01 1.91393316e+00 -1.04893339e+00
5.30843854e-01 -6.41446412e-02 -1.24835670e+00 3.57591510e-01
3.84510159e-01 -6.47785068e-01 8.07306290e-01 -6.83035702e-02
-1.49039298e-01 -1.53484881e-01 -1.33638608e+00 6.47786915e-01
8.08251560e-01 -9.34674025e-01 -1.27513301e+00 3.42200667e-01
1.49276686e+00 -1.87744021e-01 -2.62089759e-01 2.26697192e-01
2.60156810e-01 -1.11928856e+00 6.48845732e-01 -9.84947383e-01
1.90532133e-01 -1.39527738e-01 -5.73063672e-01 -7.61474371e-01
6.94255009e-02 -1.29290378e+00 -7.47292265e-02 1.41261899e+00
7.76549280e-01 -7.93428719e-01 5.43991387e-01 1.41892564e+00
-5.73399007e-01 -5.42053342e-01 -1.01702344e+00 -6.33948624e-01
-1.85514569e-01 -6.06198490e-01 7.36106873e-01 4.17834729e-01
6.43620133e-01 1.00642872e+00 1.46843031e-01 8.40662643e-02
-2.76743948e-01 -2.58154303e-01 5.07782161e-01 -8.33220303e-01
-1.40702888e-01 -3.72536391e-01 2.60125130e-01 -1.73772204e+00
-7.74270073e-02 -4.93327945e-01 2.45627478e-01 -1.88196182e+00
-5.42713821e-01 -3.82434018e-02 6.90639913e-01 4.18662161e-01
5.13760448e-02 -3.22822839e-01 -2.75700744e-02 -1.46441096e-02
-7.47449219e-01 4.77480024e-01 8.95141780e-01 1.07289739e-01
-4.52616662e-01 2.45364189e-01 -1.18997729e+00 5.54080427e-01
6.37385547e-01 6.47281557e-02 -4.01246458e-01 -5.71610987e-01
6.83275878e-01 6.36425018e-01 2.88516194e-01 -6.12291336e-01
3.75494629e-01 -1.15833163e-01 -7.19449341e-01 -5.56693137e-01
2.50883073e-01 -7.34010220e-01 -4.62768555e-01 2.59135097e-01
-5.39723217e-01 2.42413163e-01 2.26557150e-01 9.54235420e-02
-4.12454933e-01 -6.92941666e-01 1.57365456e-01 -2.58892745e-01
-5.62359869e-01 -4.94696766e-01 -8.70479643e-01 6.30906761e-01
1.64385736e-01 -1.58011973e-01 -4.88065273e-01 -1.14593244e+00
-8.51111650e-01 6.86118782e-01 -3.04822952e-01 5.94130993e-01
8.88047457e-01 -8.10538292e-01 -2.56178468e-01 1.19129360e-01
1.54752329e-01 -1.18345238e-01 4.09210056e-01 5.08653462e-01
-4.47643429e-01 9.08734322e-01 3.56686354e-01 -3.36947680e-01
-1.12884748e+00 4.24411185e-02 4.49598998e-01 5.70370667e-02
-2.11331606e-01 1.15033996e+00 -4.20573115e-01 -9.65905368e-01
4.51573461e-01 -7.84562290e-01 -2.99122185e-01 8.63788724e-02
6.95547044e-01 5.81162453e-01 5.29201567e-01 -1.19427383e-01
-3.93893868e-01 5.75205088e-02 1.95144303e-02 -5.01366675e-01
6.10585690e-01 -7.13303864e-01 -2.79792994e-01 2.59684682e-01
8.52527320e-01 3.51127595e-01 -4.41857666e-01 -2.83884943e-01
3.35630506e-01 -4.57861125e-02 -5.04013121e-01 -1.42401659e+00
7.11776093e-02 8.41688931e-01 2.38533452e-01 7.59753168e-01
8.30098569e-01 5.01612484e-01 1.58498192e+00 1.09641051e+00
3.91937524e-01 -1.17441189e+00 3.98727804e-01 1.13399994e+00
1.32031834e+00 -1.08856630e+00 -1.02654243e+00 -4.36576933e-01
-7.87281334e-01 1.03400815e+00 5.94839394e-01 6.22468293e-01
2.59241939e-01 -2.24225130e-02 5.60423732e-01 -2.56346703e-01
-8.00877094e-01 -2.78836161e-01 -1.04129188e-01 5.65612257e-01
1.68617204e-01 -1.06309801e-01 -2.13065907e-01 1.37976038e+00
-7.60857105e-01 -8.19618851e-02 4.84595418e-01 8.61362159e-01
-9.44501221e-01 -8.88674438e-01 -3.34066987e-01 4.86766011e-01
-3.66444021e-01 -3.44137907e-01 -7.64725387e-01 7.75845572e-02
-5.48932314e-01 1.96455514e+00 -1.32407889e-01 -3.45599562e-01
9.69213307e-01 7.05764771e-01 1.20245673e-01 -9.91782010e-01
-1.11090696e+00 -3.14360857e-01 7.06225216e-01 -7.17579663e-01
2.13632822e-01 -4.29210663e-01 -1.76871586e+00 -1.31838009e-01
-3.30459356e-01 6.58706248e-01 2.64639467e-01 1.05726576e+00
8.76031697e-01 2.20289603e-01 5.03557861e-01 4.84607965e-01
-1.14838481e+00 -9.43942249e-01 7.19966590e-02 2.10846409e-01
6.15501583e-01 1.07905179e-01 -5.86118758e-01 7.17139319e-02] | [11.76576042175293, 7.997804641723633] |
5b669513-95d2-46f4-90c2-e76e94550a6b | summarize-then-answer-generating-concise | 2109.06853 | null | https://arxiv.org/abs/2109.06853v1 | https://arxiv.org/pdf/2109.06853v1.pdf | Summarize-then-Answer: Generating Concise Explanations for Multi-hop Reading Comprehension | How can we generate concise explanations for multi-hop Reading Comprehension (RC)? The current strategies of identifying supporting sentences can be seen as an extractive question-focused summarization of the input text. However, these extractive explanations are not necessarily concise i.e. not minimally sufficient for answering a question. Instead, we advocate for an abstractive approach, where we propose to generate a question-focused, abstractive summary of input paragraphs and then feed it to an RC system. Given a limited amount of human-annotated abstractive explanations, we train the abstractive explainer in a semi-supervised manner, where we start from the supervised model and then train it further through trial and error maximizing a conciseness-promoted reward function. Our experiments demonstrate that the proposed abstractive explainer can generate more compact explanations than an extractive explainer with limited supervision (only 2k instances) while maintaining sufficiency. | ['Kentaro Inui', 'Niranjan Balasubramanian', 'Steven Sinha', 'Harsh Trivedi', 'Naoya Inoue'] | 2021-09-14 | null | https://aclanthology.org/2021.emnlp-main.490 | https://aclanthology.org/2021.emnlp-main.490.pdf | emnlp-2021-11 | ['multi-hop-reading-comprehension'] | ['natural-language-processing'] | [ 7.52935946e-01 1.25452709e+00 -2.28454217e-01 -4.82604235e-01
-1.16828048e+00 -5.51248133e-01 6.11865878e-01 4.77110326e-01
-2.45291263e-01 1.06581450e+00 6.23050749e-01 -6.98667765e-01
-2.85408437e-01 -4.87080425e-01 -8.34312141e-01 -6.46678880e-02
2.44541258e-01 5.57622492e-01 -4.95229661e-02 -8.10957700e-02
6.25530422e-01 3.74728888e-02 -1.43413007e+00 4.60340768e-01
1.66673219e+00 4.61488664e-01 3.88415039e-01 1.22747707e+00
-3.84775490e-01 9.94251966e-01 -8.84720266e-01 -5.15599668e-01
-2.71401078e-01 -1.09542477e+00 -1.62388062e+00 3.05326670e-01
4.31886256e-01 -4.27201182e-01 5.98555394e-02 6.24159575e-01
-4.57164496e-02 2.06917122e-01 7.82976031e-01 -9.73413587e-01
-7.07620144e-01 1.14374685e+00 -2.28819937e-01 2.87287682e-01
7.44063616e-01 9.53674912e-02 1.39057314e+00 -7.84076989e-01
4.04240280e-01 1.08010769e+00 8.78550336e-02 8.34429264e-01
-1.11638188e+00 -2.38931682e-02 5.07380545e-01 2.77500041e-02
-4.94343311e-01 -6.41177595e-01 5.96775055e-01 1.42034307e-01
1.19204140e+00 8.34207833e-01 6.22254431e-01 8.36236060e-01
1.15756802e-01 1.07565928e+00 7.41077006e-01 -5.64838886e-01
3.62488031e-01 1.74764663e-01 5.52943945e-01 8.87361586e-01
4.77416635e-01 -4.45066065e-01 -6.50026441e-01 -5.31527735e-02
1.24030590e-01 -1.33710310e-01 -5.29702187e-01 1.95305794e-02
-1.19814193e+00 1.12064099e+00 3.96338373e-01 -1.77689753e-02
-6.29780412e-01 1.33396775e-01 2.39151806e-01 4.60109144e-01
4.67608571e-01 1.11403573e+00 -4.20166910e-01 -1.06406465e-01
-1.24215925e+00 4.97485608e-01 1.11724555e+00 1.02666593e+00
5.64272225e-01 -7.18451217e-02 -3.10556531e-01 5.14126778e-01
7.58089349e-02 3.80590230e-01 3.23658317e-01 -1.21137214e+00
8.09032559e-01 6.44310892e-01 9.10965353e-02 -5.55265486e-01
-2.57578135e-01 -3.64833772e-01 -6.64851189e-01 -2.19198003e-01
1.18863165e-01 -4.11426365e-01 -7.00811446e-01 1.41678023e+00
6.67204484e-02 -5.29943883e-01 4.86271381e-01 8.65140021e-01
8.20780516e-01 9.06331658e-01 -5.29895425e-02 -4.82952088e-01
1.25919926e+00 -1.58780885e+00 -6.81675494e-01 -6.83883429e-01
5.94155073e-01 -4.27993596e-01 1.18021095e+00 4.55441236e-01
-1.56360936e+00 -2.10027665e-01 -1.13090014e+00 -3.05825114e-01
1.49561495e-01 2.54546911e-01 4.38044906e-01 8.33455548e-02
-8.90036106e-01 5.24421632e-01 -5.14997303e-01 -3.35874230e-01
3.27429444e-01 2.05771118e-01 -3.31238449e-01 -2.38261789e-01
-8.46704841e-01 9.90772307e-01 4.35203016e-01 -2.80087050e-02
-5.46033978e-01 -5.24780512e-01 -9.56213713e-01 4.13973510e-01
8.61908853e-01 -1.24032259e+00 1.66220343e+00 -8.87616634e-01
-1.43431199e+00 2.66592205e-01 -4.90988851e-01 -7.24424124e-01
1.74751326e-01 -5.51904619e-01 1.69505700e-01 7.66901731e-01
1.89084724e-01 7.77937055e-01 7.36971140e-01 -1.45234025e+00
-6.11953676e-01 -3.16592976e-02 5.47365785e-01 2.88678646e-01
-4.84489836e-02 -2.27802545e-01 3.01109888e-02 -3.79409224e-01
2.61574239e-01 -7.00442612e-01 -4.02637213e-01 -3.12082916e-01
-1.03957009e+00 -2.25089401e-01 4.09203827e-01 -9.66245651e-01
1.22865713e+00 -1.42660141e+00 3.94721746e-01 -9.00454447e-02
5.11656940e-01 1.23150468e-01 -2.78833777e-01 5.87327778e-01
8.66019577e-02 4.80738610e-01 -5.62849522e-01 -4.33383018e-01
-1.51444925e-02 1.94559947e-01 -6.00397706e-01 -2.12821171e-01
5.58493018e-01 1.04509974e+00 -1.18250859e+00 -5.68462968e-01
-1.36122689e-01 -1.92741781e-01 -9.98115659e-01 6.00996614e-01
-7.49949992e-01 6.98967203e-02 -8.26899767e-01 2.10673973e-01
3.83890510e-01 -4.29368973e-01 -8.20136070e-02 1.18558429e-01
1.17968200e-02 7.75854588e-01 -6.69926703e-01 1.51357722e+00
-6.69585168e-01 7.47280777e-01 -2.39953533e-01 -1.13166583e+00
7.79362917e-01 7.12483004e-02 -2.17304319e-01 -2.47047678e-01
-2.12252021e-01 3.06587845e-01 -9.36013013e-02 -6.70064449e-01
8.98748696e-01 -1.58629775e-01 -1.20854869e-01 9.86113846e-01
-2.18625665e-02 -3.12881529e-01 3.07740122e-01 7.97846615e-01
1.31604457e+00 8.22875425e-02 5.65987766e-01 1.63787976e-02
5.97971797e-01 3.60815257e-01 2.26356965e-02 1.09602332e+00
3.22468221e-01 8.79531860e-01 8.25993359e-01 -1.84004441e-01
-1.03253639e+00 -9.30285752e-01 5.83847523e-01 7.51784563e-01
-1.43558551e-02 -7.27443159e-01 -1.06791890e+00 -1.14565611e+00
-2.48407319e-01 1.39987528e+00 -4.04434323e-01 -2.80140758e-01
-7.07538366e-01 3.27431411e-02 3.44321847e-01 4.00278628e-01
4.18077081e-01 -1.24304104e+00 -1.27507460e+00 1.14308365e-01
-5.29272735e-01 -8.13070655e-01 -6.91685140e-01 4.99307543e-01
-1.13163483e+00 -9.17445362e-01 -4.57831144e-01 -5.34294546e-01
1.02451432e+00 4.17327225e-01 1.28463233e+00 6.71302080e-01
2.37368986e-01 3.77728015e-01 -5.88707566e-01 -4.23664898e-01
-7.16373265e-01 5.34087479e-01 -3.22195172e-01 -4.68580991e-01
-2.11284142e-02 -3.93012047e-01 -6.50760055e-01 -3.89937997e-01
-9.37760234e-01 6.43490434e-01 9.49631453e-01 8.69385004e-01
3.17755967e-01 -4.84577358e-01 1.08662355e+00 -1.17291164e+00
1.11964703e+00 -3.70850146e-01 -5.81125617e-02 6.06007159e-01
-6.71192169e-01 7.54217684e-01 1.12452495e+00 -1.83992907e-01
-1.16651034e+00 -1.51772231e-01 -6.30292147e-02 -1.81538407e-02
-7.55481422e-02 6.12218261e-01 -3.48626077e-02 6.37251496e-01
6.70507014e-01 5.05496562e-01 1.31190240e-01 -2.16673404e-01
7.02808440e-01 7.13149428e-01 6.36793494e-01 -6.05988264e-01
9.48658049e-01 -2.15523675e-01 -3.60172331e-01 -6.33962750e-01
-1.24224865e+00 -9.96816903e-02 -4.32618260e-01 -1.22194188e-02
8.12454641e-01 -3.27471524e-01 -5.37599325e-01 -5.63673377e-01
-1.60218787e+00 -7.32703879e-02 -5.77642500e-01 1.48031890e-01
-7.97134578e-01 3.51782978e-01 -2.89060533e-01 -9.46618795e-01
-7.99237370e-01 -8.08723748e-01 1.19528103e+00 4.32250887e-01
-8.49421144e-01 -7.55719006e-01 -2.82138437e-01 7.69881606e-01
3.14608753e-01 1.02717690e-01 1.29107189e+00 -1.18031168e+00
-7.19204962e-01 -1.11414358e-01 -7.49976262e-02 1.94710672e-01
1.42460525e-01 -2.30253577e-01 -7.67797589e-01 1.46163031e-01
8.41344520e-02 -6.45704150e-01 8.19676399e-01 1.54383153e-01
1.24868345e+00 -1.24632776e+00 -1.65464893e-01 7.81224594e-02
1.02509892e+00 -2.41248816e-01 2.62659550e-01 2.90167660e-01
3.64050388e-01 8.75827610e-01 7.13899612e-01 3.08786750e-01
6.03154957e-01 9.26495120e-02 3.69801611e-01 9.61556286e-02
1.31639346e-01 -6.69390976e-01 9.78088379e-02 9.04886127e-01
7.97600746e-02 -4.37793881e-01 -5.73213875e-01 6.05730712e-01
-1.87089097e+00 -1.09865189e+00 -1.51562288e-01 1.69934690e+00
9.72210705e-01 1.37606502e-01 -9.75815058e-02 2.01745734e-01
2.93640256e-01 2.16297537e-01 -6.79290116e-01 -1.11855602e+00
1.79238752e-01 1.38182506e-01 -4.50134799e-02 7.03515053e-01
-3.40070486e-01 8.68084192e-01 6.04037619e+00 3.26866090e-01
-6.96989655e-01 -3.01551700e-01 6.50770307e-01 -2.69422024e-01
-1.02039623e+00 3.80484700e-01 -3.87432367e-01 5.23004420e-02
1.17835224e+00 -5.74002922e-01 4.44954038e-01 7.34452963e-01
4.85930502e-01 -3.96761000e-01 -1.43350637e+00 3.10658932e-01
3.22740048e-01 -1.66498566e+00 6.07475400e-01 -1.74699232e-01
6.02358401e-01 -7.94461966e-01 -2.54986882e-01 3.46633464e-01
3.45344171e-02 -1.14890313e+00 7.80538440e-01 6.34554625e-01
2.86367774e-01 -7.92673469e-01 6.66529536e-01 1.05901027e+00
-5.66128314e-01 -1.41286194e-01 -4.69124347e-01 -2.45925650e-01
3.35967064e-01 4.02827233e-01 -8.65636051e-01 8.84524286e-01
1.18498340e-01 3.51130635e-01 -6.61099672e-01 7.52700329e-01
-6.04773462e-01 6.75754428e-01 9.76834074e-02 -4.76448178e-01
4.36458439e-01 -1.14555247e-01 5.61792016e-01 1.04123425e+00
3.56251001e-01 7.30648279e-01 -9.40284282e-02 1.05913949e+00
-1.95126221e-01 6.34102374e-02 -3.18085432e-01 -1.44267499e-01
4.18607146e-01 1.10687995e+00 -5.48325181e-01 -6.32870257e-01
-1.59318969e-01 1.33357525e+00 6.24946654e-01 2.54791170e-01
-6.50872111e-01 -4.71527994e-01 -1.08553998e-01 -4.05039489e-02
3.19841594e-01 1.95055798e-01 -5.59615016e-01 -1.18121934e+00
2.76774883e-01 -1.20331812e+00 3.15870613e-01 -1.31419897e+00
-8.23616505e-01 8.75633121e-01 2.46274665e-01 -9.22031522e-01
-6.51360154e-01 -1.20751984e-01 -1.16750455e+00 8.64307165e-01
-1.59272146e+00 -8.59694183e-01 -1.51861593e-01 -1.11106128e-01
1.26136899e+00 1.39530540e-01 7.37256706e-01 -7.48878956e-01
-4.39008027e-01 4.14407432e-01 -4.00570273e-01 -5.14357507e-01
2.97037154e-01 -1.69618213e+00 3.36167186e-01 8.96204114e-01
2.00460508e-01 9.70520496e-01 1.32129896e+00 -4.64925557e-01
-1.57024467e+00 -1.05088425e+00 1.44590533e+00 -3.44547838e-01
3.50239158e-01 2.47725286e-04 -1.05904067e+00 5.89837193e-01
8.03898752e-01 -6.53849661e-01 4.92879272e-01 -6.46709949e-02
1.47806220e-02 1.79022148e-01 -9.87388730e-01 8.48452866e-01
9.71232653e-01 -2.73492098e-01 -1.38214135e+00 4.59329873e-01
1.18375528e+00 -1.71213612e-01 -4.49694335e-01 -5.16336560e-02
2.23739207e-01 -7.99076200e-01 5.02895296e-01 -8.98580432e-01
1.23771012e+00 -8.76918659e-02 2.59647280e-01 -1.60962212e+00
6.87830672e-02 -8.13888729e-01 -3.21443915e-01 1.12717843e+00
9.42161024e-01 -2.93872982e-01 7.31304288e-01 7.36103296e-01
-4.13158357e-01 -1.16740954e+00 -5.70414901e-01 -4.73946601e-01
1.67525217e-01 -5.27723581e-02 6.11219764e-01 1.66649550e-01
7.54984498e-01 9.11299944e-01 -2.60436565e-01 4.70690019e-02
5.21693528e-01 5.27551472e-01 8.25653136e-01 -9.05405998e-01
-2.14893490e-01 -3.54320556e-01 5.12877822e-01 -1.57181919e+00
4.39989299e-01 -7.86672056e-01 5.26278496e-01 -2.22613978e+00
6.17233336e-01 1.87331036e-01 4.23343837e-01 4.26545799e-01
-7.23460197e-01 -5.45843899e-01 2.83191919e-01 1.35463327e-01
-8.61349046e-01 6.83088005e-01 1.30049574e+00 -2.35280007e-01
-1.25660688e-01 1.98207796e-01 -1.36150110e+00 5.44813156e-01
7.48988569e-01 -3.73851329e-01 -7.85354614e-01 -3.43650907e-01
9.64278281e-02 7.08669305e-01 1.56781659e-01 -4.71294433e-01
3.19229782e-01 -3.32002193e-01 1.13871187e-01 -6.65444076e-01
-9.50113982e-02 -3.49809438e-01 -2.94770986e-01 5.57772398e-01
-1.20944214e+00 2.81373799e-01 -1.39390096e-01 5.28224826e-01
-2.50980407e-01 -6.52539790e-01 3.20064634e-01 -1.94935173e-01
3.74261755e-04 -3.96924093e-02 -5.59090734e-01 1.84264660e-01
6.28176570e-01 -3.55537176e-01 -5.67282379e-01 -7.61052966e-01
-4.21523154e-01 4.38764751e-01 2.49365643e-01 1.38279855e-01
9.70052958e-01 -8.63505185e-01 -8.12371850e-01 -2.86690801e-01
1.96896922e-02 1.68833837e-01 -1.80167276e-02 4.82265532e-01
-3.06143135e-01 8.89494538e-01 1.39574900e-01 -1.16863474e-01
-1.11500537e+00 5.04790425e-01 3.26847658e-02 -5.16528308e-01
-9.38080549e-01 5.36470950e-01 1.60181493e-01 -3.40992749e-01
8.16510096e-02 -6.60095274e-01 -3.84563953e-01 -2.42838264e-01
4.84728098e-01 2.28937015e-01 -1.14094473e-01 1.11056576e-02
-1.14521153e-01 1.59830973e-01 -2.90337920e-01 -2.91605264e-01
1.44571495e+00 -2.86408961e-01 1.40499726e-01 2.52776116e-01
9.09107327e-01 2.58781686e-02 -1.10198343e+00 1.12054318e-01
2.97200710e-01 -1.25809565e-01 -3.30910563e-01 -8.51598799e-01
-4.00973350e-01 8.16168725e-01 -5.90304434e-01 5.92675030e-01
1.12538350e+00 3.16141278e-01 9.22036827e-01 8.60419869e-01
-1.71855524e-01 -7.62688220e-01 4.33097959e-01 3.12625736e-01
1.34346426e+00 -1.01368642e+00 1.94275931e-01 -3.96869957e-01
-9.32649016e-01 1.35860765e+00 7.25115776e-01 1.55686503e-02
-3.08689207e-01 -1.63421437e-01 -2.79250532e-01 -3.51722509e-01
-1.42255759e+00 6.57098442e-02 3.57823879e-01 2.97736585e-01
5.25783241e-01 -2.81197965e-01 -5.32505274e-01 8.82448912e-01
-7.68370569e-01 -2.24201784e-01 1.16109109e+00 8.46466243e-01
-1.00132978e+00 -7.84704447e-01 -2.31851593e-01 8.43600214e-01
-2.00988322e-01 -1.99528277e-01 -8.02122355e-01 6.13062024e-01
-6.82799041e-01 1.32677865e+00 -2.51709104e-01 -5.74994795e-02
3.89277101e-01 2.09091514e-01 3.67622256e-01 -9.24617887e-01
-6.37914479e-01 -3.81685406e-01 5.63029408e-01 -3.31651360e-01
-1.29914194e-01 -5.94509721e-01 -1.57092464e+00 -2.84669340e-01
-4.60723132e-01 7.69748330e-01 4.84142154e-01 1.46735942e+00
2.24261105e-01 6.53250635e-01 7.76935875e-01 -4.07760084e-01
-1.05850899e+00 -9.38459337e-01 2.42801029e-02 9.69277415e-03
6.92167640e-01 1.04526281e-01 -6.34990871e-01 1.93609700e-01] | [12.2748441696167, 9.26167106628418] |
38fa9cb4-c816-4aec-af11-43195c61d2ff | a-singing-voice-database-in-basque-for | null | null | https://aclanthology.org/L16-1120 | https://aclanthology.org/L16-1120.pdf | A Singing Voice Database in Basque for Statistical Singing Synthesis of Bertsolaritza | This paper describes the characteristics and structure of a Basque singing voice database of bertsolaritza. Bertsolaritza is a popular singing style from Basque Country sung exclusively in Basque that is improvised and a capella. The database is designed to be used in statistical singing voice synthesis for bertsolaritza style. Starting from the recordings and transcriptions of numerous singers, diarization and phoneme alignment experiments have been made to extract the singing voice from the recordings and create phoneme alignments. This labelling processes have been performed applying standard speech processing techniques and the results prove that these techniques can be used in this specific singing style. | ['David Tavarez', 'Daniel Erro', 'Inma Hernaez', 'Ibon Saratxaga', 'Eva Navas', 'Xabier Sarasola'] | 2016-05-01 | a-singing-voice-database-in-basque-for-1 | https://aclanthology.org/L16-1120 | https://aclanthology.org/L16-1120.pdf | lrec-2016-5 | ['singing-voice-synthesis'] | ['speech'] | [ 1.23303011e-01 -2.42481977e-01 3.18398803e-01 -4.99817431e-02
-6.01658583e-01 -8.55625451e-01 6.03768647e-01 -4.05596673e-01
-3.30109037e-02 6.56996310e-01 2.92573392e-01 -1.23511948e-01
-2.79745936e-01 -2.48369619e-01 -4.84133884e-02 -6.89821661e-01
-3.98329534e-02 7.31904805e-01 7.42049143e-02 -7.19070792e-01
1.74470842e-01 5.70635200e-01 -1.69230640e+00 2.41078287e-01
3.45812291e-01 1.82313457e-01 4.43479836e-01 1.45904744e+00
1.83297947e-01 2.55266756e-01 -1.06665862e+00 7.04952702e-02
1.75340608e-01 -1.13196731e+00 -8.79020393e-01 -3.94975916e-02
5.81723630e-01 1.85611203e-01 -6.14734292e-02 5.81901014e-01
6.18096590e-01 2.93984950e-01 6.95401609e-01 -5.09693027e-01
3.54278952e-01 9.51904535e-01 1.93449959e-01 4.90351826e-01
4.70681518e-01 -1.66460842e-01 1.00026333e+00 -8.24574709e-01
5.49171507e-01 1.07133973e+00 4.72677737e-01 7.10479021e-01
-1.15018237e+00 -7.66549289e-01 -1.04825389e+00 1.57110453e-01
-1.48045552e+00 -6.86411560e-01 9.33912277e-01 -5.52855909e-01
8.61079752e-01 8.82170439e-01 1.01290107e+00 5.59435129e-01
-1.90136746e-01 5.17967522e-01 1.02890110e+00 -9.30111110e-01
-1.48403849e-02 -2.93537471e-02 7.07227513e-02 2.47819126e-01
-4.62665617e-01 2.24234760e-01 -7.28504121e-01 -1.33010268e-01
6.43803537e-01 -8.39939058e-01 -1.99832737e-01 3.15316916e-01
-1.14296818e+00 5.42143583e-01 -2.81895489e-01 1.06854594e+00
-3.68309766e-01 5.19059338e-02 7.74421751e-01 3.97514999e-01
-5.70786446e-02 7.99776614e-01 6.46092817e-02 -7.10381925e-01
-1.54752004e+00 6.88034832e-01 9.50828671e-01 6.67536855e-01
2.03160141e-02 6.96518242e-01 2.06381306e-02 1.16222739e+00
-7.81603828e-02 5.39955020e-01 4.55453694e-01 -1.18273115e+00
9.13438648e-02 -1.40315309e-01 -1.00058489e-01 -5.88101685e-01
2.12672688e-02 -1.39349461e-01 -9.73377451e-02 1.70867994e-01
5.01910627e-01 -1.43970594e-01 -6.22086167e-01 1.29711854e+00
1.45484895e-01 8.81895497e-02 -5.23353331e-02 7.85255730e-01
8.59943867e-01 9.90737081e-01 -3.70382071e-01 -7.06657529e-01
1.13925076e+00 -8.74303818e-01 -1.05160034e+00 6.86629295e-01
-1.74497813e-02 -1.41018260e+00 1.28532648e+00 9.58677709e-01
-1.37819660e+00 -7.89029062e-01 -1.09987497e+00 1.89162895e-01
1.78618878e-01 2.80148625e-01 6.27806336e-02 1.02651489e+00
-7.88862407e-01 6.96216226e-01 -4.25507784e-01 -2.35186324e-01
-5.21537662e-01 2.97532618e-01 -1.88992560e-01 9.03432846e-01
-7.84681618e-01 4.63802755e-01 6.05892837e-01 5.83989583e-02
-1.12630665e+00 -3.07739913e-01 -3.75467867e-01 -9.05577242e-02
2.26485878e-01 -1.30168036e-01 1.60673821e+00 -8.82220924e-01
-2.07355714e+00 9.03765738e-01 -3.51521671e-01 -2.43461639e-01
2.31459707e-01 8.16418082e-02 -8.28038812e-01 1.92502230e-01
-2.16144055e-01 -4.44251634e-02 1.23891377e+00 -9.74833071e-01
-6.19910777e-01 -9.08702314e-02 -7.01574743e-01 3.27535063e-01
-5.96238002e-02 7.53608406e-01 -1.18387423e-01 -9.90457535e-01
2.18569875e-01 -9.69878078e-01 5.74352741e-01 -1.22846496e+00
-3.59206945e-01 -3.57775569e-01 9.97962654e-01 -1.12593627e+00
1.70596969e+00 -2.17434692e+00 4.56530601e-01 4.06337619e-01
-2.14989081e-01 6.09534919e-01 1.61130264e-01 7.42874980e-01
-1.61216781e-01 -2.53483206e-01 -1.40439078e-01 -1.62668616e-01
-1.87222689e-01 5.32544672e-01 -3.20087582e-01 1.36338085e-01
-4.22682256e-01 3.34957302e-01 -7.12269723e-01 -5.68004429e-01
3.00993443e-01 2.58193552e-01 -3.67122859e-01 4.62988317e-01
2.52686325e-03 7.81212866e-01 1.89322487e-01 4.80842054e-01
1.18632816e-01 1.07181120e+00 1.92895278e-01 4.56142090e-02
-5.35323143e-01 8.68817210e-01 -1.18335176e+00 1.32387042e+00
-5.20276666e-01 5.76097608e-01 3.65117759e-01 -8.86694491e-01
1.33564711e+00 9.10300434e-01 2.93380082e-01 5.38358428e-02
2.86791265e-01 7.48034775e-01 4.83564943e-01 -6.32479489e-01
8.44996393e-01 -6.01838529e-01 1.18117265e-01 2.02550858e-01
3.46447289e-01 -9.19873714e-01 6.63787007e-01 -3.27500284e-01
4.43361253e-01 1.44762650e-01 4.45106596e-01 -4.23260957e-01
9.66038048e-01 1.62438080e-01 5.19175529e-01 2.56531984e-01
3.86405215e-02 8.02682877e-01 -1.58774562e-03 -2.17590574e-02
-1.23351014e+00 -1.03779876e+00 -2.24773020e-01 1.05528712e+00
-7.61193871e-01 -7.19851732e-01 -9.96029913e-01 -1.47077097e-02
-6.48131788e-01 7.61025488e-01 -5.39428666e-02 4.11721110e-01
-1.20273256e+00 5.19113205e-02 9.13951814e-01 1.01552270e-01
1.06679097e-01 -1.60896277e+00 -1.91496432e-01 4.16924357e-01
-1.42767623e-01 -5.89750171e-01 -8.68452668e-01 -4.44281437e-02
-7.01692343e-01 -8.97665024e-01 -8.66422117e-01 -1.12562025e+00
-3.64376128e-01 -4.03181091e-03 9.99197423e-01 -2.01198056e-01
-2.15297267e-01 2.50934750e-01 -4.25932825e-01 -7.19160378e-01
-1.16347408e+00 -8.91028158e-03 3.43685955e-01 -1.83240771e-01
3.02946925e-01 -8.17965806e-01 -2.09993068e-02 2.47986972e-01
-6.12844169e-01 -3.48462105e-01 -8.22972581e-02 8.23957443e-01
5.20610809e-01 3.33685338e-01 5.93281627e-01 -7.72148371e-01
7.55164146e-01 -4.97611016e-02 -5.62355340e-01 -2.97213048e-01
-1.38584644e-01 -7.30385661e-01 8.69597316e-01 -3.63974452e-01
-7.60977626e-01 -2.41309986e-03 -6.22682273e-01 -2.91969568e-01
-3.45680922e-01 1.47614136e-01 -3.12406689e-01 2.42842305e-02
8.35898340e-01 4.02046800e-01 3.80424447e-02 -7.86954880e-01
5.27527370e-02 1.23985779e+00 1.21519458e+00 -4.91799682e-01
8.11105728e-01 -4.07070443e-02 -1.81140900e-01 -1.63864410e+00
-1.49350658e-01 -8.94747317e-01 -7.06091464e-01 -6.05586112e-01
8.45373333e-01 -3.08939666e-01 -7.55652487e-01 5.11233032e-01
-9.50405717e-01 8.62662196e-02 -5.57582498e-01 7.00703919e-01
-9.46111321e-01 3.79942507e-01 -5.05009413e-01 -1.18527818e+00
-5.63537061e-01 -7.40191579e-01 6.40891731e-01 3.42470616e-01
-7.56748497e-01 -8.22783589e-01 5.04971683e-01 4.99709874e-01
2.19792854e-02 -1.52215213e-01 7.56203830e-01 -6.49925053e-01
1.50584251e-01 -8.16069767e-02 7.90569007e-01 8.10274661e-01
4.44814414e-01 2.22524345e-01 -1.03347707e+00 -1.97485238e-01
2.69439757e-01 -1.96183138e-02 3.13790351e-01 2.16494635e-01
5.24782121e-01 -2.73573101e-01 4.60106224e-01 3.40109676e-01
8.65027070e-01 7.04867065e-01 4.61014986e-01 -5.10900207e-02
5.15882790e-01 8.59552681e-01 7.80564547e-01 1.45479605e-01
-2.60727853e-01 9.77994680e-01 -2.24556968e-01 4.33775246e-01
-5.93416691e-01 -6.05794609e-01 7.11717427e-01 1.71527421e+00
-5.92429280e-01 2.07771003e-01 -5.98067105e-01 1.03169918e+00
-1.26771140e+00 -1.22363663e+00 -4.81050849e-01 2.16503787e+00
9.58899140e-01 -8.16383883e-02 9.20278013e-01 1.07073438e+00
7.73842990e-01 3.59530985e-01 2.18117431e-01 -1.22726870e+00
-1.51316479e-01 9.89274383e-01 1.02575667e-01 7.90556490e-01
-6.65638149e-01 1.13196993e+00 7.10975409e+00 1.11231589e+00
-1.19611430e+00 -1.16646541e-02 -3.56747091e-01 -1.56107277e-01
-1.75434768e-01 9.44474265e-02 -7.68965483e-01 3.72313231e-01
1.22936928e+00 -2.07132801e-01 9.26164627e-01 3.28131437e-01
8.29774022e-01 4.93536443e-02 -8.05241108e-01 1.17585742e+00
3.91121298e-01 -1.29901743e+00 -6.58018216e-02 -1.85665190e-01
6.85015261e-01 -3.72915953e-01 -2.60101289e-01 1.46712169e-01
-4.23382401e-01 -1.03641450e+00 8.95656705e-01 1.98697850e-01
6.67314768e-01 -8.67827535e-01 3.49610001e-01 3.19961786e-01
-1.08660161e+00 1.84202060e-01 -1.25760570e-01 -4.24717329e-02
3.96704227e-01 4.18591872e-02 -1.45917284e+00 4.66010839e-01
3.66281927e-01 3.87187034e-01 -4.94642258e-02 1.00842357e+00
-2.02879563e-01 1.67408454e+00 -4.42876220e-01 -1.57767504e-01
5.70284352e-02 -7.44701684e-01 1.31103885e+00 1.35987890e+00
4.90298539e-01 -9.53419413e-03 -2.89116174e-01 7.47101068e-01
4.76898611e-01 7.49368787e-01 -4.97328311e-01 -5.50973117e-01
4.22337651e-01 1.10306168e+00 -4.07359332e-01 -3.33693802e-01
2.49247864e-01 7.77011514e-01 -6.41902983e-01 1.36814686e-02
-2.91390479e-01 -5.88826060e-01 3.72349918e-01 5.07907748e-01
1.88118488e-01 -5.15359819e-01 -2.17852369e-01 -4.96169716e-01
-3.93974155e-01 -1.19281590e+00 3.72332871e-01 -5.70797026e-01
-6.86637700e-01 7.96841502e-01 4.81797829e-02 -1.09174383e+00
-9.47581351e-01 -4.59656417e-01 -1.01684558e+00 1.19223964e+00
-5.19632041e-01 -1.01188684e+00 2.10563630e-01 4.56227213e-01
1.05933762e+00 -7.35846519e-01 1.14144981e+00 3.24827045e-01
-1.80304766e-01 2.63497293e-01 1.52676493e-01 -1.17145926e-01
6.02734745e-01 -1.57876325e+00 1.17119156e-01 6.58591807e-01
7.19724655e-01 4.22457248e-01 1.21302009e+00 -3.47260982e-01
-1.05909431e+00 -3.97297591e-01 1.53411305e+00 -2.47463599e-01
6.48672342e-01 -2.68368244e-01 -7.31735647e-01 2.49480695e-01
4.20228839e-01 -7.63419688e-01 7.31339574e-01 -9.99997333e-02
1.13371417e-01 -1.73537537e-01 -8.17631781e-01 4.45885122e-01
7.00842977e-01 -6.93113029e-01 -1.19272232e+00 2.21144125e-01
9.21759307e-02 -1.25939786e-01 -7.67001867e-01 -1.03626717e-02
5.90376735e-01 -9.84401405e-01 3.93663883e-01 -4.41607445e-01
-2.53686029e-02 -7.02101588e-01 8.85902271e-02 -1.32862914e+00
1.68507472e-02 -1.57592905e+00 3.94565701e-01 1.58368552e+00
1.92641780e-01 -2.71683753e-01 4.63564098e-01 -5.14990151e-01
-4.14815605e-01 4.58269686e-01 -1.06503212e+00 -9.54468966e-01
9.91316363e-02 -3.94810140e-01 3.08619857e-01 7.17468798e-01
1.47917435e-01 6.90014243e-01 -6.03284121e-01 -3.54284704e-01
4.34877247e-01 6.55755103e-02 8.72271836e-01 -1.11424196e+00
-7.38985837e-01 -3.81574214e-01 -2.29228169e-01 -5.33996284e-01
8.80776495e-02 -8.20721149e-01 1.53867349e-01 -8.33936572e-01
-5.53454161e-01 -3.47278893e-01 -3.61945182e-02 7.98202213e-03
2.31727868e-01 5.73448062e-01 4.14881170e-01 1.78855255e-01
4.36767578e-01 1.15080565e-01 1.22980809e+00 1.16893716e-01
-6.40124440e-01 5.82734108e-01 1.80066735e-01 7.69207060e-01
8.44479561e-01 -4.04560655e-01 -3.83787125e-01 3.88592184e-01
-2.03696162e-01 3.59052956e-01 -1.27838343e-01 -1.04367578e+00
-6.09412715e-02 3.61185931e-02 -3.56476784e-01 -8.76506209e-01
6.26367033e-01 -4.47011799e-01 6.58888459e-01 2.74942964e-01
-2.25872815e-01 -1.45326927e-01 -4.94950078e-02 -6.82809874e-02
-5.94659925e-01 -8.01258445e-01 9.59852636e-01 3.66367074e-03
-4.76504892e-01 -3.31521422e-01 -8.22627723e-01 -1.41785398e-01
8.74997616e-01 -6.06999397e-01 4.83929962e-01 -4.67023015e-01
-9.91729081e-01 -5.90027928e-01 1.87436715e-01 2.36980274e-01
2.64270246e-01 -1.27116609e+00 -1.04641628e+00 4.54854071e-01
2.84019895e-02 -5.08840561e-01 9.32846069e-02 9.29148257e-01
-1.06745183e+00 5.08863151e-01 -4.25429046e-01 -3.60600322e-01
-2.00154543e+00 1.56607285e-01 3.69453907e-01 3.19205105e-01
-6.56186938e-01 5.77420592e-01 -3.74981612e-01 -2.93168247e-01
9.45561379e-02 7.41966441e-02 -7.25556314e-01 1.22709438e-01
3.59804124e-01 8.65277648e-01 9.58709221e-04 -1.23208070e+00
-3.52127515e-02 4.99830663e-01 4.54222560e-01 -7.00413883e-01
9.19635713e-01 -8.56829584e-02 -4.19398040e-01 1.11781800e+00
8.23600233e-01 1.06069493e+00 -2.08515182e-01 4.69519377e-01
5.22326902e-02 -6.27968609e-01 1.38158193e-02 -6.40568614e-01
-3.92453253e-01 8.55073333e-01 2.94347942e-01 4.49212790e-01
1.06706023e+00 -2.38228485e-01 1.11194444e+00 -2.82472558e-02
1.08478978e-01 -1.38949239e+00 -5.66976845e-01 6.95748746e-01
1.15197539e+00 -4.49375033e-01 -6.21919036e-01 -5.26862741e-01
-6.70712411e-01 1.30818176e+00 -8.21040869e-02 -4.72121388e-01
6.31543040e-01 1.80740699e-01 3.25268120e-01 2.33020782e-01
-3.77872050e-01 -2.41623700e-01 5.05687773e-01 6.14603460e-01
7.69746065e-01 4.19777423e-01 -1.48143589e+00 4.62398916e-01
-1.22049570e+00 -2.30372846e-01 5.46758533e-01 3.84558767e-01
-4.50240999e-01 -1.54281354e+00 -9.06827033e-01 8.72922242e-02
-7.39892423e-01 -1.20733075e-01 -8.88513744e-01 5.87738037e-01
3.14985037e-01 1.17382121e+00 -9.25861895e-02 -4.84560996e-01
4.60816085e-01 3.14294398e-01 6.54062390e-01 -8.76455963e-01
-1.26333046e+00 1.07615197e+00 5.49062192e-01 1.11029550e-01
-2.68744022e-01 -7.57094502e-01 -1.23764527e+00 -2.34425396e-01
-1.66277260e-01 1.03403866e+00 8.53009522e-01 6.85129523e-01
-4.21323597e-01 4.82142806e-01 9.33930218e-01 -7.67798722e-01
-4.31321532e-01 -1.42021942e+00 -1.27441239e+00 2.62293369e-01
1.09660551e-01 -2.19975173e-01 -4.30835664e-01 2.52853841e-01] | [15.125420570373535, 6.242941379547119] |
f74018a3-6564-4fab-af82-6021aad5df4a | ballgan-3d-aware-image-synthesis-with-a | 2301.09091 | null | https://arxiv.org/abs/2301.09091v1 | https://arxiv.org/pdf/2301.09091v1.pdf | BallGAN: 3D-aware Image Synthesis with a Spherical Background | 3D-aware GANs aim to synthesize realistic 3D scenes such that they can be rendered in arbitrary perspectives to produce images. Although previous methods produce realistic images, they suffer from unstable training or degenerate solutions where the 3D geometry is unnatural. We hypothesize that the 3D geometry is underdetermined due to the insufficient constraint, i.e., being classified as real image to the discriminator is not enough. To solve this problem, we propose to approximate the background as a spherical surface and represent a scene as a union of the foreground placed in the sphere and the thin spherical background. It reduces the degree of freedom in the background field. Accordingly, we modify the volume rendering equation and incorporate dedicated constraints to design a novel 3D-aware GAN framework named BallGAN. BallGAN has multiple advantages as follows. 1) It produces more reasonable 3D geometry; the images of a scene across different viewpoints have better photometric consistency and fidelity than the state-of-the-art methods. 2) The training becomes much more stable. 3) The foreground can be separately rendered on top of different arbitrary backgrounds. | ['Youngjung Uh', 'Hyeran Byun', 'Hyunsu Kim', 'Young Sun Choi', 'Jeongmin Bae', 'Yunji Seo', 'Minjung Shin'] | 2023-01-22 | null | null | null | null | ['3d-aware-image-synthesis'] | ['computer-vision'] | [ 3.39897335e-01 2.79018104e-01 3.48664492e-01 -2.03135788e-01
-5.15469551e-01 -5.17587900e-01 5.98072946e-01 -8.66595209e-01
1.46211892e-01 7.09513009e-01 -7.35923797e-02 -1.48666292e-01
5.07288218e-01 -8.65201294e-01 -8.81338239e-01 -9.21885550e-01
6.72911167e-01 4.98257637e-01 4.37937319e-01 -1.24145253e-02
-1.30263820e-01 6.16404474e-01 -1.58046901e+00 2.45932676e-02
1.17401648e+00 1.06077170e+00 3.14260095e-01 4.58341688e-01
-2.32993677e-01 6.14839017e-01 -8.84230375e-01 -4.72451627e-01
6.95653141e-01 -9.10542428e-01 -2.94661462e-01 6.31907344e-01
6.33744240e-01 -6.01731241e-01 -2.22891033e-01 1.23170877e+00
3.43501598e-01 3.69541310e-02 6.93783879e-01 -1.26554751e+00
-6.58148527e-01 -1.46186560e-01 -7.80151725e-01 -4.12701547e-01
2.28263021e-01 9.24533457e-02 4.21807408e-01 -7.27470696e-01
6.20275795e-01 1.30064344e+00 2.72562385e-01 7.77246892e-01
-1.24636447e+00 -4.59341258e-01 4.07966226e-01 -2.73570180e-01
-1.36948931e+00 -3.41152966e-01 1.03668940e+00 -4.72534031e-01
1.28678501e-01 5.94953179e-01 8.47027600e-01 1.07387137e+00
1.02843553e-01 7.49268711e-01 1.14954805e+00 -3.48724067e-01
3.12107682e-01 1.94960698e-01 -5.78414321e-01 6.57340467e-01
3.77663255e-01 -4.97449227e-02 -2.36051947e-01 -6.87328205e-02
1.39944613e+00 -2.03287869e-04 -5.92772663e-01 -6.39070570e-01
-1.07485652e+00 5.58419228e-01 2.91687071e-01 -1.49908345e-02
-2.89860785e-01 -9.16954949e-02 -2.94803411e-01 -2.35612527e-01
5.79901695e-01 4.25263435e-01 4.44402136e-02 3.00063163e-01
-9.14679885e-01 3.07963461e-01 5.51589966e-01 1.14195907e+00
7.18483090e-01 5.65296412e-01 -2.62701809e-02 6.52226627e-01
3.44665259e-01 7.89309978e-01 1.38636947e-01 -1.24273849e+00
2.69974679e-01 6.47200644e-01 3.53498071e-01 -8.56019557e-01
2.07123771e-01 -4.14151341e-01 -9.12151217e-01 5.64164281e-01
5.39922595e-01 -7.78298080e-02 -1.05409050e+00 1.61466873e+00
7.28541136e-01 3.61548781e-01 -1.04990369e-02 1.22412610e+00
1.04144597e+00 8.75261843e-01 -5.14703929e-01 -2.32758209e-01
1.07796979e+00 -9.32933211e-01 -8.07158887e-01 -2.78173745e-01
1.86725799e-02 -8.68703961e-01 9.62208092e-01 4.68857408e-01
-1.43715179e+00 -5.22065103e-01 -1.04798090e+00 -1.30677540e-02
1.83992043e-01 9.20438319e-02 5.01029134e-01 7.33053803e-01
-9.89434898e-01 1.86840996e-01 -6.54709518e-01 1.02779843e-01
2.70252883e-01 -9.09664482e-02 -1.82935387e-01 -2.23032936e-01
-7.76703060e-01 7.83029974e-01 5.48261702e-02 9.40290317e-02
-9.84904766e-01 -5.84732413e-01 -9.02578056e-01 -2.29674876e-01
3.92606765e-01 -9.63654399e-01 8.85983467e-01 -1.29403067e+00
-1.71915472e+00 1.07687712e+00 -2.86261082e-01 2.72456735e-01
9.43926334e-01 2.31249817e-02 -1.49735957e-01 6.24378659e-02
-3.35516930e-02 3.78807515e-01 9.86229122e-01 -1.91195118e+00
-3.08872551e-01 -3.05135727e-01 2.34433725e-01 4.76084322e-01
2.39954039e-01 -2.67669201e-01 -7.56669283e-01 -6.69922411e-01
4.84142423e-01 -9.13173139e-01 -2.23224521e-01 2.58482158e-01
-6.39717638e-01 3.52672398e-01 9.13782716e-01 -5.89760959e-01
6.61145985e-01 -2.08843088e+00 2.07289323e-01 -1.60596400e-01
2.96938896e-01 2.94168055e-01 1.66777503e-02 -9.15737823e-03
1.03003092e-01 4.27760743e-02 -2.79366821e-01 -4.42938268e-01
-1.78380907e-01 2.66338497e-01 -2.67851830e-01 5.22844434e-01
1.12959698e-01 7.24310279e-01 -7.86346257e-01 -5.63644707e-01
4.03343588e-01 7.76989996e-01 -5.05054474e-01 5.26432633e-01
-3.22235435e-01 1.02758062e+00 -5.83711743e-01 5.67872882e-01
1.21925938e+00 -2.99467444e-01 9.34675336e-03 -1.94135800e-01
-1.28309652e-01 9.97929089e-03 -1.36425889e+00 1.45490897e+00
-2.90213555e-01 4.04285759e-01 2.78621554e-01 -7.00978637e-01
1.02427435e+00 1.26386344e-01 3.42788547e-01 -5.36283970e-01
1.84542984e-01 9.41696018e-02 -2.14517355e-01 -2.93210715e-01
2.30153799e-01 -2.61291295e-01 3.05889606e-01 2.35172078e-01
-3.01180005e-01 -8.43414426e-01 -1.34345919e-01 2.88849361e-02
5.52697122e-01 4.85958576e-01 -8.06028172e-02 -1.61446586e-01
4.70203102e-01 -2.73443639e-01 9.95741785e-01 4.87178534e-01
1.83050811e-01 1.20855415e+00 5.12093365e-01 -2.91245043e-01
-1.07918978e+00 -1.28028643e+00 -1.03342913e-01 2.61241645e-01
6.80153012e-01 4.49580029e-02 -1.03347480e+00 -3.94064546e-01
-2.35077396e-01 7.92486429e-01 -4.70320135e-01 -4.96425219e-02
-6.34302497e-01 -6.46779239e-01 5.46933971e-02 2.32215181e-01
7.25543022e-01 -5.80411494e-01 -6.23898804e-01 -2.26819023e-01
-4.30745870e-01 -1.19912934e+00 -5.17851830e-01 -3.36011976e-01
-8.40317905e-01 -1.07083511e+00 -1.01333117e+00 -5.59739351e-01
9.85761702e-01 5.30424535e-01 1.37456512e+00 -4.46016937e-02
-5.13836518e-02 8.77873376e-02 -2.03210607e-01 -4.34157550e-01
-6.08206511e-01 -7.09043622e-01 -4.56024818e-02 3.88430655e-01
-1.78842679e-01 -5.44907093e-01 -8.57944250e-01 7.01424181e-01
-8.73173535e-01 5.65081477e-01 2.45670706e-01 7.10017383e-01
9.32130098e-01 1.93212524e-01 9.00306031e-02 -8.40446591e-01
-1.76407937e-02 -8.50771740e-02 -8.73483062e-01 2.75686800e-01
-1.04732208e-01 -3.00765723e-01 6.69030964e-01 -5.61300814e-01
-1.43671548e+00 6.07919134e-02 1.69259459e-02 -7.41735041e-01
-1.17479309e-01 -3.63623321e-01 -7.23876536e-01 -1.51789144e-01
4.50541437e-01 3.95166010e-01 -1.39181927e-01 -3.51843327e-01
2.38657340e-01 3.38727385e-01 3.35016161e-01 -6.76210165e-01
9.78751898e-01 8.28366339e-01 1.78187385e-01 -9.29916203e-01
-9.11454260e-01 5.81131577e-02 -5.41441023e-01 -4.46751326e-01
9.58826005e-01 -9.42597926e-01 -4.02151287e-01 7.15271771e-01
-1.22312963e+00 -5.50547540e-01 -3.62937003e-01 4.63790178e-01
-5.33169687e-01 4.63536114e-01 -4.01601225e-01 -9.86345708e-01
4.25868221e-02 -1.35818470e+00 1.29170978e+00 4.49493587e-01
1.54111773e-01 -7.83809423e-01 -3.54377955e-01 5.00046730e-01
3.02596241e-01 6.79264486e-01 7.73900986e-01 2.57953763e-01
-9.76154089e-01 9.52960551e-02 -2.18593970e-01 4.55700248e-01
2.91629493e-01 4.38613474e-01 -1.21781683e+00 -5.64122349e-02
4.07166332e-01 -3.46205235e-02 5.20133018e-01 5.48068821e-01
1.19263756e+00 -2.54925847e-01 -1.49981186e-01 1.01405227e+00
1.25669599e+00 3.98287892e-01 7.48932362e-01 -5.20253256e-02
9.93982255e-01 6.41215861e-01 5.19831717e-01 2.59753048e-01
2.83275157e-01 7.88991272e-01 7.82194018e-01 -4.86994863e-01
-3.70783657e-01 -4.02749121e-01 2.02418014e-01 6.67932272e-01
-1.24844380e-01 -5.61564744e-01 -4.85116094e-01 1.87344491e-01
-1.56607902e+00 -7.19533026e-01 -3.86765271e-01 2.42117333e+00
5.64927876e-01 -2.06755940e-02 -1.01435557e-01 -4.45317253e-02
8.06030273e-01 2.02306077e-01 -8.26865554e-01 -5.70366830e-02
-4.80126888e-01 -1.99131742e-01 2.82234490e-01 4.58752632e-01
-7.39393055e-01 8.17539871e-01 6.24940538e+00 7.08387434e-01
-1.31422472e+00 -2.00188830e-02 1.02026403e+00 -2.00499266e-01
-8.29745591e-01 -7.97351003e-02 -7.45716155e-01 6.65043116e-01
3.26667391e-02 5.43606915e-02 3.04979801e-01 6.46386921e-01
1.04609586e-01 -1.53887257e-01 -9.08612192e-01 1.18652999e+00
3.07561725e-01 -1.21171904e+00 3.39277208e-01 1.24551818e-01
1.09471071e+00 -3.40416670e-01 1.54673025e-01 -1.41917363e-01
2.84581304e-01 -9.68295157e-01 1.11082697e+00 6.20685756e-01
1.07281363e+00 -4.83676493e-01 3.43877763e-01 6.39203131e-01
-8.72174382e-01 4.80656028e-01 -4.67648089e-01 1.71460807e-01
4.78211224e-01 8.01323533e-01 -2.77346879e-01 5.80019832e-01
5.32441258e-01 3.34800273e-01 -2.39892974e-01 8.77420902e-01
-4.06400800e-01 2.22270355e-01 -3.27733666e-01 2.86620378e-01
5.14700487e-02 -9.18683469e-01 6.59828961e-01 5.11708260e-01
6.52364373e-01 3.29834253e-01 1.43065393e-01 1.34424424e+00
5.42564206e-02 -4.74022999e-02 -6.93893075e-01 3.16705316e-01
2.40249708e-01 1.09497142e+00 -6.80073798e-01 -1.96570888e-01
-4.47990119e-01 1.24673188e+00 7.71480948e-02 7.29001999e-01
-9.74195302e-01 1.00259311e-01 6.07195139e-01 3.58260930e-01
2.33888149e-01 -3.50424498e-02 -2.14115500e-01 -1.61875904e+00
2.98901498e-01 -8.18288207e-01 8.35008267e-03 -1.29866755e+00
-1.11328828e+00 9.42480087e-01 -1.59636915e-01 -1.32563508e+00
3.58777791e-02 -6.54182792e-01 -7.34247625e-01 1.03818381e+00
-1.31109500e+00 -1.05083919e+00 -5.99610031e-01 5.61564744e-01
3.44590187e-01 1.89213127e-01 7.69691765e-01 1.99725151e-01
-5.23505926e-01 3.39923143e-01 -1.41241346e-02 -1.45523816e-01
5.83371758e-01 -1.13159418e+00 3.37573409e-01 1.10953104e+00
-7.52212033e-02 3.31340253e-01 5.75694561e-01 -5.85021496e-01
-1.37557828e+00 -1.00846720e+00 4.00880873e-01 -5.20073712e-01
3.56353894e-02 -5.78792036e-01 -9.07801807e-01 4.44656134e-01
5.76444110e-03 1.41221002e-01 4.46564853e-01 -5.41267157e-01
-2.48584494e-01 -1.37082949e-01 -1.41571558e+00 8.48890901e-01
1.12056231e+00 -1.99390158e-01 -9.91136357e-02 2.91250616e-01
6.81155086e-01 -8.27488959e-01 -6.40451193e-01 3.65986973e-01
2.60980695e-01 -1.32231581e+00 9.57492828e-01 -1.49175003e-01
4.42131341e-01 -6.50754094e-01 -3.25112700e-01 -1.34071302e+00
-1.55402860e-02 -8.34681511e-01 -2.21140474e-01 1.18752766e+00
7.06636757e-02 -7.22095549e-01 9.00157630e-01 7.76441097e-01
-1.42033264e-01 -6.78246498e-01 -8.62457395e-01 -7.39781916e-01
1.02159241e-02 -4.04119231e-02 8.45479071e-01 9.33484554e-01
-7.18924165e-01 2.03195468e-01 -5.69952607e-01 2.21662179e-01
8.67497444e-01 5.18777251e-01 1.03669882e+00 -1.23649967e+00
-3.43797535e-01 -2.64335603e-01 -2.03731492e-01 -1.61445785e+00
-6.53774813e-02 -4.71424043e-01 2.25347765e-02 -1.46155488e+00
1.23223044e-01 -7.90954292e-01 3.27587694e-01 -1.39906585e-01
-2.45715886e-01 3.22079331e-01 1.36644304e-01 1.86198309e-01
-1.98063001e-01 8.24010611e-01 2.14587569e+00 6.90357834e-02
-2.65445970e-02 1.28313228e-01 -7.26848960e-01 8.53031814e-01
6.50294542e-01 -1.33165225e-01 -5.68644583e-01 -6.81566119e-01
8.03883374e-02 2.49433294e-01 3.83034497e-01 -6.44420683e-01
-3.14368635e-01 -3.49718183e-01 6.97794378e-01 -6.20279312e-01
7.52555728e-01 -8.87886226e-01 6.01277590e-01 1.64809581e-02
2.13348880e-01 -3.63812566e-01 -8.77184570e-02 4.75701839e-01
-8.30122381e-02 -2.08568826e-01 1.07965755e+00 -3.99189770e-01
-7.02077001e-02 5.32717407e-01 -1.31397903e-01 2.52938777e-01
1.08588612e+00 -5.24025321e-01 -2.28441596e-01 -6.55969083e-01
-2.82236993e-01 -1.23496078e-01 1.13968909e+00 1.53627068e-01
5.32142937e-01 -1.46724761e+00 -6.76055908e-01 5.05895138e-01
-1.98006824e-01 7.47595072e-01 3.02813858e-01 4.53786314e-01
-9.15349841e-01 -1.61434337e-02 2.78474856e-02 -9.04677331e-01
-1.10139847e+00 4.08054680e-01 6.53532147e-01 1.09074675e-01
-7.73332536e-01 9.31838036e-01 1.02647913e+00 -5.67636728e-01
2.11293101e-01 -1.92002192e-01 1.02114141e-01 -3.73442233e-01
4.92773771e-01 1.48714051e-01 -2.79983312e-01 -8.17284763e-01
-1.70685336e-01 8.47279608e-01 3.48826468e-01 1.95034072e-02
1.12220156e+00 -1.25471249e-01 -4.20878716e-02 3.61994565e-01
7.82915831e-01 2.79851586e-01 -1.78077149e+00 -6.31168336e-02
-9.07508135e-01 -1.07271302e+00 -1.71012804e-02 -5.03121972e-01
-1.40070856e+00 8.39568973e-01 2.76114762e-01 2.21107885e-01
1.06781256e+00 -7.55190104e-02 7.61473417e-01 -2.34337702e-01
4.03142452e-01 -7.19482601e-01 4.29084748e-02 2.82183945e-01
9.68583703e-01 -1.11818564e+00 5.28375991e-02 -8.94208848e-01
-7.00894773e-01 8.97341311e-01 8.59660804e-01 -1.17664233e-01
3.94024581e-01 2.46279731e-01 3.17799389e-01 -4.72676419e-02
-2.75305063e-01 8.74109939e-02 4.89739835e-01 7.46073544e-01
3.08025569e-01 -3.38680893e-02 2.25758091e-01 2.91787505e-01
-1.88722372e-01 -4.95942712e-01 6.24832034e-01 4.16201353e-01
-2.76502389e-02 -7.96735048e-01 -5.78251600e-01 4.48558852e-02
-4.00397927e-01 1.20029822e-01 -4.12103206e-01 6.57976747e-01
2.18134761e-01 9.52749312e-01 1.27524197e-01 5.74231409e-02
3.58461827e-01 -3.73645633e-01 7.73615062e-01 -5.39643645e-01
6.39739111e-02 4.45637435e-01 -1.82497934e-01 -6.32400990e-01
-4.72861975e-01 -4.71699834e-01 -9.61314201e-01 -1.92419022e-01
-5.45364738e-01 2.85298545e-02 5.58712065e-01 5.56090534e-01
2.13660926e-01 5.52659929e-01 7.62315273e-01 -1.17370868e+00
-1.57176405e-01 -5.33522666e-01 -8.02477896e-01 4.09648776e-01
2.73466647e-01 -7.74386108e-01 -6.68795109e-01 7.91724548e-02] | [9.354724884033203, -3.121927261352539] |
13de4271-5372-42d5-8304-c7d8ab7e33cd | squash-root-microbiome-transplants-and | 2109.07521 | null | https://arxiv.org/abs/2109.07521v1 | https://arxiv.org/pdf/2109.07521v1.pdf | Squash root microbiome transplants and metagenomic inspection for in situ arid adaptations | Arid zones contain a diverse set of microbes capable of survival under dry conditions, some of which can form relationships with plants under drought stress conditions to improve plant health. We studied squash (Cucurbita pepo L.) root microbiome under historically arid and humid sites, both in situ and performing a common garden experiment. Plants were grown in soils from sites with different drought levels, using in situ collected soils as the microbial source. We described and analyzed bacterial diversity by 16S rRNA gene sequencing (N=48) from the soil, rhizosphere, and endosphere. Proteobacteria were the most abundant phylum present in humid and arid samples, while Actinobacteriota abundance was higher in arid ones. The Beta-diversity analyses showed split microbiomes between arid and humid microbiomes, and aridity and soil pH levels could explain it. These differences between humid and arid microbiomes were maintained in the common garden experiment, showing that it is possible to transplant in situ diversity to the greenhouse. We detected a total of 1009 bacterial genera; 199 exclusively associated with roots under arid conditions. With shotgun metagenomic sequencing of rhizospheres (N=6), we identified 2969 protein families in the squash core metagenome and found an increased number of exclusively protein families from arid (924) than humid samples (158). We found arid conditions enriched genes involved in protein degradation and folding, oxidative stress, compatible solute synthesis, and ion pumps associated with osmotic regulation. Plant phenotyping allowed us to correlate bacterial communities with plant growth. Our study revealed that it is possible to evaluate microbiome diversity ex-situ and identify critical species and genes involved in plant-microbe interactions in historically arid locations. | ['Luis D. Alcaraz', 'Daniel Piñero', 'Hugo R. Barajas', 'Miguel F. Romero', 'Rocío Cruz-Ortega', 'Felipe García-Oliva', 'Cristóbal Hernández-Álvarez'] | 2021-09-15 | null | null | null | null | ['plant-phenotyping'] | ['computer-vision'] | [ 3.44933271e-01 -1.83605418e-01 -1.63991541e-01 4.26273197e-01
5.04101753e-01 -1.00935435e+00 2.70992577e-01 5.82494617e-01
3.21517736e-02 8.93741608e-01 7.27019161e-02 -5.93106747e-01
-1.47759318e-01 -1.21363616e+00 -5.46230495e-01 -1.04786301e+00
-6.83549762e-01 2.98904926e-01 -5.75438738e-02 -5.29923081e-01
9.38996151e-02 6.60541952e-01 -2.02505493e+00 5.42720020e-01
6.76584542e-01 -1.07137375e-01 1.49539673e+00 5.83147228e-01
-1.43329471e-01 -2.71077573e-01 -1.69111937e-01 7.44342923e-01
2.17726603e-01 -1.92627504e-01 -5.92888117e-01 -5.73473051e-02
-3.63552153e-01 -5.79700023e-02 4.42277074e-01 8.30370486e-01
3.05654824e-01 -2.92557746e-01 5.17394602e-01 -6.07167900e-01
-4.19140190e-01 8.91161382e-01 -7.82121480e-01 -3.03456128e-01
1.90398529e-01 2.91582376e-01 7.87079632e-01 -9.96790409e-01
1.06495416e+00 1.50915778e+00 2.37009227e-01 1.05816178e-01
-1.34727800e+00 -2.51859635e-01 -1.46912351e-01 2.55114555e-01
-1.15149164e+00 -4.54941779e-01 -3.53543758e-01 -8.74652445e-01
1.00563717e+00 1.69695720e-01 1.34479582e+00 9.23626602e-01
8.32525909e-01 1.42861724e-01 1.10396767e+00 -1.20603666e-01
4.86222059e-01 -3.50489505e-02 -3.15471530e-01 -8.20557699e-02
1.07797265e+00 1.94812536e-01 -1.66083157e-01 -4.78008598e-01
3.46795738e-01 2.00750679e-02 -9.51503336e-01 -4.76222038e-01
-1.21737969e+00 6.41818941e-01 5.93429029e-01 4.77645844e-01
-6.82676196e-01 -1.73508406e-01 8.00788760e-01 7.06908107e-02
7.63135031e-02 6.03797615e-01 -1.05579114e+00 2.11554781e-01
4.71831076e-02 -1.49758622e-01 1.04948425e+00 4.75511551e-01
9.80652213e-01 -2.46683851e-01 5.73604345e-01 1.13503754e+00
4.72775310e-01 9.35520768e-01 6.77717328e-01 -6.51669204e-01
-4.71096039e-01 1.69103324e-01 2.68802196e-01 -1.08474743e+00
-5.50988168e-02 -8.88357833e-02 -4.52094555e-01 2.02129990e-01
2.33430997e-01 2.84771383e-01 -4.94382828e-01 1.58033919e+00
3.69419694e-01 -5.25525630e-01 2.32829392e-01 6.62182331e-01
-4.75724461e-03 5.95278144e-01 1.29630029e-01 -5.72852910e-01
2.02499485e+00 -2.48441741e-01 -5.96093118e-01 -1.40209183e-01
5.79029620e-01 -9.56554890e-01 8.33859622e-01 2.69318789e-01
-5.70269227e-01 -5.32628112e-02 -1.17414069e+00 8.34059298e-01
-5.72918177e-01 -2.98350245e-01 3.97693753e-01 9.19990361e-01
-9.84660625e-01 1.04504311e+00 -1.14248824e+00 -1.49237752e+00
-1.18888654e-01 -1.55684069e-01 -4.93975490e-01 -3.48721683e-01
-8.20026100e-01 1.13221359e+00 5.58791041e-01 1.87907845e-01
-1.08244908e+00 -3.10043961e-01 -3.05779368e-01 3.14564586e-01
-4.14628051e-02 -4.06024247e-01 7.49426603e-01 -3.24764609e-01
-1.60521650e+00 8.71073067e-01 -3.04861903e-01 1.37618527e-01
-3.53386134e-01 -2.34036624e-01 2.59293765e-02 1.49126828e-01
3.87571156e-01 2.15078250e-01 -3.46922874e-01 -1.45638943e+00
-2.47407958e-01 -6.82419956e-01 -4.73197818e-01 -3.88325572e-01
-4.07128543e-01 -7.97377601e-02 1.02211142e+00 -2.21400514e-01
1.06278479e+00 -1.04846954e+00 -4.96152937e-01 2.12579682e-01
-1.26596108e-01 5.45614243e-01 7.95709074e-01 -4.40176576e-01
-7.85602108e-02 -1.96975827e+00 2.06750184e-01 3.03447898e-02
-1.19648919e-01 1.07802242e-01 -3.58803302e-01 8.87263894e-01
-4.80840236e-01 7.99458623e-01 -1.69823706e-01 6.71400845e-01
-6.82974815e-01 4.36066568e-01 -3.98686230e-02 8.59326720e-01
3.05240661e-01 -1.08056262e-01 -1.39687693e+00 3.66772003e-02
1.37550205e-01 4.65287536e-01 -3.48147064e-01 -3.85396242e-01
-1.08968496e-01 2.01962411e-01 -2.27576435e-01 1.31822693e+00
1.50444973e+00 1.45922899e-01 1.05673385e+00 1.34872645e-02
-7.99609303e-01 8.62815008e-02 -7.61853755e-01 1.32146418e+00
-2.77105927e-01 6.36148155e-01 6.85344815e-01 -7.92082250e-01
9.05554533e-01 3.39954287e-01 3.24486762e-01 -1.34853110e-01
-6.92829043e-02 6.61326587e-01 3.07165712e-01 -4.37947631e-01
2.09512815e-01 -2.68229216e-01 8.09878170e-01 3.50690521e-02
-3.68271798e-01 2.61734333e-02 1.61143482e-01 -3.70461009e-02
1.08185947e+00 3.69757265e-01 6.47410691e-01 -1.27341914e+00
3.33308429e-01 2.27002919e-01 6.27376735e-01 1.51456445e-01
-3.70747298e-01 1.49030030e-01 6.92992747e-01 1.36745691e-01
-1.15937841e+00 -9.91443396e-01 -9.72327471e-01 9.82196510e-01
-1.12501480e-01 -1.55862719e-01 -1.53055921e-01 7.13078320e-01
2.38847867e-01 4.61640567e-01 -6.91181660e-01 9.45912674e-02
-1.21901117e-01 -1.23438549e+00 3.04228157e-01 -6.88159317e-02
2.28113264e-01 -8.23449552e-01 -5.76583147e-01 6.24714613e-01
-4.32882458e-01 -6.18508160e-01 6.79527819e-01 8.35443199e-01
-9.26596224e-01 -1.09273100e+00 -6.10779345e-01 -4.97717917e-01
1.80112019e-01 8.23489487e-01 8.19522679e-01 -1.15490280e-01
-7.13995516e-01 -3.94306958e-01 -8.25647831e-01 -5.72858751e-01
-6.60899222e-01 -2.17101142e-01 2.58243471e-01 -8.84286463e-01
5.70231713e-02 -9.92500305e-01 -5.96555233e-01 4.57281709e-01
-3.17087531e-01 -4.07530665e-01 3.52021605e-01 1.06903291e+00
4.23019439e-01 -5.58327317e-01 6.62465572e-01 -6.35365725e-01
-2.00513735e-01 -1.23800778e+00 -1.98613465e-01 2.14700475e-01
-5.21271110e-01 -4.81135309e-01 1.96270883e-01 -2.84308612e-01
-7.05453694e-01 -2.06086710e-01 2.16386572e-01 1.01261353e+00
-2.32749116e-02 9.10757244e-01 -4.43996936e-01 3.63982439e-01
1.01165295e+00 -1.06769241e-01 4.36488628e-01 -1.68426245e-01
-1.87581226e-01 4.10653830e-01 -4.45611635e-03 -6.71444595e-01
1.75712869e-01 5.26901305e-01 2.13724072e-03 -1.50651348e+00
-2.54228801e-01 -5.81247151e-01 -2.80438662e-01 -1.89172924e-01
1.09232426e+00 -1.09369409e+00 -1.46179521e+00 3.75783294e-01
-7.80460954e-01 -4.00591701e-01 2.06115052e-01 7.64043152e-01
-4.62072939e-01 2.40992516e-01 -8.08792949e-01 -8.23242724e-01
-4.19023067e-01 -9.78194356e-01 4.39738542e-01 -2.52628345e-02
-3.42660129e-01 -5.27312100e-01 7.45883286e-01 -3.37622136e-01
5.84541500e-01 9.17304635e-01 1.11649108e+00 1.80899307e-01
-3.76345128e-01 6.45516932e-01 -1.07211910e-01 1.46436855e-01
6.07377887e-01 8.40120971e-01 -1.32647002e+00 -3.81827801e-01
-3.85205611e-03 -8.74595270e-02 7.39338577e-01 4.46628749e-01
4.28805232e-01 2.32051179e-01 -7.05273271e-01 5.61670065e-01
1.94653654e+00 4.88203853e-01 8.11598957e-01 9.68182087e-01
2.94038236e-01 1.34767675e+00 9.78388965e-01 5.77817798e-01
-6.23541236e-01 -9.39161032e-02 1.09020579e+00 3.62151265e-01
3.79990816e-01 4.34056997e-01 9.50329423e-01 4.58474994e-01
-4.44621950e-01 -6.09759092e-01 -1.17366016e+00 9.19390976e-01
-1.46291363e+00 -1.40806055e+00 -7.24250972e-01 2.09326148e+00
7.49373615e-01 -4.64799345e-01 -3.40567946e-01 8.44856650e-02
7.86502361e-01 -3.93659115e-01 -4.84494150e-01 -5.04403412e-01
-7.60361493e-01 5.52973151e-01 7.30462134e-01 2.69322991e-01
-4.04380202e-01 5.51215708e-01 6.57041740e+00 -4.12982047e-01
-1.09016240e+00 -3.13116252e-01 -1.25621587e-01 -1.18284887e-02
-3.32153857e-01 8.73021841e-01 -3.74153197e-01 -1.81009084e-01
1.21167493e+00 -2.23810077e-01 1.85099587e-01 1.15470648e+00
6.44586027e-01 -4.36661422e-01 -5.92320919e-01 -3.20294470e-01
-6.34806097e-01 -1.33132589e+00 -5.80031693e-01 6.12902224e-01
4.90572423e-01 7.93237150e-01 -6.35562420e-01 -5.16662821e-02
2.84554750e-01 -9.19829428e-01 4.97939140e-01 1.57988533e-01
4.13536221e-01 -6.70493484e-01 6.92278624e-01 1.46075621e-01
-1.50640357e+00 -2.19022095e-01 -9.58837509e-01 -2.96504855e-01
1.93276741e-02 1.18349504e+00 -1.39880419e+00 5.14407456e-01
1.22598696e+00 1.03792870e+00 -2.26237774e-01 3.67648810e-01
-2.03333814e-02 4.14893389e-01 -2.75461286e-01 4.72046621e-02
-4.50147897e-01 -4.51642275e-01 4.43001390e-01 1.01386964e+00
8.14244747e-01 -2.46810555e-01 1.92756712e-01 6.05135202e-01
7.21840143e-01 4.43428993e-01 -4.23770726e-01 -5.62478840e-01
1.03485966e+00 1.25514805e+00 -1.14744568e+00 -3.56199026e-01
2.36446992e-01 4.72902954e-01 -2.43308842e-01 1.00964300e-01
-1.10641003e-01 -3.00023079e-01 1.43743932e+00 2.03607440e-01
3.86299282e-01 -5.13044715e-01 -3.26039225e-01 -8.40056777e-01
-7.10362732e-01 -8.31860721e-01 -1.26970246e-01 -7.37573028e-01
-1.10739803e+00 -4.74337935e-02 -2.22855434e-01 -6.95831656e-01
1.90848395e-01 -1.00114107e+00 -5.38772166e-01 1.37689078e+00
-1.19232321e+00 -5.05338252e-01 -8.34270358e-01 -1.38734102e-01
3.95743176e-02 2.50713438e-01 1.58306885e+00 -1.17217332e-01
-7.69132018e-01 -4.55275655e-01 7.95194328e-01 -9.06505048e-01
1.15963209e+00 -8.79150569e-01 4.92555387e-02 3.57028097e-01
-1.27778566e+00 1.14597559e+00 9.96080816e-01 -1.02480614e+00
-1.41232824e+00 -1.03261793e+00 6.55456543e-01 5.35408854e-01
7.63835251e-01 -8.22269842e-02 -9.10456717e-01 5.00732660e-01
5.00091791e-01 -6.88167214e-01 1.12781811e+00 3.18546116e-01
-1.31111935e-01 4.61266637e-01 -1.47490489e+00 7.33270824e-01
7.19995320e-01 -2.86824644e-01 2.72903830e-01 1.01228261e+00
7.80094028e-01 6.09974973e-02 -1.44574666e+00 4.84245807e-01
1.11533058e+00 -6.34150505e-01 1.17259836e+00 -1.57398790e-01
7.55387187e-01 -5.72233140e-01 -7.23466158e-01 -9.32868838e-01
-7.47685254e-01 -6.35591820e-02 5.97477019e-01 1.18740308e+00
-8.09485540e-02 -9.47017848e-01 6.89889789e-02 -6.99700892e-01
-3.68957043e-01 3.52753028e-02 -1.86729103e-01 -8.42720389e-01
3.11510354e-01 6.90473437e-01 4.24404323e-01 1.38147509e+00
2.50123531e-01 2.60056183e-02 4.58004892e-01 1.33274794e-01
4.89510089e-01 -2.05914080e-01 6.89980507e-01 -1.40837121e+00
-3.92755479e-01 -2.40927264e-01 -1.61928698e-01 1.63814932e-01
5.25829829e-02 -1.14722919e+00 3.56283486e-01 -1.62200630e+00
3.32624346e-01 -4.54728633e-01 2.44776011e-01 4.74146575e-01
-2.04759642e-01 -3.90358955e-01 -2.11329192e-01 2.91354865e-01
1.29396951e+00 2.59926409e-01 1.04040122e+00 1.51172370e-01
-2.24834546e-01 -4.34422314e-01 -7.49148071e-01 5.48537195e-01
1.09994674e+00 -5.14990509e-01 -2.79093921e-01 3.41311730e-02
4.63064700e-01 -2.76739569e-03 6.55213892e-02 -5.27775168e-01
-1.03098071e+00 -7.57513940e-01 2.87417203e-01 -5.06719828e-01
5.05534187e-02 -4.56567287e-01 6.25910640e-01 1.72000873e+00
3.29692274e-01 5.31221256e-02 4.80081916e-01 5.25496840e-01
3.34483951e-01 -2.34489173e-01 1.03141832e+00 -4.42355603e-01
-4.21600431e-01 -2.63155431e-01 -1.41882408e+00 -6.50481641e-01
9.22359049e-01 -3.64773929e-01 -8.51102173e-01 4.57472026e-01
-8.95603418e-01 -8.55986848e-02 1.05253398e+00 -1.92840293e-01
-2.42221300e-02 -9.90661263e-01 -8.95553648e-01 -1.96515471e-01
2.55356580e-01 -6.28272712e-01 1.70325413e-01 9.73395407e-01
-1.59715736e+00 9.72607173e-03 -1.15980697e+00 -1.14175236e+00
-1.04401696e+00 6.44366562e-01 2.04136103e-01 1.93063051e-01
-1.40155166e-01 6.72462642e-01 3.29773664e-01 -4.55470204e-01
-6.29440665e-01 -2.95847565e-01 -3.21171016e-01 4.65971500e-01
2.63992935e-01 3.71074259e-01 5.37221283e-02 -3.69914994e-02
-8.51766825e-01 2.18862183e-02 2.71489084e-01 4.77769561e-02
1.61557651e+00 -1.71475023e-01 -1.22526050e+00 9.33061421e-01
8.79774392e-01 -4.49511334e-02 -2.76877791e-01 4.21080202e-01
1.00769050e-01 -7.37410367e-01 -5.99978268e-01 -1.60306215e-01
-4.43617225e-01 4.21204507e-01 1.04270446e+00 1.49508417e-01
9.37509358e-01 -5.39771736e-01 -5.50653040e-02 6.84487939e-01
2.65515447e-01 -8.39617908e-01 4.20283824e-02 7.31353104e-01
9.46584344e-01 -8.80203903e-01 5.41736893e-02 -6.98809266e-01
2.59187877e-01 1.17747045e+00 5.19093096e-01 -1.12426683e-01
7.12629199e-01 3.52351874e-01 -1.71723053e-01 7.50439167e-02
-1.14529586e+00 1.18327633e-01 -1.17117548e+00 8.58952880e-01
1.53969514e+00 5.67309976e-01 -7.86420822e-01 -3.86996031e-01
-4.56977993e-01 -6.21447206e-01 1.26583803e+00 1.51759422e+00
-1.05553913e+00 -1.21604645e+00 -8.22605550e-01 4.86210547e-02
-9.37057659e-03 -4.13484015e-02 -3.44358772e-01 7.35408723e-01
1.08142443e-01 9.92766917e-01 -5.62855937e-02 -1.71674356e-01
1.41930297e-01 3.81244957e-01 3.30530286e-01 -8.64472032e-01
-4.62220639e-01 4.77244139e-01 2.31657386e-01 -3.08163732e-01
-5.47757030e-01 -1.40950215e+00 -1.26635861e+00 -6.91490710e-01
-8.41902316e-01 3.31565663e-02 1.48981667e+00 3.88315827e-01
1.02673741e-02 9.09627020e-01 8.03308189e-01 -9.72106755e-01
-2.63831407e-01 -1.33315992e+00 -1.04778004e+00 -2.71526784e-01
2.71934122e-02 -5.61969638e-01 -4.69818860e-01 2.58016735e-01] | [5.057920455932617, 4.940194606781006] |
c955b1cb-d2b7-400e-8858-280c6a692010 | mus-cdb-mixed-uncertainty-sampling-with-class | 2212.02804 | null | https://arxiv.org/abs/2212.02804v3 | https://arxiv.org/pdf/2212.02804v3.pdf | MUS-CDB: Mixed Uncertainty Sampling with Class Distribution Balancing for Active Annotation in Aerial Object Detection | Recent aerial object detection models rely on a large amount of labeled training data, which requires unaffordable manual labeling costs in large aerial scenes with dense objects. Active learning is effective in reducing the data labeling cost by selectively querying the informative and representative unlabelled samples. However, existing active learning methods are mainly with class-balanced setting and image-based querying for generic object detection tasks, which are less applicable to aerial object detection scenario due to the long-tailed class distribution and dense small objects in aerial scenes. In this paper, we propose a novel active learning method for cost-effective aerial object detection. Specifically, both object-level and image-level informativeness are considered in the object selection to refrain from redundant and myopic querying. Besides, an easy-to-use class-balancing criterion is incorporated to favor the minority objects to alleviate the long-tailed class distribution problem in model training. To fully utilize the queried information, we further devise a training loss to mine the latent knowledge in the undiscovered image regions. Extensive experiments are conducted on the DOTA-v1.0 and DOTA-v2.0 benchmarks to validate the effectiveness of the proposed method. The results show that it can save more than 75% of the labeling cost to reach the same performance compared to the baselines and state-of-the-art active object detection methods. Code is available at \href{https://github.com/ZJW700/MUS-CDB}{\textit{https://github.com/ZJW700/MUS-CDB}}. | ['Sheng-Jun Huang', 'Ying-Peng Tang', 'Jing-Wei Zhang', 'Dong Liang'] | 2022-12-06 | null | null | null | null | ['active-object-detection'] | ['computer-vision'] | [ 2.23816141e-01 5.74160330e-02 -5.12998819e-01 -3.98298979e-01
-9.60969567e-01 -4.32757646e-01 1.28689811e-01 1.89085364e-01
-4.59526211e-01 6.52293682e-01 -3.09300244e-01 -1.42793730e-01
-3.08594674e-01 -8.53178740e-01 -6.05578721e-01 -1.00860655e+00
1.44242585e-01 2.69016832e-01 5.80465078e-01 1.16074085e-01
-2.09695883e-02 2.02941477e-01 -1.56956244e+00 5.58038168e-02
1.07263553e+00 1.24795175e+00 5.62664986e-01 2.57053941e-01
5.85418604e-02 8.02650750e-01 -4.19548154e-01 -1.09992221e-01
5.69966137e-01 -1.24796264e-01 -6.81095839e-01 3.94647658e-01
4.57824349e-01 -7.50889719e-01 -2.02684313e-01 1.24337518e+00
6.03547156e-01 1.74656764e-01 4.73329365e-01 -1.18128610e+00
-2.96285093e-01 2.48391405e-01 -8.57317090e-01 3.71171772e-01
-1.72826424e-01 5.56751668e-01 1.04742956e+00 -1.18179691e+00
2.70785570e-01 8.51904750e-01 3.81415397e-01 2.98434019e-01
-1.18314612e+00 -9.34039235e-01 7.29650617e-01 1.11200370e-01
-1.67021358e+00 -4.77534771e-01 6.55378342e-01 -3.62639248e-01
2.29648784e-01 5.80492079e-01 6.75695419e-01 6.37422681e-01
-4.32430297e-01 1.32460785e+00 6.89860761e-01 -2.45812073e-01
1.62003607e-01 3.03765297e-01 1.85555786e-01 1.07606268e+00
4.88488227e-01 3.78648005e-02 -2.75112063e-01 -2.95969993e-01
6.78873301e-01 4.33951318e-01 -4.21569288e-01 -6.00019574e-01
-8.64039481e-01 8.29789996e-01 7.44131207e-01 -1.39672756e-01
-4.62563634e-01 -7.24115223e-02 2.33778819e-01 -1.56853512e-01
7.00781226e-01 4.18089449e-01 -3.73042673e-01 4.15099084e-01
-9.80389118e-01 2.43215393e-02 2.01790124e-01 1.24817228e+00
9.10203695e-01 -6.04655361e-03 -6.08244836e-01 9.56503272e-01
4.88882661e-01 5.17383933e-01 6.37004524e-02 -9.23914671e-01
4.32497531e-01 1.12003970e+00 3.11610430e-01 -8.83897007e-01
2.12441254e-02 -7.23964989e-01 -6.77267313e-01 3.62769365e-01
1.70533642e-01 -1.15211330e-01 -1.08442259e+00 1.44151235e+00
6.95965230e-01 -3.51160690e-02 -3.23019803e-01 1.11959898e+00
8.96757662e-01 7.81279683e-01 1.29851207e-01 -3.04589152e-01
1.09181964e+00 -1.12791502e+00 -2.77276486e-01 -4.97392058e-01
5.09783089e-01 -6.14875734e-01 1.21102285e+00 1.82397336e-01
-1.05111086e+00 -5.65548658e-01 -1.11403632e+00 4.42827269e-02
-7.03007057e-02 7.84404159e-01 6.87716544e-01 3.46877426e-01
-4.60827947e-01 9.20728743e-02 -8.88121486e-01 1.05031338e-02
1.20450222e+00 4.40594196e-01 -1.02291908e-02 -3.32362026e-01
-7.77454913e-01 9.51041579e-02 5.76357365e-01 2.69063771e-01
-1.39660347e+00 -7.71481991e-01 -6.09833241e-01 -1.81350224e-02
1.09076023e+00 -3.49593043e-01 1.22407770e+00 -1.05193186e+00
-9.09482479e-01 8.16308200e-01 1.59044370e-01 -2.90307164e-01
6.00813329e-01 -3.06463987e-01 6.61279485e-02 1.34568080e-01
2.56382495e-01 9.17013884e-01 7.19519138e-01 -1.18408394e+00
-1.00851846e+00 -4.74970490e-01 4.87367988e-01 3.63202631e-01
-6.47350609e-01 -3.08493137e-01 -8.08120072e-01 -7.72668839e-01
2.00004146e-01 -8.29791963e-01 -2.44457051e-01 7.34799266e-01
-5.24645805e-01 -2.19533324e-01 9.67899859e-01 -2.83676147e-01
1.21430337e+00 -2.27913618e+00 -8.99427533e-02 -1.55281320e-01
1.89769819e-01 5.55735290e-01 -1.32290170e-01 1.02240928e-01
2.61372179e-01 -4.57000397e-02 -4.38275933e-01 -3.31514686e-01
-2.34672070e-01 -9.24765393e-02 -4.16963041e-01 4.53276575e-01
4.21316534e-01 6.65994108e-01 -1.02824020e+00 -8.95055413e-01
3.44030976e-01 1.90715626e-01 -2.71884829e-01 3.74667346e-01
-4.13181096e-01 1.80931166e-01 -8.10018063e-01 1.26060736e+00
6.94936633e-01 -4.40891445e-01 -1.18964657e-01 -3.18154544e-01
-1.06355727e-01 -4.78704087e-02 -1.22054386e+00 1.57366419e+00
-7.22933784e-02 2.91421503e-01 3.00881058e-01 -8.79923642e-01
7.55550921e-01 -5.81479724e-03 4.79742438e-01 -4.40730661e-01
1.40787214e-01 9.70083922e-02 -1.85044676e-01 -3.16128463e-01
2.60022342e-01 3.76742005e-01 1.94293082e-01 6.46246746e-02
-2.21402794e-02 -4.75306995e-02 4.23159868e-01 2.76972264e-01
9.90270257e-01 2.29539081e-01 4.09400314e-01 -3.22725207e-01
3.26334655e-01 3.21175933e-01 8.85487139e-01 7.19948113e-01
-3.31404775e-01 4.43737984e-01 1.02972321e-01 -2.77949989e-01
-3.75660628e-01 -7.96668053e-01 -2.15093702e-01 1.21905792e+00
5.76908290e-01 -2.37300366e-01 -4.80583996e-01 -8.83325756e-01
-1.77787393e-01 6.74583495e-01 -4.71743375e-01 -2.57899463e-01
-1.57696858e-01 -1.07834268e+00 2.36920387e-01 4.69533443e-01
7.12221026e-01 -9.75925565e-01 -8.09678197e-01 -7.29072765e-02
-2.08109483e-01 -6.90391302e-01 -2.95030385e-01 2.68938243e-01
-9.76797700e-01 -1.12178075e+00 -8.60908329e-01 -5.84168136e-01
9.24718618e-01 7.97382057e-01 9.27885771e-01 3.63060534e-01
-7.51023233e-01 1.71541259e-01 -4.37288016e-01 -7.82465875e-01
2.81093538e-01 1.92509621e-01 -2.71410525e-01 4.42444086e-02
2.52498627e-01 -7.13809952e-02 -9.34791327e-01 6.53875887e-01
-9.73913968e-01 1.26328655e-02 7.42103755e-01 8.06973696e-01
9.76004064e-01 1.33285120e-01 3.18981647e-01 -8.34727824e-01
-1.42147675e-01 -3.68395627e-01 -1.03472340e+00 4.46467429e-01
-5.39379239e-01 -3.90576482e-01 1.90854207e-01 -5.66319227e-01
-9.13100302e-01 5.16560316e-01 3.09916556e-01 -4.88419026e-01
1.09386155e-02 2.57359028e-01 -4.36314732e-01 -8.25064331e-02
5.42723894e-01 1.20504417e-01 -3.33373398e-01 -2.14004681e-01
-1.24134175e-01 5.79470515e-01 1.66120246e-01 -3.04271311e-01
8.60525548e-01 5.59703290e-01 -4.02334511e-01 -6.32826209e-01
-1.26040184e+00 -8.03223193e-01 -3.98138076e-01 -2.79066205e-01
4.08715636e-01 -1.28134656e+00 -2.86909759e-01 3.70993137e-01
-6.75161719e-01 -5.44231474e-01 -6.16385937e-01 5.87429404e-01
-2.29762331e-01 2.65017569e-01 -2.28215322e-01 -1.13139498e+00
-5.33259332e-01 -1.16685295e+00 1.19502187e+00 4.50511694e-01
1.97964802e-01 -3.90006274e-01 -2.13795245e-01 6.01607621e-01
8.44370350e-02 -8.37044418e-02 6.08876467e-01 -4.94615883e-01
-1.20302296e+00 -4.22236651e-01 -3.84872615e-01 4.10354465e-01
1.84852436e-01 -6.72072917e-02 -9.40207839e-01 -5.39748490e-01
-2.65720487e-01 -7.44117498e-01 1.17451119e+00 3.84501219e-01
1.48118424e+00 -1.69981822e-01 -4.77986574e-01 5.04777133e-01
1.25210929e+00 2.41588533e-01 4.80433941e-01 -6.52632043e-02
5.38599789e-01 5.83010197e-01 1.47105289e+00 6.47498369e-01
-4.86711077e-02 5.86268842e-01 9.07583654e-01 -4.62698191e-01
-1.12218969e-01 -1.08151019e-01 1.42730355e-01 9.24553350e-02
-9.31393076e-03 -5.31104982e-01 -8.14857244e-01 5.69019556e-01
-1.83969545e+00 -8.03148150e-01 1.74163561e-02 2.47865033e+00
1.00938690e+00 3.27210426e-01 1.54077098e-01 1.48523375e-01
6.94943130e-01 2.52087355e-01 -9.53285038e-01 8.33742976e-01
-2.86342553e-03 -3.05103093e-01 6.00813389e-01 2.61087179e-01
-1.56300068e+00 8.84519279e-01 4.00309277e+00 1.24983478e+00
-9.42627788e-01 1.38895288e-01 7.80432403e-01 -4.01746064e-01
3.44679207e-01 2.58979518e-02 -1.15544200e+00 5.28009295e-01
9.68264714e-02 1.35792524e-01 -8.28141123e-02 1.17969811e+00
7.14347586e-02 -3.52306485e-01 -7.61530519e-01 1.00022840e+00
-1.14718147e-01 -1.11031640e+00 -1.58919603e-01 3.51364426e-02
6.41249716e-01 -2.81945989e-02 -1.67456493e-02 3.27768803e-01
2.36526147e-01 -5.37675560e-01 8.20940554e-01 2.82555461e-01
8.12763095e-01 -4.07801479e-01 6.63933516e-01 4.71748561e-01
-1.40670907e+00 -4.93935555e-01 -6.69978559e-01 2.25878134e-01
-1.07915793e-03 6.82341814e-01 -6.51386142e-01 3.90602410e-01
8.89483929e-01 6.17697418e-01 -7.79003799e-01 1.37615395e+00
-1.45614624e-01 8.54340971e-01 -3.65757734e-01 4.56796065e-02
2.52922863e-01 -4.62374203e-02 5.70458710e-01 7.30012417e-01
1.78485453e-01 4.22028959e-01 6.90011144e-01 8.17882836e-01
-1.71132773e-01 5.58479540e-02 -2.56290168e-01 -2.08133593e-01
6.78608954e-01 1.52037203e+00 -1.00199461e+00 -2.76423633e-01
-2.51868814e-01 8.10690463e-01 1.58118933e-01 2.40998000e-01
-9.60135639e-01 -3.11389923e-01 1.41855314e-01 4.75205064e-01
3.50373834e-01 5.00598624e-02 1.52420223e-01 -9.06999350e-01
8.95675793e-02 -6.21469319e-01 5.47030091e-01 -7.28863597e-01
-1.10711133e+00 4.29274112e-01 2.77482897e-01 -1.42631209e+00
3.65330309e-01 -5.01197994e-01 -4.81146693e-01 5.69339097e-01
-1.31742573e+00 -1.22945666e+00 -9.09823000e-01 4.18623418e-01
9.35582399e-01 -2.15939552e-01 5.73315680e-01 5.68774998e-01
-8.25742245e-01 4.56406862e-01 -1.40261754e-01 2.51075149e-01
6.03456259e-01 -9.66052413e-01 -6.54628351e-02 8.86943042e-01
1.22629926e-01 2.89924383e-01 2.70853013e-01 -6.34690404e-01
-1.11141920e+00 -1.48489130e+00 9.06835049e-02 -1.90567523e-01
1.77983180e-01 -4.00368124e-01 -7.74164617e-01 5.03741205e-01
-3.02486122e-01 4.76856917e-01 5.33527493e-01 -2.48250067e-01
-5.35177328e-02 -4.48065013e-01 -9.76124167e-01 3.03005219e-01
1.05309427e+00 -1.57672465e-01 -8.12646449e-02 6.62825227e-01
7.26687908e-01 -1.95607707e-01 -3.09275448e-01 8.77860188e-01
2.26656109e-01 -7.75375843e-01 1.14764845e+00 -2.65143961e-01
1.25842258e-01 -5.75657606e-01 -3.06697935e-01 -6.45981014e-01
-4.20786262e-01 -1.53025314e-01 -7.28807226e-02 1.40043759e+00
4.72340405e-01 -3.97051007e-01 8.18608224e-01 3.64283651e-01
-2.68551171e-01 -1.07933438e+00 -6.76174223e-01 -6.96080565e-01
-5.71760118e-01 8.23503584e-02 2.24887997e-01 7.18883693e-01
-8.14611435e-01 3.10236961e-01 -2.54183322e-01 3.62782657e-01
6.81954563e-01 3.74798387e-01 7.05528319e-01 -1.29660678e+00
-3.15362036e-01 3.47521761e-03 -2.02487573e-01 -1.14054811e+00
-2.92649806e-01 -5.39276719e-01 3.20317000e-01 -1.55832374e+00
4.07088369e-01 -9.21614349e-01 -3.13051492e-01 7.68330693e-01
-3.93382251e-01 4.05775875e-01 1.80076241e-01 5.12525797e-01
-1.02236271e+00 7.82709360e-01 8.70602548e-01 -4.55290377e-01
-3.12761396e-01 2.89201975e-01 -5.50775826e-01 8.01349699e-01
8.55987847e-01 -6.93069875e-01 -7.97255814e-01 -5.27335763e-01
1.56294499e-02 -2.18393803e-01 7.99327910e-01 -1.01273715e+00
2.99626946e-01 -3.28568041e-01 5.68385065e-01 -9.12892520e-01
6.75520897e-01 -8.62919867e-01 -9.01745260e-02 5.14643967e-01
-3.46392125e-01 -4.51811641e-01 1.53147936e-01 7.82777548e-01
-1.05417557e-01 -4.58209783e-01 1.04753709e+00 -3.41389716e-01
-7.55407810e-01 7.38760531e-01 4.28646877e-02 8.49899426e-02
1.35412383e+00 -2.27166697e-01 -3.10968071e-01 7.99055547e-02
-2.79152691e-01 5.15279472e-01 4.61977780e-01 1.94084436e-01
5.26230991e-01 -9.51512575e-01 -6.40118837e-01 8.37429464e-02
4.75675225e-01 6.58199489e-01 5.01897037e-01 8.79616737e-01
-4.35889870e-01 2.26610601e-01 3.69502306e-02 -6.54477179e-01
-1.52487481e+00 4.89065588e-01 2.98990995e-01 -1.17662564e-01
-2.97748178e-01 1.16004789e+00 5.71874499e-01 -3.29017341e-01
5.76870382e-01 -1.18911006e-02 4.63190898e-02 2.68265277e-01
4.54040259e-01 5.02552629e-01 3.47648039e-02 -1.83990911e-01
-4.01452988e-01 2.16917396e-01 -2.10388064e-01 4.14935201e-01
1.15171444e+00 -1.39009133e-01 1.77758783e-01 2.74087936e-01
6.77068293e-01 -1.61598518e-01 -1.49087346e+00 -4.66598064e-01
-2.73269713e-01 -7.01305151e-01 4.57260221e-01 -8.36391807e-01
-1.17838252e+00 8.91925037e-01 1.00389111e+00 4.24985727e-03
1.31404030e+00 2.78197322e-02 3.84232432e-01 6.37964308e-01
3.85478050e-01 -1.02941382e+00 3.17073524e-01 -9.47940350e-02
8.60780180e-01 -1.54649818e+00 5.74380517e-01 -6.93230450e-01
-5.72873771e-01 6.02724433e-01 9.62417305e-01 6.85656667e-02
4.74723011e-01 -1.98483896e-02 -7.39870295e-02 -3.45150590e-01
-5.39902031e-01 -3.19462895e-01 2.49706894e-01 1.64982989e-01
3.80864255e-02 1.01655805e-02 9.00179148e-03 4.22391713e-01
6.21222198e-01 -4.73284423e-02 1.86454486e-02 1.22075927e+00
-6.64315045e-01 -6.93580747e-01 -2.85762131e-01 6.53473377e-01
-1.83574393e-01 -8.83854181e-02 -5.34237623e-01 8.03668380e-01
2.86854893e-01 9.22012210e-01 -8.59961361e-02 -5.96629940e-02
2.08309293e-01 -3.26656342e-01 2.79141635e-01 -9.49152768e-01
-2.32094452e-01 1.69090390e-01 -1.53579637e-01 -4.80077684e-01
-5.28739154e-01 -4.12481815e-01 -9.75516915e-01 2.29035944e-01
-9.82705653e-01 -2.00929809e-02 1.53413489e-01 3.71973604e-01
4.18142557e-01 3.77993524e-01 5.28662562e-01 -6.87508285e-01
-7.78263986e-01 -9.46194112e-01 -5.45967519e-01 -1.40346810e-01
9.47210640e-02 -9.07002926e-01 -3.56580824e-01 4.99115065e-02] | [9.212244033813477, 1.0190373659133911] |
115d9989-d728-4b8f-b83e-3cd7c6ea45e1 | shapestacks-learning-vision-based-physical | 1804.08018 | null | http://arxiv.org/abs/1804.08018v2 | http://arxiv.org/pdf/1804.08018v2.pdf | ShapeStacks: Learning Vision-Based Physical Intuition for Generalised Object Stacking | Physical intuition is pivotal for intelligent agents to perform complex
tasks. In this paper we investigate the passive acquisition of an intuitive
understanding of physical principles as well as the active utilisation of this
intuition in the context of generalised object stacking. To this end, we
provide: a simulation-based dataset featuring 20,000 stack configurations
composed of a variety of elementary geometric primitives richly annotated
regarding semantics and structural stability. We train visual classifiers for
binary stability prediction on the ShapeStacks data and scrutinise their
learned physical intuition. Due to the richness of the training data our
approach also generalises favourably to real-world scenarios achieving
state-of-the-art stability prediction on a publicly available benchmark of
block towers. We then leverage the physical intuition learned by our model to
actively construct stable stacks and observe the emergence of an intuitive
notion of stackability - an inherent object affordance - induced by the active
stacking task. Our approach performs well even in challenging conditions where
it considerably exceeds the stack height observed during training or in cases
where initially unstable structures must be stabilised via counterbalancing. | ['Ingmar Posner', 'Andrea Vedaldi', 'Oliver Groth', 'Fabian B. Fuchs'] | 2018-04-21 | shapestacks-learning-vision-based-physical-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Oliver_Groth_ShapeStacks_Learning_Vision-Based_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Oliver_Groth_ShapeStacks_Learning_Vision-Based_ECCV_2018_paper.pdf | eccv-2018-9 | ['physical-intuition'] | ['reasoning'] | [-1.67070925e-02 5.50889552e-01 1.26285836e-01 -6.91090673e-02
-2.41355091e-01 -1.05056906e+00 1.14289725e+00 4.82362658e-01
-6.09273724e-02 3.49326253e-01 2.63593405e-01 -5.33456147e-01
-3.87323648e-01 -7.48572767e-01 -1.08708096e+00 -7.03922093e-01
-6.12117231e-01 8.13427448e-01 7.69902349e-01 -7.57048070e-01
5.17155528e-01 4.31036651e-01 -2.14869189e+00 2.26911038e-01
9.20220792e-01 9.84237969e-01 5.42160988e-01 7.82148004e-01
2.24892125e-01 6.14970863e-01 -5.61983213e-02 -1.38298899e-01
4.07978386e-01 3.67557108e-01 -8.65480602e-01 1.17856734e-01
5.09091079e-01 1.99183822e-01 1.58764929e-01 5.88468075e-01
1.20206647e-01 1.71803348e-02 8.44772637e-01 -1.03656375e+00
-6.53054178e-01 3.75513226e-01 -1.76857829e-01 2.79076695e-01
3.74613792e-01 7.28339851e-01 1.43545485e+00 -6.12195075e-01
7.82392800e-01 1.09073293e+00 6.97575271e-01 3.08300465e-01
-1.41526163e+00 1.21157587e-01 3.11259091e-01 1.00481600e-01
-7.91593373e-01 -3.09437871e-01 5.23814857e-01 -8.21288347e-01
1.11621594e+00 4.08557534e-01 9.39902186e-01 8.75173986e-01
7.14198053e-02 7.26058424e-01 1.29615712e+00 -3.73797983e-01
5.74241340e-01 -2.12623209e-01 9.84310955e-02 9.69307303e-01
2.84146011e-01 3.95756036e-01 -7.49105036e-01 7.78604820e-02
7.20404088e-01 -4.82842475e-01 -1.75658152e-01 -1.16926849e+00
-1.28524697e+00 2.33802333e-01 8.76433372e-01 1.27733307e-04
-6.77490160e-02 2.56434768e-01 3.66482973e-01 1.40070273e-02
2.08875234e-03 8.98745537e-01 -6.57037139e-01 -2.43325219e-01
-5.52320004e-01 7.02038944e-01 5.44860303e-01 6.78554654e-01
5.64845264e-01 -1.97300896e-01 2.67558873e-01 2.83967316e-01
4.22996312e-01 2.13787466e-01 2.93012798e-01 -9.44778383e-01
4.64706570e-01 8.43762457e-01 2.22656220e-01 -5.31793833e-01
-6.10490441e-01 -4.15163398e-01 -2.89083630e-01 9.15161252e-01
6.31636620e-01 4.90574330e-01 -1.09998178e+00 1.66909456e+00
2.74449885e-01 -3.77624854e-02 -3.73251140e-02 8.42766166e-01
4.36916322e-01 2.08659187e-01 1.76785395e-01 3.00835311e-01
1.22958565e+00 -7.31231570e-01 1.98649257e-01 -3.82443517e-01
6.55254185e-01 -2.83605993e-01 1.48240399e+00 3.24306160e-01
-1.20902634e+00 -3.07778448e-01 -1.24334133e+00 6.21448504e-03
-3.95604908e-01 -1.61441281e-01 9.99819219e-01 5.24861038e-01
-1.12620068e+00 1.06574130e+00 -9.78121281e-01 -3.80167186e-01
5.20893395e-01 3.59601229e-01 -3.42934012e-01 6.02631152e-01
-5.97795725e-01 1.19459522e+00 5.65962255e-01 -9.68832597e-02
-7.14322865e-01 -9.07821298e-01 -8.13432157e-01 -6.05197176e-02
6.24328256e-01 -9.18698728e-01 1.46216691e+00 -7.16390193e-01
-1.44450629e+00 1.22684944e+00 3.60859156e-01 -4.59427655e-01
6.76304221e-01 -3.70464802e-01 1.47210881e-01 -1.97903693e-01
-1.02629162e-01 4.56543744e-01 6.48056686e-01 -1.66258061e+00
-1.02559999e-01 -3.43454391e-01 3.81671906e-01 9.42525268e-02
3.72861028e-01 -5.40874124e-01 1.75204530e-01 -4.44091231e-01
3.33555549e-01 -1.07201362e+00 -2.10279524e-01 1.09221801e-01
-4.81251091e-01 -7.04815015e-02 6.49746537e-01 -3.30610335e-01
8.37431967e-01 -1.79678941e+00 7.20301330e-01 4.17729497e-01
4.48997647e-01 2.95011848e-01 2.57433206e-01 5.42159677e-01
-2.93408632e-01 4.06841487e-02 -3.30915600e-01 -8.65930468e-02
3.66581380e-01 2.13444531e-01 -4.67219859e-01 2.40842521e-01
2.91283101e-01 1.17388988e+00 -7.80750930e-01 -2.26674572e-01
4.31746483e-01 -4.56854515e-02 -6.82999551e-01 6.00926578e-02
-7.93718815e-01 5.04381001e-01 -5.26041269e-01 5.26885152e-01
4.38082725e-01 -5.37109792e-01 2.83267587e-01 3.07290964e-02
-4.09690022e-01 4.77120727e-01 -9.86257851e-01 1.52050936e+00
-2.05879703e-01 4.14792061e-01 -1.13527488e-03 -7.41410613e-01
5.59745550e-01 -2.43564188e-01 -3.73872705e-02 -6.28471613e-01
1.52217060e-01 2.93624938e-01 2.65058488e-01 -4.87400442e-01
2.93526739e-01 -4.60713595e-01 -7.60694444e-02 2.51477212e-01
-1.96233720e-01 -8.36814046e-01 -9.03865919e-02 2.47103602e-01
1.16033316e+00 5.36948860e-01 4.14113373e-01 -9.47044730e-01
4.27017450e-01 -6.42105341e-02 -7.87280351e-02 4.79229838e-01
6.13919050e-02 4.03236538e-01 4.90480691e-01 -6.09525859e-01
-1.49478734e+00 -1.66011953e+00 -3.49801779e-02 1.20316899e+00
4.30192709e-01 -7.70540178e-01 -6.77356660e-01 -1.79384559e-01
2.06267580e-01 5.78126252e-01 -9.98724520e-01 -2.94544250e-01
-5.49830079e-01 -5.20692170e-01 5.38502559e-02 6.08885884e-01
1.31636053e-01 -1.14718723e+00 -1.46644175e+00 -7.69452006e-02
3.44773501e-01 -8.57813954e-01 3.62950027e-01 1.45290136e-01
-8.45130026e-01 -1.20154214e+00 -1.95599496e-01 -3.24672103e-01
3.34110498e-01 -1.26700491e-01 1.48220718e+00 7.14706302e-01
-4.14460152e-01 2.47699350e-01 -3.16575356e-02 -4.30124521e-01
-5.49776673e-01 1.40279159e-01 2.29702488e-01 -5.42173028e-01
-4.19931829e-01 -9.33386505e-01 -7.02251256e-01 2.17941582e-01
-6.71220958e-01 4.53109354e-01 4.97838080e-01 3.05865496e-01
5.98899871e-02 -1.20734170e-01 -1.24018081e-01 -2.73515165e-01
5.11478782e-01 -3.06364745e-01 -5.77239454e-01 3.89051795e-01
-2.37453520e-01 6.05739236e-01 3.28748018e-01 -1.36028364e-01
-1.05188119e+00 -1.72945246e-01 1.81859881e-01 9.77041423e-02
-1.84375346e-01 2.02606097e-01 -2.27849349e-01 -5.44546843e-02
8.36323380e-01 -1.29744798e-01 -2.41267476e-02 -3.48540932e-01
6.07445598e-01 1.20343544e-01 5.88579655e-01 -1.21376443e+00
1.01011431e+00 5.49396634e-01 4.73420918e-01 -7.95607865e-01
-8.29399109e-01 -1.76097509e-02 -8.98666620e-01 -1.85861826e-01
8.19633722e-01 -5.44097602e-01 -1.19431841e+00 7.06642687e-01
-6.87511384e-01 -7.44115710e-01 -3.39556634e-01 -9.87145677e-02
-1.16416121e+00 4.23808277e-01 -3.48860860e-01 -8.89414072e-01
1.01073697e-01 -1.13888776e+00 1.39430940e+00 5.90303317e-02
-4.76456881e-01 -9.84178483e-01 3.64598572e-01 3.04854691e-01
1.55055866e-01 5.60823083e-01 1.21637738e+00 -2.54401773e-01
-1.16262269e+00 3.41946810e-01 -4.34836000e-02 -3.76121908e-01
-1.65653720e-01 5.09102643e-01 -9.12906885e-01 -8.50914717e-02
-2.87013322e-01 -6.25563741e-01 9.22218978e-01 1.05050020e-01
9.80300188e-01 1.14382133e-02 -2.93194950e-01 1.83844641e-01
1.31962800e+00 -4.13957655e-01 6.92635894e-01 8.77531350e-01
5.15852571e-01 8.37052345e-01 3.56336713e-01 2.60777086e-01
2.75032848e-01 9.50514853e-01 8.47108185e-01 2.63864696e-01
2.98644662e-01 -2.91981876e-01 1.19909599e-01 4.15266335e-01
-5.71885407e-01 -6.00881390e-02 -1.49439681e+00 2.75374502e-01
-1.96436644e+00 -8.81305456e-01 -2.84850657e-01 2.12954426e+00
6.29199147e-01 8.53024125e-01 2.82757193e-01 2.19978109e-01
2.23480552e-01 2.48309612e-01 -3.85625154e-01 -2.77311802e-01
-4.63383645e-02 1.15345651e-03 3.56544703e-01 4.92904603e-01
-9.14067686e-01 8.59862447e-01 6.51081610e+00 5.06262898e-01
-9.63470757e-01 -2.04745308e-01 3.57106835e-01 -1.23768605e-01
-3.19481015e-01 1.70868605e-01 -3.01777840e-01 4.41806316e-01
6.27714097e-01 3.93560946e-01 4.53725576e-01 5.87544918e-01
-8.96596834e-02 -5.10593474e-01 -1.33630097e+00 4.39350247e-01
-4.17690873e-01 -1.56653762e+00 -5.49656153e-02 1.32309049e-01
3.02167833e-01 6.17494099e-02 3.47465873e-01 -3.14557999e-02
6.35551095e-01 -1.38329971e+00 1.60481131e+00 7.44298756e-01
3.62218320e-01 -2.14290053e-01 7.26245195e-02 6.88626349e-01
-1.26243317e+00 -3.62749785e-01 2.42793694e-01 -5.38217425e-01
2.25129992e-01 2.13334218e-01 -6.81918919e-01 5.52870810e-01
7.22280383e-01 4.33548093e-01 -1.00254548e+00 8.36674988e-01
-1.45564079e-01 1.50458619e-01 -7.60462761e-01 8.45338851e-02
4.41941768e-01 -1.69748634e-01 6.04437172e-01 7.82334864e-01
-3.22755314e-02 1.78373773e-02 -1.86930940e-01 9.59053814e-01
5.09157836e-01 -4.46519703e-01 -5.68582118e-01 1.09762080e-01
1.02057487e-01 9.37933981e-01 -1.05427611e+00 -1.17369935e-01
1.53442577e-01 4.37268108e-01 8.49331617e-01 -2.22855080e-02
-7.40636587e-01 6.14910722e-01 5.90928197e-01 4.54762250e-01
5.23879051e-01 -5.17919362e-01 -6.68151021e-01 -9.65334535e-01
4.07566667e-01 -3.81401688e-01 -9.11252797e-02 -1.28477132e+00
-8.55767369e-01 1.50277510e-01 2.09770203e-01 -7.38782465e-01
-1.00193396e-02 -1.14458525e+00 -1.10535574e+00 4.73377585e-01
-9.32300448e-01 -1.44869220e+00 -2.05573171e-01 -8.66067708e-02
3.09527725e-01 5.02339564e-02 5.04728258e-01 -5.33926368e-01
-1.66382432e-01 8.44671875e-02 -1.82296487e-03 -5.62138557e-01
-2.19523124e-02 -1.66115999e+00 8.04377258e-01 5.25532544e-01
1.62531197e-01 7.65840590e-01 1.48880935e+00 -6.89492047e-01
-1.16897929e+00 -4.47550744e-01 1.88827410e-01 -1.37669861e+00
9.99509275e-01 -7.14674592e-01 -1.07872701e+00 6.76001132e-01
6.94522336e-02 -7.53998682e-02 3.49772394e-01 2.64457971e-01
-4.37488347e-01 5.19454181e-01 -8.06139231e-01 8.55452955e-01
1.71192038e+00 -3.77783388e-01 -1.04899502e+00 1.14103630e-01
4.89590704e-01 -4.18697298e-01 -5.76528549e-01 6.79898024e-01
5.55501044e-01 -1.47009897e+00 1.26487160e+00 -8.99891973e-01
6.50845289e-01 -3.83921295e-01 -1.59190565e-01 -1.15906692e+00
-2.49462306e-01 -6.72108531e-01 -2.09558576e-01 6.82907939e-01
4.44710821e-01 -3.16060394e-01 7.66630709e-01 6.90519154e-01
-3.15030992e-01 -1.15794802e+00 -9.77592111e-01 -6.43791139e-01
2.42512375e-01 -3.27240109e-01 7.03420937e-01 5.73498607e-01
1.90655991e-01 2.31756702e-01 3.77875268e-01 1.63801730e-01
6.16722167e-01 1.43687740e-01 7.50838816e-01 -1.52816689e+00
-5.46364725e-01 -7.71537840e-01 -6.06238425e-01 -8.25273156e-01
8.91334787e-02 -7.12412953e-01 -7.74142146e-02 -1.07465458e+00
5.28168753e-02 -6.32538974e-01 -6.08144607e-03 9.37437862e-02
-3.07833613e-03 -1.92359716e-01 4.11620408e-01 2.12529317e-01
-7.98439085e-01 7.48217940e-01 1.17397499e+00 3.51616442e-01
-6.22276589e-02 -2.52883375e-01 -6.28076553e-01 1.07653546e+00
6.07977867e-01 2.72409171e-01 -5.05721807e-01 -2.79923856e-01
8.36316705e-01 -2.20331937e-01 9.77352738e-01 -9.89428461e-01
-2.03189820e-01 -2.79599726e-01 1.77125320e-01 -4.68788922e-01
5.44145346e-01 -7.53149569e-01 1.95877045e-01 4.69673157e-01
-2.09837735e-01 5.59325621e-04 3.84788871e-01 6.32096708e-01
5.84461689e-01 -1.02829516e-01 6.12901747e-01 -3.67802829e-01
-8.87411952e-01 -1.94549426e-01 -2.55100578e-01 1.32869303e-01
1.14886844e+00 -5.06711721e-01 -5.28320312e-01 1.89481571e-01
-9.30883169e-01 1.50639310e-01 1.31714523e+00 4.01103377e-01
2.81847060e-01 -9.77553606e-01 -1.55071557e-01 3.29865903e-01
3.41386259e-01 3.46225709e-01 1.89147852e-02 5.92729211e-01
-8.27020407e-01 1.68962866e-01 -4.87960249e-01 -8.73091936e-01
-1.11094737e+00 7.08656609e-01 3.80859405e-01 -2.14767575e-01
-7.57106841e-01 1.02957237e+00 7.16512561e-01 -5.50620556e-01
-3.57233167e-01 -9.06735182e-01 -1.40413782e-03 -3.41209680e-01
1.97800645e-03 2.17648119e-01 1.43976301e-01 -6.35334969e-01
-2.84498185e-01 7.12948740e-01 1.21938691e-01 -1.11763708e-01
1.53250253e+00 1.39869852e-02 3.03994380e-02 5.06012440e-01
4.75624859e-01 -2.34707166e-02 -1.83073580e+00 1.92016110e-01
4.92542803e-01 -3.57599288e-01 -4.96130764e-01 -7.77928054e-01
-3.31881255e-01 8.01594794e-01 3.47173512e-01 3.90372813e-01
4.38841701e-01 6.99146211e-01 2.89444029e-01 4.63352382e-01
5.55414200e-01 -1.03398287e+00 4.19212878e-01 5.23343801e-01
1.09144545e+00 -1.15623665e+00 1.65466294e-01 -6.69290543e-01
-5.17711282e-01 8.64770710e-01 6.42574251e-01 -4.81108069e-01
4.02118236e-01 2.03345537e-01 -3.16644371e-01 -8.01643848e-01
-9.04835701e-01 -1.02467217e-01 3.69006038e-01 6.87061250e-01
-1.18903052e-02 1.33929431e-01 4.68389630e-01 3.00485522e-01
-9.30469155e-01 -3.88656020e-01 3.76521677e-01 1.25577819e+00
-8.51375222e-01 -8.32295299e-01 -2.66091228e-01 2.90325195e-01
9.89969745e-02 1.72858909e-01 -7.01121449e-01 7.12500215e-01
1.87329531e-01 1.56956628e-01 9.39078182e-02 -9.49499533e-02
3.83651316e-01 3.04412749e-02 1.02372563e+00 -6.72584057e-01
-3.67907405e-01 -5.70161939e-01 2.48377889e-01 -5.27937889e-01
-3.14165741e-01 -8.79851222e-01 -1.38128793e+00 -3.36643398e-01
-5.32214120e-02 -2.41860703e-01 4.37942833e-01 1.15829623e+00
1.59124807e-01 4.08119053e-01 -9.44223851e-02 -1.58049703e+00
-5.49075425e-01 -4.35429424e-01 -3.45698953e-01 7.31082022e-01
5.31585753e-01 -1.40449536e+00 -1.95246816e-01 2.62969546e-02] | [8.406485557556152, 0.9822599291801453] |
d01f8335-aec9-45d8-8876-f7419f83c349 | abstractive-text-summarization-using | 2302.13117 | null | https://arxiv.org/abs/2302.13117v1 | https://arxiv.org/pdf/2302.13117v1.pdf | Abstractive Text Summarization using Attentive GRU based Encoder-Decoder | In todays era huge volume of information exists everywhere. Therefore, it is very crucial to evaluate that information and extract useful, and often summarized, information out of it so that it may be used for relevant purposes. This extraction can be achieved through a crucial technique of artificial intelligence, namely, machine learning. Indeed automatic text summarization has emerged as an important application of machine learning in text processing. In this paper, an english text summarizer has been built with GRU-based encoder and decoder. Bahdanau attention mechanism has been added to overcome the problem of handling long sequences in the input text. A news-summary dataset has been used to train the model. The output is observed to outperform competitive models in the literature. The generated summary can be used as a newspaper headline. | ['Samiran Chattopadhyay', 'Debarshi Kumar Sanyal', 'Suchandan Das', 'Tohida Rehman'] | 2023-02-25 | null | null | null | null | ['abstractive-text-summarization'] | ['natural-language-processing'] | [ 4.51257735e-01 5.03461957e-01 -2.77510226e-01 -2.09979460e-01
-1.03249216e+00 -4.25432533e-01 7.74300039e-01 8.44030559e-01
-5.27827263e-01 1.33319771e+00 1.08757591e+00 -2.18778029e-01
2.18101069e-01 -4.97313231e-01 -4.35357422e-01 -2.98154473e-01
3.97111773e-01 3.07356447e-01 1.35228887e-01 -5.11183441e-01
9.42758501e-01 6.22900613e-02 -1.41279399e+00 6.19882822e-01
1.11494589e+00 7.02607512e-01 6.90996766e-01 9.31260407e-01
-6.48195028e-01 1.04867756e+00 -9.90552127e-01 -4.30767328e-01
-1.92714602e-01 -8.69487524e-01 -1.01574659e+00 8.91457424e-02
9.47276428e-02 -1.61143895e-02 -1.58811823e-01 1.03728437e+00
4.41139221e-01 2.98241079e-01 1.00977921e+00 -2.89325535e-01
-3.87510985e-01 9.98729050e-01 -5.06114721e-01 3.88622165e-01
5.12929559e-01 -5.10050178e-01 1.07118392e+00 -7.70571411e-01
6.52750969e-01 9.22118843e-01 8.55341181e-03 3.49863082e-01
-6.95817530e-01 -2.08997261e-02 -2.75635540e-01 2.33216703e-01
-6.72386944e-01 -5.12731135e-01 9.80582833e-01 -1.08723126e-01
1.25808752e+00 3.44545960e-01 6.09767616e-01 9.38649058e-01
8.45046520e-01 9.84164894e-01 6.29824519e-01 -7.72138298e-01
1.84474483e-01 2.56638765e-01 5.41824996e-01 4.30786520e-01
5.49452245e-01 -6.94193602e-01 -6.03665531e-01 2.31594056e-01
-5.88358007e-02 -3.43503326e-01 -1.71909869e-01 6.29946589e-01
-7.45276392e-01 8.90830457e-01 2.94202566e-01 5.48045814e-01
-8.15218151e-01 -1.54191419e-01 9.46826696e-01 5.08313179e-01
8.01528096e-01 7.25138426e-01 -2.44215310e-01 -4.29248214e-01
-1.18583012e+00 4.51234244e-02 9.81705666e-01 8.00085723e-01
2.90828198e-01 1.87863842e-01 -2.23613098e-01 7.91295409e-01
4.67136912e-02 1.87674180e-01 9.52276647e-01 -4.36642200e-01
9.13930178e-01 9.50043499e-01 8.56262632e-03 -1.04430091e+00
-2.07403764e-01 -5.77044964e-01 -1.01167667e+00 -2.57422417e-01
-4.03978944e-01 -3.58533442e-01 -8.63411486e-01 9.64808702e-01
-4.55624051e-02 -3.95027518e-01 3.80102336e-01 4.34491873e-01
1.07131195e+00 1.30414832e+00 -9.29077342e-02 -5.86088955e-01
1.23675442e+00 -1.07593834e+00 -9.49067593e-01 -2.42785811e-01
4.69356179e-01 -1.01475561e+00 6.10802770e-01 4.66785938e-01
-1.03630245e+00 -5.09529173e-01 -1.29710019e+00 -4.45597976e-01
-4.41594273e-01 3.46480906e-01 2.69512713e-01 2.31486946e-01
-7.84817755e-01 5.85236430e-01 -4.29398149e-01 -5.37972510e-01
3.22497934e-01 2.10099310e-01 -3.52765918e-01 1.39769256e-01
-1.09166563e+00 1.15013707e+00 1.09592819e+00 -5.72915040e-02
-2.97893494e-01 -1.05697706e-01 -6.52808964e-01 3.00564378e-01
2.67098159e-01 -7.52164543e-01 1.33295143e+00 -1.04651499e+00
-1.37676132e+00 5.73168218e-01 -3.31292689e-01 -9.15383875e-01
3.54795367e-01 -3.98502856e-01 -4.12787169e-01 3.65205824e-01
1.30313233e-01 1.66548997e-01 7.81859159e-01 -8.70214820e-01
-8.97162557e-01 -2.59799749e-01 -3.89253944e-01 4.96206254e-01
-3.00929397e-01 1.42628282e-01 -1.07589237e-01 -6.70520902e-01
-1.38690203e-01 -4.51602906e-01 -5.89831695e-02 -1.09671593e+00
-7.01750159e-01 -3.56155813e-01 8.36949766e-01 -1.30832040e+00
1.70522487e+00 -1.58358800e+00 1.58099815e-01 -1.03726961e-01
-1.03416987e-01 5.89164555e-01 2.01689199e-01 1.03949821e+00
1.49006858e-01 1.27013803e-01 -3.09659064e-01 -2.30462313e-01
-3.10887218e-01 -9.24846604e-02 -5.69123268e-01 -9.31826234e-02
3.22316527e-01 8.61816227e-01 -7.67606437e-01 -7.08096802e-01
7.50524476e-02 7.47332275e-02 -2.67409861e-01 2.70988613e-01
-4.39141661e-01 3.44603062e-01 -7.99276769e-01 2.43447348e-01
7.20053539e-02 3.80277634e-02 -2.26962730e-01 -7.72503531e-03
-3.74536127e-01 5.77666402e-01 -5.30807853e-01 1.52714431e+00
-3.81804705e-01 9.26142335e-01 -4.57640946e-01 -1.31553233e+00
1.04969168e+00 5.39965212e-01 1.97008312e-01 -6.29874825e-01
5.37138343e-01 2.74392277e-01 -3.34964879e-02 -6.79150760e-01
1.22715068e+00 2.50817649e-02 -2.90607899e-01 3.55024099e-01
4.95754629e-02 -3.48260969e-01 6.66222155e-01 4.57944691e-01
1.09924924e+00 -1.03793792e-01 9.38316226e-01 -1.76825002e-01
8.02521110e-01 5.37672698e-01 1.07270591e-02 5.50036788e-01
3.94225210e-01 4.74169612e-01 5.47379792e-01 -2.56499052e-01
-1.46948862e+00 -3.56028229e-01 1.41781092e-01 6.64837241e-01
-3.80388767e-01 -3.99234474e-01 -7.28633642e-01 -4.95612204e-01
-4.07337934e-01 1.13929725e+00 -4.59210873e-01 -2.75229722e-01
-5.89395165e-01 -3.87864023e-01 2.29302809e-01 2.66127259e-01
6.04795754e-01 -1.36119294e+00 -7.10989892e-01 6.93236172e-01
-3.21766824e-01 -7.79588759e-01 -2.50960052e-01 3.18731129e-01
-1.07228112e+00 -7.08969951e-01 -7.16794014e-01 -7.73115814e-01
4.46611106e-01 2.60800511e-01 9.70990479e-01 -2.80232340e-01
-1.86722234e-01 -1.46949753e-01 -8.49939406e-01 -9.05389428e-01
-9.82904077e-01 7.58720279e-01 -3.21680367e-01 -1.63670853e-01
2.38532692e-01 -3.87725949e-01 -3.78108352e-01 -7.61187911e-01
-1.04270601e+00 3.27601880e-01 9.25464809e-01 8.92635942e-01
1.75500885e-01 1.42039239e-01 1.28375840e+00 -1.06071556e+00
1.23822343e+00 -5.75293481e-01 -2.10693434e-01 2.52917737e-01
-2.56371588e-01 5.25927424e-01 1.01993454e+00 7.30951652e-02
-1.34113479e+00 -2.12648094e-01 -3.97043735e-01 4.37992841e-01
3.33343521e-02 1.17054486e+00 -3.24928463e-02 7.12820709e-01
5.20021141e-01 4.93700862e-01 -2.68713683e-01 -4.68668222e-01
2.11236492e-01 1.41668773e+00 4.17500854e-01 -5.52683882e-03
2.08412170e-01 -9.99816731e-02 -1.35804385e-01 -1.40709853e+00
-1.21935105e+00 -7.20731735e-01 -6.08449101e-01 -2.60974556e-01
7.39865005e-01 -6.57407284e-01 1.32066128e-03 1.88653581e-02
-1.40523267e+00 3.09766203e-01 -2.25116476e-01 4.09231097e-01
-4.26590711e-01 3.30888808e-01 -2.65240550e-01 -9.15971279e-01
-1.08797967e+00 -6.33650839e-01 8.86965811e-01 5.18714666e-01
-5.41864455e-01 -8.10817540e-01 6.20004758e-02 4.77195114e-01
2.22006634e-01 2.07511812e-01 1.00183082e+00 -1.22970748e+00
-2.34654665e-01 -6.18676424e-01 -2.40752976e-02 6.09008193e-01
3.13852489e-01 -5.94403595e-03 -7.47958004e-01 1.41164988e-01
1.90713570e-01 -3.34173024e-01 1.13475442e+00 3.02820563e-01
8.04523230e-01 -6.99492037e-01 -6.98459819e-02 -1.26452178e-01
1.45790088e+00 3.56568128e-01 7.20680356e-01 3.63954723e-01
4.43695724e-01 5.12100816e-01 6.41661346e-01 4.39532638e-01
1.93153754e-01 1.07326433e-01 3.98342460e-02 4.33950156e-01
-8.20710883e-02 -3.63173634e-01 3.58825147e-01 1.72205043e+00
7.41657838e-02 -5.68542063e-01 -5.99885166e-01 5.15749633e-01
-1.82417893e+00 -1.24698734e+00 -2.02204406e-01 1.89127517e+00
9.34298396e-01 3.87765318e-01 -1.62578046e-01 3.34342152e-01
5.61028659e-01 2.52949834e-01 -1.73650399e-01 -9.98938322e-01
-4.95983511e-02 1.80823296e-01 2.69055754e-01 4.76068854e-01
-8.49388003e-01 9.45360243e-01 5.13238621e+00 7.58121252e-01
-1.10573125e+00 -2.08347246e-01 5.62864780e-01 2.11700961e-01
-1.21466525e-01 -2.18047604e-01 -7.71832526e-01 6.75476968e-01
1.25619900e+00 -7.86532462e-01 -2.09175229e-01 7.29537368e-01
6.04790568e-01 -7.17525721e-01 -6.62514746e-01 6.67259216e-01
4.98470932e-01 -1.37650621e+00 4.27558094e-01 -2.15573847e-01
8.66996646e-01 -2.14489594e-01 -4.11957175e-01 2.55271584e-01
5.90010323e-02 -7.44324028e-01 4.40904945e-01 6.76530361e-01
3.72560173e-01 -1.13729596e+00 1.18644881e+00 6.93709314e-01
-7.74404705e-01 1.92535110e-02 -5.95418394e-01 -1.25365570e-01
3.53861421e-01 5.28878987e-01 -1.15796411e+00 8.29045057e-01
4.04792354e-02 7.46354818e-01 -5.50455332e-01 1.17633438e+00
-1.96781218e-01 7.29269862e-01 -1.11658297e-01 -6.90214872e-01
5.26129484e-01 -3.62853587e-01 7.57312477e-01 1.35299551e+00
4.40357804e-01 1.05313256e-01 -9.83888358e-02 2.12791607e-01
-3.58503342e-01 6.20972335e-01 -8.86097312e-01 -4.11583334e-01
1.39695898e-01 1.10369265e+00 -8.16965640e-01 -7.06401169e-01
-1.04998775e-01 1.21043217e+00 3.13024968e-01 3.27330790e-02
-3.39645207e-01 -6.73956454e-01 -1.88013121e-01 -1.32313654e-01
2.63132274e-01 -9.74063650e-02 -2.79182047e-01 -1.18318474e+00
1.03626158e-02 -7.97639072e-01 2.18772635e-01 -7.32025623e-01
-7.33581185e-01 8.71048987e-01 -1.57187387e-01 -1.02929568e+00
-5.70082843e-01 -2.15460733e-01 -7.81773090e-01 7.78421223e-01
-1.34163046e+00 -9.15849388e-01 5.77602498e-02 -4.89311293e-02
1.33700299e+00 -4.31967020e-01 5.91383576e-01 -4.83213319e-03
-4.44854736e-01 -5.82465045e-02 5.58872223e-01 2.34539472e-02
4.97113496e-01 -1.39833045e+00 2.03541607e-01 1.03581583e+00
2.28815660e-01 4.16060776e-01 1.18268871e+00 -9.48751152e-01
-1.31878054e+00 -1.15992713e+00 1.44214344e+00 -1.79181188e-01
6.53733075e-01 7.14227334e-02 -9.34631944e-01 4.68591154e-01
9.04828966e-01 -1.01229179e+00 4.90725785e-01 -3.63806993e-01
3.20428729e-01 -3.36078554e-02 -7.13262737e-01 6.85476482e-01
2.56496012e-01 -2.65928686e-01 -1.42116773e+00 1.78764939e-01
8.43296707e-01 -2.14529663e-01 -3.86716485e-01 -1.52442083e-01
2.31774226e-01 -6.51972413e-01 3.27752531e-01 -7.59343326e-01
9.65516865e-01 7.01463968e-02 1.22573048e-01 -1.79281425e+00
1.57305866e-01 -6.91658139e-01 -2.82960534e-01 1.44385254e+00
6.84891164e-01 -3.21086079e-01 5.71568668e-01 1.16064630e-01
-4.58922833e-01 -4.80024815e-01 -7.21676707e-01 -3.23702931e-01
-2.40083747e-02 -2.85453687e-04 4.71848026e-02 4.62572753e-01
3.95009637e-01 1.29240787e+00 -4.32398349e-01 -5.25417089e-01
4.41701561e-01 4.57657548e-03 5.72522163e-01 -1.17467010e+00
1.90054893e-01 -5.18868685e-01 -1.82181224e-01 -9.25740123e-01
5.36461361e-02 -9.23011899e-01 8.90444890e-02 -2.11171150e+00
3.22915733e-01 2.86303371e-01 3.48227732e-02 -1.76935166e-01
-3.42936486e-01 -1.40339985e-01 1.30018979e-01 4.57105525e-02
-5.34839511e-01 6.81780517e-01 1.13173580e+00 -1.13973573e-01
-2.70633370e-01 2.67625034e-01 -9.50867951e-01 5.12554407e-01
1.16428101e+00 -4.55399454e-01 -3.63545001e-01 -6.14228286e-03
3.88955355e-01 2.99183577e-01 -3.97948116e-01 -9.70068693e-01
4.39587414e-01 1.16262674e-01 2.71258861e-01 -1.09053957e+00
1.52945712e-01 -7.06430197e-01 -2.22183838e-01 3.37183803e-01
-7.38715053e-01 3.51089329e-01 4.08175401e-02 4.03334111e-01
-5.98255932e-01 -7.39537418e-01 4.28337276e-01 -2.82470703e-01
-5.25671542e-01 -1.69404894e-01 -5.01777053e-01 8.29103217e-02
8.04296732e-01 -2.04271808e-01 -2.65659720e-01 -5.66154361e-01
-2.13353038e-01 1.15367912e-01 -3.71827185e-02 3.54723126e-01
6.31142199e-01 -8.44329834e-01 -1.15722990e+00 -2.58251876e-01
4.53749113e-02 -1.48788303e-01 8.96800533e-02 6.01230323e-01
-8.05951595e-01 9.02814031e-01 -2.56281555e-01 3.89774814e-02
-1.32114470e+00 4.15217668e-01 -4.18562412e-01 -5.01682460e-01
-9.16497469e-01 2.02580974e-01 -4.91642922e-01 4.00124013e-01
9.82953683e-02 -4.00252104e-01 -7.63805151e-01 5.88466883e-01
9.31462824e-01 3.24877977e-01 2.66782343e-01 -6.20595217e-01
2.54975080e-01 1.35598153e-01 -4.27333206e-01 -2.52258807e-01
1.60604060e+00 -3.83028656e-01 -3.02424043e-01 7.33058512e-01
1.15141022e+00 1.41900510e-01 -5.92633843e-01 7.06057204e-03
7.35724390e-01 7.78041184e-02 1.17247872e-01 -7.59175897e-01
-3.27939808e-01 9.72010553e-01 -2.15141654e-01 5.67846179e-01
1.02316225e+00 -7.35518262e-02 9.86996770e-01 7.10541248e-01
-4.74839285e-02 -1.51228452e+00 -3.27604683e-03 7.54852116e-01
1.08622169e+00 -1.41907871e+00 1.96758404e-01 8.80958214e-02
-9.64845836e-01 1.34284663e+00 9.94884893e-02 -2.48873129e-01
2.48685300e-01 2.18765065e-01 -2.01851651e-01 -1.14597365e-01
-9.40693855e-01 -1.86232284e-01 4.88516957e-01 8.06350037e-02
7.38100350e-01 -1.36981279e-01 -9.77867246e-01 5.07657230e-01
-4.73550081e-01 -6.14723898e-02 8.48399162e-01 9.58287477e-01
-1.18449533e+00 -9.27429795e-01 -1.09937847e-01 9.68244255e-01
-1.04144037e+00 -2.12831244e-01 -6.12090230e-01 4.68009144e-01
-5.58039248e-01 9.85202372e-01 -2.30393022e-01 -3.47580016e-02
1.59971818e-01 3.15023899e-01 7.15113580e-02 -8.73220801e-01
-8.26962769e-01 -1.93448607e-02 4.47806031e-01 1.67714328e-01
-4.01056409e-01 -3.92392665e-01 -1.20481539e+00 -1.31375790e-02
-3.60017657e-01 7.34770298e-01 9.47069645e-01 1.01511896e+00
2.07126096e-01 8.26809227e-01 6.38437092e-01 -4.47767735e-01
-4.95991021e-01 -1.42794979e+00 -2.56266594e-01 1.80875570e-01
4.13307607e-01 6.78701699e-02 -5.08411117e-02 2.98952907e-01] | [12.522771835327148, 9.522319793701172] |
3398f88d-b7f8-4dcb-82cc-269ed7250395 | didfuse-deep-image-decomposition-for-infrared | 2003.09210 | null | https://arxiv.org/abs/2003.09210v3 | https://arxiv.org/pdf/2003.09210v3.pdf | DIDFuse: Deep Image Decomposition for Infrared and Visible Image Fusion | Infrared and visible image fusion, a hot topic in the field of image processing, aims at obtaining fused images keeping the advantages of source images. This paper proposes a novel auto-encoder (AE) based fusion network. The core idea is that the encoder decomposes an image into background and detail feature maps with low- and high-frequency information, respectively, and that the decoder recovers the original image. To this end, the loss function makes the background/detail feature maps of source images similar/dissimilar. In the test phase, background and detail feature maps are respectively merged via a fusion module, and the fused image is recovered by the decoder. Qualitative and quantitative results illustrate that our method can generate fusion images containing highlighted targets and abundant detail texture information with strong robustness and meanwhile surpass state-of-the-art (SOTA) approaches. | ['Chun-Xia Zhang', 'Junmin Liu', 'Jiangshe Zhang', 'Zixiang Zhao', 'Shuang Xu', 'Pengfei Li'] | 2020-03-20 | null | null | null | null | ['infrared-and-visible-image-fusion'] | ['computer-vision'] | [ 5.26244044e-01 -4.74046052e-01 2.07095087e-01 -7.47090131e-02
-8.42146754e-01 -2.24902168e-01 4.37915146e-01 -2.14124352e-01
-4.05793861e-02 6.37382329e-01 1.85429722e-01 1.48754239e-01
-1.47540420e-01 -8.43274057e-01 -4.76711899e-01 -1.23485672e+00
3.89140874e-01 -4.61088896e-01 4.47892137e-02 -4.20639664e-01
-9.10977572e-02 3.11397046e-01 -1.71997726e+00 3.92855614e-01
1.05035806e+00 1.27944517e+00 5.35351753e-01 3.96960139e-01
-1.52519336e-02 8.85316014e-01 -4.16458130e-01 -4.85873282e-01
4.67693537e-01 -5.47651529e-01 -1.60034671e-01 6.11397564e-01
4.63498861e-01 -4.63277906e-01 -7.28899896e-01 1.68168497e+00
5.36095560e-01 -7.18917772e-02 4.95249093e-01 -1.15976226e+00
-9.59146798e-01 1.09084219e-01 -1.00210798e+00 2.05699742e-01
1.96647689e-01 1.34957850e-01 3.86186510e-01 -7.45439410e-01
2.02193543e-01 1.17619121e+00 3.93947899e-01 1.45142838e-01
-1.25753820e+00 -5.40456712e-01 -8.62467941e-03 2.02692077e-01
-1.37918019e+00 -4.98075515e-01 9.33843195e-01 -1.69758886e-01
1.24677725e-01 3.32148105e-01 4.80369568e-01 7.39120305e-01
7.48694837e-01 7.86364377e-01 1.22482216e+00 -3.34348291e-01
-1.14474501e-02 -6.23255922e-03 -2.04855055e-01 6.04507506e-01
3.96340847e-01 3.79570097e-01 -4.04293507e-01 1.42675459e-01
5.73963404e-01 4.34572786e-01 -7.29678452e-01 -4.89693135e-03
-1.21718192e+00 4.09649879e-01 4.56953973e-01 3.76287639e-01
-6.85325265e-01 -3.08727890e-01 -9.56004411e-02 4.36385483e-01
6.74934268e-01 -4.03079599e-01 -1.04491957e-01 6.10218942e-01
-9.89098549e-01 -3.65848579e-02 3.30348700e-01 7.00772643e-01
9.91829634e-01 1.61848605e-01 -4.15097624e-01 7.08863139e-01
5.22073328e-01 9.77625489e-01 1.84170052e-01 -9.66520011e-01
3.85396689e-01 5.00560164e-01 5.89220896e-02 -1.20156825e+00
1.48119420e-01 -6.06400132e-01 -1.42227161e+00 7.35885561e-01
-5.60470521e-02 -1.98953480e-01 -8.43878031e-01 1.41775322e+00
2.17755765e-01 1.59637511e-01 4.14550811e-01 9.03483927e-01
8.97083223e-01 8.79214644e-01 -1.61550164e-01 -5.53313911e-01
1.36715293e+00 -8.42292607e-01 -1.09652722e+00 -3.75724614e-01
-3.19908679e-01 -1.16687655e+00 3.06142956e-01 3.77210259e-01
-1.22653782e+00 -1.05914605e+00 -1.07232749e+00 4.71994393e-02
5.57726100e-02 5.53249955e-01 3.39957565e-01 4.92995054e-01
-1.05552018e+00 3.95768940e-01 -3.76064837e-01 1.38276620e-02
2.61104614e-01 -3.12321354e-02 -5.78862548e-01 -5.32337427e-01
-1.02570605e+00 8.39754403e-01 5.35115063e-01 3.45253706e-01
-7.23916292e-01 -5.66823363e-01 -1.06866932e+00 7.98233747e-02
2.50765085e-01 -9.79822934e-01 8.55592132e-01 -1.04179394e+00
-1.42101300e+00 5.20763159e-01 -3.31220239e-01 -1.40622720e-01
2.20944420e-01 -5.75055070e-02 -8.72655749e-01 3.48742425e-01
2.22806126e-01 5.11780858e-01 1.30862391e+00 -1.59882760e+00
-1.08715749e+00 -2.24345118e-01 -2.49813855e-01 2.96735555e-01
-3.30429189e-02 -8.24708119e-02 -5.42014301e-01 -7.05903530e-01
2.95210272e-01 -3.81278455e-01 3.56806293e-02 1.90906242e-01
-2.67002255e-01 3.20670336e-01 1.07540524e+00 -1.25750339e+00
1.00148094e+00 -2.63638306e+00 1.77809656e-01 5.11039793e-02
4.11263406e-01 2.57012427e-01 -3.08638901e-01 1.77302063e-01
-2.69041032e-01 -4.26224738e-01 -4.66869473e-01 -5.62676266e-02
-3.78137946e-01 -1.35769084e-01 -3.38743329e-01 6.78764939e-01
1.38857737e-01 8.31266344e-01 -7.94932723e-01 -5.74438095e-01
7.16271281e-01 7.57240891e-01 1.01352846e-02 2.45746762e-01
2.73924381e-01 7.43503809e-01 -2.94483662e-01 7.49492407e-01
1.26261139e+00 -5.92847541e-02 -1.01994239e-01 -8.38493586e-01
-2.91018605e-01 -4.85779375e-01 -1.21436381e+00 1.53027880e+00
-4.02910411e-01 6.02662146e-01 4.47784722e-01 -8.41432631e-01
1.07236207e+00 3.64988804e-01 5.96199811e-01 -1.03095388e+00
1.94267809e-01 1.66810364e-01 -1.99209020e-01 -5.03133893e-01
2.60320306e-01 -1.49178401e-01 2.56976098e-01 -7.95748308e-02
4.89382744e-02 -1.06650263e-01 6.61252066e-02 -6.87402561e-02
5.74320853e-01 1.74108788e-01 1.28375590e-01 4.50864099e-02
9.63608563e-01 -2.08384275e-01 5.75044036e-01 3.36333036e-01
5.31547107e-02 6.85473859e-01 -8.61789435e-02 -2.19655275e-01
-1.06810808e+00 -1.30635405e+00 5.38673624e-02 4.73980844e-01
7.38409102e-01 2.86035035e-02 -5.31104684e-01 -1.51447713e-01
-1.17973991e-01 4.10496294e-01 -4.13036883e-01 -2.93699473e-01
-2.87733793e-01 -5.72582245e-01 8.20302665e-02 1.98905900e-01
1.30421710e+00 -6.40510499e-01 -2.84788609e-01 2.08795995e-01
-6.70730174e-01 -1.01157403e+00 -4.89234120e-01 -2.39003122e-01
-8.06441188e-01 -9.95646656e-01 -1.14795363e+00 -8.92732739e-01
6.89840496e-01 1.12093079e+00 8.10672700e-01 -1.14099689e-01
-2.89597124e-01 -3.16819437e-02 -2.63411224e-01 -1.53496414e-01
-3.81687611e-01 -7.93644130e-01 -3.71814519e-01 6.54331982e-01
1.15044722e-02 -3.66239995e-01 -7.10032523e-01 1.47865623e-01
-1.17684078e+00 3.38289082e-01 9.88421440e-01 9.02348638e-01
6.29758716e-01 9.54734743e-01 1.43375993e-01 3.24908867e-02
3.54125410e-01 -1.90055221e-01 -7.53585637e-01 4.80454743e-01
-3.41406256e-01 -9.13788453e-02 4.63760287e-01 -4.57646064e-02
-1.73992872e+00 5.03440350e-02 5.05482256e-02 -5.29527545e-01
-1.19116761e-01 3.58420223e-01 -5.74694693e-01 -2.86954224e-01
3.00103515e-01 7.77177036e-01 3.20061564e-01 -5.71612060e-01
2.32646316e-01 7.21617341e-01 1.02934051e+00 -2.06578150e-02
1.11827421e+00 7.71977782e-01 -5.66320084e-02 -5.58077931e-01
-9.28054094e-01 -4.49515939e-01 -3.55830133e-01 -3.01461309e-01
9.36152995e-01 -1.32541621e+00 -6.49591208e-01 7.13639557e-01
-1.04778790e+00 1.61285937e-01 -2.03477904e-01 6.07801914e-01
-3.38958174e-01 7.06363857e-01 -5.80209196e-01 -8.08116913e-01
-3.87296319e-01 -1.10866463e+00 1.30923665e+00 7.12584674e-01
6.51008427e-01 -7.92090416e-01 -2.47704044e-01 3.23276430e-01
5.36683440e-01 2.84240723e-01 5.11117876e-01 2.64288515e-01
-6.44076943e-01 -1.15702385e-02 -5.60957432e-01 7.52786160e-01
5.00171065e-01 -4.54817377e-02 -1.00755942e+00 -3.52634162e-01
3.78597170e-01 2.95460135e-01 1.10521948e+00 5.00409007e-01
8.34156394e-01 -2.51517951e-01 -2.98167676e-01 7.29079723e-01
1.67543638e+00 3.48609686e-01 9.77031648e-01 2.82175601e-01
3.79055947e-01 5.19056022e-01 6.06996179e-01 1.38596594e-01
1.51096493e-01 5.06836057e-01 3.95626336e-01 -6.93052173e-01
-4.85191077e-01 1.58231258e-02 5.10049343e-01 5.49381196e-01
2.47037224e-02 -3.97563964e-01 -3.55998516e-01 4.82934237e-01
-1.74474072e+00 -1.36943805e+00 -1.10169142e-01 2.17154884e+00
6.28830731e-01 -3.00521404e-01 -3.53957206e-01 3.41795951e-01
1.09190214e+00 2.53776371e-01 -3.34099531e-01 2.98129857e-01
-7.17643321e-01 1.00233108e-01 5.42935729e-01 4.13515717e-01
-1.21150613e+00 3.59407425e-01 5.86547375e+00 1.06565607e+00
-1.10140657e+00 -1.70461778e-02 6.91607118e-01 4.64890152e-01
-2.26516232e-01 1.56632683e-04 -2.40006775e-01 7.17479289e-01
4.46546078e-01 -2.60389894e-01 5.32832801e-01 3.03616166e-01
5.94142750e-02 -4.08126205e-01 -4.94173855e-01 1.15145254e+00
2.68946260e-01 -1.14021003e+00 -1.11655012e-01 -4.84026317e-03
8.13696206e-01 -1.75104812e-01 2.89717108e-01 -1.66882455e-01
3.00372899e-01 -8.00228894e-01 7.33051538e-01 1.04889691e+00
8.55792105e-01 -7.55356908e-01 1.03841472e+00 3.26703966e-01
-1.53890073e+00 -1.99873462e-01 -4.12322551e-01 2.88573235e-01
3.95691305e-01 9.41401660e-01 1.44579902e-01 1.53778994e+00
8.20313036e-01 8.64493430e-01 -4.75552768e-01 1.17046165e+00
-1.81972295e-01 1.67346403e-01 -2.35891696e-02 8.73165905e-01
-5.57410344e-02 -7.01789677e-01 5.30256331e-01 8.95208478e-01
5.68690717e-01 1.69661403e-01 2.28272423e-01 8.46157849e-01
3.30987841e-01 -3.26089948e-01 -5.92599094e-01 2.21974701e-01
1.30080819e-01 1.41778004e+00 -4.66069728e-01 -5.88888586e-01
-4.84773606e-01 1.17768633e+00 -4.60352868e-01 5.60032666e-01
-8.09481442e-01 -3.85354608e-01 4.72575337e-01 -4.37098116e-01
3.23118746e-01 5.51836449e-04 -1.92144707e-01 -1.32662606e+00
1.85591444e-01 -9.09539163e-01 2.40880921e-01 -1.15775418e+00
-1.40547407e+00 6.98627889e-01 -1.62731588e-01 -1.79217637e+00
2.14436129e-01 -3.23334664e-01 -6.30627632e-01 1.20879507e+00
-1.96627879e+00 -1.32256460e+00 -9.12097573e-01 7.19668806e-01
3.68085355e-01 -2.05942482e-01 3.32014173e-01 4.72081423e-01
-4.94005740e-01 1.49679095e-01 5.88336945e-01 5.91842122e-02
6.05017543e-01 -6.94510102e-01 9.89972502e-02 1.20185900e+00
-8.45481828e-02 1.69903994e-01 6.57676458e-01 -7.29376912e-01
-1.27284932e+00 -1.21139407e+00 5.66932499e-01 2.04526529e-01
1.69529274e-01 1.98669925e-01 -9.02267098e-01 2.03582942e-01
5.91805637e-01 1.35769024e-01 2.71438420e-01 -8.70679438e-01
-2.02722952e-01 -3.60033691e-01 -1.22517955e+00 4.27432716e-01
5.83124816e-01 -5.00228763e-01 -4.71601486e-01 1.27875492e-01
5.45477509e-01 -4.01307315e-01 -7.21418917e-01 6.75282300e-01
4.72000033e-01 -1.29769397e+00 1.17469740e+00 2.80523956e-01
4.10783648e-01 -9.47069049e-01 -3.10727060e-01 -1.34780896e+00
-6.58879638e-01 -4.14906114e-01 8.96319374e-02 1.24998486e+00
-8.86011049e-02 -6.51490927e-01 4.22373451e-02 -2.36057509e-02
-3.94821540e-02 -2.30791181e-01 -7.19854593e-01 -6.46176457e-01
-5.08092225e-01 1.16597183e-01 5.16689658e-01 7.26677299e-01
-5.35542250e-01 1.12320907e-01 -5.98721623e-01 5.93865216e-01
1.08150053e+00 3.05847377e-01 4.89571601e-01 -1.07736588e+00
-1.27865657e-01 -2.51932055e-01 -4.70830888e-01 -7.34669089e-01
-1.37803028e-03 -5.76844037e-01 8.33055601e-02 -1.79766309e+00
2.95265853e-01 1.02351658e-01 -2.64020830e-01 3.48760307e-01
-4.49447989e-01 5.23885846e-01 3.02672237e-01 2.57772744e-01
-2.78741300e-01 8.25305581e-01 1.36449754e+00 -5.21953523e-01
1.13558367e-01 -1.50589496e-01 -8.74723017e-01 5.61792552e-01
6.14727437e-01 -1.95550337e-01 -2.34155521e-01 -4.97803807e-01
-2.27707058e-01 4.04662609e-01 8.17840576e-01 -1.23111868e+00
3.63430798e-01 -1.56515419e-01 8.68319690e-01 -7.08294094e-01
3.00506681e-01 -1.13538253e+00 5.11201859e-01 4.41071540e-01
5.62733933e-02 -2.92822778e-01 1.97676763e-01 6.95299089e-01
-7.71967173e-01 1.39959544e-01 8.88066828e-01 1.65009767e-01
-7.70306349e-01 3.20553571e-01 -1.56682327e-01 -5.72877824e-01
1.10073733e+00 -4.31126446e-01 -5.58462203e-01 -4.11648631e-01
-4.50559318e-01 6.01789579e-02 5.44334769e-01 2.78679729e-01
8.45601022e-01 -1.48833311e+00 -1.12731266e+00 4.34300274e-01
4.38133515e-02 -1.19729854e-01 8.91374469e-01 1.06803775e+00
-3.41065377e-01 1.01940575e-04 -4.62129235e-01 -5.87008893e-01
-1.37174761e+00 7.75254250e-01 4.09065247e-01 9.37948450e-02
-8.02190006e-01 5.64404666e-01 7.18110979e-01 4.00498770e-02
-1.21750735e-01 4.18524630e-02 -2.40693197e-01 -1.66554809e-01
9.64149714e-01 3.40493321e-01 -3.21869925e-02 -9.10939217e-01
-8.76543000e-02 7.63509691e-01 1.67626187e-01 -3.27140763e-02
1.22976649e+00 -5.48641205e-01 -4.01117116e-01 -4.26763371e-02
1.23401093e+00 1.19340122e-01 -1.40612257e+00 -5.08950531e-01
-6.44389689e-01 -8.96442413e-01 4.66356486e-01 -6.87360108e-01
-1.49192166e+00 7.57362187e-01 1.02690661e+00 1.06408298e-01
1.83600879e+00 -1.15596771e-01 7.16003776e-01 -3.47064137e-02
1.20658986e-01 -7.00151026e-01 -1.72271989e-02 6.93347827e-02
8.26855302e-01 -1.16948307e+00 1.38936996e-01 -5.76568067e-01
-5.70105076e-01 1.16437387e+00 3.61059666e-01 3.61998230e-02
3.45716178e-01 3.88066947e-01 1.50070474e-01 -8.54047239e-02
-4.69591498e-01 -5.60603023e-01 4.37704831e-01 7.62245178e-01
1.08197413e-01 -2.62747020e-01 1.07387811e-01 1.96124494e-01
3.01632851e-01 1.25822380e-01 1.60413861e-01 8.32945704e-01
-5.96007884e-01 -8.48335624e-01 -9.60617602e-01 2.28402674e-01
-4.39162284e-01 -1.59591436e-01 1.78665705e-02 3.23793769e-01
3.93790394e-01 1.42580128e+00 3.68279964e-02 -5.29823422e-01
8.31975341e-02 -3.20946217e-01 3.81304055e-01 -3.66258360e-02
-1.89746276e-01 4.42921579e-01 -2.40149036e-01 -4.57626432e-01
-7.82743931e-01 -1.37998626e-01 -7.55952358e-01 -3.78835976e-01
-6.09989285e-01 1.35173216e-01 4.97541398e-01 7.76245713e-01
3.61991912e-01 8.38963807e-01 1.05239952e+00 -9.39187944e-01
-2.41012439e-01 -7.51091361e-01 -8.04449677e-01 4.32503194e-01
7.23920524e-01 -5.20581543e-01 -4.52092469e-01 5.44899762e-01] | [10.604340553283691, -1.9327218532562256] |
9c4c474e-9888-482d-8c4e-d57582077c36 | wavedm-wavelet-based-diffusion-models-for | 2305.13819 | null | https://arxiv.org/abs/2305.13819v1 | https://arxiv.org/pdf/2305.13819v1.pdf | WaveDM: Wavelet-Based Diffusion Models for Image Restoration | Latest diffusion-based methods for many image restoration tasks outperform traditional models, but they encounter the long-time inference problem. To tackle it, this paper proposes a Wavelet-Based Diffusion Model (WaveDM) with an Efficient Conditional Sampling (ECS) strategy. WaveDM learns the distribution of clean images in the wavelet domain conditioned on the wavelet spectrum of degraded images after wavelet transform, which is more time-saving in each step of sampling than modeling in the spatial domain. In addition, ECS follows the same procedure as the deterministic implicit sampling in the initial sampling period and then stops to predict clean images directly, which reduces the number of total sampling steps to around 5. Evaluations on four benchmark datasets including image raindrop removal, defocus deblurring, demoir\'eing, and denoising demonstrate that WaveDM achieves state-of-the-art performance with the efficiency that is comparable to traditional one-pass methods and over 100 times faster than existing image restoration methods using vanilla diffusion models. | ['Shifeng Chen', 'Jiaxi Lv', 'Yu Dong', 'Jianzhuang Liu', 'Jiancheng Huang', 'Yi Huang'] | 2023-05-23 | null | null | null | null | ['deblurring', 'image-restoration'] | ['computer-vision', 'computer-vision'] | [ 3.89232278e-01 -5.07542670e-01 1.95760369e-01 -2.19230697e-01
-1.11698663e+00 -7.10247830e-02 5.66846550e-01 -3.24500471e-01
-5.06586730e-01 6.68731153e-01 3.53266090e-01 -2.98301786e-01
-1.27489135e-01 -9.20845807e-01 -6.26913726e-01 -1.42252660e+00
5.12150396e-03 1.85414210e-01 3.60130042e-01 9.94994715e-02
3.97973657e-01 2.03752488e-01 -1.20753610e+00 2.51942873e-01
1.20848954e+00 9.00222003e-01 6.54251158e-01 8.89294386e-01
-9.38484967e-02 1.01056957e+00 -6.65997326e-01 -9.60489139e-02
7.76696503e-02 -5.67146480e-01 -3.68928283e-01 2.33434439e-01
2.89005041e-01 -7.72056162e-01 -7.04177022e-01 1.51585972e+00
8.32277954e-01 2.26300880e-01 6.79201961e-01 -4.83301163e-01
-1.11333776e+00 3.79408032e-01 -9.62867975e-01 5.42654157e-01
7.35983551e-02 1.68386355e-01 3.40064824e-01 -9.96751785e-01
4.98240829e-01 1.38086843e+00 8.33275557e-01 3.01698774e-01
-1.35456479e+00 -5.96277595e-01 7.57552832e-02 4.96335804e-01
-1.15897143e+00 -4.59933937e-01 6.78909361e-01 -3.10252815e-01
6.56200349e-01 1.20651066e-01 4.84741718e-01 1.06862617e+00
3.84559661e-01 7.46178150e-01 1.69555700e+00 -4.29679334e-01
5.19503176e-01 -4.80836332e-01 2.63251603e-01 3.96929085e-01
1.55574530e-01 3.27459008e-01 -6.14179194e-01 -3.25382531e-01
7.89570570e-01 3.65217775e-02 -8.04565728e-01 2.01952040e-01
-9.74119782e-01 9.31078613e-01 2.68886983e-01 1.26496896e-01
-7.11685479e-01 1.85897082e-01 1.21836193e-01 4.99660462e-01
1.16687787e+00 -2.94050992e-01 -9.25152302e-02 2.10698485e-01
-1.48790884e+00 2.87697494e-01 6.89379692e-01 6.16318762e-01
6.71719849e-01 2.59936750e-01 -5.08177996e-01 9.76373196e-01
2.73319006e-01 1.12345707e+00 2.26114258e-01 -1.15083075e+00
1.73321664e-01 -3.82860392e-01 3.79943937e-01 -7.52718508e-01
1.92895383e-01 -4.50313270e-01 -1.51764727e+00 5.02761781e-01
2.12181449e-01 -4.06512571e-03 -1.34031963e+00 1.11144900e+00
3.67885202e-01 7.79774845e-01 4.14523780e-02 9.74195957e-01
5.52484453e-01 1.05681574e+00 -2.52358079e-01 -6.80852354e-01
1.19112706e+00 -9.71768200e-01 -1.08455551e+00 -2.93369204e-01
-3.26898903e-01 -1.06199443e+00 6.64538920e-01 8.81515741e-01
-1.04534650e+00 -5.69217086e-01 -9.92627203e-01 -1.05107769e-01
2.07807839e-01 -1.55276677e-03 5.31643748e-01 7.80461431e-01
-1.08625722e+00 7.35546529e-01 -9.75407839e-01 1.25585586e-01
7.95476377e-01 -1.91317141e-01 -1.16932951e-01 -1.14386880e+00
-7.54297733e-01 7.87491202e-01 -3.77292395e-01 3.38365227e-01
-1.43292189e+00 -8.47340524e-01 -6.47375107e-01 -1.23721726e-01
-1.14001743e-01 -5.97626090e-01 8.25752378e-01 -5.02242267e-01
-1.66193831e+00 3.89999926e-01 -6.54009402e-01 -6.95749283e-01
6.76188171e-01 -3.93379539e-01 -3.75030547e-01 3.24479103e-01
1.23878699e-02 1.01535477e-01 1.73222876e+00 -1.37543619e+00
-4.05866325e-01 -2.79827654e-01 -2.35920459e-01 -1.95699468e-01
5.34004532e-02 -2.32503444e-01 -4.14429933e-01 -9.31705773e-01
2.96070576e-01 -5.58434069e-01 -4.15829837e-01 2.93948621e-01
-2.68384308e-01 2.74053887e-02 7.92884052e-01 -1.10917652e+00
1.11486602e+00 -2.36169505e+00 2.70037174e-01 -1.24917574e-01
2.27748126e-01 3.66501570e-01 -3.15547019e-01 4.67691153e-01
1.02523066e-01 -3.95912439e-01 -8.11519504e-01 -6.57500327e-01
-2.47630954e-01 3.65125507e-01 -6.73110425e-01 9.01183784e-01
-8.12596753e-02 3.04281980e-01 -9.28959310e-01 -2.75412112e-01
3.24340731e-01 9.69727337e-01 -4.40606654e-01 2.46268108e-01
-9.77359787e-02 4.74288613e-01 -2.35065028e-01 5.32162786e-01
1.60738730e+00 -6.29771054e-02 1.50004483e-03 -3.75309348e-01
-4.28430438e-02 -1.22339226e-01 -1.14676523e+00 1.64047444e+00
-6.69189155e-01 7.19561815e-01 3.34805250e-01 -1.17844975e+00
7.56883621e-01 2.92978376e-01 1.71418741e-01 -5.97319782e-01
-2.21740469e-01 1.20384447e-01 -5.62179327e-01 -7.43615508e-01
1.50825500e-01 -3.38277787e-01 8.15765202e-01 3.75727266e-01
1.03428354e-03 -3.79214704e-01 -5.66492043e-02 9.62188989e-02
1.20055091e+00 7.55115598e-02 -1.80731401e-01 -3.52831572e-01
2.86452919e-01 -6.08705357e-02 5.06644964e-01 1.19503701e+00
6.86222017e-02 8.71735811e-01 3.28576867e-03 -4.22549725e-01
-7.61781156e-01 -1.25252783e+00 -2.98556536e-01 3.86812806e-01
2.31785491e-01 -5.31381927e-02 -8.25584888e-01 -3.68240654e-01
-1.81903720e-01 9.10076082e-01 -6.36480987e-01 1.64404809e-01
-4.68683928e-01 -1.40267122e+00 2.12179348e-01 -3.51314954e-02
9.93270636e-01 -7.01978683e-01 -2.96231478e-01 3.43251556e-01
-5.71299672e-01 -1.07189977e+00 -4.21175599e-01 -5.27533442e-02
-1.06532109e+00 -9.73664701e-01 -1.21766293e+00 -6.73475981e-01
7.82805562e-01 8.02479148e-01 9.51481581e-01 -5.88193424e-02
-3.82364988e-01 1.98599547e-01 -4.87334758e-01 -2.59164453e-01
-1.64392233e-01 -6.68835163e-01 -3.90380949e-01 2.10473016e-01
1.79558933e-01 -7.68802226e-01 -1.12656260e+00 7.49887340e-03
-1.21086252e+00 -2.65322328e-01 5.57496071e-01 1.19101012e+00
7.52930462e-01 6.86121702e-01 3.17391723e-01 -7.72941649e-01
7.30994523e-01 -4.77746069e-01 -6.12992465e-01 6.28863946e-02
-6.92816615e-01 -1.47636890e-01 2.74747550e-01 -7.13784218e-01
-1.45059049e+00 -3.20381820e-01 -2.30284765e-01 -4.12070572e-01
1.40988454e-01 5.97720981e-01 4.93143260e-01 -1.24114610e-01
6.13971412e-01 7.51700401e-01 1.37704492e-01 -9.24488485e-01
2.85211504e-01 5.67548394e-01 6.17069542e-01 -4.11870480e-01
9.66662526e-01 1.08688211e+00 -1.48458585e-01 -1.13641310e+00
-9.65553403e-01 -2.84609586e-01 -1.69569045e-01 -1.46137044e-01
7.90590823e-01 -1.20025146e+00 -3.98059815e-01 1.03461635e+00
-1.27025628e+00 -4.48126167e-01 -3.73082548e-01 6.89834416e-01
-2.41378322e-01 7.58302987e-01 -9.06343997e-01 -9.97893095e-01
-4.34253514e-01 -1.02497482e+00 9.29854810e-01 2.71486379e-02
4.81198937e-01 -8.25777292e-01 -1.39824124e-02 3.26689303e-01
9.18497443e-01 2.46915240e-02 8.53597641e-01 4.70733374e-01
-8.30806911e-01 -2.18295857e-01 -4.37887400e-01 8.75152767e-01
6.57130778e-02 -4.44955885e-01 -8.96287322e-01 -6.82384908e-01
7.57966220e-01 -6.38897941e-02 1.47619677e+00 1.10790384e+00
1.10434282e+00 -1.09905623e-01 -9.93648842e-02 6.98656976e-01
1.72792292e+00 -1.31439537e-01 1.19350374e+00 5.81689999e-02
5.88876419e-02 1.18502557e-01 4.89324123e-01 5.11844337e-01
1.83605522e-01 2.14823246e-01 3.44559640e-01 -3.13494831e-01
-7.08079576e-01 1.14866473e-01 5.04300177e-01 8.91617656e-01
-1.47353858e-01 -3.78735781e-01 -3.11585993e-01 8.72538328e-01
-1.52753353e+00 -1.17014396e+00 -3.24665368e-01 2.21903181e+00
1.06089771e+00 -1.39661357e-01 -5.72485566e-01 4.32773307e-02
5.31175435e-01 5.82344949e-01 -5.76241791e-01 2.09681675e-01
-2.46902138e-01 6.29555166e-01 6.45599365e-01 9.94527578e-01
-8.46826792e-01 6.96261823e-01 6.84860134e+00 1.23542261e+00
-7.65856028e-01 5.69477081e-01 6.05442166e-01 1.05861500e-01
-2.62072146e-01 4.11403812e-02 -5.69582403e-01 4.45561588e-01
5.47472477e-01 2.40081519e-01 8.91862690e-01 2.88508564e-01
4.36907351e-01 -5.09278119e-01 -4.08944398e-01 8.89425874e-01
1.05811387e-01 -1.35256147e+00 -2.43393611e-02 -1.18275598e-01
8.10530603e-01 4.62457240e-01 2.92163733e-02 -1.37233377e-01
6.17345273e-01 -8.32610607e-01 4.44746137e-01 8.36740673e-01
7.52206802e-01 -4.50725853e-01 7.39671707e-01 4.08333331e-01
-6.29259646e-01 4.71855476e-02 -8.78207028e-01 1.91140056e-01
4.29099292e-01 1.60037518e+00 2.51240022e-02 6.06178701e-01
1.15889680e+00 7.06948340e-01 5.76831177e-02 1.25777304e+00
-3.45750958e-01 1.13939404e+00 -2.95511335e-01 5.96713603e-01
-4.70209727e-03 -5.39208055e-01 7.62529194e-01 1.20979691e+00
7.02302516e-01 3.12960446e-01 -6.52300343e-02 6.38105273e-01
1.30628854e-01 -6.49149954e-01 -1.60892472e-01 3.61959010e-01
3.81529212e-01 8.19530427e-01 -3.68495286e-01 -3.63258630e-01
-3.79928350e-01 1.39951229e+00 -2.98562080e-01 8.28436077e-01
-7.64998257e-01 -3.75029981e-01 5.85065901e-01 -3.55594419e-02
8.09199989e-01 -4.63585615e-01 6.34454982e-03 -1.20684111e+00
-1.46377049e-02 -9.18287158e-01 1.11774127e-04 -8.57916534e-01
-1.70355260e+00 6.60365224e-01 -1.38048574e-01 -1.12883985e+00
4.31165755e-01 -2.68879175e-01 -5.41808844e-01 1.19510877e+00
-2.34628773e+00 -8.72671247e-01 -4.91143405e-01 7.20604897e-01
8.69123459e-01 1.36783645e-01 8.00783515e-01 4.10824358e-01
-2.81858176e-01 -9.66532528e-02 7.26719618e-01 -1.50152668e-01
6.90199614e-01 -1.01218152e+00 4.80150253e-01 1.11137545e+00
-9.52102393e-02 3.79341036e-01 1.00138772e+00 -6.60432756e-01
-1.53195477e+00 -9.04609561e-01 7.27182388e-01 7.08461627e-02
4.37054694e-01 1.07414044e-01 -1.02854788e+00 2.04058170e-01
6.18808329e-01 2.79184550e-01 3.02416950e-01 -4.09985125e-01
-5.00823617e-01 -2.94192463e-01 -1.29577148e+00 3.14801455e-01
9.14331079e-01 -4.09046769e-01 -3.28842610e-01 7.66246438e-01
5.51472008e-01 -4.44096655e-01 -7.45301366e-01 1.57825902e-01
2.11005405e-01 -1.06808114e+00 1.43441761e+00 6.83088154e-02
5.10138690e-01 -5.15034974e-01 -3.85228664e-01 -1.43520141e+00
-4.90036756e-01 -8.75424087e-01 -2.31927648e-01 1.01859415e+00
-5.50616123e-02 -7.79776216e-01 2.72361010e-01 -1.68610573e-01
7.92352632e-02 -4.53721941e-01 -1.16148376e+00 -7.55699813e-01
-4.35480289e-02 -3.41600955e-01 3.18342239e-01 7.72449732e-01
-8.34340274e-01 -5.06056510e-02 -7.37501860e-01 6.90219045e-01
1.61503279e+00 2.81330615e-01 4.03878599e-01 -9.90543962e-01
-5.36417305e-01 -7.06116408e-02 1.40252918e-01 -1.68527627e+00
-1.64067894e-01 -4.74933654e-01 3.21295887e-01 -1.91060388e+00
1.47254154e-01 -2.13860601e-01 -1.17148101e-01 1.07455198e-02
-2.86889434e-01 3.15759838e-01 -1.66363731e-01 4.57315803e-01
-2.05418412e-02 7.79889941e-01 1.47366261e+00 -5.15806735e-01
1.68683603e-01 9.14658010e-02 -4.40467179e-01 6.32948935e-01
3.20716590e-01 -8.35091054e-01 -4.43410873e-01 -1.08635485e+00
-2.48780429e-01 3.32477331e-01 6.27650321e-01 -9.19917524e-01
4.37431037e-01 -1.10989057e-01 4.28956747e-01 -6.94159508e-01
5.69619358e-01 -6.98282599e-01 3.10496911e-02 5.52491128e-01
-4.18071933e-02 -4.01376367e-01 -4.29751165e-03 1.15911043e+00
-4.35641110e-01 -3.13639790e-01 1.14292943e+00 -8.48779976e-02
-2.37383872e-01 2.80874133e-01 -6.91388309e-01 -2.11970672e-01
5.05011022e-01 3.93592641e-02 -5.47370434e-01 -5.58741450e-01
-7.86996186e-01 -1.17453732e-01 8.67972672e-02 2.53344383e-02
9.38741982e-01 -9.07736778e-01 -1.20975542e+00 2.51106560e-01
-5.29299140e-01 2.14327183e-02 7.91809320e-01 9.48639929e-01
-5.75686455e-01 -2.94548690e-01 3.04219574e-01 -5.21783113e-01
-1.00121963e+00 4.17141020e-01 2.39230260e-01 -1.81760833e-01
-1.24658275e+00 1.02830982e+00 6.14651255e-02 6.27599433e-02
1.55131996e-01 -2.03410670e-01 -2.87877228e-02 -1.68409497e-01
1.03976154e+00 5.51713288e-01 -1.07018240e-01 -1.32657856e-01
1.42788067e-01 6.77648306e-01 -1.37981758e-01 -2.01366276e-01
1.78870082e+00 -3.11108679e-01 -5.42579472e-01 -2.05861032e-02
9.00580883e-01 7.36200437e-02 -1.64344025e+00 -6.34442329e-01
-4.32923019e-01 -1.03112328e+00 8.71270299e-01 -1.00889850e+00
-1.36545181e+00 9.96531069e-01 9.79448557e-01 2.84530044e-01
1.47842538e+00 -4.36581045e-01 1.04820609e+00 1.58125222e-01
2.96571106e-01 -8.21291327e-01 -7.24040419e-02 2.39621028e-01
1.11615372e+00 -1.03103650e+00 3.93063545e-01 -4.75081325e-01
-2.63961196e-01 9.98367012e-01 -2.21264601e-01 -4.20694977e-01
1.15338826e+00 3.59178215e-01 4.97022979e-02 -1.30371317e-01
-4.68265206e-01 6.43765554e-02 -1.48719743e-01 8.69214475e-01
2.95168068e-02 -1.17479675e-01 -2.71523863e-01 6.10127561e-02
4.85333234e-01 3.14792067e-01 4.41531420e-01 8.28519464e-01
-5.56990564e-01 -1.06059420e+00 -5.95962584e-01 3.63866657e-01
-5.24188817e-01 -5.28612852e-01 5.28198838e-01 1.09299354e-01
-7.00531751e-02 1.49105453e+00 -1.90087676e-01 1.37127399e-01
1.18144043e-01 -3.96328300e-01 6.47998869e-01 -5.12077473e-02
1.90315656e-02 3.74331057e-01 -1.55406758e-01 -4.87347037e-01
-8.71911585e-01 -6.47623241e-01 -6.24478877e-01 -5.13727129e-01
-5.44750690e-01 8.88033137e-02 6.70317709e-01 9.08680022e-01
3.98230940e-01 5.19880712e-01 7.20259786e-01 -1.11389101e+00
-8.44511092e-01 -1.17703140e+00 -8.22734177e-01 5.27169146e-02
8.42979908e-01 -3.95693868e-01 -7.66347349e-01 3.50302994e-01] | [11.588811874389648, -2.391327142715454] |
52a3a9f8-77d1-4ca8-b3a3-665d282dd988 | experimental-exploration-of-compact | 1911.09010 | null | https://arxiv.org/abs/1911.09010v1 | https://arxiv.org/pdf/1911.09010v1.pdf | Experimental Exploration of Compact Convolutional Neural Network Architectures for Non-temporal Real-time Fire Detection | In this work we explore different Convolutional Neural Network (CNN) architectures and their variants for non-temporal binary fire detection and localization in video or still imagery. We consider the performance of experimentally defined, reduced complexity deep CNN architectures for this task and evaluate the effects of different optimization and normalization techniques applied to different CNN architectures (spanning the Inception, ResNet and EfficientNet architectural concepts). Contrary to contemporary trends in the field, our work illustrates a maximum overall accuracy of 0.96 for full frame binary fire detection and 0.94 for superpixel localization using an experimentally defined reduced CNN architecture based on the concept of InceptionV4. We notably achieve a lower false positive rate of 0.06 compared to prior work in the field presenting an efficient, robust and real-time solution for fire region detection. | ['Ganesh Samarth C. A.', 'Toby P. Breckon', 'Neelanjan Bhowmik'] | 2019-11-20 | null | null | null | null | ['fire-detection'] | ['time-series'] | [ 4.74551946e-01 -5.49333930e-01 4.04181093e-01 1.65738583e-01
-1.06488258e-01 -6.47707880e-01 7.19036460e-01 -3.06622714e-01
-1.16475904e+00 7.62414753e-01 -2.94719070e-01 -5.88497035e-02
-1.78951845e-01 -8.61498296e-01 -2.57585168e-01 -7.29164779e-01
-2.92393446e-01 -5.85142784e-02 6.60099983e-01 -3.10826153e-01
2.48222098e-01 8.75059664e-01 -1.63531148e+00 4.42252606e-01
-3.52049172e-02 1.30900180e+00 1.71802282e-01 1.17713439e+00
6.11209035e-01 8.75821829e-01 -9.55665171e-01 3.45262557e-01
8.28929067e-01 -4.88658845e-02 -7.71557927e-01 -1.38628986e-02
8.20046186e-01 -6.97798550e-01 -6.70643508e-01 7.39063442e-01
7.36622453e-01 2.35915169e-01 3.64756554e-01 -1.04842007e+00
-1.40427738e-01 4.32317197e-01 -3.98341328e-01 1.15649521e+00
-5.31849526e-02 7.36148238e-01 4.70548958e-01 -5.10382056e-01
6.74957573e-01 7.17455387e-01 1.13328755e+00 3.20147485e-01
-1.34489763e+00 -7.18453944e-01 -1.79078355e-01 -1.36114554e-02
-1.90819311e+00 -4.42612350e-01 5.62278181e-02 -4.81987685e-01
1.58107841e+00 1.14018679e-01 8.46035659e-01 1.01668549e+00
1.51139736e-01 -2.90586073e-02 1.18885016e+00 -1.99490592e-01
3.05615723e-01 -4.59032059e-01 -8.07315037e-02 6.13638222e-01
3.28355342e-01 5.56871176e-01 -2.71199167e-01 2.58367360e-01
1.44123662e+00 -1.02905169e-01 -4.52640980e-01 5.44930458e-01
-1.33171296e+00 5.24250627e-01 8.01001549e-01 6.74965382e-01
-5.94565630e-01 6.03730381e-01 4.58262414e-01 5.89438453e-02
3.02406222e-01 3.43793064e-01 -4.75721776e-01 4.88963388e-02
-1.66835678e+00 3.93741131e-01 4.78526413e-01 7.46075690e-01
5.43054700e-01 6.73536003e-01 -4.77626234e-01 5.39237201e-01
-1.04473218e-01 3.87661040e-01 3.45612243e-02 -1.22554433e+00
9.68794376e-02 3.95834446e-01 2.01451391e-01 -7.12060332e-01
-7.25715518e-01 -8.55709374e-01 -1.02885485e+00 8.31963122e-01
5.52782118e-01 -3.56796920e-01 -1.29037917e+00 1.16895533e+00
-2.85448819e-01 3.04152787e-01 -1.08667746e-01 1.04338610e+00
8.69836032e-01 4.01358813e-01 2.56194979e-01 2.44124725e-01
1.30033922e+00 -6.14736378e-01 -1.16672680e-01 -3.69361669e-01
5.74894659e-02 -6.10060990e-01 4.25584465e-01 4.79188591e-01
-9.09520030e-01 -7.13388801e-01 -1.29931700e+00 4.44557145e-02
-5.33512115e-01 4.27312762e-01 4.97525305e-01 6.57623947e-01
-1.52842402e+00 9.04685855e-01 -7.48232543e-01 -7.37835526e-01
8.82110596e-01 5.33147752e-01 -2.20223278e-01 1.81629196e-01
-8.27513516e-01 9.39038634e-01 8.53901446e-01 3.31218362e-01
-1.42892218e+00 -5.23919344e-01 -2.28581890e-01 -1.27651903e-03
3.66141573e-02 -8.33521783e-01 1.10732460e+00 -1.10026205e+00
-1.01943898e+00 9.68792915e-01 3.41584027e-01 -9.26880598e-01
5.93692780e-01 9.92833078e-02 -3.00980121e-01 3.53650808e-01
-1.74132630e-01 1.16736686e+00 7.62990057e-01 -8.41582716e-01
-1.13932991e+00 4.34539691e-02 6.32375658e-01 -2.23870382e-01
2.56710321e-01 2.95346946e-01 1.52457908e-01 -7.47885883e-01
-3.97270508e-02 -6.16654754e-01 -3.60527635e-01 2.49057651e-01
-3.75443213e-02 1.08814046e-01 9.09040153e-01 -4.62446928e-01
9.05168355e-01 -1.87506330e+00 -2.09892154e-01 -5.29092252e-02
4.27789539e-01 6.44366801e-01 -1.21042587e-01 -9.66316834e-02
-1.58218354e-01 2.94157416e-01 -6.19253218e-01 1.46009415e-01
-6.01972103e-01 2.12168053e-01 1.42133430e-01 6.08942747e-01
4.53967780e-01 7.39892185e-01 -7.33494639e-01 -1.55682534e-01
6.42725229e-01 9.03627455e-01 -3.39402080e-01 -2.07169130e-01
2.33662874e-02 5.71036004e-02 7.79635981e-02 9.07200098e-01
7.09821463e-01 7.48713091e-02 -2.34993681e-01 -3.72803807e-01
-6.63554788e-01 -6.34758025e-02 -1.02218902e+00 1.48145211e+00
-2.69526482e-01 1.46379435e+00 4.02860045e-01 -7.48287439e-01
8.40058386e-01 4.14398789e-01 5.13871372e-01 -5.55272698e-01
6.22724593e-01 1.45381726e-02 1.11668281e-01 -1.21673830e-01
5.71764231e-01 -1.14949808e-01 6.69137657e-01 1.47821650e-01
5.66792965e-01 2.99938768e-01 4.83228475e-01 -3.07612807e-01
1.54916477e+00 4.01609875e-02 1.74733981e-01 -7.02404499e-01
2.60595322e-01 2.64615327e-01 1.72692910e-02 1.10922551e+00
-6.51334345e-01 1.06868088e+00 7.59698674e-02 -7.79300570e-01
-1.27891600e+00 -8.76884043e-01 -2.85234064e-01 8.98818314e-01
-1.85334057e-01 -3.84705245e-01 -8.34282696e-01 -2.57915616e-01
-4.12949532e-01 1.95528939e-01 -7.81417012e-01 3.22417349e-01
-7.33338475e-01 -1.11885500e+00 1.08370912e+00 7.75305986e-01
1.23501444e+00 -1.38217103e+00 -1.62970448e+00 3.54749650e-01
1.48619162e-02 -1.37766373e+00 3.09924781e-01 6.81484878e-01
-8.33931208e-01 -1.22655821e+00 -7.83690274e-01 -5.42274773e-01
1.88594222e-01 3.20762902e-01 1.28470314e+00 2.35667512e-01
-8.39774191e-01 2.60677487e-01 -5.16617954e-01 -9.82394964e-02
1.81955591e-01 2.10684687e-01 -4.42991227e-01 -4.05326694e-01
3.49037021e-01 -5.70937634e-01 -9.92003202e-01 3.40536982e-02
-1.17343032e+00 -1.75721049e-01 3.25865567e-01 5.99103332e-01
2.23089069e-01 1.61754861e-01 -2.28662387e-01 -3.15303713e-01
3.78421664e-01 -1.51877448e-01 -7.07587600e-01 -3.83352302e-02
-4.15463895e-01 -2.68726051e-01 3.75691324e-01 -1.84378982e-01
-6.75541520e-01 2.23524153e-01 -2.19579160e-01 -4.54014599e-01
-7.07735956e-01 1.08840913e-02 6.90139234e-01 -5.19538164e-01
1.20302892e+00 7.70808682e-02 -3.83986175e-01 6.32218570e-02
1.47048667e-01 2.97030628e-01 8.65695655e-01 -9.50891301e-02
6.63603008e-01 1.00716555e+00 1.43441156e-01 -1.03594446e+00
-4.48022515e-01 -5.22868574e-01 -1.02472687e+00 -6.69496596e-01
1.06247997e+00 -8.88746798e-01 -6.75165832e-01 6.03621781e-01
-1.37871897e+00 -5.49446642e-01 -3.88279319e-01 2.72579253e-01
-5.06292939e-01 -1.05630890e-01 -4.18400347e-01 -8.30573738e-01
-5.84353924e-01 -9.72398818e-01 9.88424659e-01 2.97388107e-01
-8.72429684e-02 -7.58373916e-01 -8.51919428e-02 -2.05722272e-01
9.43761349e-01 7.54298985e-01 -1.03225503e-02 2.76356433e-02
-7.25134015e-01 -8.83392990e-02 -6.67259336e-01 5.03910422e-01
-3.14953536e-01 4.42030609e-01 -1.33555794e+00 -1.29052386e-01
-2.21594796e-01 -4.48924303e-02 1.48282325e+00 6.05699778e-01
7.78297365e-01 2.50249393e-02 -3.21578123e-02 1.01077473e+00
1.95167422e+00 2.85005569e-01 1.07078981e+00 9.84266520e-01
3.39943349e-01 1.42234340e-01 4.11108993e-02 6.99010789e-01
-4.53251690e-01 3.11238796e-01 7.15815902e-01 -2.54580557e-01
-5.15397251e-01 4.15900499e-01 8.01311061e-02 -4.40696985e-01
-9.62277770e-01 -4.58892763e-01 -9.56889093e-01 5.38656950e-01
-1.60842168e+00 -1.35549092e+00 -2.43527800e-01 1.84046543e+00
1.72886997e-02 1.29345939e-01 1.64340317e-01 4.01364177e-01
7.67863214e-01 3.73375714e-01 -1.29789427e-01 2.29800399e-02
-6.88169301e-01 8.98031175e-01 1.27501011e+00 2.21060030e-02
-1.62532067e+00 9.99068737e-01 7.44728279e+00 6.14197075e-01
-1.30123103e+00 2.68780410e-01 5.82315326e-01 -3.76802713e-01
7.76703179e-01 -1.12629337e-02 -5.76026618e-01 -4.31239195e-02
8.24956894e-01 3.51693273e-01 8.14916909e-01 5.55456936e-01
2.66996294e-01 -3.63141447e-01 -4.31530863e-01 1.07672727e+00
-6.38053864e-02 -1.59003258e+00 -2.70313382e-01 -1.90426037e-01
6.34663582e-01 7.84150004e-01 -2.73613155e-01 -6.51094466e-02
1.11244582e-01 -1.28439641e+00 8.79926741e-01 5.04227698e-01
9.03814673e-01 -7.66704857e-01 8.46497536e-01 -7.12802410e-02
-1.43589163e+00 -3.29658449e-01 -5.85709274e-01 -4.89582330e-01
4.22295704e-02 1.95331827e-01 -6.02014363e-01 2.32767850e-01
1.18932426e+00 8.68650317e-01 -7.07201242e-01 1.21384335e+00
5.68335280e-02 4.53083843e-01 -6.58087969e-01 2.54022181e-01
7.91033268e-01 1.64042413e-01 4.37666237e-01 1.71865606e+00
1.23257630e-01 1.16735578e-01 -7.47647062e-02 1.05350363e+00
1.62924498e-01 -6.04879141e-01 -5.08922637e-01 8.47436488e-02
1.72329083e-01 1.64152789e+00 -1.40089643e+00 -3.04641902e-01
-5.71090057e-02 9.41078484e-01 -4.82095964e-02 4.83310223e-01
-7.81220973e-01 -4.26835597e-01 6.80144250e-01 4.12756205e-01
5.80139577e-01 -4.42907482e-01 -3.39999408e-01 -8.11736047e-01
-3.43226194e-01 -3.30702275e-01 3.86934698e-01 -1.00649226e+00
-8.04073811e-01 1.06036580e+00 1.27536952e-01 -1.02959073e+00
-7.17023015e-02 -1.12480581e+00 -7.26631701e-01 6.79419935e-01
-1.26433814e+00 -1.03464675e+00 -6.53526783e-01 6.32315516e-01
5.15970111e-01 -2.13526368e-01 7.56318510e-01 3.84447724e-01
-5.57654798e-01 1.17859207e-01 -1.65525526e-01 4.08555657e-01
1.01596482e-01 -8.33236992e-01 5.45816839e-01 1.30730963e+00
7.59717450e-02 2.39618152e-01 3.97025079e-01 -2.68382549e-01
-6.30652308e-01 -1.34682167e+00 3.14210594e-01 -2.15596303e-01
4.33705270e-01 -1.39998853e-01 -4.14049238e-01 6.25120699e-01
5.28522313e-01 1.53024390e-01 -1.77617639e-01 -3.60535234e-01
-2.60417134e-01 3.02772727e-02 -1.30836427e+00 3.51535827e-01
1.19502687e+00 -5.03672659e-01 -4.24952433e-02 2.26643488e-01
8.44835192e-02 -2.00277820e-01 -6.30863547e-01 5.55077136e-01
6.95650160e-01 -1.41142225e+00 1.16946328e+00 -2.04056546e-01
3.46365392e-01 -5.05274475e-01 -1.85252607e-01 -7.64438331e-01
-7.38121331e-01 -3.01348239e-01 1.75311536e-01 7.07123041e-01
1.66300192e-01 -4.54096019e-01 8.04028869e-01 4.66242284e-02
-1.27592370e-01 -4.53036219e-01 -1.22131205e+00 -6.30277812e-01
-2.92035073e-01 -6.25375688e-01 -9.73752514e-03 5.26011109e-01
-8.18184972e-01 -6.69088364e-02 -6.36789352e-02 1.96577802e-01
4.04455245e-01 -4.34692740e-01 3.16973090e-01 -1.11461759e+00
1.43827409e-01 -8.39832485e-01 -1.00153565e+00 -4.68586951e-01
-2.85587192e-01 -5.99282920e-01 5.73725626e-02 -1.60122037e+00
4.71920632e-02 -1.52151268e-02 -4.60422873e-01 6.98104560e-01
3.85867655e-01 1.05432677e+00 2.24322945e-01 9.21496972e-02
-4.37698841e-01 -1.18014991e-01 6.83003902e-01 -1.53171867e-01
2.56689906e-01 -3.64020228e-01 -7.23539293e-02 7.63395607e-01
1.14460862e+00 -4.91973937e-01 -5.62416448e-04 -5.85786402e-01
-2.84838714e-02 -5.21654308e-01 1.24862099e+00 -1.86651683e+00
2.65106976e-01 1.49046257e-01 7.88136780e-01 -4.33864117e-01
1.98202834e-01 -9.17837083e-01 1.56825989e-01 5.87370753e-01
-7.85885900e-02 3.23410541e-01 6.22450769e-01 1.27735108e-01
3.31778452e-02 -3.18273544e-01 1.19118762e+00 -7.79981315e-01
-1.16414285e+00 2.15778887e-01 -7.93970287e-01 -8.25589970e-02
8.66497636e-01 -6.20139480e-01 -5.76938093e-01 7.98254535e-02
-6.57157063e-01 -4.30975348e-01 3.51520270e-01 1.87988311e-01
7.17477322e-01 -7.60618031e-01 -8.59676778e-01 1.73514888e-01
-2.32661620e-01 -2.25552827e-01 8.36233348e-02 7.52307832e-01
-1.24886739e+00 5.06705940e-01 -9.02811706e-01 -7.33304679e-01
-1.07735240e+00 -7.33665703e-03 1.00461960e+00 2.11669281e-01
-6.71167314e-01 9.17222440e-01 -2.13744193e-01 1.02519639e-01
7.59748742e-02 -5.14046311e-01 -8.62066522e-02 -1.00211069e-01
5.24725795e-01 7.18717813e-01 3.10616463e-01 -6.42153680e-01
-4.73257631e-01 4.27270234e-01 3.08547080e-01 -2.47545063e-01
1.42063236e+00 2.44974792e-01 -2.35868543e-02 8.70502964e-02
7.95784235e-01 -9.75466371e-01 -1.56159008e+00 3.96186739e-01
-2.60475636e-01 -3.41863990e-01 8.00058901e-01 -9.63238239e-01
-1.46538746e+00 7.59370148e-01 1.36879551e+00 1.33354858e-01
1.42158759e+00 -2.96022594e-01 2.97568917e-01 3.71478498e-01
2.58805782e-01 -1.04239368e+00 -4.55970109e-01 8.23761642e-01
7.95828521e-01 -8.64053249e-01 2.10433424e-01 2.46117190e-02
-4.69269939e-02 1.39690685e+00 6.04497254e-01 -6.45727217e-01
6.29974902e-01 6.99300706e-01 5.95297478e-02 -4.51242983e-01
-4.66081202e-01 -8.47804129e-01 -1.08877145e-01 9.25222874e-01
4.27312851e-01 -1.64393559e-01 3.43144052e-02 -2.20817283e-01
-8.02106038e-03 2.80491769e-01 4.19276834e-01 1.18372416e+00
-6.34913385e-01 -4.20599192e-01 -4.25377667e-01 4.67439681e-01
-4.54925597e-01 -3.26496631e-01 -6.09320641e-01 1.26392066e+00
7.14192271e-01 9.81243253e-01 3.89868826e-01 -5.92176497e-01
4.19606149e-01 -1.62699416e-01 7.30010867e-01 -3.18270832e-01
-1.37769425e+00 -5.70377745e-02 -5.23078330e-02 -5.82087994e-01
-9.75207508e-01 -2.61625201e-01 -8.90551031e-01 -4.41677660e-01
-3.48461032e-01 -6.65817142e-01 5.62759042e-01 7.55583704e-01
1.24789841e-01 8.67838085e-01 -8.43827054e-02 -1.38386095e+00
1.87800035e-01 -1.11807227e+00 -7.11373508e-01 -2.21346114e-02
3.39197218e-01 -6.12894773e-01 -3.73855889e-01 1.41291484e-01] | [8.7676420211792, -1.0500773191452026] |
b6571b17-1d86-48e3-8e83-40618fbd422f | pac-hubert-self-supervised-music-source | 2304.02160 | null | https://arxiv.org/abs/2304.02160v1 | https://arxiv.org/pdf/2304.02160v1.pdf | Pac-HuBERT: Self-Supervised Music Source Separation via Primitive Auditory Clustering and Hidden-Unit BERT | In spite of the progress in music source separation research, the small amount of publicly-available clean source data remains a constant limiting factor for performance. Thus, recent advances in self-supervised learning present a largely-unexplored opportunity for improving separation models by leveraging unlabelled music data. In this paper, we propose a self-supervised learning framework for music source separation inspired by the HuBERT speech representation model. We first investigate the potential impact of the original HuBERT model by inserting an adapted version of it into the well-known Demucs V2 time-domain separation model architecture. We then propose a time-frequency-domain self-supervised model, Pac-HuBERT (for primitive auditory clustering HuBERT), that we later use in combination with a Res-U-Net decoder for source separation. Pac-HuBERT uses primitive auditory features of music as unsupervised clustering labels to initialize the self-supervised pretraining process using the Free Music Archive (FMA) dataset. The resulting framework achieves better source-to-distortion ratio (SDR) performance on the MusDB18 test set than the original Demucs V2 and Res-U-Net models. We further demonstrate that it can boost performance with small amounts of supervised data. Ultimately, our proposed framework is an effective solution to the challenge of limited clean source data for music source separation. | ['Jonathan Le Roux', 'François G. Germain', 'Gordon Wichern', 'Ke Chen'] | 2023-04-04 | null | null | null | null | ['music-source-separation'] | ['music'] | [ 4.85847384e-01 -2.47790694e-01 -1.73035219e-01 -1.13680139e-01
-1.38724566e+00 -8.38752270e-01 4.79106277e-01 8.95960480e-02
-2.91898936e-01 4.64200556e-01 4.51560348e-01 -4.53549810e-02
-5.64122617e-01 -2.23483786e-01 -5.54239154e-01 -8.25882435e-01
-1.01826154e-01 2.97172070e-01 7.01287538e-02 -2.19372675e-01
-1.88220236e-02 1.86536074e-01 -1.67300034e+00 2.96680510e-01
8.79756570e-01 8.45905542e-01 2.66881764e-01 9.85860050e-01
3.25663447e-01 7.87276745e-01 -6.12391829e-01 3.05769686e-02
3.89770329e-01 -1.03485930e+00 -6.10654652e-01 -2.26199463e-01
3.13849062e-01 1.29400253e-01 -2.18825966e-01 9.89525914e-01
9.02278185e-01 2.40485117e-01 7.41880119e-01 -1.21256840e+00
-3.22981715e-01 1.41235113e+00 -2.33025566e-01 3.57668519e-01
5.45512401e-02 -2.85230011e-01 1.16505718e+00 -7.25280464e-01
1.97283193e-01 8.81900847e-01 8.10596585e-01 3.95855874e-01
-1.35382557e+00 -7.00768173e-01 -3.67065579e-01 4.61256683e-01
-1.55530107e+00 -1.11367011e+00 1.08074749e+00 -1.53731138e-01
7.76791692e-01 4.48764294e-01 2.63601243e-01 1.18388188e+00
-5.58180571e-01 9.35695112e-01 9.72738564e-01 -7.30157256e-01
5.97855210e-01 -3.55406329e-02 -8.33710879e-02 1.68559894e-01
-8.07564035e-02 1.68687969e-01 -9.72728133e-01 -1.74427241e-01
5.69011033e-01 -5.99349141e-01 -2.24305749e-01 -2.40457430e-01
-1.31392395e+00 4.75592941e-01 2.37781405e-01 5.38282812e-01
-1.69824734e-02 1.21372290e-01 2.78985381e-01 4.18731898e-01
4.45235938e-01 5.44595361e-01 -4.87725675e-01 -2.80514240e-01
-1.62868333e+00 1.02750026e-01 7.62858152e-01 8.22516084e-01
3.34577113e-01 6.57836676e-01 -9.59332064e-02 1.15772533e+00
2.47559845e-01 6.31464005e-01 8.35168242e-01 -1.05031872e+00
1.98781118e-01 -3.15899178e-02 -2.39808127e-01 -7.03205168e-01
-3.35765660e-01 -1.12584674e+00 -8.30025554e-01 1.47649840e-01
3.54421854e-01 -7.32222274e-02 -7.09306955e-01 1.84342837e+00
1.19135074e-01 6.31611228e-01 3.37004960e-01 7.34018087e-01
6.31226003e-01 4.65281367e-01 -4.55430716e-01 -3.72142345e-01
7.78238535e-01 -1.21814096e+00 -6.87928021e-01 -6.18601441e-02
1.88252479e-01 -9.60137844e-01 7.37047255e-01 7.70052254e-01
-1.11975253e+00 -7.84888029e-01 -1.24400592e+00 1.45398527e-01
-1.19074985e-01 3.22007060e-01 2.97890455e-01 9.63435948e-01
-9.59356666e-01 9.67418075e-01 -8.29160154e-01 -1.49754688e-01
3.26555967e-01 2.83060908e-01 -1.36372164e-01 3.47703397e-01
-9.53882277e-01 3.30526888e-01 4.39929813e-01 -2.03391790e-01
-1.17736602e+00 -5.58952689e-01 -6.30870879e-01 1.79992452e-01
3.68098110e-01 -4.97076541e-01 1.29449606e+00 -1.09305549e+00
-1.70096040e+00 5.15716016e-01 -1.17398515e-01 -9.17700410e-01
1.44514278e-01 -2.28128552e-01 -7.40183830e-01 2.95948327e-01
3.54528613e-02 4.48981911e-01 1.34786987e+00 -1.31882274e+00
-6.08067870e-01 -1.89694047e-01 -6.05306506e-01 2.68614888e-01
-3.95685822e-01 1.69820383e-01 -1.38722822e-01 -1.37598002e+00
2.22730309e-01 -9.40449357e-01 1.05618298e-01 -6.60822868e-01
-4.79061842e-01 1.02919668e-01 4.33902651e-01 -6.04262710e-01
1.24961221e+00 -2.37007785e+00 3.96422267e-01 2.87573040e-01
-6.21087216e-02 3.69539350e-01 -4.73706603e-01 4.27867055e-01
-4.90705222e-01 -1.44729570e-01 -6.01227999e-01 -7.66964674e-01
1.15855418e-01 -1.81025080e-02 -5.35372972e-01 3.26652318e-01
-7.44395331e-02 6.24780893e-01 -8.34364355e-01 -3.61124277e-01
-1.87946726e-02 5.65386832e-01 -7.55907178e-01 2.16011435e-01
2.14617532e-02 7.10381150e-01 2.57310152e-01 5.22006273e-01
4.61036772e-01 2.42678002e-01 -1.09258816e-01 -8.65022019e-02
-1.13669664e-01 4.97352183e-01 -1.47908163e+00 2.20261407e+00
-1.69340804e-01 6.01706266e-01 2.37093061e-01 -1.06494772e+00
8.87480140e-01 5.28950632e-01 6.82939768e-01 -3.81276101e-01
2.32441589e-01 5.58252633e-01 2.61885881e-01 -8.30977112e-02
2.34436199e-01 -3.78442109e-01 2.56186333e-02 4.65163589e-01
7.68531621e-01 -2.52712250e-01 1.52688384e-01 1.66296512e-01
1.17309403e+00 1.29246905e-01 1.64242148e-01 -1.74968824e-01
5.44745505e-01 -2.03677699e-01 3.95647526e-01 7.09245861e-01
-2.29166836e-01 1.15874922e+00 3.12037617e-02 3.86892766e-01
-7.59068847e-01 -1.32069361e+00 -9.58097428e-02 1.34889472e+00
-3.11109781e-01 -7.46147573e-01 -9.16258574e-01 -3.24967891e-01
-2.55307853e-01 8.42513442e-01 -1.47934318e-01 -3.41337562e-01
-5.01203299e-01 -6.06730521e-01 1.22765315e+00 4.18467253e-01
1.94059476e-01 -8.95500600e-01 -2.38768622e-01 1.71573237e-01
-4.24367338e-01 -9.86702323e-01 -3.35119486e-01 7.15822339e-01
-7.90862262e-01 -7.98187256e-01 -9.58654702e-01 -8.51396024e-01
-8.56189281e-02 3.55355680e-01 8.13027143e-01 -4.53212947e-01
4.67632227e-02 5.23162305e-01 -6.76634789e-01 -6.63013101e-01
-7.31780529e-01 2.21944556e-01 5.15581667e-01 2.59551436e-01
8.11660215e-02 -1.15637052e+00 -2.06695393e-01 2.45180726e-01
-9.17826355e-01 -1.36940241e-01 3.41570705e-01 6.96004510e-01
6.19256616e-01 5.54504275e-01 1.20814848e+00 -4.06989068e-01
4.28719670e-01 -4.06244695e-01 -1.71255991e-01 -2.44650215e-01
-5.76861143e-01 -3.79505306e-02 7.02755868e-01 -3.99366349e-01
-9.55780685e-01 1.60686508e-01 -3.61769617e-01 -6.72349453e-01
-3.02270353e-01 3.41215998e-01 -3.35544735e-01 3.19848061e-02
9.22018349e-01 3.15040559e-01 -2.96561331e-01 -9.47539270e-01
4.61832434e-01 1.04644585e+00 1.14224482e+00 -4.60289747e-01
1.15045488e+00 3.96293074e-01 -2.61108994e-01 -9.94562089e-01
-8.71496379e-01 -7.33837306e-01 -8.59748781e-01 1.84453771e-01
6.91844821e-01 -1.21881092e+00 -1.93975225e-01 4.87799466e-01
-7.89712608e-01 -3.37672383e-01 -6.82333171e-01 6.50428891e-01
-9.30769145e-01 2.86225885e-01 -3.47679585e-01 -1.03424764e+00
-2.85895675e-01 -6.90207660e-01 1.05729115e+00 -3.19489203e-02
-2.37667426e-01 -7.68008113e-01 5.59409857e-01 4.95545208e-01
4.57964659e-01 -1.58654585e-01 6.51775002e-01 -1.01205695e+00
-1.85260564e-01 6.54490525e-03 3.28092784e-01 9.05459046e-01
2.24114299e-01 -5.31681836e-01 -1.60664570e+00 -1.99486285e-01
4.70497489e-01 -1.86499387e-01 1.10831118e+00 2.93509960e-01
8.82010818e-01 -2.18383342e-01 1.28896728e-01 9.45923090e-01
1.09309947e+00 1.01554692e-01 4.42743659e-01 1.45929679e-01
7.84850419e-01 1.94168448e-01 7.16734007e-02 3.57498020e-01
2.02426016e-01 6.49287343e-01 2.23806605e-01 -2.71832552e-02
-6.72016501e-01 -4.76923704e-01 5.75959921e-01 1.51106834e+00
-1.99447110e-01 2.93724127e-02 -7.64747977e-01 7.35525370e-01
-1.71180403e+00 -1.15032697e+00 8.52513239e-02 2.41005373e+00
1.14434910e+00 -6.73466828e-03 5.01866996e-01 1.13637662e+00
4.80785370e-01 1.08929440e-01 -4.36999887e-01 1.23533189e-01
-5.68213284e-01 5.62655926e-01 2.32410938e-01 3.37119997e-01
-1.30432367e+00 6.65532470e-01 5.95570469e+00 1.23994362e+00
-8.86430144e-01 3.77034456e-01 -1.32730287e-02 -3.30986440e-01
1.41552873e-02 -5.68871237e-02 -4.36832994e-01 3.09302032e-01
1.34919286e+00 -2.18561217e-01 9.54740405e-01 4.94556487e-01
1.50230914e-01 1.48989365e-01 -1.34105742e+00 1.38886702e+00
5.16437173e-01 -1.00193501e+00 -2.69652009e-01 -3.12191397e-01
7.00317740e-01 2.12105572e-01 9.64147151e-02 3.12499464e-01
-1.67676136e-02 -9.18047547e-01 1.07856536e+00 1.14997894e-01
8.00432026e-01 -8.11072230e-01 3.64980876e-01 5.99908113e-01
-1.21953619e+00 -3.01202267e-01 -3.14003825e-01 7.52971470e-02
-2.67762076e-02 4.98721868e-01 -7.75591910e-01 8.23466003e-01
5.13236165e-01 8.76032174e-01 -7.22840428e-01 1.25434196e+00
-1.13474526e-01 1.26185763e+00 -3.39584619e-01 6.64162815e-01
-1.87772423e-01 1.16889201e-01 1.09391999e+00 1.35450733e+00
3.22019935e-01 -2.36690342e-01 -1.33774325e-01 7.01715946e-01
-1.86119433e-02 1.32194549e-01 -3.15173984e-01 -1.48466989e-01
3.55941117e-01 1.07747877e+00 -7.97436833e-01 -1.81912646e-01
1.09510556e-01 1.04265964e+00 4.11867090e-02 3.64125729e-01
-7.46648312e-01 -3.86023372e-01 3.76106441e-01 -2.03778461e-01
4.39304888e-01 -2.28331923e-01 -3.18747759e-01 -1.36579788e+00
-3.06947500e-01 -1.25111997e+00 3.58745605e-01 -8.46460462e-01
-1.37724745e+00 6.81351244e-01 -5.59720695e-02 -1.51052558e+00
-3.23467523e-01 -2.84342885e-01 -5.35388291e-01 7.76495934e-01
-1.51915526e+00 -1.04892516e+00 1.65386096e-01 9.65610445e-01
6.76129878e-01 -7.30333805e-01 1.08128595e+00 4.79444683e-01
-3.85257363e-01 6.75424397e-01 4.48716938e-01 9.00375471e-02
1.05876994e+00 -1.45636499e+00 1.94027960e-01 9.26608860e-01
1.04794896e+00 5.07719755e-01 6.10026836e-01 -4.29046363e-01
-1.30066836e+00 -1.11284137e+00 5.63810587e-01 -4.99057561e-01
6.33005142e-01 -5.11869013e-01 -7.81469285e-01 3.93838376e-01
3.20649981e-01 -2.34601021e-01 1.08140075e+00 1.61152124e-01
-6.32482231e-01 -3.13531846e-01 -8.00809443e-01 4.12281990e-01
1.08595371e+00 -8.79952252e-01 -1.02586150e+00 1.42394826e-01
7.91251361e-01 -7.65000060e-02 -4.89311308e-01 3.17852437e-01
3.97297591e-01 -1.09969413e+00 1.24968004e+00 -3.28254879e-01
1.84839293e-01 -4.74317282e-01 -4.80718285e-01 -1.69612443e+00
-3.36176008e-01 -1.13206947e+00 -2.54719168e-01 1.61449814e+00
3.93501520e-01 -1.18266173e-01 3.96443248e-01 -3.54506791e-01
-3.45555305e-01 6.71988213e-03 -1.10297644e+00 -1.18128562e+00
1.59304067e-01 -9.05210316e-01 3.30031633e-01 1.01575661e+00
2.90713280e-01 6.17040575e-01 -4.29558545e-01 1.53388873e-01
9.34121847e-01 -1.19743068e-02 6.95806682e-01 -1.22869885e+00
-8.27624202e-01 -5.10067582e-01 -2.23630786e-01 -7.63325095e-01
1.72018901e-01 -1.46457469e+00 2.58184582e-01 -1.22239006e+00
-2.51990221e-02 -3.49878699e-01 -7.39986002e-01 4.69078422e-01
-3.12917344e-02 5.34285784e-01 4.72618341e-01 4.73209888e-01
-5.66794515e-01 6.24401271e-01 5.12685537e-01 -1.87816456e-01
-5.23116767e-01 3.05272490e-01 -7.71944165e-01 9.60699677e-01
7.95434594e-01 -7.16952562e-01 -5.52845597e-01 -4.65387881e-01
7.74736255e-02 1.01686157e-02 3.31300378e-01 -1.65333009e+00
5.77912390e-01 4.25335348e-01 1.74300402e-01 -4.95274663e-01
3.84520501e-01 -7.98669457e-01 7.71356421e-03 -9.47564170e-02
-5.67886174e-01 -4.65983331e-01 2.72421330e-01 6.38434529e-01
-4.07211244e-01 -1.32894412e-01 7.30159283e-01 3.01716119e-01
-1.05326138e-01 -3.05893064e-01 -2.49835819e-01 2.00171337e-01
4.33895826e-01 1.61709636e-01 -4.24157875e-03 -5.42984724e-01
-8.70062232e-01 -3.28437150e-01 5.08322828e-02 4.82277453e-01
3.79176527e-01 -1.39228630e+00 -9.86520648e-01 4.81474608e-01
1.79680493e-02 -3.25020313e-01 2.38192603e-01 7.82860637e-01
1.39815226e-01 2.76755750e-01 -1.28073096e-01 -6.07903540e-01
-1.17913043e+00 5.54944992e-01 1.10747270e-01 9.89892185e-02
-4.40658063e-01 9.25821185e-01 -1.42043363e-02 -3.45622778e-01
5.08034766e-01 -1.51135698e-01 -1.96775377e-01 2.12390527e-01
6.45798922e-01 7.12724447e-01 2.22763017e-01 -9.47602510e-01
-3.44698250e-01 4.47213531e-01 6.72536373e-01 -6.58742726e-01
1.50734711e+00 -3.01578522e-01 3.08440533e-02 9.50445473e-01
1.04841530e+00 6.95078135e-01 -8.66491616e-01 -3.29380691e-01
1.06685303e-01 -1.62407547e-01 1.81147501e-01 -1.04813504e+00
-8.02069843e-01 9.83024895e-01 6.44212544e-01 2.50491291e-01
1.48695695e+00 -2.48276703e-02 6.32587016e-01 3.86523813e-01
3.27352077e-01 -9.47965503e-01 -6.17966726e-02 5.69067299e-01
8.37831616e-01 -1.00045860e+00 -3.36227983e-01 -8.46620128e-02
-5.73329747e-01 8.56646419e-01 -1.75941944e-01 -1.79116845e-01
6.87578797e-01 5.46796560e-01 2.57731050e-01 2.68424511e-01
-5.28272212e-01 -5.75456202e-01 6.55704975e-01 8.25144470e-01
3.33766133e-01 -2.21031457e-02 2.19157800e-01 1.26461470e+00
-6.86066568e-01 -7.12610036e-02 3.95507216e-01 5.68784058e-01
-4.40579832e-01 -1.20347106e+00 -7.18399465e-01 9.01837945e-02
-5.14242291e-01 -4.52084839e-01 -5.29677093e-01 8.17711651e-02
2.18478814e-01 1.51567781e+00 -3.47173452e-01 -4.80243146e-01
2.02717483e-01 4.46861863e-01 4.68396991e-01 -7.44016588e-01
-6.80726409e-01 9.56274629e-01 -9.78611633e-02 -1.69898421e-01
-7.17512429e-01 -5.97750127e-01 -1.12213910e+00 3.41570497e-01
-2.72531450e-01 2.33969033e-01 6.38676226e-01 8.52652788e-01
3.11685145e-01 7.80110300e-01 7.73146510e-01 -1.15630364e+00
-5.59499621e-01 -1.04269159e+00 -8.29437554e-01 1.79089084e-01
6.25423133e-01 -3.64370525e-01 -6.31861210e-01 5.51968992e-01] | [15.467473030090332, 5.519401550292969] |
df52d006-25e6-47c8-813c-b2e2ce92d923 | uni6dv3-5d-anchor-mechanism-for-6d-pose | 2210.10959 | null | https://arxiv.org/abs/2210.10959v5 | https://arxiv.org/pdf/2210.10959v5.pdf | Geo6D: Geometric Constraints Learning for 6D Pose Estimation | Existing direct 6D pose estimation methods regress target 6D poses without the need for post-processing, making them effective and easy to develop. However, due to the lack of geometric constraints, the precision of these approaches remains limited, and more training data are required to achieve better performance. To this end, we propose a geometric constraint-based 6D pose estimation method (Geo6D), which establishes an explicit geometric constraint between the input and the regression target. Specifically, a Geo head is proposed to build a point-to-point relationship between the camera frame and the object frame. And additional inputs are introduced as supplementary information to facilitate the learning of the geometric correlation. The proposed Geo6D can be used as a plugin in mainstream direct 6D pose estimation methods. Extensive experimental results show that when equipped with Geo6D, the direct 6D method achieves state-of-the-art performance on multiple datasets and shows significant effectiveness with limited data volume. | ['Xiaoke Jiang', 'Liwei Wu', 'Rui Zhao', 'Guoqiang Jin', 'Donghai Li', 'Zhenyu He', 'Tianpeng Bao', 'Ye Zheng', 'Mingshan Sun', 'Jianqiu Chen'] | 2022-10-20 | null | null | null | null | ['6d-pose-estimation-1'] | ['computer-vision'] | [-3.13167542e-01 -1.90723464e-01 -2.67991513e-01 -3.94061506e-01
-6.01554096e-01 -4.04695988e-01 5.94795823e-01 -1.38954833e-01
-1.65107876e-01 1.16238944e-01 -4.49252911e-02 -5.28808795e-02
-8.32443312e-03 -5.08771837e-01 -8.58914137e-01 -5.01923859e-01
2.48760745e-01 4.37406540e-01 3.50695699e-01 -5.18910140e-02
2.73284614e-01 5.38634479e-01 -1.16846240e+00 -4.86971676e-01
6.49284482e-01 1.15185368e+00 4.52963531e-01 1.66952431e-01
2.67288774e-01 1.88314080e-01 -3.79063129e-01 -2.05334172e-01
6.05306983e-01 -5.24443686e-02 5.43689989e-02 1.93237618e-01
4.99726772e-01 -6.09411597e-01 -4.03197765e-01 7.49669909e-01
6.31076515e-01 1.22524098e-01 6.17066145e-01 -1.06476557e+00
-4.72880788e-02 -1.44866899e-01 -7.56853402e-01 -2.84877568e-01
6.17720604e-01 4.39287014e-02 5.86788356e-01 -1.47534525e+00
5.34912527e-01 1.11851203e+00 6.51261687e-01 4.00534183e-01
-8.78296494e-01 -8.76653373e-01 2.83546716e-01 -1.08034186e-01
-1.55470157e+00 -4.60910559e-01 9.95823741e-01 -3.14329535e-01
7.31346130e-01 1.02839554e-02 7.04243720e-01 1.00539970e+00
7.76114538e-02 7.73421824e-01 7.38053143e-01 -2.58735418e-01
-5.25697693e-02 -1.93373442e-01 -2.88753062e-01 6.80337310e-01
2.64227122e-01 1.47864774e-01 -5.98995268e-01 1.41581431e-01
1.17796695e+00 1.75420761e-01 -5.97805083e-02 -1.03608227e+00
-1.17029107e+00 6.47598386e-01 7.07324445e-01 -8.51575434e-02
-4.61072754e-03 1.39743805e-01 8.36336687e-02 -2.18472540e-01
5.99890411e-01 2.72913575e-01 -3.62309515e-01 -8.53755847e-02
-5.65328658e-01 4.09700930e-01 3.63106787e-01 1.26605821e+00
5.91237068e-01 -1.79492235e-01 9.99960303e-02 5.75741112e-01
5.94223499e-01 6.94725096e-01 1.25354245e-01 -7.78351247e-01
8.92205298e-01 8.90719354e-01 2.11785883e-01 -1.37919617e+00
-4.90189970e-01 -6.25524640e-01 -5.82449794e-01 7.31229559e-02
4.36847687e-01 5.04482165e-03 -6.26681805e-01 1.36440229e+00
8.17165613e-01 3.99762064e-01 -3.63938749e-01 1.27781880e+00
6.91138685e-01 4.80114996e-01 -3.54591012e-01 -1.05044007e-01
8.65512133e-01 -8.17535400e-01 -4.66678321e-01 -3.96919101e-01
5.44709980e-01 -9.99926686e-01 9.17747974e-01 7.53551573e-02
-8.10277820e-01 -6.06579065e-01 -1.01156223e+00 -1.40263855e-01
-2.53856871e-02 4.67564881e-01 6.65992141e-01 2.92111367e-01
-4.02907401e-01 2.16585528e-02 -9.82530177e-01 -1.90552115e-01
1.80951953e-01 5.23920059e-01 -4.55466181e-01 -1.14813067e-01
-6.84852302e-01 1.09115076e+00 3.74856174e-01 3.10633153e-01
-6.30932093e-01 -6.48334086e-01 -9.87010181e-01 -3.05878699e-01
7.71587789e-01 -7.00365186e-01 1.13292551e+00 4.27673161e-02
-1.61915135e+00 8.60858977e-01 4.34964634e-02 -5.97972088e-02
7.95148730e-01 -7.58806705e-01 1.59724951e-01 -9.12119262e-03
-6.66035190e-02 4.20750320e-01 8.65007460e-01 -1.24993348e+00
-4.24334317e-01 -6.08712494e-01 2.27610350e-01 5.94474733e-01
-1.77968621e-01 -2.74689704e-01 -1.27793717e+00 -7.27533221e-01
6.17911041e-01 -1.19903851e+00 -3.00155401e-01 2.95111984e-01
-4.30942774e-01 -1.50291726e-01 1.07585776e+00 -4.29511130e-01
9.80428636e-01 -2.33610868e+00 2.63311088e-01 3.26415986e-01
2.16033801e-01 3.21323484e-01 -6.11195378e-02 1.81771830e-01
1.79181859e-01 -3.97444904e-01 6.50769323e-02 -5.92793465e-01
-1.84418932e-01 -7.09332293e-03 2.98466142e-02 6.45023763e-01
9.87399742e-02 7.52711833e-01 -8.46689582e-01 -3.88294756e-01
8.29279065e-01 5.53958893e-01 -5.60626626e-01 4.40655172e-01
-7.58365616e-02 5.52854598e-01 -7.74295866e-01 5.63810050e-01
9.13713753e-01 -5.39680421e-02 -6.22712113e-02 -5.04856825e-01
-2.85384823e-02 3.39152426e-01 -1.23371434e+00 2.06376648e+00
-5.03724754e-01 1.04328431e-01 -8.27535838e-02 -6.47393584e-01
1.30238461e+00 -5.20772040e-02 5.98520160e-01 -4.94549066e-01
5.42817593e-01 1.82267562e-01 -2.81918883e-01 -1.65385231e-01
2.06548586e-01 1.96768105e-01 -1.52947053e-01 4.28978503e-02
-1.84446692e-01 -5.99465609e-01 -1.38290137e-01 -8.33193585e-03
6.48839295e-01 5.69312871e-01 5.13001204e-01 1.05420388e-01
5.50019085e-01 -1.19895354e-01 6.49090111e-01 2.21161455e-01
1.53197110e-01 8.39211762e-01 6.93888441e-02 -2.18083128e-01
-9.59244013e-01 -1.14396012e+00 8.88692122e-03 2.11293921e-01
6.34560049e-01 -5.71752071e-01 -5.39252162e-01 -6.64958894e-01
2.72164047e-01 2.51929760e-01 -3.01370144e-01 -1.39410093e-01
-7.33489811e-01 -4.54382658e-01 3.22794728e-02 6.40426278e-01
6.25035346e-01 -3.43943611e-02 -4.88990694e-01 9.62963328e-02
-1.12554833e-01 -1.41327465e+00 -6.51562035e-01 -1.52807489e-01
-1.06852615e+00 -1.17572844e+00 -6.18616819e-01 -6.44274116e-01
7.82242656e-01 7.27288604e-01 5.13846338e-01 -2.20721029e-02
7.99628049e-02 7.80645013e-02 -3.86407614e-01 -3.13088447e-01
1.97530866e-01 1.77724898e-01 3.07856470e-01 -1.73038602e-01
2.19518468e-01 -4.78804797e-01 -7.42019951e-01 8.15450788e-01
-3.72992873e-01 4.01150733e-01 6.33604527e-01 6.41938448e-01
6.50743246e-01 -2.06431702e-01 3.13572064e-02 -4.06540066e-01
3.00289062e-03 -8.98641869e-02 -8.02089751e-01 -2.33884811e-01
-3.44587177e-01 -1.66016087e-01 4.29723859e-01 -4.37164545e-01
-9.68161285e-01 6.03940368e-01 -6.98680282e-02 -8.01078618e-01
1.05330028e-01 6.08366191e-01 -5.28384209e-01 -3.27909678e-01
3.30430448e-01 -1.53188497e-01 1.74102321e-01 -7.96664357e-01
1.91335320e-01 4.67914045e-01 5.38961470e-01 -4.79397595e-01
1.38046265e+00 3.45803797e-01 2.93397188e-01 -5.85240126e-01
-1.04490960e+00 -6.43210649e-01 -9.51109290e-01 -2.53721088e-01
6.26272261e-01 -1.13536668e+00 -6.15068913e-01 5.56384742e-01
-1.07412279e+00 -3.71972062e-02 2.55273342e-01 9.28409815e-01
-4.99261498e-01 1.80520937e-01 -2.82406390e-01 -5.71995556e-01
-4.28104773e-02 -1.16402376e+00 1.40659988e+00 6.83214217e-02
-8.89680609e-02 -6.44676149e-01 -2.10886881e-01 4.55366343e-01
-9.69023407e-02 4.51366335e-01 4.87260699e-01 -8.68771002e-02
-8.01399410e-01 -5.58579326e-01 -2.56448746e-01 9.26945582e-02
1.96140811e-01 -7.48827532e-02 -5.83936751e-01 -3.94068807e-01
7.77365640e-02 -9.06621814e-02 3.40334028e-01 3.63916039e-01
1.16724515e+00 5.42005189e-02 -4.23586786e-01 9.33624208e-01
1.35972297e+00 4.63163592e-02 4.29346859e-01 4.05431241e-01
9.95246649e-01 5.04789829e-01 1.15981817e+00 4.08205837e-01
3.96165699e-01 1.20985818e+00 7.30724752e-01 6.09235317e-02
-7.96128884e-02 -7.07956135e-01 2.73411900e-01 1.03566039e+00
-2.72629738e-01 -6.71687536e-03 -7.78522253e-01 7.09561631e-02
-1.78638911e+00 -4.37523395e-01 -2.23538861e-01 2.42534089e+00
4.86893862e-01 3.74960065e-01 5.77044338e-02 1.26768202e-01
5.50364137e-01 2.84531802e-01 -7.21420467e-01 3.57635885e-01
3.28244120e-01 -3.98461781e-02 4.72378522e-01 4.43181247e-01
-1.09016669e+00 1.13037896e+00 5.61094332e+00 9.37774837e-01
-1.15767777e+00 -2.53350019e-01 8.45668912e-02 -6.24379329e-02
1.68815300e-01 2.18024254e-02 -1.07793772e+00 2.84600824e-01
1.93365604e-01 1.24711981e-02 -4.33158642e-03 1.03176284e+00
4.21694100e-01 -1.52990505e-01 -1.12189829e+00 1.39730990e+00
1.07361034e-01 -9.70519543e-01 -1.53357863e-01 5.09839244e-02
5.13429046e-01 -3.28974366e-01 3.11905332e-02 -1.09217577e-02
-1.59834728e-01 -6.95476472e-01 7.00277209e-01 2.75932848e-01
9.58576620e-01 -8.14536631e-01 5.88934779e-01 6.39921665e-01
-1.43474185e+00 1.70793682e-01 -4.14289892e-01 -2.73159742e-01
2.14216977e-01 6.75737500e-01 -9.60238874e-01 7.74818420e-01
5.43170929e-01 9.69927013e-01 -5.72885633e-01 1.12662065e+00
-5.79765260e-01 1.36051863e-01 -5.12653053e-01 -6.47103339e-02
1.96284913e-02 -1.64726436e-01 6.30607545e-01 6.76804841e-01
5.08416533e-01 1.38046056e-01 4.32446301e-01 3.79201919e-01
1.36316121e-01 5.38156144e-02 -5.86215138e-01 2.60749340e-01
5.28201640e-01 1.13586867e+00 -7.18283534e-01 6.24149628e-02
-3.60594958e-01 7.90432513e-01 1.13816924e-01 6.05365485e-02
-9.38779235e-01 -2.24201515e-01 6.81900799e-01 2.51194149e-01
2.36565724e-01 -8.21190894e-01 -3.77358198e-01 -1.33941281e+00
3.01198870e-01 -7.76848555e-01 2.02470347e-02 -9.49406385e-01
-1.08873677e+00 1.16914324e-01 2.76003510e-01 -1.83795643e+00
-1.76692665e-01 -8.19894671e-01 -2.15397134e-01 5.82248449e-01
-1.11961734e+00 -1.30652666e+00 -6.81503892e-01 5.63354194e-01
5.89818478e-01 8.43079109e-03 5.89161158e-01 2.26616055e-01
-4.28066283e-01 6.14471972e-01 -2.80369014e-01 1.67929709e-01
8.30728352e-01 -7.76324987e-01 3.80516022e-01 7.29765773e-01
2.51818094e-02 6.87029839e-01 5.71050227e-01 -7.01929986e-01
-1.89314592e+00 -1.02128649e+00 6.50636971e-01 -7.02426076e-01
3.71726513e-01 -7.35210180e-01 -3.67511511e-01 5.00577748e-01
-6.16245806e-01 1.15409508e-01 2.39561632e-01 -3.01008970e-02
-2.26296231e-01 -2.87167728e-01 -7.90071905e-01 6.72043324e-01
1.35627651e+00 -2.97557831e-01 -6.59462273e-01 1.06485933e-01
7.08655775e-01 -1.25246429e+00 -6.71027958e-01 6.53395057e-01
7.36833692e-01 -7.08497941e-01 1.34857631e+00 8.83693695e-02
3.57567459e-01 -7.51858592e-01 -1.68793827e-01 -1.13705659e+00
-1.09328307e-01 -3.30446362e-01 -4.03494030e-01 1.02997577e+00
6.68842271e-02 -3.69155884e-01 9.85328138e-01 4.32795286e-01
-1.71923384e-01 -8.14536095e-01 -8.66421521e-01 -8.59617114e-01
-3.45982969e-01 -5.49520433e-01 4.90186095e-01 6.69110358e-01
-3.58319819e-01 5.48134565e-01 -5.47768354e-01 4.94155109e-01
5.73868990e-01 1.79591283e-01 1.47808802e+00 -1.12582374e+00
-1.05712660e-01 -1.10694110e-01 -7.24794686e-01 -1.95714271e+00
-9.17657167e-02 -5.07752836e-01 1.04242869e-01 -1.35846949e+00
-7.29275495e-02 -6.27493382e-01 9.18459669e-02 2.66989172e-01
-2.63788521e-01 3.20674419e-01 3.47208560e-01 2.43796840e-01
-3.17791879e-01 8.23006749e-01 1.57747459e+00 3.91841084e-02
-3.17199111e-01 3.50277096e-01 -4.24287438e-01 8.24815035e-01
5.38333356e-01 -3.30225736e-01 -4.43035036e-01 -4.94912982e-01
9.86636430e-03 3.92688401e-02 4.16033536e-01 -1.04950142e+00
2.08725169e-01 -2.10912853e-01 7.07836688e-01 -1.21210277e+00
6.88822746e-01 -1.16501653e+00 1.12902410e-01 3.71957988e-01
1.83593065e-01 5.92390969e-02 -8.70903134e-02 4.82887506e-01
-1.75967813e-01 -9.60259233e-03 5.30503690e-01 7.42303878e-02
-6.50384665e-01 7.73605943e-01 3.00540060e-01 -2.26331472e-01
1.27479005e+00 -4.26591069e-01 1.04275361e-01 -3.48449111e-01
-2.18360052e-01 2.83251524e-01 6.28924847e-01 5.50505280e-01
7.67540276e-01 -1.50087380e+00 -3.61714900e-01 1.99553370e-01
3.33899409e-01 7.06324220e-01 -2.73005310e-02 8.91760051e-01
-5.05188227e-01 3.15005630e-01 4.86206524e-02 -1.02849424e+00
-1.37841332e+00 4.70300704e-01 7.67798647e-02 7.25408196e-02
-6.93144441e-01 6.19255960e-01 3.71445090e-01 -6.54480577e-01
2.90048689e-01 -4.15082961e-01 -7.09373355e-02 -1.04110315e-01
3.52512121e-01 2.46181250e-01 2.48400476e-02 -7.51531780e-01
-5.31897187e-01 1.13881969e+00 8.08095932e-02 8.52101594e-02
1.47304070e+00 -2.72695780e-01 3.20517778e-01 3.14810097e-01
1.28000391e+00 1.27809674e-01 -1.70668149e+00 -1.68281123e-01
-2.71701664e-01 -1.13129866e+00 3.92007343e-02 -3.15925181e-01
-1.07267821e+00 8.60705554e-01 4.63963479e-01 -5.51977873e-01
9.50997829e-01 -1.28524125e-01 8.05304646e-01 3.76488894e-01
6.69951379e-01 -9.26473200e-01 3.17055583e-01 5.96074224e-01
9.50019240e-01 -1.24469948e+00 4.08181489e-01 -1.08999527e+00
-1.06496215e-01 1.19715130e+00 9.23688948e-01 -3.23815346e-01
5.37447512e-01 1.58063203e-01 7.56926760e-02 -6.67526945e-02
-8.97610113e-02 -3.32970284e-02 6.68871820e-01 6.16390705e-01
2.15169415e-01 -1.26045182e-01 -3.36102843e-01 2.37657234e-01
-2.20636308e-01 -2.79254138e-01 -5.41230626e-02 9.24911797e-01
-2.01431781e-01 -1.22980785e+00 -4.78091568e-01 1.62619576e-01
3.27283256e-02 3.30907404e-01 -2.64056057e-01 1.10737097e+00
1.42236784e-01 7.19034910e-01 -1.44223705e-01 -6.81148708e-01
6.01693928e-01 -4.38024789e-01 7.50389338e-01 -7.39544749e-01
-4.79420461e-02 4.11618710e-01 -1.05874024e-01 -7.92889118e-01
-5.92949390e-01 -6.68704569e-01 -1.09242475e+00 -2.19735801e-01
-6.95007920e-01 -2.19629720e-01 7.53881037e-01 1.02451849e+00
3.16487730e-01 4.49726619e-02 8.15548360e-01 -1.34287202e+00
-4.41825271e-01 -7.88184524e-01 -2.05909029e-01 1.99163362e-01
1.70716301e-01 -1.17721856e+00 -2.56823331e-01 -1.21919475e-01] | [7.546899318695068, -2.6276497840881348] |
33026028-36ec-4246-88a6-a880d7e29431 | ood-cv-v2-an-extended-benchmark-for | 2304.10266 | null | https://arxiv.org/abs/2304.10266v1 | https://arxiv.org/pdf/2304.10266v1.pdf | OOD-CV-v2: An extended Benchmark for Robustness to Out-of-Distribution Shifts of Individual Nuisances in Natural Images | Enhancing the robustness of vision algorithms in real-world scenarios is challenging. One reason is that existing robustness benchmarks are limited, as they either rely on synthetic data or ignore the effects of individual nuisance factors. We introduce OOD-CV-v2, a benchmark dataset that includes out-of-distribution examples of 10 object categories in terms of pose, shape, texture, context and the weather conditions, and enables benchmarking of models for image classification, object detection, and 3D pose estimation. In addition to this novel dataset, we contribute extensive experiments using popular baseline methods, which reveal that: 1) Some nuisance factors have a much stronger negative effect on the performance compared to others, also depending on the vision task. 2) Current approaches to enhance robustness have only marginal effects, and can even reduce robustness. 3) We do not observe significant differences between convolutional and transformer architectures. We believe our dataset provides a rich test bed to study robustness and will help push forward research in this area. Our dataset can be accessed from http://www.ood-cv.org/challenge.html | ['Adam Kortylewski', 'Alan Yuille', 'Christian Theobalt', 'Oliver Zendel', 'Shaozuo Yu', 'Siwei Yang', 'Artur Jesslen', 'Wufei Ma', 'Jiahao Wang', 'Bingchen Zhao'] | 2023-04-17 | null | null | null | null | ['3d-pose-estimation'] | ['computer-vision'] | [-1.02166720e-01 -5.65027058e-01 1.82323664e-01 -2.70591527e-01
-5.95908165e-01 -1.03851509e+00 8.79486024e-01 -6.74420074e-02
-4.87730265e-01 3.59116018e-01 2.88115233e-01 -7.85375610e-02
2.30807319e-01 -3.72188747e-01 -8.75319004e-01 -6.00183010e-01
-2.18787611e-01 3.91703881e-02 6.01550937e-01 -3.07855219e-01
3.45221043e-01 7.55314887e-01 -1.58914113e+00 -2.23327614e-02
4.03497636e-01 9.14045691e-01 -1.08218184e-02 8.47682059e-01
6.34416044e-01 4.34122443e-01 -8.71362090e-01 -3.99471402e-01
7.58467019e-01 -7.46539384e-02 -4.47602540e-01 -5.58428988e-02
1.09217823e+00 -4.85608250e-01 -6.63397551e-01 9.75257218e-01
9.22759056e-01 7.76810497e-02 6.52907312e-01 -1.33611619e+00
-7.39191294e-01 1.72975093e-01 -6.76862717e-01 1.45779893e-01
3.23971719e-01 8.81525517e-01 6.22221708e-01 -6.65011525e-01
6.36347234e-01 1.39650142e+00 9.53424275e-01 6.26959562e-01
-1.30350304e+00 -5.54576159e-01 1.71237126e-01 1.26769245e-01
-1.13513982e+00 -6.52224958e-01 5.24217248e-01 -4.62461442e-01
8.93474042e-01 3.41472566e-01 2.13389605e-01 1.64007330e+00
1.25590757e-01 4.78052139e-01 1.20565808e+00 -9.88754779e-02
1.65140450e-01 -4.89762165e-02 1.10086482e-02 4.16798621e-01
3.24681759e-01 6.24201179e-01 -3.44167888e-01 9.95244179e-03
4.16403323e-01 -2.94300467e-01 -4.44900304e-01 -5.71433306e-01
-1.40627980e+00 4.87936258e-01 6.53536022e-01 -1.97368905e-01
1.40848368e-01 1.73783213e-01 5.43395340e-01 3.86377275e-01
1.46630153e-01 6.79576457e-01 -5.38752913e-01 -9.31928009e-02
-5.60877681e-01 6.85326636e-01 6.28042102e-01 7.68756628e-01
3.47401679e-01 6.25696406e-02 -3.10046494e-01 7.11413264e-01
1.06296018e-01 8.47802818e-01 1.37037158e-01 -1.09054995e+00
3.75384897e-01 1.85387537e-01 3.41770411e-01 -1.12073672e+00
-5.90648293e-01 -3.54334831e-01 -6.30720556e-01 7.53932178e-01
6.82270050e-01 -8.89707729e-02 -1.18092000e+00 1.88052392e+00
1.35510266e-01 -8.05716671e-04 -1.10644728e-01 1.13662696e+00
1.18676043e+00 1.77134201e-01 -8.22146833e-02 2.64749229e-01
1.16294599e+00 -8.36259723e-01 -2.65490264e-01 -4.02481943e-01
2.00123385e-01 -1.18991613e+00 1.26332080e+00 2.81085998e-01
-9.03039634e-01 -6.97141230e-01 -1.04047287e+00 1.56625047e-01
-5.59033751e-01 4.31331359e-02 3.99274975e-01 7.09899783e-01
-1.10756457e+00 4.89964992e-01 -6.11967027e-01 -5.67917824e-01
3.49580079e-01 9.09974426e-02 -5.12780786e-01 -1.89227343e-01
-9.22178805e-01 1.22505677e+00 1.90214187e-01 -1.79543979e-02
-1.15626025e+00 -6.98092699e-01 -7.84399033e-01 -6.09387815e-01
4.37410325e-01 -5.95481634e-01 1.22572899e+00 -5.63431084e-01
-1.11696613e+00 7.85068333e-01 2.74757296e-01 -3.79841596e-01
9.25356984e-01 -3.17281216e-01 -3.16948116e-01 -7.32382014e-02
-6.19541295e-02 1.05739868e+00 9.90071177e-01 -1.39664149e+00
-2.35996380e-01 -2.78666586e-01 1.84995338e-01 1.16248704e-01
1.48842297e-03 5.10172248e-02 -7.65029609e-01 -8.19078863e-01
-2.56575733e-01 -1.19601476e+00 -1.53268337e-01 1.84651047e-01
-5.69594860e-01 1.21000610e-01 1.03306878e+00 -4.92113113e-01
5.61399102e-01 -2.15658784e+00 1.19433291e-02 -1.81010589e-01
3.33920047e-02 3.76526624e-01 -4.37087178e-01 2.75708199e-01
-9.09067988e-02 2.82843232e-01 -1.31276160e-01 -1.60774484e-01
1.38359759e-02 -1.70716301e-01 -4.08652961e-01 6.12634361e-01
4.56282347e-01 7.75501788e-01 -5.39672494e-01 -4.70069014e-02
6.25872791e-01 4.38777715e-01 -3.57970983e-01 5.79800457e-02
-1.87891811e-01 5.06829679e-01 -5.89343682e-02 7.07021832e-01
9.72619116e-01 1.60297975e-01 -4.11049068e-01 -5.85179925e-01
-1.40474588e-01 4.07658741e-02 -1.25242496e+00 1.33096981e+00
-2.27777526e-01 7.84912109e-01 -9.78230387e-02 -4.36464399e-01
6.48576617e-01 -8.90334547e-02 1.11123458e-01 -7.57107496e-01
1.10487662e-01 -4.01360095e-02 1.47332579e-01 -4.57978249e-01
5.79724014e-01 3.93551856e-01 2.87875421e-02 1.94535837e-01
-9.49316621e-02 -4.31098104e-01 3.13911140e-01 2.36726567e-01
1.30119777e+00 1.62148222e-01 -7.14088455e-02 -2.43929625e-01
9.51623265e-03 -1.47091001e-01 3.78105998e-01 1.01255131e+00
-6.14945650e-01 1.13071215e+00 4.75078166e-01 -5.49870372e-01
-1.08229887e+00 -1.23360109e+00 -3.65105301e-01 6.78920686e-01
3.44995230e-01 -2.89955437e-01 -5.61766744e-01 -7.05111861e-01
1.29761487e-01 5.81419826e-01 -8.27598274e-01 -2.74004638e-01
-2.37393498e-01 -9.89258349e-01 8.22986484e-01 6.43318892e-01
5.25662541e-01 -8.89894903e-01 -7.57598937e-01 -4.54309613e-01
-2.27835570e-02 -1.39885342e+00 -4.18895274e-01 -6.84310347e-02
-5.98228097e-01 -1.24501646e+00 -4.65492427e-01 -3.49210560e-01
4.10969466e-01 7.11103916e-01 1.40712655e+00 1.27427205e-01
-4.25552040e-01 5.05038142e-01 -4.50887859e-01 -6.79150403e-01
-2.42663443e-01 -1.30166173e-01 3.66824269e-01 -4.19565052e-01
6.39642179e-02 -2.76838779e-01 -7.81115413e-01 7.99694121e-01
-8.60792637e-01 -2.67572522e-01 4.56880420e-01 5.13727546e-01
3.37594479e-01 -7.95795768e-02 3.85945082e-01 -4.01587218e-01
4.66084003e-01 -1.76508233e-01 -8.17689717e-01 7.04888254e-02
-3.44446450e-01 -1.04441032e-01 2.97720283e-01 -5.15810370e-01
-7.93088675e-01 1.05884895e-01 8.41562729e-03 -5.27505755e-01
-4.70022529e-01 -1.18649095e-01 -2.11824656e-01 -3.49049717e-01
1.15525365e+00 -1.24064751e-01 -1.18803367e-01 -4.31505084e-01
3.45521867e-01 2.12284744e-01 7.91563034e-01 -5.44158280e-01
1.18546188e+00 4.41974401e-01 4.46433621e-03 -9.20828640e-01
-5.09809792e-01 -9.82551053e-02 -4.20682758e-01 -2.33078107e-01
7.27945447e-01 -9.95900810e-01 -6.20280504e-01 8.87849510e-01
-1.00946844e+00 -6.22802794e-01 -3.02636879e-03 5.23915946e-01
-4.06061739e-01 4.61485445e-01 -4.52984273e-01 -5.32876194e-01
1.00823920e-02 -1.45171368e+00 1.05063224e+00 3.49651992e-01
-9.72151384e-02 -6.02841735e-01 1.03641897e-02 3.59751582e-01
5.46904147e-01 4.71054614e-01 4.74535376e-01 -2.79732198e-01
-4.53462243e-01 -2.56061196e-01 -3.97181034e-01 3.54776263e-01
-9.46296975e-02 4.54626113e-01 -1.19364107e+00 -6.06998920e-01
-3.73833120e-01 -9.12532508e-01 1.17004144e+00 3.47398579e-01
1.18456256e+00 3.70010100e-02 -3.26448902e-02 7.56315470e-01
1.28299975e+00 -3.20074141e-01 8.32497597e-01 3.54318619e-01
6.23041987e-01 6.21640086e-01 5.47820449e-01 2.21392885e-01
1.98536426e-01 9.82965291e-01 8.97424638e-01 -9.66964737e-02
-4.56062257e-01 -4.20864038e-02 5.27167976e-01 -2.71106753e-02
-2.76387751e-01 -3.77983391e-01 -1.02537906e+00 3.77022773e-01
-1.53588653e+00 -8.52551758e-01 -1.48884654e-01 2.34866023e+00
6.03874326e-01 3.36834610e-01 3.20960253e-01 3.12863737e-02
6.34369195e-01 2.55807996e-01 -6.61388934e-01 -4.32521515e-02
-3.91476125e-01 -1.73042029e-01 7.57870376e-01 3.44009817e-01
-1.38673115e+00 9.16191459e-01 6.75479269e+00 6.28183126e-01
-1.28821027e+00 -3.31290841e-01 5.18003047e-01 -3.94888222e-01
9.02673677e-02 -4.40614671e-01 -7.66355097e-01 4.68201905e-01
5.35539567e-01 4.33345377e-01 4.15312380e-01 6.92923903e-01
1.96756512e-01 -2.80334920e-01 -1.17325056e+00 8.10213327e-01
1.51939481e-01 -9.97595251e-01 -2.23599836e-01 -2.68900115e-02
7.12057114e-01 5.63859344e-01 4.77914214e-01 2.43132308e-01
4.33716565e-01 -1.02952635e+00 8.96902859e-01 3.95472527e-01
7.83707678e-01 -5.92315137e-01 7.55446374e-01 1.32384747e-01
-8.26863050e-01 5.31065650e-02 -4.71357167e-01 5.09309508e-02
-2.53365070e-01 4.28651661e-01 -6.93616271e-01 3.15163106e-01
1.37711239e+00 4.40446973e-01 -1.30182815e+00 1.28749144e+00
-2.52972513e-01 5.71979880e-01 -3.87156039e-01 1.77126974e-01
2.61088298e-03 2.47989446e-01 8.33522618e-01 1.29154980e+00
1.00415893e-01 -2.26370022e-01 1.05875514e-01 6.12086177e-01
-2.84299883e-03 -2.92882919e-01 -8.42391849e-01 2.44499043e-01
4.73748356e-01 1.17488539e+00 -8.43856871e-01 1.42401576e-01
-4.45948303e-01 8.30536246e-01 8.36136118e-02 3.54670584e-01
-9.45641518e-01 -2.66309798e-01 9.77698505e-01 -6.38330355e-02
4.06806856e-01 -3.81957054e-01 -3.57508898e-01 -1.16754603e+00
2.06073806e-01 -1.24465287e+00 2.20016509e-01 -9.14953828e-01
-1.33171105e+00 5.29214323e-01 1.22325495e-01 -1.19559562e+00
-8.18332881e-02 -7.94042110e-01 -6.18762851e-01 7.01328635e-01
-1.31448829e+00 -1.15452003e+00 -5.74398816e-01 5.49488544e-01
4.57290351e-01 -2.04919055e-01 6.95055306e-01 5.38489893e-02
-7.61416793e-01 8.22085798e-01 -2.09632918e-01 2.53420621e-01
1.18892908e+00 -1.15588236e+00 9.46215510e-01 1.11745703e+00
1.13269523e-01 4.27338600e-01 1.06176591e+00 -4.28154767e-01
-1.42654335e+00 -1.17080128e+00 3.13767910e-01 -1.11534929e+00
4.67489570e-01 -4.60149676e-01 -7.56870210e-01 3.45758140e-01
1.75955504e-01 3.14735770e-01 3.63592327e-01 1.82656884e-01
-9.42602634e-01 -9.45805609e-02 -1.19752371e+00 9.13102865e-01
1.17722619e+00 -2.55180389e-01 -3.74071807e-01 2.31385082e-01
8.32575977e-01 -5.72370887e-01 -5.50293922e-01 7.18774080e-01
6.49356246e-01 -1.36928344e+00 1.22046280e+00 -5.50797462e-01
4.12057698e-01 -6.36981249e-01 -4.70847040e-01 -1.52587593e+00
-3.26564312e-01 -1.72179535e-01 -6.17638975e-02 1.26788366e+00
5.34852624e-01 -4.71846849e-01 5.10170460e-01 5.48891068e-01
-7.87676871e-02 -4.51175302e-01 -5.98784506e-01 -1.04068792e+00
-4.72226404e-02 -5.98447680e-01 4.55651611e-01 7.08745182e-01
-6.54827893e-01 1.69854805e-01 -5.13734937e-01 3.44860673e-01
6.90479815e-01 -1.46920621e-01 1.19343650e+00 -8.61146390e-01
-2.60227919e-01 -5.34113288e-01 -5.77655673e-01 -6.00788414e-01
-1.51635528e-01 -4.38600928e-01 1.38625875e-01 -1.09322512e+00
1.66757286e-01 -3.89156282e-01 -2.28728890e-01 4.82353747e-01
-2.99514204e-01 8.68559599e-01 4.56317574e-01 1.21606536e-01
-6.29108727e-01 2.86766678e-01 1.08583260e+00 -2.19457552e-01
-5.93096949e-03 3.49565186e-02 -6.76618397e-01 6.05226338e-01
1.15614927e+00 -1.58468723e-01 -3.90611947e-01 -7.66975224e-01
7.94754252e-02 -4.74192947e-01 8.19865584e-01 -1.31938827e+00
-1.42973542e-01 -2.58351475e-01 7.06159234e-01 -3.53120983e-01
3.79980266e-01 -6.41204655e-01 -1.66320905e-01 4.85666573e-01
-2.41165698e-01 2.15267599e-01 5.79141021e-01 6.32088006e-01
1.84702694e-01 1.21647656e-01 1.00050676e+00 2.82855034e-02
-8.85158777e-01 2.72678196e-01 -1.59732834e-01 3.84270340e-01
8.63019824e-01 -9.95506793e-02 -6.53294921e-01 -3.40725005e-01
-1.42273694e-01 1.07442498e-01 9.50532079e-01 8.31508756e-01
4.50839937e-01 -1.26673198e+00 -9.95502770e-01 1.15932852e-01
5.21155715e-01 -1.76746517e-01 2.51662344e-01 6.26527309e-01
-5.93306839e-01 1.69411331e-01 -5.22386849e-01 -7.89880693e-01
-1.29061043e+00 7.46816754e-01 3.62951458e-01 1.79462448e-01
-3.09234530e-01 9.36128318e-01 1.32074460e-01 -6.47240579e-01
5.88619173e-01 -1.14463091e-01 8.92709941e-02 -1.00758895e-01
6.47559345e-01 4.79992360e-01 3.38754505e-01 -6.48189723e-01
-6.20779812e-01 7.19215572e-01 2.69529782e-02 -6.79304264e-03
1.13873696e+00 1.74861610e-01 2.22362891e-01 3.73718999e-02
9.24654365e-01 5.08209839e-02 -1.55891180e+00 -1.00527026e-01
-3.64897639e-01 -6.93795085e-01 -1.21247284e-02 -1.13807273e+00
-1.19918478e+00 8.45026612e-01 9.21847999e-01 2.13051513e-02
1.13696539e+00 5.65696657e-02 3.27093601e-01 4.48632956e-01
2.56899029e-01 -1.08208656e+00 3.02918345e-01 5.84846020e-01
1.27086413e+00 -1.62478888e+00 8.47890228e-02 -2.74576694e-01
-6.19573891e-01 8.58169913e-01 1.00602245e+00 -3.13824974e-02
3.55150461e-01 3.83673072e-01 4.58707839e-01 1.34972543e-01
-6.92666411e-01 -2.69654483e-01 3.27194542e-01 1.06094897e+00
3.25427860e-01 -6.59359843e-02 1.19722120e-01 9.66507643e-02
-3.28883439e-01 -3.79616499e-01 4.46365058e-01 6.06391311e-01
-2.28339344e-01 -8.57123137e-01 -6.82943344e-01 2.44097039e-01
-3.78282219e-01 1.25876367e-01 -8.52853000e-01 8.41486037e-01
2.52405852e-01 1.05242193e+00 -2.72057235e-01 -7.76434004e-01
5.59454620e-01 -3.63338113e-01 7.16275036e-01 -3.08469146e-01
-3.90999675e-01 -3.23341101e-01 1.67355269e-01 -7.93502212e-01
-8.36942419e-02 -7.43312478e-01 -5.64794064e-01 -5.43868363e-01
-1.89846188e-01 -4.36490476e-01 7.37825394e-01 4.70742077e-01
3.54817927e-01 3.48732889e-01 4.42855030e-01 -1.33487594e+00
-7.40593553e-01 -9.98646379e-01 -5.83557077e-02 5.55377126e-01
4.80878383e-01 -8.04812253e-01 -4.94325578e-01 -8.29623938e-02] | [8.085175514221191, -1.3930728435516357] |
9f65cf71-e3fe-4289-a531-32561952704b | reinforcement-learning-based-sequential-batch | 2112.10944 | null | https://arxiv.org/abs/2112.10944v2 | https://arxiv.org/pdf/2112.10944v2.pdf | Reinforcement Learning based Sequential Batch-sampling for Bayesian Optimal Experimental Design | Engineering problems that are modeled using sophisticated mathematical methods or are characterized by expensive-to-conduct tests or experiments, are encumbered with limited budget or finite computational resources. Moreover, practical scenarios in the industry, impose restrictions, based on logistics and preference, on the manner in which the experiments can be conducted. For example, material supply may enable only a handful of experiments in a single-shot or in the case of computational models one may face significant wait-time based on shared computational resources. In such scenarios, one usually resorts to performing experiments in a manner that allows for maximizing one's state-of-knowledge while satisfying the above mentioned practical constraints. Sequential design of experiments (SDOE) is a popular suite of methods, that has yielded promising results in recent years across different engineering and practical problems. A common strategy, that leverages Bayesian formalism is the Bayesian SDOE, which usually works best in the one-step-ahead or myopic scenario of selecting a single experiment at each step of a sequence of experiments. In this work, we aim to extend the SDOE strategy, to query the experiment or computer code at a batch of inputs. To this end, we leverage deep reinforcement learning (RL) based policy gradient methods, to propose batches of queries that are selected taking into account entire budget in hand. The algorithm retains the sequential nature, inherent in the SDOE, while incorporating elements of reward based on task from the domain of deep RL. A unique capability of the proposed methodology is its ability to be applied to multiple tasks, for example optimization of a function, once its trained. We demonstrate the performance of the proposed algorithm on a synthetic problem, and a challenging high-dimensional engineering problem. | ['Sayan Ghosh', 'Piyush Pandita', 'Yonatan Ashenafi'] | 2021-12-21 | null | null | null | null | ['policy-gradient-methods'] | ['methodology'] | [ 1.78610414e-01 -1.54201075e-01 -2.16590241e-01 -1.10034451e-01
-5.57110965e-01 -4.28295642e-01 3.31358820e-01 1.01106904e-01
-5.78400731e-01 1.02303183e+00 -4.51146275e-01 -4.59624141e-01
-6.22449338e-01 -7.54693449e-01 -8.38096142e-01 -7.29127288e-01
-1.18715443e-01 4.34925765e-01 -1.10649273e-01 5.62802479e-02
5.17750084e-01 6.83978438e-01 -1.66316915e+00 -2.34980509e-01
7.61311710e-01 1.21555281e+00 7.04689145e-01 5.90117335e-01
2.08323207e-02 5.83004177e-01 -3.55565548e-01 -7.87756965e-02
3.33554000e-01 -1.88639566e-01 -3.97233427e-01 2.21881568e-01
-3.06300581e-01 -6.01574838e-01 3.57591584e-02 7.14534342e-01
5.78626275e-01 3.73081535e-01 4.83780473e-01 -1.12065506e+00
-8.39584842e-02 3.38561147e-01 -4.19414937e-01 1.89039633e-02
3.15398991e-01 6.52675331e-01 1.05752432e+00 -6.62233233e-01
4.38309550e-01 8.87423396e-01 5.07020392e-02 2.29740575e-01
-1.34814131e+00 -3.39351654e-01 2.82588869e-01 -2.43012160e-02
-1.25754476e+00 -3.03820223e-01 8.17567587e-01 -5.11721373e-01
6.70587063e-01 -4.19086888e-02 5.16712248e-01 1.09471989e+00
3.71362776e-01 6.11879289e-01 1.22750676e+00 -3.04903835e-01
9.68380809e-01 3.54051322e-01 -3.45840335e-01 3.35932136e-01
2.37228483e-01 2.64897585e-01 -4.14515615e-01 -2.35440865e-01
6.28619015e-01 1.84547275e-01 -5.07293791e-02 -6.78082407e-01
-8.25219750e-01 8.96219313e-01 5.02174012e-02 2.46926457e-01
-8.46697807e-01 1.73343003e-01 2.35460579e-01 3.39764476e-01
2.01485574e-01 8.19892108e-01 -6.29112363e-01 -1.97445959e-01
-1.12771666e+00 5.65279722e-01 1.01882100e+00 7.26344585e-01
7.79751062e-01 -5.13945380e-03 -3.81035566e-01 4.44347322e-01
3.19221318e-01 1.81303695e-01 3.21230054e-01 -9.08144474e-01
4.58701462e-01 1.59890592e-01 6.59152031e-01 -8.35097075e-01
-2.56967783e-01 -6.47387862e-01 -6.06546640e-01 4.00568843e-01
4.04260248e-01 -2.97873944e-01 -5.78334689e-01 1.75174594e+00
4.91734087e-01 6.58187121e-02 -5.58191575e-02 9.72335219e-01
-1.32000417e-01 6.21170163e-01 9.98106226e-02 -5.50382733e-01
1.19542801e+00 -4.53663677e-01 -4.80655193e-01 -5.91929555e-02
3.89168084e-01 -7.78883159e-01 1.26625192e+00 8.09008479e-01
-1.00124741e+00 -4.35583025e-01 -1.00846827e+00 6.12062812e-01
-3.18796903e-01 3.69849831e-01 4.44456428e-01 7.51329958e-01
-7.72188127e-01 9.79118943e-01 -5.76064289e-01 -2.77689517e-01
3.15718293e-01 4.93197292e-01 2.02930287e-01 -2.26416618e-01
-1.01675868e+00 6.88512564e-01 3.08054417e-01 4.14195210e-01
-1.33524084e+00 -6.80485725e-01 -3.42049360e-01 4.03587490e-01
1.05697715e+00 -6.40429795e-01 1.18901920e+00 -7.37289906e-01
-1.96676624e+00 2.31881350e-01 3.26439887e-01 -5.14980495e-01
9.01698828e-01 -3.68676364e-01 -3.25541332e-04 9.73853767e-02
-1.61450833e-01 3.65068406e-01 9.41348374e-01 -1.22427046e+00
-5.08060098e-01 -2.69612938e-01 4.93429393e-01 1.26262397e-01
-3.06164652e-01 -1.33111611e-01 -2.21627187e-02 -2.61766404e-01
-3.67071718e-01 -1.00887811e+00 -5.90460181e-01 -1.73427373e-01
-4.16552991e-01 -4.63941284e-02 6.01964235e-01 -4.25435275e-01
1.28044403e+00 -1.86947381e+00 1.92625716e-01 3.52608442e-01
-2.72857577e-01 5.52081242e-02 1.05305254e-01 6.56156957e-01
3.20105344e-01 9.58514884e-02 -1.54640928e-01 -1.48146018e-01
2.44517744e-01 1.24179542e-01 -1.33677736e-01 4.31722671e-01
2.46493757e-01 3.13099653e-01 -7.79630005e-01 -4.69281554e-01
1.73858866e-01 4.37679626e-02 -7.19959259e-01 6.04822695e-01
-7.35235274e-01 7.36258984e-01 -1.01835656e+00 4.07953173e-01
5.16797304e-01 -3.36838067e-01 2.63255268e-01 -1.85552631e-02
-4.12423223e-01 -1.13563798e-01 -1.57844675e+00 1.70399809e+00
-1.03593946e+00 -5.63815385e-02 1.92269027e-01 -1.33158481e+00
7.58897603e-01 3.46064985e-01 7.53130376e-01 -3.74110729e-01
2.00198442e-01 4.90317911e-01 6.44058511e-02 -6.74845040e-01
2.23466039e-01 1.78761836e-02 6.86365291e-02 5.83927929e-01
-1.38844669e-01 -3.46296579e-01 4.72028762e-01 -1.70563683e-01
1.17851722e+00 4.46941555e-01 4.99682903e-01 -1.19810738e-01
4.76668000e-01 -2.22696140e-01 4.78681982e-01 9.48608756e-01
7.27362260e-02 1.21861063e-01 6.40584588e-01 -7.42395520e-02
-1.15433097e+00 -6.52118325e-01 7.84766823e-02 9.51268971e-01
-1.57557860e-01 3.05849999e-01 -5.06205916e-01 -5.99477649e-01
2.87408475e-02 8.45416844e-01 -3.06914687e-01 -7.33744577e-02
-4.26469773e-01 -4.56219226e-01 -7.63929114e-02 -5.62033951e-02
3.53997707e-01 -1.17551208e+00 -1.32410049e+00 5.16191006e-01
4.84255254e-01 -9.71000910e-01 -2.31980607e-01 4.86316860e-01
-8.35458100e-01 -8.23501348e-01 -8.40076566e-01 -2.63875306e-01
5.04417717e-01 -1.29488096e-01 9.12794113e-01 -1.78298905e-01
-4.66247648e-01 5.02864361e-01 -3.81741047e-01 -1.75151527e-01
-5.22312038e-02 1.62312254e-01 1.41325861e-01 2.57497817e-01
-9.17094760e-03 -7.77815938e-01 -7.77393520e-01 3.06516647e-01
-1.04986906e+00 -3.34676743e-01 1.08766651e+00 1.11375058e+00
4.94748920e-01 2.92793334e-01 8.96348417e-01 -7.49272227e-01
7.42963254e-01 -7.66992927e-01 -1.15489805e+00 2.17870593e-01
-8.66934061e-01 3.42310250e-01 1.05075717e+00 -6.03126705e-01
-1.07547891e+00 6.54848740e-02 1.45196527e-01 -5.84477425e-01
-2.44983181e-01 6.55786157e-01 -2.34947667e-01 -1.02319638e-03
3.82147044e-01 2.11702138e-01 -2.82928944e-02 -2.98612177e-01
2.12869346e-01 5.62715888e-01 -2.30831131e-01 -1.13622975e+00
4.98401314e-01 -5.20641096e-02 2.40574330e-01 -5.74402630e-01
-5.41611969e-01 -3.00998271e-01 -2.70683974e-01 -4.87903744e-01
6.65050149e-01 -4.56850708e-01 -1.16829669e+00 -5.60637005e-02
-9.37820137e-01 -2.78614879e-01 -2.03482240e-01 7.59900212e-01
-8.89941096e-01 9.99572128e-02 -8.16102922e-02 -1.31876206e+00
2.64181271e-02 -1.39932942e+00 8.06812644e-01 4.35360342e-01
4.51703705e-02 -6.40937448e-01 -1.18295498e-01 8.22635144e-02
4.93416876e-01 2.70925134e-01 8.28640223e-01 -7.27519989e-01
-9.40253377e-01 -1.29538760e-01 1.39705181e-01 4.53983933e-01
-3.99117321e-02 -8.75815973e-02 -6.99232757e-01 -4.52682704e-01
9.52943489e-02 -6.86524928e-01 3.01908195e-01 5.58372974e-01
1.47784686e+00 -2.84289122e-01 -8.29238445e-02 1.04644962e-01
1.64596474e+00 5.79214692e-01 2.74142891e-01 3.35083276e-01
1.40366182e-02 8.23838890e-01 1.05416024e+00 1.12443841e+00
-1.94862381e-01 8.35706174e-01 6.07934833e-01 2.22613424e-01
5.92261434e-01 -1.34069219e-01 1.99132740e-01 3.16161036e-01
-5.59837073e-02 -3.61849815e-01 -5.21928191e-01 4.34144378e-01
-1.86350715e+00 -7.29454875e-01 6.16121173e-01 2.65544152e+00
6.22609317e-01 3.09317917e-01 1.83900326e-01 -1.26121640e-02
6.84271932e-01 -1.85990408e-02 -9.24273670e-01 -4.78058130e-01
3.68182719e-01 9.01309103e-02 6.21742249e-01 1.20413788e-02
-6.65257156e-01 4.46338475e-01 4.97714043e+00 8.90771091e-01
-1.00151956e+00 -1.18964702e-01 8.54787052e-01 -2.26460323e-01
-1.97536632e-01 2.02725187e-01 -5.98872304e-01 5.63599408e-01
1.01860583e+00 -1.18971467e-01 8.16305637e-01 1.02815855e+00
8.23686659e-01 -4.93800193e-01 -1.37547469e+00 7.10998058e-01
-5.35511494e-01 -9.98283982e-01 -4.37209547e-01 2.97190189e-01
5.28987110e-01 -4.80174333e-01 3.59689534e-01 3.58391315e-01
2.80470848e-01 -8.82438660e-01 5.34774184e-01 5.14412582e-01
4.40277100e-01 -7.95867682e-01 7.12714791e-01 5.93550742e-01
-7.88227856e-01 -5.04911244e-01 -3.11401069e-01 -4.08139266e-03
1.98781878e-01 7.27673769e-01 -8.96975935e-01 6.90759003e-01
5.12193501e-01 5.25183454e-02 2.33048886e-01 1.17884219e+00
7.41315028e-03 5.74886203e-01 -3.55341345e-01 -6.53203249e-01
4.82150942e-01 -4.32952702e-01 4.74628597e-01 8.65228474e-01
7.76498735e-01 -4.15169373e-02 4.90809023e-01 1.08302140e+00
7.43270963e-02 3.34848344e-01 -6.78788424e-01 -2.06257179e-01
5.43219686e-01 1.24188066e+00 -6.63426280e-01 1.57352872e-02
-2.64610171e-01 5.41145802e-01 6.17260113e-02 5.09139895e-01
-9.13768053e-01 -3.31265688e-01 2.97558844e-01 8.95340964e-02
5.79685628e-01 -1.65764958e-01 -1.06241055e-01 -7.39154756e-01
2.24280193e-01 -1.00560033e+00 1.54402077e-01 -5.64550817e-01
-1.14399350e+00 1.83889836e-01 1.48635358e-01 -1.30221272e+00
-3.46152037e-01 -6.24796152e-01 -3.94902915e-01 8.25183153e-01
-1.55767870e+00 -6.45834625e-01 1.48084790e-01 3.87945354e-01
7.03491151e-01 -2.70967901e-01 4.21683282e-01 3.94241601e-01
-6.75550401e-01 2.43429467e-01 1.87139988e-01 -6.31402791e-01
5.05431890e-01 -1.08509493e+00 -4.75852668e-01 6.10044301e-01
-2.05420673e-01 5.01279593e-01 1.16702509e+00 -4.89477634e-01
-1.64931571e+00 -6.41363084e-01 2.69227356e-01 4.11907434e-01
8.12194765e-01 -2.39849254e-01 -5.47968030e-01 1.44235432e-01
-5.49942926e-02 -1.78586680e-03 2.28782207e-01 4.07556295e-02
3.28026652e-01 -3.69084418e-01 -1.33264029e+00 6.49316967e-01
6.31906450e-01 -1.45331740e-01 -4.45036441e-02 4.49047863e-01
5.89006126e-01 -1.99935511e-01 -9.15999293e-01 2.52629966e-01
4.98552650e-01 -8.72788846e-01 6.93911433e-01 -6.41276479e-01
4.45674568e-01 -2.28965387e-01 -1.36540189e-01 -1.36686313e+00
3.46413366e-02 -9.69302773e-01 -1.20592915e-01 1.15799809e+00
4.93523061e-01 -4.70297873e-01 9.01257098e-01 7.16101229e-01
5.78150116e-02 -1.16785419e+00 -9.42919135e-01 -8.25012028e-01
-3.23100477e-01 -2.94669807e-01 5.41601837e-01 3.25789720e-01
-3.55721205e-01 -4.70833369e-02 -6.50877416e-01 2.80830085e-01
5.41831493e-01 1.18305258e-01 8.16332817e-01 -9.19419348e-01
-9.95703459e-01 -2.12886095e-01 -1.58188846e-02 -1.05211985e+00
-6.86848117e-03 -1.21885166e-01 3.68653715e-01 -1.10419691e+00
-6.54159784e-02 -6.67682767e-01 -4.67783898e-01 -5.45431441e-03
1.45100668e-01 -6.17605686e-01 2.00882360e-01 3.79295349e-02
-4.54732418e-01 6.24996722e-01 1.20413196e+00 1.33162756e-02
-4.23494011e-01 5.33716917e-01 -3.76620233e-01 3.80029410e-01
7.06323028e-01 -5.09955943e-01 -6.91434860e-01 2.34501045e-02
4.17659074e-01 6.11423075e-01 8.80481079e-02 -9.08736289e-01
2.82237768e-01 -4.99730408e-01 4.98572402e-02 -2.55996704e-01
2.15453729e-01 -1.16764724e+00 1.89923882e-01 5.30352592e-01
-4.57088590e-01 8.41554552e-02 -1.70845479e-01 8.86689186e-01
-9.94253680e-02 -6.85923755e-01 7.04653323e-01 -2.78726965e-01
-4.93578374e-01 2.39023745e-01 -2.92056799e-01 -1.62309676e-01
1.30204749e+00 -1.05553463e-01 2.74664611e-01 -2.57072508e-01
-7.66441584e-01 3.08375120e-01 6.80119321e-02 7.28686340e-03
3.59954357e-01 -8.33344221e-01 -4.21820641e-01 -3.24446112e-01
-8.68788436e-02 -1.50321409e-01 2.80427814e-01 9.34243321e-01
-3.73520821e-01 4.06094730e-01 -1.46843836e-01 -4.11072910e-01
-7.32580066e-01 8.00179183e-01 1.63191393e-01 -8.08555245e-01
-1.75119966e-01 3.53713542e-01 1.41283059e-02 -3.25381570e-02
1.71228856e-01 -3.51331890e-01 -1.50062114e-01 3.99934277e-02
7.32176304e-02 4.14790273e-01 1.47608474e-01 2.45594293e-01
-2.54540462e-02 4.11433816e-01 -3.52555811e-02 -3.79573405e-01
1.45563471e+00 -9.37740281e-02 2.41728798e-01 5.04755199e-01
8.39728355e-01 -1.01482734e-01 -1.53479743e+00 1.38448142e-02
4.03680019e-02 -5.13795018e-01 2.35635012e-01 -7.69306123e-01
-7.98836470e-01 6.41767740e-01 6.51821613e-01 3.97855818e-01
1.20836854e+00 -3.75241011e-01 4.55026031e-01 6.53015375e-01
7.05988467e-01 -1.38574398e+00 2.52828360e-01 8.05330276e-03
6.14948332e-01 -1.14873278e+00 6.33868426e-02 4.30350229e-02
-5.43596327e-01 1.18768549e+00 4.54379946e-01 -2.35948145e-01
6.07228637e-01 1.01802103e-01 -5.25698245e-01 -1.93902235e-02
-8.19189429e-01 -1.28659114e-01 -1.32120654e-01 5.99646568e-02
2.22061068e-01 -2.84062251e-02 -7.17202604e-01 3.44770253e-01
3.68155599e-01 3.29116940e-01 3.98914129e-01 1.17878771e+00
-4.64485705e-01 -1.11230099e+00 -3.78127784e-01 4.95270252e-01
-5.46576500e-01 1.48460999e-01 1.74710259e-01 8.29441369e-01
1.23939678e-01 9.90572870e-01 -3.10990930e-01 4.43891808e-02
4.04000133e-01 -1.75332010e-01 2.93805420e-01 -5.20150244e-01
-5.36090195e-01 1.02640718e-01 5.75628430e-02 -5.98359227e-01
-3.47931564e-01 -8.10952485e-01 -8.87476325e-01 2.39870865e-02
-3.72292191e-01 2.31719568e-01 9.70787823e-01 1.05488169e+00
3.60284746e-01 6.75956070e-01 1.02110815e+00 -9.17514265e-01
-1.17158067e+00 -7.89635360e-01 -8.76416445e-01 1.16749942e-01
9.74020734e-02 -1.03737915e+00 -5.22767723e-01 -3.17735910e-01] | [4.424655437469482, 2.4504895210266113] |
64c2fe83-b110-40ce-b8de-0dc87c744c65 | spectral-reconstruction-with-deep-neural | 1905.04305 | null | https://arxiv.org/abs/1905.04305v2 | https://arxiv.org/pdf/1905.04305v2.pdf | Spectral Reconstruction with Deep Neural Networks | We explore artificial neural networks as a tool for the reconstruction of spectral functions from imaginary time Green's functions, a classic ill-conditioned inverse problem. Our ansatz is based on a supervised learning framework in which prior knowledge is encoded in the training data and the inverse transformation manifold is explicitly parametrised through a neural network. We systematically investigate this novel reconstruction approach, providing a detailed analysis of its performance on physically motivated mock data, and compare it to established methods of Bayesian inference. The reconstruction accuracy is found to be at least comparable, and potentially superior in particular at larger noise levels. We argue that the use of labelled training data in a supervised setting and the freedom in defining an optimisation objective are inherent advantages of the present approach and may lead to significant improvements over state-of-the-art methods in the future. Potential directions for further research are discussed in detail. | ['Felix P. G. Ziegler', 'Nicolas Wink', 'Manuel Scherzer', 'Alexander Rothkopf', 'Sebastian J. Wetzel', 'Julian M. Urban', 'Lukas Kades', 'Jan M. Pawlowski'] | 2019-05-10 | null | null | null | null | ['spectral-reconstruction'] | ['computer-vision'] | [ 6.01730704e-01 3.07422012e-01 2.27176994e-01 -3.44585925e-01
-6.87095404e-01 -2.88099915e-01 8.11842024e-01 -3.04532409e-01
-5.91206551e-01 1.00659466e+00 -4.22066487e-02 -4.10237074e-01
-7.80775845e-01 -6.72104836e-01 -5.01296461e-01 -1.26850379e+00
-8.73090476e-02 7.90524721e-01 1.75038241e-02 -1.90062791e-01
3.61246705e-01 7.52616286e-01 -1.43362629e+00 -2.97280878e-01
7.06308484e-01 1.14401937e+00 2.65920330e-02 3.93227071e-01
3.17577153e-01 6.89197183e-01 -1.91170052e-01 -1.82587638e-01
1.96082130e-01 -5.71241558e-01 -9.28269863e-01 1.92673095e-02
-2.36278921e-01 -2.55810339e-02 -2.94037998e-01 9.94111001e-01
5.50969899e-01 4.83548492e-01 1.11833227e+00 -4.67968553e-01
-3.71892303e-01 2.08868176e-01 -6.31407201e-02 8.54213983e-02
-4.52930760e-03 9.27761346e-02 7.62012720e-01 -6.90526426e-01
4.32374448e-01 8.42657924e-01 8.91036630e-01 3.56256276e-01
-1.79360962e+00 -7.45848343e-02 -5.72493792e-01 2.27094233e-01
-1.33287990e+00 -7.63358474e-01 9.61788476e-01 -4.72657263e-01
8.29456389e-01 2.97469288e-01 4.98311549e-01 8.68084013e-01
4.76445593e-02 1.03254527e-01 1.44899535e+00 -1.13570189e+00
4.93984997e-01 2.85453379e-01 -1.58307608e-02 6.39997661e-01
1.31118253e-01 7.22625375e-01 -2.22378463e-01 -1.60955071e-01
9.63612199e-01 -6.71222687e-01 -3.58376980e-01 -7.14990735e-01
-1.23363745e+00 1.04923618e+00 2.35822782e-01 2.60679632e-01
-6.07123017e-01 2.95951962e-01 2.88094759e-01 5.02607636e-02
7.68579781e-01 5.22001863e-01 -4.48857963e-01 -4.53662612e-02
-1.01213825e+00 2.87946641e-01 1.00778198e+00 1.41631842e-01
6.22511744e-01 2.64466405e-01 1.89927623e-01 8.59258652e-01
6.64739251e-01 5.28715491e-01 6.06293650e-03 -1.62812769e+00
-1.14602715e-01 -1.83397189e-01 2.49326721e-01 -6.76824689e-01
-4.22846735e-01 -5.25119543e-01 -7.18882322e-01 6.11519754e-01
6.34166002e-01 -2.58448094e-01 -5.37632465e-01 1.62609231e+00
3.20979953e-01 1.93961740e-01 1.33143216e-01 8.56581509e-01
5.71066260e-01 3.86426419e-01 -3.66126120e-01 -4.41351444e-01
9.29811001e-01 -3.54020357e-01 -6.57366931e-01 6.17687851e-02
2.88558125e-01 -7.17240691e-01 5.35578072e-01 5.43531775e-01
-1.07096636e+00 -3.51606905e-01 -1.13749707e+00 2.03653157e-01
-3.85059901e-02 1.60944432e-01 9.27796841e-01 8.25938761e-01
-9.33942139e-01 1.35132182e+00 -9.01249528e-01 -2.30047241e-01
6.16373420e-02 4.37739044e-01 -8.88108835e-02 2.90631115e-01
-1.01850510e+00 1.10890210e+00 5.14124811e-01 4.66325611e-01
-4.48999166e-01 -3.79933327e-01 -8.32160354e-01 -1.93968579e-01
1.32810906e-01 -6.33260429e-01 1.32930851e+00 -7.27349758e-01
-2.07696509e+00 8.95880878e-01 1.35404736e-01 -3.58860105e-01
4.07826483e-01 2.81367272e-01 -4.59199935e-01 4.59995449e-01
-2.06748679e-01 2.54563242e-01 9.41324353e-01 -1.37089622e+00
7.40574598e-02 -1.28088281e-01 -1.16840743e-01 -2.52294093e-01
2.70504177e-01 -7.43834749e-02 7.06151947e-02 -5.09628356e-01
4.85636055e-01 -9.26297843e-01 -3.53865653e-01 -1.55123800e-01
-1.51010901e-01 4.97398041e-02 3.52412403e-01 -5.98477066e-01
8.01726878e-01 -1.71227324e+00 3.25661659e-01 5.82826138e-01
3.53797972e-02 1.29165739e-01 1.45708412e-01 6.35926664e-01
-4.46945548e-01 -3.24642360e-01 -6.26385868e-01 -1.84556395e-01
9.50135887e-02 3.62436146e-01 -2.18387648e-01 9.22359705e-01
2.10113525e-01 8.06007326e-01 -7.37018824e-01 -1.66377708e-01
5.30513167e-01 6.64969146e-01 -4.20894891e-01 4.73435596e-02
-7.99234137e-02 9.73979473e-01 -3.47576857e-01 1.33575544e-01
6.22937083e-01 -2.14767233e-01 1.81984365e-01 -3.01946014e-01
-2.67811537e-01 4.22630608e-01 -1.26570070e+00 1.47878718e+00
-4.70010310e-01 6.01815999e-01 2.30361849e-01 -1.78584135e+00
7.90684402e-01 5.64691842e-01 4.98865753e-01 -6.08529389e-01
2.18444809e-01 3.53720844e-01 1.68754444e-01 -7.56313264e-01
3.51299644e-02 -8.70769441e-01 3.10926646e-01 4.76525843e-01
2.62398899e-01 -2.99752146e-01 -1.28079638e-01 -2.28395984e-01
8.86386752e-01 7.47291625e-01 3.74295592e-01 -5.81836283e-01
7.29878783e-01 -1.43582001e-01 1.62960693e-01 9.53625977e-01
2.33050082e-02 6.12114847e-01 3.46096843e-01 -4.14121002e-01
-1.01902306e+00 -9.35411632e-01 -8.21485162e-01 5.40957212e-01
-2.53131777e-01 7.40575790e-03 -7.75323868e-01 -2.31025100e-01
-1.69652998e-01 8.96321356e-01 -6.22050822e-01 -3.14357840e-02
-5.69280326e-01 -1.28380764e+00 2.90380150e-01 2.10445464e-01
4.59173620e-01 -1.28920901e+00 -5.42912424e-01 3.31995785e-01
-9.48038176e-02 -9.96441543e-01 5.28832793e-01 6.06378734e-01
-1.17442441e+00 -9.05133069e-01 -6.67329550e-01 -3.28931957e-01
3.32060993e-01 -2.32581034e-01 1.15282357e+00 -2.45297235e-02
-1.36879399e-01 6.05842471e-01 -1.17579430e-01 -1.59749180e-01
-8.04230452e-01 -3.03581119e-01 8.96337628e-02 3.70994285e-02
1.42961621e-01 -1.10996258e+00 -5.70227027e-01 1.14781968e-01
-7.77193248e-01 -2.69570917e-01 2.80220747e-01 1.06150937e+00
2.77096391e-01 1.84164524e-01 5.68428099e-01 -8.07359517e-01
5.12860358e-01 -2.91450799e-01 -8.31299007e-01 -1.53176293e-01
-8.19030523e-01 4.50896442e-01 4.67359155e-01 -9.49653834e-02
-1.38506949e+00 6.56855106e-03 -6.24385774e-01 3.51173356e-02
-4.61807787e-01 5.80645323e-01 1.41890734e-01 -7.96940327e-01
9.15852070e-01 1.02144308e-01 2.60011643e-01 -6.07639551e-01
2.36673877e-01 5.76233625e-01 7.39105761e-01 -1.00384665e+00
8.08586121e-01 5.89161515e-01 5.71139336e-01 -1.08593297e+00
-6.38743103e-01 -4.43097264e-01 -8.32701325e-01 -2.34835088e-01
5.44292629e-01 -4.23234999e-01 -8.36641312e-01 3.30937654e-01
-1.15243685e+00 -3.42393547e-01 -3.61054569e-01 7.80455470e-01
-1.06106472e+00 6.42344654e-01 -5.31301022e-01 -1.16208971e+00
-4.09890749e-02 -1.09062350e+00 9.65500474e-01 -5.83986156e-02
-6.50018156e-02 -1.52686775e+00 2.43288398e-01 4.27802175e-01
4.37505335e-01 3.58988106e-01 8.89465511e-01 -4.61541146e-01
-5.95881045e-01 -2.05393523e-01 -1.66938305e-01 6.38606906e-01
-1.68529660e-01 -1.28139302e-01 -1.36121440e+00 -1.39123306e-01
7.26592958e-01 -2.99148291e-01 9.35335636e-01 5.52490830e-01
9.91966188e-01 -1.99496467e-02 -1.69915363e-01 5.41198134e-01
1.55054677e+00 -2.99468152e-02 8.41755509e-01 2.18238965e-01
3.06020230e-01 7.65931427e-01 2.89396673e-01 3.31778944e-01
-2.41274312e-01 7.08723426e-01 3.75398040e-01 -3.03287804e-02
3.70837338e-02 3.72812718e-01 -1.88925013e-01 9.04260755e-01
-6.01667762e-01 1.55526802e-01 -6.60396338e-01 2.92644709e-01
-1.74550653e+00 -1.04436111e+00 -3.61000746e-01 2.42600131e+00
6.81773543e-01 9.43584964e-02 -9.69664901e-02 6.27020717e-01
3.69811475e-01 -5.57084680e-02 -1.48379773e-01 -2.05208585e-01
1.02855705e-01 8.06502759e-01 3.95266116e-01 6.87022150e-01
-1.00293851e+00 4.26396519e-01 7.57332563e+00 8.82817745e-01
-8.92314494e-01 2.57853150e-01 3.77433121e-01 3.96316886e-01
-1.43263832e-01 2.49762088e-01 -1.61691099e-01 3.31496179e-01
1.10872936e+00 5.12909353e-01 8.32449436e-01 2.43385091e-01
3.05355132e-01 -5.02313375e-01 -9.17752504e-01 8.04225147e-01
-7.26352483e-02 -1.41531706e+00 -6.46853864e-01 1.78659379e-01
5.14592469e-01 1.77742526e-01 -1.74538806e-01 1.02521889e-01
-6.30123615e-02 -1.04633677e+00 5.10097980e-01 7.31264353e-01
6.76699936e-01 -6.88641846e-01 6.57348871e-01 5.02041459e-01
-5.02458513e-01 3.95931035e-01 -2.45481431e-01 -3.73062044e-01
2.44102478e-01 6.93984985e-01 -4.54688847e-01 7.08273053e-01
5.01699269e-01 5.11394918e-01 -1.28097758e-01 1.08934402e+00
-2.13670358e-01 9.86534297e-01 -5.59692323e-01 -7.20548071e-03
1.71246246e-01 -7.99339116e-01 6.93314850e-01 1.15795588e+00
1.30236939e-01 1.17822953e-01 -2.03389525e-01 1.17392325e+00
4.91967559e-01 -2.13447183e-01 -5.75829864e-01 -6.00049905e-02
-1.45251587e-01 1.35396111e+00 -1.04163587e+00 -9.90797430e-02
-2.22939506e-01 5.94566405e-01 2.27260396e-01 6.86365247e-01
-6.33277297e-01 -3.97835553e-01 2.37805635e-01 -2.09842641e-02
4.75384980e-01 -1.89380363e-01 -2.91214347e-01 -1.01881063e+00
4.54323404e-02 -5.45435846e-01 5.44718504e-02 -8.28382671e-01
-1.26453090e+00 2.77510405e-01 3.33284348e-01 -9.40717876e-01
-6.32314205e-01 -1.12043273e+00 -5.22409618e-01 9.96595860e-01
-1.38570881e+00 -7.29510009e-01 7.08044134e-03 3.52375656e-02
-1.78980798e-01 -7.74563774e-02 1.23806071e+00 4.17745896e-02
-1.96067542e-01 -6.87109381e-02 7.94006586e-01 -1.30782485e-01
2.12899402e-01 -1.28394771e+00 5.45723364e-02 5.09252727e-01
2.79150516e-01 6.50408328e-01 1.24469507e+00 -2.96518505e-01
-1.17565095e+00 -2.98886806e-01 4.05716896e-01 -4.16987717e-01
7.85276413e-01 -2.34528139e-01 -1.02214527e+00 4.71899837e-01
2.32324183e-01 2.36061230e-01 7.19649911e-01 2.09762007e-01
1.04167745e-01 1.62670389e-01 -1.21840012e+00 1.92536458e-01
7.42105246e-01 -6.64132714e-01 -7.12691069e-01 4.26272511e-01
2.28235945e-01 -2.58724332e-01 -1.05585182e+00 6.38713896e-01
5.62109709e-01 -1.38805819e+00 1.32773936e+00 -4.33037996e-01
2.96866059e-01 -1.29544854e-01 -1.40769154e-01 -1.26541662e+00
-3.13760877e-01 -7.81266153e-01 -1.16111360e-01 8.86714697e-01
1.07208259e-01 -1.00199831e+00 6.86772823e-01 3.36586386e-01
2.69270819e-02 -7.49087811e-01 -1.25355053e+00 -7.73893356e-01
2.08108172e-01 -7.39094377e-01 4.13718112e-02 6.80139363e-01
-1.16180286e-01 4.53328133e-01 -3.01913559e-01 1.72355786e-01
9.49058473e-01 -6.95950538e-02 2.82074243e-01 -1.47154570e+00
-7.51970291e-01 -3.29344213e-01 -2.64906585e-01 -8.37374210e-01
4.46164191e-01 -8.83633494e-01 1.61518127e-01 -1.09382308e+00
-1.82421654e-01 -4.82567102e-01 -1.96607262e-01 -1.16938822e-01
3.57379824e-01 5.27783692e-01 -3.91092896e-01 1.93493888e-02
-1.10652067e-01 8.13760698e-01 1.10879230e+00 2.22294509e-01
-4.17905208e-03 2.65502214e-01 -3.35178286e-01 1.02178109e+00
7.67211378e-01 -5.56654513e-01 -1.94859385e-01 1.03112735e-01
1.64869621e-01 2.36337677e-01 7.80154586e-01 -1.15982270e+00
-2.38623489e-02 -4.76487773e-03 5.08488357e-01 -2.79278785e-01
6.71114862e-01 -8.54168534e-01 2.80566692e-01 2.29989856e-01
-1.28060177e-01 -5.84399700e-01 2.08314836e-01 5.89402020e-01
9.07126442e-02 -1.13323092e+00 1.02497327e+00 -3.06395143e-01
-2.86126137e-01 -3.73879433e-01 -4.82455641e-01 -8.05709213e-02
3.89821291e-01 -2.20521599e-01 3.17469329e-01 -3.26452553e-01
-8.33911240e-01 -5.96391857e-01 3.42340350e-01 -4.26518232e-01
2.46695206e-01 -1.18308675e+00 -4.96261477e-01 2.09909424e-01
-4.59613763e-02 -3.39119017e-01 -3.08131836e-02 1.07543278e+00
-7.25103378e-01 4.35878545e-01 7.97987059e-02 -7.15467513e-01
-7.60521352e-01 4.60843414e-01 7.79983759e-01 -1.21808238e-01
-6.67790890e-01 4.00673538e-01 -1.14988722e-01 -7.93796420e-01
2.14859545e-02 -9.38892290e-02 -6.19659200e-02 -3.97014916e-01
2.10792035e-01 5.17823577e-01 2.35016972e-01 -7.28993714e-01
-5.31207100e-02 6.27952039e-01 5.45885801e-01 -5.45699477e-01
1.52498865e+00 -6.96342960e-02 -5.09027779e-01 5.02903044e-01
1.03257871e+00 -1.02747567e-01 -1.11855435e+00 -3.17381263e-01
7.79018253e-02 -9.66514722e-02 3.87513161e-01 -8.64468455e-01
-8.44319582e-01 7.88162231e-01 5.42039692e-01 2.98678815e-01
1.04922748e+00 3.64049189e-02 1.87553927e-01 7.61004031e-01
4.10254389e-01 -1.14968622e+00 -9.47934613e-02 4.19932932e-01
8.83788526e-01 -1.34746552e+00 2.57922471e-01 -5.27988851e-01
4.42951731e-02 1.37051320e+00 -1.89074397e-01 -3.73388529e-01
7.26785421e-01 1.30658612e-01 -1.34186307e-02 -4.00220990e-01
-1.87010005e-01 -2.64901519e-01 4.64158952e-01 7.92906463e-01
4.27481949e-01 -1.42608538e-01 -3.37816119e-01 2.30969697e-01
-2.75688060e-02 1.08085610e-01 3.40698928e-01 7.55511224e-01
-2.64026254e-01 -1.29313099e+00 -6.26266420e-01 3.25114489e-01
-4.65850711e-01 -2.38146670e-02 9.96714979e-02 8.71793211e-01
9.15412009e-02 8.97997379e-01 -2.89049685e-01 1.24774188e-01
7.08244890e-02 3.54811311e-01 9.89934087e-01 -3.83158028e-01
-2.23173693e-01 2.90855616e-01 3.13558400e-01 -2.79860705e-01
-8.76072824e-01 -7.71418273e-01 -1.16427779e+00 -1.00965332e-02
-4.56943393e-01 3.14577699e-01 8.57373655e-01 1.32167351e+00
-7.83114731e-02 6.22037947e-01 3.01516294e-01 -1.42509830e+00
-8.31549108e-01 -1.07886028e+00 -7.13634908e-01 1.49272874e-01
3.89313281e-01 -1.06297147e+00 -6.54862046e-01 -2.51933672e-02] | [12.145406723022461, -2.478773593902588] |
14693a14-6d7f-4db1-9997-65aead2ad932 | a-hybrid-data-driven-physics-constrained | 2205.06494 | null | https://arxiv.org/abs/2205.06494v2 | https://arxiv.org/pdf/2205.06494v2.pdf | A hybrid data driven-physics constrained Gaussian process regression framework with deep kernel for uncertainty quantification | Gaussian process regression (GPR) has been a well-known machine learning method for various applications such as uncertainty quantifications (UQ). However, GPR is inherently a data-driven method, which requires sufficiently large dataset. If appropriate physics constraints (e.g. expressed in partial differential equations) can be incorporated, the amount of data can be greatly reduced and the accuracy further improved. In this work, we propose a hybrid data driven-physics constrained Gaussian process regression framework. We encode the physics knowledge with Boltzmann-Gibbs distribution and derive our model through maximum likelihood (ML) approach. We apply deep kernel learning method. The proposed model learns from both data and physics constraints through the training of a deep neural network, which serves as part of the covariance function in GPR. The proposed model achieves good results in high-dimensional problem, and correctly propagate the uncertainty, with very limited labelled data provided. | ['Tieyong Zeng', 'Cheng Chang'] | 2022-05-13 | null | null | null | null | ['gpr', 'gpr'] | ['computer-vision', 'miscellaneous'] | [ 2.32028961e-02 1.16892932e-02 2.47713774e-01 -4.98390824e-01
-9.73208845e-01 -1.60552889e-01 6.59974098e-01 1.82422489e-01
-5.45792520e-01 8.88993382e-01 -6.97317719e-02 -9.77174342e-02
-3.14008623e-01 -1.14655864e+00 -7.78358817e-01 -1.08216560e+00
2.90221781e-01 8.47173750e-01 3.42434436e-01 3.50781322e-01
3.36190909e-01 4.21744406e-01 -1.08104467e+00 -6.70405924e-01
1.03773594e+00 9.95063722e-01 1.90116048e-01 5.20324051e-01
-4.51395094e-01 6.90871775e-01 -1.45238191e-01 -1.58876404e-01
-2.15211213e-01 -7.83164278e-02 -4.18325275e-01 -3.59816670e-01
-3.86506259e-01 -1.31511986e-01 -4.46294934e-01 1.11470246e+00
4.99877393e-01 6.14689171e-01 1.25263321e+00 -1.00454497e+00
-6.01430178e-01 3.26422989e-01 -6.88704073e-01 -9.52872187e-02
-1.13519177e-01 -7.18980432e-02 5.67113221e-01 -9.75757539e-01
1.36828020e-01 1.38903058e+00 5.55988967e-01 4.84662563e-01
-1.32348275e+00 -5.73638558e-01 -3.30124125e-02 -1.05588913e-01
-1.53234291e+00 -1.06929004e-01 9.56426084e-01 -6.52516901e-01
5.31860530e-01 -1.40325174e-01 3.31004709e-01 1.00716066e+00
5.09044170e-01 6.80344760e-01 1.00220883e+00 -1.61409855e-01
8.15695941e-01 1.50097147e-01 2.45689392e-01 4.85801578e-01
2.80257434e-01 1.20554917e-01 -4.13249016e-01 -4.34872419e-01
9.82682645e-01 7.54424855e-02 -1.56772271e-01 -4.64659393e-01
-7.94416487e-01 9.47354734e-01 2.47595951e-01 -1.90894768e-01
-3.06760043e-01 4.41566378e-01 3.34544070e-02 -3.88887763e-01
7.31380582e-01 6.71247840e-02 -4.65636194e-01 -1.56851709e-01
-8.95710886e-01 4.68863547e-01 7.59019852e-01 8.37840796e-01
8.22927952e-01 -5.90875233e-03 -1.49188593e-01 8.28082383e-01
1.15327299e+00 8.37449789e-01 1.94519341e-01 -7.33079791e-01
1.76117361e-01 1.99469551e-01 3.49665672e-01 -6.95079029e-01
-3.67818773e-01 -2.00811952e-01 -1.02029717e+00 2.35955685e-01
2.14799896e-01 -4.50834453e-01 -1.00120628e+00 1.69680178e+00
5.24316251e-01 3.34617287e-01 4.42045927e-02 8.65146875e-01
6.37342274e-01 8.96770358e-01 3.27350974e-01 -1.04705289e-01
1.24208736e+00 -4.77459908e-01 -7.57137418e-01 -4.34354991e-02
2.09841922e-01 -4.77330774e-01 6.86731458e-01 5.28020442e-01
-8.39528739e-01 -4.79083896e-01 -9.01279807e-01 -1.23469427e-01
-1.89331636e-01 -1.34866521e-01 6.24380231e-01 6.25805020e-01
-6.28968954e-01 7.60652542e-01 -1.15805352e+00 -7.43854940e-02
4.33459342e-01 3.99078995e-01 -5.55540854e-03 5.60030155e-02
-1.11360455e+00 7.33781159e-01 5.29174805e-01 3.81010860e-01
-9.41199064e-01 -5.70157647e-01 -8.12995851e-01 3.45115364e-02
3.18701208e-01 -7.03605294e-01 1.17649376e+00 1.14727914e-01
-1.94100463e+00 2.52923995e-01 -9.91393253e-02 -2.46525213e-01
5.94892263e-01 -5.28295100e-01 -2.58597881e-01 -7.36878365e-02
-2.33675659e-01 3.49226803e-01 1.06088173e+00 -1.20656729e+00
-3.45696986e-01 -4.58645821e-01 -3.47965658e-01 1.20024290e-02
1.33546546e-01 -1.77848581e-02 -3.63289803e-01 -3.12415779e-01
5.16210794e-01 -9.32805836e-01 -3.91900778e-01 -1.55087665e-01
-3.85099888e-01 -2.89354593e-01 5.81790447e-01 -7.95799315e-01
8.79164219e-01 -1.96621704e+00 3.29624951e-01 2.95340359e-01
1.05621338e-01 -4.44315299e-02 5.72243392e-01 2.68247932e-01
3.66575420e-01 4.50814739e-02 -8.71540546e-01 -5.58589578e-01
2.91429877e-01 3.43722403e-01 -2.44945526e-01 4.80949312e-01
4.73351777e-01 6.29734635e-01 -7.96136737e-01 -5.59656918e-01
4.02275592e-01 6.55438960e-01 -3.24109077e-01 3.59754920e-01
-4.79882240e-01 9.95492935e-01 -8.42810154e-01 2.95262754e-01
1.23810089e+00 -1.08208984e-01 -3.16082716e-01 -8.36833417e-02
-6.03771489e-03 -1.64803624e-01 -1.40456903e+00 1.85298741e+00
-5.16839981e-01 1.34058729e-01 -1.71266086e-02 -8.13669920e-01
1.19890571e+00 2.00656429e-01 2.94017106e-01 -1.04661867e-01
2.31734887e-01 6.22678809e-02 -2.71885037e-01 -4.77651924e-01
3.75388592e-01 -5.34054518e-01 -9.35856551e-02 1.51313528e-01
1.05841450e-01 -6.28360868e-01 -2.89327353e-01 1.28120091e-02
8.49180639e-01 6.15695894e-01 3.62819768e-02 -2.68298060e-01
6.57207489e-01 -2.24388495e-01 5.03800511e-01 6.08910441e-01
5.99520281e-03 6.14496946e-01 3.31662059e-01 5.97279295e-02
-9.54740405e-01 -1.28092754e+00 -5.33638895e-01 5.33829391e-01
1.63400546e-01 3.77650633e-02 -5.75784683e-01 -4.11855578e-01
1.47591606e-01 9.55077767e-01 -3.37561429e-01 -2.88889259e-01
-3.14075857e-01 -1.23422992e+00 3.50238651e-01 4.91024762e-01
5.88600218e-01 -7.35265732e-01 3.54675576e-02 3.96482438e-01
2.83636868e-01 -1.01846862e+00 8.97652432e-02 2.50490516e-01
-9.96788621e-01 -8.16317558e-01 -6.73801541e-01 -1.34004116e-01
4.95527327e-01 -3.81558478e-01 5.27816236e-01 -7.36560166e-01
-1.61172435e-01 2.98694998e-01 -5.21728545e-02 -5.13903081e-01
-2.01572880e-01 -2.42361471e-01 1.57929122e-01 8.03765655e-03
5.75114131e-01 -7.02913582e-01 -4.93613601e-01 -5.85743301e-02
-8.10859442e-01 -2.05459341e-01 5.96876800e-01 6.36887908e-01
8.03344488e-01 4.37912434e-01 4.44189221e-01 -7.46269286e-01
7.23735094e-01 -5.17883658e-01 -8.55935395e-01 -1.05518028e-01
-5.25260746e-01 4.88551974e-01 3.86542678e-01 -2.04098791e-01
-1.67468667e+00 -5.64925000e-02 -3.22400063e-01 -4.05516982e-01
-2.76335478e-01 5.48743546e-01 -4.42182869e-01 -7.69406930e-02
4.36500520e-01 1.35274991e-01 -3.42575043e-01 -7.83343911e-01
3.86854410e-01 6.94124341e-01 3.39512050e-01 -1.13725162e+00
6.77236617e-01 4.35267031e-01 4.47665870e-01 -6.55142605e-01
-7.14948118e-01 -3.34130913e-01 -6.35574698e-01 -4.79037426e-02
1.03940821e+00 -8.83865356e-01 -7.76093900e-01 5.18769145e-01
-1.10620153e+00 -6.78043664e-02 -6.66195676e-02 1.01712382e+00
-5.90739071e-01 4.76476550e-01 -5.14614582e-01 -1.50222778e+00
-3.54110926e-01 -1.00616777e+00 1.11018670e+00 4.52275455e-01
1.74778506e-01 -1.10065758e+00 2.47975752e-01 7.24361315e-02
1.86114386e-01 4.04225498e-01 7.67674863e-01 -5.70924342e-01
-6.92063272e-01 -2.83733875e-01 -4.80746299e-01 4.99323249e-01
-2.01919563e-02 6.18927032e-02 -1.21391022e+00 6.59460053e-02
4.49216306e-01 -1.65103972e-01 9.59664226e-01 5.83270788e-01
1.39509594e+00 3.71186197e-01 -2.92099893e-01 4.91139889e-01
1.59633029e+00 2.38396227e-02 4.99590904e-01 -9.65963602e-02
8.87751997e-01 5.44556558e-01 4.80637729e-01 6.25946462e-01
2.77872771e-01 2.19118223e-01 2.75590330e-01 4.37766433e-01
3.73259068e-01 -2.28357762e-01 8.98755565e-02 9.87902641e-01
-1.48015156e-01 -1.27722859e-01 -1.07911611e+00 1.26830742e-01
-2.17949247e+00 -4.70578134e-01 -5.51373661e-01 2.24565482e+00
8.33583772e-01 3.28599215e-01 -5.39845824e-01 -1.49008960e-01
7.16076136e-01 -9.94181708e-02 -7.18780875e-01 -1.85046062e-01
2.49386847e-01 4.58686948e-01 5.43708682e-01 6.02648139e-01
-1.07323158e+00 6.90666139e-01 5.94327641e+00 1.05214870e+00
-7.13032126e-01 2.59544879e-01 2.44396791e-01 1.18531786e-01
-1.86722219e-01 5.64804189e-02 -8.96330297e-01 6.94552958e-01
9.80569184e-01 2.28624139e-02 7.85954371e-02 8.00837576e-01
4.53698575e-01 -5.30063093e-01 -1.12131429e+00 1.07386327e+00
-3.22769970e-01 -8.67672086e-01 -2.19230205e-01 1.83640316e-01
5.46818435e-01 -8.72443430e-03 -9.33031738e-02 4.94792253e-01
5.85099518e-01 -9.83312845e-01 3.43741059e-01 1.21370482e+00
4.41230774e-01 -9.19762433e-01 8.43315542e-01 6.25703096e-01
-8.58612835e-01 2.64248133e-01 -8.51737320e-01 7.02480897e-02
4.21949059e-01 1.28156960e+00 -3.75896752e-01 6.91463947e-01
5.44095874e-01 3.54723990e-01 5.42696454e-02 1.07690418e+00
-3.79245669e-01 7.76824594e-01 -6.51689827e-01 -6.56939894e-02
3.58043797e-02 -6.66590333e-01 6.01615846e-01 1.04200375e+00
4.86943841e-01 9.56110284e-02 6.74264207e-02 1.37057400e+00
3.95525508e-02 1.19937919e-01 -2.60834873e-01 -2.00121012e-02
2.71024555e-01 1.31135190e+00 -5.55192530e-01 -8.57355073e-02
-1.96910948e-01 8.21249902e-01 3.78693163e-01 4.18186069e-01
-9.20689762e-01 -2.53983498e-01 3.87829006e-01 -2.64756620e-01
2.19485551e-01 -5.69281876e-01 -2.13467196e-01 -1.14612794e+00
-1.87935710e-01 7.47907534e-02 1.10502578e-01 -9.46555674e-01
-1.60094595e+00 1.13646269e-01 2.14149475e-01 -8.30780804e-01
-6.34734780e-02 -7.63603806e-01 -6.55031800e-01 1.39423382e+00
-1.67792523e+00 -9.82412338e-01 -2.65517861e-01 4.32539225e-01
9.73910317e-02 1.06584236e-01 8.09469879e-01 1.52505621e-01
-6.73478425e-01 -7.97158256e-02 6.55296147e-01 -1.52092889e-01
8.00654471e-01 -1.55045223e+00 1.77238196e-01 5.00859380e-01
-2.93009967e-01 7.11505413e-01 9.28427398e-01 -9.24253583e-01
-1.45652413e+00 -1.11868191e+00 1.49723992e-01 -3.72591674e-01
8.42997313e-01 -3.84850383e-01 -1.09581852e+00 3.67927045e-01
-3.14906746e-01 1.05235547e-01 6.99633420e-01 4.04123310e-03
7.24550933e-02 1.97949991e-01 -1.36016965e+00 1.63875625e-01
4.73960638e-01 -4.85275686e-01 -5.90271831e-01 3.45147341e-01
8.45249712e-01 -4.88873124e-01 -1.24387002e+00 4.18466002e-01
2.09967062e-01 -4.75600690e-01 8.00895452e-01 -3.68676484e-01
2.58143574e-01 -4.48027045e-01 -2.84589916e-01 -1.33970761e+00
-2.85103559e-01 -4.04265434e-01 -6.26878798e-01 1.40116167e+00
2.33352363e-01 -5.92478514e-01 8.78092170e-01 1.11534131e+00
7.98874721e-02 -5.88707745e-01 -8.99770558e-01 -7.01702893e-01
3.15898329e-01 -7.99237251e-01 6.29932046e-01 4.68067884e-01
-3.30358505e-01 2.43719548e-01 -3.42158586e-01 5.65396309e-01
1.10288143e+00 -3.70469898e-01 5.75186968e-01 -1.67949271e+00
-3.49709302e-01 -4.70676832e-02 -3.37598175e-01 -1.03186440e+00
1.53251201e-01 -6.05957091e-01 3.36462200e-01 -1.76708543e+00
1.76600069e-01 -6.14698887e-01 -3.12423885e-01 -1.58845112e-01
-3.89815390e-01 -3.72672141e-01 -2.54538715e-01 2.86889285e-01
-2.44788513e-01 1.07996416e+00 1.15792263e+00 8.79161730e-02
-5.88239506e-02 1.87212884e-01 -2.71369427e-01 8.51994634e-01
9.51418519e-01 -5.51080167e-01 -4.50822979e-01 -1.48466170e-01
2.14817196e-01 -1.70737710e-02 3.77012223e-01 -1.04206264e+00
4.22080398e-01 -1.84358433e-01 6.06529713e-01 -7.55912423e-01
6.65037394e-01 -8.38539302e-01 6.32152557e-02 -1.01568267e-01
-6.25477284e-02 -7.18466640e-01 2.51671851e-01 1.11197937e+00
-7.08430707e-02 -6.74450397e-01 6.72357202e-01 -1.52574480e-01
-5.20040333e-01 5.73748887e-01 -1.17063656e-01 -1.02921918e-01
8.88633311e-01 1.36420518e-01 7.76225775e-02 -1.77341163e-01
-9.63522375e-01 3.01711142e-01 1.79012924e-01 -3.93836061e-03
5.76415777e-01 -1.31099665e+00 -5.54591835e-01 -4.58774865e-02
3.28593403e-02 7.40868032e-01 5.53313553e-01 6.89652145e-01
-3.46396506e-01 2.83618540e-01 3.34115595e-01 -7.41247833e-01
-4.46739376e-01 5.06665647e-01 2.19406888e-01 -2.19114035e-01
-5.66074908e-01 8.79166484e-01 1.10686064e-01 -8.87673080e-01
-3.89403431e-03 -3.51974875e-01 -2.18244076e-01 -2.36585394e-01
4.19464797e-01 4.84928131e-01 -1.41144767e-01 -4.67112064e-01
-1.41046003e-01 6.78377867e-01 1.14764035e-01 -3.92806500e-01
1.22735417e+00 -2.43224755e-01 -1.77732676e-01 7.13507712e-01
1.01733041e+00 -1.13339268e-01 -1.54468095e+00 -4.21692222e-01
3.63331474e-02 -1.83170974e-01 5.08308351e-01 -4.69039828e-01
-5.62674224e-01 1.20711112e+00 5.41028678e-01 7.28986114e-02
5.42369008e-01 -1.31582066e-01 6.90717578e-01 3.76930207e-01
4.76831377e-01 -1.26587009e+00 -2.41251439e-01 4.55333799e-01
5.81451774e-01 -1.46880782e+00 1.56789556e-01 -5.02573490e-01
-4.70720440e-01 1.08924675e+00 5.36134064e-01 -2.52064437e-01
1.17621505e+00 1.98026925e-01 -3.98952693e-01 -2.11178914e-01
-2.77289987e-01 -6.96809739e-02 1.52336165e-01 6.74882770e-01
2.07270145e-01 1.77326292e-01 -1.18038565e-01 9.55812395e-01
-2.38903854e-02 1.67125121e-01 2.30198562e-01 9.18027759e-01
-8.05335164e-01 -9.02096450e-01 -6.28066123e-01 4.90034878e-01
-2.97314733e-01 -3.03047262e-02 3.02318692e-01 4.09585059e-01
9.44470838e-02 8.73553157e-01 -2.52474900e-02 9.95689854e-02
5.26192486e-02 3.29002917e-01 5.25406122e-01 -5.69019198e-01
2.52796173e-01 2.11592659e-01 -1.92764163e-01 -2.80934811e-01
-3.36356163e-01 -6.06016636e-01 -1.60103631e+00 -2.36849591e-01
-4.41769660e-01 2.17421755e-01 1.09585118e+00 1.14360821e+00
6.96237683e-02 7.39000916e-01 2.88505852e-01 -8.40251446e-01
-7.11507440e-01 -1.18984902e+00 -9.57479894e-01 -7.48319700e-02
8.06506723e-03 -1.03647542e+00 -4.78593022e-01 -2.11398304e-01] | [6.7585859298706055, 3.6082468032836914] |
603f7736-016e-4418-a12c-6843245012d7 | npr-nocturnal-place-recognition-in-street | 2304.00276 | null | https://arxiv.org/abs/2304.00276v2 | https://arxiv.org/pdf/2304.00276v2.pdf | NPR: Nocturnal Place Recognition in Streets | Visual Place Recognition (VPR) is the task of retrieving database images similar to a query photo by comparing it to a large database of known images. In real-world applications, extreme illumination changes caused by query images taken at night pose a significant obstacle that VPR needs to overcome. However, a training set with day-night correspondence for city-scale, street-level VPR does not exist. To address this challenge, we propose a novel pipeline that divides VPR and conquers Nocturnal Place Recognition (NPR). Specifically, we first established a street-level day-night dataset, NightStreet, and used it to train an unpaired image-to-image translation model. Then we used this model to process existing large-scale VPR datasets to generate the VPR-Night datasets and demonstrated how to combine them with two popular VPR pipelines. Finally, we proposed a divide-and-conquer VPR framework and provided explanations at the theoretical, experimental, and application levels. Under our framework, previous methods can significantly improve performance on two public datasets, including the top-ranked method. | ['Hong Zhang', 'Yihong Wu', 'Jinqiang Cui', 'Feng Lu', 'Yujie Fu', 'Bingxi Liu'] | 2023-04-01 | null | null | null | null | ['visual-place-recognition'] | ['computer-vision'] | [ 2.20554933e-01 -5.87326765e-01 1.60986204e-02 -4.93672013e-01
-1.16398776e+00 -9.03761685e-01 7.20406353e-01 -8.00734852e-03
-3.10952663e-01 7.33297229e-01 -5.67144807e-03 -2.39245176e-01
1.25090033e-01 -7.63661146e-01 -8.91468465e-01 -5.30347288e-01
2.18301013e-01 3.09138328e-01 4.14295554e-01 -2.65532583e-01
5.81811488e-01 7.36758411e-01 -1.88887429e+00 2.83786714e-01
7.46090531e-01 6.73544586e-01 7.08678782e-01 5.94251752e-01
1.67986210e-02 7.81976461e-01 -5.66673756e-01 1.56898890e-02
4.85848814e-01 -3.13400030e-01 -5.02050579e-01 1.99703444e-02
6.59728050e-01 -9.57104936e-02 -4.41639304e-01 7.76788652e-01
6.51792705e-01 4.92206991e-01 4.12071705e-01 -1.54361808e+00
-1.13742876e+00 -3.65079105e-01 -5.57236612e-01 2.61882931e-01
5.77978551e-01 2.87074298e-01 8.30451071e-01 -1.00241649e+00
9.50945973e-01 7.46671319e-01 4.37615871e-01 3.11454266e-01
-1.21333456e+00 -4.10337180e-01 -1.25697613e-01 5.82143486e-01
-1.95874643e+00 -5.45043647e-01 8.06173325e-01 -1.77155375e-01
1.21047580e+00 4.96458292e-01 6.60272837e-01 1.11001515e+00
-2.65231699e-01 7.05301762e-01 1.20180583e+00 -1.41395345e-01
1.92195907e-01 5.09094039e-04 -5.05460083e-01 3.17821681e-01
-3.34193334e-02 -1.29357567e-02 -7.17939198e-01 -7.13160262e-02
5.46245098e-01 6.32080063e-02 -4.53696072e-01 -3.85856718e-01
-1.31382191e+00 4.24794346e-01 9.29350853e-01 8.18261504e-02
-3.02694112e-01 -1.54712364e-01 7.10358610e-03 4.69881035e-02
2.12652639e-01 4.76240218e-01 -2.63241112e-01 2.53840201e-02
-1.20475388e+00 9.94307697e-02 5.03847778e-01 9.61759329e-01
1.20420265e+00 -3.58065337e-01 -3.13541561e-01 9.92051721e-01
3.01778942e-01 8.37722957e-01 3.03363293e-01 -1.07164383e+00
4.66225833e-01 4.91583735e-01 3.12447339e-01 -1.27243948e+00
-1.09889343e-01 3.05708162e-02 -5.72339296e-01 -2.07318813e-01
4.10319380e-02 6.68871999e-01 -9.60874736e-01 1.50680161e+00
2.68438518e-01 3.61975998e-01 2.08162069e-01 1.19958103e+00
9.54443038e-01 9.26732600e-01 -3.43745351e-01 -3.80010307e-02
1.28776884e+00 -1.25037694e+00 -4.38546002e-01 -5.48823953e-01
3.45083565e-01 -7.87202120e-01 1.31903398e+00 -8.86689276e-02
-5.51877677e-01 -3.80235165e-01 -1.10353565e+00 -4.87731040e-01
-7.39623070e-01 2.12432995e-01 2.55746961e-01 2.10636288e-01
-1.51551378e+00 2.90681958e-01 -4.99841809e-01 -8.60589266e-01
3.04978460e-01 -6.51154518e-02 -6.08335257e-01 -3.95469248e-01
-8.99879098e-01 7.91180015e-01 1.83529764e-01 3.38505268e-01
-1.10230064e+00 -4.63090390e-01 -9.78922307e-01 -1.65076450e-01
1.24668874e-01 -6.04392588e-01 1.02196145e+00 -6.12978339e-01
-1.07904565e+00 1.40776992e+00 -6.95280254e-01 -3.93879801e-01
3.90758753e-01 2.21272379e-01 -6.21349335e-01 1.36439472e-01
5.96171439e-01 9.44779038e-01 5.82991660e-01 -1.38091838e+00
-6.26068294e-01 -4.93787467e-01 4.64173630e-02 4.61393356e-01
3.51513237e-01 -2.51869440e-01 -9.45459247e-01 -2.26603016e-01
4.07613218e-01 -1.04069817e+00 -1.36169150e-01 1.71309218e-01
-1.82554170e-01 1.34449750e-01 7.61543930e-01 -6.36599243e-01
7.32136607e-01 -2.59433031e+00 -4.86412048e-01 5.57251722e-02
4.39375415e-02 3.10151391e-02 -3.51327509e-01 5.77468395e-01
7.08397701e-02 -1.23483397e-01 -3.41680884e-01 -3.42700571e-01
-8.20684209e-02 5.26577175e-01 -6.48124754e-01 7.03865945e-01
6.84285238e-02 1.06481588e+00 -1.05977523e+00 -5.10939062e-01
4.74285781e-01 4.77087289e-01 -4.55829203e-01 1.72901258e-01
-1.16778605e-01 5.70728540e-01 1.37592897e-01 1.02458274e+00
7.98583984e-01 -2.72459954e-01 9.67207178e-03 -5.16504794e-02
-4.72461462e-01 2.33125716e-01 -9.21111763e-01 1.85921288e+00
-3.00531745e-01 1.27117085e+00 -2.99973339e-01 -6.11503839e-01
1.23926747e+00 -1.89107388e-01 2.81789452e-01 -1.53389597e+00
-3.85727674e-01 5.09090185e-01 -5.89653313e-01 -3.52522939e-01
1.09794343e+00 3.06260645e-01 4.25571762e-02 -1.14473859e-02
-2.62787282e-01 -1.21772476e-01 1.09497108e-01 -1.01260312e-01
1.07154882e+00 9.76108611e-02 2.12821722e-01 4.94529977e-02
5.04197538e-01 3.64564925e-01 7.31853604e-01 8.39628518e-01
-6.26607597e-01 1.31653774e+00 1.15534343e-01 -4.90463495e-01
-9.90084767e-01 -1.24775541e+00 -3.33373398e-02 8.94360662e-01
6.46739721e-01 -3.60406697e-01 -2.69784391e-01 -3.72312188e-01
-1.40930206e-01 8.28795314e-01 -5.75590670e-01 1.27393514e-01
-2.76220918e-01 -5.62741339e-01 5.72558105e-01 3.05105329e-01
1.02767372e+00 -1.20700085e+00 -6.28108263e-01 -2.24821657e-01
-7.67901719e-01 -1.27817059e+00 -5.18423438e-01 -1.33045718e-01
-2.32157603e-01 -1.19299448e+00 -7.25779653e-01 -8.89947593e-01
6.95270777e-01 1.09712756e+00 1.19539428e+00 -1.65025890e-01
-4.11430001e-01 4.29189295e-01 -3.00342053e-01 -1.33686349e-01
1.40032589e-01 -1.53709143e-01 6.87189698e-02 2.25906763e-02
4.07281488e-01 -7.01461494e-01 -9.50353801e-01 6.07793868e-01
-8.34399641e-01 -7.41581321e-02 5.56225002e-01 5.58142781e-01
1.11998606e+00 -1.95296809e-01 2.25911085e-02 -1.76978454e-01
2.21614420e-01 -2.83104450e-01 -9.40762699e-01 4.96869326e-01
-3.47132266e-01 -8.69316235e-02 4.25094128e-01 -1.84943289e-01
-6.69140816e-01 2.49742761e-01 -1.26131460e-01 -7.20329762e-01
-1.06642582e-01 9.24132243e-02 -2.91945934e-01 -1.83000267e-01
7.93921292e-01 8.49753380e-01 -4.82841045e-01 -1.61526948e-01
5.71363866e-01 8.05223823e-01 9.19815660e-01 -2.83538014e-01
8.84968519e-01 8.22843313e-01 -1.09640107e-01 -1.05435979e+00
-6.95016086e-01 -8.71274173e-01 -4.48671043e-01 -6.17867224e-02
9.31398511e-01 -1.20178843e+00 -5.92616022e-01 2.02585667e-01
-1.12499082e+00 -4.25696701e-01 -3.82980943e-01 2.46048242e-01
-6.64136112e-01 1.01147942e-01 1.20316029e-01 -6.79488778e-01
-2.36511841e-01 -9.35424685e-01 1.41356075e+00 3.78574520e-01
2.21207485e-01 -6.59810245e-01 4.52360392e-01 4.14581627e-01
3.93827438e-01 1.51561588e-01 2.96464592e-01 -4.39883620e-01
-1.02725196e+00 1.37988940e-01 -5.58347642e-01 -1.21796586e-01
1.20987453e-01 -1.82769448e-02 -1.21294391e+00 -1.60034254e-01
-9.91176143e-02 -6.79922551e-02 8.17193389e-01 7.15911984e-02
1.03551400e+00 -2.73280352e-01 -2.98229873e-01 9.98009086e-01
1.61818635e+00 1.12177461e-01 1.12927914e+00 8.32505465e-01
7.44673550e-01 1.57734200e-01 8.59153032e-01 2.60955662e-01
9.02884066e-01 8.50931704e-01 4.05239731e-01 -2.93260962e-01
-9.21388343e-02 -7.60021746e-01 2.34525636e-01 4.48376060e-01
1.01379007e-01 -7.30781406e-02 -1.07645428e+00 9.48670447e-01
-1.95075166e+00 -1.14264572e+00 -7.19384849e-02 2.37296128e+00
2.67802775e-01 -3.50622296e-01 -2.38466546e-01 -3.93549412e-01
6.17343545e-01 6.00536108e-01 -6.19450688e-01 8.98111425e-03
-6.34628773e-01 -3.12977999e-01 7.77527571e-01 2.26664439e-01
-1.11330009e+00 1.13204932e+00 6.08326912e+00 6.25309765e-01
-1.30721724e+00 9.11821201e-02 3.04769307e-01 1.13487691e-01
-1.80538446e-01 2.66159594e-01 -6.85182929e-01 2.37913996e-01
7.84363270e-01 -1.88465565e-01 9.92299914e-01 1.05342352e+00
2.32342675e-01 -2.05427900e-01 -1.07724357e+00 1.74803627e+00
4.86475378e-01 -1.29114068e+00 -6.76253065e-02 1.04962736e-01
9.03210819e-01 7.03010201e-01 6.67881519e-02 4.00350392e-01
1.09257556e-01 -9.58105028e-01 5.07025838e-01 4.52822298e-01
6.40950322e-01 -3.13095003e-01 3.66228789e-01 3.01191896e-01
-1.39942479e+00 7.23221246e-03 -7.10941911e-01 1.32199362e-01
-6.91712201e-02 1.04071602e-01 -1.08674037e+00 7.63303518e-01
9.05977786e-01 9.16425586e-01 -1.10218060e+00 1.34987915e+00
-4.38495338e-01 3.71066593e-02 -3.96925062e-01 2.81434119e-01
7.39356205e-02 -2.84250259e-01 4.40600544e-01 7.83732831e-01
5.33385277e-01 -3.30858529e-02 -4.35238406e-02 1.02619219e+00
-2.68928379e-01 5.63222207e-02 -1.11563301e+00 1.11461788e-01
6.15546405e-01 1.47788620e+00 -6.66439354e-01 -2.31162772e-01
-2.36986309e-01 1.48445177e+00 3.52020204e-01 5.01744926e-01
-9.94666934e-01 -2.11626083e-01 8.22623551e-01 1.00659117e-01
3.73253971e-01 -3.14738005e-01 3.71358618e-02 -1.48979509e+00
3.23069572e-01 -6.29153192e-01 3.02543998e-01 -1.50751436e+00
-1.18197215e+00 5.51373899e-01 -2.95759737e-01 -1.53173316e+00
2.03681625e-02 -2.90365875e-01 -4.35324609e-01 7.50658751e-01
-1.91702449e+00 -1.20414615e+00 -7.51120925e-01 7.92477787e-01
6.33515835e-01 1.01972453e-01 8.98542285e-01 4.34693515e-01
-3.57900441e-01 3.15253973e-01 5.12493968e-01 1.68425232e-01
9.00324106e-01 -1.03496778e+00 6.33752644e-01 1.02886081e+00
5.82173109e-01 5.54493546e-01 4.94968981e-01 -3.91407490e-01
-1.53114057e+00 -1.48808491e+00 9.44439530e-01 -6.82835400e-01
3.72399539e-01 -4.55166489e-01 -8.11306953e-01 6.77830219e-01
2.10521687e-02 4.47433323e-01 4.87393171e-01 -2.60890990e-01
-5.21472454e-01 -3.39791656e-01 -1.14513326e+00 9.64834511e-01
1.12677777e+00 -1.12161005e+00 -6.86096787e-01 5.06614506e-01
7.05611110e-01 -5.68091989e-01 -2.99475342e-01 1.39898911e-01
3.24584395e-01 -1.01267719e+00 1.13532209e+00 8.48039165e-02
2.49513373e-01 -9.22986269e-01 -8.08691740e-01 -1.11514437e+00
-1.23402469e-01 -3.88227433e-01 4.26075786e-01 1.19905567e+00
2.52994657e-01 -6.84135079e-01 4.43838388e-01 5.74304342e-01
2.29852274e-02 -1.66313991e-01 -9.71215606e-01 -9.50945497e-01
-6.11315012e-01 -3.98259223e-01 7.05335975e-01 1.05295050e+00
-5.59863329e-01 1.68482944e-01 -2.60172963e-01 6.90719426e-01
5.28473139e-01 4.93713111e-01 9.92090762e-01 -7.76818335e-01
-1.19526833e-02 1.18294060e-02 -6.82307243e-01 -1.07258439e+00
-2.20820867e-02 -1.07399952e+00 4.04583395e-01 -2.06273031e+00
2.38328770e-01 -8.29501972e-02 -3.30657333e-01 6.08896554e-01
9.15998369e-02 7.39232898e-01 2.91749448e-01 7.06850111e-01
-1.08902204e+00 7.18562067e-01 8.59122515e-01 -3.19262654e-01
-4.67514038e-01 -3.89319748e-01 -4.96213675e-01 3.16035926e-01
7.48567581e-01 -3.82821888e-01 -3.44074637e-01 -4.42269951e-01
1.89314008e-01 -2.06175685e-01 6.81170821e-01 -1.22008610e+00
4.03964132e-01 -2.04024523e-01 4.74736869e-01 -1.01937532e+00
5.77328861e-01 -7.86199868e-01 3.38288337e-01 -3.02006677e-02
1.83149111e-02 1.52009889e-01 1.76090822e-01 5.06706178e-01
-3.05802405e-01 2.43463978e-01 5.41569829e-01 -1.10637404e-01
-1.31751943e+00 1.62759632e-01 8.97515416e-02 7.13563487e-02
1.14419246e+00 -3.10002774e-01 -8.24863911e-01 -2.75588006e-01
-2.81274527e-01 3.69072229e-01 1.03814352e+00 7.08770871e-01
9.09784138e-01 -1.38786888e+00 -4.90116268e-01 3.86582315e-01
8.63317668e-01 8.30591619e-02 3.19527298e-01 7.78344333e-01
-8.33362758e-01 5.23120046e-01 -2.46684402e-01 -8.54422748e-01
-1.11390531e+00 6.96510196e-01 3.16642016e-01 1.31628543e-01
-5.93727052e-01 3.97596776e-01 1.55634895e-01 -9.07470047e-01
-8.34843069e-02 -8.06741789e-02 -2.03762278e-02 -7.69800842e-02
5.95076621e-01 3.54855396e-02 1.08609170e-01 -1.04836619e+00
-7.63636112e-01 5.41867912e-01 4.49656695e-01 -2.71011263e-01
1.38203895e+00 -4.10433978e-01 -1.87888667e-01 6.71898365e-01
1.37958515e+00 -6.13033548e-02 -1.14658248e+00 -1.04161575e-01
-1.84434548e-01 -9.12866473e-01 -3.07611078e-01 -5.75572670e-01
-6.78890526e-01 7.65396237e-01 8.87823999e-01 -7.17881322e-02
1.22011507e+00 1.80503577e-02 5.10300100e-01 8.25335205e-01
6.98729873e-01 -1.11728454e+00 -1.76381096e-01 6.81067646e-01
8.45742226e-01 -1.48072553e+00 -1.78340897e-01 -6.29863366e-02
-7.48304069e-01 6.23249829e-01 4.80673701e-01 4.20331582e-02
2.13713646e-01 -4.42744911e-01 2.24974349e-01 -1.30452663e-01
-5.94570100e-01 -5.72714627e-01 2.64851362e-01 8.01660657e-01
-1.25918403e-01 -8.45330022e-03 3.38213742e-02 4.26797532e-02
-3.67888361e-01 1.42456189e-01 3.17804635e-01 9.19729352e-01
-3.04122746e-01 -6.49384141e-01 -5.33710539e-01 -1.00796185e-01
1.58712536e-01 -2.80320585e-01 -4.50028419e-01 5.96892834e-01
-2.83243880e-02 9.33187485e-01 1.35822058e-01 -4.49971527e-01
5.13291776e-01 -1.50423929e-01 2.45302543e-01 -2.76018471e-01
-1.49241716e-01 -3.51438940e-01 -1.86961845e-01 -9.16239083e-01
-3.88983250e-01 -4.71017718e-01 -1.02229166e+00 -2.06921697e-01
2.48208940e-01 -9.32728350e-02 8.99662375e-01 6.55312002e-01
8.08746040e-01 1.91054910e-01 8.91021848e-01 -1.01746976e+00
1.14222668e-01 -4.49690014e-01 -6.06346309e-01 4.73950833e-01
2.87505627e-01 -4.29850966e-01 -3.73244762e-01 -1.16820922e-02] | [7.646080493927002, -1.8372936248779297] |
1ac07ab1-8fba-4189-81d6-42bd7f828a7b | unsupervised-intrinsic-image-decomposition | 2303.10820 | null | https://arxiv.org/abs/2303.10820v2 | https://arxiv.org/pdf/2303.10820v2.pdf | Unsupervised Intrinsic Image Decomposition with LiDAR Intensity | Intrinsic image decomposition (IID) is the task that decomposes a natural image into albedo and shade. While IID is typically solved through supervised learning methods, it is not ideal due to the difficulty in observing ground truth albedo and shade in general scenes. Conversely, unsupervised learning methods are currently underperforming supervised learning methods since there are no criteria for solving the ill-posed problems. Recently, light detection and ranging (LiDAR) is widely used due to its ability to make highly precise distance measurements. Thus, we have focused on the utilization of LiDAR, especially LiDAR intensity, to address this issue. In this paper, we propose unsupervised intrinsic image decomposition with LiDAR intensity (IID-LI). Since the conventional unsupervised learning methods consist of image-to-image transformations, simply inputting LiDAR intensity is not an effective approach. Therefore, we design an intensity consistency loss that computes the error between LiDAR intensity and gray-scaled albedo to provide a criterion for the ill-posed problem. In addition, LiDAR intensity is difficult to handle due to its sparsity and occlusion, hence, a LiDAR intensity densification module is proposed. We verified the estimating quality using our own dataset, which include RGB images, LiDAR intensity and human judged annotations. As a result, we achieved an estimation accuracy that outperforms conventional unsupervised learning methods. Dataset link : (https://github.com/ntthilab-cv/NTT-intrinsic-dataset). | ['Jun Shimamura', 'Shingo Ando', 'Takuhiro Kaneko', 'Taiga Yoshida', 'Yasuhiro Yao', 'Shogo Sato'] | 2023-03-20 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Sato_Unsupervised_Intrinsic_Image_Decomposition_With_LiDAR_Intensity_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Sato_Unsupervised_Intrinsic_Image_Decomposition_With_LiDAR_Intensity_CVPR_2023_paper.pdf | cvpr-2023-1 | ['intrinsic-image-decomposition'] | ['computer-vision'] | [ 2.75272816e-01 -4.16099101e-01 -5.80328032e-02 -5.11084199e-01
-6.35056555e-01 -2.60469586e-01 3.08346301e-01 -4.30520140e-02
-4.80764091e-01 7.37468004e-01 -3.14918816e-01 -7.94718340e-02
3.09850294e-02 -9.35843050e-01 -4.86658901e-01 -9.17527676e-01
2.96005189e-01 2.37709552e-01 -3.21846381e-02 1.42861500e-01
1.04412183e-01 5.78820467e-01 -1.87521589e+00 -3.44380587e-01
1.36114955e+00 1.03817296e+00 3.33766937e-01 3.21475714e-01
-4.94642258e-01 6.07537150e-01 -3.92543823e-01 -5.36873983e-03
5.64723432e-01 -4.77045774e-01 -3.11786205e-01 3.14650983e-01
6.67122066e-01 -4.65568841e-01 4.56677042e-02 1.21853817e+00
3.93837452e-01 1.38836235e-01 7.80028880e-01 -1.56829035e+00
-2.10085616e-01 -1.29392669e-01 -8.96903813e-01 -1.64688215e-01
3.06264848e-01 1.52359053e-01 9.40758228e-01 -1.01942992e+00
2.19364449e-01 8.73720229e-01 4.66848701e-01 1.20241255e-01
-1.32327592e+00 -6.87786758e-01 -1.67675301e-01 4.13334370e-02
-1.54828703e+00 -2.88462222e-01 9.41627681e-01 -3.75164658e-01
3.84476990e-01 1.12971991e-01 6.89742148e-01 5.96144497e-01
-4.14633267e-02 5.77543974e-01 1.67529821e+00 -6.32870436e-01
2.83071011e-01 2.75623858e-01 -2.30807155e-01 7.38948941e-01
4.33686376e-01 1.92504480e-01 -5.70928276e-01 2.52765507e-01
6.27232909e-01 1.76027536e-01 -2.96129704e-01 -3.76696050e-01
-9.45980787e-01 6.34318948e-01 6.84009433e-01 -1.13543712e-01
-3.39164466e-01 -2.39363257e-02 -1.49142593e-01 8.50000679e-02
6.08226776e-01 1.39608935e-01 -2.69187689e-01 1.77944154e-02
-1.02414346e+00 -5.96596710e-02 5.40119886e-01 6.51033461e-01
1.49717391e+00 1.33233801e-01 5.24161756e-01 8.67400944e-01
4.32837665e-01 9.22538221e-01 6.15170635e-02 -1.10569191e+00
1.74814209e-01 9.46967423e-01 1.06391467e-01 -1.06926131e+00
-2.57317930e-01 -2.14652494e-01 -9.97150421e-01 6.92728639e-01
4.86891061e-01 -6.60323398e-03 -9.57060933e-01 1.51806045e+00
4.91781235e-01 3.06652673e-02 7.16443062e-02 1.17161429e+00
6.92673147e-01 5.45813203e-01 -4.01551053e-02 -4.58310336e-01
9.90142822e-01 -6.04505420e-01 -7.93679357e-01 -2.57131368e-01
3.84673387e-01 -9.63861227e-01 1.39960611e+00 5.27037144e-01
-7.55927384e-01 -6.54024601e-01 -1.01745522e+00 -8.58689770e-02
-4.94698018e-01 1.49281755e-01 6.88194692e-01 7.22490489e-01
-6.80115342e-01 2.40099758e-01 -5.47803819e-01 -3.61014336e-01
2.47712314e-01 1.16236895e-01 -3.21949989e-01 -2.44076133e-01
-9.39656317e-01 7.25110888e-01 1.47842363e-01 1.91080257e-01
-6.00445211e-01 -5.97395837e-01 -8.04649115e-01 -2.73473799e-01
3.81059110e-01 -5.14130294e-01 7.58010447e-01 -9.84781861e-01
-1.37438333e+00 8.56344521e-01 -1.77649245e-01 -1.13280363e-01
3.71302009e-01 -6.75083771e-02 -1.53655589e-01 2.94804215e-01
2.38412634e-01 8.02072823e-01 9.94951069e-01 -1.55696774e+00
-5.68581700e-01 -4.07239169e-01 5.66588975e-02 3.20565820e-01
-3.36598724e-01 -3.91426563e-01 -2.86956549e-01 -3.97527874e-01
6.18368864e-01 -9.56319213e-01 -1.30679710e-02 2.83228427e-01
-3.71773392e-01 8.04843381e-02 9.26568985e-01 -3.41922432e-01
7.86765575e-01 -2.14777350e+00 -3.12787116e-01 1.41329259e-01
1.63819604e-02 9.46535822e-03 2.07584426e-01 7.01810941e-02
6.70417994e-02 -2.87121325e-03 -6.66706383e-01 -3.82870913e-01
-2.75095344e-01 3.81378800e-01 -1.21520124e-01 5.00984550e-01
1.29314736e-01 4.20595884e-01 -9.13147807e-01 -8.61244023e-01
6.11501634e-01 5.81812084e-01 -3.20599794e-01 3.43118906e-01
-2.09696949e-01 6.28100216e-01 -1.71091601e-01 9.71734583e-01
1.06830812e+00 8.68149400e-02 -1.29800573e-01 -5.22210360e-01
-3.51679593e-01 -1.41491536e-02 -1.36768460e+00 1.53755176e+00
-6.57533467e-01 5.74907780e-01 -1.44603560e-02 -7.39520788e-01
1.15696084e+00 -2.54987255e-02 7.56831408e-01 -7.71837294e-01
3.05517372e-02 2.86614299e-01 -3.06102157e-01 -5.46427786e-01
2.81800896e-01 -3.85490984e-01 2.57022768e-01 4.25802439e-01
-2.32452348e-01 -8.10653150e-01 1.00182690e-01 8.11224952e-02
6.54286146e-01 4.41694826e-01 1.56250209e-01 9.86227766e-03
6.70812845e-01 1.66879997e-01 7.17641234e-01 4.62205410e-01
-1.95965841e-01 9.20138240e-01 1.52035251e-01 -3.37972820e-01
-7.76521325e-01 -1.25744832e+00 -3.27006042e-01 5.82907319e-01
3.12026143e-01 -1.70986027e-01 -7.54738927e-01 -3.49805474e-01
-1.67702660e-02 6.19392931e-01 -2.36202642e-01 1.84336752e-01
-1.14395209e-01 -6.36619508e-01 3.04943532e-01 1.62863970e-01
1.12173581e+00 -7.77170599e-01 -7.03563929e-01 -1.40252516e-01
-3.18892360e-01 -1.26486707e+00 -1.63654104e-01 1.39470562e-01
-1.00863910e+00 -1.21027803e+00 -4.22263682e-01 -3.19948941e-01
8.60110879e-01 7.46751070e-01 8.97118270e-01 1.69987366e-01
-5.43094099e-01 5.18114328e-01 -3.07227731e-01 -5.30640125e-01
-1.42596997e-02 -1.58070594e-01 2.13045061e-01 1.90515399e-01
3.12906116e-01 -6.83266580e-01 -7.65217185e-01 4.23310876e-01
-1.13515913e+00 1.49492398e-01 5.56784689e-01 6.03996813e-01
8.44270051e-01 2.00250357e-01 1.11945912e-01 -7.99264669e-01
2.53712863e-01 -1.88703891e-02 -9.18357670e-01 6.49690405e-02
-7.83513188e-01 -1.60520539e-01 3.61511618e-01 -1.07480757e-01
-1.35029697e+00 5.06075144e-01 1.51709020e-01 -4.19676334e-01
-3.99744064e-01 5.02306700e-01 -3.91118258e-01 -2.08493024e-01
6.27209127e-01 2.48899549e-01 2.12487951e-01 -2.61280060e-01
2.93768406e-01 7.75418997e-01 4.33573067e-01 -6.08000755e-01
1.14619148e+00 7.29373276e-01 2.55893171e-01 -1.22287893e+00
-1.05839980e+00 -6.36802197e-01 -7.72699773e-01 -5.25393486e-01
8.42709899e-01 -1.10072386e+00 -5.35695612e-01 4.25882995e-01
-9.70741034e-01 -2.72185981e-01 -1.63851202e-01 6.20227098e-01
-3.99892002e-01 5.31788766e-01 -1.18174702e-01 -1.06729245e+00
-1.31610513e-01 -1.10574234e+00 1.00255096e+00 3.45504671e-01
2.39792436e-01 -8.50638807e-01 3.10196262e-02 6.37494504e-01
2.51701802e-01 5.46544135e-01 5.65409899e-01 3.99410993e-01
-7.85697043e-01 -1.62454378e-02 -5.63382268e-01 8.16898644e-01
4.87871617e-01 2.20420793e-01 -1.17999613e+00 -4.06149067e-02
1.07849821e-01 -4.38329399e-01 6.95747137e-01 2.66962767e-01
1.19178987e+00 9.63570699e-02 1.92398801e-01 8.20001066e-01
1.77657735e+00 -1.41325772e-01 7.85642087e-01 4.91809607e-01
7.32849538e-01 7.68783867e-01 1.00580871e+00 4.30529386e-01
3.76293093e-01 5.93891501e-01 6.22907281e-01 -5.02672732e-01
-2.89717555e-01 -1.81610242e-01 2.53325939e-01 5.71637034e-01
-3.37818056e-01 1.45763248e-01 -9.17070210e-01 2.83519149e-01
-1.53629720e+00 -7.39683568e-01 -5.45259833e-01 2.49567437e+00
7.75520563e-01 1.33204341e-01 -7.39912391e-02 3.59603226e-01
3.44150573e-01 9.04495716e-02 -4.85439330e-01 -1.04263335e-01
-2.90946573e-01 1.19407669e-01 6.72620416e-01 6.60542727e-01
-9.02094364e-01 8.44656289e-01 5.11973476e+00 5.99341869e-01
-1.29433048e+00 5.52350618e-02 5.32080233e-01 2.99788833e-01
-3.30370337e-01 1.93427533e-01 -5.99287212e-01 3.88479203e-01
3.84736419e-01 2.48630300e-01 3.99224550e-01 7.28106916e-01
6.42268360e-01 -7.63681173e-01 -6.28555775e-01 1.16331589e+00
1.85303688e-02 -6.06194615e-01 -1.33085951e-01 9.71256346e-02
7.15753734e-01 -5.52742481e-02 -4.34831269e-02 -8.14403519e-02
-1.35048777e-01 -7.47391999e-01 3.86610299e-01 5.46955764e-01
9.44079697e-01 -5.04734337e-01 4.99499172e-01 4.34383720e-01
-1.12661767e+00 1.90551892e-01 -6.17798567e-01 -1.90871790e-01
6.56360164e-02 1.18396115e+00 -7.01231062e-01 5.20480990e-01
8.13305259e-01 5.73122263e-01 -6.69510543e-01 1.02088618e+00
-7.14428186e-01 5.25989830e-01 -6.60741806e-01 3.58959019e-01
3.21284235e-02 -8.44587982e-01 3.30339462e-01 8.33260059e-01
3.67760599e-01 2.93637738e-02 2.66648382e-01 9.69190121e-01
-6.81112614e-03 2.49401942e-01 -7.28508055e-01 2.83209741e-01
3.59047383e-01 1.51795673e+00 -7.87435234e-01 -6.36674538e-02
-4.45568264e-01 8.31561983e-01 9.06509906e-02 5.51716685e-01
-8.31670761e-01 -2.40639403e-01 5.80430210e-01 2.98112720e-01
-1.28222749e-01 -4.62892979e-01 -4.67860639e-01 -1.15288353e+00
-6.46418110e-02 -6.39855921e-01 1.70599490e-01 -1.10688019e+00
-1.10170150e+00 2.57919908e-01 3.20576914e-02 -1.61878860e+00
1.34764194e-01 -6.55345857e-01 -4.98626620e-01 7.49958992e-01
-1.96852469e+00 -1.10057080e+00 -1.11694062e+00 7.03117192e-01
3.63095343e-01 2.26541251e-01 6.35392070e-01 4.39258337e-01
-6.29646659e-01 1.42426223e-01 -4.59956005e-02 -4.56950776e-02
9.32768941e-01 -1.31954753e+00 -4.24833298e-01 8.20797086e-01
1.71556041e-01 4.26040709e-01 7.35565722e-01 -3.99865210e-01
-1.45480037e+00 -9.48939502e-01 4.99764681e-01 4.90411092e-03
4.06027108e-01 -2.54020065e-01 -8.36626410e-01 2.90982306e-01
2.97827274e-02 3.59673947e-01 6.15663767e-01 -2.83144087e-01
-2.69903690e-01 -7.42264211e-01 -1.21118486e+00 4.56600815e-01
7.81522393e-01 -5.96544981e-01 -2.59792268e-01 3.22807461e-01
4.10174876e-01 -9.49540287e-02 -7.65559077e-01 5.76580167e-01
4.44170535e-01 -1.34070170e+00 1.05879569e+00 2.44742498e-01
3.56488645e-01 -7.89900124e-01 -3.05487245e-01 -1.01342332e+00
2.75745571e-01 -1.72709525e-01 1.21693090e-01 1.37467051e+00
3.15362394e-01 -7.53980279e-01 9.16082382e-01 4.97154087e-01
2.15559244e-01 -4.30853158e-01 -6.99019551e-01 -7.32400358e-01
-3.58640581e-01 -6.10576451e-01 3.43843609e-01 9.43462610e-01
-6.16832376e-01 2.42598385e-01 -3.85027558e-01 3.61180246e-01
1.22555006e+00 3.27688038e-01 1.17638254e+00 -1.24282813e+00
-2.83311028e-03 -1.51474357e-01 -2.16635615e-01 -8.37438583e-01
3.46637517e-02 -6.13791466e-01 2.75265008e-01 -1.54001284e+00
-1.90478433e-02 -7.19214320e-01 -5.28763123e-02 3.93837124e-01
-1.83578953e-01 6.57806575e-01 8.29387680e-02 5.32785654e-01
-2.86251038e-01 6.32571816e-01 1.11925817e+00 -2.34575823e-01
-2.78743833e-01 -9.41399932e-02 -3.05070311e-01 8.69253159e-01
1.13835740e+00 -5.36807239e-01 -4.33248729e-01 -4.23955053e-01
1.98023185e-01 -2.54717171e-01 3.33323121e-01 -1.15633321e+00
2.68591195e-01 -3.89360219e-01 3.66994500e-01 -7.74371684e-01
4.45008695e-01 -9.73495364e-01 6.74991459e-02 3.59418839e-01
1.93093956e-01 -3.18063378e-01 -1.56411245e-01 2.92275339e-01
-4.10135865e-01 -3.01813304e-01 8.53594124e-01 -1.81865722e-01
-7.64657676e-01 3.75769287e-01 -1.39229923e-01 -1.68368623e-01
7.23728955e-01 -5.11833966e-01 1.33398278e-02 -4.63868618e-01
-1.50324732e-01 3.89131205e-03 8.08127820e-01 -8.66734460e-02
8.53100955e-01 -1.08261585e+00 -5.23544848e-01 2.20776707e-01
3.33990842e-01 3.93643230e-01 -7.94890597e-02 1.04437137e+00
-8.88554096e-01 -1.29024625e-01 -8.36994350e-02 -8.95686924e-01
-1.29609382e+00 3.03549081e-01 2.30991766e-01 1.66002870e-01
-3.71053427e-01 3.42461646e-01 1.26519218e-01 -6.40352309e-01
2.26817757e-01 -1.11864924e-01 4.65836860e-02 1.78845853e-01
1.84882894e-01 2.87675649e-01 7.14467047e-03 -6.42089128e-01
-2.76692063e-01 9.36453581e-01 5.51617920e-01 4.72330712e-02
1.14855945e+00 -4.08645898e-01 -4.04906929e-01 4.98513043e-01
1.12382913e+00 9.70433578e-02 -1.22871375e+00 -1.24833331e-01
1.29553850e-03 -8.69190812e-01 4.42416191e-01 -3.59976530e-01
-1.18735254e+00 1.08180559e+00 8.67909491e-01 1.15178704e-01
1.51634693e+00 -4.19863731e-01 6.15327477e-01 5.35205483e-01
4.37966377e-01 -1.16792786e+00 2.52072006e-01 2.98287153e-01
5.78645408e-01 -1.76898646e+00 5.19887745e-01 -6.86264396e-01
-3.36435139e-01 1.04158998e+00 7.20148802e-01 1.98302329e-01
5.69502890e-01 1.95395604e-01 5.17915428e-01 -1.74422301e-02
-5.43577336e-02 -6.97156250e-01 3.63628305e-02 5.33793807e-01
4.73019809e-01 -5.12068644e-02 -2.58372784e-01 -2.55197763e-01
-8.54866803e-02 5.21956161e-02 4.52543885e-01 7.92022169e-01
-5.75488687e-01 -1.09781718e+00 -6.95521533e-01 2.20747411e-01
-6.00353554e-02 2.64879707e-02 -1.27416208e-01 8.23715746e-01
4.32964921e-01 1.07025909e+00 -7.58068562e-02 -2.74511755e-01
2.91503847e-01 -1.88320965e-01 4.65761334e-01 -5.08460104e-01
1.70665607e-01 1.21419959e-01 -3.08049560e-01 -4.62331384e-01
-9.14109826e-01 -4.75770265e-01 -1.26300812e+00 -1.62538797e-01
-4.67772871e-01 1.32403485e-02 9.97373343e-01 7.05169141e-01
-1.52558833e-01 1.37309313e-01 1.03655434e+00 -9.13898230e-01
-1.24883741e-01 -7.69179165e-01 -6.93347156e-01 4.65766549e-01
1.10067256e-01 -9.40029860e-01 -8.53412807e-01 1.05082713e-01] | [9.983009338378906, -2.6689023971557617] |
54f610fa-54ba-4093-adb6-0cae97a5124e | metric-learning-and-adaptive-boundary-for-out | 2204.10849 | null | https://arxiv.org/abs/2204.10849v1 | https://arxiv.org/pdf/2204.10849v1.pdf | Metric Learning and Adaptive Boundary for Out-of-Domain Detection | Conversational agents are usually designed for closed-world environments. Unfortunately, users can behave unexpectedly. Based on the open-world environment, we often encounter the situation that the training and test data are sampled from different distributions. Then, data from different distributions are called out-of-domain (OOD). A robust conversational agent needs to react to these OOD utterances adequately. Thus, the importance of robust OOD detection is emphasized. Unfortunately, collecting OOD data is a challenging task. We have designed an OOD detection algorithm independent of OOD data that outperforms a wide range of current state-of-the-art algorithms on publicly available datasets. Our algorithm is based on a simple but efficient approach of combining metric learning with adaptive decision boundary. Furthermore, compared to other algorithms, we have found that our proposed algorithm has significantly improved OOD performance in a scenario with a lower number of classes while preserving the accuracy for in-domain (IND) classes. | ['Jan Šedivý', 'Ondřej Kobza', 'Petr Marek', 'Jakub Konrád', 'Jan Pichl', 'Tommaso Gargiani', 'Petr Lorenc'] | 2022-04-22 | null | null | null | null | ['open-intent-detection'] | ['natural-language-processing'] | [-6.51694313e-02 3.24396826e-02 5.04606552e-02 -5.09292722e-01
-7.20340312e-01 -6.23713374e-01 8.28717589e-01 -2.69525703e-02
-2.52687693e-01 9.01579738e-01 2.49401674e-01 -7.19951689e-02
1.20844394e-01 -6.34018242e-01 -1.04388520e-01 -7.35414207e-01
-1.68762475e-01 9.55527663e-01 3.41719806e-01 -1.55830517e-01
3.39765280e-01 2.25737125e-01 -1.74538219e+00 1.78071737e-01
1.05328417e+00 7.49753058e-01 2.03527838e-01 7.95275629e-01
-3.55704784e-01 7.56585002e-01 -1.34729254e+00 -1.07358955e-01
2.48182565e-01 -2.93739915e-01 -8.26848865e-01 3.39950651e-01
1.10583171e-01 -5.12084901e-01 -2.93385535e-01 8.82831633e-01
8.42258215e-01 4.75434124e-01 8.95344496e-01 -1.61729085e+00
-3.62575829e-01 5.56646362e-02 -2.01078221e-01 6.26966804e-02
8.13174963e-01 6.43450022e-02 7.50589192e-01 -7.26719499e-01
7.14105725e-01 1.35610104e+00 3.45412344e-01 8.24207127e-01
-9.38027501e-01 -5.36058605e-01 2.50900000e-01 8.76562148e-02
-9.95735884e-01 -4.17446405e-01 7.24449456e-01 -4.25043017e-01
1.02007902e+00 1.43225998e-01 2.68385857e-01 1.51960313e+00
-1.39122128e-01 8.65143120e-01 1.10917342e+00 -3.42213035e-01
7.04116225e-01 5.96015632e-01 1.61544502e-01 2.94544101e-01
1.09108724e-01 -4.28807847e-02 -4.53046083e-01 -5.20768523e-01
1.10502839e-01 -3.30020301e-02 -2.41791964e-01 -4.45433170e-01
-8.90577495e-01 8.61966193e-01 -1.76767722e-01 1.86840519e-01
-1.66783035e-01 -6.87981725e-01 4.98058230e-01 6.93837047e-01
5.53080082e-01 1.93986967e-01 -5.53417385e-01 -6.87935591e-01
-2.29909688e-01 3.79256576e-01 1.65927720e+00 1.08212280e+00
5.72170675e-01 -5.15363097e-01 9.90630090e-02 1.29783976e+00
2.19935223e-01 2.46419325e-01 5.91764808e-01 -8.65931392e-01
8.26475084e-01 7.05763042e-01 5.46889424e-01 -9.07794714e-01
-4.34219569e-01 7.83424173e-03 -5.05449116e-01 1.74942449e-01
7.34111726e-01 -3.70259583e-01 -4.92506623e-01 1.74780464e+00
8.30826402e-01 -1.31262928e-01 5.58878481e-01 6.82797074e-01
8.36410701e-01 5.86855888e-01 -3.52068603e-01 -4.14143622e-01
1.12350273e+00 -9.16113257e-01 -9.58727300e-01 -2.14996785e-01
7.69538403e-01 -7.76189625e-01 1.23144770e+00 5.02966523e-01
-5.12731612e-01 -1.20098293e-01 -1.10739911e+00 2.06244081e-01
-5.45397103e-01 -6.15578055e-01 5.79697847e-01 8.25596094e-01
-6.82531834e-01 7.54698813e-02 -1.73396662e-01 -7.92746723e-01
5.08637130e-02 9.20031816e-02 -3.78583491e-01 -1.78358078e-01
-1.23479843e+00 9.39699113e-01 7.81294778e-02 -2.10501537e-01
-9.24386024e-01 -1.94387048e-01 -7.36105084e-01 -3.92662585e-01
4.24691051e-01 -2.03702822e-01 1.75775337e+00 -5.94112217e-01
-1.88395405e+00 8.28725696e-01 -2.04458386e-01 -1.64438143e-01
9.65365589e-01 -6.61142636e-03 -4.81224328e-01 -4.24246602e-02
1.59657806e-01 1.50762647e-01 7.52110720e-01 -1.26090181e+00
-8.81700039e-01 -3.03325772e-01 1.74879923e-01 5.24491608e-01
-3.58153254e-01 -1.03675276e-01 -2.60380149e-01 -1.91782206e-01
8.93348753e-02 -1.01148891e+00 5.89198470e-02 -3.01877797e-01
-3.32570314e-01 -8.25398445e-01 1.17785966e+00 -1.86684340e-01
1.09012508e+00 -2.20084405e+00 -2.59854078e-01 2.47207079e-02
3.17278057e-01 1.54298455e-01 6.57128245e-02 6.51621282e-01
3.73225778e-01 -7.39428401e-02 -7.32597057e-03 -3.36215675e-01
2.99223661e-01 3.39143634e-01 5.33083975e-02 3.69356900e-01
-1.71996042e-01 -5.72013259e-02 -1.08351946e+00 -5.91251433e-01
2.27924794e-01 -3.44233513e-02 -4.70898986e-01 9.19090092e-01
-2.72386581e-01 4.39416587e-01 -6.37314856e-01 6.19784355e-01
7.37761259e-01 6.76914901e-02 2.85164624e-01 4.67458963e-01
-1.14548944e-01 6.06306493e-01 -1.28322172e+00 1.35438144e+00
-6.67382240e-01 7.09357262e-01 2.31178805e-01 -9.72222626e-01
1.12700820e+00 6.37789845e-01 4.33625787e-01 -4.28801745e-01
3.56138200e-01 2.96356440e-01 1.17417201e-01 -7.50807941e-01
2.77990729e-01 2.38369539e-01 -2.63207406e-01 6.47096694e-01
-1.98854189e-02 -2.05725536e-01 3.93029809e-01 5.53525938e-03
1.29079616e+00 -3.87835354e-01 5.62317252e-01 -1.00055031e-01
4.44808424e-01 -2.59453714e-01 6.43962741e-01 9.47305977e-01
-9.79896724e-01 3.65106374e-01 7.07041979e-01 -3.52943152e-01
-6.78153396e-01 -9.47399199e-01 -3.26844126e-01 1.18429959e+00
1.88798636e-01 -1.15531787e-01 -9.18915212e-01 -8.36362958e-01
-1.46036848e-01 7.25180328e-01 -3.95173520e-01 1.21228158e-01
-2.85018712e-01 -7.09343553e-01 3.53956103e-01 -1.51892126e-01
6.79009438e-01 -1.02505851e+00 -2.94371665e-01 3.39911163e-01
-4.14729208e-01 -1.36039627e+00 -6.10397100e-01 2.54908025e-01
-5.05570769e-01 -1.26827836e+00 -4.94656861e-01 -1.02537775e+00
4.56591129e-01 6.41511977e-01 1.15339053e+00 -3.15503657e-01
-5.12186177e-02 2.96788186e-01 -6.10427797e-01 -5.78454792e-01
-9.55268085e-01 1.13977648e-01 3.69213194e-01 8.38941559e-02
1.02596879e+00 -5.99300206e-01 -3.41529757e-01 7.80878007e-01
-4.73038584e-01 -3.82484198e-01 2.78651834e-01 7.93505311e-01
-1.71163827e-01 2.89088666e-01 9.00516450e-01 -9.59243834e-01
1.08345580e+00 -7.27132499e-01 -2.98174560e-01 -2.42096987e-02
-4.10169363e-01 -2.73322407e-02 6.11823618e-01 -7.19414711e-01
-1.30988908e+00 -3.41608524e-01 4.74611893e-02 1.49431735e-01
-7.37742066e-01 -4.36100177e-02 -5.15534520e-01 3.15720737e-01
6.79565728e-01 -6.67218119e-02 1.82392895e-01 -4.26242560e-01
6.10270873e-02 1.82938862e+00 3.86701636e-02 -6.36890650e-01
4.57984298e-01 3.37595731e-01 -7.13373244e-01 -1.20793211e+00
-6.68933332e-01 -6.84923470e-01 -3.45862836e-01 -4.13159430e-01
5.43835282e-01 -7.23451138e-01 -7.55767226e-01 7.70299554e-01
-1.04171813e+00 -3.26573133e-01 1.49934024e-01 4.52966660e-01
-4.97485727e-01 4.44155633e-01 -4.08783853e-01 -1.08864295e+00
-4.20471246e-04 -1.16678917e+00 6.77467763e-01 4.25424933e-01
-5.96936703e-01 -9.09954488e-01 3.34280908e-01 6.15198255e-01
1.56874046e-01 1.35336071e-01 5.06774962e-01 -1.35371006e+00
-1.37372315e-01 -4.42845434e-01 -1.57274341e-03 2.40471706e-01
6.94517553e-01 -1.82842895e-01 -1.16849875e+00 -3.19291145e-01
1.57773018e-01 -5.62855244e-01 1.68439239e-01 -1.85760036e-02
7.00074673e-01 -2.39384413e-01 -2.47234702e-01 -1.81454029e-02
6.85033679e-01 6.52124584e-01 2.22798854e-01 4.03434902e-01
9.83533859e-02 9.70503688e-01 8.91993403e-01 8.77383888e-01
5.52831829e-01 6.15564764e-01 2.36543417e-01 2.10722476e-01
2.19155163e-01 -1.34685904e-01 4.92357373e-01 6.46879196e-01
5.57313323e-01 -7.29167879e-01 -6.54622734e-01 6.88858211e-01
-1.99278808e+00 -9.49487984e-01 2.16477603e-01 2.20008850e+00
1.02310538e+00 3.79486084e-01 2.87839234e-01 7.40523115e-02
9.61325049e-01 2.78963447e-01 -7.60282993e-01 -7.05201447e-01
5.49433753e-02 -5.47850251e-01 2.32451502e-02 4.61846948e-01
-1.08277416e+00 5.60912609e-01 6.36446810e+00 4.48830366e-01
-6.03028297e-01 1.96911708e-01 2.99274296e-01 1.64117441e-02
1.08821325e-01 -3.97128046e-01 -8.85857224e-01 6.76258624e-01
7.78440654e-01 -1.30187944e-01 3.12222481e-01 1.23017740e+00
3.75993133e-01 -3.24092537e-01 -1.33469999e+00 1.01862955e+00
1.92574129e-01 -6.33175433e-01 -4.62995052e-01 1.26783952e-01
7.48199761e-01 2.70669013e-02 -3.48209113e-01 5.93792081e-01
6.97848737e-01 -5.86058199e-01 2.17688099e-01 -2.98415916e-03
3.16050857e-01 -6.24189973e-01 9.79518056e-01 7.64817595e-01
-8.34665418e-01 -4.59937453e-02 -2.12522000e-01 -3.55997980e-01
-1.60681248e-01 5.65749168e-01 -1.32011044e+00 5.89586515e-03
7.39795148e-01 5.16023993e-01 6.92817792e-02 8.96234095e-01
2.71684304e-02 3.33445251e-01 -4.70804513e-01 -5.43715417e-01
1.90673172e-01 -2.75352120e-01 6.44166648e-01 1.06787562e+00
2.63757166e-02 1.34295290e-02 4.48972672e-01 2.85259217e-01
3.20320986e-02 6.85540289e-02 -1.09786701e+00 9.22689214e-02
9.13557470e-01 8.59143734e-01 -2.97172844e-01 -2.85675526e-01
-6.79486930e-01 8.39732945e-01 3.97859544e-01 2.32167885e-01
-6.56629443e-01 -5.71391582e-01 1.23548698e+00 -3.92849743e-01
3.33801545e-02 -7.64360353e-02 1.59958169e-01 -1.07005250e+00
1.89450011e-01 -1.21579182e+00 5.78508377e-01 -1.61165088e-01
-1.75832880e+00 6.40377223e-01 -1.23716451e-01 -1.61929536e+00
-5.93423784e-01 -4.33342129e-01 -6.07048571e-01 3.65177393e-01
-1.24706709e+00 -4.04549658e-01 -4.60003614e-01 4.85638380e-01
1.04758143e+00 -3.90834928e-01 9.28779006e-01 2.64100105e-01
-6.92714930e-01 6.08048081e-01 5.04628658e-01 2.01239184e-01
1.02072060e+00 -1.26521051e+00 5.69621995e-02 4.72987592e-01
-2.41789714e-01 3.47206831e-01 1.03267789e+00 -3.69926691e-01
-1.05280650e+00 -8.07665765e-01 8.50048900e-01 -5.06505132e-01
5.51862240e-01 -7.05257177e-01 -7.72941053e-01 4.63834256e-01
2.31440440e-01 -2.75314540e-01 9.51139867e-01 2.67093956e-01
-2.87455022e-01 -5.38995527e-02 -1.46797299e+00 6.98251843e-01
1.04097700e+00 -4.57403392e-01 -9.71956134e-01 6.43002748e-01
5.58441103e-01 -4.93656248e-01 -6.29867256e-01 -1.98694207e-02
4.16750699e-01 -1.18028069e+00 5.57449400e-01 -6.13304853e-01
-2.98557997e-01 -1.91649333e-01 -3.35348487e-01 -1.57086301e+00
2.25367561e-01 -1.09055245e+00 -2.27507904e-01 1.30552518e+00
2.94238627e-01 -1.03986967e+00 7.15384424e-01 7.23579109e-01
1.43116862e-02 -3.13536465e-01 -1.05233777e+00 -8.71972919e-01
-2.47605398e-01 -3.26754510e-01 5.31092763e-01 8.46027792e-01
6.13941908e-01 4.09771442e-01 -4.71537620e-01 3.01870644e-01
7.28362501e-01 1.64170653e-01 1.20779562e+00 -1.36037195e+00
-4.30517457e-02 -1.27661332e-01 -3.02678108e-01 -1.40162373e+00
3.22015226e-01 -1.85477555e-01 3.77683342e-01 -1.34239042e+00
8.64320323e-02 -5.42437553e-01 -3.77763696e-02 -4.69779186e-02
-8.93671066e-02 -1.12939104e-01 -2.03927621e-01 1.88807532e-01
-7.21244991e-01 6.23399317e-01 1.09268880e+00 -8.16365853e-02
-6.12801671e-01 4.19735819e-01 -3.82387906e-01 8.98309231e-01
9.76447880e-01 -5.70169508e-01 -5.72014034e-01 9.14168134e-02
-4.26122934e-01 -3.46522443e-02 -3.31046462e-01 -8.32975686e-01
4.62511703e-02 -4.45082098e-01 -2.03881770e-01 -6.16110325e-01
4.37195301e-01 -7.44307995e-01 -4.33694422e-01 2.70038962e-01
-3.16827774e-01 -3.07499409e-01 -3.03264737e-01 7.83379674e-01
-2.30050400e-01 -3.01173061e-01 7.80404568e-01 -1.87881082e-01
-6.98077679e-01 1.17633440e-01 -8.74511123e-01 5.03087521e-01
1.19278240e+00 -3.34300071e-01 -4.54748124e-01 -6.69123590e-01
-4.97872919e-01 5.04562676e-01 3.30332220e-01 7.72428989e-01
3.26654792e-01 -9.58727300e-01 -5.47908902e-01 2.19052613e-01
5.03853440e-01 -4.01662011e-03 6.20381087e-02 4.18465972e-01
-3.38378191e-01 3.51131082e-01 3.72538865e-02 -5.73520243e-01
-1.36195111e+00 3.53803903e-01 3.44148993e-01 -1.80929735e-01
-4.28468943e-01 6.12071276e-01 2.69900620e-01 -1.11363661e+00
8.65133345e-01 -1.72153428e-01 -2.55970985e-01 3.57398689e-01
6.92046344e-01 5.02297759e-01 1.32151410e-01 -3.46217334e-01
-3.85616660e-01 -2.08025426e-02 -2.57904351e-01 -3.58111747e-02
1.00344050e+00 -4.46211368e-01 2.10839063e-01 7.03075111e-01
1.29966092e+00 1.67951286e-02 -1.31119776e+00 -4.50509727e-01
-3.12044029e-03 -6.78352833e-01 -3.30880761e-01 -5.63237429e-01
-3.17961991e-01 5.20842075e-01 5.60957015e-01 8.43967140e-01
6.80153072e-01 5.12685776e-02 7.42162108e-01 6.08864605e-01
5.64293623e-01 -1.62647080e+00 4.25994754e-01 7.19766140e-01
7.41539598e-01 -1.65714562e+00 -3.90346617e-01 -3.69954556e-01
-6.84266627e-01 8.84912133e-01 9.81647074e-01 2.39566877e-01
6.60538971e-01 2.88093358e-01 5.05587161e-01 1.02980666e-01
-1.06607676e+00 -1.68053135e-01 -4.17685360e-01 1.05314445e+00
1.85628533e-01 1.22268774e-01 -8.89979005e-02 2.48922855e-01
-2.84617692e-01 -3.83620799e-01 7.98008204e-01 1.09891880e+00
-6.50781512e-01 -1.10959303e+00 -4.18831766e-01 2.90243357e-01
2.07891455e-03 2.97723562e-01 -6.37694478e-01 7.94995666e-01
-2.06726924e-01 1.60491216e+00 8.17265213e-02 -2.82665223e-01
4.70152259e-01 2.32252270e-01 -6.82335049e-02 -6.28050923e-01
-2.36427829e-01 -2.82297850e-01 5.80125391e-01 -4.26687866e-01
-4.07216668e-01 -7.94996917e-01 -1.03094399e+00 -4.67298299e-01
-4.54040527e-01 4.58126038e-01 4.27974105e-01 1.01101410e+00
3.94462973e-01 1.50936946e-01 1.33110130e+00 -5.80067396e-01
-8.56255174e-01 -1.26511288e+00 -6.77370429e-01 6.93533003e-01
5.05387962e-01 -9.52231586e-01 -8.46601844e-01 -3.57331336e-01] | [12.529240608215332, 7.613171100616455] |
09c004d1-5869-47c3-9734-b9fb8f479d20 | learning-to-decouple-relations-few-shot | 2010.10894 | null | https://arxiv.org/abs/2010.10894v1 | https://arxiv.org/pdf/2010.10894v1.pdf | Learning to Decouple Relations: Few-Shot Relation Classification with Entity-Guided Attention and Confusion-Aware Training | This paper aims to enhance the few-shot relation classification especially for sentences that jointly describe multiple relations. Due to the fact that some relations usually keep high co-occurrence in the same context, previous few-shot relation classifiers struggle to distinguish them with few annotated instances. To alleviate the above relation confusion problem, we propose CTEG, a model equipped with two mechanisms to learn to decouple these easily-confused relations. On the one hand, an Entity-Guided Attention (EGA) mechanism, which leverages the syntactic relations and relative positions between each word and the specified entity pair, is introduced to guide the attention to filter out information causing confusion. On the other hand, a Confusion-Aware Training (CAT) method is proposed to explicitly learn to distinguish relations by playing a pushing-away game between classifying a sentence into a true relation and its confusing relation. Extensive experiments are conducted on the FewRel dataset, and the results show that our proposed model achieves comparable and even much better results to strong baselines in terms of accuracy. Furthermore, the ablation test and case study verify the effectiveness of our proposed EGA and CAT, especially in addressing the relation confusion problem. | ['Tiejun Zhao', 'BoWen Zhou', 'Xiaodong He', 'Youzheng Wu', 'Guangyi Liu', 'Junwei Bao', 'Yingyao Wang'] | 2020-10-21 | null | https://aclanthology.org/2020.coling-main.510 | https://aclanthology.org/2020.coling-main.510.pdf | coling-2020-8 | ['few-shot-relation-classification', 'few-shot-relation-classification'] | ['methodology', 'natural-language-processing'] | [ 1.41590178e-01 5.51024735e-01 -2.92017758e-01 -5.33939123e-01
-8.23263466e-01 -2.77408540e-01 5.63798308e-01 4.41670507e-01
-1.55837074e-01 8.05861115e-01 2.60212988e-01 -4.66023862e-01
-2.13816270e-01 -7.74391830e-01 -3.72562349e-01 -6.94272339e-01
2.08347267e-03 5.89839339e-01 1.61363721e-01 -4.85066712e-01
-2.32196569e-01 2.89352722e-02 -1.21705449e+00 3.26311171e-01
1.24565578e+00 1.01260281e+00 8.00910890e-02 3.27307671e-01
-1.48991406e-01 1.24350846e+00 -6.37756586e-01 -8.64263237e-01
-6.34260476e-02 -5.52856565e-01 -1.33407569e+00 3.02538741e-03
1.36407033e-01 -1.67480856e-02 -4.06146795e-01 1.05772412e+00
4.39011037e-01 3.53040040e-01 6.41848564e-01 -1.11585271e+00
-8.19926143e-01 1.01800978e+00 -6.03168488e-01 5.08834183e-01
4.77663189e-01 -1.44998983e-01 1.55861962e+00 -7.60654747e-01
6.55916274e-01 1.18865573e+00 4.78384823e-01 3.71901035e-01
-1.13171721e+00 -5.81741214e-01 3.65400344e-01 5.56606412e-01
-1.49039936e+00 -4.63754028e-01 6.01528645e-01 -3.33789438e-01
1.49106133e+00 3.89324456e-01 1.81765139e-01 1.16762900e+00
1.58914208e-01 6.28936112e-01 5.59295654e-01 -4.92563218e-01
1.73108242e-02 2.92680189e-02 8.49064410e-01 3.22431117e-01
2.49773815e-01 -3.00084651e-01 -3.12306643e-01 2.46814024e-02
2.44411658e-02 -2.68864512e-01 -4.70819443e-01 -1.59070119e-01
-1.00138736e+00 5.82057238e-01 6.31481111e-01 5.53451955e-01
-2.26445496e-01 -3.84232402e-01 5.25202155e-01 2.83268303e-01
7.61274040e-01 7.41319954e-01 -4.97495919e-01 -1.03099257e-01
-2.64102191e-01 4.64563780e-02 8.95621717e-01 1.22424877e+00
6.95498049e-01 -6.90370321e-01 -8.10170949e-01 1.08888364e+00
-8.19835439e-02 -2.20924959e-01 3.46689582e-01 -3.75190616e-01
1.02922833e+00 9.73912895e-01 -8.35602507e-02 -1.00842083e+00
-4.00997877e-01 -5.52753448e-01 -9.45169985e-01 -4.60775793e-01
1.13279171e-01 -2.09422037e-01 -7.08866060e-01 1.66512334e+00
3.19597334e-01 2.90750742e-01 4.39621031e-01 7.90356934e-01
1.39312100e+00 4.06271726e-01 4.01695192e-01 -4.22081232e-01
1.74016535e+00 -1.22606778e+00 -1.14719725e+00 -5.44352889e-01
1.16586256e+00 -4.49240357e-01 9.39662635e-01 -3.78839642e-01
-7.77797103e-01 -4.53323662e-01 -1.17690122e+00 -3.02560836e-01
-3.91087323e-01 -3.47698517e-02 8.62371564e-01 1.92972004e-01
-4.14545298e-01 5.44833422e-01 -5.23746192e-01 -5.24850607e-01
5.09383678e-01 8.26042965e-02 -3.72888654e-01 5.35555184e-02
-1.91390693e+00 1.24052536e+00 6.82502687e-01 2.33214438e-01
-1.38045684e-01 -5.87948382e-01 -1.33049786e+00 4.89479899e-01
8.69233489e-01 -4.30092394e-01 1.08881509e+00 -3.18961173e-01
-9.37307000e-01 1.00589073e+00 -3.22924733e-01 -4.96254683e-01
7.76126012e-02 -4.36949521e-01 -4.42784905e-01 -1.13469593e-01
3.38616788e-01 1.79180831e-01 7.63045996e-02 -1.08901393e+00
-5.20072758e-01 -7.33754113e-02 3.86684537e-01 3.69235933e-01
-2.32426152e-01 1.35612085e-01 -4.31265414e-01 -5.67203820e-01
1.79333121e-01 -6.35012865e-01 3.46927755e-02 -6.86033189e-01
-8.89122307e-01 -6.97756708e-01 7.16154814e-01 -3.96556646e-01
1.45593524e+00 -2.10669541e+00 2.25465477e-01 -2.20111579e-01
2.65943736e-01 5.10337830e-01 -1.03093505e-01 3.99240851e-01
-5.80929101e-01 1.49198398e-01 -2.41584167e-01 -4.15211767e-01
-3.86786878e-01 4.72979277e-01 -3.21588099e-01 5.53780757e-02
7.17624009e-01 1.21865487e+00 -1.25991499e+00 -5.71410656e-01
-1.30690545e-01 2.20052421e-01 1.68737881e-02 3.99158537e-01
6.55378476e-02 3.74451905e-01 -3.09325784e-01 3.94336402e-01
5.92649281e-01 -5.39510131e-01 5.01238108e-01 -3.64399880e-01
5.59480011e-01 8.50426912e-01 -8.47236097e-01 1.44960487e+00
-4.14157987e-01 3.61111909e-01 -3.28522146e-01 -1.10945594e+00
9.57025051e-01 6.55443251e-01 8.51288438e-02 -3.94239843e-01
4.24567282e-01 -2.01956794e-01 5.02861679e-01 -7.24858105e-01
4.24373239e-01 1.75320450e-02 -5.31025082e-02 3.36991757e-01
2.74074197e-01 2.08812073e-01 4.17791843e-01 4.89154309e-01
1.43554735e+00 4.32302617e-02 7.85163462e-01 -3.58758755e-02
7.25296199e-01 -3.03676605e-01 8.50501418e-01 6.40738130e-01
-3.27447861e-01 4.91094738e-01 8.65756452e-01 -3.53141725e-01
-4.17127371e-01 -6.72443986e-01 -7.99046531e-02 9.51798081e-01
5.82205057e-01 -7.86011398e-01 -3.91926318e-01 -1.14643741e+00
-3.26321930e-01 8.94638836e-01 -8.56706202e-01 -5.79892457e-01
-5.05699456e-01 -9.21249151e-01 2.37531960e-01 6.42740548e-01
6.57800257e-01 -1.04057813e+00 -3.00399721e-01 1.41146317e-01
-6.57094777e-01 -1.54956043e+00 -3.44012916e-01 6.48006141e-01
-3.50370377e-01 -1.20555842e+00 -1.96999341e-01 -9.83289659e-01
5.65373242e-01 4.54585284e-01 1.27141201e+00 1.67803809e-01
2.93610822e-02 -3.77319753e-01 -7.74884522e-01 -1.01595350e-01
-2.64220059e-01 3.47822428e-01 -2.63167500e-01 -4.41963635e-02
8.53004932e-01 -5.47924638e-01 -6.06749617e-02 1.77018180e-01
-3.74924093e-01 7.67260715e-02 4.70846623e-01 1.14854336e+00
1.25279829e-01 2.58596271e-01 5.68446636e-01 -1.38245153e+00
7.75598288e-01 -5.93633592e-01 7.59238079e-02 7.04336405e-01
-4.97199565e-01 1.49363264e-01 3.41112852e-01 -5.00657022e-01
-1.17814600e+00 -2.02048719e-01 8.35994538e-03 -2.14902878e-01
-1.04832515e-01 6.58592343e-01 -6.03997946e-01 4.55972135e-01
5.90116084e-01 -1.97741911e-01 -5.00400960e-01 -2.46230543e-01
4.21972811e-01 6.85720146e-01 5.65410316e-01 -3.89104098e-01
7.20457435e-01 -3.29817459e-02 -2.13026688e-01 -3.71375829e-01
-1.67273653e+00 -6.42893136e-01 -8.17828953e-01 2.52688408e-01
1.05961645e+00 -8.11483264e-01 -4.88755465e-01 1.90894917e-01
-1.46790874e+00 -9.28985327e-02 -2.09838107e-01 2.78309375e-01
-1.15775645e-01 2.27442011e-01 -9.51561511e-01 -8.57429087e-01
-4.45584655e-01 -8.90219986e-01 8.98848772e-01 3.47713590e-01
-5.30271888e-01 -8.78692091e-01 -5.53447418e-02 4.83689129e-01
-5.94534874e-02 -1.54177053e-02 1.12353599e+00 -1.20274770e+00
-1.94877788e-01 -2.71999866e-01 -5.32374442e-01 3.64314243e-02
4.85588163e-01 -2.81904578e-01 -1.12428999e+00 7.83772841e-02
-2.07313877e-02 -4.14521128e-01 8.88308585e-01 -2.16651276e-01
7.40030706e-01 -1.04860514e-01 -7.18949914e-01 2.56276011e-01
7.68023372e-01 3.42848718e-01 7.00351119e-01 2.06874236e-01
8.91831338e-01 7.28714883e-01 9.54954982e-01 -8.09345543e-02
5.38051724e-01 8.68041098e-01 1.53636411e-01 -5.10678452e-04
-2.35988095e-01 -1.30495280e-01 -1.91172868e-01 7.08157003e-01
-8.53957757e-02 -4.09252882e-01 -8.40979517e-01 4.14338052e-01
-2.07190156e+00 -9.71960664e-01 -1.08652130e-01 1.90035820e+00
1.21005309e+00 4.89194214e-01 -3.35891634e-01 3.01715016e-01
1.00674796e+00 2.14594796e-01 -2.45230094e-01 -2.57181764e-01
-2.53518730e-01 1.49842083e-01 -7.49317855e-02 6.18456364e-01
-1.45942783e+00 1.21189260e+00 5.10036707e+00 8.12677622e-01
-7.63770223e-01 1.35118842e-01 8.26367497e-01 2.66629428e-01
1.74624044e-02 6.90643191e-02 -8.19065034e-01 1.47837624e-01
6.32780433e-01 -1.79451585e-01 -6.78571686e-02 7.06224263e-01
-3.79778892e-01 -1.44696832e-01 -1.28608298e+00 7.82035172e-01
1.17369302e-01 -1.05183613e+00 -1.37041479e-01 -2.05017775e-01
4.04492497e-01 -4.85091686e-01 -4.13297147e-01 8.01854253e-01
3.13767374e-01 -8.86227310e-01 2.88489997e-01 1.09205052e-01
6.43302798e-01 -6.35886490e-01 1.30719733e+00 2.36230969e-01
-1.33440638e+00 1.84254497e-01 -2.75817424e-01 -4.71774340e-01
1.85526654e-01 5.60327232e-01 -7.69895494e-01 9.60737944e-01
4.49668586e-01 8.18641663e-01 -5.60697377e-01 6.89290524e-01
-8.53776693e-01 4.24795091e-01 2.92712092e-01 6.70932746e-03
-3.74832153e-02 4.87040095e-02 5.23100555e-01 1.15560400e+00
-1.11193903e-01 5.22483110e-01 -3.12766992e-03 7.84207344e-01
-1.37851864e-01 1.02802455e-01 -4.43878233e-01 1.29648179e-01
6.10568225e-01 1.51699781e+00 -6.78117275e-01 -4.90371823e-01
-5.40629447e-01 1.06241643e+00 9.88406420e-01 3.41625810e-01
-8.35440755e-01 -7.31735468e-01 6.58824623e-01 -3.95150155e-01
3.44296336e-01 1.54214859e-01 -3.14808130e-01 -1.33943498e+00
1.13266602e-01 -4.28871483e-01 5.22979438e-01 -5.99336684e-01
-1.56533098e+00 1.08278477e+00 -6.52568638e-02 -1.02929819e+00
-1.66295961e-01 -2.59296328e-01 -8.31144571e-01 1.04954970e+00
-1.32840466e+00 -1.04267561e+00 -3.65729988e-01 3.61765362e-02
5.10023355e-01 7.12174326e-02 1.07854676e+00 3.56225491e-01
-1.11003983e+00 9.14017737e-01 -6.86771691e-01 5.88015676e-01
8.08248580e-01 -1.19662607e+00 5.23055673e-01 9.33701098e-01
2.53850609e-01 7.49436855e-01 6.80545330e-01 -7.34202325e-01
-7.64848351e-01 -1.05229998e+00 1.47733009e+00 -4.51106519e-01
8.00531507e-01 -5.42509973e-01 -1.39029753e+00 8.05165291e-01
2.93339461e-01 2.99860358e-01 8.60787332e-01 8.12799692e-01
-7.24803329e-01 -3.60908508e-02 -8.32068622e-01 5.64731777e-01
1.19986713e+00 -5.36370516e-01 -1.10418737e+00 3.62174660e-01
9.74251509e-01 -5.38702786e-01 -7.32498467e-01 7.00794935e-01
-1.23062441e-02 -6.56279683e-01 7.70541310e-01 -9.44588959e-01
7.98107564e-01 -5.02522774e-02 6.91341562e-03 -1.40945292e+00
-4.81999338e-01 -5.73090136e-01 -5.78679621e-01 1.81560206e+00
7.71752238e-01 -4.18762088e-01 4.06547099e-01 7.66028464e-01
-1.23277381e-02 -1.05576146e+00 -8.82414222e-01 -9.31375325e-01
-1.22678593e-01 -2.10054502e-01 4.92082357e-01 1.35211420e+00
6.74536765e-01 1.47343862e+00 -3.73757511e-01 2.23682240e-01
8.28148052e-02 3.11202288e-01 4.62714434e-01 -1.03298390e+00
-3.88561457e-01 -2.09609777e-01 -4.87179041e-01 -9.35951948e-01
4.90822047e-01 -9.52388525e-01 3.32161158e-01 -1.56378579e+00
5.31203628e-01 -5.60088992e-01 -3.74140680e-01 5.91166615e-01
-1.14048100e+00 6.00207187e-02 -5.17764352e-02 9.29242149e-02
-8.54789913e-01 6.55104816e-01 1.16004694e+00 -2.68840969e-01
-1.61154866e-01 6.45433553e-03 -1.01527262e+00 5.14313817e-01
4.53510970e-01 -3.42596918e-01 -5.26594996e-01 -2.22882241e-01
1.17535837e-01 -9.70196258e-03 -1.46037281e-01 -6.99399114e-01
1.94217876e-01 8.23462680e-02 -1.03034213e-01 -3.83492798e-01
3.34808141e-01 -5.63683569e-01 -3.20080906e-01 1.81435928e-01
-6.08008087e-01 -3.85220230e-01 -1.62606053e-02 4.87830877e-01
-3.23499322e-01 -1.83349863e-01 4.95164841e-01 1.99695364e-01
-6.42422795e-01 1.54508531e-01 3.00419107e-02 4.14054990e-01
1.22230053e+00 2.94116437e-01 -7.75578558e-01 -4.37526286e-01
-8.38632286e-01 4.87839192e-01 -1.88559383e-01 6.41205728e-01
2.82680005e-01 -1.35752070e+00 -6.80952191e-01 1.47524169e-02
6.69558287e-01 2.35541269e-01 1.70094028e-01 8.87615502e-01
4.33071256e-02 3.24760646e-01 3.39952439e-01 -2.29635134e-01
-1.49118555e+00 1.04776442e+00 3.70775938e-01 -5.93856692e-01
-7.73467481e-01 1.26864648e+00 2.90049404e-01 -1.82297021e-01
2.96009421e-01 -3.69198710e-01 -4.56462324e-01 2.23500863e-01
7.41015375e-01 -2.17644311e-02 2.24676684e-01 -7.01142251e-01
-6.47688925e-01 2.31572598e-01 -5.27882576e-01 4.04725879e-01
1.13607025e+00 -2.59160548e-01 -2.39207864e-01 5.20037651e-01
9.10250604e-01 -6.93031251e-02 -8.99377406e-01 -5.74988902e-01
4.11260635e-01 -3.25525314e-01 -3.09297204e-01 -8.59145999e-01
-8.80889654e-01 6.53136194e-01 -1.19072460e-01 5.37368596e-01
8.15230429e-01 5.67028105e-01 7.02594161e-01 3.48893374e-01
8.68308693e-02 -5.81635177e-01 -1.13384314e-01 8.81596208e-01
7.24021792e-01 -1.51494765e+00 -7.64966905e-02 -1.30955350e+00
-8.98973703e-01 7.56509662e-01 1.06015050e+00 1.44066229e-01
5.25824368e-01 4.71132636e-01 2.11019516e-01 -2.05479249e-01
-1.01510882e+00 -5.23508906e-01 3.82526666e-01 5.62633455e-01
7.96313345e-01 -1.60270873e-02 -6.03426456e-01 9.15992618e-01
-6.77020475e-02 -5.59235096e-01 2.35272914e-01 7.90181994e-01
-1.13288030e-01 -1.08292198e+00 1.54102072e-01 3.89670342e-01
-2.66595393e-01 -4.60991174e-01 -5.05222917e-01 5.08420408e-01
1.54985011e-01 1.35580003e+00 1.29961342e-01 -6.68428898e-01
4.33705449e-01 2.47732490e-01 3.11094701e-01 -1.12075853e+00
-6.30591810e-01 -2.19050482e-01 5.35171449e-01 -5.13235688e-01
-3.36946279e-01 -4.00721848e-01 -1.18889451e+00 8.36182460e-02
-8.01086664e-01 4.09664989e-01 -2.93947756e-01 1.44139302e+00
2.71316737e-01 1.12064338e+00 3.43871862e-01 -2.97824591e-01
-3.43057096e-01 -1.41665447e+00 -5.51227629e-01 7.30811417e-01
1.03240468e-01 -1.05237293e+00 -2.03654513e-01 -3.96782398e-01] | [9.332364082336426, 8.587111473083496] |
b248baff-fce1-4601-803f-053b54318d6d | rotating-features-for-object-discovery | 2306.00600 | null | https://arxiv.org/abs/2306.00600v1 | https://arxiv.org/pdf/2306.00600v1.pdf | Rotating Features for Object Discovery | The binding problem in human cognition, concerning how the brain represents and connects objects within a fixed network of neural connections, remains a subject of intense debate. Most machine learning efforts addressing this issue in an unsupervised setting have focused on slot-based methods, which may be limiting due to their discrete nature and difficulty to express uncertainty. Recently, the Complex AutoEncoder was proposed as an alternative that learns continuous and distributed object-centric representations. However, it is only applicable to simple toy data. In this paper, we present Rotating Features, a generalization of complex-valued features to higher dimensions, and a new evaluation procedure for extracting objects from distributed representations. Additionally, we show the applicability of our approach to pre-trained features. Together, these advancements enable us to scale distributed object-centric representations from simple toy to real-world data. We believe this work advances a new paradigm for addressing the binding problem in machine learning and has the potential to inspire further innovation in the field. | ['Max Welling', 'Francesco Locatello', 'Phillip Lippe', 'Sindy Löwe'] | 2023-06-01 | null | null | null | null | ['object-discovery'] | ['computer-vision'] | [-1.20439842e-01 2.28490368e-01 -2.02622071e-01 -6.04890883e-01
-4.38977420e-01 -4.79189515e-01 8.94524395e-01 2.66178578e-01
-3.61176223e-01 8.63090515e-01 3.87132838e-02 2.82413792e-02
-5.33041596e-01 -8.38930964e-01 -6.44064307e-01 -7.75131822e-01
-1.93985403e-01 7.71257460e-01 3.50906461e-01 -3.22035439e-02
2.08534732e-01 6.55806422e-01 -1.74769986e+00 1.48326740e-01
5.00971854e-01 1.22794449e+00 3.47972423e-01 1.93568081e-01
-2.04852104e-01 6.83203578e-01 -6.64752126e-01 -3.35049808e-01
1.84680402e-01 -2.88405269e-01 -9.59570706e-01 -1.38339356e-01
3.72730076e-01 -1.46423191e-01 -2.27037966e-01 9.90035117e-01
3.31489921e-01 3.52444798e-01 9.65231478e-01 -1.31332994e+00
-7.81420887e-01 5.37453711e-01 -9.86032113e-02 4.03069317e-01
-5.24932370e-02 -5.40675879e-01 1.24682105e+00 -8.00899267e-01
4.53106076e-01 1.29036069e+00 3.19985092e-01 6.31492734e-01
-1.28758740e+00 -2.43756071e-01 2.46147841e-01 4.56696898e-01
-1.33172548e+00 -2.92485803e-01 7.53022254e-01 -6.54952824e-01
1.21424246e+00 -5.32152168e-02 6.48050904e-01 1.03117168e+00
1.18698590e-01 9.72398698e-01 1.06040418e+00 -6.08482301e-01
5.94493628e-01 1.25466123e-01 2.01627463e-01 5.66680014e-01
3.53211313e-01 1.18326256e-02 -5.80487013e-01 -2.44922712e-01
9.14275050e-01 2.19964698e-01 -3.95171829e-02 -7.68089592e-01
-1.18473601e+00 1.20770586e+00 6.06769264e-01 7.28764176e-01
-4.04048681e-01 1.84164137e-01 1.55542105e-01 2.80200303e-01
4.50070202e-01 6.96754277e-01 -5.79285264e-01 -7.67258108e-02
-7.63166666e-01 3.06180596e-01 7.65852034e-01 8.46064448e-01
9.17708457e-01 -1.66696474e-01 5.06633334e-02 8.22148919e-01
2.89764136e-01 1.23282775e-01 6.77368641e-01 -1.04362202e+00
5.06156050e-02 4.47639376e-01 -1.04282320e-01 -1.02498078e+00
-4.61596757e-01 -6.11146986e-01 -5.85724115e-01 2.29297459e-01
4.12067860e-01 -3.85605209e-02 -5.34509182e-01 1.89012182e+00
1.22644424e-01 -1.76678732e-01 6.58533201e-02 1.06346607e+00
4.48889703e-01 4.35974538e-01 -7.96889812e-02 2.04374358e-01
1.19132602e+00 -9.49331224e-01 -4.91930723e-01 -1.02590732e-01
4.74871814e-01 -3.98899227e-01 5.80085874e-01 5.02018571e-01
-1.02555001e+00 -4.73890841e-01 -1.01094413e+00 -1.75160766e-01
-6.76753521e-01 7.17285797e-02 1.23883474e+00 5.43716311e-01
-1.04090130e+00 6.54110134e-01 -8.93577576e-01 -5.23869693e-01
7.35657334e-01 6.70069575e-01 -4.02418911e-01 -4.79487469e-03
-1.13324749e+00 1.07047176e+00 4.60246652e-01 -6.68507516e-02
-7.13686287e-01 -2.58125305e-01 -8.05319667e-01 3.28695804e-01
3.64560395e-01 -7.08954930e-01 1.25691867e+00 -8.57716978e-01
-1.66974819e+00 6.00496352e-01 -1.66364312e-02 -6.20141745e-01
-1.18609928e-01 -1.06513537e-01 -1.30383015e-01 1.30249843e-01
-1.41953863e-02 8.80880117e-01 8.44584942e-01 -1.05772412e+00
-5.38585305e-01 -4.88788515e-01 1.14021808e-01 4.87953871e-02
-5.62041342e-01 -1.32006913e-01 -5.90669252e-02 -6.29241884e-01
2.52503097e-01 -7.96906531e-01 -1.53713807e-01 1.06154978e-01
-3.74263413e-02 -6.94425344e-01 3.90236020e-01 6.27678335e-02
6.44361258e-01 -2.27104163e+00 5.10377824e-01 2.60270894e-01
3.66716117e-01 7.59944543e-02 -1.18132040e-01 3.12340915e-01
5.82692623e-02 -9.44851488e-02 -4.72850986e-02 -2.61769205e-01
1.92743376e-01 4.24498737e-01 -3.48294288e-01 5.44739962e-01
5.17912924e-01 9.32885766e-01 -9.59300816e-01 -4.53761458e-01
7.01029897e-02 3.44616532e-01 -6.66369498e-01 1.70267373e-01
-1.97416693e-01 1.29021972e-01 -7.09876716e-01 4.06193078e-01
4.48768079e-01 -3.15152973e-01 3.67492884e-02 1.33622661e-01
9.31789260e-03 3.16238910e-01 -9.80793536e-01 1.85599041e+00
-2.90006906e-01 7.92854309e-01 -1.53977409e-01 -1.52619874e+00
9.73493874e-01 3.06021094e-01 4.18002307e-01 -5.65135062e-01
2.83135563e-01 1.28459364e-01 3.53963822e-01 -2.40431100e-01
4.52866912e-01 -2.87847340e-01 -4.82480414e-02 5.54392934e-01
7.49302506e-01 -2.47362807e-01 3.11045703e-02 1.32129818e-01
1.10010457e+00 2.79590255e-03 2.96693236e-01 -3.71244520e-01
6.22471571e-02 -3.57271492e-01 2.76141077e-01 8.03790987e-01
-1.67287901e-01 6.69797897e-01 4.13965464e-01 -4.02220368e-01
-7.31766820e-01 -1.22002518e+00 -3.48639935e-01 1.13927305e+00
2.13566199e-02 -3.13671470e-01 -5.19465208e-01 -5.91344178e-01
2.29747966e-01 4.57471550e-01 -7.25636721e-01 -2.77039766e-01
-1.57365918e-01 -4.19063449e-01 2.61895210e-01 6.61814868e-01
1.75165161e-01 -1.07842505e+00 -7.15324402e-01 3.12726408e-01
-1.69892907e-02 -8.61555815e-01 9.11393240e-02 6.88431621e-01
-1.00339198e+00 -7.65595675e-01 -8.39722574e-01 -8.01873267e-01
6.44777060e-01 1.87153533e-01 1.11302495e+00 -8.64600539e-02
-3.24899375e-01 4.93899077e-01 -2.01917380e-01 -6.25757575e-01
2.74836600e-01 3.52920800e-01 1.17561877e-01 -8.86030793e-02
5.27129054e-01 -6.39620066e-01 -4.97931063e-01 2.24484116e-01
-9.33003783e-01 -4.02913809e-01 5.67376375e-01 9.69155729e-01
2.42061153e-01 8.79644975e-02 7.45744407e-01 -6.93620920e-01
7.56852090e-01 -7.51823127e-01 -4.93420690e-01 6.22528754e-02
-4.15066272e-01 4.45755929e-01 5.23982406e-01 -5.16221285e-01
-1.07667112e+00 3.74848023e-02 1.58602014e-01 -3.22999030e-01
-3.84284288e-01 6.48500681e-01 3.08788475e-02 6.55040964e-02
7.85053492e-01 -2.86252256e-02 -1.71118956e-02 -4.33716655e-01
5.43784201e-01 5.17545164e-01 3.06163281e-01 -1.02199697e+00
3.77695680e-01 4.90212411e-01 -5.41281179e-02 -5.99279106e-01
-9.35035408e-01 -5.08140802e-01 -6.49054348e-01 1.75114751e-01
7.08349407e-01 -7.60811627e-01 -6.23726904e-01 1.16051389e-02
-1.21862972e+00 -1.75068289e-01 -6.67394280e-01 8.17140043e-01
-9.20787573e-01 2.79147904e-02 -4.47760940e-01 -8.06092680e-01
2.49166951e-01 -7.75668800e-01 9.84397650e-01 2.88073927e-01
-3.19568634e-01 -9.87525463e-01 2.11872250e-01 6.55434153e-04
5.82668066e-01 -2.30138853e-01 1.08480871e+00 -9.67635512e-01
-5.96684396e-01 -2.57951617e-01 -1.10625543e-01 2.32358396e-01
1.33681893e-01 -2.39337876e-01 -1.04361928e+00 -1.80198133e-01
1.37532055e-01 -8.86093616e-01 1.10359228e+00 3.18468422e-01
1.30415034e+00 1.19562760e-01 -3.88113558e-01 2.04019740e-01
1.21765745e+00 7.81681091e-02 3.94751281e-01 3.54273289e-01
8.96170512e-02 7.88179338e-01 3.13441366e-01 5.72752833e-01
2.79618770e-01 6.18377686e-01 5.49926281e-01 1.79627493e-01
3.33735980e-02 -2.23085172e-02 -8.47567916e-02 8.48372281e-01
-2.98219740e-01 -2.01074779e-01 -7.25947738e-01 7.24245250e-01
-2.03810525e+00 -1.02722156e+00 5.42102039e-01 2.00383186e+00
6.55590236e-01 -1.62898526e-02 7.50842318e-02 2.98094135e-02
7.07362056e-01 -7.47427344e-02 -5.07249355e-01 -3.66332322e-01
-1.30837053e-01 4.32719588e-01 -1.21535644e-01 2.42565684e-02
-9.25308406e-01 9.85229194e-01 6.91584206e+00 7.53726780e-01
-1.10600936e+00 1.56290233e-01 3.49774927e-01 5.99309430e-02
-2.70282269e-01 -3.86507213e-02 -7.71072090e-01 9.37392339e-02
9.36400235e-01 -1.49111329e-02 5.25208771e-01 1.06207085e+00
-5.71087182e-01 -1.00797996e-01 -1.43086398e+00 9.71981645e-01
1.26989841e-01 -1.16083264e+00 -7.40772560e-02 2.06815392e-01
6.21316135e-01 1.42612979e-01 3.80445093e-01 4.81736422e-01
2.79929399e-01 -1.14218998e+00 6.46970689e-01 6.56727612e-01
2.06417739e-01 -6.43419981e-01 7.00091422e-01 3.45415831e-01
-7.56998897e-01 -2.19959795e-01 -8.35947037e-01 -3.47104490e-01
-1.73424721e-01 6.12724602e-01 -8.38143528e-01 2.68782765e-01
6.15850985e-01 6.08016968e-01 -5.33525765e-01 1.14564884e+00
-4.45208438e-02 4.63838965e-01 -5.45999646e-01 -3.32833052e-01
2.05695570e-01 1.30680233e-01 9.50597972e-02 9.53296721e-01
4.01139677e-01 1.16994649e-01 -1.36785181e-02 1.19406986e+00
-7.49632940e-02 -2.33202912e-02 -7.23942280e-01 -2.52304256e-01
4.16205794e-01 1.29141951e+00 -9.57164168e-01 -2.85585403e-01
-6.17595375e-01 6.69488966e-01 9.46958899e-01 2.19048887e-01
-4.75063413e-01 -3.02264005e-01 5.78234553e-01 -2.06660539e-01
7.15920031e-01 -4.64421183e-01 -2.78500896e-02 -1.30814910e+00
-1.18178993e-01 -5.77872992e-01 2.61434406e-01 -5.71878672e-01
-1.54683471e+00 6.94942772e-01 2.05916047e-01 -9.34248328e-01
-5.95249414e-01 -8.44250500e-01 -2.76819229e-01 5.86272359e-01
-1.44689012e+00 -6.99461699e-01 1.24835163e-01 8.13644111e-01
3.00336868e-01 -4.83779490e-01 1.25924516e+00 -5.96128255e-02
-1.72097981e-01 4.91972774e-01 3.87207955e-01 -7.24809617e-02
5.56056201e-01 -1.25404394e+00 4.43235971e-02 2.39155427e-01
7.35617399e-01 9.95488107e-01 4.89101648e-01 -1.79882735e-01
-1.15215349e+00 -6.41071916e-01 8.05900872e-01 -4.54940885e-01
6.60041094e-01 -5.17977476e-01 -8.19191277e-01 6.31959677e-01
2.56877035e-01 4.62598830e-01 9.15976405e-01 6.86720252e-01
-4.64147121e-01 1.58708040e-02 -1.14458323e+00 2.68893898e-01
8.74908149e-01 -6.05295122e-01 -1.03391731e+00 2.11079180e-01
4.25999105e-01 2.40445957e-02 -1.04610372e+00 7.38258883e-02
4.09738988e-01 -9.71166670e-01 7.43011951e-01 -6.81211710e-01
2.52091050e-01 -1.02749117e-01 -3.94410789e-01 -1.45762599e+00
-6.62968457e-01 -2.62678981e-01 -2.25383952e-01 1.05757558e+00
3.21835637e-01 -7.52528548e-01 7.29858696e-01 4.93698657e-01
-6.41394407e-02 -7.99089670e-01 -1.12821841e+00 -8.09528053e-01
2.96348929e-01 -3.63896012e-01 5.59324563e-01 8.13424349e-01
2.28495881e-01 4.89047140e-01 1.64100563e-03 -1.66384712e-01
4.06541258e-01 3.31420749e-01 4.46990728e-01 -1.70525122e+00
-3.96834791e-01 -3.87976229e-01 -9.32748854e-01 -1.16524041e+00
5.06966770e-01 -1.02937222e+00 5.15573658e-02 -1.26409161e+00
2.47685552e-01 -5.26875913e-01 -6.57156825e-01 5.74993551e-01
3.32203716e-01 5.69785982e-02 1.87943637e-01 2.53127187e-01
-7.80798793e-01 8.07883143e-01 1.06911290e+00 -1.49176672e-01
8.79154503e-02 -1.70102697e-02 -7.48333454e-01 7.10551143e-01
8.87660801e-01 -6.83593392e-01 -5.98926783e-01 -4.50142413e-01
2.34660447e-01 -3.68538558e-01 3.82248610e-01 -1.00602722e+00
4.29571688e-01 3.34651582e-02 6.77086771e-01 -2.39007816e-01
6.90336704e-01 -9.74515200e-01 -3.97647083e-01 4.79465425e-02
-4.20283616e-01 -1.28207326e-01 1.01072550e-01 7.02735007e-01
-4.31934625e-01 -4.47456747e-01 4.78726953e-01 -3.26653302e-01
-5.54281533e-01 1.94813803e-01 -4.74357367e-01 2.64773611e-02
9.69309866e-01 -8.18603784e-02 -2.82148451e-01 -3.34122360e-01
-8.47704470e-01 -1.82883546e-01 2.02501908e-01 3.58893275e-01
6.27185225e-01 -1.34413028e+00 -3.10636252e-01 2.42949963e-01
1.84549376e-01 2.22078487e-02 -1.15969427e-01 4.72934902e-01
1.00803562e-02 7.92541862e-01 -4.42335546e-01 -6.90965831e-01
-6.80817485e-01 7.45407164e-01 1.03968762e-01 9.50057209e-02
-4.27750856e-01 8.40859830e-01 1.94447562e-01 -3.25526953e-01
3.20296407e-01 -5.11689544e-01 -2.61517912e-01 1.73522830e-01
3.40880662e-01 6.85225502e-02 2.23295286e-01 -5.27905524e-01
-3.70650977e-01 3.44737828e-01 -3.55410933e-01 -3.14742744e-01
1.61139631e+00 2.82862075e-02 -2.37142101e-01 7.81102598e-01
1.15763760e+00 -3.15989524e-01 -1.18366647e+00 -4.11060721e-01
1.52680486e-01 -2.99308419e-01 3.90178114e-02 -5.48869431e-01
-9.36617732e-01 9.98419344e-01 4.98799831e-01 4.49702829e-01
7.79221952e-01 5.53466022e-01 2.37078130e-01 9.87143993e-01
6.06584549e-01 -1.09394515e+00 2.96572149e-01 6.34126365e-01
7.49011219e-01 -1.16157842e+00 -1.65174142e-01 -4.86812480e-02
-3.61652136e-01 1.36914706e+00 6.00111187e-01 -3.21932524e-01
9.63823020e-01 1.14286140e-01 -2.76191682e-01 -2.59015739e-01
-9.75457013e-01 -3.23146224e-01 2.70788372e-01 8.64892244e-01
6.04898393e-01 -5.04528470e-02 -3.77784036e-02 6.46904051e-01
-2.83681303e-01 8.81737191e-03 2.57474303e-01 1.04376662e+00
-6.22482896e-01 -1.23146844e+00 -1.64538786e-01 5.19941688e-01
-3.67573917e-01 -4.15616408e-02 -4.09857243e-01 7.55028844e-01
2.90361613e-01 8.24045956e-01 3.13502342e-01 -1.82894841e-02
1.49465539e-02 1.45655617e-01 9.45598364e-01 -9.23380136e-01
-2.39333749e-01 -3.46581340e-02 -3.28622282e-01 -3.71215671e-01
-6.75045609e-01 -6.30495906e-01 -1.31562054e+00 9.34962258e-02
-5.65928221e-01 4.89432782e-01 8.17754388e-01 1.15863156e+00
5.01335740e-01 4.58513647e-01 3.08766603e-01 -1.04080164e+00
-7.71010220e-01 -9.44104314e-01 -8.85753572e-01 1.31465971e-01
2.72663981e-01 -1.19841731e+00 -3.38544875e-01 -1.36421248e-01] | [9.418134689331055, 2.640986919403076] |
df3b6dea-1e88-4cd8-921c-e43aaa517dbe | u-tilise-a-sequence-to-sequence-model-for | 2305.13277 | null | https://arxiv.org/abs/2305.13277v1 | https://arxiv.org/pdf/2305.13277v1.pdf | U-TILISE: A Sequence-to-sequence Model for Cloud Removal in Optical Satellite Time Series | Satellite image time series in the optical and infrared spectrum suffer from frequent data gaps due to cloud cover, cloud shadows, and temporary sensor outages. It has been a long-standing problem of remote sensing research how to best reconstruct the missing pixel values and obtain complete, cloud-free image sequences. We approach that problem from the perspective of representation learning and develop U-TILISE, an efficient neural model that is able to implicitly capture spatio-temporal patterns of the spectral intensities, and that can therefore be trained to map a cloud-masked input sequence to a cloud-free output sequence. The model consists of a convolutional spatial encoder that maps each individual frame of the input sequence to a latent encoding; an attention-based temporal encoder that captures dependencies between those per-frame encodings and lets them exchange information along the time dimension; and a convolutional spatial decoder that decodes the latent embeddings back into multi-spectral images. We experimentally evaluate the proposed model on EarthNet2021, a dataset of Sentinel-2 time series acquired all over Europe, and demonstrate its superior ability to reconstruct the missing pixels. Compared to a standard interpolation baseline, it increases the PSNR by 1.8 dB at previously seen locations and by 1.3 dB at unseen locations. | ['Konrad Schindler', 'Vivien Sainte Fare Garnot', 'Corinne Stucker'] | 2023-05-22 | null | null | null | null | ['cloud-removal'] | ['computer-vision'] | [ 6.76075161e-01 -4.66669828e-01 1.43925056e-01 -5.27530611e-01
-8.29202592e-01 -5.61604321e-01 4.49774742e-01 -8.09873492e-02
-4.78888482e-01 7.53818274e-01 1.77255139e-01 -3.20462346e-01
4.62058485e-02 -9.05265510e-01 -8.92376184e-01 -1.05320215e+00
-5.25890172e-01 -1.53025985e-01 -3.23393606e-02 -2.38709571e-03
-1.45569563e-01 4.72782940e-01 -1.73179257e+00 4.83697146e-01
9.69312489e-01 1.15019834e+00 7.29649305e-01 8.07947516e-01
6.41059950e-02 8.46390784e-01 -2.58890510e-01 2.13424876e-01
3.70282263e-01 -3.11210513e-01 -5.30293167e-01 2.42387995e-01
3.44586343e-01 -6.42542481e-01 -4.32561487e-01 1.06773865e+00
2.87402242e-01 -8.33401382e-02 3.31780940e-01 -7.27444410e-01
-7.39494741e-01 8.62029195e-02 -5.29926360e-01 3.21364045e-01
-1.71641886e-01 1.49531811e-01 8.41103733e-01 -7.65982330e-01
4.62220550e-01 8.03014219e-01 7.76629686e-01 1.39002845e-01
-1.26734090e+00 -4.88368958e-01 -6.83128610e-02 2.94337928e-01
-1.46216762e+00 -3.16271663e-01 3.94472510e-01 -7.29014933e-01
9.15878475e-01 3.04074883e-01 7.29653955e-01 7.23685801e-01
2.38432482e-01 3.65864426e-01 1.06641579e+00 -2.40329981e-01
-1.97765324e-02 -3.30686122e-01 -3.35693896e-01 3.64664607e-02
-7.52868429e-02 3.76486182e-01 -2.73074001e-01 1.25601724e-01
6.65099502e-01 3.39720726e-01 -6.67903483e-01 1.23879381e-01
-1.25292218e+00 7.15480447e-01 8.68308425e-01 1.38082102e-01
-9.15527940e-01 3.64720732e-01 5.59531152e-02 3.61615121e-01
8.25938880e-01 2.96290498e-02 -4.71769989e-01 3.54646206e-01
-1.34051025e+00 2.84379780e-01 1.63311541e-01 6.79970026e-01
9.23156261e-01 3.91887009e-01 -3.62379886e-02 5.29808283e-01
9.71682221e-02 9.84449506e-01 2.86259949e-01 -8.33748341e-01
4.15984362e-01 7.37097263e-02 5.66075444e-01 -7.56516337e-01
-1.76861063e-01 -6.50039494e-01 -1.02925074e+00 2.42092177e-01
-4.82185856e-02 -1.78722903e-01 -9.33371067e-01 1.62349641e+00
7.38252550e-02 5.60000658e-01 1.90888703e-01 1.15116441e+00
3.02740514e-01 1.38787305e+00 3.68927941e-02 -3.23763311e-01
1.14249086e+00 -6.44842565e-01 -7.22552657e-01 -4.68835354e-01
2.33820423e-01 -6.74417078e-01 5.58337986e-01 -1.18024431e-01
-6.76543713e-01 -6.99958980e-01 -1.09179401e+00 7.56108910e-02
-3.92698586e-01 2.56485939e-01 2.50006825e-01 8.30845162e-03
-1.26359046e+00 6.68439031e-01 -7.55274355e-01 -6.46708533e-02
1.23916462e-01 -1.42776176e-01 -2.12011173e-01 -2.68552095e-01
-1.27327192e+00 7.59350061e-01 5.90374708e-01 6.09046996e-01
-9.34902310e-01 -8.44257951e-01 -7.16761947e-01 1.62697598e-01
-1.82767455e-02 -5.01710474e-01 9.34724867e-01 -1.50594866e+00
-8.34202707e-01 4.83192563e-01 -3.32651794e-01 -7.77634382e-01
2.46339917e-01 -7.70462379e-02 -8.34831357e-01 3.02789509e-01
2.40046963e-01 6.66008890e-01 1.20410895e+00 -1.11928642e+00
-7.04837859e-01 -2.45144829e-01 -2.96205133e-01 2.59628654e-01
-4.97415438e-02 -2.35541210e-01 -2.68333048e-01 -9.55723703e-01
2.16484457e-01 -9.27094102e-01 -2.08289832e-01 3.47722024e-01
1.43704459e-01 5.69958210e-01 9.05386925e-01 -1.27540016e+00
8.81600201e-01 -2.34926438e+00 1.33078501e-01 -1.07093766e-01
-1.98403642e-01 3.01424652e-01 -4.89899457e-01 4.92557079e-01
-3.86679083e-01 -9.53332856e-02 -7.70410240e-01 -7.29903951e-02
-5.97580731e-01 5.23863256e-01 -7.36552536e-01 6.71260715e-01
4.17014211e-01 7.41095901e-01 -8.86410832e-01 7.45212734e-02
2.85659611e-01 8.35666239e-01 -2.03201860e-01 3.28033328e-01
-4.75426674e-01 6.93639755e-01 -5.50897494e-02 4.17455167e-01
1.29950035e+00 -2.72736996e-01 1.99685916e-01 -9.14126262e-02
-5.88148534e-01 1.69043571e-01 -8.37422907e-01 1.63160121e+00
-5.32785654e-01 9.20634091e-01 9.85929519e-02 -8.78576875e-01
8.18099856e-01 3.71561795e-01 4.65759426e-01 -1.06822574e+00
-4.21032935e-01 3.51318359e-01 -3.37098420e-01 -6.65768623e-01
6.63095415e-01 -2.41583124e-01 2.01422870e-01 2.17410624e-01
-2.70731062e-01 -4.48242342e-03 -4.11468625e-01 -2.64295816e-01
6.62218809e-01 1.57494590e-01 -1.50558069e-01 -4.42234278e-02
6.05117917e-01 1.89275071e-02 4.89619732e-01 4.25029725e-01
1.46129355e-01 7.48350322e-01 -2.18186043e-02 -7.98897147e-01
-1.47890735e+00 -9.44061637e-01 -2.41167352e-01 7.11735249e-01
-1.09800942e-01 8.70899111e-02 -2.56785542e-01 -5.67896711e-03
9.74398032e-02 6.56750739e-01 -6.46368325e-01 2.14828947e-03
-4.00830775e-01 -8.32599282e-01 3.13245237e-01 3.34199667e-01
6.92796290e-01 -8.47783864e-01 -7.88957953e-01 4.57486212e-01
-6.77761734e-01 -1.28353906e+00 -1.34395331e-01 1.50402516e-01
-1.05237973e+00 -7.59225428e-01 -8.34086955e-01 -5.44332802e-01
4.60721880e-01 7.24179983e-01 1.00252020e+00 -1.17410928e-01
-2.16872618e-01 -1.41097397e-01 -5.67108929e-01 -2.22630396e-01
-2.87487775e-01 -3.42995852e-01 -3.93761039e-01 4.43032861e-01
9.09760967e-02 -6.49468482e-01 -7.26177335e-01 -1.22358613e-02
-1.37342680e+00 2.51694083e-01 4.51206714e-01 9.36140478e-01
7.14699626e-01 1.19076222e-01 1.63767621e-01 -4.67137545e-01
-9.94021073e-02 -7.15788484e-01 -9.87167895e-01 1.33401081e-01
-3.77810985e-01 4.55102958e-02 4.14413005e-01 1.03027239e-01
-9.49218750e-01 3.95856470e-01 -5.32196574e-02 -5.15744865e-01
1.37629136e-01 8.28056872e-01 1.93157762e-01 6.31627999e-03
5.43917775e-01 9.30245459e-01 -2.51096599e-02 -5.69879115e-01
3.72296065e-01 9.43160295e-01 1.05810046e+00 -3.44587080e-02
9.26638305e-01 8.92092884e-01 -2.38553017e-01 -1.22425306e+00
-7.92706847e-01 -5.48272729e-01 -5.95166326e-01 -2.95574814e-01
9.33840156e-01 -1.67349672e+00 -2.32923310e-02 6.09625518e-01
-1.33983171e+00 -3.70329678e-01 -2.29542390e-01 4.18702453e-01
-3.64605069e-01 3.92326266e-01 -2.50675231e-01 -8.04946423e-01
-4.66938227e-01 -7.04932034e-01 1.19965005e+00 -3.63956541e-02
5.48874676e-01 -6.86096907e-01 5.62921268e-05 3.29945050e-02
6.40128553e-01 6.14632905e-01 6.96317971e-01 4.63696659e-01
-1.01760793e+00 -1.79000810e-01 -5.35671473e-01 7.17616200e-01
1.72273651e-01 -1.41237140e-01 -1.10996878e+00 -6.28052592e-01
1.14971660e-01 1.03815019e-01 1.29055405e+00 5.10992825e-01
1.29573965e+00 -5.49398541e-01 1.30241774e-02 1.16283917e+00
2.01807499e+00 1.33400708e-01 1.00892389e+00 3.02899837e-01
4.30006951e-01 3.39092463e-01 3.49151671e-01 5.80436289e-01
3.60107660e-01 5.53899705e-01 1.04301322e+00 -1.53410241e-01
-1.11773290e-01 -3.81180048e-02 2.70806968e-01 5.43136716e-01
-2.67625391e-01 -2.50438899e-01 -7.70248950e-01 1.04977643e+00
-1.70108676e+00 -1.30489123e+00 -2.52774090e-01 2.30010557e+00
6.28484130e-01 -4.97479856e-01 -3.88343632e-01 1.23292752e-01
6.46332085e-01 7.24961519e-01 -5.65774381e-01 1.21588662e-01
-5.51749051e-01 2.39462361e-01 1.08792710e+00 5.92633843e-01
-1.21615839e+00 6.81015491e-01 5.77217817e+00 2.44882032e-01
-1.63152850e+00 3.49473238e-01 2.55939692e-01 3.04823723e-02
-4.47116524e-01 8.47565085e-02 -1.35878459e-01 6.38713658e-01
1.28642249e+00 8.54658615e-03 5.83870053e-01 3.15923780e-01
5.69717288e-01 -1.44302413e-01 -5.78263998e-01 9.28927839e-01
-1.83300465e-01 -1.62678432e+00 -2.26460639e-02 2.88902950e-02
8.98555458e-01 6.32845998e-01 1.06384195e-01 -2.08346099e-01
-2.53897756e-02 -1.11801434e+00 9.80683804e-01 8.66271138e-01
1.16471922e+00 -5.78074992e-01 7.80013919e-01 5.08772850e-01
-1.34564185e+00 -1.15076594e-01 -5.15395403e-01 -1.42420277e-01
7.53758773e-02 6.93017423e-01 -6.73232675e-01 9.45313752e-01
8.62037659e-01 1.11266887e+00 -2.13691264e-01 1.18754280e+00
-2.95872837e-01 5.30107975e-01 -2.19366863e-01 6.31124556e-01
5.04059732e-01 -2.54471719e-01 3.33845198e-01 1.09450042e+00
7.50133216e-01 2.76332170e-01 -1.40302613e-01 8.00357223e-01
1.92154497e-01 -3.27914625e-01 -6.59449518e-01 -9.13319960e-02
3.83456498e-01 7.70531178e-01 -5.16162924e-02 -2.01615497e-01
-5.08623838e-01 1.33393204e+00 -1.55359104e-01 6.36103868e-01
-8.90177548e-01 -1.40766501e-01 1.03301525e+00 1.60186857e-01
8.40583920e-01 -5.30514598e-01 -6.79904073e-02 -1.16368437e+00
2.58684218e-01 -6.07719839e-01 1.67153597e-01 -1.27444017e+00
-8.68589878e-01 7.67752945e-01 -3.95876884e-01 -1.73898530e+00
-2.58841276e-01 -2.04344973e-01 -2.31097013e-01 1.58429146e+00
-2.33011127e+00 -1.13732564e+00 -7.21559525e-01 6.04899228e-01
3.70212287e-01 2.39407510e-01 9.10298407e-01 4.53121573e-01
-6.07922040e-02 -1.46467432e-01 7.42427051e-01 2.85560731e-02
3.01412910e-01 -8.16092253e-01 6.54568911e-01 1.28588581e+00
2.42638998e-02 2.33543608e-02 7.05816925e-01 -4.26725030e-01
-1.29449463e+00 -1.82614648e+00 1.16868842e+00 1.69339910e-01
4.52401757e-01 -1.09388225e-01 -1.23401427e+00 7.77699649e-01
1.39024034e-01 5.78481138e-01 2.83210814e-01 -8.00059617e-01
-5.52939713e-01 -4.61510837e-01 -8.89933348e-01 -6.45489758e-03
5.49486816e-01 -9.95571196e-01 -2.77418643e-01 5.70354939e-01
8.04505587e-01 -4.62464929e-01 -5.82059741e-01 2.81112731e-01
5.02394676e-01 -9.99068081e-01 1.10098684e+00 -2.90356785e-01
6.13506436e-01 -6.78202748e-01 -5.85042655e-01 -1.32843256e+00
-4.33136970e-01 -1.69197753e-01 3.15758646e-01 6.28470898e-01
2.49993369e-01 -5.31132519e-01 3.33806127e-01 1.45120209e-03
-1.70492187e-01 -5.62471263e-02 -1.15635979e+00 -8.47583532e-01
-7.63131119e-03 -4.58020538e-01 7.63360620e-01 9.86110806e-01
-7.93582201e-01 -1.49841100e-01 -8.42753589e-01 1.04837000e+00
7.13706076e-01 5.57512224e-01 3.15399706e-01 -1.07104158e+00
-1.21424653e-01 1.23063862e-01 -2.98850834e-01 -1.08577025e+00
-6.55668974e-02 -8.13562810e-01 1.96717203e-01 -1.45333493e+00
-1.45079479e-01 -4.14847016e-01 -1.61882654e-01 4.15356904e-01
8.05555657e-03 2.58879662e-01 1.58459097e-01 5.34380317e-01
1.68580160e-01 7.16474175e-01 9.67304587e-01 -4.07179952e-01
1.50853530e-01 -2.12722629e-01 -7.88958520e-02 1.81894913e-01
5.71952164e-01 -5.81722081e-01 -1.87043890e-01 -1.20157266e+00
2.87657678e-01 6.07549191e-01 8.27878475e-01 -1.20579588e+00
9.75980461e-02 -2.02843979e-01 3.81926239e-01 -8.48527908e-01
3.26997697e-01 -1.16900182e+00 7.76690662e-01 5.41664422e-01
-9.03413668e-02 -2.68834233e-02 2.55265146e-01 7.12591231e-01
-5.53083599e-01 3.19277905e-02 9.07445431e-01 -1.52359366e-01
-1.12376535e+00 5.34384310e-01 -2.50567585e-01 -4.13743794e-01
8.01056027e-01 -6.13127556e-03 -9.44023132e-02 -4.71527427e-01
-6.04926884e-01 1.37931854e-01 4.01887059e-01 4.40733254e-01
5.98481834e-01 -1.21583521e+00 -1.14800477e+00 5.79955757e-01
2.51416028e-01 -9.00676772e-02 6.55854225e-01 5.68707407e-01
-9.37141418e-01 3.51341844e-01 -2.62385517e-01 -8.53817165e-01
-9.21329081e-01 4.10983056e-01 6.28830791e-01 2.81543195e-01
-8.10833931e-01 7.36408412e-01 -2.03448683e-01 -1.66229397e-01
-2.69874185e-01 -5.46135545e-01 1.00122154e-01 9.08914134e-02
8.10145795e-01 -1.41170651e-01 3.14041644e-01 -9.25522804e-01
-2.10580945e-01 5.39318502e-01 5.03292561e-01 -9.32267681e-02
1.74610293e+00 -2.47095108e-01 -3.72972250e-01 3.78386259e-01
1.36281526e+00 -5.32420814e-01 -1.65465522e+00 -6.59046471e-01
-1.80462509e-01 -7.89627671e-01 4.71910506e-01 -6.82298422e-01
-1.34478939e+00 1.01727057e+00 9.28213000e-01 2.02292368e-01
1.54997385e+00 -3.66950184e-01 7.80600846e-01 2.71255523e-02
1.60392001e-01 -6.53950036e-01 -4.58055705e-01 5.62243938e-01
1.05578744e+00 -1.17794931e+00 -8.43871906e-02 -1.26357317e-01
-2.79541314e-01 1.23450625e+00 -2.04224765e-01 1.05294622e-02
6.52608693e-01 1.22657217e-01 7.03539774e-02 5.51945530e-02
-7.58735418e-01 -3.94622833e-01 9.07839276e-03 5.79870284e-01
9.18637514e-02 3.91655743e-01 9.23945382e-02 -1.75964069e-02
9.29600596e-02 3.81445616e-01 5.24581909e-01 7.25380838e-01
-5.01840830e-01 -6.81814671e-01 -5.27059972e-01 2.43675470e-01
-6.11005537e-02 -2.81093836e-01 3.79211694e-01 3.60294074e-01
3.08272958e-01 6.63980961e-01 5.08336484e-01 -2.73232967e-01
8.15742388e-02 7.75134936e-02 9.72885452e-03 -2.99523950e-01
-2.18560264e-01 1.00328550e-01 -1.69694453e-01 -5.46709478e-01
-7.26310551e-01 -9.04959261e-01 -8.83069932e-01 -3.47014159e-01
2.64111366e-02 1.68316588e-02 8.96879613e-01 7.53365517e-01
3.40936005e-01 5.16979814e-01 9.66624141e-01 -9.84726310e-01
-4.73871559e-01 -7.74509728e-01 -8.02530885e-01 9.67152044e-02
1.23417342e+00 -1.74541429e-01 -3.94616276e-01 3.59052330e-01] | [9.764838218688965, -1.6961698532104492] |
9d35b856-feec-4741-a93d-f2a9ac03e7ec | improving-short-video-speech-recognition | 2210.15876 | null | https://arxiv.org/abs/2210.15876v2 | https://arxiv.org/pdf/2210.15876v2.pdf | Random Utterance Concatenation Based Data Augmentation for Improving Short-video Speech Recognition | One of limitations in end-to-end automatic speech recognition (ASR) framework is its performance would be compromised if train-test utterance lengths are mismatched. In this paper, we propose an on-the-fly random utterance concatenation (RUC) based data augmentation method to alleviate train-test utterance length mismatch issue for short-video ASR task. Specifically, we are motivated by observations that our human-transcribed training utterances tend to be much shorter for short-video spontaneous speech (~3 seconds on average), while our test utterance generated from voice activity detection front-end is much longer (~10 seconds on average). Such a mismatch can lead to suboptimal performance. Empirically, it's observed the proposed RUC method significantly improves long utterance recognition without performance drop on short one. Overall, it achieves 5.72% word error rate reduction on average for 15 languages and improved robustness to various utterance length. | ['Zejun Ma', 'Lu Lu', 'Tze Yuang Chong', 'Yerbolat Khassanov', 'Van Tung Pham', 'HaiHua Xu', 'Tao Han', 'Yist Y. Lin', 'Yi He'] | 2022-10-28 | null | null | null | null | ['activity-detection'] | ['computer-vision'] | [ 4.44625318e-01 1.38679489e-01 6.26908289e-03 -5.29107928e-01
-1.40386832e+00 -6.82213545e-01 5.50392091e-01 -3.40113372e-01
-4.53597248e-01 5.67817450e-01 4.39152092e-01 -7.62578785e-01
5.48993647e-01 -2.96835694e-03 -3.69050115e-01 -4.81012940e-01
1.93998173e-01 1.02370031e-01 -8.14208910e-02 -1.52866721e-01
1.44352153e-01 3.17521185e-01 -1.25055599e+00 1.55786633e-01
5.24403214e-01 7.72611439e-01 2.80722976e-01 1.35537374e+00
-8.18134397e-02 8.36701393e-01 -1.06165576e+00 -1.29502356e-01
8.04225430e-02 -4.97131348e-01 -7.56695211e-01 7.05880582e-01
3.17661434e-01 -5.10270298e-01 -7.28661895e-01 7.85879076e-01
8.03232551e-01 4.39218521e-01 3.48806739e-01 -8.72322381e-01
-5.44370711e-01 5.81513822e-01 -2.16907829e-01 4.06159997e-01
5.13533175e-01 3.62107635e-01 6.65777147e-01 -1.16825223e+00
2.30115667e-01 1.20172298e+00 1.81776419e-01 8.65317404e-01
-6.18010879e-01 -4.64367270e-01 2.21330687e-01 -1.55133665e-01
-1.46311533e+00 -1.19381368e+00 5.24691582e-01 -1.72139853e-01
1.40671837e+00 6.37533307e-01 3.33494507e-02 1.36244035e+00
-2.39244506e-01 8.79964471e-01 6.87091231e-01 -4.61576670e-01
1.52393043e-01 -9.46326181e-02 2.69005179e-01 3.98981392e-01
-4.17970628e-01 -4.97204661e-02 -3.61005932e-01 2.00917065e-01
6.46673143e-01 -2.53046930e-01 -3.78665179e-01 7.93943226e-01
-1.08736312e+00 6.28321409e-01 -3.55029166e-01 2.01436341e-01
-2.48540029e-01 -3.36056352e-02 7.54085958e-01 8.16210091e-01
4.73517567e-01 9.25807729e-02 -4.98879135e-01 -8.10652554e-01
-7.91365683e-01 -2.13382859e-02 4.58028138e-01 1.21473813e+00
1.21541917e-01 8.14445019e-01 -2.78745294e-01 1.31341684e+00
2.97725081e-01 6.21493578e-01 8.14876735e-01 -7.69427538e-01
7.87820041e-01 -1.39837220e-01 -6.56368956e-02 -2.35489056e-01
-6.46178499e-02 -5.19802988e-01 -6.01191819e-01 -2.76901901e-01
1.84021443e-01 -4.16334301e-01 -1.11248910e+00 1.53313220e+00
-2.89810207e-02 2.18035594e-01 5.54108799e-01 9.00201678e-01
8.17551315e-01 9.60715413e-01 -1.78445920e-01 -8.12989593e-01
1.19679296e+00 -1.33240283e+00 -1.19701993e+00 -1.73743188e-01
9.93630350e-01 -1.13376415e+00 1.37042630e+00 3.33797336e-01
-1.06851315e+00 -6.55047417e-01 -1.05617738e+00 2.54718423e-01
-9.87661653e-04 2.40175322e-01 2.53611933e-02 1.01738524e+00
-1.02266610e+00 2.20583475e-04 -5.72891831e-01 -2.29670182e-01
-1.29973367e-01 1.57007113e-01 -3.98536921e-01 -1.22264251e-01
-9.90567803e-01 4.60897893e-01 2.80081034e-01 1.30971819e-01
-1.05009270e+00 -2.29541734e-01 -9.02983308e-01 -2.52691180e-01
5.16189814e-01 4.88286875e-02 1.77793002e+00 -9.19439733e-01
-2.13025022e+00 4.88812387e-01 -4.79000330e-01 -5.26366711e-01
2.71812469e-01 -4.48781431e-01 -8.52656901e-01 -2.08525926e-01
-4.39249963e-01 4.30395544e-01 7.69678891e-01 -8.72205436e-01
-3.49812895e-01 -3.21861863e-01 -2.49055788e-01 5.00357985e-01
-2.17616096e-01 4.32138741e-01 -5.13422310e-01 -9.82665420e-01
4.29793857e-02 -9.65041935e-01 -6.24574907e-02 -8.43092263e-01
-3.98143739e-01 -3.13434899e-01 1.22140086e+00 -9.07804430e-01
1.55213284e+00 -2.20558691e+00 -2.91756451e-01 -2.02795073e-01
-3.36994290e-01 7.39397109e-01 -4.28162724e-01 4.07632798e-01
-1.60234243e-01 2.27801904e-01 1.02187507e-02 -5.11293828e-01
-3.45669210e-01 4.22028005e-01 -4.28539425e-01 4.34286237e-01
2.44761407e-01 6.58625364e-01 -7.31035173e-01 -9.92122814e-02
4.27311301e-01 3.80233735e-01 -3.09943050e-01 8.16567123e-01
2.27179732e-02 5.51869571e-01 -6.34496734e-02 7.54345834e-01
5.35433650e-01 3.62771243e-01 -1.69068322e-01 1.96812570e-01
-1.93605125e-01 6.16960704e-01 -7.93300152e-01 1.65098512e+00
-7.03103185e-01 1.11622024e+00 -6.81024343e-02 -8.36775124e-01
1.18446815e+00 1.06094456e+00 -6.38964623e-02 -6.09208822e-01
1.76066443e-01 1.39658645e-01 2.49395370e-01 -5.94639778e-01
4.80636358e-01 -3.31886718e-03 1.38122126e-01 3.02841038e-01
-7.71830827e-02 -3.74810323e-02 -1.12446370e-02 1.28080435e-02
1.22976637e+00 -3.83627832e-01 3.14141929e-01 2.87668526e-01
9.11784232e-01 -6.50111914e-01 4.20605510e-01 7.32389510e-01
-4.76745725e-01 8.94910932e-01 1.19443491e-01 -3.72863002e-02
-1.18966842e+00 -9.21373546e-01 1.46230862e-01 1.16552520e+00
-6.21991813e-01 -2.90522039e-01 -9.28698540e-01 -7.08220601e-01
-7.16378987e-01 9.42414939e-01 5.05137332e-02 1.29743814e-01
-7.60459304e-01 -1.92773297e-01 1.02172899e+00 6.16633832e-01
3.47210407e-01 -1.08915007e+00 5.07040173e-02 3.97753119e-01
-1.97957933e-01 -1.68205500e+00 -9.97348070e-01 -6.59248605e-02
-8.01068246e-01 -3.45097572e-01 -9.29210663e-01 -6.30652308e-01
5.42172670e-01 5.31218946e-01 8.34185183e-01 -1.08573008e-02
1.70273688e-02 3.44337136e-01 -8.90211880e-01 -2.73325324e-01
-8.70727599e-01 -1.24871731e-01 4.08479393e-01 1.23219796e-01
3.10290247e-01 -1.91324636e-01 -3.83791298e-01 6.09940648e-01
-7.75569856e-01 -1.42970994e-01 4.47861135e-01 8.92646134e-01
3.81085753e-01 -2.83950925e-01 1.12088764e+00 -5.60809195e-01
7.59857476e-01 -3.25434983e-01 -3.69883835e-01 3.89641047e-01
-5.12014210e-01 -1.61119759e-01 6.43288493e-01 -6.34016454e-01
-1.03055310e+00 6.20341999e-03 -7.71912456e-01 -6.20166361e-01
-4.39696044e-01 3.49721044e-01 -2.43910611e-01 3.40713501e-01
3.57064188e-01 5.56553304e-01 9.01131928e-02 -4.48729903e-01
1.23473816e-01 1.60681593e+00 5.72555363e-01 -1.94030970e-01
3.71280521e-01 -2.97617406e-01 -7.06724882e-01 -1.57945383e+00
-5.72995424e-01 -7.65123546e-01 -3.53012323e-01 -3.43335181e-01
7.11572230e-01 -1.20481241e+00 -4.92747694e-01 4.52451259e-01
-1.31661642e+00 -4.05085385e-01 2.09671631e-01 6.97442591e-01
-4.83257145e-01 4.58880007e-01 -6.74450636e-01 -1.41561794e+00
-7.41160154e-01 -1.46761787e+00 1.08293688e+00 -1.61679685e-02
-3.43010277e-01 -6.13122761e-01 -1.54254615e-01 7.22220242e-01
3.53415310e-01 -3.82301033e-01 2.23137677e-01 -1.05308521e+00
-1.62821874e-01 -4.72994864e-01 -2.58191489e-02 8.50566685e-01
3.20093274e-01 6.65587187e-02 -1.20373678e+00 -6.25221968e-01
9.59561095e-02 -2.62900978e-01 3.10681850e-01 3.05986851e-01
1.22310460e+00 -5.68313122e-01 3.61500144e-01 1.81573257e-01
1.02821481e+00 7.19382405e-01 9.02399123e-01 -3.06937933e-01
6.23669505e-01 4.91314709e-01 9.80257571e-01 4.64167565e-01
-2.82494754e-01 8.97824049e-01 -3.41283828e-02 1.60344079e-01
-3.25112939e-01 -8.26016814e-02 9.60723698e-01 1.68353915e+00
1.10774323e-01 -9.34540093e-01 -8.08936417e-01 5.32685280e-01
-1.47646415e+00 -8.92334700e-01 -1.19212918e-01 2.43015409e+00
7.79213250e-01 2.38577098e-01 3.04639608e-01 3.07593614e-01
7.00247407e-01 3.42572570e-01 -3.54218543e-01 -7.28151381e-01
-1.31106645e-01 -5.99506162e-02 4.42017496e-01 9.37450647e-01
-7.10008323e-01 9.80151534e-01 6.36385393e+00 1.00073647e+00
-1.17849469e+00 2.68930405e-01 7.15628743e-01 -2.57159948e-01
-5.43182082e-02 -3.50784540e-01 -7.83493042e-01 4.85138148e-01
1.76629376e+00 -7.11237490e-02 3.60653937e-01 7.63685644e-01
6.94050908e-01 3.03857952e-01 -9.37482953e-01 1.40714908e+00
2.38847286e-01 -1.04058397e+00 -3.72174084e-02 -3.80749279e-03
4.82558399e-01 1.89322099e-01 7.02212602e-02 5.38638234e-01
-1.29938871e-01 -1.16628051e+00 4.52608407e-01 -1.99354112e-01
1.19465506e+00 -9.32946563e-01 8.94854307e-01 3.72757733e-01
-1.24223912e+00 8.77854973e-02 -3.38884503e-01 -6.47703409e-02
4.12212402e-01 2.91261017e-01 -1.37561131e+00 2.09554151e-01
-3.46084847e-03 1.73504993e-01 -2.48714626e-01 5.45044899e-01
5.29946275e-02 1.28764868e+00 -9.98031944e-02 -7.65808448e-02
4.82260078e-01 2.81051937e-02 7.16651440e-01 1.79604280e+00
3.94015133e-01 3.83686304e-01 1.25585184e-01 -1.32910013e-01
-2.53510058e-01 2.27537170e-01 -1.02375674e+00 -3.84003669e-01
7.39634633e-01 7.46442616e-01 -3.05335850e-01 -4.95313197e-01
-6.19830310e-01 1.18236172e+00 -2.74216905e-02 6.69960320e-01
-8.97234499e-01 -4.28740412e-01 7.31421113e-01 -2.13719875e-01
1.51935322e-02 -4.76287842e-01 -1.00331202e-01 -9.33359504e-01
6.39564320e-02 -1.17519438e+00 -1.07283029e-03 -5.29485643e-01
-8.10786486e-01 9.05860186e-01 -5.67414701e-01 -1.35711527e+00
-5.64855754e-01 -4.12947476e-01 -6.29946172e-01 6.68949604e-01
-1.00540078e+00 -6.90713584e-01 -7.45934376e-04 3.74232203e-01
2.00714946e+00 -6.68057621e-01 8.51480424e-01 4.40107226e-01
-9.21100140e-01 1.26674056e+00 2.41931342e-02 3.18161428e-01
4.94691104e-01 -9.31606948e-01 7.98092842e-01 1.11679292e+00
2.97627836e-01 4.47840214e-01 8.37565303e-01 -4.62283105e-01
-1.42262077e+00 -1.14851737e+00 7.24119902e-01 -3.29726487e-01
4.17165786e-01 -3.16787988e-01 -1.05881643e+00 5.81347585e-01
3.98705840e-01 -8.92498940e-02 8.39069426e-01 -1.79921657e-01
-2.99435973e-01 1.46295309e-01 -8.43105078e-01 7.15390980e-01
8.29832792e-01 -9.62626934e-01 -4.64567035e-01 3.32590014e-01
1.22875607e+00 -3.36513042e-01 -8.59832823e-01 3.31426561e-01
4.69925731e-01 -6.23233199e-01 6.75218523e-01 -4.96185958e-01
7.24446625e-02 -1.31318718e-01 -6.58942640e-01 -1.06578970e+00
2.82281339e-01 -1.03825963e+00 -2.59203076e-01 1.45379901e+00
6.45410597e-01 -3.78934175e-01 6.96087182e-01 4.03325170e-01
-6.38748348e-01 -6.98702693e-01 -9.96923566e-01 -1.10739052e+00
-2.78109580e-01 -8.09426665e-01 1.87089190e-01 6.70643032e-01
-3.05276830e-02 6.46492064e-01 -6.57347322e-01 2.88345903e-01
1.77487433e-01 -7.89824963e-01 8.19547594e-01 -3.97360951e-01
-3.06715161e-01 -6.58850595e-02 -2.77376264e-01 -1.56570470e+00
1.60394341e-01 -3.66935521e-01 4.53811437e-01 -8.96342814e-01
-2.29478106e-01 -5.98033778e-02 -2.45102301e-01 1.22757971e-01
-1.69356912e-01 8.34248774e-03 9.24938992e-02 -1.55976087e-01
-4.75592226e-01 7.24967539e-01 9.47081923e-01 6.56140223e-02
-2.85875201e-01 1.79573342e-01 -3.62698138e-02 4.62844819e-01
7.87659168e-01 -2.70100534e-01 -7.16310143e-01 -5.49120784e-01
-6.49452090e-01 5.53350091e-01 -3.48003626e-01 -7.73441315e-01
-1.66395586e-02 -8.04701671e-02 -3.01069409e-01 -7.70155489e-01
4.43722367e-01 -4.00006324e-01 -1.55469552e-01 2.21330032e-01
-4.94533181e-01 2.53488123e-01 1.82629198e-01 5.27291298e-01
-3.92829120e-01 -3.98043275e-01 7.32399285e-01 1.12531178e-01
-4.93432224e-01 8.26734900e-02 -8.31430316e-01 -1.10705458e-01
6.92959547e-01 -3.42685461e-01 -1.17697187e-01 -7.38682270e-01
-4.62925076e-01 -5.22903949e-02 -3.14867906e-02 7.11194277e-01
9.78532970e-01 -1.25302052e+00 -1.05497539e+00 2.79878289e-01
1.85434163e-01 -1.30105674e-01 3.85018170e-01 6.32822812e-01
-4.20967430e-01 8.74689043e-01 4.90186483e-01 -5.78273952e-01
-1.88869882e+00 3.10309470e-01 -4.50460333e-03 8.33252296e-02
-4.01135623e-01 1.16938758e+00 -2.16211975e-02 -3.01195771e-01
6.46021962e-01 -2.21710190e-01 2.51770150e-02 -3.18737596e-01
8.15587580e-01 4.96101856e-01 2.43348137e-01 -7.48400569e-01
-1.36301532e-01 3.32778066e-01 -5.37980497e-01 -6.11281097e-01
8.22525978e-01 -5.22967219e-01 6.20854914e-01 6.04166329e-01
1.51441157e+00 8.00557211e-02 -1.14962125e+00 -2.17605814e-01
-9.58104134e-02 -5.74756324e-01 1.04370907e-01 -6.03532195e-01
-7.08865166e-01 9.19771075e-01 7.30353177e-01 1.37916118e-01
1.04280353e+00 -9.80067775e-02 1.12168348e+00 6.74929857e-01
9.23216790e-02 -1.15151644e+00 2.71251559e-01 8.60264540e-01
1.14722431e+00 -1.42381060e+00 -3.84324670e-01 -2.28327096e-01
-9.08809721e-01 1.02116096e+00 6.34717941e-01 3.27680558e-01
7.83975497e-02 1.87146008e-01 4.14819837e-01 2.88947791e-01
-1.07965720e+00 -1.55369982e-01 1.42325729e-01 5.71864367e-01
9.35214102e-01 1.69937372e-01 -1.58971861e-01 1.15517542e-01
-2.62820125e-01 -3.97741288e-01 7.03052580e-01 7.78269351e-01
-5.77702284e-01 -1.03038561e+00 -4.59261417e-01 1.84868038e-01
-7.37027943e-01 -2.12248087e-01 -2.00229660e-01 2.78339952e-01
-6.33630276e-01 1.41517818e+00 2.17437506e-01 -5.35897493e-01
2.40796462e-01 3.35433930e-01 7.91271776e-02 -7.85601020e-01
-3.29593301e-01 4.98314500e-01 5.36120534e-01 -3.47647399e-01
2.39592716e-02 -5.08568645e-01 -1.30874574e+00 -1.30853713e-01
-5.69969893e-01 1.00413136e-01 8.50357771e-01 1.08405435e+00
2.12419808e-01 6.67655766e-01 1.06157768e+00 -4.18019712e-01
-7.98954427e-01 -1.47931910e+00 -2.43706077e-01 2.36228943e-01
6.73841655e-01 -1.51759714e-01 -4.24628973e-01 3.10279548e-01] | [14.499856948852539, 6.723163604736328] |
7a90e737-e2fa-4994-b5b9-da3e46b4bd1c | md-manifold-a-medical-distance-based | 2305.00553 | null | https://arxiv.org/abs/2305.00553v1 | https://arxiv.org/pdf/2305.00553v1.pdf | MD-Manifold: A Medical-Distance-Based Representation Learning Approach for Medical Concept and Patient Representation | Effectively representing medical concepts and patients is important for healthcare analytical applications. Representing medical concepts for healthcare analytical tasks requires incorporating medical domain knowledge and prior information from patient description data. Current methods, such as feature engineering and mapping medical concepts to standardized terminologies, have limitations in capturing the dynamic patterns from patient description data. Other embedding-based methods have difficulties in incorporating important medical domain knowledge and often require a large amount of training data, which may not be feasible for most healthcare systems. Our proposed framework, MD-Manifold, introduces a novel approach to medical concept and patient representation. It includes a new data augmentation approach, concept distance metric, and patient-patient network to incorporate crucial medical domain knowledge and prior data information. It then adapts manifold learning methods to generate medical concept-level representations that accurately reflect medical knowledge and patient-level representations that clearly identify heterogeneous patient cohorts. MD-Manifold also outperforms other state-of-the-art techniques in various downstream healthcare analytical tasks. Our work has significant implications in information systems research in representation learning, knowledge-driven machine learning, and using design science as middle-ground frameworks for downstream explorative and predictive analyses. Practically, MD-Manifold has the potential to create effective and generalizable representations of medical concepts and patients by incorporating medical domain knowledge and prior data information. It enables deeper insights into medical data and facilitates the development of new analytical applications for better healthcare outcomes. | ['Wenli Zhang', 'Qing Li', 'Shaodong Wang'] | 2023-04-30 | null | null | null | null | ['feature-engineering'] | ['methodology'] | [ 9.54176858e-02 3.28634143e-01 -4.88481730e-01 -4.54289615e-01
-6.08845294e-01 -1.14143424e-01 3.63907844e-01 1.21289337e+00
4.49969321e-02 3.87100786e-01 8.19265723e-01 -4.77025688e-01
-7.32712030e-01 -1.02536166e+00 1.38078071e-02 -4.61504072e-01
-3.85054827e-01 8.46696615e-01 -7.50293672e-01 -2.49790147e-01
-7.39776418e-02 4.95175451e-01 -1.22290421e+00 4.70610976e-01
9.19404268e-01 7.81973362e-01 -7.85362720e-02 4.06385034e-01
-5.36597848e-01 8.32210302e-01 -4.74889457e-01 -2.17204288e-01
-4.18844670e-02 -6.24608159e-01 -6.83959126e-01 -1.07770666e-01
-1.95432842e-01 1.90799400e-01 -2.10712850e-01 7.39223659e-01
6.02839887e-01 4.05333042e-02 9.34540331e-01 -1.04288626e+00
-1.30993712e+00 4.46104765e-01 -2.87346959e-01 1.03838407e-01
3.76516879e-01 2.58378573e-02 7.32439339e-01 -7.82481372e-01
6.71733677e-01 1.28464687e+00 9.08411860e-01 8.63364220e-01
-1.09243405e+00 -4.13728386e-01 9.20118839e-02 1.67784065e-01
-1.18866706e+00 -7.73275048e-02 7.78030336e-01 -9.14836943e-01
8.27207386e-01 4.43177700e-01 7.45891511e-01 8.79661918e-01
3.24251562e-01 6.81945682e-01 3.80530477e-01 -3.79767299e-01
2.48799518e-01 2.90370196e-01 3.95284623e-01 7.86049068e-01
5.06086707e-01 2.50778973e-01 -2.54757524e-01 -5.48155427e-01
6.28733873e-01 1.05291939e+00 -2.57150203e-01 -3.71634930e-01
-1.53218305e+00 1.16839993e+00 5.61485946e-01 5.97621143e-01
-7.37757564e-01 -1.06104761e-01 6.12286150e-01 3.20625067e-01
3.94721031e-01 9.11566556e-01 -6.04139626e-01 6.34286851e-02
-7.87160397e-01 -1.13950167e-02 5.85643709e-01 8.91058207e-01
3.96919817e-01 -8.06193799e-02 -4.12967771e-01 5.77609777e-01
3.90409857e-01 3.37865293e-01 9.69756126e-01 -4.82685566e-01
3.94978315e-01 1.30162752e+00 -6.31524399e-02 -1.17353272e+00
-7.14625537e-01 -5.10791183e-01 -1.07244349e+00 -4.00178671e-01
-1.03530340e-01 -1.30205631e-01 -1.03220248e+00 1.39400411e+00
2.85361201e-01 1.81433171e-01 5.83006859e-01 7.49102175e-01
1.17008853e+00 3.04530948e-01 4.10519242e-01 -1.00490242e-01
1.72691262e+00 -4.41751331e-01 -8.94961476e-01 1.42636955e-01
1.17818010e+00 -3.77843767e-01 8.55416238e-01 -8.89221653e-02
-7.80518532e-01 -3.71259928e-01 -8.29123557e-01 7.36753270e-02
-6.70129061e-01 -9.22551453e-02 9.65599418e-01 6.19335651e-01
-5.84784865e-01 5.71741223e-01 -8.65257442e-01 -4.51208889e-01
9.20240402e-01 2.17511535e-01 -4.80210721e-01 -4.89904881e-01
-1.27086961e+00 8.35275650e-01 3.12104046e-01 -3.39348428e-02
-6.72641873e-01 -1.47905469e+00 -1.41390109e+00 1.67215884e-01
1.06028989e-02 -1.43822205e+00 6.95128500e-01 -4.46817964e-01
-9.95867014e-01 7.21523702e-01 1.36015356e-01 -3.00346524e-01
1.03059992e-01 -1.08411074e-01 -8.64155173e-01 9.78517905e-02
-2.11115368e-02 3.16511929e-01 4.44365591e-01 -8.58413100e-01
-4.34679717e-01 -6.66241467e-01 -5.34489095e-01 -5.41712902e-02
-6.07507765e-01 -3.67708743e-01 4.82810624e-02 -7.79930890e-01
1.29308075e-01 -5.18043697e-01 -7.16393948e-01 1.61375310e-02
-3.84449422e-01 4.41417983e-03 4.96897310e-01 -6.40443206e-01
1.38853550e+00 -2.12204790e+00 1.94158107e-01 4.01315480e-01
6.95087850e-01 2.57371575e-01 -1.14561737e-01 6.37375832e-01
-2.05556601e-01 1.61306232e-01 -2.76713103e-01 -8.79694298e-02
-1.98727131e-01 6.54547289e-02 5.59597723e-02 3.72051418e-01
5.97231090e-01 1.23773777e+00 -1.25512874e+00 -4.18521881e-01
4.04144078e-01 8.12434673e-01 -6.73151076e-01 2.76437372e-01
-8.12617019e-02 5.74015856e-01 -7.42995083e-01 8.74557137e-01
1.80407047e-01 -6.04665101e-01 1.89515382e-01 -1.96128219e-01
3.44228953e-01 -4.78601754e-02 -7.71533012e-01 1.83615124e+00
-3.88646036e-01 3.11082453e-01 -4.06599224e-01 -1.44769955e+00
1.05878878e+00 4.67382044e-01 1.15601218e+00 -4.25997734e-01
1.50868133e-01 -1.25782579e-01 4.66045290e-02 -9.37200606e-01
1.78939924e-02 -5.19394100e-01 -1.34466827e-01 2.45032147e-01
-5.63913733e-02 3.15085888e-01 -2.43024617e-01 1.18503980e-01
1.17494905e+00 -4.96737540e-01 5.39219081e-01 -3.09126019e-01
4.46552157e-01 3.06264043e-01 7.32627332e-01 2.94375539e-01
-1.21565372e-01 3.87403101e-01 3.26685578e-01 -7.52707601e-01
-7.52675653e-01 -1.21849418e+00 -4.12845016e-01 6.74458086e-01
-2.51276314e-01 -6.86504900e-01 -1.38327911e-01 -6.15787685e-01
6.43824458e-01 5.11177182e-01 -9.70845640e-01 -7.26977229e-01
-1.77576795e-01 -1.02023625e+00 3.09081465e-01 8.53863835e-01
-2.36325786e-01 -8.58035564e-01 -5.42024553e-01 5.16013920e-01
8.25105235e-03 -4.69027817e-01 -4.05276030e-01 -1.59769520e-01
-1.26850104e+00 -1.59268045e+00 -7.22911000e-01 -8.27986300e-01
9.16288674e-01 -2.61236727e-01 1.27070332e+00 -5.30189946e-02
-1.03191459e+00 5.72275281e-01 -3.93915653e-01 -8.44548643e-01
-6.33261144e-01 -2.09331349e-01 -3.52361263e-03 -6.05128817e-02
9.44839120e-01 -1.36598483e-01 -9.96482551e-01 -2.39670780e-02
-9.55539584e-01 -2.35548809e-01 7.26454198e-01 1.18076956e+00
6.72580361e-01 -2.92768758e-02 9.59510326e-01 -1.29182863e+00
9.57707226e-01 -1.01991558e+00 1.20478295e-01 2.87713796e-01
-1.12080252e+00 3.59143466e-01 3.86060387e-01 -2.43634939e-01
-6.24693871e-01 -2.28033647e-01 1.21774420e-01 -6.27941966e-01
-8.12851191e-02 1.03344786e+00 2.82098893e-02 5.54903090e-01
1.02656198e+00 -5.68553917e-02 6.22112453e-01 -6.78674102e-01
6.51717842e-01 8.25681388e-01 3.95054996e-01 -3.07766467e-01
3.54195058e-01 5.21927893e-01 2.19927132e-01 -5.78548908e-01
-5.96791208e-01 -8.34456503e-01 -6.24156713e-01 4.63778555e-01
1.00578332e+00 -9.17674780e-01 -5.52247584e-01 -2.53052354e-01
-6.11755073e-01 2.56697357e-01 -7.86005080e-01 7.69819260e-01
-3.32683235e-01 1.49250135e-01 -5.77744782e-01 -4.66003865e-01
-6.70095980e-01 -8.46383750e-01 8.89771223e-01 -1.53115049e-01
-6.53856277e-01 -1.61065674e+00 1.76256090e-01 2.04291016e-01
4.66599971e-01 9.32428241e-01 1.47169292e+00 -8.37998331e-01
2.26698965e-02 -3.57757807e-01 6.02162927e-02 -1.96541231e-02
8.98886144e-01 -3.25942546e-01 -5.25127590e-01 -3.44011813e-01
-1.77477613e-01 1.60324723e-01 8.47790480e-01 4.84114617e-01
1.16490662e+00 -4.17225897e-01 -7.88294733e-01 5.95546782e-01
1.22241294e+00 2.87840396e-01 3.42433780e-01 -2.59938519e-02
8.44326496e-01 9.90955710e-01 6.09672010e-01 6.64353967e-01
6.57411814e-01 2.14438647e-01 1.23615712e-01 -5.69049418e-01
1.49837285e-01 -1.42158285e-01 -2.42312506e-01 8.32128882e-01
3.26826811e-01 2.80465543e-01 -1.31467533e+00 8.34541321e-01
-1.85180426e+00 -9.18207049e-01 6.17150329e-02 2.02525020e+00
8.92527223e-01 -4.85073268e-01 1.60286903e-01 2.17008829e-01
5.58116436e-01 -5.75231552e-01 -7.02421248e-01 -3.32996368e-01
2.47494787e-01 2.84849107e-01 8.69895145e-02 7.46055245e-02
-1.03988945e+00 3.64394486e-01 6.41674423e+00 -1.83782540e-02
-8.91590297e-01 1.10003673e-01 6.04690254e-01 -1.86767414e-01
-6.29797399e-01 -5.24704158e-01 -3.49824965e-01 4.11437482e-01
1.15723848e+00 -3.81683409e-01 -1.22521341e-01 8.62842083e-01
3.06162059e-01 6.18059456e-01 -1.68277586e+00 1.34673464e+00
1.03901938e-01 -1.75699604e+00 4.18723911e-01 1.25454634e-01
6.73333108e-01 -3.55100006e-01 1.28412977e-01 1.26975864e-01
4.06220138e-01 -1.49919307e+00 -1.46386892e-01 9.95466471e-01
1.08648121e+00 -7.61152923e-01 9.60954428e-01 -7.82479495e-02
-1.11939847e+00 -4.19159412e-01 -2.78293878e-01 1.86105296e-01
7.15952516e-02 7.11248875e-01 -1.14887738e+00 8.78835738e-01
4.84362572e-01 1.21604586e+00 -4.36367929e-01 8.55388463e-01
4.19998974e-01 3.20903748e-01 3.33901882e-01 2.09006757e-01
4.57746722e-03 -1.20166004e-01 1.30935207e-01 1.27539611e+00
3.58372897e-01 2.85149306e-01 3.40464503e-01 9.27256942e-01
1.81725532e-01 3.30298930e-01 -9.01043534e-01 -5.45843363e-01
4.60489571e-01 9.27204192e-01 -3.13065767e-01 -4.76202756e-01
-5.14113665e-01 5.25513113e-01 -1.14768028e-01 1.68599769e-01
-2.89838105e-01 -4.11311179e-01 1.19076693e+00 3.47646952e-01
-1.20142952e-01 2.35596895e-01 -3.29052269e-01 -1.06035006e+00
-3.60869110e-01 -8.87267172e-01 1.06901896e+00 -6.44548610e-02
-1.85609996e+00 4.87684906e-01 -1.21321276e-01 -1.43657887e+00
-5.38597584e-01 -7.03360736e-01 -3.02157849e-01 9.10813153e-01
-1.56777608e+00 -1.17313612e+00 -3.69069338e-01 8.07976365e-01
4.25949730e-02 -7.29751527e-01 1.37679935e+00 4.96599674e-01
-5.07705271e-01 6.21740162e-01 2.82308131e-01 3.52161705e-01
6.36853337e-01 -1.31003749e+00 1.59053624e-01 6.98670372e-02
-1.39575481e-01 1.14970303e+00 2.80489981e-01 -6.11141741e-01
-1.51936376e+00 -1.53258228e+00 7.47635603e-01 -9.03333545e-01
2.79052019e-01 5.01369424e-02 -1.05527484e+00 5.63899517e-01
-4.42678452e-01 7.56260380e-02 1.86988258e+00 3.75531733e-01
-2.39235461e-01 -1.39871970e-01 -1.36084962e+00 2.16147810e-01
8.95877063e-01 -6.06031477e-01 -9.24427748e-01 4.51332241e-01
7.61726022e-01 2.20065843e-02 -1.63047421e+00 4.24336702e-01
3.04695576e-01 -1.27827466e-01 1.21001530e+00 -1.52222717e+00
3.35350335e-01 -2.46747062e-01 -3.99982557e-02 -1.40420330e+00
-6.07008338e-01 -3.61062467e-01 -4.65845227e-01 8.65853965e-01
4.67631370e-01 -7.03432560e-01 6.94207788e-01 9.02863443e-01
-1.33499846e-01 -1.02363729e+00 -5.34638345e-01 -5.33784151e-01
2.78614670e-01 -1.00763403e-01 1.06368136e+00 1.57988667e+00
4.96594399e-01 1.53704226e-01 -3.19861062e-02 1.20912246e-01
4.56660539e-01 2.25253642e-01 3.97037297e-01 -1.80443883e+00
-2.84003019e-02 -4.98588383e-01 -9.69471753e-01 2.73119425e-03
-7.47641996e-02 -1.31147873e+00 -6.13153517e-01 -2.02828383e+00
1.65311083e-01 -6.36285245e-01 -8.31504226e-01 4.37183022e-01
-3.92590225e-01 -1.77212149e-01 -1.73927441e-01 1.56458139e-01
-6.14174530e-02 5.64128637e-01 1.21529984e+00 -5.95764339e-01
-6.59952819e-01 -2.45358199e-01 -1.28308868e+00 2.38271862e-01
6.79016411e-01 -4.01620239e-01 -7.04753280e-01 -2.51213431e-01
-3.24731655e-02 1.40396189e-02 2.37540245e-01 -7.10781753e-01
-1.48048839e-02 -2.06834689e-01 6.24379575e-01 -1.35384530e-01
3.68355401e-02 -8.63781154e-01 3.70462865e-01 9.04534996e-01
-3.94144088e-01 3.01548153e-01 1.32110402e-01 8.41869831e-01
-3.82538438e-01 1.98562056e-01 4.79119718e-01 -3.11779171e-01
-7.01638520e-01 5.24199247e-01 -1.02830350e-01 1.36311218e-01
1.31645572e+00 -1.66197717e-01 -1.75945327e-01 2.40809359e-02
-1.14687157e+00 5.53133845e-01 2.52341241e-01 6.58896863e-01
9.96180654e-01 -1.57818747e+00 -9.46911931e-01 4.00440514e-01
6.98860586e-01 2.19110120e-02 3.09666097e-01 6.51133060e-01
-5.54396451e-01 5.64678609e-01 -1.69904724e-01 -6.38957858e-01
-1.06665099e+00 1.13533878e+00 3.06215048e-01 -6.44282699e-02
-9.72508311e-01 5.31883597e-01 2.33111724e-01 -4.55406725e-01
1.22365914e-01 -5.63019753e-01 -4.90839094e-01 2.98257589e-01
7.66648889e-01 3.58685762e-01 2.80634798e-02 -4.27351475e-01
-6.43711925e-01 4.79821265e-01 -5.01283929e-02 3.40485513e-01
1.55794191e+00 2.32343897e-01 4.45087478e-02 4.38252449e-01
1.26966608e+00 -3.78754109e-01 -4.59233522e-01 -3.46944809e-01
3.74397755e-01 -4.53769773e-01 5.21834679e-02 -8.14171016e-01
-9.38199818e-01 9.81817603e-01 8.41447771e-01 -1.87694073e-01
1.11994433e+00 2.73782343e-01 5.07646441e-01 3.78562212e-01
1.18955290e-02 -8.12813461e-01 3.07176411e-01 -1.41724885e-01
8.29029977e-01 -1.40996766e+00 -6.58339411e-02 -3.55553120e-01
-7.81894207e-01 1.02191031e+00 3.46268892e-01 2.41814181e-01
1.15422666e+00 -8.21769144e-03 4.57162976e-01 -8.19988012e-01
-6.43870354e-01 -7.36179352e-02 5.29749870e-01 1.03266108e+00
6.99081302e-01 4.71736908e-01 -1.11380585e-01 8.99633586e-01
-7.35645518e-02 1.29304174e-02 3.30163538e-02 9.35594976e-01
-4.37245145e-02 -1.24147093e+00 -3.34301442e-01 1.01099217e+00
-4.76493984e-01 -2.20978171e-01 -1.30725443e-01 5.99769056e-01
1.94802627e-01 8.44871342e-01 2.25294933e-01 -3.63286942e-01
5.57895243e-01 3.37267399e-01 1.06669106e-01 -1.13080442e+00
-4.66540247e-01 -3.57752144e-01 -2.55836517e-01 -4.91030723e-01
-3.38571846e-01 -6.30165458e-01 -1.41857123e+00 -1.85097650e-01
-1.48353726e-01 3.42415869e-01 4.54479337e-01 5.48655450e-01
1.18649101e+00 9.52259243e-01 4.29857492e-01 -2.13358000e-01
-4.67383444e-01 -8.84020805e-01 -4.65901166e-01 8.40649068e-01
6.01841390e-01 -6.80112660e-01 1.56453386e-01 9.88548473e-02] | [7.871105194091797, 6.758248329162598] |
89a67385-19e2-4524-83a6-0411dd3c2cc9 | controlling-extra-textual-attributes-about | 2205.04747 | null | https://arxiv.org/abs/2205.04747v2 | https://arxiv.org/pdf/2205.04747v2.pdf | Controlling Extra-Textual Attributes about Dialogue Participants -- A Case Study of English-to-Polish Neural Machine Translation | Unlike English, morphologically rich languages can reveal characteristics of speakers or their conversational partners, such as gender and number, via pronouns, morphological endings of words and syntax. When translating from English to such languages, a machine translation model needs to opt for a certain interpretation of textual context, which may lead to serious translation errors if extra-textual information is unavailable. We investigate this challenge in the English-to-Polish language direction. We focus on the underresearched problem of utilising external metadata in automatic translation of TV dialogue, proposing a case study where a wide range of approaches for controlling attributes in translation is employed in a multi-attribute scenario. The best model achieves an improvement of +5.81 chrF++/+6.03 BLEU, with other models achieving competitive performance. We additionally contribute a novel attribute-annotated dataset of Polish TV dialogue and a morphological analysis script used to evaluate attribute control in models. | ['Carolina Scarton', 'Loïc Barrault', 'Sebastian T. Vincent'] | 2022-05-10 | null | null | null | null | ['morphological-analysis'] | ['natural-language-processing'] | [ 2.85874099e-01 2.93549359e-01 -3.25900048e-01 -6.37846112e-01
-1.27272177e+00 -9.69079375e-01 1.07248378e+00 4.04914320e-02
-5.64117074e-01 1.08649325e+00 5.46461701e-01 -4.42615926e-01
1.86007485e-01 -6.29035890e-01 -2.48294368e-01 -4.06752497e-01
6.06895268e-01 1.41833806e+00 -2.27313757e-01 -5.90704620e-01
2.86608875e-01 1.46777526e-01 -9.46169913e-01 4.29334491e-01
7.70475209e-01 1.43309563e-01 -2.76511330e-02 4.94793206e-01
-5.12743652e-01 4.52203304e-01 -7.18132794e-01 -1.17606723e+00
1.99798748e-01 -3.50180656e-01 -1.07874632e+00 1.38653055e-01
5.07636666e-01 1.50653198e-01 1.34533465e-01 9.61884022e-01
7.00900793e-01 -1.80621654e-01 7.31945276e-01 -9.62174594e-01
-4.39748287e-01 1.10874510e+00 -1.58896163e-01 9.38044768e-03
8.05017054e-01 8.25387537e-02 1.21346831e+00 -8.03785563e-01
1.12952065e+00 1.47802246e+00 5.93337774e-01 7.22319841e-01
-1.70060086e+00 -4.05969650e-01 -3.12516034e-01 2.93603931e-02
-1.19646001e+00 -8.91330302e-01 5.56632936e-01 -2.61466265e-01
1.12191927e+00 6.49901032e-01 4.67391491e-01 1.33425093e+00
-3.84872989e-03 5.49311459e-01 1.63464725e+00 -7.63914466e-01
-1.89121291e-01 5.34973204e-01 -2.87944883e-01 2.87170261e-01
-1.96166873e-01 -2.32874438e-01 -7.79914796e-01 -2.85267085e-01
3.34320903e-01 -1.02822161e+00 6.62904754e-02 2.05590785e-01
-1.56008148e+00 8.62509251e-01 -5.57624638e-01 2.63150334e-01
-1.32915229e-01 -3.93812478e-01 6.00653231e-01 7.46041417e-01
5.21240294e-01 8.17289054e-01 -9.04956579e-01 -6.70645535e-01
-5.87661684e-01 4.91949409e-01 1.26299632e+00 1.27163470e+00
5.95361173e-01 -3.00236970e-01 -3.19520175e-01 1.22932005e+00
6.06122613e-02 7.69090354e-01 3.54311347e-01 -1.00977123e+00
8.66002977e-01 5.20654321e-01 4.50512730e-02 -5.06552577e-01
-2.98276603e-01 -1.99688017e-01 -2.88988233e-01 -3.59010726e-01
9.59001005e-01 -6.11383766e-02 -4.55630481e-01 1.74761283e+00
5.47272682e-01 -8.74335051e-01 3.92479062e-01 5.05342901e-01
6.89152479e-01 5.39916277e-01 2.84307599e-01 -5.56491315e-01
1.72215629e+00 -6.40838444e-01 -9.24340487e-01 -4.56706405e-01
7.92881548e-01 -1.41918898e+00 1.46266425e+00 1.36831775e-01
-1.43750000e+00 -9.00639147e-02 -4.90072250e-01 -2.87654221e-01
-3.11468273e-01 2.24761203e-01 5.44182360e-01 8.57375383e-01
-9.50347900e-01 3.04281622e-01 -4.10624713e-01 -7.43814290e-01
-1.69895053e-01 6.79586828e-01 -5.41254818e-01 1.77308947e-01
-1.32173073e+00 1.19335771e+00 6.26195446e-02 -3.25780869e-01
-1.26445815e-01 -6.56066239e-01 -8.18176270e-01 -3.50846618e-01
2.39399165e-01 -4.76577550e-01 1.52123117e+00 -7.12114632e-01
-1.71505404e+00 1.51883626e+00 -3.37133497e-01 2.15654038e-02
8.68802190e-01 1.35040715e-01 -2.80857921e-01 -2.66600102e-01
4.62304324e-01 4.14010197e-01 4.40769672e-01 -9.31596398e-01
-7.63089061e-01 -5.81842303e-01 -8.27553123e-02 4.69414890e-01
-2.13230640e-01 7.33895421e-01 -4.52082515e-01 -6.01247013e-01
1.51979342e-01 -1.18390298e+00 1.09966479e-01 -6.69728816e-01
-4.26518828e-01 -3.21099937e-01 2.87194252e-01 -9.82980371e-01
1.15288138e+00 -1.75730133e+00 1.78344369e-01 2.53844094e-02
-2.36792833e-01 -7.99173787e-02 1.06535360e-01 6.58017695e-01
3.40904027e-01 3.08582842e-01 -8.16611499e-02 -6.03219271e-01
2.61274308e-01 5.97822309e-01 -4.83042188e-02 4.18069988e-01
2.68248737e-01 8.00670624e-01 -6.80535078e-01 -8.79607022e-01
-5.04675470e-02 2.54865587e-01 -2.13577062e-01 7.55965710e-02
-1.61251277e-01 6.18198872e-01 -2.46039733e-01 8.15509558e-01
4.55116302e-01 5.51618040e-01 5.15753269e-01 1.40315503e-01
-3.65369290e-01 1.09100437e+00 -9.98485029e-01 1.49236917e+00
-8.18374574e-01 5.93464792e-01 3.12438488e-01 -2.68225282e-01
9.01217163e-01 5.20684600e-01 3.62100303e-02 -6.12302482e-01
1.46715656e-01 7.28118896e-01 2.96052188e-01 -4.43091422e-01
7.12292969e-01 -2.92881280e-01 -5.43178439e-01 5.40458798e-01
3.20177265e-02 -5.14236987e-01 3.70945841e-01 -1.52157724e-01
7.08375335e-01 2.10720748e-01 3.70956838e-01 -5.59331000e-01
7.40140736e-01 4.54423368e-01 7.51988351e-01 3.61174017e-01
-1.06919356e-01 5.17674625e-01 6.25010550e-01 -2.80968487e-01
-1.37797523e+00 -5.11856556e-01 -4.66011852e-01 1.56740117e+00
-5.39101362e-01 -3.75462443e-01 -9.40704882e-01 -6.37280524e-01
-3.94907057e-01 9.95379806e-01 -4.63853717e-01 1.60448492e-01
-1.09428096e+00 -9.21003222e-01 9.32820141e-01 1.64957583e-01
1.85566589e-01 -9.74601805e-01 -1.65609121e-01 4.52820659e-01
-8.15451860e-01 -1.41663110e+00 -5.40693104e-01 1.21612266e-01
-6.75919473e-01 -6.48750782e-01 -2.11799011e-01 -8.68893087e-01
2.61617273e-01 -6.11963272e-01 1.48229849e+00 -8.33700970e-02
3.06149423e-01 -1.45442900e-04 -3.05201709e-01 -5.80619872e-01
-1.12969649e+00 6.98973060e-01 3.40387747e-02 -2.40992919e-01
8.74420464e-01 -3.28250796e-01 2.62455456e-02 4.60102797e-01
-3.55036288e-01 3.92084941e-02 3.79775971e-01 9.13641870e-01
2.40326092e-01 -8.62958252e-01 3.14247221e-01 -1.25229323e+00
7.70586789e-01 -9.47379768e-02 -3.58903915e-01 2.54308462e-01
-5.99835694e-01 9.66347903e-02 4.19716597e-01 -5.66804826e-01
-1.12568629e+00 8.00754875e-02 -3.26214105e-01 5.30947387e-01
-1.52991161e-01 3.31533730e-01 -6.94531083e-01 1.08918555e-01
6.23617411e-01 2.82241821e-01 1.30950645e-01 -4.33185190e-01
1.92111924e-01 1.04416585e+00 5.63028574e-01 -9.61455166e-01
7.39851534e-01 5.28617427e-02 -1.12266451e-01 -6.75512671e-01
-2.30418980e-01 -4.23928142e-01 -9.30537224e-01 2.05185637e-03
5.39366245e-01 -8.28202009e-01 -4.59160686e-01 2.86648184e-01
-1.29178250e+00 -4.25998658e-01 -2.42935255e-01 5.20303249e-01
-8.49625170e-01 1.50333911e-01 -7.98223734e-01 -5.09831250e-01
-3.24335694e-01 -1.27675402e+00 1.06734169e+00 -2.36821011e-01
-1.09731495e+00 -1.14579713e+00 1.34457022e-01 7.84072220e-01
4.45881367e-01 -5.08714207e-02 1.44314361e+00 -1.22974956e+00
5.70070371e-03 1.31850038e-02 2.53583461e-01 -2.94665545e-01
9.86652672e-02 -7.24042673e-03 -9.23755884e-01 3.10452580e-02
3.76192741e-02 -3.44229311e-01 1.92639641e-02 -2.36354366e-01
1.68354623e-02 -6.73613548e-01 -5.73278852e-02 3.72819245e-01
9.01417375e-01 -5.00098756e-03 5.23789644e-01 7.93141127e-01
5.32549560e-01 1.16690159e+00 7.94847131e-01 2.31836453e-01
8.19410682e-01 1.00505173e+00 -1.03161475e-02 1.65338784e-01
-1.79822877e-01 -8.25967863e-02 5.42756021e-01 1.17234921e+00
-1.18698910e-01 -3.22965905e-02 -1.09080267e+00 7.22094357e-01
-1.49730539e+00 -6.25997484e-01 -5.38352609e-01 2.19210029e+00
1.60753942e+00 -4.74292822e-02 1.62042409e-01 2.77514625e-02
7.58347929e-01 -1.44286260e-01 7.86213577e-02 -9.09945548e-01
-3.36817294e-01 1.29414991e-01 3.79102677e-01 8.47021818e-01
-8.27658355e-01 1.49318218e+00 6.34622478e+00 8.83363903e-01
-8.27133358e-01 1.69898301e-01 5.19017756e-01 2.12173730e-01
-5.63641667e-01 9.35458615e-02 -9.17949677e-01 1.95138946e-01
1.14554965e+00 -1.76601455e-01 5.44725060e-01 3.80907178e-01
3.13146025e-01 2.07479559e-02 -1.41829813e+00 7.50170529e-01
5.15528955e-02 -9.19360042e-01 7.03416467e-02 1.97336107e-01
5.14699757e-01 -6.70704693e-02 -7.96210840e-02 3.90031606e-01
2.96519369e-01 -1.07406807e+00 8.86728168e-01 3.48316692e-02
1.04870713e+00 -7.90999234e-01 8.29038560e-01 2.54948586e-01
-6.56849682e-01 3.87808174e-01 -3.04124147e-01 -1.03241928e-01
-1.13945967e-02 -1.00111261e-01 -1.42245114e+00 2.68165529e-01
3.89291316e-01 2.14911446e-01 -4.69776988e-01 4.95011091e-01
-4.09456283e-01 7.39590108e-01 -1.85555547e-01 -2.47253403e-01
1.73618212e-01 -5.03697455e-01 8.84392142e-01 1.60775399e+00
2.30131283e-01 -3.67313460e-03 -3.12758125e-02 4.57577407e-01
-1.19270913e-01 8.52220714e-01 -4.71683919e-01 -3.51287611e-02
6.40645325e-01 1.40002656e+00 -6.08945310e-01 -3.29698324e-01
-3.80492836e-01 8.62994134e-01 1.83334842e-01 4.30282429e-02
-3.70313585e-01 -8.15487579e-02 7.17378497e-01 1.84220791e-01
-1.00673981e-01 -1.83706179e-01 -6.36721313e-01 -8.99843812e-01
1.44236302e-02 -1.56936693e+00 2.60070950e-01 -3.61755788e-01
-1.12700117e+00 7.29942620e-01 -1.85759768e-01 -9.25208867e-01
-6.82018757e-01 -5.21184385e-01 -2.83743143e-01 1.18175709e+00
-1.12080550e+00 -1.58003151e+00 3.76233131e-01 6.20815694e-01
8.38298678e-01 -5.19471824e-01 1.18653059e+00 3.34081948e-01
-1.87186494e-01 9.01348948e-01 1.02227076e-03 2.45522007e-01
1.28677237e+00 -1.55664647e+00 5.31598032e-01 4.56483632e-01
1.52939811e-01 5.63698351e-01 1.09307027e+00 -6.21518254e-01
-1.59448969e+00 -8.78210068e-01 2.08813143e+00 -1.05689240e+00
8.79458904e-01 -6.95390999e-01 -7.48763323e-01 6.85751855e-01
6.06766045e-01 -6.56659186e-01 9.11801398e-01 4.74090427e-01
-2.10468531e-01 1.23441003e-01 -1.36317277e+00 8.36656690e-01
9.75453556e-01 -7.02326119e-01 -8.17078769e-01 5.48782527e-01
4.62170392e-01 -6.78569674e-01 -1.22205198e+00 8.64541829e-02
5.01078248e-01 -3.56919169e-01 5.78898489e-01 -7.19621599e-01
3.09333384e-01 7.36229792e-02 -3.23822647e-01 -1.49992347e+00
-1.36377558e-01 -1.14677119e+00 5.44532835e-01 1.81895435e+00
9.99749720e-01 -4.25671816e-01 5.12896180e-01 9.05036867e-01
-1.42798156e-01 -3.29130918e-01 -1.09468770e+00 -4.09279138e-01
5.22416472e-01 -4.80531991e-01 8.13278913e-01 1.23590696e+00
2.78810114e-01 7.86953568e-01 -2.61506736e-01 -1.41454697e-01
3.14826369e-01 -1.14233106e-01 8.54658186e-01 -1.18932199e+00
-1.48838431e-01 -3.49229366e-01 -2.02572480e-01 -3.72451752e-01
4.23283368e-01 -9.86856282e-01 1.64434523e-03 -1.25181043e+00
2.17915088e-01 -5.20842135e-01 5.83882272e-01 5.19168437e-01
-1.47979170e-01 4.34343517e-01 1.32970497e-01 3.53863716e-01
-2.43120432e-01 2.45885089e-01 1.08828020e+00 -1.54973909e-01
-2.13592410e-01 2.21837550e-01 -6.55811429e-01 6.13274872e-01
9.53317642e-01 -7.57279336e-01 -3.21157509e-03 -5.56905746e-01
4.65148836e-01 1.18230402e-01 -1.92888245e-01 -2.84899950e-01
1.61287338e-01 -4.32605386e-01 6.58230186e-02 -1.52789623e-01
3.30233246e-01 -7.04029381e-01 -5.17889066e-03 2.59591639e-01
-4.41194415e-01 6.08345747e-01 1.07348591e-01 -1.47817582e-01
-1.49666190e-01 -3.94207031e-01 3.00193131e-01 -3.87858629e-01
-2.32972488e-01 -4.71644923e-02 -8.51580679e-01 4.57657754e-01
6.05548978e-01 -2.24335358e-01 -2.11207479e-01 -3.71040612e-01
-6.66640520e-01 7.57244080e-02 7.41770804e-01 2.77196020e-01
2.21705530e-02 -1.24217927e+00 -1.19411647e+00 6.60113245e-02
3.19988996e-01 -5.09080172e-01 -5.09089828e-01 9.09763694e-01
-5.10271549e-01 2.97017306e-01 -2.13384673e-01 -4.01289195e-01
-1.56120241e+00 4.60715406e-02 5.57434745e-02 -2.99205840e-01
-2.12162316e-01 5.83703101e-01 -4.96615380e-01 -1.22879589e+00
2.79975291e-02 2.39901468e-02 -2.53113687e-01 4.05287325e-01
1.91657975e-01 4.19789076e-01 4.66108859e-01 -1.25349903e+00
-3.82395178e-01 3.80583316e-01 -7.76394978e-02 -8.30569088e-01
9.97734785e-01 -5.47532678e-01 -5.47477782e-01 5.53285718e-01
9.03033078e-01 8.15915704e-01 -7.08863735e-01 -3.51807058e-01
3.26543391e-01 -4.33402866e-01 -4.61402625e-01 -1.07102382e+00
-3.88958365e-01 6.60806060e-01 9.76295769e-02 -4.28117476e-02
6.71231210e-01 7.40130469e-02 6.07688248e-01 5.35206437e-01
3.57728988e-01 -1.56086659e+00 -5.83426356e-01 1.05886674e+00
8.52429986e-01 -1.38181365e+00 -2.33508423e-01 -5.25617480e-01
-1.04265642e+00 1.11051524e+00 3.68746489e-01 6.18719339e-01
-1.38209045e-01 2.94216782e-01 7.96358168e-01 1.55437570e-02
-7.60702252e-01 -4.38462012e-02 9.09134522e-02 7.18414128e-01
9.59996164e-01 3.07802141e-01 -8.95539284e-01 4.03366417e-01
-1.16580486e+00 -6.29500389e-01 6.42150223e-01 6.47594810e-01
1.07230231e-01 -1.93273091e+00 -5.21000624e-01 2.09017918e-01
-1.08603907e+00 -4.67751712e-01 -8.14192057e-01 1.11156011e+00
7.30624273e-02 1.19651973e+00 1.26449406e-01 -1.90847263e-01
3.89755011e-01 5.49798369e-01 4.69153523e-01 -7.57395983e-01
-1.24111748e+00 2.72569209e-01 1.06535697e+00 -1.05790675e-01
-4.47234184e-01 -1.36665773e+00 -9.14013982e-01 -5.08892059e-01
-1.42713279e-01 3.98701310e-01 7.70950079e-01 1.00549757e+00
-1.44935802e-01 -8.28982294e-02 4.68500376e-01 -2.45501891e-01
-6.64028764e-01 -1.27954853e+00 -3.19048733e-01 3.85792851e-01
-1.02644585e-01 -1.73442885e-01 -2.40764096e-01 4.81153354e-02] | [11.429649353027344, 10.188562393188477] |
2651403a-fb39-441e-902c-e91f7d017ca7 | ad-nev-a-scalable-multi-level-neuroevolution | 2305.16497 | null | https://arxiv.org/abs/2305.16497v1 | https://arxiv.org/pdf/2305.16497v1.pdf | AD-NEV: A Scalable Multi-level Neuroevolution Framework for Multivariate Anomaly Detection | Anomaly detection tools and methods present a key capability in modern cyberphysical and failure prediction systems. Despite the fast-paced development in deep learning architectures for anomaly detection, model optimization for a given dataset is a cumbersome and time consuming process. Neuroevolution could be an effective and efficient solution to this problem, as a fully automated search method for learning optimal neural networks, supporting both gradient and non-gradient fine tuning. However, existing methods mostly focus on optimizing model architectures without taking into account feature subspaces and model weights. In this work, we propose Anomaly Detection Neuroevolution (AD-NEv) - a scalable multi-level optimized neuroevolution framework for multivariate time series anomaly detection. The method represents a novel approach to synergically: i) optimize feature subspaces for an ensemble model based on the bagging technique; ii) optimize the model architecture of single anomaly detection models; iii) perform non-gradient fine-tuning of network weights. An extensive experimental evaluation on widely adopted multivariate anomaly detection benchmark datasets shows that the models extracted by AD-NEv outperform well-known deep learning architectures for anomaly detection. Moreover, results show that AD-NEv can perform the whole process efficiently, presenting high scalability when multiple GPUs are available. | ['Roberto Corizzo', 'Kamil Faber', 'Dominik Zurek', 'Marcin Pietron'] | 2023-05-25 | null | null | null | null | ['time-series-anomaly-detection'] | ['time-series'] | [-1.27337929e-02 -5.70422053e-01 4.27620292e-01 -5.54904155e-02
-1.01119377e-01 -3.02122414e-01 3.35488975e-01 3.11793655e-01
-4.01149422e-01 5.25582433e-01 -5.67900300e-01 -3.30921024e-01
-4.38348711e-01 -7.83159971e-01 -3.79223287e-01 -9.37841356e-01
-3.26155245e-01 6.32538378e-01 1.78443447e-01 -4.20812041e-01
5.58661222e-01 9.61241663e-01 -2.01534986e+00 -2.94520825e-01
1.12506604e+00 1.11948824e+00 -2.87909985e-01 7.07683146e-01
-1.71502307e-01 2.81056106e-01 -6.69110179e-01 -7.54080266e-02
4.21960533e-01 -4.80700940e-01 -2.43436560e-01 -1.36747256e-01
2.88237035e-01 -8.67640153e-02 1.12417221e-01 1.10830712e+00
6.21782601e-01 3.60494524e-01 6.41430736e-01 -1.46847558e+00
-2.27792382e-01 1.53611958e-01 -5.20601273e-01 5.06923497e-01
-1.20365985e-01 3.40669215e-01 6.69623256e-01 -6.62125766e-01
6.63278997e-02 9.48447883e-01 7.72888660e-01 6.03837848e-01
-1.18941939e+00 -5.67779839e-01 8.49142000e-02 6.95305109e-01
-1.26023149e+00 -4.90441471e-02 9.50865448e-01 -4.25358117e-01
1.32430482e+00 5.41430354e-01 9.98293519e-01 8.17643940e-01
6.33985639e-01 4.14723784e-01 6.97014391e-01 -4.25951183e-01
7.27462053e-01 -1.76032424e-01 4.17799592e-01 7.60671616e-01
5.12324870e-01 2.05485374e-01 -4.18173492e-01 -5.03618419e-01
5.40747464e-01 4.11187708e-02 -3.36861610e-02 -2.66551107e-01
-7.30587602e-01 7.48371482e-01 1.94323942e-01 3.47373992e-01
-8.79446387e-01 1.67208195e-01 7.71303713e-01 5.13028920e-01
4.05244708e-01 8.93378139e-01 -4.58973438e-01 -3.99782360e-01
-7.27005243e-01 4.31822240e-01 6.91560090e-01 9.36544761e-02
5.51752508e-01 8.00126910e-01 -2.22460687e-01 7.79491901e-01
1.51862934e-01 2.79790431e-01 9.24279690e-01 -4.96371806e-01
-9.61996987e-02 1.07793069e+00 -1.32533059e-01 -1.14686203e+00
-6.21230483e-01 -6.56654954e-01 -9.65751231e-01 7.90161610e-01
2.67824054e-01 -5.19116372e-02 -9.25058126e-01 1.40576494e+00
6.79983795e-01 4.93443131e-01 -6.45265430e-02 7.14173973e-01
5.74905127e-02 4.00238574e-01 6.25141561e-02 -2.48577178e-01
1.08893263e+00 -8.54170799e-01 -7.07791686e-01 -3.27675864e-02
7.19810665e-01 -2.98558891e-01 8.64630163e-01 7.95892954e-01
-7.36099958e-01 -4.24538285e-01 -1.35116971e+00 5.64081192e-01
-5.73077738e-01 -1.08190691e-02 5.37426829e-01 9.87696171e-01
-9.95244026e-01 8.57691526e-01 -1.13864529e+00 -2.84786820e-01
3.14347148e-01 6.14327192e-01 -1.03854001e-01 4.63684976e-01
-7.24769890e-01 8.21752131e-01 8.53708267e-01 4.07610506e-01
-6.59715891e-01 -5.09811997e-01 -4.55636650e-01 1.86038822e-01
2.55038738e-01 -5.85503042e-01 7.92822719e-01 -1.08948815e+00
-1.68899620e+00 4.34436589e-01 5.93983717e-02 -7.92077720e-01
2.69298106e-01 -4.19308662e-01 -5.57262897e-01 -2.22597286e-01
-6.18287742e-01 6.80044815e-02 1.14225686e+00 -8.72700095e-01
-3.64008158e-01 -5.84051311e-01 -5.17325222e-01 -2.82184090e-02
-9.15380478e-01 -6.71605542e-02 1.77054435e-01 -7.95051098e-01
3.51828337e-01 -8.05373609e-01 -3.71824116e-01 -3.99443805e-02
-3.62588614e-01 -2.75404453e-01 1.26519299e+00 -7.55782723e-01
1.44360304e+00 -1.95155466e+00 3.56592059e-01 6.50509655e-01
5.82068861e-02 8.28627527e-01 3.13431881e-02 2.14160070e-01
-2.83274680e-01 -2.42650360e-01 -6.53697848e-01 -2.94340312e-01
-6.92457110e-02 4.16201621e-01 -2.90908366e-01 3.79551083e-01
3.10041457e-01 6.03124201e-01 -5.26346862e-01 -2.31758177e-01
3.68851304e-01 2.92169303e-01 -7.63472974e-01 3.52295011e-01
-2.96683967e-01 3.65899980e-01 -3.76393706e-01 7.90457070e-01
5.86361170e-01 1.22132532e-01 -1.38839588e-01 1.26369238e-01
1.22874891e-02 -5.33100188e-01 -1.37112010e+00 1.35691667e+00
-2.32026964e-01 4.09108043e-01 -1.04750238e-01 -1.26637781e+00
1.28341424e+00 9.15755555e-02 5.49243748e-01 -5.72663426e-01
1.86927274e-01 5.99047422e-01 2.74972171e-01 -5.77035546e-01
3.97297382e-01 1.98707864e-01 1.71313360e-01 5.78381002e-01
7.22536966e-02 1.56432942e-01 2.25791007e-01 -4.70318198e-01
1.13768733e+00 3.05632446e-02 4.11604971e-01 -1.47545949e-01
1.07607412e+00 -1.74635977e-01 7.34809160e-01 8.63110304e-01
-2.91588873e-01 2.89823443e-01 3.70299608e-01 -9.55218852e-01
-8.89478028e-01 -8.71937215e-01 -3.23940329e-02 9.32272971e-01
-2.59986043e-01 -1.12792410e-01 -8.82200301e-01 -7.66029060e-01
-7.55306333e-02 9.45448458e-01 -5.22129953e-01 -5.96901476e-01
-7.35897779e-01 -1.09217858e+00 6.53511405e-01 4.83547658e-01
4.70161617e-01 -1.33456194e+00 -9.26787317e-01 3.51252764e-01
3.85745317e-01 -4.07271087e-01 2.72866845e-01 2.42004216e-01
-1.21640861e+00 -8.86864543e-01 -3.36781770e-01 -4.27407175e-01
5.90874553e-01 -4.21225935e-01 9.89902735e-01 5.70018291e-01
-8.02119136e-01 2.84595639e-01 -2.46868104e-01 -6.99720442e-01
-4.53863263e-01 -9.14457515e-02 3.73852462e-01 2.02713534e-01
5.95234334e-01 -9.18734193e-01 -3.92364651e-01 1.37414888e-01
-9.84360874e-01 -5.12886882e-01 4.66141641e-01 1.14453042e+00
6.21765852e-01 8.65876302e-02 7.53096938e-01 -3.73560399e-01
8.09080303e-01 -6.06315613e-01 -9.80153799e-01 3.98604423e-01
-9.93850350e-01 1.42289236e-01 7.55959392e-01 -3.50555778e-01
-8.04267228e-01 -2.28071343e-02 -3.62114131e-01 -6.56777620e-01
-3.98749501e-01 3.01507592e-01 1.93220511e-01 -3.44141990e-01
8.18934739e-01 5.18694639e-01 2.12130174e-01 -4.81501848e-01
-1.08981103e-01 3.76907796e-01 4.87211525e-01 -7.30820179e-01
7.27699220e-01 1.37062848e-01 4.08912003e-01 -8.64218235e-01
-2.88126945e-01 -2.94872344e-01 -5.67707956e-01 -4.29082185e-01
7.61065960e-01 -1.16117828e-01 -6.67384148e-01 9.30464089e-01
-9.84399498e-01 2.14517713e-02 -3.99003923e-01 1.81860805e-01
-2.81542152e-01 5.12434721e-01 -1.49567887e-01 -1.16342580e+00
-7.56641686e-01 -1.11095607e+00 6.58324003e-01 3.25459868e-01
-1.37617588e-01 -1.02246654e+00 3.51391673e-01 -1.04088664e-01
7.74357378e-01 5.35255492e-01 1.04888999e+00 -1.34932911e+00
-2.15435430e-01 -6.20745301e-01 2.70079762e-01 6.14105046e-01
-1.07462145e-01 2.60585606e-01 -8.66878450e-01 -3.74036163e-01
2.64744550e-01 -8.34919885e-02 4.76614416e-01 2.71456033e-01
1.62052870e+00 -2.27931082e-01 -1.28501460e-01 6.55627966e-01
1.33802843e+00 3.34819883e-01 5.13852477e-01 7.13057995e-01
6.95878923e-01 3.90135527e-01 4.32341456e-01 6.77192688e-01
-2.19498232e-01 7.08711565e-01 7.84277797e-01 2.42652968e-01
3.43053728e-01 3.31012249e-01 2.17045471e-01 8.48867774e-01
-1.77923858e-01 -1.53294101e-01 -1.20899928e+00 2.66236573e-01
-2.02729821e+00 -8.83491099e-01 -1.68870464e-01 2.26717472e+00
2.77039379e-01 1.20244637e-01 1.86899066e-01 6.80636346e-01
5.19552827e-01 -2.09051564e-01 -9.15154099e-01 -7.96341658e-01
-1.76351473e-01 4.15809631e-01 -8.13472122e-02 2.70216838e-02
-9.82029319e-01 4.66330320e-01 5.56535816e+00 7.75527239e-01
-1.31203818e+00 -1.03502117e-01 3.75380605e-01 -3.34960580e-01
4.67629991e-02 -2.51868635e-01 -4.62113112e-01 5.75379014e-01
1.13571084e+00 -7.72599801e-02 3.62160623e-01 1.07734621e+00
-2.27720402e-02 5.88296130e-02 -8.84427309e-01 1.04816473e+00
7.57203922e-02 -1.14769423e+00 2.20995560e-01 6.87722638e-02
7.21819758e-01 -1.23479113e-01 7.56224967e-04 3.83009076e-01
-2.46760488e-01 -9.24716711e-01 4.02361900e-01 6.90023959e-01
-7.86293447e-02 -1.06244433e+00 1.00222719e+00 3.40673923e-01
-8.58559847e-01 -7.03237772e-01 -3.07632297e-01 8.33974704e-02
8.59792084e-02 5.80603123e-01 -5.43586075e-01 6.68291271e-01
8.41724694e-01 3.04032236e-01 -8.56879175e-01 1.41827071e+00
1.65621191e-01 6.38489604e-01 -4.86574054e-01 -1.38755769e-01
1.71764314e-01 -5.00624716e-01 1.10928524e+00 8.81037772e-01
6.51057065e-01 -3.52486432e-01 -2.02274863e-02 7.95115888e-01
5.66368341e-01 2.12566972e-01 -4.98951375e-01 -1.73273623e-01
3.73796225e-01 1.11459756e+00 -8.41874301e-01 -5.16730845e-02
9.63691324e-02 8.99712265e-01 3.19866896e-01 1.15482435e-01
-7.99051702e-01 -5.41621089e-01 8.45099509e-01 -3.01015526e-01
1.35656148e-01 -2.15075731e-01 -5.71345508e-01 -9.22754109e-01
1.91779677e-02 -1.10875595e+00 7.33804345e-01 -2.62580305e-01
-1.25095594e+00 7.95620739e-01 -1.90982014e-01 -1.31541848e+00
-6.36520147e-01 -8.13465178e-01 -1.00194812e+00 7.67638206e-01
-9.86027062e-01 -8.12600434e-01 -4.48654354e-01 5.28620005e-01
5.41453481e-01 -7.75013864e-01 1.11155510e+00 1.14165321e-01
-1.19058132e+00 7.10866213e-01 3.63957793e-01 -3.97804022e-01
2.02238530e-01 -1.27031553e+00 2.54367948e-01 1.22805560e+00
8.11929852e-02 4.84705865e-01 9.00135398e-01 -6.67018116e-01
-1.35685563e+00 -8.14133823e-01 2.41992176e-01 -4.21747267e-02
5.85338175e-01 -6.45580888e-02 -1.48305357e+00 2.66622394e-01
-4.11689915e-02 2.35230718e-02 8.40830684e-01 1.03164814e-01
3.47538702e-02 -2.56324857e-01 -1.35213363e+00 6.35818362e-01
6.23381615e-01 -6.15577474e-02 -3.80402595e-01 2.37511769e-01
3.26559603e-01 -3.64138842e-01 -6.45578980e-01 5.82030892e-01
3.38298470e-01 -1.20402813e+00 9.36170757e-01 -6.77783012e-01
5.72296903e-02 -4.23217952e-01 7.20508844e-02 -1.30812752e+00
-9.19557065e-02 -5.64290881e-01 -8.60428452e-01 1.02131248e+00
-9.27282963e-03 -1.12032080e+00 7.13341236e-01 6.01235449e-01
-2.79222131e-01 -1.18396115e+00 -1.27751172e+00 -9.56502259e-01
-1.85695022e-01 -5.22636473e-01 8.17304015e-01 8.27716589e-01
-5.13134837e-01 -3.75514686e-01 -1.79969519e-01 3.11855733e-01
7.41207957e-01 -2.04932734e-01 6.48034096e-01 -1.59499943e+00
-4.63987529e-01 -9.55934107e-01 -9.38492835e-01 -3.98453977e-03
1.65855244e-01 -5.77221751e-01 -1.88007578e-01 -9.13861394e-01
-3.72643620e-01 -2.09623247e-01 -7.27223516e-01 5.26625872e-01
-2.69106239e-01 4.61134650e-02 -2.26771846e-01 1.98681261e-02
-2.84540087e-01 8.51449728e-01 5.11876643e-01 1.81261614e-01
-3.38594198e-01 4.85207774e-02 -9.98509899e-02 7.48600900e-01
1.02078688e+00 -5.31644762e-01 -3.77340615e-01 -1.54973909e-01
5.33107519e-02 -3.20403993e-01 5.08087516e-01 -1.60151863e+00
3.24385136e-01 -2.49090139e-02 4.45616126e-01 -5.63666999e-01
2.59377480e-01 -9.29630101e-01 1.04767881e-01 8.98970306e-01
7.16130584e-02 6.70451581e-01 7.08227336e-01 7.65392959e-01
-2.13144705e-01 -4.65382397e-01 5.76376915e-01 2.60115027e-01
-9.67838049e-01 1.67783037e-01 -4.85606492e-01 -3.51296276e-01
1.28883374e+00 -5.65945625e-01 -9.34652686e-02 1.93672612e-01
-6.28206730e-01 1.27785400e-01 2.42830575e-01 5.41248083e-01
6.71281815e-01 -1.21867454e+00 -5.83743572e-01 6.21461868e-01
-4.63143103e-02 -1.26220077e-01 4.90792960e-01 7.19157994e-01
-7.04993784e-01 6.29160255e-02 -7.65729070e-01 -8.21682334e-01
-1.30225861e+00 3.41327548e-01 6.88021481e-01 -3.96982580e-01
-5.27232468e-01 8.79032612e-01 -4.36976016e-01 -4.74162519e-01
2.59211749e-01 2.36896917e-01 -3.52121502e-01 -2.74246261e-02
4.98148799e-01 8.58019769e-01 3.89148623e-01 -2.93731183e-01
-3.82510126e-01 4.81713653e-01 -1.34455770e-01 2.16651425e-01
1.52274799e+00 2.86033839e-01 -5.30965805e-01 5.33097565e-01
6.99537575e-01 -6.37747526e-01 -9.27863538e-01 9.80569348e-02
3.46590549e-01 -3.78863573e-01 2.28771120e-01 -6.98810875e-01
-1.04124165e+00 7.64843822e-01 1.20455217e+00 3.09430808e-01
1.59092045e+00 -7.68920541e-01 5.46368718e-01 8.26307774e-01
1.92991912e-01 -1.20623362e+00 3.15169722e-01 5.72361290e-01
1.04123342e+00 -9.99899745e-01 -3.29884179e-02 2.32150242e-01
-3.19909632e-01 1.47250247e+00 1.23467994e+00 -3.34100157e-01
5.68178117e-01 5.72889373e-02 -5.36926882e-03 -3.86642575e-01
-6.16079390e-01 3.08094054e-01 5.86917698e-01 4.74359781e-01
1.71212599e-01 -2.53740132e-01 -2.60738403e-01 4.09828365e-01
6.66878652e-03 -3.35823059e-01 1.70573816e-01 1.11069012e+00
-3.73678684e-01 -1.09273231e+00 -6.20365798e-01 6.98103189e-01
-2.30767310e-01 2.46068195e-01 -1.18853673e-01 4.83276397e-01
4.06236611e-02 6.02945328e-01 1.71807051e-01 -5.04214108e-01
3.19457144e-01 4.35822517e-01 2.56183445e-01 -1.43963292e-01
-1.03824782e+00 -4.55338269e-01 -3.46441567e-01 -6.03168249e-01
1.61885679e-01 -6.08045518e-01 -1.10787249e+00 -1.45293638e-01
-3.36133689e-01 6.01182655e-02 9.44887757e-01 1.09576142e+00
5.77568531e-01 8.18356037e-01 4.38342035e-01 -1.05046523e+00
-9.85739708e-01 -8.82313192e-01 -4.79785413e-01 2.37371936e-01
2.23243013e-01 -8.28983486e-01 -4.95518744e-01 -4.68863785e-01] | [7.6291584968566895, 2.51727032661438] |
55cd2b45-7353-45b9-8bba-c7544ce92223 | architecturing-binarized-neural-networks-for | 2303.15005 | null | https://arxiv.org/abs/2303.15005v1 | https://arxiv.org/pdf/2303.15005v1.pdf | Architecturing Binarized Neural Networks for Traffic Sign Recognition | Traffic signs support road safety and managing the flow of traffic, hence are an integral part of any vision system for autonomous driving. While the use of deep learning is well-known in traffic signs classification due to the high accuracy results obtained using convolutional neural networks (CNNs) (state of the art is 99.46\%), little is known about binarized neural networks (BNNs). Compared to CNNs, BNNs reduce the model size and simplify convolution operations and have shown promising results in computationally limited and energy-constrained devices which appear in the context of autonomous driving. This work presents a bottom-up approach for architecturing BNNs by studying characteristics of the constituent layers. These constituent layers (binarized convolutional layers, max pooling, batch normalization, fully connected layers) are studied in various combinations and with different values of kernel size, number of filters and of neurons by using the German Traffic Sign Recognition Benchmark (GTSRB) for training. As a result, we propose BNNs architectures which achieve more than $90\%$ for GTSRB (the maximum is $96.45\%$) and an average greater than $80\%$ (the maximum is $88.99\%$) considering also the Belgian and Chinese datasets for testing. The number of parameters of these architectures varies from 100k to less than 2M. The accompanying material of this paper is publicly available at https://github.com/apostovan21/BinarizedNeuralNetwork. | ['Mădălina Eraşcu', 'Andreea Postovan'] | 2023-03-27 | null | null | null | null | ['traffic-sign-recognition'] | ['computer-vision'] | [ 1.64724495e-02 7.62906224e-02 -1.60232589e-01 -4.58994895e-01
1.45441275e-02 -2.31483430e-01 7.16594040e-01 -3.96266818e-01
-9.10709083e-01 6.89674795e-01 -4.63118345e-01 -7.85570741e-01
-2.21514210e-01 -8.70331705e-01 -7.16862142e-01 -7.30316460e-01
9.29705426e-02 3.02493908e-02 5.35639882e-01 -3.39042515e-01
1.61315769e-01 1.02563608e+00 -2.00973535e+00 1.14015862e-01
7.47339010e-01 1.38665342e+00 4.55810241e-02 7.73910820e-01
-2.40881573e-02 6.64502680e-01 -5.96756637e-01 -6.35799289e-01
4.90721285e-01 -5.35147116e-02 -4.33296353e-01 -3.77480775e-01
8.18951368e-01 -2.80195445e-01 -6.69858098e-01 9.93375063e-01
5.17762423e-01 9.77392346e-02 5.49668491e-01 -1.50191927e+00
-2.55053550e-01 2.09623680e-01 -8.00709501e-02 4.80958015e-01
-6.52147055e-01 5.18954813e-01 8.56188655e-01 -6.02098703e-01
4.56692427e-01 9.20988321e-01 4.54544365e-01 6.40246749e-01
-8.79403830e-01 -9.79407489e-01 -1.94438666e-01 7.39166260e-01
-1.19821525e+00 -7.00441301e-01 5.82670271e-01 -3.39489996e-01
1.29310572e+00 2.96712309e-01 6.92510724e-01 7.88517952e-01
1.14590563e-01 6.07356131e-01 1.11843896e+00 -4.13599551e-01
1.55893296e-01 3.03314537e-01 4.50329304e-01 7.31794238e-01
7.97155023e-01 1.95558161e-01 -3.71524066e-01 4.13418502e-01
4.61233437e-01 -2.67788261e-01 2.32674628e-01 -5.02233431e-02
-9.27924991e-01 6.58172727e-01 6.72876179e-01 3.94269317e-01
-4.55629498e-01 7.06015468e-01 5.06149948e-01 2.35431626e-01
-1.48513064e-01 1.19505763e-01 -4.46473986e-01 -3.26710612e-01
-7.93949246e-01 2.76234329e-01 6.01126194e-01 8.90312374e-01
7.28207946e-01 4.95883048e-01 -8.11287165e-02 9.06096995e-01
1.19862914e-01 8.23152125e-01 3.32073510e-01 -7.96599030e-01
5.22914112e-01 7.89858818e-01 -1.12063289e-01 -7.53542006e-01
-4.79643494e-01 -1.57829329e-01 -9.32059824e-01 7.66588449e-01
7.16019094e-01 -1.89547688e-01 -1.32667565e+00 1.28228009e+00
-1.53967664e-01 -1.97748411e-02 -5.51831797e-02 6.21919036e-01
8.65353584e-01 4.97797638e-01 9.94500965e-02 5.23362637e-01
1.56168485e+00 -8.25390935e-01 -5.13982594e-01 -3.56634617e-01
5.28040051e-01 -4.78241175e-01 7.94391632e-01 4.20869589e-01
-1.06983316e+00 -8.55432749e-01 -1.27267182e+00 -1.02824263e-01
-1.05753672e+00 4.98378992e-01 6.37962937e-01 1.39070892e+00
-1.16879869e+00 4.15190578e-01 -6.86670065e-01 -3.19530547e-01
8.08301747e-01 7.63929486e-01 -1.67431846e-01 -1.16339892e-01
-1.04339468e+00 1.16325355e+00 3.86546552e-01 6.13533437e-01
-4.62716818e-01 -3.14463049e-01 -6.22779012e-01 1.54543761e-02
3.39347906e-02 -2.59956986e-01 1.05498970e+00 -7.40546227e-01
-1.39911830e+00 8.73946846e-01 2.60594320e-02 -9.03060913e-01
4.70683306e-01 2.30825424e-01 -7.04607427e-01 -6.59891739e-02
-5.50009489e-01 1.16580832e+00 6.52231097e-01 -6.04273200e-01
-9.10355508e-01 6.19783662e-02 1.26369759e-01 -3.77269953e-01
-2.00550154e-01 7.09920153e-02 -2.53447413e-01 -2.81015515e-01
-1.33495137e-01 -9.98492539e-01 -1.31113991e-01 1.91566929e-01
-1.30078584e-01 -2.56837994e-01 1.06331933e+00 -5.97393870e-01
1.02380764e+00 -1.89138603e+00 -5.48873901e-01 4.30865169e-01
5.40224835e-02 9.18653548e-01 -4.79811393e-02 -1.58683747e-01
-7.52075166e-02 -4.38572392e-02 -1.59773424e-01 -4.32650223e-02
3.05910319e-01 4.72631902e-01 1.30133107e-01 4.38447028e-01
3.06773126e-01 1.22614932e+00 -3.22079301e-01 -2.60928810e-01
7.79193044e-01 5.85026443e-01 -3.05763721e-01 -4.14957941e-01
2.89478660e-01 -1.62224308e-01 -2.46773496e-01 7.94496894e-01
8.44567537e-01 2.36508250e-01 -1.90916672e-01 -2.01592326e-01
-3.97814542e-01 1.70989960e-01 -1.20583367e+00 1.15028942e+00
-3.19317073e-01 1.39650691e+00 5.30136302e-02 -1.25874758e+00
1.20601606e+00 1.28471419e-01 1.83051974e-01 -1.26421046e+00
5.26132941e-01 6.18457675e-01 5.60227275e-01 -4.66903925e-01
5.06065667e-01 8.03384036e-02 5.14046736e-02 -6.63198829e-02
1.72051743e-01 8.99181888e-02 6.68875575e-01 -1.98910072e-01
1.10016775e+00 -2.63979048e-01 -1.10447273e-01 -3.15917879e-01
7.45740891e-01 -8.40980709e-02 3.14142734e-01 5.35830200e-01
-5.99482596e-01 1.52027503e-01 5.78037381e-01 -5.96801698e-01
-1.27786136e+00 -7.77148724e-01 -4.05597746e-01 5.96863210e-01
-2.64520980e-02 2.11243391e-01 -8.10159743e-01 -3.58817369e-01
7.48391002e-02 6.99623406e-01 -5.01846433e-01 -2.06865236e-01
-9.58259404e-01 -7.03796446e-01 9.95004296e-01 8.44986439e-01
1.09137690e+00 -1.40509129e+00 -1.21117866e+00 -6.06523529e-02
2.63754785e-01 -1.26075613e+00 2.28675768e-01 3.63918990e-01
-8.06241572e-01 -1.07205319e+00 -7.47654915e-01 -6.82376623e-01
6.09914482e-01 -1.03993662e-01 8.03016305e-01 9.77326632e-02
-5.85346222e-01 -1.24540076e-01 -7.00992346e-02 -8.17742765e-01
-1.81227103e-01 9.66833755e-02 -2.10008517e-01 7.28986040e-03
6.58002496e-01 -2.52828658e-01 -7.27687120e-01 3.82050902e-01
-6.77257240e-01 1.97490789e-02 9.38669145e-01 6.30241990e-01
1.80659652e-01 -5.35094328e-02 4.07545775e-01 -4.91753906e-01
3.13463002e-01 9.41962972e-02 -9.93405342e-01 -2.16054847e-03
-5.87915838e-01 6.89620711e-03 4.08889323e-01 -1.62326574e-01
-7.99168229e-01 -2.00995672e-02 -3.42123270e-01 -1.83823153e-01
-5.60795307e-01 -1.60501599e-01 -5.74350059e-02 -4.38354790e-01
5.37293673e-01 3.35261524e-02 2.80440804e-02 -2.94218123e-01
1.55957490e-01 7.51017988e-01 3.19009721e-01 -1.18066348e-01
8.29606056e-01 6.35516465e-01 3.13295186e-01 -9.51711416e-01
1.02174073e-01 -2.37848520e-01 -5.83783627e-01 -5.97095251e-01
7.95715749e-01 -5.29840052e-01 -1.15302610e+00 8.99798691e-01
-7.85862625e-01 -4.53606486e-01 -4.84709412e-01 5.96861541e-01
-3.03815782e-01 -6.82150796e-02 -2.59495527e-01 -8.99104774e-01
-3.19742680e-01 -1.22581398e+00 4.46932077e-01 5.71296990e-01
2.01538987e-02 -6.18466377e-01 -6.13329470e-01 5.30710161e-01
8.94333541e-01 2.23543674e-01 6.21744692e-01 -4.46783751e-01
-7.66633034e-01 -5.27450144e-01 -6.83944046e-01 8.41434479e-01
-2.51032472e-01 6.05343143e-03 -1.27592278e+00 1.72367856e-01
-6.11847878e-01 -5.63047007e-02 1.16473174e+00 7.44557977e-01
9.88006473e-01 -2.22232360e-02 -2.16294527e-01 4.70413506e-01
1.37087274e+00 6.48852229e-01 1.16370416e+00 5.39798737e-01
4.07379240e-01 5.21245182e-01 2.94757783e-01 8.55488628e-02
1.36817664e-01 5.55267155e-01 5.47748804e-01 -2.15894848e-01
-5.08780301e-01 2.26218805e-01 2.16685608e-01 2.28572205e-01
-4.22200710e-01 -2.23162144e-01 -1.04659486e+00 6.78473353e-01
-1.40829515e+00 -9.74387348e-01 -5.00034869e-01 1.88317716e+00
2.04737768e-01 6.57809138e-01 1.79577291e-01 7.21082389e-01
6.27514243e-01 6.02986626e-02 -5.19883990e-01 -1.06527126e+00
-2.73801506e-01 6.89730346e-01 1.27706742e+00 1.74618348e-01
-1.05550945e+00 7.81687677e-01 5.15183020e+00 8.97477567e-01
-1.38963401e+00 2.31778715e-02 5.69200039e-01 -1.88469246e-01
3.18062901e-01 -3.51078808e-01 -9.74630296e-01 4.96723711e-01
1.39349723e+00 2.80662537e-01 3.08510900e-01 7.18273640e-01
2.70238936e-01 -4.29916769e-01 -6.59468591e-01 1.04344201e+00
-2.14068457e-01 -1.44987082e+00 -2.56360382e-01 3.18123668e-01
6.06409967e-01 4.67404544e-01 1.49704903e-01 2.46764213e-01
3.73338051e-02 -1.08377266e+00 9.82016742e-01 5.45134127e-01
8.92857850e-01 -8.45449388e-01 1.03672898e+00 -1.48752732e-02
-1.18313396e+00 -4.67865050e-01 -4.17783678e-01 -1.13829590e-01
2.07649499e-01 5.33250570e-01 -4.73977983e-01 2.54202217e-01
9.61801946e-01 4.25240308e-01 -6.75339997e-01 1.25347292e+00
-1.22396238e-01 6.74379289e-01 -6.42275035e-01 -5.09089649e-01
5.47207534e-01 -1.21932529e-01 2.12533638e-01 1.45129418e+00
3.37312251e-01 -1.88685015e-01 -6.82171106e-01 7.73636162e-01
-9.70043465e-02 -1.35641739e-01 -4.03903902e-01 9.07603204e-02
1.50423929e-01 1.24596488e+00 -9.13424671e-01 -4.13614720e-01
-3.95325005e-01 4.12920058e-01 -2.93371379e-01 3.59158725e-01
-1.06816232e+00 -8.90585780e-01 7.99865603e-01 1.55753151e-01
5.78819335e-01 -2.30770558e-01 -7.71362901e-01 -5.20935535e-01
2.32137039e-01 -3.83273005e-01 9.63286981e-02 -7.58840084e-01
-5.60595095e-01 5.47794819e-01 -3.81931104e-02 -1.14740515e+00
2.06077352e-01 -1.39971840e+00 -5.78880548e-01 1.03996217e+00
-1.92016172e+00 -1.02161896e+00 -6.13157511e-01 3.50856692e-01
3.22434753e-01 -3.39036018e-01 4.35549229e-01 7.62399256e-01
-7.21454322e-01 7.82270908e-01 2.31904965e-02 3.87997031e-01
2.36892581e-01 -8.28568995e-01 6.72223508e-01 9.16452289e-01
-2.48506919e-01 1.60333052e-01 3.86096954e-01 -1.15725785e-01
-1.18349731e+00 -1.00505698e+00 1.16469276e+00 -9.63079855e-02
4.52858806e-01 -3.81869107e-01 -4.81196284e-01 2.58413792e-01
2.73075998e-01 2.24223256e-01 9.80590433e-02 -4.11214590e-01
-2.09880993e-01 -6.69197977e-01 -1.36071098e+00 5.37057817e-01
1.02318513e+00 -2.19748333e-01 -1.30698085e-01 -2.20920220e-01
-1.60164267e-01 -1.77723378e-01 -4.33684826e-01 3.91573280e-01
8.16161692e-01 -1.06755519e+00 9.20555294e-01 -2.71517485e-01
9.04648975e-02 -3.02044839e-01 1.05557346e-03 -7.55996406e-01
-4.10638787e-02 -1.63174435e-01 -1.97245553e-02 8.49491000e-01
6.44876540e-01 -1.06667054e+00 1.18581200e+00 7.77366102e-01
-4.62026924e-01 -7.47082174e-01 -1.37827754e+00 -9.52711642e-01
-5.31675816e-02 -7.02953219e-01 3.72583568e-01 3.60206068e-01
-4.69385356e-01 -1.77847937e-01 -2.17613056e-02 -2.61458576e-01
4.53573614e-01 -4.34440851e-01 7.19876587e-01 -1.29610264e+00
3.61576766e-01 -1.08020532e+00 -1.11051631e+00 -7.47654498e-01
-1.11938596e-01 -6.52728140e-01 -1.63463905e-01 -1.60090995e+00
-3.80354226e-01 -6.80183709e-01 -3.93792957e-01 7.58720517e-01
5.24609387e-01 6.32282674e-01 1.21122569e-01 -3.03278804e-01
-1.24374200e-02 1.36826402e-02 1.02089190e+00 -3.25685501e-01
2.06275448e-01 5.61498664e-02 -1.87276602e-01 5.82412064e-01
1.33302712e+00 -2.06172526e-01 -9.87046063e-02 -1.96320787e-01
-3.25454213e-02 -7.29715765e-01 8.04759860e-01 -1.38680720e+00
3.56654048e-01 9.78277922e-02 2.43121713e-01 -8.03076744e-01
5.39099872e-01 -9.92647529e-01 1.00908421e-01 1.01075661e+00
-1.39235090e-02 -7.84175843e-02 5.51908851e-01 8.69491920e-02
-2.55116522e-01 -2.59854585e-01 9.69559789e-01 2.02949211e-01
-1.25951231e+00 5.35003170e-02 -4.85247284e-01 -4.03117508e-01
1.09014976e+00 -1.10120630e+00 -5.16526878e-01 -2.77073681e-02
-4.12111610e-01 6.06152341e-02 -9.14281756e-02 4.34543908e-01
4.51270580e-01 -1.16355991e+00 -5.01906991e-01 4.92091507e-01
-1.30831962e-03 -2.96040326e-01 4.98850405e-01 1.01098275e+00
-9.87799227e-01 9.79955018e-01 -7.62169898e-01 -3.74725282e-01
-1.39962947e+00 3.60606872e-02 6.77388489e-01 7.05682188e-02
-4.86015797e-01 8.74461472e-01 -4.16239023e-01 -1.51576534e-01
3.71579140e-01 -7.13280141e-01 -3.08222830e-01 1.27850771e-01
2.80096561e-01 9.10907745e-01 5.05723000e-01 -7.48783290e-01
-5.50814629e-01 5.93974113e-01 2.47237578e-01 8.99886489e-02
1.34237576e+00 3.56889278e-01 1.77775115e-01 -8.91990140e-02
1.12532365e+00 -6.41643524e-01 -1.20475590e+00 1.59483507e-01
1.92638952e-02 -7.03830719e-02 1.84873432e-01 -8.44350398e-01
-1.51414764e+00 1.14770854e+00 9.96661067e-01 -2.99334712e-02
1.15509593e+00 -3.07602167e-01 8.53363812e-01 6.64495409e-01
6.68469295e-02 -1.44570756e+00 -5.24388134e-01 5.17812014e-01
5.14722705e-01 -1.08764553e+00 -2.28828356e-01 -8.02382901e-02
-3.05058539e-01 1.19528091e+00 6.23387575e-01 -2.72304416e-01
7.44530618e-01 3.92299592e-01 8.59972611e-02 -2.01420277e-01
-4.61323023e-01 -5.31644881e-01 3.13760996e-01 5.18167853e-01
1.67844191e-01 2.14184985e-01 -5.42037964e-01 1.91045985e-01
-2.34892234e-01 1.61440507e-01 3.12975734e-01 1.04025912e+00
-4.60680634e-01 -9.44713175e-01 -2.67534882e-01 6.93014026e-01
-2.93150693e-01 -9.78107285e-03 -2.43091881e-01 9.41944003e-01
7.21606433e-01 8.73584449e-01 2.67838597e-01 -4.01876688e-01
8.46082270e-01 2.20409542e-01 4.30387795e-01 1.13504946e-01
-7.72596598e-01 -4.77993011e-01 4.85032439e-01 -4.40656900e-01
-4.21947867e-01 -6.53801739e-01 -1.22110033e+00 -6.66006565e-01
-2.89106965e-01 -2.90306538e-01 1.25185311e+00 7.07624078e-01
1.31904274e-01 7.62160659e-01 1.33320093e-01 -9.12288368e-01
-5.97894974e-02 -9.35467064e-01 -4.89368111e-01 5.28017804e-02
1.11993693e-01 -6.69707835e-01 -1.56317234e-01 1.09493524e-01] | [7.996129035949707, -0.7704320549964905] |
782264f0-aa81-436d-8371-df1c55c9533a | networked-restless-bandits-with-positive | 2212.05144 | null | https://arxiv.org/abs/2212.05144v1 | https://arxiv.org/pdf/2212.05144v1.pdf | Networked Restless Bandits with Positive Externalities | Restless multi-armed bandits are often used to model budget-constrained resource allocation tasks where receipt of the resource is associated with an increased probability of a favorable state transition. Prior work assumes that individual arms only benefit if they receive the resource directly. However, many allocation tasks occur within communities and can be characterized by positive externalities that allow arms to derive partial benefit when their neighbor(s) receive the resource. We thus introduce networked restless bandits, a novel multi-armed bandit setting in which arms are both restless and embedded within a directed graph. We then present Greta, a graph-aware, Whittle index-based heuristic algorithm that can be used to efficiently construct a constrained reward-maximizing action vector at each timestep. Our empirical results demonstrate that Greta outperforms comparison policies across a range of hyperparameter values and graph topologies. | ['John P. Dickerson', 'Christine Herlihy'] | 2022-12-09 | null | null | null | null | ['multi-armed-bandits'] | ['miscellaneous'] | [ 1.15407921e-01 2.77935505e-01 -1.40252864e+00 2.21326083e-01
-3.02132040e-01 -7.86686957e-01 5.07348001e-01 -9.98731144e-03
-2.54003227e-01 1.36921191e+00 2.94213206e-01 -6.80165291e-01
-8.22633088e-01 -9.89054739e-01 -6.72061503e-01 -6.20350838e-01
-4.47462589e-01 1.12821913e+00 -1.63857147e-01 7.59271830e-02
-8.13451260e-02 4.78941470e-01 -5.83445728e-01 -8.85274261e-02
4.83517081e-01 6.96238279e-01 -1.43173784e-01 6.47736371e-01
7.42522925e-02 9.09016073e-01 -3.76429319e-01 -3.16458821e-01
6.95559204e-01 -5.29690206e-01 -9.74523067e-01 3.87098461e-01
-4.37149405e-01 -5.79289734e-01 -6.38167024e-01 6.62537932e-01
2.14206174e-01 4.50470358e-01 7.30394363e-01 -1.57629848e+00
-4.92155790e-01 1.35920405e+00 -1.14174914e+00 5.29549003e-01
3.00777078e-01 2.57172555e-01 1.34450889e+00 1.51197851e-01
5.86230695e-01 1.08564949e+00 1.46511674e-01 4.18421298e-01
-1.51591480e+00 -4.99683410e-01 6.60327733e-01 -1.08784355e-01
-5.91505110e-01 -3.33546162e-01 8.24946463e-01 -7.44072795e-02
1.20428216e+00 4.83060896e-01 1.47909093e+00 6.37308121e-01
-2.68376976e-01 9.18753445e-01 1.08998394e+00 -3.94892722e-01
4.18159366e-01 -5.63736558e-01 1.68424875e-01 4.97072756e-01
6.46513999e-01 -2.90228371e-02 -6.25692368e-01 -5.68984032e-01
9.99321401e-01 2.95167118e-01 -1.22554339e-01 -6.92001343e-01
-1.09845078e+00 8.44084263e-01 5.79883516e-01 -1.17623799e-01
-1.05922794e+00 9.58625972e-01 6.25366392e-03 3.25245380e-01
6.48600221e-01 2.46055797e-01 -5.13200164e-02 -1.21530138e-01
-7.30750263e-01 1.51756570e-01 7.51611292e-01 1.04178703e+00
7.84765065e-01 -3.20632830e-02 -7.12681711e-01 6.39181495e-01
1.58490613e-01 5.22683740e-01 -1.48704588e-01 -1.07142484e+00
8.19898069e-01 3.98433477e-01 8.49850178e-01 -5.74231565e-01
-5.22217333e-01 -7.26454139e-01 -6.27856493e-01 -2.40048900e-01
3.76655668e-01 -5.74683785e-01 -9.47642863e-01 1.77196777e+00
4.54657882e-01 1.51702613e-01 -1.98635623e-01 9.29276764e-01
-2.11954474e-01 7.27533698e-01 -1.81232736e-01 -8.42131317e-01
9.13026392e-01 -1.30531895e+00 -5.71722686e-01 -7.21926570e-01
3.39521021e-01 -1.95823193e-01 5.49698472e-01 -1.59280710e-02
-1.60070288e+00 7.59894192e-01 -5.10943055e-01 7.07294941e-01
8.72473344e-02 -7.08204746e-01 1.07098329e+00 8.09278488e-01
-1.01621091e+00 3.49927753e-01 -9.05934274e-01 -1.26484573e-01
7.69769609e-01 3.56473625e-01 1.49184704e-01 -1.44168869e-01
-7.49461353e-01 5.42576194e-01 2.38312464e-02 5.93717583e-02
-1.13201058e+00 -5.52253902e-01 -1.79266438e-01 3.85162711e-01
1.01429570e+00 -1.07136786e+00 1.36869824e+00 -9.55193996e-01
-1.22475743e+00 4.31491286e-01 7.39017949e-02 -7.70975947e-01
5.90497017e-01 3.05761844e-01 3.02878588e-01 1.39458463e-01
8.00413638e-02 1.48218632e-01 8.13380361e-01 -8.37958515e-01
-6.12213492e-01 -2.84889847e-01 4.70015615e-01 8.12179029e-01
-6.34317875e-01 -1.44104689e-01 -3.20730895e-01 -5.73611438e-01
-2.18875140e-01 -1.31919312e+00 -4.74699885e-01 -5.17634928e-01
-8.12844634e-01 -1.53205335e-01 1.06071029e-02 1.03126764e-01
1.19159520e+00 -1.37139452e+00 3.37636560e-01 5.59607625e-01
1.73343137e-01 -2.60257542e-01 -3.50817144e-01 7.81306982e-01
3.52774858e-01 4.66870070e-01 1.58773035e-01 -1.51769683e-01
1.11452080e-01 1.92689806e-01 -3.72155346e-02 6.96871638e-01
-4.59615767e-01 9.06783879e-01 -9.44834590e-01 -1.44347876e-01
-2.21750826e-01 -4.81611580e-01 -6.90128565e-01 -2.95291990e-02
-4.50835526e-01 -1.48196682e-01 -1.09264660e+00 5.60455322e-01
2.95661986e-01 -7.79755175e-01 5.61232805e-01 6.76555634e-01
1.89737841e-01 2.09877774e-01 -9.64006245e-01 1.01368570e+00
-3.29104543e-01 2.52389610e-01 4.52964872e-01 -1.00658429e+00
3.37482467e-02 7.99794272e-02 6.15422487e-01 -2.71731287e-01
7.20744729e-02 -7.22569674e-02 3.91509607e-02 8.01632479e-02
4.34299439e-01 1.44836707e-02 1.45655110e-01 1.37231708e+00
-6.03667974e-01 -1.82440683e-01 2.52264053e-01 7.62500286e-01
1.52299619e+00 -3.58211070e-01 2.67068416e-01 -6.32769018e-02
-3.65175426e-01 2.49663517e-01 4.62305367e-01 1.29159093e+00
-3.29497576e-01 -1.52505547e-01 7.25486696e-01 -3.99862319e-01
-9.58061159e-01 -7.36672759e-01 5.50457656e-01 1.42925978e+00
2.75706440e-01 3.57695580e-01 -3.32147986e-01 -5.11588395e-01
5.41425228e-01 5.09654582e-01 -7.46366322e-01 1.83814719e-01
-1.70114130e-01 -8.05096865e-01 -1.79010436e-01 2.26524070e-01
2.80085206e-01 -1.08263218e+00 -6.30577624e-01 5.03972173e-01
-1.21935725e-01 -7.02470243e-01 -1.02087712e+00 2.44897470e-01
-9.22039330e-01 -1.03163517e+00 -9.78920937e-01 -2.77507007e-01
7.64271677e-01 8.73919249e-01 1.03960931e+00 3.51598263e-01
-3.02566197e-02 8.62358510e-01 -1.99404314e-01 -2.20603764e-01
1.42931119e-01 3.11480463e-01 -5.80664165e-02 -3.28956731e-02
-2.27942362e-01 -5.06008446e-01 -1.01120913e+00 2.77015597e-01
-7.01796591e-01 3.36822480e-01 3.42963308e-01 8.68385732e-01
5.77107430e-01 -3.00273031e-01 1.01839375e+00 -9.05110180e-01
1.03236437e+00 -8.74908268e-01 -8.48212242e-01 5.67088664e-01
-4.63005632e-01 -7.18092695e-02 2.43050590e-01 -6.36575997e-01
-9.60529089e-01 -1.71720497e-02 9.29955781e-01 -2.38187030e-01
6.42234266e-01 8.13114047e-01 6.92490280e-01 3.95759195e-02
4.45786804e-01 5.05650081e-02 6.90965354e-02 4.52061594e-02
4.62071121e-01 5.09456933e-01 -1.88580573e-01 -8.99688005e-01
4.11913991e-01 4.19521540e-01 2.79608101e-01 -3.23136806e-01
-7.53061950e-01 -2.35948011e-01 4.41025436e-01 -5.94521165e-01
1.05727263e-01 -8.13025475e-01 -1.22354782e+00 3.13708931e-01
-7.50458658e-01 -1.09256315e+00 -3.15920502e-01 5.27360737e-01
-9.13172305e-01 6.70323893e-02 -6.22648358e-01 -1.55097651e+00
-3.38266522e-01 -6.60054684e-01 2.12595835e-01 3.05150211e-01
1.50301456e-01 -7.66615093e-01 1.03064954e-01 6.77090466e-01
5.22836447e-01 -7.78510198e-02 7.73294806e-01 -5.16328275e-01
-1.03639925e+00 -2.77667101e-02 -6.59825131e-02 -5.37670314e-01
2.58265585e-01 -1.42277613e-01 -1.39029026e-01 -7.72842586e-01
-8.02922130e-01 -6.34365201e-01 7.72711754e-01 1.01160455e+00
9.79778588e-01 -9.98024404e-01 -7.40144253e-01 2.10469261e-01
1.10405099e+00 2.54109830e-01 6.64519891e-02 3.97203028e-01
4.72395122e-02 2.03565404e-01 5.50353110e-01 9.23652589e-01
3.57671857e-01 4.51518744e-01 1.01929128e+00 1.17835142e-01
1.51260808e-01 -2.38735348e-01 1.07585162e-01 -6.11777306e-02
-3.34578842e-01 -9.25672591e-01 -7.72415340e-01 1.06446695e+00
-2.28357053e+00 -1.02554429e+00 2.50496000e-01 2.21552467e+00
8.80449235e-01 2.91630864e-01 8.60431373e-01 -3.00593287e-01
9.74918604e-01 2.38495156e-01 -1.38084662e+00 -2.96510816e-01
-3.36624426e-03 -3.03616792e-01 1.15960813e+00 6.07017756e-01
-5.80004275e-01 9.52215254e-01 7.16667175e+00 8.14773917e-01
-4.41598207e-01 -3.54428887e-02 1.06949675e+00 -1.11631703e+00
-5.05449772e-01 -9.95503217e-02 -2.84836084e-01 2.92047739e-01
8.49426627e-01 -7.45515764e-01 1.39276671e+00 7.08117843e-01
3.25708359e-01 -3.25275809e-01 -9.02818084e-01 5.52850366e-01
-5.39667428e-01 -1.76851153e+00 -2.71361917e-01 3.21196318e-01
1.31854880e+00 3.20510656e-01 1.70925334e-01 -1.46850199e-01
1.57188189e+00 -7.31627703e-01 6.15311801e-01 2.02594146e-01
6.02598548e-01 -1.14385235e+00 -3.12303025e-02 5.65195262e-01
-9.18909550e-01 -5.10538101e-01 -1.86110109e-01 -5.14189243e-01
2.87665904e-01 4.91661519e-01 -9.70530033e-01 2.78583109e-01
3.93415868e-01 3.39439511e-01 3.76364261e-01 1.15874875e+00
-1.55822068e-01 7.42846549e-01 -5.55218756e-01 -4.89857167e-01
5.56367099e-01 -4.82258022e-01 7.67884552e-01 4.58253622e-01
2.30869442e-01 5.08782506e-01 5.15960932e-01 5.73646486e-01
-4.25276935e-01 -2.74571665e-02 -6.81142151e-01 -5.44752598e-01
9.39078510e-01 1.12462926e+00 -1.13877308e+00 -3.24514270e-01
-1.81172267e-01 5.34090698e-01 5.85661411e-01 7.89777398e-01
-6.80768847e-01 1.47346249e-02 8.04133713e-01 1.50764836e-02
3.86610240e-01 1.30843729e-01 -1.42138660e-01 -9.37914073e-01
-4.02936816e-01 -4.93375748e-01 6.92059159e-01 -1.05157757e+00
-1.25870228e+00 1.89140975e-01 6.13361271e-03 -4.95044917e-01
-3.03592473e-01 1.56574145e-01 -5.98612368e-01 5.25811970e-01
-1.60523617e+00 -1.03427899e+00 3.54001224e-01 7.06672251e-01
3.03100079e-01 -3.65459211e-02 2.16951355e-01 -3.54940712e-01
-8.62003386e-01 2.99104661e-01 3.71966302e-01 -2.28538036e-01
4.05854471e-02 -1.01715612e+00 -5.42823523e-02 7.57181287e-01
-7.95915350e-02 4.27717537e-01 7.52853036e-01 -8.32926631e-01
-1.80304635e+00 -9.50468898e-01 5.44933341e-02 1.82183847e-01
9.13921654e-01 7.20075965e-02 -2.73375921e-02 1.12656415e+00
3.83972347e-01 1.49009064e-01 5.62062263e-01 3.20013285e-01
1.65447257e-02 -8.25635344e-02 -1.33377111e+00 7.67473698e-01
1.52618062e+00 -5.70935123e-02 -1.60087217e-02 8.58060181e-01
6.94099665e-01 -1.33106634e-01 -6.71336114e-01 -1.84436321e-01
3.87231678e-01 -6.60713255e-01 8.44443917e-01 -1.02482712e+00
1.80953130e-01 2.38864586e-01 1.96666941e-01 -1.92208850e+00
-6.65278137e-01 -1.38386834e+00 -4.28258270e-01 5.97871363e-01
7.53441215e-01 -1.04263711e+00 1.27868903e+00 8.92225742e-01
2.34439537e-01 -5.79427719e-01 -1.14013112e+00 -9.52103555e-01
1.20167412e-01 1.60214573e-01 1.03706741e+00 8.13669503e-01
5.95730484e-01 3.33047092e-01 -6.97454512e-01 -1.79517195e-01
9.01656687e-01 6.55883908e-01 5.51232517e-01 -9.47311044e-01
-4.77466851e-01 -7.08951354e-01 6.13355279e-01 -1.16257274e+00
1.66663259e-01 -7.33687103e-01 -1.27542332e-01 -1.99029768e+00
9.47321653e-01 -8.38208020e-01 -3.63797843e-01 1.01802254e+00
2.94396793e-03 1.49296224e-01 4.00028974e-01 3.05383742e-01
-9.90785241e-01 3.59458148e-01 1.48214710e+00 -4.02041465e-01
-5.30946553e-01 3.72199297e-01 -9.44658816e-01 1.02481470e-01
8.40899587e-01 -5.35415411e-01 -6.37083888e-01 -2.95375109e-01
9.06736851e-01 1.04116893e+00 -1.07163683e-01 2.19884403e-02
2.84796506e-01 -1.13673389e+00 -9.43766162e-02 -5.76574385e-01
3.50854486e-01 -8.50053191e-01 6.00633144e-01 8.65722418e-01
-7.70479381e-01 -1.21050790e-01 -2.50674248e-01 1.17333758e+00
6.44457996e-01 6.13731630e-02 4.11952704e-01 -1.93292633e-01
2.44913161e-01 7.00769603e-01 -7.56273985e-01 9.21175852e-02
1.35344279e+00 -1.06493115e-01 -6.12683773e-01 -1.07111895e+00
-6.15354240e-01 8.33710968e-01 4.36537504e-01 -6.66671246e-02
2.48299956e-01 -1.23630142e+00 -7.61478305e-01 -4.73523796e-01
-3.35229069e-01 -1.87944219e-01 2.09246933e-01 6.26247466e-01
-1.09983929e-01 3.75859678e-01 -2.22402155e-01 -9.42150727e-02
-1.05193067e+00 7.81127036e-01 3.16621870e-01 -7.60323644e-01
-1.77836403e-01 6.47404492e-01 -1.77211240e-01 2.65416056e-01
1.46588758e-01 4.32139672e-02 1.29405990e-01 1.64269254e-01
3.22058111e-01 7.10615337e-01 -3.37162167e-01 -2.99615879e-02
-2.07339719e-01 -1.72528654e-01 -1.21156320e-01 -4.95830655e-01
1.78611982e+00 -2.56430060e-01 -3.03006738e-01 -1.34957343e-01
2.49551341e-01 -2.37233475e-01 -1.41800082e+00 -4.59382415e-01
-3.65341306e-01 -8.21866512e-01 3.55015367e-01 -8.61939192e-01
-1.50578356e+00 -4.59900573e-02 -2.36611053e-01 1.22376835e+00
9.36500132e-01 -8.80387425e-03 3.13075274e-01 4.08410251e-01
7.64121830e-01 -1.17137599e+00 1.87752426e-01 2.11226538e-01
7.07923114e-01 -7.98096955e-01 3.40267390e-01 -2.62952566e-01
-6.15901053e-01 4.63376284e-01 3.51338297e-01 -6.76951036e-02
5.64581037e-01 -7.92205986e-03 -8.37697625e-01 -9.67159271e-02
-1.24671650e+00 -3.95322591e-01 -3.33037645e-01 6.40071034e-01
-8.27538371e-02 5.90331495e-01 -4.77129310e-01 2.67998368e-01
2.43067905e-01 1.50730954e-02 8.77161324e-01 1.05245280e+00
-6.76305413e-01 -9.38208938e-01 -3.54635268e-01 1.17589819e+00
-5.67075551e-01 -8.50401912e-03 -6.50744200e-01 3.85935426e-01
-1.04190326e+00 1.26490796e+00 2.90051043e-01 3.61881256e-01
-1.98663101e-01 -3.13873321e-01 6.12314224e-01 -6.09325290e-01
-5.84424257e-01 3.83995235e-01 3.85501832e-01 -3.73558760e-01
-5.31510830e-01 -4.72886413e-01 -9.41563249e-01 -7.68440306e-01
-5.64864874e-01 3.94596189e-01 2.74175763e-01 6.16033077e-01
5.59047043e-01 5.45848787e-01 1.02627659e+00 -7.24732876e-01
-6.25958681e-01 -7.18315363e-01 -8.45949352e-01 -4.64158505e-03
5.37245750e-01 -6.56090915e-01 -3.66464168e-01 -6.08030319e-01] | [4.487096309661865, 3.261888265609741] |
1bfbfd7a-ece2-4eb2-b564-b4de0431e253 | persian-keyphrase-generation-using-sequence | 2009.12271 | null | https://arxiv.org/abs/2009.12271v1 | https://arxiv.org/pdf/2009.12271v1.pdf | Persian Keyphrase Generation Using Sequence-to-Sequence Models | Keyphrases are a very short summary of an input text and provide the main subjects discussed in the text. Keyphrase extraction is a useful upstream task and can be used in various natural language processing problems, for example, text summarization and information retrieval, to name a few. However, not all the keyphrases are explicitly mentioned in the body of the text. In real-world examples there are always some topics that are discussed implicitly. Extracting such keyphrases requires a generative approach, which is adopted here. In this paper, we try to tackle the problem of keyphrase generation and extraction from news articles using deep sequence-to-sequence models. These models significantly outperform the conventional methods such as Topic Rank, KPMiner, and KEA in the task of keyphrase extraction. | ['Ehsan Doostmohammadi', 'Mohammad Hadi Bokaei', 'Hossein Sameti'] | 2020-09-25 | null | null | null | null | ['keyphrase-generation'] | ['natural-language-processing'] | [ 5.15411377e-01 1.49019256e-01 -5.04799366e-01 2.12182701e-01
-8.97441208e-01 -7.11890936e-01 1.08958840e+00 1.04740024e+00
-4.96552855e-01 1.26928771e+00 9.63956416e-01 -2.13007271e-01
-1.15174472e-01 -7.36052990e-01 -6.87853634e-01 -5.67420423e-01
1.30048901e-01 4.09162581e-01 2.77835995e-01 -2.97659159e-01
8.58259857e-01 2.00634494e-01 -1.42939162e+00 4.98992980e-01
9.58395898e-01 6.05688930e-01 3.49648535e-01 6.09282017e-01
-9.11977232e-01 1.02915478e+00 -9.38023984e-01 -3.10452461e-01
-2.26886049e-01 -6.18164241e-01 -8.78697813e-01 2.27892194e-02
3.73254955e-01 -4.11512315e-01 -5.22603214e-01 1.09763634e+00
3.20278585e-01 1.65901437e-01 7.80937195e-01 -1.11586523e+00
9.04537067e-02 1.23335600e+00 -7.86602139e-01 6.23215675e-01
4.01985615e-01 -4.69676912e-01 1.33990383e+00 -5.99834859e-01
7.53644407e-01 9.95688379e-01 3.13541628e-02 1.69232979e-01
-6.26527786e-01 -4.34691131e-01 1.60870150e-01 1.16866149e-01
-1.00787783e+00 -2.76860863e-01 8.86131406e-01 -2.26181477e-01
7.41619766e-01 3.69580537e-01 7.71799624e-01 7.55462289e-01
5.32410145e-01 1.36640072e+00 6.92267954e-01 -3.50740194e-01
1.75242037e-01 -7.08097592e-02 5.66401958e-01 2.33462811e-01
5.28875649e-01 -6.82801068e-01 -7.94460952e-01 -5.20185828e-01
3.04745436e-01 5.68787120e-02 -4.88944650e-01 2.27495104e-01
-1.52205849e+00 8.53112161e-01 -1.66970134e-01 3.54693055e-01
-7.16241717e-01 2.41171774e-02 7.71438599e-01 3.07721734e-01
6.54002368e-01 7.33762741e-01 -4.32915360e-01 -1.24341749e-01
-1.24216223e+00 8.74994278e-01 1.10045862e+00 9.60596800e-01
5.47368467e-01 -2.93885112e-01 -4.68341589e-01 5.32828093e-01
-8.72477666e-02 6.46374151e-02 6.64500117e-01 -3.24449837e-01
7.91538000e-01 7.19535112e-01 3.83678675e-01 -1.14292777e+00
-2.77516007e-01 -3.08320314e-01 -8.57443213e-01 -6.96890354e-01
-1.88908502e-01 -4.42167580e-01 -7.42071331e-01 1.09281993e+00
3.19275498e-01 -4.54374738e-02 1.23180971e-01 3.22568655e-01
1.28714347e+00 1.32394564e+00 -6.98748007e-02 -7.55290926e-01
1.54107988e+00 -8.18811655e-01 -1.09877145e+00 -2.64487803e-01
2.28629887e-01 -9.70969677e-01 4.13333029e-01 2.75914907e-01
-7.93542206e-01 -6.64443374e-02 -9.03827608e-01 -1.62098497e-01
-4.08685744e-01 1.17706813e-01 6.45555556e-01 3.36339734e-02
-6.20814204e-01 5.09974778e-01 -3.54119658e-01 -3.78635824e-01
3.09692323e-01 -1.02051891e-01 -1.75462887e-01 4.92615029e-02
-1.32616603e+00 6.09192193e-01 1.01332617e+00 -3.03876907e-01
-5.46995044e-01 -7.43880689e-01 -7.17866600e-01 2.02811792e-01
1.03881633e+00 -6.74295723e-01 1.53042245e+00 -2.89583534e-01
-1.15309298e+00 5.79727352e-01 -3.24128717e-01 -6.96509302e-01
3.58894676e-01 -7.30315745e-01 -2.55841613e-01 2.85249859e-01
2.19837144e-01 2.89984286e-01 9.75394726e-01 -6.96450174e-01
-9.70677972e-01 -4.27352749e-02 8.39737877e-02 3.64565194e-01
-2.53407598e-01 3.27424616e-01 -3.83888781e-01 -1.10031271e+00
-3.48591693e-02 -5.86991370e-01 -3.90308797e-01 -6.93225026e-01
-1.01142669e+00 -5.78953147e-01 8.21038127e-01 -8.69667351e-01
1.66621673e+00 -1.78673649e+00 1.74882367e-01 -3.71326208e-02
4.01142061e-01 7.13702068e-02 3.31742406e-01 1.13671374e+00
1.69405103e-01 3.50428283e-01 -1.60010681e-01 2.99089849e-01
-9.86243039e-03 -1.93177834e-01 -1.07583785e+00 1.67515874e-01
-6.99946210e-02 8.07407856e-01 -9.92812097e-01 -7.76386559e-01
-9.40278769e-02 -1.42239735e-01 8.68133083e-03 4.10148352e-01
-7.32447624e-01 6.30616620e-02 -8.51565003e-01 1.84362382e-01
1.66584447e-01 -1.59942582e-01 -1.95567191e-01 -1.52830347e-01
-3.75791252e-01 8.12695861e-01 -1.04771805e+00 1.35457373e+00
-8.55289921e-02 8.53330910e-01 -2.41608247e-01 -8.48809481e-01
5.19926131e-01 5.15296817e-01 8.18992734e-01 -1.37897894e-01
1.29354715e-01 2.20780566e-01 -1.83108985e-01 -1.17261268e-01
1.27418041e+00 1.58931330e-01 -3.95852566e-01 8.86023045e-01
-4.25121821e-02 -4.28919256e-01 8.80620241e-01 8.71547759e-01
9.43682432e-01 -4.41231936e-01 1.07516086e+00 -2.32132137e-01
4.14507151e-01 3.30777258e-01 2.61162579e-01 9.82925415e-01
4.46243107e-01 5.00617683e-01 8.46747696e-01 -1.62252441e-01
-1.10195303e+00 -2.33859852e-01 1.48869395e-01 8.24915826e-01
1.67127252e-02 -1.08718669e+00 -6.44055665e-01 -7.49552667e-01
-5.55360019e-02 7.28968740e-01 -4.47230518e-01 5.17861210e-02
-5.59885263e-01 -7.24039257e-01 2.84357488e-01 1.81592792e-01
3.77809852e-01 -1.14194584e+00 -5.25523484e-01 6.07856214e-01
-4.98116255e-01 -1.12408102e+00 -7.02889562e-01 1.41738877e-01
-7.02080071e-01 -9.24959362e-01 -1.09027994e+00 -4.58501309e-01
5.26952207e-01 4.71474111e-01 9.74511206e-01 -2.00197443e-01
-1.09116659e-01 1.46983609e-01 -6.30813479e-01 -7.54492879e-01
-5.82180679e-01 5.94061971e-01 -1.34999841e-01 -1.38270363e-01
1.05982088e-01 -3.90190184e-01 -2.64054835e-01 -4.63602811e-01
-1.17572153e+00 3.79828185e-01 7.89010882e-01 7.48614609e-01
4.94482607e-01 5.54705083e-01 5.50527036e-01 -1.25933433e+00
1.20507717e+00 -5.68318248e-01 -3.90370458e-01 2.81436831e-01
-3.04504395e-01 3.57148647e-01 5.67781270e-01 -3.90387088e-01
-8.74453366e-01 -3.58812153e-01 -4.17133048e-02 2.79552519e-01
-7.95735791e-02 1.04607916e+00 -1.41120717e-01 7.10278749e-01
2.79615462e-01 7.86551952e-01 -5.20827591e-01 -6.51944101e-01
4.69177246e-01 6.60682976e-01 2.57912636e-01 -3.62961113e-01
7.12209702e-01 1.45566538e-01 -1.73517570e-01 -1.30498505e+00
-1.20763409e+00 -9.15798783e-01 -3.63970071e-01 -9.83318761e-02
4.38121736e-01 -8.47966075e-01 -1.69681191e-01 4.95043546e-01
-1.42437327e+00 3.26652676e-01 -3.01138043e-01 1.81682095e-01
-2.22769544e-01 6.11411452e-01 -4.76630002e-01 -5.73439598e-01
-9.78522599e-01 -6.14600539e-01 1.13310456e+00 6.35781348e-01
-5.67697763e-01 -5.51671267e-01 7.05892779e-03 -7.52083138e-02
8.02264065e-02 2.02147320e-01 1.07976151e+00 -1.21132636e+00
-4.31945801e-01 -2.80920625e-01 -1.99603979e-02 -6.44137412e-02
4.26407933e-01 1.92222428e-02 -3.95895213e-01 -7.46225268e-02
-1.73353463e-01 -1.56243488e-01 1.28421116e+00 3.82623523e-01
1.02295530e+00 -1.03124976e+00 -4.38993126e-01 -9.97902360e-04
1.01455820e+00 1.62539765e-01 4.18485135e-01 3.53493303e-01
7.82964528e-01 7.63948500e-01 6.45409822e-01 6.12508953e-01
3.27882320e-01 3.39136004e-01 -1.12819634e-01 2.27009222e-01
2.79567719e-01 -4.67879891e-01 3.24615598e-01 1.19068539e+00
3.57392728e-01 -5.04884362e-01 -6.57854438e-01 6.32547617e-01
-1.88491726e+00 -1.11971974e+00 -2.37353519e-01 1.71365035e+00
1.31759274e+00 4.24972862e-01 2.51181331e-02 2.71707207e-01
6.16297543e-01 8.10170352e-01 -3.54980230e-01 -9.88151431e-02
-9.95822996e-02 7.55485520e-02 4.21688348e-01 2.76557833e-01
-1.23533297e+00 1.11495221e+00 5.21050167e+00 9.09734368e-01
-8.50748658e-01 -3.25960398e-01 4.44368154e-01 3.64063710e-01
-3.74793231e-01 1.72701880e-01 -1.17685258e+00 3.93717319e-01
6.34413779e-01 -9.86666143e-01 -1.88928068e-01 9.46905851e-01
1.75450906e-01 -5.77431381e-01 -9.46369052e-01 7.83645809e-01
1.53284952e-01 -1.51624632e+00 5.92096269e-01 -6.03485294e-02
7.75184155e-01 -3.88086468e-01 -4.61534888e-01 1.18455850e-01
2.81180441e-01 -6.15503848e-01 5.14929175e-01 3.72929960e-01
2.73667872e-01 -9.09059942e-01 8.10895026e-01 6.02333724e-01
-9.73174214e-01 1.73483387e-01 -2.83137560e-01 2.47109801e-01
3.25634569e-01 1.10642540e+00 -9.40110624e-01 7.35386848e-01
3.47038716e-01 6.20754778e-01 -2.38250107e-01 1.32233858e+00
-5.02047956e-01 8.24782789e-01 -2.42315903e-01 -6.70849800e-01
4.15155113e-01 -4.16469434e-03 8.64691496e-01 1.36697149e+00
2.45547950e-01 2.51554012e-01 3.15445662e-01 3.48091185e-01
-4.09281284e-01 5.38794637e-01 -5.32037616e-01 -7.91568518e-01
3.21280390e-01 1.21013403e+00 -1.09242666e+00 -7.67866611e-01
-8.67535099e-02 8.84782314e-01 -3.42590868e-01 2.44302794e-01
-2.93658406e-01 -8.51883829e-01 3.52029264e-01 6.41209781e-02
1.80712730e-01 -3.37352246e-01 1.37892649e-01 -1.30180073e+00
-1.63147077e-02 -1.17827094e+00 4.09480631e-01 -5.57531774e-01
-1.00554001e+00 3.25074852e-01 3.97352874e-01 -9.95721877e-01
-6.31037235e-01 -1.69093519e-01 -7.59550750e-01 5.85090578e-01
-1.36483669e+00 -7.37269580e-01 -3.87148336e-02 8.37742910e-02
1.11866689e+00 -6.20083734e-02 3.62263322e-01 -2.07369685e-01
-3.78674597e-01 -8.65459070e-02 1.85115039e-01 3.23976099e-01
5.54736137e-01 -1.41833591e+00 7.46632814e-01 9.72320735e-01
4.15095598e-01 8.87157917e-01 1.14881802e+00 -9.55609679e-01
-1.32337856e+00 -9.44849014e-01 1.27700448e+00 1.68008022e-02
7.70329893e-01 -2.61646092e-01 -9.12681878e-01 2.71517903e-01
6.93618834e-01 -7.66888678e-01 5.92096210e-01 -1.06520571e-01
9.06793177e-02 6.35480508e-02 -3.29191655e-01 9.13731456e-01
4.39176589e-01 -3.86774629e-01 -1.18768919e+00 5.37310719e-01
1.09277821e+00 -5.26965439e-01 -2.32654914e-01 6.85317740e-02
2.90133178e-01 -2.26611823e-01 6.08089566e-01 -7.11302102e-01
6.42029643e-01 -2.71878779e-01 2.88028806e-01 -1.48908293e+00
1.82393938e-02 -1.11592770e+00 -7.31471896e-01 1.43711829e+00
3.12354803e-01 -2.75622964e-01 4.67735380e-01 6.31932765e-02
1.54046252e-01 -5.29115140e-01 -4.77853268e-01 -2.08623543e-01
-3.32481325e-01 -1.23703472e-01 4.94671643e-01 7.83924580e-01
2.17040882e-01 9.68648553e-01 -5.20201623e-01 -3.82897764e-01
4.81762201e-01 4.20661837e-01 8.51334155e-01 -1.41741300e+00
7.37982690e-02 -5.17133474e-01 1.45002201e-01 -1.22447312e+00
-3.25259753e-02 -5.14801562e-01 1.13313742e-01 -2.21712685e+00
4.52748120e-01 1.26788884e-01 2.20305827e-02 2.04140782e-01
-6.02914333e-01 -7.36458182e-01 -1.36833265e-01 3.36139053e-01
-6.32613003e-01 6.22956038e-01 1.10744154e+00 -4.16706622e-01
-4.02320325e-01 3.30037743e-01 -1.08956909e+00 5.64624488e-01
7.86486804e-01 -6.90171301e-01 -4.06799257e-01 1.69972882e-01
6.04668319e-01 2.09899873e-01 -9.45754424e-02 -5.14308631e-01
5.21125317e-01 -5.50703704e-01 -5.63447457e-03 -1.15784681e+00
-8.92468616e-02 -3.54373664e-01 -1.72494099e-01 1.97968856e-01
-5.96606612e-01 2.14480281e-01 1.87178761e-01 5.15205503e-01
-4.09426659e-01 -5.21827817e-01 1.79036319e-01 -5.24583519e-01
-7.27134347e-01 3.07116032e-01 -6.97194278e-01 3.22346151e-01
4.91418988e-01 3.53785545e-01 -4.74773228e-01 -6.85562134e-01
5.80478497e-02 2.81319529e-01 5.00586554e-02 3.82347643e-01
7.21847534e-01 -8.70201230e-01 -1.07921505e+00 -5.00123024e-01
1.75595075e-01 2.22039685e-01 1.38411466e-02 6.56021237e-01
-3.55932355e-01 7.50295997e-01 7.96195641e-02 3.24107781e-02
-1.14257789e+00 4.31633592e-01 -3.87514263e-01 -6.56370521e-01
-8.96984041e-01 5.86582124e-01 1.18056037e-01 1.91258356e-01
2.35681042e-01 -5.58579624e-01 -8.53131652e-01 7.83621907e-01
1.04090095e+00 4.31238972e-02 4.74984236e-02 -4.16957945e-01
3.71008441e-02 2.47428901e-02 -8.37056339e-01 -3.10379893e-01
1.38279629e+00 4.50812504e-02 -5.64214528e-01 5.96463203e-01
9.73925054e-01 3.47935528e-01 -5.83481550e-01 -4.94468868e-01
5.58866262e-01 2.38101259e-02 2.30735615e-01 -4.47372347e-01
-3.81494135e-01 6.85754776e-01 -3.07340115e-01 4.98279244e-01
9.46037412e-01 2.41039395e-01 1.13152337e+00 8.29662204e-01
1.05202369e-01 -1.24740064e+00 -2.57510915e-02 7.47602999e-01
9.54816937e-01 -9.44253087e-01 6.56531334e-01 -1.85729384e-01
-4.87596393e-01 1.04218793e+00 3.20148528e-01 2.12479353e-01
4.50384080e-01 1.58246130e-01 -2.88787037e-01 -1.82231605e-01
-8.52777839e-01 -3.34358573e-01 5.65360665e-01 -1.57743961e-01
4.34943289e-01 -1.94153756e-01 -7.93802619e-01 5.49130082e-01
-4.30585831e-01 -3.46812636e-01 7.94827282e-01 1.16744208e+00
-9.38902915e-01 -9.07473207e-01 -2.46688396e-01 9.23237503e-01
-1.01989627e+00 -4.16655362e-01 -9.26947117e-01 4.96375859e-01
-5.06685257e-01 7.63332248e-01 -1.76747248e-01 -1.60242423e-01
2.99124862e-03 1.42232820e-01 6.60009757e-02 -9.41807806e-01
-6.39504492e-01 2.78688788e-01 2.27297589e-01 7.98587427e-02
-3.64176184e-01 -7.59578764e-01 -1.13569736e+00 -1.99724153e-01
-4.15427804e-01 8.04699183e-01 5.78700960e-01 1.09087682e+00
2.25106701e-01 6.40082181e-01 5.60120106e-01 -4.06875849e-01
-3.63406628e-01 -1.25379455e+00 -2.85516053e-01 -7.33425617e-02
6.31307721e-01 -1.56498700e-01 -9.24174935e-02 2.63763785e-01] | [12.304964065551758, 8.994425773620605] |
3939f942-774a-49f1-af57-44df44d7aa24 | a-corpus-for-commonsense-inference-in-story | null | null | https://aclanthology.org/2022.lrec-1.375 | https://aclanthology.org/2022.lrec-1.375.pdf | A Corpus for Commonsense Inference in Story Cloze Test | The Story Cloze Test (SCT) is designed for training and evaluating machine learning algorithms for narrative understanding and inferences. The SOTA models can achieve over 90% accuracy on predicting the last sentence. However, it has been shown that high accuracy can be achieved by merely using surface-level features. We suspect these models may not truly understand the story. Based on the SCT dataset, we constructed a human-labeled and human-verified commonsense knowledge inference dataset. Given the first four sentences of a story, we asked crowd-source workers to choose from four types of narrative inference for deciding the ending sentence and which sentence contributes most to the inference. We accumulated data on 1871 stories, and three human workers labeled each story. Analysis of the intra-category and inter-category agreements show a high level of consensus. We present two new tasks for predicting the narrative inference categories and contributing sentences. Our results show that transformer-based models can reach SOTA performance on the original SCT task using transfer learning but don’t perform well on these new and more challenging tasks. | ['Mei Si', 'Julian Lioanag', 'Ethan Joseph', 'Bingsheng Yao'] | null | null | null | null | lrec-2022-6 | ['cloze-test'] | ['natural-language-processing'] | [ 1.88987941e-01 2.01677829e-01 -3.03536594e-01 -4.75889087e-01
-1.35131681e+00 -7.65596390e-01 9.10281837e-01 1.44241691e-01
-1.98131874e-01 8.74692500e-01 9.21240091e-01 -3.31243038e-01
-1.13707289e-01 -7.83228040e-01 -5.38490653e-01 -2.04528138e-01
4.29477155e-01 6.99940383e-01 2.00709462e-01 -2.73756534e-01
7.51095653e-01 -1.89838246e-01 -1.29995966e+00 1.24169981e+00
7.75598109e-01 8.29218805e-01 1.22604437e-01 8.13190818e-01
-1.47486240e-01 1.99343002e+00 -8.65005612e-01 -9.70646620e-01
-2.59173274e-01 -7.74139524e-01 -1.44205642e+00 -1.89164147e-01
3.65857899e-01 -3.28274578e-01 -2.18473718e-01 5.30797184e-01
2.94894248e-01 -5.35531854e-03 1.02621520e+00 -1.17015004e+00
-8.58688474e-01 1.28175616e+00 -1.05794504e-01 5.07371068e-01
9.41811800e-01 4.25773971e-02 1.26562500e+00 -9.73349631e-01
1.07585335e+00 1.22094500e+00 9.24061358e-01 6.43306494e-01
-8.24948072e-01 -6.45856857e-01 -1.76889300e-01 8.17318141e-01
-1.07059252e+00 -5.36440730e-01 7.33191907e-01 -9.25935745e-01
1.20268381e+00 2.97909170e-01 4.73629117e-01 1.42686367e+00
1.35802135e-01 8.69885325e-01 1.47744703e+00 -4.05445367e-01
3.18639994e-01 2.01608017e-01 2.64031231e-01 4.37585115e-01
-2.29458734e-01 -4.14559871e-01 -1.13890660e+00 -2.62624770e-01
2.32163280e-01 -6.40532315e-01 -1.34328172e-01 3.72424632e-01
-1.41394222e+00 1.10338640e+00 4.10652637e-01 5.27654827e-01
-7.53675550e-02 -2.85989158e-02 6.41776741e-01 1.74122587e-01
6.50875568e-01 7.19856024e-01 -1.29188016e-01 -7.03194320e-01
-7.75216043e-01 7.56909549e-01 9.50474083e-01 7.59911060e-01
1.39577776e-01 -3.74501020e-01 -3.71114194e-01 7.87837148e-01
-3.13625902e-01 1.04763746e-01 3.13394368e-01 -1.32517231e+00
7.83708155e-01 4.73226786e-01 1.96432307e-01 -1.34521306e+00
-2.52391189e-01 -2.15699479e-01 -3.89734447e-01 -7.30769988e-03
7.78145730e-01 4.31454647e-03 -1.69780001e-01 1.61106968e+00
-1.44806415e-01 -1.83536217e-01 4.67730723e-02 9.35278952e-01
9.25459504e-01 5.10706544e-01 6.05456866e-02 -1.93548277e-01
1.41017890e+00 -6.32257581e-01 -8.59373033e-01 -5.31647682e-01
8.49317431e-01 -5.59684396e-01 1.45503581e+00 5.02151608e-01
-1.20508707e+00 -1.76459357e-01 -1.26042187e+00 -5.06354153e-01
-2.56390542e-01 -1.54815968e-02 7.42576838e-01 4.06464487e-01
-6.84634328e-01 5.97187936e-01 -3.76236618e-01 -3.04476112e-01
8.73277605e-01 -4.26172137e-01 -2.48393744e-01 -2.80141294e-01
-1.29615605e+00 1.34999382e+00 4.02663767e-01 -1.83907643e-01
-9.24624264e-01 -7.35015571e-01 -6.18666768e-01 -7.67554119e-02
4.77442414e-01 -5.02101541e-01 1.38532376e+00 -7.08755076e-01
-1.06322134e+00 1.41202736e+00 -4.92866516e-01 -2.49700978e-01
6.88084304e-01 -2.78998405e-01 -2.24870488e-01 1.65248454e-01
7.04796076e-01 1.87480718e-01 3.43899727e-01 -1.22970343e+00
-5.86123765e-01 -2.58279055e-01 2.13240147e-01 1.24919929e-01
-1.19071506e-01 5.03675282e-01 4.64017540e-01 -5.58805883e-01
2.21681632e-02 -4.91648823e-01 4.43282396e-01 -3.79336953e-01
-3.37848306e-01 -5.98954618e-01 3.86611521e-01 -9.53827739e-01
1.16340613e+00 -1.87712550e+00 8.37336034e-02 -4.79489535e-01
3.92791122e-01 -3.26653928e-01 3.57329279e-01 3.91326427e-01
2.02615470e-01 4.74093705e-01 -2.20024332e-01 -2.08569616e-01
1.65343314e-01 8.84856135e-02 -7.03009307e-01 4.52462472e-02
2.93568224e-01 8.64951015e-01 -1.21660626e+00 -6.93249881e-01
-2.95356333e-01 -4.84259576e-02 -4.52546418e-01 9.14985910e-02
-3.59031886e-01 2.07720906e-01 -3.15151244e-01 2.96004117e-01
5.63795045e-02 -3.96303803e-01 9.01249275e-02 2.80136287e-01
9.90795717e-02 9.87756729e-01 -4.56325948e-01 1.45969319e+00
-3.17014664e-01 1.42166007e+00 -4.42568451e-01 -6.48298025e-01
7.92749643e-01 3.32986802e-01 -3.41867924e-01 -4.27735209e-01
2.12620884e-01 2.31549010e-01 1.97843704e-02 -8.50671589e-01
5.00816643e-01 -7.39848971e-01 -7.02780306e-01 8.24677646e-01
-9.94230434e-02 -5.06143570e-01 2.52413571e-01 5.58137894e-01
1.27744114e+00 -1.02914795e-01 2.97749341e-01 -2.79018581e-01
3.13279539e-01 6.77654028e-01 4.77970719e-01 8.97913933e-01
-5.69002070e-02 6.49318814e-01 1.01602316e+00 -5.98777771e-01
-1.14187121e+00 -1.08302975e+00 -3.45414318e-02 1.37014604e+00
-1.02051705e-01 -5.05734742e-01 -9.86265421e-01 -5.55008769e-01
-3.76999676e-01 1.63762689e+00 -8.57623994e-01 1.18827455e-01
-5.58770359e-01 -3.62335384e-01 7.67021239e-01 8.11824441e-01
4.83290851e-01 -1.13020074e+00 -6.51107132e-01 8.80138874e-02
-9.46970224e-01 -1.35346079e+00 -8.16730186e-02 -8.13236237e-02
-1.61918640e-01 -1.19361949e+00 -1.70315728e-02 -5.85190058e-01
1.04932956e-01 -6.01916797e-02 1.25592232e+00 2.52805334e-02
1.21327015e-02 3.20707224e-02 -5.97431302e-01 -5.56364298e-01
-6.87908351e-01 2.76757162e-02 -1.56361639e-01 -3.46003234e-01
7.62073457e-01 -5.10083616e-01 1.62881687e-01 9.74202976e-02
-4.04469520e-01 3.95491511e-01 -7.55381286e-02 8.41746986e-01
-1.11342750e-01 7.86565542e-02 6.67047203e-01 -9.65209186e-01
1.06957257e+00 -6.43893659e-01 3.70006293e-01 2.27948561e-01
-3.00642163e-01 -2.35750288e-01 5.27775407e-01 -3.56298745e-01
-1.30814087e+00 -6.75717533e-01 8.43196660e-02 1.23211518e-01
-2.56310981e-02 5.41349173e-01 -1.22465380e-01 8.33969653e-01
1.08107889e+00 -9.90824550e-02 -4.07102227e-01 -1.40824005e-01
1.73560366e-01 7.54380822e-01 8.84192944e-01 -8.72186065e-01
6.05926037e-01 3.94490480e-01 -5.61209381e-01 -3.39325339e-01
-1.83517873e+00 -5.70572838e-02 -7.26777852e-01 -6.03931129e-01
1.15106142e+00 -8.76736403e-01 -7.61540294e-01 3.24580997e-01
-1.64627004e+00 -5.71492851e-01 -1.30389839e-01 2.96760142e-01
-9.45286095e-01 -1.65824294e-01 -7.96894133e-01 -8.04789245e-01
-1.52211443e-01 -6.79655731e-01 7.07029760e-01 -3.50457802e-02
-1.08964002e+00 -9.19642448e-01 -8.35055392e-03 1.01884460e+00
1.09130144e-01 4.93599117e-01 1.33069468e+00 -7.94733584e-01
-2.08671361e-01 -1.48469493e-01 -2.45090947e-01 -1.13768183e-01
-2.78030753e-01 -2.06572101e-01 -1.14100778e+00 4.31978643e-01
2.53282309e-01 -9.86299574e-01 7.26146400e-01 3.21449414e-02
8.46782446e-01 -3.85837674e-01 -1.93902597e-01 6.46320209e-02
1.06151605e+00 4.57777828e-02 8.13962877e-01 5.65770447e-01
5.18927395e-01 7.10894048e-01 5.14484882e-01 3.71797085e-01
7.81111121e-01 4.35848475e-01 -9.94215310e-02 7.50629127e-01
-1.19989835e-01 -6.94541931e-01 5.08736432e-01 5.51491618e-01
-2.40937859e-01 -3.91246766e-01 -1.22490001e+00 7.88612127e-01
-1.93294024e+00 -1.54042101e+00 -4.11001831e-01 1.34056664e+00
1.12831438e+00 5.93683004e-01 -8.34263191e-02 5.20194888e-01
5.97317874e-01 3.25418591e-01 -3.22214186e-01 -6.79312468e-01
-4.10761923e-01 9.41065140e-03 -1.90543517e-01 6.85961962e-01
-7.60005593e-01 1.05062330e+00 7.10250759e+00 1.05150700e+00
-4.95759428e-01 5.61074972e-01 5.93051851e-01 -1.92989171e-01
-5.03708541e-01 9.77340192e-02 -5.31849265e-01 2.77065188e-01
7.08245516e-01 -5.45712531e-01 4.23156977e-01 7.68267632e-01
1.61140203e-01 -4.53757674e-01 -1.36044979e+00 8.12203169e-01
5.91593981e-01 -1.55280292e+00 -9.73952934e-03 -2.89449126e-01
7.92933464e-01 -4.73109454e-01 -3.97629976e-01 5.14470696e-01
4.49864089e-01 -1.38971829e+00 1.20739079e+00 5.56232929e-01
7.54839122e-01 -6.22743249e-01 8.50493431e-01 7.02188194e-01
-5.21547139e-01 -1.17142305e-01 -5.14608771e-02 -9.88560081e-01
2.11447582e-01 3.85898411e-01 -1.04647005e+00 -8.76590237e-02
6.55077755e-01 6.94700062e-01 -6.37588382e-01 2.12474898e-01
-8.46498728e-01 9.24875319e-01 2.10905999e-01 -4.88789886e-01
1.16923533e-01 4.06346142e-01 5.23423851e-01 1.20837879e+00
4.70708311e-02 6.33510351e-01 -1.25534743e-01 1.35265839e+00
-1.89280480e-01 -2.35935003e-01 -6.15517020e-01 -1.77625716e-01
6.29546225e-01 8.09461057e-01 -6.39222503e-01 -5.81145048e-01
1.25362203e-01 7.66146898e-01 7.18041718e-01 -3.44620161e-02
-7.92931437e-01 -2.57759064e-01 3.58716697e-01 3.42625469e-01
-6.64917156e-02 -1.07350484e-01 -1.14841628e+00 -1.00520873e+00
2.09118292e-01 -6.36501074e-01 4.04985696e-01 -1.51934075e+00
-1.62016392e+00 4.76464331e-01 1.58020034e-01 -6.72871351e-01
-5.52629650e-01 -5.36948383e-01 -9.39711154e-01 5.97443163e-01
-9.22726154e-01 -1.18981266e+00 -4.34183240e-01 3.42480510e-01
8.11810136e-01 -1.41809970e-01 7.82647908e-01 -3.07891726e-01
-1.57981679e-01 3.60496104e-01 -4.47522700e-01 6.13228142e-01
7.59139538e-01 -1.28911316e+00 2.43862092e-01 5.31503260e-01
-8.49157572e-02 3.38643044e-01 1.00931489e+00 -7.35836625e-01
-6.74646497e-01 -5.57374358e-01 1.53788483e+00 -1.26212049e+00
9.74773049e-01 -4.50725943e-01 -8.97869170e-01 9.14565742e-01
3.40596497e-01 -7.51272261e-01 9.60687459e-01 4.31599140e-01
-8.39264691e-01 4.65005547e-01 -1.08043683e+00 5.95332563e-01
1.33594811e+00 -1.00396121e+00 -1.41732323e+00 6.40958786e-01
6.23978078e-01 -4.26068515e-01 -7.60388970e-01 -5.80680184e-02
5.33961356e-01 -1.11098266e+00 5.64928293e-01 -9.74294960e-01
1.75252461e+00 4.95314561e-02 -2.48281956e-01 -1.22207057e+00
-3.50773454e-01 -3.32366079e-01 2.25608334e-01 1.30352652e+00
6.01609826e-01 -1.01350084e-01 6.09760582e-01 9.84442353e-01
-1.19328864e-01 -4.97796714e-01 -9.08574104e-01 -6.36755705e-01
5.81397355e-01 -9.92774665e-01 3.33999842e-01 1.25081968e+00
8.37486386e-01 8.58091176e-01 -1.02487266e-01 -3.00073981e-01
5.48577249e-01 1.97914578e-02 6.41776145e-01 -1.05316341e+00
-7.91369304e-02 -5.36705732e-01 -1.43811658e-01 -4.31096464e-01
5.52701831e-01 -1.11247349e+00 1.39197931e-01 -1.71933091e+00
8.66420269e-01 -8.77646580e-02 4.54677373e-01 5.29364288e-01
-3.64895999e-01 1.85937688e-01 2.07093865e-01 1.64375171e-01
-5.96692681e-01 1.31849378e-01 1.05954385e+00 -1.72033504e-01
1.20683379e-01 -4.96242911e-01 -9.14106667e-01 1.02672732e+00
8.80556285e-01 -5.14652908e-01 -2.64352232e-01 -6.45857811e-01
7.14597762e-01 1.00889876e-01 8.06768715e-01 -8.83517683e-01
3.82288665e-01 -3.31108302e-01 4.42399830e-01 -6.06884480e-01
3.60266417e-01 2.50019133e-02 8.01687390e-02 1.71714909e-02
-8.76477599e-01 -2.71069735e-01 1.85820647e-02 2.61880547e-01
-9.47017148e-02 -4.86437023e-01 4.68581587e-01 -3.58006984e-01
-6.04803681e-01 -6.57673955e-01 -6.31571591e-01 6.29685819e-01
8.60897124e-01 -3.66004586e-01 -7.15753973e-01 -7.21080601e-01
-7.50961304e-01 -1.89247783e-02 3.82767648e-01 4.05487925e-01
7.29769528e-01 -1.44885242e+00 -1.30730414e+00 -6.22372389e-01
2.46595785e-01 -1.63470522e-01 2.02504098e-01 5.65523088e-01
-3.94014835e-01 2.26439968e-01 -1.53746545e-01 -2.73843586e-01
-1.00936663e+00 4.55609232e-01 8.00571591e-02 -2.64257371e-01
-5.04787624e-01 9.81232643e-01 -1.05827563e-01 -5.42683750e-02
-2.33562931e-01 1.08620292e-02 -2.95975178e-01 2.35330150e-01
9.89834130e-01 5.23442984e-01 -2.43984088e-01 -6.21450782e-01
-3.20736319e-01 2.87597507e-01 -2.20135394e-02 -4.97372121e-01
1.43768084e+00 -2.12548207e-02 -2.16754809e-01 1.03784502e+00
9.98213112e-01 -4.38614599e-02 -7.36454964e-01 -1.56321317e-01
2.67033547e-01 -4.98378187e-01 -1.28814235e-01 -1.04649115e+00
-1.34072542e-01 9.61561918e-01 -5.69961905e-01 3.07032883e-01
6.50220931e-01 4.16989326e-01 8.76235604e-01 2.30402365e-01
5.62519133e-01 -1.23746622e+00 2.70704836e-01 7.39844918e-01
1.27240145e+00 -1.15872169e+00 6.73503280e-02 -4.85398293e-01
-1.18916214e+00 1.06233966e+00 5.86927176e-01 -4.48326655e-02
1.19777419e-01 4.15787727e-01 -1.46728262e-01 -4.21016634e-01
-1.11662078e+00 1.23775244e-01 1.39589254e-02 6.53389513e-01
4.91628140e-01 2.55700022e-01 -3.03255707e-01 1.20204854e+00
-1.05063212e+00 4.42939065e-02 8.36867929e-01 6.03678286e-01
-5.05251050e-01 -3.15664828e-01 -5.22136211e-01 5.05815327e-01
-3.76047313e-01 3.39096673e-02 -1.06226122e+00 6.52794242e-01
1.30734935e-01 1.43421280e+00 7.24372491e-02 -6.97365046e-01
9.49196666e-02 3.56577605e-01 6.99624836e-01 -5.72268903e-01
-6.76091611e-01 -6.64125919e-01 8.41969788e-01 -2.89112598e-01
-3.95567656e-01 -9.33653831e-01 -1.25584352e+00 -7.83510566e-01
-1.41240526e-02 1.25111058e-01 1.18575893e-01 1.49587262e+00
-1.57028943e-01 3.56396288e-01 3.98907095e-01 -5.93273759e-01
-3.05614501e-01 -1.20977557e+00 -2.57349133e-01 6.26052856e-01
-6.23926446e-02 -5.90821207e-01 -6.87336624e-01 4.85978782e-01] | [11.14405345916748, 8.832623481750488] |
6e801b58-dcdc-444e-93ab-6ce8949e6ecb | transcmd-cross-modal-decoder-equipped-with | 2112.02363 | null | https://arxiv.org/abs/2112.02363v3 | https://arxiv.org/pdf/2112.02363v3.pdf | CAVER: Cross-Modal View-Mixed Transformer for Bi-Modal Salient Object Detection | Most of the existing bi-modal (RGB-D and RGB-T) salient object detection methods utilize the convolution operation and construct complex interweave fusion structures to achieve cross-modal information integration. The inherent local connectivity of the convolution operation constrains the performance of the convolution-based methods to a ceiling. In this work, we rethink these tasks from the perspective of global information alignment and transformation. Specifically, the proposed \underline{c}ross-mod\underline{a}l \underline{v}iew-mixed transform\underline{er} (CAVER) cascades several cross-modal integration units to construct a top-down transformer-based information propagation path. CAVER treats the multi-scale and multi-modal feature integration as a sequence-to-sequence context propagation and update process built on a novel view-mixed attention mechanism. Besides, considering the quadratic complexity w.r.t. the number of input tokens, we design a parameter-free patch-wise token re-embedding strategy to simplify operations. Extensive experimental results on RGB-D and RGB-T SOD datasets demonstrate that such a simple two-stream encoder-decoder framework can surpass recent state-of-the-art methods when it is equipped with the proposed components. Code and pretrained models will be available at \href{https://github.com/lartpang/CAVER}{the link}. | ['Huchuan Lu', 'Lihe Zhang', 'Xiaoqi Zhao', 'Youwei Pang'] | 2021-12-04 | null | null | null | null | ['rgb-d-salient-object-detection'] | ['computer-vision'] | [ 3.02047372e-01 -1.02888420e-01 -4.59447987e-02 -4.36497092e-01
-7.93752730e-01 -3.76912355e-01 5.90245605e-01 4.80380282e-02
-6.60918653e-01 3.41873080e-01 1.82836562e-01 -3.14174205e-01
8.99265707e-03 -6.81674540e-01 -9.47065234e-01 -7.65861869e-01
2.46595189e-01 -6.42348826e-02 4.83930379e-01 -5.20205200e-01
1.42494813e-01 2.05922306e-01 -1.64319992e+00 6.01695895e-01
8.39899302e-01 1.28582573e+00 5.79945385e-01 5.45795679e-01
-4.20659631e-01 6.44382238e-01 -1.79651350e-01 -7.25414574e-01
1.60075724e-01 -2.38658920e-01 -6.87172234e-01 -1.13014325e-01
4.20245022e-01 -4.80126798e-01 -4.05118167e-01 1.10765934e+00
7.35267937e-01 -5.02978489e-02 3.68542790e-01 -1.30294240e+00
-9.97570097e-01 6.71018481e-01 -8.48483622e-01 4.00212079e-01
1.27249256e-01 2.19775468e-01 1.05470514e+00 -1.25440705e+00
5.71019828e-01 9.55691278e-01 4.64310199e-01 3.35780650e-01
-9.28734541e-01 -7.53657699e-01 6.00692689e-01 4.86781120e-01
-1.50369465e+00 -2.21443489e-01 9.83700037e-01 -1.93588719e-01
1.04579282e+00 3.06333214e-01 7.64737308e-01 9.50821459e-01
3.64283174e-02 1.06118453e+00 9.50829685e-01 -1.19872279e-01
-2.84610182e-01 7.48633668e-02 -6.87595084e-02 8.85162532e-01
-2.61485297e-02 -2.18898486e-02 -8.69985878e-01 4.22406077e-01
8.64586413e-01 1.07437067e-01 -2.67131805e-01 -1.71675906e-01
-1.48632550e+00 5.36919415e-01 9.62108672e-01 4.36714977e-01
-2.27648363e-01 3.52133483e-01 4.90849555e-01 8.47261697e-02
4.12131280e-01 -1.19193822e-01 -3.68592232e-01 -6.54247329e-02
-9.87930477e-01 -6.85124993e-02 1.37589216e-01 1.31036103e+00
8.19750965e-01 7.24166483e-02 -2.90962487e-01 6.96153700e-01
3.49694759e-01 2.05413952e-01 4.18150067e-01 -4.36416268e-01
7.07003415e-01 6.45058692e-01 -2.18094468e-01 -7.21818626e-01
-4.58204776e-01 -7.17112124e-01 -8.33561778e-01 1.39112532e-01
2.28338450e-01 1.63364649e-01 -9.50119078e-01 1.62673318e+00
5.15290439e-01 1.85591698e-01 -1.79234773e-01 1.20229495e+00
1.16708112e+00 6.94899440e-01 1.39622808e-01 1.84380859e-01
1.84071100e+00 -1.20497823e+00 -6.69421017e-01 -2.38069728e-01
3.23529124e-01 -8.29431117e-01 1.26320732e+00 2.26027127e-02
-1.17743111e+00 -7.56966293e-01 -1.19990587e+00 -8.00865591e-01
-6.43626511e-01 2.74357170e-01 6.08962417e-01 3.04927468e-01
-1.01488435e+00 2.44760796e-01 -7.73054659e-01 -1.55895546e-01
5.39071679e-01 1.95865124e-01 -3.92927498e-01 1.70309767e-02
-1.20661390e+00 8.30597043e-01 4.97426510e-01 5.08090258e-01
-7.57392228e-01 -8.40673566e-01 -8.65848005e-01 -3.80609669e-02
4.06587213e-01 -7.69631386e-01 1.02742183e+00 -7.28419304e-01
-1.32480609e+00 8.86033595e-01 -2.54670411e-01 -3.89736034e-02
5.01844943e-01 -2.55378157e-01 -4.80820924e-01 1.65556327e-01
4.09385040e-02 8.75790596e-01 7.88975596e-01 -1.26923692e+00
-8.26109827e-01 -3.73418391e-01 1.25813887e-01 4.24018979e-01
-3.26788455e-01 1.60636947e-01 -7.67021060e-01 -9.61668253e-01
2.05174953e-01 -5.29477656e-01 1.86300367e-01 2.81384408e-01
-5.04816413e-01 1.39666982e-02 7.50017881e-01 -8.06419790e-01
1.36363459e+00 -2.25670147e+00 2.55891651e-01 -1.13542728e-01
2.95038462e-01 1.41569436e-01 -7.12456405e-02 4.47035909e-01
-2.00544730e-01 2.59505417e-02 -3.73521298e-01 -5.91168702e-01
1.66308448e-01 -5.49737550e-02 -7.90698752e-02 4.43990082e-01
4.55022991e-01 1.07566977e+00 -8.00291121e-01 -6.14939809e-01
3.11779648e-01 7.52936363e-01 -4.99367565e-01 -3.51611786e-02
-2.96114460e-02 3.38021219e-01 -2.92570919e-01 9.01982188e-01
7.89900184e-01 -2.28265628e-01 -2.54398376e-01 -6.84674919e-01
-4.72260296e-01 3.46408576e-01 -1.11461842e+00 2.21834183e+00
-4.23453450e-01 4.79341179e-01 1.27782926e-01 -9.76365209e-01
8.05382133e-01 6.92250654e-02 2.96894252e-01 -8.72892737e-01
3.04703414e-01 2.87210852e-01 -1.20718352e-01 -3.02037656e-01
7.64662325e-01 2.50166107e-04 -1.25880256e-01 2.54448280e-02
2.58375078e-01 9.90432575e-02 -3.40881348e-02 2.43054241e-01
6.85013235e-01 5.41503489e-01 7.31394766e-03 -6.85114488e-02
5.28304160e-01 -2.30812967e-01 5.17397761e-01 4.42588329e-01
-1.41069502e-01 7.23727524e-01 2.57023782e-01 -2.42423028e-01
-9.95813668e-01 -1.14742422e+00 -1.44963235e-01 1.32612932e+00
5.24226129e-01 -3.74602705e-01 -4.86605912e-01 -4.12289172e-01
-2.14125767e-01 5.80861747e-01 -7.70985842e-01 -2.06992820e-01
-5.08528829e-01 -6.38243020e-01 4.98416156e-01 6.72611356e-01
8.60647857e-01 -8.38207603e-01 -7.51745522e-01 3.08337748e-01
-3.97768319e-01 -1.09580851e+00 -7.87974536e-01 3.25732529e-01
-5.66728652e-01 -5.37838519e-01 -7.83737063e-01 -8.82781684e-01
5.04040778e-01 1.65701225e-01 7.95327783e-01 2.22718492e-02
-4.11320448e-01 1.36531845e-01 -4.62976009e-01 -3.75660509e-01
1.23079233e-01 1.03769757e-01 -3.84256899e-01 1.55332163e-01
2.89592236e-01 -7.06420302e-01 -8.65453005e-01 2.58188903e-01
-1.02669919e+00 5.65858245e-01 6.98606551e-01 8.71811450e-01
8.77570391e-01 -2.99190342e-01 4.06924903e-01 -2.28177831e-01
2.15007454e-01 -3.08430225e-01 -3.42566371e-01 4.12221432e-01
-2.45777190e-01 5.53181320e-02 3.47518206e-01 -4.39365774e-01
-1.17212057e+00 1.95988510e-02 -3.03758562e-01 -6.74925447e-01
-9.40687768e-03 3.38345885e-01 -3.96592617e-01 1.19162463e-01
8.29999447e-02 6.59611940e-01 -2.40900412e-01 -5.35649717e-01
7.47024417e-01 4.35477406e-01 7.05390573e-01 -4.61300522e-01
7.32224584e-01 6.16012394e-01 -2.32338563e-01 -5.85052848e-01
-5.34900904e-01 -2.65878737e-01 -5.78117371e-01 -4.08540189e-01
1.15021086e+00 -1.08631277e+00 -7.59421289e-01 6.60844505e-01
-1.27172911e+00 -7.97842517e-02 -4.31074768e-01 3.75536323e-01
-3.86154801e-01 2.23721802e-01 -5.46659589e-01 -5.38329005e-01
-5.57064295e-01 -1.44742453e+00 1.32312644e+00 3.68821710e-01
2.04240456e-01 -4.69067603e-01 -4.50581759e-01 3.91680568e-01
4.97612536e-01 7.59634599e-02 7.20861852e-01 -3.29100907e-01
-8.99524629e-01 6.68380782e-02 -5.73080122e-01 2.75605649e-01
1.56105252e-03 -6.30203411e-02 -1.04253972e+00 -3.12422104e-02
-2.23068848e-01 -3.98021266e-02 1.16368842e+00 1.16999023e-01
1.24208987e+00 -2.11340412e-01 -1.77406102e-01 9.47864473e-01
1.40871143e+00 -3.39148492e-02 6.74996555e-01 3.64262015e-01
1.02717972e+00 3.93989950e-01 4.12816107e-01 5.06097257e-01
8.91932249e-01 7.45554984e-01 5.98835170e-01 -3.08635324e-01
-4.49440479e-01 -3.20270360e-01 3.13503534e-01 9.36655164e-01
-2.69964665e-01 -8.03961158e-02 -6.46466374e-01 6.12354994e-01
-1.70078802e+00 -1.04937708e+00 -8.17614049e-02 1.93121028e+00
1.09400487e+00 2.08401844e-01 1.48946547e-03 3.75465751e-02
6.76089764e-01 3.07198465e-01 -5.69343030e-01 -2.99962997e-01
-3.51054788e-01 1.51438519e-01 5.26282549e-01 4.11795676e-01
-1.09141421e+00 9.33588803e-01 3.97497177e+00 1.12372172e+00
-1.25065553e+00 4.48871642e-01 4.22598630e-01 -3.27060491e-01
-5.91896594e-01 -7.85059482e-02 -8.74861479e-01 5.15707076e-01
3.72361928e-01 1.78101435e-01 2.46744692e-01 3.93776208e-01
-6.19574860e-02 -1.22371323e-01 -8.79814982e-01 1.04694343e+00
-3.26093994e-02 -1.39071000e+00 9.09848288e-02 -2.00801238e-01
2.20544755e-01 1.68046117e-01 1.56669796e-01 2.29526043e-01
-1.57689396e-02 -5.99926293e-01 1.50784338e+00 5.10785699e-01
9.12406564e-01 -5.81436515e-01 4.94648367e-01 4.70234565e-02
-1.86289680e+00 1.66991167e-02 3.51450825e-03 3.18511963e-01
5.11106551e-01 5.92867374e-01 -2.84862369e-01 9.56970632e-01
1.09405291e+00 7.54347265e-01 -5.38344145e-01 8.24982643e-01
-2.75104731e-01 2.14536816e-01 -4.67750371e-01 3.01741082e-02
3.05883586e-01 -3.13493758e-02 6.63778365e-01 1.46465898e+00
2.71768749e-01 1.14689901e-01 -5.57487570e-02 1.03503311e+00
-5.10559268e-02 -1.92092676e-02 -1.20702676e-01 2.09740028e-01
4.07824904e-01 1.28742301e+00 -7.79818356e-01 -2.89403498e-01
-5.32431662e-01 1.18547642e+00 2.94839919e-01 3.21367145e-01
-1.31106758e+00 -6.03320062e-01 7.51514733e-01 2.22298317e-02
7.40398765e-01 -2.67219573e-01 -4.24640417e-01 -1.20500839e+00
3.82169485e-01 -4.34343338e-01 2.22839400e-01 -8.81693602e-01
-1.03226578e+00 7.47724235e-01 5.98242991e-02 -1.50001526e+00
5.19876003e-01 -4.44794118e-01 -3.79891187e-01 8.54775965e-01
-1.89221001e+00 -1.65403485e+00 -3.65118057e-01 8.88120711e-01
5.10731101e-01 2.10100576e-01 4.47836995e-01 6.88453972e-01
-6.41590178e-01 8.68188500e-01 -2.85032064e-01 1.45054623e-01
3.87685210e-01 -1.00899577e+00 2.10834876e-01 1.06748140e+00
9.83438045e-02 4.97746438e-01 4.07172948e-01 -4.46287423e-01
-1.45363712e+00 -1.04108739e+00 6.26373470e-01 -2.27713175e-02
6.98839366e-01 -5.38161218e-01 -8.41254711e-01 6.12170517e-01
5.12210369e-01 1.91523150e-01 5.23604512e-01 -3.34350049e-01
-5.98938107e-01 -2.38773853e-01 -9.40230787e-01 6.23453081e-01
1.33054686e+00 -6.95266843e-01 -4.74951744e-01 -6.05276972e-02
1.11424816e+00 -5.52997828e-01 -1.00070059e+00 5.25936544e-01
6.44577801e-01 -1.05482650e+00 1.09057188e+00 -1.96262524e-01
5.90517163e-01 -6.92272186e-01 -3.12443018e-01 -8.32422972e-01
-2.67440945e-01 -5.57601750e-01 -2.83689555e-02 1.41102207e+00
4.61818904e-01 -5.85310936e-01 2.05382764e-01 2.73461550e-01
-5.26298642e-01 -1.02471137e+00 -1.20373964e+00 -4.12883997e-01
-1.48064345e-01 -5.43898940e-01 6.86906695e-01 7.84587324e-01
-1.53183326e-01 1.20263055e-01 -2.44052380e-01 2.34660268e-01
5.90557277e-01 1.47072077e-01 3.38528514e-01 -6.75900638e-01
-2.12333381e-01 -6.25020146e-01 -3.43562841e-01 -1.25324345e+00
-3.33301753e-01 -1.18808305e+00 -1.42761886e-01 -1.62991440e+00
7.09624365e-02 -5.16499579e-01 -5.25302708e-01 7.74730504e-01
-1.56211808e-01 3.19982290e-01 4.50535387e-01 -6.65553287e-03
-5.35076261e-01 9.31100428e-01 1.38568544e+00 -1.86223373e-01
-4.91969101e-03 -4.75805253e-01 -7.01268911e-01 4.57125276e-01
7.49931276e-01 -2.79930443e-01 -3.25050265e-01 -8.45107675e-01
2.13240832e-01 -2.25373805e-01 6.48260891e-01 -7.80974746e-01
4.32002455e-01 1.26514599e-01 3.43204230e-01 -9.49343503e-01
6.32362306e-01 -8.43829036e-01 1.87497437e-02 1.06507033e-01
-2.40014181e-01 3.04638833e-01 4.13719952e-01 3.53164524e-01
-2.99235642e-01 2.04492480e-01 6.98218405e-01 -1.41729459e-01
-8.59322846e-01 4.13486302e-01 -9.60882157e-02 4.18234691e-02
1.00344586e+00 -4.58496571e-01 -5.61770082e-01 -2.05028746e-02
-6.05819881e-01 1.66308850e-01 2.09343195e-01 6.82327747e-01
8.19380343e-01 -1.35797155e+00 -6.24038458e-01 2.52979666e-01
2.20693514e-01 2.03302547e-01 6.90822542e-01 1.21744120e+00
-3.44566464e-01 1.21610455e-01 -3.47929776e-01 -5.58199704e-01
-1.10865617e+00 4.99380350e-01 3.30823034e-01 -9.15238485e-02
-7.37710834e-01 1.29409468e+00 2.54498273e-01 -2.67714530e-01
2.16655537e-01 -5.07022560e-01 5.98348752e-02 2.90291578e-01
4.46925193e-01 1.37361154e-01 9.82012078e-02 -8.33842695e-01
-5.37044346e-01 5.47317624e-01 -1.78505376e-01 -2.93742102e-02
1.35725808e+00 -3.26092869e-01 -2.39119098e-01 3.35425079e-01
1.33393002e+00 -3.11167657e-01 -1.23235512e+00 -3.93208385e-01
-4.67151940e-01 -3.02468389e-01 1.93285793e-01 -7.49256849e-01
-1.44288790e+00 1.11860120e+00 7.14597344e-01 -9.67022926e-02
1.41032970e+00 5.32219410e-02 7.94337034e-01 -8.14118311e-02
2.56389797e-01 -1.07516146e+00 -2.76319720e-02 2.06947640e-01
1.02587402e+00 -1.09949875e+00 -4.99711521e-02 -4.88702893e-01
-7.24759817e-01 8.96718025e-01 6.53975964e-01 1.83168277e-01
8.41287613e-01 4.33700711e-01 -2.26542264e-01 -1.92915991e-01
-6.22862339e-01 -4.71245766e-01 2.50869483e-01 3.19877386e-01
3.11914742e-01 5.51213734e-02 -2.60331988e-01 7.35994518e-01
-1.13651842e-01 9.29383710e-02 1.36885092e-01 1.04515469e+00
-2.82965600e-01 -9.15763080e-01 -1.29827857e-01 1.96858138e-01
-2.45229855e-01 -5.15350163e-01 1.38587818e-01 7.60219097e-01
5.06703496e-01 7.48536527e-01 1.47326896e-02 -4.63141561e-01
4.55501676e-01 7.92153627e-02 4.34111983e-01 -5.35732992e-02
-7.98247159e-01 2.50859112e-01 -7.42787868e-02 -6.18274093e-01
-6.46922469e-01 -5.94322205e-01 -1.48409331e+00 -1.84921145e-01
-2.88298935e-01 -4.68750209e-01 6.18721426e-01 6.94016457e-01
5.46149135e-01 9.52058911e-01 2.64196157e-01 -9.39455152e-01
-1.60720900e-01 -8.66128922e-01 -4.19888109e-01 2.54610389e-01
4.37496722e-01 -7.47310638e-01 -6.47256523e-02 1.55072182e-01] | [9.701240539550781, -0.7358705401420593] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.