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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
f646edb7-d0c1-433d-b6ad-060b99ef946a | zhou-qiaoli-a-divide-and-conquer-strategy-for | null | null | https://aclanthology.org/S12-1072 | https://aclanthology.org/S12-1072.pdf | Zhou qiaoli: A divide-and-conquer strategy for semantic dependency parsing | null | ['Guiping Zhang', 'Fei Liu', 'Qiaoli Zhou', 'Ling Zhang', 'Dongfeng Cai'] | 2012-07-01 | null | null | null | semeval-2012-7 | ['semantic-dependency-parsing'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.280211448669434, 3.6812686920166016] |
fa4f48e4-6908-4a21-b3d1-2a81b4eb6fc1 | trace-encoding-in-process-mining-a-survey-and | 2301.02167 | null | https://arxiv.org/abs/2301.02167v1 | https://arxiv.org/pdf/2301.02167v1.pdf | Trace Encoding in Process Mining: a survey and benchmarking | Encoding methods are employed across several process mining tasks, including predictive process monitoring, anomalous case detection, trace clustering, etc. These methods are usually performed as preprocessing steps and are responsible for transforming complex information into a numerical feature space. Most papers choose existing encoding methods arbitrarily or employ a strategy based on a specific expert knowledge domain. Moreover, existing methods are employed by using their default hyperparameters without evaluating other options. This practice can lead to several drawbacks, such as suboptimal performance and unfair comparisons with the state-of-the-art. Therefore, this work aims at providing a comprehensive survey on event log encoding by comparing 27 methods, from different natures, in terms of expressivity, scalability, correlation, and domain agnosticism. To the best of our knowledge, this is the most comprehensive study so far focusing on trace encoding in process mining. It contributes to maturing awareness about the role of trace encoding in process mining pipelines and sheds light on issues, concerns, and future research directions regarding the use of encoding methods to bridge the gap between machine learning models and process mining. | ['Gabriel M. Tavares', 'Rafael S. Oyamada', 'Paolo Ceravolo', 'Sylvio Barbon Jr.'] | 2023-01-05 | null | null | null | null | ['predictive-process-monitoring'] | ['time-series'] | [ 6.97032332e-01 -4.29570898e-02 -1.55481592e-01 -2.71757215e-01
-3.58258903e-01 -3.85590166e-01 6.17381632e-01 9.91007864e-01
-2.98663259e-01 5.34730017e-01 -2.16796383e-01 -5.87308407e-01
-6.70884550e-01 -1.07954383e+00 -5.16277589e-02 -4.56267476e-01
-2.22819835e-01 7.05403805e-01 3.19368660e-01 2.87202805e-01
6.79126263e-01 5.97101510e-01 -1.78924239e+00 3.21815670e-01
5.25816023e-01 1.03502166e+00 -3.62180650e-01 4.41688687e-01
-5.69929123e-01 1.17915428e+00 -7.60622680e-01 -5.86141288e-01
-8.02295879e-02 -5.26841581e-01 -8.37921441e-01 1.87881440e-01
-3.35633457e-01 9.80282724e-02 -5.20505719e-02 8.75332594e-01
2.11404413e-01 1.80490129e-02 5.37076294e-01 -1.32446980e+00
-4.43678468e-01 8.38382661e-01 -4.93431300e-01 5.11450708e-01
7.27951825e-01 2.44628042e-01 9.01291192e-01 -2.51502603e-01
5.52252293e-01 1.03577459e+00 6.07298672e-01 3.35199267e-01
-1.22749245e+00 -5.54058373e-01 7.83481002e-02 3.32802594e-01
-1.20427632e+00 -3.62528086e-01 5.11336684e-01 -5.00884950e-01
1.30334055e+00 4.10090894e-01 4.95051235e-01 9.89473820e-01
2.85848320e-01 7.46040046e-01 1.09637952e+00 -5.01919508e-01
5.64770043e-01 7.75834769e-02 3.49924713e-01 5.38848400e-01
6.54409349e-01 -1.63056910e-01 -7.65263617e-01 -4.78864759e-01
5.49196839e-01 1.04578778e-01 -1.35392159e-01 -1.89385563e-01
-9.79783475e-01 7.55385935e-01 -6.41179442e-01 4.78789866e-01
-4.28109884e-01 -7.86206871e-02 7.16360033e-01 6.16169751e-01
1.34401053e-01 6.09911323e-01 -4.67213452e-01 -7.60584176e-01
-1.00719845e+00 2.74435133e-01 1.36692834e+00 8.61128986e-01
6.26991153e-01 -1.31145939e-01 -4.51093674e-01 4.77869838e-01
3.85761321e-01 -2.46727899e-01 5.93290269e-01 -6.04757011e-01
2.48970285e-01 1.00439131e+00 -2.05701277e-01 -7.47482777e-01
-2.62584448e-01 -2.83331424e-01 -6.12289310e-01 1.52745962e-01
3.53101939e-01 -1.30223423e-01 -5.77695370e-01 1.24790776e+00
-9.77781508e-03 9.56605747e-02 -1.69639841e-01 3.08694750e-01
2.99097568e-01 3.17961961e-01 3.47534955e-01 -5.58515847e-01
1.53208697e+00 -5.38723528e-01 -1.00524735e+00 -1.87792733e-01
5.06245077e-01 -7.48038709e-01 8.62640440e-01 7.07202554e-01
-1.19390559e+00 -3.11152130e-01 -1.09283054e+00 4.26818073e-01
-4.72501576e-01 -1.06720716e-01 1.11059606e+00 1.10100591e+00
-6.26021206e-01 9.56080019e-01 -1.24950123e+00 -5.92973351e-01
5.02692044e-01 3.12186748e-01 -1.13022096e-01 1.57947242e-01
-1.00135016e+00 6.99258983e-01 5.87489963e-01 -1.57943815e-01
-5.23314953e-01 -6.66864038e-01 -7.21617222e-01 9.56298485e-02
7.66998291e-01 -6.51602864e-01 1.22364998e+00 -2.63424635e-01
-1.41193509e+00 6.16774738e-01 -1.85414404e-01 -7.35790849e-01
5.85776389e-01 -1.47627532e-01 -6.79355979e-01 -1.12180457e-01
-9.13028866e-02 -3.74284774e-01 4.45991248e-01 -7.11889446e-01
-8.43715310e-01 -3.82231802e-01 -2.12194443e-01 -2.84505844e-01
-2.14271083e-01 3.53769779e-01 -3.87841046e-01 -4.89983112e-01
1.85575411e-01 -5.76388240e-01 -4.73426163e-01 -3.88678789e-01
-4.23130423e-01 -3.13870311e-01 7.19228566e-01 -2.33695775e-01
1.78833604e+00 -2.08615088e+00 -1.95669591e-01 4.08561856e-01
1.26361132e-01 3.95372622e-02 4.87166971e-01 1.04064572e+00
1.70286700e-01 4.17387307e-01 -3.44830036e-01 -2.96855122e-01
1.81738704e-01 2.87713438e-01 -2.48461217e-01 3.26354355e-01
4.36444253e-01 5.96067488e-01 -8.72687399e-01 -5.36628842e-01
2.65544504e-01 1.69908911e-01 -6.16885088e-02 1.81760505e-01
-7.19653443e-02 2.47529045e-01 -4.63458419e-01 8.74864817e-01
3.28812301e-01 -2.70267636e-01 3.41252655e-01 2.11434379e-01
-2.63948500e-01 4.30830657e-01 -1.60186291e+00 1.39483643e+00
-1.57508001e-01 3.96536738e-01 -3.18948507e-01 -8.41460168e-01
1.04646146e+00 4.19450730e-01 8.55969489e-01 -4.52706218e-01
5.02924584e-02 5.28561100e-02 2.05400154e-01 -5.49352586e-01
4.67242628e-01 -1.96561124e-02 3.54013668e-04 8.44206393e-01
1.50174124e-03 2.85535574e-01 8.61701190e-01 -2.06332207e-01
1.62402439e+00 8.21563378e-02 6.98684990e-01 -2.11034995e-02
8.00332844e-01 -8.82706419e-02 6.18828773e-01 8.80999506e-01
-1.88693345e-01 3.12315226e-01 1.03918481e+00 -4.06434298e-01
-7.26852417e-01 -8.94558966e-01 -1.76714480e-01 8.74008775e-01
-2.61097193e-01 -1.03523135e+00 -4.57572728e-01 -6.30461633e-01
5.79865184e-03 8.84619653e-01 -6.53757632e-01 -2.73229539e-01
-6.28578365e-01 -9.88822937e-01 9.77636695e-01 7.22574949e-01
4.20011103e-01 -1.10891223e+00 -1.06449378e+00 6.21864736e-01
1.75764829e-01 -9.53925490e-01 2.48771146e-01 5.42062163e-01
-1.22473896e+00 -1.33071053e+00 2.46673375e-01 -1.46326602e-01
1.83112681e-01 -4.11266953e-01 1.22165561e+00 -3.23979147e-02
-3.12275290e-01 1.18758298e-01 -4.96884793e-01 -7.48329282e-01
-5.54196000e-01 2.10785359e-01 -2.96817124e-01 -1.22588903e-01
1.08446467e+00 -6.23146355e-01 -2.09274843e-01 2.43956745e-01
-1.12311780e+00 -5.25209248e-01 8.96801353e-01 4.19805557e-01
4.69334900e-01 6.01654410e-01 4.20014679e-01 -1.27483940e+00
1.19808865e+00 -5.33139169e-01 -2.71131694e-01 3.14509988e-01
-1.15472949e+00 3.22656542e-01 4.18128878e-01 -3.55734468e-01
-1.20254338e+00 -1.58351064e-01 2.09589805e-02 -9.88656506e-02
-6.47426665e-01 5.06415009e-01 -2.48086795e-01 3.18153888e-01
4.55826432e-01 3.50818157e-01 -5.57757244e-02 -3.38227510e-01
-1.46829784e-01 3.35495621e-01 1.24623597e-01 -6.58730507e-01
4.65365082e-01 5.78169644e-01 1.94935188e-01 -5.20113111e-01
-3.14771563e-01 -6.64507926e-01 -5.26435673e-01 7.36332461e-02
4.98725474e-01 -1.35981575e-01 -7.52334535e-01 3.33929211e-01
-7.88279355e-01 4.43227589e-02 -5.45044184e-01 3.32207382e-01
-5.59939802e-01 4.53414857e-01 -9.64287937e-01 -1.07086110e+00
-4.05274481e-01 -1.06132317e+00 8.09301615e-01 -1.48375612e-02
-8.12661588e-01 -9.62546527e-01 6.64846599e-02 2.07327098e-01
3.74107987e-01 3.23068649e-01 9.63665843e-01 -1.16366982e+00
-3.42885405e-01 -2.96915472e-01 1.88240603e-01 -3.84012680e-03
3.40203553e-01 1.92154333e-01 -1.01070845e+00 -1.89149927e-03
1.83136821e-01 5.53339310e-02 5.74181318e-01 1.49451736e-02
1.30042672e+00 -8.23881403e-02 -3.91108900e-01 3.87641490e-01
1.34436512e+00 5.99128425e-01 7.79160559e-01 7.34524071e-01
1.43039748e-01 6.68290377e-01 7.74179161e-01 1.00378180e+00
-2.48520868e-03 1.92240566e-01 3.33633989e-01 4.56855983e-01
3.27059090e-01 -3.65922153e-02 1.72731370e-01 3.07662487e-01
-4.40956682e-01 -2.24164516e-01 -1.03880119e+00 3.62180799e-01
-1.81251526e+00 -1.24761081e+00 -3.04090321e-01 2.15425587e+00
6.66756094e-01 5.52800179e-01 2.40230262e-02 9.40065980e-01
4.46526825e-01 -1.48551418e-02 -2.44244441e-01 -6.77522957e-01
2.12602064e-01 4.58245307e-01 5.29543459e-01 -1.23869427e-01
-1.15909457e+00 4.96940762e-01 5.98828220e+00 5.82856238e-01
-7.98449039e-01 -5.27116247e-02 2.53157765e-01 -1.72213167e-02
-9.99540314e-02 1.73520833e-01 -8.06264699e-01 4.61057603e-01
1.22989416e+00 -2.69992024e-01 1.35675579e-01 6.56983614e-01
2.11752117e-01 -1.98725581e-01 -1.57342792e+00 8.99887800e-01
-9.69320983e-02 -1.16182613e+00 -7.23534599e-02 2.20290795e-01
3.21060240e-01 -4.83261794e-01 -1.72929108e-01 3.62192720e-01
2.50822425e-01 -1.22136092e+00 5.67040920e-01 6.18113458e-01
5.60399331e-02 -9.10183311e-01 9.73115981e-01 9.20063034e-02
-1.22124600e+00 -2.22346604e-01 -5.38998619e-02 -3.13389927e-01
5.13736010e-01 8.18859577e-01 -9.45026040e-01 9.29684281e-01
8.26707244e-01 5.31438231e-01 -5.57763934e-01 1.04706633e+00
-5.57580441e-02 9.02042747e-01 -1.15107596e-01 -4.74403873e-02
1.46364555e-01 -1.71343580e-01 5.53522527e-01 1.54919803e+00
2.23368600e-01 -4.34923977e-01 1.71394497e-01 8.71118069e-01
4.10193950e-01 1.93455130e-01 -5.66588819e-01 -2.77431965e-01
5.69436312e-01 1.00074911e+00 -9.81068254e-01 -2.07168967e-01
-5.69254518e-01 7.24388421e-01 -3.33938785e-02 9.42539722e-02
-6.75702155e-01 -3.79831016e-01 7.20851004e-01 4.19742644e-01
1.62285060e-01 -1.37065336e-01 -6.65726244e-01 -7.28981018e-01
1.39694020e-01 -9.13101792e-01 8.59378457e-01 -1.90170370e-02
-1.28255773e+00 4.06622291e-01 1.12763450e-01 -9.82282996e-01
-4.36524749e-01 -4.55876857e-01 -5.94978750e-01 5.67828655e-01
-1.26700342e+00 -8.17401171e-01 -2.61072338e-01 3.81009758e-01
5.80607951e-01 -7.58090988e-02 8.80789876e-01 3.22261602e-01
-6.64958835e-01 3.11362237e-01 -3.19085211e-01 -8.50443393e-02
7.75242209e-01 -1.34770775e+00 2.23396614e-01 7.78067410e-01
-5.66008687e-02 8.63729417e-01 7.70302653e-01 -7.61832833e-01
-1.44088006e+00 -9.91100669e-01 8.12924623e-01 -4.54113156e-01
8.54556262e-01 -6.35681301e-02 -1.17708993e+00 8.06890130e-01
1.97514012e-01 -4.04130578e-01 1.27324390e+00 3.12184185e-01
6.08329512e-02 4.21944782e-02 -1.21557415e+00 5.15398085e-01
9.87912118e-01 -2.66630113e-01 -5.26898324e-01 -2.22187743e-01
6.51241019e-02 -4.63353842e-02 -1.37432170e+00 3.54338914e-01
3.29846054e-01 -1.09362650e+00 6.21962130e-01 -6.25218689e-01
2.61712790e-01 -3.36822242e-01 3.99210751e-02 -7.58711815e-01
-3.99824530e-01 -7.05093801e-01 -6.07296050e-01 1.66108930e+00
3.81035000e-01 -5.72841525e-01 7.13839591e-01 5.41878939e-01
6.53234720e-02 -8.50346625e-01 -5.90457439e-01 -6.67841434e-01
-3.34997475e-01 -8.25488865e-01 1.00381029e+00 9.81120884e-01
1.21084154e-01 2.22485453e-01 -3.51790413e-02 3.43858451e-02
6.00516021e-01 5.66482842e-02 5.12176812e-01 -1.62745559e+00
-3.14888477e-01 -6.44221067e-01 -6.75160587e-01 -2.88887441e-01
-1.38251811e-01 -5.92565775e-01 -4.54171896e-01 -1.57138717e+00
5.43531450e-03 -3.47070605e-01 -4.62122411e-01 4.74064946e-01
7.51232058e-02 -1.87479213e-01 -6.71523586e-02 2.70194024e-01
-6.08059943e-01 -5.91952093e-02 9.56532836e-01 2.75728479e-02
-3.89123917e-01 2.58703798e-01 -8.55194807e-01 8.07612181e-01
1.27361834e+00 -6.02817178e-01 -6.62207007e-01 -4.77635413e-02
4.75576341e-01 -8.84889439e-02 1.47773504e-01 -1.02658558e+00
6.34806275e-01 -4.13334876e-01 2.82652944e-01 -4.34621722e-01
-2.89489869e-02 -8.70430112e-01 4.75723326e-01 5.99833965e-01
-3.58538568e-01 3.63747478e-01 7.43234381e-02 7.20236599e-01
-3.85047346e-01 -6.84742689e-01 2.80381352e-01 -4.74384993e-01
-9.94807184e-01 9.68836099e-02 -8.08090925e-01 -1.05711453e-01
1.38109159e+00 -6.52757823e-01 -4.40481566e-02 -8.58014897e-02
-6.10318661e-01 -9.96920988e-02 4.06603158e-01 3.19498301e-01
3.88463318e-01 -7.68918693e-01 -4.83181149e-01 2.53009886e-01
3.44519585e-01 -1.03079803e-01 -6.24444522e-02 9.75476265e-01
-3.81916702e-01 2.48018056e-01 -2.23802537e-01 -4.28087831e-01
-1.16145074e+00 7.08794773e-01 -1.55043276e-02 -8.25697064e-01
-6.00716174e-01 2.01269493e-01 -5.89373410e-01 -8.50920528e-02
1.91819578e-01 -3.23247612e-01 -3.35034549e-01 2.06131116e-01
6.63911104e-01 9.13884819e-01 5.14183700e-01 1.56353906e-01
-5.54228365e-01 9.66982245e-02 -3.55604924e-02 -1.11006968e-01
1.07760894e+00 4.52798158e-02 -4.32555467e-01 7.29038417e-01
4.47345227e-01 1.01399064e-01 -7.96156168e-01 1.01370476e-02
8.32007766e-01 -4.52745795e-01 -2.80475259e-01 -6.23893678e-01
-8.05193961e-01 6.22104347e-01 3.59132320e-01 5.70142925e-01
1.21458805e+00 6.41490221e-02 2.74339825e-01 3.69529247e-01
3.33173305e-01 -1.22003031e+00 -1.16389662e-01 4.75134313e-01
2.76378810e-01 -8.50307107e-01 7.46440217e-02 -6.26015007e-01
-6.30838454e-01 1.30439198e+00 5.47680259e-01 2.33676910e-01
6.43280149e-01 8.11676562e-01 -2.37823159e-01 -2.74383783e-01
-1.01663530e+00 -1.78060710e-01 -2.16038302e-01 6.17442131e-01
8.21687579e-01 -6.16225712e-02 -6.80894077e-01 6.09753847e-01
-2.43183866e-01 2.51096785e-01 4.30831671e-01 1.63305748e+00
-3.31729531e-01 -1.69588733e+00 -4.31365669e-01 8.64262640e-01
-1.01544130e+00 9.34759825e-02 -3.66445035e-01 8.38202477e-01
2.02875689e-01 1.16205859e+00 1.13548525e-01 -1.07210658e-01
5.65913260e-01 3.73110473e-01 4.36575502e-01 -5.78324318e-01
-9.13051426e-01 -2.94133157e-01 4.44697350e-01 -4.87682760e-01
-2.89974660e-01 -1.08456481e+00 -1.03781939e+00 -4.37806427e-01
-2.01575737e-02 2.54393697e-01 4.00364906e-01 8.34925711e-01
2.03301206e-01 9.86983836e-01 -6.13242425e-02 5.65007180e-02
-6.18073523e-01 -1.02729356e+00 -5.61302781e-01 4.62434769e-01
-3.19170654e-01 -6.29077315e-01 -2.49270033e-02 8.56613740e-02] | [8.59028434753418, 6.01471471786499] |
63b033d9-556e-410d-b145-a2df0f0fcb93 | stop-words-for-processing-software | 2303.10439 | null | https://arxiv.org/abs/2303.10439v2 | https://arxiv.org/pdf/2303.10439v2.pdf | Stop Words for Processing Software Engineering Documents: Do they Matter? | Stop words, which are considered non-predictive, are often eliminated in natural language processing tasks. However, the definition of uninformative vocabulary is vague, so most algorithms use general knowledge-based stop lists to remove stop words. There is an ongoing debate among academics about the usefulness of stop word elimination, especially in domain-specific settings. In this work, we investigate the usefulness of stop word removal in a software engineering context. To do this, we replicate and experiment with three software engineering research tools from related work. Additionally, we construct a corpus of software engineering domain-related text from 10,000 Stack Overflow questions and identify 200 domain-specific stop words using traditional information-theoretic methods. Our results show that the use of domain-specific stop words significantly improved the performance of research tools compared to the use of a general stop list and that 17 out of 19 evaluation measures showed better performance. Online appendix: https://zenodo.org/record/7865748 | ['Christoph Treude', 'Chetan Arora', 'Yaohou Fan'] | 2023-03-18 | null | null | null | null | ['general-knowledge'] | ['miscellaneous'] | [ 2.15494066e-01 -6.86822459e-02 -2.38492742e-01 -1.09943278e-01
-6.47577524e-01 -6.93269014e-01 2.15905741e-01 5.19213915e-01
-5.39330184e-01 5.63506961e-01 4.01010990e-01 -8.72703135e-01
-4.86752152e-01 -5.76984823e-01 -2.50742644e-01 2.44819261e-02
2.96903670e-01 8.24231580e-02 3.48675013e-01 -1.48124084e-01
1.15033233e+00 -2.65276618e-02 -1.64730215e+00 1.61721572e-01
1.14372182e+00 2.88960993e-01 3.05968434e-01 5.60231626e-01
-8.26630473e-01 5.82361341e-01 -7.65342951e-01 -3.96236062e-01
2.98778284e-02 -2.96149999e-01 -9.63718772e-01 -2.16463879e-01
3.64893287e-01 1.13680810e-01 -6.56718835e-02 1.17195988e+00
3.61916184e-01 3.15103047e-02 3.91651541e-01 -9.42973733e-01
-6.11755311e-01 8.37350428e-01 -3.35325092e-01 5.83252788e-01
7.04651237e-01 1.42971650e-01 1.26074171e+00 -8.36835206e-01
7.71957755e-01 9.65039909e-01 6.06581271e-01 3.27129990e-01
-1.01859319e+00 -6.63673282e-01 -1.14910966e-02 9.92024541e-02
-1.39029539e+00 -3.00354987e-01 6.05037689e-01 -7.70591617e-01
1.30572546e+00 2.46827886e-01 4.48629171e-01 7.37673759e-01
3.89390171e-01 4.59426582e-01 7.38903224e-01 -1.07498658e+00
1.73908487e-01 5.24792261e-02 1.01905286e+00 4.65667278e-01
9.07719851e-01 -2.19646588e-01 -4.33010876e-01 -5.17194748e-01
2.99354970e-01 -1.10745445e-01 -1.73179582e-01 -5.01382835e-02
-1.04049397e+00 7.27336884e-01 -6.47905409e-01 8.31054926e-01
-2.33021215e-01 -1.27259225e-01 3.83087188e-01 7.07058430e-01
3.82396519e-01 1.10119593e+00 -5.77260554e-01 -6.54210210e-01
-9.61536288e-01 2.55594343e-01 1.34431362e+00 1.15512419e+00
7.24942565e-01 -1.72709867e-01 -2.75656164e-01 8.01725805e-01
3.84181619e-01 2.58271694e-01 5.61190724e-01 -1.03736496e+00
3.80722761e-01 9.71773803e-01 1.80418283e-01 -8.57755005e-01
-2.92800874e-01 -3.02571267e-01 9.12854373e-02 -5.98488599e-02
4.39566314e-01 -2.59797156e-01 -7.32007563e-01 1.40373755e+00
-2.26827890e-01 -4.28570122e-01 -1.93405837e-01 4.29746866e-01
5.65362930e-01 2.46135771e-01 2.28649497e-01 -4.60598290e-01
1.43962443e+00 -5.36411285e-01 -1.02863312e+00 -5.95854998e-01
9.70727265e-01 -1.15581596e+00 1.40748525e+00 5.25173664e-01
-8.47256660e-01 -1.64327081e-02 -9.98491228e-01 -3.71885765e-03
-4.45104331e-01 -1.14595383e-01 6.83758497e-01 9.55040038e-01
-7.86304712e-01 3.21709335e-01 -5.09444773e-01 -4.47064400e-01
4.43749353e-02 -8.93968195e-02 -1.20436708e-02 -1.15421616e-01
-1.18331170e+00 9.08623219e-01 4.49841827e-01 -7.59902954e-01
-2.18362361e-01 -7.51024067e-01 -6.88081264e-01 6.92561492e-02
9.33902442e-01 -5.26539803e-01 1.37773776e+00 -7.18922794e-01
-9.45386767e-01 8.37306082e-01 -2.58364081e-01 -1.12233363e-01
3.18096317e-02 -5.21144986e-01 -3.74349177e-01 -2.69049332e-02
2.26555362e-01 -2.87738949e-01 6.68107986e-01 -7.71818221e-01
-6.68191552e-01 -1.81050375e-01 6.88651726e-02 -1.17798142e-01
-5.89365244e-01 4.79189664e-01 -3.26516062e-01 -6.53225780e-01
-1.03185929e-01 -7.09189653e-01 -2.48033434e-01 -5.25773287e-01
2.12316262e-03 -3.73914868e-01 2.87138313e-01 -6.52764916e-01
2.28770971e+00 -2.20953631e+00 -4.15433288e-01 1.67490944e-01
3.70031357e-01 2.53356189e-01 -6.74783811e-02 5.73468029e-01
-2.75639594e-01 7.10482776e-01 -2.14509666e-01 3.00548643e-01
1.26950704e-02 -2.18223378e-01 -1.55250847e-01 1.88522875e-01
-8.99816453e-02 6.31611168e-01 -1.04513001e+00 -6.64027572e-01
2.07447678e-01 2.27628332e-02 -5.76468825e-01 -2.99274415e-01
-3.28514516e-01 -4.38128471e-01 -6.92209363e-01 7.25258470e-01
2.22468585e-01 -1.63266048e-01 2.39395007e-01 4.33274716e-01
-4.03140992e-01 7.77924180e-01 -1.11936748e+00 1.42861676e+00
-4.27078098e-01 7.50496805e-01 -2.03635603e-01 -4.35508698e-01
8.44520390e-01 3.01613539e-01 3.02079827e-01 -5.68450332e-01
1.90823764e-01 2.40626052e-01 1.82189599e-01 -6.98648930e-01
6.74500048e-01 -1.50325865e-01 -1.81552514e-01 6.00839496e-01
-6.21655844e-02 -3.46204072e-01 7.69591689e-01 1.75679520e-01
1.63248634e+00 -2.43512884e-01 6.67272866e-01 -4.26998883e-01
5.84837794e-01 2.26748943e-01 5.94745100e-01 9.74025488e-01
-1.60916746e-01 4.68964219e-01 6.03254378e-01 -7.47726038e-02
-7.75889933e-01 -6.86260641e-01 -1.37027815e-01 1.27018476e+00
-1.90814883e-01 -1.18111169e+00 -8.14331293e-01 -8.19672704e-01
1.31416291e-01 1.51602042e+00 -9.87175331e-02 -2.22420961e-01
-5.27686954e-01 -3.88797581e-01 5.45680821e-01 1.77040994e-01
-4.82092127e-02 -1.09965229e+00 -5.80218136e-01 3.82073700e-01
-1.97186813e-01 -8.35198522e-01 -5.15824795e-01 2.31633812e-01
-9.03237283e-01 -1.15750909e+00 -3.46419752e-01 -4.85731661e-01
4.61940229e-01 3.45686913e-01 1.41831434e+00 5.53989708e-01
-2.64213324e-01 6.62070990e-01 -6.98240876e-01 -7.48159409e-01
-7.18234062e-01 3.74999553e-01 -2.41890833e-01 -9.94223535e-01
1.42434239e+00 -3.01560968e-01 -1.16983719e-01 1.66014910e-01
-1.06071901e+00 -5.46177268e-01 6.05230570e-01 4.37200963e-01
-2.40062978e-02 4.00642790e-02 6.23417735e-01 -8.99626374e-01
1.31209826e+00 -3.90893430e-01 -6.58254266e-01 3.40756088e-01
-1.09831870e+00 1.85373142e-01 3.85906011e-01 -3.55028361e-01
-8.28030109e-01 -3.17372143e-01 -8.64339154e-03 1.65103719e-01
-2.14005768e-01 9.04863477e-01 1.29435599e-01 -2.89685503e-02
8.93457770e-01 1.21205077e-02 -6.05607294e-02 -4.74278212e-01
5.24193682e-02 8.83524776e-01 -2.29091242e-01 -6.48105085e-01
6.56055331e-01 -1.97340325e-01 -4.35177445e-01 -1.11071646e+00
-6.82385981e-01 -7.14096844e-01 -3.00844073e-01 -1.19833022e-01
4.71412122e-01 -3.82221848e-01 -2.20004290e-01 -4.34544086e-02
-9.12303329e-01 2.10392885e-02 -3.50011408e-01 6.10911489e-01
-3.49199206e-01 7.91717827e-01 -1.76336795e-01 -8.36422145e-01
-3.04403633e-01 -9.21781957e-01 5.86761713e-01 1.58117071e-01
-9.03345227e-01 -8.21931124e-01 9.42493454e-02 3.65666211e-01
3.30345333e-01 -3.66579473e-01 1.25370419e+00 -1.03944290e+00
-2.07426429e-01 -4.83596772e-01 2.36792535e-01 4.62439328e-01
2.77466923e-01 1.98708832e-01 -3.50496620e-01 1.31719401e-02
2.67556369e-01 2.47826040e-01 6.91519201e-01 2.08970159e-01
8.22741687e-01 -8.94911736e-02 -2.73790836e-01 2.47795526e-02
1.37128115e+00 2.96930790e-01 5.48343182e-01 7.60502279e-01
2.87010640e-01 7.18874812e-01 1.04727221e+00 5.45097113e-01
6.64639845e-02 1.06785543e-01 -1.08326174e-01 7.20667124e-01
1.06653623e-01 6.53480589e-02 3.07257116e-01 1.07733440e+00
2.03528300e-01 -2.77534306e-01 -1.53489542e+00 9.27627325e-01
-1.40162063e+00 -6.47270560e-01 -3.43837678e-01 2.25877929e+00
1.03162944e+00 6.31482959e-01 -5.78450970e-02 2.51466602e-01
7.40430832e-01 2.22661737e-02 -6.16245568e-02 -4.86392587e-01
2.40630254e-01 2.31517076e-01 6.73519909e-01 4.08901602e-01
-8.23587894e-01 8.99990499e-01 6.78704882e+00 7.96154141e-01
-6.78103626e-01 -6.25111088e-02 -1.23395115e-01 -7.01636523e-02
-5.49194992e-01 3.34936708e-01 -7.58283257e-01 4.42168891e-01
1.25621927e+00 -1.06201637e+00 1.76507920e-01 1.06859124e+00
1.54534355e-01 -3.32398891e-01 -1.04167140e+00 6.17351532e-01
1.33064473e-02 -9.99557436e-01 -1.30682230e-01 -6.85756356e-02
4.50000256e-01 -1.33861974e-01 -2.13034272e-01 4.22209054e-01
1.94268033e-01 -7.93791533e-01 4.59996015e-01 5.33266962e-01
3.01972359e-01 -5.77759922e-01 7.23035216e-01 4.17557925e-01
-6.32483482e-01 -8.21420550e-02 -2.40940213e-01 -4.37513024e-01
-1.52576059e-01 9.07871246e-01 -9.02231097e-01 3.90365183e-01
5.16573429e-01 3.17591250e-01 -6.98382020e-01 1.19894969e+00
-2.50929207e-01 1.04300463e+00 -3.08399528e-01 -4.68142748e-01
6.02486022e-02 -1.21357232e-01 8.52829635e-01 1.58822739e+00
1.24470681e-01 -1.43739238e-01 -1.98001012e-01 6.52126253e-01
1.05669841e-01 3.86714458e-01 -7.35995591e-01 -6.09045684e-01
7.35601902e-01 8.70124757e-01 -8.23539555e-01 -1.64434955e-01
-8.07627380e-01 2.32281208e-01 -2.34885871e-01 2.88542002e-01
-3.93780977e-01 -9.96510506e-01 6.50787354e-01 3.77382874e-01
2.54259735e-01 -4.36515093e-01 -7.06365466e-01 -1.04626417e+00
4.15244877e-01 -1.20711660e+00 3.62389356e-01 -3.56458038e-01
-1.10849237e+00 1.33667558e-01 1.23007193e-01 -1.14809263e+00
-2.07807064e-01 -5.42257786e-01 -3.15366179e-01 7.41293252e-01
-1.09570098e+00 -1.65205792e-01 -2.02029020e-01 -1.23266883e-01
4.90371466e-01 -9.62720290e-02 5.98735154e-01 3.44543695e-01
-2.32974678e-01 4.46650594e-01 1.45211399e-01 3.54579724e-02
8.46852899e-01 -1.21133542e+00 5.16050398e-01 9.01850700e-01
-7.81946331e-02 1.47154880e+00 1.07549608e+00 -1.11862683e+00
-1.47629237e+00 -6.74611807e-01 1.22802401e+00 -6.79414332e-01
9.54558969e-01 -1.58596821e-02 -1.22412312e+00 6.32035494e-01
2.27599189e-01 -6.54740870e-01 8.55858564e-01 3.26279432e-01
-2.77422786e-01 3.61584485e-01 -1.02930975e+00 7.80038714e-01
9.95546639e-01 -5.55811703e-01 -1.46285295e+00 1.89138293e-01
8.03110659e-01 9.45135728e-02 -8.47754180e-01 2.14560553e-01
3.87674689e-01 -5.25763869e-01 6.47925019e-01 -5.10190427e-01
3.39247227e-01 -2.66306609e-01 -2.54981238e-02 -8.65154386e-01
-3.23980600e-01 -7.83785403e-01 8.78723860e-02 1.19295633e+00
5.89295447e-01 -6.22739017e-01 5.16955256e-01 8.23227465e-01
-1.01183832e-01 -3.24366033e-01 -7.22944856e-01 -9.30877388e-01
3.64233971e-01 -6.82293177e-01 2.75360137e-01 9.85902727e-01
4.76783007e-01 3.63039553e-01 3.08484316e-01 -3.76693428e-01
4.43284303e-01 -1.35203615e-01 4.51940060e-01 -1.39874721e+00
1.76100805e-01 -6.77309692e-01 -3.62571537e-01 -7.67502844e-01
2.29419991e-01 -8.14968526e-01 3.23793441e-02 -1.58193433e+00
7.57661611e-02 4.98914346e-02 -2.32748821e-01 3.69031340e-01
-3.84934068e-01 -5.07378519e-01 1.87894404e-02 -3.90021242e-02
-7.28500724e-01 1.11965872e-01 7.89297998e-01 1.27771363e-01
-3.90129775e-01 -4.84936833e-02 -1.10135090e+00 1.01211762e+00
7.88912177e-01 -8.18646312e-01 -3.07912081e-01 -2.39291131e-01
7.21695602e-01 -1.96224287e-01 -2.20613003e-01 -7.68656552e-01
3.09516817e-01 -4.09483910e-01 -3.94286573e-01 -3.92492175e-01
-4.33806539e-01 -8.42469335e-01 -1.62985578e-01 3.49797785e-01
-4.71890539e-01 1.68683343e-02 3.92524064e-01 2.08214700e-01
-1.41405150e-01 -1.25112092e+00 3.68898600e-01 -2.51798421e-01
-7.63822317e-01 -4.78399336e-01 -9.88246500e-01 5.81619859e-01
9.61078703e-01 -4.60567594e-01 -3.58743936e-01 -6.38060868e-02
-2.89444089e-01 1.64142400e-01 6.28645360e-01 7.06715465e-01
5.30391037e-01 -6.21891439e-01 -6.03002012e-01 -1.17839888e-01
4.36816514e-01 -4.39769000e-01 -2.50853121e-01 6.49982333e-01
-5.70191324e-01 6.59871578e-01 1.74936801e-01 -9.13715661e-02
-1.40491295e+00 5.60624480e-01 -1.24212623e-01 -7.63518810e-02
-6.20705247e-01 3.61128479e-01 -2.22707942e-01 -2.16539845e-01
3.22776258e-01 -5.11902630e-01 -2.17299432e-01 3.40958655e-01
6.65480077e-01 5.00911474e-01 2.98433572e-01 3.59775946e-02
-6.51555777e-01 3.00931722e-01 -4.22392815e-01 -1.84113324e-01
1.23934627e+00 -1.41050816e-01 -3.36559892e-01 7.23516047e-01
8.85736108e-01 3.38489890e-01 -9.77736637e-02 -2.18534410e-01
1.09962678e+00 -6.74203694e-01 1.07601546e-01 -7.98507452e-01
-4.63788539e-01 3.19779932e-01 2.44036019e-01 5.32760799e-01
9.58765805e-01 -1.24631703e-01 3.34573239e-01 8.11190069e-01
4.55811054e-01 -1.37409663e+00 -2.13261545e-01 1.04012728e+00
6.69759512e-01 -8.85726571e-01 1.53746590e-01 -4.99150127e-01
-3.37877423e-01 1.18621254e+00 5.25624037e-01 2.04762593e-01
7.14320064e-01 3.90818268e-01 -3.31729166e-02 -2.94411987e-01
-8.68525028e-01 -2.48924121e-01 -2.09723487e-02 3.17283809e-01
9.58001435e-01 -2.58402973e-01 -1.27304709e+00 8.59516144e-01
-3.09368104e-01 4.50770438e-01 9.64373410e-01 1.63713324e+00
-7.27712691e-01 -1.23022711e+00 -4.09756750e-01 9.30985391e-01
-1.00588953e+00 -6.68393135e-01 -6.40661716e-01 7.52990901e-01
-3.04071218e-01 1.22130501e+00 -1.66825220e-01 -3.02633613e-01
4.38182443e-01 6.73259497e-01 1.99262515e-01 -1.07509911e+00
-8.60811234e-01 -2.13235736e-01 3.80056620e-01 -2.44943842e-01
-1.06068186e-01 -9.55612957e-01 -1.13225484e+00 -3.67413431e-01
-7.08186150e-01 5.02972722e-01 6.95208013e-01 9.71711397e-01
4.68985409e-01 7.45163620e-01 2.45391689e-02 3.37035596e-01
-6.07497275e-01 -1.04729688e+00 -3.06369960e-01 2.10289463e-01
2.18268156e-01 -7.32614040e-01 -8.10853899e-01 1.64843183e-02] | [7.8046488761901855, 7.9750189781188965] |
fa1937a2-41a6-407f-a8d2-5de528c7f918 | personalized-ppg-normalization-based-on | 2202.11465 | null | https://arxiv.org/abs/2202.11465v1 | https://arxiv.org/pdf/2202.11465v1.pdf | Personalized PPG Normalization based on Subject Heartbeat in Resting State Condition | Physiological responses are nowadays widely used to recognize the affective state of subjects in real-life scenarios. However, these data are intrinsically subject-dependent, making machine learning techniques for data classification not easily applicable due to inter-subject variability. In this work, the reduction of inter-subject heterogeneity is considered in the case of PhotoPlethysmoGraphy (PPG), which is successfully used to detect stress and evaluate experienced cognitive load. To face the inter-subject heterogeneity, a novel personalized PPG normalization is here proposed. A subject-normalized discrete domain where the PPG signals are properly re-scaled is introduced, considering the subject's heartbeat frequency in resting state conditions. The effectiveness of the proposed normalization is evaluated in comparison with other normalization procedures in a binary classification task, where cognitive load and relaxing state are considered. The results obtained on two different datasets available in the literature confirm that applying the proposed normalization strategy permits to increase classification performance. | ['Stefania Bandini', 'Marta Giltri', 'Alessandra Grossi', 'Francesca Gasparini'] | 2022-02-23 | null | null | null | null | ['photoplethysmography-ppg'] | ['medical'] | [ 5.70033967e-01 -5.20121567e-02 -8.28944817e-02 -4.09081250e-01
-1.01014577e-01 -3.43021870e-01 3.46760184e-01 2.61873245e-01
-6.75115287e-01 8.52094173e-01 1.04157545e-01 4.34190482e-01
-2.73477584e-01 -4.24790293e-01 1.18958140e-02 -9.23734009e-01
2.53817458e-02 -1.42701089e-01 -3.79563481e-01 -4.42136414e-02
3.22730929e-01 5.48809469e-01 -1.86089313e+00 -2.19551921e-01
1.00673723e+00 1.06368446e+00 -1.03425361e-01 3.24266553e-01
2.33747661e-01 -6.58966899e-02 -9.26709890e-01 -3.06738447e-02
1.11803688e-01 -7.40459263e-01 -3.41886997e-01 9.76518989e-02
1.54456303e-01 3.86715889e-01 1.69782147e-01 1.02443326e+00
9.62242901e-01 3.99779558e-01 4.68766063e-01 -9.20977771e-01
-3.46756428e-02 5.85963614e-02 -2.72979885e-01 5.10534167e-01
5.23297846e-01 -1.55721426e-01 5.86370289e-01 -5.69680512e-01
4.53285456e-01 6.61501884e-01 5.32840252e-01 3.72615725e-01
-1.52543890e+00 -4.13116604e-01 -2.72197515e-01 6.96533501e-01
-1.37161577e+00 -4.08260375e-01 1.28794777e+00 -5.81052721e-01
5.42474926e-01 4.57476705e-01 8.25530112e-01 1.13751268e+00
6.31262302e-01 -1.43159762e-01 1.88261306e+00 -4.61330652e-01
5.33028245e-01 4.17065889e-01 2.88891286e-01 -2.58395225e-02
3.42602104e-01 -9.95715931e-02 -6.56729281e-01 -2.10364148e-01
1.39414415e-01 -2.78435528e-01 -5.12240052e-01 -2.66333967e-01
-9.23172653e-01 4.63084459e-01 -6.13753684e-02 9.93035138e-01
-7.22618997e-01 -4.39760208e-01 6.85691655e-01 2.38110617e-01
6.92212403e-01 4.74777073e-01 -1.50747418e-01 -2.99669623e-01
-1.08230197e+00 1.36267068e-02 7.57671773e-01 5.11564016e-02
4.58640128e-01 2.87109971e-01 -3.96635056e-01 7.77874231e-01
-1.85559884e-01 5.69035947e-01 9.06845868e-01 -5.50434887e-01
-8.14648941e-02 6.75483406e-01 4.56744730e-02 -1.28458345e+00
-9.59792912e-01 -8.41348827e-01 -9.94005263e-01 1.85297668e-01
3.86325747e-01 -1.53987378e-01 -3.51249546e-01 1.90165234e+00
4.35691595e-01 2.32209802e-01 2.04365954e-01 1.08560050e+00
6.29119575e-01 2.25334272e-01 5.95595121e-01 -8.43332589e-01
1.68448830e+00 2.00870950e-02 -1.19935083e+00 -2.10974053e-01
1.25782982e-01 -7.05524683e-01 1.11034155e+00 4.93023962e-01
-7.76566744e-01 -7.59083092e-01 -1.12741613e+00 4.31731582e-01
-4.27139670e-01 2.61582106e-01 1.11188807e-01 1.07332170e+00
-6.96916878e-01 7.31275797e-01 -6.07648849e-01 -7.10277855e-01
3.76541317e-02 1.97851107e-01 -3.96336704e-01 3.50078821e-01
-1.49818790e+00 1.16514075e+00 4.50571209e-01 4.04924512e-01
8.54968000e-03 -6.33461058e-01 -4.71747160e-01 2.21048653e-01
1.57884315e-01 -4.01885122e-01 5.80768108e-01 -1.02445686e+00
-1.97200429e+00 1.07129860e+00 -1.91839561e-01 -8.26761350e-02
5.31237304e-01 2.03858688e-01 -8.43555152e-01 2.22530484e-01
-3.01492423e-01 -2.09341660e-01 9.69327748e-01 -6.82401001e-01
2.84775764e-01 -7.04347074e-01 -4.12772328e-01 3.83826613e-01
-4.85769480e-01 -1.16404504e-01 3.23898166e-01 -4.13790137e-01
3.40770274e-01 -9.87369716e-01 1.42591700e-01 -6.38466299e-01
-1.14229180e-01 8.23300108e-02 3.94315958e-01 -6.70192897e-01
1.15728295e+00 -2.18784165e+00 1.17712073e-01 5.77761114e-01
-1.93618312e-02 3.47431511e-01 2.31359124e-01 2.36540869e-01
-3.27295363e-01 -3.57357651e-01 -2.36685917e-01 -1.49859622e-01
8.10924545e-02 1.55614883e-01 1.16972469e-01 7.79694498e-01
-1.29889563e-01 4.72379088e-01 -4.65687990e-01 -3.41140389e-01
5.86883664e-01 6.54371560e-01 -5.07824728e-03 2.52692580e-01
3.58447582e-01 8.12644958e-01 -1.34441614e-01 2.23794103e-01
7.90527523e-01 2.94302911e-01 5.33975661e-01 -5.68589211e-01
-1.51302621e-01 -2.46400908e-01 -1.01678097e+00 1.38757885e+00
-5.24477065e-01 5.26902676e-01 -5.90019040e-02 -1.19748628e+00
1.57372332e+00 6.04126990e-01 6.74544334e-01 -1.07895696e+00
6.18853390e-01 2.14363694e-01 3.63866091e-01 -7.51478136e-01
2.99843818e-01 -5.63581526e-01 -3.21211740e-02 1.69534028e-01
-1.47435322e-01 -3.43573350e-03 2.76341587e-01 -5.02454281e-01
6.32097065e-01 -3.79753597e-02 7.97641218e-01 -7.44265854e-01
1.08222926e+00 -6.00977898e-01 5.54931819e-01 3.89378399e-01
-5.39207876e-01 3.44338030e-01 6.33995295e-01 -2.99823791e-01
-6.03104532e-01 -6.70660317e-01 -3.47264260e-01 7.46688068e-01
-1.78843699e-02 1.04669213e-01 -9.18013752e-01 1.49794519e-02
-8.42304230e-02 6.32130921e-01 -7.82350242e-01 -4.56258625e-01
-1.94936171e-01 -1.43136978e+00 2.94658303e-01 1.30234614e-01
4.17452782e-01 -1.06637657e+00 -1.09537578e+00 4.18453813e-01
-3.09262872e-01 -1.13571465e+00 1.90573558e-01 8.57350901e-02
-9.97805893e-01 -1.00290704e+00 -7.65427589e-01 -4.64213043e-02
2.65537232e-01 -2.63486415e-01 9.64107811e-01 -3.18121701e-01
-6.45496905e-01 6.30122960e-01 -3.09004992e-01 -3.48643661e-01
-3.85831356e-01 2.02226400e-01 2.39937022e-01 5.50843418e-01
5.74376523e-01 -9.85082448e-01 -7.33817399e-01 2.83548921e-01
-7.57540703e-01 -4.78383213e-01 5.07046521e-01 3.39805245e-01
4.14035797e-01 -2.12507963e-01 1.07044911e+00 -7.98751175e-01
9.82017279e-01 -2.35465869e-01 -3.42914492e-01 7.17198327e-02
-9.69105899e-01 -4.08481270e-01 5.70962906e-01 -3.30742329e-01
-1.17378294e+00 -2.23311529e-01 5.23701310e-02 6.07429855e-02
-5.57676136e-01 3.92790467e-01 -2.65791357e-01 -3.43856007e-01
9.09528315e-01 1.69532537e-01 1.61960110e-01 -1.98685840e-01
-2.40242973e-01 4.73080486e-01 6.64624631e-01 -4.47299033e-01
3.23545307e-01 1.20095290e-01 4.96526867e-01 -1.07359207e+00
-5.12443721e-01 -3.30964953e-01 -8.07538211e-01 -5.75042725e-01
8.52925718e-01 -5.47472179e-01 -9.92192090e-01 4.28636312e-01
-8.25688899e-01 1.08813904e-01 -3.15642774e-01 9.11166668e-01
-5.65409124e-01 4.12469625e-01 -1.14386141e-01 -1.01034772e+00
-6.71964347e-01 -7.17571914e-01 2.88935870e-01 5.25042832e-01
-4.00248200e-01 -9.51984644e-01 2.65434384e-01 2.97123224e-01
6.66490495e-01 6.94996417e-01 6.86787128e-01 -6.22938931e-01
3.73694897e-01 -4.80174661e-01 2.04034045e-01 5.74435174e-01
2.53293663e-01 -4.42233950e-01 -1.33124208e+00 -1.11071669e-01
5.45300782e-01 1.24754362e-01 2.42669225e-01 4.41107422e-01
8.68450463e-01 1.59543335e-01 2.61605054e-01 1.27742633e-01
1.40861785e+00 2.64529109e-01 9.97801304e-01 3.01429275e-02
1.20431520e-01 8.36493194e-01 5.15172541e-01 7.51105309e-01
-1.07989959e-01 6.60128236e-01 9.04581547e-02 -2.37932160e-01
2.54601508e-01 5.78194141e-01 2.37082288e-01 6.48334324e-01
-5.52845061e-01 -4.74999845e-02 -4.66780096e-01 7.11945072e-02
-1.47110653e+00 -1.00688612e+00 -2.55784690e-01 2.51057339e+00
5.75627089e-01 -4.21581119e-02 1.60087749e-01 5.51599085e-01
9.08804715e-01 2.15818599e-01 -4.37885851e-01 -7.14960873e-01
-3.57500464e-01 5.80099642e-01 2.46497080e-01 3.39870363e-01
-8.46411467e-01 8.11782554e-02 5.79718018e+00 2.69014478e-01
-1.56973040e+00 1.59610823e-01 3.81782383e-01 -1.52202800e-01
1.98988929e-01 -4.20366615e-01 -2.32214585e-01 7.12254047e-01
1.23579049e+00 -6.37729883e-01 3.01032782e-01 5.64423025e-01
7.86328852e-01 -4.52325076e-01 -6.83635771e-01 1.24983799e+00
4.84435081e-01 -3.90799493e-01 -8.15004587e-01 -1.10287957e-01
2.65401602e-01 -6.04906738e-01 -2.02166699e-02 2.64406409e-02
-1.27189827e+00 -5.62932312e-01 1.83405146e-01 9.74233627e-01
5.43856084e-01 -6.88079357e-01 9.98728931e-01 -3.32760699e-02
-6.61919117e-01 1.26284257e-01 -9.53482017e-02 -2.77493864e-01
1.20736696e-01 1.12163746e+00 -7.39034593e-01 7.11896598e-01
2.73147255e-01 4.48187202e-01 -5.72359145e-01 8.99553955e-01
-6.53934926e-02 5.57627976e-01 -1.89705387e-01 -6.23232266e-03
-4.07535791e-01 -4.97342259e-01 7.65016735e-01 1.17526686e+00
5.11942983e-01 2.13678285e-01 -2.63061255e-01 9.34985876e-01
3.48181874e-01 7.42740214e-01 -3.74358386e-01 -4.87326831e-03
4.67789680e-04 1.57739079e+00 -9.36971188e-01 -2.81245589e-01
-5.45551740e-02 7.66372025e-01 -5.18840313e-01 2.30670258e-01
-7.32046664e-01 -5.11031270e-01 5.16229808e-01 1.58956900e-01
-3.93181562e-01 5.97039424e-02 -3.78158897e-01 -1.02319229e+00
8.61846432e-02 -4.62169975e-01 4.15333897e-01 -5.55607319e-01
-1.06489480e+00 5.73088765e-01 2.44358256e-01 -1.25092137e+00
-2.13856936e-01 -2.98339844e-01 -5.58493197e-01 1.18033576e+00
-1.38036263e+00 -5.42355657e-01 -6.57983720e-01 4.59383339e-01
-4.15607058e-02 2.13298276e-01 9.13767874e-01 6.90265656e-01
-6.46055281e-01 5.55075407e-01 -2.70429939e-01 -7.02899218e-01
9.28481400e-01 -1.01339674e+00 -5.52109897e-01 7.64713943e-01
-5.87908089e-01 5.74541152e-01 1.07476461e+00 -4.64886516e-01
-9.44202125e-01 -4.73969609e-01 9.40009177e-01 2.32625172e-01
2.14380220e-01 -3.48256201e-01 -1.17104089e+00 -6.11829907e-02
1.85210630e-01 1.18522108e-01 9.55603123e-01 4.19153571e-02
9.56170335e-02 -5.42519748e-01 -1.46199262e+00 2.34678224e-01
4.23474878e-01 -4.73157555e-01 -7.27333069e-01 8.44100937e-02
-1.92187473e-01 -2.38543704e-01 -1.42337525e+00 4.10174936e-01
6.88719451e-01 -1.15483499e+00 6.66130781e-01 -1.35915503e-01
-1.59887105e-01 -1.19068585e-01 2.17837706e-01 -1.43789136e+00
-1.41058847e-01 -5.34218848e-01 2.44821683e-01 1.24206793e+00
-1.11477256e-01 -1.12398779e+00 6.44632757e-01 9.69172001e-01
-3.78446765e-02 -2.33525217e-01 -1.03181374e+00 -6.06710196e-01
-4.67953831e-01 9.05902416e-04 2.16283560e-01 9.97608006e-01
4.83253151e-01 1.35810420e-01 -4.93174434e-01 -2.94442654e-01
5.61377227e-01 8.66137818e-02 4.58126277e-01 -1.66439390e+00
-1.16273284e-01 -2.62028217e-01 -8.67579281e-01 3.35555077e-01
3.02347660e-01 -7.71627188e-01 -9.84926224e-02 -8.31426382e-01
7.16997460e-02 3.80565487e-02 -7.47212887e-01 2.13159457e-01
-9.41408649e-02 6.92395866e-01 1.10410362e-01 -1.49338320e-01
1.12607203e-01 5.82405269e-01 9.81626749e-01 3.64814043e-01
-7.01631606e-01 9.93483961e-02 -5.31787813e-01 4.10737365e-01
9.57592010e-01 -3.62704664e-01 -5.12787580e-01 5.73261559e-01
-5.39013371e-03 3.11366171e-02 2.46443704e-01 -1.46774423e+00
-1.85352087e-01 3.19690704e-02 4.47937638e-01 -1.95784464e-01
4.49810833e-01 -8.76344979e-01 5.70066750e-01 5.37520587e-01
-2.49328896e-01 -1.25217542e-01 2.69688159e-01 2.38577977e-01
-2.08447218e-01 -2.61326075e-01 1.05729067e+00 6.09511323e-02
-4.64941651e-01 -2.52409011e-01 -4.68846411e-01 -1.49809837e-01
1.02835321e+00 -3.92278641e-01 -1.78872511e-01 -2.24858701e-01
-1.09606612e+00 -5.50029933e-01 1.31633893e-01 1.63970575e-01
1.54071614e-01 -1.08103359e+00 -4.93831784e-01 1.34980857e-01
2.35410452e-01 -9.57863331e-01 1.12063348e+00 1.61925757e+00
-6.36797473e-02 2.47799322e-01 -8.29017162e-01 -5.19151449e-01
-1.34110272e+00 4.64356631e-01 3.79675955e-01 4.45954986e-02
-3.81615609e-01 -9.35968533e-02 -1.94634631e-01 2.22888827e-01
-1.68670580e-01 -2.04191506e-01 -7.11981118e-01 6.61278665e-01
3.83589238e-01 7.11462438e-01 6.26673996e-01 -6.49673223e-01
-3.60009700e-01 6.44269466e-01 4.89274144e-01 3.76443602e-02
1.05128801e+00 -4.60914373e-01 -2.36298516e-01 6.32033288e-01
9.58211601e-01 -2.74922382e-02 -6.30473197e-01 4.49690223e-02
4.88151200e-02 -4.78463322e-01 8.76935348e-02 -8.48909080e-01
-8.98666859e-01 9.25914288e-01 1.26043653e+00 2.37341166e-01
1.51650083e+00 -7.17897177e-01 2.28777096e-01 1.87808990e-01
2.01093122e-01 -1.41757023e+00 -2.40119472e-01 -1.79863021e-01
8.61861527e-01 -7.35434532e-01 1.24832608e-01 -4.48616832e-01
-6.62962258e-01 1.26462483e+00 2.97757953e-01 -6.12680726e-02
5.14700830e-01 -2.48089600e-02 9.78910998e-02 7.79556250e-03
-2.52265096e-01 -1.78778544e-01 2.43447855e-01 5.42204916e-01
5.68059683e-01 9.25923064e-02 -1.34170425e+00 7.00961888e-01
-6.46513179e-02 4.05093759e-01 6.81603134e-01 7.15204716e-01
-1.57449916e-01 -9.50141609e-01 -5.87013185e-01 2.92672962e-01
-5.40609300e-01 4.13711935e-01 -1.23018734e-01 7.04770565e-01
1.63684666e-01 9.96160746e-01 -7.30426311e-02 -1.77369595e-01
8.58602822e-01 7.90915787e-01 4.53783721e-01 -2.86981791e-01
-7.93349683e-01 8.52180496e-02 -1.67857468e-01 -5.55013776e-01
-9.37237084e-01 -6.86970711e-01 -9.66764152e-01 -3.29592079e-02
1.65831856e-03 2.09448978e-01 9.68767822e-01 8.92752707e-01
4.63995069e-01 5.28412879e-01 7.12025881e-01 -9.02348816e-01
-2.22091570e-01 -1.09212816e+00 -1.01690245e+00 7.17728317e-01
-1.15525415e-02 -9.64826882e-01 -5.61245620e-01 1.26564816e-01] | [13.678200721740723, 3.0299031734466553] |
87575c6b-e0af-4ba5-8383-e69031522d84 | target-specific-and-selective-drug-design-for | 2004.01215 | null | https://arxiv.org/abs/2004.01215v2 | https://arxiv.org/pdf/2004.01215v2.pdf | CogMol: Target-Specific and Selective Drug Design for COVID-19 Using Deep Generative Models | The novel nature of SARS-CoV-2 calls for the development of efficient de novo drug design approaches. In this study, we propose an end-to-end framework, named CogMol (Controlled Generation of Molecules), for designing new drug-like small molecules targeting novel viral proteins with high affinity and off-target selectivity. CogMol combines adaptive pre-training of a molecular SMILES Variational Autoencoder (VAE) and an efficient multi-attribute controlled sampling scheme that uses guidance from attribute predictors trained on latent features. To generate novel and optimal drug-like molecules for unseen viral targets, CogMol leverages a protein-molecule binding affinity predictor that is trained using SMILES VAE embeddings and protein sequence embeddings learned unsupervised from a large corpus. CogMol framework is applied to three SARS-CoV-2 target proteins: main protease, receptor-binding domain of the spike protein, and non-structural protein 9 replicase. The generated candidates are novel at both molecular and chemical scaffold levels when compared to the training data. CogMol also includes insilico screening for assessing toxicity of parent molecules and their metabolites with a multi-task toxicity classifier, synthetic feasibility with a chemical retrosynthesis predictor, and target structure binding with docking simulations. Docking reveals favorable binding of generated molecules to the target protein structure, where 87-95 % of high affinity molecules showed docking free energy < -6 kcal/mol. When compared to approved drugs, the majority of designed compounds show low parent molecule and metabolite toxicity and high synthetic feasibility. In summary, CogMol handles multi-constraint design of synthesizable, low-toxic, drug-like molecules with high target specificity and selectivity, and does not need target-dependent fine-tuning of the framework or target structure information. | ['Jannis Born', 'Matteo Manica', 'Aleksandra Mojsilovic', 'Payel Das', 'Kar Wai Lim', 'Inkit Padhi', 'Hendrik Strobelt', 'Benjamin Hoover', 'Teodoro Laino', 'Samuel C. Hoffman', 'Vijil Chenthamarakshan'] | 2020-04-02 | null | http://proceedings.neurips.cc/paper/2020/hash/2d16ad1968844a4300e9a490588ff9f8-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/2d16ad1968844a4300e9a490588ff9f8-Paper.pdf | neurips-2020-12 | ['retrosynthesis'] | ['medical'] | [ 2.18051255e-01 -1.74045980e-01 -2.91020811e-01 -1.56314194e-01
-8.12680900e-01 -1.03448522e+00 2.87322104e-01 5.03217340e-01
-2.07259357e-01 1.52396154e+00 2.99874842e-02 -3.82807016e-01
9.09977034e-02 -5.54374516e-01 -8.35235536e-01 -9.61926818e-01
-2.57232964e-01 7.63392866e-01 -1.74491733e-01 -2.07835451e-01
1.51134625e-01 9.20065463e-01 -7.36793160e-01 2.93897927e-01
1.19609070e+00 3.73229295e-01 3.08187276e-01 4.13349569e-01
2.61188835e-01 1.16343886e-01 -2.83339262e-01 -2.61497080e-01
1.28263474e-01 -4.11120534e-01 -3.51421207e-01 -5.71099758e-01
1.33050352e-01 -3.52768563e-02 3.70300978e-01 5.41901171e-01
8.04250360e-01 8.75108466e-02 1.32872641e+00 -6.07013166e-01
-8.74955893e-01 -3.23746145e-01 -2.15877946e-02 -7.94762149e-02
4.79105085e-01 5.50711811e-01 1.12281144e+00 -1.45960939e+00
1.10716939e+00 9.22983229e-01 6.73940361e-01 7.62299418e-01
-1.60194886e+00 -6.20588183e-01 -3.85182530e-01 2.20321920e-02
-1.37552881e+00 -2.62473345e-01 1.77546531e-01 -9.83761728e-01
1.73309886e+00 1.18508734e-01 4.32600558e-01 1.30368125e+00
7.54608572e-01 4.28325497e-02 6.23929977e-01 9.43575427e-02
6.73708677e-01 1.89009249e-01 -3.15865964e-01 7.33307242e-01
4.18574303e-01 2.24679381e-01 -3.40868860e-01 -1.00908995e+00
1.74818411e-01 3.32491457e-01 -2.82327384e-01 -6.45471394e-01
-8.42028201e-01 1.17752028e+00 2.39218324e-01 -1.34021148e-01
-7.76964962e-01 -2.77835488e-01 3.19243670e-01 -2.10721865e-01
1.63573578e-01 9.58366156e-01 -1.17665255e+00 2.56416887e-01
-5.04374027e-01 3.26626211e-01 7.15546489e-01 4.61597443e-01
7.61563778e-01 1.54763803e-01 -1.04725026e-01 3.63127112e-01
6.44826531e-01 4.82177824e-01 7.40724206e-02 -1.20954633e-01
-2.11528763e-01 4.48084742e-01 2.39429459e-01 -5.15952229e-01
-6.95523202e-01 -4.10875350e-01 -3.92589033e-01 2.42211863e-01
-6.76115900e-02 -2.41534144e-01 -1.00037217e+00 2.03885150e+00
6.61523342e-01 1.30419493e-01 4.98477191e-01 5.53447723e-01
6.66738749e-01 1.06058145e+00 5.56936681e-01 -9.34177458e-01
1.39771307e+00 -6.54984832e-01 -4.15847063e-01 4.12958443e-01
5.73170066e-01 -6.51652515e-01 7.38038480e-01 2.47971818e-01
-6.45762444e-01 -1.84182584e-01 -1.16157413e+00 3.60058427e-01
-4.63158131e-01 -3.02193128e-02 6.03312194e-01 7.24068761e-01
-5.29986560e-01 8.48482251e-01 -7.19014823e-01 -2.50799417e-01
4.60229188e-01 9.31563675e-01 -6.02752030e-01 1.08418569e-01
-1.14242733e+00 1.12807584e+00 6.91690028e-01 -3.59217167e-01
-1.47418058e+00 -1.17730093e+00 -6.71459913e-01 -8.63765329e-02
-8.39258777e-04 -1.16202164e+00 4.94983673e-01 -5.03345251e-01
-1.78703761e+00 3.61604065e-01 -6.47308156e-02 -3.37068647e-01
-1.91095009e-01 -5.76852970e-02 -3.23780626e-01 7.62457401e-02
4.02352959e-02 4.75470692e-01 5.96063197e-01 -9.56145585e-01
-2.82226205e-02 -4.59002882e-01 -4.09040868e-01 2.60520905e-01
3.02640975e-01 1.51702404e-01 3.60171735e-01 -5.49096525e-01
-5.44251740e-01 -9.73830342e-01 -4.97634321e-01 -5.70205927e-01
-4.40468162e-01 -2.23742828e-01 4.14536238e-01 -3.45794171e-01
7.01182842e-01 -1.60823858e+00 3.95121098e-01 4.19380158e-01
2.16801107e-01 4.74124372e-01 -3.31582785e-01 1.17204916e+00
-3.23728114e-01 2.13792473e-01 -5.78452908e-02 5.22923172e-01
-4.49262440e-01 -4.12171304e-01 -3.75488877e-01 6.94862187e-01
4.12774831e-01 8.00486088e-01 -9.25381303e-01 1.38517097e-01
3.05635370e-02 8.13693702e-01 -9.86087501e-01 3.52055997e-01
-8.43747675e-01 7.91965365e-01 -8.71174634e-01 9.59442735e-01
7.62679160e-01 -2.36675963e-01 5.82622468e-01 -2.07146674e-01
-9.95638147e-02 2.35271037e-01 -5.75152576e-01 1.26831996e+00
1.09022642e-02 -2.10483238e-01 -6.52376890e-01 -3.38813365e-01
9.11056459e-01 5.00750780e-01 5.07715642e-01 -2.69014001e-01
7.60837495e-02 3.98417056e-01 1.25088140e-01 -3.50434750e-01
-1.78559035e-01 -4.60854828e-01 2.93318808e-01 -3.44477477e-03
2.46177971e-01 2.42440999e-01 8.59218985e-02 2.65296809e-02
1.02020705e+00 5.62027812e-01 4.76745576e-01 -3.60674471e-01
5.73632240e-01 2.76314944e-01 9.91580725e-01 1.45983532e-01
1.18853249e-01 2.92194039e-01 2.26615027e-01 -5.77641010e-01
-1.29809237e+00 -1.00624990e+00 -3.50986183e-01 1.17430532e+00
-2.64299870e-01 -6.69730246e-01 -8.00836802e-01 -4.52623516e-01
-6.43982887e-02 5.58953643e-01 -5.21857083e-01 -1.29936159e-01
-4.51476723e-01 -1.01879179e+00 3.82301509e-01 -1.22292764e-01
-5.09605229e-01 -1.01418614e+00 -9.31826904e-02 6.58807337e-01
5.12197435e-01 -6.65348113e-01 -3.19203436e-01 3.84648085e-01
-6.59449160e-01 -1.33783543e+00 -5.93555510e-01 -6.99676096e-01
5.02116978e-01 -3.96690041e-01 4.80860263e-01 -3.49275947e-01
-3.18295270e-01 -3.16512793e-01 -1.32092834e-01 -6.46729529e-01
-5.02095819e-01 -2.89304733e-01 8.33183587e-01 -1.77347735e-01
3.81136179e-01 -6.85935795e-01 -8.37931931e-01 7.94040710e-02
-3.90231431e-01 -1.66721329e-01 6.57056630e-01 1.11161208e+00
1.05662203e+00 -6.32337749e-01 9.41967607e-01 -7.94457376e-01
5.85571885e-01 -7.57802069e-01 -6.37622595e-01 2.16041774e-01
-7.45969772e-01 2.06686661e-01 1.03780508e+00 -4.10148293e-01
-7.22840905e-01 2.93884605e-01 -3.79313916e-01 -6.94005638e-02
5.63263185e-02 5.76881886e-01 -4.32564139e-01 -2.67369840e-02
9.66247916e-01 3.20316195e-01 -7.52550829e-03 -4.55842018e-01
2.84790248e-01 3.54270875e-01 -3.35195963e-03 -6.67030811e-01
7.36411035e-01 7.91636780e-02 2.22192198e-01 -7.45083511e-01
-1.28041372e-01 -1.60114944e-01 -2.98279613e-01 5.27504861e-01
1.09602964e+00 -1.02958584e+00 -1.21687639e+00 -3.67156453e-02
-1.28516734e+00 -1.48725696e-02 2.29374826e-01 8.80709589e-01
-6.63990796e-01 2.63263643e-01 -4.75075185e-01 -4.22260046e-01
-9.05256510e-01 -1.59038055e+00 8.41518342e-01 4.55663130e-02
-4.22781855e-01 -6.35900915e-01 6.63245082e-01 3.43017817e-01
9.11201015e-02 7.88647294e-01 1.46863544e+00 -1.25203729e+00
-9.03101563e-01 3.41151804e-02 9.96143892e-02 7.74185136e-02
1.66391194e-01 1.29416287e-01 -6.05097950e-01 -6.10976815e-01
-4.40412849e-01 -4.47940677e-01 6.62791967e-01 6.01825893e-01
3.04694474e-01 -5.05547225e-01 -5.74789226e-01 8.28495920e-01
1.65474546e+00 9.66409802e-01 5.78414261e-01 2.48449549e-01
6.44459784e-01 1.95789367e-01 6.28566742e-01 6.30879581e-01
-1.52472556e-01 7.54676282e-01 3.90754759e-01 9.67915878e-02
4.05879259e-01 -2.21567005e-01 4.87313241e-01 1.35892808e-01
-2.94511348e-01 -4.69017863e-01 -7.55018294e-01 1.89728811e-01
-1.30554724e+00 -1.04657543e+00 -1.31221935e-01 2.27703214e+00
1.29180908e+00 -3.39107901e-01 1.06476620e-01 -7.01537728e-01
4.15323198e-01 -4.55284774e-01 -7.83890069e-01 -7.68150330e-01
-2.33279571e-01 6.76930308e-01 3.27862769e-01 6.43613994e-01
-7.75286078e-01 1.10966158e+00 5.86617136e+00 9.85777617e-01
-9.54781055e-01 -7.83192217e-02 5.11023700e-01 9.02141258e-02
-5.22460759e-01 4.26954091e-01 -1.03630972e+00 3.00393820e-01
1.15656209e+00 -1.54614493e-01 1.38391852e-01 8.64562035e-01
5.21930575e-01 3.68363500e-01 -1.20235002e+00 7.04083085e-01
-3.57143223e-01 -2.22435284e+00 2.18763530e-01 2.88419247e-01
8.60629439e-01 1.27361581e-01 3.36505808e-02 -3.38433459e-02
-5.27056828e-02 -1.31463516e+00 1.49366722e-01 4.10987496e-01
1.11286736e+00 -1.03389132e+00 4.93557125e-01 8.61635432e-02
-7.49876797e-01 1.06534153e-01 -3.96480173e-01 4.46390241e-01
1.00818247e-01 2.11931184e-01 -1.40184593e+00 2.05958620e-01
4.72790748e-02 3.16556454e-01 7.80574977e-02 7.57090747e-01
9.59952734e-03 2.65384167e-01 -2.19389245e-01 -2.56447643e-01
2.48428002e-01 -4.84253615e-01 6.09371424e-01 1.13821292e+00
-4.15683836e-02 4.81090456e-01 5.19284964e-01 8.48891854e-01
-1.27989009e-01 6.94451988e-01 -6.11650050e-01 -4.02464449e-01
4.72601205e-01 9.89174902e-01 -2.71405101e-01 -2.97455907e-01
-1.19477347e-01 9.55975354e-01 -9.32563543e-02 5.99070132e-01
-6.67233944e-01 -3.05339813e-01 1.25522435e+00 -7.78265372e-02
4.26713705e-01 4.05905657e-02 1.88193649e-01 -7.19934642e-01
-5.94311595e-01 -1.17731380e+00 1.85657665e-01 -4.58577007e-01
-1.13738751e+00 7.56404400e-01 -2.59561092e-01 -9.14182961e-01
-8.19875821e-02 -6.00089848e-01 -5.21400809e-01 1.19328022e+00
-1.01199663e+00 -1.29415965e+00 5.09888351e-01 2.59311348e-01
5.13566732e-01 -7.00093210e-01 1.43772924e+00 2.53537111e-03
-5.19501150e-01 4.54824060e-01 7.28343487e-01 -7.22558975e-01
7.44017541e-01 -9.02178586e-01 1.78779691e-01 2.13695794e-01
-4.55516279e-01 1.26240146e+00 9.80456948e-01 -1.22615433e+00
-1.49805272e+00 -1.24062216e+00 7.56466627e-01 -5.80299556e-01
3.81752938e-01 -5.49382031e-01 -6.86160266e-01 2.66447604e-01
-6.20808871e-03 -2.89412767e-01 1.56130874e+00 -1.03180528e-01
-4.31061119e-01 4.95270640e-01 -1.44755518e+00 6.77277088e-01
6.17479324e-01 -3.89475077e-01 -3.27194184e-01 8.32971454e-01
9.18810487e-01 -3.23844194e-01 -9.78531003e-01 6.74012125e-01
6.46456599e-01 -5.71820617e-01 1.15188956e+00 -1.37624955e+00
-3.75833102e-02 -7.35011876e-01 -3.36020708e-01 -1.13210762e+00
-6.78467572e-01 -9.17764008e-01 -8.51628631e-02 4.97808516e-01
8.42191994e-01 -2.88672239e-01 8.28587294e-01 1.98453248e-01
-1.90515563e-01 -1.21673644e+00 -7.85871208e-01 -7.49614120e-01
9.51429233e-02 1.02827258e-01 4.26743925e-01 8.38087082e-01
1.61125381e-02 8.83373678e-01 -5.97970963e-01 3.04695845e-01
2.99641460e-01 6.07418194e-02 6.48358524e-01 -1.27873349e+00
-3.95896554e-01 -1.00493498e-01 -3.02118331e-01 -4.07367706e-01
5.37476651e-02 -1.05352390e+00 -6.30856812e-01 -1.45791686e+00
4.29610103e-01 -1.60981312e-01 -2.09392071e-01 2.70145357e-01
1.64130926e-01 -7.85647035e-02 -1.69347495e-01 1.26935467e-01
-1.43801138e-01 7.84634292e-01 8.58708084e-01 -2.31370240e-01
-8.21018577e-01 2.51377141e-03 -6.32367969e-01 3.50885451e-01
6.90335751e-01 -7.00471580e-01 -4.46491987e-01 5.93497455e-01
5.58450162e-01 -1.22061804e-01 -1.47318378e-01 -3.29769820e-01
-4.86048967e-01 -7.41414368e-01 5.66198945e-01 -7.05000997e-01
5.20921886e-01 -5.31986594e-01 7.52362847e-01 7.43149042e-01
3.36620450e-01 -2.92445779e-01 2.91824371e-01 7.32440174e-01
4.15115833e-01 3.09720188e-02 1.06894720e+00 7.25848600e-02
-2.63706505e-01 7.02797115e-01 -7.35601187e-01 -3.56168568e-01
1.09594345e+00 -7.15732455e-01 -1.25796407e-01 2.07446769e-01
-8.34790170e-01 -2.82982498e-01 5.70744812e-01 1.96724102e-01
8.24374318e-01 -1.00015759e+00 -7.81183064e-01 1.58782274e-01
5.30779779e-01 -5.13710260e-01 3.17267656e-01 6.17172062e-01
-9.30475473e-01 7.63018787e-01 -1.93423301e-01 -4.01903123e-01
-1.26024520e+00 9.93414700e-01 4.82900590e-01 -1.35367794e-03
-8.55262429e-02 8.87852252e-01 6.43260896e-01 -3.14210832e-01
-1.80672660e-01 3.48362066e-02 -1.80685967e-01 1.09278299e-01
6.42125487e-01 5.22710495e-02 1.11938581e-01 -6.35046721e-01
-8.79329205e-01 5.48663199e-01 -3.48146677e-01 7.11416066e-01
1.73467577e+00 4.48850960e-01 -7.95418248e-02 -2.10592702e-01
1.29242206e+00 2.50733614e-01 -9.65918422e-01 8.38328972e-02
-1.60019174e-02 -1.71617284e-01 -2.95222789e-01 -1.18560541e+00
-7.25198314e-02 2.90073544e-01 9.15948689e-01 -8.82358849e-01
5.80933750e-01 -8.23657140e-02 7.43827820e-01 6.26428604e-01
3.18467200e-01 -7.76946425e-01 1.70387298e-01 6.60966456e-01
9.90324199e-01 -1.01565516e+00 1.59210190e-01 -7.44483247e-02
-6.18725777e-01 9.59743440e-01 4.27016944e-01 2.56512254e-01
2.35469237e-01 -2.87581444e-01 -2.93913394e-01 -6.48668528e-01
-1.02058375e+00 2.00886041e-01 3.21779907e-01 9.78772938e-01
5.95607996e-01 -2.68900022e-03 -6.70561492e-01 5.71467102e-01
3.92438710e-01 -1.81191921e-01 2.91575462e-01 5.78477383e-01
-5.51625371e-01 -1.59132433e+00 -1.61344379e-01 -5.23917153e-02
-6.77335680e-01 -6.80400610e-01 -5.93034744e-01 3.26378316e-01
3.49997550e-01 6.00182056e-01 -5.38090467e-01 -2.15114236e-01
1.81970417e-01 1.67257980e-01 4.38583851e-01 -8.83273602e-01
-5.68099856e-01 4.42629695e-01 9.68429148e-02 -1.40352130e-01
8.41577277e-02 -1.96838334e-01 -1.60669351e+00 -1.35598004e-01
-6.86556995e-01 4.97793078e-01 7.33683944e-01 8.36112380e-01
1.03577220e+00 3.74628529e-02 8.84677768e-01 -6.90137923e-01
-2.78735518e-01 -7.10711420e-01 -4.95090306e-01 1.72635522e-02
2.98290342e-01 -5.77246070e-01 2.19812691e-01 8.22115541e-02] | [4.874614715576172, 5.611928939819336] |
94c1eb46-75cf-4f5f-9e9b-3e0c8163ed4e | cross-lingual-vision-language-navigation-1 | 1910.11301 | null | https://arxiv.org/abs/1910.11301v3 | https://arxiv.org/pdf/1910.11301v3.pdf | Cross-Lingual Vision-Language Navigation | Commanding a robot to navigate with natural language instructions is a long-term goal for grounded language understanding and robotics. But the dominant language is English, according to previous studies on vision-language navigation (VLN). To go beyond English and serve people speaking different languages, we collect a bilingual Room-to-Room (BL-R2R) dataset, extending the original benchmark with new Chinese instructions. Based on this newly introduced dataset, we study how an agent can be trained on existing English instructions but navigate effectively with another language under a zero-shot learning scenario. Without any training data of the target language, our model shows competitive results even compared to a model with full access to the target language training data. Moreover, we investigate the transferring ability of our model when given a certain amount of target language training data. | ['Lei LI', 'William Yang Wang', 'An Yan', 'Xin Eric Wang', 'Jiangtao Feng'] | 2019-10-24 | null | null | null | null | ['vision-language-navigation'] | ['computer-vision'] | [-1.19095601e-01 1.29769564e-01 -1.71061009e-01 -4.49690759e-01
-3.74321312e-01 -4.80223149e-01 7.41014242e-01 -5.31314731e-01
-1.14086568e+00 6.91480279e-01 3.08608174e-01 -7.23851442e-01
6.91253841e-01 -8.22248638e-01 -7.40341067e-01 -4.60633963e-01
2.41575778e-01 6.38309240e-01 2.08887890e-01 -8.98550451e-01
1.08014077e-01 9.29273367e-02 -1.30725706e+00 3.26139703e-02
1.01567841e+00 3.01334023e-01 1.11279237e+00 4.35291827e-01
-3.14254433e-01 1.14892006e+00 -1.28917182e-02 1.46369621e-01
2.96734095e-01 -3.49811882e-01 -1.08991230e+00 -1.67396024e-01
-1.82180516e-02 -4.78384048e-01 -6.32170081e-01 1.06166196e+00
3.93083572e-01 4.56126720e-01 5.85620165e-01 -1.01452982e+00
-9.11529720e-01 5.21685183e-01 1.07121095e-01 -3.57875109e-01
8.24971020e-01 4.43593591e-01 7.29870915e-01 -9.47124898e-01
1.06108665e+00 1.39206755e+00 2.72434652e-01 1.08080959e+00
-8.50121617e-01 -3.41332257e-01 4.64682937e-01 5.03907442e-01
-1.33167195e+00 -3.73421371e-01 4.35196459e-01 -2.19039455e-01
1.06177640e+00 -4.62391108e-01 4.78277802e-01 1.53162420e+00
1.45297199e-01 8.39604378e-01 8.48199368e-01 -4.42849159e-01
4.09791321e-01 1.71261296e-01 2.42093623e-01 1.06155610e+00
-8.46729800e-03 1.92101181e-01 -4.13335383e-01 5.49271405e-01
3.92898202e-01 -1.47867039e-01 -6.66977584e-01 -8.44059825e-01
-1.61793053e+00 1.00118375e+00 7.97302544e-01 4.00771260e-01
-2.30254129e-01 5.67367673e-02 3.82355154e-01 4.21557933e-01
-2.25992084e-01 4.10321265e-01 -4.27458495e-01 -2.10062951e-01
-7.00978711e-02 -1.72956079e-01 9.40325916e-01 1.51682663e+00
1.02696836e+00 2.07262233e-01 -1.16168611e-01 6.98899686e-01
2.10947618e-01 8.49861026e-01 8.27330649e-01 -9.76929426e-01
5.97055376e-01 4.14200425e-01 3.58378440e-01 -5.62545121e-01
-6.59267008e-01 1.93098605e-01 -8.37929010e-01 1.24445818e-01
5.18976450e-01 -2.53652960e-01 -8.87495279e-01 1.94044900e+00
-6.65743276e-02 -2.05465421e-01 8.46099734e-01 1.10253417e+00
9.72690046e-01 7.48923004e-01 -2.78620452e-01 1.53117374e-01
1.26959300e+00 -1.66528428e+00 -5.69394410e-01 -8.98033619e-01
1.15977287e+00 -1.00402430e-01 1.69833267e+00 2.49383077e-01
-2.59894043e-01 -7.47838557e-01 -9.77416813e-01 -6.06114030e-01
-8.41905594e-01 2.71788031e-01 5.89971244e-01 1.68656498e-01
-1.38743877e+00 -2.53081679e-01 -6.80528641e-01 -1.05410814e+00
-7.13621601e-02 4.71457392e-02 -8.33761990e-01 -8.13126981e-01
-1.34421968e+00 1.28970122e+00 4.45633382e-01 7.32850656e-02
-1.30370021e+00 -5.52378111e-02 -1.54447091e+00 -4.21702713e-01
6.86681271e-01 -6.66266620e-01 1.34730673e+00 -3.63599390e-01
-1.65735209e+00 8.62818718e-01 -4.40126985e-01 -4.77711469e-01
5.43176591e-01 -2.81763762e-01 4.52060625e-02 -4.36822511e-02
5.71251571e-01 1.06932425e+00 2.98780382e-01 -1.35821354e+00
-5.91543317e-01 -2.82335162e-01 6.14950418e-01 4.67674762e-01
8.63086656e-02 -5.83779871e-01 -6.44342184e-01 8.76365881e-03
3.44523117e-02 -1.25270522e+00 -5.41244209e-01 -3.21917415e-01
-3.69298637e-01 -1.54224649e-01 4.10496980e-01 -5.47252655e-01
4.10633832e-01 -2.02076554e+00 4.36373770e-01 -3.63714397e-01
-1.50980443e-01 -1.35149136e-01 -6.13010228e-01 3.00302476e-01
4.29480284e-01 -3.13321054e-01 -2.84036517e-01 -2.89274812e-01
1.73962489e-01 4.93712872e-01 -3.77538383e-01 3.57238084e-01
-3.32848907e-01 1.24159145e+00 -1.32927406e+00 -2.55858719e-01
3.47773671e-01 2.91130185e-01 -7.06561267e-01 2.07096606e-01
-6.44715726e-02 9.37722564e-01 -4.26001906e-01 3.81266385e-01
2.93945849e-01 -8.09490085e-02 1.08065814e-01 7.40488693e-02
-2.64349759e-01 1.39579311e-01 -7.08251178e-01 2.45183706e+00
-1.17782593e+00 6.79766655e-01 2.23990157e-01 -1.02662051e+00
9.06547606e-01 4.33486924e-02 3.07765957e-02 -1.11975121e+00
7.66923353e-02 3.40562373e-01 6.22368045e-02 -7.26279736e-01
5.04408419e-01 8.43856037e-02 -5.71142554e-01 2.55349904e-01
1.39151588e-01 -3.22864294e-01 1.47556156e-01 2.95205474e-01
1.14470172e+00 3.49622071e-01 4.32488114e-01 -3.10566783e-01
6.84442461e-01 4.42709774e-01 1.44685045e-01 1.17248917e+00
-5.09960294e-01 3.24525148e-01 -1.30424663e-01 -5.85327029e-01
-6.33494139e-01 -8.38733613e-01 3.29505831e-01 1.51643050e+00
6.34416699e-01 -2.47843847e-01 -5.93962908e-01 -6.92970693e-01
-5.31944335e-01 1.31667197e+00 -4.15481061e-01 -2.49086305e-01
-5.83332121e-01 -2.71621466e-01 3.96325588e-01 2.69204110e-01
1.03738892e+00 -1.50201702e+00 -1.37488830e+00 1.10716950e-02
-7.63447583e-01 -1.70436549e+00 -5.05465865e-01 4.29165035e-01
-2.06661016e-01 -1.08375430e+00 -7.83432305e-01 -1.38866806e+00
7.26024628e-01 5.99829018e-01 1.22237313e+00 -1.43521100e-01
-5.76739609e-02 1.01699162e+00 -6.81913018e-01 -2.01006055e-01
-5.08659780e-01 2.77249008e-01 2.79943377e-01 -3.89493108e-01
6.02021754e-01 1.37647241e-02 -2.34581515e-01 3.22852939e-01
-3.65137070e-01 4.14913386e-01 4.22656953e-01 1.01273620e+00
5.29069500e-03 -2.75125384e-01 2.78108418e-01 -3.39876622e-01
5.02690494e-01 -2.45928645e-01 -5.65052867e-01 3.57971251e-01
-2.44392484e-01 3.98525566e-01 6.86686397e-01 -3.99755836e-01
-1.06072974e+00 2.10426152e-01 -3.67543399e-02 -5.22794649e-02
-4.74601477e-01 4.08202767e-01 -4.25948471e-01 1.54624898e-02
5.83451152e-01 6.54070616e-01 -1.51061147e-01 -1.10696293e-01
8.93686056e-01 6.48737073e-01 8.41153979e-01 -6.51506484e-01
5.52073956e-01 5.61301112e-01 -2.31443763e-01 -9.35417116e-01
-3.82730663e-01 -5.51578164e-01 -1.03005886e+00 6.78266573e-04
1.22923803e+00 -1.16493165e+00 -8.38047147e-01 3.13705385e-01
-1.51364434e+00 -1.04384255e+00 -4.83195065e-03 8.11944127e-01
-8.53210270e-01 1.52311102e-01 -4.83689457e-01 -6.24915600e-01
1.09823123e-01 -1.56148577e+00 1.02108097e+00 -5.68794049e-02
4.85850796e-02 -9.43456590e-01 6.91495184e-03 -3.34660485e-02
4.58564758e-01 -4.50951129e-01 8.45651984e-01 -6.42628193e-01
-5.49568415e-01 1.45827994e-01 -1.56757966e-01 -2.04075724e-02
3.19577396e-01 -1.07363188e+00 -6.51037276e-01 -3.49912524e-01
2.69694358e-01 -7.13978291e-01 9.32764947e-01 8.21809173e-02
5.35673618e-01 -1.73543364e-01 -4.10005867e-01 5.47158122e-01
1.43728971e+00 4.47932273e-01 4.86148000e-01 7.55284607e-01
9.52088892e-01 8.75311553e-01 8.11487436e-01 2.44527124e-02
1.07070041e+00 5.31462371e-01 5.33630490e-01 1.35804027e-01
2.23565981e-01 -4.52611029e-01 8.19791317e-01 8.48891318e-01
1.30903363e-01 -5.77464066e-02 -1.26869226e+00 6.14345729e-01
-1.87845778e+00 -5.70646524e-01 2.17800617e-01 2.06877518e+00
5.26351213e-01 -1.78396687e-01 -4.79434490e-01 -6.16079807e-01
3.61645371e-01 6.38808832e-02 -6.98948562e-01 -2.10333094e-01
-1.44404158e-01 -2.97499001e-01 5.63342750e-01 8.93110096e-01
-9.71785963e-01 1.67659247e+00 5.67290258e+00 4.37500268e-01
-1.16475594e+00 1.62032917e-01 1.23922043e-01 2.99890131e-01
-9.34173837e-02 -7.51533806e-02 -7.97353208e-01 1.23073965e-01
7.18239129e-01 -1.41042382e-01 9.32719886e-01 1.07301927e+00
2.23043501e-01 -3.65086913e-01 -1.52550125e+00 1.21639764e+00
3.73970091e-01 -8.37452412e-01 1.27889216e-01 -4.44738008e-03
4.02922660e-01 5.30694187e-01 -2.79310822e-01 1.10462701e+00
7.75726676e-01 -1.21070647e+00 6.99528575e-01 3.76724303e-01
7.01074421e-01 -3.04917514e-01 7.47062445e-01 9.68072772e-01
-1.26516247e+00 -2.30709285e-01 -6.39598489e-01 -1.73155710e-01
4.65224028e-01 -3.08045179e-01 -8.95218194e-01 6.80746496e-01
7.18393266e-01 8.20148408e-01 -6.21481955e-01 3.81213516e-01
-4.78794485e-01 -1.18325099e-01 -1.38867021e-01 -2.61998624e-01
6.19484425e-01 -6.64095223e-01 3.30384225e-01 8.17739904e-01
6.30378425e-01 1.10134423e-01 5.77501595e-01 7.66201615e-01
1.20491624e-01 1.68163165e-01 -1.58172548e+00 2.47550011e-01
1.73486516e-01 9.06882942e-01 -5.75749993e-01 -4.07355249e-01
-7.71550179e-01 1.53702307e+00 4.10654634e-01 8.02718461e-01
-7.76608050e-01 -3.70395869e-01 2.70827264e-01 -4.37492818e-01
9.98855606e-02 -6.27235293e-01 8.82189423e-02 -1.31953454e+00
-1.54620871e-01 -8.44535410e-01 -1.26007155e-01 -1.10314465e+00
-8.28938723e-01 8.66940439e-01 -7.50440732e-02 -1.35930574e+00
-5.45221686e-01 -1.06681919e+00 -3.56552303e-01 6.33803368e-01
-1.64784360e+00 -1.28224921e+00 -6.43485367e-01 7.16586947e-01
1.04172814e+00 -6.04228973e-01 1.13934422e+00 -2.24848818e-02
-1.53231785e-01 1.73993021e-01 -9.71737280e-02 3.20563525e-01
7.68546283e-01 -1.15681481e+00 3.89005929e-01 8.26670051e-01
2.64795899e-01 7.38905132e-01 6.01051450e-01 -4.21786815e-01
-1.74169719e+00 -9.69949067e-01 6.89832151e-01 -6.06213987e-01
5.67731380e-01 -6.56464338e-01 -6.37211025e-01 1.08250058e+00
5.41414678e-01 -1.02810018e-01 2.29545131e-01 -1.22469082e-01
-2.53501534e-01 3.34466636e-01 -8.79841685e-01 1.22025251e+00
1.47913992e+00 -6.36602461e-01 -7.92748451e-01 2.95081139e-01
1.20895696e+00 -3.13362062e-01 -4.43141088e-02 3.11325133e-01
2.68670827e-01 -7.75424600e-01 8.80475044e-01 -5.23852706e-01
1.95380047e-01 -3.66754621e-01 -6.68266594e-01 -1.59883785e+00
-3.16586420e-02 -2.19779104e-01 5.19748092e-01 6.73626125e-01
3.05124938e-01 -7.24356651e-01 4.43275392e-01 4.02675658e-01
-2.99572319e-01 -2.26189256e-01 -9.27067280e-01 -8.92423332e-01
2.30959609e-01 -6.16681159e-01 3.31904978e-01 6.49523139e-01
2.09625512e-01 8.92650664e-01 -4.34573919e-01 8.38656425e-02
3.19724262e-01 1.77668408e-02 1.13064373e+00 -9.25080061e-01
1.78933159e-01 -2.39923716e-01 4.05449346e-02 -1.77088261e+00
9.34935689e-01 -1.19374263e+00 8.88982177e-01 -2.11089015e+00
-1.58734005e-02 -1.15945391e-01 1.24839783e-01 5.72365761e-01
1.12681568e-01 -1.93161190e-01 4.04921293e-01 1.19911851e-02
-8.83644044e-01 1.09171474e+00 1.33808684e+00 -6.02453351e-01
-3.30167681e-01 -3.90816838e-01 -5.63994646e-01 7.38182724e-01
5.84314704e-01 8.40972736e-02 -6.66493058e-01 -7.02106237e-01
-1.94795094e-02 -1.59528106e-01 2.86497980e-01 -9.36145723e-01
6.56386077e-01 -1.59148693e-01 -2.24803448e-01 -4.59796846e-01
3.71860653e-01 -9.03141439e-01 -5.52408397e-01 7.36596823e-01
-2.50448823e-01 1.65848762e-01 8.42233896e-02 6.17792189e-01
-1.01246111e-01 -3.15052867e-01 4.66289133e-01 -6.26218379e-01
-1.77277052e+00 2.68111024e-02 -8.15312207e-01 1.19119406e-01
9.17248726e-01 -1.98986428e-03 -4.34333950e-01 -6.91840947e-01
-6.07595444e-01 7.23568141e-01 7.78681576e-01 7.89161980e-01
6.89548910e-01 -1.40192437e+00 -4.02180851e-01 2.13241875e-01
6.73984826e-01 2.56233573e-01 5.80453351e-02 7.13230729e-01
-6.34042978e-01 9.95945573e-01 -2.70343661e-01 -6.33332729e-01
-5.43252289e-01 1.07009828e+00 3.66661221e-01 -2.00972706e-02
-8.60968947e-01 5.53304017e-01 7.42669523e-01 -1.40760744e+00
3.80283862e-01 -4.65643704e-01 -4.86515522e-01 -3.33212942e-01
3.05270135e-01 2.11466886e-02 -3.93809110e-01 -9.54098463e-01
-5.47396362e-01 6.61574423e-01 3.40391636e-01 -5.66009521e-01
7.74853945e-01 -5.86286426e-01 -1.29968286e-01 9.15058255e-01
1.05402732e+00 -1.33530349e-01 -9.48169351e-01 -2.25502327e-01
1.66137770e-01 1.24165870e-01 -4.01207387e-01 -4.08224225e-01
-6.56602740e-01 1.10468078e+00 5.99167466e-01 -4.22964871e-01
5.44579446e-01 9.53610390e-02 4.21215683e-01 1.44835937e+00
1.33127749e+00 -1.07691777e+00 2.73132741e-01 1.31572258e+00
8.99981499e-01 -1.95717108e+00 -6.88920975e-01 -4.79231179e-02
-9.58637238e-01 8.20897996e-01 9.85104680e-01 1.82786316e-01
5.53633988e-01 -1.79635882e-01 4.39634681e-01 8.63948613e-02
-5.18245935e-01 -7.19634473e-01 -1.44170612e-01 1.15978956e+00
8.58713910e-02 -4.21773903e-02 5.60087487e-02 6.91768885e-01
-4.21729177e-01 3.16941030e-02 6.83216572e-01 8.81432474e-01
-4.96565312e-01 -7.86210656e-01 -1.02392688e-01 -2.37725407e-01
3.95596027e-01 -2.48179227e-01 -1.51289836e-01 1.02613020e+00
-7.18826726e-02 1.32348156e+00 -6.59617186e-02 -2.83039480e-01
3.43484312e-01 1.78637058e-01 2.11986288e-01 -9.48288977e-01
1.91386893e-01 -3.92925322e-01 -6.39381558e-02 -7.92649209e-01
-3.21922004e-01 -2.50883341e-01 -1.74338770e+00 1.75428450e-01
2.55827427e-01 1.33985907e-01 5.52092791e-01 1.11639154e+00
1.04836524e-01 5.97102821e-01 3.34698737e-01 -1.11777723e+00
-2.93166786e-01 -1.06100130e+00 -1.79309502e-01 2.28590369e-01
5.37295222e-01 -6.50839508e-01 -3.14260781e-01 -3.22937481e-02] | [4.447103500366211, 0.5526334047317505] |
49cad5d7-803b-4868-a4dd-b9e06157d330 | a-comparative-study-on-crime-in-denver-city | 2001.02802 | null | https://arxiv.org/abs/2001.02802v1 | https://arxiv.org/pdf/2001.02802v1.pdf | A Comparative Study on Crime in Denver City Based on Machine Learning and Data Mining | To ensure the security of the general mass, crime prevention is one of the most higher priorities for any government. An accurate crime prediction model can help the government, law enforcement to prevent violence, detect the criminals in advance, allocate the government resources, and recognize problems causing crimes. To construct any future-oriented tools, examine and understand the crime patterns in the earliest possible time is essential. In this paper, I analyzed a real-world crime and accident dataset of Denver county, USA, from January 2014 to May 2019, which containing 478,578 incidents. This project aims to predict and highlights the trends of occurrence that will, in return, support the law enforcement agencies and government to discover the preventive measures from the prediction rates. At first, I apply several statistical analysis supported by several data visualization approaches. Then, I implement various classification algorithms such as Random Forest, Decision Tree, AdaBoost Classifier, Extra Tree Classifier, Linear Discriminant Analysis, K-Neighbors Classifiers, and 4 Ensemble Models to classify 15 different classes of crimes. The outcomes are captured using two popular test methods: train-test split, and k-fold cross-validation. Moreover, to evaluate the performance flawlessly, I also utilize precision, recall, F1-score, Mean Squared Error (MSE), ROC curve, and paired-T-test. Except for the AdaBoost classifier, most of the algorithms exhibit satisfactory accuracy. Random Forest, Decision Tree, Ensemble Model 1, 3, and 4 even produce me more than 90% accuracy. Among all the approaches, Ensemble Model 4 presented superior results for every evaluation basis. This study could be useful to raise the awareness of peoples regarding the occurrence locations and to assist security agencies to predict future outbreaks of violence in a specific area within a particular time. | ['Md. Aminur Rab Ratul'] | 2020-01-09 | null | null | null | null | ['crime-prediction'] | ['miscellaneous'] | [-2.20781595e-01 -4.05740321e-01 -2.47195482e-01 -2.81674057e-01
-2.75980949e-01 -3.21229994e-01 2.91979402e-01 5.41031837e-01
-4.53142315e-01 7.88866341e-01 2.33092591e-01 -7.66128838e-01
-5.42819858e-01 -1.13626456e+00 6.60791025e-02 -6.03766263e-01
-1.58579350e-01 1.71690017e-01 1.10502794e-01 -1.53922662e-01
8.42272699e-01 8.30422759e-01 -1.42100680e+00 3.46487314e-01
1.17663813e+00 5.96584857e-01 -8.56947675e-02 2.75884748e-01
-9.38555449e-02 7.33279526e-01 -4.68088895e-01 -2.76404470e-01
-5.36819771e-02 -2.00158626e-01 -6.64694607e-01 -5.53775072e-01
-3.83903146e-01 -4.70280021e-01 4.38877046e-02 8.67518365e-01
2.66610831e-01 1.29166052e-01 8.76187086e-01 -1.18776953e+00
-4.67342526e-01 2.20018357e-01 -7.19338953e-01 6.33303642e-01
4.85363692e-01 8.80036578e-02 1.55268058e-01 -7.56815434e-01
3.11774671e-01 1.10616934e+00 7.37644970e-01 2.71786779e-01
-8.66531670e-01 -1.00344527e+00 -1.25252083e-01 6.60931051e-01
-1.30293024e+00 -9.92027596e-02 6.39049232e-01 -8.70535851e-01
9.08369601e-01 3.17071438e-01 6.76465034e-01 6.86522782e-01
3.52920562e-01 2.49252528e-01 1.32598591e+00 -4.09187913e-01
1.80762392e-02 2.90463656e-01 7.13101268e-01 6.65580869e-01
6.18493080e-01 1.59979433e-01 -1.37043193e-01 -3.51075351e-01
2.02352107e-01 6.39387429e-01 -6.42149150e-02 4.41696972e-01
-4.31587696e-01 9.58229363e-01 1.30823061e-01 5.07991195e-01
-3.69346142e-01 -7.01505959e-01 5.70947349e-01 -8.25995058e-02
6.06848955e-01 5.62959872e-02 -2.91682929e-01 -2.93694109e-01
-8.13693106e-01 1.99460745e-01 3.38163435e-01 -2.32178941e-02
6.06015563e-01 -1.22516215e-01 -1.09568819e-01 6.97963059e-01
1.42058119e-01 6.22283399e-01 2.44314969e-01 -3.54135185e-01
7.86036670e-01 1.04029012e+00 -1.04090080e-01 -1.66442299e+00
-6.29323483e-01 -1.78082716e-02 -1.07957220e+00 2.40514994e-01
3.02629441e-01 -2.51151681e-01 -5.95380068e-01 1.14025617e+00
2.78930455e-01 3.36623229e-02 -5.29941618e-02 5.79625487e-01
6.06550753e-01 7.84331858e-01 5.08685231e-01 -4.93021578e-01
1.15094590e+00 -1.42734692e-01 -7.82124817e-01 1.01581082e-01
7.19445407e-01 -4.56655800e-01 5.82535148e-01 4.73977417e-01
-5.18570185e-01 -3.85514021e-01 -4.87477839e-01 5.69809318e-01
-6.14538968e-01 2.92495579e-01 4.71877038e-01 8.18729520e-01
-3.18678677e-01 6.55580759e-01 -7.98317432e-01 -5.09675741e-01
4.93386060e-01 -3.18879820e-02 -4.92105484e-01 -1.80626556e-01
-1.15501261e+00 1.23607123e+00 4.09103513e-01 1.53620645e-01
-2.29997024e-01 -4.67262447e-01 -7.28214145e-01 -5.71321510e-02
-8.22318271e-02 -6.18395805e-02 2.64854908e-01 -2.09792465e-01
-4.16664928e-01 7.05243647e-01 -2.64284909e-01 -1.14103064e-01
1.54481128e-01 6.16657138e-02 -7.32243776e-01 -9.88176316e-02
5.51104307e-01 -2.40825579e-01 -2.46281107e-03 -9.10967171e-01
-1.02800357e+00 -1.10685551e+00 -2.77525723e-01 -1.76403791e-01
-3.58887136e-01 6.53056979e-01 4.66353983e-01 -4.08340901e-01
1.24543175e-01 -4.41106498e-01 -2.69694626e-01 -6.79794788e-01
-2.68795073e-01 -3.27829659e-01 1.07121205e+00 -1.23658061e+00
1.74218678e+00 -2.14755774e+00 -5.60173750e-01 4.63877499e-01
-9.95371789e-02 5.79102576e-01 5.40998042e-01 5.03127098e-01
-1.93110004e-01 1.86259627e-01 -2.60858119e-01 3.25997293e-01
-5.33991396e-01 8.94656479e-02 -1.67592719e-01 3.58304113e-01
1.72621757e-01 2.24138483e-01 -6.65323496e-01 -5.12933850e-01
4.59131151e-01 2.07416162e-01 -2.47196794e-01 1.10226929e-01
7.77137399e-01 3.81829798e-01 -5.41030645e-01 7.42331147e-01
7.39075780e-01 3.49883169e-01 -1.16681471e-01 1.32081941e-01
-4.94664848e-01 4.08979552e-03 -1.17855179e+00 6.80291474e-01
-2.66248226e-01 5.44744134e-01 -1.71494126e-01 -1.40150809e+00
1.40563142e+00 3.17175537e-01 3.21173310e-01 -8.08901429e-01
-3.91835384e-02 3.05809706e-01 -2.29045212e-01 -1.04590166e+00
1.41660407e-01 1.44669220e-01 9.59526896e-02 4.76241142e-01
-6.32506788e-01 6.46635234e-01 2.70262629e-01 -1.28274500e-01
9.67185318e-01 -3.72529119e-01 5.41777790e-01 -1.61321715e-01
5.28529882e-01 3.02939981e-01 6.19098127e-01 4.07129407e-01
-3.19770545e-01 1.21787734e-01 5.25337338e-01 -1.05392933e+00
-5.94887912e-01 -7.45291293e-01 -4.54042315e-01 7.00120330e-01
-4.67338562e-02 7.34791979e-02 -6.03240728e-01 -5.39246380e-01
-2.18336090e-01 1.00911760e+00 -4.60287929e-01 -2.12195918e-01
-4.71844137e-01 -1.26226604e+00 4.70554680e-01 2.46173993e-01
1.04337800e+00 -1.04290891e+00 -9.00310040e-01 1.40914023e-01
-2.42429212e-01 -4.52230126e-01 3.91679496e-01 -1.68903068e-01
-8.07784915e-01 -1.55889571e+00 -2.29188100e-01 -3.93671900e-01
4.97301906e-01 2.94890940e-01 5.45303702e-01 3.25245470e-01
-4.91163164e-01 -2.85265334e-02 -5.59468210e-01 -5.31228364e-01
-1.63465977e-01 -2.65770555e-01 1.41397819e-01 -1.10327192e-01
8.21329176e-01 -6.96884871e-01 -4.47950900e-01 6.17278591e-02
-4.96751249e-01 -1.60742044e-01 2.03350410e-01 4.76897806e-01
1.06244050e-01 5.45733333e-01 4.68974084e-01 -5.88082612e-01
8.89950752e-01 -8.79048407e-01 -2.37594143e-01 4.85224068e-01
-7.86031961e-01 -4.93092865e-01 4.57929790e-01 2.91058477e-02
-9.57445264e-01 -2.90436596e-01 6.41573519e-02 2.28473276e-01
-5.41327000e-01 4.84588653e-01 4.63296771e-02 4.88729566e-01
8.05539250e-01 2.72700578e-01 -1.90200970e-01 -5.10949731e-01
-5.01089811e-01 1.05279052e+00 3.18810046e-01 -4.32944447e-01
4.72324699e-01 2.66409427e-01 5.71801737e-02 -7.32433975e-01
-4.95658189e-01 -5.82622170e-01 -6.75613344e-01 -5.99403381e-01
1.17859137e+00 -4.72749025e-01 -8.45311821e-01 5.93102276e-01
-1.63384807e+00 2.92374283e-01 5.92507482e-01 6.55669272e-01
2.27231756e-01 2.15481520e-01 -2.27991521e-01 -1.49410379e+00
-4.87800956e-01 -8.96997750e-01 3.40202868e-01 4.95285422e-01
-1.98936403e-01 -8.31277072e-01 2.45970175e-01 4.34964865e-01
2.48023733e-01 6.24817550e-01 1.20201480e+00 -8.01881075e-01
1.46669820e-01 -2.31643677e-01 -3.96852851e-01 2.10176110e-01
3.40119213e-01 5.24706185e-01 -7.32273877e-01 8.65466669e-02
-2.79972821e-01 1.99954018e-01 7.26465166e-01 4.91151959e-01
1.26498675e+00 -5.81261218e-01 -5.99320173e-01 3.39497149e-01
1.26255000e+00 1.10795927e+00 7.83458292e-01 6.48993731e-01
3.32609743e-01 9.84191418e-01 6.91744804e-01 5.24221182e-01
5.52979589e-01 3.80150616e-01 3.08220267e-01 -1.27014801e-01
5.10187447e-01 -2.13035107e-01 1.45823404e-01 2.95737743e-01
-8.10732484e-01 2.89721072e-01 -1.49953067e+00 3.86509418e-01
-1.87334406e+00 -1.65708041e+00 -6.26589298e-01 2.26148391e+00
2.73069292e-01 -3.58016714e-02 3.36361706e-01 8.17540765e-01
9.32249129e-01 -3.03844124e-01 -1.28266914e-02 -6.81713581e-01
1.59193948e-01 2.05522195e-01 3.71312350e-01 2.49964878e-01
-1.08989406e+00 4.67091858e-01 5.79118013e+00 8.64731729e-01
-1.13453341e+00 -3.78497988e-02 1.14544618e+00 1.58533409e-01
8.34466219e-02 1.15154527e-01 -6.37570798e-01 8.25020075e-01
7.14069545e-01 -8.77871662e-02 3.38746607e-01 9.15792704e-01
7.85190046e-01 -4.75587934e-01 -2.59952635e-01 8.94414008e-01
-2.26662412e-01 -1.35700929e+00 -3.00021023e-01 1.33352771e-01
4.17781979e-01 -5.41043162e-01 -2.79609293e-01 3.51230264e-01
1.61113247e-01 -1.11730623e+00 2.38691062e-01 9.43125784e-01
5.63568115e-01 -1.06827307e+00 9.05211210e-01 8.50359380e-01
-9.59832191e-01 -5.46100497e-01 -2.14360520e-01 -6.70507789e-01
4.55301516e-02 6.39412999e-01 -5.95796764e-01 4.98864442e-01
1.19463813e+00 4.41927135e-01 -5.22103071e-01 8.67522836e-01
-1.31776601e-01 7.37244129e-01 -2.05939896e-02 -1.95329785e-01
1.13014355e-01 -3.71236384e-01 1.16103351e-01 1.19107795e+00
4.27299172e-01 8.17734480e-01 -8.14210717e-03 5.43305814e-01
8.99349868e-01 3.48209381e-01 -9.48098481e-01 5.85478067e-01
6.68403566e-01 8.94812346e-01 -6.14652693e-01 -1.19087972e-01
-2.56008357e-01 3.09497893e-01 3.12023401e-01 2.17387706e-01
-6.05455399e-01 -6.31327987e-01 4.96885031e-01 4.35208499e-01
-4.69532996e-01 -5.37972450e-02 -6.49722159e-01 -6.75055146e-01
5.37453359e-03 -6.41290545e-01 5.76432467e-01 -5.23765326e-01
-1.00579512e+00 3.91160518e-01 5.32105803e-01 -9.67939734e-01
-1.59377396e-01 -5.61485827e-01 -1.19946420e+00 1.01958454e+00
-1.01214111e+00 -7.72166491e-01 -1.72767952e-01 6.34298980e-01
2.03031421e-01 -4.77385879e-01 9.14607942e-01 3.57641071e-01
-1.12199211e+00 2.23601252e-01 1.62736505e-01 5.46549618e-01
1.68935567e-01 -5.96807778e-01 -2.89942831e-01 8.71682465e-01
-3.94752383e-01 6.18494987e-01 5.17301023e-01 -8.82267594e-01
-4.05836463e-01 -8.00451994e-01 1.22004569e+00 -1.52662665e-01
3.70351791e-01 1.26269534e-01 -8.65841806e-01 2.34142110e-01
-3.48210782e-01 -2.51017034e-01 7.92437732e-01 3.35192829e-01
-7.50810206e-02 -2.75952905e-01 -1.57325244e+00 3.29731464e-01
5.48513353e-01 -1.49044544e-01 -5.55302441e-01 2.28123397e-01
-9.68285091e-03 4.92827632e-02 -7.67753184e-01 3.67908478e-01
5.87146640e-01 -1.51182783e+00 7.17939794e-01 -9.01161015e-01
6.31611764e-01 1.77698676e-02 7.32306344e-03 -9.92734373e-01
-3.90525460e-01 2.73023307e-01 4.68439311e-01 1.41201341e+00
4.91822779e-01 -7.70155489e-01 5.76416373e-01 1.07797396e+00
1.78474784e-01 -9.52142775e-01 -1.15236115e+00 -4.48336661e-01
1.10653266e-01 -7.86261261e-01 8.43825519e-01 1.20995927e+00
3.34826767e-01 -1.23961024e-01 -3.45991611e-01 2.40001172e-01
5.78194559e-01 -1.45458147e-01 6.76267862e-01 -1.23870301e+00
2.07196802e-01 -3.76788259e-01 -6.67825222e-01 5.69996573e-02
-2.12460123e-02 -5.17780304e-01 -8.29277396e-01 -1.69988966e+00
4.98137057e-01 -7.00107753e-01 -6.33769706e-02 8.71166050e-01
-3.25407267e-01 -1.94871858e-01 -4.79949377e-02 2.04471499e-01
-9.63773113e-03 1.72183603e-01 8.03314149e-01 -1.22801609e-01
-2.92689711e-01 4.20266360e-01 -5.80695570e-01 9.01716411e-01
1.02430308e+00 -6.08767569e-01 -1.69660479e-01 -3.41783553e-01
-6.16143420e-02 3.70206237e-01 4.69903141e-01 -1.05574822e+00
2.40506694e-01 -9.49492991e-01 6.08290434e-01 -6.41991496e-01
-9.20423195e-02 -9.46570992e-01 3.43216687e-01 8.74084532e-01
7.43617192e-02 3.48963708e-01 1.43523201e-01 6.28665984e-02
-2.18101859e-01 -4.25720632e-01 5.00578284e-01 -1.41603779e-02
-5.64688444e-01 1.98312417e-01 -5.92944562e-01 -3.70429218e-01
1.46929562e+00 -5.99662364e-01 -5.94063997e-01 -8.95377696e-02
-5.77132523e-01 1.98217526e-01 -1.30973801e-01 1.25489399e-01
6.73388898e-01 -1.26503575e+00 -8.79171968e-01 6.92688748e-02
-4.48166095e-02 -4.20941651e-01 5.70482016e-01 8.43578279e-01
-7.03205585e-01 3.37688118e-01 -4.92550254e-01 -1.31857842e-01
-1.50039566e+00 3.72016728e-01 1.15730397e-01 -5.29200256e-01
-2.63387442e-01 3.33062232e-01 -4.37061310e-01 -2.67258555e-01
-1.30228728e-01 1.90589279e-01 -1.03294957e+00 2.51463950e-02
8.29869270e-01 1.08157742e+00 -8.54656547e-02 -8.04730117e-01
-6.35261416e-01 6.23770118e-01 1.10400036e-01 1.76299781e-01
1.52037036e+00 3.43105257e-01 -3.85128945e-01 1.10162094e-01
8.63450468e-01 -1.32007107e-01 -4.34369147e-01 3.17376107e-01
1.60694689e-01 -7.79982448e-01 -1.61133692e-01 -8.24565828e-01
-9.72428083e-01 8.80548179e-01 6.62535131e-01 4.50685263e-01
1.43668532e+00 -3.66093308e-01 2.95471281e-01 1.80554569e-01
5.42010427e-01 -1.06168699e+00 -5.90329647e-01 3.20037037e-01
6.16202593e-01 -1.32975602e+00 4.05545020e-03 -1.73351541e-01
-5.69184780e-01 1.38181353e+00 5.70233524e-01 9.62041132e-03
8.32210660e-01 1.07312627e-01 -3.25498125e-03 -2.06523716e-01
-4.97622550e-01 4.80508879e-02 -1.88099872e-02 9.53303397e-01
4.49586511e-01 2.90157199e-01 -8.31897497e-01 8.83613169e-01
-1.61232695e-01 -1.43402964e-01 1.22791886e-01 6.53410196e-01
-8.50251019e-01 -8.40497971e-01 -9.48532224e-01 8.11515570e-01
-4.64345813e-01 6.37281314e-02 -2.42845595e-01 9.05279934e-01
6.58642888e-01 1.41264999e+00 1.37844190e-01 -6.97869956e-01
5.36745310e-01 -7.32143447e-02 -1.80966318e-01 -1.48589209e-01
-4.69008952e-01 -6.69770956e-01 -3.60714179e-03 -1.88961521e-01
-3.42828542e-01 -7.81958163e-01 -1.14337850e+00 -8.46491218e-01
-2.46701285e-01 4.34984058e-01 7.63255835e-01 1.13150430e+00
3.15364748e-02 -6.66505471e-03 9.35917854e-01 -4.26264107e-01
-1.10617355e-01 -1.11542892e+00 -6.03610516e-01 2.29433894e-01
-1.33476570e-01 -7.81818986e-01 -3.46999049e-01 -4.12783891e-01] | [6.792037487030029, 1.9769750833511353] |
38587b3f-b47b-4d62-844e-ace927af5d39 | deconstruct-to-reconstruct-a-configurable | 2011.00483 | null | https://arxiv.org/abs/2011.00483v1 | https://arxiv.org/pdf/2011.00483v1.pdf | Deconstruct to Reconstruct a Configurable Evaluation Metric for Open-Domain Dialogue Systems | Many automatic evaluation metrics have been proposed to score the overall quality of a response in open-domain dialogue. Generally, the overall quality is comprised of various aspects, such as relevancy, specificity, and empathy, and the importance of each aspect differs according to the task. For instance, specificity is mandatory in a food-ordering dialogue task, whereas fluency is preferred in a language-teaching dialogue system. However, existing metrics are not designed to cope with such flexibility. For example, BLEU score fundamentally relies only on word overlapping, whereas BERTScore relies on semantic similarity between reference and candidate response. Thus, they are not guaranteed to capture the required aspects, i.e., specificity. To design a metric that is flexible to a task, we first propose making these qualities manageable by grouping them into three groups: understandability, sensibleness, and likability, where likability is a combination of qualities that are essential for a task. We also propose a simple method to composite metrics of each aspect to obtain a single metric called USL-H, which stands for Understandability, Sensibleness, and Likability in Hierarchy. We demonstrated that USL-H score achieves good correlations with human judgment and maintains its configurability towards different aspects and metrics. | ['Akiko Aizawa', 'Yang Zhao', 'Vitou Phy'] | 2020-11-01 | null | https://aclanthology.org/2020.coling-main.368 | https://aclanthology.org/2020.coling-main.368.pdf | coling-2020-8 | ['dialogue-evaluation'] | ['natural-language-processing'] | [-4.64066267e-01 6.99776709e-02 -1.21528029e-01 -7.01787055e-01
-2.40405932e-01 -6.98165357e-01 5.32998681e-01 3.78447950e-01
-4.66265500e-01 5.82539976e-01 4.20678318e-01 1.09544672e-01
-4.77174640e-01 -6.36084318e-01 4.11636442e-01 -2.72971153e-01
5.86975574e-01 3.50150377e-01 2.28331864e-01 -6.75231993e-01
4.33527052e-01 1.88490361e-01 -1.23411679e+00 2.67777473e-01
1.42100525e+00 8.32582474e-01 2.24437788e-01 2.83755600e-01
-4.44208324e-01 7.72119641e-01 -7.61169314e-01 -4.88132179e-01
-2.15215564e-01 -9.30384219e-01 -1.17583776e+00 4.10475433e-02
-3.61449212e-01 -3.22728395e-01 2.14720488e-01 9.89880383e-01
3.51822585e-01 5.37547804e-02 7.94670105e-01 -1.22947633e+00
-6.29988194e-01 7.28484511e-01 -4.23385650e-02 -1.32440269e-01
8.33892226e-01 2.88626254e-01 1.29507267e+00 -4.21259165e-01
2.06932425e-01 1.18974209e+00 3.78891528e-01 5.46618938e-01
-9.51433063e-01 -3.24628025e-01 1.74369946e-01 5.55099919e-02
-7.66803741e-01 -2.24894211e-01 7.54494131e-01 -5.89432895e-01
4.89659965e-01 4.85667735e-01 6.94401622e-01 7.56493211e-01
-1.12706106e-02 4.54199731e-01 1.42658424e+00 -1.18923351e-01
3.79605621e-01 6.10646725e-01 5.28902471e-01 3.31378758e-01
1.02029197e-01 -3.13235343e-01 -5.12675166e-01 -1.29621372e-01
6.12308800e-01 -1.05253890e-01 -5.67785501e-01 -2.41028249e-01
-1.47561741e+00 9.40948844e-01 1.60322726e-01 5.64342260e-01
-1.96532309e-01 -5.83172202e-01 5.52886724e-01 6.10074401e-01
9.28168595e-02 8.64408135e-01 -4.88231301e-01 -5.36949754e-01
-5.43298483e-01 1.12146676e-01 1.18123722e+00 6.43069565e-01
6.14351571e-01 -3.50525409e-01 -5.67923009e-01 1.11352801e+00
2.80107707e-01 2.65239060e-01 7.64551222e-01 -9.96320367e-01
8.03185105e-02 1.04864562e+00 3.12671691e-01 -1.06512761e+00
-7.66591787e-01 -1.95865780e-01 -7.77456701e-01 2.10595373e-02
5.02023458e-01 -6.38585314e-02 -1.05255157e-01 1.99330378e+00
3.89731944e-01 -9.39420223e-01 1.11920454e-01 1.19537830e+00
1.40797949e+00 3.01931262e-01 5.28424531e-02 -4.12518293e-01
1.38350260e+00 -9.82950091e-01 -1.03969765e+00 -1.43099399e-02
5.88170230e-01 -7.19344795e-01 1.65485299e+00 3.96975100e-01
-9.18070078e-01 -4.98885453e-01 -1.02140570e+00 1.10116974e-01
-2.51187801e-01 -1.34023577e-01 5.13583302e-01 6.47147238e-01
-7.45634913e-01 5.52243829e-01 -6.43184707e-02 -4.37082082e-01
-6.41488791e-01 9.91128236e-02 -3.18403214e-01 5.94726205e-01
-1.54198885e+00 1.13797963e+00 3.36091042e-01 -1.81475505e-01
-3.25706482e-01 1.51673118e-02 -4.45975602e-01 1.20619252e-01
2.42311046e-01 -5.33490360e-01 1.37377429e+00 -9.77325320e-01
-1.98113441e+00 7.50605106e-01 3.37345093e-01 9.38608646e-02
4.72780794e-01 1.39119625e-01 -5.38386822e-01 2.01502115e-01
5.45443501e-03 4.24283296e-01 4.60093409e-01 -9.63121533e-01
-5.35958588e-01 -5.41822314e-02 7.90383577e-01 3.75452965e-01
-6.56746030e-01 4.03080255e-01 -5.65277971e-03 -5.48583567e-01
1.37655094e-01 -5.41002512e-01 1.30126104e-01 -6.87850639e-02
-1.09923109e-01 -5.64695716e-01 3.55529815e-01 -4.78673339e-01
1.62536049e+00 -2.01577115e+00 4.23376402e-03 -2.01925915e-03
5.05039990e-01 3.60151708e-01 -1.44893229e-01 6.53610587e-01
1.89664990e-01 3.04218195e-02 -5.98177090e-02 5.50342202e-01
3.98006916e-01 1.71909109e-02 2.92362213e-01 -4.00748327e-02
4.99718711e-02 6.99188948e-01 -1.22145796e+00 -5.21813631e-01
-7.76912421e-02 7.70870894e-02 -5.65801620e-01 4.09791589e-01
-7.73465633e-02 4.50715184e-01 -7.87749767e-01 2.77456492e-01
3.43723595e-01 -3.49351287e-01 2.31081739e-01 -1.32399574e-01
-6.32518381e-02 4.63646561e-01 -1.08577144e+00 1.22787476e+00
-6.03034616e-01 6.69357330e-02 5.90483099e-03 -6.29839063e-01
1.38578808e+00 3.29064816e-01 2.24519283e-01 -6.80691123e-01
2.53510088e-01 4.18546498e-01 3.36050093e-01 -6.55375421e-01
6.44168913e-01 -1.54530317e-01 -3.35125893e-01 8.24565113e-01
-9.55975056e-02 -2.94556558e-01 2.66803026e-01 2.18771100e-01
8.42208207e-01 -6.81024790e-02 7.85925984e-01 -3.66091758e-01
8.38935673e-01 -1.87172532e-01 4.98383909e-01 4.19007272e-01
-5.61880350e-01 3.78192306e-01 8.14197302e-01 -2.29048997e-01
-6.77126467e-01 -7.87999332e-01 4.40032557e-02 1.09157217e+00
3.33093166e-01 -5.74915528e-01 -8.66173744e-01 -7.35659719e-01
-4.34085280e-01 6.93553686e-01 -2.63572842e-01 -2.88096189e-01
-1.01422347e-01 -3.34945679e-01 4.36681718e-01 1.69307813e-01
8.29237282e-01 -1.11879587e+00 -8.61328602e-01 1.27639443e-01
-7.39497781e-01 -7.34913826e-01 -6.15658164e-01 -2.69177761e-02
-4.66562510e-01 -1.18663120e+00 -5.00691473e-01 -5.38250566e-01
3.29712510e-01 3.84778798e-01 1.27797437e+00 2.89037257e-01
4.66856837e-01 3.18863571e-01 -9.45799649e-01 7.36964960e-03
-5.78006744e-01 -3.30278762e-02 8.87976587e-02 -6.23845905e-02
4.19074178e-01 -2.66239047e-01 -6.17738843e-01 7.35684812e-01
-8.12325418e-01 7.54734129e-02 5.23047924e-01 8.76270890e-01
1.06344521e-01 -3.38464975e-01 7.79482484e-01 -5.69517732e-01
1.45987904e+00 -3.47165823e-01 1.72330245e-01 4.76038605e-01
-7.87948072e-01 -1.05487078e-01 7.68797398e-01 -4.46630359e-01
-9.42087173e-01 -4.40173239e-01 -2.52708137e-01 2.85981536e-01
-1.38505369e-01 5.92193902e-01 -2.48468518e-01 -1.50817826e-01
4.86148894e-01 3.15345973e-02 2.50898242e-01 -2.45885909e-01
2.23617554e-01 1.09459257e+00 1.68733150e-01 -7.29593873e-01
4.08180416e-01 -1.76438361e-01 -3.78780931e-01 -5.69190145e-01
-9.42417145e-01 -3.49593818e-01 -4.88410175e-01 -4.46417272e-01
7.35756040e-01 -4.65625286e-01 -1.09383881e+00 3.74817789e-01
-1.17741132e+00 -1.08009689e-02 -1.97690576e-01 6.39798403e-01
-6.43804014e-01 7.51711965e-01 -5.54513156e-01 -7.49282241e-01
-4.97403443e-01 -1.00341928e+00 4.74375606e-01 3.79148871e-01
-8.50702226e-01 -8.81389499e-01 -2.53001712e-02 4.45556611e-01
6.91591859e-01 4.99025807e-02 9.36964393e-01 -1.04643965e+00
2.44537264e-01 -1.29925415e-01 -2.41464779e-01 5.14039576e-01
2.93513894e-01 -7.76538253e-02 -5.18873692e-01 -7.36172404e-03
5.46253502e-01 -7.20706701e-01 3.91123295e-01 -8.81813988e-02
7.17224956e-01 -6.59169197e-01 2.61538565e-01 8.34261924e-02
8.26832175e-01 3.81120503e-01 4.82028604e-01 5.03013909e-01
1.82912305e-01 9.93563414e-01 1.05374002e+00 5.64051509e-01
5.84551394e-01 6.96055770e-01 8.93703550e-02 5.21305352e-02
2.59154648e-01 -1.30729049e-01 3.31888676e-01 1.14521146e+00
9.32912976e-02 -1.60371751e-01 -6.25115097e-01 -2.57246383e-02
-1.71488893e+00 -8.09540272e-01 -1.61848053e-01 2.15985250e+00
1.01678252e+00 1.63395137e-01 4.36087966e-01 2.21491665e-01
6.86430395e-01 2.32319534e-01 -3.02373946e-01 -8.19344699e-01
-1.54347783e-02 -3.38728130e-01 -2.45661914e-01 4.73619461e-01
-5.43776989e-01 8.10006082e-01 6.35229588e+00 6.89451218e-01
-8.87468040e-01 6.22784793e-02 4.36146110e-01 2.84325659e-01
-5.50379574e-01 -4.71700877e-02 -4.03915524e-01 5.09510934e-01
6.28929675e-01 -5.20762742e-01 3.34449887e-01 7.19936728e-01
3.09893906e-01 -4.79132459e-02 -1.00463486e+00 8.27465296e-01
-5.65297902e-02 -4.03258801e-01 2.62841731e-02 -4.08850580e-01
1.65933579e-01 -7.20851541e-01 -1.61749437e-01 2.66818821e-01
1.01970389e-01 -7.97429681e-01 5.65294027e-01 5.50300360e-01
4.89303768e-01 -5.77461958e-01 9.30791974e-01 6.28840625e-01
-6.84100688e-01 2.84753352e-01 -2.66550422e-01 -3.46448988e-01
1.03425533e-02 4.47431445e-01 -2.94755578e-01 5.45460224e-01
3.36703748e-01 2.31246114e-01 -3.30745995e-01 8.44576478e-01
-4.88998473e-01 1.15060799e-01 -6.97806701e-02 -6.28176212e-01
2.50680923e-01 -3.67316782e-01 3.87146652e-01 1.02792120e+00
2.37441227e-01 3.59587461e-01 2.35015437e-01 8.30729842e-01
4.11758900e-01 6.97193623e-01 -2.70199955e-01 -2.25150302e-01
7.40664840e-01 1.49878669e+00 -6.09436393e-01 -2.17660189e-01
-3.11797142e-01 9.07986760e-01 2.11283714e-01 -6.41502962e-02
-8.64171624e-01 -5.94393730e-01 4.98281181e-01 -2.29677975e-01
-1.80804372e-01 -5.97538911e-02 -2.66276300e-01 -1.04014111e+00
-6.91069365e-02 -1.13516438e+00 2.84127593e-01 -5.78678846e-01
-1.28730762e+00 9.39991176e-01 -9.68580917e-02 -1.48885059e+00
-1.06750935e-01 -4.28788781e-01 -6.17979228e-01 8.84405196e-01
-1.18547606e+00 -7.95232594e-01 -6.74601912e-01 4.65990156e-01
3.19441497e-01 3.10727581e-02 9.63859439e-01 3.49790463e-03
-4.11924064e-01 6.27647698e-01 -3.09527755e-01 -1.21393673e-01
1.09238887e+00 -1.12825680e+00 -1.47595137e-01 3.47660542e-01
-3.80962580e-01 8.62739086e-01 7.28155315e-01 -3.75300884e-01
-8.40303302e-01 -3.63456726e-01 1.22345304e+00 -1.91595510e-01
6.73931003e-01 6.68108091e-02 -9.06186700e-01 3.95813882e-02
2.44569764e-01 -7.22511888e-01 9.36549485e-01 2.77462453e-01
-3.58567774e-01 -9.57683697e-02 -1.08465755e+00 3.99718165e-01
7.19161987e-01 -3.93732697e-01 -9.10033822e-01 2.10655063e-01
7.27085829e-01 -2.50545859e-01 -1.14019835e+00 2.03934550e-01
5.79399168e-01 -1.54188442e+00 5.13911486e-01 -4.70034420e-01
5.00304818e-01 -3.27707291e-01 6.05330653e-02 -1.43027401e+00
-6.92463696e-01 -5.99432230e-01 1.97846442e-01 1.50139785e+00
3.57961535e-01 -8.06949139e-01 2.48440683e-01 8.19157124e-01
-1.40307784e-01 -8.18402827e-01 -5.14679313e-01 -7.89276004e-01
3.60126197e-02 3.92604694e-02 1.02541173e+00 1.13357031e+00
8.57449234e-01 8.09623957e-01 -5.38286805e-01 -4.27766174e-01
2.62800381e-02 2.08438650e-01 7.75189638e-01 -1.48022509e+00
-2.71386027e-01 -8.74946535e-01 -1.08934194e-01 -1.17367280e+00
-2.31907256e-02 -5.55855870e-01 -2.30732232e-01 -1.85822916e+00
3.46090853e-01 -3.68030906e-01 -3.49248558e-01 3.90980840e-01
-4.28516060e-01 -1.35013282e-01 2.80233800e-01 3.96459222e-01
-7.51601636e-01 5.59083223e-01 1.49659073e+00 -7.60813504e-02
-4.04089719e-01 -3.29352543e-02 -1.20934987e+00 8.38181913e-01
9.68487859e-01 -5.63447252e-02 -4.74932522e-01 -2.19466999e-01
2.70270050e-01 3.10242563e-01 3.45167294e-02 -8.27707052e-01
1.71409458e-01 -6.50599003e-01 -9.67613831e-02 -1.15049556e-01
1.32267803e-01 -5.15102625e-01 -2.16401294e-02 4.84410942e-01
-5.43363273e-01 1.45940762e-02 -6.47298515e-01 1.28669634e-01
-5.46379805e-01 -4.27193284e-01 8.04556370e-01 -1.87741309e-01
-7.34421730e-01 -4.84921634e-02 -2.16929778e-01 1.38227701e-01
8.95488262e-01 -4.39183712e-01 -3.67934704e-01 -6.97043896e-01
-4.61224824e-01 5.06832242e-01 6.19933069e-01 5.56314468e-01
4.93134856e-01 -1.27222967e+00 -7.15828538e-01 -6.19048774e-02
3.77375185e-01 -5.87444246e-01 1.35756880e-01 1.03169608e+00
-4.38016176e-01 4.54383433e-01 -4.53445554e-01 -2.25288361e-01
-1.21077967e+00 4.17871535e-01 2.31303811e-01 -3.90553355e-01
-1.84473038e-01 5.37756622e-01 3.40826273e-01 -5.48695803e-01
2.72272199e-01 -2.40969032e-01 -8.65003586e-01 2.69520640e-01
5.27514696e-01 4.24317002e-01 -1.88343182e-01 -6.68565035e-01
-1.96377486e-01 5.77611744e-01 1.09625138e-01 -2.61164419e-02
8.77301455e-01 -3.72982949e-01 -2.90957510e-01 6.34242594e-01
5.99702954e-01 8.55559409e-02 -6.87606633e-01 -1.36185467e-01
-8.90266709e-03 -3.36966485e-01 -3.00204456e-01 -9.73351598e-01
-7.28873372e-01 6.71413779e-01 1.13794520e-01 7.93764591e-01
1.14758801e+00 -2.06608102e-01 7.59910285e-01 5.18641949e-01
5.72129250e-01 -1.34976506e+00 3.63089681e-01 7.64654040e-01
8.96909058e-01 -1.15042317e+00 -2.38517642e-01 -3.35782588e-01
-1.26018918e+00 1.29344749e+00 8.94305706e-01 3.95483136e-01
3.49623263e-01 -3.27192545e-01 2.88546324e-01 -1.54471040e-01
-5.99169970e-01 -2.46771246e-01 3.12154591e-01 3.62669110e-01
8.72909188e-01 2.34189332e-01 -1.24364650e+00 9.70225155e-01
-4.79565531e-01 -2.33008862e-01 3.69843543e-01 5.36057591e-01
-8.35491061e-01 -1.08282995e+00 -9.56202596e-02 4.28515136e-01
-1.85460865e-01 2.57614732e-01 -9.03550386e-01 5.94838262e-01
-8.31907988e-02 1.52210212e+00 -4.42863643e-01 -7.65149355e-01
5.98693013e-01 -1.15555279e-01 2.28087693e-01 -6.22830868e-01
-8.85535836e-01 -1.58193588e-01 3.48027647e-01 -4.70760822e-01
-6.49596930e-01 -1.67679042e-01 -1.13940752e+00 -5.94480157e-01
-5.67403316e-01 6.75120294e-01 2.90153563e-01 1.11593139e+00
-3.82600278e-02 2.69962400e-01 9.76250529e-01 -1.82685927e-01
-9.13340807e-01 -1.23414254e+00 -7.47540355e-01 6.07584298e-01
-1.06138997e-02 -7.63712168e-01 -5.51730156e-01 -6.26008391e-01] | [12.823145866394043, 8.208946228027344] |
afbf454a-077e-4b3d-aeb3-a8aa2bf18d06 | a-stable-multi-scale-kernel-for-topological | 1412.6821 | null | http://arxiv.org/abs/1412.6821v1 | http://arxiv.org/pdf/1412.6821v1.pdf | A Stable Multi-Scale Kernel for Topological Machine Learning | Topological data analysis offers a rich source of valuable information to
study vision problems. Yet, so far we lack a theoretically sound connection to
popular kernel-based learning techniques, such as kernel SVMs or kernel PCA. In
this work, we establish such a connection by designing a multi-scale kernel for
persistence diagrams, a stable summary representation of topological features
in data. We show that this kernel is positive definite and prove its stability
with respect to the 1-Wasserstein distance. Experiments on two benchmark
datasets for 3D shape classification/retrieval and texture recognition show
considerable performance gains of the proposed method compared to an
alternative approach that is based on the recently introduced persistence
landscapes. | ['Ulrich Bauer', 'Stefan Huber', 'Roland Kwitt', 'Jan Reininghaus'] | 2014-12-21 | a-stable-multi-scale-kernel-for-topological-1 | http://openaccess.thecvf.com/content_cvpr_2015/html/Reininghaus_A_Stable_Multi-Scale_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Reininghaus_A_Stable_Multi-Scale_2015_CVPR_paper.pdf | cvpr-2015-6 | ['3d-shape-retrieval'] | ['computer-vision'] | [ 6.79224804e-02 -3.93241763e-01 -2.17309535e-01 -2.07122833e-01
-2.89880008e-01 -5.77242017e-01 6.88210011e-01 6.20060742e-01
-4.00324136e-01 6.14184976e-01 -2.38878503e-01 -5.29842913e-01
-7.42825866e-01 -7.45121300e-01 -3.69322091e-01 -1.14891911e+00
-4.49786752e-01 1.47442505e-01 7.85430253e-01 -1.74369335e-01
5.71295202e-01 9.44167793e-01 -1.50600815e+00 -2.26634666e-01
8.73140395e-01 1.00818408e+00 -2.44561210e-01 6.12468541e-01
-2.76119560e-02 2.83025593e-01 -2.33030081e-01 -1.35161221e-01
2.73451302e-02 -7.48016834e-02 -5.95763803e-01 -1.45128936e-01
4.62871790e-01 3.84162903e-01 -2.38247827e-01 1.12183392e+00
2.56025225e-01 9.25775245e-02 1.18051744e+00 -1.17420530e+00
-7.00707257e-01 -2.03398809e-01 -4.51327145e-01 3.50563377e-01
1.53468087e-01 -2.96393782e-01 8.37545931e-01 -6.88260317e-01
7.96273530e-01 7.99750865e-01 7.95812905e-01 4.36244644e-02
-1.55925882e+00 1.79784924e-01 -4.67534006e-01 4.77755994e-01
-1.13410473e+00 -6.59929514e-02 1.10441184e+00 -6.97821379e-01
5.55049241e-01 4.12777364e-01 6.31957173e-01 7.23033428e-01
5.65979838e-01 5.30335248e-01 1.63884723e+00 -5.57537615e-01
3.01289678e-01 3.22348982e-01 6.52193487e-01 1.08721828e+00
4.38643873e-01 -9.21278819e-02 -2.21485019e-01 -3.79781544e-01
9.92856860e-01 -3.27897489e-01 -1.75830647e-01 -1.18980622e+00
-1.14672828e+00 9.31737781e-01 3.84696007e-01 5.62868059e-01
-1.73541550e-02 -2.00089350e-01 3.98726493e-01 3.49469543e-01
6.21234238e-01 9.35915709e-02 -2.94671115e-02 -4.97554503e-02
-4.95157570e-01 9.55466256e-02 8.16252410e-01 4.51527208e-01
6.47696555e-01 -1.17960975e-01 1.47422597e-01 7.22528636e-01
1.51934698e-02 6.93083942e-01 2.50225902e-01 -4.01207030e-01
-2.22841203e-02 7.85099983e-01 1.30865648e-01 -1.45909989e+00
-5.71417511e-01 -1.68316007e-01 -9.72549438e-01 4.93946642e-01
7.42187679e-01 5.63476801e-01 -3.86094898e-01 1.39924002e+00
5.03609359e-01 1.79712757e-01 5.36962003e-02 8.06999803e-01
4.43960220e-01 4.65611219e-01 -4.04791176e-01 -1.43400028e-01
9.27603662e-01 -6.37930691e-01 -4.91964251e-01 6.58205271e-01
3.64922345e-01 -5.35975754e-01 1.14044762e+00 3.80724192e-01
-5.08949280e-01 -1.39585465e-01 -1.00464475e+00 3.29763889e-01
-6.74475133e-01 3.33434612e-01 8.23529720e-01 8.72355759e-01
-9.78778422e-01 7.55107105e-01 -8.84771943e-01 -6.90880597e-01
2.23970473e-01 1.15280487e-01 -6.59031630e-01 3.37409794e-01
-7.70662010e-01 9.77925301e-01 2.69897074e-01 -8.78078025e-03
-5.72099946e-02 -2.33361974e-01 -5.74202836e-01 -2.88975596e-01
-2.36166939e-02 -3.11456084e-01 4.75068957e-01 -3.28853607e-01
-1.56939375e+00 9.75010991e-01 2.73525510e-02 -4.41574663e-01
5.30222654e-01 1.98915824e-01 -1.91253781e-01 4.57636833e-01
-1.50780261e-01 -2.50060827e-01 1.06596744e+00 -8.74547899e-01
-4.43319008e-02 -4.86516237e-01 -2.41672680e-01 -3.80574077e-01
-4.94650304e-01 -2.81738877e-01 2.53489763e-01 -4.55070615e-01
3.12909186e-01 -8.47309768e-01 -1.03069633e-01 7.50292540e-02
-3.14418703e-01 -4.29389834e-01 8.88228238e-01 -3.03540170e-01
1.05236471e+00 -1.96995866e+00 2.94538200e-01 6.25564873e-01
1.34899914e-01 3.31475168e-01 9.54561830e-02 5.91015816e-01
-1.61090598e-01 -1.00167878e-01 -3.48285496e-01 1.46254674e-01
7.09321052e-02 3.90878320e-02 -4.38746452e-01 1.04551816e+00
2.44403213e-01 6.61374867e-01 -8.77336204e-01 -4.83157605e-01
4.56210852e-01 5.41312337e-01 -6.29215017e-02 -2.74251163e-01
1.62794903e-01 1.29622534e-01 -6.34894967e-01 6.89938426e-01
7.40158439e-01 -2.24803552e-01 -1.77255005e-01 -2.26804942e-01
-3.17325056e-01 -1.88461810e-01 -1.08380949e+00 1.26013768e+00
-2.92622764e-02 7.67020583e-01 -2.19138891e-01 -1.55035424e+00
1.10132647e+00 -1.18027702e-01 5.16478181e-01 -3.22277933e-01
9.49551314e-02 2.27329344e-01 -4.24755424e-01 -4.19463247e-01
2.40028948e-01 -7.79988393e-02 4.32747230e-02 2.09615067e-01
-1.61048204e-01 1.64901644e-01 6.76461905e-02 -9.64460075e-02
1.28458059e+00 -1.09630831e-01 2.29389623e-01 -8.10005426e-01
9.65319514e-01 -1.52905202e-02 -4.17525098e-02 5.41586816e-01
-3.51629883e-01 1.66790932e-01 8.73239517e-01 -5.39600909e-01
-1.00509822e+00 -1.42097878e+00 -7.72693813e-01 3.57042313e-01
1.64543986e-01 -3.17672223e-01 -6.88352346e-01 -5.89276016e-01
3.38014066e-01 -8.51345062e-02 -7.83923328e-01 -1.60471112e-01
-2.93714911e-01 -7.51857221e-01 5.92873871e-01 1.61422893e-01
4.20424461e-01 -6.58506513e-01 -7.91861236e-01 -5.82037047e-02
2.61709541e-01 -8.94861817e-01 1.25794625e-02 1.18334256e-01
-1.26072454e+00 -1.34973085e+00 -7.33006358e-01 -7.18601525e-01
6.82516158e-01 2.98838139e-01 4.83793736e-01 -2.18885735e-01
-6.00016594e-01 5.92597961e-01 -4.22661841e-01 4.94671166e-02
-3.54842156e-01 4.15463448e-02 2.99727768e-01 3.23688269e-01
1.27647445e-01 -7.29328156e-01 -3.60594064e-01 4.17405307e-01
-8.64405990e-01 -1.58237293e-01 5.79319298e-01 1.04818130e+00
5.33379197e-01 3.40368122e-01 5.55085838e-01 -4.65485543e-01
6.87068284e-01 -6.77243248e-02 -8.61419559e-01 4.39382583e-01
-7.97521472e-01 3.15013766e-01 5.36116302e-01 -4.82178122e-01
-5.99482417e-01 1.16436794e-01 4.03789401e-01 -2.77173817e-01
-1.20884836e-01 5.28992355e-01 9.59501863e-02 -7.39192724e-01
5.64464092e-01 5.65559983e-01 2.48344168e-01 -5.28332114e-01
3.95735979e-01 5.74077189e-01 4.79497582e-01 -7.10616291e-01
9.74174976e-01 7.27270782e-01 6.08213186e-01 -1.52042425e+00
-2.31502414e-01 -5.42845130e-01 -8.43745351e-01 -1.75265044e-01
6.90671861e-01 -9.85795259e-02 -8.90227497e-01 6.19342864e-01
-9.49220955e-01 5.70775904e-02 -5.92619069e-02 4.38676000e-01
-1.00570929e+00 6.10626101e-01 -4.08565581e-01 -1.06455290e+00
4.68660891e-02 -6.32333577e-01 8.64797771e-01 1.45451903e-01
2.42757365e-01 -1.28892398e+00 4.99787509e-01 6.08658791e-02
5.62762916e-01 7.24170804e-01 1.31699252e+00 -4.50624436e-01
-4.44646209e-01 -5.15458822e-01 -4.82565880e-01 3.57096195e-01
8.78904164e-02 3.24691892e-01 -6.09031320e-01 -2.12342888e-01
-4.77611385e-02 5.87776788e-02 1.06648576e+00 3.54020536e-01
7.62790382e-01 -3.43909338e-02 -2.74030209e-01 3.84908259e-01
1.50774419e+00 -9.10116211e-02 3.84269625e-01 4.69560236e-01
4.57283199e-01 6.41158462e-01 4.78843689e-01 2.75149733e-01
1.03376359e-01 7.22535312e-01 7.13656396e-02 7.88560733e-02
3.04871231e-01 -7.07823783e-02 1.60154656e-01 6.43331945e-01
-3.76383305e-01 5.13944924e-01 -1.00759482e+00 3.30769509e-01
-2.21194410e+00 -9.11246657e-01 -4.82689410e-01 2.58917689e+00
5.92357218e-01 2.28599861e-01 2.28710875e-01 3.69696885e-01
6.79269791e-01 2.02291578e-01 -1.96767285e-01 -3.80483896e-01
-5.42327225e-01 2.64959604e-01 5.51045775e-01 5.48781037e-01
-1.45615149e+00 7.43815184e-01 6.42064571e+00 8.52815628e-01
-1.26258993e+00 -8.18783715e-02 3.39414030e-02 4.59169745e-01
-3.63629237e-02 -2.51180809e-02 -3.06730121e-01 2.55852431e-01
6.31661355e-01 -2.38825783e-01 9.10836533e-02 8.92523706e-01
2.76927110e-02 -4.34958458e-01 -7.57333934e-01 1.12218285e+00
8.18491057e-02 -1.37772191e+00 6.03203140e-02 2.63120443e-01
4.64361399e-01 -2.70972043e-01 2.00137272e-01 -3.16213429e-01
-4.08521175e-01 -7.08502233e-01 2.66967684e-01 7.88793206e-01
4.66051012e-01 -7.52905309e-01 5.64531446e-01 2.43189573e-01
-1.18342412e+00 2.23002702e-01 -6.06104255e-01 1.34353675e-02
-1.34833694e-01 9.60618675e-01 -5.39446414e-01 7.12199330e-01
3.47947657e-01 8.29195559e-01 -9.33679700e-01 1.46806431e+00
1.53658330e-01 3.69057298e-01 -3.72495711e-01 -4.27017301e-01
1.76766187e-01 -6.93588972e-01 9.35795546e-01 1.24026692e+00
1.33090943e-01 -1.70111328e-01 -1.76848188e-01 6.52923763e-01
5.16252518e-01 4.59864199e-01 -1.14615571e+00 -2.57092834e-01
-8.85275081e-02 1.29090154e+00 -1.35794556e+00 -5.33023402e-02
-2.03767717e-01 1.02476382e+00 3.16302985e-01 2.36904070e-01
-6.19306743e-01 -6.67207062e-01 4.95526433e-01 1.15661912e-01
2.81507641e-01 -7.76892602e-01 -2.51127720e-01 -1.40580547e+00
3.10232252e-01 -1.13791108e-01 7.60299936e-02 -3.91425937e-01
-1.39839244e+00 3.60877037e-01 1.80337474e-01 -1.18074822e+00
2.88196862e-01 -1.15206158e+00 -5.95561326e-01 5.84911704e-01
-1.37137663e+00 -1.20697224e+00 1.95450746e-02 6.50586784e-01
-2.19208926e-01 -7.61822611e-02 9.25774574e-01 -3.89017761e-02
-3.57327729e-01 2.11289316e-01 6.42876983e-01 2.03828607e-02
6.59485877e-01 -1.59945238e+00 1.83669515e-02 5.69477975e-01
3.38111699e-01 5.92123449e-01 7.23369062e-01 -5.12305558e-01
-1.76347101e+00 -4.81028080e-01 9.15162146e-01 -6.55363977e-01
1.06124282e+00 -2.51754194e-01 -9.59327817e-01 -1.61582362e-02
-3.58310372e-01 2.72841662e-01 6.70349538e-01 1.05131291e-01
-5.51986217e-01 -2.15574116e-01 -1.05581629e+00 2.37125337e-01
7.34639406e-01 -6.60791636e-01 -7.39678204e-01 1.77891493e-01
8.85567889e-02 1.26480153e-02 -9.11302507e-01 3.93339097e-01
8.63687515e-01 -1.29866850e+00 1.08623064e+00 -6.88585997e-01
-6.97249966e-03 -3.66003066e-01 -2.76302010e-01 -9.49000001e-01
-2.44184300e-01 -3.87586355e-01 -1.09452449e-01 8.35431039e-01
7.18125775e-02 -9.71047699e-01 7.36033499e-01 1.13807410e-01
2.63604969e-01 -9.74336982e-01 -1.39170229e+00 -1.35749269e+00
1.72916993e-01 -1.65319189e-01 -3.96088921e-02 9.17167902e-01
2.87623048e-01 4.15378958e-02 -5.04359938e-02 1.39338970e-01
1.07735860e+00 3.85360301e-01 7.17218637e-01 -1.73305392e+00
-6.52774423e-02 -8.72050107e-01 -1.22224414e+00 -5.12241483e-01
3.14313859e-01 -8.83199692e-01 -5.04860580e-01 -1.33915329e+00
3.34975077e-03 -6.06903553e-01 -5.35465658e-01 1.26553059e-01
2.96264380e-01 2.50681698e-01 -2.35512897e-01 1.24219351e-01
-4.90501165e-01 5.78313291e-01 1.00568259e+00 -1.22336149e-02
-3.28927070e-01 9.09812525e-02 -1.28261253e-01 7.53670871e-01
6.92994893e-01 -1.91059664e-01 -3.28830659e-01 2.47039318e-01
1.89688638e-01 -2.52619013e-02 8.28108966e-01 -1.02464676e+00
2.04413459e-01 -1.92784071e-01 1.66395754e-01 -5.16196072e-01
1.95846960e-01 -7.08255291e-01 9.69613250e-03 5.19960403e-01
-6.88045621e-02 -1.21378787e-01 1.46731272e-01 9.12705481e-01
-1.98756427e-01 -6.17842302e-02 8.13143194e-01 4.55447197e-01
-5.05184829e-01 4.41255122e-02 -3.29098552e-01 -2.22297192e-01
1.07990658e+00 -4.26855922e-01 -4.70880181e-01 5.37315756e-02
-8.13262820e-01 -2.91690737e-01 8.38453352e-01 2.99262315e-01
7.77408779e-01 -1.47752607e+00 -2.91125685e-01 2.80078471e-01
3.31308454e-01 -7.65595317e-01 1.15224369e-01 1.35367954e+00
-5.95731854e-01 7.31495619e-01 -4.86692071e-01 -8.75998914e-01
-1.46988785e+00 5.45783758e-01 1.03220440e-01 -6.95373341e-02
-9.34388578e-01 3.02453220e-01 -2.97996849e-01 -1.00084633e-01
9.50091630e-02 -5.37888467e-01 -2.18724348e-02 1.57480434e-01
3.63489240e-01 5.78769326e-01 8.97491425e-02 -7.00468659e-01
-7.22115219e-01 1.13188994e+00 3.97851050e-01 -6.82304725e-02
1.36896908e+00 9.81326699e-02 -4.72207814e-01 8.97369444e-01
1.29146683e+00 -3.30440886e-02 -9.63435471e-01 -1.59413606e-01
3.89165670e-01 -6.76108718e-01 -9.65324938e-02 -4.23837304e-01
-3.82896036e-01 8.77954185e-01 7.35581815e-01 8.51747751e-01
9.82476532e-01 1.34663716e-01 3.22060317e-01 7.38077819e-01
5.55821121e-01 -8.81428421e-01 -2.00526934e-04 3.65811497e-01
8.19133818e-01 -1.06939149e+00 -1.63784683e-01 -4.44450825e-01
-1.94834918e-01 1.48219943e+00 3.76372645e-03 -5.66035807e-01
1.03925323e+00 -1.20440580e-01 -1.93020344e-01 -1.25254676e-01
-3.81543756e-01 -4.25601393e-01 4.28499699e-01 7.25433946e-01
7.28046447e-02 2.23252654e-01 -6.26474142e-01 1.73122332e-01
3.06939408e-02 -1.51872262e-01 4.18938816e-01 9.10691500e-01
-8.02432060e-01 -1.20611000e+00 -2.35989645e-01 3.20368707e-01
-1.47941581e-03 2.96943724e-01 -5.19672990e-01 7.67434061e-01
-3.71082067e-01 4.13660973e-01 -3.70951802e-01 -3.47517878e-01
2.39102975e-01 2.67263740e-01 8.92782211e-01 -9.76117104e-02
1.73913613e-01 -1.22032084e-01 -1.44429028e-01 -4.31409150e-01
-6.36970758e-01 -6.33891404e-01 -9.34698105e-01 -2.86031961e-01
-2.87870497e-01 3.18350941e-01 8.74721467e-01 6.65381968e-01
1.32503316e-01 -6.86130598e-02 7.82342255e-01 -7.12446213e-01
-6.39617801e-01 -4.52445030e-01 -9.73597288e-01 1.40183061e-01
4.88405138e-01 -1.13062322e+00 -6.50059640e-01 -1.21853374e-01] | [7.5692596435546875, 4.073818206787109] |
92254e34-4685-4f9c-93b5-8575b40276c6 | mitigating-motion-sickness-with-optimization | 2301.07977 | null | https://arxiv.org/abs/2301.07977v1 | https://arxiv.org/pdf/2301.07977v1.pdf | Mitigating Motion Sickness with Optimization-based Motion Planning | The acceptance of automated driving is under the potential threat of motion sickness. It hinders the passengers' willingness to perform secondary activities. In order to mitigate motion sickness in automated vehicles, we propose an optimization-based motion planning algorithm that minimizes the distribution of acceleration energy within the frequency range that is found to be the most nauseogenic. The algorithm is formulated into integral and receding-horizon variants and compared with a commonly used alternative approach aiming to minimize accelerations in general. The proposed approach can reduce frequency-weighted acceleration by up to 11.3% compared with not considering the frequency sensitivity for the price of reduced overall acceleration comfort. Our simulation studies also reveal a loss of performance by the receding-horizon approach over the integral approach when varying the preview time and nominal sampling time. The computation time of the receding-horizon planner is around or below the real-time threshold when using a longer sampling time but without causing significant performance loss. We also present the results of experiments conducted to measure the performance of human drivers on a public road section that the simulated scenario is actually based on. The proposed method can achieve a 19\% improvement in general acceleration comfort or a 32% reduction in squared motion sickness dose value over the best-performing participant. The results demonstrate considerable potential for improving motion comfort and mitigating motion sickness using our approach in automated vehicles. | ['Tamas Keviczky', 'Barys Shyrokau', 'Yanggu Zheng'] | 2023-01-19 | null | null | null | null | ['motion-planning'] | ['robots'] | [ 5.09258062e-02 3.50034714e-01 -1.09920576e-01 -7.32366368e-02
-7.21422553e-01 -3.30373198e-01 5.30185461e-01 1.84388682e-01
-7.59965599e-01 6.61650836e-01 8.41099396e-02 -7.23667979e-01
-3.27720851e-01 -6.44833624e-01 -5.80287457e-01 -7.94723213e-01
-4.67404835e-02 -4.95251939e-02 3.62011105e-01 -3.94338638e-01
2.37489492e-01 6.69583559e-01 -1.86228788e+00 -4.21165317e-01
9.86738622e-01 7.56177008e-01 3.25933278e-01 6.02996707e-01
6.71292484e-01 4.50087845e-01 -4.97277230e-01 2.36199066e-01
3.43104959e-01 -2.26244897e-01 -3.62367392e-01 1.73438329e-03
-1.95400223e-01 -4.17233914e-01 -2.17048839e-01 4.89706129e-01
6.04149163e-01 6.63033605e-01 5.30736804e-01 -1.61696196e+00
5.12358308e-01 -5.70717216e-01 -2.90061444e-01 3.70452762e-01
2.95976341e-01 5.19322395e-01 6.69456124e-02 -3.04567009e-01
2.14571118e-01 8.38475108e-01 3.61988366e-01 3.79217565e-01
-7.03778982e-01 -2.43563697e-01 -2.63079852e-01 3.36627424e-01
-1.53155410e+00 -3.32624346e-01 5.04315138e-01 -3.38355690e-01
1.35338175e+00 7.67430544e-01 8.47159624e-01 1.89063251e-01
9.56380069e-01 2.90760845e-01 8.09758544e-01 -1.41218066e-01
4.86001432e-01 2.08728492e-01 -5.53425662e-02 4.70645875e-01
4.64104772e-01 6.09258711e-01 5.36547303e-02 -1.42011911e-01
-9.86424014e-02 -3.04299086e-01 -1.00348383e-01 -1.20277435e-01
-7.99805701e-01 8.02095234e-01 1.28919005e-01 -4.21559550e-02
-5.75279236e-01 2.17286468e-01 4.76799935e-01 -1.91469759e-01
2.16249749e-01 2.14193910e-01 -1.31387562e-01 -3.50604028e-01
-7.88481176e-01 8.61710668e-01 5.05280495e-01 8.16095233e-01
3.94799709e-01 1.34121865e-01 -4.87551212e-01 2.03141093e-01
2.46807948e-01 6.17800117e-01 1.63619742e-01 -1.16047823e+00
2.90213138e-01 3.42051834e-01 6.93067908e-01 -7.29004145e-01
-7.11333871e-01 -3.09057951e-01 -1.13228783e-01 6.47260845e-01
1.64865479e-01 -5.18320203e-01 -6.53931618e-01 1.44616127e+00
5.94816864e-01 -1.42339677e-01 6.68272227e-02 8.94101322e-01
1.97371200e-01 9.17218029e-01 1.26807973e-01 -6.35090649e-01
1.41074991e+00 -7.92753875e-01 -1.04888272e+00 -2.49975979e-01
9.49371219e-01 -7.08160579e-01 8.50410998e-01 3.49121153e-01
-1.12303174e+00 -5.11635900e-01 -1.52122116e+00 3.52575839e-01
-1.04431860e-01 -2.82430142e-01 3.52335200e-02 1.29116809e+00
-1.06778991e+00 2.20032960e-01 -6.92231417e-01 -2.41573766e-01
1.36231855e-02 3.14762175e-01 1.41199619e-01 -3.12286639e-03
-1.11051607e+00 1.52921164e+00 1.84079126e-01 3.33800502e-02
-7.86677003e-01 -7.32574582e-01 -9.02645469e-01 -1.33199170e-01
4.53801006e-01 -4.59499031e-01 1.33211398e+00 -2.32427210e-01
-1.26140189e+00 8.72159973e-02 -2.62921304e-01 -6.82895422e-01
6.45650923e-01 -2.74768144e-01 -5.94675422e-01 -5.84875001e-04
4.45292369e-02 7.03370035e-01 3.14376086e-01 -9.83263493e-01
-7.20368445e-01 1.67324096e-02 7.56555200e-02 6.04696453e-01
1.12887226e-01 -1.50942504e-01 4.41389829e-02 -3.41510884e-02
-5.37157178e-01 -1.46260560e+00 -6.85686588e-01 -5.34088552e-01
-9.00072306e-02 -2.81478286e-01 9.47939038e-01 -5.84213436e-01
1.28306174e+00 -1.93711734e+00 -4.20654356e-01 1.51517466e-01
-4.29939687e-01 2.44692609e-01 2.72116601e-01 4.32952225e-01
1.22604303e-01 -2.62992889e-01 -2.66722739e-01 2.06211269e-01
-1.61649778e-01 2.89952517e-01 -7.87031874e-02 8.26439798e-01
1.68418568e-02 5.61795533e-01 -8.13294053e-01 -1.31308377e-01
7.25800276e-01 5.64977527e-01 -4.73389238e-01 1.35049954e-01
2.25603282e-01 2.36336499e-01 -2.05271885e-01 2.33827844e-01
6.97873712e-01 7.79094338e-01 -4.34385031e-01 1.57131568e-01
-5.43347239e-01 2.03916822e-02 -9.12802458e-01 9.91829813e-01
-5.27429402e-01 7.61791408e-01 -3.29875916e-01 -4.65616226e-01
5.93624294e-01 3.95902336e-01 6.20959580e-01 -1.05498433e+00
2.57911175e-01 1.55772507e-01 6.18918389e-02 -6.46976292e-01
8.98376942e-01 -3.52780610e-01 -2.13768139e-01 5.44987582e-02
-8.90993893e-01 -3.27620715e-01 -6.82091862e-02 -7.32690319e-02
1.01458693e+00 -2.02438787e-01 2.85320222e-01 -7.82253563e-01
6.18085861e-01 3.20171744e-01 3.00262421e-01 1.35809839e-01
-6.91883922e-01 -1.88908726e-01 2.64109761e-01 -2.35022441e-01
-9.22101855e-01 -8.31925929e-01 -1.93506494e-01 5.48963666e-01
6.40351892e-01 1.39437243e-03 -1.03795767e+00 -1.47806808e-01
-8.26428905e-02 1.71304405e+00 -3.14244270e-01 -6.81849897e-01
-5.42784214e-01 -7.79507816e-01 4.45376337e-01 3.12704176e-01
4.39125150e-01 -5.29756725e-01 -1.74334157e+00 5.15111387e-02
-1.98477417e-01 -9.39633667e-01 -4.99651104e-01 -1.39659613e-01
-5.52921057e-01 -6.65321827e-01 -5.44333220e-01 -2.15831727e-01
6.49712861e-01 5.67982078e-01 5.04690170e-01 1.51184395e-01
-2.25241020e-01 4.68584359e-01 9.07875746e-02 -9.27174509e-01
-2.72092342e-01 -3.35152924e-01 8.62519667e-02 -2.84783036e-01
2.89087117e-01 3.57557982e-01 -1.04548419e+00 6.95550323e-01
-8.06814671e-01 -1.29272491e-01 1.40824631e-01 3.42569083e-01
2.46189415e-01 7.37047851e-01 6.52832925e-01 -4.84263189e-02
7.83355117e-01 -5.05778670e-01 -6.86451018e-01 -3.33980232e-01
-8.67831230e-01 -6.37619793e-02 1.27066627e-01 -7.36739710e-02
-1.19613016e+00 1.80078849e-01 2.68505588e-02 1.65114522e-01
-9.74636152e-02 5.46481535e-02 -2.12590635e-01 -2.33184621e-01
5.16729534e-01 -1.98788002e-01 1.41761690e-01 2.87043214e-01
1.76134586e-01 3.80848736e-01 4.11070287e-01 2.65510436e-02
6.81562185e-01 5.72960198e-01 4.99303967e-01 -9.79812205e-01
-1.37441894e-02 -6.52152240e-01 -5.19941859e-02 -9.32358384e-01
1.11895418e+00 -8.16935062e-01 -9.64516580e-01 2.43118703e-01
-7.80138314e-01 -2.94735551e-01 -1.38266131e-01 6.41386509e-01
-9.31108415e-01 3.54430765e-01 1.33772895e-01 -1.37790799e+00
3.91654894e-02 -1.44366193e+00 7.84500778e-01 1.90154746e-01
-4.19061512e-01 -7.12688446e-01 -9.39723030e-02 3.88266534e-01
4.56034929e-01 7.22247064e-01 4.53180104e-01 1.12166345e-01
-4.78870571e-01 -3.89852881e-01 2.52832800e-01 3.50012444e-02
-1.27720803e-01 -3.37227732e-01 -8.77716601e-01 -4.97825593e-01
3.57380599e-01 3.24849904e-01 4.22649711e-01 7.39296615e-01
5.52204609e-01 -2.97209084e-01 -4.56116617e-01 5.13387211e-02
1.54859090e+00 8.94237936e-01 9.30122375e-01 5.54455459e-01
1.70412585e-01 1.13718712e+00 1.56490993e+00 5.32160103e-01
2.84526914e-01 1.01636279e+00 6.68163180e-01 -2.20377415e-01
3.66296351e-01 1.83761463e-01 4.82659787e-01 1.28842279e-01
-1.83524251e-01 -2.97378331e-01 -8.63087893e-01 9.00323510e-01
-1.55754495e+00 -9.35237944e-01 -5.11718631e-01 2.67811346e+00
1.23780265e-01 5.10661423e-01 4.88229990e-01 4.61622804e-01
5.94815314e-01 -9.72064435e-02 -2.24173605e-01 -1.23418033e+00
5.31528890e-01 -3.45692426e-01 1.37665308e+00 9.68798816e-01
-6.67521417e-01 2.16506094e-01 6.52379417e+00 8.03665340e-01
-6.95897877e-01 1.20135486e-01 6.83659196e-01 -5.53852320e-01
-2.77358592e-01 1.00544378e-01 -6.76713884e-01 3.83872896e-01
1.81865561e+00 -7.74137378e-01 9.34588090e-02 7.28255630e-01
1.04918945e+00 -1.04752171e+00 -7.78439522e-01 3.85586083e-01
-7.49158859e-02 -7.73091793e-01 -5.16347826e-01 4.65387940e-01
4.38210756e-01 -3.85037065e-01 -4.80917729e-02 1.66025534e-01
-4.20224816e-01 -9.30071115e-01 8.41221809e-01 4.01200175e-01
4.91773129e-01 -1.64924586e+00 8.23602915e-01 6.41209304e-01
-1.12662446e+00 -1.19041957e-01 -8.54997784e-02 -2.98636258e-01
9.02762532e-01 4.06683505e-01 -1.15429282e+00 3.77136141e-01
2.87663490e-01 -4.61020708e-01 -2.28862777e-01 1.15016508e+00
3.54676157e-01 3.10519785e-01 -3.64276052e-01 -3.71390998e-01
5.63225925e-01 -1.15848444e-01 7.20699728e-01 1.07692230e+00
4.67019707e-01 2.61341691e-01 -1.44125417e-01 2.75359482e-01
1.04499435e+00 -1.20351478e-01 -9.27659094e-01 4.20990586e-01
1.71048716e-01 8.18514287e-01 -6.89241111e-01 -2.92821646e-01
-2.69106716e-01 6.64646745e-01 -7.57845938e-01 3.02619696e-01
-1.42691445e+00 -5.43980241e-01 7.11305618e-01 6.49399519e-01
-1.60303637e-01 -3.04267824e-01 -5.74682832e-01 1.72057942e-01
1.08496398e-01 -5.11661246e-02 8.09729397e-02 -7.11274087e-01
-4.75734770e-02 5.03287673e-01 8.49687099e-01 -1.43701041e+00
-5.12768984e-01 -2.96936363e-01 -4.56603050e-01 1.15655684e+00
-1.37470198e+00 -2.77197093e-01 -2.21082434e-01 2.67821848e-01
7.08904147e-01 1.49014220e-01 2.44876325e-01 4.15599912e-01
-3.15448105e-01 4.22813863e-01 -2.10515305e-01 -1.03762317e+00
8.77275765e-02 -6.54642940e-01 1.67329103e-01 9.57747698e-01
-1.25120163e+00 2.15379581e-01 1.47152758e+00 -8.07222784e-01
-1.26147330e+00 -1.16528583e+00 1.05451548e+00 -4.27187890e-01
8.10668245e-02 -2.64764186e-02 -5.95305085e-01 7.91482255e-02
3.80838215e-01 -5.88755012e-01 2.96368837e-01 -6.64173543e-01
7.76251554e-01 -1.09963961e-01 -1.59781218e+00 9.37371731e-01
5.18625796e-01 -1.99799880e-01 -3.25925320e-01 2.17239961e-01
8.55154335e-01 -4.67503816e-01 -5.23441255e-01 5.61066091e-01
4.79745865e-01 -9.64052498e-01 7.93440163e-01 6.62549287e-02
-1.63322791e-01 -4.90141988e-01 4.44370806e-02 -1.08819294e+00
-1.13676459e-01 -8.07775319e-01 3.78193259e-01 6.06255114e-01
3.67789507e-01 -6.87817395e-01 5.71858525e-01 1.05530190e+00
-5.28307319e-01 -9.36484396e-01 -1.49226272e+00 -1.02705038e+00
-2.47557089e-01 -8.34007859e-01 3.26276898e-01 1.64919659e-01
8.28627050e-02 -1.40652299e-01 -3.32911700e-01 4.32979763e-01
5.83661795e-01 -7.69301414e-01 6.93977118e-01 -4.48070168e-01
9.46179479e-02 -1.77510798e-01 -5.91118753e-01 -3.19007397e-01
-2.97468483e-01 -3.18674654e-01 5.43751478e-01 -1.65711617e+00
-1.90547377e-01 1.07636184e-01 -9.55828726e-02 -7.11141452e-02
-1.13473609e-01 -2.28322402e-01 4.05917987e-02 -3.55386347e-01
1.07351646e-01 4.99845535e-01 1.12703419e+00 1.37185648e-01
-4.31728482e-01 3.25148940e-01 -2.87561417e-01 5.20583332e-01
8.12459469e-01 -3.45393896e-01 -1.13949227e+00 3.13946158e-01
3.74858007e-02 4.90003884e-01 4.43985194e-01 -1.27155161e+00
-7.22911283e-02 -5.46689928e-01 -2.11993158e-01 -1.05349457e+00
3.84335101e-01 -1.08632040e+00 6.41038656e-01 1.22831571e+00
3.51258703e-02 4.47899878e-01 7.82092035e-01 7.41039813e-01
3.30387294e-01 -1.01875663e-01 8.38491738e-01 4.21298385e-01
-7.62103021e-01 -3.00795496e-01 -1.24687660e+00 -5.98075330e-01
1.73389757e+00 -5.89473307e-01 -9.82601494e-02 -5.08411527e-01
-2.85416514e-01 3.64209145e-01 4.06210542e-01 5.92802763e-01
3.76294166e-01 -1.33672845e+00 -4.30691749e-01 -1.01936191e-01
9.47492272e-02 -5.05947053e-01 6.23525262e-01 1.03319764e+00
-8.03210020e-01 9.89476204e-01 -3.25307637e-01 -5.79053164e-01
-1.27636266e+00 7.93184280e-01 5.03061950e-01 5.29199955e-04
-5.22877216e-01 2.64867425e-01 2.40538523e-01 1.44360155e-01
-1.13521121e-01 -4.00062084e-01 -1.35456353e-01 -2.37894848e-01
4.46172923e-01 1.19596183e+00 4.40782607e-01 -8.57823789e-01
-6.70847178e-01 2.00920299e-01 5.41430652e-01 -6.18939996e-01
4.62295830e-01 -4.27012056e-01 4.89260972e-01 2.87016779e-01
1.04120183e+00 -8.21983814e-03 -1.51617694e+00 7.88262904e-01
-2.85957503e-04 -4.61563736e-01 3.90698105e-01 -6.71410859e-01
-5.09149849e-01 3.32751960e-01 1.11459112e+00 7.63675272e-02
1.28276610e+00 -4.29343313e-01 9.26728368e-01 1.13880605e-01
4.84400421e-01 -1.35488784e+00 -4.88268644e-01 1.86832286e-02
8.74434412e-01 -8.89107466e-01 4.79556881e-02 -4.75802600e-01
-8.40400517e-01 4.42181617e-01 5.52710176e-01 -6.25110492e-02
5.43137312e-01 2.80458599e-01 -4.69999313e-02 -1.42449528e-01
-7.27244318e-01 -1.83508933e-01 3.03499460e-01 4.67700422e-01
8.82911161e-02 4.40670282e-01 -1.20890582e+00 1.07961081e-01
-1.29338011e-01 -7.37438202e-02 8.01370442e-01 1.07872236e+00
-8.72142017e-01 -4.46181715e-01 -6.63107395e-01 1.81477830e-01
-8.27710256e-02 5.56243598e-01 2.53029883e-01 1.05767262e+00
3.15515548e-01 1.53895509e+00 2.32374161e-01 -3.66024882e-01
8.17923129e-01 -3.75226922e-02 1.12660341e-01 -8.26433599e-02
-3.56318533e-01 8.89114141e-02 6.22982383e-01 -6.92123353e-01
-3.48448485e-01 -7.53064752e-01 -1.67902100e+00 -4.44441080e-01
-2.04370454e-01 1.25990793e-01 9.58629608e-01 1.15295100e+00
1.49238914e-01 7.93285429e-01 7.86089420e-01 -8.30974162e-01
-1.51596770e-01 -4.46128815e-01 -3.40130955e-01 -1.90578118e-01
5.73909581e-01 -8.32319558e-01 -4.68442380e-01 -4.48691458e-01] | [5.510324478149414, 1.4764595031738281] |
3095dc3f-c7ef-4e53-8492-12f8e7c55cf2 | natural-language-descriptions-of-human | null | null | https://aclanthology.org/W16-3205 | https://aclanthology.org/W16-3205.pdf | Natural Language Descriptions of Human Activities Scenes: Corpus Generation and Analysis | null | ['Yoshihiko Gotoh', 'Nouf Alharbi'] | 2016-08-01 | null | null | null | ws-2016-8 | ['video-description'] | ['computer-vision'] | [-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] | [-7.390260696411133, 3.621987819671631] |
0941f7e6-f4da-43c4-8cb8-3b6b325d965f | t-star-lite-a-fast-time-risk-optimal-motion | 2008.13048 | null | https://arxiv.org/abs/2008.13048v1 | https://arxiv.org/pdf/2008.13048v1.pdf | T$^{\star}$-Lite: A Fast Time-Risk Optimal Motion Planning Algorithm for Multi-Speed Autonomous Vehicles | In this paper, we develop a new algorithm, called T$^{\star}$-Lite, that enables fast time-risk optimal motion planning for variable-speed autonomous vehicles. The T$^{\star}$-Lite algorithm is a significantly faster version of the previously developed T$^{\star}$ algorithm. T$^{\star}$-Lite uses the novel time-risk cost function of T$^{\star}$; however, instead of a grid-based approach, it uses an asymptotically optimal sampling-based motion planner. Furthermore, it utilizes the recently developed Generalized Multi-speed Dubins Motion-model (GMDM) for sample-to-sample kinodynamic motion planning. The sample-based approach and GMDM significantly reduce the computational burden of T$^{\star}$ while providing reasonable solution quality. The sample points are drawn from a four-dimensional configuration space consisting of two position coordinates plus vehicle heading and speed. Specifically, T$^{\star}$-Lite enables the motion planner to select the vehicle speed and direction based on its proximity to the obstacle to generate faster and safer paths. In this paper, T$^{\star}$-Lite is developed using the RRT$^{\star}$ motion planner, but adaptation to other motion planners is straightforward and depends on the needs of the planner | ['Zongyuan Shen', 'Thomas A. Wettergren', 'James P. Wilson', 'Shalabh Gupta'] | 2020-08-29 | null | null | null | null | ['optimal-motion-planning'] | ['robots'] | [-1.47797152e-01 1.63026094e-01 -2.47713909e-01 8.62318352e-02
-6.81040525e-01 -3.10793161e-01 4.16416556e-01 -1.07466713e-01
-6.68448746e-01 9.70879614e-01 -6.43176317e-01 -9.62681592e-01
-7.49385536e-01 -1.12730408e+00 -5.53978205e-01 -7.70984411e-01
-4.88329083e-01 4.69719738e-01 4.83584374e-01 -6.85534179e-01
4.01829422e-01 7.27795184e-01 -1.51041746e+00 -8.65543008e-01
1.03589118e+00 7.43482232e-01 5.08040965e-01 7.52221406e-01
-4.09078114e-02 -7.30284825e-02 -1.35294393e-01 1.20534800e-01
4.13110763e-01 -6.23901963e-01 -6.42252445e-01 -3.09137285e-01
-2.38883421e-01 -4.16571833e-02 -1.55608222e-01 8.58238161e-01
2.14546725e-01 7.00918198e-01 7.66492844e-01 -1.42622316e+00
3.58758658e-01 -2.61032254e-01 -5.32968402e-01 2.96599716e-01
1.89409494e-01 5.26380718e-01 5.72156131e-01 -7.23434687e-01
7.41910100e-01 9.79530156e-01 5.43815732e-01 4.09612566e-01
-8.70773315e-01 -6.09872282e-01 3.31912860e-02 9.14939791e-02
-1.72773170e+00 -4.81187627e-02 6.44945383e-01 -4.17324275e-01
1.15870118e+00 1.54892460e-01 7.05959380e-01 1.58243567e-01
7.35292077e-01 3.56348455e-01 6.40816510e-01 -3.21448535e-01
6.01799905e-01 -3.63038987e-01 -1.60499960e-01 9.51477766e-01
3.95284206e-01 6.71151102e-01 1.93454072e-01 -1.19859613e-01
8.15969408e-01 -4.25187200e-01 -8.55289474e-02 -6.09618008e-01
-8.39949787e-01 1.09818912e+00 1.75957128e-01 1.48470685e-01
-3.21330160e-01 3.39430034e-01 1.36490479e-01 -4.19520997e-02
3.03319860e-02 4.29209322e-01 -2.38930434e-01 -4.24298853e-01
-7.68694937e-01 8.52617264e-01 3.77386272e-01 1.29621911e+00
9.35608149e-01 3.78232241e-01 7.02781975e-02 3.55999410e-01
4.86881942e-01 7.66623199e-01 -1.83443546e-01 -1.38421941e+00
6.21635377e-01 4.45579499e-01 6.83527470e-01 -8.36450994e-01
-8.05657744e-01 -2.21604686e-02 -5.43219447e-01 9.71135616e-01
4.23765838e-01 -4.66878682e-01 -6.65174186e-01 1.67602229e+00
5.82863152e-01 3.32474448e-02 3.77765834e-01 7.73139179e-01
1.36008769e-01 8.37133348e-01 -3.73487435e-02 -6.12556458e-01
9.69615698e-01 -7.02826440e-01 -5.33828080e-01 -1.75935224e-01
1.00812006e+00 -3.92186910e-01 9.30473387e-01 1.05722658e-01
-1.09757507e+00 -2.35104322e-01 -1.27572417e+00 5.58602989e-01
-2.94650882e-01 -3.06213796e-01 2.25308552e-01 5.66787064e-01
-1.19657826e+00 7.38423526e-01 -7.86761940e-01 -4.40505087e-01
1.63966026e-02 3.39294463e-01 -8.41667727e-02 -1.09520331e-01
-1.10709965e+00 1.22323704e+00 1.36417329e-01 -3.28735225e-02
-8.04531932e-01 -2.75551826e-01 -1.10064912e+00 -2.94044197e-01
5.33428907e-01 -4.25041676e-01 1.24879253e+00 9.80958417e-02
-1.91788971e+00 5.65883406e-02 -5.30662596e-01 -1.40596390e-01
4.99478489e-01 2.44847924e-01 -2.32210040e-01 2.33935013e-01
4.27389234e-01 4.77214813e-01 2.97818065e-01 -1.21343815e+00
-7.94772327e-01 2.86358278e-02 -2.15179980e-01 3.18933666e-01
4.70665514e-01 -2.35984668e-01 -2.23349139e-01 -2.77468234e-01
1.24046259e-01 -9.91562247e-01 -8.03804874e-01 -1.89451545e-01
2.90893614e-02 -4.12076354e-01 8.77843022e-01 -9.84576941e-02
1.38092017e+00 -1.72954762e+00 1.36265129e-01 4.93983209e-01
-1.78718835e-01 3.63139689e-01 4.72940020e-02 6.61117792e-01
3.33478153e-01 1.02343559e-01 -5.23098290e-01 5.15637770e-02
-1.64899945e-01 3.62727433e-01 1.10986352e-01 5.76757371e-01
-3.32653485e-02 4.55915362e-01 -1.20122302e+00 -3.51466119e-01
5.84955513e-01 8.45751017e-02 -5.89048088e-01 -2.25674599e-01
-3.35770011e-01 3.96988899e-01 -8.70886028e-01 4.40756828e-01
6.38599813e-01 4.42029089e-01 -1.25038356e-01 4.96061981e-01
-1.00942099e+00 -1.50661126e-01 -1.27790678e+00 1.30534685e+00
-3.00208211e-01 4.54688638e-01 4.45603400e-01 -8.56902540e-01
1.19294035e+00 2.22602293e-01 7.20119298e-01 -7.37735391e-01
3.45221400e-01 5.34058630e-01 -1.82351306e-01 -4.16059613e-01
7.78255284e-01 -2.59158701e-01 -2.55702317e-01 4.03572530e-01
-6.19446039e-01 -7.96864510e-01 3.88695747e-01 -2.26300862e-02
1.11746097e+00 2.97724247e-01 4.45161074e-01 -7.12404549e-01
6.56617939e-01 6.25572383e-01 7.22858727e-01 6.12732351e-01
-5.69836140e-01 -2.88953315e-02 3.30430031e-01 -3.45967501e-01
-1.02644360e+00 -8.65467072e-01 -2.11037308e-01 4.30690527e-01
8.82413805e-01 -3.27607542e-01 -6.60577834e-01 -3.09215307e-01
-8.62582624e-02 1.21068704e+00 -2.38889471e-01 -1.51598593e-02
-1.04256058e+00 -4.68576491e-01 2.88553655e-01 3.17001671e-01
5.73303759e-01 -8.62391412e-01 -1.13217497e+00 6.03684783e-01
4.99090217e-02 -7.10753024e-01 -3.41944844e-01 -5.50888702e-02
-6.99012756e-01 -1.11419952e+00 -4.80006278e-01 -5.49658537e-01
6.28837466e-01 5.00432134e-01 4.67152506e-01 1.45015977e-02
-1.73268825e-01 3.22596431e-01 -5.19548118e-01 -4.04658407e-01
-3.39502484e-01 -2.38247827e-01 1.93581998e-01 -4.49616224e-01
1.14910953e-01 -2.50481158e-01 -6.60659730e-01 7.64714122e-01
-5.93662083e-01 -1.94294862e-02 2.64803439e-01 5.58295667e-01
9.08142507e-01 5.01936436e-01 5.49672544e-01 -5.31507609e-03
4.67967600e-01 -5.44338644e-01 -1.05498004e+00 -4.08363521e-01
-8.35798144e-01 2.35818759e-01 8.69449377e-01 -1.47062883e-01
-5.87362587e-01 4.47528176e-02 -3.94574761e-01 -3.17335159e-01
5.00157475e-02 3.67156655e-01 -1.26167133e-01 -2.10289001e-01
4.82312113e-01 1.49618030e-01 1.76361382e-01 -2.65184522e-01
2.67614275e-01 2.26930708e-01 3.15481246e-01 -4.14429426e-01
9.01530802e-01 4.46872979e-01 5.91899633e-01 -6.61426663e-01
2.46003240e-01 -2.16372281e-01 -4.17216569e-01 -5.30761838e-01
8.99336457e-01 -1.94712460e-01 -1.04613602e+00 1.15296744e-01
-9.09034967e-01 -8.51767600e-01 -2.77554512e-01 6.62997663e-01
-1.13637578e+00 2.81673372e-01 6.97079450e-02 -1.17230773e+00
1.22460322e-02 -1.57603121e+00 7.22952604e-01 4.21293497e-01
-1.93328515e-01 -8.25411439e-01 9.59613472e-02 -2.49003723e-01
5.01023114e-01 6.72142923e-01 8.01331282e-01 1.54007822e-01
-6.61005974e-01 -3.75248134e-01 9.71254334e-02 -3.35027188e-01
-1.53939769e-01 4.97295968e-02 2.55544540e-02 -5.28842270e-01
-9.53363627e-02 3.34598124e-01 4.05605167e-01 7.00785041e-01
4.45082843e-01 -3.83089870e-01 -8.50056589e-01 4.24531728e-01
1.49475968e+00 9.83998537e-01 5.45734644e-01 8.68982315e-01
1.91750497e-01 6.14880562e-01 1.18390155e+00 3.53796929e-01
7.51683891e-01 9.33146775e-01 7.31344223e-01 1.02672733e-01
3.08578283e-01 -1.25816733e-01 3.66524518e-01 1.97509024e-02
-2.28222981e-01 -1.05718017e-01 -1.10817683e+00 5.84162951e-01
-1.73824060e+00 -1.00473106e+00 -3.29721272e-01 2.24222112e+00
3.36140126e-01 8.45879391e-02 3.21648598e-01 2.20797405e-01
6.30162179e-01 1.95633888e-01 -5.68147898e-01 -8.16504955e-01
3.05344403e-01 9.19834152e-02 9.81289864e-01 1.15828717e+00
-6.90702438e-01 9.58781362e-01 5.63425112e+00 1.05617273e+00
-1.05989158e+00 -1.69584110e-01 7.75038451e-02 2.89378706e-02
-3.09995502e-01 2.84316927e-01 -9.61770833e-01 5.17606020e-01
1.00042701e+00 -7.20332086e-01 2.14682370e-01 1.01804614e+00
8.04446936e-01 -7.05441236e-01 -4.08688515e-01 5.34396350e-01
-4.23791498e-01 -1.33029413e+00 -5.41051805e-01 2.81439066e-01
4.64680403e-01 -2.58124202e-01 -1.07986525e-01 2.38741085e-01
5.39635956e-01 -9.44692373e-01 8.28486085e-01 4.54451144e-02
1.09107709e+00 -1.19572186e+00 5.00708640e-01 7.14639187e-01
-1.92479920e+00 -1.85491532e-01 -1.88245922e-01 -8.69821534e-02
1.00716209e+00 5.83251864e-02 -7.21750438e-01 7.47134268e-01
7.20121920e-01 1.18725531e-01 2.35229313e-01 9.08716261e-01
1.71132921e-03 9.09036770e-02 -4.72791106e-01 -4.33010966e-01
6.94160223e-01 -5.19209802e-01 1.14665747e+00 8.83452237e-01
8.13793778e-01 5.38334668e-01 3.06293398e-01 6.73658669e-01
8.93145561e-01 -1.53897600e-02 -9.36919868e-01 4.22408760e-01
4.86356646e-01 7.99797356e-01 -8.07719350e-01 1.12672299e-02
7.61581883e-02 2.53101945e-01 -7.63854086e-02 5.04544497e-01
-9.53589797e-01 -9.14765358e-01 9.27293837e-01 4.21512306e-01
3.41105312e-01 -9.02442932e-01 -3.79159242e-01 -3.02064180e-01
-2.51526654e-01 -1.85845211e-01 1.84001490e-01 -5.02825856e-01
-3.32171112e-01 7.51232743e-01 6.93581164e-01 -1.59713078e+00
-6.18177295e-01 -5.01405358e-01 -5.96560538e-01 9.62805033e-01
-1.43195856e+00 -3.19231182e-01 -1.23641640e-03 5.58355868e-01
6.47132695e-01 -2.04055935e-01 5.87128043e-01 -2.20375717e-01
-5.92051268e-01 4.73211646e-01 3.87100220e-01 -3.40778887e-01
-2.54274398e-01 -7.84372449e-01 4.08597350e-01 9.29995418e-01
-9.19081867e-01 2.61747360e-01 1.19472754e+00 -7.75813282e-01
-1.56473732e+00 -1.18588400e+00 7.81904697e-01 -5.15309051e-02
3.99629891e-01 1.93527535e-01 -5.68977177e-01 3.52637947e-01
-1.07416548e-01 -2.31394604e-01 8.57789740e-02 -7.03258574e-01
5.36038399e-01 -1.16576269e-01 -1.66029513e+00 1.09944594e+00
9.85081911e-01 2.30104029e-01 -1.19940534e-01 -4.45564538e-02
6.33298159e-01 -4.40836906e-01 -6.42902255e-01 7.66993284e-01
3.82848591e-01 -8.17407846e-01 6.70823157e-01 5.53747416e-02
-3.71647745e-01 -5.99970579e-01 -1.40862674e-01 -1.22762465e+00
-3.49768221e-01 -1.17901278e+00 2.90913284e-01 5.26213884e-01
4.53486800e-01 -9.46982801e-01 7.60200679e-01 5.32591462e-01
-6.55336142e-01 -1.10714591e+00 -1.76930845e+00 -1.18056226e+00
4.22843546e-01 -6.53612196e-01 4.55363840e-01 3.42217118e-01
2.64594883e-01 -2.38922223e-01 1.45104434e-02 3.33079010e-01
7.30865717e-01 -1.99724019e-01 7.32253790e-01 -7.92795777e-01
1.83510373e-03 -6.91568196e-01 -1.37027830e-01 -1.13827085e+00
-8.57912563e-03 -4.72892463e-01 4.64114636e-01 -1.82812476e+00
-6.05601013e-01 -1.03550804e+00 3.30333918e-01 2.37531781e-01
1.03262112e-01 -1.97466925e-01 1.25587836e-01 7.50843957e-02
-1.43922061e-01 7.44352937e-01 1.34983361e+00 1.75576389e-01
-6.91999793e-01 1.47989094e-01 -2.06260011e-01 6.76534951e-01
9.32396114e-01 -3.18430424e-01 -6.60357058e-01 -8.82226899e-02
4.32436801e-02 7.96097577e-01 9.06105191e-02 -1.03843558e+00
2.12146357e-01 -1.00921106e+00 -4.82978851e-01 -7.98240006e-01
5.21687090e-01 -5.78536630e-01 3.37632418e-01 7.98465550e-01
2.32980311e-01 4.06108648e-01 4.75327849e-01 6.27580762e-01
9.88665447e-02 -3.10460627e-01 1.06473756e+00 -8.25691968e-02
-7.33465791e-01 3.77855033e-01 -9.05239761e-01 -2.40301490e-01
1.65987468e+00 -7.63415515e-01 -1.21394314e-01 -3.51105511e-01
-3.55506957e-01 5.90173781e-01 5.70777833e-01 4.15423103e-02
6.83239937e-01 -1.24996018e+00 -4.04484957e-01 1.82308242e-01
-1.23418182e-01 4.03766423e-01 2.33341545e-01 7.80559361e-01
-8.31776500e-01 6.22489512e-01 9.87330079e-02 -5.95588923e-01
-7.86980093e-01 4.70531464e-01 4.99125987e-01 -2.23027200e-01
-7.56854177e-01 7.87922382e-01 -2.57147491e-01 -1.96756288e-01
-2.33176351e-01 -1.82097733e-01 -1.92934722e-01 -4.54098731e-01
3.91086638e-01 8.89089286e-01 -2.73747146e-01 -1.04179144e+00
-4.74174052e-01 9.77809608e-01 5.60325682e-01 -7.03513324e-01
9.42028880e-01 -3.83505464e-01 3.11412185e-01 -2.47559547e-02
9.30536091e-01 -2.62850881e-01 -1.66830540e+00 3.35937560e-01
-3.30664739e-02 -5.00552237e-01 -7.81619251e-02 -2.25473806e-01
-7.31392980e-01 3.90890121e-01 2.23404855e-01 -8.85118917e-03
8.01687956e-01 -1.38415858e-01 8.18191767e-01 1.71935074e-02
1.01827335e+00 -1.28654456e+00 -1.81913495e-01 8.86999488e-01
7.42571831e-01 -7.05622017e-01 -2.13869587e-02 -3.48485887e-01
-4.71882463e-01 9.88865197e-01 7.38014519e-01 -1.74369097e-01
6.14248514e-01 2.01047674e-01 -3.43616717e-02 -5.57164066e-02
-4.30340022e-01 -2.09607884e-01 -2.46489957e-01 6.23564124e-01
-2.67786831e-01 2.47476529e-02 -1.04800677e+00 4.19375271e-01
-2.68126726e-01 -1.71518829e-02 6.23840749e-01 1.21738803e+00
-9.75940824e-01 -1.16060221e+00 -5.51543236e-01 1.70565043e-02
3.97819966e-01 3.67302865e-01 3.61525178e-01 1.09758782e+00
2.43970633e-01 1.35150850e+00 9.71892029e-02 -4.06913370e-01
5.53596914e-01 -1.03454560e-01 1.20303914e-01 -1.61907881e-01
-5.25829159e-02 -3.82270366e-02 2.60370672e-01 -6.72817647e-01
-7.34651787e-03 -6.62009776e-01 -2.12998176e+00 -5.89806080e-01
-1.77838221e-01 5.61074793e-01 7.57120967e-01 1.06104088e+00
4.02710736e-01 -1.28891198e-02 9.34081614e-01 -1.40315080e+00
-3.29054445e-01 -4.50540811e-01 -3.76654863e-01 -3.15293133e-01
1.59035742e-01 -1.14833677e+00 -4.18973118e-01 -6.37342095e-01] | [5.1613593101501465, 1.6341394186019897] |
2d67de18-3b13-4c1f-8dfe-c98b87c2671e | modality-aware-contrastive-instance-learning | 2207.05500 | null | https://arxiv.org/abs/2207.05500v1 | https://arxiv.org/pdf/2207.05500v1.pdf | Modality-Aware Contrastive Instance Learning with Self-Distillation for Weakly-Supervised Audio-Visual Violence Detection | Weakly-supervised audio-visual violence detection aims to distinguish snippets containing multimodal violence events with video-level labels. Many prior works perform audio-visual integration and interaction in an early or intermediate manner, yet overlooking the modality heterogeneousness over the weakly-supervised setting. In this paper, we analyze the modality asynchrony and undifferentiated instances phenomena of the multiple instance learning (MIL) procedure, and further investigate its negative impact on weakly-supervised audio-visual learning. To address these issues, we propose a modality-aware contrastive instance learning with self-distillation (MACIL-SD) strategy. Specifically, we leverage a lightweight two-stream network to generate audio and visual bags, in which unimodal background, violent, and normal instances are clustered into semi-bags in an unsupervised way. Then audio and visual violent semi-bag representations are assembled as positive pairs, and violent semi-bags are combined with background and normal instances in the opposite modality as contrastive negative pairs. Furthermore, a self-distillation module is applied to transfer unimodal visual knowledge to the audio-visual model, which alleviates noises and closes the semantic gap between unimodal and multimodal features. Experiments show that our framework outperforms previous methods with lower complexity on the large-scale XD-Violence dataset. Results also demonstrate that our proposed approach can be used as plug-in modules to enhance other networks. Codes are available at https://github.com/JustinYuu/MACIL_SD. | ['Yuejie Zhang', 'Rui Feng', 'Ying Cheng', 'Jinyu Liu', 'Jiashuo Yu'] | 2022-07-12 | null | null | null | null | ['anomaly-detection-in-surveillance-videos', 'anomaly-detection-in-surveillance-videos'] | ['computer-vision', 'methodology'] | [ 1.64120257e-01 -1.32569417e-01 -2.84899026e-01 -1.78015843e-01
-1.04541230e+00 -6.62009954e-01 8.01382780e-01 1.48636132e-01
-2.92468160e-01 4.24778908e-01 3.78574580e-01 -8.95786658e-02
-4.77445684e-03 -5.15000939e-01 -6.88326955e-01 -1.02378619e+00
-2.13223472e-02 1.70097023e-01 -1.39083723e-02 -4.61690538e-02
-9.74176079e-02 2.39543170e-01 -1.73514211e+00 7.68807709e-01
5.55360317e-01 8.94014299e-01 -2.17307374e-01 8.01423490e-01
-1.81109563e-01 1.16178107e+00 -3.77332181e-01 -4.30021614e-01
7.03975633e-02 -5.98950982e-01 -7.05190182e-01 7.42536783e-02
5.57463408e-01 -3.09064716e-01 -5.36996245e-01 8.78678858e-01
7.57118046e-01 2.22719133e-01 8.10007870e-01 -1.88297200e+00
-2.80650437e-01 6.68033957e-01 -1.02060008e+00 2.71623880e-01
5.42514443e-01 4.40649807e-01 9.60527599e-01 -1.12363207e+00
4.07158941e-01 1.47388399e+00 4.15833116e-01 5.65323293e-01
-1.25118315e+00 -1.08832490e+00 3.10230583e-01 5.17649591e-01
-1.40688837e+00 -5.48764408e-01 1.24436998e+00 -6.43675983e-01
4.66070652e-01 5.66422939e-01 7.32039571e-01 1.79312861e+00
-5.79796255e-01 1.14377952e+00 8.88485432e-01 -1.30734727e-01
-3.92740518e-02 2.99918979e-01 4.79132831e-02 6.47904217e-01
-2.22277537e-01 1.18696496e-01 -7.75246620e-01 -1.90579295e-01
4.64228511e-01 2.05962718e-01 -2.05204964e-01 -2.57440299e-01
-9.90770519e-01 8.55203331e-01 2.20088884e-01 2.80033767e-01
-1.58370793e-01 3.32804471e-02 8.02089691e-01 1.72558129e-01
5.31301796e-01 -1.48631796e-01 1.00649171e-01 -2.18819574e-01
-8.93565118e-01 1.49487806e-02 2.74571508e-01 5.58567584e-01
6.82444274e-01 -1.30932471e-02 -4.79254544e-01 9.91460860e-01
1.94523752e-01 4.84199226e-01 2.03930333e-01 -8.79773855e-01
6.66642070e-01 5.09974539e-01 -4.72295403e-01 -9.94993746e-01
-4.10822779e-01 -1.63208872e-01 -1.00331092e+00 1.83571167e-02
3.87010872e-01 -5.48478141e-02 -7.05435693e-01 1.85487926e+00
5.79277813e-01 5.81298232e-01 -3.32624912e-02 1.01520753e+00
1.37932348e+00 7.83495247e-01 3.30334246e-01 -4.04320747e-01
1.19634414e+00 -8.58391643e-01 -6.49661541e-01 -1.06828762e-02
4.70765531e-01 -4.24542636e-01 1.25880444e+00 3.53646457e-01
-1.14118004e+00 -4.31035787e-01 -5.57180464e-01 8.56009945e-02
-2.12695971e-01 -1.21331565e-01 2.29931414e-01 4.11575973e-01
-6.52301848e-01 1.67923257e-01 -5.67930877e-01 -1.25604957e-01
6.75583959e-01 -4.35486138e-02 -5.93357623e-01 4.55411673e-02
-1.18006146e+00 2.20349118e-01 4.39740986e-01 1.79275975e-01
-1.19029260e+00 -6.67024255e-01 -1.16175878e+00 -2.19923228e-01
5.56153834e-01 -4.38572854e-01 8.45156789e-01 -1.35676825e+00
-1.03221905e+00 9.17043567e-01 -6.78514764e-02 5.32056764e-02
5.60958564e-01 1.24982782e-01 -3.77117813e-01 6.56596601e-01
-7.07364753e-02 6.49779081e-01 1.26787734e+00 -1.69090259e+00
-5.41155696e-01 -3.51899594e-01 9.80808362e-02 4.21487927e-01
-7.17102170e-01 4.37642671e-02 -6.74741983e-01 -7.90938675e-01
-7.77544007e-02 -5.43975532e-01 3.02050948e-01 -3.86561118e-02
-6.47846639e-01 -1.40333802e-01 8.36812258e-01 -5.87884605e-01
1.38200605e+00 -2.54213786e+00 3.66540670e-01 3.50816876e-01
5.13095617e-01 -1.84159596e-02 -3.09383631e-01 3.21000904e-01
-3.96058351e-01 4.74906730e-04 -2.03869119e-01 -6.03517711e-01
5.53323142e-02 1.74961433e-01 -3.23657632e-01 5.30077398e-01
4.14459527e-01 6.41355157e-01 -1.10604382e+00 -9.85431671e-01
3.91341090e-01 6.33752644e-01 -4.98075902e-01 2.90869981e-01
2.56093174e-01 7.39411056e-01 8.52623507e-02 1.04762793e+00
6.76167071e-01 1.02206841e-02 5.35057746e-02 -4.11006063e-01
1.50222182e-01 -1.78319335e-01 -1.10586023e+00 1.37974501e+00
-2.44141459e-01 5.44229209e-01 2.71443158e-01 -1.26761198e+00
5.07194459e-01 3.79646808e-01 5.71687639e-01 -8.24820518e-01
2.27561489e-01 -1.70903921e-01 -3.58352304e-01 -9.37148929e-01
2.39106357e-01 -4.07872021e-01 -2.14912608e-01 9.87425148e-02
2.78908074e-01 1.95099771e-01 1.32813171e-01 5.94630778e-01
9.73954737e-01 9.72004514e-03 -5.63752316e-02 5.07415652e-01
5.30023396e-01 -4.07866359e-01 4.44422901e-01 7.19740808e-01
-4.97810721e-01 8.41745973e-01 8.69947255e-01 6.76132366e-02
-5.79731226e-01 -1.46556222e+00 -1.09883711e-01 1.46629989e+00
2.48628363e-01 -4.67145085e-01 -4.54458863e-01 -8.78175378e-01
-1.81935150e-02 4.22687501e-01 -7.09364176e-01 -4.04423654e-01
-2.70267755e-01 -5.29410362e-01 7.73503184e-01 5.49581051e-01
4.44447547e-02 -1.12171388e+00 -1.21981196e-01 -2.10674599e-01
-4.67581481e-01 -9.30139244e-01 -1.71460569e-01 1.61009714e-01
-3.64323914e-01 -1.07273829e+00 -7.91418135e-01 -6.60674453e-01
4.25359726e-01 2.35979751e-01 8.09728503e-01 -8.88116807e-02
-3.05799067e-01 7.78515041e-01 -5.38895071e-01 3.65412943e-02
-3.24290216e-01 -3.33917707e-01 5.83614595e-02 4.95513380e-01
1.59468397e-01 -6.84350431e-01 -4.80252564e-01 4.21800986e-02
-1.06167233e+00 1.25031963e-01 2.11476877e-01 9.07112777e-01
4.25630838e-01 -1.31352156e-01 4.67798442e-01 -4.99710381e-01
3.93066466e-01 -9.24067438e-01 1.60049628e-02 2.11975630e-02
-3.32834050e-02 -3.64440084e-01 5.76265812e-01 -8.71899188e-01
-9.33635414e-01 1.20503679e-01 -1.79148745e-02 -1.03914344e+00
-3.58548403e-01 4.59334046e-01 -3.20272923e-01 2.77460665e-01
4.26544636e-01 1.31108344e-01 -5.81006557e-02 -2.61403322e-01
5.30217648e-01 7.37432003e-01 7.66924381e-01 -6.21347189e-01
9.90559936e-01 6.48506820e-01 -3.49847704e-01 -9.25401330e-01
-4.94123101e-01 -6.17901862e-01 -3.97521913e-01 -8.98021460e-01
8.54010582e-01 -1.09036589e+00 -7.67769873e-01 5.82857788e-01
-9.21519935e-01 -3.13229322e-01 -2.51802951e-01 3.99264157e-01
-3.00316930e-01 5.94309986e-01 -7.69333959e-01 -1.16761351e+00
9.12782922e-03 -9.23567474e-01 1.25464344e+00 2.11885810e-01
-1.82723224e-01 -8.07022572e-01 1.00249812e-01 6.68744802e-01
-2.83253580e-01 4.05703843e-01 6.70548797e-01 -6.44578040e-01
6.00678958e-02 -4.17389572e-02 -4.60276395e-01 3.13202828e-01
-1.89236820e-01 2.03183487e-01 -1.39487624e+00 -2.48528838e-01
-5.33433855e-01 -7.63693750e-01 1.17200935e+00 2.26658925e-01
1.34878445e+00 -2.91850001e-01 7.22498260e-03 5.59791267e-01
1.00185037e+00 -9.23101753e-02 4.26374584e-01 1.05024658e-01
1.11254549e+00 7.30942309e-01 5.26995778e-01 8.75745833e-01
4.12166804e-01 5.22232950e-01 7.67320633e-01 -4.29038167e-01
-1.04328990e-01 -3.05634588e-01 6.34323537e-01 7.22042561e-01
-1.54274836e-01 -1.56828865e-01 -9.27080095e-01 6.48420274e-01
-1.88231552e+00 -1.55012882e+00 -1.41447797e-01 2.04477167e+00
9.02322114e-01 2.13561375e-02 7.29892671e-01 4.77164388e-01
9.07747209e-01 5.06295860e-01 -2.96718687e-01 -1.79614499e-01
-3.10937613e-01 4.79881018e-02 -1.40783146e-01 4.63180542e-01
-1.41820741e+00 7.04905629e-01 4.50552320e+00 1.19958746e+00
-1.09507203e+00 3.89924109e-01 6.51511669e-01 -6.94214880e-01
-4.17977899e-01 -2.55864263e-01 -1.94928199e-01 5.87565243e-01
5.47393501e-01 2.26936191e-01 4.54993963e-01 4.36386883e-01
2.46207133e-01 -7.50778243e-03 -1.07348752e+00 1.32831323e+00
3.03175420e-01 -9.09009695e-01 8.46515745e-02 -2.72154599e-01
3.08552980e-01 -3.26519072e-01 3.21269333e-01 6.18141711e-01
-1.70723408e-01 -8.97545218e-01 1.08993447e+00 4.74386901e-01
8.81601751e-01 -9.46559966e-01 4.56966341e-01 2.56590426e-01
-1.54772079e+00 -5.05630732e-01 2.12228537e-01 5.53961731e-02
1.53976157e-01 3.49447250e-01 -3.64965141e-01 5.62984169e-01
9.34611738e-01 8.18538189e-01 -6.32632077e-01 7.38359272e-01
-3.47685963e-02 8.32342625e-01 -1.85352251e-01 3.45741719e-01
1.96725264e-01 -4.90750012e-04 8.31642270e-01 1.56664431e+00
-7.96276331e-02 -3.18277255e-02 2.06489459e-01 7.36412823e-01
4.07550558e-02 8.66850838e-02 -7.14910686e-01 -1.15929760e-01
4.56662476e-01 1.39950538e+00 -5.00478506e-01 -4.57153708e-01
-4.09321040e-01 9.85415757e-01 2.96787888e-01 5.55044174e-01
-1.30385315e+00 -2.65171170e-01 3.88783008e-01 -9.90654454e-02
6.00206899e-03 2.89230943e-01 -1.31967753e-01 -1.25319505e+00
8.44619870e-02 -8.76830280e-01 8.28272641e-01 -8.59671533e-01
-1.54469824e+00 3.53615224e-01 2.75382936e-01 -1.52560186e+00
-1.50360897e-01 -2.09427759e-01 -7.56654024e-01 3.34584594e-01
-1.24823248e+00 -1.35046268e+00 -5.00758290e-01 1.15193796e+00
4.69033033e-01 -5.34646325e-02 3.37411791e-01 5.51544607e-01
-9.47692990e-01 9.13921535e-01 -2.07783803e-01 3.19728166e-01
8.25155854e-01 -9.96654272e-01 -6.90481842e-01 7.44270682e-01
2.04638168e-01 1.38752580e-01 6.54631257e-01 -5.29559553e-01
-1.17803979e+00 -1.17557919e+00 4.20446724e-01 -2.62600392e-01
8.26122642e-01 -5.94548702e-01 -9.79511499e-01 3.05806071e-01
2.47391492e-01 1.82472944e-01 8.04857194e-01 3.54451523e-03
-8.63938093e-01 -1.74467579e-01 -1.01572001e+00 6.78277135e-01
1.07692516e+00 -9.65858996e-01 -4.52891678e-01 2.60480225e-01
5.61702728e-01 -1.52013958e-01 -5.87289512e-01 4.54495251e-01
5.41814208e-01 -1.04522455e+00 1.11163151e+00 -6.51904702e-01
8.07381988e-01 -8.56681317e-02 -1.75335824e-01 -9.37599897e-01
-1.44576997e-01 -4.90230143e-01 -4.00656313e-01 1.66670549e+00
2.27357730e-01 -3.64459664e-01 4.32733297e-01 6.72561377e-02
-3.20799612e-02 -5.30831575e-01 -1.12693608e+00 -6.94283664e-01
4.89724660e-03 -7.41732776e-01 1.08651586e-01 1.38299894e+00
4.78841215e-01 3.28768700e-01 -5.73057294e-01 2.90848225e-01
6.71161950e-01 -3.48358452e-02 6.52027667e-01 -9.60199594e-01
-3.98723751e-01 -5.89087903e-01 -4.46965128e-01 -4.55539048e-01
3.64326149e-01 -1.15680778e+00 -1.71341121e-01 -9.95129168e-01
7.39256918e-01 1.73437856e-02 -5.51388443e-01 8.34299207e-01
-3.04153830e-01 7.38117218e-01 4.39991266e-01 2.61001308e-02
-9.81330156e-01 7.27847874e-01 8.58226538e-01 -4.81361479e-01
-2.72551060e-01 -2.83019751e-01 -4.83186573e-01 7.05148637e-01
5.73622286e-01 -4.25785929e-01 -3.46000135e-01 2.37046238e-02
7.54161179e-02 1.87963530e-01 9.34935331e-01 -7.54317224e-01
-3.61182131e-02 -1.15517177e-01 3.17639112e-01 -6.34992838e-01
6.04639471e-01 -7.97180593e-01 -6.36317655e-02 7.74713829e-02
-4.48631525e-01 -2.99709737e-01 2.33136237e-01 6.65480018e-01
-3.50320995e-01 7.38463625e-02 6.58324420e-01 1.45355940e-01
-4.66129035e-01 2.49394864e-01 -5.05688488e-01 3.05035204e-01
1.11269987e+00 -2.66390413e-01 -4.21045899e-01 -6.45311296e-01
-1.01326334e+00 3.51742417e-01 1.59254715e-01 5.35647810e-01
8.03054929e-01 -1.64898765e+00 -6.68164253e-01 2.22684950e-01
4.73912686e-01 -2.14573890e-01 8.40913355e-01 1.32394493e+00
-8.08046723e-04 -3.25398982e-01 6.26866594e-02 -9.47154462e-01
-1.62228847e+00 6.09946549e-01 5.25126606e-02 6.71928376e-02
-3.70081335e-01 9.05771196e-01 4.16721672e-01 -4.44688469e-01
6.59376860e-01 2.91660875e-01 -3.66455108e-01 6.41253591e-01
5.55450022e-01 4.50298816e-01 -3.31175417e-01 -1.04471242e+00
-4.83017951e-01 3.76065820e-01 2.56675303e-01 -2.40439385e-01
1.24563539e+00 -6.55639917e-02 4.52175215e-02 8.11494410e-01
1.39447808e+00 2.32506786e-02 -1.19293821e+00 -1.09296024e-01
-5.27085423e-01 -2.99761415e-01 -3.10976710e-02 -4.71721679e-01
-1.32625890e+00 1.14372504e+00 7.41677344e-01 3.17796201e-01
1.37396359e+00 4.92046028e-01 5.17647326e-01 -7.15647787e-02
-2.37026408e-01 -1.08612728e+00 4.70218092e-01 2.82802761e-01
8.59407306e-01 -1.26603830e+00 -3.20387751e-01 -2.78178841e-01
-1.02113056e+00 9.03396487e-01 7.61585832e-01 2.02687249e-01
4.69089329e-01 3.25130701e-01 -1.41083598e-02 -2.13975161e-01
-6.75240040e-01 -4.53767776e-01 2.86482036e-01 6.55543089e-01
1.35395110e-01 -3.55279520e-02 3.53770866e-03 9.32789266e-01
2.91623622e-01 -3.50839257e-01 8.71207342e-02 8.74196529e-01
-6.41134903e-02 -5.40260255e-01 -7.56808162e-01 2.42206797e-01
-2.31723264e-01 -1.62237823e-01 -6.42128408e-01 6.68494761e-01
3.73735100e-01 1.05897355e+00 2.31708258e-01 -8.52508068e-01
1.77064493e-01 1.22561157e-01 4.21957940e-01 -2.83448398e-01
-7.53398776e-01 5.08502662e-01 1.45986348e-01 -6.42661512e-01
-4.51587796e-01 -5.96975327e-01 -1.28025532e+00 -1.39261752e-01
-1.13132179e-01 -1.44693673e-01 1.33095935e-01 8.83888781e-01
1.20399244e-01 4.62976575e-01 9.47602510e-01 -1.31870985e+00
-2.36108571e-01 -7.07002878e-01 -5.50894022e-01 8.27919722e-01
6.48915350e-01 -7.97742844e-01 -9.79851305e-01 1.50396805e-02] | [13.467679023742676, 4.733566761016846] |
7e244342-1fcd-434a-b78c-840ff4d52867 | inverse-cooking-recipe-generation-from-food | 1812.06164 | null | https://arxiv.org/abs/1812.06164v2 | https://arxiv.org/pdf/1812.06164v2.pdf | Inverse Cooking: Recipe Generation from Food Images | People enjoy food photography because they appreciate food. Behind each meal there is a story described in a complex recipe and, unfortunately, by simply looking at a food image we do not have access to its preparation process. Therefore, in this paper we introduce an inverse cooking system that recreates cooking recipes given food images. Our system predicts ingredients as sets by means of a novel architecture, modeling their dependencies without imposing any order, and then generates cooking instructions by attending to both image and its inferred ingredients simultaneously. We extensively evaluate the whole system on the large-scale Recipe1M dataset and show that (1) we improve performance w.r.t. previous baselines for ingredient prediction; (2) we are able to obtain high quality recipes by leveraging both image and ingredients; (3) our system is able to produce more compelling recipes than retrieval-based approaches according to human judgment. We make code and models publicly available. | ['Xavier Giro-i-Nieto', 'Michal Drozdzal', 'Amaia Salvador', 'Adriana Romero'] | 2018-12-14 | inverse-cooking-recipe-generation-from-food-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Salvador_Inverse_Cooking_Recipe_Generation_From_Food_Images_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Salvador_Inverse_Cooking_Recipe_Generation_From_Food_Images_CVPR_2019_paper.pdf | cvpr-2019-6 | ['recipe-generation'] | ['miscellaneous'] | [ 2.74971873e-01 3.05023137e-02 -2.48808190e-02 -4.84715372e-01
-7.12076902e-01 -9.07771766e-01 6.29050732e-01 5.88304043e-01
-1.56050637e-01 2.40197703e-01 7.70693600e-01 -1.48169650e-02
2.85444170e-01 -9.93038356e-01 -1.28721178e+00 -3.82543057e-01
2.32525975e-01 3.63597929e-01 -1.43182978e-01 -2.98071742e-01
8.53919312e-02 -9.83723029e-02 -1.55113697e+00 7.57439017e-01
6.94004655e-01 5.82269907e-01 5.78968644e-01 9.21059966e-01
-7.36169219e-02 1.07318974e+00 -4.72628884e-02 -5.03434241e-01
4.18235570e-01 -6.73111498e-01 -5.51192522e-01 4.73098367e-01
8.43095243e-01 -7.29821026e-01 -2.88305525e-02 9.25815225e-01
-4.52928506e-02 9.03294906e-02 7.23008156e-01 -7.51671255e-01
-1.37048376e+00 1.01387417e+00 -2.03362584e-01 -5.59826970e-01
7.54406452e-01 3.66862953e-01 9.87465739e-01 -8.17605793e-01
5.84953368e-01 1.04995310e+00 6.02053642e-01 6.06810808e-01
-1.58158791e+00 -1.51111707e-01 3.53121340e-01 1.28026426e-01
-8.70042503e-01 -4.24977779e-01 8.21648479e-01 -3.09569567e-01
9.90314245e-01 1.56017751e-01 1.08241689e+00 1.13495457e+00
-2.39218161e-01 9.68199909e-01 1.02670932e+00 -6.00441515e-01
1.07606284e-01 1.65819913e-01 2.22701043e-01 9.00605977e-01
3.80173437e-02 1.73442125e-01 -3.52044940e-01 3.61558497e-01
6.05710864e-01 1.27387509e-01 -1.95040271e-01 -4.72939372e-01
-1.57904720e+00 8.23754668e-01 6.24771357e-01 6.25565574e-02
-8.47524047e-01 2.64627248e-01 2.28047464e-02 2.58619487e-01
3.41302395e-01 3.38837177e-01 -3.40037495e-01 3.87718678e-01
-8.92863512e-01 3.84532481e-01 9.47201431e-01 9.55644846e-01
6.77425086e-01 -2.73005694e-01 5.21350615e-02 7.35997200e-01
2.98463762e-01 8.41839194e-01 -9.29599330e-02 -1.02566254e+00
2.61127621e-01 3.31683934e-01 4.23481047e-01 -1.02083516e+00
-4.58354622e-01 2.86585391e-01 -7.74733603e-01 2.39965945e-01
4.97308135e-01 1.68775663e-01 -9.67230737e-01 1.88461637e+00
3.34432483e-01 -8.68796650e-03 -1.68517716e-02 1.06102896e+00
1.17107487e+00 8.61371398e-01 7.07259238e-01 2.07583651e-01
1.54910469e+00 -1.34059274e+00 -7.67550588e-01 -2.10827410e-01
3.34026843e-01 -1.01023567e+00 1.28680599e+00 6.38834596e-01
-1.24216020e+00 -8.60159755e-01 -1.20794737e+00 -5.18344581e-01
-5.10425687e-01 4.13435280e-01 7.33028114e-01 4.83091235e-01
-1.12701559e+00 9.07780707e-01 -7.34801948e-01 -4.96472836e-01
2.34535322e-01 7.58000985e-02 -4.84201074e-01 -6.70800269e-01
-7.74988353e-01 9.04212296e-01 4.38178509e-01 3.44315707e-03
-9.67404306e-01 -8.01961720e-01 -1.30699587e+00 -3.70219462e-02
3.65367979e-01 -1.23436725e+00 1.30762684e+00 -1.25468373e+00
-1.58873749e+00 8.91740263e-01 5.79943322e-02 -2.06377015e-01
2.40126461e-01 -4.92191046e-01 -4.47559804e-01 3.40649843e-01
7.72485808e-02 1.35253334e+00 7.06075490e-01 -1.41523802e+00
-4.62413937e-01 -1.10660441e-01 4.82984304e-01 1.58470556e-01
3.27009745e-02 -2.69981533e-01 -1.37750193e-01 -6.01327300e-01
-1.57294631e-01 -8.59116018e-01 -2.94426233e-01 3.62416476e-01
-4.58225876e-01 1.50325432e-01 4.59337085e-02 -9.08074021e-01
8.80358994e-01 -2.08967257e+00 7.89343342e-02 -2.17459962e-01
2.19120197e-02 1.24309339e-01 -5.94273090e-01 5.00620723e-01
-5.85803390e-02 7.29770586e-02 -1.82731092e-01 -4.75580156e-01
5.43667078e-01 9.71063375e-02 -2.53800422e-01 4.47639108e-01
1.74685270e-01 1.05060101e+00 -1.17113590e+00 -3.71150523e-01
7.70592868e-01 7.52267778e-01 -7.84194887e-01 3.25347811e-01
-9.49390173e-01 1.07982352e-01 -2.65137464e-01 4.51756597e-01
7.05844522e-01 -3.36060375e-01 3.55572551e-01 -7.44170845e-01
-2.16644824e-01 3.26049834e-01 -1.10459733e+00 2.27353144e+00
-4.95323390e-01 4.16191183e-02 -1.66030452e-02 -6.69979095e-01
5.09256303e-01 4.44912881e-01 2.00316906e-01 -6.20910347e-01
-1.27752811e-01 -4.50538173e-02 -5.25486946e-01 -6.10481203e-01
5.98862231e-01 -1.74519077e-01 9.12143569e-03 8.04204106e-01
9.05432999e-02 -1.78601474e-01 5.37680089e-01 8.86678547e-02
6.25418723e-01 9.19345438e-01 4.14678782e-01 -3.11223686e-01
2.86672860e-01 4.34622496e-01 -1.25871256e-01 5.95053494e-01
5.10795563e-02 6.85296357e-01 -1.84034675e-01 -9.47698593e-01
-1.43374598e+00 -1.23526490e+00 4.56780940e-01 1.26713574e+00
1.28458545e-01 -5.49303532e-01 -7.32911825e-01 -4.53373820e-01
-1.44702122e-01 1.12352180e+00 -8.13791871e-01 1.37552723e-01
-7.76512682e-01 -3.90222549e-01 -1.73511729e-01 4.69629049e-01
4.04412299e-01 -1.37185180e+00 -6.79511487e-01 3.35473537e-01
-5.02846003e-01 -8.75723064e-01 -9.87595975e-01 -2.05095634e-02
-5.70717514e-01 -1.25191152e+00 -6.18801117e-01 -9.95302975e-01
5.70646584e-01 7.15214550e-01 1.86044681e+00 3.23040515e-01
-3.95828158e-01 5.64899683e-01 -3.91743004e-01 -1.77699238e-01
-9.58078325e-01 -2.35381067e-01 -4.54076350e-01 -3.51038009e-01
6.33730114e-01 -4.34967518e-01 -9.92327034e-01 -9.19616222e-02
-1.06530356e+00 7.46452928e-01 4.86474335e-01 3.93136770e-01
6.64732635e-01 -1.76309898e-01 2.02339455e-01 -9.59916532e-01
1.68955505e-01 -3.87586206e-01 -4.69101429e-01 5.22549510e-01
-4.39076424e-01 2.40525454e-01 6.79494858e-01 -5.01429081e-01
-1.13250935e+00 6.78416014e-01 -1.34279445e-01 -4.49283384e-02
-7.27278054e-01 4.78747711e-02 -4.59816912e-03 6.16484106e-01
6.18968368e-01 -1.80789065e-02 -1.74820065e-01 -6.59599662e-01
1.28155768e+00 1.46023631e-01 6.45984411e-01 -6.63494766e-01
5.76182544e-01 3.22123528e-01 -2.42761388e-01 -4.78016108e-01
-9.99834538e-01 -3.73365581e-01 -6.12192035e-01 -2.49356210e-01
1.24616337e+00 -1.13142049e+00 -9.28796470e-01 1.75059587e-01
-1.13084495e+00 -5.74682295e-01 -2.98967153e-01 5.68801820e-01
-6.67986274e-01 2.36488655e-01 -9.53151107e-01 -5.61299384e-01
-2.98250020e-01 -9.07484353e-01 1.15505481e+00 1.12790510e-01
-3.73746127e-01 -9.61339831e-01 2.34397173e-01 3.69375885e-01
2.72993296e-01 4.53312069e-01 1.02136743e+00 -3.00339721e-02
-6.85474157e-01 2.76830882e-01 -2.01676756e-01 7.54617453e-02
3.93176377e-01 1.38003323e-02 -9.27249730e-01 1.05930440e-01
-1.30362481e-01 -5.44870496e-01 1.00312257e+00 5.15146196e-01
5.69342732e-01 -4.17978406e-01 3.12222056e-02 3.37474197e-01
1.82418764e+00 -1.89317495e-01 4.80526596e-01 1.49774402e-01
6.90569758e-01 9.77532566e-01 2.90672719e-01 1.27501860e-01
7.60170281e-01 6.04228020e-01 4.51739490e-01 -5.32880843e-01
-6.28728092e-01 -8.50926995e-01 6.66068852e-01 6.07300639e-01
3.95556055e-02 -1.90499976e-01 -1.36859298e-01 6.88156188e-01
-1.91260254e+00 -1.17140818e+00 -4.02768910e-01 2.04700279e+00
8.65220785e-01 -4.23489004e-01 3.88289601e-01 -2.48340234e-01
5.27432621e-01 4.46662791e-02 -4.90926981e-01 -3.21387321e-01
-1.01593919e-01 1.75870568e-01 3.24961990e-01 6.03506207e-01
-1.48053062e+00 8.98335814e-01 6.67529440e+00 2.68606544e-01
-5.01271009e-01 -2.53304727e-02 7.05515146e-01 -1.03613235e-01
-5.08161545e-01 -2.78005779e-01 -3.33565980e-01 6.57705367e-02
7.35014915e-01 4.01520759e-01 1.11178577e+00 5.75130343e-01
1.06220149e-01 -1.32269770e-01 -1.59098291e+00 8.14420104e-01
2.80509412e-01 -1.25727010e+00 8.13915730e-02 -2.79730797e-01
6.25592053e-01 -2.46552199e-01 -1.26194775e-01 2.22833857e-01
1.01086593e+00 -1.04378438e+00 1.26629460e+00 4.93504226e-01
4.89496589e-01 -4.61279869e-01 1.38877034e-01 2.40082741e-01
-1.36288273e+00 1.89006209e-01 -3.92787725e-01 -2.76597198e-02
3.81995708e-01 4.56200004e-01 -4.49044973e-01 5.69130123e-01
4.94751364e-01 8.94992411e-01 -4.76243675e-01 6.83382332e-01
-7.06031442e-01 2.86914766e-01 -4.43418592e-01 1.06672108e-01
2.03246206e-01 -5.82775772e-01 -9.30233300e-02 1.39234567e+00
4.55360532e-01 2.77456790e-01 4.32054907e-01 1.26193678e+00
-2.99007539e-03 3.02411854e-01 -5.88339865e-01 -2.21954823e-01
-1.41153872e-01 1.40522981e+00 -7.75197744e-01 -4.57510710e-01
-6.26540482e-01 1.31835401e+00 2.43246883e-01 2.83867836e-01
-7.57797539e-01 5.04234970e-01 4.99549031e-01 -1.49876714e-01
5.34907997e-01 6.44947588e-02 -3.09137329e-02 -1.25426769e+00
-2.23478407e-01 -9.40948665e-01 3.39661747e-01 -1.21758723e+00
-1.50625288e+00 4.97772068e-01 -1.60760120e-01 -7.85557985e-01
-2.83082992e-01 -6.30085588e-01 -1.08442575e-01 6.07729912e-01
-1.62101007e+00 -1.58778977e+00 -1.77022785e-01 4.74172413e-01
7.04529762e-01 5.58836997e-01 1.40977788e+00 3.52485925e-01
-7.22807944e-02 -1.06888667e-01 -3.52757931e-01 1.20640909e-02
7.72712588e-01 -1.55066121e+00 5.04640698e-01 6.25750661e-01
8.43841851e-01 4.40630496e-01 1.00166106e+00 -6.34509206e-01
-1.52397478e+00 -1.06977415e+00 1.09141088e+00 -5.88887930e-01
3.01614642e-01 -4.34624702e-01 -5.98286927e-01 8.05864275e-01
7.74602234e-01 -3.64406377e-01 8.96701396e-01 1.80167362e-01
-8.16730559e-01 9.49191526e-02 -1.19016588e+00 6.32305443e-01
9.74699676e-01 -4.37790394e-01 -6.01181805e-01 6.35690629e-01
7.42935836e-01 -1.77471533e-01 -8.29120517e-01 -7.42546171e-02
7.71722674e-01 -1.12419558e+00 1.35438597e+00 -4.61746424e-01
8.35451901e-01 -3.48714828e-01 -3.91086251e-01 -1.39939308e+00
-6.70803249e-01 -1.85490966e-01 6.85392618e-02 9.03009772e-01
4.81876403e-01 8.83989558e-02 4.73994255e-01 7.37407207e-01
-1.05957188e-01 -1.32944465e-01 5.63395992e-02 -2.09500462e-01
-1.34200677e-01 -1.41435847e-01 9.18719530e-01 7.70027995e-01
-2.90013338e-03 4.62809086e-01 -6.38874829e-01 1.42022207e-01
6.69800580e-01 7.21410573e-01 6.38321638e-01 -7.70954907e-01
-6.91532195e-01 -4.26812410e-01 6.11054242e-01 -1.22889245e+00
-8.85204226e-02 -8.80334854e-01 5.12252748e-01 -1.90011895e+00
6.40055835e-01 -1.42002299e-01 -2.56818026e-01 6.14225149e-01
-3.33950967e-01 5.59033155e-01 5.53045094e-01 -5.16687222e-02
-7.72013605e-01 1.17504396e-01 1.31892526e+00 -3.73454869e-01
8.70097429e-02 -5.70952117e-01 -8.23771894e-01 6.83232844e-01
7.98779666e-01 -3.25822979e-01 -4.89432305e-01 -6.95787489e-01
4.02983785e-01 4.70955223e-02 5.40822506e-01 -7.00837135e-01
-7.94440433e-02 -1.23971686e-01 6.48796141e-01 -4.91886884e-01
3.14663023e-01 -9.31055427e-01 8.10755372e-01 4.27516043e-01
-4.96716678e-01 -6.95223957e-02 2.23546356e-01 5.32381177e-01
2.88481653e-01 -2.45444596e-01 4.60583299e-01 -6.00004494e-01
-6.88856125e-01 8.87541473e-02 -1.98104262e-01 -5.55312395e-01
5.41262627e-01 2.47817710e-01 -4.20244008e-01 -3.72327268e-01
-8.35286498e-01 -3.33533548e-02 9.79349971e-01 3.97389591e-01
3.88433188e-01 -1.57266045e+00 -8.03830624e-01 9.05680060e-02
1.93263650e-01 -3.75687838e-01 4.82082874e-01 3.83117348e-01
-7.82806218e-01 3.66281658e-01 -9.69780609e-02 -3.80525023e-01
-1.08068836e+00 1.35110903e+00 -5.53700551e-02 -3.61081034e-01
-9.25195873e-01 4.54328507e-01 5.89732826e-01 -5.82121372e-01
3.39018926e-02 -6.30789399e-01 -2.88309157e-01 -3.41129124e-01
1.05489099e+00 -9.11998078e-02 -2.89626867e-01 -5.78406930e-01
8.91810581e-02 7.39888489e-01 2.59207994e-01 1.45919845e-01
1.58705318e+00 -2.67913967e-01 1.34558544e-01 3.65257382e-01
8.49417746e-01 -1.67296216e-01 -1.72547710e+00 -9.08689573e-02
-3.01658273e-01 -3.32057327e-01 -8.26572999e-02 -1.36449790e+00
-1.04405928e+00 7.92779803e-01 5.79825819e-01 1.82108119e-01
1.31574249e+00 -5.15354089e-02 1.03437483e+00 2.74718970e-01
2.77909219e-01 -8.20771456e-01 -3.23931780e-03 -1.93859711e-01
7.84079671e-01 -1.48355150e+00 8.66467953e-02 -5.65981030e-01
-5.28000414e-01 1.23600161e+00 -1.55511862e-02 -3.96358997e-01
3.96382451e-01 1.74041674e-01 2.89550304e-01 -3.64192098e-01
-7.43962765e-01 -2.20371976e-01 3.88740987e-01 6.57599390e-01
5.33013105e-01 4.73218441e-01 1.08665731e-02 4.95515615e-01
-2.97706909e-02 3.60929102e-01 2.12392658e-01 4.54685152e-01
-5.23946404e-01 -1.35997915e+00 -2.83977509e-01 -1.18598580e-01
-1.79339260e-01 -4.51228470e-01 -2.41666928e-01 7.29601204e-01
3.32514524e-01 1.18473947e+00 -8.02311897e-02 -3.25735584e-02
3.00999850e-01 3.85878235e-02 1.10341716e+00 -5.70424676e-01
-8.03735971e-01 3.05674762e-01 3.15326244e-01 -8.10198367e-01
-9.59399939e-01 -5.92743218e-01 -1.07476711e+00 -5.69359601e-01
1.47786021e-01 -8.13690946e-02 7.89547801e-01 6.33386075e-01
1.68280706e-01 3.44192654e-01 2.66511470e-01 -1.05677426e+00
-2.78919488e-01 -7.63250947e-01 -4.93567050e-01 1.03549683e+00
2.14748889e-01 -1.78244829e-01 2.71972325e-02 9.78266895e-01] | [11.519472122192383, 4.47749137878418] |
23e4824a-b7b4-4eb9-8a16-b3efac88e5d2 | multi-task-learning-in-histo-pathology-for | 2005.08645 | null | https://arxiv.org/abs/2005.08645v1 | https://arxiv.org/pdf/2005.08645v1.pdf | Multi-Task Learning in Histo-pathology for Widely Generalizable Model | In this work we show preliminary results of deep multi-task learning in the area of computational pathology. We combine 11 tasks ranging from patch-wise oral cancer classification, one of the most prevalent cancers in the developing world, to multi-tissue nuclei instance segmentation and classification. | ['Navid Alemi Kooohbanani', 'Jevgenij Gamper', 'Nasir Rajpoot'] | 2020-05-09 | null | null | null | null | ['oral-cancer-classification'] | ['medical'] | [ 2.98272848e-01 2.89094836e-01 -3.59829009e-01 -2.77089793e-02
-1.73116410e+00 -7.26235732e-02 2.76490211e-01 5.99411666e-01
-6.78225100e-01 7.58695066e-01 1.65900096e-01 -3.43843669e-01
-1.43357918e-01 -3.30666929e-01 -3.47156465e-01 -1.02988684e+00
1.98068723e-01 8.74591172e-01 3.65491390e-01 -9.52135101e-02
4.31870036e-02 4.89880830e-01 -7.51900315e-01 8.07649732e-01
4.55927134e-01 7.64363766e-01 1.00739256e-01 9.63751376e-01
-4.93992306e-02 4.82154667e-01 -3.76292974e-01 -3.29345435e-01
-5.67475915e-01 8.83071050e-02 -1.22148061e+00 1.87489122e-01
4.86935854e-01 -5.66650108e-02 2.52159506e-01 9.21525955e-01
6.14166319e-01 -3.65566939e-01 7.44194865e-01 -6.59016609e-01
-3.53702195e-02 2.98098058e-01 -7.23403096e-01 3.30806792e-01
1.21693313e-02 -8.81793723e-02 9.77968514e-01 -3.65907878e-01
9.05387223e-01 9.24852073e-01 9.76521969e-01 6.07438505e-01
-9.10820305e-01 -2.22406909e-01 -4.47940648e-01 -4.60324176e-02
-1.07350564e+00 -3.36176395e-01 2.50114232e-01 -3.38641584e-01
9.33690667e-01 2.61062592e-01 4.35172617e-01 9.52121735e-01
7.67232239e-01 1.13117075e+00 1.12692201e+00 -2.27969751e-01
1.17715478e-01 -3.92373614e-02 1.47999495e-01 5.83569348e-01
2.03581676e-01 -6.27463222e-01 8.64382014e-02 -2.77221978e-01
4.79740530e-01 8.59115571e-02 2.00916842e-01 2.92417824e-01
-9.99298394e-01 7.99343288e-01 3.27484101e-01 8.50387692e-01
-3.81993726e-02 4.29753095e-01 6.81139112e-01 -1.02049656e-01
6.72137141e-01 -1.22805454e-01 -5.43361485e-01 3.76858681e-01
-8.28904688e-01 1.35973319e-01 3.61224800e-01 -1.20208733e-01
2.53526688e-01 -5.17674208e-01 1.82755128e-01 1.04917943e+00
7.54233420e-01 -5.19756973e-02 1.07263470e+00 -8.16103816e-01
-1.61905929e-01 6.51993215e-01 -4.62450296e-01 -9.16027948e-02
-1.16837943e+00 -5.07443845e-01 -7.93840289e-01 2.13526472e-01
8.23518634e-01 -6.86800554e-02 -1.16351831e+00 1.00318599e+00
3.66029501e-01 -1.34659111e-02 4.29343209e-02 3.61576647e-01
1.04509234e+00 2.46004239e-01 6.27733648e-01 2.83845067e-02
1.65812051e+00 -8.13401580e-01 -5.62030733e-01 -1.66816548e-01
1.29159379e+00 -5.77252507e-01 3.30462962e-01 1.28608152e-01
-7.62614250e-01 1.08521856e-01 -5.24152160e-01 -5.91690183e-01
-5.93290806e-01 1.87515333e-01 9.46835101e-01 3.83228719e-01
-1.22631621e+00 -3.54609303e-02 -9.96803701e-01 -6.50780022e-01
1.02627182e+00 5.62362254e-01 -6.83826506e-01 -2.34024838e-01
-5.55338442e-01 9.45712745e-01 1.22655869e-01 -2.72454292e-01
-6.96527600e-01 -7.54166722e-01 -5.19894123e-01 -2.39497423e-01
-2.22432707e-02 -6.13805473e-01 1.39461064e+00 -6.03171289e-01
-1.02322936e+00 1.57110405e+00 -1.62227392e-01 -1.91092402e-01
4.06542093e-01 3.10930818e-01 -1.61365852e-01 -1.95563436e-01
3.46520424e-01 8.57230067e-01 3.30858231e-02 -9.12408710e-01
-8.32945943e-01 -6.32619321e-01 -4.43653375e-01 5.10785319e-02
-1.54446706e-01 1.14180051e-01 -5.21132410e-01 -3.68826210e-01
-2.88823158e-01 -9.44594860e-01 -9.06988144e-01 3.99448723e-01
-6.43979251e-01 -5.23933470e-01 6.71497881e-01 -6.68110728e-01
3.46306086e-01 -2.29878211e+00 2.86995590e-01 1.45080850e-01
3.13247621e-01 -5.48210181e-02 -4.09954861e-02 -1.65843621e-01
4.13025431e-02 2.65581012e-01 1.33554507e-02 -7.56441474e-01
-3.41605186e-01 2.61368543e-01 5.94171226e-01 7.40985572e-01
-2.10210327e-02 9.80716348e-01 -6.68520093e-01 -9.78649318e-01
-7.61888996e-02 2.67124027e-01 -1.90461680e-01 -4.07152981e-01
-4.88052756e-01 3.35320681e-01 -3.74220848e-01 1.16315496e+00
1.68651938e-01 -5.18714547e-01 6.43505901e-02 -2.53584415e-01
1.78449035e-01 -1.10818319e-01 -4.17584628e-01 1.61601317e+00
-4.18000281e-01 6.04321539e-01 5.70463479e-01 -8.44692230e-01
-1.11634731e-01 6.01204574e-01 1.13547575e+00 -6.29032671e-01
2.73347199e-01 3.13995749e-01 1.65789604e-01 -3.05077821e-01
2.83388734e-01 -2.77933806e-01 -2.86045313e-01 3.50550264e-01
2.37269849e-01 -2.97103971e-01 2.38152415e-01 2.69968044e-02
1.08898556e+00 -3.01595181e-01 5.81936955e-01 -3.74311864e-01
4.39074337e-01 2.30184108e-01 5.39341509e-01 -5.88879501e-03
-5.61322570e-01 5.34550071e-01 6.75414503e-01 -4.69930798e-01
-6.94021046e-01 -8.13105226e-01 -5.29095054e-01 1.18476832e+00
-3.83411348e-01 6.76657036e-02 -4.72805828e-01 -5.53836226e-01
1.54357150e-01 -5.21298833e-02 -1.02046609e+00 4.57935601e-01
-6.35311007e-01 -1.65467525e+00 7.65100956e-01 2.68423855e-01
1.81602865e-01 -8.74701500e-01 -2.02099830e-01 2.22169653e-01
1.38068810e-01 -1.04851258e+00 -2.42825851e-01 5.33525348e-01
-8.81542385e-01 -1.20542681e+00 -1.18099940e+00 -1.28935540e+00
4.39156502e-01 -4.34397906e-01 8.69388580e-01 2.79284000e-01
-1.10624874e+00 2.08061319e-02 1.55660987e-01 -7.34941602e-01
-7.71081090e-01 3.04017931e-01 -6.75872982e-01 -2.88228810e-01
9.39208120e-02 -1.45007506e-01 -2.56461352e-01 -6.54477701e-02
-8.38761508e-01 1.01114236e-01 8.81262362e-01 7.48292267e-01
1.23656142e+00 -3.60298231e-02 4.11903292e-01 -1.15382612e+00
3.35260004e-01 -4.05800641e-01 -5.75866401e-02 5.52640557e-01
1.72773033e-01 -5.31794906e-01 1.00683160e-01 -5.73806167e-02
-6.70372784e-01 5.53107917e-01 -7.07285106e-01 5.96580803e-01
-3.78659576e-01 4.81136590e-01 2.60729700e-01 -3.45363826e-01
6.28877819e-01 5.58715463e-02 2.34664321e-01 -5.07549606e-02
1.21166594e-01 3.92085791e-01 3.83264869e-01 -9.07421857e-02
1.99713279e-02 7.67805219e-01 4.14352149e-01 -1.16083586e+00
-9.12087977e-01 -6.78854048e-01 -5.25777876e-01 -2.46349201e-01
1.37020504e+00 -7.44440377e-01 -4.47293550e-01 1.08684576e+00
-8.79135728e-01 -5.76303065e-01 -3.04829050e-02 -1.06298052e-01
-2.73419201e-01 -9.10778195e-02 -1.34875679e+00 -1.97581872e-01
-4.41792935e-01 -1.09386611e+00 1.66149247e+00 3.12608749e-01
-1.71056673e-01 -1.34191871e+00 4.69728738e-01 5.98900020e-01
2.96953827e-01 4.53355640e-01 1.13595927e+00 -6.44164860e-01
-1.41567960e-01 -1.74254686e-01 -1.92033187e-01 -1.88777477e-01
3.44527066e-01 3.72692347e-01 -1.04965520e+00 -1.88075334e-01
-5.06313264e-01 -5.30348957e-01 1.31186855e+00 1.00228000e+00
1.35712433e+00 4.63435382e-01 -1.30664217e+00 4.82521117e-01
1.43219697e+00 5.47134019e-02 3.82793367e-01 3.98894817e-01
7.55674839e-01 3.39758098e-01 4.54254672e-02 -1.74564674e-01
5.77133358e-01 3.77745599e-01 5.19495130e-01 -7.30850339e-01
-4.98242587e-01 6.58758759e-01 -3.89421970e-01 3.21896881e-01
1.98990270e-01 -2.66866893e-01 -1.18049121e+00 9.36033368e-01
-1.14004934e+00 -4.95254427e-01 -1.52902231e-01 1.07395577e+00
7.58360445e-01 1.40141353e-01 9.99784619e-02 2.31367007e-01
3.67355585e-01 -1.03196733e-01 -5.06420553e-01 -1.52092502e-01
-2.45773583e-03 2.83908159e-01 5.49403489e-01 3.67377818e-01
-1.39700305e+00 6.77645266e-01 7.11501980e+00 8.77065361e-01
-1.02952695e+00 6.53778374e-01 1.42223942e+00 1.33786365e-01
-2.44912561e-02 -8.09624851e-01 -8.39259446e-01 6.95424825e-02
8.54987919e-01 2.88109839e-01 -5.00814736e-01 5.42722344e-01
-3.74947267e-04 -2.89326489e-01 -7.97513783e-01 4.79053706e-01
-2.37012744e-01 -1.25121391e+00 -4.69055846e-02 4.36510026e-01
8.17034721e-01 5.38060546e-01 3.17588300e-01 9.98160690e-02
3.75817090e-01 -1.22558784e+00 9.21702012e-02 3.87687296e-01
6.18628323e-01 -4.66728926e-01 1.13928425e+00 9.83272716e-02
-8.02174687e-01 -1.70165762e-01 1.46022081e-01 7.03898013e-01
2.82999426e-01 7.63391495e-01 -9.45325375e-01 1.54703408e-01
5.93222082e-01 5.52545846e-01 -6.68811381e-01 1.35251045e+00
4.74511772e-01 3.45579088e-01 -4.68140930e-01 -1.47953406e-01
2.08482727e-01 4.93365139e-01 1.20442152e-01 1.36863232e+00
2.15380669e-01 9.92868748e-03 3.32540929e-01 1.21076748e-01
-3.87895197e-01 1.12935774e-01 3.23902331e-02 1.40984118e-01
-1.90302774e-01 1.51167274e+00 -1.22358632e+00 -1.66795462e-01
-1.02635190e-01 7.15586841e-01 9.27300826e-02 -1.68979599e-03
-4.69674587e-01 1.16696112e-01 5.22512436e-01 -1.97508171e-01
-6.40242621e-02 5.06556109e-02 -6.18492723e-01 -5.10572195e-01
-8.50924969e-01 -5.20150006e-01 6.94762111e-01 -2.05518365e-01
-1.29611647e+00 1.73551485e-01 -4.94965404e-01 -5.65167248e-01
-5.45865670e-02 -8.06986928e-01 -7.50236094e-01 4.28713113e-01
-1.62941635e+00 -1.65393126e+00 -3.95228155e-02 3.30975801e-01
6.89688683e-01 4.51748036e-02 1.17306590e+00 4.59591240e-01
-7.33082116e-01 7.24912584e-01 3.16060007e-01 9.40400735e-02
5.13467014e-01 -1.55627418e+00 2.56150991e-01 8.07310548e-03
-2.18534037e-01 -2.12152988e-01 1.90709457e-01 -4.57604736e-01
-8.96697342e-01 -1.09316850e+00 6.23551667e-01 -2.14189991e-01
6.78481102e-01 -1.35266036e-01 -6.49548829e-01 5.77343345e-01
-8.23114906e-03 4.44923580e-01 1.07943141e+00 -1.79584190e-01
2.88477182e-01 3.51231359e-02 -1.54616094e+00 1.05430461e-01
-5.81987202e-04 -5.64972579e-01 2.74592578e-01 9.99139428e-01
4.98688638e-01 -8.33993852e-01 -1.18878841e+00 2.30004415e-01
5.27596414e-01 -4.73966807e-01 9.09737885e-01 -1.94186673e-01
3.75321597e-01 1.27898559e-01 3.34751271e-02 -1.17575324e+00
-1.75133273e-01 2.84443319e-01 2.36197159e-01 7.78359294e-01
8.88448417e-01 -2.38789439e-01 1.20273209e+00 -5.33632711e-02
-5.86169243e-01 -1.02083838e+00 -1.38195097e+00 -2.50232574e-02
7.47495294e-01 -2.20485896e-01 -2.14432850e-01 5.05569816e-01
-2.65597820e-01 2.36577373e-02 4.53432977e-01 -2.34214403e-02
6.69265866e-01 -3.28846008e-01 -1.66764427e-02 -1.26710737e+00
-2.65341729e-01 -1.00990558e+00 -2.72259772e-01 -1.56121626e-01
1.77195042e-01 -8.56939316e-01 5.57946488e-02 -2.03974676e+00
4.90369648e-01 -2.48093694e-01 -4.01050359e-01 8.50896180e-01
-5.15411161e-02 6.26780093e-01 -3.86125356e-01 1.82025470e-02
-7.36494124e-01 -3.50392997e-01 1.37443054e+00 -6.58570230e-01
1.16410412e-01 1.83362246e-01 -6.10200584e-01 4.85601962e-01
7.65729368e-01 -4.38977152e-01 2.66440481e-01 -4.44688171e-01
-1.20489918e-01 1.68548953e-02 2.29870111e-01 -9.07343388e-01
1.71664611e-01 -1.04042664e-01 6.37340128e-01 -5.14918745e-01
4.48416710e-01 -5.46162069e-01 2.90548317e-02 8.36270154e-01
-2.26742610e-01 -3.43524128e-01 6.33349955e-01 3.64732295e-01
-2.78157920e-01 7.86819384e-02 9.43664432e-01 -5.27514935e-01
-5.07336497e-01 2.76894599e-01 -8.08868289e-01 -4.93705958e-01
1.19917107e+00 -4.14936543e-02 -7.74612427e-01 1.92778870e-01
-9.98959243e-01 4.78109032e-01 1.07280100e-02 -1.03860097e-02
1.68446079e-01 -7.50619829e-01 -8.76213729e-01 -4.01350349e-01
-1.29202884e-02 2.32313529e-01 4.41414654e-01 1.05275035e+00
-6.26673102e-01 6.52860463e-01 -1.69090375e-01 -7.68717825e-01
-1.48492265e+00 -8.81402008e-03 9.61034715e-01 -9.58099186e-01
-2.57115930e-01 1.22249353e+00 -1.68503877e-02 -5.23140669e-01
2.18411222e-01 -5.13883054e-01 -7.08102107e-01 2.15417936e-01
2.39451677e-01 1.36888236e-01 5.53590953e-01 -5.41734576e-01
-4.67696607e-01 5.85876346e-01 -6.32751048e-01 2.38421515e-01
1.25818193e+00 4.51699466e-01 -4.60817099e-01 2.86604702e-01
1.21769071e+00 -4.40343529e-01 -5.82169354e-01 1.94557577e-01
3.30441982e-01 6.26070976e-01 4.52090114e-01 -1.14150465e+00
-1.08299232e+00 5.65430284e-01 1.03249061e+00 2.58827582e-04
9.32297409e-01 4.71905500e-01 9.92581248e-01 1.46298453e-01
1.86677098e-01 -7.86969244e-01 -5.55641688e-02 2.05148146e-01
3.27331096e-01 -1.46631587e+00 9.96323302e-02 -5.72877347e-01
-9.89413634e-02 1.05438519e+00 5.45982003e-01 -5.95152332e-03
8.91421854e-01 5.54377198e-01 2.93996036e-01 -5.06527305e-01
-8.81313264e-01 -2.66001076e-01 7.88133815e-02 3.51618826e-01
6.79829478e-01 2.65035927e-01 -1.51816621e-01 2.90217012e-01
2.34935418e-01 4.33524624e-02 3.45214367e-01 7.69281209e-01
-5.24939299e-01 -1.12651277e+00 -1.68424964e-01 8.48022759e-01
-1.10100782e+00 2.40100771e-01 -5.81455886e-01 8.83130610e-01
2.61262298e-01 3.54680687e-01 8.15839171e-02 2.45533630e-01
-1.66985914e-01 -1.91538706e-01 4.37049389e-01 -8.30234647e-01
-7.80467808e-01 2.31063232e-01 3.35966051e-02 -1.81829944e-01
-5.39206326e-01 -8.56149554e-01 -1.58267188e+00 1.33915529e-01
-1.17280565e-01 -2.94556797e-01 1.07908225e+00 1.33744264e+00
-3.96318972e-01 1.05016792e+00 -7.07333684e-02 -7.46878684e-01
2.28556935e-02 -1.17188680e+00 -5.41243970e-01 -5.55781368e-03
5.19100130e-01 -1.90413728e-01 -1.24638148e-01 -1.13656439e-01] | [15.046255111694336, -3.000272035598755] |
44b8d0a5-0b12-44f0-b5dd-ae39815e0f27 | multi-scale-attention-for-audio-question | 2305.17993 | null | https://arxiv.org/abs/2305.17993v1 | https://arxiv.org/pdf/2305.17993v1.pdf | Multi-Scale Attention for Audio Question Answering | Audio question answering (AQA), acting as a widely used proxy task to explore scene understanding, has got more attention. The AQA is challenging for it requires comprehensive temporal reasoning from different scales' events of an audio scene. However, existing methods mostly extend the structures of visual question answering task to audio ones in a simple pattern but may not perform well when perceiving a fine-grained audio scene. To this end, we present a Multi-scale Window Attention Fusion Model (MWAFM) consisting of an asynchronous hybrid attention module and a multi-scale window attention module. The former is designed to aggregate unimodal and cross-modal temporal contexts, while the latter captures sound events of varying lengths and their temporal dependencies for a more comprehensive understanding. Extensive experiments are conducted to demonstrate that the proposed MWAFM can effectively explore temporal information to facilitate AQA in the fine-grained scene.Code: https://github.com/GeWu-Lab/MWAFM | ['Di Hu', 'Yixin Xu', 'Guangyao Li'] | 2023-05-29 | null | null | null | null | ['scene-understanding'] | ['computer-vision'] | [ 3.19054127e-02 -4.56551075e-01 2.65235782e-01 -4.59311217e-01
-1.16674173e+00 -5.18985391e-01 3.89783263e-01 3.65289897e-01
-1.97062165e-01 2.05086693e-01 6.15464330e-01 -2.26248220e-01
-3.51564944e-01 -6.12416506e-01 -4.98396993e-01 -4.95475531e-01
-1.39174476e-01 -1.44288805e-03 5.85143983e-01 -2.34793708e-01
2.04434365e-01 2.05413893e-01 -1.75887036e+00 7.99513400e-01
6.21342480e-01 1.12055397e+00 4.64317471e-01 9.47803319e-01
-2.48396367e-01 8.58455241e-01 -5.54153442e-01 -7.74390176e-02
-1.57508478e-01 -3.90286356e-01 -1.23585486e+00 -2.26667419e-01
3.79761130e-01 -2.10287392e-01 -4.59515184e-01 8.44987333e-01
6.73654258e-01 5.00529468e-01 -1.03623774e-02 -1.57498777e+00
-5.61111212e-01 5.83902657e-01 -4.04349029e-01 8.55394781e-01
7.09226489e-01 3.33883107e-01 1.15172935e+00 -8.59385729e-01
1.13419697e-01 1.68369043e+00 3.37209404e-01 1.21942237e-01
-8.77682328e-01 -6.78267896e-01 7.12424397e-01 9.87181246e-01
-1.27660179e+00 -2.57186025e-01 9.29485977e-01 -3.14217538e-01
8.24078560e-01 4.77515727e-01 7.67554641e-01 1.05954099e+00
8.39698687e-02 1.00496709e+00 1.09446752e+00 -9.60896760e-02
2.32972041e-01 -4.40260023e-01 4.92005616e-01 3.13473463e-01
-4.24665153e-01 -3.33077639e-01 -9.22524869e-01 -1.72950059e-01
5.72946012e-01 5.09302318e-02 -3.26809108e-01 1.56868592e-01
-1.37122428e+00 5.37227809e-01 5.48018336e-01 3.95313531e-01
-4.49161053e-01 2.79215425e-01 5.34784138e-01 3.20494473e-01
3.42632025e-01 1.85752958e-01 -2.45370880e-01 -3.18158507e-01
-5.50807416e-01 2.97119141e-01 3.08020264e-01 5.85582316e-01
6.41887009e-01 2.86267269e-02 -5.87442935e-01 7.60443866e-01
3.30144554e-01 1.78893700e-01 5.24836361e-01 -1.08708978e+00
3.52845937e-01 4.61816400e-01 -1.89883895e-02 -8.85852396e-01
-4.19003695e-01 -1.43115625e-01 -6.13543153e-01 9.73110199e-02
3.52721483e-01 8.27773884e-02 -8.17487538e-01 1.89126623e+00
5.76213956e-01 4.66026217e-01 -2.76843905e-01 1.12409294e+00
1.00242603e+00 1.04861224e+00 4.27786976e-01 -1.98246658e-01
1.86584210e+00 -1.12426269e+00 -8.85845900e-01 -2.59794146e-01
1.96349733e-02 -7.43540943e-01 1.44428682e+00 3.04389834e-01
-1.25387108e+00 -1.11417198e+00 -8.95144045e-01 -4.61753011e-01
-5.26602149e-01 -2.32333854e-01 4.36528713e-01 -6.54940307e-02
-9.31907713e-01 1.15448788e-01 -6.31443620e-01 -3.97488803e-01
3.92856717e-01 -8.47636312e-02 -2.00382471e-01 -2.30267551e-02
-1.50799656e+00 3.92593563e-01 3.59424114e-01 4.12103027e-01
-1.13992774e+00 -7.53939331e-01 -7.46469557e-01 2.61539608e-01
6.35815084e-01 -7.32811809e-01 1.62171435e+00 -6.33756459e-01
-1.25743341e+00 4.79753047e-01 -5.93990386e-01 -4.06164289e-01
6.94953743e-03 -6.63599432e-01 -4.65105653e-01 5.16643107e-01
1.71294063e-01 6.63914621e-01 1.02580810e+00 -1.07233500e+00
-6.34819627e-01 -3.92914712e-01 5.44509232e-01 4.48787719e-01
-1.05451278e-01 1.56923369e-01 -5.19046009e-01 -9.06415045e-01
-2.73230765e-02 -3.79814982e-01 -2.96606682e-02 1.95297766e-02
-1.00348726e-01 -4.36883599e-01 1.17128122e+00 -6.61178291e-01
1.66304457e+00 -2.32562232e+00 6.77127913e-02 -2.76981801e-01
1.48549989e-01 1.09175406e-01 -2.97208518e-01 7.47121692e-01
-4.15563643e-01 -5.17854914e-02 -1.54146910e-01 -2.98585951e-01
8.68782699e-02 3.70731726e-02 -7.56213367e-01 -7.35420063e-02
2.61011124e-01 9.62020516e-01 -1.07415986e+00 -5.94262481e-01
1.84015259e-01 3.33753675e-01 -3.70407552e-01 4.35160786e-01
-3.41081530e-01 4.89857286e-01 -5.04778683e-01 8.11697185e-01
4.31304485e-01 -3.27147841e-01 -3.38491648e-01 -3.70826960e-01
-9.92444456e-02 1.40237615e-01 -1.25324154e+00 2.09050417e+00
-3.57431352e-01 6.80885851e-01 2.50383407e-01 -8.03912401e-01
6.79832160e-01 4.71044481e-01 3.10892075e-01 -8.64452124e-01
1.28950365e-02 -3.16785336e-01 6.72590360e-03 -9.35141444e-01
5.16396701e-01 -1.94916055e-01 -1.54480800e-01 4.77487773e-01
1.40712142e-01 -7.40739405e-02 1.15233213e-01 3.73622090e-01
1.12228489e+00 -1.67303085e-02 3.24577183e-01 -1.05306797e-01
7.14053452e-01 -2.81382263e-01 5.44780374e-01 7.34213829e-01
-6.62356973e-01 6.55340791e-01 3.06502610e-01 -5.37481010e-01
-6.45353198e-01 -1.23890531e+00 1.69149414e-01 1.66140699e+00
4.20363933e-01 -6.27280891e-01 -5.48418581e-01 -4.53026056e-01
-1.43586069e-01 4.96700257e-01 -7.52476633e-01 -1.50667965e-01
-4.57023531e-01 -2.75739521e-01 4.70348537e-01 6.68027341e-01
6.55167699e-01 -1.34404433e+00 -8.61205459e-01 1.66729212e-01
-7.44867086e-01 -9.22408760e-01 -7.26102293e-01 -1.24108478e-01
-3.84674966e-01 -1.03705215e+00 -5.53933263e-01 -6.02804124e-01
4.49622981e-02 5.48709273e-01 9.93353486e-01 -2.01652303e-01
-5.45184851e-01 7.76160896e-01 -6.10793173e-01 -6.94040656e-01
2.49406964e-01 -2.76500255e-01 -1.10211194e-01 5.33663452e-01
1.27216816e-01 -7.77303874e-01 -8.46292198e-01 3.75465631e-01
-1.23523557e+00 9.34637934e-02 3.09571028e-01 6.23034120e-01
6.04189932e-01 -2.88465749e-02 9.29324687e-01 -4.65772510e-01
9.28799450e-01 -6.51513100e-01 -2.06224993e-01 3.03207725e-01
1.03394412e-01 -1.53946489e-01 4.53006864e-01 -7.18116701e-01
-1.31617928e+00 -3.53472799e-01 -3.53215575e-01 -6.07538044e-01
-4.66175050e-01 4.71921414e-01 -2.95906872e-01 3.95194054e-01
3.66582751e-01 9.19616446e-02 -4.07666475e-01 -4.55082417e-01
4.22102004e-01 3.95692796e-01 7.74459183e-01 -6.21049285e-01
5.98281026e-01 6.45514607e-01 -3.56219023e-01 -6.40341520e-01
-8.79955173e-01 -7.82293320e-01 -4.29135919e-01 -6.31511867e-01
1.12988770e+00 -8.86108041e-01 -1.02340853e+00 3.98306698e-01
-1.18797493e+00 -4.26681817e-01 -4.02165622e-01 4.77034688e-01
-5.39104283e-01 3.82428110e-01 -6.02116883e-01 -1.03514493e+00
-1.69599026e-01 -1.01914692e+00 1.37381923e+00 4.32508111e-01
-3.06972831e-01 -6.68476880e-01 1.90001011e-01 5.32803535e-01
4.27345991e-01 1.78809285e-01 8.74750733e-01 -3.40098083e-01
-6.65899575e-01 1.76454827e-01 -3.23601305e-01 -2.19094500e-01
-5.43536525e-03 -2.57776201e-01 -1.38187683e+00 -1.42464325e-01
1.57037809e-01 -5.20526528e-01 9.57193911e-01 4.10380840e-01
1.66128898e+00 -9.92168263e-02 1.17767088e-01 2.06647635e-01
1.03163755e+00 5.17756045e-01 4.98021424e-01 2.34662309e-01
4.39727634e-01 5.60095251e-01 9.03977036e-01 6.48577154e-01
5.90547562e-01 5.79831123e-01 7.82899678e-01 -2.19321530e-02
-8.56144875e-02 -1.99366137e-01 3.70351434e-01 9.22875226e-01
1.07391469e-01 -1.85779080e-01 -9.01629329e-01 7.80975163e-01
-2.04445720e+00 -1.26169538e+00 7.72364587e-02 1.74875069e+00
5.84320664e-01 1.52772432e-02 2.25845560e-01 4.51960951e-01
7.21731484e-01 6.38500988e-01 -5.83403945e-01 -4.16283637e-01
9.65023134e-03 1.17463067e-01 -5.98867774e-01 4.93130445e-01
-1.09922469e+00 5.81246018e-01 5.74140930e+00 1.07345116e+00
-8.63048911e-01 3.01527143e-01 3.13328594e-01 -1.81172594e-01
-3.98451388e-01 7.47013986e-02 -2.53449202e-01 3.88945699e-01
8.61913383e-01 -2.08766520e-01 2.84883261e-01 4.26303297e-01
3.83613348e-01 -2.31998101e-01 -9.24377441e-01 1.12445581e+00
-1.09107412e-01 -1.03220046e+00 1.62863180e-01 -5.17859101e-01
1.27130285e-01 -2.99754918e-01 1.30610168e-01 4.52716112e-01
-3.70850153e-02 -8.59780490e-01 9.34117258e-01 9.15571213e-01
4.24116313e-01 -5.84372401e-01 3.59443426e-01 2.95971096e-01
-1.95923400e+00 -4.44642901e-01 6.38802424e-02 -3.04072797e-01
5.82080483e-01 2.44378537e-01 -3.42228949e-01 7.86966383e-01
1.33438528e+00 4.90999818e-01 -7.01159060e-01 1.16092563e+00
3.83059122e-02 6.76443696e-01 -2.01077193e-01 1.33705780e-01
2.87922978e-01 1.88328311e-01 7.74496615e-01 1.21884823e+00
2.85704464e-01 4.95416850e-01 3.84949669e-02 5.73552728e-01
1.96730956e-01 -6.87345071e-03 -3.39626729e-01 4.17510308e-02
4.47725773e-01 1.24098659e+00 -6.40773177e-01 -3.50359261e-01
-4.83143121e-01 8.36042404e-01 2.32262418e-01 6.09375060e-01
-1.08304679e+00 -4.56478357e-01 7.40780234e-01 -6.67702928e-02
2.28189394e-01 -1.46212593e-01 7.00316653e-02 -1.05274844e+00
1.16223782e-01 -8.68976712e-01 1.03240383e+00 -1.64848650e+00
-1.22902596e+00 6.90007567e-01 1.89024478e-01 -1.33266950e+00
1.57668456e-01 -2.36252397e-01 -9.61209655e-01 7.74894357e-01
-1.46735764e+00 -1.10672235e+00 -7.07742453e-01 9.97568429e-01
1.16382277e+00 3.55883390e-01 6.31168604e-01 2.78950274e-01
-5.01603305e-01 3.07102740e-01 -5.60193956e-01 -1.56148717e-01
8.02574515e-01 -1.20397162e+00 1.13021471e-01 7.97431886e-01
3.00788581e-01 4.65400934e-01 6.83929741e-01 -3.32195103e-01
-1.36126924e+00 -9.21337903e-01 5.44943631e-01 -3.43877435e-01
8.54298055e-01 -1.68052480e-01 -1.40687287e+00 6.39768302e-01
6.38906181e-01 -2.36137323e-02 6.62632227e-01 1.05022900e-01
-6.45738482e-01 -3.49971354e-01 -6.18317306e-01 5.28382123e-01
8.71977389e-01 -1.09171653e+00 -9.28177774e-01 -4.11385112e-02
1.07134938e+00 -2.45641649e-01 -9.28694308e-01 6.11715078e-01
5.75148344e-01 -1.01481843e+00 1.05294192e+00 -5.81989169e-01
2.54406810e-01 -6.97387874e-01 -2.43416846e-01 -9.16371465e-01
-4.08436984e-01 -6.67728245e-01 -2.25623608e-01 1.28417909e+00
-7.00411424e-02 -4.29020256e-01 1.10180877e-01 1.23641081e-01
-2.87315696e-01 -7.72631526e-01 -1.06418967e+00 -4.25190449e-01
-4.44357932e-01 -9.27051485e-01 6.69832289e-01 8.30906034e-01
1.38684928e-01 5.63552022e-01 -3.39510769e-01 5.76547801e-01
2.55240530e-01 5.16176283e-01 5.18131733e-01 -1.07348132e+00
-3.36416066e-01 -6.54142797e-01 -2.97223032e-01 -1.02737844e+00
-9.44935009e-02 -4.58250523e-01 1.22852959e-01 -1.55330682e+00
2.65150338e-01 2.24679053e-01 -6.07500374e-01 4.16997313e-01
-3.99528503e-01 1.62478685e-01 3.08230698e-01 5.83245084e-02
-1.07911611e+00 7.18364477e-01 1.24353755e+00 -1.69501886e-01
-4.50212285e-02 -9.37914550e-02 -6.29318833e-01 7.94768691e-01
7.34911323e-01 -4.67043445e-02 -6.72639012e-01 -5.61171591e-01
1.13199353e-01 4.59451437e-01 6.89017415e-01 -1.00093198e+00
4.35470104e-01 -1.66658044e-01 2.19137743e-01 -1.03626335e+00
6.73741341e-01 -6.00063205e-01 1.44966230e-01 1.52377829e-01
-5.68828523e-01 3.41494948e-01 5.86881340e-01 8.81651044e-01
-7.56038070e-01 2.34739184e-01 5.43834209e-01 -2.61896253e-02
-1.06817806e+00 5.45980453e-01 -4.72830683e-01 8.59584436e-02
8.78313899e-01 3.53899486e-02 -3.06551725e-01 -6.86129272e-01
-1.10087252e+00 6.20025396e-01 -2.26156607e-01 7.47952878e-01
8.32296073e-01 -1.61245608e+00 -5.13110638e-01 -7.25022182e-02
3.44104230e-01 6.14723824e-02 1.08398151e+00 8.00667107e-01
-1.16858140e-01 3.72966856e-01 -1.55927882e-01 -6.94826663e-01
-1.35800397e+00 7.92009234e-01 9.90502611e-02 -1.79538459e-01
-4.80933815e-01 9.90099490e-01 7.17099428e-01 1.60450395e-03
4.96419817e-01 -5.94191134e-01 -3.84000808e-01 3.10832292e-01
8.01553547e-01 3.09828997e-01 -2.07506016e-01 -5.46959281e-01
-2.82230854e-01 7.04033256e-01 2.47725010e-01 -2.46457979e-01
9.72458541e-01 -6.25169456e-01 9.92703512e-02 1.06572759e+00
9.22072291e-01 -2.34358564e-01 -1.30192375e+00 -5.00142634e-01
-2.85324246e-01 -3.63588393e-01 -1.53910935e-01 -5.68737507e-01
-6.09269142e-01 1.38111436e+00 5.12197673e-01 7.46932089e-01
1.74002564e+00 3.50977331e-01 7.75711238e-01 5.52375987e-02
2.17416901e-02 -8.68548274e-01 6.16503060e-01 5.27456224e-01
1.38282204e+00 -1.05034292e+00 -3.45876873e-01 -1.25170678e-01
-6.68382525e-01 1.03957355e+00 7.68753052e-01 1.93888903e-01
7.75118530e-01 1.10605555e-02 1.07659444e-01 -2.60247111e-01
-1.13623834e+00 -6.00457311e-01 4.63514447e-01 3.02403092e-01
1.05787918e-01 -1.50262520e-01 1.47693262e-01 9.52353477e-01
3.86573784e-02 -2.11025566e-01 4.50929403e-02 9.04051840e-01
-4.31717455e-01 -6.37087047e-01 -5.21289170e-01 2.51719840e-02
-3.70707214e-01 2.91326921e-02 -1.11923084e-01 6.37273967e-01
1.66454926e-01 1.00538456e+00 1.10822387e-01 -4.61645871e-01
4.54402834e-01 3.03814471e-01 1.05115779e-01 -3.21902007e-01
-5.44374824e-01 2.67947972e-01 -2.03982815e-01 -9.71245050e-01
-6.14532471e-01 -6.32329404e-01 -1.18488371e+00 7.84490258e-02
5.96578717e-02 1.90689012e-01 9.11790282e-02 9.62601244e-01
3.52002323e-01 9.18304980e-01 4.66576368e-01 -7.80319512e-01
7.65172690e-02 -9.68621612e-01 -4.33362007e-01 4.05307472e-01
6.62836194e-01 -6.39016569e-01 -1.45797491e-01 1.34004086e-01] | [10.524393081665039, 1.1094987392425537] |
7a38c970-c7dc-468e-8158-3e5d50dd9fd9 | graphreg-dynamical-point-cloud-registration | 2302.01109 | null | https://arxiv.org/abs/2302.01109v1 | https://arxiv.org/pdf/2302.01109v1.pdf | GraphReg: Dynamical Point Cloud Registration with Geometry-aware Graph Signal Processing | This study presents a high-accuracy, efficient, and physically induced method for 3D point cloud registration, which is the core of many important 3D vision problems. In contrast to existing physics-based methods that merely consider spatial point information and ignore surface geometry, we explore geometry aware rigid-body dynamics to regulate the particle (point) motion, which results in more precise and robust registration. Our proposed method consists of four major modules. First, we leverage the graph signal processing (GSP) framework to define a new signature, (i.e., point response intensity for each point), by which we succeed in describing the local surface variation, resampling keypoints, and distinguishing different particles. Then, to address the shortcomings of current physics-based approaches that are sensitive to outliers, we accommodate the defined point response intensity to median absolute deviation (MAD) in robust statistics and adopt the X84 principle for adaptive outlier depression, ensuring a robust and stable registration. Subsequently, we propose a novel geometric invariant under rigid transformations to incorporate higher-order features of point clouds, which is further embedded for force modeling to guide the correspondence between pairwise scans credibly. Finally, we introduce an adaptive simulated annealing (ASA) method to search for the global optimum and substantially accelerate the registration process. We perform comprehensive experiments to evaluate the proposed method on various datasets captured from range scanners to LiDAR. Results demonstrate that our proposed method outperforms representative state-of-the-art approaches in terms of accuracy and is more suitable for registering large-scale point clouds. Furthermore, it is considerably faster and more robust than most competitors. | ['Huang Tiejun', 'Yan Dong-Ming', 'Jia Xiaohong', 'Ma Lei', 'Zhao Mingyang'] | 2023-02-02 | null | null | null | null | ['point-cloud-registration'] | ['computer-vision'] | [ 2.42988259e-01 -6.04964733e-01 1.37099147e-01 -2.73914188e-01
-6.87970877e-01 -3.72005492e-01 6.67493045e-01 2.73852021e-01
-2.24420875e-01 2.44924039e-01 -2.58785993e-01 5.50593920e-02
-4.71254498e-01 -9.03455436e-01 -7.34144449e-01 -7.96710610e-01
-3.64837833e-02 7.03801870e-01 4.60188866e-01 -1.70768425e-01
5.51599860e-01 1.19154096e+00 -1.52806568e+00 -7.24729061e-01
1.17563438e+00 8.56771529e-01 7.18180602e-03 4.50397842e-02
-1.18582286e-01 -2.42682397e-01 -2.77546048e-01 5.42182215e-02
4.60749686e-01 -8.19191486e-02 -2.95181930e-01 1.00722648e-01
6.00404322e-01 1.29780397e-01 -6.89133853e-02 1.26140893e+00
6.23541594e-01 5.17064095e-01 7.72589743e-01 -9.94309664e-01
-6.43103898e-01 -1.94297224e-01 -1.02099586e+00 4.56127152e-02
4.43001807e-01 1.77985176e-01 6.69065654e-01 -1.13139033e+00
5.29323757e-01 1.36799753e+00 8.09007168e-01 2.23928407e-01
-1.33153319e+00 -7.28160501e-01 5.08246273e-02 -1.93233624e-01
-1.66388631e+00 -2.64455438e-01 1.21646011e+00 -4.75520581e-01
5.05392015e-01 4.37093228e-01 7.19091773e-01 5.39130151e-01
2.54409015e-01 -5.84144481e-02 1.08043611e+00 -2.52070040e-01
2.96858042e-01 -5.58112204e-01 1.80113241e-01 5.19125521e-01
5.93548119e-01 2.76880115e-01 -4.15688664e-01 -6.06782258e-01
8.73665869e-01 2.83508688e-01 -3.55270863e-01 -7.90440738e-01
-1.36373091e+00 6.46115243e-01 5.57141483e-01 1.57935247e-02
-5.54451585e-01 2.96618883e-02 1.18347686e-02 -1.99473098e-01
5.24091303e-01 1.13971755e-01 -1.05161346e-01 1.68475747e-01
-8.21970403e-01 4.32530999e-01 4.97690529e-01 7.77010679e-01
9.62912023e-01 -1.39145982e-02 -4.09385785e-02 6.92990422e-01
8.19530666e-01 1.02487540e+00 1.95895746e-01 -9.40535247e-01
3.09962273e-01 7.69620538e-01 -1.35995178e-02 -1.40255499e+00
-3.77986014e-01 -2.59793848e-01 -9.72670138e-01 3.65182370e-01
-6.55825669e-03 4.34281945e-01 -1.04994488e+00 1.40598798e+00
8.03858340e-01 7.90636063e-01 -1.47987366e-01 1.01638830e+00
7.59938002e-01 4.56582397e-01 -1.43748835e-01 -4.46969688e-01
1.21014428e+00 -2.23156661e-01 -4.55248922e-01 1.00948207e-01
1.57925878e-02 -8.38827252e-01 8.92936647e-01 -2.21126787e-02
-1.04817688e+00 -6.17501318e-01 -1.04856038e+00 1.73920363e-01
-3.74216177e-02 -4.52019691e-01 3.90711814e-01 3.37843895e-01
-7.67370164e-01 7.80506849e-01 -1.12287712e+00 -3.84250075e-01
1.93995565e-01 5.27297556e-01 -2.13862345e-01 2.22084403e-01
-8.11715305e-01 6.38525784e-01 6.81516975e-02 2.70744503e-01
-2.27143794e-01 -8.72160196e-01 -8.23977053e-01 -3.96897256e-01
1.68572605e-01 -9.16205943e-01 6.04761660e-01 -2.04188883e-01
-1.68446219e+00 8.87075543e-01 -4.23004538e-01 -1.38411745e-01
3.54978919e-01 3.58927436e-02 -2.99761117e-01 1.58665851e-02
1.31574035e-01 3.28429878e-01 8.29567313e-01 -1.60352254e+00
-1.30252257e-01 -6.47978067e-01 -6.11315370e-01 3.55401903e-01
1.88483611e-01 -3.98580246e-02 -6.76247299e-01 -6.10646427e-01
9.37440574e-01 -1.15652049e+00 -4.26128119e-01 -3.36415172e-02
-3.24492276e-01 -2.04365745e-01 8.07992518e-01 -1.69566199e-01
8.08543205e-01 -2.33021879e+00 1.18344821e-01 8.92452657e-01
2.84194857e-01 7.66262263e-02 1.04841933e-01 2.54721969e-01
6.75093308e-02 -3.39262076e-02 -5.80380380e-01 -4.16103333e-01
-3.25912312e-02 3.06804836e-01 -3.22009265e-01 9.46174979e-01
1.16953798e-01 7.86286712e-01 -7.41278052e-01 -4.46565986e-01
5.64023137e-01 6.15166426e-01 -4.01202589e-01 -1.00368090e-01
1.16678350e-01 8.95434260e-01 -7.39088297e-01 8.26318145e-01
1.22180367e+00 8.15113354e-03 -4.10907358e-01 -3.49742472e-01
-3.34795356e-01 -3.31683755e-02 -1.63885081e+00 1.64040792e+00
3.75662595e-02 -1.61537379e-01 1.11246563e-01 -6.86414719e-01
1.44148004e+00 -7.42537901e-02 9.51515913e-01 -3.67472023e-01
1.19234174e-01 3.97694796e-01 -2.26984605e-01 -4.29147563e-04
4.69482899e-01 -2.53053643e-02 3.12696397e-02 9.35722440e-02
-5.41678309e-01 -7.15862513e-01 -3.37844014e-01 -1.40130430e-01
9.42261696e-01 1.26940593e-01 2.38713428e-01 -3.07137787e-01
8.41351151e-01 3.42871435e-02 6.70189798e-01 5.01595080e-01
-2.10595563e-01 7.63919771e-01 -3.78260821e-01 -2.09272653e-01
-7.41814733e-01 -1.34587204e+00 -4.86026704e-01 1.52327299e-01
8.37237239e-01 -1.29888088e-01 -4.00504678e-01 -1.25896633e-01
5.09162128e-01 3.69350225e-01 -1.87480614e-01 -1.09850273e-01
-7.73289323e-01 -8.11990798e-01 1.37253493e-01 1.60448864e-01
5.34961820e-01 -7.25145698e-01 -1.94046319e-01 1.52508140e-01
1.13938786e-01 -1.04688835e+00 -5.27151823e-01 -3.58332187e-01
-1.37776291e+00 -9.67244685e-01 -3.34046721e-01 -4.36801016e-01
8.14091980e-01 5.97414374e-01 8.64098012e-01 2.48130932e-01
-6.40576631e-02 5.45677125e-01 -1.68237865e-01 -3.36488813e-01
-1.46432102e-01 -2.06668273e-01 4.80073512e-01 9.48278904e-02
3.88131320e-01 -8.45480025e-01 -6.48006797e-01 4.88730013e-01
-6.86716676e-01 -4.79570955e-01 5.03668785e-01 4.66866404e-01
1.37547529e+00 -4.87975329e-02 1.06972069e-01 -5.17401755e-01
5.81283510e-01 -2.24774644e-01 -8.03472042e-01 -1.47946402e-01
-6.78979099e-01 -1.05297051e-01 2.76486754e-01 -2.84134239e-01
-7.29420424e-01 3.78144383e-01 4.15780861e-03 -6.79825842e-01
-2.46370524e-01 3.12858999e-01 -1.21594757e-01 -9.07214761e-01
5.25698364e-01 2.65132457e-01 1.23666719e-01 -4.66626167e-01
3.49932939e-01 4.07306194e-01 7.18052566e-01 -8.76306713e-01
1.50623035e+00 9.63905871e-01 5.85635424e-01 -9.82337058e-01
-2.56557316e-01 -8.26720238e-01 -8.98602962e-01 -1.67413890e-01
7.51767933e-01 -7.84889877e-01 -9.04043138e-01 4.92680669e-01
-1.17050242e+00 3.48480225e-01 -3.66224349e-01 6.35518670e-01
-5.69711328e-01 8.01086962e-01 -1.95547998e-01 -7.34511256e-01
-3.77288431e-01 -1.13120270e+00 1.47290456e+00 2.03191996e-01
2.80888248e-02 -8.64486158e-01 4.63532895e-01 2.27297276e-01
2.07983240e-01 6.40996397e-01 6.03030324e-01 -3.99777502e-01
-9.36262608e-01 -1.72795132e-01 -4.75941710e-02 -3.83878648e-02
2.16433346e-01 2.92255431e-01 -6.10550702e-01 -3.79043519e-01
3.08716625e-01 4.18069422e-01 4.08736110e-01 4.27209377e-01
8.55119705e-01 2.34250471e-01 -5.54584503e-01 9.06481206e-01
1.51525056e+00 3.99424024e-02 6.23655140e-01 4.31962848e-01
8.38538229e-01 4.15835053e-01 7.89755106e-01 2.87546754e-01
4.44936156e-01 8.64945054e-01 6.05952740e-01 -1.55297965e-01
1.22884005e-01 -1.27172247e-01 2.01414511e-01 1.26176608e+00
-5.06919026e-01 2.68703371e-01 -9.02757823e-01 2.70360142e-01
-1.75041330e+00 -7.69023538e-01 -5.70899367e-01 2.42689419e+00
5.37934124e-01 -5.73676825e-02 -2.16918334e-01 -3.68646868e-02
9.84156191e-01 1.72840789e-01 -6.17761016e-01 -3.42268567e-03
5.29076345e-02 4.72350329e-01 7.62114942e-01 4.86496389e-01
-9.89529014e-01 9.06635463e-01 5.90142155e+00 6.04848325e-01
-1.06469810e+00 -6.38291463e-02 -2.14195952e-01 2.83883303e-01
-2.91350245e-01 1.61411166e-01 -7.13686824e-01 4.68339711e-01
5.49933851e-01 -2.88129777e-01 3.29863936e-01 6.13694429e-01
5.83240449e-01 1.34961382e-01 -6.60364926e-01 1.06508636e+00
4.74731550e-02 -1.09069264e+00 2.10287765e-01 2.61831373e-01
6.57858074e-01 2.46743679e-01 -8.70116204e-02 -3.23855013e-01
1.37486234e-01 -6.71974719e-01 5.23829997e-01 7.99128771e-01
4.80751425e-01 -5.83195508e-01 4.10848349e-01 3.14473301e-01
-1.56338322e+00 4.76933509e-01 -5.82421482e-01 1.12321660e-01
3.97701800e-01 7.97890544e-01 -3.29530746e-01 1.07508981e+00
7.41634607e-01 8.07318211e-01 -3.41213614e-01 1.23238814e+00
1.58862548e-03 2.03916773e-01 -6.74430728e-01 1.54936731e-01
-1.09398380e-01 -9.03045774e-01 1.10142028e+00 6.20937705e-01
5.99337816e-01 3.98828506e-01 5.04475713e-01 9.25780416e-01
1.19947560e-01 3.28178972e-01 -6.51286364e-01 6.21186197e-01
8.71396303e-01 1.14093471e+00 -7.91751683e-01 -2.04360466e-02
-2.75678068e-01 6.76223040e-01 -8.53195637e-02 3.26160371e-01
-7.29749203e-01 -1.77609324e-01 7.72452295e-01 2.39518896e-01
8.26716572e-02 -7.47504771e-01 -3.81657243e-01 -1.05479825e+00
2.42823422e-01 -5.11525989e-01 4.81078587e-02 -6.55717969e-01
-1.50089490e+00 3.24170709e-01 1.77594647e-01 -1.58737600e+00
1.22360632e-01 -4.11884248e-01 -7.80894995e-01 9.47508335e-01
-1.70344579e+00 -1.15584028e+00 -6.48868680e-01 7.91046023e-01
1.04017101e-01 2.31573984e-01 5.12069583e-01 2.96231300e-01
-2.69747913e-01 6.44236729e-02 1.70895904e-01 -1.99801072e-01
6.96623266e-01 -9.39060450e-01 6.31660283e-01 9.23827291e-01
1.35664225e-01 1.00563264e+00 5.96361816e-01 -1.03586435e+00
-1.70255482e+00 -1.06416917e+00 3.75758559e-01 -5.42209387e-01
6.87091887e-01 -1.36501119e-01 -1.15646219e+00 4.57543284e-01
-5.21224201e-01 2.51177430e-01 2.76925385e-01 -2.69348204e-01
5.08669205e-03 -1.57677412e-01 -1.23720837e+00 3.95019770e-01
1.26009333e+00 -2.77970195e-01 -9.10583317e-01 2.11717010e-01
6.17225111e-01 -8.10548007e-01 -1.01916730e+00 9.18660045e-01
1.59713060e-01 -7.22855330e-01 1.32517064e+00 4.95304652e-02
-3.83249789e-01 -9.26992178e-01 -1.07823536e-01 -1.01429522e+00
-4.64516371e-01 -7.00754941e-01 -4.42131087e-02 1.29889870e+00
-1.54455021e-01 -1.05805171e+00 7.66721308e-01 4.66716588e-01
-3.94956768e-01 -4.90872324e-01 -1.28097260e+00 -9.72570062e-01
-1.35032043e-01 -4.66934770e-01 8.58313560e-01 1.08364093e+00
-6.37557864e-01 -1.22733399e-01 2.51590665e-02 8.85543227e-01
1.12706995e+00 3.75014812e-01 1.06200612e+00 -1.86029673e+00
1.71953872e-01 -3.13496739e-01 -8.10845554e-01 -1.05263281e+00
1.71676487e-01 -9.51367736e-01 1.92122117e-01 -1.34459829e+00
-5.47486804e-02 -8.04704010e-01 -1.39998898e-01 -4.57228795e-02
-1.90004885e-01 3.47157061e-01 9.62498598e-03 8.73750865e-01
-1.48631692e-01 7.43662655e-01 1.20914114e+00 9.39232707e-02
-5.28130352e-01 1.91821814e-01 -3.63426715e-01 8.49071205e-01
4.72031057e-01 -4.50399220e-01 -1.38888694e-02 -2.27377370e-01
-1.61423862e-01 -3.17378551e-01 6.22790515e-01 -1.10912490e+00
4.43293273e-01 -4.00640666e-01 1.06419370e-01 -9.44656134e-01
3.12411159e-01 -1.16558385e+00 5.52328646e-01 2.96096981e-01
3.77907187e-01 2.38172814e-01 1.01788960e-01 8.14675510e-01
-1.63094699e-01 -5.27841412e-02 8.75902951e-01 1.99745432e-01
-5.24955750e-01 7.75102079e-01 3.13059092e-01 -2.39898115e-01
1.15268981e+00 -5.56668818e-01 -9.53733400e-02 7.50784278e-02
-3.74151617e-01 2.23570749e-01 1.00807190e+00 2.26389483e-01
8.27701151e-01 -1.56372690e+00 -6.17406964e-01 4.65679079e-01
2.25432545e-01 4.66981024e-01 8.17484483e-02 8.70270908e-01
-6.62650049e-01 -5.27303144e-02 1.59212872e-01 -1.21636248e+00
-1.14328420e+00 2.82763332e-01 2.67514527e-01 1.83323368e-01
-8.21681857e-01 4.06293303e-01 -2.03885257e-01 -7.15033531e-01
-1.56303883e-01 -5.09374559e-01 7.21308216e-02 -2.74597496e-01
-2.58813743e-02 4.25257564e-01 1.52662635e-01 -1.15837634e+00
-7.14081109e-01 1.72523665e+00 3.70975196e-01 6.64138570e-02
1.24799263e+00 -1.89644381e-01 -3.78795385e-01 2.76064128e-01
7.85720468e-01 4.89428431e-01 -1.12327349e+00 -2.38389745e-01
1.78417880e-02 -7.32973158e-01 2.90093590e-02 -8.90618004e-03
-9.46730852e-01 5.07854283e-01 7.01965094e-01 1.09910734e-01
9.05822396e-01 2.43898276e-02 8.29218388e-01 1.79151371e-01
6.40610456e-01 -7.33911693e-01 -3.81286085e-01 4.87216055e-01
7.61635661e-01 -1.12561977e+00 5.13529837e-01 -8.12244058e-01
-1.26241431e-01 9.59652007e-01 3.48701924e-01 -5.14388919e-01
6.78255260e-01 -8.38551223e-02 -6.04256503e-02 -5.54225802e-01
5.12268431e-02 -1.73827678e-01 5.00924826e-01 6.39029562e-01
-3.35024185e-02 -1.27505958e-01 -3.17914099e-01 3.22677977e-02
-1.67870879e-01 -1.50385067e-01 -9.24057607e-03 8.48055243e-01
-5.79497099e-01 -1.12882984e+00 -8.72659564e-01 2.24364489e-01
-2.46937759e-02 6.38567060e-02 -2.19168350e-01 8.11640263e-01
1.74143724e-02 6.43183947e-01 1.61511764e-01 -2.16231734e-01
8.06173563e-01 -2.23005489e-01 3.41839969e-01 -6.17105603e-01
-2.74470717e-01 1.31999716e-01 -5.54815888e-01 -6.69537544e-01
-9.08059299e-01 -8.80849421e-01 -1.53087604e+00 -3.57456505e-01
-5.60503483e-01 1.31975874e-01 7.98689544e-01 8.03823233e-01
5.68703711e-01 1.50034800e-01 8.68259668e-01 -1.31840813e+00
-5.16841173e-01 -4.54767793e-01 -5.55487990e-01 5.87675273e-01
1.65973306e-01 -1.06386209e+00 -7.06681132e-01 -1.85022607e-01] | [7.6797075271606445, -2.9339849948883057] |
e504413d-39f6-42c4-a4c1-f54b5f073a5e | 2dpass-2d-priors-assisted-semantic | 2207.04397 | null | https://arxiv.org/abs/2207.04397v3 | https://arxiv.org/pdf/2207.04397v3.pdf | 2DPASS: 2D Priors Assisted Semantic Segmentation on LiDAR Point Clouds | As camera and LiDAR sensors capture complementary information used in autonomous driving, great efforts have been made to develop semantic segmentation algorithms through multi-modality data fusion. However, fusion-based approaches require paired data, i.e., LiDAR point clouds and camera images with strict point-to-pixel mappings, as the inputs in both training and inference, which seriously hinders their application in practical scenarios. Thus, in this work, we propose the 2D Priors Assisted Semantic Segmentation (2DPASS), a general training scheme, to boost the representation learning on point clouds, by fully taking advantage of 2D images with rich appearance. In practice, by leveraging an auxiliary modal fusion and multi-scale fusion-to-single knowledge distillation (MSFSKD), 2DPASS acquires richer semantic and structural information from the multi-modal data, which are then online distilled to the pure 3D network. As a result, equipped with 2DPASS, our baseline shows significant improvement with only point cloud inputs. Specifically, it achieves the state-of-the-arts on two large-scale benchmarks (i.e. SemanticKITTI and NuScenes), including top-1 results in both single and multiple scan(s) competitions of SemanticKITTI. | ['Zhen Li', 'Shenghui Cui', 'Ruimao Zhang', 'Chao Zheng', 'Chaoda Zheng', 'Jiantao Gao', 'Xu Yan'] | 2022-07-10 | null | null | null | null | ['robust-3d-semantic-segmentation', 'lidar-semantic-segmentation'] | ['computer-vision', 'computer-vision'] | [ 1.61936253e-01 4.83294465e-02 -3.74516875e-01 -5.92086911e-01
-1.06455207e+00 -4.92644429e-01 5.64464808e-01 -4.55425158e-02
-4.94688720e-01 2.97606021e-01 -2.61538476e-01 -2.24603012e-01
8.71485472e-02 -8.57472718e-01 -1.26813006e+00 -5.83665133e-01
5.96923053e-01 7.50462115e-01 6.23555720e-01 -2.94840723e-01
1.10178359e-01 3.25651765e-01 -1.73991859e+00 -9.71094295e-02
1.07256758e+00 1.32692087e+00 5.40782809e-01 1.46625906e-01
-5.39640009e-01 3.70196581e-01 1.55827448e-01 -4.52716738e-01
5.45777321e-01 2.49100968e-01 -4.60387349e-01 7.40523860e-02
7.47765303e-01 -3.20562959e-01 -4.63543832e-01 1.27140558e+00
2.68527210e-01 2.42553934e-01 3.91931653e-01 -1.31647635e+00
-4.33329076e-01 3.01260322e-01 -9.62704182e-01 -1.51869148e-01
-7.54677877e-02 4.38994706e-01 8.85752857e-01 -1.11451924e+00
2.07939208e-01 1.53028762e+00 6.26347005e-01 3.15415621e-01
-1.12361574e+00 -8.48612249e-01 5.08371413e-01 2.20716894e-01
-1.43285239e+00 -3.32684994e-01 9.15635705e-01 -3.04471284e-01
8.60714436e-01 -6.75883815e-02 6.61412239e-01 9.09490287e-01
-2.96249062e-01 1.09929979e+00 1.11328971e+00 1.03611894e-01
1.38319924e-01 2.72181779e-02 1.69156417e-01 5.66218972e-01
2.33135819e-01 2.18963474e-01 -6.44921660e-01 1.85916469e-01
7.47453451e-01 3.83261561e-01 8.17296654e-02 -4.57786798e-01
-1.12611961e+00 6.87282503e-01 7.51326919e-01 -2.24047378e-01
-3.65042806e-01 3.50574195e-01 2.01795101e-01 -1.29226595e-01
4.95281875e-01 -7.97501728e-02 -5.09768188e-01 6.85084388e-02
-9.47706699e-01 2.45325118e-01 2.80991584e-01 1.22278094e+00
1.38261080e+00 -2.53810644e-01 2.19034433e-01 7.90230393e-01
6.29632950e-01 1.18122673e+00 1.19126953e-01 -1.13240898e+00
8.80263090e-01 7.45733321e-01 -4.11386155e-02 -7.74087131e-01
-2.75100112e-01 -2.46002942e-01 -6.50375664e-01 1.64485231e-01
7.47899264e-02 1.30886659e-01 -1.45560956e+00 1.62599885e+00
5.60304046e-01 7.82414258e-01 2.82705158e-01 1.19022238e+00
9.38958108e-01 5.89454293e-01 8.66842717e-02 2.55631208e-01
1.30482197e+00 -1.09654880e+00 -3.70266408e-01 -7.80455709e-01
3.83145154e-01 -4.14679587e-01 1.02433193e+00 2.38223761e-01
-8.67789686e-01 -7.35437930e-01 -1.09690511e+00 -4.04976130e-01
-4.37978476e-01 -1.33064106e-01 6.48105323e-01 3.12237829e-01
-7.78433204e-01 2.71619767e-01 -1.11542225e+00 -2.04653576e-01
8.11105490e-01 1.86871722e-01 -4.32357013e-01 -5.96564054e-01
-9.99884963e-01 7.60563493e-01 5.36473095e-01 1.97040617e-01
-9.34959531e-01 -9.34986532e-01 -9.21557724e-01 -2.43922547e-01
7.82942235e-01 -8.62889707e-01 1.21538639e+00 -4.55290973e-01
-1.28099954e+00 8.61050427e-01 -1.62427425e-01 -4.88416672e-01
5.35610795e-01 -4.34753835e-01 -2.13027447e-01 2.11991847e-01
2.57195204e-01 1.33098447e+00 6.57351971e-01 -1.65660501e+00
-9.99768913e-01 -7.76373684e-01 2.33078092e-01 4.44700152e-01
1.31121099e-01 -4.99090821e-01 -1.04517126e+00 -2.20583528e-01
5.41991889e-01 -1.07402205e+00 -3.32283169e-01 1.38076529e-01
-4.08770621e-01 -2.36431003e-01 9.78128314e-01 -4.59746659e-01
3.63851279e-01 -2.32707095e+00 1.96216628e-01 4.23415378e-02
2.64816642e-01 1.69675037e-01 -1.43954739e-01 -1.23990893e-01
3.20502490e-01 -1.34310588e-01 -6.64699137e-01 -9.02487516e-01
3.03257167e-01 7.30193019e-01 -4.00330335e-01 2.94740736e-01
3.31466675e-01 1.17975271e+00 -8.94329607e-01 -7.83005893e-01
5.27983963e-01 4.37617570e-01 -4.46822286e-01 3.26943211e-03
-5.34688354e-01 5.66873431e-01 -8.65078747e-01 7.09882677e-01
1.15516639e+00 -1.81456625e-01 -4.48756516e-01 -2.84773499e-01
-1.07239038e-01 1.09312683e-01 -9.59771037e-01 2.52343965e+00
-5.56273580e-01 2.54201919e-01 2.37421505e-02 -8.89948547e-01
8.38379383e-01 -5.05467989e-02 4.67539310e-01 -8.68479431e-01
-2.27222741e-02 3.97077978e-01 -5.17647207e-01 -2.39734367e-01
4.65675771e-01 -2.15055555e-01 -1.98149785e-01 -7.60342628e-02
3.70089933e-02 -6.43494606e-01 -1.96474716e-01 2.95875251e-01
7.62160540e-01 4.81802762e-01 -2.77379811e-01 1.05869651e-01
4.24583495e-01 4.19601500e-01 8.98226678e-01 5.12553632e-01
-9.69430655e-02 7.88196087e-01 1.26736417e-01 -3.23011614e-02
-7.34082043e-01 -1.22670507e+00 -7.65816793e-02 6.14727318e-01
9.52189744e-01 -1.22429781e-01 -5.27561784e-01 -7.12026238e-01
4.67360079e-01 7.60593951e-01 -3.08356434e-01 -2.63430923e-01
-4.70979273e-01 -4.45855737e-01 5.42237937e-01 7.51686633e-01
8.14010859e-01 -4.98054385e-01 -6.30447388e-01 4.86499704e-02
-2.30201691e-01 -1.73420179e+00 -2.47211188e-01 1.54567391e-01
-9.76671040e-01 -9.63009357e-01 -4.21139061e-01 -3.50553185e-01
2.88983047e-01 8.49253714e-01 8.67872894e-01 -2.40155548e-01
1.26584336e-01 3.31565022e-01 -3.25781137e-01 -5.12725830e-01
2.19019666e-01 3.74280922e-02 -9.53576341e-02 -2.82363221e-02
5.07783890e-01 -7.24056244e-01 -6.23920441e-01 3.51120770e-01
-8.80175054e-01 4.18245405e-01 9.83145118e-01 4.88317609e-01
1.12830079e+00 -1.50904611e-01 2.09317088e-01 -7.03056395e-01
-2.25694776e-01 -5.36959887e-01 -6.92556202e-01 2.12386176e-02
-4.34387416e-01 3.32012540e-04 2.48795912e-01 4.16518711e-02
-1.15832734e+00 2.62624770e-01 -1.70160338e-01 -1.20444202e+00
-3.01893920e-01 5.81880033e-01 -5.95184326e-01 -2.06162706e-01
3.40876907e-01 2.84601599e-01 1.05335124e-01 -7.13100910e-01
6.60538793e-01 6.34161651e-01 8.92257988e-01 -6.34544730e-01
1.04740334e+00 8.61816406e-01 1.24700807e-01 -5.17478049e-01
-1.10079837e+00 -7.65820801e-01 -6.72747552e-01 -3.91065143e-02
1.04173601e+00 -1.50648820e+00 -3.61284256e-01 7.03506410e-01
-1.14603484e+00 -2.31981218e-01 -1.98629990e-01 5.78965068e-01
-5.77729285e-01 2.91713357e-01 -5.33093996e-02 -6.00944221e-01
-6.33347034e-02 -1.35854816e+00 1.62743711e+00 3.22831154e-01
6.21496201e-01 -6.06853187e-01 -2.44326681e-01 8.12453568e-01
9.04034916e-03 2.43110448e-01 5.16554475e-01 -3.70707422e-01
-1.08510935e+00 1.01757944e-02 -6.71061456e-01 5.11078775e-01
-1.19306622e-02 -3.37245286e-01 -1.27526486e+00 -3.02547365e-02
-1.56343311e-01 -4.11726952e-01 1.25191319e+00 9.38501582e-02
1.27685392e+00 3.24204713e-01 -5.61802030e-01 9.64969575e-01
1.35589218e+00 -1.56483635e-01 3.72436941e-01 3.67094204e-02
1.24225628e+00 4.77967888e-01 1.03134978e+00 2.17222422e-01
1.05456960e+00 6.93233192e-01 9.61819053e-01 -8.10866579e-02
-2.67825097e-01 -3.86771291e-01 1.56824142e-01 6.59707129e-01
1.43928573e-01 8.87304079e-04 -1.04571676e+00 5.64209342e-01
-2.16920662e+00 -3.93212527e-01 -2.31209204e-01 2.01997447e+00
5.79607904e-01 3.82832289e-01 -1.62944928e-01 -2.23979369e-01
6.20084584e-01 3.68407905e-01 -1.18919420e+00 4.67689931e-01
-2.64651597e-01 5.64566366e-02 1.05815351e+00 2.87740558e-01
-1.19831932e+00 1.23149836e+00 4.30722427e+00 1.05805707e+00
-1.14927185e+00 3.97775263e-01 3.09355110e-01 -1.05801828e-01
-4.59899157e-01 1.46473929e-01 -9.08131063e-01 5.30900717e-01
7.73967862e-01 1.02744251e-01 3.29215139e-01 8.17644179e-01
9.96767357e-02 -1.75382867e-01 -8.77408862e-01 1.23687267e+00
-1.57309532e-01 -1.22782660e+00 -1.04934432e-01 9.65035856e-02
6.59197032e-01 8.05900455e-01 6.95309713e-02 3.08751464e-01
4.32865530e-01 -6.82835460e-01 1.02991104e+00 5.32026649e-01
7.63466358e-01 -6.86946630e-01 6.89596057e-01 5.99069595e-01
-1.36875284e+00 -4.69065877e-03 -4.05795008e-01 2.47418568e-01
3.35747004e-01 8.47260356e-01 -3.41514200e-01 1.19537783e+00
8.28555346e-01 1.10296309e+00 -3.43377084e-01 9.89649355e-01
-2.28523776e-01 4.34978187e-01 -8.68528843e-01 5.46876550e-01
4.96995598e-01 -2.63241291e-01 4.85346049e-01 7.71637559e-01
3.84266555e-01 2.43686408e-01 5.97282827e-01 9.70419109e-01
-1.43327534e-01 -4.06638950e-01 -3.40545565e-01 5.86948954e-02
6.68540478e-01 1.40332413e+00 -5.69025040e-01 -3.89388859e-01
-6.38009906e-01 6.27006173e-01 1.04302242e-01 2.35898569e-01
-9.30241764e-01 -6.12598099e-03 1.03776073e+00 -6.17407560e-02
3.79753053e-01 -5.50625503e-01 -5.75007677e-01 -1.11840558e+00
1.84807733e-01 -3.79510969e-01 1.71444669e-01 -8.70209098e-01
-1.27459621e+00 2.21010566e-01 2.77320862e-01 -1.15766907e+00
2.91317612e-01 -5.15757143e-01 -3.52543801e-01 1.07180381e+00
-2.22428036e+00 -1.69073677e+00 -8.33808184e-01 6.68361366e-01
4.69949603e-01 2.76067317e-01 9.04026106e-02 4.99767125e-01
-5.46293437e-01 2.89132446e-01 -2.89582580e-01 -1.49284482e-01
5.86540341e-01 -1.25392616e+00 7.28427231e-01 7.21346676e-01
1.06361404e-01 1.31192535e-01 3.04174870e-01 -6.78861618e-01
-1.75199509e+00 -1.58357453e+00 2.01662213e-01 -5.12736738e-01
5.20611823e-01 -2.77810335e-01 -1.03162503e+00 5.65947354e-01
-1.60151139e-01 2.99253494e-01 2.31369451e-01 -1.61008194e-01
-3.89108598e-01 -3.26846123e-01 -9.92037714e-01 2.84932345e-01
1.43575156e+00 -6.27182961e-01 -6.20888472e-01 3.87394786e-01
1.27197611e+00 -7.81070650e-01 -7.72584140e-01 7.36062169e-01
1.68107361e-01 -7.37455904e-01 1.21504498e+00 -1.63845241e-01
2.74493217e-01 -7.04261839e-01 -5.18424988e-01 -1.08143449e+00
1.11054227e-01 -1.63074449e-01 -5.99478111e-02 1.01985157e+00
2.87925780e-01 -7.82909989e-01 8.02575111e-01 4.61909831e-01
-8.27045083e-01 -5.29224813e-01 -1.16006017e+00 -8.19231927e-01
1.13683991e-01 -8.71525645e-01 1.00219166e+00 6.60211802e-01
-7.40056992e-01 2.69002765e-01 -8.30564797e-02 6.62038743e-01
8.30274403e-01 3.28684688e-01 1.00579715e+00 -1.39101100e+00
-3.66387963e-02 -3.10512364e-01 -5.02471685e-01 -1.55226970e+00
3.50898087e-01 -1.00630569e+00 2.95608282e-01 -1.60288072e+00
-8.79633874e-02 -7.58218288e-01 -3.96633267e-01 6.43499196e-01
-3.45007390e-01 2.40625963e-01 3.57701957e-01 3.61018211e-01
-7.09999561e-01 9.06534970e-01 1.28327560e+00 -3.59395921e-01
4.86421734e-02 -1.57568231e-01 -6.60281241e-01 7.26323426e-01
5.99738240e-01 -4.90110964e-01 -5.48388779e-01 -8.87510717e-01
1.61381066e-01 5.94225526e-02 8.10370028e-01 -1.22176385e+00
3.64084005e-01 -4.20672685e-01 -1.99646223e-04 -1.18024266e+00
8.96514237e-01 -1.01339650e+00 6.13165833e-02 -1.69885121e-02
1.58135235e-01 -2.57012308e-01 2.88473487e-01 9.31121111e-01
-3.94595385e-01 1.03331774e-01 6.59074306e-01 -1.08733416e-01
-1.23482740e+00 7.94752598e-01 5.13291061e-01 -6.42142519e-02
9.77874160e-01 -3.91735613e-01 -3.55720341e-01 1.90300554e-01
-3.75854820e-01 8.32602382e-01 7.19392300e-01 6.15080237e-01
6.11927211e-01 -1.27045035e+00 -4.87433821e-01 1.47637457e-01
5.18704593e-01 1.21633816e+00 6.45117342e-01 1.12751329e+00
-2.43109226e-01 2.78448224e-01 -3.80148701e-02 -1.16433787e+00
-8.35248590e-01 4.47899520e-01 9.83941406e-02 1.29357845e-01
-8.30483198e-01 1.00339878e+00 4.89062846e-01 -7.33615935e-01
-1.90136638e-02 -5.59665143e-01 1.39366299e-01 -6.80733174e-02
3.62323485e-02 3.25645730e-02 2.50487387e-01 -6.88692033e-01
-4.47675526e-01 8.16768110e-01 -3.67969926e-03 -1.29090875e-01
1.13883770e+00 -3.14513803e-01 8.03904310e-02 5.89830399e-01
1.14425313e+00 -4.72152352e-01 -1.72084725e+00 -6.79708481e-01
-2.33981311e-01 -6.83384180e-01 5.65885901e-01 -6.05225801e-01
-1.44112241e+00 1.22604120e+00 4.43509758e-01 -3.27175379e-01
1.00189960e+00 2.15170905e-01 1.22520185e+00 4.98873830e-01
7.10577369e-01 -1.04003370e+00 -2.27748856e-01 4.25142258e-01
5.60260892e-01 -1.71563745e+00 -1.28110155e-01 -6.71745121e-01
-7.72931278e-01 5.92769980e-01 6.97874427e-01 -4.48351055e-02
6.22993112e-01 -4.05987091e-02 2.90816631e-02 -2.38856688e-01
-5.95834911e-01 -5.17868280e-01 2.68294781e-01 5.15613377e-01
-5.56561649e-01 2.33176395e-01 3.48192543e-01 6.22583866e-01
2.81284805e-02 1.63952366e-01 -2.97646802e-02 1.00377607e+00
-6.14226699e-01 -8.73313189e-01 -2.88816035e-01 4.79806602e-01
3.44537079e-01 1.31103257e-03 -1.88404471e-01 7.10571527e-01
4.67486113e-01 8.83201599e-01 1.49462074e-01 -6.53292716e-01
5.09767592e-01 -1.69758439e-01 3.55955660e-01 -5.44035912e-01
-8.18971023e-02 -3.08198333e-02 -2.21667424e-01 -8.46357524e-01
-4.50785249e-01 -8.05753946e-01 -1.65721202e+00 -2.74650604e-01
-3.11816305e-01 5.60106570e-03 9.92436588e-01 1.20986414e+00
6.03949964e-01 5.00358939e-01 4.50459361e-01 -1.15535092e+00
-4.83533800e-01 -8.02806139e-01 -5.01260102e-01 1.63996890e-01
3.19525450e-01 -1.12424803e+00 -1.36234209e-01 -1.97593376e-01] | [8.290787696838379, -2.718316078186035] |
0aa448a5-b016-43ba-bfee-9deea9bf8486 | 2305-14814 | 2305.14814 | null | https://arxiv.org/abs/2305.14814v1 | https://arxiv.org/pdf/2305.14814v1.pdf | What functions can Graph Neural Networks compute on random graphs? The role of Positional Encoding | We aim to deepen the theoretical understanding of Graph Neural Networks (GNNs) on large graphs, with a focus on their expressive power. Existing analyses relate this notion to the graph isomorphism problem, which is mostly relevant for graphs of small sizes, or studied graph classification or regression tasks, while prediction tasks on nodes are far more relevant on large graphs. Recently, several works showed that, on very general random graphs models, GNNs converge to certains functions as the number of nodes grows. In this paper, we provide a more complete and intuitive description of the function space generated by equivariant GNNs for node-tasks, through general notions of convergence that encompass several previous examples. We emphasize the role of input node features, and study the impact of node Positional Encodings (PEs), a recent line of work that has been shown to yield state-of-the-art results in practice. Through the study of several examples of PEs on large random graphs, we extend previously known universality results to significantly more general models. Our theoretical results hint at some normalization tricks, which is shown numerically to have a positive impact on GNN generalization on synthetic and real data. Our proofs contain new concentration inequalities of independent interest. | ['Samuel Vaiter', 'Nicolas Keriven'] | 2023-05-24 | null | null | null | null | ['graph-classification'] | ['graphs'] | [ 2.98877120e-01 5.35616457e-01 -1.77987233e-01 -1.04588099e-01
1.04534425e-01 -6.46408856e-01 4.44035709e-01 3.87271106e-01
-3.18769753e-01 7.26254821e-01 -3.00737798e-01 -6.04856074e-01
-4.59159762e-01 -1.11945641e+00 -1.00017154e+00 -9.15288150e-01
-9.80322361e-01 2.60890931e-01 2.65511394e-01 -7.81290650e-01
-1.97010428e-01 7.12743580e-01 -1.20171523e+00 -4.08876419e-01
5.05404592e-01 6.51150644e-01 -2.12137416e-01 1.01744926e+00
2.45902568e-01 5.37723184e-01 -3.70348841e-01 -7.76606977e-01
5.75717807e-01 -3.53389651e-01 -9.58505571e-01 -1.13457792e-01
7.29366064e-01 3.88028651e-01 -6.67506516e-01 1.41406202e+00
3.55237335e-01 2.38521233e-01 4.85692203e-01 -1.55078721e+00
-9.30115640e-01 1.18059134e+00 -3.61848772e-01 3.32928926e-01
-3.17867815e-01 -7.68300965e-02 1.42321897e+00 -2.46498451e-01
6.41094744e-01 1.18801951e+00 1.16502666e+00 6.12558365e-01
-1.39867330e+00 -5.68859637e-01 2.88476869e-02 6.57661557e-02
-1.15756869e+00 2.13721348e-03 8.10805500e-01 -2.58865088e-01
6.83908582e-01 3.03287596e-01 6.44679785e-01 1.02158320e+00
2.71633357e-01 1.52065903e-01 6.31179392e-01 -6.09080791e-01
-1.22635281e-02 -9.84265357e-02 5.38123608e-01 9.69285011e-01
8.71941209e-01 3.55774686e-02 7.30884746e-02 1.20279752e-01
9.73775864e-01 -1.36236995e-01 -5.18941879e-01 -8.17965865e-01
-8.66901696e-01 1.14537799e+00 9.77865517e-01 6.83236539e-01
-1.09420866e-01 5.85580349e-01 5.05053341e-01 7.91759193e-01
4.84669715e-01 4.19913709e-01 -3.99739057e-01 4.28296506e-01
-2.78770745e-01 1.23974653e-02 1.18999696e+00 1.04163361e+00
7.74286807e-01 1.11365981e-01 1.69504061e-01 5.39816320e-01
-1.72936469e-01 2.87363380e-01 8.47564191e-02 -5.91442585e-01
4.56542164e-01 5.45716405e-01 -6.23029292e-01 -1.16210043e+00
-8.77261519e-01 -9.39488471e-01 -1.44647336e+00 -1.16434298e-01
7.12519765e-01 -6.08120188e-02 -6.66260004e-01 2.31888962e+00
-1.26938567e-01 2.47956291e-02 -2.03858372e-02 4.80508089e-01
5.27221680e-01 3.12637419e-01 -5.23281321e-02 -6.45212010e-02
8.89770389e-01 -6.91730440e-01 -3.52298170e-01 1.60194095e-02
1.00294769e+00 5.52559979e-02 1.02874422e+00 9.73113328e-02
-9.21355963e-01 -2.83443362e-01 -9.70777988e-01 9.99718606e-02
-6.14541352e-01 -3.13497186e-01 9.90508020e-01 1.00695872e+00
-1.68864000e+00 1.22763646e+00 -5.78696191e-01 -8.74045908e-01
5.37049055e-01 7.17756510e-01 -5.30983627e-01 -1.02112025e-01
-1.37273407e+00 6.45298362e-01 4.72466886e-01 2.01630622e-01
-5.36525369e-01 -5.82285881e-01 -9.95132983e-01 3.79520327e-01
4.59049523e-01 -7.99630344e-01 9.34738576e-01 -1.21829510e+00
-9.63010073e-01 7.58740008e-01 3.73714983e-01 -8.45961332e-01
3.17979068e-01 5.27493000e-01 -3.01036894e-01 7.12182745e-02
-2.83835173e-01 3.41030240e-01 3.99964899e-01 -8.00390422e-01
-3.48470844e-02 -4.92335498e-01 5.12112439e-01 -1.88874096e-01
-7.36882389e-01 -4.11257952e-01 6.29143864e-02 -5.37544787e-01
-2.47992337e-01 -9.78332400e-01 -5.23556113e-01 -6.76729009e-02
-4.38418984e-01 -4.62767482e-01 3.08040053e-01 -1.77598163e-01
1.17842996e+00 -1.94192028e+00 4.89609927e-01 6.07142270e-01
8.63259912e-01 9.14939493e-02 -3.91618907e-01 5.91800690e-01
-4.19435441e-01 5.04983366e-01 -2.57195085e-01 3.02096792e-02
1.78375125e-01 3.06727231e-01 -2.43529808e-02 8.31376374e-01
-1.92065665e-03 1.26481307e+00 -7.64984250e-01 -1.33426383e-01
-5.13511226e-02 3.39638829e-01 -6.01030886e-01 -4.30935055e-01
-9.78870019e-02 -1.54608965e-01 -2.04260230e-01 1.42115980e-01
5.77455401e-01 -7.93738961e-01 4.21901196e-01 6.80934116e-02
4.15527254e-01 -7.40172528e-03 -8.67882431e-01 1.33056414e+00
-2.11503178e-01 7.84288883e-01 2.36969918e-01 -1.57990074e+00
5.99308312e-01 2.75044492e-03 1.98513448e-01 -2.98112750e-01
3.08379978e-01 6.14701770e-02 5.21032393e-01 -3.19200195e-02
2.62454718e-01 -2.63126791e-01 -1.69707507e-01 2.67873853e-01
3.31648558e-01 2.93591946e-01 5.74996471e-01 4.96940851e-01
1.43035161e+00 -5.02737105e-01 4.72123802e-01 -7.82892287e-01
4.40660536e-01 -4.00102317e-01 -7.34503642e-02 1.07697845e+00
1.44412005e-02 5.32378256e-02 1.04761195e+00 -4.09700155e-01
-1.17277288e+00 -8.49645674e-01 -4.46480215e-02 1.28256714e+00
1.27522603e-01 -5.56659877e-01 -9.04244959e-01 -5.55374682e-01
-1.55799678e-02 3.31735224e-01 -1.17081392e+00 -4.08866495e-01
-5.86426377e-01 -9.74167585e-01 7.51490474e-01 5.57169557e-01
1.22542575e-01 -9.86325264e-01 -1.01386271e-01 -1.09520935e-01
5.68036854e-01 -1.18359923e+00 -1.55904830e-01 4.47739035e-01
-1.11424541e+00 -1.27616692e+00 -9.07019615e-01 -9.78057206e-01
6.76439285e-01 2.09153742e-01 1.30683637e+00 4.77434099e-01
-5.33180274e-02 6.29981756e-01 -1.57693818e-01 -9.17574018e-02
-7.52584755e-01 7.09597945e-01 1.15957171e-01 -1.92423001e-01
2.59868763e-02 -1.09445441e+00 -1.65643856e-01 1.76677331e-01
-1.06571233e+00 -3.03561315e-02 4.87345457e-01 7.40411639e-01
4.08915132e-02 3.34003717e-02 5.07820249e-01 -1.25016725e+00
8.51265490e-01 -5.54100811e-01 -7.23152459e-01 2.83182442e-01
-6.32655501e-01 5.02399921e-01 1.20045745e+00 -3.78709942e-01
-6.42817244e-02 -3.31214368e-01 -5.86310774e-02 -3.21883053e-01
1.94893435e-01 7.04710066e-01 -8.93378630e-02 -5.60461104e-01
8.27666759e-01 1.29182160e-01 6.51251599e-02 -1.64138928e-01
5.53536355e-01 3.88564751e-03 4.14044410e-01 -5.46563983e-01
1.08054590e+00 2.09376439e-01 8.66780818e-01 -9.71545637e-01
-5.09556592e-01 -1.63357079e-01 -6.45325184e-01 -2.06545200e-02
4.65895921e-01 -3.84583831e-01 -1.00364339e+00 2.81097710e-01
-9.04241681e-01 -5.59788465e-01 -2.97432035e-01 1.25021443e-01
-4.12372351e-01 5.06741762e-01 -8.09162319e-01 -5.66288948e-01
-2.55209923e-01 -7.87017465e-01 4.79932100e-01 5.60561707e-03
2.57648915e-01 -1.63392127e+00 1.43382788e-01 -4.93734568e-01
5.42643070e-01 2.25600675e-01 1.45282233e+00 -1.05834734e+00
-4.77117747e-01 -1.21363856e-01 -2.37599820e-01 3.08443099e-01
-1.97788581e-01 -2.72483259e-01 -6.51260614e-01 -5.58101952e-01
-3.35091650e-01 8.57706591e-02 1.11366379e+00 4.05991763e-01
1.19667721e+00 -4.33077246e-01 -4.52037990e-01 7.48405039e-01
1.63195574e+00 -4.53483254e-01 5.20462215e-01 -8.94566774e-02
1.03387749e+00 5.00743210e-01 -4.12586093e-01 -5.55096641e-02
3.67816463e-02 3.93130720e-01 7.94485688e-01 -1.51140057e-02
-9.90467295e-02 -1.25347629e-01 6.59327731e-02 9.61028576e-01
-5.10106266e-01 -6.55368388e-01 -6.92968965e-01 3.00244093e-01
-1.60379779e+00 -7.71323264e-01 -3.72775882e-01 2.12620783e+00
3.91864300e-01 1.71495691e-01 3.26685786e-01 2.60563225e-01
1.16394317e+00 3.28060120e-01 -4.54346627e-01 -2.74242431e-01
-3.62708926e-01 5.77480793e-01 1.08266354e+00 5.82992554e-01
-9.16191936e-01 7.94735312e-01 6.32432556e+00 8.08031797e-01
-9.81707335e-01 1.30554423e-01 5.53513587e-01 1.94573432e-01
-5.45102537e-01 -9.04842317e-02 -4.29691881e-01 7.00530857e-02
1.10702908e+00 -3.29797029e-01 6.69970095e-01 8.43182266e-01
-4.82128620e-01 5.86128831e-01 -1.42645419e+00 8.14255416e-01
6.08966826e-03 -1.36117864e+00 1.14573874e-01 4.53434765e-01
7.46732473e-01 1.97454080e-01 1.73642971e-02 4.29859102e-01
5.35598278e-01 -1.32360387e+00 1.59540713e-01 4.56016250e-02
7.39091873e-01 -9.21961486e-01 5.80897212e-01 1.22645892e-01
-1.36438656e+00 -1.55991182e-01 -7.30411053e-01 -2.00960368e-01
-2.90175468e-01 4.85603899e-01 -6.10948682e-01 8.03723216e-01
3.45623463e-01 8.39540064e-01 -8.27796459e-01 7.96482444e-01
-2.68580560e-02 7.47531533e-01 -4.98048484e-01 -5.26782215e-01
3.25170040e-01 -1.09878354e-01 6.17374897e-01 1.08783436e+00
1.11057974e-01 -1.71181738e-01 -2.11778834e-01 9.38967705e-01
-6.10402822e-01 3.06850046e-01 -1.04054046e+00 -1.84815153e-01
4.60388102e-02 1.39218616e+00 -1.07435417e+00 1.39072418e-01
-2.48437330e-01 7.16976404e-01 7.52873957e-01 3.36854577e-01
-7.18354762e-01 -5.02170205e-01 3.72553110e-01 3.18929166e-01
4.71514553e-01 -2.77981400e-01 1.18897550e-01 -1.00629091e+00
1.18488684e-01 -5.00428081e-01 4.94709224e-01 -2.90250540e-01
-1.41174448e+00 6.63650572e-01 -4.66692410e-02 -7.30320752e-01
-1.70808986e-01 -1.15975082e+00 -5.94867289e-01 5.22771895e-01
-1.12293744e+00 -8.71498525e-01 -1.36014134e-01 5.40659785e-01
-1.20417312e-01 -6.93837777e-02 7.21180081e-01 2.52889633e-01
-5.27062774e-01 8.51363599e-01 1.69408977e-01 4.07277524e-01
8.93993750e-02 -1.32579696e+00 8.42418790e-01 9.66145039e-01
4.26341116e-01 5.88320076e-01 6.65350497e-01 -3.40240836e-01
-1.63977444e+00 -9.31697309e-01 6.29296184e-01 -3.51710320e-01
9.96227205e-01 -7.10138023e-01 -9.14584935e-01 9.73144174e-01
-1.95565587e-03 4.05794710e-01 3.20434570e-01 4.46435958e-01
-4.49227035e-01 -3.02014016e-02 -8.77953768e-01 7.01331854e-01
1.62311089e+00 -3.64698082e-01 -1.39899366e-02 3.32726777e-01
9.69525576e-01 -1.36148423e-01 -9.08724189e-01 4.13284183e-01
4.25843507e-01 -8.57676685e-01 9.24903095e-01 -1.05295599e+00
3.11979204e-01 2.54136384e-01 -9.10972282e-02 -1.46833086e+00
-4.81053293e-01 -8.73098373e-01 -1.10583864e-01 7.09658980e-01
5.51776946e-01 -1.10163426e+00 7.46862292e-01 1.21663421e-01
2.85047852e-03 -8.86529267e-01 -1.00561035e+00 -9.70956802e-01
4.43361908e-01 -3.39828849e-01 3.99187088e-01 1.04365039e+00
1.20787971e-01 7.48060346e-01 -1.82487413e-01 5.01707979e-02
5.26642442e-01 -1.53201431e-01 6.04805946e-01 -1.74696720e+00
-3.81052911e-01 -9.23980236e-01 -1.06229115e+00 -9.46497917e-01
4.80551481e-01 -1.63211274e+00 -4.38959509e-01 -1.23983204e+00
2.72840589e-01 -3.47879499e-01 -3.23247164e-01 2.36217365e-01
2.87044924e-02 2.96854556e-01 1.39168680e-01 -1.70797840e-01
-6.41398668e-01 2.93752372e-01 1.24264467e+00 -8.30021426e-02
8.65833908e-02 2.16860086e-01 -8.90684426e-01 4.87056702e-01
8.72824192e-01 -4.46010917e-01 -5.46886623e-01 -1.45830482e-01
7.32847869e-01 -1.78780749e-01 7.09418535e-01 -1.06258154e+00
2.10956216e-01 2.34167963e-01 4.93007340e-02 1.36490352e-02
6.80344086e-03 -6.83831871e-01 -1.34661406e-01 6.63101017e-01
-5.99913418e-01 3.35147142e-01 8.80373269e-02 7.24590719e-01
2.94387132e-01 -3.97495717e-01 6.65914953e-01 -3.21456268e-02
-3.15161049e-01 7.97344625e-01 4.90288027e-02 3.73303205e-01
8.08452666e-01 -1.54505908e-01 -4.17196453e-01 -7.96066344e-01
-8.13015819e-01 1.20685725e-02 4.22471941e-01 1.41942814e-01
1.45867944e-01 -1.30662584e+00 -7.58113563e-01 9.73473191e-02
4.53341380e-02 -1.41341358e-01 1.03974789e-01 9.37544703e-01
-4.69759494e-01 4.10582066e-01 -1.15198836e-01 -5.26673496e-01
-1.07011557e+00 8.63272965e-01 5.73059201e-01 -4.87913638e-01
-7.26971328e-01 7.36631095e-01 4.78691310e-01 -4.08711791e-01
3.06330144e-01 -5.21496952e-01 -2.67346073e-02 -3.03365529e-01
7.35553354e-02 4.24798876e-01 7.79501572e-02 -4.49324131e-01
-1.04968898e-01 5.33539236e-01 5.54217957e-02 3.15339863e-01
1.39933825e+00 2.21557096e-01 -2.62171060e-01 3.79922420e-01
1.41331458e+00 -1.23395376e-01 -8.07027340e-01 -3.30336541e-01
-6.89485595e-02 2.62565404e-01 -5.10008693e-01 -1.96481124e-01
-1.36579013e+00 9.96841729e-01 3.17928433e-01 9.66061115e-01
1.07710969e+00 4.64869827e-01 3.59925747e-01 8.25642347e-01
3.25529516e-01 -5.02400577e-01 -2.69690990e-01 6.72300696e-01
6.78830385e-01 -8.43263924e-01 -2.11759627e-01 -5.63908696e-01
6.76513389e-02 1.33955419e+00 3.17240059e-01 -4.13159579e-01
7.38844037e-01 1.48470402e-01 -7.81984806e-01 -3.06343049e-01
-5.48029065e-01 -2.40730196e-01 1.92446619e-01 6.23981535e-01
2.50392884e-01 2.73539364e-01 -1.03968062e-01 5.20182908e-01
-4.60794181e-01 -3.71991634e-01 6.58417642e-01 2.70799518e-01
-3.04338336e-01 -1.10781443e+00 2.38432959e-01 5.80991149e-01
-5.10288000e-01 -4.57791358e-01 -6.75334752e-01 1.39503348e+00
-2.99475849e-01 3.02380770e-01 2.26821862e-02 -4.14954752e-01
5.72381951e-02 -7.29059801e-02 1.04011571e+00 -4.12029564e-01
-5.04931211e-01 -6.44671619e-01 3.79273929e-02 -2.78965503e-01
-3.83280396e-01 -2.42969066e-01 -9.47204590e-01 -7.08347678e-01
-4.43406463e-01 1.77597418e-01 3.67589056e-01 7.35722601e-01
1.29466459e-01 6.21944070e-01 3.21743250e-01 -7.73799896e-01
-7.59022593e-01 -1.09560227e+00 -7.84544170e-01 2.69084394e-01
3.62590104e-01 -2.88095266e-01 -6.56370223e-01 -5.46922445e-01] | [6.830629348754883, 6.112415790557861] |
a5775c4e-08db-4a0e-8a3f-982cffffaec7 | swin-transformer-hierarchical-vision | 2103.14030 | null | https://arxiv.org/abs/2103.14030v2 | https://arxiv.org/pdf/2103.14030v2.pdf | Swin Transformer: Hierarchical Vision Transformer using Shifted Windows | This paper presents a new vision Transformer, called Swin Transformer, that capably serves as a general-purpose backbone for computer vision. Challenges in adapting Transformer from language to vision arise from differences between the two domains, such as large variations in the scale of visual entities and the high resolution of pixels in images compared to words in text. To address these differences, we propose a hierarchical Transformer whose representation is computed with \textbf{S}hifted \textbf{win}dows. The shifted windowing scheme brings greater efficiency by limiting self-attention computation to non-overlapping local windows while also allowing for cross-window connection. This hierarchical architecture has the flexibility to model at various scales and has linear computational complexity with respect to image size. These qualities of Swin Transformer make it compatible with a broad range of vision tasks, including image classification (87.3 top-1 accuracy on ImageNet-1K) and dense prediction tasks such as object detection (58.7 box AP and 51.1 mask AP on COCO test-dev) and semantic segmentation (53.5 mIoU on ADE20K val). Its performance surpasses the previous state-of-the-art by a large margin of +2.7 box AP and +2.6 mask AP on COCO, and +3.2 mIoU on ADE20K, demonstrating the potential of Transformer-based models as vision backbones. The hierarchical design and the shifted window approach also prove beneficial for all-MLP architectures. The code and models are publicly available at~\url{https://github.com/microsoft/Swin-Transformer}. | ['Baining Guo', 'Stephen Lin', 'Zheng Zhang', 'Yixuan Wei', 'Han Hu', 'Yue Cao', 'Yutong Lin', 'Ze Liu'] | 2021-03-25 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Liu_Swin_Transformer_Hierarchical_Vision_Transformer_Using_Shifted_Windows_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Liu_Swin_Transformer_Hierarchical_Vision_Transformer_Using_Shifted_Windows_ICCV_2021_paper.pdf | iccv-2021-1 | ['thermal-image-segmentation', 'real-time-object-detection'] | ['computer-vision', 'computer-vision'] | [ 1.82372183e-01 1.57639908e-03 -4.85054813e-02 -3.58138800e-01
-7.29461253e-01 -6.15851343e-01 6.42376363e-01 -2.69251287e-01
-7.77675867e-01 3.67500484e-01 -2.36155838e-01 -3.73217285e-01
1.51218578e-01 -6.09723985e-01 -8.61212671e-01 -5.43475509e-01
2.19778359e-01 2.48686418e-01 6.85864091e-01 -6.12739548e-02
-4.94533591e-02 3.05841416e-01 -1.51297891e+00 5.78214586e-01
6.20469570e-01 1.38933003e+00 5.64177394e-01 6.96616054e-01
-6.42511249e-02 6.07455730e-01 -5.06017506e-01 -5.71002185e-01
3.33034515e-01 7.47725666e-02 -7.72623301e-01 -4.06442657e-02
1.05021942e+00 -1.67166740e-01 -4.03607845e-01 9.61229920e-01
4.55953538e-01 -1.46034211e-01 4.79017586e-01 -1.17081702e+00
-1.03054464e+00 5.70169151e-01 -7.66602039e-01 6.18168533e-01
-1.45007506e-01 3.25163037e-01 1.31351495e+00 -1.32981944e+00
4.62108701e-01 1.30443454e+00 9.15570796e-01 5.53767145e-01
-1.28292537e+00 -6.27563477e-01 3.31383288e-01 3.71149361e-01
-1.26903760e+00 -4.65942293e-01 2.11726248e-01 -4.75573987e-01
1.38809872e+00 3.41435015e-01 7.21069455e-01 1.01800573e+00
1.84896648e-01 1.03347909e+00 1.35938323e+00 -3.34871620e-01
-5.78509122e-02 2.03806773e-01 2.69485891e-01 7.85833180e-01
9.34083536e-02 -2.45299667e-01 -5.36119103e-01 2.57182300e-01
8.48902166e-01 -4.36863713e-02 -2.49845624e-01 -3.23753990e-02
-1.22725272e+00 7.77529538e-01 6.96584463e-01 2.61454582e-01
-1.03470027e-01 4.06179100e-01 3.00094545e-01 1.18462525e-01
3.82729650e-01 4.11373734e-01 -6.23650789e-01 3.92272286e-02
-9.82863486e-01 7.85177052e-02 4.14246917e-01 9.87535834e-01
6.18531466e-01 2.75566339e-01 -2.65374005e-01 9.70757067e-01
2.68663555e-01 7.16058075e-01 5.04544854e-01 -9.76784110e-01
3.82305294e-01 5.90530872e-01 -2.35696033e-01 -5.62051415e-01
-3.39081049e-01 -5.74185073e-01 -6.35546803e-01 4.04756337e-01
6.40059054e-01 5.15532196e-02 -1.46394658e+00 1.57036614e+00
1.36469826e-02 -1.37072355e-01 -1.62359163e-01 6.36205971e-01
1.00712490e+00 7.20777154e-01 1.96874410e-01 3.72691482e-01
1.85396767e+00 -1.18102646e+00 -1.89564869e-01 -6.65434062e-01
2.71312773e-01 -8.43935728e-01 1.44511139e+00 2.25522026e-01
-1.23046207e+00 -6.90235734e-01 -9.04288590e-01 -3.43961060e-01
-5.90786636e-01 3.61904472e-01 4.06689137e-01 4.45207119e-01
-1.53419948e+00 1.85330719e-01 -7.79472649e-01 -6.93118453e-01
6.97380364e-01 4.87591475e-01 -2.27084845e-01 4.07309011e-02
-9.24095988e-01 1.06100082e+00 3.78526032e-01 3.21739465e-02
-6.77278221e-01 -8.50627780e-01 -8.24653745e-01 1.42607614e-01
2.66161114e-01 -5.72428823e-01 1.50797331e+00 -1.02524149e+00
-1.15497077e+00 1.15540290e+00 -7.48594403e-02 -8.68276060e-01
5.87303817e-01 -2.68431872e-01 -2.00447306e-01 1.73077986e-01
3.06433558e-01 1.27323246e+00 1.01978171e+00 -8.18886876e-01
-8.73066425e-01 -2.99286932e-01 2.45574817e-01 2.13060334e-01
-5.05074084e-01 1.22856073e-01 -9.79528248e-01 -7.30384886e-01
-2.28113517e-01 -9.42622542e-01 -9.95794162e-02 2.06611946e-01
-2.56207526e-01 -2.26149663e-01 9.68084812e-01 -5.91597140e-01
9.28107917e-01 -2.04902339e+00 -1.86361745e-01 -1.54838130e-01
4.24207568e-01 5.33282936e-01 -2.03518361e-01 4.94402535e-02
4.51660864e-02 1.15409773e-02 -2.26420730e-01 -3.87302339e-01
-3.60946283e-02 2.05947384e-01 -1.99236587e-01 2.46151224e-01
3.27821672e-01 1.21900499e+00 -4.32880223e-01 -4.71279293e-01
3.27305436e-01 5.31437576e-01 -3.30821902e-01 -2.20373407e-01
-1.38388082e-01 -1.62919968e-01 -1.71539694e-01 8.23196411e-01
5.29656291e-01 -5.66656470e-01 -1.75719336e-01 -4.68358576e-01
-2.57457227e-01 1.89150330e-02 -9.60187197e-01 1.48909533e+00
-3.20222795e-01 1.01780486e+00 2.82888800e-01 -8.38192761e-01
7.42021441e-01 1.46814249e-02 1.46619409e-01 -8.64844382e-01
8.56985375e-02 1.54445479e-02 -3.99233885e-02 -2.32441708e-01
4.84805554e-01 2.03699246e-01 9.29830223e-02 -2.96551120e-02
3.53022128e-01 -1.08986236e-01 3.64640981e-01 1.06728245e-02
9.65283215e-01 -4.19869199e-02 1.99328922e-02 -4.61283505e-01
4.00062293e-01 1.73276603e-01 5.27255535e-01 8.06062222e-01
-2.82016814e-01 6.89068437e-01 3.97093773e-01 -4.02726024e-01
-1.10872877e+00 -1.12628925e+00 -3.74012411e-01 1.21604562e+00
5.64571172e-02 -4.63463455e-01 -9.57574129e-01 -5.73238432e-01
-5.96979372e-02 4.27350312e-01 -7.05132484e-01 1.68897972e-01
-4.98509228e-01 -8.59038591e-01 8.00813198e-01 1.04041135e+00
8.41846228e-01 -1.11570072e+00 -8.77796054e-01 8.41120165e-03
-5.49320541e-02 -1.45261979e+00 -5.46075642e-01 2.41498560e-01
-6.67713881e-01 -9.82767999e-01 -8.08548868e-01 -9.80472624e-01
4.64462817e-01 2.59869039e-01 1.30284131e+00 -2.93268055e-01
-5.89287639e-01 5.60842395e-01 -5.12560904e-02 -6.21359766e-01
2.80096754e-02 9.38065946e-02 -2.22454339e-01 -5.72752543e-02
3.47118050e-01 -2.53714740e-01 -8.00314605e-01 3.49133551e-01
-7.56712079e-01 2.13031814e-01 6.90153480e-01 8.65289867e-01
7.33106375e-01 -3.65271211e-01 2.71748096e-01 -6.41958416e-01
2.88892955e-01 -1.04388677e-01 -7.48608530e-01 3.00848603e-01
-6.80456758e-01 -2.43299320e-01 4.16825414e-01 -4.38773751e-01
-8.21488678e-01 1.39439292e-02 -4.18568961e-02 -4.47180003e-01
-2.38848552e-02 1.32815287e-01 7.73182735e-02 -3.72043438e-02
5.09079814e-01 2.18762949e-01 -1.29320726e-01 -4.70117182e-01
5.23006439e-01 5.14729619e-01 6.03387535e-01 -2.32217833e-01
5.97496033e-01 4.91485149e-01 -4.18034285e-01 -9.16954517e-01
-7.49337256e-01 -3.03156495e-01 -4.65598345e-01 -1.51113356e-02
1.19113898e+00 -1.12063324e+00 -5.81400752e-01 7.29862630e-01
-9.72090423e-01 -5.41048408e-01 -3.30612272e-01 2.15087757e-01
-3.47883970e-01 8.92576352e-02 -8.18937957e-01 -4.75103170e-01
-7.54849315e-01 -1.27535737e+00 9.70467985e-01 4.35513318e-01
-1.48326829e-01 -8.26868415e-01 -5.19041777e-01 6.54321194e-01
4.31796700e-01 -4.21274677e-02 7.48409331e-01 -5.23796558e-01
-6.55410767e-01 4.42255940e-03 -7.15418398e-01 7.02295303e-01
-2.34207377e-01 -1.72588509e-02 -1.27198076e+00 -2.44147807e-01
-2.65015632e-01 -2.86595523e-01 1.16201401e+00 5.09527922e-01
1.04245806e+00 -3.95111041e-03 -3.05063397e-01 7.37969816e-01
1.44594562e+00 1.48978010e-01 3.58174056e-01 5.21699011e-01
8.16768527e-01 3.22531044e-01 3.44483405e-01 1.36871770e-01
5.32142818e-01 6.74870074e-01 5.41387498e-01 -4.42523003e-01
-4.21129107e-01 1.06507927e-01 4.64640051e-01 4.44163144e-01
-1.46703878e-02 -1.42575413e-01 -1.22500956e+00 8.06703746e-01
-1.81660211e+00 -7.26765513e-01 -1.18311055e-01 1.87408233e+00
7.12814689e-01 4.03026730e-01 5.39110750e-02 -2.00402424e-01
7.04343498e-01 1.72555998e-01 -5.85190177e-01 -4.84702885e-01
-2.88709581e-01 3.96701664e-01 7.94679701e-01 3.80453229e-01
-1.21631134e+00 1.17927849e+00 5.95385218e+00 1.01713312e+00
-1.28077722e+00 2.27723777e-01 8.23898315e-01 -2.44865552e-01
1.45390674e-01 -3.17831486e-01 -1.11596656e+00 4.15564239e-01
7.61247993e-01 1.67864084e-01 2.22582385e-01 8.82598996e-01
-5.44684066e-04 -1.04733132e-01 -1.06496668e+00 1.15928817e+00
9.85736698e-02 -1.60808933e+00 6.13273792e-02 -1.09189853e-01
4.90372956e-01 7.91177630e-01 3.97480011e-01 2.80337691e-01
1.81756496e-01 -1.15288889e+00 1.07585669e+00 7.78351426e-02
1.09328341e+00 -3.42417091e-01 5.43196261e-01 5.26017062e-02
-1.44094098e+00 -1.39583170e-01 -2.66018778e-01 1.19749874e-01
-2.11363673e-01 3.47234726e-01 -8.46447766e-01 6.16739616e-02
1.44559968e+00 7.17475295e-01 -8.22118044e-01 9.32667732e-01
-1.04376860e-02 5.74002862e-01 -4.90034670e-01 1.79157704e-01
6.04430497e-01 -5.40256165e-02 4.25029308e-01 1.64257348e+00
1.71208158e-01 -1.84688449e-01 1.64307758e-01 8.79903913e-01
-2.31002226e-01 -8.54309648e-02 -3.83642495e-01 2.21536160e-01
3.74435484e-01 1.34799087e+00 -9.38421905e-01 -3.71558696e-01
-7.11591125e-01 1.04576242e+00 3.08005154e-01 4.11028475e-01
-1.07309365e+00 -4.04999673e-01 7.35273480e-01 1.74818143e-01
9.14246857e-01 -1.94592327e-01 -4.49446023e-01 -1.01353109e+00
1.91655234e-01 -8.77812386e-01 3.20405126e-01 -9.02131379e-01
-1.09414363e+00 9.07542765e-01 -4.96312082e-02 -8.25528145e-01
2.61678398e-01 -1.06143486e+00 -5.65392017e-01 7.95624614e-01
-1.55451405e+00 -1.57172477e+00 -3.01928163e-01 7.65845239e-01
9.65574801e-01 -5.06955497e-02 6.69942141e-01 2.85117328e-01
-6.57856822e-01 7.18927920e-01 1.02417305e-01 3.25730652e-01
5.89902818e-01 -1.48867571e+00 7.19399989e-01 8.84766519e-01
3.65933985e-01 4.02059346e-01 3.89188707e-01 -1.37523010e-01
-1.12420857e+00 -1.17260110e+00 7.38623261e-01 -7.97887266e-01
7.32994497e-01 -5.16989470e-01 -7.83137143e-01 8.48053515e-01
4.77795601e-01 3.12599689e-01 2.66786605e-01 -7.65705183e-02
-6.49725080e-01 -3.68275642e-01 -1.05728936e+00 6.77849412e-01
8.43078077e-01 -5.36975682e-01 -5.51802218e-01 2.02273160e-01
7.19821751e-01 -4.53795791e-01 -7.94588506e-01 4.02982384e-01
5.58495224e-01 -9.80163693e-01 1.11836076e+00 -1.55577511e-01
2.06374004e-01 -2.07800478e-01 -2.03067392e-01 -8.65528405e-01
-4.43061560e-01 -4.63059902e-01 -5.07493131e-02 1.22167158e+00
6.61382496e-01 -6.64694488e-01 6.53730571e-01 5.06154120e-01
-1.61317840e-01 -1.05269599e+00 -9.69786525e-01 -6.73787534e-01
4.40169685e-02 -5.83507419e-01 1.55798003e-01 6.47728980e-01
-5.49878001e-01 4.93809432e-01 -1.94523986e-02 6.76372275e-02
4.75815088e-01 -2.14713048e-02 3.15504313e-01 -1.07525253e+00
-2.99650818e-01 -6.71361923e-01 -3.30534786e-01 -1.17539310e+00
-1.84879467e-01 -7.96430349e-01 -2.02130660e-01 -1.63979638e+00
1.20290197e-01 -3.80706370e-01 -4.63135213e-01 8.48695278e-01
-3.96014974e-02 8.10117304e-01 6.35852933e-01 1.77019119e-01
-5.11226952e-01 2.28125274e-01 9.94426668e-01 -3.16883028e-01
5.59271537e-02 -1.54954910e-01 -6.91760540e-01 1.06530416e+00
9.31456804e-01 -2.06912026e-01 -4.25046951e-01 -8.68915319e-01
-1.36799246e-01 -4.49940443e-01 6.11327946e-01 -1.12193167e+00
1.88436761e-01 1.40818238e-01 5.08512318e-01 -4.71237957e-01
6.18766606e-01 -7.27372646e-01 -2.88571537e-01 4.04885292e-01
-1.27970532e-01 4.90074754e-01 6.22341454e-01 4.19989109e-01
-2.04092190e-01 -1.88605080e-03 1.00669658e+00 -3.83495480e-01
-1.12791121e+00 2.20500499e-01 -3.93796980e-01 2.83027738e-01
1.11122906e+00 -6.23116553e-01 -5.27916431e-01 5.59007004e-03
-7.27482319e-01 3.02789927e-01 3.72170031e-01 5.71164846e-01
6.23821318e-01 -8.80459785e-01 -6.35249794e-01 2.19960153e-01
1.49047375e-01 -8.83958593e-04 9.02602598e-02 8.86612475e-01
-5.14226556e-01 5.48494160e-01 -3.26231509e-01 -8.90928805e-01
-1.52238560e+00 2.09523216e-01 4.79242116e-01 -2.02831402e-01
-7.78589070e-01 1.30134201e+00 5.41941941e-01 -1.04634821e-01
3.40599149e-01 -6.25244796e-01 -3.40996608e-02 1.91142946e-01
5.17214239e-01 2.36936316e-01 1.44928321e-01 -5.16264141e-01
-5.09229124e-01 8.05042028e-01 -4.42651719e-01 -5.77112213e-02
1.32428765e+00 -5.41489646e-02 -1.22903094e-01 3.10630947e-01
8.31344783e-01 -2.73369431e-01 -1.51420653e+00 -3.74241322e-01
-3.41943949e-02 -8.91224220e-02 1.20447010e-01 -1.01077366e+00
-1.10750282e+00 8.76629472e-01 8.35452080e-01 1.38074398e-01
1.16768885e+00 2.45389640e-01 7.52779603e-01 3.27085435e-01
1.92173943e-02 -1.13425720e+00 -2.16407981e-02 5.92095912e-01
8.70650411e-01 -1.17678261e+00 -1.57545298e-01 -2.97659963e-01
-8.10961485e-01 8.59754682e-01 9.48427439e-01 -1.57379374e-01
3.97474617e-01 3.46087962e-01 2.37580851e-01 -1.60487384e-01
-7.63277113e-01 -3.31442565e-01 4.46792603e-01 6.10446632e-01
3.53430510e-01 -1.18509177e-02 1.56954393e-01 2.22118467e-01
-1.19468972e-01 -2.27065384e-01 3.76151085e-01 6.84515536e-01
-4.45089489e-01 -9.14129019e-01 -3.09990823e-01 4.67537373e-01
-5.94542384e-01 -5.29290378e-01 -2.37826452e-01 9.18873131e-01
4.14308757e-01 8.09746444e-01 1.77308381e-01 -9.70844254e-02
3.08858454e-01 1.58984572e-01 4.28648651e-01 -6.09200060e-01
-8.34129214e-01 1.19973622e-01 5.92422597e-02 -5.91424763e-01
-3.01425219e-01 -5.29008687e-01 -1.19853139e+00 -1.55260831e-01
-1.70345291e-01 -2.64758348e-01 5.54132342e-01 6.73451483e-01
4.39997673e-01 5.64426482e-01 -9.40648913e-02 -7.46665359e-01
-5.01871645e-01 -8.37779522e-01 -3.92325222e-01 1.38458282e-01
4.26581621e-01 -3.53535563e-01 3.80164385e-02 3.14611673e-01] | [9.465435981750488, 1.0924276113510132] |
8e40ee2f-3042-4a1f-a677-9acb0dc60165 | distinguishable-speaker-anonymization-based | 2211.03038 | null | https://arxiv.org/abs/2211.03038v1 | https://arxiv.org/pdf/2211.03038v1.pdf | Distinguishable Speaker Anonymization based on Formant and Fundamental Frequency Scaling | Speech data on the Internet are proliferating exponentially because of the emergence of social media, and the sharing of such personal data raises obvious security and privacy concerns. One solution to mitigate these concerns involves concealing speaker identities before sharing speech data, also referred to as speaker anonymization. In our previous work, we have developed an automatic speaker verification (ASV)-model-free anonymization framework to protect speaker privacy while preserving speech intelligibility. Although the framework ranked first place in VoicePrivacy 2022 challenge, the anonymization was imperfect, since the speaker distinguishability of the anonymized speech was deteriorated. To address this issue, in this paper, we directly model the formant distribution and fundamental frequency (F0) to represent speaker identity and anonymize the source speech by the uniformly scaling formant and F0. By directly scaling the formant and F0, the speaker distinguishability degradation of the anonymized speech caused by the introduction of other speakers is prevented. The experimental results demonstrate that our proposed framework can improve the speaker distinguishability and significantly outperforms our previous framework in voice distinctiveness. Furthermore, our proposed method also can trade off the privacy-utility by using different scaling factors. | ['Jie Liu', 'Namin Wang', 'Lei Xie', 'Pengcheng Guo', 'Yi Lei', 'Qing Wang', 'Jixun Yao'] | 2022-11-06 | null | null | null | null | ['speaker-verification'] | ['speech'] | [-1.37877733e-01 2.72567123e-01 1.27920330e-01 -4.51628149e-01
-7.01691568e-01 -8.65385175e-01 4.17593986e-01 -6.60303831e-02
-2.02512965e-01 4.76860046e-01 6.65176392e-01 -3.63979489e-01
2.03077659e-01 -4.11320686e-01 -3.79822612e-01 -7.49261498e-01
1.27187788e-01 -3.62989217e-01 -7.82843456e-02 5.54486997e-02
-4.46359105e-02 5.47425985e-01 -1.53986275e+00 -3.58597599e-02
1.03662646e+00 9.02721286e-01 -2.43153960e-01 4.11960900e-01
-2.55495846e-01 1.52291268e-01 -1.05437410e+00 -6.28570437e-01
5.71957886e-01 -4.97293204e-01 -6.22572541e-01 -8.56824219e-02
8.91642720e-02 -4.52454448e-01 -6.17531955e-01 1.34457421e+00
7.08903611e-01 -1.00390352e-01 2.15843506e-02 -1.63236344e+00
-6.90854013e-01 6.62148714e-01 -3.75917494e-01 1.17737547e-01
3.85351628e-01 -4.58899587e-02 6.23291552e-01 -3.83346647e-01
1.14988960e-01 1.44105017e+00 5.37072062e-01 7.25422680e-01
-9.70484495e-01 -1.02073336e+00 1.78224236e-01 -9.33032949e-03
-1.84131777e+00 -9.21813965e-01 8.88393342e-01 -2.12374538e-01
1.91464514e-01 8.30098331e-01 5.00396729e-01 9.58115220e-01
-2.68047303e-01 5.67853987e-01 9.02779400e-01 -2.34038338e-01
1.97157383e-01 4.71747398e-01 1.34719506e-01 2.45421693e-01
1.92796454e-01 1.71385244e-01 -4.27848428e-01 -6.60042048e-01
3.83915007e-01 9.02368687e-03 -6.73549414e-01 4.02987376e-02
-1.00750518e+00 4.82374579e-01 9.70644057e-02 8.20552185e-02
-8.19455460e-02 -2.34912083e-01 5.49480796e-01 3.67485642e-01
3.29828352e-01 1.14238217e-01 -1.45794507e-02 2.35299300e-02
-7.12630987e-01 1.98030084e-01 8.41041327e-01 1.00150728e+00
5.89281201e-01 1.43915042e-01 -1.21084295e-01 8.46023440e-01
3.59298974e-01 6.16118968e-01 8.81630480e-01 -7.85945237e-01
5.54330170e-01 1.33624241e-01 2.62288392e-01 -1.25029051e+00
6.18077628e-03 -2.39689052e-01 -7.01094329e-01 -1.85665622e-01
4.79831129e-01 -3.81046236e-01 -4.71723855e-01 1.91786575e+00
4.86492544e-01 1.19308218e-01 2.57312447e-01 9.43487287e-01
5.77943146e-01 6.47373974e-01 1.78418830e-01 -3.61767799e-01
1.38901222e+00 -5.87103248e-01 -1.26217926e+00 4.71442156e-02
2.72345632e-01 -8.68855834e-01 1.07388830e+00 -4.47608493e-02
-5.95765412e-01 -3.10283154e-01 -9.96371210e-01 1.51828349e-01
-1.67771399e-01 -2.48858735e-01 3.32926869e-01 1.59623849e+00
-8.54870379e-01 2.42974952e-01 -6.91302598e-01 -2.17780456e-01
2.79099464e-01 4.02791679e-01 -5.22340477e-01 4.41158444e-01
-1.55379927e+00 2.33508632e-01 8.19891989e-02 1.56700928e-02
-4.35697436e-01 -6.97465897e-01 -7.55842030e-01 1.93811148e-01
1.25902772e-01 -2.96291530e-01 1.24905014e+00 -1.01869607e+00
-1.80777776e+00 6.82056069e-01 -2.37343550e-01 -4.57962692e-01
7.26352453e-01 -6.02660654e-03 -1.18249786e+00 -2.26128101e-01
-5.31229340e-02 2.91703671e-01 1.11467052e+00 -1.20324516e+00
-6.05858743e-01 -5.79593658e-01 -1.31337434e-01 1.97946042e-01
-8.44046175e-01 2.13489190e-01 -1.42421111e-01 -8.68326545e-01
2.60605812e-01 -8.72000754e-01 1.97593987e-01 3.18162926e-02
-5.59350729e-01 3.32429290e-01 1.18177783e+00 -1.13232183e+00
1.33391118e+00 -2.70824313e+00 -3.07988107e-01 1.54007122e-01
1.17499329e-01 4.17563617e-01 2.30846837e-01 2.78066695e-01
-1.41938299e-01 4.24673855e-01 -3.09691995e-01 -3.90227586e-01
9.79583114e-02 5.09994105e-02 -5.33605814e-01 5.99729657e-01
-2.18374193e-01 3.67833078e-01 -6.47571981e-01 -2.72197574e-01
5.81585169e-02 5.60563684e-01 -4.83407915e-01 1.19896956e-01
3.26718390e-01 3.32901478e-01 -4.29056615e-01 6.23598635e-01
1.29597962e+00 4.97922897e-01 2.08821997e-01 -1.11351991e-02
-4.54975851e-02 3.12417418e-01 -1.39822578e+00 1.21028578e+00
-2.62981445e-01 4.05036300e-01 6.63997650e-01 -4.62709159e-01
9.81766701e-01 5.68927765e-01 3.51577550e-01 -3.36793005e-01
2.65094757e-01 9.32341591e-02 1.64557043e-02 -3.33725721e-01
4.29904252e-01 -3.83462757e-02 -1.00671962e-01 4.19671744e-01
-4.68384594e-01 2.19519600e-01 -7.18053877e-01 -2.87221149e-02
6.34257257e-01 -5.25125623e-01 3.25701177e-01 -4.16226834e-01
8.51596832e-01 -7.84896195e-01 8.45380545e-01 2.52057672e-01
-8.61726224e-01 5.13173938e-01 1.45580098e-01 6.63628206e-02
-7.49202728e-01 -1.11189985e+00 -2.36820504e-01 7.10578382e-01
1.84308141e-01 -3.63678306e-01 -1.27509618e+00 -6.16999686e-01
2.64994591e-01 6.46937072e-01 -1.45169050e-01 -3.83791655e-01
-3.84537965e-01 -4.18540120e-01 1.13935876e+00 5.83430305e-02
7.77154922e-01 -4.45613533e-01 2.22758070e-01 6.86943382e-02
-3.51884812e-01 -1.05466008e+00 -1.10777617e+00 -4.40646559e-01
-3.67706418e-01 -5.85632443e-01 -7.26185381e-01 -6.42236352e-01
3.42073709e-01 4.63123828e-01 1.48647144e-01 -2.72883683e-01
1.60822701e-02 1.56371459e-01 -2.08910525e-01 -6.35565460e-01
-8.10291290e-01 5.53919859e-02 5.16676486e-01 7.27667809e-01
2.27401406e-01 -7.29254425e-01 -5.61800122e-01 4.61571544e-01
-9.00277436e-01 -4.57458079e-01 -1.38442948e-01 4.04973477e-01
1.61195531e-01 3.33232045e-01 9.38682675e-01 -6.56033695e-01
9.27410066e-01 -4.56991494e-01 -3.93526435e-01 1.81332976e-01
-4.61558908e-01 -1.60708576e-01 8.60963464e-01 -4.12369043e-01
-1.09496808e+00 6.94874451e-02 -2.67406791e-01 -2.76620120e-01
-2.27893263e-01 -1.00510772e-02 -1.03571260e+00 -2.81201810e-01
3.87683928e-01 3.35514098e-01 3.90730351e-01 -5.75058401e-01
4.90123779e-01 1.52575922e+00 6.35108948e-01 -2.73704797e-01
7.76365399e-01 4.75981891e-01 -5.21696091e-01 -1.15354359e+00
-9.23170894e-02 -4.23664063e-01 -7.92642757e-02 -8.62778910e-03
4.98693019e-01 -1.04198384e+00 -9.97799933e-01 9.14078534e-01
-1.04216182e+00 6.37458384e-01 -1.04309246e-01 4.82169479e-01
-1.98683992e-01 9.61555421e-01 -3.84871304e-01 -1.08796740e+00
-4.19927627e-01 -1.08620024e+00 7.64619529e-01 1.71871498e-01
-2.15174958e-01 -5.99948108e-01 -1.74131930e-01 4.78156388e-01
5.11692107e-01 9.63525772e-02 5.95618427e-01 -7.06214368e-01
-3.17890108e-01 -3.67839277e-01 5.80092333e-02 5.33091843e-01
8.38207006e-01 -1.93229198e-01 -1.43619883e+00 -3.88693124e-01
4.95939642e-01 4.08439666e-01 2.48258904e-01 -1.66384177e-03
1.12982547e+00 -9.70917523e-01 -1.46814555e-01 8.28941464e-01
7.53559351e-01 2.68335521e-01 6.45949960e-01 7.07232207e-02
8.32986295e-01 8.99345040e-01 1.26556799e-01 5.70787311e-01
2.48058930e-01 6.95633471e-01 2.28589788e-01 1.10725194e-01
-1.16944611e-02 -5.08200705e-01 4.35061783e-01 8.46848428e-01
4.69996959e-01 -2.45453790e-01 -6.03615701e-01 5.34480393e-01
-1.45196044e+00 -9.25309956e-01 1.42715618e-01 2.56460643e+00
8.04044068e-01 -2.89481699e-01 4.19487953e-01 4.32792336e-01
1.18340659e+00 3.20461094e-01 -4.75020289e-01 -3.66292953e-01
-1.40182748e-01 -2.66867906e-01 8.01901639e-01 7.34596193e-01
-9.30253267e-01 6.63491547e-01 6.15029478e+00 7.77292609e-01
-1.31216812e+00 6.69109523e-02 5.87067306e-01 -3.78291421e-02
-4.82395619e-01 -9.39031616e-02 -5.83982348e-01 8.20658684e-01
1.06307578e+00 -7.95921862e-01 6.22193098e-01 7.08594501e-01
4.66395766e-01 6.26457751e-01 -7.50387073e-01 1.10218668e+00
-7.11951032e-02 -8.54867578e-01 9.48637277e-02 2.86483914e-01
2.06661463e-01 -5.99338651e-01 1.62419260e-01 -7.11481646e-02
-1.18580654e-01 -7.28394330e-01 7.65476525e-01 2.08789855e-01
7.45846331e-01 -9.14477050e-01 4.53046381e-01 2.67293036e-01
-1.18565631e+00 -1.11191563e-01 -1.54032543e-01 1.28677145e-01
1.80931538e-01 5.15990019e-01 -8.19718778e-01 6.38526678e-01
5.68422437e-01 1.35620728e-01 -2.48954237e-01 7.72711873e-01
4.51100506e-02 6.33315325e-01 -4.55289572e-01 9.17174816e-02
-3.10586482e-01 -1.07615106e-01 8.79243076e-01 7.72239923e-01
4.52287793e-01 9.55100507e-02 -1.43267229e-01 7.69704759e-01
-2.83712566e-01 3.61560076e-01 -4.65514690e-01 -2.21097231e-01
1.01730216e+00 8.34592283e-01 -1.35667071e-01 1.65365756e-01
-1.45116061e-01 1.15873027e+00 -1.52913988e-01 3.76400173e-01
-7.79052079e-01 -5.73126376e-01 1.23032343e+00 4.08145070e-01
-1.81641862e-01 1.20298006e-02 -3.02887917e-01 -1.11913085e+00
4.27890271e-01 -9.61276650e-01 2.76462048e-01 -1.15266435e-01
-1.19817352e+00 5.50051153e-01 -3.53871286e-01 -1.21484733e+00
-5.95009811e-02 5.95130175e-02 -4.66419786e-01 1.15703273e+00
-1.10908401e+00 -9.11365449e-01 -3.72967087e-02 6.75513566e-01
6.21624850e-02 -2.95083314e-01 9.09205616e-01 5.58858633e-01
-6.17605448e-01 1.46306944e+00 3.22766930e-01 1.47690907e-01
6.57790363e-01 -8.18774283e-01 8.45756769e-01 8.73720586e-01
-1.39071018e-01 9.06715512e-01 7.05141187e-01 -5.35306156e-01
-1.24120557e+00 -1.04546857e+00 1.01128411e+00 -1.06629267e-01
4.46620494e-01 -6.18748605e-01 -1.17339480e+00 5.51934183e-01
2.33787969e-02 -9.57116336e-02 9.40861523e-01 -2.52798110e-01
-5.25188923e-01 -3.87162983e-01 -1.69021213e+00 5.49771488e-01
8.29015970e-01 -1.03993499e+00 -4.04107064e-01 9.58291441e-02
1.14448059e+00 -1.72386169e-01 -7.85271406e-01 -4.76819240e-02
4.68903810e-01 -9.05977368e-01 8.61064970e-01 -3.64990830e-01
-5.52280784e-01 -5.60313821e-01 -3.38533938e-01 -1.06881225e+00
-6.81536496e-02 -1.24453580e+00 3.50599289e-02 2.06745315e+00
1.75329238e-01 -1.24882531e+00 8.29970062e-01 9.71168220e-01
1.83779851e-01 8.81969631e-02 -1.21439743e+00 -1.03903568e+00
-6.63514361e-02 -1.06374025e-01 1.43389201e+00 1.14570594e+00
1.29199222e-01 -1.96771845e-01 -6.23559713e-01 7.29062259e-01
6.50337100e-01 -3.32238257e-01 7.64432371e-01 -9.52402115e-01
-6.72131702e-02 -1.89497411e-01 -5.77449679e-01 -8.29813540e-01
2.62426883e-01 -7.61587679e-01 -2.39413604e-01 -6.47944868e-01
-2.39279717e-01 -2.81503230e-01 -1.96944430e-01 1.13593452e-01
-3.19200486e-01 -5.01188874e-01 1.51974902e-01 2.57403404e-01
3.06630105e-01 7.71534920e-01 8.41302216e-01 -2.92925630e-03
-3.89849156e-01 3.89509410e-01 -8.93832505e-01 4.50757444e-01
8.48103821e-01 -4.80253071e-01 -5.63583732e-01 -2.41164148e-01
-6.19450390e-01 1.45589769e-01 2.85123140e-01 -1.02487111e+00
1.71913356e-01 -8.35066885e-02 -2.28684098e-01 -8.85080323e-02
2.20577449e-01 -1.07474542e+00 3.63032520e-01 3.38713109e-01
-2.75075585e-01 -9.53673050e-02 3.44196945e-01 8.57037604e-01
-3.44249159e-01 1.25702441e-01 7.76281655e-01 3.04854989e-01
-3.47071350e-01 3.58341336e-01 -4.40471351e-01 -1.27143621e-01
1.01140451e+00 -1.44403368e-01 -3.15358430e-01 -6.12673342e-01
-5.66803575e-01 -2.43302397e-02 6.75178409e-01 6.95277214e-01
3.21769118e-01 -1.37242770e+00 -6.33022726e-01 6.15038335e-01
1.27576113e-01 -4.19288516e-01 5.75858414e-01 1.56502470e-01
-3.76412839e-01 2.04681307e-01 -6.84942156e-02 -3.01030010e-01
-1.48336995e+00 6.92708313e-01 5.26937664e-01 4.34108138e-01
-4.01254147e-01 7.25925982e-01 1.94031313e-01 -6.58697844e-01
4.13363427e-01 -1.17325135e-01 1.74556509e-01 -1.14021130e-01
8.27414930e-01 5.46672940e-01 1.94170773e-01 -9.46449518e-01
-5.52331090e-01 1.08080819e-01 -1.31660745e-01 -2.70229340e-01
7.33474851e-01 -7.21811295e-01 -2.66342834e-02 2.58183051e-02
1.48827136e+00 5.36565006e-01 -9.28474844e-01 -1.44167185e-01
-2.93354720e-01 -8.95999134e-01 3.21349390e-02 -4.09725368e-01
-1.04907668e+00 5.67370057e-01 8.18469286e-01 6.31489217e-01
9.65632498e-01 -4.82794255e-01 1.08383429e+00 -3.33720446e-02
3.52384925e-01 -9.17239130e-01 -8.11583102e-01 1.23485766e-01
6.28407180e-01 -1.06638646e+00 -3.61450970e-01 -8.34102988e-01
-6.96992338e-01 6.09271467e-01 2.39126608e-01 4.48757350e-01
9.05983865e-01 1.62389472e-01 4.60513055e-01 3.60127777e-01
7.51156406e-03 4.30165589e-01 1.27186382e-03 8.25714648e-01
1.93211630e-01 4.21784878e-01 -3.36121827e-01 1.05117190e+00
-8.44459295e-01 -4.11614716e-01 4.33230132e-01 8.52999628e-01
-2.11600393e-01 -1.24124730e+00 -7.49163449e-01 5.70690520e-02
-6.02606177e-01 -3.90821137e-02 -6.37256742e-01 3.10113430e-01
-1.15300611e-01 1.31078458e+00 -1.19686775e-01 -6.11696422e-01
4.45093930e-01 1.89589366e-01 -1.67680845e-01 -2.91993201e-01
-6.31727576e-01 -1.76074281e-01 7.99842998e-02 -3.12858939e-01
-7.87594691e-02 -7.32157111e-01 -1.29978383e+00 -9.25453961e-01
-3.40573758e-01 3.90894085e-01 8.17985773e-01 6.85228348e-01
6.44090414e-01 1.95949346e-01 1.26024628e+00 -1.48400620e-01
-9.51297820e-01 -5.72959900e-01 -1.02557635e+00 4.56037343e-01
7.17333972e-01 -1.90505490e-01 -8.06396246e-01 -5.18721230e-02] | [13.994216918945312, 5.85652494430542] |
06165e74-895d-4a8d-b5e2-6e85f470d64b | learning-a-discriminative-feature-network-for | 1804.09337 | null | http://arxiv.org/abs/1804.09337v1 | http://arxiv.org/pdf/1804.09337v1.pdf | Learning a Discriminative Feature Network for Semantic Segmentation | Most existing methods of semantic segmentation still suffer from two aspects
of challenges: intra-class inconsistency and inter-class indistinction. To
tackle these two problems, we propose a Discriminative Feature Network (DFN),
which contains two sub-networks: Smooth Network and Border Network.
Specifically, to handle the intra-class inconsistency problem, we specially
design a Smooth Network with Channel Attention Block and global average pooling
to select the more discriminative features. Furthermore, we propose a Border
Network to make the bilateral features of boundary distinguishable with deep
semantic boundary supervision. Based on our proposed DFN, we achieve
state-of-the-art performance 86.2% mean IOU on PASCAL VOC 2012 and 80.3% mean
IOU on Cityscapes dataset. | ['Jingbo Wang', 'Changqian Yu', 'Nong Sang', 'Gang Yu', 'Chao Peng', 'Changxin Gao'] | 2018-04-25 | learning-a-discriminative-feature-network-for-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Yu_Learning_a_Discriminative_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Yu_Learning_a_Discriminative_CVPR_2018_paper.pdf | cvpr-2018-6 | ['thermal-image-segmentation'] | ['computer-vision'] | [ 1.25620455e-01 -5.06215729e-02 -1.76082537e-01 -6.70672417e-01
-7.01956928e-01 -4.04532880e-01 3.79977971e-01 -3.02523881e-01
-4.93220150e-01 4.14060324e-01 -2.56354120e-02 -1.88957229e-02
2.76150201e-02 -6.79562747e-01 -6.10266447e-01 -4.79493022e-01
2.83212751e-01 -3.62688070e-03 8.51694465e-01 2.51476150e-02
2.61804461e-01 2.95582414e-01 -1.39526951e+00 4.39703017e-01
1.21255136e+00 1.40292001e+00 9.23904702e-02 1.79989114e-01
-3.28365862e-01 2.82060444e-01 -5.84288239e-01 -1.03082806e-01
2.97698319e-01 -2.86208540e-01 -9.92641568e-01 1.52469724e-01
5.96469462e-01 -3.05769622e-01 -2.56705225e-01 1.27742898e+00
4.34852511e-01 1.99522510e-01 5.97833812e-01 -1.23170733e+00
-5.54661930e-01 3.43406439e-01 -8.10199022e-01 3.64621401e-01
-1.62625872e-02 1.15741804e-01 1.04619753e+00 -7.01953650e-01
3.97340745e-01 1.37027276e+00 7.16803730e-01 4.47506189e-01
-9.84157801e-01 -7.15307534e-01 5.53446710e-01 3.23403001e-01
-1.46549928e+00 -1.84592724e-01 6.51515722e-01 -3.60152155e-01
6.61110699e-01 1.17148027e-01 4.75287497e-01 6.93442047e-01
-1.22571662e-01 1.15912783e+00 1.04766226e+00 3.49595137e-02
3.54155749e-02 -1.61017179e-01 2.34283790e-01 6.43081069e-01
1.97281152e-01 -1.03520155e-01 -2.55280100e-02 3.67586344e-01
8.74752283e-01 7.45762885e-02 -3.10657829e-01 -6.27905354e-02
-7.26006448e-01 6.64249957e-01 7.70365536e-01 2.70983398e-01
1.74816087e-01 1.25736073e-01 3.86602223e-01 5.56616522e-02
3.51713151e-01 5.17742559e-02 -6.62539124e-01 4.49796468e-02
-7.74033308e-01 8.49878043e-02 3.54010046e-01 9.07102287e-01
8.47687960e-01 -1.74855620e-01 -3.01753938e-01 1.06256568e+00
4.58066672e-01 3.17798108e-01 4.42138135e-01 -9.67232823e-01
7.76776075e-01 7.86538959e-01 -1.97974682e-01 -1.04856086e+00
-5.32961845e-01 -4.57109928e-01 -7.06600130e-01 6.87643960e-02
3.42826337e-01 9.00086481e-03 -1.38788545e+00 1.58126354e+00
3.24253440e-01 3.00850630e-01 -1.42345056e-01 1.07447731e+00
1.18070662e+00 5.50299108e-01 1.00109592e-01 2.35923618e-01
1.33457816e+00 -1.53366911e+00 -5.88884652e-01 -3.61293644e-01
5.72763383e-01 -5.87829888e-01 1.07208323e+00 2.15745956e-01
-8.64169836e-01 -8.21836174e-01 -1.06040335e+00 -3.63265932e-01
-2.92901456e-01 3.00834507e-01 4.78912145e-01 4.14481461e-01
-8.46373439e-01 5.04128575e-01 -7.83094645e-01 -8.05159472e-03
7.70510793e-01 5.80589056e-01 -2.66669095e-01 -1.15019068e-01
-1.09332561e+00 2.54977793e-01 4.48256999e-01 3.56608301e-01
-5.68087816e-01 -5.13942301e-01 -1.06761372e+00 1.99258223e-01
3.92410308e-01 -4.45030272e-01 1.10818446e+00 -1.23053181e+00
-1.48872352e+00 9.27624881e-01 -3.79177183e-02 -1.42266676e-01
5.01987338e-01 -2.96753615e-01 -5.01292825e-01 1.86330721e-01
2.59379536e-01 9.01300013e-01 4.64163840e-01 -1.13905859e+00
-9.35760915e-01 -4.39401209e-01 -2.28292793e-01 1.51087970e-01
-2.38299578e-01 -2.02095419e-01 -1.00500619e+00 -8.01682770e-01
4.82212603e-01 -6.44464076e-01 -2.39707053e-01 3.56613658e-02
-6.52306616e-01 -4.90489721e-01 1.25657809e+00 -7.98632443e-01
1.28415418e+00 -2.38228488e+00 -2.16796979e-01 8.40691626e-02
2.27882534e-01 6.31564200e-01 -3.95226866e-01 -4.20014679e-01
1.04849704e-01 2.71588922e-01 -3.62037808e-01 -3.08966190e-01
-5.97884431e-02 1.53808847e-01 -4.88630645e-02 3.45880419e-01
3.02563369e-01 8.72901618e-01 -7.05887973e-01 -6.08460903e-01
1.79134637e-01 4.69547778e-01 -5.27321696e-01 1.16613090e-01
-1.03107229e-01 4.59085047e-01 -7.85287619e-01 7.13322580e-01
1.11484909e+00 -9.45957974e-02 -2.14788243e-01 -2.93428183e-01
-1.47829488e-01 2.93547183e-01 -1.38295031e+00 1.80105984e+00
-1.82412580e-01 4.40951049e-01 2.60471284e-01 -9.34060633e-01
9.88523424e-01 -5.60613163e-02 1.54688060e-01 -9.59355354e-01
4.51886863e-01 4.18519408e-01 -8.96189660e-02 -5.10620236e-01
3.47475678e-01 1.97496861e-01 5.39183058e-02 -1.18482612e-01
3.49776484e-02 -4.29275259e-02 -1.17962971e-01 -1.60311922e-01
9.72495317e-01 7.13801235e-02 -3.00780416e-01 -3.87308598e-01
6.72547400e-01 -5.05779564e-01 1.27504802e+00 4.38021958e-01
-6.36245847e-01 1.15286207e+00 7.73998916e-01 -6.21219873e-01
-5.92107594e-01 -1.05221617e+00 -2.82390952e-01 7.39612460e-01
6.61215186e-01 -1.48419917e-01 -9.70638931e-01 -1.10595226e+00
-8.95505771e-02 2.30367482e-01 -6.43821895e-01 -2.55842246e-02
-5.91030896e-01 -5.51159799e-01 4.63235497e-01 9.82399762e-01
1.25920165e+00 -9.78649020e-01 -3.18603009e-01 7.04156458e-02
-1.15207523e-01 -1.37991023e+00 -7.00877666e-01 2.53380887e-04
-7.23744273e-01 -1.08011758e+00 -7.82351494e-01 -1.06213582e+00
6.35826528e-01 3.02627921e-01 8.93829525e-01 3.25039439e-02
-1.88095570e-01 -2.51904547e-01 -3.43095720e-01 -4.44227643e-02
3.43544662e-01 4.28002089e-01 -5.17168999e-01 2.49564461e-02
2.64001191e-01 -4.20555502e-01 -9.78807926e-01 6.20882511e-01
-9.11126256e-01 9.97902602e-02 4.16635990e-01 7.64212608e-01
9.23584998e-01 8.39440078e-02 4.99741584e-01 -8.72089505e-01
1.31064087e-01 -2.07630247e-01 -6.53615057e-01 1.85453907e-01
-4.33391631e-01 -1.61915123e-01 6.24214590e-01 -2.57130321e-02
-1.05220568e+00 1.29602462e-01 -4.06734109e-01 -2.29637995e-01
-1.89141318e-01 2.34338250e-02 -7.24285603e-01 -4.21212129e-02
8.26399624e-02 -4.72471975e-02 -3.03515792e-01 -6.42415404e-01
2.59069175e-01 1.05778122e+00 6.98787868e-01 -4.26049680e-01
4.13603753e-01 5.44294178e-01 -3.62650543e-01 -4.95354980e-01
-1.13409352e+00 -6.01583362e-01 -5.95539629e-01 1.38947651e-01
1.33143783e+00 -1.06148469e+00 -5.03294528e-01 9.98105407e-01
-1.06988084e+00 -5.21821976e-01 -3.33664864e-02 2.76450753e-01
-3.46911639e-01 3.78542304e-01 -6.98955059e-01 -2.04500884e-01
-5.01400828e-01 -1.41326761e+00 1.19987750e+00 9.05252337e-01
2.15626627e-01 -7.64511228e-01 -2.77492881e-01 5.59976637e-01
3.89388382e-01 2.21808761e-01 6.01015091e-01 -4.74010468e-01
-4.68008816e-01 -5.91271445e-02 -9.25467074e-01 6.44485295e-01
2.58332074e-01 3.85116437e-03 -1.08453286e+00 -1.61769629e-01
-3.47835183e-01 -1.71165720e-01 1.30392313e+00 4.26594734e-01
1.59561479e+00 1.23545773e-01 -4.82637227e-01 1.10223186e+00
1.31649399e+00 1.91717118e-01 8.16826582e-01 4.60909247e-01
1.09520519e+00 4.47522610e-01 5.83337784e-01 2.25096643e-01
5.26163578e-01 6.84933126e-01 3.70797157e-01 -3.33530366e-01
-3.11424494e-01 -1.27980694e-01 -6.35573491e-02 5.38647115e-01
1.40818134e-01 -4.39943634e-02 -7.82831371e-01 7.02880740e-01
-1.86037469e+00 -4.61628318e-01 -1.63903594e-01 1.93661976e+00
6.65373981e-01 4.23976392e-01 3.99461798e-02 -1.77104995e-01
9.33115184e-01 2.33141914e-01 -7.26700723e-01 -3.13714087e-01
-2.31662259e-01 1.53688654e-01 4.96769100e-01 4.14719582e-01
-1.56146693e+00 1.20019853e+00 5.76075554e+00 1.23337269e+00
-1.28557718e+00 9.42426994e-02 1.14434254e+00 2.57897675e-01
-4.05060768e-01 -7.52817243e-02 -1.00173295e+00 6.49818361e-01
3.44366908e-01 5.37748575e-01 -2.77285390e-02 8.11739981e-01
1.47655010e-02 -1.80391952e-01 -5.82361817e-01 9.90123391e-01
-6.13913983e-02 -1.08195591e+00 -9.28242058e-02 -1.56013608e-01
9.73230541e-01 1.56973436e-01 -1.47577226e-02 3.12691867e-01
1.36939228e-01 -1.13430631e+00 5.56159258e-01 1.42284125e-01
8.35499942e-01 -8.72111142e-01 9.14723575e-01 -6.19400246e-03
-1.49959600e+00 -3.19639817e-02 -3.11858743e-01 9.62151587e-02
1.73668459e-01 6.97758079e-01 1.27149105e-01 5.53387940e-01
9.93624926e-01 9.21890676e-01 -5.66940725e-01 1.22786224e+00
-4.91140783e-01 4.48794365e-01 -3.32187146e-01 3.68758947e-01
6.64170921e-01 -2.39583194e-01 6.62453473e-02 1.17311180e+00
5.85633442e-02 -1.03321290e-02 3.44167113e-01 8.51963758e-01
-1.61620617e-01 -2.37613693e-02 -1.32629380e-01 3.63469124e-01
4.11784768e-01 1.41585219e+00 -1.03396022e+00 -3.55329752e-01
-4.55479324e-01 1.26520526e+00 3.25607270e-01 2.85114318e-01
-8.79366815e-01 -9.05821681e-01 8.24799120e-01 -1.23089433e-01
7.18142927e-01 -9.37702954e-02 -5.08413136e-01 -1.13092148e+00
2.35903874e-01 -6.30678058e-01 4.68574822e-01 -5.06046176e-01
-1.08507621e+00 5.59088290e-01 -3.34048092e-01 -9.78023469e-01
5.82798421e-01 -7.02793300e-01 -9.22935605e-01 7.36690938e-01
-1.80557823e+00 -1.13560355e+00 -5.12950778e-01 4.17979360e-01
5.88736892e-01 2.05685943e-01 3.00268799e-01 5.68799615e-01
-9.75134492e-01 8.72885346e-01 5.17905280e-02 4.52803582e-01
5.55290699e-01 -1.22599852e+00 5.85142553e-01 6.79609299e-01
-2.42526233e-01 2.99592942e-01 9.12499204e-02 -4.83960032e-01
-6.55648768e-01 -1.48771846e+00 6.58111274e-01 2.00520065e-02
2.91350484e-01 -2.93403924e-01 -9.54341948e-01 5.09093881e-01
-2.07813010e-02 5.63170969e-01 2.65179932e-01 -7.68083632e-02
-4.98630673e-01 -2.75422454e-01 -1.19211626e+00 4.25306827e-01
1.22535169e+00 -2.83711076e-01 -4.32404310e-01 1.65041238e-01
9.94246840e-01 -5.16012430e-01 -6.30712211e-01 7.43003190e-01
4.06190872e-01 -1.04859281e+00 8.39153528e-01 -2.06295848e-01
3.65434021e-01 -4.80687141e-01 -1.08808346e-01 -1.06725454e+00
-2.31104955e-01 -3.59006912e-01 4.01471019e-01 1.49805880e+00
3.51158142e-01 -6.87017143e-01 9.12029922e-01 3.69123667e-01
-5.27521908e-01 -1.06133306e+00 -1.06332695e+00 -8.49857688e-01
2.04761133e-01 -3.64343315e-01 6.39245689e-01 8.36569369e-01
-4.77926284e-01 3.36462528e-01 -3.78427729e-02 7.34233782e-02
2.71183401e-01 1.81936815e-01 5.56855261e-01 -1.06607103e+00
-1.23765029e-01 -6.49268091e-01 -5.72727501e-01 -1.45571578e+00
1.79149508e-01 -6.29909635e-01 2.24958524e-01 -1.59907925e+00
2.72161156e-01 -5.94716907e-01 -4.78275299e-01 5.25037885e-01
-3.50478977e-01 3.38094592e-01 1.41221449e-01 2.96257138e-02
-9.72626626e-01 7.59921610e-01 1.55076146e+00 -1.98595613e-01
-3.70772719e-01 -9.49648768e-02 -6.50994837e-01 9.26192462e-01
8.24054837e-01 -4.03718740e-01 -2.67906636e-01 -7.95533597e-01
-1.71592608e-01 -3.22164983e-01 1.92338452e-01 -1.12053120e+00
-6.58611804e-02 -1.52349681e-01 3.94116968e-01 -6.63907707e-01
2.80744787e-02 -5.93902111e-01 -4.85289097e-01 3.37479532e-01
-8.51190239e-02 -1.42447904e-01 2.51636416e-01 4.83605772e-01
-5.99996924e-01 -7.51215890e-02 8.93100858e-01 7.64756203e-02
-1.07484698e+00 5.04574180e-01 4.42969007e-03 4.59338784e-01
1.07463515e+00 -2.76645184e-01 -5.52728236e-01 1.27598867e-01
-5.43314278e-01 7.71192908e-01 2.48120293e-01 4.28813577e-01
4.82277066e-01 -1.32083070e+00 -4.00574446e-01 3.54618907e-01
-8.54412280e-03 6.07604265e-01 5.91408253e-01 8.57798815e-01
-6.60426080e-01 2.58144140e-01 -1.39619842e-01 -5.54349005e-01
-9.93066192e-01 9.63565037e-02 5.13135910e-01 -1.91699281e-01
-6.75944984e-01 1.22595704e+00 5.50808668e-01 -5.58458567e-01
2.12514058e-01 -4.15867746e-01 -3.41611803e-01 -4.38454337e-02
4.69307035e-01 3.03368419e-01 1.98195514e-04 -6.73376977e-01
-5.29588103e-01 9.97094810e-01 -2.39441365e-01 2.16480881e-01
1.05892062e+00 -2.11641490e-01 -2.17399690e-02 8.94878060e-02
1.48088884e+00 -2.37291038e-01 -1.70074642e+00 -1.55405462e-01
-9.48383585e-02 -3.57977539e-01 1.94557220e-01 -7.70977437e-01
-1.57181501e+00 1.10494852e+00 7.79725432e-01 -4.36252467e-02
1.26914692e+00 -1.83526501e-02 1.31257355e+00 -1.25662303e-02
-6.74174577e-02 -1.46367657e+00 -3.61067764e-02 7.32760370e-01
5.54734647e-01 -1.41476417e+00 -1.93772733e-01 -8.92926872e-01
-4.78425860e-01 9.99355435e-01 1.01953685e+00 -3.22172850e-01
8.88130963e-01 1.00850478e-01 3.05596888e-01 -1.88713759e-01
-2.16799006e-01 -4.60010678e-01 4.99510020e-01 3.80175650e-01
3.25463414e-01 9.80378240e-02 -5.71990371e-01 9.74318683e-01
1.16075367e-01 -8.36222768e-02 1.17177837e-01 6.92932367e-01
-6.50232017e-01 -9.61927533e-01 -6.16692454e-02 2.52684653e-01
-5.05337417e-01 -6.35609205e-04 -2.07189769e-01 7.81514525e-01
4.90151584e-01 8.66471052e-01 4.18096721e-01 -6.35234416e-01
4.80780423e-01 -2.88274676e-01 1.23052821e-01 -5.38151026e-01
-5.57890177e-01 2.13153884e-01 -1.13415271e-01 -7.89557517e-01
-3.28106761e-01 -3.96812499e-01 -1.61059535e+00 1.98370200e-02
-4.73592252e-01 -1.17116176e-01 3.76800895e-01 1.04958928e+00
4.82018918e-01 6.96837485e-01 5.89523017e-01 -5.82772911e-01
-2.04564005e-01 -9.51946318e-01 -6.78676546e-01 4.63257641e-01
1.47687763e-01 -6.13101900e-01 -2.82473564e-01 -1.96012512e-01] | [9.54922103881836, 0.1640278846025467] |
713a3986-52ee-4065-aa63-9a870494aad0 | variational-disentangled-graph-auto-encoders | 2306.11315 | null | https://arxiv.org/abs/2306.11315v1 | https://arxiv.org/pdf/2306.11315v1.pdf | Variational Disentangled Graph Auto-Encoders for Link Prediction | With the explosion of graph-structured data, link prediction has emerged as an increasingly important task. Embedding methods for link prediction utilize neural networks to generate node embeddings, which are subsequently employed to predict links between nodes. However, the existing embedding methods typically take a holistic strategy to learn node embeddings and ignore the entanglement of latent factors. As a result, entangled embeddings fail to effectively capture the underlying information and are vulnerable to irrelevant information, leading to unconvincing and uninterpretable link prediction results. To address these challenges, this paper proposes a novel framework with two variants, the disentangled graph auto-encoder (DGAE) and the variational disentangled graph auto-encoder (VDGAE). Our work provides a pioneering effort to apply the disentanglement strategy to link prediction. The proposed framework infers the latent factors that cause edges in the graph and disentangles the representation into multiple channels corresponding to unique latent factors, which contributes to improving the performance of link prediction. To further encourage the embeddings to capture mutually exclusive latent factors, we introduce mutual information regularization to enhance the independence among different channels. Extensive experiments on various real-world benchmarks demonstrate that our proposed methods achieve state-of-the-art results compared to a variety of strong baselines on link prediction tasks. Qualitative analysis on the synthetic dataset also illustrates that the proposed methods can capture distinct latent factors that cause links, providing empirical evidence that our models are able to explain the results of link prediction to some extent. All code will be made publicly available upon publication of the paper. | ['Dali Chen', 'Shuang Li', 'Xiaojuan Zhang', 'Jun Fu'] | 2023-06-20 | null | null | null | null | ['link-prediction', 'disentanglement'] | ['graphs', 'methodology'] | [ 2.64224820e-02 4.08800036e-01 -8.14530671e-01 2.18138602e-02
-3.42043079e-02 -5.43197989e-01 8.03587496e-01 5.69841824e-02
2.95200378e-01 6.77269220e-01 5.53679347e-01 -4.38200742e-01
-3.73643547e-01 -1.05821145e+00 -5.45829177e-01 -6.45728648e-01
-5.07532537e-01 2.78066158e-01 7.09235435e-03 -1.56325519e-01
-2.23253891e-01 1.82532549e-01 -9.33990121e-01 -1.52652904e-01
8.27627301e-01 5.45801580e-01 -1.68272704e-01 4.84531522e-01
-4.32898141e-02 6.56583786e-01 9.88563597e-02 -9.03855801e-01
2.41627961e-01 -2.72789806e-01 -4.28853840e-01 -3.25374573e-01
3.52613479e-02 -1.94038048e-01 -1.11815298e+00 1.03733754e+00
1.93733796e-01 -3.60002577e-01 7.09491551e-01 -1.58217859e+00
-1.03367472e+00 9.49805796e-01 -4.48085666e-01 1.56712532e-01
2.61402845e-01 -1.79169014e-01 1.95483184e+00 -8.01860332e-01
6.02623582e-01 1.04810262e+00 6.34378612e-01 4.21519816e-01
-1.61496699e+00 -8.06357145e-01 2.78077573e-02 3.29527766e-01
-1.37345290e+00 -2.86340952e-01 1.27910745e+00 -4.73698795e-01
6.67739034e-01 2.29019091e-01 7.39162982e-01 1.26042211e+00
3.49103689e-01 6.67136550e-01 7.65632212e-01 -1.02651581e-01
-2.36029714e-01 4.19051982e-02 6.50620610e-02 9.41012800e-01
8.37486386e-01 4.06954259e-01 -6.64435208e-01 -4.17321742e-01
6.94465578e-01 2.13338867e-01 -5.04623532e-01 -7.43042946e-01
-1.25427234e+00 1.11338735e+00 7.88205802e-01 1.45941647e-02
-2.89196134e-01 3.43071848e-01 4.12163109e-01 2.16832817e-01
4.92818952e-01 1.79067045e-01 -1.76534653e-01 1.00009784e-01
-3.80859315e-01 -3.17687658e-03 8.90552163e-01 7.40893722e-01
8.40222895e-01 -2.65280515e-01 -5.79935238e-02 3.67300540e-01
8.21416974e-01 2.33525306e-01 -1.47161022e-01 -5.54971159e-01
7.04676569e-01 8.27790737e-01 -2.46710107e-01 -1.54740012e+00
-1.09533325e-01 -7.09177554e-01 -1.03175771e+00 -2.25654721e-01
-3.18445340e-02 5.33856824e-02 -4.42031771e-01 1.93658829e+00
1.13438316e-01 5.03033698e-01 -2.20105778e-02 7.25889206e-01
8.28641713e-01 6.20885491e-01 -4.79587503e-02 -8.32493529e-02
1.08390033e+00 -1.00005949e+00 -8.90015602e-01 -6.68426156e-02
5.19226670e-01 -4.80548024e-01 5.35299361e-01 -1.92923054e-01
-7.38071561e-01 -2.52544761e-01 -1.28162634e+00 4.56790477e-02
-1.55071214e-01 -1.78066552e-01 1.29983723e+00 4.41804707e-01
-8.46579611e-01 6.45056367e-01 -9.23676491e-01 -1.02853790e-01
5.44275939e-01 3.59005332e-01 -6.83765769e-01 -1.84946522e-01
-1.51337647e+00 5.51205456e-01 2.63366222e-01 2.55372196e-01
-6.64133072e-01 -5.72400391e-01 -9.24722373e-01 3.29422951e-01
3.95630032e-01 -8.10711861e-01 3.42574298e-01 -3.74598563e-01
-1.18730485e+00 3.43690425e-01 -1.69030449e-03 -2.60940224e-01
1.90522730e-01 1.29933795e-02 -5.82693875e-01 2.13171527e-01
1.03992559e-01 3.75849396e-01 6.41611338e-01 -1.30902719e+00
-6.28099367e-02 -6.98197037e-02 1.67239502e-01 -2.52101980e-02
-7.40975797e-01 -4.33744550e-01 -5.57615042e-01 -6.15622103e-01
1.59732178e-01 -8.99496019e-01 3.40710878e-02 1.71371296e-01
-8.35238874e-01 -2.52443939e-01 6.74551249e-01 -4.06131685e-01
1.54501510e+00 -1.85761845e+00 7.44168162e-01 2.47489691e-01
1.02986729e+00 6.11482039e-02 -3.82554173e-01 8.67903829e-01
-2.55458385e-01 4.78890270e-01 -2.65105558e-03 -4.94439840e-01
2.30324686e-01 5.12070000e-01 -2.63834715e-01 4.98202682e-01
2.94934094e-01 1.19557750e+00 -1.01864338e+00 -4.86805469e-01
3.12523209e-02 7.54919410e-01 -7.11583734e-01 2.62964547e-01
5.19835278e-02 3.26019883e-01 -6.31432891e-01 3.96129578e-01
5.83742201e-01 -5.99231005e-01 5.36646008e-01 -5.80202758e-01
3.03103894e-01 4.95720714e-01 -8.64853203e-01 1.59225762e+00
-1.70042232e-01 6.10564888e-01 -1.47487283e-01 -1.03695345e+00
8.18507552e-01 4.01552320e-01 7.01266944e-01 -3.63669962e-01
-1.23128910e-02 3.59484069e-02 2.89459139e-01 -3.79459649e-01
2.50134200e-01 -1.76760796e-02 6.59435913e-02 4.85275358e-01
2.36643374e-01 6.18624628e-01 1.02243520e-01 7.80014694e-01
1.42281544e+00 -7.33734667e-03 2.62462109e-01 -4.90532778e-02
3.66318434e-01 -4.26479071e-01 6.68128252e-01 3.41523200e-01
-2.08013877e-01 1.05589546e-01 9.69596028e-01 -3.47519934e-01
-1.03562033e+00 -1.36968267e+00 -5.22292592e-03 4.11361694e-01
2.62490422e-01 -9.02606726e-01 -9.02164280e-02 -8.06503117e-01
1.89410046e-01 3.24754953e-01 -7.67359972e-01 -5.73138058e-01
-2.59567827e-01 -7.47226238e-01 5.55470526e-01 4.49488074e-01
7.03422353e-02 -4.66891199e-01 3.32184255e-01 5.14956824e-02
-1.80113375e-01 -1.18131530e+00 -3.36563736e-01 7.93107301e-02
-7.38366067e-01 -1.22482228e+00 -2.62413323e-01 -4.62398350e-01
7.59654284e-01 3.84983987e-01 1.03969169e+00 3.23775321e-01
-9.02766809e-02 1.82108879e-01 -3.28327805e-01 3.45127791e-01
-2.80015230e-01 2.37481073e-01 1.31053656e-01 5.51001681e-03
3.15325767e-01 -9.75197673e-01 -7.92698026e-01 1.39063850e-01
-6.92290127e-01 2.74670333e-01 6.67734385e-01 1.02095318e+00
2.34297916e-01 5.12931496e-02 5.69662690e-01 -9.48564351e-01
7.67149746e-01 -9.54801738e-01 -2.35055432e-01 1.85577616e-01
-1.05662692e+00 4.05724913e-01 7.30141282e-01 -2.57102251e-01
-5.55200160e-01 -3.57629031e-01 2.40113989e-01 -5.45442581e-01
5.07555902e-01 8.88803840e-01 -3.23466688e-01 -1.81578383e-01
1.04852110e-01 5.75841451e-03 3.55508327e-02 -2.84934700e-01
6.98402703e-01 2.38832131e-01 1.20495997e-01 -4.25128490e-01
1.31106806e+00 3.74548584e-01 3.93143594e-01 -2.07947403e-01
-5.58017373e-01 -2.18323782e-01 -5.89652777e-01 -1.40184939e-01
6.85174406e-01 -1.07699203e+00 -7.29022324e-01 -2.11555079e-01
-1.21049166e+00 2.13212833e-01 9.73570496e-02 5.47007859e-01
-1.05048388e-01 5.74861825e-01 -7.94217825e-01 -5.29209614e-01
-1.66704997e-01 -1.01815081e+00 6.43975496e-01 6.60715550e-02
-8.34610313e-02 -1.51623476e+00 4.17786419e-01 3.63126218e-01
1.73574865e-01 2.36034051e-01 1.22536957e+00 -5.37984490e-01
-9.36856806e-01 -5.02907157e-01 -5.09433746e-01 3.66010629e-02
2.28842750e-01 4.30541039e-02 -6.13836706e-01 -1.88938677e-01
-7.82279432e-01 -1.08791888e-01 9.95941460e-01 -1.67276204e-01
6.21580005e-01 -4.22117800e-01 -7.39115655e-01 5.53133905e-01
1.41795266e+00 -4.27746505e-01 5.96474051e-01 -1.54277027e-01
1.25647140e+00 2.92442322e-01 3.31565412e-03 2.75563389e-01
8.31364095e-01 7.58740306e-01 7.70268261e-01 1.00283913e-01
-1.44045815e-01 -9.31293786e-01 3.06780964e-01 1.41556823e+00
-2.65343189e-01 -4.89525765e-01 -7.10890770e-01 4.63900000e-01
-1.96331382e+00 -9.47153032e-01 -4.99642760e-01 1.91896868e+00
6.86562538e-01 2.31153622e-01 -1.19970575e-01 9.27107334e-02
6.81343257e-01 6.48001432e-01 -2.86472142e-01 -1.49644777e-01
1.14009470e-01 9.42936316e-02 4.89147156e-01 6.81129456e-01
-8.35676968e-01 7.01348603e-01 5.88805532e+00 5.24187863e-01
-8.18101048e-01 2.13675499e-01 3.29886004e-02 2.90491492e-01
-1.05589187e+00 4.75323945e-01 -4.23546314e-01 4.89964217e-01
6.88531160e-01 -2.35090122e-01 5.68147182e-01 4.57781523e-01
1.42348604e-02 6.08007669e-01 -1.16170704e+00 6.48800790e-01
-1.40519559e-01 -1.54361439e+00 2.12510973e-01 4.43104833e-01
7.56055534e-01 6.14269786e-02 -1.80799030e-02 2.19002828e-01
3.28158528e-01 -1.05837679e+00 1.83782488e-01 6.61202431e-01
5.77485502e-01 -6.53967798e-01 7.51914561e-01 8.79509076e-02
-1.49118745e+00 5.77431098e-02 -4.01972622e-01 -3.62435073e-01
2.19682470e-01 7.44528770e-01 -7.07770884e-01 1.07346940e+00
6.37391433e-02 1.23223746e+00 -5.25257111e-01 7.06860065e-01
-5.75888813e-01 7.37621367e-01 -7.36234412e-02 -1.08450286e-01
-3.09639703e-02 -3.02295178e-01 7.88512349e-01 6.69724107e-01
6.45978898e-02 -1.69934213e-01 1.26295000e-01 1.19236767e+00
-4.19864714e-01 -8.06842744e-02 -8.35577071e-01 -5.85252881e-01
6.77658439e-01 1.40907550e+00 -5.30735075e-01 1.51574314e-01
-7.65095949e-01 9.66617227e-01 7.45494545e-01 5.92638016e-01
-8.47849548e-01 -2.27654710e-01 7.60948062e-01 1.91633729e-03
3.33960801e-01 -4.79266226e-01 -2.39300922e-01 -1.48410702e+00
-1.73426280e-03 -2.94769585e-01 1.16060100e-01 -2.59628117e-01
-1.60897124e+00 5.06721854e-01 -1.40557319e-01 -1.25911820e+00
1.63360965e-02 -5.38307428e-01 -8.65867198e-01 8.51927936e-01
-1.69736814e+00 -1.32530284e+00 -1.15166731e-01 2.35306844e-01
-1.77821323e-01 -1.71013236e-01 9.91126418e-01 5.90732992e-01
-8.24894547e-01 7.14761734e-01 2.00461000e-01 2.92863458e-01
4.15906310e-01 -1.10585535e+00 2.20756501e-01 8.08106840e-01
5.61139464e-01 8.18224251e-01 5.95059454e-01 -7.98549712e-01
-1.73010969e+00 -1.03493321e+00 9.45526302e-01 -4.14751172e-01
1.22361195e+00 -6.24380648e-01 -7.58057654e-01 8.33425641e-01
2.19626546e-01 5.16630352e-01 1.07512152e+00 4.12477791e-01
-7.85718858e-01 -8.29772055e-02 -5.87289572e-01 5.67757189e-01
1.34870553e+00 -8.59290659e-01 -2.41367370e-01 5.09648174e-02
9.97340143e-01 1.29076004e-01 -1.10686886e+00 4.45149869e-01
6.45568252e-01 -8.91540051e-01 1.07808828e+00 -6.67382479e-01
9.16528642e-01 -1.09923214e-01 -1.41908705e-01 -1.38379002e+00
-7.30173767e-01 -6.75756752e-01 -8.44459414e-01 1.36346114e+00
4.73984331e-01 -9.00478542e-01 7.62400389e-01 3.11076790e-01
7.45956823e-02 -1.18585014e+00 -9.40653265e-01 -4.31970954e-01
-1.58265740e-01 -2.64524460e-01 4.87350464e-01 1.26912880e+00
2.48943061e-01 6.71051741e-01 -6.01241350e-01 4.03200150e-01
9.27267671e-01 1.75271094e-01 5.58756053e-01 -1.38879418e+00
-5.57404399e-01 -4.47733045e-01 -8.74495625e-01 -1.07224870e+00
5.12896836e-01 -1.47347891e+00 -5.38151681e-01 -1.61394227e+00
5.18903553e-01 -5.64297915e-01 -4.45868969e-01 3.44195247e-01
-3.98872852e-01 2.97090292e-01 1.46427527e-01 4.08343613e-01
-5.31142354e-01 1.02493834e+00 1.36580515e+00 -2.61340052e-01
2.17535511e-01 -2.42155328e-01 -8.97509098e-01 3.76894265e-01
5.08801997e-01 -5.14711320e-01 -6.60545468e-01 -3.48637551e-01
7.80830801e-01 4.96750288e-02 6.04607463e-01 -5.02164066e-01
1.85928747e-01 4.32430245e-02 9.19718146e-02 -2.65137076e-01
2.80466378e-01 -9.58197355e-01 3.21123272e-01 4.50547218e-01
-2.21642375e-01 -2.05986917e-01 -1.49561778e-01 1.15310013e+00
-1.22685008e-01 7.26949126e-02 -2.44061444e-02 4.57305759e-01
-4.14507359e-01 7.94995487e-01 2.11469188e-01 -1.48208037e-01
9.15148735e-01 1.07317381e-01 -5.85759461e-01 -5.35505295e-01
-6.44353986e-01 3.58615309e-01 3.42319787e-01 4.65072304e-01
6.03470445e-01 -1.91528821e+00 -6.48482203e-01 1.46259114e-01
2.59879500e-01 -4.21575397e-01 7.12186620e-02 1.00335848e+00
-1.93429455e-01 4.56553817e-01 -1.35965981e-02 -2.81568795e-01
-8.53851676e-01 6.12231195e-01 6.77326578e-04 -6.10563397e-01
-5.74719906e-01 6.69502378e-01 5.94594292e-02 -3.92097145e-01
-4.46774773e-02 -1.87434349e-02 -1.78553909e-01 7.17500923e-03
-1.79314259e-02 2.78182447e-01 -5.44151545e-01 -7.06773162e-01
-3.61248076e-01 4.41168725e-01 -9.51947346e-02 3.54358584e-01
1.36024237e+00 -2.03963414e-01 -3.21396291e-01 3.97948980e-01
1.55223501e+00 3.07368696e-01 -1.08591676e+00 -3.36594611e-01
-2.42931172e-01 -4.57532376e-01 1.94196612e-01 -3.96610677e-01
-1.21191645e+00 1.05804038e+00 9.07938257e-02 6.32983804e-01
6.41537726e-01 2.45702490e-01 8.11122298e-01 -1.56708602e-02
3.28655899e-01 -3.04845572e-01 2.00340629e-01 2.19645754e-01
5.33302724e-01 -1.13287663e+00 2.12615609e-01 -1.01356971e+00
-2.89576113e-01 1.17558908e+00 4.53188300e-01 -2.46720746e-01
8.49948168e-01 -1.02406600e-02 -5.71753561e-01 -3.65651608e-01
-9.98243809e-01 -4.26289923e-02 5.48857450e-01 4.41953361e-01
4.96313035e-01 3.81071121e-01 -5.00485539e-01 8.49395156e-01
1.26658648e-01 -2.89147556e-01 3.78707528e-01 3.21223706e-01
1.22026145e-01 -1.51354218e+00 2.14087486e-01 4.92041439e-01
-1.67230889e-01 -1.85028568e-01 -3.79241019e-01 7.43983030e-01
-2.57366262e-02 6.85858548e-01 -9.79091227e-02 -9.07000065e-01
-1.56510472e-01 -1.36085361e-01 3.82260591e-01 -5.24445713e-01
1.30611941e-01 -1.98282972e-01 1.20267138e-01 -4.47280496e-01
-4.81413245e-01 -3.06882560e-01 -1.15595150e+00 -5.97834826e-01
-6.16007686e-01 2.01252297e-01 2.48366728e-01 8.02709401e-01
6.02749348e-01 7.61893392e-01 7.99383283e-01 -5.56069136e-01
-4.30888563e-01 -6.66444659e-01 -4.75301504e-01 3.79727930e-01
3.08256209e-01 -1.06775355e+00 -5.50381720e-01 -4.16448206e-01] | [7.26071310043335, 6.199878215789795] |
5740b52c-8bd0-4b7b-aafa-247b37698075 | a-practical-guide-to-cnns-and-fisher-vectors | 1508.02496 | null | http://arxiv.org/abs/1508.02496v3 | http://arxiv.org/pdf/1508.02496v3.pdf | A Practical Guide to CNNs and Fisher Vectors for Image Instance Retrieval | With deep learning becoming the dominant approach in computer vision, the use
of representations extracted from Convolutional Neural Nets (CNNs) is quickly
gaining ground on Fisher Vectors (FVs) as favoured state-of-the-art global
image descriptors for image instance retrieval. While the good performance of
CNNs for image classification are unambiguously recognised, which of the two
has the upper hand in the image retrieval context is not entirely clear yet. In
this work, we propose a comprehensive study that systematically evaluates FVs
and CNNs for image retrieval. The first part compares the performances of FVs
and CNNs on multiple publicly available data sets. We investigate a number of
details specific to each method. For FVs, we compare sparse descriptors based
on interest point detectors with dense single-scale and multi-scale variants.
For CNNs, we focus on understanding the impact of depth, architecture and
training data on retrieval results. Our study shows that no descriptor is
systematically better than the other and that performance gains can usually be
obtained by using both types together. The second part of the study focuses on
the impact of geometrical transformations such as rotations and scale changes.
FVs based on interest point detectors are intrinsically resilient to such
transformations while CNNs do not have a built-in mechanism to ensure such
invariance. We show that performance of CNNs can quickly degrade in presence of
rotations while they are far less affected by changes in scale. We then propose
a number of ways to incorporate the required invariances in the CNN pipeline.
Overall, our work is intended as a reference guide offering practically useful
and simply implementable guidelines to anyone looking for state-of-the-art
global descriptors best suited to their specific image instance retrieval
problem. | ['Olivier Morère', 'Antoine Veillard', 'Jie Lin', 'Hanlin Goh', 'Vijay Chandrasekhar'] | 2015-08-11 | null | null | null | null | ['image-instance-retrieval'] | ['computer-vision'] | [-1.28400236e-01 -5.40913165e-01 -1.87114164e-01 -3.69872600e-01
-7.06086874e-01 -7.30786443e-01 9.48283553e-01 3.83996665e-01
-6.18169069e-01 2.04461113e-01 -3.17738280e-02 9.07998011e-02
-5.70864499e-01 -8.37378383e-01 -5.94707072e-01 -7.03839064e-01
-1.02706268e-01 2.36886710e-01 4.84824657e-01 -4.81591254e-01
4.41934943e-01 1.23478389e+00 -1.97250044e+00 1.47528335e-01
-2.44110897e-02 1.39235914e+00 -4.16715704e-02 5.41812539e-01
5.10705858e-02 4.78876561e-01 -6.38725698e-01 -1.89298868e-01
5.73262751e-01 -8.06826800e-02 -7.26908207e-01 -3.04303288e-01
8.92107069e-01 -3.48272771e-01 -4.31038976e-01 8.76413584e-01
7.59276450e-01 3.10974289e-02 9.09808338e-01 -1.00635481e+00
-6.64521158e-01 -9.45309177e-02 -1.77228123e-01 2.97373563e-01
1.74354374e-01 5.37665002e-02 1.08815610e+00 -8.41653705e-01
8.71095777e-01 1.14167714e+00 8.52065146e-01 1.20299697e-01
-1.00892878e+00 -3.03447455e-01 -1.94858700e-01 1.63862079e-01
-1.53434825e+00 -4.96036083e-01 6.17915750e-01 -3.10312748e-01
1.27132773e+00 2.25072101e-01 4.62321937e-01 7.40033150e-01
2.71933705e-01 5.38990319e-01 8.27592254e-01 -5.32834709e-01
1.33474156e-01 -1.85140576e-02 1.23477645e-01 6.55728519e-01
4.09350604e-01 2.67894059e-01 -4.78509068e-01 -4.76623289e-02
8.77662480e-01 1.49233833e-01 -5.07973582e-02 -6.88995838e-01
-1.13944674e+00 9.87361729e-01 9.07616198e-01 8.10168087e-01
-2.93947995e-01 3.42564821e-01 6.72853708e-01 6.07665896e-01
3.06939363e-01 5.98740518e-01 -4.16078269e-01 -5.65199032e-02
-1.24515498e+00 4.63004261e-01 6.35128856e-01 7.27038026e-01
8.14669847e-01 -4.22888510e-02 -1.74504951e-01 9.76655006e-01
4.39908728e-02 4.17173952e-01 4.86706316e-01 -6.45331621e-01
-1.18548505e-01 6.83828712e-01 -2.13212252e-01 -1.41256595e+00
-4.68466192e-01 -5.80843985e-01 -7.06585288e-01 2.68078595e-01
5.02640009e-01 5.05009055e-01 -9.48285818e-01 1.36172950e+00
5.97017165e-03 -3.30656677e-01 -2.08434656e-01 8.98389935e-01
9.24450696e-01 2.10094258e-01 -2.11811081e-01 2.89526641e-01
1.35843623e+00 -7.51524448e-01 -2.50683546e-01 4.43111621e-02
7.42743850e-01 -9.62734163e-01 5.87320566e-01 2.47011095e-01
-8.80620837e-01 -6.42064273e-01 -1.27395082e+00 -4.39743966e-01
-8.90170455e-01 1.17711924e-01 5.36802351e-01 5.86760700e-01
-1.35319114e+00 9.71480429e-01 -7.50798881e-01 -7.79451430e-01
3.35132509e-01 6.39189601e-01 -7.17370152e-01 -6.46858886e-02
-1.07838762e+00 1.16481280e+00 1.57328323e-01 1.13506615e-01
-5.56070209e-01 -4.53030825e-01 -8.08754385e-01 1.69281974e-01
-4.39466611e-02 -5.28799117e-01 9.53451931e-01 -9.40272689e-01
-1.24777234e+00 1.16919386e+00 7.65701234e-02 -5.33506215e-01
4.63149220e-01 -1.10794380e-01 -1.12492561e-01 3.51065755e-01
-5.76809570e-02 8.22074473e-01 8.61516714e-01 -1.01238227e+00
-2.96271294e-01 -3.20086092e-01 1.46075234e-01 -1.02517135e-01
-4.25040483e-01 1.91621423e-01 -3.94105703e-01 -5.21931052e-01
6.13853559e-02 -9.23007965e-01 8.91868994e-02 4.40746486e-01
-4.53858227e-02 -4.34346676e-01 8.57574463e-01 -8.05173591e-02
1.02501559e+00 -2.06725836e+00 -2.33233310e-02 2.60612458e-01
4.23680991e-02 6.30246222e-01 -4.15644616e-01 7.86654830e-01
-1.99855477e-01 8.92396569e-02 8.11882392e-02 -2.92624265e-01
6.84153810e-02 2.33769745e-01 -1.75985470e-01 8.41382563e-01
4.27595109e-01 9.79484558e-01 -6.41256571e-01 -2.00951353e-01
4.97570843e-01 9.12854552e-01 -3.57724398e-01 -1.58830240e-01
3.95663232e-01 -3.27337570e-02 -3.54666412e-01 6.73536062e-01
6.12753332e-01 -9.82832015e-02 -1.70226082e-01 -3.94589126e-01
-3.07375252e-01 1.22410387e-01 -9.87175167e-01 1.66605604e+00
-3.69585872e-01 9.48297739e-01 -1.35154456e-01 -1.22564685e+00
1.12736642e+00 7.72855133e-02 2.92159945e-01 -9.09865618e-01
2.85417587e-01 5.94428062e-01 -3.95435281e-02 -1.55701697e-01
5.05015194e-01 -2.40950473e-02 1.28464207e-01 4.08827923e-02
5.16311884e-01 -5.70804924e-02 1.85780451e-01 -6.05720095e-02
1.21271431e+00 -6.08240440e-02 4.55397844e-01 -5.84821105e-01
6.20486438e-01 -1.32843480e-01 -1.29347965e-01 9.46993530e-01
-2.09181145e-01 1.07988060e+00 3.64741534e-01 -8.62690747e-01
-1.12126148e+00 -6.58095181e-01 -4.41984296e-01 9.70839322e-01
9.77727547e-02 -4.90096331e-01 -4.13516521e-01 -3.62876356e-01
8.84878188e-02 -1.24532521e-01 -8.42669308e-01 -1.51871920e-01
-4.93614733e-01 -4.94839877e-01 7.51835823e-01 3.75198156e-01
4.47793841e-01 -1.08382189e+00 -9.48232591e-01 -1.99647918e-02
4.07815486e-01 -1.00544965e+00 -3.66889723e-02 3.43156606e-01
-7.97091186e-01 -1.11566830e+00 -1.07453573e+00 -6.87434971e-01
4.24222559e-01 4.40088093e-01 1.10604930e+00 3.33542347e-01
-5.56712210e-01 6.79924309e-01 -5.17078340e-01 -2.86491752e-01
-2.51752008e-02 4.75989759e-01 -8.40706378e-02 -1.54825926e-01
3.96779448e-01 -5.09623408e-01 -9.25929964e-01 2.20217556e-01
-1.31026709e+00 -7.12494135e-01 7.33786225e-01 7.96374500e-01
3.48446518e-01 -3.05463672e-01 6.13372549e-02 -5.85370243e-01
6.15242302e-01 -2.00951248e-01 -5.02948999e-01 8.87836367e-02
-5.18890023e-01 2.56014168e-01 4.58015919e-01 -2.27470711e-01
-2.41230682e-01 1.46968439e-01 -2.70892560e-01 -5.44858813e-01
-2.91558921e-01 4.86964017e-01 2.94863045e-01 -7.23734856e-01
8.49280417e-01 1.00312687e-01 4.58477028e-02 -4.71727163e-01
2.46132746e-01 4.32463408e-01 2.36042827e-01 -4.99501973e-01
8.45020354e-01 6.74097300e-01 3.11154306e-01 -1.17336142e+00
-4.17887390e-01 -8.45710397e-01 -8.32431018e-01 -6.47949195e-03
7.52058804e-01 -6.89591229e-01 -5.52414954e-01 5.96572995e-01
-1.11486304e+00 8.10734779e-02 -2.59292126e-01 1.07820891e-01
-4.95187968e-01 3.37894619e-01 -4.30962056e-01 -5.15330195e-01
-3.06772679e-01 -1.37888873e+00 1.47483492e+00 2.57711243e-02
-1.53458282e-01 -1.06222320e+00 1.62460461e-01 -4.13323678e-02
9.91739154e-01 2.45611250e-01 5.95896780e-01 -8.96856785e-01
-5.61903775e-01 -7.48426735e-01 -4.25191045e-01 4.75217938e-01
2.18395446e-03 1.97312489e-01 -1.13377905e+00 -5.21414578e-01
-1.81530014e-01 -2.96601593e-01 1.22718406e+00 3.08344334e-01
8.38663697e-01 -6.20045811e-02 -1.25075713e-01 7.76041806e-01
1.81959581e+00 -9.86125618e-02 9.85766590e-01 7.28688717e-01
4.41099375e-01 6.40246749e-01 2.18904629e-01 1.06575757e-01
-1.06501035e-01 9.22639430e-01 4.92620379e-01 -2.44416296e-01
-2.54555553e-01 1.24074295e-01 2.36323506e-01 5.54673016e-01
-2.99129963e-01 -9.34308097e-02 -8.61142695e-01 6.00547016e-01
-1.54578733e+00 -8.03947449e-01 3.07149664e-02 2.29869342e+00
2.83475071e-01 -2.87999995e-02 -4.93518189e-02 2.59678096e-01
4.36654747e-01 3.22647572e-01 -1.31769642e-01 -4.01748031e-01
-3.05288017e-01 7.08365738e-01 7.43803561e-01 2.02662930e-01
-1.32972860e+00 7.73737252e-01 6.37808180e+00 9.60425079e-01
-1.57166374e+00 -1.24745235e-01 3.51359606e-01 -2.09356900e-02
1.62590370e-01 -2.71212421e-02 -6.74389303e-01 -8.08549020e-03
9.19637263e-01 3.65398347e-01 1.07779406e-01 8.05197775e-01
-2.35725179e-01 -1.58215798e-02 -1.09903622e+00 1.13988829e+00
1.72562554e-01 -1.36982977e+00 2.40063816e-01 1.66181013e-01
5.56446016e-01 4.62273359e-01 1.66929752e-01 1.39053673e-01
-3.03566337e-01 -1.19287455e+00 7.26914108e-01 4.25137639e-01
7.12957501e-01 -6.15995288e-01 1.03207088e+00 -5.26330844e-02
-1.04331172e+00 1.10439874e-01 -6.05741441e-01 4.91296202e-02
-3.57779801e-01 4.28245813e-01 -3.25983286e-01 6.11839712e-01
6.77460134e-01 7.19021738e-01 -9.66351330e-01 1.19865191e+00
2.29412809e-01 8.48517120e-02 -4.52503026e-01 2.50706286e-03
5.99086702e-01 1.00325204e-01 4.40353394e-01 1.56232369e+00
3.62393379e-01 -2.60862321e-01 -1.83539823e-01 5.79098105e-01
8.12585056e-02 3.07597667e-01 -8.86804283e-01 -5.32469414e-02
1.40096843e-01 1.46077335e+00 -9.77497637e-01 -1.56433120e-01
-4.78289843e-01 8.41556728e-01 3.31416845e-01 2.80165792e-01
-2.74079829e-01 -5.53881288e-01 7.31409371e-01 3.17926884e-01
6.70916080e-01 -2.30872914e-01 1.48942888e-01 -1.14073730e+00
-2.24835798e-02 -7.28580415e-01 1.83401123e-01 -6.97267592e-01
-1.34726417e+00 6.86518610e-01 7.95898214e-03 -1.22381496e+00
-2.30847865e-01 -1.08898044e+00 -3.81610423e-01 7.68862963e-01
-1.87669039e+00 -1.26302421e+00 -3.54393601e-01 6.18182778e-01
2.48309955e-01 -1.37692392e-01 1.01200783e+00 2.93745518e-01
-2.92957332e-02 6.30781710e-01 2.87250042e-01 2.74591982e-01
9.03800368e-01 -8.93581629e-01 2.41532311e-01 4.87566292e-01
4.39666569e-01 1.15391004e+00 6.35482788e-01 -3.61458659e-02
-1.35515594e+00 -5.93049467e-01 7.89042473e-01 -4.30922806e-01
6.27773881e-01 -3.38487238e-01 -8.80831897e-01 2.34204531e-01
1.12431414e-01 4.42650825e-01 3.33160520e-01 3.62817012e-02
-8.00725162e-01 -1.70600742e-01 -8.86216700e-01 7.86384642e-02
7.31055975e-01 -8.59116852e-01 -3.45806479e-01 2.71528631e-01
1.31589040e-01 -3.31200123e-01 -8.58326018e-01 5.45979142e-01
8.35247099e-01 -1.37884521e+00 1.19252527e+00 -4.31720197e-01
4.91774499e-01 -8.58715847e-02 -3.14913481e-01 -9.00050879e-01
-3.48203868e-01 -2.51040965e-01 3.50927889e-01 8.98168325e-01
1.15229517e-01 -7.80177832e-01 5.92318416e-01 2.60922551e-01
-5.68578281e-02 -7.69526482e-01 -1.06425583e+00 -1.01214314e+00
3.42782378e-01 -3.06694001e-01 2.56981403e-01 7.91594565e-01
-4.51179534e-01 8.59379023e-02 -4.34965827e-02 -1.99979514e-01
1.29355714e-01 -8.10101107e-02 8.12180638e-01 -1.37538779e+00
3.27941440e-02 -8.16033304e-01 -1.31145859e+00 -7.83039629e-01
-1.04248039e-01 -8.70128095e-01 -1.72491938e-01 -1.44003797e+00
6.38535321e-02 -1.78846464e-01 -5.52674294e-01 4.44348216e-01
3.30111146e-01 6.43073618e-01 4.40760136e-01 5.02424359e-01
-5.70758224e-01 2.43720829e-01 9.40328181e-01 -8.44915882e-02
1.09522954e-01 -2.39267886e-01 -4.46990490e-01 4.63766754e-01
6.02536023e-01 -3.58921587e-01 -1.93555743e-01 -4.02323693e-01
3.75747025e-01 -4.87299591e-01 6.93284810e-01 -1.21754313e+00
3.76633525e-01 4.65186417e-01 3.87483209e-01 -2.55095989e-01
3.49012256e-01 -8.05887461e-01 -4.79940809e-02 3.92378241e-01
-3.61566097e-01 2.66601115e-01 3.52604628e-01 4.20284152e-01
-6.03498042e-01 -4.12021726e-01 8.04070950e-01 -2.94486076e-01
-6.39194846e-01 2.88987428e-01 -1.58124700e-01 -2.25351840e-01
6.26177490e-01 -4.27414745e-01 -2.00188532e-01 -4.10197586e-01
-4.03439373e-01 -4.14869189e-01 5.41073918e-01 5.89104652e-01
5.02279758e-01 -1.24377537e+00 -5.34249544e-01 2.46538430e-01
3.56891423e-01 -2.05344841e-01 1.26736090e-01 6.92146480e-01
-9.29719865e-01 9.94781673e-01 -3.39527249e-01 -6.82188690e-01
-1.32260180e+00 5.15430987e-01 5.07817090e-01 -2.80738413e-01
-3.69808048e-01 8.78094435e-01 1.71202719e-01 -2.80125886e-01
1.68416247e-01 -4.21773344e-01 -1.55139819e-01 3.95162970e-01
3.85482967e-01 1.23257235e-01 7.36211002e-01 -9.62841749e-01
-5.28241515e-01 9.46494699e-01 -2.87954748e-01 1.47945702e-01
1.60569692e+00 1.87115908e-01 -2.43816644e-01 3.23229909e-01
1.81370044e+00 -3.46216172e-01 -8.32556188e-01 -6.28987104e-02
-2.28882935e-02 -3.90372157e-01 2.38632709e-01 -4.90609825e-01
-1.25179553e+00 1.10733151e+00 8.90101135e-01 3.68104994e-01
1.02816951e+00 -1.63840484e-02 5.66333115e-01 4.28430825e-01
3.26282710e-01 -7.91854858e-01 -4.42132354e-02 6.04976833e-01
1.07440078e+00 -1.10246634e+00 2.59971619e-01 -7.78766647e-02
-1.39194876e-01 1.46187961e+00 -2.98949406e-02 -7.97012508e-01
8.05440605e-01 -3.49516124e-02 5.29145934e-02 -6.13512933e-01
-4.72579867e-01 -4.19218779e-01 6.97286367e-01 5.71202934e-01
5.27624547e-01 -3.52843493e-01 -2.59757161e-01 -1.73311412e-01
-1.20652020e-01 -1.57858342e-01 3.70893367e-02 1.01440978e+00
-3.49929750e-01 -1.21159804e+00 -2.70310551e-01 3.27274412e-01
-5.78298986e-01 -1.58290178e-01 -7.04027414e-01 1.21576750e+00
8.92238021e-02 4.77661490e-01 7.63468351e-03 -2.88415819e-01
4.72071588e-01 -2.59650312e-02 6.64645076e-01 -4.64744121e-01
-8.96648586e-01 -1.19676292e-01 -3.18963885e-01 -7.77307749e-01
-7.44664729e-01 -5.96816063e-01 -6.74337804e-01 -2.79591918e-01
-5.10390162e-01 -2.71638483e-01 1.01118279e+00 9.32723105e-01
3.84059995e-01 2.64968038e-01 1.06926255e-01 -1.20125532e+00
-5.20812809e-01 -7.68898427e-01 -5.38162708e-01 4.99545485e-01
5.62807620e-01 -7.98733354e-01 -5.10517538e-01 -3.41173112e-01] | [10.55742359161377, 0.4003359377384186] |
9af76571-e913-4fb2-927c-07e7840fc679 | deep-burst-super-resolution | 2101.10997 | null | https://arxiv.org/abs/2101.10997v2 | https://arxiv.org/pdf/2101.10997v2.pdf | Deep Burst Super-Resolution | While single-image super-resolution (SISR) has attracted substantial interest in recent years, the proposed approaches are limited to learning image priors in order to add high frequency details. In contrast, multi-frame super-resolution (MFSR) offers the possibility of reconstructing rich details by combining signal information from multiple shifted images. This key advantage, along with the increasing popularity of burst photography, have made MFSR an important problem for real-world applications. We propose a novel architecture for the burst super-resolution task. Our network takes multiple noisy RAW images as input, and generates a denoised, super-resolved RGB image as output. This is achieved by explicitly aligning deep embeddings of the input frames using pixel-wise optical flow. The information from all frames are then adaptively merged using an attention-based fusion module. In order to enable training and evaluation on real-world data, we additionally introduce the BurstSR dataset, consisting of smartphone bursts and high-resolution DSLR ground-truth. We perform comprehensive experimental analysis, demonstrating the effectiveness of the proposed architecture. | ['Radu Timofte', 'Luc van Gool', 'Martin Danelljan', 'Goutam Bhat'] | 2021-01-26 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Bhat_Deep_Burst_Super-Resolution_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Bhat_Deep_Burst_Super-Resolution_CVPR_2021_paper.pdf | cvpr-2021-1 | ['multi-frame-super-resolution', 'burst-image-super-resolution'] | ['computer-vision', 'computer-vision'] | [ 5.34737468e-01 -3.21385860e-01 4.51990031e-02 -3.30484629e-01
-1.12767017e+00 -8.93464983e-02 5.17614961e-01 -3.78335804e-01
-3.99301291e-01 9.34747934e-01 4.13374692e-01 3.47299933e-01
-7.27879778e-02 -6.32481694e-01 -8.59556973e-01 -8.23806465e-01
1.67078108e-01 -1.71282619e-01 1.75900340e-01 -2.23343879e-01
7.74817541e-02 4.33771104e-01 -1.86619699e+00 4.59433585e-01
9.87285256e-01 1.08907390e+00 7.22514629e-01 5.81896365e-01
1.34008348e-01 9.60276484e-01 -5.03583074e-01 -1.63767397e-01
2.65054613e-01 -5.81525028e-01 -4.53110665e-01 1.65678814e-01
5.78718603e-01 -8.25719833e-01 -5.57481170e-01 9.19311583e-01
5.58392942e-01 4.73678946e-01 -1.02231137e-01 -8.18007290e-01
-8.94844711e-01 3.57154787e-01 -5.43234408e-01 7.49540210e-01
4.39207017e-01 3.69209290e-01 8.90845478e-01 -9.45158362e-01
5.70813596e-01 1.22429073e+00 3.66391093e-01 4.87372667e-01
-1.36468983e+00 -6.88167751e-01 9.12670270e-02 3.07457119e-01
-1.25784624e+00 -6.37047231e-01 8.53024542e-01 7.45328665e-02
7.39481330e-01 9.39203128e-02 5.98507881e-01 1.30919850e+00
1.07086338e-01 4.15979892e-01 1.31285644e+00 3.24856378e-02
1.16253711e-01 -1.30148157e-01 -2.49007836e-01 2.56856978e-01
1.98417604e-01 1.40295163e-01 -8.84526849e-01 2.26236984e-01
1.26020324e+00 3.13102663e-01 -6.26046300e-01 2.62726635e-01
-1.27750027e+00 4.25629228e-01 5.91426313e-01 4.79120612e-01
-6.36735916e-01 2.28611410e-01 -1.43679857e-01 -6.68486133e-02
6.53597355e-01 1.50206745e-01 3.46243605e-02 -1.07552081e-01
-1.05935812e+00 5.83310276e-02 2.46940088e-02 4.99022961e-01
7.22408533e-01 4.08114314e-01 2.86959000e-02 8.62287700e-01
4.75466326e-02 4.43725705e-01 5.87338626e-01 -1.44694507e+00
3.14142317e-01 7.61380494e-02 5.04280388e-01 -9.19688642e-01
-1.03748687e-01 -5.92244446e-01 -1.11376572e+00 1.34235546e-01
6.23673573e-02 2.21887097e-01 -6.57731354e-01 1.66273081e+00
2.79287905e-01 7.64556050e-01 1.51880637e-01 1.42895329e+00
7.38886118e-01 9.46887672e-01 -1.76175460e-01 -4.60304707e-01
1.16107905e+00 -6.85438931e-01 -1.02180529e+00 -7.09643587e-02
-3.81694257e-01 -5.87842166e-01 9.67565119e-01 5.01698792e-01
-1.29522514e+00 -1.00609481e+00 -9.90956366e-01 -5.16226590e-01
1.62395045e-01 -1.72019854e-01 2.38972396e-01 1.01938680e-01
-1.20561171e+00 6.93722069e-01 -9.64332640e-01 -5.13045155e-02
5.11894524e-01 1.95851130e-03 -4.41466361e-01 -4.46754187e-01
-1.21134853e+00 6.84122682e-01 2.06039667e-01 2.15749562e-01
-8.74113381e-01 -7.36643314e-01 -8.03363740e-01 6.90713972e-02
2.29176104e-01 -6.95575953e-01 9.27077293e-01 -8.86137247e-01
-1.69041264e+00 5.19923270e-01 -3.72622877e-01 -6.32170677e-01
3.40627044e-01 -4.06341940e-01 -5.89915752e-01 6.32758021e-01
7.28271753e-02 5.42119205e-01 1.14444029e+00 -1.36421871e+00
-8.92950773e-01 -1.90848529e-01 2.30406374e-01 3.29712301e-01
-1.85841978e-01 -5.65955304e-02 -2.76275218e-01 -7.51970053e-01
9.09216050e-03 -4.94291037e-01 -2.28652775e-01 -1.78647593e-01
4.41832282e-02 1.96847275e-01 7.19904065e-01 -9.11314607e-01
1.03460550e+00 -2.32895422e+00 3.31916034e-01 -5.27653217e-01
4.24402356e-01 9.79810953e-02 -1.84131280e-01 -6.44764230e-02
-8.16943049e-02 -1.63425550e-01 -3.05265963e-01 -6.86148226e-01
-4.57644612e-01 2.47582152e-01 -6.57951713e-01 5.43138206e-01
4.49286073e-01 7.03160882e-01 -1.07881582e+00 -3.05408716e-01
6.16579652e-01 1.25067163e+00 -4.16229635e-01 4.24176276e-01
-1.01384602e-01 9.57240582e-01 -1.33315772e-01 3.90703738e-01
7.45844781e-01 -4.93626684e-01 -1.53323263e-01 -5.96988201e-01
-2.74481237e-01 1.57068089e-01 -1.16114080e+00 1.89764059e+00
-8.21030796e-01 6.74508750e-01 1.10331913e-02 -6.22148991e-01
7.48306274e-01 1.24240316e-01 6.41044736e-01 -1.10145330e+00
-8.44243169e-02 2.46867806e-01 -4.25287932e-01 -4.60664451e-01
8.73282909e-01 -1.89986587e-01 2.43193135e-01 3.18847388e-01
-3.68094118e-03 -3.13703194e-02 -3.03531867e-02 2.87358999e-01
8.48861337e-01 2.36996919e-01 6.31355494e-02 3.24077010e-01
4.91123796e-01 -4.43638235e-01 6.18501842e-01 3.35305095e-01
-9.37623903e-03 1.13276863e+00 -1.53359890e-01 -4.36460316e-01
-1.12728429e+00 -1.24621129e+00 -6.49066865e-02 6.18936896e-01
5.61883271e-01 -2.47823343e-01 -5.87403834e-01 -1.19527355e-01
-3.19047034e-01 7.25465059e-01 -5.57434916e-01 8.93920138e-02
-7.69727647e-01 -6.86116755e-01 -2.18296307e-03 2.19621599e-01
8.48754227e-01 -8.26030076e-01 -9.04365480e-01 3.92638087e-01
-9.31492209e-01 -1.68104827e+00 -4.41621959e-01 -2.27626026e-01
-7.40048885e-01 -8.12034309e-01 -7.81938016e-01 -2.91603059e-01
2.69764841e-01 8.06760609e-01 1.03608835e+00 -6.45034835e-02
-4.43922758e-01 4.06307489e-01 -3.43747020e-01 3.31582457e-01
-3.32534224e-01 -3.27019393e-01 3.70943956e-02 7.03547418e-01
-2.98627555e-01 -8.78412902e-01 -9.76032853e-01 1.53892592e-01
-1.29121304e+00 2.30261326e-01 7.67875254e-01 7.49558032e-01
9.81259108e-01 2.42801040e-01 5.84148407e-01 -4.44957584e-01
3.82835329e-01 -4.39585775e-01 -5.26249468e-01 -1.57648861e-01
-1.81654990e-01 -4.88252798e-03 7.58210301e-01 -4.47232693e-01
-1.47252834e+00 -1.22493111e-01 -1.14749275e-01 -8.59648764e-01
-1.02810539e-01 1.08644135e-01 -1.62005052e-01 8.95086527e-02
4.66572940e-01 4.74126935e-01 1.58137686e-04 -5.66148162e-01
5.08708239e-01 6.59128189e-01 1.02040696e+00 -1.60757959e-01
6.99938416e-01 1.05465150e+00 -9.55010802e-02 -9.65569615e-01
-1.01834095e+00 -2.78587162e-01 -3.93688440e-01 -3.12394172e-01
9.79895234e-01 -1.37091041e+00 -5.23970544e-01 4.29078758e-01
-1.11509621e+00 -3.05108458e-01 -4.65066344e-01 4.93748635e-01
-6.01848662e-01 3.49673361e-01 -5.74819148e-01 -7.88010061e-01
-2.01263532e-01 -1.24452484e+00 1.22031617e+00 4.01459575e-01
1.88366324e-01 -5.62597275e-01 -1.30673021e-01 5.40318131e-01
7.22601593e-01 3.39284629e-01 1.96091264e-01 1.09996274e-01
-1.13087893e+00 1.99079618e-01 -4.35485989e-01 5.94080329e-01
3.72506201e-01 -3.61540705e-01 -1.13892853e+00 -2.90318668e-01
3.67992938e-01 -1.64365932e-01 1.00594211e+00 4.77807760e-01
1.13316357e+00 -1.33907303e-01 9.07980558e-03 8.78143013e-01
1.62157309e+00 -1.65694460e-01 7.75569797e-01 1.75739363e-01
8.71747494e-01 3.38314801e-01 6.12779677e-01 4.51629877e-01
5.63609481e-01 8.63750160e-01 6.44094229e-01 -1.72393888e-01
-7.51471519e-01 -6.49647936e-02 4.26914215e-01 6.39156997e-01
-3.75413537e-01 -1.93974555e-01 -2.82567322e-01 4.57001209e-01
-1.56475759e+00 -1.28238320e+00 9.72045213e-02 2.16618705e+00
7.81374991e-01 -1.69179544e-01 -4.38977331e-02 -1.71970725e-02
7.46732354e-01 6.15735948e-01 -7.14315832e-01 2.75417030e-01
-5.84142983e-01 3.24219614e-01 2.46202141e-01 6.23730958e-01
-8.88320625e-01 7.07404137e-01 5.03335762e+00 8.64472389e-01
-1.32973993e+00 4.75796789e-01 8.06103289e-01 -3.52176458e-01
-4.21917081e-01 -2.41955400e-01 -6.39183521e-01 7.31772184e-01
1.04544997e+00 -4.18572761e-02 9.23795760e-01 2.00360373e-01
5.51881909e-01 -3.43954653e-01 -6.94594800e-01 1.47171509e+00
2.08505258e-01 -1.47540331e+00 1.50615172e-02 -8.41096789e-02
8.91308188e-01 9.13705230e-02 3.87652814e-01 -1.87666684e-01
1.91114724e-01 -9.65412974e-01 7.54802465e-01 7.42596805e-01
1.06766558e+00 -7.42696106e-01 3.75863254e-01 1.55684829e-01
-1.25233757e+00 -2.88512260e-01 -3.85588020e-01 3.75087596e-02
5.68422556e-01 8.25193703e-01 -2.53664792e-01 8.49368095e-01
1.11487305e+00 1.06561828e+00 -3.45667034e-01 7.10263968e-01
-1.70118004e-01 3.45942616e-01 -2.82918394e-01 7.09462047e-01
-2.03100204e-01 -2.99194336e-01 5.89094937e-01 8.71631265e-01
4.54333007e-01 3.99010897e-01 -1.86938331e-01 1.00569308e+00
-1.00140214e-01 -4.14156467e-01 -3.39499682e-01 8.42863694e-02
4.17853564e-01 1.48256719e+00 -5.10620415e-01 -2.80159682e-01
-5.89569807e-01 1.31974101e+00 9.80915874e-02 5.22638679e-01
-9.65495825e-01 -6.04007803e-02 8.77251446e-01 2.11333945e-01
3.79421264e-01 -3.66877109e-01 4.37097996e-02 -1.42808318e+00
1.02976896e-01 -7.61301160e-01 1.31125525e-01 -1.12605655e+00
-1.08508945e+00 8.98054957e-01 -2.45138690e-01 -1.38851452e+00
-2.15149909e-01 1.04147820e-02 -1.84151918e-01 9.53729749e-01
-2.03866720e+00 -1.06507969e+00 -7.70176232e-01 5.66656411e-01
7.67523885e-01 2.78796881e-01 4.03356135e-01 5.67122877e-01
-6.07533455e-01 1.72274143e-01 6.67406172e-02 -2.05977455e-01
7.41685510e-01 -8.29491436e-01 3.73411536e-01 1.22913873e+00
1.30104125e-01 3.49930435e-01 8.89931142e-01 -4.96236533e-01
-1.35190356e+00 -1.39498317e+00 3.42728853e-01 -3.49298447e-01
5.07390559e-01 -2.10701868e-01 -1.17684042e+00 4.75199312e-01
2.25199819e-01 5.61503768e-01 1.57543615e-01 -7.46418834e-01
-2.79612273e-01 -4.76799160e-01 -1.01714826e+00 2.88372815e-01
1.14548910e+00 -6.33680284e-01 -2.03738227e-01 -3.65158804e-02
1.07952893e+00 -5.39585233e-01 -9.32075262e-01 3.36477518e-01
3.44408512e-01 -1.47538805e+00 1.28096807e+00 1.02606736e-01
5.92503548e-01 -5.94642401e-01 -3.89076859e-01 -1.28575563e+00
-1.12394676e-01 -7.61423886e-01 -2.57867426e-01 1.16482854e+00
-2.18580723e-01 -7.51675963e-01 2.91911215e-01 2.55031973e-01
-2.02591151e-01 -5.62152267e-01 -9.83028471e-01 -5.31866670e-01
-4.75313514e-01 -2.88924754e-01 6.26455665e-01 7.88776100e-01
-6.06784225e-01 1.65269107e-01 -7.43829668e-01 5.04970372e-01
9.97329950e-01 3.48205686e-01 6.42695308e-01 -8.63832653e-01
-4.87611741e-01 -1.79617614e-01 -2.24898249e-01 -1.10088181e+00
5.02433255e-02 -4.27316785e-01 9.00403261e-02 -1.52256119e+00
1.11194722e-01 -9.66082215e-02 -2.24226832e-01 -5.95386364e-02
-3.11991572e-01 7.36189663e-01 2.25459456e-01 3.53567779e-01
-7.40321994e-01 7.03107893e-01 1.27106047e+00 2.36486062e-01
-1.83980122e-01 -2.94006795e-01 -6.68265104e-01 4.27559584e-01
6.24659479e-01 -7.75818080e-02 -3.85858804e-01 -6.63139403e-01
1.54385790e-01 3.63235742e-01 7.27589607e-01 -1.07406330e+00
1.54903173e-01 -1.87237728e-02 4.80502695e-01 -3.45743448e-01
8.22787285e-01 -8.06351006e-01 4.90427881e-01 1.05207287e-01
-2.92791426e-01 -3.50426435e-02 4.42577666e-03 8.16983581e-01
-4.63159114e-01 4.03886974e-01 1.03592217e+00 -7.49173388e-02
-7.38773465e-01 4.39480662e-01 3.95295732e-02 -9.54698473e-02
7.16347516e-01 -2.08373398e-01 -3.73962343e-01 -4.31918323e-01
-5.76147676e-01 -7.26742446e-02 5.91590524e-01 3.90570611e-01
9.42764580e-01 -1.16885209e+00 -7.96803534e-01 3.21978062e-01
-2.44739220e-01 3.49714011e-01 8.96670818e-01 8.23897362e-01
-2.77323306e-01 6.63204342e-02 -2.94660389e-01 -6.52670085e-01
-8.78906012e-01 5.77424586e-01 3.19256663e-01 4.94080894e-02
-1.02290261e+00 6.66905701e-01 3.41436565e-01 3.75546634e-01
-4.79409993e-02 -5.15482426e-01 -2.26442501e-01 -8.55647773e-02
9.48380589e-01 3.98986250e-01 -5.22902086e-02 -9.24661040e-01
6.61810674e-03 5.15618026e-01 9.44103599e-02 -2.69214898e-01
1.66857612e+00 -7.23519444e-01 1.92304626e-02 5.44775426e-01
1.15162790e+00 5.91009222e-02 -1.78939676e+00 -5.26812017e-01
-4.45993781e-01 -8.28773558e-01 4.09440756e-01 -5.34037292e-01
-1.25368464e+00 9.78357553e-01 6.68185592e-01 6.63696527e-02
1.53736818e+00 -2.74190158e-02 1.12777364e+00 -2.91544676e-01
5.45320988e-01 -6.69170558e-01 4.24409300e-01 -1.77479461e-02
8.25985610e-01 -1.23619878e+00 -8.46552700e-02 -3.15433502e-01
-5.01639426e-01 1.03376329e+00 4.24626052e-01 -2.32481822e-01
1.62638411e-01 2.35238388e-01 -7.14517608e-02 2.67389596e-01
-7.91547716e-01 -2.97372550e-01 -8.19805935e-02 5.27533829e-01
8.33163634e-02 -4.12945181e-01 1.54308945e-01 6.44987881e-01
4.09213128e-03 2.76704490e-01 7.32088089e-01 4.56582248e-01
-3.91096979e-01 -6.56948686e-01 -4.84805822e-01 1.08106352e-01
-5.09340346e-01 -5.19172661e-03 3.46652865e-01 4.66868728e-01
4.79848422e-02 9.45713401e-01 3.59906256e-01 -2.65506506e-01
2.29722843e-01 -3.83916080e-01 6.24491334e-01 -3.04872692e-01
-9.09183919e-02 2.23804608e-01 -3.39725673e-01 -1.08836699e+00
-7.55854487e-01 -6.27482414e-01 -1.14455092e+00 -1.31460771e-01
7.05685169e-02 -8.44646543e-02 5.34234405e-01 7.74783790e-01
5.60833931e-01 8.28665376e-01 7.19604909e-01 -1.40675724e+00
-1.93693444e-01 -7.27029741e-01 -5.83860397e-01 5.47330260e-01
9.22725737e-01 -5.19767761e-01 -6.28350079e-01 3.48894179e-01] | [10.914033889770508, -2.002272367477417] |
c92bbcb7-4441-47da-ab76-f474806de74d | egocentric-activity-recognition-and | 2105.09544 | null | https://arxiv.org/abs/2105.09544v3 | https://arxiv.org/pdf/2105.09544v3.pdf | Egocentric Activity Recognition and Localization on a 3D Map | Given a video captured from a first person perspective and the environment context of where the video is recorded, can we recognize what the person is doing and identify where the action occurs in the 3D space? We address this challenging problem of jointly recognizing and localizing actions of a mobile user on a known 3D map from egocentric videos. To this end, we propose a novel deep probabilistic model. Our model takes the inputs of a Hierarchical Volumetric Representation (HVR) of the 3D environment and an egocentric video, infers the 3D action location as a latent variable, and recognizes the action based on the video and contextual cues surrounding its potential locations. To evaluate our model, we conduct extensive experiments on the subset of Ego4D dataset, in which both human naturalistic actions and photo-realistic 3D environment reconstructions are captured. Our method demonstrates strong results on both action recognition and 3D action localization across seen and unseen environments. We believe our work points to an exciting research direction in the intersection of egocentric vision, and 3D scene understanding. | ['Chao Li', 'James M. Rehg', 'Kristen Grauman', 'Yin Li', 'Kiran Somasundaram', 'Lingni Ma', 'Miao Liu'] | 2021-05-20 | null | null | null | null | ['egocentric-activity-recognition'] | ['computer-vision'] | [ 2.18409926e-01 -1.66241065e-01 -2.21765965e-01 -4.14793789e-01
-4.93314385e-01 -6.27666831e-01 6.47267878e-01 -4.46703494e-01
-1.90616429e-01 1.67959750e-01 9.94964659e-01 3.16231936e-01
2.34727368e-01 -4.83277053e-01 -8.51070225e-01 -4.88605171e-01
4.46932809e-03 4.92704272e-01 7.04836026e-02 4.12497729e-01
3.00909668e-01 4.93414015e-01 -1.43869889e+00 4.57940668e-01
5.22962287e-02 7.38306880e-01 2.44340792e-01 8.83222580e-01
4.89762127e-01 9.93254721e-01 -1.63200662e-01 9.70240235e-02
2.94385284e-01 -1.96544066e-01 -8.72095287e-01 9.44096744e-01
7.50454187e-01 -8.28320742e-01 -1.08106601e+00 7.89656103e-01
2.13721231e-01 4.37155992e-01 5.99477351e-01 -1.23520279e+00
-3.83538812e-01 4.51913849e-02 -5.73045552e-01 4.25461233e-01
1.11776006e+00 4.12034899e-01 6.59826517e-01 -8.02083910e-01
9.67046380e-01 1.46328676e+00 2.98245549e-01 3.63286138e-01
-9.52950001e-01 -6.87177032e-02 5.57504535e-01 4.82725859e-01
-1.52767920e+00 -5.42931020e-01 8.65204155e-01 -7.45411694e-01
1.20477617e+00 -1.33135617e-01 9.09162700e-01 1.67684066e+00
2.67725945e-01 1.14774144e+00 7.76335418e-01 -2.48091087e-01
3.17773074e-01 -1.70059934e-01 -2.09952816e-01 7.46994495e-01
-3.18399012e-01 -3.21531653e-01 -8.67103755e-01 -9.41775739e-02
1.14341283e+00 5.13343155e-01 -3.14985901e-01 -9.73694265e-01
-1.49862003e+00 2.58025557e-01 1.49317235e-01 1.08696729e-01
-7.68404245e-01 5.19664109e-01 5.67115843e-02 -3.03855717e-01
5.69508314e-01 -4.60672230e-02 -2.30928585e-01 -6.11990094e-01
-6.39911175e-01 3.26530486e-01 2.78527200e-01 9.33759272e-01
5.20721555e-01 -2.26445988e-01 -1.83582246e-01 1.60772279e-01
3.19015920e-01 7.49929428e-01 2.08079919e-01 -1.46830797e+00
5.64389825e-01 8.47765565e-01 4.64579463e-01 -1.03766406e+00
-1.99757352e-01 -2.31671799e-02 -7.06534162e-02 -2.08246112e-01
4.29541081e-01 -3.01483236e-02 -8.95660043e-01 1.78473139e+00
6.48067653e-01 7.54009008e-01 -1.05828948e-01 1.01783264e+00
3.70476544e-01 4.88501370e-01 -3.38557027e-02 9.61578339e-02
1.26706231e+00 -6.61687672e-01 -5.10682285e-01 -4.80110198e-01
7.08965123e-01 -2.08818376e-01 7.38528967e-01 -1.44209526e-02
-1.16136122e+00 -5.12701690e-01 -5.21322072e-01 -2.68516690e-01
-6.87059462e-02 1.61936715e-01 3.71187925e-01 2.10400268e-01
-9.87615407e-01 1.50328308e-01 -1.20281446e+00 -8.92722189e-01
3.68719280e-01 -1.19033284e-01 -9.64079261e-01 -3.89587194e-01
-8.07143271e-01 6.97440624e-01 4.46308367e-02 -1.74621008e-02
-1.56442535e+00 -1.97851613e-01 -1.08802319e+00 -8.63298848e-02
5.75280130e-01 -9.07010317e-01 1.03427851e+00 -7.03768730e-01
-1.06471705e+00 1.17061734e+00 -5.21629393e-01 -3.44945729e-01
4.63908702e-01 -5.34873128e-01 -1.88550621e-01 4.73827690e-01
2.56279677e-01 4.85047996e-01 7.23117411e-01 -9.52853024e-01
-5.99448621e-01 -1.04494107e+00 4.45451200e-01 8.16097260e-01
3.59646417e-02 -1.33775488e-01 -7.98844457e-01 -3.10937792e-01
4.74118114e-01 -1.11731017e+00 -9.70230997e-02 8.55081156e-02
-2.97815651e-01 -6.44552484e-02 7.44232893e-01 -6.33061409e-01
6.25688612e-01 -2.27972651e+00 5.40479600e-01 2.84502301e-02
1.45884827e-01 -3.21726769e-01 1.51753381e-01 3.97307158e-01
6.86830729e-02 -1.63436979e-01 1.87764406e-01 -4.74677116e-01
-5.15874773e-02 2.07870454e-01 -4.13243562e-01 9.58811402e-01
-2.75299370e-01 1.00029838e+00 -1.16715026e+00 -3.76533836e-01
4.92908359e-01 7.14423716e-01 -6.82358265e-01 3.55094522e-01
-1.24358982e-01 6.39312744e-01 -7.91328967e-01 8.16976011e-01
2.38769323e-01 -2.02486977e-01 2.53193587e-01 -1.18950717e-01
3.24378647e-02 1.39337093e-01 -1.22847748e+00 2.14317679e+00
-6.82360157e-02 7.91767001e-01 -1.48185074e-01 -8.58486414e-01
5.33117115e-01 5.61393440e-01 8.21818531e-01 -4.48739648e-01
2.17745919e-03 -4.90085900e-01 -6.99779868e-01 -1.12452078e+00
4.38823372e-01 1.67354107e-01 -1.98755458e-01 6.38318300e-01
2.18220979e-01 3.34991038e-01 -3.23831111e-01 4.40257400e-01
1.47906637e+00 7.44317770e-01 3.16673726e-01 2.41126955e-01
2.34855816e-01 -1.49885952e-01 4.26020920e-01 8.51791859e-01
-8.14656436e-01 4.97261673e-01 4.92337167e-01 -6.91818535e-01
-9.16721702e-01 -1.30696166e+00 3.55968684e-01 9.58060026e-01
4.31155115e-01 -5.31890571e-01 -5.90586007e-01 -8.66208076e-01
-1.80564523e-01 6.40099227e-01 -8.85316491e-01 -9.73127112e-02
-5.61162114e-01 6.43277839e-02 2.03238904e-01 7.04772353e-01
6.59510553e-01 -8.23947728e-01 -6.91993058e-01 -2.79698759e-01
-8.00752103e-01 -1.59121978e+00 -5.37169814e-01 -4.74723369e-01
-7.94612467e-01 -1.16496849e+00 -5.73186517e-01 -6.34801090e-01
6.77957237e-01 8.26121032e-01 9.48670983e-01 -6.08938813e-01
-1.91445529e-01 1.43295264e+00 -2.86869943e-01 4.39853221e-02
1.75600022e-01 -6.22154117e-01 5.27933240e-01 5.32993734e-01
8.55974436e-01 -7.66136289e-01 -8.54139328e-01 4.32179034e-01
-2.47025937e-01 3.35006341e-02 1.61113441e-01 1.59396753e-01
7.62291133e-01 3.45430113e-02 -1.56776711e-01 -5.31035662e-01
-1.27741739e-01 -7.34729648e-01 3.99436057e-03 2.11228475e-01
3.45510513e-01 -4.52352494e-01 -1.60608981e-02 -2.22985163e-01
-1.12615883e+00 5.98459005e-01 4.16835845e-01 -9.44141209e-01
-7.87780881e-01 -5.09832520e-03 -5.76732457e-01 5.18690407e-01
6.48178160e-01 4.03575778e-01 -3.24471205e-01 -3.49131554e-01
5.00506222e-01 4.76514280e-01 6.15597188e-01 -3.28197002e-01
6.43454909e-01 1.23239911e+00 -1.81671545e-01 -8.46087337e-01
-7.65194118e-01 -7.51871109e-01 -1.29460621e+00 -6.48370624e-01
1.31963646e+00 -1.42895234e+00 -6.93287969e-01 4.12494451e-01
-1.06017005e+00 -3.67239743e-01 -1.32811591e-01 8.10484767e-01
-1.11236322e+00 5.40020943e-01 -5.64726181e-02 -9.24926639e-01
4.42697078e-01 -1.03163576e+00 1.65972531e+00 -5.77518828e-02
-4.33916032e-01 -9.74797189e-01 1.56870648e-01 7.57907271e-01
-2.88873702e-01 4.00062889e-01 3.49017262e-01 -9.21369120e-02
-1.00222111e+00 -4.23508048e-01 7.56262243e-02 -9.19411033e-02
2.32356265e-01 -3.41925651e-01 -1.06997681e+00 4.30561863e-02
2.44696617e-01 -1.84734821e-01 6.88942432e-01 7.99442172e-01
1.00163960e+00 -1.36179641e-01 -6.38031721e-01 6.55843258e-01
9.69608545e-01 5.31636290e-02 5.07249475e-01 9.58409309e-02
9.61914062e-01 6.23357713e-01 7.52261996e-01 7.02979207e-01
6.78230524e-01 9.89548385e-01 5.75952232e-01 2.13604227e-01
3.69791165e-02 -6.73800647e-01 6.66656911e-01 3.58489379e-02
-3.41596961e-01 -3.09951276e-01 -8.22660506e-01 6.63254976e-01
-2.06437898e+00 -1.52345455e+00 4.29848611e-01 2.06156826e+00
1.00982912e-01 -2.34664977e-03 2.24770650e-01 -3.73040408e-01
6.73424542e-01 4.97700542e-01 -8.04892898e-01 1.35265753e-01
1.59363985e-01 -6.82610333e-01 3.84214789e-01 4.38333124e-01
-1.47853923e+00 1.04161584e+00 6.17065239e+00 -2.87845563e-02
-7.68605173e-01 -7.51833757e-03 3.13591808e-01 -6.17850661e-01
-9.37641561e-02 -4.50651981e-02 -6.21789455e-01 1.94192290e-01
6.84863269e-01 9.47312489e-02 3.30317646e-01 1.07626355e+00
7.36605048e-01 -1.96140915e-01 -1.43735480e+00 1.30394399e+00
6.06461227e-01 -1.14580119e+00 5.87207451e-02 3.23294073e-01
6.93398535e-01 1.85090914e-01 1.07184052e-01 2.22713590e-01
1.48686081e-01 -8.39366138e-01 6.78853810e-01 1.01312637e+00
3.91231745e-01 -3.70475322e-01 2.15032682e-01 6.53429806e-01
-1.11333752e+00 -2.26901487e-01 -1.15545146e-01 -3.02586734e-01
5.53220749e-01 -1.22487888e-01 -1.02664232e+00 -5.65936007e-02
8.42771173e-01 1.50203288e+00 -4.34931338e-01 7.74824917e-01
-2.93205023e-01 2.46203944e-01 -5.35685606e-02 4.41905648e-01
1.72665969e-01 -1.35705501e-01 1.02597535e+00 9.00833488e-01
2.57413507e-01 4.08881724e-01 4.60147619e-01 7.96494424e-01
7.19924495e-02 -2.72402316e-01 -1.32052135e+00 -1.64456740e-02
1.86137781e-01 7.91434050e-01 -6.39701068e-01 -2.43127123e-01
-2.93391138e-01 1.33870184e+00 2.51885891e-01 7.50085831e-01
-9.05542076e-01 2.95718461e-01 9.60793972e-01 2.72196919e-01
3.91337186e-01 -7.09352791e-01 2.36802354e-01 -1.60017979e+00
2.43899703e-01 -3.96060705e-01 3.05637330e-01 -1.33043408e+00
-6.30066037e-01 1.14761293e-01 1.40069172e-01 -1.56261361e+00
-5.19776523e-01 -4.21336889e-01 -4.63001192e-01 4.90580082e-01
-8.50051105e-01 -1.13442302e+00 -6.01918697e-01 8.69116664e-01
8.72932613e-01 -1.01137616e-01 6.26560807e-01 7.79463956e-03
-3.35315824e-01 1.40668571e-01 2.95487717e-02 1.30401835e-01
6.07493281e-01 -1.13942349e+00 6.66881263e-01 9.16791499e-01
6.68930769e-01 5.59850514e-01 4.89801735e-01 -7.96967268e-01
-1.81112063e+00 -1.13355577e+00 8.06953192e-01 -1.48704374e+00
4.47095037e-01 -5.92984378e-01 -4.59697843e-01 1.38525331e+00
-3.52428555e-01 8.14316720e-02 6.59399271e-01 6.99242279e-02
-2.65533239e-01 2.91416675e-01 -9.85365629e-01 7.63317466e-01
1.77699459e+00 -9.10601318e-01 -7.01405466e-01 6.62779927e-01
5.37498415e-01 -4.89491194e-01 -6.72973216e-01 4.90178727e-03
7.50192165e-01 -1.03511465e+00 1.18463016e+00 -8.38924229e-01
3.50284249e-01 -3.26663703e-01 -6.78090513e-01 -1.02938151e+00
-3.81455243e-01 -3.02290440e-01 -4.65021461e-01 7.65322089e-01
-1.96092010e-01 -1.69189759e-02 1.00286484e+00 9.43525493e-01
-3.20410170e-02 -4.57424283e-01 -9.99930024e-01 -3.80139917e-01
-6.42915905e-01 -7.57954597e-01 2.16571704e-01 6.82478309e-01
8.90739784e-02 8.00116435e-02 -7.22427726e-01 4.61671054e-01
6.32800162e-01 1.17323726e-01 1.32524669e+00 -7.34899104e-01
-2.53508806e-01 1.80268377e-01 -1.09607613e+00 -1.65958834e+00
3.48486423e-01 -6.19917631e-01 -4.34757248e-02 -1.55150008e+00
4.99485165e-01 3.48704755e-01 -1.00137003e-01 2.31423080e-01
1.23021044e-01 1.88118517e-01 -3.64765562e-02 4.50193048e-01
-1.13059819e+00 7.18170762e-01 1.10042393e+00 -1.73070639e-01
-2.65683293e-01 -1.20765883e-02 -4.01392519e-01 1.06051099e+00
3.33588451e-01 -1.93710044e-01 -6.91099524e-01 -6.95179939e-01
-1.23647355e-01 2.20420584e-01 8.20205629e-01 -8.94983649e-01
2.01396853e-01 -5.19110084e-01 6.16767466e-01 -8.69759083e-01
1.00069249e+00 -8.89669001e-01 -6.56618327e-02 4.05816250e-02
-3.85392874e-01 -1.19415082e-01 -2.84404039e-01 1.20280218e+00
1.35500610e-01 4.05665636e-01 1.39025703e-01 -5.39298952e-01
-1.24080467e+00 6.22286499e-01 -5.22274315e-01 3.05371042e-02
1.30713999e+00 -6.80171788e-01 -1.64205655e-01 -7.07675457e-01
-1.03638744e+00 2.41913661e-01 9.43780720e-01 6.17670059e-01
9.27819967e-01 -1.27741325e+00 -5.08776724e-01 3.58280301e-01
3.18812251e-01 -3.40258211e-01 8.18702042e-01 8.18864286e-01
-3.19677025e-01 6.70026481e-01 -2.36321971e-01 -1.01118767e+00
-1.31360936e+00 6.65462494e-01 4.84256446e-01 3.52564037e-01
-1.10298693e+00 7.57532954e-01 7.57216513e-01 -1.63011760e-01
4.82300997e-01 -1.52542219e-01 -3.63143355e-01 -2.84334630e-01
6.78563416e-01 4.81911600e-01 -5.43962061e-01 -1.44692338e+00
-5.49037516e-01 6.98728204e-01 1.62915334e-01 -3.01434755e-01
1.15559816e+00 -6.26357019e-01 3.10027719e-01 6.62609875e-01
1.20230258e+00 -1.19249158e-01 -1.85869229e+00 -5.09537756e-01
-4.32559550e-01 -1.11992049e+00 -1.04237255e-02 -3.20090503e-01
-7.32766390e-01 7.65820146e-01 6.25602126e-01 -3.95124853e-01
7.65906036e-01 4.65470910e-01 5.78755140e-01 4.75685120e-01
6.09025419e-01 -1.02647579e+00 4.03423697e-01 4.49114412e-01
7.29987919e-01 -1.25030768e+00 6.60556257e-02 -1.83097813e-02
-9.16293859e-01 7.14239597e-01 6.63520277e-01 -2.37997085e-01
6.55238807e-01 -4.80645955e-01 -1.79037824e-01 -5.18845081e-01
-7.55607426e-01 -1.92521140e-01 2.08493575e-01 8.13832819e-01
-5.48977964e-02 1.67689413e-01 6.10418558e-01 1.39112756e-01
6.29192069e-02 -8.29477087e-02 5.75745761e-01 9.44604039e-01
-3.48276466e-01 -4.02182579e-01 -3.36620599e-01 -1.26634017e-01
-1.26344010e-01 4.91458297e-01 -7.04726160e-01 6.00484312e-01
1.53193191e-01 8.73696327e-01 4.15587962e-01 -4.74719793e-01
3.74076337e-01 3.18343490e-01 6.38802052e-01 -7.35057592e-01
3.28820944e-01 4.53440510e-02 -5.98563068e-02 -1.24920940e+00
-6.16420209e-01 -1.19245195e+00 -1.00528872e+00 -1.67493266e-03
3.35758269e-01 -2.42047772e-01 5.58850467e-01 1.17304313e+00
5.68289459e-01 7.08912387e-02 6.53468430e-01 -1.22932649e+00
1.37620149e-02 -6.88714862e-01 -7.88714051e-01 6.86682999e-01
4.14521873e-01 -1.01289725e+00 -2.83485860e-01 3.57738733e-01] | [8.135749816894531, 0.40318143367767334] |
6ab7d7fa-74d1-44b3-90f1-57f0b50710ac | vpit-real-time-embedded-single-object-3d | 2206.02619 | null | https://arxiv.org/abs/2206.02619v1 | https://arxiv.org/pdf/2206.02619v1.pdf | VPIT: Real-time Embedded Single Object 3D Tracking Using Voxel Pseudo Images | In this paper, we propose a novel voxel-based 3D single object tracking (3D SOT) method called Voxel Pseudo Image Tracking (VPIT). VPIT is the first method that uses voxel pseudo images for 3D SOT. The input point cloud is structured by pillar-based voxelization, and the resulting pseudo image is used as an input to a 2D-like Siamese SOT method. The pseudo image is created in the Bird's-eye View (BEV) coordinates, and therefore the objects in it have constant size. Thus, only the object rotation can change in the new coordinate system and not the object scale. For this reason, we replace multi-scale search with a multi-rotation search, where differently rotated search regions are compared against a single target representation to predict both position and rotation of the object. Experiments on KITTI Tracking dataset show that VPIT is the fastest 3D SOT method and maintains competitive Success and Precision values. Application of a SOT method in a real-world scenario meets with limitations such as lower computational capabilities of embedded devices and a latency-unforgiving environment, where the method is forced to skip certain data frames if the inference speed is not high enough. We implement a real-time evaluation protocol and show that other methods lose most of their performance on embedded devices, while VPIT maintains its ability to track the object. | ['Alexandros Iosifidis', 'Anastasios Tefas', 'Nikolaos Passalis', 'Paraskevi Nousi', 'Illia Oleksiienko'] | 2022-06-06 | null | null | null | null | ['3d-single-object-tracking'] | ['computer-vision'] | [-2.68960688e-02 -3.84793997e-01 -2.04511061e-01 2.00664163e-01
-4.14702922e-01 -6.89061761e-01 4.57213491e-01 -9.42320153e-02
-6.56826138e-01 3.91220003e-01 -7.02835619e-01 -2.28549838e-01
1.36480227e-01 -5.58737099e-01 -8.24187279e-01 -6.79947138e-01
-9.90235880e-02 9.91273046e-01 1.31873834e+00 6.68748468e-02
3.13035518e-01 9.03753757e-01 -1.66116822e+00 -5.50534017e-02
3.82633209e-01 1.09966993e+00 4.54663724e-01 8.70382130e-01
-1.81363627e-01 2.61819750e-01 -6.61246598e-01 7.07451105e-02
6.80023611e-01 8.84358212e-03 -4.30418909e-01 -1.27279237e-01
6.75108552e-01 -3.50535184e-01 -9.82770920e-02 9.70150828e-01
4.35022742e-01 5.36727486e-04 2.95147121e-01 -1.50903749e+00
4.19264659e-02 1.32214531e-01 -7.64251828e-01 3.10252577e-01
4.58227873e-01 2.16069400e-01 4.64050829e-01 -1.06172144e+00
1.00370061e+00 1.29862201e+00 6.41556859e-01 4.92337137e-01
-1.15569985e+00 -6.96290910e-01 3.00291061e-01 4.15391624e-02
-1.59359479e+00 -6.78945184e-02 6.22369409e-01 -2.28387341e-01
9.22051489e-01 6.60057783e-01 9.56587434e-01 5.75366199e-01
3.94059122e-01 8.08163524e-01 9.79664266e-01 -2.95215964e-01
3.61335874e-01 -2.75200736e-02 2.80144699e-02 7.46762991e-01
3.72938246e-01 3.44911009e-01 -5.49146056e-01 -2.85190403e-01
1.01350403e+00 1.27137810e-01 -1.15424760e-01 -1.05992007e+00
-1.71417165e+00 5.34659624e-01 7.32097864e-01 2.48869270e-01
-1.73007637e-01 5.71585774e-01 1.84113711e-01 1.14145018e-01
4.04320449e-01 8.41184855e-02 -4.93508756e-01 -1.13884799e-01
-1.08213460e+00 5.14215052e-01 4.76183861e-01 1.11961663e+00
4.20868188e-01 -7.63037950e-02 -1.63731784e-01 1.05847165e-01
3.53025585e-01 9.53585207e-01 3.44380528e-01 -9.36722815e-01
2.04777151e-01 4.93343413e-01 1.88248590e-01 -9.31460917e-01
-4.59173650e-01 -2.27261618e-01 -3.44426960e-01 8.98158789e-01
4.65966195e-01 3.63363385e-01 -1.06213951e+00 1.31928039e+00
1.03568602e+00 2.19645083e-01 -2.70157188e-01 1.21005833e+00
7.51850605e-01 5.68592191e-01 -2.88977236e-01 -2.23492175e-01
1.48525345e+00 -9.43067670e-01 -4.62701946e-01 5.77195082e-03
4.97181028e-01 -7.79454529e-01 6.43568277e-01 4.47272062e-01
-1.14012671e+00 -6.81766033e-01 -9.92021084e-01 1.83338657e-01
-3.94356966e-01 -2.21578151e-01 4.43459749e-01 5.99197149e-01
-9.15972769e-01 4.26472336e-01 -1.18564236e+00 -2.89248407e-01
3.48687768e-01 7.26495981e-01 -1.01751283e-01 1.23010308e-01
-6.44896388e-01 8.56742799e-01 3.00624818e-01 -1.42715126e-01
-7.20623374e-01 -7.76015222e-01 -5.28227568e-01 -2.56089866e-01
6.78222001e-01 -6.15781248e-01 1.18875384e+00 -4.99719888e-01
-1.34497094e+00 8.15044820e-01 -2.54705369e-01 -4.44492877e-01
7.99205124e-01 -1.94652006e-01 -6.18878901e-02 1.41278729e-01
2.02628523e-01 8.14921141e-01 1.08219695e+00 -1.28454959e+00
-7.54847825e-01 -6.81552470e-01 -1.63386047e-01 2.91168511e-01
2.00903550e-01 -7.76567683e-02 -1.09543002e+00 -3.85668486e-01
5.09358168e-01 -1.34546542e+00 -2.88816005e-01 6.13728762e-01
-3.22635800e-01 -1.53622821e-01 1.44792867e+00 -5.65196462e-02
8.96681368e-01 -2.16595602e+00 8.14171694e-03 3.54080051e-01
3.69240701e-01 2.75815398e-01 1.43271670e-01 8.41035694e-03
2.61204869e-01 -2.76270598e-01 1.51528955e-01 -2.86890656e-01
-1.77217677e-01 2.19199926e-01 -1.36559054e-01 5.62670827e-01
-2.72562921e-01 8.17448199e-01 -1.00364363e+00 -8.66263747e-01
4.61042017e-01 5.02837956e-01 -4.66234505e-01 -2.58117795e-01
-2.73876071e-01 4.18954015e-01 -7.57189214e-01 8.08557928e-01
1.07410455e+00 -2.03496248e-01 -8.08431953e-02 -2.96029031e-01
-5.25448143e-01 -1.41064495e-01 -1.49817538e+00 1.52266157e+00
-6.69013262e-02 4.51149762e-01 -8.16136673e-02 -1.99072763e-01
8.40803266e-01 -5.39505035e-02 7.03800917e-01 -7.59511352e-01
1.25749618e-01 2.48931348e-01 -6.91299215e-02 -1.92658473e-02
6.88455343e-01 3.63584191e-01 3.07253413e-02 3.48705232e-01
-5.08493662e-01 -4.74923760e-01 -5.64382263e-02 1.52449086e-01
9.75317180e-01 4.47134554e-01 1.89995185e-01 -1.50826260e-01
3.33158195e-01 5.41336000e-01 4.52784926e-01 8.54566634e-01
-3.02838326e-01 5.77371061e-01 6.85499907e-02 -4.58909541e-01
-9.60975111e-01 -1.18541265e+00 -4.36514884e-01 8.44982624e-01
7.61686504e-01 -2.55314440e-01 -5.82673371e-01 -6.68620944e-01
4.61682230e-01 3.53751838e-01 -4.46413577e-01 1.81569561e-01
-8.95852685e-01 -1.73386231e-01 7.04542622e-02 4.99041110e-01
4.91950601e-01 -7.93818772e-01 -1.57796609e+00 2.18302622e-01
3.04100215e-01 -1.09740210e+00 -7.01616228e-01 3.76477599e-01
-1.21959054e+00 -1.03813255e+00 -6.98816478e-01 -6.70960665e-01
6.77634597e-01 6.88092709e-01 8.34826767e-01 1.70029894e-01
-2.37035975e-01 3.00515592e-01 -2.61677027e-01 -3.52136075e-01
-1.86170131e-01 -2.31521174e-01 3.36938411e-01 -4.96121585e-01
1.24167487e-01 -2.08065495e-01 -6.04704142e-01 6.50663912e-01
-6.04176521e-01 5.73002212e-02 3.08808774e-01 6.54157698e-01
1.12211275e+00 1.07101031e-01 -2.43279487e-01 -5.38290620e-01
7.30434731e-02 2.14040861e-01 -1.31037426e+00 -9.73275304e-03
-5.01730084e-01 -2.41360925e-02 1.31619558e-01 -9.59848523e-01
-5.28673351e-01 5.29078543e-01 2.59810299e-01 -1.18283045e+00
2.92379200e-01 -5.88925667e-02 6.79735988e-02 -5.30451417e-01
4.30109441e-01 8.02347586e-02 1.16157211e-01 -4.29284662e-01
1.67514250e-01 2.10258380e-01 6.90902770e-01 -2.93723434e-01
1.16133296e+00 7.92262375e-01 2.75954694e-01 -6.69280827e-01
-4.56453770e-01 -4.85369116e-01 -8.07678819e-01 -3.96909714e-01
8.03499699e-01 -5.42918861e-01 -1.07906282e+00 3.18853229e-01
-1.20665753e+00 -3.93241495e-01 -4.73499924e-01 6.05626822e-01
-2.96589434e-01 2.14618713e-01 -1.77779585e-01 -8.46603870e-01
-3.16597372e-01 -1.37218487e+00 1.48238897e+00 1.20695911e-01
-1.58304144e-02 -5.87303698e-01 1.32861331e-01 -1.40240356e-01
1.42105907e-01 2.85284698e-01 3.46140176e-01 -4.15149987e-01
-1.07234061e+00 -1.94974959e-01 -1.87610716e-01 -3.62593263e-01
-1.59374997e-01 5.64510264e-02 -7.36581147e-01 -5.00889003e-01
9.64090303e-02 2.59646922e-01 4.27847594e-01 4.93857175e-01
1.00963354e+00 1.44566104e-01 -9.49159443e-01 6.61379099e-01
1.40124404e+00 4.59680974e-01 2.28317603e-01 5.98613858e-01
6.71921015e-01 6.31451383e-02 1.12769687e+00 2.13225916e-01
1.86387643e-01 1.32558596e+00 7.83824503e-01 -9.54795852e-02
-3.67913187e-01 -8.65360722e-03 2.61176676e-01 4.89222527e-01
-1.82941988e-01 2.25663371e-02 -7.90047407e-01 2.25226104e-01
-1.80644596e+00 -9.21063781e-01 -5.36003351e-01 2.55125332e+00
4.82429802e-01 5.74180186e-01 2.41600230e-01 1.23308059e-02
5.97361565e-01 -8.73111635e-02 -9.65845048e-01 4.83775102e-02
1.70990884e-01 1.49580270e-01 8.51341784e-01 3.04341227e-01
-9.19625163e-01 9.87001657e-01 6.21447563e+00 6.72686636e-01
-1.20541608e+00 2.15190247e-01 -8.92081931e-02 -2.02245563e-01
2.35164389e-01 -2.65592914e-02 -1.06440949e+00 5.02092540e-01
4.67763901e-01 -5.32104038e-02 1.94632307e-01 1.11382926e+00
3.73843312e-02 -5.13772845e-01 -1.02281916e+00 1.23576939e+00
-2.10056100e-02 -1.07778049e+00 -2.78921455e-01 2.79687434e-01
3.93877864e-01 2.51527548e-01 2.24024206e-01 7.43062347e-02
8.04819241e-02 -3.97024095e-01 1.20143509e+00 2.31461853e-01
7.39088416e-01 -4.81961817e-01 4.20102000e-01 4.14957404e-01
-1.59532166e+00 1.71241224e-01 -3.81110162e-01 2.73968458e-01
1.43547624e-01 2.42867023e-01 -9.73865092e-01 2.88157582e-01
1.06956995e+00 2.15245739e-01 -6.36497498e-01 1.32845080e+00
1.87704355e-01 8.67875144e-02 -8.26976180e-01 -2.20767379e-01
1.54278129e-01 5.51897883e-02 1.03138793e+00 7.76580691e-01
3.97547603e-01 4.40315623e-03 3.59024048e-01 5.73339283e-01
2.76791573e-01 -2.54518151e-01 -5.76694667e-01 6.30308151e-01
6.63282394e-01 9.38396871e-01 -1.27048171e+00 -5.41669786e-01
-1.17774323e-01 9.57646012e-01 -1.92473620e-01 -6.12413697e-02
-1.12069428e+00 -7.85125047e-02 5.05427778e-01 2.83177048e-01
8.51102352e-01 -3.29645455e-01 -8.44467282e-02 -9.33313191e-01
2.50261370e-02 -5.23263931e-01 2.91775525e-01 -1.11373079e+00
-6.96361959e-01 7.50813127e-01 3.37368160e-01 -1.68679559e+00
-8.73958617e-02 -6.02257013e-01 -2.41290718e-01 4.38629538e-01
-1.36368370e+00 -9.75001276e-01 -4.78363186e-01 8.23619545e-01
6.04493380e-01 2.01982990e-01 4.04224038e-01 1.67767212e-01
-1.03143647e-01 4.36593294e-01 1.13664448e-01 -1.73184663e-01
4.54956055e-01 -1.08449209e+00 3.88229638e-01 5.76616824e-01
2.30017841e-01 3.46413016e-01 8.13991845e-01 -9.82871830e-01
-1.83752036e+00 -1.11962104e+00 4.15132850e-01 -9.32390690e-01
4.39359039e-01 -5.30770719e-01 -7.99642265e-01 6.40554130e-01
-2.57184207e-01 5.11843622e-01 -7.79683068e-02 -3.30192596e-01
-4.81019877e-02 -1.75673932e-01 -1.29935372e+00 5.75763583e-01
1.21031499e+00 -1.04066893e-01 -4.54751581e-01 3.18327397e-01
9.70064938e-01 -1.12802494e+00 -6.47799492e-01 3.17044616e-01
7.40708709e-01 -7.58836567e-01 1.34046316e+00 1.00143803e-02
-5.01452446e-01 -8.16496491e-01 -1.11107431e-01 -7.91033268e-01
-1.83735877e-01 -4.77350444e-01 -4.10440266e-01 6.85379684e-01
-1.44654870e-01 -5.77543318e-01 1.12601364e+00 2.72072434e-01
-1.16036370e-01 -6.06550694e-01 -1.34168684e+00 -1.11298430e+00
-4.30210859e-01 -4.41631913e-01 8.37664843e-01 5.95544636e-01
-4.82192874e-01 8.40246230e-02 -1.55829266e-02 4.43036139e-01
9.82496381e-01 4.12332773e-01 1.11361384e+00 -1.33912337e+00
1.63335241e-02 -2.14892089e-01 -6.58602118e-01 -1.25261128e+00
-3.73241127e-01 -6.56480849e-01 3.13244946e-02 -1.06681931e+00
8.62240717e-02 -8.01005840e-01 3.28116156e-02 3.36337775e-01
1.47764236e-01 4.07594979e-01 4.32758033e-01 5.31764090e-01
-6.99602246e-01 3.21675569e-01 1.40757549e+00 -1.86524168e-01
-4.68561172e-01 2.26329818e-01 1.95332229e-01 6.95335865e-01
4.10689682e-01 -7.53586709e-01 -2.55081832e-01 -4.36575770e-01
-2.43778884e-01 1.71370581e-01 6.96206212e-01 -1.28640258e+00
5.20455539e-01 -1.66237444e-01 6.19060934e-01 -1.46223235e+00
6.40621424e-01 -1.38511348e+00 4.53928083e-01 9.67989743e-01
2.92376637e-01 3.88365269e-01 2.09348604e-01 4.90072578e-01
2.86605984e-01 -1.22730657e-01 7.59864092e-01 -6.26278818e-02
-8.08619320e-01 5.28863192e-01 -2.63205785e-02 -3.96724254e-01
1.41999543e+00 -9.80212510e-01 -1.75972775e-01 1.77077264e-01
-5.90665400e-01 2.91764736e-01 9.36166942e-01 2.66670585e-01
8.05304945e-01 -1.52207851e+00 -2.31761739e-01 2.51008868e-01
4.86653484e-03 2.95148671e-01 2.11294852e-02 9.62845206e-01
-6.67353928e-01 5.66839039e-01 -5.95270135e-02 -1.37581825e+00
-1.74263787e+00 9.75534320e-01 2.27505222e-01 -2.97913790e-01
-1.08414507e+00 5.64990222e-01 1.87068656e-01 -2.76170224e-01
2.84313917e-01 -5.82128704e-01 9.69196334e-02 -2.91156858e-01
4.62637872e-01 2.72024363e-01 1.94768459e-01 -7.01918602e-01
-6.19977057e-01 9.75622118e-01 2.97214743e-02 -2.11484104e-01
1.08466244e+00 3.67111905e-04 3.00005764e-01 3.63538414e-01
8.92506480e-01 1.14254788e-01 -1.31087220e+00 -1.61469489e-01
-1.51443258e-01 -9.22959924e-01 2.04940990e-01 -3.56057316e-01
-1.08931708e+00 4.69592065e-01 1.11105812e+00 2.01878414e-01
8.05524647e-01 2.23407343e-01 7.43547380e-01 3.33563566e-01
9.26672876e-01 -6.79408550e-01 1.19092546e-01 4.05417085e-01
5.83294690e-01 -9.59036648e-01 3.67198914e-01 -4.86452341e-01
-3.53298515e-01 9.95785594e-01 8.32274079e-01 -1.39534473e-01
5.99971533e-01 4.67431158e-01 -1.10461392e-01 -4.82819557e-01
-5.13072848e-01 -1.51872486e-01 1.92028791e-01 7.16341555e-01
-3.47438782e-01 -2.00897992e-01 3.27632241e-02 -1.76042482e-01
-1.98575914e-01 -1.17247120e-01 8.32735375e-03 1.28972912e+00
-4.43676293e-01 -9.38829064e-01 -9.98963773e-01 2.14167997e-01
-3.74518931e-02 3.65864426e-01 -1.46660516e-02 1.25566053e+00
2.70167626e-02 3.21664363e-01 4.23733801e-01 -2.62597740e-01
4.99976456e-01 -4.70079213e-01 6.63984179e-01 -4.28457767e-01
-6.08593047e-01 4.26898986e-01 -4.34627146e-01 -9.44133043e-01
-5.12738943e-01 -8.08268011e-01 -1.60601127e+00 -2.67475069e-01
-5.84595501e-01 5.87950237e-02 1.00057328e+00 4.84032393e-01
3.05779785e-01 4.31799322e-01 5.83351552e-01 -1.37124538e+00
-3.01505029e-01 -3.66872191e-01 -3.36371601e-01 6.30090237e-02
3.93621713e-01 -1.16199315e+00 -1.75451502e-01 -1.77460730e-01] | [6.886017322540283, -2.280820369720459] |
d7543189-cff6-435a-bd60-6ba9b3be7197 | lets-lie-together-co-presence-effects-on | null | null | https://aclanthology.org/W12-0409 | https://aclanthology.org/W12-0409.pdf | Let's Lie Together: Co-Presence Effects on Children's Deceptive Skills | null | ['Marc Swerts'] | 2012-04-01 | null | null | null | ws-2012-4 | ['deception-detection'] | ['miscellaneous'] | [-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] | [-7.3394060134887695, 3.8966827392578125] |
92272d20-56c1-476b-88b5-356c4969b92d | temporal-patterns-in-insulin-needs-for-type-1 | 2211.07393 | null | https://arxiv.org/abs/2211.07393v2 | https://arxiv.org/pdf/2211.07393v2.pdf | Temporal patterns in insulin needs for Type 1 diabetes | Type 1 Diabetes (T1D) is a chronic condition where the body produces little or no insulin, a hormone required for the cells to use blood glucose (BG) for energy and to regulate BG levels in the body. Finding the right insulin dose and time remains a complex, challenging and as yet unsolved control task. In this study, we use the OpenAPS Data Commons dataset, which is an extensive dataset collected in real-life conditions, to discover temporal patterns in insulin need driven by well-known factors such as carbohydrates as well as potentially novel factors. We utilised various time series techniques to spot such patterns using matrix profile and multi-variate clustering. The better we understand T1D and the factors impacting insulin needs, the more we can contribute to building data-driven technology for T1D treatments. | ['Zahraa S. Abdallah', 'Isabella Degen'] | 2022-11-14 | null | null | null | null | ['type'] | ['speech'] | [ 3.77312660e-01 -4.86404866e-01 -6.74176216e-01 -3.87514681e-01
2.70406842e-01 -7.62619019e-01 9.80290994e-02 8.55809093e-01
9.16186199e-02 5.51231265e-01 6.20282948e-01 -3.23976070e-01
-3.65791261e-01 -7.69154966e-01 -4.29048747e-01 -6.43856645e-01
-5.29893637e-01 6.01498306e-01 -2.88534850e-01 -1.76035896e-01
2.98103005e-01 2.35448867e-01 -1.41175616e+00 1.97312042e-01
9.20342922e-01 1.11017621e+00 -2.37601519e-01 5.82104385e-01
-2.31884941e-01 2.82668561e-01 -9.85480174e-02 1.59845725e-01
5.10674417e-01 -1.07136655e+00 1.38424918e-01 8.42128508e-03
1.29176388e-02 4.64275479e-02 1.62177861e-01 8.41685057e-01
5.40499985e-01 -1.84922621e-01 3.85632187e-01 -1.09705186e+00
-2.48414293e-01 4.28464055e-01 -2.42283255e-01 2.08240032e-01
2.28760138e-01 5.19405425e-01 1.89877272e-01 1.13906637e-02
5.26517153e-01 1.22866356e+00 6.96130574e-01 2.59980470e-01
-1.62251544e+00 -5.90590298e-01 -1.34480819e-01 2.27004543e-01
-1.08698130e+00 -5.79443753e-01 4.18579280e-01 -5.73310792e-01
6.52301013e-01 4.53754961e-01 1.17055452e+00 9.50827241e-01
6.55227423e-01 1.09026641e-01 1.42698395e+00 -4.34065044e-01
3.21400523e-01 1.52381342e-02 -1.18747801e-01 9.67346802e-02
2.34879747e-01 2.45808169e-01 -5.97179592e-01 -1.25902876e-01
9.02768254e-01 4.32726562e-01 -1.90286353e-01 -5.10325193e-01
-1.24497104e+00 6.60052061e-01 -3.55184749e-02 1.21946529e-01
-5.55260241e-01 -1.54109895e-01 4.06436294e-01 5.60919464e-01
3.09096605e-01 1.62958592e-01 -1.07839620e+00 -4.03425932e-01
-4.29288626e-01 1.59258679e-01 8.45897377e-01 5.83358645e-01
4.15312290e-01 -3.29133481e-01 -6.48770342e-03 7.84972608e-01
5.21837234e-01 5.74312925e-01 3.76937985e-01 -6.89858973e-01
-7.34739825e-02 1.17654157e+00 4.45019752e-02 -9.82168436e-01
-6.68113053e-01 -2.52754986e-01 -8.14766526e-01 -5.44691645e-02
9.33449864e-01 -1.26315385e-01 -8.85454178e-01 1.58875751e+00
5.43407798e-01 2.05615774e-01 -1.30971059e-01 1.08247566e+00
2.72413462e-01 1.22429110e-01 3.93556803e-01 -6.66418910e-01
1.53326201e+00 -8.19357857e-02 -8.09455097e-01 3.29596996e-02
4.27567273e-01 -7.88926005e-01 3.63719851e-01 4.73967820e-01
-7.77239561e-01 -2.57180035e-01 -8.55774999e-01 2.34849304e-01
-7.13310659e-01 -4.64480639e-01 9.50863421e-01 7.98190832e-01
-3.92281413e-01 8.51925969e-01 -9.31286931e-01 -8.78297150e-01
2.40871355e-01 1.82336241e-01 -3.67204361e-02 -6.36944354e-01
-1.06657171e+00 9.82473433e-01 -3.76954488e-02 4.92088236e-02
-4.12324190e-01 -1.33612132e+00 -6.68032825e-01 -3.78599793e-01
4.73770738e-01 -9.55781877e-01 6.84029698e-01 -5.04088461e-01
-1.30576849e+00 5.80145776e-01 -4.18105163e-02 -4.16670889e-01
4.15128499e-01 8.06491375e-02 -6.09539628e-01 -2.49984965e-01
-8.87375921e-02 -1.06083222e-01 3.24447572e-01 -6.59158766e-01
-4.26387310e-01 -1.09911001e+00 -7.42809296e-01 -1.05854742e-01
6.18666470e-01 8.29387009e-02 1.19523421e-01 -4.73152101e-01
4.53464895e-01 -9.39829230e-01 -5.99261999e-01 5.84094860e-02
-7.39786401e-02 9.89727303e-03 3.24822515e-01 -5.94164193e-01
8.27746749e-01 -1.83817184e+00 3.85699481e-01 4.10805792e-01
1.45037398e-01 1.38648832e-03 4.96187538e-01 7.64443219e-01
-2.12296873e-01 3.83064419e-01 2.48017997e-01 6.65440857e-01
2.11744800e-01 4.24415141e-01 2.54543841e-01 6.33236885e-01
2.45237425e-01 7.00769663e-01 -9.90923405e-01 4.13774773e-02
5.80710828e-01 3.95082027e-01 -2.84408510e-01 3.05375364e-03
-5.44667721e-01 6.50038421e-01 -2.57552415e-01 9.03111517e-01
5.90357780e-01 1.80676460e-01 7.38068223e-01 -3.81058872e-01
-5.98800957e-01 -3.26081030e-02 -1.22764969e+00 1.48511386e+00
2.30553433e-01 1.20113112e-01 7.39695206e-02 -9.02174234e-01
1.02752197e+00 1.91340059e-01 1.10047591e+00 -1.32058883e+00
3.28704745e-01 1.65040553e-01 4.39780146e-01 -8.36924493e-01
-6.34952664e-01 -5.38275361e-01 2.32046202e-01 8.05061385e-02
-3.75917614e-01 4.91024464e-01 4.01715398e-01 -2.55118489e-01
1.28184736e+00 9.52039734e-02 5.39315343e-01 -4.60038245e-01
1.63899273e-01 2.41512343e-01 1.01642191e+00 4.27380763e-02
-5.39368451e-01 3.19155864e-02 8.09001625e-01 -7.39415586e-01
-1.12957907e+00 -9.63257194e-01 -3.86934668e-01 4.57440436e-01
-1.39628842e-01 -1.70549586e-01 3.36728632e-01 1.83700442e-01
7.50261903e-01 3.87652040e-01 -7.10986018e-01 -1.63634270e-01
-5.66119105e-02 -1.14410579e+00 -5.99328168e-02 1.31666839e-01
2.05401912e-01 -3.98319691e-01 -4.08742130e-01 7.97620952e-01
1.56304047e-01 -4.77541298e-01 -2.46051490e-01 5.78471899e-01
-1.03786433e+00 -1.23916888e+00 -3.56647432e-01 -2.27810711e-01
3.26925516e-01 -1.91745296e-01 1.15872526e+00 -2.04859748e-01
-8.83972347e-01 -8.53164047e-02 -3.98199022e-01 -9.13827240e-01
-2.84103245e-01 -5.76728404e-01 2.33734220e-01 6.77881315e-02
9.84799683e-01 -9.13988292e-01 -1.12299526e+00 5.18430948e-01
-6.52015448e-01 3.81455437e-04 5.81193686e-01 8.86450186e-02
8.78566742e-01 2.59148091e-01 6.35012865e-01 -5.94926834e-01
3.11075121e-01 -1.04380357e+00 -4.43378597e-01 -9.64096636e-02
-9.70314801e-01 -1.46382540e-01 -4.31152806e-02 -4.81770217e-01
-2.08166391e-01 7.12598022e-03 4.62572932e-01 -4.83080819e-02
-3.17888647e-01 5.40276706e-01 -2.19014883e-01 1.92600891e-01
4.34656382e-01 1.33943066e-01 5.46912253e-01 -5.29156685e-01
8.16599932e-03 3.30956072e-01 1.62180215e-01 -5.59113801e-01
2.23748207e-01 2.28070378e-01 3.57520729e-01 -7.34464109e-01
-4.05985385e-01 -7.21245289e-01 -5.08663058e-01 -3.60266715e-01
9.23533380e-01 -8.89667988e-01 -1.06919503e+00 3.67892146e-01
-1.96051568e-01 -5.69789886e-01 1.24941066e-01 6.59272194e-01
-4.30601746e-01 -2.72850215e-01 -1.97160736e-01 -7.79385567e-01
-3.46823931e-02 -6.62755728e-01 2.92743027e-01 3.65952224e-01
-3.06501657e-01 -9.75728095e-01 4.48063105e-01 2.60131001e-01
8.66638899e-01 1.06655240e+00 1.34782958e+00 -4.10398751e-01
-4.57233727e-01 2.39154339e-01 -1.14305995e-01 -8.01126659e-02
4.92345572e-01 6.30145147e-02 -7.68700242e-02 1.67138070e-01
5.87679707e-02 5.40690348e-02 3.56929779e-01 8.50483119e-01
4.10334140e-01 -2.49286830e-01 -2.71887064e-01 6.30794048e-01
1.88382268e+00 7.93505669e-01 8.85978222e-01 4.63251948e-01
3.22549611e-01 5.88324726e-01 4.05678332e-01 7.42513299e-01
4.03094590e-01 5.05368352e-01 1.35040268e-01 -2.67354965e-01
1.42481342e-01 6.50314540e-02 -3.91617268e-02 3.44383210e-01
-5.71964569e-02 2.15461552e-01 -8.51924777e-01 4.85824406e-01
-1.75284839e+00 -9.69183803e-01 -7.62613237e-01 2.25549269e+00
1.18100703e+00 -1.52449608e-01 4.17570740e-01 1.15258722e-02
3.85715246e-01 -7.11132228e-01 -1.01754153e+00 -5.64996481e-01
-6.17682524e-02 2.96470225e-01 8.53748798e-01 -1.08956754e-01
-5.48184752e-01 1.73723832e-01 6.36722708e+00 -3.83707106e-01
-1.11066985e+00 -4.13295180e-01 5.15947402e-01 -3.86231929e-01
6.09716438e-02 -7.10496232e-02 -3.36787373e-01 6.21010423e-01
1.40189409e+00 -3.04809511e-01 7.53316343e-01 2.94412822e-01
1.14840329e+00 -5.69128931e-01 -1.09470010e+00 8.70379448e-01
-3.02404583e-01 -1.09359646e+00 -6.59731746e-01 2.50658900e-01
4.88719314e-01 1.07830629e-01 -4.66132075e-01 2.11134225e-01
2.94929594e-01 -8.06159496e-01 3.56864668e-02 8.72522712e-01
4.51851308e-01 -6.65976584e-01 5.89163482e-01 -1.43352628e-01
-9.48679090e-01 -1.34604737e-01 -1.31832197e-01 -2.58507133e-01
1.54132292e-01 1.03065586e+00 -6.09103739e-01 3.26176673e-01
6.78711236e-01 8.98465395e-01 -5.28354645e-01 1.48075461e+00
4.29616868e-01 4.51247245e-01 -2.94012517e-01 2.98131734e-01
-1.61000282e-01 -8.26890945e-01 3.95511180e-01 6.70461297e-01
3.12387258e-01 4.26495850e-01 3.39801759e-01 6.52893901e-01
3.84496629e-01 3.08859050e-01 -4.56182480e-01 -4.41383958e-01
9.09091532e-02 9.94510293e-01 -7.97274649e-01 -8.37222561e-02
-4.07793522e-01 5.39055347e-01 -4.70141739e-01 1.87643930e-01
-5.38257897e-01 1.12193421e-01 1.34135818e+00 4.19727445e-01
4.10436764e-02 -3.70315790e-01 -3.33009720e-01 -8.89290750e-01
-3.43813926e-01 -1.13076603e+00 7.48502195e-01 -4.05511409e-01
-1.48132074e+00 -5.09317338e-01 -2.48256043e-01 -8.90204728e-01
7.66439810e-02 -4.79709566e-01 -1.73558548e-01 1.09625340e+00
-1.47058606e+00 -8.78112078e-01 -3.33561778e-01 4.85327065e-01
3.46290648e-01 3.52049589e-01 9.52218592e-01 3.40428621e-01
-8.17004204e-01 -2.28260845e-01 3.10519278e-01 -3.40023667e-01
1.09851587e+00 -1.29896247e+00 -4.58477736e-01 2.56245494e-01
-7.35986829e-01 7.57031918e-01 9.62551296e-01 -1.08237553e+00
-2.15682316e+00 -9.12175179e-01 6.49414420e-01 -3.19701552e-01
7.09078312e-01 -3.12122583e-01 -6.83348954e-01 3.17424297e-01
-1.19828982e-02 -3.01796198e-01 1.43625331e+00 2.63040185e-01
-1.40906394e-01 -3.04206014e-01 -1.44190085e+00 3.08109611e-01
7.74195313e-01 2.52289951e-01 -2.42911652e-01 2.34117433e-01
3.19164813e-01 -2.12038055e-01 -1.78174579e+00 1.94165617e-01
9.36016500e-01 -9.27901030e-01 8.07698786e-01 -8.21702123e-01
2.48070657e-01 -3.58336866e-01 -2.09569216e-01 -1.42669332e+00
-1.59922525e-01 -6.86894238e-01 4.51577455e-02 1.17418623e+00
3.04310560e-01 -6.42796338e-01 5.42734802e-01 1.20607805e+00
3.09681073e-02 -5.77022254e-01 -6.87327564e-01 -5.75932682e-01
-2.12516531e-01 -1.41959026e-01 7.57200718e-01 1.23336267e+00
2.63682991e-01 -7.16298968e-02 -2.31078893e-01 3.36024649e-02
8.83526623e-01 1.39755175e-01 6.04670525e-01 -1.42079616e+00
1.47431105e-01 -4.23547059e-01 -6.07046366e-01 9.82785746e-02
-8.85290563e-01 -8.56490254e-01 -2.63251096e-01 -1.49349761e+00
1.64043102e-02 -7.18668997e-01 -4.75125581e-01 3.53890687e-01
1.90127522e-01 -1.87053695e-01 -3.72542925e-02 -2.61146575e-01
-8.96836594e-02 5.91325946e-02 1.17161155e+00 7.75851309e-02
-6.91341102e-01 -1.65679961e-01 -9.32987332e-01 6.55675754e-02
9.72604156e-01 -5.99401534e-01 -4.23322976e-01 1.00537531e-01
3.15912724e-01 3.89661372e-01 4.10318375e-02 -7.87366927e-01
5.11522368e-02 -9.04724658e-01 9.73404944e-01 -6.14169538e-01
-4.05171156e-01 -1.07030666e+00 8.61774385e-01 7.54722595e-01
-1.37079090e-01 -9.11432132e-02 3.02018821e-01 5.76167285e-01
4.58683163e-01 4.63136762e-01 6.06917620e-01 -4.19164002e-01
-4.66448903e-01 2.02941209e-01 -5.35080552e-01 -4.80239019e-02
1.16328764e+00 -6.26910552e-02 -3.49155307e-01 3.93403322e-02
-1.01162314e+00 6.40658915e-01 6.70249760e-01 6.30496383e-01
1.14837430e-01 -1.44640028e+00 -8.59723866e-01 3.97554219e-01
2.17348382e-01 -3.98300767e-01 4.68034208e-01 1.28675902e+00
-3.14713985e-01 3.06178421e-01 -6.60971701e-01 -7.47576594e-01
-9.77093339e-01 8.02088380e-01 4.02759612e-01 1.89813301e-01
-7.10751653e-01 4.66493145e-02 -6.60189986e-01 -1.76387221e-01
1.69230159e-02 -6.37039840e-01 -6.53642118e-02 3.29469442e-01
7.29876041e-01 5.83475947e-01 -1.12651542e-01 2.59095244e-02
-3.99306059e-01 5.88572979e-01 4.23672557e-01 6.09441161e-01
1.67695594e+00 -5.15334547e-01 -4.68325764e-01 9.27351892e-01
8.07014346e-01 -3.92519593e-01 -8.88117194e-01 1.30777275e-02
2.49820799e-01 -6.59045398e-01 1.20900661e-01 -1.41542554e+00
-7.62625217e-01 3.74697261e-02 1.22850811e+00 3.60884428e-01
1.08923948e+00 -1.45575538e-01 6.34783506e-01 -3.96841675e-01
2.49338046e-01 -8.47918391e-01 -3.59231502e-01 -1.41617700e-01
7.46754766e-01 -1.11631656e+00 6.81265667e-02 -2.15132460e-01
-2.91940421e-01 1.08952463e+00 1.91733211e-01 5.54486103e-02
6.90697312e-01 1.98359549e-01 4.92380649e-01 -3.13475370e-01
-1.06043577e+00 -3.69707853e-01 -2.44832337e-01 7.49657989e-01
4.21247184e-01 2.93009520e-01 -5.56227684e-01 3.93591136e-01
1.62018821e-01 5.85378587e-01 2.28833392e-01 8.14328015e-01
-4.21390384e-01 -1.37842810e+00 -4.15291756e-01 1.00569880e+00
-3.37779671e-01 2.52367586e-01 -2.32808098e-01 6.07259393e-01
5.38044155e-01 1.01990664e+00 1.06013976e-01 9.13678557e-02
6.91034198e-01 5.42630613e-01 4.05127972e-01 -3.57340187e-01
-3.66776824e-01 3.84556204e-01 1.74678206e-01 -8.02002847e-01
-4.28199261e-01 -1.09710491e+00 -1.15677702e+00 -5.27019799e-01
2.78933227e-01 -3.67103189e-01 1.28947949e+00 8.32511783e-01
5.79197347e-01 5.88329732e-01 5.67985058e-01 -5.83064675e-01
-2.08523363e-01 -8.99745166e-01 -9.94098306e-01 5.61686158e-01
1.78249702e-01 -7.33712256e-01 -2.91978627e-01 2.39774778e-01] | [8.123228073120117, 5.58479642868042] |
7e0e4e70-19c9-46e1-b6b1-1d610c04c686 | rtm3d-real-time-monocular-3d-detection-from | 2001.03343 | null | https://arxiv.org/abs/2001.03343v1 | https://arxiv.org/pdf/2001.03343v1.pdf | RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving | In this work, we propose an efficient and accurate monocular 3D detection framework in single shot. Most successful 3D detectors take the projection constraint from the 3D bounding box to the 2D box as an important component. Four edges of a 2D box provide only four constraints and the performance deteriorates dramatically with the small error of the 2D detector. Different from these approaches, our method predicts the nine perspective keypoints of a 3D bounding box in image space, and then utilize the geometric relationship of 3D and 2D perspectives to recover the dimension, location, and orientation in 3D space. In this method, the properties of the object can be predicted stably even when the estimation of keypoints is very noisy, which enables us to obtain fast detection speed with a small architecture. Training our method only uses the 3D properties of the object without the need for external networks or supervision data. Our method is the first real-time system for monocular image 3D detection while achieves state-of-the-art performance on the KITTI benchmark. Code will be released at https://github.com/Banconxuan/RTM3D. | ['PengFei Liu', 'Feidao Cao', 'Peixuan Li', 'Huaici Zhao'] | 2020-01-10 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/1054_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123480647.pdf | eccv-2020-8 | ['vehicle-pose-estimation'] | ['computer-vision'] | [-4.03429925e-01 -3.35398793e-01 -3.83688807e-01 -3.22715975e-02
-2.34691426e-01 -5.65157771e-01 3.58081430e-01 -3.87702316e-01
-3.76213908e-01 -2.42824033e-02 -9.97923985e-02 -3.42098087e-01
4.50892746e-01 -5.41145146e-01 -7.88518906e-01 -4.36347604e-01
2.51660734e-01 4.70577776e-01 8.15007389e-01 1.28956530e-02
4.04057831e-01 7.68860698e-01 -1.43189168e+00 -1.78997710e-01
2.22625330e-01 1.22712374e+00 4.76576626e-01 6.64072573e-01
2.34944165e-01 3.45038354e-01 -3.09197634e-01 1.98299319e-01
8.75207782e-01 -7.00438768e-02 -1.22020818e-01 3.11759859e-01
6.48417652e-01 -9.15027618e-01 -8.69944751e-01 1.18503702e+00
5.75210571e-01 -2.65946329e-01 5.80315351e-01 -1.10057449e+00
-3.86425316e-01 -5.10407053e-02 -9.44270074e-01 -4.25940473e-03
5.62451780e-01 2.36721471e-01 7.18470633e-01 -1.38174105e+00
8.09268534e-01 1.06159520e+00 4.80583578e-01 3.75467211e-01
-7.05608964e-01 -7.14113891e-01 2.91490287e-01 2.15005144e-01
-1.64902949e+00 -3.10837537e-01 7.40713656e-01 -3.55458587e-01
1.00900149e+00 -9.00867209e-02 8.69886279e-01 6.98896945e-01
1.67525206e-02 1.06530905e+00 7.95086443e-01 -3.58190805e-01
-4.62464094e-02 7.08233267e-02 8.56133252e-02 1.02456224e+00
3.41264516e-01 3.37610751e-01 -5.04550099e-01 7.62254298e-02
1.02457607e+00 4.36081558e-01 -2.75617748e-01 -1.01045334e+00
-1.13575661e+00 5.43206871e-01 7.14576185e-01 -1.28746659e-01
-8.47610980e-02 7.78683349e-02 -3.23232519e-03 -3.94088775e-02
3.45583707e-01 1.06920756e-01 -2.20441878e-01 1.33198518e-02
-5.33397257e-01 1.87570065e-01 4.30452675e-01 1.42806900e+00
6.09126389e-01 -2.20764652e-01 2.06932575e-01 2.87130803e-01
2.03343809e-01 7.95144141e-01 1.48001954e-01 -9.06180501e-01
4.83190805e-01 9.46333170e-01 3.39125603e-01 -1.02350473e+00
-6.07588053e-01 -3.00102741e-01 -5.38356185e-01 4.12712216e-01
4.87243056e-01 2.28653233e-02 -9.90725219e-01 1.12792349e+00
6.79682672e-01 -8.93295854e-02 -3.90130788e-01 1.28148341e+00
8.47346723e-01 5.21389842e-01 -9.70960617e-01 1.21606454e-01
1.26193810e+00 -9.01856005e-01 -1.13413304e-01 -5.27833819e-01
5.17928243e-01 -8.58402908e-01 8.24704349e-01 2.46796936e-01
-9.79010165e-01 -4.53276068e-01 -1.16574931e+00 -2.38098681e-01
-1.10431738e-01 5.27435243e-01 4.93060768e-01 2.37836123e-01
-6.43174231e-01 -6.59783790e-03 -8.23091149e-01 -3.75663131e-01
3.56002986e-01 2.49750957e-01 -3.57390195e-01 -2.86168993e-01
-6.71620011e-01 1.10650098e+00 3.39468896e-01 1.71678495e-02
-6.99804842e-01 -3.98747891e-01 -8.19704056e-01 -1.71124153e-02
8.50854635e-01 -7.37927198e-01 1.20415831e+00 -1.05215073e-01
-1.29449069e+00 1.11066616e+00 -1.85955212e-01 -3.06129515e-01
8.06709647e-01 -4.68277246e-01 2.53720939e-01 2.30971649e-01
5.57094812e-02 6.73558891e-01 8.14460158e-01 -1.03571045e+00
-7.83993065e-01 -7.46769190e-01 1.55890629e-01 4.70429808e-01
1.97380707e-02 -7.46682957e-02 -8.41852546e-01 -3.48443314e-02
8.85757148e-01 -1.03220189e+00 -3.30898017e-02 5.34513712e-01
-6.64662480e-01 -2.45981663e-01 9.99285400e-01 -1.48648337e-01
7.65601218e-01 -2.16609883e+00 -9.81128961e-02 -2.36389711e-01
3.13953608e-01 3.86102945e-01 2.83513814e-01 1.74330547e-01
3.63309979e-01 -1.85851485e-01 2.21278474e-01 -2.10730776e-01
-1.27898052e-01 -3.20034653e-01 -2.46878907e-01 9.51068938e-01
-8.88673589e-02 7.82595575e-01 -8.18775475e-01 -2.41098031e-01
5.05467892e-01 3.88131768e-01 -5.29617131e-01 1.99624017e-01
-4.89073060e-02 1.32010266e-01 -6.34429634e-01 9.02112782e-01
8.50853205e-01 -2.95600712e-01 -3.27594072e-01 -2.72417188e-01
-2.78777391e-01 3.41829807e-01 -1.29533780e+00 1.60440361e+00
-1.47926584e-01 5.90755224e-01 -1.22889623e-01 -6.41216099e-01
1.09528506e+00 -4.22546603e-02 2.58677125e-01 -5.16591489e-01
3.68178219e-01 2.36227944e-01 -2.02222005e-01 -2.70231187e-01
3.35350722e-01 2.56736040e-01 -5.95092438e-02 3.95803660e-01
-2.22100273e-01 -5.37470400e-01 -8.82513821e-02 7.65090361e-02
7.63970554e-01 3.05511415e-01 7.40895808e-01 2.19283044e-01
2.04414681e-01 1.44822737e-02 7.10346818e-01 8.21738064e-01
-3.04107994e-01 8.13936472e-01 4.35442060e-01 -5.31884015e-01
-1.17283046e+00 -9.49696958e-01 -1.84438750e-01 2.97705263e-01
7.19405770e-01 -3.74153435e-01 -3.74052674e-01 -7.99619079e-01
2.65352309e-01 2.52979428e-01 -4.00120556e-01 -9.50598493e-02
-5.33879578e-01 -3.28230739e-01 6.30379990e-02 4.82138455e-01
5.14833450e-01 -3.74911875e-01 -1.14744711e+00 -3.06041211e-01
2.78372485e-02 -1.39699101e+00 -7.06375897e-01 1.56129196e-01
-7.81128943e-01 -1.40759456e+00 -5.41336894e-01 -7.54945815e-01
7.66173959e-01 1.11091626e+00 5.74628770e-01 5.39989918e-02
-3.86983424e-01 4.31744270e-02 -2.96242118e-01 -4.71254170e-01
1.81330964e-01 -2.40070954e-01 3.10057759e-01 -3.65394711e-01
6.15586698e-01 -3.29010636e-01 -8.34024549e-01 7.39745378e-01
-1.70255482e-01 3.88525039e-01 4.61132109e-01 7.67009079e-01
7.58527935e-01 -1.25832647e-01 -1.54339477e-01 -4.79666024e-01
-3.81011277e-01 -1.69091508e-01 -1.10721612e+00 -2.22356066e-01
-3.52234900e-01 -3.47669929e-01 4.54824924e-01 -4.38873738e-01
-6.58436358e-01 7.01072693e-01 3.51972401e-01 -7.95281231e-01
-1.94394633e-01 -3.64981107e-02 -9.05572996e-02 -1.56054467e-01
7.14745581e-01 2.81595796e-01 -1.47966832e-01 -5.98928154e-01
2.87272990e-01 7.27781415e-01 3.70047390e-01 2.43721046e-02
1.02318335e+00 1.00708425e+00 2.54424691e-01 -6.93524420e-01
-1.14709234e+00 -8.41462433e-01 -1.11875510e+00 -2.14759156e-01
5.67153811e-01 -1.23085511e+00 -6.82772875e-01 5.31983137e-01
-1.28677535e+00 -2.68031470e-02 4.62405048e-02 7.41089523e-01
-5.50279558e-01 4.50947404e-01 -5.48809052e-01 -7.81395376e-01
-2.39924729e-01 -1.17003000e+00 1.23127615e+00 1.38403475e-01
1.29043967e-01 -2.16723338e-01 -3.65861386e-01 2.26648524e-01
-3.65317687e-02 1.17774509e-01 3.91493857e-01 -1.44354716e-01
-1.12995553e+00 -7.17993140e-01 -4.52592403e-01 7.64465779e-02
-5.70897907e-02 -8.23475197e-02 -8.92619371e-01 -2.47598454e-01
2.92242348e-01 -2.61619627e-01 9.57208753e-01 4.61512625e-01
8.30669165e-01 1.35086834e-01 -5.11185408e-01 1.01656806e+00
1.21225321e+00 2.07298249e-01 8.50907192e-02 3.09993088e-01
8.40750754e-01 2.74906546e-01 9.82156932e-01 5.45976996e-01
3.81016552e-01 9.50972378e-01 6.48717761e-01 1.36331275e-01
-1.74939796e-01 -5.49413025e-01 3.65754277e-01 4.10343647e-01
1.05251484e-01 -2.69231088e-02 -9.30453956e-01 2.54257053e-01
-1.76165926e+00 -7.32103109e-01 -1.23720840e-01 2.28921151e+00
3.40299457e-01 5.05548000e-01 1.52900249e-01 -1.02124423e-01
6.57349169e-01 1.36690900e-01 -9.45881665e-01 2.68774509e-01
1.88841857e-02 -6.31384134e-01 8.04542661e-01 4.64970767e-01
-1.15859520e+00 1.07639539e+00 5.51509714e+00 5.02424479e-01
-1.13477504e+00 -2.14904994e-01 2.18940035e-01 -5.32104254e-01
3.74376655e-01 2.04032436e-01 -1.47941434e+00 2.90096581e-01
1.25240739e-02 -2.30577569e-02 2.52315044e-01 1.10459518e+00
1.75396934e-01 -4.22468066e-01 -1.22886634e+00 1.33044767e+00
3.20863217e-01 -1.16094434e+00 -1.91021353e-01 1.58715978e-01
6.04478657e-01 4.29231793e-01 5.69273382e-02 -1.40427738e-01
-1.27715886e-01 -5.75136721e-01 9.49823141e-01 1.71112224e-01
8.07229757e-01 -5.55782318e-01 6.28094316e-01 9.48752701e-01
-1.22129738e+00 -1.43815324e-01 -8.48359227e-01 -3.66848499e-01
1.30705401e-01 5.49012482e-01 -1.07640445e+00 6.42880425e-02
7.56192029e-01 8.93353820e-01 -4.45691794e-01 1.31411028e+00
-5.82709968e-01 4.39628214e-02 -6.06364071e-01 -1.10368520e-01
5.67364022e-02 -1.03987135e-01 1.00267971e+00 7.72553265e-01
4.86565948e-01 2.40609333e-01 4.05835181e-01 8.10414374e-01
-1.14058480e-01 -1.79635689e-01 -9.72149670e-01 3.33823651e-01
7.08264351e-01 1.09832561e+00 -7.62795150e-01 -2.55332828e-01
-5.54657876e-01 7.82917321e-01 2.74252594e-01 3.94316278e-02
-6.68744147e-01 -3.25577050e-01 4.48318303e-01 3.01568478e-01
6.04591906e-01 -5.20578682e-01 -2.87930518e-01 -1.43880045e+00
3.92967284e-01 -4.89028454e-01 1.00644700e-01 -9.45660472e-01
-8.51936996e-01 4.50198442e-01 -1.12225458e-01 -1.68564951e+00
6.45949924e-03 -8.70300531e-01 -3.38018149e-01 6.71687663e-01
-1.40734065e+00 -9.86372232e-01 -5.32337785e-01 3.13227087e-01
4.00505394e-01 -1.01097062e-01 4.34585750e-01 -1.68615226e-02
-3.81295621e-01 2.89401501e-01 -6.65543750e-02 2.09989756e-01
7.41605401e-01 -9.56247211e-01 7.26173818e-01 9.27702010e-01
2.23959342e-01 3.62535506e-01 4.48031783e-01 -5.94369709e-01
-1.73154616e+00 -7.64019430e-01 7.87069321e-01 -8.61929655e-01
4.66528893e-01 -8.22458923e-01 -5.85192680e-01 6.02573156e-01
-4.98796791e-01 3.44874948e-01 -1.45694278e-02 -1.39715031e-01
-5.67937016e-01 -4.61610742e-02 -7.84978867e-01 6.48203254e-01
1.54113722e+00 -5.28775275e-01 -5.43193460e-01 5.16592860e-01
7.09327996e-01 -9.63034332e-01 -3.26970965e-01 4.15704668e-01
6.33194983e-01 -1.17524600e+00 1.12709653e+00 -1.61714211e-01
5.96144833e-02 -6.55651450e-01 -1.09170631e-01 -8.96704078e-01
-1.42034084e-01 -3.57588828e-01 -3.84794176e-01 4.90386009e-01
9.43063870e-02 -4.75612849e-01 1.10459459e+00 2.79901534e-01
-9.51364711e-02 -8.51006210e-01 -9.73743618e-01 -9.23777401e-01
-4.18435991e-01 -4.13186401e-01 3.21089238e-01 4.14111763e-01
-3.08553036e-02 4.36738282e-01 -5.40388823e-01 3.78246665e-01
5.90902030e-01 7.19723046e-01 1.20513511e+00 -1.03671026e+00
-2.88899422e-01 -3.78914177e-01 -7.15146363e-01 -2.01758170e+00
-2.42635533e-01 -7.01334536e-01 1.35291582e-02 -1.20438707e+00
3.98488671e-01 -2.86563754e-01 4.94715460e-02 2.23240212e-01
5.78859868e-03 3.32939148e-01 4.17092115e-01 3.94814700e-01
-6.18941486e-01 5.58267534e-01 1.40046835e+00 9.00414810e-02
-2.48907596e-01 2.19550386e-01 -3.87272686e-01 1.10930824e+00
6.73962653e-01 -4.47797656e-01 -1.54602721e-01 -5.32503307e-01
1.40371462e-02 1.22690484e-01 4.65237945e-01 -8.09373915e-01
5.41235805e-01 -1.48793608e-01 7.44706154e-01 -1.25765049e+00
6.78474009e-01 -8.63941312e-01 -3.39731991e-01 5.25540233e-01
2.22629353e-01 -1.35827959e-01 -1.03756681e-01 5.10531902e-01
7.70901442e-02 -1.04064681e-01 8.58275115e-01 -2.77771264e-01
-7.73675561e-01 6.32522821e-01 1.23543195e-01 -7.10094124e-02
1.18419039e+00 -4.90631789e-01 -5.15541494e-01 -3.66062164e-01
-3.00747931e-01 4.07895684e-01 9.73599851e-01 4.61611003e-01
1.05976152e+00 -1.16440737e+00 -5.52959859e-01 4.02821481e-01
1.72681555e-01 4.28042978e-01 1.22972742e-01 8.12331259e-01
-6.45438910e-01 6.34125769e-01 -8.39827880e-02 -9.59559023e-01
-1.33917665e+00 8.47244620e-01 5.57313263e-01 3.72311205e-01
-1.02568471e+00 8.65899682e-01 6.01266980e-01 -2.43965402e-01
4.01306808e-01 -2.73865998e-01 1.53072670e-01 -1.46201387e-01
7.17094064e-01 2.41996318e-01 -1.87180042e-01 -6.65521383e-01
-4.90992665e-01 1.01703346e+00 -1.81198850e-01 2.70593584e-01
1.20362806e+00 -2.34142512e-01 3.12543184e-01 4.52135533e-01
1.16392386e+00 -4.91438955e-02 -1.62972367e+00 -5.61544776e-01
-3.75888377e-01 -9.92823958e-01 1.73899516e-01 -3.63032907e-01
-8.10717463e-01 1.08555579e+00 4.56218839e-01 -2.35322937e-01
8.04747999e-01 2.84810245e-01 8.10431421e-01 5.46426654e-01
5.82861662e-01 -8.26703668e-01 1.62227347e-01 7.88558364e-01
8.26770246e-01 -1.36672068e+00 3.89490575e-01 -5.82475245e-01
-3.62848967e-01 1.17313838e+00 8.20503473e-01 -3.57395232e-01
6.47501647e-01 2.65473444e-02 -1.08046448e-02 -1.80856988e-01
-5.11707246e-01 -2.34016195e-01 4.44141388e-01 4.47025180e-01
-1.13296129e-01 6.22463450e-02 1.26990795e-01 1.61623657e-01
-1.47501335e-01 -3.42292100e-01 5.21935105e-01 8.44695866e-01
-7.90857971e-01 -6.73247993e-01 -3.67137551e-01 1.78184316e-01
-6.42238138e-03 7.01533929e-02 -5.36568642e-01 9.41166222e-01
4.33607586e-02 6.40828371e-01 1.34632923e-02 -5.27812898e-01
5.29380441e-01 -2.56057233e-01 6.22406363e-01 -8.73769104e-01
1.23004086e-01 3.30779195e-01 -2.09564194e-01 -7.17078745e-01
-8.23083334e-03 -6.91019654e-01 -1.25623429e+00 -2.75530428e-01
-6.45026505e-01 -3.65376472e-01 7.37516224e-01 6.90769017e-01
4.21565741e-01 -3.06281060e-01 8.20476890e-01 -1.06941795e+00
-7.26653039e-01 -6.98298216e-01 -4.98062909e-01 4.35127169e-02
5.25911748e-01 -8.46317649e-01 -3.57106477e-01 -4.44959462e-01] | [7.794622898101807, -2.475839138031006] |
bd4e9d48-5ea2-47f7-8fe2-f2431dca25c3 | geometry-and-convergence-of-natural-policy | 2211.02105 | null | https://arxiv.org/abs/2211.02105v1 | https://arxiv.org/pdf/2211.02105v1.pdf | Geometry and convergence of natural policy gradient methods | We study the convergence of several natural policy gradient (NPG) methods in infinite-horizon discounted Markov decision processes with regular policy parametrizations. For a variety of NPGs and reward functions we show that the trajectories in state-action space are solutions of gradient flows with respect to Hessian geometries, based on which we obtain global convergence guarantees and convergence rates. In particular, we show linear convergence for unregularized and regularized NPG flows with the metrics proposed by Kakade and Morimura and co-authors by observing that these arise from the Hessian geometries of conditional entropy and entropy respectively. Further, we obtain sublinear convergence rates for Hessian geometries arising from other convex functions like log-barriers. Finally, we interpret the discrete-time NPG methods with regularized rewards as inexact Newton methods if the NPG is defined with respect to the Hessian geometry of the regularizer. This yields local quadratic convergence rates of these methods for step size equal to the penalization strength. | ['Guido Montúfar', 'Johannes Müller'] | 2022-11-03 | null | null | null | null | ['policy-gradient-methods'] | ['methodology'] | [-2.84624249e-01 4.88474041e-01 -4.17786688e-01 -1.12361044e-01
-7.75612473e-01 -7.43060172e-01 5.69182575e-01 2.28799298e-01
-8.37556660e-01 1.18490970e+00 3.80740136e-01 -4.73745197e-01
-3.95928800e-01 -5.02439439e-01 -7.38935828e-01 -8.87436807e-01
-5.59929788e-01 3.93562406e-01 -9.45862308e-02 -4.09245402e-01
4.65993464e-01 3.21072042e-01 -9.05453622e-01 -5.23237705e-01
1.05252540e+00 1.00192702e+00 -1.93067357e-01 1.07030773e+00
2.61200786e-01 6.54091895e-01 -1.44884840e-01 -3.44286948e-01
7.56342292e-01 -3.59999239e-01 -8.60952139e-01 -4.60997857e-02
1.76984429e-01 -4.88882214e-01 -2.42041022e-01 1.55005217e+00
4.01267350e-01 6.05011165e-01 7.42447972e-01 -1.05233526e+00
-3.82297128e-01 5.16581297e-01 -4.42503750e-01 3.05265307e-01
-7.48422444e-02 2.65311360e-01 1.10469162e+00 -4.17031854e-01
5.18466055e-01 1.47308660e+00 5.15102863e-01 9.55255568e-01
-1.24839747e+00 1.28950775e-01 4.71096426e-01 -2.06702352e-01
-6.22453928e-01 -6.14531673e-02 3.76393320e-03 -5.53585768e-01
8.22204530e-01 2.02237248e-01 5.31099498e-01 7.52767622e-01
3.59213412e-01 7.56993115e-01 1.16664207e+00 -2.49373928e-01
6.78844154e-01 -4.27737366e-03 1.35979161e-01 9.30520177e-01
3.49691212e-01 4.61452574e-01 1.87994346e-01 -1.81403294e-01
9.18262482e-01 -3.66608500e-02 -3.34166884e-01 -4.66674626e-01
-1.06016743e+00 1.11891615e+00 6.65982738e-02 -1.60555065e-01
-6.20507061e-01 2.84192920e-01 3.54007423e-01 3.20227981e-01
5.18375754e-01 5.02896667e-01 -2.17994615e-01 -4.73536670e-01
-4.70622480e-01 6.05234921e-01 1.14773202e+00 9.40556645e-01
3.81771535e-01 2.46081650e-01 -5.94131768e-01 9.84854251e-02
2.66650051e-01 8.37499917e-01 1.86550468e-01 -1.45780551e+00
9.18802977e-01 7.82808512e-02 7.39085674e-01 -5.12815237e-01
-4.32131797e-01 -2.00426519e-01 -5.83291769e-01 5.08604825e-01
9.77655888e-01 -7.16622829e-01 -6.45565033e-01 1.78316259e+00
3.32432568e-01 -3.32921147e-01 3.97795558e-01 9.96585310e-01
-4.69741136e-01 7.44463146e-01 1.83964707e-02 -6.44087076e-01
8.10488641e-01 -6.90818727e-01 -7.07445085e-01 1.49801105e-01
7.31607676e-01 -2.02474043e-01 1.15735388e+00 2.99889565e-01
-1.38023865e+00 5.49202524e-02 -9.43015456e-01 1.99166715e-01
6.64944127e-02 -2.56100483e-02 1.55313954e-01 5.40504992e-01
-1.03846753e+00 1.35020971e+00 -1.05395973e+00 -1.60870235e-02
8.57523978e-02 2.50686556e-01 3.76338035e-01 5.91763854e-01
-1.05206180e+00 1.09096742e+00 4.14075315e-01 2.80437887e-01
-1.36997211e+00 -5.39502323e-01 -5.12592614e-01 7.09118135e-03
5.60622334e-01 -2.41306871e-01 1.57621121e+00 -8.33050251e-01
-1.99365628e+00 2.55295515e-01 3.47150117e-01 -8.79988313e-01
1.21415615e+00 -4.69273925e-01 1.20287277e-01 2.35470042e-01
4.48419005e-02 9.09548402e-02 8.97423983e-01 -6.56657994e-01
-6.76977873e-01 -3.57149124e-01 5.07081091e-01 6.67481065e-01
-1.80838197e-01 -2.48650149e-01 5.04620075e-01 -1.72640607e-01
-3.43025237e-01 -9.24273252e-01 -7.25380719e-01 -4.23528105e-02
-2.99980968e-01 -1.01571769e-01 4.95854288e-01 -7.79830933e-01
9.54538822e-01 -1.81905544e+00 4.80199009e-01 2.10692629e-01
8.21372960e-03 7.33824596e-02 2.14225695e-01 1.88704789e-01
2.01807976e-01 1.86973378e-01 -4.40715522e-01 -4.06256281e-02
2.86881894e-01 2.91884303e-01 -5.18612683e-01 8.95542443e-01
3.23098898e-03 6.14011705e-01 -1.20200062e+00 -2.07008839e-01
1.88635178e-02 -1.73922464e-01 -4.64066535e-01 6.25953600e-02
-5.02472758e-01 2.90133834e-01 -7.80663133e-01 -2.07636543e-02
4.76889431e-01 9.47761908e-02 5.09644710e-02 5.34884691e-01
-5.48445940e-01 -1.29877664e-02 -1.11268568e+00 1.02769399e+00
-3.41302693e-01 2.35235751e-01 5.94219029e-01 -1.12443113e+00
4.68878508e-01 2.79610425e-01 4.12246257e-01 -4.37617488e-02
1.19754173e-01 4.31326598e-01 -3.65865618e-01 -2.57054776e-01
6.08125269e-01 -5.87211370e-01 1.35275811e-01 2.21815065e-01
-1.74006850e-01 7.94843584e-03 2.64993817e-01 1.51074558e-01
8.29149008e-01 1.16303660e-01 9.67509001e-02 -8.28354478e-01
5.26460707e-01 -6.87209964e-02 5.13015509e-01 9.06192005e-01
-5.38233578e-01 1.78229064e-01 1.09332025e+00 -1.27334848e-01
-1.26171541e+00 -1.02052236e+00 -1.49938464e-01 9.51620162e-01
-5.03198430e-02 -1.25121072e-01 -8.94767404e-01 -6.08218610e-01
1.65713966e-01 6.78911150e-01 -7.06687629e-01 -4.54039685e-02
-5.19087911e-01 -7.68617630e-01 2.72240788e-01 3.98102582e-01
7.26955235e-01 -7.67289340e-01 -7.54910409e-01 3.79443854e-01
1.12991430e-01 -8.10302973e-01 -7.42577493e-01 8.28769654e-02
-1.17992187e+00 -9.97642040e-01 -1.12710452e+00 -1.02266006e-01
5.71259737e-01 -3.58022094e-01 5.76425254e-01 -8.05824995e-01
2.61431783e-01 7.24350154e-01 1.50606677e-01 -2.80904949e-01
-5.47043741e-01 -1.47877663e-01 3.19083124e-01 8.79627988e-02
-4.44608599e-01 -3.46854031e-02 -6.55820131e-01 3.05542678e-01
-6.73215687e-01 -3.14282656e-01 4.45106179e-02 6.09066844e-01
6.45450830e-01 -2.36919969e-01 2.75135517e-01 -5.12943804e-01
1.08189487e+00 -4.62290853e-01 -1.50410843e+00 3.03222269e-01
-8.48746479e-01 7.44621098e-01 8.26152027e-01 -3.53113472e-01
-1.24489963e+00 -1.03138313e-01 3.12868387e-01 -3.90074193e-01
4.80731577e-01 2.96596646e-01 3.12345564e-01 1.54560376e-02
6.29680932e-01 5.09834150e-03 1.30919874e-01 -3.26290280e-01
5.06135941e-01 2.99240500e-01 2.58861572e-01 -9.96056795e-01
2.47738689e-01 7.04566061e-01 3.60779583e-01 -5.99161923e-01
-7.90662408e-01 -1.05287820e-01 -1.64828300e-01 -3.00722897e-01
1.10760093e+00 -4.19244260e-01 -1.12152708e+00 3.62761915e-01
-9.22266483e-01 -6.01302862e-01 -9.73907113e-01 8.66652668e-01
-1.23813009e+00 5.47200263e-01 -8.21254134e-01 -1.49846470e+00
-2.19430357e-01 -9.44768012e-01 5.85713804e-01 2.81987280e-01
4.72945303e-01 -1.37105453e+00 3.52572471e-01 -3.49570394e-01
3.95220757e-01 3.83707583e-01 5.98280013e-01 -3.57720077e-01
-2.66163260e-01 -6.61700815e-02 1.15808070e-01 6.23494327e-01
-3.56942981e-01 -4.13698778e-02 -4.37986523e-01 -4.58795696e-01
5.24107873e-01 -2.47010328e-02 7.41411746e-01 8.69860411e-01
6.02514088e-01 -9.31314826e-01 -1.79679543e-02 4.70626533e-01
1.50167787e+00 2.86132962e-01 1.03266113e-01 3.22234511e-01
5.44190347e-01 6.43146217e-01 6.09041810e-01 7.47590423e-01
-3.99712920e-02 1.80928484e-01 4.45659190e-01 4.91062433e-01
7.42604315e-01 -2.42055357e-01 8.57921004e-01 4.56400365e-01
-4.78648305e-01 2.52455652e-01 -6.29038095e-01 5.31496048e-01
-2.33518076e+00 -1.00623119e+00 -1.49026662e-01 2.69894981e+00
9.41531003e-01 1.65984452e-01 4.85807121e-01 -4.66115952e-01
9.28865194e-01 6.94880411e-02 -1.00719762e+00 -8.39921236e-01
-4.30716313e-02 -7.61771053e-02 1.49267972e+00 1.15013862e+00
-9.78199840e-01 6.15811169e-01 6.83566713e+00 6.31706059e-01
-7.56688535e-01 1.77278295e-01 6.66714609e-01 -2.89723217e-01
-1.16848834e-01 1.10561617e-01 -9.73497152e-01 3.84277433e-01
1.11520517e+00 -4.59785998e-01 6.36843443e-01 1.10705090e+00
5.27773082e-01 -1.17357150e-01 -7.78402090e-01 4.24788743e-01
-7.14668512e-01 -1.02759457e+00 -5.87038934e-01 4.07334059e-01
1.06688893e+00 -3.00998800e-03 1.57657579e-01 2.08388135e-01
8.52047563e-01 -6.61429644e-01 7.41757452e-01 3.91536564e-01
4.26472545e-01 -9.27299798e-01 4.88016963e-01 2.39635006e-01
-9.11156476e-01 -4.39135194e-01 -6.19463444e-01 -2.46909767e-01
4.17994380e-01 4.04300034e-01 -4.36831743e-01 2.87659466e-01
7.94900805e-02 6.12786114e-01 2.97675371e-01 8.25492024e-01
-1.34628832e-01 5.56890130e-01 -6.04284525e-01 -2.11978510e-01
8.75698388e-01 -8.71628106e-01 9.16590810e-01 8.16115260e-01
9.05537382e-02 3.82924303e-02 1.90030918e-01 8.11316967e-01
3.08599174e-01 6.37155306e-03 -5.23611605e-01 -2.58566558e-01
-1.72652200e-01 1.00760794e+00 -6.26996756e-01 -2.81704068e-01
-4.25176099e-02 7.54955292e-01 2.49394894e-01 6.41932547e-01
-6.83148324e-01 -3.67239684e-01 9.16556835e-01 -1.20171309e-01
2.21701965e-01 -5.08144557e-01 4.46581729e-02 -1.25081491e+00
2.45828062e-01 -3.23518991e-01 4.52311903e-01 -2.76472326e-03
-1.12183428e+00 3.34760845e-01 3.68688822e-01 -7.89607406e-01
-5.20534515e-01 -7.01181948e-01 -4.84231770e-01 8.65524471e-01
-1.12782741e+00 -1.27230242e-01 6.63224816e-01 7.87634611e-01
1.68770313e-01 1.64596632e-01 3.73320401e-01 -1.95179746e-01
-5.42621434e-01 2.99514718e-02 9.47709978e-01 -1.80321813e-01
-1.51047871e-01 -1.73070490e+00 3.08247596e-01 9.29319084e-01
-5.44385850e-01 1.82972178e-01 1.03168774e+00 -6.59923911e-01
-1.39818525e+00 -9.32583988e-01 2.74828821e-01 -1.49943367e-01
1.09411252e+00 -6.65733367e-02 -5.44501781e-01 7.98203886e-01
7.16143847e-02 -1.08156621e-01 -3.11101675e-01 -1.25476137e-01
3.38629037e-01 1.21008702e-01 -1.20075178e+00 5.64878166e-01
8.39858472e-01 -2.92819202e-01 -1.60216331e-01 7.29712248e-01
5.74670851e-01 -4.76522833e-01 -7.30476737e-01 -4.72474396e-02
1.91658854e-01 -5.90662301e-01 4.88441229e-01 -1.14901304e+00
1.81472778e-01 3.12684067e-02 -8.55217874e-03 -1.31818211e+00
1.06758691e-01 -1.38824427e+00 -1.42132685e-01 5.78321457e-01
3.42261970e-01 -9.86030042e-01 5.28361320e-01 8.24427962e-01
-9.39927325e-02 -9.01251197e-01 -1.24514735e+00 -1.23548651e+00
5.27776957e-01 -7.71735162e-02 -4.69144955e-02 4.60876554e-01
4.84267890e-01 4.26102988e-03 -4.31939632e-01 -1.50988281e-01
9.40044403e-01 -2.14859977e-01 2.66114902e-03 -7.19412744e-01
-4.21810985e-01 -6.01713657e-01 8.50550234e-02 -1.10257542e+00
1.39959067e-01 -8.24642539e-01 2.26184294e-01 -1.33411169e+00
-1.43289685e-01 -3.69742244e-01 -8.14029798e-02 5.97156435e-02
3.87193970e-02 -8.19169939e-01 4.27478522e-01 1.97929636e-01
-5.15775204e-01 8.81407976e-01 1.52235997e+00 1.83615834e-01
-5.85897267e-01 4.67653364e-01 -2.81881362e-01 7.39794970e-01
1.00420547e+00 -3.62209707e-01 -5.18634856e-01 -1.54186925e-02
4.23139066e-01 4.37233269e-01 2.23742440e-01 -7.27546990e-01
-1.51051402e-01 -6.11582696e-01 -3.60946774e-01 -1.47019178e-01
1.48816928e-01 -3.03881019e-01 -2.38878742e-01 6.97071552e-01
-9.08616781e-01 2.78225601e-01 -2.04126075e-01 1.00927949e+00
2.43119463e-01 -6.84316874e-01 9.36421037e-01 -3.23670655e-01
-7.89657384e-02 3.76090080e-01 -6.56727731e-01 7.16391623e-01
9.43680108e-01 1.08276717e-01 -1.70745894e-01 -7.75577843e-01
-9.48780656e-01 5.65469801e-01 8.09738860e-02 -2.47297347e-01
4.31988150e-01 -1.16976643e+00 -5.86943209e-01 -3.49420905e-01
-7.44391680e-01 -1.46848217e-01 -9.19387713e-02 1.10451925e+00
-4.77003008e-01 4.50649828e-01 -1.79711267e-01 -1.89728558e-01
-5.96206427e-01 6.00316107e-01 7.79132545e-01 -5.97595215e-01
-4.69027340e-01 5.59676468e-01 6.04750402e-02 -2.04416975e-01
2.81670004e-01 -8.60046685e-01 1.45580307e-01 -9.36169364e-03
3.51099610e-01 9.37690973e-01 -3.32679749e-01 -1.83876306e-01
-4.58736792e-02 3.67244691e-01 -4.74403873e-02 -8.90106559e-01
1.02841783e+00 -1.19847409e-01 2.82093495e-01 3.82746130e-01
1.25439024e+00 -3.92995745e-01 -2.11507559e+00 6.00841530e-02
1.41078204e-01 -2.14096025e-01 -8.93211812e-02 -3.60964596e-01
-9.28651571e-01 8.36260557e-01 5.75177789e-01 4.56454664e-01
5.55742443e-01 -3.96348268e-01 3.23907554e-01 5.10900438e-01
2.94925898e-01 -1.60868895e+00 -2.12433070e-01 8.57467711e-01
6.90991998e-01 -9.28842306e-01 -1.35983184e-01 2.04213262e-01
-9.31758285e-01 1.17589533e+00 1.37134954e-01 -4.76400614e-01
6.58373177e-01 -1.07134320e-01 -4.08765107e-01 1.21577226e-01
-5.92155099e-01 -2.69261688e-01 -1.05247349e-01 3.62522990e-01
-5.62173165e-02 4.63461310e-01 -9.60660875e-01 5.49306497e-02
-3.67581472e-02 -1.52685583e-01 7.33130336e-01 9.65596616e-01
-6.43840730e-01 -7.37734199e-01 -1.59009203e-01 3.09894741e-01
-6.02121830e-01 -1.13083400e-01 1.02942344e-03 5.54727793e-01
-7.92787790e-01 7.26239920e-01 -1.55907646e-01 2.22414896e-01
3.00877929e-01 -3.57946530e-02 5.95540464e-01 -2.12237775e-01
-5.47469735e-01 -7.06973150e-02 3.58771421e-02 -6.23455286e-01
-1.69888213e-01 -7.93386757e-01 -1.44509172e+00 -3.18404645e-01
-2.26475447e-01 4.64783758e-01 7.06924915e-01 1.01139545e+00
1.49365723e-01 -7.91252032e-02 7.13984370e-01 -6.69819236e-01
-1.81823730e+00 -8.61519098e-01 -9.05745327e-01 1.02524720e-01
6.21373892e-01 -4.84027267e-01 -7.27155626e-01 -3.88882488e-01] | [4.269306182861328, 2.625560760498047] |
dec27325-789d-450f-8236-4d019f2801b4 | ccnet-criss-cross-attention-for-semantic | 1811.11721 | null | https://arxiv.org/abs/1811.11721v2 | https://arxiv.org/pdf/1811.11721v2.pdf | CCNet: Criss-Cross Attention for Semantic Segmentation | Contextual information is vital in visual understanding problems, such as semantic segmentation and object detection. We propose a Criss-Cross Network (CCNet) for obtaining full-image contextual information in a very effective and efficient way. Concretely, for each pixel, a novel criss-cross attention module harvests the contextual information of all the pixels on its criss-cross path. By taking a further recurrent operation, each pixel can finally capture the full-image dependencies. Besides, a category consistent loss is proposed to enforce the criss-cross attention module to produce more discriminative features. Overall, CCNet is with the following merits: 1) GPU memory friendly. Compared with the non-local block, the proposed recurrent criss-cross attention module requires 11x less GPU memory usage. 2) High computational efficiency. The recurrent criss-cross attention significantly reduces FLOPs by about 85% of the non-local block. 3) The state-of-the-art performance. We conduct extensive experiments on semantic segmentation benchmarks including Cityscapes, ADE20K, human parsing benchmark LIP, instance segmentation benchmark COCO, video segmentation benchmark CamVid. In particular, our CCNet achieves the mIoU scores of 81.9%, 45.76% and 55.47% on the Cityscapes test set, the ADE20K validation set and the LIP validation set respectively, which are the new state-of-the-art results. The source codes are available at \url{https://github.com/speedinghzl/CCNet}. | ['Wenyu Liu', 'Humphrey Shi', 'Yunchao Wei', 'Xinggang Wang', 'Thomas S. Huang', 'Lichao Huang', 'Zilong Huang'] | 2018-11-28 | ccnet-criss-cross-attention-for-semantic-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/Huang_CCNet_Criss-Cross_Attention_for_Semantic_Segmentation_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Huang_CCNet_Criss-Cross_Attention_for_Semantic_Segmentation_ICCV_2019_paper.pdf | iccv-2019-10 | ['thermal-image-segmentation', 'human-parsing'] | ['computer-vision', 'computer-vision'] | [ 1.68978378e-01 -6.03388511e-02 -3.02089393e-01 -3.85801166e-01
-9.14540648e-01 -2.57918715e-01 2.59053290e-01 1.66885510e-01
-5.41582048e-01 5.01180649e-01 -5.55516742e-02 -2.09860161e-01
4.65533212e-02 -6.59051538e-01 -9.34688985e-01 -8.25063288e-01
4.20401424e-01 6.51638396e-03 6.53931856e-01 9.02806595e-03
2.48297736e-01 1.92422822e-01 -1.47639310e+00 3.27198625e-01
1.06389439e+00 1.31645823e+00 4.46388513e-01 5.23719847e-01
-1.37779519e-01 4.36943561e-01 -2.88067520e-01 -3.95047367e-01
3.28851835e-04 -1.93777859e-01 -7.40922928e-01 1.08743802e-01
4.66417849e-01 -1.84461832e-01 -1.20595478e-01 1.44932330e+00
4.18233633e-01 -2.57295603e-03 3.89235079e-01 -1.24052787e+00
-5.05055070e-01 4.47219610e-01 -1.15877008e+00 1.89374834e-01
3.34229991e-02 2.04998150e-01 9.99936223e-01 -7.82216609e-01
5.24167180e-01 1.16332674e+00 4.27443296e-01 3.59931350e-01
-8.17538679e-01 -8.22604239e-01 4.08341438e-01 3.36481184e-01
-1.24892104e+00 -1.72627375e-01 6.80125833e-01 -2.23281011e-01
5.32049596e-01 2.73499399e-01 6.74006045e-01 8.68282914e-01
-6.58351108e-02 1.22878003e+00 1.01048839e+00 -2.32905596e-01
4.52647619e-02 -1.10170707e-01 4.81027514e-01 8.45033407e-01
1.83470473e-01 -1.95810333e-01 -2.96305686e-01 2.09207907e-01
6.67760789e-01 3.47397253e-02 -2.23750487e-01 -8.62393621e-03
-8.64474714e-01 6.53252721e-01 7.13492095e-01 2.61184365e-01
-2.63785332e-01 2.62381464e-01 5.94582796e-01 -2.16439486e-01
3.81948262e-01 -3.26044351e-01 -6.55311823e-01 2.49017193e-03
-6.89396203e-01 -4.63985354e-02 2.68195987e-01 1.06863201e+00
8.39740098e-01 -1.32335797e-01 -8.17444623e-02 1.01730883e+00
3.45595479e-01 4.53734189e-01 3.04045260e-01 -9.33796763e-01
6.76954925e-01 6.47099733e-01 -2.53837496e-01 -8.73273373e-01
-2.98277080e-01 -4.62957442e-01 -9.01505053e-01 -3.77302757e-03
4.07797664e-01 -1.06112309e-01 -1.17558694e+00 1.66080642e+00
5.43829381e-01 3.29107702e-01 -1.54441653e-03 9.27419603e-01
1.00125957e+00 8.53710473e-01 4.38967943e-01 1.37854768e-02
1.67670453e+00 -1.18227184e+00 -4.59920704e-01 -2.53162414e-01
4.52107638e-01 -9.47081029e-01 1.13747263e+00 1.70559406e-01
-9.43346500e-01 -7.59850442e-01 -8.65953445e-01 -2.65301019e-01
-2.23128200e-01 3.46367925e-01 6.29556477e-01 3.00377905e-01
-7.94229448e-01 3.53674322e-01 -9.08015132e-01 -3.19457680e-01
7.81454027e-01 3.36959511e-01 -2.97866583e-01 -2.17136279e-01
-8.76920164e-01 1.64336801e-01 4.95376110e-01 2.21945807e-01
-5.35278678e-01 -7.74005055e-01 -9.33381557e-01 4.22933362e-02
5.86270690e-01 -4.24192131e-01 1.13101315e+00 -1.15131342e+00
-1.18130028e+00 8.95929337e-01 -2.29334056e-01 -2.03276455e-01
5.39605021e-01 -2.15062931e-01 -2.53401011e-01 2.67971784e-01
3.90270710e-01 9.65298235e-01 3.80977690e-01 -1.07002723e+00
-9.54470098e-01 -4.81159866e-01 -1.39900878e-01 1.60810694e-01
-6.09398782e-02 -4.94295266e-03 -1.20582008e+00 -6.54543102e-01
2.77882695e-01 -1.00856948e+00 -5.37517592e-02 -1.04546808e-02
-7.74539530e-01 -4.04166818e-01 1.00022519e+00 -7.08051920e-01
1.07821190e+00 -2.40321779e+00 -2.58265853e-01 -4.40671146e-02
-4.84965034e-02 7.24867702e-01 -8.01285356e-02 -1.02547891e-01
4.22778958e-03 2.19249755e-01 -3.36599499e-01 -3.15374136e-01
-1.61230117e-01 5.43118715e-02 1.71991363e-01 2.73176223e-01
3.41799378e-01 8.97178650e-01 -7.76530504e-01 -7.49853194e-01
4.91361499e-01 5.57032645e-01 -4.80887920e-01 -5.68715967e-02
-1.61198735e-01 3.35490406e-01 -6.02495909e-01 8.63216996e-01
8.93755138e-01 -1.94318056e-01 9.41382628e-03 -3.90261173e-01
-1.76576629e-01 -3.27913128e-02 -1.15194225e+00 1.71857953e+00
-2.49120101e-01 6.00099564e-01 5.91839701e-02 -9.21717227e-01
7.95249045e-01 1.43229188e-02 2.03903854e-01 -1.13979328e+00
1.76307753e-01 2.30517596e-01 -2.28187978e-01 -5.17530203e-01
2.61795312e-01 2.59943694e-01 -2.15171859e-01 -6.32167459e-02
1.92954391e-02 4.56033885e-01 3.03663343e-01 1.55132025e-01
6.00374460e-01 5.86716421e-02 1.60717636e-01 -4.11870211e-01
7.75504827e-01 -1.16899930e-01 9.49248791e-01 4.79633689e-01
-3.88082176e-01 6.58510506e-01 6.90957904e-01 -3.09017390e-01
-7.32464135e-01 -9.45489764e-01 -1.41772583e-01 9.81170356e-01
4.49394614e-01 -2.80801624e-01 -1.07537901e+00 -7.12216675e-01
-1.74414530e-01 4.71897542e-01 -6.62582934e-01 9.94904488e-02
-6.63141668e-01 -6.31955445e-01 3.62316251e-01 8.53237510e-01
1.08298349e+00 -1.19542861e+00 -5.07416725e-01 8.79542455e-02
-2.71437198e-01 -1.15881741e+00 -8.63786161e-01 -1.75026342e-01
-7.96523452e-01 -1.14812458e+00 -7.70468116e-01 -1.08165753e+00
6.58332348e-01 3.87227416e-01 8.25732827e-01 1.75134599e-01
-4.37556326e-01 -4.14144397e-02 -2.12338939e-01 -1.73060849e-01
9.17158648e-02 4.04929370e-03 -5.81857145e-01 1.67148575e-01
3.49228561e-01 -3.20523143e-01 -9.36936021e-01 4.46138591e-01
-7.42011368e-01 3.69603485e-01 5.36088169e-01 8.57552946e-01
1.13969195e+00 -1.79880843e-01 4.25814956e-01 -8.09336662e-01
-6.25708029e-02 -3.75676543e-01 -7.72991240e-01 1.17636472e-01
-2.93419719e-01 -2.03206763e-01 4.70962435e-01 -2.42110595e-01
-1.17608166e+00 3.43600094e-01 -3.70027184e-01 -3.29052091e-01
-2.73908794e-01 6.29433021e-02 -5.95769405e-01 2.53699541e-01
-6.95667267e-02 2.41184130e-01 -3.11786443e-01 -6.64326072e-01
2.42856517e-01 6.47038639e-01 6.41568184e-01 -4.73663270e-01
1.80306807e-01 4.42407966e-01 -1.43715426e-01 -8.61269832e-01
-7.72207618e-01 -5.44501960e-01 -4.18462932e-01 -7.63276219e-02
1.28821278e+00 -9.46308434e-01 -8.56640935e-01 7.87698865e-01
-1.03798115e+00 -3.60947818e-01 5.44206463e-02 3.41310143e-01
-2.10005000e-01 3.18961084e-01 -6.25422299e-01 -6.80738628e-01
-5.63829362e-01 -1.53802657e+00 1.24173486e+00 8.62163544e-01
1.27648011e-01 -7.09320605e-01 -4.60102737e-01 5.84396124e-01
-1.13537997e-01 3.28039557e-01 8.15070868e-01 -4.01757658e-01
-7.50054538e-01 6.82813898e-02 -8.27602088e-01 4.65205729e-01
-1.32043839e-01 1.48871556e-01 -9.83315647e-01 -4.33638282e-02
-5.44986129e-01 -1.32357493e-01 1.07729065e+00 6.92766309e-01
1.47564054e+00 -3.93701121e-02 -3.16415489e-01 7.98457980e-01
1.64433515e+00 4.90552843e-01 6.54332399e-01 2.38547742e-01
8.79250288e-01 5.10736644e-01 8.97582054e-01 1.08502500e-01
5.28787494e-01 6.62012398e-01 4.11527246e-01 -5.17176032e-01
-3.86578143e-01 -4.10672128e-01 1.39554486e-01 6.65109038e-01
1.33624718e-01 -2.21258327e-01 -8.43252122e-01 6.93019807e-01
-1.90146077e+00 -6.24792755e-01 -4.53193456e-01 2.00975251e+00
5.10465860e-01 2.57267356e-01 8.69222656e-02 1.43464103e-01
1.06513834e+00 2.37974331e-01 -6.25567257e-01 -3.88541281e-01
-4.38476168e-02 2.59841591e-01 5.46136737e-01 5.24766207e-01
-1.30447662e+00 1.25295258e+00 3.85903168e+00 1.13717282e+00
-1.02629459e+00 2.29918301e-01 1.10375977e+00 1.45538360e-01
1.67363733e-01 -1.59975275e-01 -9.73810375e-01 7.61519730e-01
4.79068190e-01 2.71130323e-01 -4.27199081e-02 9.22414064e-01
1.44367188e-01 -3.47598404e-01 -6.28750205e-01 9.96360779e-01
-1.06892936e-01 -1.24863529e+00 -8.83866921e-02 -3.23206969e-02
6.38712227e-01 1.94891840e-01 9.57112610e-02 8.08718055e-02
1.78175420e-02 -6.92400753e-01 7.30739415e-01 9.19983685e-02
8.51355910e-01 -1.02331865e+00 7.31367111e-01 1.81299746e-02
-1.53745794e+00 6.92062229e-02 -2.38429710e-01 4.70365554e-01
8.14766437e-02 5.68008006e-01 -2.10550591e-01 5.21192789e-01
1.11257958e+00 6.52831674e-01 -5.08628607e-01 9.79959071e-01
-3.35052431e-01 6.62923872e-01 -3.57552737e-01 -1.10588660e-02
6.78176224e-01 -2.71865696e-01 3.02418470e-01 1.42650771e+00
1.10043921e-01 2.28614360e-01 1.17005430e-01 5.90560973e-01
-1.91799819e-01 2.86071867e-01 -3.61843146e-02 4.26932454e-01
3.08935344e-01 1.18424690e+00 -1.02452409e+00 -4.92624968e-01
-3.65495890e-01 9.15826023e-01 2.11100742e-01 3.52613270e-01
-1.03703594e+00 -6.42827332e-01 7.19098628e-01 -1.10788494e-01
6.67733371e-01 1.63472816e-01 -3.37794691e-01 -9.67086911e-01
2.33202919e-01 -7.17937171e-01 4.51763213e-01 -7.74988055e-01
-9.07912195e-01 6.08935475e-01 -2.00905055e-01 -9.34065521e-01
2.74054557e-01 -5.72848201e-01 -7.11458743e-01 7.72593915e-01
-1.71560204e+00 -1.25480056e+00 -6.11012936e-01 6.08092070e-01
7.80018866e-01 2.17214659e-01 3.56268167e-01 4.49550450e-01
-1.04276574e+00 6.76717341e-01 8.06120690e-03 4.44607109e-01
4.27268267e-01 -1.12471604e+00 4.39781874e-01 8.50689054e-01
9.61650815e-03 3.04381698e-01 2.05287471e-01 -5.21015167e-01
-9.62327242e-01 -1.28335428e+00 6.54158950e-01 2.14721352e-01
4.19747323e-01 -2.53104746e-01 -9.05465662e-01 4.90804970e-01
8.19936693e-02 3.05377960e-01 3.12408775e-01 -1.78173155e-01
-4.45024699e-01 -4.46765453e-01 -1.01271331e+00 5.00261486e-01
1.05460548e+00 -3.02108943e-01 -2.95097113e-01 2.03225330e-01
7.89099693e-01 -4.66609061e-01 -5.83966911e-01 4.41885233e-01
6.94947720e-01 -1.10500205e+00 9.94531691e-01 -1.18991949e-01
4.99521941e-01 -4.47858602e-01 -2.24889860e-01 -7.10269094e-01
-2.55177654e-02 -4.76361930e-01 2.92633563e-01 1.58887279e+00
4.51531798e-01 -6.57195032e-01 7.75680304e-01 2.99741417e-01
-3.33037376e-01 -1.12373698e+00 -8.92527163e-01 -5.67214191e-01
8.08229372e-02 -5.47980964e-01 4.77480561e-01 5.77580571e-01
-6.01497471e-01 2.63350397e-01 -1.34717628e-01 2.11923033e-01
6.96183264e-01 6.70803934e-02 5.85284710e-01 -9.41381454e-01
-1.82404131e-01 -4.35980469e-01 -4.21194881e-01 -1.18759167e+00
6.35992885e-02 -6.73482299e-01 -3.41167599e-02 -1.58576477e+00
4.22596723e-01 -5.42596638e-01 -4.67581511e-01 6.34814382e-01
-4.54748631e-01 4.34273869e-01 4.59337264e-01 -7.35090747e-02
-6.34087920e-01 2.90223569e-01 1.35911608e+00 -8.51244777e-02
-1.98371068e-01 -8.65728706e-02 -7.85035193e-01 9.06727552e-01
1.01063561e+00 -4.64440554e-01 -3.84254098e-01 -4.92925823e-01
-3.68582666e-01 -6.24772348e-02 4.64228153e-01 -9.19695199e-01
2.37450913e-01 4.57484461e-02 3.63110065e-01 -8.90894473e-01
3.13186437e-01 -4.74222988e-01 -1.17822193e-01 6.03537142e-01
-5.81386276e-02 -1.17330424e-01 4.48438257e-01 5.72032750e-01
-3.79251510e-01 -9.08482820e-02 1.14605951e+00 -1.79674476e-03
-9.64981616e-01 4.43698436e-01 1.24531910e-01 3.30700248e-01
1.11685514e+00 -2.36782551e-01 -3.50478679e-01 5.68166897e-02
-5.63652158e-01 3.29979241e-01 2.63418704e-01 3.62170041e-01
5.00000775e-01 -1.11344099e+00 -5.65569997e-01 2.44738951e-01
7.68329948e-02 2.93398827e-01 7.02258646e-01 8.36573601e-01
-6.00529134e-01 3.21869284e-01 -9.71768647e-02 -9.61993039e-01
-1.54386854e+00 3.30907792e-01 1.30618602e-01 -8.88739824e-02
-7.59250939e-01 1.07331944e+00 6.75499618e-01 -1.03659995e-01
1.61236644e-01 -4.63990629e-01 -1.12101555e-01 7.69459978e-02
4.90025073e-01 3.49934876e-01 -6.89443201e-02 -8.29568267e-01
-6.53946877e-01 1.11215758e+00 -1.25509202e-01 7.59419873e-02
1.11413753e+00 -4.51117568e-02 4.71357554e-02 -9.19881370e-03
1.52731752e+00 -1.70821846e-01 -1.47968519e+00 -1.20889761e-01
-1.38533935e-01 -4.18394387e-01 1.60987228e-01 -8.24975371e-01
-1.58981502e+00 1.16011953e+00 7.79308558e-01 -2.22978354e-01
1.44223881e+00 1.12839438e-01 1.02320910e+00 -2.66269863e-01
1.17745444e-01 -1.05959201e+00 -8.01419988e-02 2.95124143e-01
6.94360197e-01 -1.31722713e+00 -1.86223596e-01 -7.57004440e-01
-8.35290134e-01 8.35017741e-01 7.40716934e-01 -1.71348438e-01
6.43387258e-01 9.70871150e-02 1.32283522e-02 -2.55521033e-02
-4.29428101e-01 -4.52948362e-01 2.07220405e-01 3.50965589e-01
2.57905334e-01 2.40928203e-01 -3.74000847e-01 5.57992339e-01
1.22897670e-01 -1.42680809e-01 5.32621779e-02 6.92714632e-01
-2.28653044e-01 -8.52006912e-01 -8.14593881e-02 2.15187892e-01
-5.40059984e-01 -1.06685944e-01 -6.65057078e-02 1.13804352e+00
4.35720980e-01 1.02526045e+00 1.63822144e-01 -2.32613519e-01
3.97572160e-01 -1.14010200e-01 1.11948863e-01 -2.85143644e-01
-4.76210475e-01 4.64391798e-01 1.73842963e-02 -7.82585382e-01
-2.55046427e-01 -7.49345303e-01 -1.56633699e+00 -1.67245552e-01
-2.61164218e-01 -2.41342224e-02 7.69269943e-01 6.67196393e-01
4.94175404e-01 6.55853629e-01 4.82171714e-01 -4.95914727e-01
3.35838422e-02 -7.53059030e-01 -2.71816283e-01 4.79556918e-01
1.62170276e-01 -5.57870090e-01 -1.67964265e-01 -6.34735897e-02] | [9.440918922424316, -0.19131898880004883] |
ceeecc1d-194a-406c-8170-2b1c30d0e99c | weakly-aligned-feature-fusion-for-multimodal | 2204.09848 | null | https://arxiv.org/abs/2204.09848v1 | https://arxiv.org/pdf/2204.09848v1.pdf | Weakly Aligned Feature Fusion for Multimodal Object Detection | To achieve accurate and robust object detection in the real-world scenario, various forms of images are incorporated, such as color, thermal, and depth. However, multimodal data often suffer from the position shift problem, i.e., the image pair is not strictly aligned, making one object has different positions in different modalities. For the deep learning method, this problem makes it difficult to fuse multimodal features and puzzles the convolutional neural network (CNN) training. In this article, we propose a general multimodal detector named aligned region CNN (AR-CNN) to tackle the position shift problem. First, a region feature (RF) alignment module with adjacent similarity constraint is designed to consistently predict the position shift between two modalities and adaptively align the cross-modal RFs. Second, we propose a novel region of interest (RoI) jitter strategy to improve the robustness to unexpected shift patterns. Third, we present a new multimodal feature fusion method that selects the more reliable feature and suppresses the less useful one via feature reweighting. In addition, by locating bounding boxes in both modalities and building their relationships, we provide novel multimodal labeling named KAIST-Paired. Extensive experiments on 2-D and 3-D object detection, RGB-T, and RGB-D datasets demonstrate the effectiveness and robustness of our method. | ['Hong Qiao', 'Zhen Lei', 'Xu Yang', 'Zhan Song', 'Xiangyu Zhu', 'Zhiyong Liu', 'Lu Zhang'] | 2022-04-21 | null | null | null | null | ['robust-object-detection'] | ['computer-vision'] | [ 2.03781322e-01 -4.62861210e-01 -5.07704094e-02 -4.96856809e-01
-6.48412108e-01 -5.29467106e-01 2.69718409e-01 5.35880588e-02
-3.75596732e-01 2.95260131e-01 4.08777855e-02 1.44772321e-01
-2.02543885e-01 -4.23180044e-01 -6.11693740e-01 -1.04271626e+00
4.38106924e-01 -4.40482125e-02 4.83414829e-01 -1.95640363e-02
3.54061812e-01 7.28394866e-01 -1.40290082e+00 3.81292552e-01
6.31012201e-01 1.41243815e+00 3.85898352e-01 5.03597260e-02
-1.52162030e-01 3.50717127e-01 -3.90327901e-01 -1.17515132e-01
4.47895169e-01 -2.61401206e-01 -4.44163054e-01 3.95743817e-01
3.87456328e-01 -4.65429604e-01 -3.83330733e-01 1.00052631e+00
7.18591511e-01 2.58171052e-01 4.80051666e-01 -1.49060225e+00
-3.88759732e-01 3.96428794e-01 -1.17964852e+00 2.37903357e-01
2.54012764e-01 1.44676387e-01 5.73636949e-01 -9.84233320e-01
3.78855973e-01 1.30514121e+00 3.69081497e-01 4.26746905e-01
-9.93237257e-01 -7.26157725e-01 3.91060352e-01 1.53625384e-01
-1.57841229e+00 -2.25550801e-01 1.15114355e+00 -2.13769794e-01
4.40957129e-01 1.91856444e-01 4.87809867e-01 9.33769703e-01
-2.40496267e-02 9.01451409e-01 9.87959623e-01 -2.34881759e-01
-1.23347640e-01 1.18152380e-01 -1.88950256e-01 6.76758170e-01
9.76762772e-02 1.35888457e-01 -5.58420718e-01 -4.69813775e-03
7.02238858e-01 4.07869160e-01 -3.75359178e-01 -6.73108876e-01
-1.51262033e+00 4.13835764e-01 7.44190037e-01 3.56953412e-01
-9.60945040e-02 2.28902400e-02 2.41941214e-01 -1.23471104e-01
3.43628302e-02 -1.82045181e-03 -3.82611930e-01 3.34507495e-01
-4.87791330e-01 -5.18392548e-02 -1.30376607e-01 9.18572962e-01
8.43637705e-01 -3.67178470e-01 -2.49973163e-01 9.88085866e-01
5.90193033e-01 5.30402362e-01 5.07994354e-01 -6.51020646e-01
6.27038658e-01 9.55392003e-01 1.88981555e-02 -1.34810686e+00
-7.42067337e-01 -3.06375027e-01 -1.00892460e+00 5.02255745e-02
5.03348112e-01 1.66677922e-01 -8.81640077e-01 1.60284317e+00
5.97773194e-01 5.33796810e-02 -1.51363179e-01 1.46748233e+00
9.32484865e-01 4.60562736e-01 -1.52545244e-01 -2.04996362e-01
1.34316361e+00 -7.51727343e-01 -4.94880587e-01 -2.42187083e-01
1.74404159e-01 -9.20262992e-01 6.66261852e-01 1.91560403e-01
-8.98078561e-01 -7.36133039e-01 -9.89895403e-01 -4.67371847e-03
-3.49796772e-01 2.77556211e-01 4.29245859e-01 2.92765349e-01
-4.30129439e-01 1.47443756e-01 -6.34190202e-01 -1.56818509e-01
2.89191753e-01 4.61238682e-01 -5.79820096e-01 -2.61929423e-01
-1.02585590e+00 6.49037540e-01 5.97827733e-01 5.78483284e-01
-4.76371497e-01 -2.73467720e-01 -7.93877900e-01 -1.79433823e-01
4.98519778e-01 -3.65631849e-01 9.23068643e-01 -1.06617296e+00
-1.26271200e+00 6.98908448e-01 -1.05296277e-01 2.34302536e-01
3.62453043e-01 5.81577122e-02 -5.50019681e-01 6.08713850e-02
-1.04278855e-01 7.62212455e-01 1.00449574e+00 -1.48030674e+00
-8.51683021e-01 -5.55205584e-01 -2.25103512e-01 3.85938466e-01
-2.48891532e-01 9.62858871e-02 -6.83763266e-01 -4.74126667e-01
8.05405736e-01 -7.61029184e-01 -1.13672383e-01 1.21505797e-01
-5.79209387e-01 -2.20031768e-01 8.39604259e-01 -3.52222443e-01
1.07937145e+00 -2.48101377e+00 2.88309902e-01 5.55762827e-01
1.55530348e-01 4.43449840e-02 -1.78225726e-01 -6.21413030e-02
-1.38704300e-01 -1.02469452e-01 -1.47590294e-01 -1.92164585e-01
-1.93803564e-01 1.76997706e-01 -7.35359490e-02 6.80977464e-01
3.38539451e-01 6.97217226e-01 -6.90457463e-01 -7.89023280e-01
3.46759230e-01 3.71751547e-01 -1.24232970e-01 6.83650374e-02
1.84925824e-01 6.41829312e-01 -5.96219957e-01 1.07222199e+00
1.07919288e+00 -1.54404476e-01 -1.58438891e-01 -9.56299305e-01
-2.63387948e-01 -3.41941178e-01 -1.47976112e+00 1.74421978e+00
-1.45433784e-01 4.38923061e-01 -1.06609374e-01 -8.04851472e-01
1.03966165e+00 -4.85496819e-02 5.76686621e-01 -9.58301425e-01
4.41755176e-01 3.02818000e-01 -5.98887959e-03 -6.26965284e-01
5.33172369e-01 4.66416292e-02 -4.66714948e-02 3.20340276e-01
-2.04670191e-01 2.05064952e-01 -1.77011296e-01 -1.65761441e-01
5.47150493e-01 4.95171547e-02 1.14994727e-01 3.51442754e-01
6.36809051e-01 -2.65676051e-01 7.61593640e-01 4.73860264e-01
-4.09147531e-01 9.55077410e-01 3.52183640e-01 -4.56732571e-01
-7.91854322e-01 -9.35905755e-01 -2.21960440e-01 8.94959092e-01
9.46773767e-01 1.68496326e-01 -1.61334395e-01 -8.23000431e-01
2.04367712e-01 1.09090030e-01 -5.55401146e-01 -3.38120848e-01
-6.35730028e-01 -7.41270840e-01 2.74182141e-01 6.41101122e-01
6.78999305e-01 -6.67124569e-01 -5.81899583e-01 -6.21564686e-03
-3.26627254e-01 -9.81783688e-01 -6.99818671e-01 1.84592113e-01
-7.34824538e-01 -1.05842710e+00 -7.34076738e-01 -7.86224306e-01
9.24384892e-01 7.40594327e-01 5.58553040e-01 -2.06886809e-02
-1.27009317e-01 2.64357656e-01 -4.56855088e-01 -5.65459765e-02
1.40956163e-01 -1.53986141e-01 -2.03301441e-02 3.91160965e-01
2.81551600e-01 -1.46979585e-01 -8.58154655e-01 7.29529083e-01
-1.04223168e+00 5.88677265e-02 9.11264062e-01 8.92771661e-01
6.63509190e-01 3.19304243e-02 2.94979095e-01 4.81141731e-03
2.48034582e-01 -2.44936049e-01 -5.70177555e-01 5.56024790e-01
-2.22309962e-01 -9.75376144e-02 1.19749844e-01 -6.67548597e-01
-1.01902878e+00 4.01060224e-01 3.79985571e-01 -6.04730427e-01
-1.99885979e-01 3.41102690e-01 -4.36641276e-01 -2.60990411e-01
3.11192840e-01 2.04050586e-01 -6.63171243e-03 -3.02438468e-01
3.04532021e-01 8.24674487e-01 5.49025953e-01 -3.45317125e-01
9.07242477e-01 5.76072335e-01 1.32053010e-02 -3.15497845e-01
-5.35984516e-01 -4.33703452e-01 -6.25377119e-01 -5.64412773e-01
7.06923902e-01 -8.78723979e-01 -9.10631955e-01 5.52615345e-01
-1.23755634e+00 2.69991100e-01 1.58497721e-01 6.49045408e-01
-1.34487867e-01 4.52293426e-01 -2.39303842e-01 -7.70290852e-01
-6.34692535e-02 -1.28722358e+00 1.35094094e+00 7.84511626e-01
2.96794772e-01 -4.45521653e-01 -1.72465771e-01 2.80703396e-01
2.66538739e-01 2.13455379e-01 6.78803205e-01 -4.90389913e-01
-6.87129200e-01 -2.18350813e-01 -7.17085898e-01 1.31140590e-01
2.32283235e-01 1.95472479e-01 -7.86021292e-01 -1.28922090e-01
-3.51248741e-01 -7.46072978e-02 8.98778319e-01 2.97243655e-01
1.15531850e+00 1.95066944e-01 -4.70118135e-01 4.68186945e-01
1.27532899e+00 3.44956905e-01 3.75417084e-01 3.68318051e-01
7.78515697e-01 6.60667479e-01 8.45908701e-01 5.54717600e-01
1.91557184e-01 8.77380848e-01 6.43097758e-01 -2.68046558e-01
7.14367777e-02 -3.40586230e-02 1.89587936e-01 3.90717894e-01
1.01641797e-01 -1.56970069e-01 -7.03986704e-01 2.72657722e-01
-2.00454736e+00 -6.21699393e-01 9.46905464e-02 2.21540046e+00
6.04900002e-01 -7.66052157e-02 1.07847877e-01 -5.61373197e-02
1.11950541e+00 -4.47161235e-02 -6.42660737e-01 2.83000469e-01
-5.06039798e-01 -4.31293577e-01 5.44406772e-01 2.55037099e-02
-1.21010387e+00 4.11073804e-01 4.80611706e+00 8.74990225e-01
-1.36199725e+00 -6.05943538e-02 5.42423606e-01 -1.57296970e-01
-2.07915187e-01 -1.69098794e-01 -6.70859218e-01 6.16364181e-01
-1.01519801e-01 3.12981784e-01 2.70968467e-01 6.23029649e-01
1.23653412e-01 -2.74307758e-01 -9.52650309e-01 1.28949440e+00
2.18187958e-01 -8.67603242e-01 -1.54635414e-01 -1.31838292e-01
5.59838474e-01 -1.70940414e-01 1.34591281e-01 -7.42759258e-02
-3.36588860e-01 -6.22325540e-01 9.00755465e-01 7.82756627e-01
4.36548799e-01 -8.97139430e-01 9.15285885e-01 5.01731336e-02
-1.30661559e+00 -2.72814274e-01 -3.59160572e-01 6.06111109e-01
1.15627564e-01 5.19444704e-01 -4.51087058e-01 7.75831938e-01
8.45370710e-01 6.67353153e-01 -7.23750114e-01 1.27189958e+00
1.25626072e-01 -2.92174399e-01 -4.94796932e-01 5.06252982e-02
5.33637851e-02 -3.35862264e-02 5.72646499e-01 8.96548688e-01
4.66177523e-01 1.22654736e-02 1.86067641e-01 8.21372986e-01
1.27859488e-01 -3.85358222e-02 -3.56523633e-01 2.06299365e-01
5.64039230e-01 1.40034604e+00 -8.63988459e-01 4.12254483e-02
-4.85635549e-01 1.02755117e+00 8.76140967e-02 4.42178160e-01
-9.41860199e-01 -3.65296334e-01 3.46018404e-01 -3.39343280e-01
4.96970415e-01 -1.75122678e-01 -1.19009435e-01 -1.07226396e+00
1.14654310e-01 -5.63536227e-01 4.23879117e-01 -9.64056790e-01
-1.56728685e+00 4.46243852e-01 3.49740754e-03 -1.80226707e+00
2.50919074e-01 -6.54989123e-01 -3.66814256e-01 7.04214334e-01
-1.49186456e+00 -1.26285183e+00 -5.46754539e-01 7.93043673e-01
1.28332943e-01 -5.90961799e-02 1.51590466e-01 5.46034455e-01
-8.49157691e-01 7.41402149e-01 -1.04838520e-01 1.73459113e-01
9.81976867e-01 -8.04038882e-01 -3.74350071e-01 7.19616234e-01
-1.61537468e-01 6.12232625e-01 3.84902209e-01 -4.84850585e-01
-1.48909843e+00 -9.23422575e-01 3.19269925e-01 -2.33661234e-02
3.15241218e-01 -4.58174609e-02 -8.66204441e-01 2.17335254e-01
-2.39333138e-02 2.79476196e-01 5.78598201e-01 -2.41865933e-01
-5.05332351e-01 -5.29527605e-01 -1.12970603e+00 5.12161613e-01
7.92228997e-01 -4.68906999e-01 -3.58243942e-01 2.01265112e-01
6.27964675e-01 -6.99298561e-01 -7.08283126e-01 8.00574839e-01
7.65429914e-01 -9.94494796e-01 9.70189333e-01 -3.43110770e-01
1.54487088e-01 -9.27900791e-01 -3.34304482e-01 -8.91315043e-01
-1.77975729e-01 -1.86908424e-01 -2.15601865e-02 1.34981501e+00
2.42411897e-01 -5.56977391e-01 5.25413930e-01 5.95056593e-01
-1.26470566e-01 -7.67575383e-01 -1.24584270e+00 -6.12619817e-01
-5.09520411e-01 -3.35703582e-01 6.97087765e-01 8.09002995e-01
-1.17111139e-01 6.71842024e-02 -4.14347917e-01 4.67082947e-01
3.87725353e-01 4.20060724e-01 6.33605063e-01 -8.78357410e-01
-4.20862101e-02 -7.62071848e-01 -5.28186798e-01 -1.10089457e+00
-2.19365016e-01 -5.82216561e-01 2.43447423e-01 -1.14797604e+00
3.76850367e-01 -4.25440907e-01 -6.97713614e-01 5.36862135e-01
-2.25576371e-01 3.46843392e-01 2.43893877e-01 2.39136145e-01
-7.74967790e-01 6.65140688e-01 1.27815592e+00 -1.80955544e-01
-2.56731540e-01 -8.80605821e-03 -5.07462382e-01 5.06624877e-01
5.92350602e-01 -3.11897367e-01 -4.44562584e-02 -4.38050300e-01
3.13174635e-01 -3.47661600e-02 6.31555915e-01 -8.91212940e-01
5.15635788e-01 -1.02987930e-01 8.31250131e-01 -1.09238541e+00
1.90135330e-01 -1.38305724e+00 -6.10712580e-02 1.97784975e-01
-1.87156737e-01 1.07365899e-01 1.83949322e-01 6.59287930e-01
-3.29416156e-01 -6.69365600e-02 7.84515083e-01 1.60290778e-01
-8.49978149e-01 4.07918960e-01 3.29785720e-02 -4.72413749e-01
1.10529554e+00 -4.35628563e-01 -4.36560631e-01 7.49350665e-03
-4.34059381e-01 5.20854712e-01 3.04320127e-01 6.46666706e-01
9.55636442e-01 -1.71973050e+00 -3.84093195e-01 3.76620144e-01
4.75269496e-01 1.65170223e-01 8.01415384e-01 1.22879720e+00
-7.64418021e-02 1.04921550e-01 -1.63646534e-01 -9.55423772e-01
-1.27682734e+00 5.19224524e-01 5.51881611e-01 2.62986273e-01
-2.29436770e-01 9.10967588e-01 1.14515543e-01 -3.44577581e-01
3.71687055e-01 -3.34520310e-01 -2.13435382e-01 2.88089395e-01
5.40195942e-01 1.73329413e-01 6.04064427e-02 -7.45224953e-01
-6.58005893e-01 9.17771578e-01 -1.24501392e-01 1.95101164e-02
9.09586608e-01 -5.56255043e-01 -1.91214919e-01 4.04645860e-01
1.15148306e+00 -9.31555703e-02 -1.35381484e+00 -5.80096364e-01
-3.11295003e-01 -7.29950368e-01 -1.71085261e-02 -6.65376008e-01
-1.40961576e+00 6.90187633e-01 1.05566466e+00 4.96035367e-02
1.43342090e+00 1.38687596e-01 7.11285591e-01 2.26533785e-01
1.09448925e-01 -1.04400122e+00 2.50402272e-01 2.13947788e-01
7.75826514e-01 -1.55449736e+00 2.17379145e-02 -2.59198904e-01
-7.40340114e-01 1.35788691e+00 9.37490582e-01 3.84247571e-01
4.49319661e-01 -5.03328815e-02 8.13965425e-02 -1.31143779e-01
-2.61626750e-01 -3.17205369e-01 5.67787588e-01 4.21052665e-01
7.46018067e-02 -2.35330850e-01 -8.55221748e-02 6.09920084e-01
5.23786128e-01 -3.22596222e-01 3.52050066e-02 1.04371977e+00
-4.92415279e-01 -8.71221244e-01 -7.29361057e-01 1.24779820e-01
-9.91548747e-02 2.39699274e-01 -3.28832179e-01 8.74383509e-01
5.49060702e-01 9.37328160e-01 2.48544842e-01 -7.80506849e-01
4.48919684e-01 -2.92573988e-01 4.54987913e-01 -1.45998761e-01
-3.61403078e-01 5.31115592e-01 -5.06501734e-01 -5.96649885e-01
-8.20202410e-01 -7.18614101e-01 -1.34684312e+00 -4.74330932e-02
-7.14869022e-01 -2.78288603e-01 7.03557909e-01 9.76935685e-01
3.41246367e-01 5.06140649e-01 9.69031274e-01 -1.00255215e+00
-1.54180989e-01 -8.36845994e-01 -5.74313939e-01 2.91834831e-01
4.84332800e-01 -9.28920627e-01 -2.79901981e-01 -3.05943131e-01] | [9.711146354675293, -0.9855990409851074] |
540b8fb8-0ca2-457e-9b23-ebe7a7ec9edd | structured-3d-features-for-reconstructing | 2212.06820 | null | https://arxiv.org/abs/2212.06820v3 | https://arxiv.org/pdf/2212.06820v3.pdf | Structured 3D Features for Reconstructing Controllable Avatars | We introduce Structured 3D Features, a model based on a novel implicit 3D representation that pools pixel-aligned image features onto dense 3D points sampled from a parametric, statistical human mesh surface. The 3D points have associated semantics and can move freely in 3D space. This allows for optimal coverage of the person of interest, beyond just the body shape, which in turn, additionally helps modeling accessories, hair, and loose clothing. Owing to this, we present a complete 3D transformer-based attention framework which, given a single image of a person in an unconstrained pose, generates an animatable 3D reconstruction with albedo and illumination decomposition, as a result of a single end-to-end model, trained semi-supervised, and with no additional postprocessing. We show that our S3F model surpasses the previous state-of-the-art on various tasks, including monocular 3D reconstruction, as well as albedo and shading estimation. Moreover, we show that the proposed methodology allows novel view synthesis, relighting, and re-posing the reconstruction, and can naturally be extended to handle multiple input images (e.g. different views of a person, or the same view, in different poses, in video). Finally, we demonstrate the editing capabilities of our model for 3D virtual try-on applications. | ['Cristian Sminchisescu', 'Andrei Zanfir', 'Eduard Gabriel Bazavan', 'Thiemo Alldieck', 'Mihai Zanfir', 'Enric Corona'] | 2022-12-13 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Corona_Structured_3D_Features_for_Reconstructing_Controllable_Avatars_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Corona_Structured_3D_Features_for_Reconstructing_Controllable_Avatars_CVPR_2023_paper.pdf | cvpr-2023-1 | ['virtual-try-on'] | ['computer-vision'] | [ 1.13094352e-01 2.51362681e-01 2.45988771e-01 -2.44902253e-01
-2.41424695e-01 -4.68644768e-01 5.70571423e-01 -3.90137643e-01
8.08348060e-02 4.58897650e-01 3.70218515e-01 3.11665922e-01
4.01514024e-01 -6.91502869e-01 -9.34012413e-01 -3.08749855e-01
3.06456268e-01 7.72014380e-01 4.81964052e-02 -1.22050978e-01
-1.48054421e-01 8.54176104e-01 -1.86113298e+00 1.75545961e-02
6.23490274e-01 9.42233980e-01 9.75346565e-02 6.14976287e-01
1.83973715e-01 4.05964494e-01 -3.73230904e-01 -6.75868273e-01
5.05489409e-01 -1.72631487e-01 -3.64025533e-01 9.53959286e-01
1.15547669e+00 -7.35360086e-01 -3.46166581e-01 5.26873469e-01
5.29647171e-01 8.10146630e-02 6.26297355e-01 -9.75627184e-01
-5.72220266e-01 -2.34993473e-01 -5.93118370e-01 -6.61044717e-01
1.03417742e+00 1.65043756e-01 7.18908072e-01 -1.36683714e+00
1.05549073e+00 1.59175718e+00 7.07757354e-01 6.89900637e-01
-1.40783095e+00 -1.38324961e-01 2.49241427e-01 -1.81803554e-01
-1.20156109e+00 -3.81152421e-01 1.00096679e+00 -5.97899318e-01
6.80433452e-01 3.19856614e-01 1.43274236e+00 1.27209520e+00
1.69304252e-01 8.57039392e-01 1.19325292e+00 -5.05280495e-01
-3.83637585e-02 1.35770157e-01 -1.85428739e-01 9.23532844e-01
-1.33949351e-02 7.79901817e-02 -8.49550664e-01 -8.92077014e-02
1.23368073e+00 3.22192281e-01 -5.88833690e-01 -1.17303276e+00
-1.20985484e+00 4.44189429e-01 3.56923789e-01 -3.45678508e-01
-3.52189273e-01 2.44751036e-01 -8.75943676e-02 -3.17977779e-02
7.67290056e-01 -1.67787701e-01 -2.82492578e-01 2.27114320e-01
-8.03328812e-01 6.35433614e-01 6.92098796e-01 1.12606454e+00
8.93023431e-01 1.30431086e-01 -4.83709648e-02 6.66056633e-01
4.25498575e-01 1.04534364e+00 -1.50753453e-01 -1.21171737e+00
2.66920716e-01 5.61052620e-01 4.80240643e-01 -7.73252547e-01
-3.66392493e-01 -3.10509562e-01 -7.54104376e-01 6.28154993e-01
1.29538924e-01 1.02124304e-01 -1.02838409e+00 1.68855929e+00
9.44218934e-01 -2.09595859e-02 -3.52105618e-01 1.12924528e+00
9.09246743e-01 3.12175035e-01 -5.70921600e-01 -4.01583174e-03
1.39989114e+00 -9.12469685e-01 -5.78234434e-01 -2.46716887e-01
-1.19327873e-01 -6.37757242e-01 1.16581786e+00 5.48725843e-01
-1.51201010e+00 -6.04874194e-01 -8.34120572e-01 -5.97137451e-01
-4.15580580e-03 -5.07136323e-02 4.72204030e-01 6.49996102e-01
-1.06214261e+00 5.79278886e-01 -7.34168708e-01 -5.88861227e-01
2.56535888e-01 1.22126862e-01 -5.83146811e-01 -2.98660159e-01
-7.83030927e-01 9.28736508e-01 -3.06738704e-01 1.72217205e-01
-8.34494829e-01 -7.03294039e-01 -1.24086070e+00 -1.89971298e-01
3.61067861e-01 -1.55035257e+00 8.44386756e-01 -9.33576643e-01
-1.68753231e+00 1.46814430e+00 -3.05456549e-01 -1.10371061e-01
9.41068888e-01 -6.68294072e-01 1.56430110e-01 2.04323977e-01
-2.60275919e-02 6.54372811e-01 1.11345255e+00 -1.60708058e+00
1.01290189e-01 -7.90872276e-01 2.61553675e-01 7.80774474e-01
1.30874783e-01 -2.27989107e-01 -7.80417442e-01 -6.70071840e-01
1.58280253e-01 -9.81399894e-01 -1.18503079e-01 8.14059973e-01
-5.49612582e-01 3.38270873e-01 3.91211748e-01 -8.24866235e-01
3.95050853e-01 -2.07915473e+00 7.28657722e-01 2.03555509e-01
2.52981395e-01 -1.83306888e-01 2.64897257e-01 2.74809629e-01
9.20379683e-02 -2.00828120e-01 -3.46972644e-01 -1.03539920e+00
1.39423311e-01 5.76174222e-02 -1.50620386e-01 7.38697052e-01
6.63049286e-03 7.60123014e-01 -6.58641040e-01 -4.10058171e-01
6.96881175e-01 9.49263513e-01 -6.50149107e-01 5.40889025e-01
-8.38682875e-02 7.39668608e-01 -2.05747947e-01 7.86152899e-01
8.59525919e-01 -1.70270175e-01 4.14697379e-02 -1.82562947e-01
6.02949820e-02 -6.56592194e-03 -1.48737788e+00 2.24109936e+00
-6.67813420e-01 2.43826777e-01 3.55668813e-01 -4.08770770e-01
7.80143976e-01 1.88727066e-01 5.34276843e-01 -3.67742509e-01
5.21475216e-03 6.99116215e-02 -8.53923023e-01 -2.78847396e-01
3.81888092e-01 -9.21199992e-02 5.20479642e-02 3.06808710e-01
-6.37374446e-02 -5.18210649e-01 -3.76762122e-01 2.34857537e-02
5.39659679e-01 7.89513469e-01 3.60351950e-01 -8.66227150e-02
4.10128653e-01 -3.82625282e-01 3.61847073e-01 2.77996033e-01
2.92858332e-01 1.19887447e+00 1.21786982e-01 -6.03372395e-01
-1.11906624e+00 -1.47299671e+00 -9.41973329e-02 5.49254954e-01
3.68869632e-01 -1.12046488e-01 -7.46955395e-01 -4.98515069e-01
3.48984957e-01 4.68093336e-01 -6.99083507e-01 3.14557970e-01
-5.65401375e-01 1.80521552e-02 -5.63710369e-02 3.20250154e-01
3.33751380e-01 -8.46468866e-01 -1.04863524e+00 2.94606872e-02
-2.29687512e-01 -1.14594758e+00 -6.89690471e-01 -1.80522293e-01
-8.62073362e-01 -9.83218431e-01 -1.15963721e+00 -4.50135946e-01
8.08761239e-01 4.52599436e-01 1.50164115e+00 -7.14350585e-03
-3.39811385e-01 1.01704121e+00 -1.16908096e-01 -1.81217581e-01
-6.33386075e-02 -5.33261538e-01 1.74812332e-01 3.44680965e-01
-2.11382404e-01 -7.80102432e-01 -8.09917867e-01 2.79971242e-01
-5.07212520e-01 5.26225328e-01 5.50942793e-02 9.30910170e-01
8.70094657e-01 -7.33777404e-01 -1.50099233e-01 -8.05515528e-01
-2.12783128e-01 7.18539581e-03 -4.71038818e-01 1.20206133e-01
-3.45666371e-02 -2.52863735e-01 3.10034126e-01 -3.01085591e-01
-1.03321254e+00 2.67506570e-01 -1.44825997e-02 -1.02737224e+00
-1.32677212e-01 -1.63865313e-01 -5.29792309e-01 -2.48206668e-02
5.89906454e-01 1.41713768e-01 2.27245823e-01 -6.11035109e-01
6.36932850e-01 3.34717184e-01 5.81932425e-01 -4.49635834e-01
1.05255342e+00 9.57144558e-01 1.50676504e-01 -1.00925565e+00
-8.38407874e-01 -1.74084067e-01 -1.11883783e+00 -3.91502053e-01
8.33747268e-01 -1.20749032e+00 -6.91344202e-01 6.25535727e-01
-1.26515782e+00 -4.00131047e-01 -5.94714403e-01 3.36331934e-01
-9.20428455e-01 5.09236157e-01 -4.61786956e-01 -8.79954934e-01
-3.27478081e-01 -9.22168314e-01 1.73388588e+00 -1.85071342e-02
-1.88979745e-01 -9.14149165e-01 -1.38224915e-01 7.00661302e-01
-6.85626641e-02 7.68261611e-01 6.26280546e-01 2.68185884e-01
-8.29447508e-01 5.89186251e-02 1.93356171e-01 2.62666970e-01
-3.33204158e-02 -2.08507329e-01 -1.20974243e+00 -4.01785076e-01
-9.75493565e-02 -4.50080752e-01 5.95058620e-01 5.44744015e-01
9.56734002e-01 -1.80698559e-01 -1.40339389e-01 1.02893090e+00
1.24742794e+00 -5.38945019e-01 5.80756903e-01 -2.00223513e-02
1.06815898e+00 7.91388750e-01 5.40092826e-01 6.44502819e-01
6.28026426e-01 1.09983087e+00 7.66636670e-01 -4.19036925e-01
-5.20989776e-01 -4.59856033e-01 3.61828923e-01 4.60109085e-01
-4.77352560e-01 -1.60959974e-01 -3.86234999e-01 3.60082448e-01
-1.45244741e+00 -7.37213433e-01 -1.45693123e-01 2.67967129e+00
4.26307589e-01 -3.15395504e-01 1.94981143e-01 -7.98587967e-03
4.22289580e-01 2.58123875e-01 -7.55299151e-01 -2.29168504e-01
-1.52343974e-01 2.86559135e-01 2.68246531e-01 6.81246042e-01
-8.50857019e-01 8.25062931e-01 5.89493895e+00 1.88648999e-01
-8.90170872e-01 3.28769162e-02 3.62979114e-01 -3.63056690e-01
-8.06248724e-01 -2.40385041e-01 -5.72590172e-01 9.81581882e-02
8.67804885e-02 2.47500792e-01 6.76974297e-01 7.68762171e-01
2.06588492e-01 2.50891019e-02 -1.36796284e+00 1.36451077e+00
5.82943022e-01 -1.09762311e+00 1.01572715e-01 3.24948758e-01
7.41092920e-01 -3.34576100e-01 4.46960814e-02 -1.00293547e-01
-5.38508184e-02 -9.13795054e-01 1.25825143e+00 7.06307948e-01
1.13647485e+00 -5.10996103e-01 2.78766900e-01 3.49647343e-01
-1.04695916e+00 2.07766026e-01 -2.36027777e-01 1.33295938e-01
5.32301664e-01 6.97233021e-01 -4.02929038e-01 7.04416633e-01
7.82767773e-01 8.10571969e-01 -2.64936805e-01 6.53932631e-01
-3.05625170e-01 -2.04377592e-01 -3.49252135e-01 3.07795107e-01
-3.66874546e-01 -3.40175122e-01 7.67414331e-01 8.28777432e-01
5.53255618e-01 7.69284219e-02 3.07287812e-01 9.46543097e-01
-1.05664551e-01 2.01999918e-01 -8.48854959e-01 7.32555747e-01
1.66456670e-01 9.97385383e-01 -2.60278285e-01 -2.28760898e-01
-4.38068509e-01 1.61968160e+00 4.13744837e-01 4.82671946e-01
-7.96961486e-01 1.66788712e-01 7.22106814e-01 7.03309476e-01
3.65352482e-01 -2.10847884e-01 -1.19494282e-01 -1.63859761e+00
4.63298053e-01 -7.82280028e-01 -1.11259624e-01 -1.24858522e+00
-1.12622380e+00 4.82508391e-01 -7.95199070e-03 -1.49685180e+00
-3.15859169e-01 -5.22849977e-01 -1.87140271e-01 9.07197356e-01
-1.36550879e+00 -1.55531073e+00 -7.29266405e-01 7.87220895e-01
6.59335554e-01 2.17092261e-01 8.14166486e-01 7.05651045e-02
-7.79440477e-02 3.29236984e-01 -2.27624580e-01 -3.19147348e-01
7.34231532e-01 -1.43777359e+00 5.42262316e-01 6.26965225e-01
2.25048289e-01 4.88514960e-01 5.09633005e-01 -4.98974264e-01
-1.87268281e+00 -1.00595820e+00 6.70534909e-01 -7.17921495e-01
-1.28687859e-01 -7.86838531e-01 -6.01663351e-01 9.03515697e-01
4.30468693e-02 1.83411255e-01 3.88798058e-01 1.15684785e-01
-3.04788113e-01 -1.17755756e-01 -1.34639776e+00 8.22778702e-01
1.57681036e+00 -3.00075442e-01 -4.99413788e-01 3.56684357e-01
5.91163635e-01 -1.09749842e+00 -8.83670568e-01 1.07331596e-01
9.06926215e-01 -1.43965673e+00 1.43568313e+00 -1.86547428e-01
3.09673786e-01 -2.18585476e-01 -3.18936467e-01 -1.34718060e+00
-3.14719915e-01 -7.81023622e-01 -4.72514242e-01 8.65954757e-01
-1.56399027e-01 -3.78126144e-01 7.65659690e-01 6.52962625e-01
-2.11494967e-01 -7.67346084e-01 -9.09046531e-01 -4.80294704e-01
-3.41049850e-01 -3.93845320e-01 5.96975505e-01 5.70630491e-01
-5.78892469e-01 1.83176517e-01 -9.64813471e-01 2.80607402e-01
1.10057378e+00 4.45564717e-01 1.42383552e+00 -1.48106003e+00
-4.65679675e-01 2.87198592e-02 -4.11991149e-01 -1.53961492e+00
1.90298453e-01 -7.13639200e-01 -1.58365428e-01 -1.55695462e+00
1.07175089e-01 -1.17383748e-01 5.67336619e-01 2.60910690e-01
-1.19554892e-01 4.11932081e-01 3.99859071e-01 3.23185548e-02
-3.57001610e-02 8.43200922e-01 1.65329897e+00 -2.52043381e-02
-2.30268732e-01 -1.65958330e-02 -3.93041372e-01 9.83362198e-01
4.29437496e-02 1.08922776e-02 -4.29977864e-01 -6.26363933e-01
9.80160013e-02 2.88747013e-01 8.13663960e-01 -7.48673856e-01
-4.38977093e-01 -1.12957917e-01 8.34531546e-01 -7.26541400e-01
1.09000850e+00 -9.77996409e-01 4.80272472e-01 2.64309108e-01
1.08118705e-01 -1.51842296e-01 -1.59957498e-01 6.52225435e-01
3.43504459e-01 1.58679754e-01 7.49613166e-01 -3.97779822e-01
-3.11362416e-01 6.50086343e-01 1.24521382e-01 1.86373871e-02
8.56821597e-01 -6.71119988e-01 2.82148957e-01 -4.47113723e-01
-9.51081753e-01 7.71577582e-02 1.23961198e+00 2.34844953e-01
9.48498726e-01 -1.53980529e+00 -8.67599845e-01 5.16161621e-01
1.06978744e-01 3.68723691e-01 5.17573595e-01 4.93649274e-01
-7.09572852e-01 -8.94579366e-02 -1.36232212e-01 -1.02891886e+00
-1.40749776e+00 4.67672557e-01 4.52559769e-01 -3.82492468e-02
-1.22840357e+00 4.34636950e-01 6.15392923e-01 -6.76634848e-01
2.66792476e-01 -3.93674076e-01 1.10950619e-01 -2.54121423e-01
3.76964539e-01 2.69738615e-01 -6.49493262e-02 -8.14127624e-01
-3.23572665e-01 1.30586255e+00 5.06186962e-01 -3.12865555e-01
1.21720076e+00 -3.87808591e-01 1.13671176e-01 5.78942358e-01
7.88006902e-01 3.46857369e-01 -1.68839955e+00 -2.69617110e-01
-9.98270810e-01 -9.65493381e-01 -2.46303126e-01 -6.27507508e-01
-1.07157302e+00 9.61979687e-01 3.81979495e-01 -2.34643430e-01
1.01196909e+00 1.52904373e-02 6.16472542e-01 4.22730856e-02
9.26845074e-01 -6.23103678e-01 6.34576082e-02 1.69402689e-01
1.35405087e+00 -9.64535713e-01 4.14363742e-01 -7.39181280e-01
-7.62739122e-01 9.11574423e-01 4.15465832e-01 -2.12993562e-01
5.70825219e-01 -6.40997812e-02 -8.30090493e-02 -2.35837460e-01
-4.31459039e-01 -1.13510296e-01 4.89574343e-01 9.67691302e-01
2.14243948e-01 8.71452913e-02 1.80358455e-01 1.59683213e-01
-2.70729661e-01 -1.93959847e-01 5.47808528e-01 6.95958853e-01
-1.24686427e-01 -8.45747948e-01 -7.14266658e-01 3.35449316e-02
-1.25185147e-01 1.40732080e-01 -3.25140804e-01 8.07269096e-01
2.07358629e-01 5.50422072e-01 9.03528407e-02 4.13527712e-02
6.95328832e-01 -9.40190554e-02 1.00956857e+00 -7.90469944e-01
-3.79471660e-01 2.94108689e-01 9.91882756e-02 -8.69479716e-01
-5.35104811e-01 -7.80498683e-01 -8.44557226e-01 -2.10066035e-01
-7.40540773e-02 -4.94783521e-01 4.59748656e-01 5.73083103e-01
3.52891654e-01 1.23636752e-01 9.22790766e-01 -1.66881704e+00
-2.51607060e-01 -7.59299815e-01 -8.26834142e-01 7.96502888e-01
4.37178344e-01 -1.02484667e+00 -3.25326324e-01 3.12472433e-01] | [7.2329583168029785, -1.3026989698410034] |
2dd4899c-8102-40ee-95cb-e3c64deacc76 | chinese-idiom-paraphrasing | 2204.07555 | null | https://arxiv.org/abs/2204.07555v2 | https://arxiv.org/pdf/2204.07555v2.pdf | Chinese Idiom Paraphrasing | Idioms, are a kind of idiomatic expression in Chinese, most of which consist of four Chinese characters. Due to the properties of non-compositionality and metaphorical meaning, Chinese Idioms are hard to be understood by children and non-native speakers. This study proposes a novel task, denoted as Chinese Idiom Paraphrasing (CIP). CIP aims to rephrase idioms-included sentences to non-idiomatic ones under the premise of preserving the original sentence's meaning. Since the sentences without idioms are easier handled by Chinese NLP systems, CIP can be used to pre-process Chinese datasets, thereby facilitating and improving the performance of Chinese NLP tasks, e.g., machine translation system, Chinese idiom cloze, and Chinese idiom embeddings. In this study, CIP task is treated as a special paraphrase generation task. To circumvent difficulties in acquiring annotations, we first establish a large-scale CIP dataset based on human and machine collaboration, which consists of 115,530 sentence pairs. We further deploy three baselines and two novel CIP approaches to deal with CIP problems. The results show that the proposed methods have better performances than the baselines based on the established CIP dataset. | ['Xindong Wu', 'Yi Zhu', 'Yunhao Yuan', 'Yun Li', 'Chaowei Zhang', 'Yang Li', 'Jipeng Qiang'] | 2022-04-15 | null | null | null | null | ['paraphrase-generation', 'paraphrase-generation'] | ['computer-code', 'natural-language-processing'] | [ 1.59656614e-01 -1.54502302e-01 -2.99593806e-01 -3.46295029e-01
-3.73080641e-01 -8.04364860e-01 4.71264660e-01 -3.09318602e-01
-2.60161757e-01 6.59582973e-01 7.44817674e-01 -3.69325906e-01
2.92740524e-01 -6.45681679e-01 -2.30049163e-01 -3.52384269e-01
6.27711236e-01 6.69482768e-01 -3.45765688e-02 -5.49299657e-01
4.62918103e-01 4.34898352e-03 -8.66268158e-01 1.04952109e+00
1.37449014e+00 3.89494598e-01 3.36114019e-01 -2.41233390e-02
-3.76671642e-01 9.79911089e-01 -8.16914797e-01 -9.28772032e-01
-5.86062409e-02 -9.57850575e-01 -1.10128987e+00 -3.60632896e-01
-2.54148282e-02 -1.55313239e-01 1.66047156e-01 1.24308288e+00
3.85001034e-01 -1.86779767e-01 4.81387854e-01 -8.44974875e-01
-1.58346415e+00 1.23793805e+00 -4.09498453e-01 5.61751276e-02
7.25952446e-01 -5.11024117e-01 1.05152488e+00 -1.00432670e+00
8.78130138e-01 1.19564569e+00 5.01885712e-01 8.86195540e-01
-9.50637937e-01 -8.51505399e-01 -7.21419416e-03 6.04743719e-01
-1.28792131e+00 -2.05211207e-01 1.28366387e+00 -4.48872186e-02
9.74797189e-01 4.94614363e-01 4.42898899e-01 1.26646888e+00
-4.04124409e-02 8.80425751e-01 1.14679897e+00 -5.72062671e-01
-2.71432579e-01 3.93519938e-01 -2.17393078e-02 3.22564170e-02
-2.60895342e-01 -3.83410543e-01 -4.17109162e-01 3.46793532e-01
2.25432977e-01 -1.18829034e-01 -3.24086905e-01 3.43367934e-01
-1.34327471e+00 8.85334551e-01 4.07612771e-01 6.53248668e-01
1.16028354e-01 -6.34913683e-01 8.24687481e-01 7.29179502e-01
3.86709630e-01 6.67725921e-01 -5.34319997e-01 -3.46886843e-01
-2.99926579e-01 3.06220740e-01 6.50723219e-01 1.38558424e+00
1.82026803e-01 -3.84755880e-01 2.83999830e-01 1.36773658e+00
-1.84866160e-01 5.50378084e-01 1.18850338e+00 -5.31243205e-01
8.62491310e-01 7.96626449e-01 -3.04676324e-01 -1.15507889e+00
1.52777672e-01 5.86270876e-02 -8.52207780e-01 -8.82143438e-01
-4.33312953e-02 -4.55536917e-02 -1.36081755e-01 1.80584538e+00
3.72921932e-03 -3.83635372e-01 4.33417857e-01 9.23443556e-01
1.09091091e+00 9.71706271e-01 -4.25138205e-01 -3.72900635e-01
1.47849452e+00 -1.41261327e+00 -7.73577869e-01 -1.31596282e-01
5.90478361e-01 -1.23200476e+00 1.63428175e+00 2.42188916e-01
-1.11959100e+00 -4.77620780e-01 -8.89581442e-01 -4.11543220e-01
-4.50505644e-01 1.81740299e-01 3.06160152e-01 5.96228302e-01
-5.70658624e-01 3.26122999e-01 -5.46230674e-01 -1.54081374e-01
1.43709198e-01 -1.60439432e-01 -3.87701094e-01 -1.62568957e-01
-1.32004774e+00 1.17073357e+00 7.86048532e-01 1.01947322e-01
-6.74015507e-02 -6.77170873e-01 -8.66383374e-01 -3.61893512e-02
-9.92415100e-02 -2.50831395e-01 1.17830706e+00 -1.41547990e+00
-1.78704345e+00 1.25419271e+00 -3.15514416e-01 -7.35967532e-02
2.39826262e-01 -3.88157666e-01 -6.25570118e-01 2.86793313e-03
5.18504322e-01 4.91109073e-01 4.39442903e-01 -7.26053774e-01
-2.86141425e-01 -2.34849960e-01 5.13996296e-02 3.90735984e-01
-3.96287054e-01 9.27496433e-01 -2.45284781e-01 -9.28637326e-01
3.43258202e-01 -1.02535939e+00 2.89931327e-01 -6.67163908e-01
-4.37719762e-01 -1.31012917e-01 1.04825175e+00 -9.89182413e-01
1.24107015e+00 -2.10182452e+00 4.36341256e-01 -4.44515228e-01
-3.45874608e-01 3.33090454e-01 -1.59297243e-01 7.96516538e-01
-1.17845185e-01 2.05140144e-01 -3.09013009e-01 -1.95839703e-01
-1.34643510e-01 3.28005344e-01 -5.84910154e-01 -1.94639757e-01
3.85277152e-01 1.24646771e+00 -9.48064864e-01 -4.66123581e-01
-8.74205083e-02 -3.91412154e-02 -6.62711620e-01 2.54519403e-01
-1.01612695e-01 8.08924496e-01 -1.14518411e-01 6.27970278e-01
9.05984581e-01 1.83190718e-01 6.13098502e-01 2.35255927e-01
-2.40539327e-01 1.16626060e+00 -3.74152154e-01 1.79077065e+00
-8.51984084e-01 4.49799001e-01 -3.53580892e-01 -1.14896107e+00
9.48510408e-01 2.80753225e-01 -1.05603762e-01 -9.45687711e-01
-1.00718429e-02 4.49411541e-01 3.67516398e-01 -4.00243938e-01
4.19172585e-01 -4.73225564e-01 -4.24060583e-01 6.79637671e-01
-4.03408170e-01 -3.72634768e-01 1.81631118e-01 -2.44163256e-02
7.64102161e-01 -1.01204701e-01 6.19883597e-01 -6.28416061e-01
1.06494927e+00 1.32226095e-01 8.32012713e-01 1.23300299e-01
2.68184662e-01 6.52014077e-01 5.34063280e-01 -5.63124001e-01
-8.48865271e-01 -1.42901778e+00 -3.11758459e-01 8.48071814e-01
3.45213264e-01 -5.58507740e-01 -8.13403964e-01 -7.89695144e-01
-5.05336106e-01 7.65905797e-01 -4.14865106e-01 -8.93833414e-02
-1.31872296e+00 -7.70766199e-01 8.17019284e-01 5.31099141e-01
8.68722558e-01 -1.60301244e+00 -3.61987144e-01 1.69065490e-01
-7.41252661e-01 -1.22081041e+00 -7.20813870e-01 -3.85426611e-01
-4.86919045e-01 -8.67689312e-01 -4.36949164e-01 -1.58203685e+00
8.22420418e-01 4.03330475e-01 1.05420482e+00 2.72611082e-01
2.09338099e-01 -6.32848978e-01 -9.35912251e-01 -3.59817326e-01
-7.37538099e-01 1.39158189e-01 -1.57053918e-01 -5.73716521e-01
6.19105577e-01 -7.39761591e-01 -1.47805333e-01 1.83301032e-01
-8.40097666e-01 7.55145431e-01 3.96509916e-01 1.25984895e+00
1.46153659e-01 -1.42921180e-01 6.71522141e-01 -1.18697953e+00
8.39785755e-01 -3.81995618e-01 -2.20673129e-01 3.26669812e-01
-1.97168007e-01 -1.79607764e-01 1.31518066e+00 -7.04149961e-01
-1.49061513e+00 -4.64196742e-01 -3.56012404e-01 1.61640063e-01
-2.96329670e-02 7.60187864e-01 -6.31563723e-01 2.80807137e-01
2.76551455e-01 4.57101703e-01 -4.89242114e-02 -6.63328946e-01
5.19243062e-01 1.05439734e+00 6.60476208e-01 -9.06229675e-01
5.90077102e-01 1.29065132e-02 -5.50252855e-01 -6.86984420e-01
-1.02765024e+00 -1.30051240e-01 -8.17862034e-01 4.92809474e-01
7.78310120e-01 -8.18564773e-01 -2.33352005e-01 5.76423764e-01
-1.85255635e+00 8.15515369e-02 1.26799345e-01 1.44143522e-01
-2.70378500e-01 6.16851568e-01 -1.22122490e+00 4.09848541e-02
-7.50122786e-01 -1.00165796e+00 7.35590637e-01 1.41487524e-01
-4.81943369e-01 -1.01595616e+00 2.70532817e-01 6.09315038e-01
4.73983645e-01 -4.31926623e-02 1.65191936e+00 -7.52974868e-01
-1.55513659e-01 1.88067093e-01 -4.84193742e-01 6.62314177e-01
5.05718589e-01 -2.80211002e-01 -6.32810473e-01 -5.04917949e-02
4.49771911e-01 -3.38832170e-01 3.14153671e-01 -3.67525220e-01
7.80904114e-01 -8.71917605e-01 1.61865741e-01 5.71696401e-01
1.13474417e+00 4.37993377e-01 7.43168592e-01 3.67192417e-01
5.62270999e-01 7.53515720e-01 5.07323205e-01 -5.53799719e-02
7.37096131e-01 8.51908803e-01 -2.18768328e-01 2.14892745e-01
-1.03875302e-01 -6.39323175e-01 6.65592670e-01 1.93942356e+00
2.45285362e-01 2.45707884e-01 -9.48588729e-01 7.04913020e-01
-1.78709829e+00 -9.52757776e-01 -2.14646548e-01 1.72001004e+00
1.48532569e+00 -5.88440523e-02 -2.66309172e-01 -5.36089242e-02
5.69669366e-01 4.90893237e-02 1.18353479e-01 -8.61220956e-01
-4.25128430e-01 6.02968156e-01 -4.05076921e-01 4.25544977e-01
-8.57057154e-01 1.49574578e+00 4.85908556e+00 8.57792020e-01
-1.25906694e+00 4.74257380e-01 1.31825954e-01 4.23801273e-01
-4.74615842e-01 2.96204180e-01 -4.48307276e-01 9.07140851e-01
4.63098705e-01 -3.99094790e-01 6.98849499e-01 8.37499976e-01
4.68788389e-03 2.37725750e-01 -1.14084578e+00 7.97619462e-01
5.28749108e-01 -1.31921721e+00 3.31559598e-01 -5.98634362e-01
1.02003515e+00 -1.36805564e-01 -1.38624370e-01 4.15232331e-01
-1.33074939e-01 -9.39259827e-01 5.88762820e-01 -2.37372339e-01
8.21466625e-01 -8.25334966e-01 1.00628662e+00 6.00477815e-01
-1.07551730e+00 1.23262018e-01 -6.74250543e-01 -6.62917495e-01
1.74992427e-01 1.01650879e-01 -4.93564457e-01 7.38363564e-01
3.92085612e-01 1.09787095e+00 -6.47211492e-01 3.77551734e-01
-9.89491999e-01 3.58423471e-01 5.19005693e-02 -3.31034362e-01
6.29797801e-02 -5.28065443e-01 3.35127801e-01 1.36490726e+00
1.99969694e-01 1.89490631e-01 -2.81160437e-02 1.17965472e+00
5.04625849e-02 5.55234969e-01 -6.98664904e-01 1.17981108e-02
7.08490133e-01 1.01961184e+00 -4.97350514e-01 -3.08410436e-01
-5.36692917e-01 1.41061294e+00 5.59951484e-01 -1.17089683e-02
-7.78628469e-01 -6.53828025e-01 6.17726862e-01 -3.53537351e-01
3.23307910e-03 2.23193672e-02 -5.91956079e-01 -1.66227818e+00
3.69679660e-01 -1.21968746e+00 1.05134934e-01 -7.30599761e-01
-1.57194746e+00 7.43343353e-01 3.05753723e-02 -1.29386973e+00
-3.47128063e-01 -6.39863610e-01 -1.33224940e+00 8.10666859e-01
-1.21573412e+00 -1.69458973e+00 3.27550317e-03 3.49190205e-01
8.90243530e-01 -8.55955034e-02 1.13980997e+00 3.34712297e-01
-5.99451125e-01 7.51343787e-01 -8.64462182e-02 4.44975495e-01
5.02023101e-01 -9.69414473e-01 5.35808444e-01 1.05375850e+00
2.45198712e-01 1.08848155e+00 2.57530987e-01 -4.74337876e-01
-9.29274321e-01 -8.62442672e-01 1.66923571e+00 -5.26607871e-01
8.16608787e-01 -5.88129640e-01 -9.44431782e-01 6.39239669e-01
6.45563722e-01 -7.01272190e-01 7.26059675e-01 1.27591535e-01
-4.49038774e-01 1.22335486e-01 -8.28731298e-01 9.20813918e-01
1.23896194e+00 -7.49178350e-01 -1.42745876e+00 4.11418378e-01
1.06454778e+00 -5.21963120e-01 -5.38471162e-01 4.82186973e-01
4.40029442e-01 -8.02897573e-01 7.25318670e-01 -9.06134427e-01
1.28158498e+00 -1.70583591e-01 -2.88900677e-02 -1.40447438e+00
-5.23514673e-02 -5.24155676e-01 5.58462262e-01 1.62580013e+00
4.32453096e-01 -5.26499212e-01 4.67337281e-01 1.02966078e-01
-3.03031921e-01 -5.02680480e-01 -9.85436499e-01 -8.50206673e-01
6.14023507e-01 -1.23122200e-01 7.59689510e-01 1.46065867e+00
8.47155154e-01 9.63361919e-01 -2.09057972e-01 -3.01635176e-01
-7.25962669e-02 6.74360573e-01 6.07626379e-01 -6.33894145e-01
-3.03358018e-01 -5.18428743e-01 -1.06969357e-01 -1.07266486e+00
8.29703391e-01 -1.27608836e+00 -1.93715543e-01 -7.92297423e-01
4.30743545e-01 -4.13407236e-01 1.71442658e-01 2.96046942e-01
-1.97683826e-01 2.05804676e-01 3.57342869e-01 3.60478759e-01
-2.94709295e-01 7.59132743e-01 1.22236645e+00 3.48031819e-02
-2.44391322e-01 -1.29043788e-01 -7.27758944e-01 7.27497220e-01
9.93884921e-01 -4.41399127e-01 -4.55396324e-01 -9.06189203e-01
2.23465949e-01 -8.70501176e-02 -7.86981136e-02 -3.19198102e-01
4.15633991e-02 -4.45232391e-01 -7.23379701e-02 -5.80957115e-01
-1.03890598e-02 -4.69199210e-01 1.06804989e-01 4.98427123e-01
-3.10706884e-01 8.13196361e-01 6.21374696e-02 -5.12063205e-01
-6.10431552e-01 -5.85505247e-01 7.06071079e-01 -3.08962673e-01
-4.87613022e-01 -4.91844684e-01 -1.37290120e-01 5.97091496e-01
9.23920393e-01 -6.48510829e-02 -6.95374906e-01 -2.19173416e-01
6.96277022e-02 1.22553907e-01 6.99228823e-01 7.88411558e-01
8.53395581e-01 -1.55287254e+00 -7.12938368e-01 4.83220249e-01
2.64353901e-01 1.70745403e-02 9.60448608e-02 8.74896884e-01
-9.60943639e-01 3.48279029e-01 -5.58364153e-01 -5.43980626e-03
-1.24809539e+00 6.01516724e-01 -3.04068830e-02 -2.67669350e-01
-6.88274264e-01 6.96104348e-01 3.31627011e-01 -1.04308903e+00
-3.37526470e-01 -1.67836234e-01 -1.74583465e-01 -2.66093016e-01
6.71143651e-01 1.36379162e-02 -1.38837054e-01 -7.66640067e-01
-4.59853977e-01 5.16664088e-01 -5.41732073e-01 -5.17148115e-02
1.05661404e+00 -1.99276999e-01 -1.09689212e+00 2.07806185e-01
1.21390140e+00 6.09821200e-01 -3.20480503e-02 -2.14514911e-01
3.33586186e-01 -5.67676544e-01 -9.23600256e-01 -8.14448416e-01
-5.55323422e-01 1.01956820e+00 -4.52848434e-01 -2.02827528e-01
1.18818641e+00 8.28266665e-02 1.23579764e+00 3.80378902e-01
2.69166142e-01 -1.03761470e+00 1.32569224e-01 1.07766879e+00
1.10008609e+00 -9.83642042e-01 -4.25849259e-01 -7.40094602e-01
-9.94001925e-01 1.06547105e+00 8.83694708e-01 -2.57806659e-01
3.79045427e-01 1.38203129e-01 4.70770687e-01 2.38666788e-01
-8.02054882e-01 3.74561876e-01 2.09256694e-01 6.21118009e-01
6.98798597e-01 2.70803601e-01 -1.09353817e+00 1.25642908e+00
-1.01913989e+00 -5.11154890e-01 6.12594843e-01 5.29511631e-01
9.71570313e-02 -1.55658948e+00 8.94454196e-02 -6.43549860e-02
-4.60620612e-01 -6.60748124e-01 -8.43312085e-01 7.44019926e-01
2.78055936e-01 7.44039297e-01 3.41879427e-02 -4.75348383e-01
1.22095689e-01 -1.20413922e-01 3.37936223e-01 -9.78700936e-01
-6.36019111e-01 -4.68577921e-01 2.46748149e-01 -1.73461944e-01
-3.08275223e-01 -5.23974001e-01 -1.17958999e+00 -4.66404647e-01
-1.89535469e-01 3.08507383e-01 2.16176182e-01 1.21545517e+00
5.22744544e-02 -9.94574949e-02 7.95172036e-01 -3.59094888e-01
-5.85127652e-01 -1.12447155e+00 -2.39013493e-01 8.41953814e-01
-3.68459642e-01 -1.33301601e-01 1.77999586e-02 4.76513710e-03] | [11.038776397705078, 9.282607078552246] |
6b2877a0-5c7b-4984-b5e4-af6e32110eff | i-can-t-believe-there-s-no-images-learning | 2211.09778 | null | https://arxiv.org/abs/2211.09778v3 | https://arxiv.org/pdf/2211.09778v3.pdf | I Can't Believe There's No Images! Learning Visual Tasks Using only Language Data | Many high-level skills that are required for computer vision tasks, such as parsing questions, comparing and contrasting semantics, and writing descriptions, are also required in other domains such as natural language processing. In this paper, we ask whether it is possible to learn those skills from textual data and then transfer them to vision tasks without ever training on visual training data. Key to our approach is exploiting the joint embedding space of contrastively trained vision and language encoders. In practice, there can be systematic differences between embedding spaces for different modalities in contrastive models, and we analyze how these differences affect our approach and study strategies to mitigate this concern. We produce models using only text training data on four representative tasks: image captioning, visual entailment, visual question answering and visual news, and evaluate them on standard benchmarks using images. We find these models generally perform close to models trained on images, while surpassing prior work for captioning and visual entailment in this text only setting by over 9 points, and outperforming all prior work on visual news by over 30 points. We also showcase a variety of stylistic image captioning models that are trained using no image data and no human-curated language data, but instead using readily-available text data from books, the web, or language models. | ['Aniruddha Kembhavi', 'Christopher Clark', 'Sophia Gu'] | 2022-11-17 | null | null | null | null | ['visual-entailment'] | ['reasoning'] | [ 5.09654999e-01 3.91921461e-01 1.68486256e-02 -4.06971723e-01
-7.39900827e-01 -8.98945451e-01 1.10958445e+00 1.25246376e-01
-7.72467017e-01 4.43041265e-01 4.43017632e-01 -6.56693697e-01
3.28101486e-01 -4.86269325e-01 -1.09073865e+00 -1.32736892e-01
4.39552069e-01 4.23753560e-01 1.41789809e-01 -1.74675167e-01
3.01349431e-01 2.30841696e-01 -1.44702446e+00 5.93800664e-01
4.86186743e-01 8.62223625e-01 2.64149517e-01 8.99528086e-01
-3.90278041e-01 1.35701251e+00 -4.88794833e-01 -8.60724628e-01
1.06430557e-02 -4.86288011e-01 -1.18154716e+00 3.51448894e-01
1.27445149e+00 -4.67259616e-01 -4.58901048e-01 9.63559270e-01
2.35451251e-01 -1.01415306e-01 9.00005877e-01 -1.21412277e+00
-1.46127331e+00 4.74564135e-01 -5.81279635e-01 3.17281216e-01
4.59368140e-01 4.64812666e-01 1.24135673e+00 -1.00618827e+00
7.90739834e-01 1.46997726e+00 4.38570142e-01 8.14925075e-01
-1.39470518e+00 -4.27750349e-02 1.61494613e-01 2.42058337e-01
-8.01683307e-01 -6.06807232e-01 5.25493503e-01 -7.05971122e-01
1.00749433e+00 1.97102442e-01 2.87611663e-01 1.34269392e+00
-1.18027404e-01 1.01103210e+00 1.22593915e+00 -5.11793733e-01
-9.96037349e-02 4.30699110e-01 3.86867598e-02 8.55372965e-01
1.14037357e-01 8.81713107e-02 -5.47605038e-01 2.22530693e-01
8.35538387e-01 -3.14492434e-01 -3.72251719e-01 -3.87740463e-01
-1.61154151e+00 8.90844822e-01 6.05535924e-01 5.06496876e-02
-1.16272755e-01 4.15057272e-01 4.96552199e-01 3.28475714e-01
3.37828010e-01 6.68809950e-01 -2.83729702e-01 8.94047394e-02
-8.58833253e-01 1.69870496e-01 7.75982141e-01 1.03411937e+00
6.06854260e-01 6.25651181e-02 -2.97918141e-01 6.49205565e-01
3.29174310e-01 6.28120363e-01 3.68046939e-01 -1.12005758e+00
6.26300752e-01 2.56228328e-01 1.27625942e-01 -9.41348791e-01
-1.77610159e-01 2.30860729e-02 -4.01442200e-01 2.42352739e-01
8.21892202e-01 3.07269003e-02 -1.19004774e+00 1.77823782e+00
-1.15890466e-01 -1.50767773e-01 2.43426427e-01 8.21879506e-01
1.26992869e+00 7.87218153e-01 3.86061579e-01 1.60565972e-01
1.78890848e+00 -1.13553691e+00 -5.17384589e-01 -8.42929542e-01
5.07297575e-01 -9.47565079e-01 1.71823347e+00 1.16357632e-01
-1.46010387e+00 -5.47142684e-01 -9.33499694e-01 -7.31040835e-01
-5.33993661e-01 9.85358879e-02 4.83301163e-01 2.42569879e-01
-1.31670141e+00 2.47799784e-01 -4.30370420e-01 -6.90703511e-01
4.69117463e-01 -1.67327046e-01 -4.87767130e-01 -2.70254642e-01
-9.96291816e-01 1.41163182e+00 1.18665457e-01 -1.50643110e-01
-9.10162210e-01 -6.06992900e-01 -1.28597033e+00 -6.42876104e-02
2.95880437e-01 -9.12788987e-01 1.59149301e+00 -1.34874547e+00
-1.01556671e+00 1.58544743e+00 -7.92213306e-02 -4.83342767e-01
5.03849804e-01 -7.48215169e-02 -3.88655178e-02 3.54192764e-01
7.05883652e-02 1.24775326e+00 9.44184065e-01 -1.41991639e+00
-3.02898109e-01 -1.38030067e-01 5.78163147e-01 2.92587399e-01
-2.80114830e-01 2.82891225e-02 -5.82859218e-01 -4.75326926e-01
-3.96895915e-01 -7.83194900e-01 8.59182328e-02 5.90258121e-01
-3.30466986e-01 -1.39695644e-01 5.59947431e-01 -8.37091148e-01
4.68242645e-01 -2.11259270e+00 1.70440212e-01 -3.00069153e-01
2.71879345e-01 2.77127586e-02 -5.24980843e-01 2.85220951e-01
4.00819583e-03 1.63139462e-01 -3.29810500e-01 -4.22836572e-01
2.14952126e-01 3.47233713e-01 -5.62046826e-01 3.41636389e-01
6.04173422e-01 1.34329283e+00 -9.34668422e-01 -8.07557940e-01
4.04379278e-01 3.89565557e-01 -5.36634266e-01 2.52577603e-01
-5.47716856e-01 9.21602100e-02 4.42383103e-02 4.74870086e-01
3.39725912e-01 -5.86678743e-01 -7.27482140e-02 -4.40160871e-01
6.39389157e-02 2.00038061e-01 -6.35723710e-01 1.65452957e+00
-6.67457283e-01 1.33158600e+00 2.81954389e-02 -1.12963140e+00
4.59293783e-01 1.26507819e-01 -1.02107152e-01 -1.05801976e+00
-4.35321219e-02 -9.55716148e-02 -1.52576327e-01 -8.50770116e-01
4.20744300e-01 -1.90113276e-01 1.00382805e-01 5.40404499e-01
2.34635860e-01 -6.12885177e-01 3.38151813e-01 4.73797172e-01
8.25961053e-01 1.34345919e-01 1.58452436e-01 -1.04079358e-01
2.60925055e-01 3.46118808e-01 -3.38511378e-01 8.61781001e-01
-2.75980055e-01 7.63447821e-01 6.51724279e-01 -3.18589568e-01
-1.48781765e+00 -1.24699187e+00 -1.09448254e-01 1.34680951e+00
1.64028779e-01 -2.96066701e-01 -7.19016671e-01 -6.40422463e-01
-3.20256874e-03 9.43175554e-01 -8.78275216e-01 3.80883664e-02
-2.91883528e-01 -3.00329417e-01 5.62960804e-01 6.86776161e-01
3.52771342e-01 -1.40155172e+00 -7.34475434e-01 -2.84132302e-01
-1.24424167e-01 -1.59732866e+00 -5.33702135e-01 9.29078087e-02
-5.14249384e-01 -1.01244152e+00 -9.48820651e-01 -1.05933261e+00
7.01047242e-01 1.94826081e-01 1.82513630e+00 1.74771622e-01
-2.94605494e-01 1.05776215e+00 -1.76617905e-01 -6.83704615e-01
-5.77615857e-01 -2.15163618e-01 -5.04815400e-01 -2.75268108e-01
3.70799124e-01 -7.46998563e-02 -5.86024702e-01 3.91886197e-03
-1.23828745e+00 4.27126199e-01 8.42644513e-01 8.90676141e-01
3.33827436e-01 -8.47028911e-01 1.93431675e-01 -8.08194697e-01
6.66166067e-01 -3.79358351e-01 -3.93046796e-01 4.83774811e-01
-3.99494410e-01 3.45269054e-01 4.44155604e-01 -4.88184094e-01
-7.70209193e-01 -5.53555228e-02 -4.78553325e-02 -4.09871966e-01
-2.61290938e-01 4.05776054e-01 1.90157115e-01 7.80531913e-02
8.50199044e-01 3.01377267e-01 1.89127475e-01 -1.16994299e-01
9.62017059e-01 5.47165096e-01 8.14293802e-01 -5.47881007e-01
8.53044271e-01 6.35535359e-01 -2.25328192e-01 -8.77701461e-01
-1.11902153e+00 -1.17586039e-01 -3.59488577e-01 -6.33087084e-02
1.18311584e+00 -8.71411979e-01 -6.06957614e-01 4.69577275e-02
-1.52821779e+00 -4.06347781e-01 -4.11255091e-01 2.73333251e-01
-8.54469776e-01 4.57840323e-01 -7.03253329e-01 -4.44799304e-01
-1.80655867e-01 -1.13863981e+00 1.22520626e+00 4.16252315e-02
-2.31469557e-01 -1.19912112e+00 -1.39395759e-01 5.63120067e-01
4.76190448e-01 1.54118970e-01 1.07928455e+00 -5.64196408e-01
-5.22052348e-01 2.05193102e-01 -7.31204629e-01 5.84431052e-01
-1.05669975e-01 -1.80387888e-02 -1.07747400e+00 -5.83980307e-02
-2.33903915e-01 -1.05939233e+00 1.04385638e+00 2.70331860e-01
1.22759032e+00 -2.99611688e-01 1.40370382e-02 4.37758386e-01
1.31656063e+00 -3.21205854e-01 6.84717357e-01 3.60495001e-01
8.34197760e-01 1.01139879e+00 1.67066529e-01 -1.61038145e-01
6.89438701e-01 4.83076513e-01 6.13888860e-01 -4.45034504e-01
-3.32595080e-01 -2.90097713e-01 4.49365765e-01 4.00759190e-01
2.28556409e-01 -3.14760536e-01 -1.01482737e+00 8.31529379e-01
-1.68699181e+00 -1.08700860e+00 8.37389156e-02 1.75029075e+00
1.12654877e+00 6.64717555e-02 -1.80119071e-02 -2.78074741e-01
4.16579276e-01 2.69114703e-01 -5.62021673e-01 -7.37414777e-01
-5.12410067e-02 1.56571075e-01 5.06515741e-01 5.66252172e-01
-1.09956181e+00 9.63874280e-01 7.13789368e+00 3.40519816e-01
-1.12228811e+00 5.71119413e-02 8.40988100e-01 -5.17598838e-02
-6.35164976e-01 -6.44440502e-02 -2.00920328e-01 1.17345206e-01
8.17153215e-01 5.45635410e-02 4.61148590e-01 7.03825533e-01
-7.06430450e-02 -2.12458875e-02 -1.65598488e+00 1.25670934e+00
6.24968767e-01 -1.43441284e+00 3.17739308e-01 -2.91135877e-01
5.78092635e-01 7.64294714e-02 4.52348739e-01 2.50072539e-01
2.97860771e-01 -1.57442594e+00 9.76466238e-01 4.48234230e-01
8.59788477e-01 -2.25266829e-01 4.13235486e-01 7.70934373e-02
-6.46405995e-01 2.68324256e-01 -2.71959335e-01 -9.49426740e-02
5.25044613e-02 1.44100517e-01 -7.48735964e-01 9.11137760e-02
6.76239848e-01 7.32770622e-01 -9.26326334e-01 5.84465802e-01
-2.50717491e-01 4.71977741e-01 -3.26084048e-02 -1.90828696e-01
4.79694784e-01 1.68383747e-01 2.05007002e-01 1.28618467e+00
-2.14065760e-02 -3.08671862e-01 -1.57524258e-01 1.07769930e+00
-2.28927359e-01 2.08536647e-02 -8.81111860e-01 -3.62612039e-01
8.19880590e-02 1.11153412e+00 -5.51353633e-01 -4.79502022e-01
-1.12039649e+00 1.10905707e+00 3.75922382e-01 6.42920554e-01
-8.83047163e-01 -3.01581621e-01 4.17012155e-01 1.72049135e-01
3.01516533e-01 -3.40373516e-01 -2.85182536e-01 -1.29189241e+00
-7.68392533e-03 -1.10791564e+00 2.48872653e-01 -1.40453529e+00
-1.61050081e+00 5.43748379e-01 2.81747747e-02 -7.96049416e-01
-4.78748202e-01 -1.20475101e+00 -4.76048917e-01 8.36888015e-01
-1.69071639e+00 -1.30179644e+00 -4.56275970e-01 7.27204144e-01
7.41440475e-01 5.28080203e-02 5.37563324e-01 -1.06397882e-01
-8.30577239e-02 4.81605798e-01 -1.94100842e-01 3.04711193e-01
8.82864416e-01 -1.52408397e+00 6.39643431e-01 6.87180221e-01
7.39625394e-01 4.84008968e-01 1.01711357e+00 -8.18288401e-02
-1.40952599e+00 -7.99063921e-01 8.28729153e-01 -1.09470308e+00
8.56102884e-01 -4.65247452e-01 -8.35914552e-01 9.76219714e-01
8.95812869e-01 6.73390329e-02 4.37003911e-01 -3.96634564e-02
-7.54203856e-01 3.14219654e-01 -8.45541775e-01 9.36070025e-01
9.26056027e-01 -9.59367752e-01 -9.99503732e-01 5.87892115e-01
8.58518720e-01 -3.65572006e-01 -4.71354038e-01 9.28894803e-02
4.14142221e-01 -7.28567302e-01 1.20507562e+00 -9.70977426e-01
1.04534054e+00 -9.40903276e-02 -2.21241817e-01 -1.16939199e+00
7.97578879e-03 -1.29419953e-01 5.82302473e-02 1.02120876e+00
6.57772541e-01 -3.23620796e-01 5.32937467e-01 6.58746302e-01
6.06101751e-02 -5.03309250e-01 -5.66204965e-01 -5.36253333e-01
4.97184038e-01 -4.00615335e-01 1.68368146e-02 9.46313083e-01
-2.21556410e-01 7.43052065e-01 -3.32049459e-01 -2.21787140e-01
5.44788599e-01 5.47228754e-02 8.30842257e-01 -7.81085134e-01
-4.02952611e-01 -5.79203069e-01 -3.46724659e-01 -1.19920349e+00
3.41016769e-01 -9.70654428e-01 8.71719643e-02 -1.99374402e+00
4.50106949e-01 2.54882574e-01 7.07628354e-02 5.31836569e-01
-1.24453239e-01 4.81474102e-01 3.94469857e-01 7.45626986e-02
-6.71064377e-01 2.86655754e-01 1.49735928e+00 -4.60742980e-01
4.61040765e-01 -6.86981320e-01 -1.01340747e+00 8.21578622e-01
4.36633706e-01 -2.18692292e-02 -6.79492593e-01 -1.03764224e+00
4.62399274e-01 -2.86410779e-01 8.55407417e-01 -5.53119779e-01
7.10767284e-02 -3.03278774e-01 5.68104267e-01 -1.87965825e-01
3.31050187e-01 -6.97479784e-01 -6.23894036e-01 1.92750514e-01
-7.84537554e-01 3.52941275e-01 4.21485782e-01 4.14228588e-01
-2.65090257e-01 -3.33985120e-01 7.09507108e-01 -3.31451476e-01
-1.02109861e+00 6.37480766e-02 -3.14556807e-01 5.42400956e-01
9.21819508e-01 -1.20025292e-01 -7.64513195e-01 -7.75102794e-01
-6.86119497e-01 2.77359635e-01 6.63571894e-01 5.22550881e-01
7.24762261e-01 -1.20028210e+00 -8.57412755e-01 -1.89641312e-01
5.60982764e-01 -1.42666683e-01 3.60197248e-03 6.08013391e-01
-7.19663382e-01 4.38156664e-01 -3.12504172e-01 -7.37216115e-01
-1.11846566e+00 7.75371373e-01 3.17980230e-01 1.40263557e-01
-4.68466908e-01 7.99954116e-01 6.48101211e-01 -2.85003901e-01
2.88576961e-01 -5.16507626e-01 -2.08721265e-01 -4.07009944e-02
5.99184513e-01 -3.32631558e-01 -2.90235937e-01 -6.72119617e-01
-3.36280465e-01 6.95186138e-01 -5.64679131e-02 -3.85715723e-01
9.87485051e-01 -2.37487793e-01 9.37636942e-02 5.25675714e-01
1.21082270e+00 -1.27139032e-01 -1.29460323e+00 -2.66385585e-01
3.15988399e-02 -2.38996848e-01 -1.48879454e-01 -7.90270984e-01
-7.99612463e-01 1.27816224e+00 3.54749769e-01 3.56625795e-01
9.58164871e-01 5.87697923e-01 5.16594052e-01 4.96884227e-01
-2.24884480e-01 -9.24504220e-01 6.44508362e-01 5.12311757e-01
1.10597348e+00 -1.63402545e+00 -1.15899362e-01 -3.24940383e-02
-1.03265929e+00 9.52241600e-01 6.87046707e-01 -1.06410548e-01
1.41341537e-01 8.09110850e-02 2.25878090e-01 -3.21433872e-01
-9.09877956e-01 -5.02620220e-01 5.86931646e-01 7.94861555e-01
7.05833137e-01 -3.17595184e-01 1.01561710e-01 1.63165331e-01
-1.28664628e-01 -3.24214160e-01 5.79368711e-01 8.37538719e-01
-3.34115356e-01 -7.29647279e-01 -3.14123482e-01 4.45054293e-01
-4.17604834e-01 -5.72185934e-01 -7.38713086e-01 9.10437524e-01
-1.39920071e-01 7.98609018e-01 3.95421773e-01 1.62850935e-02
2.67226189e-01 5.36636636e-02 8.85169744e-01 -6.09197259e-01
-1.39893115e-01 -4.72915769e-01 2.09887460e-01 -4.82793927e-01
-5.32809258e-01 -4.29458916e-01 -1.00550067e+00 -2.33871281e-01
1.94483772e-01 -2.29544327e-01 7.63476431e-01 9.99228477e-01
8.19102749e-02 4.05418932e-01 -2.00922955e-02 -7.31713593e-01
-5.73230445e-01 -8.33819985e-01 -8.50626454e-02 9.08296168e-01
5.46101451e-01 -3.74501884e-01 -4.92520392e-01 5.55449843e-01] | [10.834157943725586, 1.7077449560165405] |
0ec043df-99d8-4840-bf20-07a5dfb7c8ab | perceive-interact-predict-learning-dynamic | 2212.02181 | null | https://arxiv.org/abs/2212.02181v1 | https://arxiv.org/pdf/2212.02181v1.pdf | Perceive, Interact, Predict: Learning Dynamic and Static Clues for End-to-End Motion Prediction | Motion prediction is highly relevant to the perception of dynamic objects and static map elements in the scenarios of autonomous driving. In this work, we propose PIP, the first end-to-end Transformer-based framework which jointly and interactively performs online mapping, object detection and motion prediction. PIP leverages map queries, agent queries and mode queries to encode the instance-wise information of map elements, agents and motion intentions, respectively. Based on the unified query representation, a differentiable multi-task interaction scheme is proposed to exploit the correlation between perception and prediction. Even without human-annotated HD map or agent's historical tracking trajectory as guidance information, PIP realizes end-to-end multi-agent motion prediction and achieves better performance than tracking-based and HD-map-based methods. PIP provides comprehensive high-level information of the driving scene (vectorized static map and dynamic objects with motion information), and contributes to the downstream planning and control. Code and models will be released for facilitating further research. | ['Chang Huang', 'Wenyu Liu', 'Qian Zhang', 'Helong Zhou', 'Jiajie Chen', 'Tianheng Cheng', 'Bencheng Liao', 'Xinggang Wang', 'Shaoyu Chen', 'Bo Jiang'] | 2022-12-05 | null | null | null | null | ['motion-prediction'] | ['computer-vision'] | [-1.59387484e-01 2.07460150e-01 -2.96228141e-01 -5.70325911e-01
-6.86966538e-01 -4.91850674e-01 9.53153133e-01 2.60623395e-01
-3.91066521e-01 3.57909262e-01 2.98612326e-01 -5.14627285e-02
-1.43579900e-01 -9.93831694e-01 -6.14403963e-01 -3.31559807e-01
-3.93688560e-01 8.07841718e-01 1.11021602e+00 -4.38994527e-01
3.29565108e-01 4.27609324e-01 -2.12262893e+00 1.54200807e-01
6.36758745e-01 9.72529829e-01 1.17162204e+00 1.04424417e+00
1.28009409e-01 1.21259737e+00 4.55324315e-02 1.73281990e-02
1.58333346e-01 1.94006518e-01 -5.80219686e-01 -2.53336251e-01
1.84606612e-01 -6.77982807e-01 -7.17098057e-01 6.03175282e-01
5.69258213e-01 4.81957048e-01 3.04664224e-01 -1.64094341e+00
-9.98373628e-02 1.65445670e-01 -1.06173135e-01 2.59382606e-01
4.51146215e-01 6.36620045e-01 8.15374792e-01 -8.60415816e-01
7.66800463e-01 1.21436441e+00 2.61465132e-01 1.19471900e-01
-6.15394831e-01 -4.04619902e-01 3.27206701e-01 9.21025753e-01
-1.37807572e+00 -5.37390351e-01 7.21581578e-01 -6.00251794e-01
1.16568613e+00 3.80201280e-01 7.15255141e-01 6.43713295e-01
3.59050274e-01 1.01361096e+00 3.78101379e-01 1.91136777e-01
1.22066312e-01 2.77396608e-02 -1.32316023e-01 7.24389315e-01
-5.22886217e-01 6.04907393e-01 -8.69207978e-01 1.00121416e-01
4.86611664e-01 -1.10915631e-01 1.02597974e-01 -7.09842801e-01
-1.74003851e+00 4.85491186e-01 4.32436764e-01 -2.73966581e-01
-8.02072823e-01 3.56083155e-01 2.08411038e-01 4.46700118e-03
1.53829470e-01 -1.08219258e-01 -2.55786508e-01 -5.62365770e-01
-5.49322248e-01 5.35486519e-01 4.33804601e-01 1.43549454e+00
1.23112559e+00 -1.39856473e-01 -4.95648801e-01 3.63474011e-01
3.64410102e-01 9.03608799e-01 7.98452571e-02 -1.42974126e+00
5.13238549e-01 5.10223389e-01 6.75971866e-01 -1.20111883e+00
-6.80382252e-01 1.64742637e-02 -2.44354591e-01 2.02959716e-01
-9.93801206e-02 1.17960148e-01 -6.51473045e-01 1.64044607e+00
7.74084389e-01 4.73811835e-01 2.85734594e-01 1.06017196e+00
7.52636254e-01 7.88525879e-01 1.33081540e-01 5.45220077e-02
1.35241818e+00 -1.26663208e+00 -8.10090542e-01 -5.78391135e-01
5.88227212e-01 -5.05553126e-01 8.29246581e-01 -2.19589978e-01
-1.04369676e+00 -9.82108712e-01 -8.18867326e-01 -2.11024761e-01
-4.03526872e-01 9.36707184e-02 3.20570618e-01 -5.05608739e-03
-1.14713848e+00 7.70019442e-02 -1.03736150e+00 -4.15145338e-01
1.50883451e-01 3.59662950e-01 -1.65566560e-02 -1.58114374e-01
-1.09653366e+00 1.31190455e+00 5.71589649e-01 2.23183203e-02
-1.31699800e+00 -7.33206391e-01 -1.10334241e+00 -2.49282688e-01
4.28870738e-01 -8.16635668e-01 1.47048998e+00 -1.98052321e-02
-1.53696358e+00 4.37838435e-01 -5.64266086e-01 -5.35889089e-01
6.77979171e-01 -2.88917989e-01 -4.43333149e-01 1.29913734e-02
4.32253748e-01 1.20734513e+00 3.99034083e-01 -1.20320976e+00
-1.54762340e+00 -2.60925025e-01 9.43354890e-02 6.85320258e-01
2.04652235e-01 -3.75886351e-01 -9.80988443e-01 8.17876905e-02
9.10746753e-02 -8.25594008e-01 -4.66832340e-01 2.00514235e-02
-2.54471451e-01 -3.61417860e-01 1.16887951e+00 -4.33193415e-01
1.06103563e+00 -2.16906047e+00 3.32195871e-02 -1.17476070e-02
1.94666803e-01 -2.46748939e-01 -1.55812100e-01 6.91787362e-01
7.57427037e-01 -7.36176312e-01 2.32204527e-01 -4.68638629e-01
4.77502227e-01 3.30742061e-01 -3.72388601e-01 3.63236517e-01
-1.47954270e-01 1.14097440e+00 -1.19230568e+00 -5.51071703e-01
8.62728179e-01 3.45033079e-01 -3.15475434e-01 4.12852764e-01
-5.27182758e-01 7.25482762e-01 -8.18745553e-01 4.88323838e-01
6.11545384e-01 2.28901003e-02 -7.98783526e-02 -2.55464673e-01
-7.32600629e-01 3.07283908e-01 -1.06063187e+00 1.90987659e+00
-3.91397059e-01 8.17935884e-01 1.18347118e-02 -3.48083377e-01
8.66858125e-01 5.45787141e-02 5.95663488e-01 -1.22235823e+00
-4.00820494e-01 -8.22006688e-02 -4.99339163e-01 -6.66148961e-01
1.20897031e+00 5.75456798e-01 -3.88584107e-01 9.70407054e-02
-4.61236000e-01 -1.53512314e-01 1.17514491e-01 1.49268031e-01
1.22613192e+00 3.66329700e-01 1.04212143e-01 -3.66347507e-02
4.65186656e-01 7.33909428e-01 5.00024259e-01 8.83283556e-01
-4.69605327e-01 1.05473466e-01 -1.33332402e-01 -6.37376130e-01
-9.31668162e-01 -1.13783979e+00 1.23217411e-01 1.55114388e+00
1.10860705e+00 -4.60776746e-01 -2.04343811e-01 -1.93706661e-01
1.55084863e-01 1.10241008e+00 -3.70371938e-01 -1.20364957e-01
-8.59664679e-01 -2.37744108e-01 1.19947210e-01 5.78834653e-01
6.76353395e-01 -1.00489116e+00 -1.35952032e+00 5.30694664e-01
-4.98397827e-01 -1.28082347e+00 -6.37407780e-01 -1.27682611e-01
-2.05243751e-01 -8.27705562e-01 -1.21659651e-01 -6.45389140e-01
2.45619342e-01 5.84062636e-01 9.07317042e-01 -2.46208146e-01
8.06755796e-02 7.38945067e-01 -1.81179345e-01 -3.58574927e-01
-4.42960322e-01 -2.19640464e-01 -7.78697580e-02 -1.46232322e-01
2.22541854e-01 -3.83191764e-01 -8.78890514e-01 8.00924540e-01
-2.46136189e-01 6.98374391e-01 4.70643491e-01 2.24083096e-01
8.19587588e-01 -4.96068224e-02 2.48058468e-01 -1.28788710e-01
5.04068099e-02 -5.21523416e-01 -9.84028995e-01 -1.91211812e-02
-5.03638864e-01 -2.24958047e-01 1.69685513e-01 -2.47694641e-01
-1.03959239e+00 4.57946748e-01 -3.22573483e-02 -4.34069633e-01
-2.63860285e-01 4.24264550e-01 -1.69976786e-01 1.30009130e-01
3.50854218e-01 5.15303552e-01 -1.33696094e-01 -1.15588397e-01
5.54048359e-01 2.95659751e-01 1.15913057e+00 -1.25475630e-01
7.42279351e-01 7.61622488e-01 -1.60058424e-01 -4.61124808e-01
-3.93082857e-01 -8.36900175e-01 -6.77301884e-01 -7.46818185e-01
1.20611703e+00 -1.24801934e+00 -1.13958859e+00 3.37135732e-01
-1.31955338e+00 -7.95569658e-01 -9.79575515e-02 5.73171854e-01
-1.13641262e+00 -2.68748496e-02 -3.28705817e-01 -9.32926893e-01
1.43346444e-01 -1.35329640e+00 1.38571334e+00 7.34345913e-02
1.03604309e-01 -8.14897180e-01 1.64524645e-01 1.59936145e-01
3.93876255e-01 1.30983517e-01 3.81470472e-01 -2.66681463e-01
-1.51784003e+00 -1.21354997e-01 -1.80311859e-01 -7.84814000e-01
-1.57626599e-01 -4.46357846e-01 -8.89488757e-01 -1.74420923e-01
-4.41586286e-01 5.46498857e-02 7.05111623e-01 4.54218596e-01
7.01606929e-01 -1.20875686e-01 -8.98238599e-01 4.94626194e-01
1.34672356e+00 4.86384302e-01 5.86625099e-01 6.09078825e-01
7.95846879e-01 8.63466442e-01 1.41287827e+00 7.74521649e-01
1.53167844e+00 1.17297637e+00 9.73994374e-01 2.86493860e-02
-4.18285504e-02 -4.85393792e-01 4.99880522e-01 4.96755749e-01
1.75766140e-01 -3.77450794e-01 -1.04193175e+00 7.09437191e-01
-2.34602070e+00 -1.28251052e+00 -3.41475308e-01 1.98034418e+00
4.71601356e-03 1.83404192e-01 2.72463828e-01 -5.38790703e-01
7.14825094e-01 3.78396064e-01 -9.57494140e-01 2.26371571e-01
1.83915690e-01 -8.54624331e-01 7.65323877e-01 8.44034791e-01
-1.19239950e+00 1.31966400e+00 6.25062132e+00 6.29916787e-01
-6.91351593e-01 3.72997969e-01 5.53057576e-03 -8.92185606e-03
-3.20165724e-01 -5.66394674e-03 -1.09851456e+00 4.25993443e-01
9.46808398e-01 -2.78784245e-01 3.55056494e-01 8.69946361e-01
7.24802911e-01 -3.30246091e-01 -1.21129954e+00 8.93877029e-01
-4.18552876e-01 -1.80216932e+00 -4.38583463e-01 1.37410849e-01
2.41417229e-01 6.96927071e-01 1.85580086e-02 4.46005136e-01
7.35648453e-01 -5.43084741e-01 1.32597303e+00 7.03097641e-01
4.03662175e-01 -6.59452260e-01 3.82638991e-01 8.97881925e-01
-1.86503339e+00 -3.25409055e-01 -2.53686428e-01 -7.72758648e-02
7.50686407e-01 3.73356231e-02 -1.08540154e+00 6.62446558e-01
5.69510460e-01 9.35931325e-01 -3.72499198e-01 9.12690580e-01
2.58607715e-01 3.00136465e-03 -3.64528537e-01 5.69038019e-02
3.48499119e-01 3.18879485e-02 8.20996165e-01 1.11143064e+00
3.29891413e-01 1.12376019e-01 7.20394850e-01 8.54300976e-01
7.21455336e-01 -3.91205013e-01 -6.65099025e-01 5.11940241e-01
8.27763677e-01 1.07499480e+00 -4.31837410e-01 -5.70921898e-01
-4.61338282e-01 7.98491061e-01 2.55544007e-01 4.68746990e-01
-1.08757341e+00 1.13548242e-01 1.11866367e+00 1.20348983e-01
5.55552542e-01 -5.61743915e-01 1.34747744e-01 -4.79622364e-01
-3.53591368e-02 -3.55694443e-02 2.01567829e-01 -9.85878706e-01
-5.15927732e-01 5.49608707e-01 2.44324595e-01 -1.48058212e+00
-6.70923531e-01 -4.51886766e-02 -6.15062118e-01 6.74844682e-01
-1.78212678e+00 -1.38575757e+00 -7.78825760e-01 5.65902829e-01
7.11733937e-01 -9.08049941e-02 5.39852023e-01 1.92880601e-01
-3.46726328e-01 -1.91017352e-02 -8.94213691e-02 -3.66235703e-01
3.56350601e-01 -1.10596681e+00 6.64181352e-01 7.57432103e-01
-3.04260999e-01 -3.20234656e-01 9.87101495e-01 -8.16855669e-01
-1.78285348e+00 -1.47701466e+00 6.94798410e-01 -7.87491381e-01
4.92469341e-01 -2.76284993e-01 -5.63900530e-01 8.10116947e-01
7.09429011e-02 -5.19676358e-02 2.10954752e-02 -4.21415359e-01
3.69166523e-01 -1.57945752e-01 -7.95060158e-01 5.60575247e-01
1.29077947e+00 -4.38038558e-01 -1.77085429e-01 4.81462300e-01
1.04429638e+00 -9.44476724e-01 -7.44174957e-01 5.17987132e-01
5.20143688e-01 -1.12462485e+00 1.19810188e+00 -1.01205312e-01
-1.75949827e-01 -7.65493512e-01 -4.92166907e-01 -9.18638051e-01
-6.71918452e-01 -5.24046302e-01 -3.09814543e-01 7.09500909e-01
3.32438797e-01 -4.57584143e-01 8.90061796e-01 6.17794752e-01
-8.33696187e-01 -4.84069943e-01 -1.05584109e+00 -5.59735715e-01
-7.42866874e-01 -8.46178114e-01 8.36135805e-01 4.61864114e-01
-1.96771100e-01 6.90229163e-02 -5.41416049e-01 9.16744947e-01
4.72096115e-01 4.07323688e-01 1.35427356e+00 -8.14205527e-01
1.73673388e-02 -3.78248632e-01 -7.49339163e-01 -1.75226557e+00
1.97233379e-01 -6.01771891e-01 5.97386360e-01 -1.78170764e+00
-1.41700894e-01 -4.95904416e-01 -2.57944711e-03 2.78097332e-01
1.70760006e-02 -1.54017016e-01 2.24529132e-01 4.85161096e-01
-1.13839996e+00 7.77395964e-01 1.41184056e+00 -1.33771166e-01
-4.84614581e-01 3.27345133e-02 5.88503592e-02 4.46531475e-01
5.49161017e-01 -1.66990250e-01 -6.58127010e-01 -5.86482704e-01
-5.46969660e-02 6.58588707e-01 7.79914677e-01 -1.07775664e+00
8.93414855e-01 -5.52520216e-01 1.01626264e-02 -1.63182259e+00
1.02639997e+00 -8.30114186e-01 3.94890845e-01 4.15836513e-01
-2.68607706e-01 3.96991223e-01 2.25209102e-01 8.94834757e-01
-1.25091299e-01 3.80169868e-01 2.24221438e-01 -1.62373744e-02
-1.79690933e+00 7.73246825e-01 -5.96807361e-01 -3.47010523e-01
1.40731823e+00 -3.91389430e-01 -4.13724810e-01 -5.70353091e-01
-7.76817143e-01 9.63197291e-01 4.63860959e-01 5.95053256e-01
7.73353100e-01 -1.47031295e+00 -6.07343078e-01 1.92558348e-01
4.94162917e-01 2.33785436e-01 5.59715629e-01 1.04269361e+00
-3.20962459e-01 4.43022907e-01 -1.66427225e-01 -1.06713295e+00
-9.35762048e-01 6.41521513e-01 1.72517672e-01 4.67049927e-02
-9.93355751e-01 3.88814986e-01 5.35979033e-01 -4.55332309e-01
1.90383315e-01 -1.17592573e-01 -3.19946289e-01 -3.77508640e-01
7.87774026e-01 6.49491608e-01 -2.30289623e-01 -1.06072843e+00
-3.84150058e-01 4.15378213e-01 2.28697851e-01 -2.63233781e-01
1.05391932e+00 -9.07813013e-01 2.76275933e-01 4.46984828e-01
8.15258563e-01 -2.62177140e-01 -1.91218483e+00 -1.49258107e-01
3.92295094e-03 -4.96103734e-01 1.68244153e-01 -5.57371378e-01
-5.52473307e-01 5.44736803e-01 7.56325424e-01 5.63151203e-02
7.43257880e-01 4.23114806e-01 8.49512339e-01 3.98504555e-01
8.34579229e-01 -1.00963664e+00 -2.20814301e-03 6.38769746e-01
7.76534796e-01 -1.34779787e+00 -4.32759970e-01 -5.25931180e-01
-8.49640489e-01 6.84854865e-01 1.07406652e+00 4.06042755e-01
3.62103850e-01 2.79063255e-01 1.12369051e-02 -2.92055577e-01
-1.22922695e+00 -6.76327288e-01 2.72661120e-01 1.08369958e+00
-5.11881173e-01 1.11630708e-01 5.84943831e-01 1.21407777e-01
-2.65377373e-01 -1.68098167e-01 8.16623122e-02 8.91341627e-01
-9.24622536e-01 -6.43493354e-01 -2.66930550e-01 2.67931938e-01
6.62447989e-01 2.38365263e-01 2.63774514e-01 6.62077427e-01
1.80211470e-01 9.83211219e-01 5.21759212e-01 -6.54609263e-01
4.72642034e-01 -3.61268967e-01 9.36906934e-02 -2.73198545e-01
-9.56901312e-02 -8.96923542e-02 3.83353621e-01 -1.18156803e+00
-2.53332883e-01 -7.20994830e-01 -1.72781253e+00 -4.25111353e-01
1.05541252e-01 -2.28456438e-01 6.67880058e-01 9.85538661e-01
8.59396338e-01 4.33219373e-01 4.43457842e-01 -1.51704383e+00
-5.34011349e-02 -5.45256078e-01 -3.00457105e-02 1.06517360e-01
6.56767905e-01 -9.22791898e-01 2.67819852e-01 -1.15113705e-02] | [5.870194435119629, 0.7817896604537964] |
e7b7adfe-4bf3-4d93-bd9f-e17b35911136 | bayesian-variable-selection-in-a-million | 2208.01180 | null | https://arxiv.org/abs/2208.01180v2 | https://arxiv.org/pdf/2208.01180v2.pdf | Bayesian Variable Selection in a Million Dimensions | Bayesian variable selection is a powerful tool for data analysis, as it offers a principled method for variable selection that accounts for prior information and uncertainty. However, wider adoption of Bayesian variable selection has been hampered by computational challenges, especially in difficult regimes with a large number of covariates P or non-conjugate likelihoods. To scale to the large P regime we introduce an efficient MCMC scheme whose cost per iteration is sublinear in P. In addition we show how this scheme can be extended to generalized linear models for count data, which are prevalent in biology, ecology, economics, and beyond. In particular we design efficient algorithms for variable selection in binomial and negative binomial regression, which includes logistic regression as a special case. In experiments we demonstrate the effectiveness of our methods, including on cancer and maize genomic data. | ['Martin Jankowiak'] | 2022-08-02 | null | null | null | null | ['variable-selection'] | ['methodology'] | [ 5.57042480e-01 -3.13459814e-01 -2.68454820e-01 -2.33199939e-01
-6.27406836e-01 -6.40985668e-01 4.54842448e-01 3.70253384e-01
-5.89964151e-01 1.19259143e+00 -2.83556908e-01 -5.02799094e-01
-2.83364892e-01 -9.37578738e-01 -6.38565361e-01 -1.01967645e+00
-2.22222596e-01 6.99842572e-01 1.18321672e-01 1.65274099e-01
2.66962111e-01 4.80285466e-01 -1.33173168e+00 -6.95075691e-01
6.34416103e-01 2.76624948e-01 2.70298868e-01 7.22615838e-01
1.43132910e-01 -2.90119410e-01 -1.09353401e-01 -3.97385240e-01
-8.72368962e-02 -2.86460668e-01 -4.42130715e-01 -1.53061152e-01
-2.06319213e-01 -1.50293708e-01 2.51057804e-01 9.07432258e-01
7.33601034e-01 -2.27287114e-02 1.02085543e+00 -1.20534885e+00
-2.20064729e-01 6.93799138e-01 -1.31927884e+00 -1.04508430e-01
7.95321946e-04 5.45035340e-02 9.17027414e-01 -7.40827441e-01
5.25482178e-01 1.62834644e+00 7.96884716e-01 1.61613137e-01
-1.92648339e+00 -6.38791144e-01 -6.02682903e-02 -2.15001419e-01
-1.40315592e+00 -3.28711569e-01 2.35773608e-01 -5.76245308e-01
6.58170164e-01 2.62737662e-01 7.47702181e-01 9.15655971e-01
3.72121036e-01 7.38588095e-01 9.75295901e-01 -6.61706567e-01
6.39426589e-01 -1.88697234e-01 1.96525529e-01 3.04219753e-01
6.23671830e-01 3.17167938e-01 -5.13001740e-01 -8.96304905e-01
8.25527370e-01 1.89596843e-02 5.69745013e-03 -3.34487528e-01
-9.82779741e-01 1.40749121e+00 -2.44318351e-01 -2.21035957e-01
-4.52331632e-01 4.09414172e-01 1.85044065e-01 -8.89828950e-02
6.41888738e-01 3.22116315e-01 -6.98678195e-01 -8.91026631e-02
-7.59628236e-01 4.55936909e-01 8.94102693e-01 1.00273979e+00
6.74686015e-01 -1.81183845e-01 -2.13693723e-01 9.16819274e-01
4.03394312e-01 8.67032290e-01 -1.33398354e-01 -8.57172489e-01
-1.44072488e-01 8.72920603e-02 3.35559309e-01 -5.96927166e-01
-5.32673478e-01 -1.64124757e-01 -1.04585147e+00 -1.09099768e-01
4.21906590e-01 -2.13963613e-01 -8.27679217e-01 1.92195272e+00
7.12243855e-01 1.19868040e-01 -5.05575836e-01 3.02862048e-01
8.04623365e-02 4.85661358e-01 3.05782139e-01 -8.48693848e-01
1.37087333e+00 -9.06675160e-02 -5.89912176e-01 -9.48064402e-02
2.42149144e-01 -6.03308856e-01 7.47731388e-01 5.36787331e-01
-8.51758897e-01 2.06311777e-01 -4.87916261e-01 1.55864224e-01
-9.66626406e-02 -2.20623836e-01 1.14971673e+00 8.41939330e-01
-9.14093792e-01 3.45992953e-01 -1.01857948e+00 -5.67123532e-01
5.14365613e-01 4.83234614e-01 9.64877009e-02 -1.67500507e-02
-8.74107957e-01 5.25054395e-01 1.67518586e-01 9.47124362e-02
-7.72343516e-01 -6.64482236e-01 -7.56870627e-01 1.96899742e-01
3.55730295e-01 -7.81365156e-01 1.02296031e+00 -1.12749770e-01
-1.49311912e+00 7.18953490e-01 -4.81792241e-01 -2.68534064e-01
3.95912021e-01 1.00390136e-01 3.13098192e-01 -1.90135226e-01
1.06849238e-01 5.15535295e-01 5.81314623e-01 -8.10832560e-01
-5.27272165e-01 -6.81330025e-01 -4.60959315e-01 -1.31389737e-01
-9.82274115e-03 3.40205669e-01 -2.37919047e-01 -6.17273986e-01
4.12171662e-01 -1.02844477e+00 -7.38324225e-01 2.81459212e-01
-3.38365704e-01 -1.10080652e-01 1.23588212e-01 -4.36668068e-01
8.82814825e-01 -1.95498228e+00 4.05333430e-01 2.77569473e-01
2.37844847e-02 -2.80314296e-01 -3.68792284e-03 4.13023323e-01
4.11175579e-01 3.27416271e-01 -7.61392534e-01 -1.52455077e-01
-1.35774529e-02 1.00502238e-01 -1.05651975e-01 7.94085383e-01
3.82507265e-01 5.76964140e-01 -9.20541286e-01 -5.37587941e-01
7.06593469e-02 6.21180415e-01 -7.53162146e-01 -3.18921804e-02
-1.75384432e-01 3.13960105e-01 -4.39158797e-01 8.57222855e-01
9.32014525e-01 -3.25179905e-01 4.00440276e-01 6.44899309e-01
-3.53895515e-01 8.59817583e-03 -1.21041536e+00 1.21129382e+00
-2.22462982e-01 4.16136652e-01 4.47309047e-01 -8.91571701e-01
7.36345530e-01 9.79215577e-02 3.05776536e-01 1.88546136e-01
1.29274204e-01 1.88329220e-01 -1.88615873e-01 -1.52545452e-01
1.89365745e-01 -4.34655398e-01 -6.08881786e-02 3.45776379e-01
-6.14804327e-02 -4.97251511e-01 2.22514570e-01 -5.67948110e-02
9.94950533e-01 1.19777128e-01 9.83202219e-01 -4.94008452e-01
7.24473000e-02 -3.35453413e-02 8.44623864e-01 1.18436956e+00
-8.91598985e-02 3.61639053e-01 7.82820404e-01 4.10115942e-02
-7.67431319e-01 -9.60861504e-01 -8.89276803e-01 1.12151814e+00
-1.09134465e-01 -1.25069559e-01 -2.03838319e-01 -2.14447901e-02
2.88034201e-01 7.10423648e-01 -6.36728525e-01 1.69502169e-01
-1.97110951e-01 -1.87317288e+00 3.59922737e-01 2.46204838e-01
-2.24898439e-02 -7.66012490e-01 -4.81088758e-01 3.59207094e-01
1.42383546e-01 -6.27528131e-01 9.22295377e-02 6.73242867e-01
-9.67141569e-01 -1.05739200e+00 -7.07840741e-01 -3.19274545e-01
4.48604554e-01 2.54515857e-01 9.78140056e-01 -1.36271641e-01
-4.94357288e-01 1.43383518e-01 -2.05121279e-01 -5.93610883e-01
-2.11726516e-01 -6.59212172e-02 3.86698656e-02 -4.11319107e-01
6.77512169e-01 -5.94395816e-01 -4.57160056e-01 2.62762189e-01
-8.35303962e-01 -1.55997962e-01 4.72936064e-01 1.28312635e+00
7.14981437e-01 -8.02075341e-02 3.95273775e-01 -1.13602924e+00
2.38325879e-01 -7.13262916e-01 -1.23338211e+00 3.86676937e-01
-4.71844167e-01 7.12976381e-02 1.01480141e-01 -5.72309911e-01
-1.07801473e+00 2.56820083e-01 -6.53062537e-02 4.46367979e-01
-8.38167071e-02 8.01320732e-01 -1.45814434e-01 -2.22729504e-01
4.97602910e-01 -3.90438922e-02 -9.54079926e-02 -4.50515032e-01
1.03421159e-01 5.86362898e-01 5.62728941e-02 -6.21495068e-01
5.17644763e-01 4.98381734e-01 8.01788688e-01 -1.19685602e+00
-1.79596886e-01 -1.71176121e-01 -6.68398917e-01 2.12057799e-01
5.65709233e-01 -8.28493178e-01 -9.89021599e-01 4.66671139e-01
-9.45485115e-01 -3.47623974e-01 7.61153698e-02 8.56765866e-01
-5.99933684e-01 2.81418592e-01 -6.61838174e-01 -1.30151629e+00
-1.37227941e-02 -1.19667459e+00 1.14757764e+00 2.74728239e-01
-1.01339705e-01 -1.08687532e+00 3.35876524e-01 -2.35245034e-01
2.60117799e-01 3.16122562e-01 1.00063193e+00 -4.09802943e-01
-4.32726115e-01 -1.29238665e-01 -9.74631980e-02 -5.19386590e-01
-6.57717437e-02 4.60671604e-01 -8.88769627e-01 -3.40535522e-01
-1.63319737e-01 -2.55481839e-01 1.05023646e+00 1.17148936e+00
1.04382753e+00 1.50147930e-01 -7.00513482e-01 7.21000135e-01
1.37371218e+00 -4.01427746e-02 1.74592063e-01 1.01944119e-01
2.97336966e-01 7.62621939e-01 6.19210780e-01 8.51334810e-01
7.74062797e-02 7.31399477e-01 1.80578500e-01 -5.93966199e-03
5.37654281e-01 -2.52189394e-02 7.62982760e-03 4.13425297e-01
1.88990951e-01 -2.65210658e-01 -9.23800766e-01 5.21934211e-01
-1.86516094e+00 -7.88856566e-01 -5.23089468e-01 2.65392494e+00
1.29019451e+00 -2.29255110e-01 3.13429266e-01 1.69727635e-02
8.72414410e-01 -3.52374226e-01 -7.50870764e-01 -1.87381133e-01
-3.95125300e-01 2.75321901e-01 8.18301737e-01 6.14124179e-01
-1.01061451e+00 7.88204432e-01 8.13027668e+00 9.80972111e-01
-5.76153040e-01 3.25799361e-02 7.27028966e-01 -7.53790289e-02
-4.03102815e-01 3.13983560e-01 -1.07964540e+00 4.28155065e-01
8.12496424e-01 -1.28581807e-01 4.25820112e-01 5.87151766e-01
3.99638325e-01 -6.68188930e-01 -9.71645355e-01 6.57505929e-01
-4.81864989e-01 -7.55810440e-01 -5.23384392e-01 3.75088006e-01
5.46210647e-01 -1.37119457e-01 -2.20238660e-02 8.35871175e-02
9.51859176e-01 -9.48726892e-01 3.49455364e-02 -1.12154253e-01
6.99678004e-01 -8.40955555e-01 5.14181852e-01 3.70342374e-01
-7.46998727e-01 1.86315447e-01 -9.14230227e-01 -2.76898682e-01
3.51251930e-01 1.09977782e+00 -8.27980161e-01 2.12135650e-02
6.18040919e-01 4.61365819e-01 -3.76015723e-01 1.25890124e+00
-2.94360191e-01 9.24649358e-01 -8.48938942e-01 -2.86643684e-01
-3.36775720e-01 -4.87299949e-01 5.00541031e-01 1.24603403e+00
4.94592011e-01 1.06661700e-01 -1.65586993e-01 8.74103189e-01
2.86829233e-01 6.56859353e-02 -4.98980463e-01 1.10189237e-01
9.62717831e-01 1.10936761e+00 -1.17016888e+00 9.56208911e-03
-3.11092705e-01 4.71055120e-01 1.99618965e-01 4.08320457e-01
-5.47125757e-01 -1.00836791e-01 5.09088039e-01 -2.65406698e-01
4.33568954e-01 -1.38207644e-01 -4.84780341e-01 -1.13804746e+00
-4.54967558e-01 -6.16342068e-01 5.83372176e-01 -2.44436339e-01
-1.39029217e+00 -3.21933180e-01 4.62934047e-01 -5.64552903e-01
-4.38634217e-01 -7.68906653e-01 -9.12422836e-02 1.01266766e+00
-1.32223284e+00 -9.34436262e-01 1.96792111e-01 5.52515946e-02
2.49184892e-01 2.36782819e-01 7.86547959e-01 -6.76605478e-02
-7.18921006e-01 3.44955683e-01 6.87013865e-01 -5.75534046e-01
6.52577758e-01 -1.31593835e+00 4.29825902e-01 7.62278318e-01
-3.17315310e-01 7.83897161e-01 1.05917585e+00 -8.63709450e-01
-1.67765820e+00 -5.93976140e-01 7.72333384e-01 -1.15987383e-01
7.60062635e-01 -5.25622129e-01 -6.43075585e-01 7.49335945e-01
-3.68815005e-01 -3.75810832e-01 1.10252130e+00 6.62099361e-01
-1.05018519e-01 2.92548239e-01 -1.37669444e+00 6.49404883e-01
8.32865119e-01 -7.67218098e-02 1.02871090e-01 5.00664949e-01
5.45766175e-01 -1.30671397e-01 -8.59019816e-01 3.50509316e-01
8.81414115e-01 -4.43838477e-01 9.64491844e-01 -5.25419831e-01
1.49277896e-01 -2.84834504e-01 -3.19599926e-01 -1.03030610e+00
-4.31552172e-01 -8.60062480e-01 2.47858226e-01 1.42556667e+00
3.32563788e-01 -6.75139666e-01 4.61954325e-01 6.65580630e-01
5.90117633e-01 -3.00818592e-01 -1.10162020e+00 -5.01394629e-01
4.87244576e-01 -3.09724927e-01 4.92783427e-01 7.35934973e-01
-1.80264011e-01 2.42699847e-01 -6.04593635e-01 7.21805319e-02
1.10155725e+00 1.49266079e-01 7.12847650e-01 -1.58497107e+00
-6.20958626e-01 -3.84000212e-01 -1.83256313e-01 -1.01906633e+00
-1.17321254e-03 -4.42192882e-01 3.41216952e-01 -8.20629656e-01
8.88789833e-01 -4.47012216e-01 1.77573711e-01 2.19582543e-01
-7.43414223e-01 1.15224592e-01 -3.28765064e-01 -8.16957057e-02
2.08688900e-02 4.20438945e-01 7.23754108e-01 -5.95041597e-03
-3.75595540e-01 2.99954563e-01 -5.99709332e-01 5.91437995e-01
6.26877010e-01 -8.96860301e-01 1.09170787e-02 -1.96652144e-01
5.23723304e-01 3.08360815e-01 3.40013623e-01 -1.04075156e-01
2.29645167e-02 -7.33610511e-01 4.63500410e-01 -6.57072544e-01
3.47472996e-01 -5.47429681e-01 4.49944347e-01 4.97713059e-01
-3.29471976e-01 -1.94914676e-02 1.90234274e-01 8.00157607e-01
2.39000350e-01 -5.46668470e-01 7.03751147e-01 -7.26448223e-02
6.42614737e-02 2.23226160e-01 -6.50981903e-01 -1.90526083e-01
7.17245102e-01 1.75661087e-01 -2.66809791e-01 -2.33950585e-01
-5.88905871e-01 4.90962714e-01 6.27982140e-01 -4.22566503e-01
4.42695349e-01 -8.92813683e-01 -1.02709401e+00 5.99047169e-02
-1.05716683e-01 1.83868986e-02 -9.61046244e-05 1.03946340e+00
-5.21041214e-01 2.48201683e-01 8.24236497e-02 -8.84909630e-01
-1.28474534e+00 5.63434362e-01 -2.37489462e-01 -1.57041550e-01
-1.67786375e-01 8.17504585e-01 2.49291092e-01 -4.67016667e-01
1.45468846e-01 4.61964048e-02 -7.12650865e-02 1.21867955e-01
5.89773178e-01 5.30657828e-01 -4.53166157e-01 -2.07236066e-01
-4.47081208e-01 4.94359255e-01 2.12395504e-01 -4.43617225e-01
1.72144008e+00 -3.07352573e-01 -5.95389903e-01 7.80245900e-01
6.49584591e-01 -7.08101690e-02 -1.25362527e+00 -8.88495743e-02
2.04255134e-01 -5.72337151e-01 2.68864799e-02 -3.51607472e-01
-5.84608793e-01 8.44604850e-01 2.68828869e-01 1.81574136e-01
1.01810443e+00 -1.90126061e-01 -1.29279599e-01 4.01295096e-01
4.99142140e-01 -6.94309294e-01 -7.10249424e-01 2.83892244e-01
4.87863541e-01 -1.35525966e+00 5.86591125e-01 -6.24681711e-01
-1.34471446e-01 9.72732067e-01 7.00224042e-02 1.50192782e-01
8.92886698e-01 5.22657931e-01 -5.30582011e-01 4.27123271e-02
-8.94565523e-01 -1.32656440e-01 -2.49674767e-01 8.33641946e-01
6.58851027e-01 2.86763161e-01 -9.04462278e-01 3.60241711e-01
1.56999659e-02 9.83773321e-02 7.16277301e-01 9.70087886e-01
-4.28514361e-01 -1.26229405e+00 -6.35573387e-01 7.04068363e-01
-6.34396136e-01 -3.52765322e-01 -2.80809164e-01 7.32773125e-01
-4.43508297e-01 1.05030370e+00 -1.05103232e-01 2.78091282e-01
-3.85154575e-01 -4.22913581e-02 5.02705574e-01 -4.84078825e-01
-3.51712465e-01 7.02711403e-01 -1.02393180e-01 -1.53275207e-01
-4.47177827e-01 -1.23675525e+00 -6.80476129e-01 -5.37412286e-01
-7.83680201e-01 9.53141451e-02 7.73748279e-01 7.46229589e-01
4.66900170e-02 5.48547730e-02 4.60591972e-01 -9.00551438e-01
-8.28074992e-01 -9.42932904e-01 -9.92891192e-01 -2.21567214e-01
2.21194491e-01 -9.07368600e-01 -5.95600009e-01 -3.64025496e-02] | [7.325560092926025, 4.564911842346191] |
0497ec47-d349-48e2-99c8-73ec0e0cebf1 | ambernet-a-compact-end-to-end-model-for | 2210.15781 | null | https://arxiv.org/abs/2210.15781v1 | https://arxiv.org/pdf/2210.15781v1.pdf | AmberNet: A Compact End-to-End Model for Spoken Language Identification | We present AmberNet, a compact end-to-end neural network for Spoken Language Identification. AmberNet consists of 1D depth-wise separable convolutions and Squeeze-and-Excitation layers with global context, followed by statistics pooling and linear layers. AmberNet achieves performance similar to state-of-the-art(SOTA) models on VoxLingua107 dataset, while being 10x smaller. AmberNet can be adapted to unseen languages and new acoustic conditions with simple finetuning. It attains SOTA accuracy of 75.8% on FLEURS benchmark. We show the model is easily scalable to achieve a better trade-off between accuracy and speed. We further inspect the model's sensitivity to input length and show that AmberNet performs well even on short utterances. | ['Boris Ginsburg', 'Jagadeesh Balam', 'Nithin Rao Koluguri', 'Fei Jia'] | 2022-10-27 | null | null | null | null | ['spoken-language-identification'] | ['speech'] | [-2.57109672e-01 -1.55030936e-01 2.40460739e-01 -8.26855898e-01
-9.44909632e-01 -6.35986984e-01 4.13146436e-01 -1.33015528e-01
-8.14963400e-01 1.14558034e-01 2.87209660e-01 -3.80667001e-01
3.24347258e-01 -3.15125018e-01 -4.73623961e-01 -5.23083746e-01
-3.39208603e-01 5.25605500e-01 4.15219329e-02 -1.64182171e-01
-3.07655990e-01 2.41178110e-01 -1.29124105e+00 5.04198611e-01
3.00212324e-01 9.72880185e-01 4.64939363e-02 1.33272743e+00
-1.02238201e-01 4.78041291e-01 -6.37440205e-01 -1.83295220e-01
1.58736750e-01 -1.34200931e-01 -9.43943143e-01 -2.45691523e-01
6.82342529e-01 -4.18028235e-01 -5.41156769e-01 5.93231499e-01
1.07351470e+00 3.01433951e-01 3.48828197e-01 -8.77727389e-01
-8.69083881e-01 1.02378201e+00 -6.12121969e-02 4.31383520e-01
1.45875528e-01 3.40012223e-01 1.13488364e+00 -1.23173499e+00
8.73354003e-02 1.64269090e+00 9.21253681e-01 8.28661323e-01
-1.15679324e+00 -8.07252228e-01 1.85199291e-01 -4.03524876e-01
-1.45240819e+00 -1.04157114e+00 1.97942749e-01 1.09210901e-01
1.79750824e+00 3.13816398e-01 2.24444315e-01 1.20001769e+00
-2.70762265e-01 1.17051280e+00 7.08998859e-01 -2.26261705e-01
3.90969694e-01 -6.37520989e-03 2.25497723e-01 7.16642976e-01
-4.46498126e-01 -6.75978959e-02 -7.34651923e-01 -1.51175946e-01
4.56493825e-01 -5.80378890e-01 -8.29206631e-02 1.98427707e-01
-1.21777558e+00 7.04404950e-01 3.57446194e-01 3.40063483e-01
7.21584558e-02 4.32711661e-01 7.42104411e-01 5.08816361e-01
3.94768953e-01 4.77550656e-01 -8.73804331e-01 -4.83849078e-01
-9.95809972e-01 2.33763590e-01 1.04420340e+00 9.96677160e-01
1.45478919e-01 5.27908146e-01 -1.60234287e-01 1.27257609e+00
1.48359135e-01 5.75458527e-01 7.36350238e-01 -7.57145703e-01
4.64927822e-01 2.56223027e-02 -4.02952284e-01 -1.66380584e-01
-6.44947171e-01 -5.45968056e-01 -7.72943020e-01 -5.39393798e-02
5.50984263e-01 -5.15472412e-01 -1.31877708e+00 1.77786231e+00
-2.44827136e-01 1.94078133e-01 3.47129524e-01 8.33574295e-01
1.19614518e+00 8.66701782e-01 2.53633827e-01 3.25576872e-01
1.25987458e+00 -1.12349617e+00 -5.18409312e-01 -6.75682366e-01
6.57015979e-01 -6.02733672e-01 1.31195831e+00 3.89035434e-01
-1.21307063e+00 -5.43816507e-01 -8.83013725e-01 -3.47831547e-01
-5.64709067e-01 2.38619521e-01 5.35872340e-01 8.11268985e-01
-1.44472897e+00 3.64446640e-01 -9.19059873e-01 -6.25686586e-01
1.66587919e-01 6.31263137e-01 -2.73448139e-01 2.66456336e-01
-1.14696205e+00 6.11060679e-01 4.55639124e-01 2.39488572e-01
-9.81881320e-01 -6.23806417e-01 -1.04358733e+00 3.14533025e-01
1.55165166e-01 -4.49184895e-01 1.69795632e+00 -5.59885740e-01
-2.04874086e+00 8.28050256e-01 -2.66388208e-01 -6.75835431e-01
2.55835265e-01 -2.14124843e-01 -5.20467937e-01 -1.26458079e-01
-2.85886347e-01 9.13422883e-01 5.67114890e-01 -7.20175505e-01
-4.45236892e-01 -8.98919702e-02 -4.53133136e-01 1.31301790e-01
-4.47460711e-01 3.01990807e-01 -4.98406202e-01 -5.92059135e-01
-3.89646478e-02 -6.18543983e-01 -1.26901343e-01 -1.81384668e-01
-4.95455891e-01 -4.43807542e-01 7.99370229e-01 -6.26577616e-01
1.16004801e+00 -2.30035138e+00 -9.93657708e-02 5.14908060e-02
4.89040315e-02 6.82538271e-01 -6.34911001e-01 2.87619412e-01
8.89321715e-02 2.19367251e-01 -2.54598886e-01 -1.01124513e+00
4.14897680e-01 1.78712636e-01 -1.94849551e-01 2.66275644e-01
3.89360160e-01 1.03213048e+00 -7.90054858e-01 -4.33445983e-02
9.99526978e-02 5.61234713e-01 -5.63196898e-01 4.68410969e-01
-9.07874405e-02 1.01818174e-01 -9.50988606e-02 7.58748233e-01
6.35549426e-01 5.80024607e-02 -1.08722344e-01 2.72678584e-01
-7.71563947e-02 6.34591758e-01 -9.20328498e-01 1.74702466e+00
-6.35771096e-01 8.63288105e-01 5.34393966e-01 -6.77446485e-01
1.11477959e+00 3.37428838e-01 -2.57667780e-01 -4.36857224e-01
3.90111148e-01 3.36846083e-01 -5.23746498e-02 -3.42506647e-01
4.78139907e-01 9.44824070e-02 -2.37580672e-01 2.81818271e-01
5.84763587e-01 -2.34724760e-01 -3.41653585e-01 2.49860249e-02
1.01424170e+00 -5.13469696e-01 5.44998385e-02 -3.74212742e-01
2.34880254e-01 -5.40827096e-01 1.56483009e-01 1.03620601e+00
-3.29749972e-01 7.38003910e-01 1.69932410e-01 -5.87292552e-01
-1.01223481e+00 -1.18480647e+00 -1.99137494e-01 1.64704537e+00
-4.10937518e-01 -4.48406547e-01 -8.89870822e-01 -4.06566322e-01
2.26601809e-02 6.42507195e-01 -5.32012224e-01 2.41236966e-02
-4.15876776e-01 -6.99564874e-01 1.40986264e+00 8.44972551e-01
8.07569146e-01 -1.15521276e+00 -3.81282121e-01 2.67158508e-01
2.30888918e-01 -1.29168022e+00 -7.80750573e-01 6.01230443e-01
-4.93407786e-01 -3.60137582e-01 -7.22204149e-01 -1.16672897e+00
1.45678259e-02 -1.67543009e-01 1.28958035e+00 -2.76652366e-01
-1.15874372e-01 1.73305228e-01 -1.37145981e-01 -3.73771548e-01
-5.77993572e-01 5.99948704e-01 3.16054672e-01 -2.04596370e-01
7.54105449e-01 -3.22250038e-01 -3.55392605e-01 2.59345651e-01
-6.45903111e-01 -3.33828628e-01 2.68221974e-01 1.14072549e+00
1.95262566e-01 -2.59634584e-01 6.84379637e-01 -5.14452636e-01
9.06219959e-01 -2.11866364e-01 -5.91001987e-01 2.10957631e-01
-1.55531928e-01 1.27551809e-01 6.55694127e-01 -5.30227780e-01
-7.60081828e-01 1.85671344e-01 -7.30504453e-01 -3.00085366e-01
-4.11356121e-01 2.79756337e-01 -1.81413502e-01 5.17672393e-04
5.71641922e-01 3.15153122e-01 6.53820932e-02 -6.19248152e-01
4.54013675e-01 1.20816100e+00 8.28505635e-01 -2.74168313e-01
2.10214257e-01 1.26525208e-01 -7.78998852e-01 -1.33171856e+00
-5.10825157e-01 -5.34366012e-01 -5.34078717e-01 3.36304069e-01
7.51347899e-01 -1.05457509e+00 -9.83140528e-01 9.93330359e-01
-9.94409800e-01 -8.39847326e-01 -1.21493004e-01 3.36830884e-01
-3.42119515e-01 6.43329173e-02 -1.07932305e+00 -1.09264624e+00
-8.28432024e-01 -1.08230615e+00 1.18507576e+00 2.01647311e-01
-3.48917007e-01 -9.86952007e-01 -6.32858053e-02 -7.01509342e-02
7.25688100e-01 -4.04377639e-01 3.88198674e-01 -1.06766284e+00
-1.22294739e-01 -1.21695973e-01 -1.61720946e-01 6.80553496e-01
-2.14773253e-01 -2.06813719e-02 -1.65996838e+00 -6.24728799e-01
-3.60698998e-01 -7.04894125e-01 1.24623775e+00 4.30801481e-01
1.11616635e+00 -4.40271616e-01 -6.58691749e-02 9.63040352e-01
8.07591081e-01 1.06751084e-01 3.80922377e-01 5.79813756e-02
6.07396185e-01 3.38055909e-01 -1.41897023e-01 3.45142573e-01
5.00305533e-01 4.28320885e-01 2.06332818e-01 -2.53937155e-01
-7.10011795e-02 -1.71746388e-01 4.67745721e-01 1.01733863e+00
4.86014873e-01 -6.02651536e-01 -1.04955101e+00 8.62102926e-01
-1.56763864e+00 -7.24706769e-01 3.80684763e-01 1.97341466e+00
7.60702014e-01 1.32870942e-01 3.39254081e-01 -2.91475537e-03
3.16471994e-01 2.64133871e-01 -5.39190173e-01 -1.01399291e+00
-4.07766640e-01 4.83093500e-01 4.75838661e-01 9.14938211e-01
-1.19964540e+00 1.30241621e+00 7.51986885e+00 8.11379135e-01
-1.25844097e+00 -6.75642639e-02 9.40278590e-01 -2.48799026e-01
-9.89809111e-02 -6.16568565e-01 -1.04114056e+00 1.62601933e-01
1.22882009e+00 3.32804024e-01 6.44729316e-01 8.57501566e-01
-8.13591108e-02 2.82508135e-01 -1.19817901e+00 9.69110847e-01
1.45284191e-01 -1.11052918e+00 -3.16793233e-01 -3.35163444e-01
2.86728352e-01 8.85544121e-01 2.97559142e-01 6.54068351e-01
4.67502654e-01 -1.48419261e+00 6.08922839e-01 -1.37543276e-01
1.12029004e+00 -8.36807430e-01 9.06230748e-01 2.84393996e-01
-1.04622459e+00 -6.78075626e-02 -3.26629221e-01 -9.12343562e-02
4.57014237e-03 7.53170028e-02 -1.38816547e+00 -2.76514649e-01
9.40721452e-01 3.77649724e-01 -3.78573149e-01 7.06766665e-01
6.17295802e-02 9.79241073e-01 -8.59174371e-01 -5.09379029e-01
6.62598908e-01 4.54551935e-01 5.05909383e-01 1.99247801e+00
1.70714349e-01 1.37082890e-01 5.60599566e-02 8.09088171e-01
-2.38199890e-01 -7.89044648e-02 -4.52892184e-01 -2.83815116e-01
5.93414426e-01 9.20590699e-01 -2.23767817e-01 -4.51624691e-01
-2.71203995e-01 1.27258110e+00 4.78928298e-01 7.11193621e-01
-2.31266111e-01 -5.50749660e-01 1.03073192e+00 -3.22929353e-01
5.63189626e-01 -4.04073566e-01 -1.80022284e-01 -1.02166712e+00
-3.59435707e-01 -8.29584360e-01 3.82862926e-01 -4.28131729e-01
-1.42847359e+00 1.16222847e+00 -3.59001040e-01 -4.40255344e-01
-3.84735018e-01 -1.16162241e+00 -6.39094830e-01 1.00977314e+00
-1.34539616e+00 -1.17596745e+00 -8.29254091e-02 5.25788665e-01
8.69861186e-01 -4.71372962e-01 1.31525612e+00 2.04406127e-01
-5.79708755e-01 1.27557397e+00 2.17291877e-01 6.12969518e-01
6.22151792e-01 -1.33359897e+00 1.35197854e+00 7.62372255e-01
1.78163245e-01 5.02795458e-01 6.43526793e-01 -5.17730713e-02
-1.12235081e+00 -1.05337548e+00 1.12653911e+00 -4.95021671e-01
6.60405695e-01 -1.08024025e+00 -1.03454387e+00 7.86801457e-01
5.84743679e-01 2.05876842e-01 6.59229696e-01 4.59593087e-01
-6.72174811e-01 -1.14423886e-01 -1.06666696e+00 6.23404801e-01
9.89540160e-01 -8.95336330e-01 -3.86705548e-01 -3.69323082e-02
9.45449948e-01 -7.31449604e-01 -6.46976948e-01 2.64630526e-01
7.97629178e-01 -8.77697229e-01 9.16363955e-01 -7.47505963e-01
-1.30385369e-01 1.89728975e-01 -2.56573230e-01 -1.51790726e+00
-2.93976575e-01 -1.04217744e+00 1.11962348e-01 1.10259128e+00
7.18671679e-01 -7.20983505e-01 5.80055296e-01 4.66579229e-01
-2.85764188e-01 -7.51443267e-01 -1.30805242e+00 -8.31211269e-01
3.32160681e-01 -8.18828285e-01 7.01394260e-01 8.52823496e-01
2.29272574e-01 5.58841169e-01 -3.12261432e-01 3.97751957e-01
2.52917081e-01 -3.64936292e-01 6.15654886e-01 -5.60378611e-01
-4.39124137e-01 -6.34837091e-01 -2.54896879e-01 -1.53463650e+00
2.40702689e-01 -6.70920074e-01 3.14587563e-01 -9.29250360e-01
-1.31155849e-01 -3.28477651e-01 -4.70432580e-01 7.94125319e-01
-1.26948774e-01 3.44614178e-01 1.33505628e-01 -4.30435002e-01
-6.03751540e-01 5.74930966e-01 6.91653609e-01 -3.24630141e-01
-5.75958550e-01 -9.05772597e-02 -3.05366814e-01 5.23582160e-01
9.97983634e-01 5.81458360e-02 -2.25439951e-01 -9.30476665e-01
-2.31654599e-01 -2.86024213e-01 1.46269083e-01 -9.21803176e-01
2.85160631e-01 3.46364826e-01 1.95216537e-01 -6.06575310e-01
5.78119755e-01 -3.00221741e-01 -5.69084346e-01 1.31771252e-01
-7.20573008e-01 8.89973417e-02 7.69114256e-01 7.97226056e-02
-2.99258888e-01 1.59390375e-01 6.57375634e-01 1.67483278e-02
-8.22527468e-01 2.26116151e-01 -6.98070586e-01 1.73666105e-01
1.90103099e-01 6.95390766e-03 -3.67008060e-01 -5.48587024e-01
-8.14087212e-01 5.14172554e-01 -1.00662276e-01 6.19090855e-01
7.63439476e-01 -1.08254874e+00 -1.01022816e+00 6.77821159e-01
2.02677682e-01 1.01464503e-01 1.45494074e-01 2.96573579e-01
-5.11588275e-01 5.22955537e-01 2.85433680e-01 -6.95375979e-01
-1.33063853e+00 2.03503549e-01 7.23231852e-01 1.51520714e-01
-4.42152649e-01 1.69353724e+00 1.07674286e-01 -1.00772190e+00
8.45908523e-01 -1.02447319e+00 1.33899271e-01 -3.15830231e-01
7.92188108e-01 3.32309976e-02 6.55437559e-02 -3.97666991e-01
-6.06887877e-01 2.42270470e-01 -6.47828728e-02 -1.95292056e-01
1.17052972e+00 -1.84553653e-01 2.60697246e-01 6.65452242e-01
1.34943068e+00 -1.11335749e-02 -1.23040640e+00 -3.39135647e-01
-1.65636331e-01 1.63537916e-02 2.87796408e-01 -1.12251198e+00
-8.60590935e-01 1.19582379e+00 6.86353385e-01 1.31224558e-01
1.02671945e+00 1.34679332e-01 1.12186110e+00 6.69927359e-01
-1.22426428e-01 -1.01216280e+00 -1.00994982e-01 9.90037382e-01
8.61486018e-01 -1.43636930e+00 -5.93218386e-01 -1.15028471e-01
-6.31192803e-01 1.08919406e+00 4.96214867e-01 3.62706445e-02
7.57708132e-01 7.79105365e-01 5.45331597e-01 -3.16004828e-02
-9.19039488e-01 -1.52546272e-01 2.39146769e-01 4.84380752e-01
5.79070866e-01 3.20512921e-01 5.54508805e-01 6.90530002e-01
-4.65273917e-01 -3.26403856e-01 -2.65749861e-02 4.82287467e-01
-3.31979603e-01 -6.60267830e-01 -2.18707085e-01 1.50458679e-01
-3.81260365e-01 -5.46855271e-01 -5.06588042e-01 6.39373064e-01
-9.64719355e-02 9.81540620e-01 4.59132820e-01 -6.30080342e-01
3.68410438e-01 2.14439601e-01 1.19314231e-01 -5.92558146e-01
-8.69053483e-01 3.46139252e-01 4.08713877e-01 -4.46711749e-01
7.59295076e-02 -5.03913283e-01 -1.36136138e+00 -4.57382113e-01
-2.08282158e-01 1.94176976e-02 5.18388212e-01 7.76216626e-01
4.47622150e-01 4.57350850e-01 4.35346693e-01 -6.81477129e-01
-7.15470493e-01 -1.28323996e+00 -5.69500625e-01 -4.18936349e-02
1.00499499e+00 7.20035508e-02 -4.59834129e-01 -3.08389664e-01] | [14.290417671203613, 6.4580278396606445] |
3b1d799e-e34a-47ff-ba9e-7c40ee3f3b18 | epidemic-control-on-a-large-scale-agent-based | 2304.04475 | null | https://arxiv.org/abs/2304.04475v1 | https://arxiv.org/pdf/2304.04475v1.pdf | Epidemic Control on a Large-Scale-Agent-Based Epidemiology Model using Deep Deterministic Policy Gradient | To mitigate the impact of the pandemic, several measures include lockdowns, rapid vaccination programs, school closures, and economic stimulus. These interventions can have positive or unintended negative consequences. Current research to model and determine an optimal intervention automatically through round-tripping is limited by the simulation objectives, scale (a few thousand individuals), model types that are not suited for intervention studies, and the number of intervention strategies they can explore (discrete vs continuous). We address these challenges using a Deep Deterministic Policy Gradient (DDPG) based policy optimization framework on a large-scale (100,000 individual) epidemiological agent-based simulation where we perform multi-objective optimization. We determine the optimal policy for lockdown and vaccination in a minimalist age-stratified multi-vaccine scenario with a basic simulation for economic activity. With no lockdown and vaccination (mid-age and elderly), results show optimal economy (individuals below the poverty line) with balanced health objectives (infection, and hospitalization). An in-depth simulation is needed to further validate our results and open-source our framework. | ['Janani Venugopalan', 'Harshal Hayatnagarkar', 'Jayanta Kshirsagar', 'Gaurav Deshkar'] | 2023-04-10 | null | null | null | null | ['epidemiology'] | ['medical'] | [-1.47561818e-01 1.42909244e-01 -5.81012309e-01 1.38460055e-01
-2.30958968e-01 -2.88608372e-01 5.51162243e-01 7.49881923e-01
-7.65705884e-01 1.10887170e+00 6.26926303e-01 -7.93084502e-01
-3.84314805e-01 -1.07632923e+00 -5.37095249e-01 -4.38838422e-01
-6.03827178e-01 9.57989514e-01 -3.59970033e-02 -3.93522233e-01
-2.31721282e-01 2.02957079e-01 -1.44201195e+00 -3.15305531e-01
1.28672862e+00 -7.04310164e-02 1.63621441e-01 7.58562267e-01
3.91064972e-01 2.54977524e-01 -6.56297266e-01 -3.59250419e-02
3.85954559e-01 -4.27002460e-01 -2.37626806e-01 -3.24922651e-01
-5.10666251e-01 -8.43975186e-01 3.07338119e-01 5.69955349e-01
9.49813128e-01 -1.22403204e-01 7.11248636e-01 -1.35685694e+00
-4.21398342e-01 4.86270756e-01 -4.54962641e-01 1.72880530e-01
4.59125340e-01 6.03121042e-01 3.15879136e-01 -5.82188778e-02
5.69363594e-01 1.71740484e+00 8.83335531e-01 5.56130290e-01
-1.23535156e+00 -4.12944973e-01 8.35408196e-02 -2.59637117e-01
-8.39805424e-01 -1.42305955e-01 9.59680602e-02 -3.88497233e-01
1.28397775e+00 3.36887658e-01 1.23815203e+00 9.63625669e-01
4.32900101e-01 1.57549813e-01 1.33543849e+00 -2.94807047e-01
4.12197828e-01 7.87295252e-02 -1.63380980e-01 5.75124800e-01
6.64270759e-01 5.64803541e-01 9.94217619e-02 -7.63830483e-01
6.09448433e-01 3.14352363e-01 -1.11948170e-01 1.11690424e-02
-1.05268633e+00 1.07080770e+00 6.71294183e-02 -1.87448964e-01
-1.00356257e+00 -1.92838740e-02 3.29781026e-01 3.92969400e-01
6.57032371e-01 4.07959133e-01 -8.33529890e-01 1.40892463e-02
-7.60865629e-01 8.87698174e-01 9.42649424e-01 1.38315380e-01
4.55686837e-01 2.43877177e-03 -3.62098426e-01 3.40917945e-01
3.17196369e-01 1.18464816e+00 -7.52662942e-02 -1.13808095e+00
4.25726116e-01 3.38368207e-01 6.42470777e-01 -8.69776964e-01
-8.97096097e-01 -3.48617941e-01 -7.24553585e-01 1.42325103e-01
5.98802567e-01 -1.19352245e+00 -7.11091459e-01 1.98866284e+00
6.55841947e-01 2.73098886e-01 -3.21586281e-02 5.43854535e-01
7.94427395e-02 7.18720198e-01 5.40852606e-01 -8.27989399e-01
1.41412628e+00 -4.81025904e-01 -5.41370094e-01 -2.08172485e-01
9.56649363e-01 -3.30504298e-01 6.16156936e-01 -3.87229562e-01
-1.16327679e+00 2.69395500e-01 -3.45072597e-01 9.40614045e-01
-6.49131775e-01 -6.11405015e-01 3.12091529e-01 7.01788127e-01
-1.26150084e+00 3.55027974e-01 -9.52666819e-01 -7.47993290e-01
1.12920895e-01 4.84057993e-01 2.03751773e-01 2.52359584e-02
-1.47656417e+00 1.10779262e+00 1.42699838e-01 -4.57694858e-01
-7.65458405e-01 -1.12244296e+00 -6.58052206e-01 2.29244918e-01
2.09072247e-01 -1.30023944e+00 7.62370765e-01 -8.54662657e-01
-9.93326068e-01 5.07683158e-01 1.95339426e-01 -7.54873872e-01
4.87828016e-01 3.59331220e-01 -8.81168153e-03 5.39511722e-03
1.41535610e-01 6.80118978e-01 3.03237289e-01 -8.19029748e-01
-6.85389638e-01 -6.30639136e-01 1.03797503e-01 7.00557530e-01
-2.72432327e-01 8.41206014e-01 6.28597379e-01 -7.31144130e-01
-9.91291344e-01 -7.95580387e-01 -7.20709801e-01 -1.00449634e+00
3.93487588e-02 -7.12343454e-02 3.33497256e-01 -1.21655357e+00
1.05421746e+00 -1.35988641e+00 2.97916196e-02 -1.38401791e-01
-6.62945732e-02 1.02376059e-01 -1.33719206e-01 7.92051136e-01
4.30335164e-01 1.27154112e-01 -3.53240430e-01 -5.37914596e-02
2.12992821e-02 1.39798880e-01 1.92354918e-01 4.78862733e-01
2.48980045e-01 7.62469530e-01 -9.06713009e-01 -3.24028164e-01
2.26295993e-01 4.66687590e-01 -1.01742446e+00 2.14180887e-01
-3.51522237e-01 4.83871639e-01 -5.83734035e-01 5.22730052e-01
6.61920249e-01 -5.44434041e-03 3.15489560e-01 6.51462495e-01
-3.04988772e-01 6.84220493e-02 -7.92672276e-01 5.21450579e-01
-3.36131245e-01 -2.87739243e-02 5.72171926e-01 -1.29387081e+00
1.92225024e-01 3.09212267e-01 8.75464916e-01 -6.51224554e-01
6.88264668e-02 -1.67382695e-02 2.87344843e-01 -5.31844378e-01
4.43590246e-03 -1.15826979e-01 2.15992071e-02 8.18185985e-01
-3.75912875e-01 1.16530269e-01 2.22225949e-01 7.89891556e-02
1.15365982e+00 -3.37953031e-01 1.89005941e-01 -1.00219119e+00
-1.80893362e-01 3.62488657e-01 7.72362232e-01 7.59611428e-01
-4.46511418e-01 -1.97665721e-01 6.53978288e-01 -1.90061793e-01
-1.10958612e+00 -8.76685619e-01 1.29584065e-02 1.19053793e+00
-1.82377204e-01 2.33732268e-01 -1.00681674e+00 -3.96766752e-01
3.26622427e-01 6.57815814e-01 -6.72090948e-01 3.73671278e-02
-7.51872003e-01 -1.68396926e+00 3.82949620e-01 1.70614123e-02
3.91545445e-01 -1.07907450e+00 -1.12185252e+00 4.78416890e-01
1.09949663e-01 -5.72600126e-01 -2.37350434e-01 -3.59618992e-01
-8.98211122e-01 -1.15768766e+00 -1.26978266e+00 -6.16767168e-01
9.09983933e-01 -2.06656143e-01 1.06322658e+00 1.36087209e-01
8.77372324e-02 6.99242949e-01 2.17664734e-01 -7.02942431e-01
-4.77123350e-01 -9.43955854e-02 4.35885310e-01 -6.69922173e-01
2.06300646e-01 -1.23580970e-01 -1.10867548e+00 5.98206818e-02
-7.12679088e-01 7.09499419e-02 8.93704519e-02 6.52508855e-01
5.81176504e-02 -1.08307935e-01 1.30852115e+00 -4.70223218e-01
1.10896909e+00 -9.51106310e-01 -6.66071415e-01 5.33873975e-01
-7.53823936e-01 -1.99661717e-01 4.24552292e-01 -6.33417189e-01
-8.59565854e-01 -5.23367107e-01 -8.81783590e-02 5.64907491e-01
1.09066619e-02 6.09441400e-01 3.96502465e-01 3.88718158e-01
4.00856853e-01 -1.37902424e-01 3.99112880e-01 -3.89992744e-01
-5.28567098e-03 6.13481939e-01 -8.12679827e-02 -6.57375216e-01
2.75495619e-01 3.40031952e-01 -1.15300603e-01 -6.96035564e-01
1.44629981e-02 1.51968911e-01 1.38628557e-01 -2.47252882e-01
7.25989401e-01 -1.07437742e+00 -9.60687459e-01 6.04708314e-01
-6.38932943e-01 -1.13855791e+00 -8.44896585e-02 8.54733050e-01
-4.46474999e-01 -2.96737272e-02 -5.89976847e-01 -1.11014450e+00
-5.98954737e-01 -1.03369951e+00 5.55105865e-01 4.60571796e-01
-1.55486882e-01 -1.44405329e+00 5.38798213e-01 7.48037547e-02
8.20133150e-01 7.89771676e-01 9.74886000e-01 -4.05877203e-01
-9.17234346e-02 3.00290495e-01 1.25556262e-02 -3.25593174e-01
2.86577523e-01 9.98257548e-02 8.03473517e-02 -8.69635284e-01
-1.20877095e-01 -2.07123369e-01 4.12059635e-01 1.12888241e+00
3.61322820e-01 -9.53671157e-01 -5.39659381e-01 3.93447399e-01
1.26530695e+00 1.54406607e-01 2.79887225e-02 4.72297937e-01
1.03540272e-01 1.05060375e+00 6.46512926e-01 7.07269192e-01
1.05733347e+00 5.36525548e-01 3.62860799e-01 -4.26524818e-01
2.33467206e-01 -7.69778192e-02 5.83913028e-01 3.70838165e-01
-2.80040383e-01 -3.92542817e-02 -1.24084985e+00 8.24677587e-01
-1.65496004e+00 -1.16827512e+00 2.50869274e-01 2.54742026e+00
9.18973625e-01 1.03196412e-01 1.00652242e+00 -5.15109420e-01
6.67425275e-01 -1.68131858e-01 -5.48264265e-01 -8.84830356e-01
-1.45121023e-01 -1.32916778e-01 9.32497025e-01 8.34672332e-01
-6.23953402e-01 5.06290078e-01 7.30796146e+00 3.54052305e-01
-9.45512235e-01 1.99371547e-01 1.03332508e+00 -8.79125968e-02
-3.40749919e-01 -2.43024081e-01 -7.00693190e-01 5.68718433e-01
1.58600843e+00 -4.09870118e-01 7.99635112e-01 8.76072794e-02
1.08479106e+00 -2.39862591e-01 -4.83454376e-01 -4.20905463e-02
-3.80514413e-01 -1.23779845e+00 -6.64685369e-01 2.52799153e-01
1.03476036e+00 4.64679569e-01 2.50877813e-02 4.34521347e-01
9.90870416e-01 -8.91229987e-01 2.18510047e-01 2.87949681e-01
6.66053236e-01 -1.03801119e+00 6.25387907e-01 7.50035286e-01
-7.90630460e-01 -2.77178884e-01 1.75654918e-01 -5.25363564e-01
3.34746480e-01 3.91825020e-01 -8.19349587e-01 8.49909559e-02
5.51693738e-01 1.77441806e-01 -1.51687652e-01 7.67395079e-01
3.00561070e-01 7.69183159e-01 -7.55696476e-01 -2.32383907e-01
2.98069835e-01 -2.45898262e-01 5.07290602e-01 1.09041667e+00
2.93915719e-01 3.80257189e-01 3.21016788e-01 4.60936546e-01
1.92221552e-01 8.02099891e-03 -6.50580525e-01 5.32222427e-02
4.53954756e-01 5.54133892e-01 -7.71923184e-01 -3.86512607e-01
-2.09910870e-01 2.47152984e-01 -5.71614243e-02 6.76244915e-01
-6.45369411e-01 -1.56005368e-01 9.76104200e-01 4.67375875e-01
-2.04631835e-01 -4.73572686e-02 -1.63570091e-01 -8.89653921e-01
-5.28453052e-01 -1.11627626e+00 4.74100262e-01 -9.90088508e-02
-8.06753635e-01 -3.24642360e-01 6.58849478e-01 1.36568481e-02
-6.32350624e-01 -1.77170590e-01 -8.91186535e-01 9.81339872e-01
-1.38380098e+00 -6.88135028e-01 6.07608557e-01 2.60964751e-01
2.53735930e-01 3.64132106e-01 6.44546807e-01 3.27360690e-01
-8.78587484e-01 3.24824989e-01 5.59323251e-01 -2.66031712e-01
1.77448928e-01 -9.76401865e-01 2.12236285e-01 6.26826465e-01
-1.22908032e+00 4.69975054e-01 9.07356679e-01 -1.29703116e+00
-1.15620601e+00 -9.80334163e-01 9.14423227e-01 -7.65694082e-02
5.30208468e-01 -2.67456442e-01 -5.83534062e-01 5.81465483e-01
5.13300776e-01 -7.73395956e-01 5.65562546e-02 -1.45681158e-01
4.88655508e-01 1.66474894e-01 -1.84820890e+00 9.06377435e-01
7.76413679e-01 6.21875674e-02 -3.05868119e-01 6.43901289e-01
1.02062809e+00 1.08366199e-01 -9.07185912e-01 6.28434539e-01
4.55448121e-01 -6.61886394e-01 1.21012485e+00 -8.75933290e-01
4.14763726e-02 4.61142451e-01 2.21893236e-01 -1.64292395e+00
-1.17034009e-02 -6.72105372e-01 -1.55899093e-01 9.04414475e-01
2.54892945e-01 -1.32625616e+00 4.40885723e-01 7.07740605e-01
4.87939507e-01 -1.16632581e+00 -8.69689941e-01 -7.49778450e-01
5.59774578e-01 4.89737153e-01 1.12733090e+00 1.02954400e+00
-1.25420302e-01 -2.39287928e-01 -3.62526923e-01 3.10151517e-01
8.64157021e-01 -2.43390635e-01 4.06873167e-01 -7.61551797e-01
-2.85421401e-01 -4.85326648e-01 2.69814730e-01 -4.64325212e-02
-8.06161165e-02 -1.67753473e-01 -3.29330444e-01 -1.75359941e+00
4.21242267e-01 -6.15567923e-01 -1.23115644e-01 4.96417463e-01
-3.92452508e-01 -5.89068949e-01 5.81022128e-02 -2.85026580e-01
-2.65714508e-02 5.98007679e-01 7.25453913e-01 -5.38279228e-02
-6.19510710e-01 8.44843015e-02 -4.44572985e-01 5.18759668e-01
1.20665884e+00 -6.31934047e-01 -2.84741700e-01 -2.77078241e-01
1.93712592e-01 5.88948786e-01 3.61570269e-01 -4.77077574e-01
-3.28742713e-01 -9.79479015e-01 -2.11853206e-01 -2.91070074e-01
-1.09507374e-01 -3.95387441e-01 4.03025776e-01 1.66031861e+00
-1.23699673e-01 5.96366704e-01 1.80911914e-01 1.98266879e-01
6.09164000e-01 2.13353753e-01 7.85676301e-01 2.54072845e-02
2.96002150e-01 2.60899276e-01 -7.28244960e-01 4.28072959e-01
1.03504801e+00 2.43220925e-01 -6.35733247e-01 -3.80645007e-01
-3.70168328e-01 7.68462658e-01 8.25039446e-01 3.82096879e-02
1.88898087e-01 -8.32185090e-01 -1.36554468e+00 -2.40801275e-01
-5.63306212e-01 -4.34655100e-01 7.75384530e-02 8.53092849e-01
-7.40439892e-01 4.97429937e-01 -4.01907831e-01 -8.46062750e-02
-1.08033109e+00 7.07893968e-01 5.62831223e-01 -7.79482782e-01
-2.88687468e-01 1.74353331e-01 -1.56323006e-03 -7.97943532e-01
5.91032915e-02 -5.82332611e-02 -1.61407515e-02 8.11686814e-02
5.52453697e-01 8.86262774e-01 -4.99148935e-01 -3.94639641e-01
-4.96621460e-01 2.30358139e-01 4.55576003e-01 -3.44606489e-01
1.54428542e+00 -2.26593122e-01 -1.08034097e-01 -1.02428660e-01
7.19286442e-01 -2.42789194e-01 -1.43041706e+00 2.84008771e-01
-2.06953377e-01 1.32771078e-02 -2.66846903e-02 -8.67263913e-01
-5.98370016e-01 3.89201313e-01 7.90328026e-01 6.05561674e-01
1.24233651e+00 -5.23842394e-01 6.82758570e-01 8.64325278e-03
1.89040884e-01 -1.14056301e+00 -5.64061046e-01 6.67159632e-02
6.39267504e-01 -1.23365355e+00 1.30376637e-01 3.70102644e-01
-2.44447216e-01 7.41024017e-02 4.02508765e-01 -6.97935447e-02
9.15831685e-01 2.04532713e-01 -2.48056963e-01 -1.48447782e-01
-1.05405343e+00 -2.13688269e-01 -3.63799751e-01 8.70421827e-01
-9.72888544e-02 7.18727529e-01 -9.18459117e-01 9.32047889e-02
2.39083052e-01 6.22370504e-02 5.03139973e-01 8.01338375e-01
-5.39710462e-01 -8.40787470e-01 -7.69002438e-01 9.13397133e-01
-7.51450419e-01 -1.14322335e-01 -4.06771749e-02 8.04492116e-01
1.04649879e-01 1.06148970e+00 3.08791637e-01 3.53323787e-01
-5.36949150e-02 -2.73040146e-01 2.51788020e-01 -3.30170780e-01
-8.64073515e-01 -2.16684073e-01 1.26719564e-01 -1.27074178e-02
-4.52474773e-01 -8.37963343e-01 -1.29562080e+00 -1.05808425e+00
6.63714856e-02 1.09710708e-01 6.29277468e-01 7.77230561e-01
6.44955993e-01 3.07530850e-01 8.46230030e-01 -7.27115870e-01
-6.20694339e-01 -8.09889376e-01 -3.30175787e-01 1.19937174e-01
3.74433488e-01 -4.86383051e-01 -4.30745035e-01 -6.91639245e-01] | [6.006906986236572, 4.378661632537842] |
5e145b2d-4fa0-4201-8b1e-a9ba7fb6d0ca | hyperparameter-optimization-in-black-box | null | null | https://www.cs.princeton.edu/~fheide/proxyopt | https://www.cs.princeton.edu/~fheide/ProxyOpt.pdf | Hyperparameter Optimization in Black-box Image Processing using Differentiable Proxies | Nearly every commodity imaging system we directly interact with, or indirectly rely on, leverages power efficient, application-adjustable black-box hardware image signal processing (ISPs) units, running either in dedicated hardware blocks, or as proprietary software modules on programmable hardware. The configuration parameters of these black-box ISPs often have complex interactions with the output image, and must be adjusted prior to deployment according to application-specific quality and performance metrics. Today, this search is commonly performed manually by "golden eye" experts or algorithm developers leveraging domain expertise. We present a fully automatic system to optimize the parameters of black-box hardware and software image processing pipelines according to any arbitrary (i.e., application-specific) metric. We leverage a differentiable mapping between the configuration space and evaluation metrics, parameterized by a convolutional neural network that we train in an end-to-end fashion with imaging hardware in-the-loop. Unlike prior art, our differentiable proxies allow for high-dimension parameter search with stochastic first-order optimizers, without explicitly modeling any lower-level image processing transformations. As such, we can efficiently optimize black-box image processing pipelines for a variety of imaging applications, reducing application-specific configuration times from months to hours. Our optimization method is fully automatic, even with black-box hardware in the loop. We validate our method on experimental data for real-time display applications, object detection, and extreme low-light imaging. The proposed approach outperforms manual search qualitatively and quantitatively for all domain-specific applications tested. When applied to traditional denoisers, we demonstrate that—just by changing hyperparameters—traditional algorithms can outperform recent deep learning methods by a substantial margin on recent benchmarks. | ['Felix Heide', 'Jean-François Lalonde', 'Derek Nowrouzezahrai', 'Karl St. Arnaud', 'Fahim Mannan', 'Yuting Yang', 'Felix Yu', 'Ethan Tseng'] | 2019-04-01 | null | null | null | siggraph-2019-2019-4 | ['image-quality-estimation'] | ['computer-vision'] | [ 3.95948499e-01 -3.42966557e-01 -8.51691365e-02 -5.18050194e-01
-9.73567545e-01 -8.38893890e-01 1.67343348e-01 -1.57272846e-01
-6.40666723e-01 -1.47009417e-01 -2.26048186e-01 -5.89081049e-01
2.53446959e-03 -2.67347366e-01 -8.73265922e-01 -7.37148702e-01
-8.11390281e-02 2.33236864e-01 1.33685008e-01 -3.10389176e-02
2.32965797e-01 6.00885570e-01 -1.41491544e+00 5.53952111e-03
5.40654898e-01 1.27484143e+00 3.05480421e-01 1.16181052e+00
1.88338920e-01 5.15680492e-01 -5.40562212e-01 -2.09243998e-01
8.18748057e-01 4.59841788e-02 -1.95601866e-01 1.44438475e-01
4.77692008e-01 -4.85974908e-01 -7.30332136e-02 1.10118914e+00
7.44897664e-01 -4.71626520e-01 4.56988141e-02 -1.03985155e+00
-4.95068192e-01 6.31883502e-01 -7.11026669e-01 1.15239397e-01
-1.24447800e-01 1.26727796e+00 5.43831229e-01 -6.94382846e-01
3.04684520e-01 8.35025847e-01 6.77565932e-01 2.53118664e-01
-1.60630906e+00 -6.02042019e-01 -4.03584205e-02 -2.00184919e-02
-1.35187435e+00 -7.30756700e-01 5.95729768e-01 -4.71358865e-01
1.10582912e+00 1.46978900e-01 4.41089213e-01 8.69092882e-01
4.45127189e-01 3.59462917e-01 9.01238739e-01 -1.76764816e-01
4.54588950e-01 -1.85816318e-01 8.89588594e-02 7.45237648e-01
1.19163178e-01 -3.79817188e-02 -5.71222126e-01 -1.77582324e-01
9.08941746e-01 -2.72283703e-01 -4.26087260e-01 -5.55648327e-01
-1.42166483e+00 3.82169515e-01 3.09343398e-01 -1.95007101e-01
-3.13866824e-01 7.12396920e-01 4.23298061e-01 2.50610381e-01
-5.66198491e-02 7.37618268e-01 -8.80753398e-01 -2.67068595e-01
-1.13149679e+00 -2.49424785e-01 7.27165580e-01 9.62177396e-01
8.64033937e-01 2.45973200e-01 -3.33824277e-01 4.72474307e-01
1.50674105e-01 5.01651585e-01 4.43074554e-01 -1.24112868e+00
1.62281230e-01 4.12994117e-01 -1.45815033e-02 -5.43398321e-01
-7.49727905e-01 -6.78673804e-01 -8.41316223e-01 7.01423943e-01
2.98427194e-01 -3.47143590e-01 -1.07652080e+00 1.61359966e+00
2.43578136e-01 1.48007423e-01 -3.29264224e-01 1.26442790e+00
3.32595646e-01 5.00416934e-01 -4.87424396e-02 -5.67132607e-02
1.77415824e+00 -1.00725114e+00 -2.62803942e-01 -3.70523840e-01
2.32170433e-01 -9.69432116e-01 1.71142578e+00 7.62261152e-01
-1.31638312e+00 -4.94712263e-01 -1.40777981e+00 -2.50411958e-01
1.71757996e-01 6.10933721e-01 5.56813002e-01 7.81950057e-01
-1.46120238e+00 6.70090258e-01 -1.29471600e+00 -7.40326047e-02
3.99324238e-01 8.43243837e-01 1.21889748e-01 4.77493167e-01
-3.18865269e-01 4.90318686e-01 -1.29295304e-01 2.78319061e-01
-1.14214170e+00 -1.37801576e+00 -5.32528222e-01 3.58262748e-01
5.41487098e-01 -1.20265019e+00 1.41641438e+00 -1.03227615e+00
-2.15378571e+00 8.62789512e-01 2.35460758e-01 -5.43030679e-01
5.11884451e-01 -3.08419168e-01 -3.93234968e-01 1.03631601e-01
-2.66141236e-01 4.72821534e-01 1.48681605e+00 -8.30834329e-01
-2.20853597e-01 -2.03900233e-01 8.18957686e-02 -1.59521565e-01
-3.41061294e-01 2.55501121e-01 -1.01897931e+00 -4.92561638e-01
-7.33254701e-02 -1.03019512e+00 -4.40832257e-01 4.10977751e-01
-4.33548510e-01 5.73447227e-01 7.28168905e-01 -2.49805480e-01
1.06798339e+00 -2.31104326e+00 -1.04102559e-01 3.45013618e-01
3.12771946e-01 7.33523965e-02 -1.51653096e-01 -2.62714833e-01
-1.66493386e-01 2.71595977e-02 -2.65298635e-01 -2.62230039e-01
1.11250125e-01 -4.37619723e-02 -2.23983943e-01 6.39682472e-01
2.83829212e-01 8.01408470e-01 -6.31369948e-01 -3.44788700e-01
3.22066337e-01 4.04308319e-01 -8.84896457e-01 1.97089106e-01
-3.23253423e-01 5.06487668e-01 -2.59592116e-01 8.00819278e-01
5.63466370e-01 -7.95612395e-01 2.77269810e-01 -9.20221567e-01
-2.98365355e-01 3.75266410e-02 -1.23454261e+00 1.97041368e+00
-8.85058343e-01 8.25130582e-01 5.47522545e-01 -6.01000428e-01
3.72301668e-01 -1.88335657e-01 4.25619274e-01 -3.54583830e-01
2.47506604e-01 1.38311729e-01 9.82832983e-02 -3.57309401e-01
3.87143970e-01 5.69497705e-01 -3.31841521e-02 5.11462688e-01
6.10501282e-02 -2.84891903e-01 -4.99376543e-02 -7.76417255e-02
1.65614843e+00 1.51815400e-01 2.80876666e-01 -3.97434801e-01
2.36912295e-01 -1.34661272e-02 4.67644632e-01 9.76350009e-01
-4.62566428e-02 7.90268719e-01 6.03420734e-01 -3.95916343e-01
-1.12119412e+00 -1.05315900e+00 -2.98552096e-01 1.21251011e+00
1.06618434e-01 -4.82895374e-01 -8.88104498e-01 -1.90313205e-01
-3.33570361e-01 4.62583870e-01 -2.71095842e-01 7.95701742e-02
-7.98200071e-01 -8.97308409e-01 4.93420422e-01 3.09953213e-01
4.89899218e-01 -5.84124744e-01 -1.34337234e+00 2.68359095e-01
6.43964171e-01 -1.28372312e+00 -8.47480416e-01 4.57559377e-01
-7.47334719e-01 -8.61598074e-01 -2.69787610e-01 -2.45345190e-01
7.18906760e-01 1.71764866e-01 1.28876472e+00 3.22596543e-02
-8.18766832e-01 4.61412609e-01 1.68817669e-01 2.63168477e-02
-1.99198842e-01 3.34211253e-02 -5.88886328e-02 1.67837828e-01
-1.66492105e-01 -8.43234479e-01 -1.23343933e+00 3.77011240e-01
-9.58023369e-01 2.27371171e-01 9.17582631e-01 9.04923439e-01
8.97356570e-01 -1.20952882e-01 -2.64561474e-01 -6.56821489e-01
3.50972772e-01 -1.74373731e-01 -1.57276773e+00 1.97572708e-01
-6.94625854e-01 6.29746020e-01 7.04020381e-01 -6.17470503e-01
-7.73250699e-01 5.50283849e-01 1.02560163e-01 -8.50224495e-01
1.67820588e-01 2.91693747e-01 -3.84338737e-01 -6.71499610e-01
9.31976378e-01 -2.28765700e-02 -1.00054204e-01 -3.05650175e-01
4.15754139e-01 3.57405424e-01 1.09259665e+00 -6.41385138e-01
9.12624180e-01 5.35079718e-01 8.74123797e-02 -5.28721988e-01
-4.01458830e-01 -6.46047816e-02 -2.10269719e-01 -9.75295380e-02
9.53390539e-01 -9.16953087e-01 -1.21097505e+00 4.78365332e-01
-1.04411781e+00 -7.71127999e-01 -1.36751816e-01 1.87904686e-01
-4.77655739e-01 3.39504406e-02 -4.45907474e-01 -2.14677721e-01
-5.90442240e-01 -1.90882051e+00 1.54794908e+00 3.04072380e-01
5.72481100e-03 -4.10182595e-01 -4.82476890e-01 1.16963275e-01
8.22151542e-01 9.19030979e-02 8.11947346e-01 4.50510755e-02
-1.19432509e+00 3.68292294e-02 -2.38217428e-01 3.84883285e-01
-2.01761082e-01 4.22626406e-01 -1.02202821e+00 -2.15271622e-01
2.68199563e-01 1.11675143e-01 5.46242177e-01 6.37928247e-01
1.89691925e+00 -2.51467198e-01 -7.99712837e-02 1.56454551e+00
1.55479467e+00 -2.91448087e-01 5.20113647e-01 3.51105928e-01
5.57021797e-01 -1.18506327e-01 1.72536254e-01 6.66967094e-01
1.55496478e-01 9.70652997e-01 5.97600341e-01 -2.29893222e-01
-9.37627032e-02 2.57018834e-01 5.58933437e-01 4.40918088e-01
2.49893636e-01 4.87003922e-02 -1.04524386e+00 1.37866691e-01
-1.51865590e+00 -5.22466362e-01 -5.60283512e-02 2.34512329e+00
1.07815635e+00 2.39384234e-01 -2.54613847e-01 -4.24403101e-01
5.67481160e-01 -1.58723313e-02 -1.18639255e+00 -1.79504469e-01
-4.24788659e-03 5.88086307e-01 1.35404193e+00 2.78624684e-01
-1.12033081e+00 6.10539019e-01 5.72758579e+00 6.06185794e-01
-1.77514088e+00 2.27243841e-01 7.51516879e-01 -8.10136795e-01
-2.38717012e-02 2.47937188e-01 -5.90096533e-01 4.97008890e-01
1.06107378e+00 -1.59438297e-01 9.27387357e-01 1.09224200e+00
5.25615573e-01 -1.04718497e-02 -1.46887505e+00 1.46239460e+00
-2.39064157e-01 -1.59057879e+00 -5.63070655e-01 2.66188718e-02
4.09182400e-01 2.75556296e-01 3.68968785e-01 2.19503976e-03
3.55085105e-01 -8.58779848e-01 8.95450056e-01 2.65080869e-01
1.05290985e+00 -3.81511301e-01 2.12741330e-01 -3.78715396e-02
-8.88228536e-01 -7.15232491e-02 -1.30167589e-01 3.80790293e-01
3.67184043e-01 7.65803039e-01 -6.38566673e-01 2.77462844e-02
1.00827897e+00 3.75957764e-03 -7.56754756e-01 9.10237432e-01
-2.32791096e-01 7.25746572e-01 -5.57204962e-01 1.97003976e-01
3.42489523e-03 -8.86995420e-02 4.97178137e-01 1.16723037e+00
3.96300375e-01 1.05435975e-01 -9.97876823e-02 8.97805214e-01
-2.25890279e-01 -2.56770194e-01 6.70833439e-02 2.31505677e-01
2.89067626e-01 1.77461040e+00 -6.53909564e-01 -1.41875356e-01
-3.04389149e-01 9.29017425e-01 -2.88835466e-02 3.36076081e-01
-1.22055054e+00 -3.98899198e-01 1.22905755e+00 2.23442391e-01
2.73842841e-01 -4.30251777e-01 -7.52206266e-01 -1.14861774e+00
2.67739505e-01 -1.09958589e+00 -3.55098732e-02 -8.93598199e-01
-1.10066700e+00 5.17248869e-01 -1.53416932e-01 -1.25401926e+00
-2.55506158e-01 -7.82302439e-01 -5.96787930e-01 7.66660452e-01
-1.37225974e+00 -8.30243707e-01 -6.05822146e-01 5.84597826e-01
3.40798974e-01 -1.52099252e-01 5.29791594e-01 4.79976714e-01
-7.68590271e-01 7.89202094e-01 3.55370380e-02 -1.00933366e-01
8.00800383e-01 -1.10163236e+00 6.78285539e-01 1.08343589e+00
-7.65411109e-02 6.65656149e-01 8.84969234e-01 -1.95672646e-01
-2.37391138e+00 -8.80695879e-01 -2.77702928e-01 -1.81433275e-01
1.01656616e+00 -6.23508453e-01 -7.28237629e-01 4.04289365e-01
4.10584390e-01 5.44038177e-01 5.30879140e-01 -1.59647673e-01
-4.02673066e-01 -5.43649852e-01 -9.84759271e-01 8.95469785e-01
1.05841982e+00 -4.26556319e-01 1.91711232e-01 6.75799191e-01
8.21877241e-01 -9.29472089e-01 -7.29948997e-01 2.92484790e-01
5.21788895e-01 -8.64379168e-01 1.13842785e+00 -2.59865791e-01
3.49234909e-01 -8.18402171e-01 -3.06885969e-03 -9.94812727e-01
-1.72850043e-01 -1.35062635e+00 -1.14598788e-01 7.11650789e-01
3.87803495e-01 -4.07518566e-01 7.24042892e-01 8.72642100e-01
-4.80291754e-01 -6.34302437e-01 -8.18911970e-01 -6.27460957e-01
-5.62574923e-01 -8.04014266e-01 7.78842032e-01 4.09156680e-01
-6.26849115e-01 2.18954608e-01 -3.73366028e-02 8.45863938e-01
8.95841300e-01 9.35259908e-02 9.54767108e-01 -4.60482031e-01
-1.12673163e+00 -8.87274981e-01 -4.23444748e-01 -1.11514843e+00
-1.44318298e-01 -4.02311772e-01 1.56452850e-01 -6.78388715e-01
-2.18376800e-01 -4.20135677e-01 -1.71312660e-01 5.53697288e-01
-4.30710055e-02 1.76824927e-01 1.94400981e-01 2.27469668e-01
-4.69681531e-01 1.36477500e-01 6.83922350e-01 -1.88213602e-01
-2.48573408e-01 -2.38472223e-01 -7.31877327e-01 8.67463112e-01
5.98293245e-01 -2.97136337e-01 -2.84118533e-01 -1.09229076e+00
3.13374847e-01 -3.29608135e-02 7.05266535e-01 -1.34152997e+00
6.75999880e-01 -1.60655305e-01 2.07452253e-01 -1.07307658e-01
2.59465635e-01 -8.38330090e-01 4.50357020e-01 2.07423314e-01
-2.30353668e-01 1.61423966e-01 3.30377221e-01 1.14878096e-01
2.12673247e-01 -1.93908112e-03 1.12059879e+00 2.71579206e-01
-7.59092212e-01 3.29324901e-01 -3.61181200e-02 -9.44549814e-02
8.92037570e-01 -8.82800072e-02 -5.19864202e-01 -1.87825501e-01
-5.19974411e-01 9.70802233e-02 6.74570918e-01 6.27595782e-02
2.90426582e-01 -8.35502565e-01 -5.39632142e-01 3.29979718e-01
8.82568136e-02 -2.65174750e-02 3.28684211e-01 9.44802642e-01
-1.07514572e+00 -8.40573236e-02 -3.52270193e-02 -1.10010934e+00
-9.80349481e-01 5.43921590e-01 5.98638892e-01 -1.86686158e-01
-6.19456708e-01 6.83597863e-01 2.88165063e-01 -2.88011134e-02
1.58161685e-01 -6.22934937e-01 6.63506508e-01 -5.31489611e-01
5.40275037e-01 1.16279824e-02 5.15183926e-01 1.30159199e-01
-4.08390462e-01 6.23233080e-01 2.52380073e-01 -1.68107554e-01
1.44961417e+00 4.08353731e-02 -1.38330027e-01 -7.82018825e-02
1.17330837e+00 -6.21108226e-02 -1.66215730e+00 -6.95044026e-02
-1.99041545e-01 -4.42107171e-01 6.51207328e-01 -8.61316264e-01
-1.52248836e+00 6.83828115e-01 1.02955675e+00 -1.53635889e-01
1.73886263e+00 -1.81212157e-01 6.03879929e-01 4.34297860e-01
5.41986167e-01 -1.06359053e+00 -8.32001418e-02 2.96510896e-03
7.63935447e-01 -1.21874070e+00 1.32181540e-01 -2.03318238e-01
-3.56449395e-01 1.23302972e+00 5.40966511e-01 -3.48819196e-02
7.75428832e-01 1.10439503e+00 1.76063269e-01 -7.03686252e-02
-9.06240463e-01 2.86419511e-01 5.50955944e-02 3.65089953e-01
2.19710812e-01 -2.11916361e-02 1.75234869e-01 5.43786705e-01
-2.52279341e-01 4.79330309e-02 5.44668555e-01 8.33779871e-01
-8.20024982e-02 -1.02007651e+00 -5.16153097e-01 5.27068317e-01
-3.67882639e-01 -3.53614986e-01 3.42139989e-01 4.69162464e-01
-7.14632496e-02 6.44506752e-01 1.18517101e-01 -2.94745922e-01
2.64934361e-01 -3.92986089e-01 3.66568804e-01 -4.03989315e-01
-1.00160336e+00 1.19915210e-01 -2.41409257e-01 -1.24118710e+00
1.94941908e-01 -6.12784207e-01 -1.10010302e+00 -2.65050262e-01
8.56454857e-03 -6.73408508e-01 1.24629521e+00 6.24918520e-01
9.80656207e-01 6.48687363e-01 5.11459768e-01 -1.21585548e+00
-8.87155175e-01 -3.60168755e-01 -8.19384083e-02 -1.16726220e-01
3.75436008e-01 -2.93198377e-01 -1.78273723e-01 1.89649388e-01] | [10.44153118133545, -1.9796040058135986] |
52c5c5e8-d8d7-4ca1-bae3-f35bacc4d01a | representation-learning-with-function-call | 2205.06918 | null | https://arxiv.org/abs/2205.06918v3 | https://arxiv.org/pdf/2205.06918v3.pdf | Representation learning with function call graph transformations for malware open set recognition | Open set recognition (OSR) problem has been a challenge in many machine learning (ML) applications, such as security. As new/unknown malware families occur regularly, it is difficult to exhaust samples that cover all the classes for the training process in ML systems. An advanced malware classification system should classify the known classes correctly while sensitive to the unknown class. In this paper, we introduce a self-supervised pre-training approach for the OSR problem in malware classification. We propose two transformations for the function call graph (FCG) based malware representations to facilitate the pretext task. Also, we present a statistical thresholding approach to find the optimal threshold for the unknown class. Moreover, the experiment results indicate that our proposed pre-training process can improve different performances of different downstream loss functions for the OSR problem. | ['Philip K. Chan', 'Jingyun Jia'] | 2022-05-13 | null | null | null | null | ['open-set-learning'] | ['miscellaneous'] | [ 3.91009241e-01 -3.70094895e-01 -3.92979562e-01 -2.31989041e-01
-5.11201084e-01 -6.26885772e-01 4.52740759e-01 1.99618444e-01
-1.10689022e-01 6.01970673e-01 -3.79544437e-01 -7.25311100e-01
-9.31658968e-02 -8.39379072e-01 -4.58644658e-01 -5.85114717e-01
-2.16207325e-01 4.19326693e-01 1.68302104e-01 -2.86084354e-01
2.79968858e-01 6.31929874e-01 -1.45062220e+00 4.55155939e-01
1.01258087e+00 1.06771672e+00 1.23685926e-01 6.36515796e-01
-2.85205424e-01 5.63847303e-01 -8.90580714e-01 -3.54294360e-01
4.27051753e-01 -2.44848177e-01 -7.35222638e-01 -1.14883423e-01
4.80210148e-02 -3.07935998e-02 -4.36998218e-01 1.43373656e+00
1.04319289e-01 1.99516222e-01 9.78785634e-01 -1.60799241e+00
-5.45215189e-01 5.80509901e-01 -5.57333171e-01 5.02842784e-01
1.27403826e-01 2.07669944e-01 9.16689575e-01 -4.77364242e-01
4.38357532e-01 1.25742841e+00 5.93616307e-01 8.03431630e-01
-8.58962178e-01 -8.42862606e-01 1.20860934e-01 3.59620303e-01
-1.36052990e+00 -1.67993352e-01 6.76222026e-01 -4.90858406e-01
7.47215211e-01 4.88542676e-01 7.39088431e-02 1.29632521e+00
2.19322339e-01 5.04505992e-01 1.01836348e+00 -6.64612651e-02
4.59572896e-02 3.78313184e-01 6.97408080e-01 7.28355467e-01
5.47220409e-01 1.05853803e-01 1.04688190e-01 -4.94245112e-01
3.05265725e-01 3.95435601e-01 -3.93825084e-01 8.63158703e-02
-6.00574970e-01 1.02874947e+00 3.88995081e-01 4.54063743e-01
2.33658999e-01 -1.69390533e-02 8.76009464e-01 5.53612709e-01
6.01413131e-01 4.43223983e-01 -5.45951605e-01 1.40410274e-01
-5.66284299e-01 -4.01260018e-01 8.63994241e-01 5.42342663e-01
6.74942493e-01 1.13706253e-01 -1.71955645e-01 7.80877292e-01
3.60833704e-01 3.15202057e-01 6.93613708e-01 -1.58303857e-01
5.52618980e-01 6.51443362e-01 -5.67305446e-01 -1.03569138e+00
1.31296650e-01 -6.18938744e-01 -8.29907238e-01 -1.56907648e-01
4.25201833e-01 9.11138058e-02 -8.25308502e-01 1.30439103e+00
2.86162466e-01 6.14578784e-01 1.42012775e-01 3.80344033e-01
6.84940696e-01 1.07993400e+00 -2.06535801e-01 -4.75237191e-01
1.27084076e+00 -7.73535788e-01 -5.16093254e-01 -1.54013082e-01
8.09567273e-01 -4.26897049e-01 1.07891405e+00 4.33207661e-01
5.87925240e-02 -3.18157673e-01 -1.04307783e+00 3.80532295e-01
-5.10747731e-01 8.20866600e-02 5.37784696e-01 9.26197886e-01
-4.96308357e-01 6.27139390e-01 -5.72266698e-01 -3.11264217e-01
5.28406024e-01 3.92093450e-01 -3.48770738e-01 -2.03728259e-01
-9.05380249e-01 3.58836561e-01 7.33433664e-01 -1.35531679e-01
-1.25101519e+00 -3.81026536e-01 -7.35016644e-01 2.54525840e-01
6.35691166e-01 -1.66197866e-02 7.19448090e-01 -1.01875246e+00
-1.19708490e+00 7.31097221e-01 1.23855524e-01 -5.61879039e-01
3.08450282e-01 9.64427516e-02 -4.29786354e-01 -3.67899127e-02
-2.49917254e-01 4.18711938e-02 1.38636065e+00 -1.27291763e+00
-3.79897565e-01 -5.83778083e-01 -1.92564260e-02 -2.30229214e-01
-6.93042576e-01 2.06402257e-01 4.01684716e-02 -6.69523120e-01
-1.55252129e-01 -6.80276096e-01 -1.93624780e-01 -6.04033113e-01
-5.97263515e-01 -3.91785920e-01 1.50138748e+00 -8.37080359e-01
1.33758080e+00 -2.34524965e+00 -1.56891625e-02 4.07175988e-01
2.75621623e-01 4.44840223e-01 -1.82475954e-01 6.96932198e-04
-2.97213286e-01 3.02098304e-01 -4.75399733e-01 -8.93877298e-02
-2.73436934e-01 2.26983890e-01 -8.51361096e-01 6.46817029e-01
2.70986468e-01 4.35198605e-01 -7.17185676e-01 -3.55526358e-01
-7.76460096e-02 3.39573354e-01 -3.82355779e-01 4.50611740e-01
-5.51464140e-01 6.08495235e-01 -4.35690016e-01 8.45099628e-01
5.29317498e-01 -2.03430817e-01 8.44517648e-02 8.67271349e-02
4.23876762e-01 1.42882377e-01 -8.35501373e-01 7.43626058e-01
-4.51213628e-01 4.59356517e-01 -1.69770923e-02 -1.18162906e+00
1.33952594e+00 1.98174976e-02 3.27253640e-01 1.15458913e-01
6.27086580e-01 3.46885353e-01 2.84262806e-01 -3.47035348e-01
3.23680729e-01 2.94004325e-02 -3.01950984e-02 2.58975416e-01
1.39206033e-02 3.06829870e-01 -5.35319299e-02 -2.36362219e-02
1.19951177e+00 -4.21440989e-01 4.24537927e-01 -1.42765433e-01
1.03475392e+00 -6.76177070e-02 7.72349000e-01 4.67763543e-01
-3.85093033e-01 1.66616946e-01 6.90645933e-01 -3.73425841e-01
-6.61266625e-01 -5.47363460e-01 -1.27425283e-01 8.79354000e-01
6.31628335e-02 -3.74896973e-01 -7.73037732e-01 -1.43431640e+00
5.39389811e-02 5.98004341e-01 -3.98496032e-01 -4.17078435e-01
-5.76753020e-01 -8.78707767e-01 7.47481048e-01 1.34135619e-01
3.02742600e-01 -8.71585190e-01 -7.79707134e-02 -2.57664919e-01
-1.31761238e-01 -1.01674974e+00 -4.63710666e-01 3.69150698e-01
-7.87311137e-01 -1.43511105e+00 -2.80580819e-01 -8.34448636e-01
9.57978010e-01 2.89543033e-01 4.24707204e-01 3.93954962e-01
-3.67973089e-01 9.05272141e-02 -6.83915138e-01 -1.30870983e-01
-9.55463588e-01 1.25995472e-01 1.73305959e-01 3.76665473e-01
3.05446714e-01 -1.92298383e-01 4.18406352e-02 4.39293236e-01
-8.93823981e-01 -6.55542970e-01 2.81452537e-01 8.25350940e-01
5.68908870e-01 5.98611712e-01 6.63507998e-01 -1.05531299e+00
7.33834445e-01 -9.14139271e-01 -7.86636889e-01 3.75991225e-01
-4.37532276e-01 1.15840219e-01 1.33273017e+00 -9.05890107e-01
-7.83789456e-01 -8.38912576e-02 -2.51306146e-01 -8.73042047e-01
3.38534974e-02 1.00315057e-01 -5.11748910e-01 -2.34532818e-01
6.46868408e-01 3.28541338e-01 1.11851115e-02 -3.95271868e-01
-4.31407131e-02 1.09900165e+00 1.18477166e-01 -4.61413056e-01
1.11742294e+00 2.39614725e-01 1.59749109e-02 -1.00501359e+00
-6.35549486e-01 -6.88527346e-01 -1.96925446e-01 1.42232865e-01
6.54023290e-01 -4.29660589e-01 -7.07966030e-01 4.58292633e-01
-1.01874757e+00 -9.18679312e-02 -1.79667681e-01 1.43658042e-01
-1.95227146e-01 7.94023454e-01 -4.38978076e-01 -8.48828018e-01
-7.29637861e-01 -1.59241116e+00 8.47238183e-01 9.16600227e-02
2.39973396e-01 -9.40485537e-01 -1.30622953e-01 3.19076777e-01
1.24681689e-01 6.71590865e-02 1.04294980e+00 -1.58782995e+00
-3.45754206e-01 -3.75020355e-01 -1.60167933e-01 7.68070400e-01
2.93157548e-01 1.14731647e-01 -9.70415711e-01 -6.83391571e-01
3.58366311e-01 -2.73483127e-01 1.09520948e+00 -8.17574561e-02
1.76124048e+00 -6.14452481e-01 -6.41899824e-01 8.16479981e-01
1.32096314e+00 4.06825662e-01 4.87415582e-01 -3.56099159e-02
9.01383936e-01 4.06402946e-01 5.97359836e-01 3.52943003e-01
-3.15460443e-01 2.84361601e-01 3.88032913e-01 5.24285674e-01
2.08908871e-01 -2.24364534e-01 8.48873079e-01 7.85582781e-01
4.52357531e-01 -4.20874119e-01 -8.31900477e-01 -2.37840321e-02
-1.44961345e+00 -8.60014379e-01 -8.72082114e-02 2.40456939e+00
6.60106897e-01 1.59095019e-01 -1.77378088e-01 4.28317636e-01
1.06241286e+00 1.52917534e-01 -5.20223558e-01 -4.87374634e-01
2.30088141e-02 1.85969770e-01 5.33981204e-01 4.28062767e-01
-1.33806443e+00 9.15510595e-01 5.66255903e+00 1.47886133e+00
-1.23264515e+00 3.48508984e-01 9.42850769e-01 5.58207750e-01
-1.53374687e-01 7.39249587e-02 -1.20883179e+00 5.70926189e-01
8.83488297e-01 -3.55437696e-02 6.64143026e-01 1.07832360e+00
-2.04559147e-01 2.70914108e-01 -1.08902347e+00 1.04925358e+00
2.80980408e-01 -9.74094093e-01 8.04628879e-02 2.14095935e-01
5.25102317e-01 -1.32440493e-01 1.26554176e-01 6.64962590e-01
1.18583933e-01 -1.08818233e+00 3.05720642e-02 -5.91498986e-02
9.22359586e-01 -7.95747876e-01 6.67073667e-01 6.58667505e-01
-1.29838455e+00 -6.95520461e-01 -3.85959059e-01 2.79509097e-01
-5.17640293e-01 4.75937307e-01 -1.26599145e+00 4.70592529e-01
3.29337746e-01 5.69790542e-01 -7.21108377e-01 9.13134933e-01
-1.19896486e-01 8.62917840e-01 -1.84717104e-01 -1.47498697e-01
1.98137332e-02 -3.01761001e-01 8.47315729e-01 9.17507827e-01
8.49406272e-02 -1.42438889e-01 4.24515784e-01 5.95518053e-01
-2.11636022e-01 1.84793234e-01 -9.52946723e-01 -4.96893227e-01
4.12752092e-01 1.23791635e+00 -8.97973239e-01 -1.65968016e-01
2.57858168e-02 8.27511787e-01 4.21042055e-01 -7.40534440e-02
-1.02802813e+00 -5.85332990e-01 7.06957459e-01 1.14160828e-01
1.07913926e-01 -1.95681155e-01 -7.78708681e-02 -1.30131042e+00
-2.38658652e-01 -1.16732681e+00 8.84986937e-01 4.40379828e-02
-1.52240098e+00 7.81092227e-01 -2.58222967e-01 -1.17810500e+00
8.09478387e-02 -9.27855313e-01 -8.70331824e-01 2.52508372e-01
-1.37999809e+00 -8.25183749e-01 7.37014040e-02 4.71140951e-01
5.62059045e-01 -5.32233417e-01 5.41211426e-01 3.96894455e-01
-1.03854108e+00 9.49383080e-01 4.31412667e-01 3.45280081e-01
3.27681720e-01 -7.44721234e-01 9.40407142e-02 1.03293073e+00
5.78387007e-02 5.69029391e-01 3.72572809e-01 -1.04240155e+00
-1.55108941e+00 -1.56997180e+00 2.45268062e-01 -3.05152059e-01
7.33434379e-01 -4.59348828e-01 -1.09497547e+00 7.98253179e-01
-4.65196699e-01 2.26452589e-01 6.03046834e-01 -4.76095714e-02
-7.89274156e-01 -2.02099010e-01 -1.50899029e+00 1.14694424e-01
7.88632452e-01 -4.97757733e-01 -3.63020152e-01 5.03848851e-01
1.07092059e+00 -4.15864028e-02 -5.68840265e-01 4.71971780e-01
-8.89055878e-02 -3.68330657e-01 8.88528168e-01 -8.80488753e-01
1.20254792e-01 -2.56126732e-01 -4.07257587e-01 -1.22099388e+00
1.03049681e-01 -6.37657940e-01 -4.46570784e-01 1.24672246e+00
1.11640282e-01 -1.03581595e+00 5.69182575e-01 -2.09806174e-01
-7.07854256e-02 -7.75280476e-01 -1.01826274e+00 -9.85311866e-01
-1.24436475e-01 -3.66800934e-01 4.59039062e-01 1.05025697e+00
-1.26778498e-01 3.47555131e-01 -2.07172647e-01 3.77184212e-01
6.42651081e-01 -8.03646166e-03 5.49795091e-01 -1.50353074e+00
-5.41057527e-01 -3.41431230e-01 -6.59797788e-01 -5.37371814e-01
6.57014787e-01 -1.35645568e+00 -1.70390248e-01 -9.47784066e-01
2.70770550e-01 -7.90328324e-01 -2.62368083e-01 4.69174802e-01
-2.61876434e-01 -1.32994309e-01 3.48117016e-02 3.93416256e-01
-4.81323749e-01 4.83117223e-01 9.27995265e-01 -3.74859482e-01
-2.55197942e-01 2.86692291e-01 -5.37013590e-01 5.55547655e-01
1.08068454e+00 -6.82237566e-01 -3.76200020e-01 1.02778986e-01
-2.30837777e-01 -1.47029340e-01 1.52482852e-01 -8.82175505e-01
-3.20210189e-01 -3.10359180e-01 1.27328679e-01 -5.00729918e-01
2.44150728e-01 -8.75297248e-01 -9.83545482e-02 9.06117201e-01
-8.63767788e-02 -2.40629718e-01 1.28391296e-01 8.31717014e-01
8.87241587e-02 -6.02478743e-01 1.02933252e+00 6.68950602e-02
-4.20309395e-01 5.57837009e-01 -2.22398236e-01 2.04765111e-01
1.42467296e+00 -7.09160343e-02 -5.63381374e-01 7.18332455e-02
-2.06587687e-01 2.12490052e-01 2.70503938e-01 4.69137669e-01
8.49082351e-01 -9.09464359e-01 -5.48533440e-01 3.92733485e-01
1.72251657e-01 -1.47562429e-01 -1.46720499e-01 5.25634885e-01
-5.02834499e-01 3.41788977e-01 -5.48156463e-02 -4.57577467e-01
-1.55179667e+00 9.10889864e-01 3.49513739e-01 -5.94107509e-01
-4.02852178e-01 7.93394327e-01 1.25117242e-01 -7.40341902e-01
4.78354424e-01 -6.99515939e-02 -4.60995078e-01 -1.83148682e-01
6.36178493e-01 5.22083759e-01 5.41679449e-02 -6.87938690e-01
-2.58401245e-01 -3.39386463e-02 -2.53000468e-01 5.93609333e-01
1.11729920e+00 3.91932070e-01 -3.43991488e-01 2.47728482e-01
1.71045125e+00 -9.78792906e-02 -6.75616205e-01 -1.00893103e-01
8.51211771e-02 -6.10693574e-01 -1.30565360e-01 -3.42988044e-01
-9.70271647e-01 1.00554276e+00 6.85599983e-01 4.22911525e-01
1.02419484e+00 -9.89846289e-02 8.72703671e-01 5.52908480e-01
3.84579688e-01 -5.92688441e-01 1.24902599e-01 7.94795752e-01
4.84401733e-01 -1.14292109e+00 -2.88654059e-01 -7.63878286e-01
-3.95011067e-01 1.20976973e+00 8.61663282e-01 -6.93157464e-02
6.51106656e-01 2.79749781e-01 -5.58131337e-01 -2.52252053e-02
-7.31468499e-01 7.21686482e-02 1.44339696e-01 6.68967068e-01
-1.27028320e-02 2.58852959e-01 -1.84478521e-01 7.13416100e-01
6.68870434e-02 -5.97733855e-01 4.43807811e-01 7.88572669e-01
-6.27995372e-01 -1.10341120e+00 -8.07880819e-01 9.63354409e-01
-4.71385926e-01 4.06370610e-02 -6.50019407e-01 2.54408360e-01
7.70723969e-02 9.04398978e-01 -3.12972208e-03 -9.51568425e-01
-1.95495993e-01 1.79582641e-01 3.51399720e-01 -7.79807329e-01
-6.14072859e-01 -3.72511417e-01 -5.38666807e-02 -1.34716496e-01
4.83556986e-01 -1.67288229e-01 -1.44083977e+00 -1.79384798e-01
-8.41980994e-01 2.23367915e-01 5.37666917e-01 8.17746997e-01
2.49220639e-01 5.67204595e-01 1.29157269e+00 -2.47944683e-01
-1.28451467e+00 -7.54340291e-01 -3.36199850e-01 3.16231519e-01
1.26773551e-01 -6.74357414e-01 -7.10929155e-01 -2.49170825e-01] | [14.34919261932373, 9.598189353942871] |
319c6e61-e166-4687-82c6-8f35e6a0cb38 | only-train-once-mr-fingerprinting-for | 2206.04383 | null | https://arxiv.org/abs/2206.04383v1 | https://arxiv.org/pdf/2206.04383v1.pdf | Only-Train-Once MR Fingerprinting for Magnetization Transfer Contrast Quantification | Magnetization transfer contrast magnetic resonance fingerprinting (MTC-MRF) is a novel quantitative imaging technique that simultaneously measures several tissue parameters of semisolid macromolecule and free bulk water. In this study, we propose an Only-Train-Once MR fingerprinting (OTOM) framework that estimates the free bulk water and MTC tissue parameters from MR fingerprints regardless of MRF schedule, thereby avoiding time-consuming process such as generation of training dataset and network training according to each MRF schedule. A recurrent neural network is designed to cope with two types of variants of MRF schedules: 1) various lengths and 2) various patterns. Experiments on digital phantoms and in vivo data demonstrate that our approach can achieve accurate quantification for the water and MTC parameters with multiple MRF schedules. Moreover, the proposed method is in excellent agreement with the conventional deep learning and fitting methods. The flexible OTOM framework could be an efficient tissue quantification tool for various MRF protocols. | ['HyunWook Park', 'Hye-Young Heo', 'Beomgu Kang'] | 2022-06-09 | null | null | null | null | ['magnetic-resonance-fingerprinting'] | ['medical'] | [ 5.56365490e-01 -1.97365195e-01 -1.66859403e-01 -3.21921170e-01
-7.24624097e-01 -2.35381871e-01 4.28130955e-01 5.34203202e-02
-6.42895937e-01 6.96246862e-01 6.47077709e-02 -2.10899189e-01
-4.39758539e-01 -3.39841008e-01 -4.39085662e-01 -1.12937748e+00
-3.54242951e-01 6.57720327e-01 3.59425068e-01 2.28959113e-01
1.05037384e-01 6.74925685e-01 -1.09054399e+00 1.63326502e-01
7.49266922e-01 1.09446394e+00 7.61253178e-01 5.75031400e-01
-5.61895184e-02 1.16227734e+00 -3.02435488e-01 2.13192120e-01
5.35705760e-02 -2.68661290e-01 -8.43877733e-01 -2.99556881e-01
4.53280747e-01 -4.65004086e-01 -5.75363398e-01 1.00024652e+00
9.16616440e-01 2.58691043e-01 7.88076520e-01 -5.80629230e-01
-2.80192226e-01 9.64340150e-01 -5.22760391e-01 3.20647836e-01
2.44530868e-02 -2.33537644e-01 -1.62124075e-02 -5.26543319e-01
6.30750358e-01 6.23478830e-01 1.02026606e+00 6.75638258e-01
-1.13583148e+00 -4.61587042e-01 -3.55195343e-01 6.30555972e-02
-1.14694154e+00 -4.00960416e-01 5.39857268e-01 -6.78420722e-01
7.86625922e-01 3.47406596e-01 4.54670131e-01 7.97107935e-01
9.28435206e-01 4.12236065e-01 1.58022261e+00 -3.34432095e-01
5.97266182e-02 -1.80903032e-01 1.09814681e-01 8.33580196e-01
-1.15575083e-02 1.15510806e-01 -1.26206368e-01 -4.50932592e-01
9.41502035e-01 3.13296989e-02 -5.47933817e-01 -3.85110319e-01
-1.50681210e+00 4.62294400e-01 1.56964049e-01 8.14183950e-01
-4.89049047e-01 5.82867265e-01 4.91864800e-01 7.96840340e-02
2.94271886e-01 3.38663280e-01 -2.25587606e-01 2.79850602e-01
-1.31166947e+00 7.40307868e-02 5.33486962e-01 5.47780335e-01
3.24273676e-01 2.23780259e-01 -2.93213874e-01 1.07160807e+00
2.65042156e-01 7.73578167e-01 1.03175712e+00 -1.11817276e+00
-4.54561040e-02 -2.38126710e-01 2.47674987e-01 -8.53112817e-01
-9.41395938e-01 -3.23488295e-01 -1.05887234e+00 -1.76996708e-01
4.76876587e-01 -3.87577824e-02 -1.01314497e+00 1.56434989e+00
4.40019429e-01 3.11777741e-01 -3.54910523e-01 1.08252335e+00
9.13184226e-01 3.51051658e-01 1.74237564e-01 -4.82422203e-01
1.23486781e+00 -8.55478168e-01 -8.45731974e-01 2.43719935e-01
6.30343974e-01 -6.15599275e-01 5.26366115e-01 1.25285253e-01
-8.50716174e-01 -2.24241719e-01 -8.53935063e-01 2.55901307e-01
-1.35703146e-01 1.77327124e-03 9.13063884e-01 8.82399321e-01
-1.05899239e+00 1.12228727e+00 -1.17166471e+00 1.06549434e-01
-8.47038627e-02 6.83607817e-01 -5.37815869e-01 -6.35455474e-02
-1.26455402e+00 9.62782323e-01 3.00123036e-01 5.88082433e-01
-9.78202999e-01 -1.07502210e+00 -6.90793812e-01 -3.48053992e-01
-5.63115329e-02 -5.75384617e-01 1.20619881e+00 -4.32352513e-01
-1.81673658e+00 7.83500552e-01 -9.63827670e-02 -4.61248189e-01
6.35625184e-01 1.29390389e-01 -4.91264671e-01 3.60268116e-01
4.86961603e-02 5.00561237e-01 6.74237847e-01 -9.76744533e-01
2.25275680e-01 -3.36122394e-01 -4.85767186e-01 -1.33780017e-01
3.00871432e-01 1.06311396e-01 2.51050562e-01 -4.05653894e-01
4.28921282e-01 -8.90930831e-01 -4.56118613e-01 -3.95514928e-02
-3.69841784e-01 3.95208061e-01 2.59059221e-01 -7.66014993e-01
7.87280560e-01 -1.62350523e+00 -1.93059757e-01 3.26498866e-01
6.09598517e-01 1.52236983e-01 -2.03587338e-01 8.11474696e-02
-2.69431204e-01 -2.06968695e-01 -2.80469894e-01 4.06834409e-02
-1.32927015e-01 -1.39662206e-01 9.24958363e-02 9.60067272e-01
-3.72727334e-01 8.43968153e-01 -1.09721518e+00 -6.71580076e-01
2.38397077e-01 6.34023726e-01 8.46007653e-03 -4.46731858e-02
1.75590128e-01 9.42882061e-01 -5.59901297e-01 6.61031008e-01
9.79712009e-01 -2.85424292e-01 6.09121919e-01 -4.48126018e-01
-2.48420879e-01 -2.70736486e-01 -7.15161920e-01 1.64001596e+00
-3.58921826e-01 3.29771072e-01 7.36269057e-02 -9.14276600e-01
9.14619505e-01 5.56670606e-01 1.20813894e+00 -9.80297565e-01
2.42430896e-01 5.96833169e-01 8.51402506e-02 -7.36873984e-01
3.29982191e-01 -5.98309457e-01 2.15455860e-01 5.76453865e-01
2.75001496e-01 2.18086407e-01 -2.30267104e-02 -4.37787414e-01
1.08565223e+00 1.65025428e-01 4.82212566e-02 -8.45845401e-01
7.02567935e-01 -5.84382892e-01 4.71843064e-01 9.97044981e-01
-5.91104209e-01 4.85074788e-01 1.61323428e-01 -6.95812583e-01
-9.77632344e-01 -9.23811674e-01 -7.00331807e-01 7.60659873e-01
1.03596643e-01 2.24059671e-01 -8.98160040e-01 -2.40165830e-01
-4.42859717e-02 5.02663068e-02 -9.19396996e-01 2.96486497e-01
-7.64367223e-01 -1.14764893e+00 6.33212447e-01 2.49071404e-01
3.20025295e-01 -9.20916021e-01 -6.47026956e-01 5.82988858e-01
-4.52072501e-01 -1.20505404e+00 -6.54852867e-01 5.00325322e-01
-9.99615729e-01 -9.88726020e-01 -1.29823995e+00 -6.88097179e-01
5.41692317e-01 1.47816539e-01 8.21253061e-01 -2.68417802e-02
-3.09913725e-01 2.94323843e-02 -4.58980277e-02 4.52291444e-02
-7.06071138e-01 1.28436217e-03 9.63790491e-02 -1.04186330e-02
-1.58234119e-01 -7.33463466e-01 -9.43489969e-01 3.96425933e-01
-7.94983566e-01 -5.11280857e-02 5.46582520e-01 9.07113433e-01
9.37531948e-01 -3.52442920e-01 4.89502817e-01 -1.11780024e+00
4.27812815e-01 -4.48378563e-01 -6.34108543e-01 6.34899020e-01
-4.98145401e-01 2.99297273e-01 4.24993813e-01 -7.32325017e-01
-7.74152637e-01 1.35808513e-01 -1.67051628e-02 -4.61963385e-01
8.49731863e-02 4.57308948e-01 4.05349344e-01 -7.04630554e-01
6.07006848e-01 5.77980161e-01 1.68297186e-01 -3.28363180e-01
-1.13400072e-01 5.02553046e-01 7.95012355e-01 -7.55069852e-01
2.42265046e-01 4.00719583e-01 3.87775153e-01 -6.05913877e-01
-4.41796184e-01 -4.21817005e-01 -6.69414818e-01 -4.89851624e-01
7.61543930e-01 -5.25228083e-01 -9.47153270e-01 8.71761560e-01
-9.78781343e-01 -6.90628886e-01 2.32577801e-01 9.07068789e-01
-7.60693073e-01 6.90390706e-01 -1.10487759e+00 -6.46621883e-01
-8.04956079e-01 -1.61893845e+00 8.91667306e-01 6.93692267e-02
6.46714941e-02 -1.30592835e+00 2.33929172e-01 1.62885115e-01
9.20770288e-01 7.85341382e-01 8.02345932e-01 -2.02348158e-01
-2.82156020e-01 -1.07023761e-01 -5.40674105e-02 -1.32430300e-01
9.89825428e-02 -5.13681829e-01 -9.86914396e-01 -3.92970443e-01
2.98947364e-01 -1.79509982e-01 8.15177262e-01 1.00489247e+00
1.32098675e+00 -9.76096839e-02 -3.08174521e-01 7.82129347e-01
1.38914847e+00 3.50999445e-01 7.57221997e-01 2.19714433e-01
5.09717464e-01 4.16545272e-01 3.06284279e-01 3.80959392e-01
6.02385215e-02 6.91153288e-01 5.34993261e-02 -1.17790870e-01
-1.32355317e-01 -9.82382149e-02 7.20175877e-02 1.06054056e+00
-1.99761644e-01 5.75461946e-02 -8.41263115e-01 2.76664764e-01
-1.52605796e+00 -9.51217413e-01 -2.28304803e-01 2.10505986e+00
9.15605724e-01 -4.73530412e-01 -5.83711490e-02 -1.03022374e-01
9.51677918e-01 5.72991110e-02 -6.46358609e-01 -8.64027143e-02
7.78778940e-02 3.78376305e-01 8.97567451e-01 6.17925644e-01
-9.84135866e-01 3.33744019e-01 7.22376060e+00 7.46749818e-01
-1.74468708e+00 4.54379350e-01 5.68377554e-01 2.79402465e-01
-3.37783068e-01 -3.00243199e-01 -3.43811631e-01 6.39659226e-01
1.16544843e+00 2.73447663e-01 6.43871963e-01 3.71793300e-01
2.02586606e-01 -2.61686087e-01 -6.55669987e-01 8.85096073e-01
-2.05715045e-01 -1.25100172e+00 -3.66568297e-01 -1.72935873e-02
3.38641226e-01 3.96108866e-01 -7.72413686e-02 -3.12113203e-02
-1.74926832e-01 -9.88015592e-01 6.92484856e-01 7.80219674e-01
1.31465042e+00 -2.75583267e-01 8.30114484e-01 1.41003758e-01
-1.00836992e+00 1.54038474e-01 -5.49861968e-01 7.30883420e-01
8.19423571e-02 8.17939103e-01 -8.00196409e-01 5.55158973e-01
4.08550709e-01 2.95423090e-01 -2.64013916e-01 1.10374510e+00
3.35487247e-01 2.43463993e-01 -2.09168375e-01 6.44827262e-02
9.65084210e-02 -4.84894775e-02 2.78786212e-01 1.31473982e+00
4.04228240e-01 -1.44002080e-01 -5.86035312e-04 8.80480230e-01
1.96637273e-01 1.52824581e-01 -2.63622314e-01 -4.30685878e-02
4.78600770e-01 1.43020844e+00 -1.06275403e+00 -1.89201832e-01
1.34294987e-01 7.35026181e-01 -3.11453231e-02 3.37488353e-01
-9.20373380e-01 -2.79936224e-01 -1.67882532e-01 3.16833586e-01
3.98755111e-02 -1.51569113e-01 3.50649267e-01 -1.19039810e+00
-3.09374839e-01 -4.78484988e-01 6.11970620e-03 -6.82492375e-01
-1.05942106e+00 8.47127616e-01 1.66683450e-01 -1.07437932e+00
-1.85807377e-01 -6.71315372e-01 -3.78035367e-01 7.92777956e-01
-1.70609725e+00 -1.03808951e+00 -2.92129517e-01 5.29984534e-01
-2.52457947e-01 5.72925955e-02 1.01425457e+00 5.94204605e-01
-2.93295026e-01 6.32525921e-01 4.56648290e-01 1.03465293e-03
5.94849348e-01 -1.27617526e+00 -2.60878146e-01 2.88426518e-01
-4.17275220e-01 8.60124111e-01 6.90914452e-01 -4.95864987e-01
-1.36375570e+00 -8.84132564e-01 6.30848944e-01 -1.24200946e-02
7.10469425e-01 -1.71838969e-01 -8.09974253e-01 3.20900172e-01
-1.63244739e-01 5.87179124e-01 9.03534710e-01 -3.26460987e-01
-1.16495974e-01 -1.93338796e-01 -1.48023307e+00 4.68694046e-02
4.32435721e-01 -6.88164353e-01 -2.05600001e-02 4.93040264e-01
4.45192635e-01 -9.72897112e-01 -1.72310793e+00 5.39153993e-01
1.08792126e+00 -7.51741052e-01 7.60287642e-01 -1.53234005e-01
5.27458712e-02 -3.62943292e-01 -1.80552632e-01 -9.91696894e-01
-4.31698203e-01 -7.64254034e-01 -1.93325028e-01 5.23932695e-01
1.01943716e-01 -7.76618361e-01 6.80408418e-01 6.05378389e-01
-2.71924049e-01 -7.71290958e-01 -1.20308816e+00 -6.08716786e-01
2.47745499e-01 -7.92234913e-02 7.61077642e-01 1.07783461e+00
-1.24719024e-01 -3.01284522e-01 -5.61919630e-01 -2.85687577e-02
9.34412897e-01 8.59017521e-02 9.41026434e-02 -1.00663972e+00
-5.12467206e-01 -2.28747696e-01 -2.95743465e-01 -6.38943255e-01
2.64372826e-01 -1.12232780e+00 1.41362831e-01 -9.89493966e-01
5.99006891e-01 -7.24722385e-01 -7.50184655e-01 4.37924117e-01
1.76533177e-01 1.91875517e-01 -1.07902050e-01 4.51479644e-01
-3.74640137e-01 1.56545877e-01 1.81444120e+00 -2.64677972e-01
1.99691713e-01 -2.37462863e-01 -1.89997122e-01 2.41736978e-01
5.37759066e-01 -7.98636854e-01 -3.41083616e-01 -2.65895933e-01
-2.14459196e-01 9.20769811e-01 3.72141063e-01 -1.09691930e+00
2.17419192e-01 -6.65763766e-02 6.03771508e-01 -3.66481990e-01
3.38948146e-02 -6.80305064e-01 7.11473346e-01 8.67869079e-01
-3.21424246e-01 -1.89519912e-01 7.77365714e-02 3.49249750e-01
2.11843010e-02 -4.90101218e-01 1.11129630e+00 -3.84168804e-01
-2.73649365e-01 5.11845410e-01 -6.45125389e-01 -3.29481453e-01
5.12576580e-01 -2.17422381e-01 -3.66675198e-01 -9.07983109e-02
-1.01962805e+00 -9.34611782e-02 6.41612262e-02 -1.65875360e-01
4.86808270e-01 -1.32070935e+00 -4.74051654e-01 7.68005429e-03
-2.56883919e-01 -4.15104061e-01 9.45025921e-01 1.51606631e+00
-8.92635703e-01 6.13661528e-01 -3.80381703e-01 -7.86865950e-01
-6.11360312e-01 3.91357511e-01 1.21051443e+00 -7.03083396e-01
-5.85213661e-01 3.56367439e-01 5.83632700e-02 -9.58969831e-01
-1.02363348e-01 -5.17779946e-01 -2.32514322e-01 -2.99372107e-01
6.60645485e-01 2.51295209e-01 2.82921731e-01 -6.70653820e-01
-2.89708495e-01 6.67712331e-01 -8.74358714e-02 1.99712768e-01
1.34850943e+00 -7.85387307e-02 -3.57676893e-01 3.96429211e-01
1.18697405e+00 -3.32039386e-01 -1.08506835e+00 -1.78786293e-01
1.51179865e-01 -9.79460850e-02 3.92582923e-01 -9.83749628e-01
-1.32705748e+00 8.20935607e-01 1.11609674e+00 -2.88875669e-01
8.54159951e-01 -3.89742345e-01 9.62486148e-01 2.71186680e-01
5.69595277e-01 -8.09418678e-01 -3.49978209e-01 1.81331635e-01
7.57648170e-01 -8.41423154e-01 -5.85698746e-02 -2.34519586e-01
-1.60002142e-01 1.52244592e+00 2.56898344e-01 1.33981526e-01
6.31988347e-01 4.90509748e-01 1.37690648e-01 -2.90020257e-01
-2.59240091e-01 5.59887588e-01 2.44625613e-01 6.35914624e-01
8.93048227e-01 2.01684967e-01 -3.34979981e-01 4.35029715e-01
1.81810036e-01 4.68221903e-01 5.00935256e-01 7.73809373e-01
-3.48156661e-01 -1.04624975e+00 -1.58137456e-01 4.76045817e-01
-8.02248895e-01 -1.58719476e-02 3.24276268e-01 5.34537256e-01
-9.79522690e-02 3.89383286e-01 -3.10985744e-01 -2.38254935e-01
-7.93812349e-02 -1.26932696e-01 1.04828715e+00 -1.23607442e-01
-5.92533112e-01 3.20435554e-01 -1.75630972e-01 -7.12110341e-01
-8.11440170e-01 -5.20404935e-01 -1.41312659e+00 -2.37320259e-01
-5.82409680e-01 2.82538682e-01 7.82259405e-01 9.09847140e-01
4.75627109e-02 3.65540832e-01 6.80715621e-01 -1.02741611e+00
-5.40470719e-01 -1.06246495e+00 -1.20275033e+00 2.75460761e-02
4.60583210e-01 -7.82118201e-01 -7.01732337e-02 -1.91849381e-01] | [13.519079208374023, -2.4051969051361084] |
f1644114-bbdc-435e-a5b6-e9a658e445e1 | twin-identification-over-viewpoint-change-a | 2207.05316 | null | https://arxiv.org/abs/2207.05316v1 | https://arxiv.org/pdf/2207.05316v1.pdf | Twin identification over viewpoint change: A deep convolutional neural network surpasses humans | Deep convolutional neural networks (DCNNs) have achieved human-level accuracy in face identification (Phillips et al., 2018), though it is unclear how accurately they discriminate highly-similar faces. Here, humans and a DCNN performed a challenging face-identity matching task that included identical twins. Participants (N=87) viewed pairs of face images of three types: same-identity, general imposter pairs (different identities from similar demographic groups), and twin imposter pairs (identical twin siblings). The task was to determine whether the pairs showed the same person or different people. Identity comparisons were tested in three viewpoint-disparity conditions: frontal to frontal, frontal to 45-degree profile, and frontal to 90-degree profile. Accuracy for discriminating matched-identity pairs from twin-imposters and general imposters was assessed in each viewpoint-disparity condition. Humans were more accurate for general-imposter pairs than twin-imposter pairs, and accuracy declined with increased viewpoint disparity between the images in a pair. A DCNN trained for face identification (Ranjan et al., 2018) was tested on the same image pairs presented to humans. Machine performance mirrored the pattern of human accuracy, but with performance at or above all humans in all but one condition. Human and machine similarity scores were compared across all image-pair types. This item-level analysis showed that human and machine similarity ratings correlated significantly in six of nine image-pair types [range r=0.38 to r=0.63], suggesting general accord between the perception of face similarity by humans and the DCNN. These findings also contribute to our understanding of DCNN performance for discriminating high-resemblance faces, demonstrate that the DCNN performs at a level at or above humans, and suggest a degree of parity between the features used by humans and the DCNN. | ["Alice J. O'Toole", 'Carlos D. Castillo', 'Jacqueline G. Cavazos', 'Ying Hu', 'Vivekjyoti Banerjee', 'Virginia E. Strehle', 'Connor J. Parde'] | 2022-07-12 | null | null | null | null | ['face-identification'] | ['computer-vision'] | [ 1.88853279e-01 -2.42125943e-01 2.61107177e-01 -7.43142247e-01
-2.42691383e-01 -8.46583664e-01 7.61262059e-01 -3.19964178e-02
-5.67471862e-01 1.28110170e-01 1.99214322e-03 6.28252327e-02
-5.06105535e-02 -5.75189948e-01 -2.44266808e-01 -3.64852011e-01
-1.45537704e-01 3.68597507e-01 -5.09198666e-01 -3.91929522e-02
3.09833974e-01 8.89053881e-01 -1.81843030e+00 5.54369926e-01
3.04992139e-01 1.16581261e+00 -2.70646185e-01 5.99423885e-01
4.05646473e-01 9.20953974e-02 -4.96660322e-01 -1.01399231e+00
8.47723484e-01 -7.30649114e-01 -4.42643791e-01 -1.18955120e-01
1.60304415e+00 -5.41300297e-01 -2.07282126e-01 7.92226732e-01
7.19975233e-01 1.09192319e-01 7.09499657e-01 -1.52746093e+00
-9.89757419e-01 2.47837622e-02 -6.54267967e-01 3.12958181e-01
6.94490552e-01 5.17382801e-01 6.87600613e-01 -1.07879090e+00
4.13669884e-01 1.70884442e+00 1.07247055e+00 7.43851304e-01
-1.37591410e+00 -1.39718199e+00 -3.25548410e-01 2.09647179e-01
-1.64295018e+00 -1.03999019e+00 1.43595546e-01 -6.69372857e-01
1.03928900e+00 5.36796579e-04 9.35106456e-01 1.07036829e+00
-1.53196976e-03 -2.77746975e-01 1.37936699e+00 -8.61879140e-02
-1.31039456e-01 1.12056114e-01 -2.15381607e-01 3.07464302e-01
4.09834713e-01 5.64311206e-01 -4.73211676e-01 -3.87147039e-01
9.69643891e-01 -2.14419678e-01 -1.17775045e-01 1.22296847e-01
-9.16004777e-01 7.20936537e-01 5.38557053e-01 4.76836264e-01
-2.96165377e-01 -3.68484467e-01 3.76972169e-01 4.89577383e-01
1.14517964e-01 8.62028182e-01 -7.83886537e-02 -7.08802091e-03
-8.08244288e-01 3.35817993e-01 6.81285441e-01 4.95014071e-01
5.88716626e-01 9.82549861e-02 -2.02609316e-01 1.04565620e+00
-1.19725026e-01 4.22378004e-01 6.03677094e-01 -1.19251275e+00
5.38707897e-02 2.44594052e-01 3.59549746e-02 -1.55249119e+00
-5.62878847e-01 -6.46381155e-02 -7.86258221e-01 6.01111591e-01
7.43204772e-01 1.22524843e-01 -5.57199597e-01 2.09628749e+00
1.17758587e-01 -1.32602498e-01 -3.49482149e-01 1.23599541e+00
6.37059748e-01 1.45130619e-01 3.40165287e-01 -1.76776901e-01
1.81130016e+00 -2.17167705e-01 -2.39450455e-01 -4.94802088e-01
1.07991457e-01 -9.12884593e-01 9.46852624e-01 8.66188183e-02
-1.17191267e+00 -1.15393567e+00 -1.04993916e+00 3.92144732e-02
-1.75186321e-01 -2.83311438e-02 3.89180183e-01 1.13195252e+00
-1.51258433e+00 8.94829512e-01 -1.07767366e-01 -7.60127366e-01
5.75307369e-01 4.60523099e-01 -1.05599880e+00 -1.38749301e-01
-1.06578350e+00 1.31914532e+00 -1.34678930e-01 1.38560846e-01
-7.53917456e-01 -8.60139668e-01 -9.56102550e-01 -2.98613943e-02
-5.32167315e-01 -6.45625710e-01 1.13263428e+00 -1.88493848e+00
-9.68936205e-01 1.71608555e+00 -2.44184002e-01 -1.10886447e-01
2.55761504e-01 1.39137268e-01 -9.10333514e-01 3.73095363e-01
3.60784680e-01 8.53617489e-01 8.64509761e-01 -1.05894375e+00
-2.47428745e-01 -1.08709371e+00 -2.32088134e-01 3.49026114e-01
-1.14234842e-01 6.21094704e-01 4.37675625e-01 -3.56921703e-01
6.17675185e-02 -7.59711087e-01 5.41732728e-01 6.38086498e-01
8.82230252e-02 -2.88018346e-01 5.34400880e-01 -7.54244387e-01
5.71834624e-01 -2.09964585e+00 -5.63775420e-01 2.45852500e-01
3.87414992e-01 4.06521887e-01 -5.77790022e-01 3.68523598e-01
-6.86667323e-01 2.85915196e-01 3.09037883e-02 -3.09204996e-01
-1.72547586e-02 -3.15388381e-01 2.41261095e-01 8.63905132e-01
4.60884273e-01 7.65061855e-01 -6.56424344e-01 -1.88350275e-01
-1.01502158e-01 4.53431875e-01 -3.02893996e-01 3.11105907e-01
6.27674162e-01 2.66936511e-01 5.37577987e-01 4.90542799e-01
1.04848611e+00 2.15025648e-01 1.98189825e-01 -4.21242744e-01
-1.16503231e-01 -1.26119301e-01 -7.75628090e-01 9.25064802e-01
-4.21875417e-01 1.00699270e+00 3.17500293e-01 -5.23436844e-01
9.89078343e-01 3.38224292e-01 -3.89492214e-02 -9.34850097e-01
1.91983894e-01 3.46648932e-01 8.51006269e-01 -3.74573201e-01
1.53389499e-01 -6.47392511e-01 3.96938503e-01 5.69049776e-01
1.26423150e-01 -1.11792117e-01 -3.07329912e-02 -8.62353593e-02
5.93554914e-01 -4.06963080e-01 4.52747166e-01 -3.02544057e-01
2.82173067e-01 -5.45098960e-01 5.46999335e-01 7.90523946e-01
-7.23442554e-01 9.10439789e-01 5.37184238e-01 -6.70419455e-01
-1.21325552e+00 -1.39222348e+00 -4.20330495e-01 9.95770454e-01
-2.02806860e-01 5.24413846e-02 -7.63239443e-01 -3.26909542e-01
4.15914476e-01 5.31352997e-01 -1.03268719e+00 -3.83550286e-01
-2.53103644e-01 -2.68687546e-01 9.80596125e-01 4.67650026e-01
5.04844964e-01 -9.81086373e-01 -3.35275203e-01 -4.60122317e-01
1.32903591e-01 -1.09805954e+00 -7.68637180e-01 -6.59796357e-01
-3.06433052e-01 -1.18953943e+00 -6.56414330e-01 -9.21082497e-01
6.73321366e-01 3.13260585e-01 1.25196946e+00 2.23893598e-01
-5.89981496e-01 4.65982735e-01 3.29182073e-02 -2.20910385e-01
-3.58776599e-01 -8.86757314e-01 7.55422175e-01 2.47642577e-01
6.91506028e-01 -5.82229555e-01 -8.68521810e-01 7.14254498e-01
-3.98812264e-01 -3.32040638e-01 4.85914886e-01 7.02102542e-01
-1.64406851e-01 -4.52517331e-01 6.35066748e-01 -5.69445901e-02
4.96071458e-01 -4.10221666e-01 -1.92093551e-01 8.71832669e-02
-3.88187528e-01 -7.28441179e-01 7.00654924e-01 -6.31332934e-01
-9.06094313e-01 -2.85208791e-01 1.62672669e-01 -5.04534900e-01
-5.63707829e-01 -2.72492886e-01 -1.28663540e-01 -3.73575777e-01
9.09621179e-01 -1.74013138e-01 4.78937328e-01 -7.66841844e-02
-1.69830263e-01 7.25035429e-01 7.57291973e-01 -4.71130699e-01
7.39193499e-01 4.40126538e-01 -9.24306065e-02 -8.13122213e-01
-3.69370610e-01 -8.71828496e-02 -7.16781676e-01 -4.15586114e-01
9.66918707e-01 -9.59399760e-01 -1.24997592e+00 8.23094130e-01
-1.08537090e+00 -1.58214763e-01 1.04307994e-01 7.39229143e-01
-2.81534642e-01 3.45346063e-01 -4.51522380e-01 -7.42258549e-01
-2.83914983e-01 -9.13318694e-01 8.22871745e-01 2.86806762e-01
-7.34975338e-01 -5.77959716e-01 2.14597490e-02 3.14120203e-01
3.93756628e-01 1.42791182e-01 8.33034158e-01 -9.22759116e-01
2.30961785e-01 -2.68226475e-01 -7.57104874e-01 3.77922773e-01
5.12117088e-01 1.00656375e-01 -1.20260870e+00 -3.69477212e-01
1.11164838e-01 -6.18056953e-01 4.57603931e-01 3.67242843e-01
8.18444848e-01 -3.21314752e-01 -9.65357274e-02 5.42804360e-01
1.09678364e+00 3.53913993e-01 6.38227999e-01 -2.15636805e-01
4.12289053e-01 1.03265572e+00 1.09863900e-01 2.42258776e-02
1.23915866e-01 8.05257380e-01 -2.60897428e-02 -1.97910275e-02
-1.56363621e-01 -2.52553046e-01 3.45978439e-01 -1.99058920e-01
3.25785652e-02 5.36740478e-03 -8.80202889e-01 4.68595147e-01
-9.55220520e-01 -1.48131824e+00 7.15315621e-03 2.46830130e+00
4.68401730e-01 -3.72920662e-01 4.49950069e-01 -1.12752959e-01
1.32499766e+00 -1.28127500e-01 -7.00086415e-01 -8.69512439e-01
-3.40441018e-01 1.83372229e-01 -1.49664715e-01 8.99913013e-02
-8.59203219e-01 5.67519307e-01 6.60773706e+00 4.50454623e-01
-1.04586101e+00 -1.94174841e-01 9.84531403e-01 -2.73665845e-01
1.07451528e-01 -1.81641951e-01 -4.15498376e-01 4.26513135e-01
9.48138833e-01 -3.59425664e-01 7.22261548e-01 5.96997261e-01
1.39613912e-01 -1.31705061e-01 -1.53717947e+00 1.36256146e+00
5.19957006e-01 -7.52174020e-01 2.42283978e-02 -4.14403901e-02
6.06916964e-01 -3.52544755e-01 5.04151762e-01 -2.53276178e-03
2.73453444e-02 -1.59183073e+00 4.99717206e-01 2.93126523e-01
1.22335839e+00 -5.77502668e-01 5.51700950e-01 -8.27844143e-02
-1.08550119e+00 -2.07119137e-01 -4.10848647e-01 -6.28619552e-01
-1.91442832e-01 1.11105978e-01 -8.24354589e-01 -6.57695979e-02
1.03135645e+00 3.36249918e-01 -6.15874529e-01 6.53145850e-01
4.24603075e-01 -6.35818765e-02 -2.06432477e-01 3.12492460e-01
-1.64130598e-01 -6.01050481e-02 2.00878263e-01 1.03694427e+00
4.28764313e-01 4.22261924e-01 -4.50603634e-01 1.13082016e+00
-1.04650617e-01 3.13497707e-02 -8.71966958e-01 -1.28134623e-01
8.12445700e-01 1.55223179e+00 -5.56530118e-01 -1.68828040e-01
-3.51033598e-01 1.08563197e+00 3.31545085e-01 1.57783136e-01
-3.20287943e-01 -2.69362360e-01 1.24903274e+00 2.74779052e-01
-1.58392295e-01 1.25666365e-01 -4.39886689e-01 -6.56579375e-01
4.04230133e-02 -1.03181946e+00 3.62364620e-01 -1.05890298e+00
-1.89656842e+00 7.63047218e-01 -1.32439286e-01 -1.04738927e+00
-1.27690122e-01 -7.02098489e-01 -8.12484562e-01 1.41085124e+00
-5.98536730e-01 -1.05289125e+00 -6.65770411e-01 5.80225527e-01
-6.02867119e-02 -2.29780883e-01 9.36051846e-01 5.32366633e-02
-3.98787707e-01 1.01828885e+00 -3.43923509e-01 5.14204443e-01
1.09739995e+00 -7.14072168e-01 6.56379580e-01 4.97448564e-01
-7.66466632e-02 1.15575016e+00 1.11409619e-01 -4.84869033e-01
-8.56585801e-01 -6.16585314e-01 9.28681791e-01 -5.21930158e-01
2.54200816e-01 -3.50636870e-01 -6.90524399e-01 4.48595673e-01
3.06560695e-01 3.32259536e-02 9.84159410e-01 -2.03904733e-02
-1.06488621e+00 -1.56197503e-01 -1.82213187e+00 5.61851978e-01
1.35640085e+00 -1.13307941e+00 -4.24017251e-01 5.24435118e-02
-2.54314318e-02 1.29200473e-01 -9.74228024e-01 4.36511964e-01
1.44531929e+00 -1.66343975e+00 1.07113361e+00 -7.08805680e-01
5.22754252e-01 -3.29886079e-02 -6.41999021e-02 -1.39520836e+00
-7.25226581e-01 -1.24656394e-01 5.97143233e-01 1.16635895e+00
-2.23845569e-03 -9.04940605e-01 1.49305344e-01 8.52152348e-01
8.72838348e-02 -3.76095623e-01 -1.16748691e+00 -9.24501777e-01
3.16182286e-01 -4.43042330e-02 6.01519048e-01 1.24987125e+00
-1.95772778e-02 -1.63321476e-02 1.81498691e-01 7.55902380e-02
5.06910384e-01 -1.18565142e-01 4.39047933e-01 -1.33217323e+00
1.91563666e-02 -9.85768735e-01 -8.54606628e-01 -7.35984370e-02
5.89086115e-01 -9.59615409e-01 -3.19123119e-01 -1.04015827e+00
3.50963771e-01 -1.66253716e-01 3.33074518e-02 3.67488742e-01
-1.66410446e-01 7.44286478e-01 4.22689497e-01 1.72052190e-01
3.73057097e-01 9.65248421e-03 1.02535057e+00 1.25150278e-01
2.74092644e-01 -2.50106394e-01 -1.00674915e+00 6.12694681e-01
8.00527513e-01 8.68265796e-03 -8.97158831e-02 -3.71531904e-01
-3.05836290e-01 -9.39486250e-02 9.73852277e-01 -1.24131155e+00
-5.62381595e-02 4.48623747e-02 1.34272909e+00 2.96339393e-01
6.13560855e-01 -5.10208964e-01 3.57469022e-01 4.53608245e-01
-2.33731940e-01 5.63433111e-01 4.50299710e-01 -1.44110620e-01
2.34367803e-01 3.27300727e-02 1.47929215e+00 -2.02691242e-01
-4.96244431e-01 2.67910898e-01 -1.15309559e-01 1.81084812e-01
8.27461600e-01 -5.76909482e-01 -4.91683662e-01 -6.51633382e-01
-5.13614357e-01 -3.80000502e-01 8.54455352e-01 5.76879859e-01
6.85932696e-01 -1.51950192e+00 -9.94474053e-01 5.83517492e-01
1.56505466e-01 -1.09589946e+00 4.47480887e-01 8.69196951e-01
-2.19777688e-01 7.46298954e-02 -8.72903168e-01 -3.68959308e-01
-1.32013023e+00 5.06745338e-01 8.52787375e-01 6.97638988e-01
1.69229671e-01 1.09287512e+00 5.47307372e-01 -3.35623831e-01
-3.43169928e-01 4.69908327e-01 3.10867317e-02 2.43388310e-01
1.00103188e+00 4.88911211e-01 -1.17987223e-01 -1.31129122e+00
-5.46052337e-01 8.97244632e-01 -2.86796567e-04 -1.30144998e-01
5.84065974e-01 1.59160465e-01 -2.51651734e-01 2.33064473e-01
1.60489631e+00 -2.34754890e-01 -7.48392820e-01 2.16861010e-01
-7.03334689e-01 -9.52258110e-01 -4.15730625e-01 -9.51073408e-01
-9.50672448e-01 8.89691472e-01 1.01166892e+00 5.24544157e-02
9.06614363e-01 -1.63298443e-01 3.31579089e-01 1.88704934e-02
3.21111083e-01 -8.43915343e-01 -3.68201807e-02 1.06144920e-01
1.22810411e+00 -1.41335666e+00 -5.05449027e-02 4.33786213e-02
-6.31978989e-01 1.06480968e+00 1.20822275e+00 -8.06340948e-02
3.34172338e-01 -2.00535804e-01 6.34214282e-02 -1.85203850e-01
-4.24563080e-01 -1.48936927e-01 5.15582860e-01 1.11908531e+00
6.69923365e-01 -2.27530450e-02 -1.61071435e-01 4.20570463e-01
-6.67633057e-01 -3.39372158e-01 1.25749305e-01 2.29905844e-01
-1.32584363e-01 -5.21119356e-01 -7.11175025e-01 8.46896827e-01
-1.53713435e-01 -1.02553301e-01 -8.30470562e-01 8.57188463e-01
3.62892121e-01 1.15596414e+00 9.96311307e-01 -5.07944167e-01
3.09696406e-01 2.02494264e-01 7.85747707e-01 -5.11453271e-01
-9.65231121e-01 -5.88768840e-01 1.48011521e-01 -5.82182646e-01
-1.34143949e-01 -1.16399109e+00 -6.93170547e-01 -7.19151676e-01
5.19168340e-02 -1.27545208e-01 3.29702258e-01 5.60621083e-01
3.66300046e-01 -4.86424893e-01 7.83842802e-01 -1.15226090e+00
-3.25404257e-01 -1.17220056e+00 -6.98615134e-01 7.89703190e-01
3.59521359e-01 -7.57822812e-01 -7.65413582e-01 -2.82053560e-01] | [12.988640785217285, 1.203847885131836] |
f5a844bb-d70f-4298-9293-2e5de6db1bb6 | multi-class-model-fitting-by-energy | 1706.00827 | null | http://arxiv.org/abs/1706.00827v2 | http://arxiv.org/pdf/1706.00827v2.pdf | Multi-Class Model Fitting by Energy Minimization and Mode-Seeking | We propose a general formulation, called Multi-X, for multi-class
multi-instance model fitting - the problem of interpreting the input data as a
mixture of noisy observations originating from multiple instances of multiple
classes. We extend the commonly used alpha-expansion-based technique with a new
move in the label space. The move replaces a set of labels with the
corresponding density mode in the model parameter domain, thus achieving fast
and robust optimization. Key optimization parameters like the bandwidth of the
mode seeking are set automatically within the algorithm. Considering that a
group of outliers may form spatially coherent structures in the data, we
propose a cross-validation-based technique removing statistically insignificant
instances. Multi-X outperforms significantly the state-of-the-art on publicly
available datasets for diverse problems: multiple plane and rigid motion
detection; motion segmentation; simultaneous plane and cylinder fitting; circle
and line fitting. | ['Jiri Matas', 'Daniel Barath'] | 2017-06-02 | multi-class-model-fitting-by-energy-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Daniel_Barath_Multi-Class_Model_Fitting_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Daniel_Barath_Multi-Class_Model_Fitting_ECCV_2018_paper.pdf | eccv-2018-9 | ['motion-detection'] | ['computer-vision'] | [ 2.23182589e-01 -7.80595988e-02 -1.41465440e-01 -3.04326743e-01
-1.40008295e+00 -4.77136225e-01 3.84648472e-01 2.85080463e-01
-9.01453122e-02 3.83987069e-01 -3.60580236e-01 4.38320339e-02
-4.24641550e-01 -9.80530530e-02 -9.44689929e-01 -1.03836477e+00
1.41573489e-01 1.02040458e+00 3.43889236e-01 4.84511763e-01
4.64159995e-01 5.21235883e-01 -1.53124106e+00 2.71705031e-01
9.63404000e-01 7.09184349e-01 -2.91923404e-01 5.85615873e-01
1.52287111e-01 7.29881898e-02 -5.77281952e-01 -7.94575736e-02
3.03195059e-01 -2.12839454e-01 -6.11172855e-01 5.24892867e-01
9.31537271e-01 2.39684850e-01 4.65539157e-01 9.76992488e-01
3.78782541e-01 5.44603050e-01 1.07823384e+00 -1.30356038e+00
5.50428219e-02 8.41211304e-02 -1.11732042e+00 2.66551506e-02
2.02400103e-01 1.60709266e-02 6.24784887e-01 -1.16064620e+00
6.08158350e-01 1.22936594e+00 9.01816785e-01 2.30669737e-01
-1.66208279e+00 -3.53026658e-01 8.26481879e-02 6.62418604e-02
-1.57271636e+00 -4.52782035e-01 7.23085761e-01 -8.11513245e-01
8.56503248e-01 5.85618436e-01 3.00653785e-01 8.46344054e-01
2.76089720e-02 5.91720164e-01 8.39325190e-01 -4.51644093e-01
2.08102465e-01 4.57935421e-05 2.84130126e-01 5.04164875e-01
2.28015602e-01 -3.86620648e-02 -1.67442620e-01 -7.05323458e-01
7.93230712e-01 -3.39671522e-01 -1.23269133e-01 -6.74640656e-01
-1.12657249e+00 6.93502367e-01 -1.90239251e-01 1.07714929e-01
-7.72176012e-02 -4.99928147e-02 3.67303699e-01 3.42517756e-02
7.56242752e-01 4.51650709e-01 -5.87337494e-01 7.67114758e-02
-1.27348733e+00 5.41905522e-01 5.68822443e-01 9.14292574e-01
6.46037817e-01 -1.26784131e-01 1.09252393e-01 8.87462020e-01
5.02876937e-01 2.45252594e-01 4.21084136e-01 -1.06156135e+00
4.62855697e-01 5.05721211e-01 1.07536219e-01 -1.01318598e+00
-7.26363540e-01 -4.70771044e-01 -6.69159472e-01 2.23809123e-01
9.08332586e-01 9.09089968e-02 -9.76322830e-01 1.23058784e+00
9.82238054e-01 8.70122552e-01 -2.72962868e-01 7.86201060e-01
6.37613118e-01 3.49136770e-01 -1.04215592e-01 -5.75828850e-01
9.24670398e-01 -1.15553606e+00 -4.18811768e-01 -1.16984747e-01
7.72932589e-01 -1.10380065e+00 7.86668777e-01 6.56229675e-01
-1.01971424e+00 -6.28540874e-01 -7.18191504e-01 1.64690867e-01
-9.45677515e-03 1.93870932e-01 2.19055936e-01 3.65411460e-01
-4.14683014e-01 9.74700093e-01 -7.06589878e-01 4.90549393e-02
2.71094531e-01 4.42880541e-01 -4.30251122e-01 5.80665134e-02
-1.62768438e-01 6.44778252e-01 2.44087994e-01 1.69774201e-02
-3.77152264e-01 -9.18727040e-01 -7.77545333e-01 -3.20723891e-01
5.27827978e-01 -6.09067857e-01 1.05159628e+00 -6.94278896e-01
-1.25143516e+00 1.16271830e+00 -3.98656934e-01 1.21914595e-02
7.25008249e-01 -2.90076405e-01 -3.22977006e-01 1.54405713e-01
1.47085279e-01 3.71738315e-01 1.40111923e+00 -1.68522966e+00
-4.26921219e-01 -4.23175544e-01 -8.37469757e-01 -4.54397351e-02
4.70554858e-01 1.39421120e-01 -6.45949900e-01 -8.84363234e-01
7.97995031e-01 -1.19650722e+00 -4.87392545e-01 -3.13401252e-01
-8.82549405e-01 -2.24534854e-01 9.12411809e-01 -5.98258913e-01
1.24182343e+00 -2.28870463e+00 2.67337531e-01 7.17974305e-01
3.02378863e-01 4.77434099e-02 4.88782115e-02 -1.66822448e-01
-6.40488207e-01 1.38735816e-01 -4.44430113e-01 -7.75486767e-01
-1.02876693e-01 1.68021351e-01 -9.47193429e-02 9.57675755e-01
1.25526875e-01 5.21955907e-01 -6.24434590e-01 -5.19837856e-01
4.55957025e-01 1.25702918e-01 -3.78830999e-01 3.36963609e-02
-2.87207276e-01 9.97604191e-01 -2.26114348e-01 5.95381558e-01
1.02435470e+00 -3.69457603e-01 -2.78341770e-01 -1.28991187e-01
5.44604026e-02 -3.92237574e-01 -1.88906085e+00 1.46226406e+00
1.70297101e-01 2.43676484e-01 -3.50816958e-02 -1.00105894e+00
8.84905279e-01 4.11201209e-01 1.08664787e+00 3.14457938e-02
6.04493320e-02 4.30996388e-01 -1.31733119e-01 -6.61862075e-01
3.09996039e-01 1.03858471e-01 1.55943140e-01 4.60238904e-02
-4.29842062e-02 -3.63879174e-01 -2.06666850e-02 -2.34413445e-01
6.55779898e-01 2.53274620e-01 3.45948875e-01 -3.27728361e-01
5.13500750e-01 -7.08174543e-04 7.50540435e-01 7.81643510e-01
-1.26594439e-01 1.19645834e+00 4.17361289e-01 -3.23262572e-01
-1.08965254e+00 -8.44522417e-01 -6.35206640e-01 7.18789160e-01
6.35609329e-02 -4.04845327e-01 -8.27552855e-01 -4.77228850e-01
2.98345417e-01 5.89873493e-01 -4.39827085e-01 6.75345734e-02
-8.49173963e-01 -1.08942997e+00 2.46003449e-01 2.82327503e-01
-2.04431608e-01 -6.84274733e-01 -2.39206344e-01 3.30019802e-01
-1.99863136e-01 -1.34918511e+00 -5.92678964e-01 2.62408316e-01
-1.13356662e+00 -1.39128387e+00 -5.23299158e-01 -5.36196709e-01
8.93784344e-01 -9.67216045e-02 1.11069870e+00 2.42950827e-01
-4.56399471e-01 3.79379153e-01 -9.13638324e-02 -2.46944487e-01
-4.99276847e-01 -2.67856807e-01 5.98516315e-02 2.43534669e-01
1.06428944e-01 -3.28903794e-01 -2.31725276e-01 6.41784489e-01
-6.27345860e-01 -1.78604826e-01 -5.18113151e-02 6.81046128e-01
1.13099766e+00 7.85485581e-02 1.72604457e-01 -8.63752007e-01
1.99823946e-01 -5.73404014e-01 -7.27239668e-01 1.65979132e-01
-4.45208102e-01 -9.46299583e-02 4.97162282e-01 -6.99752569e-01
-5.93867779e-01 4.00049329e-01 2.16749370e-01 -9.87737775e-01
-6.27222776e-01 1.05399959e-01 6.04526326e-02 -4.76992577e-01
5.35541058e-01 -2.74418890e-01 -2.46584237e-01 -7.22070217e-01
2.97485381e-01 3.19699824e-01 9.14163351e-01 -7.62186766e-01
7.46708989e-01 4.67359781e-01 4.87561971e-01 -1.00019538e+00
-6.44841552e-01 -1.27018523e+00 -1.15469015e+00 -2.45404512e-01
7.62756705e-01 -6.01972401e-01 -4.39934522e-01 7.16168225e-01
-1.26835775e+00 -8.90062079e-02 -3.48396122e-01 4.86462682e-01
-8.73026192e-01 6.73375189e-01 -2.71872699e-01 -8.38479102e-01
7.17141554e-02 -1.45840263e+00 1.55070674e+00 1.52169302e-01
-3.42940003e-01 -1.00151229e+00 1.37095153e-01 5.86223781e-01
-2.94611067e-01 6.71633780e-01 8.57101083e-01 -9.62954402e-01
-5.51327825e-01 -3.85555863e-01 1.42396718e-01 1.35493368e-01
-2.09532917e-01 5.82277358e-01 -1.10761416e+00 -2.71229923e-01
2.49342993e-02 9.61296931e-02 6.91440642e-01 9.83473420e-01
1.44362962e+00 2.21995115e-02 -6.82035565e-01 9.33877349e-01
1.37066209e+00 -5.34842126e-02 3.45350713e-01 2.33035609e-01
8.45203996e-01 6.06383801e-01 7.64148891e-01 6.37422919e-01
-1.14396133e-01 1.04998469e+00 2.99930215e-01 -2.30089217e-01
3.15028399e-01 2.80619234e-01 -1.42071158e-01 6.79674983e-01
-1.77268703e-02 -9.25623178e-02 -1.04539359e+00 3.86772543e-01
-2.23700118e+00 -7.90133715e-01 -7.47397959e-01 2.59072900e+00
4.51934308e-01 9.59820077e-02 3.77842277e-01 2.95610577e-01
8.20184827e-01 -1.69489786e-01 -5.37080765e-01 -4.94979285e-02
-1.33843049e-01 -4.98401234e-03 5.02738953e-01 6.08527720e-01
-1.51560056e+00 6.40416920e-01 6.60053921e+00 1.08772254e+00
-7.44313538e-01 -4.98083606e-02 7.08072186e-01 6.80809692e-02
-5.23119234e-02 -7.52690881e-02 -1.03108203e+00 4.44303900e-01
8.27231526e-01 1.82116106e-01 1.07553579e-01 7.95250595e-01
4.23646867e-01 -2.92287081e-01 -1.07795727e+00 1.06063485e+00
8.87156948e-02 -1.09606731e+00 -3.96656543e-01 5.19952923e-03
9.57449973e-01 -5.64051494e-02 -3.34125757e-02 -2.70651311e-01
-2.22800523e-01 -9.96991277e-01 6.99055374e-01 7.87590563e-01
7.05910444e-01 -5.21286726e-01 3.43755543e-01 5.20820796e-01
-1.22304761e+00 7.51442909e-02 -2.09795162e-01 6.28813863e-01
2.94574618e-01 8.25231969e-01 -6.28279090e-01 6.88424230e-01
5.38497388e-01 5.45996606e-01 -6.64741755e-01 1.67231846e+00
3.26661438e-01 5.77803075e-01 -7.47318566e-01 7.02414811e-01
-6.62713647e-02 -7.46405840e-01 9.97827291e-01 1.25878143e+00
2.06961587e-01 -1.54277653e-01 5.88823199e-01 7.61124790e-01
3.93063396e-01 2.12870821e-01 -3.99299413e-01 6.65065646e-01
2.68713087e-01 1.20006192e+00 -1.07032180e+00 -3.23905826e-01
-4.05258477e-01 6.43665373e-01 1.13977809e-02 4.24818575e-01
-8.69544685e-01 2.67858952e-01 4.72186416e-01 1.26071036e-01
4.28374857e-01 -1.68999687e-01 -5.59112728e-01 -1.03060520e+00
2.11279094e-01 -9.52833354e-01 4.56889600e-01 -6.51948810e-01
-1.32861245e+00 1.95825249e-01 3.97623181e-01 -1.36995542e+00
-3.88870329e-01 -6.71456993e-01 -7.52711654e-01 6.85245991e-01
-1.01817310e+00 -1.00039852e+00 -2.13607147e-01 3.94577444e-01
6.48826480e-01 -1.22226693e-01 8.03180873e-01 3.41550738e-01
-7.88085520e-01 4.53187466e-01 6.13198042e-01 -4.28477347e-01
8.86891365e-01 -1.44129682e+00 3.00220281e-01 8.10625136e-01
1.27520934e-01 4.63590145e-01 9.44640100e-01 -8.17025542e-01
-6.67243183e-01 -1.04449987e+00 3.53010029e-01 -6.49300277e-01
3.96229625e-01 -9.09404680e-02 -1.44445074e+00 7.62342453e-01
-3.89114976e-01 4.59295273e-01 6.67222083e-01 -9.66708288e-02
7.61418939e-02 3.60086977e-01 -1.26065207e+00 2.46503279e-01
6.96523488e-01 -1.15151443e-01 -3.29071552e-01 7.06619203e-01
4.47465539e-01 -8.74711215e-01 -1.05296934e+00 5.57998359e-01
2.44978443e-01 -7.58602321e-01 1.34926558e+00 -7.61953413e-01
-3.34426109e-03 -3.86348367e-01 1.89776123e-02 -1.26067233e+00
-2.52121806e-01 -1.07591021e+00 -3.09270382e-01 1.07529235e+00
3.02762121e-01 -3.87699783e-01 8.08123469e-01 7.12703526e-01
-3.52810562e-01 -8.96257102e-01 -1.29189885e+00 -7.07894146e-01
1.48563102e-01 -5.51073790e-01 4.09341574e-01 1.06680381e+00
-4.42039520e-01 1.69480704e-02 -2.32209370e-01 4.70447809e-01
1.00720787e+00 -8.54842830e-03 9.46319222e-01 -1.63334799e+00
-3.56450528e-01 -4.00118381e-01 -2.69463837e-01 -8.73077035e-01
1.85031906e-01 -6.31762028e-01 3.17518529e-03 -7.90917516e-01
-1.37362471e-02 -5.47053814e-01 1.85318425e-01 1.53214708e-01
-2.54109412e-01 -2.47828066e-02 -2.97658127e-02 5.12146771e-01
-6.44479513e-01 2.40911752e-01 1.23308289e+00 1.79844156e-01
-3.02639723e-01 5.84498823e-01 -1.19876951e-01 1.17575586e+00
4.74229187e-01 -7.75306284e-01 1.46583188e-02 4.05227952e-03
-7.67431408e-02 2.17738822e-01 4.62558806e-01 -1.04449463e+00
2.05962688e-01 -2.06409469e-01 3.15791577e-01 -9.98395145e-01
3.25885057e-01 -9.78197694e-01 3.47114652e-01 -1.33552104e-01
-7.60949776e-02 1.93796724e-01 4.06221628e-01 6.70004964e-01
7.24956801e-04 -6.29552782e-01 1.09509552e+00 6.31387681e-02
-3.95990312e-01 2.24110663e-01 -9.33687538e-02 1.98338628e-01
1.20275450e+00 -4.18788850e-01 -1.67706490e-01 8.20575729e-02
-1.28961694e+00 1.63479686e-01 5.28525651e-01 2.06390396e-01
4.77136940e-01 -1.11873496e+00 -5.19207776e-01 4.73073334e-01
7.29910433e-02 5.29098153e-01 1.06460355e-01 1.15315700e+00
-3.19652349e-01 1.23311274e-01 4.00794297e-01 -1.17108357e+00
-1.59102499e+00 2.28395402e-01 6.58583462e-01 -1.82003006e-01
-8.12820494e-01 8.65258694e-01 -1.84551179e-02 -6.32960618e-01
1.88359380e-01 -6.59674108e-01 -1.01049565e-01 -7.95841515e-02
1.39667511e-01 7.39099562e-01 2.09377229e-01 -1.04064727e+00
-2.89831728e-01 1.09471333e+00 2.21147358e-01 8.26110616e-02
1.17854559e+00 -1.28954574e-01 -1.29376024e-01 7.37588227e-01
1.14199162e+00 4.29387167e-02 -1.27721500e+00 -8.05024616e-03
2.34344155e-01 -6.09991372e-01 -1.26741111e-01 -4.79561418e-01
-7.66906321e-01 3.82995129e-01 5.96780300e-01 8.65267441e-02
7.00549245e-01 9.80865434e-02 3.29663366e-01 -1.13185626e-02
-1.14080191e-01 -1.34330630e+00 -2.43157316e-02 3.83014262e-01
7.97531664e-01 -1.42533469e+00 2.48970896e-01 -7.63179541e-01
-3.78636599e-01 1.26075125e+00 6.63093925e-01 -1.42158031e-01
7.10432768e-01 2.52951711e-01 1.44494012e-01 -4.25280541e-01
-3.89219344e-01 2.18947873e-01 9.26512122e-01 4.81627136e-01
1.65974483e-01 1.71442404e-02 5.72333783e-02 4.42121357e-01
-3.89768593e-02 -4.01840001e-01 3.26043129e-01 5.77167571e-01
-2.98845500e-01 -9.62363005e-01 -1.10631967e+00 7.02687085e-01
-4.36246186e-01 2.57412970e-01 3.74234803e-02 8.81861448e-01
4.01747137e-01 7.18864501e-01 2.68958628e-01 4.64385189e-03
4.99597847e-01 2.70118713e-01 4.16613489e-01 -6.53263807e-01
-5.05044699e-01 8.51222754e-01 -8.08468014e-02 -6.33676410e-01
-4.69720274e-01 -1.29800117e+00 -1.24950695e+00 -8.05549920e-02
-8.96233797e-01 -1.27744347e-01 5.24622977e-01 1.19845891e+00
1.68345720e-01 4.36082840e-01 4.93852824e-01 -1.05740249e+00
-6.53795123e-01 -8.58861804e-01 -5.90769172e-01 7.52195358e-01
5.74912250e-01 -8.81522655e-01 -7.44059145e-01 1.07577197e-01] | [8.000880241394043, -2.6807377338409424] |
2b35bce5-0592-4634-bfca-7db874c81d87 | searching-for-efficient-multi-scale | 1809.04184 | null | http://arxiv.org/abs/1809.04184v1 | http://arxiv.org/pdf/1809.04184v1.pdf | Searching for Efficient Multi-Scale Architectures for Dense Image Prediction | The design of neural network architectures is an important component for
achieving state-of-the-art performance with machine learning systems across a
broad array of tasks. Much work has endeavored to design and build
architectures automatically through clever construction of a search space
paired with simple learning algorithms. Recent progress has demonstrated that
such meta-learning methods may exceed scalable human-invented architectures on
image classification tasks. An open question is the degree to which such
methods may generalize to new domains. In this work we explore the construction
of meta-learning techniques for dense image prediction focused on the tasks of
scene parsing, person-part segmentation, and semantic image segmentation.
Constructing viable search spaces in this domain is challenging because of the
multi-scale representation of visual information and the necessity to operate
on high resolution imagery. Based on a survey of techniques in dense image
prediction, we construct a recursive search space and demonstrate that even
with efficient random search, we can identify architectures that outperform
human-invented architectures and achieve state-of-the-art performance on three
dense prediction tasks including 82.7\% on Cityscapes (street scene parsing),
71.3\% on PASCAL-Person-Part (person-part segmentation), and 87.9\% on PASCAL
VOC 2012 (semantic image segmentation). Additionally, the resulting
architecture is more computationally efficient, requiring half the parameters
and half the computational cost as previous state of the art systems. | ['Liang-Chieh Chen', 'Yukun Zhu', 'Jonathon Shlens', 'Florian Schroff', 'Maxwell D. Collins', 'Barret Zoph', 'Hartwig Adam', 'George Papandreou'] | 2018-09-11 | searching-for-efficient-multi-scale-1 | http://papers.nips.cc/paper/8087-searching-for-efficient-multi-scale-architectures-for-dense-image-prediction | http://papers.nips.cc/paper/8087-searching-for-efficient-multi-scale-architectures-for-dense-image-prediction.pdf | neurips-2018-12 | ['street-scene-parsing'] | ['computer-vision'] | [ 5.29579282e-01 3.36133838e-01 -7.56266043e-02 -6.14671707e-01
-9.35241461e-01 -3.01338196e-01 5.49598932e-01 -1.75545007e-01
-6.61720932e-01 3.25302124e-01 -8.33482146e-02 -2.92412728e-01
-8.13245401e-03 -7.55036652e-01 -8.73141170e-01 -3.01340431e-01
6.83331564e-02 9.23220992e-01 3.88223022e-01 -1.05880484e-01
-8.78680404e-03 4.30410057e-01 -1.65433896e+00 6.58821344e-01
6.41519189e-01 1.08345056e+00 5.02410352e-01 8.50812495e-01
-3.22627544e-01 7.61742473e-01 -4.73240793e-01 -5.29262722e-01
3.14190447e-01 -3.31473976e-01 -1.31333673e+00 2.33979627e-01
9.38601553e-01 -2.04220504e-01 -1.94384173e-01 9.02435720e-01
1.89740956e-01 9.52549204e-02 5.15065730e-01 -8.61153305e-01
-4.19225067e-01 3.97032946e-01 -2.99209714e-01 1.71852469e-01
-1.30698949e-01 1.37248412e-01 1.02471232e+00 -7.31179237e-01
5.87394178e-01 1.18737197e+00 9.21463668e-01 8.35675359e-01
-1.36067891e+00 -4.19933230e-01 2.21118242e-01 2.36500368e-01
-1.10341311e+00 -4.36929464e-01 5.15150070e-01 -5.24172306e-01
1.45465541e+00 1.36588603e-01 6.56720281e-01 8.44459534e-01
-5.36908917e-02 1.00765693e+00 9.41027939e-01 -4.04566109e-01
1.13187581e-01 1.10348545e-01 2.50561059e-01 9.38637197e-01
1.41532868e-01 -3.91729362e-02 -2.28758931e-01 1.88344210e-01
7.58679092e-01 -1.79228798e-01 1.72472894e-01 -4.35335010e-01
-9.31080222e-01 8.70724559e-01 7.07160115e-01 2.18374193e-01
-2.46125653e-01 2.05752894e-01 3.62427205e-01 -6.34592548e-02
4.02772605e-01 7.44960070e-01 -8.62770796e-01 -2.04704027e-03
-1.06545055e+00 2.29820728e-01 8.13283920e-01 8.05389702e-01
9.44361329e-01 1.14070982e-01 1.80933520e-01 1.04466832e+00
7.94981141e-03 1.49773091e-01 5.12568772e-01 -1.25354230e+00
6.24041736e-01 6.58836901e-01 -6.25058487e-02 -6.66320980e-01
-6.55133009e-01 -5.61456621e-01 -9.34730530e-01 1.41025335e-01
5.27348638e-01 -1.20953888e-01 -1.34757781e+00 1.55339718e+00
2.03390181e-01 -9.90068987e-02 -4.20433842e-02 5.12905777e-01
7.59947002e-01 7.02967823e-01 4.14797157e-01 2.04322487e-01
1.32002974e+00 -1.21349883e+00 7.60184228e-02 -9.92448151e-01
7.18023777e-01 -4.76485014e-01 1.00605392e+00 3.24419379e-01
-1.05298293e+00 -1.05922043e+00 -7.78281331e-01 -1.97668880e-01
-3.87366712e-01 1.02282293e-01 8.30292404e-01 7.31754363e-01
-1.12210393e+00 6.93672419e-01 -8.05387139e-01 -5.53246081e-01
7.95225024e-01 4.71928328e-01 -2.98706204e-01 -1.95677981e-01
-6.40516579e-01 1.01808548e+00 7.08616853e-01 -6.77112192e-02
-6.85980618e-01 -7.20306873e-01 -8.65021884e-01 8.11572522e-02
1.02779523e-01 -8.00588191e-01 1.45452762e+00 -1.32182419e+00
-1.12652469e+00 1.19601822e+00 -1.59911484e-01 -8.78951788e-01
2.37465590e-01 -3.17086995e-01 -2.21458375e-01 1.81880578e-01
1.78478166e-01 1.36305952e+00 7.30692804e-01 -1.12853992e+00
-1.01167655e+00 -3.71215761e-01 -1.16598807e-01 2.07393080e-01
-2.05428123e-01 -1.34103537e-01 -6.35716975e-01 -4.13251579e-01
1.44826174e-01 -1.10019243e+00 -6.63365841e-01 -2.79679954e-01
-2.81394124e-01 -2.48580649e-01 7.29505956e-01 -6.53164923e-01
7.36783922e-01 -2.06139421e+00 6.45295456e-02 -8.24900940e-02
3.21195982e-02 4.57935065e-01 -3.06793958e-01 -8.10896605e-03
-5.62528856e-02 1.21804669e-01 -2.13909745e-01 -5.06233454e-01
-8.94254819e-02 3.12373161e-01 -2.89732188e-01 1.10338770e-01
2.14951664e-01 9.91642356e-01 -4.61403102e-01 -5.95641851e-01
3.17536443e-01 2.54693776e-01 -6.96215570e-01 1.94044352e-01
-5.43887436e-01 2.16287822e-01 -3.52399439e-01 6.71645641e-01
2.98731804e-01 -6.24609768e-01 2.12073743e-01 -2.48818696e-01
-6.95769787e-02 1.97781160e-01 -8.65869701e-01 1.76621115e+00
-3.77326339e-01 6.04736745e-01 1.14230610e-01 -1.33107936e+00
7.48687208e-01 -1.68788396e-02 4.56101000e-01 -8.96597683e-01
5.06009907e-02 1.63817257e-01 -4.38909978e-03 -3.42863470e-01
5.34261107e-01 -2.96370108e-02 -1.59651041e-01 1.26830071e-01
4.22715396e-01 -3.36259753e-01 1.98456794e-01 -1.13027602e-01
1.15852654e+00 8.21666718e-02 1.29992180e-02 -3.01066071e-01
2.03325346e-01 6.42845571e-01 4.35143143e-01 1.04161191e+00
-3.25568169e-01 5.90939701e-01 1.67758480e-01 -9.60155547e-01
-1.31214774e+00 -9.03925002e-01 -1.80042595e-01 1.39268517e+00
-7.18046352e-02 -2.02784210e-01 -1.11372900e+00 -6.06614411e-01
-1.89905971e-01 6.09449327e-01 -5.02178729e-01 2.42472664e-01
-9.38899219e-01 -8.49245906e-01 5.83438694e-01 8.30306053e-01
8.05941045e-01 -1.24799478e+00 -7.56605506e-01 2.93257982e-01
-8.47805068e-02 -1.37192392e+00 1.65256917e-01 4.63725835e-01
-1.12468338e+00 -1.10795462e+00 -6.22456491e-01 -1.13814378e+00
4.72082883e-01 1.31304726e-01 1.49825001e+00 1.34781348e-02
-6.54303372e-01 4.27184999e-01 -9.92738381e-02 -3.61033857e-01
-3.71174276e-01 4.56504047e-01 -4.43959057e-01 -5.02846420e-01
5.01770794e-01 -2.55137235e-01 -5.97164214e-01 2.36906707e-01
-6.52703166e-01 5.07193863e-01 8.73823345e-01 8.79430473e-01
8.07179987e-01 1.65547386e-01 3.31345737e-01 -1.12259519e+00
4.41266708e-02 -3.42235774e-01 -6.91193283e-01 1.77909017e-01
-4.40574765e-01 1.16239287e-01 5.68071365e-01 -7.42112175e-02
-1.15078664e+00 5.94133675e-01 -4.86771822e-01 -6.19113296e-02
-6.51931942e-01 1.34266704e-01 3.81410010e-02 -7.61685744e-02
7.87648857e-01 1.45521432e-01 -1.75764248e-01 -5.94357729e-01
5.39568424e-01 3.81974936e-01 6.66765273e-01 -5.70989251e-01
5.27545989e-01 4.34206814e-01 -4.36203443e-02 -9.71415043e-01
-1.27782202e+00 -5.16049683e-01 -8.43286395e-01 1.64406583e-01
1.42375183e+00 -1.00949740e+00 -1.87912732e-01 5.38822651e-01
-1.09255075e+00 -7.92354822e-01 -2.37893164e-01 7.11404309e-02
-7.16658711e-01 1.86322834e-02 -6.90714717e-01 -4.48083967e-01
-4.03150856e-01 -1.10322094e+00 1.16810143e+00 1.34474874e-01
-3.10912728e-01 -8.30566168e-01 -5.82255609e-02 8.73443961e-01
2.93438911e-01 7.57392943e-02 1.10160732e+00 -6.15975857e-01
-7.61643469e-01 -6.92119822e-02 -4.76298720e-01 4.17273939e-01
-3.49418223e-01 -4.16411996e-01 -1.03631377e+00 3.33735831e-02
-1.38418674e-01 -6.12026870e-01 1.08784533e+00 5.61431527e-01
1.53098500e+00 -1.27391815e-01 -4.37181681e-01 8.68403256e-01
1.49736440e+00 2.04751655e-01 6.60776138e-01 6.01663768e-01
8.52708876e-01 7.29730606e-01 3.32384437e-01 -3.32618169e-02
4.84947413e-01 5.85213006e-01 2.45085150e-01 -2.08895162e-01
-3.09199631e-01 -2.59919018e-01 -1.07570432e-01 2.98090041e-01
1.04029551e-01 5.16191917e-03 -1.25398183e+00 7.16457307e-01
-1.82241440e+00 -9.29961085e-01 -3.52090485e-02 1.74500823e+00
7.09248662e-01 4.51657027e-01 1.42882079e-01 -1.38000667e-01
5.27012050e-01 1.31154165e-01 -5.60733438e-01 -4.77850080e-01
3.69602698e-04 4.83948290e-01 6.27616882e-01 3.11979204e-01
-1.54154217e+00 1.36436439e+00 6.99689627e+00 7.96091139e-01
-8.08917105e-01 6.03193529e-02 1.13056600e+00 6.10182062e-03
1.52065724e-01 -2.22407147e-01 -1.22819138e+00 1.25496425e-02
1.19552445e+00 5.71098268e-01 2.18309939e-01 1.38799298e+00
-3.11977655e-01 -1.39136255e-01 -1.18852842e+00 1.06534314e+00
6.49448857e-02 -1.71274090e+00 -4.69968058e-02 -1.10968143e-01
8.53477776e-01 5.80533385e-01 3.84607390e-02 6.00843847e-01
4.52269137e-01 -1.34204125e+00 6.36052430e-01 1.32367849e-01
5.95977008e-01 -5.85025549e-01 5.44942081e-01 4.37160760e-01
-1.14369917e+00 -3.18101078e-01 -4.62577194e-01 -4.26784344e-02
1.20555878e-01 3.78062129e-01 -8.10883224e-01 -4.43130508e-02
1.01336861e+00 4.53337431e-01 -8.68582368e-01 9.54837978e-01
5.91319567e-03 5.72138906e-01 -3.65569800e-01 1.83844805e-01
5.51453352e-01 1.17573343e-01 2.71811988e-02 1.44619727e+00
9.27915275e-02 7.51118436e-02 3.57126534e-01 6.57372355e-01
-1.62768424e-01 -2.07976118e-01 -5.50358713e-01 1.20555118e-01
1.59962088e-01 1.06972826e+00 -8.91017437e-01 -5.02397895e-01
-4.73082393e-01 9.99408126e-01 6.11868978e-01 2.68457979e-01
-6.92480028e-01 -8.93658772e-02 5.74982822e-01 1.73478916e-01
6.41597152e-01 -2.64384955e-01 -6.86648011e-01 -8.56971502e-01
-2.22788379e-01 -9.20410395e-01 4.88742679e-01 -6.84074044e-01
-1.12767541e+00 8.49345624e-01 4.92938757e-02 -5.88629663e-01
-5.00174344e-01 -9.26582754e-01 -3.87988091e-01 6.31341100e-01
-1.28506327e+00 -1.36977208e+00 -3.26039255e-01 5.43663025e-01
9.83512640e-01 -3.43311310e-01 9.84521627e-01 1.92515224e-01
-5.91434836e-01 3.07728380e-01 -9.47901309e-02 3.37891310e-01
1.73490182e-01 -1.24263644e+00 7.20039487e-01 7.87347674e-01
5.25558114e-01 1.56840503e-01 4.81478035e-01 -5.02865732e-01
-1.13945425e+00 -1.33100009e+00 7.67869532e-01 -5.95057666e-01
4.31164682e-01 -3.56955707e-01 -7.28169680e-01 9.12304163e-01
9.63179171e-02 5.82325682e-02 5.95715582e-01 4.43725586e-01
-4.81358349e-01 -1.25049189e-01 -1.07773232e+00 5.64151645e-01
1.15414035e+00 -4.47580427e-01 -6.38480723e-01 3.86295229e-01
7.60725439e-01 -3.60358417e-01 -6.09396100e-01 5.36903501e-01
4.89219040e-01 -1.19382787e+00 1.26820982e+00 -6.18267715e-01
4.25077975e-01 1.24007784e-01 -2.73787588e-01 -8.79927337e-01
-5.50630391e-01 -1.46530524e-01 -3.61506194e-02 8.56759727e-01
5.44749856e-01 -2.64821291e-01 1.41010618e+00 9.59957361e-01
-3.65666717e-01 -8.65555465e-01 -7.57834315e-01 -6.63181901e-01
2.71295398e-01 -6.17922246e-01 2.99441397e-01 5.41378081e-01
-5.91754258e-01 5.14217734e-01 -9.80329141e-02 6.34213239e-02
7.49327898e-01 2.26881504e-01 8.36771429e-01 -1.23526382e+00
-4.98174995e-01 -6.16281509e-01 -3.89145106e-01 -1.18017292e+00
3.30816358e-01 -5.73780298e-01 1.40200391e-01 -1.66780555e+00
2.72065431e-01 -5.52028000e-01 -1.27423376e-01 5.88328004e-01
3.25420648e-02 3.49740386e-01 2.94800103e-01 3.50030601e-01
-8.06126952e-01 2.09036037e-01 8.21934104e-01 -2.09510654e-01
-1.05377696e-01 -7.52349896e-03 -6.60724282e-01 1.16843796e+00
9.30147171e-01 -3.04616064e-01 -4.11761820e-01 -8.57626796e-01
-3.76994126e-02 -2.01685637e-01 5.35718560e-01 -1.37845623e+00
1.01517856e-01 -8.18786621e-02 7.40704656e-01 -6.97316706e-01
6.37683332e-01 -7.20274627e-01 3.64356525e-02 5.38717508e-01
-2.69940585e-01 3.74150202e-02 4.97979254e-01 4.23767537e-01
-2.66167879e-01 -3.26040447e-01 1.09328437e+00 -6.97453201e-01
-1.42089117e+00 2.60378420e-01 -1.53980866e-01 2.13180855e-01
8.86616349e-01 -4.34028327e-01 -2.99547344e-01 6.38023466e-02
-7.31109083e-01 -2.60017775e-02 3.23061317e-01 3.76860052e-01
5.13948441e-01 -8.30833077e-01 -5.40601432e-01 1.08523384e-01
-8.82119089e-02 2.29445443e-01 3.03033143e-01 3.02359372e-01
-7.68983126e-01 6.95309281e-01 -1.99762315e-01 -7.29961514e-01
-1.16707659e+00 3.21349233e-01 4.40495640e-01 -4.87406433e-01
-7.59082139e-01 1.25568449e+00 3.92077237e-01 -4.75780368e-01
2.46292219e-01 -1.80582628e-01 -8.73696804e-03 -2.19605386e-01
3.93070370e-01 1.07208714e-01 1.52370669e-02 -5.50762951e-01
-2.74918169e-01 5.54289520e-01 -2.07325608e-01 9.14251804e-02
1.41654396e+00 1.36679932e-01 7.53474683e-02 3.95838656e-02
1.29246461e+00 -6.51721239e-01 -1.46345305e+00 -8.51632804e-02
1.66330516e-01 -2.33061045e-01 4.96665351e-02 -9.65740681e-01
-8.68241370e-01 9.42286551e-01 6.85442150e-01 8.07818547e-02
1.02966297e+00 2.84368843e-01 1.01046908e+00 6.97429001e-01
4.31935757e-01 -1.31461203e+00 1.97936594e-01 7.96225131e-01
5.16341865e-01 -1.54595113e+00 -3.44585963e-02 -3.63980711e-01
-6.00344300e-01 9.19711471e-01 9.07327235e-01 -2.28886172e-01
6.14148915e-01 2.31150463e-02 -1.96196288e-01 -3.27077091e-01
-5.90870440e-01 -3.14595073e-01 4.85511243e-01 6.55081332e-01
3.97247046e-01 1.04374595e-01 3.91375661e-01 4.53787088e-01
-4.20339912e-01 -2.18150809e-01 5.20632416e-02 7.32199073e-01
-8.46102118e-01 -1.05356884e+00 -1.72467858e-01 6.95602953e-01
-4.50869620e-01 -2.19147444e-01 -2.51564920e-01 9.55759525e-01
3.79094154e-01 7.38642454e-01 4.08240736e-01 -1.94230065e-01
2.05109760e-01 2.26410359e-01 6.16665542e-01 -9.39138532e-01
-6.19399309e-01 -1.12390779e-01 4.65754211e-01 -6.83847010e-01
-3.47245485e-01 -5.36285400e-01 -1.08392036e+00 -5.88365644e-02
-1.39487637e-02 -1.61124364e-01 7.30288863e-01 1.08599305e+00
3.37177843e-01 3.67246985e-01 -8.13816860e-02 -1.04118454e+00
-3.31654340e-01 -7.63116896e-01 -2.14007080e-01 3.86433125e-01
3.01788971e-02 -2.88643837e-01 8.71850103e-02 1.76900923e-01] | [9.640812873840332, 0.8921961188316345] |
4213fefe-6de6-4f53-a35b-30be6b9c1617 | time-series-anomaly-detection-based-on | 2303.17802 | null | https://arxiv.org/abs/2303.17802v2 | https://arxiv.org/pdf/2303.17802v2.pdf | Time-series Anomaly Detection based on Difference Subspace between Signal Subspaces | This paper proposes a new method for anomaly detection in time-series data by incorporating the concept of difference subspace into the singular spectrum analysis (SSA). The key idea is to monitor slight temporal variations of the difference subspace between two signal subspaces corresponding to the past and present time-series data, as anomaly score. It is a natural generalization of the conventional SSA-based method which measures the minimum angle between the two signal subspaces as the degree of changes. By replacing the minimum angle with the difference subspace, our method boosts the performance while using the SSA-based framework as it can capture the whole structural difference between the two subspaces in its magnitude and direction. We demonstrate our method's effectiveness through performance evaluations on public time-series datasets. | ['Kazuhiro Fukui', 'Atsuto Maki', 'Naoya Sogi', 'Takumi Kanai'] | 2023-03-31 | null | null | null | null | ['time-series-anomaly-detection'] | ['time-series'] | [ 1.48339465e-01 -5.98597944e-01 1.70216650e-01 -8.86322260e-02
-2.01960847e-01 -7.29497373e-01 4.69223529e-01 2.10275471e-01
-3.31421802e-03 3.92159857e-02 2.64585525e-01 -1.98404536e-01
-2.54347116e-01 -5.66895366e-01 -2.86663920e-01 -5.89232087e-01
-6.76456213e-01 -5.21036565e-01 3.89454305e-01 -6.16176784e-01
2.57741034e-01 5.04728436e-01 -1.15050697e+00 6.41740263e-02
7.05716074e-01 1.31968975e+00 -6.74493313e-01 3.09227586e-01
1.95527405e-01 9.83411670e-02 -6.20623291e-01 2.22451687e-01
6.04736209e-01 -3.63405615e-01 -1.90975174e-01 9.13269073e-02
2.51359493e-01 -2.32077520e-02 -3.92881930e-01 1.39620829e+00
1.67629570e-01 1.03031248e-01 3.60150307e-01 -1.34296465e+00
-1.88084781e-01 -1.09457807e-03 -7.48004913e-01 8.98208439e-01
7.61596859e-01 -3.97037491e-02 9.67784405e-01 -8.77013743e-01
2.58491039e-01 9.92970169e-01 1.04005194e+00 -3.79067026e-02
-1.48618078e+00 -3.07552427e-01 1.06259443e-01 4.63595778e-01
-1.38269126e+00 -1.60265058e-01 1.48876226e+00 -5.01531005e-01
6.60166800e-01 6.55584335e-01 7.24434555e-01 8.27414930e-01
3.84543419e-01 4.56223339e-01 8.76584291e-01 -3.08049977e-01
3.29120815e-01 -4.83944863e-01 2.91766942e-01 2.41397709e-01
2.31961091e-03 -4.01038565e-02 -3.99973094e-01 -8.19068134e-01
5.22841573e-01 3.70202750e-01 -4.90581661e-01 -6.06562614e-01
-1.41065919e+00 4.68097836e-01 6.74652234e-02 5.34904301e-01
-5.73990285e-01 -4.15063053e-01 6.46193445e-01 6.07587099e-01
5.98260522e-01 2.27400407e-01 -3.47908378e-01 -1.63901031e-01
-7.78392971e-01 1.96957931e-01 6.51189566e-01 3.99999082e-01
2.62659907e-01 2.82374769e-01 -1.50783136e-01 5.35801470e-01
2.69722104e-01 4.24038649e-01 8.67642939e-01 -5.55922568e-01
4.10246074e-01 8.78241658e-01 -5.99428564e-02 -1.51251745e+00
-3.70685458e-01 -7.01736033e-01 -7.80162096e-01 -8.12128186e-02
5.27446091e-01 -9.72180814e-02 -3.45975548e-01 1.74849951e+00
6.06127262e-01 6.90467715e-01 -6.12496622e-02 6.90002918e-01
6.69049919e-02 6.52473152e-01 -4.35087115e-01 -7.69408345e-01
1.20421219e+00 -3.08869094e-01 -9.74629700e-01 1.09313913e-02
2.96101272e-01 -6.96400464e-01 9.56411541e-01 3.66562665e-01
-6.96755469e-01 -3.72791946e-01 -1.43605995e+00 7.18430161e-01
-2.22351968e-01 -2.05375671e-01 1.13531888e-01 5.05283296e-01
-7.30546951e-01 8.88972282e-01 -8.53002906e-01 -3.69547635e-01
-3.40215653e-01 -2.19595253e-01 -2.52924055e-01 5.91900468e-01
-1.33664191e+00 3.05150896e-01 6.01888523e-02 5.46533652e-02
-2.80507416e-01 -6.04023635e-01 -7.67173469e-01 -4.23653536e-02
2.16907144e-01 1.22197255e-01 8.79860222e-01 -6.77787066e-01
-1.17085528e+00 3.68459314e-01 -1.76607385e-01 -4.76044416e-01
4.04275090e-01 -6.94024786e-02 -1.33546937e+00 3.55198421e-02
-4.39554676e-02 -5.45173049e-01 1.31132853e+00 -6.44252419e-01
-3.90725255e-01 -6.99208498e-01 -4.97450262e-01 -5.44472635e-02
-6.30762339e-01 -9.91146043e-02 5.12422547e-02 -1.21359980e+00
9.75935698e-01 -8.89541030e-01 -3.80133092e-02 -2.45168865e-01
-3.33453268e-01 -1.60422370e-01 1.28898597e+00 -7.56416142e-01
1.93531883e+00 -2.82692671e+00 -1.36032552e-01 6.18885219e-01
5.80331944e-02 1.73817009e-01 3.64478007e-02 7.52562284e-01
-6.45994663e-01 -3.61838013e-01 -4.44255680e-01 3.37159663e-01
-3.00595194e-01 -7.13490024e-02 -9.74794924e-01 9.96907055e-01
6.89496100e-02 3.64365280e-01 -9.91163313e-01 1.41244322e-01
-6.30408078e-02 1.43767715e-01 -1.56200439e-01 6.26060590e-02
2.16008246e-01 5.13354003e-01 -4.00824368e-01 6.08655930e-01
7.09401548e-01 3.28902066e-01 2.00795978e-01 -2.78377473e-01
-2.69891232e-01 1.13224261e-01 -1.57965088e+00 1.36039150e+00
3.43305498e-01 6.42690182e-01 -2.16346085e-01 -1.14963555e+00
1.09431279e+00 5.48391938e-01 8.93038750e-01 -7.07317054e-01
-2.39891410e-01 2.75198132e-01 9.03993472e-02 -4.26963747e-01
5.22548668e-02 -7.33524859e-02 -1.65217012e-01 2.60529339e-01
-2.24639848e-01 1.80314913e-01 1.44865602e-01 -4.58439514e-02
1.21117091e+00 -1.25283152e-01 6.02445722e-01 -4.64304805e-01
8.79921496e-01 -5.14188945e-01 9.41164255e-01 2.82027632e-01
-8.28797221e-01 2.86097765e-01 7.76009798e-01 -6.54484749e-01
-9.45722878e-01 -1.31803524e+00 -2.75976241e-01 6.32579744e-01
-1.28440494e-02 -7.16256320e-01 -5.36364496e-01 -5.88461995e-01
2.92659611e-01 4.87006843e-01 -6.23551786e-01 -4.84096467e-01
-6.29795134e-01 -6.75227344e-01 4.78339553e-01 3.91095638e-01
3.60482156e-01 -3.85059267e-01 -5.19706130e-01 1.41390577e-01
-3.04880172e-01 -8.73313606e-01 -8.72536719e-01 -1.97752625e-01
-1.29561865e+00 -9.39713836e-01 -3.72525036e-01 -1.97072282e-01
2.75175065e-01 4.45600271e-01 6.26447439e-01 -3.40182662e-01
-1.34354815e-01 4.42603558e-01 -2.41148427e-01 -3.82027268e-01
-2.14287713e-01 -8.66527498e-01 5.50017118e-01 7.08599746e-01
3.63824517e-01 -1.00203001e+00 -6.52742803e-01 5.36789060e-01
-9.08599734e-01 -6.67117715e-01 -4.87966649e-02 5.31931579e-01
5.37308037e-01 2.82255352e-01 6.93769336e-01 -2.05885340e-02
9.33727920e-01 -7.81161666e-01 -3.93292606e-01 -2.26096064e-03
-9.63861406e-01 3.63545725e-03 7.49522209e-01 -5.04919410e-01
-4.58735645e-01 -3.50418180e-01 3.95894945e-01 -6.37333393e-01
-6.01331517e-02 5.48806548e-01 4.85440940e-02 2.10418046e-01
6.47547245e-01 4.92463320e-01 7.34423921e-02 -7.46897697e-01
-1.64804123e-02 4.55288798e-01 7.13655770e-01 -2.01705351e-01
8.72902095e-01 7.44547248e-01 1.92291662e-01 -8.74022841e-01
-4.67735797e-01 -1.02618611e+00 -7.29626834e-01 -2.50963748e-01
2.64502823e-01 -5.86557388e-01 -4.40301925e-01 5.49734533e-01
-8.47420573e-01 5.20678163e-01 -4.22368139e-01 5.25511265e-01
-3.41874421e-01 1.06344569e+00 -3.42429817e-01 -9.40317392e-01
-3.91106218e-01 -4.81303394e-01 9.72495735e-01 -9.45161954e-02
-4.00743663e-01 -8.69624615e-01 7.17188299e-01 -4.13713694e-01
2.72561252e-01 8.53965044e-01 9.65040684e-01 -8.78874719e-01
3.28840107e-01 -7.23215938e-01 4.69445914e-01 4.94535983e-01
4.47229296e-01 9.52124894e-02 -6.71100736e-01 -6.51314378e-01
7.06796288e-01 6.93124592e-01 3.87372404e-01 6.49699271e-02
8.06756258e-01 -3.59251678e-01 -9.08043608e-02 5.12628376e-01
1.22218084e+00 5.25455773e-01 4.09567118e-01 3.63950968e-01
3.13359529e-01 2.90976226e-01 6.47663176e-01 6.58197522e-01
-2.01687768e-01 9.36234415e-01 2.36929268e-01 3.86591405e-01
3.70023817e-01 -1.52083367e-01 8.78445446e-01 1.12629199e+00
-2.35464200e-02 3.94512624e-01 -9.57897246e-01 5.16470671e-01
-1.88472784e+00 -1.26849401e+00 -2.50471860e-01 2.62349534e+00
5.26194453e-01 2.76490629e-01 4.97676432e-01 7.84076095e-01
7.35242486e-01 4.89533931e-01 -6.69485569e-01 -3.45279753e-01
-1.04083039e-01 -1.46061867e-01 1.28082361e-03 -6.03801981e-02
-1.25293660e+00 1.12322740e-01 7.13844633e+00 5.99736452e-01
-1.27834773e+00 -2.42235288e-01 -3.59459035e-02 -6.70127571e-02
5.02687320e-02 -5.65966293e-02 5.29338531e-02 6.64620459e-01
1.01391423e+00 -5.53254306e-01 3.89973193e-01 6.40960395e-01
4.65156972e-01 2.98579603e-01 -9.48432028e-01 9.98914182e-01
-3.35401967e-02 -6.63738489e-01 -1.02705650e-01 3.20203006e-02
3.67864221e-01 -5.39702475e-02 1.19264640e-01 1.29884796e-03
-5.58367193e-01 -3.19663703e-01 7.43045449e-01 5.93137145e-01
5.08047462e-01 -6.75289214e-01 4.73296821e-01 2.84160644e-01
-1.59584546e+00 -4.19567168e-01 6.47727549e-02 -3.36489528e-01
7.91777968e-02 8.99547517e-01 -6.61317229e-01 6.58835649e-01
7.70416141e-01 9.73662615e-01 -4.86569732e-01 9.93837297e-01
1.88129455e-01 8.10280979e-01 -2.02066451e-01 5.36613464e-01
9.87931248e-03 -8.57087910e-01 1.58201838e+00 8.38586450e-01
5.24021327e-01 -3.96884345e-02 2.55993187e-01 4.63189334e-01
5.50863445e-01 1.18603565e-01 -8.04675400e-01 -1.42899022e-01
4.68102813e-01 8.88659060e-01 -4.64045703e-01 -1.20066427e-01
-5.73418260e-01 8.11869502e-01 -2.20081866e-01 4.16487336e-01
-6.25247419e-01 -5.92626154e-01 8.63016784e-01 -1.26398737e-02
9.77798402e-02 -4.13983703e-01 -1.63092956e-01 -1.30443764e+00
6.80513442e-01 -1.02626741e+00 7.96363533e-01 -3.50846887e-01
-1.37606657e+00 4.80878741e-01 -8.16393644e-02 -2.13716483e+00
-4.26876277e-01 -3.44149262e-01 -9.44248676e-01 6.85063362e-01
-7.12923825e-01 -4.21839207e-01 -1.11665837e-01 9.50869024e-01
3.57403964e-01 -2.25387007e-01 7.85189092e-01 1.66257322e-01
-6.67314529e-01 3.59949857e-01 4.13098156e-01 1.47412539e-01
5.01340151e-01 -1.24672842e+00 6.00111008e-01 1.26768672e+00
2.15597406e-01 7.56539941e-01 1.12896228e+00 -7.31370628e-01
-1.40761518e+00 -6.90412939e-01 3.87841433e-01 -3.95392954e-01
1.21425271e+00 -1.54801786e-01 -9.90116179e-01 5.41972935e-01
-1.92705411e-02 2.17030331e-01 7.62988448e-01 1.85936447e-02
-6.21296167e-01 -3.42775166e-01 -1.16326320e+00 6.52150750e-01
8.95783067e-01 -6.98062658e-01 -9.97270286e-01 -3.23091820e-02
5.81682861e-01 -5.07504269e-02 -1.06631291e+00 8.34721625e-01
5.51515222e-01 -9.58263338e-01 1.00527370e+00 -7.02491879e-01
-2.90497601e-01 -7.34051824e-01 -4.75980014e-01 -1.45877361e+00
-4.69705194e-01 -9.68036294e-01 -6.57311380e-01 8.53573382e-01
-9.44232941e-02 -9.52100217e-01 1.97371185e-01 1.67068075e-02
1.17962331e-01 -6.60797000e-01 -1.33298039e+00 -1.21237814e+00
-3.63020211e-01 -4.07612264e-01 7.63709486e-01 1.20614183e+00
6.11241341e-01 5.58432192e-02 -1.04405947e-01 4.38700676e-01
6.15239441e-01 1.31632075e-01 4.55118358e-01 -1.39734554e+00
-1.85245797e-01 -4.61345822e-01 -1.01553237e+00 -4.70654100e-01
-3.72263402e-01 -6.60090864e-01 -3.07386518e-01 -6.71572268e-01
-2.00319976e-01 3.20278466e-01 -8.03150296e-01 5.27826622e-02
-4.44547534e-02 -5.44178262e-02 -5.07098772e-02 4.91359353e-01
-1.51521817e-01 7.15234756e-01 6.44488752e-01 5.56039810e-03
-4.49919730e-01 2.50010133e-01 -2.89295256e-01 8.49909961e-01
6.04081511e-01 -2.47463122e-01 -4.67771709e-01 2.48692170e-01
-4.10327204e-02 2.64184773e-02 3.81557971e-01 -1.43473387e+00
-5.08654676e-02 -1.89071432e-01 3.12374622e-01 -6.21315002e-01
1.85681097e-02 -9.64179099e-01 2.22363919e-01 5.97353995e-01
1.55003257e-02 6.21324062e-01 2.32395396e-01 1.06061029e+00
-4.56292212e-01 3.32749367e-01 5.83866179e-01 3.86864215e-01
-7.42212653e-01 1.25683516e-01 -2.37764537e-01 -1.93994045e-01
1.15405416e+00 -3.20848495e-01 -9.11229923e-02 -3.42274219e-01
-8.80142629e-01 -4.03397754e-02 3.20414662e-01 5.77132821e-01
6.26023471e-01 -1.77537274e+00 -5.29258549e-01 6.14129126e-01
3.39809358e-01 -7.26911426e-01 3.30912888e-01 1.24836957e+00
-1.43149436e-01 3.65764380e-01 -2.52135187e-01 -8.33372891e-01
-1.20787060e+00 8.17941964e-01 4.63763118e-01 -9.64165851e-02
-8.44096780e-01 1.68782279e-01 -1.09910876e-01 -1.12134755e-01
6.30841926e-02 -3.62055779e-01 -4.32661176e-01 1.66745394e-01
7.59167254e-01 8.00506294e-01 2.51311451e-01 -7.88100004e-01
-6.43382013e-01 5.88174045e-01 3.94821078e-01 -3.08886051e-01
1.14055049e+00 -3.23366284e-01 -4.22409683e-01 1.06154704e+00
1.28344500e+00 4.53424275e-01 -1.14245570e+00 -3.33010346e-01
4.72593993e-01 -5.69978893e-01 -6.27908297e-03 -2.86125064e-01
-7.92518795e-01 4.29639995e-01 1.00459385e+00 9.47254241e-01
1.40798187e+00 -2.88906008e-01 7.22417295e-01 1.66757733e-01
3.18804175e-01 -1.11574566e+00 1.32752165e-01 4.29095924e-01
1.09505367e+00 -7.10886121e-01 -1.71518520e-01 -2.24617109e-01
-4.34773356e-01 1.16820419e+00 1.32066488e-01 -3.46452326e-01
9.66370821e-01 -2.27027431e-01 3.18050049e-02 -2.91359037e-01
-5.60127378e-01 1.02971382e-01 7.44648278e-01 3.25671256e-01
3.50102067e-01 2.03185335e-01 -5.94808459e-01 4.15500283e-01
1.38852885e-02 -5.20860374e-01 4.05631512e-01 1.04741240e+00
-3.33254069e-01 -9.05267358e-01 -7.48332441e-01 2.99155027e-01
-4.83886629e-01 2.68359423e-01 -3.73646319e-01 3.34413141e-01
-1.64634928e-01 1.15668178e+00 2.97693282e-01 -6.18414700e-01
8.53536785e-01 5.89396238e-01 -6.10260665e-03 -2.49079004e-01
-3.24380957e-02 2.83401877e-01 -3.58668715e-01 -1.00349140e+00
-1.18769787e-01 -1.11956120e+00 -9.74067390e-01 -1.01804450e-01
-6.47288710e-02 8.31031650e-02 3.96382958e-01 9.01540518e-01
6.06427372e-01 2.99325466e-01 1.34798741e+00 -2.64147043e-01
-1.01256788e+00 -9.45332825e-01 -9.45410371e-01 9.03042614e-01
8.38114738e-01 -5.88133633e-01 -5.92602968e-01 -1.53178513e-01] | [7.245658874511719, 3.01836895942688] |
8caaaacc-3504-4f8e-ab17-7eb28e3055fc | counterfactual-collaborative-reasoning | 2307.00165 | null | https://arxiv.org/abs/2307.00165v1 | https://arxiv.org/pdf/2307.00165v1.pdf | Counterfactual Collaborative Reasoning | Causal reasoning and logical reasoning are two important types of reasoning abilities for human intelligence. However, their relationship has not been extensively explored under machine intelligence context. In this paper, we explore how the two reasoning abilities can be jointly modeled to enhance both accuracy and explainability of machine learning models. More specifically, by integrating two important types of reasoning ability -- counterfactual reasoning and (neural) logical reasoning -- we propose Counterfactual Collaborative Reasoning (CCR), which conducts counterfactual logic reasoning to improve the performance. In particular, we use recommender system as an example to show how CCR alleviate data scarcity, improve accuracy and enhance transparency. Technically, we leverage counterfactual reasoning to generate "difficult" counterfactual training examples for data augmentation, which -- together with the original training examples -- can enhance the model performance. Since the augmented data is model irrelevant, they can be used to enhance any model, enabling the wide applicability of the technique. Besides, most of the existing data augmentation methods focus on "implicit data augmentation" over users' implicit feedback, while our framework conducts "explicit data augmentation" over users explicit feedback based on counterfactual logic reasoning. Experiments on three real-world datasets show that CCR achieves better performance than non-augmented models and implicitly augmented models, and also improves model transparency by generating counterfactual explanations. | ['Yongfeng Zhang', 'Hao Wang', 'Yingqiang Ge', 'Juntao Tan', 'Max Xiong', 'Shuyuan Xu', 'Zelong Li', 'Jianchao Ji'] | 2023-06-30 | null | null | null | null | ['logical-reasoning'] | ['reasoning'] | [ 6.09431043e-02 7.36681819e-01 -5.43907821e-01 -4.42045510e-01
6.26322106e-02 -2.83139825e-01 6.89141154e-01 -2.48582706e-01
-2.29421586e-01 1.08018219e+00 6.50663197e-01 -8.38817775e-01
-3.27773780e-01 -9.56363440e-01 -6.57848001e-01 -2.64067709e-01
1.95770785e-01 1.62579075e-01 -4.51624632e-01 -4.49035496e-01
2.73717523e-01 -3.78144830e-02 -1.43656564e+00 6.52165532e-01
1.72922885e+00 4.13261265e-01 -1.59971178e-01 1.61562160e-01
-2.64700234e-01 1.34622157e+00 -5.17282724e-01 -9.89006102e-01
2.74542481e-01 -2.60757089e-01 -8.42226744e-01 -4.01062906e-01
5.60910068e-02 -4.34724569e-01 -1.86770603e-01 9.03665245e-01
1.81634247e-01 3.00818026e-01 4.32140350e-01 -1.50747800e+00
-1.48805690e+00 1.09517372e+00 -3.02671224e-01 -1.26269415e-01
4.72218692e-01 1.92070171e-01 1.03019953e+00 -5.99379539e-01
2.21257597e-01 1.40053487e+00 6.14673972e-01 9.10339773e-01
-1.14608252e+00 -1.13152492e+00 7.81491280e-01 4.13994044e-01
-8.59117627e-01 -1.43482506e-01 1.14636850e+00 -3.83485496e-01
6.52788222e-01 7.36969113e-01 8.42211664e-01 9.54699278e-01
-4.42435801e-01 1.06380081e+00 1.27017736e+00 -3.85716885e-01
1.63395479e-01 5.96829057e-01 1.32218927e-01 5.15686572e-01
6.20079041e-01 3.96038085e-01 -5.71970455e-02 -1.64319262e-01
7.95185447e-01 4.00043339e-01 -6.06248617e-01 -3.34558934e-01
-1.09455824e+00 8.60033512e-01 8.33826482e-01 5.53644001e-02
-4.93400484e-01 -1.68008134e-01 8.70053917e-02 4.63070333e-01
3.96586806e-01 1.07780254e+00 -8.96406472e-01 2.79767007e-01
-3.11551064e-01 5.09335160e-01 7.94078052e-01 7.45512068e-01
4.18804109e-01 9.39125791e-02 -4.83996630e-01 8.86535764e-01
5.34662545e-01 5.19368768e-01 7.75346816e-01 -8.50956261e-01
7.27093279e-01 1.30922878e+00 4.27823663e-01 -8.97794068e-01
-2.72576004e-01 -4.28392947e-01 -9.65399384e-01 6.96381330e-02
2.03953594e-01 -3.11431438e-01 -4.71028388e-01 1.95351851e+00
3.90932649e-01 3.03951472e-01 4.49953824e-01 1.04649222e+00
8.17213595e-01 2.27092519e-01 3.07335377e-01 -5.18782556e-01
1.32574081e+00 -1.02904189e+00 -9.70817208e-01 1.48355244e-02
9.25879478e-01 -2.40618750e-01 1.44237220e+00 3.76180470e-01
-7.70389855e-01 -4.16373461e-01 -9.00172234e-01 1.25211522e-01
-3.85677665e-01 7.29414970e-02 1.35523820e+00 7.78376937e-01
-4.48143005e-01 4.27064866e-01 -1.64535880e-01 3.78598839e-01
6.45922661e-01 3.13236237e-01 -1.35929376e-01 -1.17992252e-01
-1.73540664e+00 7.36223698e-01 3.92516643e-01 -6.37513921e-02
-1.92312434e-01 -1.14135742e+00 -7.28369951e-01 1.79049239e-01
7.18645930e-01 -1.05599642e+00 1.28941190e+00 -1.03163528e+00
-1.39103246e+00 1.13042243e-01 7.18488470e-02 -5.79690814e-01
7.59167016e-01 -4.64673460e-01 -6.65647686e-01 -5.43801844e-01
-9.31070745e-02 3.88801426e-01 4.02361095e-01 -1.32456326e+00
-4.89632130e-01 -3.42993915e-01 6.86585665e-01 4.42806900e-01
-4.99646485e-01 -4.14438188e-01 1.52960205e-02 -9.00129497e-01
-1.50166154e-01 -6.55670464e-01 -4.24870431e-01 -3.77395451e-01
-4.33597565e-01 -3.47040474e-01 6.21486902e-01 -4.84927535e-01
1.40492761e+00 -1.76576042e+00 -4.07133460e-01 9.67168212e-02
3.07262123e-01 5.72834194e-01 -1.35235339e-01 -8.92704949e-02
-4.76595521e-01 6.46010756e-01 -3.49954423e-03 1.87791362e-01
9.01919529e-02 1.31247044e-01 -7.65415370e-01 -3.22676331e-01
7.22193345e-02 1.23402655e+00 -9.55043733e-01 -1.15010425e-01
1.50338799e-01 8.84978026e-02 -1.06271768e+00 2.56279588e-01
-4.76210058e-01 3.20786059e-01 -6.03254616e-01 3.01014870e-01
8.33024621e-01 -1.30542129e-01 5.49044430e-01 -2.56767988e-01
-1.15037113e-02 5.72109044e-01 -1.11567867e+00 1.05262005e+00
-6.21537387e-01 4.90550175e-02 -5.90158522e-01 -8.14077020e-01
7.77033031e-01 3.84034991e-01 1.91717654e-01 -7.79722512e-01
-6.90900013e-02 -8.52541178e-02 2.74139166e-01 -7.39929736e-01
2.70599097e-01 -2.68458933e-01 1.51767626e-01 6.39975309e-01
-4.83858198e-01 1.32200643e-01 8.59119818e-02 1.90195695e-01
6.54656768e-01 2.39447802e-01 7.06526875e-01 -6.50214823e-03
7.81965733e-01 -4.67688479e-02 7.67596781e-01 7.06434309e-01
4.95889746e-02 -3.08350921e-02 4.68574882e-01 -4.72554326e-01
-7.05926239e-01 -7.77068496e-01 2.73938090e-01 9.15532708e-01
1.08710743e-01 -3.19039047e-01 -5.02381086e-01 -1.28254056e+00
3.36258203e-01 1.35429096e+00 -8.63421023e-01 -4.57503676e-01
-1.95645481e-01 -9.84344244e-01 3.71053010e-01 7.21852481e-01
9.62949455e-01 -1.03549147e+00 -9.78824776e-03 -4.41333279e-02
-4.79748189e-01 -4.52556103e-01 -1.75293252e-01 -5.14593422e-01
-1.05322993e+00 -1.38704824e+00 -1.23031601e-01 1.08601227e-01
7.22994924e-01 5.82368851e-01 9.40572858e-01 6.67899191e-01
4.17783052e-01 5.12013026e-02 -3.95604551e-01 -7.35765517e-01
-2.35951036e-01 -5.08002758e-01 4.38282311e-01 -1.36059567e-01
5.26884675e-01 -6.72365963e-01 -6.40695393e-01 2.81195343e-01
-7.13958442e-01 5.83949089e-01 8.15068126e-01 9.36877668e-01
1.66600987e-01 1.33058310e-01 9.21234548e-01 -1.55483210e+00
1.09349704e+00 -6.06193781e-01 -2.50896215e-01 4.33909178e-01
-1.28478122e+00 2.39200771e-01 1.03980947e+00 -5.73775411e-01
-1.88082302e+00 -4.88314986e-01 5.83546534e-02 -1.51211381e-01
-8.88723880e-02 5.98286450e-01 -5.49053490e-01 2.90821344e-01
7.31604099e-01 -6.08144049e-03 -1.47614360e-01 -4.75082189e-01
8.61449480e-01 6.50403440e-01 4.85784471e-01 -6.68336034e-01
8.34935665e-01 3.39448750e-01 -3.80855292e-01 5.15273549e-02
-1.29491496e+00 -5.91696985e-02 -2.38059133e-01 1.38834834e-01
3.99800688e-01 -8.50096762e-01 -1.23626006e+00 -1.46099776e-01
-1.00577021e+00 -1.32347465e-01 -4.49307680e-01 6.96922481e-01
-4.28678215e-01 1.27042934e-01 -1.17831975e-01 -1.33126307e+00
-3.14710379e-01 -6.44580066e-01 1.35879129e-01 3.85394663e-01
-2.59095818e-01 -1.03585458e+00 -7.78315887e-02 8.86795640e-01
1.73101813e-01 -1.68674290e-01 9.90265250e-01 -9.93393421e-01
-5.17857969e-01 -4.80019189e-02 -3.78538996e-01 1.21314712e-01
1.08835891e-01 -1.79547131e-01 -1.10167730e+00 1.98834881e-01
8.10045674e-02 3.10581215e-02 6.88512385e-01 3.23697515e-02
1.40684390e+00 -1.18716395e+00 -2.44696349e-01 3.06668878e-01
9.81766045e-01 4.02629942e-01 7.26089716e-01 2.37037748e-01
6.67504311e-01 6.39364660e-01 8.96290421e-01 4.19180453e-01
8.54833007e-01 3.99369657e-01 4.04884368e-01 -1.36724517e-01
2.25179613e-01 -7.21537590e-01 1.25454411e-01 2.58624047e-01
-7.18080938e-01 1.70502871e-01 -6.67629957e-01 2.56377131e-01
-2.23850513e+00 -1.44247162e+00 -4.39583421e-01 2.17558050e+00
9.86364305e-01 -5.65777794e-02 -5.86227048e-03 3.08888853e-01
3.96666974e-01 -2.44211093e-01 -6.46600783e-01 -4.49976653e-01
1.12554960e-01 -2.32648239e-01 -1.40054361e-03 3.65093917e-01
-7.67368138e-01 7.91896999e-01 5.55023670e+00 4.52479243e-01
-7.00415909e-01 1.42851457e-01 5.49091816e-01 -2.29551449e-01
-1.08849776e+00 8.53109509e-02 -3.47215027e-01 5.49828768e-01
5.88534713e-01 -4.18394625e-01 5.69909096e-01 1.03177643e+00
4.63727325e-01 2.32131734e-01 -1.13828921e+00 6.01609230e-01
-3.16204280e-01 -1.54290807e+00 5.39485753e-01 4.61442061e-02
1.01333606e+00 -7.02987313e-01 1.18487656e-01 1.05850399e+00
9.87132668e-01 -9.23884392e-01 4.17264044e-01 9.02668715e-01
5.65255880e-01 -7.18156278e-01 9.40062702e-01 7.86769450e-01
-7.34944046e-01 -7.13218749e-01 -2.69778490e-01 -6.38799667e-01
-2.33160153e-01 5.93526065e-01 -8.12293828e-01 8.77134323e-01
4.01690394e-01 2.94924974e-01 -5.16169012e-01 6.87331378e-01
-7.89244413e-01 4.59868819e-01 2.07491860e-01 -1.55593365e-01
-2.11992636e-01 -1.24505505e-01 1.11335881e-01 7.84922779e-01
-1.02268578e-02 6.91929340e-01 -2.36670263e-02 1.29983950e+00
-3.50010365e-01 1.87507749e-01 -6.83055401e-01 4.47510518e-02
6.94051147e-01 9.64138210e-01 1.60873637e-01 -6.49726868e-01
-5.64614773e-01 4.83441055e-01 5.74088871e-01 4.59247172e-01
-8.45324397e-01 6.65726559e-03 9.31732059e-01 -4.62214239e-02
-3.65761012e-01 5.04286230e-01 -7.72224426e-01 -1.55882525e+00
-1.03247404e-01 -9.66406703e-01 5.87410569e-01 -1.00697803e+00
-1.39158535e+00 7.04333633e-02 -6.22453801e-02 -1.14325333e+00
-1.85329005e-01 -3.84820819e-01 -8.76983821e-01 9.18005228e-01
-1.61695051e+00 -1.13522494e+00 -3.50752801e-01 4.87314671e-01
2.72412390e-01 -1.87359408e-01 7.75124729e-01 3.14089842e-02
-5.89584947e-01 6.87767088e-01 -3.20595324e-01 2.77643371e-03
5.22102594e-01 -1.32114255e+00 2.27360964e-01 7.33111024e-01
2.84725744e-02 1.35761547e+00 5.27616501e-01 -8.03150713e-01
-9.27080393e-01 -1.24969018e+00 7.95459867e-01 -6.79372787e-01
3.37441415e-01 8.75113755e-02 -1.09361815e+00 7.49119222e-01
-8.47614780e-02 -3.85404319e-01 1.12155211e+00 7.18672633e-01
-6.59385741e-01 2.77019851e-03 -1.35314071e+00 1.30821311e+00
1.28818178e+00 -2.82724172e-01 -1.16399324e+00 2.26341113e-01
1.16516209e+00 -4.13503274e-02 -6.12514734e-01 5.74694693e-01
5.72475910e-01 -1.02838671e+00 1.06361127e+00 -1.28459501e+00
7.42774308e-01 -3.34343255e-01 9.54498574e-02 -1.44724631e+00
-5.57262480e-01 -4.90013927e-01 -6.73847795e-01 1.28003633e+00
5.59987426e-01 -8.58385980e-01 7.24816978e-01 1.27145946e+00
5.76997809e-02 -9.02929485e-01 -2.22500548e-01 -4.82433319e-01
-5.59287183e-02 -7.33099699e-01 1.52951324e+00 1.62307465e+00
5.82421660e-01 3.70773375e-01 -3.67168576e-01 1.47601083e-01
1.90043479e-01 4.28589582e-01 1.07477283e+00 -1.38963628e+00
-2.46907294e-01 -6.14202023e-01 2.53285319e-01 -1.00752234e+00
1.83251247e-01 -7.98736095e-01 -6.22428834e-01 -1.48425436e+00
2.85456538e-01 -5.76564372e-01 -4.34583366e-01 8.28992367e-01
-7.45843589e-01 -1.49512574e-01 1.49217248e-01 2.17342257e-01
-3.74845684e-01 6.61810160e-01 1.62996852e+00 8.36152360e-02
-4.26001847e-01 2.96082944e-01 -1.64151204e+00 9.61736977e-01
7.35214710e-01 -3.48175727e-02 -1.01441872e+00 4.27270196e-02
4.27990019e-01 1.38568252e-01 5.62303603e-01 -5.25120616e-01
-1.12385876e-01 -7.43527889e-01 3.66591543e-01 -1.41233653e-01
-7.76815740e-03 -9.30373192e-01 5.04639558e-02 4.48185295e-01
-6.21422350e-01 -2.27777272e-01 2.18882430e-02 7.81855643e-01
9.81128067e-02 -9.30888653e-02 2.00916946e-01 -2.10653543e-01
-5.67253113e-01 1.43476397e-01 7.55397081e-02 -2.07966000e-01
9.47289348e-01 -5.91353960e-02 -6.42151892e-01 -3.42059880e-01
-6.19607627e-01 5.16076565e-01 6.07978962e-02 5.63319206e-01
6.09143972e-01 -1.64103854e+00 -5.38092554e-01 3.94030720e-01
1.82991028e-01 -1.93005949e-01 4.81385559e-01 9.28770781e-01
2.03153685e-01 6.07376933e-01 -6.48806989e-02 2.43304223e-01
-1.04654205e+00 1.05175269e+00 2.92528391e-01 -4.67345417e-01
-1.97633997e-01 6.85323596e-01 3.78335059e-01 -9.80094314e-01
-1.32065371e-01 -3.09076130e-01 -6.60597146e-01 -2.22527906e-01
7.08618581e-01 3.28691661e-01 -2.96720952e-01 2.24360064e-01
-2.27473956e-02 -2.97030628e-01 -2.99751610e-01 1.04416534e-01
1.12510908e+00 -6.75116554e-02 -3.14798951e-02 1.60011441e-01
3.14423591e-01 1.99450523e-01 -1.08209801e+00 -3.35989118e-01
-7.19183609e-02 -8.15434575e-01 -2.40619808e-01 -1.45717013e+00
-7.21029341e-01 6.89368963e-01 4.28613387e-02 4.01564658e-01
1.18074238e+00 -4.80111420e-01 6.35356009e-02 6.87643111e-01
2.27634177e-01 -9.07967210e-01 -1.55518681e-01 1.81038380e-01
1.12551510e+00 -1.48235846e+00 9.41215754e-02 -6.55285180e-01
-9.96453881e-01 6.80818737e-01 1.22486436e+00 1.70234740e-01
3.54803026e-01 -2.74024233e-02 -1.24270283e-01 1.05097979e-01
-1.18337691e+00 -3.91849168e-02 5.98179281e-01 6.85994506e-01
8.50935996e-01 3.92414808e-01 -5.44856548e-01 1.60288978e+00
-5.61821818e-01 4.49636251e-01 4.76521283e-01 2.26048157e-01
-5.77585809e-02 -8.35157931e-01 -4.09933478e-01 8.11157107e-01
-1.12163380e-01 -3.21238667e-01 -5.17125666e-01 9.46136892e-01
4.09082562e-01 1.02592194e+00 -2.82501429e-01 -6.36447966e-01
5.25366426e-01 1.43575907e-01 1.98697641e-01 -6.03539348e-01
-5.75852036e-01 -6.39720440e-01 2.30179727e-01 -4.43559140e-01
-3.90362889e-01 -3.58806759e-01 -1.40811098e+00 -6.68688953e-01
-5.88387132e-01 4.61954564e-01 2.53317595e-01 1.21782029e+00
3.98150086e-01 7.70351648e-01 5.30937254e-01 -7.82524571e-02
-8.52365792e-01 -1.10812557e+00 -2.12795407e-01 6.81000769e-01
-4.89771692e-03 -8.46830487e-01 -2.10966870e-01 -2.18929991e-01] | [9.384330749511719, 5.697742938995361] |
45c2a052-1826-40e2-84da-c529fa649120 | diffbfr-bootstrapping-diffusion-model-towards | 2305.04517 | null | https://arxiv.org/abs/2305.04517v1 | https://arxiv.org/pdf/2305.04517v1.pdf | DiffBFR: Bootstrapping Diffusion Model Towards Blind Face Restoration | Blind face restoration (BFR) is important while challenging. Prior works prefer to exploit GAN-based frameworks to tackle this task due to the balance of quality and efficiency. However, these methods suffer from poor stability and adaptability to long-tail distribution, failing to simultaneously retain source identity and restore detail. We propose DiffBFR to introduce Diffusion Probabilistic Model (DPM) for BFR to tackle the above problem, given its superiority over GAN in aspects of avoiding training collapse and generating long-tail distribution. DiffBFR utilizes a two-step design, that first restores identity information from low-quality images and then enhances texture details according to the distribution of real faces. This design is implemented with two key components: 1) Identity Restoration Module (IRM) for preserving the face details in results. Instead of denoising from pure Gaussian random distribution with LQ images as the condition during the reverse process, we propose a novel truncated sampling method which starts from LQ images with part noise added. We theoretically prove that this change shrinks the evidence lower bound of DPM and then restores more original details. With theoretical proof, two cascade conditional DPMs with different input sizes are introduced to strengthen this sampling effect and reduce training difficulty in the high-resolution image generated directly. 2) Texture Enhancement Module (TEM) for polishing the texture of the image. Here an unconditional DPM, a LQ-free model, is introduced to further force the restorations to appear realistic. We theoretically proved that this unconditional DPM trained on pure HQ images contributes to justifying the correct distribution of inference images output from IRM in pixel-level space. Truncated sampling with fractional time step is utilized to polish pixel-level textures while preserving identity information. | ['Xuecheng Nie', 'Tiande Guo', 'Bonan Li', 'ZiCheng Zhang', 'Congying Han', 'Xinmin Qiu'] | 2023-05-08 | null | null | null | null | ['blind-face-restoration'] | ['computer-vision'] | [ 5.38487554e-01 1.18396692e-01 2.09477171e-01 -1.16115876e-01
-7.63513684e-01 -2.88057774e-01 5.06959856e-01 -6.55523896e-01
-1.95213541e-01 8.57661903e-01 2.14911014e-01 -9.12560374e-02
-1.19558960e-01 -9.83783960e-01 -8.85862529e-01 -1.23709667e+00
5.30267537e-01 2.56175362e-02 3.34576070e-02 -9.26663056e-02
9.40130800e-02 5.12912989e-01 -1.72655010e+00 3.21366012e-01
1.02767980e+00 8.17575812e-01 3.78506094e-01 4.13330704e-01
-1.67231575e-01 7.47838795e-01 -5.45289218e-01 -6.41448557e-01
4.38868642e-01 -7.47438371e-01 -4.24810797e-01 2.45180413e-01
2.66530007e-01 -6.07992530e-01 -1.44986838e-01 1.31455350e+00
8.04027677e-01 -3.66083309e-02 8.64652693e-01 -9.16219354e-01
-1.08031976e+00 3.51901978e-01 -9.63607073e-01 -2.59010673e-01
1.89414665e-01 2.62961984e-01 5.24256885e-01 -6.69082582e-01
5.36880732e-01 1.35508680e+00 6.32070184e-01 8.38477552e-01
-1.30552387e+00 -6.58797145e-01 -2.10563526e-01 1.84647918e-01
-1.19863474e+00 -4.62365687e-01 1.05032241e+00 -1.29624799e-01
3.64694536e-01 3.33840370e-01 3.93499792e-01 1.19756746e+00
1.22308992e-01 5.16898274e-01 1.65744174e+00 -4.99755204e-01
1.67680472e-01 2.06164703e-01 -2.68857151e-01 3.82898778e-01
2.26466611e-01 2.72075862e-01 -4.52365875e-01 7.05553591e-02
1.00847912e+00 -9.05829221e-02 -6.57261968e-01 5.71954660e-02
-6.07063353e-01 5.92028260e-01 2.60028452e-01 4.30139661e-01
-6.09179676e-01 -1.29484266e-01 -5.42200021e-02 1.03565797e-01
4.90816832e-01 -2.12881826e-02 -1.94938928e-01 3.30610514e-01
-1.05884910e+00 1.10630412e-02 3.25032860e-01 6.33411169e-01
7.81758130e-01 2.39627793e-01 -5.36739290e-01 1.08692050e+00
2.97679424e-01 8.27594161e-01 3.86177272e-01 -1.07636440e+00
7.46093318e-02 2.56964773e-01 6.59224391e-02 -1.04305065e+00
1.82469442e-01 -6.12575114e-01 -1.20457315e+00 5.71046829e-01
3.97147417e-01 1.46824300e-01 -1.03984022e+00 1.83206069e+00
4.25651014e-01 1.80592269e-01 -1.41888892e-03 9.20166552e-01
6.22481525e-01 8.21396410e-01 -1.59014434e-01 -4.21920151e-01
1.37665915e+00 -6.75522447e-01 -9.40101087e-01 1.72521606e-01
-1.46302851e-02 -8.63942921e-01 1.22463775e+00 5.02047718e-01
-1.23168468e+00 -7.37146497e-01 -9.52332616e-01 -1.72914326e-01
1.15634501e-01 4.24628496e-01 1.83129847e-01 9.92444456e-01
-1.27518868e+00 6.29578590e-01 -3.89100164e-01 -1.33974969e-01
6.65921032e-01 1.36290982e-01 -4.37881023e-01 -3.98669362e-01
-1.03611374e+00 6.44859552e-01 -1.25931278e-01 4.79845762e-01
-7.56753504e-01 -6.59716129e-01 -5.93601108e-01 5.64110354e-02
-2.91927462e-03 -9.29189920e-01 5.68084896e-01 -1.24322081e+00
-1.84202874e+00 7.47579992e-01 -3.18174154e-01 -1.66468039e-01
7.54697978e-01 3.71895428e-03 -2.30432525e-01 3.72082710e-01
8.29592124e-02 5.18546999e-01 1.53683031e+00 -1.66943133e+00
-2.79781431e-01 -3.94280106e-01 -2.19909757e-01 5.64868934e-02
-2.46068671e-01 -2.30804175e-01 -2.63688594e-01 -8.08563948e-01
1.07136726e-01 -4.03794110e-01 5.72963730e-02 -1.26738161e-01
-3.18488449e-01 1.95143580e-01 6.55478179e-01 -1.09353757e+00
8.81478846e-01 -2.22601414e+00 -1.37627989e-01 1.42019883e-01
9.84948277e-02 2.49588385e-01 -2.90813982e-01 8.42220858e-02
-8.74525681e-02 1.02970600e-02 -6.75711334e-01 -6.51906610e-01
-9.43751410e-02 2.27620140e-01 -4.53760177e-01 6.48308575e-01
3.64040256e-01 7.22777426e-01 -5.17364144e-01 -4.33982730e-01
2.32320100e-01 1.02513433e+00 -5.35963714e-01 3.04239333e-01
-2.08519418e-02 8.54948521e-01 -7.33852312e-02 5.43175817e-01
1.49819434e+00 1.61614686e-01 5.65020852e-02 -5.87921739e-01
5.20707369e-02 -3.11123937e-01 -1.22837937e+00 1.33500957e+00
-4.69945610e-01 4.67306003e-02 5.65843344e-01 -9.41788077e-01
1.03816891e+00 2.31273428e-01 1.20571226e-01 -9.68833148e-01
7.02327415e-02 2.12698177e-01 -2.16972739e-01 -4.02373314e-01
1.66767001e-01 -5.76362371e-01 6.03557169e-01 2.20597118e-01
5.78668201e-03 -1.14580110e-01 -1.58828676e-01 2.02460308e-02
8.32259238e-01 4.53453839e-01 -1.74663186e-01 -2.88509429e-01
8.86059105e-01 -5.65212190e-01 6.51520550e-01 6.33778870e-01
3.34242149e-03 9.55590606e-01 3.24661136e-01 -1.32091129e-02
-1.18259704e+00 -1.07787633e+00 -2.28639498e-01 5.24043381e-01
2.31262818e-01 9.21636000e-02 -1.03371143e+00 -4.74389136e-01
-3.16031754e-01 7.79311359e-01 -6.78689897e-01 -7.95442834e-02
-4.46606070e-01 -1.03453994e+00 4.87299770e-01 9.53207817e-03
8.99120629e-01 -1.09137070e+00 -2.34634444e-01 -1.72528893e-01
-4.08488214e-01 -8.58802378e-01 -3.19740206e-01 -1.02153011e-01
-6.95526719e-01 -9.68366325e-01 -1.10523975e+00 -5.71626008e-01
9.04438138e-01 9.17405784e-02 7.57704794e-01 -1.74691412e-03
-2.06816822e-01 1.42397940e-01 -2.74867505e-01 1.45166025e-01
-6.53225780e-01 -4.26818579e-01 -2.20794320e-01 5.24199843e-01
-1.88650370e-01 -9.13472652e-01 -9.05392289e-01 3.07958931e-01
-1.32163894e+00 1.24525331e-01 8.70159209e-01 1.08271837e+00
6.29866660e-01 5.50399184e-01 7.08775342e-01 -7.59095371e-01
3.83430749e-01 -2.38697574e-01 -3.84651631e-01 2.01530635e-01
-6.60428762e-01 6.08717538e-02 8.18729997e-01 -4.86597598e-01
-1.57049668e+00 -2.65275627e-01 -5.98267794e-01 -4.63652730e-01
-2.32343346e-01 -1.94344178e-01 -6.71450555e-01 3.35101550e-03
4.66364950e-01 7.70288229e-01 2.96994567e-01 -5.91007888e-01
4.26763326e-01 5.65085292e-01 6.36351645e-01 -6.67684197e-01
9.08714533e-01 8.20925772e-01 7.71550462e-02 -7.20655739e-01
-5.63878119e-01 -4.64340439e-03 -2.25872353e-01 -2.68482059e-01
7.12377727e-01 -7.15913713e-01 -8.98813903e-01 7.96782017e-01
-1.01305223e+00 -2.84877032e-01 -4.66998219e-01 2.55575895e-01
-4.52712923e-01 6.43545628e-01 -8.01986098e-01 -1.09802318e+00
-4.84112263e-01 -1.05781221e+00 1.09424686e+00 2.94286728e-01
4.52471673e-01 -7.41796732e-01 -2.83727527e-01 5.59383810e-01
6.48128569e-01 6.83207586e-02 8.64338994e-01 -1.43238045e-02
-4.50537622e-01 1.25682265e-01 -4.64463234e-01 9.77652431e-01
3.18409175e-01 -1.18928976e-01 -1.29703319e+00 -2.97439098e-01
7.06672490e-01 4.92521301e-02 9.79893982e-01 6.50560439e-01
1.00799203e+00 -4.89065200e-01 4.95105498e-02 6.44099951e-01
1.61236787e+00 -1.12569788e-02 1.24270642e+00 1.69426024e-01
5.96489549e-01 8.62720549e-01 2.81600267e-01 1.54028535e-01
2.36413963e-02 4.71514910e-01 4.73192364e-01 -4.85868126e-01
-7.81796932e-01 -3.30392033e-01 6.77314341e-01 5.49117744e-01
-6.46901727e-02 -2.38848716e-01 -1.33955106e-01 4.73581523e-01
-1.43490005e+00 -1.08201730e+00 -2.53112227e-01 2.30817103e+00
9.94063437e-01 -5.55506572e-02 -1.58302248e-01 2.86466092e-01
1.01635456e+00 2.23662760e-02 -3.69630694e-01 -8.79605412e-02
-6.92131996e-01 3.57572436e-01 4.25420791e-01 6.36360049e-01
-6.11197352e-01 6.67080700e-01 5.17797947e+00 1.56030726e+00
-1.06567681e+00 3.38540316e-01 9.24993157e-01 2.84831554e-01
-6.93532050e-01 1.33557515e-02 -6.67142212e-01 7.38386631e-01
4.13889170e-01 4.30661172e-01 5.49674094e-01 3.51334810e-01
3.26317370e-01 -2.58462280e-01 -4.93143588e-01 1.07344007e+00
1.57831445e-01 -9.51901734e-01 3.60486656e-01 1.03665091e-01
7.71706402e-01 -6.33743882e-01 3.46056759e-01 1.37230325e-02
3.56768295e-02 -1.11123455e+00 7.59963632e-01 8.39371443e-01
9.48537707e-01 -7.24027038e-01 8.41107965e-01 3.25246185e-01
-7.59380102e-01 -3.08482461e-02 -5.21247804e-01 1.50099456e-01
1.96968272e-01 1.25845039e+00 -4.04611081e-01 9.20564175e-01
6.74221933e-01 4.77601975e-01 -3.77044410e-01 5.19460797e-01
-5.31985343e-01 7.25780547e-01 -2.69738853e-01 7.18421757e-01
-2.96841860e-01 -5.94368041e-01 6.55293524e-01 8.53240073e-01
4.92437214e-01 -3.03478949e-02 -3.93662244e-01 1.29935634e+00
-1.44289481e-03 2.05016267e-02 -3.77986252e-01 4.02002722e-01
2.50120372e-01 1.43174458e+00 -7.23071575e-01 -1.68612346e-01
-1.09763265e-01 1.29172063e+00 -1.43333212e-01 5.10472178e-01
-9.04401779e-01 -6.48433343e-02 1.65275887e-01 4.41796720e-01
4.33716089e-01 1.43902510e-01 -4.29031909e-01 -9.37615395e-01
2.17351019e-01 -9.00890648e-01 3.62697095e-02 -8.57359409e-01
-1.60548353e+00 6.48012817e-01 -3.86397570e-01 -1.03269041e+00
2.42334872e-01 -2.34460443e-01 -5.25831401e-01 1.27856565e+00
-1.90654910e+00 -1.54962575e+00 -2.52688348e-01 7.60011971e-01
3.62084895e-01 1.57752410e-01 4.90275562e-01 5.34044385e-01
-5.82334697e-01 6.89634323e-01 1.14632159e-01 -1.77180961e-01
8.25036168e-01 -1.00204480e+00 -7.00358599e-02 1.10601079e+00
-2.41822317e-01 6.31153882e-01 6.75451458e-01 -8.33754420e-01
-1.27179372e+00 -1.00060952e+00 4.85501796e-01 -1.38472870e-01
1.16544418e-01 -1.63674697e-01 -8.90548706e-01 1.30559653e-01
1.33146480e-01 -2.00789750e-01 3.08450252e-01 -4.83704716e-01
-3.66725594e-01 -3.60045642e-01 -1.65307510e+00 4.95399565e-01
8.37366581e-01 -4.82694238e-01 -3.56053293e-01 -2.04424262e-02
4.48341727e-01 -2.84071080e-03 -6.80630386e-01 5.97276330e-01
4.29504514e-01 -1.31975007e+00 9.24912751e-01 1.92400172e-01
7.37971663e-01 -7.29354143e-01 -3.15895379e-01 -1.01073039e+00
-2.52127618e-01 -6.81071639e-01 8.60708803e-02 1.82792330e+00
5.60497306e-02 -8.43619525e-01 5.15060246e-01 9.88787413e-02
-2.83838958e-02 -5.36817253e-01 -8.98686469e-01 -7.28549123e-01
8.56551633e-04 -1.24512717e-01 6.01546824e-01 7.64035463e-01
-7.03701496e-01 1.67191729e-01 -6.21723354e-01 3.27598304e-01
9.88757789e-01 -2.39051785e-02 6.36492252e-01 -8.05411875e-01
-3.94970149e-01 -3.14473212e-01 3.14828716e-02 -9.77279842e-01
1.04410881e-02 -7.07992315e-01 1.16196729e-01 -1.49249554e+00
3.59011173e-01 -3.89231771e-01 -7.00121373e-02 2.19823733e-01
-2.01663703e-01 6.44688070e-01 5.89132085e-02 4.49912757e-01
1.34805635e-01 7.69736826e-01 1.62754750e+00 -5.26286140e-02
7.09349886e-02 -9.64792445e-03 -7.38355160e-01 4.86810178e-01
5.38988292e-01 -3.83659720e-01 -4.84357983e-01 -1.91440538e-01
7.86154345e-02 -9.84456986e-02 6.63034141e-01 -9.32303786e-01
2.26449538e-02 1.30398557e-01 6.28533006e-01 -3.56743723e-01
2.82151639e-01 -7.65380621e-01 4.04789716e-01 2.67138749e-01
5.19476421e-02 -4.76029217e-01 5.04467078e-02 5.02761722e-01
-2.14957580e-01 -1.75618425e-01 1.25438178e+00 2.86873430e-02
-1.54882208e-01 7.21518174e-02 -1.09616227e-01 -4.45946872e-01
6.47507191e-01 -5.14882565e-01 -3.47533047e-01 -3.92435014e-01
-6.15844965e-01 -3.11454505e-01 6.78480387e-01 -5.32814935e-02
5.31706452e-01 -1.20198643e+00 -9.63909745e-01 4.81577307e-01
-4.94674951e-01 -4.84020529e-05 1.00129557e+00 9.46756184e-01
-4.26804543e-01 -2.46945962e-01 -2.29421198e-01 -4.92226958e-01
-9.32597101e-01 6.60192490e-01 3.83145660e-01 -3.04983735e-01
-7.92952001e-01 8.55651259e-01 5.72621882e-01 -3.06277335e-01
-1.04607102e-02 1.13839686e-01 -2.20701784e-01 -8.97000823e-03
6.59083724e-01 2.83608258e-01 1.55059248e-01 -5.66073895e-01
-8.00902620e-02 7.85957396e-01 9.57409292e-02 -2.28865087e-01
1.32092953e+00 -4.80498135e-01 -5.40212631e-01 -2.49588955e-02
1.04368055e+00 4.90216076e-01 -1.53925478e+00 1.97814763e-01
-6.25325561e-01 -6.05603814e-01 1.54911429e-01 -9.25880671e-01
-1.30459225e+00 8.05708945e-01 8.59916270e-01 1.20555684e-02
1.70586777e+00 -2.24137887e-01 9.05508220e-01 -5.78690708e-01
1.73568264e-01 -8.68216157e-01 7.60595053e-02 -2.71618608e-02
9.84128952e-01 -7.93760717e-01 -1.36726111e-01 -5.70726275e-01
-4.34409648e-01 1.00511181e+00 2.92351305e-01 -1.25117630e-01
3.93531352e-01 3.06519955e-01 -7.96140060e-02 5.51531352e-02
-2.27975413e-01 -1.82665765e-01 6.17835373e-02 8.13018024e-01
5.28851487e-02 -2.42894515e-01 -3.87794495e-01 8.38975251e-01
-8.14956203e-02 6.67056143e-02 3.81505162e-01 3.99657160e-01
-1.76033184e-01 -1.13026416e+00 -7.34759510e-01 1.08759321e-01
-5.94618857e-01 -3.21683675e-01 1.79268673e-01 3.18876207e-01
5.39254963e-01 1.06691372e+00 -1.10580653e-01 -1.88486606e-01
4.74227704e-02 -7.98343495e-03 6.59901857e-01 -6.73451126e-02
-4.03982878e-01 3.65254521e-01 -3.57515782e-01 -3.57955158e-01
-5.65931976e-01 -4.22375917e-01 -8.19641113e-01 -4.60977018e-01
-3.31713319e-01 4.82393466e-02 6.11417592e-01 8.10041070e-01
3.47115576e-01 6.17043614e-01 8.13808382e-01 -6.50205910e-01
-5.00194609e-01 -1.00901711e+00 -9.49235916e-01 4.47941959e-01
2.56108522e-01 -5.22160649e-01 -7.64018595e-01 2.87105173e-01] | [11.684853553771973, -1.7530957460403442] |
46e4b8c9-1327-4aeb-9ff5-ca76d88a49f7 | delivering-speaking-style-in-low-resource | 2211.08857 | null | https://arxiv.org/abs/2211.08857v2 | https://arxiv.org/pdf/2211.08857v2.pdf | Delivering Speaking Style in Low-resource Voice Conversion with Multi-factor Constraints | Conveying the linguistic content and maintaining the source speech's speaking style, such as intonation and emotion, is essential in voice conversion (VC). However, in a low-resource situation, where only limited utterances from the target speaker are accessible, existing VC methods are hard to meet this requirement and capture the target speaker's timber. In this work, a novel VC model, referred to as MFC-StyleVC, is proposed for the low-resource VC task. Specifically, speaker timbre constraint generated by clustering method is newly proposed to guide target speaker timbre learning in different stages. Meanwhile, to prevent over-fitting to the target speaker's limited data, perceptual regularization constraints explicitly maintain model performance on specific aspects, including speaking style, linguistic content, and speech quality. Besides, a simulation mode is introduced to simulate the inference process to alleviate the mismatch between training and inference. Extensive experiments performed on highly expressive speech demonstrate the superiority of the proposed method in low-resource VC. | ['Yuping Wang', 'Qiao Tian', 'Yuanzhe Chen', 'Lei Xie', 'Xinsheng Wang', 'Zhichao Wang'] | 2022-11-16 | null | null | null | null | ['voice-conversion', 'voice-conversion'] | ['audio', 'speech'] | [-1.82356425e-02 -3.07069182e-01 -2.39971548e-01 -3.53324920e-01
-5.99960923e-01 -2.67766476e-01 1.70036957e-01 -2.63862044e-01
-1.55840456e-01 5.56913793e-01 4.29346383e-01 -2.21714661e-01
1.11665845e-01 -3.71996939e-01 -2.36463457e-01 -6.90314233e-01
6.57112777e-01 -2.39724725e-01 -1.71637133e-01 -9.47655141e-02
-2.08085068e-02 2.65621215e-01 -1.51804841e+00 -3.72453593e-02
1.13873231e+00 9.80507076e-01 7.36124039e-01 5.75785875e-01
-2.96414495e-01 5.13966680e-01 -7.45982945e-01 -1.28830709e-02
-2.08814248e-01 -7.47591197e-01 -3.43469322e-01 3.56170952e-01
-1.30295441e-01 -1.26200035e-01 -9.21427906e-02 1.20247447e+00
8.93125415e-01 5.47759950e-01 5.99418759e-01 -8.76328707e-01
-5.80492616e-01 6.27488434e-01 -1.93528399e-01 2.10160166e-01
2.36572251e-01 5.36059774e-02 8.29498231e-01 -8.83016229e-01
1.63049519e-01 1.52568984e+00 3.14158887e-01 7.56916940e-01
-1.04435349e+00 -8.66736352e-01 3.97183836e-01 1.75373554e-01
-1.44080007e+00 -9.65757012e-01 1.24860489e+00 -1.91432327e-01
4.64258343e-01 4.61495608e-01 4.79512483e-01 9.54831123e-01
-4.70975101e-01 7.22541392e-01 9.26632583e-01 -5.43529332e-01
1.42030835e-01 4.86856729e-01 -6.89921826e-02 1.86581895e-01
-4.50538754e-01 1.14807658e-01 -4.76062894e-01 1.36405066e-01
6.07283473e-01 -5.06197333e-01 -6.34483814e-01 1.68008968e-01
-9.36165750e-01 7.41793036e-01 2.18924247e-02 4.46885556e-01
-6.92363530e-02 -2.40638465e-01 5.99549830e-01 1.57962292e-01
5.14984548e-01 -5.24649583e-02 -2.54814297e-01 -3.24394941e-01
-8.84050012e-01 -2.33503595e-01 5.91915190e-01 1.15397489e+00
2.20650598e-01 8.03604424e-01 -2.97430813e-01 1.42952275e+00
6.28701925e-01 7.25528061e-01 6.73140645e-01 -7.75618076e-01
5.24753571e-01 9.72605683e-03 4.03409638e-02 -9.98967648e-01
-7.43661225e-02 -7.39732385e-01 -1.05222237e+00 -2.92320222e-01
-2.14913473e-01 -3.38037133e-01 -4.94568020e-01 1.98961747e+00
3.66184443e-01 1.98223308e-01 3.76542181e-01 1.12045074e+00
9.06347573e-01 1.13699841e+00 -3.85592356e-02 -1.14095628e+00
1.17897797e+00 -8.58789623e-01 -1.30943477e+00 -1.48920402e-01
-1.19562291e-01 -1.06777990e+00 1.63393354e+00 2.61363178e-01
-1.09354281e+00 -1.12700045e+00 -1.03189743e+00 1.33540541e-01
2.39783585e-01 5.36106706e-01 2.95418113e-01 1.00421250e+00
-5.01350224e-01 1.08317442e-01 -4.68270928e-01 8.38415772e-02
3.26443426e-02 1.03211306e-01 2.69539475e-01 2.49952987e-01
-1.43154800e+00 4.34931397e-01 2.53839403e-01 4.25870031e-01
-7.18859911e-01 -5.68846583e-01 -6.57345831e-01 1.35954067e-01
2.86595345e-01 -2.02592790e-01 1.32821822e+00 -1.15384459e+00
-2.20927858e+00 2.70218790e-01 -4.43683654e-01 4.39354219e-02
4.44849402e-01 -1.27076551e-01 -9.19342518e-01 3.78809199e-02
-2.67894149e-01 2.92357683e-01 1.17323661e+00 -1.27667630e+00
-5.77055454e-01 6.76855892e-02 -4.80023682e-01 5.67036569e-01
-6.47171855e-01 1.74187988e-01 -6.99953437e-01 -7.85221994e-01
1.11699164e-01 -5.87778091e-01 2.19072104e-01 -2.50403494e-01
-4.24864292e-01 -1.78385884e-01 8.23422670e-01 -8.13023806e-01
1.41759372e+00 -2.62682438e+00 1.94208384e-01 3.02933693e-01
-3.39492172e-01 4.44004834e-01 1.44936666e-01 -1.74080431e-02
2.07436606e-01 -1.31416053e-01 -2.52731711e-01 -5.60024142e-01
1.18048623e-01 1.84165090e-01 -3.74467283e-01 2.87040323e-01
5.49523570e-02 1.67038754e-01 -6.37349188e-01 -8.01660538e-01
3.27902734e-01 6.69204414e-01 -7.15400159e-01 5.35796583e-01
-8.52996930e-02 6.79905474e-01 -2.57645845e-01 2.89934009e-01
7.31052101e-01 4.56779897e-01 6.19874857e-02 -3.56119275e-01
-1.73652008e-01 3.25429618e-01 -1.33333623e+00 1.46551895e+00
-9.00647521e-01 4.04742748e-01 7.68393517e-01 -7.32856631e-01
1.27290559e+00 6.12464011e-01 1.27980486e-01 -5.88631153e-01
2.45428488e-01 1.76501915e-01 4.06600118e-01 -5.31378329e-01
3.81397992e-01 -4.20232713e-01 1.88909084e-01 1.45255372e-01
-1.31227121e-01 -4.24956679e-01 -1.32927597e-01 -3.64349127e-01
4.89028357e-02 -1.37890726e-01 1.24589063e-01 -2.32826233e-01
8.40057552e-01 -5.29846430e-01 9.37306046e-01 9.95833650e-02
-4.01678234e-01 5.11055470e-01 -5.13691343e-02 6.53046727e-01
-6.77720249e-01 -1.00912857e+00 -2.03906387e-01 1.12808013e+00
2.19354078e-01 -1.29824430e-01 -8.40986669e-01 2.89001279e-02
-3.98079872e-01 1.00291777e+00 7.48888850e-02 -4.30229306e-01
-4.67809945e-01 -4.44049567e-01 6.47480309e-01 3.10136676e-01
5.08739293e-01 -1.16148710e+00 -2.22306430e-01 3.75049949e-01
-5.76104581e-01 -9.53267992e-01 -1.05382085e+00 -1.65394887e-01
-5.20806730e-01 -4.53260154e-01 -6.67846978e-01 -1.14310980e+00
4.58370417e-01 1.73316181e-01 4.23859477e-01 -1.23713380e-02
2.44741410e-01 1.65899005e-02 -3.34982961e-01 -3.35004687e-01
-7.01972246e-01 -3.25147361e-02 4.69347715e-01 2.31454462e-01
5.65010272e-02 -5.55042326e-01 -3.46748233e-01 2.94481337e-01
-6.48564577e-01 1.33281037e-01 3.10968220e-01 8.69248390e-01
5.75834870e-01 5.69148421e-01 1.00328505e+00 -4.51700151e-01
9.68509793e-01 -1.95283771e-01 -3.66466880e-01 -1.18150748e-02
-3.43366832e-01 -4.09665555e-01 1.12257874e+00 -1.01274896e+00
-1.63906908e+00 -1.85375690e-01 -5.02464116e-01 -5.69645166e-01
-1.29777104e-01 5.63474953e-01 -9.51507628e-01 4.56485569e-01
1.70550168e-01 5.04293561e-01 8.39016512e-02 -6.05803788e-01
2.92722672e-01 1.30535328e+00 7.75905967e-01 -5.90428412e-01
6.98417008e-01 -1.38078749e-01 -5.96640706e-01 -1.12805796e+00
-6.35121703e-01 -3.04125875e-01 -3.27416778e-01 -2.55909503e-01
5.64905047e-01 -9.52702284e-01 -8.18104565e-01 4.92546916e-01
-1.07369173e+00 -1.01343922e-01 -9.31847766e-02 1.05718064e+00
-5.78160405e-01 3.69920790e-01 -6.42414689e-01 -1.27916396e+00
-4.66033697e-01 -1.19644964e+00 5.40614009e-01 3.68002594e-01
7.42824599e-02 -7.80867159e-01 -1.81913257e-01 2.80902952e-01
5.37258804e-01 -4.04925853e-01 1.04516912e+00 -3.33382636e-01
1.49843588e-01 3.68782878e-01 -1.95472613e-02 8.11365902e-01
6.40123844e-01 7.06473738e-02 -1.26145661e+00 -2.25591093e-01
5.28605580e-01 -1.15568943e-01 2.88014621e-01 3.06510389e-01
1.16436136e+00 -4.49346185e-01 9.57499146e-02 4.74597812e-01
9.31251347e-01 5.10721684e-01 3.58570993e-01 -4.08221602e-01
6.14551842e-01 6.82341278e-01 7.90697396e-01 4.50886369e-01
1.37424275e-01 6.15078807e-01 -7.41244555e-02 -1.40026003e-01
-3.43673736e-01 -4.05886561e-01 4.96064931e-01 1.82596266e+00
2.50087827e-01 -2.07603589e-01 -3.28355163e-01 2.39599988e-01
-1.35293448e+00 -8.09925377e-01 1.95146874e-01 2.08531523e+00
1.41144311e+00 3.22965235e-01 -1.43603548e-01 4.06532884e-01
1.08637202e+00 2.06856683e-01 -6.30241752e-01 -4.37149197e-01
-1.11732639e-01 -1.12982474e-01 -2.88068831e-01 7.67204225e-01
-7.34563828e-01 9.83504295e-01 5.66775942e+00 1.38446891e+00
-1.46696663e+00 2.04336867e-01 4.70796406e-01 -7.62292668e-02
-4.68819648e-01 -2.72705466e-01 -6.68483913e-01 7.15245724e-01
7.49034882e-01 -3.81471902e-01 7.35985518e-01 7.80475855e-01
8.60337436e-01 3.23387980e-01 -8.43759120e-01 1.19585598e+00
1.01086982e-01 -7.16614604e-01 -1.29746482e-01 -3.63589883e-01
2.75960118e-01 -6.25001252e-01 2.11102083e-01 4.50485617e-01
-3.33446413e-01 -8.77400339e-01 9.62414503e-01 3.78643006e-01
1.23145866e+00 -9.83329296e-01 4.35120374e-01 5.18835366e-01
-1.45832276e+00 4.19247262e-02 -4.29179221e-01 1.61953479e-01
1.81522459e-01 3.32440406e-01 -5.38837373e-01 5.10876894e-01
2.35338748e-01 5.16315699e-01 -1.00426234e-01 6.06796801e-01
-1.27608404e-01 1.09479165e+00 -2.45150894e-01 -1.08663894e-01
-1.20430067e-01 -2.54068106e-01 6.55784190e-01 1.32149184e+00
2.99871236e-01 2.36256570e-01 1.89046010e-01 9.10070658e-01
1.10932235e-02 6.35372519e-01 -8.28556940e-02 -9.83242244e-02
9.83503759e-01 9.71957624e-01 -1.90990210e-01 -5.95082180e-04
-1.52058631e-01 8.42796743e-01 -2.67770160e-02 5.69555998e-01
-9.87480700e-01 -4.70663816e-01 5.97931445e-01 -3.25908273e-01
1.97366893e-01 -1.29572153e-01 -8.35038573e-02 -9.01902795e-01
-6.62450343e-02 -1.06279600e+00 -1.13693930e-01 -7.36342788e-01
-1.08947802e+00 5.69949090e-01 -2.97736973e-01 -1.39807415e+00
-2.64885664e-01 -2.13409975e-01 -8.72278810e-01 1.16230547e+00
-1.49325871e+00 -1.02460170e+00 -6.06064498e-02 7.56958902e-01
8.87346983e-01 -3.09163243e-01 7.51018643e-01 5.20079494e-01
-8.51562142e-01 9.11821723e-01 2.75955021e-01 -2.18016163e-01
7.24676847e-01 -8.33467245e-01 -3.19596827e-01 7.28851140e-01
-1.56683773e-01 7.23898292e-01 8.01223099e-01 -3.11222851e-01
-1.17698014e+00 -1.07277191e+00 5.17537773e-01 5.27870834e-01
3.05616975e-01 -4.42131877e-01 -1.00662839e+00 1.98213667e-01
6.54156581e-02 -2.66982853e-01 8.31320167e-01 2.53194328e-02
-2.59091370e-02 -4.31465179e-01 -9.87874448e-01 7.14395165e-01
7.68418491e-01 -7.88280487e-01 -5.73103309e-01 2.91052330e-02
1.10167754e+00 -4.58581716e-01 -7.97705591e-01 2.41425410e-01
3.30033988e-01 -5.64838231e-01 7.93865383e-01 -2.25957856e-01
-2.65175803e-03 -4.93082881e-01 -4.79380637e-01 -1.60621047e+00
-1.50830090e-01 -8.46813500e-01 6.83302358e-02 1.99925756e+00
2.12258816e-01 -4.19158846e-01 2.02278063e-01 1.64823487e-01
-4.68071610e-01 -2.63979256e-01 -9.57310021e-01 -7.16073632e-01
-1.05504282e-01 -5.61545312e-01 6.38011932e-01 9.16679919e-01
6.63019195e-02 6.50173008e-01 -5.71789205e-01 2.62670010e-01
2.88134545e-01 5.00117727e-02 5.48737228e-01 -9.53783095e-01
-3.40189129e-01 -4.56133008e-01 2.90557861e-01 -1.23096907e+00
4.80568945e-01 -5.70761144e-01 5.31961024e-01 -1.12652349e+00
-1.63227811e-01 -5.59092462e-01 -3.72368008e-01 -1.60765618e-01
-3.51286650e-01 -2.98299670e-01 2.36109465e-01 2.04302639e-01
-2.94266734e-02 1.19589615e+00 1.48812473e+00 -1.24369353e-01
-4.70755219e-01 4.16614562e-01 -5.41797459e-01 7.27822423e-01
7.10851908e-01 -1.19210012e-01 -8.95110190e-01 -2.20607236e-01
-5.24177074e-01 4.78622913e-01 -1.26592025e-01 -8.29660058e-01
1.83619902e-01 -3.77861142e-01 9.79796350e-02 -5.94708681e-01
7.44040668e-01 -8.70292008e-01 3.78152281e-02 3.22993159e-01
-5.14028072e-01 -5.22293925e-01 2.72277862e-01 4.61538821e-01
-4.58234340e-01 -2.49289334e-01 1.05357659e+00 2.85386741e-01
-3.13212574e-01 1.05027996e-01 -4.22902673e-01 1.14093959e-01
5.75816929e-01 3.88657600e-02 -6.95217960e-03 -5.75670958e-01
-6.44028068e-01 1.46758735e-01 4.78623956e-02 3.81370515e-01
7.31094897e-01 -1.50779426e+00 -7.10310161e-01 3.98768902e-01
-4.04427610e-02 -4.19628806e-02 5.83509803e-01 7.13620722e-01
-7.94598833e-02 3.71711314e-01 8.25413465e-02 -4.82632667e-01
-1.33587766e+00 5.56160331e-01 4.17016178e-01 4.04578447e-01
-4.12297517e-01 7.02275753e-01 1.78828642e-01 -5.57705760e-01
5.41275501e-01 -3.25837821e-01 -3.68943214e-01 4.54355776e-02
4.70357060e-01 2.38046616e-01 -1.19386084e-01 -9.46511686e-01
-3.21789593e-01 3.60774517e-01 1.82502031e-01 -3.82105559e-01
8.04139018e-01 -7.21132159e-01 5.07106222e-02 8.04233968e-01
1.11093616e+00 5.95093608e-01 -1.28948796e+00 -3.62534404e-01
-4.78810906e-01 -5.16616821e-01 3.76914203e-01 -5.50440311e-01
-9.74876344e-01 1.07615292e+00 6.13307059e-01 -8.00614581e-02
1.21635890e+00 -3.90881836e-01 8.24605882e-01 2.47805968e-01
1.47765419e-02 -1.50624359e+00 1.14544496e-01 3.90239388e-01
8.80598843e-01 -9.78596628e-01 -4.34361279e-01 -5.14698625e-01
-8.38298321e-01 8.71239364e-01 8.28296244e-01 3.36886376e-01
7.60957479e-01 1.15322694e-01 3.05024683e-01 4.78912652e-01
-6.38385594e-01 -1.28813982e-01 2.23984182e-01 6.51542783e-01
5.68169296e-01 1.91787004e-01 -1.60843685e-01 9.80711102e-01
-5.58594704e-01 -4.24382150e-01 3.45946580e-01 3.68839234e-01
-5.79215825e-01 -6.94872677e-01 -5.53019285e-01 6.97051957e-02
-4.31093663e-01 -2.18299121e-01 -1.17291868e-01 4.04553950e-01
6.38001561e-02 1.57888234e+00 9.30693522e-02 -3.99536222e-01
4.13576394e-01 3.71607319e-02 1.37872919e-01 -6.64773405e-01
-3.25111926e-01 9.06840503e-01 3.40206400e-02 -8.09039641e-03
-5.36853790e-01 -4.76908177e-01 -1.52322137e+00 -3.07129800e-01
-7.12846279e-01 5.29644310e-01 6.14961386e-01 9.34990942e-01
-1.03962131e-01 8.81632149e-01 1.19658947e+00 -4.81693208e-01
-7.20710933e-01 -1.30662954e+00 -9.06983018e-01 2.67652541e-01
4.05617744e-01 -4.60610420e-01 -4.96683508e-01 1.64577112e-01] | [14.909741401672363, 6.450470447540283] |
4699bcac-3887-4f96-8881-44a546a63066 | unified-named-entity-recognition-as-word-word | 2112.10070 | null | https://arxiv.org/abs/2112.10070v1 | https://arxiv.org/pdf/2112.10070v1.pdf | Unified Named Entity Recognition as Word-Word Relation Classification | So far, named entity recognition (NER) has been involved with three major types, including flat, overlapped (aka. nested), and discontinuous NER, which have mostly been studied individually. Recently, a growing interest has been built for unified NER, tackling the above three jobs concurrently with one single model. Current best-performing methods mainly include span-based and sequence-to-sequence models, where unfortunately the former merely focus on boundary identification and the latter may suffer from exposure bias. In this work, we present a novel alternative by modeling the unified NER as word-word relation classification, namely W^2NER. The architecture resolves the kernel bottleneck of unified NER by effectively modeling the neighboring relations between entity words with Next-Neighboring-Word (NNW) and Tail-Head-Word-* (THW-*) relations. Based on the W^2NER scheme we develop a neural framework, in which the unified NER is modeled as a 2D grid of word pairs. We then propose multi-granularity 2D convolutions for better refining the grid representations. Finally, a co-predictor is used to sufficiently reason the word-word relations. We perform extensive experiments on 14 widely-used benchmark datasets for flat, overlapped, and discontinuous NER (8 English and 6 Chinese datasets), where our model beats all the current top-performing baselines, pushing the state-of-the-art performances of unified NER. | ['Fei Li', 'Donghong Ji', 'Chong Teng', 'Meishan Zhang', 'Shengqiong Wu', 'Jiang Liu', 'Hao Fei', 'Jingye Li'] | 2021-12-19 | null | null | null | null | ['relation-classification', 'nested-named-entity-recognition', 'chinese-named-entity-recognition'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [-7.69518986e-02 -9.06748250e-02 -1.29607946e-01 -2.37334758e-01
-9.21515703e-01 -5.81913650e-01 4.03659672e-01 2.45608553e-01
-7.56313384e-01 7.98219621e-01 4.84610945e-01 -6.91656530e-01
-9.06798765e-02 -8.98608088e-01 -4.57446069e-01 -5.63387513e-01
7.39943981e-02 4.46193278e-01 1.95275515e-01 -3.08929741e-01
6.92411065e-02 2.50811994e-01 -1.12547970e+00 1.51118577e-01
1.21754873e+00 8.27142596e-01 1.31672397e-01 4.96475101e-01
-7.07867324e-01 4.39639360e-01 -6.51022851e-01 -7.41071939e-01
3.49265859e-02 -4.04468536e-01 -1.14479089e+00 -4.99675632e-01
2.64698386e-01 1.48157358e-01 -2.26963326e-01 9.74485219e-01
7.87977040e-01 5.41212499e-01 4.71605808e-01 -6.27789199e-01
-1.03686857e+00 9.22923565e-01 -5.93973994e-01 3.52376580e-01
2.43275061e-01 -3.38657230e-01 1.43324113e+00 -1.10791171e+00
5.21716714e-01 1.07337999e+00 1.10842872e+00 4.88385230e-01
-1.07278049e+00 -5.94568312e-01 2.69825935e-01 2.08006769e-01
-1.72064388e+00 -2.26607695e-01 2.78361171e-01 -4.66726981e-02
1.60712886e+00 2.73485392e-01 7.18174651e-02 9.13377106e-01
7.74477050e-02 5.61898291e-01 8.81999612e-01 -5.34372389e-01
2.72476152e-02 -4.88184482e-01 6.48444891e-01 4.48110074e-01
3.01075101e-01 -1.61305651e-01 -2.33726844e-01 -1.12209216e-01
6.54828608e-01 -1.85986683e-01 -3.68355483e-01 4.15974706e-01
-1.11478961e+00 6.22532845e-01 5.08875906e-01 7.12094665e-01
-4.80689049e-01 -2.96763629e-01 4.23106194e-01 -5.48440591e-02
6.57932580e-01 3.55351448e-01 -9.26976025e-01 -1.36626050e-01
-8.24382186e-01 -5.81063144e-03 9.79839504e-01 1.00539446e+00
6.86153591e-01 -2.00283244e-01 -5.43344915e-01 1.11779404e+00
6.42981753e-02 8.49540457e-02 6.64327562e-01 -1.32383510e-01
8.17155182e-01 7.01571167e-01 1.77438572e-01 -6.20984852e-01
-6.55017734e-01 -3.84198129e-01 -1.20252812e+00 -4.38367397e-01
3.98816139e-01 -5.16545832e-01 -1.11684287e+00 1.63535821e+00
2.69131452e-01 5.64380407e-01 2.43053719e-01 7.16026127e-01
1.22988904e+00 7.61728823e-01 4.25623059e-01 -3.08110118e-02
1.86291039e+00 -1.20700467e+00 -9.39148903e-01 -1.56590834e-01
8.81258190e-01 -6.57465398e-01 9.17517900e-01 -1.93226829e-01
-9.35536325e-01 -6.45440996e-01 -6.79537833e-01 -3.95678073e-01
-9.71949637e-01 2.69261390e-01 3.66614550e-01 5.58794439e-01
-1.01071548e+00 6.76056921e-01 -6.57959223e-01 -4.83921200e-01
3.81625667e-02 2.59712279e-01 -5.22004247e-01 2.97511108e-02
-1.90091991e+00 1.23182547e+00 7.39913642e-01 5.42014718e-01
-3.17044854e-02 -7.98243821e-01 -1.04256427e+00 2.73292810e-01
2.75329232e-01 -6.66987956e-01 1.01689541e+00 -9.33512449e-02
-1.15565872e+00 8.51730645e-01 -4.64672625e-01 -3.06440324e-01
1.10001862e-01 -5.48535705e-01 -9.11818624e-01 -4.67652112e-01
1.10917285e-01 3.66145611e-01 3.39318849e-02 -1.08616054e+00
-7.19796240e-01 -1.37204647e-01 -2.31994212e-01 1.85808599e-01
-3.33160967e-01 4.99917269e-01 -4.20915365e-01 -7.56363988e-01
-3.14980485e-02 -5.10074496e-01 -3.85383874e-01 -8.96996677e-01
-7.16290414e-01 -8.93364608e-01 4.05883461e-01 -7.88795173e-01
1.94505835e+00 -1.88037384e+00 -4.10132855e-02 -8.69785994e-02
8.14544335e-02 6.83477104e-01 -2.92455137e-01 7.89535105e-01
-4.48743522e-01 5.89305639e-01 -2.07796574e-01 -5.79117715e-01
1.11133285e-01 3.33725244e-01 -3.36985797e-01 4.52819467e-02
4.56012368e-01 1.22609830e+00 -8.54644954e-01 -5.04757524e-01
-2.48175874e-01 5.79981863e-01 6.44129515e-02 4.36092198e-01
1.88702956e-01 1.16011187e-01 -2.19991207e-01 4.93003309e-01
7.97242165e-01 -1.75656870e-01 2.89175093e-01 -2.66480267e-01
-5.12612224e-01 7.17253327e-01 -1.21889126e+00 1.56319880e+00
-5.10574758e-01 5.81920035e-02 -5.70056587e-02 -7.13793635e-01
1.09521329e+00 5.77087939e-01 6.91826046e-02 -4.81509835e-01
7.29552507e-02 3.70173305e-01 -2.23336145e-02 -3.84020388e-01
8.25788975e-01 -5.93962260e-02 -2.49094605e-01 2.39597149e-02
1.86005011e-01 4.01097655e-01 2.76892632e-01 -2.84501761e-01
1.28849399e+00 2.01469496e-01 4.58975911e-01 -3.21500182e-01
4.03051794e-01 -1.72605127e-01 9.02028978e-01 6.55785978e-01
-2.33401775e-01 6.80045128e-01 4.31028724e-01 -4.36596423e-01
-7.83638000e-01 -8.60733271e-01 -2.26833180e-01 1.21610796e+00
2.95672536e-01 -4.67589080e-01 -7.57306218e-01 -9.60638821e-01
-3.72938424e-01 8.22770655e-01 -6.82955205e-01 1.49905458e-01
-1.10410023e+00 -8.94669235e-01 1.09826410e+00 9.33184564e-01
6.57339334e-01 -1.32641876e+00 -1.56511724e-01 5.32110929e-01
-4.05921817e-01 -1.24056530e+00 -9.13997889e-01 7.80058265e-01
-6.12597346e-01 -7.54011989e-01 -8.73802602e-01 -1.13234448e+00
3.12085152e-01 1.21924482e-01 1.35451853e+00 1.21979984e-02
-5.13363667e-02 -3.03048015e-01 -7.57626653e-01 -2.82254964e-02
1.99515581e-01 6.50084615e-01 -3.70608307e-02 -1.60043240e-01
5.76626003e-01 -4.31114018e-01 -4.06691551e-01 3.86723965e-01
-8.85492921e-01 -3.07502300e-01 7.54195929e-01 1.21027207e+00
5.42406499e-01 -2.24064980e-02 6.65679932e-01 -9.00068820e-01
6.84556842e-01 -6.91896439e-01 -1.48889989e-01 4.85573351e-01
-5.07941306e-01 1.27630934e-01 7.50438929e-01 -5.87340415e-01
-1.28408563e+00 -1.60989761e-01 -5.98607361e-01 -2.27642268e-01
-5.47999680e-01 8.39495718e-01 -5.49995363e-01 4.14162219e-01
3.71049702e-01 1.85879275e-01 -8.31077397e-01 -8.64357948e-01
6.23631954e-01 7.09515810e-01 6.59541428e-01 -7.28237808e-01
5.90354264e-01 -8.66218507e-02 -4.50004458e-01 -6.76596701e-01
-1.12379539e+00 -7.40474463e-01 -8.58709335e-01 4.21337396e-01
1.21475005e+00 -7.49514997e-01 -3.90225768e-01 4.21502888e-01
-1.66302335e+00 -3.18217486e-01 -1.32571802e-01 2.78423727e-01
1.95648298e-01 4.54000592e-01 -1.04374051e+00 -8.13475847e-01
-6.53523028e-01 -8.56119752e-01 1.09947562e+00 5.31165242e-01
-3.00397903e-01 -1.15794599e+00 2.74021149e-01 8.28486308e-02
4.38758522e-01 -1.03388675e-01 1.10134840e+00 -1.31246901e+00
6.74833879e-02 8.05966333e-02 -5.61154604e-01 8.28415975e-02
5.36649637e-02 -4.09895182e-01 -7.08207250e-01 1.11286297e-01
-4.64804858e-01 -1.10222034e-01 1.05052733e+00 9.91791338e-02
7.91422188e-01 2.96079386e-02 -4.39984888e-01 6.20646000e-01
1.24933851e+00 2.33709782e-01 7.79975355e-01 4.05783772e-01
8.15502524e-01 5.57775080e-01 4.83070791e-01 2.13048831e-01
7.40079284e-01 4.81650263e-01 2.25286007e-01 -3.31307918e-01
-1.19446233e-01 -2.58319110e-01 1.73850417e-01 1.07679880e+00
-2.88114727e-01 -6.12933040e-01 -1.17047441e+00 7.55872250e-01
-1.94595504e+00 -7.83347428e-01 -4.66018885e-01 1.80134892e+00
9.09299016e-01 9.19746142e-03 -2.50778973e-01 -8.81834328e-02
1.25760341e+00 2.19481841e-01 -2.88371295e-01 -5.23750842e-01
-3.31616849e-01 7.51157761e-01 4.07011837e-01 2.62242585e-01
-1.55066466e+00 1.28782046e+00 5.52189064e+00 1.13739908e+00
-6.99523687e-01 1.50150850e-01 5.32287061e-01 7.44909823e-01
-1.74721643e-01 -1.25610083e-02 -1.45966911e+00 2.28908911e-01
9.54917192e-01 2.63227910e-01 1.22764722e-01 5.71076334e-01
-2.99916625e-01 2.72878945e-01 -8.76504302e-01 8.99350703e-01
-1.49727106e-01 -1.12471783e+00 -1.03459999e-01 -2.26123221e-02
6.94627345e-01 6.48587123e-02 -4.13052946e-01 7.13960350e-01
4.87535387e-01 -1.19382679e+00 4.99420941e-01 4.13130462e-01
9.17029083e-01 -6.91376865e-01 1.28607380e+00 3.38015318e-01
-1.93718648e+00 3.10651839e-01 -2.49984264e-01 4.00060154e-02
3.43718946e-01 6.30492508e-01 -1.69833750e-01 1.23214996e+00
6.28607213e-01 2.94585675e-01 -8.89251232e-02 1.00757813e+00
-3.51685911e-01 7.46988177e-01 -2.99195170e-01 -1.11933433e-01
2.78776914e-01 -2.08149135e-01 2.40099341e-01 1.77905822e+00
4.64526355e-01 3.57841820e-01 2.14898333e-01 7.55749524e-01
-2.01738879e-01 3.60236704e-01 -8.35099369e-02 7.60271549e-02
7.75898755e-01 1.50109923e+00 -9.04500067e-01 -1.41511470e-01
-5.70698798e-01 1.29208314e+00 8.68021905e-01 3.61993045e-01
-8.30808461e-01 -1.00764990e+00 7.78401375e-01 -4.99957919e-01
7.65401840e-01 -3.28648180e-01 -3.67159516e-01 -1.33970332e+00
-1.60383612e-01 -4.58651334e-01 7.63533056e-01 -3.71239573e-01
-1.91874802e+00 1.13837135e+00 -2.41743147e-01 -7.55422294e-01
1.68165654e-01 -7.22492516e-01 -8.88083041e-01 1.31198561e+00
-1.74232340e+00 -1.24015999e+00 -4.08767536e-02 3.01855534e-01
4.69370008e-01 2.36909837e-01 1.19250941e+00 5.80249190e-01
-1.05464339e+00 8.90636921e-01 -2.13731676e-02 7.24395692e-01
7.62937248e-01 -1.35602200e+00 8.69129956e-01 8.41220438e-01
3.82724196e-01 8.29220474e-01 1.89917400e-01 -8.33013773e-01
-1.02769649e+00 -1.18332124e+00 1.63590336e+00 -4.11576033e-01
8.94820929e-01 -3.85659307e-01 -1.14812720e+00 5.92842221e-01
4.77663696e-01 3.42198700e-01 8.22363555e-01 5.45798123e-01
-5.76143563e-01 2.17275307e-01 -7.66847610e-01 7.34392107e-01
1.24621332e+00 -4.08593893e-01 -9.68218744e-01 -2.04736039e-01
9.96935189e-01 -4.36065555e-01 -1.05304396e+00 4.31480139e-01
2.22390532e-01 -7.13629484e-01 8.92445385e-01 -8.08508217e-01
2.28215471e-01 -3.26444954e-01 -2.13830750e-02 -1.51738155e+00
-5.84247112e-01 -4.90876555e-01 -1.69528961e-01 1.82487631e+00
6.70650780e-01 -5.96586585e-01 5.06399930e-01 2.81984001e-01
-5.42605579e-01 -9.65559840e-01 -1.03247237e+00 -8.11370194e-01
3.76990885e-01 -3.49555999e-01 8.59545588e-01 1.14273703e+00
8.68341550e-02 6.32627010e-01 -2.10374862e-01 2.67128348e-01
1.09610938e-01 5.35236821e-02 3.37013602e-02 -1.14520872e+00
8.28298647e-03 -6.47866666e-01 -6.53358772e-02 -1.39096296e+00
3.68175834e-01 -9.36041236e-01 3.51929247e-01 -1.70721459e+00
8.44031572e-02 -4.28437471e-01 -7.22533047e-01 7.97685683e-01
-8.88532758e-01 1.88349366e-01 -2.37216093e-02 7.85874650e-02
-7.32475579e-01 5.42904377e-01 8.04743290e-01 3.75639290e-01
-2.49291927e-01 -3.08143824e-01 -6.53252780e-01 6.06771052e-01
6.61964953e-01 -2.92717218e-01 2.55585968e-01 -6.16808534e-01
2.64757395e-01 -1.47531182e-01 -1.02093942e-01 -6.75862372e-01
5.37285745e-01 1.17937744e-01 2.09490150e-01 -6.43659770e-01
-8.03511739e-02 -6.04629993e-01 -1.10032119e-01 -8.84844586e-02
-2.22779006e-01 2.61488140e-01 1.11668535e-01 4.39952582e-01
-3.61466855e-01 -2.67468691e-01 4.19878960e-01 1.87745690e-02
-8.24442744e-01 3.57699901e-01 -1.34086013e-01 4.71517384e-01
7.70874977e-01 -6.32820278e-02 -4.48094398e-01 1.03082001e-01
-7.01206148e-01 3.75777245e-01 -2.67112523e-01 4.23048764e-01
4.49150205e-01 -1.35635841e+00 -8.54594052e-01 -2.78612971e-02
1.02864996e-01 1.75987303e-01 3.49061251e-01 7.23648190e-01
-1.23653308e-01 3.31263632e-01 3.44502956e-01 -5.22952154e-02
-9.35794592e-01 5.04086018e-01 2.82820612e-01 -1.02012849e+00
-5.43467700e-01 1.11386156e+00 3.91635858e-02 -7.71526039e-01
1.97808743e-01 -4.70836759e-01 -4.59257364e-01 2.37864718e-01
5.13557553e-01 4.48139966e-01 2.85037726e-01 -8.15797269e-01
-4.35534954e-01 5.77780306e-01 -1.08860321e-01 1.99924260e-01
1.17512381e+00 -1.43041775e-01 -3.69989514e-01 4.19540465e-01
9.41596985e-01 6.85402974e-02 -6.91450775e-01 -4.02951747e-01
6.69331908e-01 2.66637504e-01 -1.56912223e-01 -8.20544779e-01
-6.63496017e-01 8.65173101e-01 1.88044030e-02 5.08982301e-01
1.01047385e+00 6.70563728e-02 1.40938497e+00 2.13315651e-01
2.31361702e-01 -8.89766514e-01 -5.82877398e-01 1.36827087e+00
3.77408445e-01 -1.01418102e+00 -6.08136952e-01 -4.91480589e-01
-5.82142413e-01 1.25692046e+00 7.49660492e-01 -2.43478082e-02
7.24564552e-01 5.86407363e-01 2.13247746e-01 9.61050913e-02
-5.53732336e-01 -7.84246922e-01 4.71122324e-01 3.48961085e-01
8.16870987e-01 2.50538960e-02 -6.52997255e-01 1.32511151e+00
2.51357686e-02 -3.12123179e-01 -2.51432568e-01 7.10778356e-01
-2.10267812e-01 -1.42234194e+00 -2.97554761e-01 2.74862111e-01
-7.58693635e-01 -7.77549565e-01 -2.01523766e-01 7.03912020e-01
4.69384611e-01 9.73949254e-01 2.08017051e-01 -4.86049294e-01
7.23591924e-01 5.27919173e-01 -1.67904086e-02 -6.40672266e-01
-1.11657584e+00 1.86681654e-02 4.45285022e-01 -3.37705910e-01
-2.10738003e-01 -3.69131029e-01 -1.57485092e+00 4.87757061e-04
-6.89840853e-01 4.66451049e-01 2.45770708e-01 1.11903882e+00
5.57173848e-01 6.38948441e-01 2.68589228e-01 -6.91544354e-01
-5.70180655e-01 -1.21725738e+00 -7.03225017e-01 3.48224699e-01
-9.71634930e-04 -5.73203921e-01 -1.91375718e-01 -3.20694774e-01] | [9.641348838806152, 9.532496452331543] |
8733f9a3-55e6-4170-b52a-8944cb1cccef | transition-based-semantic-dependency-parsing | 2005.13344 | null | https://arxiv.org/abs/2005.13344v2 | https://arxiv.org/pdf/2005.13344v2.pdf | Transition-based Semantic Dependency Parsing with Pointer Networks | Transition-based parsers implemented with Pointer Networks have become the new state of the art in dependency parsing, excelling in producing labelled syntactic trees and outperforming graph-based models in this task. In order to further test the capabilities of these powerful neural networks on a harder NLP problem, we propose a transition system that, thanks to Pointer Networks, can straightforwardly produce labelled directed acyclic graphs and perform semantic dependency parsing. In addition, we enhance our approach with deep contextualized word embeddings extracted from BERT. The resulting system not only outperforms all existing transition-based models, but also matches the best fully-supervised accuracy to date on the SemEval 2015 Task 18 English datasets among previous state-of-the-art graph-based parsers. | ['Carlos Gómez-Rodríguez', 'Daniel Fernández-González'] | 2020-05-27 | null | null | null | null | ['semantic-dependency-parsing'] | ['natural-language-processing'] | [ 3.99235114e-02 5.85881174e-01 -2.86657333e-01 -4.08523589e-01
-6.85620248e-01 -6.96904242e-01 5.01270771e-01 3.75909120e-01
-7.00357735e-01 5.60096323e-01 5.40368855e-01 -9.79407787e-01
-1.42619731e-02 -9.94154453e-01 -6.84537470e-01 -2.10141182e-01
-3.94773096e-01 9.22443390e-01 5.41187465e-01 -2.69886255e-01
1.14452146e-01 4.13153261e-01 -8.57466042e-01 1.04117267e-01
5.82652152e-01 4.22622472e-01 1.48661911e-01 8.59203935e-01
-5.72513819e-01 7.11098015e-01 -4.88053471e-01 -9.34385121e-01
-2.91590184e-01 -1.70802668e-01 -1.23824775e+00 -7.82800138e-01
4.43797767e-01 -5.39460741e-02 -1.92322224e-01 9.21274662e-01
3.78614873e-01 -8.06598645e-03 2.21551389e-01 -7.64905930e-01
-1.24559891e+00 1.47633934e+00 2.09381729e-02 5.66643476e-01
4.79653060e-01 -2.38285109e-01 1.55039752e+00 -5.79413176e-01
8.96115243e-01 1.50911856e+00 8.67989719e-01 9.22202289e-01
-1.25870848e+00 -5.84494770e-01 3.81021500e-01 7.72226825e-02
-4.27810580e-01 -1.99726880e-01 7.94442952e-01 7.27043003e-02
1.88831174e+00 -3.80943120e-01 5.02001762e-01 1.44216573e+00
3.82169336e-01 7.08838165e-01 9.67299283e-01 -7.93653965e-01
1.42611980e-01 -5.86297810e-01 7.36499667e-01 1.00026393e+00
4.96782511e-01 2.49942824e-01 -5.16354620e-01 1.54467985e-01
4.70478863e-01 -5.82770526e-01 1.15412951e-01 -1.44347027e-01
-9.34582949e-01 1.15312958e+00 5.23848772e-01 5.50633907e-01
-2.48075739e-01 3.51221442e-01 6.44858539e-01 1.58777177e-01
6.45138919e-01 6.58433020e-01 -1.03380346e+00 -2.00092465e-01
-5.44655442e-01 -2.63781864e-02 1.23259008e+00 9.95773435e-01
4.13253069e-01 1.34167343e-01 -1.01173818e-01 5.58326066e-01
4.35105085e-01 1.67174086e-01 4.31815207e-01 -7.01916337e-01
1.07253444e+00 6.26721442e-01 -4.11599189e-01 -4.18070436e-01
-9.32345331e-01 -1.42253771e-01 -3.38724375e-01 5.83598344e-03
5.97999752e-01 -4.36850995e-01 -1.30612409e+00 1.79748821e+00
5.22028096e-02 5.05555533e-02 4.35801953e-01 2.97716320e-01
1.14163017e+00 6.51581824e-01 8.59456718e-01 1.59997523e-01
1.55485380e+00 -9.49267864e-01 -6.78752422e-01 -6.95522726e-01
1.11457205e+00 -2.87770122e-01 8.22618484e-01 -1.23307500e-02
-1.07621992e+00 -5.51913083e-01 -1.07016230e+00 -1.94907397e-01
-6.72250867e-01 -2.99911886e-01 1.20633411e+00 1.10288358e+00
-1.51857471e+00 8.87796700e-01 -1.14238429e+00 -4.71854717e-01
6.72858298e-01 4.67114508e-01 -8.43556643e-01 -2.00937316e-01
-1.42199874e+00 1.21160710e+00 9.77478623e-01 1.10121705e-01
-5.35967708e-01 -7.18628585e-01 -1.46521938e+00 3.72901171e-01
3.36965472e-01 -5.52197635e-01 1.45543349e+00 -3.06432098e-01
-1.61584723e+00 7.86262691e-01 9.41270664e-02 -7.88961649e-01
-2.10282683e-01 -3.60967070e-01 -3.96429449e-01 7.03866482e-02
9.90079120e-02 8.60132515e-01 3.23644340e-01 -8.20130527e-01
-6.08692765e-01 -2.65792906e-01 2.32385635e-01 -1.44833893e-01
-3.24000955e-01 4.18424577e-01 -1.24791846e-01 -1.96115687e-01
-1.49756759e-01 -7.50094056e-01 -5.14963150e-01 -6.89892173e-01
-4.44974184e-01 -8.34066927e-01 5.79410911e-01 -7.87448704e-01
9.85190332e-01 -1.92536688e+00 1.68470949e-01 -3.32551360e-01
-3.14050801e-02 6.50371432e-01 -5.11040270e-01 6.00939691e-01
-3.57780397e-01 5.27274966e-01 -4.75864321e-01 -6.46970868e-01
1.07736334e-01 7.95271695e-01 1.01580128e-01 -1.16401082e-02
8.32927525e-01 1.24857247e+00 -1.12847042e+00 -6.58169568e-01
1.09044813e-01 4.22620267e-01 -5.40295124e-01 3.14442456e-01
-4.34150785e-01 2.23826364e-01 -2.82717079e-01 5.04069865e-01
4.29720670e-01 1.59160510e-01 7.37112761e-01 2.54615664e-01
-2.07931865e-02 7.18833089e-01 -5.92706978e-01 2.07756996e+00
-7.42740214e-01 4.42282170e-01 -3.05594444e-01 -1.13976490e+00
8.89593899e-01 4.88045603e-01 -1.59017280e-01 -8.85987759e-01
1.85544819e-01 1.72331482e-01 2.87792772e-01 -2.21740693e-01
4.68187243e-01 -1.78583980e-01 -5.45366466e-01 1.95620999e-01
9.94228184e-01 -1.15342841e-01 6.34797275e-01 2.96898633e-01
1.52653384e+00 6.81543946e-01 1.44004121e-01 -3.21474582e-01
4.33022141e-01 1.56186461e-01 5.58577359e-01 5.68393230e-01
-2.30447918e-01 3.93075615e-01 9.71278667e-01 -5.63309133e-01
-9.30478275e-01 -1.09857869e+00 5.33079281e-02 1.42778480e+00
-4.55664545e-01 -4.95936036e-01 -9.41150248e-01 -1.30311179e+00
-3.78446817e-01 1.04751742e+00 -7.53268838e-01 1.12004355e-01
-1.24526012e+00 -7.96730936e-01 8.53064477e-01 1.22636318e+00
1.80948094e-01 -1.84065688e+00 -4.12056774e-01 7.76172519e-01
1.26881480e-01 -1.63035226e+00 1.81768328e-01 1.05919182e+00
-1.03660703e+00 -1.09132171e+00 -1.52608037e-01 -1.38844860e+00
2.98394442e-01 -5.58032691e-01 1.76346052e+00 3.48464698e-02
3.99588905e-02 -3.04641090e-02 -7.71220267e-01 -3.75730395e-01
-8.10963869e-01 6.63733423e-01 -4.48457479e-01 -9.21467602e-01
4.02023315e-01 -5.49602985e-01 6.86229169e-02 -4.71013784e-01
-7.45676816e-01 -2.94615328e-01 4.65629876e-01 9.21729863e-01
1.54922187e-01 -2.08697379e-01 5.21845818e-01 -1.45683897e+00
8.20037544e-01 -3.73941660e-01 -6.81567013e-01 4.34806719e-02
-4.93306398e-01 6.24944866e-01 1.10366619e+00 1.04358867e-01
-1.44402909e+00 2.00159043e-01 -9.22824681e-01 3.12136263e-01
-4.21573132e-01 6.14156544e-01 3.81766483e-02 2.80428559e-01
4.91842479e-01 -3.33522886e-01 -6.09705150e-01 -5.73890686e-01
8.65090311e-01 1.01274401e-01 7.88575351e-01 -8.01761568e-01
5.69709480e-01 -7.47621432e-02 3.72966528e-01 -4.03913826e-01
-1.03901863e+00 -1.78591833e-01 -1.19097137e+00 5.54096997e-01
1.47075629e+00 -5.54915071e-01 -2.45385572e-01 3.11954886e-01
-1.59813404e+00 -6.28209293e-01 -1.93489745e-01 1.73713505e-01
-3.75679463e-01 4.04587954e-01 -1.16796315e+00 -4.06958133e-01
-4.77455139e-01 -7.94235647e-01 8.22817862e-01 2.43204564e-01
-1.51281983e-01 -1.52739561e+00 5.42015672e-01 -1.43675394e-02
2.26526663e-01 2.14544997e-01 1.41005230e+00 -1.05465639e+00
-4.83626164e-02 -2.20279917e-01 -3.37474585e-01 3.14160347e-01
-2.25599587e-01 7.49475583e-02 -7.88161218e-01 7.99821019e-02
-6.54907823e-01 -2.27489874e-01 1.04137301e+00 2.93447495e-01
5.15188992e-01 2.22025678e-01 -4.76128817e-01 6.27189755e-01
1.41466606e+00 4.80284318e-02 5.35224557e-01 5.10589480e-01
9.05888200e-01 6.33594215e-01 1.79059103e-01 -2.04792440e-01
4.78701234e-01 1.43937692e-01 7.38526940e-01 -5.12559479e-03
-5.24790287e-01 -5.11884511e-01 3.24080229e-01 9.72646594e-01
-1.37558237e-01 -5.51755786e-01 -1.14571965e+00 8.05517972e-01
-1.70206213e+00 -4.17591959e-01 -5.06796420e-01 1.55581188e+00
5.92411816e-01 6.16510868e-01 -1.30042732e-01 1.57669596e-02
5.91571987e-01 4.55846876e-01 -2.02340502e-02 -1.38142371e+00
1.57092497e-01 1.49430239e+00 6.80378854e-01 3.54159772e-01
-1.32813108e+00 1.78549492e+00 6.67110300e+00 1.06317662e-01
-5.18079162e-01 4.70388621e-01 2.12931126e-01 7.27779686e-01
-2.35480703e-02 2.60908216e-01 -1.14333653e+00 8.36024806e-02
1.92967284e+00 4.69882071e-01 1.87085327e-02 9.56750154e-01
-5.35138369e-01 1.03334062e-01 -1.41774631e+00 1.19398139e-01
-1.56175166e-01 -1.46568966e+00 -1.26693800e-01 -3.64243656e-01
3.29548508e-01 5.19414604e-01 -4.10596460e-01 7.61553943e-01
1.28019619e+00 -1.12960732e+00 4.49800044e-01 -2.07067296e-01
5.43897927e-01 -7.19307721e-01 1.10336709e+00 1.11973651e-01
-1.28943098e+00 -3.78563330e-02 -4.13212091e-01 -2.79177070e-01
5.61301053e-01 2.72762865e-01 -6.52791202e-01 8.41290295e-01
8.34790707e-01 5.37048161e-01 -7.22741902e-01 6.29607797e-01
-1.26160967e+00 1.01120532e+00 -1.73037902e-01 -2.62894243e-01
6.53293848e-01 1.61242515e-01 1.22358449e-01 1.71278501e+00
-3.33979465e-02 -2.20455974e-01 -4.55131531e-02 5.75404882e-01
-3.93214822e-01 2.32711192e-02 -6.27993405e-01 -3.05293828e-01
4.31827456e-01 1.29357958e+00 -9.64375138e-01 -2.35122323e-01
-7.20046580e-01 8.81470919e-01 1.13425338e+00 -1.28025413e-01
-5.91388285e-01 -5.37915409e-01 3.17126304e-01 -5.58151424e-01
7.14000046e-01 -7.29551971e-01 -5.77904731e-02 -8.46403897e-01
-2.34152645e-01 -3.67813349e-01 6.98637545e-01 -5.78633368e-01
-1.28884304e+00 9.75719035e-01 2.95103472e-02 -3.25363398e-01
-3.92926306e-01 -1.25029576e+00 -9.06512797e-01 8.74453783e-01
-1.85156167e+00 -1.33307111e+00 3.37415338e-01 3.30067933e-01
4.27558154e-01 9.57134292e-02 1.50402629e+00 1.01403296e-02
-6.64082646e-01 4.22156125e-01 -3.71940404e-01 7.28790283e-01
3.53979200e-01 -1.69755459e+00 1.60321474e+00 1.14375663e+00
5.45924366e-01 4.27811682e-01 4.09575462e-01 -4.90130365e-01
-1.47488546e+00 -9.54469264e-01 1.33411002e+00 -7.70442784e-01
1.07039571e+00 -5.91362417e-01 -8.75261307e-01 1.08686972e+00
6.11877978e-01 3.89537960e-01 3.42270732e-01 6.35677457e-01
-6.97339058e-01 3.53522390e-01 -9.00143862e-01 2.90797621e-01
1.39347279e+00 -3.14192206e-01 -1.26492131e+00 2.52310187e-01
1.25444639e+00 -4.87640589e-01 -9.62713122e-01 3.10745448e-01
2.05354556e-01 -9.35311794e-01 8.51901531e-01 -1.05347085e+00
7.47534931e-01 4.52933550e-01 6.54351115e-02 -1.77418840e+00
-6.17086887e-01 -3.93298835e-01 -4.00498621e-02 1.51730192e+00
9.28112745e-01 -8.74829054e-01 1.07759047e+00 2.55488306e-01
-6.25775933e-01 -4.69451457e-01 -1.14704955e+00 -9.11934674e-01
8.16974461e-01 -6.56810462e-01 4.96447206e-01 6.71345174e-01
1.12569936e-01 7.72972107e-01 1.76904351e-01 2.05726489e-01
4.93385673e-01 -1.93600655e-01 2.34146163e-01 -1.55092037e+00
-1.77769989e-01 -3.92870516e-01 -3.35862249e-01 -5.82912385e-01
9.83665287e-01 -1.23169100e+00 2.72105604e-01 -2.08416557e+00
-2.58669704e-01 -4.29566771e-01 -3.99483770e-01 8.29709589e-01
-2.57990748e-01 6.17589951e-02 2.37876996e-01 -4.10744518e-01
-4.11696166e-01 4.55732159e-02 9.55800116e-01 1.72586665e-01
-3.49858627e-02 -3.35047305e-01 -7.11120486e-01 6.96583688e-01
9.57175195e-01 -9.77084517e-01 3.67025845e-02 -7.87564278e-01
4.61991757e-01 2.69730270e-01 -1.60518259e-01 -8.74427617e-01
1.15344986e-01 4.09576923e-01 1.55768439e-01 -2.97105700e-01
8.35234374e-02 -5.05076289e-01 -5.96501410e-01 4.68566179e-01
-1.38165176e-01 4.86273110e-01 3.79966497e-01 3.84421438e-01
-1.60087943e-01 -6.35364711e-01 4.26212102e-01 -4.45730448e-01
-1.07354915e+00 3.25185776e-01 -4.71185088e-01 4.18394715e-01
5.16742647e-01 7.41824135e-02 -6.50510192e-01 3.61599147e-01
-8.37016106e-01 3.59303579e-02 -1.64370820e-01 5.62538624e-01
1.97011501e-01 -8.10323179e-01 -6.51120365e-01 1.80556014e-01
1.19441550e-03 8.74018148e-02 -2.46413931e-01 2.58668393e-01
-7.68177450e-01 6.55254483e-01 -3.68273079e-01 -1.83315471e-01
-1.00398076e+00 6.43045366e-01 -1.69834107e-01 -9.86489534e-01
-8.05207729e-01 1.21181476e+00 -4.03895348e-01 -1.00204182e+00
-9.37477052e-02 -7.72687674e-01 -4.81229067e-01 -7.03033656e-02
1.89092718e-02 -1.03672065e-01 3.44487131e-01 -4.04994965e-01
-5.87287188e-01 4.00025517e-01 -1.39490277e-01 1.96324252e-02
1.74130571e+00 6.47311866e-01 -1.16382375e-01 -1.12157643e-01
1.05437589e+00 -1.34687871e-01 -9.80475187e-01 3.98464908e-05
6.87225759e-01 1.52023032e-01 -3.34527567e-02 -8.57293606e-01
-9.27303731e-01 1.07018030e+00 1.75565332e-02 3.89920592e-01
5.83414376e-01 2.99508184e-01 9.97437298e-01 6.05625153e-01
3.32654864e-01 -8.55954826e-01 -1.71337739e-01 1.14128721e+00
1.90589577e-01 -1.13195515e+00 -1.99524596e-01 -5.17695487e-01
-4.09469604e-01 1.41576242e+00 4.81079668e-01 -7.49434292e-01
6.09135568e-01 6.57100201e-01 2.91499440e-02 -2.83920825e-01
-7.34339297e-01 -4.34579343e-01 -2.40297362e-01 1.09672081e+00
8.39559078e-01 3.07294548e-01 -4.20630038e-01 6.65964007e-01
-3.66740733e-01 -2.27280185e-01 5.35374463e-01 1.05740952e+00
-2.24563703e-01 -1.75779402e+00 5.01590297e-02 1.46401569e-01
-8.45239162e-01 -7.25731373e-01 -2.90871829e-01 1.19572151e+00
-1.67630017e-01 1.06453466e+00 4.35403138e-02 -8.75578597e-02
5.91087162e-01 4.95302916e-01 7.78445780e-01 -1.21939814e+00
-1.00703561e+00 -7.20072985e-01 9.29538906e-01 -5.39494872e-01
-4.45409983e-01 -4.69407469e-01 -1.55072308e+00 3.17748897e-02
-4.15909171e-01 1.61305457e-01 8.93987536e-01 1.06402194e+00
-2.20539011e-02 8.31548810e-01 -2.23311540e-02 -1.05657554e+00
-3.38757485e-01 -1.10169935e+00 -2.12145433e-01 1.99568316e-01
-2.04149842e-01 -3.90925229e-01 -2.06395552e-01 -3.10348451e-01] | [10.326864242553711, 9.586577415466309] |
dd3d8cfc-481f-4d83-9644-a19972608ec4 | pixel-wise-orthogonal-decomposition-for-color | 1407.0010 | null | http://arxiv.org/abs/1407.0010v2 | http://arxiv.org/pdf/1407.0010v2.pdf | Pixel-wise Orthogonal Decomposition for Color Illumination Invariant and Shadow-free Image | In this paper, we propose a novel, effective and fast method to obtain a
color illumination invariant and shadow-free image from a single outdoor image.
Different from state-of-the-art methods for shadow-free image that either need
shadow detection or statistical learning, we set up a linear equation set for
each pixel value vector based on physically-based shadow invariants, deduce a
pixel-wise orthogonal decomposition for its solutions, and then get an
illumination invariant vector for each pixel value vector on an image. The
illumination invariant vector is the unique particular solution of the linear
equation set, which is orthogonal to its free solutions. With this illumination
invariant vector and Lab color space, we propose an algorithm to generate a
shadow-free image which well preserves the texture and color information of the
original image. A series of experiments on a diverse set of outdoor images and
the comparisons with the state-of-the-art methods validate our method. | ['Zhi Han', 'Jiandong Tian', 'Yandong Tang', 'Liangqiong Qu'] | 2014-06-30 | null | null | null | null | ['shadow-detection'] | ['computer-vision'] | [ 5.53790987e-01 -6.23718917e-01 2.85628974e-01 -3.83707583e-01
-9.52993184e-02 -4.40662712e-01 1.62255064e-01 -7.82328665e-01
-1.76695675e-01 7.32751667e-01 -1.75657541e-01 -3.27698618e-01
-5.07981926e-02 -6.97004914e-01 -4.77708012e-01 -1.16281939e+00
2.76211470e-01 -1.58691645e-01 5.97417533e-01 -2.98444837e-01
2.92340398e-01 6.10209703e-01 -1.68927145e+00 -5.76347373e-02
9.86515284e-01 1.03227198e+00 3.24635148e-01 6.78049803e-01
-2.86332428e-01 3.51386040e-01 -6.10086620e-01 1.60783052e-01
4.92842972e-01 -7.18096375e-01 -2.28520215e-01 6.04766011e-01
6.48797274e-01 -2.78811604e-01 -4.26188290e-01 1.16070485e+00
2.16654658e-01 2.86208183e-01 7.73726583e-01 -1.28050852e+00
-6.83537424e-01 -5.54314852e-01 -6.22690380e-01 -2.64846444e-01
3.72268260e-01 -2.68741846e-01 4.40048546e-01 -8.51108730e-01
7.51295209e-01 1.16721654e+00 5.33806086e-01 4.93613034e-02
-1.07507849e+00 -5.32694042e-01 2.00635359e-01 2.80341089e-01
-1.36518669e+00 -7.48910084e-02 1.12951887e+00 3.61771546e-02
2.30881944e-02 6.87494874e-01 8.63004148e-01 7.70798683e-01
6.26198471e-01 6.09141171e-01 1.95691955e+00 -7.43558824e-01
3.07539880e-01 1.57524288e-01 1.43852562e-01 1.24536550e+00
2.95758396e-01 7.57408068e-02 -4.34398234e-01 -3.59643586e-02
6.80829942e-01 4.13314223e-01 -6.21224463e-01 -7.43074715e-01
-1.03921199e+00 4.14582193e-01 4.23128486e-01 6.61504045e-02
-8.57518092e-02 -3.12975585e-01 -4.59949166e-01 -4.61085849e-02
9.73877981e-02 -7.44074732e-02 -2.99070954e-01 4.80446905e-01
-8.31206977e-01 -9.19864252e-02 9.73092914e-01 6.13556504e-01
1.30510092e+00 6.15742169e-02 -9.11367387e-02 6.03043914e-01
3.45738053e-01 1.38109422e+00 1.42190143e-01 -9.01885986e-01
-1.59766465e-01 6.06958687e-01 2.81518489e-01 -1.32776797e+00
-2.33008906e-01 -5.74464090e-02 -8.25777590e-01 6.58006966e-01
2.57859498e-01 -4.30230796e-02 -1.08158374e+00 1.24280286e+00
4.67174381e-01 3.26353431e-01 3.02316964e-01 9.54390407e-01
5.92274010e-01 9.35535491e-01 -8.54454339e-01 -5.93482912e-01
1.05291522e+00 -9.07836080e-01 -8.74870300e-01 -9.38219279e-02
-2.55450875e-01 -1.03713500e+00 1.21373653e+00 6.05534196e-01
-1.99614093e-01 -4.82436389e-01 -1.31125510e+00 4.01707441e-01
-4.67683822e-01 2.08350852e-01 6.94726229e-01 8.20832610e-01
-6.75090671e-01 1.53768718e-01 -3.26340586e-01 -3.60196590e-01
-1.79542214e-01 2.53295135e-02 -4.84543860e-01 -3.86840910e-01
-6.33323252e-01 6.58777475e-01 6.29799515e-02 3.36841315e-01
-5.54189622e-01 -1.96162567e-01 -7.04151154e-01 -3.86418849e-01
6.16808891e-01 -2.09638193e-01 5.00771642e-01 -1.28111827e+00
-1.70077908e+00 4.73301262e-01 -6.25649571e-01 3.52775455e-01
1.96928427e-01 -4.36299518e-02 -6.60886705e-01 5.68663478e-02
-3.59774791e-02 -5.96761890e-02 1.16864741e+00 -2.16104579e+00
-5.01978695e-01 -1.30444020e-01 -1.61793619e-01 2.66460478e-01
-2.65718669e-01 -2.49990463e-01 -7.33114064e-01 -2.77336627e-01
6.31589949e-01 -1.23616469e+00 -1.59815833e-01 2.23132759e-01
-5.85211575e-01 4.44158435e-01 1.41888845e+00 -6.18002057e-01
1.03239024e+00 -2.23426151e+00 -7.95323700e-02 5.30170202e-01
-1.40996173e-01 1.41242728e-01 -4.78602275e-02 2.75452077e-01
1.37543693e-01 -3.62336427e-01 -6.85771465e-01 3.27952094e-02
-2.87548959e-01 5.39800942e-01 -2.55608022e-01 7.23276675e-01
-2.65659124e-01 3.79615873e-01 -7.42588937e-01 -8.19358587e-01
5.26221037e-01 5.21634459e-01 -5.61558455e-02 4.02446419e-01
2.81262901e-02 2.90693581e-01 -3.98498952e-01 8.59303117e-01
1.24810159e+00 3.07272315e-01 2.28199497e-01 -4.54446554e-01
-3.84638637e-01 -7.84851074e-01 -1.67236853e+00 1.19441485e+00
-4.02029246e-01 8.41366291e-01 2.35419683e-02 -6.36451244e-01
1.20744145e+00 -2.57908016e-01 3.58216792e-01 -6.31478071e-01
8.32222849e-02 6.32556677e-02 -3.32704335e-01 -5.95780015e-01
3.69070172e-01 -1.06229454e-01 7.49410316e-02 3.29448253e-01
-3.61849695e-01 -2.47422665e-01 -5.64025044e-02 -1.31052705e-02
9.30938900e-01 4.15185481e-01 1.64888985e-02 -3.96214843e-01
9.72561181e-01 -1.76562488e-01 7.90574908e-01 6.69488192e-01
-3.87176424e-02 8.75312865e-01 1.74126670e-01 -3.51319730e-01
-4.98644561e-01 -1.36320901e+00 -3.81024063e-01 6.96471334e-01
9.58062589e-01 -1.10768393e-01 -7.87396967e-01 -7.33954549e-01
5.84017113e-02 2.44207844e-01 -4.82630372e-01 1.27165005e-01
-4.71533984e-01 -7.94580400e-01 -1.89655960e-01 -4.16243076e-02
1.08137012e+00 -7.87580550e-01 -4.99513298e-01 -1.09857284e-01
-2.10207820e-01 -1.07657361e+00 -2.34262094e-01 1.84006989e-02
-4.37422842e-01 -1.18954861e+00 -6.69519484e-01 -6.78698957e-01
1.15001762e+00 9.98130620e-01 6.51311874e-01 1.16077766e-01
-5.95522344e-01 4.43186253e-01 -3.45552325e-01 -3.47452164e-01
5.33688180e-02 -5.75790226e-01 2.53493991e-02 8.23152900e-01
-1.86696067e-01 -3.27353328e-01 -6.37197137e-01 5.65952241e-01
-9.72388923e-01 1.48348331e-01 7.77938008e-01 7.93771446e-01
9.58097100e-01 3.75741601e-01 -3.85850459e-01 -8.14396203e-01
2.58130789e-01 8.16675127e-02 -9.80682850e-01 6.57321751e-01
-6.94226503e-01 2.68908679e-01 6.77590728e-01 -2.45717108e-01
-1.48457396e+00 6.22693598e-01 5.10531068e-01 -1.74714923e-01
-1.49261057e-01 -3.43072671e-03 -4.30923611e-01 -7.66141772e-01
5.04125893e-01 7.11825252e-01 7.08790869e-02 -3.43178570e-01
4.54187810e-01 5.98583460e-01 7.13941455e-01 -4.04011518e-01
1.42478848e+00 1.06269920e+00 4.88202035e-01 -1.45753217e+00
-9.09459472e-01 -5.65324545e-01 -8.49250972e-01 -6.59626722e-01
8.53937328e-01 -3.66904646e-01 -5.87325633e-01 7.51710176e-01
-9.00901318e-01 -1.73049644e-01 1.41499341e-01 2.76921898e-01
-3.31631124e-01 6.75212860e-01 1.94057763e-01 -9.82266605e-01
1.32511869e-01 -9.48578835e-01 9.43002403e-01 6.27999663e-01
5.18047988e-01 -8.87136281e-01 3.20419788e-01 1.78490087e-01
3.15511793e-01 4.94238466e-01 5.65660417e-01 4.97932285e-01
-1.07046962e+00 -2.55631238e-01 -4.01606917e-01 5.63291311e-01
3.18742573e-01 3.61189485e-01 -8.57004046e-01 -5.68627939e-02
8.80187228e-02 2.45768368e-01 9.06129241e-01 2.89604962e-01
9.74436522e-01 -2.00715899e-01 -4.07318264e-01 1.06543028e+00
2.01471329e+00 1.91368371e-01 7.13548243e-01 3.53267550e-01
7.99710155e-01 3.31910819e-01 8.59555185e-01 1.63581744e-01
2.40834728e-01 5.67509174e-01 3.16635549e-01 -7.15837061e-01
-2.96485335e-01 3.75687853e-02 4.22317207e-01 5.96858919e-01
-3.26686949e-01 -1.99269757e-01 -3.45471770e-01 8.09776410e-02
-1.93212795e+00 -8.17795694e-01 -3.86350602e-01 2.21539330e+00
3.63046765e-01 -1.33356363e-01 -3.71252179e-01 1.57117888e-01
5.24093449e-01 3.90577108e-01 -4.10666287e-01 -3.31503212e-01
-5.97212613e-01 1.67049423e-01 9.50836182e-01 6.72436714e-01
-9.80444252e-01 1.07020628e+00 6.87875366e+00 5.41101396e-01
-1.38206565e+00 -9.63638052e-02 6.13493621e-02 4.98884052e-01
-4.78846848e-01 4.73262250e-01 -6.44779503e-01 3.01893383e-01
2.30457366e-01 1.14421947e-02 6.31837666e-01 6.27214968e-01
7.93415904e-02 -7.84499288e-01 -3.74802530e-01 1.04273474e+00
6.31031692e-01 -1.02929068e+00 -4.07633930e-02 1.29993096e-01
1.18854618e+00 -5.07255733e-01 8.35718401e-03 -2.50020176e-01
3.02444045e-02 -7.30786741e-01 4.82816219e-01 1.10878694e+00
5.53504527e-01 -4.66062665e-01 6.54550374e-01 6.97063506e-02
-1.25182903e+00 6.14338741e-02 -5.24522722e-01 -7.12610111e-02
-9.73098651e-02 6.73879266e-01 -2.12258041e-01 5.70412636e-01
6.97861671e-01 5.96063495e-01 -8.34535301e-01 8.84642601e-01
-5.23460925e-01 4.17156667e-01 -2.74551868e-01 -1.32355958e-01
3.48255597e-02 -8.47620785e-01 4.10923928e-01 1.16946566e+00
2.65788883e-01 2.85598725e-01 3.55918735e-01 5.55801690e-01
4.00307715e-01 2.65915036e-01 -7.55745113e-01 4.34012800e-01
1.85686797e-01 1.55868530e+00 -9.63360071e-01 -1.74596325e-01
-4.04823393e-01 1.51406288e+00 -2.17121750e-01 8.12624514e-01
-7.52214313e-01 -7.61018932e-01 5.32163620e-01 -1.15438662e-01
2.56218612e-01 -5.75405061e-01 -2.28544191e-01 -1.17853522e+00
1.70641333e-01 -6.88567638e-01 -1.30407557e-01 -1.06375027e+00
-8.46746624e-01 4.02367055e-01 -9.93779451e-02 -1.34252596e+00
3.19813818e-01 -9.66083288e-01 -8.17890406e-01 8.25702071e-01
-1.75054896e+00 -1.17285740e+00 -9.17855620e-01 1.06890595e+00
1.30624250e-01 -1.90647870e-01 9.64153349e-01 -2.27033570e-01
-6.02714360e-01 3.25337201e-01 7.10368156e-01 -1.50892496e-01
8.47434998e-01 -1.11676085e+00 -2.96317935e-01 1.28604972e+00
4.36856374e-02 4.93715256e-01 7.30287254e-01 -5.35128713e-01
-1.95686960e+00 -6.38840675e-01 3.46445292e-01 -3.40959132e-01
3.20297331e-01 -3.57852459e-01 -5.35173059e-01 1.98959619e-01
7.66427889e-02 1.32427365e-01 4.32656020e-01 -2.11023107e-01
-2.80923963e-01 -7.10159063e-01 -1.05004287e+00 6.78257704e-01
9.45490599e-01 -3.31323206e-01 -3.66168320e-01 4.32819963e-01
4.43310559e-01 -3.53108793e-01 -3.37378800e-01 3.84719819e-01
9.60711598e-01 -1.39947104e+00 1.11287427e+00 5.88594042e-02
-3.94056849e-02 -9.15323317e-01 -4.92330998e-01 -9.90185618e-01
-3.51802170e-01 -5.92783809e-01 3.05278599e-01 1.10738766e+00
8.04684386e-02 -8.61929119e-01 6.55720592e-01 1.23099975e-01
-2.53353044e-02 -6.34788573e-01 -6.49811268e-01 -1.01993358e+00
-5.82195044e-01 -9.23021287e-02 4.72700000e-01 4.96766597e-01
-7.57663310e-01 -4.66596074e-02 -5.51564693e-01 6.20449662e-01
1.21229684e+00 8.23681533e-01 1.15301824e+00 -1.22277951e+00
-1.12124860e-01 1.32816248e-02 -3.90749186e-01 -5.95384777e-01
3.31418067e-01 -3.39404136e-01 5.12207747e-01 -1.70697474e+00
3.91406447e-01 -5.13852477e-01 -3.74192983e-01 3.22873950e-01
-1.71391383e-01 6.37116611e-01 2.29635730e-01 2.47394845e-01
-5.29101193e-01 5.22723317e-01 1.33259082e+00 -5.14092036e-02
-2.79967099e-01 -1.24576099e-01 -5.15202403e-01 8.54725301e-01
5.54547727e-01 -2.60138094e-01 -2.34671071e-01 7.89755732e-02
-3.30563575e-01 -2.63874203e-01 3.31014603e-01 -1.22642457e+00
-1.38652340e-01 -9.26738441e-01 6.83741629e-01 -5.37662983e-01
5.90475619e-01 -1.02051818e+00 1.01050884e-01 5.06113470e-01
4.45339799e-01 -3.91103625e-01 -1.13229670e-01 6.29059315e-01
-5.43461833e-03 -6.74127564e-02 9.24693465e-01 -5.14073893e-02
-1.12053418e+00 9.71580297e-02 -2.90544420e-01 -3.06840986e-01
1.07365716e+00 -6.17530763e-01 -4.37789708e-01 -3.79036188e-01
-1.52197406e-01 -2.85904169e-01 9.16359246e-01 2.44254097e-01
8.89986157e-01 -1.27943885e+00 -3.94263059e-01 7.13123143e-01
8.45966935e-02 -5.51501632e-01 2.28693381e-01 4.58248287e-01
-9.03206050e-01 5.90164214e-02 -4.99253362e-01 -6.32743657e-01
-1.51253521e+00 3.00754577e-01 3.65978032e-01 3.10016602e-01
-6.57138348e-01 4.69325125e-01 3.86023909e-01 -3.63852799e-01
-1.45000219e-02 -4.24110740e-02 7.55518153e-02 -3.92678738e-01
2.56306022e-01 3.87286395e-01 -2.92367578e-01 -9.16262746e-01
-5.07548690e-01 1.31963575e+00 6.39712155e-01 -4.43186581e-01
1.02885926e+00 -1.05352819e-01 -4.97489005e-01 5.65637469e-01
1.22597671e+00 5.65313697e-01 -1.25374389e+00 -9.69928950e-02
-6.45954669e-01 -1.09848285e+00 1.94655314e-01 -7.26970255e-01
-1.04842496e+00 4.59954619e-01 1.10704851e+00 3.73088382e-02
1.42736220e+00 -4.08639461e-01 7.07653224e-01 6.21830940e-01
3.19130301e-01 -1.09551179e+00 1.50974467e-01 4.86842662e-01
8.41432452e-01 -1.32379138e+00 5.06507993e-01 -6.41315758e-01
-3.57299238e-01 1.20131361e+00 4.15757209e-01 -2.35065341e-01
8.53967369e-01 1.97683886e-01 4.87508029e-01 1.77758589e-01
7.74074271e-02 -4.52420741e-01 4.47288036e-01 9.11762476e-01
-3.94225940e-02 2.36160219e-01 -3.90126616e-01 2.17703246e-02
-6.75536841e-02 -2.92181462e-01 6.32100046e-01 9.22962904e-01
-8.11303318e-01 -1.15613198e+00 -9.91137743e-01 -3.97387482e-02
2.62517929e-01 1.23744152e-01 -5.03716290e-01 7.30631411e-01
3.28766167e-01 1.06351554e+00 -3.04312348e-01 -4.40085113e-01
2.41773710e-01 2.38561127e-02 6.12805486e-01 -2.11742029e-01
2.65443951e-01 3.48728783e-02 -3.89785528e-01 -7.67505050e-01
-4.10591781e-01 -6.24403417e-01 -1.29143894e+00 -1.34100765e-01
-3.23612571e-01 -1.08430870e-01 1.05077708e+00 8.77542853e-01
-7.00353459e-02 2.83934802e-01 1.14036679e+00 -9.54814017e-01
1.56242624e-01 -2.70188898e-01 -9.42817688e-01 3.06105614e-01
4.55967307e-01 -6.92490101e-01 -6.48298144e-01 1.46558031e-01] | [10.501219749450684, -2.77299427986145] |
e25c608d-3968-472d-88ef-5908201d8715 | partial-domain-adaptation-using-selective | 2101.02275 | null | https://arxiv.org/abs/2101.02275v1 | https://arxiv.org/pdf/2101.02275v1.pdf | Partial Domain Adaptation Using Selective Representation Learning For Class-Weight Computation | The generalization power of deep-learning models is dependent on rich-labelled data. This supervision using large-scaled annotated information is restrictive in most real-world scenarios where data collection and their annotation involve huge cost. Various domain adaptation techniques exist in literature that bridge this distribution discrepancy. However, a majority of these models require the label sets of both the domains to be identical. To tackle a more practical and challenging scenario, we formulate the problem statement from a partial domain adaptation perspective, where the source label set is a super set of the target label set. Driven by the motivation that image styles are private to each domain, in this work, we develop a method that identifies outlier classes exclusively from image content information and train a label classifier exclusively on class-content from source images. Additionally, elimination of negative transfer of samples from classes private to the source domain is achieved by transforming the soft class-level weights into two clusters, 0 (outlier source classes) and 1 (shared classes) by maximizing the between-cluster variance between them. | ['Hemanth Venkateswara', 'Baoxin Li', 'Arunabha Sen', 'Riti Paul', 'Sandipan Choudhuri'] | 2021-01-06 | null | null | null | null | ['partial-domain-adaptation'] | ['methodology'] | [ 5.02356112e-01 2.65256405e-01 -3.57361287e-01 -8.17905366e-01
-5.36037385e-01 -5.70238590e-01 3.39851201e-01 3.18143666e-01
-3.96731317e-01 7.34033287e-01 -2.24609852e-01 2.36783892e-01
9.37871709e-02 -5.94420671e-01 -7.22022474e-01 -8.74471903e-01
3.35191071e-01 6.35225296e-01 2.55555451e-01 1.27990335e-01
9.94422436e-02 2.34838381e-01 -1.67879081e+00 4.14489388e-01
1.07076645e+00 1.20640445e+00 -2.03644950e-02 7.51290694e-02
-3.50318789e-01 5.52747488e-01 -7.32065558e-01 -5.06752074e-01
5.98526418e-01 -5.11534035e-01 -5.69662035e-01 4.95347619e-01
5.81813514e-01 -1.96582317e-01 2.42733225e-01 1.54750156e+00
4.06195134e-01 -8.26677680e-02 9.56083953e-01 -1.76523304e+00
-7.03741848e-01 3.69990736e-01 -7.86313713e-01 -1.53828785e-01
-2.06948176e-01 -1.28416389e-01 8.19380224e-01 -8.39753330e-01
6.42931640e-01 8.46772850e-01 5.74776709e-01 5.80795646e-01
-1.42994368e+00 -8.81840646e-01 5.16794264e-01 9.48637798e-02
-1.43860936e+00 -3.03854316e-01 1.02246082e+00 -7.20429838e-01
2.97895759e-01 -1.16149694e-01 1.84704140e-01 1.48218155e+00
-2.66105086e-01 6.19624615e-01 1.15034080e+00 -4.94262099e-01
5.92024624e-01 9.09502625e-01 1.60289094e-01 4.43513468e-02
3.40769142e-01 -2.15270832e-01 -5.79976976e-01 -1.73184752e-01
3.45249921e-01 1.15537688e-01 -1.72081947e-01 -1.04353929e+00
-7.98370421e-01 9.97560203e-01 7.34879151e-02 2.05159739e-01
-1.16276085e-01 -5.33740520e-01 6.26861453e-01 4.35768306e-01
7.05699265e-01 9.19707194e-02 -7.38776743e-01 3.70596021e-01
-8.49073827e-01 3.15122083e-02 6.35983646e-01 1.31745183e+00
1.00169158e+00 -1.88982621e-01 4.08437885e-02 1.01796901e+00
2.48036340e-01 4.10203785e-01 7.58113861e-01 -5.68253279e-01
4.22853619e-01 9.10376370e-01 1.46679729e-01 -1.00887954e+00
-1.94899127e-01 -5.33612430e-01 -8.58496308e-01 2.76414037e-01
6.73755765e-01 -1.70943774e-02 -8.85402799e-01 2.02122664e+00
6.27629161e-01 8.36568996e-02 1.10457562e-01 8.47048759e-01
4.62355524e-01 2.26182431e-01 2.39216164e-01 -1.48415208e-01
9.92579281e-01 -7.44215608e-01 -5.06816208e-01 -4.43018377e-01
7.97283590e-01 -5.83215714e-01 1.28976488e+00 4.35944796e-01
-4.59634155e-01 -5.72651088e-01 -1.03146195e+00 1.05976194e-01
-5.75231373e-01 3.51964176e-01 1.91560403e-01 6.95845902e-01
-5.40538192e-01 4.12218839e-01 -3.41809273e-01 -4.95854706e-01
6.14608169e-01 4.08721775e-01 -5.17967463e-01 -1.16391480e-01
-1.04204035e+00 5.91865599e-01 7.39543080e-01 -2.86839515e-01
-8.35252941e-01 -5.93331873e-01 -6.66508079e-01 -1.34115905e-01
4.69411731e-01 -2.35222548e-01 1.05174339e+00 -1.88960898e+00
-1.19858432e+00 1.55008268e+00 2.01285690e-01 -2.71423280e-01
6.00028753e-01 -1.18781269e-01 -4.69660997e-01 -1.43122628e-01
4.39577729e-01 5.38292825e-01 1.29199135e+00 -1.67659974e+00
-9.47486520e-01 -7.30563939e-01 -3.02470416e-01 1.68920636e-01
-8.06654692e-01 -1.26843423e-01 -1.78879946e-01 -6.61597490e-01
1.42700803e-02 -9.23776090e-01 1.01737291e-01 1.84401318e-01
-4.46090847e-01 -2.80478746e-01 9.10295069e-01 -4.28482711e-01
1.05775917e+00 -2.50798655e+00 5.80440387e-02 2.98040301e-01
1.33819327e-01 2.48470277e-01 -1.13248661e-01 -9.56817903e-03
-4.19659466e-01 -1.44548118e-01 -4.62386489e-01 -3.11127156e-01
6.11687824e-02 2.57787943e-01 -4.84992325e-01 5.79891801e-01
4.02855575e-01 9.28321257e-02 -7.94844389e-01 -5.11089683e-01
-3.86102200e-02 1.19638242e-01 -4.95466232e-01 4.99894381e-01
-2.24156067e-01 6.89134896e-01 -3.71206522e-01 5.11951625e-01
9.85691547e-01 -1.59570009e-01 2.64217943e-01 -1.89328939e-01
2.65755922e-01 -2.19804823e-01 -1.45317876e+00 1.62294841e+00
-1.04066163e-01 1.60713285e-01 2.16257781e-01 -1.38142061e+00
1.11330104e+00 7.43250400e-02 7.16649592e-01 -4.86491710e-01
1.50286555e-01 3.92942607e-01 -4.03178930e-02 -4.41879570e-01
2.34228194e-01 -4.07499522e-01 -1.50670514e-01 3.81946295e-01
4.37325299e-01 1.23700850e-01 1.19470350e-01 -7.18113557e-02
7.15840280e-01 1.23231031e-01 3.14321607e-01 -2.32276037e-01
4.51115251e-01 -5.54681383e-03 9.62422848e-01 6.60299242e-01
-5.81655324e-01 7.37574756e-01 5.49753189e-01 -3.21319789e-01
-1.33270991e+00 -1.14524949e+00 -3.53199422e-01 1.33393335e+00
1.72963455e-01 8.79288241e-02 -8.39999795e-01 -1.21802938e+00
1.63202107e-01 6.01897061e-01 -6.82419240e-01 -5.64734221e-01
-2.38143072e-01 -7.39828587e-01 3.57858598e-01 4.05477881e-01
3.38072270e-01 -8.91102970e-01 -3.42761576e-01 9.05790646e-03
-1.22987740e-01 -1.10537219e+00 -2.81561613e-01 5.40518463e-01
-7.10489929e-01 -1.09382153e+00 -8.09829116e-01 -9.81787324e-01
9.08953726e-01 -6.66013137e-02 1.24545729e+00 -2.19866052e-01
-1.63993433e-01 1.22567318e-01 -5.66140115e-01 -5.46039164e-01
-3.95112127e-01 1.75087512e-01 2.53167212e-01 5.38828969e-01
8.97144794e-01 -4.78115797e-01 -2.44610250e-01 4.93198603e-01
-1.02586663e+00 -2.56291509e-01 4.63164628e-01 7.33280957e-01
6.78679228e-01 3.49977762e-01 7.95092821e-01 -1.18967164e+00
1.53320834e-01 -8.85445416e-01 -4.14086372e-01 1.69497177e-01
-5.16981363e-01 -2.48389900e-01 7.30677187e-01 -7.45151818e-01
-1.17096078e+00 3.65006238e-01 2.36410886e-01 -4.89796817e-01
-6.72162116e-01 5.24532087e-02 -7.51492679e-01 2.61026710e-01
8.58181894e-01 -1.43063609e-02 -6.71481341e-02 -5.50919473e-01
2.53688365e-01 9.69400525e-01 5.29384434e-01 -6.11002028e-01
8.43223393e-01 6.15013480e-01 -3.67453307e-01 -5.51029146e-01
-1.26197970e+00 -6.21893227e-01 -1.11410570e+00 -4.59586009e-02
6.89958036e-01 -9.91811872e-01 5.02920188e-02 5.32426000e-01
-8.76337469e-01 -2.59655118e-01 -6.88463211e-01 2.94309199e-01
-5.08409560e-01 3.36845726e-01 -1.14425935e-01 -6.53194904e-01
2.27023169e-01 -1.03894413e+00 1.09684694e+00 1.59609586e-01
-1.89562991e-01 -8.93184662e-01 3.20448726e-02 7.58665055e-02
1.61369443e-02 3.70162070e-01 9.07189429e-01 -1.31199551e+00
-9.32470988e-03 -3.45417053e-01 -3.66205484e-01 8.24889123e-01
2.63191640e-01 -2.92115211e-01 -1.23534989e+00 -3.71736795e-01
3.67033958e-01 -5.99419713e-01 5.92034757e-01 1.73461989e-01
1.18155694e+00 -6.35362566e-02 -3.88106614e-01 5.06852090e-01
1.38288796e+00 7.14935139e-02 1.56105652e-01 4.38589007e-01
7.19282925e-01 1.06374955e+00 7.21172810e-01 5.68241894e-01
9.99697894e-02 5.47600150e-01 4.73396122e-01 -1.31390512e-01
6.04241863e-02 -2.07573339e-01 3.05829376e-01 4.24003333e-01
5.88589549e-01 -3.16000015e-01 -8.48357856e-01 7.30201840e-01
-1.84102786e+00 -6.18413091e-01 -2.49088854e-02 2.48364997e+00
1.06520319e+00 2.19676763e-01 1.96529776e-01 1.13624915e-01
1.11694324e+00 -3.08356255e-01 -9.07898188e-01 -1.59255583e-02
-3.01327497e-01 -2.86833286e-01 4.60315973e-01 -2.96820886e-02
-1.47092485e+00 7.08715260e-01 5.70199966e+00 8.42624784e-01
-9.29812729e-01 1.61592066e-01 6.62898660e-01 -8.35603699e-02
-1.38948876e-02 -1.40617430e-01 -9.44446802e-01 7.91783631e-01
8.46014559e-01 1.05807833e-01 -4.20850106e-02 1.20064628e+00
-1.80887178e-01 -5.66921420e-02 -1.46661234e+00 9.29912627e-01
2.38942504e-01 -4.66743201e-01 9.84288380e-03 1.28839195e-01
8.36791694e-01 -1.86538145e-01 2.03816175e-01 3.74245971e-01
4.52625245e-01 -7.41427481e-01 7.56807029e-01 1.90723956e-01
8.95081401e-01 -7.94491947e-01 9.06928957e-01 5.18279016e-01
-6.97431803e-01 -2.93082744e-01 -5.13277590e-01 6.39631227e-02
-1.12508848e-01 9.78724957e-01 -5.18008292e-01 3.55860382e-01
1.01324558e+00 8.04281116e-01 -7.12830782e-01 7.64654338e-01
1.67406593e-02 4.31764245e-01 -2.92280376e-01 6.69273198e-01
-7.84775615e-02 -3.36611390e-01 2.60817528e-01 1.05351031e+00
2.15985686e-01 -4.18517619e-01 4.91845459e-01 8.02444816e-01
-1.81248009e-01 1.83000267e-01 -8.33655834e-01 1.09140918e-01
4.70469236e-01 1.11411703e+00 -7.44876087e-01 -4.40907717e-01
-6.22670650e-01 1.22929561e+00 4.93491709e-01 2.75428891e-01
-8.44565988e-01 -3.77363563e-01 6.98921382e-01 1.26187399e-01
1.55494988e-01 1.96664363e-01 -5.70304036e-01 -1.32355571e+00
1.46598607e-01 -8.59525502e-01 6.63685679e-01 -3.99064869e-01
-1.95795107e+00 4.20453876e-01 3.27500398e-03 -1.61159718e+00
-3.68948057e-02 -7.85122633e-01 -1.45535812e-01 6.73715353e-01
-1.55718195e+00 -1.16318691e+00 -4.05285537e-01 8.81602585e-01
3.38974446e-01 -4.23421115e-01 7.75107205e-01 6.58888400e-01
-6.02871299e-01 8.82440925e-01 5.02055764e-01 2.38742858e-01
1.36886847e+00 -1.18455410e+00 -3.05900365e-01 7.09519625e-01
-1.03561752e-01 3.90615046e-01 7.10984647e-01 -6.07726991e-01
-5.02780855e-01 -1.37625146e+00 6.52495444e-01 -5.78914225e-01
6.19037986e-01 -5.14645696e-01 -1.16602170e+00 8.01986754e-01
-5.80399260e-02 2.57793218e-01 1.01510632e+00 5.75409783e-03
-6.87680304e-01 -1.49262100e-01 -1.45461345e+00 2.38202482e-01
8.62434149e-01 -4.08029735e-01 -4.43491399e-01 4.79165792e-01
4.55444515e-01 -2.16616601e-01 -8.42551827e-01 3.31658930e-01
1.36947155e-01 -1.04689968e+00 7.19691098e-01 -7.07814932e-01
4.29603994e-01 -4.81029421e-01 -3.47925812e-01 -1.28524029e+00
-2.47784350e-02 9.53937769e-02 7.81240240e-02 1.63392127e+00
1.63772330e-01 -4.18646842e-01 9.18908536e-01 6.34152889e-01
-8.86965096e-02 -1.70469850e-01 -9.08722520e-01 -9.94714737e-01
3.14235508e-01 -1.43423140e-01 5.07388294e-01 1.33063102e+00
-2.97329426e-01 3.92340779e-01 -3.71092498e-01 2.43230104e-01
7.23303735e-01 4.05921750e-02 6.96601391e-01 -1.67690790e+00
-5.26530519e-02 -1.76922947e-01 -3.69391501e-01 -5.78245819e-01
4.27352756e-01 -9.40227151e-01 2.14380145e-01 -8.84387314e-01
4.35361803e-01 -6.33552790e-01 -3.42423856e-01 4.86741871e-01
3.27687301e-02 3.68782818e-01 -8.89893323e-02 5.63070893e-01
-6.43548071e-01 6.67718172e-01 7.06320286e-01 -5.34113795e-02
-4.68888842e-02 -1.32857659e-03 -6.96793318e-01 1.07440257e+00
7.66848624e-01 -7.93435454e-01 -5.89094818e-01 -4.38972354e-01
1.18373021e-01 -6.84534609e-01 2.13479117e-01 -1.01826620e+00
-8.72567222e-02 -2.75294542e-01 4.72951829e-01 -2.63842702e-01
3.65303531e-02 -1.42699242e+00 -5.30174077e-02 -4.41354932e-03
-5.23888826e-01 -3.59417677e-01 -5.31207919e-02 7.79625177e-01
-3.31764519e-01 -4.90206152e-01 1.22733939e+00 -1.53202787e-01
-7.53126681e-01 2.76244015e-01 -7.67756850e-02 3.44040841e-01
1.44250500e+00 -3.83411795e-01 2.36106967e-03 -2.68540680e-02
-8.67772520e-01 3.31890061e-02 8.39502454e-01 4.91329908e-01
1.28901273e-01 -1.47972405e+00 -5.41951299e-01 3.89179945e-01
6.89852417e-01 2.87954658e-01 2.25281134e-01 5.39181113e-01
5.80637380e-02 -2.05701310e-02 -3.96893710e-01 -8.05052102e-01
-1.05616403e+00 8.52062106e-01 2.67862886e-01 -1.30756915e-01
-4.10298973e-01 8.67137253e-01 6.74169362e-01 -8.88332188e-01
4.29867208e-01 1.52844980e-01 -1.58655852e-01 3.80376518e-01
4.35089469e-01 1.28792912e-01 1.20989971e-01 -8.36385489e-01
-2.92243958e-01 3.79382223e-01 -2.67424107e-01 1.57550901e-01
1.19400430e+00 -3.61949265e-01 -3.88662927e-02 6.04429305e-01
1.31281304e+00 -2.64694661e-01 -1.49706614e+00 -5.48038006e-01
3.66507113e-01 -4.83474135e-01 -3.47175330e-01 -7.53987432e-01
-1.03836977e+00 9.46870446e-01 9.52204585e-01 1.56686947e-01
1.23975730e+00 8.35605785e-02 4.81209219e-01 -2.13886648e-02
1.62232980e-01 -1.67240918e+00 2.39587888e-01 3.58942479e-01
5.48338532e-01 -1.69321001e+00 -2.85858423e-01 -4.08601999e-01
-8.90379190e-01 8.19704950e-01 1.06224620e+00 -6.08620867e-02
6.78743064e-01 6.63065314e-02 1.86423928e-01 -1.38116870e-02
-2.56131977e-01 -7.67581016e-02 1.62315786e-01 1.08611357e+00
4.15978462e-01 -6.70475094e-03 2.35838797e-02 8.69472861e-01
1.58643976e-01 -4.49963175e-02 3.41651589e-01 8.82958353e-01
-4.70395476e-01 -1.14472067e+00 -4.22321260e-01 4.46022153e-01
-4.17198807e-01 1.22615740e-01 -5.08681536e-01 6.27666235e-01
6.40509725e-01 8.39295924e-01 2.85231590e-01 -1.96832672e-01
3.71061862e-01 3.36583555e-01 7.22154975e-02 -9.47394907e-01
-2.86224723e-01 -7.82937333e-02 -2.91639507e-01 -2.85495907e-01
-5.45352817e-01 -6.50640249e-01 -1.02952588e+00 1.62363034e-02
-3.34211707e-01 -9.29808896e-03 5.15512109e-01 8.74468982e-01
3.59738827e-01 2.13379905e-01 6.77991331e-01 -6.11377180e-01
-8.41596842e-01 -8.60916078e-01 -1.04293883e+00 1.08718264e+00
2.39690140e-01 -9.01439369e-01 -5.79427958e-01 4.78973746e-01] | [10.324519157409668, 3.117658853530884] |
3931924f-3d0f-4d16-a6da-baf81cfd749d | neural-joint-entropy-estimation | 2012.11197 | null | https://arxiv.org/abs/2012.11197v1 | https://arxiv.org/pdf/2012.11197v1.pdf | Neural Joint Entropy Estimation | Estimating the entropy of a discrete random variable is a fundamental problem in information theory and related fields. This problem has many applications in various domains, including machine learning, statistics and data compression. Over the years, a variety of estimation schemes have been suggested. However, despite significant progress, most methods still struggle when the sample is small, compared to the variable's alphabet size. In this work, we introduce a practical solution to this problem, which extends the work of McAllester and Statos (2020). The proposed scheme uses the generalization abilities of cross-entropy estimation in deep neural networks (DNNs) to introduce improved entropy estimation accuracy. Furthermore, we introduce a family of estimators for related information-theoretic measures, such as conditional entropy and mutual information. We show that these estimators are strongly consistent and demonstrate their performance in a variety of use-cases. First, we consider large alphabet entropy estimation. Then, we extend the scope to mutual information estimation. Next, we apply the proposed scheme to conditional mutual information estimation, as we focus on independence testing tasks. Finally, we study a transfer entropy estimation problem. The proposed estimators demonstrate improved performance compared to existing methods in all tested setups. | ['Irad Ben-Gal', 'Amichai Painsky', 'Yuval Shalev'] | 2020-12-21 | null | null | null | null | ['mutual-information-estimation'] | ['methodology'] | [ 4.40538377e-01 4.70616333e-02 -1.89175487e-01 -5.91971099e-01
-4.90551621e-01 -1.76838428e-01 3.58237118e-01 3.94145131e-01
-6.86678886e-01 1.33378291e+00 -1.99261695e-01 -1.94670752e-01
-5.39422810e-01 -8.70762765e-01 -4.89775062e-01 -5.82157850e-01
-4.84697759e-01 2.69576520e-01 5.49083240e-02 5.45504689e-02
3.39525878e-01 3.93170208e-01 -1.70632243e+00 -4.28109139e-01
9.72229838e-01 1.36696482e+00 1.76316127e-01 6.45146489e-01
1.78063184e-01 8.05541158e-01 -6.88689053e-01 -5.81516325e-01
1.96815550e-01 -5.06136417e-01 -9.10265088e-01 -4.42482680e-01
1.61451939e-02 -4.83502924e-01 -4.70058858e-01 1.32155406e+00
5.23521960e-01 2.84981191e-01 1.00971103e+00 -1.45176542e+00
-2.62754858e-01 8.64663243e-01 -4.44964439e-01 2.62834340e-01
-1.08584568e-01 -7.01434255e-01 1.18633235e+00 -3.05834860e-01
3.14425319e-01 9.49233294e-01 6.95927978e-01 2.10968941e-01
-7.37118185e-01 -9.53156233e-01 -4.93630826e-01 6.11409009e-01
-1.48951709e+00 -3.03129852e-01 5.82776606e-01 -2.19549984e-01
8.94453764e-01 1.71648040e-01 3.13865811e-01 7.27846563e-01
3.51234466e-01 9.34431911e-01 9.24208164e-01 -5.90333283e-01
3.44149500e-01 3.67304683e-01 2.00448632e-01 4.95735615e-01
4.26750422e-01 -2.81919595e-02 -1.65607974e-01 -2.05228943e-02
6.48695827e-01 1.72738079e-02 -2.07114756e-01 -1.05027936e-01
-8.81404042e-01 8.96391213e-01 5.54293580e-02 6.12436056e-01
-1.26567006e-01 -8.31502452e-02 4.80458647e-01 4.99674529e-01
5.11965513e-01 1.50721833e-01 -4.88830119e-01 -3.67287904e-01
-1.06154597e+00 5.68417571e-02 1.29004741e+00 1.16400719e+00
5.86570323e-01 -2.01549917e-01 -8.11395571e-02 1.02630961e+00
2.40466177e-01 3.58735979e-01 5.84075451e-01 -8.01701486e-01
6.87454104e-01 6.21009851e-03 -3.21335286e-01 -1.13623333e+00
-3.37736785e-01 -5.78170776e-01 -1.53918672e+00 -2.15825766e-01
2.47137800e-01 -2.12624446e-01 -3.43326390e-01 1.92950666e+00
-3.18486616e-02 -1.47706568e-01 1.55310243e-01 4.00565058e-01
5.10148168e-01 3.96378279e-01 -2.60904044e-01 -6.06691599e-01
9.24077988e-01 -5.64694047e-01 -9.77289438e-01 8.86554345e-02
5.59683800e-01 -6.04585469e-01 1.89882070e-01 4.88225996e-01
-1.10458469e+00 -3.78928214e-01 -1.24507797e+00 1.20190293e-01
-3.78625870e-01 2.40036640e-02 4.89570141e-01 8.71863842e-01
-1.07738578e+00 1.04797554e+00 -6.50329471e-01 -2.88854212e-01
4.97201741e-01 4.78871822e-01 -3.40099186e-01 2.33535126e-01
-1.68364096e+00 8.16625834e-01 1.11459768e+00 -1.80093020e-01
-1.50069863e-01 -7.80414566e-02 -8.35410178e-01 3.81794840e-01
1.02370018e-02 -4.12791073e-01 1.14679563e+00 -6.96304500e-01
-1.49100578e+00 3.85445952e-01 3.69373523e-02 -8.44030380e-01
5.08704364e-01 -2.06128463e-01 -4.09843802e-01 -3.40045057e-02
-2.08800331e-01 3.45414072e-01 4.25743759e-01 -5.40171027e-01
-6.55232906e-01 -2.94130176e-01 -1.08670756e-01 7.20470771e-02
-8.11423838e-01 -9.95344892e-02 -1.42773509e-01 -6.29303277e-01
7.31906071e-02 -6.94995999e-01 -5.12352176e-02 -1.69404209e-01
-5.45374393e-01 -1.32849604e-01 5.99703372e-01 -6.52991235e-01
1.50465107e+00 -2.07171035e+00 5.77339530e-02 4.33446199e-01
2.61263818e-01 2.27596059e-01 2.55869806e-01 4.82401848e-01
-1.85090154e-01 1.94424510e-01 -7.05492258e-01 -3.69567782e-01
1.75797746e-01 1.88013941e-01 6.45364635e-03 5.42974174e-01
-5.67758046e-02 5.76028228e-01 -5.65982521e-01 -7.81746268e-01
1.57618031e-01 4.93901193e-01 -4.55058783e-01 2.39523396e-01
2.43927076e-01 1.92476898e-01 -3.00769389e-01 2.72356898e-01
9.13068473e-01 -3.58430237e-01 2.00561523e-01 -8.54021236e-02
1.16927572e-01 1.66797549e-01 -1.38219035e+00 1.24474442e+00
-6.71303213e-01 1.16670632e+00 -3.75259429e-01 -1.49721503e+00
9.92562652e-01 2.49823168e-01 5.19978166e-01 -3.39236170e-01
5.20334721e-01 2.78771788e-01 3.05352777e-01 -4.25040603e-01
6.66807950e-01 -1.67106375e-01 6.00500703e-02 5.74353874e-01
2.44453236e-01 4.64904234e-02 5.40089250e-01 6.58253655e-02
1.11124897e+00 -5.25906324e-01 1.12113619e+00 -1.29432961e-01
5.73588908e-01 -1.00645292e+00 2.60678053e-01 9.03142393e-01
-3.19159597e-01 3.89373541e-01 4.74304706e-01 3.23527642e-02
-1.10104704e+00 -8.73448789e-01 -7.89066434e-01 6.90833688e-01
1.21667888e-02 -3.65678281e-01 -8.43263388e-01 -4.77990061e-01
-1.42130703e-01 7.87682414e-01 -4.63026136e-01 -1.39242768e-01
2.06144601e-02 -8.90469015e-01 8.12879860e-01 3.99353147e-01
1.25685263e+00 -1.00651085e+00 -6.22843981e-01 7.01057538e-02
-3.94157797e-01 -1.09810662e+00 -6.63896929e-03 4.04171288e-01
-9.24078107e-01 -7.49475718e-01 -8.13170969e-01 -5.87458313e-01
2.57886916e-01 -2.98735619e-01 1.05806768e+00 -1.52483702e-01
-1.43222392e-01 1.82330713e-01 -3.62611562e-01 -3.59061390e-01
-6.16903901e-01 3.32128942e-01 1.93521678e-01 -1.99618578e-01
3.61540437e-01 -8.83405507e-01 -3.77867967e-01 3.22627693e-01
-1.23576152e+00 -2.89428473e-01 9.56375241e-01 9.01091099e-01
2.90834397e-01 4.63340223e-01 5.86288512e-01 -7.53817976e-01
8.62279177e-01 -6.49808884e-01 -5.76763332e-01 1.83934972e-01
-6.01347923e-01 5.66664338e-01 7.27427006e-01 -2.76005000e-01
-7.99506903e-01 -2.65860349e-01 -4.25575852e-01 -1.31893056e-02
-8.74871686e-02 5.69397628e-01 -2.24436879e-01 -6.93434849e-02
2.83842206e-01 4.64553565e-01 -1.20797023e-01 -2.95605659e-01
-1.49551928e-01 1.14207900e+00 2.33482197e-01 -1.92467585e-01
3.67983848e-01 -4.27552760e-02 2.82422543e-01 -9.60731447e-01
-6.82712734e-01 -3.75034750e-01 -6.31795347e-01 -1.19311688e-02
5.16662836e-01 -4.39694732e-01 -1.01407206e+00 5.93143106e-01
-1.24519658e+00 8.85355696e-02 -8.32994878e-02 7.87175655e-01
-8.02294075e-01 6.38024092e-01 -4.59994316e-01 -1.08751106e+00
-3.19029361e-01 -9.96927381e-01 7.53517747e-01 2.04810485e-01
-2.61885449e-02 -1.35698783e+00 1.74197897e-01 -2.61630982e-01
4.54397619e-01 9.35894027e-02 6.12242579e-01 -1.15566146e+00
-2.15300277e-01 -3.44808459e-01 -3.98173660e-01 8.40955555e-01
2.05201417e-01 -2.86685079e-01 -7.97095299e-01 -2.73887843e-01
2.18529806e-01 -3.39306086e-01 9.26451981e-01 4.66482460e-01
1.64827955e+00 -4.92815554e-01 -2.14788780e-01 6.31242037e-01
1.30072522e+00 1.92443326e-01 7.71438897e-01 2.64232278e-01
1.53538123e-01 4.39050525e-01 3.31078649e-01 9.19447184e-01
7.24086761e-02 3.81651461e-01 2.41729736e-01 2.45209932e-01
4.37476188e-01 1.10714525e-01 -5.23070730e-02 1.29870701e+00
-1.94182158e-01 -6.33637488e-01 -7.08146095e-01 2.18917593e-01
-1.62838411e+00 -1.24294901e+00 4.25375029e-02 2.43941355e+00
1.06197214e+00 1.22867368e-01 -1.22580640e-01 6.68746769e-01
9.23792541e-01 2.02962860e-01 -5.76196849e-01 -2.84868032e-01
-9.88130271e-02 2.95546412e-01 7.39794850e-01 4.02678818e-01
-1.31727529e+00 3.17951411e-01 6.65544510e+00 1.13777721e+00
-6.34840667e-01 5.81845716e-02 8.63279343e-01 2.53914654e-01
7.98860788e-02 -2.80691266e-01 -8.22956979e-01 6.29886389e-01
1.18221390e+00 -4.04357821e-01 2.43051857e-01 7.51358211e-01
-4.83286411e-01 -2.34448463e-01 -1.13907552e+00 1.18612981e+00
1.80950508e-01 -7.64121830e-01 -3.39395553e-01 1.35998175e-01
7.52056718e-01 -8.12509134e-02 1.34104207e-01 2.96366185e-01
1.77219108e-01 -1.05334914e+00 8.27415660e-02 5.09208560e-01
9.31809545e-01 -1.01261735e+00 1.19497263e+00 5.36751211e-01
-9.95178401e-01 -1.04856580e-01 -6.75249100e-01 -2.65713364e-01
7.33829066e-02 1.12591350e+00 -8.08214962e-01 5.94823241e-01
5.09524047e-01 7.78535306e-01 -2.42924407e-01 1.26236522e+00
1.64302476e-02 4.92272049e-01 -5.82171142e-01 -5.04994452e-01
-1.90876022e-01 -1.28266394e-01 3.73102576e-01 1.23068428e+00
5.92402756e-01 -3.07557918e-02 -4.88009065e-01 6.56307578e-01
-3.48282725e-01 1.93154141e-01 -8.23006511e-01 3.05539067e-03
6.11754894e-01 8.87508571e-01 -6.30230308e-01 -4.08635527e-01
-3.14317554e-01 9.87080872e-01 3.46802324e-01 1.47809640e-01
-9.31530118e-01 -9.48000789e-01 3.51090550e-01 -4.20161843e-01
3.55214775e-01 -6.56864047e-02 -5.20161353e-02 -1.00497651e+00
8.21914673e-02 -5.19200563e-01 2.25183368e-01 -1.04342572e-01
-1.17975688e+00 4.46636021e-01 3.43518913e-01 -1.44558048e+00
-6.42710984e-01 -4.87958491e-01 -1.14281818e-01 7.69250572e-01
-1.27459216e+00 -8.46957490e-02 -1.42816782e-01 2.93595642e-01
1.97555885e-01 -4.02172565e-01 7.27362096e-01 5.89865804e-01
-4.07971472e-01 1.08312786e+00 1.02428043e+00 3.47694993e-01
5.45179963e-01 -1.14258218e+00 3.27465296e-01 5.94925046e-01
1.77046537e-01 5.12236893e-01 7.65645087e-01 -4.06615645e-01
-8.33948195e-01 -9.05414402e-01 9.39907551e-01 3.71364616e-02
6.35474324e-01 -2.00969309e-01 -7.36947000e-01 5.11088371e-01
1.73260912e-01 -3.41002643e-02 8.37494135e-01 2.83594914e-02
-8.62361980e-04 -3.96194197e-02 -1.34164023e+00 1.73030078e-01
9.68378544e-01 -3.49610418e-01 -3.37501585e-01 3.09497297e-01
3.86104554e-01 -2.70632297e-01 -1.14180005e+00 5.84568679e-01
8.74898672e-01 -1.37250364e+00 6.28031790e-01 -2.64401019e-01
6.61354899e-01 3.48010540e-01 -4.09532845e-01 -1.25181949e+00
-6.77621663e-02 -3.79977256e-01 -2.85747379e-01 1.10306835e+00
1.88784376e-01 -8.45944703e-01 3.68889451e-01 4.12897825e-01
3.10713410e-01 -7.04156995e-01 -1.20504975e+00 -1.00041759e+00
2.56737918e-02 -5.36445379e-01 4.02216226e-01 8.22981298e-01
2.56219327e-01 1.40233859e-01 -4.91324425e-01 -2.03984201e-01
6.95494592e-01 -2.44148567e-01 4.77052718e-01 -1.48667467e+00
-5.62538922e-01 -5.88337004e-01 -9.08939421e-01 -1.28085268e+00
2.04356775e-01 -9.87923384e-01 2.45528668e-01 -1.21291029e+00
5.30547917e-01 -2.08830923e-01 -5.64255297e-01 7.67395571e-02
3.79118063e-02 2.47089826e-02 9.08295587e-02 3.42211500e-02
-7.12880194e-01 7.19135761e-01 9.07386959e-01 -8.91730413e-02
8.24534073e-02 3.94795448e-01 -4.28181469e-01 5.92331409e-01
1.09728754e+00 -4.92854893e-01 -4.02195901e-01 4.20459062e-02
1.97853968e-01 2.11430460e-01 6.72576353e-02 -1.38711607e+00
2.51819968e-01 2.15103731e-01 3.19111675e-01 -5.75910985e-01
1.76275969e-01 -8.89776468e-01 1.97864454e-02 4.18193907e-01
-5.34844875e-01 -3.03429104e-02 1.20520115e-01 5.77750862e-01
-5.15017986e-01 -6.59979045e-01 7.92369187e-01 2.08625793e-01
-3.25317830e-01 3.91771048e-01 -2.43598163e-01 2.63294995e-01
9.48025465e-01 -5.43889180e-02 -1.23944208e-01 -6.89150095e-01
-4.88609105e-01 -8.75571445e-02 4.00511846e-02 2.62066610e-02
6.60815418e-01 -1.32461488e+00 -6.45213068e-01 2.57786125e-01
-1.13747589e-01 -1.86913759e-01 1.95260942e-01 1.01874387e+00
-3.28463405e-01 6.45348728e-01 -2.09562838e-01 -4.99446332e-01
-1.11892736e+00 2.20181599e-01 1.24390878e-01 -5.47439098e-01
-1.59713328e-01 7.01833725e-01 4.09383960e-02 -2.08280206e-01
4.68767405e-01 -2.91814625e-01 -3.19326699e-01 -8.00214708e-02
5.77450454e-01 5.00851512e-01 8.99122376e-03 -4.84898388e-01
-2.81030267e-01 2.66502798e-01 -1.40366673e-01 -2.19809264e-01
1.25394046e+00 -1.76417843e-01 -1.64355591e-01 6.90786481e-01
1.88733852e+00 -3.47901642e-01 -7.35159755e-01 -3.79887849e-01
-2.98780873e-02 -2.61614352e-01 4.26202416e-02 -4.38572496e-01
-9.99305785e-01 1.07941055e+00 7.58360565e-01 6.11326635e-01
1.36431229e+00 -1.79098055e-01 6.60769463e-01 8.48104656e-01
3.66349518e-01 -1.24563944e+00 -3.81749094e-01 8.48386824e-01
4.82457101e-01 -1.37219620e+00 7.64749944e-02 -7.04748183e-02
-2.30006069e-01 1.25609875e+00 1.36921316e-01 1.99508831e-01
1.01716304e+00 3.91506672e-01 -7.75700450e-01 3.75671893e-01
-5.90602398e-01 -1.40579730e-01 1.99109524e-01 4.74839628e-01
7.65160978e-01 2.07648166e-02 -6.34413481e-01 4.87807900e-01
-5.43548882e-01 1.22815512e-01 4.14134592e-01 5.87299347e-01
-5.38951218e-01 -1.00087667e+00 -1.72003284e-01 8.95724952e-01
-5.32488704e-01 -4.32979017e-01 -1.46363422e-01 8.86139035e-01
-1.21215194e-01 7.22239912e-01 1.69947341e-01 -5.04936635e-01
-2.62628168e-01 -2.03840155e-02 5.00813484e-01 -2.21526958e-02
1.33930758e-01 -5.70230365e-01 -9.12130922e-02 -9.40407738e-02
-6.57331288e-01 -7.52133429e-01 -7.77707338e-01 -3.95950496e-01
-6.07847869e-01 2.16969132e-01 6.34762824e-01 1.20668316e+00
4.22156528e-02 5.33253849e-01 6.87419176e-01 -4.91240084e-01
-7.28439033e-01 -1.33087182e+00 -9.89891887e-01 1.21223100e-01
4.21400726e-01 -7.78797805e-01 -5.30625939e-01 -2.23490775e-01] | [7.636730670928955, 3.7512238025665283] |
53b5869f-dc3b-4cd1-8f7d-c89a664dcdfb | fcl-gan-a-lightweight-and-real-time-baseline | 2204.07820 | null | https://arxiv.org/abs/2204.07820v2 | https://arxiv.org/pdf/2204.07820v2.pdf | FCL-GAN: A Lightweight and Real-Time Baseline for Unsupervised Blind Image Deblurring | Blind image deblurring (BID) remains a challenging and significant task. Benefiting from the strong fitting ability of deep learning, paired data-driven supervised BID method has obtained great progress. However, paired data are usually synthesized by hand, and the realistic blurs are more complex than synthetic ones, which makes the supervised methods inept at modeling realistic blurs and hinders their real-world applications. As such, unsupervised deep BID method without paired data offers certain advantages, but current methods still suffer from some drawbacks, e.g., bulky model size, long inference time, and strict image resolution and domain requirements. In this paper, we propose a lightweight and real-time unsupervised BID baseline, termed Frequency-domain Contrastive Loss Constrained Lightweight CycleGAN (shortly, FCL-GAN), with attractive properties, i.e., no image domain limitation, no image resolution limitation, 25x lighter than SOTA, and 5x faster than SOTA. To guarantee the lightweight property and performance superiority, two new collaboration units called lightweight domain conversion unit(LDCU) and parameter-free frequency-domain contrastive unit(PFCU) are designed. LDCU mainly implements inter-domain conversion in lightweight manner. PFCU further explores the similarity measure, external difference and internal connection between the blurred domain and sharp domain images in frequency domain, without involving extra parameters. Extensive experiments on several image datasets demonstrate the effectiveness of our FCL-GAN in terms of performance, model size and reference time. | ['Meng Wang', 'Yi Yang', 'Mingliang Xu', 'Richang Hong', 'Zhao Zhang', 'Suiyi Zhao'] | 2022-04-16 | null | null | null | null | ['blind-image-deblurring'] | ['computer-vision'] | [ 1.46442205e-01 -4.60312992e-01 -9.71047431e-02 -2.14457706e-01
-5.42846382e-01 -4.29682434e-01 6.35726392e-01 -8.98231924e-01
-1.31472781e-01 8.84937108e-01 4.13026482e-01 6.83009764e-03
-2.85950303e-01 -4.70452517e-01 -5.41993678e-01 -9.89605546e-01
4.24988300e-01 -1.79863274e-01 2.02547893e-01 -1.30839646e-01
5.08318190e-03 5.01889400e-02 -1.15020335e+00 -1.60246462e-01
1.41303289e+00 1.02908981e+00 5.86160958e-01 2.29166836e-01
5.10039143e-02 8.17661166e-01 -4.20462340e-01 -2.34988481e-01
3.42027307e-01 -8.06377172e-01 -3.83177936e-01 1.06619373e-01
2.07995251e-01 -7.75463462e-01 -6.67912662e-01 1.29912484e+00
7.15052783e-01 5.33490675e-03 4.99657840e-01 -1.15981627e+00
-1.09028637e+00 3.03969443e-01 -8.75281990e-01 1.60414636e-01
1.05214074e-01 5.33555686e-01 4.49234307e-01 -7.48015285e-01
3.03038061e-01 1.09459972e+00 5.43932021e-01 6.42407119e-01
-1.15277028e+00 -8.99127185e-01 5.28173037e-02 3.13769907e-01
-1.21344304e+00 -5.54755569e-01 8.85279834e-01 -3.10302526e-01
4.03451979e-01 1.57325119e-01 3.43050092e-01 1.19924533e+00
-1.36751324e-01 5.39779425e-01 1.58370388e+00 -1.41093343e-01
1.08631223e-01 3.19847316e-02 -2.08299175e-01 4.60114211e-01
2.84352154e-01 3.53974432e-01 -2.95090079e-01 1.83723509e-01
1.20882034e+00 2.17870902e-02 -9.21629488e-01 -2.50103861e-01
-1.21805620e+00 3.96092355e-01 5.71576238e-01 9.32063088e-02
-8.73824283e-02 6.86042830e-02 2.80657917e-01 2.94449985e-01
5.32197416e-01 7.95204118e-02 -3.00070852e-01 -1.13630861e-01
-1.07758689e+00 1.50340702e-02 1.62303090e-01 1.13950396e+00
6.54947698e-01 2.59510428e-01 -2.29273692e-01 9.65150535e-01
2.00769052e-01 6.89955056e-01 8.10943425e-01 -9.34378386e-01
3.34541440e-01 2.13236511e-01 1.77102551e-01 -8.28259170e-01
8.46566409e-02 -4.29822296e-01 -1.35417271e+00 8.51262808e-02
6.46444038e-02 -2.04255298e-01 -8.96181166e-01 1.84676480e+00
1.88273017e-03 3.86977613e-01 2.91273594e-02 1.35800600e+00
7.42068708e-01 8.81532669e-01 -2.06170097e-01 -5.78970790e-01
1.31783605e+00 -1.02866197e+00 -9.63759005e-01 -2.45374098e-01
-1.79407727e-02 -9.34239864e-01 1.21978724e+00 4.24137622e-01
-1.20037401e+00 -7.72911072e-01 -1.00784349e+00 -1.83279708e-01
9.22406837e-02 1.36755437e-01 5.90506077e-01 6.14804089e-01
-8.35917711e-01 2.49802187e-01 -4.28426325e-01 -5.01192436e-02
6.76241338e-01 1.81741968e-01 -1.43358141e-01 -3.42933148e-01
-1.33751094e+00 6.64681435e-01 2.85197854e-01 1.56521097e-01
-8.42019141e-01 -7.73760676e-01 -6.55983269e-01 1.16390642e-02
3.01160038e-01 -7.83202171e-01 1.16115046e+00 -9.88472700e-01
-1.80473447e+00 6.15461111e-01 -8.97826403e-02 -4.54663604e-01
8.17873299e-01 -3.34266782e-01 -6.83390975e-01 6.07701913e-02
-6.51644692e-02 4.91570354e-01 1.40493679e+00 -1.12210631e+00
-3.91691566e-01 -5.76394349e-02 -7.21885189e-02 2.48645693e-01
-5.10083735e-01 -7.12387869e-03 -5.25827587e-01 -9.91883814e-01
-4.88275290e-02 -6.03804290e-01 1.54579476e-01 2.15136945e-01
-1.58432409e-01 8.11036453e-02 1.09075153e+00 -8.26479912e-01
1.22525179e+00 -2.43472934e+00 -1.03814481e-02 -3.94250840e-01
3.45005989e-01 5.60705602e-01 -4.16182214e-03 -4.10958007e-02
-6.95958659e-02 -1.17046393e-01 -5.87604403e-01 -2.96004787e-02
-9.09708738e-02 7.58715793e-02 -3.94030392e-01 4.99087960e-01
1.02620378e-01 9.07493949e-01 -9.15403903e-01 -4.90233243e-01
2.92352617e-01 4.64161366e-01 -5.36457956e-01 4.82804179e-01
-2.66590267e-01 5.78636169e-01 -2.10953474e-01 4.68743443e-01
1.35116267e+00 -2.52757967e-01 -1.96749628e-01 -6.21557474e-01
-2.64663547e-01 -7.63746053e-02 -9.98742521e-01 1.75143707e+00
-6.54410422e-01 6.15681529e-01 2.43118837e-01 -8.89440894e-01
9.04255629e-01 2.32080072e-01 2.47220434e-02 -9.44761992e-01
1.78568244e-01 4.17005897e-01 -1.39697924e-01 -6.60187244e-01
1.80537641e-01 -2.13807344e-01 2.83458948e-01 3.96941543e-01
-2.91579552e-02 -2.09257692e-01 -1.89065248e-01 2.74393652e-02
7.07900524e-01 3.50307792e-01 1.25962719e-01 -4.02849078e-01
7.88784087e-01 -4.05598700e-01 7.94248521e-01 2.05376431e-01
-2.71910667e-01 1.02878833e+00 1.97401032e-01 -3.89255062e-02
-1.05681598e+00 -1.04851019e+00 -1.08571902e-01 4.41385269e-01
8.32973719e-01 1.13568073e-02 -8.74293923e-01 -5.12032211e-01
-3.53827745e-01 3.48451048e-01 -2.31136963e-01 -3.47263753e-01
-6.65610671e-01 -7.87861347e-01 4.80024368e-01 3.84819120e-01
1.52296317e+00 -7.31845319e-01 -2.13482186e-01 -6.83932826e-02
-3.55668902e-01 -1.26482463e+00 -1.07888460e+00 -3.78166676e-01
-7.51807094e-01 -8.19861770e-01 -1.30079329e+00 -1.14344370e+00
6.72354460e-01 8.20335388e-01 8.01464677e-01 -3.05135489e-01
-4.53879274e-02 -1.46158472e-01 -2.15368867e-01 -5.83950989e-02
-7.97819048e-02 -3.70274991e-01 1.41732126e-01 2.58546501e-01
1.24943359e-02 -7.67980695e-01 -1.16209185e+00 5.14943242e-01
-1.06496942e+00 2.99241632e-01 9.61955726e-01 1.21036279e+00
3.35080534e-01 3.27907711e-01 7.02298999e-01 -4.87486541e-01
7.19949007e-01 -2.23380104e-01 -7.65423179e-01 1.53335601e-01
-7.05506086e-01 -2.50322614e-02 9.36365008e-01 -5.23754299e-01
-1.66529155e+00 -3.95897746e-01 1.07867241e-01 -6.58810019e-01
1.29222516e-02 2.15443596e-01 -5.56999624e-01 -7.08471537e-02
5.63528776e-01 6.85810328e-01 2.08282113e-01 -5.84648609e-01
3.87574673e-01 9.04681981e-01 9.22375023e-01 -3.24363858e-01
1.07076573e+00 3.63567054e-01 -4.33228344e-01 -5.96287549e-01
-5.19537807e-01 -1.08370729e-01 -2.96385977e-02 -5.26960231e-02
7.10791767e-01 -1.29574978e+00 -5.86971402e-01 1.06545925e+00
-1.16533411e+00 -3.05211306e-01 -3.03073451e-02 6.31516278e-01
-4.05700475e-01 6.71802342e-01 -6.30794287e-01 -4.82491761e-01
-4.71215218e-01 -1.14892471e+00 6.58239424e-01 6.10114336e-01
3.23018640e-01 -7.54553854e-01 -3.50362659e-01 4.85065520e-01
8.35578442e-01 -6.52007982e-02 7.64330626e-01 2.57096887e-01
-6.76343918e-01 3.01061988e-01 -9.06978071e-01 8.15685451e-01
5.05486727e-01 -4.54164326e-01 -1.14510262e+00 -4.21773314e-01
3.82397145e-01 -2.32423767e-01 7.77373374e-01 4.04894799e-01
1.41739488e+00 -4.23062414e-01 -7.10643530e-02 9.71974552e-01
1.53487873e+00 1.98199928e-01 8.75912726e-01 8.38691294e-02
6.41093612e-01 2.04124942e-01 5.96851349e-01 2.41944298e-01
1.14423648e-01 6.28773928e-01 4.16830510e-01 -3.26702148e-01
-6.57672226e-01 -3.11993241e-01 4.91242945e-01 9.62583184e-01
-6.80275038e-02 -2.02912301e-01 -3.31551969e-01 4.76485163e-01
-1.56869924e+00 -8.80919993e-01 2.14644074e-02 2.17439485e+00
1.19556189e+00 9.02801529e-02 -7.88323358e-02 -1.01084396e-01
9.81169462e-01 2.64164388e-01 -8.48417401e-01 8.93914979e-03
-3.98346186e-01 1.13008790e-01 4.88897562e-01 3.61205995e-01
-8.63025010e-01 6.60805762e-01 4.81627178e+00 1.24093080e+00
-1.13576472e+00 3.30526471e-01 7.59118319e-01 4.90924194e-02
-3.08675915e-01 2.36008782e-02 -3.24158221e-01 1.14083850e+00
4.18145984e-01 -2.14258894e-01 8.52531135e-01 5.75110793e-01
3.78638387e-01 8.36303085e-03 -6.26379311e-01 1.45020008e+00
-1.07454330e-01 -1.22704291e+00 -8.86399001e-02 8.61367956e-02
8.47200155e-01 -1.93008199e-01 3.74712974e-01 9.22341868e-02
3.19744051e-02 -8.80247831e-01 6.59897327e-01 3.40177119e-01
1.45752501e+00 -5.99727988e-01 7.70160913e-01 2.28562668e-01
-9.16442275e-01 -7.71480501e-02 -4.32739496e-01 -3.98277268e-02
6.66788667e-02 8.98108780e-01 -1.00609906e-01 7.91953743e-01
8.23631942e-01 9.72885907e-01 -2.61924654e-01 9.60304260e-01
-1.99631095e-01 6.83627486e-01 -9.23315138e-02 3.19395959e-01
1.16533808e-01 -4.65243161e-01 5.27663469e-01 9.88281369e-01
5.06538033e-01 2.22221673e-01 -3.06764245e-01 1.07769215e+00
-1.95429057e-01 -3.39749753e-01 -3.44981194e-01 2.72130787e-01
6.59678221e-01 1.09674144e+00 -2.15532854e-01 -1.37349457e-01
-5.32480717e-01 1.33738327e+00 -2.78355092e-01 4.41022307e-01
-1.17497563e+00 -5.15060782e-01 7.18733966e-01 2.81232962e-04
2.62842208e-01 -1.19825408e-01 -2.74440199e-01 -1.46025610e+00
2.31548041e-01 -1.01807749e+00 -4.00152244e-02 -1.07467675e+00
-1.44476390e+00 6.03437543e-01 -1.62894547e-01 -1.79948819e+00
2.24475816e-01 -3.47729713e-01 -6.79835737e-01 1.19550097e+00
-1.88461256e+00 -1.16028762e+00 -8.99104416e-01 8.31350207e-01
6.64569438e-01 -1.09621614e-01 3.16639692e-01 4.77938861e-01
-6.84976459e-01 7.31934488e-01 4.26269174e-01 4.43640389e-02
1.11787021e+00 -9.35361862e-01 2.68308073e-01 1.05856574e+00
-5.46323121e-01 5.37179708e-01 6.34267032e-01 -4.05292571e-01
-1.43036413e+00 -1.19481599e+00 2.60456800e-01 1.21990263e-01
4.78554189e-01 -2.40974545e-01 -9.23508227e-01 1.89078093e-01
4.97502118e-01 2.26779103e-01 1.15552634e-01 -6.42048955e-01
-3.44822764e-01 -4.95303363e-01 -1.25557625e+00 6.44697070e-01
1.21583962e+00 -5.36990464e-01 -3.99594247e-01 2.55365610e-01
1.00079393e+00 -3.93656731e-01 -6.04840338e-01 4.12594289e-01
3.50855023e-01 -1.29296362e+00 1.01395488e+00 2.14576304e-01
6.03003979e-01 -6.32563412e-01 1.10474363e-01 -1.36302018e+00
-4.78128791e-01 -1.03861868e+00 -2.74285078e-01 1.53684998e+00
-1.59487039e-01 -1.04938340e+00 4.64770496e-01 2.81388432e-01
-2.51551896e-01 -3.63454193e-01 -6.79910898e-01 -1.12173331e+00
-2.12984364e-02 7.74098113e-02 8.33153784e-01 9.90978301e-01
-3.84067953e-01 3.19383442e-01 -7.75040150e-01 1.70956343e-01
6.94001138e-01 2.13583559e-01 5.73452592e-01 -8.87447000e-01
-4.06944543e-01 -4.16216880e-01 -2.02393949e-01 -1.57896411e+00
-2.67877907e-01 -4.01271075e-01 2.52896603e-02 -1.23631287e+00
1.21128231e-01 -6.02743328e-01 -1.29748061e-01 1.15164272e-01
-2.75974095e-01 3.35084260e-01 4.92651798e-02 6.10100985e-01
-2.49633536e-01 7.86913276e-01 1.63829756e+00 -2.18411803e-01
-5.65661676e-03 -3.50670546e-01 -7.05742121e-01 5.52938879e-01
7.57073641e-01 -1.85106453e-02 -6.82380855e-01 -8.62285674e-01
-2.37538785e-01 9.48884785e-02 4.56656575e-01 -1.04246104e+00
2.97790229e-01 -2.18897417e-01 3.47998440e-01 -1.67008057e-01
3.03162575e-01 -6.54072940e-01 2.09316984e-01 3.65322053e-01
5.18375374e-02 -3.37615699e-01 -7.43673742e-03 5.64718664e-01
-5.10617077e-01 4.71378043e-02 1.10427964e+00 -4.44291644e-02
-8.61333013e-01 4.36908931e-01 1.38375938e-01 1.44999474e-01
8.28104973e-01 -3.87794018e-01 -6.68388784e-01 -4.56304997e-01
-1.40920460e-01 7.75725916e-02 7.16775596e-01 2.44704053e-01
5.81442177e-01 -1.37206233e+00 -7.96006620e-01 4.56250757e-01
-2.85647198e-04 3.35062325e-01 7.41182685e-01 9.61402595e-01
-5.68509400e-01 2.53491014e-01 -2.01550215e-01 -3.91493827e-01
-8.92601788e-01 6.02404654e-01 1.89615056e-01 7.49805719e-02
-6.98933840e-01 9.10255253e-01 6.68334961e-01 -3.28101777e-02
4.22515199e-02 -1.27476439e-01 7.43572488e-02 -2.74977654e-01
6.36955678e-01 3.55388641e-01 -2.61834599e-02 -3.88545275e-01
-1.29353553e-01 6.73088253e-01 -8.26323479e-02 1.77998379e-01
1.17870295e+00 -3.98590982e-01 -2.58903116e-01 -2.79642373e-01
1.20661294e+00 1.57973189e-02 -1.66524541e+00 -4.77489769e-01
-5.42182148e-01 -8.00057232e-01 3.47135246e-01 -9.08676922e-01
-1.29044330e+00 8.37232351e-01 8.14847350e-01 5.09038977e-02
1.64761281e+00 -2.72899628e-01 1.22430992e+00 -3.39445889e-01
2.79343158e-01 -8.54870737e-01 1.89360395e-01 1.51211500e-01
8.67613554e-01 -1.17165267e+00 5.79874441e-02 -5.45106888e-01
-5.05944371e-01 8.28400075e-01 8.25513542e-01 -1.17467999e-01
4.07892108e-01 9.12942365e-02 6.37190603e-03 2.98165947e-01
-4.23228115e-01 3.46136205e-02 1.02663063e-01 7.88982868e-01
1.72655106e-01 -2.27195412e-01 -4.08193320e-01 5.94708085e-01
-2.26174891e-02 2.28837743e-01 5.28649569e-01 5.32052875e-01
-1.60498187e-01 -8.03368866e-01 -4.82323021e-01 3.86770427e-01
-2.81570226e-01 -2.70483494e-01 1.84299380e-01 5.22640646e-01
3.88856411e-01 9.18382108e-01 -6.64415807e-02 -3.35357517e-01
1.13153666e-01 -5.84343255e-01 4.75306213e-01 -7.50512779e-02
7.71875819e-03 1.95351288e-01 -4.00070012e-01 -2.56921530e-01
-5.01336396e-01 -3.76250684e-01 -8.88626158e-01 -4.35855716e-01
-5.65784574e-01 1.29753441e-01 2.28881046e-01 7.78554142e-01
4.52418089e-01 4.26975399e-01 9.53585625e-01 -7.28162467e-01
-6.03927135e-01 -9.68237281e-01 -4.94844109e-01 2.39441514e-01
5.77142060e-01 -6.18182361e-01 -4.78095293e-01 2.55926013e-01] | [11.422733306884766, -2.608238458633423] |
2c32c64b-a033-48c0-a132-e78fc33a8910 | patch-based-progressive-3d-point-set | 1811.11286 | null | http://arxiv.org/abs/1811.11286v3 | http://arxiv.org/pdf/1811.11286v3.pdf | Patch-based Progressive 3D Point Set Upsampling | We present a detail-driven deep neural network for point set upsampling. A
high-resolution point set is essential for point-based rendering and surface
reconstruction. Inspired by the recent success of neural image super-resolution
techniques, we progressively train a cascade of patch-based upsampling networks
on different levels of detail end-to-end. We propose a series of architectural
design contributions that lead to a substantial performance boost. The effect
of each technical contribution is demonstrated in an ablation study.
Qualitative and quantitative experiments show that our method significantly
outperforms the state-of-the-art learning-based and optimazation-based
approaches, both in terms of handling low-resolution inputs and revealing
high-fidelity details. | ['Olga Sorkine-Hornung', 'Daniel Cohen-Or', 'Shihao Wu', 'Hui Huang', 'Wang Yifan'] | 2018-11-27 | patch-based-progressive-3d-point-set-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Yifan_Patch-Based_Progressive_3D_Point_Set_Upsampling_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Yifan_Patch-Based_Progressive_3D_Point_Set_Upsampling_CVPR_2019_paper.pdf | cvpr-2019-6 | ['point-cloud-super-resolution', 'point-set-upsampling'] | ['computer-vision', 'computer-vision'] | [ 4.77932245e-01 1.28881857e-01 3.07037354e-01 -2.86562592e-01
-1.22432959e+00 -1.89369656e-02 6.50084496e-01 -3.04941297e-01
1.67461634e-01 6.88259065e-01 2.53040433e-01 5.31431427e-03
-9.59759951e-02 -1.16054296e+00 -1.11060059e+00 -1.14405632e-01
-1.39562666e-01 2.99238205e-01 3.10037404e-01 -5.58411837e-01
3.47144425e-01 1.09825742e+00 -1.92940307e+00 8.52822006e-01
5.82909167e-01 1.08579743e+00 1.38215758e-02 7.20906019e-01
-2.27355167e-01 5.17259598e-01 -2.51783848e-01 -1.37605309e-01
4.03938323e-01 -2.22334385e-01 -7.53962517e-01 -7.39673432e-03
1.15350819e+00 -8.30436707e-01 -2.48602152e-01 5.41942894e-01
5.69372892e-01 6.17246330e-03 4.19850081e-01 -2.93331504e-01
-6.93239331e-01 5.51470518e-02 -7.98226058e-01 5.93615025e-02
2.73924887e-01 1.18957520e-01 7.81194150e-01 -1.30864668e+00
8.45980525e-01 1.38939071e+00 1.21506965e+00 4.15120244e-01
-1.66739106e+00 -5.34169674e-01 3.78696360e-02 -3.42961937e-01
-1.26256776e+00 -5.69748402e-01 1.20730376e+00 -2.24040613e-01
1.18912804e+00 2.48699248e-01 8.23255718e-01 7.33279943e-01
1.68409750e-01 3.28019828e-01 1.15110004e+00 -2.90929884e-01
3.10305059e-01 -4.49413836e-01 -3.55181366e-01 6.15890861e-01
-5.69260567e-02 5.46172559e-01 -7.14037061e-01 -2.69182742e-01
2.11205292e+00 -2.24284548e-02 -2.27102578e-01 -3.65577728e-01
-1.07985497e+00 6.53336346e-01 8.60042393e-01 1.90437868e-01
-6.91105425e-01 5.86059570e-01 7.44755641e-02 5.23325130e-02
1.03831339e+00 6.85826242e-01 -4.43778187e-01 7.52853677e-02
-1.32322443e+00 5.15296400e-01 1.58738777e-01 7.51264334e-01
1.06816614e+00 4.57841456e-01 5.77986240e-03 8.77948701e-01
5.25029562e-02 1.15478449e-01 -1.37085095e-01 -1.53722525e+00
-3.68435495e-02 4.73344922e-01 3.28307509e-01 -9.50084150e-01
-3.37344497e-01 -7.56214797e-01 -8.56410265e-01 1.07382047e+00
-8.68440419e-02 2.13750482e-01 -1.07421088e+00 1.20787668e+00
3.19316030e-01 3.48988026e-01 -3.32226425e-01 9.20354962e-01
1.08506155e+00 5.12813926e-01 -1.14773311e-01 2.03416929e-01
1.19020450e+00 -7.02662349e-01 -4.58578259e-01 2.06808783e-02
-7.81999342e-03 -7.54191816e-01 1.29359186e+00 3.78295809e-01
-1.68997526e+00 -6.07354462e-01 -1.14713264e+00 -5.80332100e-01
1.33590817e-01 -1.67412296e-01 7.93635249e-01 1.35457128e-01
-1.38298357e+00 1.24933445e+00 -7.80002058e-01 1.17623091e-01
1.09009004e+00 1.89171463e-01 -2.39751950e-01 -2.31209770e-02
-7.58368373e-01 3.72846752e-01 -2.54571438e-01 -8.10303837e-02
-8.63171816e-01 -1.56092787e+00 -7.09872842e-01 1.88191950e-01
-1.96475774e-01 -1.30610132e+00 1.32742202e+00 -9.07331884e-01
-1.64807141e+00 9.75485802e-01 -1.48109093e-01 -3.62552911e-01
6.21430159e-01 -2.12780446e-01 -2.10454222e-03 4.90574211e-01
-4.81678993e-02 7.09662735e-01 9.49234009e-01 -1.76739812e+00
-6.12631261e-01 -2.41791412e-01 1.10261716e-01 1.62966758e-01
2.01346934e-01 -3.06507554e-02 -1.15564518e-01 -6.48292303e-01
2.28563607e-01 -8.06339085e-02 -4.44318920e-01 6.32615209e-01
2.91001592e-02 2.34101936e-01 9.24647331e-01 -6.72002256e-01
6.39194250e-01 -1.98248625e+00 -2.02803195e-01 -1.22092418e-01
6.02374077e-01 2.57818270e-02 -2.00283304e-02 3.97721648e-01
-2.11831361e-01 1.75610363e-01 -4.86988187e-01 -7.71494985e-01
-4.03548717e-01 -2.27960482e-01 -4.60208893e-01 2.60270208e-01
6.43712640e-01 9.77842212e-01 -7.02639401e-01 -1.90667287e-01
7.05095530e-01 1.50935864e+00 -7.14553416e-01 1.38406038e-01
-3.48456413e-01 6.85892224e-01 -3.47615093e-01 6.47780299e-01
9.09471750e-01 -5.34872472e-01 -3.54632229e-01 -6.81576669e-01
-4.09106374e-01 2.97076851e-01 -9.95722651e-01 2.15785122e+00
-8.37751389e-01 7.16420352e-01 2.65214443e-01 -2.63122171e-01
8.69038939e-01 2.25033596e-01 4.10326809e-01 -8.02004755e-01
-1.54289931e-01 1.86820090e-01 -5.74789166e-01 6.43313229e-02
6.72819257e-01 -4.54878688e-01 5.02204597e-01 2.89979279e-02
-1.01018138e-01 -4.99580592e-01 -6.29365146e-01 2.50370838e-02
8.79123449e-01 6.56104505e-01 1.48684189e-01 -4.15743083e-01
2.34633103e-01 1.39702754e-02 1.98840797e-01 4.13562626e-01
3.05151612e-01 1.48538876e+00 1.30020514e-01 -9.84765887e-01
-1.68115699e+00 -1.05226517e+00 -4.48218882e-01 8.59743059e-01
-4.81073074e-02 -2.48612225e-01 -5.49402475e-01 6.37645498e-02
-9.76229981e-02 4.84062344e-01 -8.64516020e-01 4.14770126e-01
-9.48635399e-01 -1.93524942e-01 5.99557422e-02 7.10390985e-01
7.03446269e-01 -9.26234901e-01 -8.55386496e-01 2.79754043e-01
2.29620054e-01 -9.46247935e-01 6.65366799e-02 -2.42967024e-01
-1.28096437e+00 -6.81691825e-01 -9.18572843e-01 -6.15751266e-01
4.85214353e-01 3.12065721e-01 1.86317968e+00 3.05544436e-01
-3.52037162e-01 2.11649448e-01 1.74770832e-01 -9.20712799e-02
-3.65013272e-01 -3.54466289e-02 -4.13507253e-01 -2.51149267e-01
-4.24199790e-01 -1.44253731e+00 -1.21741855e+00 5.92446737e-02
-8.88158917e-01 5.73065877e-01 5.96270382e-01 6.21650457e-01
1.10874009e+00 -1.62005737e-01 2.18964219e-01 -6.84037566e-01
4.23329949e-01 5.64186051e-02 -5.66321075e-01 -3.46283972e-01
-2.29878008e-01 -8.41246080e-03 5.81728816e-01 -1.28496215e-01
-1.12529242e+00 -9.63469893e-02 -5.40784955e-01 -7.75062442e-01
-2.85434693e-01 9.11083296e-02 1.83270976e-01 -5.99272966e-01
1.02848554e+00 -4.06645276e-02 -4.73652929e-02 -7.13320971e-01
3.82417679e-01 -7.09310174e-03 6.04065180e-01 -5.53057194e-01
7.09622622e-01 1.11046767e+00 3.66741180e-01 -9.96808171e-01
-6.75275505e-01 1.09000511e-01 -5.79763174e-01 -2.54785389e-01
6.23417139e-01 -1.01526582e+00 -5.15289605e-01 2.41111279e-01
-1.26412058e+00 -5.32153249e-01 -7.36418724e-01 -2.00017393e-01
-8.14572752e-01 -5.57410382e-02 -7.92278528e-01 -6.15944445e-01
-6.24366045e-01 -1.06653512e+00 1.70142341e+00 1.34591207e-01
-1.06722526e-01 -7.29443431e-01 3.46380286e-02 2.40963474e-02
9.20277357e-01 7.23220825e-01 6.64445162e-01 3.90794694e-01
-1.06881332e+00 5.88263683e-02 -5.98470807e-01 1.70111626e-01
1.38910562e-02 1.51081219e-01 -1.26840246e+00 -1.87321246e-01
-5.35498597e-02 -1.60225183e-01 7.33472824e-01 7.61801839e-01
1.09618890e+00 -8.53210315e-02 -1.98144063e-01 1.24111581e+00
1.85210145e+00 -4.19542730e-01 9.52155769e-01 3.97580892e-01
9.54138756e-01 3.72669697e-01 1.69079781e-01 1.84885263e-01
5.63669018e-02 6.61823153e-01 6.40122175e-01 -5.95898211e-01
-7.10087061e-01 -4.38361883e-01 -4.55329150e-01 1.46863446e-01
-4.30424809e-01 4.11935210e-01 -8.23371768e-01 4.52100813e-01
-1.47580290e+00 -8.59985888e-01 -9.83899534e-02 2.09149075e+00
6.69521987e-01 -4.83865291e-02 9.24105272e-02 -1.45406211e-02
4.46346134e-01 4.37178254e-01 -5.91698110e-01 -3.55854630e-01
-1.63518280e-01 6.13454878e-01 1.93834007e-01 8.78196061e-01
-8.41154873e-01 9.71731424e-01 7.47691679e+00 8.94266784e-01
-1.22213829e+00 7.47528374e-02 9.67480361e-01 -2.16005400e-01
-8.18433225e-01 -2.59008467e-01 -3.84232491e-01 -6.11490756e-02
7.24100590e-01 4.60345000e-02 5.22122681e-01 8.57541800e-01
1.75770223e-01 2.35179767e-01 -1.00669837e+00 1.12697029e+00
-1.15214743e-01 -2.21717930e+00 3.00732881e-01 5.07084802e-02
1.00553381e+00 3.44241202e-01 2.13937521e-01 -8.11342821e-02
2.22831428e-01 -1.30362952e+00 6.23650789e-01 5.90683341e-01
1.40539038e+00 -9.95442629e-01 1.25686929e-01 -5.04191630e-02
-1.25286007e+00 3.36893231e-01 -3.38288665e-01 -2.45793924e-01
2.92395234e-01 7.84683824e-01 -1.41025364e-01 5.40510416e-01
9.04553652e-01 6.57605827e-01 -1.78481162e-01 7.77863443e-01
-5.61928339e-02 3.33675832e-01 -3.92398596e-01 6.21541858e-01
9.30140018e-02 -7.99224600e-02 5.28136790e-01 9.77036595e-01
6.16247393e-02 2.68850207e-01 -3.29183102e-01 1.60279119e+00
-1.87405631e-01 -2.00227454e-01 -5.89767575e-01 4.88928378e-01
3.41712207e-01 1.26040697e+00 -6.21163607e-01 -3.44312906e-01
-1.95776120e-01 9.83996272e-01 5.74985266e-01 4.73542631e-01
-4.11311358e-01 -1.59181297e-01 8.35411906e-01 8.17024469e-01
5.38587272e-01 -3.16322148e-01 -8.98299634e-01 -7.80634165e-01
9.18591488e-03 -5.85455358e-01 -2.81063199e-01 -1.09827971e+00
-1.25703871e+00 9.94123042e-01 -9.22747403e-02 -1.27868223e+00
-1.44207209e-01 -2.56504804e-01 -4.73707080e-01 1.41375947e+00
-1.85128725e+00 -1.37068057e+00 -5.51349103e-01 2.43610919e-01
5.34401059e-01 3.11756253e-01 9.73726034e-01 2.10419133e-01
1.41452715e-01 2.23116562e-01 -1.48721740e-01 -1.57215536e-01
1.31627664e-01 -8.74367058e-01 1.04427946e+00 7.59328544e-01
-3.07660729e-01 5.93474269e-01 7.27254033e-01 -5.96356928e-01
-1.24672914e+00 -9.97499883e-01 2.29724705e-01 -5.08480251e-01
4.34477851e-02 -2.47138366e-01 -1.19882929e+00 5.36688685e-01
-6.36982918e-03 3.50321680e-01 1.86160073e-01 9.42119062e-02
-3.49907279e-01 -1.39891938e-03 -1.61908245e+00 7.79278457e-01
1.23279738e+00 -6.63744271e-01 -4.17363256e-01 -9.27791186e-03
9.64243352e-01 -8.21861148e-01 -1.13194776e+00 6.85974836e-01
6.32956147e-01 -1.62564003e+00 1.44215405e+00 -2.11139604e-01
1.25092149e+00 -2.09135771e-01 -1.83308143e-02 -1.11323118e+00
-9.10004616e-01 -7.34171033e-01 -3.85718673e-01 7.43572891e-01
-2.28458606e-02 -4.70158041e-01 1.14056528e+00 4.27742720e-01
-3.83076578e-01 -1.14915168e+00 -1.00738049e+00 -2.88408041e-01
3.26187670e-01 -3.13105315e-01 9.89991724e-01 8.67674470e-01
-6.51638329e-01 2.11752281e-01 -1.61295965e-01 2.44476721e-01
9.45550919e-01 2.66478688e-01 6.78904772e-01 -1.18834138e+00
-8.18118155e-02 -6.60068512e-01 -3.80818456e-01 -1.15827107e+00
-4.64973092e-01 -2.25055039e-01 -1.15290031e-01 -1.97240913e+00
-2.22302020e-01 -3.74337673e-01 -1.06155477e-01 -1.00796819e-02
-8.50016549e-02 8.08236420e-01 -1.89671114e-01 2.55678326e-01
-2.31867582e-02 6.27076864e-01 1.52781367e+00 2.86482364e-01
-2.91866928e-01 -3.03403646e-01 -7.58134604e-01 8.30878675e-01
6.77544773e-01 -2.64733192e-03 -3.28374356e-01 -8.21447551e-01
2.90494025e-01 7.88844079e-02 6.14587665e-01 -1.24767959e+00
-3.19334149e-01 9.04245377e-02 7.91408300e-01 -9.21321213e-01
9.18727219e-01 -7.25480914e-01 3.19390237e-01 6.76798448e-02
-3.17146569e-01 -7.83938318e-02 4.98754710e-01 2.96841174e-01
8.75504538e-02 4.45486277e-01 1.16927254e+00 -4.77475911e-01
-4.74405348e-01 6.01907194e-01 2.00804010e-01 -1.94323808e-01
3.70099127e-01 -6.03674948e-01 -1.36512592e-01 -4.12980109e-01
-3.96448612e-01 -4.07718778e-01 1.02147424e+00 9.04429629e-02
9.96446073e-01 -1.34983265e+00 -9.77010965e-01 4.86793935e-01
-1.61159143e-01 7.39990652e-01 6.57256842e-01 2.23265812e-01
-9.34028327e-01 6.52191341e-02 -3.71558189e-01 -8.36333156e-01
-8.40586483e-01 2.24862173e-01 7.32738733e-01 -1.25096485e-01
-1.37699127e+00 9.86517668e-01 3.63950580e-01 -3.02259713e-01
-1.28090993e-01 -4.38910365e-01 2.14405790e-01 -6.86313093e-01
7.19441235e-01 3.22370559e-01 2.30121821e-01 -4.55114365e-01
-1.76875368e-02 9.97398138e-01 1.34056536e-02 -2.91113019e-01
1.67365575e+00 1.63881868e-01 9.90009829e-02 3.37579370e-01
1.18100405e+00 -8.03317726e-02 -1.81661272e+00 -1.43726021e-01
-7.59878576e-01 -8.13318431e-01 5.90247452e-01 -7.14472055e-01
-1.01731920e+00 8.98843884e-01 5.32320380e-01 -7.98582733e-02
9.75716174e-01 -8.30796957e-02 8.77272367e-01 -1.94074675e-01
6.30684078e-01 -5.56677103e-01 -1.62460431e-01 2.11571589e-01
1.38333118e+00 -1.03998399e+00 4.85045105e-01 -7.30068147e-01
-5.37546054e-02 9.43593740e-01 3.74480546e-01 -8.40440929e-01
6.65392160e-01 4.65947896e-01 -1.01612762e-01 -5.81039190e-01
-6.08017027e-01 2.06454888e-01 4.27988380e-01 7.99323201e-01
6.03015661e-01 -2.04696879e-01 6.11762851e-02 5.57912439e-02
-3.85845363e-01 2.87895620e-01 1.94070682e-01 9.54526365e-01
-3.24962914e-01 -6.39967918e-01 -4.62106586e-01 4.45544571e-01
-2.73730874e-01 -3.03246558e-01 -3.20364796e-02 6.48823977e-01
-1.81989357e-01 2.55471975e-01 4.54469353e-01 -1.89572737e-01
5.11929810e-01 -4.14685488e-01 7.01344788e-01 -4.34686065e-01
-6.44055903e-01 1.57557651e-02 8.79703909e-02 -1.08801734e+00
-3.35245967e-01 -2.66251564e-01 -1.03112996e+00 -4.10668224e-01
1.94909468e-01 -3.66264790e-01 6.85352147e-01 4.31810707e-01
9.10379648e-01 1.02446318e+00 5.53759336e-01 -1.96040344e+00
7.30110779e-02 -7.63520002e-01 -2.12844744e-01 3.13577920e-01
5.99910975e-01 -5.33416212e-01 -3.73284459e-01 -2.38918945e-01] | [9.181977272033691, -3.2107176780700684] |
373e651a-4ce9-412b-b87f-2df4d3844118 | provably-learning-nash-policies-in | 2306.07749 | null | https://arxiv.org/abs/2306.07749v1 | https://arxiv.org/pdf/2306.07749v1.pdf | Provably Learning Nash Policies in Constrained Markov Potential Games | Multi-agent reinforcement learning (MARL) addresses sequential decision-making problems with multiple agents, where each agent optimizes its own objective. In many real-world instances, the agents may not only want to optimize their objectives, but also ensure safe behavior. For example, in traffic routing, each car (agent) aims to reach its destination quickly (objective) while avoiding collisions (safety). Constrained Markov Games (CMGs) are a natural formalism for safe MARL problems, though generally intractable. In this work, we introduce and study Constrained Markov Potential Games (CMPGs), an important class of CMGs. We first show that a Nash policy for CMPGs can be found via constrained optimization. One tempting approach is to solve it by Lagrangian-based primal-dual methods. As we show, in contrast to the single-agent setting, however, CMPGs do not satisfy strong duality, rendering such approaches inapplicable and potentially unsafe. To solve the CMPG problem, we propose our algorithm Coordinate-Ascent for CMPGs (CA-CMPG), which provably converges to a Nash policy in tabular, finite-horizon CMPGs. Furthermore, we provide the first sample complexity bounds for learning Nash policies in unknown CMPGs, and, which under additional assumptions, guarantee safe exploration. | ['Andreas Krause', 'Niao He', 'Giorgia Ramponi', 'Pragnya Alatur'] | 2023-06-13 | null | null | null | null | ['multi-agent-reinforcement-learning', 'safe-exploration'] | ['methodology', 'robots'] | [-2.30163574e-01 5.07813811e-01 -6.30566001e-01 3.77242237e-01
-1.17638946e+00 -7.72209466e-01 3.39537114e-01 2.31556684e-01
-5.95127523e-01 1.42274296e+00 -1.18020765e-01 -7.37866938e-01
-4.81070101e-01 -9.70305264e-01 -7.97678351e-01 -9.82800066e-01
-5.14707386e-01 9.57107842e-01 4.00221273e-02 -4.17394012e-01
1.48475822e-03 4.60999370e-01 -9.45260763e-01 -4.02796179e-01
1.04921758e+00 7.53689706e-01 3.03282529e-01 7.63849020e-01
1.04468539e-01 9.94418383e-01 -3.44372183e-01 -4.05286252e-01
4.48644191e-01 -3.02835554e-01 -1.01152384e+00 2.28968874e-01
-3.46656591e-01 -5.07532299e-01 -1.33246437e-01 1.22276807e+00
1.90138876e-01 4.81968999e-01 4.45701480e-01 -1.94647241e+00
-6.30541965e-02 8.44386518e-01 -8.97933304e-01 -2.34197557e-01
1.18803419e-01 3.63337666e-01 1.35866761e+00 -6.49792105e-02
5.37418187e-01 1.43119848e+00 5.77414297e-02 7.90786445e-01
-1.23201919e+00 -2.49482229e-01 6.22502983e-01 6.77764118e-02
-8.20575476e-01 -2.48571988e-02 4.37757462e-01 8.39238148e-03
5.75678587e-01 4.26660478e-01 7.70500481e-01 6.81488454e-01
2.42794529e-01 1.22564769e+00 1.16406214e+00 -3.79248887e-01
6.87403440e-01 -2.37571634e-02 -2.44889826e-01 6.44988894e-01
3.81934702e-01 4.51107711e-01 -1.55968681e-01 -3.25957179e-01
4.98966962e-01 -2.08732665e-01 -2.06839927e-02 -6.88350499e-01
-1.18130970e+00 1.29051685e+00 1.21586733e-01 -2.85180867e-01
-5.72261274e-01 4.49350744e-01 2.20284238e-01 4.92893845e-01
-6.25681423e-04 5.98344147e-01 -2.31438741e-01 -3.12407255e-01
-3.93283516e-01 8.80228043e-01 1.12056887e+00 9.31847274e-01
7.01037943e-01 2.56827176e-01 -8.40989575e-02 3.34721029e-01
2.00478539e-01 6.25652850e-01 -2.10499480e-01 -1.51605892e+00
7.06182599e-01 -1.16884515e-01 7.91031361e-01 -6.49722099e-01
-4.72618252e-01 -1.33943006e-01 -6.13093555e-01 7.62884617e-01
6.46076024e-01 -6.69739783e-01 -4.37345803e-01 1.92900431e+00
3.77856582e-01 -1.51914269e-01 4.16418523e-01 8.99608731e-01
-5.97861819e-02 9.43304777e-01 -1.87898260e-02 -8.73455048e-01
9.58017826e-01 -9.62053418e-01 -4.88258660e-01 -2.23648131e-01
6.66974545e-01 -1.67625934e-01 6.50027215e-01 6.09002650e-01
-1.46846437e+00 4.22257692e-01 -7.19184399e-01 6.39316738e-01
1.07426578e-02 -6.21021152e-01 7.44970858e-01 6.01530552e-01
-1.04206359e+00 3.76563758e-01 -8.57146740e-01 -5.75242601e-02
2.46895537e-01 5.80487549e-01 8.02246109e-02 1.11474238e-01
-1.13386285e+00 9.78923321e-01 5.99181354e-01 -9.08933729e-02
-1.52604234e+00 -3.38001341e-01 -7.50064671e-01 1.35741994e-01
1.47673166e+00 -3.14342707e-01 1.73400319e+00 -8.22961867e-01
-1.63262963e+00 3.26293379e-01 6.10189550e-02 -6.91165626e-01
8.22372019e-01 2.62197435e-01 6.85014725e-02 1.35666251e-01
2.96868235e-01 6.04814172e-01 5.55704415e-01 -1.56874716e+00
-1.12455297e+00 7.34314835e-03 1.02241611e+00 6.01046681e-01
-3.79095040e-02 -6.61822781e-02 -5.16163558e-02 -1.51132315e-01
-4.88319039e-01 -1.15901542e+00 -1.14260948e+00 -2.97108501e-01
-5.29756188e-01 -2.37639993e-01 4.30965632e-01 -1.90414011e-01
9.44640875e-01 -1.53016198e+00 4.29762959e-01 4.31197405e-01
1.59758985e-01 -5.70868067e-02 -1.96996033e-01 5.60231090e-01
4.58766371e-01 1.46244749e-01 -2.63833761e-01 -2.56927907e-01
4.45247620e-01 7.32457221e-01 -2.46904403e-01 5.04837632e-01
-1.99870124e-01 9.86054242e-01 -1.29584432e+00 -4.65086460e-01
4.42069955e-02 -5.09246647e-01 -7.33916938e-01 -2.87905540e-02
-7.81522036e-01 2.51116216e-01 -8.26355219e-01 5.78100681e-01
6.56062603e-01 5.82493991e-02 5.32109737e-01 6.34434998e-01
-3.02066177e-01 -2.42145002e-01 -1.45336366e+00 1.08379281e+00
-5.91991007e-01 2.75020421e-01 6.77195728e-01 -1.25817430e+00
2.52085358e-01 9.21326205e-02 9.07224000e-01 -7.21935272e-01
1.50189176e-02 2.80687153e-01 -1.51161745e-01 -1.75549671e-01
7.23543704e-01 -2.89341867e-01 -4.55047041e-01 7.35281467e-01
-4.31223631e-01 -1.82086393e-01 5.45899868e-01 3.78702313e-01
9.00622308e-01 -1.26078933e-01 4.25885409e-01 -3.74351978e-01
4.32930857e-01 3.15030187e-01 9.01585340e-01 1.13231885e+00
-2.75534809e-01 -1.18007287e-01 1.24472475e+00 -1.55218706e-01
-8.03520977e-01 -1.06907725e+00 5.01760244e-01 8.98388803e-01
4.16511685e-01 -1.19487070e-01 -7.01612651e-01 -7.84920156e-01
1.69534281e-01 8.73036325e-01 -3.43458265e-01 1.71989217e-01
-6.56165004e-01 -8.23813081e-01 8.94097984e-02 4.92186695e-02
4.52601880e-01 -8.54396820e-01 -8.14824283e-01 6.25518143e-01
-1.10499412e-01 -1.06171799e+00 -4.64594543e-01 1.13463648e-01
-4.67528701e-01 -1.17943668e+00 -7.33394325e-01 -4.27657127e-01
6.01044476e-01 3.05478245e-01 9.45607781e-01 -9.89515930e-02
2.11277500e-01 5.77165723e-01 -2.11293086e-01 -5.06507158e-01
-6.14846349e-01 2.05361634e-01 1.00390106e-01 -8.73110518e-02
-1.68019712e-01 -2.98544824e-01 -4.30561870e-01 2.28206486e-01
-7.24599421e-01 1.12466499e-01 3.79364073e-01 8.02846611e-01
7.15977311e-01 4.94008392e-01 5.38406670e-01 -6.58289135e-01
8.81065547e-01 -5.58125734e-01 -1.41546214e+00 4.08600450e-01
-5.23437083e-01 1.43343896e-01 8.99007320e-01 -2.70338565e-01
-8.73805046e-01 3.84283252e-02 4.69039865e-02 -2.58184195e-01
2.32071817e-01 5.88674843e-01 -2.65408337e-01 -1.85766950e-01
2.17964485e-01 2.89365947e-01 2.85220295e-01 3.31910215e-02
2.95821875e-01 2.38946617e-01 1.91361830e-01 -1.25289202e+00
7.92174995e-01 3.58866841e-01 5.61441422e-01 -6.27017498e-01
-6.35435700e-01 5.83428331e-02 -2.59213764e-02 -4.54730332e-01
5.57519972e-01 -6.48961902e-01 -1.70490682e+00 1.85619935e-01
-9.60820079e-01 -8.12677801e-01 -3.65393668e-01 2.42566079e-01
-1.21742213e+00 3.11959714e-01 -4.22844797e-01 -1.50757194e+00
1.19396970e-01 -1.28512526e+00 4.94548738e-01 3.23411316e-01
3.51821393e-01 -1.10250461e+00 -8.92549381e-03 1.31680965e-01
2.60023147e-01 5.52423894e-01 7.74095476e-01 -2.52922654e-01
-9.24851179e-01 2.47128353e-01 2.21681014e-01 -1.92138061e-01
-2.30135858e-01 -1.39209345e-01 -2.11205244e-01 -7.22768843e-01
-2.10868746e-01 -5.38980544e-01 3.94381702e-01 6.23101413e-01
1.03925288e+00 -9.23274159e-01 -3.68486226e-01 1.95969135e-01
1.64386940e+00 5.74356019e-01 3.27458113e-01 7.91140378e-01
3.01367909e-01 5.96930504e-01 1.12138677e+00 8.53902996e-01
7.14657009e-01 7.08684444e-01 9.76838589e-01 1.08953811e-01
8.04181933e-01 -1.29874930e-01 5.34025013e-01 -3.84179540e-02
-2.88392484e-01 -4.88148868e-01 -8.43059659e-01 4.80698138e-01
-2.34927464e+00 -1.09050834e+00 7.30414391e-02 2.25053406e+00
7.57950068e-01 3.13249320e-01 6.57665670e-01 -2.20175922e-01
4.91999120e-01 -1.06053855e-02 -9.65356529e-01 -8.48903477e-01
-1.38867795e-01 -1.64070383e-01 1.13834822e+00 9.77193713e-01
-8.74199212e-01 9.53633249e-01 6.08387709e+00 8.95558238e-01
-6.04032636e-01 2.15016529e-01 8.55691671e-01 -3.90182436e-01
-4.33395594e-01 9.57650840e-02 -7.31832981e-01 4.61963326e-01
9.26919699e-01 -7.51524687e-01 1.00563073e+00 9.64500189e-01
6.67020142e-01 -4.86307591e-01 -8.81099343e-01 5.45185983e-01
-7.00122833e-01 -1.44478869e+00 -2.94318438e-01 4.89190727e-01
1.04950953e+00 -2.76269466e-01 -8.38212296e-03 4.95371968e-01
1.30596125e+00 -8.65141392e-01 8.54542196e-01 -4.17002887e-02
3.85486782e-01 -1.64964044e+00 3.31000537e-01 5.56323647e-01
-1.13808274e+00 -6.23784721e-01 -2.15448439e-01 -7.61103928e-02
5.21687925e-01 1.74536854e-01 -3.19215387e-01 6.32638276e-01
2.77424961e-01 1.40433356e-01 3.12967926e-01 9.64742124e-01
-1.78608119e-01 2.29110762e-01 -3.40043455e-01 -3.14693660e-01
1.02796865e+00 -5.52575469e-01 7.69247174e-01 7.11090147e-01
1.00843564e-01 2.79131800e-01 8.69269013e-01 7.94359624e-01
1.82870612e-01 -1.03898317e-01 -4.76740062e-01 -1.80514500e-01
4.16573226e-01 1.16393197e+00 -7.71227658e-01 -5.33571728e-02
-3.29874426e-01 6.83836520e-01 3.12680781e-01 5.46382070e-01
-1.01796782e+00 -1.62817359e-01 1.18376350e+00 -2.98885435e-01
1.65843479e-02 -3.41652364e-01 -8.92920122e-02 -9.59983885e-01
-6.51358441e-02 -1.06245065e+00 5.71456432e-01 -1.77122474e-01
-9.52256978e-01 1.45871997e-01 8.38461891e-02 -1.09720278e+00
-7.55212069e-01 -4.47001517e-01 -6.48352385e-01 5.78459501e-01
-1.85423279e+00 -8.50853145e-01 3.56118441e-01 6.40656233e-01
6.22168183e-01 -1.03350341e-01 3.70178103e-01 -1.74136192e-01
-6.31829917e-01 3.01809818e-01 4.66685355e-01 -2.72960067e-01
-5.04056513e-02 -1.63381886e+00 2.97867786e-02 8.53990972e-01
-4.45224077e-01 3.95669602e-02 8.86499703e-01 -5.19802690e-01
-1.95525885e+00 -1.01895869e+00 2.22606719e-01 2.08880976e-01
1.00402606e+00 -1.29857495e-01 -2.95782685e-01 6.52652264e-01
2.16933921e-01 -2.45666832e-01 -1.05419522e-02 -9.92628038e-02
4.60707903e-01 -1.02885723e-01 -1.07931507e+00 1.10941923e+00
8.20107639e-01 1.21518724e-01 1.37056699e-02 6.58268690e-01
6.75178349e-01 -7.20819652e-01 -5.00261605e-01 -1.18209243e-01
1.13869220e-01 -6.49913669e-01 8.30894351e-01 -8.02871764e-01
2.62631893e-01 -1.82369798e-01 -6.52910173e-02 -1.72216415e+00
-1.69421643e-01 -1.33276081e+00 -3.57392311e-01 7.68970311e-01
4.31618035e-01 -6.68179870e-01 1.09509385e+00 7.66963422e-01
-7.46530369e-02 -7.96052277e-01 -1.19512987e+00 -1.36383188e+00
6.53785706e-01 -4.38168824e-01 5.26381850e-01 5.23115814e-01
2.82806486e-01 -2.25186005e-01 -6.76890671e-01 3.12342882e-01
1.15219522e+00 2.23246112e-01 5.94266951e-01 -6.86009645e-01
-5.65397263e-01 -6.97567105e-01 2.70704627e-01 -9.40363109e-01
5.37997723e-01 -6.49717510e-01 2.97652096e-01 -1.61690927e+00
1.27608553e-01 -8.64389896e-01 -3.97880282e-03 4.89577532e-01
1.90639973e-01 -3.97283047e-01 6.79336965e-01 -1.41637802e-01
-1.05139935e+00 5.17347991e-01 1.47269726e+00 -3.44259918e-01
-3.64881933e-01 4.05010641e-01 -7.71653295e-01 4.16000575e-01
9.30455446e-01 -2.76145279e-01 -5.91171563e-01 -1.53873637e-01
3.33452821e-01 9.01273847e-01 1.61877483e-01 -5.32536983e-01
2.71799177e-01 -1.37498140e+00 -5.36611795e-01 -4.77204263e-01
2.41366193e-01 -7.29797721e-01 9.69176814e-02 1.04255617e+00
-3.23032141e-01 7.80335739e-02 -4.28249501e-02 6.18612766e-01
1.36943012e-02 -5.88133335e-01 7.99378872e-01 -4.41878885e-01
-7.81961918e-01 6.65287197e-01 -9.88049328e-01 3.43889207e-01
1.54577792e+00 1.61369056e-01 -1.66479200e-01 -8.64827633e-01
-7.24562407e-01 1.23308456e+00 2.05032080e-01 -4.69490103e-02
5.23160338e-01 -1.17334163e+00 -6.30292654e-01 -4.74971265e-01
-3.30306798e-01 1.91401504e-02 1.36252314e-01 6.46746278e-01
-2.67849356e-01 4.92317945e-01 -1.47453338e-01 -5.34703843e-02
-9.10515428e-01 8.07866275e-01 6.28402829e-01 -6.90080702e-01
-2.48235911e-01 3.99863303e-01 1.00743547e-01 -1.93151638e-01
2.76823848e-01 1.16201781e-01 -4.97130603e-02 1.15583174e-01
4.60675538e-01 6.04080439e-01 -3.92392188e-01 -1.45126328e-01
-7.03785270e-02 3.79794389e-02 -1.46935076e-01 -6.67240202e-01
1.36707675e+00 -3.07955384e-01 3.35488655e-02 -1.31150007e-01
7.06515014e-01 -3.09876263e-01 -1.49839330e+00 -4.11397889e-02
6.45450279e-02 -3.90714496e-01 -1.52286962e-02 -5.70632756e-01
-1.18775070e+00 3.32235605e-01 -1.66487060e-02 5.05994678e-01
9.15233910e-01 -2.75556654e-01 7.14489996e-01 6.17982984e-01
9.79735672e-01 -1.40422857e+00 -1.80990864e-02 6.06763899e-01
6.58004701e-01 -1.25065136e+00 -2.51788169e-01 -1.45797923e-01
-9.65668499e-01 1.12454545e+00 7.32793808e-01 1.06580496e-01
2.39007652e-01 4.01599824e-01 -3.63392234e-01 1.37519181e-01
-1.04781282e+00 -6.35061800e-01 -5.85948944e-01 6.87571108e-01
-5.19461453e-01 3.60477954e-01 -4.85733747e-01 2.72751510e-01
-5.17075919e-02 -1.15208365e-01 1.02364850e+00 1.14878726e+00
-6.50947571e-01 -1.28188384e+00 -5.02053022e-01 3.96037817e-01
-3.79291981e-01 2.68618613e-01 -5.02603129e-02 8.39800715e-01
-3.69403869e-01 1.17937601e+00 -5.68966791e-02 1.37389556e-01
8.65735412e-02 -6.99387312e-01 4.30810601e-01 -2.49084279e-01
-1.51317418e-01 3.54590826e-02 3.24792325e-01 -5.88192582e-01
-2.36275822e-01 -8.49334061e-01 -1.32392430e+00 -7.82968581e-01
2.45509092e-02 4.95757729e-01 3.54156911e-01 1.00381708e+00
-2.25935340e-01 2.81857371e-01 1.05878305e+00 -6.04045987e-01
-9.80129421e-01 -3.25159468e-02 -7.77667224e-01 -2.16122255e-01
4.36972946e-01 -6.34064138e-01 -2.80272841e-01 -5.95732570e-01] | [4.217959880828857, 2.5429840087890625] |
66532f5b-b961-4656-9eee-22c0369626a7 | evaluating-neural-morphological-taggers-for | 2005.10893 | null | https://arxiv.org/abs/2005.10893v1 | https://arxiv.org/pdf/2005.10893v1.pdf | Evaluating Neural Morphological Taggers for Sanskrit | Neural sequence labelling approaches have achieved state of the art results in morphological tagging. We evaluate the efficacy of four standard sequence labelling models on Sanskrit, a morphologically rich, fusional Indian language. As its label space can theoretically contain more than 40,000 labels, systems that explicitly model the internal structure of a label are more suited for the task, because of their ability to generalise to labels not seen during training. We find that although some neural models perform better than others, one of the common causes for error for all of these models is mispredictions due to syncretism. | ['Amrith Krishna', 'Ashim Gupta', 'Oliver Hellwig', 'Pawan Goyal'] | 2020-05-21 | evaluating-neural-morphological-taggers-for-1 | https://aclanthology.org/2020.sigmorphon-1.23 | https://aclanthology.org/2020.sigmorphon-1.23.pdf | ws-2020-7 | ['morphological-tagging'] | ['natural-language-processing'] | [ 4.11014199e-01 2.08055556e-01 -1.44860089e-01 -5.43822050e-01
-4.50624287e-01 -1.05626166e+00 6.58228874e-01 4.11970496e-01
-6.94775403e-01 8.74740362e-01 3.93947542e-01 -7.07440317e-01
1.60671279e-01 -4.43134248e-01 -2.58386672e-01 -5.10231435e-01
7.07755387e-02 7.94478714e-01 3.86118203e-01 -3.10629666e-01
3.18189621e-01 4.77694958e-01 -1.10797071e+00 2.16686219e-01
5.24817288e-01 3.24066967e-01 2.47558758e-01 5.25254071e-01
-3.36572319e-01 9.49566245e-01 -7.13634849e-01 -7.63508558e-01
-5.09504564e-02 -4.72537816e-01 -1.37951481e+00 -1.13821968e-01
4.22718495e-01 2.03474686e-01 -4.08708081e-02 1.29709840e+00
4.46519107e-01 8.92491564e-02 8.35152328e-01 -6.61610603e-01
-7.55164862e-01 7.99436152e-01 -3.07789832e-01 4.07389820e-01
2.52694100e-01 -1.34315342e-01 1.09155202e+00 -5.66057324e-01
9.25767541e-01 1.42415285e+00 9.26517487e-01 7.71641433e-01
-1.28702545e+00 -5.45788944e-01 4.37908247e-02 -7.35935494e-02
-1.31396532e+00 -6.85076594e-01 3.91629785e-01 -4.73877341e-01
1.14359260e+00 1.63505167e-01 4.85268682e-01 6.14989996e-01
5.03395647e-02 6.01893604e-01 1.52926874e+00 -6.46309257e-01
6.07715845e-02 -6.91596642e-02 2.36818403e-01 6.71282589e-01
-6.09554462e-02 1.60646290e-01 -2.77908951e-01 -3.38661015e-01
6.26067996e-01 -6.63617074e-01 -3.64630669e-02 -6.38783397e-03
-9.34200823e-01 8.61703515e-01 1.39458165e-01 5.75018287e-01
-1.09732412e-01 -1.31324738e-01 7.49522150e-01 2.66067803e-01
5.36916554e-01 7.27769792e-01 -8.39518130e-01 -1.80439919e-01
-8.81139457e-01 -7.48942643e-02 7.05899179e-01 6.28747880e-01
5.24451554e-01 1.05451316e-01 3.34070235e-01 1.22734559e+00
2.30789006e-01 1.83982179e-02 8.42097104e-01 -9.15734231e-01
3.25058959e-02 4.57022190e-01 -2.87545264e-01 -6.18849516e-01
-6.37232959e-01 -2.64536113e-01 -4.20642763e-01 1.57353878e-01
6.52429998e-01 1.47661818e-02 -1.41067243e+00 1.76020479e+00
3.57347950e-02 -1.00869738e-01 1.72623605e-01 4.55835134e-01
7.37134159e-01 3.87109965e-01 7.45769382e-01 -2.75453091e-01
1.41340017e+00 -6.11649215e-01 -5.68765640e-01 -5.26739419e-01
1.04640329e+00 -8.00085723e-01 8.32346618e-01 1.73913643e-01
-9.13812101e-01 -3.44512254e-01 -8.65020990e-01 2.17894688e-02
-5.56826949e-01 -1.83883727e-01 7.27661252e-01 9.54022765e-01
-1.16496050e+00 7.30014801e-01 -7.79909790e-01 -6.71919286e-01
8.25744346e-02 4.27844405e-01 -5.01898944e-01 9.74095762e-02
-1.13589025e+00 1.43402541e+00 1.06252682e+00 4.77640219e-02
-3.36133420e-01 -1.76411241e-01 -9.44831848e-01 -2.72778898e-01
1.32152021e-01 1.60793871e-01 1.49997270e+00 -9.30220306e-01
-1.04972720e+00 1.42994654e+00 6.99335360e-04 -2.65646696e-01
-8.82328376e-02 2.34776244e-01 -6.98623836e-01 7.46423528e-02
4.96084504e-02 8.64997983e-01 1.85372382e-01 -1.18580556e+00
-6.70123458e-01 -1.66777194e-01 -2.31222227e-01 1.38833061e-01
6.35880083e-02 7.88679779e-01 -8.62853788e-03 -8.04748476e-01
1.65629596e-01 -1.03346336e+00 -3.21035385e-01 -8.07038605e-01
-3.57969012e-03 -5.12139022e-01 4.93755907e-01 -1.12174761e+00
1.09531677e+00 -2.03445101e+00 -2.26626068e-01 3.39969546e-02
-1.15081869e-01 6.95664525e-01 -4.78597656e-02 4.94918942e-01
-1.46381080e-01 6.76529765e-01 -3.82328093e-01 1.00635439e-01
-6.52363598e-02 6.83777511e-01 -3.35501917e-02 4.65804845e-01
3.64479214e-01 9.24445570e-01 -9.06603634e-01 -6.31232262e-01
1.22826368e-01 3.16844970e-01 7.62832984e-02 -2.47231141e-01
-1.02684878e-01 4.44045126e-01 -3.07840034e-02 5.63871264e-01
3.67130876e-01 -8.12225565e-02 5.80485880e-01 4.00570333e-01
6.94374666e-02 7.55633056e-01 -8.55849981e-01 1.38777101e+00
-1.13795325e-01 4.54805672e-01 1.00034662e-01 -6.34303510e-01
7.98271954e-01 6.27905428e-01 6.86811432e-02 -4.93129283e-01
7.87067488e-02 5.87373734e-01 4.52213705e-01 -1.86134115e-01
6.09223366e-01 -7.13278890e-01 -1.62726432e-01 4.05294627e-01
1.89018920e-01 7.54692629e-02 2.68565446e-01 6.00016080e-02
1.28038764e+00 3.14824462e-01 7.69304276e-01 -1.82019323e-01
4.39268738e-01 2.21109420e-01 1.01567793e+00 4.91623431e-01
-2.63636559e-01 3.07423383e-01 1.36632740e-01 -5.77733636e-01
-1.29230261e+00 -6.75497591e-01 -2.87319809e-01 1.47397411e+00
-2.97276825e-01 -1.80989772e-01 -6.18969619e-01 -1.06009674e+00
-4.47794944e-01 7.94624984e-01 -4.52519447e-01 1.74171925e-01
-9.19217706e-01 -8.72928858e-01 1.08503914e+00 6.38465524e-01
3.37905698e-02 -1.70288086e+00 -5.81631124e-01 6.05842471e-01
-3.08403652e-02 -9.85658050e-01 -3.29791345e-02 7.27678597e-01
-6.86889827e-01 -7.44231880e-01 -5.36124945e-01 -1.18124139e+00
5.35511374e-01 -3.72682840e-01 1.28864312e+00 6.56934798e-01
1.00776546e-01 -5.30047268e-02 -5.90485275e-01 -2.55688101e-01
-1.18871140e+00 3.52355726e-02 -1.81545943e-01 -7.31867731e-01
5.46469748e-01 -8.07018280e-02 1.21161781e-01 1.82866156e-01
-9.45104957e-01 -2.47480199e-01 4.55421507e-01 8.42492044e-01
3.49325866e-01 1.16079293e-01 6.89672709e-01 -1.29799390e+00
5.78267038e-01 -2.82900780e-01 -1.83732942e-01 2.96365917e-01
-4.91709560e-01 5.97589538e-02 3.95368367e-01 -4.49969113e-01
-1.12831151e+00 3.56784880e-01 -7.08061099e-01 4.43448454e-01
-8.10120523e-01 5.25599599e-01 -1.17716417e-01 -2.56566882e-01
5.54938257e-01 -7.56451208e-03 -3.23557287e-01 -7.32704520e-01
3.33631545e-01 6.30210996e-01 6.18370295e-01 -5.85117102e-01
3.12503189e-01 -3.67096066e-02 -1.12913154e-01 -8.81012797e-01
-6.58801377e-01 -5.00866175e-01 -1.04537511e+00 -5.02692610e-02
8.98980737e-01 -5.99830091e-01 -3.49742979e-01 4.03838962e-01
-9.95464265e-01 -4.04561192e-01 -9.11463276e-02 2.06415832e-01
-4.64933276e-01 6.96146011e-01 -1.17480290e+00 -8.13717246e-01
-5.75878061e-02 -9.70059216e-01 7.53567636e-01 3.02138869e-02
-6.50067687e-01 -1.30919313e+00 2.56366938e-01 -3.23032998e-02
6.36293814e-02 2.34852254e-01 1.50943100e+00 -1.35196030e+00
4.19779271e-01 1.36822918e-02 -5.25884554e-02 2.56440789e-01
5.03722429e-02 -1.15752831e-01 -8.13114524e-01 -3.49751532e-01
5.17531484e-02 -6.08568728e-01 7.38581359e-01 -3.70238684e-02
2.58018523e-01 -1.68298855e-01 -3.54064882e-01 3.14810015e-02
1.39313960e+00 6.42493248e-01 4.54165190e-01 5.68913519e-01
5.53192258e-01 8.77666235e-01 4.48597997e-01 -3.74153644e-01
3.07490408e-01 6.20749056e-01 1.30148664e-01 -3.92187648e-02
-3.29811603e-01 -2.55228728e-01 3.14618051e-01 9.90459085e-01
2.32484400e-01 -3.64361107e-01 -1.40935433e+00 8.17444623e-01
-1.60484445e+00 -6.63894534e-01 -3.89693469e-01 1.89944065e+00
1.11594462e+00 1.82744637e-01 2.30080873e-01 2.25805685e-01
9.25300717e-01 -3.24240215e-02 -1.17739394e-01 -8.77087295e-01
-3.39050144e-01 5.04055500e-01 5.79618990e-01 3.59662801e-01
-1.00011528e+00 1.30006957e+00 7.85234404e+00 8.42205048e-01
-7.60627210e-01 2.17800677e-01 6.08586311e-01 5.74271977e-01
-2.19426248e-02 1.52977988e-01 -8.47381830e-01 5.20640135e-01
1.31409466e+00 5.62234342e-01 2.60691494e-01 2.74673074e-01
-2.37880960e-01 -1.17560230e-01 -8.52020860e-01 3.38297725e-01
1.16479948e-01 -9.18099344e-01 -1.25202134e-01 2.65841961e-01
5.70975363e-01 1.70840770e-01 -3.75844330e-01 2.65479863e-01
9.62725401e-01 -1.26997530e+00 9.99479175e-01 -6.69453666e-02
7.97774911e-01 -8.18962336e-01 9.97229159e-01 6.07646167e-01
-8.60613644e-01 1.05574079e-01 -5.36688626e-01 -2.93140620e-01
5.46144187e-01 1.57019168e-01 -1.00303495e+00 1.86196014e-01
2.45631665e-01 2.38513380e-01 -9.95373011e-01 1.17567825e+00
-6.73708200e-01 1.05318856e+00 -2.38147259e-01 3.35407481e-02
5.38094163e-01 9.06036347e-02 2.56342322e-01 1.57949567e+00
-2.06005462e-02 2.41322458e-01 5.02194762e-01 -1.55888349e-01
2.53167808e-01 3.64334732e-01 -4.77947891e-01 -2.74739087e-01
6.12889826e-01 9.97534096e-01 -1.41134310e+00 -4.53634441e-01
-4.31540459e-01 9.19905603e-01 7.16047943e-01 6.29917532e-02
-3.35629255e-01 -1.38291687e-01 3.04761291e-01 -1.94804639e-01
3.61587465e-01 -2.24157840e-01 -3.77590805e-01 -6.30326211e-01
-4.44823474e-01 -9.92072821e-01 6.22098148e-01 -6.76887214e-01
-1.30074239e+00 6.68356538e-01 -2.76542693e-01 -3.59181076e-01
-4.08261418e-01 -8.21514130e-01 -4.40492257e-02 9.28925335e-01
-1.03960133e+00 -1.47029364e+00 8.71087790e-01 6.28596246e-02
4.72905159e-01 -4.90103811e-02 1.20951807e+00 1.41575903e-01
-1.46106750e-01 2.10378021e-01 -6.78153336e-02 4.99956012e-01
5.44733047e-01 -1.63002992e+00 9.58463788e-01 8.97210538e-01
6.84552372e-01 4.65005189e-01 6.50644004e-01 -1.02082622e+00
-4.10073280e-01 -7.03461468e-01 1.36367846e+00 -6.09602332e-01
7.23758519e-01 -2.45920405e-01 -1.28693426e+00 8.46087396e-01
2.29771525e-01 -2.86847562e-01 9.05469537e-01 2.79892027e-01
-4.84117359e-01 8.99551213e-01 -1.25421453e+00 3.65754932e-01
1.09938323e+00 -5.97863436e-01 -1.07371020e+00 1.02282904e-01
4.55817372e-01 -2.45179206e-01 -1.03184128e+00 4.57697541e-01
3.18568736e-01 -5.64371049e-01 4.54959959e-01 -8.53246152e-01
7.25446567e-02 -4.10693616e-01 4.19904292e-03 -1.38847601e+00
-8.13383758e-01 -4.65259433e-01 2.83927739e-01 1.67731857e+00
6.76588416e-01 -3.75773340e-01 8.90434504e-01 4.35775250e-01
-2.23109633e-01 -3.07459325e-01 -9.79984164e-01 -9.23077166e-01
3.47528577e-01 -3.90985608e-01 5.84311664e-01 1.32839310e+00
1.74176604e-01 8.61789465e-01 -4.03937519e-01 -1.42057493e-01
2.16022104e-01 -4.72299188e-01 -1.11270361e-01 -1.40398109e+00
-3.05806100e-01 -4.80247021e-01 -6.59054458e-01 -5.66530228e-01
5.55278480e-01 -1.15380347e+00 3.65386307e-01 -1.53959775e+00
8.78833234e-02 -6.88556075e-01 -2.71829814e-01 9.90302265e-01
-1.36191308e-01 7.69804001e-01 3.37332904e-01 2.24537387e-01
-3.95805806e-01 -3.27260971e-01 9.35232043e-01 2.22355559e-01
2.98670623e-02 -4.09887820e-01 -6.95531249e-01 1.08612943e+00
9.54577148e-01 -8.69258881e-01 3.57849672e-02 -5.06848633e-01
5.28681457e-01 -9.94156674e-02 -2.23981664e-01 -7.19310403e-01
-1.35125458e-01 -1.77360103e-01 4.07513291e-01 -4.19689655e-01
1.77259222e-01 -6.77445710e-01 4.53878671e-01 5.00921845e-01
-5.09688795e-01 2.77954102e-01 2.60186434e-01 3.49485487e-01
-7.43768662e-02 -8.98715198e-01 9.85336602e-01 -5.64718127e-01
-1.04295397e+00 -2.27021933e-01 -8.27960730e-01 2.50508189e-01
6.97190702e-01 -4.30274278e-01 -2.66664803e-01 1.06303118e-01
-9.04659629e-01 -1.88339844e-01 9.25261796e-01 2.91052699e-01
9.67665538e-02 -1.05743504e+00 -5.64080656e-01 -1.97647847e-02
3.17240134e-02 -4.54770476e-01 -4.19102013e-01 3.08360487e-01
-5.45079291e-01 3.74311954e-01 -2.66393751e-01 -8.00194591e-02
-1.38191986e+00 7.28294730e-01 6.75094128e-02 -3.74298006e-01
-4.57320243e-01 1.00359881e+00 2.83253985e-03 -7.96005547e-01
-3.70656885e-02 1.36437744e-01 -6.21640503e-01 2.81272619e-03
3.55783403e-01 3.03190917e-01 2.89075524e-01 -1.52588665e+00
-5.00325918e-01 2.14825317e-01 -3.91676277e-01 -2.88066357e-01
1.15146577e+00 -1.20665990e-02 -2.89450109e-01 4.68915850e-01
7.11274862e-01 3.35136056e-02 -8.86096239e-01 -2.68984437e-01
7.38871753e-01 -2.19270065e-02 -1.84503421e-01 -1.30138278e+00
-5.01575232e-01 6.77256703e-01 3.66814524e-01 2.94501245e-01
8.10879529e-01 7.44555816e-02 7.06452072e-01 1.47005901e-01
3.59704137e-01 -1.09021950e+00 -6.63737714e-01 1.03383005e+00
1.44776732e-01 -7.26073146e-01 -9.84349921e-02 -4.59583163e-01
-7.05049157e-01 8.77629101e-01 2.42261752e-01 1.32199312e-02
4.13653076e-01 3.73736441e-01 4.65831131e-01 -3.55859697e-01
-5.87116897e-01 -3.62465680e-01 -4.10787500e-02 9.21440780e-01
9.21758711e-01 2.91747987e-01 -9.12027597e-01 5.00215068e-02
-4.50240314e-01 -6.05362236e-01 5.03090262e-01 1.08802712e+00
-5.82526505e-01 -1.49894977e+00 -3.05265784e-01 5.03201127e-01
-1.25375485e+00 -2.63865530e-01 -8.60023797e-01 6.88559234e-01
4.15227443e-01 1.01866162e+00 -4.49803062e-02 -1.27973467e-01
-1.79235172e-02 6.48593783e-01 5.29446721e-01 -1.19791865e+00
-1.20543361e+00 1.51596412e-01 6.96021020e-01 1.55071542e-01
-6.61578953e-01 -9.42922354e-01 -1.61348212e+00 1.14135919e-02
-4.76980209e-01 4.15562898e-01 6.17635727e-01 1.22013998e+00
-2.94567138e-01 9.27072689e-02 -1.21809207e-01 -4.42079246e-01
-3.45166266e-01 -1.09514260e+00 -8.10197055e-01 5.61557353e-01
-1.59938514e-01 -5.54577112e-01 9.68748182e-02 2.37549514e-01] | [10.327393531799316, 10.023092269897461] |
e900110f-854d-4525-a324-0621fd63e9ee | density-based-feasibility-learning-with | 2307.01317 | null | https://arxiv.org/abs/2307.01317v2 | https://arxiv.org/pdf/2307.01317v2.pdf | Density-based Feasibility Learning with Normalizing Flows for Introspective Robotic Assembly | Machine Learning (ML) models in Robotic Assembly Sequence Planning (RASP) need to be introspective on the predicted solutions, i.e. whether they are feasible or not, to circumvent potential efficiency degradation. Previous works need both feasible and infeasible examples during training. However, the infeasible ones are hard to collect sufficiently when re-training is required for swift adaptation to new product variants. In this work, we propose a density-based feasibility learning method that requires only feasible examples. Concretely, we formulate the feasibility learning problem as Out-of-Distribution (OOD) detection with Normalizing Flows (NF), which are powerful generative models for estimating complex probability distributions. Empirically, the proposed method is demonstrated on robotic assembly use cases and outperforms other single-class baselines in detecting infeasible assemblies. We further investigate the internal working mechanism of our method and show that a large memory saving can be obtained based on an advanced variant of NF. | ['Rudolph Triebel', 'Stephan Günnemann', 'Maximilian Durner', 'Ismael Rodríguez', 'Matan Atad', 'Jianxiang Feng'] | 2023-07-03 | null | null | null | null | ['ood-detection'] | ['computer-vision'] | [ 1.82619944e-01 1.08067408e-01 -6.62699997e-01 -2.49011174e-01
-6.74786806e-01 -9.10969973e-01 2.37792164e-01 7.17309117e-02
1.10874966e-01 6.57531261e-01 6.01291656e-03 -6.44460320e-01
-1.29692793e-01 -7.25438774e-01 -1.05473912e+00 -4.28233206e-01
-1.15369096e-01 6.59212410e-01 -7.64162764e-02 1.05085284e-01
3.21392745e-01 4.71694738e-01 -1.45096898e+00 4.25589025e-01
1.11284125e+00 9.08804774e-01 6.39815509e-01 7.17283666e-01
-2.50463467e-02 6.81332231e-01 -8.30587029e-01 -1.70514375e-01
2.52390772e-01 -2.33403057e-01 -6.24580920e-01 1.83796346e-01
1.32458471e-02 -4.44453061e-01 2.85355877e-02 8.80121827e-01
2.65004247e-01 1.30894139e-01 1.00070786e+00 -1.52101684e+00
-5.55463791e-01 6.94678843e-01 -4.81488645e-01 8.88737962e-02
4.24118549e-01 3.86305332e-01 9.16599631e-01 -7.87405312e-01
4.63865429e-01 1.24420691e+00 4.05939430e-01 4.38786656e-01
-1.03176820e+00 -5.93769789e-01 5.20207107e-01 -7.83655494e-02
-1.07638168e+00 -3.48430008e-01 6.19714200e-01 -6.24073327e-01
1.23995256e+00 9.16702300e-02 4.50123638e-01 1.16057682e+00
2.08798438e-01 1.05040002e+00 7.20230818e-01 -3.17714661e-01
6.25946224e-01 -1.88737586e-02 -4.21992354e-02 6.73309147e-01
5.12314975e-01 2.16353774e-01 -3.56253564e-01 -4.84013893e-02
8.27050090e-01 -2.18409032e-01 -1.79605290e-01 -7.46846437e-01
-9.69078362e-01 8.73961031e-01 1.58522874e-01 -1.84987277e-01
-2.88949668e-01 -2.19925065e-02 3.69045585e-01 7.69286752e-02
1.93170875e-01 7.56738007e-01 -7.47615457e-01 -4.75369036e-01
-5.95686913e-01 2.13711888e-01 9.37316239e-01 1.64675772e+00
5.88036001e-01 6.14202060e-02 -1.71586275e-01 7.07506776e-01
3.39145988e-01 2.66295314e-01 1.91706046e-01 -7.82041192e-01
7.50595808e-01 4.69833016e-01 4.80236113e-01 -6.15182102e-01
-4.19027925e-01 -3.99555981e-01 -3.17417413e-01 1.31241590e-01
4.53346848e-01 -4.15655285e-01 -9.58259404e-01 1.50049675e+00
3.19467276e-01 1.55316323e-01 3.20782185e-01 7.53015637e-01
3.52912456e-01 8.90667796e-01 -3.27996425e-02 -2.87637115e-01
1.01391172e+00 -1.04004276e+00 -5.12135983e-01 -4.41693097e-01
9.66372371e-01 -7.95137346e-01 1.11353195e+00 6.52914822e-01
-9.49625254e-01 -5.56667924e-01 -1.22089398e+00 3.56283128e-01
-6.49608895e-02 4.62722301e-01 1.12439239e+00 1.01260006e+00
-5.84178448e-01 6.04412556e-01 -9.27312255e-01 -8.54513422e-02
6.93338931e-01 3.95187587e-01 2.46897293e-03 -5.01541317e-01
-3.36662054e-01 8.50761831e-01 6.95140481e-01 2.53945291e-01
-1.39170849e+00 -6.53648674e-01 -1.18800437e+00 1.46556780e-01
7.29045510e-01 -4.42650676e-01 1.34259093e+00 -5.92351198e-01
-1.65738475e+00 -2.16042232e-02 -4.18865196e-02 -2.05093026e-01
3.06727320e-01 -5.58385491e-01 -2.04223633e-01 -3.04344624e-01
-1.90555274e-01 7.82363713e-01 9.67231333e-01 -1.42371011e+00
-7.51161516e-01 1.59817964e-01 4.82005000e-01 3.30136865e-02
-1.50577039e-01 -3.02182585e-01 -1.94104299e-01 -4.37092274e-01
-1.39576390e-01 -1.00887740e+00 -1.88576043e-01 -3.24920088e-01
-5.51892161e-01 -3.70435417e-01 3.95369977e-01 -6.37320936e-01
1.15460491e+00 -2.07700682e+00 1.71268117e-02 1.83307126e-01
-2.98084676e-01 1.91531628e-01 -5.68787396e-01 3.70658904e-01
1.11895829e-01 -9.36276913e-02 -2.91167758e-02 1.00283317e-01
3.45817059e-01 4.33594316e-01 -2.99463630e-01 2.70356864e-01
7.60940015e-01 7.60864198e-01 -1.19366050e+00 -1.40364468e-01
3.30212504e-01 -1.11673795e-01 -8.65671694e-01 3.10591638e-01
-6.50444567e-01 3.96268606e-01 -2.39823386e-01 8.49766135e-01
8.18937123e-01 -1.50816232e-01 6.13719225e-01 -1.01980470e-01
8.45126808e-02 4.03482139e-01 -9.62827265e-01 1.67300916e+00
-8.07059705e-01 1.40789777e-01 -3.21006894e-01 -1.01694477e+00
1.05601442e+00 -1.49513349e-01 6.97321966e-02 -3.08129221e-01
2.99984105e-02 2.26784557e-01 3.81062776e-01 -4.85252589e-01
4.55685109e-01 1.21698529e-01 -2.50981122e-01 2.33941957e-01
7.39898309e-02 -2.37939045e-01 5.38885415e-01 -2.00922087e-01
1.10730553e+00 6.83722198e-01 4.00984466e-01 -1.34846628e-01
1.23950675e-01 9.48704109e-02 8.16810787e-01 7.53965855e-01
3.70117463e-02 2.87613690e-01 8.38465512e-01 -3.70890379e-01
-9.47174966e-01 -1.10927999e+00 3.91872197e-01 9.58605170e-01
1.97677746e-01 -3.51796001e-01 -5.17788768e-01 -1.19265127e+00
-3.33028566e-03 1.07800436e+00 -2.24055704e-02 -2.63737053e-01
-7.27073014e-01 -6.70538306e-01 2.71563288e-02 9.47181404e-01
-1.97114110e-01 -9.89503324e-01 -7.37256765e-01 2.32898712e-01
-6.26395270e-02 -1.16930377e+00 -4.52798665e-01 3.16621095e-01
-8.71730924e-01 -1.03180385e+00 -4.78021264e-01 -9.28940237e-01
9.19668317e-01 4.25624438e-02 1.10132301e+00 -1.52743012e-01
-1.98553875e-01 1.61975190e-01 -4.03135836e-01 -2.99165726e-01
-8.10599387e-01 1.22603372e-01 5.49440682e-02 -6.48402572e-01
8.39841738e-02 -3.85608166e-01 -3.08133036e-01 5.57321429e-01
-4.27028030e-01 7.98606500e-02 9.62650657e-01 9.53865349e-01
4.21624869e-01 3.02479476e-01 8.67010236e-01 -6.41433656e-01
6.63289726e-01 -4.51926589e-01 -9.52837229e-01 4.92236435e-01
-3.50102484e-01 2.14245617e-01 6.79774761e-01 -6.38751149e-01
-1.21773052e+00 4.28443581e-01 -1.37500301e-01 -4.19635236e-01
-2.91331887e-01 5.08382201e-01 -2.49339178e-01 2.95170754e-01
3.57215911e-01 2.93037156e-03 -2.45392889e-01 -2.08165392e-01
4.87189651e-01 4.94334489e-01 2.69644827e-01 -1.02083731e+00
5.35862088e-01 -2.03005761e-01 -1.78460926e-01 -4.89996225e-01
-7.60985196e-01 -3.79878193e-01 -5.22250950e-01 -5.24442717e-02
2.84561723e-01 -8.52856219e-01 -9.03953373e-01 -1.36966957e-02
-1.21778107e+00 -6.52277529e-01 -1.99854702e-01 6.53235137e-01
-7.64768183e-01 3.57860059e-01 -4.77465242e-01 -1.28087199e+00
1.02177285e-01 -1.32922935e+00 9.94446933e-01 1.37985557e-01
-3.72397006e-01 -6.17990196e-01 -2.63916790e-01 9.95782390e-02
7.47555271e-02 -1.98633727e-02 1.35140395e+00 -5.27905703e-01
-7.17690527e-01 -1.60072759e-01 -3.32903750e-02 4.23407018e-01
3.39803696e-01 -1.26128793e-02 -7.09328949e-01 -3.67573231e-01
-9.93208736e-02 -4.34820175e-01 5.10717750e-01 4.81792659e-01
1.19233537e+00 -5.16446650e-01 -4.30123568e-01 1.53401539e-01
1.09152508e+00 3.02734137e-01 6.62242651e-01 -6.82256892e-02
3.44500244e-01 6.22887313e-01 1.30802274e+00 7.88321137e-01
2.76691765e-01 4.45962071e-01 6.04950666e-01 2.66599298e-01
2.61235666e-02 -3.50886732e-01 6.65735841e-01 6.08069003e-01
2.11176854e-02 -7.39370227e-01 -7.78100908e-01 4.82257187e-01
-1.90046513e+00 -5.65544665e-01 2.92263329e-01 2.17799616e+00
6.38968527e-01 4.10282761e-01 1.85081124e-01 -3.88978515e-03
6.68209910e-01 -1.81960985e-01 -7.04863608e-01 -4.94260460e-01
4.03911859e-01 5.34200929e-02 2.90980488e-01 2.04234704e-01
-1.10063088e+00 7.03332245e-01 6.52692699e+00 8.65086257e-01
-7.26868451e-01 -7.54875988e-02 7.49459922e-01 -1.98557023e-02
-3.53489250e-01 -1.82635278e-01 -8.83712351e-01 3.91443312e-01
6.45817578e-01 2.87264287e-01 4.59170103e-01 1.44148076e+00
-1.78581879e-01 -2.96399236e-01 -1.55208254e+00 8.31998467e-01
-6.00367635e-02 -1.28648627e+00 -5.77932522e-02 -9.87350717e-02
7.56854057e-01 -3.92539054e-01 9.63475257e-02 6.79081380e-01
6.28552258e-01 -9.44979846e-01 7.70467162e-01 3.70703787e-02
5.90280175e-01 -8.49444628e-01 8.62040281e-01 5.91061175e-01
-1.08325863e+00 -5.56293905e-01 -6.11925244e-01 -1.42897859e-01
3.77113611e-01 5.66613615e-01 -1.34397864e+00 8.12912881e-01
1.75057769e-01 5.44981778e-01 -3.98365874e-03 1.00397861e+00
-4.10262734e-01 5.02434969e-01 -3.90292823e-01 -3.10634613e-01
1.23939395e-01 1.07085770e-02 2.83153445e-01 1.07093394e+00
6.62803471e-01 -2.70600230e-01 4.55780417e-01 1.20178723e+00
1.52468473e-01 -4.18345600e-01 -4.21324760e-01 -4.90370691e-01
5.05803049e-01 1.14628994e+00 -9.71475303e-01 -1.16529763e-02
-3.04857790e-01 7.93137968e-01 3.66608024e-01 3.28280091e-01
-1.15971899e+00 4.58182432e-02 5.54332674e-01 -3.26651037e-01
8.10184181e-01 -4.88444626e-01 -2.14890629e-01 -7.65820384e-01
1.38834566e-01 -6.24505877e-01 1.67738676e-01 -5.23705304e-01
-1.53914452e+00 2.22957984e-01 1.64345682e-01 -1.26248729e+00
-3.38378310e-01 -9.28818643e-01 -4.65603888e-01 5.74242830e-01
-1.79862452e+00 -9.56934571e-01 -1.39080301e-01 -7.77649134e-02
9.50448573e-01 -1.19988725e-01 6.43485606e-01 8.76758546e-02
-5.24947703e-01 7.30440497e-01 -3.03408176e-01 -1.94386333e-01
2.40711272e-01 -1.03086424e+00 3.87557358e-01 8.30455363e-01
1.16148246e-02 5.90968728e-01 5.63585639e-01 -9.71647263e-01
-1.77493382e+00 -1.17081964e+00 2.98728973e-01 -3.47874671e-01
3.29832375e-01 -5.40794730e-01 -4.85074222e-01 6.52083337e-01
-2.00153425e-01 -1.04235597e-01 5.08262515e-01 3.10131639e-01
-2.46409625e-01 2.74929821e-01 -1.07278395e+00 4.61344808e-01
1.24638283e+00 1.55626327e-01 -4.00321931e-01 5.36888301e-01
8.75017643e-01 -5.87068737e-01 -8.73107612e-01 6.39225125e-01
5.68086624e-01 -6.56332076e-01 9.89420176e-01 -6.16573334e-01
6.14975691e-01 -3.03497076e-01 -1.24237642e-01 -1.24971330e+00
-2.84633845e-01 -7.48093843e-01 -5.02682269e-01 1.13776720e+00
5.87288022e-01 -5.20281196e-01 8.16557229e-01 4.21647638e-01
-6.99564993e-01 -9.50832367e-01 -6.57494962e-01 -1.35602272e+00
-1.50014341e-01 -4.40444589e-01 8.81165981e-01 6.62709117e-01
2.60642976e-01 1.47415131e-01 -3.03293705e-01 5.21415532e-01
1.39595168e-02 2.37079442e-01 8.06101143e-01 -8.14806879e-01
-5.51888585e-01 -2.58966088e-01 -1.20880567e-01 -1.42363226e+00
2.28321373e-01 -6.82557762e-01 5.57501495e-01 -1.50860918e+00
4.37698849e-02 -8.66483986e-01 -1.80695742e-01 3.72813076e-01
-3.31400968e-02 -4.90449756e-01 1.34358153e-01 -1.91124156e-01
-6.40059650e-01 6.22123420e-01 1.19100475e+00 -1.69348910e-01
-3.15648973e-01 2.46927112e-01 -4.64750558e-01 7.37981260e-01
9.59020019e-01 -2.81914145e-01 -7.29570866e-01 -3.24836195e-01
4.29958522e-01 9.95832533e-02 5.79833686e-02 -8.60693872e-01
-1.49707153e-01 -5.20710528e-01 3.06189746e-01 -8.47962379e-01
1.56073242e-01 -6.69344783e-01 5.85437492e-02 6.11123204e-01
-2.09449247e-01 4.83163632e-03 4.22592521e-01 7.92803645e-01
3.37042600e-01 -6.44874990e-01 2.50857592e-01 2.63799690e-02
-8.50477457e-01 1.11349098e-01 -4.49539512e-01 -2.18172520e-01
1.12021565e+00 -6.37430102e-02 -4.45363551e-01 -3.55093926e-02
-4.26768690e-01 2.17011258e-01 4.27767605e-01 6.20552003e-01
7.64749646e-01 -9.44930017e-01 -3.85550857e-01 1.51270628e-01
2.04250038e-01 2.38317713e-01 2.14190021e-01 6.44087493e-01
-4.83609825e-01 3.67986023e-01 -1.38866559e-01 -6.40238285e-01
-8.01937699e-01 1.00301504e+00 -1.85178950e-01 -4.80075538e-01
-4.26743358e-01 1.03454745e+00 3.79675537e-01 -6.75850809e-01
3.58878285e-01 -8.85823965e-01 7.36406026e-03 -4.08728123e-01
2.47067750e-01 3.53737235e-01 -3.21623906e-02 1.05993658e-01
-3.70446980e-01 8.23885873e-02 -1.35581613e-01 3.61585796e-01
1.23747075e+00 7.80428424e-02 2.22884357e-01 2.47243598e-01
7.12048829e-01 -2.82554030e-01 -1.67429638e+00 3.50742012e-01
1.74085721e-01 -6.31668746e-01 -4.41416889e-01 -1.03109968e+00
-7.39227772e-01 8.34028780e-01 3.74363631e-01 -2.26755682e-02
8.34296286e-01 7.99254328e-02 7.21991241e-01 6.00908756e-01
6.27072513e-01 -1.12465465e+00 3.01923066e-01 4.48854268e-01
8.44435632e-01 -1.23945653e+00 2.63035391e-03 -1.02199852e+00
-6.64186478e-01 1.21202159e+00 8.29752743e-01 8.89452472e-02
1.16586894e-01 5.88649631e-01 -5.96309423e-01 1.65168997e-02
-7.43497849e-01 -4.78774197e-02 1.50691986e-01 1.05901921e+00
1.98141962e-01 2.45200753e-01 2.13524103e-01 6.72055721e-01
-1.48181275e-01 1.54332556e-02 6.33450806e-01 1.20930350e+00
-3.32495332e-01 -9.47683096e-01 -6.24558851e-02 5.95051050e-01
1.11037038e-01 1.13436198e-02 2.33918756e-01 6.83177829e-01
5.22949956e-02 1.04121399e+00 1.09231062e-01 -2.67216384e-01
1.78960875e-01 -4.31972966e-02 9.67496037e-01 -1.01293707e+00
-1.64238989e-01 3.46973054e-02 5.70657611e-01 -5.53483665e-01
-1.39228672e-01 -6.17586493e-01 -1.18566477e+00 7.33031258e-02
-8.48593056e-01 -2.17081174e-01 6.44942105e-01 9.78379548e-01
4.68162775e-01 7.99404860e-01 4.99642730e-01 -9.65518177e-01
-7.91222394e-01 -8.07480752e-01 -2.78160483e-01 -8.23873580e-02
9.04333740e-02 -9.67422307e-01 -6.15335954e-03 3.24824527e-02] | [4.996531009674072, 2.7051491737365723] |
e99aac64-0728-44b5-89d7-f3a98efa30c0 | didn-t-see-that-coming-a-survey-on-non-verbal | 2203.02480 | null | https://arxiv.org/abs/2203.02480v1 | https://arxiv.org/pdf/2203.02480v1.pdf | Didn't see that coming: a survey on non-verbal social human behavior forecasting | Non-verbal social human behavior forecasting has increasingly attracted the interest of the research community in recent years. Its direct applications to human-robot interaction and socially-aware human motion generation make it a very attractive field. In this survey, we define the behavior forecasting problem for multiple interactive agents in a generic way that aims at unifying the fields of social signals prediction and human motion forecasting, traditionally separated. We hold that both problem formulations refer to the same conceptual problem, and identify many shared fundamental challenges: future stochasticity, context awareness, history exploitation, etc. We also propose a taxonomy that comprises methods published in the last 5 years in a very informative way and describes the current main concerns of the community with regard to this problem. In order to promote further research on this field, we also provide a summarised and friendly overview of audiovisual datasets featuring non-acted social interactions. Finally, we describe the most common metrics used in this task and their particular issues. | ['Cristina Palmero', 'Isabelle Guyon', 'Wei-Wei Tu', 'Zhen Xu', 'Sergio Escalera', 'Johnny Núñez', 'German Barquero'] | 2022-03-04 | null | null | null | null | ['human-behavior-forecasting'] | ['time-series'] | [ 2.48163402e-01 4.21683639e-01 -1.12820745e-01 -3.25311542e-01
-2.10542828e-01 -3.06315631e-01 9.55676138e-01 -1.26787603e-01
-4.81643647e-01 8.54806602e-01 6.36722863e-01 4.69459802e-01
-2.96381056e-01 -2.95805365e-01 -3.85434814e-02 -1.03960037e+00
-6.96656525e-01 4.71733510e-01 2.88310111e-01 -6.19218230e-01
1.78069398e-01 3.93819332e-01 -2.11494875e+00 1.64580196e-01
3.43361109e-01 5.98522663e-01 1.77918777e-01 9.62035358e-01
2.54038423e-01 1.02894664e+00 -8.07383835e-01 -3.65764678e-01
-2.77478993e-01 -5.39311171e-01 -1.02125943e+00 2.30527833e-01
-3.58604699e-01 1.51831657e-01 -3.89943719e-01 5.84518075e-01
7.14974523e-01 6.19253933e-01 9.63100791e-01 -1.85066402e+00
-5.41970171e-02 7.45010614e-01 -2.11045057e-01 5.43030873e-02
1.11132681e+00 1.29811838e-01 8.87165248e-01 -2.92594850e-01
8.93250287e-01 1.38857388e+00 7.37787187e-01 8.13831508e-01
-8.21023941e-01 -1.88434288e-01 4.18472916e-01 6.58947229e-01
-1.11698020e+00 -4.95386124e-01 6.66351020e-01 -6.58373058e-01
1.00346708e+00 6.05586112e-01 8.85836899e-01 1.53312206e+00
-3.87817025e-02 1.32213056e+00 6.28751814e-01 -2.97094226e-01
3.14544261e-01 2.71413494e-02 1.29547611e-01 4.05964375e-01
-2.64362693e-01 1.02075368e-01 -7.55351365e-01 -4.37162817e-01
3.32719058e-01 -4.14183527e-01 -1.59262002e-01 -5.27014017e-01
-1.54455817e+00 9.01379406e-01 2.87789106e-02 6.30521119e-01
-5.35244584e-01 2.93304354e-01 5.08486390e-01 2.29029477e-01
4.47956413e-01 1.07903704e-01 -6.60242736e-02 -8.18091393e-01
-3.10740113e-01 8.22104812e-01 1.10719121e+00 6.77506506e-01
2.00354546e-01 1.13506883e-01 3.40862535e-02 8.14197540e-01
4.02209669e-01 2.50151634e-01 4.79391128e-01 -1.37604499e+00
-4.92803492e-02 -2.34816093e-02 2.85821229e-01 -1.29926097e+00
-9.66526270e-01 1.26126111e-01 -8.02646041e-01 3.59265506e-03
3.99138391e-01 -4.34724182e-01 4.78938483e-02 1.69206107e+00
4.10176367e-01 2.36244470e-01 3.41430455e-01 7.50776529e-01
9.42750216e-01 8.70442569e-01 1.70427367e-01 -7.04782784e-01
1.03887856e+00 -1.05149651e+00 -8.66198421e-01 -2.92022899e-02
6.12527847e-01 -4.94782835e-01 5.23655295e-01 3.87932330e-01
-1.02827239e+00 -1.87528834e-01 -7.48792589e-01 3.33900422e-01
-3.53641599e-01 -1.88728333e-01 9.57699180e-01 5.22460222e-01
-1.16962516e+00 7.03335941e-01 -8.48689318e-01 -9.79977489e-01
-3.38405758e-01 4.13436174e-01 -1.21888302e-01 6.02719605e-01
-1.15353394e+00 9.76345420e-01 1.28984302e-02 -4.86131944e-02
-4.35985893e-01 1.19949438e-01 -8.04942846e-01 -4.91033942e-01
3.06617826e-01 -5.77857435e-01 1.65468693e+00 -8.34682703e-01
-2.04068041e+00 9.66398597e-01 -1.65478647e-01 -5.03925204e-01
9.01132524e-01 -1.51978269e-01 -4.29991275e-01 -5.70043288e-02
6.35973215e-02 6.17387056e-01 6.04882121e-01 -1.14037812e+00
-1.06929481e+00 -2.13020965e-01 -1.05624251e-01 6.21544242e-01
-1.84122726e-01 3.78061473e-01 -2.99587190e-01 -9.31647182e-01
-4.12682086e-01 -1.25916469e+00 -5.48442900e-01 -4.24704045e-01
-2.34445617e-01 -7.05210507e-01 5.00251472e-01 -1.19295433e-01
1.38998806e+00 -2.09999871e+00 8.66105318e-01 5.65263852e-02
1.24697067e-01 -1.40694216e-01 -5.37376590e-02 9.04667556e-01
3.04313272e-01 -3.49667996e-01 -2.67037153e-01 -6.58772171e-01
2.56633401e-01 2.05099449e-01 -2.68169433e-01 7.46448398e-01
-4.56363469e-01 7.71916866e-01 -1.23198211e+00 -4.60659713e-01
4.71964717e-01 4.78764862e-01 -2.60582209e-01 1.40982538e-01
-1.11528330e-01 7.71985471e-01 -5.07117331e-01 2.32062817e-01
4.94870543e-02 -4.62613143e-02 3.44887495e-01 4.29433435e-01
-2.04561532e-01 -1.88535154e-01 -1.16470110e+00 1.35952759e+00
-1.99494228e-01 8.93757224e-01 -5.36222477e-03 -8.88542235e-01
6.77712381e-01 5.25321126e-01 1.14370716e+00 -1.47492802e-02
1.96216211e-01 1.68465093e-01 -1.79610059e-01 -7.46454000e-01
7.75638521e-01 1.34144321e-01 -5.32700419e-01 5.73470294e-01
-8.04117322e-02 -7.61113465e-02 2.09984332e-01 4.52578366e-02
9.44513321e-01 1.61726743e-01 8.55330408e-01 4.52081719e-03
8.47385764e-01 -4.47405092e-02 1.90292194e-01 7.31515110e-01
-6.36899829e-01 1.29993618e-01 3.84012818e-01 -5.04932821e-01
-3.75746071e-01 -6.22160316e-01 2.12319210e-01 1.56155884e+00
2.91312516e-01 -5.84380686e-01 -6.81583762e-01 -3.52378875e-01
-2.71339953e-01 5.18435419e-01 -7.43072331e-01 1.34082973e-01
-7.55956590e-01 -9.26286757e-01 4.28662419e-01 3.15481186e-01
1.91677541e-01 -1.71000433e+00 -1.03440535e+00 4.18810517e-01
-6.40358150e-01 -1.18162644e+00 4.45877761e-02 -2.81502247e-01
-5.23406982e-01 -8.28239858e-01 -1.22547078e+00 -9.07763422e-01
-6.81671649e-02 3.98697853e-01 9.66311157e-01 4.22950163e-02
-5.03959041e-03 1.09592891e+00 -7.60495722e-01 -4.68047708e-01
-5.91118693e-01 1.14840843e-01 4.29356098e-01 3.07440132e-01
3.87671202e-01 -6.40542865e-01 -3.19554418e-01 5.24865389e-01
-3.26526374e-01 -1.14516906e-01 -1.36255044e-02 4.23515946e-01
-1.75768673e-01 -1.80125311e-01 4.93157059e-01 -5.04102707e-01
9.03583467e-01 -7.21236706e-01 -3.05344999e-01 3.08132470e-02
-6.05351143e-02 -2.38137811e-01 4.44739014e-02 -6.24265790e-01
-1.03292656e+00 4.02208775e-01 -7.37790540e-02 3.51292014e-01
-4.75576311e-01 2.46413916e-01 2.36356691e-01 -1.07686080e-01
6.43471301e-01 2.37632826e-01 1.02530308e-01 -1.03252225e-01
5.58655322e-01 7.88078964e-01 4.87806499e-01 -7.47355223e-02
3.13208312e-01 6.81376934e-01 8.52149818e-03 -1.45043862e+00
-1.63301513e-01 -6.91105902e-01 -6.26557350e-01 -1.03774309e+00
9.25078392e-01 -5.59293270e-01 -1.47169030e+00 9.79639351e-01
-1.39775479e+00 -5.50754011e-01 -1.54360205e-01 4.90064383e-01
-1.23922670e+00 5.05515099e-01 -7.39690840e-01 -1.53399312e+00
1.51497006e-01 -9.49569166e-01 1.05588913e+00 -6.10817224e-03
-8.73547852e-01 -1.22161591e+00 4.32732671e-01 2.45016634e-01
1.68551281e-01 5.92996836e-01 2.03594893e-01 -6.09333992e-01
-1.96956694e-01 -1.57834068e-01 4.61174011e-01 -4.24830049e-01
8.83728173e-03 -1.30847422e-03 -1.02181661e+00 1.04935184e-01
-1.42374128e-01 -2.57375091e-01 6.59964085e-01 6.55492902e-01
2.61882037e-01 -9.87578183e-02 -7.03916371e-01 -1.30371034e-01
5.45110822e-01 4.86930192e-01 2.89683700e-01 3.94375443e-01
4.65338677e-01 1.51660538e+00 8.65312636e-01 8.95298183e-01
8.09352756e-01 1.16886067e+00 1.84200898e-01 4.39481884e-01
3.39255124e-01 1.91881344e-01 6.91297352e-01 8.49117756e-01
-7.01202810e-01 -7.18672514e-01 -7.85511792e-01 3.32999289e-01
-2.54849172e+00 -1.33456421e+00 -3.07397723e-01 1.74474502e+00
1.49347782e-01 -3.84678841e-01 9.55899417e-01 3.60947907e-01
7.94221699e-01 5.39974809e-01 -1.07523687e-01 -2.61402309e-01
-1.60375386e-01 -8.19524527e-01 7.59595558e-02 8.51546407e-01
-1.29893792e+00 1.04487085e+00 6.98345375e+00 6.68674469e-01
-7.13213325e-01 -4.06136885e-02 2.61645228e-01 9.52178463e-02
2.25986540e-01 -4.98618722e-01 -6.11531317e-01 3.68243635e-01
8.76459658e-01 -2.07415581e-01 4.80646074e-01 7.27047205e-01
4.97651666e-01 -3.29052031e-01 -1.00702286e+00 1.34339321e+00
3.29502672e-01 -9.53337312e-01 -3.01120937e-01 1.25435248e-01
3.59088749e-01 -1.16735049e-01 1.02861650e-01 1.55483410e-01
4.50183094e-01 -8.87960196e-01 8.61924410e-01 6.45937324e-01
5.10472506e-02 -6.11657381e-01 6.82167590e-01 5.82991004e-01
-1.33797777e+00 -3.09315503e-01 5.47371320e-02 -4.97261792e-01
1.01918542e+00 -4.52748910e-02 -4.49728012e-01 3.43281776e-01
7.28310347e-01 1.22255385e+00 -7.89490119e-02 1.03141177e+00
1.00550786e-01 1.18609153e-01 -2.22024903e-01 -8.61878872e-01
3.80809456e-01 -1.97678909e-01 1.08631337e+00 1.20909798e+00
1.92524254e-01 2.89791524e-01 1.77721918e-01 -5.58844358e-02
7.47333884e-01 3.80022496e-01 -9.00649965e-01 6.19774908e-02
1.01000853e-01 8.66352439e-01 -8.06424499e-01 -1.20882750e-01
-3.49637389e-01 1.24154007e+00 4.90658656e-02 2.24612251e-01
-7.48462617e-01 -9.75190550e-02 7.88506448e-01 -2.20882282e-01
-1.24097884e-01 -3.52240920e-01 3.51709694e-01 -9.67393994e-01
-3.56776893e-01 -6.58566058e-01 5.04675686e-01 -5.53286254e-01
-1.04008567e+00 8.78668845e-01 4.74682301e-01 -1.36820316e+00
-1.15995920e+00 -5.27675867e-01 -2.41531760e-01 -4.05744202e-02
-7.61923313e-01 -9.13540244e-01 -2.52607912e-01 4.36908782e-01
7.30912089e-01 -3.44836086e-01 1.00022352e+00 3.31284218e-02
-4.12977278e-01 4.24672402e-02 1.46577824e-02 -3.45097244e-01
4.71892774e-01 -9.88742590e-01 4.19121534e-01 2.33623177e-01
1.19497821e-01 8.30620825e-02 1.24713528e+00 -5.19856989e-01
-9.69202816e-01 -6.00417852e-01 1.17801750e+00 -6.69134974e-01
7.07535565e-01 -1.73636451e-01 -4.47276592e-01 6.80356443e-01
3.17531437e-01 -6.44471288e-01 7.17256367e-01 -1.85179859e-01
4.84003931e-01 4.68585700e-01 -9.24242139e-01 1.08481979e+00
1.40319157e+00 1.79251023e-02 -2.80085593e-01 5.13982892e-01
3.40443313e-01 -3.51506807e-02 -5.06148756e-01 3.26955646e-01
7.42407858e-01 -1.30616724e+00 9.12294209e-01 -4.24313277e-01
-3.34928781e-01 9.00414884e-02 -1.84709400e-01 -1.23737419e+00
-3.16321671e-01 -1.45310354e+00 -2.81369805e-01 8.58546019e-01
7.78781325e-02 -6.22748911e-01 8.94459009e-01 4.44334209e-01
1.99576393e-01 -4.44374233e-01 -8.71790171e-01 -6.47673905e-01
-1.44707218e-01 -6.84890449e-01 1.93643361e-01 7.38260448e-01
7.94301331e-01 4.11147803e-01 -9.81826365e-01 -3.19194406e-01
4.40900207e-01 -2.31244773e-01 1.04276359e+00 -1.45097971e+00
-1.30549654e-01 -7.75997818e-01 -9.54869509e-01 -1.33981240e+00
6.47584379e-01 -4.40254092e-01 3.08246493e-01 -1.19302237e+00
-9.59998667e-02 1.36123896e-01 3.48781288e-01 -2.20879853e-01
2.41903067e-01 2.88361549e-01 3.55409354e-01 1.44425243e-01
-9.27605271e-01 7.25284338e-01 8.25861931e-01 1.07946932e-01
-5.37206948e-01 7.19523609e-01 -2.48246670e-01 1.24773192e+00
8.05445969e-01 -2.01450825e-01 -4.86508876e-01 5.82611524e-02
2.75433809e-01 3.74035954e-01 3.02068681e-01 -1.09310663e+00
4.60344136e-01 -5.02743006e-01 -4.31397974e-01 -3.78303796e-01
9.56655204e-01 -6.91080391e-01 1.71004236e-01 6.90825760e-01
-4.98180687e-01 3.86881791e-02 -2.44210243e-01 9.11735952e-01
-1.30643725e-01 -1.53094426e-01 5.08923054e-01 -1.17747232e-01
-1.16923416e+00 7.62636401e-03 -1.41994071e+00 -2.75433630e-01
1.52019608e+00 -2.92036802e-01 3.23940128e-01 -1.33239734e+00
-1.17286229e+00 2.76561260e-01 1.16668254e-01 7.33371258e-01
4.11417633e-01 -1.15399361e+00 -7.15738177e-01 -4.69965309e-01
2.79748619e-01 -8.24330330e-01 5.70020795e-01 8.33535671e-01
-3.91497761e-01 3.98803949e-01 -1.35240793e-01 -5.53197920e-01
-1.56343114e+00 3.93642128e-01 1.79594114e-01 5.41157350e-02
-5.05648077e-01 7.92291582e-01 -7.66058713e-02 -3.61268699e-01
7.97679007e-01 -2.22117342e-02 -1.06457365e+00 2.47926563e-01
6.03671551e-01 1.05536187e+00 -6.20592833e-01 -1.50113547e+00
-5.00644565e-01 6.34685934e-01 5.60419381e-01 -6.47577763e-01
1.35259867e+00 -6.61140740e-01 -3.98238264e-02 1.09898806e+00
9.25727665e-01 -2.07228735e-01 -1.00006437e+00 -1.59677833e-01
4.06225204e-01 3.08550131e-02 -5.28076589e-01 -2.61237770e-01
-5.84608495e-01 5.54787040e-01 4.69612837e-01 8.18205595e-01
7.80694246e-01 4.10951585e-01 5.50017655e-01 5.03463209e-01
7.59028494e-01 -1.02406847e+00 2.52904147e-01 6.52332008e-01
1.24120462e+00 -1.02403629e+00 -1.38744146e-01 -8.10467064e-01
-1.21228433e+00 1.07728112e+00 3.05158913e-01 7.74420938e-03
8.10787797e-01 -2.53126509e-02 1.15061425e-01 -4.87393290e-02
-7.76120245e-01 -4.14064854e-01 5.95261194e-02 1.15461099e+00
4.07778352e-01 1.83959842e-01 -2.04835430e-01 5.15379310e-01
-4.81382310e-01 -2.03209981e-01 6.58541143e-01 9.57661033e-01
-4.01627928e-01 -1.00104117e+00 -2.80194134e-01 6.00890778e-02
-1.17239282e-01 6.00604355e-01 -7.21685171e-01 6.20797813e-01
-2.88840026e-01 1.43960178e+00 -9.88750681e-02 -6.81880951e-01
2.72766232e-01 -2.43623018e-01 3.13686639e-01 -3.37898463e-01
-5.91278255e-01 -1.27918301e-02 4.55265999e-01 -7.67245352e-01
-1.20656550e+00 -1.08431697e+00 -9.55126345e-01 -5.97161293e-01
2.53400928e-03 2.31965423e-01 3.56380105e-01 8.75413477e-01
1.65955238e-02 8.33689198e-02 5.11234999e-01 -1.63528943e+00
-2.24213853e-01 -9.61523414e-01 -6.85082674e-01 1.79028243e-01
4.08304930e-01 -8.84868145e-01 -4.18007314e-01 1.46181434e-01] | [7.184111595153809, 0.0874871015548706] |
74e11a3d-4c4e-47aa-9cc1-86eb2486374f | robust-boosting-forests-with-richer-deep | 2210.16451 | null | https://arxiv.org/abs/2210.16451v1 | https://arxiv.org/pdf/2210.16451v1.pdf | Robust Boosting Forests with Richer Deep Feature Hierarchy | We propose a robust variant of boosting forest to the various adversarial defense methods, and apply it to enhance the robustness of the deep neural network. We retain the deep network architecture, weights, and middle layer features, then install gradient boosting forest to select the features from each layer of the deep network, and predict the target. For training each decision tree, we propose a novel conservative and greedy trade-off, with consideration for less misprediction instead of pure gain functions, therefore being suboptimal and conservative. We actively increase tree depth to remedy the accuracy with splits in more features, being more greedy in growing tree depth. We propose a new task on 3D face model, whose robustness has not been carefully studied, despite the great security and privacy concerns related to face analytics. We tried a simple attack method on a pure convolutional neural network (CNN) face shape estimator, making it degenerate to only output average face shape with invisible perturbation. Our conservative-greedy boosting forest (CGBF) on face landmark datasets showed a great improvement over original pure deep learning methods under the adversarial attacks. | ['Jianqiao Wangni'] | 2022-10-29 | null | null | null | null | ['adversarial-defense', 'face-model'] | ['adversarial', 'computer-vision'] | [ 8.24313909e-02 3.81987751e-01 1.06678233e-01 -5.25977135e-01
-2.38759592e-01 -7.23132432e-01 4.11970645e-01 -6.46331489e-01
-1.77175879e-01 7.26628244e-01 -2.77406834e-02 -6.91075206e-01
1.24842837e-01 -1.00937736e+00 -8.09137881e-01 -1.07823002e+00
-3.95060778e-01 -6.97216466e-02 1.46828289e-03 -2.69604385e-01
9.82889831e-02 1.05026805e+00 -1.24009919e+00 3.42744529e-01
4.95464712e-01 1.42283356e+00 -8.22413623e-01 5.23557901e-01
1.02441818e-01 7.08319128e-01 -4.84691769e-01 -1.16047430e+00
8.31501245e-01 -7.69638196e-02 -7.20340371e-01 -4.58759218e-01
6.88274503e-01 -6.26289904e-01 -2.63901919e-01 1.19675446e+00
1.00292671e+00 -4.89137739e-01 5.79169095e-01 -1.82255626e+00
-5.89066207e-01 5.81820965e-01 -8.86033714e-01 4.91218567e-02
3.33096571e-02 2.83367962e-01 3.90988708e-01 -7.03964174e-01
2.76613623e-01 1.49974775e+00 1.14359188e+00 1.18346941e+00
-1.04952693e+00 -1.48303831e+00 2.98988789e-01 8.66753235e-02
-1.07837296e+00 -8.54697049e-01 1.06725907e+00 -2.01337308e-01
2.67915547e-01 3.79368067e-01 3.78800839e-01 1.44326282e+00
1.55393556e-01 3.81535143e-01 1.27477872e+00 -1.83406204e-01
2.23001555e-01 2.03417048e-01 -7.33344406e-02 1.09452879e+00
2.65970111e-01 6.25143647e-01 -2.87889272e-01 -7.48750508e-01
5.08805811e-01 6.68777600e-02 -1.44800648e-01 -5.10338306e-01
-2.63677508e-01 9.43462074e-01 6.91058338e-01 -2.03753605e-01
-1.51408508e-01 1.00337006e-01 5.24746239e-01 4.98486370e-01
4.70707119e-01 -9.97943059e-02 -9.24187243e-01 4.98338193e-01
-6.16267502e-01 3.64641875e-01 7.49303162e-01 6.36916995e-01
7.01512277e-01 2.12573975e-01 -1.38581529e-01 4.79694396e-01
3.84872556e-01 4.79522675e-01 2.57595718e-01 -8.16350937e-01
2.51716584e-01 4.31322217e-01 -2.87736982e-01 -1.28111851e+00
-3.71916950e-01 -4.73927468e-01 -1.25299215e+00 1.06092930e+00
5.16877055e-01 -4.00819182e-01 -1.05583966e+00 1.95150042e+00
6.93232834e-01 1.01607561e-01 -2.17163786e-01 6.46271706e-01
7.39677548e-01 2.77547184e-02 -3.19404751e-02 -2.74264842e-01
1.13183069e+00 -6.92673028e-01 -2.49816686e-01 -2.96689142e-02
4.71362144e-01 -5.26684821e-01 8.98501039e-01 4.41243738e-01
-8.05636764e-01 -4.81116444e-01 -1.12565935e+00 2.35502839e-01
-4.80076760e-01 -1.98490962e-01 7.74341047e-01 1.59449053e+00
-1.07235110e+00 9.17644262e-01 -5.38487494e-01 9.80759338e-02
1.11623025e+00 6.44868195e-01 -6.20002091e-01 2.13461146e-01
-1.09342086e+00 6.62196934e-01 -2.12578207e-01 3.93990010e-01
-1.10896349e+00 -6.46374404e-01 -6.34617567e-01 -1.76591933e-01
5.36722988e-02 -6.43768191e-01 8.41024697e-01 -1.10287690e+00
-1.70173860e+00 1.05634928e+00 -6.02771230e-02 -5.80640972e-01
7.97709584e-01 -1.15963683e-01 -2.66304225e-01 -2.71027595e-01
-1.96291655e-01 5.97160816e-01 1.55034268e+00 -1.20415568e+00
-1.12880230e-01 -9.96656179e-01 6.36439398e-02 -4.17335838e-01
-3.54625612e-01 2.37562165e-01 4.10445124e-01 -6.24812901e-01
1.87922567e-01 -9.16257560e-01 -2.76137084e-01 4.64203626e-01
-4.93978560e-01 9.28324759e-02 1.10947871e+00 -6.55174494e-01
9.30884361e-01 -2.08577585e+00 -3.66704047e-01 4.03303325e-01
3.56038958e-01 4.52419400e-01 -1.82954445e-01 -1.37000248e-01
-4.13953811e-01 4.27901208e-01 -2.86276758e-01 -1.53214753e-01
-1.89035952e-01 -1.09662667e-01 -8.26884389e-01 8.02055120e-01
7.14043230e-02 7.63839424e-01 -3.90917361e-01 -4.23893094e-01
-3.93456370e-01 3.79478186e-01 -8.46533239e-01 1.31216317e-01
7.54580200e-02 3.49595368e-01 -4.77433264e-01 1.01482260e+00
1.66660237e+00 4.76946861e-01 -1.37183815e-01 -2.02067316e-01
4.77034837e-01 -1.08357873e-02 -9.56114948e-01 9.58699286e-01
-3.15756053e-01 1.98873878e-01 5.28821528e-01 -9.58855867e-01
1.21503353e+00 9.89852026e-02 1.71478719e-01 -2.33835861e-01
2.65537024e-01 4.21130694e-02 1.38553277e-01 -2.88579315e-01
-3.23992312e-01 -9.28068981e-02 2.27136478e-01 3.14937025e-01
-1.13276906e-01 1.04604535e-01 -9.52258229e-01 1.57003161e-02
1.15669930e+00 2.74380893e-01 2.18278617e-01 -2.72600293e-01
8.03073287e-01 -4.93021727e-01 7.32311904e-01 6.11365676e-01
-8.12243164e-01 4.13344681e-01 8.68615389e-01 -9.71746445e-01
-5.34497380e-01 -9.16071594e-01 -2.08015129e-01 1.39911532e+00
-2.64917046e-01 -2.17047632e-01 -1.08040309e+00 -1.53295135e+00
4.74619150e-01 1.98051661e-01 -1.01003850e+00 -4.62027431e-01
-7.50448525e-01 -9.00856197e-01 1.24023616e+00 3.32247078e-01
9.26423252e-01 -9.85719025e-01 -1.36632979e-01 -2.64605343e-01
3.78517807e-01 -6.40921950e-01 -2.05015033e-01 2.81599730e-01
-9.09196854e-01 -1.15715802e+00 -3.21885794e-01 -6.73430979e-01
6.77515447e-01 -1.55342937e-01 8.88853073e-01 3.24634701e-01
-1.94226652e-01 -2.83854097e-01 1.12345582e-02 -6.93734407e-01
-5.15258074e-01 3.17342393e-02 3.90247613e-01 1.52655199e-01
2.36048430e-01 -8.87815297e-01 -6.95350945e-01 4.11335498e-01
-4.60526079e-01 -5.32204866e-01 4.04281765e-01 9.78411257e-01
-6.46415055e-02 -1.94587156e-01 7.58467436e-01 -1.10741639e+00
4.61895138e-01 -3.00221503e-01 -4.70673233e-01 4.20593508e-02
-7.81187713e-01 2.58394424e-02 7.76732981e-01 -5.17654896e-01
-8.73549163e-01 2.85285324e-01 -5.81669152e-01 -5.48304975e-01
-1.85973689e-01 -3.97469580e-01 -5.46359837e-01 -8.54248166e-01
1.02161908e+00 6.00790679e-02 2.24606410e-01 -4.79715079e-01
4.33104068e-01 6.99498296e-01 2.40290582e-01 -6.13218844e-01
1.05689895e+00 5.55255890e-01 4.43460792e-01 -1.77344158e-01
-6.38474047e-01 4.14009660e-01 -5.39406717e-01 -1.25686467e-01
3.90028059e-01 -5.66820443e-01 -1.30615091e+00 8.03524196e-01
-1.21161807e+00 8.23482946e-02 9.04419571e-02 -1.63920775e-01
-3.83662820e-01 3.21583241e-01 -5.87552011e-01 -8.69080484e-01
-7.45865166e-01 -9.30466175e-01 8.18580687e-01 -7.41141513e-02
2.79178977e-01 -4.55311626e-01 -3.80323939e-02 1.73254505e-01
6.46091104e-01 6.13877833e-01 9.03244972e-01 -8.35622728e-01
-2.26539373e-01 -3.54674459e-01 -3.30055624e-01 6.24354303e-01
7.97265917e-02 7.32783675e-02 -1.72108889e+00 -5.89382172e-01
5.21426678e-01 -4.61642146e-01 1.06389999e+00 2.08926603e-01
1.65319908e+00 -8.74370635e-01 -2.72424906e-01 1.22988284e+00
1.07997906e+00 -3.22366059e-02 6.07817352e-01 2.80648917e-01
6.16214514e-01 7.03843474e-01 1.35237217e-01 1.85603186e-01
-2.19540268e-01 4.52960014e-01 1.02468693e+00 -9.54364240e-02
6.12128489e-02 -3.16611081e-01 5.43596685e-01 4.82372940e-03
-2.23149166e-01 4.30529684e-01 -5.51068783e-01 -4.93155085e-02
-1.36905968e+00 -9.54645514e-01 2.76572734e-01 2.02252698e+00
9.53451633e-01 2.96142638e-01 2.77696699e-01 3.09977029e-02
7.59289265e-01 3.33473593e-01 -7.34217346e-01 -7.97751248e-01
6.89231157e-02 3.90430212e-01 7.09499538e-01 3.22384834e-01
-1.27510881e+00 1.01530588e+00 6.59791803e+00 1.01776874e+00
-1.46522403e+00 1.99868709e-01 1.08940959e+00 -1.38132393e-01
4.30144742e-02 -9.63613167e-02 -9.99909818e-01 4.09990132e-01
5.12165964e-01 2.11571887e-01 5.88301957e-01 1.41919005e+00
-2.80710012e-01 6.68327928e-01 -8.98977041e-01 8.20875466e-01
-1.27572164e-01 -1.12902820e+00 1.18601916e-03 2.55609632e-01
4.42432255e-01 -1.86551452e-01 2.91307807e-01 4.09994453e-01
4.77623314e-01 -1.31290495e+00 4.90975082e-01 1.33952439e-01
7.71918476e-01 -8.50232661e-01 5.64500570e-01 4.63945568e-01
-8.37834060e-01 -4.45892572e-01 -6.75185204e-01 -1.03195570e-01
-5.57502627e-01 5.76945901e-01 -7.94020951e-01 2.04665616e-01
1.00729227e+00 4.39840853e-02 -6.33086026e-01 4.79284763e-01
-2.03087047e-01 7.04595745e-01 -4.15390909e-01 1.15340047e-01
-1.65425062e-01 1.72063082e-01 5.91987789e-01 9.99576330e-01
-3.83847370e-03 -2.06695348e-01 -1.30598009e-01 6.54328108e-01
-3.89146000e-01 5.64029105e-02 -9.86980617e-01 6.38105273e-01
6.01047397e-01 1.38673592e+00 -3.46792907e-01 1.02493286e-01
7.46630132e-02 6.94383323e-01 4.93429691e-01 7.05194026e-02
-8.24294984e-01 -2.00245589e-01 9.49389338e-01 1.13178462e-01
2.37345234e-01 2.20870256e-01 -7.07793474e-01 -9.00595188e-01
2.33743891e-01 -1.13686776e+00 4.44842637e-01 -2.57547647e-01
-1.39795506e+00 8.95493388e-01 -3.39404702e-01 -1.03324199e+00
-2.26319820e-01 -7.47426152e-01 -9.30034339e-01 8.86611700e-01
-1.27852511e+00 -1.49562919e+00 -6.17155991e-02 1.00860465e+00
-1.11442424e-01 -5.69349825e-01 9.06724393e-01 3.32351252e-02
-5.82768857e-01 1.35942757e+00 -3.15438539e-01 3.97153437e-01
7.33288229e-01 -8.54734004e-01 6.87864542e-01 8.19200277e-01
-1.71125397e-01 6.76434517e-01 4.00503486e-01 -5.23128748e-01
-1.21078169e+00 -1.19184077e+00 4.85613465e-01 -4.42954421e-01
3.64685506e-01 -7.41774619e-01 -7.86302269e-01 6.26465142e-01
-3.93882543e-02 3.69363576e-01 5.27192533e-01 1.96639612e-01
-9.93997633e-01 -6.11911833e-01 -2.09263849e+00 5.61490357e-01
1.45065880e+00 -3.69130969e-01 -2.16181993e-01 1.27836347e-01
8.74593675e-01 -2.24159047e-01 -3.48272741e-01 9.12976146e-01
8.91974747e-01 -1.24943602e+00 1.15586007e+00 -9.88435030e-01
2.46415153e-01 -1.07352167e-01 -1.66117057e-01 -1.04658222e+00
-4.30203825e-01 -8.67999792e-01 -3.17721039e-01 1.31908703e+00
2.83974499e-01 -9.63648200e-01 1.41466820e+00 4.70904201e-01
3.20943117e-01 -1.09877253e+00 -1.25039542e+00 -6.50508404e-01
5.24044156e-01 -2.70609617e-01 1.17542660e+00 8.46915603e-01
-4.20562088e-01 -1.70375630e-01 -6.37433052e-01 3.46979022e-01
7.66123891e-01 -4.82090898e-02 1.05825984e+00 -1.18682671e+00
-1.17115468e-01 -4.09765989e-01 -6.03514254e-01 -5.53790390e-01
3.10380191e-01 -7.83221245e-01 -2.66295761e-01 -3.66289228e-01
-1.30498514e-01 -4.25863087e-01 -3.23429763e-01 8.58696520e-01
-8.15708712e-02 5.13921797e-01 6.91907629e-02 -3.19702104e-02
2.88741320e-01 3.82682562e-01 1.15321839e+00 -2.25146294e-01
6.33962080e-02 6.30116522e-01 -1.13318503e+00 9.63876545e-01
7.56958485e-01 -7.65891969e-01 -1.80758014e-01 -2.01086141e-02
-3.18044424e-02 -1.46823257e-01 6.37724936e-01 -8.87126863e-01
1.88783899e-01 -6.24715909e-02 9.42423999e-01 -2.74015754e-01
1.94238394e-01 -1.04262817e+00 -1.11393057e-01 7.80665874e-01
-3.54665190e-01 -1.77823484e-01 1.45619079e-01 3.33390534e-01
1.99332237e-01 -7.20336065e-02 1.25834262e+00 2.25208551e-02
-8.12724382e-02 7.54227042e-01 2.28271723e-01 -2.93161392e-01
9.62808251e-01 -2.31657580e-01 -3.90210569e-01 -2.45717913e-01
-8.59292865e-01 -2.08148271e-01 3.89874101e-01 2.03359008e-01
4.69777554e-01 -1.44927454e+00 -8.59086037e-01 9.28181708e-01
-3.03316921e-01 -2.31834635e-01 2.31385436e-02 2.30293572e-01
-3.75020564e-01 -2.77350187e-01 -4.52389657e-01 -3.11867982e-01
-1.50664139e+00 9.71899152e-01 6.24966025e-01 -2.09132120e-01
-3.92016500e-01 1.16952956e+00 2.12894276e-01 -8.56145561e-01
5.54529607e-01 2.19335824e-01 -1.59710929e-01 -1.13654606e-01
5.59062183e-01 3.92847300e-01 3.04022521e-01 -3.22177202e-01
-6.82599545e-01 5.57147801e-01 -1.38265237e-01 2.68724322e-01
1.18453336e+00 1.35684893e-01 -4.34939384e-01 -5.57296157e-01
1.35216558e+00 4.03568864e-01 -1.17223287e+00 1.05910618e-02
-3.84801626e-01 -7.09740698e-01 -1.15540482e-01 -8.54685605e-01
-1.57313383e+00 8.67847204e-01 9.46642578e-01 3.97108108e-01
1.32884026e+00 -3.74918193e-01 6.71415925e-01 4.20635104e-01
4.98077959e-01 -3.32334608e-01 -3.14239770e-01 3.40146214e-01
1.11730969e+00 -1.11763942e+00 1.40692696e-01 -5.32317638e-01
-3.20390731e-01 1.23853624e+00 8.96772921e-01 -4.10760194e-01
1.33790278e+00 5.85891604e-01 2.87659675e-01 1.13352507e-01
-6.96370423e-01 3.18228394e-01 -9.28523168e-02 1.06106102e+00
-5.52375838e-02 -8.26689824e-02 9.03094709e-02 7.22208917e-01
-5.33329785e-01 1.02374516e-01 -1.39132306e-01 5.75343847e-01
-2.67377526e-01 -1.03930676e+00 -6.42407119e-01 4.02927876e-01
-8.38954508e-01 -3.24642882e-02 -5.52795053e-01 5.26052833e-01
4.34802055e-01 6.47357941e-01 -2.43928686e-01 -8.28737557e-01
1.56774268e-01 1.86404437e-01 3.02613407e-01 2.62869895e-02
-8.57137382e-01 -5.57486057e-01 -5.49489670e-02 -7.62647569e-01
-1.54715888e-02 -4.04623926e-01 -5.47991812e-01 -9.26422477e-01
-3.48384202e-01 -8.72521624e-02 7.66299129e-01 7.14094758e-01
4.42764133e-01 -8.29452500e-02 1.44697940e+00 -8.93609583e-01
-1.24258935e+00 -9.53451991e-01 -4.14225817e-01 2.20100999e-01
5.95321298e-01 -5.76271415e-01 -7.54135072e-01 -2.96872199e-01] | [5.524381160736084, 7.902141571044922] |
291661e7-233e-4765-bcd9-474a9d719d0a | twitter-sentiment-analysis | 1509.04219 | null | http://arxiv.org/abs/1509.04219v1 | http://arxiv.org/pdf/1509.04219v1.pdf | Twitter Sentiment Analysis | This project addresses the problem of sentiment analysis in twitter; that is
classifying tweets according to the sentiment expressed in them: positive,
negative or neutral. Twitter is an online micro-blogging and social-networking
platform which allows users to write short status updates of maximum length 140
characters. It is a rapidly expanding service with over 200 million registered
users - out of which 100 million are active users and half of them log on
twitter on a daily basis - generating nearly 250 million tweets per day. Due to
this large amount of usage we hope to achieve a reflection of public sentiment
by analysing the sentiments expressed in the tweets. Analysing the public
sentiment is important for many applications such as firms trying to find out
the response of their products in the market, predicting political elections
and predicting socioeconomic phenomena like stock exchange. The aim of this
project is to develop a functional classifier for accurate and automatic
sentiment classification of an unknown tweet stream. | ['Afroze Ibrahim Baqapuri'] | 2015-09-14 | null | null | null | null | ['twitter-sentiment-analysis'] | ['natural-language-processing'] | [-1.63326725e-01 1.68717146e-01 -6.74372256e-01 -4.27140534e-01
-2.32365765e-02 -7.89345086e-01 9.24389124e-01 8.58506739e-01
-4.99314874e-01 6.48511410e-01 4.73388672e-01 -5.05967736e-01
7.67049909e-01 -1.12545276e+00 -9.12081636e-03 -3.66688818e-01
1.43110633e-01 4.25828457e-01 1.03926703e-01 -9.42597449e-01
8.10732782e-01 3.27469438e-01 -1.78630173e+00 5.07752359e-01
2.10735589e-01 1.11789382e+00 -1.55241236e-01 7.04345524e-01
-6.91251457e-01 1.17136621e+00 -6.79325104e-01 -2.89859772e-01
-2.02577021e-02 -2.45859593e-01 -7.64669895e-01 -1.50252968e-01
-2.37865344e-01 2.12533370e-01 5.97660065e-01 1.11149538e+00
1.83545619e-01 -3.39885354e-01 2.47351691e-01 -1.10164845e+00
1.45047933e-01 6.51085258e-01 -6.66357338e-01 5.98251164e-01
7.95933545e-01 -3.55554461e-01 1.06167293e+00 -6.59916878e-01
6.68037474e-01 8.61363649e-01 5.05638480e-01 8.46896917e-02
-7.21489847e-01 -9.79918599e-01 1.10551268e-01 -4.27500039e-01
-8.72678876e-01 -3.58302444e-01 4.12133217e-01 -8.13984692e-01
7.57713854e-01 6.91995084e-01 8.97563398e-01 7.91857004e-01
5.70638239e-01 5.08273721e-01 1.40091980e+00 -2.72823244e-01
2.83484042e-01 1.05871105e+00 3.00369203e-01 7.79294670e-02
5.10493554e-02 -7.19693720e-01 -7.23686397e-01 -4.85721529e-01
-2.22333193e-01 6.86702132e-02 2.09774718e-01 5.74854016e-01
-1.14198112e+00 1.30790889e+00 6.49329722e-02 6.80635393e-01
-5.01903951e-01 -5.27401209e-01 7.94184089e-01 8.35701466e-01
1.40941656e+00 5.08248627e-01 -7.60740399e-01 -4.99385446e-01
-8.71948659e-01 3.27168912e-01 1.21286833e+00 3.58833015e-01
9.69831705e-01 -3.41286659e-01 6.01958513e-01 6.02414608e-01
4.31784213e-01 7.26282716e-01 1.04560471e+00 -3.68570536e-02
3.09664309e-01 1.18077624e+00 2.11856127e-01 -1.53282738e+00
-6.45956337e-01 -2.03396067e-01 -4.35803562e-01 -2.39140796e-03
2.02539802e-01 -5.25209248e-01 -2.22008035e-01 1.06273556e+00
3.76865774e-01 -4.37353760e-01 -1.21271208e-01 3.06331336e-01
7.70921588e-01 9.22108233e-01 2.60316163e-01 -4.92412508e-01
1.49370813e+00 -2.33018920e-01 -8.20101202e-01 -2.97906876e-01
7.38796592e-01 -1.14058137e+00 7.58933008e-01 4.66794491e-01
-7.40762055e-01 -1.50583073e-01 -9.10514295e-01 4.63265091e-01
-1.13640165e+00 -4.67522621e-01 7.55715370e-01 1.03855050e+00
-6.95809543e-01 6.03230894e-01 -4.70322847e-01 -3.88216406e-01
9.40129012e-02 4.89876240e-01 -2.35037938e-01 7.98195183e-01
-1.22422588e+00 7.02500403e-01 -2.00983539e-01 -3.84512067e-01
2.76431382e-01 -5.39069116e-01 -6.64779246e-01 -4.14410233e-01
-1.75299570e-01 -4.61168326e-02 1.24638164e+00 -1.56452072e+00
-1.25799716e+00 1.42695200e+00 -5.76829910e-01 -3.06792736e-01
4.60872144e-01 2.07165405e-01 -7.10381269e-01 -3.43786716e-01
5.77783883e-01 -4.07443531e-02 4.05041873e-01 -4.17376012e-01
-1.10056329e+00 -6.68329537e-01 -3.38310868e-01 -5.62894046e-02
-5.81553459e-01 7.20954001e-01 3.05917650e-01 -4.49491590e-01
7.80761167e-02 -1.04023111e+00 -1.93326965e-01 -1.13182688e+00
-2.42996767e-01 -4.62417036e-01 1.08957994e+00 -2.23508179e-01
1.27315152e+00 -2.04944849e+00 -6.11807108e-01 5.32766104e-01
5.95342591e-02 -3.66649381e-03 5.78711748e-01 9.63581085e-01
-2.26535797e-01 5.59406042e-01 4.17145073e-01 -4.11717147e-02
-9.55767035e-02 -2.66310692e-01 -7.05421925e-01 6.30632639e-01
-8.34751800e-02 3.29818010e-01 -9.00792062e-01 -2.25633204e-01
4.72364165e-02 1.59421742e-01 -5.10383807e-02 -1.66504800e-01
-1.10380657e-01 4.23311710e-01 -8.83503258e-01 4.46241498e-01
5.10910392e-01 -3.18558693e-01 2.42758915e-01 4.15493041e-01
-9.14724231e-01 6.44601166e-01 -9.41805959e-01 5.09129524e-01
-4.36706722e-01 1.22726142e+00 -1.41131550e-01 -7.21710205e-01
1.08049166e+00 2.79777884e-01 6.92527652e-01 -7.09324002e-01
7.31671572e-01 3.95656496e-01 -3.84832174e-01 -1.42865852e-01
9.61692274e-01 -5.26410043e-01 -4.80176121e-01 7.84349322e-01
-7.10262656e-01 3.81179247e-03 5.18165410e-01 2.34437227e-01
7.00018048e-01 -5.69193125e-01 4.65013415e-01 -5.42463958e-01
8.17850471e-01 2.48255625e-01 1.02951884e-01 1.10063791e-01
-2.02520415e-01 2.03748628e-01 9.27039385e-01 -7.32818365e-01
-5.66795945e-01 1.24511579e-02 -3.11219633e-01 1.41286623e+00
-8.64718296e-03 -5.81837773e-01 -5.22439897e-01 -1.40296504e-01
-2.63094157e-02 3.14154953e-01 -7.26729631e-01 5.45440435e-01
-1.97233379e-01 -1.21488452e+00 -1.62951231e-01 -1.52718753e-01
3.27468455e-01 -1.23494089e+00 -3.85624170e-01 2.34740362e-01
-3.00095707e-01 -8.39497387e-01 2.57315282e-02 1.21972874e-01
-5.28488398e-01 -1.03671491e+00 -3.17196310e-01 -7.43074298e-01
6.43305421e-01 1.81435019e-01 1.05182326e+00 -6.50613606e-02
2.56653190e-01 -1.18934577e-02 -5.83449125e-01 -1.35082579e+00
-6.35507882e-01 5.18140793e-01 9.00795311e-02 4.39498395e-01
8.65649760e-01 -2.66394794e-01 -4.37164724e-01 2.68331915e-01
-7.89364636e-01 -1.48457959e-01 -1.93700105e-01 -1.91346735e-01
1.17313467e-01 7.37193748e-02 7.50285447e-01 -1.64073813e+00
9.01895583e-01 -1.08009923e+00 -4.03757900e-01 -3.85922998e-01
-6.79707766e-01 -5.55994749e-01 4.37129050e-01 -9.31651425e-03
-6.01917326e-01 -1.23390593e-01 -2.65449762e-01 9.44850445e-01
1.60173267e-01 8.25614631e-01 6.21486664e-01 1.30242899e-01
6.01020634e-01 -1.78909048e-01 1.99611440e-01 -1.06376089e-01
-3.40444595e-01 1.31144106e+00 -1.52222499e-01 2.78158307e-01
4.29906964e-01 8.33337247e-01 -3.16453129e-01 -1.17756641e+00
-1.13761675e+00 -1.20277786e+00 -1.95315659e-01 -5.94760120e-01
6.76869512e-01 -1.09958065e+00 -1.17375112e+00 6.66615486e-01
-7.22237349e-01 1.27830759e-01 -1.25719890e-01 1.80999368e-01
-3.35084610e-02 -3.63821685e-02 -6.03747904e-01 -8.72429013e-01
-5.03469825e-01 -7.53252864e-01 6.48527265e-01 3.53955209e-01
-9.61108148e-01 -1.21869338e+00 5.12996495e-01 4.27543014e-01
5.86424887e-01 5.67304850e-01 2.29110196e-01 -1.01592410e+00
3.41474891e-01 -8.69629085e-01 4.20993194e-02 1.99616447e-01
2.10053399e-01 2.77159005e-01 -9.50053394e-01 -2.29789615e-01
2.15685904e-01 -2.43861258e-01 4.38023627e-01 2.74088860e-01
5.29831648e-01 -3.18850964e-01 -3.69663388e-01 -2.18716890e-01
1.23056102e+00 1.28579155e-01 5.27635396e-01 1.07140243e+00
1.21808294e-02 8.27366412e-01 8.76467228e-01 9.40803826e-01
3.46321315e-01 -1.15738297e-03 4.40860242e-01 5.06338887e-02
8.38640511e-01 -8.10901150e-02 6.13140047e-01 9.44642425e-01
4.61626938e-03 -1.29654616e-01 -9.77687955e-01 5.06360412e-01
-1.61264169e+00 -1.28899121e+00 -8.16576183e-01 2.04824042e+00
7.18616068e-01 4.87098157e-01 3.95973653e-01 4.78975624e-01
7.85770297e-01 4.55393940e-01 5.70283942e-02 -8.30036938e-01
-3.73069867e-02 3.25786114e-01 8.99294317e-01 4.98131037e-01
-1.00514066e+00 5.11459172e-01 6.12045574e+00 2.89837807e-01
-1.87432206e+00 -1.92871094e-02 9.37223554e-01 1.06968500e-01
-3.11626136e-01 -1.02589376e-01 -1.07516348e+00 7.54896522e-01
1.28704226e+00 -4.10291076e-01 -2.92942405e-01 8.56987894e-01
6.67341173e-01 -6.20558381e-01 -2.82482117e-01 5.19931793e-01
5.00802323e-02 -1.55858183e+00 -2.62724549e-01 3.98798257e-01
8.90324175e-01 6.35296822e-01 1.22743182e-01 6.49791583e-02
4.05265912e-02 -6.74079597e-01 8.94191682e-01 2.19395533e-01
4.52562243e-01 -8.40690911e-01 9.33510363e-01 6.62727356e-01
-8.95116210e-01 9.34029967e-02 1.13624081e-01 -8.61000180e-01
2.52328426e-01 9.16992009e-01 -8.33751857e-01 -1.61067396e-01
7.43710339e-01 9.50783253e-01 -4.76752937e-01 4.39066857e-01
2.40849763e-01 7.08269119e-01 -3.26673388e-01 -6.85116887e-01
3.54131311e-01 -3.53214532e-01 4.15159106e-01 1.44515312e+00
5.95374815e-02 -1.12593099e-01 -5.34109026e-02 -1.17616765e-01
4.34400104e-02 6.34783685e-01 -6.37430727e-01 -2.80536503e-01
-1.90715361e-02 1.45116663e+00 -1.34412956e+00 -2.93818712e-01
-3.76595318e-01 4.10523504e-01 -2.82007277e-01 -3.60380769e-01
-3.95407856e-01 -5.15380502e-01 6.07090950e-01 8.30545783e-01
5.81225822e-06 7.36298710e-02 -1.77174672e-01 -8.96345556e-01
-2.24353999e-01 -8.39923680e-01 1.38899997e-01 -3.45281541e-01
-1.24831796e+00 7.04833388e-01 -4.84795988e-01 -1.17645037e+00
-3.76205444e-01 -6.67830825e-01 -9.07823384e-01 7.76006937e-01
-1.53174806e+00 -6.36252344e-01 -9.94683355e-02 4.04663384e-01
2.09165052e-01 -3.75456065e-01 7.39363909e-01 2.11929798e-01
-1.72131106e-01 -1.54638037e-01 2.92426106e-02 9.41535756e-02
6.16276324e-01 -1.30840278e+00 4.32048231e-01 7.25054601e-03
-2.88785666e-01 5.38729966e-01 1.14033234e+00 -5.97867250e-01
-8.80581439e-01 -7.70121992e-01 1.94985175e+00 -6.47396326e-01
1.24974108e+00 -1.49973363e-01 -1.57792896e-01 5.15288711e-01
3.76868904e-01 -4.78037953e-01 1.27243626e+00 3.03966165e-01
-1.77325476e-02 -6.60674870e-02 -9.93980944e-01 3.11565399e-01
-8.56028795e-02 -5.05128443e-01 -3.12042207e-01 7.42542803e-01
1.02984823e-01 -1.98283136e-01 -7.05851197e-01 -1.09968476e-01
7.07091808e-01 -1.25581563e+00 2.27565616e-01 -3.77009958e-01
5.50754964e-01 -1.82986364e-01 1.47911340e-01 -1.16659248e+00
1.92267135e-01 -8.95137787e-01 8.09224725e-01 1.25007606e+00
8.59335840e-01 -1.19161439e+00 1.09327519e+00 6.02821946e-01
4.45850939e-01 -4.62115645e-01 -3.60470414e-01 5.99072687e-02
-2.42937401e-01 -6.57416403e-01 4.61344361e-01 1.31530130e+00
5.95227718e-01 6.42004967e-01 -4.31530736e-02 -4.48584408e-01
-8.18861127e-02 7.82146007e-02 9.11214948e-01 -1.53268445e+00
3.14931065e-01 -5.83888352e-01 -5.36419034e-01 -4.85747635e-01
-1.22910701e-01 -5.70966065e-01 -6.60496175e-01 -1.09347117e+00
-3.66360396e-02 -4.23645467e-01 1.01606727e-01 -3.32653848e-03
4.43429977e-01 7.74519086e-01 -6.86703026e-02 3.17100435e-01
-4.43518817e-01 -3.15786451e-01 9.18562531e-01 -9.03737321e-02
-3.43814582e-01 6.02508366e-01 -1.08173180e+00 8.24775994e-01
9.04138923e-01 -6.00624681e-01 3.85460146e-02 5.01466930e-01
1.71394622e+00 -3.49119127e-01 -4.06089455e-01 -4.69578356e-01
1.15869090e-01 -2.68056780e-01 9.17107835e-02 -9.07996595e-01
-7.43457824e-02 -5.72655618e-01 1.73577011e-01 4.53432322e-01
-2.11417824e-01 4.20468211e-01 -5.65295778e-02 3.59911621e-01
-7.53135741e-01 -3.01369429e-01 5.40009022e-01 -3.61354649e-01
-3.94779444e-01 -3.28151858e-03 -1.00006020e+00 -2.74577364e-02
1.22259438e+00 -3.49298090e-01 -3.42911392e-01 -7.80636370e-01
-5.69500506e-01 1.35774463e-01 4.66186315e-01 5.50508261e-01
-1.99416161e-01 -9.39243436e-01 -7.83615172e-01 2.30199724e-01
2.11864457e-01 -4.48754311e-01 -1.94275007e-01 8.17922235e-01
-8.51025701e-01 3.37401092e-01 1.38202514e-02 -2.14946926e-01
-1.48690057e+00 -1.33589789e-01 5.72400168e-02 -3.06759238e-01
-5.13866171e-02 7.64473557e-01 -2.66806215e-01 -3.47180992e-01
-2.06438422e-01 2.89640371e-02 -9.56409216e-01 1.16884422e+00
9.98082340e-01 2.73410082e-01 2.99843073e-01 -1.17666960e+00
-3.06176484e-01 2.89889991e-01 -1.09198324e-01 -1.66459680e-01
1.53005040e+00 -1.98276833e-01 -7.41000533e-01 1.01980448e+00
1.46000683e+00 6.59041882e-01 -1.32171124e-01 2.93557197e-01
3.55347425e-01 -4.75401253e-01 1.67178541e-01 -1.92852810e-01
-9.13512170e-01 2.96829313e-01 1.03831701e-01 1.39221752e+00
6.17597222e-01 -3.82970385e-02 8.55294704e-01 8.85550752e-02
1.60767451e-01 -1.51474178e+00 -1.64536312e-01 6.95255399e-01
5.26920676e-01 -1.47314191e+00 3.50267887e-01 -2.23368734e-01
-7.81114221e-01 1.11730611e+00 -1.96189076e-01 -8.94858092e-02
1.34648669e+00 2.33481824e-01 4.47618097e-01 -5.48481405e-01
-7.40449071e-01 5.18332496e-02 -1.59007221e-01 1.25480548e-01
1.07486355e+00 2.73197562e-01 -8.59755754e-01 2.09947333e-01
-9.26674485e-01 -1.48521945e-01 9.85121429e-01 9.83406544e-01
-7.52128124e-01 -1.18596911e+00 -2.78774500e-01 7.79406428e-01
-1.28665805e+00 2.23915786e-01 -6.57437325e-01 5.64319909e-01
-1.81556538e-01 1.35552061e+00 3.67119133e-01 -3.59272271e-01
2.84359485e-01 -6.69138432e-02 -4.47623223e-01 -7.07367837e-01
-1.28314722e+00 -2.94079840e-01 5.24764121e-01 -2.10461527e-01
-8.84130359e-01 -1.06815684e+00 -1.10918784e+00 -9.50545549e-01
-4.69917417e-01 6.34957612e-01 1.43698680e+00 8.34164679e-01
8.97870027e-03 -1.93417877e-01 1.42474508e+00 -7.45440364e-01
-2.59859353e-01 -9.34393942e-01 -8.93567622e-01 3.09506595e-01
5.14299273e-01 -4.57308590e-02 -7.60903001e-01 1.27827987e-01] | [10.818168640136719, 6.934776782989502] |
5e5b4938-ad83-4d94-a4dd-e1a0608c5f61 | generative-adversarial-networks-for-dental | 2307.02019 | null | https://arxiv.org/abs/2307.02019v1 | https://arxiv.org/pdf/2307.02019v1.pdf | Generative Adversarial Networks for Dental Patient Identity Protection in Orthodontic Educational Imaging | Objectives: This research introduces a novel area-preserving Generative Adversarial Networks (GAN) inversion technique for effectively de-identifying dental patient images. This innovative method addresses privacy concerns while preserving key dental features, thereby generating valuable resources for dental education and research. Methods: We enhanced the existing GAN Inversion methodology to maximize the preservation of dental characteristics within the synthesized images. A comprehensive technical framework incorporating several deep learning models was developed to provide end-to-end development guidance and practical application for image de-identification. Results: Our approach was assessed with varied facial pictures, extensively used for diagnosing skeletal asymmetry and facial anomalies. Results demonstrated our model's ability to adapt the context from one image to another, maintaining compatibility, while preserving dental features essential for oral diagnosis and dental education. A panel of five clinicians conducted an evaluation on a set of original and GAN-processed images. The generated images achieved effective de-identification, maintaining the realism of important dental features and were deemed useful for dental diagnostics and education. Clinical Significance: Our GAN model and the encompassing framework can streamline the de-identification process of dental patient images, enhancing efficiency in dental education. This method improves students' diagnostic capabilities by offering more exposure to orthodontic malocclusions. Furthermore, it facilitates the creation of de-identified datasets for broader 2D image research at major research institutions. | ['Eugene Loh', 'Kelvin Weng Chiong Foong', 'Wilson Weixun Lu', 'Mingchuan Tian'] | 2023-07-05 | null | null | null | null | ['de-identification'] | ['natural-language-processing'] | [ 4.26657379e-01 6.63877785e-01 -1.75961941e-01 -3.68778229e-01
-1.31044960e+00 -3.35955799e-01 6.31020367e-02 -1.48544744e-01
-5.02752438e-02 4.14192408e-01 2.63870448e-01 -2.82817602e-01
2.82626357e-02 -8.22726488e-01 -4.61767346e-01 -1.04370224e+00
2.29124948e-01 3.53394836e-01 -3.86976600e-01 5.29203145e-03
3.08813244e-01 7.33389556e-01 -1.84830189e+00 2.61909395e-01
8.50359142e-01 5.64227223e-01 -1.50355831e-01 6.31969571e-01
9.78191122e-02 4.44827735e-01 -9.76635814e-01 -8.98985088e-01
1.87662035e-01 -3.64475995e-01 -5.11889279e-01 3.17246877e-02
8.27799320e-01 -9.09384727e-01 -2.08603799e-01 9.69843388e-01
1.27758956e+00 -3.33878458e-01 4.62368399e-01 -1.15573466e+00
-1.12185800e+00 2.01225221e-01 -5.95824540e-01 5.85782714e-02
3.87517691e-01 3.44696492e-01 4.26861733e-01 -3.41704696e-01
7.84726739e-01 1.05817878e+00 5.37875056e-01 1.02544785e+00
-1.12679601e+00 -1.11263263e+00 -4.62218493e-01 6.61950335e-02
-1.09910834e+00 -4.52848017e-01 9.65078652e-01 -2.81153232e-01
2.65405476e-01 3.33488852e-01 1.14485478e+00 1.19409263e+00
4.06488746e-01 7.30671525e-01 1.23667502e+00 -4.42278564e-01
1.49644703e-01 1.41026303e-01 -2.76756287e-01 6.55290604e-01
3.06154579e-01 4.79655981e-01 -2.54197717e-01 -7.69997239e-02
8.41670454e-01 -1.88962698e-01 -1.32839426e-01 9.43823531e-03
-6.78686559e-01 1.10408771e+00 9.12059918e-02 1.17984414e-01
-3.16274643e-01 -3.01882159e-02 2.54558057e-01 1.63575590e-01
3.82444054e-01 2.24313978e-02 3.97136718e-01 -6.39692992e-02
-7.22894251e-01 -8.76845047e-02 4.23180312e-01 6.18771791e-01
2.68480510e-01 3.04491013e-01 1.10820666e-01 1.08293581e+00
4.54475582e-01 7.01626360e-01 7.95603156e-01 -1.18717837e+00
-1.06901839e-01 3.60815763e-01 -6.76985443e-01 -9.38416421e-01
4.55786241e-03 -2.33940527e-01 -6.09260798e-01 7.34792113e-01
5.91436774e-03 1.59313887e-01 -1.42755282e+00 1.70615232e+00
7.58090675e-01 -1.01907598e-02 9.27294120e-02 2.21815333e-01
1.10562098e+00 8.20204318e-02 2.37793520e-01 2.13007301e-01
1.70937538e+00 -4.51947629e-01 -9.07719314e-01 -7.94301834e-03
5.02346337e-01 -1.15393567e+00 9.41282272e-01 2.62346864e-01
-1.27823675e+00 -3.40413094e-01 -9.23496842e-01 -2.86294192e-01
-9.18257385e-02 -3.73676568e-02 4.76480693e-01 1.44628525e+00
-1.40263510e+00 -2.89731342e-02 -8.34301174e-01 -5.30702956e-02
1.05832577e+00 6.89541519e-01 -6.15963101e-01 -3.12084645e-01
-7.53810108e-01 7.05792665e-01 -1.36881873e-01 1.03699192e-01
-7.08964407e-01 -9.91490066e-01 -9.39564347e-01 -2.74849623e-01
-4.30034518e-01 -9.66778517e-01 1.24066544e+00 -7.77812302e-01
-1.90890849e+00 1.79346144e+00 2.23283485e-01 7.94594921e-03
7.65134454e-01 1.44026920e-01 -6.44817889e-01 4.68304694e-01
3.60851943e-01 9.68021452e-01 7.82952130e-01 -1.31437063e+00
-5.19953787e-01 -7.18204319e-01 -2.52188891e-01 -1.69428755e-02
5.28887808e-02 -2.63080269e-01 -2.76612222e-01 -1.03566802e+00
1.70857742e-01 -1.10144174e+00 -6.26082346e-02 3.41191024e-01
-4.78910118e-01 2.55079687e-01 9.21548069e-01 -9.44298267e-01
7.16603100e-01 -2.42850137e+00 -6.13043129e-01 3.81073534e-01
3.96065623e-01 3.53592068e-01 -1.51459798e-01 2.31714755e-01
-2.94077426e-01 3.00159395e-01 -2.56274045e-01 -5.24371564e-01
-2.11133733e-01 2.20671147e-01 1.35314286e-01 6.45223677e-01
-3.00726831e-01 9.21878636e-01 -6.00859702e-01 -8.02305639e-01
5.29606044e-01 1.13004947e+00 -8.69928181e-01 1.72166631e-01
3.72200519e-01 7.69641817e-01 -2.32594967e-01 9.80766654e-01
1.09517086e+00 2.29286164e-01 -6.35007918e-02 -2.83754289e-01
4.89130616e-01 -2.57011820e-02 -8.41160536e-01 1.38234067e+00
-7.39072502e-01 6.33355081e-01 3.13777655e-01 -5.12625396e-01
8.14386189e-01 5.18583715e-01 8.54891658e-01 -9.53842819e-01
3.75320613e-01 2.71845281e-01 -4.44099791e-02 -7.57971227e-01
-3.74919735e-02 -4.13104147e-01 2.10940003e-01 5.92753053e-01
-1.64784357e-01 -5.80160975e-01 -2.51713902e-01 -4.79646735e-02
3.64840388e-01 -2.40751803e-01 -1.24658933e-02 3.40576991e-02
3.91196430e-01 -4.59823221e-01 1.69362947e-01 2.25431383e-01
-3.45373154e-01 8.76284719e-01 4.83466685e-02 -1.32448196e-01
-7.62923598e-01 -1.38329279e+00 -4.38795567e-01 4.31959063e-01
-2.82267362e-01 1.57865390e-01 -9.21596944e-01 -4.82429951e-01
-1.30407840e-01 8.51876378e-01 -8.43304873e-01 -4.39161807e-01
-6.38494313e-01 -4.87217516e-01 7.68941581e-01 2.57491380e-01
3.52975637e-01 -1.03459215e+00 -5.20781696e-01 9.95869264e-02
-2.07411885e-01 -7.03135014e-01 -7.52972782e-01 -4.79345739e-01
-8.41898620e-01 -1.18928695e+00 -9.38610017e-01 -1.43521333e+00
1.07112217e+00 -1.81653410e-01 8.01374078e-01 5.98513484e-02
-8.91790807e-01 8.29130054e-01 1.33910447e-01 -6.65746689e-01
-1.14781487e+00 -2.02765048e-01 -3.64928365e-01 -2.78323621e-01
4.20266002e-01 -8.77341866e-01 -1.18267691e+00 2.01688349e-01
-1.33401346e+00 -3.88882339e-01 5.70394635e-01 8.37207377e-01
7.32975841e-01 -1.20604143e-01 6.03000224e-01 -9.56445813e-01
6.59772158e-01 -2.31075689e-01 -4.10297513e-01 2.03202605e-01
-7.76367188e-01 -1.13961026e-01 1.04916967e-01 -3.63836914e-01
-1.23349535e+00 -2.34732211e-01 -7.25436568e-01 4.88606058e-02
-1.14447288e-01 -8.88745263e-02 -3.74319851e-01 -4.53592479e-01
3.78950328e-01 2.57838905e-01 6.79178596e-01 -2.04492509e-01
2.94622302e-01 8.65352094e-01 8.72304738e-01 -4.77355093e-01
5.57440460e-01 6.99780822e-01 -6.09976202e-02 -7.08729625e-01
-1.83045551e-01 -4.65469994e-02 -4.13728684e-01 -1.81197494e-01
9.61842120e-01 -9.25196946e-01 -9.66873527e-01 8.63061905e-01
-5.37445188e-01 2.98798293e-01 -4.71739858e-01 5.36909521e-01
-3.15738052e-01 4.67669338e-01 -6.96441829e-01 -4.21201169e-01
-7.03480840e-01 -1.63745749e+00 1.28652883e+00 3.53669912e-01
-3.22493613e-01 -1.23913538e+00 3.99602443e-01 1.05833137e+00
3.61079842e-01 5.73039353e-01 1.11854529e+00 -2.56632864e-01
-3.13863665e-01 -4.97502238e-01 3.78808498e-01 4.78175372e-01
6.66932166e-01 2.40576252e-01 -1.25311136e+00 -3.59370857e-01
3.07379305e-01 -7.49870911e-02 3.86337399e-01 6.19920552e-01
1.11755168e+00 -4.15207297e-01 -2.26989999e-01 9.36660409e-01
1.08165026e+00 8.37527335e-01 1.18381417e+00 4.23569024e-01
6.41077042e-01 8.20480466e-01 4.03651483e-02 -3.75887901e-02
4.78934169e-01 3.78421009e-01 4.09240335e-01 -4.91362840e-01
-5.72132170e-01 -4.28608805e-01 1.36499569e-01 7.59603024e-01
1.76263019e-01 -1.86616201e-02 -6.10816717e-01 7.66770780e-01
-5.76182604e-01 -8.18641424e-01 4.14539903e-01 2.01787281e+00
8.40156555e-01 -4.37225282e-01 9.59782675e-02 1.53312966e-01
9.26813483e-01 -2.37128839e-01 -7.44607568e-01 -5.13473868e-01
-7.93829001e-03 9.51442063e-01 3.88673276e-01 4.10034537e-01
-6.74312055e-01 5.50668418e-01 7.02502966e+00 6.53129637e-01
-1.43863809e+00 -7.29252547e-02 9.66140270e-01 -1.28093809e-01
-8.22381973e-01 -5.41791618e-01 -4.48365331e-01 5.01286387e-01
7.67492175e-01 -2.99687862e-01 -2.63029456e-01 9.12340164e-01
-4.62075835e-03 2.52088279e-01 -8.25376570e-01 9.67216849e-01
1.87082320e-01 -1.37654638e+00 3.22884023e-01 5.37917078e-01
7.59268403e-01 -5.33346653e-01 8.52824509e-01 -1.63041160e-01
7.34977245e-01 -1.29723430e+00 2.31725022e-01 -9.51681808e-02
1.26016927e+00 -8.47122014e-01 6.33126438e-01 -6.40137792e-01
-6.15226626e-01 2.39624873e-01 3.18297833e-01 7.70594716e-01
2.56781757e-01 4.63394225e-01 -1.28927612e+00 2.14906052e-01
4.59488004e-01 2.65860736e-01 -2.92589158e-01 7.78089523e-01
-3.23107809e-01 4.55657333e-01 -3.33807021e-01 4.24344063e-01
-1.34441853e-01 -2.00988144e-01 4.15017217e-01 5.34399390e-01
4.61559445e-01 1.38666272e-01 -5.26548564e-01 4.74807471e-01
-9.74635780e-02 2.85445035e-01 -4.82528567e-01 2.63756245e-01
6.78753555e-01 7.94913828e-01 -6.49662673e-01 1.56792790e-01
-7.58872926e-02 7.57642567e-01 -4.20840412e-01 -3.55274603e-02
-6.01053715e-01 -4.06947918e-02 8.69308114e-01 2.85428733e-01
2.72134513e-01 4.08665001e-01 -3.44153613e-01 -6.87897980e-01
-1.79226637e-01 -1.29054761e+00 6.20285571e-01 -6.32034957e-01
-1.13831043e+00 4.88345236e-01 -5.85124493e-02 -1.16475296e+00
-4.89910364e-01 -3.08721006e-01 -6.64508522e-01 4.77210134e-01
-1.34352553e+00 -1.78040099e+00 -2.53755927e-01 7.07445264e-01
1.83232248e-01 -2.20654964e-01 1.01595521e+00 6.14187062e-01
-5.90006709e-01 1.23682189e+00 2.35338286e-01 8.77205506e-02
6.42955720e-01 -9.43609655e-01 8.78650919e-02 5.89148700e-01
-3.10899434e-03 8.51105750e-01 4.33632702e-01 -5.41181445e-01
-1.02893221e+00 -6.92238510e-01 3.50522995e-01 -1.59661502e-01
-2.81996522e-02 1.94574445e-01 -4.51420516e-01 8.59654665e-01
2.42101893e-01 -4.29764867e-01 1.45328760e+00 -6.34963453e-01
-1.53220907e-01 -2.50859112e-01 -2.02648854e+00 6.84775889e-01
5.46606004e-01 -7.33046710e-01 -4.21228319e-01 2.46800557e-01
3.31807107e-01 -6.81180239e-01 -1.27241147e+00 1.64642334e-01
1.00651515e+00 -8.39279592e-01 1.36241972e+00 -5.92876822e-02
7.00276911e-01 1.60491824e-01 5.24277762e-02 -1.13768852e+00
1.47963256e-01 -7.29162157e-01 5.11658132e-01 1.42168415e+00
1.19992353e-01 -1.02993214e+00 1.20244467e+00 7.32248902e-01
-1.33610561e-01 -5.48715413e-01 -1.29532206e+00 -2.77403325e-01
4.14606154e-01 -2.34402686e-01 8.46839130e-01 8.35416973e-01
-5.58420002e-01 -3.50564063e-01 -1.86738566e-01 1.23235203e-01
8.59641731e-01 -1.44509956e-01 7.35048711e-01 -8.00083578e-01
7.43544102e-02 -6.29055738e-01 -8.50837827e-01 -4.12217289e-01
5.30368164e-02 -9.67846572e-01 -6.08119667e-01 -1.29129100e+00
-1.67867318e-01 -3.88767242e-01 -4.75510918e-02 4.47965413e-01
-7.86538124e-02 6.88570619e-01 -1.38809487e-01 1.60857990e-01
9.03597713e-01 3.45369756e-01 1.81607866e+00 -2.61713803e-01
-2.31763478e-02 3.56925309e-01 -1.09054279e+00 3.16842347e-01
7.43723273e-01 -5.84793627e-01 -7.61468828e-01 -3.07955116e-01
-4.64132458e-01 -2.27026880e-01 4.05472815e-01 -7.15500772e-01
3.38688828e-02 2.49939993e-01 1.16431564e-01 -5.38570523e-01
3.04677635e-01 -9.89727676e-01 6.24527335e-01 9.51711357e-01
-5.92708923e-02 5.56450412e-02 4.89136994e-01 1.08004428e-01
-4.44719464e-01 9.20374990e-02 1.14060843e+00 -1.56237096e-01
-4.08773631e-01 1.94579601e-01 -1.58594474e-01 -5.02455533e-02
1.39979994e+00 -7.70591497e-01 -1.76514521e-01 -8.17475677e-01
-8.74400377e-01 -3.02276939e-01 6.36954665e-01 1.78487957e-01
8.17461550e-01 -1.13620782e+00 -7.54865527e-01 7.36814380e-01
-3.94710563e-02 1.34968832e-01 8.28016818e-01 5.32444835e-01
-1.32146823e+00 -6.98157698e-02 -5.03474534e-01 -7.77985036e-01
-1.72739863e+00 1.86480463e-01 4.11590248e-01 -1.38749287e-01
-7.63292551e-01 1.03206992e+00 3.54850590e-01 -5.00721633e-01
9.16921198e-02 -2.75971472e-01 1.76726840e-02 6.87921420e-02
4.66525674e-01 2.91644305e-01 1.70553863e-01 -7.32123137e-01
-2.27433190e-01 8.96082044e-01 -4.52811271e-01 -3.89779173e-02
1.48369002e+00 -3.13418478e-01 5.68760596e-02 -3.54253471e-01
1.63458884e+00 4.76541221e-01 -9.00535583e-01 3.67395103e-01
-7.49832630e-01 -6.75305009e-01 2.11521059e-01 -9.53605533e-01
-1.95131993e+00 7.73570597e-01 1.38126767e+00 -3.28379154e-01
1.48588157e+00 -1.25355333e-01 1.10393298e+00 -6.45912826e-01
1.09188206e-01 -7.72197604e-01 1.62762776e-01 -5.57663441e-01
6.43013120e-01 -1.07864547e+00 3.50687392e-02 -6.23367250e-01
-4.45453674e-01 1.05866241e+00 3.20935369e-01 3.00088555e-01
8.06005538e-01 5.33162236e-01 8.83384943e-01 -5.51344514e-01
2.48131707e-01 6.27150357e-01 1.86124459e-01 1.26015007e+00
2.57757723e-01 -9.04235542e-02 -1.72825992e-01 -1.99380934e-01
-9.22085762e-01 7.27151185e-02 3.64223897e-01 1.09745419e+00
3.68660837e-01 -1.54734671e+00 -5.29291630e-01 1.60545439e-01
-8.98095965e-01 -9.92353782e-02 -6.02169894e-02 1.10019124e+00
1.45287141e-01 8.29132080e-01 -1.76316295e-02 -2.53860783e-02
1.77011862e-01 -3.57773125e-01 5.04810154e-01 -3.78191888e-01
-6.87236190e-01 -7.37148672e-02 -1.20340489e-01 -2.33481750e-01
-4.21585321e-01 -7.77955830e-01 -7.88606763e-01 -6.25583589e-01
-1.28842533e-01 -2.51128018e-01 1.00158846e+00 5.15848875e-01
5.64513147e-01 6.58307850e-01 7.61835694e-01 -2.79636055e-01
-2.87241936e-01 -5.76418459e-01 -5.11961162e-01 6.18701279e-01
4.73244399e-01 -4.05545533e-01 -5.40506601e-01 1.94460079e-01] | [13.906986236572266, -1.7960929870605469] |
16ee1030-fd08-4a3f-b9f9-f7a5d9edd788 | 3d-multi-object-tracking-based-on-uncertainty | 2303.01786 | null | https://arxiv.org/abs/2303.01786v1 | https://arxiv.org/pdf/2303.01786v1.pdf | 3D Multi-Object Tracking Based on Uncertainty-Guided Data Association | In the existing literature, most 3D multi-object tracking algorithms based on the tracking-by-detection framework employed deterministic tracks and detections for similarity calculation in the data association stage. Namely, the inherent uncertainties existing in tracks and detections are overlooked. In this work, we discard the commonly used deterministic tracks and deterministic detections for data association, instead, we propose to model tracks and detections as random vectors in which uncertainties are taken into account. Then, based on the Jensen-Shannon divergence, the similarity between two multidimensional distributions, i.e. track and detection, is evaluated for data association purposes. Lastly, the level of track uncertainty is incorporated in our cost function design to guide the data association process. Comparative experiments have been conducted on two typical datasets, KITTI and nuScenes, and the results indicated that our proposed method outperformed the compared state-of-the-art 3D tracking algorithms. For the benefit of the community, our code has been made available at https://github.com/hejiawei2023/UG3DMOT. | ['Xiyang Wang', 'Chunyun Fu', 'JiaWei He'] | 2023-03-03 | null | null | null | null | ['3d-multi-object-tracking'] | ['computer-vision'] | [-3.57178092e-01 -6.38773024e-01 -6.36194646e-02 -6.81387782e-02
-5.20912111e-01 -7.50596523e-01 6.00844204e-01 2.49127746e-01
-3.20147902e-01 7.51432598e-01 -2.15017781e-01 -1.24032609e-01
-4.28560078e-01 -7.38803804e-01 -5.71734071e-01 -8.52730393e-01
3.15438434e-02 4.76699978e-01 5.97561240e-01 3.03817511e-01
2.20628634e-01 6.57462537e-01 -1.77112949e+00 -5.24014115e-01
6.62172616e-01 1.02252316e+00 1.72875360e-01 3.62473428e-01
-2.08019465e-01 3.69403899e-01 -4.32844222e-01 -2.91079551e-01
4.88748640e-01 -1.12098120e-01 7.54373372e-02 1.82420183e-02
2.68963993e-01 -3.21828038e-01 -3.01270247e-01 1.22308338e+00
6.48901641e-01 8.27155337e-02 6.78735912e-01 -1.40287101e+00
-3.83116037e-01 3.08775514e-01 -8.78905714e-01 3.77465695e-01
1.50291264e-01 5.78921139e-02 7.11984456e-01 -9.24601257e-01
4.29960966e-01 1.14164543e+00 9.10519123e-01 1.13798521e-01
-9.52190518e-01 -1.07444453e+00 2.85307616e-01 -2.38476843e-02
-1.84493709e+00 -3.37046295e-01 5.87574601e-01 -7.93379068e-01
3.53406906e-01 3.20556134e-01 6.41670763e-01 5.40712833e-01
4.17159021e-01 6.57920122e-01 8.58453929e-01 -3.41605306e-01
-7.11180735e-03 1.44206464e-01 1.57925352e-01 4.77843672e-01
9.13443506e-01 5.78607857e-01 -2.26220325e-01 -2.48175398e-01
6.34168744e-01 9.56730768e-02 2.65711016e-04 -5.67633271e-01
-1.06914616e+00 7.09618747e-01 1.84579089e-01 8.49906281e-02
-3.10661167e-01 -1.15170978e-01 3.33463311e-01 -4.57393117e-02
5.15102565e-01 -3.88555616e-01 7.18701929e-02 1.19530715e-01
-7.47239351e-01 5.29505491e-01 3.31837744e-01 1.30288589e+00
5.03416121e-01 -8.41446444e-02 -3.14249873e-01 4.29760963e-01
7.78870463e-01 9.03557539e-01 -1.31099746e-01 -4.44883525e-01
3.33435535e-01 5.56472778e-01 6.21027052e-01 -9.90373373e-01
-3.77620071e-01 -5.43370605e-01 -6.16210222e-01 3.58481288e-01
4.50514674e-01 -2.29368776e-01 -7.70125389e-01 1.34619474e+00
8.37812662e-01 3.92267197e-01 -2.93308385e-02 1.03918350e+00
7.45328307e-01 2.45043308e-01 -2.36880034e-02 -3.31513852e-01
1.19578385e+00 -4.83309180e-01 -9.72757697e-01 1.88330218e-01
3.34838867e-01 -1.19272614e+00 4.05957967e-01 6.19134642e-02
-8.79105568e-01 -7.61758268e-01 -1.08666027e+00 5.04922211e-01
-2.39349827e-01 3.27913940e-01 4.32802588e-01 6.78860307e-01
-5.34120262e-01 3.92000586e-01 -8.74702990e-01 -4.89459485e-01
2.48813003e-01 1.18292578e-01 2.08986565e-01 1.81464389e-01
-1.07995236e+00 8.55721891e-01 5.29321551e-01 2.57523239e-01
-7.32081592e-01 -5.48573256e-01 -4.60016251e-01 -5.05626559e-01
5.69848001e-01 -4.78854150e-01 1.17900670e+00 -4.49895784e-02
-1.09566927e+00 6.57737911e-01 5.27681187e-02 -2.12761477e-01
7.25975335e-01 -3.53372216e-01 -7.24782765e-01 -4.28787470e-01
2.12983981e-01 1.01812840e-01 4.35648412e-01 -1.38280141e+00
-1.05901575e+00 -2.11630464e-01 -2.60320529e-02 1.80699974e-01
1.87743798e-01 7.77447596e-02 -7.05313087e-01 -6.25816345e-01
2.72687703e-01 -1.11064625e+00 6.78528249e-02 2.68180251e-01
-3.98836821e-01 -1.59529880e-01 7.99578428e-01 -2.59137899e-01
1.41668415e+00 -2.20496202e+00 -1.38536856e-01 2.74216712e-01
1.02678664e-01 1.15286157e-01 4.16551530e-01 4.17090595e-01
4.78211045e-01 -2.37490773e-01 5.72662577e-02 -2.81612575e-01
1.54685363e-01 -5.21173663e-02 1.62474532e-02 8.80986273e-01
-1.06992409e-01 2.93765306e-01 -9.47120786e-01 -6.43863320e-01
5.44954062e-01 3.66358876e-01 2.85502169e-02 6.86978325e-02
-1.28562197e-01 3.85883391e-01 -7.90247619e-01 9.39405560e-01
1.07558429e+00 1.96976047e-02 -1.68395653e-01 -2.90247768e-01
-7.59122133e-01 -1.66842565e-01 -1.70729601e+00 1.46200633e+00
7.94530138e-02 2.61298627e-01 -2.99711134e-02 -2.87402749e-01
1.06785476e+00 3.10024738e-01 7.59024203e-01 -2.50606090e-01
4.08210993e-01 2.47396559e-01 6.73850104e-02 4.13851216e-02
7.17643440e-01 2.25851852e-02 5.38081564e-02 3.74773324e-01
-3.85766119e-01 -1.32760629e-02 2.85497189e-01 8.70763063e-02
6.32958829e-01 4.42994356e-01 2.71800965e-01 -3.22392434e-01
4.76235718e-01 3.77005428e-01 7.38198996e-01 8.33879948e-01
-3.41665059e-01 4.40196276e-01 -2.59618342e-01 -6.98661711e-03
-9.03031707e-01 -1.38878953e+00 -5.30016720e-01 5.06681740e-01
7.05964386e-01 -2.12752119e-01 -1.90690398e-01 -3.75254393e-01
5.29500484e-01 5.05655825e-01 -4.61289555e-01 7.82224238e-02
-7.58698285e-02 -6.50095701e-01 5.35094798e-01 4.73964900e-01
2.06257388e-01 -5.22381306e-01 -4.73565459e-01 4.00494844e-01
2.81544805e-01 -8.78458381e-01 -6.03998780e-01 -1.16431981e-01
-7.55394697e-01 -1.16140938e+00 -6.53659105e-01 -3.11479539e-01
6.13859892e-01 5.68284094e-01 6.73045337e-01 1.15764119e-01
-3.59266967e-01 2.23768339e-01 -5.26212692e-01 -8.50542724e-01
-1.20186284e-01 -3.49883556e-01 5.48268557e-01 -2.27033541e-01
5.69301844e-01 -1.85204074e-01 -4.54760253e-01 7.68913150e-01
-4.74484622e-01 -1.35016844e-01 4.88349557e-01 4.96533900e-01
9.38989162e-01 2.58996964e-01 3.29215527e-01 -4.46581185e-01
3.53199482e-01 -5.12323081e-01 -1.29823029e+00 1.44328877e-01
-5.50777256e-01 -1.04837745e-01 1.09390542e-01 -5.76198161e-01
-1.03486657e+00 1.19759575e-01 -7.43078440e-03 -7.63767183e-01
-3.14426748e-03 2.33758524e-01 -1.77130848e-01 -2.69314408e-01
3.09386164e-01 1.26983553e-01 -7.11421594e-02 -4.97731298e-01
1.60701752e-01 6.70020223e-01 3.33656162e-01 -4.82930273e-01
1.07942557e+00 6.14558816e-01 1.13795832e-01 -4.71289903e-01
-7.85491765e-01 -6.96520567e-01 -6.69817328e-01 -6.75419629e-01
7.31310487e-01 -9.99225140e-01 -7.74422944e-01 5.22770107e-01
-9.15759027e-01 4.12921697e-01 -1.03891745e-01 9.34755385e-01
-2.37627357e-01 3.67782176e-01 -1.68511704e-01 -1.35780478e+00
-1.84732750e-01 -1.02373409e+00 8.07893097e-01 4.94031638e-01
7.48584196e-02 -8.37083042e-01 2.29506373e-01 -7.30894729e-02
2.53803968e-01 3.94863814e-01 2.65150547e-01 -4.62657988e-01
-7.72705913e-01 -5.75885475e-01 -2.77336687e-01 -1.66573331e-01
4.15871561e-01 3.02345306e-01 -7.58788109e-01 -4.03724402e-01
-1.75812259e-01 1.91496730e-01 4.20443356e-01 5.78968167e-01
6.93853021e-01 2.21746206e-01 -6.97718322e-01 1.83227628e-01
1.42895341e+00 5.22916496e-01 2.94841200e-01 3.92932862e-01
5.48879504e-01 2.92095870e-01 1.24353218e+00 8.39007795e-01
3.67908180e-01 8.28257322e-01 4.88402694e-01 2.44977012e-01
-8.32862854e-02 -2.14049384e-01 1.23845913e-01 6.62589610e-01
-1.18495479e-01 -4.93678629e-01 -7.82801628e-01 4.24487233e-01
-2.08218598e+00 -1.03426337e+00 -5.61106682e-01 2.51914191e+00
3.89984757e-01 5.51164150e-01 2.38694653e-01 -9.76264775e-02
1.16164911e+00 3.31969522e-02 -5.83557189e-01 3.84036839e-01
-3.63582708e-02 -5.33087790e-01 9.85331297e-01 2.64251739e-01
-1.25923693e+00 4.93732035e-01 5.55101776e+00 8.77557635e-01
-7.12472618e-01 7.59932697e-02 -2.75404602e-01 -3.20943534e-01
1.17447572e-02 1.54630961e-02 -1.43428326e+00 6.63968325e-01
5.58447123e-01 -4.16851819e-01 -5.59842028e-02 5.16782880e-01
3.83564413e-01 -2.96970636e-01 -7.52109885e-01 8.09392393e-01
-3.84693712e-01 -1.13451111e+00 -2.51963109e-01 1.87346891e-01
5.34216702e-01 2.33074054e-02 -6.43354580e-02 9.14625227e-02
5.12800276e-01 -3.07883501e-01 1.05365896e+00 9.20252442e-01
3.57571840e-01 -7.10019469e-01 8.64806950e-01 3.01616043e-01
-1.79760468e+00 1.55758813e-01 -4.29263502e-01 8.31597969e-02
5.29568255e-01 7.11776733e-01 -4.71464962e-01 1.15775228e+00
6.95354521e-01 6.64365649e-01 -2.53443837e-01 1.62978482e+00
1.44063801e-01 2.34978870e-01 -5.93121052e-01 -1.48247868e-01
4.24312316e-02 -2.33700231e-01 9.00285363e-01 8.72116566e-01
6.71947002e-01 5.87603673e-02 5.09084880e-01 7.36944973e-01
3.10956985e-01 -9.47756544e-02 -4.91916239e-01 7.00539425e-02
1.00476873e+00 1.08621633e+00 -7.36498654e-01 -1.01735018e-01
-6.61129713e-01 2.38652110e-01 -2.71024317e-01 -1.51222125e-01
-1.26242518e+00 -2.38678217e-01 7.38433480e-01 9.91612002e-02
5.20346701e-01 -2.15635940e-01 -1.39009506e-01 -8.34924877e-01
5.33382297e-02 -2.24467784e-01 5.26862323e-01 -5.78039944e-01
-1.36197042e+00 3.94786149e-01 3.57400954e-01 -1.95933986e+00
8.71524587e-02 -3.82360041e-01 -5.62770307e-01 9.60469723e-01
-1.10497653e+00 -1.05412316e+00 -3.38729650e-01 2.91682899e-01
2.77035713e-01 -5.63213639e-02 2.84818977e-01 8.45169783e-01
-6.44303918e-01 4.26589936e-01 6.06146216e-01 -8.18320960e-02
6.62347019e-01 -8.23690891e-01 2.26482391e-01 1.04285145e+00
-1.56721517e-01 4.85755622e-01 1.02612412e+00 -1.10999954e+00
-1.48561180e+00 -1.13925087e+00 4.23942059e-01 -3.38700384e-01
9.58685100e-01 -2.40069538e-01 -6.84059262e-01 5.34453571e-01
-1.31206229e-01 7.56308138e-02 4.91307378e-01 -1.09560534e-01
2.30342358e-01 -2.21491493e-02 -1.15200973e+00 3.27494293e-01
1.20306599e+00 -9.30974856e-02 -4.09161717e-01 2.00405270e-01
4.96347666e-01 -7.00535357e-01 -1.08122349e+00 4.87258971e-01
8.88956189e-01 -5.21661878e-01 9.40064847e-01 2.41850335e-02
-4.38309997e-01 -1.22179782e+00 -2.97100753e-01 -1.00639701e+00
-5.37114441e-01 1.36069655e-02 -2.52955198e-01 1.53572261e+00
1.45630732e-01 -5.08519948e-01 5.54350436e-01 1.12935908e-01
-2.47824058e-01 -2.73495793e-01 -8.88441145e-01 -1.22595859e+00
-3.80561411e-01 -2.79006332e-01 7.67055392e-01 4.83263671e-01
-7.07861066e-01 -1.90328851e-01 -5.56960285e-01 7.22507417e-01
1.22425258e+00 3.36547434e-01 7.96715319e-01 -1.53494155e+00
-1.72528759e-01 -2.40370527e-01 -3.86346728e-01 -9.71762359e-01
-3.02613944e-01 -6.76895559e-01 2.23719135e-01 -1.34664392e+00
1.38090938e-01 -7.12847769e-01 -2.86364883e-01 2.47225724e-02
-4.84917127e-02 7.68432468e-02 2.04759106e-01 5.46734333e-01
-6.28178775e-01 5.59144914e-01 1.15950871e+00 -1.09665804e-01
-1.25814587e-01 4.18203950e-01 -3.70625198e-01 5.69363177e-01
6.85752869e-01 -8.17260325e-01 -1.27617180e-01 -3.85239720e-01
-3.39005709e-01 -1.28424287e-01 3.33536863e-01 -1.22202873e+00
4.78646934e-01 -1.74759611e-01 3.39576781e-01 -1.47455490e+00
3.18294466e-01 -1.07115710e+00 8.66586149e-01 4.85052377e-01
1.89421996e-01 1.16701342e-01 5.15844285e-01 7.41304159e-01
3.94544154e-02 -2.82713085e-01 5.00767887e-01 1.98929280e-01
-7.85669625e-01 4.87867951e-01 -2.35626563e-01 -2.75935084e-01
1.40904593e+00 -4.82758045e-01 -3.36545140e-01 1.01496764e-01
-5.35367668e-01 5.86768389e-01 5.28353453e-01 4.67719018e-01
3.93454105e-01 -1.75758386e+00 -7.27118194e-01 -8.59982669e-02
3.75649482e-01 -7.07555488e-02 3.48767012e-01 1.06858587e+00
-2.31487408e-01 4.55248058e-01 -1.00784481e-01 -9.34200466e-01
-1.39466834e+00 5.05795300e-01 1.18864685e-01 1.62054464e-01
-5.30783474e-01 4.17126834e-01 -9.94347036e-02 -3.35590094e-01
4.01425630e-01 -2.63767213e-01 -2.31171660e-02 1.21326447e-01
4.27421927e-01 4.78984177e-01 -1.14277013e-01 -7.80788541e-01
-6.86034799e-01 7.52720356e-01 -1.33302733e-02 1.46915153e-01
9.59399164e-01 -4.29494053e-01 4.32706237e-01 6.36546195e-01
5.26380897e-01 1.43880397e-01 -1.42869854e+00 -3.07779759e-01
1.74899951e-01 -7.79710591e-01 1.74043447e-01 -5.32249749e-01
-8.54872525e-01 3.44788522e-01 1.03859544e+00 2.16392726e-01
7.57359445e-01 -5.17485570e-03 3.58374447e-01 -4.37478498e-02
6.84935331e-01 -8.20348084e-01 -6.32478476e-01 4.65327442e-01
4.47428346e-01 -1.28341031e+00 3.95251542e-01 -4.50073928e-01
-4.56119180e-01 8.01772356e-01 6.31152332e-01 2.52295081e-02
6.98729575e-01 5.44486642e-01 2.92124227e-02 -9.11492035e-02
-3.54868799e-01 -5.97467601e-01 4.02925313e-01 4.32562470e-01
3.39639664e-01 -2.20695194e-02 -4.96093452e-01 3.75171930e-01
1.82552084e-01 1.74204871e-01 2.27466464e-01 1.26844120e+00
-5.64835608e-01 -9.37702954e-01 -8.87801170e-01 4.98508602e-01
-1.28776357e-01 2.69075096e-01 8.02123696e-02 1.05590570e+00
2.68521667e-01 9.53716099e-01 1.17752612e-01 -5.32308221e-01
6.96583986e-01 -3.20017070e-01 4.82071966e-01 -4.08586025e-01
-3.14508915e-01 2.87685871e-01 4.77929041e-02 -1.94700554e-01
-6.13656819e-01 -1.04935730e+00 -1.27644598e+00 -5.04715323e-01
-9.84977186e-01 2.75140911e-01 6.17846966e-01 7.32281029e-01
2.49091744e-01 6.07457399e-01 4.76140797e-01 -8.71429741e-01
-5.42225420e-01 -8.08617353e-01 -7.94273973e-01 -8.03897902e-03
1.25823632e-01 -1.54113686e+00 -3.80570501e-01 -2.86206573e-01] | [6.545280456542969, -2.0419468879699707] |
9937a907-790e-4cfc-9ccc-c10e525847f1 | face-evolve-a-high-performance-face | 2107.08621 | null | https://arxiv.org/abs/2107.08621v4 | https://arxiv.org/pdf/2107.08621v4.pdf | Face.evoLVe: A High-Performance Face Recognition Library | In this paper, we develop face.evoLVe -- a comprehensive library that collects and implements a wide range of popular deep learning-based methods for face recognition. First of all, face.evoLVe is composed of key components that cover the full process of face analytics, including face alignment, data processing, various backbones, losses, and alternatives with bags of tricks for improving performance. Later, face.evoLVe supports multi-GPU training on top of different deep learning platforms, such as PyTorch and PaddlePaddle, which facilitates researchers to work on both large-scale datasets with millions of images and low-shot counterparts with limited well-annotated data. More importantly, along with face.evoLVe, images before & after alignment in the common benchmark datasets are released with source codes and trained models provided. All these efforts lower the technical burdens in reproducing the existing methods for comparison, while users of our library could focus on developing advanced approaches more efficiently. Last but not least, face.evoLVe is well designed and vibrantly evolving, so that new face recognition approaches can be easily plugged into our framework. Note that we have used face.evoLVe to participate in a number of face recognition competitions and secured the first place. The version that supports PyTorch is publicly available at https://github.com/ZhaoJ9014/face.evoLVe.PyTorch and the PaddlePaddle version is available at https://github.com/ZhaoJ9014/face.evoLVe.PyTorch/tree/master/paddle. Face.evoLVe has been widely used for face analytics, receiving 2.4K stars and 622 forks. | ['Jian Zhao', 'Haoyi Xiong', 'Pengfei Zhang', 'Qingzhong Wang'] | 2021-07-19 | null | null | null | null | ['face-alignment'] | ['computer-vision'] | [-5.00837624e-01 -2.87978500e-01 6.41083391e-03 -4.89153713e-01
-6.04762375e-01 -4.48722363e-01 3.36709023e-01 -3.38684857e-01
-1.72329426e-01 2.59695023e-01 -6.39524683e-02 -1.90372586e-01
4.87076864e-02 -5.66064954e-01 -5.85516989e-01 -6.43070281e-01
-2.00964019e-01 6.43096983e-01 -3.19662631e-01 -3.05459321e-01
3.17616127e-02 5.71810544e-01 -1.96423841e+00 2.31373236e-01
2.88622826e-01 9.87139523e-01 -2.01062262e-01 7.92571604e-01
2.16389090e-01 2.89836019e-01 -2.78555065e-01 -1.09400618e+00
5.64910650e-01 -1.53948441e-01 -6.43375576e-01 -2.78791130e-01
9.29879606e-01 -5.99869609e-01 -4.36354518e-01 8.16485882e-01
1.14668691e+00 -2.35093221e-01 1.96573451e-01 -1.73501325e+00
-4.53240335e-01 8.88670534e-02 -7.97456622e-01 4.18547004e-01
3.46097380e-01 5.29417872e-01 5.20834327e-01 -1.34646511e+00
3.91112834e-01 1.24642408e+00 1.01906502e+00 8.69883835e-01
-9.03659999e-01 -9.64945138e-01 -3.19903642e-01 2.09773406e-01
-1.47626269e+00 -1.34031403e+00 4.77867424e-01 -3.88389707e-01
1.06637144e+00 3.75711292e-01 6.32740438e-01 1.39269292e+00
-1.15883671e-01 7.22496033e-01 8.02586019e-01 -2.20952570e-01
-1.49455935e-01 -1.64746106e-01 2.38504648e-01 1.22234607e+00
6.75491020e-02 1.33497372e-01 -6.87916160e-01 -4.79100049e-01
6.68122172e-01 -7.75291473e-02 -2.42737696e-01 -9.33329687e-02
-7.67481089e-01 6.00580812e-01 -7.77742118e-02 1.07204914e-01
-1.09129854e-01 7.28975385e-02 6.90827608e-01 4.66856748e-01
5.40599465e-01 -7.64370561e-02 -4.83373761e-01 -5.64562976e-01
-9.08257544e-01 3.50104332e-01 1.01904833e+00 1.06949341e+00
7.89474785e-01 1.20476253e-01 -8.64960477e-02 1.11432457e+00
2.08721444e-01 3.18820655e-01 2.93383807e-01 -1.10186744e+00
1.77399173e-01 3.12096626e-01 -3.55418801e-01 -8.17139387e-01
-3.38463366e-01 -2.09592134e-01 -8.35878074e-01 3.58316630e-01
4.21098381e-01 -3.35021853e-01 -7.81888068e-01 1.55563331e+00
6.39360487e-01 5.98082066e-01 -4.12564307e-01 7.75268793e-01
1.26525307e+00 3.30280870e-01 -2.93531269e-01 1.14977963e-01
1.61115444e+00 -1.18995273e+00 -2.56379008e-01 -4.02035899e-02
6.39833331e-01 -9.39139724e-01 1.11153746e+00 3.22910964e-01
-1.19459224e+00 -5.58881760e-01 -7.27966666e-01 -1.80438295e-01
-3.37147951e-01 1.18503019e-01 7.13775873e-01 9.22150791e-01
-1.75406563e+00 7.27798104e-01 -1.06952500e+00 -6.16062760e-01
9.90554929e-01 5.65121591e-01 -7.39270866e-01 -2.14568183e-01
-5.38541555e-01 5.18397152e-01 -1.66869566e-01 1.47003874e-01
-9.39769685e-01 -9.94188726e-01 -7.68696308e-01 -1.39892355e-01
2.47938782e-01 -7.82844365e-01 1.38764811e+00 -7.72669673e-01
-1.41034365e+00 1.29111767e+00 -2.82700837e-01 6.20703809e-02
6.15251541e-01 -2.54019558e-01 -4.32837367e-01 -6.21639527e-02
-1.47315189e-01 7.06202686e-01 9.18248832e-01 -6.09108448e-01
-1.47132993e-01 -6.26453936e-01 -3.64190310e-01 -1.45477176e-01
-6.26465499e-01 6.18499756e-01 -1.15728712e+00 -4.20275033e-01
-6.79128468e-01 -9.82463777e-01 3.33469182e-01 2.86950558e-01
-3.19917500e-01 -3.18461806e-01 9.57309365e-01 -7.70058632e-01
9.41695213e-01 -2.27658153e+00 -1.88847989e-01 -1.34156689e-01
4.36449796e-01 6.62816823e-01 -5.53213835e-01 4.55574930e-01
-4.17957157e-01 1.25738652e-02 -1.47451207e-01 -1.02735293e+00
7.21093342e-02 -1.83179490e-02 2.60704011e-03 5.79623938e-01
4.16331232e-01 9.37499762e-01 -4.72828656e-01 -1.50934845e-01
-5.11204153e-02 7.14711487e-01 -7.86915481e-01 2.70615876e-01
9.22045857e-02 3.90048087e-01 -1.78849906e-01 1.09337509e+00
1.08464158e+00 -1.60770819e-01 -3.77020687e-02 1.25625599e-02
-8.10760036e-02 2.17045043e-02 -1.04664505e+00 1.63727093e+00
-1.00285336e-01 6.14401817e-01 5.02650976e-01 -6.51390374e-01
7.94267356e-01 2.45433688e-01 4.84824330e-01 -4.57607955e-01
1.01602457e-01 1.45641387e-01 -1.82211563e-01 -3.46486658e-01
3.84340644e-01 5.40379882e-01 4.20842797e-01 6.34479105e-01
2.58633852e-01 2.56516248e-01 2.48709857e-01 1.43194929e-01
1.29069316e+00 2.84288079e-01 4.35536914e-02 -1.21786676e-01
2.81937540e-01 -2.31417581e-01 5.73348582e-01 4.55545455e-01
-4.49948788e-01 6.82436764e-01 4.91538584e-01 -7.23447800e-01
-1.06418777e+00 -8.17006171e-01 -3.57285440e-01 1.38421953e+00
-6.13113344e-01 -8.70647848e-01 -9.02879596e-01 -4.89572048e-01
1.86421275e-01 4.26589921e-02 -5.71821928e-01 3.12706046e-02
-4.23104763e-01 -8.30968618e-01 1.02262676e+00 4.36569661e-01
4.20722246e-01 -1.17661846e+00 -1.29622743e-01 -1.99934959e-01
2.09704593e-01 -7.58584261e-01 -5.23573697e-01 -3.00606787e-01
-5.57680905e-01 -1.26580882e+00 -7.22173691e-01 -6.82512105e-01
6.28864467e-01 1.99631527e-01 1.40076113e+00 7.93413162e-01
-8.18624496e-01 6.52465284e-01 -2.89532900e-01 -3.82338464e-01
-1.00936428e-01 9.79381520e-03 3.14049333e-01 -1.03104778e-01
5.07607818e-01 -5.92996955e-01 -5.86588562e-01 3.19149464e-01
-5.31644583e-01 -4.07907031e-02 1.39825091e-01 8.72623980e-01
2.41731510e-01 -4.09950227e-01 2.97731847e-01 -7.47885168e-01
3.90622348e-01 -5.85753679e-01 -8.05648863e-01 1.86410770e-01
-4.10370499e-01 -3.90475810e-01 3.34274501e-01 -4.49469537e-02
-7.89689660e-01 -1.16500966e-01 -6.00376368e-01 -6.62902892e-01
-2.33035147e-01 1.12951294e-01 -1.86251655e-01 -2.64083892e-01
4.78182405e-01 2.02763483e-01 3.89643490e-01 -6.61836147e-01
1.92518368e-01 7.56122172e-01 4.23115075e-01 -7.53951967e-01
7.30488718e-01 3.16781789e-01 -3.46770376e-01 -8.75072002e-01
-2.48718709e-01 -2.25913957e-01 -4.62286830e-01 -1.50057703e-01
3.27761441e-01 -8.96464229e-01 -1.12817955e+00 1.15915012e+00
-1.01013219e+00 -5.72935402e-01 1.41665369e-01 5.42648323e-02
-3.37263674e-01 2.96888173e-01 -9.79347467e-01 -5.55720150e-01
-7.93110788e-01 -1.29365373e+00 1.22128522e+00 4.33945358e-01
-4.62796353e-02 -8.04099619e-01 1.13962695e-01 4.93824214e-01
6.05124116e-01 6.70951158e-02 4.21581686e-01 -6.50052786e-01
-3.43391985e-01 -8.28684345e-02 -1.66452542e-01 3.49774033e-01
-8.17162320e-02 6.09969139e-01 -1.25583446e+00 -8.77679467e-01
-2.72296906e-01 -6.41780734e-01 7.54213572e-01 8.71210992e-02
1.53888774e+00 -4.00466800e-01 -3.97420943e-01 1.21749997e+00
1.00029349e+00 -5.12727499e-02 8.28209400e-01 2.06111789e-01
6.42503202e-01 6.44209921e-01 2.90549725e-01 7.73645639e-01
4.47119325e-01 8.97678554e-01 3.81055653e-01 -5.09344526e-02
-6.24550879e-02 1.10652857e-01 5.53026617e-01 7.81531811e-01
-2.79930413e-01 7.55582526e-02 -1.18315899e+00 1.27113342e-01
-1.56981397e+00 -1.15060997e+00 -1.76553831e-01 2.13002086e+00
7.60918736e-01 -3.80200297e-01 4.80221957e-01 -2.45098859e-01
5.82479000e-01 1.38127739e-02 -6.59875691e-01 -4.04567629e-01
-2.97712013e-02 6.58538818e-01 -6.51504612e-03 3.39400589e-01
-1.02985299e+00 1.01667643e+00 5.76146221e+00 8.34028363e-01
-1.25895607e+00 2.21892163e-01 1.10974550e+00 -5.16124368e-01
1.60598695e-01 -2.01557219e-01 -1.20030093e+00 6.98312402e-01
9.83132184e-01 -2.67710418e-01 7.30997622e-01 9.84389782e-01
-6.03468977e-02 4.59826589e-02 -1.22524488e+00 1.28548992e+00
2.36024767e-01 -1.46479964e+00 -3.43953371e-01 3.17607373e-01
4.11348552e-01 5.19800246e-01 2.66389310e-01 3.18917394e-01
2.54836321e-01 -1.24129605e+00 4.49455559e-01 4.48789746e-01
1.20547605e+00 -7.29753196e-01 6.02248132e-01 -6.10023290e-02
-1.12277722e+00 1.66376024e-01 -4.42400187e-01 2.91573126e-02
-2.75531054e-01 6.97099388e-01 -6.86434984e-01 5.92877626e-01
1.12308133e+00 7.75336146e-01 -8.60251844e-01 9.97407138e-01
2.47701854e-01 4.79548544e-01 -4.69410330e-01 4.41252410e-01
-3.80407363e-01 -2.59459853e-01 3.49917173e-01 1.22139430e+00
5.23690939e-01 4.92380559e-02 -4.52397391e-03 5.32778203e-01
-3.66560876e-01 -2.62408555e-02 -5.89853048e-01 -1.38885930e-01
8.14914525e-01 1.66336930e+00 -4.31386590e-01 -2.28330955e-01
-5.61063349e-01 1.00703669e+00 5.75632751e-01 2.79136822e-02
-9.96112704e-01 -1.38236120e-01 1.28066337e+00 2.09250376e-02
1.62063286e-01 -1.46052822e-01 1.59339696e-01 -1.12225652e+00
9.48628858e-02 -1.43285394e+00 5.07399857e-01 -5.64131856e-01
-1.45804203e+00 1.02346432e+00 -2.24385738e-01 -7.74756134e-01
-8.75576958e-02 -8.19974601e-01 -8.54006588e-01 9.68460202e-01
-1.21476114e+00 -1.15176272e+00 -6.45820439e-01 9.37749445e-01
3.68447840e-01 -7.10648000e-01 1.06581187e+00 6.99171185e-01
-1.39577627e+00 1.20031321e+00 -1.72535628e-02 4.39133704e-01
1.01425052e+00 -7.13796318e-01 1.00189638e+00 6.60391569e-01
1.53763831e-01 7.73468614e-01 3.31733435e-01 -3.81949574e-01
-1.71589816e+00 -1.04000592e+00 4.87902880e-01 -7.20490396e-01
5.44135034e-01 -6.80742621e-01 -8.67815971e-01 8.30789149e-01
2.10575908e-01 4.31302547e-01 8.32112968e-01 1.83667019e-01
-5.66510320e-01 -3.33548635e-01 -1.12544644e+00 6.75341249e-01
1.32307100e+00 -5.46379626e-01 3.08406334e-02 4.91424024e-01
1.73587129e-01 -6.02205932e-01 -7.88082063e-01 2.03773364e-01
7.12691665e-01 -1.30728030e+00 8.72589707e-01 -4.66343433e-01
3.31814080e-01 -1.24921687e-02 7.10295066e-02 -1.14973152e+00
-3.68640542e-01 -9.86252248e-01 -3.02065283e-01 1.62106061e+00
1.18661366e-01 -9.36819255e-01 9.64187562e-01 5.41471958e-01
-2.79606253e-01 -1.04046369e+00 -9.78334546e-01 -6.88594103e-01
-1.00746013e-01 -4.13145572e-01 8.51296425e-01 1.05679011e+00
-4.07699764e-01 -2.61897221e-03 -3.14020425e-01 -1.20507173e-01
6.70136452e-01 -1.67714521e-01 1.27797592e+00 -9.74887848e-01
-4.13816690e-01 -6.00189388e-01 -3.47127497e-01 -6.30930901e-01
1.05138749e-01 -9.94914889e-01 -5.14115453e-01 -9.82923090e-01
3.30129355e-01 -5.41989326e-01 1.12395853e-01 1.13255477e+00
-8.21106881e-02 6.20580018e-01 2.50942498e-01 3.44929487e-01
-1.94117919e-01 5.58952451e-01 9.31107879e-01 1.13940299e-01
9.09810811e-02 -1.52697816e-01 -7.68842876e-01 5.35000741e-01
9.09734309e-01 -3.31522793e-01 2.31274460e-02 -6.15280747e-01
-2.39104256e-01 -4.18867111e-01 5.73202491e-01 -1.07659245e+00
2.61435211e-01 3.01059395e-01 6.41774952e-01 -2.12058663e-01
6.47347450e-01 -3.21851194e-01 4.14925694e-01 2.75714874e-01
3.07235122e-01 6.57800853e-01 6.38329268e-01 -6.28886670e-02
4.85832542e-02 5.00048026e-02 7.61516631e-01 -3.79927605e-02
-7.49236345e-01 9.63975608e-01 1.87379479e-01 -2.54399311e-02
1.04077446e+00 -2.57825226e-01 -8.07961106e-01 -2.90554732e-01
-5.11900187e-01 1.64703727e-01 7.06220031e-01 6.06312454e-01
4.55534816e-01 -1.33581805e+00 -9.69133973e-01 7.01842844e-01
1.87183231e-01 -3.24532211e-01 4.89112049e-01 1.15247989e+00
-7.28572130e-01 1.71065882e-01 -5.27109087e-01 -5.11447430e-01
-1.81084764e+00 2.53221452e-01 3.71102244e-01 9.18260887e-02
-6.22909069e-01 1.31723738e+00 -8.45369548e-02 -6.32868171e-01
1.93273857e-01 3.66022170e-01 1.28310874e-01 7.97237456e-02
9.88523602e-01 4.22659248e-01 4.12003219e-01 -7.62259126e-01
-6.13398731e-01 3.92113864e-01 -1.97763219e-01 3.06392431e-01
1.70289683e+00 3.45293820e-01 -6.33812129e-01 -1.27972484e-01
1.33900893e+00 7.42230713e-02 -1.26212847e+00 6.67742044e-02
-2.26500452e-01 -6.32933259e-01 -1.53658897e-01 -6.96141601e-01
-1.54631770e+00 7.43330598e-01 7.62076497e-01 -3.57288957e-01
1.25895631e+00 -7.65586197e-02 7.66129732e-01 8.36301818e-02
4.99702036e-01 -6.48272991e-01 1.62722602e-01 6.46169662e-01
9.76248324e-01 -1.24034762e+00 -1.36968475e-02 -4.16816503e-01
-3.63284290e-01 1.14958370e+00 9.73743975e-01 1.35795161e-01
7.02558875e-01 7.17732251e-01 6.81277364e-02 -4.12278593e-01
-9.09876347e-01 -2.11618599e-02 -1.44297546e-02 6.49110854e-01
6.52872741e-01 -8.02092925e-02 1.36615813e-01 5.64585209e-01
-3.91738057e-01 6.71481639e-02 2.45300248e-01 1.02657473e+00
2.50926107e-01 -1.53612256e+00 -5.94460726e-01 5.51181614e-01
-4.39904064e-01 -3.37439448e-01 -3.43654126e-01 5.70196807e-01
1.18660659e-01 7.25267410e-01 3.11727285e-01 -5.43230593e-01
8.52649435e-02 1.75133854e-01 5.90631306e-01 -6.38531744e-01
-7.21476138e-01 -2.25114465e-01 8.20555985e-02 -8.75545859e-01
1.07567511e-01 -6.60640895e-01 -7.69625902e-01 -1.29387200e+00
8.46021920e-02 -3.52966301e-02 6.13037348e-01 5.20853579e-01
1.00562608e+00 -1.61483679e-02 5.41784644e-01 -1.14017081e+00
-4.06512529e-01 -9.45110261e-01 -4.51613098e-01 1.38888597e-01
1.33598864e-01 -5.75109005e-01 -2.09977165e-01 -9.69217271e-02] | [13.295279502868652, 0.7324897646903992] |
968ef101-bbc4-4b7c-9a7c-0d937b4380a6 | stage-conscious-attention-network-scan-a | 2112.02278 | null | https://arxiv.org/abs/2112.02278v1 | https://arxiv.org/pdf/2112.02278v1.pdf | Stage Conscious Attention Network (SCAN) : A Demonstration-Conditioned Policy for Few-Shot Imitation | In few-shot imitation learning (FSIL), using behavioral cloning (BC) to solve unseen tasks with few expert demonstrations becomes a popular research direction. The following capabilities are essential in robotics applications: (1) Behaving in compound tasks that contain multiple stages. (2) Retrieving knowledge from few length-variant and misalignment demonstrations. (3) Learning from a different expert. No previous work can achieve these abilities at the same time. In this work, we conduct FSIL problem under the union of above settings and introduce a novel stage conscious attention network (SCAN) to retrieve knowledge from few demonstrations simultaneously. SCAN uses an attention module to identify each stage in length-variant demonstrations. Moreover, it is designed under demonstration-conditioned policy that learns the relationship between experts and agents. Experiment results show that SCAN can learn from different experts without fine-tuning and outperform baselines in complicated compound tasks with explainable visualization. | ['Winston H. Hsu', 'Yi-Ting Chen', 'Hung-Ting Su', 'Chi-Ming Chung', 'Jia-Fong Yeh'] | 2021-12-04 | null | null | null | null | ['few-shot-imitation-learning'] | ['methodology'] | [-2.80587710e-02 4.72921915e-02 5.57037108e-02 -2.65327469e-02
-4.25521702e-01 -5.45840263e-01 5.18537819e-01 -8.16005826e-01
-5.82554460e-01 8.59608114e-01 -1.08224489e-01 -1.61556918e-02
-1.50115803e-01 -4.73877266e-02 -9.43534195e-01 -8.19538772e-01
-1.63780183e-01 6.07946694e-01 4.08666044e-01 -1.77765548e-01
4.14282888e-01 4.29484636e-01 -1.36636436e+00 1.67289953e-04
1.22110164e+00 5.70033610e-01 9.67354715e-01 6.74901664e-01
2.04555854e-01 1.15291083e+00 -7.56308973e-01 9.51614976e-02
5.67856848e-01 -6.62595153e-01 -7.12114990e-01 -1.29876420e-01
1.94879726e-01 -8.43364298e-01 -4.14728284e-01 1.01354098e+00
5.62861860e-01 4.91222054e-01 6.03529096e-01 -1.50268209e+00
-8.75535667e-01 8.72211695e-01 -4.86441731e-01 3.82571518e-01
2.02528656e-01 1.15719306e+00 6.22374237e-01 -7.29319453e-01
3.76166135e-01 1.66847825e+00 2.10812137e-01 9.27067935e-01
-1.00171542e+00 -7.91229963e-01 6.07676268e-01 6.09205008e-01
-1.02973127e+00 -2.39782080e-01 4.91768479e-01 -4.88916814e-01
1.24686444e+00 -2.44990766e-01 8.15115809e-01 1.74799097e+00
1.30368829e-01 1.11070609e+00 1.24380696e+00 5.99019751e-02
4.19349402e-01 -1.59603819e-01 1.69429928e-02 9.12724972e-01
2.89838433e-01 4.45619315e-01 -6.65115058e-01 2.49679342e-01
1.13690531e+00 3.21953803e-01 -4.88019764e-01 -6.98106706e-01
-1.57492316e+00 5.07309258e-01 5.32897651e-01 1.38456047e-01
-3.26515883e-01 5.54855347e-01 1.60233051e-01 9.23071086e-01
-3.41391504e-01 1.05803847e+00 -4.72311288e-01 -2.52421319e-01
-3.13686818e-01 1.86477795e-01 8.56574178e-01 1.55200326e+00
7.16622055e-01 4.72986460e-01 -3.90856594e-01 2.02403799e-01
-1.65567055e-01 4.83053148e-01 8.79266143e-01 -1.47391367e+00
4.04586643e-01 4.27154213e-01 4.64900434e-01 -2.30813384e-01
-3.50011140e-01 -3.69501889e-01 -4.56024259e-01 7.11372912e-01
2.85105675e-01 -4.07467902e-01 -1.11254632e+00 1.84116828e+00
8.17829296e-02 4.99463946e-01 2.28350237e-01 1.25132227e+00
6.84164822e-01 3.53746057e-01 2.06428189e-02 2.50668544e-03
1.04316986e+00 -1.75648987e+00 -5.14230072e-01 -3.37532252e-01
3.55896860e-01 -7.28033558e-02 1.50170946e+00 4.25613075e-01
-8.65154147e-01 -8.20196331e-01 -1.12823582e+00 -2.16562480e-01
-2.96508312e-01 1.91738382e-01 5.88569343e-01 -3.61735001e-02
-1.01398563e+00 8.43140900e-01 -1.07029927e+00 -4.69158322e-01
4.81269717e-01 4.16504115e-01 -3.10515463e-02 1.29103482e-01
-8.04757535e-01 1.12989819e+00 3.83079052e-01 4.16439734e-02
-2.05428076e+00 -6.39529347e-01 -7.78006911e-01 2.62531191e-01
1.07665884e+00 -1.05834770e+00 1.54614604e+00 -1.03051543e+00
-2.06983852e+00 4.61980522e-01 1.55151471e-01 -3.38689268e-01
6.46241486e-01 -6.08254433e-01 1.34749413e-01 2.81051267e-02
2.84393221e-01 7.74077892e-01 1.21864891e+00 -1.13070118e+00
-6.74708128e-01 -3.08494061e-01 5.25116682e-01 4.46399629e-01
8.42651874e-02 -3.90548795e-01 -2.20803604e-01 -3.54857653e-01
-2.61509448e-01 -9.56972122e-01 -1.11022048e-01 2.26305410e-01
-2.90993601e-01 -5.89929163e-01 6.98422790e-01 -3.25131685e-01
4.33906674e-01 -1.97332156e+00 9.21589851e-01 -5.66986024e-01
4.71408814e-01 3.64667565e-01 -4.76220846e-01 3.60618144e-01
2.11092249e-01 -3.53938192e-01 -7.67283738e-02 -2.90338516e-01
3.07958245e-01 3.39412123e-01 -2.86779672e-01 2.99750090e-01
2.81507790e-01 1.51587141e+00 -1.37137496e+00 -4.32852536e-01
1.76862001e-01 3.61288935e-02 -3.42746347e-01 6.94347262e-01
-3.65848958e-01 9.51169014e-01 -7.19707072e-01 6.35278940e-01
1.79281071e-01 -3.49843442e-01 -9.02442727e-04 9.42434743e-02
-1.53125688e-01 -1.29896507e-01 -8.56566846e-01 2.37101340e+00
-4.11069691e-01 5.50909936e-01 3.20912600e-01 -1.04425931e+00
4.62833673e-01 1.95974901e-01 -1.53780386e-01 -5.92253983e-01
1.47973821e-01 4.11411822e-02 4.33157086e-01 -1.10614383e+00
-8.62289965e-02 1.73956901e-01 1.73002519e-02 5.42078316e-01
5.95441699e-01 -1.95655391e-01 2.91815195e-02 2.18070805e-01
1.40959275e+00 5.82603931e-01 3.83899957e-01 1.23977549e-02
1.73224211e-01 9.26759467e-02 7.16889918e-01 1.30672622e+00
-7.59654582e-01 4.89731245e-02 3.72239292e-01 -3.90667021e-01
-9.40874517e-01 -1.09583497e+00 6.95849180e-01 1.32731056e+00
5.65162003e-01 1.06613629e-01 -5.27516484e-01 -7.44636595e-01
7.02050030e-02 7.62956321e-01 -8.46818507e-01 -2.87955612e-01
-8.03206384e-01 1.62512779e-01 2.84017920e-01 6.31751478e-01
6.96852624e-01 -1.62309575e+00 -1.42868495e+00 -9.86931399e-02
1.92139715e-01 -8.07588696e-01 -5.13587534e-01 3.62703800e-01
-6.93870366e-01 -1.35137141e+00 -1.10097516e+00 -1.04274154e+00
7.11466551e-01 7.05171108e-01 8.12339842e-01 -5.81483394e-02
-3.72902274e-01 7.17753291e-01 -4.20518577e-01 -3.41967046e-01
8.85599852e-02 1.16527155e-02 2.33323187e-01 -5.13893068e-01
1.79726020e-01 -8.93503308e-01 -5.91883183e-01 4.02629316e-01
-4.84721214e-01 1.99136630e-01 1.25439179e+00 1.12867653e+00
1.39442161e-01 -2.67201275e-01 5.86379468e-01 -4.59293664e-01
9.41939771e-01 -4.33600128e-01 -6.65680170e-01 4.63907212e-01
-3.96583557e-01 3.67409706e-01 9.18790996e-01 -1.00577426e+00
-1.22492707e+00 -1.16144465e-02 5.99955678e-01 -1.05660105e+00
-3.59360009e-01 8.07116181e-02 -3.99137512e-02 -1.02952801e-01
6.33319616e-01 4.61890727e-01 3.70513275e-02 -6.06041551e-01
5.33477545e-01 2.82704830e-01 6.33992434e-01 -6.54652834e-01
9.40745234e-01 2.58893162e-01 -3.55716765e-01 -4.17140514e-01
-6.61839604e-01 -2.89466441e-01 -7.61411488e-01 -5.64951636e-02
8.22690129e-01 -9.81570721e-01 -1.18883789e+00 3.96667600e-01
-1.08611953e+00 -1.01083004e+00 -2.59964019e-01 6.59802556e-01
-8.32057118e-01 1.66710481e-01 -5.07804930e-01 -8.32838953e-01
-7.15930015e-02 -1.44831765e+00 8.51422668e-01 5.63736022e-01
-4.27709743e-02 -3.88189495e-01 -2.57865228e-02 1.30411461e-01
5.33338428e-01 -5.24362810e-02 6.98239684e-01 -6.24823689e-01
-9.82304752e-01 5.07878065e-01 -2.39010558e-01 3.60952248e-03
5.61400913e-02 -5.20096719e-01 -7.25643396e-01 -6.99466705e-01
1.35374129e-01 -9.39756393e-01 1.04829001e+00 1.11388013e-01
1.08702290e+00 -3.70503396e-01 -4.79409426e-01 6.71373188e-01
9.84183908e-01 2.98491359e-01 2.40178660e-01 2.37907603e-01
7.55598068e-01 5.47428966e-01 6.57628894e-01 2.13407710e-01
2.96245784e-01 4.41418320e-01 6.73317134e-01 4.14220333e-01
-3.25552970e-01 -4.01077420e-01 7.32820511e-01 7.36288846e-01
-2.40524754e-01 1.49880916e-01 -4.71538633e-01 5.50769210e-01
-2.20207214e+00 -1.10921586e+00 5.67355275e-01 1.67655873e+00
6.71632290e-01 1.29273236e-02 1.76616862e-01 -7.66203582e-01
5.72376668e-01 -6.56639785e-02 -1.64380169e+00 -3.68848369e-02
1.45732135e-01 2.02326737e-02 1.46688998e-01 2.41635874e-01
-6.91914558e-01 1.28839254e+00 5.51892614e+00 5.82038522e-01
-8.17033231e-01 1.05100892e-01 -1.65020019e-01 -3.80680650e-01
3.04091960e-01 1.40404344e-01 -5.62905312e-01 4.82500255e-01
2.90347993e-01 -3.55183393e-01 1.01118767e+00 1.16104650e+00
-9.88290366e-03 -2.12147325e-01 -1.49215937e+00 1.04833293e+00
2.39134461e-01 -5.70079148e-01 -1.30794898e-01 -1.51632458e-01
9.22182560e-01 1.35621592e-01 2.64925003e-01 1.08737564e+00
7.43940830e-01 -1.09113920e+00 6.68132722e-01 5.77964902e-01
5.84962368e-01 -1.87258914e-01 3.03120881e-01 9.39978778e-01
-8.45420480e-01 -6.74881518e-01 -6.22216463e-01 -3.16158593e-01
9.48139001e-03 -5.12720644e-01 -8.22238088e-01 3.01155716e-01
8.47449124e-01 9.80379581e-01 -4.23393697e-01 1.10336673e+00
-7.33997524e-01 1.73232332e-01 4.14338373e-02 -5.00891984e-01
4.44546193e-01 -1.46713108e-01 7.54444063e-01 5.70674241e-01
2.47036204e-01 3.08772773e-01 2.38380149e-01 1.42389488e+00
1.68767907e-02 -6.06443346e-01 -6.36710823e-01 -4.15993296e-02
6.50115073e-01 1.03492761e+00 -4.49487656e-01 -4.71470296e-01
-2.24434495e-01 1.41341162e+00 8.97208989e-01 6.38506770e-01
-9.22091961e-01 -4.31622237e-01 6.05641127e-01 -5.53076088e-01
7.93407202e-01 -3.76713485e-01 3.08471590e-01 -1.40482605e+00
-2.39787810e-02 -1.14384401e+00 2.57425699e-02 -1.08719707e+00
-1.26621234e+00 4.28311318e-01 -6.54198900e-02 -1.36464453e+00
-1.77207708e-01 -7.27864683e-01 -8.00980866e-01 6.87295020e-01
-1.79945290e+00 -1.05669999e+00 -5.79948127e-01 8.03514957e-01
1.17833114e+00 -4.98122811e-01 7.78512120e-01 -3.31440389e-01
-8.32051516e-01 4.32693481e-01 2.12561842e-02 -1.13875642e-01
7.33716130e-01 -1.43696427e+00 2.86769688e-01 6.02963686e-01
-1.10147119e-01 9.69152331e-01 6.38183177e-01 -6.11101568e-01
-1.72310650e+00 -6.71624839e-01 -1.06647633e-01 -4.64787096e-01
6.82937860e-01 -2.18509853e-01 -7.92308331e-01 1.01860225e+00
4.65580642e-01 -2.64440596e-01 1.15475439e-01 -5.16553409e-02
-4.62846965e-01 1.23459265e-01 -7.45519519e-01 9.46512282e-01
1.49151766e+00 -4.54529673e-01 -1.20030248e+00 2.65488535e-01
9.51020420e-01 -4.65340257e-01 -3.11457634e-01 4.31664661e-02
7.27372706e-01 -9.10878301e-01 9.63850319e-01 -1.08268249e+00
4.68742490e-01 -4.37368751e-01 3.99835229e-01 -1.75847685e+00
-4.33520019e-01 -9.36511278e-01 -3.50226045e-01 5.46217144e-01
9.55836400e-02 -6.39197171e-01 3.14102381e-01 4.62786257e-01
-4.20239478e-01 -5.59546769e-01 -6.92781270e-01 -1.29274988e+00
-3.52750011e-02 3.05841804e-01 4.69771862e-01 7.18574226e-01
2.05858007e-01 6.73571765e-01 -7.14170456e-01 1.03079073e-01
7.26900280e-01 5.21818578e-01 1.25857794e+00 -1.24611354e+00
-5.51238358e-01 -5.77344418e-01 1.13506697e-01 -1.65393198e+00
4.57120210e-01 -7.39568114e-01 4.36146289e-01 -1.32955611e+00
3.34333599e-01 -4.33315523e-02 -3.67972195e-01 5.70794463e-01
-3.02942216e-01 -5.79132378e-01 3.06222647e-01 5.39100289e-01
-1.28730297e+00 9.25855875e-01 1.85947871e+00 -2.31709391e-01
-2.08272770e-01 -1.74602672e-01 -6.35842681e-01 7.29271829e-01
8.16590488e-01 -4.09549534e-01 -7.37791717e-01 -7.66231298e-01
-8.54817703e-02 4.36393693e-02 6.15117550e-01 -1.20680439e+00
6.86862350e-01 -3.28273684e-01 3.84999126e-01 -3.52895200e-01
4.92983669e-01 -8.18890393e-01 -2.13405713e-01 7.93347418e-01
-4.75348502e-01 6.98653236e-02 -7.25604072e-02 1.06524098e+00
2.93374658e-01 -2.50282496e-01 4.50060606e-01 -7.70959973e-01
-1.35831380e+00 1.79468378e-01 -2.83232540e-01 1.81803793e-01
1.19158244e+00 -2.68290550e-01 -6.22883201e-01 -3.45071048e-01
-6.49089038e-01 9.84966159e-01 5.15942097e-01 4.34759170e-01
6.86322331e-01 -9.38066423e-01 -4.74658310e-01 -6.98497472e-03
1.71257526e-01 1.96199343e-01 2.89821744e-01 9.19330239e-01
-2.25591883e-01 4.80032861e-01 -6.50907815e-01 -3.35072428e-01
-8.09054375e-01 9.83328164e-01 3.28622758e-01 3.48318310e-04
-8.77073705e-01 1.11666453e+00 5.69749236e-01 -5.05789161e-01
6.75131917e-01 -3.50400835e-01 -1.83732897e-01 -4.20857012e-01
4.81574327e-01 4.95638788e-01 -7.96580076e-01 5.55657083e-03
-9.79013517e-02 4.47800517e-01 -1.85315073e-01 9.11419466e-03
1.44268930e+00 -1.52097464e-01 2.34775409e-01 7.06216931e-01
6.56297028e-01 -7.16619670e-01 -2.06067371e+00 -2.77835846e-01
-1.93024993e-01 -3.67545754e-01 -3.57116550e-01 -1.17883587e+00
-7.10352004e-01 1.09900725e+00 3.13497424e-01 -2.36657128e-01
5.63192070e-01 4.74501960e-02 6.10311151e-01 1.15026987e+00
7.10744023e-01 -1.19476318e+00 8.09551001e-01 5.54483414e-01
1.19296265e+00 -1.57390368e+00 -1.70543745e-01 1.71412885e-01
-8.63859653e-01 8.85270357e-01 1.31660831e+00 -4.23359603e-01
1.03465743e-01 -1.25402167e-01 -1.62212208e-01 -4.19440985e-01
-1.05712366e+00 -5.76187611e-01 -1.57047495e-01 9.15365398e-01
-4.55828995e-01 -1.38470933e-01 1.39751241e-01 7.98550248e-01
9.70295221e-02 1.24566415e-02 3.70243013e-01 9.56597149e-01
-6.52276218e-01 -4.91681457e-01 -5.93062676e-03 3.20676327e-01
2.26787746e-01 7.80584142e-02 -5.76817870e-01 8.86082768e-01
-8.89395103e-02 8.29478621e-01 -3.14768523e-01 -2.54246771e-01
2.46112481e-01 1.16086928e-02 6.62581563e-01 -7.54214466e-01
-6.21335387e-01 -2.62508929e-01 -4.77769792e-01 -1.03881860e+00
-3.44968557e-01 -3.39657903e-01 -1.15413666e+00 -2.01531239e-02
-2.84779906e-01 1.60541534e-02 2.77609140e-01 9.54598606e-01
5.30395031e-01 8.74690831e-01 3.00742775e-01 -1.10091853e+00
-1.17880702e+00 -1.28537464e+00 -5.63138366e-01 3.12314451e-01
5.93384802e-01 -1.24977231e+00 -6.50693655e-01 -1.83404386e-01] | [4.33360481262207, 1.1033991575241089] |
8969a0ac-bc9e-4fd4-8a40-36a7f765f956 | imperfect-segmentation-labels-how-much-do | 1806.04618 | null | http://arxiv.org/abs/1806.04618v3 | http://arxiv.org/pdf/1806.04618v3.pdf | Imperfect Segmentation Labels: How Much Do They Matter? | Labeled datasets for semantic segmentation are imperfect, especially in
medical imaging where borders are often subtle or ill-defined. Little work has
been done to analyze the effect that label errors have on the performance of
segmentation methodologies. Here we present a large-scale study of model
performance in the presence of varying types and degrees of error in training
data. We trained U-Net, SegNet, and FCN32 several times for liver segmentation
with 10 different modes of ground-truth perturbation. Our results show that for
each architecture, performance steadily declines with boundary-localized
errors, however, U-Net was significantly more robust to jagged boundary errors
than the other architectures. We also found that each architecture was very
robust to non-boundary-localized errors, suggesting that boundary-localized
errors are fundamentally different and more challenging problem than random
label errors in a classification setting. | ['Joshua Dean', 'Nikolaos Papanikolopoulos', 'Nicholas Heller'] | 2018-06-12 | null | null | null | null | ['liver-segmentation'] | ['medical'] | [ 2.12820679e-01 3.02723527e-01 -1.31882995e-01 -6.11630321e-01
-1.17161071e+00 -7.23767340e-01 2.03397229e-01 3.45025510e-01
-3.38852733e-01 8.26334894e-01 1.79314151e-01 -5.30500710e-01
1.74987733e-01 -3.34050566e-01 -8.48141193e-01 -6.99972272e-01
-2.13148177e-01 7.05242157e-01 2.88655609e-01 2.64598906e-01
1.08560309e-01 1.84981883e-01 -5.53969562e-01 2.39450768e-01
7.99520791e-01 6.80003881e-01 -2.07970217e-01 6.28488421e-01
1.05325192e-01 7.42396951e-01 -7.80085027e-01 -3.31412144e-02
3.34623516e-01 -6.45331442e-01 -1.18041670e+00 2.81331539e-01
8.22153270e-01 -1.31620109e-01 -7.24347532e-02 1.23314190e+00
7.13941693e-01 2.13137064e-02 6.64912343e-01 -7.62574553e-01
-3.93970549e-01 7.23348916e-01 -4.72476929e-01 6.56832576e-01
1.20690718e-01 4.69923109e-01 5.09357274e-01 -4.61576253e-01
8.50941896e-01 1.01830769e+00 1.42056501e+00 4.58512336e-01
-1.59199476e+00 -4.30376083e-01 2.61633843e-02 -3.82482260e-01
-1.27054441e+00 -3.48587513e-01 2.77533084e-01 -7.31428623e-01
7.19122648e-01 1.78192094e-01 3.24731499e-01 9.55269814e-01
4.98185277e-01 4.66383964e-01 1.20871127e+00 -2.98277825e-01
1.75460145e-01 -2.40473058e-02 5.78571022e-01 7.91014671e-01
5.42869210e-01 2.92251259e-02 2.25081995e-01 -3.22264403e-01
7.25842416e-01 -4.24564064e-01 -5.03457546e-01 -1.21149205e-01
-1.40030777e+00 6.93283975e-01 6.99583113e-01 5.11846304e-01
-2.54077584e-01 3.21565211e-01 5.18596172e-01 1.43344402e-01
5.32316625e-01 1.01007390e+00 -5.94675064e-01 1.29404038e-01
-1.10875511e+00 6.67436048e-02 7.26198316e-01 8.25379491e-01
3.57606202e-01 9.73820314e-02 -3.78675610e-01 7.34488070e-01
-2.95658549e-03 2.28836364e-03 6.03695631e-01 -1.06977797e+00
1.89097762e-01 2.90627599e-01 7.71619603e-02 -1.03964865e+00
-9.92911816e-01 -7.19242692e-01 -6.01144493e-01 1.40288159e-01
1.16016471e+00 -5.04922211e-01 -1.57139742e+00 1.67457366e+00
8.34246948e-02 1.42531395e-01 -2.47636482e-01 1.03976846e+00
1.08308876e+00 2.38640737e-02 4.95103866e-01 3.94793488e-02
1.19230640e+00 -9.78055060e-01 -6.64393842e-01 -3.92812610e-01
9.93531048e-01 -8.12619627e-01 1.11560667e+00 4.57681380e-02
-1.13399470e+00 -1.63895771e-01 -7.95111358e-01 1.14561796e-01
-4.41434719e-02 -1.88697115e-01 4.78470922e-01 7.80704141e-01
-1.16080642e+00 7.22741187e-01 -8.93812060e-01 -4.86409634e-01
7.75877774e-01 1.87150806e-01 -3.14426541e-01 -5.91630936e-02
-8.70173991e-01 1.21008718e+00 3.81355673e-01 -7.30792508e-02
-9.18638527e-01 -1.04949343e+00 -8.19758952e-01 -2.73898840e-01
3.51146996e-01 -5.70184171e-01 1.47424376e+00 -1.11242628e+00
-7.33314633e-01 1.26933372e+00 -8.92368928e-02 -4.70262378e-01
8.30451190e-01 3.64361495e-01 -1.81500584e-01 1.95700541e-01
2.19264492e-01 7.64742136e-01 3.93603027e-01 -1.48795521e+00
-1.16764531e-01 -4.33294833e-01 -3.00190777e-01 1.21547140e-01
4.75687444e-01 4.80322633e-03 -3.71334329e-03 -8.57817709e-01
5.42601228e-01 -1.06537294e+00 -5.39193511e-01 -4.36920226e-02
-5.59197247e-01 6.63442314e-02 4.15750057e-01 -7.29690015e-01
9.44316328e-01 -2.06742239e+00 -4.26708937e-01 1.86059713e-01
3.46006900e-01 1.12013996e-01 7.99748451e-02 -2.80414999e-01
-2.27988139e-01 7.29568064e-01 -6.15535378e-01 -8.15351605e-02
-4.25208479e-01 2.66410351e-01 2.72678852e-01 5.47587693e-01
-4.65251729e-02 1.08395493e+00 -1.09572327e+00 -6.04257643e-01
7.79655874e-02 1.33579895e-01 -5.62814534e-01 -3.20842475e-01
-1.08563691e-01 8.45427215e-01 -3.20169896e-01 8.83944094e-01
4.27368432e-01 -7.00214088e-01 8.65005106e-02 -1.95725530e-01
2.72000641e-01 1.00032330e-01 -6.59520030e-01 1.70809209e+00
-1.23693079e-01 6.05550170e-01 1.59514815e-01 -7.50819147e-01
4.16501492e-01 3.92353594e-01 5.84296167e-01 -4.55514133e-01
2.90273339e-01 4.12254721e-01 4.16689664e-01 -4.38808769e-01
2.21569926e-01 -6.25093400e-01 1.46990672e-01 3.82396638e-01
-1.45966351e-01 -2.69176513e-01 -1.14007220e-02 5.79650924e-02
1.31266487e+00 -3.29186022e-01 4.33611535e-02 -7.50650942e-01
-5.90711161e-02 5.77426553e-01 7.49348223e-01 1.08805692e+00
-8.60476971e-01 1.24947691e+00 5.45619905e-01 -6.15885198e-01
-8.79608393e-01 -8.88787329e-01 -6.44491494e-01 6.93343282e-01
2.18834803e-01 -5.69394790e-02 -1.09176147e+00 -1.14212728e+00
6.97769448e-02 7.67864406e-01 -7.09812045e-01 -6.33858070e-02
-5.27851820e-01 -1.35981846e+00 8.64035189e-01 5.71196973e-01
3.48039389e-01 -1.10828817e+00 -6.01387084e-01 3.25807303e-01
-2.90178299e-01 -1.19150758e+00 -6.45892441e-01 2.14525655e-01
-1.29495680e+00 -1.38134694e+00 -7.42564499e-01 -9.04267430e-01
1.15557027e+00 -1.40832886e-01 1.55686963e+00 4.32776660e-01
-4.23018008e-01 3.01555485e-01 -1.40455306e-01 -3.18428755e-01
-7.86627948e-01 2.92116672e-01 -4.15285319e-01 -6.25364602e-01
1.82702303e-01 8.61234963e-03 -6.66352987e-01 5.59331000e-01
-7.94754088e-01 -3.23273510e-01 1.52442053e-01 9.92044687e-01
6.37262881e-01 -3.66463736e-02 6.12684190e-01 -1.38363385e+00
6.11535192e-01 -6.18021548e-01 -1.26895681e-01 1.29672170e-01
-7.01389015e-01 -2.63545007e-01 3.08191389e-01 -2.62087733e-01
-7.55616367e-01 -3.95698287e-02 -9.55853686e-02 -3.02360326e-01
-4.26740974e-01 4.18064892e-01 4.99880940e-01 -4.66591597e-01
1.18856931e+00 -2.96192557e-01 2.12877706e-01 -3.24050963e-01
-1.08083494e-01 1.75714746e-01 4.08272624e-01 -5.62671125e-01
1.75857902e-01 4.54046965e-01 -1.33907855e-01 -6.68007076e-01
-9.49757159e-01 -2.34869868e-01 -4.05826360e-01 -3.81876901e-02
1.05511558e+00 -6.19877100e-01 1.05033807e-01 7.38781154e-01
-7.78207242e-01 -1.01137602e+00 -4.65254933e-01 3.88650388e-01
-3.74422282e-01 3.79305273e-01 -9.00186718e-01 3.90246175e-02
-3.53521407e-01 -1.73660791e+00 7.19139159e-01 2.83885747e-01
-4.60281193e-01 -1.39465928e+00 -2.30879143e-01 3.98159087e-01
5.52917659e-01 5.90243936e-01 9.77081656e-01 -8.66535187e-01
-2.66445220e-01 -2.41927966e-01 -2.42438614e-01 3.66532505e-01
3.17493975e-01 -1.19016320e-01 -8.19701791e-01 -3.82923990e-01
-1.34810107e-02 -4.75476354e-01 8.02084863e-01 9.55662191e-01
1.27292502e+00 -6.32117093e-02 -4.24394369e-01 7.24135518e-01
1.27794921e+00 2.58393615e-01 5.17636001e-01 3.86203885e-01
5.90483785e-01 3.18887800e-01 3.38306665e-01 -2.03504756e-01
2.00855702e-01 3.19422603e-01 2.31445000e-01 -5.06742418e-01
-3.58195961e-01 1.40172958e-01 -3.59854341e-01 1.62165672e-01
3.24170500e-01 -2.55865932e-01 -1.52454638e+00 6.73082232e-01
-1.41777432e+00 -4.80192095e-01 -3.54427636e-01 1.89490926e+00
9.30543244e-01 2.37705603e-01 -6.97332099e-02 -2.30131939e-01
8.10880423e-01 7.50918388e-02 -5.64527988e-01 -2.22356878e-02
8.70200172e-02 -7.53083155e-02 8.43454897e-01 5.07415235e-01
-1.22930753e+00 9.33388591e-01 8.36890507e+00 2.88888037e-01
-1.12446916e+00 2.66883314e-01 1.29288983e+00 9.31576639e-02
-1.70920238e-01 -1.77517891e-01 -2.94289917e-01 5.74207485e-01
5.83747327e-01 2.94463277e-01 -1.94107089e-02 6.44684434e-01
1.18923515e-01 -3.90454382e-01 -9.64774668e-01 5.05169272e-01
-4.23406698e-02 -1.24222398e+00 -4.06283945e-01 -2.12255493e-01
1.09612322e+00 4.62405950e-01 -1.22339204e-01 5.09149395e-02
5.57270050e-01 -1.25188267e+00 4.69620198e-01 2.12564677e-01
8.45026195e-01 -1.71351314e-01 9.50425148e-01 1.53542474e-01
-4.05378968e-01 2.80912995e-01 -1.59502625e-01 2.44676009e-01
2.06789315e-01 6.87144578e-01 -1.08677459e+00 6.44179583e-02
6.57797277e-01 3.96816880e-01 -6.78706586e-01 1.26955533e+00
-6.34531081e-02 9.14342821e-01 -4.67941850e-01 5.47125101e-01
5.37325442e-01 1.64737791e-01 6.11189008e-01 1.41573620e+00
-9.69329923e-02 1.86124086e-01 3.58524442e-01 7.44680166e-01
-1.78354189e-01 -1.27513245e-01 -4.02267665e-01 1.89174429e-01
3.71408463e-01 1.15231574e+00 -1.50505352e+00 -5.66764057e-01
-2.02893600e-01 6.41826451e-01 -1.31481439e-01 5.20564556e-01
-8.36935401e-01 1.95767224e-01 4.65047330e-01 2.90810198e-01
-1.66143715e-01 7.51825199e-02 -8.56835186e-01 -1.05264127e+00
-2.01685160e-01 -9.57545877e-01 7.69807756e-01 -8.45250368e-01
-1.30335104e+00 5.06700695e-01 -1.65985763e-01 -5.63599885e-01
2.98300143e-02 -4.14830387e-01 -6.35976732e-01 7.10285127e-01
-1.33589196e+00 -6.59809709e-01 -5.73223472e-01 2.34696984e-01
4.53805566e-01 1.89240500e-01 7.76549339e-01 3.17461163e-01
-5.51654458e-01 7.23301291e-01 -4.57565971e-02 5.39544761e-01
9.36782479e-01 -1.37889314e+00 4.60855782e-01 7.45095551e-01
-1.48717269e-01 5.33672988e-01 7.02013016e-01 -8.80863190e-01
-6.16715550e-01 -1.11748612e+00 4.35643464e-01 -6.49198711e-01
4.32746410e-01 3.13414872e-01 -1.01015425e+00 1.19867826e+00
-2.21818853e-02 5.27683616e-01 6.77713931e-01 1.45623777e-02
-3.77801619e-02 4.13230926e-01 -1.76726091e+00 2.58903712e-01
9.89517152e-01 -1.80759072e-01 -5.23310065e-01 6.17551923e-01
5.27216673e-01 -1.06642282e+00 -1.06937110e+00 6.33592665e-01
1.23714902e-01 -8.36737812e-01 8.53617966e-01 -8.04378450e-01
3.44625145e-01 -8.88299420e-02 1.60692096e-01 -1.47199559e+00
-4.60036665e-01 -1.82106808e-01 5.14597654e-01 7.26856053e-01
7.57447064e-01 -7.82276809e-01 9.59484458e-01 9.08309221e-01
-6.97523773e-01 -6.52895212e-01 -9.17858303e-01 -4.45224822e-01
5.23059726e-01 -2.57940650e-01 2.52534151e-01 1.51494622e+00
-3.91442925e-01 -1.22657917e-01 3.28852296e-01 1.21388055e-01
5.67194521e-01 -2.08500504e-01 2.26720139e-01 -8.68154168e-01
-1.31209955e-01 -6.48701727e-01 -4.56363559e-01 -3.75136405e-01
2.40140371e-02 -1.08238816e+00 3.31101209e-01 -1.59359658e+00
5.18861189e-02 -1.02274835e+00 -2.60659009e-01 6.79809928e-01
-5.26678920e-01 5.80466568e-01 7.27685615e-02 2.98981905e-01
-4.08774495e-01 -2.43359730e-01 1.54255116e+00 -1.53160378e-01
2.31125224e-02 -4.77572950e-03 -7.00783372e-01 8.97113979e-01
8.27332437e-01 -6.49538517e-01 -2.67187119e-01 -6.87206805e-01
-2.03607813e-01 4.01902683e-02 4.14072216e-01 -9.90685940e-01
1.74305309e-02 1.53993011e-01 4.95063752e-01 -2.62626082e-01
-3.64506900e-01 -6.73330426e-01 1.36315271e-01 5.22739470e-01
-4.95204687e-01 7.38663077e-02 4.80736971e-01 2.18231395e-01
3.12708579e-02 -3.85703653e-01 1.25368941e+00 -6.79013252e-01
-5.51546991e-01 5.75806051e-02 -3.26258302e-01 7.92441070e-01
1.06484151e+00 -2.29606360e-01 -1.95847929e-01 -3.86678018e-02
-1.30948150e+00 3.75773698e-01 7.62236476e-01 -1.55177424e-02
1.44307047e-01 -8.23118687e-01 -6.22090936e-01 8.88320953e-02
-1.76478937e-01 4.70181674e-01 2.23139286e-01 1.21913326e+00
-9.91660595e-01 1.04851410e-01 4.51102629e-02 -1.03425193e+00
-9.45033252e-01 1.43521741e-01 1.15082526e+00 -3.90152365e-01
-7.73268700e-01 1.17257285e+00 1.53783903e-01 -5.69230437e-01
3.75277162e-01 -6.48504198e-01 2.74771929e-01 -4.61196639e-02
1.94625065e-01 3.27860385e-01 3.21521640e-01 -4.04550463e-01
-4.64934766e-01 4.20192868e-01 -2.02183411e-01 2.63946414e-01
8.54267657e-01 3.69610563e-02 9.26849172e-02 3.40161681e-01
8.03885996e-01 -6.64133966e-01 -1.17802477e+00 -1.28951937e-01
1.11519255e-01 -1.96195111e-01 3.95497084e-01 -1.37180841e+00
-1.46513009e+00 4.49891448e-01 6.09456062e-01 8.40454027e-02
7.04039931e-01 -7.32801408e-02 7.45692253e-01 -1.07895076e-01
4.31535184e-01 -9.10549998e-01 -2.07278088e-01 3.44020873e-01
4.58395839e-01 -1.49834645e+00 -3.83850671e-02 -3.49589825e-01
-6.96967125e-01 6.85104191e-01 6.47782505e-01 -1.42387360e-01
7.05534279e-01 3.74246508e-01 5.92469156e-01 -4.33384806e-01
-2.18320593e-01 1.13356180e-01 1.18651800e-01 4.88356233e-01
6.49997115e-01 2.29529291e-01 -2.64517099e-01 2.15516284e-01
-1.78881995e-02 8.52917656e-02 5.54318845e-01 9.57416534e-01
-1.99233875e-01 -5.90237796e-01 -5.03219843e-01 7.65066147e-01
-8.90500963e-01 -1.58723012e-01 -3.63643169e-01 1.10296524e+00
2.56497948e-03 6.30838275e-01 2.16888249e-01 1.60432619e-03
3.10840935e-01 3.96920890e-01 4.98269618e-01 -8.36831391e-01
-9.55556333e-01 2.85258126e-02 1.38426617e-01 -6.32757664e-01
-1.68428183e-01 -7.73195803e-01 -1.71418452e+00 -1.96157023e-01
-3.95342052e-01 -4.08954099e-02 4.95222598e-01 8.25456679e-01
1.85365230e-01 8.64114404e-01 7.04790419e-03 -5.51675081e-01
-6.34131908e-01 -1.05951822e+00 -4.25249696e-01 8.68225098e-01
3.42816889e-01 -5.61874449e-01 -6.34383976e-01 3.04121897e-02] | [14.735795974731445, -2.2806904315948486] |
62a7a060-835f-4e65-aaac-674b933b2707 | skillqg-learning-to-generate-question-for | 2305.04737 | null | https://arxiv.org/abs/2305.04737v1 | https://arxiv.org/pdf/2305.04737v1.pdf | SkillQG: Learning to Generate Question for Reading Comprehension Assessment | We present $\textbf{$\texttt{SkillQG}$}$: a question generation framework with controllable comprehension types for assessing and improving machine reading comprehension models. Existing question generation systems widely differentiate questions by $\textit{literal}$ information such as question words and answer types to generate semantically relevant questions for a given context. However, they rarely consider the $\textit{comprehension}$ nature of questions, i.e. the different comprehension capabilities embodied by different questions. In comparison, our $\texttt{SkillQG}$ is able to tailor a fine-grained assessment and improvement to the capabilities of question answering models built on it. Specifically, we first frame the comprehension type of questions based on a hierarchical skill-based schema, then formulate $\texttt{SkillQG}$ as a skill-conditioned question generator. Furthermore, to improve the controllability of generation, we augment the input text with question focus and skill-specific knowledge, which are constructed by iteratively prompting the pre-trained language models. Empirical results demonstrate that $\texttt{SkillQG}$ outperforms baselines in terms of quality, relevance, and skill-controllability while showing a promising performance boost in downstream question answering task. | ['Lingfei Wu', 'Siliang Tang', 'Bang Liu', 'Xiaoqiang Wang'] | 2023-05-08 | null | null | null | null | ['reading-comprehension', 'question-generation', 'machine-reading-comprehension'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 6.15677714e-01 5.31538785e-01 3.07326615e-01 -6.28866196e-01
-1.17318976e+00 -8.82268131e-01 4.99743193e-01 3.58732164e-01
-1.79088920e-01 5.86993039e-01 5.63981652e-01 -7.32599318e-01
-3.47547561e-01 -1.36914015e+00 -7.38528907e-01 -1.24852359e-02
4.56296712e-01 6.10383272e-01 3.22172672e-01 -8.37636590e-01
3.76028925e-01 -2.78887957e-01 -1.67289877e+00 4.95795429e-01
1.70977187e+00 1.01167881e+00 5.11795759e-01 8.42230260e-01
-6.19657159e-01 1.12818229e+00 -7.05531538e-01 -6.71788275e-01
-2.05943361e-01 -1.05328476e+00 -1.42011642e+00 -4.14640427e-01
5.58827937e-01 -3.43049735e-01 -1.23510007e-02 8.79006982e-01
2.36588851e-01 4.08924282e-01 5.60262024e-01 -6.97694182e-01
-9.32606041e-01 7.51127064e-01 2.56369323e-01 4.80624616e-01
1.02826905e+00 4.72550511e-01 1.35009253e+00 -6.44927561e-01
2.55522728e-01 1.14256072e+00 1.54408678e-01 8.08171332e-01
-1.02803648e+00 -4.21474010e-01 4.67021167e-01 2.92144567e-01
-7.31537223e-01 -1.69608310e-01 4.76930171e-01 -3.43926936e-01
1.01464212e+00 5.85194290e-01 4.94752616e-01 8.53109837e-01
-3.39015797e-02 7.99101949e-01 1.27163768e+00 -6.11131370e-01
1.79405212e-01 -2.93852776e-01 5.78396916e-01 8.98166001e-01
-1.23563213e-02 2.20967294e-03 -6.69086039e-01 2.46684849e-01
6.39594913e-01 -3.92472595e-01 -2.49450624e-01 3.29650909e-01
-1.20471048e+00 8.82409096e-01 2.82238573e-01 2.29013711e-01
-2.87325859e-01 1.91352367e-01 1.14541374e-01 5.32515645e-01
2.23643184e-01 1.27838087e+00 -6.01883769e-01 -3.01739842e-01
-7.11867571e-01 5.57250261e-01 7.83741355e-01 1.28772056e+00
8.42862546e-01 5.04218042e-03 -1.21992707e+00 7.29948521e-01
1.34922907e-01 9.45840895e-01 4.69701320e-01 -1.30901349e+00
7.89494514e-01 1.03239751e+00 9.45009440e-02 -6.26294911e-01
-2.13508636e-01 -4.36370969e-01 -3.87230277e-01 -3.35861266e-01
5.13511777e-01 -3.00453514e-01 -9.98166382e-01 2.07008958e+00
1.89976599e-02 -3.66725922e-01 -3.43962424e-02 5.98296642e-01
1.23672760e+00 6.46974564e-01 4.47972804e-01 5.72973974e-02
1.71493685e+00 -1.04956543e+00 -6.74519658e-01 -7.00039804e-01
7.06769943e-01 -3.19900513e-01 1.91104925e+00 1.85231999e-01
-1.50100362e+00 -7.31158435e-01 -6.80396676e-01 -3.17475468e-01
-2.37985015e-01 -3.02642137e-01 3.56951922e-01 5.60338795e-01
-1.09767008e+00 2.26121694e-01 -1.40332311e-01 -9.51800719e-02
1.80547297e-01 -6.91132108e-03 3.46843213e-01 -3.70966673e-01
-1.70076251e+00 9.07033920e-01 3.45563859e-01 -2.86707938e-01
-1.09006310e+00 -8.36011112e-01 -1.01521397e+00 3.46178323e-01
7.16142118e-01 -1.12936258e+00 1.69290245e+00 -5.26862204e-01
-1.54797864e+00 8.28427553e-01 -3.69337350e-01 -1.86248228e-01
8.65399763e-02 -1.93647236e-01 -2.25628570e-01 3.91270280e-01
4.35551167e-01 7.71734238e-01 6.79427266e-01 -1.01064646e+00
-7.99393296e-01 -1.50685668e-01 9.00633454e-01 4.87494648e-01
-2.03113496e-01 -1.27850279e-01 -8.64213333e-02 -6.39315784e-01
1.89214915e-01 -4.16190684e-01 -8.89109075e-02 -6.12616539e-01
-2.78201163e-01 -6.39998794e-01 -1.06357671e-01 -7.38879919e-01
1.55071616e+00 -1.55086291e+00 1.17481664e-01 1.15807459e-01
3.25118423e-01 2.20459893e-01 -5.03412485e-01 4.51721132e-01
2.06017897e-01 3.35396737e-01 -3.01326454e-01 8.45443904e-02
3.14272553e-01 1.50024116e-01 -3.55638891e-01 -7.48754740e-01
3.55242044e-01 1.45794153e+00 -1.37049425e+00 -4.49620128e-01
-2.92527229e-02 -3.84170622e-01 -7.99455822e-01 8.00457418e-01
-1.14312291e+00 4.36712533e-01 -7.88355649e-01 4.44980174e-01
2.86846217e-02 -3.54731560e-01 -2.67165691e-01 2.57949769e-01
3.92641693e-01 1.02781618e+00 -6.64125979e-01 1.82590246e+00
-7.75563896e-01 -1.07866511e-01 -2.85985142e-01 -6.74014509e-01
1.04252315e+00 1.94057435e-01 -2.25068569e-01 -1.02259338e+00
-2.35197367e-03 9.34558660e-02 4.25588787e-02 -7.76718438e-01
6.64721906e-01 -4.26445335e-01 -4.05034930e-01 7.45145082e-01
2.57043958e-01 -9.02873337e-01 5.68363249e-01 4.98321891e-01
1.25654650e+00 2.58622587e-01 -1.01665571e-01 -4.12741512e-01
5.93428969e-01 1.21641636e-01 -1.59315839e-01 1.21025133e+00
-3.44256312e-02 5.64515829e-01 5.64788222e-01 2.84191787e-01
-6.27783537e-01 -1.29702306e+00 3.30869019e-01 1.79834449e+00
6.03593327e-03 -4.16295320e-01 -1.13859844e+00 -8.00649226e-01
-3.86399835e-01 1.31678295e+00 -6.68011248e-01 -4.97639805e-01
-6.27038598e-01 -1.96039394e-01 7.73201585e-01 6.70072794e-01
7.49481797e-01 -1.62782109e+00 -7.14225888e-01 3.09802383e-01
-7.97570348e-01 -8.60334575e-01 -5.67675531e-01 -7.93689415e-02
-8.43158841e-01 -8.74591053e-01 -3.82316530e-01 -8.12117994e-01
6.12175047e-01 1.85812458e-01 1.88689852e+00 4.64953184e-01
3.43783855e-01 7.14563310e-01 -8.11424851e-01 -4.73565847e-01
-4.84377772e-01 3.62727672e-01 -5.74594557e-01 -5.03501475e-01
4.58789229e-01 -3.99187773e-01 -8.94383311e-01 7.70646185e-02
-1.29604256e+00 2.17155933e-01 4.05164242e-01 6.56543374e-01
3.79688263e-01 -4.01012450e-01 9.86497045e-01 -9.79681551e-01
1.37095010e+00 -4.30369973e-01 -5.90544455e-02 5.00986576e-01
-6.84738457e-01 2.71432549e-01 7.30049253e-01 -1.04015119e-01
-1.35376263e+00 -6.79626226e-01 -5.40587127e-01 2.86908418e-01
-3.71624291e-01 7.71617711e-01 -2.67973959e-01 5.89875579e-01
9.95829999e-01 6.31614089e-01 -1.72365576e-01 -1.34931564e-01
9.55897391e-01 1.86013058e-01 6.35411799e-01 -1.22750878e+00
7.97803819e-01 -1.40862867e-01 -3.89319539e-01 -2.06556737e-01
-1.50108373e+00 -2.00066492e-01 -1.35641247e-01 -2.09619254e-01
1.06743562e+00 -5.35767436e-01 -7.23067999e-01 3.00967753e-01
-9.00492847e-01 -6.68866217e-01 -7.66514540e-01 -1.68306082e-01
-7.16287494e-01 2.28582159e-01 -5.67063212e-01 -6.61722362e-01
-5.91303349e-01 -9.25849617e-01 1.11298525e+00 4.23861265e-01
-6.35602891e-01 -1.15674996e+00 -2.23172665e-01 8.73221040e-01
6.29248261e-01 -3.35872211e-02 1.62819052e+00 -7.56646276e-01
-6.03768110e-01 -8.17431509e-02 -1.96021065e-01 5.05520046e-01
1.07656129e-01 -7.84384131e-01 -7.57551551e-01 1.27583370e-01
2.27036342e-01 -7.56750226e-01 6.36246860e-01 9.46078449e-02
1.32408774e+00 -6.50431335e-01 3.06829274e-01 1.24020033e-01
1.00327218e+00 1.65743619e-01 7.65219867e-01 1.48361266e-01
4.24329907e-01 9.38864529e-01 5.73497355e-01 4.69440129e-03
9.75106239e-01 1.47484720e-01 3.84224176e-01 2.46035218e-01
-2.16142952e-01 -8.08857381e-01 2.12766707e-01 7.13835955e-01
1.62169620e-01 -1.85651362e-01 -9.81998742e-01 7.05739915e-01
-1.39571106e+00 -8.38485301e-01 -2.32710719e-01 1.85822356e+00
1.40976346e+00 1.30184814e-01 -1.45667240e-01 -4.77931043e-03
1.63163722e-01 2.11784422e-01 -4.69078183e-01 -4.54246342e-01
-6.98631182e-02 9.60900486e-01 -4.17744994e-01 8.62194419e-01
-2.40708396e-01 1.18343735e+00 5.37461710e+00 7.97674656e-01
-5.14667511e-01 2.05880683e-03 5.84127247e-01 1.30086198e-01
-1.13542020e+00 4.27716523e-02 -6.40626431e-01 3.62425327e-01
8.01334918e-01 -3.38358730e-01 4.02154714e-01 2.63885319e-01
1.05063431e-01 -2.39418358e-01 -1.24936664e+00 3.47846568e-01
2.12126642e-01 -1.04316616e+00 6.04696631e-01 -4.79427755e-01
6.08440280e-01 -6.86635792e-01 1.02978081e-01 8.79785419e-01
5.56444347e-01 -1.35007000e+00 7.69774854e-01 6.99465036e-01
8.56791794e-01 -4.32415247e-01 4.75847334e-01 6.70883358e-01
-9.98079062e-01 -1.38313234e-01 1.16827125e-02 -5.10793686e-01
2.80193925e-01 3.20352167e-01 -7.04279840e-01 6.45287752e-01
6.02168620e-01 4.28773463e-03 -9.56329823e-01 3.33247662e-01
-9.52598214e-01 9.90168631e-01 1.02948703e-01 -5.20071745e-01
1.09537639e-01 -1.28924713e-01 2.47894585e-01 9.27652001e-01
2.81209856e-01 8.53929818e-01 4.00747806e-02 1.07159793e+00
-1.88505471e-01 1.71704888e-01 -1.16328016e-01 -5.45361312e-03
5.43623209e-01 8.35184395e-01 -1.67210504e-01 -6.80718124e-01
-5.16555533e-02 7.60237694e-01 3.89824748e-01 4.40482050e-01
-4.43024158e-01 -5.18003047e-01 2.46019557e-01 3.67094517e-01
-6.90988777e-03 -5.34961335e-02 -6.16213322e-01 -1.23032808e+00
1.93226069e-01 -1.26819718e+00 5.25353312e-01 -1.19925928e+00
-1.12979078e+00 6.03416800e-01 3.21265727e-01 -4.60661143e-01
-4.60952878e-01 -5.05356789e-01 -5.95840991e-01 1.24689853e+00
-1.52693892e+00 -1.02484727e+00 -5.40859163e-01 6.40163004e-01
8.09379637e-01 1.14198215e-01 8.83423805e-01 -3.73732567e-01
-2.98357643e-02 6.68086946e-01 -5.50976098e-01 -1.70062874e-02
3.28474760e-01 -1.62706721e+00 5.70706785e-01 9.10120308e-01
-2.03746855e-01 8.65108371e-01 6.54323876e-01 -6.58615589e-01
-1.26630998e+00 -1.11621869e+00 1.33116424e+00 -9.35985804e-01
5.11057377e-01 -2.66512573e-01 -1.13878810e+00 6.32169425e-01
5.27723312e-01 -7.78166413e-01 7.11208522e-01 7.16876388e-02
-4.11411375e-01 -1.32082682e-02 -1.14451385e+00 7.76506901e-01
1.44399250e+00 -6.57499135e-01 -1.42195559e+00 3.02669227e-01
1.27470171e+00 -4.02678281e-01 -8.65878522e-01 4.16108519e-01
-1.53087713e-02 -9.61232245e-01 7.17836976e-01 -9.15528893e-01
9.10510659e-01 -1.06745854e-01 -7.13589713e-02 -1.34025300e+00
-4.72547382e-01 -3.84620488e-01 -1.54174715e-01 1.21204591e+00
7.66765833e-01 -4.94278908e-01 5.74911058e-01 8.32983434e-01
-4.52453315e-01 -8.12170446e-01 -6.37111008e-01 -3.47343892e-01
7.06090569e-01 -5.07256269e-01 1.05925715e+00 5.89350641e-01
1.58312887e-01 6.89299166e-01 2.40041167e-01 -1.72885403e-01
2.60251909e-01 1.64973140e-01 4.51700300e-01 -9.75794613e-01
-4.11061317e-01 -5.57989776e-01 5.65415025e-01 -1.61876822e+00
-4.26975898e-02 -1.01081049e+00 2.98415959e-01 -2.15552044e+00
2.57446207e-02 -4.10643309e-01 -3.20451148e-02 4.64839518e-01
-9.97568429e-01 -3.34769905e-01 2.90832341e-01 -3.00267667e-01
-7.37689018e-01 4.47765529e-01 1.81630611e+00 -2.45188922e-03
1.70634463e-01 -5.32305129e-02 -1.40017724e+00 5.08477807e-01
7.23003924e-01 2.20164694e-02 -9.00999308e-01 -8.57260525e-01
7.24271476e-01 4.99509424e-01 3.03913295e-01 -7.50743032e-01
1.50108173e-01 -4.59580958e-01 -6.19196296e-02 -2.07111239e-01
6.30222037e-02 -1.11085542e-01 -6.82268143e-01 2.79469997e-01
-9.44205642e-01 2.64946252e-01 1.07512586e-01 2.54994869e-01
-1.77961707e-01 -6.24162674e-01 3.47395271e-01 -5.97256243e-01
-3.55609715e-01 9.09105912e-02 -4.39576596e-01 8.83977890e-01
4.45501477e-01 -1.31085828e-01 -5.47819495e-01 -8.00481379e-01
-4.63202596e-01 7.46858835e-01 2.75414977e-02 4.50141907e-01
7.76548207e-01 -1.03637612e+00 -1.00038731e+00 -1.43418079e-02
4.54593331e-01 5.98094165e-01 4.12217140e-01 8.19566101e-02
-2.48691559e-01 5.75262070e-01 1.25462532e-01 -1.97154924e-01
-6.34308517e-01 3.03429246e-01 5.27652621e-01 -7.30597496e-01
9.47190374e-02 1.26315236e+00 2.54085720e-01 -6.52272284e-01
6.41656816e-02 -8.19610715e-01 -4.99893308e-01 -8.87400880e-02
5.16612411e-01 2.48376518e-01 3.71239185e-02 1.05892159e-01
1.84461460e-01 2.94531226e-01 2.41190288e-02 -3.84132028e-01
6.40749156e-01 -2.04985842e-01 -2.43616804e-01 7.40900263e-02
6.44773722e-01 -1.02577291e-01 -9.97273743e-01 -2.72175819e-01
1.05340473e-01 -1.52819559e-01 -6.78878069e-01 -1.52786899e+00
-4.49367762e-01 9.21018183e-01 2.12083682e-02 4.65606362e-01
1.34773982e+00 2.97417045e-01 9.25938129e-01 4.45636868e-01
3.95900846e-01 -8.10118854e-01 7.32727587e-01 1.04940438e+00
1.08308589e+00 -8.44704449e-01 -6.52010024e-01 -4.18527633e-01
-5.69850624e-01 7.15804160e-01 1.18689251e+00 1.11905709e-01
1.35641778e-02 -3.38691831e-01 2.35556617e-01 -2.86871135e-01
-7.85111189e-01 -5.12831211e-01 3.78685355e-01 5.64686954e-01
5.87504089e-01 -2.59835683e-02 -5.19109249e-01 1.05597520e+00
-1.00348055e+00 -3.77577730e-02 3.19791317e-01 9.83562768e-01
-8.38659763e-01 -1.17818260e+00 -1.63837507e-01 9.36312735e-01
-1.97012737e-01 -6.69354558e-01 -3.93481225e-01 1.53920472e-01
1.32852852e-01 1.50747788e+00 -3.46774220e-01 -2.51317769e-01
8.88810575e-01 4.68839496e-01 6.00166440e-01 -1.19675994e+00
-1.15923226e+00 -6.88787162e-01 3.34542066e-01 -2.61781067e-01
6.47895932e-02 -4.15998220e-01 -1.44316685e+00 -3.75065161e-03
-1.16684221e-01 5.76302052e-01 2.16671266e-02 1.34746301e+00
3.06955725e-01 8.29649925e-01 3.91537249e-01 1.48651049e-01
-8.57875288e-01 -1.29615521e+00 -1.34543076e-01 8.19420934e-01
1.16148695e-01 -2.44466722e-01 -1.86181799e-01 7.36558214e-02] | [11.397351264953613, 8.103897094726562] |
938bdf9e-c3da-465a-9b42-9f73961b0152 | improving-aspect-extraction-based-on-rules | null | null | https://openreview.net/forum?id=20Lgo1A-aSh | https://openreview.net/pdf?id=20Lgo1A-aSh | Improving Aspect Extraction based on Rules through Deep Syntax-Semantics Communication | Recent studies show integrating language resources which consist of lexical resources, syntactic resources and semantic resources can improve the performance of natural language processing (NLP) tasks. The existing methods mostly perform simple integration through concatenating these resources successively, seldom consider complementary relationship among them, such as the deep communication of syntactic and semantic relations between words. To enhance deep syntax-semantics communication, this paper takes aspect term extraction (ATE) task as an example and explores four integration strategies of language resources. These strategies, based on Answer Set Programming (ASP) rules, have interpretability. Experiments on eight ATE datasets show that our strategies achieve superior performance, demonstrating that they are highly effective in integrating language resources. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['term-extraction', 'aspect-extraction'] | ['natural-language-processing', 'natural-language-processing'] | [-3.85663748e-01 -1.17378412e-02 -4.89485204e-01 -2.88663685e-01
-2.91063935e-01 -5.56647182e-01 5.24642646e-01 2.40176871e-01
-5.69434702e-01 6.60057485e-01 5.54269671e-01 -5.44352710e-01
-1.94128275e-01 -1.16951787e+00 -3.47239941e-01 -7.67858922e-02
1.27718955e-01 4.12501007e-01 6.48627162e-01 -5.11363387e-01
1.30622849e-01 2.78841890e-02 -1.24073398e+00 6.10290170e-01
1.15084553e+00 9.16090190e-01 1.06127635e-01 -1.96577758e-01
-1.46446919e+00 8.30995023e-01 -5.14810324e-01 -6.02242768e-01
2.77752560e-02 -7.54420236e-02 -1.14752996e+00 -4.85345274e-01
-6.62158370e-01 5.43686152e-02 1.70703694e-01 1.36209619e+00
2.72654057e-01 6.52562678e-02 4.46598887e-01 -1.24687219e+00
-7.99699843e-01 1.09469450e+00 -3.55146974e-01 2.15820804e-01
7.88992167e-01 -1.19235784e-01 1.33002841e+00 -9.13734019e-01
6.67393029e-01 1.73116648e+00 6.47164345e-01 2.46170878e-01
-6.80975318e-01 -6.72147274e-01 3.58513534e-01 4.40512151e-01
-1.29010046e+00 -6.04915880e-02 6.85425580e-01 1.90053843e-02
1.85134399e+00 2.62649328e-01 4.68821377e-01 6.36363328e-01
9.08951014e-02 7.41340041e-01 1.05192006e+00 -6.60464883e-01
-2.35991716e-01 3.41636151e-01 8.08106244e-01 7.40696669e-01
5.65836668e-01 -2.33786300e-01 -4.11282927e-01 -1.92397073e-01
4.09136891e-01 -1.05688125e-01 -1.01168871e-01 3.88273925e-01
-9.84875441e-01 6.84663653e-01 3.48402590e-01 6.79751396e-01
-1.92608297e-01 -2.74717033e-01 6.73097253e-01 2.69167155e-01
3.46493483e-01 5.21445572e-01 -8.95715535e-01 3.29665065e-01
-1.13761380e-01 6.73490344e-03 1.17069697e+00 1.36062193e+00
7.70546675e-01 -3.49950016e-01 -1.31229490e-01 9.50071335e-01
7.27193058e-01 5.78020275e-01 9.10201132e-01 -6.13366067e-01
6.82687640e-01 1.30473256e+00 -2.11942598e-01 -1.21327186e+00
-4.69757199e-01 -6.86810538e-02 -4.67457920e-01 -5.42334616e-01
-1.51815951e-01 -2.75226235e-01 -5.99207342e-01 1.54219472e+00
4.07682180e-01 -2.41266385e-01 3.15550804e-01 6.26813114e-01
1.68165588e+00 8.71247530e-01 5.93221307e-01 -4.59063530e-01
1.97505033e+00 -1.13458514e+00 -1.13069391e+00 -4.27225173e-01
7.73434401e-01 -8.98182452e-01 1.24466848e+00 9.69516337e-02
-9.20836985e-01 -3.31178874e-01 -9.38706756e-01 -3.46012563e-01
-1.21429157e+00 5.66195399e-02 8.77243698e-01 3.34631741e-01
-6.00079775e-01 2.71019965e-01 -3.46617877e-01 -4.10281479e-01
1.55858234e-01 2.86317915e-01 -2.20625013e-01 1.15587376e-01
-1.96700346e+00 8.39267015e-01 1.13429761e+00 -8.79793018e-02
4.70147133e-02 -5.13760567e-01 -9.23174202e-01 1.85739949e-01
8.95662844e-01 -9.58217680e-01 9.61000085e-01 -8.85417342e-01
-1.31864488e+00 8.45528185e-01 -1.81949511e-01 -4.03091311e-01
-2.03583047e-01 -4.11302805e-01 -8.51898193e-01 4.52722609e-02
1.89941376e-01 5.52643716e-01 1.99273359e-02 -7.35583842e-01
-5.49971461e-01 -3.02568495e-01 3.80942911e-01 3.62951726e-01
-5.28964221e-01 8.95919621e-01 -5.82695127e-01 -5.12431920e-01
1.05477951e-01 -2.92957067e-01 -7.09846988e-02 -3.66671622e-01
-4.71292466e-01 -8.44267607e-01 4.33779269e-01 -6.95224166e-01
1.74483204e+00 -1.72120512e+00 2.91808788e-03 1.76545784e-01
7.84322768e-02 4.41903502e-01 8.12095329e-02 5.44986665e-01
1.80813879e-01 7.92242885e-01 -2.75494307e-02 1.56205103e-01
2.55024582e-01 5.42571366e-01 -4.53727543e-01 -4.83903915e-01
8.88812914e-02 1.30272865e+00 -6.51497006e-01 -9.92989600e-01
-3.55177969e-02 -1.09113388e-01 -3.90820980e-01 9.68408585e-02
-7.61540055e-01 -1.24200769e-01 -8.83940518e-01 6.35293305e-01
3.88229907e-01 -2.51783729e-01 1.56375557e-01 -2.95596868e-01
2.27559492e-01 7.40502715e-01 -9.81379390e-01 1.52945983e+00
-6.14122212e-01 -1.60715859e-02 -2.68748343e-01 -7.43765354e-01
1.12245071e+00 3.29517812e-01 1.39627501e-01 -7.62958288e-01
2.89213151e-01 3.39565456e-01 -3.52065004e-02 -1.11152160e+00
1.84156269e-01 -3.59869242e-01 -6.78245351e-02 3.50800157e-01
1.69225410e-01 -2.65936311e-02 3.09538305e-01 1.47994086e-01
8.49391460e-01 2.24117190e-01 8.83860111e-01 -3.73164266e-01
1.11964464e+00 2.74501175e-01 7.62747407e-01 1.28070787e-01
-3.89468297e-02 -6.34092316e-02 6.06586874e-01 -3.32112700e-01
-6.91250920e-01 -7.01092899e-01 1.52574982e-02 1.01216662e+00
3.94663393e-01 -9.95161533e-01 -5.58760583e-01 -9.30832803e-01
-3.22261155e-01 8.35380912e-01 -1.57370642e-01 -1.32372186e-01
-6.56068325e-01 -7.78439939e-01 4.88571495e-01 6.88267052e-01
1.05856311e+00 -1.28217697e+00 -1.31115764e-01 1.60282612e-01
-3.57631743e-01 -1.38980174e+00 -8.34520832e-02 -1.43721834e-01
-7.32716322e-01 -1.06041193e+00 1.01687029e-01 -9.85230267e-01
3.68416488e-01 2.30055735e-01 1.39264250e+00 5.63519597e-01
2.61539686e-02 6.27660304e-02 -5.58379471e-01 -6.33675992e-01
-1.41003236e-01 7.27863237e-02 -1.66016966e-01 -4.12340403e-01
1.17816019e+00 -5.52430332e-01 -2.18571082e-01 1.79985315e-01
-1.03280866e+00 1.49754047e-01 4.56021696e-01 4.89902318e-01
5.50415158e-01 8.72110948e-02 6.09301686e-01 -1.08475924e+00
9.78764832e-01 -7.27526784e-01 -3.26923698e-01 6.99765265e-01
-7.51334667e-01 3.07262182e-01 7.58192122e-01 -8.48212019e-02
-1.49956262e+00 -6.04695320e-01 -2.75028884e-01 4.21623848e-02
-1.46040902e-01 1.01092660e+00 -7.35544205e-01 2.43029118e-01
3.94026339e-01 9.14668590e-02 -2.71860540e-01 -6.18234813e-01
4.87998515e-01 6.40966594e-01 7.13573545e-02 -1.04997325e+00
3.80887359e-01 1.75260603e-01 -1.08480491e-01 -2.32207760e-01
-9.92965281e-01 -5.45655310e-01 -4.97531682e-01 1.92347065e-01
1.01831174e+00 -6.59317255e-01 -6.41155779e-01 -2.03471139e-01
-1.65652657e+00 5.45843005e-01 -1.24447450e-01 5.13561070e-01
-8.30733255e-02 3.82279068e-01 -5.78386724e-01 -4.31600362e-01
-6.35643423e-01 -8.70767295e-01 9.53603089e-01 3.73710811e-01
-2.41788834e-01 -1.30840492e+00 -1.66113839e-01 1.86560541e-01
2.61769056e-01 -8.32438394e-02 1.55624509e+00 -1.35247731e+00
-4.18208569e-01 5.18522747e-02 -3.44607532e-01 2.40157858e-01
1.31693766e-01 -1.07512191e-01 -5.38169444e-01 4.63895500e-01
2.11178258e-01 -4.77819778e-02 7.15978980e-01 -7.13377893e-02
1.08674717e+00 -7.36313879e-01 -4.66728657e-01 2.50755757e-01
1.59864831e+00 3.82650852e-01 5.95219314e-01 5.09049535e-01
4.13959265e-01 1.04314983e+00 6.25045002e-01 -3.07860784e-02
5.71316898e-01 2.11579621e-01 6.78448156e-02 3.47457200e-01
-1.91665351e-01 -3.99314582e-01 3.79331619e-01 1.48059356e+00
-9.09643695e-02 -1.24676414e-01 -1.12765515e+00 1.29252031e-01
-1.97297716e+00 -5.62315047e-01 -1.38032526e-01 1.60258210e+00
1.19379354e+00 2.52571255e-01 -1.06771015e-01 -1.21997975e-01
5.84424853e-01 3.30436490e-02 -1.91717088e-01 -4.43140298e-01
-4.40184504e-01 1.06470242e-01 6.48800135e-02 5.92121720e-01
-8.49154174e-01 1.19267142e+00 6.01535940e+00 1.27869081e+00
-4.55942482e-01 2.36842722e-01 2.84751236e-01 3.32671165e-01
-8.01029861e-01 1.51892543e-01 -1.18898487e+00 3.66009861e-01
7.90484905e-01 -7.05719352e-01 2.26552799e-01 6.32675171e-01
-2.35025108e-01 1.29389673e-01 -9.05945957e-01 7.89964736e-01
2.31372677e-02 -1.31543934e+00 6.08011544e-01 -3.83495599e-01
2.20475852e-01 -2.84328520e-01 -4.32858646e-01 6.47820413e-01
4.41444695e-01 -1.02940714e+00 1.75400779e-01 6.00257397e-01
1.13415383e-01 -6.01515174e-01 1.06213450e+00 4.43638384e-01
-1.52053607e+00 9.37099531e-02 -5.69275260e-01 -1.37316972e-01
2.10871577e-01 5.21382213e-01 -3.12684864e-01 1.10053790e+00
6.36044383e-01 6.84753835e-01 -6.39605939e-01 6.28258109e-01
-7.53516555e-01 3.81836146e-01 -5.65934002e-01 -5.21314561e-01
3.07390958e-01 -3.28203499e-01 5.86539865e-01 1.22059357e+00
7.66145736e-02 5.94895780e-01 3.04765135e-01 1.18715382e+00
2.32379045e-02 8.15639675e-01 -7.40235686e-01 -3.81858833e-02
6.77433014e-01 1.07593179e+00 -7.10108697e-01 -6.94767058e-01
-8.01348805e-01 1.77250400e-01 3.75960112e-01 3.49260747e-01
-8.63071918e-01 -5.97507417e-01 3.05220723e-01 -3.13167840e-01
-1.03143193e-01 -1.97611928e-01 -4.91329014e-01 -1.27081919e+00
3.64821583e-01 -8.35426450e-01 8.82261038e-01 -1.04344344e+00
-1.36909950e+00 5.76503873e-01 4.15119410e-01 -1.10699856e+00
4.77124974e-02 -7.57451475e-01 -5.62189639e-01 9.72579956e-01
-1.53606379e+00 -1.16200674e+00 6.04838096e-02 5.55537879e-01
6.61386967e-01 -2.47323632e-01 8.33753049e-01 2.38512531e-01
-6.23115659e-01 1.13704659e-01 -6.21557355e-01 1.16549172e-01
1.88813150e-01 -9.19390380e-01 -6.06156029e-02 7.57659018e-01
3.51980239e-01 1.19317102e+00 3.65399361e-01 -7.17097640e-01
-1.43052042e+00 -7.80359805e-01 1.38705361e+00 -2.00287834e-01
9.07186627e-01 1.43363968e-01 -1.15681469e+00 8.36550832e-01
7.08497226e-01 -2.74236828e-01 1.06420946e+00 2.83690840e-01
-5.70756376e-01 3.25648263e-02 -9.89265978e-01 7.03429103e-01
1.19165921e+00 -4.20196265e-01 -1.52212358e+00 4.44994777e-01
1.54281545e+00 -6.75535873e-02 -7.60774791e-01 7.77805567e-01
1.88277841e-01 -5.86501300e-01 1.09732807e+00 -1.16700828e+00
6.31415009e-01 -4.68368083e-01 -1.43232569e-01 -1.05913186e+00
-4.65257689e-02 -3.55147600e-01 -2.50796258e-01 1.73975813e+00
7.32105494e-01 -9.21112418e-01 1.03106074e-01 8.17496896e-01
1.62200734e-01 -9.14811194e-01 -5.59175789e-01 -9.33372378e-01
1.90174520e-01 -5.67458212e-01 1.14039457e+00 9.70249355e-01
4.98646170e-01 1.02408397e+00 3.79595280e-01 1.43301869e-02
-7.30350167e-02 3.78709942e-01 2.88054854e-01 -1.11429536e+00
-2.89326698e-01 -7.60841906e-01 -1.33719379e-02 -1.02400994e+00
5.68253458e-01 -1.28302097e+00 -4.28376138e-01 -1.79760826e+00
2.08791599e-01 -3.81131381e-01 -3.35873276e-01 7.43825018e-01
-4.07631636e-01 -4.44084287e-01 -3.04973852e-02 2.59805858e-01
-7.45363891e-01 6.41454935e-01 1.25748062e+00 -1.95296079e-01
4.16544527e-02 -3.37268949e-01 -8.46987069e-01 1.09805274e+00
7.64662683e-01 -3.51890951e-01 -5.18360794e-01 -7.50093341e-01
8.33281934e-01 -1.38231456e-01 -1.69013351e-01 -4.60927010e-01
5.18004477e-01 -4.22352523e-01 -1.99643791e-01 -3.69202137e-01
-1.14183418e-01 -1.18257785e+00 -1.46555424e-01 3.93815160e-01
-3.21133703e-01 2.10810542e-01 4.64169830e-01 3.00012112e-01
-5.96091092e-01 -8.24608326e-01 9.15707648e-02 -4.74319726e-01
-1.04829943e+00 2.02596217e-01 9.66719631e-03 4.36844677e-01
9.31684196e-01 1.60829842e-01 -4.66878593e-01 1.44820839e-01
-6.79390371e-01 5.29054105e-01 -2.93125868e-01 5.80919027e-01
4.83217478e-01 -1.45533061e+00 -3.93063694e-01 1.02999441e-01
2.81107239e-02 1.74948260e-01 -2.45128140e-01 6.88367069e-01
-7.86702573e-01 7.45173693e-01 1.51363891e-02 -1.73837692e-01
-1.31100667e+00 8.55345249e-01 3.39488894e-01 -7.02998877e-01
-4.89452362e-01 9.36300993e-01 2.35047057e-01 -6.32940888e-01
1.67954460e-01 -5.71554422e-01 -9.22335982e-01 1.01375498e-01
5.90798676e-01 1.26426235e-01 -1.19433858e-01 -3.98874968e-01
-5.75931072e-01 7.91857183e-01 2.18262058e-02 -1.07366562e-01
9.75895524e-01 -2.55136192e-01 -9.41187322e-01 4.74563718e-01
1.08181405e+00 1.37608014e-02 2.91389935e-02 -7.67128408e-01
5.64127803e-01 -3.12068641e-01 -3.04377407e-01 -8.39859545e-01
-9.25033510e-01 8.80865276e-01 -1.40921518e-01 4.62908745e-01
1.15287542e+00 1.87137514e-01 1.08744168e+00 5.87433338e-01
5.49810946e-01 -9.73219037e-01 -2.27525994e-01 7.85460830e-01
9.56776381e-01 -1.04239237e+00 5.19001931e-02 -1.11065757e+00
-3.22342098e-01 1.40228975e+00 8.95493746e-01 8.81006103e-03
9.45785105e-01 4.18539584e-01 -2.38565251e-01 -4.51391250e-01
-9.65268552e-01 -4.33324903e-01 3.74896437e-01 2.69500226e-01
5.44020474e-01 2.11210810e-02 -1.13551426e+00 1.47325337e+00
-2.25066423e-01 -7.37621486e-02 -2.02995971e-01 8.45957339e-01
-6.54961169e-01 -1.33704507e+00 -2.02766880e-02 3.23719978e-01
-4.73432839e-01 -6.70637906e-01 -5.76380074e-01 7.82866299e-01
3.61879796e-01 8.86514604e-01 -4.15390953e-02 -2.92811006e-01
4.04020011e-01 2.95974284e-01 2.19080165e-01 -8.75814676e-01
-8.69303703e-01 -1.61977991e-01 4.64245796e-01 -5.75168133e-01
-6.61818326e-01 -2.08328411e-01 -1.66377962e+00 -2.68145531e-01
-2.82735795e-01 3.90217543e-01 1.46964878e-01 1.43399596e+00
3.53877515e-01 7.21344173e-01 1.73406631e-01 4.16085273e-01
-3.51994038e-01 -8.89736116e-01 -2.63991535e-01 3.45734656e-01
-4.31253076e-01 -4.70917195e-01 -2.07562625e-01 -7.09241256e-02] | [9.928942680358887, 8.898032188415527] |
5225bb0a-f4d9-41b8-844e-bcb48e104db8 | msw-transformer-multi-scale-shifted-windows | 2306.12098 | null | https://arxiv.org/abs/2306.12098v1 | https://arxiv.org/pdf/2306.12098v1.pdf | MSW-Transformer: Multi-Scale Shifted Windows Transformer Networks for 12-Lead ECG Classification | Automatic classification of electrocardiogram (ECG) signals plays a crucial role in the early prevention and diagnosis of cardiovascular diseases. While ECG signals can be used for the diagnosis of various diseases, their pathological characteristics exhibit minimal variations, posing a challenge to automatic classification models. Existing methods primarily utilize convolutional neural networks to extract ECG signal features for classification, which may not fully capture the pathological feature differences of different diseases. Transformer networks have advantages in feature extraction for sequence data, but the complete network is complex and relies on large-scale datasets. To address these challenges, we propose a single-layer Transformer network called Multi-Scale Shifted Windows Transformer Networks (MSW-Transformer), which uses a multi-window sliding attention mechanism at different scales to capture features in different dimensions. The self-attention is restricted to non-overlapping local windows via shifted windows, and different window scales have different receptive fields. A learnable feature fusion method is then proposed to integrate features from different windows to further enhance model performance. Furthermore, we visualize the attention mechanism of the multi-window shifted mechanism to achieve better clinical interpretation in the ECG classification task. The proposed model achieves state-of-the-art performance on five classification tasks of the PTBXL-2020 12-lead ECG dataset, which includes 5 diagnostic superclasses, 23 diagnostic subclasses, 12 rhythm classes, 17 morphology classes, and 44 diagnosis classes, with average macro-F1 scores of 77.85%, 47.57%, 66.13%, 34.60%, and 34.29%, and average sample-F1 scores of 81.26%, 68.27%, 91.32%, 50.07%, and 63.19%, respectively. | ['Jingfeng Guo', 'Lei Xie', 'Shuxin Zhuang', 'Zhemin Zhuang', 'Renjie Cheng'] | 2023-06-21 | null | null | null | null | ['ecg-classification', 'classification-1'] | ['medical', 'methodology'] | [ 1.90557390e-01 -4.76916879e-01 -7.31764734e-02 -4.04782534e-01
-7.78302252e-01 -2.73814499e-01 -8.61987993e-02 2.59238541e-01
-2.88712502e-01 6.29939616e-01 -4.24789004e-02 -3.00175339e-01
-4.50384468e-01 -6.09818399e-01 -1.17086865e-01 -8.26926231e-01
-3.71310860e-01 2.20862195e-01 1.53822392e-01 -1.66078046e-01
-7.72685111e-02 4.21882898e-01 -1.02944052e+00 7.40990996e-01
9.07110035e-01 1.44002736e+00 -1.30187003e-02 6.35443866e-01
2.18175426e-01 3.78755718e-01 -9.05532360e-01 2.97973311e-04
-1.07662797e-01 -6.85468733e-01 -3.70683551e-01 -1.92771628e-01
-1.14185758e-01 -5.10728508e-02 -1.95790559e-01 8.58852804e-01
1.04365981e+00 -2.26089254e-01 6.06707692e-01 -9.72096324e-01
-5.10614872e-01 5.54055393e-01 -5.60271323e-01 6.26916289e-01
-6.61253333e-02 4.84122150e-02 8.08076620e-01 -7.13837266e-01
1.78646669e-01 8.63437831e-01 9.01370823e-01 3.27942431e-01
-1.20961654e+00 -9.70835865e-01 -7.13818669e-02 4.55784172e-01
-1.59437203e+00 -8.45893770e-02 7.22184718e-01 -3.91247272e-01
7.82544971e-01 4.93955702e-01 8.30091536e-01 7.86331117e-01
5.44650912e-01 5.99191308e-01 1.02757609e+00 -1.21034803e-02
-1.83398291e-01 -1.21152699e-01 1.90717191e-01 4.47575420e-01
1.59907937e-01 -1.87180087e-01 -2.24864379e-01 -1.41448125e-01
8.51243675e-01 4.96532202e-01 -6.65961564e-01 1.35309607e-01
-1.40862250e+00 5.84795654e-01 5.44524193e-01 4.54596013e-01
-4.64975208e-01 -3.71610850e-01 6.79286540e-01 3.03502589e-01
3.73279154e-01 3.38842213e-01 -7.58219659e-01 -8.96665230e-02
-6.27814114e-01 1.63865462e-01 1.65083408e-01 4.41711336e-01
4.00492251e-01 1.61138535e-01 -5.62225759e-01 9.33233440e-01
8.69330168e-02 6.38839483e-01 8.30898345e-01 -5.68545818e-01
5.26766241e-01 7.43201017e-01 -1.31951481e-01 -1.03694594e+00
-7.91153431e-01 -1.01403022e+00 -1.47843421e+00 -1.44786865e-01
8.53804201e-02 -1.36076445e-02 -8.18642020e-01 1.56534052e+00
1.07912116e-01 2.86847413e-01 8.21357295e-02 8.26796114e-01
1.10430682e+00 5.57268441e-01 6.39092624e-02 -4.51198131e-01
1.51007986e+00 -3.37827891e-01 -7.13019013e-01 4.68482524e-02
4.25877243e-01 -4.85421896e-01 8.85815501e-01 3.94450188e-01
-9.54720855e-01 -7.81829417e-01 -1.08241618e+00 2.99085885e-01
-8.48160312e-03 4.32581961e-01 1.97813079e-01 4.45859164e-01
-5.66636741e-01 7.08631396e-01 -7.89514065e-01 -1.63225643e-02
7.36193538e-01 3.75767559e-01 -9.58005115e-02 1.09037265e-01
-1.53870094e+00 5.80297768e-01 2.46741012e-01 3.51730555e-01
-5.77624977e-01 -8.13790619e-01 -6.84862256e-01 3.52867395e-01
-7.63721988e-02 -4.89507705e-01 8.13112020e-01 -7.49444008e-01
-1.05430722e+00 6.87836587e-01 -7.47396722e-02 -4.31032062e-01
3.65145028e-01 -7.08957687e-02 -7.10673690e-01 3.13845903e-01
1.03180602e-01 3.84455398e-02 6.56694889e-01 -6.17111564e-01
-6.38496339e-01 -6.29830718e-01 -1.69884816e-01 5.05955517e-02
-4.04033124e-01 8.70865136e-02 -1.57282054e-01 -8.97039056e-01
1.59519896e-01 -6.95670187e-01 -2.08318770e-01 -2.93001145e-01
-2.90799797e-01 -8.44048932e-02 7.44991660e-01 -7.39089549e-01
1.58632374e+00 -2.34130502e+00 9.68467817e-02 2.12258428e-01
6.76694751e-01 5.54204345e-01 -3.13264877e-02 -5.66053241e-02
-2.22777262e-01 1.21270269e-01 -2.77938396e-01 1.20689645e-01
-3.97111952e-01 -5.64543996e-03 -7.56690279e-02 2.90499896e-01
2.85756111e-01 8.16160738e-01 -5.71795523e-01 -4.59502131e-01
1.90030053e-01 6.72350764e-01 -2.28379712e-01 8.77152160e-02
5.64249218e-01 4.28199351e-01 -6.34593546e-01 6.47385836e-01
6.00042462e-01 -4.18347955e-01 1.50082469e-01 -5.18982768e-01
1.87103093e-01 7.06005096e-02 -1.05638993e+00 1.52421415e+00
-2.98477054e-01 4.63177621e-01 -1.64964989e-01 -1.30042374e+00
1.02824044e+00 6.97604537e-01 7.65262723e-01 -6.75746381e-01
2.92695880e-01 1.99220210e-01 6.59744680e-01 -6.59118235e-01
-2.53609329e-01 -3.73825669e-01 -1.52279576e-03 2.37917468e-01
-1.17866807e-01 4.50786620e-01 2.20239274e-02 -1.67438909e-01
1.04159844e+00 -3.47279400e-01 3.40664327e-01 -2.65535742e-01
5.72590053e-01 -5.80426455e-01 1.08409667e+00 6.92247450e-01
-3.93983334e-01 8.99833202e-01 4.19790685e-01 -8.89144838e-01
-4.65339869e-01 -1.02678394e+00 -6.01288736e-01 5.40642738e-01
8.97201598e-02 -4.80128288e-01 -4.14185107e-01 -5.30758560e-01
-1.23534523e-01 7.36259948e-03 -5.84394395e-01 -5.91372132e-01
-6.69081688e-01 -1.11201477e+00 6.92185104e-01 8.98474216e-01
5.48029542e-01 -1.26644886e+00 -1.12771714e+00 4.21127707e-01
-5.03936231e-01 -8.28920364e-01 -3.22700471e-01 8.54280815e-02
-1.09145963e+00 -1.19348884e+00 -9.24018025e-01 -7.13351130e-01
2.62577772e-01 -1.87862944e-02 1.13007903e+00 1.28628358e-01
-5.84249496e-01 -2.38322660e-01 -2.64791042e-01 -5.74452996e-01
2.19493657e-02 9.76386741e-02 -5.54286912e-02 3.28086406e-01
1.79796055e-01 -5.20878017e-01 -8.39371979e-01 4.15808082e-01
-5.31542063e-01 -9.72123295e-02 7.20839620e-01 1.20380127e+00
6.52730286e-01 -6.34617060e-02 9.52505350e-01 -6.83537364e-01
7.70305574e-01 -3.82628322e-01 -3.17683518e-02 2.45834410e-01
-6.52485669e-01 -4.67867821e-01 8.22711110e-01 -5.50870657e-01
-5.80465198e-01 -3.13112438e-01 -2.40471184e-01 -5.31814337e-01
-1.61565363e-01 5.33446074e-01 -2.22148106e-01 2.57716835e-01
6.42471731e-01 3.95104289e-01 -9.05043110e-02 -2.94965118e-01
-4.48353052e-01 8.03730130e-01 2.01146320e-01 -3.06479633e-01
4.11437452e-01 1.98466226e-01 -1.24141619e-01 -6.07988000e-01
-5.49424887e-01 -2.98983574e-01 -3.95726830e-01 2.97610294e-02
9.55052018e-01 -7.01232374e-01 -5.94212115e-01 5.67190945e-01
-9.32313204e-01 4.22830656e-02 -2.09586337e-01 5.67103326e-01
-7.42650554e-02 3.29584152e-01 -7.47025728e-01 -5.61429679e-01
-9.41357136e-01 -1.16512823e+00 9.44399655e-01 2.54525274e-01
-2.66052008e-01 -6.89145863e-01 -2.82053858e-01 1.78875253e-02
6.23567045e-01 5.11510849e-01 1.11173153e+00 -7.03445911e-01
-9.79690105e-02 -3.01232338e-01 -2.71392852e-01 4.95031118e-01
4.34625030e-01 -2.81415492e-01 -8.20448816e-01 -3.56652796e-01
-5.82524501e-02 -5.12189604e-02 7.47727275e-01 6.69503510e-01
1.49887323e+00 4.30078730e-02 -4.81883854e-01 7.60663152e-01
1.00650251e+00 8.08192909e-01 6.46291316e-01 1.03926957e-02
5.81804514e-01 7.61133432e-02 3.43898684e-01 4.91534770e-01
2.76415646e-01 4.17999804e-01 1.31746307e-01 -4.50810522e-01
1.25043258e-01 4.76376444e-01 -6.55947402e-02 8.12216759e-01
-3.77938628e-01 1.22420453e-01 -1.02413511e+00 5.51492810e-01
-1.64934063e+00 -1.00152600e+00 -1.22904681e-01 2.21095514e+00
7.48553038e-01 3.18942875e-01 6.96827024e-02 5.85537314e-01
8.10927510e-01 6.79404438e-02 -6.79733992e-01 -2.09955156e-01
-1.31114304e-01 6.00263000e-01 -1.14317194e-01 -4.29777913e-02
-1.28751850e+00 7.60867968e-02 5.09467077e+00 6.86513364e-01
-1.43083477e+00 -6.56278571e-03 8.56532753e-01 -4.96447757e-02
2.36333966e-01 -3.71701151e-01 -4.31311488e-01 6.40837848e-01
7.98898578e-01 -7.91298002e-02 3.12372409e-02 4.83068734e-01
2.64236629e-01 5.11990726e-01 -8.41244280e-01 1.31925476e+00
-9.06810090e-02 -1.22390401e+00 -1.03634216e-01 -1.90552458e-01
2.55153507e-01 2.25464944e-02 8.34052116e-02 3.42652291e-01
-6.02240920e-01 -1.07202268e+00 3.51495892e-01 5.18571973e-01
1.12870836e+00 -7.95863569e-01 1.12356508e+00 1.95608631e-01
-1.54773140e+00 -2.37465814e-01 -2.01335773e-01 1.87506303e-02
-8.91224444e-02 6.86340928e-01 -4.67718631e-01 6.83858573e-01
1.13171172e+00 1.03412032e+00 -4.74492162e-01 1.01982093e+00
1.04194634e-01 7.09445000e-01 -1.57424510e-01 3.49243395e-02
-2.56225407e-01 1.05076976e-01 4.32834625e-01 1.19659483e+00
4.03069168e-01 1.98696434e-01 3.74856144e-01 5.65521955e-01
5.69675267e-02 1.42539337e-01 -1.51781365e-01 2.80144095e-01
3.42818379e-01 1.27872825e+00 -6.26809955e-01 -5.29653013e-01
-2.34971270e-01 5.62446594e-01 -1.03060916e-01 4.29237038e-01
-1.08638287e+00 -1.19341969e+00 5.95316887e-01 1.21043749e-01
1.53847560e-01 2.94821382e-01 -3.87955487e-01 -1.28265238e+00
1.35458097e-01 -1.13312423e+00 5.46135783e-01 -5.27448595e-01
-1.22023749e+00 1.02925098e+00 -1.84878007e-01 -1.48707569e+00
1.02908991e-01 -4.47280407e-01 -8.66729319e-01 1.18219221e+00
-1.39681983e+00 -8.30043852e-01 -5.49138546e-01 6.18449569e-01
4.48743641e-01 -1.77963167e-01 1.01735663e+00 6.55889332e-01
-6.53637290e-01 8.54183078e-01 1.38236098e-02 4.42365021e-01
7.37698734e-01 -1.06649911e+00 4.17650156e-02 5.80875337e-01
-2.27697387e-01 7.57279694e-01 1.00674711e-01 -1.98965549e-01
-8.65877867e-01 -1.27507365e+00 8.70589972e-01 1.23777415e-03
1.76620618e-01 -8.48933756e-02 -1.21636045e+00 3.03639978e-01
-8.03104788e-02 4.44619238e-01 8.54160249e-01 1.84730291e-01
-1.13623075e-01 -6.02138698e-01 -9.54476178e-01 3.71507257e-01
7.01266885e-01 -4.10982609e-01 -5.50605714e-01 -2.25779274e-03
2.55825728e-01 -4.34909135e-01 -1.29946852e+00 8.52650523e-01
9.58584666e-01 -8.56574535e-01 1.06974149e+00 -6.75012887e-01
3.60730112e-01 -3.04988265e-01 1.27096891e-01 -1.25339031e+00
-6.45277143e-01 -4.28202391e-01 4.33277600e-02 9.63503063e-01
4.47830766e-01 -1.02485704e+00 4.37517554e-01 1.02579929e-01
-3.97671551e-01 -1.36321938e+00 -9.62140441e-01 -3.42703462e-01
-6.11574985e-02 -3.97746384e-01 5.98581612e-01 1.04417694e+00
-9.36567709e-02 4.82680619e-01 -2.96258390e-01 3.93200293e-02
4.41890538e-01 4.89704400e-01 8.25764909e-02 -1.51976705e+00
-3.78063917e-01 -5.47928274e-01 -6.17170990e-01 -4.37274992e-01
-3.67300481e-01 -9.01589394e-01 -3.27430576e-01 -1.51441896e+00
2.95470834e-01 -4.78985518e-01 -1.22575605e+00 6.23428345e-01
-4.76409912e-01 4.07393008e-01 3.53820436e-02 1.48814425e-01
-3.39070737e-01 3.16188425e-01 1.26117826e+00 -2.82391101e-01
-2.94573188e-01 2.70733654e-01 -6.50345325e-01 7.19063520e-01
1.13405180e+00 -4.97038156e-01 -5.54963827e-01 -2.32734978e-01
-1.73332915e-01 2.52085745e-01 2.43592829e-01 -1.22639823e+00
-2.04299539e-01 1.45776674e-01 8.94193530e-01 -4.47163194e-01
1.26235828e-01 -4.60535735e-01 7.06010014e-02 8.05103600e-01
-3.56552482e-01 3.28868389e-01 2.17924833e-01 3.00732702e-01
-3.89748871e-01 2.93179750e-01 7.62953341e-01 -4.36745249e-02
-2.69740969e-01 5.46728849e-01 -3.50135595e-01 4.01639074e-01
8.20921302e-01 -1.86478972e-01 8.93209409e-03 6.52251914e-02
-9.20169830e-01 1.55163422e-01 -2.94990420e-01 5.21870375e-01
9.64165747e-01 -1.42485845e+00 -9.50035512e-01 4.12157804e-01
2.84205109e-01 1.04888655e-01 7.08298743e-01 1.20659065e+00
-2.92435437e-01 1.83424592e-01 -4.51987416e-01 -9.11790788e-01
-1.40706170e+00 1.31046519e-01 6.27407670e-01 -4.52412128e-01
-8.32402706e-01 7.29566276e-01 3.14671099e-01 6.78278729e-02
2.97678895e-02 -7.28509009e-01 -4.10599917e-01 3.47773246e-02
7.40050137e-01 3.70451570e-01 3.39979827e-01 -4.39346582e-01
-6.08503401e-01 9.43401456e-01 -1.45545244e-01 5.48669338e-01
1.19521379e+00 1.88313738e-01 7.99211785e-02 4.97203857e-01
1.08179176e+00 -4.24288183e-01 -7.32762158e-01 -3.18336070e-01
-3.73231798e-01 -1.45060256e-01 -7.95547590e-02 -9.42716956e-01
-1.35450816e+00 1.47218800e+00 1.09454894e+00 3.16423982e-01
1.61303723e+00 -2.80481160e-01 1.01955116e+00 -2.91821901e-02
8.39935988e-02 -7.51345575e-01 -2.08668932e-01 2.01582775e-01
7.94773638e-01 -9.58286822e-01 -8.42275918e-02 -7.01110391e-03
-7.15303600e-01 1.20564330e+00 4.17011142e-01 -4.04659547e-02
7.44361103e-01 1.42030314e-01 2.22871140e-01 -1.14067949e-01
-6.28245711e-01 1.72163069e-01 5.22601545e-01 4.29934502e-01
5.98003209e-01 1.62709057e-01 -4.38113302e-01 1.27706087e+00
2.98518032e-01 5.62738553e-02 8.42742622e-02 7.87415504e-01
-2.35305458e-01 -8.39544058e-01 -2.17343807e-01 8.07584047e-01
-9.76738989e-01 -1.44913629e-01 1.34265572e-01 5.87664127e-01
4.83392447e-01 8.34096909e-01 -1.76955283e-01 -6.97485864e-01
4.75867897e-01 1.59630865e-01 2.20072776e-01 -3.85749668e-01
-8.00715864e-01 4.68215466e-01 -2.14721918e-01 -3.14809144e-01
-2.43194073e-01 -5.39956748e-01 -1.33492148e+00 2.91882128e-01
-2.16658637e-01 3.44447613e-01 1.48171797e-01 8.06080163e-01
6.81674898e-01 1.22525024e+00 5.79091728e-01 -4.86770242e-01
-6.20634437e-01 -1.16675460e+00 -6.95340335e-01 4.04428780e-01
5.63558280e-01 -5.74216843e-01 -5.77405952e-02 1.03593193e-01] | [14.270788192749023, 3.232184648513794] |
a13efb84-7e48-4f02-99aa-91f0320c0f1a | damo-nlp-at-semeval-2022-task-11-a-knowledge | 2203.00545 | null | https://arxiv.org/abs/2203.00545v3 | https://arxiv.org/pdf/2203.00545v3.pdf | DAMO-NLP at SemEval-2022 Task 11: A Knowledge-based System for Multilingual Named Entity Recognition | The MultiCoNER shared task aims at detecting semantically ambiguous and complex named entities in short and low-context settings for multiple languages. The lack of contexts makes the recognition of ambiguous named entities challenging. To alleviate this issue, our team DAMO-NLP proposes a knowledge-based system, where we build a multilingual knowledge base based on Wikipedia to provide related context information to the named entity recognition (NER) model. Given an input sentence, our system effectively retrieves related contexts from the knowledge base. The original input sentences are then augmented with such context information, allowing significantly better contextualized token representations to be captured. Our system wins 10 out of 13 tracks in the MultiCoNER shared task. | ['Yong Jiang', 'Wei Lu', 'Kewei Tu', 'Yueting Zhuang', 'Weiming Lu', 'Fei Huang', 'Pengjun Xie', 'Xiaobin Wang', 'Tao Wang', 'Jiong Cai', 'Yongliang Shen', 'Xinyu Wang'] | 2022-03-01 | null | https://aclanthology.org/2022.semeval-1.200 | https://aclanthology.org/2022.semeval-1.200.pdf | semeval-naacl-2022-7 | ['multilingual-named-entity-recognition'] | ['natural-language-processing'] | [-2.05329508e-01 -6.64310828e-02 -1.93348899e-01 -4.91031796e-01
-1.32798600e+00 -1.15953434e+00 4.30354536e-01 3.97356987e-01
-9.98697221e-01 1.05991232e+00 6.51167452e-01 -2.25707605e-01
7.32004121e-02 -7.59456396e-01 -5.77105105e-01 7.32270703e-02
1.09343484e-01 5.81013083e-01 2.63090640e-01 -5.98599792e-01
6.61481768e-02 4.53170627e-01 -1.05358624e+00 7.27471888e-01
1.17333317e+00 4.44042951e-01 5.59257746e-01 4.65914875e-01
-7.47114241e-01 6.27085268e-01 -6.72667682e-01 -6.16615593e-01
-1.96603574e-02 1.30059063e-01 -1.41204035e+00 -6.41621530e-01
4.94754910e-01 1.48347914e-01 -1.55439049e-01 8.44206035e-01
5.47627807e-01 4.00899708e-01 3.08426619e-01 -6.61907792e-01
-7.50365317e-01 1.00031960e+00 1.85374528e-01 3.66804898e-01
7.56124735e-01 -2.83001631e-01 1.40846944e+00 -1.39610124e+00
1.08984840e+00 1.12004054e+00 7.53644824e-01 4.54430699e-01
-9.23265159e-01 -5.33364058e-01 3.31852883e-01 9.48941037e-02
-1.73880506e+00 -4.73051697e-01 2.95287222e-02 3.08974460e-02
1.70556080e+00 2.04668120e-01 2.53724873e-01 9.74762619e-01
-3.85016829e-01 7.08907187e-01 7.68104374e-01 -6.83356464e-01
-2.15986581e-03 8.65437165e-02 3.55612516e-01 3.64409536e-01
5.38632333e-01 -2.01134533e-01 -6.72335982e-01 -1.54544070e-01
2.58609980e-01 -3.04025501e-01 -1.17121749e-01 3.47111613e-01
-1.33795929e+00 3.56865555e-01 4.62845504e-01 8.31376553e-01
-3.29150259e-01 -2.19561964e-01 4.15818304e-01 1.49835646e-01
2.10877061e-01 1.15279663e+00 -1.04783499e+00 -4.93672751e-02
-8.94898951e-01 1.81941077e-01 9.31933880e-01 1.22164893e+00
9.77113128e-01 -3.61495197e-01 -3.17981601e-01 1.17576754e+00
2.47424304e-01 7.43878782e-01 5.21885037e-01 -6.14547312e-01
9.64404881e-01 8.12823653e-01 4.71153110e-01 -6.85909986e-01
-6.82465255e-01 -1.85142845e-01 -1.17595740e-01 -5.74822545e-01
4.59135890e-01 -3.85391116e-01 -1.03480661e+00 1.62625957e+00
3.32554549e-01 7.78006855e-03 6.42564654e-01 6.11511111e-01
1.13751447e+00 5.32394409e-01 5.31614184e-01 3.63753349e-01
1.68456542e+00 -7.50726640e-01 -6.52619183e-01 -6.77866578e-01
1.09917045e+00 -1.03496492e+00 7.39676297e-01 -1.79256856e-01
-4.86438155e-01 -1.93571895e-01 -6.71848059e-01 -4.07836050e-01
-9.99245882e-01 3.50866884e-01 4.32641298e-01 4.28285956e-01
-8.98193061e-01 2.60264575e-01 -4.95338500e-01 -6.22484207e-01
-1.34215012e-01 1.02027655e-01 -7.57018626e-01 -4.52056140e-01
-1.78335845e+00 1.37786841e+00 1.05926466e+00 1.55316517e-01
-5.60797393e-01 -7.15213418e-01 -1.27819407e+00 -1.20275781e-01
6.14137471e-01 -5.55345535e-01 1.17758560e+00 -6.16480112e-01
-7.91170180e-01 1.06110370e+00 -4.17088300e-01 -3.15610111e-01
-2.39081010e-02 -5.51745236e-01 -8.33954275e-01 7.80089796e-02
7.81435549e-01 3.71090263e-01 4.66075204e-02 -8.91309619e-01
-9.38653946e-01 -2.53197402e-01 2.43265718e-01 4.36673135e-01
1.69252809e-02 4.84029204e-01 -5.19576252e-01 -6.69917643e-01
5.10450266e-02 -8.27891648e-01 -3.89816880e-01 -9.81366873e-01
-4.87979084e-01 -3.21065068e-01 2.71005660e-01 -9.10820305e-01
1.18679166e+00 -1.93738651e+00 -2.88681954e-01 8.19049180e-02
-9.75419804e-02 3.11791033e-01 -2.96888530e-01 6.24659657e-01
-1.16006047e-01 4.51419622e-01 -2.73397565e-02 -2.24490128e-02
-1.57723144e-01 4.33682710e-01 -4.46642101e-01 -3.17012459e-01
4.81046855e-01 8.77209246e-01 -1.28804612e+00 -6.79417908e-01
-1.56025097e-01 1.62760183e-01 -3.27876121e-01 4.39101495e-02
-1.44877434e-01 2.69797564e-01 -5.56218565e-01 7.13354468e-01
3.97889763e-01 1.06069662e-01 5.53296447e-01 -1.85389772e-01
-4.68145460e-01 8.10974538e-01 -1.41412461e+00 1.78461516e+00
-9.07279432e-01 3.75565827e-01 -1.07517771e-01 -4.68399256e-01
7.05460846e-01 6.06888354e-01 3.65205482e-02 -5.90281963e-01
-4.57242727e-01 6.23613060e-01 -5.16615331e-01 -6.19634688e-01
1.20410323e+00 -1.88256413e-01 -6.95231557e-01 1.21328952e-02
4.05640393e-01 1.70956388e-01 2.94646412e-01 4.56833422e-01
1.18844020e+00 9.69430879e-02 7.32145727e-01 -1.10172674e-01
5.56077659e-01 4.48266953e-01 9.76152420e-01 9.64400172e-01
-2.37823918e-01 3.55098456e-01 -1.16025522e-01 -3.82082671e-01
-8.85643661e-01 -1.04651678e+00 -1.92214191e-01 1.52276838e+00
-5.95771568e-03 -8.00576627e-01 -2.43012711e-01 -1.03770578e+00
7.88492803e-03 8.76853526e-01 -3.39530110e-01 3.68515134e-01
-8.02337945e-01 -4.44508940e-01 1.24921083e+00 7.16369808e-01
2.16702342e-01 -1.13093424e+00 5.57950698e-02 5.10491908e-01
-7.99385726e-01 -1.53176498e+00 -3.92068744e-01 2.33008474e-01
-3.60509574e-01 -1.09370792e+00 -3.73002559e-01 -1.03789783e+00
4.39762652e-01 7.17453361e-02 1.34589708e+00 -1.36025831e-01
-1.91618636e-01 4.21541393e-01 -6.29135907e-01 -2.49109030e-01
-3.91789496e-01 3.56816381e-01 1.86170246e-02 -4.91873085e-01
6.44178808e-01 -5.05023822e-02 -1.27172530e-01 3.75526279e-01
-9.23885703e-01 -3.43804210e-01 6.07858062e-01 8.02068889e-01
6.79806948e-01 -4.44651783e-01 9.01872873e-01 -8.89370084e-01
3.67169768e-01 -6.21709824e-01 -4.41788316e-01 8.04394901e-01
-1.57426372e-01 3.24089020e-01 6.27110898e-01 -4.09371518e-02
-1.64862335e+00 8.78058821e-02 -2.96171278e-01 2.61561036e-01
-3.45026225e-01 8.68968427e-01 -4.60347950e-01 2.56018490e-01
8.76705587e-01 7.49282837e-02 -1.17899847e+00 -6.74946547e-01
7.69077003e-01 9.07809317e-01 7.78267980e-01 -1.04609561e+00
5.67438841e-01 7.48934597e-02 -6.20052814e-01 -9.36031044e-01
-1.30738211e+00 -8.94904792e-01 -8.66455853e-01 1.11637577e-01
1.01777387e+00 -1.31126726e+00 -7.71707529e-03 2.11188439e-02
-1.33883286e+00 4.77992930e-02 -2.88063020e-01 6.70668721e-01
-1.23234522e-02 3.31463665e-01 -5.83016992e-01 -5.76797247e-01
-3.24407488e-01 -6.09197915e-01 8.99829566e-01 3.11226726e-01
-3.49799275e-01 -9.61350024e-01 3.32960576e-01 5.42780459e-01
2.68184274e-01 -1.65243387e-01 7.27489531e-01 -1.58554780e+00
-3.27414602e-01 -1.61632255e-01 -1.05886653e-01 1.61503434e-01
1.01481661e-01 -2.13184625e-01 -7.59399593e-01 7.45476112e-02
-7.21046150e-01 -3.51657361e-01 7.53135979e-01 -3.73268932e-01
4.35694635e-01 -1.46644488e-01 -4.93784845e-01 2.14308828e-01
1.37984741e+00 -3.06443628e-02 4.34386700e-01 6.63271487e-01
6.52109146e-01 4.41329688e-01 8.19296420e-01 8.67982358e-02
7.18613863e-01 4.58213270e-01 -1.56946391e-01 1.58335268e-01
4.60429192e-02 -3.65459621e-01 2.83427984e-01 7.46141911e-01
1.09749444e-01 1.26742916e-02 -1.51962554e+00 1.15467083e+00
-1.60822308e+00 -1.05622995e+00 -3.45987326e-04 2.09559774e+00
1.17053509e+00 -1.42826825e-01 -4.23907042e-01 -6.77203953e-01
1.04676628e+00 -5.66273257e-02 -1.86906934e-01 -2.61160523e-01
-4.82895315e-01 3.91516000e-01 6.26851499e-01 5.30003130e-01
-1.22992265e+00 1.65618289e+00 6.57713747e+00 7.82069743e-01
-6.83040321e-01 1.00545986e-02 -2.03641385e-01 6.53241351e-02
-4.95509148e-01 3.88137490e-01 -1.48696482e+00 1.71644583e-01
1.04302311e+00 -5.35436332e-01 1.27806500e-01 8.83885980e-01
-1.17950745e-01 -7.77285546e-02 -9.65945303e-01 6.18198574e-01
3.19699757e-02 -1.29043388e+00 1.59341335e-01 -3.54045391e-01
7.06382573e-01 6.77961230e-01 -4.29450065e-01 7.02112854e-01
8.24391365e-01 -8.85796189e-01 5.01436234e-01 5.00444233e-01
6.72152996e-01 -7.78320193e-01 9.06097174e-01 1.90523177e-01
-1.46772850e+00 1.18017845e-01 -3.71806622e-01 2.19635740e-01
3.50074172e-01 5.89821756e-01 -1.17432117e+00 1.06491411e+00
6.79143369e-01 3.51445526e-01 -6.71523392e-01 1.16588688e+00
-6.11101925e-01 5.42056382e-01 -5.64614117e-01 1.27286986e-01
2.98791379e-01 2.31919110e-01 8.59832287e-01 1.83746469e+00
1.03836432e-01 2.40144432e-01 4.07904625e-01 4.27549541e-01
-6.91798627e-01 6.23479724e-01 -8.39431643e-01 -1.64414614e-01
9.78354037e-01 1.36585975e+00 -6.19183600e-01 -6.81352735e-01
-4.81043220e-01 9.55201268e-01 8.87332082e-01 2.67759740e-01
-2.92726964e-01 -8.86916935e-01 7.02773273e-01 -5.90670288e-01
3.94172519e-01 -3.50839049e-01 4.06168923e-02 -1.72499967e+00
1.37062386e-01 -7.77957976e-01 9.47275817e-01 -7.20327377e-01
-1.51778710e+00 6.80914342e-01 -1.21471390e-01 -7.81093955e-01
-2.01204568e-01 -5.74492395e-01 -3.58858079e-01 9.13926482e-01
-1.67380929e+00 -1.25193489e+00 5.89514710e-02 5.39038002e-01
3.68917316e-01 -6.29289122e-03 1.15705609e+00 6.09662354e-01
-5.12890995e-01 5.50676823e-01 6.31312653e-02 7.94645131e-01
1.34723127e+00 -1.43468070e+00 5.36742389e-01 1.14879131e+00
4.07726437e-01 1.16217124e+00 5.13658583e-01 -1.01596844e+00
-8.89217973e-01 -1.34681511e+00 1.70071459e+00 -8.24602127e-01
8.43395531e-01 -1.32702351e-01 -1.02506959e+00 1.05315936e+00
7.01744482e-02 1.13138281e-01 1.28825009e+00 6.55236781e-01
-8.35261822e-01 2.83577353e-01 -1.02526379e+00 4.75951463e-01
1.08664656e+00 -1.02629340e+00 -1.56851339e+00 2.07478300e-01
9.02419090e-01 -5.20292282e-01 -1.07639587e+00 6.20718524e-02
2.63111442e-01 1.43279321e-03 8.00199568e-01 -1.14674437e+00
-5.20351715e-02 -5.46738088e-01 -4.57603663e-01 -1.42034543e+00
1.23773830e-03 -2.63283700e-01 4.27027851e-01 1.58984160e+00
1.13161993e+00 -5.64091623e-01 2.05066174e-01 8.26185822e-01
-2.88760483e-01 1.19155729e-02 -1.06447875e+00 -8.90320003e-01
1.02666073e-01 -6.85641110e-01 6.86053514e-01 1.42291200e+00
3.25241625e-01 5.64673066e-01 5.04933223e-02 7.11548090e-01
2.16833547e-01 2.12141484e-01 2.00497746e-01 -1.01311576e+00
2.23574117e-01 2.83437759e-01 -3.22093993e-01 -7.70737469e-01
4.65820700e-01 -1.38421214e+00 2.14642733e-01 -1.73346806e+00
2.40895152e-02 -8.44989717e-01 -4.20946598e-01 1.02351487e+00
-4.83058453e-01 1.88533124e-02 3.42971653e-01 1.45400569e-01
-1.22957051e+00 1.82712927e-01 4.23632920e-01 6.37762696e-02
-3.17026317e-01 -4.07965720e-01 -8.11056197e-01 6.80402935e-01
5.93883932e-01 -7.78504252e-01 3.37829441e-01 -7.74527788e-01
7.07776666e-01 -2.06405371e-01 1.85976997e-02 -8.97411644e-01
4.05508071e-01 -2.33262584e-01 4.43767279e-01 -5.46577394e-01
-7.28209019e-02 -5.54258227e-01 -2.15061113e-01 -5.56787737e-02
-4.71477777e-01 1.04659565e-01 3.28792930e-01 3.47853333e-01
-3.56418550e-01 -4.11324650e-01 3.93892854e-01 -3.88166696e-01
-1.14681518e+00 -7.25266337e-02 -5.14021337e-01 7.74380505e-01
7.35233605e-01 2.51671463e-01 -4.58476543e-01 1.76964000e-01
-1.13961339e+00 6.26932502e-01 4.03207064e-01 6.24312699e-01
3.57518196e-01 -1.32738900e+00 -7.01784790e-01 -3.92738581e-01
5.97968221e-01 1.41962394e-01 3.09624374e-01 2.26744607e-01
-4.03718978e-01 7.49633670e-01 -1.68298155e-01 -2.33099069e-02
-1.01493013e+00 3.40396434e-01 4.03059274e-01 -5.85644603e-01
-3.87764990e-01 6.66657448e-01 -1.77395612e-01 -1.09096265e+00
-2.99968094e-01 -8.11215267e-02 -5.11027992e-01 3.30535799e-01
7.79790044e-01 2.42352262e-01 4.10210848e-01 -8.69584322e-01
-8.70753646e-01 1.57932654e-01 -3.59847307e-01 -3.84834081e-01
1.28519762e+00 -2.76506662e-01 -5.46199046e-02 2.46235862e-01
9.53966260e-01 6.85191691e-01 -3.64638686e-01 -6.30107343e-01
8.53090405e-01 -1.32515907e-01 -2.50607878e-01 -1.21065986e+00
-3.98922652e-01 3.05074334e-01 6.80973008e-02 -1.49116367e-01
5.74758708e-01 3.29810619e-01 8.26733768e-01 1.13279045e+00
5.96654534e-01 -1.48147106e+00 -4.47313160e-01 1.33430481e+00
6.40929341e-01 -1.21600151e+00 -3.49225283e-01 -3.67688358e-01
-7.13396728e-01 9.67140853e-01 7.54450917e-01 3.43171179e-01
2.65668452e-01 8.83732662e-02 3.44291091e-01 -2.21935064e-01
-5.47381520e-01 -7.11999476e-01 2.09331974e-01 6.95909083e-01
5.36943436e-01 1.12810805e-01 -4.35755998e-01 1.07639277e+00
-1.61003709e-01 -3.02933604e-01 3.81487906e-01 1.17247677e+00
-5.91910362e-01 -1.27943695e+00 -3.52455646e-01 2.05222011e-01
-8.00557554e-01 -6.78236902e-01 -4.83440548e-01 6.38162971e-01
1.87000379e-01 1.02076352e+00 -1.81055799e-01 5.80073670e-02
7.30199039e-01 5.41144907e-01 -2.35709716e-02 -1.08669400e+00
-8.60986292e-01 -5.55141747e-01 8.30582857e-01 -6.19588256e-01
-2.68834651e-01 -5.00589371e-01 -1.53708684e+00 3.68385792e-01
-3.97457868e-01 5.77519298e-01 6.75118268e-01 1.21059358e+00
6.84332728e-01 1.34134114e-01 2.04344362e-01 -2.87882775e-01
-2.03551948e-01 -7.69822240e-01 -2.16331318e-01 5.07565439e-01
-2.03598477e-02 -3.91060412e-01 -5.56263179e-02 -1.15528673e-01] | [9.776711463928223, 9.531672477722168] |
468f9ea5-8d21-4760-90bc-525487c5ba03 | vietnamese-word-segmentation-with-svm | 2006.07804 | null | https://arxiv.org/abs/2006.07804v1 | https://arxiv.org/pdf/2006.07804v1.pdf | Vietnamese Word Segmentation with SVM: Ambiguity Reduction and Suffix Capture | In this paper, we approach Vietnamese word segmentation as a binary classification by using the Support Vector Machine classifier. We inherit features from prior works such as n-gram of syllables, n-gram of syllable types, and checking conjunction of adjacent syllables in the dictionary. We propose two novel ways to feature extraction, one to reduce the overlap ambiguity and the other to increase the ability to predict unknown words containing suffixes. Different from UETsegmenter and RDRsegmenter, two state-of-the-art Vietnamese word segmentation methods, we do not employ the longest matching algorithm as an initial processing step or any post-processing technique. According to experimental results on benchmark Vietnamese datasets, our proposed method obtained a better F1-score than the prior state-of-the-art methods UETsegmenter, and RDRsegmenter. | ['Ngan Luu-Thuy Nguyen', 'Duc-Vu Nguyen', 'Kiet Van Nguyen', 'Dang Van Thin'] | 2020-06-14 | null | null | null | null | ['vietnamese-word-segmentation', 'vietnamese-datasets'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.79231137e-01 -2.82382786e-01 -3.76992136e-01 -3.44414651e-01
-3.22631329e-01 -8.87601674e-01 4.26145881e-01 3.54572952e-01
-8.79694283e-01 8.23404908e-01 -2.25262165e-01 -6.88724875e-01
1.91211030e-01 -9.15620267e-01 -3.10787857e-01 -6.89537764e-01
1.26736239e-01 5.39474905e-01 5.71201801e-01 -4.71721828e-01
6.37515903e-01 5.09845793e-01 -1.48018718e+00 1.14206873e-01
1.17812014e+00 6.04936242e-01 1.66061252e-01 3.48544359e-01
-3.31584364e-01 -1.88401848e-01 -6.70498490e-01 -4.73552883e-01
2.18021244e-01 -4.47935641e-01 -6.59092486e-01 7.22430572e-02
-1.60769209e-01 4.25772630e-02 -2.28678301e-01 1.10690355e+00
2.09484741e-01 2.88531691e-01 9.28966820e-01 -9.27131712e-01
-5.19528568e-01 8.58963013e-01 -6.60631657e-01 5.24906397e-01
4.02790189e-01 -4.31839302e-02 1.01646376e+00 -1.07356477e+00
6.05375528e-01 1.03224826e+00 3.93966079e-01 1.69168040e-01
-7.04717398e-01 -6.35822356e-01 1.33081123e-01 4.08362657e-01
-1.64444792e+00 -1.28526405e-01 6.63511634e-01 -2.98792571e-01
1.21010578e+00 3.89129460e-01 6.03467882e-01 7.04580903e-01
4.05699871e-02 8.48344088e-01 1.26240039e+00 -7.75458097e-01
2.38865390e-01 -4.70157005e-02 6.49869680e-01 7.35543907e-01
4.12830383e-01 7.81426672e-03 -1.04322195e-01 1.58898477e-02
6.33230686e-01 -3.67564976e-01 -6.02961965e-02 3.47995162e-01
-1.10232496e+00 1.03442001e+00 -1.56435356e-01 6.45095170e-01
-3.67499113e-01 -6.37151897e-01 4.40894663e-01 -1.55121526e-02
-1.50382118e-02 2.29778424e-01 -6.54970467e-01 -2.08159760e-01
-1.04537857e+00 9.94943008e-02 7.39417434e-01 1.03874505e+00
7.17133224e-01 1.00325443e-01 -1.12233989e-01 7.47039616e-01
1.55266315e-01 4.29969281e-01 7.92239904e-01 -3.03862065e-01
4.05950904e-01 4.69960213e-01 -2.45485321e-01 -7.42610693e-01
-2.31884837e-01 -2.03913659e-01 -6.01485968e-01 -1.64901093e-01
5.89639962e-01 -1.90104812e-01 -1.36149764e+00 1.26778460e+00
2.93740898e-01 -1.79771602e-01 1.56301036e-01 7.41645575e-01
8.03436637e-01 1.05578196e+00 3.32763344e-02 -5.83006799e-01
1.58909488e+00 -1.05580330e+00 -9.83086050e-01 -1.68923587e-01
5.61259508e-01 -1.16959214e+00 9.24035728e-01 7.31658518e-01
-7.52776384e-01 -6.15229845e-01 -1.21349573e+00 1.32393077e-01
-7.93036819e-01 2.00828120e-01 7.87089229e-01 1.10503709e+00
-3.16315383e-01 5.44539750e-01 -7.71390915e-01 -2.79949516e-01
-1.85201004e-01 4.38715011e-01 -4.41719621e-01 2.14986518e-01
-1.22931015e+00 7.69928336e-01 1.01447093e+00 1.25507489e-01
-2.77824193e-01 -6.25353456e-02 -1.06725085e+00 6.91246474e-04
5.27399838e-01 -2.42367219e-02 9.66527641e-01 -6.64394915e-01
-1.46412611e+00 7.35913932e-01 -4.68421847e-01 -1.45336762e-01
2.61118203e-01 4.96692434e-02 -7.03682125e-01 5.25142625e-02
8.84288996e-02 3.38412374e-01 5.88644028e-01 -1.10065293e+00
-8.47416639e-01 -3.47956628e-01 -5.85339189e-01 1.58956051e-01
-2.73354143e-01 1.47796020e-01 -7.21534967e-01 -8.67603242e-01
5.07063627e-01 -9.02611613e-01 -1.22521378e-01 -7.75730789e-01
-5.03106713e-01 -6.18230522e-01 9.76313889e-01 -1.01138747e+00
1.55632436e+00 -1.95486927e+00 -6.04929179e-02 6.75175309e-01
-2.36414239e-01 7.88701594e-01 -5.36084129e-03 4.68887985e-01
-1.72349270e-02 3.39767873e-01 -4.00564998e-01 1.47920102e-01
2.18618847e-03 6.12829685e-01 6.46400591e-03 5.69014847e-01
2.83334225e-01 4.94413137e-01 -8.51626277e-01 -9.03111994e-01
1.22312672e-01 2.39247918e-01 -1.41637221e-01 3.21195349e-02
4.59945872e-02 -4.99253385e-02 -2.04092085e-01 8.57422590e-01
7.21725464e-01 4.74865347e-01 3.69374394e-01 -5.49909398e-02
-4.13620979e-01 3.51770073e-01 -1.43444085e+00 1.13671875e+00
-9.03039798e-02 2.82642394e-01 -3.29787314e-01 -1.17624366e+00
1.25826025e+00 2.93987572e-01 8.28143731e-02 -6.85899556e-01
5.26149869e-01 5.97457528e-01 3.59704256e-01 -4.42567289e-01
7.81037390e-01 1.28776506e-01 -4.09474909e-01 7.72026973e-03
1.48417383e-01 1.06545627e-01 9.30970788e-01 -1.86754212e-01
5.15313983e-01 8.19242895e-02 7.61115968e-01 -3.00023675e-01
8.43796968e-01 1.44653186e-01 9.02549565e-01 3.97239506e-01
-2.99908936e-01 6.19207323e-01 4.24160779e-01 -1.00102350e-01
-9.31645572e-01 -1.01251578e+00 -1.94812253e-01 1.10833549e+00
3.98756564e-02 -2.35111102e-01 -9.74898458e-01 -8.29362512e-01
-2.40210280e-01 7.86599636e-01 -2.63512284e-01 2.22744465e-01
-1.04467285e+00 -7.60330141e-01 7.68359780e-01 5.81235647e-01
3.86265367e-01 -1.07000768e+00 -3.55768353e-01 2.79985368e-01
-1.69121727e-01 -1.13695312e+00 -7.49673486e-01 4.94606435e-01
-5.88409483e-01 -8.71107101e-01 -6.67525709e-01 -1.40210831e+00
4.97159719e-01 3.23201045e-02 7.28415847e-01 2.36296237e-01
-1.04915768e-01 -4.63537633e-01 -8.19233656e-01 -2.59848088e-01
-2.05079898e-01 3.77901882e-01 6.83728382e-02 -1.76637664e-01
6.51026547e-01 -1.66448876e-01 -3.26201797e-01 1.81554168e-01
-7.89160550e-01 -4.28272128e-01 7.32228756e-01 7.88904071e-01
8.14675510e-01 3.38579029e-01 3.77991796e-01 -9.45470750e-01
5.17140448e-01 -4.43199039e-01 -6.54294312e-01 2.26698160e-01
-8.54442239e-01 3.59408148e-02 8.06317806e-01 -6.30485117e-01
-8.87245536e-01 1.12810791e-01 -7.23839402e-01 1.05179437e-01
-3.99500310e-01 5.89005530e-01 -5.10363817e-01 1.37887135e-01
9.58097950e-02 5.17683327e-01 -3.83540452e-01 -4.56768543e-01
2.30243772e-01 1.05620551e+00 7.20046163e-01 -4.41241860e-01
7.63953030e-01 2.84249801e-02 -2.39236161e-01 -1.24312770e+00
-1.08444646e-01 -7.45745003e-01 -1.11533356e+00 2.45426431e-01
1.05871999e+00 -4.46869612e-01 -4.88386095e-01 4.47464079e-01
-1.29893541e+00 3.25467706e-01 1.90958574e-01 5.42888641e-01
-8.91941711e-02 9.16132569e-01 -6.78056002e-01 -8.76562238e-01
-4.71893579e-01 -1.24987781e+00 7.39300013e-01 6.05445027e-01
-2.77796298e-01 -6.01073742e-01 -1.13330014e-01 3.68209422e-01
-1.43641561e-01 5.36929481e-02 1.13377154e+00 -1.33721292e+00
-7.95908198e-02 -1.33279398e-01 1.39533002e-02 2.36518964e-01
1.36835116e-03 3.51117373e-01 -5.16065955e-01 -6.80294484e-02
-2.25754529e-01 2.08910137e-01 9.01537776e-01 1.15593232e-01
7.05147624e-01 -3.04550737e-01 -3.30394417e-01 4.95658875e-01
1.35670102e+00 9.90093470e-01 8.07023466e-01 5.25978804e-01
4.59128559e-01 4.97463256e-01 1.20901716e+00 3.20020616e-01
4.16520894e-01 4.38472360e-01 1.19173586e-01 -1.48830533e-01
1.08855546e-01 -2.21848622e-01 2.06669599e-01 1.29805112e+00
-2.57303059e-01 -6.04689181e-01 -9.15419400e-01 7.92168438e-01
-1.46442235e+00 -7.79524565e-01 -3.87041926e-01 2.22437882e+00
8.75613511e-01 2.77448237e-01 3.97906959e-01 7.59298444e-01
8.37737501e-01 1.93692416e-01 -7.89849013e-02 -9.56721246e-01
-2.24849612e-01 7.94912636e-01 6.70198441e-01 4.88775283e-01
-1.25691938e+00 1.56604791e+00 5.52258253e+00 1.20912206e+00
-1.08677602e+00 5.01372572e-03 2.91585505e-01 5.80983043e-01
-1.98948145e-01 9.50946063e-02 -1.12668014e+00 4.11056936e-01
5.28340518e-01 1.40051380e-01 2.79211611e-01 5.88402867e-01
-1.16461441e-01 -2.01314941e-01 -7.94908583e-01 7.95693576e-01
1.75022110e-01 -9.07188654e-01 -4.50242609e-02 -1.54809887e-02
4.50996429e-01 -4.12562430e-01 -1.94772184e-01 4.19699371e-01
-2.04783469e-01 -8.73039484e-01 8.01367521e-01 -1.21806383e-01
5.61588526e-01 -1.04518616e+00 1.05055308e+00 2.90586412e-01
-1.35226285e+00 3.57201457e-01 -3.40677947e-01 -2.23975237e-02
2.20514104e-01 1.78060710e-01 -8.08942139e-01 6.46883905e-01
3.37437093e-01 7.17302635e-02 -5.01962841e-01 9.79184687e-01
-4.58332449e-01 9.46387291e-01 -3.33738327e-01 -5.50226212e-01
5.10139823e-01 -5.72209120e-01 5.64305782e-01 1.68085194e+00
3.08655083e-01 2.78212249e-01 4.79014277e-01 2.22655967e-01
3.27070475e-01 6.99970186e-01 -2.12517872e-01 -2.47173220e-01
7.59440482e-01 1.03459978e+00 -1.36792588e+00 -4.91630375e-01
-3.84373963e-01 1.24224424e+00 6.27146289e-02 2.53911614e-01
-8.01709294e-01 -1.12700200e+00 4.86674607e-01 1.93189532e-02
7.43758619e-01 -6.66293979e-01 -4.71231520e-01 -9.86913621e-01
9.13590416e-02 -1.11376393e+00 4.12456870e-01 -1.08033322e-01
-8.77351284e-01 7.42908597e-01 3.38610560e-02 -8.95469487e-01
-3.28483731e-01 -8.17638159e-01 -5.73776007e-01 7.48136818e-01
-1.43324077e+00 -1.07958746e+00 3.73934925e-01 3.99875075e-01
9.03219819e-01 -2.11837426e-01 7.30440915e-01 1.80396259e-01
-8.34987342e-01 8.67887318e-01 2.61573717e-02 4.71535832e-01
5.12543082e-01 -1.08719516e+00 3.70996118e-01 1.23038161e+00
2.91694045e-01 7.31996000e-01 7.76124954e-01 -9.54807758e-01
-1.30878007e+00 -5.83941936e-01 1.26843965e+00 3.32078516e-01
5.09721518e-01 -3.08990061e-01 -9.27027225e-01 6.26786172e-01
4.24757749e-01 -3.74301881e-01 7.32529402e-01 3.53249200e-02
-3.12660187e-01 2.08729476e-01 -1.19245553e+00 5.66874266e-01
8.53728116e-01 -1.29163504e-01 -1.11511183e+00 7.84245580e-02
6.22505546e-01 -3.22284102e-01 -7.89763033e-01 3.48456085e-01
7.12475955e-01 -7.05771685e-01 7.37366736e-01 -4.57634568e-01
1.42002804e-02 -3.62615436e-01 -2.63444960e-01 -1.18935168e+00
-1.97275013e-01 -5.65064311e-01 2.96981394e-01 1.48265636e+00
7.72559106e-01 -6.84894502e-01 7.42217839e-01 1.87919382e-02
-3.67155612e-01 -4.61278588e-01 -9.34836745e-01 -1.05685604e+00
1.87030047e-01 -1.63985655e-01 5.19038439e-01 1.15043044e+00
7.72872791e-02 2.59068668e-01 -1.61249951e-01 2.54660219e-01
4.83567327e-01 1.32709682e-01 2.43469760e-01 -1.00964701e+00
-3.63089561e-01 -5.17888069e-01 -4.02691573e-01 -8.66473258e-01
2.81660318e-01 -7.85494864e-01 1.53424338e-01 -1.37836552e+00
-1.28604904e-01 -2.31924742e-01 -7.99466893e-02 6.52694881e-01
-2.89068907e-01 3.41964841e-01 1.17706053e-01 -2.02715009e-01
-1.32256761e-01 2.38523275e-01 9.79386210e-01 -7.03300014e-02
-5.34885049e-01 -1.67966709e-01 -3.89158785e-01 7.94230342e-01
8.69244754e-01 -5.56547701e-01 -1.33686125e-01 -3.54240805e-01
-2.28158817e-01 -1.40871257e-01 -4.59475487e-01 -6.90226197e-01
5.59267849e-02 -4.00164902e-01 3.83372545e-01 -1.02371931e+00
1.89527526e-01 -5.69319546e-01 -1.56260923e-01 6.88685596e-01
2.84758002e-01 4.42961156e-01 7.39607289e-02 3.96515988e-02
-2.40952089e-01 -8.01595390e-01 6.02514565e-01 -8.73817652e-02
-1.03897989e+00 -6.93892539e-02 -9.46640611e-01 -8.24703872e-02
1.12507975e+00 -5.62674940e-01 -3.33060533e-01 2.14097992e-01
-4.32344615e-01 -8.36476758e-02 2.66039252e-01 4.49565858e-01
7.12482572e-01 -9.13622200e-01 -5.36762714e-01 5.64369321e-01
-9.35764145e-03 -4.08519328e-01 -2.43205294e-01 4.55165893e-01
-8.89262855e-01 4.37691838e-01 -3.09045821e-01 -2.19078243e-01
-1.54329550e+00 6.65350139e-01 -4.12864089e-01 -1.80064186e-01
-1.28937066e-01 6.46881282e-01 -4.81264263e-01 -4.41573411e-01
1.13270424e-01 -3.36387306e-01 -5.70971429e-01 3.20556670e-01
1.91137716e-01 5.18086135e-01 2.49721438e-01 -1.11982679e+00
-5.97694337e-01 6.44853652e-01 -2.60451585e-01 -3.10149491e-01
8.83314669e-01 1.40919536e-02 -1.64737493e-01 2.65428722e-01
1.04368007e+00 2.99456179e-01 -2.61501431e-01 1.87961176e-01
2.81352073e-01 -3.96604180e-01 -3.60483900e-02 -5.25999308e-01
-7.89560318e-01 6.28810227e-01 5.99521518e-01 2.40536436e-01
1.10037851e+00 -2.28362441e-01 1.25056970e+00 2.62341410e-01
4.30150867e-01 -1.39899552e+00 -4.35398579e-01 8.06619167e-01
3.87605548e-01 -1.24098361e+00 -9.20212790e-02 -9.80237365e-01
-6.98934317e-01 1.13736165e+00 5.11755705e-01 -1.39917582e-01
6.16880417e-01 4.02978927e-01 1.10070281e-01 3.70985001e-01
-1.40449166e-01 -7.62503803e-01 3.70869845e-01 7.03997254e-01
6.16064012e-01 2.88188219e-01 -1.54319131e+00 7.96559393e-01
-5.85383713e-01 -4.87156570e-01 4.90863591e-01 1.15726614e+00
-8.11250985e-01 -1.56305075e+00 -5.18266797e-01 2.92672664e-01
-6.19958818e-01 -3.16607058e-01 -3.33411217e-01 1.15884984e+00
4.19996083e-01 1.15405524e+00 1.25381529e-01 -5.57122290e-01
4.08257931e-01 2.04555228e-01 4.92148489e-01 -5.67229271e-01
-6.83648527e-01 5.13352752e-01 3.39406043e-01 -3.38270627e-02
5.68290520e-03 -8.55997980e-01 -1.72344863e+00 -1.32924080e-01
-5.88509023e-01 3.60847890e-01 8.13165188e-01 1.12557101e+00
-2.72881836e-01 1.49755150e-01 5.67722082e-01 -5.51213622e-01
-4.34412569e-01 -9.65693712e-01 -7.10821211e-01 2.95770139e-01
-1.01037592e-01 -5.59382081e-01 -1.36090413e-01 -4.24307585e-02] | [10.263628005981445, 10.143024444580078] |
ae72ffd5-987a-460c-a2cf-cc33d9861474 | boundary-aware-information-maximization-for | 2202.02371 | null | https://arxiv.org/abs/2202.02371v2 | https://arxiv.org/pdf/2202.02371v2.pdf | Boundary-aware Information Maximization for Self-supervised Medical Image Segmentation | Unsupervised pre-training has been proven as an effective approach to boost various downstream tasks given limited labeled data. Among various methods, contrastive learning learns a discriminative representation by constructing positive and negative pairs. However, it is not trivial to build reasonable pairs for a segmentation task in an unsupervised way. In this work, we propose a novel unsupervised pre-training framework that avoids the drawback of contrastive learning. Our framework consists of two principles: unsupervised over-segmentation as a pre-train task using mutual information maximization and boundary-aware preserving learning. Experimental results on two benchmark medical segmentation datasets reveal our method's effectiveness in improving segmentation performance when few annotated images are available. | ['Christian Desrosiers', 'Marco Pedersoli', 'Ping Wang', 'Jizong Peng'] | 2022-02-04 | null | null | null | null | ['unsupervised-pre-training'] | ['methodology'] | [ 5.96357644e-01 2.24717379e-01 -5.19708455e-01 -8.25006187e-01
-1.10465086e+00 -3.53170365e-01 4.36352581e-01 2.06415787e-01
-7.15935588e-01 7.24482000e-01 1.29278572e-02 -2.45470092e-01
2.11503245e-02 -6.53956532e-01 -5.09798527e-01 -8.28067482e-01
1.80563822e-01 4.85692441e-01 4.17278111e-01 -4.75702621e-02
3.19200814e-01 2.73818105e-01 -1.13743985e+00 3.12454224e-01
1.12091446e+00 6.56283021e-01 4.72918600e-01 2.45098874e-01
-2.25047931e-01 6.51513875e-01 -2.25294977e-01 -2.83536553e-01
4.37505364e-01 -6.44865334e-01 -1.17071581e+00 4.41345960e-01
4.78919968e-02 2.09110640e-02 1.95355356e-01 1.16638124e+00
4.28662211e-01 6.07979298e-02 7.71721244e-01 -8.66356850e-01
-2.46469006e-01 6.11865103e-01 -9.06453490e-01 2.88756311e-01
-1.95041105e-01 -5.46869170e-03 1.19245112e+00 -8.67481828e-01
5.99298000e-01 8.54922175e-01 6.50671303e-01 6.02456450e-01
-1.26776671e+00 -4.58920509e-01 8.41891319e-02 -2.70185079e-02
-1.34807897e+00 -1.21482752e-01 9.32333827e-01 -3.82206887e-01
4.79190111e-01 2.50564009e-01 5.86056054e-01 4.51082855e-01
-7.71052465e-02 1.21224844e+00 1.34774971e+00 -4.53759700e-01
1.01716876e-01 2.25475937e-01 3.93620312e-01 7.13543236e-01
-2.60120388e-02 2.49162149e-02 -2.25065172e-01 3.86277796e-03
6.94073260e-01 -3.64939636e-03 -1.51914626e-01 -5.92477918e-01
-1.05301833e+00 7.03043699e-01 6.34131432e-01 5.59270203e-01
-2.54628599e-01 -2.65842140e-01 3.77740949e-01 2.84614682e-01
7.13780820e-01 5.39001465e-01 -5.71286321e-01 2.96268880e-01
-1.15120757e+00 -2.06661269e-01 4.34287846e-01 7.61179626e-01
1.10051703e+00 -3.02206308e-01 -1.96449026e-01 9.16729033e-01
1.90263137e-01 -1.20762646e-01 7.06157804e-01 -5.98132372e-01
2.63756126e-01 7.84571707e-01 -3.33203852e-01 -7.57691443e-01
-4.52354789e-01 -5.97547531e-01 -9.77614820e-01 -1.30427003e-01
4.94325548e-01 -2.53357798e-01 -1.09608078e+00 1.62502861e+00
3.28509778e-01 2.76887029e-01 4.90341261e-02 7.81920254e-01
8.75923753e-01 3.11332583e-01 3.11403900e-01 -4.20544773e-01
8.47178578e-01 -1.20242441e+00 -6.83961511e-01 -2.66072303e-01
9.45252180e-01 -7.67738223e-01 9.01149452e-01 2.62991428e-01
-8.21821630e-01 -4.96302843e-01 -8.23312879e-01 1.43457123e-03
-2.15579316e-01 1.01654105e-01 8.24870944e-01 5.22285759e-01
-9.33508098e-01 5.46450794e-01 -7.96274245e-01 -8.96202326e-02
8.70912135e-01 6.08065128e-01 -4.32318062e-01 -8.70326720e-03
-7.93431044e-01 6.35430396e-01 6.94960535e-01 1.11498386e-01
-7.01523304e-01 -5.82316220e-01 -7.88330495e-01 -1.73661590e-01
4.43918347e-01 -5.81565797e-01 1.07227111e+00 -1.28084171e+00
-1.28447187e+00 1.22256517e+00 -5.08616306e-02 -4.15694237e-01
6.58363938e-01 -1.97454050e-01 1.05270170e-01 8.93406793e-02
2.45269209e-01 9.10527349e-01 6.81781888e-01 -1.46281040e+00
-5.81306159e-01 -4.81873006e-01 -1.50711253e-01 4.91188318e-01
-3.89564186e-01 -2.59311795e-01 -6.27242148e-01 -6.48047566e-01
5.63163042e-01 -8.25857282e-01 -8.67714167e-01 -1.99891821e-01
-5.71598709e-01 -3.84551108e-01 7.05564797e-01 -4.65222627e-01
9.06162620e-01 -2.04392338e+00 -2.89326571e-02 4.24130142e-01
1.89913169e-01 3.23251724e-01 1.02514513e-02 -5.56867057e-03
-1.34191439e-01 1.45731121e-01 -8.02201867e-01 -2.82992840e-01
-4.06791925e-01 2.24273831e-01 1.83836162e-01 4.35303032e-01
3.67719680e-01 9.45759594e-01 -1.12993717e+00 -1.08443344e+00
2.49594957e-01 2.51595050e-01 -5.37675679e-01 3.08916509e-01
-2.52936054e-02 8.68617177e-01 -5.32416463e-01 6.18686497e-01
6.38652682e-01 -3.57297897e-01 4.07358706e-01 -1.50552168e-01
3.32577713e-02 1.04448706e-01 -8.84195805e-01 1.90116012e+00
-8.08507428e-02 3.72511089e-01 -1.60807490e-01 -1.66025889e+00
9.72133398e-01 1.91021532e-01 6.26257241e-01 -5.81913948e-01
2.06878990e-01 2.02456415e-01 -7.35292807e-02 -5.62901020e-01
2.24849015e-01 -5.25503218e-01 3.39938849e-02 2.05738679e-01
8.06315988e-02 -3.45331192e-01 2.84533918e-01 1.10269748e-01
7.83625722e-01 2.12514192e-01 4.42736775e-01 -5.37684321e-01
6.37003779e-01 9.35202986e-02 9.55532372e-01 6.32738829e-01
-3.90324414e-01 7.89969683e-01 3.87977213e-01 -1.90457582e-01
-5.62330127e-01 -9.79451060e-01 -1.36587918e-01 1.05990243e+00
2.47279674e-01 -3.80498558e-01 -9.39141929e-01 -1.18090796e+00
-4.63510901e-01 3.77834976e-01 -6.28357589e-01 7.20025506e-03
-6.39799893e-01 -1.06866813e+00 2.94788957e-01 6.06366098e-01
7.12361872e-01 -9.53361332e-01 -3.92198116e-01 -4.62482162e-02
-2.24207923e-01 -8.49580348e-01 -4.27481353e-01 5.35283864e-01
-1.24222505e+00 -1.10858381e+00 -8.18512619e-01 -1.48275352e+00
1.17700839e+00 5.09355605e-01 1.15379953e+00 2.68717557e-01
-3.05845082e-01 9.98627394e-02 -2.12430909e-01 -2.37677932e-01
-1.58264890e-01 1.64156541e-01 -3.93745005e-01 -3.80515829e-02
2.72800416e-01 -5.37811399e-01 -7.40528703e-01 3.32390517e-01
-8.52647960e-01 1.80888340e-01 9.52780485e-01 1.07067990e+00
1.08624470e+00 -5.39275967e-02 6.59540117e-01 -1.57552862e+00
3.62039655e-01 -4.43899274e-01 -2.65482992e-01 4.01880920e-01
-6.49283051e-01 5.93757555e-02 3.30311984e-01 -8.73093829e-02
-1.39102876e+00 6.13325834e-01 -2.38421068e-01 -1.54846072e-01
-2.24015430e-01 5.21031499e-01 -1.49400473e-01 4.45732102e-02
4.90272135e-01 2.32277244e-01 6.08477630e-02 -5.21601856e-01
3.41891825e-01 5.49815357e-01 5.74973881e-01 -5.46432495e-01
6.67668641e-01 4.62459654e-01 -6.60513863e-02 -7.82181680e-01
-1.10620344e+00 -9.46909785e-01 -1.23170531e+00 -5.27571440e-02
9.49576259e-01 -7.78479993e-01 -1.42161757e-01 4.10965770e-01
-6.84930682e-01 -3.65171641e-01 -3.58537614e-01 4.40910161e-01
-4.12421912e-01 6.24709845e-01 -4.20784116e-01 -3.94590080e-01
-4.08710569e-01 -1.12894845e+00 8.49031448e-01 3.48298222e-01
-5.64520098e-02 -1.09777427e+00 1.48715913e-01 6.19815767e-01
-3.13447453e-02 3.46213765e-02 6.35822177e-01 -8.90889049e-01
-3.60478222e-01 -1.56029969e-01 -3.07935297e-01 5.85049868e-01
4.65710789e-01 -2.36082613e-01 -9.77727592e-01 -1.39589041e-01
-5.17390668e-02 -5.41186273e-01 1.13562644e+00 4.06276554e-01
1.35013795e+00 1.74291097e-02 -5.52414060e-01 4.88965064e-01
1.39927101e+00 1.40729053e-02 5.41443765e-01 6.80149794e-02
7.70597339e-01 8.34350467e-01 9.58697498e-01 1.11147717e-01
2.67778397e-01 2.00563043e-01 1.23316303e-01 -5.62134385e-01
1.53420428e-02 -1.43675163e-01 -1.91132382e-01 8.96096647e-01
-8.29074755e-02 1.59419134e-01 -1.05292082e+00 7.12329984e-01
-1.84362245e+00 -7.62946546e-01 -2.49486461e-01 2.03569460e+00
1.23181045e+00 2.66378850e-01 1.13402382e-01 1.31276280e-01
7.45998085e-01 -2.31652349e-01 -3.79940778e-01 1.23827048e-01
-7.95408338e-02 3.63290042e-01 2.88776189e-01 3.22390527e-01
-1.44298458e+00 1.13036585e+00 6.33882475e+00 7.88988829e-01
-1.07393014e+00 2.02413902e-01 1.11309922e+00 3.86629224e-01
-3.35406333e-01 3.50901075e-02 -5.35984516e-01 2.80182838e-01
3.64969999e-01 1.29322425e-01 -4.42966670e-01 7.07423091e-01
4.14907970e-02 -2.21286207e-01 -8.73130083e-01 9.56029356e-01
-9.88718122e-02 -1.25952685e+00 -1.39483109e-01 -2.07356572e-01
9.19648349e-01 1.44207738e-02 2.70525813e-02 3.11594158e-01
2.83790976e-01 -8.23579788e-01 2.16241002e-01 2.79761881e-01
3.69900316e-01 -7.05981851e-01 8.35861325e-01 5.63303828e-01
-9.70340073e-01 2.77033985e-01 -2.00092793e-01 -4.48304825e-02
5.23003489e-02 8.07659030e-01 -1.08829606e+00 6.11613691e-01
6.09031081e-01 9.12404835e-01 -5.77199817e-01 1.17299664e+00
-3.58471036e-01 8.16725373e-01 -1.85435206e-01 2.15260431e-01
4.74892616e-01 -4.15158212e-01 1.97948471e-01 1.37067008e+00
-2.97214866e-01 1.80082306e-01 5.66689491e-01 4.81983393e-01
-1.51347280e-01 5.19768119e-01 -5.77023625e-01 1.84265330e-01
3.25830914e-02 1.42436457e+00 -1.33956909e+00 -3.68454337e-01
-3.66389960e-01 9.90735531e-01 5.06005824e-01 6.22781105e-02
-5.69038391e-01 -2.37130132e-02 -2.10187081e-02 -1.31371528e-01
1.17086902e-01 5.99348284e-02 -5.73230922e-01 -9.12580967e-01
-1.99226514e-01 -6.16081536e-01 6.96410298e-01 -2.22149938e-01
-1.25524688e+00 4.14872020e-01 -1.88064370e-02 -1.28341901e+00
-1.13483407e-01 -2.20248610e-01 -5.82836211e-01 4.75262851e-01
-1.59107351e+00 -1.24470055e+00 -1.30459920e-01 4.65342790e-01
6.67452455e-01 8.52312297e-02 6.92275524e-01 2.52057791e-01
-7.41948247e-01 5.28757334e-01 5.94750158e-02 2.63633221e-01
7.31735289e-01 -1.53143764e+00 -1.47075906e-01 7.89864123e-01
2.38729969e-01 5.97531140e-01 4.58336264e-01 -5.99156141e-01
-7.77249694e-01 -1.03427052e+00 7.21714854e-01 2.88251769e-02
2.59410053e-01 4.44064150e-03 -1.00213075e+00 7.53034711e-01
3.34853321e-01 2.49369130e-01 1.16145194e+00 1.70033768e-01
-3.11797243e-02 -1.01726249e-01 -1.18574059e+00 3.74269009e-01
1.01548648e+00 -2.75884539e-01 -6.89823151e-01 5.73277771e-01
3.99539977e-01 -1.70456588e-01 -7.48089433e-01 7.25671947e-01
1.88144118e-01 -9.50861394e-01 8.69720459e-01 -4.38008040e-01
2.90437639e-01 -1.36527672e-01 6.89618811e-02 -1.20713031e+00
-7.37094367e-03 -7.20069230e-01 5.13017774e-01 1.43934059e+00
5.99667430e-01 -4.10736769e-01 1.03258443e+00 4.41895068e-01
-2.44840696e-01 -9.26541150e-01 -5.88503480e-01 -5.77238679e-01
1.24376647e-01 -2.91186929e-01 9.68589634e-02 1.19543660e+00
1.77416429e-01 6.70354247e-01 -1.77722603e-01 -3.41251604e-02
6.10440433e-01 2.22469017e-01 5.98051488e-01 -1.16771173e+00
-2.29258582e-01 -3.07020605e-01 -2.29102269e-01 -1.14466012e+00
3.47656459e-01 -1.10018671e+00 5.01705527e-01 -1.44857264e+00
6.15340173e-01 -7.61220574e-01 -6.62253857e-01 6.43126667e-01
-4.32299644e-01 4.42244738e-01 -1.96464673e-01 3.28356236e-01
-8.42442870e-01 3.99388760e-01 1.44842887e+00 -2.91156381e-01
-3.24793935e-01 1.93320185e-01 -7.30134368e-01 9.65838730e-01
1.18686390e+00 -4.82459068e-01 -7.05625117e-01 -2.72470206e-01
-1.86267823e-01 -1.38908967e-01 2.18601227e-02 -6.93963826e-01
1.07131608e-01 -1.55687779e-01 2.62375802e-01 -8.43382418e-01
-1.33446325e-03 -6.73490345e-01 -3.25728357e-01 5.18581152e-01
-4.89201695e-01 -2.60530710e-01 -1.26462411e-02 5.34196258e-01
-4.28911060e-01 -4.83547419e-01 1.02111006e+00 -3.53835553e-01
-1.07141876e+00 2.06091359e-01 -7.90167004e-02 2.35102251e-01
1.07296753e+00 -1.59238428e-01 2.70535201e-01 -1.36297857e-02
-8.35534513e-01 3.43821108e-01 2.04172164e-01 -1.54318223e-02
6.70447111e-01 -1.00962639e+00 -6.46146536e-01 9.77899879e-02
1.50001440e-02 3.73971403e-01 1.28701121e-01 9.47109222e-01
-4.14918453e-01 2.64456362e-01 -2.20779076e-01 -8.88040245e-01
-1.53237796e+00 4.00323719e-01 1.11597434e-01 -6.15329325e-01
-4.92041469e-01 1.13746858e+00 5.18486977e-01 -6.32710636e-01
8.15191790e-02 -1.35002673e-01 -4.77058589e-01 8.76438916e-02
1.68985695e-01 2.01436758e-01 4.76554446e-02 -6.53160393e-01
-2.58299977e-01 4.36594546e-01 -3.85561466e-01 2.72262786e-02
1.43032229e+00 -1.66358262e-01 -2.46204972e-01 4.08976406e-01
1.35961092e+00 -3.74860913e-01 -1.08519900e+00 -3.83903384e-01
4.40692425e-01 -3.60635906e-01 1.77332163e-01 -5.21881104e-01
-1.30079997e+00 7.99323082e-01 7.04757392e-01 2.34397668e-02
1.29838002e+00 4.82588932e-02 8.12200308e-01 4.11634445e-01
1.74904346e-01 -1.25384772e+00 2.56385177e-01 1.85175091e-01
4.68834192e-01 -1.72587478e+00 -2.76412442e-03 -9.44227815e-01
-6.58425450e-01 7.83139288e-01 7.71076500e-01 -8.23480934e-02
9.28916812e-01 8.40121433e-02 1.72909260e-01 -2.23775670e-01
-2.95875549e-01 -6.35564387e-01 5.53452790e-01 4.42280829e-01
7.61700451e-01 1.56737596e-01 -7.86468983e-01 5.31353295e-01
1.56077579e-01 6.24002516e-03 3.89495753e-02 1.19787526e+00
-4.03647065e-01 -1.30726159e+00 8.67645815e-02 5.36222041e-01
-6.44550562e-01 -1.05974182e-01 -5.95725596e-01 8.05794358e-01
1.89831182e-01 8.98953497e-01 -2.76187155e-02 -2.62131661e-01
-8.41732621e-02 -4.44205701e-02 5.56960046e-01 -9.25998211e-01
-5.88726461e-01 4.30629313e-01 -1.19038954e-01 -2.63745606e-01
-9.23313975e-01 -7.05217183e-01 -1.64889681e+00 4.45027679e-01
-5.31188965e-01 1.91702485e-01 1.23914875e-01 1.20469820e+00
1.88116096e-02 5.01456439e-01 7.42781579e-01 -5.81980050e-01
-3.46055388e-01 -8.36150467e-01 -4.35545921e-01 6.61303043e-01
5.27205504e-02 -5.08286536e-01 -7.71092921e-02 3.68148863e-01] | [14.741292953491211, -2.0901145935058594] |
4b3a4f9d-7c07-4567-9042-bc845f9a10d9 | benchmarking-scalable-methods-for-streaming | null | null | https://aclanthology.org/2021.acl-long.364 | https://aclanthology.org/2021.acl-long.364.pdf | Benchmarking Scalable Methods for Streaming Cross Document Entity Coreference | Streaming cross document entity coreference (CDC) systems disambiguate mentions of named entities in a scalable manner via incremental clustering. Unlike other approaches for named entity disambiguation (e.g., entity linking), streaming CDC allows for the disambiguation of entities that are unknown at inference time. Thus, it is well-suited for processing streams of data where new entities are frequently introduced. Despite these benefits, this task is currently difficult to study, as existing approaches are either evaluated on datasets that are no longer available, or omit other crucial details needed to ensure fair comparison. In this work, we address this issue by compiling a large benchmark adapted from existing free datasets, and performing a comprehensive evaluation of a number of novel and existing baseline models. We investigate: how to best encode mentions, which clustering algorithms are most effective for grouping mentions, how models transfer to different domains, and how bounding the number of mentions tracked during inference impacts performance. Our results show that the relative performance of neural and feature-based mention encoders varies across different domains, and in most cases the best performance is achieved using a combination of both approaches. We also find that performance is minimally impacted by limiting the number of tracked mentions. | ['Dan Bikel', 'Sameer Singh', 'Andrew McCallum', 'Robert L Logan IV'] | 2021-08-01 | null | null | null | acl-2021-5 | ['entity-disambiguation'] | ['natural-language-processing'] | [-0.14950727 0.10599585 -0.26451612 -0.37802586 -0.95764047 -0.99253255
0.8093878 0.86304474 -0.82890904 0.88727623 0.57304615 -0.16141509
-0.21146601 -0.8097355 -0.752261 -0.17745863 -0.4778144 0.78119683
0.54673487 -0.14367571 0.06183771 0.45603278 -1.4442861 0.01690047
0.85429186 0.47433433 -0.1271521 0.37923196 -0.49604735 0.4673107
-0.90208507 -0.78756773 0.11406159 -0.04087157 -1.2354467 -0.49454713
0.53533137 -0.2043205 -0.23287724 0.8720985 0.6869744 0.3334884
0.6035454 -1.0153404 -0.545295 1.150447 -0.2735005 0.62811536
0.42939794 -0.2662783 1.2526257 -0.561187 1.1461341 1.2025759
0.8057913 0.2516949 -1.2278781 -0.7304818 0.4043638 0.13942066
-1.5190171 -0.8293823 0.16105695 -0.2514151 1.2875283 0.08161006
0.07565214 0.94695663 -0.3851531 0.57996035 0.39016446 -0.27199826
0.3318924 -0.02308217 0.31936622 0.13145909 0.6534771 -0.12165095
-0.49892768 -0.4249798 0.24763404 -0.49076864 -0.38066417 -0.20216383
-1.2511735 0.73871934 0.4313928 0.670457 -0.3623641 -0.06685063
0.56805253 0.14748967 0.4728432 0.777619 -0.72119796 -0.37729132
-1.1772707 0.45058548 1.1977313 1.1286107 0.63146716 -0.3491824
-0.11948239 0.67768395 0.07612178 0.16217414 0.38332123 -1.0192451
0.6622488 0.47401416 0.37381953 -0.936186 -0.6845125 -0.16116473
-0.24898599 -0.364217 0.60065544 -0.57821 -0.6737902 1.9865716
0.5113504 0.5238302 0.25401747 0.78605723 1.0366522 0.41288278
0.41833693 -0.30923155 1.5046165 -0.39107677 -0.7381925 -0.29840913
0.693434 -0.72242665 0.35570705 -0.25859317 -0.9719792 -0.11727913
-0.82253337 -0.07993876 -0.6159173 -0.32920092 0.755738 0.5261538
-0.9532974 0.7195105 -1.1419173 -0.7469311 0.1697678 0.47984022
-0.4474221 0.1474105 -1.7836363 1.0012792 0.7496895 -0.31996045
-0.24531339 -0.94260377 -1.0029798 0.30596563 0.3083295 -0.5906211
1.3308841 -0.5551658 -0.85441613 0.5482649 -0.41533843 -0.74715126
0.2557654 -0.09061641 -0.6479757 0.11049194 0.431625 0.7614313
0.07862792 -1.1368588 -0.9940239 -0.19024579 0.31018206 0.2802519
-0.35534903 0.29414615 -0.76250404 -0.5007278 -0.04802179 -0.7193156
-0.17171149 -0.6150886 -0.2937994 -0.47817704 0.6650625 -0.4014272
1.3365412 -2.0641255 -0.21019353 -0.0163748 -0.01806813 0.38658515
0.0374111 0.6048986 0.05448547 0.51255995 -0.07884458 -0.36607087
0.13124989 0.09990685 -0.1189724 0.23465629 0.30424234 0.7716062
-1.1351378 -0.7387024 -0.17965956 0.63732296 -0.6243089 -0.03541696
-0.21552543 0.25998196 -0.40270942 0.466038 0.5195496 -0.33097988
0.50874 -0.2766441 -0.13039215 0.9236135 -1.6236537 1.6965036
-0.27524883 0.5811091 0.1301316 -0.63750815 0.5943563 0.5672195
0.5558932 -0.48810184 -0.22778338 0.18139637 -0.15371554 -0.3087464
0.90397495 -0.03822749 -0.20310868 0.4990504 0.13961475 0.51227486
0.6514866 0.55554605 1.512684 -0.21361262 0.3620628 -0.13126577
-0.02574402 0.3833561 0.8844031 0.6959712 -0.2790405 0.43407732
0.262445 -0.02012225 -0.621607 -0.9664815 -0.45626003 1.1220864
0.39027318 -0.61647254 -0.5657886 -0.81712997 0.24442631 0.7757407
-0.35164496 0.12596509 -0.64262956 -0.7783195 0.81575197 0.5816894
0.2375959 -0.9457618 -0.5278361 0.48898357 -0.43911523 -1.349416
-0.43605635 0.25066358 -0.62278473 -1.0710715 -0.44447383 -0.7090256
0.2709158 0.14049657 1.484966 -0.02308321 0.08107118 0.4402301
-0.5065836 -0.26806262 -0.17844178 0.63607424 0.1439168 -0.4041184
0.84604955 -0.484295 -0.4563357 0.01154557 -0.7002872 -0.71334505
0.4542555 0.5503028 0.20355958 0.1679024 0.721225 -1.3559566
0.58450055 -0.8549754 -0.47424957 0.2593315 -0.66180277 0.2266163
0.36657202 -0.30254492 -1.2284193 -0.07769175 -0.05404055 -0.1344677
-0.5245478 0.76111543 -0.09798305 0.4286283 0.71918994 -0.353319
-0.5824926 -0.55512327 0.5063227 0.6831159 0.6195839 -0.6952221
0.61627567 0.28182805 -0.63502735 -0.60137314 -0.63742757 -0.7683968
-0.78856415 0.2575131 0.77459854 -1.1957904 -0.49011785 -0.13029486
-1.3048265 0.03429481 -0.1730062 0.6159245 0.02088172 0.45581883
-0.8074988 -0.5084688 -0.27288517 -0.93687403 1.018094 0.29523733
-0.6607283 -1.0636324 0.2589409 0.03853345 0.39219004 0.262341
0.8677027 -1.3157097 -0.54603696 0.01514085 -0.16658698 -0.48615012
0.08504414 0.01115567 -0.81330234 -0.40800864 -0.63919485 -0.00976613
0.8165471 0.21373938 0.24586663 -0.3155766 -0.92525923 0.425909
1.3806571 0.22412786 0.21982959 0.72059333 0.37710673 0.5595188
0.62298447 0.3661041 0.8673149 0.7247206 0.16911082 0.16344011
-0.03427836 -0.1243277 0.1457338 0.5904977 0.30244118 -0.45606485
-1.193401 1.0960448 -1.6595583 -1.0832859 -0.03222446 2.1657562
1.127628 0.13393973 0.26980418 -0.11110774 1.0780408 0.04709182
-0.48174298 -0.12740426 -0.08018988 0.23827088 0.7143515 0.5099719
-1.3332626 0.9861972 6.617903 0.30873618 -0.90560615 0.06060216
0.13873488 -0.19555616 -0.31921855 0.20817812 -1.1578633 0.590572
1.3077731 -0.41850516 0.0924762 0.6239784 -0.18731707 0.02308388
-1.2732642 0.45036778 -0.17498638 -1.2583271 -0.16003151 -0.20856252
0.5910782 0.45040026 -0.41353628 0.46554363 0.8239197 -0.70815474
0.4819321 0.12103385 0.5839384 -0.7804176 0.81157625 0.11612102
-1.2954599 0.09496453 -0.14453064 0.22091827 0.43111533 0.63076806
-1.0002989 0.7861519 0.7683086 0.551421 -0.47066087 1.19652
-0.09212852 0.6992272 -0.650699 0.05705301 0.19329698 0.46871647
0.51522577 1.7114533 0.20043276 0.13198496 0.05968182 0.57284486
-0.2767802 0.0970467 -0.47206146 0.07213519 1.4275131 1.0962251
-0.7936058 -0.37295502 -0.6400856 0.5947233 0.6229609 0.3996437
-0.7036539 -0.6796408 0.90793216 0.1086873 0.6936436 -0.14863601
-0.01724106 -1.1911873 -0.15488835 -0.69099724 0.98489463 -0.2767633
-1.2558758 0.41932818 0.18820938 -0.79218304 -0.53392875 -0.0456268
-0.45350543 0.6115271 -1.5497702 -0.67965806 0.02519416 0.39285636
0.07821413 0.2669652 0.8360321 0.76561254 -0.50920355 0.8556455
0.12443317 0.52797616 1.1421959 -1.435759 0.7968199 1.0208107
0.44384742 0.9208257 0.8468327 -0.72092056 -1.0665278 -1.2143759
1.259441 -0.6493129 0.57476 -0.25660074 -1.1756642 0.9897581
0.18137565 -0.06268989 0.69840264 0.813118 -0.48020524 0.14752284
-1.038863 0.42222908 1.0936292 -0.3287346 -0.89052576 0.15192497
1.0048758 -0.5220752 -1.38776 0.39750305 0.17511919 -0.7551827
0.96902907 -0.694531 0.05496308 -0.37678906 -0.08042599 -1.3591795
-0.35733667 -0.39675736 -0.3534338 1.9886831 0.7965639 -0.5703995
0.6923693 0.90431637 0.11462165 -0.19499436 -0.9225763 -0.7550465
0.14293376 -0.26140445 0.92134273 1.4740229 0.3484088 0.56157774
0.10369273 0.597957 0.35899144 0.22388074 0.4846894 -1.3032941
-0.08455569 -0.41087398 -0.45932165 -0.7985919 0.37491196 -0.9900493
0.01056177 -1.5953156 0.01183755 -0.7924282 -0.32159048 0.6367776
-0.51422894 -0.2059602 0.13451594 0.22505988 -0.90706706 0.10233846
0.415604 -0.02866763 -0.3528222 -0.28107536 -0.92883277 0.3948498
0.5367866 -0.5888168 -0.14493084 -0.69465226 0.05626921 0.03157069
-0.07864168 -0.88113266 0.7697039 0.13752486 0.2175699 -0.4776298
0.06009116 -0.5724202 0.10437538 0.02885869 -0.42005423 0.12926099
0.2814353 0.6786527 -0.47352293 -0.2090864 0.5012911 -0.11149549
-1.0478162 0.163263 -0.16248132 0.72322917 0.94465786 0.15413342
-0.40794718 -0.28943732 -0.6408716 0.4473168 0.5488711 0.53140146
-0.13573027 -1.2213689 -0.691311 -0.33318627 0.14182517 0.21069387
-0.05809732 0.58122724 -0.13887544 0.56965685 0.2333398 -0.4366614
-1.1701047 0.5432783 0.18503943 -0.27776104 -0.51177543 0.88874817
-0.2648501 -0.636605 0.4227663 -0.16289203 -0.44565272 0.5019319
0.5865327 0.38300496 0.37450108 -0.79653823 -0.6791252 0.1023008
-0.3670377 0.05569297 1.385578 -0.25480226 0.0305891 0.23028189
1.0173924 0.02990672 -0.9489618 -0.35799834 0.5413411 -0.16185525
-0.14453846 -0.70371765 -0.94234073 0.39647225 0.36110952 0.31478864
0.8809875 0.18302755 0.8632969 0.44019106 0.4904455 -0.9390975
-0.713847 0.67944473 0.20224015 -1.1848046 0.01743243 -0.33166245
-0.5269849 0.75534475 0.61920375 0.26851878 0.5856871 0.59652793
0.10587201 -0.23610535 -0.9417176 -0.4881134 -0.05022879 0.57093567
0.72053427 -0.07570242 -0.42061 0.6454817 -0.28466564 -0.29076946
0.461943 1.0439363 -0.35808274 -1.1576132 -0.304734 0.49419987
-0.89573485 -0.27139542 -0.38000637 0.93775755 0.11032215 1.1084877
0.325244 -0.12023053 0.54491675 0.2465838 0.01067415 -0.8596434
-0.7173152 -0.37970743 0.50972563 -0.36218128 -0.604109 -1.088797
-1.5333035 -0.44497213 -0.5764074 0.6302009 0.4352843 0.9600468
0.78603524 0.3431778 0.06156168 -0.44043133 -0.34951332 -0.8267814
-0.36623704 0.6398351 0.17887901 -0.80941755 -0.1849429 -0.07381047] | [9.372690200805664, 8.934479713439941] |
804cba3a-7572-4845-b8fd-44fbc4d33ab0 | contrasting-centralized-and-decentralized | 2102.04402 | null | https://arxiv.org/abs/2102.04402v2 | https://arxiv.org/pdf/2102.04402v2.pdf | Contrasting Centralized and Decentralized Critics in Multi-Agent Reinforcement Learning | Centralized Training for Decentralized Execution, where agents are trained offline using centralized information but execute in a decentralized manner online, has gained popularity in the multi-agent reinforcement learning community. In particular, actor-critic methods with a centralized critic and decentralized actors are a common instance of this idea. However, the implications of using a centralized critic in this context are not fully discussed and understood even though it is the standard choice of many algorithms. We therefore formally analyze centralized and decentralized critic approaches, providing a deeper understanding of the implications of critic choice. Because our theory makes unrealistic assumptions, we also empirically compare the centralized and decentralized critic methods over a wide set of environments to validate our theories and to provide practical advice. We show that there exist misconceptions regarding centralized critics in the current literature and show that the centralized critic design is not strictly beneficial, but rather both centralized and decentralized critics have different pros and cons that should be taken into account by algorithm designers. | ['Christopher Amato', 'Brett Daley', 'Yuchen Xiao', 'Xueguang Lyu'] | 2021-02-08 | null | null | null | null | ['misconceptions'] | ['miscellaneous'] | [-5.69430470e-01 1.55278683e-01 -2.88656890e-01 -1.74356729e-01
-4.95149434e-01 -8.04304421e-01 6.24168634e-01 2.46417031e-01
-5.66752851e-01 7.63610661e-01 3.50753158e-01 -5.48948050e-01
-1.69294283e-01 -3.44714254e-01 -3.33824426e-01 -7.91917622e-01
-1.07632354e-01 6.31043375e-01 1.21938311e-01 -3.88591588e-01
9.49702859e-02 4.49837536e-01 -1.22000825e+00 -3.48045856e-01
6.22672975e-01 2.47640535e-01 -2.67953366e-01 9.79399025e-01
2.70579249e-01 1.26844370e+00 -1.03378963e+00 -3.53384912e-01
5.80166996e-01 -7.66891539e-01 -6.29813373e-01 2.12176502e-01
-2.22136855e-01 -7.27631867e-01 4.84098420e-02 8.89465153e-01
7.22404659e-01 7.32768178e-02 4.06583369e-01 -1.54600251e+00
-3.99843127e-01 9.21234608e-01 -5.32848835e-01 2.01418862e-01
9.32925493e-02 8.02798569e-01 1.11175716e+00 -1.29320115e-01
4.32190418e-01 1.30659580e+00 5.18453479e-01 6.06622338e-01
-1.15533400e+00 -3.79186451e-01 6.48173690e-01 -2.63816684e-01
-9.55938339e-01 -3.80809963e-01 7.82523811e-01 -4.02891606e-01
9.55891609e-01 -1.11473184e-02 9.77346718e-01 1.01354682e+00
1.37966141e-01 8.64914834e-01 1.17803609e+00 -5.42934418e-01
5.57369947e-01 2.43056700e-01 -3.48768950e-01 5.10963559e-01
8.10929537e-01 3.47931474e-01 -1.26770571e-01 -4.43129480e-01
9.66865480e-01 -1.68705791e-01 1.50279805e-01 -6.15030706e-01
-1.14402211e+00 1.07580793e+00 9.49797183e-02 4.17451173e-01
-5.48287511e-01 6.43285811e-01 7.59917736e-01 6.24653816e-01
2.12371156e-01 7.39657283e-01 -4.06465560e-01 -3.91735613e-01
-5.62462926e-01 6.00896657e-01 1.07108557e+00 5.85155845e-01
6.60498679e-01 6.67511284e-01 3.22983354e-01 3.42132568e-01
6.24455094e-01 4.82261211e-01 3.98541003e-01 -1.52399886e+00
2.44656026e-01 3.44065338e-01 5.72708070e-01 -8.09415817e-01
-3.67644936e-01 -2.44767934e-01 -3.03810388e-01 8.81833553e-01
7.22089708e-01 -9.69623804e-01 -1.01105943e-01 1.61907864e+00
2.60035187e-01 -2.99458504e-01 3.05590928e-01 1.09809959e+00
7.74676073e-03 2.93182343e-01 1.63940221e-01 -2.33203113e-01
7.67860234e-01 -1.25329924e+00 -4.73594934e-01 -3.02791208e-01
7.53014624e-01 -5.04521430e-01 8.18630219e-01 4.22251761e-01
-1.07634950e+00 2.29546670e-02 -9.78750408e-01 4.31913376e-01
-5.13466746e-02 -8.03130716e-02 8.39816332e-01 7.56217301e-01
-1.14910030e+00 6.54585600e-01 -1.28315067e+00 -3.35808724e-01
-1.11441866e-01 2.79098898e-01 2.76672482e-01 4.68278408e-01
-8.36999118e-01 1.20839036e+00 1.44653514e-01 -8.02487694e-03
-1.20987535e+00 5.77298142e-02 -4.94913995e-01 1.98118776e-01
5.87960899e-01 -7.42971420e-01 1.98734426e+00 -1.85686779e+00
-2.01990294e+00 3.26079607e-01 2.24682540e-01 -4.55707103e-01
8.25828969e-01 -2.24492978e-02 1.81062575e-02 3.01245358e-02
-6.11789376e-02 1.84654906e-01 8.59345734e-01 -1.38314331e+00
-6.30470455e-01 3.99560221e-02 5.99114060e-01 6.77829087e-01
-2.23396923e-02 1.86527267e-01 3.43900651e-01 -3.37378383e-01
-5.65770030e-01 -9.30896640e-01 -5.60436487e-01 -2.09143370e-01
1.55146420e-01 -3.73447567e-01 4.68337089e-01 -1.12316899e-01
9.54067886e-01 -1.87724566e+00 -8.53087306e-02 1.45226434e-01
4.14785504e-01 -1.28212515e-02 -1.03027165e-01 9.84468460e-01
2.67949939e-01 3.15478981e-01 2.76777685e-01 -1.84135243e-01
3.47415268e-01 3.20640475e-01 -8.42692927e-02 8.16846132e-01
-1.41507071e-02 5.66915929e-01 -1.30084097e+00 -4.48850155e-01
2.00990409e-01 1.18681990e-01 -6.90486193e-01 1.07815146e-01
-3.73904735e-01 4.83293295e-01 -9.42229033e-01 5.54361343e-01
-2.95645408e-02 -3.80288750e-01 7.92067468e-01 6.28964603e-01
-4.10423905e-01 2.85855472e-01 -1.28342342e+00 1.16298068e+00
-1.45992026e-01 6.29648089e-01 6.00714922e-01 -1.04803646e+00
5.77291369e-01 8.21688652e-01 7.59390473e-01 -5.37678480e-01
1.65169418e-01 2.92089969e-01 5.30039549e-01 -4.07737821e-01
3.58632922e-01 -2.63044834e-01 1.08626038e-01 1.06126344e+00
-2.54831582e-01 -1.70276180e-01 3.62716049e-01 1.26687884e-01
1.22361875e+00 2.53971189e-01 5.08837461e-01 -2.84892917e-01
4.06621881e-02 4.38433558e-01 9.04500544e-01 1.04671752e+00
-6.08018696e-01 9.42163095e-02 7.63248205e-01 -5.60241461e-01
-1.44628096e+00 -5.35867155e-01 5.61150610e-01 1.19291341e+00
8.33886415e-02 -3.68783593e-01 -5.25975406e-01 -7.28567958e-01
2.33242959e-01 4.02749181e-01 -2.91072011e-01 2.39037544e-01
-6.89858556e-01 -4.06739861e-01 5.42636752e-01 5.46093702e-01
1.35566339e-01 -1.11927545e+00 -1.64256263e+00 4.14455831e-01
3.56251061e-01 -5.26372731e-01 -1.22073650e-01 1.53439373e-01
-8.06252301e-01 -1.30175555e+00 -6.76598489e-01 -3.16954374e-01
7.06276536e-01 4.28936958e-01 1.28330910e+00 4.15574670e-01
2.66113132e-01 1.09345436e+00 -4.84082133e-01 -6.52305067e-01
-8.87462735e-01 -7.40255713e-02 2.35726848e-01 -4.50839460e-01
-8.62118304e-02 -4.69678164e-01 -7.43198514e-01 2.87127197e-01
-7.06035376e-01 -1.25497073e-01 8.01798284e-01 8.61594498e-01
-2.14145824e-01 -1.26986265e-01 8.50920856e-01 -8.75830531e-01
1.10975969e+00 -3.76874387e-01 -1.03580606e+00 1.90970942e-01
-9.93083179e-01 8.42579454e-02 9.56351101e-01 -3.59137356e-01
-9.98262942e-01 -1.57271013e-01 8.86986256e-02 -3.19772869e-01
-1.99821934e-01 6.11297846e-01 3.68267983e-01 -3.17854434e-02
7.36874104e-01 -1.58639595e-01 3.62445086e-01 -1.52641833e-01
3.64807039e-01 2.11429030e-01 -3.78169835e-01 -9.87150848e-01
6.28079355e-01 4.37952936e-01 -2.69052088e-01 -6.76227450e-01
-2.31752470e-01 -2.25883529e-01 -1.61462709e-01 -4.59987342e-01
4.15759951e-01 -1.02684760e+00 -1.20358729e+00 1.05195209e-01
-9.75545943e-01 -8.72286618e-01 -5.96444488e-01 5.21979451e-01
-8.98779988e-01 3.00374687e-01 -7.29477823e-01 -1.22223341e+00
-1.18317418e-01 -1.51418257e+00 5.04527688e-01 4.69052225e-01
-3.13877314e-01 -1.40052485e+00 5.49180329e-01 -1.05604202e-01
7.54238069e-01 2.59382546e-01 5.12155652e-01 -8.63143623e-01
-3.84435743e-01 4.64801356e-04 3.42403442e-01 2.39259586e-01
9.56855416e-02 4.27889258e-01 -7.49149561e-01 -6.61204517e-01
-7.45026171e-02 -6.20105028e-01 7.10152462e-02 3.36002648e-01
2.78926194e-01 -6.23629928e-01 3.87938432e-02 -1.55436486e-01
1.68057847e+00 2.13142961e-01 2.77706049e-02 7.73900807e-01
2.44208872e-01 4.62691903e-01 3.87900472e-01 9.03113008e-01
4.21600014e-01 2.19878778e-01 4.51350033e-01 1.23163946e-01
3.79966468e-01 -2.07337081e-01 7.64528275e-01 8.24638486e-01
-1.83097035e-01 -1.66245818e-01 -1.10900402e+00 5.81761956e-01
-2.22164416e+00 -1.09918535e+00 4.56409991e-01 2.09714651e+00
7.12260842e-01 1.07938349e-01 7.13867009e-01 -1.74865782e-01
5.83107829e-01 3.02352250e-01 -6.92988575e-01 -7.09024251e-01
1.06919102e-01 -2.39646420e-01 6.69404805e-01 3.97516191e-01
-6.90523446e-01 7.56232083e-01 7.78287363e+00 1.66539982e-01
-1.21341872e+00 1.41240552e-01 3.63862425e-01 -1.58912808e-01
-2.20185831e-01 6.53321147e-01 -3.59175920e-01 1.95974395e-01
1.08819902e+00 -6.29721403e-01 6.45211756e-01 1.19178069e+00
6.23623550e-01 -2.76417911e-01 -1.08175683e+00 5.27382076e-01
-3.12078059e-01 -1.03510749e+00 -5.11149228e-01 6.20427318e-02
8.47571731e-01 1.59388945e-01 -1.87942579e-01 3.75947148e-01
1.36627340e+00 -8.13764811e-01 9.48814273e-01 4.11666594e-02
-2.70023067e-02 -6.61624312e-01 5.06782234e-01 6.55635357e-01
-9.19351220e-01 -3.82860333e-01 -2.74789572e-01 -7.03701198e-01
9.26712975e-02 1.54032096e-01 -7.16860116e-01 1.96778312e-01
4.28951532e-01 5.28890193e-01 -2.25186348e-01 1.04090798e+00
-4.53179955e-01 7.97003925e-01 -3.15236151e-01 -3.11497539e-01
6.33610487e-01 -2.02911988e-01 4.20728624e-01 8.07947040e-01
-2.08501354e-01 -7.27474988e-02 4.48259771e-01 6.24930680e-01
4.26866919e-01 -8.30165073e-02 -7.46391475e-01 -4.31627303e-01
5.09512544e-01 1.21127605e+00 -9.61526513e-01 -3.41274321e-01
-7.89645314e-01 2.60186970e-01 3.98752272e-01 5.88880479e-01
-7.34626591e-01 -1.04697095e-02 7.24834681e-01 -1.06010683e-01
2.09124252e-01 -4.10580635e-01 -8.30029622e-02 -1.31435192e+00
-3.34297329e-01 -1.42318141e+00 2.67673075e-01 -3.54606658e-01
-1.22695065e+00 -3.22199278e-02 3.79760377e-02 -1.17306089e+00
-7.12096751e-01 -3.63933951e-01 -8.00063372e-01 4.12709951e-01
-1.50518739e+00 -7.42299795e-01 1.16864979e-01 3.87823552e-01
5.34409046e-01 2.08127312e-03 6.82579458e-01 -6.30503744e-02
-6.97382450e-01 2.58438349e-01 5.54252446e-01 1.96256176e-01
6.52741194e-01 -1.29596806e+00 2.25930840e-01 7.22653568e-01
-1.78297803e-01 6.58185244e-01 1.01663947e+00 -4.98888314e-01
-1.77621949e+00 -4.33137804e-01 3.06336999e-01 -3.40001404e-01
8.65655899e-01 1.92631066e-01 -3.81238401e-01 1.06786776e+00
7.54083753e-01 -2.86711901e-01 3.89786720e-01 1.69568554e-01
-1.61844268e-01 7.95332938e-02 -8.41907978e-01 8.06855977e-01
3.33449513e-01 -2.67718852e-01 -3.80160660e-01 3.38416070e-01
3.10737610e-01 -2.64019072e-01 -7.89745569e-01 -1.65438622e-01
6.19640887e-01 -1.16081226e+00 5.03033757e-01 -5.84437490e-01
8.54425877e-02 -1.98310062e-01 2.52008915e-01 -1.55958521e+00
-1.64683342e-01 -8.62338543e-01 9.83428583e-02 8.60710442e-01
3.52491558e-01 -1.07460439e+00 8.54482770e-01 8.89690638e-01
-1.23085156e-02 -5.61664581e-01 -7.55674660e-01 -9.67360675e-01
4.90938246e-01 4.97403257e-02 2.45744735e-01 1.01459813e+00
3.58102351e-01 2.77992517e-01 -3.43399137e-01 1.80447195e-02
4.78750587e-01 1.14783294e-01 1.06912613e+00 -8.33366096e-01
-6.09637022e-01 -7.74752498e-01 -3.24132442e-02 -8.81458580e-01
1.87321275e-01 -3.28402460e-01 6.44745901e-02 -1.46772873e+00
-6.93093091e-02 -5.68511724e-01 -4.58125323e-02 4.52867091e-01
2.97477156e-01 -5.07955968e-01 4.77665275e-01 5.22515953e-01
-9.60332394e-01 5.30061722e-01 1.11485755e+00 -3.19250636e-02
-2.67490804e-01 -1.68238506e-01 -8.09487700e-01 8.89170110e-01
9.35525239e-01 -6.49233103e-01 -5.31643569e-01 -7.24171221e-01
5.36293864e-01 3.04835111e-01 2.22480774e-01 -7.86738873e-01
3.31290036e-01 -8.18186700e-01 3.08195432e-03 2.82107443e-01
-4.70017232e-02 -9.18088675e-01 -2.39478592e-02 7.68827081e-01
-4.76199895e-01 7.84665942e-01 -6.68788776e-02 5.00101507e-01
-4.43267822e-02 -3.24700266e-01 7.14018345e-01 -5.98110914e-01
-4.16696787e-01 -1.14276983e-01 -8.07501256e-01 2.11809233e-01
1.29557920e+00 -1.95463095e-02 -2.91808128e-01 -8.87993157e-01
-4.42284018e-01 5.68418264e-01 8.51017058e-01 3.03293753e-04
2.87576497e-01 -8.82995307e-01 -5.07556975e-01 -2.73954272e-01
-2.93184221e-01 -1.55452192e-01 -3.11917961e-01 9.01418686e-01
-7.67158449e-01 2.74991125e-01 -2.26192221e-01 -3.18828523e-01
-8.05237532e-01 4.18598264e-01 5.23162663e-01 -3.10354233e-01
-8.16090822e-01 9.50542018e-02 -2.00989321e-01 -2.07534537e-01
2.27015465e-01 -1.58603802e-01 4.81113084e-02 -1.36688694e-01
1.61999956e-01 4.11184192e-01 -3.38320196e-01 -2.06195578e-01
-1.52157545e-01 1.00706786e-01 -5.26742190e-02 -6.25779450e-01
1.27615559e+00 -2.20296115e-01 6.85223043e-02 6.79576337e-01
3.83925587e-01 -8.24123770e-02 -1.50975120e+00 -3.18982862e-02
3.10821850e-02 -4.14504915e-01 -3.15561622e-01 -4.86943841e-01
-9.89572704e-01 5.60206294e-01 6.18051328e-02 6.68527246e-01
6.72071934e-01 -2.66406745e-01 1.25254616e-01 6.24823570e-01
4.98727590e-01 -1.44843292e+00 1.46078944e-01 4.87413764e-01
5.96362650e-01 -1.19378877e+00 4.02632415e-01 4.42172825e-01
-1.18334627e+00 1.07710385e+00 8.44774544e-01 -6.96772575e-01
4.08025920e-01 2.40905866e-01 3.95641893e-01 -1.95531577e-01
-1.28462875e+00 -9.19361413e-02 -7.06757724e-01 3.01653773e-01
4.65853691e-01 3.69698405e-02 -6.15501106e-01 1.10427529e-01
1.83467492e-01 -1.30968377e-01 8.93949687e-01 1.61721838e+00
-3.87149721e-01 -1.21335959e+00 -4.39100206e-01 1.83178917e-01
-4.39850807e-01 4.57796037e-01 -4.97758418e-01 1.12034202e+00
-4.16765839e-01 1.05657613e+00 -2.56341904e-01 1.12567075e-01
1.51054831e-02 1.51383961e-02 3.21247220e-01 -6.55413270e-01
-1.17864776e+00 3.49084795e-01 5.83414659e-02 -3.14286113e-01
-6.69817030e-01 -6.29747391e-01 -1.21164620e+00 -5.72296679e-01
-3.28011394e-01 5.89060664e-01 6.88005507e-01 9.02529478e-01
3.75062019e-01 4.05265778e-01 7.35096812e-01 -9.82858479e-01
-1.44860291e+00 -6.02442145e-01 -5.77952862e-01 2.28904411e-01
5.03480256e-01 -5.08398473e-01 -5.65732121e-01 -2.24913090e-01] | [3.825136423110962, 2.0878751277923584] |
f26051ea-0f02-4850-a66b-1b6be67fae9c | csecu-dsg-at-semeval-2021-task-5-leveraging | null | null | https://aclanthology.org/2021.semeval-1.135 | https://aclanthology.org/2021.semeval-1.135.pdf | CSECU-DSG at SemEval-2021 Task 5: Leveraging Ensemble of Sequence Tagging Models for Toxic Spans Detection | The upsurge of prolific blogging and microblogging platforms enabled the abusers to spread negativity and threats greater than ever. Detecting the toxic portions substantially aids to moderate or exclude the abusive parts for maintaining sound online platforms. This paper describes our participation in the SemEval 2021 toxic span detection task. The task requires detecting spans that convey toxic remarks from the given text. We explore an ensemble of sequence labeling models including the BiLSTM-CRF, spaCy NER model with custom toxic tags, and fine-tuned BERT model to identify the toxic spans. Finally, a majority voting ensemble method is used to determine the unified toxic spans. Experimental results depict the competitive performance of our model among the participants. | ['Abu Nowshed Chy', 'Radiathun Tasnia', 'Fareen Tasneem', 'Jannatun Naim', 'Tashin Hossain'] | 2021-08-01 | null | null | null | semeval-2021 | ['toxic-spans-detection'] | ['natural-language-processing'] | [-2.42217839e-01 -1.95084795e-01 -2.36406058e-01 -2.25278050e-01
-8.84686410e-01 -9.64884937e-01 4.21202451e-01 8.75584856e-02
-6.34824753e-01 8.78000975e-01 5.29686272e-01 -3.92615020e-01
4.28911656e-01 -4.44769502e-01 -5.68221249e-02 -2.93275297e-01
-2.44600579e-01 1.49509162e-01 -1.12897471e-01 -3.89445096e-01
6.50460660e-01 5.09149134e-01 -4.98651683e-01 7.70986140e-01
6.74572766e-01 7.94873714e-01 -5.24513841e-01 7.88620710e-01
-1.19539194e-01 1.18225539e+00 -1.10514975e+00 -1.04315257e+00
3.18272784e-02 -2.58773804e-01 -8.16382229e-01 -4.66390312e-01
4.15912837e-01 -3.70891571e-01 -2.63564736e-01 1.14913523e+00
5.56707501e-01 2.50210494e-01 6.03818178e-01 -1.19819748e+00
-8.72541368e-01 1.33677626e+00 -8.86554480e-01 9.03418541e-01
4.18405771e-01 9.23330337e-02 1.13055658e+00 -5.33644557e-01
7.34870374e-01 1.46432304e+00 1.01503801e+00 4.39513743e-01
-7.44969428e-01 -1.01134920e+00 -5.45961745e-02 9.58402455e-02
-1.19596541e+00 -2.07790211e-01 8.97587717e-01 -6.44591987e-01
1.28649485e+00 2.75575340e-01 4.43194866e-01 1.94722259e+00
5.52253306e-01 4.45651323e-01 1.00961661e+00 -1.58971950e-01
-1.99668571e-01 -6.29094988e-02 6.83605850e-01 6.78594589e-01
2.27467835e-01 -2.25488588e-01 -7.42504656e-01 -9.01023567e-01
-1.97307631e-01 2.04056334e-02 -8.61000493e-02 1.14769399e+00
-4.10574287e-01 1.20988929e+00 3.08896095e-01 7.29207218e-01
-3.34854066e-01 1.16755985e-01 1.02488887e+00 3.52865845e-01
1.05988598e+00 7.41706729e-01 -2.74219513e-01 -1.58223405e-01
-1.15084231e+00 2.29107797e-01 1.28473616e+00 6.21708810e-01
2.31635883e-01 -1.16945893e-01 -2.64575124e-01 7.16648936e-01
3.28285366e-01 3.72952670e-01 3.52850705e-01 -6.96558595e-01
5.69169998e-01 2.41774783e-01 -5.86607046e-02 -1.26682305e+00
-6.51706636e-01 -4.65247691e-01 -5.04109740e-01 -3.01359028e-01
2.84027785e-01 -7.15670764e-01 -6.67809069e-01 1.67404258e+00
1.85056161e-02 -3.72420847e-02 -6.40619278e-01 4.50831831e-01
2.68158585e-01 5.92291415e-01 6.04566097e-01 -4.95883822e-01
1.49053764e+00 -5.38887501e-01 -9.61961091e-01 -1.30608035e-02
8.72433782e-01 -8.04851294e-01 5.84274948e-01 7.36338079e-01
-6.68066144e-01 8.06464478e-02 -9.97902334e-01 -1.92942441e-01
-4.90637690e-01 -6.38355017e-01 2.29341924e-01 1.01045418e+00
-4.82561290e-01 7.46373951e-01 -4.38877046e-01 -1.25533462e-01
5.41189313e-01 -1.74424544e-01 -6.05977215e-02 2.42959693e-01
-1.91081762e+00 1.15880895e+00 3.42150033e-01 -4.51109372e-02
-6.85373247e-01 -8.06586146e-01 -6.51070416e-01 -1.32127598e-01
-1.23085365e-01 -3.38016331e-01 1.35437560e+00 -5.63724697e-01
-8.71980667e-01 9.02972937e-01 3.55727434e-01 -6.67820632e-01
5.47413707e-01 -7.13854432e-01 -7.72095263e-01 2.41761297e-01
2.09098756e-01 2.54115045e-01 5.38133919e-01 -5.51534355e-01
-8.82351212e-03 -4.30074722e-01 -2.98984796e-01 -1.70143396e-01
-7.60257065e-01 8.77153873e-01 4.45047826e-01 -5.74846685e-01
-6.26961648e-01 -6.52494848e-01 -9.67240632e-02 -5.54450154e-01
-1.12094116e+00 -5.70272624e-01 8.20494950e-01 -1.31600177e+00
1.80472636e+00 -1.83837008e+00 -5.80418706e-01 2.34738752e-01
6.32790625e-01 -3.18263993e-02 9.20415297e-02 1.02203214e+00
3.13751161e-01 8.83402765e-01 6.35115281e-02 -5.27705014e-01
1.54177561e-01 -1.56768888e-01 -6.43873930e-01 6.98052585e-01
-3.69191736e-01 6.28868818e-01 -7.67997682e-01 -7.34316707e-01
-5.42755544e-01 2.62213558e-01 -1.93057179e-01 3.94564420e-02
-1.26482829e-01 1.82276256e-02 -5.59142113e-01 5.56418240e-01
6.18358076e-01 -6.26988187e-02 1.53207272e-01 1.92269042e-01
-2.35800445e-01 7.70263135e-01 -1.63462892e-01 1.33712113e+00
-1.30982166e-02 6.82024300e-01 1.10933356e-01 -7.25068673e-02
1.01429939e+00 2.15077072e-01 1.36060014e-01 -3.30886871e-01
4.81125921e-01 1.46293044e-01 -5.38315140e-02 -8.18954945e-01
9.37447429e-01 -6.92443848e-01 -8.24068666e-01 6.04658544e-01
-7.78269842e-02 4.29164231e-01 8.46224874e-02 6.37011826e-01
1.36776292e+00 -3.97241414e-01 4.89014775e-01 -1.79816812e-01
1.27702221e-01 -7.92943314e-02 7.99258530e-01 7.68098891e-01
-7.02662349e-01 1.82311937e-01 8.65038812e-01 -3.96774262e-01
-8.48206401e-01 -7.02528715e-01 1.03510134e-01 1.47592366e+00
-1.53939515e-01 -3.50736588e-01 -7.66317368e-01 -1.17182255e+00
-1.81999266e-01 1.07648349e+00 -5.40627480e-01 1.05133176e-01
-7.09936261e-01 -7.34575689e-01 1.46998549e+00 3.19262475e-01
4.89321679e-01 -1.30616927e+00 -3.59805107e-01 2.30410352e-01
-6.56383157e-01 -1.01552188e+00 -7.92721868e-01 1.47002831e-01
-2.59795934e-01 -9.13721383e-01 -2.01807201e-01 -3.50840718e-01
-1.31239921e-01 -2.05654457e-01 8.45758617e-01 1.72351673e-01
-2.53562242e-01 -3.44220430e-01 -3.72258157e-01 -5.95880330e-01
-3.88189942e-01 2.27289960e-01 1.91606492e-01 -3.47288996e-01
5.71757615e-01 -8.09934735e-01 -2.62032986e-01 -3.64565164e-01
-7.76447535e-01 -6.23043418e-01 -4.71317656e-02 3.05426359e-01
-2.81362176e-01 -5.11299789e-01 1.05785942e+00 -1.48051453e+00
1.14606428e+00 -1.02376592e+00 3.98792267e-01 7.50532076e-02
-5.00571012e-01 -4.32031602e-01 8.01164806e-01 -7.23629355e-01
-1.15592933e+00 -4.90731239e-01 -4.64460403e-01 -3.12301099e-01
9.84389707e-02 5.62339544e-01 -1.04325123e-01 3.36370230e-01
9.86122251e-01 -3.86411577e-01 -5.49253821e-01 -8.30610216e-01
5.38997591e-01 1.02342725e+00 4.22569215e-01 -1.87826589e-01
5.95402837e-01 1.12470850e-01 -6.58474982e-01 -8.42404306e-01
-1.23358619e+00 -6.98904812e-01 -4.52549130e-01 -2.85962790e-01
7.96620786e-01 -5.92695296e-01 -7.83110917e-01 5.85663140e-01
-1.82316697e+00 1.96681812e-01 1.27307326e-01 1.16817020e-01
1.28199205e-01 8.25927079e-01 -1.57607949e+00 -1.14163291e+00
-1.13822913e+00 -1.66443989e-01 4.01964307e-01 2.17285261e-01
-9.15097594e-01 -9.63401377e-01 7.41107702e-01 5.17395854e-01
2.81201303e-01 7.55830288e-01 8.09677839e-01 -1.55341470e+00
3.73528779e-01 -4.12405014e-01 4.95970435e-03 2.73956746e-01
-4.71408576e-01 6.76084906e-02 -1.07906473e+00 6.09649383e-02
3.34837870e-03 -7.36334682e-01 1.15507078e+00 -2.19569001e-02
7.34406650e-01 -8.79594445e-01 -4.19860482e-01 6.89813569e-02
1.05853379e+00 1.26821786e-01 5.26865005e-01 3.20426017e-01
5.82040668e-01 5.49152613e-01 2.71897372e-02 8.37090790e-01
-1.28681203e-02 -8.31276998e-02 4.17052299e-01 5.36825836e-01
2.15752646e-01 -6.13363266e-01 7.40643680e-01 9.77766693e-01
5.25556982e-01 -9.64683235e-01 -7.63229251e-01 5.85676968e-01
-1.41239643e+00 -1.50748932e+00 -5.43113410e-01 1.27370811e+00
7.52566338e-01 4.25334781e-01 2.62884080e-01 -5.42693250e-02
1.11372066e+00 6.85696065e-01 -2.42284387e-01 -1.27694440e+00
-2.35377271e-02 -8.11901912e-02 3.39187294e-01 5.48617721e-01
-1.00425768e+00 8.71524215e-01 6.79925585e+00 1.06044734e+00
-7.47858047e-01 5.38618386e-01 6.94241166e-01 -2.43413687e-01
-4.26254421e-01 -2.56522596e-01 -9.28192973e-01 7.99699783e-01
1.37809622e+00 -1.36027351e-01 -1.32732978e-02 7.90503085e-01
4.36263025e-01 1.71729743e-01 -4.86474454e-01 5.28741658e-01
5.38432896e-01 -1.05090499e+00 -2.06979379e-01 1.55670583e-01
4.39653903e-01 3.12831938e-01 -1.92673989e-02 4.07029808e-01
5.88813245e-01 -8.40215266e-01 8.12455893e-01 7.89043069e-01
5.55694580e-01 -8.66230249e-01 6.58116400e-01 7.02639341e-01
-5.02831340e-01 -1.48397788e-01 -1.70100093e-01 -3.68261278e-01
6.90558314e-01 1.06545973e+00 -9.42826688e-01 1.73107296e-01
5.54877102e-01 5.51148415e-01 -4.59470779e-01 8.66986513e-01
-4.83307630e-01 1.19905806e+00 -3.12403172e-01 -3.68747324e-01
2.51282305e-01 1.59961373e-01 9.22413230e-01 1.91596210e+00
2.50815988e-01 8.50292966e-02 1.46529123e-01 8.35839391e-01
-9.44420815e-01 -1.18419128e-02 -6.86125755e-01 -4.11144316e-01
7.94691086e-01 1.71977925e+00 -7.19283342e-01 -1.80482596e-01
1.64593399e-01 7.88586140e-01 6.22978210e-01 7.68791465e-03
-7.93878913e-01 -4.67409372e-01 3.99340987e-01 -5.28120883e-02
-1.12084165e-01 -1.34202778e-01 -7.06515610e-01 -8.68893743e-01
-3.20009828e-01 -6.15590572e-01 8.77520025e-01 -6.58551037e-01
-2.04362679e+00 8.90520513e-01 -2.56574750e-01 -6.90075994e-01
-1.68294489e-01 -2.09067836e-01 -1.22928035e+00 7.22317636e-01
-8.85625243e-01 -1.20750821e+00 2.86609113e-01 2.38403063e-02
3.85347754e-01 1.78960919e-01 5.09569228e-01 2.67456681e-01
-7.79713988e-01 6.09913707e-01 -3.44648689e-01 5.11556447e-01
9.27245021e-01 -8.89381647e-01 4.21717405e-01 9.45142984e-01
-9.39557701e-02 6.26093805e-01 8.84404302e-01 -1.24021208e+00
-2.98784852e-01 -1.12337482e+00 1.70635986e+00 -8.44329894e-01
1.24910617e+00 -4.51668203e-01 -9.81446385e-01 7.76104987e-01
7.60163784e-01 -4.60621357e-01 1.16917229e+00 1.75336897e-01
-1.07768834e+00 6.10005200e-01 -1.22198844e+00 2.35675499e-01
1.02780557e+00 -6.18234515e-01 -8.69015574e-01 3.74850243e-01
8.91245782e-01 1.90969065e-01 -8.04402590e-01 -8.69882479e-02
5.14724970e-01 -7.05479026e-01 4.06482995e-01 -1.03086329e+00
8.29016984e-01 4.04345244e-01 7.73060843e-02 -1.06739557e+00
-5.14821827e-01 -9.69903409e-01 -2.87248373e-01 1.84481728e+00
5.20446181e-01 -3.39112192e-01 4.13002998e-01 6.25441790e-01
-2.92148322e-01 -4.07075405e-01 -1.05205297e+00 -5.96519709e-01
3.65448266e-01 -5.91291964e-01 1.12722903e-01 1.24862969e+00
5.03536701e-01 7.15839922e-01 -8.42220783e-01 -1.50000453e-01
5.22966802e-01 -5.54295816e-02 -1.11998208e-01 -1.19371116e+00
-6.15515634e-02 -3.11373264e-01 4.11256365e-02 -6.33691430e-01
5.15454948e-01 -1.01700306e+00 -5.24083376e-02 -9.66182113e-01
6.86390221e-01 1.44832641e-01 -4.31814969e-01 6.79054797e-01
-1.42345903e-02 6.37215436e-01 2.67974824e-01 2.60715485e-01
-1.10739434e+00 1.82364568e-01 7.10599482e-01 -1.40027761e-01
-2.07513496e-01 -2.81523943e-01 -1.16411948e+00 8.24212372e-01
1.03315687e+00 -1.10934138e+00 2.99827963e-01 9.70162302e-02
8.07873547e-01 6.05798736e-02 1.40484959e-01 -4.03691500e-01
6.35010153e-02 -5.03277592e-02 4.34887052e-01 -8.50278556e-01
6.19079638e-03 -1.75911821e-02 -2.45466888e-01 5.63870370e-01
-8.06778193e-01 1.38833150e-01 7.43414685e-02 6.74157739e-01
3.32754999e-01 -5.90218306e-01 7.81090975e-01 3.85652180e-03
-8.41015056e-02 6.23406917e-02 -9.85181391e-01 3.87197584e-01
6.15030766e-01 2.05617309e-01 -8.11767995e-01 -4.80125427e-01
-6.26751125e-01 1.71370342e-01 -4.60258909e-02 3.83754998e-01
4.34135348e-01 -9.32944298e-01 -9.89998817e-01 -5.59015930e-01
-2.72830248e-01 -9.44882214e-01 5.52619338e-01 8.68554294e-01
-3.49964678e-01 -1.06244378e-01 -2.01181278e-01 3.62563908e-01
-1.50542080e+00 6.59730136e-01 9.59776491e-02 -9.24303889e-01
-5.09258032e-01 1.02362955e+00 -1.63948953e-01 -1.99624181e-01
-9.60033610e-02 2.42530450e-01 -5.21737933e-01 5.04704773e-01
9.84507322e-01 1.07707512e+00 -1.80513412e-01 -6.68814898e-01
-5.51459193e-01 -4.48622286e-01 -5.17880082e-01 -1.68660134e-01
1.18551886e+00 -1.85092092e-01 -3.83581638e-01 4.41051751e-01
1.56522083e+00 3.26187760e-01 -4.85522598e-01 -6.11130781e-02
4.86455828e-01 9.47083533e-02 -1.57330364e-01 -1.11068141e+00
-5.26890695e-01 7.70229936e-01 -1.93666995e-01 5.31252384e-01
6.71099126e-01 -2.59821326e-01 1.54646075e+00 1.02969930e-01
2.15411372e-02 -9.46486294e-01 1.15818284e-01 8.22857201e-01
7.72728205e-01 -5.90777040e-01 -3.53690907e-02 -3.59604180e-01
-7.25176036e-01 1.07670522e+00 5.18712580e-01 -3.27473760e-01
4.60002571e-01 5.75415909e-01 -8.35766550e-04 -3.96964371e-01
-9.11624253e-01 3.86564374e-01 4.22115847e-02 3.30730736e-01
4.02988553e-01 -4.67870496e-02 -9.20220971e-01 1.23057806e+00
-4.14346188e-01 -3.36150885e-01 7.86716104e-01 5.06457508e-01
-8.00339878e-01 -6.18558705e-01 -3.41746479e-01 6.43105507e-01
-1.41483057e+00 -3.86839569e-01 -1.12758505e+00 5.57267606e-01
3.29092741e-01 1.30362046e+00 -1.54071897e-01 -8.27660441e-01
-1.24731250e-01 5.91048896e-01 -1.50536627e-01 -4.63361502e-01
-1.59157848e+00 1.88159779e-01 7.82900274e-01 -2.26754904e-01
2.36654487e-02 -4.94335592e-01 -1.23626244e+00 -7.59816408e-01
-4.82848018e-01 5.11820257e-01 2.95750827e-01 9.44243968e-01
2.49896631e-01 7.73595273e-02 6.86770797e-01 -3.36632431e-01
-8.87322545e-01 -1.64668608e+00 -8.19828928e-01 3.58665526e-01
1.05643407e-01 -8.63770023e-02 -6.75414085e-01 -9.57795382e-02] | [8.903630256652832, 10.617754936218262] |
ae631cc2-2dc3-429a-9bc8-a4117c14e144 | encoder-decoder-architecture-for-3d-seismic | 2207.14789 | null | https://arxiv.org/abs/2207.14789v1 | https://arxiv.org/pdf/2207.14789v1.pdf | Encoder-Decoder Architecture for 3D Seismic Inversion | Inverting seismic data to build 3D geological structures is a challenging task due to the overwhelming amount of acquired seismic data, and the very-high computational load due to iterative numerical solutions of the wave equation, as required by industry-standard tools such as Full Waveform Inversion (FWI). For example, in an area with surface dimensions of 4.5km $\times$ 4.5km, hundreds of seismic shot-gather cubes are required for 3D model reconstruction, leading to Terabytes of recorded data. This paper presents a deep learning solution for the reconstruction of realistic 3D models in the presence of field noise recorded in seismic surveys. We implement and analyze a convolutional encoder-decoder architecture that efficiently processes the entire collection of hundreds of seismic shot-gather cubes. The proposed solution demonstrates that realistic 3D models can be reconstructed with a structural similarity index measure (SSIM) of 0.8554 (out of 1.0) in the presence of field noise at 10dB signal-to-noise ratio. | ['Mauricio Araya-Polo', 'Yen Sun', 'Amir Adler', 'Maayan Gelboim'] | 2022-07-29 | null | null | null | null | ['seismic-inversion'] | ['miscellaneous'] | [ 9.01046544e-02 6.73482791e-02 9.66243982e-01 -2.20678434e-01
-1.25253701e+00 -3.33669364e-01 5.28430194e-02 1.54413432e-01
-4.77374375e-01 3.00761044e-01 3.25577885e-01 -4.42457020e-01
-3.60891461e-01 -1.10346770e+00 -8.91701996e-01 -5.00673592e-01
-8.00952196e-01 5.29352903e-01 2.56010324e-01 -3.88978750e-01
2.54434079e-01 8.48987997e-01 -1.53044915e+00 2.65747577e-01
3.75927806e-01 1.29212689e+00 2.87361681e-01 9.04072821e-01
-1.87814720e-02 8.42748344e-01 -5.11144042e-01 -2.34392911e-01
2.44467705e-01 -6.95191994e-02 -7.12240219e-01 -1.45260870e-01
3.33290786e-01 -8.80952954e-01 -6.37802243e-01 8.01734388e-01
1.01624656e+00 -1.15896627e-01 4.82858241e-01 -4.96263444e-01
1.60707697e-01 5.19967735e-01 -8.16454589e-01 2.60763079e-01
2.47500896e-01 -6.34020716e-02 6.07528150e-01 -1.15591931e+00
1.14366360e-01 9.34263229e-01 1.12938964e+00 -2.32963562e-01
-1.07630277e+00 -5.23223162e-01 -8.54404092e-01 1.33107767e-01
-1.52773654e+00 -7.09542811e-01 7.23956168e-01 -5.19349456e-01
1.18538034e+00 8.23675692e-02 6.26620531e-01 5.99568665e-01
1.67350397e-01 2.65620172e-01 4.08049792e-01 -1.10967264e-01
4.30384845e-01 -8.55509043e-01 -2.58009404e-01 4.57729220e-01
5.03579915e-01 -1.16536446e-01 -8.61956358e-01 -9.28880051e-02
8.08996081e-01 -3.72476369e-01 -1.13293573e-01 4.96087670e-01
-9.28196847e-01 5.71604967e-01 2.11180210e-01 -2.18706410e-02
-6.51399016e-01 7.25344419e-01 5.32412052e-01 3.39277118e-01
2.86022782e-01 3.20626080e-01 3.01301163e-02 -4.12721664e-01
-1.33893967e+00 5.77900529e-01 6.86521530e-01 8.29018652e-01
7.28926241e-01 7.95948923e-01 7.78699160e-01 4.42209572e-01
3.81480396e-01 1.05238354e+00 1.67564690e-01 -1.19301760e+00
6.41110778e-01 1.63630337e-01 1.64367020e-01 -1.29004776e+00
-5.70110202e-01 -5.41975319e-01 -1.32446849e+00 2.29395851e-01
1.30178779e-01 -2.70882905e-01 -4.10537124e-01 1.11548042e+00
4.77148257e-02 4.09402251e-01 1.92902714e-01 6.63005888e-01
8.68971467e-01 6.41513526e-01 -5.33444762e-01 1.57541424e-01
9.97189760e-01 8.56452361e-02 -3.92830014e-01 -6.91872716e-01
7.66085088e-01 -6.87425315e-01 8.07231545e-01 3.91293824e-01
-1.31065047e+00 -2.85584152e-01 -1.26362932e+00 1.70781389e-01
4.31826741e-01 -3.09932560e-01 3.60898614e-01 3.08056325e-01
-1.06916058e+00 6.49576604e-01 -1.06321788e+00 4.43988830e-01
5.00857949e-01 1.41255483e-01 -4.50879276e-01 -1.55192778e-01
-1.16937006e+00 5.37285507e-01 -1.78727973e-02 5.69442391e-01
-1.26593518e+00 -9.13955152e-01 -7.71940947e-01 4.34608817e-01
-6.96960017e-02 -2.89272040e-01 1.13576794e+00 1.08662121e-01
-9.17563498e-01 5.37612796e-01 2.99368978e-01 -5.47755301e-01
4.48336750e-01 -3.02816451e-01 -2.04366833e-01 4.45814818e-01
2.49402791e-01 -1.58035114e-01 3.76579344e-01 -1.05807579e+00
-2.65093118e-01 -3.95648152e-01 -3.39424729e-01 -1.00967616e-01
-1.74691211e-02 -1.62188672e-02 -1.74386486e-01 -3.93625230e-01
6.94792688e-01 -5.06743193e-01 -5.64999163e-01 -1.59289852e-01
-1.59626871e-01 8.64753187e-01 4.73742634e-01 -1.08180749e+00
9.73864853e-01 -2.17233014e+00 -1.15049288e-01 3.34455699e-01
1.71175256e-01 -1.08496910e-02 1.05518885e-02 7.99172759e-01
4.17366549e-02 -4.16268706e-02 -8.46370995e-01 -2.59691358e-01
-2.53347039e-01 3.42736170e-02 -3.71976793e-01 6.12406194e-01
-4.39116694e-02 4.35874790e-01 -5.96260190e-01 -1.71600193e-01
-7.64312074e-02 4.27982837e-01 -4.71964180e-01 1.38755292e-01
2.86186129e-01 2.69667208e-01 -5.18817484e-01 5.39134562e-01
1.16665637e+00 -1.56408757e-01 -5.49517237e-02 -2.21409291e-01
-3.84605348e-01 2.98415750e-01 -1.44105029e+00 1.75477087e+00
-5.67950308e-01 7.45526910e-01 3.94442052e-01 -1.22687018e+00
1.16199362e+00 4.30926740e-01 5.16117394e-01 -8.72761846e-01
1.63897183e-02 7.61671364e-01 -2.68226247e-02 -8.48736763e-01
4.95665759e-01 -4.44607615e-01 -4.34411943e-01 6.59277856e-01
-9.93555635e-02 -7.20659137e-01 -6.35467395e-02 2.07971796e-01
1.67808795e+00 -4.13836896e-01 -2.53496647e-01 -1.77450836e-01
-4.44694655e-03 -6.82937205e-02 3.12849998e-01 6.07785761e-01
4.80496824e-01 8.39865565e-01 3.66141468e-01 -5.92960715e-01
-1.42066395e+00 -1.00755835e+00 3.75891402e-02 1.02368645e-01
-1.06870376e-01 -4.79467539e-03 -4.12986249e-01 4.64802802e-01
-1.32559508e-01 4.59710300e-01 -2.76063144e-01 -8.75640363e-02
-8.31902623e-01 -9.24908280e-01 1.35584784e+00 6.93937719e-01
6.22002184e-01 -5.95059276e-01 -1.26797819e+00 5.89147866e-01
-4.22802120e-01 -1.03171599e+00 4.08729106e-01 2.30069160e-01
-1.05588317e+00 -6.42040908e-01 -4.21140820e-01 -5.69956064e-01
2.66535968e-01 2.34614447e-01 1.10497534e+00 5.12755737e-02
-2.88730025e-01 -1.38240635e-01 -3.57802600e-01 -1.94061249e-01
-2.89889216e-01 -3.35181683e-01 2.13729460e-02 -3.81009541e-02
-1.25690438e-02 -1.21465635e+00 -4.26770419e-01 4.50201668e-02
-1.07707739e+00 -8.68347101e-03 5.62974334e-01 6.34385645e-01
3.52971375e-01 3.46409798e-01 4.89472061e-01 -7.26019666e-02
3.51541072e-01 -5.24844587e-01 -7.86026597e-01 -2.22042188e-01
4.17213917e-01 -9.44554619e-03 4.48557794e-01 -5.34081981e-02
-9.65052962e-01 -6.38584420e-02 -8.06507409e-01 -1.46645293e-01
9.45265368e-02 1.23976398e+00 -1.98857319e-02 -2.04918861e-01
8.39581370e-01 2.51121521e-01 -7.69244432e-02 -6.71046913e-01
-1.94642603e-01 8.23739886e-01 8.07649732e-01 -2.12704569e-01
8.31713378e-01 8.42420459e-01 4.42650944e-01 -1.35671294e+00
-5.70861220e-01 -2.35036716e-01 -5.43815255e-01 -5.04298389e-01
4.20567155e-01 -1.14699876e+00 -8.54822695e-01 6.83181345e-01
-1.25292647e+00 -3.60180497e-01 -1.05820417e-01 8.41781199e-01
-4.68594551e-01 5.38191855e-01 -6.62623167e-01 -8.56051147e-01
-5.50971150e-01 -9.91568208e-01 1.17804635e+00 -2.81798273e-01
-1.79429695e-01 -5.16398787e-01 -1.93298738e-02 2.39715427e-01
5.24325252e-01 7.08457589e-01 7.05800712e-01 4.93233986e-02
-6.07711256e-01 -6.25231683e-01 -1.52125731e-01 6.51801378e-02
-1.61479846e-01 -4.24013287e-01 -1.02243674e+00 -1.69634685e-01
5.67890942e-01 -3.24065536e-01 6.98614955e-01 6.21065438e-01
8.28684688e-01 -7.61688799e-02 1.85693115e-01 8.08404922e-01
1.57953751e+00 1.37278467e-01 9.27150607e-01 2.29774982e-01
4.41659987e-01 5.99677682e-01 2.29076296e-01 1.14323604e+00
8.78657401e-02 1.69287458e-01 7.92871058e-01 9.03393105e-02
6.79964125e-02 1.01477973e-01 -1.51495159e-01 1.22941387e+00
-2.30871618e-01 -3.23501319e-01 -1.55520833e+00 8.56096208e-01
-1.20573795e+00 -1.15203369e+00 -5.26042700e-01 2.10753703e+00
4.27909106e-01 2.42248431e-01 -6.35494769e-01 7.69666255e-01
2.34956816e-01 2.25109100e-01 -4.49670106e-01 -6.33686036e-02
-3.29582810e-01 4.12928730e-01 8.18723679e-01 5.40998101e-01
-7.11888313e-01 3.08928847e-01 5.79087019e+00 4.63600367e-01
-1.16386771e+00 -1.80356592e-01 2.88423419e-01 -7.21923113e-02
-4.61508960e-01 -2.73590982e-01 -3.76135290e-01 3.01972389e-01
1.32189369e+00 -5.86641990e-02 1.24521613e-01 4.22103435e-01
4.59370971e-01 -3.45918149e-01 -6.61587119e-01 1.12668848e+00
-5.91848679e-02 -1.81348014e+00 -1.36997283e-01 1.66295692e-01
3.62571657e-01 3.35451871e-01 -1.71656430e-01 -3.50117385e-01
1.64940685e-01 -1.12166238e+00 9.93871093e-01 6.02160335e-01
1.01972032e+00 -9.23566163e-01 9.70424294e-01 5.18767118e-01
-1.25518930e+00 -2.32988503e-02 -4.74927545e-01 -2.40827858e-01
9.63244557e-01 1.04958999e+00 -6.73535407e-01 7.03456521e-01
1.14140081e+00 4.07520950e-01 5.62547473e-03 1.21370208e+00
2.24780500e-01 9.42613840e-01 -8.24619532e-01 3.96582752e-01
4.53560680e-01 2.35893801e-02 4.34821337e-01 9.84952629e-01
1.19570661e+00 4.04412419e-01 -4.12327945e-01 4.13765252e-01
-1.09246306e-01 -3.59474540e-01 -6.10294640e-01 3.04418534e-01
5.68903327e-01 5.45402706e-01 -3.89314294e-01 -1.05420403e-01
-3.20004433e-01 4.65691686e-01 -1.42867029e-01 -1.72846597e-02
-6.20633185e-01 -7.36388266e-01 3.95104915e-01 5.78749597e-01
4.11207378e-01 -6.79415643e-01 -6.47110879e-01 -4.51767862e-01
2.17976704e-01 -6.59144759e-01 -1.96037162e-02 -8.70544672e-01
-9.06935930e-01 5.21539211e-01 -6.64621815e-02 -1.54557776e+00
-1.95564494e-01 -3.16828519e-01 -5.78239381e-01 7.88136721e-01
-1.08590102e+00 -8.28945279e-01 -6.24341607e-01 2.03041404e-01
1.60398513e-01 -5.12132198e-02 7.47174263e-01 9.90882993e-01
1.04221702e-01 4.60859463e-02 4.10812736e-01 3.17452133e-01
2.99783479e-02 -5.80931962e-01 1.13208592e+00 8.64398360e-01
-4.09543782e-01 -5.77092133e-02 1.01519918e+00 -6.59501672e-01
-1.65825701e+00 -1.00669849e+00 8.54682028e-01 1.75789654e-01
7.30066538e-01 -2.34900609e-01 -1.39304733e+00 4.29921746e-01
-2.26741105e-01 2.08487045e-02 7.72439003e-01 -5.89701891e-01
-1.68759212e-01 -1.63696378e-01 -9.81761575e-01 4.11881506e-01
9.48974848e-01 -4.43609685e-01 -4.58697617e-01 2.65146662e-02
5.01506686e-01 -4.11469012e-01 -1.00275564e+00 5.85683525e-01
5.56438208e-01 -1.03525281e+00 1.20080054e+00 -2.37947069e-02
6.67212427e-01 -3.27231765e-01 -6.87179983e-01 -1.04135597e+00
-1.43507704e-01 -5.63833356e-01 1.44400388e-01 5.05078912e-01
3.49416286e-01 -2.07772180e-01 8.38648915e-01 2.74735630e-01
-7.07758665e-01 -3.91467065e-01 -1.50685179e+00 -6.89943314e-01
4.06811014e-02 -9.71715987e-01 7.66789794e-01 5.80856919e-01
-3.57182950e-01 5.19177280e-02 -4.55795079e-01 7.55831182e-01
9.91365731e-01 -1.11531019e-01 5.72203159e-01 -1.21220446e+00
-3.02245080e-01 -2.15139873e-02 -6.77863657e-01 -1.13212335e+00
-4.37889159e-01 -4.14113194e-01 1.94481373e-01 -1.68502283e+00
-4.55309272e-01 -5.50013542e-01 2.18734041e-01 9.63813365e-02
5.01307368e-01 4.57583725e-01 -2.70481497e-01 9.10059735e-03
2.10025385e-01 6.18760347e-01 7.49529958e-01 -2.58194804e-01
1.57478958e-01 -1.30104005e-01 -2.47722149e-01 7.65290618e-01
6.42653823e-01 -6.82335913e-01 -1.99732229e-01 -1.46122098e+00
8.51443291e-01 7.00394869e-01 4.66354549e-01 -1.42912030e+00
3.30956459e-01 1.90337554e-01 1.73289195e-01 -1.20005369e+00
6.80032849e-01 -5.27674377e-01 6.87745810e-01 6.87957644e-01
2.27434114e-02 1.29443005e-01 3.40428948e-01 3.91719013e-01
-6.40724599e-01 -5.14672935e-01 7.79199183e-01 -2.77093232e-01
-5.65649271e-01 6.68412745e-02 -6.77482605e-01 1.60472885e-01
4.57480788e-01 -3.10822129e-01 1.32800013e-01 -4.47537631e-01
-3.96394730e-01 -5.93447909e-02 -9.61994752e-03 -3.91274035e-01
1.13387406e+00 -1.07625246e+00 -1.31651163e+00 3.30623180e-01
-1.47786215e-01 5.39777637e-01 9.50969398e-01 6.16596639e-01
-1.55799532e+00 -1.00274034e-01 -1.23355508e-01 -6.24381363e-01
-7.95920908e-01 -2.95118779e-01 4.29231286e-01 1.88267864e-02
-9.50358629e-01 1.28028548e+00 -1.60595506e-01 -1.70843974e-01
-1.46284088e-01 -2.58668065e-01 1.90610111e-01 1.32727891e-01
8.60808074e-01 6.83165431e-01 5.67864954e-01 -7.23404229e-01
-2.14919478e-01 6.66659772e-01 5.64304829e-01 -4.57663327e-01
1.95742917e+00 1.18396029e-01 -7.66451145e-03 9.95072573e-02
1.25332141e+00 -1.69817194e-01 -1.03435588e+00 -2.89043695e-01
4.98199789e-03 -5.81643045e-01 3.04654092e-01 -9.83793288e-02
-9.94838834e-01 1.36607277e+00 4.18055892e-01 -9.23200324e-03
1.03739583e+00 -6.70126602e-02 1.00007558e+00 7.99708128e-01
4.98162776e-01 -8.96024704e-01 3.19370031e-02 7.90067255e-01
1.05723059e+00 -6.67911172e-01 1.29370213e-01 9.56989080e-02
-3.17606479e-01 1.00097132e+00 -1.04485698e-01 -2.61590570e-01
9.12014544e-01 1.08770812e+00 -8.62144604e-02 -4.11883503e-01
-4.16861743e-01 2.93897837e-01 -5.64046621e-01 4.55483794e-01
1.46570697e-01 -1.96642935e-01 4.05740231e-01 4.36498135e-01
-4.86879319e-01 -2.14888498e-01 8.10522079e-01 1.25128317e+00
-7.73689687e-01 -3.03352654e-01 -6.25096083e-01 5.78668952e-01
-3.54085922e-01 -1.95886791e-01 3.58091503e-01 2.92897820e-01
-3.43809545e-01 8.62499475e-01 4.68949050e-01 -4.01926249e-01
4.88364816e-01 -4.29985046e-01 1.90937206e-01 -3.52721572e-01
-2.48104975e-01 2.23782375e-01 3.72980624e-01 -2.81587720e-01
-2.33514532e-01 -6.07848227e-01 -1.55985343e+00 -4.69961852e-01
2.23131366e-02 1.34075612e-01 7.42277682e-01 7.31510282e-01
1.79127365e-01 5.05064964e-01 5.29998243e-01 -1.28453207e+00
-3.66301537e-01 -1.00811911e+00 -8.96800578e-01 -2.75824615e-03
4.49032933e-01 -5.00283360e-01 -5.10917127e-01 2.16814250e-01] | [6.87230920791626, 2.506937265396118] |
427e2620-1503-46d3-aee5-8d18180fc1d1 | a-mountain-shaped-single-stage-network-for | 2305.05146 | null | https://arxiv.org/abs/2305.05146v1 | https://arxiv.org/pdf/2305.05146v1.pdf | A Mountain-Shaped Single-Stage Network for Accurate Image Restoration | Image restoration is the task of aiming to obtain a high-quality image from a corrupt input image, such as deblurring and deraining. In image restoration, it is typically necessary to maintain a complex balance between spatial details and contextual information. Although a multi-stage network can optimally balance these competing goals and achieve significant performance, this also increases the system's complexity. In this paper, we propose a mountain-shaped single-stage design base on a simple U-Net architecture, which removes or replaces unnecessary nonlinear activation functions to achieve the above balance with low system complexity. Specifically, we propose a feature fusion middleware (FFM) mechanism as an information exchange component between the encoder-decoder architectural levels. It seamlessly integrates upper-layer information into the adjacent lower layer, sequentially down to the lowest layer. Finally, all information is fused into the original image resolution manipulation level. This preserves spatial details and integrates contextual information, ensuring high-quality image restoration. In addition, we propose a multi-head attention middle block (MHAMB) as a bridge between the encoder and decoder to capture more global information and surpass the limitations of the receptive field of CNNs. Extensive experiments demonstrate that our approach, named as M3SNet, outperforms previous state-of-the-art models while using less than half the computational costs, for several image restoration tasks, such as image deraining and deblurring. | ['Depeng Dang', 'Jingfan Yang', 'Ning Wang', 'Ying Zhang', 'Jing Yang', 'Hu Gao'] | 2023-05-09 | null | null | null | null | ['deblurring', 'single-image-deraining'] | ['computer-vision', 'computer-vision'] | [ 4.65018779e-01 -1.17313847e-01 2.63329782e-02 -9.72320959e-02
-5.77239513e-01 -1.15234070e-01 3.12685698e-01 -2.81090975e-01
-3.48111898e-01 5.87628663e-01 3.77343029e-01 -1.92621037e-01
1.33557156e-01 -7.39841104e-01 -8.49688649e-01 -8.97490919e-01
5.42775333e-01 -6.49976194e-01 4.24389809e-01 -2.33212799e-01
2.83301383e-01 5.27111888e-01 -1.41277063e+00 4.08703417e-01
1.21766114e+00 1.15358484e+00 5.66991508e-01 4.75979090e-01
8.97649378e-02 9.39302981e-01 -4.60225761e-01 -2.82268792e-01
7.15210959e-02 -3.67941737e-01 -5.40363967e-01 2.31086224e-01
6.47287309e-01 -7.32200503e-01 -6.90153778e-01 1.32610559e+00
5.40352821e-01 -2.41678371e-03 2.42936715e-01 -7.34633863e-01
-1.29313517e+00 3.33027542e-01 -7.73619533e-01 2.72368819e-01
-7.14507699e-02 3.28803301e-01 4.69832480e-01 -9.65100765e-01
2.82737613e-01 1.17140090e+00 5.22083938e-01 5.64860046e-01
-1.39224553e+00 -7.04568207e-01 2.07089409e-01 4.33251500e-01
-1.30729556e+00 -7.52647519e-01 7.51275420e-01 -2.54599720e-01
9.39294636e-01 1.79904804e-01 3.56526107e-01 6.91179931e-01
6.01341426e-01 6.72662735e-01 8.04224193e-01 -1.92309693e-01
-1.70861650e-02 -2.05344945e-01 1.00112800e-02 3.95521373e-01
3.06003451e-01 6.93272352e-02 -5.33408105e-01 3.92946243e-01
1.19798923e+00 2.28360087e-01 -9.23243940e-01 -5.67548685e-02
-9.98827457e-01 2.91011095e-01 9.91021514e-01 5.00938892e-01
-5.19348264e-01 1.75351754e-01 1.86775550e-02 2.65360862e-01
4.15019989e-01 1.74306363e-01 -1.59713820e-01 4.35158193e-01
-1.12173307e+00 -1.20791793e-01 4.53110188e-02 7.50283837e-01
8.31305921e-01 2.97811866e-01 -3.16179812e-01 8.62147570e-01
2.12030143e-01 1.53296039e-01 5.77718854e-01 -9.09507573e-01
4.65754628e-01 6.60942852e-01 1.29440308e-01 -1.11486721e+00
-7.09379166e-02 -8.01639676e-01 -1.55884266e+00 4.34601575e-01
-6.93747550e-02 2.28448436e-01 -1.22357547e+00 1.69583631e+00
4.99215759e-02 2.11876288e-01 8.93214121e-02 1.21220696e+00
7.95661390e-01 7.38207221e-01 -2.93814410e-02 -2.08311319e-01
1.42342722e+00 -1.13245964e+00 -1.00469422e+00 -5.30528188e-01
-5.81201203e-02 -8.86880338e-01 9.52950776e-01 2.19501987e-01
-1.45214617e+00 -1.01532555e+00 -1.40429473e+00 -5.90268910e-01
-3.63202728e-02 1.88799366e-01 2.45268166e-01 3.50778997e-01
-1.36202037e+00 7.08286464e-01 -7.00388551e-01 1.24290138e-01
5.68775952e-01 3.10416311e-01 -5.79595566e-01 -3.93827826e-01
-1.03395367e+00 8.62679780e-01 2.60701716e-01 5.14697134e-01
-7.34860241e-01 -6.14212692e-01 -1.02536356e+00 3.89471263e-01
2.30355442e-01 -8.63370001e-01 9.43856418e-01 -8.72558057e-01
-1.51263618e+00 4.16017830e-01 -2.26309866e-01 -2.67395586e-01
3.44545811e-01 -2.75308102e-01 -4.76753533e-01 1.27986953e-01
-9.33176950e-02 6.83589935e-01 1.13153565e+00 -1.36767912e+00
-5.41148782e-01 -3.09757411e-01 3.35400179e-02 1.42465472e-01
-3.53277832e-01 -1.03817515e-01 -7.82656372e-01 -1.00756419e+00
5.41253448e-01 -4.33084309e-01 -1.14946581e-01 1.08327702e-01
-3.44086230e-01 3.64599258e-01 9.18274760e-01 -1.01223195e+00
1.41383505e+00 -2.37928820e+00 3.73900950e-01 -2.52995372e-01
4.08584863e-01 4.97551203e-01 -2.41611868e-01 2.86710374e-02
-2.92624772e-01 6.12313151e-02 -5.11516631e-01 -5.47128618e-01
-4.54122037e-01 7.14949444e-02 -3.09245437e-01 5.96319914e-01
3.56465101e-01 1.12968194e+00 -5.72124124e-01 -1.94099948e-01
4.16256219e-01 8.54084253e-01 -5.89176714e-01 1.57443374e-01
1.81530386e-01 5.36196709e-01 4.82288711e-02 5.66798985e-01
1.08879638e+00 -3.88731986e-01 5.58744371e-02 -6.38637125e-01
-2.79086590e-01 7.18386918e-02 -1.10138261e+00 1.72118413e+00
-5.62667131e-01 6.95559025e-01 5.41157126e-01 -8.74336064e-01
7.77804613e-01 2.36889794e-01 1.09502561e-01 -1.02342844e+00
1.76283702e-01 2.43922457e-01 -2.24891026e-02 -3.18512082e-01
5.98011672e-01 -9.87079516e-02 1.85285583e-01 -3.00941467e-02
1.29161347e-02 2.58876503e-01 -1.44680768e-01 -2.56318059e-02
8.58376265e-01 8.86561573e-02 3.18751037e-01 -1.91421136e-01
8.24283719e-01 -4.63851929e-01 7.41889000e-01 4.85888153e-01
-1.49072021e-01 1.11954081e+00 2.80916065e-01 -3.85214329e-01
-1.01552022e+00 -9.72161233e-01 -7.54129514e-02 5.67418694e-01
5.76440930e-01 -2.59662837e-01 -8.92009616e-01 -2.71356970e-01
-3.18193316e-01 3.99557263e-01 -4.35506135e-01 -4.67064202e-01
-7.92808533e-01 -7.27262795e-01 2.33748749e-01 3.92857581e-01
1.07507348e+00 -1.05628300e+00 -4.08253133e-01 3.62473547e-01
-5.28100431e-01 -1.04972231e+00 -8.79124701e-01 9.66015607e-02
-7.61829615e-01 -8.62711012e-01 -8.52288485e-01 -9.14569557e-01
8.59575212e-01 7.25171387e-01 8.77970278e-01 2.73309946e-01
-1.07243463e-01 -5.37630796e-01 -1.99048892e-01 3.58717829e-01
-2.59223431e-01 -1.36408195e-01 -2.14626536e-01 3.29518884e-01
-3.21070850e-01 -9.19524848e-01 -1.03910792e+00 3.64820808e-01
-1.50379646e+00 3.83082092e-01 1.01289701e+00 1.00681674e+00
5.35903871e-01 3.60388339e-01 4.21070039e-01 -3.35629851e-01
4.33256328e-01 -1.98135212e-01 -3.99593890e-01 1.24368347e-01
-4.24753517e-01 -3.37383226e-02 8.51879478e-01 -4.88147348e-01
-1.16057420e+00 -1.05425060e-01 -1.42051488e-01 -5.62655985e-01
9.63639170e-02 2.54905134e-01 -6.97276831e-01 -2.93968171e-01
3.71442616e-01 6.70012712e-01 5.11958897e-02 -7.35683322e-01
3.58404309e-01 7.43754923e-01 9.99751449e-01 1.26584889e-02
7.18863010e-01 4.36426401e-01 -2.43340403e-01 -5.44772029e-01
-5.82146347e-01 -1.77118573e-02 -5.65728068e-01 -2.26410836e-01
7.45659232e-01 -1.14263690e+00 -7.48364806e-01 9.16879952e-01
-1.26715410e+00 -1.51839569e-01 -9.38521549e-02 2.40473181e-01
-2.79142022e-01 4.80751634e-01 -8.09437394e-01 -3.14993352e-01
-4.12588418e-01 -1.40736580e+00 9.22506452e-01 5.24169922e-01
2.82278240e-01 -4.13735747e-01 -5.19515932e-01 4.90888864e-01
8.15113783e-01 -4.01978791e-02 9.74918783e-01 2.66576886e-01
-9.43913043e-01 7.72169009e-02 -6.70298934e-01 7.72701144e-01
2.62098789e-01 -5.33908844e-01 -9.30218577e-01 -5.48935175e-01
2.74073392e-01 1.52910026e-02 1.29987276e+00 4.97393340e-01
1.19702756e+00 -4.88789290e-01 -7.38395303e-02 8.46604526e-01
1.47209477e+00 2.15418458e-01 1.33842647e+00 3.73184383e-01
8.28199387e-01 3.53256941e-01 2.04081237e-01 -5.32125570e-02
4.10358429e-01 7.04506636e-01 5.15809000e-01 -5.18234730e-01
-7.69859791e-01 -2.27393970e-01 4.53456938e-01 8.23215127e-01
2.29967162e-01 -3.24993819e-01 -5.70596576e-01 5.29780269e-01
-1.95455098e+00 -8.85140598e-01 9.21620503e-02 2.09435201e+00
1.03477669e+00 4.39874008e-02 -5.60985744e-01 2.01903775e-01
8.10758889e-01 4.08802271e-01 -6.70918524e-01 1.86516941e-02
-2.54163861e-01 -7.96288028e-02 5.14736652e-01 7.75361657e-01
-1.00542343e+00 9.02500451e-01 5.96811390e+00 1.03237844e+00
-1.19375789e+00 1.34779587e-01 9.78132069e-01 -6.04259782e-02
-1.32136777e-01 -1.53555915e-01 -4.88058865e-01 6.08110309e-01
5.45144379e-01 2.60087758e-01 5.63024104e-01 3.97353858e-01
3.31207633e-01 -1.96345434e-01 -7.62091815e-01 1.10851109e+00
1.38461083e-01 -1.41578162e+00 1.64952531e-01 1.21472195e-01
7.08149552e-01 -2.71905035e-01 1.10177189e-01 -5.25104702e-02
-7.92730376e-02 -1.09605980e+00 8.97765875e-01 6.18888974e-01
8.57118309e-01 -8.30625117e-01 6.67475939e-01 3.31892014e-01
-1.19488752e+00 -2.21055061e-01 -3.05050731e-01 9.81571525e-02
3.33016217e-01 8.78667176e-01 1.22186877e-01 7.66190529e-01
8.62600803e-01 7.20496476e-01 -3.35105717e-01 1.00799012e+00
-2.94279993e-01 7.67254978e-02 1.39773712e-01 7.84596682e-01
1.02477834e-01 -1.60547301e-01 4.55650896e-01 9.34115112e-01
3.21135938e-01 1.63231105e-01 -1.32656246e-01 9.72077131e-01
-2.06454515e-01 -3.01415056e-01 -2.52930492e-01 5.10885537e-01
2.95092314e-01 1.21147561e+00 -4.66846287e-01 -3.86207342e-01
-4.63034987e-01 1.40090001e+00 2.71926075e-01 4.45698947e-01
-8.04528773e-01 -4.61323202e-01 9.63838279e-01 2.27466524e-01
4.55602318e-01 -1.80625498e-01 -4.15134132e-01 -1.32137084e+00
4.36877519e-01 -9.57082987e-01 -1.66561902e-01 -8.61053824e-01
-8.91629875e-01 8.72006238e-01 -5.34638524e-01 -1.35211468e+00
1.87716216e-01 -3.24756175e-01 -3.88282150e-01 1.21605933e+00
-2.13467598e+00 -1.13935828e+00 -5.42091966e-01 5.94023943e-01
6.03023648e-01 2.02984318e-01 2.92155921e-01 7.61956453e-01
-8.46265018e-01 4.26638395e-01 -2.01724749e-02 4.58195694e-02
6.70819640e-01 -8.07907164e-01 5.46630442e-01 1.40398979e+00
-3.45397741e-01 7.26647615e-01 4.28559184e-01 -7.01662362e-01
-1.21755004e+00 -1.25473225e+00 7.50694633e-01 2.08099678e-01
1.67434290e-01 -2.12360442e-01 -1.33319092e+00 3.04887563e-01
2.99670428e-01 1.61274478e-01 1.71478063e-01 -5.71455300e-01
-4.37953949e-01 -3.18115920e-01 -1.09818232e+00 7.39992738e-01
9.53421593e-01 -5.32609642e-01 -4.08413529e-01 -2.62567341e-01
8.24119508e-01 -4.60904568e-01 -7.17250824e-01 4.81972247e-01
3.95985216e-01 -1.21537066e+00 1.19415724e+00 2.45662425e-02
7.53203690e-01 -8.18321049e-01 -1.94934726e-01 -1.31880307e+00
-8.23704183e-01 -7.21977592e-01 -1.64440334e-01 1.22986937e+00
1.44830614e-01 -4.91004378e-01 4.50290143e-01 3.92820776e-01
-3.32511365e-01 -6.94140732e-01 -1.00688457e+00 -5.59265971e-01
-1.44300550e-01 -6.48876652e-02 5.64269662e-01 6.99176729e-01
-3.53580415e-01 3.32927227e-01 -7.10131526e-01 4.34390098e-01
6.90502763e-01 -7.56585076e-02 3.12521011e-01 -7.75256634e-01
-9.81750712e-02 -5.13380527e-01 -2.85466433e-01 -1.49302411e+00
-1.54627085e-01 -5.81967473e-01 1.61484659e-01 -1.68206620e+00
3.19240957e-01 -4.99915145e-02 -3.58657509e-01 5.86021364e-01
-3.71885836e-01 7.30365276e-01 2.24963233e-01 5.38121104e-01
-2.34826326e-01 8.50942254e-01 1.66190362e+00 -1.17379792e-01
-2.12594837e-01 -4.32669729e-01 -1.16272438e+00 5.93848109e-01
5.71588099e-01 -2.44714946e-01 -3.49347323e-01 -8.42448354e-01
-1.97775811e-01 1.44187078e-01 5.83065987e-01 -1.10694885e+00
5.44085622e-01 3.50074247e-02 6.91038549e-01 -4.74680871e-01
2.94193625e-01 -9.40300524e-01 1.78568214e-01 4.55828696e-01
8.71610455e-03 -6.71121776e-02 2.12342754e-01 4.66203570e-01
-5.71710885e-01 2.07797915e-01 1.33609366e+00 -1.03953881e-02
-6.84726477e-01 2.85675377e-01 -2.35170141e-01 -5.61012030e-01
8.69681776e-01 -3.91138315e-01 -5.30368447e-01 -2.57487953e-01
-5.30847371e-01 6.64046556e-02 6.45036757e-01 5.57130337e-01
9.22171533e-01 -1.32487583e+00 -6.57822788e-01 5.65375745e-01
-2.77028292e-01 1.40061826e-01 8.55042636e-01 8.94126713e-01
-4.81672764e-01 3.42718095e-01 -4.37221080e-01 -3.60646158e-01
-1.04415500e+00 5.42719245e-01 5.45682073e-01 -2.46271923e-01
-9.65016901e-01 5.50694525e-01 5.80484271e-01 7.35197216e-02
1.16131291e-01 -2.70103067e-01 -2.30672032e-01 -6.63921386e-02
1.06172597e+00 1.19600765e-01 2.06879869e-01 -7.15609312e-01
-2.07697257e-01 6.17111802e-01 -2.84410506e-01 9.02988985e-02
1.36082947e+00 -6.41204000e-01 -4.26319450e-01 -2.38866523e-01
1.17260420e+00 -1.99395776e-01 -1.68573165e+00 -4.13393915e-01
-4.18369502e-01 -6.86334789e-01 6.81518793e-01 -8.32403779e-01
-1.58126867e+00 7.07636297e-01 6.29499733e-01 -1.10326791e-02
1.69298494e+00 -2.88913667e-01 9.57775056e-01 -1.98292688e-01
1.88074093e-02 -6.67232037e-01 2.04497263e-01 1.91784233e-01
1.17510200e+00 -1.03641963e+00 3.00898626e-02 -4.97183472e-01
-3.64164531e-01 8.97260487e-01 8.06722879e-01 -1.33632675e-01
4.66816723e-01 4.27059025e-01 -5.33487797e-02 1.11385085e-01
-5.97456694e-01 -1.27300397e-01 3.64231050e-01 4.40886855e-01
2.67997622e-01 -2.82941520e-01 -8.70846957e-02 6.50664032e-01
1.44700840e-01 7.67554715e-02 4.93838310e-01 7.25681007e-01
-5.39452851e-01 -9.51513827e-01 -4.79207069e-01 5.19400015e-02
-6.02120221e-01 -5.54570258e-01 -9.09836125e-03 5.51419795e-01
3.48130763e-01 1.07846022e+00 1.15824766e-01 -4.93220657e-01
3.84983838e-01 -4.79630411e-01 2.92845458e-01 -2.15817302e-01
-5.24877787e-01 2.58811384e-01 -3.66863728e-01 -8.14075828e-01
-3.51104110e-01 -9.73645151e-02 -8.09433699e-01 -5.29189706e-01
-2.96482056e-01 -3.49113226e-01 4.08203930e-01 7.44667232e-01
6.86474741e-01 9.78106141e-01 6.11519992e-01 -1.11416936e+00
-2.29096308e-01 -1.02277648e+00 -5.26564240e-01 6.07108399e-02
9.34000075e-01 -3.75016630e-01 -2.30356500e-01 2.19473660e-01] | [11.204832077026367, -2.1805310249328613] |
1bce8d43-206e-4a14-bca1-506c20ba96ee | predicting-sector-index-movement-with | null | null | https://aclanthology.org/Y15-1065 | https://aclanthology.org/Y15-1065.pdf | Predicting Sector Index Movement with Microblogging Public Mood Time Series on Social Issues | null | ['Kotaro Sakamoto', 'Jinlong Guo', 'Hideyuki Shibuki', 'Tatsunori Mori', 'Yujie Lu'] | 2015-10-01 | predicting-sector-index-movement-with-1 | https://aclanthology.org/Y15-1065 | https://aclanthology.org/Y15-1065.pdf | paclic-2015-10 | ['stock-prediction'] | ['time-series'] | [-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] | [-7.406898021697998, 3.7285566329956055] |
f1c763b0-3f69-437d-9af3-1693aca198b7 | strong-generalization-and-efficiency-in | 2007.03629 | null | https://arxiv.org/abs/2007.03629v2 | https://arxiv.org/pdf/2007.03629v2.pdf | Strong Generalization and Efficiency in Neural Programs | We study the problem of learning efficient algorithms that strongly generalize in the framework of neural program induction. By carefully designing the input / output interfaces of the neural model and through imitation, we are able to learn models that produce correct results for arbitrary input sizes, achieving strong generalization. Moreover, by using reinforcement learning, we optimize for program efficiency metrics, and discover new algorithms that surpass the teacher used in imitation. With this, our approach can learn to outperform custom-written solutions for a variety of problems, as we tested it on sorting, searching in ordered lists and the NP-complete 0/1 knapsack problem, which sets a notable milestone in the field of Neural Program Induction. As highlights, our learned model can perform sorting perfectly on any input data size we tested on, with $O(n log n)$ complexity, whilst outperforming hand-coded algorithms, including quick sort, in number of operations even for list sizes far beyond those seen during training. | ['Oriol Vinyals', 'Yujia Li', 'Felix Gimeno', 'Pushmeet Kohli'] | 2020-07-07 | null | null | null | null | ['program-induction'] | ['computer-code'] | [ 2.64125437e-01 1.39853433e-01 -4.77389246e-01 -1.36737481e-01
-5.74463904e-01 -9.63375747e-01 1.63566917e-01 5.02580345e-01
-8.04878771e-01 5.48001528e-01 -2.33840346e-01 -7.62987852e-01
-2.67874181e-01 -1.19113040e+00 -1.66143334e+00 -4.31468844e-01
-6.95861518e-01 7.14669526e-01 2.38611296e-01 -5.37032448e-02
6.59126222e-01 3.66339564e-01 -1.85754716e+00 4.38107431e-01
7.94453502e-01 6.18889570e-01 2.44950980e-01 1.22383833e+00
-2.59411335e-01 1.09158862e+00 -4.31238413e-01 -2.40232557e-01
2.12107494e-01 7.65211806e-02 -1.06843376e+00 -5.88624299e-01
5.65357208e-01 -4.01170105e-01 -4.84130949e-01 9.30774570e-01
1.25202104e-01 3.12193073e-02 1.83686212e-01 -1.10122013e+00
-6.28094971e-01 1.26941741e+00 -5.08404225e-02 1.24553159e-01
5.10235667e-01 3.75081539e-01 1.18450344e+00 -1.92818418e-01
5.85454047e-01 1.19358981e+00 7.48649240e-01 7.46007264e-01
-1.61110461e+00 -6.80068672e-01 -5.87965660e-02 1.45238580e-03
-7.28603005e-01 -9.15802941e-02 -4.49930551e-03 -4.25372273e-01
1.21537411e+00 2.72216499e-01 7.49159515e-01 3.80467653e-01
-3.41236815e-02 1.24297202e+00 1.01771605e+00 -4.12302583e-01
2.58160740e-01 -2.41724569e-02 2.19776571e-01 9.66896415e-01
2.08141461e-01 4.49488789e-01 -4.07478034e-01 -2.41945639e-01
5.03001869e-01 -8.79022200e-03 -2.34963432e-01 -5.66771746e-01
-1.16574562e+00 6.52504206e-01 6.13640428e-01 2.24004999e-01
1.42418772e-01 7.77222157e-01 5.64988315e-01 8.59817028e-01
-3.47848594e-01 8.12555969e-01 -8.08092535e-01 -3.43194246e-01
-9.77766097e-01 4.36601579e-01 1.38511443e+00 1.20553577e+00
8.87391508e-01 1.83180887e-02 1.37459666e-01 2.81289577e-01
-3.93191814e-01 5.45235693e-01 3.88773978e-01 -1.23283803e+00
3.77712816e-01 7.91685641e-01 -8.80725756e-02 -6.89684987e-01
-3.12608093e-01 -2.45211393e-01 -4.79794472e-01 2.71990150e-01
6.47118151e-01 -1.19882546e-01 -5.38023233e-01 1.77473462e+00
-4.35036756e-02 -1.59706339e-01 -1.12272136e-01 3.91899288e-01
3.46855074e-01 8.95861685e-01 -6.63749725e-02 -2.30092093e-01
8.65234733e-01 -9.52927947e-01 8.70146304e-02 -2.18466416e-01
1.30897748e+00 -1.45492435e-01 1.37159944e+00 8.14222097e-01
-1.43263626e+00 -4.90889162e-01 -1.07249689e+00 2.10048944e-01
-3.03747535e-01 -2.11824164e-01 1.30810547e+00 7.30517924e-01
-1.45166349e+00 1.18488812e+00 -8.83561552e-01 -1.15721308e-01
5.02507150e-01 9.48431313e-01 -1.93400130e-01 -2.65832156e-01
-3.90738815e-01 6.29063010e-01 7.80369163e-01 -3.45079184e-01
-1.03005683e+00 -1.01575279e+00 -4.50264663e-01 2.63088822e-01
5.01313269e-01 -6.72400713e-01 1.52861464e+00 -1.03168619e+00
-1.37850165e+00 9.36951518e-01 -8.71849433e-02 -7.80228257e-01
2.55915344e-01 -5.38761541e-02 3.50744069e-01 -1.38330966e-01
-4.28170264e-01 7.79898465e-01 3.18659097e-01 -1.13591599e+00
-8.06106925e-01 -3.24115843e-01 5.78648865e-01 -2.70111680e-01
-4.11445618e-01 4.37393188e-02 -6.68720379e-02 -1.34796783e-01
-2.28814706e-01 -1.10699117e+00 -4.40199494e-01 -6.68997830e-03
3.13452557e-02 -4.30926651e-01 3.69527519e-01 -3.80507410e-01
1.05564058e+00 -1.93881023e+00 3.96226048e-01 2.73670018e-01
3.25978816e-01 2.34885588e-01 -1.93696722e-01 4.27983224e-01
9.95805711e-02 1.95305139e-01 -2.66279310e-01 3.14712882e-01
3.85539502e-01 4.61739987e-01 -4.45719630e-01 2.19567150e-01
-1.09560236e-01 1.12536073e+00 -9.57315326e-01 -3.50349247e-01
-3.86329085e-01 -3.47003192e-01 -1.40189612e+00 3.68062884e-01
-8.53293657e-01 5.91996163e-02 -2.71956205e-01 3.86498600e-01
3.06504965e-01 -4.46826577e-01 2.17993721e-01 5.01010895e-01
-7.15502948e-02 2.47297421e-01 -1.16339624e+00 1.83112156e+00
-8.24626505e-01 6.76593602e-01 2.62275279e-01 -1.26613796e+00
6.20420814e-01 -1.47080302e-01 1.60042033e-01 -7.34472156e-01
-8.91129002e-02 3.72732908e-01 2.70852298e-01 -5.85999727e-01
2.63606727e-01 1.75475210e-01 -4.71560769e-02 8.71789694e-01
7.48013481e-02 -4.38656956e-01 4.22782063e-01 1.87889755e-01
1.69110882e+00 -4.79901358e-02 -2.73036391e-01 -2.98686534e-01
3.82413775e-01 3.19620460e-01 2.05545947e-01 1.40847933e+00
2.60149330e-01 -1.49571434e-01 1.02376807e+00 -6.65031493e-01
-1.21318352e+00 -7.54224896e-01 1.78886726e-01 1.84003472e+00
-2.09486350e-01 -3.28318626e-01 -8.61898541e-01 -5.32279789e-01
2.30390027e-01 5.98817587e-01 -5.42755604e-01 -7.13967979e-02
-1.12191367e+00 -4.63643670e-01 7.13779807e-01 6.06422067e-01
8.31951946e-02 -1.37488377e+00 -8.44555199e-01 2.25806117e-01
5.84288955e-01 -6.45080030e-01 -2.79219389e-01 7.48300850e-01
-1.33157659e+00 -1.25404096e+00 -4.12306041e-01 -1.38735187e+00
7.11123705e-01 -2.55888730e-01 1.33117712e+00 6.62692904e-01
-2.92413384e-01 5.18534184e-01 1.59067169e-01 -2.19579220e-01
-7.75991857e-01 2.71638155e-01 -2.69769758e-01 -8.80913138e-01
1.09790467e-01 -8.50417495e-01 -3.15848142e-01 -4.90296148e-02
-1.01107562e+00 -1.91129581e-03 7.61703491e-01 9.09580708e-01
2.62248039e-01 3.20220649e-01 2.41688877e-01 -1.25029993e+00
6.11016870e-01 -2.61062652e-01 -1.17505217e+00 2.43078098e-01
-6.78454340e-01 6.06683612e-01 1.16170597e+00 -6.10587239e-01
-2.96556294e-01 1.11446798e-01 -3.38616855e-02 4.51427279e-03
-1.35310605e-01 4.55102503e-01 9.70476121e-02 -4.86881495e-01
8.09524357e-01 5.85560560e-01 7.30582848e-02 -1.85858622e-01
4.03571129e-01 2.59354711e-01 6.48154438e-01 -1.23597443e+00
7.10637331e-01 1.68807462e-01 8.62314031e-02 -3.09909254e-01
-4.80312377e-01 -1.42838746e-01 -4.60760951e-01 3.20916682e-01
2.53734946e-01 -4.89832401e-01 -1.52014279e+00 3.63751262e-01
-1.06530094e+00 -1.06210816e+00 -2.83293992e-01 -5.34290560e-02
-9.14623559e-01 5.99998981e-02 -8.89972389e-01 -5.84653616e-01
-2.97633827e-01 -1.24860895e+00 6.59206212e-01 1.99238479e-01
-1.74669787e-01 -8.88479352e-01 7.75499176e-03 -2.50363704e-02
5.82824349e-01 -1.29370987e-01 1.83843350e+00 -8.42112362e-01
-1.15587068e+00 -1.16567172e-01 -2.01231912e-01 3.44379067e-01
-7.33285904e-01 -6.92301840e-02 -3.98531765e-01 -3.67567897e-01
-3.54303300e-01 -6.65880322e-01 7.94392884e-01 1.42840087e-01
1.77818656e+00 -6.79522753e-01 -3.10180336e-01 8.43010306e-01
1.47843778e+00 1.77468315e-01 4.04167473e-01 7.84060836e-01
4.13891375e-01 3.45941037e-01 1.31812066e-01 2.30106622e-01
1.91854268e-01 1.00483336e-01 4.15557384e-01 3.64773452e-01
2.07252413e-01 -4.18347687e-01 1.93780825e-01 5.92711806e-01
4.31044810e-02 4.59657162e-02 -1.23739100e+00 5.73135376e-01
-1.64804757e+00 -7.09101796e-01 1.51406065e-01 2.32563710e+00
1.19658089e+00 4.28394794e-01 3.28239948e-01 2.69206703e-01
2.87920475e-01 -2.68412083e-01 -6.56306326e-01 -8.97483885e-01
1.80678904e-01 9.47297156e-01 8.78081024e-01 3.73831153e-01
-7.14503646e-01 1.03966403e+00 6.87215519e+00 5.84091842e-01
-1.00472403e+00 -3.86378437e-01 3.47956657e-01 3.08258068e-02
-5.48066556e-01 1.35291979e-01 -5.50081968e-01 2.26049632e-01
1.34127402e+00 -3.24620217e-01 1.26220536e+00 1.27834976e+00
-3.80276948e-01 -2.30307221e-01 -1.80208361e+00 7.23189831e-01
-2.96096623e-01 -1.57576156e+00 -1.51723713e-01 -7.80287683e-02
7.71249235e-01 -9.40815732e-02 3.83800305e-02 7.86153257e-01
7.19238877e-01 -1.39003587e+00 5.95620990e-01 9.38498005e-02
4.72226381e-01 -9.66553211e-01 4.66550261e-01 7.81282425e-01
-5.52197158e-01 -6.26973569e-01 -3.27143818e-01 -2.54069716e-01
-6.32067263e-01 2.83695400e-01 -9.46394503e-01 -4.16799575e-01
7.01350808e-01 3.82533759e-01 -5.47223508e-01 1.32775843e+00
-2.16426075e-01 7.43933916e-01 -4.90579367e-01 -6.63474798e-01
1.92416444e-01 1.98132262e-01 2.30697483e-01 1.39228380e+00
2.14782044e-01 2.38220096e-01 1.41919166e-01 8.46558690e-01
-4.72046912e-01 -9.11732689e-02 -6.12900555e-01 -2.32861489e-01
2.89963663e-01 8.38163257e-01 -5.95617056e-01 -3.61084759e-01
9.50064063e-02 6.50355995e-01 6.28071666e-01 1.67619035e-01
-7.04394698e-01 -6.06735170e-01 1.22838780e-01 7.63538852e-02
5.94712615e-01 -4.38636303e-01 -4.72225994e-01 -7.72111237e-01
1.20052777e-01 -1.33211064e+00 3.29297155e-01 -4.46629405e-01
-5.42145252e-01 2.09983513e-01 1.64475534e-02 -4.27074850e-01
-2.65861899e-01 -9.72718537e-01 -6.56993032e-01 4.44003940e-01
-1.22543538e+00 -5.69764137e-01 5.82666025e-02 2.48689786e-01
8.91785100e-02 -1.92688219e-02 7.59486437e-01 1.70073554e-01
-1.82949483e-01 9.26382840e-01 2.58585870e-01 7.21099526e-02
1.19950235e-01 -1.49671674e+00 4.58276391e-01 5.18697262e-01
1.12488985e-01 8.58596742e-01 7.89912343e-01 -1.89100057e-01
-2.45545077e+00 -7.46138275e-01 6.95443332e-01 -3.11743289e-01
1.00135779e+00 -3.35140884e-01 -8.89235377e-01 7.34528840e-01
-2.14278564e-01 -4.56977487e-02 3.30063969e-01 3.98849189e-01
-6.35669768e-01 -1.24425828e-01 -9.15926516e-01 5.55340648e-01
1.43145609e+00 -4.77432489e-01 -5.33352196e-01 5.51117539e-01
8.37726772e-01 -7.71068513e-01 -7.34166920e-01 1.93724930e-01
6.16671562e-01 -8.98640513e-01 9.92952466e-01 -1.03504646e+00
8.66492212e-01 6.61128238e-02 -4.25792225e-02 -1.23303866e+00
-6.92998245e-02 -7.15852678e-01 -5.01273572e-01 8.47769022e-01
4.88347918e-01 -5.35068214e-01 1.23048103e+00 4.58470881e-01
-3.83107215e-02 -9.89993453e-01 -6.70571148e-01 -6.73566580e-01
5.03028035e-01 -4.37234372e-01 7.04069674e-01 4.72999960e-01
2.46093214e-01 -5.76814748e-02 8.62381458e-02 9.11968574e-02
5.99495649e-01 5.86732090e-01 1.01520622e+00 -9.60148990e-01
-1.11235583e+00 -9.89086866e-01 -1.03315756e-01 -1.48255813e+00
4.38304514e-01 -1.38054097e+00 1.86724409e-01 -1.03052378e+00
3.41204315e-01 -9.52463925e-01 -4.71944176e-02 5.95847785e-01
1.41642481e-01 -3.92259151e-01 2.25531831e-01 4.93910275e-02
-7.19129801e-01 -2.93237511e-02 1.09974205e+00 -3.59950095e-01
-1.84371904e-01 2.01001123e-01 -7.61247754e-01 7.45104253e-01
5.54843605e-01 -8.00628603e-01 1.53527156e-01 -6.87332392e-01
8.06990206e-01 4.51757759e-01 2.72153825e-01 -1.24801075e+00
6.16271973e-01 -3.33185613e-01 -3.24034542e-02 -2.45653480e-01
-4.06697273e-01 -7.09507287e-01 -2.60410547e-01 9.43169951e-01
-9.81461167e-01 3.22242409e-01 4.38955247e-01 4.57394183e-01
1.48101792e-01 -6.12611175e-01 5.94750464e-01 -3.47263694e-01
-6.66066229e-01 2.96546459e-01 -4.03801501e-01 2.66090751e-01
7.77581751e-01 -1.11142643e-01 -4.25696254e-01 -2.16347858e-01
-3.82539779e-01 4.76725847e-01 6.64229751e-01 -2.31795944e-02
3.13519061e-01 -6.60407186e-01 -3.83547962e-01 2.58163273e-01
-6.75531179e-02 2.07893193e-01 -1.80498511e-01 5.56407273e-01
-1.00092018e+00 5.13159633e-01 -2.24487394e-01 -6.00668371e-01
-1.07489431e+00 9.27160501e-01 1.22211106e-01 -5.08943856e-01
-4.40779448e-01 9.46534872e-01 -1.49160549e-01 -8.98359656e-01
8.13126087e-01 -7.39960432e-01 3.68859231e-01 -5.81102610e-01
5.42080939e-01 1.82314038e-01 6.39463142e-02 5.72482646e-01
-3.78450640e-02 2.41722479e-01 -1.17159477e-02 3.42484474e-01
1.79048455e+00 7.04490721e-01 -5.20134330e-01 2.07183167e-01
1.18971169e+00 -8.91335979e-02 -1.02634454e+00 -1.89157486e-01
4.49046701e-01 -2.12389514e-01 -2.46608183e-01 -6.45678222e-01
-8.01436305e-01 8.94931376e-01 3.98533434e-01 2.72456586e-01
1.11021829e+00 -4.86843660e-02 9.71210361e-01 1.29448760e+00
7.00864494e-01 -9.28482831e-01 9.45775136e-02 7.95800090e-01
4.11528468e-01 -1.00435722e+00 -1.42781422e-01 5.67238517e-02
7.46567696e-02 1.47415435e+00 5.81336439e-01 -6.39633358e-01
1.96112156e-01 7.54404843e-01 -6.42074525e-01 1.54129773e-01
-1.15128863e+00 1.49046287e-01 -1.56400576e-01 4.77099478e-01
2.33261496e-01 -9.19999834e-03 -3.60531285e-02 4.19543415e-01
-4.66964036e-01 2.62672037e-01 6.27692759e-01 1.26793289e+00
-8.62716377e-01 -1.18965614e+00 -1.26137629e-01 6.32420421e-01
-3.82477641e-01 -1.41021818e-01 7.35218218e-03 8.59226286e-01
-8.06734562e-02 3.17021698e-01 -2.36047730e-02 -4.16695565e-01
2.54517853e-01 8.51595700e-02 8.33729625e-01 -5.32798886e-01
-9.38855886e-01 -7.48369455e-01 -7.57455220e-03 -6.95905566e-01
5.62549308e-02 -3.98762167e-01 -1.39265907e+00 -6.57103658e-01
5.40425582e-03 2.37303808e-01 6.88788354e-01 7.77878761e-01
1.59560174e-01 2.00296983e-01 4.58414406e-01 -8.69859934e-01
-9.74710047e-01 -2.90142596e-01 -2.36600593e-01 1.95081383e-01
5.39380610e-01 6.06706105e-02 -4.42366004e-01 -8.82382840e-02] | [8.663511276245117, 7.28926944732666] |
f6b0520e-b415-4e1a-b05e-13ecaffd03fc | multi-target-tracking-and-occlusion-handling | 1511.01726 | null | http://arxiv.org/abs/1511.01726v1 | http://arxiv.org/pdf/1511.01726v1.pdf | Multi-Target Tracking and Occlusion Handling with Learned Variational Bayesian Clusters and a Social Force Model | This paper considers the problem of multiple human target tracking in a
sequence of video data. A solution is proposed which is able to deal with the
challenges of a varying number of targets, interactions and when every target
gives rise to multiple measurements. The developed novel algorithm comprises
variational Bayesian clustering combined with a social force model, integrated
within a particle filter with an enhanced prediction step. It performs
measurement-to-target association by automatically detecting the measurement
relevance. The performance of the developed algorithm is evaluated over several
sequences from publicly available data sets: AV16.3, CAVIAR and PETS2006, which
demonstrates that the proposed algorithm successfully initializes and tracks a
variable number of targets in the presence of complex occlusions. A comparison
with state-of-the-art techniques due to Khan et al., Laet et al. and Czyz et
al. shows improved tracking performance. | ['Ata-ur-Rehman', 'Syed Mohsen Naqvi', 'Lyudmila Mihaylova', 'Jonathon Chambers'] | 2015-11-05 | null | null | null | null | ['occlusion-handling'] | ['computer-vision'] | [ 1.84579432e-01 -4.72746640e-01 2.44159356e-01 -9.31788683e-02
-6.68551147e-01 -5.57932913e-01 1.08290505e+00 4.02506560e-01
-6.63010299e-01 8.55910122e-01 -1.23072363e-01 1.49828315e-01
-4.74639297e-01 -3.48639667e-01 -4.69526619e-01 -7.25020587e-01
-2.40690455e-01 1.04308164e+00 9.96906996e-01 4.92293090e-02
1.65281251e-01 7.13568807e-01 -1.85680199e+00 -2.73877233e-01
2.37297446e-01 7.48937011e-01 5.40314257e-01 1.18445945e+00
3.93485218e-01 5.49867690e-01 -6.81042016e-01 -1.19724296e-01
3.49685997e-01 -9.70568508e-03 -1.25803307e-01 2.78595984e-01
6.49627745e-01 1.70760199e-01 2.11373866e-01 1.08481264e+00
5.66027999e-01 2.87486076e-01 9.24416482e-01 -1.19291496e+00
1.57636359e-01 1.33686319e-01 -7.51198351e-01 8.14706445e-01
3.33431482e-01 1.97919980e-02 5.63117146e-01 -5.15536010e-01
5.95315874e-01 1.36365247e+00 7.41925299e-01 1.67326063e-01
-1.38214600e+00 -5.75223505e-01 -1.34962946e-01 5.37468612e-01
-1.28456295e+00 -2.36227274e-01 3.26385319e-01 -9.55309391e-01
4.97686923e-01 4.14216876e-01 6.10607624e-01 1.08622849e+00
2.17965156e-01 3.27533752e-01 1.02304709e+00 -3.30001175e-01
2.57192641e-01 7.36587681e-03 5.26198506e-01 2.87601531e-01
6.15516365e-01 5.79796791e-01 -2.43986189e-01 -4.89697963e-01
2.53972828e-01 -6.17993213e-02 1.24598518e-01 -4.85008836e-01
-1.09382284e+00 8.15810025e-01 -1.38738766e-01 4.93538529e-01
-7.50711799e-01 -2.09269021e-02 3.56945068e-01 -9.52053815e-02
3.54440182e-01 -3.09138387e-01 -1.48086682e-01 -8.91155452e-02
-1.13127673e+00 6.28456652e-01 6.63316190e-01 6.62379622e-01
2.82832384e-01 8.09920728e-02 -3.13464761e-01 2.27723718e-01
8.14770222e-01 1.08815706e+00 2.16368109e-01 -6.65114999e-01
1.74075216e-01 1.52969450e-01 8.32106411e-01 -1.15033877e+00
-5.24675071e-01 -4.24001336e-01 -4.80029225e-01 7.02214420e-01
4.48246032e-01 -2.90726900e-01 -7.19672084e-01 1.49013841e+00
1.14378858e+00 7.86974669e-01 1.59517005e-01 8.06006074e-01
5.11783898e-01 4.36982661e-01 1.75983250e-01 -5.52738905e-01
1.36350858e+00 -6.48101687e-01 -9.57521081e-01 1.50655255e-01
-1.07446298e-01 -1.26205647e+00 -1.02219999e-01 5.49697518e-01
-6.68133855e-01 -1.10366058e+00 -8.27894926e-01 9.28785980e-01
-2.00806007e-01 2.57666439e-01 -2.24286746e-02 9.27431941e-01
-8.87004852e-01 5.94514728e-01 -9.55150902e-01 -8.37750137e-01
-6.13293238e-02 5.14451325e-01 -1.35776311e-01 3.00319850e-01
-8.10325563e-01 1.15175021e+00 6.46104157e-01 6.55743405e-02
-1.17822552e+00 -4.10840929e-01 -4.29724425e-01 -3.05388510e-01
4.42377537e-01 -5.10395825e-01 1.02816164e+00 -6.68544292e-01
-1.30808413e+00 4.78454173e-01 -1.74692377e-01 -7.87295103e-01
8.38331342e-01 -5.38057566e-01 -5.41416287e-01 -9.64788944e-02
5.09508550e-02 2.29777113e-01 9.97221351e-01 -1.24890018e+00
-1.07818615e+00 -5.33746660e-01 -4.17475432e-01 3.55563074e-01
3.64631951e-01 2.55380929e-01 -2.72920668e-01 -3.93440813e-01
-3.13440710e-01 -1.05818248e+00 -3.09837967e-01 -4.85482365e-01
-2.29612634e-01 -1.74406067e-01 1.05002940e+00 -5.88137746e-01
8.17573071e-01 -1.74984467e+00 4.20775980e-01 2.97990948e-01
1.59952819e-01 5.53737104e-01 5.52447140e-02 5.70196807e-01
9.05391574e-02 -5.68221271e-01 1.46976814e-01 -6.61493778e-01
8.29241201e-02 1.57767981e-01 2.16279775e-01 1.03607380e+00
-4.55794603e-01 2.23093510e-01 -6.88211679e-01 -7.48216271e-01
7.24535286e-01 6.22543395e-01 7.73343891e-02 3.94675702e-01
-2.88792342e-01 7.69998491e-01 -5.09252906e-01 3.20483655e-01
8.70052755e-01 8.09229985e-02 8.70747492e-02 -2.08622724e-01
-4.71255183e-01 -7.06500232e-01 -1.93043423e+00 1.03215206e+00
4.09490287e-01 5.19686043e-01 2.29617625e-01 -8.48401487e-01
9.04198050e-01 4.84703392e-01 8.96752357e-01 -7.68282339e-02
5.67095757e-01 -1.43442914e-01 9.53446925e-02 -2.95496166e-01
5.01406133e-01 8.89560580e-02 3.13264459e-01 1.88529015e-01
-5.37743904e-02 4.04861182e-01 5.23014009e-01 -4.72559370e-02
1.18067443e+00 2.22252786e-01 4.44636941e-01 -3.08769494e-01
9.24877048e-01 1.96127772e-01 5.15707970e-01 1.03624392e+00
-5.46551585e-01 1.63654551e-01 -3.60412270e-01 -1.54547185e-01
-9.59797263e-01 -1.18033707e+00 -7.19398782e-02 7.82643855e-01
2.32256696e-01 -1.41059071e-01 -4.54516500e-01 -1.61485478e-01
7.08679780e-02 3.58118206e-01 -8.61061454e-01 3.12382013e-01
-3.71968985e-01 -8.55840862e-01 1.64335892e-01 -1.64522007e-01
1.81686834e-01 -9.11091328e-01 -1.08506060e+00 5.40071428e-01
4.84411046e-02 -1.22714353e+00 6.66096434e-02 -3.11008543e-01
-7.37437427e-01 -1.38322163e+00 -5.08787632e-01 -1.59917399e-01
1.15586087e-01 1.43547371e-01 1.04966187e+00 -2.13463217e-01
-6.16461456e-01 8.34439337e-01 -5.23075998e-01 -7.56104112e-01
-6.57733619e-01 -5.09502530e-01 3.10347795e-01 2.18598932e-01
4.30679649e-01 -3.59122008e-01 -3.43690634e-01 3.79309535e-01
-4.22718704e-01 -5.15172541e-01 3.52707535e-01 5.74319005e-01
5.36843061e-01 2.67253011e-01 -1.55374482e-02 -4.92809534e-01
3.11193764e-01 -4.48713422e-01 -1.38021529e+00 2.66062438e-01
-3.81424963e-01 -2.98439115e-01 4.26215716e-02 -6.13992035e-01
-1.00069726e+00 4.73031253e-01 1.46584183e-01 -6.61289394e-01
-4.08297330e-01 -7.97392428e-02 2.18374461e-01 -3.30426335e-01
6.05924785e-01 1.97869837e-01 1.78655818e-01 -5.06750345e-01
1.82377651e-01 2.30858311e-01 3.72433633e-01 -3.53047192e-01
1.18976784e+00 5.10280609e-01 4.71204191e-01 -8.61771464e-01
-5.69115400e-01 -1.00430989e+00 -8.25280130e-01 -8.93179417e-01
1.11507738e+00 -8.77283931e-01 -1.00190902e+00 5.73777080e-01
-1.14526439e+00 1.49524778e-01 -6.52212203e-02 9.27747846e-01
-4.49764609e-01 6.11361861e-01 -2.36217439e-01 -1.43577397e+00
-2.11617842e-01 -1.08462119e+00 9.42092359e-01 3.15880835e-01
-8.17771181e-02 -1.02830684e+00 7.04800487e-01 3.27730715e-01
2.24657446e-01 6.33193910e-01 -2.58157432e-01 -8.46083164e-01
-5.57276309e-01 -4.05752599e-01 2.44354755e-01 1.13987990e-01
2.50720549e-02 1.51729971e-01 -5.60447216e-01 -5.54086506e-01
1.73415057e-02 2.19552621e-01 4.85763669e-01 7.57783532e-01
-2.74691009e-03 2.16740444e-01 -6.25916123e-01 -1.96844175e-01
1.61098409e+00 5.02853036e-01 2.34672114e-01 4.07822847e-01
5.15056610e-01 3.33627492e-01 1.19984436e+00 7.78505147e-01
6.94959685e-02 1.01227677e+00 7.06046402e-01 3.12036961e-01
1.10889710e-01 5.61402321e-01 3.25997680e-01 2.37541705e-01
-2.37984136e-01 -4.24574286e-01 -7.78899968e-01 4.46613431e-01
-2.23727560e+00 -1.41357613e+00 -6.75420523e-01 2.21636343e+00
-1.68508217e-02 1.82641953e-01 5.57337821e-01 1.01193637e-01
1.17052281e+00 -2.46763062e-02 -5.29581830e-02 4.21832800e-01
9.18969512e-02 3.12403589e-02 7.44578421e-01 9.12267745e-01
-1.48343801e+00 5.79870939e-01 5.95472670e+00 8.17747653e-01
-5.30113459e-01 4.64189410e-01 -3.62969249e-01 2.76558176e-02
7.32898057e-01 -2.00694934e-01 -1.36491895e+00 5.71415126e-01
1.17797649e+00 -1.76461801e-01 -1.01296589e-01 3.90374035e-01
4.08946246e-01 -5.82546830e-01 -6.29777968e-01 7.98715949e-01
2.72599578e-01 -8.81636202e-01 -3.64268124e-01 1.19161725e-01
3.75859082e-01 3.07963490e-01 -1.23129794e-02 -3.47287417e-03
6.38223767e-01 -6.07686937e-01 7.54180372e-01 8.56785774e-01
-4.71933544e-01 -5.01799703e-01 8.14938664e-01 4.87730324e-01
-1.50360477e+00 3.77154872e-02 -3.46549809e-01 4.83454764e-02
5.15412986e-01 3.79871905e-01 -9.25941229e-01 9.11441386e-01
7.38686681e-01 6.64541125e-01 -6.59046829e-01 1.54964447e+00
1.97701901e-01 5.49554408e-01 -7.17011273e-01 -1.52386546e-01
2.14857608e-01 -2.43105844e-01 1.25628376e+00 1.08218682e+00
5.06006062e-01 -1.42815392e-02 5.50098658e-01 2.83329606e-01
1.06431448e+00 1.68518171e-01 -4.78324682e-01 2.72533864e-01
4.39152032e-01 1.33689463e+00 -9.15806770e-01 -4.25508887e-01
-1.45194575e-01 4.26031619e-01 4.44151051e-02 -2.42332537e-02
-1.05823576e+00 5.37825108e-01 3.87246668e-01 -2.14626882e-02
8.03092241e-01 -2.22270951e-01 4.27701175e-01 -6.96399987e-01
-3.20041180e-01 -5.55592537e-01 7.29347467e-01 -5.62962949e-01
-1.11202919e+00 5.95649958e-01 3.76174241e-01 -1.42936409e+00
-3.69280517e-01 -4.51882035e-01 -4.57706630e-01 8.70925844e-01
-8.57653558e-01 -1.21875489e+00 -2.53252029e-01 7.16931403e-01
4.29116875e-01 -6.16434455e-01 5.50108612e-01 4.95947301e-01
-3.33399832e-01 -1.91865414e-01 3.29622835e-01 -3.43016326e-01
4.05738950e-01 -1.09511447e+00 -5.05474620e-02 1.20425558e+00
1.79666355e-01 9.77737904e-02 1.61462808e+00 -1.09704900e+00
-1.09106696e+00 -1.01731765e+00 3.52719873e-01 -6.64178967e-01
8.01083386e-01 -1.12007461e-01 -4.92741108e-01 7.60477602e-01
4.81539607e-01 -1.80239379e-01 6.13461912e-01 -2.68559698e-02
2.48422712e-01 3.07657383e-02 -1.12227559e+00 1.18386455e-01
4.83520627e-01 2.95725644e-01 -8.25988412e-01 5.52298486e-01
-2.36742944e-02 -4.26000684e-01 -9.41336930e-01 4.88854617e-01
3.25639606e-01 -1.28966558e+00 1.10703278e+00 -2.01217964e-01
-8.26223254e-01 -7.49243259e-01 -3.04310411e-01 -1.09258175e+00
-5.45462489e-01 -5.23807049e-01 -2.14457676e-01 1.23447847e+00
-6.59016371e-02 -3.01997215e-01 8.25003505e-01 3.16278301e-02
2.51535982e-01 1.22428343e-01 -1.35270727e+00 -8.87633681e-01
-6.03152931e-01 -2.17836797e-01 -4.23054956e-02 3.98863554e-01
-7.58526325e-01 2.22451240e-01 -7.72222102e-01 7.62290716e-01
1.63233876e+00 -1.60172716e-01 1.11176395e+00 -1.94310641e+00
-6.05913401e-01 -1.16462283e-01 -1.01333177e+00 -2.91861713e-01
4.80151623e-02 -2.41114125e-01 7.98827857e-02 -1.30102527e+00
2.76151597e-01 -6.59153610e-02 -2.48237267e-01 -3.03528070e-01
-2.18871266e-01 1.60291299e-01 4.70312029e-01 1.53160587e-01
-8.21486473e-01 2.22202003e-01 7.39102244e-01 1.96248963e-02
1.59592256e-01 4.45413172e-01 2.31203794e-01 5.97031415e-01
6.77226424e-01 -7.92041302e-01 -1.97099149e-01 1.56934500e-01
-4.62312490e-01 1.37906268e-01 7.28105783e-01 -1.66018927e+00
4.54878151e-01 -1.06404416e-01 4.74928528e-01 -1.32424915e+00
6.33004785e-01 -1.28857219e+00 1.06777573e+00 9.27224636e-01
2.40674242e-01 3.35165113e-01 2.68396676e-01 1.12773752e+00
-4.09033988e-03 -4.49753195e-01 9.83748436e-01 -7.84205049e-02
-9.45009112e-01 1.97315156e-01 -5.18574357e-01 -2.65520006e-01
1.73832321e+00 -2.09308639e-01 1.19836321e-02 -2.76297688e-01
-1.35120666e+00 2.52087593e-01 1.83115005e-02 4.10774142e-01
1.18143260e-01 -1.20183206e+00 -9.03477669e-01 -3.57011229e-01
-2.20376641e-01 -5.96516073e-01 5.48564076e-01 9.29043114e-01
-2.59468079e-01 2.81810611e-01 -3.70435059e-01 -1.12801802e+00
-2.32055020e+00 6.14564717e-01 2.42121607e-01 -5.83839118e-01
-4.70059365e-01 4.60501105e-01 -2.96723425e-01 -2.99375635e-02
4.28627729e-01 2.99975753e-01 -8.10460091e-01 2.92725772e-01
5.28864801e-01 6.57023013e-01 -4.13632959e-01 -1.31712651e+00
-4.33779716e-01 9.43165302e-01 1.95302799e-01 -3.99288237e-01
1.24415481e+00 -3.50677192e-01 2.41277263e-01 5.45661986e-01
4.64095980e-01 -9.07168910e-02 -1.33219135e+00 -1.16642989e-01
2.32179508e-01 -5.57976186e-01 -1.59671083e-01 -6.18486285e-01
-7.09665537e-01 4.40030515e-01 1.48435903e+00 2.04716623e-01
5.80630660e-01 -2.18685657e-01 5.20119704e-02 2.04592511e-01
4.03213322e-01 -8.17160726e-01 -2.11463664e-02 4.42739218e-01
6.16424561e-01 -1.24614167e+00 3.01404685e-01 -4.07374978e-01
-3.31375241e-01 8.90870333e-01 5.53936183e-01 -2.47836486e-01
8.58123243e-01 4.84077573e-01 4.29848731e-02 -2.43897706e-01
-5.37950635e-01 -7.50264287e-01 3.61999780e-01 8.94365966e-01
-7.17746764e-02 1.14935212e-01 -5.62154949e-01 -1.17703445e-01
1.16028368e-01 6.03147596e-03 3.75342518e-01 1.11221087e+00
-7.18904734e-01 -1.00046456e+00 -1.21879029e+00 1.29919976e-01
-4.34758276e-01 3.89522970e-01 2.49075633e-03 1.12482381e+00
3.36057276e-01 1.26069260e+00 -6.76245093e-02 -2.74967670e-01
5.39226174e-01 -1.35546505e-01 6.06923282e-01 -3.58745158e-01
-8.07120919e-01 5.78701854e-01 1.34077206e-01 -4.01723891e-01
-1.30787826e+00 -1.31856620e+00 -6.71960652e-01 -6.28894269e-02
-4.42790210e-01 5.53763628e-01 7.18513727e-01 9.47119832e-01
9.24644917e-02 7.16359377e-01 2.11821064e-01 -1.22523725e+00
-5.65691113e-01 -1.29756355e+00 -7.86187887e-01 4.56672937e-01
2.09021688e-01 -1.34735417e+00 -3.94802421e-01 2.15151891e-01] | [6.586842060089111, -1.989241361618042] |
e4c5f796-af85-4e6d-a617-27eaf3796fa5 | 3d-speaker-a-large-scale-multi-device-multi | 2306.15354 | null | https://arxiv.org/abs/2306.15354v2 | https://arxiv.org/pdf/2306.15354v2.pdf | 3D-Speaker: A Large-Scale Multi-Device, Multi-Distance, and Multi-Dialect Corpus for Speech Representation Disentanglement | Disentangling uncorrelated information in speech utterances is a crucial research topic within speech community. Different speech-related tasks focus on extracting distinct speech representations while minimizing the affects of other uncorrelated information. We present a large-scale speech corpus to facilitate the research of speech representation disentanglement. 3D-Speaker contains over 10,000 speakers, each of whom are simultaneously recorded by multiple Devices, locating at different Distances, and some speakers are speaking multiple Dialects. The controlled combinations of multi-dimensional audio data yield a matrix of a diverse blend of speech representation entanglement, thereby motivating intriguing methods to untangle them. The multi-domain nature of 3D-Speaker also makes it a suitable resource to evaluate large universal speech models and experiment methods of out-of-domain learning and self-supervised learning. https://3dspeaker.github.io/ | ['Qian Chen', 'Hui Wang', 'Yafeng Chen', 'Luyao Cheng', 'Siqi Zheng'] | 2023-06-27 | null | null | null | null | ['self-supervised-learning', 'disentanglement'] | ['computer-vision', 'methodology'] | [ 3.72360423e-02 2.85324991e-01 -4.87924308e-01 -1.99033692e-01
-1.07159305e+00 -9.12162364e-01 9.27337229e-01 -3.98176610e-01
1.69134721e-01 3.01793545e-01 9.27212119e-01 -3.70566905e-01
-1.55867174e-01 -1.63786635e-01 -1.90039456e-01 -1.00022995e+00
-2.00166851e-01 7.02986121e-01 -3.42321664e-01 -3.22093725e-01
-4.25638020e-01 3.21607411e-01 -1.43211138e+00 2.51895398e-01
2.63962895e-01 4.03874218e-01 2.76306152e-01 8.56729209e-01
3.83980349e-02 5.57775974e-01 -9.34473276e-01 -3.07918519e-01
1.77343845e-01 -3.38922769e-01 -4.68698651e-01 -2.29157358e-02
3.34860653e-01 -4.55326997e-02 -1.16397643e+00 9.55416918e-01
9.05446589e-01 7.43368417e-02 7.06150055e-01 -1.23650682e+00
-8.05460215e-01 1.08595049e+00 -6.74177632e-02 5.68373203e-01
2.58350700e-01 -3.44010703e-02 1.25462627e+00 -6.55029714e-01
4.68985327e-02 1.41560507e+00 3.68957669e-02 7.78731763e-01
-1.37896764e+00 -1.00286448e+00 -1.16972461e-01 6.81683943e-02
-1.45019293e+00 -1.48475885e+00 1.11947358e+00 -3.49330395e-01
9.10859346e-01 7.22392499e-01 4.77686286e-01 1.96406472e+00
-4.28007007e-01 8.72909009e-01 9.01955962e-01 -4.16584909e-01
1.92242831e-01 3.01228553e-01 1.58475742e-01 3.61868054e-01
5.77154309e-02 4.41159934e-01 -1.14417827e+00 -3.90719563e-01
4.56484497e-01 -2.41416067e-01 -2.31118411e-01 -1.92508608e-01
-1.49281573e+00 5.61171770e-01 -2.95609385e-01 4.08999413e-01
2.29585066e-01 -1.90041482e-01 4.53156799e-01 8.49865019e-01
3.17188561e-01 1.98458940e-01 -3.61253709e-01 -3.34925443e-01
-7.83995390e-01 1.00796349e-01 8.37610304e-01 1.36840773e+00
4.58773494e-01 6.75880075e-01 2.51757711e-01 9.87794340e-01
4.45242763e-01 1.03799856e+00 6.32602632e-01 -8.29172730e-01
8.99060786e-01 -1.62012875e-01 -4.36560094e-01 -7.83860385e-01
-1.68355033e-01 -4.96297032e-01 -1.20569360e+00 -2.59703934e-01
3.48468214e-01 -1.94494382e-01 -4.70352858e-01 1.98994291e+00
5.53197376e-02 4.00190979e-01 4.24100906e-01 6.74305201e-01
9.16854680e-01 5.55826843e-01 -2.91330963e-01 -4.28655297e-01
1.50026882e+00 -7.61191189e-01 -8.79320979e-01 -1.62616923e-01
3.83841366e-01 -7.67638803e-01 8.75459850e-01 3.47774416e-01
-8.96952271e-01 -4.15150076e-01 -1.11620402e+00 -9.73473396e-03
-2.62699097e-01 7.20820129e-02 6.41076803e-01 1.01314259e+00
-7.55155325e-01 8.10105205e-02 -7.44168580e-01 1.31828472e-01
3.77576113e-01 2.69966304e-01 -7.18168259e-01 3.44346672e-01
-1.25166273e+00 6.30532742e-01 2.10117791e-02 -1.96825907e-01
-1.20802724e+00 -6.87559664e-01 -6.03688419e-01 -3.41489092e-02
1.67379424e-01 -4.80276138e-01 1.14838719e+00 -2.70304531e-01
-1.73303771e+00 9.74529922e-01 -2.34798923e-01 -1.83634326e-01
1.46868259e-01 -1.02024198e-01 -1.11756635e+00 -4.01786417e-02
3.16799544e-02 -1.51975441e-03 1.15316296e+00 -8.77493680e-01
1.43644661e-01 -5.39094567e-01 -3.53318542e-01 3.07587624e-01
-3.78338158e-01 4.64629792e-02 4.53832112e-02 -8.47131431e-01
4.48321909e-01 -1.03330028e+00 3.24055612e-01 -4.51061726e-01
-8.12417507e-01 -1.72096938e-01 8.84448707e-01 -3.13196361e-01
9.23169672e-01 -2.49779582e+00 4.56691712e-01 -2.86155403e-01
6.57384992e-01 2.88372368e-01 -1.02217384e-01 7.55898476e-01
-4.40248042e-01 2.88237810e-01 1.57616347e-01 -8.85509014e-01
1.84068978e-01 2.17631549e-01 -8.58221531e-01 5.72330952e-01
-3.49248722e-02 4.73037988e-01 -7.82766283e-01 -3.63087475e-01
3.06401700e-01 3.90774161e-01 -3.53633761e-01 3.42353553e-01
4.69332114e-02 8.39410663e-01 -4.82563794e-01 5.85701942e-01
7.61000872e-01 7.29207918e-02 4.15631294e-01 -8.37496817e-02
1.68045327e-01 1.23724437e+00 -8.89549077e-01 1.98563921e+00
-3.79736096e-01 1.17449903e+00 3.70006830e-01 -1.10681891e+00
8.28972816e-01 8.45707834e-01 2.08673045e-01 -4.36436862e-01
4.84770723e-02 1.17463572e-02 4.24946547e-01 -5.18792212e-01
3.12786371e-01 -3.34295779e-01 -3.69129747e-01 6.09098017e-01
5.30458272e-01 -2.54181951e-01 -4.72757757e-01 3.95584702e-01
1.12438047e+00 -9.29638028e-01 2.96228915e-01 -2.52304345e-01
1.73369661e-01 -9.00581598e-01 3.10091078e-01 7.71461546e-01
-7.31749535e-01 5.51084936e-01 4.45155054e-01 2.11058408e-02
-8.86743844e-01 -1.79718363e+00 -3.46781194e-01 1.07001388e+00
-9.89325568e-02 -6.16270721e-01 -2.82871246e-01 -2.16479734e-01
-1.02737524e-01 6.71864927e-01 -2.13049456e-01 -3.34026992e-01
-3.51417780e-01 -4.46685404e-01 1.31785083e+00 6.37035668e-02
1.49003386e-01 -5.28380573e-01 4.62000221e-01 -2.75728881e-01
-3.47811311e-01 -1.34532917e+00 -4.56260502e-01 4.71467614e-01
-4.80815887e-01 -5.82107782e-01 -4.26219761e-01 -6.72626734e-01
6.60328269e-02 6.25766635e-01 1.07020760e+00 -5.39597571e-01
1.82371870e-01 1.84571430e-01 -1.52536035e-01 -2.72707403e-01
-8.58259559e-01 3.94102365e-01 9.11720216e-01 -1.49657233e-02
4.03998852e-01 -1.25405252e+00 7.41834119e-02 2.64722973e-01
-3.36445123e-01 -9.47616845e-02 3.17109019e-01 6.90367520e-01
4.53649536e-02 -8.24915990e-02 7.98667550e-01 -4.13816035e-01
7.79795289e-01 -8.62106681e-01 -1.73632905e-01 9.07969400e-02
2.54627496e-01 2.29374290e-01 4.58477885e-01 -7.55324125e-01
-8.63561749e-01 -4.03430223e-01 -7.13738948e-02 -5.08181691e-01
-4.35434103e-01 -4.32795174e-02 -7.07872093e-01 5.91870844e-01
5.71108460e-01 4.64830250e-01 1.41963094e-01 -5.54026484e-01
7.09600925e-01 1.29507327e+00 4.29496169e-01 -6.65265679e-01
8.68621111e-01 1.92920595e-01 -4.68979061e-01 -1.76697373e+00
-5.73328197e-01 -3.46709788e-01 -6.04373991e-01 1.09911226e-01
6.35933816e-01 -1.38622451e+00 -8.07871640e-01 3.31056267e-01
-1.17651784e+00 2.37808172e-02 -2.60006458e-01 8.92836988e-01
-4.90336776e-01 1.84445545e-01 -3.85094196e-01 -1.13347995e+00
5.66282310e-02 -1.25060844e+00 1.12384319e+00 -2.07945660e-01
-4.17004794e-01 -7.70652711e-01 3.47840935e-01 8.03041279e-01
1.54066145e-01 -5.20733953e-01 8.47610474e-01 -1.28820062e+00
-6.45771563e-01 1.02414824e-01 3.39835376e-01 3.06063145e-01
5.74878812e-01 -3.02377284e-01 -1.59115314e+00 -1.89317614e-01
2.48552874e-01 -3.78624052e-01 6.25984848e-01 -6.40216051e-03
8.42429698e-01 -5.40147185e-01 -2.92759299e-01 5.66489220e-01
2.79103607e-01 -1.94475371e-02 2.10605413e-01 -4.92541164e-01
7.84890175e-01 5.74855030e-01 -3.56746316e-01 3.58502030e-01
3.80639166e-01 8.85870337e-01 7.04189166e-02 5.14493883e-01
-3.99025828e-01 -3.92768502e-01 6.91291451e-01 2.12094998e+00
2.24555716e-01 -5.33803880e-01 -6.53976977e-01 3.89676809e-01
-1.34892380e+00 -1.37076986e+00 1.06053501e-01 2.04740930e+00
8.17907453e-01 1.75458342e-02 3.61825824e-01 3.24246258e-01
7.07901835e-01 7.68007457e-01 -4.84317482e-01 -3.13154571e-02
-5.69155216e-01 -3.93533520e-02 3.10476329e-02 5.14574289e-01
-1.00225854e+00 8.32279682e-01 6.87979507e+00 1.10625970e+00
-9.63494480e-01 2.92559952e-01 2.55650848e-01 -5.59945583e-01
-7.08142757e-01 -1.36497930e-01 -6.34052455e-01 4.78714854e-01
1.32878482e+00 -3.80123436e-01 6.30118787e-01 5.88693678e-01
3.74954455e-02 6.39102161e-01 -1.24990344e+00 1.42382729e+00
5.63791627e-03 -1.37545335e+00 -4.02632579e-02 2.85109043e-01
3.15666795e-01 6.32060170e-01 4.64405656e-01 2.85267740e-01
3.79470825e-01 -9.40770864e-01 8.95249367e-01 1.80040956e-01
8.35275650e-01 -4.40969437e-01 -1.27456501e-01 6.99623942e-01
-1.00886035e+00 1.11562796e-01 -1.76572174e-01 -1.91090018e-01
1.29058838e-01 5.32268763e-01 -8.36919725e-01 2.43063807e-01
3.77454996e-01 6.32444739e-01 -6.51284978e-02 1.02974311e-01
-1.83578536e-01 8.35765183e-01 -3.20784181e-01 -1.78087175e-01
-4.43645358e-01 -2.62530804e-01 1.16415048e+00 1.07400012e+00
1.27855986e-01 -2.41634604e-02 -2.13140160e-01 1.09111249e+00
-2.97602892e-01 -4.30016160e-01 -1.23084736e+00 -5.63705385e-01
1.14457214e+00 9.23544526e-01 -2.00009599e-01 -1.68955669e-01
-3.83209020e-01 6.86852276e-01 4.54715014e-01 4.14276779e-01
-5.57124436e-01 7.25336894e-02 1.54755056e+00 -7.38522783e-02
3.62149291e-02 -9.34566855e-01 -7.84845650e-02 -1.56754088e+00
1.24077862e-02 -1.12203145e+00 -1.81883872e-01 -4.60798383e-01
-1.53577518e+00 7.35286057e-01 -6.60569593e-02 -1.33726215e+00
-3.41698050e-01 -3.86833012e-01 -4.24461633e-01 7.58226693e-01
-8.43566537e-01 -8.96030068e-01 3.47179681e-01 6.45049214e-01
6.85145557e-01 -1.10136938e+00 1.23040962e+00 3.84361446e-01
-8.98812473e-01 8.21855009e-01 2.68633991e-01 1.92250073e-01
5.09750009e-01 -9.50022638e-01 7.52287626e-01 3.48732412e-01
9.78773475e-01 9.20026183e-01 7.36264706e-01 -1.03087671e-01
-1.61831319e+00 -5.84642887e-01 8.03573191e-01 -9.67434168e-01
9.50032771e-01 -1.31166828e+00 -6.63915455e-01 7.24357247e-01
3.75595182e-01 4.24494259e-02 1.26817214e+00 3.99672478e-01
-8.69006753e-01 -1.10220127e-01 -6.83138430e-01 9.25349176e-01
1.44237423e+00 -1.64842892e+00 -8.84409726e-01 3.93090397e-01
1.18781734e+00 -1.79566428e-01 -6.43378556e-01 -3.01873058e-01
5.51827371e-01 -1.03724623e+00 1.22873890e+00 -7.72257030e-01
-6.87051862e-02 -1.98974349e-02 -8.21665704e-01 -1.46730828e+00
1.09241446e-02 -1.31163967e+00 -4.08107847e-01 1.50609016e+00
4.73800242e-01 -9.12327826e-01 5.97048461e-01 3.30475271e-01
-3.23747665e-01 -1.60491049e-01 -1.42758060e+00 -9.46364999e-01
2.96576142e-01 -8.60918880e-01 7.78167546e-01 1.01837146e+00
4.92085367e-01 8.48492742e-01 -6.25953913e-01 4.68653053e-01
7.94591486e-01 -1.65934727e-01 8.44524622e-01 -1.08032024e+00
-4.84406203e-01 -4.70980614e-01 -5.68111122e-01 -1.58291066e+00
7.10608304e-01 -1.26371932e+00 -4.36143517e-01 -3.30859035e-01
-1.48789436e-01 -5.49944997e-01 1.97211141e-03 1.15041919e-02
3.95445794e-01 -4.23428148e-01 1.07138835e-01 4.05364603e-01
-3.03140461e-01 9.13588464e-01 1.14286172e+00 -4.27658677e-01
-2.06526164e-02 3.39164227e-01 -9.28940654e-01 5.42306602e-01
7.70379066e-01 -6.82658017e-01 -8.49736392e-01 -5.10714948e-01
-1.30499214e-01 4.24029320e-01 2.92974412e-01 -7.29434252e-01
2.89382339e-01 -2.77177729e-02 -2.68735856e-01 -3.67543370e-01
9.50424790e-01 -5.61920524e-01 -1.51683921e-02 -1.30155087e-01
-6.58138812e-01 -2.44617507e-01 5.71233220e-02 6.88905001e-01
-3.45217764e-01 1.09934151e-01 4.05340999e-01 -3.51928100e-02
-8.33771825e-02 2.21563190e-01 -5.45994937e-01 6.63446724e-01
5.70064843e-01 2.14796856e-01 -5.97246051e-01 -5.90866148e-01
-8.40321064e-01 -5.02435982e-01 -3.70641977e-01 8.85330379e-01
3.87990624e-01 -1.48404551e+00 -7.03122020e-01 7.93538034e-01
1.08865537e-01 -2.54237026e-01 3.74457270e-01 5.07647932e-01
2.79414803e-01 5.51868498e-01 3.35719705e-01 -7.07148492e-01
-1.17569077e+00 3.45285922e-01 2.98105687e-01 3.36139441e-01
-5.05088091e-01 9.93453145e-01 2.99286544e-01 -7.97204852e-01
1.24616817e-01 -2.18527883e-01 3.51935387e-01 1.42206788e-01
4.84957069e-01 3.71812016e-01 6.19794652e-02 -1.03350854e+00
-5.30678391e-01 3.19675989e-02 -1.46436449e-02 -4.74135488e-01
1.15564716e+00 -3.97468835e-01 2.65740693e-01 9.61513758e-01
1.58856618e+00 5.99399626e-01 -7.92283118e-01 -5.00386596e-01
-3.24418575e-01 -2.89498180e-01 -5.70850484e-02 -2.85662264e-01
-8.27407479e-01 1.15361559e+00 2.23146245e-01 7.37588406e-01
5.97046494e-01 6.30295873e-01 4.21056777e-01 4.80092287e-01
3.57191473e-01 -7.24308670e-01 3.26759458e-01 5.28463960e-01
1.10518146e+00 -1.31410384e+00 -3.46155107e-01 -4.41021949e-01
-7.38995492e-01 7.24417686e-01 8.51228386e-02 1.30751580e-01
1.12115002e+00 5.69048703e-01 1.72799811e-01 -1.42483294e-01
-1.04754567e+00 -2.14714870e-01 1.68497026e-01 9.84992743e-01
3.64887208e-01 5.49192727e-01 6.20954216e-01 8.74125898e-01
-1.00925791e+00 -1.00492668e+00 3.68439406e-01 2.78589338e-01
-6.45273924e-02 -1.18120205e+00 -1.37761921e-01 9.34977159e-02
-3.71320285e-02 -2.47675151e-01 -5.05908608e-01 5.54262578e-01
-1.95111990e-01 1.43381453e+00 6.83733374e-02 -8.89852822e-01
1.96277708e-01 2.04755083e-01 3.94784987e-01 -5.77480793e-01
-3.68732065e-02 4.38737012e-02 3.73568892e-01 8.84944759e-03
-2.74846137e-01 -9.41297531e-01 -9.65031028e-01 -4.84057665e-01
-4.24901098e-01 1.51162773e-01 5.75102210e-01 1.06130505e+00
6.17516875e-01 5.10302246e-01 1.01202977e+00 -8.32229614e-01
-8.72465849e-01 -1.06585586e+00 -9.49747980e-01 1.87464386e-01
1.15575826e+00 -7.73639798e-01 -9.34059083e-01 -1.47959501e-01] | [14.809454917907715, 6.474897861480713] |
fc68eaac-6a1c-4cdd-a07d-fcfe3139fadb | spectral-spatial-classification-of-2 | null | null | https://doi.org/10.3390/rs9010067 | https://www.mdpi.com/2072-4292/9/1/67/pdf?version=1484303115 | Spectral–Spatial Classification of Hyperspectral Imagery with 3D Convolutional Neural Network | Recent research has shown that using spectral–spatial information can considerably improve the performance of hyperspectral image (HSI) classification. HSI data is typically presented in the format of 3D cubes. Thus, 3D spatial filtering naturally offers a simple and effective method for simultaneously extracting the spectral–spatial features within such images. In this paper, a 3D convolutional neural network (3D-CNN) framework is proposed for accurate HSI classification. The proposed method views the HSI cube data altogether without relying on any preprocessing or post-processing, extracting the deep spectral–spatial-combined features effectively. In addition, it requires fewer parameters than other deep learning-based methods. Thus, the model is lighter, less likely to over-fit, and easier to train. For comparison and validation, we test the proposed method along with three other deep learning-based HSI classification methods—namely, stacked autoencoder (SAE), deep brief network (DBN), and 2D-CNN-based methods—on three real-world HSI datasets captured by different sensors. Experimental results demonstrate that our 3D-CNN-based method outperforms these state-of-the-art methods and sets a new record. | ['Qiang Shen', 'Haokui Zhang', 'Ying Li'] | 2017-01-13 | null | null | null | remote-sensing-2017-1 | ['few-shot-image-classification'] | ['computer-vision'] | [ 2.96882659e-01 -6.41199291e-01 2.66254216e-01 -4.06105459e-01
-6.56709433e-01 -2.97462493e-01 3.42738926e-01 -8.99140313e-02
-2.33469442e-01 5.53110540e-01 7.35518634e-02 -2.96041697e-01
-4.27822709e-01 -1.24689531e+00 -5.78968227e-01 -1.16560745e+00
-2.39487574e-01 -2.03867823e-01 6.12917319e-02 -1.95273772e-01
-1.13089994e-01 8.77701759e-01 -1.92186427e+00 7.09981993e-02
1.14562392e+00 1.81506217e+00 1.60903707e-01 2.78432488e-01
-2.20554676e-02 5.04917085e-01 -2.68634081e-01 1.49008799e-02
5.87710500e-01 -7.26583302e-02 -2.77101815e-01 3.98066789e-01
5.32756388e-01 -4.93936092e-01 -6.70021117e-01 1.15575361e+00
5.87720692e-01 5.17429292e-01 7.46340036e-01 -9.61291432e-01
-9.78284657e-01 4.10069734e-01 -8.03610802e-01 -2.14504987e-01
-8.89480263e-02 -5.74060269e-02 6.16966009e-01 -1.08931625e+00
-2.11910546e-01 9.22962427e-01 1.09916747e+00 -9.16723385e-02
-8.48901570e-01 -6.91887319e-01 -1.94011740e-02 3.48718733e-01
-1.59779954e+00 -2.07127869e-01 9.45363641e-01 -4.24671501e-01
8.48215163e-01 4.04364049e-01 1.00495172e+00 7.09742963e-01
2.60524638e-03 7.66145885e-01 1.18653142e+00 -4.44748163e-01
2.36197114e-01 -2.82916367e-01 2.25853577e-01 5.25906801e-01
3.75340909e-01 3.95485878e-01 -2.15252981e-01 8.98484439e-02
6.05667412e-01 5.10405362e-01 -5.14684319e-01 -6.39005080e-02
-7.71606266e-01 8.60166788e-01 1.11664355e+00 3.73369515e-01
-6.19485676e-01 -3.91402364e-01 1.16962433e-01 -7.62354210e-02
6.81847095e-01 1.53799862e-01 -1.73417702e-01 6.66895151e-01
-9.52297330e-01 -1.10518420e-02 3.65456551e-01 7.26250768e-01
9.45234358e-01 4.05131936e-01 1.67981252e-01 8.17169070e-01
8.82886127e-02 5.98135948e-01 4.11284864e-01 -4.91391838e-01
6.46052957e-02 8.38303089e-01 -2.15208381e-02 -1.35444868e+00
-6.43868744e-01 -4.84022588e-01 -1.59617186e+00 2.61902571e-01
1.18470965e-02 -7.57597312e-02 -8.87622178e-01 1.14542842e+00
1.13195851e-01 2.63441563e-01 2.88409233e-01 1.19230497e+00
1.22421086e+00 1.00881159e+00 -1.00435548e-01 4.70737591e-02
1.02908576e+00 -7.57777274e-01 -6.05365872e-01 -2.02278674e-01
1.50902152e-01 -3.43205124e-01 8.84472728e-01 3.96434486e-01
-5.72474420e-01 -8.52313519e-01 -1.42850506e+00 -2.99483892e-02
-9.21385944e-01 3.96075845e-01 8.23416948e-01 6.27994537e-01
-8.27226698e-01 6.75486624e-01 -6.77643299e-01 -4.13853168e-01
6.17722988e-01 2.18348369e-01 -4.23167884e-01 -1.66759089e-01
-1.17805398e+00 6.22572720e-01 8.22612941e-01 5.42024374e-01
-5.31460345e-01 -6.98911309e-01 -9.06040609e-01 3.53340179e-01
3.25228453e-01 -3.33836466e-01 6.83167636e-01 -8.27609479e-01
-1.37699640e+00 4.83210027e-01 1.27705894e-02 -1.20386869e-01
-2.42297664e-01 -1.11059517e-01 -6.95459187e-01 1.92453533e-01
-3.82785559e-01 3.17061186e-01 9.39069271e-01 -1.24817932e+00
-5.60385168e-01 -7.53945291e-01 -2.74383649e-02 2.10947394e-01
-9.30159330e-01 -4.54555452e-01 1.10746175e-01 -3.82975549e-01
4.31745917e-01 -6.21159852e-01 -8.09457451e-02 1.24314047e-01
-4.60520148e-01 -1.25323385e-01 1.13994908e+00 -8.02883804e-01
7.57613361e-01 -2.47047019e+00 -1.45671293e-01 3.13009411e-01
2.01658696e-01 5.53616762e-01 -1.62610948e-01 3.19798827e-01
-2.85890028e-02 2.38292277e-01 -7.14233994e-01 2.07298562e-01
2.52606370e-03 -1.27796605e-01 -1.64846823e-01 5.31703353e-01
3.21481943e-01 6.89379275e-01 -7.01059282e-01 -3.18127394e-01
6.74319386e-01 7.51700103e-01 -1.39051706e-01 3.68449420e-01
2.35111937e-02 1.95197999e-01 -2.84499109e-01 1.08999407e+00
1.29311371e+00 -1.17173381e-01 -5.88604137e-02 -8.47366512e-01
-5.53422868e-01 -3.87967050e-01 -1.00577509e+00 1.36506820e+00
-2.90021926e-01 5.55699944e-01 -9.03410390e-02 -1.32686305e+00
1.18956101e+00 1.44831404e-01 4.67668265e-01 -6.32987440e-01
2.41205111e-01 2.31337130e-01 -3.27201545e-01 -5.63036144e-01
2.35775530e-01 3.95352766e-02 1.82428315e-01 1.15064327e-02
1.41056210e-01 -2.84363240e-01 -2.31944516e-01 -5.88010311e-01
4.27972943e-01 -7.32022896e-02 4.83963698e-01 -3.55026066e-01
6.89388692e-01 2.63093021e-02 4.20406371e-01 6.39082909e-01
-2.48447761e-01 3.46060753e-01 -1.23442762e-01 -6.18642569e-01
-8.88598442e-01 -7.97206581e-01 -3.33718002e-01 7.92114615e-01
1.49684414e-01 1.69191197e-01 -5.73237360e-01 -2.50528127e-01
2.19477028e-01 4.71690148e-01 -5.78880370e-01 -1.13153614e-01
-1.12372659e-01 -9.89427090e-01 5.57073653e-01 6.58571243e-01
1.24649847e+00 -7.42034674e-01 -5.92203975e-01 -9.61013418e-03
-1.50208727e-01 -9.47249293e-01 9.83987078e-02 2.61744082e-01
-8.37117612e-01 -1.03626060e+00 -4.85164970e-01 -7.66167641e-01
2.71499038e-01 8.96913946e-01 7.23241806e-01 -1.73495904e-01
-1.91531390e-01 -1.53752878e-01 -5.68863630e-01 -7.70310163e-01
3.65287930e-01 -1.18667632e-01 -1.29254639e-01 2.05746025e-01
6.98118448e-01 -6.79457724e-01 -7.24908531e-01 -3.48531343e-02
-1.26948774e+00 -2.29017884e-02 6.72098935e-01 9.74030674e-01
6.94490016e-01 7.44156718e-01 3.56547832e-01 -3.14625889e-01
2.56217033e-01 -3.73655885e-01 -7.28445828e-01 2.10369542e-01
-2.71330833e-01 -4.01867509e-01 9.09589767e-01 -8.35222155e-02
-1.27666986e+00 3.44065964e-01 -1.47356406e-01 -5.19916236e-01
-6.34988129e-01 1.10988402e+00 -3.93939823e-01 -5.73201776e-01
6.58372581e-01 5.24437428e-01 -2.06316281e-02 -5.12828529e-01
1.22485131e-01 1.05306542e+00 6.05014205e-01 -8.81790221e-02
1.08115506e+00 6.65882289e-01 2.00518176e-01 -1.31879508e+00
-1.04947853e+00 -4.20653790e-01 -9.17620838e-01 -3.57644856e-01
9.39408004e-01 -1.16701102e+00 -7.23784149e-01 9.13940847e-01
-7.16188908e-01 -2.06540108e-01 5.87264113e-02 5.12197971e-01
-1.03939600e-01 3.63674432e-01 -2.86069810e-01 -9.04163778e-01
-4.12431270e-01 -9.30546224e-01 1.11238468e+00 5.86226046e-01
7.04898834e-01 -9.04766083e-01 -3.90289545e-01 1.27741307e-01
6.30118191e-01 7.09930778e-01 1.01939714e+00 -4.65749145e-01
-3.93273979e-01 -3.71922493e-01 -6.50753140e-01 5.26146650e-01
3.51005971e-01 1.70744002e-01 -1.38383293e+00 -1.97125062e-01
2.48305738e-01 -4.42819834e-01 1.10017025e+00 6.54540300e-01
1.93919683e+00 -1.30838081e-01 -4.12480673e-03 1.13521385e+00
2.00743580e+00 3.61038297e-01 5.49227357e-01 3.52294594e-01
7.31301606e-01 5.17294049e-01 3.79276007e-01 4.55455899e-01
7.44980127e-02 5.81380799e-02 8.71123374e-01 -4.67544794e-01
3.50105882e-01 1.80091504e-02 -7.36820102e-02 7.21648693e-01
-3.13747823e-01 -4.48414803e-01 -7.77529001e-01 1.83791712e-01
-1.71703982e+00 -1.05883408e+00 -1.57997102e-01 1.86323762e+00
2.11571053e-01 -3.98829371e-01 -7.42885768e-02 6.94209933e-01
6.11607969e-01 4.53869939e-01 -7.66194046e-01 2.63963342e-01
-5.17223477e-01 4.73786980e-01 5.96835077e-01 1.01568229e-01
-1.72381818e+00 6.54686034e-01 5.45239115e+00 6.15383506e-01
-1.44016910e+00 -1.64440393e-01 4.27454650e-01 3.95939291e-01
1.41449288e-01 -5.67583799e-01 -2.73234487e-01 1.21532483e-02
7.44121552e-01 -5.16113825e-02 5.86309135e-01 6.94565237e-01
1.26113221e-01 -1.32751673e-01 -7.65450120e-01 1.08488333e+00
2.12543890e-01 -1.32634974e+00 7.92893693e-02 -5.31250685e-02
6.38877928e-01 -9.18049887e-02 2.50873685e-01 2.63772637e-01
-1.59241512e-01 -9.52081323e-01 5.77391922e-01 7.34583497e-01
8.24213028e-01 -9.30731416e-01 1.19997227e+00 2.62531757e-01
-1.45790124e+00 -5.47552824e-01 -7.85189271e-01 -9.48884860e-02
-3.51882696e-01 9.61608529e-01 -1.57249421e-01 1.10616136e+00
1.02093947e+00 9.91797626e-01 -6.65833771e-01 1.14415979e+00
2.13595763e-01 6.02803469e-01 -1.76980391e-01 -1.12821668e-01
6.07731879e-01 -4.63389099e-01 2.04862610e-01 1.13548434e+00
9.04661775e-01 7.42015421e-01 3.56950611e-01 7.87569106e-01
9.82256047e-03 -8.58101770e-02 -8.14133644e-01 -2.54988045e-01
3.99979740e-01 1.29411948e+00 -3.48427385e-01 -2.31700584e-01
-6.48487449e-01 6.35115266e-01 -1.24824569e-01 4.46142286e-01
-6.18490875e-01 -7.88001239e-01 5.93575835e-01 -2.87838638e-01
4.91378784e-01 -2.44251907e-01 -2.07489744e-01 -1.13280213e+00
-1.67646363e-01 -7.94687510e-01 4.12866801e-01 -1.13525331e+00
-1.59919405e+00 7.86722481e-01 -1.27677932e-01 -1.48771906e+00
2.07312778e-01 -1.21657598e+00 -4.06487763e-01 9.60944593e-01
-2.05244756e+00 -1.20209670e+00 -1.27163100e+00 8.94335091e-01
2.10874364e-01 -1.73157439e-01 1.14039361e+00 2.46734992e-01
-7.62684166e-01 2.00853437e-01 4.20996636e-01 4.84698296e-01
3.69634032e-01 -1.11812520e+00 -3.26632485e-02 9.37179208e-01
-3.27939034e-01 1.03175148e-01 1.24500290e-01 -2.43904650e-01
-1.67931449e+00 -1.56713295e+00 2.78063267e-01 2.09286317e-01
4.93014187e-01 -1.50903985e-01 -1.05468774e+00 3.80370021e-01
2.15107113e-01 1.31252900e-01 1.20561385e+00 -1.57670483e-01
-4.41126674e-01 -7.49240875e-01 -1.28938723e+00 2.28538260e-01
9.07004952e-01 -6.15922451e-01 -2.61400491e-01 4.55512822e-01
6.98399961e-01 -1.97976038e-01 -1.16582179e+00 7.44992077e-01
4.24246192e-01 -1.30777347e+00 1.06644750e+00 -1.58297256e-01
4.97576952e-01 -3.33718330e-01 -7.38169312e-01 -1.62448716e+00
-7.83222854e-01 2.17514381e-01 -1.28224388e-01 7.69828260e-01
8.21776316e-02 -7.39469588e-01 5.41054487e-01 7.00211525e-02
-6.12944961e-01 -4.18436497e-01 -6.84800923e-01 -8.23020458e-01
2.94627082e-02 -2.96967238e-01 1.23821461e+00 1.03740311e+00
-2.71257699e-01 -1.04037561e-01 -1.37449428e-01 8.54769588e-01
9.79750872e-01 3.83303136e-01 3.75406682e-01 -1.75740242e+00
3.34397554e-01 -5.62469423e-01 -4.55828458e-01 -5.88516712e-01
2.30909809e-01 -7.10354984e-01 -1.28994090e-02 -1.49176228e+00
4.69608456e-02 -1.62086189e-01 -5.46357512e-01 5.95053256e-01
-5.73008396e-02 3.33195359e-01 1.16178907e-01 5.97603954e-02
4.10987362e-02 8.94167244e-01 1.24188328e+00 -5.73435247e-01
-1.18657857e-01 -1.45473540e-01 -7.20906436e-01 5.58180809e-01
1.00976646e+00 1.07245766e-01 -9.61893797e-02 -6.41164720e-01
-2.75608778e-01 2.88375672e-02 5.78467727e-01 -1.55149209e+00
3.91685843e-01 -2.17891574e-01 1.03670406e+00 -1.06221032e+00
3.87574703e-01 -1.22191918e+00 6.44655898e-02 1.31884947e-01
6.51599914e-02 -4.27942634e-01 4.65167642e-01 3.57787788e-01
-3.73616755e-01 -3.65316868e-02 7.49938369e-01 -2.34100536e-01
-1.08449709e+00 7.07117260e-01 -3.82173479e-01 -7.88883626e-01
8.20888519e-01 -4.94332284e-01 -2.14795485e-01 -2.23772392e-01
-4.94068325e-01 -5.76027595e-02 6.00761883e-02 -1.83528867e-02
8.61883998e-01 -1.48347604e+00 -7.57667840e-01 5.53928435e-01
2.23798990e-01 1.99296311e-01 4.92330402e-01 4.21217024e-01
-5.75340509e-01 5.65708160e-01 -4.08699751e-01 -8.65504801e-01
-7.57775068e-01 5.29056966e-01 6.09882712e-01 2.19766781e-01
-4.42005247e-01 7.50857532e-01 -1.81719884e-02 -8.31537902e-01
3.87821615e-01 -2.27604508e-01 -3.50040346e-01 1.92223549e-01
5.49039245e-01 4.03566092e-01 3.22479248e-01 -6.09641790e-01
-2.95043051e-01 5.84383845e-01 6.16465092e-01 2.73018181e-01
1.83202887e+00 1.43560737e-01 -3.07955652e-01 4.50431883e-01
1.22595656e+00 -5.93751729e-01 -1.33647716e+00 -4.66611028e-01
-4.19605047e-01 -8.05399179e-01 6.76642299e-01 -7.29831278e-01
-1.29855371e+00 1.09649074e+00 9.20614541e-01 6.18066370e-01
1.77484691e+00 -5.87515831e-01 6.52703822e-01 7.07063735e-01
2.33610556e-01 -8.75551999e-01 -1.30228728e-01 4.42536473e-01
8.57017040e-01 -1.47647643e+00 1.23688392e-02 -3.54499578e-01
-2.15432912e-01 1.45525646e+00 8.77034962e-01 3.85406092e-02
1.14445555e+00 2.02723332e-02 -7.64742047e-02 -1.47002637e-01
-8.39383602e-02 -4.76661175e-01 3.91998649e-01 9.23454583e-01
2.88380146e-01 4.16063666e-01 5.84436893e-01 6.03540778e-01
-1.59771726e-01 1.75051000e-02 2.11103708e-01 8.08524847e-01
-7.40162432e-01 -2.28613600e-01 -6.30973697e-01 7.14354217e-01
8.23873207e-02 -1.16101369e-01 -4.10537392e-01 8.56913447e-01
2.02935144e-01 1.24666154e+00 2.99542576e-01 -7.06125855e-01
4.85980451e-01 4.31346968e-02 3.02607864e-01 -8.04273337e-02
-2.25369945e-01 1.06867574e-01 -4.77651060e-01 -3.00993472e-01
-8.41223419e-01 -5.09968221e-01 -7.44103134e-01 -6.49717748e-01
-6.11586213e-01 -1.78807713e-02 6.32199883e-01 8.82643819e-01
1.78937793e-01 5.50774395e-01 1.01993191e+00 -1.52037489e+00
-5.25085866e-01 -1.01928389e+00 -9.94896352e-01 3.89494859e-02
6.18296027e-01 -5.83014190e-01 -5.12712777e-01 -1.35156468e-01] | [9.91744327545166, -1.6334120035171509] |
9f587561-575d-4cd2-935b-919bcddbd5f5 | freeseg-unified-universal-and-open-vocabulary | 2303.17225 | null | https://arxiv.org/abs/2303.17225v1 | https://arxiv.org/pdf/2303.17225v1.pdf | FreeSeg: Unified, Universal and Open-Vocabulary Image Segmentation | Recently, open-vocabulary learning has emerged to accomplish segmentation for arbitrary categories of text-based descriptions, which popularizes the segmentation system to more general-purpose application scenarios. However, existing methods devote to designing specialized architectures or parameters for specific segmentation tasks. These customized design paradigms lead to fragmentation between various segmentation tasks, thus hindering the uniformity of segmentation models. Hence in this paper, we propose FreeSeg, a generic framework to accomplish Unified, Universal and Open-Vocabulary Image Segmentation. FreeSeg optimizes an all-in-one network via one-shot training and employs the same architecture and parameters to handle diverse segmentation tasks seamlessly in the inference procedure. Additionally, adaptive prompt learning facilitates the unified model to capture task-aware and category-sensitive concepts, improving model robustness in multi-task and varied scenarios. Extensive experimental results demonstrate that FreeSeg establishes new state-of-the-art results in performance and generalization on three segmentation tasks, which outperforms the best task-specific architectures by a large margin: 5.5% mIoU on semantic segmentation, 17.6% mAP on instance segmentation, 20.1% PQ on panoptic segmentation for the unseen class on COCO. | ['Xingang Wang', 'Xin Pan', 'Shilei Wen', 'Rui Wang', 'Yitong Wang', 'Xuefeng Xiao', 'Ren Yuxi', 'Ming Li', 'Pengxiang Yan', 'Jie Wu', 'Jie Qin'] | 2023-03-30 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Qin_FreeSeg_Unified_Universal_and_Open-Vocabulary_Image_Segmentation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Qin_FreeSeg_Unified_Universal_and_Open-Vocabulary_Image_Segmentation_CVPR_2023_paper.pdf | cvpr-2023-1 | ['panoptic-segmentation'] | ['computer-vision'] | [ 3.21595430e-01 -7.31254220e-02 -4.72407162e-01 -5.73715270e-01
-9.14022505e-01 -6.80967450e-01 3.95732075e-01 -1.77321345e-01
-5.71813107e-01 3.59026462e-01 -2.55453825e-01 -1.77222624e-01
-8.50761496e-03 -6.22911751e-01 -4.10634220e-01 -8.32495213e-01
2.45146185e-01 7.47780919e-01 6.06467068e-01 -1.62248164e-01
6.50084317e-02 1.08209364e-02 -1.47366190e+00 1.28158554e-01
1.21184337e+00 1.09633327e+00 5.72836101e-01 5.14900446e-01
-6.03306770e-01 5.07132523e-03 -6.65410101e-01 -5.33520222e-01
2.38538653e-01 -7.85277039e-03 -8.36974502e-01 2.15822250e-01
6.94086730e-01 -6.57043420e-03 6.17925711e-02 1.15104461e+00
4.78344351e-01 1.13782413e-01 8.63903463e-01 -9.82229173e-01
-8.61900508e-01 9.55013394e-01 -5.25169313e-01 1.56754717e-01
-2.63218731e-01 1.30727738e-01 1.20667851e+00 -6.33486390e-01
4.28580672e-01 1.12486577e+00 6.62620425e-01 6.28123701e-01
-1.13759971e+00 -6.44723177e-01 5.83551049e-01 -7.94544965e-02
-1.62538564e+00 -1.09328173e-01 4.32321191e-01 -4.46460098e-01
8.21968496e-01 4.16595995e-01 5.24744868e-01 1.13835621e+00
1.32253077e-02 1.22999263e+00 7.63869941e-01 -1.00457154e-01
8.48081112e-02 1.63024992e-01 5.73187411e-01 2.60960728e-01
3.44828904e-01 -4.49663132e-01 -8.85030627e-02 5.38389206e-01
6.17317557e-01 -1.61499560e-01 -4.13901173e-02 -3.90327752e-01
-1.08894372e+00 8.43286812e-01 5.24997652e-01 3.68806362e-01
-1.38191720e-02 -9.81330350e-02 5.98345518e-01 2.76660845e-02
9.09142971e-01 5.49554825e-01 -8.32905293e-01 2.85146624e-01
-1.11399448e+00 3.73856336e-01 7.40729153e-01 1.20904565e+00
7.87925303e-01 1.98218629e-01 -5.15465260e-01 1.18707228e+00
2.85050184e-01 5.71359217e-01 5.08998215e-01 -6.55901015e-01
4.96769994e-01 5.78164995e-01 -3.58740807e-01 -6.06294751e-01
-6.01780057e-01 -8.41038048e-01 -7.35976338e-01 -4.48690653e-01
3.75466347e-01 -1.63911715e-01 -1.62415206e+00 1.58078158e+00
2.85046816e-01 6.84560835e-02 -6.54968396e-02 8.40086997e-01
1.16500592e+00 6.40155435e-01 6.07820690e-01 1.31915078e-01
1.46708453e+00 -1.31341302e+00 -6.90440476e-01 -6.54214203e-01
5.36673546e-01 -6.32275224e-01 1.33062243e+00 1.86710566e-01
-6.50698483e-01 -8.05145085e-01 -1.01913631e+00 -1.13379866e-01
-8.59752655e-01 -1.16295777e-01 6.49364531e-01 8.34374011e-01
-8.67914796e-01 2.72875279e-01 -4.58561540e-01 -4.01205361e-01
7.50225425e-01 2.02223882e-01 1.56841636e-01 1.65472738e-02
-1.26933706e+00 6.60367906e-01 1.00073028e+00 -4.25493903e-02
-8.30353022e-01 -8.97252977e-01 -8.19817305e-01 9.35564861e-02
8.35335493e-01 -5.63418448e-01 1.24583411e+00 -8.60683143e-01
-1.34096920e+00 1.10756242e+00 1.51376560e-01 -4.96695489e-01
4.76270109e-01 -3.32064986e-01 -4.71057981e-01 -2.16333885e-02
4.29957420e-01 1.30393016e+00 8.48456562e-01 -1.17184591e+00
-7.29795992e-01 -2.84741342e-01 -2.87954882e-02 3.88452888e-01
-3.10033649e-01 -1.99174300e-01 -1.06453085e+00 -8.70988011e-01
2.15077534e-01 -7.88472295e-01 -3.98499906e-01 -3.51267993e-01
-6.00634873e-01 -4.89873260e-01 7.68180490e-01 -4.61504489e-01
1.33745861e+00 -2.07345057e+00 1.32412776e-01 -1.62671134e-01
1.82605907e-01 4.33058232e-01 -1.53343499e-01 -4.56538945e-02
2.61181563e-01 3.60509425e-01 -6.32050157e-01 -5.20095468e-01
2.48425707e-01 3.98025006e-01 -1.64386064e-01 2.15134576e-01
1.43610328e-01 1.14024997e+00 -6.40665114e-01 -9.08378899e-01
3.53482515e-01 1.22919850e-01 -4.45088178e-01 1.14715904e-01
-6.41446471e-01 4.64447021e-01 -6.85480297e-01 9.30586934e-01
7.43557513e-01 -3.88699949e-01 -2.34247193e-01 3.41483168e-02
-2.07517240e-02 -2.01507062e-01 -1.08375287e+00 2.15564466e+00
-3.09336364e-01 4.15565461e-01 1.30900458e-01 -1.11553192e+00
8.37538600e-01 1.60836697e-01 3.05607021e-01 -6.96075380e-01
3.63033146e-01 3.16493779e-01 -1.29836664e-01 -4.79342759e-01
6.79487169e-01 8.70281532e-02 -5.23276091e-01 3.61127183e-02
3.85526627e-01 -5.03766298e-01 4.75037575e-01 5.52612916e-02
1.81741193e-01 1.10759303e-01 2.10691690e-01 -6.37243569e-01
4.62454081e-01 1.22910686e-01 4.66167569e-01 1.06387532e+00
-4.77167904e-01 9.37459409e-01 8.84352624e-02 -2.52758354e-01
-8.11456859e-01 -1.11280358e+00 -7.57635653e-01 1.67447937e+00
5.06337702e-01 -1.08710885e-01 -1.16511357e+00 -7.35494256e-01
-1.22595072e-01 7.95784831e-01 -5.38871288e-01 1.66670650e-01
-3.60561222e-01 -1.10373902e+00 5.57213485e-01 5.26232898e-01
9.15036380e-01 -1.12104309e+00 -2.94468224e-01 1.82890669e-01
-2.72423506e-01 -1.24510169e+00 -4.67031747e-01 1.79927707e-01
-6.95714414e-01 -8.80403459e-01 -9.89686966e-01 -9.42285180e-01
2.09496841e-01 3.45589787e-01 1.15401530e+00 -3.43611121e-01
-3.43634188e-01 1.87069729e-01 -3.16450804e-01 -5.09268522e-01
-1.66117758e-01 7.68133938e-01 -4.37753260e-01 3.52789089e-02
5.25805950e-01 -5.50997853e-02 -4.42122459e-01 4.76194710e-01
-9.56443548e-01 1.33402660e-01 4.20048147e-01 6.34769559e-01
8.26175690e-01 -2.09793612e-01 8.26695621e-01 -1.04102135e+00
4.41938639e-01 -6.71045542e-01 -6.18535697e-01 4.69516873e-01
-6.87608421e-01 -2.84382939e-01 4.32520986e-01 -4.63242650e-01
-1.13995397e+00 -1.07804894e-01 -2.37145647e-01 -2.56598055e-01
-4.09479797e-01 4.03205663e-01 -4.07988816e-01 1.96474701e-01
5.64837277e-01 3.24453086e-01 -3.76277149e-01 -5.89554727e-01
7.49362946e-01 8.80454361e-01 5.07001460e-01 -4.82581258e-01
3.50235432e-01 2.82840580e-01 -6.11252785e-01 -8.77859771e-01
-1.33290172e+00 -8.17827642e-01 -8.27336848e-01 -4.90567880e-03
1.44760263e+00 -1.23251486e+00 1.25298006e-02 9.24067557e-01
-7.96452463e-01 -5.51345706e-01 -1.70667216e-01 5.04184514e-02
-3.77907991e-01 2.10650235e-01 -3.10799718e-01 -3.96378756e-01
-4.94893432e-01 -1.49405146e+00 1.31148744e+00 5.25403917e-01
-1.50906801e-01 -1.19671071e+00 -2.60439485e-01 6.52908504e-01
5.60236156e-01 -1.12712763e-01 6.61336422e-01 -1.17203403e+00
-6.56611681e-01 -1.62167683e-01 -4.79530841e-01 5.85774362e-01
-4.53900136e-02 -1.43916160e-01 -1.25008512e+00 -2.06430927e-01
-1.77381590e-01 -4.13576484e-01 1.14690411e+00 7.60210991e-01
1.69836605e+00 2.26581573e-01 -4.98481721e-01 1.04089344e+00
1.30639982e+00 1.22935474e-01 4.28446233e-01 2.90319979e-01
1.03325224e+00 5.18755972e-01 6.05370164e-01 -1.01875216e-01
4.85368490e-01 5.78689754e-01 3.31806660e-01 -2.33779669e-01
-2.76440889e-01 -2.40341723e-02 -2.21521243e-01 7.96097755e-01
3.78259629e-01 -5.18117905e-01 -1.08557153e+00 8.74732554e-01
-1.75206554e+00 -4.85852212e-01 -1.27515242e-01 1.75506365e+00
9.33052778e-01 3.30756664e-01 -9.77718756e-02 -3.55736494e-01
8.94061446e-01 5.63064277e-01 -8.52821708e-01 -2.07491800e-01
-1.93428338e-01 1.88769102e-01 6.89170361e-01 3.14676315e-01
-1.72500813e+00 1.64782894e+00 6.51566553e+00 1.64712310e+00
-1.15383995e+00 3.62784475e-01 8.47800076e-01 1.03656724e-01
-1.86752379e-01 -3.48607033e-01 -1.10266304e+00 4.29843038e-01
7.35450864e-01 1.67768598e-02 9.55047435e-04 9.82728362e-01
-2.15115324e-01 4.42172810e-02 -6.37088180e-01 9.44801927e-01
1.03398882e-01 -1.27156723e+00 2.51368850e-01 -2.36248031e-01
1.04780877e+00 4.23356295e-01 1.63588971e-01 7.09066749e-01
1.59546182e-01 -1.03056324e+00 6.17399216e-01 1.01907231e-01
8.34823012e-01 -4.97804672e-01 7.03966975e-01 2.22768798e-01
-1.14185536e+00 1.02888033e-01 -3.24286550e-01 2.48700857e-01
2.72577107e-01 3.60347003e-01 -6.37547135e-01 5.11177182e-01
8.26547086e-01 7.42360353e-01 -7.75076330e-01 1.02644026e+00
-1.08995095e-01 7.58804917e-01 -2.33104095e-01 2.67801303e-02
7.78067052e-01 -2.45583937e-01 5.40026665e-01 1.47606790e+00
-1.43978029e-01 -2.63397664e-01 6.26812756e-01 9.40320432e-01
-9.40936729e-02 3.14791828e-01 -3.20318639e-01 1.54159293e-01
4.46662396e-01 1.27770972e+00 -1.12259865e+00 -4.73825723e-01
-4.45246279e-01 7.16085970e-01 9.64317843e-02 5.60832560e-01
-1.01195693e+00 -4.74562526e-01 6.52621150e-01 -2.48646885e-01
5.29497445e-01 1.53667228e-02 -7.46384501e-01 -1.06185830e+00
-2.32179612e-01 -6.73849404e-01 5.10434926e-01 -3.91771853e-01
-1.29037309e+00 5.40581286e-01 3.04761261e-01 -8.29606175e-01
2.30282754e-01 -6.68263376e-01 -4.65064496e-01 7.64835835e-01
-1.60074520e+00 -1.46257269e+00 -3.01795870e-01 4.21285868e-01
1.34275985e+00 -2.38908306e-01 6.42657280e-01 2.98226506e-01
-1.00428104e+00 6.00065291e-01 1.42671064e-01 2.13296428e-01
7.76238561e-01 -1.48421335e+00 7.63222694e-01 9.45647120e-01
1.14711501e-01 3.44399720e-01 4.00901586e-01 -5.68030596e-01
-7.66078830e-01 -1.29349446e+00 6.06159747e-01 -4.22069132e-01
6.96700335e-01 -5.58799446e-01 -1.09662974e+00 6.67010367e-01
1.72534227e-01 -4.93082479e-02 6.76887989e-01 2.73279399e-01
-3.40308934e-01 -1.00155964e-01 -9.80113864e-01 7.73970485e-01
1.04759681e+00 -3.22481006e-01 -6.40583217e-01 4.05296892e-01
1.28574359e+00 -5.80714822e-01 -9.00263071e-01 5.01242220e-01
3.58299792e-01 -5.88558316e-01 1.01598847e+00 -5.67616940e-01
1.69687986e-01 -1.28599435e-01 -2.73119926e-01 -1.05976903e+00
-2.14714721e-01 -3.55686516e-01 2.06113443e-01 1.24761176e+00
7.11407423e-01 -6.33585513e-01 5.72360873e-01 4.17837679e-01
-6.38511837e-01 -6.88484311e-01 -8.93308342e-01 -6.46904886e-01
4.69796598e-01 -7.77667403e-01 6.52637184e-01 8.48560214e-01
-5.81885934e-01 3.91055286e-01 -3.20984609e-02 1.11528799e-01
5.74393034e-01 1.28600657e-01 5.88883519e-01 -1.24454355e+00
4.62504178e-02 -7.86832690e-01 -7.55549073e-02 -1.46910691e+00
2.00919166e-01 -1.09615767e+00 2.72199273e-01 -1.57992268e+00
2.12479904e-01 -6.64788485e-01 -3.88895750e-01 3.56711298e-01
-5.53992629e-01 3.56485575e-01 3.00240517e-01 1.94862336e-01
-1.04704702e+00 4.28311616e-01 1.49301696e+00 -4.48099971e-01
-1.88365072e-01 2.16751069e-01 -9.56588089e-01 6.52190149e-01
8.36223722e-01 -3.03149909e-01 -7.32318223e-01 -7.27440417e-01
-2.61686504e-01 -5.50331593e-01 1.22189023e-01 -9.93616104e-01
7.57581443e-02 -9.00926217e-02 4.51414958e-02 -7.44164169e-01
6.53340816e-02 -5.82374811e-01 -2.45828494e-01 9.23339054e-02
-2.78668284e-01 -6.46434486e-01 4.24569279e-01 6.21374428e-01
-2.06003711e-01 -4.08917427e-01 8.61679316e-01 -2.36428261e-01
-1.35378158e+00 5.80408871e-01 -6.87082410e-02 7.19595373e-01
1.09104061e+00 -4.27972198e-01 -2.89572477e-01 1.44284576e-01
-8.77224982e-01 6.81713402e-01 2.11423084e-01 7.98472404e-01
6.44967929e-02 -8.25163722e-01 -8.05163801e-01 1.87504724e-01
4.09395993e-01 6.06124401e-01 6.10767722e-01 6.56848073e-01
-4.24161673e-01 6.73684716e-01 5.99010661e-02 -1.23109460e+00
-8.43500674e-01 3.82470042e-01 4.41975385e-01 -2.94371039e-01
-5.79426944e-01 1.21352494e+00 7.01960206e-01 -7.66446888e-01
4.02799547e-01 -4.40682322e-01 -3.70538026e-01 3.98860127e-01
2.75078803e-01 1.45136550e-01 6.54050440e-04 -5.33487916e-01
-1.32751048e-01 7.69977212e-01 -3.97198558e-01 3.41631263e-01
8.98022771e-01 -3.92687708e-01 1.68651000e-01 6.31398797e-01
1.10269737e+00 -6.13581717e-01 -1.60265934e+00 -2.99761146e-01
5.11117093e-03 -7.55727291e-02 2.46611372e-01 -1.10868847e+00
-1.14906025e+00 8.67119014e-01 5.12771428e-01 3.87424201e-01
9.31021929e-01 1.05133265e-01 8.77673388e-01 1.69606864e-01
2.61480242e-01 -1.36797345e+00 -2.39879400e-01 6.38921797e-01
5.69395602e-01 -1.53306711e+00 -1.36434197e-01 -6.05305672e-01
-9.36659157e-01 7.54439056e-01 7.31981695e-01 1.01678319e-01
6.11651182e-01 -1.16821162e-01 3.42612654e-01 -1.98275462e-01
-2.39821002e-01 -5.51217675e-01 5.86743534e-01 5.95841289e-01
2.22634166e-01 4.34403092e-01 -1.97945476e-01 7.39011168e-01
-5.84220737e-02 -4.47553068e-01 -6.43338822e-03 4.17698115e-01
-6.89605832e-01 -8.32005262e-01 -1.41332313e-01 4.81339395e-01
-4.32070911e-01 -1.97054297e-01 -1.75842285e-01 1.17162907e+00
3.49666804e-01 8.47504258e-01 3.84416372e-01 4.54914570e-02
1.53145403e-01 2.96568245e-01 1.18341058e-01 -8.07531059e-01
-6.48449779e-01 2.96854734e-01 2.45322939e-02 -3.38816911e-01
-3.15006673e-01 -4.93156224e-01 -1.06271303e+00 9.89179090e-02
-5.58585525e-01 -1.61011815e-01 5.94367206e-01 1.06257439e+00
2.47252718e-01 8.14939380e-01 1.15610495e-01 -8.30498159e-01
-5.51542044e-01 -1.14232528e+00 -6.30438209e-01 3.95595998e-01
3.35533842e-02 -7.13140309e-01 -2.23920792e-01 5.94603072e-04] | [9.648784637451172, 0.6401199102401733] |
d5d10fda-4565-4f5a-8fc6-0c5e42b36cd8 | quantum-gaussian-process-regression-for | 2304.12923 | null | https://arxiv.org/abs/2304.12923v1 | https://arxiv.org/pdf/2304.12923v1.pdf | Quantum Gaussian Process Regression for Bayesian Optimization | Gaussian process regression is a well-established Bayesian machine learning method. We propose a new approach to Gaussian process regression using quantum kernels based on parameterized quantum circuits. By employing a hardware-efficient feature map and careful regularization of the Gram matrix, we demonstrate that the variance information of the resulting quantum Gaussian process can be preserved. We also show that quantum Gaussian processes can be used as a surrogate model for Bayesian optimization, a task that critically relies on the variance of the surrogate model. To demonstrate the performance of this quantum Bayesian optimization algorithm, we apply it to the hyperparameter optimization of a machine learning model which performs regression on a real-world dataset. We benchmark the quantum Bayesian optimization against its classical counterpart and show that quantum version can match its performance. | ['Marco Roth', 'Frederic Rapp'] | 2023-04-25 | null | null | null | null | ['hyperparameter-optimization'] | ['methodology'] | [ 3.01319718e-01 1.05737001e-01 1.63464829e-01 -1.52039781e-01
-1.16307724e+00 -4.62251216e-01 7.75035083e-01 9.02249664e-02
-6.57906294e-01 6.64650202e-01 -1.69781446e-01 -2.96227753e-01
-1.64464220e-01 -9.31379378e-01 -7.07317173e-01 -1.24363017e+00
3.22619021e-01 8.29142153e-01 2.47255355e-01 -7.55252466e-02
5.70416152e-01 2.92389512e-01 -8.47412109e-01 -3.12282622e-01
5.71884215e-01 7.56981730e-01 -6.21782839e-01 1.01285362e+00
4.62829798e-01 2.17512414e-01 -1.71303451e-01 -4.28215325e-01
2.87575066e-01 -2.16482744e-01 -4.43111688e-01 -6.87277853e-01
8.49776864e-02 -1.63347572e-01 -8.86860669e-01 1.47469139e+00
5.48713863e-01 2.23027498e-01 1.23250771e+00 -7.13812232e-01
-4.63854223e-01 6.52460396e-01 -1.11691870e-01 1.55339176e-02
7.86399767e-02 1.21754125e-01 1.19499290e+00 -3.14421475e-01
4.49237168e-01 1.31992674e+00 5.01865149e-01 5.20215333e-01
-2.24581385e+00 -7.77631819e-01 -8.69961560e-01 7.45847821e-02
-1.67621803e+00 -4.12268072e-01 4.69813436e-01 -3.85022730e-01
7.89584577e-01 -2.03265667e-01 2.99969673e-01 1.11079299e+00
8.45049441e-01 3.02548319e-01 1.50120175e+00 -5.59339881e-01
7.72578418e-01 2.39225343e-01 7.62928784e-01 9.72045898e-01
3.67763072e-01 8.23462069e-01 -8.36902857e-01 -6.62064910e-01
4.37914670e-01 -4.50869560e-01 -1.89205572e-01 -5.89644134e-01
-1.12540090e+00 1.18434203e+00 -2.11178660e-02 -1.88285545e-01
-1.38590202e-01 1.13789141e+00 1.02478927e-02 9.69759971e-02
2.06781596e-01 4.16531414e-01 -3.50946516e-01 -3.80544990e-01
-5.34137785e-01 3.03516477e-01 1.45218265e+00 8.82129908e-01
1.06891906e+00 -2.94385850e-01 -4.93240923e-01 -6.83670267e-02
7.29128122e-01 1.23732507e+00 -1.55373886e-01 -8.91058445e-01
-1.14441976e-01 -2.99292237e-01 3.45296979e-01 -3.28455210e-01
-2.31642962e-01 3.80045623e-02 -5.61347187e-01 1.68508515e-01
6.43971026e-01 -1.81708708e-01 -7.50026584e-01 1.59398949e+00
1.84149727e-01 1.50868118e-01 1.23200580e-01 4.20438707e-01
1.16379447e-01 6.30756915e-01 -7.13863894e-02 -2.11976528e-01
1.16681075e+00 -4.44605619e-01 -8.17339957e-01 4.59631145e-01
5.82146525e-01 -7.06260383e-01 4.94517922e-01 6.18515193e-01
-6.18487716e-01 7.32424483e-02 -1.37733424e+00 1.25600174e-01
-6.24392815e-02 -2.03624934e-01 9.55507696e-01 1.59655154e+00
-7.36900151e-01 1.31218207e+00 -1.18322301e+00 -1.14124857e-01
3.06375176e-01 7.54914761e-01 -2.41615251e-01 2.70656884e-01
-1.01056230e+00 9.73814905e-01 5.21146894e-01 2.71247234e-03
-1.14269674e+00 -4.69311029e-01 -2.42384568e-01 2.37316713e-02
1.69024616e-01 -8.17422152e-01 1.30838096e+00 3.02426275e-02
-2.38159847e+00 5.20949185e-01 -2.43492201e-01 -5.29116035e-01
1.62178442e-01 -2.46385857e-01 -2.18529746e-01 2.32204795e-01
-4.75838304e-01 -1.12214737e-01 1.48272991e+00 -4.74851936e-01
-3.10020655e-01 -5.54033220e-01 -4.22296256e-01 -3.94513696e-01
1.53412685e-01 2.52201501e-02 8.40530023e-02 2.00163946e-01
4.09962326e-01 -1.42176366e+00 -4.42218721e-01 -5.67448676e-01
-5.25940537e-01 -1.44324601e-01 2.48186439e-01 -1.20062754e-01
9.25296128e-01 -1.93321168e+00 2.62136549e-01 7.64720678e-01
2.50757992e-01 -3.82039756e-01 2.95940131e-01 3.59781206e-01
2.39175662e-01 2.47451827e-01 -1.95517421e-01 -1.32960007e-01
5.58404565e-01 8.16773772e-02 -5.00723958e-01 1.23084533e+00
1.51098341e-01 8.08912992e-01 -6.87732160e-01 -1.69641554e-01
-2.00275350e-02 4.46179867e-01 -6.17442727e-01 -6.30769208e-02
-1.11408450e-01 5.19597590e-01 -5.58805883e-01 2.64642715e-01
6.40824258e-01 -8.82374346e-02 -3.17518115e-02 -4.96220201e-01
2.08588406e-01 2.66184181e-01 -1.38603961e+00 1.39957452e+00
-8.08759034e-02 4.20889020e-01 -2.10015684e-01 -4.97419834e-01
6.63919687e-01 1.30705848e-01 1.63982347e-01 -9.55591053e-02
4.86724794e-01 2.43765011e-01 1.97771266e-01 1.82800516e-01
3.73976529e-01 -6.46722734e-01 -4.16692019e-01 7.23681927e-01
5.75607419e-01 -9.38556910e-01 -1.37799308e-01 1.95585176e-01
1.27210009e+00 3.81941170e-01 4.42919672e-01 -7.55134344e-01
4.05694216e-01 -3.81618410e-01 1.19775489e-01 1.36754429e+00
-5.01477182e-01 2.25360408e-01 7.17948437e-01 -1.97873995e-01
-1.14608705e+00 -1.62986684e+00 -5.85652292e-01 8.39845836e-01
8.14661607e-02 -7.56723404e-01 -7.22606182e-01 -3.24369043e-01
-2.54583289e-03 8.73150408e-01 -7.89399028e-01 -5.38131595e-01
-2.27207646e-01 -1.60451698e+00 7.93419123e-01 -1.63151510e-02
1.01649642e-01 -1.16004743e-01 -1.05986753e-02 2.00016230e-01
6.79429591e-01 -1.12470710e+00 -9.89098623e-02 7.96401978e-01
-8.31120074e-01 -9.58586335e-01 1.47478536e-01 1.86441705e-01
2.56429225e-01 -4.67507005e-01 5.50515413e-01 -7.52466559e-01
-8.34974647e-02 6.24087036e-01 5.85846715e-02 -4.50446278e-01
-1.09978759e+00 -3.35595291e-03 3.50609958e-01 2.25212262e-03
6.71994328e-01 -6.24391079e-01 -2.41774604e-01 -3.83398235e-02
-4.55373853e-01 -2.91076541e-01 5.10932386e-01 1.07786047e+00
4.51446027e-01 -1.50156677e-01 -1.72773644e-01 -7.87180960e-01
6.74912155e-01 1.21725820e-01 -1.30908203e+00 1.25081405e-01
-9.87757683e-01 1.12224114e+00 1.74189448e-01 -2.66969442e-01
-8.31407070e-01 2.42369696e-01 2.58598924e-01 6.81409612e-02
1.70526251e-01 1.05588719e-01 5.10649681e-02 -6.46805048e-01
9.60951149e-01 1.81617420e-02 -1.14561945e-01 -1.34362010e-02
8.58038962e-01 5.06874084e-01 3.00408989e-01 -9.35424209e-01
1.21866453e+00 8.07268202e-01 1.04692113e+00 -8.80833328e-01
-8.03755999e-01 -3.81068408e-01 -7.18459666e-01 3.27700973e-01
1.20116818e+00 -7.41534472e-01 -1.36179233e+00 9.02348310e-02
-1.08935285e+00 -1.64227858e-01 -1.01004660e-01 1.09038532e+00
-1.09136105e+00 3.48902792e-01 -8.58872294e-01 -1.33044493e+00
-2.49249324e-01 -1.21914327e+00 1.20033920e+00 2.17010900e-01
9.07957554e-02 -9.22867835e-01 4.51701760e-01 -2.47372091e-02
1.68616951e-01 -1.49978682e-01 8.64134848e-01 -9.28071201e-01
-9.91476655e-01 -4.68173385e-01 -1.63674548e-01 3.27719629e-01
-4.12359059e-01 3.22211295e-01 -1.30147588e+00 4.35285410e-03
3.61955464e-01 -2.93809660e-02 1.05748332e+00 2.31640935e-01
5.10109842e-01 4.01864827e-01 -3.17751229e-01 8.33423555e-01
1.44749868e+00 -2.89041698e-01 7.43221939e-01 2.19323263e-02
7.49024332e-01 5.80872074e-02 2.19953924e-01 3.47295582e-01
-2.02664211e-01 5.60743809e-01 -5.82191348e-02 9.81853247e-01
5.26540637e-01 -2.77857870e-01 5.99058449e-01 1.00897849e+00
-3.16282302e-01 5.72575808e-01 -7.20700562e-01 -4.17660207e-01
-1.88968670e+00 -9.79204416e-01 -6.13312960e-01 2.56032705e+00
9.76721823e-01 1.96084067e-01 -3.86037946e-01 -1.25880256e-01
4.78781521e-01 -3.39855999e-01 -1.23917617e-01 -3.95700485e-01
-1.27749786e-01 1.09915650e+00 1.19399023e+00 6.60542727e-01
-1.54535758e+00 1.10942972e+00 7.64218092e+00 1.17206836e+00
-6.06139660e-01 6.66349173e-01 -1.08790826e-02 2.57468939e-01
-6.77659363e-02 6.48272991e-01 -1.31710184e+00 2.73330837e-01
1.67612481e+00 -3.11068118e-01 6.98887527e-01 7.33152151e-01
1.76912814e-01 -5.03998578e-01 -1.53125477e+00 1.27218163e+00
-2.83154219e-01 -1.13406765e+00 -2.78333068e-01 3.14409792e-01
7.78932154e-01 1.83477744e-01 5.29517382e-02 6.00130498e-01
5.12427688e-01 -1.24195480e+00 5.64158678e-01 7.25041270e-01
2.48714343e-01 -8.14443767e-01 5.71825564e-01 1.11189574e-01
-5.01628518e-01 7.31925815e-02 -6.52588964e-01 7.88023248e-02
1.99140668e-01 6.10462129e-01 -9.15018618e-01 2.75557637e-01
2.73180723e-01 8.40358138e-02 -4.79411870e-01 8.32072616e-01
-5.08208513e-01 9.91860569e-01 -8.58552158e-01 -4.10451978e-01
3.15228477e-02 -9.08050120e-01 7.26053655e-01 1.08465958e+00
2.51791239e-01 -1.30810499e-01 -2.55499184e-01 1.32934809e+00
1.75803572e-01 1.53331652e-01 -5.28259635e-01 -4.60353583e-01
2.17096508e-02 1.11952329e+00 -6.52844012e-01 -3.26153301e-02
-6.66768849e-02 1.00325537e+00 9.14196670e-02 4.04703230e-01
-8.93607795e-01 -2.09468171e-01 4.96527106e-01 -6.50931060e-01
5.22903204e-01 -5.46917856e-01 -3.76719147e-01 -1.31450260e+00
-5.98054528e-01 -4.14945573e-01 -1.75850496e-01 -3.73702854e-01
-1.20205152e+00 -4.18114774e-02 -4.80270945e-02 -5.82999647e-01
-1.34225696e-01 -1.03144753e+00 -1.40779465e-01 1.14624476e+00
-9.82304990e-01 -9.21365380e-01 3.97321969e-01 2.48961329e-01
-7.37921596e-01 -1.85794815e-01 1.25934839e+00 -2.69918054e-01
-5.69766641e-01 3.18751574e-01 8.68350029e-01 -1.91567123e-01
9.87254560e-01 -1.56374204e+00 2.24365875e-01 6.19342029e-01
3.16364259e-01 9.82527494e-01 1.27637875e+00 -6.89260960e-01
-2.10724902e+00 -6.07373893e-01 4.61811200e-02 -9.61389363e-01
1.44054234e+00 -5.67140579e-01 -3.50455195e-01 6.48331225e-01
-2.40778580e-01 -7.58963972e-02 9.58585024e-01 3.30597997e-01
-5.63148379e-01 3.40061821e-02 -9.68167841e-01 6.33677125e-01
3.63943130e-01 -1.12070024e+00 -6.22642636e-01 5.14316797e-01
6.77393377e-01 -1.41418368e-01 -1.15150297e+00 1.68495387e-01
7.31470346e-01 -5.40600061e-01 6.86667383e-01 -6.80117607e-01
-1.56756744e-01 -4.29678440e-01 -6.30571485e-01 -1.02526402e+00
-7.77299553e-02 -1.58018374e+00 -4.82282966e-01 5.28668225e-01
4.66486812e-01 -7.10126758e-01 8.22658718e-01 7.60375321e-01
2.86313385e-01 -1.97342321e-01 -1.28987205e+00 -8.21449578e-01
4.28535670e-01 -6.79889619e-01 -3.43936943e-02 2.60109782e-01
1.97654963e-02 6.59028292e-01 -3.25960070e-01 4.48784679e-01
1.26326776e+00 -3.07997644e-01 8.41689646e-01 -1.29741192e+00
-7.26666987e-01 -4.44600731e-01 -7.85155833e-01 -8.57680857e-01
1.61881506e-01 -9.35247600e-01 2.92752951e-01 -5.36990583e-01
3.93538356e-01 -8.84487554e-02 -9.79176089e-02 -6.01654291e-01
-1.74307823e-01 9.23384875e-02 -1.51338875e-01 6.53681457e-02
-4.11004275e-01 7.80352354e-01 7.88510561e-01 -5.06297909e-02
-1.25091627e-01 2.43487731e-01 -2.82491356e-01 6.93509459e-01
6.39163315e-01 -9.96583343e-01 2.19664618e-01 5.74050069e-01
7.00268328e-01 -2.62627870e-01 2.99192518e-01 -9.80393887e-01
4.90065008e-01 1.00723550e-01 1.12922236e-01 -3.70774895e-01
6.15642130e-01 -3.32277566e-01 -7.33991340e-02 4.46390748e-01
-2.05833569e-01 -6.92465723e-01 -2.33844504e-01 1.15939641e+00
3.30564171e-01 -8.10229719e-01 9.41628635e-01 3.20653528e-01
-2.01775104e-01 3.31931919e-01 -6.03362799e-01 -2.71061152e-01
7.79257834e-01 3.31669629e-01 -3.02774519e-01 -3.87457728e-01
-8.41627896e-01 -2.90030241e-01 5.85137904e-01 -5.33827245e-01
2.07383320e-01 -9.73038912e-01 -5.15488148e-01 -6.59121852e-03
1.39589489e-01 -6.67231202e-01 -1.57094151e-01 1.32126784e+00
-6.83647156e-01 4.94093120e-01 3.34663093e-01 -6.53842330e-01
-7.25717187e-01 5.44207871e-01 4.39445794e-01 -2.51262933e-01
-1.87493593e-01 5.60043454e-01 4.68316078e-02 -2.67203093e-01
-1.61574706e-01 -4.11020368e-01 4.94333655e-01 -5.55623710e-01
6.14079773e-01 4.13184971e-01 -1.74535394e-01 -3.94436657e-01
-9.16922987e-02 5.12390673e-01 2.37615198e-01 -8.35496604e-01
9.44642723e-01 -9.03459340e-02 -5.07230639e-01 8.51937234e-01
1.10194576e+00 2.48389661e-01 -1.03770697e+00 -1.93324000e-01
3.20266664e-01 3.99284884e-02 3.11728090e-01 -8.73023868e-02
3.92935844e-03 9.99414682e-01 8.37733984e-01 -3.11071277e-02
2.59452671e-01 -3.76273878e-02 1.86817765e-01 1.34199262e+00
6.86230779e-01 -1.34675848e+00 -2.01965734e-01 7.27988601e-01
1.71102509e-01 -1.12253559e+00 2.53863782e-01 -6.24390006e-01
-6.14165738e-02 1.49553931e+00 -3.57651919e-01 -6.25009537e-01
1.18002212e+00 7.23036230e-02 -5.67653775e-01 -8.94238055e-02
-5.74987769e-01 -3.14177603e-01 2.31186569e-01 6.20846093e-01
5.19369394e-02 4.02996659e-01 -1.75664231e-01 6.33015990e-01
-2.27907315e-01 -2.49177039e-01 8.34298193e-01 6.65123284e-01
-5.54876328e-01 -1.31009376e+00 -5.84844589e-01 4.05241430e-01
-6.40787065e-01 -5.18983424e-01 8.48881528e-02 5.04110634e-01
-4.19964254e-01 8.74136150e-01 -4.63019103e-01 -3.14650953e-01
-1.22125894e-01 5.40485024e-01 1.21327472e+00 -6.74102902e-01
-2.55280137e-01 2.66813725e-01 1.55936837e-01 -6.93816245e-01
-2.21653923e-01 -7.49342859e-01 -9.89417911e-01 -4.51958507e-01
-7.20201313e-01 1.09682590e-01 1.20578587e+00 1.06855738e+00
-1.32618681e-01 1.14835419e-01 2.91523308e-01 -6.18280053e-01
-1.72742510e+00 -1.06656110e+00 -1.04600894e+00 -7.21399263e-02
-4.12601866e-02 -7.25757003e-01 -6.71082854e-01 -1.84312224e-01] | [5.637090682983398, 4.872980117797852] |
c5e3ef3c-7f1a-4d9c-9044-c8bfa1188eff | improving-human-image-synthesis-with-residual | 2205.12022 | null | https://arxiv.org/abs/2205.12022v2 | https://arxiv.org/pdf/2205.12022v2.pdf | Improving Human Image Synthesis with Residual Fast Fourier Transformation and Wasserstein Distance | With the rapid development of the Metaverse, virtual humans have emerged, and human image synthesis and editing techniques, such as pose transfer, have recently become popular. Most of the existing techniques rely on GANs, which can generate good human images even with large variants and occlusions. But from our best knowledge, the existing state-of-the-art method still has the following problems: the first is that the rendering effect of the synthetic image is not realistic, such as poor rendering of some regions. And the second is that the training of GAN is unstable and slow to converge, such as model collapse. Based on the above two problems, we propose several methods to solve them. To improve the rendering effect, we use the Residual Fast Fourier Transform Block to replace the traditional Residual Block. Then, spectral normalization and Wasserstein distance are used to improve the speed and stability of GAN training. Experiments demonstrate that the methods we offer are effective at solving the problems listed above, and we get state-of-the-art scores in LPIPS and PSNR. | ['Jing Xiao', 'Jianzong Wang', 'Shijing Si', 'Jianhan Wu'] | 2022-05-24 | null | null | null | null | ['pose-transfer'] | ['computer-vision'] | [ 2.34675571e-01 -3.63794640e-02 1.27876818e-01 4.92029414e-02
-4.67931390e-01 -1.70035020e-01 4.56327856e-01 -7.34089613e-01
-1.71981975e-02 7.95885801e-01 1.42460540e-01 7.77504593e-02
4.41003829e-01 -8.46633255e-01 -5.46596944e-01 -8.48350644e-01
4.10612881e-01 1.59899354e-01 4.81513649e-01 -3.72236699e-01
3.22836637e-02 1.28576592e-01 -1.54166377e+00 4.80764434e-02
1.27447128e+00 1.07187057e+00 1.72509953e-01 2.55449057e-01
-3.00979495e-01 7.39731908e-01 -8.31409872e-01 -4.36262399e-01
2.69033939e-01 -9.74874735e-01 -4.12995040e-01 6.41402826e-02
-1.01345368e-01 -5.97456157e-01 -4.80093420e-01 1.10382986e+00
8.05354595e-01 6.38341680e-02 4.46783513e-01 -1.32277024e+00
-8.53760362e-01 1.98290646e-01 -9.16221440e-01 -1.96332052e-01
3.40490639e-01 2.34103918e-01 4.99326885e-01 -8.77316058e-01
6.67769253e-01 1.34645641e+00 4.48037237e-01 7.63162553e-01
-9.88026679e-01 -9.86271679e-01 5.52667566e-02 1.01598591e-01
-1.44648874e+00 -2.80812621e-01 8.96024764e-01 -2.70742714e-01
3.00128639e-01 3.67030829e-01 8.59872937e-01 1.09636009e+00
1.40233696e-01 8.03897500e-01 1.18664312e+00 -4.09347385e-01
9.95203108e-02 1.33729202e-03 -7.54827321e-01 6.71941340e-01
-2.85100862e-02 1.71366751e-01 -2.88796157e-01 8.76188502e-02
1.24803257e+00 -6.93489239e-02 -3.80442917e-01 -2.40220249e-01
-1.29455912e+00 7.10551500e-01 5.15630782e-01 2.55228043e-01
-2.07359344e-01 1.20134965e-01 2.98082024e-01 4.79844818e-03
6.41720712e-01 3.88982624e-01 7.97817186e-02 -1.38709769e-01
-9.96986449e-01 2.86430120e-01 3.82670730e-01 9.49556470e-01
6.61293685e-01 2.76113570e-01 -2.34766379e-01 9.26749766e-01
2.99871087e-01 4.72542346e-01 5.17432451e-01 -1.09753895e+00
5.13594806e-01 4.62432653e-01 5.01055904e-02 -1.09206593e+00
1.48409950e-02 -3.22790980e-01 -1.19485557e+00 3.86943907e-01
1.95464313e-01 -2.40954429e-01 -1.12367022e+00 1.75012720e+00
4.05183613e-01 3.72757196e-01 -1.82077840e-01 9.73114431e-01
8.85867894e-01 9.50667918e-01 -1.70415536e-01 -3.66555214e-01
1.05268836e+00 -1.15890622e+00 -1.15815461e+00 -4.42121364e-02
9.58821997e-02 -9.69754636e-01 1.13522470e+00 3.08657140e-01
-1.13525057e+00 -6.96038246e-01 -1.15078354e+00 1.03041567e-01
-2.38496903e-02 -1.53707966e-01 5.25114238e-01 5.05325973e-01
-7.70902693e-01 4.74708259e-01 -6.90071285e-01 -1.27903134e-01
4.00264114e-01 -4.64682607e-03 -1.06242500e-01 -1.63381830e-01
-1.29263258e+00 8.58396292e-01 2.55494982e-01 2.38916099e-01
-7.40489304e-01 -6.05314016e-01 -6.57790363e-01 -5.37018031e-02
4.55426842e-01 -8.07250261e-01 9.47491407e-01 -1.32237697e+00
-1.94064379e+00 5.61839163e-01 -4.84699607e-02 1.20691270e-01
1.10891438e+00 -1.85769528e-01 -4.83789504e-01 -4.63051423e-02
2.87800916e-02 5.29023290e-01 8.30218673e-01 -1.55058086e+00
-3.47391486e-01 -8.89827460e-02 -9.23990980e-02 2.92819053e-01
-2.89531231e-01 1.05525509e-01 -8.06565106e-01 -8.70418668e-01
1.55893356e-01 -9.99196351e-01 -3.18100005e-01 8.82838964e-02
-3.16746771e-01 3.65388170e-02 1.04309130e+00 -8.63485396e-01
1.42082310e+00 -2.16778183e+00 2.60420024e-01 7.23101199e-02
2.31885567e-01 4.23769265e-01 -5.70663847e-02 3.29618365e-01
2.75000045e-03 2.83005595e-01 -2.42688790e-01 -3.38646114e-01
-2.77835101e-01 1.99363649e-01 -3.77085119e-01 1.68309242e-01
-5.28885201e-02 8.88542712e-01 -8.00638556e-01 -7.46114612e-01
2.62408882e-01 7.55192995e-01 -4.00536984e-01 4.71521825e-01
-1.32156208e-01 9.12890196e-01 -5.49474657e-01 3.28657776e-01
9.12254155e-01 -9.45215151e-02 -1.60180911e-01 -1.16885208e-01
-2.79456284e-02 -6.18178062e-02 -1.23057473e+00 1.73780549e+00
-1.49020091e-01 3.21581483e-01 -5.79688214e-02 -6.16789043e-01
8.52588177e-01 3.98941755e-01 4.22120869e-01 -7.90890217e-01
1.47224978e-01 3.16773295e-01 -6.57977462e-02 -5.61527789e-01
2.16213033e-01 -1.96328744e-01 3.84604096e-01 2.14269787e-01
-3.65749449e-01 -3.71389538e-01 4.42017289e-03 1.09508015e-01
7.58737922e-01 3.81562889e-01 7.89702311e-03 -2.75620371e-02
4.81234282e-01 -3.88087183e-01 8.78506899e-01 2.08807424e-01
6.91499887e-03 1.16045237e+00 5.61840892e-01 -3.66659611e-01
-1.25509429e+00 -1.01954138e+00 2.22684547e-01 5.17144680e-01
4.66856003e-01 -4.21083689e-01 -1.07634532e+00 -3.80909294e-01
-4.71649379e-01 4.81007904e-01 -4.54976827e-01 -1.73389092e-01
-8.10047686e-01 -5.86657166e-01 2.82140464e-01 5.71629047e-01
9.23013091e-01 -1.30532205e+00 -4.82926905e-01 2.22785056e-01
-4.17652041e-01 -1.01028383e+00 -5.14274716e-01 -6.88549757e-01
-8.05132926e-01 -8.01573515e-01 -1.28148067e+00 -6.73027217e-01
7.93662786e-01 2.36098126e-01 9.44602132e-01 4.21851873e-01
-2.32761711e-01 -3.00176471e-01 -3.81537110e-01 -4.04955596e-01
-4.00005490e-01 -1.84848264e-01 -1.42874524e-01 7.05799367e-03
-4.81765330e-01 -6.67119801e-01 -8.84514809e-01 5.47849715e-01
-1.07961941e+00 5.57815433e-01 6.01894617e-01 9.26391304e-01
7.22902954e-01 4.01705682e-01 4.87525523e-01 -7.99751878e-01
4.82392788e-01 -4.20392491e-03 -5.27367949e-01 2.86403090e-01
-5.06151319e-01 -1.26931325e-01 7.66477048e-01 -5.22996366e-01
-1.26273084e+00 -1.16164483e-01 -3.12298834e-01 -5.43546915e-01
1.45933986e-01 3.30203712e-01 -5.17646551e-01 -1.36856243e-01
3.73149395e-01 2.61257559e-01 1.57723114e-01 -2.83570319e-01
2.88684905e-01 5.40024936e-01 4.52824563e-01 -4.55296934e-01
9.92794335e-01 3.61654818e-01 -9.38457251e-03 -6.67666316e-01
-6.81402087e-01 1.01846151e-01 -1.50016591e-01 -4.19903040e-01
1.00065815e+00 -9.16121662e-01 -4.77097154e-01 8.12678397e-01
-1.19889235e+00 -3.58687848e-01 -1.63937777e-01 3.66071522e-01
-5.87192059e-01 4.65238392e-01 -7.05908239e-01 -7.11035788e-01
-2.66802818e-01 -1.46504831e+00 8.97018075e-01 4.21850979e-01
8.17602277e-02 -6.43119574e-01 -6.96830153e-02 3.16961437e-01
6.39914632e-01 6.04338527e-01 7.53011584e-01 4.17858034e-01
-6.88316166e-01 -9.67880059e-03 -1.60180166e-01 4.85205501e-01
1.02349974e-01 1.94670707e-01 -8.00341666e-01 -3.99110258e-01
8.42065960e-02 -1.54870003e-01 5.24633884e-01 2.52104491e-01
1.47985995e+00 -2.32885480e-01 -2.33331874e-01 7.99318850e-01
1.23562098e+00 4.69541878e-01 1.24069834e+00 2.32322991e-01
8.73984754e-01 3.99490416e-01 7.07494438e-01 2.09704891e-01
2.50235081e-01 7.08446681e-01 4.24075931e-01 -5.44167876e-01
-3.11910808e-01 -5.08208990e-01 1.74824417e-01 1.09806550e+00
-4.86907989e-01 -2.29734018e-01 -5.99571824e-01 1.22997068e-01
-1.92568958e+00 -8.93621862e-01 1.42975733e-01 2.15178061e+00
7.40568757e-01 6.30863681e-02 7.50179682e-03 3.49612311e-02
9.22143519e-01 3.29923183e-01 -5.93853831e-01 3.79399993e-02
-4.92849387e-02 3.17222066e-02 2.38353118e-01 2.36464486e-01
-7.09503949e-01 8.54350328e-01 6.88622808e+00 9.92970407e-01
-1.28047299e+00 8.11774954e-02 7.98714101e-01 8.65141675e-02
-5.01591086e-01 -2.72964640e-03 -3.17810208e-01 8.48165035e-01
2.67236799e-01 -2.26625279e-02 6.76970840e-01 7.71747470e-01
3.68497074e-02 5.16986847e-02 -6.52378738e-01 1.24342620e+00
2.72060245e-01 -1.10550296e+00 1.23404577e-01 1.20669827e-01
1.01151371e+00 -4.48281288e-01 2.03699663e-01 2.08453715e-01
-1.91207044e-02 -1.14508891e+00 8.55571091e-01 3.83274496e-01
1.03595352e+00 -9.74845231e-01 6.72851503e-01 3.94602895e-01
-1.17373931e+00 3.39912742e-01 -4.17655736e-01 1.75684497e-01
3.83095890e-01 6.51892126e-01 9.14389491e-02 7.24374950e-01
6.11416936e-01 5.26827157e-01 -4.22589153e-01 1.00192308e+00
-6.29501760e-01 3.38393033e-01 -1.69335559e-01 1.60271317e-01
-2.56350581e-02 -4.41096336e-01 4.14802879e-01 5.71509778e-01
7.50866115e-01 2.85696357e-01 1.31275192e-01 1.06446457e+00
-5.66247925e-02 8.13200697e-02 -4.52329606e-01 1.84800938e-01
3.31679255e-01 1.14079702e+00 -6.40249968e-01 -3.35019678e-01
-3.48594636e-01 1.14707208e+00 3.59556149e-03 4.80048805e-01
-1.21947813e+00 -4.14485097e-01 4.03194636e-01 2.76086867e-01
2.14247257e-01 -1.21533602e-01 -2.19995990e-01 -1.37451446e+00
2.75842816e-01 -1.17005670e+00 8.22859183e-02 -9.99780774e-01
-1.01857269e+00 7.21903622e-01 -1.57304481e-01 -1.43227959e+00
-2.17365608e-01 -1.73671991e-01 -8.08605194e-01 8.50492239e-01
-1.29142797e+00 -1.06365097e+00 -6.33461654e-01 5.26495159e-01
4.02382940e-01 -6.94803149e-02 6.68067873e-01 5.35846412e-01
-5.89316547e-01 5.90437889e-01 3.51956114e-02 -7.28689507e-02
8.22039664e-01 -7.33887851e-01 4.44514185e-01 9.75844383e-01
-2.95145661e-01 4.65526104e-01 7.41462827e-01 -6.15727842e-01
-1.17559052e+00 -8.45367789e-01 4.70764130e-01 -7.07447231e-02
2.27831170e-01 -4.01250273e-01 -8.62435043e-01 4.82900620e-01
2.97292948e-01 2.84407437e-02 2.77181566e-01 -3.33781242e-01
-1.40911460e-01 -1.74131453e-01 -1.08672690e+00 8.25140893e-01
1.09296310e+00 -2.93429285e-01 -2.84342796e-01 1.55617371e-01
7.50121355e-01 -7.67634988e-01 -6.66987479e-01 5.95517635e-01
4.94425505e-01 -1.14601779e+00 9.73333657e-01 -1.40737221e-01
8.40640485e-01 -5.79330385e-01 1.92731708e-01 -1.51714575e+00
-3.24173927e-01 -8.69392693e-01 -7.00744838e-02 1.46821499e+00
1.63907781e-01 -7.08312333e-01 7.59271443e-01 4.42334622e-01
1.39177144e-02 -8.63559067e-01 -8.48121524e-01 -8.10666502e-01
-3.87392640e-02 8.07811916e-02 8.97847116e-01 9.16205704e-01
-3.58741075e-01 3.86349380e-01 -8.58723044e-01 -2.79860526e-01
5.07574439e-01 1.00205459e-01 9.95768249e-01 -1.01390588e+00
-3.23775351e-01 -2.55860955e-01 -3.15069854e-01 -1.08370149e+00
-1.21530436e-01 -2.57774532e-01 1.40094414e-01 -1.69458592e+00
1.98872596e-01 -4.94484544e-01 -1.54266521e-01 2.00323328e-01
-4.77407604e-01 3.35373282e-01 3.70380819e-01 3.65099132e-01
-2.86044568e-01 8.71228158e-01 2.00873780e+00 5.27732261e-02
-7.67717659e-02 -2.29757145e-01 -6.74708545e-01 6.83670759e-01
6.36439621e-01 -2.42652163e-01 -5.31041324e-01 -5.46642661e-01
5.94366118e-02 1.18431985e-01 1.82241157e-01 -1.16323411e+00
-2.04167403e-02 -2.05962524e-01 5.83373904e-01 -4.11896855e-01
5.11134565e-01 -7.35599697e-01 6.27408981e-01 5.57428300e-01
8.10675845e-02 1.00217901e-01 -1.24030568e-01 3.11057508e-01
-4.48368043e-01 2.07202941e-01 1.13068807e+00 -1.50880963e-01
-4.41470355e-01 5.65587640e-01 7.67299756e-02 1.97505653e-01
1.10177815e+00 -1.47536129e-01 -3.37481588e-01 -7.04720736e-01
-1.62022337e-01 1.77526444e-01 6.93761408e-01 4.24014747e-01
5.90074658e-01 -1.67843390e+00 -7.79922962e-01 2.91040331e-01
-1.51501611e-01 2.61820167e-01 3.91916513e-01 6.09238744e-01
-8.14025283e-01 -1.13414459e-01 -3.76737416e-01 -4.10885513e-01
-1.03603232e+00 5.78161478e-01 1.36838675e-01 -2.62308657e-01
-8.28981638e-01 7.21133530e-01 6.69134855e-01 -2.43079022e-01
1.64423138e-01 1.81497186e-01 -3.24057341e-02 -1.95814565e-01
5.92983007e-01 3.45543027e-01 -2.89799303e-01 -5.69663882e-01
-1.60328805e-01 7.68731356e-01 1.74131706e-01 -1.64938077e-01
1.25615251e+00 -6.51725680e-02 -1.54028520e-01 1.16703726e-01
9.29041386e-01 -3.46440002e-02 -1.33137667e+00 5.67096286e-02
-8.20805669e-01 -9.04681861e-01 -1.07534006e-01 -6.43695533e-01
-1.45728970e+00 9.25522625e-01 5.65439224e-01 1.37162849e-01
1.28999174e+00 -3.65957797e-01 1.24357331e+00 -3.03173095e-01
3.47262591e-01 -1.15931034e+00 2.04542175e-01 3.18170816e-01
1.07387888e+00 -9.70015407e-01 1.71790466e-01 -7.46342242e-01
-7.43926406e-01 7.95816123e-01 9.22830045e-01 -9.83828902e-02
3.36416483e-01 1.69956446e-01 3.48564327e-01 1.47383437e-01
-3.34274143e-01 1.09759562e-01 2.25774735e-01 4.48678374e-01
3.57354581e-01 -1.42477944e-01 -6.24809861e-01 2.51035810e-01
-2.53700078e-01 -7.09422678e-02 3.27420473e-01 6.96001649e-01
-1.25839651e-01 -1.18282199e+00 -3.82200032e-01 2.63665587e-01
-5.45213163e-01 6.27740696e-02 -1.75032511e-01 7.95864642e-01
2.34709367e-01 7.86898732e-01 -1.86737776e-01 -4.79199767e-01
4.69201982e-01 -2.64160931e-01 4.55875993e-01 -2.59871066e-01
-1.93699092e-01 2.75321096e-01 -2.48995319e-01 -7.12527215e-01
-3.59857321e-01 -3.12204897e-01 -1.15430832e+00 -4.71569985e-01
-6.16566718e-01 3.64527963e-02 5.46546161e-01 6.81631446e-01
1.99273452e-01 7.89289296e-01 6.72504127e-01 -9.41837490e-01
-3.86250854e-01 -8.22746456e-01 -6.87899649e-01 5.75752139e-01
8.80865157e-02 -7.30320334e-01 -3.38369846e-01 -4.10529930e-04] | [11.418187141418457, -0.8984583616256714] |
b8a5dc4c-59da-4d9d-940c-0e63aef50f49 | one-shot-imitation-learning-via-interaction | 2306.12392 | null | https://arxiv.org/abs/2306.12392v1 | https://arxiv.org/pdf/2306.12392v1.pdf | One-shot Imitation Learning via Interaction Warping | Imitation learning of robot policies from few demonstrations is crucial in open-ended applications. We propose a new method, Interaction Warping, for learning SE(3) robotic manipulation policies from a single demonstration. We infer the 3D mesh of each object in the environment using shape warping, a technique for aligning point clouds across object instances. Then, we represent manipulation actions as keypoints on objects, which can be warped with the shape of the object. We show successful one-shot imitation learning on three simulated and real-world object re-arrangement tasks. We also demonstrate the ability of our method to predict object meshes and robot grasps in the wild. | ['Robert Platt', 'Lawson L. S. Wong', 'Jan-Willem van de Meent', 'Thomas Kipf', 'Robin Walters', 'Elise van der Pol', 'Abhinav Kumar', 'Kishore Reddy Pagidi', 'Skye Thompson', 'Ondrej Biza'] | 2023-06-21 | null | null | null | null | ['imitation-learning'] | ['methodology'] | [-7.88646340e-02 1.75269201e-01 -9.04614776e-02 -1.05807118e-01
-4.41134870e-01 -7.89201736e-01 6.15649998e-01 -2.64300793e-01
-3.46945792e-01 7.62466550e-01 -2.15887874e-01 1.28703654e-01
-9.34650376e-02 -2.87576765e-01 -1.34603906e+00 -4.74604875e-01
-2.89318234e-01 1.10277128e+00 4.53722566e-01 -2.18993038e-01
4.69331563e-01 1.07223940e+00 -1.38582611e+00 1.95005108e-02
5.31755626e-01 3.95992011e-01 6.99067473e-01 9.34030414e-01
2.35369176e-01 3.71709764e-01 -2.83928275e-01 2.08382785e-01
7.08474040e-01 1.23768888e-01 -6.00849032e-01 3.05022418e-01
9.88505036e-02 -6.31653547e-01 -6.00544631e-01 8.16219091e-01
-1.33941993e-01 4.25472468e-01 9.11447704e-01 -1.69008458e+00
-3.79471779e-01 4.73020881e-01 -5.59497476e-01 -4.44064468e-01
5.14279962e-01 7.71513999e-01 5.01979828e-01 -7.30859637e-01
1.18540466e+00 1.52382481e+00 4.13078189e-01 5.88517606e-01
-1.12631559e+00 -5.01487136e-01 -6.66363835e-02 1.25519469e-01
-8.52491617e-01 -1.20664090e-01 6.27955377e-01 -6.70573652e-01
1.01741922e+00 -1.49537325e-01 8.72497320e-01 1.19239676e+00
4.86017406e-01 6.63195312e-01 8.55667770e-01 -3.39121759e-01
2.10689679e-01 -2.49602064e-01 -3.80974501e-01 8.04387510e-01
-5.90610143e-04 4.15784657e-01 -3.56398731e-01 -2.85373896e-01
1.34374654e+00 3.60004276e-01 -4.72293310e-02 -1.36530995e+00
-1.72931361e+00 3.23329180e-01 5.86067796e-01 -9.43013057e-02
-5.74728429e-01 7.01737702e-01 5.18784784e-02 4.87722129e-01
-3.12441103e-02 8.46634269e-01 -5.27204216e-01 -3.34388852e-01
1.01652384e-01 7.99219489e-01 1.11613798e+00 1.54007530e+00
5.59857965e-01 -7.12279677e-02 1.46560118e-01 2.73441821e-01
1.84298426e-01 7.08587646e-01 -6.42627701e-02 -1.64711022e+00
2.61755079e-01 4.12020594e-01 7.47191072e-01 -3.83304179e-01
-1.42878756e-01 6.49421990e-01 -3.36137302e-02 1.06264675e+00
2.99344301e-01 -5.48527315e-02 -1.02091873e+00 1.33473086e+00
4.10911113e-01 1.76292121e-01 3.91062237e-02 6.91874027e-01
9.26487148e-02 3.60110849e-01 -1.04841560e-01 1.25376642e-01
7.43952811e-01 -1.05726480e+00 -2.41776526e-01 4.55846488e-02
3.92707258e-01 -4.44749027e-01 8.87472272e-01 3.30741465e-01
-9.76614535e-01 -4.21563864e-01 -8.48508298e-01 -5.72443642e-02
-2.41721272e-01 -2.34508008e-01 4.27611113e-01 -5.67855835e-01
-6.27937794e-01 1.25385845e+00 -1.30654800e+00 -3.97017926e-01
3.49997401e-01 5.88283062e-01 -5.94915152e-01 2.36455247e-01
-1.24150626e-01 1.49272180e+00 4.30007339e-01 -1.64545149e-01
-1.48481607e+00 -5.59144497e-01 -8.87366712e-01 -1.40964776e-01
6.00131869e-01 -6.80064142e-01 1.67729092e+00 -4.07248884e-01
-1.90580523e+00 7.53617585e-01 1.41280264e-01 -2.38411769e-01
6.26217365e-01 -5.52151620e-01 3.96293879e-01 -7.05377478e-03
-2.44292174e-03 1.01091278e+00 1.19031620e+00 -1.75712693e+00
-4.80984747e-01 -4.15674746e-01 1.49417654e-01 4.01866585e-01
4.89881188e-01 -3.75730962e-01 3.32046114e-02 -3.82668763e-01
1.45051762e-01 -1.30898380e+00 -3.36833149e-01 7.23706722e-01
-2.89542824e-01 -4.31663901e-01 1.20440412e+00 -4.91969168e-01
-2.39425302e-01 -2.09337258e+00 1.01916051e+00 5.17570302e-02
2.18323469e-02 1.01063922e-02 -2.57151961e-01 6.56868637e-01
3.21867317e-01 -3.28022957e-01 -3.11523527e-02 -1.48575366e-01
2.31244951e-01 5.82633555e-01 -5.89485705e-01 5.64555705e-01
1.99891347e-02 1.19603825e+00 -1.44112837e+00 -9.58899930e-02
4.40033466e-01 1.44091383e-01 -3.43877405e-01 6.73689425e-01
-8.01199198e-01 9.37797487e-01 -5.65583348e-01 6.24756098e-01
2.26737320e-01 1.72181040e-01 1.79459929e-01 -7.20025972e-03
-2.75172234e-01 -2.04662666e-01 -7.69314528e-01 2.06127524e+00
-4.28214520e-01 4.10629541e-01 2.67317116e-01 -7.94896483e-01
8.43912601e-01 2.60123909e-01 5.24702966e-01 1.50801644e-01
1.37457505e-01 3.59859586e-01 1.81003362e-02 -9.40737247e-01
3.93471450e-01 9.54166874e-02 -1.40709281e-01 3.99106652e-01
2.28910312e-01 -1.36745656e+00 -1.70120284e-01 -1.29517972e-01
1.02140200e+00 9.70651507e-01 3.42132837e-01 4.75322604e-02
-2.31345564e-01 3.24764520e-01 -1.04452960e-01 6.49197280e-01
4.50669676e-02 2.53009319e-01 1.68162465e-01 -4.13764089e-01
-1.69806850e+00 -1.52997100e+00 3.97442192e-01 7.92718709e-01
5.04537582e-01 1.28148422e-01 -3.47204655e-01 -5.65109134e-01
7.59408951e-01 7.20296800e-01 -5.22379160e-01 7.43362531e-02
-1.04966021e+00 5.35511672e-01 4.11432721e-02 7.11918414e-01
-6.51364252e-02 -1.65343928e+00 -1.32341754e+00 8.22618976e-02
3.06920320e-01 -1.03150785e+00 -4.19044286e-01 7.38638565e-02
-9.59037304e-01 -1.39641345e+00 -6.81162059e-01 -1.05544043e+00
8.26539993e-01 3.59334379e-01 4.96801108e-01 -3.95700037e-02
-4.07702297e-01 8.42382312e-01 -3.63918126e-01 -5.16690016e-01
-8.26044559e-01 -2.85420388e-01 4.28943038e-01 -7.41554379e-01
-2.40279034e-01 -8.77782285e-01 -2.25486800e-01 2.45579883e-01
-2.55393982e-01 1.16403423e-01 4.99764055e-01 7.28887379e-01
5.88544786e-01 -5.48538744e-01 1.25050157e-01 -1.61221270e-02
4.52858478e-01 -1.83975637e-01 -8.62353921e-01 2.48847038e-01
1.56704098e-01 1.10161543e-01 3.01978827e-01 -9.45683181e-01
-6.54977322e-01 4.01500702e-01 4.90707844e-01 -1.00276387e+00
-3.09844494e-01 -1.78478807e-02 3.23199332e-01 -2.57050008e-01
5.20678341e-01 -2.12737415e-02 4.84476984e-01 -5.56109607e-01
7.72352099e-01 4.64577615e-01 8.99147153e-01 -9.62233841e-01
8.83398354e-01 5.19777656e-01 1.66767076e-01 -6.32246494e-01
-1.40662372e-01 -3.19516748e-01 -1.23922944e+00 -2.27898821e-01
8.87710035e-01 -5.26899576e-01 -1.31518781e+00 5.59871912e-01
-1.36434627e+00 -1.01132238e+00 -6.97975516e-01 5.61153054e-01
-1.42009282e+00 2.09100753e-01 -3.91290694e-01 -6.63328946e-01
-4.43517119e-02 -1.28551972e+00 1.31686509e+00 1.17479995e-01
-3.88203800e-01 -4.08517689e-01 1.08245522e-01 -3.56487781e-01
-1.35791242e-01 5.47424912e-01 7.52852261e-01 -4.97218192e-01
-7.93179691e-01 4.37621549e-02 6.72310516e-02 -1.30094364e-01
3.19767714e-01 8.69905204e-02 -3.14385831e-01 -5.81513882e-01
-7.45923668e-02 -6.61437809e-01 1.67353123e-01 1.73127592e-01
1.08726037e+00 -3.23668867e-02 -8.58923793e-01 2.32848465e-01
9.80664194e-01 3.63129705e-01 2.66232967e-01 3.21883440e-01
8.29158664e-01 5.88103116e-01 1.07454205e+00 3.75577718e-01
5.84134320e-03 9.45405602e-01 6.94729328e-01 4.61502343e-01
-2.00193096e-02 -5.17439365e-01 2.72590578e-01 3.78047824e-01
-2.75241733e-01 1.51931986e-01 -8.88034582e-01 5.05473435e-01
-2.10044146e+00 -6.39393270e-01 2.39301145e-01 1.93960667e+00
4.60590392e-01 -1.22868363e-02 -2.76848655e-02 -5.88758230e-01
4.70989168e-01 -3.37355077e-01 -1.17856109e+00 -1.20519541e-01
6.74235404e-01 9.93643105e-02 5.09248912e-01 6.61867440e-01
-6.55702472e-01 1.19059265e+00 6.43100595e+00 6.62998632e-02
-8.24428797e-01 -1.42551988e-01 -5.38271189e-01 7.31269130e-03
2.61489272e-01 1.51272833e-01 -2.98495948e-01 1.31810591e-01
8.76674652e-02 -2.42739514e-01 8.42185020e-01 1.04756033e+00
-4.93006548e-03 -2.01790467e-01 -1.81511259e+00 7.48644412e-01
-1.86218709e-01 -1.21737111e+00 -2.15964347e-01 -6.97388947e-02
6.24262333e-01 2.51701415e-01 -1.90757915e-01 4.31282580e-01
8.82523000e-01 -6.71061039e-01 8.48626614e-01 6.56790018e-01
7.51309216e-01 -2.58467436e-01 6.42825887e-02 8.50827634e-01
-9.36821938e-01 -3.22306633e-01 -2.34488115e-01 -2.72228420e-01
5.25114179e-01 -6.24925971e-01 -1.27314401e+00 2.36580402e-01
6.38173282e-01 7.12280154e-01 2.60476232e-01 9.29838300e-01
-4.36072916e-01 -3.27855378e-01 -4.68805045e-01 -1.86126426e-01
7.14690238e-02 -3.21574152e-01 1.03120244e+00 5.00824451e-01
2.94763535e-01 2.96676517e-01 4.89559352e-01 9.81523156e-01
-4.65621334e-03 -7.09666848e-01 -1.15026832e+00 4.91812639e-02
5.67737103e-01 8.50034654e-01 -6.30042255e-01 -4.31167006e-01
2.70770758e-01 9.16244090e-01 5.56952000e-01 3.41807127e-01
-7.23435700e-01 -2.07648575e-01 7.07346320e-01 -2.63622224e-01
6.12122059e-01 -9.17542577e-01 1.37168989e-01 -9.75948989e-01
7.97752365e-02 -7.67830789e-01 -4.46671367e-01 -1.25170326e+00
-1.03119302e+00 -7.82638602e-03 5.59402704e-01 -1.44318414e+00
-4.17118222e-01 -8.02545488e-01 -5.12761354e-01 3.96393895e-01
-9.11361992e-01 -1.15666950e+00 -3.80886018e-01 2.61162937e-01
1.06655407e+00 -6.50078580e-02 7.89196789e-01 -7.51853347e-01
2.63026476e-01 -1.21546775e-01 1.83348820e-01 -1.61360279e-01
5.12139916e-01 -1.17058671e+00 6.93599403e-01 -2.33286414e-02
9.99290869e-02 5.12678444e-01 8.60579848e-01 -1.00691700e+00
-1.80638015e+00 -6.27873838e-01 -7.35537708e-02 -7.19797671e-01
8.12996924e-01 -5.29217005e-01 -9.40860391e-01 1.15996981e+00
4.69788201e-02 -3.35304514e-02 -4.48159486e-01 -3.58179271e-01
-1.52686208e-01 3.71399522e-01 -1.33869755e+00 8.57141852e-01
1.31139445e+00 -2.48209059e-01 -1.19025552e+00 6.30060554e-01
9.58320737e-01 -1.05742168e+00 -8.34434748e-01 4.94961292e-01
9.93474603e-01 -1.49281681e-01 9.96743381e-01 -9.68446076e-01
3.86237681e-01 -1.36492506e-01 -4.49106507e-02 -1.73825824e+00
-1.01518571e-01 -7.81662583e-01 -3.95698816e-01 3.51619542e-01
-6.33247048e-02 -4.80808973e-01 7.53209174e-01 3.59756887e-01
-1.51784167e-01 -5.98387003e-01 -8.68477166e-01 -1.08505762e+00
8.55421349e-02 1.83926627e-01 4.69834566e-01 5.90334773e-01
3.41190100e-01 -1.33757785e-01 -1.97931036e-01 3.37342829e-01
5.99548757e-01 3.49844217e-01 1.43279123e+00 -1.08108974e+00
-4.08812851e-01 -2.53254563e-01 -2.36066312e-01 -1.14690435e+00
7.49413967e-01 -7.37206757e-01 6.95966363e-01 -1.34882319e+00
8.20128713e-03 -5.64667881e-01 4.33946192e-01 5.83956003e-01
2.97893673e-01 -3.16843182e-01 6.10835791e-01 5.32601535e-01
-2.66424894e-01 6.70495212e-01 1.74997973e+00 -1.04566246e-01
-3.41514349e-01 -5.71360737e-02 5.79065859e-01 9.97699618e-01
7.77071416e-01 -5.21294773e-01 -1.94557756e-01 -6.89888239e-01
-3.48055243e-01 4.58248556e-01 7.90516376e-01 -8.87389660e-01
2.19738841e-01 -6.32759869e-01 1.08134508e-01 -5.66474676e-01
7.75502145e-01 -1.15868485e+00 2.45743811e-01 8.87098312e-01
-3.81638288e-01 1.88449681e-01 2.61828661e-01 7.62800097e-01
5.16290486e-01 -3.09660852e-01 3.75360578e-01 -4.25027400e-01
-7.65526175e-01 3.36394042e-01 -1.09106243e-01 -4.07460451e-01
1.49768651e+00 -1.22774981e-01 -7.64504746e-02 -5.39579056e-02
-8.83773327e-01 5.23599088e-01 1.05648327e+00 7.36630857e-01
8.07719409e-01 -1.00025964e+00 -4.60377663e-01 1.49857014e-01
-3.27477269e-02 4.54707831e-01 -2.94552058e-01 4.23341811e-01
-6.48305893e-01 -1.19785674e-01 -6.43981278e-01 -8.61424863e-01
-1.27219677e+00 6.98251963e-01 2.26738453e-01 1.94495231e-01
-1.20350599e+00 6.00929499e-01 -2.30205245e-03 -8.05809498e-01
2.50408143e-01 -5.33206880e-01 2.83121914e-01 -7.63142049e-01
-7.22569600e-03 4.42699701e-01 -6.77601874e-01 -2.42136031e-01
9.02856812e-02 7.11045086e-01 -2.60921493e-02 -4.39936191e-01
1.68605900e+00 2.98795789e-01 -1.89472586e-01 7.99859643e-01
8.58275414e-01 -3.70463431e-01 -2.07207346e+00 -5.59886284e-02
-1.35703878e-02 -6.86239243e-01 -8.65718305e-01 -4.88125563e-01
-2.12436274e-01 5.84239304e-01 2.54341006e-01 -1.82853460e-01
1.04049891e-01 3.57695252e-01 6.14111125e-01 1.12411702e+00
1.11390674e+00 -9.03036118e-01 6.01886690e-01 6.43719614e-01
1.60499883e+00 -1.10923862e+00 5.75403646e-02 -2.97200620e-01
-5.28402507e-01 1.20618725e+00 6.68194711e-01 -8.39811981e-01
5.11416674e-01 5.86435974e-01 -2.72175401e-01 -3.32876593e-01
-4.19691473e-01 3.64953578e-01 2.03219242e-02 8.85599792e-01
-5.93076825e-01 2.53726780e-01 3.16748053e-01 -3.51214170e-01
-1.45410880e-01 2.11973786e-01 4.48017120e-01 1.51005864e+00
-7.54017115e-01 -9.32507038e-01 -2.53105789e-01 1.81145728e-01
4.10816252e-01 6.48756206e-01 -3.71545970e-01 1.05894935e+00
-1.41352311e-01 2.93782473e-01 1.48230180e-01 -2.77325392e-01
5.77750385e-01 -8.75177532e-02 1.36205137e+00 -1.01110148e+00
-2.67480463e-01 -3.46566260e-01 -2.79866278e-01 -7.54043043e-01
-2.99939275e-01 -7.51701832e-01 -1.56350994e+00 4.74031866e-02
-1.39654607e-01 -2.12617576e-01 8.66794109e-01 1.02056277e+00
2.47784108e-01 2.96477020e-01 4.79770094e-01 -1.92877960e+00
-1.18350017e+00 -1.03107357e+00 -4.15485501e-01 5.24842680e-01
6.41700089e-01 -1.12576389e+00 -1.89836025e-01 1.03427954e-01] | [4.733774662017822, 0.5542905926704407] |
28c53dc0-e4fa-4002-b6c8-ab1d8745be1d | negvsr-augmenting-negatives-for-generalized | 2305.14669 | null | https://arxiv.org/abs/2305.14669v1 | https://arxiv.org/pdf/2305.14669v1.pdf | NegVSR: Augmenting Negatives for Generalized Noise Modeling in Real-World Video Super-Resolution | The capability of video super-resolution (VSR) to synthesize high-resolution (HR) video from ideal datasets has been demonstrated in many works. However, applying the VSR model to real-world video with unknown and complex degradation remains a challenging task. First, existing degradation metrics in most VSR methods are not able to effectively simulate real-world noise and blur. On the contrary, simple combinations of classical degradation are used for real-world noise modeling, which led to the VSR model often being violated by out-of-distribution noise. Second, many SR models focus on noise simulation and transfer. Nevertheless, the sampled noise is monotonous and limited. To address the aforementioned problems, we propose a Negatives augmentation strategy for generalized noise modeling in Video Super-Resolution (NegVSR) task. Specifically, we first propose sequential noise generation toward real-world data to extract practical noise sequences. Then, the degeneration domain is widely expanded by negative augmentation to build up various yet challenging real-world noise sets. We further propose the augmented negative guidance loss to learn robust features among augmented negatives effectively. Extensive experiments on real-world datasets (e.g., VideoLQ and FLIR) show that our method outperforms state-of-the-art methods with clear margins, especially in visual quality. | ['Yukai Shi', 'Yuming Fan', 'Zhijing Yang', 'Xiaoyu Xian', 'Meilin Wang', 'Yexing Song'] | 2023-05-24 | null | null | null | null | ['video-super-resolution'] | ['computer-vision'] | [ 3.75568390e-01 -6.53615773e-01 9.00224224e-02 -1.38555542e-02
-8.87908161e-01 -2.04766750e-01 2.67164409e-01 -7.76952326e-01
-5.97647764e-02 9.05489981e-01 2.05862030e-01 1.40527681e-01
-8.86346176e-02 -5.37272155e-01 -5.59041560e-01 -7.75835812e-01
2.70843655e-01 -3.27453524e-01 3.65976721e-01 -5.67797065e-01
-8.94821063e-02 3.86790276e-01 -1.55305994e+00 2.89308548e-01
1.16280711e+00 9.03333724e-01 6.42341435e-01 4.64090556e-01
1.11151814e-01 8.87394130e-01 -6.34261906e-01 -1.20216593e-01
5.21503985e-01 -5.20239234e-01 -1.19563252e-01 2.64482588e-01
3.22191119e-01 -4.72595066e-01 -8.42509091e-01 1.55301809e+00
7.28312075e-01 2.48170406e-01 1.79452509e-01 -9.05721128e-01
-9.71304059e-01 3.60468119e-01 -9.90293026e-01 6.11602902e-01
5.46868801e-01 4.46133554e-01 6.18311942e-01 -1.04269481e+00
5.28389812e-01 1.54389691e+00 5.73968291e-01 6.49600208e-01
-1.20270586e+00 -8.55316579e-01 1.83417141e-01 3.37611347e-01
-1.48577106e+00 -4.79963809e-01 9.42569733e-01 -1.95918486e-01
1.73024818e-01 4.05595303e-01 4.91081119e-01 1.40935266e+00
-1.35794565e-01 6.05319262e-01 1.51178539e+00 1.06372483e-01
2.71718539e-02 -4.18417268e-02 -2.19083980e-01 1.50667429e-01
3.06883901e-01 2.41243929e-01 -5.42002082e-01 9.32560936e-02
1.12046039e+00 -2.21068576e-01 -8.56259644e-01 4.33188789e-02
-1.14150763e+00 2.62871057e-01 1.84289694e-01 1.71640828e-01
-3.20298761e-01 -1.86733305e-01 3.72698396e-01 3.05033177e-01
5.76550305e-01 1.64409384e-01 -2.61210144e-01 -1.32799968e-01
-8.03708553e-01 2.45884538e-01 1.48257092e-01 1.03267300e+00
5.21843135e-01 5.19558787e-01 -4.16884810e-01 1.26510072e+00
-9.11686849e-03 5.64049423e-01 3.84185731e-01 -1.17242980e+00
6.08050108e-01 -2.75033223e-03 4.47028279e-01 -1.07440281e+00
-1.24079607e-01 -8.39449108e-01 -1.40941513e+00 1.97283030e-01
2.01057345e-01 9.65054613e-03 -8.15939963e-01 1.63264310e+00
2.13106319e-01 7.56859481e-01 1.12079002e-01 1.36810386e+00
8.83352637e-01 6.96461022e-01 -2.11470187e-01 -8.34088206e-01
1.05370057e+00 -7.34254718e-01 -1.16153717e+00 -6.56007081e-02
-1.89977616e-01 -7.22562611e-01 1.32045484e+00 7.72849739e-01
-1.13341224e+00 -8.83841693e-01 -9.61745739e-01 -1.72949024e-02
3.57394964e-01 1.40541360e-01 1.65033728e-01 4.62613761e-01
-9.28528428e-01 4.97872144e-01 -5.99249482e-01 9.51876268e-02
5.53703129e-01 -2.15984523e-01 -1.17727727e-01 -5.70606411e-01
-1.44923162e+00 4.43156928e-01 1.21437557e-01 5.52663147e-01
-9.39109147e-01 -7.34977901e-01 -7.03091800e-01 -2.83547878e-01
8.86496603e-01 -5.92816174e-01 7.98007965e-01 -9.14897978e-01
-1.57684219e+00 3.38218033e-01 -1.11484133e-01 -1.36199906e-01
8.88188183e-01 -4.64038014e-01 -8.76188517e-01 1.15580313e-01
-2.46673096e-02 -1.52507827e-01 1.16345799e+00 -1.76784933e+00
-4.13735330e-01 -1.88168898e-01 1.03919409e-01 3.38504076e-01
-2.80710250e-01 1.58861831e-01 -5.85653424e-01 -1.14718390e+00
3.40344936e-01 -5.19084036e-01 -2.98572391e-01 -1.27220368e-02
-3.10531348e-01 2.52558023e-01 7.60127664e-01 -1.02312100e+00
1.43024945e+00 -2.22806215e+00 2.15500012e-01 -1.39862433e-01
3.73356521e-01 5.59005022e-01 -3.19848835e-01 -9.14365128e-02
-2.85616606e-01 1.49201050e-01 -1.47929132e-01 -5.96094206e-02
-4.58351374e-01 3.73699740e-02 -3.60000312e-01 3.91530365e-01
2.37054393e-01 3.77289444e-01 -1.14364076e+00 -4.62765127e-01
2.08910808e-01 7.04855263e-01 -3.44762355e-01 3.27302784e-01
-2.06243936e-02 8.18910956e-01 -3.88129830e-01 5.47307551e-01
1.06165373e+00 -1.10336475e-01 -1.83934957e-01 -7.29657233e-01
3.33184451e-02 -3.23331028e-01 -1.52175128e+00 1.46048081e+00
-4.24836695e-01 4.07732397e-01 2.47573107e-01 -6.08572602e-01
9.07074451e-01 1.40177533e-01 2.96005070e-01 -7.97640264e-01
-1.00275703e-01 3.14405829e-01 -1.52089536e-01 -8.31014991e-01
4.14370775e-01 -2.20557049e-01 6.01542711e-01 -2.19033450e-01
-2.85611778e-01 4.71493155e-02 5.58704510e-02 5.45608178e-02
1.09108138e+00 2.14347795e-01 1.36720046e-01 1.32271215e-01
8.71945024e-01 -5.81428826e-01 1.11310613e+00 5.14606893e-01
-4.34001952e-01 1.21523154e+00 2.70130575e-01 3.44523229e-02
-1.14418173e+00 -1.15048933e+00 7.50750601e-02 6.10592902e-01
4.74936694e-01 -2.71972835e-01 -7.30441093e-01 -2.88222164e-01
-5.70166111e-01 4.67791110e-01 -3.01907361e-01 -1.06141614e-02
-8.12546372e-01 -8.97497952e-01 2.66226977e-01 3.11566830e-01
8.69723499e-01 -7.93496728e-01 -3.53002101e-02 2.93694027e-02
-6.67188823e-01 -1.55581915e+00 -5.20379901e-01 -6.09818399e-01
-4.94325906e-01 -9.91624892e-01 -8.90067875e-01 -3.83535236e-01
3.83564949e-01 7.27906108e-01 1.03820932e+00 6.90544471e-02
-2.33238459e-01 1.65370792e-01 -4.92432207e-01 2.47133419e-01
-5.28845966e-01 -5.91773808e-01 2.74215639e-01 4.04302359e-01
-9.32981595e-02 -7.36091316e-01 -8.54178667e-01 4.69561070e-01
-1.21826828e+00 9.06517357e-02 5.98191381e-01 1.04398775e+00
7.85202801e-01 6.46533310e-01 7.59332001e-01 -5.96249223e-01
6.59316361e-01 -4.51260716e-01 -5.14253974e-01 9.80841666e-02
-2.48904586e-01 -3.23261648e-01 1.05092812e+00 -9.20082510e-01
-1.44310319e+00 -4.38774705e-01 -1.53481051e-01 -1.05776262e+00
-4.61908579e-02 2.32330859e-01 -6.24702752e-01 3.29150967e-02
7.16606140e-01 5.61412752e-01 -1.28109813e-01 -5.72350621e-01
1.87769100e-01 6.65146887e-01 7.09397376e-01 -4.85402495e-01
1.38554585e+00 5.86603224e-01 1.03456534e-01 -1.02073359e+00
-8.66110146e-01 -2.11616889e-01 -6.47997484e-02 -2.08108261e-01
5.99802434e-01 -1.24140775e+00 -3.31889063e-01 7.65272379e-01
-9.03550446e-01 -1.68754339e-01 -2.68355817e-01 4.19068426e-01
-6.21873617e-01 8.08442175e-01 -7.30672479e-01 -1.03229046e+00
-8.68117139e-02 -1.32575595e+00 8.72480989e-01 2.05993146e-01
5.06998360e-01 -3.77675325e-01 -2.78839856e-01 4.55121428e-01
6.31182313e-01 -1.65993404e-02 4.98391002e-01 8.06109905e-02
-8.42649817e-01 2.81481504e-01 -5.54784536e-01 8.83901238e-01
3.13572645e-01 -8.17303360e-02 -7.93654323e-01 -4.69690323e-01
4.28626090e-01 -1.97603822e-01 8.84230793e-01 3.35715055e-01
1.46292293e+00 -8.84307772e-02 1.00528814e-01 7.68335342e-01
1.53187490e+00 3.96164134e-02 1.01195610e+00 3.66112620e-01
7.63326287e-01 3.88561666e-01 1.09251249e+00 5.67983449e-01
-6.18015528e-02 8.63455951e-01 3.05573881e-01 -7.16750324e-02
-5.15422583e-01 -8.11894983e-02 5.57752311e-01 6.96640134e-01
-2.63295501e-01 -1.56648159e-01 -4.27349716e-01 2.52588481e-01
-1.59180558e+00 -9.69427228e-01 -5.30343950e-02 2.13301802e+00
9.11752701e-01 1.97205707e-01 -7.16698393e-02 1.52227446e-01
9.89834428e-01 4.63199377e-01 -6.44170225e-01 5.97945035e-01
-6.29974008e-01 -1.55782998e-01 2.54585952e-01 2.03689694e-01
-9.13891912e-01 7.04565525e-01 4.98112059e+00 1.42177451e+00
-9.89081919e-01 2.16186717e-01 7.24461257e-01 -1.11737132e-01
-4.17066932e-01 -3.18023413e-01 -5.40891588e-01 6.87101901e-01
4.02008176e-01 -1.93100065e-01 8.91696513e-01 4.26827431e-01
7.66552329e-01 2.14781165e-01 -5.97685993e-01 1.51538527e+00
6.36974871e-02 -8.07209909e-01 2.05472246e-01 -3.37155104e-01
7.87003458e-01 -2.71364897e-01 2.87936062e-01 2.87453830e-01
2.33093891e-02 -9.15346444e-01 4.82650995e-01 7.10119963e-01
1.24394906e+00 -5.63196540e-01 6.67231381e-01 2.61242449e-01
-1.20401788e+00 -3.61041903e-01 -5.54535627e-01 3.30547065e-01
3.71719956e-01 8.09315920e-01 6.70395941e-02 9.43895280e-01
9.61855173e-01 8.41682971e-01 -5.73152900e-01 1.02733004e+00
-4.56803516e-02 6.51583016e-01 -6.08416311e-02 6.62070274e-01
-2.76505619e-01 -4.05969441e-01 1.03973269e+00 9.04960930e-01
4.72864419e-01 4.79191482e-01 1.13058500e-01 8.41947198e-01
-4.51800562e-02 -2.65027713e-02 -3.87733728e-01 2.55902827e-01
4.36759561e-01 1.20677042e+00 -2.13882908e-01 -7.61652216e-02
-5.60853958e-01 1.02910805e+00 -7.40400180e-02 8.90835583e-01
-9.95353401e-01 -7.64959231e-02 8.93262208e-01 3.23328942e-01
1.59119695e-01 -8.35122839e-02 2.84035399e-04 -1.59555280e+00
4.15257037e-01 -1.35491323e+00 -2.12900098e-02 -1.05010557e+00
-1.60601389e+00 8.03578138e-01 -1.59006223e-01 -1.76055062e+00
2.47068197e-01 -2.52621174e-01 -2.97499716e-01 8.61145735e-01
-1.79689467e+00 -8.65579665e-01 -7.53660858e-01 6.46718085e-01
9.09516931e-01 -2.38479838e-01 3.58222090e-02 6.24122202e-01
-7.88411975e-01 4.71918344e-01 1.55747026e-01 -1.27981156e-02
7.19988346e-01 -8.56052577e-01 2.10232988e-01 1.16142607e+00
-4.30074990e-01 4.11423624e-01 1.07979095e+00 -7.79428124e-01
-1.37896609e+00 -1.39218438e+00 -2.56054312e-01 -5.48999235e-02
7.95590103e-01 -2.41421580e-01 -1.41345608e+00 3.40359747e-01
-1.38138339e-01 4.91937906e-01 -3.20066921e-02 -5.18364072e-01
-3.84151995e-01 -4.49661046e-01 -1.17321241e+00 9.18709695e-01
1.51054430e+00 -3.54833096e-01 -2.07572743e-01 1.37368575e-01
1.23181057e+00 -5.46254694e-01 -7.92664468e-01 7.88273692e-01
1.60733223e-01 -1.06923532e+00 1.23438191e+00 -4.44002360e-01
6.10419452e-01 -7.17039406e-01 -3.86457115e-01 -1.33422661e+00
-2.84680426e-01 -8.75122488e-01 -3.36262763e-01 1.40083194e+00
-9.49078873e-02 -5.85559905e-01 4.05759364e-01 1.90395325e-01
-8.66554976e-02 -6.58840239e-01 -7.40107954e-01 -9.93577659e-01
-3.86739284e-01 -4.69720542e-01 4.18640733e-01 9.45293069e-01
-7.44216979e-01 1.92041010e-01 -7.90224671e-01 4.67825294e-01
1.09077978e+00 -3.15341324e-01 6.49261713e-01 -8.59030366e-01
-4.04425293e-01 -1.26466915e-01 -3.64085108e-01 -1.17443120e+00
3.18599418e-02 -1.34737536e-01 5.92900813e-02 -1.33843052e+00
3.03400666e-01 -4.51536000e-01 -4.01749700e-01 -3.53704244e-01
-6.00245893e-01 2.51406163e-01 1.71258107e-01 2.91931093e-01
-5.38823247e-01 9.03096914e-01 1.68092525e+00 -3.31129530e-03
-1.15225226e-01 -5.33601008e-02 -6.37298465e-01 7.19626963e-01
5.45563102e-01 -2.61490308e-02 -6.21855080e-01 -2.81393588e-01
2.84770727e-01 4.13375050e-01 4.44050312e-01 -1.09929526e+00
-1.20449327e-01 -3.76963407e-01 3.12020600e-01 -4.87939119e-01
4.87288624e-01 -7.66732275e-01 2.46979281e-01 -7.71995038e-02
-1.59232661e-01 -2.42164642e-01 -1.84037760e-02 1.05018628e+00
-4.08488959e-01 1.50843576e-01 1.08359540e+00 -1.21511318e-01
-6.64320529e-01 5.70247591e-01 7.69077912e-02 4.16252792e-01
7.41429985e-01 -2.49761656e-01 -5.36842346e-01 -5.22388756e-01
-6.59272790e-01 1.29610360e-01 6.01295769e-01 4.56149995e-01
9.66574073e-01 -1.31269646e+00 -1.15378106e+00 1.06008187e-01
-1.77734256e-01 1.42103538e-01 8.85952473e-01 7.57833719e-01
-2.35946283e-01 -2.43386149e-01 -1.42938606e-02 -5.17621160e-01
-1.24082708e+00 9.05642092e-01 4.72264409e-01 -1.08009383e-01
-8.87512326e-01 5.12465894e-01 4.82355624e-01 1.05468847e-03
1.31262749e-01 -1.08412154e-01 -3.81869346e-01 -3.03076744e-01
9.01778340e-01 6.26156628e-01 -1.75086215e-01 -7.26096570e-01
5.17260134e-02 5.74949086e-01 1.42249808e-01 9.44138840e-02
1.31905973e+00 -6.18273079e-01 -3.52122299e-02 3.81272554e-01
9.35519040e-01 1.89614177e-01 -1.48934352e+00 -5.68677366e-01
-3.93720180e-01 -9.74606395e-01 -4.37660702e-02 -5.36333561e-01
-1.35495138e+00 6.60222471e-01 8.71451139e-01 6.01874888e-02
1.73407161e+00 -4.45597261e-01 9.15583670e-01 5.95782837e-03
4.71560359e-01 -1.00472975e+00 4.34394419e-01 1.39618576e-01
1.08102596e+00 -1.24269879e+00 9.93985496e-03 -8.95406842e-01
-6.99599624e-01 8.25922549e-01 8.55973601e-01 3.60103659e-02
4.39919621e-01 3.11073333e-01 9.59918350e-02 3.00420493e-01
-6.48961008e-01 -2.14080885e-01 1.29130026e-02 7.26669610e-01
6.29720837e-02 -2.98291266e-01 -2.62459576e-01 7.83282697e-01
1.72688559e-01 3.11593443e-01 9.53296542e-01 3.47666234e-01
-3.48350406e-01 -5.97879469e-01 -6.23468399e-01 3.80165160e-01
-6.86344743e-01 -1.51845455e-01 3.18613589e-01 4.70689148e-01
-6.28135959e-03 1.13298798e+00 -4.24629509e-01 -4.89571929e-01
5.52504480e-01 -5.62431276e-01 3.37885052e-01 -3.11681151e-01
8.65590349e-02 3.27881038e-01 -7.63924718e-02 -6.86950147e-01
-5.06910741e-01 -4.61937845e-01 -7.94646859e-01 -9.28700641e-02
-4.58302498e-01 -1.35941073e-01 6.21371008e-02 5.67162335e-01
1.26401544e-01 8.24436128e-01 8.15573096e-01 -7.88472772e-01
-8.15607548e-01 -1.04868972e+00 -8.94967735e-01 6.94603324e-01
5.45916080e-01 -8.23801935e-01 -7.20295608e-01 -4.28573005e-02] | [11.092811584472656, -2.1140694618225098] |
23c10ff2-3b60-4850-bf23-807ae409d194 | encoder-decoder-based-unified-semantic-role | null | null | https://ojs.aaai.org/index.php/AAAI/article/view/17514 | https://ojs.aaai.org/index.php/AAAI/article/view/17514/17321 | Encoder-decoder based unified semantic role labeling with label-aware syntax | Currently the unified semantic role labeling (SRL) that achieves predicate identification and argument role labeling in an end-to-end manner has received growing interests. Recent works show that leveraging the syntax knowledge significantly enhances the SRL performances. In this paper, we investigate a novel unified SRL framework based on the sequence-to-sequence architecture with double enhancement in both the encoder and decoder sides. In the encoder side, we propose a novel label-aware graph convolutional network (LA-GCN) to encode both the syntactic dependent arcs and labels into BERT-based word representations. In the decoder side, we creatively design a pointer-network-based model for detecting predicates, arguments and roles jointly. Our pointer-net decoder is able to make decisions by consulting all the input elements in a global view, and meanwhile it is syntactic-aware by incorporating the syntax information from LA-GCN. Besides, a high-order interacted attention is introduced into the decoder for leveraging previously recognized triplets to help the current decision. Empirical experiments show that our framework significantly outperforms all existing graph-based methods on the CoNLL09 and Universal Proposition Bank datasets. In-depth analysis demonstrates that our model can effectively capture the correlations between syntactic and SRL structures. | ['Donghong Ji', 'Bobo Li', 'Fei Li', 'Hao Fei'] | 2021-05-18 | null | null | null | conference-2021-5 | ['semantic-role-labeling'] | ['natural-language-processing'] | [ 5.32987237e-01 4.23021138e-01 -4.34616029e-01 -6.25387788e-01
-8.91842365e-01 -8.88940752e-01 2.22388953e-01 1.90269500e-01
-3.44969511e-01 4.17393297e-01 6.64701164e-01 -5.42720854e-01
5.94695397e-02 -9.49905097e-01 -7.37275064e-01 -3.32926214e-01
2.44766235e-01 4.13434684e-01 6.39035404e-01 -4.07073110e-01
1.26172602e-01 7.11407587e-02 -1.19622493e+00 5.75251281e-01
8.35319161e-01 9.68184590e-01 4.52385455e-01 4.35892701e-01
-4.42488372e-01 1.35613382e+00 -2.62647748e-01 -4.94008094e-01
-8.47273394e-02 -2.97898173e-01 -8.49480689e-01 -2.09235966e-01
1.00721300e-01 -1.91016689e-01 -3.07944804e-01 9.64965641e-01
5.37662566e-01 2.01743096e-02 3.52775790e-02 -1.06498826e+00
-9.26888704e-01 1.29695606e+00 -5.51073372e-01 2.27438390e-01
4.04931158e-01 2.63386995e-01 1.97419834e+00 -6.06075287e-01
5.83493769e-01 1.72422886e+00 2.78940439e-01 4.80460733e-01
-7.21593022e-01 -4.84022826e-01 8.69343400e-01 5.12324989e-01
-9.29537058e-01 -1.14708379e-01 1.09538341e+00 -2.91767325e-02
1.12188220e+00 -2.82962844e-02 4.21110123e-01 7.63376236e-01
-3.36287707e-01 1.20504570e+00 4.89643842e-01 -3.17153543e-01
-1.86953396e-01 -6.10995293e-01 5.64391136e-01 1.03682017e+00
1.61855400e-01 -2.84967840e-01 -5.23812175e-01 6.95863143e-02
6.74566865e-01 -3.50631982e-01 -6.18487410e-02 -7.28388205e-02
-1.21108365e+00 7.44700372e-01 5.86368263e-01 4.20092970e-01
-1.01287574e-01 5.63384533e-01 7.63692081e-01 1.25018671e-01
2.70604670e-01 2.64931202e-01 -7.64332533e-01 -1.20149583e-01
-2.82518685e-01 -4.56075966e-02 6.22758210e-01 1.03892982e+00
5.32305717e-01 -2.33524680e-01 -5.87113500e-01 8.54676843e-01
6.52088881e-01 1.36062682e-01 2.80539304e-01 -7.61924803e-01
8.24331880e-01 1.32090724e+00 -1.97999015e-01 -9.23044682e-01
-5.84074438e-01 -6.11353219e-01 -3.40225279e-01 -5.56245387e-01
1.91370085e-01 7.34286234e-02 -7.69801438e-01 2.16157055e+00
4.07006532e-01 1.87169001e-01 1.48453668e-01 8.85653436e-01
1.02879918e+00 7.05624759e-01 3.93323451e-01 1.25705793e-01
1.78571784e+00 -1.33186841e+00 -7.67566085e-01 -5.08148253e-01
1.03703964e+00 -2.71989256e-01 1.18794429e+00 -2.37363487e-01
-9.99065399e-01 -5.31425238e-01 -1.07444525e+00 -7.65534878e-01
-8.55099931e-02 2.59249359e-01 6.30528748e-01 3.22958469e-01
-9.54681754e-01 2.25299492e-01 -6.12071931e-01 -2.58292127e-02
5.58610439e-01 3.54852587e-01 -2.35528275e-01 -2.32854754e-01
-1.58744252e+00 5.47824800e-01 8.33472192e-01 3.66387725e-01
-7.43582904e-01 -5.39279878e-01 -1.40958059e+00 4.75883693e-01
9.10979331e-01 -7.03707993e-01 1.31182182e+00 -6.08818650e-01
-1.34025776e+00 8.27156007e-01 -4.02917743e-01 -3.64868730e-01
7.86169246e-02 -6.06463349e-04 -3.17531019e-01 3.13949376e-01
2.87346989e-01 5.45746088e-01 3.81674856e-01 -1.03593981e+00
-9.22626376e-01 -4.44163650e-01 5.82502723e-01 5.42044163e-01
-9.74696428e-02 2.72789240e-01 -8.86679411e-01 -7.22910941e-01
2.57356286e-01 -5.80684125e-01 -1.19809382e-01 -9.13922116e-02
-5.79419494e-01 -7.93038189e-01 6.21747434e-01 -8.35561395e-01
1.55958283e+00 -2.29105830e+00 2.97460914e-01 9.82926115e-02
4.16779220e-01 3.26151252e-01 -3.09545666e-01 4.40242857e-01
-1.00632787e-01 3.39330435e-01 -4.08267587e-01 -2.73891240e-01
2.87973702e-01 5.96534431e-01 -9.05255601e-02 2.45784342e-01
3.03226441e-01 1.33669817e+00 -1.23101032e+00 -5.81694186e-01
-4.84572314e-02 6.68409616e-02 -7.57512927e-01 1.95061252e-01
-6.58355355e-01 2.94636101e-01 -5.59518337e-01 7.18797624e-01
5.50187290e-01 -5.88244617e-01 8.75974834e-01 -2.20702216e-01
2.58544274e-02 9.80036020e-01 -9.10569727e-01 1.84787691e+00
-4.48873878e-01 2.69508749e-01 1.26724720e-01 -1.10307741e+00
7.45417655e-01 1.89000070e-01 3.91566642e-02 -9.27920580e-01
4.98318560e-02 3.85402620e-01 3.78289431e-01 -2.72799134e-01
2.22343877e-01 -1.36440340e-02 -5.01379192e-01 3.45572263e-01
2.18488891e-02 6.10812724e-01 4.40000802e-01 3.37144345e-01
1.21296525e+00 2.50194401e-01 4.40598577e-01 -1.70479417e-01
1.24480963e+00 -6.28137290e-02 9.87740993e-01 5.37482142e-01
-7.89336339e-02 1.44701689e-01 1.10223389e+00 -1.72239527e-01
-6.23970568e-01 -5.98856091e-01 4.89827842e-01 1.60990942e+00
4.18353260e-01 -6.35252059e-01 -4.84245032e-01 -1.14373124e+00
1.00698225e-01 6.89856231e-01 -5.04441917e-01 -2.47235462e-01
-1.19867635e+00 -2.81075507e-01 7.57721841e-01 9.62705135e-01
7.32609093e-01 -1.06253040e+00 -2.39746064e-01 5.40686071e-01
-1.45945206e-01 -1.44272757e+00 -5.34207821e-01 1.29295602e-01
-5.37430525e-01 -1.38434422e+00 -1.90835491e-01 -1.15204990e+00
5.30598760e-01 -2.13925075e-02 9.79297042e-01 3.88824016e-01
2.36448511e-01 -1.69602349e-01 -5.63025773e-01 9.87023115e-02
-2.57783771e-01 3.53172809e-01 -6.10369921e-01 1.16584495e-01
2.62696266e-01 -4.65322435e-01 -4.71247494e-01 3.10346156e-01
-6.52412593e-01 3.18776727e-01 4.51559752e-01 7.66734242e-01
6.27033591e-01 -5.85519783e-02 6.15330994e-01 -1.38899016e+00
5.02866507e-01 -4.63327587e-01 -7.47382998e-01 4.68739748e-01
-2.89875805e-01 4.24299091e-01 7.52838016e-01 2.37751916e-01
-1.26228523e+00 -1.04980007e-01 -5.27749002e-01 4.55690622e-02
6.45104945e-02 7.49705672e-01 -9.39771473e-01 3.60680491e-01
4.28713337e-02 2.11261153e-01 -2.69817919e-01 -7.95513391e-01
6.40434802e-01 4.49550092e-01 6.08108103e-01 -8.86278272e-01
6.69503808e-01 2.41332501e-01 -3.19065265e-02 -1.87825799e-01
-1.21955168e+00 -4.99213725e-01 -5.62843263e-01 2.01187134e-01
8.91902685e-01 -1.07449567e+00 -9.76722538e-01 5.22663355e-01
-1.37059653e+00 -2.92175502e-01 -1.26619071e-01 -1.38771817e-01
-3.63001853e-01 5.04032373e-01 -7.50797153e-01 -5.22098303e-01
-3.21870565e-01 -1.26503754e+00 1.28607047e+00 2.81476639e-02
1.66311458e-01 -1.06067693e+00 -3.81052971e-01 4.96250898e-01
6.97908476e-02 -1.00074932e-01 1.59271622e+00 -1.06534648e+00
-6.73716307e-01 1.76721692e-01 -7.51812577e-01 2.92072475e-01
7.82996938e-02 -5.80318749e-01 -5.75116277e-01 -1.66714452e-02
-4.16478455e-01 -8.48248787e-03 1.05880022e+00 -2.56137569e-02
1.21671855e+00 -3.11021060e-01 -4.08806890e-01 5.71328402e-01
1.35862315e+00 2.62787342e-01 4.46198463e-01 2.50270784e-01
1.26981664e+00 6.52564704e-01 6.48168385e-01 1.40254185e-01
1.05774677e+00 5.67614794e-01 7.86845744e-01 3.27100120e-02
-4.54806894e-01 -7.50158012e-01 4.19848293e-01 6.52082503e-01
2.60481238e-01 -6.60108089e-01 -9.09282327e-01 4.43799466e-01
-2.10405207e+00 -7.03119099e-01 -2.73327649e-01 1.47366619e+00
7.48759151e-01 1.51167065e-01 -8.25706646e-02 1.91225648e-01
9.61834610e-01 5.48030913e-01 -5.82561851e-01 -2.03472465e-01
-2.00168863e-01 5.04809879e-02 6.42210364e-01 5.60232401e-01
-1.12485015e+00 1.43474853e+00 5.15076160e+00 9.37105477e-01
-6.31292284e-01 3.47708046e-01 3.24876517e-01 2.63578534e-01
-7.86850631e-01 1.87297985e-01 -1.12927341e+00 4.83132958e-01
6.81407213e-01 -1.27876341e-01 1.13263182e-01 7.09817052e-01
3.69818211e-02 1.74335346e-01 -9.56048012e-01 8.33638072e-01
2.76071187e-02 -1.50993919e+00 2.14346483e-01 -1.56802148e-01
7.10009262e-02 1.88067257e-02 -3.32010448e-01 5.08700311e-01
4.69158351e-01 -7.38369644e-01 9.50680256e-01 5.95175549e-02
8.96753073e-01 -7.76206017e-01 6.89399958e-01 3.00450444e-01
-1.76296234e+00 -4.04271305e-01 -1.82782263e-01 -9.27921310e-02
3.87601465e-01 3.36059272e-01 -7.15118289e-01 9.15681362e-01
5.27488962e-02 1.26130867e+00 -5.13539553e-01 6.13034844e-01
-1.18180645e+00 7.95758963e-01 -1.48545071e-01 -8.37382302e-02
5.78317285e-01 6.02877736e-02 5.32612681e-01 1.15120709e+00
-1.84623320e-02 1.53514236e-01 4.04480934e-01 7.55061090e-01
-3.78394574e-01 4.45892364e-02 -2.42581844e-01 -2.97072828e-01
5.45502067e-01 1.09993267e+00 -7.54965484e-01 -3.46180737e-01
-5.49816668e-01 9.61817682e-01 6.59347594e-01 3.35772261e-02
-8.93036306e-01 -2.50773787e-01 6.33739710e-01 -6.43957034e-02
4.07771528e-01 -2.57476449e-01 -2.15864435e-01 -1.12251544e+00
1.79265700e-02 -7.46136010e-01 9.08581674e-01 -7.94204295e-01
-9.80015337e-01 3.70661527e-01 -9.07356814e-02 -7.35938132e-01
1.70209855e-02 -8.89019847e-01 -3.55229437e-01 9.65203941e-01
-2.04796028e+00 -1.49547279e+00 1.24693841e-01 4.84367967e-01
5.84204376e-01 -3.19484808e-02 4.39349622e-01 5.35934389e-01
-7.45567501e-01 6.33467555e-01 -5.67067742e-01 5.01093328e-01
1.61827505e-01 -1.22476113e+00 6.88510835e-01 1.10517037e+00
-4.40101177e-02 6.58677280e-01 1.33551523e-01 -7.08315194e-01
-1.46403635e+00 -1.17834115e+00 1.24906850e+00 -1.49044454e-01
6.32003725e-01 -4.37272578e-01 -9.41361606e-01 9.48349953e-01
-6.27675354e-02 8.02822933e-02 5.97158432e-01 1.80426344e-01
-5.44292152e-01 7.86796361e-02 -7.53146946e-01 4.46251154e-01
1.72818995e+00 -7.65658259e-01 -7.77802587e-01 1.86396867e-01
1.43363190e+00 -6.18611455e-01 -4.47296858e-01 5.20265758e-01
2.35739574e-01 -5.94609737e-01 8.20761561e-01 -6.37959719e-01
5.25887132e-01 -6.54477060e-01 -2.63408512e-01 -9.33769524e-01
-4.87788290e-01 -4.68467504e-01 -1.99353576e-01 1.31393135e+00
6.43429041e-01 -5.71885884e-01 5.17447293e-01 1.35611013e-01
-6.07759774e-01 -9.04835522e-01 -8.74255896e-01 -5.56671977e-01
-4.32017565e-01 -7.77972400e-01 8.62515926e-01 5.36648691e-01
-1.08262561e-01 9.43603694e-01 -1.04458340e-01 3.74659181e-01
2.51438826e-01 4.11697149e-01 2.20864594e-01 -1.22690344e+00
-2.63293058e-01 -3.46913576e-01 -1.29385337e-01 -1.54199040e+00
6.64407670e-01 -1.55333507e+00 -8.11503902e-02 -2.07554173e+00
1.03098780e-01 -4.99157339e-01 -4.85027611e-01 1.00273299e+00
-4.56818193e-01 -2.68592477e-01 3.42323929e-01 1.30517622e-02
-9.15269971e-01 5.35042584e-01 1.39885020e+00 -7.39947855e-02
1.34834141e-01 -3.93412292e-01 -1.22582257e+00 7.45154679e-01
6.46711349e-01 -4.27907020e-01 -4.86121416e-01 -9.41330492e-01
7.57004440e-01 9.92647782e-02 1.60169065e-01 -5.28694987e-01
2.80124962e-01 1.15462564e-01 -2.36706555e-01 -6.64278388e-01
7.95377642e-02 -7.12439179e-01 -5.07331192e-01 2.21523181e-01
-5.80194771e-01 5.72604015e-02 -5.33863530e-02 8.25651705e-01
-3.47618908e-01 -2.55017251e-01 4.07174081e-01 -3.55598003e-01
-1.21350050e+00 3.52519929e-01 3.15936804e-02 3.36142361e-01
7.32489645e-01 1.37517914e-01 -5.70708930e-01 -8.52584168e-02
-5.84935486e-01 8.20647240e-01 1.64693996e-01 3.91763330e-01
4.68884617e-01 -1.14667690e+00 -4.71672028e-01 2.72981763e-01
2.65128136e-01 2.88153708e-01 2.74434566e-01 5.95218718e-01
-4.78805900e-01 4.80127335e-01 1.61490723e-01 -1.24163479e-01
-1.15465343e+00 5.23993552e-01 2.57845968e-01 -8.59226346e-01
-3.80802631e-01 1.24947989e+00 3.38833719e-01 -5.99105179e-01
1.29480839e-01 -4.95314360e-01 -6.04702413e-01 1.39358655e-01
3.43800992e-01 7.18261078e-02 -2.14742824e-01 -7.85389900e-01
-6.47850215e-01 3.88897628e-01 -1.05525352e-01 9.86411944e-02
1.22143722e+00 -7.17468411e-02 -5.78588605e-01 5.80994748e-02
1.15201128e+00 1.15189277e-01 -1.11922824e+00 -5.26796103e-01
5.12101233e-01 -1.51341990e-01 -6.45065159e-02 -9.37731504e-01
-1.15774035e+00 1.01120043e+00 -2.64405936e-01 -1.22080065e-01
1.09397602e+00 2.59514004e-01 1.07367623e+00 2.11086944e-01
2.08396435e-01 -9.37741220e-01 4.32579182e-02 1.00797951e+00
6.85320199e-01 -8.69937718e-01 -5.29262125e-01 -1.18071771e+00
-6.60038948e-01 1.01268411e+00 7.44998395e-01 6.46124929e-02
2.50124365e-01 1.89742461e-01 -3.08387429e-01 -3.36057037e-01
-9.09738362e-01 -7.52957404e-01 2.41204128e-01 3.77676278e-01
4.35488641e-01 2.99161598e-02 -5.99662423e-01 1.10363889e+00
5.91278896e-02 -2.62337118e-01 1.27385110e-01 9.97131884e-01
-5.10063052e-01 -1.52438712e+00 2.87156880e-01 1.71037778e-01
-5.36084890e-01 -5.31407416e-01 -2.11585835e-01 5.57028174e-01
2.60792613e-01 8.90258312e-01 1.21939942e-01 -2.12031126e-01
5.62902868e-01 3.07108372e-01 2.56630808e-01 -1.19819987e+00
-7.89065659e-01 -3.95935103e-02 7.63091862e-01 -6.59355938e-01
-4.29849327e-01 -3.52658063e-01 -2.04357862e+00 1.70479491e-01
-7.08179623e-02 -2.02109423e-02 1.90153494e-01 1.05463207e+00
5.65288067e-01 1.07442510e+00 4.00315762e-01 -1.22492328e-01
-3.87189955e-01 -6.87743068e-01 -4.08594400e-01 2.08045870e-01
1.17576472e-01 -6.71271801e-01 -6.18457720e-02 -3.98590386e-01] | [10.328024864196777, 9.25981330871582] |
75d03e69-39ba-46fe-aa60-ab6d8d3ed0e0 | ditch-the-gold-standard-re-evaluating-1 | 2112.08812 | null | https://arxiv.org/abs/2112.08812v2 | https://arxiv.org/pdf/2112.08812v2.pdf | Ditch the Gold Standard: Re-evaluating Conversational Question Answering | Conversational question answering aims to provide natural-language answers to users in information-seeking conversations. Existing conversational QA benchmarks compare models with pre-collected human-human conversations, using ground-truth answers provided in conversational history. It remains unclear whether we can rely on this static evaluation for model development and whether current systems can well generalize to real-world human-machine conversations. In this work, we conduct the first large-scale human evaluation of state-of-the-art conversational QA systems, where human evaluators converse with models and judge the correctness of their answers. We find that the distribution of human machine conversations differs drastically from that of human-human conversations, and there is a disagreement between human and gold-history evaluation in terms of model ranking. We further investigate how to improve automatic evaluations, and propose a question rewriting mechanism based on predicted history, which better correlates with human judgments. Finally, we analyze the impact of various modeling strategies and discuss future directions towards building better conversational question answering systems. | ['Danqi Chen', 'Manan Goenka', 'Tianyu Gao', 'Huihan Li'] | 2021-12-16 | null | https://aclanthology.org/2022.acl-long.555 | https://aclanthology.org/2022.acl-long.555.pdf | acl-2022-5 | ['question-rewriting'] | ['natural-language-processing'] | [ 6.63927123e-02 7.07322061e-01 2.21577913e-01 -6.49043858e-01
-1.32370949e+00 -8.49042594e-01 1.05268836e+00 1.08885564e-01
-3.17177385e-01 7.93737769e-01 7.55695164e-01 -8.08772683e-01
1.17238350e-01 -5.83749592e-01 -3.95264365e-02 1.89277772e-02
3.39189559e-01 1.28251648e+00 4.71912205e-01 -9.92888510e-01
7.63001367e-02 -2.47993723e-01 -1.32062685e+00 1.00127339e+00
9.98788655e-01 4.17419076e-01 -2.22715661e-01 1.25456166e+00
-2.37471938e-01 1.39017415e+00 -8.99975836e-01 -1.07522404e+00
-8.34729448e-02 -1.00658548e+00 -1.80734110e+00 -1.83317825e-01
5.40269375e-01 -3.26874107e-01 -3.61291692e-02 5.60015976e-01
4.00923431e-01 9.24455002e-02 6.64784253e-01 -1.41138530e+00
-5.68528354e-01 4.75035369e-01 3.39678288e-01 1.21957891e-01
1.14621663e+00 5.99187315e-01 1.30861688e+00 -3.65425557e-01
8.16406608e-01 1.54849565e+00 5.12484491e-01 1.03460670e+00
-8.73220861e-01 1.14534202e-03 -1.29089430e-01 5.35680950e-01
-7.70188093e-01 -5.62479436e-01 3.33340108e-01 -4.89922971e-01
1.18238854e+00 8.35513353e-01 2.59707123e-01 1.13739848e+00
6.63044676e-02 5.94961047e-01 1.04873240e+00 -6.60647333e-01
3.91628081e-03 3.38279217e-01 6.83989167e-01 6.65760815e-01
-2.92807728e-01 -3.99168193e-01 -4.25078422e-01 -8.40568483e-01
-2.37313777e-01 -7.65293181e-01 -5.26410937e-01 7.90744424e-02
-1.04214180e+00 9.07250702e-01 5.51256947e-02 3.63225281e-01
-3.86819273e-01 -3.35798085e-01 3.59529227e-01 6.55750930e-01
2.39076003e-01 1.09784997e+00 -6.71631455e-01 -6.58042073e-01
-3.58786315e-01 7.30047524e-01 1.75663888e+00 7.17022657e-01
7.14010596e-01 -8.45788360e-01 -5.59882283e-01 9.30168390e-01
1.40489608e-01 4.61878181e-01 3.31380814e-01 -1.73650491e+00
6.00087166e-01 9.30158019e-01 6.23547912e-01 -8.45389485e-01
-2.78802723e-01 7.88314790e-02 -1.97047591e-01 -4.11744863e-01
8.61759007e-01 -3.72777760e-01 -1.96128696e-01 1.54411328e+00
2.84798861e-01 -4.38157678e-01 3.35634977e-01 7.21578598e-01
9.50465798e-01 6.25835776e-01 5.35549745e-02 -1.79567620e-01
1.72127318e+00 -1.35296500e+00 -6.46684527e-01 -2.76157945e-01
9.25848961e-01 -7.65207827e-01 1.44755900e+00 1.55478835e-01
-1.09467280e+00 -3.26267362e-01 -4.52808976e-01 -3.39433849e-01
-1.39761129e-02 -2.60436386e-01 3.98511946e-01 7.43605554e-01
-1.24344409e+00 9.23446715e-02 -4.64074433e-01 -7.35991776e-01
-4.79106128e-01 5.49295619e-02 3.57676074e-02 -4.51058000e-02
-1.60575294e+00 1.36172402e+00 -4.22830462e-01 -5.71504831e-02
-6.72055423e-01 -4.01691139e-01 -7.85307527e-01 -3.43117975e-02
2.36678720e-01 -1.16263521e+00 2.42022920e+00 -6.40034556e-01
-1.59483194e+00 1.04634750e+00 -9.00734305e-01 -5.84793687e-01
4.59425688e-01 5.14940992e-02 -4.14325297e-01 1.06155038e-01
1.52551442e-01 4.85368460e-01 8.74999091e-02 -1.36480772e+00
-7.82207131e-01 -1.51315749e-01 8.68604779e-01 4.11583364e-01
8.53043236e-03 5.93297258e-02 -3.75102699e-01 5.23504913e-01
-3.26543331e-01 -1.09845173e+00 -2.09689870e-01 -3.62933069e-01
-1.85338050e-01 -8.48061740e-01 3.05161268e-01 -8.89185071e-01
1.32323778e+00 -1.33655524e+00 -1.62186265e-01 -3.03235441e-03
2.87650585e-01 2.54438818e-01 -3.16995263e-01 9.66610968e-01
4.99093950e-01 1.14065982e-01 1.14838835e-02 -5.06309211e-01
3.19887489e-01 1.65202051e-01 -4.13002372e-01 -2.22575098e-01
-2.50845924e-02 1.10659134e+00 -1.16829348e+00 -7.37547517e-01
-1.61194697e-01 -1.06654100e-01 -5.27193666e-01 6.53125048e-01
-7.04216659e-01 4.36505556e-01 -4.85206395e-01 2.79144853e-01
5.83639853e-02 -5.78942776e-01 1.70262426e-01 1.48327380e-01
3.09050828e-01 1.12744319e+00 -3.30625802e-01 1.39280593e+00
-6.33038580e-01 7.92774558e-01 2.46445253e-01 -2.22714871e-01
6.39223754e-01 5.96222520e-01 -8.29905048e-02 -9.15802896e-01
7.83634558e-02 4.64405008e-02 3.31569552e-01 -8.95022631e-01
7.80328691e-01 -1.91832688e-02 -1.28727168e-01 9.09367383e-01
-1.89254612e-01 -5.18376708e-01 5.11760712e-01 4.81044739e-01
1.36916435e+00 -4.20016021e-01 1.55957669e-01 -7.87263811e-02
9.61418688e-01 6.24747276e-01 4.45257761e-02 1.13487840e+00
-5.15963554e-01 4.63726819e-01 7.11769044e-01 -3.18780392e-01
-6.85544848e-01 -6.67165756e-01 2.52056837e-01 1.45760334e+00
1.79093242e-01 -6.37843728e-01 -1.25127053e+00 -8.55501354e-01
-3.65630835e-01 1.23909259e+00 -4.75501955e-01 3.95384468e-02
-7.55192816e-01 -3.85979176e-01 7.83967257e-01 -4.82099131e-02
2.02330798e-01 -1.19279313e+00 -3.46332401e-01 3.06087613e-01
-1.12531972e+00 -1.12946892e+00 -3.58175993e-01 -6.35585725e-01
-4.68825668e-01 -1.38367999e+00 -4.25288647e-01 -6.62941515e-01
1.10290408e-01 2.70236164e-01 1.87468338e+00 5.80273330e-01
4.18843389e-01 8.18218708e-01 -5.74422956e-01 -3.07826936e-01
-1.21328962e+00 4.76961851e-01 -3.37046713e-01 -3.58348906e-01
8.61686289e-01 -1.62488475e-01 -7.08939791e-01 8.19837868e-01
-3.46161336e-01 9.51406211e-02 1.14545804e-02 7.74625659e-01
-5.13501823e-01 -6.45888627e-01 6.00002944e-01 -1.07076871e+00
1.52921319e+00 -4.44456458e-01 -3.36300656e-02 8.52132142e-01
-5.99925518e-01 6.96613416e-02 4.50097322e-01 1.20311687e-02
-1.37897909e+00 -6.54877663e-01 -4.67993110e-01 5.85912526e-01
-1.88011676e-01 3.22340518e-01 1.45593211e-01 3.54936391e-01
1.23821521e+00 -1.64165959e-01 2.98590846e-02 -1.81444377e-01
6.03968322e-01 1.06580615e+00 5.23169637e-01 -9.94528592e-01
3.76333743e-01 1.11246958e-01 -7.71486402e-01 -5.60847759e-01
-9.71402526e-01 -7.52647460e-01 -1.09807581e-01 -3.40838015e-01
8.52723598e-01 -6.79216206e-01 -1.22795820e+00 1.39708951e-01
-1.76350152e+00 -6.86520159e-01 1.66465007e-02 -1.87639799e-02
-6.15568817e-01 5.58445752e-01 -9.58038986e-01 -1.11468184e+00
-7.14817405e-01 -1.10104346e+00 1.06725204e+00 1.39424011e-01
-1.24859262e+00 -1.02445257e+00 6.47580028e-01 1.38099861e+00
5.23052037e-01 -4.58281845e-01 1.04271710e+00 -9.06123936e-01
-2.45667651e-01 -9.89978760e-02 -6.09779265e-04 2.15627536e-01
-2.54167058e-02 -1.45359218e-01 -1.04791975e+00 1.81273166e-02
9.78064686e-02 -5.85153341e-01 1.85418144e-01 -2.13355169e-01
3.88788402e-01 -5.29932261e-01 -1.43371299e-01 -5.53784668e-01
5.15514195e-01 1.58558264e-01 7.62195230e-01 2.06028655e-01
4.63657826e-02 1.24243140e+00 5.65941989e-01 -1.30174816e-01
1.13410580e+00 6.15315914e-01 -7.96941742e-02 2.29367211e-01
-1.50700703e-01 -2.69074112e-01 2.10353777e-01 1.08380020e+00
5.60430661e-02 -7.31899798e-01 -1.22127819e+00 7.69353271e-01
-1.95830274e+00 -1.01550436e+00 -4.52169210e-01 1.70621216e+00
1.13135111e+00 1.34063825e-01 1.24959916e-01 -3.98069024e-01
5.90549231e-01 -7.07204044e-02 -1.92208663e-01 -5.75512767e-01
1.48621857e-01 9.46147963e-02 -2.77469426e-01 1.25739825e+00
-4.14348036e-01 6.88598514e-01 6.89760494e+00 2.78424174e-01
-5.42707026e-01 3.66844594e-01 7.25476444e-01 3.77431989e-01
-5.82770586e-01 3.17996472e-01 -6.67948008e-01 2.03489363e-01
1.41184878e+00 -3.99197519e-01 4.30188030e-01 7.44688690e-01
1.81493968e-01 -7.26513267e-02 -1.40033877e+00 5.47813118e-01
2.21290737e-01 -1.20246208e+00 8.53026882e-02 -1.40521303e-01
6.27262294e-01 2.23070458e-02 -6.36443675e-01 9.81396258e-01
7.07292199e-01 -9.00309861e-01 2.32856885e-01 6.96167707e-01
-1.19548745e-03 -1.82096466e-01 1.11166418e+00 8.73233557e-01
-6.71256363e-01 -8.42587370e-03 -5.84265450e-03 -4.53817755e-01
6.00028038e-01 2.42125411e-02 -1.46354485e+00 4.02340740e-01
3.35831970e-01 -7.92481676e-02 -7.90293992e-01 5.65275788e-01
-1.58476859e-01 8.05619597e-01 -5.93475662e-02 -6.27688646e-01
1.78192317e-01 -1.97398707e-01 3.56911898e-01 1.24926317e+00
-1.10269688e-01 2.93171555e-01 6.57935590e-02 5.66897750e-01
-4.19631787e-02 9.32671204e-02 -3.99788797e-01 6.24002255e-02
5.54434419e-01 1.11946940e+00 1.20644175e-01 -6.46712244e-01
-4.14239824e-01 8.04116607e-01 4.57560778e-01 3.39369625e-01
-3.01156968e-01 -9.85528901e-02 5.97773433e-01 2.69055843e-01
-5.63957751e-01 -5.33897169e-02 -1.69551849e-01 -1.13019037e+00
2.68397957e-01 -1.69233537e+00 6.21793568e-01 -1.12474775e+00
-1.60374379e+00 1.12362540e+00 -7.67995566e-02 -8.18345726e-01
-1.13680673e+00 -4.41501498e-01 -8.80118668e-01 8.13374639e-01
-1.07037115e+00 -8.14726174e-01 -5.02131164e-01 4.16746795e-01
6.66504323e-01 4.06551287e-02 1.12714529e+00 1.25243843e-01
-5.25398776e-02 5.82726061e-01 -4.02701288e-01 1.16882183e-01
9.67814684e-01 -1.20653319e+00 8.34069490e-01 4.24824476e-01
-1.90654770e-03 1.05863547e+00 1.19000137e+00 -3.29572439e-01
-1.14943969e+00 -5.67960441e-01 1.59908926e+00 -1.48507392e+00
7.30852365e-01 -2.30215579e-01 -1.18852055e+00 6.15156949e-01
9.31319356e-01 -1.01709783e+00 6.89458013e-01 6.16473377e-01
-2.66176581e-01 1.44337296e-01 -9.73938882e-01 8.41323972e-01
8.90755534e-01 -9.12392318e-01 -1.17501962e+00 6.70456588e-01
1.00227976e+00 -3.04136932e-01 -6.67527258e-01 4.00813282e-01
4.22649413e-01 -9.55473244e-01 5.77462018e-01 -9.20407712e-01
3.02759439e-01 -1.81909695e-01 -2.11471785e-02 -1.20256805e+00
1.59638926e-01 -7.85170197e-01 2.17760634e-02 1.08075523e+00
9.10682082e-01 -7.04988718e-01 7.25630522e-01 1.40178144e+00
-7.05429390e-02 -6.49808109e-01 -7.25273371e-01 -2.71662474e-01
1.48172021e-01 -2.39060774e-01 6.28802419e-01 8.59655440e-01
7.19905317e-01 1.07671773e+00 -1.46648154e-01 -1.39688492e-01
2.09658351e-02 6.53263032e-02 1.33518851e+00 -1.05547154e+00
-5.67927778e-01 -3.80209535e-01 1.23785272e-01 -1.51431096e+00
3.25984299e-01 -4.04876351e-01 3.64724070e-01 -1.93323874e+00
2.52166778e-01 -4.29366045e-02 5.04852176e-01 2.09712237e-01
-5.09868860e-01 -1.66732043e-01 2.03920109e-03 2.55110145e-01
-1.06219220e+00 5.49921393e-01 1.37887812e+00 -8.69275555e-02
-1.46912292e-01 1.88674957e-01 -8.40210915e-01 5.08256018e-01
6.42117500e-01 -2.34284714e-01 -5.55536807e-01 -5.04011333e-01
3.55063736e-01 4.79457587e-01 2.12436497e-01 -7.02570438e-01
5.79710305e-01 -2.26233616e-01 -6.08818114e-01 -2.29805470e-01
4.58785444e-01 -2.32344985e-01 -2.65976250e-01 2.80046403e-01
-7.31504560e-01 2.77374595e-01 -1.94328055e-01 2.25458354e-01
-3.86226445e-01 -4.70694453e-01 2.61369318e-01 -2.37510756e-01
-2.68744886e-01 -2.54951000e-01 -7.84490287e-01 4.01181042e-01
4.21827286e-01 2.23812953e-01 -1.00281215e+00 -1.37536919e+00
-3.99270356e-01 7.68120527e-01 3.33333254e-01 5.09592712e-01
1.69655588e-02 -7.78591096e-01 -1.03209209e+00 -6.09335840e-01
5.01593232e-01 -4.22414452e-01 1.91795602e-01 5.28924108e-01
-4.93628412e-01 8.10461104e-01 3.31193268e-01 -5.71759045e-01
-1.32869506e+00 3.04012392e-02 6.49093926e-01 -6.72184467e-01
1.28451020e-01 5.16388059e-01 -2.66787373e-02 -1.21592569e+00
1.64654210e-01 -1.50530234e-01 -2.46255636e-01 -2.09957153e-01
7.65400589e-01 2.55859941e-01 3.23899418e-01 -4.21304673e-01
-2.15858608e-01 -1.09677305e-02 -1.13093480e-01 -6.40331089e-01
4.79754746e-01 -3.34386289e-01 -3.76205385e-01 3.57181489e-01
1.07506108e+00 -1.49241360e-02 -5.08509457e-01 -3.74517679e-01
2.46883258e-01 -1.82791725e-01 -7.71353662e-01 -1.01653314e+00
-2.23495476e-02 8.47694457e-01 2.09970489e-01 7.04554379e-01
5.90016544e-01 2.64333516e-01 1.11822224e+00 1.09022927e+00
5.60216129e-01 -8.21978211e-01 2.25116029e-01 1.00560629e+00
1.15706623e+00 -1.35430050e+00 -3.75671268e-01 -2.80476242e-01
-9.93719101e-01 7.81112790e-01 8.93804848e-01 4.20686454e-01
1.79133892e-01 -3.72145534e-01 7.90623724e-01 -4.77977574e-01
-1.42789268e+00 -2.31982216e-01 3.95307004e-01 4.12391216e-01
6.96810067e-01 5.26434071e-02 -4.94853377e-01 5.37713706e-01
-8.74772012e-01 -1.61049783e-01 5.31567931e-01 7.13864505e-01
-4.95316595e-01 -1.35856795e+00 -2.21140414e-01 2.61580110e-01
-4.78922725e-02 -8.72127190e-02 -1.11274421e+00 6.30153537e-01
-4.83505547e-01 1.99263430e+00 -1.82303339e-01 -4.04073477e-01
6.42184377e-01 5.73589206e-01 1.82008386e-01 -7.53510118e-01
-1.12881935e+00 -7.71924555e-01 1.20203507e+00 -4.13249612e-01
-2.79316068e-01 -4.89031881e-01 -9.39058959e-01 -5.19144535e-01
-3.61214727e-01 9.76265192e-01 3.71456265e-01 1.28609729e+00
4.55339819e-01 -1.56018138e-01 4.28496987e-01 -7.45675042e-02
-7.71574497e-01 -1.34997535e+00 4.35548574e-01 7.94981658e-01
2.55231202e-01 9.63342469e-03 -5.31871557e-01 1.19385451e-01] | [12.062211990356445, 7.982449531555176] |
788f5bdb-f1ad-4ff3-bcaa-0d0e615fe0be | meta-sage-scale-meta-learning-scheduled | 2306.02688 | null | https://arxiv.org/abs/2306.02688v2 | https://arxiv.org/pdf/2306.02688v2.pdf | Meta-SAGE: Scale Meta-Learning Scheduled Adaptation with Guided Exploration for Mitigating Scale Shift on Combinatorial Optimization | This paper proposes Meta-SAGE, a novel approach for improving the scalability of deep reinforcement learning models for combinatorial optimization (CO) tasks. Our method adapts pre-trained models to larger-scale problems in test time by suggesting two components: a scale meta-learner (SML) and scheduled adaptation with guided exploration (SAGE). First, SML transforms the context embedding for subsequent adaptation of SAGE based on scale information. Then, SAGE adjusts the model parameters dedicated to the context embedding for a specific instance. SAGE introduces locality bias, which encourages selecting nearby locations to determine the next location. The locality bias gradually decays as the model is adapted to the target instance. Results show that Meta-SAGE outperforms previous adaptation methods and significantly improves scalability in representative CO tasks. Our source code is available at https://github.com/kaist-silab/meta-sage | ['Jinkyoo Park', 'Hyeonah Kim', 'Minsu Kim', 'Jiwoo Son'] | 2023-06-05 | null | null | null | null | ['combinatorial-optimization'] | ['methodology'] | [-1.83214828e-01 -1.11355104e-01 -5.67774892e-01 -2.74465173e-01
-1.01445377e+00 -4.54849064e-01 3.25021118e-01 2.66580462e-01
-7.17099845e-01 8.52175415e-01 2.71527261e-01 -2.06588030e-01
-3.11777204e-01 -7.23099589e-01 -7.91653097e-01 -6.06192887e-01
-3.68940145e-01 7.90580571e-01 1.22833006e-01 -2.11762920e-01
2.76144594e-01 4.28705752e-01 -1.12324333e+00 3.63773853e-01
9.55617964e-01 8.78966749e-01 5.04253149e-01 7.68043876e-01
-1.35488570e-01 4.86590713e-01 -5.48884153e-01 5.11292554e-02
3.48631620e-01 -5.91137290e-01 -8.45869184e-01 -2.14102551e-01
8.47085118e-02 -1.24652535e-01 -1.58693537e-01 8.03794920e-01
5.15871406e-01 5.39286137e-01 4.00978088e-01 -1.24266016e+00
-3.49139035e-01 8.49557996e-01 -4.40113068e-01 4.91095424e-01
-2.01114789e-01 3.46142471e-01 1.07088196e+00 -1.07002103e+00
4.78382200e-01 1.08158326e+00 5.48099339e-01 6.23129845e-01
-1.24265325e+00 -6.41109228e-01 4.96180058e-01 1.93914652e-01
-1.14262402e+00 -1.95876271e-01 8.56604695e-01 8.05949494e-02
1.05262923e+00 6.29724190e-02 9.13644254e-01 1.01512289e+00
3.03717516e-02 9.62266445e-01 1.10095692e+00 -4.75270748e-01
8.69413733e-01 -7.71371126e-02 -1.44970283e-01 8.24801564e-01
-8.95052999e-02 2.12427691e-01 -7.65369654e-01 -1.92573890e-01
7.78115988e-01 -2.29690969e-01 1.99709103e-01 -6.92581952e-01
-1.06004500e+00 9.82337713e-01 7.08442688e-01 9.98261124e-02
-5.50187409e-01 4.76858348e-01 4.81420904e-01 4.20391291e-01
4.66449648e-01 9.61023092e-01 -8.18994999e-01 -3.23907018e-01
-6.91186607e-01 4.06059861e-01 5.71161747e-01 9.08972859e-01
8.73676062e-01 -4.11411934e-02 -2.63740122e-01 1.07727897e+00
1.22216567e-01 1.03821576e-01 6.75965250e-01 -9.62758482e-01
7.27356970e-01 6.06314778e-01 1.03776276e-01 -3.77357990e-01
-5.59970498e-01 -8.92779350e-01 -2.78725803e-01 2.77286053e-01
1.85225993e-01 -3.59845400e-01 -9.76671278e-01 1.89561903e+00
6.22925997e-01 8.30588192e-02 -1.59477234e-01 8.58598173e-01
4.61535901e-01 6.55097604e-01 2.64779121e-01 6.16270229e-02
9.08962190e-01 -1.68048298e+00 -2.54519433e-01 -5.10252059e-01
7.73642898e-01 -3.11540425e-01 1.47770357e+00 3.35762709e-01
-1.15850282e+00 -4.54308093e-01 -1.01823974e+00 3.13925833e-01
-5.54240406e-01 1.91644430e-01 6.45243347e-01 2.92255312e-01
-1.14528203e+00 6.41494572e-01 -8.71576607e-01 -1.97616532e-01
6.10707164e-01 4.62334096e-01 5.38431183e-02 7.39598945e-02
-1.17094493e+00 6.94336951e-01 5.96706033e-01 -2.70743258e-02
-1.29287672e+00 -6.80247605e-01 -6.94479823e-01 2.17256531e-01
6.30886674e-01 -5.16969979e-01 1.44001222e+00 -1.13000381e+00
-1.81092715e+00 3.45960885e-01 -1.66863531e-01 -4.88613009e-01
4.44288284e-01 -4.31012630e-01 -1.98114738e-01 -8.05943385e-02
3.59629169e-02 8.36905062e-01 1.04966938e+00 -1.16130555e+00
-7.31774032e-01 -2.13810787e-01 5.78501113e-02 5.23191035e-01
-3.83871287e-01 -1.22547373e-01 -7.08084345e-01 -7.79100895e-01
-1.22279420e-01 -1.04756439e+00 -5.79675257e-01 -3.78328621e-01
-1.60365418e-01 -3.35930347e-01 3.94238025e-01 -3.92664075e-01
1.61890137e+00 -2.05909181e+00 6.32215858e-01 3.17199588e-01
1.11787036e-01 1.95951268e-01 -7.78457463e-01 5.94558060e-01
-7.52181336e-02 7.61370584e-02 -1.71246797e-01 -3.94867122e-01
-2.67077107e-02 1.94707185e-01 8.39065947e-03 1.73345972e-02
1.89123750e-01 1.05884647e+00 -1.09115279e+00 -2.51770824e-01
3.06282993e-02 -4.27663438e-02 -8.59103799e-01 2.09940121e-01
-7.41800010e-01 3.65975320e-01 -6.01974308e-01 7.07724869e-01
3.56285304e-01 -5.49796879e-01 2.67752975e-01 1.11219555e-01
-4.23950609e-03 5.04197836e-01 -1.19436049e+00 1.83116591e+00
-7.16226041e-01 2.52280504e-01 -1.95978925e-01 -8.33028436e-01
7.79802620e-01 -1.39846742e-01 3.67592454e-01 -8.37187707e-01
-8.73846188e-02 1.65477768e-01 -5.66554405e-02 -2.32934058e-01
3.77815157e-01 4.30918008e-01 -9.32614431e-02 4.71238106e-01
8.25397894e-02 -2.38147359e-02 4.01297837e-01 7.80027360e-02
1.16757452e+00 4.29621816e-01 3.50668788e-01 -2.24073201e-01
2.84341633e-01 9.34098363e-02 6.46714926e-01 8.01716268e-01
-2.56784618e-01 6.64453488e-03 3.69618058e-01 -7.24746525e-01
-1.04648423e+00 -1.20465362e+00 2.14684486e-01 1.60912538e+00
-8.59358460e-02 -7.03820646e-01 -6.73192561e-01 -1.22579515e+00
1.90189496e-01 7.79720068e-01 -9.17674422e-01 -2.00582206e-01
-8.90324891e-01 -6.40533507e-01 7.35650435e-02 6.77665949e-01
5.56777298e-01 -1.38950586e+00 -6.38851047e-01 3.64340425e-01
1.30977869e-01 -4.46078062e-01 -7.23644078e-01 6.66530848e-01
-1.16149509e+00 -7.98813045e-01 -5.15262187e-01 -7.46270895e-01
7.58029878e-01 -2.25917608e-01 1.08504593e+00 1.05017066e-01
-2.02333406e-01 9.84877869e-02 -3.45172256e-01 -2.64538348e-01
-3.16252232e-01 5.36911428e-01 -8.87124985e-02 -4.12377387e-01
1.98941424e-01 -3.69123787e-01 -7.38450110e-01 3.54032129e-01
-6.55790985e-01 -9.73573700e-02 6.44901514e-01 1.19295371e+00
8.17150533e-01 -1.58945084e-01 7.35527277e-01 -8.24779093e-01
8.30003977e-01 -4.83531922e-01 -8.21028650e-01 2.88177073e-01
-9.36117232e-01 3.89625341e-01 6.95174634e-01 -6.36494875e-01
-7.46665895e-01 2.89416146e-02 9.67052355e-02 -4.50863808e-01
-6.16853014e-02 7.10248590e-01 2.03207523e-01 6.26432821e-02
7.28999436e-01 5.41960560e-02 -2.13556126e-01 -4.70068276e-01
3.39179873e-01 1.60611913e-01 -5.22970185e-02 -7.93277144e-01
6.16628170e-01 -1.49418972e-02 -5.64776324e-02 -3.61692607e-01
-9.15439248e-01 -3.17240745e-01 -4.73086983e-01 -1.88567817e-01
4.55975801e-01 -6.84979677e-01 -3.36586505e-01 1.28480643e-01
-5.21589100e-01 -1.35636985e+00 -5.00905514e-01 3.87141079e-01
-6.53405845e-01 -2.02773795e-01 -5.68395853e-01 -4.68192071e-01
-1.75036341e-01 -1.16953599e+00 8.87828231e-01 3.04028124e-01
-1.17369570e-01 -1.21285331e+00 3.64746779e-01 1.71942666e-01
6.12489641e-01 -1.56847432e-01 1.09070516e+00 -6.78682446e-01
-5.52689135e-01 7.65125602e-02 2.10111573e-01 1.15768075e-01
4.74795438e-02 -2.96251595e-01 -5.91722250e-01 -5.45399785e-01
-4.78119075e-01 -7.34909654e-01 8.97924781e-01 4.84834135e-01
1.46863675e+00 -4.50322717e-01 -2.69346923e-01 8.34254563e-01
1.40170157e+00 2.86380857e-01 3.18987012e-01 9.96404171e-01
2.95766950e-01 -3.44598144e-02 9.32471573e-01 6.41587675e-01
1.64698616e-01 7.02056944e-01 4.37965125e-01 -3.75079773e-02
-7.77883232e-02 -4.87207681e-01 3.05406898e-01 4.96592194e-01
8.57315511e-02 -1.44134134e-01 -9.01397347e-01 5.89656353e-01
-1.99859786e+00 -6.38545752e-01 7.38129020e-01 2.22525430e+00
1.05683410e+00 2.31354699e-01 4.28108543e-01 -4.36051399e-01
4.72849876e-01 2.37614721e-01 -9.98468518e-01 -5.77644825e-01
1.82380289e-01 4.46504623e-01 6.18912518e-01 6.43865705e-01
-1.04020894e+00 1.33402741e+00 6.08867168e+00 9.50616479e-01
-1.05033517e+00 3.77692848e-01 8.20435107e-01 -5.33450663e-01
-1.39270797e-01 6.51568323e-02 -7.78384328e-01 2.24580422e-01
8.96268189e-01 -2.50639990e-02 9.29879785e-01 1.12966466e+00
1.59455970e-01 -3.87956202e-02 -9.81070220e-01 5.01700819e-01
-1.19897291e-01 -1.38783669e+00 -7.94633478e-02 -3.92303243e-02
1.02686357e+00 3.30080658e-01 3.31285775e-01 6.81955338e-01
6.37131572e-01 -9.23087060e-01 4.82559204e-01 2.71213919e-01
4.12841886e-01 -9.63116765e-01 3.84035081e-01 2.66280562e-01
-1.13758206e+00 -5.72739184e-01 -3.39655310e-01 2.43958339e-01
-2.15511426e-01 1.85462058e-01 -1.06681740e+00 1.38392106e-01
8.61294210e-01 3.70171875e-01 -8.88950944e-01 1.07266176e+00
-4.22844023e-01 8.22214067e-01 -2.62484670e-01 -2.91051865e-01
6.15199506e-01 -1.60522223e-01 6.09068453e-01 1.01383173e+00
1.01137720e-01 -3.95719826e-01 4.18975115e-01 8.08948278e-01
-1.26612246e-01 1.89849690e-01 -2.28466138e-01 7.85194933e-02
8.09076548e-01 1.13116300e+00 -5.88044286e-01 -2.33564734e-01
-1.01804577e-01 9.71475363e-01 9.91239071e-01 6.56906962e-01
-9.23606336e-01 -2.78626770e-01 4.20876771e-01 -1.32090256e-01
5.91841042e-01 -2.68706441e-01 -2.98845172e-01 -7.63563097e-01
-6.96759671e-02 -1.02518916e+00 7.85579264e-01 -5.29715955e-01
-7.60202348e-01 6.05370224e-01 -7.83476140e-03 -9.82790291e-01
-2.40518928e-01 -3.99976343e-01 -6.53029084e-01 6.45621479e-01
-1.47987258e+00 -9.67198789e-01 -8.28421488e-02 3.34837765e-01
1.04864538e+00 -4.98175472e-01 8.54143620e-01 -2.66470641e-01
-7.68671572e-01 9.28981543e-01 2.65813857e-01 -1.57378912e-01
7.57786036e-01 -1.43061793e+00 5.19329250e-01 5.59286773e-01
1.52307674e-01 6.39026463e-01 4.46622968e-01 -6.71880305e-01
-1.21037364e+00 -1.10955656e+00 6.39768660e-01 -1.95396692e-01
7.99301565e-01 -2.50067860e-01 -6.51719928e-01 6.30125284e-01
1.32987991e-01 -9.33736265e-02 5.11032522e-01 5.07487237e-01
-3.29768747e-01 -2.68556505e-01 -8.91971648e-01 7.51306891e-01
9.43886042e-01 -2.02777922e-01 -2.29320988e-01 5.58773160e-01
6.45843446e-01 -5.54440379e-01 -9.07772064e-01 2.11515382e-01
3.39734703e-01 -7.21590936e-01 8.41185689e-01 -9.05228436e-01
4.40816790e-01 -8.51731226e-02 2.11199410e-02 -1.83218193e+00
-6.46132052e-01 -6.80470288e-01 -4.41199332e-01 5.27959526e-01
9.31534171e-01 -5.88214099e-01 1.00929141e+00 3.79553974e-01
-2.16321751e-01 -1.37647247e+00 -9.45800126e-01 -9.17663574e-01
1.64306358e-01 -6.67570978e-02 8.86113584e-01 7.59216368e-01
8.94391835e-02 1.56826437e-01 -1.34300798e-01 1.82815343e-01
4.99836147e-01 1.69194952e-01 7.24427819e-01 -6.83234036e-01
-6.76687300e-01 -6.24792933e-01 1.85675859e-01 -9.56356823e-01
-2.13539638e-02 -8.74635816e-01 -1.31279692e-01 -1.40648532e+00
-2.42493162e-03 -6.49296761e-01 -6.85016632e-01 7.70485282e-01
-3.63585413e-01 -1.61936283e-01 1.54133230e-01 1.13992266e-01
-9.33978796e-01 7.75369525e-01 1.24580491e+00 -6.14746399e-02
-7.13949502e-01 2.02676386e-01 -4.87746179e-01 4.74936604e-01
1.45117199e+00 -5.34003973e-01 -4.82182235e-01 -4.04945076e-01
3.67396623e-01 8.26912150e-02 1.47351056e-01 -9.79143441e-01
2.36599281e-01 -4.57102090e-01 6.25843704e-01 -3.53848130e-01
2.25393280e-01 -5.74472308e-01 -3.77272636e-01 6.99093521e-01
-9.32661414e-01 6.17047727e-01 5.15802801e-01 5.41377187e-01
9.64706615e-02 -3.13035041e-01 8.05266500e-01 -2.11976454e-01
-7.63148308e-01 4.36766982e-01 -2.24598125e-01 2.48529688e-01
8.65528882e-01 1.70043021e-01 -2.14469492e-01 -2.47722670e-01
-1.00029647e+00 6.54275954e-01 3.40350121e-01 4.71859783e-01
7.52368510e-01 -1.31559455e+00 -4.68342543e-01 2.15360954e-01
2.06845433e-01 2.26794854e-02 4.12086304e-03 6.53201103e-01
-3.46499860e-01 3.78840625e-01 -2.08985992e-02 -2.99290240e-01
-1.06882715e+00 7.41832793e-01 4.27394092e-01 -5.94432473e-01
-5.32141387e-01 1.11619115e+00 -1.71663418e-01 -5.71273744e-01
5.26416302e-01 -4.75767970e-01 7.95304496e-03 -4.14215982e-01
3.13365012e-01 3.29643726e-01 -2.09876150e-02 8.09946805e-02
-2.35859126e-01 2.29913116e-01 -4.53673035e-01 -2.98583478e-01
1.48952675e+00 1.19836614e-01 4.51450683e-02 4.29222584e-01
1.03149211e+00 -2.40006357e-01 -1.62118173e+00 -3.53764802e-01
8.94125402e-02 -4.23508823e-01 1.28359005e-01 -1.27278268e+00
-1.01450145e+00 4.54739928e-01 6.10803485e-01 -3.45288694e-01
1.21805406e+00 -6.96857870e-02 4.81990129e-01 7.26100385e-01
2.87308812e-01 -1.70158458e+00 5.78262091e-01 6.70654833e-01
1.08907115e+00 -1.01481760e+00 2.53302976e-02 2.09157899e-01
-8.74892533e-01 1.01217198e+00 1.05211818e+00 -1.96025595e-01
6.91875219e-01 9.62042585e-02 -1.41172215e-01 -2.29778752e-01
-1.18512511e+00 -6.84533790e-02 1.44624561e-01 3.92818749e-01
1.54473811e-01 1.01635203e-01 -1.35804430e-01 3.13318551e-01
-8.48102048e-02 -1.66969374e-01 -2.91725453e-02 1.03022528e+00
-5.45305073e-01 -1.23360109e+00 -2.26505145e-01 4.75003213e-01
1.01487041e-01 -2.46138304e-01 -3.92688662e-01 7.77713835e-01
-1.15881175e-01 5.21673858e-01 5.38324239e-03 -3.32064539e-01
1.57901958e-01 2.30474174e-01 4.99288082e-01 -5.60580194e-01
-7.23533452e-01 2.53363609e-01 7.26609975e-02 -8.33858311e-01
1.15489839e-02 -6.87776268e-01 -1.25644636e+00 9.11433548e-02
-1.97466493e-01 3.89376223e-01 7.16260970e-01 6.86879814e-01
6.05397165e-01 6.79685771e-01 8.58553767e-01 -7.59231031e-01
-9.50140536e-01 -9.63145792e-01 -2.98254758e-01 7.59406760e-02
2.58646846e-01 -5.98243058e-01 -2.68479496e-01 -4.40651208e-01] | [4.135000705718994, 1.954849362373352] |
b10d2d35-dfbc-4867-b206-c983fb20b34a | no-reference-image-quality-assessment-in-the | null | null | https://ieeexplore.ieee.org/document/6272356 | https://live.ece.utexas.edu/publications/2012/TIP%20BRISQUE.pdf | No-Reference Image Quality Assessment in the Spatial Domain | We propose a natural scene statistic-based distortion-generic blind/no-reference (NR) image quality assessment (IQA) model that operates in the spatial domain. The new model, dubbed blind/referenceless image spatial quality evaluator (BRISQUE) does not compute distortion-specific features, such as ringing, blur, or blocking, but instead uses scene statistics of locally normalized luminance coefficients to quantify possible losses of “naturalness” in the image due to the presence of distortions, thereby leading to a holistic measure of quality. The underlying features used derive from the empirical distribution of locally normalized luminances and products of locally normalized luminances under a spatial natural scene statistic model. No transformation to another coordinate frame (DCT, wavelet, etc.) is required, distinguishing it from prior NR IQA approaches. Despite its simplicity, we are able to show that BRISQUE is statistically better than the full-reference peak signal-to-noise ratio and the structural similarity index, and is highly competitive with respect to all present-day distortion-generic NR IQA algorithms. BRISQUE has very low computational complexity, making it well suited for real time applications. BRISQUE features may be used for distortion-identification as well. To illustrate a new practical application of BRISQUE, we describe how a nonblind image denoising algorithm can be augmented with BRISQUE in order to perform blind image denoising. Results show that BRISQUE augmentation leads to performance improvements over state-of-the-art methods. A software release of BRISQUE is available online: http://live.ece.utexas.edu/research/quality/BRISQUE_release.zip for public use and evaluation. | ['and Alan Conrad Bovik', 'Anush Krishna Moorthy', 'Anish Mittal'] | 2012-08-17 | null | null | null | ieee-transacations-on-image-processing-2012-8 | ['no-reference-image-quality-assessment'] | ['computer-vision'] | [ 3.48975956e-01 -7.69561827e-01 3.06315720e-01 -1.51993468e-01
-1.28634751e+00 -7.32045949e-01 6.60411239e-01 -1.48646096e-02
-3.60761017e-01 4.57722366e-01 4.77713376e-01 -3.04566503e-01
-4.65048194e-01 -5.14854252e-01 -4.37554389e-01 -9.81823862e-01
-3.67582738e-02 -4.71675694e-01 9.25337300e-02 -2.63934672e-01
4.13288116e-01 6.57610118e-01 -1.69265652e+00 1.33817628e-01
8.32627535e-01 1.01050973e+00 1.78568214e-01 1.05301642e+00
5.06582916e-01 7.21460760e-01 -6.78474903e-01 -3.75565022e-01
5.46364188e-01 -6.94308698e-01 -5.70966780e-01 5.40881837e-03
5.50755322e-01 -4.27343547e-01 -3.17846537e-01 1.30194306e+00
8.02247465e-01 -4.05906737e-02 6.76870823e-01 -8.97120297e-01
-7.55889475e-01 -2.93646157e-01 -2.68411636e-01 4.35663462e-01
7.11567283e-01 3.72693956e-01 7.49641120e-01 -1.06432509e+00
3.37508559e-01 1.14623904e+00 7.20118225e-01 2.63125915e-02
-1.39417255e+00 -4.08842683e-01 -4.38020974e-01 4.07385290e-01
-1.56391072e+00 -8.23808134e-01 4.91014659e-01 -3.98329079e-01
5.75764179e-01 7.69333124e-01 2.68107235e-01 6.21073306e-01
5.90230301e-02 4.15330023e-01 1.60058415e+00 -6.09201133e-01
3.04518968e-01 -1.26497403e-01 -4.64746431e-02 4.33011532e-01
1.93146020e-01 6.34089530e-01 -4.16509002e-01 -1.67652324e-01
7.51439273e-01 -3.38126451e-01 -7.08836317e-01 -3.14912766e-01
-1.25979102e+00 5.17259657e-01 3.89096767e-01 5.12477160e-01
-3.41016799e-01 7.04790326e-03 1.47646055e-01 6.51248395e-01
4.41983968e-01 3.22870404e-01 -1.41261443e-02 -1.73515379e-01
-1.17401087e+00 2.73374021e-01 4.14229184e-01 6.07169628e-01
7.34422147e-01 9.56489593e-02 -3.95291150e-01 9.76309240e-01
1.61904082e-01 7.72070408e-01 3.20841402e-01 -1.44211066e+00
8.04188699e-02 -3.51051278e-02 2.84253091e-01 -1.08121157e+00
-1.54892132e-01 -4.81356204e-01 -1.00712550e+00 7.07165956e-01
3.85807753e-01 3.73654008e-01 -7.48338342e-01 1.44318640e+00
-2.18580216e-02 -8.56769606e-02 3.70740145e-02 1.05500102e+00
5.61225772e-01 5.98686993e-01 -4.05502439e-01 -4.63603318e-01
1.20256221e+00 -4.53620583e-01 -7.63458967e-01 8.98591429e-02
1.83291286e-01 -1.15287566e+00 9.79383349e-01 8.45729411e-01
-1.30546904e+00 -8.40362906e-01 -1.03523958e+00 -3.49891186e-02
-1.77508995e-01 -7.95095414e-02 1.26182362e-01 1.22067165e+00
-1.61074698e+00 5.62089682e-01 -5.02445340e-01 -3.93842161e-01
1.35320112e-01 7.97215924e-02 -4.69957739e-01 -5.77443242e-01
-8.61653626e-01 9.06351626e-01 -1.67430326e-01 -7.31812604e-03
-8.64582300e-01 -6.07075810e-01 -8.87833059e-01 -2.18541265e-01
7.96146467e-02 -6.21992111e-01 1.06012082e+00 -9.96481895e-01
-1.31055009e+00 8.39446664e-01 -5.67459106e-01 -2.74632066e-01
4.81439650e-01 1.96962114e-02 -7.83642113e-01 3.90853405e-01
1.56160623e-01 1.89538151e-01 1.12538540e+00 -1.51892269e+00
-2.83460230e-01 -2.95269668e-01 -7.70771503e-02 1.83721080e-01
-1.37018422e-02 4.93149936e-01 -5.89847386e-01 -1.06342411e+00
2.38649413e-01 -3.11742067e-01 1.33776367e-01 1.97644502e-01
-2.11490795e-01 3.11890334e-01 4.69846815e-01 -9.99723196e-01
1.15296018e+00 -2.39054918e+00 -2.99329251e-01 2.16466099e-01
5.74442707e-02 4.35345411e-01 -4.84780431e-01 3.64787489e-01
-3.19887429e-01 -1.39612621e-02 -7.05965936e-01 -1.04359463e-01
-6.23989142e-02 -4.30484004e-02 -3.21590081e-02 1.00501072e+00
-1.43270278e-02 4.46857154e-01 -8.85782957e-01 -1.49029270e-01
6.14892185e-01 6.76951766e-01 -2.90169477e-01 1.49287552e-01
5.54881811e-01 5.67992926e-01 1.44812718e-01 7.60713041e-01
1.18637133e+00 2.40678027e-01 -3.92173529e-01 -4.12677377e-01
-3.46850842e-01 1.00166656e-01 -1.36662662e+00 1.46931374e+00
-5.12441039e-01 7.92508543e-01 4.00042921e-01 -9.22468841e-01
8.43221009e-01 4.16915029e-01 2.08209604e-01 -1.18680406e+00
-1.89084202e-01 4.12344337e-01 -1.03351407e-01 -4.17496383e-01
3.15298587e-01 -1.55630842e-01 3.75943780e-01 1.42927960e-01
1.35445699e-01 -4.69388813e-01 2.31008768e-01 1.44635111e-01
1.23067713e+00 -1.52453348e-01 4.33951527e-01 -3.24851573e-01
8.52726758e-01 -4.57343906e-01 3.16183239e-01 8.00549805e-01
-4.80999559e-01 1.09530306e+00 1.64298579e-01 8.95958468e-02
-1.19401073e+00 -1.50081205e+00 -5.09200931e-01 7.15716958e-01
2.40367681e-01 -2.89430052e-01 -7.62350321e-01 -4.96896990e-02
-2.46307552e-01 6.15304351e-01 -1.65116981e-01 5.78388423e-02
-2.21797347e-01 -7.95729399e-01 4.39795434e-01 6.43269718e-02
6.22042239e-01 -7.31391907e-01 -1.97794944e-01 2.94323992e-02
-2.95750052e-01 -9.82210577e-01 -4.72157180e-01 -2.89671600e-01
-6.72415853e-01 -1.16590750e+00 -1.23106992e+00 -3.65704924e-01
3.98703367e-01 6.13838196e-01 1.01491332e+00 -6.05366193e-02
-3.95685524e-01 7.60288358e-01 -4.24618900e-01 8.35297555e-02
-6.06952310e-01 -8.71037662e-01 1.27635077e-01 3.03019822e-01
3.26056257e-02 -6.33397162e-01 -9.57264066e-01 5.82519531e-01
-1.16632032e+00 -3.91553044e-01 4.93126839e-01 7.82610357e-01
5.23075640e-01 4.46076274e-01 2.99880207e-01 -2.40619197e-01
7.65087187e-01 2.15966883e-03 -6.81168020e-01 -2.18373630e-03
-6.46513224e-01 -2.67187953e-01 5.26941121e-01 1.49050616e-02
-1.03210914e+00 -2.98522174e-01 -2.71182060e-01 -1.84274137e-01
-4.37465757e-01 2.37460554e-01 -2.37977386e-01 -4.88641649e-01
8.87396574e-01 4.54542994e-01 -2.35225502e-02 -6.37300491e-01
3.17094117e-01 8.36005509e-01 1.07047927e+00 -2.06778780e-01
1.00793505e+00 5.97993612e-01 -2.06605252e-02 -1.00543582e+00
-2.78047949e-01 -9.37281787e-01 -2.74798632e-01 -1.98920578e-01
4.92533177e-01 -1.04958355e+00 -6.46344483e-01 8.51716280e-01
-9.66105819e-01 -1.87050670e-01 -2.86247909e-01 5.62206149e-01
-7.47279406e-01 7.37332582e-01 -4.24586207e-01 -9.65981603e-01
-4.14153300e-02 -1.22448397e+00 8.94079328e-01 -6.09574886e-03
5.52776381e-02 -7.94733822e-01 4.77134548e-02 3.95757914e-01
7.30948150e-01 1.16125733e-01 6.18861258e-01 1.61171425e-02
-4.85749513e-01 -2.34030694e-01 -5.40992439e-01 1.02751577e+00
2.80674309e-01 -2.48720571e-01 -1.24949455e+00 -5.66286504e-01
4.59122926e-01 1.20279089e-01 8.23292673e-01 6.29405856e-01
9.61912155e-01 -3.81735176e-01 4.17517245e-01 8.50969732e-01
1.77838063e+00 1.78889886e-01 1.09728277e+00 3.60149980e-01
2.69468188e-01 4.78577584e-01 5.14981627e-01 3.31423849e-01
3.28014642e-02 9.06393826e-01 4.35770065e-01 -4.65354502e-01
-6.67967498e-01 2.15376616e-01 6.07772529e-01 7.04592526e-01
-2.75924236e-01 -1.71186283e-01 -7.43933320e-01 7.51545191e-01
-1.22533381e+00 -1.11670530e+00 -4.69442993e-01 2.58218765e+00
6.12290978e-01 -1.98422462e-01 2.77180411e-03 6.81198180e-01
7.83477068e-01 2.27525026e-01 -2.73207635e-01 -4.46096569e-01
-6.07858181e-01 4.16512042e-01 8.29652965e-01 7.87771642e-01
-1.03945005e+00 3.50899994e-01 6.08170700e+00 1.08646643e+00
-7.07679510e-01 3.75421345e-01 5.09727895e-01 2.89377064e-01
-1.00739591e-01 -2.12121289e-02 5.92901967e-02 3.77372861e-01
9.17157531e-01 -1.98467463e-01 7.25687504e-01 2.66664714e-01
8.00385475e-01 -4.70554113e-01 -7.17766464e-01 1.38847554e+00
2.27827892e-01 -8.74660313e-01 -9.41286013e-02 1.66965365e-01
6.30931795e-01 -9.01551247e-02 4.19056267e-01 -4.34872836e-01
4.41459417e-02 -1.01802206e+00 8.33122492e-01 7.14143455e-01
1.17694962e+00 -6.50511801e-01 8.63102555e-01 -3.93119790e-02
-9.74169016e-01 1.32532408e-02 -2.93091059e-01 1.27911642e-01
1.49842560e-01 7.81263709e-01 -1.62984967e-01 8.75259936e-01
9.08457637e-01 6.14358127e-01 -7.63701022e-01 1.48969078e+00
-2.27499679e-01 6.21351600e-01 -1.17446549e-01 7.15270519e-01
-6.75959215e-02 -2.33076096e-01 9.45440292e-01 1.25757766e+00
5.76588094e-01 5.56754470e-02 -4.26490009e-01 7.15346575e-01
2.41420716e-01 9.94140655e-02 -4.57475275e-01 4.13865149e-01
2.76472569e-01 9.37336028e-01 -3.51658553e-01 -2.03142926e-01
-4.32430267e-01 1.38639224e+00 -7.29412436e-01 6.82553530e-01
-3.65730733e-01 -6.00532234e-01 6.84728622e-01 1.12855442e-01
2.45002121e-01 -2.08804965e-01 -2.27946132e-01 -1.00686145e+00
2.82590300e-01 -1.22159994e+00 3.94367166e-02 -1.27880585e+00
-1.41031647e+00 5.29278696e-01 -1.16143197e-01 -1.61563933e+00
-2.25422867e-02 -6.99754834e-01 -5.31262696e-01 1.32109571e+00
-1.66041625e+00 -7.85121024e-01 -3.24825376e-01 9.87558305e-01
2.90570647e-01 -8.69107991e-02 7.95371473e-01 3.05601776e-01
-9.65861231e-02 5.56438088e-01 4.93023396e-01 1.06545009e-01
9.54156339e-01 -1.29275501e+00 3.52867484e-01 1.43254960e+00
4.39846925e-02 4.32067692e-01 9.56001401e-01 -1.82444468e-01
-1.26672888e+00 -8.17028224e-01 7.38988876e-01 -2.55731910e-01
5.05185187e-01 -9.46256667e-02 -8.56838226e-01 3.70287597e-02
1.88343406e-01 2.92939156e-01 4.78073508e-01 -4.01623368e-01
-5.40594637e-01 -4.35633779e-01 -1.33005083e+00 3.35162550e-01
8.17118704e-01 -8.91781449e-01 -4.67755228e-01 1.79897442e-01
3.96964312e-01 1.22761289e-02 -8.29325378e-01 4.93690461e-01
3.47322166e-01 -1.68894064e+00 1.27408278e+00 4.00417745e-01
1.08446712e-02 -7.21414924e-01 -5.10385334e-01 -1.14431739e+00
-3.79062980e-01 -7.49282181e-01 1.79401696e-01 1.20916879e+00
1.28497094e-01 -7.36486733e-01 5.68022132e-02 8.38766918e-02
-1.04914129e-01 4.42840420e-02 -1.09151518e+00 -1.13260317e+00
-1.50171548e-01 -7.50423610e-01 3.41001958e-01 7.98488677e-01
-3.76970679e-01 -1.52398482e-01 -3.29916984e-01 4.15611207e-01
9.92157698e-01 -2.85979629e-01 6.04402125e-01 -8.66929293e-01
-2.74872184e-01 -4.67322111e-01 -7.79965281e-01 -8.43853116e-01
-5.21281779e-01 -5.41858196e-01 2.19854284e-02 -1.65288997e+00
-2.01023892e-02 -2.27604553e-01 -4.33030099e-01 1.02862053e-01
-2.38974430e-02 7.64361262e-01 2.51014024e-01 3.21300387e-01
-3.43808681e-01 3.52998257e-01 1.04856682e+00 -1.61126792e-01
-1.07169069e-01 1.48359224e-01 -6.53681397e-01 4.62177068e-01
6.98463917e-01 -2.74789959e-01 -1.07417285e-01 -3.40472311e-01
4.50419188e-02 1.36564225e-01 9.52101052e-01 -1.36073267e+00
2.65039504e-01 2.84091055e-01 2.78194666e-01 -3.28068316e-01
3.31778944e-01 -6.96568072e-01 1.64002001e-01 2.89481074e-01
5.61613031e-02 7.29485825e-02 2.90947985e-02 5.13922095e-01
-6.84732616e-01 -2.40121037e-01 1.07126141e+00 -4.43130247e-02
-6.44095123e-01 -2.56699435e-02 -5.93809724e-01 -3.55525792e-01
4.22026664e-01 -6.07250929e-01 -4.00977105e-01 -6.63162649e-01
-5.00041902e-01 -5.35063148e-01 8.37054372e-01 -6.12438843e-03
7.89597571e-01 -1.17018521e+00 -1.05827796e+00 2.61881769e-01
3.22437465e-01 -6.27281845e-01 5.85479677e-01 1.04291952e+00
-7.69448638e-01 1.45242363e-01 -1.61554053e-05 -5.51519036e-01
-1.23212314e+00 7.70413220e-01 4.54969019e-01 4.24167030e-02
-3.96674514e-01 5.74458539e-01 1.83245748e-01 -1.55574444e-03
9.63293463e-02 -2.29187012e-02 4.85180430e-02 -3.07137340e-01
9.17157292e-01 6.43309355e-01 4.75659400e-01 -1.00930190e+00
-3.12145472e-01 7.37783253e-01 4.85571444e-01 -5.95928133e-01
1.09756970e+00 -6.27442420e-01 -3.27223301e-01 1.77145347e-01
1.37493742e+00 3.56197357e-01 -9.68560457e-01 -1.98727638e-01
-1.81145296e-01 -9.06619251e-01 5.18212318e-01 -1.07883215e+00
-8.99311304e-01 8.90242994e-01 1.38495660e+00 3.54261845e-01
1.95380199e+00 -2.56213248e-01 2.26512507e-01 -8.97407085e-02
2.35254705e-01 -6.46777034e-01 -1.50081456e-01 2.12158635e-01
1.17253172e+00 -1.08112931e+00 8.62003118e-02 -3.19804370e-01
-2.28805885e-01 8.36088359e-01 -3.39042395e-01 -1.70413870e-02
5.59447825e-01 1.28350049e-01 2.98250258e-01 1.39968544e-01
-1.46726891e-01 -5.63801885e-01 4.04610693e-01 1.10900497e+00
3.27423483e-01 -1.23307534e-01 -2.87654042e-01 -1.26461321e-02
-2.06982777e-01 -1.24990806e-01 5.40553927e-01 5.96080959e-01
-2.97828168e-01 -1.07034755e+00 -1.03538537e+00 1.04952931e-01
-5.87240577e-01 -4.09057915e-01 2.60585919e-02 4.32692349e-01
2.76116699e-01 1.52547669e+00 -1.59643814e-01 -2.18026191e-01
4.15269375e-01 -2.89056927e-01 4.58244354e-01 -1.85558379e-01
-4.67064917e-01 3.07215184e-01 -1.37928307e-01 -8.40258360e-01
-8.73610795e-01 -5.37408531e-01 -4.24471110e-01 -5.31080008e-01
-1.04895838e-01 -1.30913973e-01 7.95806646e-01 5.20365119e-01
6.90153167e-02 2.29234725e-01 8.45390797e-01 -7.92217433e-01
-3.71878259e-02 -8.14913034e-01 -8.44841301e-01 5.18117189e-01
9.58200574e-01 -2.60816127e-01 -6.79024458e-01 3.62811238e-01] | [11.719024658203125, -1.981359839439392] |
45b9e042-8465-425d-b342-250a407e2b79 | optimized-preprocessing-and-tiny-ml-for | 2303.11371 | null | https://arxiv.org/abs/2303.11371v1 | https://arxiv.org/pdf/2303.11371v1.pdf | Optimized preprocessing and Tiny ML for Attention State Classification | In this paper, we present a new approach to mental state classification from EEG signals by combining signal processing techniques and machine learning (ML) algorithms. We evaluate the performance of the proposed method on a dataset of EEG recordings collected during a cognitive load task and compared it to other state-of-the-art methods. The results show that the proposed method achieves high accuracy in classifying mental states and outperforms state-of-the-art methods in terms of classification accuracy and computational efficiency. | ['Van-Tam Nguyen', 'Pavlo Mozharovskyi', 'Enzo Tartaglione', 'Rémi Nahon', 'Yinghao Wang'] | 2023-03-20 | null | null | null | null | ['eeg', 'eeg'] | ['methodology', 'time-series'] | [ 3.60761732e-01 -3.76345813e-01 -1.40829101e-01 -5.57143748e-01
-4.28108215e-01 -4.12249118e-02 4.62467492e-01 3.60991567e-01
-6.69275403e-01 1.09869325e+00 -2.04658896e-01 -1.28211394e-01
-3.53266269e-01 -4.87913162e-01 -7.26470053e-02 -4.76655930e-01
-4.08509165e-01 2.68087804e-01 9.40529406e-02 -6.76113516e-02
7.33873785e-01 4.22484756e-01 -1.65551591e+00 4.11044717e-01
9.64128673e-01 1.24601185e+00 7.40156919e-02 2.57757336e-01
2.41276503e-01 6.64973021e-01 -9.44256544e-01 2.18313839e-02
-1.74668297e-01 -5.75149357e-01 -9.49643373e-01 -5.92859164e-02
-1.84280753e-01 3.30271363e-01 -2.29302704e-01 9.65499878e-01
7.95416832e-01 3.59689057e-01 6.78257227e-01 -1.16270077e+00
-2.83496827e-01 1.61298126e-01 -2.82641172e-01 1.09794974e+00
9.47689593e-01 -3.00558090e-01 3.43701661e-01 -6.94878936e-01
9.64255929e-02 8.57004344e-01 4.07517791e-01 2.72147506e-01
-1.44764304e+00 -1.13114655e+00 1.52761452e-02 7.97977269e-01
-1.49706256e+00 -6.64991081e-01 9.33751106e-01 -5.59680223e-01
1.20289898e+00 1.29412964e-01 9.21194255e-01 9.03750360e-01
7.28795350e-01 7.30614662e-01 2.02956963e+00 -4.27895755e-01
7.70465374e-01 1.11768469e-01 8.99794757e-01 5.41046023e-01
6.52725920e-02 -1.46719053e-01 -9.18262899e-01 -3.99935037e-01
2.76942700e-01 -2.38782749e-01 -2.04000592e-01 8.02433044e-02
-1.08159590e+00 4.47043270e-01 -9.75914001e-02 6.29269719e-01
-9.70610619e-01 -2.82073945e-01 4.15045977e-01 3.71256500e-01
6.00170910e-01 4.52290833e-01 -2.95894325e-01 -4.57371920e-01
-1.36267173e+00 4.01605181e-02 6.56482935e-01 4.31715429e-01
3.89411777e-01 1.75967082e-01 -4.07818943e-01 7.79529512e-01
-2.51816437e-02 3.02759111e-01 9.81611311e-01 -3.18248719e-01
9.38006639e-02 5.83298326e-01 -8.05515423e-03 -6.85668170e-01
-8.76643598e-01 -4.75468308e-01 -6.73010051e-01 -1.20303230e-02
-4.63663526e-02 -1.62435398e-01 -6.09455943e-01 1.50934315e+00
-4.19406116e-01 4.16905940e-01 1.61305979e-01 6.17866099e-01
6.41435325e-01 6.15872025e-01 1.78548411e-01 -8.08536947e-01
1.34864044e+00 -2.47791365e-01 -1.37847888e+00 -3.21279854e-01
2.39875130e-02 -3.20677251e-01 7.48992324e-01 1.06287014e+00
-1.26224554e+00 -8.37685883e-01 -1.20552230e+00 6.48962736e-01
-3.71861130e-01 3.95312726e-01 8.38464260e-01 9.35576379e-01
-8.88311565e-01 4.87946212e-01 -9.44317877e-01 -5.28097630e-01
4.86566037e-01 5.94615459e-01 -2.09718063e-01 3.31466943e-01
-1.15264392e+00 1.22219896e+00 5.97595692e-01 -1.15678087e-01
-7.51505554e-01 -1.84189722e-01 -5.75216472e-01 8.74035135e-02
-2.51609702e-02 -2.03033909e-01 1.02624214e+00 -7.20730841e-01
-1.79780972e+00 7.10757017e-01 -6.41316295e-01 -3.86628211e-01
-9.95522439e-02 -4.73468676e-02 -9.29191411e-01 3.39785606e-01
-2.99233347e-01 1.72428638e-02 7.18424141e-01 -7.45654762e-01
-3.30450267e-01 -6.19947016e-01 -3.12043786e-01 -8.99924245e-03
-2.93093145e-01 2.38088682e-01 2.74233460e-01 -3.22665930e-01
3.80146027e-01 -8.00745189e-01 3.41362655e-01 -7.45503068e-01
-1.88761931e-02 -4.51026022e-01 4.64726090e-01 -6.50823772e-01
1.40943754e+00 -1.87745571e+00 1.09869391e-01 4.82294202e-01
-1.30508959e-01 3.28826874e-01 3.20283294e-01 2.33224288e-01
-4.31524634e-01 -3.93372387e-01 -7.94186965e-02 -2.32862592e-01
-3.59000848e-03 2.16540750e-02 7.49024600e-02 6.61153376e-01
-1.34546399e-01 6.75014555e-01 -8.32973540e-01 -2.79135138e-01
2.98616111e-01 1.52353510e-01 5.48153669e-02 3.32673669e-01
5.97033262e-01 4.11148995e-01 -1.96462512e-01 5.40890455e-01
4.17661875e-01 2.09345132e-01 4.02302682e-01 -1.78545907e-01
-6.27989694e-02 4.96833086e-01 -1.08226049e+00 1.62149668e+00
-2.37730160e-01 8.02535594e-01 -4.44691479e-01 -1.48845589e+00
1.08319318e+00 6.76871657e-01 4.42767978e-01 -8.10980976e-01
5.44861317e-01 -3.54454517e-02 2.45243892e-01 -7.94778228e-01
5.24146594e-02 -1.98293030e-01 -6.01501353e-02 6.77336395e-01
4.35780883e-01 1.66174974e-02 3.61416221e-01 -2.43613839e-01
1.09950852e+00 -2.52641946e-01 8.21954668e-01 -8.14958751e-01
8.23246419e-01 -5.60500026e-01 3.85010004e-01 6.08392298e-01
-5.61353505e-01 -3.02725554e-01 3.49779844e-01 -3.16569179e-01
-3.33663732e-01 -1.07265425e+00 -3.38252723e-01 1.06613243e+00
-1.68638855e-01 -3.04253489e-01 -1.18338788e+00 -1.29850730e-01
-3.89890015e-01 9.51367617e-01 -6.66775703e-01 -3.15913677e-01
-2.94965766e-02 -1.23173153e+00 4.02553797e-01 5.08397698e-01
6.90205216e-01 -1.19515705e+00 -9.27913725e-01 3.01105112e-01
-3.85967493e-01 -1.20592141e+00 3.67023230e-01 2.49624953e-01
-1.18331516e+00 -8.92913759e-01 -3.24724853e-01 -5.98352551e-01
5.06403565e-01 -6.75196499e-02 9.09355342e-01 -2.57957041e-01
-1.53170958e-01 1.96262226e-01 -2.94944465e-01 -8.15742075e-01
-7.64150843e-02 5.50950021e-02 5.04191399e-01 3.20330650e-01
8.58191669e-01 -8.82828712e-01 -4.93151218e-01 6.32467493e-02
-6.37863517e-01 -3.01868647e-01 6.41530871e-01 4.07040387e-01
4.11971271e-01 5.04220903e-01 1.19972575e+00 -6.12654448e-01
1.45101738e+00 -4.09008056e-01 -4.01412964e-01 1.30513251e-01
-9.07518446e-01 1.28087057e-02 3.14355403e-01 -5.53982615e-01
-9.89612818e-01 2.34248545e-02 -2.25056498e-03 6.19321950e-02
-5.25374889e-01 6.04046404e-01 8.60954896e-02 -2.35619023e-01
7.12006211e-01 7.39197075e-01 -2.54052967e-01 -4.09002304e-01
-2.84971237e-01 1.03819919e+00 6.34444952e-01 -3.07119846e-01
-1.24839783e-01 2.39203751e-01 -1.80702098e-02 -8.68478596e-01
-6.74711645e-01 -4.04020578e-01 -8.78660858e-01 -4.70381767e-01
7.99663603e-01 -7.28131831e-01 -7.96590686e-01 7.09649861e-01
-7.98397839e-01 5.56937270e-02 2.61000782e-01 8.07156622e-01
-8.47030222e-01 1.18616141e-01 -6.48710847e-01 -1.29754627e+00
-5.52128732e-01 -9.59075093e-01 5.35554111e-01 2.13076435e-02
-5.87357640e-01 -8.06606233e-01 2.33162627e-01 3.46077591e-01
4.11014557e-01 1.69059541e-02 8.46458912e-01 -7.27139771e-01
4.51248258e-01 -4.00703341e-01 1.81036696e-01 1.31479308e-01
1.21948905e-01 -6.96110785e-01 -1.36779881e+00 -3.16207647e-01
5.09972692e-01 -2.25124627e-01 5.33869743e-01 3.65307271e-01
1.32353771e+00 1.55373603e-01 -1.77033946e-01 5.77559471e-02
1.25128043e+00 7.74095953e-01 9.50624943e-01 2.37583533e-01
-2.03264505e-01 3.13745022e-01 3.31104189e-01 4.60525304e-01
2.30137929e-01 3.66416395e-01 -9.04604606e-03 1.96458295e-01
4.00706708e-01 4.16927427e-01 2.30421722e-01 8.54047835e-01
-4.08909857e-01 9.85051468e-02 -9.30677891e-01 2.87624657e-01
-1.94324255e+00 -1.18774295e+00 9.78014395e-02 2.05518222e+00
6.48682654e-01 3.94626081e-01 3.33610386e-01 1.00473440e+00
5.06864965e-01 -9.50166732e-02 -3.22796285e-01 -6.02613270e-01
2.05885306e-01 7.13904023e-01 -8.29564929e-02 5.84716424e-02
-1.00629127e+00 6.13208592e-01 8.31015778e+00 4.68793899e-01
-1.03597999e+00 4.85582113e-01 2.09642544e-01 -2.66294122e-01
7.83775151e-01 -7.32559383e-01 -1.45168230e-01 5.82896948e-01
1.43698430e+00 -2.09601000e-01 6.84279859e-01 3.86060596e-01
4.12267715e-01 -6.82923853e-01 -1.12396765e+00 1.45984912e+00
5.96826255e-01 -7.32856750e-01 -5.12892962e-01 -2.14186847e-01
4.17695582e-01 -1.76182091e-01 -2.92871147e-01 3.29638571e-01
-7.52338648e-01 -8.99168730e-01 5.79129040e-01 8.74411523e-01
3.85507703e-01 -8.58096063e-01 1.04400849e+00 5.55967987e-01
-9.00726914e-01 -3.13529074e-01 -1.96541265e-01 -8.29026699e-01
-5.83931617e-02 7.01254070e-01 -3.68810445e-01 4.34927881e-01
6.08882129e-01 5.79251826e-01 -7.59552181e-01 1.04631793e+00
-2.03047231e-01 7.65230000e-01 9.76887494e-02 -3.16562444e-01
-1.44663855e-01 -7.57983513e-03 1.30083650e-01 1.48210692e+00
2.77614504e-01 7.08977282e-01 -2.91665420e-02 7.56182432e-01
2.97981530e-01 4.10937928e-02 -4.76030856e-01 1.80772450e-02
3.86046618e-01 1.12992346e+00 -9.80363011e-01 -7.38206744e-01
-2.53566146e-01 8.87825429e-01 2.03273699e-01 9.04185846e-02
-5.18431306e-01 -5.42235851e-01 3.66253316e-01 -2.17264026e-01
-2.05008373e-01 -1.46417603e-01 -4.30871874e-01 -1.21877146e+00
4.73477095e-02 -6.42152131e-01 2.55637407e-01 -1.07688868e+00
-1.25689995e+00 8.55727971e-01 4.92135346e-01 -1.03151643e+00
-2.66987890e-01 -5.86129427e-01 -5.75344920e-01 8.26784074e-01
-1.37815487e+00 -2.61270732e-01 -3.45761061e-01 8.65018904e-01
4.63485956e-01 -4.49882597e-01 1.10846269e+00 3.28992933e-01
-5.16773522e-01 3.22402716e-01 -1.49632335e-01 -6.77561164e-02
5.58661461e-01 -8.53694975e-01 -1.13703437e-01 7.89338887e-01
-6.59545213e-02 5.21310091e-01 5.92118740e-01 -2.86142677e-01
-1.32223964e+00 -5.03223062e-01 8.30270588e-01 -1.79876521e-01
3.93231958e-01 -4.55184549e-01 -8.41211259e-01 4.02780265e-01
4.56134498e-01 -3.82085830e-01 1.09106195e+00 1.11679003e-01
2.67120898e-01 -9.73988771e-02 -1.38938999e+00 4.13316116e-02
6.63705766e-01 -7.28600383e-01 -1.43278193e+00 2.18134359e-01
-4.46751744e-01 2.25635841e-02 -9.44019556e-01 2.57914513e-01
7.01467216e-01 -9.90100443e-01 6.75156415e-01 -4.43638861e-01
-2.44697064e-01 -1.93203874e-02 -2.11728550e-02 -1.77228296e+00
-6.69096410e-01 -3.20493639e-01 -3.02985430e-01 7.42113411e-01
1.94140360e-01 -8.16205859e-01 3.80806595e-01 4.99978542e-01
2.02152357e-02 -7.10057616e-01 -8.56752038e-01 -8.38422954e-01
-3.37290168e-01 -7.49850690e-01 4.47357595e-01 1.00504017e+00
1.08338571e+00 6.22137606e-01 -1.24288984e-01 -9.18280333e-02
5.32715797e-01 1.33628339e-01 8.03069100e-02 -1.63219309e+00
6.42151311e-02 -4.20686454e-01 -1.04688561e+00 -1.19590305e-01
6.86622798e-01 -8.36306751e-01 -2.29462311e-01 -1.65564013e+00
4.31970239e-01 2.05362529e-01 -1.00741100e+00 6.63931549e-01
-1.63900167e-01 5.93044460e-01 4.89117205e-02 6.53425530e-02
-6.03643239e-01 2.16390610e-01 5.93783915e-01 1.36822034e-02
-3.82219225e-01 -2.45166086e-02 -5.44043005e-01 6.72305226e-01
1.22853529e+00 -4.95239884e-01 -5.47885418e-01 1.61264718e-01
-1.10493481e-01 2.65655249e-01 2.75891945e-02 -1.69315088e+00
1.44486636e-01 9.68019590e-02 7.34768927e-01 -5.22764385e-01
6.64243758e-01 -5.56571901e-01 1.17846489e-01 6.61511302e-01
-4.07607377e-01 2.20513865e-01 2.99731553e-01 1.56929985e-01
-9.91490483e-02 -6.75291643e-02 9.09471691e-01 6.93235639e-03
-5.91571510e-01 -2.48020872e-01 -1.03069758e+00 -3.26830506e-01
1.07691538e+00 -2.11503416e-01 -3.77880186e-01 -2.97921538e-01
-1.01330507e+00 -3.12763035e-01 -1.66949168e-01 3.96786779e-01
7.02673018e-01 -1.19110298e+00 -4.52519327e-01 7.48688698e-01
1.58001006e-01 -1.24685204e+00 -8.90085176e-02 1.30745614e+00
-3.12170084e-03 6.00861907e-01 -1.02423537e+00 -4.35315818e-01
-1.25630188e+00 5.03245950e-01 2.33936310e-01 -1.56415384e-02
-2.92558640e-01 1.94676191e-01 -4.25327450e-01 1.00307778e-01
1.16763838e-01 -2.63608754e-01 -5.81130147e-01 2.27486342e-01
9.10474062e-01 7.95845032e-01 6.32611036e-01 -4.70534801e-01
-6.10944629e-01 2.39582255e-01 4.93681669e-01 -3.45986903e-01
1.23447835e+00 2.03046221e-02 -3.32158744e-01 1.04680228e+00
7.72472203e-01 -4.94608432e-01 -3.64189565e-01 3.57484780e-02
5.64976707e-02 -3.94000232e-01 4.99756396e-01 -1.03236783e+00
-8.32188845e-01 9.40600753e-01 1.46970415e+00 3.24557990e-01
1.54406452e+00 -3.71253818e-01 4.17730451e-01 8.73869002e-01
8.57711315e-01 -1.49452865e+00 -5.53446263e-02 3.02151471e-01
3.68017793e-01 -1.07164252e+00 2.47874513e-01 -1.11394569e-01
-6.50656879e-01 9.75978076e-01 2.06257522e-01 -4.51179504e-01
1.01733792e+00 1.84290320e-01 -8.26063231e-02 -2.60184884e-01
-8.76186430e-01 -1.60763040e-01 4.59194273e-01 6.39034569e-01
6.67221248e-01 2.68949270e-01 -8.80523384e-01 9.17340159e-01
-2.87145317e-01 6.14165485e-01 5.00240386e-01 1.26050377e+00
-5.94812989e-01 -7.51810789e-01 -5.35029531e-01 7.98308074e-01
-8.15904975e-01 4.18537632e-02 -2.84614980e-01 5.79578340e-01
4.33173962e-02 1.43219614e+00 -5.38638905e-02 -4.87747610e-01
5.41562319e-01 8.78150463e-01 8.85465026e-01 -4.06252205e-01
-6.56302571e-01 -1.47524709e-02 2.84974761e-02 -7.19039738e-01
-8.80493999e-01 -8.22710633e-01 -9.94981587e-01 -5.26057109e-02
-2.28593677e-01 2.28077322e-01 8.54121327e-01 1.21018791e+00
3.69705200e-01 8.71724427e-01 4.15272623e-01 -1.19269359e+00
-7.28099421e-02 -1.36269176e+00 -8.87103736e-01 3.79995227e-01
7.91186690e-02 -1.08556473e+00 -1.41567245e-01 1.48663789e-01] | [13.24582290649414, 3.384467601776123] |
ebe5a5e2-6aa8-4a50-996a-ac3f13c78acf | docextractor-an-off-the-shelf-historical | 2012.08191 | null | https://arxiv.org/abs/2012.08191v1 | https://arxiv.org/pdf/2012.08191v1.pdf | docExtractor: An off-the-shelf historical document element extraction | We present docExtractor, a generic approach for extracting visual elements such as text lines or illustrations from historical documents without requiring any real data annotation. We demonstrate it provides high-quality performances as an off-the-shelf system across a wide variety of datasets and leads to results on par with state-of-the-art when fine-tuned. We argue that the performance obtained without fine-tuning on a specific dataset is critical for applications, in particular in digital humanities, and that the line-level page segmentation we address is the most relevant for a general purpose element extraction engine. We rely on a fast generator of rich synthetic documents and design a fully convolutional network, which we show to generalize better than a detection-based approach. Furthermore, we introduce a new public dataset dubbed IlluHisDoc dedicated to the fine evaluation of illustration segmentation in historical documents. | ['Mathieu Aubry', 'Tom Monnier'] | 2020-12-15 | null | null | null | null | ['document-layout-analysis'] | ['computer-vision'] | [ 2.39839509e-01 1.15841195e-01 1.66511923e-01 -6.73377514e-02
-1.06949282e+00 -1.06762075e+00 1.01058376e+00 2.31786489e-01
-4.04218763e-01 5.06239533e-01 5.71786538e-02 -4.90958244e-01
-8.45864043e-02 -6.43425703e-01 -1.02978694e+00 -1.68227807e-01
-2.29188800e-03 6.06295228e-01 4.64678168e-01 -3.03364068e-01
3.83063912e-01 8.48026276e-01 -1.47827494e+00 4.94637877e-01
7.59764254e-01 8.42835665e-01 -7.54024908e-02 8.68569613e-01
-2.14774460e-01 2.85721511e-01 -8.78348708e-01 -8.57907057e-01
3.76230270e-01 -1.43836558e-01 -9.26108778e-01 3.22448224e-01
8.01089346e-01 -3.20174903e-01 2.20671035e-02 5.85143387e-01
5.26223600e-01 -2.74742275e-01 8.80271018e-01 -9.83672380e-01
-6.18353844e-01 6.91766322e-01 -6.60530388e-01 -1.57247595e-02
3.28208148e-01 2.03519955e-01 1.03834605e+00 -8.74070406e-01
1.24121714e+00 9.29909766e-01 8.65196407e-01 2.47813463e-01
-1.26107287e+00 -1.89809367e-01 9.47533920e-02 -1.77705348e-01
-1.19739091e+00 -4.88916755e-01 6.03669703e-01 -7.30258226e-01
6.80182636e-01 3.61219972e-01 4.59224343e-01 1.23765063e+00
-1.64523885e-01 1.16300726e+00 1.03204441e+00 -7.08748877e-01
7.63964206e-02 2.05602512e-01 2.41197586e-01 8.14376295e-01
2.83476353e-01 -4.18862253e-01 -1.20725535e-01 2.54919343e-02
6.79990172e-01 -6.08684480e-01 -2.80662179e-01 -5.38547158e-01
-1.24747550e+00 5.22714853e-01 1.05738997e-01 4.61197108e-01
-1.39071137e-01 1.99322090e-01 5.72854638e-01 7.32517242e-03
3.06561589e-01 7.13849783e-01 -4.58313376e-01 -2.63008296e-01
-1.59890318e+00 4.56154078e-01 9.40105855e-01 9.53195751e-01
3.22199792e-01 -2.30824336e-01 -3.48789930e-01 9.24480796e-01
-7.85637572e-02 1.95070788e-01 9.24922973e-02 -8.74478400e-01
5.15267670e-01 6.58379555e-01 1.51706085e-01 -8.97271395e-01
-5.19300520e-01 -6.60522044e-01 -4.50747132e-01 4.58809167e-01
6.66529298e-01 -6.43205941e-02 -1.16777945e+00 1.28984499e+00
9.43726078e-02 -3.88205886e-01 -2.31582940e-01 6.86335385e-01
8.78758371e-01 5.38715124e-01 -1.77779898e-01 1.47169814e-01
1.53184891e+00 -1.01581943e+00 -4.60661292e-01 -1.39848813e-01
5.04881144e-01 -1.05905533e+00 1.56085968e+00 8.74798238e-01
-1.25053668e+00 -4.91940588e-01 -1.31234539e+00 -4.05999869e-01
-7.16244459e-01 4.82826591e-01 6.23417139e-01 7.73056090e-01
-8.25163245e-01 5.48545480e-01 -5.09997189e-01 -5.61593473e-01
6.18521273e-01 1.90704148e-02 -3.68639380e-01 2.29852811e-01
-6.05818570e-01 6.64934397e-01 1.94956198e-01 -2.21200272e-01
-5.71903586e-01 -8.21531653e-01 -5.84072292e-01 1.40132293e-01
5.61397374e-01 -5.46467185e-01 1.37399900e+00 -7.30112970e-01
-1.18025446e+00 1.25028813e+00 3.66034895e-01 -4.43813831e-01
1.29163587e+00 -5.74101448e-01 -4.81333464e-01 4.34311181e-01
7.02107623e-02 7.29216337e-01 8.35136414e-01 -1.62401783e+00
-4.83630091e-01 2.01497991e-02 2.31005661e-02 -3.44678283e-01
-4.25578117e-01 3.53392288e-02 -1.10810363e+00 -9.46278811e-01
-4.78795052e-01 -8.90820444e-01 -4.28886432e-03 7.35979602e-02
-9.64575827e-01 1.12829357e-01 8.73168588e-01 -8.56681585e-01
1.28501427e+00 -1.97191048e+00 -2.10172936e-01 3.88490945e-01
3.80546488e-02 3.36352736e-01 -2.16760486e-01 5.41449070e-01
-4.65444662e-02 4.15571719e-01 -3.75398636e-01 -3.89372706e-01
3.09593618e-01 -3.68387908e-01 -3.89875978e-01 3.06344241e-01
2.91643262e-01 9.28000808e-01 -6.05364859e-01 -8.85001242e-01
1.98141232e-01 4.50860888e-01 -3.65379542e-01 -1.10897593e-01
-5.69671094e-01 -7.22798184e-02 -2.96251774e-01 6.87071502e-01
6.84544742e-01 -4.49023902e-01 5.52882701e-02 -3.06690603e-01
-9.96725857e-02 -7.50021413e-02 -1.28501940e+00 1.91657269e+00
-3.22375506e-01 1.12421334e+00 -7.71152554e-03 -4.25868422e-01
7.28303909e-01 1.99982096e-02 2.47183695e-01 -5.81870794e-01
6.93452656e-02 2.31676400e-01 -2.89249063e-01 -3.62397552e-01
9.74716127e-01 5.26419938e-01 -3.70877713e-01 5.16088307e-01
2.86998153e-02 -1.72126651e-01 8.20954084e-01 5.95663130e-01
1.16359544e+00 5.72846770e-01 2.11663563e-02 -4.21961188e-01
3.89485717e-01 3.12243134e-01 5.68812527e-03 8.14567745e-01
2.29075149e-01 1.13039315e+00 8.27769458e-01 -3.13934445e-01
-1.36434746e+00 -8.32212627e-01 -2.67079949e-01 8.49506736e-01
-1.97191343e-01 -6.30090654e-01 -1.24030054e+00 -7.73526967e-01
-1.19941503e-01 6.75362051e-01 -6.77798450e-01 6.29758596e-01
-6.52650356e-01 -7.10439146e-01 8.28534186e-01 4.35090244e-01
4.94197577e-01 -9.75786269e-01 -6.96965754e-01 5.39977662e-02
3.11643630e-01 -1.38495600e+00 -3.33933085e-01 9.72483158e-02
-2.79333949e-01 -1.21407723e+00 -1.03499305e+00 -5.75913131e-01
5.53732812e-01 -1.96861312e-01 1.51565826e+00 2.10416406e-01
-5.88736594e-01 4.49931324e-01 -2.62273669e-01 -3.58786911e-01
-5.40819585e-01 5.63299537e-01 -7.78555810e-01 -2.73238450e-01
1.20639823e-01 -1.69043407e-01 -6.64461076e-01 4.81662005e-02
-1.18056607e+00 1.17340110e-01 5.46051741e-01 4.64757413e-01
3.00769955e-01 -6.14997484e-02 1.99830592e-01 -1.42400551e+00
9.03683066e-01 -5.64268716e-02 -7.35019863e-01 4.39131737e-01
-4.27804768e-01 4.72145192e-02 5.56674898e-01 -9.07071903e-02
-9.91235733e-01 1.06330037e-01 -2.69995332e-01 1.11838609e-01
-3.27618867e-01 2.76358098e-01 -2.68096745e-01 2.63284415e-01
8.41220140e-01 -2.10345253e-01 -4.50055212e-01 -7.62577236e-01
7.72105992e-01 6.80697620e-01 9.23123300e-01 -6.92925692e-01
8.69946420e-01 4.81924653e-01 6.87641464e-03 -7.83122599e-01
-6.51836991e-01 -1.85310930e-01 -7.52880454e-01 -1.71922818e-01
8.44670057e-01 -5.51142931e-01 -4.96119678e-01 4.10541087e-01
-1.12209845e+00 -4.87910062e-01 -2.95157701e-01 -2.09531397e-01
-5.48781753e-01 4.44736123e-01 -7.18214095e-01 -5.26882231e-01
-3.91116351e-01 -1.03264475e+00 1.33108175e+00 1.78363398e-01
-3.67628872e-01 -7.21720278e-01 -9.59815308e-02 3.07901859e-01
-4.10440229e-02 8.41968119e-01 9.54255879e-01 -6.72815502e-01
-7.51265109e-01 -4.02389705e-01 -4.92425531e-01 1.41727597e-01
-3.52072567e-01 8.91983211e-01 -1.07410634e+00 5.98545112e-02
-8.47387135e-01 -1.98017642e-01 9.65194345e-01 1.63421437e-01
1.33177078e+00 -7.28868917e-02 -4.95179087e-01 5.46283662e-01
1.53260505e+00 -2.20769450e-01 8.55656207e-01 8.21276665e-01
8.12600195e-01 6.03474855e-01 4.33619887e-01 4.68974471e-01
2.54700780e-01 5.54057002e-01 2.13829383e-01 -5.10858119e-01
-3.09324563e-01 -1.65540352e-01 -8.89425725e-02 3.63293558e-01
1.25681698e-01 -7.54953921e-01 -1.17659473e+00 9.04944003e-01
-1.86023295e+00 -7.80410409e-01 -4.29982871e-01 1.83261144e+00
8.77238870e-01 6.17771924e-01 4.18319434e-01 4.04489517e-01
4.74503964e-01 1.12855084e-01 -1.35355487e-01 -5.96757710e-01
-3.00842702e-01 3.80793780e-01 6.10756993e-01 1.40669838e-01
-1.19820583e+00 9.62875724e-01 7.06310749e+00 1.11149085e+00
-9.26432371e-01 -7.14728162e-02 7.88770497e-01 5.07809371e-02
-3.04489434e-01 -1.69277668e-01 -8.31721485e-01 4.96934533e-01
9.33955789e-01 3.85828793e-01 1.01702278e-02 8.90093565e-01
1.18030794e-01 -6.34387583e-02 -1.14297760e+00 6.85557127e-01
6.99947625e-02 -1.82088625e+00 -4.83464450e-02 1.36621460e-01
8.20382297e-01 -2.18416497e-01 -8.65346119e-02 1.76855195e-02
3.56420308e-01 -1.04918134e+00 1.05234146e+00 5.87379038e-01
7.74953485e-01 -7.22502828e-01 3.90950710e-01 -8.86353664e-03
-8.41753542e-01 3.43333960e-01 1.55126229e-01 6.21505558e-01
2.71355838e-01 7.81906128e-01 -7.84339011e-01 4.32399958e-01
6.46236062e-01 2.96573728e-01 -1.27974844e+00 1.28200555e+00
-2.89597511e-01 5.18767595e-01 -3.05367678e-01 1.18143320e-01
3.54583204e-01 1.12684228e-01 3.31879348e-01 2.06074882e+00
2.36725554e-01 -5.48672497e-01 -1.54360920e-01 7.04725683e-01
-4.16727036e-01 2.78722584e-01 -3.36375594e-01 -3.45463187e-01
2.26059824e-01 1.64226949e+00 -1.38803947e+00 -5.22301495e-01
-2.40289599e-01 1.04371023e+00 2.72746384e-01 2.04263166e-01
-7.33911335e-01 -8.30560267e-01 8.88361782e-02 3.72635067e-01
6.80135489e-01 -2.76043057e-01 -4.84242082e-01 -1.00434422e+00
1.23526379e-01 -1.05759621e+00 2.78261691e-01 -1.12476540e+00
-1.07374072e+00 7.08337009e-01 -9.87806991e-02 -1.21668625e+00
-4.79692459e-01 -9.54027355e-01 -4.70117837e-01 4.78709102e-01
-1.48531926e+00 -1.50242126e+00 -4.27443296e-01 2.80912578e-01
5.12411714e-01 8.23504031e-02 5.99254191e-01 3.06739897e-01
-4.99661624e-01 6.02127135e-01 2.31091276e-01 4.64588940e-01
8.38756382e-01 -1.61480105e+00 8.27773869e-01 1.05880404e+00
3.77886325e-01 6.26918197e-01 8.29906702e-01 -5.32133639e-01
-1.17765164e+00 -8.37840676e-01 4.75529224e-01 -6.55576289e-01
5.91199815e-01 -7.51248300e-01 -8.18428576e-01 4.65036899e-01
5.74217319e-01 -1.84534788e-01 5.78301609e-01 -2.71315258e-02
-3.00537765e-01 2.71486640e-01 -9.93494928e-01 8.56654108e-01
9.94755149e-01 -3.85095596e-01 -5.67074835e-01 5.00002027e-01
5.21533966e-01 -4.68507588e-01 -9.68180656e-01 1.41171724e-01
7.50479460e-01 -1.19682646e+00 9.30831909e-01 -3.34550381e-01
8.88513923e-01 -3.51478636e-01 1.98107660e-01 -9.89153445e-01
-1.60744209e-02 -8.31927657e-01 -2.05577552e-01 1.66150928e+00
7.10327446e-01 -2.22733039e-02 8.57806146e-01 3.87677312e-01
-1.30270153e-01 -5.49780786e-01 -1.73843578e-01 -7.26428211e-01
8.88479650e-02 -6.03332758e-01 7.58708596e-01 8.10500205e-01
-3.49483579e-01 3.64008933e-01 -1.36807874e-01 -1.63893342e-01
3.28304797e-01 2.79702365e-01 1.21302176e+00 -1.23163640e+00
-3.26296002e-01 -7.62030184e-01 -1.46909371e-01 -8.36452127e-01
-1.34722948e-01 -5.15938163e-01 -8.42321813e-02 -1.95709598e+00
-3.88149656e-02 -2.40519479e-01 1.37448549e-01 2.49895707e-01
-5.00417501e-02 7.12988615e-01 3.23818088e-01 2.20034510e-01
-6.88659430e-01 -4.09216434e-02 1.14729989e+00 -7.64052942e-02
-3.37966606e-02 -3.14962298e-01 -7.89492190e-01 8.42957258e-01
4.96381253e-01 -1.05875470e-01 -1.40669942e-01 -2.86450595e-01
5.29116929e-01 -5.10287583e-01 3.03839415e-01 -1.04756308e+00
-8.42507854e-02 1.82921588e-01 8.16544473e-01 -9.28064704e-01
2.13819966e-01 -6.46695137e-01 -9.10584480e-02 4.40239459e-02
-2.78019041e-01 -1.83517020e-02 4.80203599e-01 2.93313891e-01
1.57984868e-01 -5.57907283e-01 5.16562581e-01 -2.94022769e-01
-7.87243426e-01 -1.57947496e-01 -2.59873092e-01 1.80884346e-01
8.08120370e-01 -4.08768296e-01 -7.92168438e-01 -2.64177531e-01
-4.76144105e-01 -2.22593918e-02 9.99788821e-01 4.73073602e-01
4.03277166e-02 -9.41594124e-01 -5.41983783e-01 -3.85401510e-02
2.79851139e-01 -2.58992881e-01 -2.00313568e-01 2.65055001e-01
-1.13638604e+00 3.66478562e-01 -2.78470039e-01 -4.79019552e-01
-1.15275848e+00 5.45059621e-01 -6.48534223e-02 -5.48543215e-01
-7.59375989e-01 5.24025202e-01 -1.05890863e-01 -1.92733958e-01
1.45330966e-01 -4.19922650e-01 -1.01512782e-01 3.73307824e-01
4.64159608e-01 3.68499786e-01 4.85273570e-01 -4.06444997e-01
-3.04824859e-01 5.39384127e-01 -5.83746061e-02 -4.07810420e-01
1.50575101e+00 7.00468943e-02 3.34638476e-01 1.67578384e-01
1.10173178e+00 5.12512743e-01 -1.41634834e+00 9.08701345e-02
3.89916152e-01 -4.63352561e-01 3.86238322e-02 -1.11758852e+00
-9.26841736e-01 8.03149462e-01 4.57958311e-01 6.68247163e-01
9.55221415e-01 -1.19640701e-01 6.76066458e-01 3.47531736e-01
-4.75334413e-02 -1.53698552e+00 2.46425923e-02 1.03619015e-02
1.03047073e+00 -1.07887769e+00 5.10558546e-01 -4.98309225e-01
-5.17029285e-01 1.26641428e+00 2.68161803e-01 -1.93422973e-01
3.03937525e-01 6.54450774e-01 8.05432256e-03 -3.35174441e-01
-3.89831543e-01 -3.27151507e-01 6.12403870e-01 6.29305065e-01
5.59637070e-01 -2.70259768e-01 -2.83258468e-01 3.61746490e-01
-5.18946767e-01 5.08351764e-03 7.22726405e-01 1.09711564e+00
-1.86655432e-01 -1.21333516e+00 -4.13765937e-01 5.26740730e-01
-7.82527745e-01 -1.49456844e-01 -7.23668814e-01 1.38396001e+00
1.10212587e-01 4.74968344e-01 1.63614124e-01 2.07285389e-01
4.52463508e-01 1.46699809e-02 5.60394824e-01 -4.67415065e-01
-9.14804101e-01 3.97518665e-01 5.70175707e-01 -3.77282411e-01
-3.42522740e-01 -6.59659803e-01 -1.10534060e+00 -2.17713833e-01
-3.73838767e-02 -1.94590896e-01 9.93169427e-01 7.70853937e-01
5.15136838e-01 7.73779094e-01 -1.64311454e-02 -8.08464110e-01
8.03793967e-02 -8.43041480e-01 -2.43828386e-01 5.14803648e-01
2.76570529e-01 -2.84988970e-01 -8.02409723e-02 5.63459992e-01] | [11.696634292602539, 2.649200439453125] |
3ffb02d0-4f1b-4793-aad6-5ef16c8b9f00 | particlesfm-exploiting-dense-point | 2207.09137 | null | https://arxiv.org/abs/2207.09137v1 | https://arxiv.org/pdf/2207.09137v1.pdf | ParticleSfM: Exploiting Dense Point Trajectories for Localizing Moving Cameras in the Wild | Estimating the pose of a moving camera from monocular video is a challenging problem, especially due to the presence of moving objects in dynamic environments, where the performance of existing camera pose estimation methods are susceptible to pixels that are not geometrically consistent. To tackle this challenge, we present a robust dense indirect structure-from-motion method for videos that is based on dense correspondence initialized from pairwise optical flow. Our key idea is to optimize long-range video correspondence as dense point trajectories and use it to learn robust estimation of motion segmentation. A novel neural network architecture is proposed for processing irregular point trajectory data. Camera poses are then estimated and optimized with global bundle adjustment over the portion of long-range point trajectories that are classified as static. Experiments on MPI Sintel dataset show that our system produces significantly more accurate camera trajectories compared to existing state-of-the-art methods. In addition, our method is able to retain reasonable accuracy of camera poses on fully static scenes, which consistently outperforms strong state-of-the-art dense correspondence based methods with end-to-end deep learning, demonstrating the potential of dense indirect methods based on optical flow and point trajectories. As the point trajectory representation is general, we further present results and comparisons on in-the-wild monocular videos with complex motion of dynamic objects. Code is available at https://github.com/bytedance/particle-sfm. | ['Yong-Jin Liu', 'Wenping Wang', 'Hengkai Guo', 'Shaohui Liu', 'Wang Zhao'] | 2022-07-19 | null | null | null | null | ['motion-segmentation'] | ['computer-vision'] | [-2.59185523e-01 -5.25896549e-01 -1.73058156e-02 -1.79906607e-01
-7.24879980e-01 -8.13989520e-01 4.01422650e-01 -3.90911758e-01
-5.77057421e-01 4.35833037e-01 1.95570782e-01 1.36020109e-01
1.57908916e-01 -4.08973783e-01 -1.10543847e+00 -6.22835815e-01
-1.21205606e-01 5.67389548e-01 4.56239820e-01 6.85482472e-02
1.12628549e-01 6.51554883e-01 -1.30525613e+00 1.74937230e-02
4.34438467e-01 7.45419860e-01 3.79975498e-01 1.03962767e+00
2.64574558e-01 6.80451751e-01 -1.41139269e-01 -2.01552406e-01
6.51233554e-01 -3.46944332e-02 -8.09418499e-01 2.65645981e-01
1.19449854e+00 -9.05180037e-01 -6.45208716e-01 9.34199154e-01
2.86301076e-01 3.15200746e-01 1.72199667e-01 -1.09332228e+00
-2.63087824e-02 7.12214038e-02 -4.46965247e-01 2.99309433e-01
4.71392542e-01 4.90699261e-01 7.93446183e-01 -1.05553544e+00
9.74408090e-01 1.20345116e+00 9.14044499e-01 3.04796576e-01
-1.17932177e+00 -4.20933813e-01 3.84639174e-01 1.96169823e-01
-1.44700003e+00 -5.07310748e-01 6.41209781e-01 -7.01611936e-01
9.94982481e-01 -1.08875278e-02 7.68175840e-01 8.72913778e-01
1.90428853e-01 7.44477212e-01 4.66757089e-01 2.91191563e-02
-9.51816589e-02 -4.20719832e-01 -1.08417846e-01 9.03736830e-01
1.03233434e-01 3.47902477e-01 -3.74142468e-01 -2.75118649e-02
9.15210187e-01 2.51778573e-01 -6.74819708e-01 -9.48327124e-01
-1.62730348e+00 6.58630908e-01 5.75163424e-01 2.59613413e-02
-3.35085750e-01 5.96517205e-01 6.16485663e-02 -1.47446319e-01
4.74545121e-01 2.80642480e-01 -5.75934589e-01 -4.60222542e-01
-1.23651063e+00 4.47316498e-01 6.51542783e-01 1.08304107e+00
9.15304065e-01 7.44295344e-02 1.42873377e-01 3.80178034e-01
4.02359664e-01 7.31661320e-01 5.03995456e-02 -1.64858484e+00
4.63538021e-01 2.78807908e-01 2.65763670e-01 -1.15471017e+00
-4.62444007e-01 -4.38662320e-01 -5.47714949e-01 2.25647569e-01
5.61576426e-01 -1.44364178e-01 -7.62001574e-01 1.45473683e+00
5.54718018e-01 7.74773240e-01 -1.72552630e-01 1.33246601e+00
6.78232729e-01 8.41722488e-01 -3.54332536e-01 -4.28797677e-02
7.81270564e-01 -1.18170488e+00 -4.11366642e-01 -2.22666636e-01
4.85567123e-01 -8.98081362e-01 5.70288301e-01 2.18126863e-01
-1.24360657e+00 -6.15229130e-01 -6.48405075e-01 -2.72887647e-01
1.17489904e-01 1.43475905e-01 2.93283820e-01 1.98841944e-01
-1.11740398e+00 7.14136183e-01 -1.35994577e+00 -2.75189161e-01
4.24335361e-01 4.47669536e-01 -4.96139258e-01 -2.04622343e-01
-6.22880816e-01 5.26105106e-01 2.81622529e-01 3.12578231e-01
-9.51388180e-01 -1.01162171e+00 -1.11389804e+00 -1.40273541e-01
4.09346402e-01 -1.09120369e+00 1.20325863e+00 -7.72344410e-01
-1.42176902e+00 5.06288350e-01 -3.91420871e-01 -4.80697036e-01
8.69103849e-01 -7.74503231e-01 3.38488400e-01 5.85790575e-01
1.15543872e-01 1.11180329e+00 8.42406154e-01 -1.24814880e+00
-6.92998886e-01 -1.05826251e-01 6.16332889e-03 3.74569267e-01
9.20277536e-02 -2.09150180e-01 -8.87806177e-01 -2.93035090e-01
1.23783410e-01 -1.43103087e+00 -3.21687490e-01 2.92279094e-01
-2.29457065e-01 2.89711475e-01 1.15075982e+00 -7.19996929e-01
8.29623044e-01 -1.89940143e+00 5.07419109e-01 -1.87528744e-01
2.93985039e-01 2.85647750e-01 -9.49399322e-02 1.67440325e-01
1.17555186e-01 -2.67634809e-01 -3.05254906e-01 -6.35890722e-01
-2.89393544e-01 9.15482789e-02 -1.70258939e-01 9.07500446e-01
1.13372251e-01 9.23721910e-01 -1.01396120e+00 -2.69853204e-01
7.86336362e-01 6.77731156e-01 -6.94745183e-01 3.81201535e-01
-1.80719391e-01 7.58250833e-01 -1.30668178e-01 6.86051071e-01
7.65643299e-01 -3.13769251e-01 -1.65859610e-01 -6.03690028e-01
-3.54553878e-01 -4.04916964e-02 -1.30156147e+00 2.07993460e+00
-2.07980648e-01 8.60915244e-01 3.45551372e-01 -5.05659759e-01
2.75964290e-01 -1.42791308e-03 8.85894716e-01 -8.08776263e-03
1.25596344e-01 5.69492280e-02 -2.63033003e-01 -3.93584520e-01
6.82523489e-01 2.92024642e-01 3.56446922e-01 -1.45408008e-02
-1.56064366e-03 -4.32546467e-01 2.97266066e-01 3.34546179e-01
1.12352979e+00 7.15467393e-01 -6.33909553e-02 -3.01441867e-02
5.47024250e-01 2.50540912e-01 7.90805936e-01 4.06239182e-01
-2.04743043e-01 1.18557179e+00 1.03251301e-01 -5.87875724e-01
-1.21854806e+00 -1.10528660e+00 -8.10496658e-02 2.63579726e-01
5.09756625e-01 -4.17737633e-01 -6.77249134e-01 -4.70200807e-01
1.06338318e-02 2.27453664e-01 -2.08367065e-01 3.45734537e-01
-8.39333594e-01 -4.45310831e-01 7.38376006e-02 4.97307450e-01
5.88586509e-01 -6.40112162e-01 -7.03679860e-01 4.02984172e-01
-3.92399549e-01 -1.88456178e+00 -7.67245233e-01 -4.96952683e-01
-9.91195440e-01 -1.27105772e+00 -8.31473589e-01 -4.70587373e-01
5.47802508e-01 6.08223796e-01 1.18120551e+00 5.21242507e-02
-1.13549255e-01 7.67222703e-01 -1.19207978e-01 2.28445113e-01
-2.53807455e-01 1.35782640e-02 1.54789999e-01 1.52415344e-02
1.01029754e-01 -3.88681769e-01 -1.03860211e+00 5.31100273e-01
-6.78296268e-01 2.18652993e-01 2.40820065e-01 5.73414862e-01
6.67924821e-01 -3.14320385e-01 -3.12300622e-01 -4.97905970e-01
-2.58211613e-01 -4.03356612e-01 -1.01771879e+00 -3.34202141e-01
-1.27983153e-01 -2.52623647e-01 4.07689065e-01 -2.34757587e-01
-7.88748384e-01 7.23863661e-01 -8.35300386e-02 -1.09044409e+00
-2.69137293e-01 1.29602700e-01 8.49330202e-02 -3.97497326e-01
3.24719787e-01 5.52815832e-02 2.52663702e-01 -2.52454013e-01
4.14918959e-01 1.71063140e-01 7.05760419e-01 -3.21968794e-01
1.02825773e+00 1.05983675e+00 6.92813918e-02 -1.06755781e+00
-6.29390895e-01 -9.24859107e-01 -1.12637210e+00 -5.11937439e-01
1.22061121e+00 -1.31795895e+00 -7.18440056e-01 7.37837374e-01
-1.44853520e+00 -5.90345085e-01 1.14974201e-01 8.92212987e-01
-7.09809721e-01 7.34931409e-01 -8.96871030e-01 -3.19876522e-01
-1.48526579e-01 -1.58561563e+00 1.45562375e+00 1.48694022e-02
-8.52036476e-02 -1.16848171e+00 2.54305869e-01 4.36937809e-01
-1.65843517e-02 4.97402996e-01 5.82291037e-02 1.58874243e-01
-1.48750412e+00 -5.44346981e-02 -7.89605677e-02 3.84055316e-01
-8.69516581e-02 3.01819533e-01 -6.77683055e-01 -5.12675405e-01
-1.77551180e-01 2.81604230e-02 8.37864816e-01 7.86369681e-01
7.89048910e-01 -2.94109341e-02 -3.87499392e-01 1.31047583e+00
1.53131223e+00 -2.67269194e-01 5.72888076e-01 3.13655704e-01
1.50338590e+00 3.64932507e-01 6.42873824e-01 2.22850814e-01
5.69669843e-01 9.77722168e-01 6.55696750e-01 -9.21947435e-02
-1.61116436e-01 -9.82667506e-02 5.76926351e-01 7.53974915e-01
-1.36399835e-01 -7.16678053e-02 -9.21064436e-01 6.43294573e-01
-2.05976725e+00 -1.07484341e+00 -5.23403645e-01 2.10792851e+00
3.73459280e-01 -1.39015362e-01 1.58514857e-01 -3.88541937e-01
6.11104369e-01 2.20512331e-01 -4.58543539e-01 1.83539465e-01
-9.64182988e-02 -2.95110226e-01 7.58149147e-01 8.30672204e-01
-1.29382432e+00 1.11210620e+00 5.26572371e+00 2.71095693e-01
-1.14565814e+00 1.67265758e-01 3.86959046e-01 -4.14625466e-01
9.91625637e-02 1.11826599e-01 -1.10893083e+00 4.66588646e-01
6.88129365e-01 3.32479060e-01 3.15813661e-01 5.79237342e-01
5.78974247e-01 -2.30662048e-01 -1.19184101e+00 1.23912382e+00
5.61979003e-02 -1.73363471e+00 -1.71801358e-01 2.30609670e-01
1.11431217e+00 6.20608032e-01 -2.21848786e-01 -2.34398589e-01
1.18840538e-01 -6.54254019e-01 8.92997384e-01 5.57614803e-01
4.11194026e-01 -5.55753529e-01 6.41684592e-01 3.04312348e-01
-1.32734394e+00 7.83489943e-02 -4.42032814e-01 -7.71802291e-02
7.02545464e-01 5.84354758e-01 -6.03628933e-01 4.53520775e-01
9.03471291e-01 1.38908076e+00 -4.11427855e-01 1.30144894e+00
3.86387929e-02 3.42375308e-01 -6.38350606e-01 4.30365086e-01
5.52248240e-01 -4.02408332e-01 8.79785597e-01 1.18576133e+00
4.85995322e-01 1.30399495e-01 4.14006680e-01 9.15069997e-01
-5.89991584e-02 -2.78462648e-01 -7.36414135e-01 3.05870086e-01
2.76779652e-01 1.43602562e+00 -6.80788517e-01 -2.40438327e-01
-5.96871853e-01 1.07103956e+00 2.76897758e-01 4.06476527e-01
-7.99678445e-01 2.86320359e-01 9.17457998e-01 3.16938758e-01
5.90367436e-01 -9.20559227e-01 1.58403233e-01 -1.46578884e+00
1.29933655e-01 -6.06686532e-01 8.72410759e-02 -9.41581428e-01
-9.52014685e-01 4.48565394e-01 1.50335908e-01 -1.53659475e+00
-3.48480970e-01 -5.99352598e-01 -5.05108833e-01 5.18621027e-01
-1.41855729e+00 -1.11121702e+00 -6.67439163e-01 6.22106135e-01
8.11580598e-01 7.86485374e-02 2.26204857e-01 2.06672058e-01
-3.13120455e-01 6.52371123e-02 2.03074113e-01 3.09167385e-01
5.83846867e-01 -1.07028353e+00 6.50026381e-01 1.27170134e+00
3.57843190e-01 4.56423789e-01 5.04914522e-01 -7.21452832e-01
-1.64938939e+00 -1.33424723e+00 3.34234059e-01 -9.60821211e-01
5.71217954e-01 -2.99474686e-01 -8.33334804e-01 7.83934951e-01
1.50934950e-01 4.13756102e-01 9.49353874e-02 -4.53519046e-01
2.01940492e-01 -5.82865812e-02 -6.87453032e-01 5.49439311e-01
1.15569735e+00 -2.90461719e-01 -1.66964933e-01 4.85576421e-01
6.87635541e-01 -9.87481236e-01 -7.91847348e-01 3.35972846e-01
3.78726840e-01 -1.10919559e+00 1.26164055e+00 -6.38572574e-02
4.33090478e-01 -7.05248713e-01 -1.97350718e-02 -1.20044327e+00
-1.96213096e-01 -9.94513452e-01 -3.67936164e-01 7.77496994e-01
3.80002074e-02 -2.40164101e-01 1.01609910e+00 5.03177881e-01
-2.77851880e-01 -5.33512056e-01 -8.59774649e-01 -7.65339136e-01
-7.06931949e-02 -5.08950353e-01 1.48008525e-01 8.25507104e-01
-6.76342845e-01 1.19799912e-01 -3.65733325e-01 5.80353916e-01
9.60679889e-01 3.86450626e-02 1.23649287e+00 -8.53461325e-01
-2.59543866e-01 -1.83974251e-01 -7.42203116e-01 -1.60505819e+00
5.07871389e-01 -5.97727001e-01 1.73355520e-01 -1.46524096e+00
7.21088797e-03 -1.50249630e-01 3.84916395e-01 -6.91598877e-02
-1.73223495e-01 5.12319565e-01 5.50828874e-01 4.54818517e-01
-6.36983991e-01 5.15551448e-01 1.16809583e+00 8.16023559e-04
-2.65637755e-01 -2.28480343e-02 2.29822367e-01 9.95134294e-01
5.06187677e-01 -4.13990170e-01 -2.59285599e-01 -6.44667149e-01
3.82646546e-02 1.34251595e-01 9.10389483e-01 -1.30589712e+00
3.74811500e-01 1.98242031e-02 4.04486984e-01 -9.19228852e-01
6.47166908e-01 -1.01013768e+00 3.32021266e-01 4.40937817e-01
4.75967340e-02 4.47729826e-01 2.26982728e-01 6.75945699e-01
-5.76814897e-02 -2.61946321e-02 7.55234182e-01 -1.97626293e-01
-1.05327713e+00 7.91220844e-01 -2.33467534e-01 1.64243117e-01
9.56591070e-01 -3.81785870e-01 -1.26959071e-01 -4.93393689e-01
-5.73530674e-01 2.49429852e-01 8.85344148e-01 5.26407003e-01
8.18251967e-01 -1.19367683e+00 -7.72599280e-01 8.89354423e-02
-1.78304598e-01 4.32026118e-01 1.88866973e-01 1.04633105e+00
-1.32296610e+00 3.67548048e-01 -4.16628522e-04 -1.34718025e+00
-1.39653313e+00 4.03699964e-01 6.30232453e-01 1.90297365e-01
-9.17074442e-01 7.06834495e-01 3.29055637e-01 -3.71049613e-01
6.22216798e-02 -5.31211317e-01 2.22638577e-01 -3.51382941e-01
3.57412308e-01 4.83711690e-01 -5.58990762e-02 -1.09087873e+00
-3.97995174e-01 1.12933910e+00 3.99719104e-02 -1.72757924e-01
1.28506613e+00 -4.19599086e-01 -2.25924142e-02 2.15264410e-01
1.59175980e+00 -1.50304264e-03 -2.11720300e+00 -1.59111917e-02
-5.04366159e-01 -9.23638403e-01 2.66257852e-01 -2.08163098e-01
-1.27261770e+00 9.24023867e-01 5.79526544e-01 -4.64915693e-01
5.94670892e-01 -1.91531688e-01 1.04194212e+00 3.68815690e-01
3.60540241e-01 -8.54284763e-01 -1.34470675e-03 6.41130209e-01
6.13613963e-01 -1.36082208e+00 2.15507299e-01 -6.32982850e-01
-4.15482104e-01 1.37514091e+00 6.72506988e-01 -3.18637192e-01
5.18921793e-01 1.72589064e-01 2.15183645e-01 -6.71500191e-02
-4.74412680e-01 -1.19998448e-01 3.46511036e-01 5.18651366e-01
1.74201354e-01 -3.23010147e-01 2.11780638e-01 -4.83145565e-01
-5.98773621e-02 -6.77173361e-02 7.96558321e-01 7.93748736e-01
-1.22420944e-01 -7.38936782e-01 -5.17961144e-01 -3.67503287e-03
-4.14632052e-01 -3.73266228e-02 -2.29364336e-02 9.38587904e-01
-8.69927406e-02 7.87693679e-01 3.34228128e-01 -1.22252762e-01
1.64678365e-01 -4.57813799e-01 5.63913465e-01 -4.85230446e-01
-4.44821000e-01 2.50283629e-01 -6.00790558e-03 -1.20987213e+00
-7.17016637e-01 -1.01674080e+00 -1.26647127e+00 -4.52021539e-01
-6.80688471e-02 -1.99141309e-01 6.03462458e-01 8.63831997e-01
5.19823372e-01 2.25885004e-01 3.65460426e-01 -1.67411172e+00
-2.42987603e-01 -4.39188689e-01 -7.27637336e-02 5.96714616e-01
6.75045729e-01 -6.09043181e-01 -4.41208065e-01 4.12713945e-01] | [8.380518913269043, -1.9897937774658203] |
879145be-9a88-456c-a50f-653bc14e41c1 | a-novel-augmented-reality-ultrasound | 2205.04350 | null | https://arxiv.org/abs/2205.04350v1 | https://arxiv.org/pdf/2205.04350v1.pdf | A Novel Augmented Reality Ultrasound Framework Using an RGB-D Camera and a 3D-printed Marker | Purpose. Ability to locate and track ultrasound images in the 3D operating space is of great benefit for multiple clinical applications. This is often accomplished by tracking the probe using a precise but expensive optical or electromagnetic tracking system. Our goal is to develop a simple and low cost augmented reality echography framework using a standard RGB-D Camera. Methods. A prototype system consisting of an Occipital Structure Core RGB-D camera, a specifically-designed 3D marker, and a fast point cloud registration algorithm FaVoR was developed and evaluated on an Ultrasonix ultrasound system. The probe was calibrated on a 3D-printed N-wire phantom using the software PLUS toolkit. The proposed calibration method is simplified, requiring no additional markers or sensors attached to the phantom. Also, a visualization software based on OpenGL was developed for the augmented reality application. Results. The calibrated probe was used to augment a real-world video in a simulated needle insertion scenario. The ultrasound images were rendered on the video, and visually-coherent results were observed. We evaluated the end-to-end accuracy of our AR US framework on localizing a cube of 5 cm size. From our two experiments, the target pose localization error ranges from 5.6 to 5.9 mm and from -3.9 to 4.2 degrees. Conclusion. We believe that with the potential democratization of RGB-D cameras integrated in mobile devices and AR glasses in the future, our prototype solution may facilitate the use of 3D freehand ultrasound in clinical routine. Future work should include a more rigorous and thorough evaluation, by comparing the calibration accuracy with those obtained by commercial tracking solutions in both simulated and real medical scenarios. | ['Laurent Launay', 'Albert Murienne', 'Pierre Le Gargasson', 'Guillaume Pasquier', 'Boris Labbé', 'Gaétan Lelu', 'Yitian Zhou'] | 2022-05-09 | null | null | null | null | ['point-cloud-registration'] | ['computer-vision'] | [ 1.38930529e-02 1.39072835e-01 4.52360958e-01 -1.31486692e-02
-8.99444103e-01 -7.89123237e-01 -1.25444874e-01 -7.85209760e-02
-4.44752961e-01 1.26583263e-01 -1.10994913e-01 -7.01253355e-01
-3.36363055e-02 -3.24193954e-01 -6.42813861e-01 -5.10837197e-01
-5.01291990e-01 3.42904776e-01 3.81856143e-01 -6.82728877e-03
1.56084867e-02 8.14060986e-01 -1.01401091e+00 -3.04933339e-01
7.14872420e-01 1.08273101e+00 7.10379601e-01 8.83269548e-01
2.22444847e-01 3.46419603e-01 -3.33132386e-01 -1.35157779e-01
5.07247090e-01 -2.03611001e-01 -6.62708804e-02 -3.56623270e-02
2.39559248e-01 -6.82751119e-01 -1.76725551e-01 7.49435663e-01
1.05315065e+00 -2.13569596e-01 -7.36235231e-02 -5.06325245e-01
-5.65305687e-02 2.26180255e-01 -6.87320530e-01 -4.73362207e-02
9.70103741e-01 -2.11244971e-02 -2.81927496e-01 -5.97933054e-01
8.25394690e-01 5.77103555e-01 1.05453646e+00 4.87373948e-01
-7.25347638e-01 -7.38955379e-01 -6.68888211e-01 -4.95141387e-01
-1.48000729e+00 -1.06084056e-01 5.76254010e-01 -6.22326136e-01
4.72634882e-01 6.62689328e-01 9.96855795e-01 7.14522243e-01
9.40818548e-01 -1.53013259e-01 1.31114495e+00 -6.32918775e-01
2.35156596e-01 5.86279333e-01 -7.96042681e-02 1.08109784e+00
5.10755539e-01 2.00181380e-01 5.67464940e-02 -2.79603243e-01
1.52213478e+00 2.00731933e-01 -8.70944560e-01 -9.58497584e-01
-1.01492059e+00 2.63501793e-01 4.31673795e-01 6.23476148e-01
-6.22241259e-01 1.35287225e-01 8.14551935e-02 -3.21208894e-01
-2.18519382e-02 4.87286866e-01 -2.58748055e-01 -7.33214259e-01
-4.64669496e-01 -6.10896766e-01 8.96656692e-01 1.08392441e+00
6.41914457e-03 -2.11536616e-01 2.49639928e-01 2.22725675e-01
6.43877745e-01 6.67540848e-01 5.18167436e-01 -8.53294790e-01
9.27293003e-02 3.29181731e-01 4.69171822e-01 -1.09582376e+00
-7.82819927e-01 -5.28617799e-01 -2.68032879e-01 4.45564687e-01
1.42966807e-01 -3.09302837e-01 -9.43401873e-01 1.00445056e+00
6.76826596e-01 5.83337665e-01 -1.13737322e-01 1.30977154e+00
1.07263565e+00 1.38176396e-01 -4.39108580e-01 -3.48840714e-01
1.51346004e+00 -2.71172315e-01 -9.58683133e-01 3.28909785e-01
9.21235204e-01 -1.17058778e+00 1.06997073e+00 4.16710764e-01
-1.29318154e+00 -4.37587082e-01 -1.03614163e+00 6.18279576e-01
3.46838295e-01 1.67783231e-01 5.97777247e-01 1.38180053e+00
-1.04504526e+00 3.65662932e-01 -1.60176849e+00 -3.84078115e-01
-2.27097332e-01 5.65608859e-01 -5.85971117e-01 -1.68047715e-02
-6.31044805e-01 1.11107433e+00 -1.00279758e-02 4.01951760e-01
-3.26383501e-01 -8.24242115e-01 -8.95069718e-01 -3.72170866e-01
1.72443554e-01 -6.61828279e-01 1.26284444e+00 -2.82243729e-01
-2.10716152e+00 8.90518844e-01 2.09829181e-01 1.23503402e-01
4.61311072e-01 -4.78836417e-01 -5.35466373e-01 6.06837571e-01
-1.32805184e-01 -2.99636632e-01 1.75102994e-01 -1.37355983e+00
-7.80134797e-02 -5.79892397e-01 3.38095753e-03 1.56302020e-01
1.92384601e-01 1.08073846e-01 -7.91398466e-01 -2.88407683e-01
8.78584802e-01 -1.32806063e+00 -2.82684296e-01 3.12303543e-01
1.94928810e-01 8.42490256e-01 4.50641841e-01 -7.88720846e-01
1.01431167e+00 -1.96584010e+00 -2.20527902e-01 5.02623320e-01
7.73956180e-02 2.20742181e-01 5.04911125e-01 -4.47558612e-02
-2.33286187e-01 -2.25315064e-01 3.64420652e-01 7.96670094e-03
-4.76898938e-01 -1.11785099e-01 2.59318978e-01 9.55596745e-01
-9.21878457e-01 5.99776924e-01 -9.66412008e-01 -4.16797578e-01
6.16544485e-01 7.91000545e-01 -3.51733029e-01 4.55174327e-01
7.20293403e-01 1.02943027e+00 -4.58999574e-01 9.19286966e-01
8.88410211e-01 -1.03620738e-01 2.30442017e-01 -4.36373949e-01
-3.35673362e-01 -6.31124079e-02 -1.54317725e+00 2.38803029e+00
-8.69515240e-01 1.31903946e-01 4.06421870e-01 -2.35977881e-02
7.66293526e-01 7.05081940e-01 7.14405477e-01 -7.19191134e-01
5.75570285e-01 4.95388657e-01 4.57398631e-02 -1.10678697e+00
2.11753294e-01 -7.80804604e-02 2.79669702e-01 3.79503071e-01
-2.67907590e-01 -3.51847798e-01 -6.04641020e-01 5.04178815e-02
9.92751420e-01 4.79811162e-01 2.27978274e-01 -9.37826857e-02
2.33896330e-01 1.53675810e-01 2.35733818e-02 6.80917323e-01
-5.87894991e-02 8.52505684e-01 -2.14647532e-01 -1.96407974e-01
-4.52844083e-01 -9.69557524e-01 -3.96145940e-01 1.13791071e-01
6.33618712e-01 -1.65803462e-01 -5.17400682e-01 -3.83571595e-01
-2.56646514e-01 4.15890455e-01 -3.50248247e-01 1.86986804e-01
-6.69226170e-01 -4.43870667e-03 1.78407595e-01 3.10404509e-01
1.15614356e-02 -3.57355177e-01 -1.32485127e+00 3.50777477e-01
2.62590826e-01 -1.06553185e+00 -1.76772490e-01 -6.64240569e-02
-1.09075999e+00 -1.00449347e+00 -8.94389808e-01 -5.60402989e-01
7.59674609e-01 3.33931714e-01 6.69290423e-01 -2.55414844e-01
-3.23582619e-01 1.18673265e+00 -6.24116004e-01 -3.63110423e-01
-3.49448800e-01 -5.02640963e-01 1.70204088e-01 -5.06847560e-01
-1.95604831e-01 -2.46862680e-01 -1.04809105e+00 4.34553325e-01
-4.59738672e-01 -1.39469072e-01 4.37890023e-01 2.16380224e-01
3.77636969e-01 -6.32327199e-01 -2.02918246e-01 -6.05399489e-01
3.58137131e-01 -7.55656436e-02 -8.87450993e-01 1.09965459e-01
-1.77647233e-01 -4.84729230e-01 1.19872175e-01 -3.47053438e-01
-9.81285691e-01 2.67722309e-01 -3.78865719e-01 -7.00399756e-01
-3.50601450e-02 5.62657356e-01 3.09512347e-01 -7.63243079e-01
7.81805098e-01 -1.98095724e-01 4.45201576e-01 -1.78597003e-01
7.16717392e-02 9.31954503e-01 6.36250019e-01 -3.23913664e-01
4.45000827e-01 3.70275795e-01 -5.25692180e-02 -5.76850653e-01
5.71709359e-03 -5.10210395e-01 -4.69473928e-01 -5.50861239e-01
6.68496430e-01 -8.97860408e-01 -9.71618533e-01 -2.31585372e-02
-1.13316643e+00 -6.93353415e-02 -4.71718907e-02 1.42983329e+00
-4.03961271e-01 4.63069439e-01 -5.79656363e-01 -7.49340713e-01
-3.49802852e-01 -1.52619362e+00 1.03850472e+00 4.00283813e-01
-1.86600521e-01 -8.93151402e-01 3.59824002e-01 1.40312493e-01
4.91288841e-01 5.74224114e-01 -4.27554771e-02 -9.13203955e-02
-5.35086930e-01 -7.49190211e-01 2.97535598e-01 -2.95349866e-01
4.31858808e-01 -1.27717778e-01 -9.02678370e-01 -3.89548242e-01
5.67619979e-01 1.78495914e-01 -2.35212818e-01 8.34247231e-01
8.77368331e-01 2.56950438e-01 -5.85977495e-01 8.98055673e-01
1.53844833e+00 8.32771361e-01 5.64083040e-01 1.92033738e-01
5.67588210e-01 7.80003145e-02 1.12892938e+00 2.24544704e-01
1.85087565e-02 8.33971381e-01 4.75949585e-01 -3.28362375e-01
2.77467668e-02 1.73173994e-01 -3.84171717e-02 1.12847924e+00
-4.17736292e-01 2.31636465e-01 -1.11095750e+00 -1.59985393e-01
-1.16879869e+00 -2.85600066e-01 -2.35121608e-01 2.58903027e+00
6.12150252e-01 -1.50314316e-01 -4.69669670e-01 -2.81840473e-01
5.47010243e-01 -7.17671275e-01 -2.99542770e-02 -3.50867063e-01
7.08113909e-01 5.78495026e-01 6.45857990e-01 5.78991652e-01
-7.62847483e-01 1.10346444e-01 5.53362942e+00 -1.50317315e-03
-1.80360281e+00 3.82887572e-01 -1.43719926e-01 6.14409372e-02
-1.69502527e-01 -2.57144660e-01 -2.60960370e-01 3.68180662e-01
6.61189258e-01 1.67822987e-01 -2.08988711e-01 8.82057786e-01
3.96744519e-01 -5.56734025e-01 -9.23970640e-01 1.29210150e+00
1.88903436e-01 -1.25583470e+00 -8.05400074e-01 4.77259085e-02
3.48695964e-01 -1.29420087e-01 -1.01423651e-01 -1.03131920e-01
-4.70674813e-01 -6.10790849e-01 4.13326800e-01 6.24329269e-01
1.33029759e+00 -2.44363889e-01 9.28823590e-01 3.46698880e-01
-8.94041121e-01 6.47082686e-01 -2.37742160e-02 2.32371077e-01
4.42270845e-01 1.99114904e-01 -1.23698485e+00 6.38929486e-01
6.19997323e-01 -4.00409997e-02 -2.00499579e-01 1.48537385e+00
2.15549096e-01 1.90867648e-01 -4.27953988e-01 2.61982679e-02
-7.02022051e-04 -2.92945206e-01 4.96259302e-01 1.04822648e+00
8.61451447e-01 6.89159811e-01 -2.43534744e-01 3.33892494e-01
5.37299216e-01 4.35082801e-02 -6.81951404e-01 5.09968579e-01
2.77674466e-01 1.45974207e+00 -1.00167680e+00 3.37249450e-02
-4.02151734e-01 8.32194984e-01 -6.06404006e-01 3.66026647e-02
-1.20128059e+00 -5.53018928e-01 -4.33426946e-02 4.42482144e-01
1.02197647e-01 -4.25314426e-01 -1.96497574e-01 -1.04657459e+00
1.17591731e-01 -4.37551588e-01 -6.59563914e-02 -1.33735192e+00
-4.87469196e-01 8.98911536e-01 -2.21233293e-01 -1.89211512e+00
-3.26331347e-01 -6.48326933e-01 -1.51662841e-01 1.10942614e+00
-1.04918957e+00 -7.90811598e-01 -8.32674861e-01 4.41193163e-01
-1.20143488e-01 2.22478330e-01 1.29760122e+00 3.43989611e-01
-5.23197278e-02 3.90489548e-01 1.81427866e-01 -1.97107986e-01
7.24480629e-01 -1.00456452e+00 -2.80893207e-01 5.49291849e-01
-4.21006173e-01 1.10314965e+00 6.64449930e-01 -7.74450719e-01
-2.04579878e+00 -4.62011874e-01 -2.90566310e-02 -5.56682765e-01
4.38524634e-01 -1.76233143e-01 -7.05678225e-01 8.52497578e-01
1.61614805e-01 5.44673502e-01 8.75115693e-01 -3.73426825e-01
3.40822309e-01 -1.91239864e-02 -1.68429077e+00 2.90836275e-01
7.54418910e-01 -1.93053097e-01 -4.76854831e-01 1.36589900e-01
6.29865706e-01 -1.84041047e+00 -1.23095179e+00 2.79109389e-01
1.11050963e+00 -8.30748498e-01 7.76789606e-01 2.67586380e-01
-1.50122102e-02 -4.91179317e-01 8.31900015e-02 -1.17730927e+00
6.90885633e-03 -8.42507005e-01 9.78627205e-02 5.15189767e-01
4.90834452e-02 -9.46008384e-01 8.39033961e-01 9.36274171e-01
-3.66761088e-01 -6.60250902e-01 -1.07232344e+00 -4.51997310e-01
-6.35118783e-01 -6.10341787e-01 8.30559283e-02 7.62720108e-01
4.06759292e-01 -2.66943842e-01 2.10671291e-01 9.22649741e-01
4.64678377e-01 2.88799051e-02 6.95156991e-01 -7.87283361e-01
-4.67879623e-01 2.96735436e-01 -7.59267390e-01 -8.56182694e-01
-7.36623406e-01 -5.88897407e-01 -2.04955637e-01 -1.40407848e+00
-2.51982808e-01 -9.11300004e-01 1.69946179e-01 -1.81522653e-01
2.13949278e-01 1.53913140e-01 4.94363867e-02 1.36797801e-02
-1.17053278e-01 -8.12413692e-02 1.66105866e+00 6.38264418e-01
-4.60686415e-01 3.76363456e-01 -2.43147820e-01 7.08985865e-01
4.03550833e-01 -5.57195544e-01 -2.61408716e-01 -3.99472505e-01
4.49041836e-02 8.22898269e-01 2.94492245e-01 -1.01585710e+00
4.52821821e-01 4.57240164e-01 4.34462041e-01 -6.10026896e-01
5.38469434e-01 -1.72010851e+00 7.56579399e-01 1.01649225e+00
4.86926734e-01 8.76348838e-02 4.22571242e-01 7.30632320e-02
1.16990514e-01 -2.60204315e-01 5.56668282e-01 -3.27490121e-01
-2.65468836e-01 -1.37957662e-01 -5.52586466e-02 -5.79081118e-01
1.43493807e+00 -8.01342010e-01 3.68267894e-02 -3.33802491e-01
-1.12018561e+00 -3.25430959e-01 5.35242975e-01 3.52408774e-02
9.16051090e-01 -9.94026661e-01 -1.94573686e-01 5.93013942e-01
-1.23006456e-01 -6.13200478e-02 3.93625498e-01 1.30597889e+00
-1.49985313e+00 5.19075155e-01 -7.50292912e-02 -1.26330757e+00
-1.27331603e+00 3.98452580e-01 9.13574934e-01 2.52340436e-01
-5.88492274e-01 5.59128463e-01 -2.58080155e-01 -5.21566212e-01
2.45305777e-01 -7.30203569e-01 3.92747000e-02 -5.94606519e-01
4.26590502e-01 1.97498366e-01 5.56742847e-01 -3.02252710e-01
-6.19087279e-01 1.19564807e+00 3.46296996e-01 -2.19599962e-01
1.06883621e+00 -2.68105537e-01 3.31167400e-01 2.33762547e-01
8.22218776e-01 6.15979195e-01 -8.05022299e-01 1.43675223e-01
-5.71031451e-01 -9.61825788e-01 8.50792155e-02 -9.62559223e-01
-1.07076240e+00 4.87586290e-01 1.29409885e+00 1.63367912e-02
1.16061819e+00 -4.50677145e-03 2.60327786e-01 -2.01060265e-01
1.09194732e+00 -9.79705080e-02 -2.58765399e-01 -6.63131848e-03
8.98592651e-01 -9.17668939e-01 7.78383836e-02 -7.55495310e-01
-5.51640213e-01 1.17580581e+00 3.94073039e-01 1.92319229e-02
7.49052048e-01 9.80457067e-01 5.74171305e-01 -3.71410340e-01
2.57542104e-01 1.96250230e-01 1.23959683e-01 4.53617036e-01
7.96880960e-01 1.84947141e-02 -5.19282281e-01 4.26107258e-01
-3.30365915e-03 5.71404755e-01 7.59131908e-01 1.41389215e+00
-1.14334106e-01 -5.88601649e-01 -8.51949930e-01 7.54759312e-02
-7.56330431e-01 3.74677092e-01 6.28956318e-01 1.00050056e+00
3.56070399e-02 7.36266375e-01 9.13512614e-03 -1.97042629e-01
8.82431328e-01 -6.07020855e-01 9.17593956e-01 -6.78106189e-01
-9.10163403e-01 6.31566286e-01 -3.51883948e-01 -8.77545774e-01
-3.13449115e-01 -4.74851966e-01 -1.43827140e+00 4.67513576e-02
-5.78676879e-01 1.46269411e-01 1.42998672e+00 1.91457465e-01
5.61094999e-01 5.77082276e-01 5.63406169e-01 -9.10523534e-01
-3.54644090e-01 -7.82074690e-01 -7.97163844e-01 -1.55700341e-01
2.14164823e-01 -6.09785795e-01 -1.62689343e-01 -2.86259234e-01] | [13.783023834228516, -2.9872264862060547] |
86d5ee29-264b-4a6c-b903-402f3dc715df | data-driven-and-automatic-surface-texture | 2110.10005 | null | https://arxiv.org/abs/2110.10005v1 | https://arxiv.org/pdf/2110.10005v1.pdf | Data-driven and Automatic Surface Texture Analysis Using Persistent Homology | Surface roughness plays an important role in analyzing engineering surfaces. It quantifies the surface topography and can be used to determine whether the resulting surface finish is acceptable or not. Nevertheless, while several existing tools and standards are available for computing surface roughness, these methods rely heavily on user input thus slowing down the analysis and increasing manufacturing costs. Therefore, fast and automatic determination of the roughness level is essential to avoid costs resulting from surfaces with unacceptable finish, and user-intensive analysis. In this study, we propose a Topological Data Analysis (TDA) based approach to classify the roughness level of synthetic surfaces using both their areal images and profiles. We utilize persistent homology from TDA to generate persistence diagrams that encapsulate information on the shape of the surface. We then obtain feature matrices for each surface or profile using Carlsson coordinates, persistence images, and template functions. We compare our results to two widely used methods in the literature: Fast Fourier Transform (FFT) and Gaussian filtering. The results show that our approach yields mean accuracies as high as 97%. We also show that, in contrast to existing surface analysis tools, our TDA-based approach is fully automatable and provides adaptive feature extraction. | ['Firas A. Khasawneh', 'Melih C. Yesilli'] | 2021-10-19 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 6.63558900e-01 -2.20451862e-01 5.41598916e-01 -7.42050409e-02
-6.90144420e-01 -4.99848574e-01 4.84494954e-01 6.78725898e-01
-9.79095548e-02 4.10715878e-01 -4.35839087e-01 -3.25240701e-01
-4.28517133e-01 -1.40263391e+00 -4.50783879e-01 -6.74903810e-01
-7.14310333e-02 4.74024504e-01 7.08946407e-01 -5.20610392e-01
7.56864071e-01 9.00418103e-01 -2.08857942e+00 2.26086035e-01
1.04539931e+00 1.27014112e+00 1.49045929e-01 6.78369224e-01
-3.50719303e-01 -1.79757893e-01 -3.68500352e-01 -1.02543212e-01
9.89378244e-03 -1.26373887e-01 -5.60198665e-01 -3.27570550e-02
4.08591866e-01 1.27499372e-01 4.98538464e-01 9.97665465e-01
1.07802786e-02 -1.09022900e-01 1.00586236e+00 -5.43123245e-01
-1.70529351e-01 -2.88945973e-01 -3.01616967e-01 -3.41960430e-01
5.96163630e-01 -2.70179152e-01 6.44409716e-01 -9.91595149e-01
7.11333096e-01 1.07106709e+00 9.29470003e-01 -6.70222715e-02
-1.37356126e+00 -1.63542047e-01 -3.50077778e-01 5.28184585e-02
-1.37934470e+00 -3.57653558e-01 9.50352013e-01 -7.12361753e-01
5.81628382e-01 5.46504974e-01 5.66705823e-01 2.54060566e-01
7.15735853e-01 -3.70891914e-02 1.24119103e+00 -7.34250009e-01
4.27598834e-01 -1.23503208e-01 2.33364388e-01 7.32176065e-01
5.20623267e-01 -1.37789041e-01 1.69701837e-02 -3.26619744e-01
1.03057730e+00 -1.55486539e-01 -1.41066879e-01 -5.13596416e-01
-1.06100249e+00 5.27923703e-01 -8.95390138e-02 2.92073667e-01
-5.70606947e-01 -1.33814186e-01 2.34664112e-01 5.11429191e-01
4.82960999e-01 5.63684106e-01 -2.54542291e-01 -2.37271756e-01
-5.88848650e-01 3.29049319e-01 7.56404698e-01 4.58269030e-01
1.24415362e+00 -3.34374070e-01 4.73373562e-01 7.57882714e-01
1.03132159e-01 7.44724393e-01 -1.28127471e-01 -1.08118951e+00
-8.15230086e-02 9.22381878e-01 3.61230373e-01 -1.68795788e+00
-4.08680439e-01 2.61141986e-01 -5.74844360e-01 7.69657493e-01
4.46778566e-01 4.67904925e-01 -7.38652587e-01 9.56246853e-01
3.13791662e-01 -3.92089665e-01 -1.51801974e-01 3.75181615e-01
2.40143105e-01 3.19621354e-01 -4.95483339e-01 -1.72243103e-01
1.17883408e+00 -2.10965335e-01 -5.77133775e-01 4.55358803e-01
5.32528698e-01 -1.10238993e+00 1.29409063e+00 7.89247930e-01
-1.02186871e+00 -2.84614712e-01 -1.13057768e+00 2.71813393e-01
-5.00497162e-01 3.09310526e-01 3.85155648e-01 6.74674571e-01
-9.34540331e-01 1.19465494e+00 -9.97438967e-01 -2.25243703e-01
6.42826334e-02 4.11431074e-01 -3.50938797e-01 1.96908072e-01
-7.69474626e-01 7.74809361e-01 -3.50153059e-01 -4.95317690e-02
2.65176762e-02 -6.12620533e-01 -5.87580085e-01 -1.65924609e-01
1.44721165e-01 -2.62945831e-01 8.70384455e-01 -5.84160805e-01
-1.82417905e+00 5.67050695e-01 -2.73135960e-01 -5.29287346e-02
4.61240590e-01 -7.89286569e-02 -2.14824513e-01 2.71117270e-01
-1.78947486e-02 -1.75920069e-01 7.01085329e-01 -1.17418122e+00
-2.47811675e-01 -4.33033407e-01 -1.89695776e-01 -3.57549131e-01
-8.71251244e-03 -2.98759490e-01 -1.06557153e-01 -2.49387264e-01
4.68588114e-01 -7.80901134e-01 -2.19592288e-01 1.15240335e-01
-1.97156891e-01 -4.80610179e-03 8.86590719e-01 -7.09121585e-01
1.36809719e+00 -2.04248667e+00 -2.19528705e-01 7.86971033e-01
8.66301134e-02 1.29616097e-01 1.13481939e-01 8.11349690e-01
2.71515638e-01 1.42189160e-01 -5.51242232e-01 1.48447901e-01
-2.28942677e-01 -1.28592804e-01 -8.09977278e-02 5.27644575e-01
1.85236111e-01 2.19675243e-01 -6.67766750e-01 -3.30166489e-01
5.25192142e-01 6.58392489e-01 -5.07928252e-01 -1.80981860e-01
-1.12813763e-01 2.59031683e-01 -5.51166236e-01 7.37381637e-01
7.28648543e-01 4.37848382e-02 1.65778607e-01 -4.61635709e-01
-6.69792652e-01 1.51342779e-01 -1.29110682e+00 9.54414308e-01
-6.87636733e-01 5.38867712e-01 3.08325291e-01 -6.59029126e-01
1.47250938e+00 1.41265420e-02 3.78730744e-01 -6.02842093e-01
2.26035148e-01 7.23570764e-01 -2.38771185e-01 -2.80139953e-01
4.86066222e-01 5.56635931e-02 1.68275058e-01 1.75759763e-01
-7.82967567e-01 -5.38000762e-01 4.54218872e-02 -3.58268112e-01
1.12449586e+00 4.73921858e-02 1.89843565e-01 -7.37465680e-01
8.15107524e-01 1.16374597e-01 4.83301803e-02 1.72733665e-01
3.60624045e-01 6.45222127e-01 6.55975342e-01 -5.96167803e-01
-1.21267641e+00 -1.01997197e+00 -3.65916133e-01 3.46516490e-01
3.65185708e-01 -5.28736949e-01 -9.18427289e-01 1.23810014e-02
1.71757519e-01 3.86527598e-01 -7.07179487e-01 1.46093130e-01
-6.36705518e-01 -3.24388891e-01 -5.81068546e-02 1.17060564e-01
1.44454539e-01 -8.42717528e-01 -6.85289383e-01 1.62103340e-01
2.00850710e-01 -6.50679231e-01 -4.91215214e-02 -3.85664910e-01
-1.11216152e+00 -1.35903513e+00 -4.22592491e-01 -4.52106774e-01
8.75155449e-01 2.93773949e-01 9.23890233e-01 2.58635521e-01
-5.82636178e-01 3.97228152e-01 -2.89094388e-01 -1.76829711e-01
-6.63545728e-01 -1.47957802e-01 5.35806082e-02 1.53989136e-01
-3.54428887e-02 -7.57673323e-01 -6.89559996e-01 7.40987718e-01
-8.34535360e-01 2.17667427e-02 4.77079839e-01 2.89114058e-01
1.06765795e+00 1.73589438e-01 6.07594132e-01 -9.63972211e-01
6.86934233e-01 -5.22298925e-02 -8.36181760e-01 1.02789253e-01
-7.65443683e-01 5.86486533e-02 6.10924304e-01 -1.55819774e-01
-6.04718626e-01 1.00610211e-01 -3.05947214e-01 -1.30665556e-01
-3.02920374e-03 4.55099463e-01 -4.32461128e-02 -5.27912617e-01
7.14821935e-01 1.29326910e-01 6.41737103e-01 -5.98423839e-01
-1.02639616e-01 6.81913793e-01 3.19559634e-01 -5.03521502e-01
5.32647789e-01 8.01341176e-01 3.59794974e-01 -1.59064031e+00
-1.26846939e-01 -3.65158916e-01 -6.69522703e-01 -4.82194811e-01
6.53540313e-01 -1.66926637e-01 -9.63715136e-01 6.17888868e-01
-1.08802855e+00 -2.02166036e-01 -1.30039304e-01 1.06961370e-01
-8.05435777e-01 6.26474202e-01 -1.49943814e-01 -9.35321689e-01
-2.84744591e-01 -1.12548161e+00 1.23802221e+00 -2.07859963e-01
-4.50890630e-01 -9.55759704e-01 1.25276580e-01 -1.70553960e-02
8.14963102e-01 7.95055509e-01 1.10558116e+00 1.98004812e-01
-4.93199617e-01 -6.25511467e-01 -8.24911147e-02 2.05174685e-01
4.63714540e-01 5.70985198e-01 -5.38717926e-01 -9.01005641e-02
-1.40411928e-01 2.94827253e-01 4.66782361e-01 3.19244266e-01
8.56042266e-01 4.40111570e-02 -3.15136641e-01 1.93249375e-01
1.50675547e+00 2.24110745e-02 1.02959692e+00 5.51002800e-01
3.86214435e-01 1.02773714e+00 9.01859403e-01 2.87649810e-01
-4.10584025e-02 8.82754683e-01 4.04829562e-01 8.52752626e-02
-3.39338817e-02 -1.32406224e-02 3.85752171e-01 8.65935683e-01
-5.50253630e-01 2.89085239e-01 -1.03787816e+00 3.52247149e-01
-1.46184385e+00 -6.25911534e-01 -7.90765226e-01 2.59231830e+00
4.47885394e-01 8.37290138e-02 9.02282745e-02 4.98925477e-01
7.45922387e-01 -2.82088339e-01 -2.19365418e-01 -6.24351442e-01
3.64921354e-02 4.80751157e-01 5.79393625e-01 7.85482347e-01
-7.31164098e-01 6.58129394e-01 6.57820320e+00 6.18413568e-01
-1.33199346e+00 -3.80526215e-01 1.11332163e-01 6.06101036e-01
-5.21371067e-01 -9.05150622e-02 -4.33803499e-01 2.51998901e-01
8.92967343e-01 -2.33196244e-01 2.01142505e-01 8.48446310e-01
2.91556448e-01 -3.46331328e-01 -6.12185538e-01 7.57803798e-01
-2.73190230e-01 -1.30677390e+00 7.25110620e-02 3.05624723e-01
3.96218002e-01 -2.97452778e-01 -1.73583448e-01 -4.37661707e-01
-3.01435649e-01 -7.88375914e-01 5.84196687e-01 9.14640307e-01
1.13011134e+00 -7.38557875e-01 7.99373209e-01 8.34968761e-02
-1.48568988e+00 2.78899819e-01 -2.25159645e-01 -1.61962762e-01
2.07129776e-01 8.84601057e-01 -6.23812139e-01 4.51791644e-01
5.00624478e-01 3.40721190e-01 -4.59235638e-01 7.96123147e-01
1.85726017e-01 2.98134059e-01 -5.97190559e-01 -1.08723290e-01
-2.87559420e-01 -6.96697533e-01 5.64681768e-01 8.29612136e-01
7.03896284e-01 -1.13083877e-01 -1.13960028e-01 7.36242950e-01
6.25954330e-01 5.58651447e-01 -7.68879592e-01 -1.45751145e-02
4.49086428e-01 1.10434866e+00 -1.02467775e+00 1.94902830e-02
-1.57723948e-01 6.41001821e-01 -9.02339667e-02 -7.16625154e-02
-2.96609163e-01 -8.80167723e-01 6.73918605e-01 9.37688589e-01
1.97951406e-01 -6.96321368e-01 -5.96815705e-01 -5.56041420e-01
6.08949773e-02 -5.62499940e-01 -2.57749766e-01 -4.61616725e-01
-1.02507782e+00 5.10275602e-01 -1.42496347e-01 -1.24901783e+00
1.06982477e-01 -9.84296322e-01 -6.03794873e-01 9.15470243e-01
-1.17270732e+00 -6.72956526e-01 -3.62232298e-01 1.04559831e-01
1.82028726e-01 1.77497968e-01 1.00040638e+00 2.48420686e-01
-6.79170787e-02 3.00594158e-02 2.52206773e-01 -4.73520696e-01
3.04899722e-01 -1.06990278e+00 6.14073634e-01 3.39933962e-01
-5.49692810e-01 7.48069584e-01 9.25901830e-01 -8.76522839e-01
-1.77226210e+00 -6.73203349e-01 7.12787330e-01 -4.04721171e-01
7.32691228e-01 -1.75989792e-01 -1.21099603e+00 -9.64664817e-02
-4.83988732e-01 -4.22432542e-01 3.60972971e-01 2.20269263e-02
1.18842028e-01 -2.44984645e-02 -1.11613226e+00 5.45223117e-01
6.85049891e-01 -2.96560436e-01 -3.63728106e-01 -7.46113658e-02
2.10619524e-01 -4.27922793e-02 -1.18478096e+00 6.69796765e-01
9.38361764e-01 -1.34432745e+00 7.80508518e-01 1.60512418e-01
2.80789226e-01 -4.62168068e-01 -2.50498503e-01 -8.35848272e-01
-2.50328869e-01 -6.18666470e-01 4.46630836e-01 8.00017953e-01
4.02952194e-01 -9.71510768e-01 9.67817903e-01 3.85105759e-01
-2.48676077e-01 -9.95008469e-01 -8.97392333e-01 -8.61977696e-01
-4.74329740e-02 -3.53288263e-01 4.83713984e-01 6.02416992e-01
-1.75807327e-01 -1.35222614e-01 2.50190437e-01 2.64111787e-01
6.76662147e-01 3.15716177e-01 8.50697279e-01 -1.90741527e+00
2.62213230e-01 -4.86844242e-01 -6.72966242e-01 -4.03695017e-01
-2.41021454e-01 -5.13435066e-01 6.91242144e-02 -1.53329968e+00
-4.35655922e-01 -8.50683510e-01 1.70860797e-01 7.41122141e-02
2.64039814e-01 1.30457401e-01 -4.43614304e-01 2.96345711e-01
4.52473015e-02 2.53912926e-01 1.11490309e+00 3.14139873e-01
-4.58842069e-01 3.57589163e-02 -2.29809746e-01 8.81811082e-01
8.76577318e-01 3.89071703e-02 -6.13189377e-02 3.85593362e-02
6.21873558e-01 -1.69287547e-01 2.11136028e-01 -1.19451129e+00
-1.31373376e-01 -1.84714824e-01 -1.40102521e-01 -5.85075319e-01
2.03304276e-01 -7.42736638e-01 4.65476900e-01 5.14109313e-01
1.68920323e-01 1.17629975e-01 1.35640606e-01 5.50922453e-01
-1.56892270e-01 -1.84994847e-01 9.23496246e-01 9.81497169e-02
-2.19571188e-01 -7.63622895e-02 -4.19835955e-01 -6.23084307e-01
8.87001991e-01 -6.38630390e-01 -2.98212230e-01 -2.22855285e-01
-6.13234520e-01 -2.99481004e-01 1.36783612e+00 -1.19376063e-01
7.46554494e-01 -1.11116254e+00 -3.95235389e-01 4.75938350e-01
-2.53113527e-02 -1.43431947e-01 3.04430634e-01 9.05240893e-01
-1.30571687e+00 1.60506442e-01 -2.78394699e-01 -7.08743930e-01
-1.38552439e+00 2.33748943e-01 -2.26469990e-02 1.25065194e-02
-6.31274760e-01 2.26655841e-01 -2.96817757e-02 -1.82418168e-01
-4.81045157e-01 -4.37502503e-01 -1.71676770e-01 8.42906833e-02
4.09148216e-01 8.90092671e-01 5.84433854e-01 -5.84843636e-01
-2.83598959e-01 1.36826241e+00 2.31340423e-01 -5.02105951e-02
1.29698765e+00 -8.47892836e-02 -3.60276341e-01 6.51678085e-01
1.07679904e+00 5.86719215e-01 -8.28043163e-01 5.20371497e-01
8.69755372e-02 -5.05203664e-01 -7.63697624e-02 -2.12775409e-01
-4.21973377e-01 9.58209574e-01 5.37919879e-01 8.32410514e-01
1.02703178e+00 -1.55887619e-01 7.59865403e-01 2.76101708e-01
6.86926782e-01 -1.08336389e+00 -1.78662255e-01 4.35896575e-01
1.09105361e+00 -6.58211529e-01 2.23040730e-01 -1.15058434e+00
-4.53176573e-02 1.39811814e+00 -3.08730584e-02 -4.49530602e-01
6.77981317e-01 2.68126816e-01 1.75042748e-01 -5.06934941e-01
-3.77552241e-01 -4.86805253e-02 2.28172064e-01 4.42460954e-01
3.96202147e-01 7.04355910e-02 -4.99876916e-01 2.65196532e-01
-3.47609222e-01 -1.66390359e-01 6.04442894e-01 1.17035270e+00
-1.01192498e+00 -1.39744592e+00 -6.17480397e-01 6.12296104e-01
-2.88511813e-01 2.38835305e-01 -3.92475963e-01 7.12419093e-01
-2.99033195e-01 5.22512376e-01 6.67226389e-02 -5.61588287e-01
6.85999930e-01 -4.24998067e-02 5.50392210e-01 -5.25907397e-01
-8.57152417e-02 -2.41490766e-01 2.72863746e-01 -6.30751252e-01
-1.57779008e-01 -8.26996624e-01 -1.22167277e+00 -2.19455570e-01
-5.12072444e-01 1.91447079e-01 1.17636073e+00 4.53949124e-01
6.13330543e-01 2.74451643e-01 7.07563698e-01 -1.09524119e+00
-8.84470865e-02 -5.90553582e-01 -6.09754324e-01 3.18850219e-01
1.11447178e-01 -1.11390960e+00 -5.16058028e-01 -9.94360074e-04] | [8.607358932495117, -2.528127431869507] |
3c1b992a-956e-46c7-b2b6-8d07f0952cbc | learning-activation-functions-for-sparse | 2305.10964 | null | https://arxiv.org/abs/2305.10964v2 | https://arxiv.org/pdf/2305.10964v2.pdf | Learning Activation Functions for Sparse Neural Networks | Sparse Neural Networks (SNNs) can potentially demonstrate similar performance to their dense counterparts while saving significant energy and memory at inference. However, the accuracy drop incurred by SNNs, especially at high pruning ratios, can be an issue in critical deployment conditions. While recent works mitigate this issue through sophisticated pruning techniques, we shift our focus to an overlooked factor: hyperparameters and activation functions. Our analyses have shown that the accuracy drop can additionally be attributed to (i) Using ReLU as the default choice for activation functions unanimously, and (ii) Fine-tuning SNNs with the same hyperparameters as dense counterparts. Thus, we focus on learning a novel way to tune activation functions for sparse networks and combining these with a separate hyperparameter optimization (HPO) regime for sparse networks. By conducting experiments on popular DNN models (LeNet-5, VGG-16, ResNet-18, and EfficientNet-B0) trained on MNIST, CIFAR-10, and ImageNet-16 datasets, we show that the novel combination of these two approaches, dubbed Sparse Activation Function Search, short: SAFS, results in up to 15.53%, 8.88%, and 6.33% absolute improvement in the accuracy for LeNet-5, VGG-16, and ResNet-18 over the default training protocols, especially at high pruning ratios. Our code can be found at https://github.com/automl/SAFS | ['Marius Lindauer', 'Mehdi Asadi', 'Aditya Mohan', 'Mohammad Loni'] | 2023-05-18 | null | null | null | null | ['automl', 'hyperparameter-optimization'] | ['methodology', 'methodology'] | [-8.62194151e-02 1.16720088e-01 -2.05230683e-01 -3.57919753e-01
-3.50985110e-01 -3.58281255e-01 2.77852297e-01 -1.96961373e-01
-5.81615210e-01 9.73687053e-01 -1.87879950e-02 -3.85753542e-01
-2.02644497e-01 -8.39144647e-01 -8.15963447e-01 -7.59211540e-01
-7.30158836e-02 4.03710812e-01 2.24411532e-01 -2.35523954e-02
-9.47075635e-02 5.70023656e-01 -1.28164601e+00 -9.08635482e-02
8.04913580e-01 1.13813329e+00 1.88229159e-01 3.15353364e-01
2.73461565e-02 9.00080383e-01 -7.23487377e-01 -4.55748707e-01
3.73505741e-01 8.23876914e-03 -6.49447381e-01 -5.24080575e-01
5.61311901e-01 -4.44724083e-01 -6.04426265e-01 1.11824131e+00
6.70932829e-01 2.27069125e-01 2.31056809e-01 -1.20268238e+00
-1.77326486e-01 1.15337014e+00 -5.20421863e-01 3.66082370e-01
-4.32080030e-01 2.97318608e-01 8.66129160e-01 -7.19725966e-01
3.25454324e-01 1.11592293e+00 7.88806617e-01 7.34330177e-01
-1.28671718e+00 -1.14250088e+00 2.48082235e-01 1.63471013e-01
-1.73150516e+00 -8.02369535e-01 3.75438809e-01 -3.80898779e-03
1.25425136e+00 -2.98912041e-02 6.85672045e-01 1.26732707e+00
-1.72928423e-01 5.68318963e-01 6.22826695e-01 -1.36606798e-01
5.44043720e-01 -1.51401563e-02 3.44343841e-01 6.06666505e-01
8.53707910e-01 7.91504234e-03 -4.38528121e-01 -2.92459689e-02
7.24719465e-01 5.02914675e-02 -3.41789842e-01 5.19443080e-02
-7.10345685e-01 7.77727604e-01 7.79987395e-01 2.52890795e-01
-4.71910566e-01 7.20800579e-01 3.76889974e-01 1.31171957e-01
3.32279325e-01 4.83966708e-01 -4.88619149e-01 -2.45455369e-01
-1.17923617e+00 2.04806745e-01 9.31051314e-01 9.34843957e-01
8.21335137e-01 6.10947371e-01 -1.15260378e-01 9.72315490e-01
1.57399639e-01 3.53950143e-01 4.22953725e-01 -1.00016630e+00
6.19716346e-01 4.11626577e-01 -3.34087640e-01 -6.67299151e-01
-4.61336821e-01 -7.96353281e-01 -1.10306609e+00 -7.37961680e-02
1.82292551e-01 -6.08005524e-01 -1.17708302e+00 1.91845405e+00
8.19018632e-02 4.19294357e-01 -5.69621883e-02 7.50516593e-01
1.15579140e+00 6.47178173e-01 3.02272290e-01 1.72846854e-01
1.12604177e+00 -1.10082924e+00 -4.43145812e-01 -5.14377892e-01
5.74106812e-01 -2.66465694e-01 8.60159695e-01 3.15630466e-01
-1.27214813e+00 -2.96222180e-01 -1.25915396e+00 2.05257293e-02
-2.31053814e-01 1.71847597e-01 7.71018326e-01 5.97943842e-01
-1.40780795e+00 8.11235726e-01 -1.15802586e+00 -3.38542521e-01
9.48680818e-01 8.29555869e-01 7.81250652e-03 -1.15336761e-01
-1.06820226e+00 7.88901627e-01 4.90959913e-01 1.08588353e-01
-1.06250620e+00 -1.06232834e+00 -6.21011913e-01 5.43914199e-01
2.74267256e-01 -7.78840721e-01 1.20033872e+00 -7.77499557e-01
-1.34267187e+00 4.09743845e-01 5.61639713e-03 -1.00074029e+00
7.91830942e-02 -2.24072680e-01 -2.54018635e-01 1.84764937e-01
-3.10115457e-01 1.08207965e+00 5.46244979e-01 -9.71698046e-01
-3.05938661e-01 -1.91831976e-01 2.17645139e-01 1.06493883e-01
-4.94375885e-01 -2.35557631e-01 -4.10594821e-01 -5.86189628e-01
4.71194461e-02 -8.85990143e-01 -4.11828101e-01 -2.28134498e-01
-5.40132284e-01 -4.48754504e-02 7.69276798e-01 -3.24788779e-01
1.36015391e+00 -2.12263036e+00 -1.53754637e-01 3.82353127e-01
4.60199684e-01 6.37851715e-01 -2.05827907e-01 -1.49830118e-01
4.36751880e-02 3.40842515e-01 -1.01617977e-01 -4.42435265e-01
-2.13516980e-01 3.31819922e-01 -3.39518711e-02 2.58795500e-01
1.19902171e-01 8.46437335e-01 -5.80124974e-01 -3.28941107e-01
1.08779669e-01 7.12555408e-01 -7.92803109e-01 -1.92988157e-01
-7.15929568e-02 -3.29228342e-02 -3.53435487e-01 8.24748337e-01
6.22986615e-01 -3.90762925e-01 2.89398551e-01 -4.10856247e-01
2.71854132e-01 5.12350619e-01 -9.24822450e-01 1.48374438e+00
-4.20206815e-01 6.25970900e-01 2.53154546e-01 -1.01680815e+00
8.39191496e-01 2.81900108e-01 3.31787139e-01 -7.62767732e-01
3.86142999e-01 2.93572158e-01 2.49184951e-01 -2.37324852e-02
3.46678466e-01 1.49465725e-01 2.89688587e-01 2.30996102e-01
4.90623921e-01 4.18729782e-01 3.15201640e-01 3.12982142e-01
1.42476559e+00 -4.59595859e-01 -4.82265614e-02 -4.21925068e-01
6.01498857e-02 -3.39696139e-01 6.78256154e-01 8.57910156e-01
-1.03029698e-01 4.72451597e-01 5.78656137e-01 -3.22665334e-01
-9.61092293e-01 -9.42811966e-01 -2.17174008e-01 9.63228285e-01
-8.05594772e-02 -5.37138879e-01 -8.65107715e-01 -5.65570533e-01
-4.75621969e-02 7.27045774e-01 -3.31981897e-01 -2.88370937e-01
-7.42829263e-01 -9.86827374e-01 1.00720513e+00 6.76183224e-01
8.84418964e-01 -9.28383112e-01 -4.64323848e-01 1.16162203e-01
5.28010093e-02 -1.27913427e+00 -1.56004265e-01 7.35964179e-01
-1.24646294e+00 -7.70483851e-01 -4.88259554e-01 -4.87099528e-01
7.57826507e-01 1.17954202e-01 1.24424994e+00 3.56873840e-01
-2.54779458e-02 -4.39673103e-02 -2.01288804e-01 -1.53710142e-01
-3.05970255e-02 6.90907121e-01 -4.57345415e-03 -3.72395068e-01
2.67077923e-01 -1.00062454e+00 -7.82484770e-01 1.58149585e-01
-7.59411752e-01 -3.57029177e-02 6.97130263e-01 6.41309023e-01
6.22200072e-01 -6.20639659e-02 7.71073103e-01 -1.14774978e+00
2.96691775e-01 -7.50157416e-01 -7.12568700e-01 2.84954980e-02
-8.62607777e-01 1.56753659e-01 7.14315355e-01 -4.46670920e-01
-8.17777634e-01 3.61037664e-02 -5.17003953e-01 -7.68341243e-01
-1.51482532e-02 3.59599143e-01 -2.27306411e-01 -3.86961281e-01
8.07295561e-01 -7.95659348e-02 -3.08517545e-01 -5.29860497e-01
7.67668709e-02 3.00202459e-01 3.95914018e-01 -6.34850740e-01
6.82555377e-01 3.06507498e-01 -9.13562477e-02 -7.13719904e-01
-7.89517820e-01 -9.65078175e-02 -1.88016072e-02 1.70627475e-01
4.05504704e-01 -1.21034360e+00 -5.25620639e-01 3.65785837e-01
-8.51763129e-01 -7.21604943e-01 -3.10427368e-01 4.23510373e-01
2.16558985e-02 -1.79066381e-04 -7.22703397e-01 -4.53311145e-01
-8.89121950e-01 -1.16447628e+00 6.37309372e-01 4.65593278e-01
-1.53155729e-01 -8.78906250e-01 -3.40657830e-01 2.23924145e-01
8.21233988e-01 4.41263281e-02 9.17658508e-01 -8.19098055e-01
-7.07118988e-01 1.54785380e-01 -4.51490372e-01 5.03539860e-01
-3.06651294e-01 -8.62401351e-02 -1.05902982e+00 -4.53884512e-01
-2.35087872e-01 -4.79068875e-01 1.11839259e+00 7.36164451e-01
1.49584782e+00 -5.75021684e-01 -3.62314701e-01 1.32449639e+00
1.65346873e+00 5.61945662e-02 8.59064519e-01 1.77276433e-01
7.88928390e-01 -1.93222165e-01 -5.48122674e-02 3.70214731e-01
1.62579760e-01 4.99050111e-01 7.01257944e-01 -8.20487216e-02
-3.38053048e-01 -1.57230183e-01 2.94974446e-01 7.46478140e-01
-1.81770429e-01 -3.49552155e-01 -7.37779438e-01 4.81200129e-01
-1.58604860e+00 -8.23626459e-01 2.36219972e-01 2.03015876e+00
9.06172812e-01 3.40115517e-01 -4.76407148e-02 1.54954746e-01
6.27196610e-01 3.09137881e-01 -8.62304509e-01 -3.15157086e-01
-1.26051188e-01 6.47008181e-01 9.34755385e-01 2.44962752e-01
-8.65234554e-01 9.79586005e-01 5.63739920e+00 1.01965618e+00
-1.32814300e+00 3.07865322e-01 9.13582623e-01 -7.79879212e-01
-1.98114842e-01 -1.44385561e-01 -1.32875896e+00 5.54163694e-01
1.30800271e+00 8.87380019e-02 7.17078686e-01 9.39639628e-01
3.84524427e-02 -3.34888920e-02 -1.00245821e+00 1.11882174e+00
-2.94469208e-01 -1.55725121e+00 5.94970733e-02 -4.07957919e-02
8.24783623e-01 7.09277034e-01 -3.93566228e-02 5.95897019e-01
4.08660352e-01 -1.27580273e+00 7.34378815e-01 -9.72388238e-02
8.42427850e-01 -8.17019343e-01 7.22105742e-01 4.28261682e-02
-1.07228923e+00 -1.17414139e-01 -6.46295607e-01 1.51747853e-01
-3.99395972e-02 9.31293368e-01 -6.19418263e-01 8.74565728e-03
9.75769520e-01 6.32814527e-01 -4.91742045e-01 1.25969958e+00
-2.86342084e-01 1.17485285e+00 -6.86848283e-01 -6.68996386e-03
2.94981927e-01 1.75709650e-01 4.21465814e-01 1.19294095e+00
3.28834027e-01 1.92965269e-02 -3.45188111e-01 1.08015013e+00
-6.57690227e-01 -4.10199195e-01 -3.01431239e-01 2.17256062e-02
1.04401219e+00 1.46060741e+00 -7.77479410e-01 -2.11616993e-01
-1.19180568e-01 3.78349662e-01 5.15516698e-01 6.64165795e-01
-1.11406934e+00 -4.58334088e-01 8.26722860e-01 2.46177405e-01
4.68094260e-01 -3.23560573e-02 -5.25242746e-01 -9.87803638e-01
9.71915666e-04 -8.13033164e-01 1.88159674e-01 -4.85136151e-01
-8.59356821e-01 7.85386205e-01 3.45965214e-02 -5.70857108e-01
3.37085314e-02 -3.09683144e-01 -7.25383341e-01 6.05363011e-01
-1.54053211e+00 -6.49971247e-01 -5.13111055e-01 5.95071435e-01
4.16297615e-01 -1.52498484e-01 4.93689954e-01 7.87502110e-01
-9.35494125e-01 1.06358206e+00 -1.40620206e-04 9.07474458e-02
2.23700479e-01 -8.35164905e-01 4.55855131e-01 7.18654275e-01
5.82585670e-02 5.84461629e-01 4.19841290e-01 -2.53948987e-01
-1.31029737e+00 -1.43582189e+00 5.44364095e-01 1.12792484e-01
4.59055185e-01 -5.27303934e-01 -9.21788812e-01 7.66383290e-01
6.81420416e-02 1.01605356e-01 3.45456183e-01 2.03588396e-01
-1.63196906e-01 -3.75696957e-01 -1.30659997e+00 6.03610992e-01
1.25350618e+00 -4.35048789e-02 2.98595987e-02 2.27954194e-01
9.30342138e-01 -5.08166254e-01 -9.22382653e-01 5.41536629e-01
4.02659237e-01 -8.97744298e-01 1.05804801e+00 -1.80414557e-01
3.20526958e-01 -3.57077792e-02 -2.13553786e-01 -1.25935304e+00
-2.43271574e-01 -5.08373678e-01 -4.70586032e-01 1.17785692e+00
5.49404144e-01 -8.35001409e-01 1.12416053e+00 5.32867312e-01
-2.69751161e-01 -1.15344393e+00 -1.00064826e+00 -7.63439715e-01
-5.01728542e-02 -3.15871537e-01 7.95607507e-01 6.32267892e-01
-5.84498107e-01 3.16351354e-01 -1.19576894e-01 1.62896261e-01
5.10677755e-01 -4.61804509e-01 3.65162402e-01 -1.06774592e+00
-2.58981109e-01 -6.13659859e-01 -2.23103493e-01 -1.08323538e+00
1.54965267e-01 -8.10142457e-01 -2.35803515e-01 -1.38115776e+00
1.15695801e-02 -9.08337057e-01 -3.26538771e-01 9.01953816e-01
1.72877222e-01 3.19509476e-01 1.84304476e-01 2.04338640e-01
-5.59083045e-01 4.68098760e-01 8.95533025e-01 -7.03408346e-02
-1.51240587e-01 -1.17549844e-01 -8.54770422e-01 6.78264797e-01
1.01561737e+00 -6.29247665e-01 -6.10322833e-01 -8.83345485e-01
2.83631921e-01 -2.88439125e-01 5.37133098e-01 -1.36284339e+00
4.48615968e-01 1.82140648e-01 3.49537432e-01 -4.00074542e-01
4.11701024e-01 -7.52925038e-01 2.13338688e-01 4.53980833e-01
-1.88079879e-01 -4.57139909e-02 5.58238089e-01 2.33126327e-01
-1.46809652e-01 -3.92192602e-01 9.23975408e-01 -7.12149590e-02
-6.53479576e-01 5.03929079e-01 -1.76371977e-01 3.02633137e-01
6.68078482e-01 -3.30384254e-01 -7.91702271e-01 -1.89240605e-01
-5.56365252e-01 3.57077092e-01 8.03257152e-02 1.87513325e-02
5.30559599e-01 -1.11524498e+00 -4.81582552e-01 2.32408956e-01
-5.23640275e-01 4.41940904e-01 3.21828991e-01 9.44233239e-01
-6.64426327e-01 4.37234908e-01 -9.47033912e-02 -3.84013593e-01
-9.28668678e-01 6.25571888e-03 6.53193176e-01 -3.42335165e-01
-5.82280457e-01 1.12292111e+00 1.35833949e-01 -1.40657842e-01
7.24463046e-01 -4.15977031e-01 1.02454692e-01 -2.49166101e-01
1.74326703e-01 5.31241238e-01 3.98718625e-01 -1.82513148e-01
-4.97194171e-01 1.56891361e-01 -2.04689249e-01 4.24862623e-01
1.53440404e+00 2.29546443e-01 7.09038675e-02 -2.27558762e-01
1.16474402e+00 -4.14752036e-01 -1.17998993e+00 -1.97921529e-01
-2.60457069e-01 1.15052145e-02 4.93357599e-01 -7.91451573e-01
-1.90944982e+00 6.86446428e-01 5.81664562e-01 -5.20098917e-02
1.35409212e+00 -3.30273546e-02 1.04166424e+00 6.37525380e-01
3.29363316e-01 -9.14737880e-01 -2.68191755e-01 6.69356525e-01
4.38847214e-01 -8.62987816e-01 1.16322935e-01 -1.15041696e-01
-2.48324528e-01 7.74632871e-01 9.56668615e-01 -2.66532302e-01
7.84109652e-01 6.89586163e-01 -3.74173820e-01 -3.75643641e-01
-1.06575680e+00 8.38651210e-02 -1.79696009e-01 2.54782796e-01
3.00752074e-01 -3.01870592e-02 1.00932732e-01 6.98710144e-01
-3.93908620e-01 9.68289003e-02 2.51475930e-01 7.13370383e-01
-5.15771270e-01 -7.18615651e-01 1.65899247e-02 9.63341415e-01
-6.34840369e-01 -5.64064801e-01 1.95082594e-02 7.26778448e-01
7.38326162e-02 6.08283520e-01 2.18276322e-01 -5.31000555e-01
2.69110382e-01 -1.87598914e-01 4.23497975e-01 -5.73944867e-01
-9.41896498e-01 -2.74342567e-01 2.93570429e-01 -6.50444090e-01
-1.84278607e-01 -2.17648223e-01 -1.36923015e+00 -7.03310788e-01
-4.32820886e-01 9.19404030e-02 7.36762285e-01 7.12236464e-01
5.93919337e-01 6.99697196e-01 1.59289747e-01 -8.21488440e-01
-6.93243265e-01 -9.27627683e-01 -5.11749148e-01 -8.82788301e-02
6.69102296e-02 -6.56880796e-01 -5.73516548e-01 -4.36995864e-01] | [8.599345207214355, 3.1686863899230957] |
7010d59c-930c-47aa-bf3e-233d5cc3b1ff | jointly-learning-topic-specific-word-and | null | null | https://openreview.net/forum?id=Vx8l4vwv94 | https://openreview.net/pdf?id=Vx8l4vwv94 | JOINTLY LEARNING TOPIC SPECIFIC WORD AND DOCUMENT EMBEDDING | Document embedding generally ignores underlying topics, which fails to capture polysemous terms that can mislead to improper thematic representation. Moreover, embedding a new document during the test process needs a complex and expensive inference method. Some models first learn word embeddings and later learn underlying topics using a clustering algorithm for document representation; those methods miss the mutual interaction between the two paradigms. To this point, we propose a novel document-embedding method by weighted averaging of jointly learning topic-specific word embeddings called TDE: Topical Document Embedding, which efficiently captures syntactic and semantic properties by utilizing three levels of knowledge -i.e., word, topic, and document. TDE obtains document vectors on the fly simultaneously during the jointly learning process of the topical word embeddings. Experiments demonstrate better topical word embeddings using document vector as a global context and better document classification results on the obtained document embeddings by the proposed method over the recent related models. | ['Zuping Zhang', 'Farid Uddin'] | 2021-09-29 | null | null | null | null | ['document-embedding'] | ['methodology'] | [-1.10053793e-01 -9.25195962e-02 -4.59848285e-01 -4.13321495e-01
-4.35109198e-01 -4.33903426e-01 8.85077238e-01 5.40829480e-01
-3.16071659e-01 3.81381214e-01 6.26057684e-01 -9.56260711e-02
-3.99453491e-01 -8.50163639e-01 -2.08565474e-01 -8.40809286e-01
1.28144369e-01 5.04569054e-01 -3.93538699e-02 1.00717060e-01
5.61532855e-01 -1.52843386e-01 -1.61596751e+00 2.15865389e-01
8.66465211e-01 6.95939898e-01 3.35662127e-01 3.43250006e-01
-8.35860968e-01 1.15875050e-01 -8.04804027e-01 -2.79261768e-01
-3.76536697e-01 -1.67106360e-01 -5.38634539e-01 2.22976789e-01
1.42036989e-01 -5.87599054e-02 -1.26457065e-01 1.02832603e+00
2.83675969e-01 3.24071974e-01 1.05484211e+00 -1.21565652e+00
-1.24740613e+00 5.76520503e-01 -4.92013812e-01 1.01466857e-01
4.77255471e-02 -4.83084053e-01 1.51692486e+00 -1.28852081e+00
3.56024832e-01 1.29555678e+00 3.69219214e-01 3.90784591e-01
-1.28012300e+00 -5.49945533e-01 5.15133381e-01 3.38105857e-01
-1.41528416e+00 1.83040604e-01 1.18801332e+00 -5.70080340e-01
9.18390989e-01 1.72218174e-01 7.14925826e-01 1.38777232e+00
3.62087190e-01 6.80021167e-01 8.43342721e-01 -2.91217893e-01
2.50613689e-01 7.53984749e-01 8.04089129e-01 4.53088760e-01
7.62211323e-01 -3.42186719e-01 -4.84351814e-01 -4.02298808e-01
2.34483838e-01 4.88328874e-01 -3.11665714e-01 -4.74238515e-01
-1.05198681e+00 1.14247656e+00 -9.11993980e-02 6.26791775e-01
-2.70301044e-01 1.05316788e-01 5.40196955e-01 2.75219887e-01
8.61006141e-01 2.81591713e-01 -4.25170153e-01 9.38111264e-03
-8.85209382e-01 2.49636889e-01 5.87158024e-01 7.20245123e-01
9.13175762e-01 -8.50792900e-02 -3.03532273e-01 9.85368729e-01
7.22341180e-01 2.92675585e-01 1.07561982e+00 -1.48799255e-01
2.67848343e-01 6.60346925e-01 -5.17100580e-02 -1.45920229e+00
1.87507812e-02 -4.74901766e-01 -4.35974568e-01 -3.55601460e-01
-3.39854181e-01 1.89453304e-01 -7.77963340e-01 1.61910009e+00
3.00685406e-01 1.94156334e-01 1.95714667e-01 5.36256909e-01
7.30258346e-01 1.02399254e+00 2.11540118e-01 -2.54190058e-01
1.67025280e+00 -1.05297172e+00 -1.12622678e+00 -1.80021569e-01
7.71837711e-01 -5.85179389e-01 9.71436560e-01 2.95116395e-01
-7.11454153e-01 -5.98802269e-01 -1.12505364e+00 -1.14546083e-01
-8.67850721e-01 8.11688229e-02 6.09897912e-01 7.18835175e-01
-8.44158828e-01 1.06899872e-01 -5.32148361e-01 -3.93891960e-01
7.87037984e-02 1.25686809e-01 -2.62416601e-01 -2.75755733e-01
-1.36728454e+00 7.54756927e-01 7.76096165e-01 -1.35193691e-01
-7.19650865e-01 -8.11652422e-01 -1.09098101e+00 3.63055676e-01
7.07218498e-02 -4.77368563e-01 5.67706645e-01 -3.48716617e-01
-1.19563711e+00 6.12167418e-01 -5.02442956e-01 3.60703543e-02
-2.03723192e-01 -3.52929324e-01 -5.36808550e-01 -5.54122403e-02
1.63287908e-01 3.47303122e-01 1.16935647e+00 -1.23515117e+00
-3.58695209e-01 -5.32027245e-01 -2.56773323e-01 1.97066367e-01
-1.27667952e+00 -3.61232400e-01 -2.20660180e-01 -7.19153821e-01
2.74366528e-01 -4.36984122e-01 2.66342074e-01 -1.65506408e-01
-1.78526446e-01 -1.06751764e+00 1.47438109e+00 -5.44017851e-01
1.50520289e+00 -2.34490609e+00 3.19716901e-01 1.43045276e-01
3.63076299e-01 1.74034461e-02 -2.09394008e-01 7.68248200e-01
-2.45984480e-01 3.24925750e-01 -1.29848141e-02 -5.55846930e-01
3.53711098e-01 3.11348051e-01 -7.98362792e-01 3.74966800e-01
1.28123015e-01 7.20637202e-01 -9.73596036e-01 -6.83791935e-01
2.15497822e-01 6.23191774e-01 -6.95933104e-01 1.64770529e-01
-1.89433530e-01 -3.57264996e-01 -6.18771017e-01 4.17600751e-01
6.60878181e-01 -7.37906918e-02 2.84404755e-01 -2.22881630e-01
1.27433628e-01 3.79964948e-01 -1.08377469e+00 1.82513225e+00
-6.23925030e-01 6.54419422e-01 -4.01991755e-01 -1.37964284e+00
1.21334338e+00 5.30457437e-01 2.78448313e-01 -7.01350048e-02
-3.04730702e-02 -1.22226626e-02 -3.52795899e-01 -6.96225822e-01
8.92342865e-01 -2.89195240e-01 -1.58895120e-01 8.24188232e-01
5.39564729e-01 1.88980684e-01 -7.58769363e-02 3.14687729e-01
6.30271852e-01 -9.06095430e-02 1.82645768e-01 -5.24748087e-01
3.50060791e-01 -3.00043881e-01 3.86383474e-01 4.43319201e-01
1.99569687e-01 2.19990373e-01 4.90094364e-01 -2.13853449e-01
-7.84883261e-01 -1.10478377e+00 -4.68680173e-01 1.19675052e+00
2.15415090e-01 -8.36167395e-01 -2.61965871e-01 -7.19825029e-01
1.89669609e-01 9.78295505e-01 -8.57849717e-01 -5.95071495e-01
1.85638946e-02 -8.64253044e-01 1.63672388e-01 6.35393322e-01
-1.20356068e-01 -7.41090000e-01 -1.10629626e-01 2.07524210e-01
-9.10874754e-02 -6.67442262e-01 -4.34550405e-01 1.63717553e-01
-1.06429660e+00 -6.07100606e-01 -7.45871365e-01 -8.78200471e-01
7.46770561e-01 4.13503379e-01 7.98688054e-01 9.74089373e-03
-2.45794848e-01 5.80418825e-01 -5.71000755e-01 -2.53843099e-01
1.91872463e-01 -1.81888491e-02 1.66199744e-01 1.71336159e-01
1.10415351e+00 -6.34645879e-01 -4.89716083e-01 -2.65872758e-02
-1.13270712e+00 -2.66106725e-01 3.09512228e-01 1.15783596e+00
2.38961890e-01 2.83599406e-01 6.68841898e-01 -8.69075656e-01
1.04479837e+00 -8.60564470e-01 -6.10721372e-02 3.10685843e-01
-9.20953691e-01 2.61511475e-01 1.73814520e-01 -5.44093013e-01
-1.02011442e+00 -9.25220728e-01 3.94009560e-01 -4.46436077e-01
-2.86454102e-03 7.31281340e-01 -8.61496255e-02 6.45711660e-01
2.18565837e-01 5.72013378e-01 -1.65650234e-01 -6.28670990e-01
5.03811300e-01 9.12200034e-01 -4.79839444e-02 -8.32011104e-01
7.81570017e-01 4.28997815e-01 -5.29648721e-01 -9.55267489e-01
-8.79121184e-01 -6.88913584e-01 -5.25358200e-01 1.22190960e-01
1.00928688e+00 -8.12524259e-01 -3.72758776e-01 8.90399795e-03
-1.37286258e+00 5.15949965e-01 -1.80661321e-01 7.13714957e-01
-1.63549315e-02 4.71611530e-01 -2.56623298e-01 -7.34965146e-01
-2.15579242e-01 -8.62854362e-01 1.16136098e+00 -1.69531584e-01
-3.91724944e-01 -1.75603616e+00 4.04902488e-01 1.06252246e-01
3.20385337e-01 -4.58299182e-02 1.46346569e+00 -9.37941790e-01
-1.29734710e-01 -3.64798397e-01 -2.63911963e-01 5.21329463e-01
4.09721822e-01 -2.24289093e-02 -1.10245132e+00 -5.51345348e-01
2.15562865e-01 -1.67114004e-01 1.20769346e+00 5.34632504e-02
1.20829833e+00 -2.76650190e-01 -5.59794784e-01 2.50724852e-01
1.59834576e+00 6.86581582e-02 3.88318121e-01 1.12796783e-01
7.38445163e-01 7.74610877e-01 4.19506848e-01 4.71667081e-01
3.49981844e-01 4.06511247e-01 3.88748497e-02 3.16833854e-01
2.45315805e-01 -3.63212258e-01 4.76792097e-01 1.49064791e+00
2.51749575e-01 -3.21858257e-01 -6.89143181e-01 9.55184937e-01
-1.61398268e+00 -8.22576821e-01 2.13232771e-01 1.92858505e+00
8.43959689e-01 5.52961193e-02 -4.86966848e-01 1.74695566e-01
7.35183239e-01 6.38971925e-01 -2.07965061e-01 -6.22524798e-01
2.36121580e-01 2.49986097e-01 -2.28371724e-01 4.81522113e-01
-7.54665136e-01 8.80470335e-01 5.67072010e+00 9.04405236e-01
-8.83821487e-01 4.28958058e-01 -7.90242627e-02 -2.03597695e-02
-1.02136803e+00 6.33986071e-02 -8.65214765e-01 5.29393137e-01
7.08875418e-01 -5.70402086e-01 -2.11319789e-01 1.00048304e+00
-1.63057283e-01 2.53996372e-01 -1.26400578e+00 9.19179201e-01
6.98837757e-01 -1.11708784e+00 6.88581824e-01 2.41607219e-01
6.93284214e-01 -5.82551181e-01 4.24983710e-01 5.09249687e-01
1.72534473e-02 -9.27685738e-01 2.90638685e-01 5.75078189e-01
4.62903589e-01 -8.15955222e-01 7.70742297e-01 1.38846010e-01
-1.15928149e+00 -5.97905554e-02 -8.62447619e-01 1.51090369e-01
-1.00541294e-01 7.62581468e-01 -6.22831702e-01 5.12469828e-01
4.06178415e-01 1.06170881e+00 -5.62674761e-01 5.39278090e-01
-1.65442869e-01 7.01408565e-01 1.55655324e-01 -2.40312442e-01
4.05996978e-01 -3.37127984e-01 6.56312466e-01 1.23309374e+00
4.87336606e-01 -2.87688375e-01 1.45929381e-01 1.09269631e+00
2.35028919e-02 2.12122977e-01 -8.13805521e-01 -3.51032764e-01
6.07268631e-01 1.16174448e+00 -4.42710310e-01 -4.44041222e-01
-3.99739414e-01 9.22343433e-01 2.41387561e-01 5.25965154e-01
-6.75015688e-01 -5.39238036e-01 9.47296798e-01 -2.59839952e-01
4.26318645e-01 -3.86989564e-01 -2.37473566e-02 -1.30593538e+00
1.09575965e-01 -2.40726054e-01 3.00691545e-01 -5.01212180e-01
-1.47597587e+00 3.90425891e-01 2.94228941e-01 -1.09032977e+00
-1.15238942e-01 -7.33836889e-01 -9.08237278e-01 7.91219294e-01
-1.55215335e+00 -9.17889714e-01 -7.59728625e-02 4.87772852e-01
8.57877672e-01 -2.99476653e-01 1.08368063e+00 1.59404218e-01
-3.81350666e-01 5.62232494e-01 3.45232695e-01 -1.10140957e-01
8.34620357e-01 -1.37509096e+00 -9.90326032e-02 2.70263910e-01
3.74206245e-01 1.02542472e+00 5.47441185e-01 -5.09395182e-01
-1.31679213e+00 -1.02402306e+00 1.27623081e+00 -6.46842480e-01
6.82769716e-01 -5.18860519e-01 -1.13177609e+00 7.11685240e-01
4.11275208e-01 -3.67140830e-01 1.29871285e+00 5.87597489e-01
-6.08881950e-01 -7.79513344e-02 -7.47492671e-01 4.96672094e-01
5.15739262e-01 -8.84367883e-01 -1.32463753e+00 3.44446391e-01
1.28972054e+00 4.49260622e-01 -9.07988489e-01 -1.05120651e-01
4.89909887e-01 -4.51927155e-01 8.51496100e-01 -6.96167469e-01
4.24497038e-01 -2.03010187e-01 -4.28795040e-01 -1.41718054e+00
-3.65773469e-01 1.07498635e-02 -3.69652987e-01 1.53483081e+00
2.83836037e-01 -7.48758674e-01 4.33471113e-01 1.82863414e-01
8.83108750e-03 -7.66183197e-01 -7.24574983e-01 -7.73104787e-01
1.32086441e-01 -6.14079654e-01 5.71674109e-01 1.39879251e+00
3.29346746e-01 4.81130153e-01 -1.74933523e-01 2.19749779e-01
7.91399837e-01 1.09388642e-01 3.43471766e-01 -1.57326674e+00
2.57588439e-02 -3.04687530e-01 -7.82364726e-01 -8.54829848e-01
5.33530593e-01 -1.23495746e+00 -3.11500788e-01 -1.49773705e+00
4.39058036e-01 -7.32425079e-02 -6.95439041e-01 2.33301073e-01
-3.06492299e-01 -8.87845755e-02 -2.38373548e-01 1.32300571e-01
-5.38162291e-01 9.81492639e-01 9.42364514e-01 -3.82588893e-01
9.14775729e-02 -5.28097034e-01 -7.84046233e-01 6.46105051e-01
6.09509289e-01 -7.19494522e-01 -8.23630571e-01 -6.26220465e-01
1.71116039e-01 -4.10642296e-01 1.56095922e-01 -6.29809618e-01
2.64233500e-01 -1.92291159e-02 3.43279690e-01 -6.87487066e-01
4.64453667e-01 -9.64632511e-01 -4.01122540e-01 1.90831631e-01
-4.73840415e-01 4.41739075e-02 8.02393332e-02 1.08717632e+00
-5.79348743e-01 -5.78325748e-01 2.60178566e-01 9.53756571e-02
-6.53155327e-01 2.13988677e-01 -4.24866974e-01 3.95115279e-03
8.42403352e-01 -3.40457469e-01 -1.56877458e-01 3.28903757e-02
-7.85635293e-01 2.49068767e-01 5.42021990e-02 9.16549861e-01
9.67271864e-01 -1.66686118e+00 -5.93707204e-01 3.50522310e-01
4.65844691e-01 -2.92335927e-01 3.65777850e-01 5.04819691e-01
1.08102441e-01 6.38927042e-01 1.91363335e-01 -6.48610294e-01
-1.17592049e+00 6.23399913e-01 -1.89168423e-01 -2.40633786e-01
-5.56225061e-01 7.76722372e-01 4.37547356e-01 -3.96860480e-01
2.90161699e-01 -2.96658963e-01 -4.54921424e-01 5.98174512e-01
5.34000516e-01 9.94800329e-02 -1.42515495e-01 -3.41310769e-01
-2.33572662e-01 6.50494218e-01 -5.37900746e-01 -1.80796728e-01
1.29140019e+00 -1.54829547e-01 -4.94482875e-01 1.02941716e+00
1.67234778e+00 -2.51705617e-01 -5.03960907e-01 -5.93735516e-01
2.06036523e-01 -7.01620460e-01 3.62289995e-01 -2.71369964e-01
-6.76583230e-01 1.27690589e+00 5.47074378e-01 3.70316982e-01
5.38718045e-01 1.34317368e-01 7.54403174e-01 5.58054090e-01
-5.13581969e-02 -1.22154510e+00 4.25162494e-01 2.76405632e-01
7.46524811e-01 -9.33904529e-01 1.11641036e-02 -1.40156865e-01
-4.08825904e-01 1.26621985e+00 5.55777311e-01 -3.83929797e-02
1.02576554e+00 -2.67534614e-01 -3.14965189e-01 -5.40642083e-01
-1.09917235e+00 2.05210894e-01 5.23678064e-01 4.82010067e-01
5.04650295e-01 1.00410998e-01 -5.49987733e-01 8.25487852e-01
-1.16448790e-01 -5.17521262e-01 2.12840483e-01 8.27598095e-01
-7.20034242e-01 -1.16561580e+00 -2.26102516e-01 3.33738625e-01
-2.09920168e-01 -2.02058598e-01 -1.61916837e-01 5.51983237e-01
1.56100199e-01 6.82807326e-01 4.48960334e-01 -3.40998501e-01
-6.63404912e-02 6.53651059e-01 1.88142955e-01 -1.03535628e+00
-3.50418650e-02 -1.22004330e-01 -3.42650473e-01 -1.57487527e-01
-2.95894384e-01 -4.77646440e-01 -8.57144892e-01 -1.05440700e-02
-6.39713109e-01 7.23443627e-01 8.74876142e-01 1.08306766e+00
3.24432731e-01 6.99545860e-01 6.15131676e-01 -5.70285261e-01
-4.78850812e-01 -1.21889973e+00 -8.90396774e-01 3.79462630e-01
2.08088934e-01 -9.31484342e-01 -6.57516122e-01 -4.56988961e-02] | [10.442865371704102, 7.107545375823975] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.