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
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
efc2639e-f0b6-4283-9758-7cf646cd3a28 | unsupervised-extractive-summarization-via | null | null | https://aclanthology.org/P15-2138 | https://aclanthology.org/P15-2138.pdf | Unsupervised extractive summarization via coverage maximization with syntactic and semantic concepts | null | ['Anders S{\\o}gaard', 'Natalie Schluter'] | 2015-07-01 | unsupervised-extractive-summarization-via-1 | https://aclanthology.org/P15-2138 | https://aclanthology.org/P15-2138.pdf | ijcnlp-2015-7 | ['unsupervised-extractive-summarization'] | ['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] | [-7.371952056884766, 3.6428489685058594] |
0f07e9ad-33af-4487-b4da-6183298032cb | learning-to-abstract-for-memory-augmented | null | null | https://aclanthology.org/P19-1371 | https://aclanthology.org/P19-1371.pdf | Learning to Abstract for Memory-augmented Conversational Response Generation | Neural generative models for open-domain chit-chat conversations have become an active area of research in recent years. A critical issue with most existing generative models is that the generated responses lack informativeness and diversity. A few researchers attempt to leverage the results of retrieval models to strengthen the generative models, but these models are limited by the quality of the retrieval results. In this work, we propose a memory-augmented generative model, which learns to abstract from the training corpus and saves the useful information to the memory to assist the response generation. Our model clusters query-response samples, extracts characteristics of each cluster, and learns to utilize these characteristics for response generation. Experimental results show that our model outperforms other competitive baselines. | ['Zhiliang Tian', 'Nevin L. Zhang', 'Xiaopeng Li', 'Wei Bi'] | 2019-07-01 | null | null | null | acl-2019-7 | ['conversational-response-generation'] | ['natural-language-processing'] | [-1.52085349e-02 5.89902773e-02 -1.95332989e-01 -6.08087480e-01
-1.50227737e+00 -4.53185260e-01 6.62669480e-01 -4.32918549e-01
-8.75654817e-02 9.51991439e-01 6.89043820e-01 1.06110245e-01
1.07643768e-01 -8.93211484e-01 -2.29721859e-01 -5.58062911e-01
3.32359463e-01 1.06149089e+00 2.43865788e-01 -4.69761878e-01
3.03556800e-01 -2.79188186e-01 -1.13591206e+00 9.23611879e-01
1.03314173e+00 7.96073794e-01 2.97789067e-01 7.81619668e-01
-4.36001569e-01 1.06012475e+00 -1.04678440e+00 -4.51866269e-01
-3.16680551e-01 -8.81858170e-01 -1.16891778e+00 -2.75345594e-01
-7.75858611e-02 -6.76717281e-01 -6.16142273e-01 6.54303432e-01
7.23619640e-01 3.64343494e-01 7.47273505e-01 -9.64907944e-01
-1.16566372e+00 1.06913102e+00 9.83684510e-02 2.32197180e-01
4.82901841e-01 -5.33809029e-02 1.09119511e+00 -1.19492197e+00
7.94709504e-01 1.58291459e+00 9.95071456e-02 9.95809078e-01
-9.88071501e-01 -7.06738651e-01 1.52103081e-01 1.91841513e-01
-1.30014563e+00 -4.66156840e-01 7.79297531e-01 -4.75688577e-02
1.02428222e+00 4.03563678e-01 3.95180672e-01 1.56569552e+00
-1.47417128e-01 1.30649555e+00 8.16879451e-01 -2.54446268e-01
1.01086602e-01 2.11243019e-01 3.11236173e-01 1.58037618e-01
-2.77567118e-01 -3.05376977e-01 -6.58694625e-01 -3.04378092e-01
4.96455371e-01 3.33043396e-01 -2.70101607e-01 3.06150556e-01
-5.78028202e-01 1.34256864e+00 3.74448329e-01 2.42336854e-01
-3.82511258e-01 -4.05979753e-02 7.34565407e-02 4.28308964e-01
1.10679436e+00 4.62265849e-01 1.61333755e-01 -5.02491355e-01
-8.67529750e-01 5.34869492e-01 1.18977416e+00 1.28432047e+00
7.36496270e-01 -1.03132129e-01 -5.84881902e-01 1.22456622e+00
3.03767890e-01 5.95035791e-01 6.54363453e-01 -6.78398252e-01
7.48827279e-01 7.86661983e-01 -1.27318993e-01 -6.60285234e-01
5.63510135e-02 -2.58066803e-01 -6.26576781e-01 -8.75446677e-01
-3.89525779e-02 -4.11245048e-01 -6.66541457e-01 1.49200475e+00
-4.23309579e-03 -2.94266433e-01 1.44147754e-01 1.01710045e+00
1.26454866e+00 1.02576268e+00 -8.43711644e-02 -1.23310313e-01
7.01810420e-01 -1.20011818e+00 -8.41700613e-01 -4.10476953e-01
4.42777723e-01 -9.22322750e-01 1.04983222e+00 -8.76117274e-02
-1.26395714e+00 -3.58383566e-01 -4.62337911e-01 -5.61775118e-02
7.53814504e-02 -9.31460336e-02 5.60766160e-01 3.05336893e-01
-1.00695252e+00 1.15820691e-01 -5.80280840e-01 -2.77783632e-01
1.79867819e-01 1.12376876e-01 2.68449754e-01 -2.95957446e-01
-1.50068104e+00 5.05848169e-01 3.47597115e-02 -2.51783907e-01
-9.34310257e-01 -5.28492987e-01 -4.89869207e-01 3.27837557e-01
3.92697155e-01 -8.14349234e-01 1.86773038e+00 -5.78480840e-01
-1.61440909e+00 3.47490013e-01 -4.01095867e-01 -2.90492088e-01
3.66233170e-01 -3.30351323e-01 -2.39562407e-01 2.49595165e-01
-1.02589726e-01 7.94754267e-01 5.12530625e-01 -1.08337796e+00
-4.12979275e-01 -1.47758096e-01 5.09880334e-02 2.79835910e-01
-5.88616788e-01 1.09338216e-01 -6.89828217e-01 -5.98299444e-01
-1.20756999e-01 -1.12422085e+00 -1.51708409e-01 -1.01591706e+00
-4.33542967e-01 -8.35706294e-01 6.50958359e-01 -4.23972070e-01
1.62490439e+00 -1.88217664e+00 2.46703979e-02 -9.42886341e-03
2.24654421e-01 4.69044149e-02 -2.71554351e-01 1.12539065e+00
5.64119637e-01 2.95611709e-01 2.91269958e-01 -5.40086389e-01
1.94643661e-01 1.10662326e-01 -9.30290937e-01 -3.36517543e-01
1.56749725e-01 1.22283411e+00 -7.89281964e-01 -4.96577322e-01
-3.92834306e-01 3.55425596e-01 -8.08973253e-01 8.30724299e-01
-5.47530591e-01 2.40016699e-01 -1.03872204e+00 3.70793074e-01
2.76119232e-01 -5.86170852e-01 2.99989820e-01 4.01989549e-01
3.48675638e-01 1.08647668e+00 -5.55806518e-01 1.49587452e+00
-3.58308613e-01 4.95960236e-01 -2.96811104e-01 -5.11406779e-01
9.88553405e-01 4.29335594e-01 2.15099350e-01 -7.36776888e-01
8.29516128e-02 9.08661783e-02 -1.00786634e-01 -5.69849014e-01
9.86758292e-01 7.88016021e-02 -3.28258902e-01 1.18412173e+00
6.90409392e-02 -1.99831769e-01 2.30835125e-01 7.13815629e-01
9.48198199e-01 -2.88828015e-01 -1.83481917e-01 2.86335230e-01
-1.86318196e-02 2.13854127e-02 2.87829369e-01 9.88586187e-01
3.86397779e-01 7.28519678e-01 1.90203160e-01 -1.86139196e-01
-7.03629076e-01 -8.16760063e-01 4.32344139e-01 1.45343792e+00
-9.44127515e-02 -5.38663924e-01 -8.87094676e-01 -6.49329603e-01
-3.54037255e-01 5.82175493e-01 -4.45298910e-01 -5.36309242e-01
-7.59156644e-01 -5.21979928e-01 5.52643001e-01 8.10116529e-01
3.09816539e-01 -1.25865400e+00 -8.87321457e-02 3.11583549e-01
-1.05077827e+00 -7.97315598e-01 -8.48611295e-01 -9.35941190e-02
-9.24791753e-01 -5.10871470e-01 -6.61327004e-01 -7.43652761e-01
4.81839627e-01 6.78422511e-01 1.46973205e+00 1.73683643e-01
1.99670851e-01 4.07740712e-01 -7.75981724e-01 -3.42422336e-01
-7.60167480e-01 6.89333856e-01 -2.98156679e-01 -1.69247165e-01
6.47683144e-01 -4.16491151e-01 -5.90218246e-01 3.57968986e-01
-1.00282025e+00 -7.30415732e-02 7.03599513e-01 9.59879637e-01
3.87464225e-01 -5.39304674e-01 1.07005525e+00 -1.19560874e+00
1.38718164e+00 -8.55235040e-01 -3.17910756e-03 3.51914614e-01
-7.51723766e-01 6.68633580e-02 3.89734864e-01 -7.38803864e-01
-1.35860848e+00 -5.74536562e-01 -1.17245987e-01 -3.61593843e-01
4.16794494e-02 8.51442814e-01 -7.00219423e-02 4.54113841e-01
4.74296510e-01 4.66470689e-01 -2.26930246e-01 -6.76000416e-01
3.54267806e-01 1.05622101e+00 1.73782513e-01 -6.39512479e-01
3.53097916e-01 3.21203731e-02 -8.78432810e-01 -5.22425532e-01
-8.57338727e-01 -5.33026993e-01 4.29647639e-02 -2.57683128e-01
4.70943868e-01 -8.49080145e-01 -3.22501868e-01 2.00734198e-01
-1.31599879e+00 -2.22079858e-01 -7.83324391e-02 4.07832980e-01
-3.34074706e-01 8.73352587e-02 -1.17315507e+00 -1.03107047e+00
-7.45278299e-01 -9.26683724e-01 1.00194728e+00 4.04524803e-01
-5.67529321e-01 -9.84318972e-01 3.69918704e-01 6.39165878e-01
7.67360806e-01 -5.81864893e-01 8.03486288e-01 -1.08813894e+00
-1.05098832e+00 -4.86440152e-01 9.51150581e-02 3.50907184e-02
-1.74142152e-01 -2.46858418e-01 -1.14572287e+00 -1.49536595e-01
1.29046351e-01 -9.46681201e-01 1.10494864e+00 1.87126368e-01
1.00984490e+00 -7.85167754e-01 -3.79709005e-01 2.07640320e-01
7.54858851e-01 4.15186912e-01 7.28486180e-01 -3.68716940e-02
2.60882050e-01 6.19906902e-01 4.93641675e-01 5.51930010e-01
6.06888652e-01 6.14620507e-01 9.01914909e-02 1.52542695e-01
9.60808073e-04 -7.13975132e-01 3.83392185e-01 1.38493490e+00
5.19983210e-02 -7.10644960e-01 -3.92825097e-01 5.07152200e-01
-1.90169704e+00 -1.25241804e+00 1.02652900e-01 1.71543777e+00
1.08560765e+00 -6.76793903e-02 2.05736071e-01 -5.10320663e-01
4.36050504e-01 2.19156429e-01 -2.77057201e-01 -1.88432142e-01
-1.00946017e-01 3.51403117e-01 -3.07514578e-01 3.69292498e-01
-5.76096952e-01 1.00696194e+00 7.06923914e+00 8.28335166e-01
-9.45828259e-01 1.31865889e-01 6.20488286e-01 -3.89543295e-01
-6.26075506e-01 -3.51135456e-03 -1.09605598e+00 4.86310929e-01
1.05890596e+00 -4.94641811e-01 1.88366383e-01 1.03103328e+00
2.59654745e-02 1.43394947e-01 -1.22659945e+00 6.25588536e-01
4.54757899e-01 -1.30086970e+00 3.95231992e-01 1.11810945e-01
8.24420452e-01 -6.54398501e-02 3.69133472e-01 8.95717978e-01
8.69737267e-01 -9.92564499e-01 2.50625968e-01 6.81954920e-01
4.22272176e-01 -6.29452705e-01 7.77750254e-01 5.49000859e-01
-6.65756464e-01 7.01279566e-02 -4.54905629e-01 -1.55729160e-01
2.27365047e-01 1.45256877e-01 -1.35777116e+00 3.10711563e-01
4.88469332e-01 3.17924827e-01 -4.34668422e-01 8.28464150e-01
-2.78102905e-01 1.05750680e+00 -5.39713018e-02 -7.00958014e-01
3.38891625e-01 -7.17537552e-02 3.81878674e-01 1.27634799e+00
3.77124995e-01 2.69258022e-01 3.90391946e-01 1.29061007e+00
-4.53985959e-01 1.16668299e-01 -5.04031122e-01 -4.21897352e-01
9.14995611e-01 1.10332894e+00 -2.10668489e-01 -4.38491702e-01
-3.11174482e-01 7.38751352e-01 4.38705385e-01 5.69972813e-01
-5.73734939e-01 -2.88555473e-01 4.46319997e-01 -9.91261844e-03
2.69002497e-01 4.27310579e-02 8.58974308e-02 -1.18618762e+00
2.21773367e-02 -9.92336094e-01 6.53107584e-01 -7.80479670e-01
-1.54575217e+00 7.48441100e-01 8.21570680e-02 -1.12953627e+00
-1.31366718e+00 2.08016306e-01 -6.87229276e-01 1.10096514e+00
-1.10150731e+00 -1.09027433e+00 -2.30433971e-01 5.05141079e-01
1.17172074e+00 -2.03117892e-01 1.00124288e+00 2.42944956e-01
-3.93108785e-01 8.67706716e-01 1.32308647e-01 4.17641342e-01
8.87194693e-01 -8.94110680e-01 3.61920387e-01 5.51976264e-01
2.83907950e-01 1.23596108e+00 4.11203593e-01 -6.70099020e-01
-1.30138278e+00 -9.82933760e-01 1.35115099e+00 -6.56856477e-01
3.39729279e-01 -3.81277591e-01 -1.10316825e+00 6.94930315e-01
5.76363027e-01 -6.73363447e-01 1.12523437e+00 3.14610928e-01
-3.15078795e-01 8.37089047e-02 -3.94910514e-01 5.37665367e-01
7.04779744e-01 -6.71605647e-01 -7.23786294e-01 2.61887163e-01
9.67854142e-01 -4.08199102e-01 -5.21873057e-01 -1.42485812e-01
4.48907822e-01 -6.65185511e-01 7.30658412e-01 -8.46554458e-01
6.22797608e-01 4.13983703e-01 1.25238681e-02 -1.59404969e+00
-3.69390965e-01 -7.58528531e-01 -2.48166144e-01 1.46370864e+00
5.07072151e-01 -2.82404065e-01 7.16414332e-01 8.80910695e-01
-4.59840328e-01 -6.35275722e-01 -5.41522264e-01 -5.41752875e-01
2.87745595e-01 -1.60677433e-01 7.25072026e-01 6.86415672e-01
3.71108353e-01 9.97019410e-01 -6.11736953e-01 -5.60327947e-01
7.21893162e-02 6.32963538e-01 1.07150769e+00 -1.14995408e+00
-5.03605485e-01 -5.15978992e-01 3.92663985e-01 -1.70291567e+00
1.04740754e-01 -7.33893514e-01 1.60558313e-01 -1.63553500e+00
6.92075074e-01 -3.93055975e-01 6.03245907e-02 2.49481827e-01
-5.38864315e-01 6.59082038e-03 8.55264366e-02 4.87961859e-01
-1.06250036e+00 7.61851192e-01 1.16608393e+00 -1.67271033e-01
-3.56792718e-01 3.84210318e-01 -1.00095987e+00 2.00776368e-01
5.91980696e-01 -6.03500605e-01 -7.11061001e-01 -6.00708544e-01
2.40768805e-01 3.78824741e-01 -2.58387737e-02 -4.35594648e-01
4.25062954e-01 -2.56605893e-01 1.35379955e-01 -8.15118134e-01
5.31019807e-01 -2.17159510e-01 9.87614621e-04 -1.15907239e-02
-9.15242314e-01 8.85420367e-02 -3.25255334e-01 5.47742128e-01
-5.39710879e-01 -1.57893777e-01 2.13089153e-01 -2.33573735e-01
-2.71188915e-01 3.06751579e-01 -5.26660264e-01 4.95817631e-01
2.30189413e-01 1.43361941e-01 -6.80150092e-01 -1.22221792e+00
-2.45794356e-01 3.40677500e-01 2.01701745e-02 9.06866312e-01
9.28067267e-01 -1.45307541e+00 -8.99750888e-01 1.45962564e-02
3.02115232e-01 1.46533791e-02 4.93357092e-01 5.23584366e-01
1.78416222e-01 8.36554646e-01 1.79149523e-01 -4.04725969e-01
-1.28111994e+00 3.55916172e-01 -4.92089167e-02 -6.02962434e-01
-3.84636253e-01 9.68935311e-01 3.83696258e-01 -1.34271860e-01
2.84699827e-01 -6.46305382e-02 -3.34003657e-01 -4.53555910e-03
8.62986863e-01 2.16613919e-01 5.72461672e-02 -2.04127580e-01
4.68370132e-02 -2.69650638e-01 -6.07796788e-01 -3.85541350e-01
1.44927311e+00 -2.12582022e-01 4.63379547e-02 3.55690062e-01
1.18648911e+00 -4.81858663e-02 -8.97428036e-01 -7.72887588e-01
-1.16470233e-01 -5.17194450e-01 -3.57761711e-01 -7.61923194e-01
-9.67861235e-01 1.06636405e+00 -1.34253994e-01 3.15789223e-01
8.46041441e-01 3.54942530e-01 1.22521830e+00 5.81463456e-01
2.27445319e-01 -1.25535858e+00 7.73382545e-01 1.00997925e+00
8.25766027e-01 -1.07924461e+00 -5.82905889e-01 -2.87175775e-01
-8.61381948e-01 9.27227259e-01 9.49266016e-01 1.08891144e-01
1.61411881e-01 2.44665638e-01 2.41346329e-01 -1.65311337e-01
-1.40154302e+00 -3.16424854e-02 4.66085941e-01 5.42162955e-01
8.41967046e-01 -7.16794282e-02 -3.44057888e-01 1.09679675e+00
-3.00688535e-01 -2.09661752e-01 2.81876355e-01 9.01659012e-01
-6.66674137e-01 -1.28213799e+00 -5.79341389e-02 4.41081673e-01
-3.89458686e-01 -3.34147215e-01 -1.19850576e+00 4.07311052e-01
-6.76452219e-01 1.35590327e+00 -8.74632299e-02 -7.60123372e-01
1.40829489e-01 5.49098372e-01 9.85709056e-02 -8.79224598e-01
-7.12376416e-01 5.77080846e-01 2.93377757e-01 -1.72814220e-01
-1.06420323e-01 -5.01413941e-01 -9.53994513e-01 -4.22390580e-01
-5.00908077e-01 7.85934627e-01 2.05263659e-01 8.29994202e-01
5.25375962e-01 4.96759087e-01 7.73085535e-01 -2.85128444e-01
-1.01393771e+00 -1.57532382e+00 -1.83565855e-01 4.38476443e-01
-1.04510143e-01 -3.53054434e-01 -1.50909796e-01 -2.98005324e-02] | [12.51496410369873, 8.307197570800781] |
4f1c91ed-3e03-4fb4-a867-3858ad5b93d8 | mitosis-detection-from-partial-annotation-by | 2307.04113 | null | https://arxiv.org/abs/2307.04113v1 | https://arxiv.org/pdf/2307.04113v1.pdf | Mitosis Detection from Partial Annotation by Dataset Generation via Frame-Order Flipping | Detection of mitosis events plays an important role in biomedical research. Deep-learning-based mitosis detection methods have achieved outstanding performance with a certain amount of labeled data. However, these methods require annotations for each imaging condition. Collecting labeled data involves time-consuming human labor. In this paper, we propose a mitosis detection method that can be trained with partially annotated sequences. The base idea is to generate a fully labeled dataset from the partial labels and train a mitosis detection model with the generated dataset. First, we generate an image pair not containing mitosis events by frame-order flipping. Then, we paste mitosis events to the image pair by alpha-blending pasting and generate a fully labeled dataset. We demonstrate the performance of our method on four datasets, and we confirm that our method outperforms other comparisons which use partially labeled sequences. | ['Ryoma Bise', 'Shinichiro Chuma', 'Ami Katanaya', 'Kazuya Nishimura'] | 2023-07-09 | null | null | null | null | ['mitosis-detection'] | ['medical'] | [ 5.12730420e-01 4.13478352e-02 -2.99827904e-01 -4.37634498e-01
-9.82633770e-01 -5.96113443e-01 4.41670984e-01 2.54428416e-01
-7.15683699e-01 1.01477575e+00 -3.38494740e-02 -3.73560451e-02
6.93716109e-01 -7.18583524e-01 -6.98402822e-01 -1.12399769e+00
3.67246985e-01 6.05651200e-01 5.28021514e-01 4.34670717e-01
2.07817450e-01 3.16564947e-01 -1.12367058e+00 3.86486202e-01
6.11032963e-01 4.78079557e-01 2.59977609e-01 9.23032999e-01
-5.66340536e-02 7.02557206e-01 -6.28481030e-01 -2.21498773e-01
1.26310587e-01 -7.58383393e-01 -7.91292310e-01 3.80254060e-01
-1.49218254e-02 -4.47135925e-01 -3.20213139e-01 8.80222917e-01
6.07658565e-01 -3.58014643e-01 7.56692052e-01 -1.19183600e+00
4.87603806e-02 4.20343697e-01 -7.64637113e-01 3.64829570e-01
6.40313178e-02 7.56782964e-02 4.48116750e-01 -7.51480460e-01
9.48604941e-01 6.78323090e-01 4.47981417e-01 9.06090081e-01
-1.19523263e+00 -5.20107687e-01 -5.07537246e-01 -1.15851715e-01
-1.28599298e+00 -3.30867469e-01 5.38361311e-01 -4.89935368e-01
4.22367841e-01 5.30509278e-02 8.03035915e-01 9.75568116e-01
2.78049231e-01 9.92578089e-01 1.09563410e+00 -4.44084227e-01
3.34234297e-01 -3.45791906e-01 1.00601159e-01 7.70175874e-01
2.36056298e-01 -1.18143857e-01 -4.66137946e-01 1.31881446e-01
8.79204273e-01 1.59943387e-01 -2.52091080e-01 -2.05029938e-02
-1.80193841e+00 5.68162560e-01 -3.33555527e-02 3.99524808e-01
-7.02886954e-02 8.93881768e-02 5.54281533e-01 -1.22297287e-01
4.41955805e-01 1.86952129e-01 -3.46994787e-01 8.89930725e-02
-1.33087945e+00 2.40949812e-04 7.17782676e-01 8.10518742e-01
7.84295976e-01 -4.13695097e-01 -4.99003977e-01 4.36143696e-01
3.01677659e-02 2.28438810e-01 6.65061176e-01 -9.35370743e-01
-1.47719935e-01 6.47320032e-01 -4.88472208e-02 -5.72844207e-01
-5.60496688e-01 -1.93544239e-01 -1.09484017e+00 -1.86145268e-02
1.13281035e+00 -1.98418871e-01 -1.17215538e+00 1.52565050e+00
5.06313741e-01 3.63645822e-01 1.80748060e-01 8.76890063e-01
7.94369161e-01 5.97639263e-01 9.61519107e-02 -4.90000665e-01
1.40542269e+00 -1.08708727e+00 -9.37965989e-01 3.60040441e-02
8.28809738e-01 -8.37380111e-01 6.83807254e-01 2.41291940e-01
-8.44202638e-01 -3.79839063e-01 -1.06232333e+00 -4.05625492e-01
-2.92185843e-01 4.99788761e-01 3.66014302e-01 3.26521724e-01
-9.67392981e-01 3.33587259e-01 -1.16918457e+00 -5.61167479e-01
3.70348364e-01 4.10627276e-01 -2.97295988e-01 3.69451046e-02
-5.78428805e-01 4.59413916e-01 7.57053554e-01 -6.15277216e-02
-1.24369133e+00 -3.08959454e-01 -8.89110625e-01 -5.76937497e-02
2.50843704e-01 -7.28934944e-01 1.43771613e+00 -5.94973922e-01
-1.27852392e+00 1.39045215e+00 -5.33095300e-01 -3.28084111e-01
4.76408303e-01 2.59759396e-01 2.01443747e-01 2.93881387e-01
1.05400518e-01 1.30359435e+00 4.41683561e-01 -1.15895605e+00
-8.74490976e-01 -1.30357459e-01 -4.00679260e-01 -1.85569838e-01
-5.76741584e-02 -1.73316419e-01 -7.75675893e-01 -7.20947921e-01
2.34945536e-01 -9.70845938e-01 -3.37867051e-01 1.88687921e-01
-5.12065589e-01 3.23271635e-03 8.79093766e-01 -7.89472044e-01
8.40415001e-01 -2.12721729e+00 -3.77313676e-03 -2.11781070e-01
3.82008910e-01 2.12303430e-01 -4.44777049e-02 1.09814718e-01
2.19697922e-01 6.20895699e-02 -3.16070259e-01 -6.37957752e-01
-3.17939818e-01 2.26107210e-01 6.15608804e-02 4.94921982e-01
2.24119499e-01 1.16364825e+00 -1.07059479e+00 -1.12102640e+00
9.55064036e-03 2.76691288e-01 -2.94562459e-01 3.20613533e-01
-4.10443515e-01 7.34021246e-01 -1.08860554e-02 9.01727736e-01
5.14494181e-01 -5.65988898e-01 3.61209273e-01 -2.51412213e-01
6.71710521e-02 -1.59612596e-01 -7.45045006e-01 1.64967597e+00
2.28564575e-01 8.20758224e-01 -1.67359516e-01 -8.52238834e-01
8.93110693e-01 4.05059904e-01 5.75392306e-01 -1.46183640e-01
4.34744179e-01 3.10106516e-01 -1.60618931e-01 -5.65371871e-01
3.57892483e-01 -2.04756290e-01 -1.14748403e-01 5.29457569e-01
2.12615505e-01 -1.90628156e-01 5.27872622e-01 1.57994464e-01
1.26107717e+00 4.60452437e-02 3.34330797e-01 4.74554859e-02
4.93702680e-01 1.24602839e-01 9.90925133e-01 3.79331112e-01
-5.94089806e-01 8.45216513e-01 7.07663536e-01 -5.80610454e-01
-1.44537258e+00 -7.58872986e-01 8.16784948e-02 6.81982160e-01
1.18987240e-01 -2.15735629e-01 -1.06211245e+00 -9.08917367e-01
-4.62321013e-01 -1.87024605e-02 -7.10355282e-01 1.13028876e-01
-6.20139301e-01 -1.05907106e+00 6.83166623e-01 5.91626406e-01
6.81200862e-01 -9.53969061e-01 -6.17582560e-01 1.18063010e-01
-4.36089993e-01 -1.07730746e+00 -6.18222356e-01 3.35069299e-01
-7.92558134e-01 -1.20202148e+00 -1.09610665e+00 -1.43527973e+00
1.05639648e+00 2.51936406e-01 8.74386191e-01 2.75285631e-01
-4.14532870e-01 -3.23381215e-01 -1.91334307e-01 -2.36658663e-01
-4.91927266e-01 1.40549093e-01 -3.34173799e-01 -1.69037685e-01
4.70848948e-01 -9.57173258e-02 -6.30041540e-01 3.25952679e-01
-1.10732281e+00 3.54247302e-01 6.46847725e-01 1.06385899e+00
1.08828056e+00 2.88244858e-02 5.47480166e-01 -9.68073428e-01
1.73451621e-02 -1.40795633e-01 -6.45920217e-01 1.98694691e-01
2.21345834e-02 -7.09034503e-02 2.54008412e-01 -4.91827309e-01
-1.07546175e+00 7.78548539e-01 2.17749109e-03 -3.00534278e-01
-2.51631171e-01 4.70822006e-01 -2.34077722e-01 -1.08339852e-02
3.91453266e-01 2.59918958e-01 2.44603697e-02 -8.19500685e-02
-3.19223478e-02 5.99898577e-01 9.86017287e-01 -3.03025365e-01
3.56795996e-01 7.10576236e-01 7.42121935e-02 -6.17887795e-01
-8.75642955e-01 -5.05884528e-01 -7.32843518e-01 -2.62832791e-01
1.10321879e+00 -7.04724431e-01 -4.33936626e-01 9.51689839e-01
-1.27890134e+00 -6.49435520e-01 -2.03676328e-01 5.41010439e-01
-5.45854390e-01 4.86095965e-01 -1.20121944e+00 -4.09118414e-01
-2.72920132e-01 -1.01529133e+00 1.36896431e+00 5.01272678e-01
-8.61617774e-02 -8.44447255e-01 3.47210675e-01 4.72959518e-01
-2.05491990e-01 4.46259320e-01 7.08744645e-01 -7.75566578e-01
-8.23994517e-01 -1.51142776e-01 -2.17453316e-01 -2.29965895e-01
2.43657261e-01 2.93966860e-01 -1.03183353e+00 -2.55913228e-01
-1.96587458e-01 -3.61652344e-01 1.06618536e+00 5.68930268e-01
1.17544901e+00 1.49474472e-01 -9.01437163e-01 5.18292844e-01
1.21302462e+00 4.15108830e-01 8.18298638e-01 2.47348130e-01
3.19881320e-01 6.44211709e-01 6.84059083e-01 9.69144180e-02
2.18072906e-01 1.67972833e-01 1.14699997e-01 -4.65970159e-01
-2.71437496e-01 -2.55322814e-01 -6.58296719e-02 6.56279027e-01
3.22396815e-01 -6.72977269e-01 -9.26179826e-01 7.18208432e-01
-1.85124254e+00 -7.84582436e-01 -1.33410454e-01 1.92733669e+00
1.18278790e+00 1.66142359e-01 1.70244709e-01 4.70084697e-01
9.54802454e-01 -3.37408453e-01 -5.78895926e-01 2.60761499e-01
-7.38992840e-02 6.93782195e-02 3.05831552e-01 3.32517833e-01
-1.32063484e+00 9.66887891e-01 6.97755814e+00 7.26971090e-01
-1.13710725e+00 2.43178047e-02 1.19351542e+00 -1.33156264e-02
1.97871006e-03 -6.69378117e-02 -9.19062972e-01 5.89798570e-01
5.70228875e-01 -1.52840987e-01 -2.74491280e-01 3.72732699e-01
2.04670608e-01 -5.14349043e-01 -1.25756347e+00 9.13776040e-01
9.94803477e-03 -1.44612575e+00 -2.59554893e-01 2.97511667e-01
8.70653868e-01 -4.81982648e-01 -3.61259580e-01 1.20607182e-01
3.33727896e-01 -7.47621894e-01 1.77251831e-01 4.77142960e-01
7.57930338e-01 -5.05781710e-01 1.17938101e+00 5.77115476e-01
-9.92412627e-01 3.65466624e-01 -2.37224236e-01 5.36704436e-02
3.41417223e-01 1.03535485e+00 -1.19269872e+00 7.90739357e-02
2.98500836e-01 5.68770289e-01 -4.83293623e-01 1.52201366e+00
-2.76212990e-01 6.43258393e-01 -2.98344195e-01 1.60101359e-03
-2.54587591e-01 -5.38831800e-02 2.90327985e-02 1.37336600e+00
4.33995873e-01 1.95761725e-01 3.38145971e-01 6.37638271e-01
-3.46629769e-01 -1.82609797e-01 -4.84116018e-01 -2.03103229e-01
4.86930698e-01 1.52992165e+00 -1.60117650e+00 -6.10063374e-01
-2.64858872e-01 1.07029307e+00 2.34489694e-01 1.01808242e-01
-8.14068258e-01 -3.31858784e-01 1.32938355e-01 -9.17120725e-02
1.03466511e-01 -8.14724639e-02 -3.52281988e-01 -1.06208980e+00
-3.16167176e-01 -5.35555601e-01 4.57104266e-01 -7.14910626e-01
-9.18998599e-01 2.17063904e-01 -3.45594078e-01 -1.08959091e+00
-9.67189297e-02 -2.65391082e-01 -5.60551286e-01 3.10952783e-01
-1.20106292e+00 -1.06970453e+00 -5.07111609e-01 2.02090949e-01
6.87921107e-01 2.33219758e-01 5.92391253e-01 4.96612549e-01
-8.49032104e-01 5.57999253e-01 -1.92977443e-01 4.62443382e-01
9.07074809e-01 -1.27454305e+00 3.02841216e-01 8.78734648e-01
-1.01093985e-02 2.59691477e-01 5.45870543e-01 -8.80559802e-01
-7.99747884e-01 -1.15824533e+00 1.31042206e+00 -1.02508776e-01
1.66450039e-01 -3.67376804e-01 -8.40504706e-01 8.04371059e-01
3.93364906e-01 1.63410082e-01 8.54340315e-01 -7.61472046e-01
1.95177823e-01 1.48301765e-01 -1.30059719e+00 5.17146826e-01
7.08028078e-01 -3.23919922e-01 -3.04041117e-01 4.79791969e-01
6.88250422e-01 -6.90056145e-01 -6.38660848e-01 4.45394814e-01
3.10715914e-01 -6.43902183e-01 4.95316178e-01 -1.17034934e-01
5.49974322e-01 -8.18947256e-01 3.46344620e-01 -9.35498059e-01
-9.95626748e-02 -4.47296411e-01 2.53493078e-02 1.36065304e+00
4.18452084e-01 -1.46270275e-01 1.38237858e+00 2.41949931e-01
-1.82562456e-01 -7.14637280e-01 -7.48834074e-01 -4.03324783e-01
-5.18227294e-02 3.31423014e-01 3.18351567e-01 8.99053156e-01
2.62571901e-01 3.52028817e-01 -1.82853147e-01 6.60529407e-03
5.83308637e-01 4.14646491e-02 8.32271397e-01 -1.03534055e+00
-7.98226520e-02 4.46532853e-02 -3.56806487e-01 -1.01827621e+00
1.41312063e-01 -6.46690249e-01 4.28000361e-01 -1.40679085e+00
8.00845146e-01 -1.07529163e-01 -3.14060152e-02 7.17044234e-01
-4.16439950e-01 7.83636749e-01 -1.35302991e-01 3.21279109e-01
-7.03206778e-01 8.82760957e-02 1.41883445e+00 -2.54094571e-01
3.61603824e-03 -3.41482699e-01 -9.25828815e-02 7.94675469e-01
1.01108551e+00 -4.77636248e-01 -1.51076272e-01 -1.40936956e-01
5.09996712e-02 2.72229314e-01 1.87987968e-01 -1.19731855e+00
4.96752888e-01 -1.00230336e-01 6.99985445e-01 -9.60555494e-01
2.36382097e-01 -5.22597373e-01 2.06946045e-01 6.75343692e-01
-3.33508760e-01 -2.13548124e-01 -1.68880168e-02 5.13043046e-01
-2.29565516e-01 -2.68057555e-01 8.70203257e-01 -2.34862417e-01
-4.94476080e-01 1.58216819e-01 -7.00495720e-01 -7.66338110e-02
1.39317274e+00 -2.79802024e-01 -5.33647001e-01 -6.03494570e-02
-8.08359683e-01 2.85409749e-01 9.20824230e-01 -2.14213893e-01
4.63150769e-01 -1.24905157e+00 -4.30112928e-01 3.07448775e-01
9.39406920e-03 4.27322567e-01 -1.08653426e-01 7.22814322e-01
-8.43749821e-01 3.02449405e-01 -2.16496363e-01 -8.87057185e-01
-1.50961101e+00 7.59552181e-01 2.26187930e-01 -4.87391204e-01
-2.27625683e-01 7.80044854e-01 2.57697493e-01 -3.18669379e-01
2.37518013e-01 -4.85398948e-01 -1.37074769e-01 -3.72139178e-02
5.26813149e-01 2.53353506e-01 1.66914649e-02 -2.63078600e-01
-1.39417574e-01 4.41869050e-01 -2.12749183e-01 -1.43704087e-01
9.05784309e-01 1.91873480e-02 -1.68710932e-01 2.39234701e-01
1.05362344e+00 -3.01354468e-01 -1.49043798e+00 7.01007843e-02
9.85869691e-02 -2.62921393e-01 -2.39311725e-01 -5.99080801e-01
-1.31629813e+00 7.87328482e-01 5.29590428e-01 -1.90370336e-01
1.20508564e+00 9.31905881e-02 1.09855521e+00 2.20784307e-01
3.66366923e-01 -8.75030160e-01 4.40189332e-01 2.98149645e-01
2.18609218e-02 -1.26372993e+00 -2.04428554e-01 -6.46937370e-01
-2.63426602e-01 8.85720074e-01 7.84314811e-01 1.68854624e-01
2.96264648e-01 7.92431474e-01 3.18416119e-01 -9.43464413e-02
-8.10716093e-01 -2.45694920e-01 -4.11758959e-01 6.62183821e-01
5.80634773e-01 -1.66432112e-01 -8.66092861e-01 3.30087811e-01
1.18196204e-01 6.21615589e-01 6.98076904e-01 1.18508804e+00
-5.39239049e-01 -1.31160486e+00 -3.75295848e-01 3.86546075e-01
-6.10048831e-01 1.97786137e-01 -5.64751148e-01 5.41878939e-01
3.16075593e-01 7.54696071e-01 2.14750320e-01 -2.02662989e-01
-3.24024737e-01 1.60954595e-01 5.27271807e-01 -6.88031137e-01
-2.55142570e-01 5.93957603e-01 -1.83574647e-01 -9.77160186e-02
-5.50970972e-01 -6.88859046e-01 -1.54079342e+00 -2.24153489e-01
-3.32808614e-01 6.34935796e-02 2.46324375e-01 8.95789802e-01
2.16749907e-01 5.38935661e-01 2.36900300e-01 -9.20081615e-01
6.54432252e-02 -8.52855742e-01 -6.11780524e-01 3.25101763e-01
3.51800829e-01 -3.63306314e-01 -2.87538588e-01 9.45274055e-01] | [14.651463508605957, -3.1859095096588135] |
e655da11-970c-4399-8277-beec3ea27b2f | zoo-guide-to-network-embedding | 2305.03474 | null | https://arxiv.org/abs/2305.03474v1 | https://arxiv.org/pdf/2305.03474v1.pdf | Zoo Guide to Network Embedding | Networks have provided extremely successful models of data and complex systems. Yet, as combinatorial objects, networks do not have in general intrinsic coordinates and do not typically lie in an ambient space. The process of assigning an embedding space to a network has attracted lots of interest in the past few decades, and has been efficiently applied to fundamental problems in network inference, such as link prediction, node classification, and community detection. In this review, we provide a user-friendly guide to the network embedding literature and current trends in this field which will allow the reader to navigate through the complex landscape of methods and approaches emerging from the vibrant research activity on these subjects. | ['Ginestra Bianconi', 'Anaïs Baudot', 'Rubén J. Sánchez-García', 'Anthony Baptista'] | 2023-05-05 | null | null | null | null | ['community-detection', 'network-embedding'] | ['graphs', 'methodology'] | [ 3.10788512e-01 2.90043503e-01 -4.59671050e-01 3.18100601e-02
4.22858715e-01 -7.04907775e-01 6.88897610e-01 4.58055228e-01
-2.64318556e-01 5.65010190e-01 -1.60776258e-01 -5.16085565e-01
-6.46835744e-01 -9.75274384e-01 -1.04020126e-01 -5.69466174e-01
-8.26843500e-01 2.90055245e-01 4.93199192e-02 -1.54601455e-01
1.34076506e-01 5.02307117e-01 -1.08961475e+00 -3.12202632e-01
5.23949444e-01 5.78415513e-01 -2.13262349e-01 7.43400991e-01
-8.70390236e-02 4.57004040e-01 -2.69285202e-01 -7.02847123e-01
-5.01163714e-02 -3.99649739e-01 -7.35905290e-01 -2.11061984e-01
7.13975802e-02 2.51947373e-01 -9.99494374e-01 1.00554490e+00
2.53814816e-01 8.17847997e-03 7.63145447e-01 -1.66318142e+00
-5.68543077e-01 3.99479479e-01 -3.19479793e-01 3.83329451e-01
3.68880898e-01 -3.62512887e-01 1.55595589e+00 -9.17491794e-01
7.52141237e-01 1.15318632e+00 7.74060011e-01 2.15721533e-01
-1.60407972e+00 -2.76220202e-01 8.51750001e-02 3.73964578e-01
-1.43720853e+00 -1.35989428e-01 1.10329449e+00 -6.10049784e-01
5.06857514e-01 4.00173157e-01 1.11082590e+00 8.37096632e-01
1.12935834e-01 5.45934081e-01 7.51053691e-01 -4.56340253e-01
2.80848205e-01 9.07414854e-02 3.26178223e-01 6.58321679e-01
4.73112673e-01 -1.23055931e-02 -3.88911486e-01 -4.30384785e-01
8.33828390e-01 2.96762735e-01 -6.36025816e-02 -9.84107554e-01
-1.39464104e+00 1.22531617e+00 5.42380691e-01 5.56840718e-01
-6.53401092e-02 2.54864216e-01 3.51848245e-01 4.17166978e-01
4.06694710e-01 4.36098069e-01 1.53861661e-02 -1.55747030e-03
-5.50636113e-01 1.73801214e-01 1.24581921e+00 7.67252207e-01
5.10046482e-01 -1.06828034e-01 6.00212634e-01 6.62321448e-01
4.67561692e-01 1.86970904e-01 -3.53431344e-01 -9.30389404e-01
3.06211174e-01 5.47098279e-01 -3.26206595e-01 -1.72394657e+00
-3.97625864e-01 -5.23291707e-01 -1.32122970e+00 -4.88102511e-02
2.85508841e-01 -6.20566076e-03 2.55720527e-03 1.54902077e+00
4.07702953e-01 1.33187190e-01 -1.68868139e-01 4.45126206e-01
5.63296974e-01 4.89583433e-01 -2.47189417e-01 -5.12737110e-02
1.16579688e+00 -7.26607382e-01 -7.41250575e-01 2.71923393e-02
5.26576519e-01 -4.18831140e-01 3.44332993e-01 1.08795561e-01
-1.07316720e+00 -7.76361823e-02 -1.09638083e+00 1.83363289e-01
-6.53898060e-01 -3.77259940e-01 1.03561163e+00 6.19258881e-01
-1.22928631e+00 7.81491339e-01 -9.54377592e-01 -9.92392480e-01
5.18819213e-01 4.16055232e-01 -4.48289126e-01 -2.02281833e-01
-1.27060366e+00 1.00739503e+00 6.11358322e-02 4.51486439e-01
-3.18595141e-01 -5.44981778e-01 -9.24686432e-01 -1.11621313e-01
4.72057313e-01 -6.38626754e-01 5.11893868e-01 -3.54364485e-01
-9.54901278e-01 5.03946900e-01 -1.41587153e-01 -3.55050564e-01
2.81250775e-01 2.65316397e-01 -4.48993355e-01 2.52322108e-01
-2.74010658e-01 1.30396709e-01 3.99779409e-01 -8.71859789e-01
-1.49781108e-01 -3.60442489e-01 3.04646194e-01 1.44934908e-01
-5.54798007e-01 1.04562797e-01 2.74888184e-02 -5.51286638e-01
3.83168280e-01 -9.06112790e-01 -5.30686975e-01 5.68382800e-01
-5.47247589e-01 -2.88188994e-01 7.65705884e-01 -8.61244947e-02
1.40993130e+00 -2.04066157e+00 5.51013947e-01 7.27672875e-01
1.00787532e+00 -5.77754341e-02 -3.87797169e-02 1.05067372e+00
-2.50927389e-01 5.01686156e-01 -3.12635630e-01 -1.03658482e-01
1.71380267e-01 2.02116385e-01 3.28720361e-02 8.86172354e-01
1.80659458e-01 8.58125627e-01 -1.27785718e+00 -4.01111662e-01
4.84448880e-01 7.59506047e-01 -4.02867883e-01 -1.64987385e-01
1.73543334e-01 4.45739329e-02 -5.74073136e-01 4.09163952e-01
1.73181698e-01 -5.38463175e-01 5.35734355e-01 1.00991561e-03
-2.39520799e-02 2.85716802e-01 -1.26523089e+00 1.20080590e+00
-5.85095771e-02 1.16593575e+00 3.89335871e-01 -1.44240379e+00
6.48364961e-01 4.68752444e-01 7.45110095e-01 -5.57831191e-02
-3.05254553e-02 -3.80015932e-02 4.48668659e-01 -1.26721829e-01
2.67565966e-01 1.14982247e-01 1.82619452e-01 7.27579713e-01
-2.07145900e-01 1.20301597e-01 2.36609548e-01 5.58484912e-01
1.46348763e+00 -7.19656944e-01 4.47083384e-01 -2.97228664e-01
5.63065171e-01 -2.46096745e-01 8.21485221e-02 5.54141223e-01
-3.12704593e-01 1.03305895e-02 7.68788815e-01 -3.67726058e-01
-1.26551867e+00 -1.24718428e+00 -3.47322643e-01 6.46254838e-01
9.84484404e-02 -7.97391653e-01 -4.63727474e-01 -3.17788422e-01
1.36726350e-01 -2.62241900e-01 -8.27384293e-01 -5.45661934e-02
-3.85676086e-01 -7.45223224e-01 4.79961455e-01 3.43932897e-01
3.94069515e-02 -9.08465445e-01 2.64580041e-01 2.01451018e-01
1.92624666e-02 -8.72430861e-01 -2.90515751e-01 3.53438617e-03
-1.11782885e+00 -1.32065582e+00 -4.31001455e-01 -9.25084472e-01
9.01081920e-01 3.54339302e-01 1.19846261e+00 5.17740667e-01
-5.89855313e-01 5.96425891e-01 -2.12875545e-01 1.42598301e-01
-1.81339353e-01 2.92824030e-01 4.02163297e-01 -8.07498991e-02
3.88984799e-01 -8.79855275e-01 -5.63126683e-01 4.63553250e-01
-9.23948526e-01 -2.43769482e-01 3.30690563e-01 8.61535490e-01
1.79521397e-01 2.42233679e-01 5.93125463e-01 -7.21222937e-01
8.93406510e-01 -8.88793468e-01 -5.82407832e-01 1.16732277e-01
-6.52326107e-01 -2.42434815e-01 4.70864445e-01 -1.16976261e-01
1.26387686e-01 -2.54003465e-01 1.60251766e-01 2.00899348e-01
2.35606477e-01 9.97261345e-01 -1.03822254e-01 -4.76492077e-01
4.01955128e-01 -4.71101031e-02 2.90915906e-01 -2.83263743e-01
5.01893759e-01 3.84131223e-01 1.06422745e-01 -2.38048524e-01
1.03987420e+00 3.65845352e-01 5.90985358e-01 -1.17011619e+00
-2.24819332e-01 -3.77624661e-01 -1.00625074e+00 -2.59591252e-01
5.68629384e-01 -3.86894256e-01 -8.67759526e-01 1.29513457e-01
-1.08009112e+00 -2.70595942e-02 -6.59182072e-02 3.57027352e-01
-3.52540135e-01 5.16480207e-01 -7.11694717e-01 -8.52569938e-01
5.53476885e-02 -7.55409718e-01 7.79325143e-02 -3.59346019e-03
-4.96505260e-01 -2.01724482e+00 3.55479449e-01 -1.75610170e-01
5.57046413e-01 4.35650527e-01 1.14491820e+00 -5.34406066e-01
-7.15034783e-01 -5.01259387e-01 -7.61472583e-02 1.25888363e-01
2.70367712e-01 5.08810878e-01 -5.16225219e-01 -2.15391263e-01
-4.89759713e-01 8.72556120e-02 5.81993401e-01 2.09777609e-01
9.23336983e-01 -2.97349423e-01 -7.14047432e-01 3.83242458e-01
1.36461020e+00 -3.08024585e-01 3.58059019e-01 2.51208078e-02
7.23917305e-01 7.53973544e-01 2.87901540e-03 1.72845513e-01
3.70542437e-01 7.08042860e-01 4.56847608e-01 -1.15216441e-01
1.50775671e-01 -1.68887019e-01 7.47637972e-02 1.16972184e+00
-1.64546192e-01 -1.91530094e-01 -9.98301268e-01 5.51063180e-01
-1.88065434e+00 -1.25711036e+00 -3.17309678e-01 1.98567212e+00
6.01785481e-01 1.17722206e-01 3.13875407e-01 3.14824641e-01
1.10677791e+00 3.21303993e-01 -4.53299850e-01 -2.06305921e-01
-3.49826783e-01 1.99111048e-02 4.16749954e-01 6.95994318e-01
-1.04598057e+00 5.75742543e-01 8.10847187e+00 3.69560927e-01
-6.08251691e-01 -1.61835283e-01 2.99834162e-01 1.54095575e-01
-3.44198376e-01 1.04342714e-01 -2.53667921e-01 2.88873136e-01
8.02568018e-01 -5.41265726e-01 5.53344369e-01 6.91339195e-01
2.39250839e-01 9.62202922e-02 -1.37885451e+00 8.69727910e-01
-2.24926516e-01 -1.45090938e+00 -3.82030308e-01 6.28069639e-01
7.02504694e-01 1.13769814e-01 3.01653802e-01 -2.65280366e-01
4.13078785e-01 -1.29736865e+00 -9.01631191e-02 3.56183112e-01
5.34915864e-01 -5.66171765e-01 4.57301587e-01 1.09768249e-01
-1.36712480e+00 1.21621937e-01 -3.18498641e-01 -4.45987195e-01
2.60879755e-01 9.33120728e-01 -6.97872341e-01 4.16108519e-01
3.64993900e-01 1.33393598e+00 -2.19560549e-01 1.28889012e+00
3.97804528e-02 6.82980835e-01 -4.73570019e-01 -4.29431021e-01
-5.28388657e-02 -6.29832923e-01 7.36406624e-01 7.94974506e-01
-1.16200551e-01 5.80419153e-02 3.54023278e-02 8.26391995e-01
-5.86802922e-02 6.95672706e-02 -1.07909405e+00 -4.93395805e-01
8.20437610e-01 1.31195104e+00 -9.11328614e-01 -5.33243455e-02
-5.66376328e-01 5.26690900e-01 2.79223591e-01 4.13159847e-01
-4.46336657e-01 -6.88600242e-01 9.85447526e-01 3.08742285e-01
7.43010491e-02 -7.80457795e-01 -2.75922298e-01 -9.91237640e-01
-4.74928916e-02 -5.34254730e-01 1.18447118e-01 -3.12237412e-01
-1.56108499e+00 5.28686285e-01 -9.67784300e-02 -8.05859029e-01
-1.63227916e-01 -7.94821024e-01 -7.00889885e-01 7.43416727e-01
-9.80807364e-01 -4.68204290e-01 -6.12524338e-02 3.81493777e-01
-1.94097593e-01 -5.15973195e-02 1.02048850e+00 3.22350234e-01
-7.72302032e-01 2.03388721e-01 5.69081664e-01 5.18660545e-01
1.93618774e-01 -1.27255845e+00 6.07436776e-01 5.84741950e-01
2.49932557e-01 9.90436554e-01 6.56090915e-01 -5.19372702e-01
-1.63878191e+00 -8.43514800e-01 9.64755118e-01 -6.97346270e-01
1.31235063e+00 -7.62230098e-01 -6.96746707e-01 6.76030934e-01
4.42083701e-02 3.06603998e-01 9.64868009e-01 4.48219359e-01
-2.87050784e-01 -3.67760360e-02 -1.05593240e+00 8.64401281e-01
1.22930026e+00 -7.47259617e-01 -1.41842812e-01 4.35570210e-01
3.09505373e-01 2.40298167e-01 -1.27196968e+00 1.35338873e-01
6.33779943e-01 -6.03159606e-01 1.22619021e+00 -6.98268592e-01
1.06227875e-01 -1.35506168e-01 -1.04425875e-02 -1.18353641e+00
-5.53260148e-01 -8.03487182e-01 -4.16683853e-01 9.69276786e-01
3.80133659e-01 -1.00546122e+00 1.05863154e+00 4.43622947e-01
3.71653259e-01 -1.10215652e+00 -8.41078877e-01 -5.64143062e-01
4.83639129e-02 -1.85332939e-01 3.60605687e-01 1.00385916e+00
6.42243683e-01 4.76588905e-01 -2.47449931e-02 -2.04492986e-01
1.03035176e+00 -4.06214744e-01 5.40420413e-01 -1.86181712e+00
9.70842466e-02 -7.35032439e-01 -1.06036007e+00 -9.74124730e-01
1.73140138e-01 -1.10322344e+00 -3.80222589e-01 -1.66071284e+00
1.91495940e-01 -8.39750886e-01 -1.26296371e-01 3.20814885e-02
1.65640905e-01 5.30815065e-01 -8.23059529e-02 2.52840221e-01
-6.40349865e-01 3.87170464e-01 1.03921807e+00 -3.72016057e-02
1.41965002e-01 3.01313043e-01 -7.65489161e-01 6.96029544e-01
8.44114721e-01 -3.72025222e-01 -5.64315438e-01 -1.05982602e-01
8.38715017e-01 -1.93306118e-01 5.35746157e-01 -5.31989634e-01
3.04422379e-01 -7.78171653e-03 1.01229846e-01 -4.90189195e-01
5.57200313e-01 -1.01033616e+00 2.22288877e-01 5.12653828e-01
-3.25891435e-01 5.83258569e-01 -2.74663508e-01 9.69911277e-01
-7.15071857e-02 -1.94120631e-01 3.70179534e-01 1.09237455e-01
-3.19592088e-01 5.82117438e-01 -3.74859601e-01 1.75128374e-02
1.14079928e+00 -2.88160682e-01 -3.65129024e-01 -6.38304412e-01
-9.10157621e-01 2.48012573e-01 6.81035042e-01 2.45695040e-01
6.40288472e-01 -1.53616500e+00 -5.53854942e-01 -1.58278227e-01
-2.49997731e-02 -2.18251854e-01 -7.46599585e-02 1.09662354e+00
-5.83907723e-01 5.75765848e-01 3.70393209e-02 -6.05565786e-01
-1.15883112e+00 5.62461913e-01 3.08114022e-01 -1.25370488e-01
-5.32921672e-01 4.92283762e-01 -1.77631065e-01 -3.23434263e-01
3.04306448e-01 1.99274247e-04 -2.70187497e-01 -6.41739443e-02
4.33194429e-01 7.52035797e-01 -3.29267949e-01 -6.28930271e-01
-5.41092396e-01 4.36179310e-01 1.62056312e-01 -1.03615016e-01
1.50516403e+00 -2.63989985e-01 -7.89970815e-01 6.47754252e-01
1.43847978e+00 9.62610636e-03 -7.99366236e-01 -3.21533680e-01
1.00322433e-01 -2.16691986e-01 -1.87996641e-01 -1.68698907e-01
-9.75493193e-01 7.92480409e-01 5.46048256e-03 1.13936067e+00
4.38500583e-01 3.67099345e-01 1.82084650e-01 3.12400550e-01
2.58885264e-01 -7.41238892e-01 1.30296439e-01 6.08688235e-01
6.06128812e-01 -9.44821894e-01 1.42197937e-01 -8.59930992e-01
1.30326271e-01 1.14337039e+00 2.48608980e-02 -4.07476932e-01
1.30499148e+00 3.39811817e-02 -4.68544453e-01 -3.33253473e-01
-7.45261490e-01 2.68941205e-02 2.08082184e-01 8.83103490e-01
6.27946556e-01 1.64986607e-02 -2.46859834e-01 -4.15726472e-03
-2.94271469e-01 -4.88766342e-01 6.19625092e-01 6.91475093e-01
-4.86807257e-01 -1.35473323e+00 -1.76060304e-01 6.77087843e-01
-1.49191037e-01 -2.04099670e-01 -5.67727685e-01 7.14364409e-01
-3.22130680e-01 9.41164851e-01 6.31346107e-02 -3.18884730e-01
-1.14894286e-01 -2.60455906e-01 5.52213907e-01 -6.04589701e-01
-1.32311918e-02 -5.09566367e-01 5.73632366e-04 -2.44544461e-01
-5.18711925e-01 -9.65019405e-01 -7.81360567e-01 -1.00988460e+00
-4.95497137e-01 4.08540398e-01 8.19576800e-01 7.40470290e-01
2.62635767e-01 3.86346459e-01 7.38266945e-01 -7.35092163e-01
-1.87797278e-01 -7.82431722e-01 -8.82430971e-01 -2.19913542e-01
4.36014414e-01 -7.43628144e-01 -4.63323623e-01 -3.50576639e-01] | [7.075277805328369, 5.823516845703125] |
e666bb3f-e0c0-4971-a27f-8f25aedcf57f | nadi-2021-the-second-nuanced-arabic-dialect | 2103.08466 | null | https://arxiv.org/abs/2103.08466v2 | https://arxiv.org/pdf/2103.08466v2.pdf | NADI 2021: The Second Nuanced Arabic Dialect Identification Shared Task | We present the findings and results of the Second Nuanced Arabic Dialect Identification Shared Task (NADI 2021). This Shared Task includes four subtasks: country-level Modern Standard Arabic (MSA) identification (Subtask 1.1), country-level dialect identification (Subtask 1.2), province-level MSA identification (Subtask 2.1), and province-level sub-dialect identification (Subtask 2.2). The shared task dataset covers a total of 100 provinces from 21 Arab countries, collected from the Twitter domain. A total of 53 teams from 23 countries registered to participate in the tasks, thus reflecting the interest of the community in this area. We received 16 submissions for Subtask 1.1 from five teams, 27 submissions for Subtask 1.2 from eight teams, 12 submissions for Subtask 2.1 from four teams, and 13 Submissions for subtask 2.2 from four teams. | ['Nizar Habash', 'Houda Bouamor', 'AbdelRahim Elmadany', 'Chiyu Zhang', 'Muhammad Abdul-Mageed'] | 2021-03-04 | null | https://aclanthology.org/2021.wanlp-1.28 | https://aclanthology.org/2021.wanlp-1.28.pdf | eacl-wanlp-2021-4 | ['dialect-identification'] | ['natural-language-processing'] | [-2.46678725e-01 -1.88171506e-01 -7.75506273e-02 -4.66641515e-01
-1.27100992e+00 -1.06415057e+00 1.27929485e+00 2.23057166e-01
-4.32545394e-01 8.83335054e-01 4.12632227e-01 -4.51717347e-01
8.86851549e-03 -4.38622355e-01 -2.41443783e-01 -4.65125501e-01
-3.16333920e-01 7.47777939e-01 4.27252054e-02 -7.86284626e-01
7.12124884e-01 2.27012053e-01 -1.00748956e+00 5.69924533e-01
1.24924934e+00 3.96738470e-01 -1.80771947e-01 5.69221735e-01
1.44202605e-01 8.66282165e-01 -4.41870421e-01 -5.93020797e-01
1.54475138e-01 -1.91832438e-01 -1.29660618e+00 -1.62654668e-01
5.76517105e-01 -2.16453671e-01 2.12822706e-01 1.03003871e+00
4.49223399e-01 7.05510080e-02 9.53181028e-01 -1.21266115e+00
-6.15855157e-01 8.58961701e-01 -7.69194782e-01 -6.21946417e-02
6.44691527e-01 -7.13068470e-02 1.13862872e+00 -1.30591929e+00
7.50007033e-01 1.23251402e+00 7.56223559e-01 1.30131245e-02
-9.17353630e-01 -1.10023177e+00 2.31613144e-01 7.18462691e-02
-1.76476204e+00 -7.89400399e-01 1.62400022e-01 -7.57807732e-01
7.27942705e-01 3.09817612e-01 2.61824280e-01 8.46847355e-01
-2.56320149e-01 7.97716856e-01 1.69040596e+00 -2.72506058e-01
-2.22020194e-01 1.74050838e-01 3.55751574e-01 2.77395159e-01
-1.37518449e-02 -5.33001900e-01 -2.84937054e-01 -2.72082061e-01
6.05401099e-01 -6.00414991e-01 1.68286294e-01 5.21140695e-01
-1.63885307e+00 9.15925920e-01 1.62740037e-01 2.32301578e-01
-4.57199782e-01 -5.24082780e-01 4.95923489e-01 7.06925988e-01
7.73586988e-01 3.72746944e-01 -3.89596999e-01 -4.60257381e-01
-7.35955536e-01 6.17077291e-01 7.91520000e-01 8.79757524e-01
8.19867969e-01 -5.46555482e-02 3.85635556e-03 1.50864983e+00
3.65185559e-01 8.51385713e-01 2.23389909e-01 -8.08515906e-01
6.30098045e-01 5.82360446e-01 5.66179395e-01 -1.07522416e+00
-7.95783043e-01 -7.54900184e-03 -6.29122436e-01 -1.39902961e-02
1.02683079e+00 -7.04042971e-01 -7.30628252e-01 1.57988942e+00
1.09900415e-01 -3.30243260e-01 -2.28651464e-01 1.07363117e+00
9.66610551e-01 6.93011522e-01 -2.36264527e-01 1.01386935e-01
1.60360682e+00 -1.04656672e+00 -3.43036681e-01 -2.58719295e-01
7.31683612e-01 -1.28905809e+00 8.67404521e-01 3.26656193e-01
-9.41474557e-01 -2.78124303e-01 -6.21551037e-01 3.28501463e-01
-5.95514953e-01 1.48628220e-01 2.41001174e-01 7.80334115e-01
-1.47132051e+00 -3.79013717e-01 -3.11784834e-01 -6.52679682e-01
6.35175928e-02 3.77418041e-01 -6.38532460e-01 -1.73425287e-01
-1.27703714e+00 8.59950066e-01 -1.86627164e-01 -6.21417947e-02
-7.65583158e-01 -3.91521066e-01 -7.25226820e-01 -5.10058820e-01
-2.47930482e-01 -8.87840241e-03 1.20116782e+00 -1.18897581e+00
-1.33417273e+00 1.47972810e+00 -2.23209932e-01 -4.37294170e-02
3.37704062e-01 4.03089598e-02 -6.63529038e-01 -3.04403931e-01
8.45257640e-01 5.34567535e-01 6.62797809e-01 -1.16237497e+00
-1.10701466e+00 -5.01417041e-01 9.79262292e-02 4.08391237e-01
-5.15795611e-02 7.43530095e-01 -1.09845564e-01 -1.01630557e+00
-5.15503548e-02 -1.05934799e+00 -8.29476416e-02 -1.16421664e+00
-3.37548345e-01 -3.63773972e-01 5.86248517e-01 -1.36151123e+00
1.17465436e+00 -2.05738235e+00 4.29977775e-02 3.34853351e-01
3.11870545e-01 4.46003042e-02 -4.40713823e-01 6.83419406e-01
-1.29858077e-01 3.53820361e-02 -1.88266814e-01 -4.80364591e-01
9.99997258e-02 -6.21160865e-01 -1.61234871e-01 5.44860482e-01
2.18128115e-01 6.17502987e-01 -1.11115313e+00 -5.57834189e-03
-3.94975692e-01 -2.20196471e-01 -8.44472274e-02 -3.32190156e-01
1.95725664e-01 5.52162886e-01 -1.37416333e-01 1.16561127e+00
1.13822269e+00 -1.86100025e-02 2.01898023e-01 1.55774683e-01
-7.51827717e-01 4.53172445e-01 -9.51235712e-01 1.00534475e+00
-2.76528776e-01 9.18946862e-01 7.10354686e-01 -5.51536918e-01
1.10514092e+00 3.16075683e-01 4.17664737e-01 -7.51447439e-01
-2.02411443e-01 6.47758722e-01 1.13603495e-01 2.66192168e-01
9.22568440e-01 2.49352351e-01 -5.97570002e-01 1.30840957e+00
-2.19340339e-01 -7.81825092e-03 4.07993436e-01 2.18695149e-01
6.85626864e-01 -2.83644944e-01 2.42294699e-01 -8.63685071e-01
8.35857332e-01 3.59203458e-01 7.08221495e-02 7.51633942e-01
-6.18894935e-01 5.09850919e-01 3.64663005e-01 -7.50488877e-01
-7.93727517e-01 -8.31384778e-01 -1.96637139e-01 1.81855810e+00
-3.64567310e-01 -4.49111968e-01 -7.89000094e-01 -6.71443284e-01
-1.02036320e-01 1.91047207e-01 -5.62907577e-01 4.97752994e-01
-7.01538086e-01 -1.03200567e+00 1.10546350e+00 4.21537720e-02
8.95860434e-01 -8.10823083e-01 1.13203607e-01 1.05311535e-01
-1.04501545e+00 -8.88235211e-01 -8.05869341e-01 -3.06514978e-01
-1.53339162e-01 -9.97207165e-01 -1.34763789e+00 -1.22300088e+00
4.00714785e-01 4.36848462e-01 1.27182209e+00 -1.74328163e-01
4.28994626e-01 1.73078284e-01 -3.96325171e-01 -5.08511484e-01
-3.61194372e-01 7.18852520e-01 2.10874021e-01 3.13619047e-01
6.42666459e-01 8.32101107e-02 -4.31179022e-03 8.23341012e-01
-1.81766257e-01 1.72128342e-02 1.52089596e-01 5.31143785e-01
1.05323493e-02 -4.71565336e-01 9.61580813e-01 -6.99251056e-01
8.55543792e-01 -7.83761442e-01 -7.01954246e-01 1.44386247e-01
-1.68039367e-01 -8.47535431e-01 2.78783262e-01 6.89261183e-02
-1.08905649e+00 -1.70241706e-02 -4.86919694e-02 7.78888822e-01
-3.44337761e-01 9.55287576e-01 2.58515835e-01 -8.55183452e-02
7.92548656e-01 3.54745418e-01 2.99646646e-01 -1.81557670e-01
1.88800305e-01 1.54625404e+00 4.61273342e-01 -7.17816293e-01
6.77408338e-01 2.02983707e-01 -6.32142663e-01 -1.11650121e+00
-5.95239639e-01 -5.33840835e-01 -8.39591622e-01 -3.77531260e-01
7.65774369e-01 -1.63221812e+00 -4.48084116e-01 1.25836384e+00
-9.35324788e-01 -8.69840086e-01 4.90363985e-01 2.84892857e-01
-1.86961859e-01 1.32242486e-01 -6.76202178e-01 -6.21636450e-01
-2.48404190e-01 -1.26163757e+00 9.08879459e-01 -1.72454089e-01
-7.33843744e-01 -1.08358443e+00 3.68900448e-01 5.60502052e-01
4.34197485e-01 2.75553048e-01 5.86335361e-01 -7.98638880e-01
-2.20838949e-01 -3.82214934e-02 -7.01965332e-01 -2.23784804e-01
5.20535946e-01 5.28763942e-02 -1.05935788e+00 -5.45694888e-01
-6.40112579e-01 -5.46968877e-01 5.18367290e-01 3.81926328e-01
1.63718909e-01 -2.00868681e-01 1.58514142e-01 7.16341585e-02
6.69096053e-01 7.66396150e-02 3.30380768e-01 8.45738351e-01
6.41319752e-01 7.68147767e-01 6.15516365e-01 4.81267989e-01
1.28642797e+00 7.76035190e-01 1.84921861e-01 3.71508487e-02
9.05043185e-02 3.67547154e-01 8.55434835e-01 1.08013797e+00
-5.55790782e-01 3.72711718e-01 -1.78059030e+00 1.11677408e+00
-1.84419358e+00 -8.61471534e-01 -8.00519407e-01 2.23762536e+00
9.47061479e-01 -3.88179392e-01 1.13247609e+00 -8.61870423e-02
7.97701895e-01 1.53141364e-01 -2.70908307e-02 -6.18648589e-01
-3.78870636e-01 -3.83163720e-01 4.25263494e-01 7.53503144e-01
-1.50334239e+00 1.35261130e+00 6.43731642e+00 7.27901638e-01
-9.41006005e-01 5.63323274e-02 6.94962680e-01 3.44558895e-01
-1.62892491e-02 -7.89000764e-02 -8.25071573e-01 5.06105721e-01
1.06504142e+00 -3.58251005e-01 6.98560059e-01 2.19020665e-01
2.26725806e-02 -1.05384640e-01 -5.38307130e-01 7.74908483e-01
-2.88722701e-02 -1.09479654e+00 1.39547303e-01 2.91736692e-01
1.04161978e+00 7.35034645e-01 1.81480467e-01 3.13506693e-01
7.10336089e-01 -1.25375426e+00 9.26153541e-01 1.81237146e-01
1.15108430e+00 -9.76512849e-01 9.30012405e-01 3.06341350e-01
-1.29760909e+00 -1.05898552e-01 2.41881181e-02 -5.94423175e-01
5.60599342e-02 2.43076667e-01 -9.29735124e-01 6.88736439e-01
5.78456461e-01 1.18249762e+00 -6.97625995e-01 7.42190301e-01
-1.35477241e-02 8.28775585e-01 -2.77733415e-01 1.21207722e-01
8.01953733e-01 -4.96678352e-01 7.07446337e-01 1.74292421e+00
1.45695850e-01 -2.88437009e-01 1.23531088e-01 1.21056683e-01
-9.94772464e-02 3.66030574e-01 -6.41675413e-01 -1.09825954e-01
7.05942631e-01 1.16386402e+00 -5.97513676e-01 -1.66069567e-01
-5.51111937e-01 8.63319993e-01 1.53199181e-01 4.91400242e-01
-3.11673135e-01 -6.55232608e-01 1.00555849e+00 -4.95962277e-02
-8.35362896e-02 -3.62175822e-01 -5.10641038e-01 -8.95085216e-01
-4.44568932e-01 -1.34440410e+00 7.49365747e-01 -3.66085559e-01
-1.53559184e+00 5.16359866e-01 -1.23835653e-01 -1.14002132e+00
-5.71034849e-01 -5.52657723e-01 -5.19874454e-01 1.34672272e+00
-1.22390711e+00 -1.53639698e+00 -2.68440098e-01 9.44511175e-01
4.70471859e-01 -7.77606666e-01 1.08300519e+00 6.42948031e-01
-4.33627278e-01 8.38365316e-01 2.03601390e-01 6.04630053e-01
1.41217077e+00 -1.17479789e+00 6.86912775e-01 5.33649385e-01
-2.76941806e-01 7.13976324e-01 8.70529935e-02 -6.05792582e-01
-9.55354154e-01 -1.13237524e+00 1.68128669e+00 -8.51069033e-01
8.90996218e-01 -3.36778849e-01 -4.02244091e-01 9.99067187e-01
4.45846468e-01 -7.63128877e-01 7.00520515e-01 4.87940401e-01
-3.24908555e-01 1.82432547e-01 -1.16505373e+00 6.68865800e-01
5.01425683e-01 -9.22719300e-01 -1.85560703e-01 5.12930036e-01
1.60582289e-01 -5.28207958e-01 -1.06007779e+00 -1.81575432e-01
8.76387715e-01 -5.25578976e-01 8.62212658e-01 -3.95578980e-01
4.14011627e-01 -1.94065869e-01 -4.43260491e-01 -1.59132969e+00
-7.06779778e-01 -7.07399845e-01 5.92148423e-01 1.03105867e+00
8.46104503e-01 -1.10755563e+00 1.40102997e-01 2.13162959e-01
-1.96100518e-01 -1.39709925e-02 -8.76242459e-01 -6.38553977e-01
7.84035563e-01 -1.63227826e-01 8.70216727e-01 1.71083033e+00
1.16165467e-01 5.49104333e-01 -1.88916236e-01 2.07141295e-01
4.43451107e-01 9.67900530e-02 1.10721922e+00 -1.23653686e+00
5.61116099e-01 -1.06889045e+00 5.58595685e-03 -8.20317924e-01
1.32701531e-01 -1.07033968e+00 -2.65554279e-01 -1.51924229e+00
7.45626539e-02 -6.18441284e-01 4.40355502e-02 5.79282403e-01
-1.49506509e-01 8.19748878e-01 2.53274083e-01 5.66675067e-01
-5.42682111e-01 -1.21179447e-01 9.29547727e-01 -2.67590702e-01
-3.14811468e-01 9.14795771e-02 -1.09194350e+00 5.45934677e-01
1.10862923e+00 -1.03886731e-01 8.05324987e-02 -6.52683377e-01
4.53368843e-01 -3.93913686e-01 2.21453384e-01 -4.71809238e-01
1.90326810e-01 -2.49705657e-01 5.78633621e-02 -6.14230275e-01
6.79652393e-02 -1.61867470e-01 -2.78524011e-01 5.61175644e-02
2.33412117e-01 5.52603841e-01 3.57158005e-01 -1.99057207e-01
-4.40807462e-01 1.16904519e-01 5.76423883e-01 -1.16300069e-01
-7.82637179e-01 1.86457351e-01 -1.16765666e+00 4.30402726e-01
7.50793755e-01 -2.33618915e-01 -4.71835196e-01 -7.24877477e-01
-6.50107265e-01 5.54510474e-01 4.77789938e-01 5.08325040e-01
1.99004248e-01 -1.49075961e+00 -1.70901501e+00 2.85649121e-01
2.29772985e-01 -3.08151633e-01 -9.86056402e-02 1.17792070e+00
-6.29969180e-01 5.92618287e-01 -4.82971340e-01 -4.33385521e-01
-1.05579519e+00 -5.69605410e-01 3.76554579e-02 -2.54696280e-01
1.20134160e-01 8.87196064e-01 4.66721095e-02 -1.35214329e+00
-2.80929357e-01 5.01881599e-01 -5.84863663e-01 6.94872856e-01
9.26539242e-01 7.61672795e-01 1.72263905e-01 -1.57595861e+00
-6.59984887e-01 4.88728732e-01 -4.62669760e-01 -6.19603872e-01
1.18575692e+00 -3.52360278e-01 -8.24298918e-01 4.69606727e-01
9.06876326e-01 2.68554240e-01 -5.05843818e-01 -1.82726830e-01
3.75814885e-01 -2.30508417e-01 -3.78972948e-01 -1.22049522e+00
-4.11739618e-01 3.84223700e-01 1.15262866e-01 2.72104174e-01
8.39097321e-01 -3.19164544e-01 6.26141071e-01 1.28811881e-01
6.54921114e-01 -1.23746657e+00 -3.23386371e-01 1.64361763e+00
9.86766756e-01 -1.27372277e+00 -3.43545705e-01 -9.58688557e-02
-8.42680514e-01 8.99304867e-01 4.73366946e-01 2.51340538e-01
7.49340653e-01 -1.81907460e-01 6.76774383e-01 -1.32909030e-01
-2.48789966e-01 -7.12675154e-02 2.70699739e-01 9.69055891e-01
7.27466166e-01 5.12488246e-01 -2.84617156e-01 8.72998416e-01
-3.22067440e-01 -4.80098486e-01 5.91847837e-01 7.09752381e-01
-1.67414501e-01 -1.01481593e+00 -7.42079616e-01 3.52642149e-01
-3.19952935e-01 -3.85714948e-01 -1.07423699e+00 7.58742511e-01
-1.85510114e-01 1.28955746e+00 3.18338245e-01 -5.53961575e-01
7.39252418e-02 -3.71518843e-02 2.21659154e-01 -4.74352688e-01
-1.02622211e+00 -3.02619010e-01 7.67023087e-01 -2.57777989e-01
-2.43168712e-01 -1.12056983e+00 -9.26122367e-01 -7.92652309e-01
2.18707725e-01 5.67762144e-02 3.83046091e-01 6.66254222e-01
3.09351921e-01 -4.75769341e-01 8.06002557e-01 -8.63057613e-01
-3.69872630e-01 -1.27744448e+00 -6.64731562e-01 -1.61382202e-02
5.92524171e-01 -4.04959410e-01 1.59002077e-02 -8.77422839e-03] | [10.17859172821045, 10.773660659790039] |
23a34946-ecc9-44ae-8b15-de63cc8eb3fc | msvd-turkish-a-comprehensive-multimodal | 2012.07098 | null | https://arxiv.org/abs/2012.07098v1 | https://arxiv.org/pdf/2012.07098v1.pdf | MSVD-Turkish: A Comprehensive Multimodal Dataset for Integrated Vision and Language Research in Turkish | Automatic generation of video descriptions in natural language, also called video captioning, aims to understand the visual content of the video and produce a natural language sentence depicting the objects and actions in the scene. This challenging integrated vision and language problem, however, has been predominantly addressed for English. The lack of data and the linguistic properties of other languages limit the success of existing approaches for such languages. In this paper we target Turkish, a morphologically rich and agglutinative language that has very different properties compared to English. To do so, we create the first large scale video captioning dataset for this language by carefully translating the English descriptions of the videos in the MSVD (Microsoft Research Video Description Corpus) dataset into Turkish. In addition to enabling research in video captioning in Turkish, the parallel English-Turkish descriptions also enables the study of the role of video context in (multimodal) machine translation. In our experiments, we build models for both video captioning and multimodal machine translation and investigate the effect of different word segmentation approaches and different neural architectures to better address the properties of Turkish. We hope that the MSVD-Turkish dataset and the results reported in this work will lead to better video captioning and multimodal machine translation models for Turkish and other morphology rich and agglutinative languages. | ['Lucia Specia', 'Pranava Madhyastha', 'Aykut Erdem', 'Erkut Erdem', 'Menekse Kuyu', 'Ozan Caglayan', 'Begum Citamak'] | 2020-12-13 | null | null | null | null | ['video-description', 'multimodal-machine-translation'] | ['computer-vision', 'natural-language-processing'] | [ 1.83834255e-01 -9.52091888e-02 -1.12510927e-01 -3.32987964e-01
-7.63356924e-01 -8.67669642e-01 8.57484877e-01 -5.96673563e-02
-3.99104923e-01 6.64381742e-01 3.06654364e-01 -4.22351986e-01
5.82053244e-01 -3.24731797e-01 -8.63168716e-01 -4.07125473e-01
2.60718137e-01 6.81075990e-01 2.81031430e-03 -2.45792598e-01
-8.69766250e-02 1.63542643e-01 -1.32556129e+00 6.96950436e-01
6.69045746e-01 5.50957978e-01 4.34732676e-01 7.70353973e-01
-1.08197346e-01 7.18231678e-01 -4.06834424e-01 -6.33697748e-01
2.24883147e-02 -7.50252426e-01 -9.00383592e-01 4.57164139e-01
8.28133285e-01 -4.32015568e-01 -3.16971630e-01 8.13658953e-01
2.54984617e-01 1.29279178e-02 6.81492329e-01 -1.38115454e+00
-1.11393261e+00 5.81511855e-01 -3.93808872e-01 1.02125414e-01
6.78775430e-01 9.49150845e-02 8.69154930e-01 -7.45139778e-01
1.05387437e+00 1.26671612e+00 9.51298699e-02 8.62403095e-01
-9.57959712e-01 -1.63600802e-01 1.69020176e-01 2.75705874e-01
-1.24129856e+00 -4.18390781e-01 5.76006591e-01 -3.42218369e-01
9.15162325e-01 2.08377436e-01 5.35154521e-01 1.60878956e+00
-2.78385520e-01 1.10616302e+00 7.54932284e-01 -7.80723631e-01
-1.09147847e-01 3.66337925e-01 -2.63414502e-01 6.20505929e-01
2.22537443e-01 -2.18226343e-01 -2.30799571e-01 3.63610208e-01
7.27751195e-01 -5.29086530e-01 -4.45912838e-01 -3.58727306e-01
-1.64336205e+00 8.23788583e-01 -7.42956698e-02 5.69680333e-01
-2.37843201e-01 1.21556163e-01 6.29487932e-01 2.32395172e-01
2.11422756e-01 3.11552435e-01 -4.48352486e-01 -2.73661792e-01
-7.13965356e-01 1.04777358e-01 7.51817942e-01 1.15714836e+00
5.33960104e-01 1.36155739e-01 -5.88486008e-02 6.99445546e-01
2.79993683e-01 7.53995478e-01 4.53556567e-01 -9.22695637e-01
7.91960955e-01 3.09086889e-01 7.26841465e-02 -9.58535254e-01
2.04407014e-02 4.80161577e-01 -9.99373049e-02 -3.41758966e-01
6.72579587e-01 -4.15171027e-01 -9.76704478e-01 1.76656497e+00
6.25582710e-02 -1.74886957e-02 5.97761452e-01 1.31687832e+00
1.03732193e+00 9.70744431e-01 3.30839306e-01 -1.96349159e-01
1.59374774e+00 -8.61144543e-01 -7.55044818e-01 -2.54751652e-01
8.85426998e-01 -9.81919169e-01 1.03695309e+00 -1.36919618e-01
-9.96823609e-01 -3.66283417e-01 -6.85720563e-01 -1.90676570e-01
-4.29290920e-01 2.15065271e-01 5.08240223e-01 5.59120834e-01
-1.20379174e+00 -3.22977424e-01 -7.31528938e-01 -8.53561997e-01
9.05477330e-02 1.92811191e-01 -6.14999175e-01 -4.03290033e-01
-1.14416718e+00 9.75237668e-01 6.15608096e-01 -8.34030882e-02
-7.04061270e-01 -1.50276646e-01 -1.35254884e+00 -3.89452696e-01
4.55487549e-01 -6.26834810e-01 1.40885592e+00 -1.80073607e+00
-1.10798872e+00 1.03794992e+00 -2.99058646e-01 -5.83232939e-01
2.55859375e-01 -6.75998405e-02 -2.40777895e-01 6.78542495e-01
9.54553857e-02 1.30689430e+00 6.40292346e-01 -1.18080580e+00
-4.29108351e-01 -2.55277097e-01 7.20400959e-02 5.50317943e-01
-1.94597945e-01 3.70037585e-01 -7.91760385e-01 -6.42872214e-01
-4.82115626e-01 -1.03033638e+00 1.89963318e-02 -3.43799293e-01
2.84798462e-02 5.89102879e-02 8.92989039e-01 -1.01003492e+00
7.15544939e-01 -2.18698287e+00 4.00826991e-01 -4.76017624e-01
-2.17642635e-01 3.27784121e-01 -6.46022677e-01 5.37004590e-01
-3.15122530e-02 2.76828289e-01 -2.26696972e-02 -2.46515781e-01
-1.01383328e-01 4.41896588e-01 -3.60778868e-01 2.57038444e-01
4.84171212e-01 1.30565429e+00 -8.78620863e-01 -8.22434843e-01
1.87222540e-01 6.64899766e-01 -4.20346707e-01 1.46467268e-01
-6.32479608e-01 4.23269838e-01 -5.21152079e-01 7.35473156e-01
2.85396248e-01 1.60704806e-01 1.97433069e-01 -2.63251454e-01
-8.15703943e-02 -1.32791832e-01 -4.35971379e-01 1.69864452e+00
-4.17294741e-01 1.20680761e+00 -4.82892320e-02 -9.38516140e-01
6.16826534e-01 7.74806499e-01 4.21063393e-01 -9.76424694e-01
3.61812949e-01 3.08354199e-01 -1.44126177e-01 -1.04546535e+00
8.32237661e-01 -1.96286410e-01 -2.76501596e-01 3.38021189e-01
-2.18735170e-02 -1.74988165e-01 6.72431886e-01 2.58549631e-01
5.60027599e-01 5.44268608e-01 -8.41138735e-02 2.96772957e-01
3.42842013e-01 6.02202177e-01 6.17617555e-02 3.13753843e-01
-3.35653812e-01 9.60545063e-01 4.52166796e-01 -3.02721947e-01
-1.24263966e+00 -8.03270578e-01 1.81427091e-01 9.41305637e-01
-2.31772847e-02 -2.15633452e-01 -1.01106298e+00 -5.59794426e-01
-5.99057019e-01 9.44052339e-01 -3.40272367e-01 1.16474316e-01
-7.45864213e-01 -4.24749881e-01 7.07501173e-01 4.88493621e-01
4.15359229e-01 -1.20195889e+00 -5.31463742e-01 9.23815593e-02
-7.35857189e-01 -2.11226368e+00 -5.94716370e-01 -2.34353974e-01
-7.14534283e-01 -1.02009892e+00 -1.00684011e+00 -1.31328034e+00
5.92850089e-01 4.38714176e-02 1.07612681e+00 -2.97283560e-01
-1.42765855e-02 9.95731950e-01 -9.16396976e-01 -3.58565658e-01
-6.67212427e-01 -2.00399011e-01 -7.47639462e-02 6.70738518e-02
6.02293551e-01 2.76750505e-01 3.64622176e-02 1.44717559e-01
-1.23773611e+00 4.59972769e-01 6.04384124e-01 4.24326777e-01
3.66391242e-01 -5.00854075e-01 1.30065337e-01 -4.76232469e-01
4.39764351e-01 -3.68322402e-01 -4.53653842e-01 4.06716794e-01
2.69886404e-01 -3.69096734e-02 4.14643198e-01 -7.83845901e-01
-9.28414583e-01 2.74795353e-01 9.34536085e-02 -5.72503805e-01
-5.61128497e-01 6.33955598e-01 -1.11515485e-01 3.02708328e-01
2.32783318e-01 4.16707158e-01 1.06323883e-01 -6.61623701e-02
5.54137528e-01 7.55098999e-01 5.66153944e-01 -5.97265899e-01
5.41136026e-01 2.96248227e-01 -2.63470471e-01 -1.29551172e+00
-4.56136346e-01 -1.84025884e-01 -6.86226666e-01 -3.28855991e-01
1.57133400e+00 -1.22939825e+00 -2.08190486e-01 2.20499843e-01
-1.41329837e+00 -3.62589717e-01 2.56770160e-02 6.52551591e-01
-9.21817243e-01 4.62903529e-01 -5.62442005e-01 -6.02016211e-01
-9.06558856e-02 -1.32204664e+00 1.32306540e+00 1.25418037e-01
-2.92327523e-01 -1.22832215e+00 6.42568395e-02 8.39186251e-01
-5.77413067e-02 4.62810785e-01 9.11799192e-01 -6.48260474e-01
-5.68934560e-01 3.17762536e-03 -2.96217412e-01 2.80175358e-01
-3.65698524e-02 1.99973017e-01 -3.92097265e-01 -5.16344383e-02
-1.79019526e-01 -5.92550635e-01 6.01036787e-01 2.65117615e-01
1.74776256e-01 -2.68354595e-01 1.95579007e-02 2.52939820e-01
1.44655418e+00 3.08027565e-01 7.34857619e-01 4.90544796e-01
8.68077278e-01 9.01859105e-01 9.31598485e-01 -1.61796868e-01
6.59899414e-01 6.73148692e-01 2.84539789e-01 -1.03257857e-01
-2.36454532e-01 -1.60271302e-01 8.94929826e-01 9.41324830e-01
1.18382297e-01 -6.25460327e-01 -1.09274423e+00 8.56536865e-01
-1.89539409e+00 -8.76975834e-01 -2.96822697e-01 1.84104025e+00
5.71427464e-01 -5.30072212e-01 6.15170822e-02 -3.93116474e-01
8.92315328e-01 -1.92890987e-01 5.07838503e-02 -9.56458271e-01
-3.88534546e-01 -2.28534192e-01 2.20933199e-01 3.32020283e-01
-1.21237040e+00 1.30917513e+00 5.70470095e+00 5.59547246e-01
-1.38801241e+00 1.71079598e-02 6.83640122e-01 7.84309655e-02
-1.08226448e-01 -1.15493588e-01 -5.77562749e-01 1.54043347e-01
1.33968258e+00 1.81732729e-01 4.35885578e-01 3.68052065e-01
5.79387546e-01 -1.87203139e-01 -1.30320609e+00 1.10403585e+00
7.31371701e-01 -1.16292703e+00 6.03024304e-01 -1.09232105e-01
6.41882777e-01 9.67148244e-02 4.77381274e-02 2.86990792e-01
-3.28505009e-01 -9.44933414e-01 1.08608758e+00 1.77521512e-01
8.39005589e-01 -4.36556995e-01 7.56064594e-01 8.36623684e-02
-9.31899905e-01 1.25699833e-01 -1.56622529e-01 -2.22956464e-02
3.00180525e-01 -2.55070537e-01 -1.06150448e+00 4.60391283e-01
4.39073086e-01 7.37164915e-01 -7.29570687e-01 7.21725523e-01
-6.85469285e-02 4.36976016e-01 -1.05066009e-01 -3.32940549e-01
6.14154041e-01 -3.87497634e-01 6.64945006e-01 1.24350226e+00
2.92168349e-01 -4.27862667e-02 1.96356148e-01 6.86780453e-01
1.01142898e-01 4.55261379e-01 -9.21875238e-01 -8.71963441e-01
-1.25357971e-01 9.55459595e-01 -9.56196249e-01 -2.53624141e-01
-9.98902082e-01 1.28700221e+00 8.60595703e-02 6.30956173e-01
-1.06756711e+00 1.10911176e-01 3.84854019e-01 1.90275684e-01
4.43713874e-01 -5.61064959e-01 1.07376300e-01 -1.47876418e+00
4.27940339e-02 -1.16495669e+00 2.37264633e-01 -1.34713113e+00
-9.33051288e-01 9.77202117e-01 3.62780899e-01 -1.20874417e+00
-5.62523782e-01 -8.33371699e-01 -2.19587639e-01 4.74036425e-01
-1.24079657e+00 -1.71918106e+00 -4.27271379e-03 6.31025255e-01
9.41183090e-01 -1.31659508e-01 6.73739254e-01 2.52915621e-01
-5.64637303e-01 2.00847447e-01 -9.48920548e-02 5.25323153e-01
8.75833869e-01 -9.26384389e-01 2.16474026e-01 9.41315293e-01
5.21862686e-01 3.56976390e-01 8.54431629e-01 -7.18000710e-01
-1.58071244e+00 -1.03090179e+00 9.12222564e-01 -7.39748657e-01
7.59490013e-01 -2.94936001e-01 -6.21235967e-01 8.68763447e-01
8.79952133e-01 -5.79428375e-01 5.48866868e-01 -6.27204895e-01
-1.42121509e-01 4.89716440e-01 -8.40963542e-01 1.03252780e+00
5.22935331e-01 -4.29730177e-01 -8.27846110e-01 2.73725778e-01
8.74841213e-01 -2.66950041e-01 -5.35882473e-01 1.32930115e-01
3.15475881e-01 -5.99089861e-01 7.31151104e-01 -7.51216173e-01
7.02120066e-01 -2.51722485e-01 -4.15729254e-01 -9.28540885e-01
2.68985182e-01 -4.10583317e-01 3.07884276e-01 1.37262106e+00
6.51449680e-01 -1.05908409e-01 6.24921501e-01 6.60962403e-01
-1.27460837e-01 -3.08352351e-01 -7.30138898e-01 -5.11523724e-01
1.32914513e-01 -5.90063095e-01 4.39375527e-02 9.73204970e-01
-9.19554383e-02 5.93775451e-01 -3.18310350e-01 -1.06952675e-01
1.53533012e-01 9.18517187e-02 6.11078799e-01 -6.06952488e-01
1.64064914e-01 -1.81868449e-01 -5.01118839e-01 -7.43851423e-01
4.94976938e-01 -8.97942007e-01 7.69475549e-02 -1.73597860e+00
2.43978560e-01 1.05756409e-01 5.39949715e-01 6.05412304e-01
4.80239093e-02 7.31142223e-01 4.90096688e-01 8.28520805e-02
-8.14611018e-01 3.69483471e-01 1.37275541e+00 -2.57270694e-01
-1.86864913e-01 -6.66204214e-01 -4.94543523e-01 5.27233660e-01
7.03668237e-01 -1.99840859e-01 -3.50680977e-01 -9.68069673e-01
4.89495099e-02 1.95526123e-01 3.47038448e-01 -5.60041726e-01
-1.52859718e-01 -2.55111575e-01 2.43017882e-01 -3.53373259e-01
5.17230034e-01 -9.62856293e-01 -4.75260615e-02 2.13660464e-01
-1.78080320e-01 3.40489745e-01 4.24448729e-01 2.84395069e-01
-5.22065580e-01 -2.81376719e-01 4.88768905e-01 -2.98689038e-01
-1.07722437e+00 8.24057460e-02 -9.57431793e-01 2.08026052e-01
1.40145075e+00 -6.21790051e-01 -2.22803444e-01 -7.04945505e-01
-7.01380610e-01 3.18372995e-01 8.59353364e-01 8.94683063e-01
8.30562234e-01 -1.22581255e+00 -9.40134943e-01 -4.00253460e-02
2.91262627e-01 -3.86003464e-01 -5.19651771e-02 8.92060637e-01
-1.02365649e+00 7.59281397e-01 -4.15474057e-01 -5.62762320e-01
-1.42633474e+00 7.22598493e-01 6.76260516e-02 2.71618336e-01
-2.49074012e-01 2.66242921e-01 1.47425234e-01 -1.82758182e-01
-5.04615493e-02 -1.15888633e-01 -4.81854498e-01 1.12474144e-01
3.45067978e-01 -1.39168411e-01 -4.12094533e-01 -1.55867493e+00
-2.44352967e-01 5.86162210e-01 7.62993768e-02 -4.53738064e-01
1.10448813e+00 -5.05049527e-01 -2.77331740e-01 4.20899063e-01
1.31811523e+00 -5.50257452e-02 -6.48242116e-01 2.85301000e-01
-1.72462165e-02 -1.11279212e-01 -2.00347751e-01 -6.90654337e-01
-7.65088141e-01 7.16928005e-01 3.79272252e-01 -8.48898739e-02
1.09208417e+00 1.64086208e-01 1.05162311e+00 2.92959601e-01
2.65158355e-01 -9.00278807e-01 2.24872872e-01 7.17024148e-01
7.85205662e-01 -1.32774150e+00 -4.99465019e-01 -4.15014029e-01
-1.28365648e+00 1.26808035e+00 7.58384645e-01 3.19391489e-01
-1.22712553e-01 1.62660822e-01 5.94715357e-01 3.96660231e-02
-7.34543741e-01 -5.11410952e-01 3.06399137e-01 9.94576275e-01
6.44296706e-01 -9.53542367e-02 -3.54687631e-01 3.82952571e-01
-3.26417424e-02 -1.51168585e-01 9.58136380e-01 8.37643147e-01
-7.74144754e-02 -1.20685256e+00 -6.69968009e-01 -2.81678094e-03
-6.20864272e-01 -1.95617154e-01 -9.52723384e-01 9.41124737e-01
6.80832937e-02 1.15343702e+00 3.07291865e-01 -1.94245785e-01
4.25298251e-02 1.85309619e-01 7.10490704e-01 -6.98049426e-01
-3.92057955e-01 3.70949358e-01 3.25937361e-01 -2.17687428e-01
-7.21567929e-01 -8.19835246e-01 -1.37087297e+00 4.67354693e-02
-1.95717998e-03 2.78304577e-01 8.03095579e-01 9.61034536e-01
6.83302358e-02 2.35431239e-01 -5.38128952e-04 -8.38783920e-01
2.61299610e-01 -8.18364918e-01 -1.01710595e-01 5.24474561e-01
1.34731278e-01 -1.34319663e-01 -8.86556879e-02 6.24179304e-01] | [11.087691307067871, 1.2733968496322632] |
5c2afe22-b5e4-4122-a3bf-7247944285ff | joint-generative-and-contrastive-learning-for | 2012.09071 | null | https://arxiv.org/abs/2012.09071v2 | https://arxiv.org/pdf/2012.09071v2.pdf | Joint Generative and Contrastive Learning for Unsupervised Person Re-identification | Recent self-supervised contrastive learning provides an effective approach for unsupervised person re-identification (ReID) by learning invariance from different views (transformed versions) of an input. In this paper, we incorporate a Generative Adversarial Network (GAN) and a contrastive learning module into one joint training framework. While the GAN provides online data augmentation for contrastive learning, the contrastive module learns view-invariant features for generation. In this context, we propose a mesh-based view generator. Specifically, mesh projections serve as references towards generating novel views of a person. In addition, we propose a view-invariant loss to facilitate contrastive learning between original and generated views. Deviating from previous GAN-based unsupervised ReID methods involving domain adaptation, we do not rely on a labeled source dataset, which makes our method more flexible. Extensive experimental results show that our method significantly outperforms state-of-the-art methods under both, fully unsupervised and unsupervised domain adaptive settings on several large scale ReID datsets. | ['Francois Bremond', 'Antitza Dantcheva', 'Benoit Lagadec', 'Yaohui Wang', 'Hao Chen'] | 2020-12-16 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Chen_Joint_Generative_and_Contrastive_Learning_for_Unsupervised_Person_Re-Identification_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Chen_Joint_Generative_and_Contrastive_Learning_for_Unsupervised_Person_Re-Identification_CVPR_2021_paper.pdf | cvpr-2021-1 | ['unsupervised-person-re-identification'] | ['computer-vision'] | [ 3.14215332e-01 5.75432926e-02 -1.07916474e-01 -6.87878132e-01
-7.48413384e-01 -7.05896258e-01 1.14145052e+00 -3.77115577e-01
-2.93679029e-01 7.83023655e-01 4.56035793e-01 4.17948961e-01
2.44080082e-01 -9.57507074e-01 -7.26767480e-01 -7.41504669e-01
4.78670895e-01 7.94539392e-01 -4.05770570e-01 -1.51413620e-01
-7.45070502e-02 4.96460259e-01 -1.36840045e+00 -1.08938396e-01
9.31287646e-01 3.87094289e-01 -2.74963766e-01 3.17486733e-01
1.54589796e-02 4.16749179e-01 -7.04334795e-01 -9.34665918e-01
6.42446160e-01 -7.68442392e-01 -6.13899469e-01 4.37453389e-01
8.19417179e-01 -4.58433360e-01 -4.68543738e-01 8.70247364e-01
6.71312988e-01 2.50124753e-01 9.74496424e-01 -1.41745710e+00
-9.73294854e-01 3.94035637e-01 -7.26631522e-01 -8.42258409e-02
5.78595281e-01 2.03918234e-01 5.25114655e-01 -8.05208743e-01
8.00545454e-01 1.32810771e+00 5.72225332e-01 1.07637453e+00
-1.41260135e+00 -9.36551452e-01 2.04562470e-01 -7.74617866e-02
-1.30860007e+00 -5.46446860e-01 1.41284752e+00 -3.54098082e-01
1.42417401e-01 5.26990555e-02 6.67994738e-01 1.70465350e+00
-2.06327140e-01 7.62608409e-01 1.48334503e+00 -3.64904523e-01
2.08643824e-01 2.11918011e-01 -2.49074712e-01 4.35606509e-01
2.37335548e-01 1.84393093e-01 -5.84934711e-01 -1.69767350e-01
1.03549647e+00 2.38312826e-01 2.97599621e-02 -8.14733148e-01
-1.08647585e+00 6.80239975e-01 4.44199324e-01 -1.13357224e-01
-1.85425490e-01 -2.02304289e-01 3.82952929e-01 3.52245122e-01
5.81922233e-01 1.16824135e-01 9.61368158e-02 2.44107753e-01
-8.15682054e-01 4.02225405e-01 4.86049056e-01 1.13813853e+00
7.41257846e-01 2.87730098e-01 -2.97670305e-01 8.87171984e-01
2.24099517e-01 6.77131593e-01 4.31846499e-01 -6.71589434e-01
5.99970877e-01 6.69308782e-01 -1.33822232e-01 -6.70459211e-01
3.47339511e-02 -6.32606447e-01 -1.24588287e+00 2.73397267e-01
3.44765157e-01 -1.29074186e-01 -1.04282939e+00 1.97638440e+00
4.44267720e-01 2.91656077e-01 1.74250454e-01 6.07767940e-01
7.69199193e-01 1.87968329e-01 5.35521545e-02 -3.35846143e-03
1.07563543e+00 -8.83918822e-01 -5.00051916e-01 -1.69348657e-01
5.38654998e-02 -4.71779287e-01 8.44182849e-01 2.22015962e-01
-1.24008703e+00 -8.91121924e-01 -1.07973123e+00 -2.22057700e-02
-3.94204736e-01 2.37034690e-02 3.18259835e-01 7.39548385e-01
-9.51588035e-01 1.35268763e-01 -5.40769160e-01 -3.50927711e-01
6.71521842e-01 2.63418049e-01 -7.46059418e-01 -2.01932654e-01
-8.18406224e-01 5.82663774e-01 1.77727669e-01 -2.96917468e-01
-8.55214715e-01 -5.43174744e-01 -1.12367034e+00 -2.25424007e-01
1.42169833e-01 -1.28434277e+00 8.53482664e-01 -1.01265371e+00
-1.63579452e+00 1.27825999e+00 -2.09371865e-01 -4.55645651e-01
1.07984233e+00 -7.68130496e-02 -3.30432683e-01 1.08282156e-01
2.88309991e-01 8.83938313e-01 1.34487748e+00 -1.76099789e+00
-2.37454355e-01 -5.64561903e-01 8.76003504e-02 4.26287025e-01
-4.41322893e-01 -2.61115015e-01 -3.90464544e-01 -1.06793261e+00
-2.19207987e-01 -1.01142597e+00 -3.50923054e-02 4.97652255e-02
-5.22314370e-01 -1.56974539e-01 7.16719151e-01 -6.70756757e-01
5.50253510e-01 -1.93985879e+00 2.44503349e-01 1.90952078e-01
3.00610781e-01 5.47452904e-02 -1.18547566e-01 2.67334789e-01
-1.36770442e-01 -1.50175631e-01 -3.96625131e-01 -9.35426414e-01
-5.65107390e-02 1.15513936e-01 -2.01881886e-01 5.96835673e-01
1.89611182e-01 1.01003766e+00 -9.28335845e-01 -5.31847954e-01
3.20672333e-01 5.76813281e-01 -4.87919599e-01 4.63272005e-01
2.48612687e-01 1.07026207e+00 -2.42170081e-01 4.55056787e-01
9.90140140e-01 4.10287976e-02 8.08824599e-03 4.53836434e-02
2.60176361e-01 -6.87933564e-02 -9.54141200e-01 1.99566722e+00
-4.23663557e-01 1.40604973e-01 -1.49314269e-01 -9.55596864e-01
1.17324877e+00 1.78389132e-01 2.52476394e-01 -6.78743303e-01
-5.83117269e-02 -1.22123443e-01 -4.94024247e-01 5.61650768e-02
4.03037280e-01 -3.06234181e-01 -2.37573817e-01 5.81749082e-01
2.91351199e-01 1.57181263e-01 -1.67918131e-02 2.42908567e-01
7.24015713e-01 4.60118473e-01 3.02300960e-01 -4.31281291e-02
6.52076066e-01 -3.61173183e-01 6.83879018e-01 6.12660110e-01
2.21130084e-02 9.85432625e-01 4.62553762e-02 -3.88435870e-01
-1.23924994e+00 -1.52174723e+00 -3.10305413e-02 8.13071132e-01
3.08792084e-01 -1.18847042e-01 -8.48420024e-01 -1.14227545e+00
-8.24317802e-03 5.07495046e-01 -7.46484697e-01 -6.21550642e-02
-7.58914113e-01 -4.31707233e-01 5.25677204e-01 8.42583358e-01
8.04319441e-01 -7.37633228e-01 1.03199773e-01 3.55360545e-02
-1.83656380e-01 -1.03095388e+00 -8.06825221e-01 -4.15462881e-01
-7.39429712e-01 -9.06144083e-01 -1.17550933e+00 -9.51921403e-01
1.20401120e+00 4.44865227e-01 1.10373473e+00 -3.87457252e-01
-5.97422868e-02 8.01023543e-01 -2.28274971e-01 -5.16902387e-01
-5.29393077e-01 6.39984244e-03 3.14998507e-01 2.63741553e-01
2.65847325e-01 -8.70196462e-01 -6.77239954e-01 4.07047719e-01
-8.62077534e-01 2.33053982e-01 5.32292247e-01 1.02087748e+00
6.87282145e-01 -2.23278269e-01 7.60632575e-01 -1.09252548e+00
4.08598632e-01 -4.13940102e-01 -4.59037930e-01 1.95051968e-01
-6.47817969e-01 8.69120955e-02 9.06158924e-01 -4.35687482e-01
-1.51623499e+00 9.10783559e-02 1.10880360e-01 -5.92394471e-01
-4.20118690e-01 -2.22511828e-01 -6.02935970e-01 5.92399538e-02
6.82177365e-01 4.97199446e-01 1.18725650e-01 -4.72730368e-01
5.36798000e-01 4.65743631e-01 9.50940192e-01 -5.75582802e-01
1.69980085e+00 8.26301157e-01 -1.27652973e-01 -3.55514526e-01
-4.42764968e-01 -2.96508104e-01 -1.10953498e+00 -1.86538950e-01
8.67796540e-01 -1.31342959e+00 -3.25481504e-01 6.76333368e-01
-8.27669084e-01 -6.97218180e-02 -4.90247667e-01 1.30935460e-01
-6.51332021e-01 4.48888212e-01 -2.72017181e-01 -5.73350191e-01
-5.47393024e-01 -7.24576712e-01 1.17601883e+00 4.46641564e-01
-1.03101633e-01 -1.12264729e+00 2.48945549e-01 7.40895748e-01
1.66705757e-01 6.14008486e-01 4.04140413e-01 -6.79607809e-01
-3.86762470e-01 -2.59474844e-01 2.27183104e-02 4.02933687e-01
4.53762084e-01 -6.98590338e-01 -1.11881447e+00 -7.78680444e-01
-9.10768062e-02 -4.19128656e-01 7.96107829e-01 -5.51208444e-02
1.24421060e+00 -5.47965050e-01 -4.42151427e-01 8.83579671e-01
1.19942081e+00 -8.81477669e-02 6.11752391e-01 2.19202995e-01
1.04937315e+00 6.83238745e-01 3.55132639e-01 4.37808901e-01
7.41291285e-01 7.06578016e-01 5.53036630e-02 -4.05200601e-01
-2.75769681e-01 -7.47220993e-01 2.84190953e-01 4.89593238e-01
-2.96041489e-01 -2.58273363e-01 -5.47275901e-01 5.51137090e-01
-1.55843508e+00 -1.26141584e+00 4.63980466e-01 2.44838238e+00
8.21722627e-01 -6.93187956e-03 4.80974406e-01 -1.67777747e-01
9.18576241e-01 2.27037966e-01 -8.31738532e-01 3.38641815e-02
-1.38213009e-01 2.59913504e-01 3.22941124e-01 7.30719194e-02
-1.17398500e+00 7.66952038e-01 5.44846678e+00 5.41029453e-01
-7.84602404e-01 1.03038222e-01 5.47158003e-01 -1.08246602e-01
-4.90689635e-01 -2.85547346e-01 -7.12221801e-01 6.09996080e-01
3.64307165e-01 -3.73321861e-01 3.50629389e-01 9.03784990e-01
-6.47321567e-02 4.31563735e-01 -1.36180151e+00 1.26798201e+00
5.28142571e-01 -1.10684657e+00 2.93946713e-01 3.93186688e-01
1.03385198e+00 -5.41251659e-01 3.57483059e-01 2.19309106e-01
5.06334484e-01 -8.68668735e-01 5.58820307e-01 6.58605039e-01
1.12134659e+00 -1.05095339e+00 5.74961066e-01 2.02603206e-01
-1.22267413e+00 1.57674104e-01 -1.51460782e-01 2.97383696e-01
1.34408474e-01 1.49891466e-01 -6.92475140e-01 8.58965278e-01
5.15830219e-01 7.96478868e-01 -6.76644683e-01 6.19512022e-01
-3.03496987e-01 2.81697899e-01 -1.13353645e-02 5.89473844e-01
-3.44838917e-01 -2.43609831e-01 6.75421238e-01 1.00142813e+00
-6.82409247e-03 -5.52956723e-02 2.82644480e-01 1.01933503e+00
-5.62089860e-01 -1.83129653e-01 -8.56974602e-01 4.04009908e-01
6.21886015e-01 1.17472267e+00 -3.37323248e-01 -3.59359652e-01
-3.97125334e-01 1.51288509e+00 3.32966477e-01 3.72181058e-01
-6.50324106e-01 -9.13306028e-02 5.56583166e-01 2.40725935e-01
1.52292475e-01 -2.94549838e-02 -1.83047399e-01 -1.51571727e+00
2.98676699e-01 -9.20978129e-01 4.82486635e-01 -4.78155702e-01
-1.98143923e+00 3.27797532e-01 2.74844825e-01 -1.70860147e+00
-6.70853198e-01 -4.92269993e-02 -8.67363036e-01 9.61951494e-01
-1.54789042e+00 -1.91003466e+00 -5.13709664e-01 8.59008789e-01
6.70388460e-01 -6.50458276e-01 8.02487969e-01 -3.82748470e-02
-3.99448931e-01 1.23785841e+00 1.03707030e-01 5.35096824e-01
1.21533298e+00 -1.57395172e+00 7.75168061e-01 1.01584089e+00
9.57644954e-02 8.33622873e-01 3.62462133e-01 -7.60753751e-01
-1.35550976e+00 -1.38816261e+00 3.72739464e-01 -5.20287752e-01
9.74749550e-02 -5.82111418e-01 -6.90815330e-01 8.53051484e-01
3.51630449e-01 2.75117159e-02 9.55454588e-01 -1.14513576e-01
-5.80874383e-01 -2.07872123e-01 -1.52616036e+00 5.98450005e-01
1.46663415e+00 -5.28018713e-01 -6.92408621e-01 9.53704417e-02
2.70561486e-01 -4.56491351e-01 -9.13680196e-01 2.94385642e-01
6.20197296e-01 -8.94434452e-01 1.21930873e+00 -3.79347056e-01
3.10691148e-01 -2.25886285e-01 4.13957119e-01 -1.49849474e+00
-3.65729064e-01 -5.80228031e-01 -1.66682586e-01 1.77939856e+00
-1.96787596e-01 -9.16319311e-01 9.06196177e-01 5.31948209e-01
5.54477684e-02 -2.72969544e-01 -6.54816568e-01 -9.88903284e-01
2.51374602e-01 2.72342831e-01 8.39254379e-01 1.11996114e+00
-4.76465851e-01 3.98611814e-01 -5.50710440e-01 1.90080628e-01
1.31689167e+00 1.01883426e-01 1.42706931e+00 -1.35293698e+00
-3.98968518e-01 -1.34858966e-01 -5.92715323e-01 -1.02043521e+00
5.26072323e-01 -9.78977025e-01 -3.39428276e-01 -1.30095315e+00
5.89830399e-01 -3.35346609e-01 -1.45079389e-01 4.11244214e-01
-3.39478523e-01 3.74036372e-01 2.06056923e-01 4.31318909e-01
-3.00697923e-01 7.86095917e-01 1.15909040e+00 -4.62099433e-01
-1.75673798e-01 9.15473029e-02 -9.40046310e-01 6.75719440e-01
7.99247265e-01 -2.65718877e-01 -6.47065103e-01 -3.63112837e-01
-3.74399513e-01 -2.40845010e-01 6.77253366e-01 -1.05251968e+00
1.80833206e-01 6.46885857e-02 1.02422833e+00 -5.37233114e-01
3.79258722e-01 -5.41320801e-01 1.15460806e-01 9.66779962e-02
-3.68662298e-01 2.65902188e-02 -1.32141516e-01 8.13803673e-01
-6.32364079e-02 1.57493547e-01 8.86825919e-01 -1.80902883e-01
-4.78202701e-01 6.31274521e-01 9.43380594e-02 8.34788606e-02
9.02353108e-01 -4.45662856e-01 -2.92759776e-01 -5.71994543e-01
-5.53636789e-01 2.52706915e-01 1.06036007e+00 5.96270978e-01
6.83375478e-01 -1.62923658e+00 -9.47530389e-01 4.78290260e-01
4.46574032e-01 2.62089968e-01 3.31540853e-01 2.42836684e-01
8.65885057e-03 -6.02939948e-02 -4.87456888e-01 -4.93242949e-01
-1.38356102e+00 5.41071653e-01 1.05704986e-01 -2.67109483e-01
-7.04630852e-01 6.33346558e-01 6.00356698e-01 -6.89269185e-01
4.57905568e-02 4.58000600e-01 -2.78449506e-01 -7.86256716e-02
5.86472034e-01 4.55231518e-01 -3.23795676e-01 -8.54645610e-01
-1.49463221e-01 5.53993881e-01 -3.67467016e-01 -2.70200729e-01
1.31838214e+00 -1.23217002e-01 2.41305530e-01 1.89103276e-01
1.06560302e+00 2.45523065e-01 -1.50569510e+00 -4.65408325e-01
-5.52182376e-01 -4.83669192e-01 -4.90121692e-01 -5.66916227e-01
-9.02741611e-01 5.54950058e-01 7.24740684e-01 -1.99815929e-01
1.12403846e+00 4.43201810e-02 9.07379746e-01 1.37209833e-01
6.01124883e-01 -1.01385367e+00 2.99153507e-01 -3.22139896e-02
9.02605057e-01 -1.48171401e+00 1.45418540e-01 -3.92586410e-01
-6.33313835e-01 8.07189107e-01 8.89589250e-01 -1.91048592e-01
1.41821653e-01 -1.04918368e-01 4.07712273e-02 1.04464620e-01
-2.03668267e-01 -1.05991594e-01 4.41730708e-01 1.12577641e+00
1.20652273e-01 1.12809956e-01 -4.00831699e-02 3.48234445e-01
-5.67297935e-01 -2.18371853e-01 3.41644049e-01 8.50742042e-01
2.36132562e-01 -1.59917116e+00 -4.81605142e-01 3.33669722e-01
-1.24559410e-01 2.17050046e-01 -6.74489617e-01 7.50214696e-01
1.37806907e-01 7.57933676e-01 1.46326661e-01 -3.50301743e-01
1.94750011e-01 7.98476413e-02 7.68225193e-01 -7.11710215e-01
-4.24354136e-01 -1.70329884e-01 -3.01127344e-01 -1.55466795e-01
-5.98951936e-01 -9.15089905e-01 -8.45303833e-01 -3.55287403e-01
-6.68632463e-02 -1.16013430e-01 2.59698093e-01 9.70646799e-01
3.93322498e-01 2.09244907e-01 1.04634833e+00 -9.36166644e-01
-4.38041836e-01 -8.63620579e-01 -3.96030188e-01 8.67467344e-01
1.97190702e-01 -7.40982413e-01 -1.49508923e-01 4.52558875e-01] | [14.662910461425781, 0.966994047164917] |
8a80cdd4-95cc-4c01-905c-5012b7f83898 | latent-semantic-search-and-information | 1912.00180 | null | https://arxiv.org/abs/1912.00180v1 | https://arxiv.org/pdf/1912.00180v1.pdf | Latent Semantic Search and Information Extraction Architecture | The motivation, concept, design and implementation of latent semantic search for search engines have limited semantic search, entity extraction and property attribution features, have insufficient accuracy and response time of latent search, may impose privacy concerns and the search results are unavailable in offline mode for robotic search operations. The alternative suggestion involves autonomous search engine with adaptive storage consumption, configurable search scope and latent search response time with built-in options for entity extraction and property attribution available as open source platform for mobile, desktop and server solutions. The suggested architecture attempts to implement artificial general intelligence (AGI) principles as long as autonomous behaviour constrained by limited resources is concerned, and it is applied for specific task of enabling Web search for artificial agents implementing the AGI. | ['Anton Kolonin'] | 2019-11-30 | null | null | null | null | ['entity-extraction'] | ['natural-language-processing'] | [-2.22077176e-01 3.19922715e-01 -4.77968127e-01 -1.71524286e-01
1.09151542e-01 -9.20349658e-01 9.40964341e-01 -3.04115620e-02
-7.98080325e-01 8.52106333e-01 -2.71837711e-01 -3.53488594e-01
-8.99133682e-01 -7.86600232e-01 6.84688101e-04 -2.04694912e-01
-8.12295359e-04 1.02655435e+00 5.41868806e-01 3.42256278e-02
3.60807210e-01 6.81085885e-01 -1.85535991e+00 -3.41147810e-01
8.02044094e-01 1.30471301e+00 5.63050151e-01 3.83746296e-01
-2.09395573e-01 5.05660892e-01 -3.64319354e-01 -1.05001060e-02
6.39918685e-01 4.89325166e-01 -1.46446693e+00 -7.48154521e-01
-3.64831477e-01 -2.88681895e-01 -4.81756717e-01 1.25868642e+00
2.32578456e-01 1.20455630e-01 5.59420526e-01 -1.78232419e+00
-6.62179112e-01 -1.33744329e-01 5.96404016e-01 2.46751145e-01
6.41542852e-01 1.96924910e-01 6.07472837e-01 -6.60091102e-01
9.14333105e-01 1.15609634e+00 4.96349156e-01 4.83177185e-01
-6.96206450e-01 -6.44912601e-01 -2.89909780e-01 4.59592342e-01
-1.57593822e+00 -4.20890599e-01 2.29779452e-01 -1.25534654e-01
1.77133989e+00 5.90467393e-01 4.18728352e-01 8.97844613e-01
9.18311402e-02 2.70487905e-01 8.60838711e-01 -3.74718994e-01
5.43805301e-01 7.59099305e-01 3.35576802e-01 8.49334419e-01
5.20210564e-01 3.42482060e-01 -6.82258010e-01 -5.84947944e-01
4.86982137e-01 -4.32595238e-03 2.88035482e-01 -5.03667951e-01
-1.04303396e+00 7.24374413e-01 1.90631032e-01 1.33114278e-01
-7.71668613e-01 9.31575894e-02 4.54736501e-01 4.56842959e-01
5.97003251e-02 6.76476121e-01 -8.43385160e-01 -1.39020368e-01
-4.64793295e-01 4.27823484e-01 1.05474603e+00 1.41882551e+00
8.61256123e-01 -2.73849398e-01 2.46323511e-01 6.03275955e-01
7.48344183e-01 4.32965070e-01 1.19662917e+00 -1.00985610e+00
1.19084932e-01 1.08273840e+00 4.17585403e-01 -7.87417352e-01
-5.67941070e-01 2.08976105e-01 -3.09934579e-02 3.35204512e-01
-4.00730483e-02 4.30766828e-02 -1.08731806e+00 1.36393356e+00
5.13296008e-01 -7.27976263e-01 6.34576082e-01 8.03409517e-01
8.27285945e-01 4.68846381e-01 6.24720275e-01 -3.07980496e-02
1.84078944e+00 -9.01178658e-01 -7.59429514e-01 -2.66322643e-01
5.21147311e-01 -4.88889843e-01 7.33645916e-01 -3.66041660e-02
-9.45229471e-01 2.67428737e-02 -9.17799950e-01 -1.20297104e-01
-1.57660139e+00 -5.54976426e-02 1.14636958e+00 5.20402789e-01
-1.02109885e+00 2.13061064e-01 -6.40185714e-01 -1.17536724e+00
-3.75110209e-02 9.60006833e-01 -6.01913095e-01 5.25934041e-01
-1.49824655e+00 1.18458998e+00 1.06329787e+00 -3.95306647e-01
-4.69980508e-01 7.55579118e-03 -5.88583589e-01 1.76433548e-01
4.12393093e-01 -8.37667167e-01 8.24078977e-01 -4.62872416e-01
-1.58105385e+00 8.42217028e-01 1.20340362e-01 -7.60401726e-01
5.15577614e-01 1.69254199e-01 -5.93495965e-01 4.08979446e-01
4.45436329e-01 6.96831405e-01 5.67385614e-01 -4.75043982e-01
-7.25246608e-01 -6.16036057e-01 -4.94610928e-02 5.39053082e-01
-3.23460788e-01 4.68046427e-01 -4.11824405e-01 -1.75554127e-01
2.49125183e-01 -8.79756272e-01 1.14040812e-02 -3.60790826e-02
1.58764303e-01 -7.19969451e-01 1.37236226e+00 -5.90473771e-01
1.00112057e+00 -1.91164064e+00 -3.93126696e-01 4.22631979e-01
-2.59725720e-01 4.34824675e-01 2.86011934e-01 6.71584964e-01
2.70318121e-01 2.61118561e-01 5.17247736e-01 4.62291956e-01
4.73085999e-01 3.86163622e-01 3.68455835e-02 1.68600410e-01
-5.30031085e-01 9.87679303e-01 -6.38914883e-01 -9.57633615e-01
2.59652287e-01 -2.61598140e-01 -2.55441159e-01 -3.99046950e-02
-2.90353239e-01 -2.97420770e-01 -1.12719297e+00 1.25442779e+00
2.84030735e-01 -4.86677885e-01 3.05290945e-04 7.58129433e-02
-3.56327713e-01 3.06237310e-01 -1.01712477e+00 1.61650789e+00
-1.75794393e-01 3.14835399e-01 3.43867868e-01 -8.28416705e-01
7.45655298e-01 6.73102915e-01 4.52848166e-01 -8.32171261e-01
-1.74982786e-01 6.74739838e-01 -7.60088265e-01 -7.83400953e-01
6.56733632e-01 7.93265939e-01 -1.16142377e-01 2.61102170e-01
1.29162192e-01 3.03286165e-01 -5.89544773e-02 2.34183241e-02
1.32215536e+00 2.96629399e-01 4.20310706e-01 -4.43520069e-01
6.53998554e-01 6.48146152e-01 1.72913119e-01 7.13594913e-01
-4.42592800e-01 -4.50156540e-01 -4.10881341e-01 -6.47633016e-01
-1.10173357e+00 -6.60424411e-01 -1.06751874e-01 1.10153401e+00
7.12309837e-01 -3.34019303e-01 -6.48880124e-01 -7.39863694e-01
-6.93832117e-04 6.33128703e-01 4.05294122e-03 -1.39053494e-01
-7.92774186e-02 -2.73627937e-01 4.94645923e-01 2.30889656e-02
9.93486166e-01 -1.47436643e+00 -1.42119920e+00 1.48112804e-01
7.49345636e-03 -8.66251111e-01 6.97880462e-02 6.67300105e-01
-6.19240582e-01 -8.01869810e-01 -2.42074147e-01 -8.46819818e-01
5.57294369e-01 -1.65599510e-01 6.01568401e-01 1.31003603e-01
-7.36231089e-01 7.76224732e-01 -3.23610395e-01 -4.75041926e-01
-7.10433647e-02 4.27766204e-01 3.57242465e-01 -7.75367796e-01
8.35585475e-01 -4.19059098e-01 -8.41957211e-01 4.94673371e-01
-6.62426412e-01 -3.07722569e-01 4.49819356e-01 8.32983792e-01
1.32278487e-01 2.40936846e-01 5.26069880e-01 -1.66931540e-01
9.07078981e-01 -6.85236931e-01 -9.72148359e-01 5.25597692e-01
-1.76542008e+00 2.07324073e-01 3.75457928e-02 -1.21873528e-01
-1.04280770e+00 2.17033938e-01 4.03223574e-01 -2.69488007e-01
-3.35826218e-01 3.83258224e-01 -2.24411041e-01 -3.17829221e-01
6.54511631e-01 3.07225645e-01 2.68561393e-01 -6.66334510e-01
2.11816877e-01 1.12295341e+00 5.27275681e-01 -1.77027881e-01
4.59688455e-01 3.01684439e-01 -9.67834964e-02 -5.30437052e-01
1.52802691e-01 -8.19745064e-01 -2.26598233e-01 1.58354431e-01
9.66850519e-01 -6.78011656e-01 -1.06363165e+00 1.67393580e-01
-9.96759295e-01 5.31218275e-02 -2.65762955e-01 4.53927398e-01
-9.15998816e-01 1.82465270e-01 -8.54252428e-02 -1.18478107e+00
-8.46640229e-01 -9.34936225e-01 7.37028062e-01 4.19587493e-01
-4.96327966e-01 -7.50344694e-01 -3.78256857e-01 4.37643558e-01
6.46124780e-01 -4.27738726e-01 9.25814390e-01 -1.26453876e+00
-9.27025855e-01 -7.41574526e-01 -4.97132480e-01 -3.98796737e-01
1.58051521e-01 -5.76191068e-01 -6.01574719e-01 7.24445507e-02
-3.34573805e-01 -3.86130780e-01 1.11631043e-01 -1.78657889e-01
6.30025029e-01 -1.10652411e+00 -9.31530893e-01 5.92495382e-01
1.63350224e+00 7.18535841e-01 3.83382410e-01 1.10055542e+00
-2.08878607e-01 6.84084833e-01 1.00397956e+00 2.37028867e-01
8.74378681e-02 8.55865121e-01 4.62967336e-01 6.58772469e-01
1.86654583e-01 -2.91509807e-01 -1.10748187e-01 -8.43061805e-02
-5.33341244e-02 -1.58936143e-01 -9.14038181e-01 6.73843563e-01
-2.23789310e+00 -1.03097463e+00 6.46275759e-01 2.12027740e+00
4.09999967e-01 -2.22742349e-01 -2.88225189e-02 -3.78295273e-01
5.30491948e-01 -4.58818704e-01 -8.12231719e-01 -3.86272043e-01
2.30032936e-01 -2.32998341e-01 1.34591138e+00 3.86383772e-01
-8.55096102e-01 1.39655268e+00 7.34925127e+00 9.82416034e-01
-6.36836350e-01 4.08652306e-01 -2.84600705e-02 1.67742625e-01
-1.50793761e-01 3.93179178e-01 -7.97280073e-01 5.51700771e-01
8.89363825e-01 -4.08518255e-01 7.96172917e-01 1.50838792e+00
-2.55475994e-02 -1.89627558e-01 -7.62478471e-01 1.18175590e+00
-7.05090344e-01 -1.40588629e+00 -2.15061046e-02 4.82250303e-01
-1.10503189e-01 1.94252074e-01 -4.33895364e-03 1.05614271e-02
3.34320366e-01 -8.19365323e-01 6.69559240e-01 4.32954073e-01
7.92409718e-01 -3.17450374e-01 5.91897249e-01 4.86385018e-01
-1.01389790e+00 -5.12889087e-01 -3.04563046e-01 1.27431720e-01
-1.58149213e-01 -6.66107893e-01 -1.07427764e+00 3.00647646e-01
1.20088291e+00 -1.84800088e-01 -3.57733965e-01 8.48846734e-01
5.29524744e-01 -1.31973058e-01 -1.05733585e+00 -6.55001462e-01
4.05011237e-01 -3.72149527e-01 9.16690767e-01 8.04133534e-01
4.00846720e-01 1.35411561e-01 -1.03760853e-01 8.81439388e-01
5.20022035e-01 1.82305351e-01 -8.66209269e-01 -4.54978466e-01
1.06285727e+00 9.79917645e-01 -1.00133061e+00 -2.99830496e-01
3.11981775e-02 1.31864524e+00 -5.39959520e-02 4.91345555e-01
-3.61684382e-01 -5.54586947e-01 4.85062152e-01 1.60743687e-02
1.06364235e-01 -1.31891802e-01 -1.94672674e-01 -6.82204604e-01
4.09748126e-03 -4.83475566e-01 8.67038012e-01 -1.01643085e+00
-7.80679047e-01 5.47649384e-01 4.82637465e-01 -8.30511570e-01
-8.22495520e-01 -7.13368535e-01 -1.64960012e-01 9.13502395e-01
-1.01944125e+00 -1.44529235e+00 -3.00139070e-01 8.93283427e-01
6.52441502e-01 -1.11810505e+00 1.29822373e+00 -1.02105434e-03
5.30511849e-02 4.18357491e-01 1.36410072e-01 -4.57596302e-01
1.43314719e-01 -7.87598372e-01 -1.03481025e-01 3.09728622e-01
-1.74189717e-01 1.05982804e+00 6.84317470e-01 -1.03641140e+00
-1.72991097e+00 -4.26570147e-01 1.08185577e+00 -4.71519530e-01
6.13312840e-01 -8.50512609e-02 -3.01798850e-01 5.58811188e-01
1.05404653e-01 -2.60926425e-01 3.55766207e-01 -1.36227414e-01
2.58992892e-02 1.75450087e-01 -1.80868900e+00 5.12977004e-01
9.96962368e-01 -5.57386160e-01 -5.63708842e-01 7.67583132e-01
7.58373678e-01 1.30572811e-01 -7.60380983e-01 2.46341005e-01
7.44795799e-01 -4.22155946e-01 1.23971486e+00 -3.88004422e-01
-6.97107553e-01 -2.80793995e-01 -1.09468989e-01 -2.83242792e-01
-2.70890057e-01 -8.43645692e-01 -1.75565332e-01 9.26180184e-01
6.65938556e-01 -1.30221701e+00 9.74913120e-01 1.57950258e+00
2.43365288e-01 -3.27971905e-01 -1.10556245e+00 -1.02070749e+00
-7.98829079e-01 7.09006414e-02 8.01353037e-01 6.29486620e-01
4.33103591e-01 7.38588125e-02 -7.21274316e-02 1.72462180e-01
4.44695413e-01 -8.91450346e-02 3.40437353e-01 -1.34196591e+00
2.06932127e-01 -2.55214721e-01 -6.10998273e-01 -5.49925625e-01
1.07603269e-02 -6.87253892e-01 -4.33100224e-01 -1.59282744e+00
-3.25509645e-02 -8.83497357e-01 -2.76995301e-01 8.72744799e-01
7.06755340e-01 -2.23656118e-01 -1.15389258e-01 9.21847165e-01
-6.54391646e-01 1.41034588e-01 2.49826878e-01 -1.07477814e-01
-1.48268193e-01 8.82203225e-03 -1.30125403e-01 6.33905053e-01
6.77389801e-01 -4.13821310e-01 -8.49395573e-01 8.85726511e-02
3.34949136e-01 -1.24848008e-01 4.61415440e-01 -7.93202937e-01
8.48655939e-01 -3.77850801e-01 1.42371461e-01 -8.08450937e-01
5.68264544e-01 -1.60799742e+00 7.66113102e-01 5.58069348e-01
-2.67044067e-01 5.21986298e-02 -1.81878269e-01 6.64820552e-01
-1.49583116e-01 -9.29814756e-01 1.98472932e-01 -6.44774973e-01
-1.33213305e+00 3.20249289e-01 -5.43837547e-01 -8.18225563e-01
1.38243651e+00 -1.07244349e+00 -4.02164571e-02 -1.74651876e-01
-7.17865467e-01 3.97261500e-01 6.66136146e-01 5.36293209e-01
3.66436332e-01 -1.09839261e+00 2.65489966e-01 1.19731463e-01
3.07629108e-01 -2.44906262e-01 -3.06817710e-01 1.88728824e-01
-9.37994123e-01 1.21674323e+00 -4.16342854e-01 7.82377347e-02
-1.40673745e+00 1.11540008e+00 6.07073791e-02 -2.02861112e-02
-7.59823620e-01 7.33875275e-01 -2.14410499e-01 -4.22265798e-01
4.39822406e-01 3.32192302e-01 -3.03141624e-01 -2.60079145e-01
1.59110278e-01 5.42216599e-01 -2.43200913e-01 -5.14479756e-01
-8.01449120e-01 1.74185202e-01 1.58492684e-01 -3.67173910e-01
9.62320745e-01 -6.71029985e-01 -2.41539136e-01 -9.78978351e-02
1.01633549e+00 -5.10528922e-01 -4.60220695e-01 -1.25600696e-02
5.60549378e-01 -3.43332887e-01 -8.87359865e-03 -1.19595063e+00
-3.36234868e-01 -8.34165420e-03 1.22821343e+00 4.95367438e-01
7.90591776e-01 1.75905839e-01 4.58442897e-01 8.80127132e-01
9.11247432e-01 -1.88994682e+00 -6.62676632e-01 3.55671614e-01
6.40913129e-01 -1.26398098e+00 4.95954901e-02 -2.42126182e-01
-4.40752834e-01 8.98872733e-01 5.31926811e-01 4.36727732e-01
7.55751729e-01 1.78273804e-02 -1.20311528e-01 -8.11454177e-01
-8.07387471e-01 -9.74859893e-02 2.72005014e-02 8.20722044e-01
-3.50337744e-01 -1.23986378e-01 -8.42755795e-01 2.97066718e-01
-1.74977675e-01 1.49235651e-01 -4.92925107e-01 1.17780054e+00
-5.59823215e-01 -9.23939407e-01 -1.72505468e-01 2.68975973e-01
-6.05586767e-01 -2.59527899e-02 -4.86312866e-01 5.99935234e-01
3.66068780e-02 9.47060287e-01 -8.82183686e-02 -2.11318228e-02
-1.09796539e-01 6.44752264e-01 -1.51906684e-01 2.68859845e-02
-3.01431358e-01 -3.44809502e-01 6.00865126e-01 -8.74597073e-01
-1.52719855e-01 -5.67725062e-01 -1.26938438e+00 2.61422336e-01
-6.03539407e-01 7.66473830e-01 1.51396644e+00 7.96935618e-01
8.19612861e-01 -4.60227281e-01 -3.96693088e-02 -2.07233042e-01
-7.56680429e-01 -1.00968385e+00 -3.99032950e-01 1.77324742e-01
-4.43020426e-02 -8.95576537e-01 -1.89010024e-01 2.62568593e-02] | [9.635781288146973, 7.743766784667969] |
f1b9cdc9-8f76-492d-8f7a-5bfa0caac643 | optimal-positioning-of-pmus-for-fault | 2104.07211 | null | https://arxiv.org/abs/2104.07211v2 | https://arxiv.org/pdf/2104.07211v2.pdf | Optimal Positioning of PMUs for Fault Detection and Localization in Active Distribution Networks | This paper considers the problem of fault detection and localization in active distribution networks using PMUs. The proposed algorithm consists in computing a set of weighted least squares state estimates whose results are used to detect, characterize and localize the occurrence of a fault. Moreover, a criteria to minimize the number of PMUs required to correctly perform the proposed algorithm is defined. Such a criteria, based on system observability conditions, allows the design of an optimization problem to set the positions of PMUs along the grid, in order to get the desired fault localization resolution. The performances of the strategy are tested via simulations on a benchmark distribution system. | ['M. Cabiati', 'C. Bossi', 'F. Silvestro', 'G. -P. Schiapparelli', 'B. Gabriele', 'F. Conte'] | 2021-04-15 | null | null | null | null | ['fault-localization'] | ['computer-code'] | [-9.69145596e-02 -3.62571515e-02 -8.73601716e-03 1.27457559e-01
-4.38739002e-01 -4.31288421e-01 2.05630675e-01 5.47811508e-01
2.87187636e-01 1.12928128e+00 -5.17890513e-01 -1.19197540e-01
-7.65260994e-01 -7.69331217e-01 -3.77183288e-01 -9.27770793e-01
-3.98548692e-01 4.66260850e-01 1.79298595e-01 5.49636073e-02
3.63849938e-01 9.97359753e-01 -9.41762090e-01 -6.55959964e-01
1.08378983e+00 1.06271243e+00 4.63216394e-01 2.46370137e-01
5.80289602e-01 6.77860081e-01 -1.10257030e+00 8.06708574e-01
4.12806869e-02 -3.08361709e-01 -6.76067591e-01 8.33284974e-01
-3.53899479e-01 -1.17513627e-01 -5.51727377e-02 1.34613478e+00
4.49794918e-01 4.22965646e-01 8.64065051e-01 -1.24141133e+00
2.02815890e-01 3.42396498e-01 -2.79495597e-01 4.75422263e-01
5.54332137e-01 -2.31527984e-01 7.51267791e-01 -7.13567853e-01
2.18852326e-01 7.20690906e-01 1.47861630e-01 -2.46428162e-01
-1.31101239e+00 -6.23910502e-02 -2.27195174e-01 4.50086236e-01
-1.88219118e+00 -1.17452301e-01 8.13643575e-01 -4.34747666e-01
6.96626127e-01 1.45997807e-01 6.70083165e-01 1.39063457e-02
5.78408420e-01 -1.41268671e-02 8.87322068e-01 -5.44180214e-01
6.96706414e-01 -4.12624292e-02 1.98515654e-01 5.47059059e-01
7.99036384e-01 -3.92884493e-01 -6.33937716e-02 -5.01673162e-01
6.62778556e-01 -1.53437987e-01 -6.47438586e-01 -3.60467643e-01
-4.10573363e-01 6.10405266e-01 3.54182929e-01 8.10850620e-01
-7.06338584e-01 -1.59401640e-01 -7.71076307e-02 6.02698438e-02
4.87553030e-01 3.78883362e-01 6.31454075e-03 3.75319272e-01
-8.76389384e-01 -1.07052349e-01 8.85355473e-01 6.12991631e-01
7.15263665e-01 6.98415518e-01 2.46131837e-01 4.65315312e-01
3.43882740e-01 8.16795468e-01 8.67573097e-02 -4.98562574e-01
1.98384658e-01 7.14135706e-01 4.91994530e-01 -1.04603517e+00
-5.93929410e-01 -6.53740704e-01 -5.64395487e-01 2.60819405e-01
1.29655465e-01 -6.80522323e-01 -3.06706190e-01 1.27820528e+00
3.08224678e-01 4.98387277e-01 4.56684716e-02 8.11186135e-01
-3.36093426e-01 9.51247871e-01 -4.79139715e-01 -7.58841455e-01
9.69329774e-01 3.62769142e-02 -8.48405778e-01 -5.62413111e-02
4.01137203e-01 -6.90693498e-01 2.48824954e-01 4.60443914e-01
-9.93330598e-01 -3.06804627e-01 -1.36827886e+00 1.14976144e+00
6.10550940e-02 5.83173573e-01 -2.99516737e-01 3.05502355e-01
-9.33503926e-01 5.40575206e-01 -7.86538899e-01 -3.97954375e-01
-2.23575994e-01 3.80844504e-01 -3.72382067e-02 2.95447558e-01
-1.14897084e+00 1.47105563e+00 5.64321160e-01 7.25233614e-01
-1.08841550e+00 -1.62138119e-01 -7.13158965e-01 3.37749392e-01
3.80933791e-01 -1.65147915e-01 8.34078550e-01 -8.39290380e-01
-1.12183213e+00 -1.43941373e-01 -1.56572700e-01 -6.43311262e-01
2.06508726e-01 1.15743883e-01 -7.14480102e-01 4.99886245e-01
3.55474085e-01 -5.67981839e-01 7.71473467e-01 -1.09010911e+00
-7.43278265e-01 -1.75985619e-01 -2.50873357e-01 2.21123174e-01
-2.25015104e-01 -3.70860547e-01 -3.05488296e-02 -4.86027151e-02
2.79965937e-01 -6.78784311e-01 -4.80100691e-01 -7.34687686e-01
-5.24819076e-01 -2.86390573e-01 1.04969680e+00 -6.27934098e-01
1.13838112e+00 -2.02450705e+00 2.88793027e-01 9.90563750e-01
-2.26035342e-01 -8.15172717e-02 3.38508278e-01 7.97529161e-01
1.01827634e-02 -6.42544210e-01 -1.89733237e-01 2.85228401e-01
-2.14229345e-01 3.13072860e-01 -5.22614606e-02 1.04226756e+00
1.02754846e-01 -1.56062528e-01 -5.86077929e-01 -8.05055946e-02
6.17290378e-01 1.59602940e-01 -5.93773387e-02 3.63548368e-01
4.10339348e-02 2.19103754e-01 -7.31339872e-01 2.22566202e-01
5.99895537e-01 -1.95085108e-01 6.15266979e-01 -2.85761923e-01
-2.82918990e-01 -8.72956961e-03 -1.79804242e+00 6.91032410e-01
-5.60536683e-01 4.36564744e-01 3.02678049e-01 -1.48680758e+00
1.11187351e+00 7.21044302e-01 9.74725366e-01 -1.62283510e-01
2.90938228e-01 3.19727182e-01 -8.50456208e-02 -2.72664070e-01
-1.33316750e-02 3.32406521e-01 -1.39377683e-01 3.79654258e-01
1.68262467e-01 -9.30332020e-02 6.92449570e-01 5.98056167e-02
9.31878030e-01 -6.49181187e-01 5.05207539e-01 -9.09305394e-01
1.03257895e+00 -1.00028096e-02 6.37096286e-01 3.51653486e-01
2.35196888e-01 -3.49103898e-01 5.28109968e-01 -4.38996516e-02
-5.64399540e-01 -8.40689838e-01 -4.82758641e-01 -2.15194508e-01
5.32024443e-01 2.52396673e-01 -8.07010174e-01 -4.50711071e-01
1.63345873e-01 7.67928481e-01 -2.01389313e-01 -1.17291182e-01
-3.81284356e-01 -8.12099636e-01 3.69900316e-02 -1.21023366e-03
3.61790746e-01 -5.07044435e-01 -7.84469843e-01 6.09484732e-01
9.47764292e-02 -7.41408169e-01 1.27136305e-01 5.36888421e-01
-6.90014184e-01 -1.47571194e+00 -4.77659523e-01 -8.82747710e-01
1.35820365e+00 -5.22214621e-02 4.91613358e-01 2.03846008e-01
-1.57901376e-01 3.66663307e-01 -2.52026081e-01 2.80695826e-01
-2.64587760e-01 -2.03155242e-02 4.21228290e-01 1.78618416e-01
-3.02344173e-01 -4.38227206e-01 -9.72535834e-02 4.46287602e-01
-7.11129665e-01 -6.30872548e-01 4.41478640e-01 6.99752569e-01
5.02165258e-01 1.00800359e+00 9.70236599e-01 -5.66667140e-01
7.18070090e-01 -6.82286620e-01 -1.45429599e+00 4.13009942e-01
-7.10050225e-01 -1.26816124e-01 9.05582368e-01 -1.42460018e-01
-7.38791525e-01 9.53156874e-02 5.91897443e-02 7.44250566e-02
-2.22510435e-02 6.54680669e-01 -5.49800456e-01 -6.25099301e-01
3.08928668e-01 3.20234567e-01 -2.70377815e-01 -3.95975769e-01
2.96339416e-03 5.36031604e-01 3.73767436e-01 -2.17384338e-01
9.14265692e-01 9.26529691e-02 4.24429774e-01 -1.18578756e+00
-3.13454509e-01 -4.59866285e-01 -5.10999620e-01 -6.18561983e-01
2.98323601e-01 -5.78852296e-01 -8.16642165e-01 2.56107837e-01
-9.55495715e-01 6.09179139e-02 -6.54126629e-02 3.74746084e-01
-4.28582788e-01 4.32154596e-01 -5.41929066e-01 -8.94795418e-01
-2.21614555e-01 -1.15827978e+00 2.35253140e-01 1.79222599e-01
-1.77309006e-01 -1.27199280e+00 -3.16806845e-02 -2.91240185e-01
1.82544366e-01 2.83993304e-01 7.85922348e-01 -5.39833784e-01
-6.62542880e-01 -3.48196596e-01 2.42556915e-01 4.29600567e-01
4.82884467e-01 -4.09662677e-03 -4.20289710e-02 -7.03470826e-01
3.96929801e-01 3.71582717e-01 -2.11328834e-01 4.85126644e-01
3.42046678e-01 -5.40974081e-01 -5.91717780e-01 1.41687036e-01
1.93219090e+00 7.36255407e-01 1.47703186e-01 1.36138886e-01
9.47381854e-02 3.68160345e-02 8.53256464e-01 9.32301879e-01
7.56415166e-03 7.10669577e-01 5.16446114e-01 -1.15967160e-02
5.10848761e-01 3.51063907e-01 1.79424495e-01 6.14591837e-01
3.71706575e-01 -4.88521487e-01 -8.77640367e-01 5.68905175e-01
-1.77809155e+00 -5.04045963e-01 -2.84009606e-01 1.97128665e+00
-1.64559782e-02 1.66279182e-01 -1.35962322e-01 7.60800540e-01
1.02462089e+00 -1.27821624e-01 -2.55817354e-01 -2.09976345e-01
-1.15591012e-01 -2.49583237e-02 7.27815926e-01 8.18430126e-01
-8.31006825e-01 7.09765777e-02 6.54685926e+00 6.18609369e-01
-1.13568521e+00 2.06759088e-02 1.27546802e-01 4.87711757e-01
1.30053565e-01 3.37791294e-02 -6.24540746e-01 5.99775374e-01
8.97136331e-01 -5.38742661e-01 3.39168280e-01 6.76365852e-01
7.04271853e-01 -7.56417811e-01 -5.42553842e-01 4.07920748e-01
2.11985949e-02 -8.82288098e-01 -1.75007939e-01 -1.43353701e-01
9.76838529e-01 -2.56490499e-01 -5.19337654e-01 -6.36324346e-01
5.28764576e-02 -1.40499130e-01 6.45706654e-01 6.40897930e-01
-5.33136763e-02 -1.26263726e+00 9.03581917e-01 4.46972340e-01
-1.11587036e+00 -3.47065091e-01 -2.71701992e-01 -1.53444991e-01
7.33279467e-01 9.66214895e-01 -1.17871583e+00 9.49967325e-01
1.18785426e-01 5.03051698e-01 -1.40202001e-01 1.44748890e+00
-6.07139170e-01 6.64106488e-01 -4.77714866e-01 -2.76985407e-01
7.01252520e-02 -2.34065175e-01 6.87045693e-01 5.86034656e-01
5.96088707e-01 3.04275770e-02 5.79643428e-01 6.87649071e-01
4.85938996e-01 1.46678805e-01 -3.63745689e-01 4.00849253e-01
9.57990170e-01 1.10282123e+00 -9.65774298e-01 -2.94402272e-01
-9.57625434e-02 6.73889518e-01 2.27810685e-02 5.31274736e-01
-6.47522688e-01 -4.63571101e-01 3.70027214e-01 1.61941141e-01
-7.81769231e-02 -3.44762385e-01 -2.22217709e-01 -6.32341206e-01
-1.55129552e-01 -3.89954388e-01 3.58944654e-01 -5.04702806e-01
-8.53177726e-01 6.06681287e-01 8.00639242e-02 -1.27899384e+00
-5.83947361e-01 -1.79756567e-01 -8.84697020e-01 1.08349144e+00
-9.26899433e-01 -2.71913677e-01 -9.18710306e-02 6.54625714e-01
2.62202144e-01 -1.70132406e-02 5.84669590e-01 5.02935588e-01
-1.04977477e+00 -2.11742803e-01 3.46375108e-01 -1.07161932e-01
-1.03986047e-01 -1.10213542e+00 -4.63204443e-01 1.50040877e+00
-2.07628936e-01 7.73560777e-02 1.06224799e+00 -8.14840436e-01
-1.41481650e+00 -7.58735120e-01 6.31036758e-01 3.92525584e-01
6.40658200e-01 4.07294258e-02 -6.72645986e-01 6.44285083e-01
9.61871445e-02 -1.39307708e-01 -1.70179736e-02 -4.78718787e-01
7.99810052e-01 -3.54432970e-01 -1.33324361e+00 6.83852509e-02
-1.85541511e-01 -3.05607527e-01 -4.18567151e-01 4.58026886e-01
-1.58804357e-01 -2.36636266e-01 -1.05189705e+00 4.01288211e-01
-3.11388463e-01 -2.72604078e-01 6.55191183e-01 2.03317359e-01
-5.42287827e-01 -8.59308302e-01 1.41880214e-01 -1.83337235e+00
-4.32348609e-01 -3.35342705e-01 3.63155566e-02 1.17332864e+00
3.70557338e-01 -1.00542259e+00 5.14343560e-01 -9.56628248e-02
-4.56044041e-02 -4.96043921e-01 -1.34054518e+00 -6.01134181e-01
-8.26455176e-01 1.14611752e-01 4.19010073e-01 8.14186871e-01
4.97100472e-01 1.26784086e-01 -1.08821489e-01 1.08911228e+00
7.93203652e-01 2.75056716e-02 1.94325507e-01 -1.12683225e+00
-1.18281983e-01 -2.16270629e-02 -6.52038813e-01 -4.71129149e-01
2.60127902e-01 -4.02493089e-01 2.08556280e-01 -1.76545286e+00
-5.15614510e-01 -4.20703501e-01 -2.27117211e-01 3.44666570e-01
-5.88669209e-03 -7.27199689e-02 -8.40517506e-02 1.50298312e-01
-3.32189590e-01 3.03722769e-01 4.38634634e-01 -1.37559697e-03
1.12552717e-01 4.26839441e-01 1.02349684e-01 5.37488043e-01
8.54615867e-01 -3.46459359e-01 -4.78186876e-01 -5.76404259e-02
-3.48764151e-01 8.64987135e-01 1.07588805e-02 -1.44400561e+00
3.13753068e-01 -1.56243760e-02 4.02933180e-01 -7.52880991e-01
1.41772866e-01 -1.45604026e+00 5.71958244e-01 1.08709598e+00
2.52496153e-01 2.01012477e-01 -8.81497189e-02 2.20894024e-01
-3.59694004e-01 -7.04289675e-01 8.14359367e-01 5.34411550e-01
-9.08166409e-01 -1.92849919e-01 -1.10964358e+00 -5.63896120e-01
1.49321997e+00 2.75814980e-01 -5.48280664e-02 -3.35433304e-01
-8.33880663e-01 5.39855957e-01 1.28618717e-01 -1.73019707e-01
4.56939340e-01 -9.79840219e-01 -3.46602350e-01 1.56431273e-01
-3.58185172e-01 -5.26048481e-01 1.43620163e-01 7.59817004e-01
-7.66777694e-01 5.79709709e-01 -2.30977669e-01 -6.46569073e-01
-1.03389716e+00 1.86134204e-01 7.57813096e-01 -1.09182842e-01
-1.76009148e-01 3.07882391e-02 -8.11537027e-01 2.37094641e-01
-1.22060612e-01 -6.38282746e-02 -3.57239187e-01 4.66115028e-02
2.06717655e-01 8.27824533e-01 3.91338736e-01 -5.33577800e-01
-3.96208376e-01 3.52235764e-01 7.20024407e-01 -2.25000679e-02
1.21089184e+00 -4.01769787e-01 -2.82811195e-01 1.76857144e-01
7.15651274e-01 1.08334962e-02 -1.24137902e+00 4.74528894e-02
2.48960048e-01 -2.13805437e-01 1.00788452e-01 -5.41869640e-01
-1.06162202e+00 1.88670039e-01 6.41788423e-01 8.78434837e-01
1.23352003e+00 -4.14381415e-01 8.54911879e-02 2.24163517e-01
8.85183811e-01 -8.38209867e-01 -4.32351381e-01 3.07800800e-01
3.42022359e-01 -3.28791797e-01 -2.56018527e-02 -5.28095543e-01
-2.29308195e-02 1.17295802e+00 4.34898138e-01 -6.40623927e-01
7.18465567e-01 5.29266477e-01 -2.02494532e-01 -7.23386034e-02
-5.24354994e-01 -1.80915996e-01 -7.99592771e-03 2.63590991e-01
-1.83588210e-02 3.02281976e-01 -9.13616717e-01 2.05103114e-01
2.37767950e-01 -8.10352787e-02 7.06460595e-01 1.05815542e+00
-1.00166810e+00 -7.62927711e-01 -9.10708249e-01 3.16712111e-01
-3.42766076e-01 6.31238997e-01 1.18212536e-01 7.31660008e-01
1.83968082e-01 1.32578719e+00 -1.42784575e-02 9.94283482e-02
6.06566548e-01 -2.94527411e-01 2.40491256e-01 -5.17609119e-01
-1.67807624e-01 8.09487104e-02 1.64619565e-01 -8.74669328e-02
1.79242976e-02 -5.21735489e-01 -1.40998948e+00 5.87489009e-02
-8.78273427e-01 1.15090084e+00 5.04919648e-01 1.22656155e+00
-1.73041388e-01 6.56847000e-01 1.26251709e+00 -6.74270391e-01
-5.64516902e-01 -1.00988114e+00 -1.17686439e+00 -1.32696047e-01
-3.38370237e-03 -8.56659889e-01 -5.94018340e-01 -5.36038816e-01] | [5.951407432556152, 2.5444555282592773] |
37614788-cb3d-49b2-a140-4ad025acd5e3 | ff2-a-feature-fusion-two-stream-framework-for | 2211.04699 | null | https://arxiv.org/abs/2211.04699v1 | https://arxiv.org/pdf/2211.04699v1.pdf | FF2: A Feature Fusion Two-Stream Framework for Punctuation Restoration | To accomplish punctuation restoration, most existing methods focus on introducing extra information (e.g., part-of-speech) or addressing the class imbalance problem. Recently, large-scale transformer-based pre-trained language models (PLMS) have been utilized widely and obtained remarkable success. However, the PLMS are trained on the large dataset with marks, which may not fit well with the small dataset without marks, causing the convergence to be not ideal. In this study, we propose a Feature Fusion two-stream framework (FF2) to bridge the gap. Specifically, one stream leverages a pre-trained language model to capture the semantic feature, while another auxiliary module captures the feature at hand. We also modify the computation of multi-head attention to encourage communication among heads. Then, two features with different perspectives are aggregated to fuse information and enhance context awareness. Without additional data, the experimental results on the popular benchmark IWSLT demonstrate that FF2 achieves new SOTA performance, which verifies that our approach is effective. | ['Mengqi Zhang', 'Lifeng Shi', 'Hao Zhang', 'Yao Zhao', 'Kebin Fang', 'Yangjun Wu'] | 2022-11-09 | null | null | null | null | ['punctuation-restoration'] | ['natural-language-processing'] | [ 2.14781836e-01 -1.54522419e-01 -3.43811691e-01 -5.90498865e-01
-1.02276123e+00 -9.09206942e-02 4.63573724e-01 1.51267111e-01
-3.71979415e-01 5.39938033e-01 5.99274695e-01 -7.41707832e-02
2.04968244e-01 -5.60128629e-01 -6.20497763e-01 -7.86103666e-01
5.12522757e-01 -1.09833842e-02 3.43417108e-01 -2.49042958e-01
1.28476292e-01 -2.04082191e-01 -1.45129633e+00 3.71718407e-01
1.27173936e+00 9.67195153e-01 5.22410154e-01 -1.53878972e-01
-6.19958520e-01 7.07979858e-01 -6.73135638e-01 -2.44024143e-01
7.40257800e-02 -2.43369654e-01 -3.71799290e-01 1.42052382e-01
6.01052195e-02 -1.78613141e-01 -3.77278984e-01 9.89450157e-01
6.89144313e-01 6.58140108e-02 -4.55410779e-02 -1.05754960e+00
-6.06776595e-01 9.64189708e-01 -7.21900046e-01 1.60926849e-01
2.14590177e-01 1.07642345e-01 1.09695518e+00 -1.21404052e+00
8.50072205e-02 1.30137646e+00 4.36669290e-01 3.61677706e-01
-8.11065018e-01 -9.23481703e-01 7.11921632e-01 4.58685666e-01
-1.23136830e+00 -4.59025145e-01 9.59474266e-01 1.82423368e-01
6.65259063e-01 2.67373294e-01 4.77462977e-01 1.06921506e+00
-1.32157728e-01 1.34074473e+00 1.15807021e+00 -3.68432969e-01
4.68249582e-02 2.24502131e-01 2.39824638e-01 3.97946388e-01
1.60489623e-02 -3.78555357e-01 -7.28797853e-01 6.08007535e-02
3.43520105e-01 4.15858418e-01 -4.27971780e-01 -4.52084094e-02
-1.18651676e+00 5.83046556e-01 3.48203182e-01 3.88160497e-01
-4.13634092e-01 -3.22132230e-01 5.60215831e-01 6.07797727e-02
5.47495902e-01 7.28100985e-02 -5.21008313e-01 -1.61514580e-01
-9.78691697e-01 -9.90373641e-02 2.84360528e-01 9.64063108e-01
8.83251369e-01 8.84528458e-02 -6.22298896e-01 1.06909370e+00
3.43578458e-01 3.52312714e-01 7.35848248e-01 -4.98705834e-01
7.07610488e-01 8.59131753e-01 -2.76716202e-01 -6.63232803e-01
-1.09547891e-01 -8.75337243e-01 -8.26670170e-01 -5.91300309e-01
4.90177982e-02 5.02486788e-02 -9.82777476e-01 1.85060298e+00
2.62038440e-01 4.81530041e-01 -3.59138101e-02 9.26490128e-01
6.92528129e-01 8.40469480e-01 1.81979254e-01 -3.37894738e-01
1.40875721e+00 -1.25804245e+00 -9.99724150e-01 -5.32182813e-01
6.32530272e-01 -9.36975539e-01 1.55599511e+00 3.04752976e-01
-8.68037343e-01 -5.13705254e-01 -1.00390303e+00 -1.71486989e-01
-2.07166508e-01 1.92185447e-01 4.57018375e-01 4.15416032e-01
-7.25159585e-01 2.50737906e-01 -7.30778456e-01 -1.42919928e-01
4.07991976e-01 6.07520789e-02 -4.48096469e-02 -4.15333956e-01
-1.49840403e+00 5.91329336e-01 2.41109654e-01 2.58279860e-01
-6.34703875e-01 -7.34863222e-01 -7.79984951e-01 2.99224854e-01
5.02466202e-01 -3.48484576e-01 1.23882103e+00 -9.32394266e-01
-1.42302394e+00 4.82718974e-01 -6.22593164e-01 -2.41769806e-01
2.97875106e-01 -3.95900577e-01 -4.82332766e-01 -1.03626058e-01
1.32842585e-01 4.53780681e-01 8.10689926e-01 -1.23038769e+00
-8.21011841e-01 -2.74568111e-01 9.69945937e-02 4.33218092e-01
-9.48046565e-01 6.03950322e-02 -8.62583995e-01 -9.07557666e-01
1.20091721e-01 -5.29487014e-01 -1.45981282e-01 -3.35663199e-01
-5.92108548e-01 -3.53381425e-01 8.18666160e-01 -7.99188256e-01
1.83852661e+00 -2.22971869e+00 -2.38150865e-01 8.31309631e-02
5.68649396e-02 5.51350057e-01 -2.31783628e-01 3.71345669e-01
-1.47761106e-01 1.01627432e-01 -3.14541459e-01 -7.82625854e-01
6.72324141e-03 2.87520736e-01 -4.40444589e-01 9.74243507e-02
2.50682116e-01 8.65979612e-01 -7.75815427e-01 -6.32790148e-01
1.39386982e-01 4.15681392e-01 -6.12271965e-01 1.62682712e-01
-1.14736371e-02 4.36814398e-01 -7.65384376e-01 5.72558105e-01
7.67389715e-01 -3.49711090e-01 -6.98737875e-02 -3.28597426e-01
-2.20138445e-01 6.59267306e-01 -1.02298689e+00 1.93473196e+00
-5.70028484e-01 7.03426823e-02 -8.16593915e-02 -1.05233645e+00
8.45256984e-01 3.07406396e-01 4.70179677e-01 -9.70253110e-01
-3.47389258e-03 1.40427306e-01 -4.18092720e-02 -3.79561514e-01
5.38525403e-01 -1.10387705e-01 2.01340839e-02 2.72274852e-01
-9.15749669e-02 3.31583053e-01 -3.57999839e-02 2.80853122e-01
1.03345799e+00 1.69763081e-02 1.00252226e-01 -1.35559723e-01
6.91939414e-01 -2.70799249e-01 1.16857052e+00 4.61997271e-01
-2.13508204e-01 7.53005981e-01 2.47144207e-01 2.97787376e-02
-7.44651794e-01 -7.56478131e-01 1.11199334e-01 1.15307498e+00
3.72165471e-01 -7.13257134e-01 -9.04897392e-01 -7.37123489e-01
-2.79650331e-01 7.29165912e-01 -3.53475809e-01 -4.74957496e-01
-7.03134298e-01 -9.24586535e-01 3.56174737e-01 5.81872761e-01
6.61696017e-01 -8.72667074e-01 -2.42598683e-01 4.06229049e-01
-5.61784983e-01 -1.05752778e+00 -8.40813875e-01 1.09841757e-01
-6.32064283e-01 -5.88707030e-01 -7.43871570e-01 -8.54598999e-01
6.01815820e-01 6.74699843e-01 7.50352323e-01 1.32766932e-01
1.65365353e-01 -1.70122962e-02 -6.68401957e-01 -2.91651517e-01
1.24797158e-01 3.23639095e-01 -9.22625065e-02 3.55251819e-01
5.95694602e-01 -6.99467421e-01 -6.27907097e-01 3.39322239e-01
-9.53453004e-01 1.34728566e-01 8.69410396e-01 8.76443326e-01
5.09468317e-01 -2.69069895e-02 8.43738079e-01 -5.93502462e-01
4.33695257e-01 -4.66923237e-01 -9.32599232e-03 3.64428014e-01
-6.00578368e-01 -5.28093055e-02 8.20875823e-01 -5.92342913e-01
-1.23261452e+00 -3.42933297e-01 -2.79311299e-01 -4.60336477e-01
1.24572888e-01 6.40861452e-01 -8.16470325e-01 3.77310276e-01
-1.11086331e-01 6.43522143e-01 -2.44892627e-01 -7.28058934e-01
6.57882392e-02 9.35847044e-01 2.81260490e-01 -5.76698840e-01
8.10020447e-01 3.41330707e-01 -5.70443451e-01 -4.08661932e-01
-1.20179141e+00 -5.68851888e-01 -1.65630654e-01 8.05234388e-02
4.37201560e-01 -1.22517371e+00 -3.07531804e-01 6.29113019e-01
-9.52352464e-01 -1.78408414e-01 -2.68739879e-01 4.07619655e-01
1.56700164e-02 4.17658657e-01 -6.55396700e-01 -7.52726972e-01
-5.48203945e-01 -1.16186833e+00 1.14797437e+00 5.48398197e-01
1.91766798e-01 -5.84024787e-01 -3.63860905e-01 4.19925511e-01
6.60147309e-01 -4.60835278e-01 9.26177382e-01 -8.36314261e-01
-6.34532809e-01 -6.70240521e-02 -3.11828464e-01 4.54726040e-01
4.64701414e-01 -2.69871175e-01 -1.07224417e+00 -2.67651886e-01
2.52985582e-02 -2.04569086e-01 9.38922584e-01 9.16279256e-02
1.46503472e+00 -1.17422655e-01 -3.15056860e-01 3.98725361e-01
9.83831763e-01 1.12070359e-01 5.64267337e-01 3.15831423e-01
7.78142154e-01 4.99463588e-01 8.42877209e-01 5.12741864e-01
9.76167500e-01 6.01087511e-01 1.72909394e-01 -2.33866632e-01
-2.29435220e-01 -5.32112002e-01 5.83173811e-01 1.46394885e+00
4.12762314e-01 -1.62803710e-01 -7.39089429e-01 5.95532238e-01
-1.78901076e+00 -6.11111581e-01 1.06587127e-01 2.11665773e+00
1.08226740e+00 4.85089689e-01 -1.98396832e-01 3.91745508e-01
9.28978086e-01 2.91182339e-01 -6.77371502e-01 2.52960920e-01
-3.87596995e-01 1.62571892e-01 1.31383255e-01 2.53194958e-01
-8.92114520e-01 1.00077212e+00 5.09212303e+00 1.39697087e+00
-1.39760303e+00 3.28699976e-01 5.99708915e-01 -8.34793895e-02
-7.49006748e-01 2.30897456e-01 -9.33909655e-01 8.83702755e-01
6.61600769e-01 -3.19974422e-01 2.51071066e-01 4.55345213e-01
4.71819162e-01 8.49440470e-02 -5.61315238e-01 8.79374623e-01
1.86453179e-01 -8.88040960e-01 2.09927574e-01 -7.22889677e-02
3.48674595e-01 -3.08233425e-02 1.09490156e-01 6.85193479e-01
-2.31991876e-02 -5.83541453e-01 8.02692533e-01 4.11143631e-01
4.55821872e-01 -8.16487789e-01 6.08743489e-01 5.70249856e-01
-1.36822391e+00 -7.54278898e-02 -2.25004345e-01 3.99631262e-02
3.22401911e-01 9.95343387e-01 -4.80990678e-01 8.43868315e-01
6.21725738e-01 7.92381048e-01 -6.57534719e-01 1.12206614e+00
-3.34674388e-01 1.03114772e+00 -4.73525673e-01 1.03380933e-01
2.96626866e-01 -4.67091240e-02 4.80611056e-01 9.98407543e-01
4.16933775e-01 4.50511649e-03 4.52218920e-01 5.65557599e-01
-1.53658479e-01 4.40549314e-01 1.01147899e-02 1.01624742e-01
7.06950665e-01 1.46140182e+00 -3.60176057e-01 -5.17966866e-01
-7.69059122e-01 7.77419150e-01 2.29196966e-01 3.30455095e-01
-8.94630790e-01 -4.65248227e-01 7.23700821e-01 3.82513963e-02
3.17011327e-01 4.88871560e-02 -3.06668311e-01 -1.54956961e+00
3.77994001e-01 -9.49584305e-01 3.21477622e-01 -7.20262706e-01
-1.25404346e+00 5.27287602e-01 -3.05540562e-01 -1.39749134e+00
2.33503103e-01 -2.88373325e-02 -8.39565039e-01 8.34523797e-01
-2.10615754e+00 -1.34877348e+00 -2.70451576e-01 5.12798131e-01
8.46278310e-01 1.02494352e-01 4.20275658e-01 8.44304740e-01
-1.12189448e+00 1.04423249e+00 -7.18917698e-02 7.67181814e-02
1.09275544e+00 -9.39005971e-01 2.14674190e-01 1.02838600e+00
-1.33418083e-01 8.81654024e-01 5.08843660e-01 -5.61166406e-01
-1.41492081e+00 -1.19435525e+00 9.91514027e-01 -3.67266797e-02
4.71350998e-01 -4.89801168e-01 -1.41599965e+00 5.45118213e-01
2.65140295e-01 6.10806420e-02 5.81729293e-01 3.36837098e-02
-3.70477319e-01 -4.26605731e-01 -8.22947025e-01 5.86111367e-01
1.16007173e+00 -4.16145384e-01 -7.01971710e-01 1.65812653e-02
1.13847935e+00 -3.30059439e-01 -6.42589331e-01 6.01851106e-01
1.81833565e-01 -6.73681617e-01 7.41130114e-01 -4.47705150e-01
4.34226274e-01 -4.51192796e-01 -2.03719243e-01 -1.35219955e+00
-1.19115129e-01 -6.75011933e-01 -2.94111744e-02 1.86898339e+00
3.50145042e-01 -7.67737150e-01 5.67790449e-01 3.83533090e-01
-5.52846789e-01 -8.70999455e-01 -9.05353725e-01 -6.18273318e-01
-4.78538610e-02 -4.12478209e-01 9.78004813e-01 9.98483360e-01
3.93437110e-02 5.07591188e-01 -4.78877902e-01 9.30601731e-02
3.63991082e-01 2.99408466e-01 5.19657552e-01 -9.66153979e-01
-9.06176269e-02 -2.95316011e-01 1.54562712e-01 -1.41241598e+00
2.75278568e-01 -9.44603264e-01 9.48791020e-03 -1.42601752e+00
4.12163138e-01 -6.75095379e-01 -7.24586606e-01 8.32439482e-01
-8.34354579e-01 -7.75621682e-02 1.38411298e-01 2.22279698e-01
-7.97804773e-01 1.08961272e+00 1.24586618e+00 -1.17005721e-01
-8.40162933e-02 -9.36300009e-02 -1.16145182e+00 6.86132550e-01
7.76090741e-01 -4.22254324e-01 -3.67511868e-01 -6.46909952e-01
-7.84460977e-02 -3.19659978e-01 5.09622023e-02 -8.60728383e-01
4.21724528e-01 -3.72693092e-02 2.10782707e-01 -7.15633571e-01
3.00504535e-01 -6.89494908e-01 -3.79221320e-01 1.04399681e-01
-3.12138289e-01 -7.07610697e-03 6.69760555e-02 5.28291464e-01
-5.49825251e-01 2.71409899e-02 5.79899788e-01 9.61334854e-02
-6.23042822e-01 4.94764626e-01 -9.30105709e-03 1.48065209e-01
6.79740667e-01 -7.52022956e-03 -3.30212742e-01 -3.28936368e-01
-1.88319907e-01 6.80729866e-01 4.01427060e-01 6.63267434e-01
3.80497128e-01 -1.50018930e+00 -7.13049769e-01 3.02490115e-01
3.32107157e-01 1.87212989e-01 4.61759210e-01 1.09715700e+00
3.02253187e-01 3.08473349e-01 3.17470014e-01 -5.04339039e-01
-1.06293201e+00 5.15899539e-01 -2.02980433e-02 -5.12976289e-01
-5.92460215e-01 6.80432975e-01 3.70566100e-01 -4.54091161e-01
2.77596563e-01 -3.68647188e-01 -2.83488601e-01 1.88909948e-01
6.86693192e-01 -8.98409337e-02 1.60205975e-01 -5.55153608e-01
-4.33541328e-01 4.92520809e-01 -4.46052551e-01 3.69262770e-02
1.37883079e+00 -5.23146093e-01 -7.57537410e-02 3.63088489e-01
1.07370877e+00 2.64763653e-01 -1.21797299e+00 -9.15979683e-01
5.89317232e-02 -4.05900538e-01 2.15388015e-01 -6.05160296e-01
-1.24439085e+00 1.12805426e+00 2.83491105e-01 -3.59428562e-02
1.53303933e+00 -2.02468798e-01 1.40444839e+00 1.06848419e-01
3.06158781e-01 -1.16609347e+00 1.96232453e-01 6.16713703e-01
5.48171461e-01 -1.07255208e+00 -2.81230003e-01 -5.20563662e-01
-7.42144048e-01 6.48261309e-01 7.54648387e-01 2.02874139e-01
4.39932168e-01 3.33078593e-01 1.34383187e-01 3.82433355e-01
-9.09670651e-01 -3.48326951e-01 1.07870989e-01 1.34898156e-01
3.43504101e-01 -1.43007353e-01 -5.08743882e-01 1.23003507e+00
6.73591509e-04 -5.48437908e-02 2.79328167e-01 1.02987528e+00
-4.70748842e-01 -1.33717072e+00 -1.75573215e-01 5.81216514e-01
-4.90687430e-01 -5.27717471e-01 4.92664501e-02 3.15772802e-01
1.00733504e-01 9.51879978e-01 -1.56610847e-01 -3.59572053e-01
4.44814324e-01 2.15432778e-01 2.36733072e-02 -6.01018488e-01
-6.51908278e-01 3.93448055e-01 -1.56631947e-01 -5.18520713e-01
-3.03966880e-01 -6.27427578e-01 -1.28284621e+00 -2.55478807e-02
-5.27644575e-01 2.94493854e-01 3.76883805e-01 1.11198306e+00
6.08577251e-01 7.78961062e-01 9.68826056e-01 -5.38847029e-01
-5.87010682e-01 -1.14924622e+00 -4.71562117e-01 1.93839848e-01
2.93584794e-01 -4.82574105e-01 -3.78007859e-01 -1.51020199e-01] | [11.669296264648438, 5.617599010467529] |
00cefbbd-1920-4eea-80c6-9c3809f67852 | rapid-face-mask-detection-and-person | 2112.09951 | null | https://arxiv.org/abs/2112.09951v1 | https://arxiv.org/pdf/2112.09951v1.pdf | Rapid Face Mask Detection and Person Identification Model based on Deep Neural Networks | As Covid-19 has been constantly getting mutated and in three or four months a new variant gets introduced to us and it comes with more deadly problems. The things that prevent us from getting Covid is getting vaccinated and wearing a face mask. In this paper, we have implemented a new Face Mask Detection and Person Recognition model named Insight face which is based on SoftMax loss classification algorithm Arc Face loss and names it as RFMPI-DNN(Rapid Face Detection and Peron Identification Model based on Deep Neural Networks) to detect face mask and person identity rapidly as compared to other models available. To compare our new model, we have used previous MobileNet_V2 model and face recognition module for effective comparison on the basis of time. The proposed model implemented in the system has outperformed the model compared in this paper in every aspect | ['GhufranUllah', 'Mohd. Belal', 'Abdullah Ahmad Khan'] | 2021-12-18 | null | null | null | null | ['person-identification', 'person-recognition'] | ['computer-vision', 'computer-vision'] | [-3.24867219e-02 5.52953184e-02 2.31076419e-01 -4.84204620e-01
2.08199084e-01 -3.59603703e-01 3.55955780e-01 -8.17611635e-01
-6.21657372e-01 9.00506079e-01 3.34563777e-02 -6.60365447e-02
6.30397256e-03 -7.56321251e-01 -3.77287745e-01 -2.61178941e-01
1.66093141e-01 4.90047485e-01 -4.52909619e-02 -2.43478611e-01
-1.21161044e-01 6.27877593e-01 -1.63112283e+00 1.86672226e-01
3.72718692e-01 7.82686114e-01 6.58319518e-02 6.52159512e-01
1.52701914e-01 4.29410398e-01 -6.25915945e-01 -4.57558602e-01
5.71064711e-01 -2.37479851e-01 -5.32104731e-01 -5.08370161e-01
5.86101532e-01 -6.23083472e-01 -2.54959494e-01 8.76119256e-01
1.02496994e+00 -1.01133399e-01 7.15644836e-01 -1.20965731e+00
-5.09736001e-01 2.26065159e-01 -8.12530279e-01 2.89672405e-01
3.86444896e-01 2.01432109e-02 -1.67662621e-01 -6.37589276e-01
5.33528924e-01 1.72004354e+00 1.39623761e+00 1.02471876e+00
-8.29243958e-01 -1.15833950e+00 -3.74975473e-01 1.96208552e-01
-1.66479826e+00 -6.08887434e-01 3.16371769e-01 -3.87796134e-01
1.34577966e+00 2.79935747e-01 3.41548085e-01 1.08094168e+00
1.62426621e-01 2.80583888e-01 9.94862616e-01 -2.75934100e-01
-3.86426121e-01 7.04409957e-01 2.22506255e-01 9.86196935e-01
2.71801472e-01 3.76270503e-01 -8.07475299e-02 1.52584761e-01
6.17996275e-01 1.30108327e-01 1.00144602e-01 5.57480574e-01
-8.53309333e-02 6.56925201e-01 1.72598988e-01 5.51725328e-01
-4.01403815e-01 2.56394893e-01 4.63679701e-01 2.54670411e-01
1.53304994e-01 -2.44410604e-01 -4.28039551e-01 4.54697944e-02
-1.09366179e+00 1.56720370e-01 6.27130508e-01 4.41280246e-01
3.51362437e-01 3.50042790e-01 -2.51536757e-01 9.09684896e-01
5.21307409e-01 5.47143579e-01 6.33091152e-01 -4.36695307e-01
-1.02520145e-01 7.08065093e-01 -2.19447941e-01 -1.04422092e+00
-5.73792458e-01 -3.47393781e-01 -7.28099167e-01 6.68913305e-01
1.52582124e-01 -5.36991775e-01 -1.28252482e+00 1.73880875e+00
1.27099618e-01 5.83428442e-01 -1.26081824e-01 6.30319834e-01
1.15115142e+00 5.29466867e-01 5.06451353e-02 2.03665331e-01
1.42584264e+00 -5.76191962e-01 -6.76331162e-01 2.85002917e-01
9.63887572e-02 -7.39773989e-01 4.48360860e-01 5.31905353e-01
-6.98184431e-01 -8.90141070e-01 -1.15996802e+00 2.00916469e-01
-8.01282465e-01 1.24290355e-01 4.20379847e-01 1.78793323e+00
-1.46248984e+00 4.24175948e-01 -4.15191531e-01 -7.78745055e-01
7.64629364e-01 1.08268011e+00 -5.32387137e-01 1.33610412e-01
-9.52052772e-01 1.07570684e+00 2.69277215e-01 1.77792042e-01
-1.15379131e+00 -3.10218513e-01 -5.93022883e-01 -5.98873794e-02
5.33161424e-02 -6.33560777e-01 7.94065893e-01 -1.28884029e+00
-1.54448998e+00 1.05516517e+00 -2.16493711e-01 -5.34702420e-01
5.47142386e-01 -1.23660773e-01 -8.42077255e-01 -3.02318603e-01
-4.43275928e-01 9.37549531e-01 1.08161676e+00 -7.82302916e-01
-4.72894669e-01 -7.05122411e-01 -1.48352325e-01 -1.09527864e-01
-2.40534171e-01 5.78873456e-01 -8.74330476e-02 -5.05971134e-01
-5.45925498e-01 -6.35197401e-01 3.98844630e-01 -6.33822335e-03
-1.76522210e-01 -2.44096741e-01 1.59730577e+00 -1.20440543e+00
9.66053188e-01 -2.08120513e+00 -3.62385929e-01 2.59390593e-01
5.18686846e-02 1.11354876e+00 -1.32021839e-02 -9.79370400e-02
-2.59283274e-01 2.21371382e-01 -1.15674563e-01 -4.82108712e-01
-1.71648324e-01 2.13674262e-01 3.10665339e-01 5.44755340e-01
-9.26081687e-02 6.04756057e-01 -5.92118651e-02 -3.60685170e-01
2.24525362e-01 1.14861584e+00 -3.69869530e-01 9.73462593e-03
1.32956356e-01 2.64705479e-01 5.14231948e-03 7.87078083e-01
1.35698986e+00 4.77942675e-01 -3.19491029e-01 -1.09290920e-01
1.11138776e-01 -5.08499026e-01 -1.24153137e+00 1.12435293e+00
-2.00597018e-01 6.75363600e-01 4.07975584e-01 -1.02021992e+00
8.32088292e-01 6.85932934e-01 8.07404071e-02 -4.32976305e-01
7.16566682e-01 -1.68433517e-01 3.03258817e-03 -8.67778420e-01
6.54725507e-02 -1.95284173e-01 6.36021197e-01 8.01964253e-02
3.92460436e-01 8.46444845e-01 -3.24447989e-01 -2.85324901e-01
8.52543354e-01 1.88134059e-01 -1.42943889e-01 -2.47948185e-01
1.03342509e+00 -4.95982885e-01 5.40400386e-01 5.20952344e-01
-6.74174368e-01 3.08501214e-01 2.29497105e-02 -5.06258488e-01
-6.10704899e-01 -9.15499866e-01 -2.14364901e-01 7.58259535e-01
-4.54560488e-01 1.60929814e-01 -1.10497093e+00 -6.98989868e-01
1.94762051e-01 2.24165201e-01 -6.87843740e-01 1.05996944e-01
-5.65492034e-01 -6.97455227e-01 1.17611778e+00 3.78435642e-01
1.00040269e+00 -1.21420479e+00 -2.38123313e-01 -4.06292602e-02
5.09273171e-01 -6.87381864e-01 -2.21654654e-01 -1.60954475e-01
-3.49668473e-01 -9.99984741e-01 -8.19797993e-01 -1.01013064e+00
4.13835078e-01 -2.02175200e-01 6.47914350e-01 1.68534949e-01
-9.60834444e-01 1.81428403e-01 -8.17378759e-02 -1.03888905e+00
-2.37386644e-01 -1.46872342e-01 4.22041506e-01 8.06968287e-02
7.92356431e-01 -4.97679919e-01 -6.80386722e-01 -1.35912180e-01
-5.52233338e-01 -4.65938330e-01 3.12582672e-01 3.96364301e-01
-1.59025922e-01 2.67541319e-01 8.54539275e-01 -1.05459130e+00
6.05020404e-01 -4.95261669e-01 -7.81957448e-01 1.67941526e-01
-6.55345619e-01 -3.14792633e-01 1.85116813e-01 -6.47434732e-03
-1.22780824e+00 2.06172541e-02 -6.56462312e-01 -2.83583373e-01
-3.17213863e-01 -1.32437930e-01 -1.58064649e-01 -4.25927281e-01
4.39516872e-01 8.75855144e-03 4.11189824e-01 -8.26671958e-01
-8.39425325e-02 1.17457128e+00 3.03282291e-01 2.07039610e-01
6.17661655e-01 5.24914801e-01 -1.88365355e-01 -1.11537743e+00
-2.69362237e-03 -3.08855683e-01 -3.31922084e-01 -4.39300179e-01
1.19770396e+00 -9.30778623e-01 -1.43830538e+00 1.13553011e+00
-1.11952043e+00 2.91163057e-01 3.78482848e-01 4.18332189e-01
1.17710426e-01 1.05940811e-01 -6.22435212e-01 -1.27297187e+00
-7.79521406e-01 -8.49477649e-01 4.22686219e-01 7.14237392e-01
1.83277763e-02 -8.04547250e-01 9.58998129e-02 2.65746444e-01
1.03567660e+00 3.31580639e-01 3.63639623e-01 -4.54803646e-01
-1.89477444e-01 -2.92500019e-01 -3.88853014e-01 8.79214644e-01
3.49783301e-01 -1.49753287e-01 -1.45741343e+00 -3.82207602e-01
2.56096363e-01 -2.69351453e-02 1.09288859e+00 5.00584722e-01
1.07536578e+00 -2.94280648e-01 -6.21480465e-01 7.83863425e-01
1.86106753e+00 6.10011101e-01 7.98407376e-01 -2.20884383e-02
5.85947692e-01 3.93337637e-01 -2.38377556e-01 6.94290251e-02
-4.08898219e-02 6.03364646e-01 3.01535934e-01 -1.83296084e-01
-2.86557555e-01 6.22421280e-02 4.56248760e-01 2.96127424e-02
-2.67508626e-01 -2.76206344e-01 -5.74108183e-01 2.65814990e-01
-1.31079555e+00 -1.27640820e+00 2.37013116e-01 1.85460460e+00
4.77655679e-01 -6.97652102e-02 4.96231377e-01 -3.79858054e-02
8.57462943e-01 -2.66600728e-01 -3.41774940e-01 -9.06389534e-01
4.83941473e-02 8.14532518e-01 5.86850703e-01 6.59278095e-01
-1.04149413e+00 8.86958897e-01 6.81733274e+00 7.09706724e-01
-1.58372128e+00 5.50735354e-01 6.83880866e-01 -2.15945557e-01
4.30623740e-01 -6.47379220e-01 -1.23747742e+00 6.97509646e-01
8.68292212e-01 3.12969685e-01 6.13798738e-01 7.62288570e-01
1.21347919e-01 9.39420760e-02 -6.75498128e-01 1.23713708e+00
4.23404902e-01 -1.01785839e+00 -6.26636744e-02 -9.37205553e-02
2.01238349e-01 7.33272582e-02 2.85720706e-01 4.59478766e-01
7.68533871e-02 -1.69775105e+00 1.18722424e-01 6.51539385e-01
1.00574172e+00 -1.03016114e+00 1.12047696e+00 9.24537703e-02
-8.84979010e-01 -2.01839119e-01 -3.95837963e-01 -2.04149991e-01
2.94489991e-02 3.80684808e-02 -1.11895895e+00 2.50530481e-01
9.58351731e-01 1.57875657e-01 -3.73863608e-01 8.36651027e-01
4.41324741e-01 7.32943058e-01 -5.75457990e-01 -1.67556498e-02
4.71172221e-02 -3.89851183e-02 4.11690474e-01 1.31976044e+00
4.62294519e-01 -1.26029342e-01 -3.46959680e-01 9.32617366e-01
-1.43291607e-01 -7.13487044e-02 -7.79938102e-01 3.14013571e-01
2.74482429e-01 1.25543845e+00 -2.83318937e-01 -1.65903956e-01
-5.85288227e-01 1.03710687e+00 -1.37986645e-01 1.88671663e-01
-8.95692170e-01 -4.15847778e-01 7.71614492e-01 3.71125877e-01
3.55350792e-01 2.84771353e-01 8.00375346e-05 -4.02616233e-01
-2.47969091e-01 -6.94354653e-01 3.60983729e-01 -5.59766889e-01
-1.15069985e+00 8.44840169e-01 -1.48387864e-01 -5.33398569e-01
1.96389183e-02 -1.16130650e+00 -5.74154735e-01 1.29958022e+00
-1.12584460e+00 -1.35968566e+00 -1.95755139e-01 8.38713348e-01
2.20254734e-01 -8.62288713e-01 8.70465398e-01 8.38350058e-01
-8.14710200e-01 1.17715943e+00 -5.44456057e-02 3.49991232e-01
6.14228070e-01 -7.07198024e-01 5.71330003e-02 6.48840010e-01
-2.38528430e-01 6.36969686e-01 5.10448277e-01 -5.99518239e-01
-9.96862411e-01 -1.17860222e+00 6.19970202e-01 -5.06560147e-01
-2.82132030e-01 -5.41540205e-01 -5.73093116e-01 8.09335649e-01
6.68308794e-01 -3.82240415e-01 5.01270652e-01 -3.15859377e-01
-1.19013712e-01 -5.35656333e-01 -2.14148951e+00 2.43583858e-01
9.05238748e-01 -3.54536623e-01 -3.78757060e-01 2.33290181e-01
4.17111933e-01 4.86276820e-02 -1.69683158e-01 4.96665686e-01
7.31541216e-01 -1.06786573e+00 8.63589108e-01 -3.38221312e-01
-4.57826138e-01 -2.38041282e-01 -5.76307960e-02 -7.02853978e-01
-2.58271545e-01 -6.26057565e-01 1.68588981e-01 1.52300239e+00
4.14190203e-01 -1.09888840e+00 1.08875620e+00 2.91924268e-01
3.85540336e-01 -4.46456701e-01 -1.09349430e+00 -8.76007438e-01
-1.87153503e-01 -1.02870308e-01 6.63291872e-01 8.96281660e-01
-6.25084579e-01 2.27434754e-01 -6.57657981e-01 2.86447182e-02
4.37512577e-01 -7.60491967e-01 4.92796510e-01 -1.20397687e+00
-3.09794217e-01 -2.80331165e-01 -8.05628300e-01 -1.81145757e-01
-8.23357701e-02 -7.69927382e-01 -4.50181931e-01 -1.13030219e+00
4.15133566e-01 -1.44033849e-01 -4.51876730e-01 5.38855910e-01
2.02249065e-01 6.84243798e-01 2.94667645e-03 -5.98290145e-01
1.29738107e-01 -9.17293057e-02 5.05786598e-01 -6.13560639e-02
-4.49125394e-02 1.35403141e-01 -6.91803515e-01 7.12167859e-01
8.41327131e-01 -4.36321676e-01 -3.02539200e-01 -1.59450948e-01
2.64853763e-04 -2.49338657e-01 2.82691151e-01 -1.39596152e+00
7.00057149e-02 5.96100092e-01 9.74051833e-01 -5.20405173e-01
4.91849393e-01 -9.97981727e-01 5.27426183e-01 7.92289138e-01
4.01593119e-01 1.65079892e-01 6.75155878e-01 7.71685690e-02
1.83697194e-01 -3.26543897e-01 1.03659415e+00 -1.44112036e-01
-6.65912926e-01 3.55906755e-01 -5.71070731e-01 -5.68038404e-01
1.16987908e+00 -7.47975945e-01 -2.21342206e-01 -3.21727961e-01
-8.87338400e-01 7.54966959e-02 1.54738739e-01 6.52728975e-01
5.10091960e-01 -9.81996596e-01 -9.35966730e-01 5.84931016e-01
-3.69505674e-01 -7.82187581e-01 4.47376043e-01 4.17072326e-01
-8.39304805e-01 5.88893533e-01 -7.11653411e-01 -1.81411445e-01
-2.00948501e+00 5.36656678e-01 7.86239564e-01 1.19398028e-01
-3.09479654e-01 1.38635039e+00 -1.21219099e-01 -6.03711188e-01
5.00007749e-01 1.75945550e-01 -7.23915279e-01 -2.44457796e-01
7.56629825e-01 5.94172001e-01 -1.15981385e-01 -8.67162585e-01
-6.37035489e-01 5.96225500e-01 -3.05414617e-01 3.28668877e-02
1.41868758e+00 1.66153014e-01 -4.37928051e-01 -1.81327224e-01
1.13732040e+00 -1.06579073e-01 -7.13647008e-01 2.42851257e-01
-3.46534997e-01 -3.16732198e-01 1.36752715e-02 -1.27062571e+00
-1.14283013e+00 7.06804156e-01 1.95452631e+00 -2.85962582e-01
1.16405404e+00 -4.00257528e-01 6.96296334e-01 1.44674331e-01
5.36843725e-02 -1.09660006e+00 -2.98547477e-01 4.37283874e-01
6.52201593e-01 -1.31009650e+00 -3.11361760e-01 -2.40779281e-01
-8.56236219e-02 6.87634051e-01 8.32026243e-01 -2.02329397e-01
1.15959406e+00 5.98455131e-01 1.11370116e-01 -2.44167998e-01
-2.38879800e-01 -3.37432069e-03 -1.06614036e-02 1.14462698e+00
4.22389776e-01 1.48923278e-01 -3.94587725e-01 7.89858282e-01
-1.13687545e-01 4.39275473e-01 1.48953712e-02 6.08529270e-01
-4.13581729e-01 -1.23085785e+00 -4.98202980e-01 6.28970027e-01
-1.09013188e+00 3.38265672e-02 -4.12456930e-01 8.02781105e-01
9.02641952e-01 8.79406691e-01 5.24820015e-02 -6.19017720e-01
5.65415770e-02 2.55250752e-01 7.34111607e-01 -4.46866751e-01
-8.04332912e-01 -6.00404978e-01 -1.92765631e-02 -3.72275740e-01
-3.35114181e-01 -4.21425879e-01 -1.04217052e+00 -7.54451454e-01
-2.49726057e-01 -3.35896462e-01 9.32247400e-01 7.99983680e-01
3.18799108e-01 3.98920596e-01 2.64375746e-01 -4.85944420e-01
-2.57639527e-01 -1.35936010e+00 -6.39680624e-01 -3.87145244e-02
3.72526050e-01 -5.78141212e-01 7.95865059e-02 -4.69724275e-03] | [13.367984771728516, 0.9263292551040649] |
e0c44d1d-c06b-466a-9663-ab1603c56c3f | attribute-prototype-network-for-any-shot | 2204.01208 | null | https://arxiv.org/abs/2204.01208v1 | https://arxiv.org/pdf/2204.01208v1.pdf | Attribute Prototype Network for Any-Shot Learning | Any-shot image classification allows to recognize novel classes with only a few or even zero samples. For the task of zero-shot learning, visual attributes have been shown to play an important role, while in the few-shot regime, the effect of attributes is under-explored. To better transfer attribute-based knowledge from seen to unseen classes, we argue that an image representation with integrated attribute localization ability would be beneficial for any-shot, i.e. zero-shot and few-shot, image classification tasks. To this end, we propose a novel representation learning framework that jointly learns discriminative global and local features using only class-level attributes. While a visual-semantic embedding layer learns global features, local features are learned through an attribute prototype network that simultaneously regresses and decorrelates attributes from intermediate features. Furthermore, we introduce a zoom-in module that localizes and crops the informative regions to encourage the network to learn informative features explicitly. We show that our locality augmented image representations achieve a new state-of-the-art on challenging benchmarks, i.e. CUB, AWA2, and SUN. As an additional benefit, our model points to the visual evidence of the attributes in an image, confirming the improved attribute localization ability of our image representation. The attribute localization is evaluated quantitatively with ground truth part annotations, qualitatively with visualizations, and through well-designed user studies. | ['Zeynep Akata', 'Bernt Schiele', 'Jiuniu Wang', 'Yongqin Xian', 'Wenjia Xu'] | 2022-04-04 | null | null | null | null | ['few-shot-image-classification'] | ['computer-vision'] | [ 1.30107701e-01 2.30444536e-01 -5.02922952e-01 -5.78531027e-01
-6.92768633e-01 -4.63147640e-01 8.08664858e-01 4.41630602e-01
-2.04875946e-01 5.35734296e-01 2.35160097e-01 2.30121732e-01
-2.24927515e-01 -9.09103572e-01 -8.51135015e-01 -8.75346780e-01
-1.04675710e-01 4.40088063e-01 1.03518456e-01 -1.81040794e-01
-5.01507670e-02 4.68279839e-01 -2.02084184e+00 4.77741033e-01
6.04990184e-01 1.23272514e+00 1.32136002e-01 4.05256033e-01
-1.31238073e-01 1.01040208e+00 -1.93675324e-01 -2.93381125e-01
2.77066648e-01 -2.57250279e-01 -7.91319132e-01 1.83327928e-01
6.45485818e-01 -2.37599462e-01 -2.58553267e-01 8.67362082e-01
1.97383493e-01 3.62447232e-01 9.53821719e-01 -1.48402762e+00
-1.17789233e+00 2.83905208e-01 -4.65749234e-01 7.92205483e-02
2.30264515e-01 3.62406522e-01 1.39039528e+00 -1.18905413e+00
8.82234156e-01 1.08347046e+00 4.39120680e-01 4.44876850e-01
-1.45911157e+00 -4.03177708e-01 3.85123640e-01 4.13313359e-01
-1.46519172e+00 -2.37976342e-01 8.31356704e-01 -6.36522353e-01
7.11320758e-01 3.82752307e-02 6.73251390e-01 1.25684881e+00
-6.25796914e-02 9.36444938e-01 9.78936613e-01 -3.47724766e-01
3.76837879e-01 4.22772855e-01 1.47020236e-01 7.30934262e-01
8.48383456e-02 1.69643119e-01 -5.73443353e-01 1.33697912e-01
4.87053961e-01 6.85636699e-01 -1.26566410e-01 -1.02989781e+00
-1.16007209e+00 1.02594423e+00 1.00198293e+00 2.17220753e-01
-4.69771266e-01 2.27161452e-01 4.21863556e-01 2.91619450e-01
6.42794847e-01 4.70191777e-01 -3.64830434e-01 -3.80211743e-03
-6.61930919e-01 -1.92385986e-01 4.19740230e-01 1.09694958e+00
1.41249716e+00 3.26305744e-03 -4.61561322e-01 9.15485620e-01
-1.39764488e-01 3.71989161e-01 2.76512474e-01 -8.72800171e-01
-1.24042079e-01 8.80978048e-01 -1.95888236e-01 -7.23266125e-01
-1.20945260e-01 -5.83401442e-01 -9.22905803e-01 5.23318827e-01
1.45961657e-01 4.00404602e-01 -1.15636146e+00 1.80543244e+00
2.37296551e-01 2.09548727e-01 1.45070059e-02 9.30674076e-01
6.84485972e-01 5.48985541e-01 3.69686127e-01 2.55242020e-01
1.44044840e+00 -1.08182740e+00 -4.54760283e-01 -3.04916799e-01
6.25484228e-01 -2.71028161e-01 1.49466109e+00 2.13189740e-02
-6.70123100e-01 -6.42541170e-01 -1.12381482e+00 -1.46730185e-01
-8.70209992e-01 2.91980430e-02 7.93118775e-01 3.19845349e-01
-8.32086921e-01 6.44263506e-01 -5.20134032e-01 -6.88118994e-01
9.49586034e-01 -7.94698205e-03 -6.54259980e-01 -3.80336076e-01
-9.40933943e-01 8.31596673e-01 3.51276964e-01 -6.09277606e-01
-1.18456876e+00 -8.72102678e-01 -1.01737082e+00 3.92693996e-01
4.48776633e-01 -7.90026069e-01 8.59973192e-01 -1.17930877e+00
-1.13539863e+00 9.73845780e-01 -1.74202416e-02 -3.47249359e-01
1.01090506e-01 -5.55084459e-02 -1.15881972e-01 4.92654443e-01
2.48390928e-01 1.01349044e+00 9.03998792e-01 -1.39455891e+00
-5.61095774e-01 -4.32582796e-01 3.82126153e-01 5.51912710e-02
-6.86443865e-01 -3.33900988e-01 -2.11911738e-01 -6.49826288e-01
-1.56038567e-01 -7.66313851e-01 -5.15948795e-02 5.84289253e-01
-1.66045412e-01 -2.20940679e-01 8.48833561e-01 -1.34222537e-01
6.61860168e-01 -2.44956899e+00 6.16003387e-02 1.51783247e-02
2.81021923e-01 -1.62465885e-01 -3.38201016e-01 3.66186291e-01
-1.09544322e-01 1.35753974e-01 -2.13364646e-01 -4.24770206e-01
-2.52654720e-02 4.64471191e-01 -2.57909060e-01 4.34764326e-01
5.18387735e-01 1.08775556e+00 -1.00818133e+00 -2.86248952e-01
3.40059757e-01 6.30987406e-01 -6.17030621e-01 1.38554394e-01
-1.08004726e-01 3.65416557e-01 -3.00531030e-01 9.60162938e-01
3.26109558e-01 -5.61760664e-01 -2.26485580e-01 -2.86238015e-01
1.66976094e-01 -1.62443876e-01 -8.78187835e-01 1.99592590e+00
-6.40059054e-01 6.05533659e-01 -3.39775115e-01 -9.57010984e-01
8.83716881e-01 1.05941705e-01 4.11033332e-01 -8.17129254e-01
-2.84958258e-02 5.35747819e-02 -2.09853455e-01 -3.76850605e-01
3.59110951e-01 -1.32892415e-01 -2.10139882e-02 4.46119100e-01
6.61200106e-01 3.07055146e-01 8.39926582e-03 6.14943624e-01
9.41024065e-01 2.13088423e-01 6.19904757e-01 -3.49891186e-01
1.03120342e-01 -2.09415443e-02 3.69563162e-01 8.88706088e-01
-1.35492325e-01 7.98041761e-01 3.58777136e-01 -6.33320868e-01
-1.23036504e+00 -1.12179363e+00 -2.36123160e-01 1.66048682e+00
4.19702649e-01 -4.21249211e-01 -4.44815695e-01 -8.16099286e-01
1.36334181e-01 7.70710886e-01 -1.20103788e+00 -4.95939314e-01
-6.37936890e-02 -3.22906077e-01 7.04923049e-02 8.74256670e-01
2.36123934e-01 -1.05698824e+00 -7.80638754e-01 -6.36717603e-02
1.45747170e-01 -9.35180426e-01 -2.05519021e-01 4.96019006e-01
-6.04130447e-01 -9.58918929e-01 -6.07613564e-01 -5.85164249e-01
7.69113898e-01 3.63120794e-01 1.09087241e+00 -4.71991599e-02
-6.29446507e-01 7.50418544e-01 -6.61497176e-01 -2.16571316e-01
-1.11005642e-01 -7.52528897e-03 -1.00846782e-01 3.03283393e-01
3.09854150e-01 -7.83994198e-01 -7.90992737e-01 2.92346209e-01
-8.01063478e-01 8.32684115e-02 8.96628022e-01 1.09989762e+00
7.05227137e-01 -7.00465798e-01 5.28203189e-01 -1.00011170e+00
6.95261136e-02 -6.34965420e-01 -6.52214140e-02 3.21110547e-01
-5.71456015e-01 3.40535045e-01 6.34907067e-01 -5.98279119e-01
-8.08899045e-01 2.20090300e-01 2.39534765e-01 -8.28431129e-01
-3.54805857e-01 2.64472485e-01 -2.76479214e-01 -7.63922632e-02
8.63181055e-01 1.79020002e-01 1.40662685e-01 -4.30384457e-01
9.46579814e-01 4.34398413e-01 4.73306596e-01 -6.70255840e-01
5.99893510e-01 8.13881338e-01 -6.27301112e-02 -7.20604599e-01
-1.15867031e+00 -7.10855961e-01 -7.76511610e-01 -8.70376900e-02
8.97782743e-01 -1.02981341e+00 -5.64897537e-01 2.19731349e-02
-8.76116633e-01 -1.73466384e-01 -9.52029169e-01 2.28495166e-01
-8.18430841e-01 -9.80414897e-02 -4.04541820e-01 -4.56220597e-01
-5.81207275e-02 -1.10474360e+00 1.16263998e+00 1.82943940e-01
-1.59907266e-01 -8.67377698e-01 -1.77047089e-01 -2.88246777e-02
5.24883270e-01 2.24673137e-01 1.09211266e+00 -9.39311445e-01
-7.94818044e-01 -1.80385768e-01 -4.82816279e-01 2.13871211e-01
1.00716889e-01 -1.49589390e-01 -1.46192348e+00 -3.57264280e-01
-4.28662896e-01 -7.44326115e-01 1.26785135e+00 9.53666791e-02
1.54022110e+00 -3.98246109e-01 -4.32801753e-01 8.52871120e-01
1.52074242e+00 -9.99398530e-02 6.18102372e-01 2.81919211e-01
7.26655424e-01 5.19189000e-01 6.24547482e-01 5.33600271e-01
2.49657646e-01 6.78991735e-01 7.10792720e-01 -3.43230337e-01
-4.23805475e-01 -3.51152360e-01 -2.26659328e-02 3.31320554e-01
-9.94963869e-02 -6.09337352e-02 -6.73879623e-01 5.80523193e-01
-1.88540852e+00 -9.13449109e-01 3.82656276e-01 2.14120841e+00
6.54855609e-01 -3.20670456e-02 3.09016015e-02 -1.10037446e-01
5.31179130e-01 2.86488563e-01 -6.04350626e-01 -1.64095342e-01
-2.80936480e-01 7.82611594e-02 2.22981319e-01 2.12598234e-01
-1.27677333e+00 9.66654837e-01 5.38567829e+00 8.76383722e-01
-9.29191470e-01 2.58698463e-01 8.73403251e-01 -6.37224689e-02
-3.77451271e-01 1.51537701e-01 -5.20976067e-01 2.80805171e-01
5.65911651e-01 -8.50323811e-02 2.79953599e-01 1.23207307e+00
-4.25865144e-01 5.56980334e-02 -1.46848965e+00 9.48523521e-01
2.82885820e-01 -1.65708268e+00 3.97701561e-01 1.08887844e-01
6.62108064e-01 -7.66531006e-02 2.13627115e-01 5.73788643e-01
2.84936875e-01 -1.11183393e+00 6.75307572e-01 7.68984079e-01
1.08223748e+00 -7.17292726e-01 5.77513218e-01 9.01395455e-02
-1.25460184e+00 -3.35960537e-01 -6.83958352e-01 -6.23781867e-02
-2.05910474e-01 4.46847051e-01 -7.17801750e-01 4.37116623e-01
8.04232657e-01 1.07220507e+00 -8.56580794e-01 9.14057136e-01
-3.31290632e-01 3.59681398e-01 -5.07028252e-02 7.28190094e-02
3.03797722e-01 5.93771785e-02 2.47366756e-01 1.07337165e+00
2.39331245e-01 2.01260433e-01 3.60606968e-01 1.06580365e+00
-2.72622436e-01 5.53299412e-02 -8.80817175e-01 5.46806231e-02
5.58272839e-01 1.50342011e+00 -6.35608912e-01 -5.03221989e-01
-4.98363793e-01 1.10053515e+00 5.98851025e-01 5.13666928e-01
-5.29513061e-01 -4.62844998e-01 8.85281920e-01 1.23932749e-01
4.40452605e-01 1.24893136e-01 -2.39259720e-01 -1.11213243e+00
-2.60839820e-01 -5.71670830e-01 4.90335315e-01 -7.67634392e-01
-1.57467961e+00 4.37773645e-01 -8.13838169e-02 -1.32899249e+00
-2.92026043e-01 -7.00386584e-01 -6.17465854e-01 5.79721212e-01
-1.50262797e+00 -1.49473619e+00 -6.22673213e-01 6.46915734e-01
5.77136159e-01 -3.24215025e-01 1.06102049e+00 1.69147074e-01
-2.20274776e-01 6.73357904e-01 1.18130632e-01 1.34015247e-01
8.34000468e-01 -1.22139418e+00 8.97176489e-02 3.72271478e-01
5.35125196e-01 6.10904455e-01 5.69626510e-01 -2.37182260e-01
-1.27800608e+00 -1.12993371e+00 4.49153006e-01 -4.34521377e-01
6.93855882e-01 -5.88334143e-01 -1.15888762e+00 6.21922553e-01
2.08748788e-01 8.06070209e-01 7.41242051e-01 3.37874740e-01
-8.80224884e-01 -2.19989002e-01 -9.42246079e-01 2.93318778e-01
1.13820970e+00 -8.01273584e-01 -3.41040432e-01 2.22139448e-01
9.09778178e-01 2.20173404e-01 -8.54267001e-01 3.62307370e-01
4.89659756e-01 -8.94576192e-01 1.24915910e+00 -7.78630674e-01
5.37680507e-01 -1.88187853e-01 -5.19771278e-01 -1.45232785e+00
-6.06245518e-01 2.23354120e-02 -8.46122876e-02 1.31426632e+00
4.44245666e-01 -4.42010313e-01 6.43024683e-01 2.90401399e-01
-2.07114398e-01 -9.67469871e-01 -7.91643739e-01 -9.13811982e-01
6.89254468e-03 -1.74992457e-01 4.88707513e-01 1.06977785e+00
-1.87772766e-01 4.89567459e-01 -3.77596319e-01 -7.23476782e-02
6.03034616e-01 3.53660047e-01 6.31375670e-01 -1.45572078e+00
-3.11540365e-01 -2.86081642e-01 -8.68507028e-01 -7.84132242e-01
1.76315367e-01 -1.18822837e+00 -1.44034535e-01 -1.44819951e+00
7.33100593e-01 -4.26906884e-01 -6.70526862e-01 8.97892058e-01
-2.04161137e-01 4.46774751e-01 3.46396297e-01 3.02680105e-01
-8.87913883e-01 8.80466700e-01 1.16006267e+00 -4.82123047e-01
2.62638539e-01 -4.35667425e-01 -7.24421859e-01 5.90269923e-01
6.25161350e-01 -5.63835859e-01 -4.38142985e-01 -1.26688853e-01
1.19577333e-01 -3.45258147e-01 8.07637632e-01 -1.01682603e+00
2.86021292e-01 -1.08625740e-01 6.38145328e-01 -2.31669992e-01
6.32630765e-01 -7.66330004e-01 -3.38538826e-01 1.60203546e-01
-5.55455327e-01 -3.02796245e-01 -6.51543587e-03 8.80373061e-01
-3.40561301e-01 -1.11531988e-01 9.34344530e-01 -1.21076547e-01
-1.24369478e+00 5.64071834e-01 6.42745048e-02 1.09656535e-01
1.21345091e+00 -4.39625442e-01 -5.08009017e-01 -4.35611695e-01
-8.82778227e-01 1.13798670e-01 7.11184323e-01 4.44292843e-01
7.46795475e-01 -1.51776934e+00 -5.33928931e-01 3.55642647e-01
9.65808868e-01 -2.90045947e-01 3.58643442e-01 7.66359389e-01
1.27628334e-02 -2.12498624e-02 -4.92027432e-01 -9.82142925e-01
-9.00459826e-01 9.87519741e-01 3.25248800e-02 1.60475850e-01
-8.98721039e-01 8.62806261e-01 7.43893921e-01 -3.32247704e-01
2.65116483e-01 -5.25142662e-02 -1.46201953e-01 3.12600732e-01
5.11302531e-01 1.18224934e-01 -1.59141153e-01 -4.54112142e-01
-3.02617908e-01 5.62191367e-01 -1.55241147e-01 2.06630364e-01
1.32979798e+00 -9.88824591e-02 1.25772089e-01 7.86917508e-01
1.43293333e+00 -3.13200653e-01 -1.61502743e+00 -4.89153147e-01
-8.97240490e-02 -5.62536895e-01 -1.60649210e-01 -1.00051522e+00
-1.01794589e+00 1.30886960e+00 7.29256868e-01 -1.23050503e-01
1.06334102e+00 4.92854357e-01 2.90028632e-01 5.46158314e-01
3.76052141e-01 -8.87018561e-01 6.26741886e-01 3.73841614e-01
7.34956324e-01 -1.57651842e+00 -1.98513672e-01 -3.38502973e-01
-8.65047395e-01 9.99066353e-01 6.63673937e-01 -9.04283747e-02
5.26024997e-01 7.62467012e-02 -2.25425452e-01 -2.46681347e-01
-8.87671053e-01 -3.98872405e-01 3.77442122e-01 6.80210531e-01
2.15379804e-01 1.49553016e-01 3.37999284e-01 6.03148460e-01
1.95232943e-01 -3.26497644e-01 2.83773869e-01 7.26061225e-01
-6.66925311e-01 -7.87829936e-01 6.85433000e-02 4.24953282e-01
2.83942446e-02 -1.81440488e-01 -3.70806247e-01 9.10710096e-01
1.90699324e-01 5.18879831e-01 3.10094118e-01 -3.42987150e-01
8.79351720e-02 1.55998528e-01 2.76152015e-01 -7.95591116e-01
-2.87627488e-01 -2.82581180e-01 -2.05300942e-01 -7.51109064e-01
-2.53122240e-01 -2.66296297e-01 -1.10306096e+00 8.98160227e-03
1.28863035e-02 -8.13004151e-02 5.42336047e-01 7.24600613e-01
6.22458279e-01 7.03583300e-01 6.14242971e-01 -8.85338366e-01
-4.82755184e-01 -8.41298759e-01 -6.96918488e-01 8.25498879e-01
3.12223494e-01 -9.71090496e-01 -4.40930635e-01 -3.96399982e-02] | [10.00623893737793, 2.3533005714416504] |
c4654d8f-1aca-4961-b309-86bd50e42d1f | rethinking-the-encoding-of-satellite-image | 2305.02086 | null | https://arxiv.org/abs/2305.02086v1 | https://arxiv.org/pdf/2305.02086v1.pdf | Rethinking the Encoding of Satellite Image Time Series | Representation learning of Satellite Image Time Series (SITS) presents its unique challenges, such as prohibitive computation burden caused by high spatiotemporal resolutions, irregular acquisition times, and complex spatiotemporal interactions, leading to highly-specialized neural network architectures for SITS analysis. Despite the promising results achieved by some pioneering work, we argue that satisfactory representation learning paradigms have not yet been established for SITS analysis, causing an isolated island where transferring successful paradigms or the latest advances from Computer Vision (CV) to SITS is arduous. In this paper, we develop a unique perspective of SITS processing as a direct set prediction problem, inspired by the recent trend in adopting query-based transformer decoders to streamline the object detection or image segmentation pipeline, and further propose to decompose the representation learning process of SITS into three explicit steps: collect--update--distribute, which is computationally efficient and suits for irregularly-sampled and asynchronous temporal observations. Facilitated by the unique reformulation and effective feature extraction framework proposed, our models pre-trained on pixel-set format input and then fine-tuned on downstream dense prediction tasks by simply appending a commonly-used segmentation network have attained new state-of-the-art (SoTA) results on PASTIS dataset compared to bespoke neural architectures such as U-TAE. Furthermore, the clear separation, conceptually and practically, between temporal and spatial components in the panoptic segmentation pipeline of SITS allows us to leverage the recent advances in CV, such as Mask2Former, a universal segmentation architecture, resulting in a noticeable 8.8 points increase in PQ compared to the best score reported so far. | ['Roy Sterritt', 'Peter Nicholl', 'Yaxin Bi', 'Xin Cai'] | 2023-05-03 | null | null | null | null | ['panoptic-segmentation'] | ['computer-vision'] | [ 5.89735210e-01 -3.77959907e-01 1.62704498e-01 -3.43426764e-01
-9.20053244e-01 -6.14969790e-01 8.86031926e-01 -4.74390797e-02
-5.23188710e-01 3.58499825e-01 -5.42284623e-02 -5.71854115e-01
-3.14618230e-01 -7.80573666e-01 -5.56542039e-01 -7.97762394e-01
-3.91283512e-01 3.93235624e-01 2.76428163e-01 -3.84313971e-01
-3.59498560e-02 7.20120430e-01 -1.81997442e+00 1.21159263e-01
6.04746997e-01 1.28702497e+00 2.52430469e-01 6.60468102e-01
-5.88766076e-02 4.89200145e-01 -3.54446143e-01 1.04856871e-01
6.44179225e-01 -3.97397339e-01 -6.45671844e-01 5.61154597e-02
3.87377620e-01 -2.83951432e-01 -3.25657457e-01 5.42862952e-01
3.76304001e-01 3.07855129e-01 5.39697230e-01 -6.88952684e-01
-3.22302073e-01 3.62129420e-01 -7.47171342e-01 6.42976105e-01
-3.77583802e-01 5.21216691e-01 1.11807084e+00 -5.09568214e-01
4.61146683e-01 8.06263983e-01 1.02079129e+00 8.98263454e-02
-1.57449746e+00 -4.54917490e-01 3.33636314e-01 3.56115736e-02
-1.27964652e+00 -5.87111354e-01 3.09653670e-01 -5.63691437e-01
1.48809135e+00 3.79862517e-01 9.13084090e-01 8.00302088e-01
-1.64388537e-01 8.02939057e-01 1.19275236e+00 -1.60784394e-01
2.05727264e-01 -2.80330122e-01 1.27905995e-01 2.91139156e-01
-7.35670626e-02 2.97125101e-01 -1.96972504e-01 2.38794893e-01
7.84690082e-01 4.31761555e-02 -2.63317555e-01 1.51310796e-02
-1.15720940e+00 8.55691493e-01 6.51840985e-01 4.07647789e-01
-6.06482983e-01 2.79448867e-01 5.86962283e-01 4.70224172e-01
8.44043493e-01 3.96816730e-01 -7.55150497e-01 -1.35695681e-01
-1.72139025e+00 2.09724009e-01 5.22885740e-01 4.22574341e-01
8.06918323e-01 3.67436588e-01 -5.85064180e-02 6.04992330e-01
2.40750656e-01 5.80747306e-01 4.05417651e-01 -7.29209006e-01
4.21633601e-01 4.91243839e-01 -1.86027095e-01 -8.41192305e-01
-6.36682451e-01 -7.89992273e-01 -1.01995516e+00 1.75942570e-01
3.25754493e-01 -1.45548671e-01 -1.06309140e+00 1.56368399e+00
2.58637697e-01 2.72928149e-01 -4.65318486e-02 1.00598729e+00
3.81019324e-01 9.92057741e-01 1.13291174e-01 -1.90307498e-01
1.30597126e+00 -6.76550686e-01 -2.16377214e-01 -4.32663888e-01
4.28960562e-01 -4.56253886e-01 8.82544160e-01 1.92191124e-01
-7.32591152e-01 -6.20624185e-01 -1.03536654e+00 2.53269058e-02
-7.12133467e-01 5.38863391e-02 1.05913734e+00 3.16737741e-01
-1.43247306e+00 7.47321486e-01 -1.18322992e+00 -5.66871405e-01
6.96681023e-01 3.94691199e-01 -9.50066596e-02 2.80165404e-01
-9.77954507e-01 7.51541495e-01 3.48325789e-01 5.75744748e-01
-8.40504110e-01 -1.00601649e+00 -6.88022912e-01 2.00716004e-01
2.90538251e-01 -8.42864454e-01 1.17057669e+00 -1.04115200e+00
-1.52919459e+00 9.30775344e-01 -9.59376171e-02 -1.10547256e+00
3.56242597e-01 -4.14725505e-02 -2.12434828e-01 3.07777196e-01
4.54185967e-04 7.57465482e-01 9.48876500e-01 -8.15227747e-01
-6.88581407e-01 -3.24271262e-01 -1.80244759e-01 1.92185819e-01
-1.42068192e-01 8.23683590e-02 -3.26400012e-01 -6.40194714e-01
2.94168353e-01 -7.70076632e-01 -7.15033412e-01 3.52797024e-02
3.04575711e-02 7.61071965e-02 8.12218368e-01 -6.99996769e-01
1.17853272e+00 -2.19775510e+00 1.22125454e-01 1.55250862e-01
-1.20238746e-02 4.80790973e-01 -2.16542736e-01 5.18843949e-01
-1.73772931e-01 4.49487194e-02 -8.09767306e-01 -5.07606804e-01
-5.67702902e-03 3.63372207e-01 -9.65515673e-01 5.36292136e-01
4.85230118e-01 9.48530912e-01 -7.03987062e-01 -1.92491308e-01
5.44381976e-01 3.93036157e-01 -3.26076299e-01 2.91016430e-01
-4.55353737e-01 5.85603535e-01 -2.76554972e-01 6.00890219e-01
6.91416025e-01 -3.94616365e-01 -6.45078868e-02 -3.86750735e-02
-7.60119498e-01 5.18678010e-01 -1.06501412e+00 1.75899208e+00
-3.80760700e-01 7.70499408e-01 2.40152597e-01 -1.33455014e+00
7.67445028e-01 2.29522794e-01 6.42329454e-01 -9.51095700e-01
-6.82813749e-02 2.96494603e-01 -1.50862321e-01 -4.49366778e-01
5.30779302e-01 -1.85005903e-01 1.78423747e-01 4.38032627e-01
2.01968491e-01 -4.52605784e-01 2.66878694e-01 -4.48084809e-02
7.92212665e-01 5.90849459e-01 1.70813888e-01 -2.70518929e-01
3.72992158e-01 4.03613329e-01 3.03892285e-01 8.89603138e-01
-1.10436037e-01 7.63950765e-01 2.95577019e-01 -6.04892254e-01
-8.64640534e-01 -8.78043890e-01 -4.72854286e-01 1.11030054e+00
-4.14384484e-01 -3.39871079e-01 -4.39774662e-01 -2.49150947e-01
-1.51166975e-01 3.76412153e-01 -5.95712423e-01 2.78044522e-01
-6.63079917e-01 -1.35126185e+00 6.96801484e-01 4.65499252e-01
6.11027598e-01 -1.00899196e+00 -1.14987946e+00 6.06359184e-01
9.38155204e-02 -9.15456831e-01 1.90663189e-01 5.77394426e-01
-1.14108860e+00 -6.90306723e-01 -7.39521146e-01 -3.06015611e-01
-5.90495719e-03 4.61583197e-01 1.13123238e+00 -2.35676661e-01
-4.32140142e-01 1.85014740e-01 -2.24218130e-01 -3.30308288e-01
1.09002322e-01 3.56036991e-01 -3.09543222e-01 1.31670877e-01
1.88215360e-01 -9.01385486e-01 -9.01938975e-01 1.38560785e-02
-1.06918120e+00 1.83912963e-01 4.80807096e-01 6.56212807e-01
7.14619875e-01 -1.69681296e-01 4.32422340e-01 -4.48346347e-01
9.85877961e-03 -6.96147323e-01 -1.08388889e+00 -2.61093471e-02
-6.38307929e-01 -1.95574313e-01 5.05978882e-01 9.66939107e-02
-1.04284096e+00 9.44721848e-02 -3.12512875e-01 -2.89687514e-01
-3.32857996e-01 1.02090788e+00 6.69881225e-01 9.61414874e-02
5.69990516e-01 5.49097836e-01 1.56924188e-01 -5.16810775e-01
3.90223056e-01 5.67261040e-01 5.46114683e-01 -3.10775012e-01
7.62139916e-01 8.92578959e-01 -6.77914396e-02 -1.02363062e+00
-8.36344004e-01 -6.44388616e-01 -6.22452259e-01 -1.20803721e-01
8.80227089e-01 -1.22877455e+00 -4.33685541e-01 6.07633293e-01
-9.47895885e-01 -6.55965865e-01 -5.00460625e-01 3.08427334e-01
-3.83799285e-01 3.42582256e-01 -4.94026512e-01 -9.69458878e-01
-5.83276510e-01 -9.61739242e-01 1.28284454e+00 1.87669188e-01
9.61762816e-02 -9.66264665e-01 2.50671566e-01 2.05079764e-01
9.66709912e-01 4.01341885e-01 5.62795043e-01 -3.81304830e-01
-8.28181028e-01 -1.43788336e-02 -5.52625299e-01 3.55897993e-01
-1.84211418e-01 1.59988627e-01 -1.35375953e+00 -2.09595844e-01
1.21784933e-01 -2.62192786e-01 1.33068871e+00 7.12696195e-01
1.07234371e+00 8.48425701e-02 -1.94174513e-01 1.20093107e+00
1.62107909e+00 -7.65019059e-02 5.41620195e-01 4.99209821e-01
3.42578769e-01 6.77216768e-01 3.01248342e-01 5.04283190e-01
3.73232365e-01 4.36641157e-01 5.94083369e-01 -4.28598821e-01
-9.59327072e-02 1.70506135e-01 2.58098930e-01 7.16978908e-01
-2.47409478e-01 -1.67265475e-01 -1.01961696e+00 7.88185894e-01
-2.00933623e+00 -1.08591461e+00 -2.34864473e-01 2.09883904e+00
5.89577615e-01 1.02838175e-02 -9.57778841e-02 -1.93787385e-02
1.08831190e-01 8.81240726e-01 -5.13858020e-01 -2.05189094e-01
-3.93811792e-01 5.20800889e-01 7.02447057e-01 1.25949889e-01
-1.44317460e+00 1.13614368e+00 6.47166634e+00 7.29181826e-01
-1.65522707e+00 -3.52940313e-03 6.24565840e-01 -2.44422108e-01
-1.07738838e-01 -7.31425956e-02 -6.53729975e-01 3.09329599e-01
1.41771328e+00 1.82482481e-01 6.15865767e-01 4.63496327e-01
4.76571560e-01 -1.80754632e-01 -8.46581161e-01 8.66940260e-01
-3.04997981e-01 -1.65108323e+00 -7.93560818e-02 3.33610848e-02
4.88626897e-01 1.09371161e+00 2.00825393e-01 4.51739371e-01
1.51145428e-01 -1.05505490e+00 6.81605577e-01 4.76000816e-01
7.24529326e-01 -2.08581045e-01 5.65669417e-01 2.66732842e-01
-1.42720532e+00 -8.48321766e-02 -4.08072561e-01 -3.18961591e-01
3.14393789e-01 5.92314422e-01 -6.90135896e-01 8.13791871e-01
9.16871309e-01 1.08991635e+00 -4.69819218e-01 9.49963391e-01
1.53101206e-01 1.02940392e+00 -8.68849337e-01 5.33851027e-01
8.89415443e-01 -5.87775230e-01 4.58604902e-01 1.49522614e+00
3.17572534e-01 7.62440339e-02 -4.48311605e-02 9.22088981e-01
2.25611821e-01 -7.14810640e-02 -6.70535326e-01 -9.41465944e-02
7.95204714e-02 1.39821827e+00 -7.55968451e-01 -3.42198670e-01
-5.25967956e-01 6.99329138e-01 2.36725911e-01 4.83211309e-01
-8.18881035e-01 1.16205364e-01 7.60444045e-01 6.02010265e-02
8.30324709e-01 -4.71378565e-01 -4.74394023e-01 -1.12145209e+00
-8.73548314e-02 -7.19221532e-01 5.50272882e-01 -7.01261342e-01
-1.09990811e+00 7.34644651e-01 5.03547676e-02 -1.28205109e+00
-2.28229910e-01 -4.93220896e-01 -7.02649355e-01 1.01313972e+00
-2.09236526e+00 -1.44700611e+00 -4.94616665e-02 5.10573089e-01
6.48868144e-01 -5.26302718e-02 8.88278902e-01 2.78767884e-01
-6.29769266e-01 -4.39541787e-02 3.01443696e-01 -1.09794207e-01
2.17737272e-01 -1.12923276e+00 6.07390106e-01 1.15852880e+00
3.00279349e-01 4.65975016e-01 4.50002015e-01 -7.39900991e-02
-1.37833798e+00 -1.21036518e+00 8.64530265e-01 -2.47517526e-01
1.01136911e+00 -2.75737345e-01 -1.01982296e+00 8.27480853e-01
2.53731221e-01 2.36466557e-01 5.03839016e-01 8.81429240e-02
-3.80132109e-01 -3.71005446e-01 -6.50284708e-01 5.30399978e-01
9.43267107e-01 -6.11543953e-01 -6.80750787e-01 2.89285868e-01
5.00786066e-01 -2.44133040e-01 -6.82660520e-01 4.45103824e-01
3.70734036e-01 -1.12518060e+00 1.05970120e+00 -3.82783890e-01
3.45236391e-01 -4.82689083e-01 -2.71267086e-01 -9.96662796e-01
-3.18267614e-01 -6.29931390e-01 1.06920600e-01 1.07592010e+00
4.45404917e-01 -7.76072145e-01 5.11410952e-01 2.62656540e-01
-4.93378133e-01 -5.78065217e-01 -1.28578281e+00 -4.61257845e-01
-6.07405230e-02 -7.45648026e-01 4.56858873e-01 8.32883179e-01
-5.53595543e-01 1.96612492e-01 -1.02554046e-01 4.19994146e-01
4.50672448e-01 6.19911849e-01 5.64591169e-01 -1.36608148e+00
-4.44017291e-01 -6.52151644e-01 -1.03111185e-01 -1.32266402e+00
-1.44741401e-01 -8.86835694e-01 6.15886822e-02 -1.44260430e+00
-3.24499220e-01 -5.35734892e-01 -4.29979831e-01 4.68930602e-01
1.07059099e-01 4.44689780e-01 1.64630085e-01 5.25912106e-01
-4.87060964e-01 5.52519202e-01 8.34905744e-01 -6.37664199e-02
-2.73966968e-01 -1.17364330e-02 -4.38385338e-01 5.00088811e-01
6.00505531e-01 -2.72567004e-01 -2.93224007e-01 -8.71248603e-01
3.36762339e-01 1.18661113e-01 5.82742274e-01 -1.12801349e+00
1.71873346e-01 1.15343839e-01 1.23370185e-01 -7.59983897e-01
3.72678280e-01 -8.50709379e-01 1.95634961e-01 3.32382977e-01
4.21736389e-02 -1.85293853e-01 5.67129076e-01 4.38478947e-01
-4.32814091e-01 2.42833700e-02 6.99027658e-01 -3.30764830e-01
-1.08322227e+00 4.60319757e-01 -5.96125185e-01 -2.14480549e-01
7.08892822e-01 -4.32461262e-01 -7.11909682e-02 2.38125329e-03
-6.25524163e-01 2.25789279e-01 3.27670462e-02 1.76520258e-01
1.39433131e-01 -5.14743805e-01 -8.91850173e-01 4.25898582e-01
-7.05676228e-02 3.71792108e-01 5.19695759e-01 1.19040513e+00
-5.98835826e-01 5.51895440e-01 -3.42385471e-02 -1.00175154e+00
-6.64440155e-01 1.34029090e-01 4.74135429e-01 -4.82190073e-01
-9.38694000e-01 8.66453230e-01 1.39267683e-01 -4.13196683e-01
-1.24354973e-01 -6.46833599e-01 -1.61180392e-01 5.17376423e-01
4.14119959e-01 2.88338624e-02 3.26625794e-01 -4.12780821e-01
-3.24183792e-01 5.77631116e-01 9.96579826e-02 -1.45535856e-01
1.84486783e+00 -2.20630094e-01 -1.51806951e-01 4.74734545e-01
1.01482940e+00 -6.15349710e-01 -1.72579074e+00 -3.39615703e-01
1.34439975e-01 -4.60934490e-02 4.55266446e-01 -7.45405614e-01
-1.16115081e+00 1.09919727e+00 7.59988546e-01 4.02175725e-01
1.29076266e+00 -1.35287628e-01 7.20336795e-01 3.43753368e-01
1.90682575e-01 -1.02759433e+00 -5.98208129e-01 8.12104940e-01
8.45249176e-01 -1.30530465e+00 -2.00704280e-02 4.48695496e-02
-4.46622968e-01 9.90054011e-01 1.04285255e-01 -2.39004448e-01
7.13427246e-01 2.31671914e-01 1.16845019e-01 -4.09236193e-01
-8.10082555e-01 -5.98836482e-01 1.02525383e-01 5.06799817e-01
3.45842063e-01 -8.10025781e-02 9.96741876e-02 3.23319435e-01
3.77302407e-04 1.10378459e-01 8.03173929e-02 6.75757408e-01
-4.06710237e-01 -6.55098259e-01 -2.33174607e-01 5.61825752e-01
-3.17073196e-01 -4.22350913e-01 1.79085970e-01 8.50022256e-01
4.12226394e-02 6.46140575e-01 4.85999912e-01 -1.18018258e-02
2.23271400e-01 7.36424252e-02 3.65193933e-02 -5.04259646e-01
-1.11252129e+00 2.24295497e-01 -6.27767146e-02 -6.40233457e-01
-6.90827906e-01 -9.07664776e-01 -1.00091422e+00 -6.29648864e-02
1.04621546e-02 -1.02829069e-01 7.10442483e-01 1.09625733e+00
5.45405924e-01 5.40338576e-01 5.13273954e-01 -1.39583838e+00
-6.11124337e-01 -9.59896505e-01 -4.94705349e-01 2.59805899e-02
5.30774236e-01 -2.63513982e-01 -4.26691473e-01 6.43050373e-02] | [9.556787490844727, -1.4883288145065308] |
164486fd-f774-470f-a0a3-4a254ee3cce9 | how-to-select-which-active-learning-strategy | 2306.03543 | null | https://arxiv.org/abs/2306.03543v1 | https://arxiv.org/pdf/2306.03543v1.pdf | How to Select Which Active Learning Strategy is Best Suited for Your Specific Problem and Budget | In Active Learning (AL), a learner actively chooses which unlabeled examples to query for labels from an oracle, under some budget constraints. Different AL query strategies are more suited to different problems and budgets. Therefore, in practice, knowing in advance which AL strategy is most suited for the problem at hand remains an open problem. To tackle this challenge, we propose a practical derivative-based method that dynamically identifies the best strategy for each budget. We provide theoretical analysis of a simplified case to motivate our approach and build intuition. We then introduce a method to dynamically select an AL strategy based on the specific problem and budget. Empirical results showcase the effectiveness of our approach across diverse budgets and computer vision tasks. | ['Daphna Weinshall', 'Guy Hacohen'] | 2023-06-06 | null | null | null | null | ['active-learning', 'active-learning'] | ['methodology', 'natural-language-processing'] | [ 9.19224247e-02 5.65957129e-02 -8.92044842e-01 -5.22186399e-01
-1.17605793e+00 -8.26395333e-01 1.64453000e-01 3.02466184e-01
-8.15816343e-01 8.58998716e-01 -4.14129555e-01 -4.32145536e-01
-4.18810546e-01 -6.19767725e-01 -5.78240991e-01 -7.55913973e-01
2.05768153e-01 8.64366770e-01 3.51755679e-01 3.05981666e-01
2.58058906e-01 5.91894805e-01 -1.25220585e+00 -1.02742039e-01
9.58086371e-01 9.50147688e-01 2.85424739e-01 6.76706612e-01
-2.87453830e-01 1.00379777e+00 -7.48192608e-01 -3.16565812e-01
5.87522388e-01 -5.79585493e-01 -8.71805310e-01 4.78743583e-01
5.56993842e-01 -1.55153006e-01 1.58188760e-01 9.74427044e-01
5.03796101e-01 3.27821225e-01 4.34479237e-01 -1.15019763e+00
-6.21195398e-02 8.08145404e-01 -2.31437519e-01 5.47536135e-01
1.34668767e-01 5.59907742e-02 1.28959191e+00 -7.02966511e-01
4.64627147e-01 8.80075395e-01 2.57604241e-01 7.42709517e-01
-1.36973453e+00 -4.84747261e-01 7.74370253e-01 2.65104473e-01
-1.16444194e+00 -5.43258905e-01 7.84948349e-01 -9.66194347e-02
3.05162132e-01 4.86498296e-01 6.96873128e-01 6.24151051e-01
-5.83976328e-01 1.39282846e+00 1.24493098e+00 -7.45044708e-01
6.61123872e-01 4.35229957e-01 4.25526112e-01 7.17857480e-01
2.01664627e-01 -1.47487015e-01 -6.68303668e-01 -3.47471207e-01
5.06513417e-01 -9.96947065e-02 -3.76869947e-01 -9.25084233e-01
-6.61308587e-01 1.09125936e+00 1.43370986e-01 -1.14812650e-01
-1.88287750e-01 1.14520185e-01 -3.94174829e-02 6.67703390e-01
5.32521784e-01 7.54766047e-01 -5.45254946e-01 4.26181369e-02
-8.45267773e-01 3.76035392e-01 1.10520673e+00 8.80911469e-01
9.61292982e-01 -4.09222543e-01 -1.89910859e-01 1.07754433e+00
3.96885127e-01 2.40884736e-01 1.20754428e-01 -1.01254451e+00
5.42778075e-01 5.33321679e-01 4.28332329e-01 -3.36949497e-01
-1.08312540e-01 -4.17115062e-01 1.66519344e-01 1.07087441e-01
6.55154347e-01 -1.96269065e-01 -7.48584092e-01 1.52448380e+00
5.66879034e-01 -5.70820011e-02 -2.42224887e-01 8.83429527e-01
4.77630764e-01 3.49650323e-01 -8.68605450e-02 -7.31279314e-01
8.86048317e-01 -1.07370710e+00 -4.55773741e-01 -4.94297385e-01
5.52374601e-01 -4.95625585e-01 1.06806636e+00 3.74896437e-01
-1.02561057e+00 -5.85483462e-02 -9.95449007e-01 3.66607219e-01
-1.65488571e-02 -8.39446932e-02 7.46510386e-01 8.41867983e-01
-8.15959930e-01 2.14408502e-01 -9.36650574e-01 -1.69990197e-01
4.91320193e-01 6.42356217e-01 2.96430707e-01 -1.66775912e-01
-9.35575902e-01 6.31279171e-01 4.57773298e-01 -1.44939780e-01
-9.45585191e-01 -4.56175834e-01 -4.67569888e-01 1.64718598e-01
1.20344472e+00 -5.06533027e-01 1.80812132e+00 -1.28862250e+00
-1.47328866e+00 8.88578832e-01 1.34982793e-02 -7.39231706e-01
7.93368697e-01 -2.92565301e-02 1.95844725e-01 1.69988751e-01
-1.12759002e-01 5.35363138e-01 7.18247294e-01 -1.19996917e+00
-9.70654666e-01 -1.81882739e-01 8.52557480e-01 7.46288240e-01
-5.68849206e-01 8.46567228e-02 -7.47443974e-01 -4.38015401e-01
2.04534113e-01 -9.73217666e-01 -5.69013655e-01 1.86633736e-01
-3.85396443e-02 -4.93434876e-01 5.00447690e-01 3.06194812e-01
1.28818357e+00 -1.99815774e+00 -1.21956281e-02 2.25501031e-01
2.86761731e-01 8.63637868e-03 3.14439684e-01 2.49938428e-01
4.61275041e-01 6.00396693e-02 -1.63379744e-01 -3.97383630e-01
-8.60752463e-02 4.40092623e-01 -2.31907457e-01 4.91937906e-01
-8.51833597e-02 6.16033018e-01 -1.02136290e+00 -8.39343905e-01
-1.21257536e-01 -1.91890523e-01 -5.90468824e-01 5.60930312e-01
-6.04258001e-01 3.75953496e-01 -7.55454183e-01 6.90508425e-01
3.06807071e-01 -5.95072627e-01 3.48547906e-01 5.06499708e-01
6.04816750e-02 2.51152188e-01 -1.31381655e+00 1.24424994e+00
-5.11452079e-01 3.89985979e-01 1.70616165e-01 -1.36472666e+00
7.72629917e-01 1.77414283e-01 5.27518868e-01 -5.62498391e-01
-1.91570707e-02 3.27374071e-01 -6.04440942e-02 -4.04850572e-01
1.12460740e-01 -8.33436176e-02 4.18751575e-02 7.74871528e-01
-5.25967963e-02 2.18290761e-01 3.21832210e-01 4.31243330e-02
1.09133196e+00 -1.95711777e-01 5.29499352e-01 -1.82138622e-01
4.57162946e-01 7.09884688e-02 5.80461800e-01 1.39401281e+00
-5.88959754e-01 2.74818122e-01 5.01572430e-01 -4.14806396e-01
-4.17768866e-01 -6.37805998e-01 -2.14297753e-02 1.53405225e+00
3.39445174e-01 1.09902889e-01 -6.23851299e-01 -1.27472615e+00
-2.51333099e-02 4.76700813e-01 -4.04981017e-01 2.72050560e-01
-7.59067953e-01 -7.52406120e-01 -4.94509637e-02 7.99648464e-02
3.39229167e-01 -1.08224571e+00 -1.01682746e+00 1.51456758e-01
8.27057287e-02 -8.15830231e-01 -5.21830320e-01 6.93765640e-01
-9.65484858e-01 -1.09135592e+00 -6.82620227e-01 -6.49047017e-01
8.76250446e-01 3.48130196e-01 1.30493104e+00 1.77986667e-01
4.95651066e-02 6.05981946e-01 -4.00204122e-01 -5.56152582e-01
-2.75304496e-01 4.78238702e-01 -3.03980738e-01 -6.72931671e-02
2.46095583e-01 5.17163463e-02 -5.35428882e-01 3.60567033e-01
-8.04969251e-01 -1.89327434e-01 5.35441816e-01 8.39540482e-01
7.97258437e-01 9.68224648e-03 4.54933971e-01 -1.58084989e+00
6.28621757e-01 -5.34938991e-01 -1.15279520e+00 6.75020516e-01
-1.02366853e+00 1.39748231e-01 5.07245123e-01 -6.47501290e-01
-9.24536765e-01 5.14729500e-01 1.64710358e-01 -4.32886332e-01
8.32318589e-02 5.10474861e-01 -3.85760427e-01 -3.59206021e-01
6.91770196e-01 -4.52164561e-02 -3.37634593e-01 -3.93761486e-01
1.25916719e-01 4.70925182e-01 9.38961580e-02 -9.64150608e-01
4.60377783e-01 2.82566369e-01 -3.01495910e-01 -3.06228757e-01
-1.59002638e+00 -7.23676741e-01 -5.22380233e-01 -3.23088616e-01
9.02509615e-02 -6.07706606e-01 -6.38125956e-01 1.41141549e-01
-7.39243746e-01 -5.99753797e-01 -5.12067080e-01 4.16957766e-01
-7.07223475e-01 1.29905194e-01 -1.62088722e-01 -1.20276749e+00
-1.03631765e-01 -1.40054691e+00 6.31379068e-01 4.29537952e-01
-3.80431376e-02 -1.09707725e+00 2.30993614e-01 6.37250423e-01
2.48486191e-01 -2.68891007e-01 6.58303440e-01 -1.06011319e+00
-1.14608550e+00 -5.00062816e-02 8.30961317e-02 1.46993592e-01
-3.20977357e-04 -3.94794732e-01 -9.55663145e-01 -5.42783737e-01
2.16907039e-01 -5.02793312e-01 8.54559004e-01 3.15994292e-01
1.52799547e+00 -6.58187747e-01 -3.57359111e-01 5.88209867e-01
1.50748301e+00 4.28978443e-01 3.68239433e-02 5.04926085e-01
4.47788984e-01 4.70507264e-01 1.05677736e+00 2.47517288e-01
2.84321364e-02 7.72318244e-01 3.93721730e-01 4.38987426e-02
3.11950475e-01 3.12854946e-02 2.25132272e-01 3.86414140e-01
2.71023959e-01 -6.27377748e-01 -9.20992792e-01 3.39283943e-01
-1.90266585e+00 -7.53630102e-01 6.34437501e-01 2.43462157e+00
1.07263052e+00 6.61331713e-01 4.31538820e-01 2.41287604e-01
5.27841151e-01 1.07112415e-01 -1.05825508e+00 -1.86023936e-01
8.28948896e-03 2.24318013e-01 8.49408865e-01 5.29818833e-01
-1.21900058e+00 7.97098815e-01 7.13529825e+00 1.01108897e+00
-1.21997845e+00 1.60434589e-01 8.42706680e-01 -1.99626207e-01
-2.39300504e-01 1.94761127e-01 -1.05851054e+00 2.78237939e-01
6.94991589e-01 -2.87352651e-01 4.64416236e-01 1.06593072e+00
-9.69390795e-02 -2.91405946e-01 -1.35263395e+00 8.09330165e-01
-2.28917152e-02 -1.34655225e+00 -2.78184980e-01 6.32242933e-02
7.46139228e-01 6.06239140e-02 -1.01493169e-02 2.55520314e-01
5.86513460e-01 -6.65207207e-01 6.87767148e-01 1.83274478e-01
4.97261912e-01 -7.66619027e-01 4.42877859e-01 7.96973526e-01
-8.18996787e-01 -4.61818784e-01 -2.12821737e-01 5.76935485e-02
-2.44379416e-01 3.22376579e-01 -1.32524407e+00 1.31715655e-01
4.42183495e-01 1.47675812e-01 -5.61665535e-01 1.30018544e+00
-2.75634706e-01 1.18004680e+00 -3.47299993e-01 -3.12937319e-01
1.30421683e-01 9.08565056e-03 4.84017760e-01 9.47586596e-01
-1.49237797e-01 2.40348160e-01 6.19797528e-01 6.70646429e-01
-2.46794060e-01 4.51518029e-01 -3.24713260e-01 -1.48203626e-01
9.40940499e-01 1.02217555e+00 -1.16426384e+00 -2.43747577e-01
-4.57022965e-01 4.88511384e-01 7.47034252e-01 4.32947904e-01
-5.98157823e-01 3.89390886e-02 5.83794564e-02 7.36693367e-02
3.00489873e-01 3.28428787e-03 -1.11012205e-01 -1.10265505e+00
-1.89970918e-02 -9.52412605e-01 9.89556849e-01 -1.55483082e-01
-1.26993799e+00 2.30713382e-01 2.85855979e-01 -1.24446404e+00
-1.46968827e-01 -4.02278870e-01 -4.55276549e-01 4.88893121e-01
-1.33692992e+00 -5.35444379e-01 -8.21283758e-02 2.12089151e-01
8.01241517e-01 -1.53467521e-01 6.78564489e-01 2.25738809e-01
-7.47022986e-01 7.44500875e-01 2.08165899e-01 -1.74127333e-02
5.28907299e-01 -1.58640265e+00 -4.45973761e-02 7.27730691e-01
6.17705584e-01 5.40058851e-01 6.99005544e-01 -1.14904121e-01
-1.30728042e+00 -6.34350717e-01 5.72969556e-01 -2.28586912e-01
5.47127604e-01 -2.77074784e-01 -7.98375964e-01 6.65596902e-01
-2.11284801e-01 1.58457607e-01 8.22673500e-01 3.51724923e-01
-9.14858580e-02 -3.60669017e-01 -1.19083762e+00 4.40922558e-01
8.90707850e-01 -3.90334100e-01 -2.49361128e-01 5.70369720e-01
5.92097402e-01 -7.37326682e-01 -5.36743283e-01 3.39935392e-01
3.46310824e-01 -9.98756588e-01 8.03977668e-01 -4.81925815e-01
-1.95641652e-01 2.94603687e-02 2.37544142e-02 -1.08652079e+00
7.78102502e-02 -8.03362906e-01 -4.31040794e-01 8.83836269e-01
6.42817497e-01 -6.77538395e-01 1.34788287e+00 7.72422016e-01
2.80050516e-01 -1.32796872e+00 -7.96014428e-01 -7.21807063e-01
-1.81727260e-01 -4.35228556e-01 4.75053936e-01 8.87786090e-01
-3.62983942e-01 3.22263509e-01 -1.61674261e-01 4.95834202e-02
5.38680196e-01 5.16444325e-01 6.79659426e-01 -1.23999012e+00
-5.39483607e-01 -3.68407458e-01 -3.69537622e-02 -1.58760560e+00
1.53310612e-01 -5.80291152e-01 1.80014133e-01 -1.09499097e+00
3.46812308e-01 -1.11220169e+00 -6.31838500e-01 5.84779024e-01
-5.07567525e-01 -7.46527091e-02 2.76498884e-01 4.16771442e-01
-1.08662355e+00 6.19420186e-02 9.72651720e-01 -2.98205316e-01
-3.33114833e-01 6.06608033e-01 -7.92067349e-01 4.49159503e-01
7.25012600e-01 -7.81656563e-01 -8.30518246e-01 -3.13726753e-01
4.97499347e-01 2.29041561e-01 -4.85768616e-02 -5.97546756e-01
4.07777399e-01 -7.48044014e-01 -4.21571322e-02 -4.84746665e-01
6.58966973e-02 -8.49859416e-01 -1.63641632e-01 4.77997631e-01
-9.52208519e-01 -3.16227257e-01 -1.83571488e-01 7.24847674e-01
-5.38497083e-02 -8.51362109e-01 8.69302630e-01 -2.00913578e-01
-5.52784562e-01 7.23083198e-01 -5.66224456e-02 4.08490658e-01
1.08255589e+00 -9.44220647e-02 -2.43154820e-03 -3.99954826e-01
-8.05101395e-01 6.46859527e-01 4.40616637e-01 -5.17513379e-02
3.28263640e-01 -9.81762171e-01 -4.59746242e-01 -4.25909050e-02
2.35383078e-01 3.31133068e-01 -2.96629846e-01 6.37537837e-01
-4.62931901e-01 3.67199838e-01 2.54433125e-01 -6.65846467e-01
-1.51848137e+00 6.15529835e-01 6.01107895e-01 -5.24669588e-01
-3.00175190e-01 1.26934886e+00 1.03754722e-01 -2.33384803e-01
6.85269952e-01 -5.05319377e-03 -3.36778641e-01 2.22126931e-01
2.07355499e-01 2.56837398e-01 2.08442912e-01 -1.12487033e-01
-1.16988309e-01 8.16270933e-02 -2.40316004e-01 -1.88830525e-01
1.12490261e+00 -3.09181869e-01 2.87310660e-01 7.62764812e-01
8.78597260e-01 8.92894715e-02 -1.46439290e+00 -7.88507581e-01
4.84403342e-01 -7.09810078e-01 -3.95744890e-02 -5.68291545e-01
-1.17437851e+00 4.30960506e-01 6.21913612e-01 6.47644341e-01
1.22114420e+00 3.57390419e-02 3.40092838e-01 8.36840451e-01
5.70318580e-01 -1.27402663e+00 1.79656252e-01 1.49570435e-01
4.64127123e-01 -1.34748650e+00 3.20531487e-01 -4.08843279e-01
-5.92652082e-01 1.05771756e+00 6.56356990e-01 3.65728773e-02
5.69166839e-01 1.19018950e-01 4.21772242e-01 -2.55277604e-01
-1.02385700e+00 -1.83444470e-01 2.72554696e-01 4.12430689e-02
2.38737389e-01 8.80140811e-02 -3.56092453e-01 8.76225382e-02
1.86946988e-02 -1.65404826e-01 2.19544649e-01 1.25141895e+00
-6.03926003e-01 -1.37023509e+00 -3.91550601e-01 5.38417816e-01
-5.22823036e-01 3.30989361e-01 -5.97371042e-01 6.00342274e-01
1.71033725e-01 8.28325808e-01 -1.65258333e-01 2.91507334e-01
8.19583684e-02 5.93661442e-02 7.70010769e-01 -1.10303605e+00
-4.63005394e-01 9.99722928e-02 9.13934335e-02 -2.61249125e-01
-8.18917811e-01 -6.52073979e-01 -1.18882430e+00 9.00089145e-02
-7.54525244e-01 4.10311818e-01 3.24614733e-01 1.06876349e+00
-1.69233471e-01 -4.53299545e-02 1.10527241e+00 -1.01719029e-01
-1.12771201e+00 -5.41187048e-01 -5.04094958e-01 1.25413358e-01
2.81613141e-01 -7.65386462e-01 -5.09877861e-01 -1.57372326e-01] | [9.312766075134277, 4.05462646484375] |
9caf3d7c-06f1-4803-8436-a6237c0597b3 | transmission-guided-bayesian-generative-model | 2303.00900 | null | https://arxiv.org/abs/2303.00900v1 | https://arxiv.org/pdf/2303.00900v1.pdf | Transmission-Guided Bayesian Generative Model for Smoke Segmentation | Smoke segmentation is essential to precisely localize wildfire so that it can be extinguished in an early phase. Although deep neural networks have achieved promising results on image segmentation tasks, they are prone to be overconfident for smoke segmentation due to its non-rigid shape and transparent appearance. This is caused by both knowledge level uncertainty due to limited training data for accurate smoke segmentation and labeling level uncertainty representing the difficulty in labeling ground-truth. To effectively model the two types of uncertainty, we introduce a Bayesian generative model to simultaneously estimate the posterior distribution of model parameters and its predictions. Further, smoke images suffer from low contrast and ambiguity, inspired by physics-based image dehazing methods, we design a transmission-guided local coherence loss to guide the network to learn pair-wise relationships based on pixel distance and the transmission feature. To promote the development of this field, we also contribute a high-quality smoke segmentation dataset, SMOKE5K, consisting of 1,400 real and 4,000 synthetic images with pixel-wise annotation. Experimental results on benchmark testing datasets illustrate that our model achieves both accurate predictions and reliable uncertainty maps representing model ignorance about its prediction. Our code and dataset are publicly available at: https://github.com/redlessme/Transmission-BVM. | ['Nick Barnes', 'Jing Zhang', 'Siyuan Yan'] | 2023-03-02 | null | null | null | null | ['image-dehazing'] | ['computer-vision'] | [ 5.05264640e-01 -6.49705529e-02 -7.34042153e-02 -4.01668519e-01
-8.56963098e-01 -5.17347395e-01 4.44217801e-01 -3.69226694e-01
-1.60161868e-01 6.15564466e-01 -1.12258665e-01 -2.89781868e-01
7.09612761e-03 -1.06848836e+00 -7.68092573e-01 -9.50684607e-01
1.92956835e-01 3.14882934e-01 5.32357275e-01 3.92971426e-01
4.35155071e-02 1.10989571e-01 -1.01083922e+00 9.83679667e-02
1.17516458e+00 1.10170698e+00 5.35721898e-01 5.15106797e-01
-3.35433900e-01 7.52836764e-01 -2.22724542e-01 -4.80942614e-02
4.54754889e-01 -4.28217769e-01 -6.71998024e-01 -5.27338162e-02
4.58191454e-01 -6.07214332e-01 -3.83085281e-01 1.48242283e+00
1.49664611e-01 -7.84199312e-03 8.20196629e-01 -1.29048073e+00
-6.62795365e-01 6.28396153e-01 -6.63700879e-01 -6.89688995e-02
-3.12287569e-01 5.86297035e-01 7.55700946e-01 -5.39798319e-01
1.99007511e-01 9.67441022e-01 9.02503788e-01 5.50691307e-01
-1.14138365e+00 -9.44375217e-01 1.24980457e-01 1.56056568e-01
-1.62510765e+00 -1.49205804e-01 1.01031542e+00 -5.79709828e-01
2.45573029e-01 3.01075667e-01 7.68988311e-01 1.19407392e+00
1.63248822e-01 3.14754874e-01 1.37647903e+00 -3.04738246e-03
3.33321244e-01 7.79703557e-02 -9.36676189e-02 9.39267516e-01
3.28029037e-01 6.15021408e-01 -2.45215669e-01 -2.84906588e-02
8.90853405e-01 3.28229293e-02 -6.14858150e-01 6.09807968e-02
-9.24644291e-01 6.73279643e-01 1.00305235e+00 -1.00777261e-01
-2.70769507e-01 5.02778709e-01 -2.68043935e-01 -3.04861605e-01
5.35516679e-01 -5.07170781e-02 -1.87755615e-01 2.79789865e-01
-1.02401614e+00 7.36488774e-02 7.05224216e-01 8.41604352e-01
8.74556422e-01 1.82624891e-01 -3.04191172e-01 6.73833489e-01
8.49807620e-01 1.00697803e+00 -1.09568939e-01 -1.44623590e+00
1.46206245e-01 1.49311677e-01 9.66207385e-02 -8.77747536e-01
-1.57515436e-01 -3.29078257e-01 -1.10045850e+00 6.70505345e-01
5.22749424e-01 -7.79248029e-02 -1.35748625e+00 1.53759909e+00
1.75521746e-01 8.09327185e-01 -2.08988905e-01 1.15307832e+00
9.20179069e-01 8.37559104e-01 2.74911195e-01 -1.22963212e-01
1.08049047e+00 -9.51965213e-01 -4.72908080e-01 -3.71633977e-01
-2.09274422e-02 -5.97392321e-01 9.80890155e-01 2.73365915e-01
-6.35447383e-01 -3.45042378e-01 -8.87024343e-01 3.92945796e-01
-1.42721131e-01 -2.39697576e-01 4.87567425e-01 6.81942642e-01
-7.22939551e-01 7.18931258e-01 -8.46540868e-01 -4.23818342e-02
1.07633114e+00 -1.49851203e-01 4.12927747e-01 -2.69105226e-01
-1.03226531e+00 4.65010822e-01 3.09513897e-01 4.75581497e-01
-1.29312241e+00 -1.13136733e+00 -3.75633180e-01 -1.61879972e-01
4.01173979e-01 -8.11115026e-01 1.10483921e+00 -5.60577631e-01
-1.23842657e+00 5.64606905e-01 2.39856347e-01 -2.56644279e-01
7.33887672e-01 6.62550777e-02 -1.21569298e-01 2.65941292e-01
2.95122340e-03 9.05001640e-01 8.00720334e-01 -1.91642570e+00
-4.61795837e-01 -6.37938976e-02 -4.13460024e-02 -4.81512062e-02
1.12555005e-01 -3.94417614e-01 -2.01788723e-01 -4.86208141e-01
1.80927128e-01 -1.02527642e+00 -4.69348967e-01 5.58805645e-01
-8.10949028e-01 2.34129623e-01 9.24779773e-01 -5.88513374e-01
6.85489655e-01 -1.78675997e+00 -4.81156290e-01 2.21613601e-01
2.66023874e-01 2.06309035e-01 -1.36373103e-01 -7.03814104e-02
3.96752268e-01 6.33575618e-01 -1.04285169e+00 -1.01106405e-01
-1.18444353e-01 4.64024067e-01 -4.83679622e-01 4.16091383e-01
5.92293106e-02 8.61991584e-01 -9.29154575e-01 -6.32843018e-01
4.47870433e-01 8.59403551e-01 -2.68382311e-01 2.03782558e-01
-6.82875931e-01 7.27171481e-01 -7.45102108e-01 6.66237891e-01
1.14945042e+00 -2.40907878e-01 -2.12268427e-01 -5.88040173e-01
-1.81867823e-01 -1.01596214e-01 -8.33758175e-01 1.47268975e+00
-4.21556443e-01 4.69310224e-01 2.82594502e-01 -5.25154948e-01
5.22915959e-01 2.20028922e-01 6.34437203e-01 -5.08995354e-01
3.71332824e-01 -1.15305848e-01 -3.55527699e-01 -5.00667930e-01
5.73962815e-02 -3.46457720e-01 3.87149036e-01 4.65052068e-01
-3.63646358e-01 -7.41592586e-01 -2.24208221e-01 9.83966365e-02
9.18601453e-01 1.36105925e-01 -5.96044004e-01 -2.53630191e-01
-8.77236202e-02 2.12170586e-01 8.10183227e-01 9.56496060e-01
-3.37393403e-01 1.05241740e+00 -1.32814636e-02 -1.28627226e-01
-8.16112161e-01 -1.44847488e+00 -2.72114933e-01 2.64216185e-01
6.22353911e-01 -1.24202088e-01 -8.68127167e-01 -7.25757241e-01
-2.11191937e-01 1.02301538e+00 -4.54058558e-01 -1.71428457e-01
-2.58620799e-01 -1.12049782e+00 5.65554321e-01 4.85437512e-01
1.12443554e+00 -8.39183271e-01 -5.58836401e-01 -7.46909156e-02
-4.31115687e-01 -1.22294617e+00 -3.19192559e-01 7.34213963e-02
-5.52910030e-01 -1.32007253e+00 -3.81854624e-01 -3.02398801e-01
6.20586634e-01 5.19736052e-01 1.18335009e+00 3.65606785e-01
-4.61860478e-01 4.35380965e-01 -2.35381782e-01 -6.21675909e-01
-6.25425518e-01 -3.23578298e-01 -5.19691646e-01 -3.60178769e-01
-7.82412812e-02 -6.92905784e-01 -1.05565047e+00 3.30265254e-01
-1.13463211e+00 2.86428571e-01 4.11835462e-01 4.32930499e-01
7.73911893e-01 6.13112926e-01 2.87291352e-02 -9.42574561e-01
3.86106431e-01 -4.62537318e-01 -7.73210347e-01 3.10877144e-01
-8.46979022e-01 -1.80073544e-01 2.83107191e-01 -1.34140939e-01
-1.46236742e+00 1.23542435e-01 -1.62452146e-01 -5.98502219e-01
-2.92547971e-01 1.90209851e-01 -3.30575407e-02 -1.60352722e-01
5.86777925e-01 2.47209705e-02 -1.51653633e-01 -2.95714766e-01
3.91025215e-01 4.00586277e-01 5.95902085e-01 -7.18274295e-01
1.28016484e+00 7.28348136e-01 7.98892751e-02 -8.73694777e-01
-1.16998541e+00 -1.21036209e-01 -4.22281533e-01 -6.90254033e-01
1.13381660e+00 -9.02698398e-01 -5.63704371e-01 8.20620239e-01
-1.13600230e+00 -8.62867296e-01 -3.17692876e-01 3.03932339e-01
-1.52917355e-01 4.58183646e-01 -4.25619215e-01 -9.26759362e-01
-4.83753920e-01 -1.15100729e+00 9.19295251e-01 3.95953506e-01
3.35399538e-01 -9.71844792e-01 -4.57482003e-02 6.37212515e-01
6.61341071e-01 3.99773389e-01 9.06402051e-01 2.61734605e-01
-1.24786675e+00 2.01947197e-01 -5.46621919e-01 7.08603144e-01
5.20522706e-02 4.19422448e-01 -1.34695697e+00 2.17137262e-01
1.24651641e-01 -2.27021083e-01 1.32953310e+00 7.02870131e-01
1.57690561e+00 -2.55213797e-01 -3.66383672e-01 6.78372324e-01
1.54144943e+00 -4.80345450e-02 6.21464312e-01 -1.49448186e-01
1.06988204e+00 4.73850816e-01 3.25424463e-01 3.35263640e-01
3.96658897e-01 3.30464900e-01 1.03394532e+00 -1.22310348e-01
-5.44583559e-01 -2.81333148e-01 1.62196785e-01 5.51585436e-01
-4.09481190e-02 -5.68986893e-01 -8.05801213e-01 7.88294896e-02
-1.63022804e+00 -1.01055026e+00 -5.66926718e-01 2.02627587e+00
9.24086154e-01 1.86870545e-02 -2.64415085e-01 -2.86237180e-01
6.48900867e-01 3.48316818e-01 -6.55831635e-01 3.66407216e-01
-1.93094406e-02 -4.05977033e-02 8.06234002e-01 7.87903070e-01
-1.12264740e+00 9.47300315e-01 6.21039915e+00 1.02722001e+00
-1.08951020e+00 2.23261818e-01 8.45046878e-01 1.54996350e-01
-8.59463394e-01 1.82122111e-01 -4.55128282e-01 7.63879061e-01
4.67207521e-01 3.55527788e-01 7.00949728e-01 3.99427980e-01
4.48329061e-01 -3.73027980e-01 -4.08966631e-01 6.05883896e-01
-3.49914670e-01 -1.46126199e+00 -1.00324586e-01 -2.60225944e-02
9.05148745e-01 5.22055686e-01 3.17056738e-02 -9.64012071e-02
6.36575818e-01 -1.06518114e+00 7.68005848e-01 8.33142459e-01
6.70003712e-01 -3.09884220e-01 3.88968676e-01 5.07256329e-01
-1.14748883e+00 1.87977374e-01 -4.90463614e-01 2.86407471e-01
2.58075386e-01 1.16649723e+00 -7.53393054e-01 3.71239692e-01
8.11640978e-01 4.44365829e-01 -2.91315198e-01 1.05229568e+00
-5.94143450e-01 1.16262650e+00 -7.04233766e-01 2.79172122e-01
7.20631182e-02 -7.19415426e-01 4.96283144e-01 9.87751603e-01
3.80502760e-01 2.42849097e-01 4.32981580e-01 1.73298287e+00
-4.48279940e-02 -5.53257942e-01 -3.77253056e-01 7.02990070e-02
5.40805757e-01 1.37778473e+00 -9.98259306e-01 -5.75909344e-03
-1.22228011e-01 7.67566204e-01 -1.31049618e-01 6.87932611e-01
-1.11532807e+00 1.73769996e-01 6.34586990e-01 2.16286927e-01
-7.97531307e-02 -2.33879685e-01 -4.39087778e-01 -8.07825327e-01
-1.73381865e-01 -2.06889167e-01 3.76936346e-02 -1.11194777e+00
-1.54897022e+00 5.33053041e-01 1.02178186e-01 -8.77872467e-01
3.07053894e-01 -5.62610447e-01 -1.06439614e+00 7.04289854e-01
-1.78518927e+00 -1.31246483e+00 -8.64507377e-01 1.96070954e-01
2.88087279e-01 4.42022532e-01 7.17920482e-01 2.49581903e-01
-5.54518819e-01 -7.01064616e-02 -3.85196209e-02 4.48306948e-02
5.17972171e-01 -1.14380848e+00 1.11152709e-01 8.49896669e-01
2.53677428e-01 6.35226518e-02 7.35194802e-01 -6.88218653e-01
-9.19585884e-01 -1.55429423e+00 4.13525589e-02 -5.16956627e-01
4.36054558e-01 -1.16535723e-01 -9.47216630e-01 2.59558469e-01
1.62031397e-01 2.57063121e-01 3.05699348e-01 -4.51719463e-01
-3.34826231e-01 -1.46187827e-01 -1.38222241e+00 4.71356899e-01
1.17448843e+00 -4.65900928e-01 2.52290145e-02 5.67789793e-01
8.01500618e-01 -2.83583134e-01 -5.93403578e-01 6.39728010e-01
3.57617289e-01 -1.19107544e+00 1.00712323e+00 2.26714998e-01
4.87214208e-01 -5.74851930e-01 -1.29943222e-01 -1.24208331e+00
-2.59227782e-01 -1.89302638e-01 8.96778107e-02 1.50388694e+00
3.16508919e-01 -3.18243593e-01 8.73396456e-01 8.72240245e-01
-1.86215132e-01 -6.85412169e-01 -9.21162844e-01 -5.38388193e-01
1.47965923e-01 -8.16435039e-01 5.00973165e-01 8.42128813e-01
-9.81232405e-01 -7.30941445e-02 -2.86932319e-01 6.70414090e-01
1.10863531e+00 2.52720147e-01 2.84996569e-01 -1.17901123e+00
-6.25239909e-02 -4.86024290e-01 2.14631036e-01 -9.90030527e-01
2.04232454e-01 -9.49173391e-01 6.52105927e-01 -1.70747387e+00
4.67315346e-01 -8.38699698e-01 -2.95310795e-01 4.61761355e-01
-6.28716722e-02 4.58217770e-01 1.82512514e-02 1.64106622e-01
-3.64381403e-01 7.54983127e-01 1.46578968e+00 -5.45994163e-01
1.32983878e-01 2.11756349e-01 -6.14976227e-01 9.88546968e-01
1.04836380e+00 -8.40886056e-01 -5.68305194e-01 -5.96758962e-01
5.12400866e-02 -2.18943939e-01 1.10740316e+00 -1.11655927e+00
1.08280607e-01 -5.94322503e-01 2.99249202e-01 -5.77322662e-01
3.97768438e-01 -1.11665034e+00 4.34998184e-01 3.85157228e-01
-2.34324083e-01 -7.78254271e-01 3.17786983e-03 8.20461273e-01
1.91773891e-01 -4.11454260e-01 9.45695102e-01 -2.92034358e-01
-6.65605843e-01 7.49155521e-01 -2.43213132e-01 1.88461602e-01
7.57854164e-01 -3.23806442e-02 -5.85255861e-01 -3.78080010e-01
-1.41193166e-01 1.66670397e-01 4.97227669e-01 9.52546820e-02
5.85519910e-01 -1.14451957e+00 -5.95399678e-01 -7.54105523e-02
-1.53206006e-01 5.78082323e-01 5.18038571e-01 5.62464118e-01
-7.06361234e-01 -1.81190357e-01 2.64420480e-01 -7.52710581e-01
-9.21875298e-01 4.40183803e-02 6.79272771e-01 1.20133743e-01
-7.06238329e-01 1.03716505e+00 4.09782469e-01 -4.66361314e-01
2.97259912e-02 -6.54671013e-01 3.70226383e-01 -6.03598416e-01
2.56540775e-01 2.85036147e-01 -3.57451200e-01 -4.48300660e-01
-1.68067530e-01 6.23611271e-01 3.99911523e-01 1.46255553e-01
1.13573539e+00 -2.16855243e-01 -6.83146045e-02 2.27835059e-01
7.98637033e-01 -7.77212381e-02 -1.96828878e+00 -5.71377426e-02
-5.58134317e-01 -6.99219108e-01 6.60882294e-01 -1.11033654e+00
-1.61974704e+00 8.46833587e-01 8.29735577e-01 1.75909683e-01
9.30836499e-01 -2.39437316e-02 9.70924318e-01 2.04405248e-01
1.67254120e-01 -1.02694941e+00 -3.48768607e-02 2.88536429e-01
7.67781794e-01 -1.48238754e+00 7.27183521e-02 -8.47158551e-01
-4.33458149e-01 6.91906095e-01 6.90021515e-01 5.45735285e-03
1.07480621e+00 6.34000838e-01 4.97719675e-01 -2.93625563e-01
-3.96459609e-01 -1.54361781e-02 6.00703880e-02 8.51032853e-01
3.01840659e-02 1.00738958e-01 4.08484817e-01 5.24148822e-01
9.30909738e-02 -3.57998051e-02 1.95334569e-01 4.50033516e-01
-6.68608665e-01 -7.78290033e-01 -3.23109806e-01 4.90209222e-01
-1.59583673e-01 -1.45026371e-01 -2.27829680e-01 4.89431947e-01
3.75805438e-01 1.13577259e+00 -1.85098365e-01 -3.88909042e-01
-1.67742725e-02 -3.17805111e-01 5.44895887e-01 -3.75439912e-01
-8.01707283e-02 -3.14601250e-02 -1.31690025e-01 -3.98802817e-01
-6.40803158e-01 -2.26991028e-01 -1.48729968e+00 -2.72294253e-01
-3.35808218e-01 7.76126832e-02 6.21103644e-01 9.08988059e-01
-5.84427230e-02 7.14233220e-01 4.99656498e-01 -9.62456048e-01
-5.79671264e-01 -7.05620110e-01 -6.99174047e-01 1.73954934e-01
6.21751621e-02 -5.20412862e-01 -5.77298105e-01 4.86379936e-02] | [10.20374584197998, -2.498854637145996] |
08d7d2d4-3b2a-425b-9fb9-38b9a96d179b | semi-siamese-network-for-robust-change | 2212.08583 | null | https://arxiv.org/abs/2212.08583v1 | https://arxiv.org/pdf/2212.08583v1.pdf | Semi-Siamese Network for Robust Change Detection Across Different Domains with Applications to 3D Printing | Automatic defect detection for 3D printing processes, which shares many characteristics with change detection problems, is a vital step for quality control of 3D printed products. However, there are some critical challenges in the current state of practice. First, existing methods for computer vision-based process monitoring typically work well only under specific camera viewpoints and lighting situations, requiring expensive pre-processing, alignment, and camera setups. Second, many defect detection techniques are specific to pre-defined defect patterns and/or print schematics. In this work, we approach the automatic defect detection problem differently using a novel Semi-Siamese deep learning model that directly compares a reference schematic of the desired print and a camera image of the achieved print. The model then solves an image segmentation problem, identifying the locations of defects with respect to the reference frame. Unlike most change detection problems, our model is specially developed to handle images coming from different domains and is robust against perturbations in the imaging setup such as camera angle and illumination. Defect localization predictions were made in 2.75 seconds per layer using a standard MacBookPro, which is comparable to the typical tens of seconds or less for printing a single layer on an inkjet-based 3D printer, while achieving an F1-score of more than 0.9. | ['Qian Yang', 'Anson W. K. Ma', 'Ethan Chadwick', 'Yushuo Niu'] | 2022-12-16 | null | null | null | null | ['change-detection', 'defect-detection'] | ['computer-vision', 'computer-vision'] | [ 4.80716676e-01 -4.10146356e-01 5.87022841e-01 -1.10306829e-01
-6.22998774e-01 -7.03521967e-01 4.06939894e-01 4.30623531e-01
-1.26662984e-01 -3.71174701e-02 -5.38503170e-01 -8.18668529e-02
3.01364530e-02 -6.13697112e-01 -8.32998395e-01 -5.60107708e-01
3.29887182e-01 7.66756833e-01 4.55969959e-01 6.64145872e-02
7.21062243e-01 8.37477207e-01 -1.48550129e+00 6.46395758e-02
6.36564851e-01 1.31234586e+00 4.62763309e-01 1.04770064e+00
-3.03845853e-01 9.96753424e-02 -8.32176507e-01 -2.46128455e-01
5.31966269e-01 -3.42033178e-01 -2.66461700e-01 8.46257448e-01
8.23197544e-01 -1.73232988e-01 -2.54261196e-01 1.32250583e+00
5.41553259e-01 -1.71850175e-01 5.73007584e-01 -8.32093298e-01
-5.87159872e-01 -3.12895596e-01 -5.50522566e-01 1.11774243e-01
4.21557128e-01 4.28996831e-01 5.57735026e-01 -7.50675321e-01
5.26728928e-01 1.05982280e+00 6.26569569e-01 2.64879525e-01
-1.25755215e+00 -1.40065551e-01 1.70355812e-01 -2.51289904e-01
-9.47581232e-01 -2.02959523e-01 1.18085444e+00 -7.74829388e-01
6.87797070e-01 1.64408132e-01 7.19311774e-01 8.53159666e-01
6.33543611e-01 5.06732941e-01 1.23846459e+00 -4.79150712e-01
5.73840797e-01 -1.56778842e-01 -4.75916378e-02 9.14515674e-01
2.09740132e-01 -9.48821008e-02 -4.92527008e-01 1.75720751e-01
1.50290132e+00 7.86548927e-02 -2.68094450e-01 -7.95398414e-01
-1.35865271e+00 5.18063128e-01 -1.33720279e-01 2.64396697e-01
-1.48656294e-01 -5.05990759e-02 1.60609558e-01 5.78033924e-01
6.43036664e-01 8.30435574e-01 -4.43991512e-01 -2.15461120e-01
-1.09760022e+00 2.10731495e-02 1.04935884e+00 9.49011981e-01
4.56884682e-01 -6.52514994e-02 6.03423342e-02 7.97267973e-01
2.66655654e-01 5.61263502e-01 2.42596939e-01 -1.05593491e+00
4.15158063e-01 5.60369194e-01 1.83416218e-01 -1.14008176e+00
-2.04112813e-01 -1.95639566e-01 -7.57770717e-01 7.77864218e-01
5.22903502e-01 5.45649156e-02 -1.14311278e+00 7.42130458e-01
1.85943916e-01 -6.28213510e-02 -4.66377527e-01 8.38934541e-01
2.71586001e-01 4.00464803e-01 -8.64025712e-01 -1.79940209e-01
9.01327729e-01 -8.50711405e-01 -8.00704122e-01 -3.52976143e-01
6.00417070e-02 -1.18076205e+00 1.30376184e+00 7.22214758e-01
-1.13301647e+00 -6.19553983e-01 -1.26130605e+00 1.65249348e-01
-2.72634774e-01 2.84816951e-01 6.72030970e-02 5.13104916e-01
-9.31976616e-01 8.35523307e-01 -9.68466818e-01 -5.19753098e-01
3.25432807e-01 3.45827103e-01 -2.04757199e-01 -2.36484349e-01
-3.36510807e-01 8.56746852e-01 -3.30659360e-01 2.75916010e-01
-7.94386208e-01 -6.97854459e-01 -7.29659259e-01 -3.00683290e-01
4.24796730e-01 -2.35079169e-01 1.16392803e+00 -7.51208782e-01
-2.05695271e+00 9.64775383e-01 -1.09674707e-01 8.91856328e-02
1.00421917e+00 -5.05273402e-01 -3.30204457e-01 8.72630104e-02
-1.34026647e-01 9.67024341e-02 1.28663099e+00 -1.58983862e+00
-4.64213014e-01 -4.74653095e-01 -1.55158430e-01 7.75511786e-02
2.22653210e-01 -1.49473384e-01 -1.13389850e+00 -6.55288398e-01
5.45693040e-01 -7.41316855e-01 -2.52192616e-01 5.91307282e-01
-4.87919807e-01 2.86871701e-01 1.02630162e+00 -8.20502400e-01
5.88242471e-01 -2.15889430e+00 5.95632307e-02 3.32885355e-01
1.72788072e-02 1.50602475e-01 -2.45723218e-01 2.19845399e-01
1.51840672e-01 -2.88381130e-01 -4.17049497e-01 -4.48843181e-01
-7.86854401e-02 -8.44702497e-02 1.10109523e-01 6.93438888e-01
2.77475536e-01 3.95088404e-01 -5.96625745e-01 -4.41493094e-02
5.81931829e-01 7.16712475e-02 -2.09331229e-01 1.96172595e-01
-4.55658942e-01 6.09463692e-01 -8.87305569e-03 1.02597964e+00
8.74967515e-01 1.19314350e-01 -9.09620300e-02 -2.57524401e-01
-1.15312725e-01 -1.61370456e-01 -1.53933442e+00 1.71787143e+00
-4.81486261e-01 7.97510684e-01 3.10086519e-01 -8.01525533e-01
1.23621428e+00 -8.03625435e-02 6.48000598e-01 -6.51793838e-01
1.32408198e-02 1.49931803e-01 -2.61804104e-01 -6.25882328e-01
3.23413074e-01 1.39235780e-01 1.68088913e-01 4.27095294e-01
-1.05657004e-01 -9.32459950e-01 1.29586816e-01 -2.88888216e-01
1.32908082e+00 9.64912996e-02 -2.24991456e-01 1.97309982e-02
3.20563555e-01 6.15135022e-02 4.37077194e-01 6.84619248e-01
-2.98349380e-01 1.16704762e+00 5.87954462e-01 -3.98643404e-01
-1.17941773e+00 -1.12974381e+00 1.03729829e-01 2.03658119e-01
7.19123423e-01 -1.78405959e-02 -6.26396954e-01 -7.51350403e-01
2.16060326e-01 3.91015053e-01 -3.34435880e-01 -4.00825590e-02
-5.68364322e-01 -2.34345615e-01 -2.83394381e-02 3.07268471e-01
5.69244504e-01 -7.23246813e-01 -5.14081657e-01 2.28505850e-01
4.40193295e-01 -1.11330211e+00 -5.79428315e-01 2.23495573e-01
-9.65562046e-01 -1.35210121e+00 -8.28371286e-01 -9.49892700e-01
9.37711477e-01 2.69657522e-01 1.10029566e+00 -3.94030899e-01
-5.33179283e-01 7.55304635e-01 -6.24573827e-02 -3.39353204e-01
-5.39973795e-01 -5.14980614e-01 7.78178349e-02 -1.38717741e-01
9.98868495e-02 -1.59761980e-01 -4.37558472e-01 4.51433957e-01
-7.55398393e-01 -8.23376402e-02 7.76887476e-01 6.98790371e-01
9.61119890e-01 6.75106943e-01 -2.19137684e-01 -6.43733561e-01
6.40182436e-01 1.33017808e-01 -8.81295860e-01 1.96808934e-01
-5.25269032e-01 -2.18186557e-01 4.62668061e-01 -5.71260571e-01
-1.01428652e+00 3.92945677e-01 2.21541435e-01 -5.97498953e-01
-5.19588530e-01 2.13441133e-01 -3.55508208e-01 -3.01437199e-01
3.05739701e-01 2.02712372e-01 4.76938665e-01 -5.97082198e-01
-3.43741998e-02 4.53302741e-01 4.29189384e-01 -2.32471600e-01
1.06855118e+00 4.77992475e-01 -5.73091544e-02 -1.12635565e+00
-5.38915455e-01 -4.16117787e-01 -1.01619589e+00 -4.95665908e-01
6.67041600e-01 -5.13004839e-01 -5.12687564e-01 1.14666164e+00
-1.22364807e+00 -5.55319726e-01 -4.25032288e-01 3.37107599e-01
-6.89857721e-01 5.67868710e-01 -7.03437626e-01 -5.62954187e-01
-4.30451632e-02 -1.24825561e+00 1.34799469e+00 1.05901651e-01
-4.87377830e-02 -9.15403485e-01 8.45955033e-03 5.05608857e-01
2.95175314e-01 2.12127864e-01 1.04866600e+00 -9.56719369e-02
-6.61945105e-01 -5.06125748e-01 -1.01070330e-02 7.77435780e-01
6.34717822e-01 2.13498741e-01 -7.35988915e-01 -1.83682695e-01
4.20007408e-01 1.22016892e-01 5.09384394e-01 4.33652043e-01
1.12629640e+00 1.55906558e-01 -1.34671167e-01 2.15475574e-01
1.50698590e+00 4.13081080e-01 2.76190728e-01 2.49379992e-01
7.82505929e-01 4.82440382e-01 8.48184943e-01 2.92706847e-01
-1.70103535e-01 6.49893165e-01 4.77207452e-01 -2.11133301e-01
-2.75824398e-01 9.86888539e-03 5.10915697e-01 7.31090546e-01
2.06242651e-01 -1.41046360e-01 -8.50349545e-01 2.90028870e-01
-1.50614929e+00 -3.81830722e-01 -1.87429562e-01 2.24926496e+00
5.75578511e-01 3.46406847e-01 -2.48551056e-01 2.24175125e-01
8.35051239e-01 8.23726729e-02 -9.07489836e-01 -3.53703290e-01
-3.03875394e-02 2.00175136e-01 6.21385634e-01 3.83189589e-01
-1.25400770e+00 7.92207599e-01 6.33647013e+00 3.54972005e-01
-1.23486066e+00 -2.88596004e-01 3.14452529e-01 -1.67735115e-01
7.31408373e-02 -3.64483237e-01 -4.36878115e-01 5.19472957e-01
4.04940844e-01 3.47448945e-01 4.65053141e-01 4.89506006e-01
3.02492470e-01 -4.82360482e-01 -1.41494346e+00 1.24457800e+00
5.34063756e-01 -1.12816370e+00 -6.64498359e-02 1.12714052e-01
8.80315006e-01 -3.09030831e-01 2.47773126e-01 -3.86642039e-01
2.86995351e-01 -7.00574875e-01 9.67571318e-01 6.95393443e-01
5.84558249e-01 -4.55173701e-01 7.58851588e-01 1.23596013e-01
-6.43734574e-01 4.40926552e-02 -3.87100995e-01 5.88116534e-02
1.78999007e-01 1.11092544e+00 -7.33347774e-01 1.50049746e-01
7.94271708e-01 6.65259659e-01 -5.20747900e-01 1.18233490e+00
-8.46652836e-02 3.32835019e-01 -3.72400016e-01 2.67041773e-01
9.25096776e-03 -3.90680611e-01 6.27122223e-01 8.36621046e-01
6.93538189e-01 -4.50990945e-01 8.99497867e-02 8.14493179e-01
-1.17647145e-02 -3.11725259e-01 -6.36619747e-01 -8.29845965e-02
1.60965726e-01 9.16883528e-01 -9.99511242e-01 1.70221850e-01
-4.41570103e-01 1.68630266e+00 -1.85429513e-01 4.43797916e-01
-5.51399112e-01 -3.46991569e-01 7.87788510e-01 2.22806796e-01
6.10482097e-01 -7.69173265e-01 -6.62892103e-01 -8.74151409e-01
3.61847162e-01 -8.95146072e-01 -2.16522947e-01 -6.58937991e-01
-1.37105489e+00 1.84511095e-02 -6.52976274e-01 -1.44651949e+00
5.09880364e-01 -1.11944032e+00 -7.07409024e-01 6.57352030e-01
-1.35129225e+00 -9.04584825e-01 -4.69448149e-01 3.85389239e-01
1.08456421e+00 -1.29399955e-01 6.19078040e-01 4.66746166e-02
-6.91541374e-01 1.70035556e-01 5.15913785e-01 -5.11137322e-02
9.05519664e-01 -1.44829237e+00 5.48952401e-01 1.01878154e+00
1.08635761e-01 6.63259625e-02 6.93170011e-01 -8.03175986e-01
-1.75526786e+00 -1.00227475e+00 3.41233462e-01 -4.99156296e-01
5.14196098e-01 -4.91704047e-01 -7.83227146e-01 2.38585070e-01
-3.49465013e-02 -8.87195915e-02 2.92250752e-01 -1.84556425e-01
1.44934550e-01 -1.66534424e-01 -1.19263518e+00 4.60870504e-01
1.00769174e+00 -4.22948450e-01 -3.14806610e-01 5.24766207e-01
3.67447108e-01 -8.13902855e-01 -8.65190327e-01 5.03790826e-02
3.09066087e-01 -9.19551909e-01 6.96020424e-01 -4.37472155e-03
5.93554497e-01 -5.83370626e-01 -1.31492078e-01 -1.42881417e+00
-1.86252728e-01 -5.61781824e-01 -2.13097587e-01 1.18995726e+00
1.70434043e-01 -2.72081107e-01 8.78601372e-01 6.26100004e-01
-2.47092828e-01 -5.34461558e-01 -8.54834139e-01 -9.04467285e-01
-2.78291792e-01 -4.51077193e-01 1.72744170e-01 7.14729488e-01
-5.26052892e-01 -6.13062344e-02 1.34104602e-02 6.06440961e-01
6.68289125e-01 3.38073313e-01 7.32332230e-01 -1.51209629e+00
-2.78704971e-01 -4.22470689e-01 -7.47032583e-01 -1.03397024e+00
-1.49051562e-01 -3.29666138e-01 4.00913596e-01 -1.61090958e+00
-2.81202942e-01 -4.63035464e-01 1.92796905e-02 2.54342053e-02
1.55491158e-01 7.25442544e-02 1.24148550e-02 2.21472114e-01
-3.55657160e-01 3.86417240e-01 1.42481756e+00 -6.79461598e-01
-4.26783651e-01 3.65461200e-01 -4.03917760e-01 9.00956094e-01
6.55842185e-01 -7.25027546e-02 -1.17820270e-01 -8.50526690e-01
-8.43679067e-03 -1.38688669e-01 2.95920163e-01 -1.17574239e+00
4.53709096e-01 -6.08485639e-02 6.57052934e-01 -5.35915852e-01
3.34904134e-01 -1.09581852e+00 -2.16750074e-02 4.51554298e-01
-1.80628389e-01 -7.62833729e-02 1.45186752e-01 8.44397962e-01
-2.82782704e-01 -3.99377942e-01 8.01178932e-01 -3.18238497e-01
-8.11224043e-01 2.27686882e-01 -4.56914186e-01 -4.05018985e-01
1.28802466e+00 -6.68998897e-01 -2.48229504e-02 -1.37075886e-01
-5.84207535e-01 -6.90356791e-02 9.44283724e-01 4.46233571e-01
9.41018403e-01 -9.91034806e-01 -2.62736708e-01 5.93578637e-01
2.14584786e-02 3.99405211e-01 2.48751521e-01 5.58800399e-01
-1.00309753e+00 9.88011435e-02 -5.08886650e-02 -9.43111837e-01
-1.27857316e+00 3.24398041e-01 3.99757922e-01 -2.71585025e-03
-8.95431817e-01 8.51539612e-01 -2.63360143e-01 -4.16789740e-01
2.90108711e-01 -7.24542022e-01 3.13839346e-01 -1.79180190e-01
1.68776110e-01 4.87856597e-01 5.49362957e-01 -1.72219098e-01
-8.38200226e-02 1.02936149e+00 -1.41894519e-01 -1.25832453e-01
1.27358055e+00 -4.21687886e-02 3.19084562e-02 6.55025959e-01
9.08170938e-01 1.66371632e-02 -1.90338111e+00 -1.16371937e-01
-4.63243127e-02 -6.89626753e-01 2.55889803e-01 -8.53075445e-01
-1.38308823e+00 8.87194395e-01 1.06011236e+00 1.79932907e-01
9.01282728e-01 1.69109758e-02 7.25646436e-01 3.71165276e-01
2.56379575e-01 -1.53064477e+00 5.12674809e-01 5.62827706e-01
1.00950623e+00 -1.34668541e+00 -1.68863721e-02 -4.40364122e-01
-5.59900343e-01 1.32024539e+00 6.52297914e-01 -1.45482391e-01
6.78799808e-01 4.66032714e-01 3.18033218e-01 -4.21822667e-01
-6.37901872e-02 1.93199694e-01 9.11524296e-02 7.14546204e-01
5.72913848e-02 -2.86863416e-01 3.84963036e-01 7.81161636e-02
1.85453817e-01 -2.40356654e-01 5.41448414e-01 1.31068337e+00
-4.80150223e-01 -8.44223559e-01 -5.74879348e-01 6.10983014e-01
-5.36149889e-02 4.03055251e-01 -5.99602580e-01 4.80141610e-01
1.30571201e-01 8.82201016e-01 4.79267895e-01 -2.43351266e-01
8.00594151e-01 -4.41695042e-02 7.82093644e-01 -7.07645833e-01
-1.55303076e-01 1.24490701e-01 -3.42835367e-01 -8.78965616e-01
-3.16663593e-01 -9.65993881e-01 -9.16119218e-01 2.77680494e-02
-3.94069761e-01 -5.27738988e-01 9.77450371e-01 7.37622321e-01
3.25299591e-01 6.42901123e-01 7.27843642e-01 -1.10276163e+00
-3.55330676e-01 -7.29562402e-01 -1.10346878e+00 2.59057999e-01
2.45467320e-01 -6.06501579e-01 -4.18422163e-01 3.09809506e-01] | [7.387163162231445, 1.802209496498108] |
c7f57a52-bad1-46aa-8c01-2a5ae19ef896 | graph-based-semantical-extractive-text | 2212.09701 | null | https://arxiv.org/abs/2212.09701v1 | https://arxiv.org/pdf/2212.09701v1.pdf | Graph-based Semantical Extractive Text Analysis | In the past few decades, there has been an explosion in the amount of available data produced from various sources with different topics. The availability of this enormous data necessitates us to adopt effective computational tools to explore the data. This leads to an intense growing interest in the research community to develop computational methods focused on processing this text data. A line of study focused on condensing the text so that we are able to get a higher level of understanding in a shorter time. The two important tasks to do this are keyword extraction and text summarization. In keyword extraction, we are interested in finding the key important words from a text. This makes us familiar with the general topic of a text. In text summarization, we are interested in producing a short-length text which includes important information about the document. The TextRank algorithm, an unsupervised learning method that is an extension of the PageRank (algorithm which is the base algorithm of Google search engine for searching pages and ranking them) has shown its efficacy in large-scale text mining, especially for text summarization and keyword extraction. this algorithm can automatically extract the important parts of a text (keywords or sentences) and declare them as the result. However, this algorithm neglects the semantic similarity between the different parts. In this work, we improved the results of the TextRank algorithm by incorporating the semantic similarity between parts of the text. Aside from keyword extraction and text summarization, we develop a topic clustering algorithm based on our framework which can be used individually or as a part of generating the summary to overcome coverage problems. | ['Mina Samizadeh'] | 2022-12-19 | null | null | null | null | ['keyword-extraction'] | ['natural-language-processing'] | [ 5.91477454e-01 2.48063222e-01 -3.32132876e-01 -1.79127883e-02
-5.61814487e-01 -5.16973794e-01 6.84039295e-01 9.78812099e-01
-4.64145482e-01 9.62878227e-01 7.55276799e-01 -8.43086913e-02
-4.02572125e-01 -8.30438495e-01 -1.40887722e-01 -5.53546846e-01
2.98884977e-02 5.00111520e-01 4.53882694e-01 -2.00952306e-01
8.76191616e-01 3.00624728e-01 -1.85473597e+00 3.21577698e-01
1.39988446e+00 6.43334091e-01 6.50254488e-01 4.37850326e-01
-1.02664673e+00 5.48197448e-01 -8.69689524e-01 -1.10671945e-01
-2.82908410e-01 -6.49173200e-01 -1.21017635e+00 2.63215810e-01
1.88995656e-02 1.01656087e-01 1.62539274e-01 1.12457800e+00
2.33279318e-01 1.57789469e-01 6.74476385e-01 -8.95515740e-01
2.42341489e-01 8.68016124e-01 -5.35869122e-01 9.98202041e-02
6.93131149e-01 -6.75636232e-01 1.19788086e+00 -5.36133111e-01
6.31162226e-01 1.01025701e+00 1.66524008e-01 2.81182025e-02
-6.08851135e-01 -1.33188829e-01 1.06403679e-01 1.34406611e-01
-1.13587630e+00 -1.06616534e-01 8.93163800e-01 -1.94808677e-01
8.89723599e-01 6.77677393e-01 7.37139106e-01 6.72138691e-01
2.14080468e-01 9.63979006e-01 8.61373544e-01 -9.33749259e-01
3.60406905e-01 3.80072445e-01 5.70704699e-01 2.19418138e-01
7.29326904e-01 -8.87549996e-01 -4.72092599e-01 -2.67171562e-01
-2.43334733e-02 5.07954061e-02 -3.01702023e-01 -3.88722047e-02
-9.82274115e-01 8.58455896e-01 -1.95309266e-01 8.31562996e-01
-6.06354177e-01 -4.69184428e-01 6.71168566e-01 1.70505606e-02
6.14525199e-01 7.61330426e-01 -3.36824596e-01 -2.14669392e-01
-1.34120357e+00 3.23941082e-01 1.26724088e+00 7.27489650e-01
7.70599008e-01 -3.94118041e-01 -1.74691245e-01 7.60837018e-01
2.41430178e-01 1.10948510e-01 9.09761250e-01 -6.43338740e-01
4.97630239e-01 1.27384770e+00 -8.54825694e-03 -1.25870311e+00
-3.17444831e-01 -2.80152738e-01 -7.76916206e-01 -2.66586751e-01
-1.80665419e-01 -6.02296852e-02 -6.77557111e-01 1.14922893e+00
3.20021421e-01 -4.41033602e-01 1.85719222e-01 3.85067523e-01
9.46731210e-01 8.58602107e-01 -1.69031955e-02 -7.75046289e-01
1.60905969e+00 -7.50999868e-01 -9.41021442e-01 7.31312558e-02
6.46873236e-01 -9.59512591e-01 7.30377495e-01 5.62123954e-01
-6.96467578e-01 -3.27153414e-01 -1.12490141e+00 1.16412286e-02
-8.75839353e-01 2.78411321e-02 5.69690943e-01 4.97579128e-01
-9.96191382e-01 6.22336268e-01 -4.17293787e-01 -9.32703972e-01
4.63136770e-02 2.61020094e-01 -2.21208647e-01 3.15545760e-02
-1.21458209e+00 7.37290859e-01 9.34467435e-01 -4.01669979e-01
2.31197640e-01 -2.76174039e-01 -5.32470107e-01 2.02724680e-01
8.42369080e-01 -7.40100265e-01 9.46777463e-01 -9.70007420e-01
-1.08869863e+00 5.73697805e-01 -3.42246681e-01 -4.22763586e-01
1.56162232e-01 -2.37692714e-01 -3.74773145e-01 4.67799872e-01
3.71368855e-01 3.00368845e-01 6.90394819e-01 -1.14073157e+00
-1.07457149e+00 -4.42632288e-01 -2.74641961e-01 4.23088789e-01
-6.82684302e-01 1.77719906e-01 -5.27568519e-01 -7.19842136e-01
1.09999076e-01 -5.49394727e-01 -1.16219752e-01 -9.43527341e-01
-5.98744810e-01 -8.26796770e-01 9.79408503e-01 -7.60741413e-01
1.68668890e+00 -1.77206624e+00 1.07072301e-01 2.87547231e-01
3.87766212e-01 2.93585986e-01 3.12496454e-01 1.10092485e+00
1.10546850e-01 5.06163895e-01 -3.73504281e-01 -3.81181389e-02
-2.21017510e-01 8.30502957e-02 -5.45908928e-01 -2.33367696e-01
-2.54465729e-01 4.45527971e-01 -8.88453305e-01 -1.09079266e+00
1.34045342e-02 -4.33606021e-02 -1.11776948e-01 9.21992809e-02
-2.37089366e-01 3.46041545e-02 -9.65786278e-01 4.48907912e-01
3.95424217e-01 5.24500906e-02 4.01927233e-02 -6.63707480e-02
-5.26943743e-01 3.59672487e-01 -1.17651188e+00 1.48135245e+00
-7.81679973e-02 5.72605789e-01 -2.71501631e-01 -1.24703729e+00
8.04962575e-01 3.20478171e-01 8.87917340e-01 -3.38525027e-01
1.37099251e-01 3.29617798e-01 -3.27719152e-01 -6.88750327e-01
1.11961818e+00 1.70705810e-01 -2.87435114e-01 5.78004479e-01
-1.73837304e-01 -3.00922215e-01 9.75011408e-01 6.51018679e-01
9.68848228e-01 -2.82457769e-01 8.30801249e-01 -3.37126791e-01
6.24591768e-01 4.92184460e-01 3.26336361e-02 6.13422573e-01
4.76196676e-01 2.90499061e-01 5.39337933e-01 -2.36291111e-01
-8.57732475e-01 -3.79460782e-01 4.01339680e-02 6.32920504e-01
2.09597368e-02 -9.85454440e-01 -8.54410231e-01 -5.97952962e-01
-1.67296097e-01 7.80648232e-01 -3.20470423e-01 9.84910056e-02
-3.90752196e-01 -6.55814707e-01 1.82700813e-01 2.15938706e-02
5.63595057e-01 -1.17083681e+00 -6.84718966e-01 3.34543914e-01
-5.64003348e-01 -8.99055600e-01 -1.31510615e-01 2.03864112e-01
-9.93475616e-01 -1.05106032e+00 -6.99789941e-01 -6.47716701e-01
8.49753320e-01 4.43263650e-01 7.87276328e-01 -4.14736196e-02
-1.80014357e-01 3.61825824e-01 -1.09448707e+00 -7.97372162e-01
-5.34577370e-01 6.85765266e-01 -4.39307392e-02 -1.45839125e-01
2.81504899e-01 -4.37697738e-01 -2.12072849e-01 -1.72659680e-01
-1.44229138e+00 9.54811275e-02 7.68820345e-01 3.25856715e-01
3.20877373e-01 7.02490270e-01 6.51930273e-01 -9.94917333e-01
1.24987578e+00 -4.33797479e-01 -2.22868800e-01 3.78829032e-01
-8.10435951e-01 4.35981244e-01 6.04941964e-01 -2.28833809e-01
-1.03806841e+00 -2.65919536e-01 -5.06634638e-02 3.82472366e-01
-1.72988191e-01 1.01183665e+00 -2.44646639e-01 4.71164674e-01
3.04141134e-01 5.51886320e-01 -1.82301402e-01 -6.74354315e-01
2.16962487e-01 1.26036668e+00 4.00469340e-02 -3.02026808e-01
5.23479104e-01 2.75464982e-01 -5.70667200e-02 -1.42925119e+00
-7.99780428e-01 -1.09156311e+00 -7.17977583e-01 -3.29675138e-01
7.70929992e-01 -3.99508268e-01 -2.44675651e-01 2.40348518e-01
-1.07807410e+00 4.65744406e-01 -4.55445945e-01 3.23057234e-01
-2.20213845e-01 1.00376892e+00 1.25541463e-01 -7.11547673e-01
-7.64230847e-01 -5.65758467e-01 1.02363062e+00 4.37998891e-01
-7.38392055e-01 -6.81306303e-01 2.21008062e-01 3.67792815e-01
1.56520322e-01 3.37620936e-02 1.12250245e+00 -1.33158207e+00
-2.23364085e-01 -4.09868330e-01 5.02701700e-02 3.20865393e-01
6.58914030e-01 -3.76220159e-02 -4.18839753e-01 4.38275486e-02
1.98465228e-01 1.06345356e-01 1.16549551e+00 2.86991388e-01
1.04252779e+00 -6.26654148e-01 -6.17411852e-01 -3.14779840e-02
1.22829247e+00 2.84748197e-01 6.27994120e-01 5.40737927e-01
5.61935544e-01 9.79997516e-01 7.92506635e-01 4.64248896e-01
2.12134629e-01 4.90180522e-01 1.67202190e-01 6.99535850e-03
2.19913885e-01 -1.99561656e-01 1.77209184e-01 1.13433242e+00
-5.70985153e-02 -4.69179124e-01 -6.81078792e-01 4.96667057e-01
-1.84362149e+00 -1.09809864e+00 -2.68380135e-01 2.19724512e+00
1.03230917e+00 1.42438889e-01 2.01725230e-01 6.20800674e-01
6.74114525e-01 2.00530887e-01 -1.01205185e-01 -6.44245625e-01
-1.52621558e-02 -6.33690953e-02 2.45308772e-01 1.77402630e-01
-1.03620219e+00 1.01146770e+00 5.23869610e+00 1.04570460e+00
-9.33696568e-01 -4.82672215e-01 3.73870134e-01 4.05338168e-01
-3.26226830e-01 2.05384195e-01 -9.55515146e-01 4.63307887e-01
7.55709291e-01 -6.62609696e-01 -9.81008857e-02 8.53209972e-01
4.32534933e-01 -8.77154469e-01 -6.55694067e-01 5.75185418e-01
4.86437410e-01 -1.11473799e+00 6.41458392e-01 1.45571053e-01
7.64925301e-01 -5.49697578e-01 -5.04830480e-01 -1.33500779e-02
-7.75408149e-02 -5.60283244e-01 3.72304767e-01 3.76760423e-01
3.23601991e-01 -8.64844739e-01 1.08534038e+00 7.19050348e-01
-1.11968350e+00 1.37528926e-01 -3.25539708e-01 -7.53940120e-02
7.74262026e-02 9.25379574e-01 -1.00602424e+00 1.02969360e+00
5.39954662e-01 4.79098707e-01 -5.32968402e-01 1.30828381e+00
-1.45030990e-01 4.23722237e-01 -2.10911199e-01 -5.62807620e-01
2.61482984e-01 -4.01308417e-01 7.87939787e-01 1.29045367e+00
4.02208984e-01 1.15935057e-01 3.04583251e-01 2.50066221e-01
-7.39829540e-02 8.45090568e-01 -5.55517673e-01 -3.05701315e-01
2.10202947e-01 1.38129520e+00 -1.34076977e+00 -6.74723387e-01
-1.80263042e-01 9.06449199e-01 -1.09876350e-01 3.74481007e-02
-2.36432374e-01 -9.31795180e-01 4.09082212e-02 1.79569393e-01
1.06177509e-01 -2.36586124e-01 -1.84354320e-01 -9.46123779e-01
1.66594639e-01 -8.88408184e-01 4.77788329e-01 -5.70078611e-01
-8.74907851e-01 5.68973601e-01 4.50093478e-01 -1.09097421e+00
-4.34292406e-01 -1.38741940e-01 -5.89435518e-01 6.17663682e-01
-1.27702093e+00 -6.79151773e-01 -1.71656847e-01 3.52159917e-01
1.02060628e+00 6.49955049e-02 6.95003808e-01 -1.49988249e-01
-2.64109015e-01 -1.76619962e-01 2.81803131e-01 6.17787661e-03
6.48975194e-01 -1.33893096e+00 -8.76536872e-03 8.47056866e-01
2.10396945e-01 7.11930156e-01 8.93354356e-01 -1.10522854e+00
-1.06848240e+00 -7.69775093e-01 1.50357473e+00 -3.34651284e-02
6.39119923e-01 -2.06665462e-03 -9.23935831e-01 9.60995331e-02
5.39472282e-01 -1.09486043e+00 7.49728024e-01 -1.22205764e-02
2.96519995e-01 -8.71192813e-02 -7.59310603e-01 6.06846333e-01
6.02034688e-01 -4.79230583e-02 -1.28775501e+00 2.29374647e-01
7.92236865e-01 1.86015308e-01 -4.11683291e-01 1.97439283e-01
2.58290559e-01 -5.60639322e-01 5.51270843e-01 -3.14122528e-01
3.01487565e-01 -1.96053907e-01 2.66087860e-01 -1.37693179e+00
1.82886794e-01 -6.67337000e-01 -6.34731650e-02 1.55352700e+00
3.24738711e-01 -4.96994168e-01 5.79209328e-01 3.31805497e-01
-5.39342128e-02 -5.76150060e-01 -6.31545484e-01 -4.67570335e-01
-3.68068486e-01 -1.88783750e-01 2.31226504e-01 7.02247024e-01
6.07066810e-01 6.65111899e-01 -1.48804411e-01 -3.84183973e-01
4.82403219e-01 4.03359592e-01 6.47373259e-01 -1.71385789e+00
4.13421690e-01 -6.07743979e-01 -1.39579147e-01 -9.88470852e-01
-7.90766627e-02 -7.01392531e-01 -1.24001518e-01 -2.23221040e+00
5.76642811e-01 4.19827038e-03 7.42206117e-03 2.71783471e-01
-2.87655145e-01 -4.34082121e-01 -3.97469401e-02 5.35853386e-01
-7.71327019e-01 4.45957571e-01 9.84606862e-01 -9.62307975e-02
-5.67260444e-01 3.41594964e-01 -1.05676425e+00 8.07040930e-01
9.12322819e-01 -6.05527937e-01 -4.23384547e-01 1.97917968e-01
5.96761346e-01 -3.08012038e-01 -2.97769487e-01 -9.26983535e-01
5.79800248e-01 -9.89091918e-02 1.23274930e-01 -1.16402602e+00
-2.18484357e-01 -7.80583262e-01 -2.19677255e-01 3.83760929e-01
-3.78965855e-01 5.61742708e-02 -2.91881263e-02 3.57097745e-01
-5.66139221e-01 -7.55574167e-01 1.72120795e-01 -2.99859524e-01
-6.84537590e-01 -1.13670073e-01 -8.11642885e-01 -8.18855017e-02
7.80224502e-01 -4.21410829e-01 -5.52254468e-02 -3.23282868e-01
-3.97597969e-01 2.66387641e-01 2.75418192e-01 4.90082771e-01
5.74651301e-01 -7.62702703e-01 -7.19543338e-01 -2.77289122e-01
1.23240128e-01 6.71535283e-02 1.49667785e-02 6.64877057e-01
-5.23190200e-01 9.34666693e-01 6.08254876e-03 -1.53180704e-01
-1.60234547e+00 4.56180930e-01 -5.36721587e-01 -7.34254479e-01
-5.98310590e-01 1.85797900e-01 -2.52381533e-01 1.63199499e-01
2.51201421e-01 -4.79329705e-01 -1.15657151e+00 8.25190067e-01
6.77456319e-01 4.54987168e-01 2.05817714e-01 -5.17047226e-01
-1.48618326e-01 5.22320569e-01 -2.58367300e-01 -1.09210886e-01
1.45429361e+00 -3.34540486e-01 -7.28203893e-01 5.13687551e-01
1.04042220e+00 5.15550613e-01 -2.44672254e-01 -6.85408339e-02
6.97102904e-01 -5.55751240e-03 1.52780667e-01 -5.60276031e-01
-3.23318690e-01 4.22475994e-01 -8.95688012e-02 9.13793921e-01
1.29055548e+00 2.63150126e-01 7.41500199e-01 5.63326716e-01
1.58347338e-01 -1.52267480e+00 2.82658916e-03 6.39840961e-01
7.43322313e-01 -1.08410668e+00 5.30521572e-01 -5.46953678e-01
-4.84014273e-01 1.30029333e+00 6.59892196e-03 3.63324553e-01
4.27285820e-01 5.23028113e-02 -2.59176105e-01 -2.95789570e-01
-3.87436897e-01 -5.24597645e-01 5.05265415e-01 1.70289159e-01
3.86078089e-01 -1.95896700e-01 -1.19473612e+00 5.92977524e-01
-4.12814140e-01 -2.24996507e-01 6.19283497e-01 1.06909060e+00
-1.26668334e+00 -1.43600225e+00 -4.32092279e-01 8.13569486e-01
-7.85357296e-01 -9.69491303e-02 -1.03762555e+00 7.44662702e-01
-1.30171835e-01 1.30289352e+00 -2.41996884e-01 -2.10597053e-01
2.32329547e-01 1.80421963e-01 7.33031183e-02 -9.33802605e-01
-4.64909494e-01 2.03635275e-01 2.20527619e-01 -1.23691775e-01
-7.39809155e-01 -7.06818819e-01 -1.37973201e+00 3.26668024e-02
-4.78453904e-01 1.00868452e+00 1.07906640e+00 1.12080562e+00
1.80383921e-01 5.87370396e-01 7.45169997e-01 -5.99066377e-01
-2.91612417e-01 -1.13325346e+00 -5.28917968e-01 2.19004989e-01
2.93513928e-02 -2.05438912e-01 -3.95223767e-01 1.23555601e-01] | [12.264927864074707, 9.48764419555664] |
2d7906f8-45a4-478c-ac58-d6d425e0eb9c | mazajak-an-online-arabic-sentiment-analyser | null | null | https://aclanthology.org/W19-4621 | https://aclanthology.org/W19-4621.pdf | Mazajak: An Online Arabic Sentiment Analyser | Sentiment analysis (SA) is one of the most useful natural language processing applications. Literature is flooding with many papers and systems addressing this task, but most of the work is focused on English. In this paper, we present {``}Mazajak{''}, an online system for Arabic SA. The system is based on a deep learning model, which achieves state-of-the-art results on many Arabic dialect datasets including SemEval 2017 and ASTD. The availability of such system should assist various applications and research that rely on sentiment analysis as a tool. | ['Walid Magdy', 'Ibrahim Abu Farha'] | 2019-08-01 | null | null | null | ws-2019-8 | ['twitter-sentiment-analysis', 'arabic-sentiment-analysis'] | ['natural-language-processing', 'natural-language-processing'] | [-6.20795906e-01 -4.23265249e-01 1.19949892e-01 -7.13119924e-01
-3.15738261e-01 -6.97965324e-01 6.38631046e-01 3.82612258e-01
-5.60524344e-01 4.96791989e-01 -3.95136103e-02 -4.25102443e-01
3.14248592e-01 -7.89382637e-01 -1.05101265e-01 -6.74396396e-01
-1.04942068e-01 5.79013705e-01 -2.65650060e-02 -1.45466483e+00
5.35241187e-01 4.12131876e-01 -1.21853817e+00 7.18560755e-01
8.29250216e-01 1.07044828e+00 -2.98788071e-01 4.64393556e-01
-6.11950099e-01 9.20059979e-01 -9.42163706e-01 -1.07311249e+00
6.00411510e-03 -3.45378578e-01 -1.07935119e+00 -2.90218830e-01
1.42037928e-01 -8.87033716e-02 2.91024297e-01 8.41455579e-01
4.55848873e-01 -5.04223667e-02 8.15144718e-01 -1.23171079e+00
-1.00808823e+00 9.46004629e-01 -6.71095252e-01 7.87062943e-02
3.62532437e-01 -5.48504055e-01 1.03331029e+00 -1.20964944e+00
2.75407404e-01 1.06457448e+00 7.03353107e-01 2.66558379e-01
-2.07884222e-01 -6.23878598e-01 -2.92001497e-02 4.82152671e-01
-1.06610525e+00 -2.41195723e-01 9.83559728e-01 -2.02190816e-01
1.00249064e+00 -2.66212791e-01 5.87525427e-01 8.01538229e-01
5.95299423e-01 1.24671996e+00 1.32177305e+00 -9.43950236e-01
1.07945696e-01 1.30377546e-01 7.63247788e-01 7.61046112e-01
1.65369347e-01 -9.36827958e-01 -5.91778934e-01 2.62821764e-01
1.77717675e-02 -3.51602614e-01 3.18776399e-01 1.42080218e-01
-8.71352136e-01 1.20955706e+00 3.31283182e-01 5.20780921e-01
-2.78504193e-01 -5.06115019e-01 8.56950641e-01 7.62399316e-01
7.33929038e-01 3.23930651e-01 -8.17658722e-01 -2.76629359e-01
-7.68281698e-01 4.42417324e-01 9.73890185e-01 8.42547774e-01
3.17508250e-01 3.83761883e-01 4.98711765e-01 1.11586452e+00
6.86622798e-01 8.27627540e-01 8.12932193e-01 -2.25295544e-01
3.10666114e-01 9.39597964e-01 -8.97961929e-02 -1.10651243e+00
-6.27664208e-01 1.72199085e-01 -8.27095568e-01 1.82623178e-01
5.08463264e-01 -4.92267549e-01 -8.96140873e-01 1.04591894e+00
8.90121013e-02 -7.86292017e-01 4.41111952e-01 7.81904876e-01
1.31844640e+00 1.00348449e+00 4.90519293e-02 1.35684967e-01
1.52785695e+00 -1.13858366e+00 -8.79279315e-01 -3.64664704e-01
7.69938231e-01 -1.26317012e+00 1.31026614e+00 1.07847774e+00
-9.96190012e-01 -1.75404802e-01 -1.24509549e+00 -2.89307535e-01
-1.33603334e+00 4.60462272e-01 8.18007290e-01 1.18315113e+00
-1.10820639e+00 -3.93431298e-02 -6.47715151e-01 -6.89232111e-01
2.92627066e-01 4.66906607e-01 -4.47916150e-01 1.75290219e-02
-1.46879852e+00 1.26199293e+00 1.71337605e-01 5.35206497e-01
-3.58214259e-01 -4.67727035e-02 -9.72641230e-01 -3.04157048e-01
-9.53226238e-02 1.65620163e-01 1.37324286e+00 -1.41624832e+00
-1.91993034e+00 1.29617894e+00 -1.23203084e-01 -4.34077322e-01
-1.21732187e-02 -6.37738824e-01 -9.37114418e-01 -1.01373002e-01
4.68344577e-02 1.87866300e-01 6.74992800e-01 -1.03685844e+00
-5.47547638e-01 -6.41757071e-01 2.23890886e-01 2.18641192e-01
-5.92208624e-01 9.13710773e-01 -1.47836104e-01 -7.28756070e-01
-7.80053586e-02 -9.04079258e-01 7.78309852e-02 -6.29467130e-01
-8.39722306e-02 -4.32097554e-01 7.78718829e-01 -1.01081705e+00
1.37131989e+00 -1.81863821e+00 7.06317555e-03 3.14707935e-01
-3.41712445e-01 6.29805386e-01 -9.58839506e-02 9.43035901e-01
8.59501511e-02 -1.80321140e-03 -4.19949919e-01 -2.27324426e-01
4.04412337e-02 -3.11892722e-02 -5.15252173e-01 5.22534907e-01
3.41685921e-01 5.66985548e-01 -7.21655905e-01 -2.81599939e-01
8.80180076e-02 4.30805981e-01 -1.13975108e-02 -5.60445562e-02
-1.86225384e-01 -7.16362298e-02 -1.87871054e-01 9.56028342e-01
7.47696042e-01 1.95575148e-01 2.86035210e-01 -7.94648901e-02
-2.40806296e-01 2.87402898e-01 -9.10014033e-01 1.65403974e+00
-3.54377568e-01 6.67008579e-01 -3.64147499e-02 -9.93630707e-01
1.07572365e+00 1.85209811e-01 2.32159466e-01 -6.39008462e-01
7.69865930e-01 5.55709481e-01 3.14648710e-02 -3.95392805e-01
8.58376563e-01 1.35314479e-01 -3.86186302e-01 7.84025729e-01
1.28175065e-01 -2.78401196e-01 5.17747462e-01 1.93971872e-01
3.22220743e-01 1.94562867e-01 4.41786230e-01 -6.28543258e-01
1.21870816e+00 2.51474202e-01 -4.40434664e-02 2.22318515e-01
-3.86786789e-01 4.54771101e-01 4.49341476e-01 -8.42395544e-01
-9.25732672e-01 -6.59466326e-01 -2.14082226e-01 1.64243472e+00
-1.44093826e-01 -6.25227213e-01 -9.13643479e-01 -7.31843531e-01
-3.09677005e-01 5.00956714e-01 -8.16819727e-01 2.34342739e-01
-6.86078310e-01 -1.20944679e+00 6.71973050e-01 3.72167408e-01
8.42342675e-01 -1.55782568e+00 -3.66721541e-01 2.55136698e-01
-2.42116004e-01 -7.97354996e-01 1.11836269e-01 4.31744680e-02
-4.15135890e-01 -9.11409140e-01 -6.28379405e-01 -1.35832798e+00
3.93922418e-01 1.28946915e-01 1.26232779e+00 6.77048415e-02
3.06866378e-01 -1.09985165e-01 -1.13395369e+00 -1.18457806e+00
-3.97044629e-01 5.48125863e-01 8.74788091e-02 -2.99002770e-02
1.23134279e+00 2.39481255e-01 -1.73162490e-01 -1.10810265e-01
-8.17970872e-01 -3.60036343e-01 3.37073773e-01 7.11085498e-01
2.16609333e-02 -1.25389531e-01 1.04487228e+00 -1.17410135e+00
1.10115290e+00 -5.57463527e-01 -2.86688149e-01 2.65171736e-01
-4.46008325e-01 -3.59104604e-01 8.51534247e-01 2.33468056e-01
-1.11916506e+00 -3.15902889e-01 -6.60775721e-01 6.48673952e-01
-1.61737934e-01 1.19806468e+00 -1.44960344e-01 1.58100456e-01
7.11073816e-01 1.57170385e-01 -6.10265918e-02 -2.21828207e-01
3.20454985e-01 1.34013867e+00 -4.30319868e-02 -4.56401438e-01
4.27064039e-02 3.66830379e-01 -2.25303650e-01 -9.82724130e-01
-1.06636548e+00 -2.94937313e-01 -8.64775836e-01 -3.79633754e-01
6.61977351e-01 -9.05969203e-01 -4.92149472e-01 1.26921761e+00
-7.92693853e-01 -1.12524606e-01 3.41840565e-01 8.57197419e-02
-1.96843311e-01 1.99441940e-01 -9.07792926e-01 -7.60179520e-01
-9.09152746e-01 -1.03451121e+00 8.31531465e-01 3.06858748e-01
-2.47333735e-01 -9.93812919e-01 4.29143339e-01 4.21271861e-01
4.55566466e-01 -3.96837071e-02 7.88603961e-01 -9.22603548e-01
4.31066632e-01 -4.01402414e-01 4.55377139e-02 5.74987948e-01
2.85481721e-01 4.76475090e-01 -1.19976962e+00 -2.03980312e-01
-1.44906804e-01 -7.28673935e-01 5.07717013e-01 4.35428768e-01
7.26100147e-01 3.47563475e-02 5.10890424e-01 1.83862615e-02
1.21182525e+00 4.50256169e-01 7.10023701e-01 9.15104032e-01
5.34869611e-01 7.53613830e-01 9.53185618e-01 3.56549710e-01
8.99748087e-01 1.72521872e-03 8.62289891e-02 5.83011732e-02
4.17142242e-01 4.16360736e-01 9.21331525e-01 1.40198362e+00
-1.72551572e-01 -3.43024254e-01 -1.45763671e+00 6.79387927e-01
-1.79551101e+00 -6.61155641e-01 -4.25547391e-01 1.59273314e+00
8.43688130e-01 -6.63943589e-02 3.34127516e-01 6.02468610e-01
1.86295792e-01 3.06770444e-01 1.48456544e-01 -1.47781622e+00
-5.11684835e-01 7.23590672e-01 -1.21908672e-01 3.60678405e-01
-1.56697726e+00 1.59420204e+00 6.38807726e+00 4.91018474e-01
-1.29271257e+00 -8.88598263e-02 6.80975378e-01 3.83221447e-01
1.28479898e-01 -5.47651529e-01 -5.45498133e-01 2.21869513e-01
1.22813010e+00 4.41013426e-02 7.28737116e-02 8.15929651e-01
8.08015838e-02 -2.08333433e-01 -5.14763772e-01 6.53384089e-01
7.67526627e-01 -1.07938325e+00 2.28972450e-01 -5.16107857e-01
7.35358298e-01 1.41229853e-01 3.33135188e-01 5.01486182e-01
4.05152708e-01 -1.10351634e+00 6.81181490e-01 3.12667266e-02
4.05607402e-01 -1.43957055e+00 1.53774619e+00 -1.52006119e-01
-8.67261231e-01 3.11179869e-02 -3.72701436e-01 -2.57822573e-01
-1.64233744e-01 4.00103301e-01 -5.72683573e-01 5.91639340e-01
1.22014105e+00 9.65960503e-01 -7.97333658e-01 4.12651449e-01
-5.14490366e-01 1.05890405e+00 -1.47923127e-01 -6.09913111e-01
5.48110068e-01 -7.83986807e-01 -4.58061583e-02 1.64149344e+00
7.86393806e-02 -4.52073626e-02 -7.85853639e-02 -1.72380865e-01
-1.84367467e-02 1.06149840e+00 -6.77460611e-01 -2.62707949e-01
7.53390640e-02 1.29051995e+00 -9.86551940e-01 -2.33493775e-01
-7.98339069e-01 8.37001622e-01 2.64113635e-01 -2.62653343e-02
-4.60359246e-01 -8.34128797e-01 4.47054505e-01 -4.62814778e-01
8.76245797e-02 -5.19293070e-01 -7.07585931e-01 -1.25706756e+00
-1.42246559e-01 -1.40094721e+00 6.28961921e-01 -7.98168302e-01
-1.53074145e+00 9.78181481e-01 -4.31505710e-01 -1.12179160e+00
-8.24730545e-02 -1.26356399e+00 -4.71301496e-01 7.41162419e-01
-1.50178707e+00 -1.92856431e+00 6.64073303e-02 6.85769141e-01
7.42945969e-01 -9.51500058e-01 1.21300709e+00 3.47170323e-01
-5.56321204e-01 6.77649736e-01 2.92947918e-01 4.86281574e-01
1.16230726e+00 -1.42955554e+00 3.26217532e-01 9.78426993e-01
6.20310083e-02 6.74619913e-01 4.76744920e-01 -4.21235204e-01
-1.45610428e+00 -6.04319513e-01 1.43988800e+00 -8.79258752e-01
9.71740544e-01 -3.48816544e-01 -5.80645978e-01 8.22672427e-01
1.05669427e+00 -7.16609597e-01 1.15552187e+00 3.60748410e-01
-2.65432507e-01 -3.30975115e-01 -1.28072381e+00 8.21128309e-01
-8.22462142e-03 -2.85120785e-01 -7.99688220e-01 9.64442268e-02
2.02681229e-01 -3.15124452e-01 -7.58081079e-01 6.95561543e-02
6.89921081e-01 -8.17784488e-01 5.45172393e-01 -7.70624518e-01
8.86077523e-01 -2.33051792e-01 -2.41490439e-01 -1.57490766e+00
6.84206486e-02 -2.32036397e-01 9.32580233e-02 1.14231503e+00
5.95414579e-01 -3.88790488e-01 5.63645303e-01 1.54742271e-01
-2.29718462e-01 -5.34210980e-01 -3.81269723e-01 -2.54454222e-02
7.27157474e-01 -6.01077795e-01 7.40903854e-01 1.36982572e+00
3.41875196e-01 6.01966083e-01 -2.74052411e-01 -6.58026524e-03
-1.23414891e-02 7.62996450e-02 6.00940645e-01 -9.70456302e-01
6.71498239e-01 -5.61718941e-01 -2.67615139e-01 -4.00338322e-01
2.60338992e-01 -5.91753781e-01 -1.54203162e-01 -1.59082162e+00
-4.98346806e-01 -1.83390543e-01 -2.20830411e-01 6.06701672e-01
7.16997311e-02 4.92283702e-01 3.78031246e-02 -1.60505041e-01
-4.12286311e-01 3.96062672e-01 8.27812314e-01 -2.31593326e-01
-1.86723262e-01 -2.24672988e-01 -9.63182688e-01 9.14568067e-01
1.27006781e+00 -1.13004912e-02 -2.65480846e-01 -7.79269397e-01
8.37444663e-01 -7.21155167e-01 -6.90904856e-01 -6.73272729e-01
9.89324227e-02 -2.13787362e-01 3.38617444e-01 -7.63668180e-01
1.02217384e-01 -7.01075315e-01 -7.38026261e-01 1.02078237e-01
-1.19856000e-01 8.64892960e-01 4.78021771e-01 -2.84118146e-01
-6.10950649e-01 -3.89072090e-01 8.35150480e-01 -7.21640661e-02
-1.00304675e+00 6.77694306e-02 -1.08471501e+00 -1.03484802e-01
8.73857498e-01 1.89859018e-01 -4.00168806e-01 -2.67463475e-01
-3.38741481e-01 2.08362728e-01 7.46330395e-02 6.83379948e-01
5.92775404e-01 -1.09313428e+00 -1.04664624e+00 3.90042245e-01
3.41986507e-01 -1.12726703e-01 -9.33108106e-02 4.31906968e-01
-1.51995802e+00 3.79003197e-01 -6.58414662e-01 -5.20906784e-02
-1.14880323e+00 9.19987336e-02 2.32847985e-02 -1.19573563e-01
-1.22096963e-01 9.19001341e-01 -6.36225522e-01 -9.23878849e-01
-3.41079421e-02 -2.84359455e-02 -1.19532597e+00 5.15089452e-01
7.90860891e-01 2.86797583e-01 5.60438395e-01 -1.12348151e+00
-5.82501292e-01 3.84183079e-01 -5.62061667e-01 -3.07017267e-01
1.54681444e+00 -1.81487709e-01 -8.60007703e-01 7.18465388e-01
6.94599867e-01 3.40747058e-01 -8.41602013e-02 5.76182231e-02
1.50011748e-01 -2.24033296e-01 -3.47812980e-04 -1.19883990e+00
-1.02371299e+00 1.09802186e+00 6.19489074e-01 4.36441690e-01
1.18432045e+00 -6.68154895e-01 8.60887527e-01 8.82308066e-01
3.62828940e-01 -1.58712792e+00 -1.37344405e-01 1.41719460e+00
9.01983142e-01 -1.56667709e+00 -9.54080299e-02 8.53661746e-02
-1.15609550e+00 1.48044479e+00 6.61829114e-01 -3.36308539e-01
1.25939584e+00 1.25255659e-01 1.09865344e+00 -3.00809056e-01
-1.66439563e-01 -1.88061539e-02 9.79660600e-02 7.15287149e-01
1.16672397e+00 1.42844334e-01 -7.55606830e-01 1.18470299e+00
-8.20197403e-01 -7.51035586e-02 8.89278233e-01 1.34013200e+00
-4.15409535e-01 -1.34237242e+00 -2.43802682e-01 4.14790422e-01
-8.67474914e-01 -3.77931386e-01 -8.06679726e-01 9.12913740e-01
-1.33283719e-01 1.24128032e+00 -5.42321764e-02 -3.06598216e-01
2.24585876e-01 1.08929306e-01 3.52044463e-01 -4.60578442e-01
-1.10064471e+00 -2.32517630e-01 2.89278388e-01 -3.36806700e-02
-7.70469427e-01 -7.03169167e-01 -1.34035182e+00 -8.57587695e-01
-1.23955935e-01 1.51540250e-01 8.75012219e-01 1.13438916e+00
2.28531938e-02 -2.09733248e-02 6.36647284e-01 -5.02890587e-01
-1.51053572e-03 -1.17906749e+00 -8.38407397e-01 2.86571026e-01
2.44901374e-01 -2.00463995e-01 6.32979646e-02 1.87386811e-01] | [11.12539291381836, 7.044964790344238] |
bc4ff6cb-7e2f-4bf6-8f3c-3505e0c9dd54 | hierarchically-clustered-pca-and-cca-via-a | 2211.16553 | null | https://arxiv.org/abs/2211.16553v3 | https://arxiv.org/pdf/2211.16553v3.pdf | Simple and Scalable Algorithms for Cluster-Aware Precision Medicine | AI-enabled precision medicine promises a transformational improvement in healthcare outcomes by enabling data-driven personalized diagnosis, prognosis, and treatment. However, the well-known "curse of dimensionality" and the clustered structure of biomedical data together interact to present a joint challenge in the high dimensional, limited observation precision medicine regime. To overcome both issues simultaneously we propose a simple and scalable approach to joint clustering and embedding that combines standard embedding methods with a convex clustering penalty in a modular way. This novel, cluster-aware embedding approach overcomes the complexity and limitations of current joint embedding and clustering methods, which we show with straightforward implementations of hierarchically clustered principal component analysis (PCA), locally linear embedding (LLE), and canonical correlation analysis (CCA). Through both numerical experiments and real-world examples, we demonstrate that our approach outperforms traditional and contemporary clustering methods on highly underdetermined problems (e.g., with just tens of observations) as well as on large sample datasets. Importantly, our approach does not require the user to choose the desired number of clusters, but instead yields interpretable dendrograms of hierarchically clustered embeddings. Thus our approach improves significantly on existing methods for identifying patient subgroups in multiomics and neuroimaging data, enabling scalable and interpretable biomarkers for precision medicine. | ['Logan Grosenick', 'Conor Liston', 'Amanda M. Buch'] | 2022-11-29 | null | null | null | null | ['distributed-optimization'] | ['methodology'] | [ 2.66527645e-02 -1.41621470e-01 -1.77132979e-01 -1.46999270e-01
-6.40732586e-01 -6.05557621e-01 5.06483555e-01 5.45571685e-01
-2.28072524e-01 3.51300448e-01 5.80176830e-01 -1.14765748e-01
-7.39910543e-01 -2.59744644e-01 -3.12107969e-02 -1.00319922e+00
-4.10450459e-01 7.93941498e-01 -4.27986383e-01 2.31742531e-01
-1.40407845e-01 5.20369053e-01 -8.80567193e-01 -1.12790689e-01
8.38371217e-01 5.82576394e-01 -2.17112750e-01 5.63481331e-01
4.91892993e-01 2.04304501e-01 -5.23895677e-03 -1.62338018e-01
7.11672530e-02 -2.61374950e-01 -3.91090661e-01 6.82042837e-02
-5.60791865e-02 -5.04612103e-02 -4.43962336e-01 7.88477778e-01
7.57394433e-01 -5.59707880e-02 8.73766959e-01 -1.31586123e+00
-8.96894753e-01 4.69558239e-01 -5.71488261e-01 7.96388462e-02
1.03051905e-02 1.37032479e-01 9.91936564e-01 -9.29308593e-01
5.72803617e-01 1.22220135e+00 8.65315557e-01 3.71947527e-01
-1.91732788e+00 -2.59900153e-01 -2.17563182e-01 6.23745322e-02
-1.50491738e+00 -4.70250487e-01 7.76223540e-01 -1.00518501e+00
5.91289818e-01 2.90946811e-01 7.21379399e-01 8.50202560e-01
1.74590856e-01 3.76794517e-01 9.11233068e-01 -6.04935437e-02
5.77132821e-01 -1.05115054e-02 2.19966531e-01 6.50095105e-01
5.63671291e-01 4.85566556e-02 -1.44127488e-01 -7.89646745e-01
5.04294991e-01 5.19450843e-01 -3.80842209e-01 -7.75266409e-01
-1.85811925e+00 9.42407846e-01 2.24539116e-01 5.11370659e-01
-5.60252666e-01 2.96561182e-01 3.40553105e-01 -8.31766501e-02
4.25689101e-01 6.14412189e-01 -4.47665304e-01 3.44341919e-02
-9.57993090e-01 2.69531049e-02 6.26139879e-01 4.69276547e-01
4.36240971e-01 -1.40187338e-01 6.96441084e-02 4.99916136e-01
2.53196180e-01 4.45584953e-01 4.98418450e-01 -1.10716093e+00
-2.70281546e-02 8.08279872e-01 4.74241450e-02 -1.55328417e+00
-8.09995413e-01 -3.97534907e-01 -1.33882868e+00 -2.13234156e-01
3.27437043e-01 -3.49444300e-02 -4.82490391e-01 1.71542573e+00
5.16472161e-01 3.26162636e-01 -4.89320680e-02 8.44303250e-01
3.55982184e-01 1.65186509e-01 1.13469496e-01 -3.10087770e-01
1.58753181e+00 -4.66891944e-01 -8.34591031e-01 2.46908754e-01
7.08872437e-01 -2.62463212e-01 7.47992635e-01 2.18304411e-01
-7.91359305e-01 -2.05608923e-02 -8.55830312e-01 -1.88980341e-01
-3.35976958e-01 2.33242631e-01 7.61884570e-01 5.01768351e-01
-1.01426244e+00 6.41817272e-01 -1.16369641e+00 -3.29868376e-01
5.68362117e-01 4.91906226e-01 -6.39497578e-01 -1.84831202e-01
-7.08634615e-01 5.73263347e-01 -5.54291252e-03 6.83518648e-02
-4.87926424e-01 -1.24299574e+00 -6.59005105e-01 1.95644721e-01
5.39641716e-02 -1.05041361e+00 3.10336560e-01 -9.33493152e-02
-1.13980913e+00 5.87889314e-01 -2.53177822e-01 -3.52546632e-01
3.16488266e-01 -1.37642562e-01 -1.66566774e-01 5.06793618e-01
-2.73757800e-02 4.87371594e-01 6.07393324e-01 -8.91788304e-01
2.33139634e-01 -7.23388255e-01 -6.96539640e-01 7.05462545e-02
-6.82100534e-01 -1.28949136e-01 -7.60969594e-02 -4.40846294e-01
4.30868894e-01 -9.99919415e-01 -7.91069508e-01 2.97151685e-01
-4.41722184e-01 3.38234715e-02 6.70156240e-01 -6.62827432e-01
1.12598276e+00 -2.32918072e+00 8.29750776e-01 2.50083596e-01
1.04744565e+00 -1.17181480e-01 7.41264150e-02 5.40631413e-01
-2.84276247e-01 2.84298718e-01 -4.46157038e-01 -3.69128585e-01
1.06466338e-01 1.02026440e-01 8.24405253e-02 9.07393098e-01
1.72129452e-01 1.17187393e+00 -1.16430962e+00 -4.97781366e-01
3.60380590e-01 8.71610701e-01 -7.37715483e-01 -8.66206288e-02
2.81653255e-01 6.16810262e-01 -4.11162853e-01 6.12935603e-01
3.22474420e-01 -8.04910779e-01 6.24965787e-01 -6.34296656e-01
1.89200178e-01 -3.60963285e-01 -1.08385253e+00 1.58704126e+00
1.00164950e-01 3.96059662e-01 2.76710130e-02 -1.27474105e+00
5.73473096e-01 5.09957790e-01 1.25668490e+00 2.32380722e-03
1.81550264e-01 2.93451287e-02 1.30790740e-01 -4.18187052e-01
-2.81607173e-02 -2.58318126e-01 -8.99708495e-02 5.25301576e-01
-8.05722848e-02 2.39343286e-01 -1.93653941e-01 3.38418841e-01
1.53173447e+00 -4.67770994e-01 5.01109242e-01 -5.86752713e-01
1.42077252e-01 -2.05853060e-02 6.22921944e-01 2.00637147e-01
-5.45416415e-01 5.69534779e-01 6.50050163e-01 -3.63653719e-01
-1.30932832e+00 -1.24748707e+00 -4.66509432e-01 4.59866494e-01
-2.82177746e-01 -6.06941938e-01 -5.01117051e-01 -3.23701203e-01
4.55287158e-01 4.51097637e-02 -7.58152902e-01 -2.22924590e-01
-2.09787175e-01 -1.44342232e+00 3.91164273e-01 4.61813718e-01
-2.37786770e-01 -2.44787917e-01 -3.68037134e-01 2.44617596e-01
-1.96698587e-02 -1.21200681e+00 -3.39802206e-01 3.13835517e-02
-1.08846319e+00 -1.20121801e+00 -6.26013994e-01 -4.10178900e-01
8.42261195e-01 1.50861919e-01 7.58027554e-01 -2.49650270e-01
-8.17333758e-01 6.19934559e-01 1.35381566e-03 1.22655332e-01
-9.23848897e-02 -2.73921430e-01 7.28095055e-01 1.06660411e-01
4.57939267e-01 -1.01845932e+00 -9.70330358e-01 1.14212237e-01
-7.63315082e-01 -1.98559701e-01 5.76809287e-01 9.97767031e-01
5.67925751e-01 3.66910174e-03 7.51954794e-01 -5.98885059e-01
7.83347249e-01 -7.49594331e-01 -4.16197896e-01 2.35114068e-01
-9.01698887e-01 7.63238221e-02 5.80427885e-01 -5.16535521e-01
-2.07836479e-01 1.85558140e-01 4.74413246e-01 -7.24981904e-01
-7.04970676e-04 6.49988115e-01 4.35168147e-02 7.06764087e-02
6.07259452e-01 9.25031304e-02 3.26291233e-01 -6.31748199e-01
7.91435838e-01 6.23772383e-01 5.37685931e-01 -3.53284717e-01
6.39467239e-01 8.18991423e-01 4.15034413e-01 -8.46383452e-01
-1.90716356e-01 -4.88250256e-01 -1.08204699e+00 2.04130113e-01
1.12723720e+00 -1.01760340e+00 -9.65331554e-01 -1.59677297e-01
-6.69038773e-01 8.98613259e-02 -3.12456042e-01 8.24556053e-01
-4.63990360e-01 6.69610441e-01 -6.87292278e-01 -6.37259007e-01
-4.04141963e-01 -1.03202283e+00 1.12460518e+00 -3.40565801e-01
-5.57448030e-01 -1.29021990e+00 3.88698071e-01 2.86132067e-01
3.43350410e-01 8.41036201e-01 1.27512944e+00 -5.70372403e-01
-2.64977962e-01 -3.68291914e-01 -2.38139927e-01 6.99773133e-02
3.51911664e-01 -2.31511742e-02 -6.17723465e-01 -2.67038882e-01
1.50280250e-02 -1.15099717e-02 5.48833013e-01 4.37117219e-01
9.78862524e-01 -3.85823607e-01 -6.59163833e-01 6.28698289e-01
1.34092593e+00 -6.49307892e-02 2.35554636e-01 -2.88697422e-01
9.10932958e-01 4.78818595e-01 -4.41582277e-02 6.05735362e-01
5.03324449e-01 5.36492884e-01 -9.76330191e-02 -6.49352670e-02
9.91918892e-02 5.77842221e-02 -9.22556967e-02 1.30929279e+00
8.66902918e-02 2.07949340e-01 -1.04181540e+00 6.85472369e-01
-1.95957923e+00 -9.07518029e-01 -1.97182968e-01 2.08899689e+00
8.02168787e-01 -5.74334919e-01 1.65869147e-01 2.19775647e-01
5.11190593e-01 -2.07622796e-01 -6.70776188e-01 -5.41439466e-02
-1.42549694e-01 8.84168781e-03 2.86577672e-01 3.88188034e-01
-1.08642101e+00 4.48332727e-01 7.38634872e+00 3.04937720e-01
-8.16447437e-01 2.32987523e-01 6.02194130e-01 -3.07301879e-01
-3.89856309e-01 -1.59895226e-01 -1.39044702e-01 5.72235346e-01
1.12159038e+00 -3.99446726e-01 4.99536753e-01 5.56739807e-01
5.12663007e-01 3.09710145e-01 -1.40940154e+00 1.23440444e+00
-4.58085462e-02 -1.34250343e+00 -2.18465894e-01 3.91536564e-01
6.34063840e-01 1.20938733e-01 2.19591975e-01 -4.33256984e-01
3.08511347e-01 -1.11770368e+00 -1.90355062e-01 5.77790022e-01
7.50004232e-01 -4.21187818e-01 3.36405873e-01 1.81921929e-01
-9.03071225e-01 -2.38999411e-01 -4.44343984e-01 1.12277836e-01
1.59055561e-01 1.25208759e+00 -6.71456277e-01 5.01268387e-01
5.48978686e-01 1.07619488e+00 -4.73383248e-01 8.51275265e-01
3.55845422e-01 5.18813610e-01 -5.07865429e-01 2.81033218e-01
-1.49229392e-01 -4.83125240e-01 5.71285248e-01 1.01739120e+00
2.38019496e-01 3.38413417e-01 5.41481748e-02 1.00338697e+00
2.01194525e-01 3.01763993e-02 -5.67695141e-01 -6.06063247e-01
6.40195191e-01 1.38716662e+00 -6.78635001e-01 -4.45182264e-01
-1.96866661e-01 8.35848689e-01 2.28090763e-01 2.86635220e-01
-4.47513282e-01 -1.05971573e-02 8.69003057e-01 -1.24361776e-02
1.39279142e-01 -6.49449706e-01 -5.58190882e-01 -1.48271489e+00
-1.19020283e-01 -7.73877859e-01 4.45188165e-01 -2.46062130e-01
-1.48439002e+00 1.99283272e-01 -4.82423007e-02 -1.09948063e+00
-1.60241991e-01 -5.31037271e-01 -3.47284049e-01 5.33553004e-01
-1.08793414e+00 -8.50290298e-01 -2.81631611e-02 4.08194363e-01
-2.68570334e-01 -1.29252784e-02 1.17184401e+00 5.89369655e-01
-8.85888457e-01 3.37770343e-01 7.45281458e-01 -1.71694160e-02
5.23396254e-01 -1.36721706e+00 -7.61580169e-02 4.66144025e-01
-9.07092318e-02 8.18175077e-01 6.98891640e-01 -5.44106364e-01
-1.89543676e+00 -1.05179369e+00 7.52928734e-01 -7.76090503e-01
9.91484582e-01 -5.06874561e-01 -8.08086395e-01 6.28632843e-01
-2.00320140e-01 2.46393085e-01 1.37774944e+00 2.51080126e-01
-3.88197541e-01 -3.04963812e-02 -1.23509574e+00 5.67669034e-01
8.55439901e-01 -4.49082732e-01 -4.91274148e-01 7.64612317e-01
6.86585903e-01 2.12996155e-01 -1.71028638e+00 3.93010318e-01
6.35113657e-01 -4.34441119e-01 1.22572541e+00 -8.37584615e-01
2.28813335e-01 -4.40046817e-01 -3.31246316e-01 -1.17582905e+00
-7.54850984e-01 -6.88126802e-01 -5.02989292e-01 8.70328426e-01
1.74669534e-01 -6.39635265e-01 6.89291954e-01 9.27486062e-01
1.96052834e-01 -1.01866972e+00 -1.01247299e+00 -6.17203355e-01
2.31013775e-01 1.22529417e-02 4.77136433e-01 1.39696014e+00
5.68463326e-01 2.56861448e-01 -3.18439484e-01 2.98658192e-01
1.13990855e+00 7.64996260e-02 5.38821459e-01 -1.42805839e+00
-3.48245651e-01 -4.52338517e-01 -9.67642844e-01 -3.45069885e-01
-2.13043436e-01 -1.05486512e+00 -4.41488206e-01 -1.29464769e+00
4.19615895e-01 -5.03065169e-01 -3.57392490e-01 3.68702948e-01
-1.97521687e-01 3.71087015e-01 9.75358672e-03 4.40692246e-01
-5.19707978e-01 6.42440438e-01 6.86812043e-01 -8.94744918e-02
-2.82383919e-01 -6.69689059e-01 -8.70653570e-01 4.51340735e-01
5.95395327e-01 -4.38836157e-01 -3.26310366e-01 -2.03852206e-01
1.39786527e-01 2.76643131e-02 5.39907813e-01 -8.84077549e-01
2.62817383e-01 -3.32873501e-03 5.28973222e-01 -2.05117136e-01
4.00184125e-01 -8.46099555e-01 3.48470211e-01 7.05464661e-01
-2.11599439e-01 1.43486217e-01 -1.47007897e-01 8.55891585e-01
1.65402770e-01 4.31386322e-01 5.69076598e-01 1.80186868e-01
1.29670007e-02 4.64400768e-01 -1.58679202e-01 -1.06609315e-01
1.15722919e+00 5.73334470e-02 -4.43306297e-01 -2.16874182e-01
-1.12477469e+00 5.22259831e-01 3.73133838e-01 1.72074407e-01
6.44167066e-01 -1.56183791e+00 -7.74517298e-01 3.91072929e-02
6.10364266e-02 -1.54962778e-01 3.38628352e-01 1.56978714e+00
-2.09861785e-01 4.26026046e-01 5.69006503e-02 -8.69133711e-01
-1.02843738e+00 1.01497018e+00 1.11981727e-01 -1.27497941e-01
-8.49022269e-01 2.23201990e-01 8.75194669e-02 -4.36392933e-01
-7.63053074e-02 -3.37922335e-01 2.92495340e-02 2.14939192e-01
6.40698850e-01 6.55264616e-01 -3.15289080e-01 -5.58323562e-01
-4.99021709e-01 6.61434531e-01 1.69072390e-01 8.39799941e-02
1.68702769e+00 -2.01896191e-01 -4.61185515e-01 6.43813193e-01
1.53558707e+00 -2.39945889e-01 -8.96005034e-01 -1.03590608e-01
7.71559924e-02 -2.26255640e-01 8.19675066e-03 -5.92877388e-01
-8.69033396e-01 9.53623414e-01 5.32848537e-01 2.08477974e-01
8.60181093e-01 1.98352978e-01 6.74873352e-01 4.19221580e-01
-1.00037210e-01 -7.21880674e-01 7.25174695e-03 -3.04153293e-01
6.58029914e-01 -1.09740138e+00 3.29386055e-01 -2.59411782e-01
-4.79984760e-01 9.00241613e-01 -1.06611727e-02 -1.72658369e-01
1.10759497e+00 1.36079058e-01 -1.08678281e-01 -6.15389526e-01
-8.86190355e-01 1.83263168e-01 3.96161377e-01 6.26090825e-01
2.89238751e-01 4.05813158e-01 -3.73834014e-01 6.33787990e-01
-5.40352017e-02 -5.80000170e-02 2.89748967e-01 6.79695606e-01
-2.53675640e-01 -9.26178813e-01 -2.97028095e-01 6.49321437e-01
-2.85294056e-01 -2.09187493e-02 -3.99355829e-01 5.13418019e-01
-6.49675131e-02 7.06781507e-01 4.27224152e-02 -4.64839458e-01
-1.70573264e-01 1.98581129e-01 2.29260117e-01 -3.42277199e-01
-4.90783975e-02 3.01317751e-01 -3.46615583e-01 -7.61956930e-01
-4.90050673e-01 -9.82740879e-01 -1.11966240e+00 -2.77095467e-01
-1.55883953e-01 4.88064550e-02 6.84461236e-01 8.52370381e-01
9.96159911e-01 4.02062953e-01 6.40758395e-01 -8.37735713e-01
-5.97619355e-01 -6.52244806e-01 -5.97040892e-01 4.92258847e-01
5.43334484e-01 -5.80732644e-01 -5.75990319e-01 -2.04696935e-02] | [7.0783610343933105, 5.206086158752441] |
44b48036-90dd-4a21-bfda-333f75e89565 | simple-and-efficient-learning-using | 1604.01518 | null | http://arxiv.org/abs/1604.01518v1 | http://arxiv.org/pdf/1604.01518v1.pdf | Simple and Efficient Learning using Privileged Information | The Support Vector Machine using Privileged Information (SVM+) has been
proposed to train a classifier to utilize the additional privileged information
that is only available in the training phase but not available in the test
phase. In this work, we propose an efficient solution for SVM+ by simply
utilizing the squared hinge loss instead of the hinge loss as in the existing
SVM+ formulation, which interestingly leads to a dual form with less variables
and in the same form with the dual of the standard SVM. The proposed algorithm
is utilized to leverage the additional web knowledge that is only available
during training for the image categorization tasks. The extensive experimental
results on both Caltech101 andWebQueries datasets show that our proposed method
can achieve a factor of up to hundred times speedup with the comparable
accuracy when compared with the existing SVM+ method. | ['Yong liu', 'Xinxing Xu', 'Joey Tianyi Zhou', 'IvorW. Tsang', 'Zheng Qin', 'Rick Siow Mong Goh'] | 2016-04-06 | null | null | null | null | ['image-categorization'] | ['computer-vision'] | [ 2.98833370e-01 -2.85176355e-02 -5.23681283e-01 -5.25123119e-01
-4.71463233e-01 -5.94920814e-01 4.34479415e-01 1.29808500e-01
-4.57092136e-01 1.20266950e+00 -6.20948672e-01 -5.84007621e-01
-2.28594616e-01 -7.21550822e-01 -5.47709465e-01 -9.05417383e-01
1.57423884e-01 1.18953004e-01 3.84518772e-01 -2.54771203e-01
4.12375480e-01 2.03114912e-01 -1.69628465e+00 3.40741426e-01
1.31940341e+00 1.69739807e+00 8.84983167e-02 3.07955444e-01
-8.88600573e-02 6.57659829e-01 -3.37155879e-01 -3.54556769e-01
7.88179517e-01 -6.59813061e-02 -7.32647896e-01 1.71800882e-01
6.70573533e-01 -4.04106796e-01 -4.87376243e-01 1.10974216e+00
-5.90001158e-02 1.53411075e-01 4.42390084e-01 -1.60967743e+00
-3.89273167e-01 1.18703194e-01 -6.85160875e-01 3.11050653e-01
2.37831324e-01 -3.43042970e-01 9.18284357e-01 -5.37382662e-01
3.31783801e-01 7.37552941e-01 4.11885977e-01 2.38596424e-01
-9.44179654e-01 -5.94752610e-01 1.19232595e-01 6.69149935e-01
-1.16351867e+00 -6.80074617e-02 7.77877092e-01 -3.92425060e-01
7.33063519e-01 4.44703549e-01 5.79368889e-01 8.65056872e-01
1.80240661e-01 1.01965737e+00 1.79085076e+00 -5.18830359e-01
2.09576562e-01 1.05805385e+00 9.92053986e-01 1.03513443e+00
2.62928814e-01 5.44549562e-02 -3.76938134e-01 -6.26154721e-01
3.17617059e-01 1.63103417e-01 -3.49729955e-01 -7.95565188e-01
-7.85353124e-01 1.22320497e+00 4.63974327e-01 -3.13537627e-01
-1.02795184e-01 -5.76460481e-01 5.22532701e-01 6.57691300e-01
2.96258777e-01 2.13555738e-01 -6.50701225e-01 1.69973701e-01
-9.16640580e-01 3.60703911e-03 9.55360115e-01 9.67687905e-01
8.70607674e-01 4.86298241e-02 3.55475813e-01 8.42541575e-01
8.12885314e-02 4.30025429e-01 7.78881133e-01 -5.46705961e-01
5.81570983e-01 9.36558783e-01 1.34293228e-01 -6.69371068e-01
1.13826796e-01 -4.78725255e-01 -6.33590579e-01 5.86477578e-01
2.07768649e-01 5.39577454e-02 -8.36838961e-01 1.38158071e+00
5.35138547e-01 1.52876392e-01 3.46752346e-01 7.58386850e-01
5.23460567e-01 4.80696231e-01 -2.66010135e-01 -1.40009224e-01
1.23977220e+00 -1.29348004e+00 -4.51591581e-01 -1.90489113e-01
4.21708256e-01 -6.39210999e-01 1.14567077e+00 7.74343848e-01
-4.18859333e-01 -3.53156567e-01 -1.77661955e+00 1.96854725e-01
-7.75633752e-01 -4.56540436e-02 1.19283485e+00 9.59328175e-01
-5.06058395e-01 6.42835557e-01 -5.69847822e-01 -3.15943837e-01
4.70628679e-01 4.23099786e-01 -8.23443353e-01 -3.35265808e-02
-1.25607395e+00 9.81613278e-01 6.16588533e-01 -3.97654325e-02
-4.71847266e-01 -4.45102006e-01 -9.02016163e-01 3.61967161e-02
3.91649753e-01 -2.45607868e-01 8.64317000e-01 -1.11857772e+00
-1.63086104e+00 6.64575279e-01 1.94373429e-02 -5.31160414e-01
7.30394125e-01 -2.77824104e-01 -1.42636403e-01 1.88325047e-01
-1.76491916e-01 1.23907261e-01 7.30928004e-01 -1.00791705e+00
-8.20588946e-01 -6.55514121e-01 4.60821152e-01 2.61322856e-01
-6.74697697e-01 -2.34135509e-01 -1.22973397e-01 -4.48156834e-01
2.10471854e-01 -1.05332923e+00 2.59129852e-02 1.45158425e-01
-3.37835163e-01 -6.16958849e-02 1.15465391e+00 -7.89142251e-01
1.11936748e+00 -1.85860205e+00 3.18440124e-02 4.22463775e-01
-2.01825485e-01 4.19340283e-01 2.93154746e-01 2.96461791e-01
-2.15273723e-01 -3.65726411e-01 -3.41259986e-01 7.56293535e-02
-1.52966872e-01 2.03456834e-01 -5.24302900e-01 5.77565432e-01
-1.15971446e-01 3.60973299e-01 -5.89706421e-01 -6.44697487e-01
3.25696588e-01 1.08059179e-02 -2.72671431e-01 2.95855612e-01
1.61539406e-01 1.00214668e-01 -5.05976260e-01 7.88686514e-01
1.06307948e+00 -4.29755598e-01 1.78845167e-01 1.06183566e-01
2.03491539e-01 -1.15722060e-01 -1.32176948e+00 1.34042037e+00
-4.88766193e-01 3.70958745e-01 1.28089666e-01 -1.43269908e+00
7.68362701e-01 1.16823800e-01 2.05212995e-01 -3.75055671e-01
1.17678389e-01 1.94271877e-01 -2.49192864e-01 -3.56584370e-01
6.30133823e-02 -3.75347435e-02 8.58311504e-02 1.27055243e-01
8.73878524e-02 4.80427891e-01 -9.11051854e-02 7.09839240e-02
8.55130434e-01 1.11567728e-01 6.56566441e-01 -3.44392389e-01
1.04687166e+00 3.36207449e-01 5.43372750e-01 7.11861849e-01
-4.54331368e-01 4.35212115e-03 3.90981436e-01 -3.64433289e-01
-6.85999513e-01 -9.75531816e-01 -7.02175319e-01 8.53587627e-01
4.43625152e-01 -4.26538587e-02 -6.00682497e-01 -1.32096207e+00
4.24014568e-01 5.60504615e-01 -5.88019013e-01 -1.73473105e-01
-3.98770213e-01 -7.62267172e-01 2.05287054e-01 4.45561111e-01
9.73079383e-01 -5.57627618e-01 -2.84005135e-01 -2.37037048e-01
2.07849637e-01 -1.09402704e+00 3.60454656e-02 4.99936163e-01
-1.01786292e+00 -1.22797656e+00 -3.86264116e-01 -7.34113753e-01
7.43609011e-01 2.78112054e-01 2.06252351e-01 -1.13419719e-01
-2.96974331e-01 -3.02726686e-01 -5.00189066e-01 -2.14201272e-01
-4.67138216e-02 8.34591612e-02 1.35166854e-01 2.13482827e-01
4.51289386e-01 -4.11489218e-01 -4.63391989e-01 2.55928934e-01
-6.58456564e-01 1.00836672e-01 6.94754064e-01 1.43402922e+00
2.53990650e-01 1.86144840e-02 7.57632673e-01 -1.40916014e+00
2.12583885e-01 -7.80364156e-01 -8.84919226e-01 4.66329157e-01
-1.26311314e+00 2.84456331e-02 1.05644751e+00 -4.22487319e-01
-1.13524961e+00 1.85388252e-01 9.19798091e-02 -3.96255404e-01
-6.67386055e-02 3.62590104e-01 -3.16441000e-01 -6.97936654e-01
3.97551179e-01 5.50281823e-01 2.41150185e-01 -5.75882912e-01
6.75329342e-02 1.24212718e+00 1.33502990e-01 -3.83539230e-01
8.86433125e-01 4.97459054e-01 1.12430587e-01 -5.51926851e-01
-9.72202957e-01 -6.57757580e-01 -6.98978245e-01 3.00306559e-01
3.16750735e-01 -7.52930999e-01 -6.77031100e-01 5.87948859e-01
-5.40413558e-01 2.58257180e-01 1.15403965e-01 5.45560598e-01
-5.71717858e-01 7.75698245e-01 -3.94012183e-01 -7.23607183e-01
-4.33044612e-01 -1.13027930e+00 5.98422170e-01 3.52174491e-01
3.24032485e-01 -1.00735605e+00 -2.82817811e-01 9.20620322e-01
2.96140611e-01 3.11204821e-01 1.06037533e+00 -1.15923238e+00
-5.80117404e-01 -6.75464272e-01 -3.24230224e-01 7.36656189e-01
1.83276609e-01 -5.54403067e-01 -9.92071569e-01 -6.79497063e-01
2.45207161e-01 -5.73174238e-01 8.86258483e-01 -2.27584884e-01
1.24016345e+00 -4.18379635e-01 -3.14958930e-01 9.62644279e-01
1.73224473e+00 2.97282666e-01 4.44309235e-01 7.90954411e-01
4.13776875e-01 3.79548430e-01 1.18326974e+00 3.43205214e-01
1.26182199e-01 8.09964597e-01 2.59432733e-01 -8.82376060e-02
3.82966697e-01 -3.87153924e-02 2.30984613e-01 5.04180431e-01
3.29836160e-01 2.49392629e-01 -6.18277967e-01 2.10535839e-01
-2.04146075e+00 -8.68169010e-01 1.19159617e-01 2.43322825e+00
8.76981616e-01 1.46823004e-01 -1.61915943e-01 3.69315922e-01
7.08466411e-01 1.16332076e-01 -7.27591276e-01 -5.50402343e-01
-1.26432627e-01 3.68204325e-01 8.50417912e-01 6.91188812e-01
-1.52599633e+00 9.01565552e-01 6.51799202e+00 1.13128722e+00
-1.24658072e+00 1.44250646e-01 2.94850409e-01 -4.52298932e-02
3.73423934e-01 2.78374374e-01 -8.87803257e-01 6.41314030e-01
5.61647177e-01 -2.79889464e-01 3.63467813e-01 1.45104253e+00
-2.92373806e-01 -3.92718703e-01 -1.06060004e+00 9.49293435e-01
3.28324616e-01 -9.72596526e-01 2.23107040e-02 7.98959285e-02
5.36963403e-01 -2.63439327e-01 1.73825637e-01 3.96572471e-01
6.56438246e-02 -5.92691243e-01 3.31467479e-01 9.16371495e-02
6.95628166e-01 -6.96078837e-01 9.75852251e-01 5.71243823e-01
-7.87207425e-01 -3.52841079e-01 -4.63020742e-01 -1.21022828e-01
-2.91196138e-01 4.18306172e-01 -1.02393305e+00 7.80513883e-01
6.00732684e-01 3.31081629e-01 -6.89682484e-01 8.99872839e-01
-8.16844478e-02 5.57785630e-01 -2.87811458e-01 -1.15847297e-01
1.94728643e-01 -3.21170062e-01 5.54394245e-01 8.58629167e-01
-2.81574223e-02 -6.22130074e-02 5.03446281e-01 3.14417750e-01
3.57867837e-01 4.10520405e-01 -5.75422704e-01 2.52469420e-01
3.12136918e-01 1.37629974e+00 -3.56826812e-01 -4.90410030e-01
-6.73348308e-01 1.32870221e+00 4.61032361e-01 9.82118472e-02
-9.07505512e-01 -9.29568946e-01 4.98921424e-01 -4.55890363e-03
3.62042546e-01 -1.50119409e-01 -3.95854592e-01 -1.46498132e+00
4.53302532e-01 -7.73473978e-01 6.87292635e-01 -2.76547134e-01
-1.49551821e+00 6.99273288e-01 -1.88632414e-03 -1.33769500e+00
-8.56597722e-02 -9.74943936e-01 -3.89091969e-01 9.99015272e-01
-1.94843090e+00 -1.38155746e+00 -4.46227938e-01 8.37917984e-01
3.58546853e-01 -5.11343122e-01 8.88850152e-01 1.78926602e-01
-6.70147777e-01 9.40319717e-01 6.37107968e-01 -4.62484807e-02
8.41806352e-01 -1.40203440e+00 -3.36660177e-01 5.93390286e-01
-3.70943338e-01 7.49991655e-01 4.84796256e-01 -5.50186932e-01
-1.28813505e+00 -7.09376097e-01 5.66874266e-01 -3.30491006e-01
7.74672091e-01 -2.84588784e-01 -7.66786337e-01 7.18188763e-01
1.43225849e-01 -2.06857710e-03 1.22123277e+00 1.12593606e-01
-6.14325166e-01 -4.02023405e-01 -1.69346845e+00 4.71531488e-02
5.83441138e-01 -4.19412285e-01 -9.05360222e-01 3.85354906e-01
4.18283403e-01 -3.25758874e-01 -7.39265978e-01 3.75807315e-01
6.89609408e-01 -7.78631091e-01 8.47482383e-01 -9.81680155e-01
1.57919064e-01 -2.27108121e-01 -4.99392986e-01 -1.11865425e+00
-1.89096659e-01 -2.42451131e-01 -3.57755780e-01 8.35100293e-01
3.14367354e-01 -1.19515169e+00 1.01113403e+00 5.25407612e-01
2.00401053e-01 -1.13686132e+00 -1.12306297e+00 -1.05349326e+00
-1.30028561e-01 2.61908859e-01 2.63380408e-01 1.22221100e+00
3.01527530e-01 -7.41446018e-03 -5.52064061e-01 3.16703737e-01
9.66944635e-01 5.66013634e-01 5.85986376e-01 -1.43356395e+00
-5.96232414e-01 1.68433219e-01 -8.39625120e-01 -8.35065126e-01
5.40907741e-01 -1.23902881e+00 -4.54856873e-01 -9.61341023e-01
6.13047361e-01 -7.17254996e-01 -8.04254234e-01 7.41817772e-01
-3.31791848e-01 3.73407863e-02 4.25261185e-02 1.48890778e-01
-3.06302488e-01 4.96320635e-01 8.37471843e-01 -1.67279944e-01
3.37045118e-02 1.88201457e-01 -8.91128898e-01 7.55346179e-01
6.34614885e-01 -3.76390517e-01 -4.90804940e-01 1.72918313e-03
-4.67334837e-01 1.86053142e-02 3.80855352e-01 -9.93801236e-01
1.23006120e-01 -5.44777930e-01 5.39712489e-01 -3.99066269e-01
6.24796450e-01 -1.11311150e+00 -3.46219271e-01 5.49773097e-01
-1.64642468e-01 -2.68861920e-01 3.12341824e-02 7.56090224e-01
-4.71552879e-01 -3.91178131e-01 1.05224252e+00 1.23236425e-01
-7.57758200e-01 2.63090074e-01 2.39829928e-01 -4.70039189e-01
1.56388235e+00 -4.26100612e-01 -5.42288780e-01 5.38552627e-02
-7.75328636e-01 3.75539780e-01 4.57154065e-01 4.35161024e-01
4.83815342e-01 -1.40072513e+00 -2.89019048e-01 2.90457517e-01
2.88774014e-01 -7.61991262e-01 2.27859318e-01 6.66402519e-01
-5.22732556e-01 7.38921165e-01 -6.63399279e-01 -4.13411945e-01
-1.54788411e+00 8.20994616e-01 6.90093264e-02 -3.35280776e-01
-4.14105207e-01 5.76417089e-01 2.49930993e-02 -3.33037764e-01
1.59431040e-01 3.11209947e-01 -2.31823891e-01 -1.07382573e-01
4.01993901e-01 3.70158792e-01 2.23456591e-01 -4.11026090e-01
-4.60252583e-01 2.05932543e-01 -5.80012560e-01 9.57205147e-02
1.28357172e+00 1.02605902e-01 -1.53189585e-01 3.13217759e-01
1.45614731e+00 5.91671187e-03 -8.70462716e-01 -4.46073353e-01
-7.11765513e-02 -1.05383039e+00 9.43467021e-02 -8.57467175e-01
-7.83937395e-01 7.24947989e-01 9.09354627e-01 7.25478679e-02
1.05256677e+00 -3.83587629e-01 9.24560547e-01 8.36006820e-01
7.19924569e-01 -1.14069295e+00 -2.55436122e-01 1.95697904e-01
6.19882584e-01 -1.67107284e+00 1.37612790e-01 -8.44913185e-01
-6.29793465e-01 1.09512079e+00 1.03396249e+00 -2.86295533e-01
9.25800622e-01 8.27709660e-02 -5.33915311e-02 2.53968120e-01
-5.96113265e-01 9.82184857e-02 5.52513674e-02 4.12540525e-01
5.39264232e-02 5.33342771e-02 -6.70486569e-01 8.18769872e-01
-5.71135841e-02 -7.29936063e-02 4.09292072e-01 1.33976912e+00
-6.11436486e-01 -1.31198359e+00 -2.56406516e-01 7.51125634e-01
-4.65623438e-01 -3.45636159e-02 -3.23858112e-01 6.50652111e-01
8.49044621e-02 8.99795711e-01 -3.72820884e-01 -4.74683493e-01
1.45826101e-01 4.13437635e-01 4.66987789e-01 -5.62862694e-01
-3.12740058e-01 -4.94732469e-01 -2.44636796e-02 -5.28257906e-01
5.24146371e-02 -4.26823020e-01 -1.04924905e+00 -7.22349435e-02
-8.52617025e-01 3.49082381e-01 8.26335430e-01 7.92123735e-01
2.05941692e-01 6.79325163e-02 1.04788780e+00 -3.55172962e-01
-1.27122998e+00 -7.31593966e-01 -8.31109941e-01 4.40352917e-01
3.54558617e-01 -1.03799236e+00 -6.71007752e-01 -2.11701274e-01] | [8.258935928344727, 4.138667583465576] |
0e03cbf9-15cd-4d0d-ae09-3c26a8ae4638 | character-focused-video-thumbnail-retrieval | 2204.06563 | null | https://arxiv.org/abs/2204.06563v1 | https://arxiv.org/pdf/2204.06563v1.pdf | Character-focused Video Thumbnail Retrieval | We explore retrieving character-focused video frames as candidates for being video thumbnails. To evaluate each frame of the video based on the character(s) present in it, characters (faces) are evaluated in two aspects: Facial-expression: We train a CNN model to measure whether a face has an acceptable facial expression for being in a video thumbnail. This model is trained to distinguish faces extracted from artworks/thumbnails, from faces extracted from random frames of videos. Prominence and interactions: Character(s) in the thumbnail should be important character(s) in the video, to prevent the algorithm from suggesting non-representative frames as candidates. We use face clustering to identify the characters in the video, and form a graph in which the prominence (frequency of appearance) of the character(s), and their interactions (co-occurrence) are captured. We use this graph to infer the relevance of the characters present in each candidate frame. Once every face is scored based on the two criteria above, we infer frame level scores by combining the scores for all the faces within a frame. | ['Hossein Taghavi', 'Nagendra Kamath', 'Shervin Ardeshir'] | 2022-04-13 | null | null | null | null | ['face-clustering'] | ['computer-vision'] | [ 2.87413508e-01 8.78780931e-02 -2.77223229e-01 -3.12672257e-01
-3.12920690e-01 -4.87623781e-01 4.58051234e-01 1.12959497e-01
-8.25926438e-02 2.41031244e-01 5.70813954e-01 2.94723839e-01
-1.19789340e-01 -5.39326429e-01 -6.19509935e-01 -6.90735579e-01
-1.76264077e-01 -1.16620451e-01 8.01950693e-02 1.59808397e-01
4.24920201e-01 9.01822269e-01 -2.16161442e+00 8.57605219e-01
1.28623649e-01 1.37205625e+00 1.18354540e-02 5.69355607e-01
-7.30124908e-03 1.03679097e+00 -8.34067404e-01 -3.84185821e-01
-3.04161627e-02 -6.80190384e-01 -5.27974725e-01 5.09108961e-01
9.24214900e-01 -5.26559353e-01 -2.35536769e-01 1.14369547e+00
-5.11766896e-02 1.58941969e-01 6.32039249e-01 -1.35279000e+00
-3.26127678e-01 3.90461892e-01 -7.39570737e-01 4.01373088e-01
7.53480732e-01 -1.16298646e-01 8.42216730e-01 -1.26808035e+00
8.39054525e-01 1.46809697e+00 2.74498194e-01 5.94112873e-01
-6.86633945e-01 -7.59708583e-01 2.01651663e-01 4.42141205e-01
-1.54386187e+00 -8.24319720e-01 9.38391984e-01 -5.50235689e-01
5.12652934e-01 3.95615757e-01 9.69396889e-01 8.20305884e-01
-5.36970515e-03 8.01486254e-01 4.03391004e-01 -3.30695510e-01
3.39173615e-01 -1.30158231e-01 -2.17508584e-01 7.29367852e-01
-2.09892467e-01 -5.28947175e-01 -7.97676027e-01 -2.13376954e-01
7.62854517e-01 6.61543831e-02 -1.56485736e-01 2.76437700e-01
-9.93931413e-01 6.02591813e-01 1.19238451e-01 3.13796014e-01
-7.39668310e-01 1.04181953e-01 2.39351407e-01 3.96392047e-02
4.82170641e-01 -4.75420319e-02 2.63884198e-02 -8.83117318e-02
-1.16587687e+00 1.92400366e-01 4.35145408e-01 7.12684631e-01
1.05284190e+00 -1.58540592e-01 -4.90690053e-01 7.72055566e-01
2.08948538e-01 -2.99858414e-02 3.40628922e-02 -1.39368474e+00
5.35758969e-04 1.05809927e+00 -1.07663162e-01 -1.51364934e+00
1.71148047e-01 4.18855071e-01 -4.58740413e-01 1.81729466e-01
5.21744341e-02 -4.26617600e-02 -6.04244411e-01 1.56015086e+00
3.47519457e-01 3.86465102e-01 -3.95141661e-01 9.78044868e-01
1.10826099e+00 7.46458769e-01 2.58722544e-01 -5.91571450e-01
1.57442486e+00 -4.53396648e-01 -6.19784832e-01 -3.90929021e-02
-2.45501641e-02 -8.57207179e-01 8.46073210e-01 2.98550308e-01
-1.06346714e+00 -7.41989374e-01 -7.69055903e-01 1.75223112e-01
7.84129426e-02 4.70781624e-01 1.09819442e-01 3.20664465e-01
-1.15230203e+00 7.38552094e-01 -1.08103156e-01 -3.23815584e-01
4.20370430e-01 3.49568307e-01 -5.90406239e-01 2.33966872e-01
-8.67886126e-01 3.58558983e-01 1.38735920e-01 2.39263456e-02
-9.35357749e-01 -2.71535695e-01 -7.30317473e-01 8.52800533e-02
2.95429975e-01 -1.08084589e-01 7.98816621e-01 -2.25777888e+00
-1.11632133e+00 1.10840762e+00 -5.30717075e-01 4.01680619e-02
1.71538889e-01 -3.89738567e-02 -5.89522243e-01 7.45054960e-01
9.80075374e-02 8.82383227e-01 1.42804313e+00 -1.23944533e+00
-9.54074144e-01 -3.83140534e-01 1.95059121e-01 2.99304873e-01
-5.67891121e-01 8.04048717e-01 -1.00576162e+00 -6.23602509e-01
-2.71713063e-02 -7.17718005e-01 3.13663691e-01 2.37846881e-01
-3.62517834e-01 -6.41309619e-01 1.20846593e+00 -7.47855365e-01
1.37826383e+00 -2.43456531e+00 -6.41177744e-02 3.39123249e-01
2.85603166e-01 6.05720878e-02 -1.06054507e-01 9.69013274e-02
-2.09981188e-01 3.21841389e-01 3.05345356e-01 -1.19082473e-01
-4.63598043e-01 -2.21535012e-01 6.05739132e-02 4.71229941e-01
3.80831301e-01 4.07553226e-01 -8.69102895e-01 -8.35673571e-01
1.19787700e-01 6.44875765e-01 -3.55567545e-01 2.56873995e-01
-2.23425418e-01 1.95142642e-01 -3.93015414e-01 8.76344502e-01
4.70550984e-01 7.57305371e-03 1.53179273e-01 -5.83652973e-01
-1.67101230e-02 -1.95876941e-01 -9.81719553e-01 6.49734437e-01
6.68795332e-02 1.04158783e+00 1.00245558e-01 -4.67955440e-01
8.93442750e-01 3.15147936e-01 6.95230722e-01 -3.02420914e-01
4.03637104e-02 -1.15752816e-01 -1.42679617e-01 -8.67363155e-01
4.17138606e-01 3.93709958e-01 1.92777798e-01 5.01933813e-01
-1.09392248e-01 5.65870166e-01 5.40613294e-01 2.39939675e-01
8.56311917e-01 4.29432467e-02 1.02028683e-01 -2.53626853e-01
8.34759116e-01 -6.85242712e-01 5.04672229e-01 3.38824213e-01
-3.10300946e-01 6.15071058e-01 7.61859477e-01 -8.09293807e-01
-9.03144777e-01 -5.68400741e-01 2.72552729e-01 1.52884865e+00
1.53492972e-01 -8.30440462e-01 -1.05820787e+00 -5.69118321e-01
-2.46293694e-01 2.33667001e-01 -1.02731895e+00 -7.23021254e-02
-4.69661742e-01 1.19370893e-02 2.59675056e-01 3.48767251e-01
2.35688269e-01 -1.56527090e+00 -7.87978053e-01 -1.69974774e-01
-3.26068550e-01 -9.84230101e-01 -6.35613561e-01 -4.50100154e-01
-2.97006905e-01 -1.26654005e+00 -5.67160308e-01 -9.78023946e-01
1.15264785e+00 2.78913677e-01 9.51235831e-01 5.52339435e-01
-3.88833769e-02 4.25503045e-01 -5.82755089e-01 -1.03512045e-03
-4.55619752e-01 -7.26254702e-01 2.69802034e-01 7.78774977e-01
6.20707452e-01 -4.34597619e-02 -8.10092330e-01 4.29866016e-01
-6.63450956e-01 -1.45120293e-01 1.15222700e-01 1.11434788e-01
6.35016322e-01 2.85679400e-01 1.16860215e-02 -4.18005854e-01
5.52942634e-01 -4.98645008e-01 -4.02354896e-02 3.25267792e-01
1.24928348e-01 -6.24468505e-01 4.82302874e-01 -5.82093716e-01
-6.98320568e-01 2.93835610e-01 2.11843550e-01 -9.82590854e-01
-4.26532716e-01 2.90883034e-01 -1.13709494e-01 2.05222800e-01
4.68193501e-01 1.92210212e-01 -3.96236926e-02 -1.58126373e-02
1.37232267e-03 6.69710934e-01 5.34219086e-01 -4.57406223e-01
2.86215037e-01 4.61109430e-01 -2.42456004e-01 -1.24421811e+00
-5.74963748e-01 -2.50176758e-01 -5.53264201e-01 -1.33760536e+00
1.24496782e+00 -8.63191962e-01 -9.33478534e-01 1.51876450e-01
-1.26044774e+00 2.87923485e-01 1.35713428e-01 3.15058380e-01
-2.61221707e-01 5.57557404e-01 -4.27648723e-01 -1.19612682e+00
-1.85655773e-01 -1.16099393e+00 1.18163693e+00 6.14641130e-01
-8.33145380e-01 -3.52545410e-01 -4.89358336e-01 1.72065839e-01
-1.52361959e-01 3.30018491e-01 7.67846286e-01 -2.96789825e-01
-2.38736883e-01 -4.14877564e-01 -1.68423027e-01 1.39755085e-01
2.35762924e-01 1.14056635e+00 -9.49267924e-01 -1.39214754e-01
-1.05699457e-01 -8.48090351e-02 6.20727062e-01 5.86718023e-01
1.62353492e+00 -6.29628479e-01 -3.27209532e-01 3.67063791e-01
8.53200674e-01 4.87308890e-01 8.22003663e-01 1.25749484e-01
5.72104692e-01 8.79316986e-01 5.91710091e-01 8.21739554e-01
-9.69714299e-02 5.90078473e-01 6.32859468e-01 6.31239563e-02
1.00814000e-01 -2.65938640e-01 1.03930926e+00 3.49241555e-01
-3.90994221e-01 -2.43063018e-01 -6.15929604e-01 3.39771658e-01
-1.64245200e+00 -1.37413502e+00 -2.01644301e-01 2.09827662e+00
4.49551016e-01 -3.58191840e-02 4.79818165e-01 -4.11535650e-02
1.42095733e+00 1.57684967e-01 -4.39268559e-01 -3.47752064e-01
-1.08287290e-01 -4.14344855e-02 -2.02884272e-01 2.20029168e-02
-1.17170906e+00 8.16212952e-01 6.45400429e+00 7.56528080e-01
-1.20736122e+00 -4.86400276e-01 1.48529315e+00 -4.71587688e-01
-1.68553978e-01 -1.31995931e-01 -7.54652202e-01 7.19657362e-01
5.48320055e-01 -1.89907085e-02 4.38841522e-01 8.92998099e-01
3.37201297e-01 -3.78348708e-01 -1.29797065e+00 1.20227075e+00
5.51194727e-01 -1.33927059e+00 3.25024039e-01 -1.46647140e-01
5.99761367e-01 -5.78117132e-01 6.13415316e-02 -2.38724709e-01
-2.76663750e-01 -1.04434490e+00 1.10391474e+00 5.66093981e-01
1.03308177e+00 -9.28188086e-01 5.09423614e-01 3.77661586e-02
-1.34590340e+00 -2.13040262e-01 -3.11855435e-01 -6.71924278e-02
-1.53681666e-01 4.39522922e-01 -5.34971178e-01 -1.47785962e-01
1.05382478e+00 7.17683196e-01 -6.74753964e-01 5.58818042e-01
-1.07598349e-01 4.11293626e-01 -1.61989585e-01 -1.11878984e-01
2.58685667e-02 -3.99387836e-01 4.15092289e-01 1.22384596e+00
6.07895792e-01 1.92934766e-01 2.32871324e-02 8.05365384e-01
-1.01825908e-01 1.95416823e-01 -4.32004899e-01 -2.71735579e-01
8.52253437e-01 1.61741138e+00 -1.17698979e+00 -4.03136134e-01
-5.04865348e-01 9.48632777e-01 -9.22289118e-02 1.71525821e-01
-7.36762226e-01 2.74959654e-02 7.97559619e-01 4.78621185e-01
2.49338061e-01 3.85705501e-01 3.54683489e-01 -5.63064635e-01
1.25524849e-01 -9.72626150e-01 4.92527902e-01 -1.00125146e+00
-1.15830779e+00 9.96132851e-01 -3.13455105e-01 -1.63292384e+00
-2.64822364e-01 -4.04417932e-01 -9.74190533e-01 4.58252162e-01
-5.98272085e-01 -8.20622087e-01 -6.59409523e-01 7.71745503e-01
7.33250797e-01 -3.25354099e-01 5.22090793e-01 7.83187598e-02
-6.24344409e-01 4.09208745e-01 -5.40802538e-01 5.59474468e-01
4.73251164e-01 -6.74494028e-01 2.12059796e-01 7.49696434e-01
2.00109482e-01 7.94120550e-01 4.90687132e-01 -8.74655068e-01
-1.21498871e+00 -1.02083135e+00 8.92133236e-01 -2.82536477e-01
3.25942159e-01 -1.23181865e-01 -5.41083157e-01 3.58571708e-01
6.17902912e-02 -2.33215794e-01 7.83543348e-01 -1.45354450e-01
-7.83987716e-02 5.54037057e-02 -8.71024311e-01 7.63574183e-01
7.14128315e-01 -7.62717009e-01 -3.27945054e-01 1.97301716e-01
1.25605926e-01 2.14178458e-01 -5.39004624e-01 2.00213149e-01
8.46768022e-01 -1.26513219e+00 8.33384037e-01 -3.34362358e-01
9.57923412e-01 -4.04356658e-01 -8.66578594e-02 -8.18206847e-01
-5.41177094e-01 -6.60488367e-01 -2.21797004e-01 1.35386336e+00
2.05770088e-03 4.36430305e-01 9.71368909e-01 5.95636964e-01
2.50292689e-01 -5.98034680e-01 -9.48423982e-01 -2.52911597e-01
-6.97337389e-01 -2.31739208e-01 4.97284710e-01 1.00413203e+00
2.25764319e-01 1.86950669e-01 -5.35723031e-01 -2.46287242e-01
3.02098066e-01 -9.00762156e-02 7.06618190e-01 -1.41510546e+00
1.25320390e-01 -8.23046327e-01 -4.81602311e-01 -4.82938141e-01
2.61715293e-01 -6.07112825e-01 -4.13743146e-02 -1.21526349e+00
5.81992745e-01 2.92653680e-01 -1.87991768e-01 4.13891912e-01
-1.56702235e-01 5.12130797e-01 3.98129970e-01 3.87192905e-01
-7.66340315e-01 4.10409309e-02 8.98250043e-01 -9.11849961e-02
-2.26576161e-02 -9.49484184e-02 -4.89967138e-01 1.13711941e+00
4.08213645e-01 -1.57958120e-01 -8.79634395e-02 -7.22673163e-02
2.98828691e-01 3.09373766e-01 2.60651797e-01 -1.18048537e+00
1.54953510e-01 -2.60619640e-01 9.90808129e-01 -6.19339526e-01
4.21450555e-01 -7.13027537e-01 4.82900739e-01 2.94026166e-01
-4.27281231e-01 3.41606200e-01 -9.53681618e-02 2.80002892e-01
-2.24586800e-01 -3.56946468e-01 9.10990894e-01 -2.10718676e-01
-9.43903089e-01 4.63461369e-01 -7.73536444e-01 -4.56162602e-01
1.17074633e+00 -7.74874985e-01 8.93943533e-02 -8.31386626e-01
-7.75023639e-01 -2.62455970e-01 6.48090363e-01 6.09513521e-01
1.05866039e+00 -1.50797236e+00 -8.08286011e-01 4.16382968e-01
2.36739412e-01 -5.24626195e-01 3.43612403e-01 4.46842015e-01
-6.00958705e-01 -3.68362904e-01 -4.56367880e-01 -5.00785351e-01
-1.79039454e+00 4.52259988e-01 4.66538481e-02 7.04957843e-01
-3.76562178e-01 1.05906057e+00 4.20747191e-01 6.98117733e-01
3.72205973e-01 1.52821112e-02 -8.07986081e-01 6.65293097e-01
1.10418129e+00 2.72899628e-01 -2.66398877e-01 -1.40342593e+00
-5.99648476e-01 7.38249004e-01 -6.04912601e-02 9.11556780e-02
1.20394015e+00 9.03240740e-02 -4.46966738e-01 3.04173648e-01
1.35696387e+00 1.84417665e-01 -1.50185716e+00 6.91297501e-02
-9.53061804e-02 -6.20741606e-01 -1.28151789e-01 -3.27662408e-01
-1.46569240e+00 5.26755214e-01 4.63627428e-01 1.83261573e-01
1.26575434e+00 2.42520586e-01 4.16943341e-01 -2.51237042e-02
-1.19447829e-02 -1.34466517e+00 5.19346178e-01 3.90883148e-01
1.03761625e+00 -7.87089586e-01 4.06111851e-02 -2.45971769e-01
-7.43074894e-01 1.53590477e+00 6.98162019e-01 -2.28880167e-01
5.87160885e-01 1.07134707e-01 2.43140850e-02 -4.98138964e-01
-6.46283627e-01 2.98194438e-02 5.00677824e-01 4.84267086e-01
4.58231360e-01 -9.22608525e-02 -1.06109545e-01 3.73685092e-01
-1.03118129e-01 -2.17807204e-01 4.40016568e-01 4.59238708e-01
-6.55706227e-01 -3.33053291e-01 -4.55616534e-01 6.32181406e-01
-5.74214458e-01 1.28109738e-01 -1.13372052e+00 1.14315972e-01
4.12776679e-01 9.84836042e-01 6.37929559e-01 -7.07072198e-01
-2.06838828e-02 -3.86762954e-02 3.36467355e-01 -3.48749667e-01
-6.17123127e-01 4.82117474e-01 -1.22288145e-01 -6.67980313e-01
-6.80559993e-01 -7.23115802e-01 -1.10632634e+00 -3.77657503e-01
-9.32927709e-03 -2.19419673e-02 2.93648839e-01 7.41131246e-01
2.93541700e-01 1.14220738e-01 7.80965447e-01 -1.13103116e+00
3.68462503e-01 -6.35649145e-01 -5.69531918e-01 8.05363595e-01
6.09448925e-02 -6.23048961e-01 -3.71473551e-01 4.68312174e-01] | [10.204224586486816, 0.45890605449676514] |
f129b3ba-9491-4068-95ec-c03b3b9e1c50 | alime-mkg-a-multi-modal-knowledge-graph-for | 2109.07411 | null | https://arxiv.org/abs/2109.07411v1 | https://arxiv.org/pdf/2109.07411v1.pdf | AliMe MKG: A Multi-modal Knowledge Graph for Live-streaming E-commerce | Live streaming is becoming an increasingly popular trend of sales in E-commerce. The core of live-streaming sales is to encourage customers to purchase in an online broadcasting room. To enable customers to better understand a product without jumping out, we propose AliMe MKG, a multi-modal knowledge graph that aims at providing a cognitive profile for products, through which customers are able to seek information about and understand a product. Based on the MKG, we build an online live assistant that highlights product search, product exhibition and question answering, allowing customers to skim over item list, view item details, and ask item-related questions. Our system has been launched online in the Taobao app, and currently serves hundreds of thousands of customers per day. | ['Ji Zhang', 'Zhongzhou Zhao', 'Wei Zhou', 'Zhixiong Zeng', 'Yunzhou Shi', 'Fu Sun', 'Feng-Lin Li', 'Hehong Chen', 'Guohai Xu'] | 2021-09-13 | null | null | null | null | ['multi-modal-knowledge-graph'] | ['knowledge-base'] | [-2.60151446e-01 1.29861280e-01 -6.32910907e-01 -6.07434332e-01
-7.36894310e-01 -6.00572884e-01 -1.76969081e-01 5.02292633e-01
1.77971140e-01 2.62165163e-02 3.40678513e-01 -1.25449210e-01
-5.51169097e-01 -1.05451643e+00 -4.13110405e-01 -2.35087574e-01
-1.87427819e-01 7.00176597e-01 2.76851743e-01 -5.82804561e-01
1.89881429e-01 1.77220434e-01 -1.44811916e+00 6.39687002e-01
6.90767407e-01 1.36598682e+00 6.23589873e-01 4.30717379e-01
-3.56249809e-01 8.00230920e-01 -4.16874677e-01 -9.02309060e-01
-1.64213300e-01 -1.96053177e-01 -9.91852224e-01 3.60937744e-01
2.44000375e-01 -5.54200292e-01 -3.14580142e-01 1.08689415e+00
1.73637569e-01 1.13836564e-01 -7.81683922e-02 -1.34315526e+00
-9.78582799e-01 1.11893439e+00 -3.39787304e-01 6.26959324e-01
9.68077540e-01 -2.81519771e-01 1.69183648e+00 -6.88316882e-01
6.10062718e-01 9.71020579e-01 8.90761316e-02 -1.58022851e-01
-9.39176559e-01 -5.39313138e-01 4.23151791e-01 8.09808791e-01
-1.31416917e+00 -2.48484716e-01 8.70159686e-01 2.72392660e-01
6.30901396e-01 5.38180530e-01 1.01967537e+00 7.58081436e-01
-2.01973811e-01 1.19671392e+00 2.58458167e-01 -3.79681140e-02
1.72724277e-01 6.14269495e-01 3.08339804e-01 4.51095074e-01
4.47220448e-03 -2.78497219e-01 -1.35596097e+00 -1.54515162e-01
6.15626216e-01 2.33500004e-01 -3.55978161e-01 -8.18192884e-02
-7.57544935e-01 1.02435839e+00 4.43562895e-01 1.37905613e-01
-9.41751480e-01 -1.51272118e-01 1.37694255e-01 4.53859895e-01
3.15931678e-01 3.94903243e-01 -2.03428954e-01 -2.85165966e-01
-4.90401626e-01 2.25479379e-01 1.14705515e+00 1.35845888e+00
6.92447543e-01 -4.80086327e-01 5.78620099e-02 7.32835293e-01
5.55588067e-01 4.92890000e-01 2.06729263e-01 -1.11385202e+00
8.69091600e-02 7.89797425e-01 1.12725468e-03 -1.22182035e+00
-1.25718534e-01 -7.76975155e-01 -3.69542718e-01 -5.44289649e-01
-1.47121683e-01 2.90649772e-01 -1.09081104e-01 9.28159118e-01
3.83415550e-01 5.39563224e-02 -2.58438975e-01 1.11910391e+00
1.12567163e+00 8.26625705e-01 -1.28304183e-01 -3.17951590e-01
1.88235748e+00 -1.10112262e+00 -1.07546747e+00 1.81788206e-02
5.41763783e-01 -8.25760007e-01 1.18523300e+00 1.14514613e+00
-1.13647223e+00 -4.41190749e-01 -7.73284137e-01 -1.26411036e-01
-2.56226778e-01 -2.80250371e-01 9.68201160e-01 4.64180380e-01
-5.68469644e-01 4.01566446e-01 -3.00837874e-01 -4.49585646e-01
2.90033281e-01 1.96083084e-01 5.76794483e-02 -3.51102024e-01
-1.38909292e+00 3.29047203e-01 1.94060758e-01 -2.28059595e-03
-7.99657822e-01 -8.60343456e-01 -6.46754861e-01 4.54641759e-01
8.43826294e-01 -6.10003293e-01 1.58245111e+00 -7.40097523e-01
-1.39669120e+00 5.44873595e-01 -1.14922971e-01 -2.12284237e-01
-3.68058413e-01 -4.34536487e-01 -1.42660344e+00 5.51075339e-01
6.07242584e-02 3.06936771e-01 6.88898027e-01 -8.79757166e-01
-1.18203175e+00 -6.13954246e-01 5.13362467e-01 4.35953021e-01
-6.59993947e-01 -2.89389379e-02 -9.21557546e-01 -1.09823905e-01
4.08035785e-01 -4.35881466e-01 1.83337569e-01 -1.14794672e-01
-2.87593633e-01 -3.70342046e-01 1.05716467e+00 -5.83291709e-01
1.46223247e+00 -2.37121391e+00 -3.88823092e-01 5.06027341e-01
1.51631832e-01 -2.92403430e-01 -8.88246521e-02 7.60470390e-01
3.40716153e-01 -2.47068778e-01 8.87183130e-01 9.06039998e-02
2.56228000e-01 1.64260775e-01 -2.79993236e-01 -1.10708758e-01
-4.68777686e-01 1.04344451e+00 -8.27488780e-01 -2.88537949e-01
-1.95155725e-01 1.94728002e-01 -3.80783021e-01 -1.30977079e-01
-5.10703325e-01 5.24821095e-02 -8.38833511e-01 1.00770295e+00
7.17299163e-01 -8.63160789e-01 3.48913036e-02 -3.48396987e-01
2.03872576e-01 5.64321160e-01 -1.16172457e+00 1.90330982e+00
-4.84114707e-01 2.45049864e-01 4.71763998e-01 -5.92545569e-01
7.85099864e-01 1.18137084e-01 2.93062001e-01 -1.14018345e+00
-1.25152385e-02 -1.07578440e-02 -5.11822164e-01 -7.50654399e-01
9.32163954e-01 2.75933862e-01 -4.96822074e-02 5.44034898e-01
-1.66653231e-01 3.66539985e-01 3.01259339e-01 7.52483845e-01
8.78768146e-01 -3.49322736e-01 -2.16375917e-01 -7.84918517e-02
2.92898983e-01 1.88143626e-01 -8.27676132e-02 4.14448678e-01
1.72185704e-01 -5.09184487e-02 1.90317467e-01 -3.01901191e-01
-4.37560380e-01 -1.13355470e+00 1.56017974e-01 1.63954461e+00
7.90215075e-01 -6.25834465e-01 -2.97380865e-01 -5.46035767e-01
1.75353184e-01 1.11483288e+00 -1.02242000e-01 -1.84534788e-01
-1.20808885e-01 1.76891625e-01 -2.89458722e-01 3.85767281e-01
6.77576900e-01 -1.12959087e+00 -1.30683929e-01 3.18535686e-01
-7.42031336e-01 -9.12902951e-01 -1.08284116e+00 -2.03643367e-01
-7.12832212e-01 -6.55480087e-01 -4.20013368e-01 -1.04396987e+00
1.98124334e-01 7.65807569e-01 1.37543380e+00 -3.80935743e-02
-2.71244407e-01 7.82603145e-01 -6.76035345e-01 -3.13071549e-01
6.85050488e-02 2.36016452e-01 -2.34622762e-01 4.53653514e-01
8.05098355e-01 -7.33283877e-01 -1.02181137e+00 8.45068872e-01
-8.58321667e-01 -4.67166215e-01 2.96089709e-01 2.80669909e-02
6.97629869e-01 4.89391923e-01 8.29610765e-01 -9.95159507e-01
7.46052206e-01 -8.44612896e-01 -5.40067017e-01 2.58847237e-01
-8.85489523e-01 -4.95363891e-01 2.49408826e-01 -2.81998456e-01
-1.06720853e+00 -3.45197707e-01 -3.22317392e-01 -9.14921761e-02
2.64287051e-02 9.30379272e-01 -2.89675355e-01 2.20127076e-01
3.27665538e-01 2.29993939e-01 -2.92841375e-01 -6.11813903e-01
5.38161397e-01 8.70731711e-01 5.78985870e-01 2.26980090e-01
2.34034777e-01 5.12510836e-01 -5.57384551e-01 -7.80603588e-01
-1.12193263e+00 -1.28688252e+00 5.15631288e-02 -5.21858573e-01
4.92989928e-01 -9.05382216e-01 -1.78530920e+00 -1.73780471e-01
-7.01843321e-01 2.74680883e-01 -6.14780843e-01 5.42541325e-01
-1.99150756e-01 2.69775391e-01 -5.37920594e-01 -7.21959174e-01
-2.81446248e-01 -3.65305275e-01 8.69924366e-01 5.71644485e-01
-1.71982527e-01 -7.97387779e-01 -5.04650414e-01 1.03476572e+00
5.14174640e-01 -6.84774399e-01 7.54992783e-01 -8.71926069e-01
-1.20579851e+00 -4.53319162e-01 -1.01805925e-01 -2.69918032e-02
-9.28044021e-02 -7.33357787e-01 -6.04218006e-01 -1.19207777e-01
1.14566989e-01 -2.47086316e-01 4.51102108e-01 3.99877310e-01
1.21096182e+00 -3.91488850e-01 -5.07655442e-01 1.98554501e-01
1.06145573e+00 2.83355266e-01 5.29720426e-01 1.93080306e-01
3.00341491e-02 7.23619401e-01 1.03880811e+00 7.89159477e-01
7.63208985e-01 6.99392259e-01 9.16544259e-01 2.27321282e-01
2.09491208e-01 -7.60679245e-01 2.86166906e-01 9.23682868e-01
3.49447966e-01 -6.04633749e-01 8.18149522e-02 6.44914210e-01
-1.84876955e+00 -8.97198141e-01 -2.27578893e-01 1.97037792e+00
1.45939931e-01 2.69385606e-01 3.56606841e-01 -2.93322325e-01
6.76789522e-01 -1.48053885e-01 -8.03844571e-01 -1.30901039e-01
1.04520179e-01 -1.64504781e-01 4.69723552e-01 1.49746731e-01
-4.19310480e-01 5.66664696e-01 6.35436392e+00 1.19888270e+00
-6.84812129e-01 2.46843338e-01 3.69481325e-01 -3.43865842e-01
-8.41431439e-01 -9.57274958e-02 -1.04791367e+00 2.97375202e-01
8.96648467e-01 -4.53388959e-01 5.86724699e-01 1.13981438e+00
1.57163590e-01 -3.63192707e-01 -1.01445067e+00 1.13936293e+00
2.70936668e-01 -1.59381080e+00 1.35316243e-02 4.48960751e-01
3.01351577e-01 -3.54463726e-01 1.72907725e-01 3.87263060e-01
1.14006057e-01 -6.79022908e-01 4.92655158e-01 5.20536244e-01
2.21938908e-01 -9.90973353e-01 5.77719629e-01 7.45058656e-01
-1.33338082e+00 -3.74481738e-01 -6.34027198e-02 7.20799044e-02
1.02171469e+00 5.38255692e-01 -5.92481256e-01 6.84348881e-01
1.08629811e+00 4.76491570e-01 2.77515361e-03 1.15816414e+00
1.00640263e-02 5.58133721e-01 -3.11176568e-01 -4.39225197e-01
1.30396141e-02 -3.84170741e-01 2.88514167e-01 6.08226180e-01
7.06851900e-01 5.28691411e-01 1.03556931e-01 8.34013224e-01
-3.62463057e-01 2.40021303e-01 -2.14446083e-01 -4.80468541e-01
3.15413892e-01 1.59412777e+00 -6.35793328e-01 -2.51710236e-01
-5.19778490e-01 1.27849436e+00 -1.83275923e-01 9.45573151e-02
-8.57162476e-01 -4.02757764e-01 3.71802479e-01 6.08927131e-01
7.12210536e-01 2.97510196e-02 3.56184810e-01 -5.30514717e-01
1.48547575e-01 -6.82329774e-01 7.14972973e-01 -1.15527785e+00
-1.40571177e+00 3.52516025e-01 -3.17495793e-01 -9.58429635e-01
6.50172010e-02 -1.66645825e-01 -3.69787872e-01 4.93075669e-01
-1.57936013e+00 -8.36464643e-01 -3.81400406e-01 9.16420698e-01
8.45613539e-01 1.80586651e-02 7.45881736e-01 8.35099876e-01
3.49895693e-02 3.76301527e-01 -1.10919036e-01 -6.41722918e-01
3.46529901e-01 -7.34230340e-01 1.38542494e-02 -9.57867801e-02
6.84523046e-01 5.39296329e-01 5.84576368e-01 -6.10831201e-01
-1.73820496e+00 -6.99549079e-01 1.16312146e+00 -2.16110379e-01
7.15996265e-01 -2.87644327e-01 -7.78236508e-01 5.96399963e-01
2.85097331e-01 -4.66589987e-01 1.09148359e+00 5.55631578e-01
-8.89023393e-02 -6.11769199e-01 -9.49929893e-01 1.82548076e-01
1.16025126e+00 -5.44523597e-01 -2.79385716e-01 8.58287275e-01
1.09480262e+00 -8.73726904e-02 -9.06424999e-01 -2.49509633e-01
4.67133969e-01 -9.87706661e-01 7.89827943e-01 -2.25098714e-01
1.11713335e-01 -7.58282393e-02 -1.61349475e-01 -1.01219547e+00
-5.57751060e-01 -9.52518344e-01 -1.83873132e-01 1.33121300e+00
6.84549212e-01 -5.15845537e-01 1.14771998e+00 5.95199585e-01
-1.50137424e-01 -2.97693968e-01 -5.67166090e-01 -6.59575999e-01
-1.20258820e+00 -7.87361562e-01 8.76162350e-01 5.13264060e-01
5.30841768e-01 7.20194459e-01 -2.96659201e-01 5.22190690e-01
3.95722538e-01 6.35023057e-01 4.34461027e-01 -1.26255012e+00
-7.24150896e-01 -1.15888730e-01 4.74226940e-03 -2.15889788e+00
-7.12080300e-01 -8.52241099e-01 -4.92526621e-01 -1.70127332e+00
2.84874707e-01 -1.67180166e-01 -2.78363854e-01 -2.67803706e-02
7.44752884e-01 3.69207472e-01 -8.58792737e-02 4.02240008e-01
-1.40656495e+00 2.93553501e-01 1.72005546e+00 -5.02925739e-02
-1.93643883e-01 7.23107994e-01 -1.28027964e+00 5.11137724e-01
3.62953246e-01 -2.16383502e-01 -5.78668535e-01 -2.04671070e-01
1.12892878e+00 3.74925941e-01 1.65014327e-01 -2.96485126e-01
8.32434893e-01 1.59862280e-01 -7.03802258e-02 -1.21129107e+00
6.91945255e-01 -1.14275372e+00 2.50388503e-01 9.84711573e-02
-4.47122544e-01 -6.05738126e-02 -3.05126637e-01 8.26291442e-01
-5.44744313e-01 -4.38653737e-01 -1.55964002e-01 -1.01359747e-01
-9.36844885e-01 4.21468377e-01 -2.32384518e-01 -4.90078419e-01
9.58708227e-01 -2.94094682e-01 -3.81598622e-01 -1.22506738e+00
-1.07209301e+00 7.48723507e-01 -3.02975953e-01 5.20756721e-01
6.41042233e-01 -1.38456488e+00 -2.69597083e-01 2.56297916e-01
3.95634234e-01 -5.05709052e-01 7.80444264e-01 7.75643766e-01
-2.45677650e-01 6.24638915e-01 2.28638440e-01 -4.65356708e-01
-1.25185812e+00 7.27934778e-01 -4.29437429e-01 -6.50461242e-02
-5.49855709e-01 1.12491500e+00 2.35530995e-02 -5.72264334e-03
6.22499645e-01 9.46850106e-02 -4.72882271e-01 2.14403227e-01
8.88713717e-01 5.46998739e-01 1.04415610e-01 -1.64374143e-01
-1.40340552e-01 1.36858359e-01 -4.12462741e-01 -3.88623849e-02
1.14221835e+00 -9.73875165e-01 8.75194967e-02 3.79561722e-01
1.14966595e+00 7.11439401e-02 -7.63342500e-01 -4.94163454e-01
-1.52424842e-01 -5.83991826e-01 3.82746845e-01 -9.64671731e-01
-1.26277232e+00 4.11301941e-01 5.16130924e-01 1.27049339e+00
1.36339891e+00 8.31298828e-01 1.46678793e+00 3.22401047e-01
6.35473609e-01 -1.19615126e+00 4.20436800e-01 2.01465599e-02
7.69405186e-01 -1.12597895e+00 -1.06946886e-01 -9.30216908e-01
-9.17736888e-01 1.07918704e+00 2.45829254e-01 4.31833833e-01
8.98086727e-01 -1.18165873e-01 -1.48812579e-02 -9.55631971e-01
-8.92154932e-01 -2.56700188e-01 3.05714514e-02 5.62347531e-01
-1.09791659e-01 -7.91118816e-02 2.59002461e-03 9.27098870e-01
-2.44203091e-01 2.10464850e-01 9.75638777e-02 5.80218554e-01
-8.36332560e-01 -1.02276409e+00 4.45189923e-02 4.57668364e-01
-4.85945612e-01 -8.79681483e-02 -7.43510053e-02 3.44474554e-01
-1.32300094e-01 1.32801759e+00 2.22184643e-01 -3.26009035e-01
7.02747226e-01 -2.48930544e-01 7.65752643e-02 -6.53735340e-01
-5.06863952e-01 4.09135878e-01 3.91558111e-01 -8.05655181e-01
-3.09876144e-01 -4.42848593e-01 -1.22513139e+00 -4.21337515e-01
-7.82106519e-01 5.78246772e-01 9.49554265e-01 5.34703851e-01
7.65860796e-01 4.20966357e-01 7.64886796e-01 -1.11263722e-01
-3.08864206e-01 -6.14513278e-01 -1.39947987e+00 1.43729463e-01
-5.58006316e-02 -2.16650933e-01 -1.61479965e-01 5.18392399e-02] | [10.17806339263916, 5.747937202453613] |
e34eb2c9-72ed-4cc2-ba9c-1168d564975d | reproduction-study-using-public-data-of | null | null | https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0217541 | https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0217541&type=printable | Reproduction study using public data of: Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs | We have attempted to reproduce the results in Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs, published in JAMA 2016; 316(22), using publicly available data sets. We re-implemented the main method in the original study since the source code is not available. The original study used non-public fundus images from EyePACS and three hospitals in India for training. We used a different EyePACS data set from Kaggle. The original study used the benchmark data set Messidor-2 to evaluate the algorithm’s performance. We used another distribution of the Messidor-2 data set, since the original data set is no longer available. In the original study, ophthalmologists re-graded all images for diabetic retinopathy, macular edema, and image gradability. We have one diabetic retinopathy grade per image for our data sets, and we assessed image gradability ourselves. We were not able to reproduce the original study’s results with publicly available data. Our algorithm’s area under the receiver operating characteristic curve (AUC) of 0.951 (95% CI, 0.947-0.956) on the Kaggle EyePACS test set and 0.853 (95% CI, 0.835-0.871) on Messidor-2 did not come close to the reported AUC of 0.99 on both test sets in the original study. This may be caused by the use of a single grade per image, or different data. This study shows the challenges of reproducing deep learning method results, and the need for more replication and reproduction studies to validate deep learning methods, especially for medical image analysis. Our source code and instructions are available at: https://github.com/mikevoets/jama16-retina-replication. | ['Kajsa Møllersen', 'Mike Voets', 'Lars Ailo Bongo'] | 2019-06-06 | null | null | null | plos-one-2019-6 | ['diabetic-retinopathy-grading'] | ['medical'] | [-4.03899312e-01 -9.33082029e-02 -1.65610164e-01 -3.48386854e-01
-7.39919841e-01 -5.13098896e-01 -8.70000571e-02 1.13375477e-01
-6.49732828e-01 8.33172083e-01 3.76501709e-01 -8.60888302e-01
-2.61309475e-01 -7.08361685e-01 -6.35821164e-01 -5.37620306e-01
-1.35643288e-01 -6.85169222e-03 1.99635729e-01 3.69101286e-01
3.39082837e-01 3.41719896e-01 -1.48325288e+00 3.52239192e-01
1.23139012e+00 6.30614936e-01 7.16280043e-02 1.13228858e+00
4.46271181e-01 7.14444041e-01 -3.61812770e-01 -2.99051046e-01
8.00528467e-01 -4.83898759e-01 -5.64827561e-01 1.56392545e-01
1.20832062e+00 -8.14364851e-01 -4.13271785e-01 7.42399752e-01
9.89239931e-01 -3.21831584e-01 6.40982389e-01 -6.40742779e-01
-8.61503482e-01 -1.22326076e-01 -6.82961106e-01 4.71341342e-01
-3.56171429e-01 6.70100451e-01 2.82724798e-01 -3.31199199e-01
5.56308985e-01 6.88729584e-01 8.39617968e-01 3.86999279e-01
-9.21776712e-01 -7.67182231e-01 -4.85169113e-01 1.94038674e-01
-1.25340462e+00 -3.63113642e-01 -1.93721160e-01 -9.90441799e-01
6.41036391e-01 2.41552666e-02 1.04283381e+00 3.30834776e-01
4.21715319e-01 -4.94215684e-03 1.70028031e+00 -4.21819776e-01
-1.00663543e-01 9.52283107e-03 8.55623186e-02 8.07503343e-01
6.87475979e-01 4.27526563e-01 4.87183273e-01 -1.42903194e-01
1.12778437e+00 -2.00861186e-01 -3.32622260e-01 1.09329887e-01
-7.41578758e-01 8.46130669e-01 5.26108384e-01 -2.12937683e-01
-3.03559422e-01 -2.68342674e-01 3.20255786e-01 5.22235811e-01
2.71439552e-01 1.92821488e-01 -3.05444717e-01 -2.13437323e-02
-7.73636699e-01 -1.14177130e-01 5.00803769e-01 3.71983588e-01
2.11628720e-01 -3.75521541e-01 -2.53249317e-01 9.54923511e-01
3.81717086e-01 3.38008523e-01 4.94249254e-01 -1.20030808e+00
9.19390246e-02 6.45792723e-01 4.11822200e-01 -4.20846730e-01
-7.01569140e-01 -5.19362807e-01 -7.93881536e-01 9.58865345e-01
9.91896927e-01 -6.60272300e-01 -1.36818075e+00 1.11481082e+00
-1.36414587e-01 -1.65786117e-01 1.26719251e-01 1.28075933e+00
1.01982605e+00 8.80290493e-02 1.19612962e-01 6.33835420e-02
1.46297228e+00 -7.08178461e-01 -1.08457297e-01 1.38866991e-01
1.00639892e+00 -9.57470953e-01 1.09827292e+00 5.71365893e-01
-1.01471162e+00 -4.51985151e-01 -9.48274076e-01 -1.51744008e-01
-3.25630754e-02 7.89993823e-01 5.52620292e-01 6.22369230e-01
-1.54168367e+00 1.27442896e-01 -5.96196473e-01 -7.56400466e-01
7.98833787e-01 3.33737642e-01 -5.47454894e-01 -3.26502115e-01
-6.61616206e-01 9.02594507e-01 -1.36179864e-01 -6.07082099e-02
-4.12608773e-01 -6.00883365e-01 -4.64170605e-01 -5.58571577e-01
-1.51970983e-01 -8.67509127e-01 1.04446292e+00 -1.07221544e+00
-1.06431103e+00 1.45102131e+00 -1.66916505e-01 -5.46018958e-01
6.55797958e-01 5.69593534e-02 -5.79118967e-01 2.71336675e-01
-4.15092669e-02 5.51436067e-01 1.78729296e-01 -8.40851545e-01
-1.05149949e+00 -6.55945122e-01 1.37601614e-01 -1.64051816e-01
3.48632574e-01 2.58571565e-01 -2.85901785e-01 -3.65459442e-01
-2.90828764e-01 -9.89945769e-01 -5.53251728e-02 2.73193806e-01
-2.90655166e-01 -2.21213815e-03 -6.90316409e-02 -8.59281421e-01
1.07655883e+00 -2.07659841e+00 -8.23487878e-01 -1.32972151e-01
7.13592112e-01 8.33865225e-01 -3.08196664e-01 3.78809795e-02
-3.85268599e-01 4.82319981e-01 1.20486982e-01 4.06986400e-02
-6.94490433e-01 -3.91451269e-01 3.25837225e-01 8.87170017e-01
2.55666465e-01 6.06223166e-01 -6.65015459e-01 -2.32439950e-01
3.86015326e-01 3.84183973e-01 -3.55394959e-01 -1.73224539e-01
4.77754623e-01 1.54742897e-01 -1.20404763e-02 5.72612047e-01
7.36326814e-01 -4.63659048e-01 -8.87494981e-02 -4.59400415e-01
-6.83732033e-01 -5.92587367e-02 -7.98936963e-01 1.02064812e+00
-5.79500124e-02 1.35678291e+00 -3.86739820e-01 -4.98885840e-01
7.06098080e-01 2.59525537e-01 2.03084394e-01 -8.18189323e-01
6.92888498e-02 3.60350192e-01 8.33127141e-01 -8.19145322e-01
-2.63396859e-01 -1.64708942e-01 8.82290423e-01 1.13249443e-01
-2.47117907e-01 3.62512559e-01 3.03256124e-01 -2.52969533e-01
9.78396952e-01 -9.84580591e-02 3.04545224e-01 -1.01445079e-01
8.66731629e-02 2.86154807e-01 3.06887090e-01 9.07477379e-01
-5.23413539e-01 1.16649151e+00 9.37956572e-01 -6.47720158e-01
-1.23407888e+00 -1.05834365e+00 -9.61181164e-01 7.08971918e-02
-3.37123215e-01 -1.71373636e-01 -4.19903427e-01 -4.21837568e-01
1.17036887e-01 2.53253102e-01 -9.28187847e-01 2.07517222e-01
4.06551249e-02 -1.05660450e+00 6.00718319e-01 3.28178197e-01
4.81028706e-01 -4.53099430e-01 -6.66404724e-01 -6.76516071e-02
3.17825526e-01 -5.64010084e-01 -2.76627362e-01 -6.27196550e-01
-8.20943117e-01 -1.80874062e+00 -1.23359394e+00 -7.76288509e-01
6.53512895e-01 2.29017511e-01 9.03738379e-01 4.42403890e-02
-6.81468725e-01 1.53829783e-01 -1.47892982e-01 -7.49845624e-01
-4.38493878e-01 -4.55243766e-01 -1.75557703e-01 -8.78745615e-02
7.44873047e-01 -1.71678931e-01 -1.36862612e+00 2.52243817e-01
-6.21250093e-01 -9.45158601e-02 1.19529033e+00 5.37145674e-01
4.72319990e-01 -3.71543378e-01 5.34009516e-01 -6.20200872e-01
2.93608040e-01 -3.15089315e-01 -9.74178433e-01 1.15866646e-01
-9.21049356e-01 -4.89557832e-01 5.47918305e-02 -2.94365883e-01
-4.90832895e-01 -4.13750678e-01 3.87863994e-01 -1.49375409e-01
-4.58626211e-01 4.26478416e-01 8.35863233e-01 -2.69385219e-01
9.46954012e-01 -3.73789132e-01 5.99434137e-01 -5.17429054e-01
-3.31510305e-01 1.14861143e+00 3.22461516e-01 1.22849405e-01
2.25359693e-01 5.08551061e-01 2.25674268e-02 -6.63462520e-01
-6.94982052e-01 -6.31835103e-01 -1.38120398e-01 -8.01132992e-02
9.41915452e-01 -1.29843795e+00 -8.73006701e-01 9.21433091e-01
-7.71824837e-01 -6.53325081e-01 -6.11242838e-02 1.26430166e+00
-2.00342402e-01 2.12002560e-01 -4.37642515e-01 -4.89660174e-01
-3.61660928e-01 -1.25046635e+00 5.23562074e-01 6.20769382e-01
3.93051989e-02 -7.89375246e-01 3.41197610e-01 4.36824828e-01
4.22755629e-01 5.68829596e-01 9.72293079e-01 -3.08657169e-01
-3.68991166e-01 -3.55547577e-01 -9.72910702e-01 7.46116042e-01
5.56099415e-01 5.22002280e-01 -7.72436917e-01 -2.62583464e-01
-4.86545205e-01 -2.54785389e-01 9.67678249e-01 1.15812087e+00
9.95955110e-01 -1.68888047e-01 2.69676931e-02 6.71415150e-01
1.70551765e+00 5.03343821e-01 1.34240413e+00 6.32276535e-01
2.59713799e-01 5.48140645e-01 1.88150615e-01 2.02545553e-01
5.38719535e-01 4.31977779e-01 1.76048383e-01 -7.58241355e-01
-6.73218548e-01 2.54837513e-01 4.33295816e-02 -2.54545957e-01
-6.87691629e-01 -2.60167085e-02 -1.05647731e+00 7.22619772e-01
-1.45264721e+00 -6.43115342e-01 -7.67790079e-01 2.62009978e+00
7.45359659e-01 -3.24382670e-02 5.30545533e-01 -5.54695487e-01
7.50220954e-01 -5.70342243e-01 -5.86084545e-01 -3.99333835e-01
-2.77701825e-01 2.17252627e-01 8.64675343e-01 4.26245987e-01
-8.51566851e-01 4.17442918e-01 5.87145853e+00 1.09584518e-02
-1.20368087e+00 -4.32360321e-02 7.41530955e-01 -4.57625419e-01
4.53996330e-01 1.14428662e-01 -4.72349316e-01 5.73158145e-01
1.07384205e+00 -1.94865882e-01 4.04730402e-02 9.72863734e-02
8.21450770e-01 -6.12558484e-01 -6.25778139e-01 8.87383878e-01
-1.21532321e-01 -1.36510718e+00 -7.46636987e-02 4.38297838e-01
5.74647903e-01 5.57722449e-01 2.09992155e-01 -1.23653702e-01
2.80361231e-02 -1.42288637e+00 -5.84253594e-02 8.88707459e-01
1.43765318e+00 -2.83444256e-01 1.21814644e+00 -6.12099588e-01
-2.48488367e-01 1.92879383e-02 -3.19710433e-01 -2.33643740e-01
-1.76932365e-01 5.43041527e-01 -1.02706695e+00 -2.92504802e-02
9.65543807e-01 7.05697358e-01 -9.60526168e-01 2.05203509e+00
2.23750576e-01 8.85666966e-01 -3.23858298e-03 2.04703063e-01
2.50134081e-01 -3.61290246e-01 4.48653936e-01 8.15816522e-01
3.00737441e-01 1.60532549e-01 -5.99401772e-01 6.04853272e-01
1.45081043e-01 2.11914808e-01 -3.93893212e-01 -7.07492372e-03
7.32391700e-02 9.32065129e-01 -1.74051389e-01 -1.45814002e-01
-1.17011142e+00 2.83383161e-01 -1.18253529e-01 6.53501809e-01
-4.38229591e-01 -6.74068391e-01 6.16032183e-01 7.61917949e-01
-2.12438211e-01 1.40135363e-01 -3.17704320e-01 -7.97005773e-01
4.09095064e-02 -8.38584721e-01 5.91246724e-01 -1.11268520e+00
-1.21651888e+00 2.77756095e-01 -1.69982463e-01 -1.34997189e+00
3.14497948e-01 -9.31388617e-01 -3.52414101e-01 1.58292186e+00
-1.52830648e+00 -9.11283135e-01 -6.02639318e-01 3.88752222e-01
7.17696473e-02 -5.03549278e-01 6.18709445e-01 1.57821789e-01
-6.27345443e-01 5.95837533e-01 2.95047790e-01 4.14216310e-01
1.43230641e+00 -1.25775850e+00 1.30776064e-02 7.14670539e-01
-6.97615981e-01 7.52749562e-01 2.19245970e-01 -6.35275483e-01
-4.95728821e-01 -1.07666016e+00 5.12271285e-01 -3.59517455e-01
5.48409998e-01 6.00541353e-01 -6.44148707e-01 6.87590599e-01
3.01067024e-01 -1.03629060e-01 9.73996282e-01 1.03654467e-01
-1.00039370e-01 -2.47876555e-01 -1.22872365e+00 6.41888857e-01
4.95754868e-01 -4.51540500e-02 -4.32287157e-01 3.13760936e-01
1.50060341e-01 -5.25755703e-01 -1.10952520e+00 4.03014809e-01
1.01044083e+00 -1.31410348e+00 4.35955793e-01 -8.00009668e-01
5.67810357e-01 -4.26393092e-01 9.72564146e-02 -9.15957332e-01
-2.62649804e-01 -4.97055888e-01 4.41190749e-01 7.52152503e-01
6.05161965e-01 -1.39181805e+00 4.34795290e-01 4.97170240e-01
-2.20001295e-01 -9.15279686e-01 -6.19388163e-01 -3.28356594e-01
2.68090904e-01 1.05477534e-01 7.49259666e-02 6.88990057e-01
-7.00450659e-01 -2.41028398e-01 1.55168384e-01 3.94374728e-01
5.43531001e-01 -2.24325638e-02 7.34409451e-01 -1.12242460e+00
-1.40773922e-01 -3.97721350e-01 -7.39973962e-01 -1.91882581e-01
-6.26984119e-01 -6.61223888e-01 -7.12964952e-01 -2.07932711e+00
4.63039577e-01 -5.31985521e-01 -3.78069311e-01 7.30943501e-01
-3.42870451e-04 4.70516890e-01 -1.54823110e-01 4.01549369e-01
2.43195146e-01 -2.93326497e-01 1.41818392e+00 1.26131028e-01
-5.77869177e-01 1.46599188e-01 -1.22638619e+00 5.59600651e-01
1.13780046e+00 -1.79039270e-01 -4.09338623e-01 -4.55285668e-01
8.45824555e-02 -1.91936448e-01 9.44317102e-01 -1.10912359e+00
-4.76674289e-02 2.16383208e-02 6.17982864e-01 -1.29247203e-01
-1.89083531e-01 -1.33760795e-01 9.97655094e-03 5.77111483e-01
4.42077704e-02 -1.70326099e-01 5.36571741e-01 1.70433745e-01
-2.25394126e-02 9.66494530e-02 9.91752684e-01 -1.45366162e-01
-5.40291548e-01 2.72924066e-01 -4.21058238e-01 1.77828848e-01
7.55542517e-01 -5.84525883e-01 -1.05188406e+00 -1.48496240e-01
-7.73308158e-01 1.59694001e-01 9.02808011e-01 2.48013943e-01
4.14360434e-01 -7.34612763e-01 -1.31723082e+00 -1.07342973e-01
4.16279733e-01 -4.80027311e-02 4.32625324e-01 1.45866358e+00
-9.77318823e-01 3.84300381e-01 -5.07479727e-01 -4.27748948e-01
-1.21232426e+00 1.13167077e-01 9.36707675e-01 4.88890916e-01
-9.06286538e-01 2.97437012e-01 1.90024227e-01 2.17742354e-01
1.71444137e-02 -5.88875294e-01 -1.24459825e-01 1.45599141e-03
7.87039578e-01 5.37554562e-01 6.09031282e-02 -9.16267559e-02
-7.26645142e-02 9.23322201e-01 -3.07934165e-01 1.74274862e-01
1.08382428e+00 -2.39895180e-01 -1.49058059e-01 2.17326820e-01
1.07661831e+00 1.07181683e-01 -1.03992629e+00 2.05514193e-01
-6.39171064e-01 -5.15260339e-01 4.09198701e-01 -1.34455574e+00
-8.91697943e-01 5.43075085e-01 1.61407053e+00 -2.23067068e-02
1.31358814e+00 -8.56342241e-02 3.69925618e-01 -4.00138982e-02
-4.89880741e-02 -6.34410203e-01 -5.66235602e-01 3.09803393e-02
8.71350765e-01 -1.37809908e+00 -8.26064721e-02 -3.53265874e-04
-5.61486244e-01 1.05721855e+00 6.03118956e-01 -1.55433178e-01
4.48175967e-01 -1.72930196e-01 6.29314721e-01 -1.99205950e-01
-3.51941437e-01 -3.78814161e-01 4.86413628e-01 9.34618652e-01
6.86837137e-01 1.45646155e-01 -8.74197841e-01 3.67622226e-01
-1.90753881e-02 7.66715586e-01 1.20801091e+00 4.82175469e-01
-2.03641295e-01 -8.89504790e-01 -2.34683193e-02 1.24574375e+00
-6.50839686e-01 -1.10332757e-01 -3.68781745e-01 1.07984209e+00
2.14522392e-01 1.07779145e+00 3.28327358e-01 5.99399060e-02
3.74375284e-01 -2.29250163e-01 5.12235105e-01 -5.97916842e-01
-1.62559189e-02 -1.11080058e-01 4.21632826e-01 -3.61909240e-01
-6.20012283e-01 -5.10459661e-01 -7.92693913e-01 -2.31112450e-01
1.24183141e-01 -1.65030167e-01 3.48835796e-01 5.00360548e-01
6.83892667e-01 2.45999604e-01 3.67119551e-01 -1.91768724e-02
-9.27200690e-02 -1.04945850e+00 -7.97368169e-01 -6.23571947e-02
9.14090991e-01 -5.17084539e-01 -3.83931905e-01 1.18484758e-01] | [15.823871612548828, -3.998631238937378] |
c8acad60-eb49-4528-91d3-96a42a092edd | yolopose-transformer-based-multi-object-6d | 2205.02536 | null | https://arxiv.org/abs/2205.02536v1 | https://arxiv.org/pdf/2205.02536v1.pdf | YOLOPose: Transformer-based Multi-Object 6D Pose Estimation using Keypoint Regression | 6D object pose estimation is a crucial prerequisite for autonomous robot manipulation applications. The state-of-the-art models for pose estimation are convolutional neural network (CNN)-based. Lately, Transformers, an architecture originally proposed for natural language processing, is achieving state-of-the-art results in many computer vision tasks as well. Equipped with the multi-head self-attention mechanism, Transformers enable simple single-stage end-to-end architectures for learning object detection and 6D object pose estimation jointly. In this work, we propose YOLOPose (short form for You Only Look Once Pose estimation), a Transformer-based multi-object 6D pose estimation method based on keypoint regression. In contrast to the standard heatmaps for predicting keypoints in an image, we directly regress the keypoints. Additionally, we employ a learnable orientation estimation module to predict the orientation from the keypoints. Along with a separate translation estimation module, our model is end-to-end differentiable. Our method is suitable for real-time applications and achieves results comparable to state-of-the-art methods. | ['Sven Behnke', 'Arul Selvam Periyasamy', 'Arash Amini'] | 2022-05-05 | null | null | null | null | ['6d-pose-estimation-1', '6d-pose-estimation', 'robot-manipulation'] | ['computer-vision', 'computer-vision', 'robots'] | [-2.03709245e-01 1.00075901e-01 -2.12996706e-01 -4.79588389e-01
-7.35560417e-01 -1.74839109e-01 5.84549963e-01 -8.03663358e-02
-5.28819621e-01 -2.21620593e-02 -2.33399123e-01 1.34807080e-02
8.38488862e-02 -4.67555135e-01 -1.24210453e+00 -4.02415693e-01
2.76760273e-02 1.00255823e+00 3.34142178e-01 -3.02013069e-01
4.58420873e-01 8.10258269e-01 -1.58917773e+00 1.02063067e-01
4.33704376e-01 1.34160066e+00 6.39055908e-01 7.25620091e-01
9.58154798e-02 6.67135179e-01 -2.80928493e-01 -2.99883455e-01
2.45616183e-01 2.69441158e-01 -6.95533633e-01 -1.08424658e-02
1.01768482e+00 -6.91250443e-01 -4.89616573e-01 9.08931971e-01
4.10474956e-01 -6.90856203e-02 7.89109707e-01 -1.35216224e+00
-4.71332729e-01 3.28162670e-01 -5.59635222e-01 -1.60979509e-01
3.55984330e-01 2.25572780e-01 9.27932501e-01 -1.36992681e+00
5.19014239e-01 1.53293717e+00 5.64989328e-01 5.03684759e-01
-9.26262796e-01 -4.14941341e-01 3.62934947e-01 4.34033364e-01
-1.09624434e+00 -2.47834295e-01 8.57745528e-01 -4.79946315e-01
1.31761312e+00 -9.80073214e-02 7.19536901e-01 9.34115469e-01
6.25192046e-01 1.25237453e+00 5.99152923e-01 -2.81272888e-01
-2.14685485e-01 -2.27193639e-01 -3.30243200e-01 1.08616006e+00
-1.87365770e-01 -1.44808320e-02 -5.74031293e-01 3.79776746e-01
1.18429697e+00 3.27049255e-01 1.58928514e-01 -1.18199956e+00
-1.74036634e+00 5.48531055e-01 1.18259823e+00 -2.76095450e-01
-5.72794020e-01 7.42989659e-01 2.23501399e-01 2.05617156e-02
4.71113414e-01 5.46874404e-01 -7.10119545e-01 -1.00593336e-01
-4.08088475e-01 5.60959518e-01 5.78466296e-01 1.40753496e+00
6.35158598e-01 -1.76855996e-01 -1.93070054e-01 5.03398001e-01
7.44012654e-01 5.77511966e-01 2.99267083e-01 -8.86791587e-01
5.62954605e-01 5.95598936e-01 2.49831796e-01 -7.85494506e-01
-6.53196216e-01 -5.46418548e-01 -5.35247624e-01 3.99571240e-01
4.47455436e-01 4.22579318e-01 -1.15596139e+00 1.27804720e+00
4.75219995e-01 -5.03690168e-02 -3.48402500e-01 1.11380506e+00
8.12290847e-01 5.06619096e-01 -1.08417191e-01 4.35424536e-01
1.45461929e+00 -1.38688099e+00 -4.34564590e-01 -5.40797532e-01
4.01660353e-01 -6.41643465e-01 1.11940515e+00 4.33057725e-01
-1.05504572e+00 -6.54925942e-01 -9.99383867e-01 -6.61444604e-01
-5.00204682e-01 4.92022276e-01 7.02189386e-01 5.81039935e-02
-7.14328825e-01 4.54916447e-01 -1.23991847e+00 -2.70666778e-01
4.87181693e-01 5.34456372e-01 -5.29189825e-01 5.25492691e-02
-5.46139419e-01 1.63086474e+00 3.53337079e-01 4.21795547e-01
-1.16931844e+00 -5.75520992e-01 -1.11235070e+00 -4.87174913e-02
6.76715910e-01 -9.16364908e-01 1.66731822e+00 -1.58223242e-01
-1.88035262e+00 1.07713342e+00 -8.40316713e-02 -4.77141798e-01
7.79368699e-01 -9.56647217e-01 4.40906346e-01 1.18555002e-01
6.64303079e-02 9.89174962e-01 1.29359198e+00 -1.14551497e+00
-5.43449759e-01 -6.33160412e-01 4.09706868e-02 3.16377431e-01
7.01310933e-02 -3.53569165e-02 -6.62385166e-01 -4.01528418e-01
4.94929194e-01 -1.02872694e+00 -3.29244524e-01 7.92048812e-01
-4.10543203e-01 -6.24592364e-01 1.03911304e+00 -4.34699953e-01
3.53325456e-01 -1.88045073e+00 6.36312664e-01 -2.27047801e-01
3.79079878e-01 -1.27172070e-02 -1.73154131e-01 5.46041653e-02
1.19594254e-01 -4.18324500e-01 1.94853753e-01 -7.04191387e-01
3.77623588e-01 -1.57248348e-01 -1.26582146e-01 7.34568298e-01
5.38173318e-01 1.34487653e+00 -8.47759724e-01 -1.46708906e-01
8.02423954e-01 4.92824346e-01 -6.41755462e-01 5.54438591e-01
-6.18051410e-01 3.22520554e-01 -3.02352071e-01 7.82151043e-01
5.18431485e-01 -1.49575382e-01 -3.35426331e-01 -6.29794002e-01
-1.17755152e-01 2.74558634e-01 -7.87640631e-01 2.25936317e+00
-6.68211401e-01 4.49318320e-01 3.60219888e-02 -8.82761955e-01
1.01542795e+00 7.65589550e-02 3.76016527e-01 -2.75174588e-01
3.42725039e-01 2.78396875e-01 -2.40106389e-01 -5.27716637e-01
5.98530293e-01 2.49730632e-01 -2.77723283e-01 -7.84752965e-02
3.60903352e-01 -7.94501901e-01 -1.36729300e-01 -1.55040890e-01
6.62447333e-01 8.06700587e-01 3.05198073e-01 6.43300489e-02
2.54955858e-01 1.19241904e-02 4.88056019e-02 5.02600133e-01
-2.15056956e-01 8.36744249e-01 2.59981066e-01 -6.59908235e-01
-1.24114406e+00 -1.14211953e+00 1.57709777e-01 1.07418609e+00
3.12515587e-01 -3.84689987e-01 -4.42637652e-01 -6.59448326e-01
3.98457855e-01 3.74857128e-01 -5.62705398e-01 -9.90855768e-02
-8.30790877e-01 -1.05364203e-01 1.65190622e-02 8.40600014e-01
4.99392658e-01 -1.06985259e+00 -1.01644099e+00 2.58165091e-01
1.40689071e-02 -1.24598050e+00 -3.32263112e-01 3.91074270e-01
-7.78605819e-01 -9.83579278e-01 -8.99098873e-01 -9.20133352e-01
6.69412434e-01 2.45092615e-01 1.00262117e+00 -4.86080199e-01
-4.01202083e-01 2.87223488e-01 -2.70940274e-01 -6.93796933e-01
4.80055436e-02 2.19928935e-01 1.49425507e-01 -2.88699627e-01
2.44856030e-01 -2.35805348e-01 -6.39405906e-01 2.59376198e-01
-3.11722070e-01 1.26131862e-01 1.08216679e+00 6.91872060e-01
6.78463042e-01 -5.56758642e-01 8.51532817e-02 -2.82100946e-01
4.85735536e-02 1.24426158e-02 -7.30904162e-01 7.40066990e-02
-1.92719638e-01 1.63935170e-01 2.43030742e-01 -4.78835672e-01
-6.68161869e-01 6.06607914e-01 -2.53574312e-01 -8.51572216e-01
-3.16883296e-01 4.18352365e-01 -1.13995492e-01 -3.13126534e-01
3.65736306e-01 6.04173075e-03 2.16091990e-01 -7.39208639e-01
5.22019327e-01 4.91654903e-01 7.14658618e-01 -3.13751519e-01
8.27675343e-01 2.73068756e-01 1.86704069e-01 -7.70434320e-01
-9.11112130e-01 -6.30409479e-01 -1.28995979e+00 -3.76870811e-01
1.00001931e+00 -1.14584374e+00 -1.20864522e+00 8.00934494e-01
-1.43378019e+00 -4.29106534e-01 6.84660748e-02 6.18695021e-01
-1.07147682e+00 -3.10344826e-02 -4.82343584e-01 -5.90220690e-01
-4.08650368e-01 -1.29493499e+00 1.89982522e+00 9.28476602e-02
1.73504148e-02 -3.92840147e-01 -4.69350845e-01 3.46717924e-01
3.27273518e-01 1.76679090e-01 6.21113181e-01 -2.08489195e-01
-9.94227350e-01 -5.57577312e-01 -3.58117431e-01 -9.14419517e-02
-2.42324144e-01 -2.51923323e-01 -7.31668830e-01 -3.50240141e-01
-1.69597432e-01 -5.96093774e-01 8.82661521e-01 4.80580986e-01
1.45446897e+00 -7.97310472e-02 -4.76934731e-01 7.27811933e-01
1.19167733e+00 -2.53743738e-01 4.51606005e-01 5.32722592e-01
1.13312435e+00 3.47872108e-01 9.41427171e-01 3.15558881e-01
7.02343881e-01 1.03165257e+00 1.12748480e+00 9.19438228e-02
1.57373659e-02 -4.10971940e-01 3.66479903e-01 5.38634062e-01
3.06606218e-02 -2.78158374e-02 -9.00259197e-01 3.98299247e-01
-1.96261942e+00 -4.95471478e-01 6.49030283e-02 1.98609495e+00
5.10141373e-01 3.97198230e-01 9.09418762e-02 -1.66317850e-01
2.10235864e-01 1.40596390e-01 -7.79001653e-01 -5.46862036e-02
4.35419858e-01 1.60878301e-02 5.79500973e-01 2.84221768e-01
-1.47433650e+00 1.28810203e+00 5.69359589e+00 5.26436865e-01
-1.30211294e+00 -1.84003979e-01 8.04229900e-02 -3.50371636e-02
3.12372774e-01 -3.51990551e-01 -9.37630653e-01 -1.56449333e-01
3.73409569e-01 3.78224492e-01 1.12664320e-01 1.33480942e+00
-7.35517815e-02 -5.02671041e-02 -1.49455869e+00 1.16577792e+00
2.96686202e-01 -1.10789442e+00 -1.04454458e-01 -1.62290871e-01
2.71487117e-01 4.32202518e-01 1.50583401e-01 3.84660214e-01
4.29209042e-03 -9.07771885e-01 1.32665002e+00 4.85507041e-01
7.55003810e-01 -6.77802503e-01 6.97108209e-01 6.55661523e-01
-1.24038994e+00 -2.15499595e-01 -4.60975856e-01 -7.44252950e-02
2.83953816e-01 3.29472512e-01 -1.08821332e+00 4.42779839e-01
9.02868867e-01 9.15110290e-01 -6.35373414e-01 1.28821743e+00
-5.26147723e-01 -1.64908543e-01 -4.56534266e-01 -3.47581357e-01
2.43985459e-01 2.30825976e-01 6.45987272e-01 9.57651079e-01
2.71677762e-01 -3.75858635e-01 4.11623001e-01 9.22311187e-01
-8.44392776e-02 -2.60506779e-01 -5.38637578e-01 2.36210022e-02
2.10257366e-01 1.20516026e+00 -6.01542294e-01 -2.29374707e-01
-2.83033073e-01 1.19171345e+00 6.20813310e-01 1.12951621e-02
-6.80254757e-01 -6.25138044e-01 5.96548498e-01 1.31975129e-01
6.44705534e-01 -8.49606812e-01 -2.09784016e-01 -1.12846482e+00
3.26712698e-01 -5.66538155e-01 -2.41660431e-01 -1.06706357e+00
-1.10894489e+00 3.83571535e-01 1.74466655e-01 -1.47286749e+00
-3.33398521e-01 -1.30326331e+00 -2.81793118e-01 6.19470477e-01
-1.60519171e+00 -1.77937365e+00 -4.71868157e-01 2.18034089e-01
8.01924229e-01 6.31472245e-02 8.71616244e-01 -1.17195427e-01
-2.00250328e-01 4.01676059e-01 -3.06102484e-01 2.71315098e-01
7.10343361e-01 -1.37950885e+00 8.29063535e-01 3.93059433e-01
8.41041580e-02 5.85706592e-01 6.18690073e-01 -3.97974461e-01
-2.07877946e+00 -9.25757647e-01 7.93265641e-01 -7.75000453e-01
6.05343819e-01 -7.99188435e-01 -4.52494591e-01 9.03123081e-01
-8.31013545e-02 2.82482952e-01 -2.17246577e-01 5.83057664e-02
-4.00729865e-01 -8.83252993e-02 -8.62256348e-01 6.59466803e-01
1.04759943e+00 -4.82515574e-01 -7.41288543e-01 4.23995078e-01
7.58158684e-01 -1.17686367e+00 -8.51118386e-01 6.82202160e-01
8.95929039e-01 -6.01666451e-01 1.19781184e+00 -4.05107290e-01
6.07969463e-01 -4.76261079e-01 -1.65997110e-02 -1.24612272e+00
-4.64889795e-01 -3.60016167e-01 -4.98342901e-01 4.28865314e-01
2.84172624e-01 -2.25813642e-01 1.05793500e+00 3.79769355e-01
-5.08674920e-01 -9.05796111e-01 -1.00490057e+00 -5.99404454e-01
-1.67741671e-01 -3.70598733e-01 3.15401524e-01 2.32157990e-01
-7.81191364e-02 5.16290545e-01 -3.75943154e-01 2.40363955e-01
6.81207895e-01 3.72660428e-01 1.23174322e+00 -1.23774040e+00
-3.46465781e-02 -5.31439066e-01 -7.88756549e-01 -1.81535602e+00
1.59305930e-01 -6.71251953e-01 5.63030601e-01 -1.56798565e+00
-3.82571556e-02 -1.27256706e-01 1.69939607e-01 5.63769758e-01
3.87397874e-03 2.49274611e-01 5.10760009e-01 -3.35698202e-02
-6.68625772e-01 9.00918007e-01 1.52117467e+00 -3.16059202e-01
-8.93273503e-02 1.77043468e-01 -4.88333739e-02 7.60135710e-01
5.20092785e-01 -2.23921552e-01 6.68856055e-02 -7.46615231e-01
7.71126300e-02 -1.87669378e-02 7.43323505e-01 -9.79263484e-01
4.28256124e-01 1.69527773e-02 5.92955947e-01 -1.26979780e+00
8.02696764e-01 -1.01560235e+00 -3.91065896e-01 6.05174541e-01
-2.74998635e-01 1.30982429e-01 1.09347776e-01 4.89354849e-01
-5.04208654e-02 -3.83147784e-02 6.63162291e-01 -4.68871623e-01
-1.05862427e+00 7.77599216e-01 -1.97339877e-01 -5.49860954e-01
1.11188781e+00 -1.52412117e-01 -1.23136587e-01 -1.30188525e-01
-6.45630658e-01 3.81642044e-01 2.53955543e-01 8.45026493e-01
9.68807817e-01 -1.27248502e+00 -7.07769036e-01 3.08344275e-01
4.35189366e-01 7.50906587e-01 -1.15571924e-01 8.12561393e-01
-7.72290051e-01 5.65483332e-01 -2.87869155e-01 -1.10814822e+00
-1.14878762e+00 5.25142550e-01 3.12168330e-01 4.88757007e-02
-6.78014994e-01 1.19473588e+00 5.71399406e-02 -9.38359559e-01
6.24388635e-01 -7.85659909e-01 2.74004713e-02 -1.26806125e-01
2.89591193e-01 9.70311910e-02 1.01398274e-01 -5.56847155e-01
-4.41535145e-01 9.14519727e-01 -3.10117692e-01 3.55635956e-02
1.61981940e+00 3.09733041e-02 -1.52962789e-01 5.42967975e-01
1.22591162e+00 -5.85744202e-01 -1.60337746e+00 -2.20136225e-01
-1.99804574e-01 -5.43675482e-01 2.19435636e-02 -6.28084540e-01
-7.74131179e-01 1.19635820e+00 4.93290812e-01 -3.54589432e-01
5.22783220e-01 1.60517126e-01 5.66355765e-01 8.21490109e-01
6.43897891e-01 -9.36686337e-01 5.66196799e-01 9.19126570e-01
1.44517267e+00 -1.66171837e+00 1.23916350e-01 -5.25534689e-01
-4.31789279e-01 1.28241396e+00 1.03859735e+00 -3.14446777e-01
6.80141926e-01 1.64058983e-01 1.07596396e-02 -3.30401301e-01
-6.40212774e-01 -1.85180426e-01 5.78554332e-01 6.86579525e-01
2.63883501e-01 4.09714924e-03 5.05836248e-01 1.68076366e-01
-3.08368027e-01 -2.14097023e-01 -6.47659274e-03 1.01128042e+00
-6.38124168e-01 -6.28114641e-01 -2.35455275e-01 3.47599685e-01
-6.97204992e-02 1.57864302e-01 -3.28320563e-01 6.97316229e-01
-2.20499068e-01 3.94428253e-01 2.36054197e-01 -3.48026812e-01
6.36069775e-01 -1.24267310e-01 1.02661884e+00 -7.45914280e-01
-3.76599938e-01 3.66305886e-03 -2.46222302e-01 -8.24427962e-01
-2.88156211e-01 -4.61590290e-01 -1.14151359e+00 5.32649495e-02
-6.01940095e-01 -5.61584234e-01 1.04948938e+00 9.60348427e-01
2.31262714e-01 5.99313021e-01 2.66879171e-01 -1.97851253e+00
-8.85740995e-01 -1.24207461e+00 -2.32837185e-01 1.04127914e-01
6.01330519e-01 -1.01602972e+00 1.35579705e-01 -1.74420908e-01] | [7.4562578201293945, -2.6246466636657715] |
5d310e87-9e5d-4413-9ac1-8f66b6e53c79 | monocular-object-and-plane-slam-in-structured | 1809.03415 | null | https://arxiv.org/abs/1809.03415v2 | https://arxiv.org/pdf/1809.03415v2.pdf | Monocular Object and Plane SLAM in Structured Environments | In this paper, we present a monocular Simultaneous Localization and Mapping (SLAM) algorithm using high-level object and plane landmarks. The built map is denser, more compact and semantic meaningful compared to feature point based SLAM. We first propose a high order graphical model to jointly infer the 3D object and layout planes from single images considering occlusions and semantic constraints. The extracted objects and planes are further optimized with camera poses in a unified SLAM framework. Objects and planes can provide more semantic constraints such as Manhattan plane and object supporting relationships compared to points. Experiments on various public and collected datasets including ICL NUIM and TUM Mono show that our algorithm can improve camera localization accuracy compared to state-of-the-art SLAM especially when there is no loop closure, and also generate dense maps robustly in many structured environments. | ['Shichao Yang', 'Sebastian Scherer'] | 2018-09-10 | null | null | null | null | ['camera-localization'] | ['computer-vision'] | [-2.09353939e-01 -1.67412266e-01 -2.19169036e-01 -6.38135850e-01
-3.59120846e-01 -8.34025204e-01 6.88128173e-01 -9.88684222e-03
-2.77591735e-01 7.15770721e-01 -1.87141038e-02 -4.19905931e-02
-3.27913344e-01 -8.11619401e-01 -1.03172565e+00 -3.25621516e-02
6.77130744e-02 1.12990618e+00 5.65567434e-01 5.88400923e-02
5.43359637e-01 8.45556498e-01 -1.35481107e+00 -5.29836237e-01
8.46517861e-01 7.51624823e-01 7.38269985e-01 5.04092395e-01
-1.07621491e-01 3.78259182e-01 -5.05659953e-02 -1.89803261e-02
4.19538468e-01 2.11301759e-01 -4.74520981e-01 2.73522586e-01
1.01529086e+00 -3.12923640e-01 -5.09110749e-01 1.11813140e+00
1.21504694e-01 6.38953447e-02 2.12234169e-01 -1.46777356e+00
-3.06088030e-01 -1.12634681e-01 -7.84000337e-01 -4.54889953e-01
9.10729349e-01 -2.93523014e-01 7.52273560e-01 -1.20459056e+00
1.00652790e+00 1.26950407e+00 8.93906772e-01 -2.27707043e-01
-9.39336121e-01 -7.26286769e-01 2.40178004e-01 4.45293337e-02
-2.00196362e+00 -3.27920914e-01 7.52909184e-01 -2.24652648e-01
8.90489697e-01 2.52442747e-01 6.30907118e-01 6.66153252e-01
2.53240556e-01 3.64383668e-01 7.91791499e-01 -4.38754946e-01
9.83222350e-02 1.39697179e-01 -4.65494134e-02 1.12522447e+00
8.06689501e-01 -8.29508603e-02 -9.71464097e-01 -1.82297513e-01
1.26945734e+00 3.34357828e-01 -7.45000392e-02 -1.44913399e+00
-1.50185800e+00 7.93510973e-01 7.59327292e-01 -1.53592661e-01
-2.56414302e-02 2.70186722e-01 -5.00256479e-01 -2.29454532e-01
2.09991783e-01 4.67024863e-01 -2.67777711e-01 1.68972462e-01
-8.69608521e-01 2.53892899e-01 6.28714740e-01 2.03778076e+00
1.48455036e+00 -3.15697819e-01 6.93087935e-01 4.46743220e-01
5.62322199e-01 1.11731637e+00 -1.48105755e-01 -1.18391466e+00
5.92851520e-01 8.00133467e-01 4.73971695e-01 -1.61083210e+00
-6.21254325e-01 -2.21486419e-01 -4.58411783e-01 -3.24689113e-02
-1.91940129e-01 2.36886486e-01 -9.48753357e-01 1.08931553e+00
3.46505821e-01 3.48200142e-01 -2.03363612e-01 9.40628827e-01
7.14807451e-01 3.51283461e-01 -6.79313719e-01 3.50698948e-01
1.12432313e+00 -1.05419159e+00 -7.45809495e-01 -8.31846893e-01
5.95083475e-01 -9.01282907e-01 3.82245630e-01 1.36378929e-01
-5.78194141e-01 -3.97050112e-01 -1.31761312e+00 -5.26381314e-01
-4.54803199e-01 2.10747585e-01 1.03721583e+00 2.73889214e-01
-1.22129178e+00 9.24966764e-03 -9.30924177e-01 -7.96234488e-01
1.37081414e-01 4.19888705e-01 -8.59520674e-01 -4.48721230e-01
-6.23841286e-01 1.18427086e+00 5.87050855e-01 9.44951177e-02
-4.85072970e-01 -5.20561159e-01 -1.58843923e+00 -4.98627692e-01
4.64890510e-01 -9.58968401e-01 7.13610113e-01 2.09383797e-02
-1.14051008e+00 9.76617873e-01 -7.24765956e-01 -3.39706063e-01
4.70953822e-01 -6.16854012e-01 -9.13640708e-02 9.16687623e-02
4.98236895e-01 9.94699895e-01 5.00086062e-02 -1.61598563e+00
-8.80184591e-01 -7.72275448e-01 1.66390203e-02 6.68318748e-01
5.77165663e-01 -3.75626057e-01 -9.26901519e-01 2.91848838e-01
1.44890475e+00 -1.13208377e+00 -3.77113432e-01 2.94951469e-01
-4.82134938e-01 2.63466239e-01 1.15554464e+00 -4.70448524e-01
6.94848776e-01 -1.74728549e+00 1.01978704e-01 5.70424676e-01
1.63866818e-01 -5.22993326e-01 1.25986442e-01 3.23319405e-01
5.02366960e-01 -1.84730962e-01 1.30815551e-01 -6.69963777e-01
8.34975764e-02 5.90800285e-01 -1.69159293e-01 1.01103103e+00
-4.15683389e-01 9.85562682e-01 -1.01164091e+00 -4.34132785e-01
8.60130131e-01 4.85233575e-01 -4.41138297e-01 -1.16716065e-02
7.22623393e-02 3.46172363e-01 -3.92030746e-01 1.00823331e+00
1.24115872e+00 -2.60783844e-02 -4.83411960e-02 -5.87614737e-02
-4.17363822e-01 1.91856027e-01 -1.52395916e+00 2.58682704e+00
-4.30637181e-01 8.05383265e-01 -1.06616579e-02 -4.23034951e-02
1.28922713e+00 -3.16493720e-01 1.53122276e-01 -5.49984574e-01
-1.48114443e-01 2.60609508e-01 -7.24286795e-01 1.81998104e-01
9.63577807e-01 5.50460100e-01 -8.97200853e-02 -2.39629596e-01
1.15592368e-02 -7.72445083e-01 -2.85892725e-01 3.01449776e-01
7.42267370e-01 6.61560416e-01 6.21617794e-01 -4.90586311e-01
3.52264524e-01 4.59843040e-01 6.00067198e-01 8.23338091e-01
3.32484215e-01 6.80753231e-01 -9.24558267e-02 -5.18498242e-01
-1.02905560e+00 -1.33442914e+00 -2.17536986e-01 2.01174363e-01
1.22871828e+00 -6.10592604e-01 -1.00602552e-01 -2.54337430e-01
4.38662767e-01 3.98618042e-01 -2.21312702e-01 3.51232260e-01
-4.51517224e-01 -1.59818590e-01 -1.36139765e-02 4.09715354e-01
7.76349545e-01 -2.20675409e-01 -6.16859615e-01 -1.09421220e-02
-2.14093059e-01 -1.57854378e+00 -2.73647338e-01 7.59523585e-02
-7.01681256e-01 -1.15556049e+00 -7.82113001e-02 -7.31659651e-01
1.14594650e+00 9.34428990e-01 9.00666535e-01 -1.84771851e-01
-2.29778424e-01 3.84678066e-01 -2.58973628e-01 -4.06504333e-01
3.89871031e-01 -9.51116234e-02 5.03687739e-01 -3.34549785e-01
4.39864218e-01 -5.46950102e-01 -3.74678314e-01 7.26732373e-01
-1.36748031e-01 5.60434401e-01 3.78299892e-01 3.16623688e-01
1.01445127e+00 -9.20339599e-02 -5.20095348e-01 -6.22557342e-01
-3.62366349e-01 -1.70411050e-01 -1.37903953e+00 6.16244227e-02
-6.35351419e-01 -3.40792239e-02 -1.50748745e-01 2.00259149e-01
-7.66363978e-01 7.11777210e-01 5.96842945e-01 -5.57228267e-01
-1.18125670e-01 2.24204108e-01 -3.95330548e-01 -7.86627650e-01
3.68763953e-01 1.41923711e-01 -2.59046674e-01 -5.84916472e-01
4.54931557e-01 3.67638528e-01 7.65965104e-01 -5.35412788e-01
1.34340012e+00 1.03320396e+00 5.73707163e-01 -7.50798762e-01
-7.92487919e-01 -9.80926514e-01 -1.45158303e+00 -2.15878841e-02
7.57704258e-01 -1.23643625e+00 -4.34260309e-01 1.21122025e-01
-1.39029992e+00 2.18767360e-01 3.16171378e-01 8.12009215e-01
-6.64511859e-01 3.49459678e-01 -5.58310337e-02 -8.32120657e-01
1.65689856e-01 -1.22778201e+00 1.54576516e+00 2.11858213e-01
-6.57632947e-02 -1.04945707e+00 1.53036028e-01 1.39139622e-01
-1.82535797e-01 5.62771082e-01 1.30845219e-01 6.76442608e-02
-1.56280518e+00 -2.42286503e-01 -3.87273997e-01 -5.84343731e-01
1.43183976e-01 -2.91739166e-01 -6.87096953e-01 -3.35960656e-01
-3.43311191e-01 3.05024475e-01 4.65596080e-01 2.37130031e-01
4.42517310e-01 5.64395450e-02 -9.78529990e-01 1.36545062e+00
1.81140101e+00 1.28181353e-01 5.32790542e-01 8.04081678e-01
1.12642670e+00 3.92209351e-01 9.77361202e-01 2.63200045e-01
8.21134686e-01 8.96694601e-01 7.90578723e-01 -1.40760556e-01
2.08520085e-01 -8.91409874e-01 -4.14732993e-02 5.53094387e-01
2.31220692e-01 9.27476436e-02 -1.18253064e+00 3.27014327e-01
-2.21517015e+00 -3.66136104e-01 -6.42911136e-01 2.17068005e+00
6.79265708e-02 -1.75893664e-01 -6.52523398e-01 -4.83520776e-01
5.88065743e-01 9.46088135e-02 -3.59299809e-01 3.06592286e-01
-5.21931350e-01 -2.53906816e-01 1.33925951e+00 1.20393765e+00
-1.13514793e+00 1.48971772e+00 6.09757471e+00 2.13412568e-01
-5.96286535e-01 -1.16017774e-01 -3.83477092e-01 6.64334670e-02
-2.73705274e-01 7.27227032e-01 -1.33755112e+00 -5.06242253e-02
2.77066886e-01 5.83205931e-02 3.35714549e-01 1.10863197e+00
-1.17939882e-01 -6.01266742e-01 -1.13004446e+00 1.57178783e+00
4.15922284e-01 -1.61031008e+00 -9.98874009e-02 4.83070672e-01
9.51326847e-01 4.01962548e-01 -2.74788946e-01 -2.12488726e-01
4.93895143e-01 -7.26671278e-01 1.04714119e+00 3.58345389e-01
7.69659519e-01 -7.32871234e-01 7.23947823e-01 6.14238441e-01
-1.39454663e+00 3.18423063e-01 -7.23026335e-01 -3.58214349e-01
2.61964440e-01 4.27495033e-01 -1.24105394e+00 1.16987610e+00
6.17515028e-01 9.09440935e-01 -8.37247252e-01 1.28567171e+00
-3.88370782e-01 -4.38121408e-01 -8.37197661e-01 1.82114184e-01
3.44037831e-01 -4.50683683e-01 5.21049440e-01 8.08954895e-01
5.37307322e-01 -2.04060495e-01 4.60657507e-01 8.93354118e-01
2.26093963e-01 -2.32611284e-01 -9.98609841e-01 7.13408113e-01
8.90112221e-01 1.18536806e+00 -9.70532596e-01 -1.34746969e-01
-3.07809442e-01 1.19197345e+00 2.56005496e-01 2.92739213e-01
-7.71267891e-01 -1.86482325e-01 6.65057123e-01 1.32048160e-01
-5.33848144e-02 -1.00386012e+00 -4.98748213e-01 -1.32834208e+00
1.92821816e-01 -4.47940268e-03 -1.63413137e-01 -1.30719757e+00
-6.14835143e-01 3.97012413e-01 1.03060655e-01 -1.34880483e+00
-1.19029574e-01 -6.37483835e-01 8.55046138e-02 9.62605298e-01
-1.67035890e+00 -1.57145333e+00 -8.81543636e-01 5.53010464e-01
3.22717458e-01 1.14480220e-01 6.54587567e-01 -4.29145284e-02
1.82600334e-01 -7.15730935e-02 1.23714000e-01 6.91634975e-03
6.60004556e-01 -1.16217995e+00 5.23564637e-01 1.07913613e+00
6.89957201e-01 7.37114668e-01 5.93218982e-01 -1.11388540e+00
-1.82638550e+00 -1.03347242e+00 9.28585112e-01 -1.19696927e+00
3.80131572e-01 -1.09931135e+00 -2.37203792e-01 1.28308523e+00
-2.95831233e-01 3.77293490e-02 -8.18321332e-02 2.29953960e-01
-2.98960090e-01 -1.01079363e-02 -9.92667198e-01 4.52763855e-01
1.56510568e+00 -5.49083114e-01 -4.71273869e-01 5.89727998e-01
9.16260481e-01 -1.27187860e+00 -3.03593129e-01 4.64564711e-01
5.37507296e-01 -8.68410587e-01 1.26799726e+00 1.71759024e-01
-5.44714630e-01 -8.54331195e-01 -6.84678972e-01 -1.06555164e+00
-3.64898145e-01 -5.94896972e-01 9.42928717e-02 1.01005876e+00
5.98429739e-02 -6.83234215e-01 1.01555192e+00 5.20199001e-01
-1.49151951e-01 -9.75186080e-02 -9.85460341e-01 -9.45048690e-01
-8.53678524e-01 -4.18520600e-01 7.89428532e-01 9.41547334e-01
-4.70581084e-01 2.15022236e-01 -4.88497764e-01 1.03455901e+00
9.88049328e-01 2.46595651e-01 1.48043931e+00 -1.45496011e+00
3.55105519e-01 8.66666362e-02 -1.04387808e+00 -1.45728552e+00
2.83607453e-01 -7.79101074e-01 9.35871676e-02 -1.87252092e+00
9.66786593e-02 -6.47646129e-01 2.78826177e-01 1.66890681e-01
3.83004576e-01 2.10961759e-01 5.77551201e-02 3.84679228e-01
-9.67221260e-01 3.61658186e-01 8.63012195e-01 1.88537508e-01
-6.73574731e-02 -3.71981859e-01 -2.40321353e-01 1.01423097e+00
1.97057724e-01 -3.59685987e-01 -2.81951785e-01 -9.40661609e-01
2.23895878e-01 6.04892448e-02 5.74495614e-01 -1.19080889e+00
7.80042708e-01 -2.86016375e-01 7.05028296e-01 -1.51154625e+00
8.74756396e-01 -1.30464411e+00 6.19805574e-01 2.01713383e-01
3.81372541e-01 2.04123512e-01 2.36281957e-02 6.58190966e-01
-8.37928876e-02 -4.90838140e-02 2.48322710e-01 -2.58566916e-01
-1.31770837e+00 6.39036655e-01 5.84538460e-01 -6.27841651e-01
1.05458617e+00 -6.87744081e-01 -1.99973106e-01 -5.14139056e-01
-3.43421489e-01 5.93494952e-01 1.24413872e+00 6.95281208e-01
9.42630827e-01 -1.61635149e+00 -4.25923288e-01 5.57494342e-01
5.27888060e-01 6.45216823e-01 -1.86546911e-02 6.16228342e-01
-1.19066191e+00 1.00525677e+00 -1.00670770e-01 -1.32150471e+00
-1.26879036e+00 4.10867065e-01 1.00833867e-02 3.27341914e-01
-8.25286686e-01 6.05397224e-01 4.84708518e-01 -9.06511009e-01
1.08739197e-01 -3.55212063e-01 4.15568680e-01 -4.97625053e-01
3.79882932e-01 4.13561612e-01 -8.54358748e-02 -1.12638235e+00
-8.98072720e-01 1.23190808e+00 4.06584531e-01 -1.65301666e-01
1.11893582e+00 -7.62995780e-01 -3.79028141e-01 3.84402335e-01
9.75305080e-01 2.73754239e-01 -1.23732209e+00 -3.94680411e-01
2.65412271e-01 -1.14317560e+00 9.63339359e-02 -4.43612397e-01
-3.15332681e-01 5.90239584e-01 4.11897451e-01 -4.30659413e-01
3.43002498e-01 9.97171104e-02 2.72123873e-01 6.97080910e-01
1.49124074e+00 -8.13379169e-01 -4.62784082e-01 7.87202060e-01
7.68373787e-01 -1.31349552e+00 4.73979115e-01 -9.79157925e-01
-2.61918515e-01 9.76604521e-01 7.37422466e-01 -2.49401450e-01
3.63953531e-01 2.65316129e-01 -1.04203366e-01 -2.48298600e-01
8.51143226e-02 -1.37093470e-01 3.68025392e-01 9.90252316e-01
-2.64280945e-01 1.01053551e-01 4.05663997e-01 -3.69329751e-02
-4.99969184e-01 -3.13712329e-01 3.04048061e-01 1.03639936e+00
-7.49225557e-01 -8.56209397e-01 -8.15610647e-01 -1.43434033e-01
5.24518132e-01 1.32263526e-02 -4.06970233e-01 1.20263124e+00
1.22236684e-01 7.00638413e-01 3.18889499e-01 -3.37632328e-01
2.90631950e-01 -1.96482986e-01 7.88531303e-01 -6.93668962e-01
4.04845446e-01 8.98052454e-02 2.87482291e-02 -1.11946905e+00
-3.27385783e-01 -5.17384171e-01 -1.41169918e+00 -2.54515409e-01
-5.27132630e-01 -6.92351768e-03 1.34480608e+00 8.30332279e-01
5.71860611e-01 -5.48187569e-02 3.62588525e-01 -1.11348844e+00
1.29828885e-01 -5.38170099e-01 -7.17258632e-01 8.49746838e-02
2.74129838e-01 -1.16918421e+00 -1.55538887e-01 -2.55472392e-01] | [7.349660873413086, -2.267838954925537] |
df4b726a-6d63-4021-ab6b-1f0ae23768f1 | bsnlp2019-shared-task-submission-multisource | null | null | https://aclanthology.org/W19-3710 | https://aclanthology.org/W19-3710.pdf | BSNLP2019 Shared Task Submission: Multisource Neural NER Transfer | This paper describes the Cognitive Computation (CogComp) Group{'}s submissions to the multilingual named entity recognition shared task at the Balto-Slavic Natural Language Processing (BSNLP) Workshop. The final model submitted is a multi-source neural NER system with multilingual BERT embeddings, trained on the concatenation of training data in various Slavic languages (as well as English). The performance of our system on the official testing data suggests that multi-source approaches consistently outperform single-source approaches for this task, even with the noise of mismatching tagsets. | ['Tatiana Tsygankova', 'Dan Roth', 'Stephen Mayhew'] | 2019-08-01 | null | null | null | ws-2019-8 | ['multilingual-named-entity-recognition'] | ['natural-language-processing'] | [-4.12126243e-01 -1.05369084e-01 7.94543102e-02 -5.95125616e-01
-1.10592067e+00 -8.14105392e-01 7.46957123e-01 5.75106025e-01
-1.38177931e+00 9.67291296e-01 8.78404021e-01 -4.32533652e-01
1.15716130e-01 -5.01627624e-01 -3.64867389e-01 8.36227462e-03
1.06317684e-01 8.25611949e-01 2.27650329e-01 -5.11550725e-01
1.53581262e-01 4.81880158e-01 -8.57449353e-01 4.01362717e-01
7.94786751e-01 4.97967482e-01 1.35466203e-01 6.33046687e-01
-3.28241050e-01 8.27587426e-01 -5.16268730e-01 -8.59925091e-01
1.39641911e-01 1.28785759e-01 -9.80573893e-01 -9.32301939e-01
4.15985465e-01 3.60318720e-01 -2.80008852e-01 1.19547617e+00
8.02376568e-01 3.53769064e-01 3.55123967e-01 -7.81652927e-01
-1.18300116e+00 9.30063128e-01 9.46893841e-02 6.35157287e-01
2.32231945e-01 -4.48042840e-01 1.08573914e+00 -1.58871651e+00
1.22360802e+00 1.11088657e+00 1.04850197e+00 1.48954377e-01
-8.53982627e-01 -4.28170711e-01 -1.68909401e-01 3.38217318e-01
-1.81849885e+00 -9.33570564e-01 1.42365158e-01 -3.50509584e-01
1.46341121e+00 -2.83140272e-01 9.19363350e-02 9.00415480e-01
2.51990765e-01 7.28185475e-01 1.19294477e+00 -8.57722282e-01
2.17154041e-01 2.57786691e-01 4.80869144e-01 6.48789525e-01
4.13180053e-01 2.66296715e-01 -6.91336691e-01 -4.22864705e-01
1.31512970e-01 -6.94436550e-01 1.13413960e-01 9.66853052e-02
-1.35427046e+00 7.70645320e-01 8.17507729e-02 8.24855864e-01
-4.25721347e-01 1.73562951e-02 8.32378209e-01 4.09053683e-01
6.66684687e-01 6.91552877e-01 -1.02448833e+00 -1.34096250e-01
-7.68953621e-01 1.32904172e-01 9.06530976e-01 1.14860737e+00
4.80828106e-01 2.22185880e-01 -1.48351863e-01 1.19761860e+00
2.14049920e-01 4.49767083e-01 7.67286718e-01 -4.64984298e-01
7.23496795e-01 2.46417642e-01 7.28876702e-03 -6.52427673e-01
-5.46022058e-01 -2.24239990e-01 -3.01310450e-01 -2.32949674e-01
3.35580140e-01 -5.70850432e-01 -6.91074312e-01 1.75069988e+00
1.32235065e-01 -1.96336716e-01 6.89974010e-01 5.67983985e-01
9.91589069e-01 3.71725202e-01 6.25225365e-01 1.71738863e-01
1.42795360e+00 -9.14657295e-01 -7.79737532e-01 -2.58217186e-01
8.50178838e-01 -1.02792001e+00 4.10617888e-01 -1.32769510e-01
-9.21086371e-01 -5.91220737e-01 -9.76928592e-01 -4.65586454e-01
-1.29146481e+00 4.15548265e-01 6.83097839e-01 6.90225005e-01
-1.27858841e+00 1.61941335e-01 -5.68383515e-01 -6.19275689e-01
-2.63125628e-01 -9.38311890e-02 -7.12334335e-01 -3.43027472e-01
-1.63796949e+00 1.52976906e+00 1.14068460e+00 -5.98732382e-03
-4.75229979e-01 -7.74256706e-01 -1.12847531e+00 -3.97736043e-01
1.29903048e-01 6.40666559e-02 1.08535409e+00 -4.46944028e-01
-8.98981571e-01 1.27678978e+00 -2.98662353e-02 -5.36489546e-01
1.34428918e-01 -3.15095961e-01 -1.06948531e+00 -2.52726853e-01
5.53613186e-01 6.71927989e-01 -3.00920635e-01 -6.97023273e-01
-7.63745010e-01 -4.13692951e-01 -7.99444318e-02 4.10075665e-01
-1.30407959e-01 9.51934874e-01 -4.19026576e-02 -7.00539649e-01
-2.46201009e-01 -8.82644892e-01 -3.03578854e-01 -7.36175597e-01
-1.37072438e-02 -6.32032812e-01 -2.81866174e-02 -1.14765811e+00
1.01121473e+00 -2.03312683e+00 -2.87814111e-01 -1.73243303e-02
-3.00211847e-01 4.43022400e-01 -2.11537540e-01 6.23437464e-01
-1.24000728e-01 2.94135183e-01 1.72276542e-01 -1.83406010e-01
1.40761927e-01 1.54904038e-01 -3.54327634e-02 3.41292441e-01
1.07458010e-01 1.06422865e+00 -1.09766662e+00 -5.63612998e-01
-1.73192561e-01 3.91992092e-01 -1.12135880e-01 -1.87226549e-01
3.42765033e-01 -7.13418573e-02 -3.21836062e-02 4.06201571e-01
2.94533819e-01 5.55095494e-01 3.91058832e-01 1.31001770e-01
-5.13287008e-01 7.13517904e-01 -9.46885347e-01 1.94159842e+00
-4.98937696e-01 6.42755032e-01 8.68265778e-02 -4.52048540e-01
8.54907870e-01 7.71844506e-01 -2.00177897e-02 -4.54566687e-01
-8.38538036e-02 7.23778605e-01 2.08319500e-02 -1.71327531e-01
1.01466823e+00 -2.51098335e-01 -7.38388121e-01 2.54663169e-01
6.54436707e-01 6.18137196e-02 2.14450255e-01 4.37454581e-02
1.08709955e+00 1.11419760e-01 7.46156931e-01 -8.39762151e-01
3.90405536e-01 3.54842156e-01 1.03015602e+00 6.21251464e-01
-6.07225835e-01 3.37517470e-01 -3.41002434e-01 -5.33364773e-01
-1.12768292e+00 -1.08721662e+00 -3.80142748e-01 1.75628698e+00
-4.60252732e-01 -5.00549555e-01 -4.01966393e-01 -6.13261998e-01
-2.12602839e-01 1.02541423e+00 -3.54322106e-01 3.88128936e-01
-8.50408614e-01 -6.19851708e-01 1.62661278e+00 6.57037258e-01
1.15446642e-01 -1.37174928e+00 -1.78167686e-01 6.41799510e-01
-1.34455368e-01 -1.52989805e+00 -5.90894520e-01 5.89405179e-01
-2.81985313e-01 -9.90590870e-01 -5.15333891e-01 -1.44418240e+00
1.25332877e-01 -3.45831394e-01 1.07237899e+00 -3.26175481e-01
-7.78308734e-02 2.27949128e-01 -2.76675373e-01 -7.70765841e-01
-4.19942230e-01 1.78334266e-01 3.90124917e-01 -5.58744609e-01
1.06500483e+00 -1.99609160e-01 3.83180916e-01 -1.93078727e-01
-6.35661602e-01 -2.05133438e-01 3.12651545e-01 9.56651807e-01
4.61197495e-01 -3.77436608e-01 1.05706394e+00 -9.02707458e-01
8.10706198e-01 -5.74238896e-01 -5.06551147e-01 6.91708505e-01
-5.09595513e-01 2.39084810e-02 6.54009700e-01 1.43615067e-01
-1.46587098e+00 -8.05497076e-03 -2.97679752e-01 2.29336351e-01
-7.03776062e-01 8.18089008e-01 -2.96318501e-01 2.06426814e-01
7.38855541e-01 2.33834416e-01 -8.64899695e-01 -6.77146673e-01
7.21449912e-01 8.33845794e-01 1.02871466e+00 -7.95296729e-01
2.37364158e-01 -2.19208539e-01 -5.52739680e-01 -9.64495897e-01
-7.56980777e-01 -6.18945241e-01 -1.11043990e+00 3.86088714e-02
1.22922182e+00 -1.19325173e+00 1.50661469e-01 3.00227672e-01
-1.78029311e+00 1.10088252e-01 -2.16484204e-01 8.83894444e-01
-2.02769399e-01 -8.90744254e-02 -1.10405278e+00 -5.62224925e-01
-3.63366544e-01 -6.62705362e-01 6.09707654e-01 -8.69595632e-02
-1.79426715e-01 -1.44138086e+00 6.97040081e-01 3.12843055e-01
3.97472888e-01 -8.62849429e-02 1.01840425e+00 -1.73812890e+00
2.22781286e-01 -2.07097858e-01 -1.25140458e-01 3.52366149e-01
-2.38526329e-01 -5.07944345e-01 -8.72475624e-01 -4.47623655e-02
-3.50290298e-01 -5.72152495e-01 6.32399082e-01 -2.29949966e-01
1.32692158e-01 2.07757205e-02 2.11651549e-02 1.35292843e-01
1.61181486e+00 3.41826886e-01 1.14281401e-01 2.80699372e-01
4.31048304e-01 5.03000855e-01 1.63435474e-01 -2.69696981e-01
7.86201060e-01 3.51361811e-01 -5.83883166e-01 3.18255484e-01
-3.88208389e-01 -4.10744101e-01 4.14160460e-01 1.56223691e+00
-2.51251180e-02 8.76243412e-02 -1.59649241e+00 1.05937672e+00
-1.62585557e+00 -8.56263936e-01 -3.95286679e-02 1.86963797e+00
9.19469357e-01 -1.77803040e-01 -5.90342999e-01 -5.74034214e-01
9.99528944e-01 1.61792025e-01 1.32020246e-02 -9.05615211e-01
-5.39079249e-01 6.17551148e-01 6.94590688e-01 4.89149362e-01
-1.24368000e+00 1.58225727e+00 7.34037638e+00 7.83393919e-01
-4.36873019e-01 9.06884551e-01 7.67554417e-02 3.35531622e-01
-2.12191552e-01 -2.94206180e-02 -1.11156368e+00 1.69102848e-01
1.69823897e+00 -5.65323114e-01 2.84343511e-01 8.86972368e-01
-3.11634719e-01 6.06731251e-02 -9.09939766e-01 7.09699690e-01
5.48614979e-01 -1.19897068e+00 -3.66957605e-01 -2.71689624e-01
8.10448885e-01 1.16878676e+00 -7.12376177e-01 8.42217505e-01
9.75255489e-01 -8.32786679e-01 8.70356560e-01 3.19218934e-01
7.47643590e-01 -9.58721340e-01 1.12525797e+00 2.32229456e-01
-1.29417300e+00 3.60966444e-01 -5.37747741e-01 1.53018430e-01
4.25904810e-01 3.89054447e-01 -7.48122513e-01 8.28856468e-01
3.08471054e-01 1.46226719e-01 -8.00025046e-01 1.08029604e+00
-4.30513531e-01 5.79537809e-01 -1.52517423e-01 -1.87537700e-01
2.57729441e-01 2.36976027e-01 6.33840859e-01 1.94312763e+00
-1.59739777e-02 8.40352252e-02 4.29203302e-01 1.40512675e-01
-4.38361436e-01 6.51440561e-01 -8.45143855e-01 -1.81888193e-01
7.63664901e-01 9.71330881e-01 -4.21945214e-01 -6.09533668e-01
-4.01208520e-01 8.95768106e-01 1.03668928e+00 1.93836108e-01
-2.10417256e-01 -1.00185061e+00 5.43147087e-01 -6.77194297e-01
2.22783070e-02 -6.47016764e-01 -3.53800505e-01 -1.29746056e+00
-2.07395688e-01 -4.26610231e-01 6.21186495e-01 -7.92400658e-01
-1.69253016e+00 1.09967005e+00 -6.30209371e-02 -3.61137390e-01
-3.30777973e-01 -9.10188913e-01 -1.95734903e-01 1.02401626e+00
-1.33431077e+00 -1.30898225e+00 6.83528781e-01 4.25942898e-01
4.13681090e-01 -7.24862039e-01 1.34801745e+00 6.03771865e-01
-9.09069646e-03 6.37839079e-01 1.97909817e-01 8.45502675e-01
1.10663521e+00 -1.14921594e+00 6.76272273e-01 9.94761169e-01
3.92258048e-01 7.14993000e-01 2.14655504e-01 -8.44597399e-01
-8.99521291e-01 -1.13887417e+00 2.34327912e+00 -7.66120672e-01
1.01829505e+00 -2.82534957e-01 -6.25390530e-01 9.95931089e-01
7.80418515e-01 -3.81777473e-02 1.13572228e+00 4.62175131e-01
-5.86165845e-01 3.87479246e-01 -1.07041848e+00 3.22486192e-01
1.03364217e+00 -1.10325956e+00 -1.59479582e+00 3.75115484e-01
8.13108385e-01 -1.70245066e-01 -1.06453407e+00 8.38043243e-02
2.81780630e-01 -3.10656488e-01 6.39893353e-01 -1.26955807e+00
6.22675233e-02 -1.87044278e-01 -7.49704659e-01 -1.53391087e+00
-4.30651963e-01 -9.10761878e-02 6.03222132e-01 1.44002497e+00
6.73132420e-01 -5.77900589e-01 3.29245776e-02 4.58138496e-01
-4.94954377e-01 -6.44848943e-02 -1.35720909e+00 -9.95161712e-01
4.42056686e-01 -1.00264561e+00 4.68795061e-01 1.48314512e+00
3.69851440e-01 5.99769413e-01 -8.41014534e-02 1.52121186e-01
5.03983796e-01 -5.92143536e-01 9.27238166e-03 -1.19740033e+00
8.12046453e-02 -2.30053905e-02 -7.88269699e-01 -1.58705369e-01
7.84984887e-01 -1.63385737e+00 4.45961893e-01 -1.65871513e+00
4.88123074e-02 -5.50110579e-01 -6.03707910e-01 8.19708824e-01
-1.72559936e-02 2.14838639e-01 4.40423846e-01 2.47279592e-02
-6.56783938e-01 3.19191873e-01 1.93802178e-01 1.16520219e-01
-3.43686417e-02 -6.62015736e-01 -6.99190736e-01 8.90831530e-01
6.79996789e-01 -6.80244207e-01 2.81169027e-01 -1.12377262e+00
3.43262494e-01 -3.15849110e-02 -8.75335485e-02 -1.10425591e+00
5.93994260e-01 -1.92988545e-01 3.86785179e-01 -6.05684698e-01
1.16824947e-01 -3.85844350e-01 -4.78004962e-02 8.53425786e-02
-5.71101844e-01 7.73787260e-01 2.87479341e-01 1.62545443e-01
-3.62872034e-01 -3.20488513e-01 7.22815275e-01 -2.52869725e-01
-1.04380774e+00 -1.40954226e-01 -7.11880684e-01 6.09439015e-01
7.39702463e-01 4.77962464e-01 -3.63540739e-01 2.15600282e-01
-8.02210927e-01 2.04696968e-01 1.97597235e-01 8.72063339e-01
2.17316061e-01 -1.69541788e+00 -1.09556139e+00 6.61707446e-02
3.89202952e-01 -7.38913596e-01 -2.00120181e-01 3.49725425e-01
-5.56242466e-01 1.00904477e+00 -3.83114576e-01 3.15969110e-01
-7.77224422e-01 3.49144816e-01 -1.25932917e-02 -3.98406148e-01
-2.06223145e-01 7.91897118e-01 -6.23430312e-01 -1.43241417e+00
-9.89229381e-02 1.45869955e-01 -2.89486200e-01 2.13494271e-01
6.07762754e-01 5.93852937e-01 4.59704697e-01 -1.23778188e+00
-7.22003937e-01 -8.63597021e-02 6.14345260e-03 -7.60716915e-01
1.31341541e+00 -2.48892643e-02 -1.18057847e-01 8.40883672e-01
1.30266118e+00 4.43969816e-01 6.43157810e-02 -5.43042600e-01
8.75477493e-01 8.81172568e-02 6.59065098e-02 -1.12108338e+00
-4.25277978e-01 5.68532646e-01 4.10680532e-01 -5.23085356e-01
2.97148377e-01 -1.08743064e-01 8.02079141e-01 8.94076049e-01
8.35173905e-01 -1.83055842e+00 -9.14406598e-01 1.50067317e+00
6.82004750e-01 -8.58957112e-01 -2.99360305e-01 2.23163754e-01
-7.94809997e-01 1.01456738e+00 4.01476830e-01 -2.81887949e-01
9.84057188e-01 2.05225870e-01 4.28880781e-01 -1.08956456e-01
-6.94209158e-01 -3.15503269e-01 2.07401603e-01 6.20422781e-01
7.63516963e-01 3.65837991e-01 -1.04198623e+00 9.24329996e-01
-4.33472335e-01 -8.32618922e-02 5.39807260e-01 1.14017510e+00
-3.04298073e-01 -1.16938531e+00 -2.57539004e-01 1.31488279e-01
-9.95223701e-01 -8.16378772e-01 -4.33743775e-01 6.67732656e-01
5.18187881e-01 9.43753242e-01 1.52751501e-03 -3.27254161e-02
3.11058879e-01 8.55182469e-01 4.93727960e-02 -9.59458590e-01
-9.96554792e-01 -3.27141255e-01 7.64833748e-01 -2.97501653e-01
-3.31368327e-01 -5.82898915e-01 -1.28344834e+00 1.90125164e-02
2.58221813e-02 6.00376725e-01 1.14170778e+00 9.00864363e-01
3.55999321e-01 8.73945430e-02 5.48291206e-02 -5.10224044e-01
-3.20846766e-01 -1.06903350e+00 -7.08095968e-01 2.64600337e-01
-3.42954993e-01 -3.81513149e-01 3.35329846e-02 8.62491578e-02] | [9.868691444396973, 9.795039176940918] |
8fb526f8-f197-4edd-90be-1b0c3d654770 | lifting-from-the-deep-convolutional-3d-pose | 1701.00295 | null | http://arxiv.org/abs/1701.00295v4 | http://arxiv.org/pdf/1701.00295v4.pdf | Lifting from the Deep: Convolutional 3D Pose Estimation from a Single Image | We propose a unified formulation for the problem of 3D human pose estimation
from a single raw RGB image that reasons jointly about 2D joint estimation and
3D pose reconstruction to improve both tasks. We take an integrated approach
that fuses probabilistic knowledge of 3D human pose with a multi-stage CNN
architecture and uses the knowledge of plausible 3D landmark locations to
refine the search for better 2D locations. The entire process is trained
end-to-end, is extremely efficient and obtains state- of-the-art results on
Human3.6M outperforming previous approaches both on 2D and 3D errors. | ['Denis Tome', 'Chris Russell', 'Lourdes Agapito'] | 2017-01-01 | lifting-from-the-deep-convolutional-3d-pose-1 | http://openaccess.thecvf.com/content_cvpr_2017/html/Tome_Lifting_From_the_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Tome_Lifting_From_the_CVPR_2017_paper.pdf | cvpr-2017-7 | ['monocular-3d-human-pose-estimation', 'weakly-supervised-3d-human-pose-estimation'] | ['computer-vision', 'computer-vision'] | [-3.78607333e-01 2.62663931e-01 -1.44217266e-02 -3.84498805e-01
-9.26949978e-01 -2.10688472e-01 2.93117732e-01 -1.49181217e-01
-9.53837216e-01 4.84871298e-01 4.14812982e-01 7.48884603e-02
1.54910803e-01 -1.66520923e-01 -8.53039443e-01 -6.92492947e-02
-5.37898578e-02 1.16086423e+00 3.07067394e-01 -3.31075221e-01
-9.58125666e-03 6.10205472e-01 -1.21715951e+00 -3.36047173e-01
9.86096933e-02 1.09563828e+00 -5.84450811e-02 8.85216773e-01
5.38205445e-01 3.05076301e-01 -5.19855320e-01 -4.17783022e-01
5.60217142e-01 2.23596990e-02 -8.03541303e-01 4.21420395e-01
5.31285644e-01 -6.48703158e-01 -5.66540658e-01 5.61601579e-01
9.52746868e-01 2.01886132e-01 5.08500814e-01 -9.95855272e-01
-2.39651844e-01 -1.26829103e-01 -7.21167684e-01 -1.22359820e-01
8.67322385e-01 4.03320760e-01 5.65152824e-01 -1.07215142e+00
4.96611685e-01 1.57679081e+00 9.89503086e-01 5.40009618e-01
-9.46460843e-01 -1.19657308e-01 7.23546073e-02 -1.13833353e-01
-1.55398297e+00 -8.98452848e-02 6.34691358e-01 -3.11981559e-01
1.22882831e+00 -1.76883921e-01 1.03314888e+00 1.13271952e+00
3.03576201e-01 9.71998990e-01 9.42861676e-01 -4.63235527e-01
-1.30411953e-01 -4.18807864e-01 -1.92505538e-01 1.15117991e+00
2.82511681e-01 2.82506645e-01 -8.29390705e-01 9.11568627e-02
1.19200683e+00 2.32594050e-02 1.57187935e-02 -7.34772265e-01
-1.32043338e+00 5.47449827e-01 8.89806271e-01 -3.17914844e-01
-5.92668414e-01 5.46974361e-01 9.66262221e-02 -4.38257098e-01
3.25092733e-01 2.07556427e-01 -7.50209749e-01 -2.58438289e-01
-7.34337449e-01 8.72044325e-01 4.39051449e-01 9.54322338e-01
3.84413272e-01 -5.26735425e-01 -8.41865502e-03 2.49361485e-01
6.72774851e-01 6.02796972e-01 -2.88131256e-02 -1.24466324e+00
4.64273423e-01 4.94804621e-01 6.43861055e-01 -7.29613721e-01
-1.06755960e+00 -4.71592814e-01 -4.54478920e-01 3.67945611e-01
5.36043108e-01 -1.17755994e-01 -1.30777347e+00 1.48021066e+00
7.63877511e-01 -2.62561888e-01 -2.34524190e-01 1.43213725e+00
7.06363440e-01 2.81087905e-01 7.88245816e-03 6.53220713e-01
1.36380005e+00 -1.12859571e+00 -4.95856106e-01 -6.59321785e-01
2.13120580e-01 -4.29782778e-01 6.08920276e-01 4.25872743e-01
-1.34316158e+00 -8.15259039e-01 -9.88523722e-01 -6.23175204e-01
-2.04863429e-01 4.46413219e-01 6.17577791e-01 4.10670221e-01
-1.07831097e+00 5.11405528e-01 -1.10098314e+00 -4.44914550e-01
3.19776028e-01 5.65771163e-01 -7.40821838e-01 5.97159714e-02
-9.97424304e-01 1.52096260e+00 5.25367439e-01 5.41154802e-01
-7.94805825e-01 -3.80609840e-01 -1.15501380e+00 -5.09188771e-01
5.53968251e-01 -1.41874540e+00 1.44445753e+00 1.05389923e-01
-1.64906204e+00 1.14617729e+00 -1.87718630e-01 -5.65961719e-01
9.06829298e-01 -1.17195511e+00 4.54016298e-01 1.89960241e-01
1.23106889e-01 1.18441236e+00 5.63221574e-01 -1.26283979e+00
-4.91243064e-01 -8.43303084e-01 -1.96451679e-01 6.48245454e-01
6.21008933e-01 -3.35971802e-01 -9.97637868e-01 -4.51241463e-01
4.61476505e-01 -1.21836078e+00 -6.63619578e-01 4.19764340e-01
-5.67300737e-01 -1.93019375e-01 1.95814922e-01 -8.27953994e-01
5.66074073e-01 -1.54683995e+00 7.47598648e-01 2.69834846e-01
2.68069625e-01 -1.28693320e-02 2.67059296e-01 -7.65792206e-02
2.51227349e-01 -3.47700536e-01 -1.50908818e-02 -1.16399240e+00
2.00687826e-01 1.30454943e-01 1.51934966e-01 8.44498634e-01
2.96520621e-01 1.35150635e+00 -7.03527153e-01 -3.88292968e-01
6.34112656e-01 9.32956278e-01 -3.71265054e-01 4.71953124e-01
1.15269721e-02 8.17891538e-01 -3.72199714e-01 7.00333357e-01
3.26039702e-01 -2.87206799e-01 -6.79665208e-02 -3.31738651e-01
1.40190318e-01 1.93579227e-01 -1.22124541e+00 2.51380181e+00
-2.30151147e-01 1.54057786e-01 -2.17777163e-01 -3.84799719e-01
6.95864737e-01 2.72530884e-01 4.95520085e-01 -2.45365098e-01
5.28302133e-01 1.27834976e-01 -6.46247387e-01 -1.35810763e-01
5.02735555e-01 -9.34856385e-02 -5.10932148e-01 1.27289951e-01
4.03510660e-01 -3.00668240e-01 -5.62962532e-01 -1.58212766e-01
8.43109429e-01 1.01297355e+00 7.32637823e-01 1.46109790e-01
1.38112783e-01 7.67308800e-03 2.22871095e-01 4.96137559e-01
-5.19941151e-01 9.00651634e-01 1.09679274e-01 -7.80270159e-01
-1.14914525e+00 -1.52437377e+00 3.46554220e-01 7.47486472e-01
4.17631954e-01 -4.68937516e-01 -5.63619256e-01 -6.92374527e-01
2.49667600e-01 3.19037497e-01 -8.73725712e-01 -3.47843803e-02
-7.28942513e-01 -2.50875264e-01 5.05616069e-01 1.08316112e+00
5.34489870e-01 -6.63276970e-01 -1.45219970e+00 2.02857964e-02
-4.42721248e-01 -1.29939032e+00 -3.58734876e-01 3.46457630e-01
-8.39565754e-01 -1.09493375e+00 -1.12371874e+00 -4.93540764e-01
6.72455311e-01 -6.07169531e-02 1.28810394e+00 -1.94399595e-01
-5.59244275e-01 4.95898455e-01 -2.86516845e-01 -4.97719944e-01
3.23828757e-01 -1.58559993e-01 3.48117501e-01 -5.89450657e-01
2.49911219e-01 -7.30229467e-02 -6.09175682e-01 3.69364142e-01
-7.13212565e-02 -2.56085098e-02 7.96737731e-01 5.91244638e-01
8.97287369e-01 -1.42073959e-01 -2.65680283e-01 -1.99568331e-01
-2.74392613e-03 1.24841444e-01 -4.17662680e-01 6.20436706e-02
-1.35314301e-01 2.32957765e-01 -2.30439886e-01 -1.87833622e-01
-9.44870591e-01 8.27796280e-01 -3.87630343e-01 -4.84864950e-01
-5.62774122e-01 1.18023492e-02 -1.28165916e-01 3.71492817e-03
5.77579081e-01 -9.84058976e-02 7.33772218e-02 -6.86936915e-01
6.60349488e-01 1.53420344e-01 1.11299646e+00 -3.68898064e-01
8.65935326e-01 6.70674205e-01 2.40944803e-01 -3.76552433e-01
-1.08847702e+00 -6.11091197e-01 -1.46171725e+00 -4.57033277e-01
1.31781566e+00 -1.43420494e+00 -9.25088108e-01 6.56745017e-01
-1.42659175e+00 -3.17426771e-01 -8.22611973e-02 6.07104659e-01
-9.23898697e-01 2.92495459e-01 -4.87730533e-01 -8.78253102e-01
-2.76745290e-01 -1.17843723e+00 1.90620553e+00 1.66965332e-02
-7.36882329e-01 -6.64163232e-01 5.52705396e-03 3.80462706e-01
-1.40862301e-01 6.86811566e-01 2.41055444e-01 -1.89546168e-01
-4.58047420e-01 -5.88581324e-01 -1.60610199e-03 -1.55278400e-01
-3.94886047e-01 -6.05623186e-01 -7.98486948e-01 -2.09167391e-01
-2.07440406e-01 -5.69897711e-01 8.03683341e-01 7.27407217e-01
6.83512092e-01 1.68449923e-01 -4.26487625e-01 5.31792223e-01
9.55807388e-01 -6.05815947e-01 5.21693707e-01 5.07719934e-01
9.14530277e-01 6.44424975e-01 8.22537541e-01 7.36353159e-01
8.12188923e-01 8.55161488e-01 6.19374335e-01 -1.43789142e-01
-1.34025007e-01 -4.47161496e-01 1.00366682e-01 5.79402782e-02
-3.91150087e-01 3.68283503e-02 -8.87996554e-01 3.61465931e-01
-1.93718183e+00 -3.41627210e-01 1.95945024e-01 2.03819156e+00
6.24616742e-01 5.21576762e-01 6.25323057e-01 7.52803087e-02
4.68731016e-01 8.14893916e-02 -5.53809643e-01 3.55868101e-01
2.53414363e-01 2.84836859e-01 7.64745593e-01 5.71351767e-01
-1.24349606e+00 1.07673490e+00 7.05056429e+00 1.88684940e-01
-4.40142810e-01 4.05813307e-02 2.77639091e-01 -3.88485789e-01
3.10230106e-01 -2.45248556e-01 -1.01679981e+00 -1.51297539e-01
4.04237360e-01 6.35513306e-01 -2.91461088e-02 9.02758777e-01
-8.85482952e-02 -4.26023573e-01 -1.24363422e+00 1.42033350e+00
1.70769259e-01 -9.17281866e-01 -2.59786457e-01 8.52732435e-02
3.78084630e-01 1.21397227e-02 1.96668692e-02 4.82828952e-02
2.39629269e-01 -9.70540881e-01 1.43889141e+00 7.72810161e-01
5.42772949e-01 -1.01673543e+00 7.61609733e-01 6.63760066e-01
-1.25610375e+00 2.19679311e-01 -1.02659427e-01 -1.88624635e-01
6.26806438e-01 2.90851116e-01 -6.78031564e-01 5.25746465e-01
1.04289699e+00 5.12832582e-01 -7.60501146e-01 9.74590182e-01
-6.93959355e-01 -3.96800667e-01 -6.09634757e-01 1.21992402e-01
1.85524538e-01 4.57935214e-01 5.33525229e-01 9.26659048e-01
1.86362118e-01 1.19072527e-01 4.73344356e-01 6.64738655e-01
3.13278615e-01 -5.61253369e-01 -4.14026886e-01 6.12678707e-01
1.85769588e-01 9.49866414e-01 -8.05865347e-01 -1.82648480e-01
1.28617808e-01 1.39663041e+00 4.21250612e-01 4.47188839e-02
-1.06495237e+00 -1.67850554e-01 4.78989244e-01 3.09736561e-02
5.00188053e-01 -8.68750870e-01 -3.83254260e-01 -8.49795282e-01
2.51687557e-01 -4.30572212e-01 3.79056215e-01 -9.81542647e-01
-1.06442678e+00 5.11471629e-01 2.97452509e-01 -9.64595914e-01
-4.77477103e-01 -9.19000268e-01 -7.21178651e-02 9.75029171e-01
-1.15675461e+00 -1.33100390e+00 -4.30788457e-01 5.04602969e-01
4.03344959e-01 3.77834857e-01 8.04586232e-01 -2.06848085e-01
-4.78277653e-02 5.36659777e-01 -8.33275259e-01 2.36160263e-01
7.03326762e-01 -1.44433320e+00 1.05365348e+00 8.30587983e-01
4.68226441e-04 5.05189002e-01 7.71322608e-01 -8.36484492e-01
-1.37291086e+00 -8.25427175e-01 1.07923663e+00 -1.31133461e+00
-1.07748292e-01 -4.45433706e-01 -1.00921966e-01 8.04852784e-01
-2.21630648e-01 5.25217131e-02 3.79252434e-01 2.66303122e-01
-3.20844561e-01 3.39007437e-01 -1.24183476e+00 5.88839948e-01
1.36693764e+00 -4.25304204e-01 -9.62754309e-01 1.87042192e-01
7.65020967e-01 -1.09295917e+00 -8.32001925e-01 5.01005471e-01
7.71352649e-01 -8.96420956e-01 1.46118164e+00 -3.87212336e-01
-1.30161196e-01 -4.58910465e-01 -3.58667403e-01 -1.03045702e+00
-3.91963869e-01 -6.26763225e-01 -4.73611116e-01 1.66562095e-01
2.10410133e-01 -1.56967789e-02 1.03102088e+00 6.71969712e-01
1.09209314e-01 -8.69034588e-01 -1.07393372e+00 -6.07904434e-01
-3.68617743e-01 -6.62109137e-01 2.07861945e-01 -1.62609369e-01
-3.08593363e-01 3.27597886e-01 -7.38334656e-01 4.21081603e-01
1.03803062e+00 -2.24168092e-01 1.13680840e+00 -1.21454358e+00
-2.04548225e-01 -2.45234177e-01 -6.65718794e-01 -1.70411897e+00
1.44934252e-01 -1.89809814e-01 6.39731884e-01 -1.65141952e+00
-7.57837370e-02 2.35403389e-01 1.13299720e-01 4.23096925e-01
-2.84513026e-01 6.00487351e-01 3.01659495e-01 -4.80947755e-02
-1.00383019e+00 6.27646208e-01 1.10884178e+00 2.51939535e-01
-4.05587517e-02 -6.28182963e-02 -4.53490794e-01 9.18065667e-01
3.21782053e-01 -2.86687791e-01 1.08195186e-01 -5.49983621e-01
5.91257550e-02 4.63972650e-02 9.68080103e-01 -1.45740950e+00
1.29058748e-01 2.51526773e-01 1.32424176e+00 -1.40716457e+00
8.76786411e-01 -6.75789297e-01 -1.21369980e-01 7.38129854e-01
-3.17908496e-01 2.99641013e-01 2.63751186e-02 7.11674333e-01
3.38136464e-01 3.86670679e-01 7.79065728e-01 -5.49745440e-01
-9.80228543e-01 5.26522934e-01 3.08907311e-03 4.73209657e-03
9.13658202e-01 -3.33368957e-01 4.77799028e-01 -4.17883873e-01
-1.12420237e+00 2.84574568e-01 4.75334018e-01 5.37038207e-01
9.83246446e-01 -1.45514119e+00 -5.55387318e-01 1.16026491e-01
2.86018141e-02 5.45920312e-01 3.18216421e-02 7.28081942e-01
-6.68894231e-01 5.57399094e-01 -2.60759592e-01 -9.50994670e-01
-1.04591763e+00 4.58957672e-01 4.43592995e-01 -2.59323478e-01
-7.34099507e-01 1.33141899e+00 -2.47127444e-01 -6.66730583e-01
7.19885886e-01 -2.91226864e-01 1.92835435e-01 -4.57811415e-01
4.22913760e-01 4.07289803e-01 -1.27762362e-01 -1.14521980e+00
-7.14671612e-01 9.75522459e-01 2.93403953e-01 -4.65376526e-01
1.28864646e+00 -4.48657811e-01 2.65513152e-01 4.24482375e-01
1.25474095e+00 -4.80884165e-01 -1.83637333e+00 -2.99122423e-01
-2.17721105e-01 -6.25481308e-01 -1.22124419e-01 -1.10306120e+00
-6.29185081e-01 7.69980073e-01 6.38358653e-01 -8.30864131e-01
7.61578143e-01 3.82619828e-01 9.22852576e-01 5.48208535e-01
7.06349015e-01 -1.07988763e+00 5.00439107e-01 7.44628787e-01
9.63548720e-01 -1.40923154e+00 5.05774617e-01 -2.44763687e-01
-7.12037504e-01 1.00326502e+00 7.12828517e-01 -4.64167118e-01
6.90979719e-01 -5.53921089e-02 1.52956426e-01 -3.07443649e-01
-2.60234445e-01 -5.54167986e-01 8.68086874e-01 7.35953629e-01
1.91839471e-01 1.32052675e-01 1.78860873e-01 5.59144795e-01
-2.68571496e-01 -1.13310523e-01 -2.48633489e-01 1.22802365e+00
-4.85750437e-01 -6.70875490e-01 -7.26898551e-01 -3.08460444e-01
-2.79834509e-01 4.18602824e-01 -5.83538115e-01 1.03561389e+00
3.06549519e-01 6.31803095e-01 -4.77406122e-02 -6.69713616e-01
7.23670542e-01 1.10315628e-01 1.22918403e+00 -5.62432706e-01
-3.32086742e-01 3.34484369e-01 -2.52270326e-02 -1.18853641e+00
-5.70546091e-01 -5.28179705e-01 -1.48438215e+00 -3.94288376e-02
-1.68585092e-01 -4.09655571e-01 6.76673651e-01 1.26324558e+00
2.46657237e-01 4.05699939e-01 5.97493574e-02 -1.84984910e+00
-6.42450809e-01 -9.07942712e-01 -3.69940519e-01 3.24447483e-01
3.74633789e-01 -1.15734541e+00 1.77913368e-01 -1.45283237e-01] | [6.958017349243164, -0.9192792773246765] |
38852fac-2f68-4efe-a56a-b92caf34c9c9 | who-wrote-this-code-watermarking-for-code | 2305.15060 | null | https://arxiv.org/abs/2305.15060v1 | https://arxiv.org/pdf/2305.15060v1.pdf | Who Wrote this Code? Watermarking for Code Generation | Large language models for code have recently shown remarkable performance in generating executable code. However, this rapid advancement has been accompanied by many legal and ethical concerns, such as code licensing issues, code plagiarism, and malware generation, making watermarking machine-generated code a very timely problem. Despite such imminent needs, we discover that existing watermarking and machine-generated text detection methods for LLMs fail to function with code generation tasks properly. Hence, in this work, we propose a new watermarking method, SWEET, that significantly improves upon previous approaches when watermarking machine-generated code. Our proposed method selectively applies watermarking to the tokens with high enough entropy, surpassing a defined threshold. The experiments on code generation benchmarks show that our watermarked code has superior quality compared to code produced by the previous state-of-the-art LLM watermarking method. Furthermore, our watermark method also outperforms DetectGPT for the task of machine-generated code detection. | ['Gunhee Kim', 'Jamin Shin', 'Sangdoo Yun', 'Hwaran Lee', 'Ilgee Hong', 'Jaewoo Ahn', 'Seokhee Hong', 'Taehyun Lee'] | 2023-05-24 | null | null | null | null | ['code-generation'] | ['computer-code'] | [ 5.90820491e-01 1.42924944e-02 -5.33140481e-01 3.57171148e-01
-8.97556365e-01 -6.55722618e-01 7.72096455e-01 5.01648188e-01
-7.59784579e-02 4.66943532e-01 2.90006101e-02 -7.25597203e-01
6.17716014e-01 -7.22630739e-01 -5.63312113e-01 -3.62507373e-01
-2.77875423e-01 -3.64656687e-01 6.14556611e-01 9.06162262e-02
1.01889467e+00 -1.79345496e-02 -1.26420832e+00 4.45572704e-01
8.86514187e-01 4.51045185e-01 2.13900972e-02 6.23362482e-01
-4.61353213e-01 9.81465101e-01 -8.38257611e-01 -7.70945370e-01
1.70484081e-01 -5.13072729e-01 -7.19575107e-01 -1.88332513e-01
2.73645401e-01 -3.09020191e-01 -3.49837154e-01 1.58417416e+00
2.16455892e-01 -6.20395482e-01 5.23174584e-01 -1.45377982e+00
-9.29293931e-01 8.93032610e-01 -1.09506214e+00 -8.67223665e-02
4.41850275e-01 5.13343103e-02 7.27353036e-01 -7.30021000e-01
6.82846487e-01 1.09645534e+00 7.30985165e-01 6.61438048e-01
-1.13609564e+00 -7.68569946e-01 -3.39015037e-01 -1.98148966e-01
-1.17255294e+00 -3.50124836e-01 8.55845153e-01 -7.04559863e-01
7.94162273e-01 3.33207458e-01 1.50239632e-01 9.73916233e-01
6.36103034e-01 6.72914326e-01 1.00347555e+00 -7.61390507e-01
2.68564790e-01 3.03150326e-01 -6.97605684e-02 9.56609547e-01
8.14536035e-01 -4.74165678e-02 -1.71398222e-01 -8.08970451e-01
2.01998934e-01 -4.27703969e-02 -5.31194687e-01 -2.05453858e-01
-1.37510335e+00 1.02726531e+00 -1.18876450e-01 4.82837230e-01
1.42767012e-01 6.10340834e-01 9.59621251e-01 3.39047045e-01
2.14080647e-01 4.20966119e-01 -9.62248147e-02 -2.96435714e-01
-1.22842276e+00 -3.78244556e-02 5.85460603e-01 9.41237509e-01
5.45679688e-01 1.97645470e-01 8.80673155e-02 2.74742573e-01
7.50480175e-01 3.84402126e-01 8.97931099e-01 -6.01608098e-01
8.01964760e-01 7.33135998e-01 -1.36509061e-01 -1.55539608e+00
2.09623143e-01 -1.36569487e-02 -6.26580715e-01 4.12315339e-01
7.94993415e-02 1.29276618e-01 -2.82051891e-01 1.34964144e+00
-2.66824756e-02 4.66495380e-03 1.73976973e-01 2.69455165e-01
4.46481794e-01 7.30007827e-01 -1.14909485e-02 -1.72663569e-01
1.40258420e+00 -9.15415704e-01 -6.59047604e-01 -1.99331149e-01
7.74193466e-01 -1.10257745e+00 8.91917884e-01 2.19828486e-01
-7.33990967e-01 -3.21256310e-01 -1.33870399e+00 3.26035410e-01
-2.62591034e-01 9.96392742e-02 4.79592949e-01 1.46263945e+00
-8.91869545e-01 3.05552155e-01 -5.25494576e-01 -5.29917562e-03
3.61336827e-01 1.38089051e-02 -2.45600715e-01 1.85357317e-01
-8.81091177e-01 4.87497061e-01 6.21112227e-01 -4.07593399e-01
-8.86107743e-01 -3.03331614e-01 -1.00136948e+00 -1.48938939e-01
2.00572684e-01 -6.09813780e-02 1.16876304e+00 -1.01700509e+00
-1.19301033e+00 9.93866265e-01 -1.20342383e-02 -6.82501078e-01
6.15676880e-01 2.89212987e-02 -6.41223192e-01 2.68886536e-01
-7.85204545e-02 4.10763502e-01 1.48792708e+00 -1.62090433e+00
-2.81076193e-01 2.65554190e-01 8.60082135e-02 -6.99967563e-01
-8.19736958e-01 3.68332773e-01 1.37911970e-02 -1.04656160e+00
-2.58872867e-01 -7.77728558e-01 -1.24566890e-01 7.87150189e-02
-4.69688088e-01 1.31474003e-01 1.20969999e+00 -7.49834538e-01
1.74914086e+00 -2.57983470e+00 -2.74383545e-01 1.53493762e-01
4.00334954e-01 4.86447901e-01 -2.88230270e-01 7.51646101e-01
-8.57014284e-02 8.63579273e-01 -7.50210702e-01 -1.94930092e-01
1.11358970e-01 -2.83062965e-01 -7.97573864e-01 6.65851116e-01
2.73802370e-01 8.56359005e-01 -1.06197166e+00 -8.06758165e-01
-1.49845779e-01 1.63493380e-01 -5.25388777e-01 -1.07651182e-01
-3.22175592e-01 -1.18379436e-01 -4.64904934e-01 7.49142945e-01
7.40447700e-01 -1.99213326e-01 1.19552158e-01 4.36653823e-01
-1.94747403e-01 7.25187138e-02 -8.71261656e-01 1.32509840e+00
-3.70794803e-01 9.49594319e-01 -1.37519434e-01 -4.59000349e-01
9.96087253e-01 4.74015981e-01 1.12497069e-01 -2.03167394e-01
5.02183139e-02 5.48568904e-01 -1.27725199e-01 -6.92525685e-01
9.10459697e-01 2.41211370e-01 -2.19246790e-01 7.61087716e-01
-6.46741331e-01 -1.48947537e-01 2.05238163e-01 4.30735111e-01
1.43747938e+00 1.51267154e-02 3.64562154e-01 4.46434468e-02
8.83190811e-01 -1.71923395e-02 1.10312618e-01 5.39867818e-01
-3.89199376e-01 3.80687922e-01 7.72896707e-01 1.05063356e-01
-1.08026803e+00 -4.64120120e-01 7.83828571e-02 5.34024477e-01
3.57127227e-02 -7.72691011e-01 -1.03813612e+00 -1.05268133e+00
2.86878902e-03 5.26597440e-01 -3.65093499e-01 -3.94819289e-01
-7.11028099e-01 -5.58550417e-01 1.09897840e+00 1.55886665e-01
5.02198637e-01 -1.23880589e+00 -8.73280764e-01 2.87434697e-01
-9.73156765e-02 -8.56317520e-01 -7.31720269e-01 -3.63756657e-01
-8.80517244e-01 -1.16349971e+00 -8.35050166e-01 -1.09101760e+00
1.03940833e+00 4.19361591e-01 7.54078329e-01 7.31457531e-01
-4.09193158e-01 5.90266772e-02 -6.14344060e-01 8.76088440e-02
-1.45258605e+00 -7.45512033e-03 -4.12154108e-01 -5.43590300e-02
6.70991018e-02 -1.24257840e-01 -2.99125075e-01 -2.03699488e-02
-1.55858576e+00 -9.64352489e-02 6.21615350e-01 6.53913558e-01
-1.00700051e-01 4.42065477e-01 6.27883017e-01 -8.42803001e-01
7.99013674e-01 -4.95837927e-01 -8.62086117e-01 1.51378170e-01
-6.72490895e-01 2.04186782e-01 6.73080623e-01 -6.48593009e-01
-7.54668236e-01 -1.44648448e-01 -3.90820429e-02 2.32670102e-02
1.76150516e-01 6.43336535e-01 1.10957704e-01 -4.97353464e-01
6.85784400e-01 4.24355179e-01 -2.70794872e-02 -3.01764637e-01
3.38553756e-01 9.96396542e-01 5.45045376e-01 -4.12050873e-01
1.55144835e+00 4.20418173e-01 -1.93251416e-01 -4.66844559e-01
1.55426309e-01 -7.74741396e-02 -2.22590327e-01 1.14781165e-03
6.44875586e-01 -6.91694021e-01 -3.21776360e-01 6.82731926e-01
-1.41704595e+00 7.99180940e-02 5.61910830e-02 3.32960002e-02
-3.19427758e-01 1.39740849e+00 -6.89450920e-01 -9.01682377e-01
-4.00526494e-01 -1.36229360e+00 1.07392788e+00 -2.78210163e-01
-2.92202473e-01 -7.28529632e-01 2.36363962e-01 2.78629690e-01
6.20164812e-01 6.39271796e-01 1.26730359e+00 -1.84213445e-01
-8.17915261e-01 -5.36976516e-01 -3.11609477e-01 1.90905139e-01
2.35832572e-01 5.05600572e-01 -6.85454309e-01 -6.17516518e-01
-2.61338204e-02 -1.25354812e-01 7.20709622e-01 -4.37080711e-01
1.08618939e+00 -5.87711036e-01 -2.99632549e-01 2.84732312e-01
1.62979794e+00 2.81668216e-01 8.62125695e-01 6.17464721e-01
2.71391153e-01 3.15351725e-01 4.10953194e-01 6.18621945e-01
-1.35097265e-01 3.47921818e-01 7.37675965e-01 2.30861112e-01
-5.91171645e-02 -3.89324039e-01 9.28874493e-01 1.08299470e+00
4.31377083e-01 -1.38477460e-01 -8.96607459e-01 5.31295121e-01
-1.53874016e+00 -1.02976334e+00 -5.70976257e-01 2.22921395e+00
1.02550781e+00 4.35684055e-01 -2.01675311e-01 5.27576625e-01
8.55288208e-01 4.42449808e-01 -7.50420690e-02 -5.40302932e-01
2.31639445e-01 2.48433407e-02 7.63278306e-01 1.25893608e-01
-1.12112820e+00 6.77905083e-01 6.23918104e+00 1.08240545e+00
-1.03369498e+00 2.97382206e-01 1.38500452e-01 6.98093176e-01
-7.38134563e-01 3.27567279e-01 -5.60948670e-01 8.68720472e-01
9.41924274e-01 -5.58187783e-01 2.61508942e-01 1.03178930e+00
7.95304030e-02 6.29138276e-02 -6.28957927e-01 6.64746642e-01
1.84815243e-01 -1.31745839e+00 5.25769182e-02 1.62007615e-01
8.85829687e-01 -4.58068758e-01 1.86453462e-01 6.70781955e-02
6.25244752e-02 -7.47585118e-01 9.79242146e-01 -3.13486159e-02
8.60678852e-01 -6.43585503e-01 7.22342789e-01 2.90362924e-01
-1.33663690e+00 -9.29985866e-02 -3.16171616e-01 2.58242011e-01
-7.70342574e-02 6.81070507e-01 -7.13527381e-01 2.91555047e-01
-1.20085077e-02 3.93067747e-01 -8.96172106e-01 1.23230159e+00
-3.28498006e-01 7.95251369e-01 3.91695112e-01 -7.47320876e-02
1.76753089e-01 3.46028537e-01 5.43561220e-01 1.50598276e+00
5.51019490e-01 -6.63145363e-01 -2.84383018e-02 9.50975418e-01
-3.94216448e-01 2.63017476e-01 -9.41682994e-01 -7.43744373e-01
3.07117343e-01 1.10119951e+00 -1.02797961e+00 -4.25445944e-01
-6.97315693e-01 9.75481689e-01 -2.73255885e-01 5.69280684e-02
-1.09346199e+00 -8.92005622e-01 3.22288752e-01 -4.53424938e-02
1.83052599e-01 -5.17429292e-01 -4.70107757e-02 -1.28408957e+00
3.12702149e-01 -1.22640145e+00 5.77247478e-02 -3.75828892e-01
-7.69160569e-01 5.05872250e-01 -2.99797207e-01 -1.58383119e+00
-5.87907881e-02 -4.61439341e-01 -5.14626205e-01 4.04746503e-01
-1.71802199e+00 -7.92900026e-01 -3.15220244e-02 4.90112789e-02
4.75045800e-01 -3.28416288e-01 7.67454088e-01 5.03877103e-01
-2.67499894e-01 6.37636006e-01 3.54977041e-01 4.99425083e-01
6.57943666e-01 -9.38336372e-01 1.02145946e+00 1.27178419e+00
1.10510914e-02 8.48735273e-01 4.74220186e-01 -1.02122879e+00
-1.72640669e+00 -1.10923564e+00 9.78284299e-01 -2.23354891e-01
9.86519754e-01 -4.35364634e-01 -1.02197778e+00 3.95516634e-01
3.45554501e-01 -3.48395675e-01 5.57730436e-01 -9.55121338e-01
-7.87433624e-01 4.21414286e-01 -1.30701315e+00 6.29635036e-01
6.10584259e-01 -7.68031895e-01 -6.85347855e-01 3.38585526e-01
8.57480228e-01 -1.29755720e-01 -6.14022017e-01 1.91585571e-02
3.06486517e-01 -7.89306879e-01 7.05948353e-01 -1.32515416e-01
8.36064041e-01 -6.78532600e-01 -1.07864298e-01 -6.71187222e-01
1.26199797e-01 -1.03444231e+00 -3.68912965e-01 1.48561382e+00
1.99321046e-01 -8.41205895e-01 6.24667406e-01 1.16833560e-01
1.84003696e-01 -9.43799689e-02 -8.77299607e-01 -9.78333414e-01
2.80505270e-02 -2.60545731e-01 6.12511098e-01 1.15227878e+00
2.83078730e-01 -5.25882781e-01 -4.23469752e-01 1.12720668e-01
7.15657055e-01 2.17382118e-01 7.71836758e-01 -8.70306790e-01
-4.49299812e-01 -7.38519669e-01 -7.06609428e-01 -7.32624829e-01
3.38208348e-01 -1.06356823e+00 -3.08505744e-02 -1.14846194e+00
4.19178158e-01 -7.07681850e-02 -4.69228625e-03 5.37023008e-01
-2.28541732e-01 4.00953889e-01 3.33510876e-01 2.85943180e-01
-2.49850348e-01 2.25551680e-01 8.17471743e-01 -5.38172424e-01
2.59884566e-01 -2.29044035e-01 -5.32695591e-01 5.50656378e-01
7.72165298e-01 -9.25338328e-01 -2.07943201e-01 -2.15796351e-01
5.11471272e-01 8.32660217e-03 2.45325044e-01 -1.12318850e+00
-8.55129734e-02 -1.19081557e-01 -2.02393249e-01 -3.31827432e-01
-2.61110276e-01 -8.10993493e-01 2.78360337e-01 1.12898064e+00
-3.30071718e-01 2.52514362e-01 2.35101774e-01 5.58046877e-01
-2.52937883e-01 -9.65161562e-01 7.68739045e-01 -5.12695312e-02
-7.11416721e-01 -1.60374045e-01 -8.72983217e-01 5.12507521e-02
1.22963107e+00 -2.09589168e-01 -7.68508911e-01 -1.16728358e-01
1.27066866e-01 -2.33578011e-01 1.07107842e+00 6.77297413e-01
7.60113657e-01 -1.23368096e+00 -6.42951250e-01 2.25519851e-01
3.33129704e-01 -8.57469380e-01 -3.55799437e-01 4.40638691e-01
-8.62377822e-01 2.00168565e-01 4.11173999e-02 -1.93391338e-01
-1.58355749e+00 9.96198475e-01 -1.66223899e-01 -1.90655157e-01
-6.04383051e-01 2.32372895e-01 -2.03009948e-01 -5.17645776e-02
4.52442430e-02 -4.91209030e-01 1.97017550e-01 -2.45514110e-01
8.12454402e-01 3.12058836e-01 -8.46928284e-02 -6.97746456e-01
-2.00526163e-01 6.35854185e-01 2.36652326e-02 8.79866853e-02
8.25086892e-01 1.94669366e-01 -6.04684055e-01 1.34694979e-01
1.36947191e+00 7.84651816e-01 -7.16853440e-01 1.78976461e-01
4.69360024e-01 -8.75634372e-01 -2.55698413e-01 -3.53263408e-01
-9.49977398e-01 7.85439193e-01 4.36422706e-01 6.66243851e-01
7.66385734e-01 -3.98705930e-01 1.12928200e+00 2.04372749e-01
9.12864029e-01 -5.52179992e-01 2.96428502e-01 2.37861678e-01
5.30051649e-01 -1.08968079e+00 9.77219641e-02 -4.28069115e-01
-1.27644882e-01 1.28129375e+00 3.27932596e-01 1.49192303e-01
1.78958878e-01 5.98821461e-01 1.02707647e-01 1.91238567e-01
-4.12620932e-01 3.72201711e-01 3.65970912e-03 5.61133206e-01
3.99683028e-01 -6.89589977e-02 -6.56600535e-01 -2.02710271e-01
5.63797839e-02 3.94199342e-02 9.64317024e-01 1.54633939e+00
-7.16233671e-01 -1.54151142e+00 -9.79608238e-01 2.60790676e-01
-8.14561427e-01 -3.45563680e-01 -4.80959654e-01 6.99036896e-01
-4.35524881e-02 9.12307203e-01 -5.23573458e-01 -3.08860421e-01
-2.82625675e-01 5.32339476e-02 3.25504281e-02 -5.10658324e-01
-8.05655479e-01 -2.74251312e-01 -1.82648972e-01 -1.63409159e-01
-2.97599196e-01 -4.52914149e-01 -1.46658456e+00 -3.76101255e-01
-5.51229656e-01 1.75768480e-01 8.27014625e-01 3.05524707e-01
3.72628450e-01 3.26920718e-01 6.29791737e-01 -4.78636652e-01
-7.12580681e-01 -4.86209244e-01 -2.68616915e-01 4.54401106e-01
5.97333968e-01 -2.91491240e-01 -6.02026582e-01 5.41688859e-01] | [6.971961975097656, 7.824173450469971] |
53da3ebe-42f5-432e-a718-40325a686059 | augmenting-an-assisted-living-lab-with-non | 2002.05593 | null | http://arxiv.org/abs/2002.05593v1 | http://arxiv.org/pdf/2002.05593v1.pdf | Augmenting an Assisted Living Lab with Non-Intrusive Load Monitoring | The need for reducing our energy consumption footprint and the increasing
number of electric devices in today's homes is calling for new solutions that
allow users to efficiently manage their energy consumption. Real-time feedback
at device level would be of a significant benefit for this application. In
addition, the aging population and their wish to be more autonomous have
motivated the use of this same real-time data to indirectly monitor the
household's occupants for their safety. By breaking down aggregate power
consumption into its components, Non-Intrusive Load Monitoring provides
information on individual appliances and their current state of operation.
Since no additional metering equipment is required, residents are not
confronted with intrusion into their familiar environment. Our work aims to
depict an architecture supporting non-intrusive measurement with a smart
electricity meter and the handling of these data using an open-source platform
that allows to visualize and process real-time data about the total energy
consumed. As a case study, we describe a series of measurements from common
household devices and show how abnormal behavior can be detected. | [] | 2020-02-13 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [-1.23625204e-01 9.15189087e-02 2.07802221e-01 -5.06833553e-01
-1.73169702e-01 -5.44599891e-01 3.45239550e-01 6.24083281e-01
-1.47242725e-01 7.07395017e-01 3.78750749e-02 -3.08007300e-01
6.40352070e-02 -1.15780365e+00 1.45710751e-01 -6.94896877e-01
-1.24521039e-01 2.22859263e-01 2.15830043e-01 -1.90317541e-01
-1.67698324e-01 6.76740587e-01 -1.92373109e+00 -7.09805340e-02
7.53498912e-01 1.19461882e+00 -5.48429368e-03 5.57444453e-01
4.65400636e-01 4.21656400e-01 -8.99899304e-01 2.83023000e-01
-4.77878675e-02 -1.08144455e-01 -4.61167008e-01 -4.30237092e-02
-2.45038971e-01 -9.04486358e-01 2.76421666e-01 9.63192761e-01
6.49604142e-01 5.75526208e-02 2.25300968e-01 -1.62086535e+00
3.20235044e-01 5.21494687e-01 7.68951792e-03 2.92900234e-01
8.45594704e-01 3.47578198e-01 3.85568649e-01 3.76089692e-01
-5.53366840e-02 6.50377274e-01 4.14302498e-01 1.36101395e-02
-1.69120109e+00 -7.27199793e-01 2.94655804e-02 6.45599484e-01
-1.59002173e+00 -4.49823260e-01 9.64603841e-01 -2.15821445e-01
1.65095651e+00 1.01674378e+00 8.62623572e-01 1.12098157e+00
4.96849678e-02 2.48076454e-01 1.06980276e+00 -4.32501227e-01
1.00777197e+00 5.75537682e-01 1.41130283e-01 6.90590441e-02
7.27775633e-01 -1.68588296e-01 -2.70663928e-02 -3.99150223e-01
-2.10931301e-02 2.22736508e-01 -3.08616281e-01 -2.17591390e-01
-7.43605137e-01 6.71895742e-02 -4.61563691e-02 9.62840080e-01
-3.88310254e-01 1.03847096e-02 5.05594134e-01 1.16133228e-01
1.35035262e-01 1.28318295e-01 -7.01432288e-01 -7.27105260e-01
-8.12932134e-01 -2.52665013e-01 1.13904488e+00 9.04536366e-01
7.26628184e-01 -1.29816353e-01 3.12026680e-01 1.27876565e-01
5.42000830e-01 5.00624657e-01 6.96897030e-01 -7.48289824e-01
1.03146337e-01 1.22301126e+00 4.54683661e-01 -6.94368899e-01
-8.91632497e-01 2.31025890e-01 -9.76263642e-01 6.30530238e-01
-4.76372726e-02 -1.58587933e-01 -8.69854689e-02 1.45691359e+00
2.95991123e-01 -3.97343338e-01 -3.43018293e-01 3.45807552e-01
1.79829374e-01 4.21095282e-01 2.17052788e-01 -6.84995592e-01
1.69344366e+00 1.76950052e-01 -1.15004218e+00 2.59981751e-01
5.73576510e-01 -1.97664425e-01 9.63498592e-01 5.30254483e-01
-7.56361306e-01 -2.92376369e-01 -1.47976005e+00 3.02392960e-01
-9.43072498e-01 -2.82859474e-01 4.65543158e-02 1.30184031e+00
-9.67309713e-01 7.40947783e-01 -1.40134668e+00 -8.00332487e-01
2.65271783e-01 7.16390073e-01 -9.83635858e-02 7.69441903e-01
-8.50326240e-01 1.15685773e+00 4.49835122e-01 -3.49765122e-01
-1.28875583e-01 -7.57713854e-01 -6.21200562e-01 3.29064220e-01
-9.07561257e-02 -2.32268810e-01 1.33998156e+00 -7.41646141e-02
-1.54066133e+00 3.70944500e-01 1.90317601e-01 -2.17215002e-01
6.85843050e-01 -8.38522837e-02 -1.03492820e+00 -4.88522016e-02
-7.89842531e-02 -2.05582023e-01 3.02086502e-01 -8.62108350e-01
-8.33090425e-01 -5.67565024e-01 -1.79136038e-01 -4.12429154e-01
-6.46944880e-01 -3.54269624e-01 5.40404558e-01 2.12474000e-02
-2.82378614e-01 -5.06362200e-01 1.42958328e-01 -4.30343181e-01
-5.12147188e-01 -2.84622729e-01 1.56950653e+00 -6.03236020e-01
1.58535635e+00 -1.95752656e+00 -8.03389490e-01 4.20395762e-01
-1.50829315e-01 2.51862288e-01 6.70008898e-01 7.29973614e-01
-2.19277918e-01 1.61265776e-01 -1.57861542e-02 -3.88111591e-01
6.33135259e-01 2.23713726e-01 9.21437666e-02 5.63031912e-01
-3.19195777e-01 6.08357370e-01 -8.01413476e-01 -1.72652975e-01
1.23049223e+00 6.09791517e-01 1.04263343e-01 3.60608786e-01
7.10332319e-02 3.28482628e-01 -3.85231704e-01 5.43427825e-01
4.94216055e-01 1.05584808e-01 5.67746460e-01 -5.00979185e-01
-4.93284017e-01 5.10385990e-01 -1.41021442e+00 1.15219092e+00
-6.95702910e-01 4.08436149e-01 1.66667461e-01 -8.11681569e-01
5.53309381e-01 7.80764759e-01 7.58942664e-01 -1.26567674e+00
4.56874579e-01 9.38400701e-02 -5.76754749e-01 -8.83351624e-01
2.33708352e-01 2.31157005e-01 -1.51175633e-01 7.71551967e-01
-5.09725869e-01 -6.27119243e-02 3.40649426e-01 -2.21989617e-01
1.64441156e+00 5.11794761e-02 9.41700220e-01 -5.51855624e-01
5.32601774e-01 -4.82637048e-01 1.86069921e-01 -7.10469857e-02
-3.29956263e-01 -3.41775447e-01 -8.36860687e-02 -4.15763676e-01
-7.61217117e-01 -1.01083279e+00 -2.07303554e-01 7.88754582e-01
-3.05732131e-01 -5.73298216e-01 -7.71693289e-01 -3.91292095e-01
-1.15498222e-01 1.53210986e+00 -3.18693280e-01 -1.43213898e-01
-4.00363714e-01 -7.27886379e-01 -3.40318941e-02 4.77733552e-01
8.10407877e-01 -7.59114921e-01 -1.87456739e+00 4.50897217e-01
-2.42049918e-01 -8.09109032e-01 -8.47795047e-03 6.05634093e-01
-8.23000371e-01 -1.08172500e+00 2.07364812e-01 -9.79098380e-02
6.66987836e-01 -8.58223885e-02 1.23328650e+00 -1.10259838e-02
-6.03409886e-01 7.38817215e-01 -2.74756998e-01 -5.21761358e-01
-4.85007584e-01 1.08891785e-01 2.35736787e-01 -1.34469494e-01
8.15616906e-01 -1.56175709e+00 -8.74779999e-01 4.01953340e-01
-6.86560154e-01 -3.41022462e-01 -5.21432869e-02 -3.02561909e-01
6.19378574e-02 8.62014771e-01 4.59930122e-01 -3.26127321e-01
6.77561283e-01 -5.53490996e-01 -1.06736994e+00 -3.98824923e-02
-1.16898394e+00 1.26642466e-01 8.62396240e-01 -2.39174366e-01
-8.04214835e-01 1.50356993e-01 4.04243097e-02 5.49978197e-01
-9.54255462e-01 -4.84946549e-01 -7.91784883e-01 4.59047616e-01
1.94637373e-01 -1.35548905e-01 -4.88377064e-01 -7.08647072e-01
-4.04999740e-02 8.75827372e-01 6.24033451e-01 1.06265329e-01
8.04141760e-01 4.80181783e-01 6.40938357e-02 -8.40249479e-01
-7.23452866e-02 -5.43922484e-01 -6.41692996e-01 -2.38758594e-01
7.63219535e-01 -5.25430202e-01 -1.91189861e+00 1.89430743e-01
-8.94947469e-01 -4.58527803e-01 -6.62158370e-01 5.80002293e-02
-4.18530703e-01 2.68767804e-01 9.65208933e-02 -1.31807864e+00
-5.85306346e-01 -8.08747530e-01 8.99556816e-01 3.47842634e-01
-1.05770695e+00 -1.01592302e+00 2.40003616e-01 1.24326169e-01
8.18690717e-01 6.06646717e-01 6.66440845e-01 -5.16929269e-01
-4.43760693e-01 -6.74306810e-01 2.72824049e-01 2.74089903e-01
8.91545594e-01 -2.91761458e-01 -1.45047021e+00 -6.17530763e-01
4.19490278e-01 2.08927184e-01 -2.82222450e-01 -2.97988094e-02
9.26065445e-01 -6.54884279e-01 -5.48586309e-01 1.84153527e-01
1.57772887e+00 4.15584892e-01 6.90498948e-01 3.47657531e-01
4.72161099e-02 3.05765420e-01 1.91905722e-01 9.27239239e-01
4.34297800e-01 8.80970776e-01 6.82825744e-01 -1.22444816e-02
4.50032443e-01 5.23764007e-02 4.68230546e-01 4.31702554e-01
-9.40000564e-02 -2.70303100e-01 -3.75536531e-01 2.49201313e-01
-1.86265326e+00 -1.10968220e+00 -1.89967781e-01 2.32241249e+00
4.28806841e-01 -4.71078567e-02 3.57092023e-01 9.46345031e-01
4.37042087e-01 -2.08715051e-01 -7.36443460e-01 -4.75210041e-01
5.43888569e-01 2.09817976e-01 6.33055747e-01 2.29269296e-01
-7.91638792e-01 -3.78310621e-01 6.31907320e+00 1.31977633e-01
-7.63588548e-01 2.01995522e-01 2.54774660e-01 -2.76928872e-01
2.41912574e-01 -5.59237540e-01 -3.41375142e-01 1.03904212e+00
1.56855071e+00 -4.88694191e-01 3.67250532e-01 1.01536226e+00
9.50695097e-01 -8.95160556e-01 -1.57394528e+00 1.14390421e+00
-4.36303556e-01 -6.11527920e-01 -6.72417045e-01 5.06113529e-01
2.16277495e-01 -1.59434438e-01 -5.96193433e-01 -1.32482171e-01
3.14574540e-01 -6.18848026e-01 5.89587569e-01 5.27966440e-01
4.51133519e-01 -8.71730626e-01 8.00550640e-01 5.33784986e-01
-1.57318866e+00 -2.48157725e-01 3.60755205e-01 -3.81784379e-01
4.44351345e-01 7.05684245e-01 -8.12845588e-01 3.75440657e-01
1.02588546e+00 -1.13206200e-01 -6.55697465e-01 6.13120735e-01
-1.50648296e-01 4.91032690e-01 -9.50478673e-01 -1.61508039e-01
-7.15338886e-01 -2.32355177e-01 1.25354216e-01 1.12468910e+00
5.30313313e-01 1.44405618e-01 -1.80544466e-01 8.17355275e-01
4.79765683e-01 -1.81543112e-01 -5.19278705e-01 4.11502510e-01
5.33570945e-01 1.65799677e+00 -8.37255418e-01 -3.41484457e-01
-4.05658990e-01 1.17289364e+00 -4.33335871e-01 -2.96871066e-02
-5.97878397e-01 -4.78229135e-01 9.33982313e-01 5.21962762e-01
3.38354297e-02 -2.55818039e-01 -1.83274165e-01 -6.55249000e-01
2.51192331e-01 -3.27782869e-01 2.10688129e-01 -5.40658653e-01
-1.00776780e+00 7.55067021e-02 2.42316514e-01 -1.09426010e+00
-5.95955312e-01 -2.53971964e-01 -8.91666532e-01 5.26881576e-01
-1.00078940e+00 -6.19559288e-01 -6.71349823e-01 6.80273890e-01
1.80248186e-01 3.57755035e-01 1.56831229e+00 5.06393373e-01
-4.75198388e-01 2.60050297e-01 5.29037751e-02 -3.09094846e-01
8.99472535e-02 -1.55943716e+00 3.70941430e-01 6.15693450e-01
-4.56120044e-01 -3.30439396e-02 1.18099260e+00 -4.73943472e-01
-1.30762398e+00 -8.30518246e-01 6.47113025e-01 -5.54978728e-01
5.64241350e-01 -7.06318974e-01 -6.58529282e-01 5.39641440e-01
4.23192710e-01 -3.37549269e-01 9.36868012e-01 -2.69361794e-01
1.72894686e-01 -5.34426451e-01 -1.84549677e+00 3.59213412e-01
7.39156842e-01 -3.19308788e-01 -5.33999622e-01 1.38723925e-01
5.09002917e-02 4.36208963e-01 -1.08605599e+00 1.01142585e-01
3.85588765e-01 -1.52259696e+00 4.71884727e-01 3.51775408e-01
-8.92915428e-01 -5.40128350e-01 -2.40429685e-01 -1.25831258e+00
-2.63520896e-01 -9.88113701e-01 -5.10698438e-01 1.69244730e+00
-1.68869570e-01 -1.07235539e+00 4.84866977e-01 1.26182544e+00
2.97752738e-01 2.08670255e-02 -1.22656226e+00 -7.62626648e-01
-8.55664551e-01 -5.68491757e-01 1.05742943e+00 6.27691090e-01
8.80214155e-01 2.85226345e-01 2.43165776e-01 5.03172159e-01
6.20179415e-01 -3.06615233e-01 4.76693720e-01 -1.52911663e+00
4.72531915e-02 -1.73459008e-01 -9.39824343e-01 -2.21475914e-01
-4.90966290e-01 -2.91772962e-01 -1.14623442e-01 -1.66795397e+00
-1.35592893e-01 1.11066252e-01 -1.38064235e-01 5.66370308e-01
5.87770641e-01 3.45910937e-01 -1.30961701e-01 -3.57830673e-01
-4.91070956e-01 2.55272865e-01 4.61306330e-03 -1.06855549e-01
-3.16649169e-01 2.25521654e-01 -2.52672732e-01 7.94608653e-01
1.17372084e+00 -1.05573781e-01 -5.33084273e-01 3.01047325e-01
3.36771637e-01 -6.11294389e-01 3.58857304e-01 -1.65233302e+00
2.56337941e-01 7.92515650e-02 4.78694379e-01 -8.53325009e-01
2.62281120e-01 -1.88954115e+00 9.12183464e-01 8.61142993e-01
4.54096854e-01 2.24455118e-01 1.77638322e-01 2.12273881e-01
5.25497854e-01 5.57190850e-02 7.70349383e-01 6.94023296e-02
-2.51735687e-01 -2.89669216e-01 -9.97639596e-01 -7.72828043e-01
1.51751471e+00 -2.80767560e-01 -4.20072436e-01 -3.27078134e-01
-6.95718944e-01 3.24131131e-01 8.13860059e-01 3.42124999e-01
-1.94126293e-01 -1.34128010e+00 1.53699875e-01 5.24743617e-01
6.87718093e-02 -2.78040081e-01 6.28527030e-02 4.59547430e-01
-2.29766101e-01 4.36812550e-01 -1.64023116e-01 -5.00073612e-01
-1.56043792e+00 7.27823913e-01 4.86790031e-01 -2.56455660e-01
-7.47060895e-01 -2.89084733e-01 -5.85495889e-01 1.34222820e-01
2.67483771e-01 -9.83775914e-01 2.51866281e-02 4.41490233e-01
9.96768355e-01 1.19846213e+00 6.77106738e-01 -3.14848334e-01
-6.27175927e-01 3.18053901e-01 3.85829449e-01 1.67524859e-01
1.47935438e+00 -7.87507117e-01 -7.68201351e-02 9.02855873e-01
1.12952769e+00 -5.89772724e-02 -8.94664824e-01 4.56444055e-01
3.23553175e-01 5.77647537e-02 3.04655135e-02 -1.00384247e+00
-7.20755756e-01 2.57461756e-01 1.34681165e+00 1.43390334e+00
1.60066462e+00 -4.59679589e-02 7.53935516e-01 4.33883667e-01
9.23310578e-01 -1.72064877e+00 -4.94638920e-01 -3.53025883e-01
4.88925040e-01 -9.48661506e-01 2.38513336e-01 -2.02005297e-01
3.72877568e-01 1.09926879e+00 1.11583248e-01 4.31537658e-01
9.39742148e-01 8.18237305e-01 -4.62813042e-02 -1.36188343e-01
-5.49148500e-01 -1.72719643e-01 -5.95563412e-01 1.08536077e+00
1.86847985e-01 4.95079666e-01 5.63941039e-02 5.12485445e-01
-4.70390767e-01 2.08781675e-01 4.08994585e-01 1.25069606e+00
-6.58625484e-01 -1.20880508e+00 -6.56234443e-01 5.58699131e-01
-2.23365650e-01 7.97216356e-01 -4.73487116e-02 6.14675224e-01
4.12823439e-01 1.62377846e+00 2.01603428e-01 -2.52500802e-01
9.12216783e-01 4.36165720e-01 2.55341589e-01 -3.17243636e-01
-5.50018907e-01 -2.62315542e-01 1.16190575e-01 -9.80748951e-01
-3.33861977e-01 -7.16354966e-01 -1.28478980e+00 -7.08378673e-01
6.09416747e-03 -1.11508481e-01 9.93012667e-01 8.95083904e-01
2.83798516e-01 7.84206569e-01 9.48333621e-01 -1.06934142e+00
-3.58433932e-01 -1.06856370e+00 -9.79500771e-01 6.00930214e-01
3.66135657e-01 -2.88376123e-01 -5.86676300e-01 1.36898682e-01] | [5.961048603057861, 2.5485730171203613] |
e0df412d-5945-4e19-a851-8a06a04b7d05 | head-pose-estimation-based-on-multivariate | null | null | http://openaccess.thecvf.com/content_cvpr_2014/html/Geng_Head_Pose_Estimation_2014_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2014/papers/Geng_Head_Pose_Estimation_2014_CVPR_paper.pdf | Head Pose Estimation Based on Multivariate Label Distribution | Accurate ground truth pose is essential to the training of most existing head pose estimation algorithms. However, in many cases, the "ground truth" pose is obtained in rather subjective ways, such as asking the human subjects to stare at different markers on the wall. In such case, it is better to use soft labels rather than explicit hard labels. Therefore, this paper proposes to associate a multivariate label distribution (MLD) to each image. An MLD covers a neighborhood around the original pose. Labeling the images with MLD can not only alleviate the problem of inaccurate pose labels, but also boost the training examples associated to each pose without actually increasing the total amount of training examples. Two algorithms are proposed to learn from the MLD by minimizing the weighted Jeffrey's divergence between the predicted MLD and the ground truth MLD. Experimental results show that the MLD-based methods perform significantly better than the compared state-of-the-art head pose estimation algorithms. | ['Yu Xia', 'Xin Geng'] | 2014-06-01 | null | null | null | cvpr-2014-6 | ['head-pose-estimation'] | ['computer-vision'] | [-1.67586580e-01 3.60871613e-01 -1.18047319e-01 -7.91286707e-01
-9.45030510e-01 -2.27745980e-01 4.97132689e-02 7.14100599e-02
-5.35617769e-01 8.96551788e-01 1.27562225e-01 1.47520006e-01
2.48741999e-01 -2.80022413e-01 -6.96214557e-01 -7.36032188e-01
1.76251993e-01 7.47148514e-01 1.96430668e-01 2.24225730e-01
9.20443460e-02 3.86116683e-01 -1.43669093e+00 -4.07435536e-01
8.07677448e-01 9.58937287e-01 4.60714817e-01 5.59842736e-02
1.58224583e-01 2.12315902e-01 -7.69910336e-01 -5.16195178e-01
8.53486806e-02 -1.95290431e-01 -4.77161229e-01 5.22931576e-01
7.77585268e-01 -2.74070442e-01 -3.66008990e-02 1.33120775e+00
6.02211356e-01 1.57555476e-01 7.42886841e-01 -1.17092776e+00
4.86608818e-02 1.67967528e-01 -7.92558908e-01 -2.77141601e-01
7.37737417e-01 -2.25224450e-01 6.88721061e-01 -9.51075077e-01
5.78336000e-01 1.19538939e+00 6.86475098e-01 5.98630369e-01
-1.21446347e+00 -6.35876000e-01 3.23047906e-01 9.94770452e-02
-1.77470648e+00 -4.75234807e-01 8.67999554e-01 -4.75882947e-01
4.54095080e-02 1.23155005e-01 5.56029677e-01 8.68440092e-01
-8.70902538e-02 9.48360682e-01 1.20080769e+00 -4.50649828e-01
4.29144770e-01 1.87220052e-01 1.48238810e-02 9.26236093e-01
5.01930594e-01 -2.07758099e-01 -6.65786505e-01 -2.18423799e-01
5.78801632e-01 -2.42523283e-01 -5.91449499e-01 -7.70899892e-01
-8.67770135e-01 6.84502602e-01 4.62812781e-01 7.99456984e-02
-3.28098774e-01 -1.58477709e-01 6.30925074e-02 -4.47620064e-01
5.77041566e-01 2.14465454e-01 -2.35128120e-01 1.47986725e-01
-1.02336228e+00 4.29797500e-01 7.59809136e-01 9.51632500e-01
8.30484331e-01 -3.71528804e-01 1.09598637e-02 7.57152140e-01
8.70891929e-01 4.52098936e-01 2.76555568e-01 -8.74321699e-01
5.01203895e-01 4.66778517e-01 4.78771836e-01 -9.84048784e-01
-5.98591566e-01 -4.74126279e-01 -4.30309385e-01 1.83719397e-02
8.97709906e-01 -2.05960467e-01 -9.88015711e-01 1.93289697e+00
6.21733427e-01 4.23543714e-02 -4.78092104e-01 1.27339029e+00
5.61654627e-01 2.75620878e-01 -1.77540593e-02 -3.59686464e-01
1.14984322e+00 -7.47110605e-01 -1.02022004e+00 -7.32075036e-01
6.25656426e-01 -8.93604875e-01 1.01816249e+00 2.66498268e-01
-9.08032477e-01 -2.84735203e-01 -9.64427650e-01 2.69377202e-01
9.74873975e-02 5.11169136e-01 5.01098871e-01 7.90259957e-01
-7.98604131e-01 1.61548987e-01 -8.19571078e-01 -2.06900448e-01
1.82110533e-01 4.78033423e-01 -5.16184568e-01 -1.59626529e-01
-8.55110765e-01 1.05038381e+00 1.80912465e-01 3.65971595e-01
-5.50422728e-01 -2.35663071e-01 -9.94220853e-01 -4.05232400e-01
3.72910947e-01 -4.02107060e-01 1.48719311e+00 -6.17758214e-01
-1.45991945e+00 1.11416876e+00 -5.43834209e-01 -6.89683780e-02
6.34351671e-01 -4.34595317e-01 1.55957952e-01 -2.41071627e-01
2.80821681e-01 7.87139833e-01 9.33852553e-01 -1.71490133e+00
-3.10521960e-01 -7.32753396e-01 -2.37638980e-01 4.08008277e-01
5.58729805e-02 -3.14567596e-01 -8.13028157e-01 -3.90884250e-01
7.21640408e-01 -1.26501346e+00 -1.00739978e-01 1.72402352e-01
-6.01421475e-01 -3.54612440e-01 6.68012977e-01 -7.12203503e-01
9.54303741e-01 -2.16331887e+00 -5.53494096e-02 2.73238778e-01
1.81101888e-01 6.44065142e-02 2.55690187e-01 -2.78625220e-01
1.98398307e-01 -4.80544358e-01 -7.47525841e-02 -8.41808200e-01
1.35100797e-01 3.68425369e-01 5.23511805e-02 9.36890721e-01
-4.31014389e-01 3.83518368e-01 -8.44521821e-01 -7.49364734e-01
1.78366944e-01 4.27125961e-01 -5.24348259e-01 3.66812378e-01
-1.97820678e-01 7.58731425e-01 -2.71445423e-01 3.80372494e-01
8.68081510e-01 -1.67000592e-01 2.84524173e-01 -3.69985014e-01
2.19565034e-01 2.64328182e-01 -1.31151879e+00 1.67669022e+00
-1.28042981e-01 3.69798452e-01 1.62741497e-01 -6.85866594e-01
8.37187171e-01 2.93515354e-01 3.71287704e-01 -2.61296004e-01
3.68245870e-01 3.09619159e-01 -1.70870870e-01 -3.84206712e-01
3.19121152e-01 -2.87869632e-01 -7.91152716e-02 2.47760266e-01
-1.11646660e-01 -9.48947370e-02 -3.21181305e-02 -1.81382060e-01
4.04582351e-01 2.09183767e-01 3.26578736e-01 -1.11980572e-01
4.08455014e-01 -3.97870719e-01 7.15021908e-01 2.42942542e-01
-3.78553987e-01 6.67364657e-01 1.28501520e-01 -1.85085118e-01
-6.62251413e-01 -1.16343057e+00 -3.76005292e-01 9.49563801e-01
2.81256169e-01 -1.89979464e-01 -1.34461069e+00 -7.50843644e-01
-1.14245556e-01 5.69949389e-01 -4.60495234e-01 -2.65262462e-03
-5.12638152e-01 -6.50064826e-01 1.96276456e-01 4.67856318e-01
5.50153971e-01 -6.62456930e-01 -4.17134941e-01 -4.66782302e-02
-6.60554051e-01 -1.19094753e+00 -9.18043256e-01 1.29429400e-01
-7.09218740e-01 -8.38780642e-01 -8.94596934e-01 -8.54905784e-01
1.29578042e+00 1.79648712e-01 9.41896856e-01 -1.55477017e-01
3.13712023e-02 1.53030127e-01 -1.79993495e-01 -4.42821205e-01
6.91428035e-02 7.02019930e-02 3.98351729e-01 2.13609621e-01
4.38578606e-01 -2.01373577e-01 -5.17915368e-01 4.84068632e-01
-3.66704643e-01 -6.69637993e-02 5.70793271e-01 5.93368113e-01
7.41947591e-01 -1.91306561e-01 3.30243349e-01 -6.58052146e-01
3.85414243e-01 -2.14818455e-02 -6.28229082e-01 1.89215958e-01
-4.51998085e-01 3.19868743e-01 9.12576690e-02 -5.53680003e-01
-1.02610779e+00 4.87967432e-01 -8.43368694e-02 -2.55707353e-01
-3.77595067e-01 3.53107125e-01 -5.01411617e-01 -1.10539578e-01
4.38190132e-01 -5.62561341e-02 9.84930247e-02 -6.20242357e-01
1.43144667e-01 6.95300698e-01 7.30023086e-01 -6.11142755e-01
6.65233076e-01 2.03013748e-01 6.05334938e-02 -7.30245829e-01
-1.27467310e+00 -4.82282072e-01 -7.97747016e-01 -4.00826961e-01
9.51303065e-01 -7.65591025e-01 -6.32708311e-01 4.41675097e-01
-1.10044801e+00 -1.15001604e-01 1.38632357e-01 7.69940078e-01
-5.08247018e-01 4.00083333e-01 -2.30048746e-01 -8.74225676e-01
4.21275422e-02 -1.22184074e+00 1.42515445e+00 3.38971645e-01
-5.29494762e-01 -8.73526335e-01 -8.26288685e-02 5.61945856e-01
-1.30318552e-01 1.21348262e-01 6.91945970e-01 -3.93466741e-01
-4.11316872e-01 -8.02589297e-01 1.22678997e-02 5.82263544e-02
6.66247979e-02 -6.24329209e-01 -1.10580194e+00 -4.45898920e-01
1.57374710e-01 -3.40536833e-01 1.90239072e-01 6.99530602e-01
1.03937185e+00 -2.76590407e-01 -4.64372188e-01 4.05874878e-01
9.94983315e-01 -1.43073844e-02 4.41535801e-01 1.80876851e-01
5.87217093e-01 7.32652247e-01 7.56967306e-01 5.21263659e-01
6.66076124e-01 1.10917723e+00 4.28362399e-01 8.27634037e-02
-3.04972213e-02 -5.36035836e-01 3.06067765e-01 6.72047496e-01
1.21659050e-02 -1.10360026e-01 -7.19363987e-01 1.84102014e-01
-1.81293547e+00 -3.84461373e-01 7.72295380e-03 2.63131285e+00
9.24541473e-01 1.10240720e-01 1.41727462e-01 2.45216221e-01
8.81342709e-01 5.81670180e-02 -5.13799369e-01 2.75400490e-01
2.55614728e-01 -2.44816110e-01 4.34938550e-01 7.55441666e-01
-1.13112223e+00 6.08092129e-01 6.19166470e+00 4.89784986e-01
-1.05969560e+00 1.38879761e-01 2.87934214e-01 4.92780767e-02
1.91948250e-01 -2.70255476e-01 -1.18602490e+00 5.51339567e-01
3.77236068e-01 1.44991904e-01 1.97400734e-01 1.08851016e+00
2.32010126e-01 -5.68659663e-01 -1.19789004e+00 1.34225166e+00
3.82452875e-01 -4.04344350e-01 -4.22672033e-01 2.19852924e-01
6.76956415e-01 -4.20513630e-01 1.78677589e-02 3.90314236e-02
-6.31694421e-02 -8.51291418e-01 9.93494391e-01 4.12987709e-01
6.80668890e-01 -5.75290501e-01 9.22853947e-01 6.38722062e-01
-1.14135230e+00 4.06016111e-01 -3.06918472e-01 1.15365356e-01
2.60922104e-01 6.35881186e-01 -1.08381832e+00 1.05179017e-02
4.50091958e-01 1.29048631e-01 -5.84890962e-01 1.38922834e+00
-6.67818189e-01 3.26713920e-01 -5.53100646e-01 -7.23279268e-02
-1.60803661e-01 -7.63002858e-02 5.28007388e-01 7.29647934e-01
1.58040583e-01 -1.37327150e-01 5.95047832e-01 4.86808807e-01
-1.07799828e-01 1.01203687e-01 -4.97250468e-01 5.16850948e-01
6.23007298e-01 1.06541455e+00 -5.43090701e-01 -9.92466882e-02
-2.44433954e-01 9.58370864e-01 2.73180336e-01 3.04221511e-01
-6.75621390e-01 -3.41831818e-02 4.12506849e-01 4.60579395e-01
-1.86503306e-01 -2.80957043e-01 -3.26833427e-01 -8.77932608e-01
3.77441943e-01 -5.31237423e-01 1.45146012e-01 -8.30285072e-01
-1.15251887e+00 3.94431114e-01 2.39381477e-01 -1.24039340e+00
-3.78960043e-01 -4.97519493e-01 -1.83577061e-01 8.60040486e-01
-1.01301897e+00 -8.70933235e-01 -4.99419004e-01 3.19006354e-01
3.44009668e-01 3.22533011e-01 7.93805838e-01 4.42331761e-01
-4.73891646e-01 9.07147706e-01 -2.89396733e-01 9.47300270e-02
8.86621594e-01 -1.23749900e+00 -2.12039039e-01 5.44733644e-01
-9.54606533e-02 6.06488287e-01 1.19297969e+00 -5.80749571e-01
-1.05946052e+00 -8.73126030e-01 9.65932071e-01 -4.00213361e-01
9.74209383e-02 -3.35772485e-01 -9.26128745e-01 8.00763309e-01
-3.36989820e-01 1.76189825e-01 5.69707274e-01 1.91926777e-01
-6.24358617e-02 -1.97630718e-01 -1.33396065e+00 5.10024428e-01
7.77209759e-01 -3.87294024e-01 -5.82841873e-01 4.33429539e-01
4.15163338e-02 -7.89545894e-01 -4.97303307e-01 4.02189583e-01
5.29176593e-01 -6.37073398e-01 7.72712052e-01 -7.70493224e-02
-2.37276807e-01 -5.98195553e-01 -2.37488493e-01 -1.36489820e+00
-1.89160798e-02 -1.84119329e-01 -1.34461999e-01 9.35676694e-01
2.85513341e-01 -4.44133848e-01 1.26470625e+00 9.21713710e-01
-4.44318615e-02 -7.00687528e-01 -1.14974892e+00 -8.14842165e-01
-3.20137978e-01 -3.64330381e-01 4.42108303e-01 5.67831278e-01
-4.48568212e-03 5.48280895e-01 -3.71618062e-01 5.34129560e-01
1.11412919e+00 -8.95191133e-02 8.81602049e-01 -1.42589283e+00
-1.01958914e-03 -1.15471601e-01 -6.74352884e-01 -1.37503994e+00
3.55128676e-01 -6.37873828e-01 7.23112643e-01 -1.42436290e+00
3.87814254e-01 -3.68265063e-01 1.68781072e-01 4.34682399e-01
-2.92348087e-01 1.65956974e-01 3.29884142e-02 9.55390036e-02
-5.26205540e-01 5.32709599e-01 1.29316080e+00 3.22038494e-02
-1.12837173e-01 3.21266145e-01 -3.60115170e-01 1.26235580e+00
4.88142252e-01 -6.14060044e-01 -3.56454253e-01 -2.87652493e-01
-3.40141729e-02 1.61526546e-01 1.99177295e-01 -1.05468214e+00
2.58209467e-01 -2.71024313e-02 4.50934589e-01 -8.72757792e-01
5.70222914e-01 -7.97143638e-01 6.20759674e-05 2.94214427e-01
-5.99402227e-02 -2.79190540e-01 -2.01228112e-01 4.26605463e-01
-1.43796831e-01 -5.45577765e-01 1.10139477e+00 3.01558021e-02
-4.10048127e-01 1.91947937e-01 -9.93166491e-02 3.63683030e-02
8.84078622e-01 -1.78408623e-01 5.01663722e-02 -7.10059166e-01
-8.78223062e-01 1.05484948e-02 5.78883171e-01 1.57378078e-01
5.43733239e-01 -1.38637614e+00 -2.47289032e-01 3.40559274e-01
1.42658114e-01 3.25163841e-01 -1.86385721e-01 8.92718792e-01
-2.54948795e-01 3.76003563e-01 1.79085299e-01 -6.99083686e-01
-1.31747234e+00 2.32828289e-01 4.27085608e-01 -3.72576341e-02
-3.01836997e-01 8.94155264e-01 2.99007505e-01 -5.49344540e-01
7.48149574e-01 -1.42438143e-01 -1.56174183e-01 9.33146179e-02
3.52797151e-01 2.51518577e-01 5.67033030e-02 -1.00821841e+00
-4.63649780e-01 7.55166173e-01 -9.72344279e-02 -2.44428366e-01
9.81480539e-01 -2.97921091e-01 1.18639991e-01 3.50735337e-01
1.20613301e+00 4.39691066e-04 -1.32576680e+00 -2.21847624e-01
1.02042057e-01 -6.86247408e-01 3.70754823e-02 -6.75193369e-01
-1.01308107e+00 7.84813285e-01 9.35567141e-01 -3.79652083e-01
9.02472198e-01 1.84413478e-01 7.55814731e-01 3.58339876e-01
9.37953889e-01 -1.17377603e+00 2.23253861e-01 2.69853294e-01
7.57170677e-01 -1.30941975e+00 1.94440797e-01 -7.20733047e-01
-4.70437616e-01 7.16383576e-01 7.70128250e-01 9.12211835e-02
5.33819258e-01 1.49552092e-01 2.32076615e-01 -5.42157106e-02
-7.20725488e-03 -1.28783986e-01 5.25443912e-01 5.76515913e-01
5.56769133e-01 3.64432126e-01 -4.94735897e-01 5.47510266e-01
-4.23914552e-01 1.48434928e-02 1.47374868e-01 9.89534020e-01
-6.80967689e-01 -1.18859100e+00 -8.91490042e-01 3.00374240e-01
-3.98536265e-01 4.40797985e-01 -2.65780717e-01 5.32964826e-01
2.37590700e-01 7.83694327e-01 -1.13663614e-01 -2.25193277e-01
4.10807312e-01 1.80108279e-01 8.38985682e-01 -6.94719434e-01
2.45004967e-01 4.03239965e-01 -9.72840488e-02 -2.14792415e-01
-3.28492403e-01 -7.88890243e-01 -1.34784460e+00 -8.82102270e-03
-6.72761083e-01 1.40040502e-01 6.90702558e-01 1.11278224e+00
-2.19414294e-01 -1.53453555e-02 4.98842120e-01 -1.36271679e+00
-6.46744549e-01 -1.18418825e+00 -7.10692763e-01 5.40507495e-01
2.64393926e-01 -1.23737085e+00 -4.69208568e-01 1.20554782e-01] | [7.061539173126221, -1.100740671157837] |
a921f0da-a827-4624-b5d4-257ade4d2b4b | forgery-attack-detection-in-surveillance | 2201.09487 | null | https://arxiv.org/abs/2201.09487v1 | https://arxiv.org/pdf/2201.09487v1.pdf | Forgery Attack Detection in Surveillance Video Streams Using Wi-Fi Channel State Information | The cybersecurity breaches expose surveillance video streams to forgery attacks, under which authentic streams are falsified to hide unauthorized activities. Traditional video forensics approaches can localize forgery traces using spatial-temporal analysis on relatively long video clips, while falling short in real-time forgery detection. The recent work correlates time-series camera and wireless signals to detect looped videos but cannot realize fine-grained forgery localization. To overcome these limitations, we propose Secure-Pose, which exploits the pervasive coexistence of surveillance and Wi-Fi infrastructures to defend against video forgery attacks in a real-time and fine-grained manner. We observe that coexisting camera and Wi-Fi signals convey common human semantic information and forgery attacks on video streams will decouple such information correspondence. Particularly, retrievable human pose features are first extracted from concurrent video and Wi-Fi channel state information (CSI) streams. Then, a lightweight detection network is developed to accurately discover forgery attacks and an efficient localization algorithm is devised to seamlessly track forgery traces in video streams. We implement Secure-Pose using one Logitech camera and two Intel 5300 NICs and evaluate it in different environments. Secure-Pose achieves a high detection accuracy of 98.7% and localizes abnormal objects under playback and tampering attacks. | ['Qian Zhang', 'Tao Jiang', 'Wei Wang', 'Xiang Li', 'Yong Huang'] | 2022-01-24 | null | null | null | null | ['video-forensics'] | ['computer-vision'] | [ 3.80979866e-01 -6.10620618e-01 -2.01307699e-01 1.57720432e-01
-7.38046885e-01 -1.10611081e+00 1.74123093e-01 -4.18521702e-01
-2.16366500e-01 2.62548089e-01 -1.68423176e-01 -5.31027555e-01
-1.11403592e-01 -6.35524929e-01 -8.01133454e-01 -6.16660357e-01
-5.74383438e-01 -6.16309464e-01 5.17414868e-01 2.30011955e-01
1.46800831e-01 5.25288820e-01 -1.29997897e+00 4.26215857e-01
7.33183371e-03 1.39044321e+00 -2.40779340e-01 1.32987261e+00
6.01685226e-01 9.33203995e-01 -1.14343047e+00 -4.22012478e-01
6.16440535e-01 -7.67180175e-02 -1.78957254e-01 1.10379457e-01
3.34864557e-01 -9.94338751e-01 -1.21330142e+00 1.15634966e+00
2.13095769e-01 -4.07047838e-01 -1.63474903e-01 -1.90290987e+00
-5.98067865e-02 2.56136298e-01 -5.93078732e-01 8.18186522e-01
9.80915785e-01 4.02719796e-01 2.25079492e-01 -3.07009518e-01
3.61819655e-01 1.04762745e+00 1.03875101e+00 2.13813066e-01
-3.84025902e-01 -1.25537741e+00 -2.18410164e-01 4.74014938e-01
-1.72029531e+00 -5.76788723e-01 8.17616105e-01 -1.20042088e-02
5.72686315e-01 5.65948546e-01 4.07886505e-01 1.48333037e+00
4.04916734e-01 4.86880749e-01 4.64943856e-01 -2.39321351e-01
-1.25246108e-01 -1.52378798e-01 -2.01565623e-02 7.04042912e-01
6.51712239e-01 6.23137116e-01 -8.73079419e-01 -8.28145385e-01
7.68436015e-01 4.71927881e-01 -8.06142449e-01 2.84868151e-01
-1.39235461e+00 2.48294145e-01 -4.01558608e-01 2.32821137e-01
-3.67243409e-01 4.12221044e-01 6.87602878e-01 6.92729175e-01
-2.15969175e-01 1.11571491e-01 -6.67814240e-02 -3.85929614e-01
-8.28041434e-01 4.78049181e-02 7.27201402e-01 1.27241492e+00
3.44855636e-01 2.29748532e-01 1.98882550e-01 -1.53740466e-01
2.62222350e-01 9.93924499e-01 3.92773002e-01 -1.12090492e+00
5.56510806e-01 -8.39336067e-02 2.05490187e-01 -1.95843327e+00
-8.41768458e-02 -2.08220735e-01 -4.70359206e-01 -1.82406232e-01
4.91936415e-01 -4.29739922e-01 -6.58696294e-02 1.45525074e+00
2.35013202e-01 1.07009602e+00 -1.14817962e-01 9.73890185e-01
2.43131608e-01 4.12336767e-01 8.12562555e-03 -3.07416707e-01
1.51647949e+00 -1.85105711e-01 -9.04668689e-01 9.40461084e-02
3.79416168e-01 -6.14071250e-01 3.67386162e-01 8.10116231e-01
-8.09163928e-01 -4.82383043e-01 -1.20093358e+00 9.19871449e-01
-2.85120666e-01 -3.00146341e-01 4.27544147e-01 1.53168535e+00
-6.07329249e-01 2.14515626e-01 -7.08359659e-01 -3.34088840e-02
3.78134787e-01 3.27839822e-01 -5.04290700e-01 -1.92485392e-01
-1.59664762e+00 2.08463952e-01 3.30367655e-01 9.08273980e-02
-1.10240328e+00 -5.59595108e-01 -7.66357064e-01 -1.99942023e-01
6.60664022e-01 -2.27450743e-01 9.22238767e-01 -9.28938448e-01
-9.89841521e-01 6.18804932e-01 2.13542178e-01 -6.14572346e-01
5.55896103e-01 6.24951757e-02 -1.44764507e+00 1.04828072e+00
6.71693236e-02 -3.19039911e-01 1.40798426e+00 -9.08958912e-01
-9.38646913e-01 -4.79022235e-01 -4.02308144e-02 -6.12546027e-01
-6.59688354e-01 3.31167758e-01 -3.04333687e-01 -9.38287735e-01
5.92547096e-02 -6.79736376e-01 3.77363235e-01 -1.18093546e-02
-2.37504274e-01 6.30220175e-01 1.45112836e+00 -7.42817879e-01
1.51310575e+00 -2.17284632e+00 -7.05823302e-01 4.83859569e-01
3.73440117e-01 5.98668635e-01 3.61312963e-02 3.14317852e-01
1.05010174e-01 -1.30967963e-02 2.79175192e-01 1.63982466e-01
-2.72702016e-02 7.39818662e-02 -7.13372469e-01 1.26699328e+00
-3.12353164e-01 5.29568553e-01 -1.03122807e+00 -2.38929912e-01
2.90872723e-01 2.20224217e-01 -2.73870349e-01 -1.26420474e-02
4.55952883e-01 1.56098366e-01 -5.72883248e-01 1.13387239e+00
1.07554281e+00 1.64467111e-01 2.73480117e-01 -4.02121305e-01
2.89919581e-02 -3.54772896e-01 -1.27999783e+00 1.33695757e+00
6.07186519e-02 8.37633967e-01 6.06432140e-01 -8.40163589e-01
5.92996836e-01 8.60230505e-01 6.32271469e-01 -4.90854174e-01
4.67462897e-01 -2.44551548e-03 -6.90433681e-01 -8.53766501e-01
4.63723034e-01 4.56751853e-01 -5.87703645e-01 5.73710680e-01
-4.44094725e-02 9.28314567e-01 -3.79536420e-01 1.28151670e-01
2.06515121e+00 -2.59217829e-01 3.14307921e-02 1.73708662e-01
5.99790335e-01 -2.28949264e-01 5.59844375e-01 1.16951573e+00
-6.49005055e-01 7.77179748e-02 3.20033245e-02 -3.38458896e-01
-4.53753740e-01 -1.35259330e+00 1.96878344e-01 9.28419411e-01
6.17134094e-01 -6.96503460e-01 -7.67436922e-01 -9.09171402e-01
9.21633281e-03 1.38389111e-01 -9.60569382e-02 -5.32748640e-01
-7.01957405e-01 -3.70919824e-01 1.71663904e+00 2.35427469e-01
7.99342394e-01 -3.98012102e-01 -9.36132312e-01 3.93176824e-01
-3.27784568e-01 -1.85473871e+00 -4.99568313e-01 -5.49629331e-01
-3.14959139e-01 -1.74223828e+00 -5.90292178e-02 -3.84975404e-01
1.94845796e-01 1.10196006e+00 5.05648911e-01 2.77544856e-01
-6.89495385e-01 1.09309006e+00 -5.01224637e-01 -4.89456169e-02
-3.89290005e-01 -6.73887253e-01 4.40154105e-01 4.49994504e-01
7.09005117e-01 -5.01552701e-01 -4.34245437e-01 7.29849160e-01
-1.10630727e+00 -8.60104680e-01 7.98946545e-02 3.10463369e-01
-1.06049299e-01 7.65317380e-01 4.05027360e-01 -1.14580512e-01
3.66775572e-01 -7.89244831e-01 -6.55961931e-01 1.86446741e-01
-1.68511290e-02 -8.19058657e-01 6.10798240e-01 -6.86073124e-01
-8.00121427e-01 -2.70388603e-01 1.72950968e-01 -8.85586262e-01
-5.64701259e-01 8.41486268e-03 -4.64176774e-01 -6.60835922e-01
5.80729723e-01 4.31168973e-01 -1.08970724e-01 -7.62439296e-02
-1.52683668e-02 9.54285264e-01 1.26890564e+00 -3.05072844e-01
1.11021554e+00 9.03495133e-01 -1.46715850e-01 -1.15313852e+00
-1.51004821e-01 -6.69008136e-01 -2.01680511e-01 -7.73771644e-01
4.32509929e-01 -1.00987303e+00 -1.34335697e+00 9.27948654e-01
-1.30784380e+00 4.45736676e-01 4.95356590e-01 5.12336075e-01
-2.17416987e-01 1.32093716e+00 -8.62506092e-01 -8.83097291e-01
-7.04853535e-02 -7.81491578e-01 1.00335085e+00 -2.05205128e-01
-1.64475769e-01 -7.23679185e-01 -3.04420918e-01 4.65057731e-01
2.21683726e-01 6.50871515e-01 -1.61996588e-01 -4.73614365e-01
-7.90194929e-01 -1.02496624e+00 -1.30016193e-01 -3.88360210e-02
2.13016495e-01 -2.48019606e-01 -1.08847475e+00 -3.72280836e-01
5.80094874e-01 6.26547262e-02 2.59920925e-01 1.38177082e-01
1.03111684e+00 -6.18947208e-01 -5.83454788e-01 9.95648980e-01
1.02120316e+00 3.42076659e-01 9.88431990e-01 3.73931736e-01
6.60057902e-01 3.62009525e-01 6.67874098e-01 1.01267850e+00
1.25640154e-01 5.18329918e-01 5.73783100e-01 4.61592495e-01
3.73374343e-01 -2.02039525e-01 7.83985436e-01 1.44492552e-01
2.11299151e-01 -5.56198120e-01 -4.93959516e-01 1.04833350e-01
-1.61190343e+00 -1.55364418e+00 -1.71034113e-01 2.25459504e+00
1.06725290e-01 1.87341318e-01 2.73128390e-01 5.86873293e-01
1.07182670e+00 2.66816109e-01 -1.54640481e-01 2.85003543e-01
-2.14619055e-01 -2.44878799e-01 1.13689315e+00 2.32876375e-01
-1.32007480e+00 6.40721679e-01 5.57293272e+00 8.82820845e-01
-9.52185631e-01 3.84810776e-01 1.44665493e-02 -2.05718651e-01
2.60481894e-01 -2.56464154e-01 -5.05999982e-01 8.25370848e-01
1.24711764e+00 2.98054278e-01 2.17661798e-01 6.14426374e-01
3.90497237e-01 4.69766278e-03 -6.50611103e-01 1.47960281e+00
3.74251574e-01 -1.40599036e+00 -3.96711171e-01 1.98028192e-01
-1.89440176e-01 -5.37526309e-01 -3.62046622e-02 -2.38485232e-01
-1.30733654e-01 -5.45620322e-01 9.01956558e-01 4.49113160e-01
7.61640549e-01 -8.72468889e-01 5.59749663e-01 2.83044189e-01
-1.58581901e+00 -2.98921615e-01 -1.84290439e-01 -2.58772261e-02
4.86352265e-01 2.56073862e-01 -3.51422817e-01 4.78987694e-01
9.12262619e-01 5.78506172e-01 -2.49251410e-01 8.19574237e-01
1.59474865e-01 9.16600704e-01 -4.33933258e-01 3.84939224e-01
1.89617261e-01 2.59738654e-01 1.01673913e+00 1.32171357e+00
6.26260757e-01 3.78080994e-01 1.89725250e-01 5.81883013e-01
2.23675624e-01 -7.98449099e-01 -8.08046401e-01 1.95407644e-01
8.85281026e-01 8.05444956e-01 -5.08589268e-01 -5.63460290e-02
-3.00122738e-01 1.16566443e+00 -7.24633336e-01 3.99499238e-01
-1.28890419e+00 -8.34282339e-01 9.29454744e-01 1.67807713e-01
2.76994437e-01 -5.76434433e-01 3.90575588e-01 -1.31410694e+00
1.93078965e-01 -9.25210178e-01 6.55210793e-01 -4.12811339e-01
-1.12703621e+00 2.60316461e-01 -1.29422039e-01 -1.60365677e+00
-2.35328615e-01 -5.16734362e-01 -5.05936027e-01 1.31633952e-01
-1.18280697e+00 -1.11754799e+00 -4.85695809e-01 1.43729258e+00
1.50218308e-01 -3.96140307e-01 6.33258104e-01 5.46275914e-01
-5.76781213e-01 9.86048698e-01 -1.85825318e-01 6.99264646e-01
4.78002578e-01 -2.60149360e-01 4.36759889e-01 1.31428242e+00
-8.58529378e-03 4.75600302e-01 6.37314200e-01 -9.46223795e-01
-2.20472002e+00 -1.00777388e+00 3.00717831e-01 -3.24958652e-01
8.93549919e-01 -3.27721745e-01 -7.37758577e-01 5.65143526e-01
-2.14190841e-01 3.31295282e-01 7.42685974e-01 -9.15413201e-01
-6.92263603e-01 -8.71625990e-02 -1.34016216e+00 2.20063075e-01
9.82039750e-01 -8.60194325e-01 -4.63672698e-01 7.98830949e-03
5.75144947e-01 -6.97603002e-02 -7.03151345e-01 -3.11447717e-02
8.72574449e-01 -1.20538092e+00 1.29226875e+00 -2.77478456e-01
-6.29146755e-01 -4.21652257e-01 -4.67746228e-01 -3.10853809e-01
9.23111439e-02 -1.49743688e+00 -5.59838235e-01 1.11563277e+00
-4.69099522e-01 -6.06773376e-01 8.94426227e-01 2.48735711e-01
2.73679256e-01 3.41087103e-01 -1.17507255e+00 -1.12191224e+00
-8.71046484e-01 -1.20474505e+00 7.40027189e-01 1.00848079e+00
3.48540694e-01 -6.69208944e-01 -9.95686114e-01 9.83750165e-01
1.26170146e+00 -4.46376026e-01 6.96488678e-01 -7.31486082e-01
-3.19755554e-01 1.06494546e-01 -9.90349591e-01 -8.88790309e-01
5.37120663e-02 -6.67182654e-02 -2.88402170e-01 -4.26000059e-02
-4.53287363e-01 -1.37957156e-01 -9.95173529e-02 4.80960086e-02
4.45054203e-01 4.32938576e-01 -7.83414580e-03 2.05741107e-01
-8.84753823e-01 -2.38874853e-02 1.91773459e-01 -3.22479829e-02
1.16301209e-01 3.01677018e-01 -1.21956497e-01 8.39716911e-01
6.48743212e-01 -5.42955101e-01 -3.07074308e-01 -1.00099914e-01
-9.10800248e-02 8.72221887e-01 1.10171103e+00 -1.51918960e+00
5.42335689e-01 3.09685953e-02 4.21567976e-01 -3.94899100e-01
2.24888250e-01 -1.42057455e+00 2.86716104e-01 6.37754977e-01
1.09977931e-01 2.72974789e-01 1.99289382e-01 1.14947259e+00
-4.11069356e-02 4.22625132e-02 3.63371789e-01 -3.65356915e-02
-8.84394050e-01 3.72144580e-01 -1.03278172e+00 -2.61916965e-01
1.20940459e+00 -6.85923457e-01 -4.65758115e-01 -5.54129660e-01
-2.73813426e-01 -2.43500412e-01 4.50713664e-01 4.16164786e-01
1.09451306e+00 -1.21660340e+00 -2.26534590e-01 4.70229119e-01
5.84177114e-02 -9.75856245e-01 5.04915833e-01 5.13173580e-01
-6.13978505e-01 3.22917730e-01 -2.51400948e-01 -5.07505715e-01
-1.58901131e+00 8.33741009e-01 2.55560756e-01 2.92709589e-01
-6.65010452e-01 4.27377760e-01 -4.72728461e-01 4.09641951e-01
3.99154276e-01 8.88897479e-02 3.35717946e-01 -2.10821494e-01
1.37446928e+00 9.14767563e-01 -7.05325007e-02 -9.14717376e-01
-6.93358541e-01 5.27297378e-01 4.08843696e-01 -1.25455067e-01
6.20927572e-01 -7.76938736e-01 2.31979802e-01 -4.60695922e-01
1.43009543e+00 4.52382639e-02 -1.36437535e+00 -3.40308398e-01
8.28626677e-02 -1.13158214e+00 2.08831817e-01 -1.43375739e-01
-1.23244238e+00 3.47195119e-01 6.63898110e-01 4.27527279e-01
1.24439371e+00 -2.59849757e-01 1.37191033e+00 2.94626504e-01
1.00825500e+00 -6.18641078e-01 1.75202698e-01 1.06864823e-02
9.50383022e-02 -8.11850846e-01 -2.66525060e-01 -4.38767403e-01
-2.10460961e-01 1.26827788e+00 1.75148323e-01 -1.86164573e-01
6.44219697e-01 8.19887519e-01 -1.85153246e-01 -2.70861953e-01
-3.37436438e-01 3.57738435e-01 -1.93762913e-01 1.10700488e+00
-5.56679428e-01 -1.90404236e-01 5.83894551e-01 1.00178862e+00
3.68380686e-03 -2.98995581e-02 5.23895860e-01 1.21893203e+00
-5.45412481e-01 -4.38663214e-01 -1.19739497e+00 -5.36769144e-02
-8.67120445e-01 4.15471256e-01 -7.31462091e-02 6.39714837e-01
6.95749149e-02 1.70105052e+00 -1.09178782e-01 -9.68983650e-01
8.61657783e-02 -4.06201988e-01 2.28503361e-01 3.12019914e-01
-5.13031721e-01 1.29602477e-01 -1.72091350e-02 -1.25711334e+00
-2.84576327e-01 -5.81619024e-01 -7.59103119e-01 -8.63254070e-01
-2.34968349e-01 4.37237062e-02 3.61380368e-01 8.55970383e-01
3.03418219e-01 2.16963470e-01 8.49124253e-01 -7.38487542e-01
-4.74702775e-01 -1.57660961e-01 -9.68615055e-01 3.08190107e-01
7.36815810e-01 -5.83807707e-01 -7.13920772e-01 8.99042711e-02] | [12.594165802001953, 1.0755265951156616] |
7192bd53-ef15-4fc5-93ee-da07403340e3 | building-blocks-of-a-task-oriented-dialogue | null | null | https://aclanthology.org/2021.nlpmc-1.7 | https://aclanthology.org/2021.nlpmc-1.7.pdf | Building blocks of a task-oriented dialogue system in the healthcare domain | There has been significant progress in dialogue systems research. However, dialogue systems research in the healthcare domain is still in its infancy. In this paper, we analyse recent studies and outline three building blocks of a task-oriented dialogue system in the healthcare domain: i) privacy-preserving data collection; ii) medical knowledge-grounded dialogue management; and iii) human-centric evaluations. To this end, we propose a framework for developing a dialogue system and show preliminary results of simulated dialogue data generation by utilising expert knowledge and crowd-sourcing. | ['Bart Vanrumste', 'Stijn Luca', 'Dietwig Lowet', 'Heereen Shim'] | null | null | null | null | naacl-nlpmc-2021-6 | ['dialogue-management'] | ['natural-language-processing'] | [-3.55161689e-02 1.38380003e+00 1.63169235e-01 -8.33695650e-01
-7.82672644e-01 -3.19618374e-01 9.23588097e-01 6.15360975e-01
-4.23777610e-01 1.21577442e+00 9.49080169e-01 -3.53407770e-01
-5.11333235e-02 -4.04404104e-01 2.68032819e-01 -1.74660519e-01
1.40386060e-01 1.04999053e+00 1.53131053e-01 -7.24457324e-01
5.53014018e-02 9.29969400e-02 -9.57632720e-01 5.88687658e-01
9.59451616e-01 4.56255972e-01 -4.04572576e-01 9.66709733e-01
-2.37232968e-01 1.46101892e+00 -8.93575966e-01 -6.09304070e-01
5.18019311e-02 -9.97823417e-01 -1.65666485e+00 1.07811093e-01
-4.38482881e-01 -5.33956766e-01 1.42721623e-01 6.33795917e-01
1.01350284e+00 1.05080530e-02 3.45230997e-01 -1.37006724e+00
-2.64476568e-01 3.84916246e-01 5.31487942e-01 -2.23776668e-01
1.11027825e+00 2.89032727e-01 3.71304363e-01 -2.82322973e-01
1.04166174e+00 1.23511946e+00 5.73184669e-01 1.21933043e+00
-9.60572243e-01 6.77799657e-02 -4.09308493e-01 -3.09531659e-01
-1.03332663e+00 -8.31474900e-01 4.42847461e-01 -5.40318131e-01
1.03061748e+00 6.00290239e-01 8.23192298e-01 1.00610614e+00
1.35124817e-01 7.52546430e-01 1.21610153e+00 -7.28480279e-01
5.46349406e-01 1.04710209e+00 1.44126177e-01 5.31881571e-01
-1.62359759e-01 -2.01661035e-01 -6.27463996e-01 -7.42448390e-01
6.46712363e-01 -6.60478354e-01 -2.38223508e-01 -3.82099211e-01
-1.20317018e+00 1.21489465e+00 -2.40739241e-01 3.81526500e-01
-7.04879224e-01 -6.52897894e-01 8.89735103e-01 5.75603843e-01
8.14970255e-01 7.86778748e-01 -3.88726413e-01 -3.45566154e-01
-5.97593307e-01 7.12157130e-01 2.03658152e+00 1.16271472e+00
6.97067827e-02 -4.85629797e-01 -5.97811699e-01 8.32209051e-01
6.05614662e-01 -4.21911441e-02 3.48851144e-01 -1.14300108e+00
2.61262685e-01 5.75752914e-01 5.57683945e-01 -7.83748567e-01
-6.69870079e-01 7.30744660e-01 -5.32956243e-01 -1.96584940e-01
4.83133405e-01 -8.50324750e-01 -1.63298935e-01 1.27181733e+00
7.99524724e-01 -7.25241244e-01 9.96116281e-01 8.06563437e-01
1.44139254e+00 3.44165504e-01 3.19789737e-01 -5.73068321e-01
1.62773407e+00 -9.45921361e-01 -1.50283873e+00 2.28033975e-01
7.10440159e-01 -5.75973094e-01 4.03804421e-01 2.18175143e-01
-1.47159278e+00 1.66742336e-02 -3.34200412e-01 -1.92145541e-01
-3.38576823e-01 -5.66714764e-01 4.58299518e-01 1.01505017e+00
-1.46414959e+00 -5.74030392e-02 -4.70528185e-01 -9.70114052e-01
-1.22732958e-02 9.99450684e-02 -3.18332672e-01 3.77070278e-01
-1.70631778e+00 1.38918018e+00 3.37411344e-01 -1.34910911e-01
-3.81146312e-01 -4.32479560e-01 -8.84098470e-01 -4.41949099e-01
3.53731930e-01 -1.15615952e+00 1.82605410e+00 -7.48648345e-01
-1.95144367e+00 1.16823792e+00 1.46696895e-01 -7.69017458e-01
1.15933216e+00 -1.78583503e-01 -2.48021141e-01 3.33976716e-01
-1.63874235e-02 5.91693759e-01 -5.87536767e-02 -1.32412732e+00
-7.53135324e-01 -2.22608998e-01 -3.20309363e-02 6.87279224e-01
1.94767013e-01 5.38636327e-01 -9.13890153e-02 -1.98207781e-01
-7.88817644e-01 -5.23941159e-01 -7.21665680e-01 -7.21740723e-02
-5.58857501e-01 -2.92046636e-01 1.48376673e-01 -9.24566567e-01
1.27920830e+00 -1.60719478e+00 -1.79001033e-01 2.82125473e-02
2.86073416e-01 6.56955600e-01 3.69719803e-01 1.23195124e+00
4.50406015e-01 9.69441235e-02 -2.56608993e-01 -3.36758554e-01
1.21371336e-01 4.30916786e-01 3.30551453e-02 -1.97519995e-02
2.74390187e-02 8.35784078e-01 -1.11511993e+00 -1.00008333e+00
3.96383554e-01 3.38980675e-01 -3.94783884e-01 6.90945208e-01
-3.29839110e-01 8.04542601e-01 -6.85327470e-01 3.64271134e-01
1.16398945e-01 2.50953466e-01 4.09848779e-01 2.13651672e-01
-3.39388609e-01 1.98449656e-01 -8.73465776e-01 1.61289346e+00
-1.30695134e-01 1.93543643e-01 7.74372816e-01 -6.50654078e-01
9.56356347e-01 1.12614620e+00 7.05670476e-01 -4.43221539e-01
2.15644747e-01 1.08857319e-01 -8.50553289e-02 -1.35496724e+00
6.76590264e-01 -3.14548731e-01 -1.88957632e-01 7.51411438e-01
-1.02789901e-01 -3.27282757e-01 -2.17800274e-01 2.00323224e-01
1.05626941e+00 -1.45940065e-01 8.09099138e-01 -3.45478773e-01
7.25769937e-01 6.04956746e-01 2.18523115e-01 7.33022690e-01
-7.60366619e-01 3.50592315e-01 5.22617757e-01 -6.64402127e-01
-7.79725194e-01 -2.94952571e-01 -8.24072435e-02 8.37881565e-01
-6.78146929e-02 -4.48742509e-01 -1.40819299e+00 -8.81084621e-01
-6.60597682e-02 8.24468434e-01 -4.19251323e-01 2.85813034e-01
-3.41330647e-01 -5.57732940e-01 8.52144122e-01 4.04828303e-02
5.59316576e-01 -1.48895681e+00 -1.23753631e+00 5.33618569e-01
-4.90451217e-01 -9.24887300e-01 -1.68556631e-01 -2.15385586e-01
-6.36415482e-01 -1.25157452e+00 -9.41370070e-01 -6.81639314e-01
4.93799508e-01 -3.29289913e-01 1.20876002e+00 -6.66949078e-02
-3.01292151e-01 9.92628515e-01 -5.94713211e-01 -7.91112840e-01
-1.05711913e+00 -7.11244578e-03 -3.58105570e-01 -2.65827566e-01
5.32190621e-01 2.12862000e-01 -7.76753068e-01 3.67181271e-01
-8.77852380e-01 1.55590743e-01 1.74220830e-01 9.50248361e-01
-4.81445007e-02 -4.39077437e-01 9.02164221e-01 -1.52312815e+00
1.79164386e+00 -5.28375208e-01 4.33448926e-02 4.18152541e-01
-4.90213901e-01 -3.54866028e-01 1.07364550e-01 2.41968855e-01
-1.50148189e+00 2.75089532e-01 -5.82413256e-01 5.23896039e-01
-6.45651042e-01 4.35679525e-01 -1.28349677e-01 1.32994920e-01
9.80290532e-01 -7.33727664e-02 8.04174244e-01 -2.69510925e-01
6.16988420e-01 1.14201987e+00 2.06763193e-01 -5.02436340e-01
-1.40167251e-01 2.89510310e-01 -7.96297729e-01 -9.50925887e-01
-9.77163166e-02 -5.03794491e-01 -5.02302110e-01 -3.85672092e-01
1.22742665e+00 -7.47654259e-01 -1.04925704e+00 2.14279518e-01
-1.21550643e+00 -5.93135953e-01 -7.54017174e-01 2.01284423e-01
-8.21971476e-01 3.02521169e-01 -5.68206847e-01 -1.45464003e+00
-7.95744479e-01 -8.11917126e-01 9.13226545e-01 6.43509775e-02
-1.05543017e+00 -1.34923708e+00 5.64567685e-01 7.13982046e-01
5.82971275e-01 6.56049728e-01 4.58384335e-01 -1.34223640e+00
2.12387398e-01 -3.00066978e-01 9.54804271e-02 7.15247169e-02
5.22225909e-02 -3.35845321e-01 -8.94880712e-01 2.16934718e-02
4.26682860e-01 -7.12686777e-01 -2.11418852e-01 1.27084464e-01
1.10998705e-01 -7.13354349e-01 -4.12940830e-01 -3.65297675e-01
7.82998383e-01 6.21838272e-01 5.17788291e-01 1.89357936e-01
5.41751161e-02 1.60554981e+00 1.00755477e+00 7.86320269e-01
1.07128465e+00 6.04110301e-01 -1.88073128e-01 -4.58961487e-01
2.65874445e-01 -1.05599515e-01 -1.96275666e-01 4.93852556e-01
1.14103228e-01 -1.71517685e-01 -1.28784859e+00 8.78830254e-01
-2.26567411e+00 -6.64279997e-01 -1.38689920e-01 1.63380635e+00
1.26369178e+00 -4.28719729e-01 4.84183729e-01 -1.39655426e-01
5.23146629e-01 -1.84546664e-01 -2.13893965e-01 -9.71924663e-01
3.75860006e-01 -1.44570336e-01 -2.30178796e-02 7.27342904e-01
-9.95976269e-01 7.48914361e-01 7.24119616e+00 -1.21065155e-01
-2.15419412e-01 1.69146135e-01 6.77721500e-01 2.87046492e-01
-2.63806552e-01 -2.55968124e-01 -3.11724901e-01 1.64766401e-01
1.05720270e+00 -4.81287360e-01 -5.01770228e-02 7.18558729e-01
3.95509630e-01 -5.22405446e-01 -1.19508481e+00 5.40659249e-01
9.49099809e-02 -1.13351953e+00 -1.86495513e-01 1.26264662e-01
2.07468882e-01 -4.57951635e-01 -6.39135301e-01 2.01207757e-01
7.81917810e-01 -9.20029223e-01 2.65315205e-01 7.18397677e-01
4.08194661e-01 -5.66768348e-01 1.19546509e+00 7.49974132e-01
-2.47257546e-01 2.30071113e-01 1.47864267e-01 4.83659692e-02
6.81330800e-01 3.55249912e-01 -1.47376812e+00 8.89208674e-01
4.11559641e-01 1.28985703e-01 7.47683421e-02 7.92936146e-01
5.68378903e-02 9.38358903e-02 9.63276252e-02 -3.11294049e-01
8.09409991e-02 -2.09742785e-01 3.81106466e-01 1.61005819e+00
-4.07124221e-01 6.98520005e-01 4.00186688e-01 4.16694492e-01
3.10367554e-01 5.46940804e-01 -9.78601217e-01 1.43103808e-01
4.23985481e-01 9.99654293e-01 -3.62989008e-01 -4.19496953e-01
-3.78358960e-01 9.18361068e-01 -1.42883435e-01 2.99053658e-02
-2.15683326e-01 -3.97942394e-01 4.21036452e-01 -3.54259647e-02
-3.65154654e-01 4.24124360e-01 -2.56252885e-01 -8.81509960e-01
-2.64837086e-01 -1.34083474e+00 6.44675016e-01 -3.31359923e-01
-1.29880118e+00 8.12897027e-01 1.07549287e-01 -6.96728110e-01
-9.31256056e-01 -7.64951631e-02 -2.33586609e-01 1.00795996e+00
-1.07249200e+00 -9.35876191e-01 -1.49220377e-01 6.80474699e-01
4.19246763e-01 -1.87406763e-01 1.59383953e+00 1.28806099e-01
-3.20040554e-01 4.54748690e-01 -2.04386950e-01 1.38581470e-01
9.74608123e-01 -1.27951300e+00 3.37837160e-01 -6.03781417e-02
-6.97772026e-01 7.49692202e-01 8.28028560e-01 -9.47525144e-01
-1.30545378e+00 -7.65126109e-01 1.38539553e+00 -6.02163076e-01
2.15859905e-01 -3.49682510e-01 -8.97811174e-01 4.11235511e-01
8.54790926e-01 -5.90350449e-01 1.29384458e+00 -1.53705418e-01
4.67347503e-01 3.15027863e-01 -2.10401845e+00 4.76636291e-01
6.29578054e-01 -4.92008299e-01 -9.06354964e-01 7.72333086e-01
3.78810793e-01 -7.27152109e-01 -1.31714785e+00 1.52833357e-01
5.49795985e-01 -9.53533530e-01 6.94433212e-01 -7.14853525e-01
1.69148698e-01 5.92100732e-02 2.87670821e-01 -1.18596530e+00
3.36694181e-01 -1.32690263e+00 7.29959011e-02 1.24434960e+00
3.83881569e-01 -8.60814095e-01 6.82006717e-01 1.83001876e+00
5.04675582e-02 -6.93823457e-01 -7.55407631e-01 -2.10244376e-02
-2.47478317e-02 1.41572654e-01 4.03916091e-01 1.18150806e+00
8.98812950e-01 5.00485301e-01 -5.10916710e-01 -3.31509441e-01
1.29286662e-01 -5.52851915e-01 1.12793815e+00 -1.10225797e+00
4.12553959e-02 -1.22485705e-01 -1.19475797e-01 -4.60625499e-01
-2.78169900e-01 -2.36217722e-01 5.18405914e-01 -2.03486800e+00
-9.17364210e-02 -1.69806093e-01 6.64126039e-01 4.16391790e-01
-1.93366379e-01 -5.76987624e-01 8.30754116e-02 1.50598750e-01
-7.10029542e-01 3.22875977e-01 1.01341319e+00 4.26187903e-01
-5.60620308e-01 3.07341311e-02 -1.11215174e+00 3.92417848e-01
9.68605757e-01 -1.69616014e-01 -5.82747459e-01 7.98240229e-02
-7.99685046e-02 7.81379998e-01 -5.80689125e-02 -2.77210891e-01
6.54014766e-01 -8.45053419e-02 -1.83923617e-01 -3.16043824e-01
1.03712715e-01 -7.51687109e-01 3.74805689e-01 6.60564005e-01
-8.79925489e-01 -2.08740681e-01 9.47288647e-02 3.75887126e-01
-4.08607006e-01 -3.05153251e-01 5.54015696e-01 -5.54866910e-01
-2.23443732e-01 -4.34963197e-01 -1.03460872e+00 2.15166643e-01
1.41701019e+00 -2.07820073e-01 -3.33557874e-01 -7.85293460e-01
-8.66009414e-01 8.44815791e-01 3.80148530e-01 2.03680173e-01
5.62275887e-01 -7.49945402e-01 -9.69371438e-01 -1.94463972e-02
1.14477850e-01 5.99182658e-02 1.83387235e-01 5.59435189e-01
-7.25881159e-01 6.45319223e-01 -2.97529846e-01 -2.16596916e-01
-1.43497801e+00 5.07846117e-01 4.66356993e-01 -3.90031993e-01
-6.67137921e-01 5.39850354e-01 -1.85490757e-01 -8.82366300e-01
6.02694929e-01 2.48643845e-01 -7.29845405e-01 3.24236989e-01
8.09631884e-01 4.58075762e-01 -9.92831513e-02 -5.58443844e-01
-3.24044019e-01 -2.81692535e-01 -9.23219994e-02 -7.95362890e-01
1.06464767e+00 -4.73899752e-01 -1.93233594e-01 4.33222800e-01
6.04155540e-01 -3.73675585e-01 -5.45033991e-01 -2.10064024e-01
4.90277380e-01 -1.83968604e-01 -4.60621595e-01 -1.27958739e+00
-3.12664747e-01 3.46910626e-01 2.60398924e-01 9.00299370e-01
8.91452014e-01 -6.34350330e-02 6.91042840e-01 4.47162926e-01
3.58917236e-01 -1.49778736e+00 -3.47333431e-01 2.95169234e-01
1.18601298e+00 -1.25450945e+00 -1.38724193e-01 -5.37847221e-01
-1.66787219e+00 8.04412067e-01 3.61702919e-01 6.58531010e-01
8.08340132e-01 1.25904575e-01 8.58427107e-01 -7.04157770e-01
-9.50929403e-01 -3.45531590e-02 -1.53293878e-01 9.67414796e-01
6.95977867e-01 -1.82652380e-02 -9.39235628e-01 6.50870264e-01
-3.88189182e-02 3.77857238e-01 4.64842319e-01 1.53659928e+00
-1.33570224e-01 -1.13991666e+00 -4.99366701e-01 2.45418623e-01
-6.27074242e-01 1.47489056e-01 -1.38904333e+00 8.16299736e-01
-3.18369001e-01 1.53825295e+00 -4.36581314e-01 3.37985680e-02
1.00581837e+00 3.38509142e-01 -3.49314660e-02 -7.65976250e-01
-1.57257402e+00 -1.29063532e-01 1.16626585e+00 -5.64242303e-01
-7.75383472e-01 -8.42591047e-01 -9.21407044e-01 -9.38977450e-02
8.37828033e-03 7.68162012e-01 5.51433206e-01 5.30122519e-01
5.36119223e-01 2.24775672e-01 4.71743345e-01 -2.73703575e-01
-3.33577603e-01 -1.03413212e+00 -2.47539967e-01 3.98078859e-01
2.79356986e-01 4.37495299e-02 2.10526168e-01 2.08346665e-01] | [12.607176780700684, 8.27609634399414] |
e52d0435-ed2e-4f15-ade7-ac066fe86d54 | attack-is-good-augmentation-towards-skeleton | 2304.04023 | null | https://arxiv.org/abs/2304.04023v1 | https://arxiv.org/pdf/2304.04023v1.pdf | Attack is Good Augmentation: Towards Skeleton-Contrastive Representation Learning | Contrastive learning, relying on effective positive and negative sample pairs, is beneficial to learn informative skeleton representations in unsupervised skeleton-based action recognition. To achieve these positive and negative pairs, existing weak/strong data augmentation methods have to randomly change the appearance of skeletons for indirectly pursuing semantic perturbations. However, such approaches have two limitations: 1) solely perturbing appearance cannot well capture the intrinsic semantic information of skeletons, and 2) randomly perturbation may change the original positive/negative pairs to soft positive/negative ones. To address the above dilemma, we start the first attempt to explore an attack-based augmentation scheme that additionally brings in direct semantic perturbation, for constructing hard positive pairs and further assisting in constructing hard negative pairs. In particular, we propose a novel Attack-Augmentation Mixing-Contrastive learning (A$^2$MC) to contrast hard positive features and hard negative features for learning more robust skeleton representations. In A$^2$MC, Attack-Augmentation (Att-Aug) is designed to collaboratively perform targeted and untargeted perturbations of skeletons via attack and augmentation respectively, for generating high-quality hard positive features. Meanwhile, Positive-Negative Mixer (PNM) is presented to mix hard positive features and negative features for generating hard negative features, which are adopted for updating the mixed memory banks. Extensive experiments on three public datasets demonstrate that A$^2$MC is competitive with the state-of-the-art methods. | ['Mike Zheng Shou', 'Yixiao Ge', 'Guo-Sen Xie', 'Rui Yan', 'Xiangbo Shu', 'Binqian Xu'] | 2023-04-08 | null | null | null | null | ['action-recognition-in-videos'] | ['computer-vision'] | [ 7.99941778e-01 3.24998409e-01 -4.88566160e-01 -6.15980774e-02
-9.57616568e-01 -6.56962097e-02 7.72210717e-01 -2.57053345e-01
-3.13037634e-01 6.79591715e-01 4.03957486e-01 1.99726999e-01
6.47886842e-02 -1.02977979e+00 -7.40399778e-01 -9.32151020e-01
-2.64612306e-03 4.75776196e-01 3.88956368e-01 -4.47025090e-01
8.30658451e-02 2.86465764e-01 -1.57190895e+00 6.13585770e-01
6.59171760e-01 9.29548144e-01 -2.30138794e-01 2.61164695e-01
-2.34591320e-01 7.29282916e-01 -5.72861493e-01 -4.54027176e-01
5.44628263e-01 -4.51804668e-01 -6.74330175e-01 -6.01860173e-02
1.44992709e-01 -2.67273158e-01 -3.78045708e-01 1.03204215e+00
8.33344162e-01 2.24882588e-01 8.52757156e-01 -1.58626938e+00
-6.54973865e-01 5.94446123e-01 -1.12961531e+00 2.05193490e-01
5.33958673e-01 7.33890772e-01 7.94608176e-01 -9.03468788e-01
2.67536879e-01 1.52443421e+00 3.96689653e-01 1.08690751e+00
-9.04467165e-01 -9.41740990e-01 3.61791879e-01 3.86851311e-01
-1.16460204e+00 -3.11279595e-01 1.30959928e+00 -2.00600982e-01
5.34093797e-01 4.99415606e-01 8.48083377e-01 1.55721354e+00
-2.79259950e-01 1.24930871e+00 1.13915968e+00 -2.53723800e-01
1.38811275e-01 -2.32445493e-01 -9.25398432e-03 8.39864314e-01
-7.59285763e-02 2.19692737e-01 -7.26937532e-01 -2.41705373e-01
7.77732193e-01 -3.62765417e-03 -3.11122179e-01 -3.68048757e-01
-1.22184479e+00 5.67533016e-01 3.08320999e-01 4.77045923e-02
-3.52780700e-01 1.22903883e-01 5.82723320e-01 1.06079713e-01
1.36059046e-01 3.16621244e-01 -5.27911007e-01 -1.29687369e-01
-3.65982205e-01 1.34366944e-01 6.93426505e-02 5.88207185e-01
7.05792248e-01 2.51754284e-01 -3.63799334e-01 8.40770185e-01
3.56153309e-01 6.47016764e-01 1.05442119e+00 -5.17129183e-01
5.80678225e-01 9.43153918e-01 -2.80160457e-01 -1.04127169e+00
-3.61960083e-01 -3.00341070e-01 -1.03370583e+00 2.64974207e-01
2.67398626e-01 7.73989260e-02 -1.22722721e+00 2.06758404e+00
5.18260360e-01 5.07516921e-01 1.97564676e-01 7.50001013e-01
9.74640608e-01 3.53246629e-01 2.68209517e-01 -1.85462505e-01
1.30879188e+00 -1.00811541e+00 -4.15836364e-01 -4.33878094e-01
5.21264076e-01 -6.37810290e-01 1.61022449e+00 4.65943635e-01
-1.03311193e+00 -6.10453486e-01 -1.06974435e+00 2.03780249e-01
-1.94492072e-01 1.14151686e-01 5.90970159e-01 6.81944966e-01
-3.49319041e-01 4.41877663e-01 -9.17319059e-01 3.41005698e-02
8.55380177e-01 5.03344238e-01 -5.14851689e-01 2.01278180e-03
-1.34493041e+00 5.85200012e-01 2.30751529e-01 7.26811960e-02
-1.07647347e+00 -6.45865440e-01 -9.26589608e-01 -2.96422869e-01
4.91099715e-01 -6.22101605e-01 6.87572718e-01 -1.00209570e+00
-1.61146402e+00 7.74403095e-01 2.38526836e-01 -2.17540681e-01
7.36552000e-01 -2.71744102e-01 -4.78311151e-01 3.86496276e-01
6.63432330e-02 1.04816461e+00 1.26196742e+00 -1.27147222e+00
-3.54285032e-01 -5.96579909e-01 -7.13649467e-02 4.36304122e-01
-7.37386346e-01 -2.64924347e-01 -4.28801745e-01 -1.13909721e+00
3.32043618e-01 -6.48875415e-01 -4.41659600e-01 1.93074048e-01
-5.37806690e-01 -3.33819576e-02 9.37035322e-01 -4.14072096e-01
1.06122136e+00 -2.06150389e+00 2.07648814e-01 3.52956384e-01
1.33230954e-01 6.22089803e-01 -4.97125298e-01 -9.60582122e-02
-5.49635291e-01 -2.47798860e-03 -4.84461337e-01 -1.50163159e-01
-1.21696189e-01 3.90093684e-01 -3.21150631e-01 3.31862688e-01
7.48276770e-01 9.49860632e-01 -1.00142777e+00 -6.87590122e-01
2.33654737e-01 2.78375536e-01 -5.71507692e-01 1.48090690e-01
-8.59879330e-02 5.42170227e-01 -5.77091336e-01 9.44796622e-01
6.20640099e-01 3.88490781e-02 -1.20566208e-02 -4.46399897e-01
6.89977527e-01 -1.54616907e-01 -1.29013014e+00 1.37834084e+00
-6.70177564e-02 -2.95080960e-01 -2.53329068e-01 -1.35704184e+00
9.84630287e-01 7.37269223e-02 7.13158667e-01 -7.13433504e-01
1.26530871e-01 3.93406451e-02 -1.48671970e-01 -5.99261522e-01
8.42868388e-02 -5.30356705e-01 -2.15779059e-02 4.55006748e-01
-1.43631697e-01 -1.06126368e-01 -1.45747349e-01 9.24953520e-02
1.25704193e+00 4.43288714e-01 5.48805818e-02 2.99682170e-01
9.60315883e-01 -4.48258281e-01 9.07784998e-01 1.62510321e-01
-5.17225146e-01 7.79960394e-01 5.35543025e-01 -2.53475577e-01
-5.99629521e-01 -1.28267801e+00 3.25902015e-01 1.00713968e+00
1.64027721e-01 -3.26423615e-01 -7.21219540e-01 -1.34554625e+00
-2.24426225e-01 2.86248773e-01 -8.30570519e-01 -9.68815386e-01
-7.50179708e-01 -1.09721529e+00 8.16249132e-01 9.16685760e-01
8.61010849e-01 -1.17589521e+00 -9.90116969e-02 -8.95835608e-02
-4.25810397e-01 -5.81537664e-01 -3.30917507e-01 4.50917594e-02
-8.71780753e-01 -1.38694263e+00 -7.88817942e-01 -5.21594822e-01
9.54608440e-01 6.58797845e-02 6.17549241e-01 1.08760886e-01
-2.30711490e-01 2.96219081e-01 -4.38319474e-01 -1.91901058e-01
-3.88362080e-01 -1.69525072e-01 2.57952571e-01 2.37097368e-01
2.46805072e-01 -9.09746826e-01 -6.29335344e-01 4.09092069e-01
-1.40040362e+00 -4.60765734e-02 9.77343261e-01 1.09573722e+00
6.58170521e-01 -2.22207587e-02 9.19220090e-01 -7.27566183e-01
2.67820686e-01 -4.91151959e-01 2.17324570e-01 1.48176834e-01
-4.72226411e-01 1.45394951e-01 8.05664539e-01 -9.60764766e-01
-1.01423919e+00 2.62989879e-01 -4.20447290e-01 -7.03566551e-01
-6.88377395e-02 2.46521309e-01 -7.09449232e-01 -4.88078548e-03
1.06800604e+00 5.27036905e-01 2.87533671e-01 -3.58969659e-01
5.11212885e-01 4.02626574e-01 6.71391785e-01 -9.14051950e-01
1.18518448e+00 6.80867970e-01 1.47499114e-01 -5.61740875e-01
-5.46047032e-01 -2.44488105e-01 -5.80516517e-01 -1.35054454e-01
5.54836988e-01 -7.54180610e-01 -3.00959885e-01 8.34933519e-01
-4.16985244e-01 -2.76142418e-01 -8.11178029e-01 3.82259667e-01
-8.59970331e-01 9.56479013e-01 -3.68891001e-01 -4.15058762e-01
-4.45462704e-01 -1.06174195e+00 1.06415904e+00 7.36461282e-02
-1.98308632e-01 -5.17729521e-01 -1.65013988e-02 8.42684686e-01
4.77626473e-02 3.78472239e-01 9.72885191e-01 -7.03156531e-01
-1.14875831e-01 -3.27539712e-01 -9.53602195e-02 7.31571198e-01
2.85958439e-01 -1.38999820e-01 -9.62921023e-01 -1.22784011e-01
-2.01262996e-01 -7.50089765e-01 9.54219282e-01 -8.15771967e-02
1.30947053e+00 -3.36522460e-01 -2.47322991e-01 3.60253841e-01
7.40994871e-01 -1.73329830e-01 1.03202546e+00 2.18989149e-01
8.41353834e-01 3.60172242e-01 1.02076125e+00 7.53253460e-01
8.99411291e-02 3.45704406e-01 5.83345890e-01 -2.93094069e-01
-2.99457163e-01 -3.53937894e-01 7.56742001e-01 3.54307115e-01
-2.20850095e-01 2.69386470e-01 -4.90378797e-01 4.02547330e-01
-1.69312787e+00 -8.40226173e-01 3.54579128e-02 2.12276220e+00
1.32149470e+00 4.64122176e-01 3.83034557e-01 6.61414683e-01
7.61372805e-01 3.20631474e-01 -6.64433479e-01 3.19913715e-01
-2.04927742e-01 6.81709230e-01 5.32136373e-02 -2.73848418e-02
-1.09147441e+00 7.98762321e-01 4.92791605e+00 1.39045799e+00
-1.09784245e+00 -1.17429346e-02 6.02764606e-01 -2.26175338e-01
-5.56777716e-01 -2.24996015e-01 -5.19135714e-01 4.94789600e-01
4.57784563e-01 2.84444362e-01 3.85312997e-02 8.88154209e-01
-6.00705408e-02 2.15332150e-01 -8.76063168e-01 9.12302196e-01
4.69721779e-02 -1.04471195e+00 7.15742707e-01 -1.62660837e-01
5.97674429e-01 -5.19591153e-01 2.62371093e-01 6.45746291e-01
2.77005732e-01 -7.80502915e-01 5.50662041e-01 5.43894112e-01
7.95072496e-01 -8.32441330e-01 5.53707421e-01 2.23483950e-01
-1.21457970e+00 -1.81397900e-01 -1.30778819e-01 1.13244191e-01
-7.78158754e-02 3.51948440e-01 -4.79690343e-01 6.43297315e-01
5.56353986e-01 7.30886042e-01 -7.59152055e-01 4.89562571e-01
-4.32629794e-01 5.04909754e-01 -3.56771946e-01 3.32999527e-01
3.22015844e-02 9.06918570e-02 5.37742257e-01 8.15074444e-01
-2.39215810e-02 1.66800365e-01 2.25049675e-01 6.46370351e-01
-1.68142244e-02 1.58671811e-01 -4.67664182e-01 -1.41463310e-01
1.75692439e-01 1.16528583e+00 -5.63696623e-01 -5.22602141e-01
-1.69367418e-01 9.84185994e-01 1.06173784e-01 7.03167021e-02
-1.08566272e+00 -1.68209001e-01 4.82592881e-01 1.54193863e-01
-5.37464321e-02 2.66466290e-01 -2.23685518e-01 -1.26285982e+00
2.00926021e-01 -1.30603838e+00 7.44627833e-01 -4.99617636e-01
-1.41199970e+00 8.93625394e-02 -4.73007895e-02 -1.59089518e+00
8.25197622e-02 -3.37845594e-01 -7.18548059e-01 5.83722711e-01
-1.35774136e+00 -1.63534510e+00 -3.12176257e-01 1.06557584e+00
4.90582496e-01 -3.38429660e-01 6.87315404e-01 3.63761008e-01
-7.87158847e-01 1.07332742e+00 -5.87123156e-01 1.34806693e-01
7.19897032e-01 -1.02037573e+00 1.38551489e-01 7.30793655e-01
-1.66653737e-01 4.40736264e-01 3.75090837e-01 -7.67840326e-01
-1.33696377e+00 -1.24231195e+00 -1.03074737e-01 -2.23727390e-01
4.31423068e-01 1.54280379e-01 -8.06420326e-01 5.26784658e-01
-4.35354769e-01 4.05690610e-01 9.61052358e-01 -3.99551034e-01
-5.00119150e-01 -1.78891733e-01 -1.53739476e+00 8.56043637e-01
1.28664029e+00 -2.34298944e-01 -6.95514560e-01 4.00494665e-01
5.46313524e-01 -1.57649070e-01 -8.04095149e-01 1.00056672e+00
5.12523472e-01 -7.97674537e-01 1.36846960e+00 -8.19493055e-01
6.58733606e-01 -3.51830304e-01 -2.69099295e-01 -9.65349019e-01
-1.86683878e-01 -5.12504220e-01 -4.37138200e-01 1.24821544e+00
1.08491592e-01 -6.45098686e-01 1.34806693e+00 2.59280950e-01
-2.67620325e-01 -1.16811180e+00 -9.66915965e-01 -7.77569592e-01
1.31616011e-01 -5.29986501e-01 6.12934053e-01 1.15855169e+00
2.36998545e-03 1.98299557e-01 -2.98336416e-01 2.61919014e-02
3.84717017e-01 -3.65025252e-01 1.04590404e+00 -8.89701843e-01
-5.27011216e-01 -5.41327178e-01 -6.90717578e-01 -8.97763610e-01
1.88328832e-01 -7.19747066e-01 -1.93063825e-01 -1.00081146e+00
2.39622116e-01 -5.06649375e-01 -4.38533008e-01 9.78557408e-01
-5.43592513e-01 6.85869873e-01 -1.31627008e-01 1.09315336e-01
-2.56223738e-01 1.14979148e+00 1.55316401e+00 -4.68028843e-01
-1.04030579e-01 1.65891275e-01 -8.40863347e-01 1.05106878e+00
5.89269757e-01 -1.01948686e-01 -6.00312352e-01 2.25159824e-01
1.42247096e-01 -2.03444377e-01 4.48606223e-01 -1.16212273e+00
-3.83998275e-01 -3.75366122e-01 6.35760188e-01 -5.31656325e-01
6.42228186e-01 -5.62453032e-01 -3.38121414e-01 7.51906753e-01
-3.03284109e-01 -3.68552148e-01 -4.53167129e-03 8.43244433e-01
-1.02453537e-01 -2.42782205e-01 9.97031808e-01 -3.69869500e-01
-6.65553749e-01 4.34378773e-01 -9.10590962e-02 1.80938378e-01
1.16593933e+00 -4.67641741e-01 -2.96907455e-01 -7.70324096e-02
-1.02544343e+00 1.10530786e-01 2.42222935e-01 3.34631413e-01
9.55550909e-01 -1.71287537e+00 -4.96328264e-01 4.69445080e-01
2.03242928e-01 1.65274560e-01 5.17529428e-01 8.72529864e-01
-1.09220266e-01 -4.33892369e-01 -4.54398155e-01 -4.11101222e-01
-1.29913855e+00 6.40831709e-01 1.81301832e-01 -5.08748412e-01
-4.19495761e-01 1.10025382e+00 3.27764362e-01 -6.19352043e-01
2.21824795e-01 -4.46690358e-02 -3.10352087e-01 1.25062436e-01
4.80039954e-01 5.99464834e-01 -1.37557343e-01 -5.04253983e-01
-2.91460454e-01 5.61448872e-01 -8.04525316e-02 -3.01043373e-02
1.29587746e+00 1.94562599e-01 -1.99127011e-02 1.98666919e-02
9.76930678e-01 -2.77465954e-02 -1.18040395e+00 -3.77120405e-01
-5.03432155e-01 -5.54024160e-01 -3.00701916e-01 -7.01313198e-01
-1.43499947e+00 9.58318591e-01 7.85853207e-01 -3.20888907e-01
1.48989165e+00 -1.33363217e-01 1.13409567e+00 3.28711987e-01
8.68752450e-02 -1.09659970e+00 9.99215722e-01 -5.84140569e-02
9.54905152e-01 -1.18459868e+00 8.92831832e-02 -6.32451594e-01
-6.65973723e-01 9.56399381e-01 1.21547794e+00 -5.47447475e-04
4.27212447e-01 1.33673931e-02 -1.47240728e-01 -2.36456141e-01
-4.32845712e-01 -2.39332035e-01 2.12824151e-01 9.86501098e-01
-1.47279380e-02 -8.67393240e-02 -3.90630424e-01 9.13474500e-01
-4.52522971e-02 -7.28426576e-02 2.06622258e-01 1.16971421e+00
-3.96966845e-01 -1.46819270e+00 -5.72133303e-01 5.38668275e-01
-3.02465767e-01 2.78325509e-02 -5.09827316e-01 7.59996593e-01
3.17608088e-01 6.10203683e-01 -3.13106835e-01 -8.14349711e-01
3.17484438e-01 1.38421550e-01 6.30204797e-01 -6.84620500e-01
-4.28140759e-01 -3.98331974e-03 4.18395884e-02 -5.34765065e-01
-3.40406775e-01 -5.75055897e-01 -1.71001303e+00 1.02209255e-01
-2.88719565e-01 -1.97187781e-01 7.05756189e-04 9.67774272e-01
1.69345662e-01 5.53160906e-01 7.79461980e-01 -8.29018176e-01
-1.02733874e+00 -1.09799325e+00 -3.84606570e-01 7.66453266e-01
-1.58208739e-02 -1.07032073e+00 -4.08623993e-01 8.02850723e-02] | [8.376723289489746, 0.7557919025421143] |
f15d6307-54ff-48e0-b32d-3629cd733a54 | detecting-english-grammatical-errors-based-on | null | null | https://aclanthology.org/O13-1006 | https://aclanthology.org/O13-1006.pdf | Detecting English Grammatical Errors based on Machine Translation | null | ['Jian-Cheng Wu', 'Jim Chang', 'Jason S. Chang'] | 2013-10-01 | detecting-english-grammatical-errors-based-on-1 | https://aclanthology.org/O13-1006 | https://aclanthology.org/O13-1006.pdf | roclingijclclp-2013-10 | ['grammatical-error-detection'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.235448837280273, 3.619205951690674] |
daf5bc35-ee5a-45ca-a6f2-939bd29d2a35 | variational-encoding-approach-for | null | null | https://www.sciencedirect.com/science/article/pii/S0951832022000321 | https://www.sciencedirect.com/science/article/pii/S0951832022000321 | Variational encoding approach for interpretable assessment of remaining useful life estimation | A new method for evaluating aircraft engine monitoring data is proposed. Commonly, prognostics and health management systems use knowledge of the degradation processes of certain engine components together with professional expert opinion to predict the Remaining Useful Life (RUL). New data-driven approaches have emerged to provide accurate diagnostics without relying on such costly processes. However, most of them lack an explanatory component to understand model learning and/or the nature of the data. To overcome this gap we propose a novel approach based on variational encoding. The model consists of a recurrent encoder and a regression model: the encoder learns to compress the input data to a latent space that serves as a basis to build a self-explanatory map that can visually evaluate the rate of deterioration of aircraft engines. Obtaining such a latent space is regularized by a new cost function guided by variational inference and a term that penalizes prediction errors. Consequently, not only an interpretable assessment is achieved but also a remarkable prognostic accuracy, outperforming most of the state-of-the-art approaches on the popular simulation dataset C-MAPSS from NASA. In addition, we demonstrate the application of our method in a real-world scenario with data from actual Turbofan engines. | ['Luciano Sanchez', 'Nahuel Costa'] | 2022-02-23 | null | null | null | reliability-engineering-system-safety-ress | ['remaining-useful-lifetime-estimation'] | ['time-series'] | [ 1.14477128e-01 -8.53691250e-02 -9.69837233e-02 -2.80193806e-01
-6.21244669e-01 -1.36298463e-01 4.82695103e-01 3.54822576e-01
2.99970154e-04 7.43424773e-01 -9.98190865e-02 -2.94992656e-01
-6.11908257e-01 -7.98688948e-01 -5.59415042e-01 -1.03970897e+00
2.18753424e-02 6.11476958e-01 1.23278394e-01 -7.14083686e-02
6.86256588e-02 5.26904881e-01 -1.83194518e+00 2.33028412e-01
9.18179572e-01 1.27045500e+00 1.32961005e-01 5.29937685e-01
2.82217562e-01 8.23893070e-01 -5.09762585e-01 -9.70545858e-02
-1.27031446e-01 -4.54322219e-01 -4.95351195e-01 2.54160553e-01
-5.11972427e-01 6.61513675e-03 -3.74930263e-01 7.04240084e-01
2.98740804e-01 1.46368101e-01 9.90342021e-01 -1.10550463e+00
-1.45777628e-01 4.86662127e-02 1.52964637e-01 4.83049974e-02
3.20936628e-02 3.09450626e-01 7.27112174e-01 -4.97335851e-01
4.23657864e-01 8.30483735e-01 6.26698077e-01 5.01534879e-01
-1.21718681e+00 -2.87160836e-02 -9.51090008e-02 4.17402536e-01
-1.14822006e+00 -2.38482520e-01 1.01138341e+00 -8.09090137e-01
5.98932564e-01 3.07011515e-01 7.52196014e-01 1.09283447e+00
6.50575876e-01 3.79236430e-01 9.56418633e-01 -1.70232475e-01
5.03738225e-01 3.24495614e-01 -8.26153308e-02 7.68521845e-01
4.71559679e-03 4.01742578e-01 -4.91177499e-01 7.07425550e-03
4.82537717e-01 2.62931079e-01 -3.04790765e-01 -5.12727618e-01
-8.45390260e-01 5.69680333e-01 2.76352316e-01 1.69091284e-01
-7.95567036e-01 1.52582064e-01 5.38869619e-01 8.49245116e-02
7.98546553e-01 3.83190334e-01 -3.66097212e-01 -8.44994560e-02
-1.21189535e+00 3.92617546e-02 6.61663413e-01 4.33490515e-01
4.12693590e-01 2.33380422e-01 -3.88818949e-01 4.06132370e-01
5.23286343e-01 3.43116045e-01 4.73950028e-01 -8.60410154e-01
-5.19544771e-03 7.38131046e-01 1.90892935e-01 -5.97185373e-01
-2.75774509e-01 -9.31928098e-01 -8.71998549e-01 5.82446456e-01
4.66571525e-02 7.36304075e-02 -8.08011174e-01 1.37572002e+00
1.97168142e-01 2.44931385e-01 7.08047375e-02 7.59760737e-01
1.95356980e-01 6.98801100e-01 -7.48974383e-02 -4.77297693e-01
1.25661373e+00 -7.41652012e-01 -8.29861760e-01 1.54262051e-01
3.69600147e-01 -2.68783510e-01 8.88835013e-01 6.96389556e-01
-1.06756020e+00 -7.24418283e-01 -1.36943674e+00 2.91530043e-01
-4.06958938e-01 3.92177016e-01 1.47622705e-01 4.72114265e-01
-7.97181666e-01 1.15320909e+00 -1.19860911e+00 1.40756611e-02
1.89230159e-01 9.05866846e-02 -1.12301335e-01 9.36952904e-02
-1.22457194e+00 1.02329993e+00 2.72445202e-01 3.40566695e-01
-1.59662843e+00 -7.54662514e-01 -6.87420249e-01 2.09265873e-01
3.79631341e-01 -9.96145427e-01 1.05564964e+00 -4.47700143e-01
-1.64111578e+00 4.07002062e-01 -1.21339910e-01 -7.32984006e-01
7.81225860e-01 -2.99706846e-01 -5.44222772e-01 2.44037837e-01
-3.66355300e-01 -1.17098399e-01 1.23388553e+00 -1.22784007e+00
-1.36338755e-01 -2.46356592e-01 -2.71491915e-01 -1.79638356e-01
-3.95048052e-01 -3.53824288e-01 -2.58259535e-01 -5.03707469e-01
-1.55266747e-01 -6.78774476e-01 -1.90338105e-01 -4.36780788e-02
-3.11502159e-01 -1.35914758e-01 7.34186828e-01 -9.25116837e-01
1.46373975e+00 -2.07660079e+00 6.21641457e-01 9.10259411e-02
3.64240080e-01 8.28755125e-02 4.45827276e-01 6.36111081e-01
-1.94409296e-01 -1.88734934e-01 -6.08838201e-01 -6.58053458e-01
-1.69382170e-02 2.98469543e-01 -3.43673378e-01 5.14246762e-01
5.04128039e-01 6.15957081e-01 -8.85687172e-01 -4.08692747e-01
4.42757547e-01 7.13390529e-01 -2.39896595e-01 5.28909028e-01
-3.96640241e-01 6.80384278e-01 -4.66769516e-01 4.87497568e-01
6.83684573e-02 -3.67659599e-01 5.75736649e-02 -8.69034380e-02
-1.66334793e-01 -2.98060067e-02 -6.74688876e-01 1.60558379e+00
-5.72241962e-01 3.52591276e-01 -1.20520651e-01 -1.28357780e+00
1.00933802e+00 6.32588625e-01 6.83576047e-01 -4.22582507e-01
1.21195227e-01 1.92557886e-01 -3.65090579e-01 -7.16824114e-01
1.58865616e-01 -5.68550467e-01 1.03926316e-01 1.33583263e-01
-1.73458755e-02 -1.02039754e-01 2.43232865e-03 -1.33191630e-01
1.21636701e+00 2.65356749e-01 -1.92291141e-01 -2.23632395e-01
8.36904824e-01 -1.22812323e-01 6.72904372e-01 1.15322232e-01
5.02740145e-02 4.03076112e-01 8.18983078e-01 -3.23417097e-01
-1.10241425e+00 -1.04496181e+00 -2.18489051e-01 3.40563297e-01
-2.12857664e-01 -5.25040090e-01 -8.15705836e-01 -5.80832958e-01
-2.47069751e-03 9.77617562e-01 -8.23577166e-01 -7.23141372e-01
-1.33905888e-01 -7.31237054e-01 1.26982778e-01 5.96207738e-01
5.44539143e-05 -7.75979161e-01 -8.36992979e-01 2.76072323e-01
1.04724981e-01 -8.26095283e-01 1.40601918e-01 3.48579615e-01
-1.23818815e+00 -1.06304896e+00 -6.23390853e-01 -6.97805658e-02
8.05965245e-01 -4.86984104e-01 1.21627724e+00 3.49737629e-02
-4.73984450e-01 4.17862028e-01 -1.00049473e-01 -3.26535493e-01
-9.04233932e-01 -1.70864969e-01 2.47899801e-01 2.46512130e-01
-2.70862490e-01 -4.15988654e-01 -7.57085025e-01 4.26787496e-01
-9.45817649e-01 5.21402620e-02 7.91350842e-01 1.02369928e+00
8.54154468e-01 5.23503959e-01 6.23025835e-01 -6.57852888e-01
5.38219273e-01 -7.22951591e-01 -6.69579148e-01 3.89752805e-01
-1.22583032e+00 4.26700503e-01 8.05852652e-01 -8.80909264e-02
-1.08980346e+00 -8.72862935e-02 -1.12560891e-01 -9.39714670e-01
2.62728054e-02 7.42073298e-01 -1.17832847e-01 5.38686395e-01
2.60850012e-01 2.43129715e-01 2.12912723e-01 -7.84745216e-01
3.48554030e-02 5.10668099e-01 6.97559893e-01 -4.65811759e-01
5.56998074e-01 3.87649953e-01 5.06917238e-01 -5.62771440e-01
-5.69626629e-01 -3.25681061e-01 -6.30883753e-01 -6.21114910e-01
8.89654875e-01 -6.55153990e-01 -7.80036271e-01 3.08989167e-01
-9.64194417e-01 -2.17033699e-01 -7.63012469e-01 4.84138608e-01
-8.83223057e-01 1.36710554e-01 -5.73737323e-01 -1.18039930e+00
-2.63637155e-01 -1.24767363e+00 1.12329352e+00 -2.31963009e-01
1.46368861e-01 -1.12036967e+00 1.76846638e-01 2.07604215e-01
4.44083244e-01 5.75898588e-01 1.15621924e+00 -2.61484206e-01
-4.75867599e-01 -5.86891532e-01 2.31869414e-01 7.68072128e-01
2.05532126e-02 5.86270615e-02 -1.08619046e+00 -4.07908648e-01
5.78963816e-01 3.67774330e-02 9.55130160e-01 4.35100436e-01
1.45484340e+00 -5.14098033e-02 -3.02936167e-01 3.22633535e-01
1.26067543e+00 1.04491726e-01 5.23422539e-01 -9.97100770e-02
3.03980350e-01 8.08546305e-01 6.36904776e-01 5.78096151e-01
-9.24112555e-03 7.34892488e-01 7.23155320e-01 1.07700475e-01
2.67879087e-02 -2.74549186e-01 4.91656929e-01 1.16849887e+00
-3.32439661e-01 -1.87898278e-01 -1.00295198e+00 3.29814345e-01
-1.89320683e+00 -6.48670852e-01 -5.50048798e-02 2.44479537e+00
5.39403439e-01 4.83560264e-01 -1.83642745e-01 5.84972262e-01
3.72952431e-01 8.82828757e-02 -7.52867639e-01 -8.53590593e-02
2.00404748e-01 -5.14758006e-02 2.71877885e-01 3.00007999e-01
-8.57749581e-01 1.91957906e-01 6.27906466e+00 6.72999501e-01
-9.42144632e-01 1.73198387e-01 7.19774961e-01 -8.60545114e-02
-2.83020526e-01 -1.48215711e-01 -3.80410582e-01 5.26189804e-01
1.56996536e+00 -9.95839387e-02 2.22669259e-01 7.35571325e-01
4.41210538e-01 5.17227836e-02 -1.13775337e+00 6.89586043e-01
-9.82438922e-02 -1.15983510e+00 -1.77745640e-01 1.10527463e-01
3.37059766e-01 -3.75761837e-01 1.92369998e-01 1.23462491e-01
-3.50764185e-01 -1.06591094e+00 7.34306455e-01 1.54640996e+00
9.11156118e-01 -6.96585178e-01 9.13833916e-01 3.82695198e-01
-1.02318096e+00 -2.72473693e-01 -1.83395609e-01 1.83713287e-01
2.82275945e-01 9.14254069e-01 -5.98063886e-01 9.43297327e-01
4.74766403e-01 8.95062327e-01 -5.58051884e-01 9.12917018e-01
-2.55884141e-01 8.00045788e-01 7.63335358e-03 2.60067880e-01
-1.41872987e-01 -1.45152494e-01 7.54829645e-01 6.38992488e-01
5.96906602e-01 -4.30607647e-01 -1.63426667e-01 1.07944179e+00
2.84678757e-01 -3.15355748e-01 -5.00979900e-01 -3.37127864e-01
-3.46263759e-02 1.11196756e+00 -5.58793306e-01 -3.89594018e-01
-1.34246424e-01 8.71059060e-01 -6.56533018e-02 2.59557217e-01
-8.65752697e-01 -8.15086216e-02 5.84426045e-01 4.93623227e-01
1.45299360e-01 -3.49491835e-01 -8.67334753e-02 -9.31119323e-01
8.39776918e-02 -5.12654364e-01 2.00707927e-01 -7.52530456e-01
-1.12358010e+00 9.09949899e-01 1.72996774e-01 -1.56031764e+00
-6.78547025e-01 -7.54179239e-01 -5.63233018e-01 9.52855587e-01
-1.53123915e+00 -7.86211848e-01 -3.53864014e-01 3.09454232e-01
5.51021755e-01 -3.02190363e-01 7.97257304e-01 3.03541183e-01
-9.33169842e-01 1.53109813e-02 4.33579743e-01 -5.10509193e-01
4.76942271e-01 -1.41094720e+00 9.56815556e-02 7.41563857e-01
-2.21965268e-01 2.25483224e-01 1.11724734e+00 -6.27293229e-01
-1.34808636e+00 -9.14671183e-01 4.91576225e-01 -5.76017141e-01
6.75847471e-01 -1.56990662e-01 -1.08924735e+00 1.83151305e-01
8.63388404e-02 8.14240202e-02 5.61077535e-01 -1.17905945e-01
3.09032768e-01 -3.30724120e-01 -8.57258677e-01 -1.03064347e-03
6.25700533e-01 -6.84086740e-01 -4.29898709e-01 2.74547219e-01
5.61725020e-01 -2.24644616e-01 -1.19746590e+00 5.27475834e-01
3.57999653e-01 -9.29688931e-01 8.75377476e-01 -7.47039318e-01
6.01702452e-01 -3.91501456e-01 1.66866526e-01 -1.34465790e+00
-4.14681956e-02 -4.86108065e-01 -7.44464517e-01 1.05399585e+00
1.16780527e-01 -4.61504310e-01 5.33476055e-01 5.98148704e-01
-4.42878813e-01 -1.25646698e+00 -9.87107933e-01 -8.12811255e-01
-1.59696415e-01 -6.01974905e-01 5.35969675e-01 3.53855014e-01
-2.96846330e-01 -6.02056347e-02 -3.89281154e-01 2.80898452e-01
7.19356656e-01 -3.11285257e-02 9.47153717e-02 -1.50416589e+00
-3.92180711e-01 -2.69673884e-01 -5.18314481e-01 -4.26245660e-01
2.33426645e-01 -6.76444411e-01 1.10139221e-01 -1.34701395e+00
-6.45826384e-03 -2.11206019e-01 -8.71235728e-01 5.21601997e-02
-3.70666683e-02 -1.86940715e-01 -1.71297699e-01 3.52130175e-01
-4.10627067e-01 1.02483165e+00 9.59885418e-01 -6.92606866e-02
7.19975829e-02 2.57579535e-01 -1.33620232e-01 5.89163780e-01
6.05849922e-01 -5.22320151e-01 -5.47344923e-01 -4.91620973e-02
3.56124431e-01 5.52503526e-01 9.12629068e-01 -1.44598687e+00
2.71667361e-01 1.99808672e-01 3.32462281e-01 -6.53810322e-01
4.77351099e-01 -8.97235453e-01 5.04106402e-01 7.82476425e-01
-2.21657813e-01 1.26186043e-01 6.63664043e-02 1.00448871e+00
-5.72859406e-01 -3.61937821e-01 5.81392765e-01 2.38259882e-01
-3.92528266e-01 3.02356184e-01 -4.55797970e-01 -4.40857530e-01
1.02601779e+00 8.86823833e-02 -2.01538485e-02 -1.88131526e-01
-9.06209409e-01 2.93740124e-01 3.65410805e-01 3.74586195e-01
6.75601304e-01 -1.27649987e+00 -6.47943854e-01 3.68593752e-01
1.11018158e-01 -9.14502665e-02 5.86332083e-01 1.00169528e+00
-3.80356222e-01 5.55540502e-01 -1.11183882e-01 -5.94562054e-01
-8.66287947e-01 7.71137178e-01 6.04308844e-01 -5.94075203e-01
-5.94123602e-01 3.99557501e-01 4.54398897e-03 1.20633110e-01
3.58763598e-02 -4.28930104e-01 -1.98582441e-01 1.17477335e-01
2.26546675e-01 5.82764447e-01 2.90913671e-01 -4.21642125e-01
-2.40755141e-01 2.70793825e-01 3.63545120e-01 -3.89086897e-03
1.35746419e+00 6.13348139e-03 3.65634672e-02 9.89507973e-01
9.72295523e-01 -4.29086745e-01 -1.71715093e+00 4.14250530e-02
1.89978808e-01 -1.70600981e-01 4.98384327e-01 -7.56142974e-01
-1.11196148e+00 1.25128508e+00 8.46516311e-01 4.26127315e-01
1.32757354e+00 -8.72298330e-02 5.96719325e-01 2.49699168e-02
2.29320019e-01 -1.04708636e+00 1.08617023e-01 2.80785263e-02
1.19013155e+00 -1.06850052e+00 1.44191999e-02 -9.07925814e-02
-5.90686321e-01 1.20012116e+00 2.49864347e-02 1.00631207e-01
9.21245217e-01 8.34417045e-02 -2.01185390e-01 -4.02333766e-01
-1.17291164e+00 3.45309637e-02 6.42029524e-01 1.56194240e-01
2.01615512e-01 1.31556895e-02 -1.80259049e-01 9.52259123e-01
1.98401049e-01 2.50707120e-01 -1.67002659e-02 8.48792970e-01
-1.72853366e-01 -1.08606851e+00 -1.83770999e-01 4.44526702e-01
-3.34324121e-01 3.27396810e-01 1.40727758e-01 3.70424896e-01
1.01679256e-02 7.09484696e-01 -3.15924324e-02 -3.85364383e-01
5.07562697e-01 4.92963880e-01 1.97826400e-01 -3.62715185e-01
-2.89484948e-01 -9.19794738e-02 -1.28641754e-01 -7.15642333e-01
-2.38961816e-01 -7.73838282e-01 -9.83544171e-01 6.95072860e-02
-2.02019781e-01 3.97805303e-01 9.15381312e-01 9.03530657e-01
3.58651102e-01 1.46096659e+00 7.61801064e-01 -7.67511785e-01
-8.07481349e-01 -8.98880064e-01 -8.23888123e-01 2.95055717e-01
4.40865934e-01 -1.09456766e+00 -4.33186501e-01 1.15790457e-01] | [6.803813934326172, 2.5328869819641113] |
507fd53c-f86e-456b-8ceb-6db11afe40fa | a-novel-long-term-iterative-mining-scheme-for | 2206.09564 | null | https://arxiv.org/abs/2206.09564v1 | https://arxiv.org/pdf/2206.09564v1.pdf | A Novel Long-term Iterative Mining Scheme for Video Salient Object Detection | The existing state-of-the-art (SOTA) video salient object detection (VSOD) models have widely followed short-term methodology, which dynamically determines the balance between spatial and temporal saliency fusion by solely considering the current consecutive limited frames. However, the short-term methodology has one critical limitation, which conflicts with the real mechanism of our visual system -- a typical long-term methodology. As a result, failure cases keep showing up in the results of the current SOTA models, and the short-term methodology becomes the major technical bottleneck. To solve this problem, this paper proposes a novel VSOD approach, which performs VSOD in a complete long-term way. Our approach converts the sequential VSOD, a sequential task, to a data mining problem, i.e., decomposing the input video sequence to object proposals in advance and then mining salient object proposals as much as possible in an easy-to-hard way. Since all object proposals are simultaneously available, the proposed approach is a complete long-term approach, which can alleviate some difficulties rooted in conventional short-term approaches. In addition, we devised an online updating scheme that can grasp the most representative and trustworthy pattern profile of the salient objects, outputting framewise saliency maps with rich details and smoothing both spatially and temporally. The proposed approach outperforms almost all SOTA models on five widely used benchmark datasets. | ['Chong Peng', 'Yuming Fang', 'Hengsen Wang', 'Chenglizhao Chen'] | 2022-06-20 | null | null | null | null | ['video-salient-object-detection'] | ['computer-vision'] | [ 2.06684425e-01 -1.31955624e-01 -2.37280935e-01 -1.38646392e-02
-3.74263585e-01 -1.11734182e-01 5.63982964e-01 -3.28937136e-02
-5.00511885e-01 6.39043093e-01 1.35676162e-02 -2.21715644e-02
-1.60983115e-01 -5.58176994e-01 -4.69278544e-01 -7.07546234e-01
7.44308624e-03 3.78629118e-02 1.10554016e+00 -3.49910110e-01
6.30243838e-01 2.73739338e-01 -2.18384171e+00 1.42569438e-01
9.67424750e-01 1.03819072e+00 8.40814590e-01 4.74526912e-01
-3.10033828e-01 7.63591826e-01 -3.12179446e-01 -5.01228608e-02
1.90449461e-01 -4.55731869e-01 -7.19663918e-01 1.25368133e-01
2.75864720e-01 -1.45422161e-01 -1.79604664e-02 1.34993124e+00
4.65357900e-01 2.30468392e-01 1.85349762e-01 -1.45381737e+00
-5.00143230e-01 5.29313326e-01 -1.04094267e+00 7.09956527e-01
3.23712051e-01 1.59237131e-01 8.17127645e-01 -1.13076305e+00
7.25035548e-01 1.17897737e+00 3.97332668e-01 4.10934538e-01
-7.58639157e-01 -5.98965585e-01 6.38964951e-01 6.08893752e-01
-1.38403845e+00 -4.37170982e-01 1.12899685e+00 -4.66692239e-01
6.69850647e-01 4.41110104e-01 9.57877278e-01 6.19919240e-01
8.80196691e-02 1.19688618e+00 9.21338141e-01 -3.49930257e-01
2.93616921e-01 5.13827354e-02 2.65493453e-01 5.31642914e-01
2.62063354e-01 -1.05441669e-02 -6.21474922e-01 1.21925501e-02
6.45818055e-01 2.11216643e-01 -3.15861106e-01 -6.95221603e-01
-1.39389169e+00 4.43393379e-01 4.07834589e-01 5.31890094e-01
-5.21419525e-01 -2.74937660e-01 4.18374121e-01 -5.73040769e-02
2.67635167e-01 1.41255379e-01 -1.99317664e-01 -4.98191081e-02
-1.35134089e+00 4.86249238e-01 2.33787030e-01 9.04494047e-01
7.84132123e-01 7.02864826e-02 -2.63739944e-01 4.99616355e-01
2.81509429e-01 3.51863712e-01 7.19579577e-01 -4.40557301e-01
3.05273950e-01 6.43746376e-01 3.43520373e-01 -1.35854232e+00
-2.56778866e-01 -4.77481604e-01 -5.20634770e-01 4.21482325e-01
2.59376705e-01 1.64512083e-01 -8.78687561e-01 1.72418988e+00
5.96950114e-01 3.38522017e-01 -1.76090255e-01 1.35326636e+00
8.02172482e-01 7.27913439e-01 2.25517780e-01 -7.34504223e-01
1.21412551e+00 -1.10137379e+00 -1.02086425e+00 -2.85701454e-01
1.07395723e-01 -9.32198405e-01 1.05303931e+00 3.86212707e-01
-1.20273459e+00 -6.96993649e-01 -1.26946867e+00 -3.79430093e-02
-3.64168465e-01 4.67595160e-02 5.31415045e-01 1.74360186e-01
-1.14349341e+00 3.45162749e-01 -4.89287347e-01 -5.67898810e-01
4.56741393e-01 1.81139261e-01 -2.50666477e-02 2.10207090e-01
-1.13608932e+00 9.24587369e-01 6.07963860e-01 3.19738120e-01
-8.51169467e-01 -5.52588046e-01 -5.62849522e-01 2.03881040e-02
9.32697713e-01 -4.52631086e-01 1.19308770e+00 -1.26339602e+00
-1.06650257e+00 7.19698071e-01 -4.61394072e-01 -5.03670633e-01
5.69265306e-01 -2.64252961e-01 -5.44796646e-01 1.32976964e-01
2.69894898e-01 7.02120304e-01 1.09421217e+00 -1.25976849e+00
-1.11517096e+00 -2.30077460e-01 7.36853331e-02 4.82445151e-01
-2.94074804e-01 1.44635350e-01 -6.38804734e-01 -7.28471935e-01
2.65288383e-01 -5.24297059e-01 -1.18577875e-01 2.08412245e-01
-1.38444304e-01 -3.39563161e-01 1.16202521e+00 -4.91693437e-01
1.74326599e+00 -2.17199969e+00 2.13014081e-01 -1.92582190e-01
2.58512080e-01 5.16813338e-01 1.31141111e-01 3.73481065e-02
-2.67576650e-02 -2.30863854e-01 -2.31543273e-01 -1.93024054e-01
-2.78421253e-01 -1.53535619e-01 -3.16134781e-01 2.51455158e-01
6.23633154e-03 7.35323429e-01 -1.23233426e+00 -1.05302250e+00
5.38365006e-01 7.75300935e-02 -1.63773164e-01 9.22916010e-02
-2.05493599e-01 -2.32015088e-01 -4.52883363e-01 7.16899574e-01
9.19454455e-01 -2.00314298e-01 -2.32980978e-02 -2.02725872e-01
-6.05248213e-01 -1.42288774e-01 -1.41719484e+00 1.66636693e+00
7.79891163e-02 6.03256524e-01 -8.42456298e-04 -1.11106062e+00
9.64396358e-01 3.00965942e-02 6.30307674e-01 -7.84524679e-01
3.61230522e-02 2.92606622e-01 -2.99684554e-01 -6.43907666e-01
8.49538088e-01 -9.50402170e-02 2.14436069e-01 2.30964333e-01
-9.14582908e-02 3.73302639e-01 2.62487113e-01 2.80892164e-01
6.68865204e-01 3.93032372e-01 6.88823938e-01 -5.97138524e-01
6.05764806e-01 4.80472386e-01 9.13308203e-01 6.65444791e-01
-6.33534968e-01 8.86277854e-01 2.55044699e-01 -4.94741023e-01
-8.33964109e-01 -8.32820356e-01 1.90215826e-01 9.08447385e-01
9.15409803e-01 -2.25825578e-01 -7.20345438e-01 -8.29169154e-01
-1.76287487e-01 4.14253503e-01 -8.73799562e-01 -7.66753778e-02
-6.18527114e-01 -4.49425995e-01 -1.15201641e-02 2.33163655e-01
6.76794052e-01 -1.37984085e+00 -1.24411058e+00 3.54718029e-01
-2.59978980e-01 -7.85450101e-01 -6.42699957e-01 -2.19847798e-01
-8.14550638e-01 -1.08967364e+00 -9.67353642e-01 -9.24515069e-01
4.98791099e-01 1.09357953e+00 8.99322808e-01 3.06753874e-01
-2.35419959e-01 -1.59255669e-01 -4.13244635e-01 -4.46763664e-01
1.38748139e-01 -2.36376554e-01 2.53999438e-02 3.02394420e-01
4.58487451e-01 -3.41567248e-01 -7.98136532e-01 4.15153891e-01
-9.93013620e-01 4.83259171e-01 6.22536957e-01 6.43908441e-01
6.89998686e-01 2.73942370e-02 8.94779801e-01 -5.15612900e-01
3.15160483e-01 -5.39417982e-01 -6.33594215e-01 5.14373541e-01
-6.38222337e-01 -2.19930768e-01 4.99300033e-01 -4.30321783e-01
-1.10197902e+00 -2.50352360e-03 1.93747774e-01 -7.68828332e-01
5.05608805e-02 4.32181984e-01 -2.18849763e-01 -1.41958715e-02
3.98895055e-01 5.91139555e-01 4.97643612e-02 -3.99314076e-01
3.83887067e-02 4.69960004e-01 7.08047211e-01 -7.91386142e-02
7.02709556e-01 4.85577404e-01 -1.51679888e-01 -6.98473871e-01
-7.46836662e-01 -7.04742908e-01 -5.72059929e-01 -6.16602898e-01
7.13953018e-01 -7.72225797e-01 -4.27463233e-01 6.39051020e-01
-1.06031871e+00 8.66911560e-02 -2.37927705e-01 2.02764004e-01
-3.65394920e-01 6.73533797e-01 -6.97067156e-02 -8.95765722e-01
-3.57648432e-01 -1.12442791e+00 7.11979806e-01 4.85206395e-01
-4.21767980e-02 -6.66888714e-01 -1.24671258e-01 2.01520808e-02
4.57249433e-01 2.45723993e-01 3.06331664e-01 -1.42896056e-01
-7.93022633e-01 3.08269083e-01 -3.78465354e-01 -5.45787066e-02
2.21941829e-01 8.47887397e-02 -8.37789774e-01 -2.65758663e-01
1.19264886e-01 1.04687139e-01 8.29225838e-01 5.78695953e-01
9.62137640e-01 -1.47401243e-01 -4.13439572e-01 3.04712236e-01
1.53441787e+00 4.39452022e-01 6.24641061e-01 6.82001173e-01
4.63479280e-01 4.62787777e-01 1.49397981e+00 5.27539492e-01
3.99780124e-01 9.14307415e-01 5.21501482e-01 -2.69635499e-01
-2.45605364e-01 -3.86357725e-01 2.10305959e-01 6.18581295e-01
-6.83459565e-02 -4.30129319e-02 -5.62891483e-01 9.14634347e-01
-2.31251287e+00 -1.23595333e+00 -1.20610714e-01 2.14260650e+00
6.69923544e-01 2.57891327e-01 3.53466660e-01 3.35279256e-01
1.06207180e+00 4.53162432e-01 -5.15572846e-01 -8.69213194e-02
-2.47016311e-01 -3.04226428e-01 1.23033024e-01 1.86407447e-01
-1.23683774e+00 7.76402354e-01 5.67798853e+00 1.02044475e+00
-1.37672389e+00 1.78959399e-01 2.77125746e-01 -1.10124990e-01
-2.13593528e-01 1.05270319e-01 -6.38644695e-01 8.05115938e-01
2.40206942e-01 -4.83904630e-01 3.46611589e-02 9.57431972e-01
4.36646402e-01 -6.24068379e-01 -6.63353264e-01 1.21732581e+00
1.88285023e-01 -1.56212306e+00 2.33323827e-01 -2.38102585e-01
6.93713725e-01 -3.04426312e-01 1.21823199e-01 6.98679015e-02
-3.05947870e-01 -5.04378021e-01 1.35551596e+00 5.70284486e-01
5.24014294e-01 -5.76542616e-01 7.31071413e-01 3.99255097e-01
-1.44962060e+00 -2.14341640e-01 -3.69813502e-01 -5.54302968e-02
3.48318636e-01 6.80899680e-01 -4.49306190e-01 7.72529125e-01
1.03417611e+00 9.09610629e-01 -6.79487824e-01 1.48124611e+00
1.49388149e-01 2.50795066e-01 -2.04952314e-01 -2.32578278e-01
3.98021340e-01 -6.22431450e-02 9.35027182e-01 1.20105791e+00
3.69359523e-01 2.02890024e-01 1.43083349e-01 6.89447820e-01
3.87303770e-01 1.09176986e-01 -4.66041952e-01 3.68874788e-01
3.71109337e-01 1.19554579e+00 -1.00563025e+00 -5.72499573e-01
-2.73779184e-01 6.89738035e-01 1.69527039e-01 2.46232480e-01
-9.32332158e-01 -4.16860878e-01 1.70526922e-01 2.30127990e-01
5.43184280e-01 1.10436231e-01 -4.42631513e-01 -1.18730235e+00
3.51376742e-01 -7.24695504e-01 6.29569292e-01 -9.38039899e-01
-7.47821212e-01 6.10586524e-01 1.30148321e-01 -1.67207158e+00
5.80145791e-02 -1.65897235e-02 -6.73152208e-01 5.91147363e-01
-1.70681572e+00 -1.04583919e+00 -5.20899415e-01 7.02021241e-01
1.02355134e+00 6.19777516e-02 1.29633173e-01 3.17922175e-01
-6.29004717e-01 3.36235464e-01 -2.27612436e-01 -3.73304158e-01
4.60696727e-01 -9.14352834e-01 5.22097945e-02 1.44365382e+00
-5.05405739e-02 4.32233840e-01 1.04906070e+00 -7.11141288e-01
-1.22790384e+00 -7.93540061e-01 9.71822977e-01 1.27618223e-01
4.65751916e-01 -1.07073247e-01 -1.08762264e+00 2.50010461e-01
7.64775127e-02 7.46014491e-02 -1.54626504e-01 -2.50526309e-01
2.56933063e-01 -2.77075529e-01 -1.18385780e+00 7.13044703e-01
1.21865928e+00 -1.21029660e-01 -9.73370731e-01 -2.00257584e-01
6.66105986e-01 -2.58792549e-01 -7.38960803e-02 5.27647316e-01
4.58254129e-01 -1.14046848e+00 8.22706342e-01 -2.19795048e-01
3.16838354e-01 -9.43011582e-01 1.65837437e-01 -8.44330907e-01
-3.40621889e-01 -8.55786681e-01 -4.07606125e-01 1.13523531e+00
1.00095332e-01 -3.01985472e-01 6.06777608e-01 1.94744214e-01
-2.33930498e-01 -9.20716524e-01 -1.03424549e+00 -5.41309357e-01
-6.41122401e-01 -1.45042345e-01 6.65946960e-01 8.81222427e-01
7.72232115e-02 -6.85263723e-02 -4.92730260e-01 1.06646441e-01
8.32372248e-01 5.31170011e-01 4.59045678e-01 -1.08654523e+00
1.13349386e-01 -5.00974298e-01 -5.52206099e-01 -1.17240310e+00
-3.27524692e-01 -3.68321657e-01 2.66882777e-01 -1.38976407e+00
5.92752874e-01 -5.33201694e-01 -6.82569444e-01 3.02844316e-01
-5.58061957e-01 1.23633809e-01 2.54519761e-01 2.59227902e-01
-1.06780481e+00 5.76141596e-01 1.29261708e+00 -1.65798552e-02
-3.36778909e-01 -1.44545019e-01 -7.31527746e-01 7.08282709e-01
7.00118423e-01 -2.81871170e-01 -6.86956644e-01 -3.12585294e-01
-7.54581466e-02 9.12901536e-02 5.38931310e-01 -1.14635706e+00
5.88225424e-01 -6.38920367e-01 1.16682418e-01 -1.11030972e+00
1.40337899e-01 -6.67172492e-01 1.48245707e-01 3.88104379e-01
-2.63660308e-02 1.14639334e-01 2.39754975e-01 5.93758643e-01
-4.34278727e-01 -1.43074542e-01 9.04038966e-01 -1.14558496e-01
-1.51373005e+00 2.24476963e-01 -1.46002918e-01 -5.56091480e-02
1.44492424e+00 -7.18123853e-01 -2.53676176e-01 -3.37221473e-02
-5.90294898e-01 3.67270112e-01 6.24657273e-01 6.50937796e-01
9.73783910e-01 -1.26202726e+00 -4.57883805e-01 1.47734269e-01
1.11331008e-01 -7.27151334e-02 6.97453499e-01 9.57906961e-01
-2.62470007e-01 2.87743300e-01 -5.60001254e-01 -7.20280111e-01
-1.32588685e+00 1.11640286e+00 8.48366469e-02 2.96356287e-02
-7.58976161e-01 6.04131818e-01 3.07750136e-01 3.98674011e-01
1.82843298e-01 -6.43888786e-02 -7.15681255e-01 1.77516043e-01
6.44269347e-01 3.72440368e-01 -8.01937729e-02 -8.10080230e-01
-4.80200529e-01 7.32988954e-01 -9.04433057e-03 4.14945334e-02
1.09633338e+00 -5.70866585e-01 -4.10764769e-04 5.13356328e-01
7.72661209e-01 -3.54010493e-01 -1.26068497e+00 -2.94803381e-01
-3.22655961e-03 -8.48121345e-01 1.76125437e-01 -5.10742724e-01
-1.04186571e+00 6.32182479e-01 7.12854803e-01 1.34630546e-01
1.45817041e+00 -1.04360834e-01 8.64456356e-01 -3.59303802e-02
5.78098416e-01 -1.38730156e+00 -9.36043262e-02 1.55732334e-01
9.14337397e-01 -1.27118182e+00 2.73179501e-01 -4.46878880e-01
-7.31316864e-01 8.71578813e-01 8.05858612e-01 6.77263364e-02
5.61670423e-01 -6.92180470e-02 -6.62344918e-02 -1.81057245e-01
-7.95161247e-01 -2.69487977e-01 3.71718138e-01 3.26003432e-01
-1.82634503e-01 -9.50671285e-02 -6.70313597e-01 5.32231867e-01
2.41300702e-01 3.44478190e-01 5.15780985e-01 1.13722968e+00
-8.40812266e-01 -5.97971797e-01 -3.89010787e-01 3.10694456e-01
-3.19365114e-01 1.22012995e-01 2.08945312e-02 7.03720093e-01
3.06946933e-01 9.11320686e-01 -5.69735914e-02 -4.22347426e-01
3.04833680e-01 -3.16026807e-01 2.67244652e-02 -2.73377657e-01
-3.04628402e-01 -1.79956015e-02 -2.26769730e-01 -6.59511268e-01
-7.96765566e-01 -9.30663526e-01 -1.24273014e+00 -1.29417181e-01
-5.49782813e-01 1.75006568e-01 3.02424937e-01 7.98470855e-01
3.36493373e-01 3.12108427e-01 7.01024473e-01 -1.03512871e+00
-3.60341400e-01 -7.30775356e-01 -4.55619752e-01 4.98363197e-01
5.08611262e-01 -1.11662495e+00 -3.07485551e-01 2.23874953e-02] | [9.707141876220703, -0.4311903119087219] |
38e2a2de-6f59-4130-bf28-31ef6ee049e1 | license-plate-recognition-with-compressive | 1902.05386 | null | http://arxiv.org/abs/1902.05386v1 | http://arxiv.org/pdf/1902.05386v1.pdf | License Plate Recognition with Compressive Sensing Based Feature Extraction | License plate recognition is the key component to many automatic traffic
control systems. It enables the automatic identification of vehicles in many
applications. Such systems must be able to identify vehicles from images taken
in various conditions including low light, rain, snow, etc. In order to reduce
the complexity and cost of the hardware required for such devices, the
algorithm should be as efficient as possible. This paper proposes a license
plate recognition system which uses a new approach based on compressive sensing
techniques for dimensionality reduction and feature extraction. Dimensionality
reduction will enable precise classification with less training data while
demanding less computational power. Based on the extracted features, character
recognition and classification is done by a Support Vector Machine classifier. | ['Nikola Vukovic', 'Andrej Jokic'] | 2019-02-07 | null | null | null | null | ['license-plate-recognition'] | ['computer-vision'] | [ 5.39052784e-01 -8.30604851e-01 -7.12298900e-02 -1.97734237e-01
-3.31279516e-01 -5.38272738e-01 5.33386588e-01 -5.27111471e-01
-3.06226909e-01 6.33823276e-01 -2.72250444e-01 -2.74878085e-01
-1.24795660e-01 -8.61320734e-01 -1.26321912e-01 -8.94977212e-01
6.64068103e-01 3.18447381e-01 4.62631464e-01 7.91469403e-03
6.75664544e-01 1.06325078e+00 -1.88915384e+00 -1.00288875e-01
6.77859128e-01 1.06257558e+00 3.62414926e-01 4.94119972e-01
2.73941495e-02 6.15834177e-01 -3.02717090e-01 -1.98672161e-01
5.27329922e-01 -3.49886090e-01 1.86606571e-01 5.11511147e-01
1.04127534e-01 -3.68451327e-01 -7.29663372e-01 1.11324072e+00
1.96590066e-01 3.18836391e-01 1.00565732e+00 -7.85867631e-01
-2.24104181e-01 -3.48356336e-01 -2.86340892e-01 3.28598946e-01
9.26970541e-02 -8.00667424e-03 9.06339437e-02 -1.04268050e+00
2.77925581e-01 4.16097820e-01 4.49900657e-01 5.33373594e-01
-8.20847452e-01 -4.98762757e-01 -8.17290723e-01 7.63754785e-01
-1.71463954e+00 -8.94206107e-01 9.61893618e-01 -3.61804396e-01
7.12434649e-01 4.64273006e-01 3.80087703e-01 3.46099049e-01
2.67422289e-01 3.51540625e-01 1.00652218e+00 -6.18400514e-01
4.32036012e-01 2.67587811e-01 8.29514191e-02 5.20671308e-01
6.81019366e-01 -9.76086631e-02 -4.20866162e-02 1.58343062e-01
4.84389931e-01 6.37986302e-01 -1.00324474e-01 -1.44783214e-01
-5.76989472e-01 7.25354254e-01 -3.12669516e-01 4.43918705e-01
-5.22914410e-01 -2.49941155e-01 6.17984198e-02 7.08025470e-02
-6.06188625e-02 1.30439728e-01 3.89883488e-01 -1.98810667e-01
-1.13646805e+00 -2.35441178e-01 6.60555184e-01 7.44941473e-01
5.93984723e-01 3.03652823e-01 3.59525740e-01 9.17475224e-01
3.96513045e-01 1.14810574e+00 5.27130604e-01 -7.48253644e-01
4.89067763e-01 6.02423787e-01 2.84921415e-02 -9.83625948e-01
-1.30782619e-01 -2.09918898e-02 -7.79900134e-01 3.33437860e-01
-3.19578894e-03 -3.22205722e-02 -8.34743679e-01 5.14906406e-01
-5.85293695e-02 3.47918689e-01 1.95634007e-01 8.83641720e-01
4.37459499e-01 9.11163867e-01 -4.00347263e-01 -2.56186485e-01
1.32386220e+00 -5.05433083e-01 -8.01221192e-01 -3.34444731e-01
3.29176664e-01 -9.14451957e-01 4.01322812e-01 2.89242148e-01
-4.73804027e-01 -3.04169893e-01 -1.29929912e+00 3.94342124e-01
-1.90343022e-01 6.51750624e-01 2.88614154e-01 1.05788779e+00
-3.95095050e-01 1.59910813e-01 -5.29156208e-01 -2.18858123e-01
2.74035901e-01 5.33075929e-01 -4.53568786e-01 -5.51350117e-01
-7.54556537e-01 1.01945341e+00 3.36960843e-03 3.37574542e-01
3.48959193e-02 1.77435338e-01 -8.81375313e-01 1.10211059e-01
-7.52966180e-02 6.82254285e-02 6.26820743e-01 -5.33393323e-01
-1.63729751e+00 5.71167290e-01 -4.42608684e-01 -5.13079464e-01
8.24207887e-02 1.83206603e-01 -8.71716917e-01 5.21482706e-01
-1.18624181e-01 -3.02578479e-01 1.20707071e+00 -4.71119612e-01
-5.09270370e-01 -5.50328910e-01 -8.01412582e-01 1.93207026e-01
-1.69439450e-01 1.55269310e-01 -1.36197478e-01 -3.21932845e-02
4.04870093e-01 -1.15383589e+00 9.57639068e-02 -4.34297055e-01
-6.91776276e-02 -1.22787319e-01 1.58105373e+00 -5.41491985e-01
7.90497363e-01 -2.61676979e+00 -6.57279849e-01 5.63187003e-01
-1.86837893e-02 7.71839917e-01 1.85010716e-01 3.65942299e-01
2.97764599e-01 -4.43128020e-01 1.44341346e-02 -1.30833089e-02
-1.55514151e-01 1.33835733e-01 -1.96436957e-01 9.59823608e-01
9.11827683e-02 2.88736224e-01 -8.52689072e-02 -2.60682493e-01
9.28351045e-01 4.34316963e-01 3.37948613e-02 -1.50300398e-01
5.47641754e-01 -3.34700122e-02 -5.50682068e-01 6.76385760e-01
8.74733627e-01 5.61095849e-02 -1.88736379e-01 5.12525290e-02
-3.67552519e-01 -4.53527272e-02 -1.33551407e+00 5.09778976e-01
-2.45958164e-01 1.32782722e+00 8.21213648e-02 -1.24506402e+00
1.26285195e+00 4.08562481e-01 5.07133305e-01 -8.39318335e-01
3.22033733e-01 3.98783624e-01 -1.49684817e-01 -8.19333553e-01
4.28724229e-01 -2.83088207e-01 8.51591900e-02 7.75831044e-02
-5.79556048e-01 3.87143753e-02 2.36125886e-01 -1.81537971e-01
9.11198914e-01 -5.99425733e-01 1.98923662e-01 5.49275465e-02
9.56777692e-01 1.24215826e-01 5.06681621e-01 2.72588223e-01
-1.83089897e-01 2.94113666e-01 -2.86101669e-01 -4.21181887e-01
-1.08638668e+00 -6.49179101e-01 -3.29509735e-01 2.44055465e-01
4.65036660e-01 3.64228994e-01 -4.94241327e-01 -1.63748655e-02
-7.12006912e-02 4.95197296e-01 1.86332408e-02 -8.01156908e-02
-6.14940405e-01 -4.79770899e-01 5.89546800e-01 2.11341336e-01
8.94164264e-01 -5.34406364e-01 -4.58687603e-01 1.45517081e-01
1.24260470e-01 -1.38356328e+00 -2.24810451e-01 -3.29342604e-01
-7.39948273e-01 -9.53462064e-01 -4.99848932e-01 -9.94541824e-01
7.36081541e-01 8.53777587e-01 2.33241647e-01 7.96207413e-02
-2.00182766e-01 1.51477113e-01 -2.86910057e-01 -3.35818410e-01
-3.56501639e-01 -2.62405366e-01 3.50359946e-01 4.91556436e-01
9.36001778e-01 -2.82403052e-01 -4.40716356e-01 5.09366751e-01
-6.71421230e-01 -2.51788348e-01 8.97627115e-01 4.13959384e-01
3.85746717e-01 6.84046686e-01 3.08279872e-01 -3.83211613e-01
5.30270636e-01 -2.64854103e-01 -1.04894209e+00 6.36183177e-05
-5.17943680e-01 -2.23877206e-01 8.00210595e-01 -2.69700289e-02
-8.52080882e-01 4.78133142e-01 -1.87867314e-01 -3.75506520e-01
-4.22783971e-01 1.60696760e-01 -2.72547841e-01 -6.22536480e-01
5.35309315e-01 1.03910196e+00 3.33112240e-01 -3.64343673e-01
-4.06704217e-01 1.29836118e+00 5.04045486e-01 2.73167074e-01
1.00120795e+00 3.91737312e-01 5.26715457e-01 -1.76080012e+00
8.02113712e-02 -8.21312249e-01 -4.43641126e-01 -5.18103778e-01
7.36002505e-01 -7.18968809e-01 -6.47248089e-01 6.92879260e-01
-7.81377017e-01 5.91257572e-01 1.72409624e-01 1.13495314e+00
-1.65454477e-01 6.27258003e-01 -1.84757218e-01 -8.99809122e-01
-3.00540775e-02 -1.00925589e+00 5.45861304e-01 4.76901114e-01
1.48160741e-01 -6.24248743e-01 -1.99848432e-02 7.04096317e-01
6.18639886e-01 -1.08630389e-01 5.29293537e-01 -5.80738604e-01
-8.43995690e-01 -9.73439634e-01 -1.92248255e-01 5.39803445e-01
3.71651530e-01 -5.43063274e-03 -6.93794012e-01 -3.86149995e-02
2.98281908e-01 1.56962037e-01 7.94437528e-01 4.15245980e-01
6.89499557e-01 -3.34802032e-01 -2.35178083e-01 4.58843559e-01
1.62082314e+00 6.09397113e-01 1.05695248e+00 1.68710142e-01
4.15068060e-01 1.95162714e-01 4.43759918e-01 2.30823889e-01
-9.75221172e-02 5.62640190e-01 -9.60795302e-03 3.91538262e-01
5.10712601e-02 1.32096648e-01 2.72839367e-01 7.97003686e-01
-1.49146080e-01 -1.46797940e-01 -8.29622746e-01 2.27685332e-01
-1.37387300e+00 -1.52871978e+00 -3.55664194e-01 2.19275570e+00
-8.64452124e-02 -1.74233630e-01 -1.07377023e-01 7.57349253e-01
9.94865537e-01 -7.38541260e-02 -3.59244257e-01 -4.55704957e-01
-1.09168150e-01 -1.43968716e-01 8.87330770e-01 3.49896163e-01
-9.68468726e-01 5.39013505e-01 6.37228680e+00 6.68103933e-01
-1.58140469e+00 -3.60558987e-01 1.28757983e-01 1.93198115e-01
1.99982175e-03 -2.22784802e-01 -8.15255761e-01 8.22416544e-01
8.26800048e-01 -1.21173665e-01 3.98526460e-01 7.03441501e-01
2.28224084e-01 -4.60953295e-01 -5.81235588e-01 1.42871320e+00
3.56409132e-01 -1.51305830e+00 -1.29631162e-01 3.21856230e-01
4.01684493e-01 1.84674382e-01 2.51124322e-01 -9.02621597e-02
-5.86599767e-01 -6.89102054e-01 1.41802251e-01 6.42767072e-01
8.12052429e-01 -8.02214265e-01 8.91847610e-01 6.11103415e-01
-9.59925592e-01 -8.43481943e-02 -4.16419238e-01 -3.09412211e-01
-2.82978583e-02 5.75216770e-01 -9.37379837e-01 -1.32367253e-01
-2.13821270e-02 3.85129601e-01 3.02512851e-03 1.11702979e+00
1.39775351e-01 7.55180120e-01 -5.31987488e-01 -5.75090587e-01
2.46427488e-02 -6.52912199e-01 7.37492681e-01 1.06416321e+00
6.88614666e-01 4.62580562e-01 -1.25469550e-01 3.64281923e-01
9.35377255e-02 -2.07670685e-02 -9.94389713e-01 -1.57993197e-01
6.86165690e-01 8.75074327e-01 -7.78854728e-01 -2.50967503e-01
-6.31050944e-01 9.70288992e-01 -5.60817838e-01 9.74239036e-02
-5.56889057e-01 -7.82441020e-01 5.20852804e-01 5.76280534e-01
5.30123711e-01 -7.22380519e-01 -2.28709862e-01 -1.02614665e+00
1.21913768e-01 -3.37380171e-01 -1.92425728e-01 -2.88841486e-01
-7.39986420e-01 2.91905493e-01 -3.79564762e-01 -1.74870169e+00
-4.74149138e-01 -8.00919533e-01 -5.71489871e-01 8.86103630e-01
-1.30803931e+00 -4.43549961e-01 -1.46801114e-01 8.18899095e-01
6.07647955e-01 -9.12768185e-01 5.85911274e-01 3.59146774e-01
-7.37724781e-01 1.93280444e-01 8.72817993e-01 2.96019793e-01
2.66506910e-01 -4.14399922e-01 -1.67552441e-01 1.34238338e+00
-6.67112321e-02 2.50268489e-01 5.56098461e-01 -5.80121636e-01
-1.77919698e+00 -8.51244211e-01 1.02557170e+00 1.40369266e-01
1.83174059e-01 1.65287387e-02 -5.74229419e-01 6.64289147e-02
-1.98257357e-01 1.37229085e-01 1.03467631e+00 -5.21100521e-01
1.69571683e-01 -5.57114244e-01 -1.48795235e+00 2.48501465e-01
1.46065235e-01 -5.35444915e-01 -5.33354282e-01 1.84854344e-01
-3.21965873e-01 -9.26807337e-03 -4.68999535e-01 -2.65710540e-02
6.18476927e-01 -7.08826661e-01 6.14832282e-01 1.61513209e-01
-1.70907727e-03 -7.71161437e-01 -4.00756121e-01 -7.20521152e-01
-5.47703803e-01 -3.07385653e-01 1.38884008e-01 9.08610761e-01
2.36487448e-01 -8.59727621e-01 1.09080017e+00 8.62103760e-01
1.69119582e-01 -1.55167475e-01 -1.26346934e+00 -1.00121439e+00
-8.44251633e-01 -3.14246595e-01 4.38001394e-01 5.90051234e-01
-1.27453879e-01 3.06184500e-01 -5.35546362e-01 4.50165480e-01
9.31214809e-01 1.41076788e-01 4.32901978e-01 -1.34472477e+00
1.42627209e-01 -2.08962932e-01 -1.25630879e+00 -7.43635237e-01
1.62574109e-02 -5.50894856e-01 -5.30035757e-02 -1.51842308e+00
-3.25697400e-02 -3.28268975e-01 -3.08722649e-02 7.67880306e-02
4.47399855e-01 8.00805569e-01 2.23161295e-01 8.30327511e-01
-4.63132143e-01 3.96034241e-01 6.53103650e-01 -2.46819302e-01
-1.06020393e-02 4.99896318e-01 -2.87487805e-01 7.01956570e-01
9.97162223e-01 -5.90974689e-01 -3.22944403e-01 -2.90147752e-01
-1.21111363e-01 8.56221318e-02 8.26426148e-02 -1.40006387e+00
6.09676898e-01 -1.66874588e-01 6.54414117e-01 -6.21413827e-01
5.96625268e-01 -1.31536138e+00 2.34407291e-01 6.70952976e-01
4.30585414e-01 -2.67278105e-01 -7.90312961e-02 4.90826219e-01
-4.97933656e-01 -6.28345013e-01 7.25806415e-01 1.41473487e-01
-1.20474756e+00 -6.76005781e-02 -1.19506466e+00 -7.05904126e-01
1.44614398e+00 -9.23951685e-01 -4.44468074e-02 -3.22446853e-01
-2.85635173e-01 -2.76015252e-01 5.21328747e-01 1.02533452e-01
1.00133884e+00 -1.37549877e+00 -6.47144437e-01 7.34022141e-01
-8.70888494e-03 -6.65206909e-01 3.70542586e-01 4.83834028e-01
-1.03160346e+00 8.29856396e-01 -4.46261495e-01 -4.70411628e-01
-1.54820371e+00 2.93682635e-01 1.31992027e-01 3.24653387e-01
-5.47845602e-01 1.98084742e-01 -7.56286085e-01 2.82316595e-01
-3.46835077e-01 3.69568884e-01 -4.56226438e-01 -2.27007583e-01
9.09674466e-01 8.71260226e-01 3.32664728e-01 -9.86785412e-01
-5.93612194e-01 9.31741953e-01 4.98016290e-02 -5.88010997e-02
1.18982244e+00 -7.51498491e-02 1.44535983e-02 -1.19630620e-01
1.12867975e+00 4.21585351e-01 -1.08577549e+00 -1.86254486e-01
-2.02285722e-01 -8.32765460e-01 3.25846255e-01 -1.52588874e-01
-8.12913179e-01 6.44917965e-01 6.21866584e-01 2.35334367e-01
1.21339548e+00 -4.50468302e-01 6.61543369e-01 9.35120106e-01
5.27142048e-01 -1.43711340e+00 -7.05255151e-01 3.83912355e-01
4.57280636e-01 -1.24994886e+00 2.06743434e-01 -4.48443294e-01
-4.85216588e-01 1.42189777e+00 -4.50582895e-03 -4.07399833e-01
7.85427570e-01 4.10712808e-01 -4.61419597e-02 1.07691750e-01
-3.14435571e-01 -1.65211290e-01 3.28969032e-01 5.54966450e-01
-7.83051997e-02 1.80992842e-01 -4.56870973e-01 2.08580449e-01
1.21585786e-01 1.00747377e-01 5.92167854e-01 8.89956295e-01
-9.65345740e-01 -9.64773297e-01 -8.68364573e-01 6.91091537e-01
-3.62909555e-01 2.08561853e-01 -2.55869687e-01 3.56895059e-01
-2.14814469e-02 1.08144891e+00 1.73557848e-01 -5.43247283e-01
1.24091566e-01 -2.83588329e-03 4.02210176e-01 -2.58452594e-01
4.59174484e-01 -2.17690021e-01 2.80021485e-02 -8.81461129e-02
-4.03906047e-01 -9.58141327e-01 -1.13799417e+00 -6.06696069e-01
-3.05684090e-01 2.10944504e-01 1.17219186e+00 1.18306541e+00
3.32752228e-01 -7.87783787e-02 1.09004366e+00 -6.73168480e-01
-4.80247498e-01 -6.49421334e-01 -1.02791488e+00 4.90682609e-02
2.21418023e-01 -4.61331606e-01 -2.19961107e-01 4.86260541e-02] | [9.780550956726074, -5.013743877410889] |
43568814-b637-475f-a6b0-4e3359f53c42 | equivariant-spherical-cnn-for-data-efficient | 2307.03298 | null | https://arxiv.org/abs/2307.03298v1 | https://arxiv.org/pdf/2307.03298v1.pdf | Equivariant Spherical CNN for Data Efficient and High-Performance Medical Image Processing | This work highlights the significance of equivariant networks as efficient and high-performance approaches for tomography applications. Our study builds upon the limitations of Convolutional Neural Networks (CNNs), which have shown promise in post-processing various medical imaging systems. However, the efficiency of conventional CNNs heavily relies on an undiminished and proper training set. To tackle this issue, in this study, we introduce an equivariant network, aiming to reduce CNN's dependency on specific training sets. We evaluate the efficacy of equivariant CNNs on spherical signals for tomographic medical imaging problems. Our results demonstrate superior quality and computational efficiency of spherical CNNs (SCNNs) in denoising and reconstructing benchmark problems. Furthermore, we propose a novel approach to employ SCNNs as a complement to conventional image reconstruction tools, enhancing the outcomes while reducing reliance on the training set. Across all cases, we observe a significant decrease in computational costs while maintaining the same or higher quality of image processing using SCNNs compared to CNNs. Additionally, we explore the potential of this network for broader tomography applications, particularly those requiring omnidirectional representation. | ['Hamid Sabet', 'Yuemeng Feng', 'Amirreza Hashemi'] | 2023-07-06 | null | null | null | null | ['image-reconstruction', 'denoising'] | ['computer-vision', 'computer-vision'] | [ 2.74584889e-01 9.85060036e-02 5.78238308e-01 -3.77459735e-01
-7.58517444e-01 -2.18328193e-01 2.72026390e-01 -5.68439662e-02
-7.89775610e-01 4.86199021e-01 3.39185297e-01 -3.12440127e-01
-3.28893900e-01 -8.90748143e-01 -7.39711344e-01 -6.46691561e-01
-1.47597222e-02 1.60126165e-01 1.99962854e-01 -3.95211399e-01
-1.46819144e-01 9.07646298e-01 -1.04705560e+00 2.67809778e-01
4.43116486e-01 9.78301823e-01 1.25446707e-01 5.87662518e-01
3.48875642e-01 6.41058087e-01 -2.62970418e-01 -3.25586915e-01
3.75344485e-01 -8.45162570e-02 -6.14749789e-01 -1.23061158e-01
3.81130755e-01 -4.69610035e-01 -6.22247875e-01 9.51911390e-01
9.63914931e-01 -2.14712359e-02 4.36917305e-01 -7.28449821e-01
-4.38964605e-01 5.08395612e-01 -1.23811938e-01 4.87756521e-01
8.24597850e-03 2.26139799e-01 5.99750400e-01 -9.72756863e-01
6.40305758e-01 8.33336115e-01 1.29769528e+00 3.58496666e-01
-1.17359579e+00 -4.63160425e-01 -4.00462270e-01 9.58019495e-02
-1.18128204e+00 -7.13351727e-01 5.73707640e-01 -1.66866556e-01
8.40320289e-01 2.35555902e-01 6.59693956e-01 8.66730690e-01
3.90311599e-01 5.45796156e-01 9.70773339e-01 -2.52046496e-01
1.63775817e-01 -2.98712581e-01 -1.17579222e-01 7.44485140e-01
3.15826744e-01 -3.06719448e-02 -3.54455590e-01 1.05321996e-01
1.10402048e+00 -3.44218053e-02 -4.91513520e-01 -3.20371777e-01
-1.49253452e+00 7.14037657e-01 6.09425366e-01 4.67295617e-01
-7.27563620e-01 3.98869276e-01 5.50696313e-01 1.87592447e-01
5.84174395e-01 7.15134680e-01 -2.15227410e-01 6.52265623e-02
-1.16993976e+00 -3.69597636e-02 4.67329234e-01 5.91999829e-01
2.96173424e-01 3.39310378e-01 -3.31157655e-01 7.84340739e-01
-1.08525880e-01 3.00681829e-01 3.66763800e-01 -9.23294008e-01
1.67785168e-01 4.53950822e-01 -3.30949038e-01 -1.04455650e+00
-8.40489566e-01 -1.05762291e+00 -1.19310689e+00 2.54435360e-01
1.81496084e-01 5.07853292e-02 -9.98358011e-01 1.68663323e+00
2.64085412e-01 3.41090620e-01 1.27111509e-01 8.59281003e-01
1.08554685e+00 2.39048913e-01 -2.08535120e-01 5.94905615e-02
1.29170096e+00 -7.63830662e-01 -5.62455714e-01 3.06864232e-01
5.24471104e-01 -8.53928924e-01 9.97282445e-01 5.00914335e-01
-1.43562925e+00 -1.54427841e-01 -1.10553670e+00 -2.22784698e-01
-7.27723315e-02 9.22727138e-02 7.25794017e-01 7.93199301e-01
-1.55968583e+00 9.31985080e-01 -1.18693399e+00 -2.55722463e-01
6.46277785e-01 7.43245244e-01 -4.92067814e-01 -1.49676725e-01
-8.27881932e-01 6.98461890e-01 7.19977543e-02 2.85711646e-01
-9.51089859e-01 -1.11436939e+00 -8.73798728e-01 1.76574811e-01
3.01855933e-02 -1.00378942e+00 1.15072441e+00 -7.54896224e-01
-1.40858388e+00 6.09050691e-01 1.50639385e-01 -6.54010832e-01
8.65956008e-01 1.27419174e-01 -2.00272724e-01 4.74644005e-01
1.20418102e-01 6.31699860e-01 6.02740526e-01 -1.02856052e+00
-1.49080083e-01 -2.55560368e-01 1.02009676e-01 4.88514751e-02
-5.60630918e-01 -1.01855606e-01 -6.44961655e-01 -8.78317595e-01
5.45975864e-01 -8.62995625e-01 -4.82311547e-01 2.23775357e-01
-3.01691920e-01 5.76151609e-02 6.04699850e-01 -6.31569088e-01
6.26571536e-01 -2.07803106e+00 1.05913710e-02 3.54501724e-01
6.25313997e-01 2.01020285e-01 -1.59074858e-01 3.38722616e-02
-2.42032483e-01 -4.37220931e-02 -4.48518515e-01 -4.45615560e-01
-3.56550962e-01 3.52400631e-01 2.68000960e-01 7.60009408e-01
1.74697146e-01 1.06273437e+00 -7.24733949e-01 -3.92711014e-01
5.06683409e-01 9.27990079e-01 -9.48848546e-01 -2.69388109e-02
2.47571155e-01 7.06358194e-01 -2.38477603e-01 5.34960806e-01
7.92488158e-01 -4.71308887e-01 7.90225342e-02 -7.46510208e-01
-9.78095382e-02 1.32765606e-01 -9.17841196e-01 1.65801203e+00
-7.27117836e-01 6.36516333e-01 2.55030513e-01 -1.28732109e+00
4.76237744e-01 5.28448880e-01 9.95933890e-01 -9.76952255e-01
3.29383463e-01 3.72444242e-01 3.30069475e-02 -6.22380435e-01
4.54306602e-01 -3.43990713e-01 3.97333413e-01 2.99378574e-01
2.66132772e-01 -1.13507830e-01 -3.95391602e-03 1.15589984e-01
1.28294826e+00 -2.34555677e-01 4.20301817e-02 -5.48064590e-01
3.85448962e-01 -1.08527862e-01 2.97702610e-01 8.58247757e-01
-2.80212872e-02 1.10827029e+00 3.80776316e-01 -6.55811071e-01
-1.28253150e+00 -9.95546937e-01 -4.33595389e-01 6.51518226e-01
-1.96307689e-01 -9.56662670e-02 -8.19957316e-01 -2.53404319e-01
-4.73150343e-01 1.82785586e-01 -7.49723196e-01 6.31410554e-02
-9.17327464e-01 -1.10699451e+00 8.14948916e-01 5.91680467e-01
5.99373698e-01 -9.62102771e-01 -6.65887237e-01 3.02956760e-01
-3.89697284e-01 -1.40308905e+00 -8.86163041e-02 2.44214118e-01
-1.16025805e+00 -1.03054476e+00 -9.85737562e-01 -6.14415050e-01
7.87008941e-01 3.98803025e-01 1.15958881e+00 1.33232325e-01
-3.57896894e-01 5.95085680e-01 -2.07328662e-01 -2.06241652e-01
-2.72663862e-01 2.40542114e-01 -6.27657399e-02 -1.78857073e-01
-4.04616147e-01 -9.91666496e-01 -1.08369124e+00 2.48913579e-02
-1.13313544e+00 7.71774724e-02 6.31447554e-01 8.48990679e-01
5.54482639e-01 -9.73384678e-02 4.18891639e-01 -9.92354274e-01
5.05736649e-01 -4.90135223e-01 -3.01015615e-01 -8.92737508e-02
-5.03887832e-01 1.17727993e-02 7.23001242e-01 -8.75856131e-02
-8.09180439e-01 1.96490549e-02 -6.01525426e-01 -3.46752256e-01
1.99991256e-01 6.70206189e-01 4.32855099e-01 -5.84957242e-01
7.02956975e-01 2.68641323e-01 2.26192281e-01 -2.85441816e-01
1.38157494e-02 9.64474380e-02 7.77305067e-01 -4.67750043e-01
6.73643589e-01 1.09929156e+00 3.23574007e-01 -9.61944878e-01
-6.06744349e-01 -3.86893123e-01 -6.15066469e-01 -2.07600087e-01
9.99096215e-01 -8.61107290e-01 -8.06245804e-01 3.80196512e-01
-1.02330375e+00 -1.48484275e-01 -3.50218475e-01 5.83441257e-01
-4.34251964e-01 6.45243704e-01 -7.67831564e-01 -1.72959194e-01
-6.80807352e-01 -1.51441288e+00 1.00208247e+00 -3.87644880e-02
1.51700620e-02 -1.14154780e+00 -1.07967153e-01 3.66539776e-01
9.13371265e-01 4.88735914e-01 8.47115695e-01 -4.43664759e-01
-6.54753268e-01 -7.52187669e-02 -4.99542087e-01 4.70244139e-01
-1.18382052e-02 -4.63416338e-01 -9.88431811e-01 -5.16654611e-01
1.30469382e-01 -1.76957995e-01 8.71153414e-01 7.40406930e-01
1.46632791e+00 -3.27197462e-02 1.66464105e-01 1.24179888e+00
1.46148241e+00 -1.53045341e-01 7.77471244e-01 5.21527231e-01
5.96436441e-01 3.52251172e-01 -1.19509727e-01 4.35520023e-01
2.59271860e-01 4.70119029e-01 7.34655380e-01 -8.54153454e-01
-3.53413761e-01 2.20389247e-01 -4.84214909e-02 1.00950539e+00
-4.18824255e-01 -3.85904685e-02 -9.40325022e-01 6.18767619e-01
-1.53340697e+00 -4.92806822e-01 -3.64370853e-01 1.81992102e+00
5.95257223e-01 -1.88444540e-01 -3.14467490e-01 2.03903526e-01
2.79756755e-01 -3.24646421e-02 -4.16152894e-01 -6.00028299e-02
-1.55311242e-01 8.17905188e-01 6.33641481e-01 2.40420491e-01
-1.11424577e+00 5.38335145e-01 6.75325155e+00 7.54878640e-01
-1.50141895e+00 3.63505274e-01 6.27423644e-01 -1.41275197e-01
-3.01487327e-01 -4.68426704e-01 -1.34149743e-02 6.51860759e-02
8.46872032e-01 2.03756005e-01 2.27075383e-01 5.58575869e-01
4.45478171e-01 1.14859067e-01 -1.05977941e+00 9.98832583e-01
9.89363790e-02 -1.78384650e+00 1.29825205e-01 3.38831134e-02
6.43175602e-01 4.09875631e-01 2.58977950e-01 1.12258269e-04
4.35537510e-02 -1.28459799e+00 6.92077637e-01 2.78968304e-01
8.00890744e-01 -7.64133990e-01 1.14057946e+00 -1.67561173e-01
-9.83404875e-01 3.71563941e-01 -4.32173669e-01 2.26771727e-01
2.18582988e-01 8.27983916e-01 -8.79040778e-01 8.50732505e-01
9.78132606e-01 7.69409299e-01 -5.34998059e-01 1.07919884e+00
1.97136685e-01 5.66376805e-01 -2.98473775e-01 4.11679626e-01
4.86955613e-01 -1.57068565e-01 4.81217593e-01 1.44437408e+00
5.49635470e-01 2.17417151e-01 -1.35102242e-01 7.12165713e-01
-2.52007306e-01 1.55966684e-01 -5.87487936e-01 4.14901078e-01
-1.11859120e-01 1.36361217e+00 -1.05844414e+00 -2.85050511e-01
-3.17248911e-01 8.18777859e-01 1.42893836e-01 2.56918222e-01
-8.41415048e-01 -6.58626035e-02 3.46325457e-01 3.45865279e-01
1.43963620e-01 -2.82793075e-01 -6.37429476e-01 -1.05282295e+00
1.39826044e-01 -8.40992868e-01 3.38210106e-01 -8.20746720e-01
-1.01437545e+00 8.23323190e-01 -3.26564193e-01 -1.29628158e+00
2.05453664e-01 -6.71643853e-01 -5.06323576e-01 5.21253169e-01
-1.86354399e+00 -1.22864020e+00 -5.33996522e-01 9.93923247e-01
3.60864341e-01 -2.96393558e-02 6.10352516e-01 7.81096578e-01
-2.49964640e-01 5.53166687e-01 1.86587662e-01 2.78618038e-01
5.22804558e-01 -1.03932297e+00 1.19746894e-01 8.44060540e-01
-2.38589153e-01 7.90232301e-01 5.80988586e-01 -2.03059033e-01
-1.51625657e+00 -1.26124680e+00 4.42928106e-01 -1.93354473e-01
4.47997391e-01 -1.29625872e-01 -7.14700103e-01 7.29684353e-01
1.96821868e-01 3.15444231e-01 5.15025675e-01 -8.84547606e-02
-1.05084598e-01 -1.08555526e-01 -1.26304853e+00 6.70985758e-01
1.02507365e+00 -3.84375066e-01 -1.23969108e-01 4.67220813e-01
6.53041124e-01 -6.00286901e-01 -1.08673918e+00 8.88805151e-01
5.65866470e-01 -1.20960784e+00 1.44566524e+00 -2.58314341e-01
9.71157074e-01 -5.95833547e-02 -1.56216040e-01 -1.18769574e+00
-3.15267771e-01 -1.86983183e-01 4.61611718e-01 6.01035237e-01
3.15507591e-01 -7.35987902e-01 9.65293825e-01 2.27893054e-01
-7.85163462e-01 -8.26372564e-01 -1.34171391e+00 -6.51042581e-01
1.52217031e-01 -7.59397149e-01 2.55271524e-01 1.07153451e+00
-5.35404503e-01 -1.79456905e-01 -3.10793549e-01 3.05207700e-01
7.02366412e-01 -3.38714570e-01 5.20784438e-01 -8.37627888e-01
-2.03626797e-01 -3.03537279e-01 -5.39861321e-01 -7.12475717e-01
-1.46937490e-01 -1.22398150e+00 -3.93311195e-02 -1.56000328e+00
2.98021019e-01 -5.86812913e-01 -4.93570238e-01 3.65999788e-01
2.06962556e-01 9.34291482e-01 1.01504080e-01 1.88905180e-01
-5.50789475e-01 5.20045340e-01 1.46817195e+00 -1.33809954e-01
1.79856926e-01 -1.48286000e-01 -6.05137169e-01 7.64116347e-01
7.78036892e-01 -4.21926826e-01 -3.51268381e-01 -8.41362178e-01
2.27731854e-01 1.29886582e-01 6.76084399e-01 -1.28251421e+00
4.22135144e-01 4.29806560e-01 4.44550574e-01 -3.70476872e-01
3.47801417e-01 -8.16630423e-01 1.38096079e-01 7.43185163e-01
7.18744621e-02 2.62059033e-01 3.10512930e-01 2.94567227e-01
-2.98168302e-01 1.61288977e-02 1.07433593e+00 -3.49493951e-01
-2.62541205e-01 4.35450017e-01 -5.34052968e-01 -1.55377850e-01
6.25781059e-01 -3.29302460e-01 2.07258910e-02 -3.78556401e-01
-9.12508726e-01 -1.52160138e-01 2.36673757e-01 -7.51191098e-03
8.51693869e-01 -1.07844949e+00 -7.98287034e-01 2.47418910e-01
-1.81908607e-01 2.06371397e-01 4.96879429e-01 1.32131147e+00
-1.15413296e+00 3.97177219e-01 -4.10609365e-01 -9.63141322e-01
-1.15323627e+00 1.37266725e-01 7.26892352e-01 -4.89356279e-01
-9.64026868e-01 7.49966681e-01 3.00351948e-01 -7.15434670e-01
-1.06973294e-02 -6.33615553e-01 -1.75691918e-01 -3.45703602e-01
1.00311585e-01 3.40383381e-01 7.64301240e-01 -3.06391209e-01
-3.59538943e-01 4.04084325e-01 1.66826546e-02 -1.57535329e-01
1.97086275e+00 1.34904087e-01 -3.31097394e-01 -2.45335504e-01
1.11490178e+00 -1.15787320e-01 -1.07526755e+00 -2.37985045e-01
-2.68943399e-01 -2.01976866e-01 3.72422427e-01 -4.54051435e-01
-1.57105637e+00 7.73191631e-01 8.05737972e-01 -1.16748512e-01
1.18149471e+00 -1.32746369e-01 8.62280905e-01 2.71384567e-01
3.17726582e-01 -7.25796402e-01 2.22030997e-01 4.58275914e-01
9.65720952e-01 -1.21640277e+00 7.77700022e-02 -4.76180822e-01
-7.70401359e-02 1.27285063e+00 1.82284951e-01 -3.16535264e-01
7.62864172e-01 5.26269138e-01 8.02210048e-02 -6.39123738e-01
-2.68822789e-01 6.25940710e-02 8.60924348e-02 4.78214085e-01
6.69141412e-01 -9.59119424e-02 -2.71504015e-01 2.47650251e-01
-4.85269696e-01 -1.73871964e-02 6.77542090e-01 8.29296649e-01
-4.59456854e-02 -8.10252368e-01 -2.75266618e-01 4.15468991e-01
-8.40444088e-01 -4.65265125e-01 1.66691795e-01 8.35004568e-01
1.66148469e-02 6.71222866e-01 1.37884328e-02 -1.36095494e-01
5.45937657e-01 -5.91062129e-01 5.19118786e-01 -5.15385985e-01
-8.90228927e-01 -6.11736514e-02 -1.77640766e-02 -6.98726952e-01
-5.87203920e-01 -5.22193611e-01 -9.94673669e-01 -4.49544191e-01
3.11107188e-02 -2.16743097e-01 7.89661467e-01 9.09467459e-01
3.29400331e-01 1.13510346e+00 3.86852056e-01 -9.03510749e-01
-4.13052112e-01 -7.84807920e-01 -2.31872022e-01 4.65588927e-01
3.89406562e-01 -3.60519737e-01 2.31783036e-02 -3.48726436e-02] | [14.085856437683105, -2.5743587017059326] |
1487dd9a-0a0d-4dd0-be4a-c7ad0c19bdc8 | facial-landmark-points-detection-using | 2111.07047 | null | https://arxiv.org/abs/2111.07047v1 | https://arxiv.org/pdf/2111.07047v1.pdf | Facial Landmark Points Detection Using Knowledge Distillation-Based Neural Networks | Facial landmark detection is a vital step for numerous facial image analysis applications. Although some deep learning-based methods have achieved good performances in this task, they are often not suitable for running on mobile devices. Such methods rely on networks with many parameters, which makes the training and inference time-consuming. Training lightweight neural networks such as MobileNets are often challenging, and the models might have low accuracy. Inspired by knowledge distillation (KD), this paper presents a novel loss function to train a lightweight Student network (e.g., MobileNetV2) for facial landmark detection. We use two Teacher networks, a Tolerant-Teacher and a Tough-Teacher in conjunction with the Student network. The Tolerant-Teacher is trained using Soft-landmarks created by active shape models, while the Tough-Teacher is trained using the ground truth (aka Hard-landmarks) landmark points. To utilize the facial landmark points predicted by the Teacher networks, we define an Assistive Loss (ALoss) for each Teacher network. Moreover, we define a loss function called KD-Loss that utilizes the facial landmark points predicted by the two pre-trained Teacher networks (EfficientNet-b3) to guide the lightweight Student network towards predicting the Hard-landmarks. Our experimental results on three challenging facial datasets show that the proposed architecture will result in a better-trained Student network that can extract facial landmark points with high accuracy. | ['Mohammad H. Mahoor', 'Ali Pourramezan Fard'] | 2021-11-13 | null | null | null | null | ['face-alignment'] | ['computer-vision'] | [-1.03370316e-01 4.00111943e-01 -1.97074190e-01 -5.60433388e-01
-5.51905572e-01 1.22736409e-01 3.98131996e-01 -2.66601026e-01
-4.82073128e-01 3.52723122e-01 -3.24520528e-01 -4.42568436e-02
-4.22314107e-02 -8.41104448e-01 -7.23146439e-01 -8.53930116e-01
-5.77842668e-02 4.33162212e-01 3.64634007e-01 -1.61167942e-02
-1.21712163e-01 5.41288376e-01 -1.53716946e+00 -1.99751277e-02
9.80311990e-01 1.60257494e+00 1.22259662e-01 7.34078661e-02
-2.86170214e-01 5.46279550e-01 -3.51281703e-01 -5.46249688e-01
2.93487698e-01 -1.55362040e-01 -4.57103610e-01 -1.28866524e-01
5.40138483e-01 -3.60246897e-01 -5.62080741e-02 1.10409880e+00
6.14840209e-01 2.52402395e-01 5.05268157e-01 -1.45222533e+00
-2.20713437e-01 3.01484644e-01 -5.70409954e-01 -1.72371015e-01
8.51860456e-03 -7.64566511e-02 4.87369686e-01 -1.28676260e+00
2.79756963e-01 1.32016861e+00 1.08685291e+00 7.93911457e-01
-7.53480554e-01 -9.76482570e-01 2.22231120e-01 3.10106218e-01
-1.82643282e+00 -8.33343744e-01 1.02574587e+00 -1.02668636e-01
3.44922036e-01 -3.44149396e-02 6.09452844e-01 7.77764559e-01
-1.42379954e-01 8.89204204e-01 6.97399199e-01 -1.45833969e-01
1.10863455e-01 3.45059931e-01 -2.58492470e-01 1.36866593e+00
-1.82870328e-01 -1.44874364e-01 -4.15291041e-01 -1.37013942e-01
8.48417819e-01 3.52952421e-01 -2.68661022e-01 -4.63705510e-01
-3.55247051e-01 7.76759803e-01 7.67059028e-01 4.69830073e-02
-5.48463225e-01 1.58339828e-01 2.74316281e-01 1.75344691e-01
7.30776191e-01 -3.50923955e-01 -4.37327713e-01 7.23602921e-02
-1.14538550e+00 -2.32684657e-01 6.01011038e-01 7.21998215e-01
1.20763731e+00 1.47735357e-01 1.34398207e-01 1.10683382e+00
7.08165586e-01 4.35363084e-01 5.15311956e-01 -5.97309828e-01
1.75469875e-01 9.26844835e-01 -3.96595836e-01 -1.28420436e+00
-3.34863454e-01 -2.95613557e-01 -8.67282391e-01 4.16148394e-01
3.63428682e-01 -2.15414315e-01 -1.06747842e+00 1.57494271e+00
6.83085263e-01 7.79482007e-01 -2.04507768e-01 8.80986691e-01
1.03533661e+00 5.14216363e-01 1.91555396e-01 -9.16268304e-02
8.78276587e-01 -1.11202836e+00 -3.17331076e-01 2.83365455e-02
6.61125481e-01 -5.14783144e-01 9.79907513e-01 3.29159021e-01
-1.06531966e+00 -7.50557423e-01 -6.81105733e-01 9.02193263e-02
-3.34769368e-01 3.85837555e-01 5.24284065e-01 6.38841689e-01
-1.22698283e+00 5.17548442e-01 -8.03799868e-01 -3.34665358e-01
8.99792433e-01 7.47325063e-01 -3.40690911e-01 1.54940516e-01
-8.60573351e-01 6.06799841e-01 7.30851069e-02 5.51313102e-01
-1.16160643e+00 -6.71892166e-01 -9.44750845e-01 1.13350965e-01
4.32531148e-01 -4.05591428e-01 9.22584653e-01 -1.36967731e+00
-1.89395869e+00 8.20187330e-01 -1.19321104e-02 -2.43462846e-01
6.12595916e-01 -9.06767473e-02 -2.20824093e-01 3.08205724e-01
-3.41121815e-02 9.76263762e-01 1.22240186e+00 -1.22635102e+00
-6.45544708e-01 -4.06579942e-01 1.63742993e-02 9.71195698e-02
-6.04364574e-01 -2.18779981e-01 -5.81786275e-01 -3.97229463e-01
1.53527319e-01 -7.67010212e-01 -4.77396809e-02 5.46107888e-01
-4.23708409e-01 -5.75839698e-01 1.25480711e+00 -4.84710157e-01
8.68434310e-01 -2.18146014e+00 -2.88785249e-01 6.48020983e-01
2.59825826e-01 7.83743203e-01 -3.80308837e-01 -2.24815547e-01
7.24065118e-03 -1.19754173e-01 8.52236897e-02 -7.04896927e-01
-2.07929745e-01 2.73855746e-01 -1.05735719e-01 4.65942234e-01
2.32699677e-01 8.57633114e-01 -8.06985915e-01 -6.57506287e-01
2.29012907e-01 9.43723857e-01 -5.83702743e-01 2.16212973e-01
-1.56680062e-01 2.07793146e-01 -5.54042459e-01 9.29434359e-01
7.62855113e-01 -7.87826851e-02 -1.09261982e-01 -3.01165581e-01
1.41739294e-01 -1.33028373e-01 -1.11640930e+00 1.49545181e+00
-6.33475840e-01 2.23908544e-01 2.58578271e-01 -9.54292834e-01
1.32263720e+00 3.11042309e-01 3.33938748e-01 -6.49455965e-01
2.66746312e-01 1.97347775e-01 -3.29207420e-01 -5.50161242e-01
1.52850905e-02 -3.44604403e-02 5.69483817e-01 1.51503086e-01
3.20085108e-01 2.41589338e-01 -2.33980522e-01 -1.20000221e-01
6.83469474e-01 2.58544326e-01 -1.87450781e-01 -2.05317829e-02
9.02766287e-01 -6.12012088e-01 9.27366674e-01 1.70790091e-01
-2.53891945e-01 3.76495510e-01 3.83295566e-01 -7.53005683e-01
-4.41561162e-01 -7.87557602e-01 -3.86724062e-02 1.26816142e+00
1.52442709e-01 -4.52123821e-01 -9.34029996e-01 -1.15155983e+00
-6.31467104e-02 6.00707233e-02 -5.99084258e-01 -3.27314794e-01
-4.10924822e-01 -3.10282648e-01 6.80320561e-01 5.74988842e-01
7.76142955e-01 -1.02188885e+00 -2.59251654e-01 -4.82652411e-02
2.03790665e-01 -8.24817955e-01 -3.73490453e-01 -6.07394166e-02
-8.28974903e-01 -1.20270741e+00 -8.64885569e-01 -1.03899193e+00
1.24778259e+00 9.38102901e-02 6.10023797e-01 6.03191793e-01
-3.78554761e-02 3.00801933e-01 -1.11934423e-01 -5.28749228e-01
1.16296671e-01 5.90443686e-02 2.14344993e-01 6.31484151e-01
4.41870660e-01 -6.59832656e-01 -6.02166712e-01 4.89579201e-01
-6.39335632e-01 -2.28578508e-01 6.51614964e-01 8.62804711e-01
8.33176434e-01 -3.34369987e-02 5.53281486e-01 -7.75975764e-01
2.54800171e-01 -4.06768262e-01 -6.20081425e-01 1.58571541e-01
-5.57223916e-01 -3.22598964e-01 7.81574190e-01 -6.46146774e-01
-9.03951943e-01 3.20667207e-01 -5.05767286e-01 -7.97099769e-01
-1.07163683e-01 4.66068476e-01 -3.76355618e-01 -7.42062986e-01
4.06478614e-01 2.38203347e-01 2.81412840e-01 -4.92464095e-01
5.93238734e-02 4.70584184e-01 2.95323104e-01 -5.50871968e-01
8.41723800e-01 4.95048732e-01 2.97082923e-02 -8.95973384e-01
-7.94534922e-01 -1.65730327e-01 -5.60711920e-01 -3.99571002e-01
6.26021385e-01 -8.10733438e-01 -8.71777296e-01 5.71476042e-01
-9.10687327e-01 -5.93747854e-01 -1.78351745e-01 2.87610441e-01
-4.33953255e-01 1.00383960e-01 -1.60895035e-01 -8.32712948e-01
-4.81132090e-01 -1.18659961e+00 1.06108570e+00 6.35069191e-01
1.32533818e-01 -1.25420105e+00 -1.69890881e-01 1.10026620e-01
4.71408695e-01 1.13266774e-01 6.66496158e-01 -6.67307615e-01
-4.56302226e-01 -2.60193139e-01 -2.44580016e-01 4.49703753e-01
5.58409505e-02 1.08046986e-01 -1.14847887e+00 -1.02770709e-01
-2.14510843e-01 -6.24453962e-01 8.75110805e-01 3.71035486e-01
1.54353964e+00 -3.45037818e-01 -4.26566750e-01 1.15270329e+00
1.04477394e+00 5.89069165e-02 6.31676376e-01 1.52456746e-01
8.63206148e-01 4.58119929e-01 3.91224861e-01 1.97250172e-01
6.46916807e-01 5.83519518e-01 5.64171731e-01 -4.30735916e-01
-1.35392338e-01 -5.16773403e-01 3.00184339e-01 6.69673622e-01
-3.69095594e-01 3.13868999e-01 -8.56187165e-01 4.02921706e-01
-1.85346580e+00 -7.37879813e-01 2.40528107e-01 2.01785326e+00
8.58317435e-01 -7.17909914e-03 2.34875321e-01 1.05425872e-01
4.73648071e-01 1.65871024e-01 -5.13640821e-01 -1.03120632e-01
2.15924025e-01 3.21963847e-01 -6.23860098e-02 4.39739764e-01
-1.06134248e+00 1.16113770e+00 4.89227819e+00 1.07669234e+00
-1.66420579e+00 2.03839958e-01 7.79112399e-01 9.43412706e-02
6.97179064e-02 -2.26604939e-01 -1.05454588e+00 6.00786686e-01
6.57694757e-01 1.85112819e-01 9.96824726e-02 1.36503899e+00
2.74192750e-01 -6.31301850e-02 -9.44159091e-01 1.27229285e+00
1.60199404e-01 -1.11508131e+00 1.59292459e-01 -1.13745838e-01
5.41921854e-01 -2.25597650e-01 2.61149555e-01 4.40254092e-01
-1.57299563e-01 -1.08483922e+00 4.17421103e-01 7.57770181e-01
8.50070238e-01 -1.06197202e+00 8.40251386e-01 5.07667124e-01
-1.30067015e+00 2.24697971e-04 -7.12356448e-01 2.32080087e-01
-1.69414982e-01 5.50474286e-01 -8.98237944e-01 1.87190130e-01
7.51866281e-01 7.06282616e-01 -4.63991672e-01 1.15505290e+00
-3.70969325e-01 5.83767474e-01 -4.78439331e-01 1.39507689e-02
2.46601045e-01 -3.63511950e-01 2.15023607e-01 8.20925295e-01
2.89493710e-01 -6.49861470e-02 3.91882986e-01 6.34271741e-01
-3.39049041e-01 1.99496686e-01 -5.69895446e-01 2.81157762e-01
3.94738436e-01 1.66791046e+00 -7.42869079e-01 -1.69534519e-01
-3.01739067e-01 9.23979342e-01 5.60706139e-01 2.35685319e-01
-6.50732756e-01 -4.57075268e-01 6.50308609e-01 2.29409128e-01
1.28386915e-01 -7.15903938e-02 5.02293929e-02 -7.75517225e-01
-1.29768159e-02 -5.40059924e-01 2.35424876e-01 -6.50460482e-01
-1.20233619e+00 7.38790572e-01 -3.45577449e-01 -9.99070287e-01
-4.39149626e-02 -5.30062497e-01 -1.00124776e+00 8.52268100e-01
-1.67440903e+00 -1.37317157e+00 -6.04088664e-01 9.86484468e-01
2.82929748e-01 -3.94089818e-01 7.18481481e-01 5.03778577e-01
-8.10038567e-01 1.17895782e+00 -3.48657936e-01 5.39502919e-01
5.91127276e-01 -9.07240331e-01 -1.12938598e-01 3.09610128e-01
2.22321928e-01 6.78656578e-01 -8.13213065e-02 -4.22105700e-01
-1.19564199e+00 -1.37977910e+00 7.60551393e-01 -1.08118996e-01
3.43986362e-01 -2.39979535e-01 -8.23228300e-01 6.43804073e-01
-3.64946663e-01 4.63964432e-01 6.90400064e-01 8.37427601e-02
-1.75804585e-01 -6.36857033e-01 -1.45449662e+00 4.56380129e-01
9.37887251e-01 -5.32396853e-01 -1.37656391e-01 2.98907280e-01
2.95421988e-01 -3.76871437e-01 -5.78217506e-01 5.29186785e-01
5.93748808e-01 -8.39189112e-01 7.88235426e-01 -3.69514406e-01
8.87772366e-02 -1.83497429e-01 2.90660858e-01 -1.28656626e+00
-1.24132587e-02 -4.67392832e-01 -2.92709202e-01 1.25123298e+00
2.14719266e-01 -6.48215711e-01 1.37811470e+00 5.16040504e-01
-1.77927896e-01 -1.42836118e+00 -9.77283955e-01 -6.67735577e-01
-2.78023362e-01 -2.90294528e-01 6.72978818e-01 9.34466720e-01
-4.26378876e-01 9.13377032e-02 -1.76793709e-01 8.78880471e-02
5.25593281e-01 -3.95248830e-01 8.41122925e-01 -1.48491418e+00
3.38075995e-01 -4.64624822e-01 -6.87308490e-01 -1.16200268e+00
4.84074861e-01 -9.36426282e-01 -1.03245035e-01 -1.20807576e+00
-1.37083277e-01 -1.01274931e+00 -2.32403144e-01 1.00877941e+00
3.87969310e-03 6.30854189e-01 -6.60666376e-02 7.89428782e-03
-7.17496097e-01 9.15957808e-01 1.04259372e+00 -1.07447527e-01
-4.18919533e-01 3.61512452e-01 -4.17525291e-01 1.28283060e+00
5.61411679e-01 -4.30397034e-01 -5.40173411e-01 -3.62170845e-01
-1.16440412e-02 -3.10128450e-01 5.42073846e-01 -1.04408717e+00
6.67517841e-01 1.37166098e-01 6.53400779e-01 -4.87039477e-01
4.58123267e-01 -1.04150879e+00 -2.43450224e-01 2.97703117e-01
1.48693006e-02 -2.71431565e-01 6.60075173e-02 2.82346159e-01
-2.07828596e-01 -1.40655413e-01 9.20185387e-01 8.63902792e-02
-7.14932203e-01 9.41331863e-01 2.50122011e-01 -2.02126876e-01
1.05687857e+00 -5.54623246e-01 2.15371326e-02 -2.96598345e-01
-8.43747079e-01 2.91922480e-01 1.79646954e-01 2.35869020e-01
1.06443298e+00 -1.53863704e+00 -3.70107830e-01 5.19085705e-01
-1.70055732e-01 4.11978990e-01 1.29305765e-01 1.11722040e+00
-4.76066053e-01 -4.85272482e-02 -1.76216394e-01 -7.45933354e-01
-1.28822589e+00 1.79859102e-01 5.87651193e-01 1.12873666e-01
-4.81729358e-01 1.28853166e+00 1.63462982e-01 -6.16149545e-01
7.03908503e-01 -6.46570548e-02 -2.42928684e-01 1.59237042e-01
7.00104237e-01 4.27408367e-01 1.85592309e-01 -7.82409012e-01
-4.19878662e-01 7.72498429e-01 -1.15401685e-01 3.43377769e-01
1.45431232e+00 1.56976566e-01 -1.47780910e-01 1.51778860e-02
1.31851816e+00 -7.77395666e-02 -1.44021165e+00 -4.33813334e-01
-9.62999389e-02 -3.10926735e-01 1.47550255e-01 -5.46339929e-01
-1.61708021e+00 1.03094351e+00 8.58251035e-01 -1.81931004e-01
1.33145511e+00 -1.97492212e-01 9.25327003e-01 4.46344286e-01
5.35660028e-01 -1.01119137e+00 2.20095322e-01 6.02762222e-01
6.21683061e-01 -1.19801772e+00 -4.20313597e-01 -4.29148704e-01
-2.92329490e-01 1.24048114e+00 9.67408419e-01 -1.52457222e-01
1.04325795e+00 6.79147765e-02 3.53713930e-01 -1.68048993e-01
-2.94101357e-01 -2.48670101e-01 4.61470366e-01 7.53068388e-01
1.22405618e-01 -2.14561015e-01 6.22258931e-02 7.35036910e-01
-1.20744839e-01 1.45001337e-01 -1.15100347e-01 5.98946452e-01
-5.03659964e-01 -9.91538227e-01 -2.02484727e-01 4.78816569e-01
-3.26656491e-01 8.52410495e-02 -3.04224700e-01 6.86065555e-01
6.53496623e-01 6.60044253e-01 1.11446500e-01 -6.22419357e-01
1.88334987e-01 9.72282216e-02 2.90963292e-01 -7.01988876e-01
-5.02637744e-01 -4.03487571e-02 -3.35531533e-01 -6.92394257e-01
-3.53419781e-01 -1.55754119e-01 -1.40618813e+00 -3.97240341e-01
-3.34390998e-01 2.29944199e-01 6.87142909e-01 9.66172755e-01
3.39228839e-01 1.19124733e-01 8.50831449e-01 -1.12818778e+00
-3.69043440e-01 -8.04719329e-01 -5.36259294e-01 3.74236405e-02
2.70632565e-01 -9.31715071e-01 -1.99077308e-01 -2.01320350e-02] | [13.487844467163086, 0.49170851707458496] |
1ce25df2-1a6b-4888-96a6-7729792fa00c | language-models-as-zero-shot-planners-1 | 2201.07207 | null | https://arxiv.org/abs/2201.07207v2 | https://arxiv.org/pdf/2201.07207v2.pdf | Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents | Can world knowledge learned by large language models (LLMs) be used to act in interactive environments? In this paper, we investigate the possibility of grounding high-level tasks, expressed in natural language (e.g. "make breakfast"), to a chosen set of actionable steps (e.g. "open fridge"). While prior work focused on learning from explicit step-by-step examples of how to act, we surprisingly find that if pre-trained LMs are large enough and prompted appropriately, they can effectively decompose high-level tasks into mid-level plans without any further training. However, the plans produced naively by LLMs often cannot map precisely to admissible actions. We propose a procedure that conditions on existing demonstrations and semantically translates the plans to admissible actions. Our evaluation in the recent VirtualHome environment shows that the resulting method substantially improves executability over the LLM baseline. The conducted human evaluation reveals a trade-off between executability and correctness but shows a promising sign towards extracting actionable knowledge from language models. Website at https://huangwl18.github.io/language-planner | ['Igor Mordatch', 'Deepak Pathak', 'Pieter Abbeel', 'Wenlong Huang'] | 2022-01-18 | language-models-as-zero-shot-planners | https://openreview.net/forum?id=6NT1a56mNim | https://openreview.net/pdf?id=6NT1a56mNim | null | ['robot-task-planning'] | ['robots'] | [ 2.26864234e-01 8.85616720e-01 -3.17143798e-01 -4.34143841e-01
-1.02390587e+00 -6.66865349e-01 9.34761584e-01 -1.31526351e-01
-2.74926901e-01 9.39267635e-01 7.34009326e-01 -5.31607151e-01
-5.17836064e-02 -7.80557990e-01 -1.02193439e+00 -1.54038489e-01
-2.24497080e-01 8.04059148e-01 2.68474609e-01 -4.39399660e-01
1.10140368e-01 2.43316576e-01 -1.43634295e+00 6.72564507e-01
9.15780365e-01 2.45931908e-01 4.30104882e-01 8.85725439e-01
-1.31091207e-01 1.44809258e+00 -5.91979504e-01 6.74998984e-02
2.99682230e-01 -3.89203370e-01 -1.50245178e+00 1.25214458e-01
1.49075821e-01 -7.37614989e-01 -1.80500180e-01 6.08368695e-01
1.67269245e-04 3.84391218e-01 4.90071118e-01 -1.40806913e+00
-1.65871635e-01 1.10518694e+00 1.63826138e-01 -4.38001662e-01
9.19814765e-01 7.24423170e-01 9.04931664e-01 -6.71474710e-02
7.76090562e-01 1.49156249e+00 3.30362827e-01 6.86018288e-01
-1.21791482e+00 -2.56219685e-01 4.49455082e-01 7.20272809e-02
-1.11092782e+00 -5.48685789e-01 4.40241545e-01 -5.41144013e-01
1.47441912e+00 3.07524770e-01 4.25021291e-01 1.33924747e+00
8.43171254e-02 1.02218080e+00 1.01946890e+00 -5.98621249e-01
1.89455941e-01 1.07639343e-01 -9.91207287e-02 8.91143203e-01
-6.42650872e-02 2.80962825e-01 -6.51402056e-01 -8.94594342e-02
9.45076287e-01 -1.80417508e-01 -7.42442533e-02 -4.40505981e-01
-1.53861153e+00 4.11973000e-01 1.40785396e-01 2.91549385e-01
-3.98780107e-01 4.62618798e-01 4.94752139e-01 3.70673180e-01
-1.22104548e-01 8.91470075e-01 -6.44976377e-01 -5.21833479e-01
-6.87983513e-01 4.28023636e-01 1.15797055e+00 1.21910870e+00
6.78026617e-01 -1.09284513e-01 -3.27079415e-01 3.22651476e-01
2.07994252e-01 2.75964469e-01 3.65496635e-01 -1.52450991e+00
7.38751113e-01 7.04878390e-01 7.19541669e-01 -4.23898429e-01
-6.73862576e-01 2.15221584e-01 -2.34824136e-01 5.10960042e-01
5.46943128e-01 -4.32919025e-01 -6.51412308e-01 1.76313543e+00
3.16331089e-01 1.06594808e-01 2.98745573e-01 7.23388791e-01
5.70701540e-01 6.35617912e-01 4.53489155e-01 -9.53538045e-02
1.05966663e+00 -1.23822176e+00 -5.15630543e-01 -4.47800219e-01
1.15193737e+00 -2.95612633e-01 1.61667657e+00 2.76390910e-01
-1.31374502e+00 -7.49409199e-01 -7.50541627e-01 -5.05103543e-03
-2.43863150e-01 2.32428893e-01 8.52803469e-01 2.09559515e-01
-1.11105740e+00 6.60023212e-01 -1.19771564e+00 -5.22011697e-01
1.94095783e-02 3.10673058e-01 -3.18393648e-01 1.59960717e-01
-9.73930001e-01 1.03902972e+00 8.04825902e-01 -2.15601280e-01
-1.64748001e+00 -1.88283235e-01 -1.09143806e+00 -1.97072756e-02
9.88516271e-01 -5.74109852e-01 1.98142993e+00 -8.77745926e-01
-1.98677623e+00 6.24315917e-01 1.39630120e-02 -6.18029654e-01
7.63297796e-01 -5.25354743e-01 1.04605705e-02 -3.02753448e-02
2.28106573e-01 9.08786297e-01 3.26074660e-01 -1.17229068e+00
-9.46977437e-01 7.78772086e-02 1.20922399e+00 3.26158255e-01
-1.48626016e-02 -1.62179465e-03 -2.04045698e-01 -4.66331877e-02
-1.23177491e-01 -1.21783233e+00 -4.30367738e-01 -3.15664619e-01
-5.20606101e-01 -4.04525876e-01 1.57058075e-01 -5.61115205e-01
1.22303045e+00 -1.72526383e+00 2.46834859e-01 -8.89712572e-02
-1.66167431e-02 2.55081356e-02 -3.09412777e-01 8.95704806e-01
1.93971425e-01 1.23980962e-01 2.28211051e-03 -6.45105913e-02
5.90014458e-01 2.85619408e-01 -3.01936179e-01 1.58899769e-01
-6.76292703e-02 9.82517898e-01 -1.20855057e+00 -6.45027757e-01
4.83259320e-01 2.47607287e-02 -7.06395328e-01 6.69543445e-01
-8.68452072e-01 6.75842047e-01 -5.89776397e-01 4.01444972e-01
2.18628384e-02 -3.09824198e-01 5.04975677e-01 2.67308503e-01
-2.13041946e-01 8.58950078e-01 -1.17844570e+00 2.07131720e+00
-1.00483608e+00 4.40161377e-01 -9.67299566e-02 -6.02951884e-01
3.30509752e-01 4.91302431e-01 1.69501856e-01 -5.53576529e-01
-2.93548137e-01 4.66616005e-02 1.32430643e-01 -8.96340311e-01
3.50696772e-01 -2.12208610e-02 -4.07861739e-01 4.64161932e-01
2.91053988e-02 -3.80943060e-01 4.36903715e-01 2.29805619e-01
1.24309480e+00 1.15411115e+00 7.03612447e-01 1.15628624e-02
3.93142402e-01 5.46820164e-01 2.15448365e-01 9.40878689e-01
-1.58593088e-01 1.51685104e-01 5.64873636e-01 -5.47746241e-01
-7.87769556e-01 -7.95609593e-01 5.69222569e-01 1.53188324e+00
-2.14619078e-02 -8.07155192e-01 -9.43632603e-01 -7.93046296e-01
-5.70407271e-01 1.23304772e+00 -2.00079292e-01 1.07191563e-01
-8.33503544e-01 3.12873684e-02 7.00814366e-01 5.98570287e-01
6.47378802e-01 -1.60572433e+00 -1.24641299e+00 2.56145418e-01
-4.56265271e-01 -1.20235705e+00 -3.55369419e-01 2.52716720e-01
-7.61550188e-01 -1.09348214e+00 -1.20452546e-01 -7.49421179e-01
6.76734090e-01 -4.61551175e-02 1.40466785e+00 7.06447810e-02
2.32600674e-01 6.84962273e-01 -4.44619298e-01 -3.29048246e-01
-6.70999885e-01 3.29012014e-02 3.03842332e-02 -6.42841041e-01
9.13745016e-02 -5.22643924e-01 -3.09410185e-01 2.23163620e-01
-5.00988424e-01 7.68991709e-01 7.41071939e-01 4.92724061e-01
4.25399095e-01 1.26606598e-01 1.65179640e-01 -9.66080546e-01
8.15855265e-01 -3.43128555e-02 -5.00039637e-01 3.89314681e-01
-3.04769725e-01 4.58271474e-01 7.97797859e-01 -4.51241255e-01
-1.31015170e+00 3.86110336e-01 8.89645366e-04 1.77896455e-01
-8.68338466e-01 4.27168459e-01 -3.20199579e-01 3.47991467e-01
9.08867955e-01 1.46014035e-01 -5.50087869e-01 -1.72064811e-01
6.08269393e-01 4.27484959e-01 4.84981865e-01 -1.28463006e+00
8.29197288e-01 3.11188281e-01 -1.42341554e-01 -5.86299658e-01
-7.08955646e-01 -3.54728550e-01 -7.10870802e-01 -1.21372081e-01
7.92167306e-01 -8.20089817e-01 -1.10820234e+00 -2.93034501e-03
-1.17757726e+00 -1.49635410e+00 -3.48349601e-01 2.49879122e-01
-1.30816174e+00 7.95725957e-02 -5.33246815e-01 -9.88159955e-01
5.56659661e-02 -1.14722157e+00 1.19840777e+00 9.81181264e-02
-9.30006206e-01 -9.84359264e-01 -1.52027318e-02 4.34237003e-01
1.63388059e-01 3.46468031e-01 8.80338252e-01 -7.29801118e-01
-6.66192651e-01 -1.52691811e-01 2.14222953e-01 5.63369952e-02
1.38868213e-01 -2.30926856e-01 -6.81534410e-01 6.02034442e-02
-2.59617597e-01 -8.00562203e-01 4.87457868e-03 2.43691325e-01
9.28337276e-01 -8.48844111e-01 -3.82719815e-01 2.49966308e-01
1.02966952e+00 3.00356954e-01 6.43620372e-01 7.25675523e-01
5.02026021e-01 6.56865656e-01 9.66459394e-01 3.16156775e-01
6.24536097e-01 8.92681718e-01 3.06570321e-01 6.20277748e-02
-1.78868786e-01 -8.29182684e-01 7.60321379e-01 1.54691353e-01
-4.21121001e-01 -8.47608037e-03 -1.15229106e+00 5.12258112e-01
-2.11156559e+00 -1.11168432e+00 1.11908063e-01 2.01709008e+00
1.09360015e+00 4.82495129e-01 1.85331360e-01 -2.77267069e-01
9.75926071e-02 2.45593473e-01 -3.99534523e-01 -4.63449657e-01
5.24535239e-01 2.84920856e-02 2.74599552e-01 9.46763575e-01
-9.98491228e-01 1.29675341e+00 5.83683491e+00 6.96455300e-01
-7.09964514e-01 1.40485972e-01 2.40427747e-01 -7.29525387e-02
-3.97399180e-02 2.99231678e-01 -7.32570231e-01 -6.22362643e-03
1.19747412e+00 -1.81066558e-01 6.77007139e-01 1.05996633e+00
7.56370604e-01 -1.05271734e-01 -1.69349718e+00 5.55245757e-01
-3.30878228e-01 -1.16372955e+00 -4.57683578e-02 -2.65079767e-01
6.32116139e-01 -2.67713040e-01 -3.92158747e-01 1.09810126e+00
8.39469671e-01 -1.21186149e+00 9.67450857e-01 3.20358098e-01
6.10930741e-01 -3.83313894e-01 3.79215688e-01 1.16828573e+00
-1.15081728e+00 -1.50188625e-01 -1.37597509e-02 -6.63377225e-01
3.17655772e-01 -4.66585487e-01 -1.38398015e+00 3.68117124e-01
2.72614688e-01 2.62906045e-01 -2.75237918e-01 5.47956407e-01
-8.38232875e-01 4.88161236e-01 -1.96559176e-01 -1.23191662e-01
4.51216489e-01 -3.40476930e-02 3.75670433e-01 1.18560684e+00
-7.22428188e-02 3.83929282e-01 7.20722854e-01 8.25697780e-01
2.16150180e-01 -1.94325075e-01 -7.90605247e-01 -1.32921040e-01
2.31045306e-01 9.58278358e-01 -5.27009010e-01 -7.12671816e-01
-2.92239398e-01 9.79151607e-01 3.69594127e-01 5.76339424e-01
-9.59849656e-01 -8.74191895e-02 4.90577340e-01 4.33224887e-01
-1.89062178e-01 -4.31312650e-01 -1.61981300e-01 -9.47634280e-01
4.80185188e-02 -1.31359482e+00 1.72111258e-01 -1.03585637e+00
-4.55764234e-01 5.07836223e-01 5.50765455e-01 -1.26590765e+00
-6.84082568e-01 -4.25891995e-01 -6.84604704e-01 5.65115631e-01
-1.19291079e+00 -1.36253929e+00 -3.00410509e-01 3.50673169e-01
1.18916821e+00 1.66682854e-01 1.12934923e+00 -2.73005217e-01
-1.58074707e-01 1.48166224e-01 -6.32838130e-01 -9.64101031e-03
4.32754546e-01 -1.34054935e+00 5.14988899e-01 8.01928699e-01
1.37676671e-01 8.24291646e-01 1.01869535e+00 -7.85513997e-01
-1.26317716e+00 -9.36888635e-01 9.24760461e-01 -8.89583528e-01
6.14980936e-01 -2.96677351e-01 -5.42613626e-01 1.23828697e+00
2.89830834e-01 -5.64027309e-01 2.96275437e-01 1.85705841e-01
-3.47672477e-02 3.05633545e-01 -9.04694617e-01 9.98049319e-01
1.54509878e+00 -5.50379217e-01 -9.53049302e-01 8.88585329e-01
9.93088007e-01 -8.40349019e-01 -5.50659776e-01 1.41563639e-01
4.18103278e-01 -9.76554751e-01 9.90708351e-01 -1.07070482e+00
7.00864971e-01 -3.45723420e-01 1.07737130e-03 -1.24063361e+00
-5.49650863e-02 -9.36446607e-01 -2.35743269e-01 9.43331420e-01
5.60259581e-01 -3.71144861e-01 6.68191731e-01 1.20253885e+00
-4.01962310e-01 -5.05776167e-01 -4.32943612e-01 -1.08122957e+00
-2.17468426e-01 -7.08633423e-01 5.02620637e-01 4.55880731e-01
6.53875291e-01 3.03671539e-01 -3.21866214e-01 3.80336404e-01
2.96830684e-01 8.33586678e-02 1.25567257e+00 -8.28040004e-01
-6.74091578e-01 -2.59855449e-01 2.31683150e-01 -1.43268573e+00
5.67537963e-01 -8.39085817e-01 4.19763505e-01 -2.00094628e+00
4.63088527e-02 -3.50819677e-01 1.12729229e-01 9.49900210e-01
1.13448970e-01 -5.33563554e-01 2.54526675e-01 5.83574735e-02
-1.18949783e+00 1.71056360e-01 1.40045440e+00 -1.72710434e-01
-7.49613822e-01 1.70370787e-01 -5.14768064e-01 1.10299826e+00
9.74308670e-01 -3.02201748e-01 -6.94390595e-01 -5.07262111e-01
1.72056854e-01 4.52650189e-01 2.65360951e-01 -1.18476915e+00
3.56507063e-01 -9.70852137e-01 -2.24515036e-01 7.15629235e-02
3.41987759e-01 -7.94525504e-01 1.65681466e-01 4.90171760e-01
-8.39446664e-01 -1.29210576e-01 2.99243152e-01 2.50918895e-01
-1.02820247e-01 -3.09877634e-01 1.44037589e-01 -5.90443015e-01
-1.12904871e+00 -1.34098515e-01 -5.42192519e-01 -1.25753969e-01
1.13970280e+00 -2.44679242e-01 -2.01717183e-01 -6.12643242e-01
-8.18669200e-01 4.23536450e-01 4.08265918e-01 2.91017562e-01
3.57056350e-01 -1.05375051e+00 -5.47478676e-01 -2.00191572e-01
1.13544755e-01 1.93375438e-01 -2.90121473e-02 7.66990960e-01
-5.29459894e-01 8.02348435e-01 -2.11757794e-01 -2.14740157e-01
-1.12774479e+00 4.69981194e-01 3.30152750e-01 -6.27151787e-01
-7.93379664e-01 6.98344350e-01 3.27342331e-01 -6.80504143e-01
5.25867522e-01 -8.15786779e-01 -1.48907334e-01 -4.53024745e-01
4.73579049e-01 2.54410267e-01 -3.79266769e-01 -2.53131717e-01
-3.31886560e-01 1.88863993e-01 2.50833929e-01 -3.57711315e-01
1.39034891e+00 -1.86767597e-02 1.56722620e-01 5.06727636e-01
4.92271841e-01 -3.67268287e-02 -1.59340310e+00 1.18477188e-01
3.23693007e-01 -2.73363858e-01 -5.15153944e-01 -1.07877398e+00
-1.73053861e-01 6.49592698e-01 8.09096843e-02 2.07879275e-01
8.11218262e-01 3.97217236e-02 4.49750036e-01 1.03704083e+00
1.09021318e+00 -1.27194655e+00 3.78108859e-01 6.64154172e-01
1.08764589e+00 -1.35704744e+00 -1.06068827e-01 -1.57182723e-01
-9.37214553e-01 1.01730239e+00 1.01966250e+00 3.00620589e-02
-9.30059478e-02 3.97761732e-01 3.78652960e-02 -1.89653739e-01
-9.75068450e-01 -3.25704336e-01 -1.39014572e-01 9.42287862e-01
4.66081947e-01 4.20145482e-01 -1.58495173e-01 5.74536920e-01
-3.68103057e-01 3.06585431e-01 6.68585658e-01 1.22484553e+00
-6.27767026e-01 -1.10187411e+00 -3.66948694e-01 1.46787196e-01
-3.19036767e-02 1.50151640e-01 -4.72309321e-01 1.00764918e+00
5.15771545e-02 1.06852746e+00 -4.26319242e-01 -2.39116877e-01
5.95728993e-01 3.12855482e-01 5.73367834e-01 -1.12133217e+00
-5.83064675e-01 -2.34965652e-01 6.77044451e-01 -1.05450571e+00
-4.65497166e-01 -4.48374242e-01 -1.82422483e+00 -1.59610212e-01
1.60397753e-01 -8.51744413e-03 2.38397568e-01 1.13777769e+00
4.24591973e-02 7.53094137e-01 -4.97206524e-02 -1.19981444e+00
-7.76665688e-01 -9.73416984e-01 2.33683176e-02 5.29075980e-01
2.66189605e-01 -2.86883354e-01 -1.57301441e-01 4.46746260e-01] | [4.406820774078369, 0.9415757060050964] |
7bb9eacc-3e7e-4235-841d-063c936ae676 | tackling-the-story-ending-biases-in-the-story | null | null | https://aclanthology.org/P18-2119 | https://aclanthology.org/P18-2119.pdf | Tackling the Story Ending Biases in The Story Cloze Test | The Story Cloze Test (SCT) is a recent framework for evaluating story comprehension and script learning. There have been a variety of models tackling the SCT so far. Although the original goal behind the SCT was to require systems to perform deep language understanding and commonsense reasoning for successful narrative understanding, some recent models could perform significantly better than the initial baselines by leveraging human-authorship biases discovered in the SCT dataset. In order to shed some light on this issue, we have performed various data analysis and analyzed a variety of top performing models presented for this task. Given the statistics we have aggregated, we have designed a new crowdsourcing scheme that creates a new SCT dataset, which overcomes some of the biases. We benchmark a few models on the new dataset and show that the top-performing model on the original SCT dataset fails to keep up its performance. Our findings further signify the importance of benchmarking NLP systems on various evolving test sets. | ['Nasrin Mostafazadeh', 'James Allen', 'Bakhsh', 'Rishi Sharma', 'Omid eh'] | 2018-07-01 | null | null | null | acl-2018-7 | ['cloze-test'] | ['natural-language-processing'] | [ 1.70869187e-01 1.90003470e-01 -1.17512442e-01 -3.38287205e-01
-9.01129305e-01 -9.56304193e-01 1.08696628e+00 2.68862516e-01
-3.41877937e-01 7.67729700e-01 9.65254188e-01 -3.18351328e-01
-3.79751287e-02 -5.39436281e-01 -5.93679547e-01 -9.49324220e-02
1.94917977e-01 5.78783095e-01 4.38159645e-01 -5.15687644e-01
7.39654481e-01 -3.37307118e-02 -1.33964729e+00 8.19406033e-01
7.77271986e-01 3.63668382e-01 -2.30517715e-01 7.50324667e-01
-8.21243003e-02 1.61033213e+00 -9.33037460e-01 -8.48505080e-01
7.75599340e-03 -5.99079013e-01 -1.25107849e+00 -3.48378718e-01
4.29168820e-01 -3.83818895e-01 -3.56368333e-01 5.64650774e-01
6.63952112e-01 1.09813787e-01 6.04255259e-01 -1.34998786e+00
-7.80800045e-01 1.01293314e+00 -1.69742763e-01 3.63957942e-01
1.03659320e+00 3.82989019e-01 1.27762163e+00 -7.44576514e-01
1.02687633e+00 1.22095716e+00 8.73740554e-01 5.88404298e-01
-9.59743261e-01 -5.60761333e-01 -2.34313440e-02 5.21624207e-01
-1.03678751e+00 -3.76106381e-01 6.70633137e-01 -5.49269378e-01
1.09406257e+00 3.61178219e-01 5.43059587e-01 1.77271640e+00
-1.76705569e-01 1.11780131e+00 1.51274228e+00 -4.25431937e-01
2.95304745e-01 1.24888912e-01 4.04129803e-01 3.83615702e-01
3.08490515e-01 -1.69191718e-01 -1.08815134e+00 -3.62536401e-01
2.94607818e-01 -6.96225822e-01 -3.92209619e-01 3.30789573e-02
-1.42465413e+00 9.76643980e-01 9.68492627e-02 4.88899052e-01
4.45498265e-02 1.01104425e-02 7.37226963e-01 9.85248610e-02
3.69396448e-01 9.56973076e-01 -4.56609279e-02 -7.22462475e-01
-1.03259265e+00 8.58724535e-01 1.35324419e+00 7.25900948e-01
6.37634173e-02 -2.99760371e-01 -6.51037753e-01 7.39332438e-01
-9.07157660e-02 -2.16229614e-02 4.05148864e-01 -1.12347317e+00
6.95057929e-01 5.85572004e-01 2.60825962e-01 -1.03723252e+00
-2.90428728e-01 -1.26809895e-01 -1.40784666e-01 6.18388876e-02
8.24074447e-01 1.15059189e-01 -5.27202845e-01 1.60774899e+00
-7.56568313e-02 -6.68462291e-02 3.83095108e-02 9.59440291e-01
1.03562546e+00 3.52512926e-01 2.14637294e-01 2.73183137e-01
1.21165740e+00 -9.74404514e-01 -6.44614816e-01 -3.21870595e-01
8.57657254e-01 -6.15978658e-01 1.72985625e+00 5.55519938e-01
-1.03717518e+00 -7.38385841e-02 -1.34425259e+00 -3.74370933e-01
-3.98317188e-01 -2.91672170e-01 6.75581872e-01 6.11676514e-01
-8.56299341e-01 6.67165101e-01 -4.50554788e-01 -6.63996100e-01
7.00601995e-01 -3.47246945e-01 -2.58021683e-01 -2.89673209e-01
-1.28948402e+00 1.24585867e+00 4.60324615e-01 -4.60390985e-01
-1.07797337e+00 -5.61709762e-01 -5.19210339e-01 -2.37809300e-01
5.62995017e-01 -4.87217188e-01 1.52892733e+00 -6.50552869e-01
-1.21276796e+00 1.17264795e+00 -1.66160733e-01 -4.48422641e-01
9.98790085e-01 -4.10574794e-01 -1.55809969e-01 2.07055256e-01
4.80770826e-01 2.79485554e-01 2.10340202e-01 -1.20150852e+00
-2.14387208e-01 5.30517511e-02 1.88408017e-01 5.20371161e-02
-1.32725224e-01 5.00589788e-01 1.60537228e-01 -8.08113277e-01
-3.48182559e-01 -7.40551233e-01 2.66803443e-01 -5.00466287e-01
-4.04908180e-01 -5.01246452e-01 4.99026418e-01 -7.32188165e-01
1.34909654e+00 -1.73093736e+00 4.36263718e-02 -3.87798816e-01
3.49291742e-01 1.41481489e-01 -1.69936121e-01 7.98801661e-01
4.25388888e-02 6.26020432e-01 -3.20520729e-01 -3.68941903e-01
1.34075880e-01 2.42677242e-01 -7.42624342e-01 3.37583497e-02
3.31700802e-01 1.10129273e+00 -1.15569770e+00 -6.27059340e-01
-3.49729478e-01 2.89316028e-02 -5.14366448e-01 9.02292281e-02
-6.91394985e-01 3.33954096e-01 -3.42766523e-01 5.29144228e-01
2.33228981e-01 -3.17370623e-01 -6.23582006e-02 5.24776936e-01
3.57401296e-02 6.93701148e-01 -8.22432578e-01 1.66896582e+00
1.42766446e-01 1.08436513e+00 -4.44894671e-01 -3.85327429e-01
5.35682797e-01 3.47830117e-01 7.80169992e-03 -4.64348495e-01
9.96870622e-02 2.75126994e-01 4.12290841e-02 -9.25167322e-01
6.68975353e-01 -3.91567290e-01 -3.03176790e-01 9.37916219e-01
-9.19273868e-02 -3.24467659e-01 3.71068597e-01 6.06498063e-01
1.31814003e+00 2.22356245e-01 2.71233946e-01 -7.87007511e-02
1.76593736e-01 8.49467397e-01 3.02563280e-01 1.19248331e+00
-4.04329270e-01 8.48184884e-01 8.92301619e-01 -5.15656769e-01
-1.14300442e+00 -8.64126325e-01 2.23601162e-01 1.28450108e+00
-1.63925365e-01 -5.57786584e-01 -8.24902654e-01 -7.70696342e-01
-1.38470888e-01 1.38264978e+00 -7.51297534e-01 2.15964779e-01
-5.43051541e-01 -7.11704254e-01 1.24717367e+00 6.48819983e-01
5.94377697e-01 -1.05506349e+00 -8.28166425e-01 2.06909716e-01
-5.51661909e-01 -1.30379534e+00 2.46075317e-02 -1.60591036e-01
-3.74519527e-01 -1.23940337e+00 -3.33697110e-01 -3.25666457e-01
-1.15067303e-01 1.61400929e-01 1.42539358e+00 2.72127301e-01
5.33113703e-02 4.00202155e-01 -7.71490991e-01 -7.30748773e-01
-7.27937162e-01 2.35094681e-01 -2.22640455e-01 -5.13575256e-01
7.17068136e-01 -5.18941224e-01 -1.11702934e-01 1.61787733e-01
-8.48090112e-01 3.02042603e-01 8.93366635e-02 6.92607462e-01
-1.70395046e-01 -1.37930572e-01 6.73911214e-01 -9.99536395e-01
1.19790840e+00 -6.51338995e-01 -7.27645122e-04 2.12774649e-01
-4.28800255e-01 -8.85803103e-02 4.07578886e-01 -5.17915368e-01
-1.00033581e+00 -6.12256825e-01 7.65041858e-02 5.22719584e-02
-3.56253870e-02 5.85324407e-01 -1.49436370e-02 2.89838374e-01
1.03824782e+00 -4.62416559e-02 -2.70545512e-01 -4.39515233e-01
3.89605850e-01 5.08520782e-01 6.94429040e-01 -8.27241063e-01
7.92401433e-01 4.43411022e-01 -5.16402245e-01 -3.72729242e-01
-1.37373364e+00 -1.24806583e-01 -5.60094655e-01 -3.03003699e-01
1.00906324e+00 -8.68263900e-01 -6.59092426e-01 3.43493968e-01
-1.33710206e+00 -6.23764336e-01 -2.58104861e-01 2.01492727e-01
-5.96841395e-01 3.10705036e-01 -5.93729436e-01 -8.31142664e-01
-9.60832909e-02 -7.49810159e-01 7.74883509e-01 1.09274492e-01
-1.03993618e+00 -9.44539189e-01 3.64162505e-01 7.74350286e-01
4.38359708e-01 7.69391239e-01 1.20575964e+00 -1.18691957e+00
-4.12190497e-01 -2.73068756e-01 -1.20973594e-01 -5.95276803e-02
-5.48176229e-01 -1.70598458e-02 -1.15315688e+00 9.10140574e-02
2.33257413e-01 -8.75503778e-01 7.52698839e-01 -2.75575638e-01
7.07798779e-01 -1.81305006e-01 -8.44444036e-02 1.08350076e-01
1.22332275e+00 -5.71322739e-02 6.20938241e-01 9.73098755e-01
5.93672991e-01 5.95689297e-01 3.95257235e-01 3.06134045e-01
8.54437768e-01 2.83288538e-01 1.28943980e-01 3.81007373e-01
-1.01782739e-01 -6.80873096e-01 4.54664618e-01 3.60819072e-01
-1.64478257e-01 -4.71763194e-01 -1.36265361e+00 7.38543391e-01
-1.94264138e+00 -1.29314542e+00 -4.05537784e-01 1.68513238e+00
8.62011015e-01 4.47773695e-01 2.12459207e-01 2.00572297e-01
2.92161524e-01 5.72552323e-01 -3.52135211e-01 -6.15465105e-01
-5.54544747e-01 1.31771401e-01 -2.98799276e-02 4.41884249e-01
-7.48347521e-01 1.07081544e+00 7.77511692e+00 6.45465314e-01
-6.96569979e-01 2.72667348e-01 2.70602852e-01 -3.26960027e-01
-5.05747020e-01 2.89482355e-01 -4.75044906e-01 2.44222984e-01
8.32309425e-01 -4.66168404e-01 5.39032578e-01 7.60465205e-01
2.07792178e-01 -5.67630529e-01 -1.47963119e+00 6.29924178e-01
5.20759404e-01 -1.31493247e+00 1.08481258e-01 -2.91148454e-01
7.41145134e-01 1.05677880e-01 -3.05058390e-01 5.96096098e-01
7.24104047e-01 -1.51421523e+00 1.04168415e+00 3.54423791e-01
3.69778097e-01 -2.80119359e-01 6.14066422e-01 5.96810281e-01
-2.36064717e-01 -1.41674265e-01 2.37609800e-02 -7.37689078e-01
1.80618435e-01 2.08793774e-01 -1.13560653e+00 1.95784628e-01
5.92379093e-01 4.98012722e-01 -1.10982490e+00 7.30048954e-01
-6.94447100e-01 8.86800230e-01 -9.91080478e-02 -4.02695388e-01
2.34475300e-01 4.06285465e-01 9.02323604e-01 1.21740222e+00
-1.65174514e-01 2.07333088e-01 -1.98467169e-02 1.06612098e+00
-2.32245445e-01 -5.92483357e-02 -6.06263995e-01 -4.49444950e-01
5.39052606e-01 8.32634687e-01 -4.90238279e-01 -6.36761487e-01
-2.83208579e-01 8.45779479e-01 5.38177550e-01 2.28853077e-01
-8.82617593e-01 -4.20790687e-02 3.44469845e-01 3.66378397e-01
-9.61299464e-02 -3.20429891e-01 -8.65152717e-01 -1.23071194e+00
1.60622999e-01 -1.26698399e+00 5.10295868e-01 -1.28765917e+00
-1.40297663e+00 5.06326020e-01 2.87814111e-01 -6.65737927e-01
-3.37203711e-01 -4.47721660e-01 -8.97823155e-01 6.22711778e-01
-1.28811252e+00 -1.04140222e+00 -5.46586514e-01 3.01300824e-01
6.38378978e-01 -1.54663354e-01 7.32241094e-01 -2.56353319e-01
-4.24614906e-01 3.82258147e-01 -3.01128596e-01 3.02132815e-01
9.55258369e-01 -1.32347441e+00 6.54774129e-01 8.60385001e-01
9.56754237e-02 6.51618838e-01 1.13682699e+00 -9.08095479e-01
-9.09630716e-01 -5.14452934e-01 1.09119236e+00 -1.54576230e+00
1.02687204e+00 -4.31186706e-01 -1.13325536e+00 8.68129849e-01
5.97158909e-01 -7.30644822e-01 8.23323607e-01 1.41325042e-01
-8.62004101e-01 6.76011801e-01 -9.79457021e-01 7.89342701e-01
1.24464059e+00 -7.54950285e-01 -1.34421062e+00 5.07011652e-01
6.33391201e-01 -5.09272397e-01 -7.42607951e-01 3.87277305e-02
5.75447917e-01 -1.21758699e+00 5.17406762e-01 -1.02626896e+00
1.31592226e+00 -1.67695776e-01 -8.68295282e-02 -1.24019098e+00
-4.33445647e-02 -6.92317307e-01 1.94633946e-01 1.44007587e+00
3.05362135e-01 -4.71430063e-01 5.91472030e-01 1.10465872e+00
9.69942361e-02 -5.61702073e-01 -8.04351747e-01 -7.47409225e-01
5.17483652e-01 -8.11983585e-01 6.37187958e-01 1.17031765e+00
3.52019727e-01 5.53357542e-01 -2.39369124e-01 -2.13233411e-01
3.90806496e-01 -1.35887206e-01 9.70307291e-01 -9.99009788e-01
-2.11758807e-01 -5.73025346e-01 8.04338884e-03 -4.10490870e-01
1.60897061e-01 -1.05951929e+00 -2.07894668e-01 -1.79386055e+00
7.17685103e-01 2.40367278e-02 2.40292192e-01 6.57212675e-01
-4.67739016e-01 6.86007068e-02 4.64289784e-01 3.43295068e-01
-7.23153949e-01 1.76689357e-01 1.09160769e+00 1.02086931e-01
-7.85885751e-02 -5.79092979e-01 -1.10204661e+00 7.75454819e-01
7.64079928e-01 -5.98748326e-01 -2.63397306e-01 -6.91466808e-01
5.70285141e-01 -3.45429957e-01 7.84384847e-01 -1.01582992e+00
3.49516898e-01 -1.17324963e-01 2.72201955e-01 -4.64664698e-01
1.12793043e-01 -1.24020196e-01 -9.27453674e-03 2.66902242e-02
-6.42505825e-01 5.44558205e-02 1.67903081e-01 4.34609085e-01
-1.55280992e-01 -3.63774061e-01 4.63501960e-01 -3.02843153e-01
-7.23545372e-01 -5.31228125e-01 -4.68320966e-01 6.89953804e-01
8.79408419e-01 -5.19704521e-01 -8.52567732e-01 -6.17027938e-01
-4.02540743e-01 3.84380907e-01 7.38662362e-01 7.18863010e-01
2.57732123e-01 -1.15362668e+00 -1.13541758e+00 -3.67511451e-01
3.90957236e-01 -1.28968701e-01 -7.72591233e-02 6.04995668e-01
-5.28302729e-01 3.71817738e-01 -1.86634421e-01 -2.89694637e-01
-1.02389193e+00 4.05097008e-01 1.76312476e-01 -4.19962585e-01
-5.44566154e-01 7.71977723e-01 -2.75465220e-01 -1.27300382e-01
1.84318423e-02 -4.11045291e-02 -3.99425566e-01 2.15023056e-01
7.43944585e-01 6.02943420e-01 -8.34771618e-02 -4.88329470e-01
-2.44693935e-01 1.36255622e-01 -1.35000527e-01 -6.01959229e-01
1.53939033e+00 1.46982402e-01 2.15623993e-02 7.45062232e-01
7.59057283e-01 1.90157682e-01 -9.58578646e-01 -1.12978280e-01
5.46651125e-01 -5.58222532e-01 -3.80484313e-01 -1.31366670e+00
-6.24154247e-02 8.21033418e-01 -1.83662236e-01 3.51090252e-01
6.77576423e-01 3.60796303e-02 7.88264930e-01 2.70679593e-01
4.96435046e-01 -1.27374363e+00 2.20762968e-01 8.83489072e-01
1.21211267e+00 -1.23650765e+00 1.87745064e-01 -2.05022365e-01
-1.21902692e+00 9.58317280e-01 5.47652483e-01 -4.93059978e-02
-1.79509297e-01 1.90206781e-01 6.70554116e-02 -4.58057493e-01
-9.90205407e-01 -9.25358292e-03 -7.10926503e-02 5.19786894e-01
5.71821570e-01 -1.06210075e-02 -4.09449279e-01 9.65655684e-01
-7.02637851e-01 2.54749358e-01 9.51927245e-01 9.22242045e-01
-2.50231713e-01 -8.63778412e-01 -5.77103555e-01 2.14684889e-01
-4.81843740e-01 2.62384266e-02 -1.20556736e+00 1.17016590e+00
-1.47074506e-01 1.15252566e+00 -4.43146825e-01 -5.09517670e-01
5.51549911e-01 4.97166038e-01 5.29958367e-01 -5.66198170e-01
-1.01735628e+00 -5.59862137e-01 6.79191887e-01 -5.62743306e-01
-2.69332051e-01 -7.43652165e-01 -1.06944346e+00 -6.22974157e-01
2.61614677e-02 -1.96802299e-02 2.91156083e-01 1.10624611e+00
3.41044776e-02 1.86035112e-01 1.31378233e-01 -5.77380478e-01
-5.44306397e-01 -1.10783553e+00 -1.54366940e-01 7.73199201e-01
1.26304720e-02 -5.58285832e-01 -4.36594665e-01 3.67742814e-02] | [11.187148094177246, 8.806142807006836] |
dc143d7a-f210-40f6-b06f-262a063206c3 | end-to-end-models-for-chemical-protein | 2304.01344 | null | https://arxiv.org/abs/2304.01344v1 | https://arxiv.org/pdf/2304.01344v1.pdf | End-to-End Models for Chemical-Protein Interaction Extraction: Better Tokenization and Span-Based Pipeline Strategies | End-to-end relation extraction (E2ERE) is an important task in information extraction, more so for biomedicine as scientific literature continues to grow exponentially. E2ERE typically involves identifying entities (or named entity recognition (NER)) and associated relations, while most RE tasks simply assume that the entities are provided upfront and end up performing relation classification. E2ERE is inherently more difficult than RE alone given the potential snowball effect of errors from NER leading to more errors in RE. A complex dataset in biomedical E2ERE is the ChemProt dataset (BioCreative VI, 2017) that identifies relations between chemical compounds and genes/proteins in scientific literature. ChemProt is included in all recent biomedical natural language processing benchmarks including BLUE, BLURB, and BigBio. However, its treatment in these benchmarks and in other separate efforts is typically not end-to-end, with few exceptions. In this effort, we employ a span-based pipeline approach to produce a new state-of-the-art E2ERE performance on the ChemProt dataset, resulting in $> 4\%$ improvement in F1-score over the prior best effort. Our results indicate that a straightforward fine-grained tokenization scheme helps span-based approaches excel in E2ERE, especially with regards to handling complex named entities. Our error analysis also identifies a few key failure modes in E2ERE for ChemProt. | ['Ramakanth Kavuluru', 'Xuguang Ai'] | 2023-04-03 | null | null | null | null | ['chemical-protein-interaction-extraction', 'relation-classification'] | ['medical', 'natural-language-processing'] | [ 2.14610994e-01 4.74143445e-01 -1.70257702e-01 -2.25152418e-01
-9.60523486e-01 -6.26004696e-01 3.14603269e-01 1.04446447e+00
-6.13848567e-01 1.16652083e+00 2.22511455e-01 -4.84027594e-01
-2.95752436e-01 -6.51198983e-01 -6.22256339e-01 -3.28158170e-01
-1.43480003e-01 5.86506307e-01 -5.16239703e-02 -7.00017512e-02
5.75457141e-02 4.49623048e-01 -8.26186895e-01 4.04237419e-01
7.10929573e-01 6.44186914e-01 -3.35810453e-01 6.23103023e-01
-1.71307683e-01 7.54541159e-01 -5.63273966e-01 -8.22127223e-01
-2.22572070e-02 -3.33549857e-01 -1.30309010e+00 -5.43326318e-01
3.60540599e-02 5.59327304e-01 -1.40684128e-01 8.29042912e-01
7.49395549e-01 -4.50635739e-02 6.96139753e-01 -9.37530756e-01
-2.16440335e-01 7.79693842e-01 -4.57657605e-01 2.33922377e-01
6.22670949e-01 1.88579679e-01 1.13936567e+00 -9.67098236e-01
1.28522003e+00 9.40725625e-01 1.00838912e+00 3.48459870e-01
-1.41399622e+00 -7.20526934e-01 -3.83407652e-01 2.62914300e-02
-1.69481540e+00 -5.89259863e-01 -1.72604591e-01 -4.74974304e-01
1.73610985e+00 3.45665455e-01 2.46159211e-01 7.96702683e-01
4.46688801e-01 3.91449183e-01 8.28963697e-01 -3.07611108e-01
2.14684963e-01 -1.16625227e-01 4.08034086e-01 5.58037102e-01
6.18909419e-01 -2.75579005e-01 -6.41432762e-01 -5.14962912e-01
7.08590373e-02 -2.64398724e-01 -2.14369163e-01 3.25178951e-01
-1.33140814e+00 3.35200340e-01 2.31610849e-01 2.09535837e-01
-5.24019837e-01 -1.38806447e-01 7.24666953e-01 1.69949993e-01
5.10389447e-01 1.12726057e+00 -1.09859574e+00 -2.19734043e-01
-8.96196246e-01 2.49671668e-01 1.17918992e+00 1.05680656e+00
3.64928454e-01 -6.64246857e-01 -2.25185290e-01 7.78518140e-01
-3.63793932e-02 -1.76227152e-01 2.83271909e-01 -4.28618819e-01
4.16147590e-01 7.94753194e-01 1.64633766e-02 -7.64435112e-01
-8.27916503e-01 -3.76593262e-01 -7.01961100e-01 -2.45473698e-01
5.44237971e-01 -2.38742888e-01 -9.04149354e-01 1.49294710e+00
5.35495460e-01 8.46447051e-02 3.08636189e-01 3.25538486e-01
1.31448269e+00 2.71275640e-01 8.85080099e-01 -2.94473976e-01
1.84196937e+00 -5.11057436e-01 -9.67062831e-01 -9.61591601e-02
9.54486609e-01 -8.88813972e-01 2.84349173e-01 6.26921594e-01
-9.27670360e-01 6.67395890e-02 -9.05038595e-01 -5.41035950e-01
-9.50173795e-01 -1.55978426e-01 8.65210176e-01 3.96591246e-01
-5.71550369e-01 9.48833466e-01 -7.77765334e-01 -5.36158323e-01
6.13278627e-01 5.65847397e-01 -8.98187935e-01 -5.20869158e-02
-1.34218621e+00 1.30799687e+00 6.16795421e-01 1.09575719e-01
-2.48288229e-01 -1.35389304e+00 -7.84798205e-01 4.25948612e-02
5.39852798e-01 -7.24906683e-01 1.02669597e+00 2.13676721e-01
-8.75967324e-01 9.78020966e-01 -3.48104686e-01 -5.42692304e-01
2.28165999e-01 -2.93140769e-01 -5.57924271e-01 -2.39605997e-02
2.06204399e-01 5.71789682e-01 -2.48653576e-01 -6.66486561e-01
-5.70978463e-01 -2.75717556e-01 -3.14992696e-01 -7.62866288e-02
5.20609356e-02 3.38351876e-01 -3.72260034e-01 -4.97894406e-01
-1.04928166e-01 -6.26550138e-01 -4.78818476e-01 -3.34590882e-01
-8.33004653e-01 -5.35699487e-01 2.84941703e-01 -7.30370402e-01
1.45112550e+00 -1.69814277e+00 -5.17704673e-02 1.76771402e-01
7.35109806e-01 1.86616600e-01 8.88337195e-02 7.14655697e-01
-7.44286299e-01 6.68975174e-01 -2.73770273e-01 6.68577850e-02
-1.67070732e-01 -1.74409524e-01 2.10881025e-01 3.58222216e-01
6.29897594e-01 1.00154257e+00 -1.19635272e+00 -6.33986056e-01
-4.24648732e-01 5.39393127e-01 -2.37255603e-01 -2.02176452e-01
-1.47802830e-01 8.73078927e-02 -3.44309777e-01 8.57217014e-01
3.77008379e-01 -4.39836890e-01 3.34262639e-01 -5.15031397e-01
-1.61233306e-01 6.74192905e-01 -9.75120306e-01 1.52419341e+00
-8.25851262e-02 2.49715164e-01 -1.33470446e-01 -5.79141736e-01
6.85453832e-01 6.20745301e-01 8.27284038e-01 -1.56651631e-01
1.26461282e-01 3.41302812e-01 2.78260976e-01 -5.65715551e-01
4.27952826e-01 -4.29397309e-03 -2.74343994e-02 -6.23026639e-02
2.89422661e-01 -5.42265130e-03 6.13403797e-01 3.78366590e-01
1.81662703e+00 1.08796984e-01 1.17322779e+00 -3.74163300e-01
2.71090627e-01 4.11913663e-01 8.82980466e-01 3.17504168e-01
-9.03721377e-02 3.10116261e-01 7.35190153e-01 -2.43174925e-01
-9.18380737e-01 -5.02287567e-01 -5.44458151e-01 5.52121997e-01
-4.31257516e-01 -1.03855872e+00 -4.66692716e-01 -9.48046803e-01
2.04660892e-02 5.38343132e-01 -4.88915324e-01 -1.06087942e-02
-3.28293949e-01 -1.25478125e+00 1.11201000e+00 1.50522113e-01
1.28518924e-01 -8.59385252e-01 -2.10125864e-01 6.70861900e-01
-1.03927791e-01 -1.37460601e+00 -2.59856343e-01 7.84365952e-01
-5.75826764e-01 -1.53223932e+00 -3.69753540e-01 -4.31750178e-01
5.70584714e-01 -3.79269242e-01 1.36192131e+00 -2.52144992e-01
-6.57733202e-01 -2.91858673e-01 -2.88227677e-01 -8.84776354e-01
-3.78280997e-01 1.73344508e-01 -1.15430646e-01 -7.63644338e-01
8.78773332e-01 -2.05893114e-01 -6.16718113e-01 5.22227772e-02
-7.73405612e-01 -5.20422384e-02 6.68061256e-01 9.94296789e-01
8.50836277e-01 1.22900993e-01 8.73100996e-01 -1.55264997e+00
6.07815027e-01 -8.53695154e-01 -2.40370572e-01 3.03882629e-01
-9.65089798e-01 5.03311865e-02 5.64592779e-01 -1.47536620e-01
-6.56547427e-01 2.07879201e-01 -4.29522187e-01 1.88198254e-01
-2.66772747e-01 1.05440247e+00 -2.38490105e-01 2.41173163e-01
8.36131394e-01 -3.01481992e-01 -3.24728608e-01 -4.73658890e-01
9.41981822e-02 7.02220023e-01 6.06376648e-01 -3.96432281e-01
2.38128364e-01 -7.59341493e-02 4.42772865e-01 -5.67741513e-01
-7.63789892e-01 -7.05711901e-01 -4.79253232e-01 4.88380432e-01
8.75990391e-01 -1.00982082e+00 -1.06978333e+00 1.37288123e-01
-1.13293469e+00 1.77013148e-02 -2.37080708e-01 4.18462485e-01
2.33627968e-02 2.94004232e-01 -9.35007513e-01 -4.88010615e-01
-7.94752002e-01 -1.07458937e+00 9.08158004e-01 5.24560362e-02
-7.33381152e-01 -7.41342247e-01 5.57246357e-02 2.99704134e-01
-1.81222394e-01 6.55563295e-01 1.14242792e+00 -1.26919699e+00
6.74734532e-04 -3.02715033e-01 -2.64568955e-01 -1.99099198e-01
2.55118698e-01 6.23766631e-02 -9.38369334e-01 1.15227632e-01
-5.50778687e-01 -1.30814254e-01 7.92825401e-01 -4.70368303e-02
1.01604652e+00 -3.00125808e-01 -6.77842975e-01 3.57975334e-01
1.38298738e+00 2.83466846e-01 7.77335286e-01 2.70235896e-01
6.39336586e-01 6.61334634e-01 7.59162366e-01 1.57445893e-01
3.86137545e-01 3.89553010e-01 -7.18750507e-02 -3.20694059e-01
-4.58074957e-02 -5.43862768e-02 1.40628768e-02 3.94131064e-01
5.29689714e-02 -3.03023010e-01 -1.26680374e+00 6.48497880e-01
-1.50011897e+00 -6.99992239e-01 -6.54268503e-01 2.05717087e+00
1.76332724e+00 5.33654168e-02 -1.35540664e-01 1.95461690e-01
5.10352969e-01 -7.48941779e-01 -6.33217692e-01 -3.77183199e-01
-1.20027438e-01 8.41704249e-01 7.57326841e-01 2.01091081e-01
-1.05154014e+00 1.08137560e+00 6.24884462e+00 9.73883092e-01
-7.41754293e-01 -1.08750403e-01 8.25334430e-01 1.38089359e-02
1.23278484e-01 5.19354083e-02 -1.17018485e+00 2.88686633e-01
1.43494558e+00 -1.83362976e-01 -8.34309533e-02 5.76922119e-01
2.01500595e-01 -2.64766812e-01 -1.51318800e+00 8.33853781e-01
-5.27695358e-01 -1.58777320e+00 -3.20053846e-01 1.28846198e-01
3.43540698e-01 1.35647207e-01 -5.73620081e-01 1.94633186e-01
6.17809057e-01 -1.35735452e+00 2.43113115e-01 3.77911776e-01
1.06898010e+00 -6.73574924e-01 9.32934046e-01 -7.73853958e-02
-1.08032441e+00 3.19266587e-01 -1.82335913e-01 3.83722305e-01
6.62336871e-02 1.25995874e+00 -1.38322949e+00 1.01679373e+00
7.20561266e-01 6.72468841e-01 -4.82666224e-01 1.20398629e+00
-2.03498185e-01 6.13672733e-01 -3.04155022e-01 1.37334168e-01
-1.59682155e-01 2.69402027e-01 4.47620869e-01 1.84194732e+00
9.05931145e-02 5.35363853e-01 -3.11914459e-03 6.02178276e-01
-5.60208619e-01 3.98582071e-01 -3.21960300e-01 -4.94017929e-01
6.72872066e-01 1.50767982e+00 -8.52393150e-01 -4.62133259e-01
-1.50718778e-01 6.46561027e-01 3.08301181e-01 1.37725770e-01
-4.89073873e-01 -8.41982663e-01 7.09260523e-01 -8.09877515e-02
3.93808931e-02 1.40254632e-01 -3.88737321e-01 -8.89073372e-01
-4.18960690e-01 -1.07844460e+00 7.06240416e-01 -4.25063074e-01
-1.49325883e+00 6.09482765e-01 -1.44668683e-01 -8.73312414e-01
9.71380342e-03 -6.83247030e-01 8.21438134e-02 9.10961568e-01
-1.32897329e+00 -9.41090822e-01 1.19254351e-01 1.32141888e-01
1.65859357e-01 3.22676361e-01 1.22498918e+00 6.36170149e-01
-9.11747396e-01 8.92664075e-01 -7.78575242e-02 1.52204201e-01
1.17719030e+00 -1.38411868e+00 5.27023435e-01 4.77572858e-01
-1.15334287e-01 1.24164736e+00 7.97377586e-01 -1.04932642e+00
-1.31913614e+00 -1.27354288e+00 1.64540923e+00 -6.25461876e-01
7.31528163e-01 -3.93522620e-01 -7.92124152e-01 5.08270264e-01
1.09919466e-01 8.17385614e-02 1.20711386e+00 3.17040086e-01
-4.35261399e-01 3.75051200e-01 -1.47115839e+00 4.26541865e-01
1.08357692e+00 -2.65323102e-01 -4.18638378e-01 5.19938529e-01
6.48703814e-01 -6.07212782e-01 -1.80034006e+00 5.00178814e-01
3.76162469e-01 -1.36748254e-01 8.59127879e-01 -1.00628233e+00
5.42311907e-01 -3.94475341e-01 2.87863135e-01 -1.14884424e+00
-2.67765790e-01 -7.77640462e-01 -1.15726300e-01 1.45575666e+00
9.31347966e-01 -5.58866739e-01 5.76889455e-01 8.51393104e-01
-1.89774662e-01 -9.95494187e-01 -7.10683584e-01 -5.37753403e-01
1.02619976e-01 -1.79057330e-01 4.48150426e-01 1.27898538e+00
4.85744894e-01 6.56975508e-01 9.03062671e-02 2.55445242e-02
2.87472397e-01 -3.92765671e-01 3.95934433e-01 -1.25706244e+00
-2.14807585e-01 -3.46696764e-01 -4.84091759e-01 -3.80446225e-01
-1.74220145e-01 -1.10246849e+00 1.29621282e-01 -1.64288259e+00
3.80562782e-01 -4.50199485e-01 -3.34582806e-01 1.13289642e+00
-5.99963605e-01 3.32534820e-01 -2.63853401e-01 2.66501345e-02
-4.54338878e-01 -1.53185591e-01 8.22933972e-01 -1.09901927e-01
-1.62850752e-01 -5.38873911e-01 -1.04056656e+00 4.61697787e-01
5.39121866e-01 -7.72855043e-01 8.22718441e-02 1.40273064e-01
6.43153191e-01 -1.07590854e-01 3.11504733e-02 -5.44364750e-01
3.43457639e-01 9.06813075e-04 4.72376704e-01 -4.58460540e-01
-1.54934525e-01 -4.52401280e-01 5.04453599e-01 3.51468205e-01
-3.66992801e-01 8.56788531e-02 4.83347833e-01 4.12735879e-01
-8.99834856e-02 -1.36395931e-01 5.72138667e-01 -1.84226185e-01
-4.15852606e-01 2.41120875e-01 -4.12897259e-01 1.20362334e-01
7.99537122e-01 -5.72512560e-02 -6.43799305e-01 2.87976533e-01
-9.80336308e-01 3.93632889e-01 4.78325449e-02 9.07965936e-03
1.98723018e-01 -6.95782781e-01 -8.71501207e-01 -5.39907396e-01
3.74468386e-01 3.11014861e-01 -6.80199116e-02 1.12839341e+00
-5.39193869e-01 5.51919520e-01 2.13903248e-01 -1.29239351e-01
-1.56276596e+00 5.38901091e-01 1.69956923e-01 -9.02799845e-01
-4.50102866e-01 1.13314736e+00 -1.12710580e-01 -3.19080621e-01
1.49261039e-02 -3.29881191e-01 -2.94447780e-01 1.58352375e-01
5.60606241e-01 3.25733155e-01 6.32405996e-01 -2.78378218e-01
-7.32382834e-01 -1.46610931e-01 -4.84770983e-01 2.36132354e-01
1.56407917e+00 3.47281992e-01 -5.27976692e-01 1.03352703e-01
1.05270410e+00 2.74195999e-01 -3.08290899e-01 -9.01495398e-04
6.84408426e-01 1.52056456e-01 -1.17521510e-01 -1.38318110e+00
-6.79976225e-01 2.77643055e-01 9.50949043e-02 -1.87898815e-01
8.59023094e-01 2.18366757e-02 7.17175603e-01 5.19798398e-01
1.32087648e-01 -7.88660109e-01 -7.63787687e-01 5.22388577e-01
5.57893038e-01 -1.10820210e+00 5.85792303e-01 -8.79501045e-01
-3.39707106e-01 1.03192008e+00 4.12118644e-01 3.29843342e-01
5.25791407e-01 6.35025859e-01 -2.54914314e-01 -5.86976826e-01
-8.41351092e-01 -6.64290488e-02 2.86687046e-01 2.63489723e-01
1.14791048e+00 -1.35708362e-01 -9.73895609e-01 9.24459457e-01
-1.63349569e-01 2.03893125e-01 3.58181775e-01 1.09068668e+00
1.42552048e-01 -1.55596566e+00 -1.19354658e-01 6.76290393e-01
-1.19123101e+00 -7.86675334e-01 -6.69140279e-01 7.37999201e-01
3.41492832e-01 1.14934266e+00 -5.17167389e-01 -3.49324226e-01
4.38075870e-01 2.84436882e-01 2.97092438e-01 -8.39909732e-01
-1.16411459e+00 9.67070386e-02 7.37225652e-01 -5.87186635e-01
-3.07247996e-01 -5.47718346e-01 -1.85265267e+00 -2.85239220e-01
-5.46019018e-01 5.45874953e-01 6.16822004e-01 9.04290497e-01
8.21899652e-01 8.27944994e-01 -9.31906924e-02 -8.51211697e-02
-9.48766470e-02 -1.05301797e+00 -3.02961916e-01 2.98771232e-01
-1.26735836e-01 -4.44820762e-01 1.10667720e-02 2.76029229e-01] | [8.509446144104004, 8.725177764892578] |
9ab6f356-af42-41c7-93da-e7f25cea0484 | token-event-role-structure-based-multi | 2306.17733 | null | https://arxiv.org/abs/2306.17733v1 | https://arxiv.org/pdf/2306.17733v1.pdf | Token-Event-Role Structure-based Multi-Channel Document-Level Event Extraction | Document-level event extraction is a long-standing challenging information retrieval problem involving a sequence of sub-tasks: entity extraction, event type judgment, and event type-specific multi-event extraction. However, addressing the problem as multiple learning tasks leads to increased model complexity. Also, existing methods insufficiently utilize the correlation of entities crossing different events, resulting in limited event extraction performance. This paper introduces a novel framework for document-level event extraction, incorporating a new data structure called token-event-role and a multi-channel argument role prediction module. The proposed data structure enables our model to uncover the primary role of tokens in multiple events, facilitating a more comprehensive understanding of event relationships. By leveraging the multi-channel prediction module, we transform entity and multi-event extraction into a single task of predicting token-event pairs, thereby reducing the overall parameter size and enhancing model efficiency. The results demonstrate that our approach outperforms the state-of-the-art method by 9.5 percentage points in terms of the F1 score, highlighting its superior performance in event extraction. Furthermore, an ablation study confirms the significant value of the proposed data structure in improving event extraction tasks, further validating its importance in enhancing the overall performance of the framework. | ['Xiping Liu', 'Dexi Liu', 'Hui Xiong', 'Keli Xiao', 'Changxuan Wan', 'Qizhi Wan'] | 2023-06-30 | null | null | null | null | ['retrieval', 'event-extraction', 'document-level-event-extraction', 'information-retrieval'] | ['methodology', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 3.67245883e-01 -4.54611797e-03 -5.24370134e-01 -2.09239557e-01
-1.26711762e+00 -5.42553961e-01 5.69135249e-01 1.01493025e+00
-7.19366491e-01 8.32006335e-01 4.48424429e-01 -2.41243258e-01
-2.79283762e-01 -9.09123421e-01 -5.11449516e-01 -5.14206707e-01
-3.14261287e-01 1.22139305e-02 5.09348691e-01 3.67965698e-01
1.24258578e-01 4.21798170e-01 -1.44319558e+00 5.29973149e-01
5.67473710e-01 1.06944656e+00 -1.89052060e-01 3.29327196e-01
-4.40787412e-02 1.04216850e+00 -7.12561965e-01 -2.57112831e-01
-7.11444765e-02 -2.93767661e-01 -7.47290254e-01 -5.18945932e-01
-2.67934680e-01 -4.02197480e-01 -1.00492090e-01 3.71584713e-01
6.02151573e-01 6.60327300e-02 6.43811047e-01 -1.26264405e+00
1.50298327e-01 8.51607800e-01 -5.20075738e-01 5.35769045e-01
4.46451306e-01 -4.58688706e-01 1.61981046e+00 -8.89947653e-01
6.11072481e-01 7.05577135e-01 6.47203863e-01 -6.80221757e-03
-9.46269631e-01 -8.15880656e-01 2.89679438e-01 4.91460830e-01
-1.44991505e+00 -2.85336554e-01 5.82136452e-01 -1.85953766e-01
1.37125456e+00 3.51450831e-01 4.97095525e-01 8.75388563e-01
4.41848151e-02 1.12854075e+00 6.82003736e-01 -4.36761677e-01
1.70126319e-01 -1.94543242e-01 2.84667224e-01 3.62587094e-01
5.22575319e-01 3.32867424e-03 -8.11486423e-01 -4.32818681e-01
3.97023022e-01 1.25285387e-01 9.88405868e-02 1.82897016e-01
-1.23854768e+00 6.45485342e-01 2.30899197e-03 3.82683396e-01
-7.12668896e-01 8.02942179e-03 8.32363784e-01 -1.05386660e-01
3.25262010e-01 3.32694024e-01 -7.31199563e-01 -3.23156476e-01
-1.05641317e+00 4.64288414e-01 7.87077665e-01 6.87418580e-01
2.82399565e-01 -2.22838014e-01 -5.27577162e-01 6.89293325e-01
2.40943655e-01 1.30709723e-01 2.27423519e-01 -3.95693332e-01
6.49395049e-01 9.75939810e-01 1.32140024e-02 -8.85254502e-01
-7.48823583e-01 -6.47122681e-01 -4.53174591e-01 -4.70748484e-01
3.64601463e-01 -2.52589226e-01 -3.49005073e-01 1.84901702e+00
4.71295565e-01 4.25420702e-01 8.72330368e-02 5.38399279e-01
8.42480123e-01 5.04318058e-01 5.90797126e-01 -3.90977472e-01
1.89795029e+00 -5.47169149e-01 -8.12263668e-01 -1.13773011e-01
7.30891943e-01 -6.46336198e-01 2.92285025e-01 2.63215899e-01
-1.03157914e+00 -7.00549036e-02 -1.09343147e+00 -5.87615184e-03
-3.18611830e-01 1.47453249e-01 1.04525828e+00 3.16220284e-01
-1.39481023e-01 1.57490671e-01 -7.09922373e-01 1.12103689e-02
4.83095586e-01 3.61555696e-01 -2.22894043e-01 1.99738190e-01
-1.71109700e+00 8.43536139e-01 7.76477754e-01 -3.10526252e-01
-5.16927600e-01 -1.06903350e+00 -7.79800057e-01 4.97421235e-01
8.37016582e-01 -5.65463364e-01 1.27230346e+00 -1.17586518e-03
-7.18382597e-01 4.46829855e-01 -4.33028489e-01 -6.67982280e-01
1.14995837e-01 -3.14919323e-01 -7.05875754e-01 4.29975659e-01
1.94385499e-01 1.46296009e-01 5.05071402e-01 -9.00179386e-01
-1.33032060e+00 -1.52368411e-01 -3.71275656e-02 2.60180444e-01
-4.02058005e-01 3.06785107e-01 -5.54584742e-01 -9.57027853e-01
-1.28136724e-01 -4.92993474e-01 -1.42146245e-01 -5.77882171e-01
-3.70014608e-01 -5.25659561e-01 5.54021001e-01 -5.54153979e-01
1.98807657e+00 -2.07821226e+00 -3.20185423e-01 2.97870457e-01
2.65790701e-01 -1.84559654e-02 3.06405514e-01 5.64618707e-01
-2.86067188e-01 6.29531667e-02 6.83506802e-02 -1.61543146e-01
-1.39156391e-03 -2.01631896e-02 -3.76606941e-01 9.46365744e-02
4.53310728e-01 9.04748678e-01 -9.73323703e-01 -7.75717020e-01
5.15114702e-02 3.70734125e-01 -4.27814901e-01 4.26230133e-02
7.67150074e-02 6.18411414e-02 -7.86869645e-01 7.42514729e-01
2.20168531e-01 -4.46777761e-01 3.55837077e-01 -2.86540151e-01
-1.02606997e-01 7.25119889e-01 -1.57017100e+00 1.32131279e+00
-2.71068424e-01 1.77355438e-01 -3.47968072e-01 -1.07803738e+00
6.44754410e-01 7.10657954e-01 1.11072993e+00 -8.06240261e-01
-8.38891417e-02 3.27274531e-01 -6.47337809e-02 -4.00854349e-01
6.07626736e-01 -4.61074375e-02 -4.75888044e-01 5.00723660e-01
-1.20770037e-01 5.75469017e-01 6.17895901e-01 4.41244632e-01
1.47890675e+00 -2.34727368e-01 8.68935108e-01 8.77968147e-02
3.83448601e-01 -1.42496139e-01 7.63550997e-01 8.23354006e-01
3.79059575e-02 2.90311426e-01 7.00974941e-01 -1.44356176e-01
-5.64757347e-01 -9.19870198e-01 -3.43895465e-01 1.02474046e+00
-5.67930676e-02 -9.96755540e-01 -3.60759169e-01 -8.39293480e-01
7.97952488e-02 6.18052244e-01 -5.33494830e-01 -4.79653254e-02
-8.00691307e-01 -1.40710711e+00 1.03488135e+00 8.04533780e-01
2.81365365e-01 -9.53490376e-01 -8.30145001e-01 6.78869545e-01
-5.34671843e-01 -1.45313740e+00 -1.58552453e-01 4.70218927e-01
-6.84780896e-01 -1.29635990e+00 -2.48468980e-01 -4.78620321e-01
2.13854328e-01 -1.16504855e-01 1.08808851e+00 -2.34492674e-01
-3.69746953e-01 1.20166376e-01 -5.38062036e-01 -6.73160911e-01
-7.98344538e-02 2.40539968e-01 -2.74189949e-01 5.13589531e-02
5.65158308e-01 -5.04095316e-01 -7.39620566e-01 1.91935822e-01
-1.09031665e+00 -5.26161827e-02 9.12612379e-01 6.94130540e-01
6.05952919e-01 3.85618657e-01 1.14431357e+00 -9.02661920e-01
5.05268693e-01 -7.70641208e-01 -2.30433479e-01 2.48663306e-01
-8.74003708e-01 -2.75185164e-02 4.70378697e-01 -3.64268750e-01
-1.23142266e+00 -4.27036844e-02 4.50933771e-03 3.85554135e-01
-6.00067526e-02 8.92562687e-01 -1.92376375e-01 5.85343778e-01
3.52523893e-01 1.58919156e-01 -5.41999400e-01 -2.99122959e-01
8.48594606e-02 4.65975821e-01 4.22923088e-01 -5.87508082e-01
5.32604456e-01 4.29006994e-01 2.04598233e-01 -4.50399429e-01
-9.91880298e-01 -7.77328372e-01 -4.18649107e-01 6.29770905e-02
7.64972925e-01 -1.17626250e+00 -8.65194201e-01 3.49548548e-01
-1.07132709e+00 2.14382887e-01 -2.79506773e-01 6.60740852e-01
-2.12871116e-02 3.03955555e-01 -7.19785273e-01 -6.56607568e-01
-4.40051883e-01 -7.40951419e-01 1.20255053e+00 8.47740248e-02
-4.31206435e-01 -8.88425112e-01 -7.85609782e-02 3.83073181e-01
2.30148900e-02 3.41698289e-01 1.16604102e+00 -1.14036918e+00
-6.32196605e-01 -4.77158934e-01 -3.81979197e-01 -2.91993350e-01
7.12197721e-02 -3.69634658e-01 -7.47995138e-01 9.58916023e-02
-2.13345364e-01 2.78492477e-02 8.71303976e-01 2.77431220e-01
1.01353538e+00 -2.87983473e-02 -6.06425464e-01 1.60639167e-01
1.24019063e+00 3.23876768e-01 4.62764502e-01 5.73660672e-01
4.55811918e-01 3.65741074e-01 7.87770629e-01 9.69932079e-01
7.03057349e-01 8.76697004e-01 1.53301507e-01 -2.40610659e-01
2.68159192e-02 -2.04433084e-01 1.10438108e-01 4.15923148e-01
8.28771070e-02 -3.52345586e-01 -8.77472699e-01 7.12752938e-01
-2.02393246e+00 -1.10925078e+00 -2.61121184e-01 2.05714250e+00
1.07939851e+00 3.99753571e-01 1.72375545e-01 6.19261384e-01
4.65040743e-01 1.71443701e-01 -3.98509949e-01 5.04064560e-03
-1.29291341e-01 3.03969145e-01 4.37648118e-01 2.83078626e-02
-1.32901168e+00 5.94133556e-01 6.14583397e+00 1.05835927e+00
-7.90527821e-01 -1.70556493e-02 2.10632101e-01 -2.26276055e-01
-7.85486400e-02 -8.65769759e-02 -1.26701438e+00 4.53395933e-01
1.05311441e+00 -4.04344022e-01 -1.56229034e-01 4.17352051e-01
4.28468764e-01 -4.27891552e-01 -1.21856523e+00 7.44465768e-01
-7.22275227e-02 -1.42982197e+00 8.41979459e-02 1.01198763e-01
2.47813776e-01 -2.66902655e-01 -4.14979070e-01 3.92198920e-01
-1.37770206e-01 -6.17493153e-01 6.14297509e-01 2.88144886e-01
4.64523554e-01 -7.43809819e-01 6.90657496e-01 1.69435635e-01
-1.66521084e+00 -2.65792280e-01 3.59600037e-01 1.09465070e-01
5.24160504e-01 1.15293729e+00 -1.06209683e+00 1.22305822e+00
5.70636809e-01 7.61976540e-01 -4.96546805e-01 1.04054081e+00
-1.27830967e-01 9.05459166e-01 -5.63174546e-01 1.50736794e-01
3.31249759e-02 4.70320553e-01 5.38640857e-01 1.67248392e+00
1.30511031e-01 2.79674023e-01 2.94262826e-01 4.28938746e-01
-2.03947306e-01 2.34232843e-01 -1.68919668e-01 3.50290816e-03
8.02466929e-01 1.29846942e+00 -9.11751807e-01 -4.39652205e-01
-7.37756729e-01 3.41786027e-01 1.14387743e-01 4.69789505e-02
-1.04073906e+00 -6.25225186e-01 2.84048378e-01 -5.40800169e-02
5.01982272e-01 1.21299818e-01 -4.56826001e-01 -1.05368567e+00
2.90917456e-01 -7.21026361e-01 1.02254140e+00 -1.85802221e-01
-1.01176429e+00 2.75366217e-01 3.14857244e-01 -1.28223002e+00
-5.14653802e-01 -1.91891178e-01 -4.42809522e-01 5.80792367e-01
-1.57791865e+00 -1.21687126e+00 1.04129493e-01 5.76942742e-01
3.61529052e-01 9.43594053e-02 8.68254066e-01 8.62833142e-01
-6.80735290e-01 7.94845700e-01 -2.14735121e-01 3.14998835e-01
7.66868412e-01 -1.23348069e+00 6.37532212e-03 9.37349916e-01
2.50104126e-02 5.46555638e-01 2.39022851e-01 -7.52100527e-01
-1.23881960e+00 -1.04690409e+00 1.34371650e+00 -3.92450184e-01
7.39081204e-01 -3.84555548e-01 -8.35191905e-01 4.57566619e-01
-2.68728346e-01 -1.19763516e-01 1.08599365e+00 4.14582998e-01
-4.05174255e-01 -9.11143422e-02 -9.02532935e-01 4.50785756e-01
8.50514054e-01 -4.96108770e-01 -7.74821401e-01 3.55601422e-02
4.68217403e-01 -2.43304089e-01 -1.26381373e+00 6.18901253e-01
6.74372375e-01 -5.23304939e-01 1.24348855e+00 -4.97096002e-01
4.91106629e-01 -2.38015950e-01 1.86877057e-03 -7.30864465e-01
-1.56749979e-01 -5.50891101e-01 -7.64705360e-01 1.51863134e+00
7.52822995e-01 -4.19760197e-01 4.72090691e-01 5.97545147e-01
1.25687733e-01 -9.61137712e-01 -8.93817425e-01 -5.23470402e-01
-4.40051824e-01 -6.92283690e-01 6.22116685e-01 1.06040800e+00
2.38177434e-01 6.56931400e-01 -3.12829763e-01 3.37688267e-01
4.07149792e-01 3.79355282e-01 1.39774755e-01 -1.36568987e+00
-3.44448328e-01 -3.00775379e-01 -8.58714506e-02 -6.88290417e-01
2.13078707e-02 -1.02973843e+00 -2.49738976e-01 -1.53851128e+00
5.02480626e-01 -5.08926034e-01 -7.86756694e-01 7.59102881e-01
-6.88181162e-01 4.33713086e-02 -1.74146499e-02 1.20733380e-01
-9.70108449e-01 1.71618819e-01 6.97831392e-01 2.38707125e-01
-2.88530469e-01 1.05541624e-01 -8.48715305e-01 6.37422204e-01
6.09108269e-01 -7.74408340e-01 -2.51578242e-01 -4.48135622e-02
3.14958364e-01 2.17492670e-01 2.96922058e-01 -6.75714910e-01
3.40360612e-01 5.26782870e-02 5.40339947e-01 -8.33744884e-01
1.33270398e-01 -8.10014665e-01 3.32194753e-02 2.66536444e-01
-4.45305556e-01 -8.09002900e-04 4.14471328e-01 6.22656822e-01
-4.13685977e-01 -1.22373201e-01 1.83387339e-01 2.00942293e-01
-7.64896214e-01 1.67227253e-01 -4.48369712e-01 2.14813814e-01
1.08868337e+00 4.17366112e-03 -2.52270520e-01 5.34393825e-02
-7.80444086e-01 2.15504572e-01 -3.26573133e-01 5.10535300e-01
3.46023828e-01 -1.15206349e+00 -7.14802206e-01 8.27438980e-02
2.48221889e-01 -5.84347621e-02 2.49121279e-01 9.57370698e-01
1.84585869e-01 5.18693209e-01 1.59229428e-01 -3.11346561e-01
-1.46141660e+00 2.52482146e-01 -2.69232661e-01 -1.14131582e+00
-7.31228352e-01 4.39272761e-01 6.53281584e-02 8.41522142e-02
2.87572116e-01 -2.97418386e-01 -4.35902834e-01 6.14097416e-01
6.43232286e-01 5.09894609e-01 4.07267958e-01 -1.56229749e-01
-6.16695821e-01 4.66804728e-02 -3.07201475e-01 -1.00829765e-01
1.38632810e+00 2.70026736e-02 -1.67397242e-02 1.46908596e-01
8.94705415e-01 2.25077227e-01 -7.61973739e-01 -3.01127672e-01
5.09621561e-01 1.40858553e-02 7.08369166e-02 -1.15450323e+00
-7.20043063e-01 4.62044001e-01 1.38706967e-01 2.50254482e-01
1.34174395e+00 8.94015655e-02 9.52163219e-01 3.23589385e-01
2.19178408e-01 -1.00434864e+00 -1.61408782e-01 4.76488590e-01
4.09449965e-01 -1.05895805e+00 2.52827138e-01 -6.53302848e-01
-5.28977156e-01 8.29283297e-01 2.86004812e-01 2.85381407e-01
7.27593124e-01 6.15597606e-01 -3.71795148e-01 -2.11290389e-01
-9.77679312e-01 -2.73939699e-01 4.24583435e-01 -7.17937993e-03
5.06265163e-01 -4.23453515e-03 -6.03139758e-01 1.17834771e+00
2.14482009e-01 1.15644872e-01 5.07837050e-02 1.14069557e+00
-1.69895008e-01 -1.46220088e+00 -1.81586757e-01 8.19481194e-01
-1.14530075e+00 -3.53926271e-01 -3.14495564e-02 8.75840306e-01
1.10112198e-01 1.00183988e+00 -7.24418908e-02 -1.71433631e-02
6.08927429e-01 2.81546026e-01 1.81541741e-01 -4.78785425e-01
-9.97538805e-01 3.08544755e-01 4.18696940e-01 -5.50522864e-01
-4.72457290e-01 -8.74369979e-01 -1.63321626e+00 1.42147705e-01
-3.39565068e-01 3.02228779e-01 4.03588980e-01 1.18145394e+00
7.50753582e-01 9.66306508e-01 4.13131058e-01 -2.67272413e-01
-2.99574226e-01 -7.02498972e-01 -3.89223427e-01 5.14737606e-01
1.69789299e-01 -7.73349226e-01 -1.09352358e-02 1.07621305e-01] | [9.094644546508789, 9.161330223083496] |
cfe91fe9-3864-40b6-ba51-73d86209179b | structured-prediction-as-translation-between-1 | 2101.05779 | null | https://arxiv.org/abs/2101.05779v3 | https://arxiv.org/pdf/2101.05779v3.pdf | Structured Prediction as Translation between Augmented Natural Languages | We propose a new framework, Translation between Augmented Natural Languages (TANL), to solve many structured prediction language tasks including joint entity and relation extraction, nested named entity recognition, relation classification, semantic role labeling, event extraction, coreference resolution, and dialogue state tracking. Instead of tackling the problem by training task-specific discriminative classifiers, we frame it as a translation task between augmented natural languages, from which the task-relevant information can be easily extracted. Our approach can match or outperform task-specific models on all tasks, and in particular, achieves new state-of-the-art results on joint entity and relation extraction (CoNLL04, ADE, NYT, and ACE2005 datasets), relation classification (FewRel and TACRED), and semantic role labeling (CoNLL-2005 and CoNLL-2012). We accomplish this while using the same architecture and hyperparameters for all tasks and even when training a single model to solve all tasks at the same time (multi-task learning). Finally, we show that our framework can also significantly improve the performance in a low-resource regime, thanks to better use of label semantics. | ['Stefano Soatto', 'Bing Xiang', 'Cicero Nogueira dos santos', 'Rishita Anubhai', 'Alessandro Achille', 'Jie Ma', 'Jason Krone', 'Ben Athiwaratkun', 'Giovanni Paolini'] | 2021-01-14 | structured-prediction-as-translation-between | https://openreview.net/forum?id=US-TP-xnXI | https://openreview.net/pdf?id=US-TP-xnXI | iclr-2021-1 | ['joint-entity-and-relation-extraction', 'nested-named-entity-recognition'] | ['natural-language-processing', 'natural-language-processing'] | [ 4.13137347e-01 8.43846798e-01 -7.37275481e-01 -4.40117598e-01
-1.20202279e+00 -7.75703549e-01 1.06838071e+00 5.98084152e-01
-7.85883009e-01 1.18291616e+00 5.43650806e-01 -5.88774122e-02
3.84958684e-02 -4.77142334e-01 -4.82591093e-01 -1.79786980e-01
-1.49949148e-01 1.17248964e+00 3.92940581e-01 -3.64434838e-01
-3.99364918e-01 1.71396285e-01 -1.04776955e+00 6.00148976e-01
6.45269036e-01 8.56543481e-01 -8.11344758e-02 3.90773505e-01
-1.43126603e-02 1.52844429e+00 -1.98811993e-01 -6.23076200e-01
-2.93965101e-01 -1.31966710e-01 -1.75478828e+00 -1.78174034e-01
-5.15906438e-02 2.04029888e-01 -2.79339731e-01 5.48795462e-01
3.64471763e-01 3.15406680e-01 4.03048277e-01 -1.31607270e+00
-1.02423973e-01 1.02508497e+00 -2.00258225e-01 -2.12131530e-01
7.31659174e-01 -3.45404208e-01 1.50675607e+00 -7.45984793e-01
1.07405424e+00 1.39822900e+00 5.55054724e-01 6.29640341e-01
-1.27416754e+00 -5.58695138e-01 2.88439572e-01 4.14945364e-01
-9.71315801e-01 -6.75531447e-01 4.57303345e-01 -8.20802376e-02
1.43905246e+00 3.16397727e-01 1.87606871e-01 1.37728393e+00
-3.53213370e-01 1.06022286e+00 9.35610354e-01 -6.85190022e-01
-1.18872195e-01 -1.55154970e-02 6.26835346e-01 7.35197425e-01
-3.02799568e-02 -1.62226334e-01 -7.66567290e-01 -4.94372100e-01
1.77796438e-01 -5.54946363e-01 -3.26581717e-01 -3.01435381e-01
-1.65398848e+00 6.04549348e-01 -3.85489501e-02 5.43211579e-01
-2.79729366e-01 -7.92034790e-02 8.16880822e-01 3.56463730e-01
6.03275478e-01 8.22783709e-01 -1.34022498e+00 -1.86880454e-01
-2.23824352e-01 3.11251402e-01 1.47796190e+00 1.02838302e+00
6.00539744e-01 -6.37856901e-01 -3.80252808e-01 9.91858423e-01
5.33838347e-02 1.21552192e-01 3.49259198e-01 -1.16332161e+00
6.90482318e-01 7.43640184e-01 2.49386132e-01 -2.77445316e-01
-9.32756722e-01 -2.03484111e-02 -7.64652312e-01 -6.15597427e-01
5.83299279e-01 -3.31853330e-01 -5.30907393e-01 2.11209726e+00
5.84405005e-01 2.43607149e-01 5.32054901e-01 5.12025952e-01
1.07540417e+00 5.84394395e-01 6.41194403e-01 -5.91920495e-01
1.99644387e+00 -1.23146784e+00 -9.83599186e-01 -5.42315781e-01
1.11626804e+00 -3.13773274e-01 6.21263266e-01 5.49336486e-02
-8.92488539e-01 -1.28775850e-01 -7.19807625e-01 -4.38685268e-01
-3.65063936e-01 1.22420415e-01 1.09478450e+00 -3.14582922e-02
-7.18414128e-01 3.79949450e-01 -7.74096549e-01 -6.30077183e-01
-4.04465236e-02 3.64761025e-01 -8.72149825e-01 1.76709011e-01
-1.75797689e+00 1.40289521e+00 9.58519399e-01 -1.53106943e-01
-6.24271274e-01 -7.49050915e-01 -1.27796030e+00 3.84791270e-02
1.04103649e+00 -5.74385941e-01 1.66002333e+00 -4.54116315e-01
-1.49158800e+00 1.34298885e+00 -2.50531971e-01 -8.79506767e-01
6.55122474e-02 -4.77138281e-01 -3.77087653e-01 5.82046248e-03
2.65297085e-01 5.05434632e-01 2.33345460e-02 -9.33781862e-01
-8.04014444e-01 -3.24889868e-01 1.90114647e-01 4.39591616e-01
1.40677914e-02 7.10177541e-01 -4.29767609e-01 -3.36578310e-01
-1.59862056e-01 -9.42446172e-01 -4.03521836e-01 -4.65472132e-01
-6.10454857e-01 -9.49733853e-01 5.41000366e-01 -7.33521581e-01
9.17642474e-01 -1.93076789e+00 4.12253052e-01 -2.94888049e-01
2.18834892e-01 3.19762647e-01 -2.01881990e-01 4.15459186e-01
-3.21825951e-01 -4.91913520e-02 -3.85549277e-01 -5.55924296e-01
1.36302430e-02 4.19440746e-01 -1.69851884e-01 1.86447129e-01
2.92565852e-01 1.03455031e+00 -1.14185047e+00 -7.21992254e-01
4.96313125e-02 1.42966151e-01 -3.38444471e-01 3.43682736e-01
-6.14727259e-01 6.82623327e-01 -6.38153076e-01 4.10266548e-01
7.16039091e-02 -4.81944591e-01 1.00791919e+00 -2.85264105e-01
6.80911243e-02 9.70777333e-01 -1.15330493e+00 1.82112586e+00
-7.30260670e-01 3.30695957e-01 1.40593991e-01 -1.19720066e+00
6.50261164e-01 9.15971637e-01 7.16395676e-01 -5.86672366e-01
-1.60079703e-01 1.76024869e-01 -2.18427658e-01 -4.62948054e-01
3.86413902e-01 1.22457348e-01 -6.05693281e-01 5.67450821e-01
6.31012201e-01 2.31866121e-01 2.84773767e-01 2.95471221e-01
1.15669072e+00 3.44569921e-01 1.13159597e+00 -1.00749180e-01
7.22886562e-01 3.57660651e-02 7.88599730e-01 6.17234826e-01
-7.07832947e-02 -1.21242650e-01 7.43113399e-01 -5.56990921e-01
-6.74771070e-01 -5.97882867e-01 -1.18033960e-01 1.69750690e+00
-3.15987542e-02 -7.63134480e-01 -2.73743182e-01 -1.16524136e+00
-1.72461107e-01 9.24035430e-01 -4.89136010e-01 -5.21434136e-02
-1.06554210e+00 -8.59964669e-01 9.29273129e-01 3.23427111e-01
6.30248010e-01 -1.38625944e+00 -2.67709583e-01 3.96245867e-01
-7.11754322e-01 -1.90850544e+00 -2.46305004e-01 5.86625338e-01
-4.97514755e-01 -1.43826890e+00 -2.30767757e-01 -9.41064894e-01
-9.32944715e-02 -4.69546884e-01 1.61340952e+00 -3.81514102e-01
1.31608665e-01 1.66116819e-01 -3.34568232e-01 -5.55563532e-02
-4.09558713e-01 5.80544710e-01 -1.22612357e-01 1.83565933e-02
2.10445389e-01 -4.99308407e-01 1.89262807e-01 1.06200963e-01
-2.71608025e-01 2.25074649e-01 3.25513482e-01 9.43240345e-01
4.13694680e-01 -5.01464903e-01 6.19490325e-01 -1.51036620e+00
3.31502318e-01 -3.82855624e-01 -2.18646094e-01 6.56801164e-01
-5.72272778e-01 3.28842223e-01 4.10002887e-01 -2.14720756e-01
-1.43940067e+00 2.29719386e-01 -3.17230485e-02 3.04672837e-01
-4.76388931e-01 5.46442211e-01 -6.05127990e-01 4.10570383e-01
5.92205286e-01 3.01839616e-02 -3.98837447e-01 -8.19574535e-01
7.06878424e-01 4.45128292e-01 7.10825920e-01 -8.55606973e-01
4.48383629e-01 2.30369538e-01 7.30837137e-02 -5.82361281e-01
-1.65811610e+00 -5.49725413e-01 -8.74457121e-01 5.05542815e-01
1.10914254e+00 -1.26826298e+00 -9.27696288e-01 2.73935348e-01
-1.57486868e+00 -4.99844223e-01 -3.44708771e-01 4.20434594e-01
-4.84801382e-01 3.13238859e-01 -9.77809846e-01 -6.58722818e-01
-5.04324675e-01 -5.41708648e-01 1.23714769e+00 -3.75293531e-02
-4.74439740e-01 -1.14813876e+00 2.98051029e-01 5.76770484e-01
-1.52068570e-01 3.08777392e-01 1.08468652e+00 -1.48357844e+00
-1.47518322e-01 6.04812689e-02 -3.32629025e-01 -1.33538648e-01
-3.39475237e-02 -6.88867748e-01 -8.50555301e-01 2.91114524e-02
-2.28117362e-01 -7.37072468e-01 8.90567780e-01 -1.17493518e-01
6.86086893e-01 -4.79818851e-01 -6.39093161e-01 3.73029411e-01
7.44864106e-01 1.45592809e-01 3.23120713e-01 3.35790157e-01
7.60470510e-01 8.54134679e-01 8.19842041e-01 1.53063819e-01
1.01354504e+00 1.18099058e+00 -7.28349686e-02 -4.59836535e-02
-1.34657353e-01 -1.98140964e-01 2.01930672e-01 4.09816772e-01
-1.62749663e-01 -1.41305178e-01 -1.10400712e+00 5.97322822e-01
-2.41703463e+00 -8.65962744e-01 -2.30591431e-01 1.71091211e+00
1.50532889e+00 -1.37014523e-01 2.21290246e-01 -1.11931674e-01
5.74797928e-01 3.55928093e-01 -2.88720787e-01 -1.55246099e-02
-3.31567019e-01 3.14173669e-01 3.08363497e-01 5.72494090e-01
-1.67528796e+00 1.44729328e+00 5.52607298e+00 6.30991936e-01
-5.02724648e-01 5.83672166e-01 4.89895493e-01 3.64703417e-01
2.42366046e-01 3.01067501e-01 -9.75749910e-01 -5.41844219e-02
1.04990697e+00 -1.53186992e-01 2.77065188e-01 7.10405171e-01
-3.97947520e-01 3.20880301e-02 -1.47691214e+00 8.51290047e-01
-8.25928748e-02 -1.26957572e+00 -3.07263970e-01 -3.23371530e-01
4.15209562e-01 1.67675167e-01 -6.32496834e-01 7.49645293e-01
7.41033494e-01 -8.04751635e-01 3.98810953e-01 1.87392235e-01
7.64927089e-01 -3.96996200e-01 7.60439754e-01 4.43630487e-01
-1.34043086e+00 1.07249049e-02 1.68895200e-01 -5.79379778e-03
2.45342746e-01 4.71640766e-01 -9.03607190e-01 9.69053388e-01
2.98818678e-01 1.01741624e+00 -4.02670652e-01 5.28178692e-01
-6.87025011e-01 5.87831914e-01 -2.91535288e-01 1.38116524e-01
-3.70389447e-02 3.24529111e-01 7.35916197e-01 1.28789651e+00
-4.67911750e-01 2.17492461e-01 6.96403384e-01 4.12952930e-01
-5.34621656e-01 2.23658293e-01 -4.09233153e-01 -5.33739030e-02
6.31827712e-01 1.41865051e+00 -2.94816136e-01 -4.08735961e-01
-4.51348782e-01 9.07092154e-01 7.12980986e-01 1.50526419e-01
-5.13421595e-01 -6.82174936e-02 3.48098904e-01 -3.05923462e-01
-7.82290623e-02 -6.68226331e-02 -7.00445985e-03 -1.55802011e+00
-2.85215676e-01 -9.28346395e-01 9.24335003e-01 -3.86900455e-01
-1.15954256e+00 7.04687178e-01 3.06251854e-01 -6.53762996e-01
-7.56535292e-01 -4.92586046e-01 -2.86436051e-01 5.79434991e-01
-1.44406140e+00 -1.61405003e+00 2.13729054e-01 7.15650678e-01
4.49237555e-01 -1.57987341e-01 1.37233126e+00 4.24612731e-01
-5.61915934e-01 5.08237898e-01 -5.26194453e-01 6.33216739e-01
1.07402992e+00 -1.43015385e+00 4.52425033e-01 4.54467386e-01
3.20462257e-01 1.68453708e-01 4.24409717e-01 -5.79263330e-01
-1.13355625e+00 -1.19789588e+00 1.75235236e+00 -6.93832219e-01
8.69722962e-01 -6.00200772e-01 -7.06921458e-01 1.19215882e+00
2.45103449e-01 1.14111334e-01 6.64950073e-01 8.72760534e-01
-5.29204369e-01 3.56483132e-01 -9.61448848e-01 3.14795434e-01
1.37027478e+00 -7.75104046e-01 -8.57571721e-01 7.24936008e-01
1.18378007e+00 -5.92457533e-01 -1.12135398e+00 6.24060094e-01
2.45046228e-01 -2.36032560e-01 9.60660219e-01 -1.23261356e+00
1.07834049e-01 5.65926060e-02 -1.68460652e-01 -1.06229973e+00
-9.71249416e-02 -9.69901979e-01 -6.20872736e-01 1.44944870e+00
9.11660612e-01 -6.13014460e-01 4.51072723e-01 7.57536054e-01
-3.99248488e-02 -5.22721291e-01 -1.14250898e+00 -4.66403306e-01
-3.36272538e-01 -2.68136382e-01 4.27849680e-01 1.34784758e+00
3.22901815e-01 1.31623375e+00 -5.12755990e-01 8.88008103e-02
5.80449641e-01 3.67603362e-01 5.26411891e-01 -1.63576519e+00
-4.88951087e-01 1.83218077e-01 1.53775945e-01 -7.18226910e-01
9.93447602e-01 -1.11539924e+00 -6.47173971e-02 -1.36776602e+00
2.99502552e-01 -5.92413962e-01 -1.74311623e-01 1.30975807e+00
-3.39534968e-01 -1.57795638e-01 3.00096661e-01 3.40928584e-01
-1.28151560e+00 5.01660407e-01 9.87466872e-01 -9.61806700e-02
-2.23247081e-01 7.97049329e-02 -6.79891467e-01 8.53964865e-01
4.32621747e-01 -6.25997484e-01 6.56389222e-02 -1.66453287e-01
1.46920666e-01 5.85975468e-01 2.92626917e-01 -5.01749516e-01
2.69203931e-01 -1.25575751e-01 -1.63324103e-01 -1.97282270e-01
4.09769118e-01 -6.27812386e-01 -1.46973014e-01 2.62273461e-01
-8.28858435e-01 -4.23899889e-01 3.56299244e-02 4.29350138e-01
-2.48031512e-01 -1.25073999e-01 4.51320320e-01 -1.82442650e-01
-9.29706037e-01 1.66116402e-01 -1.73523545e-01 5.41048527e-01
9.91032124e-01 7.14595914e-01 -4.70631927e-01 -2.92196125e-01
-1.31001103e+00 6.00782931e-01 -1.07799381e-01 4.68104661e-01
-5.94198406e-02 -1.17822659e+00 -9.69982982e-01 -2.67136663e-01
2.80879974e-01 8.24177042e-02 7.66571984e-03 8.87560070e-01
2.45894328e-01 7.63484061e-01 1.19410992e-01 -1.81491867e-01
-1.38220000e+00 3.79850388e-01 4.16625261e-01 -1.47490036e+00
-4.12365198e-01 7.31729925e-01 1.15961477e-01 -9.39263999e-01
2.94377267e-01 -1.84384137e-02 -6.31889582e-01 4.07388210e-01
3.00896794e-01 2.63200719e-02 9.60601941e-02 -5.60049653e-01
-6.66329503e-01 8.25461224e-02 -2.47152150e-01 6.81779347e-03
1.38866389e+00 1.37865664e-02 -4.05516207e-01 4.24541563e-01
8.45720768e-01 -1.47250354e-01 -7.08445072e-01 -8.35212290e-01
7.33210325e-01 4.28849667e-01 -2.35745966e-01 -1.14095092e+00
-5.82633853e-01 3.54458690e-01 -1.31600313e-02 1.67252675e-01
7.52241075e-01 6.02149844e-01 7.85752773e-01 9.11275685e-01
5.64281285e-01 -9.47235882e-01 -1.14934638e-01 1.05316687e+00
7.77137518e-01 -1.26892197e+00 -9.95810106e-02 -8.58479619e-01
-9.16008294e-01 7.06317484e-01 6.69200599e-01 1.54994383e-01
5.56263685e-01 3.75590444e-01 -3.91182035e-01 -1.73203290e-01
-1.35659194e+00 -5.50339282e-01 3.86888444e-01 5.78092933e-01
7.89277911e-01 1.62425175e-01 -3.00198674e-01 1.03954935e+00
1.62445456e-01 -2.06222951e-01 2.87944637e-02 7.96593249e-01
-3.93299982e-02 -1.63036561e+00 1.85060114e-01 3.00237775e-01
-6.49956763e-01 -3.27425301e-01 -6.61154866e-01 9.31863964e-01
-5.42287864e-02 9.59185600e-01 -1.93066999e-01 -1.34305969e-01
4.92715359e-01 5.84778845e-01 4.18674320e-01 -1.03814638e+00
-9.12118971e-01 -2.29903951e-01 1.24626875e+00 -7.46794999e-01
-7.90521145e-01 -6.25734389e-01 -1.35639703e+00 2.03183532e-01
-2.35216275e-01 5.07644475e-01 4.31053713e-02 1.39630771e+00
3.44896704e-01 4.63200390e-01 3.35857391e-01 -4.83221024e-01
-3.91972065e-01 -1.15081894e+00 -2.21141949e-01 4.38537985e-01
5.46864467e-03 -6.56507671e-01 1.18235454e-01 -5.30155003e-02] | [9.656341552734375, 9.01121711730957] |
1b737380-4667-46d6-9fa3-2d86a78c745d | weakly-supervised-scene-text-generation-for | 2306.14269 | null | https://arxiv.org/abs/2306.14269v2 | https://arxiv.org/pdf/2306.14269v2.pdf | Weakly Supervised Scene Text Generation for Low-resource Languages | A large number of annotated training images is crucial for training successful scene text recognition models. However, collecting sufficient datasets can be a labor-intensive and costly process, particularly for low-resource languages. To address this challenge, auto-generating text data has shown promise in alleviating the problem. Unfortunately, existing scene text generation methods typically rely on a large amount of paired data, which is difficult to obtain for low-resource languages. In this paper, we propose a novel weakly supervised scene text generation method that leverages a few recognition-level labels as weak supervision. The proposed method is able to generate a large amount of scene text images with diverse backgrounds and font styles through cross-language generation. Our method disentangles the content and style features of scene text images, with the former representing textual information and the latter representing characteristics such as font, alignment, and background. To preserve the complete content structure of generated images, we introduce an integrated attention module. Furthermore, to bridge the style gap in the style of different languages, we incorporate a pre-trained font classifier. We evaluate our method using state-of-the-art scene text recognition models. Experiments demonstrate that our generated scene text significantly improves the scene text recognition accuracy and help achieve higher accuracy when complemented with other generative methods. | ['Yue Lu', 'Cong Liu', 'Bing Yin', 'Palaiahankote Shivakum', 'Hongjian Zhan', 'Xinyuan Chen', 'Yangchen Xie'] | 2023-06-25 | null | null | null | null | ['scene-text-recognition', 'text-generation'] | ['computer-vision', 'natural-language-processing'] | [ 6.76530123e-01 -4.25995409e-01 1.00357518e-01 -5.59857905e-01
-7.46863365e-01 -6.96786404e-01 8.71103287e-01 -1.59327522e-01
-1.87360004e-01 3.62670541e-01 2.08136991e-01 -2.24657089e-01
6.52068019e-01 -7.31463075e-01 -8.08527827e-01 -5.55686474e-01
1.01411271e+00 3.46774876e-01 9.59157720e-02 -2.54827023e-01
2.23464996e-01 2.92130858e-01 -1.50156021e+00 7.24924982e-01
1.16107583e+00 7.59525359e-01 5.07706285e-01 6.03838861e-01
-7.28220522e-01 1.00256801e+00 -7.74281740e-01 -5.61565936e-01
2.57751018e-01 -6.44817650e-01 -4.41505075e-01 5.41328788e-01
1.04516411e+00 -3.74184757e-01 -3.22972119e-01 9.92572486e-01
4.09873515e-01 -1.89148307e-01 6.80666029e-01 -1.04541898e+00
-9.20672178e-01 7.08693802e-01 -5.23368835e-01 -3.11833739e-01
2.30810449e-01 3.56401533e-01 8.21983278e-01 -1.08099449e+00
5.04405200e-01 1.05077660e+00 3.51720154e-01 4.41203058e-01
-1.45751035e+00 -6.35135949e-01 2.29081810e-01 -1.66510493e-01
-1.27421510e+00 -5.21422982e-01 1.13345432e+00 -6.32110775e-01
7.65271485e-01 1.38738811e-01 4.63353485e-01 1.58498216e+00
-2.47691557e-01 1.00806129e+00 1.20407104e+00 -6.74366891e-01
-2.65191519e-03 4.21220124e-01 -2.64859170e-01 7.10999012e-01
1.56664506e-01 -3.75805140e-01 -6.39468729e-01 3.08899760e-01
7.37652481e-01 -2.85917670e-02 -2.95474470e-01 -5.96850336e-01
-1.22146869e+00 7.57882357e-01 1.33029357e-01 2.49716416e-01
1.71060905e-01 -2.07828104e-01 3.00187081e-01 1.15533084e-01
4.96685088e-01 5.07048368e-01 -1.40707910e-01 3.03318584e-03
-1.07966042e+00 5.65993898e-02 7.39473045e-01 1.25771880e+00
6.03127599e-01 4.50588465e-01 -3.15842032e-01 1.19219756e+00
4.58954908e-02 1.00053787e+00 5.43656051e-01 -1.29546836e-01
1.06003892e+00 9.14311230e-01 -7.19720125e-02 -9.60644662e-01
4.86894585e-02 -1.01749331e-01 -9.65008676e-01 -2.02166475e-03
3.74134719e-01 -1.91151595e-03 -1.05933595e+00 1.51911497e+00
-3.20499614e-02 -3.70204926e-01 -3.94004323e-02 6.47728443e-01
7.38406122e-01 5.91217756e-01 -1.15842327e-01 2.26063132e-01
1.06321669e+00 -1.28844225e+00 -5.22853613e-01 -5.24537921e-01
6.30081773e-01 -1.20225549e+00 1.60587883e+00 1.20884702e-01
-8.62522185e-01 -7.29275405e-01 -1.02059937e+00 -2.52804637e-01
-3.86576086e-01 7.94642568e-01 4.11572993e-01 7.88179278e-01
-7.65483558e-01 1.58928290e-01 -6.45411074e-01 -3.48540306e-01
5.09002626e-01 -3.86107899e-02 -1.88362166e-01 -4.63659197e-01
-6.63799524e-01 5.36194205e-01 2.86218315e-01 1.92190323e-03
-6.05984151e-01 -4.42284733e-01 -1.07669139e+00 8.05923063e-03
4.15452003e-01 -7.03728676e-01 9.92046058e-01 -1.20998514e+00
-1.61390758e+00 8.73172820e-01 -8.61370638e-02 -6.12115860e-03
8.51694763e-01 -3.44304830e-01 -2.70809114e-01 6.00828379e-02
2.32838377e-01 7.56954610e-01 1.18897235e+00 -1.32693744e+00
-3.93272519e-01 -2.12604776e-01 -3.47881258e-01 4.11822796e-01
-7.36412764e-01 -5.03087640e-02 -6.96410656e-01 -1.08444309e+00
-1.18862599e-01 -8.32379997e-01 1.07572908e-02 -2.68025901e-02
-6.48622036e-01 2.21272469e-01 9.54424918e-01 -7.11216867e-01
9.05814111e-01 -2.16877723e+00 7.26005435e-02 5.09074470e-03
-1.05618216e-01 2.33990610e-01 -3.33357602e-01 4.02804583e-01
2.59702116e-01 6.52442649e-02 -1.02552392e-01 -5.77311456e-01
2.32557766e-02 3.38985287e-02 -5.96408963e-01 2.25373823e-02
5.35777509e-01 8.34866345e-01 -5.92968524e-01 -5.67331433e-01
4.65463430e-01 3.99656087e-01 -5.58960795e-01 3.94621998e-01
-5.24588287e-01 4.86669451e-01 -3.71764123e-01 7.04871833e-01
4.97637868e-01 -3.52676392e-01 3.31797525e-02 -2.70344496e-01
2.71253847e-02 2.03690052e-01 -1.08278382e+00 1.84543490e+00
-5.44831991e-01 8.91702235e-01 -1.93588927e-01 -7.19774663e-01
1.11111283e+00 -8.66862088e-02 9.94232595e-02 -7.60543466e-01
3.28040570e-01 5.19814864e-02 -2.54145801e-01 -4.40155774e-01
8.86152208e-01 8.40165541e-02 -2.10927770e-01 4.64289159e-01
-6.04908131e-02 -6.86342359e-01 4.12384510e-01 3.23991895e-01
8.15787315e-01 3.17254007e-01 -5.46949022e-02 9.07495022e-02
4.01008964e-01 6.49760365e-02 2.06217319e-01 9.59902585e-01
9.81044099e-02 9.11894679e-01 2.59067208e-01 -2.82885909e-01
-1.42495251e+00 -7.96922386e-01 7.11430758e-02 1.04542184e+00
6.53947443e-02 -4.11908448e-01 -8.51653337e-01 -7.38542378e-01
-1.17332481e-01 6.94278896e-01 -4.38133597e-01 -8.83052945e-02
-4.54852998e-01 -5.81900597e-01 7.17200816e-01 7.24523902e-01
7.05810189e-01 -9.23461914e-01 -2.46824235e-01 -1.50338441e-01
-2.75766462e-01 -1.57571387e+00 -7.48021722e-01 -5.03246449e-02
-5.92487991e-01 -6.87185585e-01 -7.09195077e-01 -8.81294072e-01
1.00570095e+00 5.30263305e-01 1.11703694e+00 -5.69103956e-02
-4.42934245e-01 3.61982435e-01 -4.85687315e-01 -3.51409435e-01
-7.07824349e-01 9.79338735e-02 -1.91215575e-01 2.79231787e-01
1.41276047e-01 -1.17209859e-01 -2.40553439e-01 2.37935722e-01
-1.09718645e+00 8.42993855e-01 6.60446703e-01 1.00096107e+00
3.83534640e-01 -3.97223514e-03 7.22753033e-02 -1.08356392e+00
4.27804649e-01 4.96250428e-02 -6.89864457e-01 4.00290728e-01
-2.75882989e-01 1.27283305e-01 1.02943861e+00 -6.75033212e-01
-1.36643982e+00 3.67691278e-01 1.87328368e-01 -4.21902686e-01
-3.62653971e-01 2.03727245e-01 -4.88229424e-01 3.67029868e-02
5.97809196e-01 5.43186903e-01 -3.52134109e-01 -4.75457698e-01
6.24713302e-01 8.10832679e-01 5.42003870e-01 -7.52168417e-01
9.71665382e-01 4.35957283e-01 -3.21035594e-01 -1.07475674e+00
-1.03793478e+00 -2.09250018e-01 -8.63574684e-01 1.29871322e-02
7.69680381e-01 -1.08411169e+00 1.21690221e-01 8.28939795e-01
-1.06147122e+00 -6.42444789e-01 -1.74168438e-01 1.98354721e-01
-4.00612980e-01 4.97016191e-01 -5.85728467e-01 -6.27047896e-01
-1.96302414e-01 -1.19170785e+00 1.48631263e+00 7.41211548e-02
-3.26264687e-02 -7.88845003e-01 -1.51224017e-01 7.46811271e-01
3.34231019e-01 5.76679222e-02 9.87084031e-01 -5.00921249e-01
-8.27171922e-01 -1.91251233e-01 -5.72080433e-01 3.51404637e-01
4.16630268e-01 4.08711433e-02 -9.78898644e-01 -1.96180984e-01
-2.58033454e-01 -6.24323428e-01 7.65987933e-01 -9.57440585e-02
1.21919250e+00 -1.78462774e-01 4.01172042e-02 7.22273827e-01
1.23475468e+00 -9.93659645e-02 3.87193829e-01 2.18281820e-01
1.48252416e+00 5.80842376e-01 3.91791165e-01 4.50574785e-01
3.82313818e-01 6.53546154e-01 -9.53473449e-02 -4.75243390e-01
-5.38792193e-01 -6.09983504e-01 4.29425508e-01 8.19972515e-01
3.56245577e-01 -5.13895035e-01 -9.13274348e-01 3.46155107e-01
-1.57600164e+00 -8.12729478e-01 -1.49390280e-01 1.99859428e+00
1.10402179e+00 1.13419153e-01 -1.49041072e-01 3.44864428e-02
7.25835204e-01 1.17010310e-01 -6.07840300e-01 -8.56931359e-02
-6.71433449e-01 2.18856372e-02 3.10972840e-01 1.20408222e-01
-1.06772852e+00 1.42335892e+00 5.68078947e+00 7.33376503e-01
-1.32522058e+00 -3.15955520e-01 5.26636183e-01 -2.16827422e-01
-2.70865232e-01 -6.05068281e-02 -9.54343259e-01 5.55367589e-01
4.25094485e-01 1.68994784e-01 4.65330154e-01 9.52683091e-01
1.60243139e-01 4.86290790e-02 -1.18727624e+00 1.21539187e+00
8.52895081e-01 -1.20517910e+00 5.45838237e-01 -1.21919543e-01
1.12374032e+00 -1.54876485e-01 1.75861537e-01 2.23364353e-01
3.67088974e-01 -9.34479415e-01 9.95262861e-01 1.96537018e-01
1.20149326e+00 -3.33570153e-01 3.12077463e-01 2.92226255e-01
-1.01814544e+00 1.44699022e-01 -3.01307380e-01 1.30355597e-01
8.74254704e-02 5.86618304e-01 -1.00455487e+00 4.02704716e-01
5.11641145e-01 7.18962073e-01 -1.13681996e+00 6.30418062e-01
-4.84663963e-01 5.20838678e-01 -1.86780199e-01 -1.37641072e-01
7.12566227e-02 -1.89246550e-01 2.13886857e-01 1.34738445e+00
3.17400426e-01 -3.69467258e-01 4.42615539e-01 1.13948178e+00
-3.29299361e-01 4.08958763e-01 -8.43851864e-01 -4.36881661e-01
2.19416276e-01 1.21169484e+00 -7.75981128e-01 -4.57738906e-01
-6.57254875e-01 1.45618939e+00 3.87631446e-01 3.40443641e-01
-7.04073489e-01 -3.87855649e-01 1.02955781e-01 -1.97456609e-02
2.71011472e-01 -2.99263179e-01 -6.60706401e-01 -1.73361468e+00
2.67619342e-01 -1.16154575e+00 8.66309479e-02 -1.10108316e+00
-1.31715977e+00 5.39604664e-01 -3.92074496e-01 -1.18665636e+00
-8.05705935e-02 -7.98953712e-01 -3.99850637e-01 8.77804697e-01
-1.33925176e+00 -1.89349914e+00 -6.46020412e-01 6.05111361e-01
1.05221248e+00 -4.99918997e-01 6.14454389e-01 1.87146112e-01
-7.89574385e-01 8.05834770e-01 1.74566939e-01 4.78632331e-01
1.09830225e+00 -1.31952465e+00 5.39935052e-01 1.17985761e+00
5.07784307e-01 4.96490240e-01 4.27780628e-01 -7.42223680e-01
-1.51567757e+00 -1.32606840e+00 6.05463505e-01 -5.56204796e-01
6.22684419e-01 -9.27719593e-01 -8.61107647e-01 5.49378157e-01
1.56859353e-01 -2.70055622e-01 7.15820849e-01 1.60807017e-02
-7.60049224e-01 6.95118234e-02 -5.57952642e-01 1.08297217e+00
9.22175467e-01 -6.93748474e-01 -4.35491055e-01 3.43661845e-01
4.46797997e-01 -4.63580489e-01 -3.70318800e-01 1.88632756e-01
3.36897582e-01 -8.93526077e-01 6.82637155e-01 -3.31437051e-01
8.39690089e-01 -2.51221418e-01 -1.91229418e-01 -1.17214751e+00
-3.55019122e-02 -4.03912067e-01 2.65500456e-01 1.65145886e+00
4.17065173e-01 -1.81227356e-01 8.46033573e-01 7.33355939e-01
-1.52451590e-01 -2.39472389e-01 -1.91145778e-01 -6.72544420e-01
1.83927361e-02 -3.54055494e-01 5.67281127e-01 1.06501627e+00
-2.12241858e-01 8.06653917e-01 -6.73334122e-01 -8.79707187e-02
4.23981220e-01 4.25053984e-01 1.28505051e+00 -8.55214238e-01
-3.98260653e-01 -5.40778100e-01 -1.41657457e-01 -1.15294993e+00
3.50439370e-01 -9.30403709e-01 1.90564632e-01 -1.35509920e+00
4.54221517e-01 -5.36359906e-01 3.30089867e-01 4.77790445e-01
-3.70973587e-01 2.50773787e-01 2.89338797e-01 1.65584326e-01
-5.99019945e-01 7.88601935e-01 1.25578535e+00 -4.15474415e-01
-1.68329567e-01 -3.23768556e-01 -6.96883678e-01 7.47358918e-01
7.97949076e-01 -1.46711946e-01 -4.80396569e-01 -9.30255830e-01
1.61986485e-01 -2.28360981e-01 1.74482509e-01 -8.58600855e-01
1.55654043e-01 -2.78257936e-01 7.21604645e-01 -7.28170514e-01
3.28011960e-01 -5.46093404e-01 -3.42330277e-01 -8.46631359e-03
-4.70144689e-01 -1.47709772e-01 1.75225794e-01 5.11634827e-01
-1.82864308e-01 -1.29200071e-01 7.30760336e-01 -3.45750302e-02
-4.66554433e-01 1.28473401e-01 -8.51829499e-02 2.36559093e-01
6.25699162e-01 -1.51218772e-01 -5.27263463e-01 -2.47959316e-01
-8.63137096e-02 2.88728569e-02 1.02555048e+00 6.61795139e-01
5.21112084e-01 -1.21577775e+00 -6.77082479e-01 4.72895622e-01
5.36238253e-01 1.89142317e-01 1.80258498e-01 3.23603481e-01
-4.75496113e-01 2.87188619e-01 -1.10392720e-01 -7.61268795e-01
-1.36916614e+00 5.20288050e-01 -5.11017488e-03 -1.97056264e-01
-5.64763427e-01 5.72005808e-01 4.67741102e-01 -3.72939765e-01
7.32959062e-02 -2.47137949e-01 6.39469028e-02 -1.56699255e-01
4.90743548e-01 -5.51626086e-02 7.65959769e-02 -7.45094180e-01
2.62085855e-01 6.41738057e-01 -2.95276493e-01 -2.16602147e-01
1.07838750e+00 -1.79117039e-01 6.57381266e-02 4.18397039e-01
8.94113541e-01 3.20384562e-01 -1.41671634e+00 -3.97712111e-01
-2.14290649e-01 -7.12457061e-01 -1.51393255e-02 -8.16096187e-01
-8.31613600e-01 1.09306049e+00 2.56301284e-01 -1.17584988e-01
1.08934033e+00 -2.49341115e-01 7.45925903e-01 4.87947255e-01
1.75906911e-01 -1.22066915e+00 5.53204477e-01 6.19556785e-01
7.87450731e-01 -1.65902972e+00 -1.29423320e-01 -5.74156046e-01
-9.31049883e-01 1.01550865e+00 9.01407659e-01 2.78638542e-01
-4.20870967e-02 3.38967502e-01 2.70886540e-01 1.46809816e-01
-4.79294777e-01 -1.66168123e-01 4.98282433e-01 5.14549851e-01
5.43168783e-01 3.25601846e-02 2.32803002e-01 4.46793854e-01
-4.17962879e-01 -4.01276708e-01 6.73019052e-01 9.55107689e-01
-1.45341486e-01 -1.27815294e+00 -3.72232795e-01 4.70855445e-01
-2.94781148e-01 -4.40506041e-01 -8.19208860e-01 4.72024888e-01
-1.77336112e-01 8.44703615e-01 -2.89800242e-02 -2.68519402e-01
3.14552933e-01 1.93893850e-01 5.83655775e-01 -8.09088588e-01
-3.00399542e-01 3.24192703e-01 3.73468362e-03 -2.40397125e-01
-2.07688436e-01 -6.01189733e-01 -7.49570429e-01 -8.06484520e-02
-3.45809579e-01 -4.21245337e-01 7.95239806e-01 8.40431094e-01
2.80234188e-01 4.94415283e-01 7.58394778e-01 -8.39329660e-01
-4.40356672e-01 -9.48438585e-01 -3.80590230e-01 8.97562146e-01
-1.24319522e-02 -3.68504465e-01 -1.61104381e-01 6.60991490e-01] | [11.8420991897583, 1.865731120109558] |
0139bf15-0c3f-4070-a780-ae860a79b9a1 | conki-contrastive-knowledge-injection-for | 2306.15796 | null | https://arxiv.org/abs/2306.15796v1 | https://arxiv.org/pdf/2306.15796v1.pdf | ConKI: Contrastive Knowledge Injection for Multimodal Sentiment Analysis | Multimodal Sentiment Analysis leverages multimodal signals to detect the sentiment of a speaker. Previous approaches concentrate on performing multimodal fusion and representation learning based on general knowledge obtained from pretrained models, which neglects the effect of domain-specific knowledge. In this paper, we propose Contrastive Knowledge Injection (ConKI) for multimodal sentiment analysis, where specific-knowledge representations for each modality can be learned together with general knowledge representations via knowledge injection based on an adapter architecture. In addition, ConKI uses a hierarchical contrastive learning procedure performed between knowledge types within every single modality, across modalities within each sample, and across samples to facilitate the effective learning of the proposed representations, hence improving multimodal sentiment predictions. The experiments on three popular multimodal sentiment analysis benchmarks show that ConKI outperforms all prior methods on a variety of performance metrics. | ['Di Niu', 'Lei Yang', 'Xiaoli Wang', 'Weidong Guo', 'Baoxun Wang', 'Feiran Sun', 'Shi-ang Qi', 'Mingjun Zhao', 'Yakun Yu'] | 2023-06-27 | null | null | null | null | ['contrastive-learning', 'multimodal-sentiment-analysis', 'contrastive-learning', 'general-knowledge', 'sentiment-analysis', 'multimodal-sentiment-analysis'] | ['computer-vision', 'computer-vision', 'methodology', 'miscellaneous', 'natural-language-processing', 'natural-language-processing'] | [ 1.87475279e-01 -7.98953101e-02 -1.86504677e-01 -4.65254992e-01
-1.08750093e+00 -7.68347204e-01 7.59634674e-01 2.00252950e-01
-1.97123274e-01 4.15403187e-01 6.63091719e-01 2.05650285e-01
2.22056592e-03 -4.28383827e-01 -6.32551372e-01 -7.94288635e-01
4.26705688e-01 1.11458145e-01 -1.87150881e-01 -3.43156099e-01
7.34459683e-02 2.90652588e-02 -1.30104840e+00 7.66060412e-01
4.92871851e-01 1.26677740e+00 -3.47507238e-01 6.01560235e-01
-3.91179919e-01 9.92414951e-01 -6.31092787e-01 -9.68628764e-01
-2.48116016e-01 -3.84359360e-01 -7.60800540e-01 -3.39061730e-02
3.60172331e-01 -8.12861845e-02 -9.36160013e-02 9.84374344e-01
4.45558190e-01 9.78111178e-02 8.24572325e-01 -1.27843690e+00
-6.11619949e-01 8.02317798e-01 -7.49396920e-01 -2.62635082e-01
7.08444118e-01 -5.21712638e-02 1.18126786e+00 -1.10128486e+00
2.58369654e-01 1.36445487e+00 5.09827137e-01 5.08440554e-01
-1.03494728e+00 -6.75863743e-01 3.66897643e-01 1.36526927e-01
-1.10126936e+00 -5.01351535e-01 1.11116934e+00 -3.98509711e-01
5.71714878e-01 1.99340492e-01 2.94591665e-01 1.33258641e+00
-1.52586237e-01 1.10469091e+00 1.03704417e+00 -3.78922254e-01
9.41148847e-02 4.49302405e-01 3.78143787e-01 7.10452914e-01
-2.04364583e-01 -3.75463426e-01 -1.23247778e+00 -2.27038845e-01
3.15665632e-01 5.46955764e-02 -1.69739753e-01 -3.58518213e-01
-1.27336681e+00 8.96979511e-01 3.61614734e-01 1.83348536e-01
-3.46254021e-01 -4.77677919e-02 7.22571552e-01 2.18684211e-01
2.76943952e-01 1.67817995e-01 -5.85651457e-01 -6.80784732e-02
-5.72048724e-01 -2.09377855e-01 7.42190063e-01 6.20813608e-01
1.11457372e+00 7.82646239e-02 -2.85873175e-01 9.66438591e-01
6.29831314e-01 7.21136272e-01 7.94326127e-01 -6.28267705e-01
6.54654682e-01 1.07245421e+00 -1.33763626e-01 -9.68637347e-01
-3.43197167e-01 -8.17286372e-02 -6.29048645e-01 -2.04390466e-01
2.06497803e-01 -5.14807820e-01 -8.43796611e-01 1.67351389e+00
3.24650049e-01 2.12259009e-01 6.45255744e-01 7.95911789e-01
1.45790875e+00 5.07214844e-01 3.10829490e-01 9.49601457e-02
1.47526681e+00 -9.88090694e-01 -6.98160410e-01 -1.88219592e-01
5.51255941e-01 -6.11728072e-01 1.03867328e+00 3.47499251e-01
-8.57741535e-01 -3.32807928e-01 -7.99753428e-01 5.73533028e-02
-7.69764900e-01 2.88200557e-01 6.85790956e-01 1.00289893e+00
-8.73731375e-01 -3.05570960e-01 -4.77732867e-01 -1.92823842e-01
4.28921819e-01 5.28612375e-01 -6.08852148e-01 -1.34627596e-01
-1.08069551e+00 6.18605971e-01 1.83619753e-01 3.11454296e-01
-8.43725860e-01 -6.62874997e-01 -1.12972641e+00 3.39219086e-02
2.36900643e-01 -8.12600017e-01 1.17233646e+00 -1.58788681e+00
-1.88786376e+00 4.32448834e-01 -3.71695191e-01 2.06937212e-02
-5.78150861e-02 -2.90007263e-01 -5.33872604e-01 3.76496613e-01
-2.16833189e-01 7.53863037e-01 1.13625836e+00 -1.67607439e+00
-4.50534910e-01 -4.10011560e-01 3.31918895e-01 5.30750275e-01
-8.28594804e-01 1.58293545e-02 -6.18522108e-01 -4.10687566e-01
-1.67454079e-01 -6.34987414e-01 7.68976882e-02 -7.55211174e-01
-5.11019766e-01 -2.81513482e-02 8.72416437e-01 -6.67271256e-01
9.56354201e-01 -2.19157362e+00 4.95352864e-01 5.40389836e-01
8.93624201e-02 -6.11292869e-02 -4.74834681e-01 4.18879271e-01
-1.95351113e-02 -1.94725484e-01 -5.07935062e-02 -6.32602274e-01
2.67597139e-01 1.22155905e-01 -4.30460513e-01 8.20016041e-02
3.86127532e-01 1.06071341e+00 -6.63057685e-01 -3.79725695e-01
1.21674724e-01 8.10501277e-01 -3.92379940e-01 1.48912340e-01
-1.49136521e-02 4.60838884e-01 -4.52112496e-01 1.11649287e+00
6.12136960e-01 -1.94211841e-01 4.66213375e-01 -7.09172010e-01
3.77296925e-01 -1.78519920e-01 -9.72969174e-01 1.60665631e+00
-6.25089407e-01 2.33375460e-01 8.22914168e-02 -8.96024525e-01
7.78488934e-01 4.15597558e-01 3.45619589e-01 -5.42148352e-01
4.51934427e-01 -2.43695512e-01 -4.50978309e-01 -3.92242044e-01
6.45989954e-01 -3.60407501e-01 -4.47263926e-01 3.88941079e-01
4.74709928e-01 1.44609839e-01 -4.57975082e-02 4.11299378e-01
7.04657733e-01 -1.02446049e-01 7.57201985e-02 4.57923234e-01
8.12091947e-01 -2.05631509e-01 2.11624756e-01 5.67261279e-01
-1.11099958e-01 3.83509398e-01 6.37066960e-01 3.07824649e-02
-1.14603691e-01 -9.18025732e-01 2.55394787e-01 1.61333752e+00
1.58311680e-01 -3.55570465e-01 -5.20556867e-01 -1.09806859e+00
2.32434738e-02 3.38098079e-01 -8.78440499e-01 -2.73646981e-01
1.20281512e-02 -8.84127140e-01 6.46619022e-01 7.18512058e-01
3.02700847e-01 -9.56232250e-01 -1.08373925e-01 -3.30242991e-01
-2.90525556e-01 -1.19722855e+00 -1.97041824e-01 9.40803364e-02
-6.92703366e-01 -1.01231158e+00 -6.82167590e-01 -7.47724652e-01
7.39838779e-01 1.75213993e-01 7.66413867e-01 -3.45877528e-01
3.26244742e-01 1.31343150e+00 -5.60370862e-01 -3.48990411e-01
-2.01936200e-01 1.31183043e-01 -2.39135608e-01 8.15572500e-01
4.34282422e-01 -7.09747151e-02 -3.44980210e-01 1.30660892e-01
-9.99141634e-01 -2.48765662e-01 8.27468097e-01 9.18376982e-01
2.80970991e-01 -1.20056979e-01 8.17691207e-01 -7.95704961e-01
6.83927000e-01 -7.77147949e-01 -1.14401653e-01 5.32732368e-01
3.87623422e-02 -1.21852443e-01 3.74126464e-01 -5.80071807e-01
-1.54984927e+00 2.57833093e-01 1.89131379e-01 -5.68247259e-01
-1.91024184e-01 1.02871358e+00 -3.87389362e-01 -1.90633848e-01
3.34937781e-01 3.03176552e-01 2.19229260e-03 -2.01051399e-01
7.29687274e-01 6.12219095e-01 3.37036312e-01 -5.88147759e-01
5.67644417e-01 4.70305592e-01 -2.35571325e-01 -7.69861102e-01
-8.44790637e-01 -5.06171823e-01 -4.88556474e-01 -4.60694909e-01
8.77371311e-01 -1.27905858e+00 -1.08513951e+00 4.20388013e-01
-6.93366349e-01 -2.50458885e-02 -4.15609777e-03 6.37500882e-01
-1.66657194e-01 4.40318733e-01 -5.41551590e-01 -8.41633558e-01
-2.36307070e-01 -1.13414514e+00 1.33760881e+00 4.56542134e-01
-2.16760382e-01 -1.39394915e+00 1.12412840e-01 9.59726870e-01
2.80404329e-01 -2.14275643e-01 8.11581075e-01 -7.98853099e-01
-1.98191747e-01 -5.13360322e-01 -3.32151860e-01 4.98991698e-01
9.58193690e-02 -6.99834004e-02 -1.42818248e+00 -8.54504108e-03
-3.56036246e-01 -7.58282900e-01 1.06805384e+00 5.53783253e-02
7.31092095e-01 -2.24102020e-01 -8.57426524e-02 1.33680940e-01
1.11535728e+00 -1.89955786e-01 4.35486108e-01 1.06229760e-01
9.52133536e-01 7.48335958e-01 3.65807861e-01 4.30193901e-01
8.27744305e-01 2.58922487e-01 4.90045518e-01 -1.06974848e-01
1.92626029e-01 3.66401449e-02 8.05052936e-01 8.45672131e-01
4.29147854e-02 -2.28879955e-02 -7.85201013e-01 5.45002341e-01
-1.91778898e+00 -7.73618042e-01 3.16489428e-01 1.80789220e+00
7.49688864e-01 -3.24729949e-01 2.70926714e-01 -1.46628767e-01
4.01148796e-01 1.21067941e-01 -4.12629485e-01 -3.58715236e-01
-5.13879657e-01 -2.85712583e-03 1.64002836e-01 3.73763204e-01
-1.28821540e+00 8.82515788e-01 6.36809301e+00 5.57624936e-01
-1.23383617e+00 1.65882275e-01 3.16312343e-01 -1.48922682e-01
-6.26004815e-01 -1.75055355e-01 -7.32899547e-01 7.03593716e-02
8.20972860e-01 1.62199154e-01 2.23846912e-01 6.04546249e-01
-2.82206416e-01 -2.16474846e-01 -8.01288247e-01 1.06179154e+00
5.41110277e-01 -1.04073715e+00 4.21104699e-01 -2.80597448e-01
8.19354117e-01 -2.31226563e-01 5.27912915e-01 6.48146212e-01
2.32297778e-01 -8.08682680e-01 3.84736508e-01 9.98649240e-01
1.94822907e-01 -1.00563705e+00 1.11671257e+00 -8.05594325e-02
-9.95283902e-01 -2.59947121e-01 1.15249880e-01 2.96791077e-01
-1.03121601e-01 2.74018556e-01 -7.38250196e-01 9.69257295e-01
6.01851583e-01 8.17446113e-01 -7.31614590e-01 4.62520927e-01
-1.02074079e-01 7.34663486e-01 -4.16735709e-02 -1.46756629e-02
9.33984146e-02 3.77642922e-03 1.28963292e-01 1.49080181e+00
1.14250429e-01 -3.35673302e-01 5.56250401e-02 2.85460144e-01
-2.81758755e-01 4.57856506e-01 -3.96306306e-01 -2.26990268e-01
7.22436681e-02 1.61528111e+00 -3.29415292e-01 -3.48176539e-01
-9.36501801e-01 8.34242105e-01 2.97051221e-01 7.03865170e-01
-6.85743153e-01 -1.84376806e-01 4.45840627e-01 -6.93672717e-01
6.10021174e-01 1.18151464e-01 -3.38725060e-01 -1.31410372e+00
-2.14743793e-01 -9.33031261e-01 7.59631872e-01 -7.93138564e-01
-1.58843076e+00 3.70152920e-01 -1.27621349e-02 -1.03698134e+00
-2.00922266e-01 -6.67474806e-01 -5.37005007e-01 8.10832441e-01
-1.75958478e+00 -1.83184481e+00 -2.09490746e-01 1.05260694e+00
1.43207446e-01 -5.27855575e-01 9.78350163e-01 1.59397557e-01
-7.68733740e-01 9.77439225e-01 -3.10712218e-01 1.97530329e-01
8.36748838e-01 -1.00830042e+00 -7.44815707e-01 4.83767271e-01
-2.78363992e-02 7.99290895e-01 2.69739211e-01 -4.54031944e-01
-1.79740250e+00 -8.70067358e-01 4.67147589e-01 -5.19078553e-01
8.57635975e-01 -1.29655361e-01 -8.98591876e-01 7.54455686e-01
6.12754047e-01 -2.74684638e-01 1.48612845e+00 5.50536394e-01
-8.43675792e-01 -2.68804163e-01 -9.54780459e-01 4.81338114e-01
1.60664633e-01 -9.26885128e-01 -4.55279589e-01 3.12578157e-02
4.61211771e-01 -7.22259507e-02 -1.07992589e+00 4.40272659e-01
7.53856122e-01 -6.94667161e-01 8.18047047e-01 -6.73872471e-01
5.27457595e-01 -4.42857802e-01 -6.50356948e-01 -1.13866770e+00
1.74263701e-01 -5.25096953e-02 -3.67483109e-01 1.58450770e+00
6.35201156e-01 -5.14572859e-01 6.13129199e-01 5.54644644e-01
1.12531096e-01 -4.47389007e-01 -8.28962564e-01 -1.80201337e-01
-8.66052061e-02 -5.22485733e-01 6.88616276e-01 1.27115309e+00
4.71781045e-01 7.64387310e-01 -4.89985853e-01 4.67724472e-01
3.14703345e-01 1.13469670e-02 9.83552158e-01 -8.18637431e-01
-2.73127556e-01 -5.78679264e-01 -4.77632731e-01 -6.79138303e-01
5.21015584e-01 -9.09785330e-01 -2.37283260e-01 -1.28912044e+00
4.36294377e-01 3.34260792e-01 -6.41577721e-01 9.03326333e-01
-2.87679732e-01 4.24373567e-01 1.76801354e-01 -4.43786867e-02
-9.57821906e-01 8.95693541e-01 1.14051533e+00 -4.28458124e-01
-2.05212116e-01 -2.78574407e-01 -1.17937267e+00 7.21094787e-01
6.22992039e-01 -2.65605835e-04 -4.46665257e-01 -2.41175070e-01
5.95006049e-01 -5.63567579e-02 4.38362867e-01 -6.21453106e-01
3.24522853e-01 9.81794968e-02 5.05364716e-01 -6.48115575e-01
8.73783827e-01 -8.22848558e-01 -2.91302443e-01 -1.80659786e-01
-2.96331942e-01 -2.50739098e-01 4.73374635e-01 6.31784081e-01
-7.33657300e-01 6.12199083e-02 3.25597107e-01 2.99929619e-01
-9.48921204e-01 -2.45440066e-01 -2.13911906e-01 -4.15753633e-01
7.58017123e-01 -1.73842516e-02 -5.48555970e-01 -6.92749023e-01
-8.64413142e-01 2.58490860e-01 -1.21909557e-02 4.80174392e-01
8.92878294e-01 -1.36711538e+00 -4.27161753e-01 1.56403199e-01
5.57854831e-01 -5.53961277e-01 7.58255363e-01 1.11110890e+00
2.47636840e-01 1.29456654e-01 -8.46964493e-03 -6.01287782e-01
-1.28848422e+00 3.53891820e-01 3.60836327e-01 -1.33460522e-01
1.94524482e-01 7.55934477e-01 4.70577985e-01 -9.34023261e-01
2.69032031e-01 1.20318070e-01 -7.87991107e-01 6.29047155e-01
5.69924355e-01 1.48198232e-01 -1.55682592e-02 -1.04995751e+00
-5.15613317e-01 7.25466609e-01 -2.18522415e-01 -2.93840528e-01
9.70098794e-01 -2.80419916e-01 -9.92574096e-02 7.97878146e-01
1.16686678e+00 2.45668903e-01 -9.14474666e-01 -4.60354269e-01
-2.54110366e-01 -9.98344421e-02 1.23689966e-02 -1.18543828e+00
-1.23246336e+00 7.80277550e-01 4.48974699e-01 -1.06212392e-01
1.39508212e+00 1.71650171e-01 4.27970946e-01 4.45302695e-01
-2.10084140e-01 -9.34332550e-01 5.77680051e-01 6.96059167e-01
7.89328694e-01 -1.48653948e+00 -1.74683437e-01 -2.40429133e-01
-1.44059813e+00 1.03417420e+00 5.25411189e-01 3.40011269e-01
7.25980997e-01 -9.61148441e-02 6.39291227e-01 -2.77284324e-01
-6.03870392e-01 -3.42553556e-01 7.16075003e-01 5.57356119e-01
4.81694162e-01 1.21502899e-01 3.95569742e-01 1.15976536e+00
7.06385076e-02 -2.99923867e-01 1.85755819e-01 9.78261769e-01
-1.20733991e-01 -1.02593899e+00 -5.37309825e-01 2.39325866e-01
-2.88261205e-01 -1.11589924e-01 -7.90616512e-01 5.86964667e-01
-1.30739644e-01 1.17350018e+00 -3.04557383e-01 -6.76621974e-01
3.51534158e-01 4.41976666e-01 3.06499302e-01 -4.40069646e-01
-8.65380943e-01 3.04459006e-01 6.01800829e-02 -3.88139993e-01
-9.80191529e-01 -7.41978049e-01 -1.15518391e+00 4.64890152e-02
-1.89429909e-01 -7.14315698e-02 7.19880581e-01 1.08896172e+00
4.19097811e-01 4.97998178e-01 7.06735849e-01 -8.45626712e-01
-1.62399113e-01 -9.15006459e-01 -5.32303691e-01 3.75553638e-01
4.53449309e-01 -6.42592549e-01 -3.50632131e-01 5.77225909e-02] | [13.138045310974121, 5.108783721923828] |
2720945a-f747-423a-b219-95323a8f7db9 | exploiting-socially-aware-tasks-for-embodied | 2212.00767 | null | https://arxiv.org/abs/2212.00767v2 | https://arxiv.org/pdf/2212.00767v2.pdf | Exploiting Proximity-Aware Tasks for Embodied Social Navigation | Learning how to navigate among humans in an occluded and spatially constrained indoor environment, is a key ability required to embodied agent to be integrated into our society. In this paper, we propose an end-to-end architecture that exploits Proximity-Aware Tasks (referred as to Risk and Proximity Compass) to inject into a reinforcement learning navigation policy the ability to infer common-sense social behaviors. To this end, our tasks exploit the notion of immediate and future dangers of collision. Furthermore, we propose an evaluation protocol specifically designed for the Social Navigation Task in simulated environments. This is done to capture fine-grained features and characteristics of the policy by analyzing the minimal unit of human-robot spatial interaction, called Encounter. We validate our approach on Gibson4+ and Habitat-Matterport3D datasets. | ['Lamberto Ballan', 'Angel X. Chang', 'Luciano Serafini', 'Tommaso Campari', 'Enrico Cancelli'] | 2022-12-01 | null | null | null | null | ['common-sense-reasoning', 'social-navigation'] | ['reasoning', 'robots'] | [-1.66897178e-01 2.78257608e-01 3.22976708e-01 -6.13207102e-01
-1.18192203e-01 -4.67105597e-01 7.89981604e-01 2.58106798e-01
-1.02840340e+00 9.84141946e-01 1.95974380e-01 -1.66363031e-01
-5.46446502e-01 -9.62647021e-01 -8.49093318e-01 -2.59109765e-01
-1.02405500e+00 4.20033693e-01 3.41899693e-01 -7.89366901e-01
1.43815309e-01 4.67741519e-01 -1.72064924e+00 -1.54789165e-01
8.32490504e-01 6.51664197e-01 6.09276712e-01 8.20831776e-01
4.92224038e-01 8.47222030e-01 -5.22723719e-02 1.96303055e-01
1.86551824e-01 -1.29599631e-01 -8.35109353e-01 -4.72249150e-01
-4.89583284e-01 -4.31606710e-01 -1.19292095e-01 5.76516271e-01
4.41281497e-01 8.38901877e-01 8.85604382e-01 -1.35569763e+00
-1.44923599e-02 5.35858035e-01 8.66671205e-02 1.29865080e-01
9.46006775e-01 5.43694317e-01 5.81300437e-01 -2.55771548e-01
7.78949857e-01 1.47666502e+00 5.18620133e-01 5.75648665e-01
-9.01667118e-01 -1.24372602e-01 3.68834585e-01 2.59686142e-01
-1.13778794e+00 -3.21661115e-01 4.51641142e-01 -3.32522899e-01
1.26564050e+00 7.51216784e-02 9.04710531e-01 1.49560368e+00
5.78900635e-01 6.52335405e-01 1.26135743e+00 -1.55930400e-01
8.80491853e-01 -2.64176816e-01 -2.56715328e-01 6.32557452e-01
1.36138976e-01 7.75450647e-01 -3.75292450e-01 -2.31776863e-01
5.12838900e-01 -2.22827047e-01 1.09697461e-01 -7.47650027e-01
-1.06819177e+00 5.81001759e-01 8.85188162e-01 1.91049933e-01
-7.89151132e-01 4.84533846e-01 8.62295032e-02 2.41085693e-01
-1.12472378e-01 4.69026297e-01 -4.79661077e-01 -4.45968539e-01
1.58585906e-01 6.94825828e-01 1.04094982e+00 1.02759039e+00
8.42448831e-01 -3.66941065e-01 2.00519606e-01 3.59097719e-01
5.44315577e-01 5.09559274e-01 1.47094131e-01 -1.21204650e+00
1.65917143e-01 4.33819473e-01 5.24114370e-01 -1.15386283e+00
-1.13322377e+00 -2.28501588e-01 -3.09242904e-01 7.48159349e-01
1.84595808e-01 -4.40741777e-01 -6.13513827e-01 1.99451375e+00
7.61250854e-01 1.59376159e-01 3.95138443e-01 9.34993625e-01
2.80924797e-01 3.70809257e-01 5.69351554e-01 3.28192472e-01
1.38968098e+00 -6.86050951e-01 -3.71622682e-01 -5.69256365e-01
7.77716041e-01 -2.81582531e-02 9.83768821e-01 1.47379324e-01
-6.87424541e-01 -2.65112966e-01 -9.50635076e-01 1.67992890e-01
-6.91360652e-01 -5.33042789e-01 7.69876897e-01 3.79557967e-01
-1.12813282e+00 4.98344511e-01 -1.10900748e+00 -1.12677956e+00
1.58023104e-01 2.59931684e-01 -5.03073871e-01 3.21206510e-01
-1.31429267e+00 1.19715047e+00 4.00738835e-01 -2.03353196e-01
-1.35865521e+00 -1.94144584e-02 -9.74041522e-01 -2.20778242e-01
4.62127507e-01 -8.92405212e-01 1.29550409e+00 -3.36542964e-01
-1.73909056e+00 6.00891411e-01 2.95530230e-01 -6.79223776e-01
6.27123237e-01 -4.93786275e-01 -1.38100490e-01 -1.64769784e-01
3.19092900e-01 8.15039158e-01 2.26640090e-01 -1.59423983e+00
-1.07102311e+00 -3.24171007e-01 4.46172267e-01 7.87186444e-01
2.67000973e-01 -4.24953610e-01 -2.66538691e-02 -3.74333858e-01
-1.34045586e-01 -1.09476948e+00 -9.16516244e-01 5.11147752e-02
-5.98597452e-02 -2.34518006e-01 3.06300044e-01 -1.73027202e-01
4.98754889e-01 -2.09870744e+00 2.34314129e-01 4.27150577e-01
-1.58793911e-01 -4.24469501e-01 -9.64643285e-02 7.65357792e-01
6.74988985e-01 -2.71485001e-01 -2.09732249e-01 -3.96976888e-01
4.18783903e-01 5.24122119e-01 -3.44061852e-02 4.50448304e-01
-2.91481525e-01 5.89981437e-01 -1.40837705e+00 -2.76890635e-01
3.70229006e-01 3.04543197e-01 -8.95607829e-01 5.04352450e-01
-2.63385534e-01 8.29274893e-01 -1.03321552e+00 3.40718985e-01
3.87373239e-01 5.60709059e-01 2.44123980e-01 5.83013535e-01
-4.81406897e-01 1.13201834e-01 -1.00458622e+00 2.06978011e+00
-6.74438775e-01 9.00416151e-02 2.40415931e-01 -5.15454054e-01
7.19948947e-01 -1.62854984e-01 2.56286442e-01 -9.07045424e-01
3.29478264e-01 1.71645105e-01 -2.14908764e-01 -7.03150511e-01
5.26182771e-01 2.97188461e-01 -6.77423537e-01 3.68352979e-01
-9.33729261e-02 -2.10516006e-01 -2.57728577e-01 3.99218127e-03
1.40218282e+00 5.79594433e-01 6.08447790e-01 -5.01929700e-01
3.64233792e-01 1.16123661e-01 2.77747393e-01 1.22020614e+00
-5.58614075e-01 -2.18927711e-01 1.23149462e-01 -5.85677803e-01
-5.92814624e-01 -1.40238988e+00 2.69235462e-01 1.49985707e+00
6.00288749e-01 -1.14142641e-01 -9.56567466e-01 -5.94864488e-01
-4.58941571e-02 1.04781401e+00 -8.72271299e-01 -2.65859783e-01
-7.02762187e-01 -3.40227187e-01 4.48600382e-01 2.81191647e-01
5.30522585e-01 -1.40957713e+00 -1.74506783e+00 3.24787080e-01
-1.04085781e-01 -1.02831447e+00 1.06698468e-01 5.65949440e-01
-2.66786873e-01 -9.27516222e-01 -6.16086163e-02 -7.55269766e-01
3.33809793e-01 2.72842627e-02 7.80090868e-01 -1.98095202e-01
-1.56999946e-01 9.23088193e-01 -6.59860492e-01 -2.25818187e-01
-1.13256052e-01 6.35246560e-02 4.93359983e-01 -2.56866932e-01
2.10881129e-01 -8.97558987e-01 -9.27618325e-01 4.00156647e-01
-3.15459967e-01 -5.97594865e-02 5.75142860e-01 4.78789002e-01
3.22990894e-01 -1.09626949e-02 2.77190328e-01 -1.27924502e-01
8.23937714e-01 -9.33072329e-01 -4.82166857e-01 7.24999458e-02
-2.21012965e-01 1.23276584e-01 3.79607201e-01 -2.32507572e-01
-1.09966230e+00 3.80986959e-01 -3.93868834e-02 5.74633956e-01
-6.28988564e-01 3.54513437e-01 -3.52133155e-01 -9.57759395e-02
8.04347217e-01 5.48465811e-02 -2.49689057e-01 -2.11589903e-01
6.65838540e-01 5.04336119e-01 4.73148435e-01 -9.62677121e-01
5.68339705e-01 6.78874016e-01 4.07266691e-02 -7.56419420e-01
-2.87747290e-02 -2.31852964e-01 -5.74850380e-01 -5.30319989e-01
1.06473410e+00 -8.48206878e-01 -1.39844894e+00 4.19373631e-01
-9.82266843e-01 -1.01696050e+00 -1.69562787e-01 5.68890870e-01
-1.41743934e+00 -8.60924199e-02 -5.86777031e-02 -1.16235638e+00
1.75598174e-01 -9.44363832e-01 9.32237148e-01 3.77960145e-01
-4.22802955e-01 -8.20493340e-01 4.83770102e-01 -3.38914782e-01
5.28958082e-01 5.54878891e-01 5.21804571e-01 -5.67224562e-01
-6.68574274e-01 3.43500316e-01 8.10981542e-02 -5.01278996e-01
-9.14113671e-02 -6.85234606e-01 -5.62249064e-01 -1.86281517e-01
-1.52661860e-01 -3.72900963e-01 3.03809494e-01 2.44533300e-01
5.04602909e-01 -1.85224786e-01 -7.58590043e-01 5.30988514e-01
1.11138272e+00 5.88603199e-01 5.08135378e-01 8.34295034e-01
2.51114905e-01 1.10029876e+00 1.12678945e+00 6.90672576e-01
1.16205978e+00 8.24125886e-01 7.74843574e-01 2.40810916e-01
2.32748196e-01 -5.29709160e-01 3.66154373e-01 1.13059007e-01
-1.96076944e-01 -3.89837742e-01 -1.07243025e+00 4.64952260e-01
-2.24646640e+00 -9.09192324e-01 1.75738737e-01 1.82805192e+00
2.39344582e-01 1.77125543e-01 8.41490328e-02 -3.66106659e-01
3.03381115e-01 -7.34517574e-02 -6.15670741e-01 -4.54350144e-01
1.83466822e-01 -2.17500433e-01 6.24780178e-01 9.75698769e-01
-1.07680690e+00 1.38071311e+00 6.29503775e+00 4.93589699e-01
-5.45755982e-01 5.36022382e-03 6.14467859e-02 2.81132847e-01
-6.04908690e-02 7.79453889e-02 -3.93063009e-01 1.28361881e-01
8.26985776e-01 1.23299539e-01 7.57377267e-01 1.09363306e+00
4.02175963e-01 -7.35782266e-01 -1.12871706e+00 4.87417728e-01
-3.70019227e-01 -4.89471346e-01 -3.35487217e-01 -3.78736295e-03
1.61526784e-01 5.95795028e-02 1.38261423e-01 5.98270595e-01
1.08344865e+00 -1.03875041e+00 1.12624860e+00 8.60184252e-01
2.38494024e-01 -8.45569491e-01 3.87675583e-01 7.17782319e-01
-1.33463466e+00 -4.30940658e-01 2.20723394e-02 -5.69011807e-01
5.19588411e-01 -3.09341103e-01 -1.11577499e+00 4.30242002e-01
9.50143039e-01 3.73555839e-01 -2.81516194e-01 8.85478258e-01
-2.03105554e-01 -1.80159166e-01 -6.05983317e-01 -6.67801678e-01
6.19501591e-01 -1.97748259e-01 6.71775162e-01 8.05811822e-01
4.82097149e-01 4.16253418e-01 3.90874475e-01 5.52595794e-01
6.77960038e-01 -1.12275533e-01 -1.04818749e+00 6.09803021e-01
5.45898139e-01 7.23114312e-01 -7.92892277e-01 1.73577294e-01
1.32224262e-01 1.08950806e+00 5.82829058e-01 4.00993615e-01
-9.02088761e-01 -3.02784771e-01 1.00388706e+00 -7.94622079e-02
6.70049489e-02 -5.17705679e-01 8.68025869e-02 -6.09496117e-01
-2.68839270e-01 -4.99292433e-01 1.58811033e-01 -7.15487659e-01
-9.45786357e-01 6.82431638e-01 2.19491750e-01 -1.12040055e+00
-4.02380854e-01 -5.39287984e-01 -3.97256643e-01 4.08620566e-01
-1.45385671e+00 -1.15876269e+00 -4.65899706e-01 7.94158578e-01
2.39079610e-01 -7.76749626e-02 9.90004182e-01 -1.64627895e-01
1.32026359e-01 2.37929821e-01 -1.58679768e-01 -3.13948095e-01
3.86223108e-01 -1.17301452e+00 6.73699081e-01 4.41236198e-01
-5.45521855e-01 4.95442152e-01 1.16097713e+00 -8.86646211e-01
-1.27826500e+00 -8.98707926e-01 4.65061814e-01 -6.42496109e-01
5.03628969e-01 -4.92021441e-01 -3.38704407e-01 7.48188078e-01
-4.61349972e-02 -2.34225377e-01 5.21926939e-01 9.23605040e-02
2.67139133e-02 1.42477736e-01 -1.55651593e+00 1.17840421e+00
1.76623273e+00 -2.77907759e-01 -7.84513950e-01 -9.15114507e-02
8.59535694e-01 -3.68733257e-01 -3.34053993e-01 4.33788538e-01
8.51066053e-01 -1.10572982e+00 1.09421468e+00 -5.86008966e-01
-1.04027033e-01 -3.19459975e-01 -6.13409281e-01 -1.60270369e+00
-3.05122435e-01 -6.60735071e-01 2.29107276e-01 6.82683051e-01
2.02996403e-01 -7.40943134e-01 7.55223095e-01 3.35702538e-01
-1.45171717e-01 -3.48987520e-01 -1.34815848e+00 -7.03974903e-01
-9.42247063e-02 -4.70958203e-01 7.68075824e-01 4.79457229e-01
1.53420031e-01 -1.98847830e-01 -4.06770498e-01 6.04160428e-01
7.83885121e-01 -5.80094755e-01 9.64431226e-01 -1.11685908e+00
-3.00932694e-02 -3.51377875e-01 -4.95739788e-01 -1.05622017e+00
2.94271111e-01 -5.42215884e-01 6.47826433e-01 -1.49855340e+00
-3.53197634e-01 -9.75893021e-01 -2.04156220e-01 1.06668852e-01
2.16059834e-01 -4.40789968e-01 -2.25152448e-02 -2.21423343e-01
-1.00783157e+00 8.90729070e-01 9.45185542e-01 1.82715192e-01
-4.67031687e-01 -1.56250242e-02 -3.38694006e-01 8.71738374e-01
8.64163160e-01 -3.92387629e-01 -5.44244587e-01 -4.37307179e-01
3.40278208e-01 2.71719366e-01 5.85569143e-01 -1.56404316e+00
6.16257071e-01 -5.93961477e-01 3.30966376e-02 -4.77182001e-01
6.62229240e-01 -1.20694470e+00 1.37172371e-01 9.49974418e-01
-3.99196774e-01 5.86657040e-02 -6.45678341e-02 1.03416657e+00
4.59906012e-01 5.72910905e-02 3.70014191e-01 -2.14461386e-01
-1.15356851e+00 2.35635582e-02 -1.06066489e+00 -1.68686002e-01
1.63913119e+00 -1.33877173e-01 -2.79160421e-02 -4.79976147e-01
-8.75355780e-01 7.51161754e-01 6.72749519e-01 4.84809697e-01
7.40334868e-01 -1.10262477e+00 -3.44018877e-01 1.40528575e-01
3.53571445e-01 -1.77369222e-01 4.61302578e-01 2.98882157e-01
-7.60250866e-01 -1.07991127e-02 -6.50349259e-01 -2.95385063e-01
-7.67709076e-01 6.02425396e-01 5.62402964e-01 -1.45080220e-02
-3.86827171e-01 8.23605299e-01 1.76189557e-01 -1.08311880e+00
3.54302168e-01 -4.48863506e-01 -5.69498122e-01 -3.44142407e-01
2.38340497e-01 3.73402208e-01 -5.06185293e-01 -7.10569680e-01
-6.35816097e-01 5.17330587e-01 4.52220857e-01 -5.95494688e-01
1.25981164e+00 -7.00295031e-01 2.69493639e-01 3.83050859e-01
6.58956289e-01 -2.11284354e-01 -1.67278600e+00 6.58155233e-02
3.35742384e-01 -2.65455484e-01 -2.38891676e-01 -8.76837671e-01
-6.78281412e-02 3.36086333e-01 9.51925933e-01 1.51312858e-01
5.62298298e-01 7.54141733e-02 6.19022727e-01 9.87429380e-01
1.30690801e+00 -1.23077035e+00 2.55498998e-02 7.95821488e-01
7.52815783e-01 -1.32479644e+00 -3.13552588e-01 -2.91845948e-02
-6.87694609e-01 6.30275846e-01 8.01020384e-01 -2.26426512e-01
7.27607071e-01 3.14264953e-01 -5.12882620e-02 -1.68824956e-01
-7.13205218e-01 -5.94744563e-01 -3.44523221e-01 1.32938647e+00
-3.87022316e-01 4.10327971e-01 -1.93966120e-01 6.35492504e-01
-4.67225254e-01 -1.94265962e-01 3.37834835e-01 1.48926914e+00
-8.44297230e-01 -7.31316984e-01 -2.61096030e-01 -2.13064402e-01
2.00100839e-01 4.29926306e-01 -2.14094132e-01 7.34683871e-01
4.40933555e-01 1.15404856e+00 8.54979381e-02 -5.21458685e-01
6.66703939e-01 -4.71421510e-01 3.29159439e-01 -3.79376948e-01
-5.01918256e-01 -6.10874414e-01 2.58518666e-01 -1.15743041e+00
-3.49872530e-01 -7.05480516e-01 -1.59829485e+00 -2.53775597e-01
4.51041132e-01 1.37500450e-01 7.78775692e-01 8.87726307e-01
2.97846943e-01 4.78554338e-01 6.36452258e-01 -1.25027919e+00
-2.83597767e-01 -6.11363590e-01 -2.38715470e-01 2.44291797e-01
5.51345944e-01 -1.08474374e+00 -2.91718930e-01 -3.82438153e-01] | [4.751788139343262, 0.8859140276908875] |
2c723c0d-ae64-4235-a535-0fed644b2b03 | automated-pancreas-segmentation-using-multi | 2009.13148 | null | https://arxiv.org/abs/2009.13148v1 | https://arxiv.org/pdf/2009.13148v1.pdf | Automated Pancreas Segmentation Using Multi-institutional Collaborative Deep Learning | The performance of deep learning-based methods strongly relies on the number of datasets used for training. Many efforts have been made to increase the data in the medical image analysis field. However, unlike photography images, it is hard to generate centralized databases to collect medical images because of numerous technical, legal, and privacy issues. In this work, we study the use of federated learning between two institutions in a real-world setting to collaboratively train a model without sharing the raw data across national boundaries. We quantitatively compare the segmentation models obtained with federated learning and local training alone. Our experimental results show that federated learning models have higher generalizability than standalone training. | ['Wei-Chung Wang', 'Kensaku MORI', 'Wei-Chih Liao', 'Kao-Lang Liu', 'Po-Ting Chen', 'Kazunari Misawa', 'Dong Yang', 'Chen Shen', 'Pochuan Wang', 'Holger R. Roth', 'Masahiro Oda', 'Daguang Xu'] | 2020-09-28 | null | null | null | null | ['pancreas-segmentation', 'automated-pancreas-segmentation'] | ['medical', 'medical'] | [-3.07416767e-01 8.64912346e-02 -2.76702255e-01 -6.73324943e-01
-1.02371502e+00 -3.15405786e-01 1.81247205e-01 1.16205379e-01
-7.89112329e-01 7.33618259e-01 1.15066446e-01 -5.93161225e-01
-7.14171827e-02 -9.35512960e-01 -7.75100231e-01 -5.42079568e-01
3.22116800e-02 3.97871912e-01 -1.37955606e-01 2.70327568e-01
-3.48553836e-01 4.80957031e-01 -9.06491518e-01 4.66419011e-01
1.05879211e+00 7.39787996e-01 -4.04681489e-02 6.79623663e-01
-2.38478780e-01 9.19542909e-01 -7.05963433e-01 -8.72915864e-01
7.63399601e-01 -3.40020031e-01 -1.01784825e+00 1.22814909e-01
7.17461407e-01 -8.37535083e-01 -4.20210660e-01 1.06831288e+00
4.68021452e-01 -1.87967330e-01 5.74844442e-02 -1.25200987e+00
-7.30678916e-01 5.22241831e-01 -1.75717667e-01 1.04105055e-01
-1.96219400e-01 3.80702972e-01 5.52370548e-01 -2.49785736e-01
9.87194598e-01 6.33903861e-01 8.45707059e-01 5.28900623e-01
-7.43592381e-01 -8.31964791e-01 -1.07784465e-01 -5.22884093e-02
-9.13675725e-01 -2.03523070e-01 4.84705359e-01 -4.10452694e-01
3.76207888e-01 1.64004266e-01 5.69236100e-01 9.44518805e-01
1.94451451e-01 7.18451977e-01 1.15885448e+00 -3.98980379e-01
1.68267161e-01 3.01289469e-01 9.35785100e-02 9.83968377e-01
5.79497755e-01 -6.16903007e-02 -2.86367685e-01 -3.68728012e-01
9.94149148e-01 4.25668001e-01 -1.11495599e-01 -2.64284790e-01
-9.96942878e-01 8.51423383e-01 5.86804867e-01 5.04776359e-01
-3.87900174e-01 -4.73320335e-02 4.32979107e-01 2.70175487e-01
6.73001468e-01 3.88087839e-01 -6.21105909e-01 2.02453006e-02
-1.24862373e+00 -1.02650285e-01 7.30715990e-01 6.83471143e-01
8.71668220e-01 -3.38662386e-01 1.86314955e-01 5.92360973e-01
-5.81951588e-02 1.52703732e-01 5.98322392e-01 -1.28484857e+00
4.57732439e-01 8.14644992e-01 -1.86834306e-01 -9.57672358e-01
-3.16719234e-01 -3.26439142e-01 -8.44631433e-01 6.65708408e-02
6.03323638e-01 -8.13630283e-01 -8.92039299e-01 1.28806043e+00
4.69569951e-01 3.87421042e-01 1.53658584e-01 7.70624280e-01
7.80679643e-01 9.52039883e-02 1.10527813e-01 3.89594346e-01
8.51833105e-01 -1.14950252e+00 -6.59321070e-01 1.70111626e-01
1.04443789e+00 -5.88178158e-01 7.34973848e-01 2.52928078e-01
-9.76370871e-01 -3.10192436e-01 -6.77746236e-01 1.21716693e-01
-5.71987033e-01 2.89290696e-02 9.86144125e-01 9.70102966e-01
-1.19161141e+00 6.06044114e-01 -1.19591653e+00 -4.50071692e-01
1.11672366e+00 3.29989552e-01 -7.22275376e-01 -2.73771107e-01
-8.33519101e-01 5.73295832e-01 9.47732329e-02 -1.42790601e-01
-8.18257451e-01 -9.73095596e-01 -6.61644459e-01 -1.58857509e-01
1.33819073e-01 -7.72543609e-01 1.26736677e+00 -1.04232502e+00
-8.24216366e-01 1.11717677e+00 5.09774506e-01 -7.66957402e-01
9.53023553e-01 6.17830781e-03 -2.89308727e-01 3.14730644e-01
2.24712312e-01 5.63759327e-01 1.63977548e-01 -1.08762074e+00
-8.61621380e-01 -4.67690408e-01 9.90520194e-02 -1.36670426e-01
-6.53897762e-01 4.76578251e-02 -5.36679149e-01 -4.44321066e-01
-3.76856923e-01 -7.62657046e-01 -8.19581151e-01 2.28238940e-01
-1.51719660e-01 2.51212865e-01 9.03391838e-01 -7.72770226e-01
7.33012021e-01 -2.15079045e+00 -7.00366199e-01 1.44687802e-01
4.62294281e-01 4.22972411e-01 -5.30415066e-02 9.65499207e-02
3.30082595e-01 4.08511728e-01 -1.29753813e-01 -3.10569882e-01
-4.58447695e-01 4.57765639e-01 1.57365471e-01 6.64697468e-01
-3.31936955e-01 9.37650204e-01 -8.30861807e-01 -1.07623947e+00
1.00026190e-01 3.36785018e-01 -7.09160984e-01 2.03078389e-01
1.23748668e-01 4.55791384e-01 -8.65107119e-01 6.31322324e-01
6.20154560e-01 -7.65734076e-01 5.35065711e-01 1.46137653e-02
1.37598515e-01 -1.45918623e-01 -7.95497179e-01 2.03467607e+00
-5.23371816e-01 4.58466142e-01 5.81256330e-01 -1.01087534e+00
5.62421322e-01 6.78894103e-01 1.08794665e+00 -5.75100362e-01
2.49723747e-01 2.70822436e-01 -1.85703978e-01 -6.67421222e-01
2.84343213e-01 -2.61528529e-02 9.51798037e-02 7.16199756e-01
4.06046808e-01 3.75918657e-01 -2.07524717e-01 3.81367415e-01
1.32268548e+00 -2.43365288e-01 -2.56342411e-01 -2.89838128e-02
1.23219937e-01 3.50733101e-01 7.71605790e-01 6.51788175e-01
-8.74657571e-01 6.53264940e-01 1.15659416e-01 -7.72679627e-01
-9.06998396e-01 -7.88773358e-01 -1.81810707e-01 9.11797464e-01
-2.60531344e-02 -1.98703289e-01 -1.12948120e+00 -1.25009644e+00
7.72595853e-02 1.25341833e-01 -6.07482135e-01 9.05442387e-02
-5.84427893e-01 -7.35308468e-01 1.00067115e+00 5.09918749e-01
9.57292676e-01 -9.04073596e-01 -9.51580524e-01 3.11906449e-03
-2.19745919e-01 -1.28915966e+00 -3.66063088e-01 -2.42259890e-01
-1.09824574e+00 -1.44357586e+00 -8.19854677e-01 -8.39094698e-01
8.13045621e-01 1.78022861e-01 9.83265817e-01 4.16231155e-01
-6.05329394e-01 6.16492450e-01 -1.52610704e-01 -5.45273840e-01
-4.48917776e-01 1.94471002e-01 -4.72089201e-01 6.05740882e-02
2.95077741e-01 -2.94913650e-02 -9.02460456e-01 2.05860391e-01
-9.40576792e-01 7.85605982e-02 6.23051941e-01 7.88652897e-01
4.75185961e-01 -8.71920213e-02 6.19411647e-01 -1.56317317e+00
4.50685799e-01 -5.94094455e-01 -4.96790886e-01 5.62709808e-01
-6.41100526e-01 -4.68111336e-01 3.93326223e-01 1.68004707e-02
-1.25751591e+00 2.50570774e-01 -1.28525253e-02 -6.25540555e-01
-5.76255739e-01 5.78232527e-01 1.26721427e-01 -2.63915718e-01
6.59419537e-01 -2.24277020e-01 3.15255404e-01 -5.51076114e-01
2.43878171e-01 8.27419639e-01 7.02843964e-01 -5.81374884e-01
4.41177636e-01 7.59367466e-01 -4.73302394e-01 -4.83515620e-01
-7.86149502e-01 -5.11903763e-01 -5.76473594e-01 -2.38711804e-01
1.07935154e+00 -1.01698208e+00 -2.07843050e-01 3.41387212e-01
-9.21676278e-01 -4.85160947e-01 -4.94316280e-01 6.90760314e-01
-3.39282632e-01 1.34851843e-01 -7.81693161e-01 -2.42931977e-01
-5.98794103e-01 -1.02878392e+00 6.11916959e-01 3.56399119e-01
9.16246027e-02 -1.36508846e+00 2.77596414e-01 9.51246798e-01
6.59983158e-01 6.57637417e-01 5.03690481e-01 -6.81457043e-01
-7.51980364e-01 -4.36441034e-01 -1.37805372e-01 3.85648042e-01
5.08765280e-01 -4.12528515e-02 -1.10281324e+00 -2.90988356e-01
1.26533896e-01 -4.89030778e-01 4.01650995e-01 5.76789618e-01
1.37459338e+00 -4.47386503e-01 -5.35602033e-01 8.95328641e-01
1.59697461e+00 -8.17291439e-02 4.86690640e-01 5.94982386e-01
5.07111609e-01 6.56928837e-01 4.23327446e-01 2.78019220e-01
3.65626812e-01 -1.46578595e-01 2.22060367e-01 -8.22112322e-01
-1.65482908e-01 -1.25682130e-01 -3.14866006e-01 6.20608628e-01
-5.44526689e-02 2.33272761e-01 -1.33246136e+00 8.55492771e-01
-1.96211469e+00 -7.99532950e-01 4.83790971e-02 1.92219496e+00
8.87973785e-01 -3.24125677e-01 -2.68632881e-02 -6.55442178e-01
8.19788337e-01 -2.85578221e-01 -5.64899981e-01 -1.65203154e-01
-1.36628091e-01 2.56916821e-01 9.84717667e-01 -8.25043246e-02
-1.16224241e+00 7.79288113e-01 7.45750952e+00 4.20857877e-01
-1.34378707e+00 6.23432040e-01 1.21836686e+00 -2.97056437e-01
-5.03309220e-02 -3.30931276e-01 -1.81869209e-01 2.55170554e-01
1.18422794e+00 -3.89704466e-01 -1.16307475e-01 1.10960329e+00
3.92278768e-02 2.35100552e-01 -8.53255093e-01 8.70580196e-01
-2.78848976e-01 -2.10588026e+00 -4.18477356e-02 5.10699093e-01
1.19843972e+00 5.72347939e-01 4.01557609e-03 8.78581963e-03
8.80917251e-01 -1.13980639e+00 1.63715780e-01 3.56928438e-01
7.96110451e-01 -6.27267838e-01 8.91258538e-01 3.71478766e-01
-6.62472725e-01 -2.43356377e-02 -2.04969034e-01 2.31856823e-01
-8.24680626e-02 5.62487900e-01 -8.43246281e-01 6.97421610e-01
1.00913453e+00 3.92839700e-01 -6.33382320e-01 1.29719937e+00
2.47282147e-01 8.99912655e-01 -1.06204569e-01 5.55819154e-01
5.08740723e-01 -2.29831919e-01 -1.49879456e-01 1.14486635e+00
-7.23550022e-02 -7.04141110e-02 3.13830107e-01 7.24073052e-01
-6.37518406e-01 3.19694966e-01 -8.17318320e-01 6.27120882e-02
1.58427328e-01 1.58360207e+00 -7.27545142e-01 -3.12677741e-01
-9.52632248e-01 6.68249249e-01 2.30049819e-01 2.28130758e-01
-7.31656373e-01 -2.50876069e-01 6.95277810e-01 9.14352387e-02
4.55357879e-02 -9.87905785e-02 -3.97348344e-01 -1.13034439e+00
-2.35052675e-01 -1.06167996e+00 9.28654969e-01 -2.19701260e-01
-1.56367648e+00 5.71792901e-01 -3.36987048e-01 -8.62447143e-01
-1.00937515e-01 -3.92372578e-01 -7.88142622e-01 7.18820035e-01
-1.45705366e+00 -1.49876225e+00 -3.48361433e-01 1.16020286e+00
1.41014785e-01 -3.39426994e-01 9.17878449e-01 6.32853389e-01
-7.70168006e-01 7.33811677e-01 2.59300321e-01 1.02942181e+00
8.08184981e-01 -8.45106304e-01 8.97339955e-02 8.23296309e-01
1.21252000e-01 7.27322340e-01 2.43751444e-02 -7.29313433e-01
-1.20141792e+00 -1.46829104e+00 5.44098556e-01 -2.14190841e-01
4.40711170e-01 -1.15292743e-01 -7.83065319e-01 1.07603705e+00
5.48883557e-01 5.50069332e-01 1.32411432e+00 2.69592702e-01
-1.08937919e-01 -1.28493160e-01 -1.69896853e+00 1.90253630e-01
6.43010974e-01 -4.40216899e-01 -1.94126517e-01 6.06582284e-01
6.63891077e-01 -2.38871515e-01 -1.17126739e+00 1.43658996e-01
2.90951759e-01 -9.01895821e-01 6.39427662e-01 -9.23678815e-01
3.85500848e-01 2.47723773e-01 -1.16715208e-02 -9.74841177e-01
-3.31020541e-03 -4.40557808e-01 4.10861194e-01 1.10353160e+00
4.28095162e-01 -8.78912747e-01 1.61209178e+00 1.21151984e+00
-4.73491922e-02 -7.03560770e-01 -9.37107205e-01 -7.02058613e-01
5.19312024e-01 -2.78200924e-01 9.77607727e-01 1.50243866e+00
-2.20682591e-01 -3.86384249e-01 1.06894761e-01 1.93276972e-01
7.25263715e-01 2.64388889e-01 8.81261408e-01 -1.00962532e+00
-2.37798303e-01 -2.09127650e-01 -3.58785927e-01 -2.07689628e-01
1.03408672e-01 -9.46237028e-01 -2.96589732e-01 -1.72658968e+00
2.75837392e-01 -1.01879299e+00 -5.26651621e-01 7.92742312e-01
1.00804225e-01 1.91135034e-01 1.35942340e-01 5.24227381e-01
-6.51384890e-01 3.33829969e-02 1.27894163e+00 -1.95459083e-01
5.05340770e-02 -1.76051646e-01 -9.17515934e-01 8.44521582e-01
9.72084284e-01 -4.15792286e-01 -2.35409454e-01 -9.80207920e-01
-2.70938247e-01 -9.24025197e-03 3.23756725e-01 -1.01648259e+00
5.93332410e-01 -2.68270612e-01 5.59195459e-01 -2.28300408e-01
-2.88524002e-01 -9.64187860e-01 2.24872395e-01 5.83307505e-01
-2.70884752e-01 -9.61647481e-02 1.32860467e-01 3.39874655e-01
-4.00698245e-01 -1.33508429e-01 6.40092075e-01 -5.77016294e-01
-4.98138428e-01 6.42096341e-01 -1.81463528e-02 5.04226461e-02
1.24474049e+00 -2.31273621e-02 -5.29715478e-01 -2.75395066e-01
-5.94506860e-01 3.80948633e-01 7.63655186e-01 -4.07635123e-02
4.66076732e-01 -9.90011334e-01 -7.38855004e-01 1.32578433e-01
-2.12271050e-01 2.91759819e-01 4.62572783e-01 7.22922623e-01
-9.33231652e-01 3.94603580e-01 -3.56395990e-01 -4.73401010e-01
-1.16418445e+00 4.17830527e-01 6.73896432e-01 -4.43370819e-01
-9.15944457e-01 6.15664005e-01 1.98950600e-02 -9.78422701e-01
2.66227633e-01 -7.64143048e-03 3.58933032e-01 -3.44587862e-01
3.85975480e-01 4.28687990e-01 3.54773521e-01 -3.21700871e-01
-1.54623955e-01 -7.70630389e-02 -1.51782632e-01 -2.00971756e-02
1.61846268e+00 1.42158270e-01 -1.32184697e-03 -1.58062264e-01
1.29483867e+00 -2.78246105e-01 -1.18939829e+00 -1.23769380e-01
-1.80596977e-01 -6.63532972e-01 2.85972059e-01 -7.86758184e-01
-1.83622241e+00 5.70291758e-01 7.43607402e-01 -4.88658734e-02
9.63042736e-01 -2.41813976e-02 1.09434187e+00 1.50284484e-01
6.77368283e-01 -1.24984181e+00 -3.14167738e-01 -1.50346473e-01
2.46378839e-01 -1.33702254e+00 -4.00032513e-02 -1.81577325e-01
-5.64902008e-01 8.66634250e-01 6.79458499e-01 -6.83056638e-02
9.23155963e-01 3.96529049e-01 7.03909755e-01 -2.45950833e-01
-4.93128717e-01 2.26338774e-01 -2.12845445e-01 6.04470313e-01
3.97455156e-01 5.40098064e-02 -4.16621529e-02 5.85375369e-01
-2.89270163e-01 5.08438826e-01 5.25965214e-01 1.31495750e+00
5.71465269e-02 -1.25857925e+00 -4.42229897e-01 7.35244811e-01
-9.61925507e-01 2.55377948e-01 -2.83319980e-01 7.89905488e-01
4.53414917e-01 9.52727795e-01 1.70527548e-01 -2.52780225e-03
4.78687435e-02 -8.46592560e-02 2.62921602e-01 -5.64471483e-01
-1.00847495e+00 -1.52514338e-01 -1.68614864e-01 -6.62003994e-01
-5.31445265e-01 -5.23109853e-01 -1.42249227e+00 -4.64135289e-01
-1.12108104e-01 2.66297638e-01 9.87502813e-01 6.80572629e-01
7.16281772e-01 3.48441958e-01 4.39961761e-01 -1.56725377e-01
-4.61470664e-01 -3.56507033e-01 -5.18099725e-01 6.73105776e-01
2.07220525e-01 2.97730397e-02 6.54280186e-02 2.37893239e-01] | [6.118232250213623, 6.4580488204956055] |
b973865b-3d0e-44ef-a15f-6b7e5d6581c8 | multi-microphone-automatic-speech | 2306.04268 | null | https://arxiv.org/abs/2306.04268v1 | https://arxiv.org/pdf/2306.04268v1.pdf | Multi-microphone Automatic Speech Segmentation in Meetings Based on Circular Harmonics Features | Speaker diarization is the task of answering Who spoke and when? in an audio stream. Pipeline systems rely on speech segmentation to extract speakers' segments and achieve robust speaker diarization. This paper proposes a common framework to solve three segmentation tasks in the distant speech scenario: Voice Activity Detection (VAD), Overlapped Speech Detection (OSD), and Speaker Change Detection (SCD). In the literature, a few studies investigate the multi-microphone distant speech scenario. In this work, we propose a new set of spatial features based on direction-of-arrival estimations in the circular harmonic domain (CH-DOA). These spatial features are extracted from multi-microphone audio data and combined with standard acoustic features. Experiments on the AMI meeting corpus show that CH-DOA can improve the segmentation while being robust in the case of deactivated microphones. | ['Jean-Hugh Thomas', 'Silvio Montrésor', 'Anthony Larcher', 'Théo Mariotte'] | 2023-06-07 | null | null | null | null | ['action-detection', 'change-detection', 'activity-detection', 'speaker-diarization'] | ['computer-vision', 'computer-vision', 'computer-vision', 'speech'] | [ 6.84819147e-02 -2.39141136e-01 2.41444290e-01 -3.87709022e-01
-1.47765625e+00 -7.94834137e-01 5.97352505e-01 2.89570123e-01
-2.96948761e-01 1.66419998e-01 5.34491241e-01 -2.10345998e-01
-8.67807940e-02 -2.63263971e-01 -1.18071727e-01 -8.32639277e-01
-1.27347559e-01 1.02519006e-01 4.75586981e-01 -8.06618109e-03
2.51886338e-01 6.87949121e-01 -1.90646219e+00 3.34306568e-01
6.14520073e-01 8.77625883e-01 2.49755129e-01 1.15626574e+00
-1.66406482e-01 1.86625659e-01 -1.16920125e+00 2.31767520e-01
5.96492924e-03 -4.95370477e-01 -5.98587334e-01 8.08326975e-02
3.27739865e-01 -4.87613417e-02 2.27323011e-01 8.20735514e-01
1.16984665e+00 1.88951060e-01 5.23167372e-01 -1.09502029e+00
4.34448242e-01 6.53694749e-01 -3.01530808e-01 9.03045297e-01
7.20996678e-01 -2.65438259e-01 7.53324509e-01 -1.12175238e+00
1.62561432e-01 1.26318324e+00 7.22086847e-01 3.37770671e-01
-8.66493046e-01 -4.39324826e-01 -3.85302156e-02 1.70225769e-01
-1.80506027e+00 -1.02737951e+00 9.54705000e-01 -5.79468012e-01
1.04525948e+00 7.86848783e-01 2.01375395e-01 7.18298972e-01
-3.37445945e-01 8.19978356e-01 6.41947806e-01 -6.58931375e-01
4.64036554e-01 8.70124437e-03 2.50896215e-01 7.34465122e-02
-4.06985968e-01 -8.73029977e-02 -8.26912105e-01 -1.45853773e-01
2.90507168e-01 -5.00642061e-01 -3.98016244e-01 2.10106656e-01
-1.23158443e+00 6.93127990e-01 -2.79615521e-01 8.65181148e-01
-3.88120919e-01 -3.81453812e-01 3.06957901e-01 1.84054479e-01
7.18895614e-01 1.41355485e-01 -3.06660354e-01 -4.07704145e-01
-1.54366100e+00 2.19821036e-01 7.71370113e-01 6.60965741e-01
4.17064220e-01 1.50590196e-01 -2.23497570e-01 1.02965832e+00
5.15316546e-01 6.98659480e-01 5.71711183e-01 -7.93965936e-01
4.35121834e-01 -1.57998368e-01 3.01417224e-02 -7.00015485e-01
-5.22577226e-01 -4.61053550e-01 -3.86760026e-01 -6.91313744e-02
4.60657477e-01 -4.39266294e-01 -6.00505829e-01 1.34946632e+00
7.18963265e-01 3.06563228e-01 3.71297100e-03 7.43422091e-01
8.75221670e-01 8.34409535e-01 -4.26273823e-01 -5.74846447e-01
1.29979205e+00 -7.68774390e-01 -1.14679766e+00 -1.14553168e-01
1.74276009e-01 -1.14289105e+00 6.06761038e-01 6.53164446e-01
-1.15178311e+00 -8.91152382e-01 -9.13630068e-01 3.41350049e-01
-3.33829671e-01 -7.15875551e-02 -1.98201120e-01 1.13359308e+00
-1.15084302e+00 -1.29346028e-01 -8.39098334e-01 -2.11855665e-01
-1.72662929e-01 2.89558321e-01 -3.90837789e-02 3.61533374e-01
-1.11036646e+00 3.02982628e-01 -1.67261109e-01 6.41706511e-02
-1.04846776e+00 -5.25799811e-01 -7.66337812e-01 5.05058048e-03
3.83968204e-02 -1.69157729e-01 1.54358602e+00 -6.73370421e-01
-1.81471741e+00 6.69347107e-01 -7.94850707e-01 -4.69589531e-01
3.36349934e-01 -1.31440416e-01 -9.06532288e-01 3.58235002e-01
8.55451748e-02 3.40929061e-01 1.32759547e+00 -1.01280868e+00
-9.58115995e-01 -5.79937160e-01 -5.87630093e-01 3.50443184e-01
-2.16195196e-01 5.07337272e-01 -3.87726575e-01 -6.79538786e-01
4.10075873e-01 -6.28288448e-01 2.61295825e-01 -7.58407593e-01
-7.14562833e-01 -4.05298233e-01 7.92333961e-01 -1.09491336e+00
1.59042239e+00 -2.58495998e+00 4.19643242e-03 2.76148736e-01
-1.03822537e-01 2.82208472e-01 1.12936281e-01 1.41671002e-01
8.32390785e-02 -1.70074522e-01 -2.11845651e-01 -5.94212830e-01
1.52915388e-01 -3.07983220e-01 -1.62760407e-01 4.14220899e-01
-3.73975858e-02 2.20632315e-01 -5.40872991e-01 -6.44576848e-01
3.14346910e-01 6.42658830e-01 -3.98881555e-01 3.90210062e-01
4.02190953e-01 5.97707987e-01 2.29463000e-02 5.29126525e-01
8.79776180e-01 7.94589639e-01 -2.27568060e-01 1.33778870e-01
-6.45280302e-01 5.32616615e-01 -1.74634743e+00 1.53350937e+00
-4.92480010e-01 8.15446675e-01 8.63357008e-01 -7.84687042e-01
1.06924486e+00 7.99673975e-01 3.69143516e-01 -3.83897275e-01
-2.08000299e-02 4.64021295e-01 -4.38110791e-02 -6.56449080e-01
3.72256726e-01 -1.86848901e-02 -1.03615828e-01 -1.82158276e-02
2.69905999e-02 -3.83891702e-01 9.68414620e-02 -2.69598782e-01
9.13973093e-01 -6.15982771e-01 1.32458925e-01 -3.19942385e-01
9.54004109e-01 -4.25518483e-01 4.25144017e-01 6.01185799e-01
-5.80709636e-01 8.94798279e-01 -4.31502908e-02 3.20330262e-01
-4.34500098e-01 -1.25944936e+00 -1.93567738e-01 1.29955137e+00
-2.52374768e-01 -2.87235081e-01 -1.32062948e+00 -3.16525817e-01
-2.90699989e-01 7.86629438e-01 -2.34882850e-02 3.07336450e-01
-8.09412003e-01 -4.07666385e-01 9.06742811e-01 3.06329995e-01
2.50885785e-01 -7.29361594e-01 -4.35637981e-01 4.11750615e-01
-4.80334461e-01 -1.00408411e+00 -8.14305902e-01 2.14868084e-01
-5.20116746e-01 -4.83199954e-01 -1.09157932e+00 -1.07906902e+00
-5.34358062e-03 3.94297153e-01 6.89478636e-01 -6.75259650e-01
-1.01572864e-01 6.85559332e-01 -3.08269322e-01 -6.21972203e-01
-6.30396128e-01 9.00200605e-02 1.33399144e-01 4.81634408e-01
4.93203670e-01 -5.28258443e-01 -4.53844070e-01 5.84343910e-01
-6.88671291e-01 -6.36522532e-01 1.55703962e-01 1.43145174e-01
4.92294759e-01 3.45083863e-01 9.06026304e-01 -2.03087106e-01
7.83592582e-01 -4.06329840e-01 -3.15035790e-01 -1.45463288e-01
-1.48118824e-01 -4.87028480e-01 2.33650252e-01 -3.08881491e-01
-1.31019199e+00 1.62979856e-01 -6.93416655e-01 -1.72883049e-01
-8.39082479e-01 2.83301353e-01 -7.13358700e-01 4.79222029e-01
6.57164991e-01 2.67967045e-01 -2.67627627e-01 -8.11859190e-01
6.13459311e-02 1.43945014e+00 6.74529135e-01 6.38339892e-02
4.51866150e-01 3.79695624e-01 -6.28572583e-01 -1.55947351e+00
-2.18891367e-01 -1.28694499e+00 -7.54759729e-01 -1.47910610e-01
1.05634356e+00 -1.08976007e+00 -2.74300158e-01 6.66872442e-01
-1.17503250e+00 -9.31360051e-02 -2.50784755e-01 8.59729826e-01
-3.51138562e-01 4.10759836e-01 -3.18424255e-01 -1.15290618e+00
-2.46362358e-01 -1.12337327e+00 1.33733761e+00 1.36917874e-01
-3.14587206e-01 -8.89070332e-01 2.87678033e-01 4.98711318e-01
4.80255127e-01 -1.27883241e-01 2.17361793e-01 -9.00583446e-01
-3.68656963e-02 -8.65594968e-02 6.35956824e-01 4.26590919e-01
3.31590056e-01 -1.06571198e-01 -1.56955266e+00 -1.40166283e-01
3.20945740e-01 6.20189428e-01 6.42258704e-01 9.69744444e-01
6.83662832e-01 -1.28616869e-01 -2.65117675e-01 2.93340385e-01
6.24465227e-01 5.44660389e-01 3.60968322e-01 -6.45074546e-02
3.45621377e-01 6.73670590e-01 7.48054504e-01 5.40950477e-01
2.34054610e-01 8.99095178e-01 1.79246828e-01 4.20432650e-02
-4.52529311e-01 1.22584864e-01 7.31795907e-01 1.28660405e+00
4.33359593e-01 -4.08854395e-01 -1.08294809e+00 9.81894553e-01
-1.26983654e+00 -9.45635736e-01 -3.94629031e-01 2.10745239e+00
7.51992047e-01 1.02246739e-01 5.44831038e-01 8.95174205e-01
1.30379105e+00 1.95579067e-01 -1.16683483e-01 -5.46945572e-01
-2.45556608e-01 2.01832116e-01 8.35172832e-02 1.04913080e+00
-1.21433747e+00 4.54749882e-01 6.08184528e+00 8.46536279e-01
-1.20120096e+00 4.53505307e-01 1.62609994e-01 -1.18901014e-01
-2.01312631e-01 -5.93517482e-01 -1.14110839e+00 4.19140637e-01
1.36457229e+00 1.41738802e-01 1.48077741e-01 4.98252958e-01
6.80471361e-01 -1.37666687e-01 -1.09164774e+00 1.27977037e+00
4.06124979e-01 -7.63791800e-01 -4.98283893e-01 -8.00465569e-02
3.87838364e-01 1.05109975e-01 1.64639279e-01 6.54846579e-02
-3.62464458e-01 -6.17244065e-01 9.79381382e-01 3.32861304e-01
5.50044417e-01 -7.32793391e-01 5.55040538e-01 4.92784679e-01
-1.44907820e+00 -5.79314977e-02 2.76162773e-01 3.14714640e-01
3.70325178e-01 8.16872835e-01 -1.51764405e+00 4.36572850e-01
8.10801446e-01 2.18252957e-01 -3.62405896e-01 1.31318045e+00
-3.18795033e-02 9.58742738e-01 -4.91480321e-01 2.35985905e-01
-1.68602169e-02 3.45290750e-01 1.14554167e+00 1.69124985e+00
5.85676730e-01 -2.66475439e-01 -1.72031194e-03 3.67518783e-01
3.52395803e-01 2.08543882e-01 -2.42011070e-01 3.05427641e-01
8.12291741e-01 1.02139175e+00 -6.62063360e-01 -8.30562711e-02
-1.58664986e-01 9.19122934e-01 -7.10592747e-01 4.15860355e-01
-7.43202031e-01 -7.03643560e-01 6.20122731e-01 1.20669633e-01
5.47313452e-01 -3.11797678e-01 -1.14680439e-01 -6.35594547e-01
4.73352708e-02 -8.18317473e-01 3.74036700e-01 -4.15158004e-01
-8.34819794e-01 5.87864518e-01 -8.88269618e-02 -1.15287089e+00
-6.11532032e-01 -1.69496015e-01 -8.20897341e-01 9.45348740e-01
-1.54394460e+00 -6.73935890e-01 -1.09748252e-01 9.58746314e-01
1.19835639e+00 -1.60139292e-01 7.64206469e-01 7.51433432e-01
-4.27296191e-01 7.07647085e-01 1.61902234e-01 4.70175184e-02
7.53893614e-01 -1.34512877e+00 3.74563307e-01 8.49740446e-01
2.92451739e-01 2.19111964e-01 8.92598093e-01 -2.90671229e-01
-9.65221763e-01 -9.69769537e-01 1.23903310e+00 -2.83788711e-01
1.51441500e-01 -3.81963640e-01 -7.35265374e-01 3.92902374e-01
7.33116791e-02 -1.07123308e-01 9.83477533e-01 -1.19719334e-01
1.20943792e-01 -2.69069880e-01 -1.08495927e+00 -8.04517642e-02
5.59332907e-01 -7.70233333e-01 -8.00586641e-01 -1.90643482e-02
4.80924428e-01 -1.42697871e-01 -7.60623515e-01 -2.83938646e-02
2.14375243e-01 -9.28253114e-01 9.89717841e-01 2.44590908e-01
-6.64706290e-01 -6.60199702e-01 -4.01861548e-01 -1.33052492e+00
2.24781651e-02 -9.90864158e-01 2.52548233e-02 1.77609408e+00
4.30971146e-01 -4.22448337e-01 2.95302361e-01 -1.26286566e-01
-4.92002964e-01 2.11414203e-01 -1.47085369e+00 -8.16109061e-01
-1.28089532e-01 -7.70305812e-01 7.26701915e-01 7.92779565e-01
1.58894360e-01 4.35662061e-01 8.21101889e-02 6.69058979e-01
2.82880813e-01 -2.30114043e-01 5.64460158e-01 -1.41431105e+00
-1.86778024e-01 -4.67736512e-01 -3.02510679e-01 -1.08970308e+00
-1.04087107e-02 -4.83197600e-01 4.75159466e-01 -1.43699622e+00
-6.41467988e-01 -1.25992432e-01 -1.22528702e-01 -1.97065651e-01
-2.75097433e-02 -1.39412731e-01 -7.59875476e-02 2.10181996e-02
-4.81059521e-01 1.83898106e-01 6.02230072e-01 -2.93309897e-01
-6.90082848e-01 6.44629478e-01 -7.98007846e-02 8.94983828e-01
8.41564894e-01 -4.28093135e-01 -2.60673553e-01 -3.16278636e-01
-3.73148233e-01 3.12162101e-01 7.10674226e-02 -1.21833479e+00
4.68227983e-01 3.57556373e-01 -3.38653624e-02 -1.16245067e+00
6.06962740e-01 -6.69178009e-01 -1.16664879e-01 2.36255348e-01
-3.19306225e-01 -9.96079519e-02 2.51837492e-01 4.01412189e-01
-6.94674730e-01 -2.19277382e-01 7.23304510e-01 1.91173032e-01
-5.28459609e-01 -2.35913530e-01 -9.80226934e-01 -7.35099092e-02
8.42340827e-01 -2.42809862e-01 1.83330536e-01 -4.99575973e-01
-1.04731131e+00 -1.91978231e-01 -2.85221845e-01 4.45917815e-01
5.73649168e-01 -1.01647961e+00 -8.51288557e-01 4.60436314e-01
-1.80615515e-01 -3.60253174e-03 4.19234544e-01 9.93112206e-01
-2.37740666e-01 6.26282692e-01 2.91395158e-01 -9.66445863e-01
-1.63672292e+00 2.95317739e-01 5.57467103e-01 2.00993046e-01
-2.77485028e-02 1.20626616e+00 6.14109375e-02 -2.30173260e-01
7.52427518e-01 -4.88102049e-01 -5.91527760e-01 5.93930423e-01
8.80572736e-01 8.42291892e-01 5.69069445e-01 -1.07961714e+00
-6.83034122e-01 4.02080566e-01 3.42046708e-01 -7.39901125e-01
9.96571302e-01 -8.04783225e-01 1.65456399e-01 7.64955878e-01
1.28335416e+00 7.45651603e-01 -1.04852545e+00 -1.70246229e-01
2.43846029e-01 -2.48228595e-01 3.42118412e-01 -6.39055312e-01
-6.74021006e-01 1.32238460e+00 1.05879319e+00 8.24551344e-01
1.11181569e+00 1.38100728e-01 7.14961588e-01 -2.05120519e-02
-1.56183705e-01 -1.32696402e+00 -3.39802317e-02 4.97955233e-01
8.25206041e-01 -8.62367034e-01 -7.83759117e-01 -4.20098603e-01
-4.36986178e-01 1.12342525e+00 1.42932892e-01 5.12550890e-01
1.01337063e+00 3.55332851e-01 4.37726200e-01 5.43359146e-02
-9.28899348e-02 -5.20538747e-01 2.65705585e-01 8.00830126e-01
4.54161286e-01 1.46392554e-01 1.65356338e-01 6.56650066e-01
-4.81724828e-01 -7.73205400e-01 2.17404708e-01 8.60918224e-01
-9.18248773e-01 -7.80524075e-01 -1.10080028e+00 -2.01712355e-01
-6.65092945e-01 3.14493142e-02 -4.83019739e-01 1.29666999e-01
2.66366392e-01 1.91150093e+00 2.76155919e-01 -1.82508379e-01
5.36525488e-01 5.68819106e-01 3.07543632e-02 -6.04855299e-01
-7.36866355e-01 9.18112397e-01 1.38205349e-01 -1.27375007e-01
-5.50737858e-01 -1.20205963e+00 -1.35412025e+00 2.17473254e-01
-4.97632027e-01 5.19487441e-01 1.08442414e+00 8.11789155e-01
3.08310151e-01 9.20040131e-01 1.01756012e+00 -8.87472332e-01
-3.63573432e-01 -1.08507216e+00 -8.53955984e-01 1.07218139e-01
9.00851011e-01 -2.17198357e-01 -7.37720490e-01 3.20683718e-01] | [14.797289848327637, 5.873857021331787] |
1e7770f2-aa7d-41e6-9388-97133f2011e3 | towards-resilient-and-secure-smart-grids | null | null | https://www.mdpi.com/2079-9292/12/12/2554 | https://www.mdpi.com/2079-9292/12/12/2554 | Towards Resilient and Secure Smart Grids against PMU Adversarial Attacks: A Deep Learning-Based Robust Data Engineering Approach | In an attempt to provide reliable power distribution, smart grids integrate monitoring, communication, and control technologies for better energy consumption and management. As a result of such cyberphysical links, smart grids become vulnerable to cyberattacks, highlighting the significance of detecting and monitoring such attacks to uphold their security and dependability. Accordingly, the use of phasor measurement units (PMUs) enables real-time monitoring and control, providing informed-decisions data and making it possible to sense abnormal behavior indicative of cyberattacks. Similar to the ways it dominates other fields, deep learning has brought a lot of interest to the realm of cybersecurity. A common formulation for this issue is learning under data complexity, unavailability, and drift connected to increasing cardinality, imbalance brought on by data scarcity, and fast change in data characteristics, respectively. To address these challenges, this paper suggests a deep learning monitoring method based on robust feature engineering, using PMU data with greater accuracy, even within the presence of cyberattacks. The model is initially investigated using condition monitoring data to identify various disturbances in smart grids free from adversarial attacks. Then, a minimally disruptive experiment using adversarial attack injection with various reality-imitating techniques is conducted, inadvertently damaging the original data and using it to retrain the deep network, boosting its resistance to manipulations. Compared to previous studies, the proposed method demonstrated promising results and better accuracy, making it a potential option for smart grid condition monitoring. The full set of experimental scenarios performed in this study is available online. | ['Yassine Amirat', 'Mohamed Benbouzid', 'Tarek Berghout'] | 2023-06-06 | null | null | null | mdpi-electronics-2023-6 | ['adversarial-attack', 'color-image-denoising', 'feature-engineering'] | ['adversarial', 'computer-vision', 'methodology'] | [-1.06638238e-01 -3.10515046e-01 6.30387738e-02 1.22326396e-01
-2.15578854e-01 -8.26070309e-01 4.71445918e-01 3.94031197e-01
8.91183391e-02 8.67763162e-01 -2.63967872e-01 -2.81550556e-01
-2.31239393e-01 -1.03819919e+00 -4.97672290e-01 -1.14449692e+00
-8.71073663e-01 2.51085609e-02 -3.17891359e-01 -6.83131590e-02
1.33851707e-01 8.59851360e-01 -1.02154112e+00 -2.02990308e-01
8.42948139e-01 1.21128345e+00 -2.88228691e-01 1.92663237e-01
5.38496435e-01 4.49042469e-01 -1.26376927e+00 2.05553651e-01
3.44668895e-01 -1.19781934e-01 -4.28958595e-01 2.86805034e-02
-5.22650480e-01 -6.12039387e-01 -4.32149857e-01 1.36931622e+00
7.49844491e-01 4.02999409e-02 9.77822393e-02 -1.62457883e+00
-4.23891246e-01 6.49473846e-01 -4.23130304e-01 4.28774029e-01
5.20869613e-01 5.59669554e-01 6.01936698e-01 -1.47422552e-01
-8.36126655e-02 7.10148036e-01 4.45898324e-01 8.91343504e-02
-1.17697585e+00 -8.58179271e-01 1.92105602e-02 6.31391943e-01
-1.12374640e+00 3.32895033e-02 1.08734620e+00 -1.50394797e-01
9.66920912e-01 2.05053583e-01 9.64578450e-01 1.23918259e+00
6.11179173e-01 6.02129877e-01 1.16871285e+00 -1.52510718e-01
6.96719646e-01 8.43912829e-03 -1.37964068e-02 -9.68641117e-02
4.20544207e-01 6.81029856e-01 2.08891323e-03 -2.44123653e-01
2.42489338e-01 4.85767834e-02 -7.22606599e-01 -1.48155943e-01
-9.42704439e-01 5.41570961e-01 5.74358642e-01 6.02461815e-01
-5.81615388e-01 -1.56547889e-01 7.60407090e-01 5.90358377e-01
8.71395692e-02 5.91762006e-01 -6.78576112e-01 -2.79975533e-01
-4.44640368e-01 -3.97481203e-01 8.35392654e-01 5.25480449e-01
2.75359541e-01 1.07661593e+00 3.86694223e-01 2.87537202e-02
4.46575917e-02 5.94043791e-01 8.10280263e-01 -2.19727054e-01
1.92674026e-01 6.58852935e-01 5.10771982e-02 -1.21890378e+00
-7.67903805e-01 -8.31301391e-01 -1.26397681e+00 5.06875575e-01
1.29578769e-01 -4.17052537e-01 -3.52803528e-01 1.47959995e+00
4.78043348e-01 3.64854574e-01 1.11845464e-01 7.06873298e-01
-7.35861957e-02 6.80685401e-01 -1.63049132e-01 -5.05521357e-01
1.14357960e+00 -4.98467078e-03 -9.27446187e-01 2.84576476e-01
4.85900491e-01 -3.20636481e-01 7.17065334e-01 9.16493356e-01
-3.87959421e-01 -4.70863223e-01 -1.63392007e+00 1.01393509e+00
-7.37441301e-01 -5.00667214e-01 5.56945503e-01 8.75506341e-01
-6.78230643e-01 9.11026120e-01 -8.91033888e-01 -5.29491454e-02
3.48641396e-01 4.24764723e-01 -2.84920752e-01 3.07581991e-01
-1.73211861e+00 1.12296867e+00 7.31762767e-01 4.21991944e-01
-1.08892977e+00 -6.21092498e-01 -6.96644962e-01 1.65492073e-01
2.24293232e-01 5.85696427e-03 6.36854589e-01 -7.80960917e-01
-1.41197526e+00 -5.61250821e-02 8.36944282e-01 -9.29553628e-01
4.48844850e-01 -1.05946042e-01 -1.20253420e+00 1.18624069e-01
-4.08136100e-01 -5.17768145e-01 1.05989170e+00 -8.38638842e-01
-3.16378683e-01 -4.85701352e-01 -6.52035326e-02 -2.82749414e-01
-5.25393069e-01 -2.25304604e-01 8.25103939e-01 -4.76499379e-01
-1.25722721e-01 -3.99145216e-01 -1.07361838e-01 -6.48636281e-01
-5.41643739e-01 -5.82506359e-02 1.45622790e+00 -7.86111116e-01
1.04372382e+00 -2.04283547e+00 -2.55890250e-01 5.64821661e-01
7.13506043e-02 6.75601184e-01 1.98376179e-01 8.62484276e-01
-5.40770590e-01 -2.14211401e-02 -2.16409534e-01 3.07649285e-01
1.96251884e-01 2.72984803e-01 -5.12890697e-01 1.01620400e+00
2.15957344e-01 6.68055534e-01 -8.09783280e-01 3.94632757e-01
7.31974423e-01 3.61766785e-01 1.30046368e-01 2.74478376e-01
-8.37296173e-02 7.02526331e-01 -5.04217982e-01 7.24449337e-01
7.74178565e-01 5.80705330e-02 1.78204715e-01 -3.47616941e-01
7.31885284e-02 -8.58216509e-02 -1.41000164e+00 8.29397202e-01
-4.34949398e-01 4.60869044e-01 8.67385119e-02 -1.65095532e+00
9.21478391e-01 5.73178113e-01 7.78717935e-01 -1.02950895e+00
5.39506495e-01 2.43266928e-04 1.12158582e-01 -4.33093578e-01
-7.84614906e-02 2.50712335e-01 -1.31409943e-01 6.48413360e-01
-1.41195944e-02 -3.40789221e-02 -7.59695424e-03 -3.55580151e-02
1.15033925e+00 -1.66545153e-01 6.26758397e-01 -2.56942421e-01
6.74672604e-01 -4.58912134e-01 6.98178232e-01 3.69419903e-01
-3.31851840e-01 -3.30148906e-01 3.77031624e-01 -6.63909256e-01
-6.05292499e-01 -1.02959156e+00 -3.65214080e-01 2.41669603e-02
-3.16164792e-02 3.05977766e-03 -3.90584916e-01 -7.31178820e-01
2.12448433e-01 1.07692993e+00 -4.92520660e-01 -6.88116491e-01
-5.20819247e-01 -1.22931743e+00 4.69281465e-01 3.90655756e-01
6.95218146e-01 -1.06834531e+00 -7.96829760e-01 5.78893602e-01
3.24769586e-01 -8.96767139e-01 4.02517989e-02 5.42138219e-01
-5.10921538e-01 -1.41576946e+00 -6.08730204e-02 -1.09292805e-01
5.60978889e-01 -1.73570439e-01 7.55210638e-01 1.77982703e-01
-3.81259263e-01 2.51948595e-01 -3.88379216e-01 -2.97122180e-01
-8.34270954e-01 -3.15479130e-01 6.76064372e-01 6.72982559e-02
2.20730349e-01 -1.03550494e+00 -5.04620135e-01 1.97367385e-01
-1.11925268e+00 -6.45246923e-01 3.71199399e-01 8.45800638e-01
7.54196569e-02 8.67011786e-01 1.18747008e+00 -3.05111468e-01
6.58175707e-01 -7.41136312e-01 -1.19373298e+00 -4.61267307e-02
-1.05457103e+00 -4.17037278e-01 1.45160067e+00 -5.95461190e-01
-4.62853491e-01 -4.04113978e-01 -7.61573911e-02 -2.50018150e-01
-4.56158340e-01 4.42884684e-01 -7.47563958e-01 -3.73302072e-01
3.02645475e-01 3.98482084e-01 1.94500566e-01 -3.10995340e-01
1.75761387e-01 7.59919882e-01 2.58927554e-01 -9.51231718e-02
1.28040063e+00 2.46632516e-01 1.31684542e-01 -8.58284831e-01
-3.05850714e-01 1.27854839e-01 -5.73797762e-01 -4.59879279e-01
3.51008654e-01 -5.86581707e-01 -1.16547143e+00 8.81965101e-01
-7.14167237e-01 4.92670760e-03 -3.04504782e-01 3.78796518e-01
-7.33039752e-02 6.70404196e-01 -6.25567198e-01 -7.09178627e-01
-4.93843555e-01 -1.04785478e+00 1.59506172e-01 2.96060562e-01
6.13625580e-03 -1.28628886e+00 -1.07086161e-02 2.93981787e-02
6.83061421e-01 8.13548446e-01 8.70227337e-01 -1.14437461e+00
-5.31823158e-01 -7.82878757e-01 2.94438809e-01 8.98394227e-01
5.93152463e-01 -1.33761376e-01 -8.41624498e-01 -1.05821812e+00
5.88572025e-01 -1.87552214e-01 -2.92101651e-01 -5.32643124e-02
8.87790680e-01 -7.36120522e-01 -2.64377650e-02 5.95153987e-01
1.45695961e+00 6.75706744e-01 5.81638515e-01 5.95334589e-01
3.47840697e-01 9.96619165e-02 3.32281739e-01 7.95507669e-01
2.90612653e-02 3.53785396e-01 1.07376480e+00 -2.28383854e-01
6.39785349e-01 1.34525359e-01 4.18265790e-01 7.90333629e-01
4.69208360e-01 -1.85680151e-01 -5.93894303e-01 2.48409256e-01
-1.27290988e+00 -9.32100415e-01 2.53229946e-01 2.25556397e+00
3.80528212e-01 3.52697462e-01 -4.08817865e-02 8.09329867e-01
5.40278018e-01 2.40283027e-01 -9.46862638e-01 -2.42959946e-01
-3.14626336e-01 2.49420516e-02 3.31967384e-01 6.42928258e-02
-9.12541866e-01 2.88756102e-01 4.78000641e+00 4.26352173e-01
-1.44674182e+00 -5.49988113e-02 5.22698462e-01 2.89205909e-01
2.40819260e-01 -1.63504377e-01 -1.48633718e-01 8.03332150e-01
1.06806040e+00 -4.38000262e-01 5.48906028e-01 6.98911011e-01
6.41327798e-01 -7.47607648e-02 -9.68268633e-01 5.67167699e-01
-6.49025142e-02 -9.09265280e-01 -2.49642327e-01 3.10662799e-02
5.46221018e-01 5.71620204e-02 -2.07012653e-01 1.89795062e-01
2.52839714e-01 -8.18524539e-01 3.69205058e-01 2.65154004e-01
2.09277436e-01 -9.79360342e-01 1.08285093e+00 3.80564779e-01
-9.71146584e-01 -5.59649110e-01 6.81339949e-02 -2.84740269e-01
4.01089787e-01 7.76077926e-01 -7.10674047e-01 9.99985158e-01
6.96992457e-01 7.23951161e-01 -3.06274235e-01 8.38756502e-01
-5.25904119e-01 9.03962672e-01 -3.63957554e-01 -4.03250046e-02
-1.31663814e-01 -1.92037448e-01 6.64042711e-01 6.14800036e-01
1.28360903e-02 -1.29000068e-01 4.12523717e-01 6.94703877e-01
1.90975621e-01 -2.94532031e-01 -6.40202701e-01 -3.71544040e-03
7.43663728e-01 1.33803761e+00 -6.72195077e-01 -1.11218482e-01
-2.22757846e-01 5.72578311e-01 -1.13048978e-01 4.13392961e-01
-7.81962454e-01 -4.02297050e-01 5.85013390e-01 -3.39461654e-01
-1.28854617e-01 -2.39636645e-01 -2.69536942e-01 -1.14510226e+00
5.57906777e-02 -1.13362992e+00 5.05143762e-01 -3.81611407e-01
-1.61403418e+00 5.05718112e-01 -1.35011673e-01 -1.56257927e+00
-4.66628045e-01 -2.18109712e-01 -9.42819238e-01 6.83793247e-01
-1.45570946e+00 -6.98873818e-01 3.21537536e-03 6.98655486e-01
2.68038422e-01 -2.44434714e-01 9.19177949e-01 2.09486872e-01
-8.72279644e-01 4.34125751e-01 2.21742287e-01 2.67261535e-01
-9.52523276e-02 -1.12825906e+00 2.64778346e-01 1.30676901e+00
-4.20971401e-02 1.24785691e-01 8.80567670e-01 -4.98232871e-01
-1.73438597e+00 -8.38292301e-01 -1.13307014e-01 4.28471193e-02
1.18642902e+00 -4.06253278e-01 -1.16472507e+00 5.15697181e-01
5.32516897e-01 -3.73292752e-02 4.90453809e-01 -5.75470328e-01
-1.78079791e-02 -4.29383397e-01 -1.57694638e+00 2.87262827e-01
1.57002568e-01 -3.79925907e-01 -6.61710560e-01 2.75993288e-01
3.99728149e-01 -7.86293894e-02 -1.24295700e+00 4.31009412e-01
-5.72086684e-02 -5.83416045e-01 7.68305600e-01 -4.26166952e-01
-4.77667958e-01 -4.48245257e-01 -4.68226559e-02 -1.90162504e+00
-1.46635205e-01 -9.98528898e-01 -5.64059913e-01 1.21642649e+00
-1.29980808e-02 -1.39024305e+00 3.65122080e-01 1.95391059e-01
-4.86438759e-02 -5.07770360e-01 -1.41108119e+00 -8.63203168e-01
5.72031289e-02 -3.65142435e-01 1.18142259e+00 1.57665586e+00
4.71354425e-01 -1.26922905e-01 -2.46196985e-01 1.05200565e+00
8.57823849e-01 1.41214952e-01 5.72520018e-01 -9.10449922e-01
-7.29523301e-02 -3.18572015e-01 -6.72885060e-01 -2.19284981e-01
1.95037499e-02 -5.49757481e-01 -3.48914623e-01 -9.91768062e-01
-6.13687456e-01 -2.27147207e-01 -5.88911653e-01 5.55140197e-01
6.72587901e-02 -5.00529222e-02 2.13161066e-01 -4.72403802e-02
-1.27620295e-01 8.07525933e-01 7.54038632e-01 -3.25518221e-01
8.25233236e-02 2.78439015e-01 -4.49165106e-02 4.69056040e-01
1.27053475e+00 -1.84195593e-01 -3.00889671e-01 4.11095843e-02
-2.83987932e-02 3.49683672e-01 2.74961323e-01 -1.31534648e+00
1.63301870e-01 8.31041038e-02 5.26469827e-01 -4.91528779e-01
-1.08360410e-01 -1.42986262e+00 2.17973977e-01 9.14393902e-01
2.62574792e-01 4.52118397e-01 3.62267256e-01 5.24164975e-01
-2.02214807e-01 4.52661179e-02 6.52854264e-01 2.20403507e-01
-6.56806052e-01 2.04014733e-01 -6.94351614e-01 -2.89569318e-01
1.35925496e+00 -1.12769864e-01 -4.17848051e-01 -2.60347068e-01
-6.90374017e-01 5.49928129e-01 2.11924866e-01 5.94358623e-01
5.17687142e-01 -1.04093993e+00 -4.36424971e-01 6.49625063e-01
-2.99665809e-01 -3.72620910e-01 4.14676607e-01 6.98989809e-01
-8.51822570e-02 1.20371275e-01 -2.59885758e-01 -5.84369898e-01
-7.62013376e-01 8.09889257e-01 6.70815587e-01 -1.38809204e-01
-7.09916830e-01 -1.13012999e-01 -6.67959571e-01 -2.45024785e-01
3.26890826e-01 -1.55803651e-01 -2.67651409e-01 1.60409093e-01
6.02993608e-01 5.44645190e-01 4.93589103e-01 -1.68233678e-01
-2.69204021e-01 -1.53174112e-03 3.71398292e-02 5.80466449e-01
1.27846599e+00 -1.56037629e-01 -1.45451799e-01 2.51705229e-01
8.99621904e-01 -3.04675370e-01 -1.24709082e+00 1.09781384e-01
1.15786046e-01 -2.79170871e-01 1.31174266e-01 -1.20585132e+00
-1.42238224e+00 6.23752534e-01 8.86560023e-01 1.15527356e+00
1.36243725e+00 -6.05977893e-01 6.18781507e-01 1.10146932e-01
9.18492496e-01 -1.00830805e+00 -6.79465681e-02 2.15104923e-01
5.37367582e-01 -9.99824762e-01 -6.13062792e-02 2.33757332e-01
-9.43616554e-02 1.17861831e+00 4.80202466e-01 -1.52831495e-01
8.84913027e-01 7.40843832e-01 6.36766851e-02 -1.57328346e-03
-6.32242084e-01 5.45004189e-01 -3.98367286e-01 1.06721568e+00
-3.81758571e-01 3.00938040e-01 5.58258407e-02 4.63878214e-01
-1.05226584e-01 -1.31576315e-01 8.38834822e-01 9.35301185e-01
-1.04349010e-01 -7.37935007e-01 -6.32427931e-01 4.13195282e-01
-4.67180252e-01 3.09334695e-01 1.76932529e-01 9.80311990e-01
-9.53614488e-02 1.16644824e+00 -1.54467300e-01 -4.29554522e-01
3.97633374e-01 -2.72663504e-01 -1.93261743e-01 1.69710964e-01
-6.39599562e-01 -3.38663995e-01 -2.83935577e-01 -5.60039878e-01
3.82135771e-02 -6.06476367e-01 -1.19130695e+00 -3.38055342e-01
-6.48357868e-01 3.51211548e-01 7.66279101e-01 1.19557071e+00
1.98341042e-01 7.07957864e-01 1.51469743e+00 -8.05239618e-01
-1.30678296e+00 -1.05794501e+00 -8.15414011e-01 4.47933465e-01
5.09064913e-01 -5.51963270e-01 -1.02160716e+00 -6.32899582e-01] | [6.064669132232666, 2.5823442935943604] |
636130b9-771e-42eb-8f69-2d19faaa8707 | transformer-based-deep-learning-model-for | 2208.08300 | null | https://arxiv.org/abs/2208.08300v1 | https://arxiv.org/pdf/2208.08300v1.pdf | Transformer-Based Deep Learning Model for Stock Price Prediction: A Case Study on Bangladesh Stock Market | In modern capital market the price of a stock is often considered to be highly volatile and unpredictable because of various social, financial, political and other dynamic factors. With calculated and thoughtful investment, stock market can ensure a handsome profit with minimal capital investment, while incorrect prediction can easily bring catastrophic financial loss to the investors. This paper introduces the application of a recently introduced machine learning model - the Transformer model, to predict the future price of stocks of Dhaka Stock Exchange (DSE), the leading stock exchange in Bangladesh. The transformer model has been widely leveraged for natural language processing and computer vision tasks, but, to the best of our knowledge, has never been used for stock price prediction task at DSE. Recently the introduction of time2vec encoding to represent the time series features has made it possible to employ the transformer model for the stock price prediction. This paper concentrates on the application of transformer-based model to predict the price movement of eight specific stocks listed in DSE based on their historical daily and weekly data. Our experiments demonstrate promising results and acceptable root mean squared error on most of the stocks. | ['Mohammad Shafiul Alam', 'Shahidul Islam Khan', 'Muhammad Ibrahim', 'Maishameem Meherin Muhu', 'Md. Mainul Ahsan', 'Anika Bintee Aftab', 'Tashreef Muhammad'] | 2022-08-17 | null | null | null | null | ['stock-price-prediction'] | ['time-series'] | [-8.94174039e-01 -4.07371551e-01 -1.00918263e-02 -1.68151215e-01
-2.45755970e-01 -7.79459357e-01 6.45631731e-01 -2.33225375e-02
-2.91056424e-01 7.67610908e-01 3.52240473e-01 -5.87549627e-01
-9.75132585e-02 -1.21874619e+00 -2.59106338e-01 -4.75979626e-01
-3.78257245e-01 2.07151279e-01 9.07119550e-03 -5.80817759e-01
7.47997761e-01 6.98090613e-01 -1.33946908e+00 8.84281248e-02
8.98450688e-02 1.29820204e+00 9.96471345e-02 2.00049520e-01
-5.14359653e-01 1.36492753e+00 -4.80984837e-01 -6.38522387e-01
8.75061989e-01 -1.66009702e-02 -3.13064098e-01 -1.21516459e-01
-2.34196842e-01 -3.82108688e-01 -2.30320424e-01 9.25080717e-01
2.10484043e-01 -1.66015178e-01 5.40216684e-01 -1.01925647e+00
-1.00588071e+00 8.32811117e-01 -6.86266661e-01 7.76414275e-01
-1.76337704e-01 -2.47139454e-01 1.38630939e+00 -1.06542897e+00
3.87328416e-01 8.90617073e-01 6.85048938e-01 -1.22558333e-01
-7.15496421e-01 -9.45870757e-01 -5.96237183e-03 1.84105188e-01
-1.00017500e+00 4.59714457e-02 8.27781022e-01 -6.08042777e-01
1.23297882e+00 7.68661126e-02 1.05815649e+00 3.15490782e-01
7.82929838e-01 5.93416691e-01 1.30497801e+00 -1.26366422e-01
5.39004616e-02 3.98984641e-01 -9.33316574e-02 1.83649451e-01
4.02422816e-01 1.71237171e-01 -3.78514171e-01 -8.75122771e-02
5.93056083e-01 2.24078849e-01 -1.46393823e-02 3.02133858e-01
-1.03036737e+00 1.16373158e+00 3.77925456e-01 6.23285592e-01
-7.42702782e-01 -5.67451641e-02 2.72321582e-01 9.37593818e-01
7.94156373e-01 3.98475230e-01 -8.62794757e-01 -4.48052824e-01
-1.13077426e+00 4.25049841e-01 9.36124384e-01 3.32098514e-01
1.44877866e-01 6.06275380e-01 3.80222619e-01 3.98051798e-01
4.44310874e-01 5.53198934e-01 9.19378817e-01 -2.58478701e-01
3.44346136e-01 6.71130478e-01 3.74297202e-01 -1.43057358e+00
-3.76202583e-01 -5.36744773e-01 -4.93770003e-01 5.42850256e-01
9.56453681e-02 -2.09812805e-01 -5.80031574e-01 1.06290674e+00
-2.51368850e-01 3.09066296e-01 4.19606924e-01 5.48540771e-01
2.37210557e-01 1.05220699e+00 -3.51810098e-01 -4.61121142e-01
9.66051459e-01 -6.78107083e-01 -9.24884439e-01 -6.63076788e-02
1.96354285e-01 -8.77475739e-01 2.65741616e-01 4.32601929e-01
-1.00771224e+00 -2.52455652e-01 -9.35354352e-01 4.23866242e-01
-6.79849148e-01 -2.42485449e-01 5.79651773e-01 3.00379544e-01
-9.00567472e-01 7.69580603e-01 -7.76734769e-01 3.72011550e-02
5.41942753e-02 1.86608642e-01 -9.88494679e-02 6.23378336e-01
-1.34475982e+00 1.29688048e+00 4.53454822e-01 2.60628372e-01
-1.25753462e-01 -6.59075141e-01 -4.86748874e-01 -3.11952289e-02
-1.04521550e-01 1.20223686e-01 1.08259928e+00 -7.71491230e-01
-1.21980309e+00 5.49827337e-01 3.94052833e-01 -9.20729041e-01
5.27890563e-01 -1.58074796e-01 -9.28081036e-01 -4.29809570e-01
6.47806004e-02 -2.65001338e-02 3.51603597e-01 -2.69954950e-01
-8.89549434e-01 -1.77509174e-01 -3.34122717e-01 -3.15613523e-02
-3.55958790e-01 1.10337980e-01 2.09480166e-01 -1.18141305e+00
1.53398857e-01 -8.53255808e-01 -1.10211782e-01 -5.91711581e-01
2.15934888e-01 -1.88515112e-01 6.15037620e-01 -9.99373257e-01
1.45605171e+00 -2.11663842e+00 -3.85128736e-01 2.15633363e-01
-2.46330619e-01 1.30454987e-01 2.98969805e-01 8.95158172e-01
-3.89078021e-01 1.94024384e-01 4.64375596e-03 4.85885561e-01
8.71563628e-02 8.60237479e-02 -9.24222350e-01 4.35285062e-01
2.84885049e-01 1.02097416e+00 -4.79423225e-01 2.50087172e-01
1.12101100e-01 1.02806948e-01 -8.05202872e-02 -7.18081817e-02
-1.34501178e-02 -2.17768043e-01 -4.10810381e-01 6.24000013e-01
7.50420928e-01 -1.38967037e-01 -1.72819540e-01 -8.65619257e-03
-7.67852187e-01 2.65830100e-01 -1.17336047e+00 6.42351747e-01
1.46199102e-02 7.19024479e-01 -5.89733303e-01 -8.23423207e-01
1.30316436e+00 3.77797455e-01 6.93901718e-01 -7.85592020e-01
6.69295341e-02 6.80284381e-01 1.26902312e-01 -2.63415933e-01
5.89868307e-01 -6.99672282e-01 -1.09890245e-01 6.45426273e-01
-4.27037328e-01 6.03171997e-04 1.52134225e-01 -1.47548795e-01
5.98154247e-01 -1.21770717e-01 5.12651682e-01 -4.50133830e-01
3.27555507e-01 -4.99327816e-02 7.03257382e-01 1.36801779e-01
-9.55555737e-02 2.46716246e-01 4.05803859e-01 -9.35127079e-01
-1.00400805e+00 -7.40200758e-01 -1.84500039e-01 5.69060862e-01
-5.32945752e-01 2.65702337e-01 -1.69830373e-03 -1.83207676e-01
7.40629792e-01 8.80909681e-01 -5.35060406e-01 2.14232743e-01
-2.78795451e-01 -9.14926529e-01 5.05922139e-02 4.22533602e-01
5.03686607e-01 -1.27219784e+00 -8.14134240e-01 6.88228905e-01
5.00841498e-01 -6.90943480e-01 -2.26223633e-01 1.90980524e-01
-9.50515091e-01 -9.48228598e-01 -9.00771081e-01 -5.54866195e-01
9.85112190e-02 8.48626718e-02 1.03843772e+00 -1.18573092e-01
5.26563972e-02 -5.45993969e-02 -3.58573139e-01 -1.11677074e+00
4.13052998e-02 -3.04805905e-01 1.58401132e-01 2.12744698e-01
7.64110804e-01 -3.75326276e-01 -4.17627901e-01 -5.57108782e-02
-8.53200436e-01 -2.89982051e-01 5.88401139e-01 5.87771893e-01
3.33917439e-01 8.09122741e-01 9.41995144e-01 -5.87153316e-01
7.73111224e-01 -8.42821777e-01 -1.01223147e+00 1.97324529e-01
-1.09489965e+00 2.43139099e-02 5.85064553e-02 -8.30570459e-02
-9.27755415e-01 -4.96532649e-01 1.58124324e-02 -3.36104363e-01
6.85934544e-01 1.30324793e+00 5.46368718e-01 8.90004784e-02
-2.13651299e-01 5.31441689e-01 -5.69063351e-02 -6.97016180e-01
-1.23038381e-01 7.23964870e-01 2.88381219e-01 2.10693493e-01
1.02379346e+00 2.86891401e-01 -7.05935508e-02 -6.08181298e-01
-7.53837466e-01 -4.32515115e-01 -5.70816934e-01 -8.51372629e-02
4.79525208e-01 -1.07775700e+00 -2.90692270e-01 1.03167427e+00
-6.99179828e-01 2.48378783e-01 -5.40462285e-02 6.49884999e-01
-6.37944862e-02 -1.05840355e-01 -7.75620461e-01 -1.18523622e+00
-4.96636003e-01 -6.26649261e-01 3.56604993e-01 1.23710781e-01
-4.92176823e-02 -1.34923065e+00 3.39432508e-01 -7.48596713e-02
8.05812895e-01 5.46470582e-01 7.04459786e-01 -9.65473473e-01
-5.01485229e-01 -6.27910435e-01 -1.05783671e-01 5.87036312e-01
6.76567674e-01 1.50921671e-02 -6.47307456e-01 -3.45012754e-01
4.07499224e-01 2.75432542e-02 7.70999551e-01 2.40838781e-01
1.99689686e-01 -4.30225641e-01 3.69336426e-01 3.26976120e-01
1.72882140e+00 7.15833902e-01 4.89455760e-01 9.84045804e-01
1.84570044e-01 4.69908714e-01 7.47070849e-01 8.32718253e-01
3.93864393e-01 1.03588171e-01 1.00140244e-01 3.19926322e-01
6.99391425e-01 -6.66673556e-02 5.48155308e-01 1.31493568e+00
-5.57991788e-02 4.56605777e-02 -1.00606787e+00 6.58686101e-01
-1.43745387e+00 -1.25944257e+00 1.17442369e-01 1.88644743e+00
6.20533526e-01 4.97699648e-01 9.16844830e-02 3.33831191e-01
2.28079617e-01 4.16561812e-01 -4.58096176e-01 -5.50181091e-01
-2.81610966e-01 6.67273030e-02 1.00509930e+00 1.95017129e-01
-9.91968453e-01 7.38222241e-01 6.11917543e+00 3.36891413e-01
-1.55500245e+00 -3.08907032e-01 6.61573827e-01 -1.18121743e-01
-5.24333417e-01 -2.84478158e-01 -7.54193306e-01 8.56820405e-01
1.13524783e+00 -8.86172771e-01 1.13233358e-01 8.03390622e-01
4.16733861e-01 8.38848725e-02 -5.33548176e-01 7.63111055e-01
-1.89001501e-01 -1.50429058e+00 6.69800267e-02 2.37806469e-01
6.10521793e-01 1.82976529e-01 3.62602264e-01 3.08643788e-01
5.25636598e-02 -8.51023734e-01 1.00684011e+00 8.11037123e-01
7.33968541e-02 -1.19512904e+00 1.21815538e+00 2.82791406e-01
-1.24077988e+00 -3.71515065e-01 -4.35540527e-01 -5.58101356e-01
3.05206358e-01 5.09447873e-01 -5.27432799e-01 2.34884575e-01
8.52024078e-01 9.46779370e-01 -2.19297558e-01 7.79097199e-01
2.99480915e-01 5.56270421e-01 -2.02330768e-01 -3.08997661e-01
8.26437712e-01 -6.39170885e-01 3.13041121e-01 7.96918809e-01
7.58527040e-01 3.53366971e-01 -1.05332859e-01 4.38321829e-01
2.03502417e-01 2.80542344e-01 -7.22568154e-01 -5.61507404e-01
3.45282286e-01 6.04382038e-01 -5.88490009e-01 -5.94042167e-02
-9.82920289e-01 6.10898495e-01 5.94371557e-02 2.36216616e-02
-4.82663155e-01 -4.70645308e-01 7.42550313e-01 1.58913344e-01
6.84915543e-01 -2.55780786e-01 -1.92016765e-01 -1.22243416e+00
1.24387003e-01 -5.45295894e-01 3.04103523e-01 -4.85433728e-01
-1.58457184e+00 5.39757490e-01 -1.19413860e-01 -1.51425374e+00
-3.89693618e-01 -7.77685046e-01 -8.72652948e-01 1.08034384e+00
-1.83820772e+00 -5.70698977e-01 6.72765136e-01 2.71983594e-01
5.47107935e-01 -1.02683258e+00 5.58734953e-01 7.43505508e-02
-4.34131294e-01 -6.60970109e-03 6.57998443e-01 3.30187500e-01
4.77419533e-02 -1.42776096e+00 9.21558142e-01 7.71189213e-01
2.26718172e-01 4.24585968e-01 7.16088176e-01 -9.50585723e-01
-1.16530299e+00 -1.11532664e+00 1.49206877e+00 -2.19636112e-01
1.56111467e+00 1.61806330e-01 -9.54664767e-01 8.56366992e-01
3.98894310e-01 -2.63526142e-01 8.03481996e-01 -4.27997500e-01
-1.20549567e-01 -3.44191372e-01 -1.14718831e+00 2.89567769e-01
-6.73891306e-02 -3.39928240e-01 -1.06537008e+00 1.21321835e-01
4.37438548e-01 2.12530613e-01 -1.05772114e+00 6.22059293e-02
7.64910877e-01 -1.00676787e+00 7.41406798e-01 -5.28501570e-01
1.15396820e-01 -1.17691476e-02 -2.87805706e-01 -1.31353676e+00
-5.74339867e-01 -2.96505034e-01 5.90285696e-02 1.25315666e+00
7.26327777e-01 -9.26160455e-01 6.63837910e-01 8.96265626e-01
3.93459260e-01 -7.37090707e-01 -9.35856044e-01 -8.23423386e-01
2.37372115e-01 -4.60303724e-01 9.63430941e-01 1.02018237e+00
-1.39777362e-01 -2.12126330e-01 -6.82122648e-01 -1.04406849e-02
4.11830813e-01 4.37610567e-01 1.63022220e-01 -1.22962713e+00
-8.43912438e-02 -6.01522267e-01 -6.98378205e-01 -3.67459387e-01
-9.32225212e-02 -6.23737395e-01 -6.26384616e-01 -1.17996848e+00
-2.27428317e-01 -1.75523043e-01 -9.21403646e-01 1.50279433e-01
2.13990927e-01 1.97219104e-01 4.50577796e-01 4.63781565e-01
2.92970985e-01 3.41365486e-01 9.58573818e-01 -1.29692838e-01
-1.15521021e-01 4.10796702e-01 -6.80912912e-01 7.04609513e-01
1.00715613e+00 -2.24259302e-01 -2.49232754e-01 -3.17363828e-01
6.96258187e-01 1.90874577e-01 -1.62028849e-01 -6.65037155e-01
1.00009620e-01 -4.28740382e-01 3.75750870e-01 -1.11892068e+00
-9.35716555e-03 -1.07017505e+00 5.76095581e-01 7.65537381e-01
-9.83454362e-02 8.80559742e-01 2.69599408e-01 4.66789007e-01
-8.34023416e-01 -1.95052192e-01 5.53475142e-01 -4.62730885e-01
-9.54847276e-01 2.19737023e-01 -5.28154075e-01 -3.97409886e-01
1.34702909e+00 -2.40826741e-01 2.27092966e-01 -4.29878235e-01
-3.91769558e-01 2.33181566e-01 2.88469922e-02 6.40517116e-01
6.95532620e-01 -1.56819093e+00 -1.01234043e+00 3.15605938e-01
-2.41791070e-01 -6.86106622e-01 -1.85286835e-01 5.68259299e-01
-1.08859575e+00 8.58361006e-01 -2.92269111e-01 2.49636084e-01
-6.71829283e-01 5.75947404e-01 2.80598283e-01 -3.58468115e-01
-6.79590940e-01 9.23098981e-01 -3.26205850e-01 1.16779871e-01
-3.94479893e-02 -4.85006720e-01 -6.51038349e-01 8.90761912e-01
8.56889844e-01 3.20164353e-01 8.75616819e-02 -8.63197446e-01
-2.68912077e-01 6.84461892e-01 -2.19772384e-01 -2.63165414e-01
2.03124809e+00 -1.51964083e-01 -3.70282680e-01 1.09709632e+00
1.13126850e+00 3.87985967e-02 -8.60495269e-01 -1.79337054e-01
8.76787126e-01 -5.37607193e-01 3.31596941e-01 -7.25985050e-01
-1.55657744e+00 5.16329944e-01 5.53501785e-01 5.54780543e-01
1.14455485e+00 -5.67563593e-01 1.05685472e+00 3.79780740e-01
5.96545219e-01 -1.11097872e+00 -4.78120297e-01 5.13471365e-01
9.54421937e-01 -1.22243440e+00 -3.42133194e-02 4.35203105e-01
-8.07502449e-01 1.29119122e+00 -1.67504847e-01 -4.19593781e-01
1.35184133e+00 4.48628873e-01 5.27733147e-01 -1.66289404e-01
-1.01826191e+00 2.51325160e-01 2.24060297e-01 7.13092387e-02
5.51885426e-01 1.73145279e-01 -3.86621803e-01 7.06376672e-01
-4.88068581e-01 1.76988825e-01 7.78505802e-01 1.25458419e+00
-7.10818768e-01 -9.34657812e-01 -3.03300411e-01 8.34895968e-01
-1.19297040e+00 -3.92972499e-01 -2.04515144e-01 9.50474560e-01
-2.50498295e-01 5.73282659e-01 6.21775031e-01 -2.58647054e-01
3.02078515e-01 1.45800918e-01 -2.62576461e-01 -2.83889234e-01
-6.84980452e-01 1.93833724e-01 -1.97394103e-01 -1.45450309e-01
-5.73027730e-01 -1.08782721e+00 -1.07730675e+00 -6.59099281e-01
-4.33777347e-02 4.38533813e-01 5.44251621e-01 8.89347017e-01
-6.03501461e-02 8.01441893e-02 1.17411268e+00 -5.47798038e-01
-1.02940285e+00 -9.65826333e-01 -1.42412698e+00 1.37637123e-01
4.64690208e-01 -5.16316533e-01 -4.26430523e-01 -7.21100047e-02] | [4.463914394378662, 4.240828990936279] |
220ed4ef-3432-4a46-906e-9b287b840b1c | cia-ssd-confident-iou-aware-single-stage | 2012.03015 | null | https://arxiv.org/abs/2012.03015v1 | https://arxiv.org/pdf/2012.03015v1.pdf | CIA-SSD: Confident IoU-Aware Single-Stage Object Detector From Point Cloud | Existing single-stage detectors for locating objects in point clouds often treat object localization and category classification as separate tasks, so the localization accuracy and classification confidence may not well align. To address this issue, we present a new single-stage detector named the Confident IoU-Aware Single-Stage object Detector (CIA-SSD). First, we design the lightweight Spatial-Semantic Feature Aggregation module to adaptively fuse high-level abstract semantic features and low-level spatial features for accurate predictions of bounding boxes and classification confidence. Also, the predicted confidence is further rectified with our designed IoU-aware confidence rectification module to make the confidence more consistent with the localization accuracy. Based on the rectified confidence, we further formulate the Distance-variant IoU-weighted NMS to obtain smoother regressions and avoid redundant predictions. We experiment CIA-SSD on 3D car detection in the KITTI test set and show that it attains top performance in terms of the official ranking metric (moderate AP 80.28%) and above 32 FPS inference speed, outperforming all prior single-stage detectors. The code is available at https://github.com/Vegeta2020/CIA-SSD. | ['Chi-Wing Fu', 'Li Jiang', 'Sijin Chen', 'Weiliang Tang', 'Wu Zheng'] | 2020-12-05 | null | null | null | null | ['birds-eye-view-object-detection'] | ['computer-vision'] | [-3.06040883e-01 -1.83983505e-01 -1.08892582e-01 -7.13711739e-01
-1.12072170e+00 -4.37891215e-01 4.33880121e-01 1.84464410e-01
-4.28071022e-01 1.82928413e-01 -3.86068225e-01 -2.87904590e-01
1.18954793e-01 -5.68801284e-01 -8.79475415e-01 -5.28309643e-01
2.27075592e-01 3.93623263e-01 9.98194218e-01 3.08563471e-01
3.53862226e-01 6.51020527e-01 -1.80617809e+00 1.60479009e-01
1.09360433e+00 1.50209963e+00 7.17108488e-01 6.12329781e-01
-1.57014895e-02 3.09001893e-01 -4.46067601e-01 -1.11974403e-01
1.77249059e-01 3.77825350e-01 -2.62110382e-01 -2.81604528e-01
4.84834194e-01 -4.01970834e-01 8.70000347e-02 1.05270863e+00
1.94897503e-01 -3.96323092e-02 7.86557555e-01 -1.31815910e+00
-5.45945823e-01 3.97373363e-02 -7.73842037e-01 2.81145394e-01
8.50791186e-02 1.07187323e-01 8.88911486e-01 -1.46174800e+00
1.27452880e-01 1.30903518e+00 8.64638269e-01 3.08391333e-01
-9.02353048e-01 -8.95094037e-01 4.80886191e-01 1.79641977e-01
-1.87480998e+00 -1.55350566e-01 3.52183461e-01 -4.15006459e-01
8.32989872e-01 2.48599529e-01 3.35455179e-01 4.53803241e-01
1.34399682e-01 8.75271320e-01 8.61516714e-01 -1.34167314e-01
2.57165223e-01 4.29571599e-01 1.92905441e-01 7.16440141e-01
5.83803236e-01 6.86881840e-02 -2.93381333e-01 -6.23578252e-03
4.29779053e-01 1.09778479e-01 1.86403453e-01 -3.39906007e-01
-1.17449021e+00 7.44838119e-01 9.24504697e-01 -7.35807344e-02
-2.10498720e-01 1.46121606e-01 -3.24259920e-04 -4.00366068e-01
6.84859216e-01 7.58984610e-02 -5.39614499e-01 2.59565920e-01
-8.13495338e-01 2.42160439e-01 2.95203596e-01 1.39089763e+00
8.06073546e-01 -2.67196774e-01 -4.34758276e-01 7.49959409e-01
7.35594571e-01 8.21471930e-01 3.79721560e-02 -7.84340560e-01
2.78914601e-01 6.14464760e-01 3.71962011e-01 -8.35700393e-01
-3.26827377e-01 -4.48279411e-01 -3.55439842e-01 4.34477836e-01
1.27985269e-01 2.13319108e-01 -1.15865088e+00 1.29725218e+00
6.23642683e-01 4.54467595e-01 -1.11921266e-01 1.16602504e+00
9.16501284e-01 7.12884307e-01 4.56778049e-01 2.12690741e-01
1.49257100e+00 -9.54636157e-01 -2.19526336e-01 -4.99396414e-01
6.37672365e-01 -7.28928089e-01 9.73696649e-01 -5.44520505e-02
-7.05696285e-01 -9.04134572e-01 -1.23324013e+00 -8.45335647e-02
-4.43365157e-01 6.73152506e-01 5.23422301e-01 4.45177227e-01
-7.00912595e-01 2.75277644e-01 -8.36781442e-01 -2.18890682e-01
6.32405102e-01 1.87954634e-01 -6.83801100e-02 -3.39783914e-02
-8.50878417e-01 8.84093881e-01 5.06006956e-01 2.39805833e-01
-6.52134776e-01 -8.27639341e-01 -6.81119859e-01 -6.26933426e-02
3.85311276e-01 -4.52592880e-01 1.42148936e+00 -2.29018092e-01
-1.11412704e+00 8.67517173e-01 -3.71885210e-01 -4.10273492e-01
5.39109409e-01 -3.83324653e-01 -3.35919470e-01 -2.11873755e-01
4.36056942e-01 1.18299234e+00 7.56930470e-01 -1.33804893e+00
-1.42936003e+00 -3.77499133e-01 -2.54034221e-01 2.10946813e-01
8.21312815e-02 -1.89270690e-01 -6.59370482e-01 -3.93408835e-01
4.51968849e-01 -8.72493088e-01 -1.32407859e-01 2.97665715e-01
-1.88514456e-01 -6.44157112e-01 1.03244364e+00 -3.36889446e-01
1.01722240e+00 -2.30003881e+00 -4.76868093e-01 1.18258387e-01
2.12223068e-01 3.21836650e-01 6.82136416e-02 -3.68203402e-01
3.46285939e-01 -1.70821920e-01 -3.16362083e-02 -5.98025322e-01
-3.34285572e-02 6.25491738e-02 -2.81537563e-01 4.88593102e-01
6.07852101e-01 7.71022856e-01 -8.59749913e-01 -6.88754678e-01
6.31143153e-01 4.60858375e-01 -5.51921606e-01 1.91490233e-01
-2.80444890e-01 3.01039457e-01 -6.94403768e-01 1.09798396e+00
1.20215213e+00 -2.01851293e-01 -3.81985158e-01 -3.64274919e-01
-4.02343571e-01 1.55296981e-01 -1.32845998e+00 1.26782703e+00
-4.05831933e-01 4.26341295e-01 -1.42127499e-01 -4.97487903e-01
1.17119169e+00 -2.65373409e-01 2.04158694e-01 -4.89193112e-01
8.05352554e-02 4.02236491e-01 -4.22821492e-01 1.00382484e-01
5.80201328e-01 3.16651881e-01 -7.54072964e-02 -2.71721125e-01
-2.17317984e-01 -2.75585234e-01 -2.52994686e-01 9.31573287e-02
7.36538589e-01 2.76893765e-01 1.15654878e-01 -4.31023151e-01
5.87822974e-01 2.32885838e-01 6.42137825e-01 8.44649553e-01
-5.91084778e-01 7.84252942e-01 3.17081213e-02 -1.85970649e-01
-9.41980481e-01 -1.23247099e+00 -5.99417746e-01 9.93251741e-01
7.61110485e-01 -2.75224656e-01 -6.89570129e-01 -7.96568096e-01
5.28844178e-01 8.74668896e-01 -3.88144493e-01 -7.62490556e-02
-2.82567352e-01 -4.56510782e-01 1.42665088e-01 9.85234439e-01
5.98629832e-01 -5.39074183e-01 -5.86093068e-01 -2.51017231e-02
1.23771019e-01 -1.19661784e+00 -4.44803894e-01 2.17127964e-01
-6.12237275e-01 -7.97636449e-01 -4.61261868e-01 -8.05117130e-01
5.76819539e-01 6.40584350e-01 7.97423661e-01 9.14672911e-02
-2.58899480e-01 2.94694584e-02 -3.74131560e-01 -7.87945032e-01
-2.69318391e-02 1.12853595e-03 1.13872312e-01 -3.23530048e-01
6.88774765e-01 2.30494933e-03 -7.74266124e-01 6.11754775e-01
-3.27113241e-01 1.21637501e-01 7.03381240e-01 5.63654125e-01
1.04486549e+00 -2.34707475e-01 5.28736770e-01 -3.28114510e-01
1.63578391e-02 -4.69961256e-01 -1.11715317e+00 1.65200025e-01
-7.47552454e-01 -1.13540292e-01 1.18053578e-01 -4.28655058e-01
-1.03707552e+00 3.66733640e-01 -2.34020025e-01 -7.08851218e-01
-1.74486578e-01 -4.71520089e-02 -1.50422275e-01 -1.78291976e-01
5.16577780e-01 -2.92154793e-02 -3.54790628e-01 -5.07461727e-01
3.87032062e-01 1.01599431e+00 7.07764030e-01 -3.13881367e-01
8.64136100e-01 4.99058127e-01 -2.40960851e-01 -5.25323033e-01
-1.20375860e+00 -8.83981824e-01 -6.19167626e-01 -3.29669654e-01
1.05061424e+00 -1.31218362e+00 -8.85589600e-01 3.19187343e-01
-1.26693773e+00 -6.10252805e-02 -7.87666142e-02 6.60587668e-01
-3.25381279e-01 4.37143147e-02 -3.57912153e-01 -1.05811501e+00
-3.23353320e-01 -1.14842927e+00 1.63593793e+00 5.10355294e-01
1.61496326e-01 -2.68702894e-01 -3.14767182e-01 2.42894515e-01
1.72257617e-01 -1.18308082e-01 2.57959932e-01 -4.81462389e-01
-9.17193592e-01 -4.22643930e-01 -8.73601675e-01 3.74074280e-01
-1.16965860e-01 1.03465997e-01 -1.12008667e+00 -1.55647397e-01
-2.68378794e-01 -5.53527549e-02 1.04288077e+00 4.78187799e-01
1.48769128e+00 2.83419043e-01 -6.93249822e-01 5.14539301e-01
1.23684561e+00 7.10558444e-02 3.05245787e-01 2.63168246e-01
6.07563794e-01 1.68494195e-01 1.39191663e+00 4.12980229e-01
5.08132696e-01 7.22139835e-01 5.78952193e-01 5.08738868e-02
-1.31206810e-01 -3.94426584e-01 2.34413538e-02 4.28841889e-01
5.38658202e-02 -1.78442016e-01 -9.21203434e-01 4.50738251e-01
-1.90740919e+00 -6.01263642e-01 -3.29504699e-01 2.12870455e+00
3.68047476e-01 5.27860940e-01 8.99956524e-02 -2.99758285e-01
9.07697618e-01 -1.67309389e-01 -6.21999085e-01 5.25734313e-02
1.19245403e-01 -2.38266245e-01 9.59175229e-01 5.52828789e-01
-1.42050791e+00 1.09538770e+00 5.53026438e+00 1.08407652e+00
-8.06805193e-01 3.28804523e-01 6.27347410e-01 1.82853699e-01
-9.80643556e-03 -1.38911158e-02 -1.50127769e+00 5.80702007e-01
7.79865026e-01 1.96613073e-01 -2.03497902e-01 1.57581615e+00
6.73309565e-02 -3.60762239e-01 -8.70809674e-01 9.03148711e-01
-2.26820230e-01 -1.22292542e+00 -2.79670596e-01 -9.32138488e-02
5.44217348e-01 3.13113838e-01 4.46851552e-02 4.74552423e-01
1.40060797e-01 -6.72683597e-01 1.31878424e+00 4.15944487e-01
9.08370078e-01 -7.20106244e-01 8.54126215e-01 4.61710364e-01
-1.65680575e+00 -1.20711692e-01 -6.89047575e-01 2.25058973e-01
1.26130328e-01 5.14784455e-01 -9.75831389e-01 3.51545990e-01
1.08583653e+00 5.67792416e-01 -7.13149667e-01 1.16843641e+00
-1.69998854e-01 3.98904830e-01 -6.97809637e-01 -2.49720335e-01
2.33154044e-01 -8.76932777e-03 4.62321609e-01 1.34278190e+00
4.23701763e-01 1.67306170e-01 2.13567883e-01 9.49700236e-01
2.02445805e-01 -1.95570424e-01 -1.18072994e-01 5.46285152e-01
9.17197466e-01 1.35183513e+00 -8.24039102e-01 -4.03396428e-01
-4.85758960e-01 7.98448741e-01 3.85620564e-01 -9.03582498e-02
-1.37804961e+00 -2.34542996e-01 9.20490921e-01 1.76360875e-01
6.56347632e-01 -2.44813755e-01 -4.65234846e-01 -9.09939706e-01
8.13647136e-02 -5.44020496e-02 2.93260694e-01 -9.99630392e-01
-1.20298731e+00 5.78985214e-01 1.47896230e-01 -1.40429688e+00
1.88478321e-01 -8.08611095e-01 -5.40105581e-01 8.06228280e-01
-1.63541174e+00 -1.31238735e+00 -7.54817367e-01 1.42697319e-01
6.47887945e-01 1.87831193e-01 3.44718575e-01 4.15824890e-01
-6.09994471e-01 6.58493042e-01 -6.62545562e-02 -1.26793623e-01
6.72611535e-01 -1.01382554e+00 5.70179641e-01 8.44156504e-01
-1.83318466e-01 3.31273526e-01 5.86616457e-01 -8.07456017e-01
-1.04622912e+00 -1.43725657e+00 6.66305423e-01 -7.27513969e-01
5.38020551e-01 -4.98246282e-01 -1.04420614e+00 5.05442202e-01
-6.50644541e-01 4.18394148e-01 -7.99638033e-02 -1.49197683e-01
-3.67825270e-01 -3.11555743e-01 -1.23817456e+00 3.58217061e-01
1.17761183e+00 -3.57487530e-01 -4.20932353e-01 4.34945017e-01
1.06808937e+00 -6.23094380e-01 -6.32032692e-01 8.64235342e-01
4.71718282e-01 -7.81551957e-01 1.15351737e+00 -2.39553005e-02
-7.54860714e-02 -9.13491011e-01 -3.57891977e-01 -6.79950178e-01
-5.42865813e-01 2.07548931e-01 -1.47256777e-01 1.12748623e+00
5.02304435e-01 -5.95829666e-01 9.16864932e-01 5.51395833e-01
-5.31110108e-01 -7.06781268e-01 -1.03607178e+00 -9.14863765e-01
-2.31829166e-01 -7.50001669e-01 6.33677900e-01 3.61926317e-01
-5.49509168e-01 -2.07410268e-02 1.61015161e-03 7.73296833e-01
7.21582949e-01 1.50088578e-01 7.00964332e-01 -1.18079734e+00
5.28137162e-02 -3.02086115e-01 -5.63810468e-01 -1.30124950e+00
-8.23148862e-02 -7.45013595e-01 6.50923073e-01 -1.43067014e+00
1.71169624e-01 -9.52071071e-01 -2.75173485e-01 2.62541443e-01
-3.60918492e-01 2.86358088e-01 9.80651230e-02 4.58178580e-01
-8.77766311e-01 5.33274233e-01 7.52571404e-01 3.41694467e-02
-1.01718545e-01 3.66439313e-01 -4.18300539e-01 7.22510636e-01
8.08077693e-01 -5.26645303e-01 -6.96871653e-02 -3.08635563e-01
-2.82591522e-01 -4.28287506e-01 6.19325459e-01 -1.41134465e+00
3.64284098e-01 -2.68498868e-01 6.10941231e-01 -1.33110797e+00
4.27187920e-01 -8.27156603e-01 -1.00146100e-01 5.46005607e-01
5.35395704e-02 -1.83320537e-01 3.06915581e-01 8.13538909e-01
-1.00895032e-01 -9.11212862e-02 9.94450748e-01 3.60113651e-01
-1.30588388e+00 3.69660616e-01 -3.91208753e-02 -4.01067406e-01
1.30080140e+00 -4.56298351e-01 -2.94652671e-01 1.40844136e-01
-2.87424624e-01 5.10972381e-01 4.36881661e-01 5.76296926e-01
6.94746315e-01 -1.25444221e+00 -5.83217382e-01 2.51190424e-01
6.36321545e-01 3.96779269e-01 3.13866109e-01 7.20670938e-01
-5.58082283e-01 6.47478402e-01 2.42605641e-01 -1.14416158e+00
-1.10593021e+00 5.56865036e-01 2.61202157e-01 2.69287109e-01
-5.64248025e-01 1.09232283e+00 3.66963267e-01 -3.29305738e-01
3.73559535e-01 -6.40602469e-01 6.66993260e-02 -3.37555200e-01
5.25694788e-01 2.70021379e-01 1.53284833e-01 -5.43532968e-01
-9.07471240e-01 6.39534175e-01 1.53567772e-02 3.77909869e-01
1.01155567e+00 -2.03336075e-01 3.71868968e-01 1.80662453e-01
9.21252608e-01 -1.10018261e-01 -1.65490758e+00 -2.00605854e-01
1.17727585e-01 -6.61013961e-01 2.02677697e-01 -6.09391749e-01
-7.01554358e-01 5.82580209e-01 9.56830859e-01 -1.98815435e-01
7.89847076e-01 5.49710512e-01 5.14203906e-01 2.71635294e-01
3.40553582e-01 -1.03660929e+00 -4.28136513e-02 5.55736363e-01
7.55900443e-01 -1.75302958e+00 1.16173334e-01 -7.16355741e-01
-7.18519747e-01 6.98901951e-01 1.09791517e+00 -1.85975254e-01
7.07998514e-01 3.15454543e-01 -1.39686257e-01 -3.40817533e-02
-4.70833510e-01 -3.93148899e-01 6.90632999e-01 5.78027964e-01
6.65165409e-02 2.79042125e-01 -6.59892708e-02 7.04231381e-01
1.73734710e-01 -2.12239221e-01 -2.59895384e-01 8.10180366e-01
-1.02914882e+00 -3.60336959e-01 -4.88906741e-01 5.16310811e-01
6.25281259e-02 7.78919458e-02 1.03609838e-01 7.35126913e-01
3.35367143e-01 9.22267139e-01 4.63531643e-01 -6.05846643e-01
4.97238219e-01 -2.84975559e-01 1.11672744e-01 -6.62695348e-01
2.35043000e-02 3.22244056e-02 -1.08559385e-01 -7.37233043e-01
-1.79017603e-01 -6.62201405e-01 -1.47900712e+00 -3.73580188e-01
-8.62540781e-01 -8.69385973e-02 1.02149212e+00 7.02337146e-01
4.69153255e-01 3.90333235e-01 4.96473610e-01 -1.18075216e+00
-5.88082314e-01 -9.52294052e-01 -3.10176343e-01 -3.43114063e-02
2.38731787e-01 -1.00979614e+00 -4.37809467e-01 -3.66786301e-01] | [8.460977554321289, -0.6140543818473816] |
df8462e7-ae8c-4b96-9c16-0142635f1a4f | qcnext-a-next-generation-framework-for-joint | 2306.10508 | null | https://arxiv.org/abs/2306.10508v1 | https://arxiv.org/pdf/2306.10508v1.pdf | QCNeXt: A Next-Generation Framework For Joint Multi-Agent Trajectory Prediction | Estimating the joint distribution of on-road agents' future trajectories is essential for autonomous driving. In this technical report, we propose a next-generation framework for joint multi-agent trajectory prediction called QCNeXt. First, we adopt the query-centric encoding paradigm for the task of joint multi-agent trajectory prediction. Powered by this encoding scheme, our scene encoder is equipped with permutation equivariance on the set elements, roto-translation invariance in the space dimension, and translation invariance in the time dimension. These invariance properties not only enable accurate multi-agent forecasting fundamentally but also empower the encoder with the capability of streaming processing. Second, we propose a multi-agent DETR-like decoder, which facilitates joint multi-agent trajectory prediction by modeling agents' interactions at future time steps. For the first time, we show that a joint prediction model can outperform marginal prediction models even on the marginal metrics, which opens up new research opportunities in trajectory prediction. Our approach ranks 1st on the Argoverse 2 multi-agent motion forecasting benchmark, winning the championship of the Argoverse Challenge at the CVPR 2023 Workshop on Autonomous Driving. | ['Yu-Kai Huang', 'Yung-Hui Li', 'JianPing Wang', 'Zihao Wen', 'Zikang Zhou'] | 2023-06-18 | null | null | null | null | ['trajectory-prediction', 'motion-forecasting'] | ['computer-vision', 'computer-vision'] | [-2.32950911e-01 -5.80455502e-03 -4.75372642e-01 -3.64501387e-01
-9.54695582e-01 -4.13835794e-01 9.33504939e-01 9.72350240e-02
-3.82900894e-01 4.76172388e-01 6.04781210e-01 -2.42838010e-01
-1.24249637e-01 -8.34930182e-01 -1.05345047e+00 -5.76704621e-01
-3.25899392e-01 6.49453223e-01 3.69097590e-01 -3.27009559e-01
9.69486907e-02 3.36101532e-01 -1.68706179e+00 2.52337217e-01
7.67258108e-01 6.11605763e-01 2.66766459e-01 1.10475945e+00
2.25682482e-01 1.24314380e+00 -5.72852306e-02 -5.31436145e-01
1.22795977e-01 -2.58759055e-02 -5.10812700e-01 -3.38412374e-01
2.95418978e-01 -6.21733069e-01 -1.03343296e+00 6.62196934e-01
2.06567034e-01 3.54294717e-01 7.54645884e-01 -1.80946243e+00
-3.35321665e-01 5.95046401e-01 -1.66071728e-01 1.75891250e-01
2.83944830e-02 4.28802460e-01 1.22264206e+00 -5.56083024e-01
6.72179401e-01 1.19239032e+00 5.71834028e-01 3.73837143e-01
-6.86767042e-01 -2.32906833e-01 4.33743179e-01 8.96245420e-01
-1.20181274e+00 -5.94257891e-01 4.61475909e-01 -6.40173018e-01
1.03442013e+00 3.50617915e-01 4.00990725e-01 9.57942307e-01
5.60682654e-01 1.26880431e+00 2.93844253e-01 3.86144102e-01
2.12652728e-01 -2.62066573e-01 1.25267301e-02 8.04687977e-01
-2.60106862e-01 1.80439681e-01 -5.62244236e-01 -1.80132985e-01
6.81186020e-02 6.60774559e-02 1.45779047e-02 -1.61893040e-01
-1.80177152e+00 7.79680312e-01 1.97474018e-01 -1.89619169e-01
-6.07693911e-01 7.00455487e-01 5.25663793e-01 1.50466353e-01
5.07044077e-01 -8.68101045e-02 -1.78798735e-01 -6.47483170e-01
-6.22837424e-01 6.33015037e-01 6.59510732e-01 1.19443500e+00
7.92156339e-01 -6.08682856e-02 -4.44519162e-01 3.31910759e-01
3.65156233e-01 7.67546177e-01 3.72245848e-01 -1.19752872e+00
7.84236908e-01 1.25374511e-01 3.60469759e-01 -9.26793337e-01
-5.18102705e-01 -1.86955512e-01 -9.26649690e-01 4.22731740e-03
1.51543826e-01 -2.60886490e-01 -2.63863802e-01 1.76364028e+00
2.97336042e-01 8.33300352e-01 4.68101770e-01 8.27698171e-01
4.00103360e-01 8.92555475e-01 5.16959392e-02 -1.27796277e-01
1.20008457e+00 -1.40526891e+00 -5.15380323e-01 5.44753484e-02
1.07807064e+00 -3.87062460e-01 5.96438706e-01 1.10704139e-01
-9.74867344e-01 -6.12782538e-01 -9.14784133e-01 -5.56015633e-02
-1.69580430e-01 1.08586952e-01 5.85633636e-01 3.01752806e-01
-9.24867988e-01 4.20983225e-01 -1.17052221e+00 -6.23190822e-03
7.63505027e-02 -3.54071031e-03 -2.28855759e-01 -1.07503429e-01
-1.05578947e+00 9.41127002e-01 3.19908708e-02 -9.27466005e-02
-9.49653864e-01 -8.52063715e-01 -8.00317645e-01 -1.10826753e-01
5.42404391e-02 -6.62912846e-01 1.17086315e+00 -1.32706493e-01
-1.55270410e+00 1.86989848e-02 -6.33167267e-01 -9.56130981e-01
7.29153454e-01 -1.54183567e-01 -7.01047301e-01 1.01983582e-03
2.96027422e-01 7.19212353e-01 8.14448059e-01 -6.65096521e-01
-1.25894701e+00 -1.87035203e-01 6.22031018e-02 3.64788175e-01
-9.42777619e-02 -4.16653275e-01 -3.92672300e-01 -2.62540609e-01
-3.11023027e-01 -1.23516119e+00 -5.22105396e-01 -3.41266304e-01
-2.64606982e-01 -4.81382102e-01 8.68338346e-01 -5.05445719e-01
1.02972817e+00 -2.25191712e+00 2.54623502e-01 -1.14659227e-01
2.99395680e-01 -7.83449039e-02 -4.30821657e-01 6.80181444e-01
4.82676595e-01 -3.29613119e-01 1.60771664e-02 -8.57405543e-01
5.85375190e-01 1.51669815e-01 -8.31943929e-01 6.62710369e-01
4.93404120e-02 1.13841963e+00 -1.06960726e+00 -1.92322075e-01
2.85244703e-01 2.39439860e-01 -6.13270998e-01 9.04879421e-02
-4.69601065e-01 3.67324829e-01 -4.21379596e-01 8.17253515e-02
6.92927301e-01 -1.08480699e-01 -1.29881606e-01 -1.18581340e-01
-4.96114403e-01 2.81724602e-01 -8.82843137e-01 1.59965825e+00
-3.71960074e-01 7.66617358e-01 -4.23994094e-01 -6.39790058e-01
6.92284763e-01 2.53536463e-01 8.62520158e-01 -7.07404971e-01
-5.90822458e-01 2.11047143e-01 -2.51017183e-01 -5.71121633e-01
1.15071833e+00 3.82654488e-01 -2.12899506e-01 5.21564543e-01
-3.98718745e-01 1.56182721e-01 2.67152667e-01 3.07483673e-01
1.22788012e+00 1.57028705e-01 -2.46215314e-01 6.87337071e-02
4.20324713e-01 1.84309393e-01 5.77897847e-01 6.80579484e-01
-3.77751440e-01 2.55634636e-01 3.57064635e-01 -7.30727494e-01
-1.24510288e+00 -8.91274154e-01 2.10980058e-01 1.05659330e+00
2.89672971e-01 -6.28785312e-01 -3.51629496e-01 -6.87594175e-01
1.62598744e-01 9.49460924e-01 -3.94970059e-01 -7.36235604e-02
-9.04749513e-01 -7.44578242e-01 6.78704798e-01 3.71425420e-01
3.26349974e-01 -5.64531803e-01 -6.43847287e-01 4.23265904e-01
-5.18826067e-01 -1.45862293e+00 -6.65002227e-01 -4.01691794e-01
-3.72811407e-01 -7.63839781e-01 -6.15297258e-01 -4.16815460e-01
2.82289702e-02 4.75828886e-01 7.01471150e-01 -3.56695294e-01
2.20136017e-01 5.42937458e-01 -2.54150122e-01 -2.81111479e-01
-5.12623072e-01 3.73648584e-01 1.75405607e-01 3.74068826e-01
2.70923167e-01 -5.46732783e-01 -7.33197331e-01 3.08018476e-01
-4.40899760e-01 4.07923490e-01 4.84736949e-01 4.38274264e-01
5.28818607e-01 2.37881113e-03 8.75025153e-01 -3.13398480e-01
3.00391376e-01 -1.02995121e+00 -7.11049318e-01 1.05303556e-01
-3.63948524e-01 1.41753018e-01 8.08903217e-01 -4.34178635e-02
-8.10274839e-01 1.00603007e-01 -3.28707933e-01 -2.54923224e-01
-2.39713773e-01 3.41760844e-01 1.89061314e-01 3.16912860e-01
1.68776959e-01 5.79364002e-01 4.70431224e-02 1.22131128e-02
7.48584688e-01 5.03050923e-01 6.28263533e-01 -2.00678170e-01
6.01226807e-01 8.65700841e-01 3.24164301e-01 -7.90701449e-01
-3.42753321e-01 -5.26869416e-01 -4.51137662e-01 -4.51378554e-01
1.01831639e+00 -1.32423794e+00 -1.12819493e+00 4.40215856e-01
-1.43359900e+00 -5.68353057e-01 -9.91864130e-02 8.42013657e-01
-9.47045505e-01 3.92346621e-01 -4.40856665e-01 -6.92016661e-01
3.18349265e-02 -1.34965622e+00 1.24649811e+00 -2.26462975e-01
1.47443458e-01 -1.00370288e+00 4.09721792e-01 2.94813931e-01
3.08588386e-01 -1.65309422e-02 7.56523371e-01 -5.02175152e-01
-1.25465655e+00 -1.53047040e-01 -4.85203378e-02 -2.03170478e-01
-2.62353450e-01 -1.12262830e-01 -6.26506686e-01 -2.25490838e-01
-3.80763561e-01 1.07893303e-01 1.06848371e+00 2.32869044e-01
7.31067896e-01 -2.93183565e-01 -5.86174488e-01 5.65926492e-01
1.10230052e+00 -1.16540007e-01 5.88018775e-01 1.55177414e-01
9.56763983e-01 5.38953066e-01 7.67337143e-01 6.94224000e-01
1.54718602e+00 9.11869586e-01 6.62548900e-01 5.93977153e-01
-1.14459664e-01 -3.86908203e-01 8.68210018e-01 1.14089024e+00
6.04373170e-03 -5.87959290e-01 -8.33317935e-01 7.58835256e-01
-2.41746283e+00 -1.21162021e+00 -4.70223546e-01 1.90862644e+00
7.81700686e-02 -1.93814486e-01 3.98241907e-01 -5.63694596e-01
3.56445283e-01 6.12782121e-01 -6.65418446e-01 -1.77587032e-01
-1.47360474e-01 -7.15441763e-01 8.33925426e-01 7.22531438e-01
-1.20839155e+00 8.35810721e-01 5.52896166e+00 8.32118511e-01
-8.56732190e-01 3.12553793e-01 3.49603921e-01 -9.57112759e-02
-5.48089802e-01 -1.10357299e-01 -1.20183635e+00 6.78234160e-01
1.48080134e+00 -2.92310447e-01 3.75305712e-01 6.84767008e-01
4.15802389e-01 9.72888395e-02 -1.15939641e+00 7.46635675e-01
-3.21616344e-02 -1.74147415e+00 3.01662255e-02 2.90476054e-01
7.74214506e-01 8.48276973e-01 1.85067400e-01 3.95709544e-01
3.71069252e-01 -7.32805431e-01 8.55427682e-01 1.02267766e+00
2.39090860e-01 -9.16296005e-01 5.45185566e-01 7.57977724e-01
-1.44407117e+00 -2.00157017e-01 -4.41235274e-01 -2.10482404e-02
8.02604854e-01 3.05801690e-01 -8.07909071e-01 8.30149353e-01
8.80851001e-02 1.12045705e+00 -1.79996744e-01 9.26820040e-01
3.55274856e-01 4.73714173e-01 -2.50218600e-01 -7.29523823e-02
4.48080182e-01 -2.41313294e-01 8.82507443e-01 1.21430695e+00
6.96723163e-01 1.70987751e-02 2.17673466e-01 3.86413753e-01
1.29566923e-01 -1.59272447e-01 -7.52385557e-01 1.42392159e-01
4.77465004e-01 9.43255723e-01 -8.44374523e-02 -5.05857706e-01
-6.18840754e-01 1.06333160e+00 3.88213247e-01 4.37457323e-01
-1.09961426e+00 5.52751310e-02 1.09259856e+00 -3.26073140e-01
5.95994115e-01 -7.42946088e-01 -3.99856232e-02 -1.08540928e+00
9.45482552e-02 -5.39596558e-01 7.11465627e-02 -3.28750938e-01
-8.91150773e-01 5.83868146e-01 -1.13190562e-01 -1.50555313e+00
-7.72010565e-01 -3.04966152e-01 -5.65957367e-01 4.68601972e-01
-1.81106985e+00 -1.35523200e+00 -7.14044261e-04 4.07899141e-01
6.12164319e-01 -4.88810360e-01 8.09436619e-01 4.57928121e-01
-4.87598091e-01 4.25203711e-01 4.92444158e-01 -2.22499117e-01
4.45246696e-01 -1.09023201e+00 1.11281657e+00 7.05040634e-01
2.97608644e-01 -6.33131787e-02 5.58267772e-01 -5.96718609e-01
-1.96412420e+00 -1.72423291e+00 1.05669153e+00 -6.01921558e-01
7.77377367e-01 -1.54268518e-01 -5.74020505e-01 8.78011763e-01
1.95421785e-01 -5.62727489e-02 5.07425308e-01 -2.24469453e-01
-2.12721244e-01 -8.79694819e-02 -5.12145162e-01 8.08910549e-01
8.23079705e-01 -4.79773223e-01 2.68357228e-02 6.53930485e-01
1.08550370e+00 -3.71782809e-01 -1.01861298e+00 2.71433413e-01
6.25861883e-01 -8.49938989e-01 8.74374986e-01 -6.55496240e-01
4.76143628e-01 -4.78202045e-01 -5.51801026e-01 -1.24126172e+00
-4.28352684e-01 -7.83023775e-01 -4.97606933e-01 7.67024338e-01
4.71026301e-01 -6.13457799e-01 8.65871072e-01 4.99997199e-01
-6.38213992e-01 -6.79572523e-01 -1.11836112e+00 -7.57820547e-01
4.36372170e-03 -8.54797959e-01 9.71744537e-01 3.46238643e-01
-1.46181703e-01 5.52845672e-02 -8.17030966e-01 6.89674854e-01
7.65341401e-01 2.00323358e-01 1.21658027e+00 -7.17803299e-01
-3.98018956e-01 -4.17031795e-01 -6.09723151e-01 -1.74722660e+00
4.86471504e-01 -8.70538116e-01 9.58953053e-02 -1.39411032e+00
-4.36386243e-02 -2.70411998e-01 1.05424359e-01 -6.25346899e-02
8.71075913e-02 9.55585763e-03 3.63888383e-01 4.80759799e-01
-1.07835662e+00 1.01928484e+00 1.01286268e+00 -2.30329514e-01
-1.53728016e-02 2.65821695e-01 -3.96680199e-02 5.16164958e-01
7.00626910e-01 -2.81611532e-01 -4.61093247e-01 -7.04313040e-01
2.60487169e-01 3.44634682e-01 6.53175592e-01 -1.10012698e+00
7.31675088e-01 -3.63975316e-01 -4.47708189e-01 -1.05725276e+00
7.79868960e-01 -5.02360582e-01 1.83058515e-01 4.33506876e-01
-3.76725286e-01 3.69284719e-01 -2.00662501e-02 1.03377581e+00
-2.06331477e-01 9.27204415e-02 1.55621707e-01 2.64120191e-01
-1.13046217e+00 8.89272094e-01 -6.53917670e-01 -1.11831516e-01
1.12950158e+00 2.43730515e-01 -6.93663538e-01 -6.30364001e-01
-5.52222550e-01 6.24043405e-01 1.73922375e-01 5.85482001e-01
6.54026687e-01 -1.57287002e+00 -1.12356484e+00 1.39137134e-01
2.75790811e-01 -2.89523870e-01 7.52333999e-01 1.16633785e+00
-2.53220499e-01 5.93075693e-01 -2.72889566e-02 -6.97608173e-01
-8.41760278e-01 4.75745380e-01 2.76831444e-02 -1.99440673e-01
-6.98876679e-01 4.55507666e-01 3.57545875e-02 -3.97892088e-01
-3.74749452e-02 -2.20073894e-01 -1.62261501e-01 -7.24299178e-02
7.37327576e-01 7.42966831e-01 -1.65783852e-01 -1.05175889e+00
-2.07827538e-01 2.40953863e-01 -1.29992947e-01 -3.96301001e-01
1.22874629e+00 -4.54448372e-01 4.06805500e-02 6.86274171e-01
1.35190892e+00 -1.01707511e-01 -1.59028530e+00 -2.10467726e-01
-8.46498236e-02 -1.95728958e-01 -1.77446797e-01 -2.20679164e-01
-5.28918624e-01 8.04900229e-01 2.40894705e-01 2.92408198e-01
4.09575224e-01 -8.04879665e-02 1.29097724e+00 6.18871689e-01
4.55216795e-01 -9.18794036e-01 -2.26287022e-01 8.85175526e-01
8.40133727e-01 -1.25651848e+00 -3.58397841e-01 -8.66473988e-02
-9.71353114e-01 9.86635625e-01 3.72093499e-01 -1.09674670e-02
6.81299925e-01 1.62669029e-02 -3.02301437e-01 9.80487838e-03
-1.47758138e+00 -2.80074000e-01 3.19839597e-01 6.50794327e-01
-9.21567380e-02 5.65993190e-01 1.96267664e-01 3.27668548e-01
-3.80992323e-01 -2.52119541e-01 4.60843056e-01 1.71661496e-01
-5.63788712e-01 -9.59543407e-01 1.79777175e-01 3.24574143e-01
2.36645028e-01 5.93667030e-02 3.54753137e-01 4.15806085e-01
-4.78027947e-02 9.15638745e-01 2.84638435e-01 -7.11862564e-01
2.09165871e-01 -9.81337354e-02 1.03533484e-01 7.49663869e-03
1.66352075e-02 -4.10564154e-01 2.98272163e-01 -8.02903950e-01
-1.98678032e-01 -1.18257165e+00 -1.19163716e+00 -7.34618604e-01
3.52980912e-01 4.06414680e-02 8.38750243e-01 1.07637632e+00
9.74527359e-01 5.17023265e-01 7.37992764e-01 -7.74101257e-01
-5.93811691e-01 -6.75417781e-01 -2.00119123e-01 5.43139353e-02
7.88139224e-01 -3.63320678e-01 9.67993587e-02 -2.96689123e-02] | [5.872568130493164, 0.8277908563613892] |
3d77177a-b524-4dd9-9d93-a1c7a8999859 | revisiting-acceptability-judgements | 2305.14091 | null | https://arxiv.org/abs/2305.14091v2 | https://arxiv.org/pdf/2305.14091v2.pdf | Revisiting Acceptability Judgements | Years have passed since the NLP community has last focused on linguistic acceptability. In this work, we revisit this topic in the context of large language models. We introduce CoLAC - Corpus of Linguistic Acceptability in Chinese, the first large-scale non-English acceptability dataset that is verified by native speakers and comes with two sets of labels. Our experiments show that even the largest InstructGPT model performs only at chance level on CoLAC, while ChatGPT's performance (48.30 MCC) is also way below supervised models (59.03 MCC) and human (65.11 MCC). Through cross-lingual transfer experiments and fine-grained linguistic analysis, we demonstrate for the first time that knowledge of linguistic acceptability can be transferred across typologically distinct languages, as well as be traced back to pre-training. | ['Rui Wang', 'Peng Zhang', 'Jiahui Huang', 'Yina Ma', 'Aini Li', 'Jackie Yan-Ki Lai', 'Weifang Huang', 'Ziyin Zhang', 'Hai Hu'] | 2023-05-23 | null | null | null | null | ['cross-lingual-transfer', 'linguistic-acceptability'] | ['natural-language-processing', 'natural-language-processing'] | [ 6.39361218e-02 4.07919437e-01 -1.08397044e-01 -7.51516163e-01
-1.43360329e+00 -8.30752730e-01 6.35217369e-01 3.78416359e-01
-6.80075228e-01 8.49570870e-01 2.70564318e-01 -7.36517787e-01
7.73489475e-02 -3.64281714e-01 -7.68846154e-01 -1.81854725e-01
1.04699597e-01 7.45437443e-01 7.72016728e-03 -3.76593411e-01
1.84604675e-01 -8.82870033e-02 -1.11578131e+00 7.39799857e-01
1.42296112e+00 6.48795664e-01 -5.35367280e-02 4.03964818e-01
4.99305613e-02 7.17121899e-01 -4.37308192e-01 -1.00121093e+00
-2.15184510e-01 -4.71049584e-02 -1.26366806e+00 -5.26757419e-01
7.28496075e-01 5.42983459e-03 6.39079392e-01 8.99926603e-01
1.42871201e-01 -1.62214860e-01 8.36351395e-01 -8.09386730e-01
-1.38505554e+00 1.15671360e+00 -1.74467638e-01 -3.29209447e-01
8.30365956e-01 6.50797933e-02 1.31348622e+00 -8.20732772e-01
7.92842984e-01 1.66081476e+00 9.12522435e-01 5.79661131e-01
-1.15994895e+00 -3.25098187e-01 2.66223460e-01 -1.32526621e-01
-1.16150188e+00 -1.93629280e-01 3.58782589e-01 -4.99697953e-01
1.36655819e+00 2.49318108e-01 4.63108331e-01 1.17231441e+00
1.83705568e-01 7.27056086e-01 1.75781190e+00 -9.43811953e-01
-3.59757543e-02 3.35429281e-01 3.66041064e-01 5.30589819e-01
4.17147530e-03 1.99906588e-01 -5.21866918e-01 -1.51692275e-02
-7.62388408e-02 -1.06627250e+00 -8.31322279e-03 3.17733496e-01
-9.98253882e-01 9.56307709e-01 2.12512434e-01 4.38957155e-01
2.18484715e-01 -2.20809020e-02 5.73120117e-01 6.99085355e-01
7.36854970e-01 5.42886615e-01 -8.85252833e-01 -2.76445061e-01
-2.39558876e-01 -8.27603936e-02 8.72498572e-01 1.03125024e+00
4.20733452e-01 8.28451244e-04 2.53514916e-01 1.13967681e+00
3.83180559e-01 7.97012866e-01 3.45207214e-01 -8.32457066e-01
4.79926884e-01 4.66819346e-01 -7.53259361e-02 -4.44250137e-01
-2.18982562e-01 3.06133360e-01 -1.79870442e-01 1.08844914e-01
7.75241435e-01 -1.23163469e-01 -5.90836823e-01 1.96657825e+00
7.16691986e-02 -8.17225039e-01 2.23047122e-01 4.56689447e-01
4.94012624e-01 7.39839017e-01 8.26964617e-01 -1.01113893e-01
1.22703350e+00 -6.08950973e-01 -2.65133768e-01 -2.82397360e-01
1.20446885e+00 -9.85814810e-01 1.96206272e+00 7.39911616e-01
-1.10673273e+00 -5.48184752e-01 -7.50633717e-01 -2.79603720e-01
-7.32877672e-01 9.52836573e-02 8.14833879e-01 1.08430910e+00
-1.42970753e+00 2.39681780e-01 -5.15219331e-01 -5.46148598e-01
1.45640761e-01 2.63520062e-01 -4.80317146e-01 -2.33606547e-01
-1.35295010e+00 1.21308076e+00 3.95115137e-01 -5.23984246e-02
-3.76677126e-01 -5.06456912e-01 -1.06804001e+00 -4.27458107e-01
6.54159635e-02 -1.49953499e-01 1.27888799e+00 -1.52582431e+00
-1.58055317e+00 1.33355379e+00 -1.13477066e-01 -1.37311473e-01
3.00105363e-01 -7.08452761e-01 -8.41305196e-01 -2.74903119e-01
1.83931082e-01 7.73113549e-01 2.12845609e-01 -1.08775270e+00
-8.37777019e-01 -2.66587228e-01 1.21253945e-01 2.60197550e-01
-2.39279449e-01 8.61546934e-01 -5.38977683e-02 -4.06794786e-01
-3.19635421e-01 -1.23241985e+00 9.99352187e-02 -3.90642494e-01
-3.22925031e-01 -7.79780209e-01 2.55691171e-01 -7.72972763e-01
1.04088640e+00 -2.00404239e+00 5.17157838e-02 2.13357255e-01
-3.36484462e-01 2.38327429e-01 -1.52932420e-01 3.79330665e-01
-7.07353977e-03 6.75294161e-01 -3.61102641e-01 -3.68976802e-01
4.02867407e-01 1.38329268e-01 -3.99381928e-02 2.77727813e-01
4.13734108e-01 1.07753921e+00 -7.99704492e-01 -5.15199482e-01
7.70643577e-02 3.59022170e-01 -7.00727642e-01 9.12861992e-03
-1.32264748e-01 4.32366371e-01 -1.53397426e-01 5.26526213e-01
5.59600711e-01 3.32279146e-01 4.29858476e-01 1.59557804e-01
-3.41253757e-01 5.25152504e-01 -4.83896464e-01 1.39284575e+00
-9.65849102e-01 4.04543757e-01 -8.43928158e-02 -4.74998295e-01
5.87030649e-01 2.35838860e-01 -4.18913186e-01 -7.77173579e-01
-3.35566024e-03 8.33847225e-01 4.32070017e-01 -2.23970547e-01
5.67174196e-01 -3.00702125e-01 -9.28444624e-01 3.25423062e-01
-1.50112092e-01 -6.31052017e-01 6.22714199e-02 9.39879268e-02
3.41770619e-01 2.19749495e-01 3.72771978e-01 -1.06822383e+00
9.16701734e-01 8.44379440e-02 2.68121034e-01 4.68537986e-01
-1.52175114e-01 2.76774228e-01 4.57648963e-01 -2.33618319e-01
-7.96974182e-01 -1.22750366e+00 -5.71635902e-01 1.55376244e+00
-1.32886589e-01 -1.19821414e-01 -1.08136296e+00 -6.79014742e-01
-1.15591377e-01 1.15039194e+00 -5.04072905e-01 7.74756819e-02
-7.52786160e-01 -4.82771486e-01 6.92242265e-01 5.62715888e-01
4.98444885e-02 -1.44611490e+00 -7.83722401e-02 1.10083684e-01
3.61100249e-02 -1.20038474e+00 -3.08703512e-01 -6.99690953e-02
-3.37463111e-01 -7.60349512e-01 -1.52310356e-01 -1.18875074e+00
3.67555052e-01 -7.91055501e-01 1.43815410e+00 2.62925893e-01
1.17960848e-01 2.22141221e-01 -4.32271451e-01 -5.02015173e-01
-9.26978946e-01 2.71936476e-01 1.50038108e-01 -4.23738271e-01
6.06878817e-01 -1.81236744e-01 5.90320826e-02 5.84060047e-03
-5.34750044e-01 -2.18166083e-01 3.19448948e-01 6.62152290e-01
4.40327913e-01 -7.00606287e-01 9.02155817e-01 -1.59677100e+00
7.80182064e-01 -2.08567187e-01 -6.45785689e-01 4.43028003e-01
-5.10447681e-01 -2.42480978e-01 9.01698232e-01 -1.44259736e-01
-1.65403914e+00 5.29127903e-02 -5.93855083e-01 8.26984167e-01
-5.76980472e-01 7.29272127e-01 -4.03635174e-01 5.05881198e-02
7.42628217e-01 -4.84268963e-01 -5.16567826e-01 -2.29699746e-01
8.60203087e-01 9.26496625e-01 6.15087450e-01 -1.17090046e+00
3.63994479e-01 -2.37097852e-02 -6.97587073e-01 -8.60384405e-01
-1.09061873e+00 -9.36406329e-02 -1.00421309e+00 1.61626451e-02
1.02740622e+00 -8.91704321e-01 -4.77432251e-01 7.60784268e-01
-1.03751981e+00 -8.08773518e-01 -3.26978341e-02 3.76354158e-01
-4.15682256e-01 2.54716843e-01 -1.27791762e+00 -6.91893816e-01
-3.10325533e-01 -8.53726566e-01 1.05134344e+00 -2.25509912e-01
-8.16824436e-01 -1.32703066e+00 1.45346135e-01 2.43092418e-01
1.94484487e-01 8.83570611e-02 1.34965158e+00 -5.75571597e-01
-1.59377158e-01 -1.53688118e-01 -6.65531904e-02 4.37872231e-01
-5.32863736e-02 2.71030366e-01 -1.06439567e+00 -1.51151374e-01
-2.05175608e-01 -1.07364666e+00 4.99330491e-01 2.96055675e-01
8.24758530e-01 -1.72462255e-01 9.52967554e-02 3.73362631e-01
1.30386591e+00 1.24670111e-01 6.73939109e-01 2.92390525e-01
7.81305015e-01 8.52285683e-01 7.53417075e-01 -3.90858114e-01
7.03249574e-01 4.29876536e-01 -3.71421397e-01 -5.12272678e-02
3.22163962e-02 -2.57344872e-01 8.13998938e-01 1.41262150e+00
-4.75963578e-02 -3.75161082e-01 -1.21902740e+00 6.55186355e-01
-1.43372726e+00 -3.59198898e-01 -3.98622215e-01 2.03461432e+00
1.23085165e+00 3.38979363e-01 -3.32369842e-02 -1.51749507e-01
4.25351411e-01 -1.88947752e-01 5.24386726e-02 -1.54045951e+00
-5.45394421e-01 4.12658066e-01 -1.00503430e-01 1.28640318e+00
-1.05579460e+00 1.59738648e+00 6.91400051e+00 8.33825350e-01
-7.80114472e-01 1.30735099e-01 9.76586580e-01 4.20280784e-01
-6.73758805e-01 8.81879330e-02 -7.17732310e-01 1.95775643e-01
1.45481062e+00 -7.70083591e-02 2.68763065e-01 7.23140359e-01
-2.21027687e-01 -3.51916611e-01 -1.41976070e+00 4.31722224e-01
1.46555543e-01 -5.24689376e-01 8.75752792e-03 -2.21631616e-01
9.50509310e-01 3.46746504e-01 2.61589110e-01 7.24533379e-01
7.49785721e-01 -1.25729430e+00 9.39460099e-01 -5.62350042e-02
1.20477366e+00 -9.80734527e-01 7.22677231e-01 1.97668478e-01
-1.10042346e+00 8.19945708e-02 -3.80920947e-01 -2.93447793e-01
3.43010366e-01 4.84211057e-01 -2.89100319e-01 2.15979546e-01
7.54894853e-01 4.71430123e-01 -7.78921664e-01 7.42735863e-02
-6.39497221e-01 9.50457931e-01 -1.93287164e-01 -2.58430302e-01
3.52615297e-01 -3.11925799e-01 4.92190123e-02 1.61866939e+00
1.50198370e-01 8.49279314e-02 1.83086917e-01 5.50430179e-01
-3.65831666e-02 6.80100262e-01 -6.05417252e-01 4.39743139e-02
3.38222772e-01 1.10366392e+00 -2.56843895e-01 -3.99320275e-01
-5.42654157e-01 8.64167273e-01 1.01053834e+00 1.20158613e-01
-6.12104952e-01 -2.06198439e-01 1.43738538e-01 -2.87766457e-01
-5.38743138e-02 -2.46322274e-01 -4.27932292e-01 -9.70201731e-01
-5.68585936e-03 -1.04587829e+00 3.77829790e-01 -6.64451778e-01
-1.72952628e+00 9.59937036e-01 -1.19712628e-01 -5.36162198e-01
-4.58025366e-01 -1.10401750e+00 -4.51103479e-01 1.08583176e+00
-1.44771886e+00 -1.54683125e+00 3.85911375e-01 3.23413849e-01
4.49790716e-01 2.17422340e-02 1.23510909e+00 1.06929801e-01
-3.56137276e-01 9.91249740e-01 -2.60457434e-02 -5.82166910e-02
9.30715740e-01 -1.51483583e+00 6.67161286e-01 6.65923059e-01
-4.98030894e-03 7.95094490e-01 5.20034730e-01 -7.99102247e-01
-9.00842488e-01 -1.02458942e+00 1.61402798e+00 -1.24999321e+00
9.77395117e-01 -6.60306931e-01 -1.15268719e+00 9.21364844e-01
6.42663598e-01 -4.74795073e-01 6.42684698e-01 7.21093059e-01
-5.19897997e-01 1.72812968e-01 -1.14640558e+00 7.71514714e-01
1.09839749e+00 -1.12496448e+00 -8.84682536e-01 2.87458807e-01
9.70576406e-01 -2.36804202e-01 -1.13384843e+00 3.19737613e-01
5.89125037e-01 -8.62594604e-01 5.86357057e-01 -5.56669474e-01
5.50909758e-01 1.62955970e-01 -3.09929490e-01 -1.41410124e+00
1.46627007e-03 -5.07419705e-01 6.49164021e-01 1.57313037e+00
1.19265425e+00 -7.71148682e-01 2.46248856e-01 9.17799175e-01
-5.77089548e-01 -5.48442662e-01 -8.08642030e-01 -8.24889302e-01
9.90314782e-01 -8.15218747e-01 4.04941231e-01 1.00607538e+00
4.31993276e-01 5.95212758e-01 1.45790912e-02 -3.79426442e-02
3.67544919e-01 7.85496682e-02 3.43100518e-01 -1.10594451e+00
-2.55318642e-01 -2.30374932e-01 8.45183656e-02 -5.76749206e-01
7.72001982e-01 -1.03956068e+00 4.33448344e-01 -1.04890299e+00
2.02995837e-01 -6.87416852e-01 -5.48538417e-02 3.82266760e-01
-3.97216082e-01 2.49700323e-01 3.35277654e-02 -1.98647037e-01
-5.79213917e-01 3.18087459e-01 7.37200499e-01 2.53032625e-01
-1.40321389e-01 -4.73728240e-01 -8.87216568e-01 1.10367310e+00
1.21412826e+00 -3.29483539e-01 -3.90876234e-01 -7.80295491e-01
4.02497679e-01 -4.36464012e-01 -2.21478477e-01 -7.20867753e-01
-2.10265175e-01 -2.52067298e-01 3.90276350e-02 -1.95752025e-01
1.08253015e-02 -2.98941702e-01 -2.77386576e-01 3.03379655e-01
-5.44399142e-01 5.06340027e-01 6.02078855e-01 1.05342530e-01
-1.70808986e-01 -7.13830069e-02 7.16845274e-01 -4.64424584e-03
-9.29184675e-01 -6.78617507e-02 -6.86186016e-01 6.77673340e-01
7.04941094e-01 1.95346788e-01 -5.55810928e-01 -1.11054204e-01
-5.15985548e-01 2.27501299e-02 7.52572656e-01 4.52990800e-01
2.11733297e-01 -1.10594308e+00 -8.90108109e-01 8.21903124e-02
4.33002234e-01 -3.77860695e-01 5.78185245e-02 4.95889634e-01
-5.36899507e-01 6.63864255e-01 5.82409687e-02 -5.80173910e-01
-1.07606089e+00 4.07584071e-01 8.14951435e-02 -4.47232425e-02
-1.33617491e-01 9.77598369e-01 2.12463781e-01 -1.24845314e+00
-2.51929641e-01 -4.23561901e-01 -9.44258198e-02 -3.73150378e-01
1.96297407e-01 -1.14522643e-01 -3.78008932e-02 -1.09159720e+00
-4.23108101e-01 8.51744533e-01 -1.76446319e-01 -3.98948848e-01
1.01354277e+00 -2.42629334e-01 -4.44429696e-01 7.80188501e-01
1.16126394e+00 6.70326412e-01 -7.08175242e-01 5.57806008e-02
6.14794374e-01 -1.54968292e-01 -5.32837689e-01 -1.22144806e+00
-3.06572944e-01 9.28398252e-01 2.78915107e-01 -1.12471089e-01
7.82218099e-01 4.01919991e-01 6.78600311e-01 5.64918041e-01
4.36602503e-01 -1.60808754e+00 -2.54462302e-01 8.67465556e-01
8.85918200e-01 -1.29041338e+00 -2.99632430e-01 -8.36738706e-01
-1.02761269e+00 5.00129342e-01 7.82546878e-01 -9.91038159e-02
5.80420375e-01 1.85576007e-01 5.24623036e-01 6.82304613e-03
-8.86695981e-01 8.61107558e-02 3.64606291e-01 9.23764765e-01
1.11957490e+00 6.85198009e-01 -5.90512455e-01 7.79725194e-01
-7.04703987e-01 -3.56291115e-01 3.90588820e-01 5.45741379e-01
-4.21325862e-01 -1.47101414e+00 4.17638496e-02 1.87091663e-01
-6.54221177e-01 -7.38101900e-01 -7.52022266e-01 1.12432837e+00
7.25919828e-02 1.03079867e+00 -2.24842466e-02 -1.08081540e-02
4.43530619e-01 1.65635705e-01 3.03022861e-01 -8.59144747e-01
-7.30872095e-01 -2.91226478e-03 8.49133313e-01 -4.00506824e-01
-3.48436862e-01 -8.50554526e-01 -1.28887689e+00 -3.97674143e-01
-2.59796083e-01 2.55430162e-01 3.40975404e-01 8.68872404e-01
-1.99259698e-01 -8.22273195e-02 2.74808019e-01 -3.51923883e-01
-5.13383746e-01 -9.01514769e-01 -5.74682832e-01 5.88101268e-01
-1.15877174e-01 3.61340120e-02 -5.91766655e-01 1.03739910e-01] | [10.841387748718262, 9.669989585876465] |
4ca7fb35-e24a-43f5-95ac-37b67a7f3dad | chupa-carving-3d-clothed-humans-from-skinned | 2305.11870 | null | https://arxiv.org/abs/2305.11870v2 | https://arxiv.org/pdf/2305.11870v2.pdf | Chupa: Carving 3D Clothed Humans from Skinned Shape Priors using 2D Diffusion Probabilistic Models | We propose a 3D generation pipeline that uses diffusion models to generate realistic human digital avatars. Due to the wide variety of human identities, poses, and stochastic details, the generation of 3D human meshes has been a challenging problem. To address this, we decompose the problem into 2D normal map generation and normal map-based 3D reconstruction. Specifically, we first simultaneously generate realistic normal maps for the front and backside of a clothed human, dubbed dual normal maps, using a pose-conditional diffusion model. For 3D reconstruction, we ``carve'' the prior SMPL-X mesh to a detailed 3D mesh according to the normal maps through mesh optimization. To further enhance the high-frequency details, we present a diffusion resampling scheme on both body and facial regions, thus encouraging the generation of realistic digital avatars. We also seamlessly incorporate a recent text-to-image diffusion model to support text-based human identity control. Our method, namely, Chupa, is capable of generating realistic 3D clothed humans with better perceptual quality and identity variety. | ['Hanbyul Joo', 'Daesik Kim', 'Sookwan Han', 'Myunggi Lee', 'Kwangho Lee', 'Patrick Kwon', 'Byungjun Kim'] | 2023-05-19 | null | null | null | null | ['3d-reconstruction'] | ['computer-vision'] | [ 1.31478846e-01 2.14322045e-01 6.28364503e-01 -6.81512132e-02
-4.22407806e-01 -4.84258085e-01 6.08471990e-01 -3.86930078e-01
2.34657094e-01 5.38079858e-01 3.18845183e-01 3.02026361e-01
4.94201869e-01 -1.08013427e+00 -6.84034586e-01 -4.48457092e-01
3.18349540e-01 7.68246889e-01 2.35137835e-01 -5.84225476e-01
-2.70401269e-01 6.07998848e-01 -1.37915075e+00 1.45438030e-01
9.44525301e-01 8.59010816e-01 -8.51065367e-02 6.97541475e-01
6.38672560e-02 3.37822258e-01 -5.13386786e-01 -7.70945072e-01
4.05201763e-01 -8.34058523e-01 -3.78555536e-01 4.05012906e-01
5.47117114e-01 -3.89684796e-01 -6.64459318e-02 8.81891370e-01
7.76364744e-01 1.72194719e-01 7.42318273e-01 -1.11199772e+00
-7.22329438e-01 1.91080526e-01 -8.39090824e-01 -9.46112454e-01
8.40493679e-01 3.99699152e-01 4.90321308e-01 -8.22467327e-01
1.19129586e+00 1.60785913e+00 8.72281134e-01 9.73663986e-01
-1.57909477e+00 -7.06353366e-01 -3.33401263e-02 -3.17987323e-01
-1.32936358e+00 -2.70260900e-01 1.17421436e+00 -5.12289405e-01
7.83587173e-02 4.01104271e-01 1.44047952e+00 1.63194168e+00
1.88352525e-01 6.22537851e-01 1.24641061e+00 -3.05938780e-01
1.41093463e-01 -2.48205349e-01 -8.86991322e-01 8.64687204e-01
-2.12718815e-01 5.98916598e-02 -6.00463033e-01 -3.74117613e-01
1.50573969e+00 -4.12541986e-01 -1.61705971e-01 -6.14314497e-01
-1.53302705e+00 5.65629005e-01 1.41906664e-01 -4.25401002e-01
-7.25324690e-01 2.13224456e-01 1.35091230e-01 -2.88221836e-01
7.07973897e-01 3.09195817e-01 7.94674903e-02 -1.38823576e-02
-7.33848333e-01 9.72194374e-01 8.47444534e-01 1.06178415e+00
5.40312350e-01 2.62961119e-01 -3.49022627e-01 1.00503111e+00
2.94747919e-01 8.67419422e-01 -2.54216343e-01 -1.45404899e+00
-7.79945180e-02 4.75666642e-01 1.94592625e-01 -1.22244191e+00
-4.03409749e-02 -4.79338840e-02 -1.15787518e+00 6.04917824e-01
3.21663052e-01 -2.81767070e-01 -9.68399584e-01 1.65020430e+00
9.64229763e-01 3.52901407e-02 -3.39784920e-01 1.24987471e+00
6.83968306e-01 5.94615221e-01 5.36898077e-02 -2.38181893e-02
1.26349485e+00 -7.23099232e-01 -8.38889837e-01 1.48389637e-01
-2.49175608e-01 -9.65435684e-01 1.06179595e+00 2.99230933e-01
-1.57727873e+00 -5.30835986e-01 -7.75165319e-01 -1.24158233e-01
3.63702655e-01 -1.74697667e-01 3.17706019e-01 4.09881502e-01
-9.48305249e-01 6.11864507e-01 -6.68333292e-01 -2.15410307e-01
3.10117990e-01 -7.89645016e-02 -2.67444253e-01 -5.94381988e-03
-1.24304497e+00 6.62257731e-01 -2.70604104e-01 -7.11254552e-02
-9.44654286e-01 -7.81138420e-01 -1.00137341e+00 -5.62946796e-01
1.03898026e-01 -1.28966773e+00 9.88002121e-01 -8.29505086e-01
-2.04686451e+00 1.04705024e+00 9.40158367e-02 9.69831422e-02
1.20211887e+00 7.69073889e-02 -9.82233956e-02 -7.09925890e-02
7.97701553e-02 9.56834257e-01 1.01267552e+00 -1.80061817e+00
-5.36347106e-02 -2.03978345e-01 -2.29176909e-01 4.33505446e-01
2.15901211e-01 -2.19032481e-01 -7.19570398e-01 -1.25037813e+00
1.52768180e-01 -1.08076489e+00 -4.27754641e-01 6.25150979e-01
-5.45502841e-01 2.72163212e-01 7.92825520e-01 -9.75809872e-01
8.19185734e-01 -2.01850176e+00 6.31462276e-01 5.04000127e-01
4.32771206e-01 -1.40235782e-01 -8.61392915e-03 2.00559482e-01
2.70800859e-01 -7.92431161e-02 -3.42888266e-01 -6.85487747e-01
1.26505002e-01 1.27714440e-01 9.24120098e-02 4.04087603e-01
8.14825445e-02 9.02457416e-01 -9.28922236e-01 -6.77338064e-01
2.36812919e-01 9.37766671e-01 -8.09649169e-01 3.32180589e-01
-5.01917720e-01 1.08709180e+00 -3.45173419e-01 7.12388694e-01
8.20040762e-01 7.99897760e-02 8.68177265e-02 -5.01423895e-01
-1.00940809e-01 -4.59253132e-01 -1.24940360e+00 1.99645448e+00
-3.42972964e-01 1.93505734e-01 6.22886896e-01 1.59061641e-01
1.18486869e+00 2.73567915e-01 6.00862443e-01 -5.46861291e-01
1.71860635e-01 9.82142389e-02 -4.32853848e-01 -2.07601219e-01
5.48339963e-01 -4.05682951e-01 -1.05755001e-01 3.49068403e-01
-3.45064789e-01 -8.07756424e-01 -2.42609933e-01 4.46826294e-02
5.36186635e-01 8.10314178e-01 -3.80244672e-01 -2.57167637e-01
1.99370965e-01 9.67599824e-03 5.96483350e-01 1.22876838e-01
1.28271997e-01 1.32828093e+00 4.30425048e-01 -4.00308013e-01
-1.72226560e+00 -1.34376764e+00 1.39079571e-01 5.30664921e-01
4.85581934e-01 -3.65786403e-01 -1.22259820e+00 -2.66308308e-01
6.34108782e-02 5.74922204e-01 -8.31125200e-01 -3.44082410e-03
-7.28316903e-01 -4.11698282e-01 5.87755799e-01 9.13998932e-02
6.14528835e-01 -9.31562245e-01 -4.41602081e-01 2.51950622e-01
-4.04020399e-01 -8.14959764e-01 -9.80778873e-01 -7.61297762e-01
-3.58245164e-01 -7.88868487e-01 -1.37003255e+00 -6.28621876e-01
6.70648754e-01 -3.49020183e-01 1.18389690e+00 -7.91797563e-02
-3.50385070e-01 3.62285048e-01 -2.71591008e-01 -2.60342300e-01
-9.16833758e-01 -3.55661213e-01 1.88613072e-01 3.79079252e-01
-6.98798954e-01 -6.98402107e-01 -1.00580025e+00 6.32047176e-01
-7.44774163e-01 8.44154894e-01 8.68108422e-02 6.50742471e-01
9.48425174e-01 -2.77687907e-01 1.44097060e-01 -6.13218188e-01
8.53485107e-01 -1.08447023e-01 -3.23712021e-01 1.36035774e-02
-3.54090407e-02 -2.40226343e-01 3.62913817e-01 -8.41563165e-01
-1.25124800e+00 2.16725171e-01 -5.83072722e-01 -6.39346421e-01
4.13892269e-02 -2.07916602e-01 -3.09534937e-01 -1.20562375e-01
7.78625369e-01 9.75555107e-02 5.00990689e-01 -3.72005016e-01
5.80324113e-01 1.99385509e-01 5.49119830e-01 -9.81512606e-01
8.95047188e-01 5.82714140e-01 1.45536913e-02 -8.49847734e-01
-2.69245714e-01 5.07876158e-01 -5.65112770e-01 -8.44428599e-01
1.17193902e+00 -7.42146730e-01 -9.99764740e-01 8.65894377e-01
-1.36292696e+00 -7.59363770e-01 -6.23522937e-01 7.41098523e-02
-8.30494583e-01 3.75260413e-01 -8.41721058e-01 -7.87835836e-01
-4.41453755e-01 -1.31360972e+00 1.45104027e+00 4.54417290e-03
-7.59342432e-01 -8.55937898e-01 1.18439145e-01 2.17505962e-01
3.95869642e-01 1.22234321e+00 7.70715237e-01 5.53691030e-01
-4.22391653e-01 5.44361537e-03 3.91898416e-02 1.39349058e-01
1.00263283e-01 1.40476316e-01 -6.12778306e-01 -5.88948540e-02
-3.19549650e-01 -2.77101845e-01 1.42267287e-01 3.05370927e-01
7.76884019e-01 -3.07694793e-01 -1.14821270e-01 7.73307204e-01
8.50221872e-01 -2.06892297e-01 6.10827446e-01 -1.75630927e-01
1.12318575e+00 8.21998060e-01 5.16205966e-01 7.75872827e-01
5.12202919e-01 9.55165446e-01 2.81184226e-01 -5.16050756e-01
-7.64227867e-01 -8.24546695e-01 3.88868041e-02 8.07510316e-01
-5.92244446e-01 -2.01050788e-02 -6.56861126e-01 1.82122514e-01
-1.58205259e+00 -7.95935631e-01 -2.78515756e-01 1.99486947e+00
9.97594953e-01 -2.07522631e-01 4.86304283e-01 -2.07831010e-01
8.89905989e-01 5.77707253e-02 -6.01739645e-01 -1.09125607e-01
-1.79816425e-01 3.51479501e-02 7.32007101e-02 6.54639482e-01
-4.50741380e-01 1.04291725e+00 6.25221825e+00 8.14747870e-01
-9.35278594e-01 -5.83715439e-02 6.48852527e-01 -4.10975888e-03
-8.65847290e-01 -2.13574454e-01 -4.25503165e-01 3.65181297e-01
1.04742363e-01 -1.33151859e-01 4.20827746e-01 6.43321276e-01
5.01406968e-01 1.33300290e-01 -6.63141727e-01 1.02230835e+00
5.81994392e-02 -1.43258059e+00 5.00990152e-01 1.94089293e-01
1.14809608e+00 -7.76778579e-01 3.58854160e-02 -2.23024100e-01
6.16053581e-01 -8.85102689e-01 1.40713346e+00 8.16918135e-01
1.40515161e+00 -8.50642741e-01 6.73448145e-02 2.65975296e-01
-1.22535920e+00 7.10001707e-01 7.29581565e-02 3.31316561e-01
7.32858360e-01 4.42303658e-01 -3.96922529e-01 4.90190089e-01
5.51766336e-01 4.48423177e-01 -8.81960019e-02 6.52304947e-01
-5.15018515e-02 -1.51204001e-02 -2.10619673e-01 2.04745412e-01
-1.90022618e-01 -5.75835884e-01 6.75137579e-01 9.78218377e-01
4.52566057e-01 2.21216097e-01 2.39441365e-01 1.32693541e+00
-5.59437880e-03 8.63427073e-02 -4.66628820e-01 3.85038733e-01
4.05753046e-01 1.04676437e+00 -5.60045660e-01 -2.01077610e-01
2.49601379e-01 1.68784201e+00 2.90627796e-02 3.56361389e-01
-1.04506350e+00 -6.90078139e-02 7.73797274e-01 6.62218571e-01
-2.86111832e-01 -4.81019467e-01 -3.26787412e-01 -1.07512760e+00
-1.24414682e-01 -1.07761860e+00 -2.30271280e-01 -1.01742935e+00
-1.36001468e+00 7.44648576e-01 -6.74088076e-02 -1.04758751e+00
-5.32589713e-03 -1.02952272e-01 -4.51772213e-01 1.15169048e+00
-7.37444818e-01 -1.62444711e+00 -5.63439965e-01 6.60548925e-01
5.64464927e-01 2.10420460e-01 8.09730291e-01 2.05726653e-01
-2.14950591e-01 4.31434810e-01 -4.54716623e-01 -6.94572031e-02
7.70447552e-01 -1.03543556e+00 1.03998065e+00 4.07858014e-01
-5.01559019e-01 3.14810187e-01 8.07806969e-01 -1.25362980e+00
-1.52707732e+00 -1.12395906e+00 2.95988947e-01 -4.77043986e-01
1.48282096e-01 -5.50123453e-01 -6.87819064e-01 3.60530585e-01
4.48222011e-02 -2.51608759e-01 3.25385213e-01 -4.30006862e-01
-5.54955378e-03 1.59480423e-01 -1.46519363e+00 1.19197798e+00
1.51716638e+00 -1.56297222e-01 4.94913571e-02 1.06160663e-01
6.06447637e-01 -1.01955760e+00 -1.18228209e+00 5.34269333e-01
9.56472456e-01 -9.45812285e-01 1.09730554e+00 9.74639058e-02
4.44067150e-01 -4.04647887e-01 1.59667030e-01 -1.58649349e+00
-4.30767775e-01 -1.15928817e+00 -6.10314589e-03 1.23145652e+00
1.59347549e-01 -1.07783198e-01 9.91378844e-01 7.34208047e-01
2.08459422e-02 -6.89565539e-01 -6.31960511e-01 -4.12924886e-01
-5.65556884e-02 -2.57023394e-01 8.35008979e-01 1.06649053e+00
-3.65960211e-01 3.23604271e-02 -8.54688466e-01 -1.81898534e-01
1.03976059e+00 -5.23564145e-02 1.09445000e+00 -1.06370330e+00
-1.31427437e-01 -3.97542238e-01 -7.20147043e-02 -1.31022000e+00
-7.18013346e-02 -6.43852949e-01 1.30205303e-01 -1.66234720e+00
-1.34774864e-01 -7.85646319e-01 8.34282219e-01 -5.48825599e-02
-9.16378424e-02 5.92037439e-01 2.19121143e-01 1.65259972e-01
-3.68907340e-02 9.15168524e-01 2.29852748e+00 1.05550893e-01
-1.99768573e-01 -1.20857373e-01 -3.31553102e-01 8.54965985e-01
4.54710484e-01 -1.83381010e-02 -2.26944447e-01 -5.19575000e-01
1.08928785e-01 4.30863649e-01 4.55456257e-01 -9.65795815e-01
1.90842096e-02 -3.78185153e-01 7.09292412e-01 -4.19516087e-01
7.73059428e-01 -6.44539416e-01 9.10055816e-01 3.50316495e-01
-2.07438350e-01 2.27303639e-01 6.98253810e-02 4.98495817e-01
2.89501756e-01 5.68017304e-01 1.19453478e+00 -3.74393404e-01
-1.33070603e-01 7.32391059e-01 -2.82814175e-01 5.02066128e-02
1.02269745e+00 -3.65084440e-01 3.41286719e-01 -6.87071681e-01
-8.78831744e-01 1.33361414e-01 1.22290027e+00 3.42750460e-01
9.30556357e-01 -1.79411530e+00 -1.00661552e+00 4.49385375e-01
-9.69139300e-03 3.23820233e-01 5.21518052e-01 4.09948468e-01
-1.05403531e+00 -6.46576405e-01 -3.82433146e-01 -5.54143012e-01
-1.13493824e+00 1.54820859e-01 4.49385792e-01 2.36649543e-01
-9.40912843e-01 7.96881914e-01 2.41950244e-01 -6.68739557e-01
1.42147597e-02 1.54396161e-01 3.12884837e-01 -2.49417990e-01
3.01851749e-01 5.76166451e-01 -4.76363450e-01 -9.48852003e-01
7.87488893e-02 9.79089618e-01 4.97699887e-01 -5.55335164e-01
1.09133148e+00 -1.82345465e-01 -9.51468796e-02 1.06992170e-01
7.25787103e-01 5.16392350e-01 -1.81257904e+00 1.64261684e-01
-7.16787159e-01 -6.22967005e-01 -4.16832983e-01 -7.14275062e-01
-1.12675762e+00 5.38280308e-01 2.51037776e-01 -1.02496929e-01
7.46718884e-01 -1.62527233e-01 1.21651280e+00 -5.17624021e-01
5.10688722e-01 -9.32698607e-01 2.08582506e-01 3.19165349e-01
1.38878059e+00 -6.72357559e-01 -9.93644297e-02 -8.20468485e-01
-8.71693134e-01 8.41187894e-01 7.28513598e-01 -1.51068151e-01
3.57055098e-01 4.19190675e-01 3.02621007e-01 -1.87290043e-01
-2.67520130e-01 2.98549622e-01 3.68139535e-01 8.93737257e-01
2.56473839e-01 1.23326413e-01 -1.11277672e-02 2.07139999e-01
-4.28212255e-01 -1.28086418e-01 2.63973415e-01 6.21528447e-01
3.02626509e-02 -1.27474773e+00 -7.52526641e-01 -7.84630328e-02
-5.17957062e-02 1.73485920e-01 -5.48238456e-01 4.53282267e-01
3.39105457e-01 6.05083704e-01 -1.01493247e-01 -5.93188047e-01
8.03670585e-01 -2.30044723e-01 7.23440230e-01 -3.51380795e-01
-4.74358290e-01 2.73824304e-01 2.26143613e-01 -5.57881594e-01
-4.74518687e-02 -5.21307945e-01 -1.07483864e+00 -9.18299437e-01
1.03173591e-01 -1.92753106e-01 6.85881078e-01 2.77478725e-01
4.81531501e-01 4.69771534e-01 4.67565179e-01 -1.60330975e+00
-1.55399302e-02 -5.99930465e-01 -6.25216186e-01 8.97247374e-01
-4.61352468e-02 -6.86639190e-01 -6.07522530e-03 2.99024045e-01] | [12.08906078338623, -0.6892684698104858] |
38f9fd8a-ccc4-44ef-a343-8aa0334f8756 | character-region-attention-for-text-spotting | 2007.09629 | null | https://arxiv.org/abs/2007.09629v1 | https://arxiv.org/pdf/2007.09629v1.pdf | Character Region Attention For Text Spotting | A scene text spotter is composed of text detection and recognition modules. Many studies have been conducted to unify these modules into an end-to-end trainable model to achieve better performance. A typical architecture places detection and recognition modules into separate branches, and a RoI pooling is commonly used to let the branches share a visual feature. However, there still exists a chance of establishing a more complimentary connection between the modules when adopting recognizer that uses attention-based decoder and detector that represents spatial information of the character regions. This is possible since the two modules share a common sub-task which is to find the location of the character regions. Based on the insight, we construct a tightly coupled single pipeline model. This architecture is formed by utilizing detection outputs in the recognizer and propagating the recognition loss through the detection stage. The use of character score map helps the recognizer attend better to the character center points, and the recognition loss propagation to the detector module enhances the localization of the character regions. Also, a strengthened sharing stage allows feature rectification and boundary localization of arbitrary-shaped text regions. Extensive experiments demonstrate state-of-the-art performance in publicly available straight and curved benchmark dataset. | ['Seung Shin', 'Hwalsuk Lee', 'Jeonghun Baek', 'Junyeop Lee', 'Youngmin Baek', 'Sungrae Park', 'Daehyun Nam'] | 2020-07-19 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/6775_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123740494.pdf | eccv-2020-8 | ['text-spotting'] | ['computer-vision'] | [ 2.75818128e-02 -2.08309203e-01 -5.51104806e-02 -4.19171363e-01
-4.28223968e-01 -4.08323765e-01 4.11387295e-01 -7.00685754e-02
-3.71482849e-01 -7.77564868e-02 1.73195690e-01 1.42072067e-01
5.77343464e-01 -7.37639308e-01 -7.15621710e-01 -6.43697798e-01
5.01243889e-01 2.95373142e-01 7.95057118e-01 -1.55018672e-01
7.31971085e-01 5.01832247e-01 -1.09764659e+00 8.70363891e-01
8.63464236e-01 8.68289530e-01 4.36318129e-01 6.52934074e-01
-3.83134544e-01 5.57838261e-01 -4.14255440e-01 -4.16040808e-01
1.74728483e-01 -2.96375602e-01 -4.27773982e-01 1.88188061e-01
4.13503170e-01 -4.74352002e-01 -4.53496188e-01 1.06501687e+00
4.72594231e-01 -3.91734624e-03 7.65411735e-01 -8.11782956e-01
-7.97406912e-01 6.39267981e-01 -1.11328137e+00 -8.89728125e-03
3.63485545e-01 3.72617155e-01 1.04558146e+00 -1.27551973e+00
3.50421339e-01 1.19503975e+00 5.84930301e-01 2.94042677e-01
-9.03917491e-01 -4.21031237e-01 5.57659805e-01 -6.89941877e-03
-1.54929852e+00 -2.83233166e-01 6.02191627e-01 -3.61315191e-01
1.01598871e+00 1.26192555e-01 4.15478289e-01 8.67103636e-01
3.40268850e-01 1.39682758e+00 5.01986563e-01 -4.00011241e-01
-1.91915661e-01 4.12584245e-01 2.91801810e-01 8.02762091e-01
2.94817910e-02 -2.65036106e-01 -6.11325324e-01 4.60552156e-01
9.46005523e-01 4.31972265e-01 -3.17290246e-01 -2.61777490e-01
-1.07221949e+00 6.08016193e-01 8.48773539e-01 2.96015918e-01
-1.53564394e-01 -1.24865443e-01 3.05162042e-01 -5.34885228e-02
1.13154374e-01 2.35886410e-01 -1.78003728e-01 1.57080844e-01
-1.34275830e+00 -1.17731467e-01 5.89437068e-01 8.36588442e-01
6.56178236e-01 -1.20415919e-01 -5.22867620e-01 7.47208893e-01
6.90920115e-01 2.60700613e-01 5.29304326e-01 -1.28020301e-01
8.08436036e-01 1.29629052e+00 -2.92804927e-01 -8.11849594e-01
-2.66843110e-01 -4.47681546e-01 -6.70994401e-01 2.85000384e-01
4.85762596e-01 -5.61998524e-02 -1.09756064e+00 1.16972816e+00
6.74124807e-02 2.79843193e-02 -1.94184050e-01 1.32011783e+00
6.69042349e-01 7.68300533e-01 -1.05284065e-01 6.30289555e-01
1.37404203e+00 -1.39128280e+00 -3.94553542e-01 -3.91696393e-01
5.34673512e-01 -1.01896417e+00 1.18571091e+00 8.87910202e-02
-1.01051474e+00 -7.34233916e-01 -1.26943445e+00 -6.44095957e-01
-5.63197792e-01 7.21010447e-01 2.78950125e-01 3.35412651e-01
-9.00719464e-01 4.53472286e-01 -7.42415428e-01 -4.55818772e-01
4.67905670e-01 2.12137178e-01 -8.66276920e-02 4.24201749e-02
-7.33763516e-01 8.00497115e-01 1.04698211e-01 4.53097671e-01
-7.73203969e-01 -3.32513988e-01 -6.73529029e-01 2.97958702e-01
1.39874388e-02 -5.51315784e-01 7.79195368e-01 -1.25348485e+00
-1.40891111e+00 8.01545441e-01 -1.85974330e-01 -3.82016897e-02
6.59143686e-01 -3.77363771e-01 -2.05137625e-01 -8.96020159e-02
2.04654969e-02 7.94529319e-01 9.74803805e-01 -9.69244480e-01
-9.85577285e-01 -5.35896599e-01 -3.29610080e-01 4.56547201e-01
-3.08702201e-01 1.99158266e-01 -8.65119576e-01 -7.43084550e-01
4.06773120e-01 -5.15148699e-01 5.40173575e-02 4.03951854e-01
-8.04994404e-01 -2.00708300e-01 9.99722362e-01 -5.60012579e-01
1.15861750e+00 -2.34461188e+00 1.55533820e-01 2.00864345e-01
8.50507170e-02 1.76484853e-01 -1.72835842e-01 3.97746891e-01
4.46415655e-02 9.94355185e-04 -2.51291692e-02 -4.90965575e-01
-1.02400832e-01 -4.17451680e-01 -4.94136930e-01 5.47554851e-01
6.12431526e-01 9.32475328e-01 -3.03913653e-01 -3.91443908e-01
2.79590786e-01 4.03877020e-01 -5.09769559e-01 1.09362766e-01
-1.69913143e-01 4.56173122e-02 -6.34790778e-01 8.72638285e-01
7.69499779e-01 -2.42105275e-01 -1.72303602e-01 -1.57232746e-01
-3.76361877e-01 1.72793433e-01 -1.33534646e+00 1.74076998e+00
1.27411157e-01 8.20832968e-01 4.91715223e-02 -8.19503903e-01
1.02544606e+00 -1.94465239e-02 7.99228251e-02 -5.78525484e-01
3.19112480e-01 4.93250564e-02 1.65253893e-01 -4.96867031e-01
6.24961138e-01 3.43697250e-01 1.39627263e-01 3.56894881e-01
-1.04968801e-01 2.96334445e-01 -1.13571554e-01 1.40048102e-01
9.10035133e-01 3.04495364e-01 -2.09723547e-01 -6.58199713e-02
5.91951013e-01 3.90360542e-02 3.43724459e-01 5.67009807e-01
-2.91101754e-01 1.08601749e+00 4.35139388e-01 -3.67658466e-01
-9.23421741e-01 -1.02918994e+00 -1.09468013e-01 1.33594334e+00
4.70931441e-01 -1.59165621e-01 -6.61172807e-01 -7.81230211e-01
2.29182392e-02 3.92257005e-01 -7.46409118e-01 -1.12731159e-01
-6.31127059e-01 -4.79765534e-01 5.41640878e-01 8.41732204e-01
8.61057222e-01 -1.07905602e+00 -5.55484414e-01 3.65480520e-02
1.43536508e-01 -8.38298321e-01 -7.88679898e-01 4.02166516e-01
-7.45436847e-01 -9.18194413e-01 -9.82012808e-01 -1.23408473e+00
1.00866771e+00 4.42500830e-01 5.70902169e-01 2.04555571e-01
-3.59742403e-01 -7.55191445e-02 -3.94167364e-01 -1.97751030e-01
1.53361317e-02 1.63731247e-01 -5.31216145e-01 3.08283836e-01
3.98200065e-01 6.55348971e-02 -8.37804258e-01 5.06687582e-01
-6.93202257e-01 3.67771298e-01 8.92183542e-01 6.37091637e-01
4.50885117e-01 -3.52403998e-01 1.92469478e-01 -4.71898556e-01
4.72442001e-01 -2.31581196e-01 -3.95342290e-01 4.96434987e-01
-1.96278498e-01 1.44265324e-01 5.68343699e-01 -4.53858167e-01
-1.06895232e+00 6.30761743e-01 -1.00912265e-01 -1.86766133e-01
-1.58172652e-01 1.13593437e-01 -3.06200951e-01 -2.20569894e-02
4.88990158e-01 6.18599355e-01 -1.53719321e-01 -5.48425853e-01
3.96842897e-01 7.98628330e-01 3.04705054e-01 -2.06192181e-01
7.10220575e-01 5.48754215e-01 -5.23802578e-01 -7.35042453e-01
-4.06347513e-01 -5.47657847e-01 -8.70521963e-01 -2.69342303e-01
1.02900350e+00 -8.69837761e-01 -5.74696958e-01 7.32289493e-01
-1.19254601e+00 -1.92062601e-01 -7.57705495e-02 2.94150382e-01
-1.06555350e-01 3.54530513e-01 -6.68486416e-01 -5.45073330e-01
-4.10827607e-01 -1.34805548e+00 1.34795582e+00 7.57491887e-01
1.18467346e-01 -6.23627365e-01 -1.58596888e-01 9.12192240e-02
4.22740519e-01 -3.70375514e-01 8.10268521e-01 -7.17970312e-01
-7.95286119e-01 -5.45560479e-01 -6.75979257e-01 2.29304507e-01
-2.32339591e-01 3.14007401e-01 -1.09316730e+00 -2.23725438e-01
-2.59577632e-01 -1.60410821e-01 1.29397857e+00 4.20402706e-01
1.13276303e+00 1.33167610e-01 -6.34834588e-01 6.78467274e-01
1.30353916e+00 1.01601265e-01 8.29914808e-01 2.51515597e-01
7.97958136e-01 5.15517771e-01 2.75527269e-01 2.32851371e-01
3.62607688e-01 4.92039979e-01 3.05282265e-01 -5.63452601e-01
-3.79704386e-01 -3.69256526e-01 6.10318244e-01 4.35166985e-01
3.58034164e-01 -3.85653228e-01 -7.50061929e-01 3.01301539e-01
-1.91605794e+00 -8.48217905e-01 -2.53501773e-01 2.14421701e+00
6.66255713e-01 2.82754809e-01 7.16327652e-02 -6.81337863e-02
1.01818871e+00 1.06307073e-02 -6.56701088e-01 -4.20921147e-01
-1.93017527e-01 -2.12697133e-01 3.08411658e-01 1.93304688e-01
-1.08074749e+00 1.13058352e+00 6.17645836e+00 6.98230565e-01
-1.46167064e+00 -3.67819548e-01 8.16124260e-01 -5.54210506e-02
9.52006429e-02 1.10665113e-02 -1.13708627e+00 4.30452108e-01
2.21795976e-01 4.47905213e-01 1.93244517e-01 7.92015493e-01
1.33031115e-01 -2.87878364e-01 -1.27854800e+00 8.26496899e-01
1.48597196e-01 -1.21098924e+00 2.39162475e-01 -1.46394014e-01
5.91011763e-01 1.99749932e-01 2.02192381e-01 2.45890915e-01
-6.65872917e-02 -1.04929996e+00 8.29823315e-01 7.52688169e-01
5.80715060e-01 -4.63519931e-01 5.15842021e-01 4.16409135e-01
-1.44361484e+00 -1.04526579e-01 -4.75676090e-01 -3.85494828e-02
-7.83655941e-02 3.07759762e-01 -9.10163224e-01 1.76291376e-01
5.76643527e-01 7.01647699e-01 -8.84396374e-01 1.42707837e+00
-3.47305745e-01 3.49614412e-01 -2.59042859e-01 -2.73554772e-01
3.44033033e-01 -2.22530544e-01 3.12921941e-01 1.44778895e+00
1.02054618e-01 -2.95649827e-01 1.88826665e-01 1.23676562e+00
-1.49133638e-01 1.75583988e-01 -3.08863103e-01 1.50116444e-01
3.09503496e-01 1.43382740e+00 -1.02766263e+00 -2.99323916e-01
-6.95194185e-01 1.42412186e+00 4.76147532e-01 3.73189270e-01
-9.57574725e-01 -8.21538746e-01 2.58081138e-01 2.48929992e-01
4.52726334e-01 -4.71571013e-02 -8.18096936e-01 -1.14941072e+00
2.02090308e-01 -8.18266094e-01 2.56554633e-01 -7.71703780e-01
-1.04741096e+00 3.66961867e-01 -7.27636456e-01 -1.13859951e+00
3.86836946e-01 -7.49204218e-01 -1.07495415e+00 1.19964373e+00
-1.49558926e+00 -1.31201875e+00 -3.95604819e-01 6.06410623e-01
9.44277823e-01 -1.20306991e-01 5.35130501e-01 1.67504400e-01
-1.23899829e+00 9.18960214e-01 8.75566006e-02 6.68590784e-01
8.66036057e-01 -1.12651634e+00 3.89505059e-01 1.22385061e+00
1.27920121e-01 6.04558527e-01 8.14703777e-02 -7.80323625e-01
-1.38441432e+00 -1.03149652e+00 5.87905645e-01 -4.58244830e-01
2.60344535e-01 -5.40355802e-01 -1.07225525e+00 4.85125393e-01
1.56062111e-01 -1.69797465e-01 2.38790900e-01 -7.71661056e-03
-4.09347445e-01 7.43570849e-02 -8.73092353e-01 7.10202038e-01
6.86187983e-01 -4.57659334e-01 -7.23040283e-01 7.86010325e-02
1.92144975e-01 -2.39647090e-01 -2.49549240e-01 -2.57773940e-02
5.90837479e-01 -7.88057148e-01 6.65154338e-01 -1.98649466e-01
6.70376360e-01 -5.96595168e-01 1.00209363e-01 -9.43531871e-01
-2.85028130e-01 -1.85296685e-01 2.49613702e-01 1.32475698e+00
7.71329463e-01 -3.41430485e-01 8.20217073e-01 5.51713824e-01
-2.92817503e-01 -7.22132444e-01 -6.72362924e-01 -2.46906787e-01
9.67533514e-02 -1.40194222e-01 2.21186742e-01 6.83735847e-01
2.16658354e-01 4.81399775e-01 -3.11777834e-02 2.77565837e-01
1.53913110e-01 2.64013767e-01 6.28235519e-01 -9.06556547e-01
-6.54198676e-02 -9.32718515e-01 -2.77012616e-01 -1.82931650e+00
-2.28351682e-01 -1.05109227e+00 3.00092340e-01 -1.63240087e+00
5.97766221e-01 -3.26414555e-01 -1.62751436e-01 5.93082130e-01
-3.47985506e-01 -7.40306824e-03 2.55237073e-01 3.81232381e-01
-7.88109541e-01 4.67395574e-01 1.39190543e+00 -1.57994732e-01
-3.33845228e-01 -7.35810176e-02 -6.32694840e-01 7.99699664e-01
6.35172486e-01 -2.60680497e-01 3.74314487e-02 -8.87952685e-01
-1.00071291e-02 -2.94824153e-01 3.24061185e-01 -1.05142272e+00
8.22257042e-01 2.88919151e-01 1.07303333e+00 -8.90244722e-01
2.07502261e-01 -7.39373147e-01 -6.11045957e-01 4.63862836e-01
-3.89590055e-01 -3.17890309e-02 1.51231751e-01 4.25921947e-01
-1.48686126e-01 -3.34081233e-01 1.02941477e+00 9.70914885e-02
-8.43973756e-01 4.22943570e-02 -3.62619877e-01 -1.50568560e-01
1.01434612e+00 -7.51009345e-01 -4.26217020e-01 2.36019026e-03
-3.47755492e-01 4.06214893e-01 4.55559611e-01 5.23347139e-01
9.43841219e-01 -8.14242780e-01 -7.74170697e-01 6.12130702e-01
1.71196274e-02 3.83327864e-02 4.64719422e-02 8.51945341e-01
-6.56599939e-01 3.82580251e-01 -2.01867759e-01 -7.06301868e-01
-1.00641263e+00 2.72278517e-01 7.41533160e-01 -2.16329452e-02
-7.54347980e-01 1.15765321e+00 5.06738365e-01 -1.92729369e-01
6.18000090e-01 -4.84819978e-01 -1.59684286e-01 -3.52535397e-02
6.34667397e-01 2.76837260e-01 -1.35164276e-01 -5.09575844e-01
-4.93957818e-01 8.05189013e-01 -4.24471021e-01 8.49388540e-02
9.97192740e-01 -1.54235333e-01 1.44147232e-01 1.97440311e-01
9.28338349e-01 -1.13813117e-01 -1.60162508e+00 -2.32972994e-01
-7.16748461e-02 -2.72152960e-01 1.36158615e-01 -9.88276422e-01
-1.18564951e+00 1.18635798e+00 6.46602511e-01 -3.06681637e-02
9.65790451e-01 -1.17898554e-01 5.44012785e-01 1.29951492e-01
-1.74488232e-01 -1.26315987e+00 3.20940226e-01 7.60335207e-01
8.57412517e-01 -1.31474209e+00 -1.72913715e-01 -3.35528851e-01
-8.35261703e-01 1.43302190e+00 1.03600097e+00 -3.73784840e-01
4.18483526e-01 4.58477229e-01 -5.60255982e-02 -1.75324023e-01
-4.65714216e-01 -2.39160925e-01 3.19840610e-01 2.51180530e-01
7.02089071e-01 -1.06789030e-01 -5.98937571e-02 8.22491229e-01
3.19538265e-01 -4.72316563e-01 2.51219660e-01 7.26406634e-01
-8.10013950e-01 -7.73852766e-01 -4.61888313e-01 3.54852647e-01
-4.17471021e-01 -2.46444792e-01 -6.77174568e-01 6.61006629e-01
2.33786087e-02 7.91701674e-01 3.27889681e-01 -4.49401528e-01
4.41509575e-01 1.43162146e-01 2.14473587e-02 -7.00491428e-01
-9.44610596e-01 2.96827346e-01 -4.82848674e-01 -4.92631763e-01
2.97162205e-01 -3.49018961e-01 -1.71735704e+00 -1.47103116e-01
-8.18632782e-01 -2.94068456e-01 6.83831453e-01 7.77351618e-01
4.58891541e-01 6.68946564e-01 7.58653641e-01 -7.50078678e-01
-5.10072827e-01 -9.34097588e-01 -4.61816341e-01 2.65319675e-01
5.12270406e-02 -2.74917901e-01 -9.30522680e-02 -2.70337705e-02] | [12.014619827270508, 2.2162253856658936] |
78a0f95b-e62e-4264-bc4f-3a37ad861df1 | approximate-fisher-kernels-of-non-iid-image | 1510.00857 | null | http://arxiv.org/abs/1510.00857v1 | http://arxiv.org/pdf/1510.00857v1.pdf | Approximate Fisher Kernels of non-iid Image Models for Image Categorization | The bag-of-words (BoW) model treats images as sets of local descriptors and
represents them by visual word histograms. The Fisher vector (FV)
representation extends BoW, by considering the first and second order
statistics of local descriptors. In both representations local descriptors are
assumed to be identically and independently distributed (iid), which is a poor
assumption from a modeling perspective. It has been experimentally observed
that the performance of BoW and FV representations can be improved by employing
discounting transformations such as power normalization. In this paper, we
introduce non-iid models by treating the model parameters as latent variables
which are integrated out, rendering all local regions dependent. Using the
Fisher kernel principle we encode an image by the gradient of the data
log-likelihood w.r.t. the model hyper-parameters. Our models naturally generate
discounting effects in the representations; suggesting that such
transformations have proven successful because they closely correspond to the
representations obtained for non-iid models. To enable tractable computation,
we rely on variational free-energy bounds to learn the hyper-parameters and to
compute approximate Fisher kernels. Our experimental evaluation results
validate that our models lead to performance improvements comparable to using
power normalization, as employed in state-of-the-art feature aggregation
methods. | ['Ramazan Gokberk Cinbis', 'Cordelia Schmid', 'Jakob Verbeek'] | 2015-10-03 | null | null | null | null | ['image-categorization'] | ['computer-vision'] | [ 3.91686969e-02 -1.22264430e-01 -3.60044271e-01 -3.95836532e-01
-9.15257633e-01 -5.96614599e-01 1.12653208e+00 3.58185560e-01
-5.07560611e-01 2.91280955e-01 4.91181552e-01 1.84010565e-01
-3.08180749e-01 -7.07186878e-01 -6.00657225e-01 -1.08770072e+00
-2.87611663e-01 2.10734963e-01 2.76977599e-01 -6.35496825e-02
3.98874044e-01 8.22692633e-01 -1.86092138e+00 1.78972319e-01
4.93715733e-01 1.12533128e+00 -1.37431860e-01 6.65420413e-01
-3.41973677e-02 5.76324224e-01 -4.12466168e-01 -2.81261235e-01
3.01997989e-01 -2.76393533e-01 -5.17435968e-01 -1.62967086e-01
7.69650459e-01 -1.87827945e-01 -4.06369418e-01 1.19880784e+00
2.04692587e-01 5.51669359e-01 1.49346232e+00 -1.08781898e+00
-1.08000648e+00 -8.22792724e-02 -5.80644906e-01 1.65892825e-01
1.80200681e-01 -2.63796687e-01 1.40762603e+00 -8.17969918e-01
6.15085900e-01 1.51900506e+00 6.29264235e-01 3.21529537e-01
-1.74500751e+00 -1.30133107e-01 1.20835871e-01 2.12927312e-01
-1.68549395e+00 -2.61469394e-01 5.96165657e-01 -6.92393839e-01
1.03070617e+00 3.90117735e-01 4.00085837e-01 8.88708293e-01
6.47675157e-01 5.64368784e-01 8.77178431e-01 -8.72684538e-01
3.72761339e-01 2.08048418e-01 1.78191841e-01 6.67529225e-01
2.14026630e-01 6.01276942e-02 -4.86232221e-01 -6.65615439e-01
7.79194415e-01 1.03068464e-01 -5.20918593e-02 -9.05075133e-01
-8.85214508e-01 1.32651472e+00 6.53836608e-01 4.47140694e-01
-4.65904891e-01 5.31160474e-01 4.05932575e-01 5.20676672e-02
8.54304671e-01 2.40310878e-02 3.85224782e-02 1.78119034e-01
-9.37098980e-01 4.05065328e-01 4.21499312e-01 6.90137625e-01
9.44445312e-01 -1.59215406e-01 -3.77520680e-01 8.63911748e-01
4.67749923e-01 4.20460910e-01 5.12611270e-01 -8.00467789e-01
-7.17948526e-02 2.58253694e-01 -4.62435978e-03 -1.14789104e+00
-1.22700706e-01 -1.90768167e-01 -6.91027522e-01 4.45821434e-01
2.22270459e-01 7.85834551e-01 -1.18455493e+00 1.73871624e+00
4.57254462e-02 -8.46975893e-02 -2.73312330e-01 6.86440229e-01
2.65288472e-01 7.66835272e-01 5.57396531e-01 -5.31950034e-02
1.40639520e+00 -6.29563928e-01 -8.00936401e-01 1.13084547e-01
5.95327795e-01 -6.45052850e-01 1.02576840e+00 2.85975993e-01
-9.16896880e-01 -5.47940373e-01 -1.04165566e+00 -2.50881642e-01
-6.92937911e-01 -4.21197042e-02 4.25506771e-01 7.00741947e-01
-1.28301632e+00 6.61660492e-01 -1.00266528e+00 -4.36105877e-01
2.86868662e-01 1.55398756e-01 -6.50017917e-01 -2.11686604e-02
-9.54154789e-01 1.04535234e+00 1.83489874e-01 -3.38814974e-01
-7.01589644e-01 -5.81304848e-01 -1.05927455e+00 2.33448491e-01
-1.99547023e-01 -5.95001996e-01 8.31344962e-01 -6.85972869e-01
-1.15395701e+00 9.31616664e-01 -3.15286130e-01 -3.53048056e-01
2.10676327e-01 -8.07001293e-02 -5.91279007e-02 3.26651275e-01
-5.15626669e-02 6.02211177e-01 1.09336293e+00 -1.31484485e+00
-2.27060899e-01 -2.91194469e-01 -5.92131540e-02 -1.13233909e-01
-4.23793644e-01 2.49304064e-02 -2.34191000e-01 -7.69824028e-01
7.19387503e-03 -8.09533119e-01 -8.82765725e-02 2.85692304e-01
-1.41735720e-02 -3.03800762e-01 4.78812277e-01 -6.45002604e-01
1.31331265e+00 -2.52782917e+00 2.04860643e-01 5.56671917e-01
2.13799298e-01 -5.01877330e-02 -2.39576355e-01 5.25565803e-01
-1.40645310e-01 3.01624924e-01 -1.71383247e-01 -5.05773842e-01
3.99340332e-01 5.07282555e-01 -4.18908954e-01 9.96617496e-01
2.46452630e-01 6.31321609e-01 -5.49758792e-01 -6.27023935e-01
3.63813996e-01 1.12237370e+00 -6.67826831e-01 1.04827605e-01
1.05556317e-01 -1.39126450e-01 -3.87092888e-01 2.17236668e-01
7.64457464e-01 -1.13486178e-01 -2.32327878e-01 -4.10215676e-01
-5.62235676e-02 2.20134091e-02 -9.29876506e-01 1.47607875e+00
-4.26766008e-01 5.30471802e-01 -2.52861291e-01 -1.08615565e+00
8.30619514e-01 9.79958922e-02 3.27154428e-01 -6.36177778e-01
2.50907727e-02 -4.15521041e-02 -4.27624315e-01 -1.08476646e-01
4.61484134e-01 -4.80611920e-01 -1.23575307e-01 3.05406842e-02
5.20965338e-01 -6.54217005e-02 1.46062821e-01 1.95441782e-01
7.79910684e-01 1.00031659e-01 6.02307618e-01 -5.96491396e-01
4.48959112e-01 -3.73671651e-01 1.66115120e-01 9.31878865e-01
1.34017169e-02 8.04149151e-01 5.94719350e-01 -2.42204756e-01
-1.09123206e+00 -1.42846704e+00 -5.72860241e-01 1.23345602e+00
-1.05976664e-01 -5.90050280e-01 -7.99318194e-01 -3.67177844e-01
2.62371123e-01 8.75461936e-01 -1.10931420e+00 -4.39773589e-01
-1.75360411e-01 -6.10324323e-01 3.38717431e-01 5.99465251e-01
-1.81572542e-01 -5.15367150e-01 -6.76981628e-01 5.84150143e-02
2.75225550e-01 -7.63957143e-01 -3.69130045e-01 4.05530781e-01
-6.26754761e-01 -6.02049828e-01 -1.07282627e+00 -3.05818945e-01
5.58823228e-01 4.99957241e-02 9.29106832e-01 -1.32974222e-01
-6.81667566e-01 7.49175251e-01 -4.09749866e-01 -1.03166759e-01
-1.39210358e-01 -2.71903932e-01 1.31108299e-01 1.75583273e-01
4.31706011e-01 -4.36618894e-01 -6.91513658e-01 1.03245355e-01
-1.27912319e+00 -5.83859622e-01 4.23142731e-01 9.25926685e-01
9.06012654e-01 -1.41711861e-01 -1.69662297e-01 -5.11468947e-01
6.37804031e-01 -3.27112943e-01 -5.40300727e-01 3.89547557e-01
-5.72528124e-01 5.25341451e-01 3.33360881e-01 -5.52469850e-01
-8.99541318e-01 -1.79994583e-01 1.90415755e-01 -6.09310269e-01
-2.38559693e-01 3.11743796e-01 1.21295132e-01 -3.21774542e-01
6.31799221e-01 8.79268870e-02 -1.08578376e-01 -5.89890301e-01
7.84646809e-01 4.66343850e-01 2.96708703e-01 -7.38505423e-01
6.68261290e-01 6.69932783e-01 2.94999391e-01 -9.54458892e-01
-7.62913406e-01 -7.68016636e-01 -6.65583134e-01 -4.78123575e-02
1.12471652e+00 -7.67422199e-01 -5.53205073e-01 3.46165150e-02
-9.99839723e-01 9.16611850e-02 -5.62683523e-01 5.36628067e-01
-8.17389548e-01 4.46996361e-01 -3.56145114e-01 -1.08568251e+00
6.47434443e-02 -1.02089930e+00 1.43761218e+00 1.67197108e-01
-2.25548416e-01 -1.31176555e+00 3.74424130e-01 -2.37627432e-01
7.20439076e-01 3.09300452e-01 1.25298178e+00 -6.20316386e-01
-1.61883324e-01 -5.28397441e-01 -2.62453586e-01 5.79360723e-01
-7.97201470e-02 1.91716939e-01 -1.20683014e+00 -2.42188886e-01
-9.11449045e-02 -1.21027939e-01 1.28647387e+00 6.16829574e-01
1.09492087e+00 -3.17713171e-01 -2.54825622e-01 6.68870449e-01
1.63124335e+00 -1.99358121e-01 6.88533902e-01 1.74063906e-01
4.63067472e-01 6.48682117e-01 2.83411622e-01 6.15288079e-01
4.17325608e-02 9.07689631e-01 2.87223995e-01 3.86581086e-02
-5.05651347e-02 -2.30171368e-01 2.95056671e-01 5.91478944e-01
-1.42222881e-01 -2.88375914e-01 -8.40048373e-01 5.66512465e-01
-1.82513011e+00 -9.75218296e-01 -3.75179090e-02 2.38962293e+00
5.75670242e-01 -8.29704553e-02 4.53044102e-02 -1.73553497e-01
4.52393442e-01 5.04599750e-01 -1.62546828e-01 -4.86720830e-01
-2.17989609e-01 5.04967868e-01 5.39658189e-01 8.36027324e-01
-1.26287401e+00 6.83053970e-01 6.66502094e+00 1.15603137e+00
-8.74672592e-01 2.23424569e-01 4.12020624e-01 -8.35307240e-02
-4.24691677e-01 1.45116925e-01 -5.98230779e-01 2.69689292e-01
9.32236314e-01 -5.17987572e-02 2.43326887e-01 9.01803434e-01
-2.49617264e-01 -1.29442021e-01 -9.62885022e-01 9.94023502e-01
2.34874219e-01 -1.06371665e+00 3.10743511e-01 4.14849222e-01
5.34915447e-01 -8.34327713e-02 3.17729950e-01 1.34410501e-01
8.43570009e-02 -1.12309039e+00 8.36238921e-01 1.02475417e+00
7.49620795e-01 -7.90439785e-01 6.26672685e-01 -1.21909276e-01
-1.23233831e+00 1.13397971e-01 -6.89865112e-01 2.31974274e-01
1.08992318e-02 6.29108191e-01 -1.77629128e-01 3.22597235e-01
6.74604118e-01 4.07451898e-01 -7.80059338e-01 9.37839389e-01
1.76042050e-01 3.18948120e-01 -4.41087812e-01 7.75394291e-02
3.51618677e-01 -3.74337673e-01 3.03127110e-01 1.33707023e+00
3.39877635e-01 -1.95891947e-01 -1.31943403e-02 9.60888863e-01
3.78532201e-01 2.76335657e-01 -6.63987875e-01 -2.23342460e-02
1.64017137e-02 1.07953286e+00 -6.86583042e-01 -2.83576369e-01
-4.22989786e-01 9.86338317e-01 3.05248350e-01 5.42293906e-01
-5.56546271e-01 -5.51001549e-01 9.68684793e-01 1.58072591e-01
6.04100883e-01 -3.02253813e-01 6.83668330e-02 -1.10614276e+00
8.13591294e-03 -3.03203553e-01 3.58917147e-01 -6.34286225e-01
-1.68635046e+00 5.86747646e-01 5.20312130e-01 -1.03979051e+00
-4.28451955e-01 -1.03380191e+00 -3.95211607e-01 9.78899956e-01
-1.54614151e+00 -1.27049196e+00 4.85163881e-03 5.49423099e-01
-1.07384287e-01 1.68318808e-01 1.05915821e+00 -3.16224284e-02
-2.21713725e-02 6.27964079e-01 4.73161370e-01 -2.02739567e-01
6.96097910e-01 -1.36319900e+00 1.76609665e-01 4.46615398e-01
3.46119672e-01 8.73397112e-01 9.17409837e-01 -3.55457515e-01
-1.01546860e+00 -5.57597458e-01 8.17146003e-01 -3.58654052e-01
7.85339117e-01 -5.56670189e-01 -1.13009822e+00 4.95939016e-01
6.65394291e-02 5.98879755e-01 8.99989307e-01 -2.07345579e-02
-8.49802613e-01 8.31299424e-02 -1.16540754e+00 2.73327142e-01
6.81567848e-01 -9.66872871e-01 -4.20642883e-01 3.30549181e-01
3.66901577e-01 2.45088369e-01 -1.03987694e+00 2.18550444e-01
6.15737200e-01 -1.11230206e+00 1.21233320e+00 -7.25372851e-01
-4.65405360e-02 -1.16533697e-01 -6.98698461e-01 -1.19168580e+00
-7.11919963e-01 -2.37098739e-01 -7.38890469e-02 1.23371875e+00
-7.40333050e-02 -6.77189291e-01 3.22309285e-01 6.55593753e-01
3.18036795e-01 -5.79596817e-01 -1.28805554e+00 -1.03431785e+00
2.86645710e-01 -1.93349868e-01 2.99942225e-01 6.65198624e-01
-7.97853693e-02 -1.73054740e-01 -1.30600974e-01 7.36619160e-03
8.33947062e-01 -3.13522816e-01 3.63690525e-01 -1.38091779e+00
-2.10525125e-01 -7.78402746e-01 -1.11955595e+00 -8.23471665e-01
4.04027790e-01 -7.94360757e-01 -7.87633434e-02 -1.24042547e+00
3.70797426e-01 -1.67807728e-01 -5.41388094e-01 4.41150248e-01
-3.22838612e-02 2.44752809e-01 1.68982461e-01 2.31994212e-01
-3.69041175e-01 8.31402004e-01 7.03317523e-01 -8.26881453e-03
2.79129922e-01 -3.91675651e-01 -4.15165484e-01 7.09406555e-01
4.81297880e-01 -5.50046444e-01 -2.04870909e-01 -5.63377403e-02
1.40568569e-01 -4.18334126e-01 6.11864090e-01 -6.89791560e-01
3.39052230e-02 -7.87197649e-02 4.31690663e-01 -2.39071548e-01
6.11362755e-01 -8.04532290e-01 -1.20234951e-01 1.81688592e-01
-5.23921013e-01 4.14205529e-02 1.17175005e-01 7.26006448e-01
-4.38100994e-01 -4.23402041e-01 8.89888704e-01 1.21254236e-01
-5.25530636e-01 5.58753200e-02 -3.44838351e-01 -1.27528861e-01
9.06477630e-01 -1.75056726e-01 -1.77156389e-01 -5.52311897e-01
-6.05178595e-01 -4.11504984e-01 6.58580363e-01 2.10469946e-01
5.49086332e-01 -1.58588362e+00 -5.45776248e-01 4.12508756e-01
4.79426920e-01 -4.91266698e-01 3.49780262e-01 7.64965534e-01
-5.89347124e-01 5.24554074e-01 -1.65490910e-01 -6.73648477e-01
-1.11411095e+00 7.88850307e-01 1.99977130e-01 -2.53951430e-01
-4.55698580e-01 7.76231647e-01 7.33454049e-01 -8.69624838e-02
2.53525734e-01 -4.19728607e-01 -4.87370603e-02 3.83515477e-01
6.21532261e-01 2.72023588e-01 9.03003439e-02 -8.69157434e-01
-6.01658583e-01 9.83779371e-01 -2.39378318e-01 -2.69438416e-01
1.34492874e+00 6.05378710e-02 -1.10861562e-01 4.51234043e-01
1.78487110e+00 -3.29690576e-02 -1.27349937e+00 -2.26944685e-01
-2.19815932e-02 -6.84507370e-01 4.52616543e-01 -3.56204838e-01
-6.78366184e-01 1.14363170e+00 9.08744991e-01 3.87223542e-01
1.02013302e+00 3.98787320e-01 1.03094175e-01 -1.40285287e-02
2.77110159e-01 -9.36457276e-01 -6.99659809e-02 2.84814715e-01
1.01835322e+00 -8.22478771e-01 2.61016697e-01 -2.04552442e-01
-5.13413310e-01 1.13870668e+00 -1.34416148e-01 -6.38807893e-01
9.15929377e-01 1.35585740e-01 -8.46246406e-02 -1.11039571e-01
-5.10744572e-01 -2.45443895e-01 8.62925768e-01 7.15288579e-01
4.40425724e-01 1.39181301e-01 -1.52798876e-01 2.51272917e-01
1.39664441e-01 -4.54797953e-01 4.98125032e-02 8.71209919e-01
-4.48802918e-01 -1.06448603e+00 -4.77626771e-01 2.98077911e-01
-4.61165130e-01 -1.32077485e-01 -1.82424888e-01 8.68052065e-01
-1.51038617e-01 5.31777024e-01 3.43414843e-01 -2.05193445e-01
2.76047289e-01 2.79397726e-01 6.64520264e-01 -3.54006976e-01
-2.37312749e-01 1.99641287e-01 -3.72980326e-01 -6.18157744e-01
-3.60816777e-01 -6.93763793e-01 -9.22441006e-01 -1.24782830e-01
-4.40141767e-01 2.24969648e-02 7.81905651e-01 7.10663378e-01
2.56746918e-01 3.30753863e-01 3.36862624e-01 -1.25248039e+00
-1.00489128e+00 -8.31656396e-01 -9.54642415e-01 6.44985914e-01
4.34794515e-01 -1.06905794e+00 -7.72714376e-01 1.42860562e-01] | [9.094504356384277, 2.776909828186035] |
7f186746-fadc-4f40-8b2c-0417658d77e8 | decoding-p300-variability-using-convolutional | null | null | http://dx.doi.org/10.3389/fnhum.2019.00201 | https://www.frontiersin.org/articles/10.3389/fnhum.2019.00201/pdf | Decoding P300 Variability using Convolutional Neural Networks | Deep convolutional neural networks (CNN) have previously been shown to be useful tools for signal decoding and analysis in a variety of complex domains, such as image processing and speech recognition. By learning from large amounts of data, the representations encoded by these deep networks are often invariant to moderate changes in the underlying feature spaces. Recently, we proposed a CNN architecture that could be applied to electroencephalogram (EEG) decoding and analysis. In this article, we train our CNN model using data from prior experiments in order to later decode the P300 evoked response from an unseen, hold-out experiment. We analyze the CNN output as a function of the underlying variability in the P300 response and demonstrate that the CNN output is sensitive to the experiment-induced changes in the neural response. We then assess the utility of our approach as a means of improving the overall signal-to-noise ratio in the EEG record. Finally, we show an example of how CNN-based decoding can be applied to the analysis of complex data. | ['Stephen M. Gordon', 'Vernon J. Lawhern', 'Jonathan Touryan', 'Anthony J. Ries', 'Jonathan R. McDaniel', 'Amelia J. Solon'] | 2019-06-14 | null | null | null | frontiers-in-human-neuroscience-2019-6 | ['eeg-decoding', 'eeg-decoding'] | ['medical', 'time-series'] | [ 4.93271649e-01 -2.47729301e-01 7.08332062e-01 -4.78455901e-01
-5.64998925e-01 -4.82933819e-01 4.23193574e-01 2.68205911e-01
-5.55736899e-01 6.38330817e-01 1.49555624e-01 -1.42002702e-01
-3.52151245e-02 -3.89845908e-01 -9.27074552e-01 -7.03168511e-01
-3.21206301e-01 -2.18106955e-01 1.22732192e-01 -2.05901951e-01
1.76378742e-01 6.46667957e-01 -1.43251777e+00 6.77694678e-01
3.75290066e-01 1.36000514e+00 8.73018578e-02 5.42226195e-01
4.85107005e-01 4.50718433e-01 -9.32761490e-01 1.97069556e-01
-6.69306144e-03 -6.05633199e-01 -4.89629447e-01 -3.53139043e-01
-9.36244279e-02 -8.98852423e-02 -6.20688975e-01 1.06288254e+00
8.98526490e-01 -4.46764976e-02 5.58565915e-01 -8.14019263e-01
-1.91987827e-01 3.73497039e-01 -8.69583618e-03 8.05564225e-01
3.46016556e-01 1.66975468e-01 4.49091017e-01 -8.33307564e-01
2.57459104e-01 6.96457267e-01 6.58251107e-01 3.72149080e-01
-1.59600437e+00 -7.76574194e-01 -2.72508293e-01 3.16165030e-01
-1.37696660e+00 -6.02224469e-01 7.80696452e-01 -4.31284964e-01
1.09044266e+00 6.16551675e-02 7.14911520e-01 1.26292968e+00
6.03550255e-01 4.69960093e-01 9.39102411e-01 -2.72460103e-01
3.71754438e-01 -1.67179108e-01 1.29280239e-01 1.15090899e-01
-9.82820317e-02 1.34346664e-01 -7.11431623e-01 -1.22942431e-02
6.18364334e-01 -1.45206928e-01 -6.96856737e-01 1.38298839e-01
-1.20435834e+00 5.81466079e-01 7.89074242e-01 5.33279121e-01
-6.08735621e-01 2.65883118e-01 4.56583619e-01 5.26290298e-01
4.33196783e-01 8.39885712e-01 -5.63593090e-01 -2.46229410e-01
-1.03181887e+00 1.10874154e-01 5.76694906e-01 3.58480901e-01
3.86765063e-01 1.85790583e-01 -2.55517870e-01 7.37843454e-01
-2.28929833e-01 1.32636979e-01 8.09207857e-01 -5.93535423e-01
1.78693429e-01 2.03911111e-01 -1.65728349e-02 -7.80081153e-01
-7.99827814e-01 -4.86977339e-01 -7.74625599e-01 1.72891989e-01
3.22440922e-01 -3.10941249e-01 -8.45566273e-01 1.82222414e+00
-4.27201748e-01 1.57046750e-01 -8.29142928e-02 7.04234302e-01
6.07441485e-01 4.18367356e-01 -1.27032667e-01 -1.38262630e-01
1.12584996e+00 -8.12512264e-02 -7.26800919e-01 -3.25368762e-01
3.34222436e-01 -3.44450176e-01 7.99543560e-01 5.05125821e-01
-1.06569266e+00 -4.73014563e-01 -1.10938740e+00 2.35501751e-01
-3.16635221e-01 2.29618121e-02 1.55469224e-01 4.02909219e-01
-1.05510604e+00 8.43610168e-01 -9.11506414e-01 -3.57254475e-01
6.99867725e-01 8.71611953e-01 -4.15920734e-01 3.38489324e-01
-1.27882349e+00 9.34115827e-01 4.73630488e-01 4.11771715e-01
-1.00960422e+00 -4.19673741e-01 -5.14797747e-01 6.09451115e-01
-2.42773607e-01 -2.40535676e-01 1.10121274e+00 -9.24837708e-01
-1.30433750e+00 6.15705729e-01 -1.65934697e-01 -5.63532472e-01
2.73305587e-02 8.25414360e-02 -4.84626889e-01 7.23378314e-03
-3.21997106e-01 3.94888699e-01 7.72038519e-01 -5.63572347e-01
-2.77456284e-01 -5.11051893e-01 -4.21079189e-01 -2.73208231e-01
-2.54759192e-01 1.18222050e-01 7.70438984e-02 -6.03534281e-01
1.42436355e-01 -6.67659342e-01 4.44894657e-02 -3.27440947e-01
-8.17144513e-02 1.69412225e-01 2.73126394e-01 -5.31079173e-01
8.62909913e-01 -2.63477874e+00 1.79900035e-01 3.59382749e-01
1.42288387e-01 1.45099044e-01 -3.55554014e-01 4.99460012e-01
-4.44418758e-01 -2.03167107e-02 -2.68449515e-01 1.31053075e-01
-2.32042149e-01 -1.52438506e-01 -3.30784470e-01 6.08922720e-01
5.49059153e-01 1.17474437e+00 -6.24504149e-01 4.77827251e-01
-8.07692781e-02 6.37551367e-01 -3.22894841e-01 3.85383427e-01
1.67437270e-01 7.89026439e-01 1.59730315e-01 2.02993199e-01
3.47753227e-01 -5.82210943e-02 -6.08157413e-03 -2.92805851e-01
1.56150125e-02 5.31509340e-01 -4.99691516e-01 1.64670455e+00
-3.95984590e-01 1.47526109e+00 -4.02150750e-02 -1.20112002e+00
8.61450672e-01 6.29829943e-01 3.06789577e-01 -1.06158161e+00
5.95927179e-01 3.13855320e-01 6.45943403e-01 -4.81535226e-01
-5.80948107e-02 -1.02286145e-01 2.09503442e-01 3.96630973e-01
3.84201586e-01 -1.73614800e-01 -1.53660193e-01 -1.93787903e-01
1.21943748e+00 -5.09447992e-01 1.32604554e-01 -3.25192243e-01
1.30533665e-01 -4.90680009e-01 7.75004998e-02 7.58679330e-01
-2.10496224e-02 7.58550644e-01 7.67407358e-01 -5.42905211e-01
-8.88855755e-01 -8.33408594e-01 -5.41804016e-01 6.75905287e-01
-2.24317595e-01 -1.00619555e-01 -7.24232495e-01 2.49301773e-02
-2.55335450e-01 4.55142796e-01 -7.38906682e-01 -7.58908868e-01
-3.73914063e-01 -1.03213489e+00 6.30208910e-01 6.12324595e-01
1.00605451e-01 -1.41665781e+00 -8.78804147e-01 6.16306424e-01
3.11681791e-03 -1.28709316e+00 1.89185247e-01 9.94795978e-01
-8.22533965e-01 -9.03828084e-01 -5.86562216e-01 -6.73682928e-01
5.14604211e-01 -3.48505557e-01 7.62165070e-01 -1.01903416e-01
-3.51803809e-01 2.57452726e-01 -2.13010728e-01 -6.15951896e-01
-1.66067958e-01 7.72394389e-02 1.81601882e-01 1.93515584e-01
4.71898735e-01 -1.01483798e+00 -6.39361799e-01 -1.01421010e-02
-1.05154145e+00 -3.18992138e-01 3.77944261e-01 9.74612474e-01
2.46999711e-01 1.17923006e-01 9.15846825e-01 -4.78899658e-01
1.24523711e+00 -4.47982699e-01 -3.93571585e-01 -9.16225836e-02
-2.68567890e-01 2.09272519e-01 6.99345708e-01 -6.05262816e-01
-3.34065437e-01 -1.61564220e-02 -4.01498944e-01 -1.81659266e-01
-2.04352543e-01 7.89995909e-01 3.15556303e-02 -4.59167093e-01
8.27704191e-01 4.11763996e-01 -1.76203340e-01 -3.32982033e-01
-2.23770842e-01 7.05481946e-01 6.59724176e-01 -2.82056987e-01
1.37331247e-01 2.35477760e-01 -6.87503219e-02 -7.94511616e-01
-4.11316752e-01 -8.53144750e-02 -5.70940971e-01 -1.10731050e-01
7.17778265e-01 -8.18109810e-01 -8.17493320e-01 6.10579789e-01
-1.36057806e+00 -4.66318637e-01 -7.06252977e-02 6.48945689e-01
-5.26047707e-01 -1.94263309e-01 -5.56733131e-01 -5.68879426e-01
-4.30849195e-01 -1.21660972e+00 8.08497310e-01 1.40195593e-01
-3.43116522e-01 -7.59809732e-01 7.78573453e-02 -4.91364717e-01
5.68605185e-01 2.08647743e-01 1.15092170e+00 -9.28793669e-01
3.57518345e-02 -5.31651974e-01 -6.58124387e-02 4.05435413e-01
5.89114986e-02 -4.33438241e-01 -1.41275251e+00 -3.78175467e-01
3.68485123e-01 -2.45194539e-01 8.49810660e-01 5.80916464e-01
1.50704455e+00 7.85764772e-03 -7.31660426e-02 7.66345978e-01
1.16792536e+00 3.07355642e-01 9.43435371e-01 1.69449419e-01
2.81848133e-01 3.37258488e-01 -4.47508276e-01 2.88723052e-01
-3.03866595e-01 5.06772399e-01 3.86128843e-01 -1.45846575e-01
1.84996560e-01 2.47043408e-02 2.91959673e-01 7.14573801e-01
-5.52146621e-02 -1.89739451e-01 -9.49900866e-01 5.76869726e-01
-1.50295603e+00 -7.95437992e-01 -1.73247345e-02 2.25575018e+00
5.85731328e-01 2.32287362e-01 -1.21725224e-01 3.42683047e-01
3.81205410e-01 -1.98812068e-01 -5.76027274e-01 -4.94850636e-01
-1.65693104e-01 8.22708905e-01 2.23058730e-01 3.35969180e-02
-6.42902374e-01 3.93560261e-01 7.44233370e+00 2.37477601e-01
-1.70421290e+00 -8.03705305e-03 4.24706787e-01 -1.80536494e-01
7.63135925e-02 -5.25144160e-01 -6.68888912e-02 5.80195606e-01
1.51826620e+00 -2.73689032e-01 7.26301074e-01 2.26756632e-01
2.54244953e-01 -1.49605781e-01 -1.58595097e+00 1.25403666e+00
-1.32329121e-01 -1.29236889e+00 -3.63712788e-01 -9.60361771e-03
4.28314626e-01 3.65682542e-01 2.18671113e-01 1.87863171e-01
-4.20309871e-01 -1.48742318e+00 5.81229925e-01 4.73163992e-01
8.88950050e-01 -5.91997087e-01 9.67937469e-01 3.93864155e-01
-7.74037480e-01 -3.16139489e-01 -2.83837438e-01 -3.68470430e-01
-1.20275855e-01 4.36909109e-01 -6.87680125e-01 3.55658196e-02
8.22862446e-01 6.34160697e-01 -6.00638092e-01 1.09350145e+00
-1.28747270e-01 7.18895376e-01 -1.68865919e-01 -3.77063304e-02
5.54399788e-02 2.14830860e-01 2.73265749e-01 1.20421898e+00
4.90699351e-01 2.85349011e-01 -4.58530366e-01 1.06087041e+00
-3.18432599e-01 -2.80195504e-01 -5.64881146e-01 -2.73114592e-01
2.90332764e-01 9.20944273e-01 -7.04381227e-01 -5.14093903e-04
-1.59499422e-01 9.02644932e-01 3.31262678e-01 7.72646427e-01
-3.13800633e-01 -6.42525256e-01 4.88983989e-01 4.81978133e-02
3.21329832e-01 -2.40407899e-01 -3.16209763e-01 -1.26446533e+00
1.75896287e-01 -8.27091455e-01 -1.83435217e-01 -1.08521473e+00
-1.03376210e+00 9.84838605e-01 -1.53469980e-01 -9.92477417e-01
-4.12547082e-01 -8.17875028e-01 -7.05048144e-01 1.11318040e+00
-1.26672077e+00 -2.11143121e-01 -2.18418285e-01 7.16700673e-01
3.09821337e-01 -1.18692942e-01 1.02178192e+00 3.09453428e-01
-5.01874566e-01 2.78738886e-01 2.54223108e-01 2.37168446e-01
4.14245754e-01 -8.89028311e-01 6.13074064e-01 8.42778265e-01
3.63136292e-01 6.23631239e-01 6.62672043e-01 -2.45612953e-03
-1.09632802e+00 -7.98282802e-01 5.19328535e-01 -2.58226305e-01
6.29065394e-01 -9.35283363e-01 -1.12832606e+00 5.49778581e-01
2.63698876e-01 1.92374766e-01 6.46439672e-01 -4.08536978e-02
-1.36732802e-01 -3.11664075e-01 -9.91658151e-01 4.58306521e-01
4.85541970e-01 -9.86449659e-01 -6.47763193e-01 1.32559454e-02
1.93006754e-01 -3.49347264e-01 -6.75112128e-01 2.83028781e-01
7.89354801e-01 -9.42450941e-01 6.66382730e-01 -6.70175850e-01
2.38706455e-01 4.96808514e-02 -1.24592721e-01 -1.86188304e+00
-4.38516140e-01 -5.20919263e-01 -9.01708938e-03 5.32361448e-01
4.52261448e-01 -6.99363649e-01 2.37613618e-01 5.21024644e-01
-1.07198499e-01 -7.75301933e-01 -1.04276824e+00 -4.88092870e-01
1.04449600e-01 -5.70782244e-01 4.65183049e-01 2.56099343e-01
3.71957600e-01 3.26828092e-01 -8.79021659e-02 2.25034058e-01
4.93548438e-02 -4.18879658e-01 1.91243634e-01 -1.24092579e+00
3.41399945e-02 -3.25565964e-01 -8.80728424e-01 -7.66371727e-01
2.03740537e-01 -9.36046064e-01 4.25925970e-01 -1.24714601e+00
2.76516806e-02 1.59150168e-01 -9.14344907e-01 4.24306065e-01
4.52290252e-02 4.32871640e-01 1.71968579e-01 9.39877033e-02
-6.98423432e-03 4.26845908e-01 6.60161972e-01 -2.28755817e-01
-2.25233838e-01 6.47729356e-03 -4.73800331e-01 2.94581890e-01
8.44980240e-01 -7.74802446e-01 -1.91101164e-01 -5.30407965e-01
3.76821429e-01 9.97111388e-03 5.33330977e-01 -1.36497140e+00
2.18966588e-01 5.58051884e-01 8.50216866e-01 -1.20864145e-01
2.94298142e-01 -9.68369424e-01 1.21639054e-02 3.61329615e-01
-6.41851544e-01 7.97363371e-02 6.66685581e-01 4.69722509e-01
-4.52159435e-01 -9.58964229e-03 5.89668751e-01 3.45554687e-02
-3.84365141e-01 1.17898032e-01 -7.58289516e-01 -1.32235527e-01
6.03080213e-01 -1.14137828e-01 -1.66876689e-01 -4.59940732e-01
-8.83294463e-01 -1.69912577e-01 -3.37639526e-02 2.51973599e-01
6.54185176e-01 -1.17329550e+00 -6.38054013e-01 8.75799298e-01
6.03613034e-02 -5.02332330e-01 5.35442270e-02 1.07956779e+00
-8.94534513e-02 4.42463547e-01 -7.38364398e-01 -8.59570503e-01
-9.14479375e-01 3.72072756e-01 6.99007988e-01 2.82058448e-01
-4.85971987e-01 7.36858904e-01 3.13913941e-01 1.66306645e-01
2.61006922e-01 -7.22056866e-01 -2.16639698e-01 -3.57297398e-02
7.51496732e-01 -2.26271525e-01 7.45233238e-01 -3.18327427e-01
-4.61200356e-01 2.25460470e-01 6.68634772e-02 -2.80727804e-01
1.72158778e+00 1.71034321e-01 -1.79197073e-01 7.35801756e-01
1.42272139e+00 -4.94304776e-01 -1.23498619e+00 1.75987229e-01
2.86487304e-02 -6.59390166e-02 2.28074551e-01 -8.07129681e-01
-1.08271968e+00 1.34935427e+00 9.40748394e-01 3.40976536e-01
1.44146824e+00 -2.05088794e-01 4.46322173e-01 5.13890862e-01
3.82396102e-01 -7.27165043e-01 -5.36432602e-02 5.14124334e-01
9.60588098e-01 -1.01455879e+00 -2.67692149e-01 3.89512986e-01
-2.67251521e-01 1.57673860e+00 -2.11175345e-02 -3.87842715e-01
7.38231480e-01 4.41213071e-01 -1.09386072e-01 -3.35960954e-01
-8.01449955e-01 3.39856632e-02 4.79216546e-01 6.89033926e-01
6.02901816e-01 -7.29215145e-02 -1.09310225e-01 8.14807296e-01
-1.49637401e-01 3.37587953e-01 6.60857975e-01 7.19382763e-01
-3.09111863e-01 -6.35964215e-01 -2.20668286e-01 6.36382282e-01
-6.22681499e-01 -3.05905581e-01 -3.94294918e-01 4.98215914e-01
-1.21772408e-01 8.03049445e-01 2.37850532e-01 -5.92396975e-01
5.85246801e-01 3.63923013e-01 6.30748451e-01 -7.62136698e-01
-7.77493179e-01 3.45112085e-02 -2.59379506e-01 -5.78021288e-01
-3.02963555e-01 -4.31803435e-01 -1.12638223e+00 2.06550255e-01
-2.23833427e-01 -5.21971285e-02 7.80415595e-01 1.11758649e+00
3.97038311e-01 9.38347518e-01 4.54495579e-01 -1.03096664e+00
-2.32891038e-01 -1.11351228e+00 -6.97560549e-01 2.83344507e-01
7.97597706e-01 -3.80007476e-01 -3.86246711e-01 4.96364832e-02] | [13.079389572143555, 3.4349663257598877] |
3ef8be0b-425a-426f-9723-b34abde85546 | geometry-aware-supertagging-with | 2203.12235 | null | https://arxiv.org/abs/2203.12235v3 | https://arxiv.org/pdf/2203.12235v3.pdf | Geometry-Aware Supertagging with Heterogeneous Dynamic Convolutions | The syntactic categories of categorial grammar formalisms are structured units made of smaller, indivisible primitives, bound together by the underlying grammar's category formation rules. In the trending approach of constructive supertagging, neural models are increasingly made aware of the internal category structure, which in turn enables them to more reliably predict rare and out-of-vocabulary categories, with significant implications for grammars previously deemed too complex to find practical use. In this work, we revisit constructive supertagging from a graph-theoretic perspective, and propose a framework based on heterogeneous dynamic graph convolutions aimed at exploiting the distinctive structure of a supertagger's output space. We test our approach on a number of categorial grammar datasets spanning different languages and grammar formalisms, achieving substantial improvements over previous state of the art scores. Code will be made available at https://github.com/konstantinosKokos/dynamic-graph-supertagging | ['Michael Moortgat', 'Konstantinos Kogkalidis'] | 2022-03-23 | null | null | null | null | ['ccg-supertagging'] | ['natural-language-processing'] | [ 4.62270901e-02 7.50402153e-01 -4.70109247e-02 -4.11400944e-01
-2.03294486e-01 -9.63387847e-01 1.07882547e+00 2.97029078e-01
-2.25990400e-01 2.17640027e-01 4.83206481e-01 -5.84217310e-01
-1.00721933e-01 -1.12846899e+00 -3.65275174e-01 -6.38881266e-01
-3.52830231e-01 5.84215343e-01 2.21677691e-01 -5.36439896e-01
-4.43651117e-02 -5.31060547e-02 -1.53325856e+00 2.99167335e-01
7.25798249e-01 6.73616171e-01 1.94576576e-01 5.34582496e-01
-3.03187639e-01 1.00195301e+00 -5.30127697e-02 -6.02959037e-01
6.15861341e-02 -3.21253955e-01 -8.41571033e-01 -9.50284153e-02
5.82048833e-01 3.04705411e-01 -5.09158731e-01 1.24355531e+00
-7.90734217e-02 1.14768274e-01 6.06028020e-01 -8.73301804e-01
-9.68160629e-01 1.24849403e+00 1.24330617e-01 2.07799509e-01
4.68369015e-02 3.81796695e-02 1.89760745e+00 -3.78747553e-01
7.69703269e-01 1.53295994e+00 6.71972573e-01 6.25526428e-01
-1.36893868e+00 -5.79842448e-01 6.30522788e-01 -1.60757646e-01
-1.28285050e+00 -2.70482093e-01 6.98445857e-01 -6.38161063e-01
1.11135983e+00 2.29922488e-01 6.82405531e-01 9.12987649e-01
9.78269652e-02 5.65927267e-01 1.00741935e+00 -5.36086202e-01
1.19756334e-01 -3.86417449e-01 6.14534974e-01 9.44093466e-01
4.54483271e-01 2.41998404e-01 -1.45137891e-01 -6.84783086e-02
5.91966033e-01 -4.47929464e-02 9.99564528e-02 -6.32413566e-01
-1.18556547e+00 1.16754663e+00 7.06865609e-01 1.00153148e+00
-1.33406386e-01 6.21300936e-01 3.96210045e-01 5.36960125e-01
7.70178199e-01 3.63049150e-01 -2.97120988e-01 1.03150018e-01
-5.40663242e-01 1.56694204e-01 8.40905070e-01 1.05482364e+00
6.31301582e-01 9.90340710e-02 2.90535480e-01 7.59616256e-01
6.29263878e-01 2.58332640e-01 3.49942476e-01 -6.17897093e-01
3.59391093e-01 1.02763653e+00 -5.88188112e-01 -8.37681711e-01
-6.69859588e-01 -6.00462556e-01 -6.10835433e-01 -2.76303917e-01
5.87842762e-01 1.87684655e-01 -1.19762433e+00 1.92642069e+00
1.68214589e-01 -2.68946476e-02 -9.09592360e-02 6.57402217e-01
8.59563112e-01 1.31415501e-01 6.46504939e-01 3.28270704e-01
1.48199153e+00 -3.10818583e-01 -2.48660058e-01 -3.01994979e-01
9.81533647e-01 -2.44712651e-01 1.06542575e+00 2.07041800e-01
-7.50578284e-01 -3.16091865e-01 -7.20328867e-01 -1.76563278e-01
-5.84198534e-01 -4.77623850e-01 1.20560026e+00 7.43370116e-01
-1.63355339e+00 6.35249972e-01 -9.19994175e-01 -6.72318995e-01
5.00554025e-01 4.63895589e-01 -3.33441347e-01 8.61454606e-02
-1.23322523e+00 6.36004806e-01 8.11006546e-01 -2.29260504e-01
-6.70372605e-01 -4.66876388e-01 -9.91334200e-01 1.09527312e-01
3.51693481e-01 -7.94683993e-01 1.36013043e+00 -9.20153499e-01
-9.06283438e-01 1.19605982e+00 2.65582442e-01 -5.06813586e-01
-8.58142041e-03 4.44199502e-01 -3.87569457e-01 -1.90938085e-01
-8.68289247e-02 5.11306345e-01 4.65569377e-01 -1.00593305e+00
-7.71865308e-01 -4.73156422e-01 5.09462595e-01 2.22074226e-01
-2.07966775e-01 4.18474749e-02 -9.13198739e-02 -6.48624539e-01
4.19791162e-01 -1.02754152e+00 -2.62087226e-01 -6.54347718e-01
-3.19354862e-01 -9.05462623e-01 2.40578622e-01 -4.54341084e-01
1.28479147e+00 -2.05139780e+00 3.27966750e-01 2.32867271e-01
9.12325680e-01 -1.71852752e-03 -3.74381803e-02 5.93142331e-01
-9.36497748e-02 4.31371838e-01 -1.08168684e-01 -3.08998674e-01
4.72087294e-01 3.61373395e-01 -2.27889970e-01 3.96400481e-01
-1.79946586e-01 1.34403205e+00 -1.25386870e+00 -3.80798817e-01
1.73426032e-01 9.85713005e-02 -6.24692559e-01 -2.34897062e-01
-4.90781665e-01 8.11775923e-02 -4.69200343e-01 7.64984429e-01
1.09329849e-01 -5.36740661e-01 8.08916092e-01 2.97417730e-01
-1.75215434e-02 7.57224083e-01 -6.13346338e-01 1.56674635e+00
-4.26167876e-01 4.45964992e-01 -1.35209486e-01 -1.16323471e+00
7.54964828e-01 1.53779060e-01 6.91124573e-02 -5.30161917e-01
4.50867802e-01 3.03878039e-01 5.59322059e-01 2.42342785e-01
4.57309753e-01 -4.07535106e-01 -5.51138520e-01 4.93887007e-01
5.39745212e-01 -6.12862594e-02 2.57159233e-01 6.28068686e-01
1.20870483e+00 1.96931977e-02 5.73984504e-01 -7.96369672e-01
1.63954899e-01 -1.22591451e-01 1.36748508e-01 8.57476950e-01
-1.92792546e-02 2.26140887e-01 6.05239153e-01 -7.79510796e-01
-9.30703282e-01 -1.19050980e+00 5.88445067e-02 1.62768269e+00
-1.97461843e-01 -5.07259369e-01 -6.12322867e-01 -7.26577222e-01
2.00533390e-01 7.14273453e-01 -8.38461280e-01 -1.95861995e-01
-5.43586671e-01 -5.92997134e-01 5.05376816e-01 3.57032120e-01
7.26783574e-02 -1.12270379e+00 -1.74551845e-01 3.52242112e-01
5.47747500e-02 -9.24641907e-01 -2.76257992e-01 2.80327499e-01
-9.30012643e-01 -1.00720382e+00 -4.04445594e-03 -7.77699769e-01
4.05984342e-01 2.08819270e-01 1.69179881e+00 6.26398146e-01
9.64639336e-02 3.55210811e-01 -4.85749245e-01 -4.13294792e-01
-8.48094761e-01 4.49187994e-01 -7.64053911e-02 -1.40109584e-01
6.55243635e-01 -8.88210654e-01 -3.05378735e-01 -8.24728683e-02
-8.72270405e-01 2.36031190e-01 2.70670533e-01 6.63091958e-01
1.66732773e-01 -4.05213356e-01 2.00558379e-01 -1.36145365e+00
5.01362979e-01 -7.46001422e-01 -7.69405365e-01 6.68413714e-02
-6.18545830e-01 1.21652201e-01 5.66172779e-01 -1.90925285e-01
-6.93754375e-01 -3.80951434e-01 -3.02666072e-02 -1.52655929e-01
-3.03871781e-01 7.43370354e-01 -1.02400698e-01 -1.97729133e-02
6.80841446e-01 2.38564298e-01 -6.84931800e-02 -3.58710527e-01
7.54468858e-01 4.84994560e-01 2.25199461e-01 -5.47253489e-01
7.74472773e-01 4.78097588e-01 -5.57799451e-02 -7.53283560e-01
-7.20781982e-01 -6.06227875e-01 -7.19373167e-01 -2.17408583e-01
7.79134750e-01 -8.35320890e-01 -5.67080081e-01 5.33845484e-01
-9.18134391e-01 -6.89195514e-01 -1.47867531e-01 1.83105543e-01
-4.50942993e-01 3.40451062e-01 -7.83965766e-01 -4.49281752e-01
-1.14518076e-01 -5.02081752e-01 9.99174833e-01 -2.08295032e-01
-3.13259780e-01 -1.71279955e+00 8.05214643e-02 1.32315218e-01
4.21417713e-01 2.33590499e-01 1.14417052e+00 -1.04047966e+00
-7.09878087e-01 -2.13928428e-02 -8.03847760e-02 1.03530131e-01
4.36272006e-03 -4.50184286e-01 -8.70994449e-01 -3.47133219e-01
-4.56312478e-01 4.39987928e-02 1.10522306e+00 1.59982413e-01
6.63216174e-01 -2.77476192e-01 -3.52006406e-01 5.88525295e-01
1.45016229e+00 8.42081979e-02 4.92381901e-01 7.92624950e-02
1.01162469e+00 4.53104824e-01 1.49326205e-01 -1.13044139e-02
7.57634699e-01 6.27426386e-01 5.21255255e-01 2.34260052e-01
-4.44657743e-01 -5.03003001e-01 1.49937630e-01 1.19867229e+00
-3.92749459e-01 -6.18542790e-01 -1.34792507e+00 7.80351818e-01
-1.73700392e+00 -9.37534451e-01 -2.27492496e-01 1.80145323e+00
5.25083899e-01 1.22629002e-01 1.05167635e-01 -1.66954428e-01
7.75477409e-01 3.24306130e-01 -1.96260899e-01 -5.16337991e-01
-8.74266922e-02 3.89857471e-01 5.39437771e-01 5.32079816e-01
-1.32118106e+00 1.56970847e+00 5.69145870e+00 7.62498736e-01
-8.80527735e-01 3.92743677e-01 3.68969917e-01 2.56034136e-01
-8.02173138e-01 2.73833513e-01 -5.47462583e-01 3.95399749e-01
1.14690685e+00 -1.98894486e-01 5.98715603e-01 6.13527179e-01
-4.00916666e-01 1.90531760e-01 -1.05004930e+00 6.01463497e-01
-2.26671189e-01 -1.37149560e+00 5.51376231e-02 6.23123467e-01
4.87340003e-01 6.62696898e-01 -1.10730626e-01 5.49674094e-01
1.29103625e+00 -8.47327888e-01 1.04430759e+00 2.38682687e-01
6.87535524e-01 -2.94700086e-01 5.09001255e-01 1.19263634e-01
-1.42407835e+00 -2.20229089e-01 -2.96095401e-01 -4.49714869e-01
-1.04037389e-01 3.11713070e-01 -8.62086236e-01 5.77236950e-01
4.02252227e-01 7.50075579e-01 -6.05983973e-01 4.14299637e-01
-4.22566384e-01 1.10606468e+00 -2.32118219e-01 -3.35597515e-01
5.06231248e-01 -2.50179142e-01 4.93739367e-01 1.12538660e+00
1.83703259e-01 3.00819665e-01 3.02626908e-01 7.85085380e-01
-2.15565458e-01 1.58444166e-01 -1.02661133e+00 -4.23155248e-01
5.37281930e-01 1.22584641e+00 -1.18027401e+00 -3.67293149e-01
-6.10683739e-01 2.48432457e-01 7.53985763e-01 4.43606153e-02
-4.92046386e-01 -4.91949320e-02 6.83352113e-01 1.93604350e-01
2.76848048e-01 -4.73454863e-01 -2.02773497e-01 -1.30572224e+00
-3.24412227e-01 -7.15034664e-01 8.73938799e-01 -4.43121016e-01
-1.21251011e+00 6.21711373e-01 6.75637368e-03 -7.67872512e-01
-5.35080492e-01 -7.06957042e-01 -3.70075077e-01 6.63785398e-01
-9.49896336e-01 -1.83726466e+00 7.92768374e-02 4.57272261e-01
1.97607592e-01 -3.11278015e-01 9.62372065e-01 -4.54786345e-02
9.81928706e-02 5.31262815e-01 -1.53110549e-01 3.55569273e-01
-9.63564739e-02 -1.58961427e+00 1.08104968e+00 1.03636849e+00
6.74262583e-01 7.69296169e-01 7.31106877e-01 -5.90859890e-01
-1.21840870e+00 -1.11646998e+00 1.17240298e+00 -7.53321230e-01
1.36922264e+00 -1.05975533e+00 -8.92708182e-01 1.04236758e+00
9.40904468e-02 -3.98245603e-02 5.98759592e-01 7.32891321e-01
-8.12950969e-01 2.75687575e-01 -7.93657839e-01 5.77786028e-01
1.87248743e+00 -7.35122144e-01 -7.65439868e-01 3.81953269e-01
9.20006096e-01 -1.85843542e-01 -9.06117737e-01 4.88050021e-02
4.90224898e-01 -7.66950727e-01 7.10804760e-01 -7.12541580e-01
9.36659425e-02 4.25441898e-02 -4.55389917e-01 -1.41517997e+00
-8.08558166e-01 -5.55393338e-01 2.67213397e-02 8.93185794e-01
3.21027219e-01 -1.07371211e+00 6.51632369e-01 2.37780020e-01
-6.48147047e-01 -2.23065302e-01 -1.10199070e+00 -7.82064259e-01
5.48894346e-01 -5.79175293e-01 7.32687354e-01 1.10981560e+00
3.24243486e-01 4.58742112e-01 1.24434702e-01 1.59944836e-02
7.23788083e-01 2.07967311e-01 4.41433311e-01 -1.73081958e+00
-1.97300091e-01 -7.17465222e-01 -1.00567377e+00 -6.21172726e-01
4.00121063e-01 -1.76083088e+00 -1.18821070e-01 -1.49117827e+00
1.18687242e-01 -6.12731874e-01 -4.17831957e-01 9.43412125e-01
9.28438753e-02 4.09586757e-01 4.17211831e-01 1.74099609e-01
-9.44884479e-01 1.14369169e-02 8.81685555e-01 -1.57884642e-01
7.35011846e-02 -3.26889575e-01 -1.27136266e+00 6.68914258e-01
1.13261724e+00 -3.11118871e-01 -4.49960172e-01 -5.12213349e-01
5.20021498e-01 -3.32985938e-01 5.29804826e-01 -8.69544923e-01
2.49516740e-02 -1.09634608e-01 -4.67087835e-01 -3.82898077e-02
-1.01862669e-01 -3.49631429e-01 3.54745716e-01 5.26735485e-01
-2.97607094e-01 -8.85709152e-02 5.16725183e-02 5.75367153e-01
-3.42685729e-02 1.98288828e-01 6.70070529e-01 -4.28450912e-01
-1.00963867e+00 3.87133092e-01 -2.71613896e-01 2.45797575e-01
6.55270338e-01 -1.53388083e-01 -4.82191980e-01 -3.42867404e-01
-1.01869667e+00 -8.70047361e-02 6.49412632e-01 7.05451071e-01
2.13355690e-01 -1.15924728e+00 -6.18526042e-01 1.17945753e-01
3.99729133e-01 -3.67012769e-01 1.70708328e-01 5.70949554e-01
-2.45841458e-01 7.57856190e-01 8.22294205e-02 -4.71539617e-01
-1.17536438e+00 5.42268276e-01 3.51166099e-01 -3.78917128e-01
-6.15825474e-01 8.79414380e-01 7.27513492e-01 -8.07706952e-01
-3.46294880e-01 -7.41642118e-01 -1.80065915e-01 3.44895222e-03
8.18122253e-02 -1.22011311e-01 2.37020209e-01 -9.12741721e-01
-2.71667957e-01 2.05122262e-01 -4.73018549e-02 2.42077872e-01
1.37812126e+00 -9.75044221e-02 -4.31972593e-01 5.35767674e-01
9.64839280e-01 -1.74287215e-01 -8.29930365e-01 -4.18727785e-01
4.91535276e-01 -1.14929453e-01 9.17937420e-03 -6.93190873e-01
-8.41791689e-01 8.16406906e-01 2.59455621e-01 8.24190140e-01
6.73806369e-01 7.70540833e-01 4.34489280e-01 1.83404461e-01
1.02107322e+00 -7.54454374e-01 -3.98994058e-01 8.79875302e-01
6.46689534e-01 -8.93814504e-01 -4.61666554e-01 -5.99787056e-01
-2.91184068e-01 8.55326653e-01 2.06716191e-02 -3.69283050e-01
7.61929691e-01 -6.43526539e-02 8.17584526e-03 -6.82774961e-01
-9.68595684e-01 -7.03175545e-01 4.04502541e-01 6.59648955e-01
5.33005416e-01 7.22227216e-01 -3.24425012e-01 6.31494224e-01
-4.53595608e-01 -3.24520469e-01 4.16048795e-01 5.85685551e-01
-5.71872115e-01 -1.27393830e+00 1.61199957e-01 7.61723757e-01
-4.01645988e-01 -5.94290376e-01 -4.45806742e-01 1.09456182e+00
1.53999597e-01 9.04529929e-01 2.29804292e-01 -5.17567575e-01
1.72553584e-01 3.08000565e-01 4.61923361e-01 -1.10546374e+00
-5.14179587e-01 -1.04600914e-01 3.56786877e-01 -5.52844763e-01
-3.39430124e-01 -1.02609766e+00 -1.16612613e+00 -5.50522149e-01
-1.21517986e-01 1.32619634e-01 3.35048825e-01 7.51281917e-01
3.35687727e-01 3.55769783e-01 7.31909722e-02 -5.95319211e-01
-4.67197686e-01 -1.16101241e+00 -8.14290226e-01 5.83544970e-01
4.73230258e-02 -8.73393416e-01 -4.29095298e-01 -1.49951175e-01] | [6.933982849121094, 6.30259370803833] |
1a12eaf4-785c-42ea-b2ce-f7f15509aac8 | prompting-large-language-models-for-zero-shot | 2306.16007 | null | https://arxiv.org/abs/2306.16007v1 | https://arxiv.org/pdf/2306.16007v1.pdf | Prompting Large Language Models for Zero-Shot Domain Adaptation in Speech Recognition | The integration of Language Models (LMs) has proven to be an effective way to address domain shifts in speech recognition. However, these approaches usually require a significant amount of target domain text data for the training of LMs. Different from these methods, in this work, with only a domain-specific text prompt, we propose two zero-shot ASR domain adaptation methods using LLaMA, a 7-billion-parameter large language model (LLM). LLM is used in two ways: 1) second-pass rescoring: reranking N-best hypotheses of a given ASR system with LLaMA; 2) deep LLM-fusion: incorporating LLM into the decoder of an encoder-decoder based ASR system. Experiments show that, with only one domain prompt, both methods can effectively reduce word error rates (WER) on out-of-domain TedLium-2 and SPGISpeech datasets. Especially, the deep LLM-fusion has the advantage of better recall of entity and out-of-vocabulary words. | ['Shujie Liu', 'Jinyu Li', 'Yu Wu', 'Yuang Li'] | 2023-06-28 | null | null | null | null | ['speech-recognition'] | ['speech'] | [ 2.10700378e-01 1.11681670e-01 -7.80799463e-02 -4.77521896e-01
-1.42791343e+00 -2.45402679e-01 6.69552803e-01 -5.80454525e-03
-8.84189487e-01 5.81929743e-01 4.30909216e-01 -5.02954900e-01
1.78809538e-01 -1.95831180e-01 -6.46970630e-01 -3.18794250e-01
5.89539349e-01 8.30193400e-01 5.63162625e-01 -4.53018904e-01
-4.26835194e-02 4.33468148e-02 -1.07743156e+00 3.91740292e-01
8.54041755e-01 8.29339147e-01 7.57219076e-01 6.31001592e-01
-4.87094760e-01 5.38505971e-01 -8.29065025e-01 -4.35527593e-01
-4.61709723e-02 -4.30638105e-01 -7.31756985e-01 -7.17024878e-02
-6.06402270e-02 -2.98990279e-01 -5.88417888e-01 9.98071492e-01
8.46245825e-01 3.55590284e-01 5.37405968e-01 -6.21731877e-01
-4.38954920e-01 1.00445175e+00 -3.02824229e-01 3.51204962e-01
1.01086654e-01 -1.04538567e-01 7.21109271e-01 -1.18815207e+00
3.45061451e-01 1.49930894e+00 2.45819420e-01 8.09648812e-01
-9.77052808e-01 -8.16953540e-01 1.21333987e-01 3.06966096e-01
-1.58416831e+00 -1.01675284e+00 5.78512311e-01 -9.56805870e-02
1.44914877e+00 -1.11604996e-01 1.17781991e-02 1.32594872e+00
-1.90593436e-01 8.75583708e-01 7.50532269e-01 -7.59011924e-01
3.61367196e-01 2.10801065e-01 3.25187683e-01 6.15171716e-02
1.88469693e-01 -1.50061445e-02 -9.65602636e-01 -1.09214187e-01
3.58527660e-01 -5.19214749e-01 -1.45588279e-01 1.16112985e-01
-1.00128150e+00 7.83562541e-01 -3.95971179e-01 5.68122268e-01
-2.56890327e-01 -2.01094896e-01 5.38336277e-01 3.64548892e-01
6.87656820e-01 2.82217234e-01 -8.48931789e-01 -4.27468061e-01
-1.10495901e+00 -7.67541900e-02 5.32850742e-01 1.02636242e+00
4.17622179e-01 1.27841830e-01 -3.55416566e-01 1.43641949e+00
4.21263158e-01 7.28308022e-01 1.28884053e+00 -1.53607026e-01
8.53622019e-01 2.58690864e-01 1.81890670e-02 -1.09294966e-01
-1.72361806e-01 -3.75032246e-01 -5.07072747e-01 -3.47138166e-01
1.29487202e-01 -2.55032957e-01 -1.21245933e+00 1.85787964e+00
4.54756729e-02 2.33313277e-01 6.70509636e-01 5.94756305e-01
8.37003767e-01 9.17047560e-01 2.23982066e-01 -3.67968500e-01
1.39013195e+00 -1.02735126e+00 -1.10477996e+00 -6.69901967e-01
8.04577589e-01 -9.40094709e-01 1.18933082e+00 2.26622611e-01
-1.21344090e+00 -6.84308052e-01 -1.06616974e+00 -9.39143226e-02
-3.73676300e-01 3.54549408e-01 -1.76564902e-01 7.10092068e-01
-8.10042202e-01 1.50959566e-01 -5.91888785e-01 -4.58566785e-01
6.70354767e-03 2.15971828e-01 -1.78621039e-01 -8.67384300e-02
-1.65631306e+00 1.03537500e+00 7.51022339e-01 -4.87343758e-01
-8.17232251e-01 -5.48773944e-01 -8.48079801e-01 2.02965036e-01
4.56352562e-01 -2.35875040e-01 1.71784687e+00 -7.04606712e-01
-1.89328897e+00 8.43782485e-01 -5.58269203e-01 -8.02396953e-01
2.32344836e-01 -4.89983469e-01 -8.12992454e-01 -2.18131512e-01
-2.29866982e-01 4.18618232e-01 8.17750335e-01 -6.36320770e-01
-7.86387742e-01 -3.30330044e-01 -5.81138492e-01 4.44066852e-01
-5.77007949e-01 3.78867179e-01 -5.67198098e-01 -7.85562634e-01
-8.32345858e-02 -6.54219270e-01 5.03543653e-02 -8.18435788e-01
-1.99594140e-01 -6.52754307e-01 8.08391690e-01 -1.11068010e+00
1.81026721e+00 -2.30709028e+00 -3.96127515e-02 -5.53052835e-02
-2.10089624e-01 9.70063627e-01 -2.97312766e-01 3.45757723e-01
1.49210438e-01 -2.48520374e-02 -1.56869218e-01 -5.76396763e-01
2.51874444e-03 1.57902807e-01 -4.32582557e-01 -4.17901203e-03
1.16391450e-01 8.33146572e-01 -6.00866735e-01 -3.25028628e-01
3.10179979e-01 3.01647753e-01 -1.95455581e-01 3.57315928e-01
-2.65473396e-01 2.31678650e-01 -1.17708303e-01 1.07872471e-01
5.78647673e-01 1.57589447e-02 1.86447486e-01 3.05794179e-01
1.32621333e-01 9.64124441e-01 -1.14149594e+00 1.81752932e+00
-6.17262542e-01 4.15226877e-01 -3.66919041e-02 -8.57210279e-01
1.05130351e+00 7.45980442e-01 -2.52156463e-02 -9.20385540e-01
9.88040045e-02 6.13397717e-01 -2.92998448e-04 -1.61980718e-01
5.57690322e-01 -1.80036262e-01 -2.82002479e-01 1.43946812e-01
4.00760293e-01 -4.48387973e-02 -2.03870796e-02 2.65241824e-02
1.15010786e+00 -3.89238417e-01 5.55811346e-01 9.12715420e-02
7.34793186e-01 -2.38965034e-01 5.34876645e-01 7.40438044e-01
-1.21904448e-01 5.75196981e-01 -2.85682995e-02 1.29020065e-01
-1.08351529e+00 -1.02707160e+00 3.64609361e-02 1.21308804e+00
-1.31404012e-01 -3.99139613e-01 -7.98966169e-01 -6.45368099e-01
-1.24644928e-01 1.30106473e+00 1.37712881e-01 -4.39029545e-01
-7.67195940e-01 -4.47548717e-01 9.25823629e-01 3.53305310e-01
3.95428807e-01 -9.19802368e-01 2.17068437e-02 4.21478897e-01
-2.98841506e-01 -1.46008527e+00 -7.59175003e-01 3.58277053e-01
-6.02596819e-01 -5.71181476e-01 -1.04007554e+00 -8.36534262e-01
2.65404910e-01 2.12604553e-01 8.02887678e-01 -4.18085635e-01
2.82614440e-01 1.08957075e-01 -7.10189402e-01 -4.44889635e-01
-1.02948511e+00 3.17589611e-01 3.81173849e-01 -7.74798542e-02
9.12768841e-01 -1.82004958e-01 -9.84954461e-02 3.17185193e-01
-6.77727580e-01 5.45218633e-03 8.47123504e-01 8.46821845e-01
4.68915701e-01 -2.23506242e-01 1.00692701e+00 -6.51706219e-01
8.79547715e-01 -3.64130855e-01 -5.44600844e-01 4.71563667e-01
-6.69102490e-01 2.42113203e-01 6.76599324e-01 -7.24549174e-01
-1.28973377e+00 -2.96781249e-02 -6.17682576e-01 -4.26335722e-01
-1.78429797e-01 3.87707055e-01 -5.06962180e-01 5.33610880e-01
5.26562512e-01 5.18576324e-01 -1.51232883e-01 -1.01154566e+00
3.57243955e-01 1.42269754e+00 5.11255860e-01 -1.66510165e-01
5.66804230e-01 -1.57127142e-01 -7.94141829e-01 -1.00586140e+00
-5.68849087e-01 -9.29858148e-01 -5.08239627e-01 2.50561297e-01
8.16526473e-01 -1.05155087e+00 -1.10757217e-01 5.63167274e-01
-1.39915252e+00 -2.23722249e-01 -2.57934153e-01 8.23894978e-01
-4.06747580e-01 4.73628938e-01 -4.25397694e-01 -9.15784776e-01
-4.57271516e-01 -1.10316563e+00 1.07374096e+00 -3.51776592e-02
-3.66882175e-01 -7.16109455e-01 8.76217932e-02 3.86714935e-01
4.25725460e-01 -9.87545371e-01 9.37131226e-01 -1.48300242e+00
9.35101658e-02 -1.26646087e-01 6.32157475e-02 7.77842999e-01
6.97218627e-02 -7.64901519e-01 -1.19028032e+00 -5.02562165e-01
6.32731616e-02 -2.98444897e-01 7.54515171e-01 3.29060465e-01
9.30951297e-01 -1.75939709e-01 -3.86377960e-01 3.19106638e-01
9.23377991e-01 6.15084469e-01 5.47125757e-01 1.37513608e-01
2.93508410e-01 3.36546510e-01 7.71549463e-01 3.32050979e-01
2.38617182e-01 1.00075698e+00 -2.69946218e-01 8.38140920e-02
-5.28836429e-01 -5.87592959e-01 8.20851743e-01 1.48655379e+00
6.37141943e-01 -6.38548672e-01 -1.01379514e+00 6.76216960e-01
-1.68811047e+00 -4.13999230e-01 2.60373682e-01 2.48492050e+00
1.04642427e+00 2.62703180e-01 -4.89975736e-02 -5.87013178e-02
9.55131412e-01 1.80326000e-01 -6.14370644e-01 -2.38752082e-01
-4.05541748e-01 2.83101201e-01 5.54101110e-01 5.36764443e-01
-8.47072601e-01 1.38595784e+00 5.37399340e+00 1.49574649e+00
-8.97125959e-01 5.59280217e-01 2.99040169e-01 -1.13426737e-01
-1.67788669e-01 -2.14579739e-02 -1.44319201e+00 7.49615014e-01
1.61984122e+00 -2.89322436e-01 2.86951929e-01 7.99301445e-01
2.22649053e-01 1.06191434e-01 -9.93626475e-01 1.28392243e+00
1.44356996e-01 -9.21245933e-01 1.85655951e-01 -1.49443030e-01
3.19647938e-01 3.04228365e-01 -2.08497643e-01 8.00854206e-01
3.52161050e-01 -8.47280264e-01 6.67628884e-01 1.04694948e-01
1.12004864e+00 -7.40519702e-01 7.02286899e-01 7.84511268e-01
-9.51169431e-01 9.86421667e-03 -5.00018179e-01 2.81024158e-01
4.51741278e-01 5.10976911e-01 -1.20578849e+00 2.69505918e-01
4.80998158e-01 1.90055728e-01 -2.72536874e-01 7.37733424e-01
-1.55944541e-01 1.06964850e+00 -2.50537544e-01 -1.78017020e-01
5.37815690e-02 3.08474123e-01 9.24325287e-01 1.45422196e+00
4.32832867e-01 -4.09212299e-02 -2.32844595e-02 3.35778207e-01
-2.66002595e-01 3.07991594e-01 -3.54914457e-01 -2.84955591e-01
9.16771233e-01 7.05203056e-01 -2.17991084e-01 -5.32572269e-01
-5.40337265e-01 1.19359863e+00 2.11117432e-01 2.37462789e-01
-5.61272323e-01 -5.44438422e-01 8.90663385e-01 1.44798279e-01
3.53867263e-01 -3.40314180e-01 -5.32938577e-02 -1.10516906e+00
2.60079000e-02 -1.00140381e+00 2.41644666e-01 -7.11407781e-01
-1.10604596e+00 6.05111301e-01 -2.48854887e-02 -1.06094360e+00
-5.62124133e-01 -5.62370002e-01 -1.39086008e-01 1.12390852e+00
-1.83635867e+00 -9.18226302e-01 2.41473034e-01 3.72474551e-01
1.34607339e+00 -7.52025664e-01 8.12247515e-01 5.45028210e-01
-4.35797274e-01 9.33488786e-01 5.43692708e-01 2.12790072e-01
1.10234356e+00 -8.78453374e-01 1.04717445e+00 9.06952739e-01
2.44081959e-01 3.92318577e-01 5.09053946e-01 -8.26376438e-01
-1.03931129e+00 -1.13069904e+00 1.31108761e+00 -2.98725575e-01
4.55033243e-01 -6.73577964e-01 -1.15068865e+00 6.15573525e-01
1.83383793e-01 -4.43989724e-01 5.33733785e-01 -1.96595192e-02
-5.03821820e-02 1.26708229e-03 -8.20406556e-01 5.56126714e-01
9.72041428e-01 -6.65220320e-01 -1.08786631e+00 2.09648982e-01
1.25639749e+00 -4.53475654e-01 -5.49156129e-01 3.85348409e-01
2.27969423e-01 -3.42910051e-01 8.47389817e-01 -5.65037310e-01
-1.63842872e-01 -2.02710062e-01 -4.00351435e-01 -1.65674937e+00
-1.04795851e-01 -6.71363115e-01 -2.36220732e-01 1.45740926e+00
5.05999863e-01 -7.08766103e-01 2.72556663e-01 1.41916841e-01
-3.53667319e-01 -2.97164589e-01 -1.31868589e+00 -1.14657581e+00
1.63239390e-01 -7.36214340e-01 7.16323376e-01 4.62017357e-01
-7.42728487e-02 7.23094285e-01 -2.46484160e-01 1.83258340e-01
1.09038003e-01 -8.12052965e-01 6.65010035e-01 -1.10842240e+00
-3.96843821e-01 -2.18461201e-01 -1.12625659e-01 -1.66186464e+00
3.64616156e-01 -8.90120387e-01 2.58969367e-01 -1.29412472e+00
-6.17110264e-03 -3.17348540e-01 -4.77082193e-01 2.72873551e-01
-3.84132564e-02 -4.73851264e-01 1.18924007e-01 2.15441287e-01
-6.07904255e-01 7.48829484e-01 6.36874199e-01 -2.15160716e-02
-2.99405515e-01 1.50436908e-01 -5.29293716e-01 6.12675190e-01
6.94379151e-01 -6.36405289e-01 -4.43424910e-01 -5.53839684e-01
-2.56131649e-01 3.07421952e-01 -1.92078680e-01 -9.38987732e-01
3.96266550e-01 4.37127724e-02 1.28235281e-01 -8.77284706e-01
4.26441193e-01 -5.87221801e-01 -3.28887939e-01 2.30516717e-01
-4.90657240e-01 -4.59394380e-02 4.15999353e-01 4.57599074e-01
-3.41615140e-01 -5.20120203e-01 9.44751322e-01 -5.66380695e-02
-1.04860783e+00 -1.39697745e-01 -7.17245877e-01 3.44653428e-01
6.93516076e-01 -1.21933356e-01 -2.12515250e-01 -4.09525990e-01
-5.05838096e-01 1.63511559e-01 6.47271350e-02 8.57534170e-01
5.81220865e-01 -1.13281620e+00 -8.65907550e-01 4.93541777e-01
3.58133942e-01 2.63937805e-02 2.24614337e-01 4.33458865e-01
1.67043000e-01 8.06346953e-01 4.69433546e-01 -4.14618075e-01
-1.38719440e+00 3.88158053e-01 2.91180819e-01 -5.19356072e-01
-3.93650353e-01 1.01025200e+00 4.23073918e-01 -5.44355035e-01
4.33739066e-01 -2.03535840e-01 -1.76306352e-01 -1.52448744e-01
7.16089308e-01 3.56581360e-01 5.10646105e-01 -6.61127627e-01
-4.30404812e-01 1.94770128e-01 -3.55285645e-01 -7.03243494e-01
9.80505228e-01 -5.90359986e-01 3.80837351e-01 6.67032778e-01
9.76250291e-01 1.24695241e-01 -9.49590147e-01 -7.05893040e-01
3.27289373e-01 -1.42826540e-02 2.37096205e-01 -1.00046504e+00
-3.12058210e-01 1.04454553e+00 6.04373872e-01 -3.19840133e-01
1.01704824e+00 1.46814153e-01 1.40894759e+00 4.10462707e-01
3.10231417e-01 -1.69277489e+00 3.80167887e-02 1.02753150e+00
7.42769063e-01 -1.26144004e+00 -5.40989518e-01 1.68934188e-04
-7.97383726e-01 9.16313231e-01 6.20437920e-01 5.00864863e-01
6.18077159e-01 2.09264547e-01 1.38722852e-01 3.10936213e-01
-9.22810376e-01 -2.67362624e-01 3.99395347e-01 5.32046080e-01
4.02517617e-01 1.85510918e-01 -3.37825358e-01 1.06997943e+00
-2.40025759e-01 -1.68337837e-01 1.96552604e-01 6.98477268e-01
-7.97463179e-01 -1.29029107e+00 -4.21246231e-01 2.98584878e-01
-4.15863454e-01 -4.52396393e-01 -2.98571646e-01 4.98266697e-01
-2.31794238e-01 1.14094412e+00 8.97766724e-02 -4.12619442e-01
5.01141250e-01 6.86052442e-01 2.88851950e-02 -9.50284779e-01
-2.65399486e-01 2.96084255e-01 9.42674130e-02 -5.29243276e-02
2.93232232e-01 -6.30226135e-01 -1.30056608e+00 2.40318939e-01
-5.87346971e-01 2.43663847e-01 8.87150168e-01 1.17354918e+00
4.98106897e-01 5.00402570e-01 4.44280326e-01 -2.29968622e-01
-1.02401555e+00 -1.52916360e+00 -5.69966972e-01 2.35808253e-01
2.71306276e-01 -5.37978470e-01 -3.12814593e-01 -6.84024915e-02] | [14.3560791015625, 6.8195624351501465] |
d332e866-0b01-4977-a80d-834ccab88f02 | s2abel-a-dataset-for-entity-linking-from | 2305.00366 | null | https://arxiv.org/abs/2305.00366v1 | https://arxiv.org/pdf/2305.00366v1.pdf | S2abEL: A Dataset for Entity Linking from Scientific Tables | Entity linking (EL) is the task of linking a textual mention to its corresponding entry in a knowledge base, and is critical for many knowledge-intensive NLP applications. When applied to tables in scientific papers, EL is a step toward large-scale scientific knowledge bases that could enable advanced scientific question answering and analytics. We present the first dataset for EL in scientific tables. EL for scientific tables is especially challenging because scientific knowledge bases can be very incomplete, and disambiguating table mentions typically requires understanding the papers's tet in addition to the table. Our dataset, S2abEL, focuses on EL in machine learning results tables and includes hand-labeled cell types, attributed sources, and entity links from the PaperswithCode taxonomy for 8,429 cells from 732 tables. We introduce a neural baseline method designed for EL on scientific tables containing many out-of-knowledge-base mentions, and show that it significantly outperforms a state-of-the-art generic table EL method. The best baselines fall below human performance, and our analysis highlights avenues for improvement. | ['Doug Downey', 'Aakanksha Naik', 'Sergey Feldman', 'Erin Bransom', 'Bailey Kuehl', 'Yuze Lou'] | 2023-04-30 | null | null | null | null | ['entity-linking'] | ['natural-language-processing'] | [-4.68013942e-01 4.53382969e-01 -6.49531186e-01 1.96850430e-02
-1.28173077e+00 -1.19982600e+00 3.28177869e-01 1.22533834e+00
-2.75734514e-01 1.48878860e+00 4.25777972e-01 -5.15444756e-01
-2.46746302e-01 -1.04804730e+00 -1.19316089e+00 -2.63667256e-01
6.64226934e-02 1.14620912e+00 1.32826582e-01 1.81924284e-01
2.51919985e-01 6.51435316e-01 -7.54943013e-01 5.48921049e-01
8.38671625e-01 7.07977712e-01 -1.08971104e-01 4.76240903e-01
-1.01053119e+00 9.54161108e-01 -9.61832881e-01 -8.86128545e-01
-3.01719517e-01 7.22129922e-03 -1.03331375e+00 -8.47345293e-01
8.28926444e-01 6.15990460e-01 -4.74038213e-01 7.96286404e-01
5.16894042e-01 -1.40697241e-01 8.13715935e-01 -1.16891253e+00
-1.01072025e+00 1.27277648e+00 -4.91685420e-01 5.33024728e-01
4.90019083e-01 -3.28467846e-01 1.34812057e+00 -7.50248075e-01
1.67900884e+00 1.25316918e+00 8.74267697e-01 2.86198080e-01
-1.14598501e+00 -9.53072369e-01 1.49096176e-01 2.79935151e-01
-1.41043949e+00 -4.06852245e-01 -1.56083629e-02 -5.67469120e-01
1.12763679e+00 2.18209535e-01 3.20998162e-01 8.15409541e-01
9.50577408e-02 5.83200514e-01 7.81053662e-01 -1.62146911e-01
2.50849962e-01 2.79531240e-01 5.04035115e-01 6.27594471e-01
1.02591372e+00 -8.54599118e-01 -9.01403189e-01 -3.33905697e-01
4.93454337e-01 -6.15243316e-01 -2.32150227e-01 9.81185213e-03
-1.71507215e+00 3.26526970e-01 4.94507402e-01 2.37144932e-01
-1.97755098e-01 -3.94634418e-02 5.52624106e-01 -9.36115533e-02
2.61185169e-01 1.31102955e+00 -1.04032302e+00 9.50527042e-02
-7.90607095e-01 5.56180775e-01 1.51692498e+00 1.46761537e+00
5.61778545e-01 -6.75501406e-01 -6.04728937e-01 5.22124171e-01
-1.57189056e-01 2.64178395e-01 -2.94218391e-01 -1.01187325e+00
5.90662956e-01 9.13559914e-01 1.13676459e-01 -8.02713752e-01
-5.26404619e-01 -6.56984508e-01 -6.33166671e-01 -4.38318729e-01
8.83569777e-01 1.38221234e-02 -6.75009072e-01 1.36050344e+00
3.98080528e-01 6.87567964e-02 3.04708838e-01 2.19968423e-01
2.10678434e+00 5.18484533e-01 5.96961200e-01 -1.93195775e-01
1.79599321e+00 -4.94054675e-01 -1.25819969e+00 2.05661431e-01
5.85777223e-01 -8.23197722e-01 4.55759227e-01 1.57173723e-01
-9.43278253e-01 -5.80771118e-02 -6.61575437e-01 -6.51534081e-01
-1.47391486e+00 1.34062096e-01 9.62993383e-01 5.18005490e-02
-8.68358254e-01 5.46881795e-01 -3.16621274e-01 -5.32534003e-01
6.27773821e-01 2.50992835e-01 -5.85476875e-01 6.19099326e-02
-1.28656435e+00 1.07759643e+00 5.88768244e-01 -1.98021941e-02
-3.98260891e-01 -1.67960095e+00 -6.52857363e-01 3.13507438e-01
6.86682343e-01 -8.78100693e-01 8.82521510e-01 4.95381534e-01
-5.35359740e-01 1.30521083e+00 -3.86297941e-01 -3.10235679e-01
3.08767892e-02 4.62661274e-02 -5.61400414e-01 -6.85000420e-02
4.54932302e-01 7.25306153e-01 -4.12262440e-01 -1.46779144e+00
-6.16833627e-01 -3.37819934e-01 -7.32691288e-02 -1.25404641e-01
-1.04464851e-02 -7.79341459e-02 -7.88720071e-01 -5.95333219e-01
7.61515927e-03 -6.47342145e-01 1.46814361e-01 -1.34760411e-02
-9.77520466e-01 -5.89105308e-01 4.89342779e-01 -9.21656489e-01
1.23029268e+00 -1.32557309e+00 8.53599831e-02 1.24908589e-01
7.10264385e-01 -2.01096356e-01 2.68375397e-01 4.90002155e-01
-5.41115142e-02 6.45216644e-01 -9.42147430e-03 -2.84112543e-02
1.65737689e-01 8.99732560e-02 -4.77807373e-01 7.76714012e-02
-1.00181028e-02 1.22343493e+00 -1.22379112e+00 -8.98176432e-01
-4.40391541e-01 3.07230055e-01 -1.93670288e-01 -5.93963861e-02
-7.67121911e-01 1.70543462e-01 -5.02881765e-01 1.25761199e+00
3.81767601e-01 -6.25631630e-01 2.34695256e-01 -5.29660165e-01
-1.57375365e-01 5.57116210e-01 -9.84421730e-01 1.56722283e+00
-4.27041240e-02 7.99821079e-01 1.26250044e-01 -3.64597529e-01
7.94502616e-01 1.64751679e-01 4.04816955e-01 -9.27749276e-02
-2.27024421e-01 3.70752454e-01 -1.87322497e-01 -3.47047150e-01
4.24572021e-01 5.17570257e-01 -1.82617113e-01 -4.68654297e-02
2.81321198e-01 -3.28088179e-02 8.86303782e-01 8.35422397e-01
1.33339465e+00 1.82890445e-01 3.35757256e-01 -6.21327698e-01
5.91666698e-01 4.31766659e-01 6.88082516e-01 9.57925141e-01
2.09094435e-01 7.73197338e-02 8.21772456e-01 -3.92620116e-01
-7.35190690e-01 -1.10096490e+00 -5.76730788e-01 7.76562572e-01
-1.78486571e-01 -7.15925276e-01 -4.24814701e-01 -7.85717607e-01
7.52930105e-01 8.21475744e-01 -8.06880295e-01 3.25981140e-01
-4.49629068e-01 -8.09999943e-01 7.86704659e-01 5.00627637e-01
1.66823968e-01 -1.04823947e+00 3.48957509e-01 2.12346464e-01
-5.16875863e-01 -1.34574699e+00 -2.42241144e-01 5.55959046e-01
-4.93629873e-01 -1.38632226e+00 -3.59464735e-01 -8.32793415e-01
6.23125315e-01 -3.87109905e-01 1.87186337e+00 -2.14271113e-01
-3.78397465e-01 8.95548165e-02 6.55413792e-02 -9.37079966e-01
-3.21861207e-01 6.41676545e-01 -1.69995353e-01 -1.03872681e+00
8.38450730e-01 -2.12875932e-01 -1.12427548e-01 -1.94290474e-01
-3.84177476e-01 -2.46360779e-01 6.10766649e-01 6.68572783e-01
1.03147519e+00 -1.93407476e-01 8.35199475e-01 -1.46614563e+00
2.87451655e-01 -7.29429483e-01 -6.89460516e-01 8.84761512e-01
-5.76176047e-01 3.93552661e-01 6.03687823e-01 8.88668001e-02
-9.68291283e-01 -2.81167656e-01 -1.47841617e-01 5.61964065e-02
-5.42960390e-02 8.15485001e-01 -5.28501153e-01 -1.04612947e-01
6.29927874e-01 -1.86780021e-01 -5.95098495e-01 -7.58011699e-01
5.71596384e-01 3.12540561e-01 1.10309851e+00 -1.06472719e+00
5.44745862e-01 8.61754715e-02 5.72067380e-01 -3.13885540e-01
-1.31268060e+00 -5.51181674e-01 -7.45702505e-01 1.77753657e-01
8.19695175e-01 -1.04102433e+00 -1.52692676e+00 -5.97056150e-02
-1.28983176e+00 1.00735798e-01 -2.27315336e-01 9.67373922e-02
1.02084592e-01 -1.22892462e-01 -1.09220243e+00 -1.43626392e-01
-3.96287620e-01 -6.18261755e-01 1.03503060e+00 3.84443820e-01
-3.70904624e-01 -1.16545725e+00 8.38849023e-02 4.32883203e-01
-5.93061633e-02 4.91235495e-01 1.52507412e+00 -1.15402555e+00
-9.27870512e-01 -4.35910821e-02 -5.47366500e-01 -7.16763258e-01
3.45507450e-02 2.68792629e-01 -7.55002439e-01 2.65811503e-01
-1.02590406e+00 -3.13801885e-01 1.16439676e+00 2.11495072e-01
1.60284913e+00 -3.93923491e-01 -1.01260078e+00 7.29209960e-01
1.30061269e+00 1.70514062e-01 4.39794213e-01 8.37783873e-01
9.82219636e-01 5.75356722e-01 1.99024901e-01 4.85666022e-02
6.52989745e-01 2.48845726e-01 -2.87924428e-02 -2.24734768e-01
-3.01145077e-01 -3.84132117e-01 -4.79272395e-01 7.84560919e-01
1.26494825e-01 -6.40412629e-01 -1.09409261e+00 6.50028586e-01
-1.45129812e+00 -9.47861850e-01 -6.01139128e-01 1.86912918e+00
1.70640540e+00 7.95020089e-02 -3.53046298e-01 -3.27285975e-01
5.42968512e-01 -3.94725740e-01 -7.46268272e-01 7.10647926e-03
-6.51549757e-01 2.32128963e-01 8.95878494e-01 4.12406862e-01
-1.10764647e+00 1.09476590e+00 7.06389427e+00 1.00982296e+00
-2.51744300e-01 -1.12131462e-01 7.84885466e-01 9.33390185e-02
-5.94885707e-01 2.68108258e-03 -1.68967783e+00 3.35967243e-01
8.82886291e-01 -4.99485940e-01 3.62567790e-02 4.16273743e-01
-3.49485338e-01 -6.30299002e-02 -1.46136785e+00 8.30294967e-01
-2.08334401e-01 -2.13921618e+00 3.91718775e-01 -6.32597459e-03
6.08633339e-01 -2.94876248e-01 -2.99410045e-01 4.99533266e-01
8.47388148e-01 -1.27134633e+00 3.31624538e-01 8.05485427e-01
9.91502106e-01 -6.43851101e-01 9.16384280e-01 -2.43728608e-01
-1.01319396e+00 5.54771125e-01 -3.99978071e-01 5.45712054e-01
-2.39562109e-01 1.10841763e+00 -8.92237127e-01 6.47362649e-01
9.59246099e-01 8.96778464e-01 -6.96963549e-01 9.33197141e-01
-3.17587346e-01 5.87674499e-01 -1.88857093e-01 -1.61876887e-01
-2.22426757e-01 8.96125883e-02 4.84543920e-01 1.80554223e+00
8.82795602e-02 2.73986936e-01 -1.01090461e-01 1.19996154e+00
-1.11293495e+00 1.64841294e-01 -5.11937976e-01 -5.83291173e-01
1.01258183e+00 1.56910491e+00 -1.01256216e+00 -6.63354456e-01
-1.12114534e-01 2.67701358e-01 6.89946651e-01 3.61156225e-01
-4.97748554e-01 -8.27113748e-01 6.98705554e-01 -2.33428180e-01
1.63320273e-01 8.13181326e-02 -7.30215788e-01 -1.02323580e+00
-2.85567760e-01 -6.16042852e-01 9.14081216e-01 -7.74355710e-01
-1.58623624e+00 2.68225014e-01 -3.10284615e-01 -4.65254515e-01
7.58086294e-02 -7.24964261e-01 8.16473812e-02 9.07157242e-01
-1.56626546e+00 -1.12802529e+00 -3.11742961e-01 1.39524370e-01
5.08960597e-02 -1.29521325e-01 1.10451806e+00 4.30615067e-01
-6.41832829e-01 7.48370647e-01 5.26531577e-01 4.62648094e-01
1.38075185e+00 -1.74671757e+00 5.48365593e-01 3.19514841e-01
8.17172304e-02 9.80582535e-01 7.65413761e-01 -1.27291739e+00
-1.70539641e+00 -1.21208179e+00 1.60550201e+00 -1.30297601e+00
9.65085089e-01 -6.17251456e-01 -1.25242770e+00 8.33846629e-01
2.64691800e-01 -1.68753803e-01 9.34160709e-01 6.08443081e-01
-6.46666110e-01 -2.10811961e-02 -1.06509244e+00 5.40279448e-01
1.30221570e+00 -4.33742911e-01 -5.95271826e-01 7.59382248e-01
8.17959547e-01 -8.97092044e-01 -1.67977512e+00 1.95421711e-01
3.83151680e-01 1.43238172e-01 1.34624875e+00 -8.63848686e-01
6.20607316e-01 -5.16866207e-01 2.82420695e-01 -1.11145353e+00
-6.08728349e-01 -3.74066263e-01 -6.65651441e-01 1.69074500e+00
9.61550534e-01 -2.56482422e-01 8.49370778e-01 5.80824494e-01
-2.71746427e-01 -6.58564568e-01 -5.34212112e-01 -7.38610446e-01
5.29021680e-01 3.53065550e-01 5.35090864e-01 1.38447261e+00
2.37078786e-01 4.97868985e-01 3.90685260e-01 7.39904791e-02
7.90711999e-01 2.85585761e-01 4.40350831e-01 -1.51280367e+00
2.23189503e-01 -6.01267397e-01 -1.02792889e-01 -5.83981991e-01
4.41435009e-01 -1.42845225e+00 -8.38204026e-02 -2.29475260e+00
4.84488994e-01 -5.96692443e-01 -3.03023100e-01 9.36882794e-01
-6.09690070e-01 3.68278503e-04 -1.72529101e-01 2.96354949e-01
-8.66145015e-01 2.25906577e-02 9.73985791e-01 -4.61524159e-01
1.53628320e-01 -7.85374165e-01 -1.06616604e+00 3.98075759e-01
1.97744995e-01 -6.27883494e-01 1.91426411e-01 -1.42754018e-01
6.01607144e-01 -3.77331465e-03 -9.19684246e-02 -8.02249074e-01
1.02039146e+00 -2.05682680e-01 9.18172240e-01 -1.07338500e+00
-4.74404693e-02 -3.02876472e-01 2.63008386e-01 6.93168491e-02
-7.62099564e-01 -1.37407528e-02 6.50794208e-01 4.58496511e-01
-2.70105861e-02 8.27387124e-02 2.27073371e-01 -4.64118034e-01
-5.98090172e-01 2.06484705e-01 -2.48555616e-01 6.51311219e-01
5.66265762e-01 3.53236318e-01 -1.28548801e+00 1.83519736e-01
-5.94866574e-01 7.99297392e-01 3.24567169e-01 2.03177631e-01
1.52852342e-01 -1.05997884e+00 -8.63140404e-01 -6.38848901e-01
3.31764907e-01 1.28551304e-01 -1.87514454e-01 4.18019831e-01
-5.89475870e-01 9.53932881e-01 -1.92917474e-02 -1.28245980e-01
-1.10094559e+00 8.34918678e-01 -1.00184813e-01 -6.05097771e-01
-3.49951863e-01 1.24471235e+00 4.78604361e-02 -6.21562719e-01
5.96621156e-01 -3.72952998e-01 -4.91035819e-01 3.36214393e-01
5.72686911e-01 4.31984901e-01 4.46219116e-01 -3.05386521e-02
-7.20106304e-01 2.40821630e-01 -3.12536091e-01 3.72609317e-01
1.28210902e+00 1.66545808e-01 -7.53371537e-01 7.75079846e-01
8.66798043e-01 5.25337875e-01 -3.76755208e-01 -2.52125919e-01
3.81470501e-01 3.39027606e-02 -1.11253269e-01 -1.62291384e+00
-9.33911860e-01 3.82754952e-01 -3.13648611e-01 -3.73227298e-01
5.03984451e-01 3.63489091e-01 6.52362764e-01 1.01579833e+00
1.35190576e-01 -8.43372822e-01 -5.74262142e-01 6.63352132e-01
7.71401346e-01 -1.20703638e+00 3.99808615e-01 -9.57204759e-01
-1.58876687e-01 1.20577133e+00 7.73258746e-01 5.92999220e-01
4.80372548e-01 7.44909465e-01 -5.45294397e-02 -6.11463428e-01
-9.89905596e-01 -6.38267696e-02 4.83039021e-01 4.68302816e-01
7.94860840e-01 -1.20790832e-01 -2.82839954e-01 8.03379357e-01
-2.81409800e-01 4.54961210e-02 5.01438260e-01 8.23226929e-01
-8.49808082e-02 -1.00773680e+00 -3.67273808e-01 9.03837800e-01
-8.54218066e-01 -5.59815943e-01 -8.48704219e-01 8.34957600e-01
2.49128565e-01 6.16651952e-01 3.39381136e-02 1.78657666e-01
3.19055647e-01 3.05470884e-01 3.62746984e-01 -5.32264411e-01
-8.70517254e-01 -3.32701385e-01 4.39456403e-01 -1.30749911e-01
-2.58452624e-01 -6.55481458e-01 -1.67344236e+00 -8.25689495e-01
4.72835684e-03 6.79236412e-01 5.62243164e-01 6.09622061e-01
6.50362492e-01 9.41392958e-01 -4.56012607e-01 2.54487991e-03
2.49331445e-01 -6.90770149e-01 -4.72548455e-01 3.00978690e-01
5.91859296e-02 -7.17159808e-01 -2.25969926e-01 3.36332560e-01] | [9.045316696166992, 8.379312515258789] |
1be6a349-1251-4229-a575-632826394686 | tp-lsd-tri-points-based-line-segment-detector-1 | 2009.05505 | null | https://arxiv.org/abs/2009.05505v1 | https://arxiv.org/pdf/2009.05505v1.pdf | TP-LSD: Tri-Points Based Line Segment Detector | This paper proposes a novel deep convolutional model, Tri-Points Based Line Segment Detector (TP-LSD), to detect line segments in an image at real-time speed. The previous related methods typically use the two-step strategy, relying on either heuristic post-process or extra classifier. To realize one-step detection with a faster and more compact model, we introduce the tri-points representation, converting the line segment detection to the end-to-end prediction of a root-point and two endpoints for each line segment. TP-LSD has two branches: tri-points extraction branch and line segmentation branch. The former predicts the heat map of root-points and the two displacement maps of endpoints. The latter segments the pixels on straight lines out from background. Moreover, the line segmentation map is reused in the first branch as structural prior. We propose an additional novel evaluation metric and evaluate our method on Wireframe and YorkUrban datasets, demonstrating not only the competitive accuracy compared to the most recent methods, but also the real-time run speed up to 78 FPS with the $320\times 320$ input. | ['Fangbo Qin', 'Xiao Liu', 'Siyu Huang', 'Pengfei Xiong', 'Yijia He', 'Ning Ding'] | 2020-09-11 | tp-lsd-tri-points-based-line-segment-detector | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/5931_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123720766.pdf | eccv-2020-8 | ['line-segment-detection'] | ['computer-vision'] | [ 1.57863468e-01 -2.30854750e-02 -3.11640091e-02 -2.92216480e-01
-7.09477246e-01 -4.68665749e-01 4.17653531e-01 6.14844978e-01
-6.34960473e-01 3.90370309e-01 -6.29148364e-01 -5.47563434e-01
1.82199925e-01 -9.16388512e-01 -7.55689859e-01 -3.36212486e-01
-1.65386960e-01 2.42555216e-01 1.04545462e+00 -1.27427161e-01
5.48995078e-01 9.39185858e-01 -1.28286707e+00 1.66143596e-01
8.88989568e-01 1.40783525e+00 -2.66541004e-01 6.14671111e-01
-3.92767936e-01 3.28270525e-01 -3.52749527e-01 -4.66708779e-01
5.35172939e-01 -3.25089172e-02 -4.14130926e-01 1.11529887e-01
4.99936551e-01 -3.91313285e-01 -1.52742431e-01 8.59436035e-01
4.44365561e-01 -2.45847553e-01 6.44363225e-01 -1.13020563e+00
3.24377477e-01 2.12837517e-01 -1.35536873e+00 5.52641042e-02
2.00216621e-01 1.01204701e-01 7.34713435e-01 -8.71894002e-01
4.89315182e-01 9.89044487e-01 1.09048724e+00 -1.76794037e-01
-8.91296387e-01 -5.93479872e-01 -8.04034173e-02 1.00128718e-01
-1.54349303e+00 -7.96915889e-02 8.19861948e-01 -6.79224908e-01
6.99466705e-01 1.49130270e-01 4.84102905e-01 4.16547567e-01
1.45502329e-01 7.89784968e-01 7.16238618e-01 -5.40804207e-01
-4.88904305e-03 -9.54026282e-02 2.76770800e-01 9.54047263e-01
7.72223547e-02 5.75610362e-02 7.06775952e-03 1.87276945e-01
1.15562952e+00 -2.80154794e-01 -2.22028270e-01 -4.72394288e-01
-1.09767830e+00 5.87649941e-01 6.15861893e-01 8.69190693e-02
-2.80757248e-01 5.23129851e-02 5.23804486e-01 -1.42515838e-01
2.39503697e-01 -2.89500542e-02 -3.49059731e-01 -3.17385942e-02
-1.38193607e+00 1.62089646e-01 5.86164296e-01 1.10949123e+00
9.32202399e-01 -3.52118194e-01 -3.69466037e-01 5.80834866e-01
1.18994378e-01 2.25494593e-01 1.94317654e-01 -4.59542155e-01
8.15347493e-01 8.61145973e-01 2.98143059e-01 -1.30125701e+00
-8.90774131e-01 -8.06423187e-01 -6.63109243e-01 5.21044433e-01
6.73337281e-01 -4.90727782e-01 -7.26987660e-01 9.69068587e-01
5.67164242e-01 1.14664473e-01 -3.90606701e-01 8.86350989e-01
6.27483785e-01 9.14189458e-01 -2.29171559e-01 -6.13932945e-02
1.45430243e+00 -1.28503251e+00 -2.71001637e-01 -3.96100320e-02
9.57522392e-01 -9.97699678e-01 7.32760429e-01 5.83102822e-01
-1.07421374e+00 -7.78301299e-01 -1.24087024e+00 -2.63526589e-01
-2.07652181e-01 8.79217863e-01 3.40998769e-01 4.99357015e-01
-6.57278538e-01 7.14539468e-01 -7.21531928e-01 -1.75712928e-01
5.02595663e-01 2.56884634e-01 -1.78668037e-01 3.69610518e-01
-7.36116469e-01 3.93626213e-01 2.56942123e-01 3.04862887e-01
-1.10109828e-01 -5.51888943e-01 -7.78395832e-01 4.55465406e-01
5.36373198e-01 -3.21921140e-01 1.03681695e+00 -9.89773631e-01
-1.41143692e+00 9.03309047e-01 -6.20880052e-02 -6.06926203e-01
1.17865837e+00 -5.26349783e-01 -1.62618682e-01 3.60557646e-01
1.07279949e-01 5.50585628e-01 6.74826026e-01 -1.17223203e+00
-1.38111448e+00 -1.27189249e-01 -1.04827188e-01 1.08370565e-01
7.27787763e-02 -2.92348623e-01 -8.88936877e-01 -5.84298074e-01
2.54784137e-01 -6.42073393e-01 -1.19804218e-01 3.69345367e-01
-9.64199722e-01 -3.93797636e-01 9.52447116e-01 -6.41292036e-01
1.51101756e+00 -2.16415524e+00 -5.62223732e-01 4.94382650e-01
1.82116762e-01 5.90439022e-01 1.04945764e-01 4.67121750e-01
-2.28041917e-01 -1.21593900e-01 -7.34399334e-02 -3.85359913e-01
-2.91226208e-01 -5.31934142e-01 -1.34338245e-01 5.37179351e-01
4.47983474e-01 5.71409225e-01 -5.90384603e-01 -7.05523551e-01
4.07742441e-01 2.41732121e-01 -1.03656918e-01 -1.27338648e-01
1.76790863e-01 2.62861531e-02 -3.53970140e-01 4.73660350e-01
1.08800054e+00 2.98771393e-02 -3.10569584e-01 -4.89413768e-01
-8.14193189e-01 -6.06269464e-02 -1.53910458e+00 1.45283377e+00
-3.83621976e-02 8.47043872e-01 -3.04348320e-01 -6.45457625e-01
1.38605690e+00 -1.33536473e-01 5.40965140e-01 -8.20578694e-01
2.91684777e-01 3.10322702e-01 -1.40546365e-02 -4.71286237e-01
5.34909725e-01 5.37295401e-01 1.14337973e-01 -4.66106907e-02
-4.79478151e-01 8.61599073e-02 4.09963250e-01 -6.34350330e-02
8.47556651e-01 6.14434898e-01 2.19586119e-01 -1.18428089e-01
7.81643629e-01 1.84760541e-01 6.24156833e-01 5.58741212e-01
-2.02857479e-01 7.71005392e-01 8.42262328e-01 -5.93098104e-01
-1.01500916e+00 -7.45132565e-01 -1.21568382e-01 6.22796178e-01
5.77561200e-01 -5.27163923e-01 -1.01271570e+00 -7.39172935e-01
-1.59844935e-01 6.17501915e-01 -3.43678087e-01 2.35998362e-01
-8.95868957e-01 -1.82851434e-01 4.09866631e-01 7.21125364e-01
8.90911043e-01 -6.89140558e-01 -1.02244031e+00 3.22575271e-01
1.38436362e-01 -1.19017851e+00 -5.10620892e-01 2.43577167e-01
-7.78066278e-01 -1.15588009e+00 -8.71739209e-01 -1.03406751e+00
6.52869463e-01 2.69985348e-01 6.59897625e-01 1.72600806e-01
-3.81160527e-01 -3.99337232e-01 -2.89561898e-01 -2.79592156e-01
1.69007540e-01 5.01540117e-02 -5.73026955e-01 2.39674121e-01
2.02558666e-01 -5.44903316e-02 -1.00100243e+00 6.10924900e-01
-3.22463930e-01 6.11160755e-01 7.69547403e-01 4.62320834e-01
8.12560022e-01 -5.39941713e-02 1.09727733e-01 -6.94297016e-01
1.51134893e-01 5.86815812e-02 -1.02743042e+00 1.85502157e-01
-3.01742822e-01 -1.97686598e-01 6.29150212e-01 -1.24142542e-02
-7.76978076e-01 5.04365385e-01 -3.08345467e-01 -3.58851492e-01
-3.51705939e-01 3.66243869e-01 1.45351604e-01 3.95544656e-02
6.55489683e-01 9.10127610e-02 -3.21014613e-01 -3.26576144e-01
3.01741719e-01 6.47113323e-01 7.75636554e-01 -2.14690015e-01
8.19697320e-01 6.18779838e-01 2.22115457e-01 -9.16819811e-01
-5.02819121e-01 -7.04156041e-01 -1.10062659e+00 -3.52381051e-01
7.26274252e-01 -5.58615983e-01 -6.15637600e-01 6.14998519e-01
-1.45584965e+00 -2.64837891e-01 -2.60511100e-01 2.55086780e-01
-4.73037660e-01 6.71831131e-01 -6.55631125e-01 -7.03208685e-01
-5.62421739e-01 -1.23737574e+00 1.21139860e+00 5.79565287e-01
2.38667756e-01 -3.57117295e-01 -1.00026771e-01 -1.73923090e-01
-7.13885874e-02 6.33379877e-01 8.69914889e-01 -6.15616739e-01
-6.30930603e-01 -6.40669405e-01 -6.59267247e-01 1.37418196e-01
-3.16644430e-01 6.51796997e-01 -7.23842561e-01 -5.21898689e-03
-5.66372335e-01 1.54866099e-01 6.86015308e-01 4.43489611e-01
1.18855250e+00 2.72246569e-01 -7.16665745e-01 8.48099828e-01
1.64322674e+00 4.10843790e-01 7.69908726e-01 5.15105307e-01
5.87556541e-01 5.90575516e-01 9.20929372e-01 4.46952254e-01
2.81482190e-01 7.85903752e-01 3.47992420e-01 -7.31529117e-01
-8.62416178e-02 -1.86868697e-01 -7.78672546e-02 2.63097107e-01
-2.01566190e-01 -4.80291039e-01 -1.02270281e+00 2.50225812e-01
-1.91547263e+00 -7.10671723e-01 -1.03884578e+00 2.51569939e+00
2.83158183e-01 7.19616592e-01 4.02742177e-01 4.93564039e-01
9.04226899e-01 -1.94896817e-01 -4.55950677e-01 -3.97306800e-01
-2.13826708e-02 6.38386756e-02 8.24722886e-01 2.02290073e-01
-1.51624596e+00 9.73705590e-01 4.92716789e+00 1.10859597e+00
-1.38271773e+00 -5.11996984e-01 8.20553243e-01 5.08800030e-01
2.96941310e-01 6.76516518e-02 -1.07181239e+00 2.62241781e-01
2.78552204e-01 4.12110597e-01 -3.23486745e-01 8.86075437e-01
4.60956633e-01 -3.62251490e-01 -8.83600533e-01 1.20623374e+00
-1.42659128e-01 -1.27684760e+00 -2.74530977e-01 -1.58725455e-01
3.80396426e-01 -1.83444515e-01 -2.92658359e-01 5.61939850e-02
-4.45618004e-01 -5.33681452e-01 1.06745362e+00 4.91027623e-01
9.56840873e-01 -9.75073218e-01 5.54047644e-01 4.41650212e-01
-1.46580768e+00 -9.41846333e-03 -1.67722315e-01 2.27701381e-01
3.00788283e-01 5.37022352e-01 -9.18189466e-01 7.44838715e-01
4.74131733e-01 5.73352039e-01 -6.74447715e-01 1.69524443e+00
-3.04636091e-01 5.71299613e-01 -6.55884922e-01 8.15312266e-02
4.56924379e-01 -3.79351407e-01 4.32855189e-01 1.67713058e+00
3.50801528e-01 -3.45186293e-01 1.80359572e-01 7.59991586e-01
2.99902827e-01 5.73220491e-01 -6.40654042e-02 4.27172452e-01
3.68167102e-01 1.55569041e+00 -1.39154351e+00 -4.48471487e-01
-4.77949113e-01 9.86714959e-01 3.01382095e-01 2.21410692e-01
-1.12267566e+00 -1.18373048e+00 -4.38727345e-03 5.72747886e-01
4.90043968e-01 -5.11662602e-01 -5.47427595e-01 -6.58926010e-01
1.10371165e-01 -2.65212357e-01 1.88554853e-01 -6.67301416e-01
-6.27948523e-01 6.55979276e-01 -1.46935076e-01 -1.52280223e+00
-7.75007010e-02 -5.80971003e-01 -8.20206046e-01 8.23281407e-01
-1.68973565e+00 -1.24407876e+00 -7.16547728e-01 3.72099668e-01
7.15802252e-01 2.87309229e-01 2.71158665e-01 3.58521283e-01
-8.78943563e-01 8.87438059e-01 1.59839153e-01 5.57491779e-01
5.91713548e-01 -1.12224317e+00 6.92166328e-01 1.06570280e+00
-1.28971919e-01 1.72967151e-01 4.04969096e-01 -7.09096014e-01
-7.07224667e-01 -7.99391687e-01 6.27772868e-01 2.24244371e-01
2.56992817e-01 -4.14723366e-01 -8.50643158e-01 4.28384513e-01
-1.09344527e-01 -7.42764771e-02 1.65275589e-01 -3.75875115e-01
-1.05427980e-01 -1.88642174e-01 -9.50874925e-01 5.34047604e-01
7.24923432e-01 2.54198968e-01 -2.01780334e-01 3.37105580e-02
4.44482684e-01 -5.97180307e-01 -5.06985366e-01 4.79791760e-01
6.75797164e-01 -1.35668349e+00 9.00917292e-01 4.58331630e-02
6.01770699e-01 -6.66893721e-01 5.76579452e-01 -7.36479759e-01
-3.09320420e-01 -7.32933402e-01 2.30733246e-01 1.25495434e+00
3.51168752e-01 -2.51078218e-01 8.81737590e-01 2.09482610e-01
-5.14782488e-01 -1.16879046e+00 -7.80702174e-01 -5.36877573e-01
-2.42376640e-01 -3.81963164e-01 5.69680870e-01 5.85843325e-01
-3.29538941e-01 2.40927935e-01 -1.24602932e-02 2.99224705e-01
4.66484815e-01 1.65509894e-01 1.08820462e+00 -1.09769583e+00
1.31694907e-02 -7.44794250e-01 -6.45460844e-01 -1.63509524e+00
-5.60063064e-01 -4.31646734e-01 2.81812325e-02 -1.49128687e+00
-4.56982106e-01 -6.84639156e-01 8.34237859e-02 3.80603224e-01
-3.37943695e-02 9.98721644e-02 1.19802885e-01 1.96151271e-01
-4.43144113e-01 1.27884552e-01 1.12620962e+00 2.14713857e-01
-4.69670802e-01 2.33020023e-01 -6.36153147e-02 1.10668385e+00
5.50667107e-01 -1.08862005e-01 -1.15445562e-01 -2.42249757e-01
3.05783339e-02 8.72910470e-02 2.15080351e-01 -1.52243769e+00
4.73704845e-01 8.93627331e-02 6.94785178e-01 -1.35314155e+00
1.63059071e-01 -7.32280910e-01 -3.89946014e-01 6.46246731e-01
-1.42477170e-01 9.31581259e-02 3.06925893e-01 3.74953896e-01
2.54456364e-02 -5.85283637e-01 9.36885893e-01 2.92718381e-01
-8.99418116e-01 2.59808421e-01 -1.33458853e-01 -4.45532173e-01
1.68571198e+00 -8.38943958e-01 -2.98459053e-01 -9.58413444e-03
-5.12025297e-01 3.37713957e-01 3.54019850e-01 1.32297382e-01
7.05788970e-01 -1.04793000e+00 -5.89354396e-01 4.19550717e-01
4.07319628e-02 4.07113582e-01 2.35469699e-01 1.09187698e+00
-1.44489276e+00 3.18255275e-01 -1.19448058e-01 -7.51832426e-01
-1.13368976e+00 5.36974072e-01 5.26080549e-01 -1.16293505e-01
-9.21078563e-01 7.65934944e-01 1.86095223e-01 4.10188824e-01
2.86505461e-01 -6.90152287e-01 -4.33959693e-01 4.89699058e-02
3.92307878e-01 6.03197992e-01 7.86660165e-02 -6.34176970e-01
-1.22022323e-01 1.07101440e+00 -1.27027348e-01 7.52439052e-02
8.86838913e-01 -4.45434116e-02 2.82321364e-01 1.97516516e-01
1.06720662e+00 1.34531349e-01 -1.52233231e+00 -8.37840885e-02
9.05121863e-02 -5.85749090e-01 -7.31086582e-02 -6.35303676e-01
-9.20724154e-01 1.00072396e+00 8.45581472e-01 2.75287837e-01
8.79264295e-01 -4.82759804e-01 1.14619386e+00 -4.65195961e-02
2.37783581e-01 -1.06785429e+00 -2.21708566e-01 2.93241590e-01
6.99252009e-01 -9.79497313e-01 5.50148673e-02 -9.97722924e-01
-3.28095496e-01 1.87214804e+00 7.39050448e-01 -4.20391113e-01
4.54437345e-01 2.27305397e-01 -3.07056606e-02 -6.09660856e-02
-4.39786017e-02 -1.93875045e-01 3.70930523e-01 3.37967902e-01
3.03207070e-01 -2.27626994e-01 -5.44454396e-01 2.82089442e-01
-1.04942217e-01 6.43774718e-02 3.10729653e-01 8.31143141e-01
-6.28166020e-01 -8.20351958e-01 -4.14766580e-01 3.77192527e-01
-1.90597549e-01 1.50399253e-01 -8.86254013e-02 1.05752540e+00
3.47817928e-01 6.63600683e-01 5.49558520e-01 -4.23778206e-01
5.64911425e-01 -3.02752316e-01 2.37304103e-02 -2.67730683e-01
-6.20300770e-01 5.34621000e-01 4.02660370e-02 -6.39969885e-01
-4.52834629e-02 -5.62791288e-01 -1.68529320e+00 -7.12231770e-02
-6.26028121e-01 -9.92274657e-02 8.28658044e-01 6.79245949e-01
4.27540123e-01 2.63619304e-01 7.93689668e-01 -1.06706202e+00
-1.68747395e-01 -7.30127692e-01 -2.87139416e-01 3.24344605e-01
2.11077064e-01 -5.65277159e-01 -1.32503211e-01 5.74462255e-03] | [8.30942440032959, -1.5278304815292358] |
6a6e18cf-ce67-44f4-90e6-4a1e1b003be4 | deep-speech-synthesis-from-mri-based | 2307.02471 | null | https://arxiv.org/abs/2307.02471v1 | https://arxiv.org/pdf/2307.02471v1.pdf | Deep Speech Synthesis from MRI-Based Articulatory Representations | In this paper, we study articulatory synthesis, a speech synthesis method using human vocal tract information that offers a way to develop efficient, generalizable and interpretable synthesizers. While recent advances have enabled intelligible articulatory synthesis using electromagnetic articulography (EMA), these methods lack critical articulatory information like excitation and nasality, limiting generalization capabilities. To bridge this gap, we propose an alternative MRI-based feature set that covers a much more extensive articulatory space than EMA. We also introduce normalization and denoising procedures to enhance the generalizability of deep learning methods trained on MRI data. Moreover, we propose an MRI-to-speech model that improves both computational efficiency and speech fidelity. Finally, through a series of ablations, we show that the proposed MRI representation is more comprehensive than EMA and identify the most suitable MRI feature subset for articulatory synthesis. | ['Gopala K. Anumanchipalli', 'Shinji Watanabe', 'Louis Goldstein', 'Alan W Black', 'Jiachen Lian', 'Yubin Zhang', 'Yijing Lu', 'Tingle Li', 'Peter Wu'] | 2023-07-05 | null | null | null | null | ['denoising', 'speech-synthesis'] | ['computer-vision', 'speech'] | [ 2.16935366e-01 1.71359584e-01 -7.91595131e-02 -2.17254698e-01
-8.97660255e-01 -6.05834723e-01 7.78090477e-01 -5.19224107e-01
-1.13537483e-01 6.36463463e-01 7.77760923e-01 -1.63497224e-01
-2.35654801e-01 -3.34912062e-01 -4.11969543e-01 -7.57295609e-01
1.40777186e-01 -1.76729802e-02 -2.94125974e-01 9.47222672e-03
-1.91946432e-01 7.31825233e-01 -1.44278395e+00 2.07838118e-01
9.10848498e-01 8.33398342e-01 6.34572029e-01 4.19306219e-01
-6.88642487e-02 2.89670289e-01 -7.16055870e-01 -3.06987971e-01
5.58853596e-02 -6.39166772e-01 -6.66076243e-01 4.31569805e-03
2.93201178e-01 -3.65082741e-01 -5.78103483e-01 8.34476471e-01
9.85891163e-01 2.23174572e-01 8.43883812e-01 -7.28991151e-01
-8.20165277e-01 6.66803002e-01 -3.74541851e-03 7.68768787e-02
8.79290625e-02 1.52250707e-01 8.35390091e-01 -9.43547368e-01
7.47040391e-01 1.06141758e+00 5.70014894e-01 1.10086977e+00
-8.85567367e-01 -4.34325129e-01 -3.00704062e-01 -4.88128923e-02
-8.80354702e-01 -8.85283589e-01 1.02897191e+00 -2.70799935e-01
8.39106321e-01 5.17202437e-01 4.92444783e-01 1.33534527e+00
2.67758936e-01 9.22880769e-01 1.19254005e+00 -3.55459780e-01
1.03635758e-01 -3.04106683e-01 -2.83079058e-01 6.76481843e-01
-1.60486966e-01 3.46011609e-01 -4.28726077e-01 -5.18159829e-02
7.60189950e-01 -3.22775304e-01 -4.47098911e-01 -7.75273219e-02
-1.38816130e+00 7.96217918e-01 5.84360398e-02 5.10526776e-01
-5.02507627e-01 3.33041906e-01 4.55664337e-01 1.83958113e-01
3.89528126e-01 7.43223906e-01 8.72160792e-02 -1.09107196e-01
-1.19739032e+00 -4.01082337e-02 4.54587162e-01 3.57437253e-01
1.08265271e-02 8.28637362e-01 -1.39402568e-01 1.47487080e+00
3.08511704e-01 9.53920782e-01 7.92801142e-01 -1.17457759e+00
-5.19157248e-03 -3.17127079e-01 -4.55471247e-01 -8.05709243e-01
-7.04298198e-01 -8.09562564e-01 -5.84987879e-01 1.05638526e-01
-2.80740652e-02 -1.61819071e-01 -8.76102090e-01 1.87049913e+00
1.32894842e-02 6.85072765e-02 -7.90467337e-02 1.11907136e+00
1.14409590e+00 4.81072038e-01 5.18084578e-02 -3.51909876e-01
1.28932989e+00 -1.20939147e+00 -1.16808736e+00 -2.61908591e-01
2.01031163e-01 -9.78779614e-01 8.69978487e-01 2.04573661e-01
-1.58975887e+00 -4.89647686e-01 -1.17033088e+00 -3.58521603e-02
-1.22651860e-01 3.44520569e-01 9.45550621e-01 8.41341496e-01
-1.22237146e+00 5.91616631e-01 -9.85817015e-01 3.90264345e-03
1.32190764e-01 3.84435952e-01 -5.27721941e-01 2.69543350e-01
-1.20725906e+00 1.17678332e+00 -2.47437283e-01 9.80428681e-02
-7.99180031e-01 -6.71673596e-01 -1.17090619e+00 -1.26857266e-01
2.48575024e-02 -9.53050196e-01 1.19907844e+00 -4.56048489e-01
-1.97838938e+00 8.04918349e-01 -3.12226057e-01 -4.49749321e-01
2.33434349e-01 1.03759542e-01 -6.90594673e-01 5.36951721e-01
-1.26478717e-01 6.48075759e-01 1.25134575e+00 -1.15232718e+00
1.96192876e-01 -2.19292641e-01 -2.94119239e-01 1.27896070e-01
-1.87244311e-01 1.72575295e-01 -2.27281511e-01 -1.16660404e+00
3.81523967e-02 -1.06131935e+00 -5.73167317e-02 6.60753101e-02
-2.39791393e-01 9.37393829e-02 3.33610833e-01 -9.11959052e-01
1.13078415e+00 -2.07093883e+00 3.17711681e-01 9.86468866e-02
4.85982597e-01 4.92354214e-01 -5.38465679e-01 2.89553732e-01
-3.39749664e-01 5.33170700e-02 -5.10986745e-01 -4.81513292e-01
5.53746708e-04 3.08603823e-01 -4.37072843e-01 5.33330202e-01
1.67696834e-01 1.23896217e+00 -7.96208799e-01 -3.04498285e-01
5.81661999e-01 6.61875725e-01 -6.27598107e-01 -1.24743260e-01
3.30610722e-01 5.89231133e-01 -4.44312930e-01 5.73694110e-01
6.07628405e-01 2.46702567e-01 -3.83191183e-03 -3.04814905e-01
4.71719392e-02 4.82181102e-01 -6.35903835e-01 2.02657437e+00
-1.11789453e+00 6.91641569e-01 5.28549254e-01 -1.19644129e+00
9.05543864e-01 5.36690176e-01 7.49095678e-01 -6.06017888e-01
9.34719145e-02 5.31692743e-01 2.73281395e-01 -7.02715158e-01
1.82052508e-01 -4.97660577e-01 1.30003691e-01 3.42899531e-01
4.09765989e-01 -5.73635757e-01 -1.66088045e-01 -2.66220897e-01
9.23166931e-01 3.69125442e-03 1.81023911e-01 -3.62461984e-01
6.38888597e-01 -5.37276685e-01 2.92887896e-01 6.86731756e-01
-3.30387145e-01 9.65350211e-01 -2.26473942e-01 6.98795691e-02
-8.98583531e-01 -1.55533814e+00 -5.92136443e-01 6.06971562e-01
-3.78183573e-01 -2.83472538e-01 -9.98901010e-01 -3.80089223e-01
-1.88842475e-01 6.48432791e-01 -3.23841721e-01 -2.57096827e-01
-8.08303654e-01 -6.24750018e-01 7.18437076e-01 4.60074008e-01
8.02142471e-02 -1.30734110e+00 -1.83780611e-01 3.69818658e-01
-5.11363506e-01 -1.16858697e+00 -7.73954749e-01 -8.81664008e-02
-8.16808164e-01 -4.45823371e-01 -1.08655047e+00 -8.06629837e-01
4.14248973e-01 8.04737490e-03 8.15406084e-01 -1.27446130e-01
-4.77632612e-01 6.27098143e-01 -1.05790891e-01 -2.64751256e-01
-9.02220309e-01 -6.09373152e-02 4.64377999e-01 -2.91025668e-01
-2.37434760e-01 -8.15013289e-01 -5.80200374e-01 1.26646131e-01
-8.60429049e-01 4.24917899e-02 6.22032762e-01 1.00945926e+00
4.42336082e-01 -4.86990124e-01 1.36993730e+00 -3.79613250e-01
1.16013551e+00 -2.83802390e-01 1.25165224e-01 1.22648723e-01
-3.17525387e-01 2.32948110e-01 7.40582466e-01 -3.84871066e-01
-1.04266810e+00 -2.13286623e-01 -9.25116062e-01 -3.43797445e-01
-1.19843133e-01 6.15985632e-01 1.05404958e-01 -1.58187732e-01
6.25207305e-01 4.90714282e-01 6.12169921e-01 -4.59263951e-01
5.76078832e-01 8.28722477e-01 9.99601662e-01 -5.87374508e-01
7.24382639e-01 5.64546347e-01 -1.61584616e-02 -1.09613109e+00
-2.84566492e-01 -2.22457796e-01 -3.95868421e-01 -3.50682706e-01
7.93288410e-01 -6.27606392e-01 -5.96515357e-01 3.42631012e-01
-1.04884827e+00 2.93217991e-02 -3.64182591e-01 9.96105671e-01
-9.50923920e-01 5.97886622e-01 -7.60779083e-01 -7.70163298e-01
-7.56463349e-01 -1.59960008e+00 1.18950605e+00 -1.97218046e-01
-5.20394087e-01 -1.02640486e+00 -5.43173440e-02 6.51600480e-01
7.72567987e-01 1.78888381e-01 8.46039534e-01 -4.24215585e-01
-7.30098784e-02 7.98883513e-02 1.77342296e-01 5.89116096e-01
5.17992616e-01 -2.37831369e-01 -9.94676173e-01 -1.25759378e-01
4.16748941e-01 -9.78258178e-02 9.07874048e-01 7.81426132e-01
1.30851185e+00 -9.58324820e-02 -6.51016384e-02 7.79639125e-01
5.56160033e-01 4.30163532e-01 5.88937461e-01 -1.45983085e-01
2.79545188e-01 6.29602075e-01 -2.98877619e-02 -8.21939334e-02
4.40867841e-02 7.39129722e-01 -3.23091559e-02 -2.53888845e-01
-1.19142294e+00 -1.72915190e-01 4.31648999e-01 1.90640724e+00
-1.76383168e-01 1.73606634e-01 -5.43745935e-01 7.18873680e-01
-1.18886745e+00 -8.71792734e-01 3.00672889e-01 1.82010198e+00
9.68741357e-01 -3.62710387e-01 -1.06289312e-01 5.53361475e-02
5.42110980e-01 4.21791762e-01 -4.47429776e-01 -5.72025955e-01
-2.58415908e-01 8.82519066e-01 7.23069757e-02 7.05269277e-01
-7.63435483e-01 8.01436603e-01 7.54127789e+00 1.18648827e+00
-1.44871163e+00 2.08083972e-01 7.13140368e-02 -2.41468605e-02
-7.13375211e-01 -7.72255957e-01 -1.46722391e-01 3.71083915e-01
9.87948239e-01 -1.57031372e-01 1.02083528e+00 6.05282784e-01
4.68905091e-01 5.07522941e-01 -8.56163263e-01 9.98586535e-01
2.35779107e-01 -1.55273473e+00 -4.13737260e-02 -1.00970104e-01
5.22518218e-01 1.50921822e-01 4.43527818e-01 5.75951859e-03
-3.11366826e-01 -1.32798946e+00 5.12472332e-01 4.61716264e-01
1.16151071e+00 -5.08933127e-01 2.89594382e-01 5.89149669e-02
-9.45700169e-01 1.42336890e-01 -1.83279783e-01 3.09025258e-01
2.57471561e-01 4.69813317e-01 -8.08057189e-01 6.60610855e-01
6.33475557e-02 4.00373161e-01 -1.50326103e-01 9.17180657e-01
-1.90805182e-01 6.28748775e-01 -1.62790880e-01 5.16059510e-02
1.34565607e-01 -1.73285767e-01 8.95625889e-01 1.21169853e+00
5.63927889e-01 -2.58768369e-02 -3.89267713e-01 1.08933079e+00
-1.51071727e-01 2.44979948e-01 -7.65781879e-01 -3.35334122e-01
3.15700978e-01 1.27552330e+00 -4.36074644e-01 5.13913222e-02
-4.66785938e-01 8.09032738e-01 -1.07802421e-01 4.18848187e-01
-6.01971328e-01 -1.93805203e-01 7.96599507e-01 -5.70077077e-02
-1.36122465e-01 -3.63965750e-01 -2.46839747e-01 -1.31750810e+00
-1.10432822e-02 -1.09244227e+00 -1.14555918e-01 -9.28333342e-01
-1.25092590e+00 7.59001970e-01 -1.07178383e-01 -1.06421041e+00
-7.93411016e-01 -7.09592760e-01 -5.16645610e-01 1.00284886e+00
-1.54526293e+00 -1.11318076e+00 2.74127603e-01 4.04171526e-01
8.46310914e-01 -3.19870591e-01 1.07200038e+00 4.22347426e-01
-1.00243874e-01 6.16248667e-01 1.22810774e-01 5.17720245e-02
5.90695560e-01 -9.57007468e-01 5.19454002e-01 6.17003083e-01
2.00665280e-01 1.08876872e+00 5.33572972e-01 -3.59306931e-01
-1.43094933e+00 -7.11324632e-01 7.11334586e-01 -1.80900231e-01
7.81534731e-01 -3.25668603e-01 -7.01706886e-01 2.64252841e-01
3.16008538e-01 -5.27570546e-02 7.67216325e-01 -1.63183123e-01
2.32400466e-02 2.35943981e-02 -1.22274625e+00 8.89826000e-01
1.26548171e+00 -9.79916334e-01 -1.02743912e+00 2.34603062e-01
5.47673225e-01 -3.13953549e-01 -1.15315616e+00 5.78890026e-01
6.10072494e-01 -4.52943087e-01 1.38646388e+00 -3.84405613e-01
4.77301270e-01 -4.47460040e-02 9.13084298e-02 -1.81469190e+00
-8.44109282e-02 -9.08149838e-01 -1.58370599e-01 8.24966192e-01
3.30491364e-01 -7.10589588e-01 4.29424644e-01 2.92481810e-01
-6.79738462e-01 -6.87124908e-01 -1.16794491e+00 -8.94389033e-01
5.48702240e-01 -5.80413640e-01 5.17099261e-01 9.33679640e-01
2.82279760e-01 4.63146344e-02 -5.46520710e-01 -2.37819269e-01
4.66174960e-01 1.83691978e-01 1.38269663e-01 -8.91928732e-01
-3.06041956e-01 -6.33738041e-01 -1.43314406e-01 -9.99271810e-01
6.89983666e-01 -1.46218920e+00 -4.11523879e-02 -1.65742314e+00
-5.44500314e-02 -1.39403462e-01 -8.17239210e-02 2.07579732e-01
7.43619129e-02 2.73198217e-01 3.04875314e-01 6.93369284e-02
3.24087024e-01 6.88428402e-01 1.98143244e+00 -1.03989378e-01
6.11886382e-02 -8.79794173e-03 -5.48757195e-01 7.23634005e-01
6.58931911e-01 -1.77228302e-01 -5.86198568e-01 -3.85348737e-01
-4.29864049e-01 5.77446222e-01 1.71030536e-01 -7.17690945e-01
1.07844852e-01 1.24155983e-01 1.53877303e-01 -3.15906823e-01
6.73769057e-01 -4.74045068e-01 5.14310040e-02 4.23963606e-01
-5.24073958e-01 -2.45434210e-01 2.85679489e-01 8.03475156e-02
-2.77231604e-01 -5.44978142e-01 6.88938618e-01 7.67729804e-03
-2.97884643e-01 2.34525353e-01 -7.39091754e-01 -1.38375640e-01
3.93109262e-01 1.02307640e-01 -3.24895650e-01 -6.77005291e-01
-1.25334382e+00 -3.17250550e-01 -3.16672735e-02 6.20479286e-01
7.11300790e-01 -1.46871591e+00 -7.76298225e-01 3.36569488e-01
-2.22382113e-01 -7.40159273e-01 4.24768746e-01 8.66534889e-01
-4.70183045e-01 7.81983614e-01 -2.78927237e-01 -3.38278592e-01
-9.91907060e-01 4.66399789e-01 4.93859679e-01 6.87836781e-02
-7.59965956e-01 4.71151233e-01 3.47729445e-01 -8.83159399e-01
-9.11184400e-02 -3.22851539e-01 -1.36898860e-01 -1.98538721e-01
4.69912857e-01 1.77677572e-01 2.76885718e-01 -7.68115520e-01
-4.69518811e-01 5.18943906e-01 4.66277897e-01 -3.88204873e-01
1.36341691e+00 -2.62699395e-01 -4.79842443e-03 3.09634339e-02
1.27509189e+00 5.19724607e-01 -9.58374560e-01 -3.70312966e-02
-3.37869674e-01 -2.50497967e-01 4.21055973e-01 -1.04761064e+00
-1.30473685e+00 1.01476240e+00 3.63374710e-01 -2.54349392e-02
1.16676772e+00 8.02163854e-02 1.05726016e+00 2.15571553e-01
1.03643514e-01 -1.09169400e+00 -1.40291274e-01 1.43307835e-01
1.44311976e+00 -7.92566836e-01 -2.52449244e-01 -5.14484704e-01
-5.25976300e-01 1.22788978e+00 -1.67426094e-02 1.84168220e-01
6.14893079e-01 7.06487656e-01 1.84196576e-01 -4.08781953e-02
-3.49533051e-01 -2.82744706e-01 7.11225867e-01 9.55911577e-01
5.31384468e-01 1.41468704e-01 -8.06249261e-01 7.35400677e-01
-6.25804186e-01 -1.17743969e-01 3.94421935e-01 2.39471525e-01
-1.41384780e-01 -1.16383076e+00 -2.67535806e-01 3.33496213e-01
-7.85090029e-01 -3.24389577e-01 -6.20210841e-02 6.53964520e-01
-3.28538448e-01 1.02455902e+00 -4.93368544e-02 -1.56010166e-01
8.27367455e-02 1.41781062e-01 7.79409826e-01 -4.33378994e-01
-4.63794172e-01 4.17575508e-01 4.01810050e-01 -4.06218022e-01
-4.73644197e-01 -6.28505886e-01 -1.42075455e+00 -1.19468287e-01
-2.32735589e-01 -2.52642035e-02 1.21452463e+00 1.12167788e+00
2.76880175e-01 7.86389947e-01 6.01477087e-01 -7.20796466e-01
-5.79264522e-01 -7.72682607e-01 -5.01854360e-01 3.53199482e-01
4.90729600e-01 -7.59211481e-01 -3.25836807e-01 -9.62603167e-02] | [14.988422393798828, 6.172701835632324] |
51dabe52-f7cc-462b-9261-4e5120afb26f | a-visual-domain-transfer-learning-approach | 2107.13237 | null | https://arxiv.org/abs/2107.13237v2 | https://arxiv.org/pdf/2107.13237v2.pdf | A Visual Domain Transfer Learning Approach for Heartbeat Sound Classification | Heart disease is the most common reason for human mortality that causes almost one-third of deaths throughout the world. Detecting the disease early increases the chances of survival of the patient and there are several ways a sign of heart disease can be detected early. This research proposes to convert cleansed and normalized heart sound into visual mel scale spectrograms and then using visual domain transfer learning approaches to automatically extract features and categorize between heart sounds. Some of the previous studies found that the spectrogram of various types of heart sounds is visually distinguishable to human eyes, which motivated this study to experiment on visual domain classification approaches for automated heart sound classification. It will use convolution neural network-based architectures i.e. ResNet, MobileNetV2, etc as the automated feature extractors from spectrograms. These well-accepted models in the image domain showed to learn generalized feature representations of cardiac sounds collected from different environments with varying amplitude and noise levels. Model evaluation criteria used were categorical accuracy, precision, recall, and AUROC as the chosen dataset is unbalanced. The proposed approach has been implemented on datasets A and B of the PASCAL heart sound collection and resulted in ~ 90% categorical accuracy and AUROC of ~0.97 for both sets. | ['Sidharth Pancholi', 'Uddipan Mukherjee'] | 2021-07-28 | null | null | null | null | ['sound-classification'] | ['audio'] | [ 9.11222119e-03 -2.93939173e-01 4.27788496e-01 -2.82202184e-01
-2.82443017e-01 -4.75251734e-01 1.66096330e-01 5.05477965e-01
-2.84725696e-01 7.48996615e-01 9.62182358e-02 -3.08275372e-01
-2.19386712e-01 -7.01105773e-01 1.21984787e-01 -3.83946657e-01
-2.67124802e-01 1.10261932e-01 2.63589978e-01 8.28856677e-02
3.63179535e-01 5.49773097e-01 -1.50868964e+00 5.47300518e-01
2.12535828e-01 9.73829985e-01 3.80163230e-02 1.67307484e+00
-1.51110306e-01 8.99432480e-01 -9.38434541e-01 1.92705467e-01
1.31040975e-01 -8.49549234e-01 -8.66400957e-01 -2.76689529e-01
2.76914597e-01 8.78885463e-02 1.52444705e-01 8.91992152e-01
1.01131940e+00 -1.22431503e-03 9.51822102e-01 -1.25351882e+00
-7.05251396e-01 1.91431448e-01 -8.88827369e-02 9.61753011e-01
4.29918915e-01 5.29372059e-02 4.50735241e-01 -5.97491860e-01
2.96932578e-01 1.09951448e+00 1.03122962e+00 4.18512851e-01
-9.58918333e-01 -7.86947250e-01 -9.50427949e-01 4.49168354e-01
-1.22969961e+00 -7.92809799e-02 9.50682759e-01 -8.27913821e-01
9.84145045e-01 4.52030331e-01 6.89715326e-01 4.92174387e-01
3.71334106e-01 -1.49348035e-01 1.38190043e+00 -7.54080713e-01
1.07682563e-01 6.32701933e-01 3.32659811e-01 6.70453668e-01
1.86403573e-01 2.57769585e-01 -1.30961016e-01 -2.08936945e-01
6.38773501e-01 1.39539585e-01 -1.07147865e-01 6.09501041e-02
-1.23006439e+00 8.30365002e-01 3.86590093e-01 8.00775826e-01
-3.79715264e-01 2.13218853e-02 7.66830862e-01 4.96027976e-01
1.11509217e-02 2.18561053e-01 -3.96562219e-01 -2.06443503e-01
-7.86723971e-01 -1.39692515e-01 6.76655412e-01 1.50615662e-01
1.93179727e-01 2.56698042e-01 -9.33912769e-02 7.96051443e-01
3.06811005e-01 7.62415648e-01 6.60445571e-01 -6.50344372e-01
-4.75124307e-02 5.56624115e-01 -2.03000531e-01 -1.37248623e+00
-7.09316373e-01 -2.26663485e-01 -9.40095365e-01 5.55407345e-01
4.09517616e-01 -2.51352727e-01 -1.15988779e+00 1.27817607e+00
1.99182510e-01 1.47197217e-01 3.13024849e-01 7.60071695e-01
1.58628428e+00 8.21383238e-01 4.93146569e-01 -9.12592039e-02
1.57560420e+00 -1.45375103e-01 -7.29745328e-01 3.45497042e-01
7.00612217e-02 -9.05110836e-01 8.63645971e-01 3.99274081e-01
-5.68022668e-01 -1.10041213e+00 -1.15823805e+00 2.92622685e-01
-5.83368599e-01 2.39332780e-01 1.21159233e-01 1.07211280e+00
-9.87182558e-01 6.04546010e-01 -5.88476479e-01 -8.43532383e-01
5.39556682e-01 1.13191627e-01 -3.55694979e-01 4.31810737e-01
-1.11644804e+00 8.53847265e-01 3.60881776e-01 -3.32332224e-01
-6.84654713e-01 -4.25540984e-01 -5.35076678e-01 1.28922790e-01
-6.67817593e-01 -5.24734020e-01 8.88267457e-01 -1.03827369e+00
-9.62475955e-01 1.01768398e+00 4.08337861e-01 -5.04459321e-01
2.95717180e-01 2.86904395e-01 -7.13171661e-01 5.61319411e-01
-4.02109474e-02 4.55937535e-01 7.16144204e-01 -8.90186787e-01
-7.42337704e-01 -8.10707435e-02 -5.48541784e-01 -9.72734541e-02
-2.43028983e-01 1.72540173e-01 4.95695293e-01 -7.37356007e-01
-1.89485922e-01 -6.57136023e-01 2.51737565e-01 8.40490684e-03
-1.57978274e-02 -2.11464301e-01 1.05331445e+00 -9.79579329e-01
1.15235507e+00 -2.39475679e+00 -5.77874064e-01 3.06399614e-01
2.62342274e-01 5.34842193e-01 2.19976470e-01 2.94021726e-01
-4.05316979e-01 2.16767177e-01 -1.14327505e-01 3.78827989e-01
-3.90398592e-01 8.97032395e-02 4.89492342e-02 3.73139739e-01
-3.08581237e-02 2.97432154e-01 -6.64927125e-01 -7.99950123e-01
4.55164999e-01 7.91457653e-01 -1.21943533e-01 2.91767389e-01
5.52477181e-01 3.93977076e-01 -1.38086751e-01 2.75110543e-01
4.23258513e-01 -1.24075241e-01 -2.48346180e-01 -3.26771349e-01
5.27306795e-02 -2.15795383e-01 -1.08378875e+00 1.03279138e+00
-1.77376494e-01 1.10557461e+00 -4.60100234e-01 -1.23382342e+00
1.22997904e+00 8.93348396e-01 3.71440470e-01 -5.69788516e-01
2.56484479e-01 6.81582317e-02 2.49261037e-01 -9.90027905e-01
-3.65148038e-01 -5.62614024e-01 1.48487866e-01 1.28230035e-01
1.85304314e-01 5.80065213e-02 -5.13198562e-02 -1.35392860e-01
8.60624254e-01 -4.53490347e-01 8.27914774e-01 -2.07434803e-01
6.28884673e-01 8.02723914e-02 4.02492940e-01 7.14130163e-01
-7.13797808e-01 7.14818716e-01 1.88975930e-01 -8.76406968e-01
-8.80599260e-01 -1.10406017e+00 -2.99327582e-01 1.00487638e+00
-2.62793630e-01 1.82422429e-01 -7.19027758e-01 -6.23030961e-01
-1.77321792e-01 3.70587885e-01 -6.79938495e-01 -1.33004010e-01
-4.24613148e-01 -4.51909393e-01 9.73927975e-01 5.67779064e-01
7.01791108e-01 -1.61854184e+00 -1.47931814e+00 1.19899102e-01
4.84897383e-02 -5.69939911e-01 7.27076605e-02 3.10413986e-01
-8.37536991e-01 -1.27200639e+00 -8.84186864e-01 -1.17047083e+00
2.30654925e-01 -2.40037769e-01 1.25479960e+00 1.95355520e-01
-1.25213230e+00 5.69988728e-01 -4.23660457e-01 -9.18609500e-01
-7.96275258e-01 -4.96596068e-01 -2.82783240e-01 -1.92952797e-01
4.80947435e-01 -6.09007657e-01 -8.19395006e-01 -2.12304398e-01
-4.63079870e-01 -2.58345306e-01 3.87038827e-01 6.81972861e-01
2.12093756e-01 1.29864290e-01 8.55162561e-01 -7.12794483e-01
8.97091389e-01 -4.59728628e-01 -8.10838956e-03 1.48979932e-01
-4.14635807e-01 -2.41348803e-01 6.07687771e-01 -3.89297903e-01
-6.17098510e-01 2.22391542e-02 1.37641713e-01 -2.69668460e-01
-7.78187811e-01 1.66960418e-01 4.45396811e-01 9.31921452e-02
1.14792597e+00 1.73922747e-01 -2.57278755e-02 -3.33028913e-01
-1.63621426e-01 9.32491779e-01 6.63313031e-01 1.16702043e-01
4.77179438e-01 2.05994442e-01 1.49127528e-01 -1.25972080e+00
-1.70486599e-01 -6.38144016e-01 -3.52780282e-01 -4.58746940e-01
1.48620605e+00 -6.08956933e-01 -4.81744587e-01 3.66397738e-01
-1.04770827e+00 2.25799054e-01 -3.10409982e-02 6.33538425e-01
-2.27084801e-01 2.33320385e-01 -3.71864647e-01 -1.04197180e+00
-9.08217251e-01 -5.25540054e-01 4.32280868e-01 4.87079948e-01
-3.92879665e-01 -1.06334198e+00 3.84203941e-01 -7.26754889e-02
8.21210980e-01 5.43930650e-01 1.19880712e+00 -9.84744310e-01
3.08751017e-01 -7.45943561e-02 -2.78179467e-01 6.09151959e-01
4.61178392e-01 8.94669592e-02 -1.18884289e+00 4.74973172e-02
-4.41291481e-02 -1.88052505e-02 6.06422901e-01 6.64507985e-01
7.91865528e-01 -1.93155840e-01 -1.76920399e-01 3.02924931e-01
1.44057989e+00 9.87005174e-01 5.37912786e-01 1.70452118e-01
3.76122206e-01 8.95629525e-02 1.91794544e-01 4.40843463e-01
-2.55304389e-02 1.52709618e-01 1.69019178e-01 -4.34511214e-01
-5.90967238e-01 1.71403423e-01 -8.80081132e-02 5.79728663e-01
-2.45133609e-01 8.52976218e-02 -1.15924370e+00 5.94348311e-01
-1.07376695e+00 -1.28734386e+00 -1.20359756e-01 2.02275920e+00
6.43160284e-01 -6.97915582e-03 4.20798689e-01 9.78133261e-01
8.41231465e-01 -2.89885372e-01 9.16677564e-02 -8.07421267e-01
3.30331445e-01 7.77617037e-01 1.19073473e-01 1.70603141e-01
-1.52795029e+00 1.45598322e-01 6.00689173e+00 2.12963209e-01
-1.64235091e+00 2.37624533e-02 7.88659036e-01 3.00615340e-01
5.04098296e-01 -4.33236629e-01 -3.87384035e-02 4.88633484e-01
1.17411411e+00 -7.04782680e-02 2.05774829e-01 8.22139621e-01
5.90648353e-02 3.39735784e-02 -8.08140039e-01 1.39146423e+00
5.25379218e-02 -1.07190716e+00 -6.44210875e-02 -5.90577006e-01
3.45345885e-01 -2.20338672e-01 -8.80005881e-02 1.94292590e-01
-2.56385535e-01 -1.30629492e+00 2.49920294e-01 7.75680780e-01
9.63474631e-01 -8.45129907e-01 8.62953842e-01 1.64480451e-02
-1.27999926e+00 -2.90340990e-01 -2.22546801e-01 -1.19174488e-01
-2.74598092e-01 1.96320623e-01 -1.39997888e+00 -5.68778664e-02
1.17402971e+00 3.64139050e-01 -7.09993005e-01 1.35921764e+00
5.13303995e-01 1.00086164e+00 -2.77365595e-01 -3.22708637e-01
-2.19238371e-01 4.22335088e-01 5.29385746e-01 1.70066416e+00
5.10033131e-01 2.99030077e-02 1.10093728e-02 5.28035462e-01
3.67417663e-01 4.12310153e-01 -8.37836564e-01 2.27085967e-02
4.00115252e-01 1.15741670e+00 -1.11603343e+00 -5.14615893e-01
-1.89175814e-01 5.41676164e-01 -4.55503583e-01 9.17150676e-02
-8.21492076e-01 -9.75190938e-01 1.30308256e-01 2.42865533e-01
7.12568313e-02 3.96248788e-01 -4.38478172e-01 -3.21201861e-01
-3.99707288e-01 -8.08058202e-01 7.57814467e-01 -7.20111668e-01
-1.08317709e+00 8.77460480e-01 3.43747847e-02 -1.39408290e+00
-3.50601450e-02 -6.07634783e-01 -8.93585145e-01 1.00207591e+00
-9.99329984e-01 -7.88444281e-01 -7.59833932e-01 6.51053727e-01
5.43735385e-01 -5.46897292e-01 1.19503176e+00 3.46285790e-01
1.06462188e-01 2.76015192e-01 -3.22423190e-01 6.77747786e-01
7.93202460e-01 -1.54414642e+00 -1.33577183e-01 5.08725047e-01
2.52462566e-01 1.96184441e-01 6.80972457e-01 -4.71988678e-01
-5.09499490e-01 -9.33768451e-01 8.43113840e-01 -1.46865636e-01
-8.04895908e-02 2.81111926e-01 -8.28906953e-01 -9.12919044e-02
4.59500909e-01 3.58519465e-01 9.08385515e-01 -5.38908184e-01
-6.88564405e-02 -2.71050304e-01 -1.54325700e+00 -9.92630050e-02
1.51943088e-01 -5.58609605e-01 -9.57929671e-01 5.20005338e-02
-1.87110975e-02 1.42105408e-02 -7.68869221e-01 3.16943407e-01
5.81260562e-01 -1.21062863e+00 1.18053007e+00 -6.51886225e-01
1.56838924e-01 -4.87017065e-01 -5.56308217e-02 -1.17506361e+00
-3.73741210e-01 -1.69354066e-01 2.63224423e-01 1.07598877e+00
2.97225773e-01 -4.03646529e-01 3.05222243e-01 -1.36338532e-01
3.64544362e-01 -1.44008920e-01 -8.89532149e-01 -2.15331227e-01
-9.39883739e-02 -1.79949790e-01 5.87829091e-02 1.06336284e+00
-1.90148294e-01 3.73258293e-01 -1.60063401e-01 1.28243621e-02
4.61831927e-01 -1.04214355e-01 4.69703734e-01 -1.78780091e+00
-1.13450401e-01 -1.86862901e-01 -1.04543078e+00 3.98256958e-01
-6.62947774e-01 -8.43858063e-01 -1.10892370e-01 -1.62117159e+00
2.81754881e-01 -2.42880568e-01 -7.89247811e-01 4.98684675e-01
-1.33285299e-01 6.10270679e-01 2.89534569e-01 -1.13898985e-01
-1.85336005e-02 -3.86227459e-01 1.02271521e+00 -1.56591713e-01
-3.10886085e-01 3.64623442e-02 -3.08869630e-01 7.85263896e-01
1.10601890e+00 -6.85946405e-01 -6.01434886e-01 2.24039927e-01
-9.83503833e-02 3.93395513e-01 7.84974158e-01 -1.65160704e+00
-1.35287464e-01 3.88204469e-03 9.23296750e-01 -3.99430215e-01
-5.31773344e-02 -9.76941764e-01 4.20952767e-01 8.37688565e-01
-4.68883395e-01 1.37148514e-01 2.49128446e-01 4.81790423e-01
-2.47192740e-01 -1.23384312e-01 1.13534570e+00 -3.15342188e-01
-6.72203898e-01 -3.43456030e-01 -4.51399595e-01 1.83829337e-01
1.02457082e+00 -4.64630932e-01 -2.51434028e-01 -3.99987310e-01
-1.08392656e+00 -5.54535925e-01 -2.25921661e-01 3.49699885e-01
6.10368013e-01 -1.24582505e+00 -8.71832907e-01 2.89071053e-01
-2.99118180e-02 -6.85311794e-01 2.65790790e-01 7.60297835e-01
-1.22391200e+00 3.04154366e-01 -8.60491931e-01 -7.20802665e-01
-2.06112671e+00 4.32817042e-01 5.57301700e-01 1.10739686e-01
-8.42130721e-01 7.30893075e-01 -1.36911020e-01 9.34887230e-02
2.28512555e-01 -6.57644331e-01 -8.87452781e-01 -4.11618054e-02
5.94914079e-01 6.04492724e-01 -1.39643922e-01 -6.51409507e-01
-5.90805173e-01 8.38956177e-01 4.35231417e-01 2.55200267e-02
1.30714750e+00 3.06877702e-01 1.59588903e-01 5.59446275e-01
1.32342088e+00 -8.40463489e-02 -4.04935211e-01 3.53118032e-01
-8.40066448e-02 -1.55458942e-01 -2.31772400e-02 -1.25401330e+00
-1.06648231e+00 1.37098277e+00 1.84380126e+00 7.08495080e-01
1.44238818e+00 -3.15061659e-01 3.35725427e-01 3.65214705e-01
-2.99764752e-01 -1.00788152e+00 7.08762035e-02 9.60924700e-02
7.75677919e-01 -1.10225928e+00 -2.22734600e-01 9.31722298e-02
-8.37567866e-01 1.47900760e+00 1.44224316e-01 -2.89653718e-01
9.93101239e-01 2.86015004e-01 6.92455471e-01 -2.17136502e-01
-3.92056078e-01 1.44488327e-02 4.55765277e-01 1.09750390e+00
7.25295007e-01 2.82131642e-01 -3.66521537e-01 6.54136419e-01
-1.06417194e-01 3.30209970e-01 2.76539892e-01 1.06249511e+00
-8.25809598e-01 -4.88629788e-01 -7.36549973e-01 4.42505240e-01
-1.08200634e+00 5.87625876e-02 -3.82669270e-01 6.39657974e-01
5.10116041e-01 1.08058834e+00 1.46712130e-02 -3.50773305e-01
1.98138610e-01 5.35682440e-01 1.34259686e-01 -2.89172769e-01
-8.27784836e-01 -2.24620208e-01 -1.06122278e-01 5.63317575e-02
-5.48163891e-01 -1.67834789e-01 -1.46642053e+00 2.56224126e-01
1.81159750e-01 2.25996077e-01 6.74248338e-01 5.47052205e-01
1.45507693e-01 8.60532284e-01 4.40357268e-01 -4.37305689e-01
-2.07340851e-01 -1.29350352e+00 -4.29091871e-01 8.01879287e-01
6.78896070e-01 -6.07130885e-01 -3.96696091e-01 7.64001787e-01] | [14.352523803710938, 3.3308584690093994] |
271e5f0f-07b0-43ee-809f-a6c8ff49fdb5 | multiple-object-tracking-in-cluttered-and | 1309.6391 | null | http://arxiv.org/abs/1309.6391v1 | http://arxiv.org/pdf/1309.6391v1.pdf | Multiple-object tracking in cluttered and crowded public spaces | This paper addresses the problem of tracking moving objects of variable
appearance in challenging scenes rich with features and texture. Reliable
tracking is of pivotal importance in surveillance applications. It is made
particularly difficult by the nature of objects encountered in such scenes:
these too change in appearance and scale, and are often articulated (e.g.
humans). We propose a method which uses fast motion detection and segmentation
as a constraint for both building appearance models and their robust
propagation (matching) in time. The appearance model is based on sets of local
appearances automatically clustered using spatio-kinetic similarity, and is
updated with each new appearance seen. This integration of all seen appearances
of a tracked object makes it extremely resilient to errors caused by occlusion
and the lack of permanence of due to low data quality, appearance change or
background clutter. These theoretical strengths of our algorithm are
empirically demonstrated on two hour long video footage of a busy city
marketplace. | ['Ognjen Arandjelović', 'Rhys Martin'] | 2013-09-25 | null | null | null | null | ['motion-detection'] | ['computer-vision'] | [ 1.80808127e-01 -6.31144345e-01 2.92059928e-01 -1.14740819e-01
-1.91618335e-02 -7.99862623e-01 5.51151812e-01 5.73892780e-02
-4.69901621e-01 6.42391503e-01 -4.44858223e-01 -3.19148488e-02
9.01804864e-02 -3.23986918e-01 -5.15730858e-01 -7.12127507e-01
-4.03239459e-01 4.22775030e-01 1.01011741e+00 -3.05600539e-02
-1.66871592e-01 8.75971019e-01 -1.63910043e+00 -1.06545873e-01
4.46657836e-01 7.79285192e-01 1.83862969e-01 1.14665473e+00
1.85900748e-01 6.58132374e-01 -4.96093631e-01 -4.17346984e-01
6.87467873e-01 -2.12692469e-01 -3.58618468e-01 8.45897615e-01
8.54813576e-01 -4.32193875e-01 -1.71820015e-01 1.10654664e+00
2.94407830e-02 4.56233561e-01 6.47550344e-01 -1.33791685e+00
9.32231247e-02 -5.18241823e-01 -1.01572812e+00 6.23841822e-01
4.09263968e-01 1.81604654e-01 3.45006734e-01 -7.20009327e-01
1.04481649e+00 1.21995306e+00 8.75285029e-01 5.11972010e-01
-1.24889028e+00 -2.44145542e-01 5.37107646e-01 -8.93902555e-02
-1.35103846e+00 -7.62441576e-01 6.54601216e-01 -7.36464322e-01
3.20295691e-01 4.45105612e-01 8.36828649e-01 6.77223325e-01
3.95219475e-01 4.58087265e-01 7.17522502e-01 -5.33471048e-01
1.82589486e-01 2.09791258e-01 4.51058820e-02 8.02756965e-01
5.00696063e-01 1.98788941e-01 -8.97485837e-02 -4.93684173e-01
1.06478000e+00 1.53097525e-01 -1.45772845e-01 -7.34262049e-01
-1.10035396e+00 2.57472277e-01 2.82172561e-01 1.73107475e-01
-4.94678617e-01 2.89717525e-01 -4.98663541e-03 3.21237594e-01
5.65813005e-01 -1.36795387e-01 -4.19655710e-01 -7.72086382e-02
-1.01668453e+00 2.36965984e-01 4.76719141e-01 1.12543929e+00
3.59692544e-01 9.18248221e-02 3.06577384e-01 5.60873628e-01
3.73874068e-01 7.47594237e-01 -4.51065116e-02 -1.03625321e+00
-8.37420896e-02 2.97692508e-01 7.13016927e-01 -1.40938842e+00
-5.10900319e-01 -2.21137315e-01 -6.60479307e-01 8.34943235e-01
8.39442611e-01 -1.92375898e-01 -8.65099132e-01 1.61977756e+00
1.12250960e+00 1.34899825e-01 -4.39391434e-01 7.83373237e-01
3.95955116e-01 4.19200748e-01 7.39611834e-02 -7.77322352e-01
1.37856519e+00 -7.36425996e-01 -7.40327895e-01 -9.06048641e-02
-4.63090315e-02 -1.14185202e+00 1.28276050e-01 3.12348753e-01
-1.27130616e+00 -8.59793305e-01 -6.76019430e-01 5.26168168e-01
-1.71219841e-01 -3.07034224e-01 4.79982018e-01 6.29821122e-01
-1.27405286e+00 4.97891128e-01 -9.01287973e-01 -7.00717211e-01
1.12220310e-01 5.58228731e-01 -3.78373504e-01 9.64837894e-02
-5.61774492e-01 8.86303246e-01 1.38438074e-02 2.45190278e-01
-4.84474182e-01 -1.84183612e-01 -6.00396812e-01 -5.54096162e-01
1.86220542e-01 -8.72912228e-01 1.28298283e+00 -1.72716475e+00
-1.05351889e+00 9.01192844e-01 -5.08297265e-01 -1.71048820e-01
1.01564240e+00 -6.47686943e-02 -7.19830275e-01 2.02811763e-01
-6.05858676e-02 4.01968539e-01 1.48812473e+00 -1.51619422e+00
-1.04737854e+00 -3.58849227e-01 -1.37695715e-01 2.88838178e-01
4.02555615e-02 4.28247809e-01 -6.48469090e-01 -5.65633178e-01
2.57462651e-01 -1.02118456e+00 -4.56366599e-01 5.44197559e-01
2.03197807e-01 2.10714623e-01 1.22830284e+00 -8.06647599e-01
1.00238562e+00 -2.12054539e+00 -8.23990181e-02 2.65407026e-01
3.07503730e-01 1.81753889e-01 -6.35002181e-02 4.68199402e-02
1.11000575e-01 -3.64434749e-01 5.27793504e-02 -2.72258162e-01
-3.60828251e-01 -2.98338272e-02 1.61371350e-01 9.20687437e-01
7.44916033e-04 5.00345588e-01 -1.14385271e+00 -7.82348454e-01
3.55014384e-01 3.87163758e-01 5.05229607e-02 2.01471984e-01
-1.28641114e-01 5.26497185e-01 -4.70538884e-01 8.72040927e-01
7.81710267e-01 -1.40165940e-01 -1.31866664e-01 -1.13509230e-01
-3.72530669e-01 -4.79875654e-01 -1.46548903e+00 1.17836010e+00
1.60283148e-02 6.60436690e-01 5.83254755e-01 -2.41160944e-01
5.97912908e-01 3.20815057e-01 8.66179287e-01 -3.68136108e-01
1.94428459e-01 -9.64041799e-02 3.81871685e-02 -4.11050916e-01
7.38696754e-01 -1.46557437e-02 4.29383457e-01 2.93777823e-01
-5.61004102e-01 9.23633352e-02 2.87395120e-01 1.42331794e-01
1.03339148e+00 4.09407437e-01 3.23996365e-01 -1.79808035e-01
3.32349360e-01 2.74329841e-01 6.77412987e-01 7.51965702e-01
-6.31655574e-01 5.73010981e-01 -4.19623673e-01 -8.37570369e-01
-1.26808381e+00 -1.06082296e+00 -2.70504862e-01 8.55202556e-01
5.79056799e-01 2.03033686e-01 -4.93184477e-01 -3.67855996e-01
-2.05121022e-02 6.32694960e-02 -7.38124847e-01 2.23397911e-01
-6.42826915e-01 -7.50774801e-01 -2.02149212e-01 3.99727792e-01
1.11424737e-01 -8.99193406e-01 -1.21318662e+00 7.23636568e-01
-6.51289821e-02 -1.19809425e+00 -6.59281194e-01 -3.27742159e-01
-1.02709877e+00 -1.18914497e+00 -7.62296617e-01 -6.01098895e-01
1.04881537e+00 7.94472456e-01 1.22090185e+00 4.52905536e-01
-6.97328866e-01 9.32409406e-01 -3.85140777e-01 -4.18780237e-01
-5.89810789e-01 -6.25628591e-01 3.55269134e-01 5.38828909e-01
6.59738928e-02 -2.79030204e-01 -6.49817228e-01 6.03849053e-01
-6.91723943e-01 1.51276952e-02 1.83490932e-01 5.11309087e-01
4.80425477e-01 3.12184036e-01 -5.45845963e-02 -3.90094966e-01
7.13717937e-02 -2.44639292e-01 -1.09267378e+00 3.69018525e-01
-1.68834087e-02 -6.24726534e-01 1.49689451e-01 -7.31890261e-01
-1.11996305e+00 4.53517109e-01 5.70670784e-01 -5.44905782e-01
-3.16878021e-01 -2.66446382e-01 2.62938470e-01 -5.17844737e-01
4.22314167e-01 -5.46284430e-02 1.67593077e-01 -1.41888171e-01
2.68071562e-01 2.66151011e-01 7.50457287e-01 -2.90639400e-01
1.17836058e+00 1.01560938e+00 8.81126449e-02 -1.27370095e+00
-1.93124726e-01 -9.89750385e-01 -1.12798488e+00 -8.95594060e-01
8.96890581e-01 -6.35980308e-01 -4.82786864e-01 5.21987081e-01
-1.29786038e+00 -2.33666644e-01 -1.38183624e-01 4.43955213e-01
-3.12928349e-01 6.60102785e-01 -3.92475545e-01 -1.39248884e+00
-1.59707442e-01 -6.73561513e-01 9.42951441e-01 3.64347994e-01
-2.53544539e-01 -1.28221083e+00 1.15326039e-01 -9.49232802e-02
4.49606270e-01 7.56475270e-01 1.11839764e-01 1.03749275e-01
-7.34625459e-01 -4.50138241e-01 -4.86699976e-02 -2.37693131e-01
5.06305754e-01 7.18819857e-01 -9.31078970e-01 -6.51248515e-01
1.54795535e-02 3.33767772e-01 4.60843682e-01 8.89244378e-01
3.09478998e-01 -5.29937027e-03 -7.40710735e-01 3.24133843e-01
1.37006366e+00 5.61102748e-01 2.19688788e-01 4.97217953e-01
6.12821758e-01 6.99854255e-01 9.12644923e-01 4.48111683e-01
2.33973726e-03 1.04318070e+00 4.11748856e-01 -6.01764083e-01
-1.02445908e-01 5.46531379e-01 3.91797990e-01 4.95345414e-01
-4.94150728e-01 -1.55973360e-01 -6.83656514e-01 5.88183284e-01
-1.98901713e+00 -1.38074267e+00 -6.80067182e-01 2.55256820e+00
2.85929292e-01 2.51732878e-02 5.68809688e-01 -2.25737870e-01
1.01826560e+00 -2.79647470e-01 -4.41225737e-01 -6.92561194e-02
-1.09547511e-01 -4.80864018e-01 6.47033393e-01 5.28791249e-01
-1.48658049e+00 5.34502029e-01 6.40944481e+00 2.45329589e-01
-6.40536070e-01 3.72101814e-02 5.10785758e-01 -9.52111185e-03
3.00678819e-01 -1.41324356e-01 -7.08568752e-01 4.23896700e-01
3.47006500e-01 4.18491624e-02 1.64621830e-01 5.12352943e-01
3.88137966e-01 -3.97815019e-01 -8.72434735e-01 9.52029228e-01
1.36725947e-01 -8.08090210e-01 -5.00270724e-01 -1.17019902e-03
8.21255624e-01 -1.84665754e-01 1.20848738e-01 -3.95867705e-01
5.48956811e-01 -5.25987327e-01 9.78511870e-01 5.71694970e-01
3.52170795e-01 -3.53590339e-01 5.26791215e-01 1.32070139e-01
-1.65361691e+00 7.42984563e-02 -3.46415162e-01 4.53480855e-02
4.81717408e-01 3.50089699e-01 -5.37624836e-01 4.63434249e-01
6.99503660e-01 5.32694161e-01 -5.59611619e-01 1.62474585e+00
3.90881717e-01 -1.44355437e-02 -6.16887510e-01 2.28592321e-01
1.70175985e-01 -3.78329754e-01 9.47475612e-01 1.18150294e+00
9.97504666e-02 1.76213056e-01 4.90907460e-01 2.92896569e-01
5.15988767e-01 -4.67831381e-02 -7.14265645e-01 5.09697795e-01
2.73806453e-01 1.47076428e+00 -1.28156328e+00 -4.67353284e-01
-5.56002617e-01 1.16191006e+00 -1.58896774e-01 5.23275793e-01
-8.59827220e-01 2.17228740e-01 6.75908089e-01 4.31976736e-01
3.87618810e-01 -3.89575988e-01 3.93929511e-01 -1.01248264e+00
2.21048743e-01 -7.08820581e-01 4.57184047e-01 -5.49927175e-01
-1.21877658e+00 8.42800081e-01 1.16360702e-01 -1.67216074e+00
-1.02067716e-01 -2.40100697e-01 -5.91994584e-01 5.20165622e-01
-1.10626841e+00 -1.19938803e+00 -4.62533742e-01 6.30414784e-01
8.06706548e-01 1.52364582e-01 3.67022932e-01 3.24830592e-01
-3.37426454e-01 1.79080799e-01 2.81624764e-01 -5.38342372e-02
5.54882646e-01 -1.03878272e+00 5.24776459e-01 1.18403089e+00
1.69472262e-01 4.58417863e-01 9.59963381e-01 -8.64630461e-01
-1.22361004e+00 -1.11076295e+00 5.71379542e-01 -8.95241439e-01
6.12495720e-01 -1.57497615e-01 -8.11676979e-01 5.15791357e-01
-1.08894063e-02 2.02287406e-01 2.68016040e-01 -2.45140254e-01
1.60331964e-01 -1.30247790e-02 -1.11481559e+00 6.00469410e-01
9.48576748e-01 6.85192496e-02 -3.70752037e-01 3.05586845e-01
1.47587940e-01 -4.92247194e-01 -5.12354076e-01 2.96199769e-01
8.91807258e-01 -8.83618712e-01 1.03767335e+00 -5.43080330e-01
-6.64132357e-01 -8.69804204e-01 1.62315428e-01 -7.90804923e-01
-6.58598363e-01 -9.23826098e-01 -2.46596076e-02 1.03425992e+00
-1.75465252e-02 -3.08250546e-01 6.53715253e-01 9.17913020e-01
2.77022153e-01 -7.42582381e-02 -1.00747490e+00 -1.17089224e+00
-6.67317092e-01 -2.32844707e-02 3.99158932e-02 9.08203125e-01
-4.77888107e-01 -1.08191624e-01 -5.36172450e-01 4.42014396e-01
1.06519222e+00 8.87259096e-02 1.15139174e+00 -1.55652678e+00
-2.51339257e-01 -3.66288364e-01 -7.27791667e-01 -7.55930781e-01
-4.11800921e-01 -2.03859564e-02 3.03858846e-01 -1.23274207e+00
3.24565470e-01 -6.44521773e-01 -3.98621559e-02 3.75029184e-02
-2.63797551e-01 2.51472235e-01 2.49880627e-01 4.62319732e-01
-8.43844950e-01 3.90602238e-02 9.23120677e-01 -1.44939810e-01
-2.78325170e-01 3.26538295e-01 2.64661789e-01 1.06302345e+00
4.80936021e-01 -6.36149228e-01 -2.26056844e-01 -3.85641754e-01
-2.04210773e-01 1.01517119e-01 5.82029283e-01 -9.90926325e-01
3.53433251e-01 -2.20131367e-01 7.49607503e-01 -5.58334291e-01
5.43902874e-01 -1.31818295e+00 8.26222539e-01 6.78334355e-01
2.53599256e-01 7.60667503e-01 3.61789048e-01 8.41995716e-01
1.06809810e-01 -3.61942723e-02 9.97507632e-01 -1.86315536e-01
-9.50305343e-01 5.23722887e-01 -5.25844216e-01 -3.70745629e-01
1.49476719e+00 -8.74164164e-01 6.45195469e-02 -4.17642474e-01
-9.21445429e-01 7.68339410e-02 1.02931833e+00 4.90699649e-01
4.04084861e-01 -1.23582590e+00 -6.82750285e-01 7.66954198e-02
-1.26621932e-01 -2.18996942e-01 1.43809885e-01 1.10599637e+00
-6.51584804e-01 -4.89427447e-02 -2.72394538e-01 -8.17544818e-01
-2.08812952e+00 7.06874430e-01 3.54821712e-01 1.31931484e-01
-8.44703674e-01 7.53759682e-01 3.71057093e-01 4.37271446e-01
3.17433387e-01 -1.50504887e-01 -5.45494556e-02 -7.80177414e-02
8.80332649e-01 4.51433212e-01 -2.10215062e-01 -1.04459095e+00
-5.17167747e-01 8.76265824e-01 -2.41311312e-01 -9.10509899e-02
8.48751724e-01 -3.69541734e-01 -3.49777676e-02 4.11055952e-01
5.89845836e-01 3.86717990e-02 -1.76062953e+00 -8.49792436e-02
-6.06776290e-02 -8.69584918e-01 -2.98546612e-01 -2.19948262e-01
-9.66780484e-01 3.16887468e-01 8.88873577e-01 4.56041187e-01
1.03660882e+00 -2.46817484e-01 4.20714855e-01 4.01063077e-02
5.94021201e-01 -9.71940041e-01 5.64081483e-02 1.07147828e-01
5.78190267e-01 -1.19453013e+00 2.48161644e-01 -4.46621627e-01
-3.45225632e-01 1.18811131e+00 4.79375243e-01 6.56596348e-02
4.40046847e-01 3.58402491e-01 4.46620017e-01 -2.14962021e-01
-5.37535787e-01 -3.28379989e-01 2.49042109e-01 9.11968291e-01
6.72862679e-02 -9.37775746e-02 -6.33346289e-02 -5.66164374e-01
5.28962135e-01 -4.39394921e-01 4.41529244e-01 1.24252057e+00
-5.91428697e-01 -8.62057805e-01 -9.57594097e-01 1.36073962e-01
-3.22704226e-01 2.73446888e-01 -3.57070029e-01 9.10601199e-01
1.73068598e-01 1.04387629e+00 1.43583417e-01 8.36903304e-02
1.66792557e-01 -2.60709465e-01 6.62319481e-01 -1.58899322e-01
-4.30018842e-01 5.17587066e-01 3.81137654e-02 -4.21470553e-01
-1.01876724e+00 -1.17755628e+00 -8.46074998e-01 -1.50900885e-01
-6.63949847e-01 -6.32706434e-02 5.49951255e-01 6.02816880e-01
-2.85498835e-02 4.10459846e-01 7.37704754e-01 -1.10263705e+00
-1.37213618e-01 -4.34259981e-01 -5.76971591e-01 7.92202711e-01
5.73601604e-01 -8.68056297e-01 -2.64157921e-01 8.68406236e-01] | [6.825959205627441, -1.81807541847229] |
2c2c7d0c-cba5-44ab-8f59-ebbcdb3cc7e1 | uhrnet-a-deep-learning-based-method-for | 2304.14503 | null | https://arxiv.org/abs/2304.14503v1 | https://arxiv.org/pdf/2304.14503v1.pdf | UHRNet: A Deep Learning-Based Method for Accurate 3D Reconstruction from a Single Fringe-Pattern | The quick and accurate retrieval of an object height from a single fringe pattern in Fringe Projection Profilometry has been a topic of ongoing research. While a single shot fringe to depth CNN based method can restore height map directly from a single pattern, its accuracy is currently inferior to the traditional phase shifting technique. To improve this method's accuracy, we propose using a U shaped High resolution Network (UHRNet). The network uses UNet encoding and decoding structure as backbone, with Multi-Level convolution Block and High resolution Fusion Block applied to extract local features and global features. We also designed a compound loss function by combining Structural Similarity Index Measure Loss (SSIMLoss) function and chunked L2 loss function to improve 3D reconstruction details.We conducted several experiments to demonstrate the validity and robustness of our proposed method. A few experiments have been conducted to demonstrate the validity and robustness of the proposed method, The average RMSE of 3D reconstruction by our method is only 0.443(mm). which is 41.13% of the UNet method and 33.31% of Wang et al hNet method. Our experimental results show that our proposed method can increase the accuracy of 3D reconstruction from a single fringe pattern. | ['Hui Li', 'Xingyang Qi', 'Canlin Zhou', 'Yixiao Wang'] | 2023-04-23 | null | null | null | null | ['3d-reconstruction'] | ['computer-vision'] | [ 2.60455400e-01 1.80662908e-02 2.83714116e-01 -3.57285053e-01
-6.66792214e-01 2.46848911e-01 2.59580672e-01 -3.38709533e-01
-3.86504322e-01 7.15163887e-01 9.28670689e-02 2.46239364e-01
-3.66316915e-01 -1.20460081e+00 -1.00258970e+00 -5.93985736e-01
1.80157572e-02 3.92319262e-01 5.71865439e-01 -1.51137292e-01
4.60836053e-01 7.65147388e-01 -1.60944164e+00 7.97516927e-02
7.54198849e-01 1.43475842e+00 3.30872536e-01 2.82266945e-01
1.18008077e-01 5.63662887e-01 -4.89480823e-01 -1.48264736e-01
5.71018577e-01 1.64143909e-02 -4.27730322e-01 -7.27729686e-03
7.22370267e-01 -6.57997549e-01 -4.98193771e-01 7.73622572e-01
9.41288650e-01 -2.14675982e-02 7.86783874e-01 -7.29636014e-01
-4.35862988e-01 1.74132034e-01 -9.32761431e-01 -1.97240829e-01
3.47034603e-01 -2.49866024e-01 4.77259666e-01 -1.09303343e+00
7.11674333e-01 1.16157734e+00 9.65437472e-01 1.81222752e-01
-7.13292241e-01 -1.01451087e+00 -3.34754139e-01 3.67672704e-02
-1.67516518e+00 -1.60596773e-01 8.60421360e-01 -2.70153254e-01
8.41184139e-01 -6.89710155e-02 8.13350081e-01 3.40246648e-01
6.83125734e-01 5.23292303e-01 1.14684379e+00 -2.70281315e-01
-2.51706570e-01 -1.46189287e-01 -2.19656646e-01 8.19612980e-01
3.26812595e-01 3.18009526e-01 -7.26922274e-01 3.80957097e-01
1.34697521e+00 -3.51929031e-02 -4.21273798e-01 -3.57423067e-01
-1.06140208e+00 6.03567779e-01 8.86825740e-01 -5.48376096e-03
-3.05955291e-01 1.10839956e-01 -3.09700314e-02 2.54170060e-01
4.91918296e-01 1.40152663e-01 -1.31547183e-01 1.20081995e-02
-1.04377806e+00 6.01932943e-01 5.50760686e-01 9.27119732e-01
6.41570091e-01 -3.44027318e-02 7.86762014e-02 8.67304385e-01
3.14006716e-01 6.96245313e-01 2.50811696e-01 -8.11378241e-01
5.33361256e-01 5.01908898e-01 9.90683585e-02 -1.07184219e+00
-4.48750734e-01 -4.45077568e-01 -8.58423471e-01 3.98961276e-01
1.23628139e-01 -7.50610343e-05 -1.05063784e+00 1.25917530e+00
2.01386660e-01 -6.23797029e-02 -6.80312216e-02 1.06303620e+00
1.03783190e+00 7.10520208e-01 -7.51811445e-01 -5.97036220e-02
7.68240392e-01 -7.01187909e-01 -6.38177514e-01 1.07024498e-01
1.89346701e-01 -1.09004259e+00 7.59866238e-01 4.61145252e-01
-1.43965709e+00 -4.32217449e-01 -1.36613381e+00 -2.56218106e-01
-5.03198653e-02 -1.29234744e-03 3.87287825e-01 1.91233441e-01
-6.81035042e-01 8.45289528e-01 -6.28102481e-01 -1.78038135e-01
5.02821445e-01 5.19774199e-01 -2.42906108e-01 -2.30269328e-01
-9.88283217e-01 7.07954943e-01 3.42008770e-01 2.86320776e-01
-5.47933996e-01 -7.00358093e-01 -6.11155868e-01 4.12385426e-02
1.20852336e-01 -5.27487874e-01 1.06380737e+00 -1.39404893e-01
-1.69312167e+00 6.05672538e-01 8.30846429e-02 -3.60882103e-01
4.17957574e-01 -3.97261679e-01 -1.71086207e-01 2.73866475e-01
-2.88765598e-02 7.30781555e-01 3.37743670e-01 -1.12256885e+00
-5.66494107e-01 -5.51957905e-01 -2.80027419e-01 3.94154876e-01
1.19394116e-01 -2.60689408e-01 -2.31347784e-01 -4.53028053e-01
9.19829845e-01 -6.05375051e-01 1.63267791e-01 4.38636601e-01
-2.26383045e-01 1.97491735e-01 1.01648140e+00 -7.12854505e-01
1.06799603e+00 -1.76373720e+00 -3.57304126e-01 3.02775383e-01
1.12856627e-01 1.93620086e-01 3.13953370e-01 3.34228396e-01
1.75420284e-01 -2.82690078e-01 -4.48916435e-01 -5.11440299e-02
-3.47682685e-01 -2.18106776e-01 1.17826752e-01 5.19879639e-01
-8.09051692e-02 9.57388222e-01 -3.23617756e-01 -3.93956035e-01
5.25907636e-01 8.44432235e-01 -5.02021849e-01 2.23702908e-01
1.77782312e-01 2.19401196e-01 -2.14991599e-01 9.56506252e-01
1.19315970e+00 -1.01108834e-01 -3.17399710e-01 -5.70121169e-01
-3.61158907e-01 8.71164128e-02 -1.30098760e+00 1.73643780e+00
-4.88834649e-01 5.55312812e-01 -7.32410757e-04 -5.48384309e-01
1.52344060e+00 1.01810858e-01 6.46815300e-01 -1.09187376e+00
2.28673562e-01 4.45323646e-01 -1.14790931e-01 -4.82283920e-01
5.55994511e-01 -4.41286564e-01 1.22966126e-01 -7.95994420e-03
-2.76080340e-01 -3.02424759e-01 -5.53661764e-01 -2.02562794e-01
9.53542054e-01 3.34642828e-01 3.15609574e-02 -1.49718344e-01
4.66265857e-01 -2.15873718e-01 5.64221561e-01 3.62210304e-01
3.66323218e-02 1.04357338e+00 -1.48903802e-01 -5.28843164e-01
-1.31797683e+00 -1.14358366e+00 -6.20019734e-01 -1.99157447e-02
4.74756241e-01 -1.37161568e-01 -2.94425756e-01 -4.17503342e-02
1.75485536e-01 7.88287744e-02 -3.11257392e-01 1.63043868e-02
-7.34897733e-01 -7.81994283e-01 3.10664415e-01 5.67883193e-01
1.31492484e+00 -8.91540766e-01 -6.92270577e-01 5.27788401e-02
-1.24587461e-01 -1.00295019e+00 -4.68340516e-02 -1.04138488e-02
-1.06768584e+00 -1.04771101e+00 -9.64460909e-01 -7.94465244e-01
4.33834881e-01 2.34419122e-01 8.16259265e-01 -1.80136725e-01
-2.15959579e-01 -1.92649812e-01 -2.45543048e-01 -1.97442114e-01
3.74241292e-01 7.22257495e-02 -1.63335726e-01 -3.05664599e-01
6.86752871e-02 -8.67658079e-01 -9.17793274e-01 3.96241665e-01
-6.98537469e-01 2.77521789e-01 8.45185399e-01 7.05589116e-01
7.32376993e-01 7.31903985e-02 3.00286740e-01 -6.91924036e-01
5.10353148e-01 -1.16845116e-01 -7.62222767e-01 -2.36435801e-01
-8.66582751e-01 -1.58017740e-01 2.55603760e-01 -1.26593918e-01
-1.11565638e+00 -1.66537538e-01 -3.53790522e-01 -5.58335721e-01
1.50266454e-01 3.82236421e-01 -9.94129032e-02 -5.56649029e-01
3.50868970e-01 1.47598073e-01 1.24209896e-01 -5.97167790e-01
-1.19541593e-01 8.13585043e-01 7.53951788e-01 -3.00325304e-01
7.41921186e-01 4.58437204e-01 4.78299439e-01 -8.82268608e-01
-6.88178480e-01 -2.21853942e-01 -5.69263875e-01 -4.45251644e-01
7.41499722e-01 -9.43499923e-01 -1.08865285e+00 7.07959890e-01
-9.10274446e-01 -7.21394792e-02 -1.26128746e-02 6.90204859e-01
-4.86959904e-01 3.53478730e-01 -7.36818671e-01 -7.64882147e-01
-7.03377306e-01 -9.55490828e-01 1.28668606e+00 3.25148940e-01
3.73002291e-01 -5.12632728e-01 -1.57733008e-01 4.27142560e-01
5.46532393e-01 6.58822238e-01 5.91628969e-01 2.07631290e-01
-1.11131227e+00 -3.66993427e-01 -6.35312915e-01 3.05164546e-01
-1.54262215e-01 -2.52728820e-01 -8.49923491e-01 -3.02331209e-01
7.91545510e-02 -4.47797954e-01 7.85349548e-01 6.39676154e-01
1.22216129e+00 8.81501958e-02 -2.46758506e-01 1.08653057e+00
2.00266242e+00 3.51990461e-01 1.14974368e+00 6.34944320e-01
7.29224324e-01 2.76733935e-01 1.01193213e+00 3.75551581e-01
2.67791122e-01 7.67353237e-01 5.03377616e-01 -6.03798293e-02
-3.78546268e-01 -3.70017946e-01 1.03042051e-01 1.11763465e+00
-4.11232054e-01 -1.89271510e-01 -7.18441308e-01 2.29665697e-01
-1.63816416e+00 -6.83799565e-01 -2.36726850e-01 2.55506325e+00
3.97519469e-01 3.66582006e-01 -3.91569585e-01 1.72223881e-01
4.87129807e-01 8.56484622e-02 -3.48529279e-01 -2.70542026e-01
-1.81796685e-01 4.35558110e-01 9.21435952e-01 6.61244869e-01
-7.62573361e-01 7.36119986e-01 5.96918154e+00 8.61272275e-01
-1.25465596e+00 -1.62706390e-01 4.85691093e-02 -1.65465906e-01
-1.07284822e-01 -2.24799588e-01 -9.71949220e-01 3.77129108e-01
5.89036047e-01 4.86057922e-02 1.32414699e-01 3.11249793e-01
3.74190100e-02 -5.02622843e-01 -6.54745162e-01 1.20266616e+00
2.47616321e-01 -1.27186728e+00 -3.64540070e-02 2.21005186e-01
6.86973155e-01 6.79955333e-02 -1.39049634e-01 -1.72046032e-02
-3.76869828e-01 -1.02831221e+00 4.78042901e-01 8.78933549e-01
9.90708411e-01 -1.06197572e+00 9.25304651e-01 3.14991772e-01
-1.35140002e+00 2.86816299e-01 -6.12569094e-01 -1.96735889e-01
3.43895793e-01 7.85196126e-01 -8.54673624e-01 7.90885210e-01
7.48773098e-01 5.94436586e-01 -3.09554338e-01 1.27490664e+00
4.03301157e-02 1.54668435e-01 -5.87028205e-01 1.36842746e-02
5.09522557e-02 -1.93542093e-01 5.22492528e-01 7.42824793e-01
6.50464475e-01 -1.52289132e-02 -7.88090453e-02 7.44469523e-01
-1.02596164e-01 1.19759336e-01 -7.24499583e-01 4.48286057e-01
3.42574090e-01 8.80018294e-01 -3.94957781e-01 -9.41004157e-02
-2.71754712e-01 7.52201438e-01 1.19727187e-01 6.05299994e-02
-7.44934380e-01 -9.13509965e-01 1.30895033e-01 7.98138261e-01
4.45683628e-01 -2.58523941e-01 -3.63849252e-01 -9.16481733e-01
4.05224472e-01 -1.73571333e-01 -1.19523801e-01 -1.17231774e+00
-8.84216547e-01 5.06382346e-01 1.05766870e-01 -1.49963856e+00
5.08802161e-02 -7.09717810e-01 -3.47778738e-01 8.55775297e-01
-1.58866990e+00 -1.21623588e+00 -6.60469472e-01 4.66975957e-01
4.09407318e-01 -8.37910697e-02 4.20990705e-01 5.53234220e-01
-1.82099968e-01 4.99219418e-01 1.12219505e-01 -1.24944352e-01
6.80103600e-01 -7.70178199e-01 8.04071128e-02 6.37834430e-01
-5.65076053e-01 2.90087938e-01 5.70858240e-01 -9.21651185e-01
-1.30620587e+00 -9.53513205e-01 8.37927401e-01 1.72614098e-01
-3.82932015e-02 -2.24864230e-01 -9.57575440e-01 5.83834767e-01
-1.71061769e-01 -9.29187834e-02 4.46724799e-03 -2.56290168e-01
1.66410044e-01 -3.57451975e-01 -1.49934697e+00 1.98779181e-01
1.09166217e+00 -2.10906953e-01 -5.23404241e-01 9.17357430e-02
6.60296381e-01 -8.59536707e-01 -1.31657064e+00 1.18356538e+00
1.12734246e+00 -1.29638267e+00 1.01721525e+00 4.62584198e-01
8.29018533e-01 -4.46669638e-01 -4.02710974e-01 -7.27031767e-01
-2.11243361e-01 -8.54721218e-02 -8.76253992e-02 8.93927574e-01
1.34129539e-01 -6.78110957e-01 1.21345448e+00 2.08152592e-01
-4.26293433e-01 -1.35480332e+00 -8.22011411e-01 -8.14569473e-01
3.86122577e-02 -2.04267874e-01 6.40286088e-01 4.82392311e-01
-6.00388467e-01 3.74892116e-01 -5.38737655e-01 2.72845328e-01
8.96178842e-01 3.05838525e-01 8.27309072e-01 -1.35974252e+00
9.72480252e-02 4.01466191e-02 -5.85046053e-01 -1.33089042e+00
-3.56608659e-01 -7.66780794e-01 6.02052063e-02 -1.68048477e+00
6.47784099e-02 -4.22966689e-01 -1.01379573e-01 -1.18891308e-02
4.55659389e-01 4.95913327e-01 8.80203862e-03 2.60317862e-01
3.82309477e-03 8.87176871e-01 1.81811023e+00 1.35294870e-01
-1.06579661e-01 3.46515216e-02 -8.10552537e-02 6.26890481e-01
6.67954147e-01 -3.67500722e-01 -2.01680854e-01 -4.34974998e-01
1.37662619e-01 3.00477624e-01 3.54044348e-01 -1.48520732e+00
2.94264108e-01 1.48773551e-01 8.04333508e-01 -1.36317790e+00
7.88608789e-01 -7.35267639e-01 2.13234827e-01 4.10019875e-01
2.23590925e-01 -5.44141121e-02 2.05279123e-02 2.19541386e-01
-3.49609703e-01 -5.39376438e-02 9.49035108e-01 -1.57014072e-01
-6.33949459e-01 5.84576309e-01 2.74013728e-01 -3.23038965e-01
7.08089650e-01 -6.67806208e-01 -1.15923315e-01 -2.49184757e-01
-4.34917808e-01 -1.12243546e-02 7.06504345e-01 -4.39953618e-03
1.25535321e+00 -1.58841145e+00 -5.79603374e-01 4.07153815e-01
-4.17445712e-02 4.73711342e-01 1.64091706e-01 7.38494933e-01
-1.05288434e+00 5.67377448e-01 -5.27267933e-01 -6.35852754e-01
-1.17653835e+00 -4.56527527e-03 4.42752004e-01 -4.03915942e-02
-1.13479364e+00 8.75302911e-01 2.98995506e-02 -5.24450064e-01
1.57944381e-01 -4.50388223e-01 4.46442254e-02 -4.27325845e-01
1.51971221e-01 6.13888562e-01 2.27143437e-01 -5.33924282e-01
-2.09084332e-01 1.18652689e+00 -5.64694442e-02 -2.18458354e-01
1.43732309e+00 -1.39599383e-01 -2.78121568e-02 3.08845192e-01
1.25562310e+00 -1.18308976e-01 -1.08160949e+00 -4.78896722e-02
-5.99118114e-01 -8.79834175e-01 3.22890937e-01 -8.36886525e-01
-1.17593563e+00 9.46171880e-01 9.49515164e-01 -3.37803423e-01
1.27774501e+00 -3.18009317e-01 1.39333165e+00 3.22352380e-01
5.66957891e-01 -9.85403955e-01 -5.04216254e-02 5.09522617e-01
1.07585287e+00 -1.12569296e+00 4.16854948e-01 -6.47020876e-01
-7.98725989e-03 1.10902762e+00 8.58382046e-01 -4.73803163e-01
8.04854393e-01 2.71188885e-01 -2.13996798e-01 -6.00500047e-01
-3.31180096e-01 -3.74182989e-03 2.68174946e-01 4.14919496e-01
4.55968350e-01 -1.04302257e-01 -5.86660981e-01 2.17923060e-01
-5.05046666e-01 3.52004468e-01 3.71463925e-01 1.11143649e+00
-8.11080396e-01 -7.38335609e-01 -3.99451882e-01 5.63417077e-01
-3.36282015e-01 1.06494889e-01 1.07609302e-01 1.02291441e+00
2.56357312e-01 3.76458406e-01 2.00216442e-01 -6.51274443e-01
6.21041238e-01 -2.52351612e-01 8.61463070e-01 -3.36271644e-01
-1.42729908e-01 2.11635932e-01 -3.74699570e-02 -5.94754934e-01
-4.89444792e-01 -3.03925365e-01 -1.18213427e+00 -3.53244424e-01
-3.08035284e-01 -2.70241708e-01 7.99570203e-01 6.91471457e-01
2.20066607e-02 4.54133451e-01 6.76906466e-01 -8.65763962e-01
-7.17139542e-02 -1.00382435e+00 -9.47083354e-01 8.72039497e-02
1.18072242e-01 -1.05590916e+00 -3.01809937e-01 -4.74628389e-01] | [8.746968269348145, -2.615817070007324] |
e712b1cc-5798-4a07-883a-067d8ebad458 | budget-constrained-interactive-search-for | 2012.01945 | null | https://arxiv.org/abs/2012.01945v3 | https://arxiv.org/pdf/2012.01945v3.pdf | Budget Constrained Interactive Search for Multiple Targets | Interactive graph search leverages human intelligence to categorize target labels in a hierarchy, which are useful for image classification, product categorization, and database search. However, many existing studies of interactive graph search aim at identifying a single target optimally, and suffer from the limitations of asking too many questions and not being able to handle multiple targets. To address these two limitations, in this paper, we study a new problem of budget constrained interactive graph search for multiple targets called kBM-IGS-problem. Specifically, given a set of multiple targets T in a hierarchy, and two parameters k and b, the goal is to identify a k-sized set of selections S such that the closeness between selections S and targets T is as small as possible, by asking at most a budget of b questions. We theoretically analyze the updating rules and design a penalty function to capture the closeness between selections and targets. To tackle the kBM-IGS-problem, we develop a novel framework to ask questions using the best vertex with the largest expected gain, which makes a balanced trade-off between target probability and benefit gain. Based on the kBM-IGS framework, we first propose an efficient algorithm STBIS to handle the SingleTarget problem, which is a special case of kBM-IGS. Then, we propose a dynamic programming based method kBM-DP to tackle the MultipleTargets problem. To further improve efficiency, we propose two heuristic but efficient algorithms kBM-Topk and kBM-DP+. kBM-Topk develops a variant gain function and selects the top-k vertices independently. kBM-DP+ uses an upper bound of gains and prunes disqualified vertices to save computations. Experiments on large real-world datasets with ground-truth targets verify both the effectiveness and efficiency of our proposed algorithms. | ['Jianliang Xu', 'Zhaonian Zou', 'Jiaxin Jiang', 'Byron Choi', 'Xin Huang', 'Xuliang Zhu'] | 2020-12-03 | null | null | null | null | ['product-categorization'] | ['miscellaneous'] | [ 2.29124039e-01 3.16355497e-01 -7.70988643e-01 -1.98347270e-01
-8.59364450e-01 -7.05421925e-01 -4.15128022e-02 5.16428709e-01
-2.70454705e-01 3.54856580e-01 -2.86011368e-01 -4.46667880e-01
-7.19992042e-01 -9.35704768e-01 -6.37468755e-01 -5.21072149e-01
-9.54073593e-02 7.60917604e-01 7.22865403e-01 -1.75662767e-02
3.23850393e-01 3.29192936e-01 -1.21504104e+00 -4.18730825e-02
1.00806391e+00 1.12416518e+00 2.32346803e-01 2.71095186e-01
7.62226656e-02 3.63764018e-01 -2.99536616e-01 -6.75428092e-01
5.54828763e-01 -4.38722074e-01 -8.93991351e-01 1.93911880e-01
2.05038533e-01 -4.50626686e-02 -6.13931902e-02 1.12855470e+00
4.04637188e-01 2.50486493e-01 6.48860812e-01 -1.67492938e+00
-3.55647802e-01 6.92665815e-01 -1.04151583e+00 2.35628784e-01
3.30086619e-01 -2.94430703e-02 1.50969696e+00 -5.02201855e-01
3.91258687e-01 1.23008120e+00 2.93918759e-01 2.48477623e-01
-1.45313096e+00 -6.84288263e-01 6.20531738e-01 2.86513954e-01
-1.66769052e+00 6.02863729e-02 7.11501002e-01 -3.15824777e-01
4.87960249e-01 6.98343813e-01 6.25444412e-01 4.40119594e-01
-8.16125944e-02 7.57256567e-01 8.97100806e-01 -6.10423982e-01
3.40259910e-01 3.95920426e-01 5.06182849e-01 8.00792038e-01
4.77248967e-01 2.63336971e-02 -5.85798800e-01 -5.57125568e-01
5.34200609e-01 -1.11857049e-01 -4.22192544e-01 -6.14087164e-01
-9.56663549e-01 1.20108104e+00 5.59586167e-01 -1.07151978e-01
-2.98543453e-01 -1.42241837e-02 5.48615344e-02 8.93863142e-02
2.73343712e-01 4.63153124e-01 -4.31194782e-01 4.90949810e-01
-6.49703801e-01 1.73111245e-01 6.46647751e-01 1.09945762e+00
8.64492536e-01 -6.13862336e-01 -4.73763227e-01 7.55432606e-01
3.77123654e-01 3.23535293e-01 -7.72414804e-02 -5.75689435e-01
4.53177959e-01 9.66025710e-01 1.35256544e-01 -1.50893223e+00
-3.14604938e-01 -3.74227464e-01 -5.67070544e-01 -8.05380419e-02
1.96028054e-01 2.83096313e-01 -9.22179043e-01 1.83802295e+00
8.50507498e-01 2.79498994e-02 -5.88057995e-01 1.05955517e+00
4.87695724e-01 6.01545453e-01 -8.88649281e-03 -4.74017829e-01
1.40747881e+00 -9.44042087e-01 -2.35624403e-01 -2.44642138e-01
6.43792391e-01 -3.15495849e-01 1.34389126e+00 2.07521021e-01
-7.37001002e-01 -4.84220646e-02 -7.99429476e-01 3.95678639e-01
-1.82484418e-01 -1.53254539e-01 5.10066986e-01 8.98447216e-01
-9.93173063e-01 1.11754976e-01 -2.27514490e-01 -2.27750629e-01
1.50658503e-01 5.84690809e-01 4.82599773e-02 -1.26836672e-01
-1.13939059e+00 5.37218153e-01 6.69890344e-01 -1.84973255e-01
-9.03333664e-01 -7.07720280e-01 -5.78179061e-01 1.05848439e-01
1.23801780e+00 -6.46319568e-01 1.00062799e+00 -7.19469309e-01
-1.16216791e+00 6.78612053e-01 2.64211502e-02 -2.61827826e-01
3.65324944e-01 3.80314857e-01 -1.46329671e-01 1.05537705e-01
3.52523923e-01 5.35770297e-01 7.16382027e-01 -1.21590269e+00
-9.42010343e-01 -5.71271777e-01 4.58382696e-01 4.69531029e-01
-6.21931076e-01 -1.04792416e-01 -1.08201778e+00 -3.70915592e-01
2.51599431e-01 -1.03019512e+00 -6.08956635e-01 -3.86643112e-01
-8.05220127e-01 -3.43868464e-01 3.49137783e-01 -3.01819533e-01
1.45274556e+00 -1.92381489e+00 2.03152791e-01 8.60066175e-01
5.22916198e-01 -5.84025308e-02 -3.02055478e-01 2.21482396e-01
2.34534517e-01 3.46814454e-01 6.05300292e-02 1.80912822e-01
-1.01563528e-01 6.43153861e-02 -3.73863950e-02 2.57143348e-01
-3.00942779e-01 8.26531231e-01 -9.05230045e-01 -6.46120310e-01
-1.77888200e-01 -2.16877908e-01 -5.70374429e-01 5.73997721e-02
-4.40966040e-01 2.02414431e-02 -6.85080051e-01 7.06843376e-01
6.47159755e-01 -7.00261116e-01 4.40874994e-01 -2.59824097e-01
2.00727716e-01 -3.20958672e-03 -1.44771349e+00 8.70198429e-01
-2.75163859e-01 -6.98894784e-02 6.21089004e-02 -9.47986722e-01
7.27154315e-01 -2.69589275e-01 4.76764828e-01 -6.39012754e-01
1.33110166e-01 1.30733132e-01 -1.56615630e-01 -9.05690491e-02
2.32853785e-01 1.27297476e-01 -2.60324329e-01 2.36529410e-01
-3.79619837e-01 3.25149715e-01 2.08269194e-01 5.40377140e-01
1.20872486e+00 -6.05813503e-01 5.28744459e-01 -2.69808799e-01
4.82719392e-01 1.67354882e-01 6.59880161e-01 1.13958824e+00
-1.27901465e-01 1.86808020e-01 7.54516959e-01 -2.54502986e-02
-3.61819744e-01 -8.23095560e-01 2.64012277e-01 1.38309491e+00
8.07542980e-01 -4.46186870e-01 -5.33191800e-01 -1.09304118e+00
2.06640422e-01 7.20734775e-01 -6.79061770e-01 -2.09866762e-01
-1.24989443e-01 -7.21497715e-01 -2.74382085e-02 6.54203594e-02
1.13762997e-01 -6.86002910e-01 -6.17613852e-01 1.89666264e-02
-2.15440989e-01 -8.33056450e-01 -1.01952517e+00 1.19784832e-01
-5.43879747e-01 -1.29866052e+00 -4.83526289e-01 -7.02448606e-01
8.78944695e-01 5.74643433e-01 1.05004466e+00 8.65776688e-02
-2.09567234e-01 4.65382338e-01 -4.17651445e-01 -3.26367795e-01
-3.86538953e-02 3.28013629e-01 -3.63276422e-01 1.59724697e-01
4.64099124e-02 -1.91642225e-01 -6.47850573e-01 7.89530993e-01
-7.41262674e-01 7.21280202e-02 6.69970572e-01 6.78752184e-01
9.78178918e-01 4.51625526e-01 4.72113281e-01 -1.00447357e+00
7.04185665e-01 -5.48287034e-01 -1.09656918e+00 6.23526514e-01
-1.12925923e+00 -1.69495121e-02 2.33074337e-01 -6.66800916e-01
-5.02805889e-01 1.12944834e-01 4.22992855e-01 -4.32056457e-01
4.45846319e-01 7.46684372e-01 -5.71928501e-01 -3.52849752e-01
4.54356998e-01 -3.48166078e-02 -3.18651825e-01 -1.35087118e-01
4.27396625e-01 3.93748879e-01 1.85110345e-01 -3.77067149e-01
7.09958851e-01 1.80974811e-01 4.30342257e-02 -4.56766397e-01
-8.49622905e-01 -8.53498280e-01 -4.97957692e-02 -4.05194491e-01
4.21238661e-01 -4.21664685e-01 -1.10792410e+00 -1.08363500e-04
-7.49227464e-01 -9.30728912e-02 -1.37177855e-01 6.80929124e-02
-2.74883866e-01 4.53301251e-01 -1.11152425e-01 -1.01733315e+00
-3.45209181e-01 -1.17372620e+00 8.85885596e-01 1.96617141e-01
1.11809252e-02 -6.80350959e-01 -2.35114604e-01 4.68770117e-01
-1.82809886e-02 2.52441525e-01 1.24893856e+00 -8.25145662e-01
-9.25433159e-01 -1.19219907e-01 -3.02521199e-01 -2.63428032e-01
1.26860468e-02 -5.88688612e-01 -1.11586928e-01 -4.32883501e-01
-2.13197678e-01 -1.35245293e-01 6.90819502e-01 4.05397415e-01
1.11323881e+00 -6.93692327e-01 -9.12238061e-01 3.41326833e-01
1.47482848e+00 4.95601505e-01 3.08916003e-01 5.25390804e-01
6.32126689e-01 5.84879458e-01 1.15821636e+00 4.70224977e-01
4.78643894e-01 9.48260307e-01 7.09772885e-01 9.74074379e-02
3.05714965e-01 -3.15074056e-01 -1.49693713e-01 3.31942618e-01
3.64955932e-01 -6.98766649e-01 -8.74287784e-01 6.14230454e-01
-1.98156047e+00 -2.71194875e-01 -7.38646910e-02 2.65326858e+00
7.58328140e-01 3.38187218e-01 4.21493351e-01 1.00551598e-01
9.56176400e-01 -8.60878229e-02 -8.97503674e-01 2.05098419e-03
2.18550190e-01 -2.11197346e-01 8.32547903e-01 5.61784744e-01
-9.46328223e-01 8.24506998e-01 5.54513645e+00 1.19547236e+00
-7.68739700e-01 9.38579887e-02 7.28802502e-01 -1.15593635e-01
-5.64409554e-01 3.29688847e-01 -1.07141340e+00 4.76777941e-01
6.95504427e-01 -4.39830393e-01 6.05152607e-01 8.39098275e-01
4.99849720e-03 -2.72312224e-01 -1.10897505e+00 1.00471735e+00
-4.27454747e-02 -1.15301204e+00 2.39805683e-01 1.52297467e-01
4.82273489e-01 -4.44189459e-01 5.34323081e-02 1.74178004e-01
4.94991392e-01 -6.43695951e-01 6.44180119e-01 1.80884786e-02
5.92531800e-01 -8.72540593e-01 3.10494721e-01 6.58934712e-01
-1.32002747e+00 -3.01174909e-01 -1.96680889e-01 5.58742106e-01
5.19978553e-02 7.13656068e-01 -9.42259431e-01 7.24388182e-01
7.67253339e-01 7.12121204e-02 -4.82746273e-01 1.15525210e+00
-1.02057070e-01 5.32779515e-01 -5.21895885e-01 -2.97672719e-01
1.52405962e-01 -1.85095057e-01 6.23861432e-01 7.88857281e-01
1.00072779e-01 3.41633707e-01 7.72629917e-01 5.45969188e-01
-1.05167471e-01 6.03686213e-01 -1.97105035e-01 3.57876457e-02
8.04727674e-01 1.28049016e+00 -1.07340431e+00 -6.60711750e-02
-1.17031150e-01 6.97609067e-01 4.87106740e-01 2.26060942e-01
-1.16223180e+00 -2.96539634e-01 1.42128736e-01 4.37311977e-01
2.56234527e-01 2.84461856e-01 -1.23768337e-01 -6.24831617e-01
-1.85443126e-02 -1.05678046e+00 1.04509556e+00 -3.85194123e-01
-1.17512763e+00 3.41763943e-01 3.32804114e-01 -9.09567356e-01
3.20377916e-01 -1.52616620e-01 -3.27224553e-01 6.22931719e-01
-1.22920001e+00 -1.03888547e+00 -3.01986367e-01 5.58500290e-01
2.83517003e-01 2.57198811e-01 2.50251055e-01 1.77136153e-01
-5.92519879e-01 9.15104866e-01 -2.44270653e-01 -4.79174733e-01
3.67739290e-01 -9.93925333e-01 5.44477664e-02 7.13505864e-01
1.02838203e-01 4.07030582e-01 6.24713480e-01 -7.59513557e-01
-1.28245580e+00 -1.17729974e+00 4.63625103e-01 -1.90528836e-02
4.40248936e-01 -3.33909869e-01 -7.18407035e-01 4.44743395e-01
-3.66526723e-01 -1.83984503e-01 4.82262105e-01 1.82360515e-01
-3.08038145e-01 -2.74522275e-01 -1.25996041e+00 6.54360294e-01
1.19140935e+00 2.50488357e-03 -1.39396731e-02 6.63441122e-01
9.78006661e-01 -3.14715832e-01 -7.14409471e-01 6.34539247e-01
1.66734040e-01 -7.98949301e-01 9.36060369e-01 -3.93036067e-01
-2.10086450e-01 -3.03133667e-01 -7.98980519e-02 -1.16637158e+00
-6.20451450e-01 -6.31335258e-01 -8.68753791e-02 1.12264895e+00
7.12865412e-01 -8.71125281e-01 9.78836298e-01 5.85399389e-01
3.32519621e-01 -1.15050936e+00 -8.44242811e-01 -9.81907666e-01
-4.89833653e-01 -1.55767605e-01 5.85611939e-01 6.87341511e-01
-1.45192370e-01 5.29212594e-01 -4.09715325e-01 6.56323254e-01
8.22234869e-01 4.60936666e-01 7.30996192e-01 -1.24681640e+00
-6.30555391e-01 -3.34493458e-01 -2.57920623e-01 -1.13823652e+00
-4.71137911e-02 -8.04893315e-01 7.89017826e-02 -1.66670501e+00
6.23398662e-01 -7.23235488e-01 -3.69176984e-01 5.93245387e-01
-3.92291576e-01 -1.90465897e-02 1.30093634e-01 2.49070838e-01
-8.05901945e-01 2.21001580e-01 1.01821947e+00 -2.36810237e-01
-5.78232706e-01 1.79993689e-01 -1.08916867e+00 4.67266560e-01
6.04707062e-01 -6.22949064e-01 -7.80028820e-01 7.63430968e-02
4.06591475e-01 2.89006531e-01 1.73899308e-01 -3.89946491e-01
3.56791645e-01 -6.75172985e-01 -9.12078544e-02 -8.69182289e-01
1.24883220e-01 -7.77394056e-01 3.64289314e-01 5.42562604e-01
-5.39569318e-01 -2.25976631e-01 -9.13010240e-02 1.10221243e+00
1.56980976e-01 -2.56184965e-01 7.33727694e-01 1.02474935e-01
-7.19912231e-01 5.41571021e-01 -9.01288446e-03 5.49600646e-02
1.53729045e+00 -1.75513431e-01 -2.40762457e-01 -4.11513209e-01
-5.83111763e-01 8.01886380e-01 2.69253254e-01 4.53885555e-01
5.40536284e-01 -1.14342690e+00 -3.46111864e-01 -1.70701995e-01
4.45571333e-01 -3.25932279e-02 1.58511966e-01 8.87323022e-01
-9.40283686e-02 3.37730527e-01 3.87674481e-01 -5.54906130e-01
-1.51018560e+00 9.94465113e-01 1.79461107e-01 -6.22179449e-01
-2.45191753e-01 9.91814673e-01 5.29789388e-01 -2.77862102e-01
4.11227077e-01 1.19058207e-01 -2.28151560e-01 3.13873924e-02
1.30353913e-01 5.18098593e-01 3.08618955e-02 -3.51632863e-01
-5.19108593e-01 3.59706104e-01 -2.91402668e-01 -1.95151139e-02
8.83900642e-01 -2.89471895e-01 -1.22763894e-01 -6.46925718e-02
9.13685143e-01 7.40134642e-02 -6.88656569e-01 -2.54088730e-01
3.34856898e-01 -7.33313501e-01 1.66604906e-01 -8.14758122e-01
-1.25039577e+00 7.04034939e-02 5.02194643e-01 4.71565366e-01
1.54788733e+00 3.06212604e-01 5.99706233e-01 1.34143919e-01
6.55808747e-01 -7.58038819e-01 1.82092369e-01 6.80914372e-02
6.96494460e-01 -9.47050273e-01 -1.93087161e-02 -1.06846309e+00
-6.52031004e-01 5.29010236e-01 9.34907496e-01 4.09994721e-01
6.15846395e-01 -1.08461648e-01 -3.62888366e-01 -4.59228933e-01
-5.85439503e-01 -3.40918720e-01 4.43442941e-01 2.35510662e-01
-2.95109719e-01 2.24835336e-01 -5.76282978e-01 5.31686962e-01
2.20325261e-01 -3.52524221e-01 2.22493932e-01 8.77587080e-01
-6.24469578e-01 -1.13813627e+00 -4.73396480e-01 7.86595881e-01
-1.97445631e-01 3.29764374e-02 -5.69730163e-01 6.15075290e-01
1.14051495e-02 1.21403623e+00 -4.34616268e-01 -6.31319046e-01
5.50134301e-01 -4.48845893e-01 2.96411335e-01 -6.18518412e-01
-3.79799753e-01 1.43569350e-01 1.74083356e-02 -5.07753372e-01
-1.01223819e-01 -2.20512837e-01 -1.04660106e+00 -1.23197511e-01
-1.05263400e+00 3.87149304e-01 3.42085063e-01 5.90576828e-01
5.18311262e-01 2.72752553e-01 9.07007992e-01 -2.13578343e-01
-8.43037307e-01 -4.43666816e-01 -7.31240034e-01 1.26900956e-01
-1.46941021e-01 -8.89004290e-01 -3.03823054e-01 -4.52516168e-01] | [7.141654014587402, 5.260804653167725] |
54f90ee9-b433-4242-bbf4-6b9925696fd6 | instant-3d-object-tracking-with-applications | 2006.13194 | null | https://arxiv.org/abs/2006.13194v1 | https://arxiv.org/pdf/2006.13194v1.pdf | Instant 3D Object Tracking with Applications in Augmented Reality | Tracking object poses in 3D is a crucial building block for Augmented Reality applications. We propose an instant motion tracking system that tracks an object's pose in space (represented by its 3D bounding box) in real-time on mobile devices. Our system does not require any prior sensory calibration or initialization to function. We employ a deep neural network to detect objects and estimate their initial 3D pose. Then the estimated pose is tracked using a robust planar tracker. Our tracker is capable of performing relative-scale 9-DoF tracking in real-time on mobile devices. By combining use of CPU and GPU efficiently, we achieve 26-FPS+ performance on mobile devices. | ['Matthias Grundmann', 'Tingbo Hou', 'Jianing Wei', 'Liangkai Zhang', 'Artsiom Ablavatski', 'Adel Ahmadyan'] | 2020-06-23 | null | null | null | null | ['3d-object-tracking'] | ['computer-vision'] | [-2.21983284e-01 -4.05523479e-01 -2.25259379e-01 1.95668638e-01
-6.21934891e-01 -7.82035112e-01 2.53297657e-01 -3.00694942e-01
-5.97754478e-01 3.91564012e-01 -3.31053555e-01 -3.64525408e-01
3.82532418e-01 -5.32138646e-01 -1.07167077e+00 -3.38776469e-01
-5.05028367e-02 4.54539329e-01 6.01681292e-01 1.26038194e-01
-1.17008224e-01 1.08907354e+00 -1.37978554e+00 -4.35872495e-01
1.34814128e-01 1.19828641e+00 2.16875091e-01 1.18958628e+00
4.30848926e-01 1.24053843e-01 -6.43105268e-01 1.33591011e-01
3.46704036e-01 3.84559453e-01 4.43529151e-02 2.08461210e-01
7.25494266e-01 -7.80604839e-01 -5.21969855e-01 1.07886922e+00
6.61487937e-01 -5.16383648e-02 1.24034233e-01 -1.13499987e+00
-3.09594989e-01 -2.01125413e-01 -7.11537778e-01 1.86890200e-01
8.19387734e-01 -8.67404118e-02 1.80998161e-01 -1.06171536e+00
6.00363731e-01 1.07055378e+00 1.10736251e+00 5.20437956e-01
-7.59199798e-01 -3.86129916e-01 2.39415035e-01 -4.57755744e-01
-1.54393077e+00 -4.16412890e-01 6.23064280e-01 -4.00022000e-01
1.01885676e+00 2.39867121e-01 9.75304484e-01 8.70577753e-01
5.08626819e-01 7.28912652e-01 2.43603766e-01 -2.58653730e-01
1.35638744e-01 -1.62783772e-01 1.34423524e-01 7.96732605e-01
6.09755933e-01 1.97882876e-01 -3.36327076e-01 -2.25662544e-01
1.59168828e+00 3.46605837e-01 -4.71986264e-01 -8.64159226e-01
-1.52564108e+00 1.95666760e-01 5.68097234e-01 -1.47964463e-01
-3.01366389e-01 8.26155961e-01 1.64087817e-01 -2.08516672e-01
2.64272988e-01 -1.21804401e-01 -5.66632509e-01 -4.06158388e-01
-4.40231591e-01 1.14016123e-01 5.95126688e-01 1.43212640e+00
8.10468420e-02 2.37482846e-01 1.84412777e-01 6.24497868e-02
7.56543398e-01 1.05055356e+00 3.32129478e-01 -1.03523970e+00
1.13066323e-01 4.49791968e-01 7.66373456e-01 -9.58804190e-01
-5.45074344e-01 -4.58002120e-01 -3.62943858e-01 3.39922696e-01
1.90962598e-01 -3.24259698e-01 -8.50877523e-01 1.23349428e+00
9.71118629e-01 5.93836904e-01 -3.76021892e-01 1.10996509e+00
8.72674882e-01 4.72909629e-01 -3.53020966e-01 -7.55942985e-02
1.47736275e+00 -7.74912298e-01 -6.73567593e-01 -1.23323821e-01
4.12970752e-01 -6.50285780e-01 5.92040539e-01 3.88004094e-01
-1.12688482e+00 -7.52178371e-01 -1.29238594e+00 -1.48280904e-01
-1.73516907e-02 5.89297354e-01 6.39073014e-01 9.18985009e-01
-9.18094754e-01 2.36793712e-01 -1.51766741e+00 -2.12064877e-01
2.17877045e-01 1.06189096e+00 -2.18479961e-01 5.74139178e-01
-5.18041492e-01 6.55910492e-01 1.97401732e-01 3.84620756e-01
-6.10546768e-01 -4.58833784e-01 -8.75616074e-01 -9.77914929e-02
3.35054308e-01 -1.03827178e+00 1.75722206e+00 -3.70941877e-01
-1.81853986e+00 7.79275954e-01 -4.25300717e-01 -2.97387868e-01
6.17559254e-01 -9.42952037e-01 -3.74813139e-01 -1.69471860e-01
-1.71482712e-01 3.29499036e-01 8.21615517e-01 -1.25348639e+00
-5.72644353e-01 -5.51021874e-01 -2.11750977e-02 3.93860608e-01
-9.24033970e-02 9.93317217e-02 -1.07821465e+00 -2.24839926e-01
6.88468158e-01 -1.31984806e+00 -1.99733049e-01 7.27877915e-01
-7.65775442e-02 9.22221392e-02 1.24451721e+00 -3.74270082e-01
8.03541720e-01 -2.00652599e+00 -3.34771335e-01 -3.94282304e-02
3.81521046e-01 5.17725468e-01 4.48959291e-01 -3.13260972e-01
2.77318269e-01 -5.12372613e-01 5.47791898e-01 -5.22052884e-01
-1.69730280e-02 -1.59028366e-01 -2.85718858e-01 8.49021614e-01
-3.28912497e-01 1.10179329e+00 -9.00804102e-01 -2.26198480e-01
4.08829421e-01 1.18394482e+00 -4.48448926e-01 5.31505756e-02
4.90319356e-03 4.25292194e-01 -5.13461471e-01 8.31514657e-01
8.23925972e-01 -4.93440539e-01 4.96679395e-02 -2.31260702e-01
-1.40248537e-01 3.81514072e-01 -1.47077048e+00 1.82323289e+00
-4.08428460e-01 7.40878522e-01 1.59657404e-01 -7.24383220e-02
9.02818263e-01 2.88164914e-01 5.48816383e-01 -1.63546175e-01
6.01769686e-01 1.13525897e-01 -3.01560551e-01 4.36747745e-02
9.64941740e-01 4.10911292e-01 -1.01842552e-01 3.94693166e-01
-2.85184324e-01 2.58463889e-01 -5.73411286e-01 -8.47554579e-02
1.01659596e+00 6.74494326e-01 4.27376509e-01 1.86014414e-01
2.62278706e-01 -1.13037705e-01 4.25659031e-01 5.92187822e-01
-4.22804922e-01 5.14161646e-01 -4.55714762e-01 -6.49075925e-01
-9.54394758e-01 -1.38002563e+00 -1.34144738e-01 7.07924306e-01
5.73283494e-01 -1.12402275e-01 -3.46250057e-01 -4.16138709e-01
3.12570572e-01 -1.73092902e-01 -2.07788721e-01 4.14874591e-02
-9.69184279e-01 -1.64022297e-01 1.10232167e-01 1.15738928e+00
1.81463122e-01 -6.72146261e-01 -1.27709091e+00 1.82991818e-01
2.92004436e-01 -1.26846087e+00 -8.06750417e-01 7.72409812e-02
-1.12345028e+00 -9.65258121e-01 -7.51653254e-01 -9.38635886e-01
8.54556084e-01 8.20438981e-01 9.02539253e-01 -4.07032520e-02
5.32570817e-02 5.97127378e-01 1.04884066e-01 -4.17542756e-01
2.54682571e-01 -3.14623594e-01 5.29852390e-01 -3.60111415e-01
4.00790900e-01 -1.40476584e-01 -7.01884210e-01 5.43782771e-01
-6.65309802e-02 -2.01040551e-01 2.54833959e-02 2.58916438e-01
8.75075161e-01 -2.39888176e-01 -1.41564637e-01 -1.49490446e-01
1.67033732e-01 7.07002729e-02 -1.32910156e+00 -8.90725702e-02
2.79607564e-01 -5.81401885e-01 -4.86254180e-03 -1.07641470e+00
-6.66813850e-01 8.64988804e-01 -1.68772023e-02 -7.90602863e-01
3.22369188e-02 1.01420596e-01 -1.77858308e-01 -5.59530199e-01
4.91743863e-01 -1.03408076e-01 -3.35628241e-01 -3.80221844e-01
2.89139569e-01 7.60180950e-01 1.02138364e+00 -3.74452770e-01
8.22285950e-01 7.36840248e-01 -1.15010917e-01 -6.46270633e-01
-5.95757425e-01 -5.76111376e-01 -9.81198907e-01 -4.07757521e-01
5.91054976e-01 -1.44265044e+00 -1.74383271e+00 3.71197850e-01
-1.42870939e+00 -3.05241682e-02 -9.53151733e-02 7.94483006e-01
-5.16649365e-01 2.02992260e-01 -4.83022511e-01 -1.03369772e+00
-5.16098380e-01 -1.10212469e+00 1.66682327e+00 4.29748327e-01
-1.65311188e-01 -6.99194193e-01 7.89823160e-02 -1.82695493e-01
4.46967781e-02 3.24619681e-01 -4.55505878e-01 3.88470627e-02
-9.35573697e-01 -7.62056887e-01 -1.93099752e-01 -4.53074634e-01
2.32965797e-01 1.02272265e-01 -8.59689474e-01 -5.43980777e-01
1.22951634e-01 2.62170434e-01 2.32483968e-01 8.22883308e-01
7.78102100e-01 1.35577992e-01 -9.35508728e-01 9.32300746e-01
1.21547222e+00 6.13947928e-01 2.16935709e-01 4.54637796e-01
1.04188216e+00 -4.28992033e-01 7.62556672e-01 3.23520601e-01
4.39365059e-01 1.10489321e+00 4.60715145e-01 1.18850254e-01
2.09878031e-02 -2.71442473e-01 4.31022614e-01 7.49835968e-01
-2.11690441e-02 -1.02769114e-01 -7.87304461e-01 1.37681350e-01
-1.82497513e+00 -5.07469714e-01 -4.54654157e-01 2.44198489e+00
3.17725331e-01 3.43066782e-01 3.62956703e-01 -1.39278919e-01
8.91000688e-01 -4.02190924e-01 -8.03938031e-01 1.09568253e-01
3.70311201e-01 -6.48880890e-03 6.67497933e-01 5.53863883e-01
-1.35367227e+00 9.26530242e-01 6.82914114e+00 -5.26910461e-02
-1.23568761e+00 1.44061074e-03 -1.41236782e-01 -2.91287750e-01
4.04459894e-01 -3.85597974e-01 -1.33464670e+00 3.15812171e-01
8.21504176e-01 1.26242697e-01 -5.89797199e-02 1.29388726e+00
5.29416744e-03 -7.74932504e-02 -1.02481651e+00 1.26754558e+00
2.84256693e-02 -1.22004616e+00 -6.51288271e-01 2.94848859e-01
6.13050342e-01 1.00484014e-01 1.41598120e-01 8.67387280e-02
3.79540145e-01 -4.65586782e-01 8.60435069e-01 4.29507941e-01
9.22702551e-01 -5.64969301e-01 4.66874093e-01 4.32815999e-01
-1.59470308e+00 2.30408102e-01 -4.30402488e-01 -2.43064091e-01
5.45800388e-01 2.37899274e-01 -1.05991149e+00 5.41672781e-02
7.27618635e-01 4.70497251e-01 -3.48525494e-02 1.43852127e+00
-5.93880331e-03 4.72424068e-02 -7.12194920e-01 -1.68605402e-01
-3.26986313e-01 2.29505599e-01 5.91031432e-01 7.35965431e-01
8.24907720e-01 1.29841194e-01 4.74152505e-01 4.57484841e-01
-1.37053937e-01 -4.89096403e-01 -5.62129974e-01 2.45299160e-01
6.95439994e-01 1.02448070e+00 -9.53815401e-01 -4.00537342e-01
-1.84169441e-01 1.01366389e+00 -7.26727769e-02 6.70962874e-03
-1.25831890e+00 -2.11918592e-01 7.56204605e-01 8.15497041e-02
6.64144814e-01 -1.03338897e+00 -2.37198919e-01 -1.36471450e+00
2.84306169e-01 -3.48375857e-01 -1.63046494e-01 -1.17312789e+00
-7.04099119e-01 5.81203103e-01 -5.78978956e-01 -1.68732584e+00
-9.22267977e-03 -9.83056724e-01 -1.20443769e-01 6.73556209e-01
-8.80991280e-01 -1.05863333e+00 -6.60139441e-01 6.53465807e-01
2.38091245e-01 8.49091560e-02 8.04928124e-01 3.53586137e-01
-3.96751583e-01 6.32149339e-01 -1.75374784e-02 1.95294142e-01
3.40066046e-01 -1.05796731e+00 1.21697974e+00 7.32882559e-01
3.09875607e-01 9.44793642e-01 5.05250275e-01 -1.04930627e+00
-2.09460759e+00 -9.01435912e-01 5.48672915e-01 -1.26482415e+00
5.67783058e-01 -6.37389421e-01 -7.28218615e-01 1.10034990e+00
-4.29327399e-01 7.73859322e-01 3.03733826e-01 -7.42628425e-02
-1.94270954e-01 9.83725116e-02 -1.09698963e+00 5.75001419e-01
1.35984004e+00 -4.96339023e-01 -3.90846014e-01 3.32915694e-01
1.13583231e+00 -1.73868203e+00 -7.45548189e-01 2.82462955e-01
1.01408768e+00 -5.02199352e-01 1.39826119e+00 -2.84898728e-01
-5.64610183e-01 -9.46997881e-01 -5.82895912e-02 -6.01320446e-01
-4.36021924e-01 -8.49896312e-01 -1.03712893e+00 4.09493119e-01
-1.81763202e-01 -3.78281057e-01 1.31257129e+00 4.99717444e-01
-7.06608742e-02 -5.73314726e-01 -1.06489146e+00 -8.90223622e-01
-8.25546980e-01 -6.91970587e-01 5.58337271e-01 5.03506780e-01
-3.74443412e-01 4.75161038e-02 -3.48095626e-01 9.21539664e-01
4.94978279e-01 -4.35393229e-02 1.33288801e+00 -1.25715292e+00
-1.19465366e-01 1.10296138e-01 -7.35242009e-01 -1.76588333e+00
-2.28535190e-01 -3.70016634e-01 1.16312340e-01 -1.22536302e+00
-2.41775006e-01 -2.29474097e-01 -7.75855184e-02 2.24295691e-01
1.20333537e-01 6.19563103e-01 1.70461640e-01 9.76708308e-02
-8.44731808e-01 3.86006117e-01 9.48229551e-01 2.87505031e-01
-5.25046706e-01 3.47210914e-01 -2.91572005e-01 1.01394272e+00
4.05418217e-01 -4.69911069e-01 7.13811442e-02 -6.53010845e-01
8.27677250e-02 1.64134100e-01 6.30713463e-01 -1.28651345e+00
4.11957115e-01 1.87976420e-01 9.36185360e-01 -1.29875982e+00
8.77465010e-01 -1.18119383e+00 4.69045728e-01 6.50625587e-01
4.39626396e-01 4.39877301e-01 4.98889565e-01 4.60890055e-01
4.36333120e-01 1.03232041e-01 2.89114803e-01 1.09014176e-01
-5.06178200e-01 4.84317303e-01 -4.37580161e-02 -7.22628534e-01
1.12314522e+00 -7.89867580e-01 -1.25268802e-01 -1.44056201e-01
-9.18933094e-01 -4.99743186e-02 8.00576627e-01 4.99065816e-01
7.41351962e-01 -1.74852538e+00 7.20368326e-02 1.62874579e-01
-2.36095652e-01 2.77990311e-01 1.23125110e-02 5.70111871e-01
-8.69222879e-01 5.51102221e-01 3.22290324e-02 -1.29334199e+00
-1.47228730e+00 6.43807650e-01 4.89391863e-01 2.67861545e-01
-8.27813566e-01 9.59468365e-01 -1.90610096e-01 -4.64630187e-01
5.60788751e-01 -5.14751196e-01 2.75115967e-01 -4.57745761e-01
7.44627535e-01 3.16212952e-01 8.95661786e-02 -7.85560191e-01
-7.08403409e-01 1.00047112e+00 1.39358997e-01 -1.46997631e-01
1.02683115e+00 -1.64607927e-01 4.52795804e-01 4.31358784e-01
1.05751729e+00 2.08847448e-01 -1.72736669e+00 -7.44453147e-02
-1.65772989e-01 -8.63814771e-01 1.51077911e-01 -3.51068228e-01
-9.49583590e-01 5.29461980e-01 1.13010406e+00 -1.44703791e-01
6.70216858e-01 -2.63595134e-01 8.61882746e-01 6.63902164e-01
8.02204132e-01 -6.00209892e-01 5.04015908e-02 7.95584798e-01
7.10015059e-01 -1.13330913e+00 2.33995542e-01 -4.73221689e-01
-1.12840161e-01 1.10456955e+00 6.99972749e-01 -4.40494448e-01
6.89805806e-01 7.26497293e-01 1.84151664e-01 -6.55727983e-02
-2.26841345e-01 2.55387593e-02 5.41498423e-01 8.48177969e-01
4.28643167e-01 6.25633895e-02 3.22042823e-01 2.21677810e-01
-1.95253506e-01 1.30572036e-01 3.84482771e-01 1.20879328e+00
-5.05347490e-01 -6.20971739e-01 -8.71489823e-01 -1.31816596e-01
-2.56121933e-01 4.26845908e-01 2.61481460e-02 8.25250924e-01
-8.23064595e-02 4.34911251e-01 3.29841167e-01 -3.86387348e-01
3.31824422e-01 -3.34186316e-01 8.27117443e-01 -5.46469390e-01
-5.51047921e-01 6.26313865e-01 -2.18304440e-01 -6.62497580e-01
-1.14783213e-01 -6.72678769e-01 -1.42615402e+00 -1.80920005e-01
-6.90478206e-01 -3.90976638e-01 1.01225829e+00 5.67671478e-01
5.44375777e-01 4.76312190e-01 2.64833719e-01 -1.55946946e+00
-2.23047778e-01 -4.32932764e-01 -1.75028190e-01 -3.89656872e-01
6.53336346e-01 -8.89300108e-01 2.44771287e-01 -5.61513044e-02] | [6.969762802124023, -2.119643449783325] |
1ab2efeb-6d1b-481b-9040-8d171ddd7122 | generating-personalized-recipes-from | 1909.00105 | null | https://arxiv.org/abs/1909.00105v1 | https://arxiv.org/pdf/1909.00105v1.pdf | Generating Personalized Recipes from Historical User Preferences | Existing approaches to recipe generation are unable to create recipes for users with culinary preferences but incomplete knowledge of ingredients in specific dishes. We propose a new task of personalized recipe generation to help these users: expanding a name and incomplete ingredient details into complete natural-text instructions aligned with the user's historical preferences. We attend on technique- and recipe-level representations of a user's previously consumed recipes, fusing these 'user-aware' representations in an attention fusion layer to control recipe text generation. Experiments on a new dataset of 180K recipes and 700K interactions show our model's ability to generate plausible and personalized recipes compared to non-personalized baselines. | ['Jianmo Ni', 'Shuyang Li', 'Julian McAuley', 'Bodhisattwa Prasad Majumder'] | 2019-08-31 | generating-personalized-recipes-from-1 | https://aclanthology.org/D19-1613 | https://aclanthology.org/D19-1613.pdf | ijcnlp-2019-11 | ['recipe-generation'] | ['miscellaneous'] | [ 3.41688156e-01 1.83620989e-01 -1.59348011e-01 -6.10338151e-01
-7.12937891e-01 -8.28502655e-01 5.07134795e-01 2.80768722e-01
8.44949409e-02 5.88042915e-01 1.39074647e+00 -6.12618700e-02
2.76883215e-01 -9.74780619e-01 -8.88399124e-01 -2.13994920e-01
2.30964184e-01 4.74119335e-01 -6.93510532e-01 -7.99076080e-01
2.14222401e-01 -3.18993419e-01 -1.41270864e+00 1.04539466e+00
1.04813373e+00 1.28399059e-01 4.74748582e-01 9.52023566e-01
-3.08694482e-01 7.99280047e-01 -3.26274693e-01 -3.50962073e-01
2.47137189e-01 -8.90530348e-01 -5.89578211e-01 1.47726148e-01
7.47624636e-01 -8.13437402e-01 -3.66045445e-01 5.75939596e-01
3.91838074e-01 7.87351370e-01 9.05261278e-01 -8.24834764e-01
-1.76124632e+00 1.62969911e+00 1.82623461e-01 -5.19587696e-01
8.82367015e-01 4.96022671e-01 1.06626928e+00 -5.17724037e-01
4.67028201e-01 1.18409741e+00 6.85034990e-01 1.06773937e+00
-1.68475449e+00 -3.95133108e-01 5.52820563e-01 -4.19821411e-01
-8.53834033e-01 -4.18255776e-01 6.05208874e-01 -3.94751430e-01
1.05995142e+00 2.65953153e-01 6.70112967e-01 1.51906538e+00
-4.12869811e-01 9.07240868e-01 5.99933922e-01 -2.38084704e-01
9.65920687e-02 -8.66509303e-02 1.95871890e-01 4.03929651e-01
2.33368129e-01 2.27809235e-01 -3.54375601e-01 -1.11174181e-01
1.04193163e+00 -2.09169716e-01 -2.76229441e-01 -2.08361626e-01
-1.73677826e+00 1.15670609e+00 4.99529868e-01 1.28933579e-01
-8.30819190e-01 3.10247779e-01 1.63246676e-01 9.04604793e-02
3.88296068e-01 1.36426163e+00 -9.67086554e-01 4.53081369e-01
-5.68721235e-01 1.07355177e+00 9.60548759e-01 1.29827845e+00
3.84682834e-01 3.14767361e-02 -1.10009980e+00 9.03275192e-01
2.34144121e-01 5.70704103e-01 3.20077002e-01 -8.42807472e-01
4.92509693e-01 3.32321912e-01 9.09048438e-01 -8.86637270e-01
-5.09190977e-01 2.23582998e-01 -7.93512523e-01 -1.89784482e-01
5.34156084e-01 -5.88300705e-01 -1.00405169e+00 1.97336817e+00
1.39899343e-01 9.67167839e-02 2.53157258e-01 8.91347587e-01
1.27441120e+00 1.00671935e+00 7.28508234e-01 3.09732437e-01
1.34143448e+00 -1.30864108e+00 -6.60197437e-01 -1.56382695e-01
4.36191678e-01 -7.89216697e-01 1.26604509e+00 2.41238415e-01
-9.54855084e-01 -9.06225979e-01 -6.61732435e-01 -5.57002783e-01
-5.82732201e-01 6.53303742e-01 1.24630761e+00 5.36664605e-01
-9.94743288e-01 1.01133847e+00 -2.01231703e-01 -3.26142758e-01
5.44255935e-02 2.35479951e-01 5.52212223e-02 -2.70566344e-01
-1.25440645e+00 6.37097538e-01 8.02217841e-01 -9.39189084e-03
-8.26585531e-01 -1.22234654e+00 -1.34177804e+00 1.05149793e-02
2.72584617e-01 -1.38855171e+00 1.53674316e+00 -9.14304376e-01
-1.83849692e+00 3.32936376e-01 3.40355277e-01 3.10067739e-02
-9.07384455e-02 -4.50704873e-01 -5.97841382e-01 -4.91212159e-01
1.64124385e-01 1.24014235e+00 4.60799307e-01 -1.13057852e+00
-5.24377704e-01 1.60085261e-01 5.40701210e-01 6.39535844e-01
3.21046561e-01 -2.94002920e-01 -2.11923812e-02 -1.11264384e+00
-2.28405297e-01 -7.85377741e-01 -8.97451520e-01 -3.93938571e-01
-5.96202254e-01 -4.50381339e-01 -3.24785441e-01 -9.26346242e-01
1.07557893e+00 -1.70502198e+00 3.22968155e-01 -2.82193482e-01
-2.72944629e-01 -9.83733833e-02 -9.08743262e-01 5.99151731e-01
-2.38145277e-01 1.57776475e-01 2.20608577e-01 -3.26852113e-01
5.74902713e-01 -1.97498634e-01 -1.84968770e-01 -1.93130091e-01
1.30523250e-01 1.10339165e+00 -1.42546678e+00 1.26208827e-01
6.39492095e-01 5.74921608e-01 -1.20806229e+00 6.56418502e-01
-1.37013781e+00 5.44741333e-01 -7.41425931e-01 1.52712181e-01
2.42724389e-01 -1.89779401e-01 3.39038670e-01 -6.38919830e-01
-9.48970690e-02 7.96679080e-01 -1.12509167e+00 2.51736093e+00
-4.87896591e-01 -2.81181455e-01 -1.83509320e-01 -1.30104333e-01
6.81466103e-01 5.93657076e-01 2.55461752e-01 -2.30685890e-01
-5.29692471e-02 -2.36736506e-01 -6.08487725e-01 -4.49818760e-01
1.13968349e+00 -4.03971374e-02 -6.07030034e-01 7.83435643e-01
4.42099199e-02 -4.36817259e-01 5.16548097e-01 3.97576869e-01
5.66734076e-01 9.78112936e-01 3.08990508e-01 -4.05088753e-01
1.13987900e-01 1.44341022e-01 -1.00528851e-01 8.65352154e-01
3.48156333e-01 6.38505280e-01 -2.15992123e-01 -6.16865993e-01
-1.18617201e+00 -9.84482586e-01 7.35513210e-01 1.75586116e+00
-3.37726504e-01 -6.29704654e-01 -6.22997582e-01 -5.87678790e-01
-7.69874379e-02 1.56054366e+00 -8.50251973e-01 -1.20873764e-01
-7.38505006e-01 -7.70537674e-01 6.03793897e-02 4.36099142e-01
1.56910837e-01 -1.47406876e+00 -6.36550263e-02 6.48627996e-01
-7.94566751e-01 -5.25438190e-01 -1.45575118e+00 -1.76316962e-01
-5.54674685e-01 -8.96646559e-01 -6.93387091e-01 -8.08536470e-01
6.17143512e-01 2.19073474e-01 1.97928905e+00 2.41277437e-03
-6.27930388e-02 4.02526885e-01 -5.80512226e-01 -1.15905181e-02
-9.46424901e-01 1.53704613e-01 -1.96965277e-01 -2.59356797e-01
4.09089059e-01 -4.05300319e-01 -9.27196383e-01 -2.31541395e-01
-7.19093263e-01 9.22297359e-01 1.87368587e-01 5.18572211e-01
2.46644884e-01 -2.85608172e-01 5.40177941e-01 -8.10384393e-01
8.30222309e-01 -8.33787978e-01 -2.78785825e-01 3.60033393e-01
-2.55796373e-01 4.10520643e-01 8.57752323e-01 -6.96411371e-01
-1.39229727e+00 6.66387439e-01 -1.26463264e-01 3.51602137e-01
-6.73577845e-01 1.14890441e-01 -3.72282803e-01 8.76503229e-01
9.63509500e-01 -8.23088083e-03 -6.51795268e-01 -8.35183859e-01
1.64777493e+00 8.58068392e-02 9.47396219e-01 -1.15306211e+00
4.76914853e-01 -5.24530888e-01 -6.93786621e-01 -2.12781236e-01
-9.48009908e-01 -2.33488068e-01 -3.84526670e-01 2.14766875e-01
1.27618766e+00 -1.05266118e+00 -1.04082787e+00 3.13904405e-01
-1.22772372e+00 -8.77569556e-01 -3.94329578e-01 4.97863650e-01
-6.69609427e-01 2.19845586e-02 -8.98919404e-01 -5.66608965e-01
-5.04086554e-01 -7.17933059e-01 1.10814679e+00 4.66266125e-01
-6.54273033e-01 -8.98692012e-01 3.35103005e-01 4.13247794e-01
4.22825783e-01 4.59379613e-01 1.04680586e+00 -5.12007773e-01
-6.22739911e-01 1.91221580e-01 -1.72254503e-01 -1.31938860e-01
5.31518698e-01 -2.46207550e-01 -6.73876941e-01 1.07994251e-01
-6.82248652e-01 -4.85223740e-01 7.44805932e-01 7.20008075e-01
7.53434122e-01 -7.11401045e-01 -1.83108971e-01 2.91980833e-01
1.42861283e+00 -6.73370063e-02 4.66965258e-01 -3.81339043e-01
1.03282726e+00 6.89800978e-01 2.31289208e-01 5.13692617e-01
6.87414408e-01 6.92637205e-01 1.23271957e-01 -1.31743655e-01
-2.55246401e-01 -1.05005133e+00 4.53592271e-01 2.52064735e-01
-4.84331027e-02 -6.75065458e-01 7.95105025e-02 6.28352821e-01
-1.71867144e+00 -1.26399302e+00 -3.48116964e-01 2.10200381e+00
1.35260403e+00 -6.63820565e-01 3.15568954e-01 -6.63262665e-01
5.18517792e-01 1.44373238e-01 -5.84870338e-01 -3.37730795e-01
1.35609791e-01 1.80661902e-01 4.30846483e-01 6.90847337e-01
-1.45409369e+00 1.28523612e+00 6.95619822e+00 3.11162442e-01
-2.22625971e-01 -2.85663217e-01 7.44664550e-01 -1.26670271e-01
-9.39084888e-01 -4.25949812e-01 -7.60982335e-01 3.26598287e-01
9.51355398e-01 1.31112605e-01 1.36925030e+00 5.88757098e-01
3.80554318e-01 2.69035876e-01 -1.64975905e+00 6.39861822e-01
7.40508363e-02 -1.53269994e+00 3.18305045e-01 -4.01122689e-01
1.08738458e+00 -2.56402880e-01 -7.34695122e-02 8.64118457e-01
1.78984046e+00 -1.40167320e+00 6.94616675e-01 5.82640350e-01
5.81590176e-01 -5.14240801e-01 -4.60048132e-02 1.03770800e-01
-1.31044853e+00 1.34483233e-01 -2.89685011e-01 -2.97686756e-01
5.26548505e-01 1.99600846e-01 -8.41358662e-01 5.75851738e-01
2.75318474e-01 8.36423635e-01 -4.04774278e-01 4.54332680e-01
-2.63127863e-01 1.93923816e-01 -2.72721261e-01 -2.31275684e-03
3.29712301e-01 -4.42600369e-01 -4.18272428e-02 1.22734404e+00
5.06981075e-01 7.08985567e-01 4.33544397e-01 1.45907462e+00
-1.40401378e-01 4.72927392e-01 -4.91198927e-01 -6.02286935e-01
1.08399272e-01 1.15576732e+00 -4.67541516e-01 -6.12698615e-01
-3.11034888e-01 1.65129244e+00 1.90821871e-01 5.29396236e-01
-6.80793524e-01 3.51952016e-01 1.00200629e+00 -4.67702150e-01
3.36105227e-01 1.12615943e-01 -2.49587059e-01 -1.39363384e+00
-8.19648504e-01 -1.23202395e+00 5.50020456e-01 -1.15594590e+00
-1.67161536e+00 3.13003302e-01 6.84907893e-03 -4.06559944e-01
-3.56702894e-01 -4.09486264e-01 -6.25073135e-01 1.15025151e+00
-7.47135103e-01 -1.63799143e+00 -3.91448513e-02 4.07754153e-01
9.69188333e-01 2.22755164e-01 1.44796813e+00 1.77490607e-01
-1.01989061e-01 3.15834314e-01 -1.07418075e-01 8.52029771e-02
7.53111780e-01 -1.72318995e+00 1.29482436e+00 5.19641280e-01
3.27944577e-01 7.44785607e-01 1.09997559e+00 -1.07690763e+00
-1.15654755e+00 -1.30198944e+00 1.15996742e+00 -1.09487021e+00
-6.72792122e-02 -4.06874955e-01 -4.76672828e-01 8.94616723e-01
5.15179455e-01 -8.67301881e-01 9.21180785e-01 3.11356217e-01
-6.20501637e-01 5.73259592e-01 -1.13142550e+00 1.15047967e+00
1.18834329e+00 -1.92519322e-01 -5.94471633e-01 6.99147582e-01
1.22557855e+00 -4.30898428e-01 -9.74751055e-01 -1.02136768e-01
6.90791190e-01 -4.73106802e-01 1.52490580e+00 -9.42974091e-01
6.06348932e-01 -3.28970850e-01 -1.86820239e-01 -1.91118526e+00
-9.33146656e-01 -7.03204691e-01 3.36180210e-01 1.03878462e+00
5.73587596e-01 4.48480844e-01 5.99639773e-01 1.22137451e+00
-4.26112413e-01 8.55169892e-02 4.49548990e-01 1.00011028e-01
1.64924487e-01 -1.16515316e-01 1.39127862e+00 1.13283896e+00
7.69490600e-02 8.20693731e-01 -1.08926845e+00 9.37940404e-02
3.95551354e-01 2.59451598e-01 8.01314414e-01 -8.01463962e-01
-7.59503961e-01 -6.55383646e-01 1.11439896e+00 -1.28051496e+00
-6.57390207e-02 -1.02226925e+00 5.60001194e-01 -2.06783271e+00
2.24620685e-01 -2.15496823e-01 -6.74401671e-02 5.14733076e-01
-6.07582569e-01 1.01380035e-01 2.46819749e-01 -6.45928025e-01
-5.32172024e-01 4.41047817e-01 1.53006434e+00 -2.65581399e-01
-7.67740011e-01 -1.26275674e-01 -1.22951055e+00 2.00547576e-01
7.99108624e-01 -1.42340019e-01 -8.37581217e-01 -4.84742165e-01
4.72252965e-01 6.30178899e-02 1.41369283e-01 -5.91772079e-01
-3.19659203e-01 -6.38212919e-01 5.58089495e-01 -5.09046257e-01
1.24302283e-01 -3.79708230e-01 7.09621966e-01 8.29183683e-02
-1.12456822e+00 -2.08654210e-01 2.99959779e-01 5.80897033e-01
7.84433126e-01 -5.39304055e-02 1.57157853e-01 -7.42629170e-01
-6.68500900e-01 3.72621506e-01 -5.40419698e-01 -1.57761976e-01
3.54293883e-01 3.53543371e-01 -1.89097002e-01 -5.61340332e-01
-9.52690303e-01 2.05406800e-01 5.12203753e-01 9.81437862e-01
1.37970790e-01 -1.79423380e+00 -1.18550253e+00 4.86280546e-02
2.98487633e-01 -2.58878976e-01 6.39407158e-01 -2.33531877e-01
-4.76952285e-01 4.40999627e-01 -9.85276029e-02 4.02650267e-01
-6.88262641e-01 1.24626207e+00 2.41669655e-01 -5.20183206e-01
-7.32713580e-01 8.16425681e-01 5.17849982e-01 -1.05310619e+00
1.36777654e-01 -9.33982790e-01 -4.32877481e-01 -3.71660411e-01
9.52869773e-01 5.09952335e-03 -7.04481542e-01 -3.30628306e-01
3.80499870e-01 1.46429017e-01 7.34436437e-02 3.21948156e-02
1.10358608e+00 -1.93409532e-01 2.02527285e-01 1.86518878e-01
5.87072730e-01 -2.08698854e-01 -1.49706459e+00 -2.01837793e-02
-2.14408234e-01 -2.63553858e-01 -3.27790648e-01 -1.61976683e+00
-7.13222563e-01 2.21405447e-01 1.79906800e-01 -1.69850856e-01
8.17019939e-01 -2.01430306e-01 1.10532689e+00 4.03423727e-01
2.10671410e-01 -8.05997729e-01 1.02290139e-01 1.79132119e-01
1.15675151e+00 -1.32043755e+00 -2.44130641e-01 -2.77743042e-01
-9.22224581e-01 8.14582348e-01 6.94183469e-01 -1.31573007e-01
3.60010922e-01 -2.72513300e-01 5.38638905e-02 5.84058687e-02
-7.84353554e-01 -1.22499056e-01 7.59299040e-01 9.65735137e-01
1.06435156e+00 6.53764069e-01 -1.38348192e-01 1.14217687e+00
-3.51545364e-01 4.67576772e-01 3.55338305e-01 1.11084253e-01
-2.71120340e-01 -1.30681431e+00 -9.77017432e-02 3.97198856e-01
1.74284149e-02 -7.72925794e-01 -2.24356666e-01 2.50149548e-01
4.26787823e-01 1.24584591e+00 -1.91700086e-01 -2.77014792e-01
3.79341096e-01 3.17141980e-01 6.23261809e-01 -1.09167790e+00
-1.20823097e+00 6.55218214e-02 6.25828683e-01 -6.59986258e-01
-3.69357586e-01 -5.16929984e-01 -1.16502309e+00 -6.58122659e-01
3.05234343e-01 9.30914059e-02 3.16337913e-01 5.58409393e-01
2.83817679e-01 5.98348081e-01 2.73109138e-01 -1.22829461e+00
-4.04857278e-01 -1.12489116e+00 -4.62972462e-01 1.13734114e+00
5.54241426e-02 2.82138437e-02 1.98719725e-01 6.48390412e-01] | [11.518588066101074, 4.550482273101807] |
991b1c90-d027-4a72-bfff-79c3c29a3946 | reveal-to-revise-an-explainable-ai-life-cycle | 2303.12641 | null | https://arxiv.org/abs/2303.12641v2 | https://arxiv.org/pdf/2303.12641v2.pdf | Reveal to Revise: An Explainable AI Life Cycle for Iterative Bias Correction of Deep Models | State-of-the-art machine learning models often learn spurious correlations embedded in the training data. This poses risks when deploying these models for high-stake decision-making, such as in medical applications like skin cancer detection. To tackle this problem, we propose Reveal to Revise (R2R), a framework entailing the entire eXplainable Artificial Intelligence (XAI) life cycle, enabling practitioners to iteratively identify, mitigate, and (re-)evaluate spurious model behavior with a minimal amount of human interaction. In the first step (1), R2R reveals model weaknesses by finding outliers in attributions or through inspection of latent concepts learned by the model. Secondly (2), the responsible artifacts are detected and spatially localized in the input data, which is then leveraged to (3) revise the model behavior. Concretely, we apply the methods of RRR, CDEP and ClArC for model correction, and (4) (re-)evaluate the model's performance and remaining sensitivity towards the artifact. Using two medical benchmark datasets for Melanoma detection and bone age estimation, we apply our R2R framework to VGG, ResNet and EfficientNet architectures and thereby reveal and correct real dataset-intrinsic artifacts, as well as synthetic variants in a controlled setting. Completing the XAI life cycle, we demonstrate multiple R2R iterations to mitigate different biases. Code is available on https://github.com/maxdreyer/Reveal2Revise. | ['Sebastian Lapuschkin', 'Wojciech Samek', 'Maximilian Dreyer', 'Frederik Pahde'] | 2023-03-22 | null | null | null | null | ['age-estimation', 'age-estimation'] | ['computer-vision', 'miscellaneous'] | [ 5.73499501e-01 6.46604180e-01 -7.62471706e-02 -1.71998754e-01
-7.45499492e-01 -4.55502391e-01 6.41862035e-01 1.52142614e-01
-2.50905097e-01 6.97433531e-01 1.34115756e-01 -5.82632780e-01
-4.59460586e-01 -4.43295956e-01 -8.44978273e-01 -4.79950905e-01
1.37063771e-01 2.34700173e-01 -1.45679563e-01 3.24159175e-01
1.88045740e-01 2.22804502e-01 -1.06448722e+00 3.91279131e-01
9.01246130e-01 7.05625951e-01 -3.01117897e-01 5.37811100e-01
4.22910094e-01 1.07344985e+00 -6.20791376e-01 -6.20923281e-01
1.01860864e-02 -2.61145562e-01 -5.83103061e-01 2.42729392e-02
2.44763240e-01 -1.92284048e-01 -1.12695545e-01 1.12049377e+00
3.87001425e-01 -3.28531384e-01 6.74394488e-01 -1.36799467e+00
-7.07791209e-01 7.80172288e-01 -6.69053435e-01 2.30379537e-01
1.17408726e-02 6.14339590e-01 8.54579031e-01 -9.53335106e-01
5.83765745e-01 1.05235076e+00 1.01657760e+00 7.27009177e-01
-1.30900896e+00 -7.53566444e-01 1.79948315e-01 -1.50244877e-01
-1.37446606e+00 -4.07387882e-01 5.92052579e-01 -5.31032801e-01
7.02286661e-01 5.28590202e-01 4.04062569e-01 1.74398267e+00
1.42057359e-01 5.93604982e-01 1.12867296e+00 -2.57537484e-01
3.14903349e-01 1.68204740e-01 4.70210552e-01 7.20315993e-01
4.70442265e-01 2.88197517e-01 -6.11209452e-01 -4.09833133e-01
5.33639848e-01 6.13270923e-02 -1.64209872e-01 3.21241617e-01
-1.04844558e+00 4.97867256e-01 4.34714377e-01 5.23622904e-04
-5.63617885e-01 4.81825471e-01 1.01341084e-01 9.76514518e-02
7.54885912e-01 8.09650123e-01 -6.50434613e-01 3.74876931e-02
-1.02935839e+00 8.39607120e-02 2.44935721e-01 6.65999413e-01
4.78242069e-01 -4.77178730e-02 -2.62493014e-01 8.13778758e-01
3.55542094e-01 9.51565504e-02 5.44700265e-01 -6.72863126e-01
1.96631491e-01 8.50097775e-01 -1.00960389e-01 -9.11030471e-01
-5.99807858e-01 -8.01301718e-01 -1.00331676e+00 1.77516863e-01
4.89681333e-01 7.84142315e-02 -1.20480704e+00 1.67591345e+00
3.12503576e-01 5.89520216e-01 -9.97259617e-02 6.98581338e-01
7.82675743e-01 -1.03478692e-01 3.33486110e-01 2.86415011e-01
1.51120019e+00 -7.92172968e-01 -4.19984460e-01 -4.56440151e-01
7.96537757e-01 -2.33957380e-01 1.24541712e+00 6.10791445e-01
-8.75193655e-01 -2.33850360e-01 -9.61847663e-01 -2.94342972e-02
-2.08332345e-01 2.92358160e-01 7.06506252e-01 5.51882207e-01
-7.91583717e-01 9.30076659e-01 -1.03864181e+00 -1.93631858e-01
8.47946465e-01 3.85195613e-01 -3.97002935e-01 4.41156663e-02
-1.08185375e+00 6.68507576e-01 6.29290864e-02 4.19760913e-01
-1.18091834e+00 -1.33075583e+00 -5.80170035e-01 -1.03310049e-01
5.02004087e-01 -9.54669535e-01 8.50548267e-01 -1.16724432e+00
-7.92608380e-01 9.78515744e-01 1.92876235e-02 -5.08371413e-01
9.87186432e-01 -4.10450518e-01 -2.06060857e-01 -3.55665386e-01
2.97210161e-02 4.00751054e-01 7.67249942e-01 -1.27618182e+00
-3.05279851e-01 -4.79910165e-01 -2.65241951e-01 -3.52637470e-01
-8.59174430e-02 1.15420036e-01 -3.33147794e-01 -8.63883793e-01
1.65486723e-01 -1.01842558e+00 -4.21741664e-01 -1.79079641e-02
-1.10656583e+00 5.13410270e-02 1.92094520e-01 -9.53664958e-01
1.18450022e+00 -2.10000777e+00 1.65300816e-02 5.17422557e-01
7.60366142e-01 5.48259020e-02 -1.32205695e-01 -1.46002799e-01
-4.89718258e-01 5.96644700e-01 -2.99631119e-01 -8.42671692e-01
-3.24350387e-01 9.36617479e-02 -2.57895052e-01 4.55458105e-01
4.29022491e-01 1.11006808e+00 -8.00154328e-01 -1.23743571e-01
-1.08331814e-01 4.44486022e-01 -5.27276218e-01 -1.82674050e-01
-1.99159160e-01 5.70932150e-01 -3.42101544e-01 9.98040676e-01
3.61694753e-01 -6.12286508e-01 1.61731765e-01 -2.79360145e-01
3.27458411e-01 3.76234055e-01 -8.65117073e-01 1.28919101e+00
-3.24126512e-01 6.54502630e-01 -3.54371160e-01 -5.04270375e-01
8.05038154e-01 9.58799347e-02 3.39072555e-01 -5.12742162e-01
2.29170155e-02 1.41610518e-01 -4.20346521e-02 -4.90860969e-01
2.95169830e-01 1.04457796e-01 1.41455799e-01 4.67850327e-01
-1.78063393e-01 4.73560601e-01 -5.02736866e-01 2.13327244e-01
1.52347302e+00 8.84548053e-02 4.30216014e-01 3.45947221e-03
1.44704431e-02 -2.57083047e-02 7.34804153e-01 1.08902061e+00
-9.29496735e-02 8.09127629e-01 6.41056836e-01 -5.44982255e-01
-9.16627944e-01 -1.07488430e+00 -1.07764825e-01 6.61396265e-01
-3.74171674e-01 -1.81981221e-01 -7.78910220e-01 -9.92997885e-01
3.01570088e-01 1.05576015e+00 -1.26833355e+00 -4.49144900e-01
-2.36345947e-01 -9.81644094e-01 9.77417409e-01 6.42236173e-01
9.54105929e-02 -1.08025229e+00 -6.61275387e-01 -3.66629176e-02
-1.19374670e-01 -7.07905889e-01 -4.43323553e-02 2.24013835e-01
-7.80567110e-01 -1.23175991e+00 -2.04246745e-01 2.04970136e-01
1.13789272e+00 -9.57957283e-02 1.19088626e+00 6.83233023e-01
-5.52583396e-01 1.74585402e-01 -2.34062016e-01 -6.15098774e-01
-5.82387745e-01 1.17004149e-01 5.87722613e-03 -1.14720710e-01
4.75461364e-01 -5.56789100e-01 -5.85406363e-01 3.28645766e-01
-9.07292008e-01 2.32260317e-01 5.64561129e-01 9.66310382e-01
6.93986177e-01 -1.17412969e-01 5.58470786e-01 -1.38085365e+00
6.18707180e-01 -7.29905188e-01 -4.56748813e-01 2.80737340e-01
-1.05261540e+00 -7.09724519e-03 3.71500492e-01 -6.27121031e-01
-7.43536353e-01 -2.73000985e-01 1.21589415e-01 -5.92558920e-01
-2.01387271e-01 7.16540813e-01 -1.68942511e-01 2.93569803e-01
1.07503021e+00 -8.41548741e-02 -5.69827147e-02 -4.41269070e-01
2.26155728e-01 5.74444056e-01 6.74718201e-01 -3.33296984e-01
8.68943334e-01 4.93412405e-01 -4.60689217e-02 -3.70141000e-01
-1.10271156e+00 -3.62583622e-02 -5.74017525e-01 -2.20682338e-01
5.11713445e-01 -8.99489582e-01 -4.41865265e-01 4.60135669e-01
-1.03858840e+00 -6.97133899e-01 -3.48158300e-01 4.00212348e-01
-1.95157439e-01 -2.27887183e-02 -2.63013601e-01 -8.09866428e-01
-4.15002823e-01 -1.03100336e+00 9.91795778e-01 1.14954628e-01
-7.60136843e-01 -1.06605995e+00 -7.34159425e-02 5.84735036e-01
1.64651155e-01 4.21043009e-01 1.05341792e+00 -1.00618052e+00
-4.85523194e-01 -2.95550615e-01 -2.63319969e-01 2.31699750e-01
1.99732780e-01 6.08963072e-02 -1.48641694e+00 -1.24215655e-01
-3.00219804e-01 -1.37276232e-01 1.02778280e+00 4.59642351e-01
1.58718836e+00 -5.29237568e-01 -4.67925936e-01 8.14072371e-01
9.67366934e-01 -2.28790537e-01 9.36404943e-01 5.00787556e-01
7.91512787e-01 6.47433102e-01 3.87559086e-01 3.58708322e-01
2.64786601e-01 4.37692881e-01 6.14004314e-01 -3.99220645e-01
-1.92444474e-01 -5.88348508e-01 8.75737220e-02 1.58083260e-01
-1.18261352e-01 -2.04281658e-02 -1.03810024e+00 6.61395609e-01
-1.92334521e+00 -7.80737579e-01 -2.31205031e-01 2.21628976e+00
7.54780650e-01 2.60720223e-01 -2.07989141e-01 8.22226554e-02
3.85992825e-01 -1.41834825e-01 -9.49138284e-01 5.47063611e-02
-8.54715779e-02 -3.77299376e-02 6.67126298e-01 1.85895905e-01
-6.85663283e-01 8.57205570e-01 5.66367483e+00 5.71693838e-01
-9.91431117e-01 2.90087044e-01 1.12995803e+00 -2.50427246e-01
-7.75622725e-01 1.14433289e-01 -3.77096057e-01 4.52121168e-01
8.40971589e-01 2.20037997e-01 5.50586998e-01 8.40083122e-01
3.69335175e-01 2.49078888e-02 -1.30755496e+00 7.23167896e-01
-6.61529228e-02 -1.35006189e+00 -2.10775435e-01 2.61148661e-01
4.94568855e-01 5.96059673e-03 2.95719504e-01 4.68543172e-02
4.38799083e-01 -1.46613860e+00 7.50209212e-01 7.58585095e-01
8.39936495e-01 -3.92887741e-01 6.60577357e-01 1.28663406e-01
-4.94256020e-01 -2.49054134e-01 -1.81620896e-01 2.33135566e-01
-2.95588970e-01 8.92680645e-01 -1.14008582e+00 5.31322777e-01
7.46100307e-01 6.78328276e-01 -1.04654515e+00 7.39564478e-01
-6.12983763e-01 1.03533411e+00 -1.83166981e-01 2.91020513e-01
-2.22358882e-01 1.74047276e-01 6.02160156e-01 9.42558944e-01
2.05625147e-01 -1.91435367e-01 -5.57957292e-01 1.36501181e+00
-2.82435417e-01 -4.92905051e-01 -3.72078985e-01 1.05616510e-01
6.38552725e-01 1.26069915e+00 -6.65702045e-01 -5.54437377e-02
-1.65762782e-01 9.00346041e-01 3.22003990e-01 4.25167352e-01
-9.39798594e-01 1.31977811e-01 6.65781319e-01 3.38463247e-01
-2.74381995e-01 3.77372831e-01 -7.96169817e-01 -1.02966344e+00
1.25732183e-01 -1.09613478e+00 4.61278975e-01 -1.07740104e+00
-1.38496947e+00 4.18609411e-01 -1.71541229e-01 -9.28639531e-01
-5.14538586e-02 -6.37334466e-01 -7.48361051e-01 7.51898468e-01
-1.52999020e+00 -1.37923038e+00 -6.60055101e-01 3.16952020e-01
1.92895755e-01 -9.92862061e-02 6.96152449e-01 7.45399818e-02
-9.19618487e-01 9.96145129e-01 -1.67627513e-01 1.36394322e-01
6.75336123e-01 -1.26121962e+00 7.32616127e-01 9.23411071e-01
1.32232621e-01 9.28152621e-01 6.40709639e-01 -1.05244160e+00
-8.17160785e-01 -1.42550325e+00 7.22224295e-01 -8.71341467e-01
7.11083412e-01 -2.75294125e-01 -1.07317805e+00 9.34035361e-01
-4.48590785e-01 9.37822163e-02 7.83069074e-01 3.70887071e-01
-5.62648773e-01 5.07751368e-02 -1.24405003e+00 6.70213878e-01
1.14812338e+00 -5.01404524e-01 -2.57733971e-01 3.52093160e-01
7.16326296e-01 -2.23535031e-01 -7.56551087e-01 4.49519336e-01
6.68143690e-01 -7.91765392e-01 8.66191983e-01 -1.07243240e+00
6.41402602e-01 -2.62273848e-01 -1.38353743e-02 -1.18897009e+00
-2.11181685e-01 -7.23917246e-01 3.81119251e-02 1.21776688e+00
9.82892513e-01 -5.94524324e-01 8.38507056e-01 1.02357471e+00
-9.09467936e-02 -8.83201122e-01 -9.25867140e-01 -4.17053729e-01
-4.70280163e-02 -8.05545092e-01 7.14163661e-01 1.14285612e+00
-3.95949721e-01 -1.72919482e-01 -5.29914677e-01 6.29313231e-01
7.35553682e-01 -4.34446454e-01 9.44744647e-01 -1.29526985e+00
-4.37671900e-01 -1.86066061e-01 -2.72919059e-01 -3.05471271e-01
-1.19882580e-02 -8.49412799e-01 -2.00049549e-01 -1.26788604e+00
4.81114298e-01 -5.94928026e-01 -5.87910771e-01 1.04407012e+00
-6.41526222e-01 1.66026577e-01 3.42448503e-02 6.07276440e-01
-2.84345180e-01 1.81350142e-01 6.24714792e-01 -1.35058790e-01
-1.27248675e-01 5.01222946e-02 -1.11899698e+00 8.10647011e-01
6.77266657e-01 -8.51178527e-01 -2.47777596e-01 -4.21280950e-01
5.59845984e-01 -4.13327873e-01 1.24054050e+00 -6.40813053e-01
8.37477371e-02 -4.93635945e-02 6.60551667e-01 -3.00948560e-01
-3.89820747e-02 -6.52179718e-01 4.30273026e-01 4.38537389e-01
-6.25967443e-01 1.49741560e-01 1.92683697e-01 4.65516388e-01
3.07539582e-01 -2.45461851e-01 6.24413013e-01 -1.37619063e-01
-9.17800814e-02 9.41643342e-02 3.71230906e-03 -3.45816128e-02
7.23989904e-01 -1.85482666e-01 -6.82805955e-01 -7.17751980e-02
-8.17310691e-01 1.80452541e-01 4.68787402e-01 3.33559722e-01
6.90489411e-01 -7.99594104e-01 -7.21630514e-01 1.95004717e-01
3.11783314e-01 1.90810889e-01 3.53232443e-01 1.11524153e+00
-9.70454141e-02 3.97801735e-02 1.95356324e-01 -5.78641713e-01
-1.25266445e+00 3.40068400e-01 6.00498497e-01 -4.11657184e-01
-6.10768318e-01 9.22105551e-01 1.34591192e-01 -4.30309623e-01
1.18249521e-01 -3.69630754e-01 1.11876853e-01 -1.64167091e-01
5.03034115e-01 3.21531415e-01 1.71123967e-01 -3.46711241e-02
-5.40884912e-01 3.38708647e-02 -3.73076439e-01 7.58070573e-02
1.35284185e+00 5.69871068e-02 4.94487993e-02 1.49549782e-01
7.09815502e-01 -6.44812956e-02 -1.43999445e+00 -5.32188229e-02
2.33314291e-01 -3.93144786e-01 1.36487797e-01 -1.24840486e+00
-9.77027714e-01 5.08723736e-01 6.30568981e-01 -1.41544446e-01
1.05754840e+00 8.56542811e-02 7.77849331e-02 -2.48352289e-02
5.31520844e-02 -9.46882069e-01 1.36798516e-01 -1.47864640e-01
8.38513017e-01 -1.20427549e+00 3.83631468e-01 -7.92944133e-02
-7.88945735e-01 7.31151521e-01 5.14947414e-01 1.26934394e-01
4.75314617e-01 3.35355163e-01 1.45336568e-01 -5.02147496e-01
-1.02520108e+00 2.82352686e-01 3.02967042e-01 5.72208405e-01
1.76245794e-01 1.35460839e-01 2.33936265e-01 1.06644213e+00
-2.48482272e-01 1.82249159e-01 5.47544062e-01 4.65677112e-01
2.95378327e-01 -6.53865814e-01 -2.35587701e-01 7.28983760e-01
-5.38856268e-01 -3.37515563e-01 -6.65686548e-01 7.63800859e-01
1.67334840e-01 8.78082454e-01 -4.38256934e-02 -5.69443524e-01
3.96747500e-01 6.05033105e-03 -9.59752500e-02 -4.22094464e-01
-8.37851226e-01 -1.20872974e-01 2.15560108e-01 -7.86885977e-01
-1.43961892e-01 -6.29514217e-01 -8.84849787e-01 1.56758968e-02
-4.52256441e-01 -3.75843436e-01 7.71129072e-01 9.57914352e-01
6.94868505e-01 1.01667821e+00 3.59360427e-01 -2.44554564e-01
-7.51213551e-01 -1.13841796e+00 -2.26795524e-01 6.03318036e-01
3.09846222e-01 -6.28264010e-01 -6.35160863e-01 -2.78151967e-03] | [8.871697425842285, 5.309967994689941] |
0ae8758e-c51d-4151-81be-385c0feb8a1a | pik-fix-restoring-and-colorizing-old-photo | 2205.01902 | null | https://arxiv.org/abs/2205.01902v3 | https://arxiv.org/pdf/2205.01902v3.pdf | Pik-Fix: Restoring and Colorizing Old Photos | Restoring and inpainting the visual memories that are present, but often impaired, in old photos remains an intriguing but unsolved research topic. Decades-old photos often suffer from severe and commingled degradation such as cracks, defocus, and color-fading, which are difficult to treat individually and harder to repair when they interact. Deep learning presents a plausible avenue, but the lack of large-scale datasets of old photos makes addressing this restoration task very challenging. Here we present a novel reference-based end-to-end learning framework that is able to both repair and colorize old, degraded pictures. Our proposed framework consists of three modules: a restoration sub-network that conducts restoration from degradations, a similarity network that performs color histogram matching and color transfer, and a colorization subnet that learns to predict the chroma elements of images conditioned on chromatic reference signals. The overall system makes uses of color histogram priors from reference images, which greatly reduces the need for large-scale training data. We have also created a first-of-a-kind public dataset of real old photos that are paired with ground truth ''pristine'' photos that have been manually restored by PhotoShop experts. We conducted extensive experiments on this dataset and synthetic datasets, and found that our method significantly outperforms previous state-of-the-art models using both qualitative comparisons and quantitative measurements. The code is available at https://github.com/DerrickXuNu/Pik-Fix. | ['Hongkai Yu', 'Alan Bovik', 'Jiaqi Ma', 'Zibo Meng', 'Jinlong Li', 'Xiaoyu Dong', 'Yuanqi Du', 'Zhengzhong Tu', 'Runsheng Xu'] | 2022-05-04 | null | null | null | null | ['colorization'] | ['computer-vision'] | [ 2.35242888e-01 -2.50098974e-01 3.59316885e-01 -1.98952049e-01
-7.73999572e-01 -3.65307242e-01 4.26428705e-01 -2.77664900e-01
-2.87460625e-01 1.00026798e+00 2.77817994e-01 9.33091342e-02
2.93636739e-01 -6.31265104e-01 -1.05579555e+00 -9.44174707e-01
1.05437279e-01 -8.88437256e-02 3.17041695e-01 -1.75771937e-01
3.68094325e-01 4.45552915e-01 -1.92668796e+00 2.88898498e-01
1.09076369e+00 9.06257749e-01 3.52251798e-01 8.03810298e-01
2.52154082e-01 9.02448356e-01 -6.80600226e-01 -4.49137032e-01
3.03802848e-01 -3.80606622e-01 -5.02598107e-01 2.02596426e-01
7.17920125e-01 -5.66408515e-01 -8.28263700e-01 1.09842753e+00
8.07567537e-01 1.67774647e-01 4.85545814e-01 -1.22289896e+00
-1.21913278e+00 1.30751953e-02 -6.04677677e-01 1.94353953e-01
4.45221037e-01 5.57311594e-01 4.28996742e-01 -8.55243862e-01
6.29312992e-01 1.16620350e+00 5.96601546e-01 6.34825468e-01
-1.29943705e+00 -7.37703681e-01 -2.10632697e-01 6.66106522e-01
-9.76553082e-01 -6.29166961e-01 8.35657716e-01 -2.24384502e-01
6.14083648e-01 3.36709842e-02 8.29303563e-01 1.29712534e+00
4.28358585e-01 5.94022334e-01 1.56100810e+00 -3.44546884e-01
2.82339931e-01 -2.52040476e-01 -2.81557292e-01 4.39971209e-01
3.15041281e-02 3.27555895e-01 -3.98772329e-01 3.23833734e-01
7.29953945e-01 1.40216872e-01 -7.20274985e-01 -3.75161827e-01
-1.08984697e+00 1.80978954e-01 7.34253824e-01 6.11388795e-02
-2.95907885e-01 1.92956418e-01 -4.55587022e-02 4.40300852e-01
4.53108191e-01 2.12504268e-01 -1.68796882e-01 1.18290648e-01
-8.78510118e-01 -4.64548878e-02 4.48962867e-01 5.92529118e-01
1.02555108e+00 1.12055801e-01 -1.58922076e-02 1.04531336e+00
-4.05963548e-02 6.55252337e-01 4.17139292e-01 -1.38962042e+00
1.50394991e-01 2.33768761e-01 2.57504225e-01 -9.23347056e-01
-2.59816557e-01 -3.61801445e-01 -1.14876473e+00 6.87386453e-01
3.36680949e-01 5.34866117e-02 -1.29056823e+00 1.73942983e+00
1.73054576e-01 3.37673426e-01 1.22577004e-01 1.13166678e+00
6.77939773e-01 8.08846295e-01 -1.02795780e-01 -3.25169295e-01
9.92204547e-01 -1.08913362e+00 -6.63459063e-01 -5.32044232e-01
-2.00644389e-01 -1.02597868e+00 1.30023253e+00 5.13717234e-01
-1.27377856e+00 -6.96905375e-01 -1.19947219e+00 -4.49991941e-01
-2.96687871e-01 6.74745739e-02 4.26387697e-01 3.17725927e-01
-1.48437989e+00 7.54446328e-01 -4.77809191e-01 -5.41095316e-01
6.07489049e-01 7.31800124e-02 -6.09965265e-01 -6.76705301e-01
-9.64475453e-01 1.05937493e+00 8.52076411e-02 2.94672400e-01
-1.26697254e+00 -5.92300892e-01 -6.83951318e-01 -1.34985866e-02
1.79713801e-01 -7.71230519e-01 8.99583936e-01 -1.09776485e+00
-1.36888540e+00 1.04309416e+00 6.67658076e-02 -3.03127524e-02
4.71666187e-01 -2.79089630e-01 -6.53578639e-01 3.68624777e-01
4.93975431e-02 7.52578318e-01 1.13703978e+00 -1.75495625e+00
-2.94090509e-01 -3.13882679e-01 -7.83939753e-03 1.23256445e-01
-1.54476270e-01 -2.16500580e-01 -6.66528106e-01 -7.79670238e-01
3.45839746e-02 -7.10117042e-01 6.43859133e-02 4.63503599e-01
-4.04084295e-01 5.08959532e-01 6.55355930e-01 -1.08369637e+00
7.16148853e-01 -2.23377109e+00 2.83445895e-01 -2.34691665e-01
1.79275632e-01 2.78016001e-01 -4.86743629e-01 5.27796686e-01
-3.90818834e-01 -1.80143625e-01 -4.66758341e-01 -4.68531311e-01
-2.16189861e-01 7.65495896e-02 -2.94311643e-01 6.73907816e-01
-3.88153642e-02 6.90254211e-01 -8.84465158e-01 -3.10540646e-01
2.93740302e-01 7.27398217e-01 -2.16189340e-01 5.76831639e-01
-1.82362467e-01 4.38557476e-01 3.33071381e-01 9.78682339e-01
1.08193815e+00 4.81807552e-02 -8.82496238e-02 -5.95704079e-01
-1.28132412e-02 -3.70831400e-01 -1.04579771e+00 1.87283969e+00
-2.70265192e-01 6.93588972e-01 1.75973892e-01 -7.54640698e-01
7.17304409e-01 -8.92241746e-02 1.98985592e-01 -1.05995786e+00
5.18984906e-02 1.20077662e-01 -4.09601778e-01 -7.54186749e-01
4.31763053e-01 -1.18671302e-02 1.33682549e-01 3.94311845e-01
9.68121514e-02 -3.48393291e-01 2.88781881e-01 1.99276730e-01
1.25959992e+00 2.52385825e-01 -1.48812875e-01 1.27383754e-01
3.87928963e-01 -3.14386129e-01 6.77139997e-01 3.84240061e-01
-2.90297002e-01 1.16505861e+00 3.13586324e-01 -3.13803226e-01
-1.29571176e+00 -1.50949025e+00 2.08180547e-01 7.06263900e-01
4.41434652e-01 6.91528916e-02 -7.00337768e-01 -1.13846682e-01
-1.07898809e-01 5.41911840e-01 -6.10430002e-01 -3.60132456e-01
-3.11355233e-01 -6.63458169e-01 2.19263181e-01 3.61495107e-01
7.52425194e-01 -1.26882267e+00 -2.50662029e-01 -3.18208337e-02
-3.50971073e-01 -1.08794439e+00 -5.24806678e-01 2.14953963e-02
-6.41453445e-01 -1.31591296e+00 -9.80235159e-01 -8.71917546e-01
8.63374293e-01 6.32234991e-01 9.95256543e-01 2.61443406e-01
-5.61212897e-01 5.52703381e-01 -2.91658223e-01 2.43736920e-03
-3.99070740e-01 -4.63049144e-01 -6.66847751e-02 9.94585529e-02
-2.61992157e-01 -1.09285033e+00 -1.06954634e+00 2.04308391e-01
-1.30596423e+00 1.76871255e-01 9.42580163e-01 9.50145423e-01
5.02145588e-01 1.11211397e-01 4.21545982e-01 -4.97590959e-01
3.45116049e-01 -2.61997491e-01 -4.28599119e-01 3.86351496e-01
-5.82469940e-01 -1.79482758e-01 6.17164433e-01 -3.85298818e-01
-1.24918318e+00 5.68525083e-02 -1.00333251e-01 -5.36087275e-01
-2.89231926e-01 6.60194606e-02 -4.61706519e-01 -2.68797219e-01
5.68954051e-01 3.02577913e-01 1.33601248e-01 -5.33071399e-01
6.10644877e-01 5.17834187e-01 1.31243765e+00 -4.38458681e-01
1.06923401e+00 6.06352270e-01 -2.65693724e-01 -7.80670047e-01
-7.62393415e-01 -1.16514131e-01 -5.07998645e-01 -5.54350674e-01
7.99305439e-01 -1.03463566e+00 -5.90377688e-01 1.17416883e+00
-9.76819873e-01 -6.43899441e-01 -2.39295930e-01 1.20253742e-01
-6.12799406e-01 6.71202600e-01 -8.26499462e-01 -3.47460836e-01
-2.88264900e-01 -8.05378556e-01 8.10955226e-01 6.93997741e-01
4.05248880e-01 -5.83371639e-01 3.20338383e-02 6.26543045e-01
5.71743250e-01 3.76929522e-01 9.14617121e-01 5.26170909e-01
-8.93166065e-01 4.87124361e-02 -4.77660626e-01 8.25656712e-01
5.26007675e-02 3.78477983e-02 -1.26221240e+00 -5.43510377e-01
-1.40191868e-01 -4.99931723e-01 1.26144814e+00 2.46201470e-01
1.20371449e+00 -7.90907294e-02 9.52100940e-03 9.49410975e-01
1.63019717e+00 -5.36456741e-02 1.46172655e+00 4.63220209e-01
6.72530353e-01 4.74705130e-01 3.95764261e-01 2.29867786e-01
5.06894827e-01 3.52745175e-01 6.44925714e-01 -4.32669282e-01
-7.24429727e-01 -1.68522507e-01 5.30155778e-01 6.10346258e-01
-6.79536313e-02 -3.40283006e-01 -6.61637485e-01 6.26510441e-01
-1.61866605e+00 -1.01809704e+00 9.32111740e-02 2.07082486e+00
1.04823780e+00 -6.81037009e-02 -4.09803778e-01 1.95803538e-01
8.92156899e-01 1.96999729e-01 -7.52089977e-01 -7.13272095e-02
-5.92793465e-01 2.00446934e-01 2.76488394e-01 3.69025469e-01
-1.06784976e+00 8.82137299e-01 5.79402542e+00 6.14325345e-01
-1.16800976e+00 3.80608104e-02 7.97581136e-01 -3.63247097e-02
-1.48587227e-01 1.08157538e-01 -2.31740382e-02 5.80210865e-01
6.89168036e-01 8.96381065e-02 8.50969791e-01 3.68823379e-01
7.33802617e-02 -5.39737165e-01 -8.60463202e-01 1.17281115e+00
4.86160278e-01 -1.22919869e+00 -1.99421227e-01 -2.80151576e-01
9.64334965e-01 -2.67051496e-02 2.20951661e-01 5.98814413e-02
1.47215769e-01 -8.34523737e-01 7.99416184e-01 1.04862225e+00
1.05604899e+00 -5.36014676e-01 5.03934205e-01 -7.46978298e-02
-7.82885373e-01 -4.29644436e-01 -5.04518390e-01 7.61421770e-02
1.53987765e-01 8.54587674e-01 -1.40038669e-01 3.97506624e-01
1.12559867e+00 9.54962552e-01 -1.10988903e+00 1.43547416e+00
-6.06479347e-01 2.61668205e-01 2.28148386e-01 6.37756169e-01
-4.43845540e-01 -1.91504627e-01 3.01981211e-01 7.78697908e-01
5.05867004e-01 6.24432825e-02 -1.62680671e-01 7.71614313e-01
-2.38996610e-01 -3.50544453e-01 -2.91593790e-01 1.68749630e-01
4.05883729e-01 1.37105787e+00 -6.74639463e-01 -2.40626231e-01
-2.86786675e-01 1.44379282e+00 2.28736877e-01 7.32020557e-01
-7.20706284e-01 -2.97229528e-01 6.36523008e-01 9.37951133e-02
1.78785205e-01 -1.89615950e-01 -1.20151334e-01 -1.36447120e+00
1.27403796e-01 -9.52197850e-01 2.29005218e-01 -1.68755579e+00
-1.58889389e+00 6.00682616e-01 -2.94827312e-01 -1.24154854e+00
1.83800340e-01 -3.82838517e-01 -7.60468543e-01 6.26355290e-01
-1.74514508e+00 -1.26700628e+00 -9.68570113e-01 7.20510125e-01
2.56938368e-01 -1.26598775e-02 5.43433368e-01 5.55957615e-01
-6.88888609e-01 3.46262574e-01 3.47606182e-01 -1.14209950e-01
1.35196483e+00 -1.12947178e+00 3.07866812e-01 1.25016630e+00
-4.13717210e-01 1.39972880e-01 9.05897021e-01 -6.02223158e-01
-1.53168917e+00 -1.06496823e+00 3.01155299e-01 -6.67339657e-03
3.64360809e-01 -2.28377298e-01 -1.08900678e+00 3.33916277e-01
4.99338239e-01 1.56792060e-01 1.64473236e-01 -4.84880745e-01
-3.90144438e-01 -5.85274875e-01 -1.17505848e+00 5.68207622e-01
1.04390812e+00 -6.51293635e-01 -3.72466058e-01 2.19841957e-01
6.64280713e-01 -3.32397074e-01 -6.71420276e-01 1.66838750e-01
5.86629689e-01 -1.38874018e+00 1.29471791e+00 2.14611501e-01
6.81101143e-01 -5.56656599e-01 -1.11496359e-01 -1.66949260e+00
-1.66212052e-01 -4.72707808e-01 -1.85743980e-02 1.44285202e+00
-3.23606730e-02 -5.89934886e-01 4.64817941e-01 5.47281742e-01
-4.12533402e-01 -3.37543637e-01 -7.78190792e-01 -6.69686377e-01
-9.69301015e-02 -4.20677736e-02 3.93269360e-01 7.70971775e-01
-5.25633454e-01 1.34624198e-01 -6.21406019e-01 1.53368875e-01
9.81380463e-01 9.02312770e-02 7.70508945e-01 -9.57797229e-01
-1.23655222e-01 -2.04535216e-01 -4.20267403e-01 -5.82604349e-01
1.28103912e-01 -5.29138207e-01 2.54467875e-01 -1.90474415e+00
2.27559045e-01 -1.56433180e-01 -2.17044875e-01 5.40026188e-01
-1.54457316e-01 6.93538249e-01 1.15518644e-01 1.72688991e-01
-4.87640768e-01 9.30551052e-01 1.49702775e+00 -3.85843158e-01
7.95265660e-02 -4.96217340e-01 -6.68339193e-01 4.65486616e-01
7.73178935e-01 -3.08537990e-01 -2.45684117e-01 -5.27597427e-01
1.19966581e-01 9.35492888e-02 7.75510609e-01 -1.39671183e+00
3.16563815e-01 -6.21956140e-02 9.33300316e-01 -4.00515944e-01
6.42608762e-01 -6.22308910e-01 3.44605386e-01 3.11308861e-01
8.35773349e-03 1.95904523e-02 2.09049746e-01 5.20920813e-01
-1.30673885e-01 1.48940682e-01 1.24317110e+00 -1.34341791e-01
-1.00836146e+00 2.27047160e-01 -1.71717763e-01 5.15928566e-02
1.00756133e+00 -1.79577723e-01 -9.90124464e-01 -6.72351062e-01
-5.41279733e-01 4.91257310e-02 1.11990821e+00 3.59658480e-01
9.84928310e-01 -1.23049450e+00 -6.94169521e-01 2.31025323e-01
-4.79233377e-02 -2.76681960e-01 8.84820342e-01 5.96560955e-01
-8.11267853e-01 -4.43445802e-01 -8.50961089e-01 -3.45791578e-01
-1.06913078e+00 9.23319757e-01 3.57166111e-01 3.13172638e-01
-8.62823904e-01 7.19412625e-01 1.10046389e-02 -1.16758384e-01
3.49589050e-01 -1.21569522e-01 4.68797870e-02 6.01680810e-03
5.57643592e-01 4.72498447e-01 -4.00101542e-02 -6.10540986e-01
-7.51333386e-02 7.39714146e-01 1.72801465e-01 5.57403266e-02
1.67067885e+00 -5.15051365e-01 -4.35741931e-01 1.34864688e-01
1.03850436e+00 -2.87392497e-01 -1.71779156e+00 -2.04899609e-01
-4.91264910e-01 -8.04654181e-01 2.28189677e-02 -1.06826627e+00
-1.46878755e+00 9.92046952e-01 1.10133314e+00 -1.24633551e-01
1.81225824e+00 -1.45246878e-01 1.00531065e+00 1.52851596e-01
2.41250589e-01 -1.14116752e+00 5.56763291e-01 9.89896730e-02
1.29043734e+00 -1.18300724e+00 1.35931775e-01 -2.82982409e-01
-4.45160627e-01 1.05088282e+00 8.51523936e-01 -1.31206706e-01
4.23050821e-01 7.41111860e-02 3.66171449e-01 1.21231668e-01
-7.17532754e-01 -2.89510727e-01 4.52287309e-02 8.90184522e-01
-3.38406451e-02 -2.58141875e-01 1.94822792e-02 2.29655385e-01
-2.01384649e-02 1.45237818e-02 1.00657582e+00 8.24269950e-01
-3.80473644e-01 -1.05437565e+00 -6.92321420e-01 1.82165489e-01
-1.15078852e-01 -3.49024758e-02 -1.56936929e-01 4.85094786e-01
3.05783987e-01 9.63894486e-01 -1.87345177e-01 -5.50988436e-01
2.56595403e-01 -1.81689009e-01 6.56472445e-01 -1.85515478e-01
-9.98763293e-02 -5.78325540e-02 -2.35639066e-01 -7.73081183e-01
-5.51759541e-01 -4.52134788e-01 -9.71485853e-01 -5.01674831e-01
1.64109349e-01 -2.83848196e-01 6.29202306e-01 5.34674168e-01
4.04021919e-01 6.21730983e-01 6.99651599e-01 -1.41458619e+00
-1.74154952e-01 -8.67313266e-01 -7.32429504e-01 6.78766251e-01
5.72763503e-01 -7.44472682e-01 -5.81770360e-01 4.05867308e-01] | [11.143068313598633, -2.1214654445648193] |
ae93856b-f898-45b8-bb13-b9cdc0684f13 | predictive-experience-replay-for-continual | 2303.06572 | null | https://arxiv.org/abs/2303.06572v1 | https://arxiv.org/pdf/2303.06572v1.pdf | Predictive Experience Replay for Continual Visual Control and Forecasting | Learning physical dynamics in a series of non-stationary environments is a challenging but essential task for model-based reinforcement learning (MBRL) with visual inputs. It requires the agent to consistently adapt to novel tasks without forgetting previous knowledge. In this paper, we present a new continual learning approach for visual dynamics modeling and explore its efficacy in visual control and forecasting. The key assumption is that an ideal world model can provide a non-forgetting environment simulator, which enables the agent to optimize the policy in a multi-task learning manner based on the imagined trajectories from the world model. To this end, we first propose the mixture world model that learns task-specific dynamics priors with a mixture of Gaussians, and then introduce a new training strategy to overcome catastrophic forgetting, which we call predictive experience replay. Finally, we extend these methods to continual RL and further address the value estimation problems with the exploratory-conservative behavior learning approach. Our model remarkably outperforms the naive combinations of existing continual learning and visual RL algorithms on DeepMind Control and Meta-World benchmarks with continual visual control tasks. It is also shown to effectively alleviate the forgetting of spatiotemporal dynamics in video prediction datasets with evolving domains. | ['Xiaokang Yang', 'Yunbo Wang', 'Siyu Gao', 'Xiangming Zhu', 'Geng Chen', 'Wendong Zhang'] | 2023-03-12 | null | null | null | null | ['video-prediction'] | ['computer-vision'] | [-3.74182940e-01 -1.60040036e-01 -1.67572871e-01 2.50946492e-01
-2.57309675e-01 -3.90244126e-01 7.28374839e-01 -9.84221324e-02
-4.56642121e-01 1.01280105e+00 -4.92813531e-03 -1.50913224e-01
-1.45008266e-01 -3.99621993e-01 -1.15334785e+00 -1.03911126e+00
-2.92960972e-01 6.06060743e-01 4.11988497e-01 -4.15771127e-01
3.18899006e-02 3.88056695e-01 -1.52669084e+00 7.39292288e-03
7.56080449e-01 7.07892001e-01 6.97460592e-01 9.60036337e-01
1.59716904e-01 1.29611027e+00 -4.18765217e-01 1.40943587e-01
1.23513460e-01 -2.15186790e-01 -4.52012777e-01 1.18936896e-01
-6.05705790e-02 -5.65174878e-01 -5.57101190e-01 5.16873181e-01
6.18270278e-01 7.93633461e-01 7.60932863e-01 -1.35913587e+00
-7.14607298e-01 1.04694344e-01 -7.08883464e-01 4.16813672e-01
-5.04174829e-02 1.01377821e+00 2.10894883e-01 -5.56632340e-01
8.54597032e-01 1.34610558e+00 6.30255461e-01 9.40990150e-01
-1.33046710e+00 -4.37530577e-01 8.28347802e-01 3.93204361e-01
-1.04529047e+00 -2.53216684e-01 7.64740705e-01 -5.38183212e-01
1.27494991e+00 -3.72165352e-01 1.20563161e+00 1.61540270e+00
8.50085020e-01 7.66356945e-01 1.21128476e+00 -1.34493321e-01
7.37231255e-01 4.39977571e-02 -4.28997636e-01 8.11709285e-01
-7.86187276e-02 6.83289111e-01 -7.62920201e-01 -1.78604648e-01
8.71260107e-01 4.70811017e-02 -1.64090946e-01 -9.64252234e-01
-1.01429462e+00 5.46001792e-01 1.63732260e-01 -1.61025837e-01
-6.54822052e-01 7.24280834e-01 3.12238842e-01 5.24140775e-01
5.42542875e-01 3.86430234e-01 -6.25644982e-01 -2.76395708e-01
-5.61138868e-01 5.50870597e-01 4.95889008e-01 8.32913756e-01
4.44761604e-01 7.48664677e-01 -3.16780776e-01 5.32814860e-01
1.59199893e-01 7.14430392e-01 6.72064424e-01 -1.34522474e+00
-1.16537236e-01 -8.35466981e-02 7.26383090e-01 -4.29169178e-01
-3.99148226e-01 -4.58457291e-01 -6.37408495e-01 9.78824735e-01
2.46845454e-01 -3.00803125e-01 -1.16331518e+00 2.05929685e+00
6.14517152e-01 7.58780003e-01 1.53764170e-02 6.90417647e-01
1.26150874e-02 9.38812792e-01 3.92297953e-01 -6.23847306e-01
4.98516500e-01 -1.16897917e+00 -7.07362592e-01 -1.76339932e-02
2.39262745e-01 -1.25514507e-01 1.49626625e+00 5.82284689e-01
-1.14219427e+00 -6.97258472e-01 -9.61856663e-01 4.02942866e-01
-2.14435264e-01 -5.04982233e-01 4.46022272e-01 6.82124868e-02
-1.32832849e+00 8.87224913e-01 -1.40549028e+00 -3.27159315e-01
2.41131350e-01 1.65968522e-01 4.08024713e-02 2.29112491e-01
-9.47962821e-01 1.16914415e+00 3.92840177e-01 -5.09806514e-01
-1.86117411e+00 -9.34157073e-01 -4.85507250e-01 -2.91554064e-01
7.36670375e-01 -1.02432394e+00 1.63472128e+00 -1.15363908e+00
-2.05490232e+00 1.75634325e-01 -5.14476746e-02 -8.41444552e-01
7.72777677e-01 -3.65696162e-01 -1.47877365e-01 -1.02516688e-01
-2.60007262e-01 6.93966746e-01 1.52476645e+00 -1.59170294e+00
-4.37442243e-01 5.00455089e-02 -1.17794313e-01 6.60217524e-01
-1.39870003e-01 -8.15697551e-01 -1.75139785e-01 -6.41127825e-01
-5.60968339e-01 -1.15080047e+00 -3.99044544e-01 6.67286441e-02
2.38214791e-01 -1.16207525e-01 9.66755092e-01 -4.33615178e-01
9.93807435e-01 -1.95147777e+00 5.66742599e-01 -2.01927647e-01
3.76146674e-01 2.12446749e-01 -1.18618272e-01 4.28118259e-01
2.35209242e-01 -2.19429821e-01 1.81731761e-01 -5.29371560e-01
-1.88712105e-01 5.06682277e-01 -8.30452561e-01 3.76474231e-01
-9.45113078e-02 9.56427395e-01 -1.07664573e+00 -2.58671343e-01
3.42897236e-01 4.80335861e-01 -5.10705352e-01 3.19604456e-01
-9.07237947e-01 8.63691270e-01 -3.47364724e-01 3.80202144e-01
4.83774930e-01 -4.16769534e-01 2.16631576e-01 3.54950994e-01
-2.66660243e-01 -3.84430915e-01 -9.01028216e-01 1.68272138e+00
-6.37699664e-01 3.30109328e-01 -1.90276071e-01 -7.32655883e-01
5.44733942e-01 1.39763191e-01 6.22851968e-01 -8.89027655e-01
-1.97596610e-01 -1.59018099e-01 -2.70726264e-01 -5.62479079e-01
3.72802109e-01 -4.06505734e-01 2.10861146e-01 4.35202539e-01
3.36948246e-01 -3.15058947e-01 -2.47766450e-01 1.91410661e-01
1.22369361e+00 7.01165915e-01 2.61125773e-01 -5.37941717e-02
-6.25797287e-02 1.44975930e-01 6.39988482e-01 1.19486868e+00
-4.63028312e-01 1.18048564e-01 2.68586010e-01 -8.15915108e-01
-1.20560050e+00 -1.40207660e+00 4.16515499e-01 1.21755779e+00
3.52070749e-01 -1.70797333e-01 -2.57829875e-01 -5.44875741e-01
2.75972098e-01 1.04098678e+00 -8.49831879e-01 -5.97481012e-01
-6.47855639e-01 -7.18019962e-01 -1.81589946e-01 4.52162027e-01
3.04749638e-01 -1.28380477e+00 -1.11801219e+00 5.41467369e-01
1.73813596e-01 -5.33728302e-01 -2.99321920e-01 3.48379821e-01
-8.63968194e-01 -7.07751334e-01 -8.48027408e-01 -4.00543749e-01
8.78997967e-02 2.69942939e-01 1.10972703e+00 -5.95693588e-02
-8.46850127e-03 1.15131962e+00 -8.89298171e-02 -3.54713470e-01
-5.38852215e-01 -1.57900035e-01 5.75154841e-01 -1.49388909e-01
-3.12114477e-01 -7.29282796e-01 -6.45594120e-01 -1.99069958e-02
-6.71502113e-01 3.27371478e-01 3.52489203e-01 1.18380988e+00
9.51652706e-01 -6.72044680e-02 6.01740658e-01 -4.01303649e-01
6.38524771e-01 -6.67553544e-01 -7.61983097e-01 4.72160786e-01
-8.91323447e-01 3.71369272e-01 8.52910221e-01 -1.32763147e+00
-1.47910464e+00 1.62669480e-01 2.84447819e-01 -1.01894283e+00
2.73310900e-01 7.86262900e-02 3.49283367e-01 -1.74603686e-01
6.65792346e-01 5.58619201e-01 1.11939363e-01 -3.49926472e-01
5.74373424e-01 -1.96269318e-01 3.05674464e-01 -8.38749468e-01
6.60389364e-01 6.13447249e-01 1.70764372e-01 -7.02699482e-01
-5.54139674e-01 5.72223291e-02 -4.66137886e-01 -8.47749174e-01
7.14700162e-01 -9.58230138e-01 -1.00092793e+00 8.90422285e-01
-9.63294029e-01 -1.42642236e+00 -8.15696537e-01 3.08914274e-01
-1.34941304e+00 1.37816146e-01 -5.94412565e-01 -1.06754053e+00
8.51551890e-02 -8.53234410e-01 6.75593555e-01 3.63785326e-01
1.66823253e-01 -1.24893606e+00 7.24595964e-01 -4.33041513e-01
6.06054723e-01 4.02954310e-01 9.18689609e-01 -2.51251813e-02
-6.32032275e-01 4.85583812e-01 5.60358942e-01 4.73307632e-02
-2.50628740e-01 -1.60726905e-01 -7.25978076e-01 -6.74167156e-01
1.89139768e-01 -7.49939322e-01 1.09478092e+00 6.47645772e-01
1.02724230e+00 -2.87254542e-01 -5.12074590e-01 6.61423564e-01
1.40449059e+00 5.85477293e-01 3.27829897e-01 4.77795064e-01
4.82735336e-01 2.20981166e-01 6.38591111e-01 9.06072199e-01
4.35559392e-01 5.03332376e-01 6.90535605e-01 1.95170894e-01
-1.50790140e-01 -7.20565438e-01 5.44849992e-01 5.12432396e-01
-1.43111810e-01 -2.84171999e-01 -8.71939301e-01 4.61350352e-01
-2.18610215e+00 -1.14915383e+00 4.54610199e-01 2.21830368e+00
7.50928760e-01 1.41925588e-01 2.47939840e-01 -5.78884721e-01
4.27767366e-01 1.93798006e-01 -1.47597420e+00 -6.05036020e-02
-1.05936192e-01 8.88128206e-03 3.46555769e-01 6.35922909e-01
-9.97132242e-01 1.23561502e+00 6.34305573e+00 7.79884517e-01
-1.13450480e+00 5.09974062e-01 7.75649309e-01 -5.63869774e-01
-5.33114597e-02 -7.20543265e-02 -6.29709244e-01 4.80267674e-01
1.04748189e+00 -3.78038198e-01 9.02252734e-01 1.00204206e+00
3.95374805e-01 -3.59065413e-01 -7.52257228e-01 8.84739459e-01
-1.04214229e-01 -1.46682131e+00 -7.48792104e-03 -1.11924917e-01
9.75381076e-01 9.72093418e-02 7.13691533e-01 7.66528368e-01
8.22122693e-01 -8.52471709e-01 9.82067406e-01 1.12101698e+00
6.42461061e-01 -5.85832417e-01 -2.91836578e-02 7.23175287e-01
-1.03910208e+00 -6.16686881e-01 -4.30852652e-01 -2.01278552e-02
4.13038224e-01 -9.85526666e-02 -5.56197643e-01 -1.15680248e-02
8.03091526e-01 7.68329501e-01 -3.78122717e-01 1.09060812e+00
1.48100674e-01 5.50044715e-01 -2.28679597e-01 -1.33849263e-01
2.19987765e-01 1.32581815e-01 8.63459527e-01 5.98687053e-01
5.08721352e-01 -1.60894006e-01 3.44635755e-01 7.26467252e-01
3.52672160e-01 -2.70857573e-01 -8.08906972e-01 2.27735937e-01
2.31278837e-01 6.84228003e-01 -5.83569527e-01 -4.68110919e-01
-7.43715689e-02 1.07606041e+00 7.75214314e-01 7.43361831e-01
-1.07098377e+00 4.59934711e-01 6.13817811e-01 2.39628643e-01
3.75160038e-01 -5.81202984e-01 2.53233790e-01 -1.21733773e+00
-3.28384489e-01 -6.97516859e-01 2.00251162e-01 -9.40896690e-01
-1.29115570e+00 4.20033038e-01 3.04641902e-01 -1.21109128e+00
-5.03279388e-01 -3.99737746e-01 -4.90170240e-01 4.48993772e-01
-1.58139014e+00 -1.07612777e+00 -1.99591592e-01 9.70294118e-01
9.62569833e-01 -3.30161184e-01 4.36325520e-01 -3.16513717e-01
-2.76042044e-01 9.77752507e-02 5.28459609e-01 -8.48814309e-01
8.46299350e-01 -1.37295902e+00 3.25741500e-01 4.96174335e-01
-2.46310562e-01 3.15655231e-01 1.09078097e+00 -1.01168418e+00
-1.53142107e+00 -1.20821667e+00 -1.79670285e-02 -5.02208292e-01
7.38319755e-01 -2.17604354e-01 -1.23392713e+00 7.21714795e-01
3.43832970e-01 9.72380340e-02 6.63274452e-02 -3.15092266e-01
-2.10687257e-02 -6.16084300e-02 -8.40469122e-01 8.66618574e-01
1.22421050e+00 -1.86684995e-03 -2.40552172e-01 4.04024720e-01
9.37181056e-01 -5.57911932e-01 -3.45109016e-01 1.60570815e-01
4.89061683e-01 -8.69387567e-01 9.29598331e-01 -1.10928416e+00
-9.56258252e-02 -2.80114204e-01 1.41619340e-01 -1.88535452e+00
-4.68988627e-01 -1.12669885e+00 -9.88609493e-01 6.35324240e-01
7.93014653e-03 -5.54063380e-01 5.63633800e-01 2.49077365e-01
-1.37637347e-01 -5.91425538e-01 -1.07292843e+00 -1.19461656e+00
3.53317142e-01 -3.48833054e-01 3.04436177e-01 5.98986506e-01
-2.76385993e-01 8.45616683e-02 -1.08495581e+00 1.38911661e-02
7.24197030e-01 -2.17178300e-01 7.33208299e-01 -8.43565762e-01
-7.92264342e-01 -9.06967074e-02 2.15527505e-01 -1.20904183e+00
4.51941937e-01 -3.68737370e-01 1.41474381e-01 -1.23220682e+00
1.70768797e-01 -2.87415862e-01 -3.93996328e-01 2.21440539e-01
-7.44030625e-02 -5.15054584e-01 3.77567619e-01 4.84164476e-01
-1.05946052e+00 1.25255430e+00 1.47936511e+00 -1.10820144e-01
-7.71875978e-01 1.03562064e-01 -1.24866664e-01 5.67659557e-01
8.17811131e-01 -5.79305053e-01 -1.02827704e+00 -2.90688872e-01
2.79371291e-01 3.43330503e-01 6.19269729e-01 -1.19228005e+00
2.68001467e-01 -6.77226722e-01 5.12364864e-01 -3.48607421e-01
5.89660764e-01 -6.05140448e-01 3.69600683e-01 8.00421774e-01
-3.97575825e-01 3.95433903e-01 4.46908057e-01 1.36593151e+00
5.07657409e-01 2.67384797e-01 1.00278997e+00 -3.88547957e-01
-1.03139436e+00 4.38745826e-01 -8.92670035e-01 1.47495285e-01
1.37975156e+00 1.64413244e-01 -4.38461661e-01 -6.48601353e-01
-1.31366539e+00 4.49064612e-01 7.12027073e-01 3.17269921e-01
7.15440869e-01 -1.22465837e+00 -1.68819577e-01 -3.81872803e-02
-1.56026676e-01 -4.76845235e-01 7.58556426e-01 7.08624959e-01
-2.42976815e-01 -9.17355642e-02 -5.10142267e-01 -5.31404138e-01
-7.03067124e-01 1.20610082e+00 7.33427227e-01 -4.92198616e-01
-8.09772789e-01 5.44453919e-01 3.56814742e-01 6.96149543e-02
2.94438392e-01 -1.34414032e-01 -8.97533447e-02 -2.99581349e-01
5.25160015e-01 4.79473203e-01 -3.97936523e-01 -2.47902796e-01
3.20214108e-02 3.33648443e-01 -1.08925037e-01 -4.87749606e-01
1.41927707e+00 -3.37736666e-01 5.46737909e-01 1.02686906e+00
4.33268815e-01 -5.98266721e-01 -2.61200595e+00 -1.38639346e-01
-3.10172588e-01 -2.58200914e-01 -4.86380495e-02 -9.41407919e-01
-6.24850869e-01 9.53306019e-01 9.84315813e-01 1.57928038e-02
9.04851794e-01 -9.64472070e-02 5.53776860e-01 5.75066626e-01
6.76851869e-01 -1.59271681e+00 9.19449806e-01 6.82185769e-01
1.10339308e+00 -1.15887368e+00 -2.29293704e-01 5.32504141e-01
-1.21296155e+00 9.49324727e-01 9.89259362e-01 -2.29220569e-01
8.71166468e-01 4.12685066e-01 -2.37684116e-01 6.73669577e-02
-1.58853197e+00 -2.36786664e-01 -1.52683616e-01 1.10080552e+00
-4.35577452e-01 -1.31864950e-01 2.96002448e-01 3.44414085e-01
3.72650445e-01 2.06049293e-01 4.67304945e-01 1.00266635e+00
-6.91399455e-01 -6.37516260e-01 -2.93191243e-02 1.20322295e-01
7.15434700e-02 2.01360658e-01 3.12496275e-02 1.03830695e+00
3.36524695e-02 5.05095720e-01 1.66389346e-02 -3.78577322e-01
1.66749597e-01 2.14110464e-01 6.04800045e-01 -4.26556468e-01
-3.75965685e-01 1.93208113e-01 -4.54733551e-01 -5.89482605e-01
-6.93550035e-02 -8.39971006e-01 -1.06270051e+00 -4.35458928e-01
1.85454890e-01 -3.79076153e-01 2.04484642e-01 6.48170948e-01
5.96617043e-01 5.72160423e-01 4.85631883e-01 -1.08271813e+00
-9.68978047e-01 -6.98599577e-01 -5.35609543e-01 2.15354919e-01
7.64067531e-01 -1.28333461e+00 -3.24205041e-01 2.82291830e-01] | [4.308984756469727, 1.456843376159668] |
23ea3dcd-ab40-437e-9acf-17220b882507 | singing-voice-synthesis-based-on-a-musical | 2212.13703 | null | https://arxiv.org/abs/2212.13703v2 | https://arxiv.org/pdf/2212.13703v2.pdf | Singing Voice Synthesis Based on a Musical Note Position-Aware Attention Mechanism | This paper proposes a novel sequence-to-sequence (seq2seq) model with a musical note position-aware attention mechanism for singing voice synthesis (SVS). A seq2seq modeling approach that can simultaneously perform acoustic and temporal modeling is attractive. However, due to the difficulty of the temporal modeling of singing voices, many recent SVS systems with an encoder-decoder-based model still rely on explicitly on duration information generated by additional modules. Although some studies perform simultaneous modeling using seq2seq models with an attention mechanism, they have insufficient robustness against temporal modeling. The proposed attention mechanism is designed to estimate the attention weights by considering the rhythm given by the musical score. Furthermore, several techniques are also introduced to improve the modeling performance of the singing voice. Experimental results indicated that the proposed model is effective in terms of both naturalness and robustness of timing. | ['Keiichi Tokuda', 'Yoshihiko Nankaku', 'Kei Hashimoto', 'Yukiya Hono'] | 2022-12-28 | null | null | null | null | ['singing-voice-synthesis'] | ['speech'] | [-1.87798534e-02 -2.11768895e-02 4.86755520e-02 -5.98062649e-02
-6.78016126e-01 -3.91270965e-01 3.06595325e-01 -4.27858651e-01
-1.90491840e-01 2.98618674e-01 4.06762332e-01 5.88075034e-02
-1.84825156e-02 -2.17645392e-01 -2.96384960e-01 -5.94433427e-01
1.12512156e-01 -5.42464443e-02 2.27698684e-01 -2.99584627e-01
2.95317739e-01 2.28986368e-01 -1.63721180e+00 1.43744737e-01
8.29738557e-01 8.12865376e-01 8.58008087e-01 1.11831343e+00
-8.08788091e-02 8.35830450e-01 -8.74094725e-01 -1.14316337e-01
2.08089367e-01 -9.34642911e-01 -2.09014505e-01 -1.87743932e-01
2.22800374e-01 -1.79474160e-01 -3.27097148e-01 7.43624866e-01
1.06024790e+00 4.63864356e-01 3.13463718e-01 -7.74119496e-01
-4.57169622e-01 7.88770080e-01 -1.64249465e-01 1.84547380e-01
3.13223213e-01 2.86385089e-01 1.22187757e+00 -8.90302300e-01
2.13294640e-01 1.10053909e+00 5.97584546e-01 6.91819370e-01
-7.62050629e-01 -6.86002672e-01 1.16945215e-01 4.32830304e-01
-1.24404800e+00 -5.62460840e-01 1.01056051e+00 -3.30045521e-01
1.06587982e+00 6.63742661e-01 8.05177748e-01 7.74822354e-01
1.00372165e-01 7.88981140e-01 6.28417611e-01 -5.56654155e-01
6.69002999e-03 -1.26205981e-01 -1.04108259e-01 2.88186103e-01
-7.37776220e-01 3.91307235e-01 -9.86236632e-01 1.74575701e-01
9.18627143e-01 -4.65746164e-01 -1.96298286e-01 1.65227339e-01
-1.02035534e+00 5.79676569e-01 1.13973565e-01 4.75646377e-01
-4.58241343e-01 3.11516881e-01 7.35255361e-01 7.71274865e-02
2.98435569e-01 6.22811794e-01 -3.20775867e-01 -4.74502176e-01
-1.36571121e+00 2.45651826e-01 4.99222934e-01 1.01092649e+00
-8.41180831e-02 9.26765561e-01 -6.41255796e-01 1.03100967e+00
3.21248919e-01 3.67268890e-01 8.06647539e-01 -1.02890217e+00
3.25648248e-01 -1.37212723e-02 3.44577953e-02 -8.07617724e-01
-2.58882463e-01 -6.41132653e-01 -5.67472279e-01 -2.02729064e-03
1.93539590e-01 -1.45010769e-01 -6.33298457e-01 1.91439271e+00
-2.52272803e-02 4.10716683e-01 -1.45834371e-01 1.05968690e+00
6.99679554e-01 9.89071250e-01 5.69856772e-03 -5.88762939e-01
1.16185760e+00 -1.37609327e+00 -1.45857453e+00 -3.35244648e-02
1.46769375e-01 -9.34062541e-01 1.13526475e+00 4.81149435e-01
-1.50932717e+00 -1.17396355e+00 -9.41944957e-01 -1.73311189e-01
3.19628656e-01 5.57891309e-01 8.24726224e-02 6.45016968e-01
-9.71146762e-01 8.06626976e-01 -6.72938287e-01 -8.84648412e-02
-4.94477391e-01 4.33025569e-01 2.78861761e-01 7.44890869e-01
-1.33124590e+00 7.12644160e-01 3.43693525e-01 2.71493345e-01
-8.28431904e-01 -6.77251101e-01 -6.40884638e-01 3.48706037e-01
5.18042371e-02 -5.72470129e-01 1.63499701e+00 -1.01564050e+00
-2.14674282e+00 7.04399124e-02 -2.19485402e-01 -4.08568472e-01
3.14396113e-01 -4.23290610e-01 -4.65320051e-01 1.30220786e-01
-3.51616740e-01 4.35091943e-01 1.08210158e+00 -8.34405303e-01
-3.16266477e-01 3.62320207e-02 -2.86833197e-01 3.21543753e-01
-4.94863480e-01 2.78090447e-01 -4.56975996e-01 -1.25769508e+00
-1.16408408e-01 -8.93588305e-01 -1.36788338e-01 -1.88092053e-01
-2.31120855e-01 -3.76794338e-02 6.30203187e-01 -1.15638125e+00
2.15234613e+00 -2.38745117e+00 4.45565552e-01 -2.47242898e-01
-4.24550533e-01 8.30579519e-01 -2.80719608e-01 7.56078184e-01
-1.76647365e-01 -1.77759692e-01 -1.70722231e-01 -5.93547165e-01
-1.25135988e-01 -3.81202549e-02 -3.93389672e-01 1.23070916e-02
1.72446400e-01 6.70895338e-01 -7.47651160e-01 -5.22552669e-01
2.16933563e-01 5.38244665e-01 -8.92540812e-01 7.01583028e-01
-2.42867664e-01 6.32412910e-01 3.10772825e-02 3.05725843e-01
2.54826397e-01 5.75777173e-01 2.62125134e-01 -2.77669102e-01
-4.11393195e-01 6.33369684e-01 -1.23027492e+00 1.74238336e+00
-7.38137186e-01 3.84094238e-01 3.04992557e-01 -5.30249655e-01
1.19376397e+00 8.55517626e-01 4.37170506e-01 -4.63869989e-01
5.39222024e-02 2.20573336e-01 5.31426311e-01 -5.98138213e-01
7.50064313e-01 -4.61596519e-01 3.17999244e-01 1.17960051e-01
2.23668993e-01 -2.82737941e-01 9.82921049e-02 -1.52723968e-01
7.41582990e-01 5.77025652e-01 2.28981346e-01 -1.49190575e-01
6.80442393e-01 -4.07501578e-01 9.74208653e-01 3.48235607e-01
-1.25086620e-01 9.69249547e-01 9.01383385e-02 -3.16443928e-02
-1.19653916e+00 -6.66347682e-01 2.67974734e-01 1.14347553e+00
-2.96854377e-01 -8.23331475e-01 -8.12228620e-01 -6.65914118e-02
-3.96849692e-01 9.49644089e-01 -2.09624320e-01 -1.88081369e-01
-7.65751362e-01 -1.68754771e-01 8.55812550e-01 8.07676494e-01
-2.00644117e-02 -1.37296069e+00 -4.26091939e-01 7.39062667e-01
-3.17132801e-01 -7.38168299e-01 -1.24907780e+00 4.81536575e-02
-9.37703431e-01 -5.85758746e-01 -1.03765595e+00 -6.98793650e-01
-7.39654377e-02 8.62709656e-02 5.41415751e-01 -2.05768067e-02
-3.21989246e-02 1.00907959e-01 -5.03060162e-01 -4.97953027e-01
-4.86544460e-01 1.26892075e-01 3.34552646e-01 1.46703601e-01
-1.11127175e-01 -7.50114858e-01 -3.49176347e-01 1.90426201e-01
-8.03250492e-01 1.18827321e-01 3.70851755e-01 8.78639936e-01
2.83612251e-01 -2.38190696e-01 1.12030768e+00 -3.31586689e-01
8.00797701e-01 -2.15556815e-01 -4.84194815e-01 8.91942251e-03
-3.47265154e-01 -2.06829030e-02 8.99525642e-01 -7.91472018e-01
-1.21053195e+00 6.01759665e-02 -5.28006613e-01 -8.17066371e-01
1.87225521e-01 2.99389690e-01 -2.78494567e-01 2.92834371e-01
1.71237603e-01 4.12998796e-01 3.08645457e-01 -6.58759356e-01
1.44699171e-01 8.71106267e-01 4.50484961e-01 -4.18984354e-01
5.32474756e-01 -3.25985849e-01 -2.08465725e-01 -1.02521765e+00
-4.35028255e-01 -5.16286969e-01 -6.26920938e-01 -3.68609160e-01
7.90689051e-01 -8.35780680e-01 -8.87578666e-01 4.84230608e-01
-1.41514611e+00 -1.64013565e-01 -3.33247662e-01 8.20404470e-01
-9.76526618e-01 5.66908002e-01 -9.32547212e-01 -1.39135504e+00
-5.01782894e-01 -1.13391471e+00 8.50171626e-01 4.71546166e-02
-5.48768878e-01 -7.33229995e-01 8.24825615e-02 3.72900844e-01
5.79223454e-01 -3.34999889e-01 8.26568007e-01 -4.64421988e-01
-4.19980079e-01 3.79744880e-02 3.26771915e-01 6.84961081e-01
8.90344977e-02 2.41028950e-01 -1.11057246e+00 -5.21694273e-02
2.36203685e-01 7.49620721e-02 5.81740975e-01 5.40578008e-01
1.08112991e+00 -4.19001937e-01 3.26734364e-01 4.07163620e-01
1.04225302e+00 6.68155849e-01 7.30813563e-01 -1.80927753e-01
6.64624214e-01 6.57449663e-01 9.20335650e-01 6.67490363e-01
-8.60004779e-03 1.07011223e+00 3.07583064e-01 1.30588830e-01
-5.73492765e-01 -4.35786307e-01 7.86631763e-01 1.83570623e+00
-2.93821305e-01 -3.29254329e-01 -4.01465982e-01 6.82581127e-01
-1.77487886e+00 -1.19638145e+00 -2.15276212e-01 2.29853678e+00
9.01968062e-01 1.71005121e-03 2.57627785e-01 4.98958349e-01
7.48762488e-01 3.87848407e-01 -3.27608883e-01 -9.76072609e-01
2.01001894e-02 2.89247811e-01 1.18569866e-01 6.16654754e-01
-5.80638587e-01 8.77394915e-01 6.77409410e+00 1.15502179e+00
-1.17044520e+00 -4.49864753e-02 -1.30238369e-01 -3.75479847e-01
-3.85124564e-01 -8.47117156e-02 -7.97641575e-01 5.86167157e-01
1.15827298e+00 -5.89385480e-02 5.90608895e-01 5.38528800e-01
7.78653622e-01 3.87533247e-01 -7.80112565e-01 7.60557294e-01
8.77553150e-02 -8.40525210e-01 2.78993268e-02 -1.96345270e-01
5.08398652e-01 -5.70947528e-01 8.85324702e-02 3.70531499e-01
-6.16095185e-01 -8.45187545e-01 1.06539929e+00 7.61463761e-01
9.05532122e-01 -7.57619560e-01 4.94824499e-01 5.52175164e-01
-1.50847220e+00 -2.58360028e-01 -8.35283753e-03 -2.56452441e-01
5.11374593e-01 1.06559753e-01 -9.67238188e-01 6.52984560e-01
1.68236971e-01 5.75790048e-01 -1.59772307e-01 1.20734966e+00
-2.31447086e-01 1.03088212e+00 -1.79439988e-02 -1.12290129e-01
9.46977884e-02 -1.20891131e-01 8.98629725e-01 1.27013838e+00
7.10465670e-01 1.15605824e-01 -1.19012982e-01 7.96125174e-01
4.36354518e-01 4.47938472e-01 -4.03564423e-01 -3.85830462e-01
4.50354397e-01 9.15548623e-01 -4.68016267e-02 -7.26255104e-02
-3.39529186e-01 9.47532356e-01 -1.67139992e-01 2.26279020e-01
-9.19068396e-01 -4.07331437e-01 6.65554643e-01 2.62603968e-01
4.61308420e-01 -5.89182556e-01 -1.64192289e-01 -7.41242468e-01
-1.92105398e-01 -9.45960701e-01 -8.95508751e-02 -1.05750978e+00
-8.09992969e-01 7.37605572e-01 -2.55756646e-01 -1.49688601e+00
-6.65427744e-01 -2.37194076e-01 -7.30393648e-01 1.06081498e+00
-1.27739275e+00 -1.02145302e+00 3.05042684e-01 1.81040987e-01
1.21953297e+00 -2.06870273e-01 8.91176879e-01 4.90836114e-01
-4.22628790e-01 6.31531656e-01 -8.98385346e-02 -3.07797670e-01
8.89394283e-01 -1.13713276e+00 2.57665843e-01 7.16897726e-01
1.81176037e-01 6.18260860e-01 7.88325250e-01 -5.90590596e-01
-1.20483232e+00 -8.31979513e-01 1.17728829e+00 1.37068555e-02
3.42664182e-01 -1.74901828e-01 -1.05149055e+00 3.01205575e-01
5.34337223e-01 -4.48096633e-01 7.15560436e-01 -2.74688788e-02
4.16474454e-02 -1.14453405e-01 -6.50997937e-01 6.03012741e-01
8.85827541e-01 -6.59941614e-01 -7.47693241e-01 -1.47588357e-01
8.30599666e-01 -3.91589463e-01 -8.32098663e-01 3.62451166e-01
5.99966645e-01 -9.10409510e-01 6.53672457e-01 -3.86388749e-01
2.67612696e-01 -5.96466839e-01 -9.36771780e-02 -1.37427127e+00
-5.90074539e-01 -9.35987830e-01 -3.56777072e-01 1.53837419e+00
1.97109595e-01 -9.20716394e-03 2.26257458e-01 2.34036297e-01
-5.38827658e-01 -4.12015140e-01 -9.53438282e-01 -1.10366499e+00
-2.60044843e-01 -5.43905556e-01 4.31346148e-01 4.96666908e-01
1.09429687e-01 4.60582584e-01 -1.09947646e+00 1.69682652e-01
1.30368888e-01 -9.52407718e-02 3.86600465e-01 -9.03686881e-01
-5.97700775e-01 -3.71825427e-01 -1.19502105e-01 -1.07786322e+00
5.36339805e-02 -5.45210898e-01 3.18006665e-01 -1.13320410e+00
-2.29032978e-01 -1.92904230e-02 -4.79633570e-01 1.55713439e-01
-3.05716127e-01 -7.29179680e-02 7.27770984e-01 -1.87827367e-02
-1.90766633e-01 1.03993964e+00 1.41065109e+00 1.62204415e-01
-4.29647714e-01 2.52306104e-01 -4.71383110e-02 5.37004530e-01
8.39562535e-01 -3.08956206e-01 -4.52016294e-01 -3.25923890e-01
-2.91659802e-01 7.09186792e-01 -6.44576177e-02 -1.23792052e+00
4.87229347e-01 4.52732295e-03 -1.68090448e-01 -7.82498300e-01
8.26030850e-01 -6.24814510e-01 2.96219528e-01 4.43909824e-01
-6.05906248e-01 1.38404861e-01 4.79751110e-01 4.68243092e-01
-6.07522309e-01 -4.56165731e-01 7.75361478e-01 7.62849972e-02
-4.80205894e-01 3.72595573e-03 -6.12663031e-01 -2.47103840e-01
7.20041335e-01 -2.00704068e-01 4.71540511e-01 -6.29771888e-01
-8.61994684e-01 -1.92214385e-01 -7.65103698e-02 6.05467200e-01
5.38093150e-01 -1.52541506e+00 -6.63671911e-01 3.35102618e-01
-6.34342432e-02 -5.43599427e-01 6.00226641e-01 6.57922089e-01
-1.75312757e-01 7.26600587e-01 -1.85715497e-01 -2.22046673e-01
-1.48743713e+00 6.69675112e-01 2.63729841e-01 -9.78171974e-02
-2.99931288e-01 8.18545341e-01 1.85772955e-01 -3.99609804e-01
5.26110649e-01 -2.65062898e-01 -4.13051873e-01 1.16784610e-01
5.36732674e-01 5.39368868e-01 -8.29701275e-02 -7.73698449e-01
-2.40491509e-01 6.37093961e-01 3.89630109e-01 -4.37506557e-01
1.02038717e+00 -2.13382140e-01 2.78059453e-01 8.07489753e-01
6.96903050e-01 4.48495477e-01 -1.29501510e+00 6.06449042e-03
-8.71717185e-02 -4.09387350e-01 1.23104937e-01 -8.51559401e-01
-9.66218948e-01 1.18873644e+00 2.96522886e-01 4.92352434e-03
1.30167270e+00 -7.60974705e-01 9.52094793e-01 -2.61525184e-01
4.39969562e-02 -1.28470814e+00 1.08789511e-01 7.43731916e-01
1.29599261e+00 -7.42251277e-01 -5.45590699e-01 -3.51824731e-01
-8.78971577e-01 1.29126310e+00 6.83764517e-01 7.52010271e-02
5.28409898e-01 3.01925153e-01 1.25424981e-01 4.93656129e-01
-1.00771952e+00 -2.26891205e-01 4.82860774e-01 4.36844260e-01
8.69655430e-01 -1.80737123e-01 -7.50690341e-01 1.00614786e+00
-1.96128651e-01 9.10133943e-02 3.67074788e-01 3.93378407e-01
-4.02347535e-01 -1.45271778e+00 -4.40633625e-01 -1.53235555e-01
-6.48713470e-01 -4.73230481e-01 -1.32906973e-01 2.79990107e-01
1.93517119e-01 9.87606823e-01 -5.62870763e-02 -5.71834862e-01
4.26988214e-01 4.02795315e-01 3.22160542e-01 -7.34289110e-01
-1.19090509e+00 6.59026086e-01 1.18928067e-01 -2.98030198e-01
-1.74308449e-01 -5.32369077e-01 -1.17068934e+00 2.18696415e-01
-4.90822434e-01 3.45329374e-01 7.62658238e-01 5.99897325e-01
4.57606673e-01 1.13216960e+00 8.77390981e-01 -7.57941663e-01
-7.45028555e-01 -1.37008357e+00 -9.81733918e-01 -1.45772165e-02
3.56520474e-01 -2.92336881e-01 -1.75487578e-01 1.12904806e-03] | [15.51501750946045, 6.172844886779785] |
ee0ed370-1b76-4a69-9a0d-739903b18dc2 | an-intelligent-algorithmic-trading-based-on-a | 2208.10707 | null | https://arxiv.org/abs/2208.10707v2 | https://arxiv.org/pdf/2208.10707v2.pdf | An intelligent algorithmic trading based on a risk-return reinforcement learning algorithm | This scientific paper propose a novel portfolio optimization model using an improved deep reinforcement learning algorithm. The objective function of the optimization model is the weighted sum of the expectation and value at risk(VaR) of portfolio cumulative return. The proposed algorithm is based on actor-critic architecture, in which the main task of critical network is to learn the distribution of portfolio cumulative return using quantile regression, and actor network outputs the optimal portfolio weight by maximizing the objective function mentioned above. Meanwhile, we exploit a linear transformation function to realize asset short selling. Finally, A multi-process method is used, called Ape-x, to accelerate the speed of deep reinforcement learning training. To validate our proposed approach, we conduct backtesting for two representative portfolios and observe that the proposed model in this work is superior to the benchmark strategies. | ['Boyi Jin'] | 2022-08-23 | null | null | null | null | ['algorithmic-trading', 'portfolio-optimization'] | ['time-series', 'time-series'] | [-3.79794806e-01 -1.33172870e-01 -8.74832049e-02 -1.45108506e-01
-5.83784223e-01 -4.60104406e-01 2.59775758e-01 -1.28978640e-01
-4.37617987e-01 9.14234817e-01 3.52618918e-02 -3.93993199e-01
-4.85105753e-01 -1.26892328e+00 -7.00839818e-01 -7.69126177e-01
5.42948544e-02 3.03102344e-01 -2.02490062e-01 -2.55377412e-01
5.66385031e-01 3.11752170e-01 -7.51327157e-01 -1.32879928e-01
7.85719872e-01 1.37149239e+00 -1.87693208e-01 3.15799445e-01
-1.78418979e-01 8.51646781e-01 -8.52969825e-01 -7.22841859e-01
5.42514145e-01 -4.51457322e-01 -9.87959057e-02 -1.92168355e-01
-4.10371184e-01 -6.02315784e-01 -2.70212680e-01 1.17520285e+00
6.78883731e-01 2.63391584e-01 6.89001203e-01 -9.69609261e-01
-6.60145640e-01 9.17699754e-01 -5.73020458e-01 2.85878778e-01
-2.92666048e-01 2.94541419e-01 1.25910687e+00 -5.21763802e-01
-1.28352642e-01 1.24381328e+00 5.68829834e-01 3.00973743e-01
-8.75511885e-01 -6.42437994e-01 6.73389956e-02 -2.63907239e-02
-7.52235949e-01 2.02705100e-01 8.40160489e-01 -1.49334311e-01
8.29686403e-01 -1.02363333e-01 9.43585277e-01 8.46000850e-01
9.83657300e-01 6.83264494e-01 8.53407443e-01 -9.83886719e-02
4.37067777e-01 7.99784716e-03 -2.24767864e-01 3.13020885e-01
3.43731910e-01 7.03094661e-01 -1.75703317e-02 -1.23599701e-01
9.64919329e-01 3.95664066e-01 1.45000890e-01 1.15941688e-02
-7.25560188e-01 9.76479709e-01 3.09503615e-01 -1.45364940e-01
-6.95500255e-01 5.71026802e-01 3.19494307e-01 5.34930646e-01
4.87424612e-01 5.08564651e-01 -3.77291769e-01 -2.37426251e-01
-6.43965006e-01 4.73895103e-01 8.51433456e-01 4.81858402e-01
2.72127748e-01 7.00835168e-01 -5.14638841e-01 6.05573058e-01
7.33005226e-01 8.37635040e-01 6.06141329e-01 -1.06304932e+00
7.22348630e-01 4.70126599e-01 3.75574142e-01 -8.94125879e-01
-1.30940363e-01 -9.34114397e-01 -6.46735847e-01 7.38627911e-01
2.63145626e-01 -7.79279768e-01 -2.82967627e-01 1.37775230e+00
2.36478113e-02 1.54046655e-01 4.37747866e-01 6.86996520e-01
1.38911083e-01 9.05756891e-01 7.19644874e-02 -4.03618962e-01
8.59274983e-01 -1.05698788e+00 -7.47858942e-01 1.35943592e-01
-2.54132897e-01 -4.88045722e-01 7.53829598e-01 4.00144219e-01
-1.30291426e+00 -4.71205086e-01 -1.19884491e+00 8.92657757e-01
2.49856152e-02 5.50419874e-02 4.91954893e-01 6.80820584e-01
-4.10541266e-01 1.09479225e+00 -5.89098692e-01 6.24601007e-01
4.30061519e-01 3.23084831e-01 4.40302223e-01 7.16790974e-01
-1.42883217e+00 7.93685734e-01 7.96456099e-01 2.28564098e-01
-1.23900461e+00 -6.80468976e-01 -2.57183164e-01 7.54142344e-01
4.26422596e-01 -6.33521616e-01 1.28445351e+00 -1.27146220e+00
-2.32624292e+00 -9.76184476e-03 8.16754222e-01 -8.61349821e-01
7.21843183e-01 -4.40120280e-01 -3.82969171e-01 -2.36476567e-02
-5.10402560e-01 -4.09405790e-02 1.13663423e+00 -6.76939428e-01
-7.75796592e-01 -1.06674572e-02 6.31191432e-02 1.80070639e-01
-5.38092554e-01 -1.83627352e-01 2.22172946e-01 -9.84999895e-01
-5.81866562e-01 -4.62109834e-01 -2.53478646e-01 -4.54114586e-01
6.03305437e-02 -2.69129664e-01 2.90687501e-01 -7.52253413e-01
1.38670075e+00 -2.06588531e+00 -1.64900064e-01 5.36949337e-01
-1.04928970e-01 1.91867650e-01 -1.22169480e-01 5.60368598e-01
-1.15865938e-01 -1.14939086e-01 -1.72804311e-01 1.82034761e-01
3.75796050e-01 -8.48907903e-02 -7.80736387e-01 1.42733470e-01
1.85378477e-01 1.18583190e+00 -7.52933502e-01 -5.71634471e-02
4.43483964e-02 1.12681966e-02 -3.51607263e-01 6.11597359e-01
-6.10796154e-01 -3.96612957e-02 -6.59475505e-01 3.35400522e-01
5.54273069e-01 -7.37890527e-02 5.21661192e-02 1.63672686e-01
-1.39992684e-01 -8.66740867e-02 -1.24161899e+00 9.44783628e-01
-4.50102121e-01 9.03792456e-02 -4.42606300e-01 -1.09099662e+00
1.43758929e+00 2.79206485e-01 4.79193330e-01 -7.34590590e-01
3.17926735e-01 2.02359974e-01 1.57247093e-02 -3.61323208e-01
2.37410948e-01 -5.17079234e-01 1.62200555e-01 8.02289188e-01
-2.19434276e-01 7.12149888e-02 -3.70021649e-02 -3.76390457e-01
8.22439790e-01 3.81244063e-01 9.41509604e-02 -1.10440955e-01
5.96458673e-01 -3.70398134e-01 6.69385493e-01 6.71737611e-01
4.74242158e-02 1.25252269e-02 1.12203872e+00 -4.48079050e-01
-1.22468042e+00 -1.19411957e+00 2.06904203e-01 7.85922170e-01
-1.24391519e-01 3.84444267e-01 -5.73536932e-01 -7.57659793e-01
4.66969937e-01 9.73538101e-01 -4.85028416e-01 -4.04217273e-01
-4.48855251e-01 -1.00815856e+00 3.27454746e-01 7.01640368e-01
6.87799990e-01 -1.53917241e+00 -6.70889616e-01 6.95376813e-01
6.11300766e-01 -2.35102773e-01 -3.31228197e-01 -7.96198845e-02
-8.91732454e-01 -9.38084781e-01 -9.67665255e-01 -3.18993181e-01
2.59149641e-01 -4.65981185e-01 9.23578858e-01 -1.92655519e-01
2.62955785e-01 2.70734251e-01 -8.73732492e-02 -7.09612608e-01
-2.39158869e-01 -1.69157628e-02 -2.18548641e-01 3.23856205e-01
-1.20323569e-01 -6.27568960e-01 -8.92611980e-01 1.30701855e-01
-7.60908902e-01 -5.03683209e-01 8.75180662e-01 1.14270496e+00
6.35261238e-01 3.04363251e-01 1.32346439e+00 -7.35958159e-01
1.34558880e+00 -5.18318713e-01 -1.36505961e+00 5.78441501e-01
-1.01936400e+00 2.91941941e-01 9.26525414e-01 -4.34747279e-01
-1.45290184e+00 -5.72094619e-01 -1.41846702e-01 -4.41775382e-01
7.56248832e-01 5.77261627e-01 -1.26284465e-01 1.23834796e-01
5.06239990e-03 2.68826097e-01 2.20998600e-01 -2.89334446e-01
1.19248115e-01 4.30779934e-01 3.48172456e-01 -5.50556302e-01
5.31837940e-01 -7.70733133e-03 2.60675788e-01 2.13501021e-01
-8.96832347e-01 3.37837577e-01 -7.81300850e-03 -4.38121557e-01
5.22202671e-01 -5.70315182e-01 -1.46816218e+00 5.64807773e-01
-8.17373514e-01 -3.60278666e-01 -6.35834455e-01 6.81771100e-01
-8.86551082e-01 1.39024081e-02 -5.85644305e-01 -1.06756747e+00
-9.87711072e-01 -7.73675442e-01 3.98051977e-01 5.56481719e-01
4.34341162e-01 -1.34111619e+00 3.58088285e-01 -2.64266759e-01
5.31110466e-01 5.73444307e-01 1.01312459e+00 -9.36805248e-01
-3.88391227e-01 -3.24214518e-01 -1.44524246e-01 8.55029225e-01
-1.94036245e-01 6.76224306e-02 -4.52685565e-01 -3.51440847e-01
4.49005723e-01 -2.29861930e-01 6.43415868e-01 4.88867223e-01
1.23068976e+00 -3.79147053e-01 3.74996454e-01 7.04325974e-01
1.80585372e+00 6.64551616e-01 6.79566383e-01 8.05490196e-01
1.83331832e-01 3.26707512e-01 9.09457505e-01 8.72559726e-01
-9.62103009e-02 1.05135299e-01 5.98252714e-01 3.62612873e-01
5.67983925e-01 -2.91419238e-01 6.53685510e-01 6.19478643e-01
-1.08692981e-01 -1.70789331e-01 -6.95086896e-01 3.36862765e-02
-2.12311339e+00 -1.26419926e+00 4.47692037e-01 2.19998550e+00
5.40800989e-01 7.45833874e-01 1.21796571e-01 -1.71708986e-01
6.01591349e-01 1.77040294e-01 -9.71678734e-01 -6.13556862e-01
2.52910592e-02 3.67699742e-01 6.72724724e-01 3.85273457e-01
-8.56431723e-01 6.60946071e-01 6.47846651e+00 9.99601305e-01
-1.13337207e+00 -2.59304315e-01 8.08149576e-01 -1.95867002e-01
-5.91669381e-01 -8.91805068e-02 -6.68672979e-01 7.93422818e-01
9.08838868e-01 -6.44192815e-01 6.54241741e-01 9.04845238e-01
3.03487211e-01 3.14563990e-01 -6.79947019e-01 5.68216205e-01
-2.06418782e-01 -1.25061333e+00 -6.09220415e-02 -1.40485331e-01
6.96891546e-01 -5.02809942e-01 4.86661643e-01 7.28533149e-01
4.82805401e-01 -1.03818011e+00 7.12962806e-01 1.32226968e+00
2.10992262e-01 -1.45739973e+00 9.96849179e-01 1.98601633e-01
-8.65039349e-01 -6.36652052e-01 -7.43972957e-01 7.37021118e-02
1.73304603e-01 4.94690090e-01 -4.50809419e-01 5.11287808e-01
3.02958429e-01 5.61530054e-01 -2.62903631e-01 1.17885721e+00
-5.19717038e-01 7.04712689e-01 8.85260664e-03 -5.25605261e-01
5.23788929e-01 -6.66055202e-01 5.15451729e-01 7.81971633e-01
5.71371794e-01 -1.04816668e-01 -7.94187710e-02 9.83381987e-01
-1.49510384e-01 8.64795521e-02 -1.68167278e-01 -1.38900161e-01
3.21581751e-01 1.24204051e+00 -3.49886060e-01 -3.03393990e-01
-2.22866103e-01 4.06570286e-01 2.14312851e-01 4.09303010e-01
-1.10280502e+00 -7.26431191e-01 1.96148574e-01 -4.81852174e-01
5.26254416e-01 7.54592335e-03 -3.89332414e-01 -7.97667146e-01
-1.15356660e-02 -8.22859883e-01 3.09843957e-01 -4.59373385e-01
-1.49617136e+00 3.07427913e-01 -2.52373338e-01 -1.31915343e+00
-3.45981091e-01 -7.24306166e-01 -1.25241506e+00 1.01634288e+00
-1.65928614e+00 -4.37048197e-01 1.37332559e-01 4.59363759e-01
3.41234148e-01 -1.08833933e+00 4.55733240e-01 1.94681138e-01
-8.12993050e-01 7.00166643e-01 9.50614929e-01 -1.53732300e-02
4.29337949e-01 -1.39327896e+00 3.60702783e-01 3.95535439e-01
-3.00152093e-01 1.61637649e-01 4.94342953e-01 -8.08904231e-01
-1.26090944e+00 -1.06963813e+00 1.31262809e-01 1.51463747e-01
1.14776242e+00 1.97077930e-01 -6.17238760e-01 5.26904345e-01
4.49610740e-01 -3.63601893e-01 4.18727696e-01 -3.99828434e-01
1.30974665e-01 -6.90971076e-01 -1.09287357e+00 4.71174389e-01
1.52530536e-01 1.31709471e-01 -5.82087636e-01 4.93209772e-02
9.20672238e-01 -2.57870615e-01 -1.21714699e+00 2.92913586e-01
6.97639942e-01 -7.74081647e-01 8.91338646e-01 -7.19970465e-01
6.74078465e-01 1.21859409e-01 2.35964060e-01 -1.62011576e+00
-3.69119883e-01 -6.98805809e-01 -5.48373342e-01 1.28339791e+00
5.18318474e-01 -1.00825965e+00 8.35654497e-01 3.17491293e-01
8.87047872e-02 -1.28026533e+00 -7.45977819e-01 -8.98211479e-01
2.44220689e-01 1.02439940e-01 1.13636780e+00 3.55832547e-01
-4.10939008e-01 5.09375818e-02 -4.99100089e-01 -9.79233235e-02
9.02934074e-01 3.75419468e-01 3.45732450e-01 -7.48919785e-01
-8.14187288e-01 -7.44107425e-01 2.64034450e-01 -6.49041533e-01
2.69462615e-01 -6.68153048e-01 -2.85977513e-01 -1.17184484e+00
-7.18528330e-02 -1.11395173e-01 -1.03208590e+00 1.53953135e-01
-3.26003343e-01 -5.18439054e-01 3.52682590e-01 1.05629284e-02
-4.97369826e-01 1.13409686e+00 1.51833677e+00 -2.45000377e-01
-1.77296266e-01 6.97768867e-01 -6.13135576e-01 5.10854602e-01
1.11504793e+00 -5.68272650e-01 -3.45190287e-01 -2.63117850e-01
5.28157651e-01 6.04329884e-01 6.96390793e-02 -7.37005770e-01
8.51939842e-02 -4.56342936e-01 7.35562742e-01 -5.37156940e-01
-9.32611376e-02 -7.87060320e-01 -1.82829667e-02 7.76704073e-01
-4.80898559e-01 5.89384437e-01 -1.79732870e-02 6.49014533e-01
-4.20788825e-01 -7.43928492e-01 5.45847476e-01 -7.69813880e-02
-2.30120957e-01 4.72807407e-01 -1.63187921e-01 8.10835436e-02
1.10508525e+00 3.64005297e-01 -1.55284598e-01 -3.25628072e-01
-5.11167526e-01 6.08872652e-01 -2.48423919e-01 1.76042348e-01
8.57819736e-01 -1.59547520e+00 -9.08992946e-01 -6.56654686e-02
-5.73816538e-01 -3.66578788e-01 6.50443956e-02 2.89003879e-01
-7.98110008e-01 1.47930712e-01 -4.81847405e-01 6.82267696e-02
-3.95329773e-01 4.55614686e-01 8.65525663e-01 -9.28565800e-01
-4.08432752e-01 4.35577989e-01 -1.55972794e-01 -1.28870308e-01
3.11002910e-01 1.68756723e-01 -4.90388960e-01 9.31218341e-02
4.39671040e-01 6.62098885e-01 -3.30838352e-01 3.79590720e-01
2.81878918e-01 4.46769208e-01 2.35868692e-02 -4.38201159e-01
1.72136891e+00 2.30877697e-01 -1.76102906e-01 2.27900296e-01
9.45921838e-01 -2.70059526e-01 -1.71264088e+00 4.18977216e-02
1.58858135e-01 -4.03487593e-01 -1.50077581e-01 -8.95692587e-01
-1.26424849e+00 8.71311307e-01 6.56799734e-01 2.27168053e-01
1.16885269e+00 -9.14953828e-01 1.01802719e+00 6.05240703e-01
6.22789115e-02 -1.37355101e+00 3.83996129e-01 6.28596604e-01
9.99889076e-01 -8.95341575e-01 1.51335180e-01 6.35483265e-01
-8.43555450e-01 1.41463733e+00 6.47099257e-01 -9.92700875e-01
7.68519342e-01 1.90442458e-01 -1.51806157e-02 2.32412331e-02
-9.20580804e-01 3.47155690e-01 6.07489273e-02 6.36152327e-02
4.35365178e-03 5.02014048e-02 -6.85250044e-01 1.02443480e+00
-2.37154409e-01 -1.30260915e-01 2.30964780e-01 5.89590311e-01
-5.10112643e-01 -1.20407331e+00 -3.92064124e-01 4.22724366e-01
-8.51885378e-01 7.43203610e-02 1.72919825e-01 6.13179386e-01
-3.27751458e-01 4.93924856e-01 1.79632053e-01 -1.22793928e-01
5.18778801e-01 -1.57618061e-01 3.37465197e-01 -2.02468842e-01
-1.12206590e+00 1.92194059e-01 -3.64167601e-01 -2.35704422e-01
1.96097381e-02 -4.75879103e-01 -1.07141352e+00 -2.42292687e-01
-2.13150024e-01 3.47364843e-01 5.96742690e-01 7.65709341e-01
-1.45679086e-01 1.01881933e+00 1.36567819e+00 -3.80141705e-01
-1.60154355e+00 -7.77353168e-01 -9.47295308e-01 1.45655915e-01
1.29042238e-01 -4.13851589e-01 -2.16468945e-01 -4.93176937e-01] | [4.492221355438232, 3.972017288208008] |
cbeb3858-add6-4d8b-999c-edb4dce5703d | solving-the-hp-model-with-nested-monte-carlo | 2301.09533 | null | https://arxiv.org/abs/2301.09533v2 | https://arxiv.org/pdf/2301.09533v2.pdf | Solving the HP model with Nested Monte Carlo Search | In this paper we present a new Monte Carlo Search (MCS) algorithm for finding the ground state energy of proteins in the HP-model. We also compare it briefly to other MCS algorithms not usually used on the HP-model and provide an overview of the algorithms used on HP-model. The algorithm presented in this paper does not beat state of the art algorithms, see PERM (Hsu and Grassberger 2011), REMC (Thachuk, Shmygelska, and Hoos 2007) or WLRE (W\"ust and Landau 2012) for better results. Hsu, H.-P.; and Grassberger, P. 2011. A review of Monte Carlo simulations of polymers with PERM. Journal of Statistical Physics, 144 (3): 597 to 637. Thachuk, C.; Shmygelska, A.; and Hoos, H. H. 2007. A replica exchange Monte Carlo algorithm for protein folding in the HP model. BMC Bioinformatics, 8(1): 342. W\"ust, T.; and Landau, D. P. 2012. Optimized Wang-Landau sampling of lattice polymers: Ground state search and folding thermodynamics of HP model proteins. The Journal of Chemical Physics, 137(6): 064903. | ['Tristan Cazenave', 'Milo Roucairol'] | 2023-01-23 | null | null | null | null | ['protein-folding'] | ['natural-language-processing'] | [ 2.44697690e-01 -2.36969993e-01 -7.51269311e-02 -6.78290948e-02
-4.52297240e-01 -4.03291941e-01 2.69934237e-01 3.55309665e-01
-2.99354047e-01 1.39799881e+00 4.39345799e-02 -5.80268621e-01
-8.85273814e-02 -7.30775654e-01 -1.01769435e+00 -1.22042656e+00
-4.89426464e-01 5.40893793e-01 3.66845131e-01 -1.07804336e-01
4.09385085e-01 5.55527091e-01 -1.49293649e+00 3.23609024e-01
1.04969454e+00 3.38218451e-01 5.87165833e-01 9.16325152e-01
2.38579810e-01 1.27551883e-01 2.72086877e-02 -8.52142498e-02
-1.10355698e-01 -9.37114060e-01 -7.48325706e-01 -7.35125899e-01
-2.27276199e-02 2.80163139e-01 3.19945335e-01 9.10874009e-01
8.20547700e-01 1.59426481e-01 1.02209020e+00 -6.39930725e-01
-3.38222712e-01 3.25627595e-01 -4.73764569e-01 6.25524372e-02
5.75473070e-01 3.92257899e-01 6.55111670e-01 -8.23206127e-01
9.76818919e-01 9.92762446e-01 7.63270497e-01 5.84894776e-01
-1.42005694e+00 -5.11217952e-01 -4.43633348e-01 4.28881109e-01
-1.26050317e+00 1.20815076e-01 2.91461766e-01 -4.01030064e-01
1.83276320e+00 5.44095576e-01 1.02066898e+00 7.62771130e-01
9.72449481e-01 3.04022461e-01 1.68022966e+00 -6.65058553e-01
6.80886447e-01 -6.32233322e-01 4.70150203e-01 6.37289703e-01
3.87887865e-01 4.97409135e-01 -5.64685166e-01 -1.00700033e+00
2.96433419e-01 -1.37357652e-01 2.18463853e-01 -4.74177003e-01
-9.73249972e-01 6.66192710e-01 7.77680278e-02 1.51815459e-01
-7.06259072e-01 1.32356673e-01 3.61116081e-01 -7.25253895e-02
1.71178862e-01 3.64829451e-02 -5.78595042e-01 -4.11121279e-01
-8.84197056e-01 8.97849619e-01 9.17260110e-01 6.22200131e-01
5.60966015e-01 -6.04295731e-01 5.73155917e-02 3.75730693e-01
4.80201721e-01 4.48286414e-01 2.32032716e-01 -9.23013151e-01
-1.87811688e-01 -7.82620683e-02 3.71091574e-01 -8.03599805e-02
-3.91667426e-01 2.48080641e-01 -8.56878579e-01 1.09361164e-01
4.24951464e-01 1.30118360e-03 -8.14971805e-01 1.41021955e+00
3.51346463e-01 -1.21970676e-01 1.39388844e-01 5.84014535e-01
6.06001198e-01 1.02748251e+00 3.99365783e-01 -9.06063974e-01
1.54378295e+00 -9.04040873e-01 -4.50740576e-01 3.86925846e-01
4.32471693e-01 -1.08588409e+00 6.01278841e-01 5.34871936e-01
-1.32860005e+00 -2.00208545e-01 -9.89292443e-01 1.58624396e-01
-3.73131156e-01 -2.71937460e-01 7.76791155e-01 8.31617236e-01
-1.06166148e+00 1.22439194e+00 -1.11500323e+00 -9.78531778e-01
-4.09386048e-05 3.96138787e-01 -2.40444429e-02 6.78381026e-02
-1.08829856e+00 1.13283408e+00 4.71892118e-01 -4.33983803e-01
-5.18090427e-01 -4.88960594e-01 -2.55224526e-01 -4.25949395e-01
-6.96257129e-02 -6.14991188e-01 8.03384364e-01 -3.01207006e-01
-1.66026008e+00 9.07587647e-01 -5.09354889e-01 -5.21034062e-01
1.70477569e-01 1.71619073e-01 -6.36768937e-02 1.23413540e-01
-3.31183344e-01 5.67805052e-01 -1.19618177e-01 -9.93950486e-01
2.59481609e-01 -4.26719159e-01 -6.04991972e-01 1.48269832e-01
6.66237235e-01 4.64332044e-01 4.88352239e-01 -4.17958587e-01
2.28511840e-01 -1.05065370e+00 -3.86730820e-01 -4.07065302e-01
-3.30262244e-01 -2.65257359e-01 5.99155277e-02 -7.37951696e-01
9.59632277e-01 -1.37735176e+00 1.25294417e-01 6.20723486e-01
-1.19247355e-01 3.16353232e-01 1.90213025e-01 1.32148015e+00
-5.27334213e-01 -7.65179396e-02 -8.03521454e-01 3.12290817e-01
-1.20804548e-01 3.32711637e-01 -6.65172637e-02 5.61957240e-01
-3.26268315e-01 8.11027348e-01 -7.96905518e-01 -3.60559434e-01
1.66737929e-01 5.47189176e-01 -6.19815767e-01 -2.20047027e-01
-4.43878740e-01 4.97865349e-01 -1.58552319e-01 5.65048039e-01
9.43884552e-01 -2.94482231e-01 7.37931073e-01 -1.51825547e-01
-7.18916714e-01 3.75135005e-01 -1.07776690e+00 1.21325636e+00
5.11580467e-01 -5.70065416e-02 -2.53032446e-01 -5.75777650e-01
9.46752429e-01 3.63909751e-01 7.20221639e-01 -2.40852833e-01
6.87831640e-02 5.80067158e-01 1.71031699e-01 -1.97698250e-01
2.26817340e-01 -4.54301119e-01 2.81479061e-01 5.89794099e-01
-2.54027367e-01 -1.66444197e-01 6.06532574e-01 3.92930768e-02
1.11595881e+00 8.17933977e-01 8.11474681e-01 -7.62557924e-01
4.30520445e-01 3.45615894e-01 4.39312965e-01 6.99329972e-01
-5.05388856e-01 3.23701650e-01 4.91073042e-01 -5.50418675e-01
-1.40299273e+00 -1.03323722e+00 -7.01599598e-01 8.42768490e-01
4.34049815e-02 -8.00380290e-01 -1.18270063e+00 1.04579501e-01
-9.05702189e-02 5.14695048e-01 -2.46049792e-01 -8.45885202e-02
-6.08580172e-01 -1.50515342e+00 4.78188932e-01 -1.99237466e-01
2.38691196e-01 -1.50264049e+00 -7.93123662e-01 4.82107878e-01
2.22977884e-02 -2.18111604e-01 -1.33638889e-01 3.37630719e-01
-1.22335041e+00 -1.10237277e+00 -7.28735387e-01 -4.65593487e-01
4.64872599e-01 -2.36068189e-01 9.79005814e-01 1.39980122e-01
-6.15159690e-01 -1.60953149e-01 -3.28827530e-01 -8.95415917e-02
-8.84959519e-01 -2.99422681e-01 2.71443486e-01 -8.43881428e-01
6.49954140e-01 -7.95531273e-01 -8.29270542e-01 1.29594952e-01
-5.72743237e-01 2.05693066e-01 6.14366353e-01 1.06381607e+00
1.02956080e+00 -3.14244807e-01 -1.02041751e-01 -5.09221315e-01
4.87303555e-01 -1.45079374e-01 -4.97782707e-01 2.47546807e-01
-9.53789234e-01 1.61841005e-01 4.43613231e-01 -3.43122818e-02
-6.92657173e-01 1.59442544e-01 -5.48462272e-01 4.72008169e-01
2.68403310e-02 2.47190282e-01 2.02353776e-01 -2.48622328e-01
7.09097564e-01 4.35236633e-01 9.79764462e-02 -7.95540094e-01
7.58766234e-02 3.11512113e-01 -2.11651638e-01 -7.60156095e-01
2.45878875e-01 1.57979563e-01 4.08461630e-01 -8.12119007e-01
-7.37593472e-02 -1.59682602e-01 -5.88856161e-01 -8.36873278e-02
1.04049337e+00 -3.03100675e-01 -1.02089977e+00 6.01072252e-01
-1.06358099e+00 -1.79390430e-01 -1.00528523e-02 6.16411328e-01
-9.15803730e-01 8.89863908e-01 -8.87531579e-01 -1.11800587e+00
-3.92293900e-01 -1.06340528e+00 7.27794707e-01 1.23109393e-01
-4.10049796e-01 -6.60537064e-01 8.16559911e-01 3.94199699e-01
8.95063654e-02 5.48009634e-01 1.03079581e+00 -3.28125954e-01
-4.01092142e-01 1.55712530e-01 1.33141786e-01 -1.58043250e-01
-4.22799200e-01 4.13885593e-01 -5.66548705e-01 -4.58553582e-01
-2.99289882e-01 1.28826022e-01 9.10241425e-01 8.00244808e-01
7.58040488e-01 -2.91590095e-01 -5.11252522e-01 2.05288589e-01
1.53209245e+00 5.34677207e-01 1.01645041e+00 5.28301537e-01
8.54092017e-02 3.82178307e-01 6.67542160e-01 3.39960873e-01
-1.24366745e-01 5.79242766e-01 1.88879520e-01 2.97982544e-01
2.09851101e-01 -3.26212138e-01 6.30020916e-01 7.75656998e-01
-8.07839215e-01 1.34901451e-02 -1.06560457e+00 -7.85931051e-02
-1.85001707e+00 -1.22735548e+00 -6.40226662e-01 2.15564346e+00
1.20948684e+00 -2.92951949e-02 4.69239771e-01 3.55636030e-02
7.49169290e-01 -2.99201645e-02 -4.23481286e-01 -8.80557001e-01
-1.97809845e-01 7.82511294e-01 7.98876643e-01 7.32960582e-01
-6.49190247e-01 7.30589151e-01 5.81110907e+00 8.89525294e-01
-5.02692401e-01 1.82656124e-01 3.65472108e-01 -4.51486707e-02
-1.46926656e-01 5.85774422e-01 -7.05215216e-01 8.64933729e-01
1.30202997e+00 -5.06252870e-02 4.32925850e-01 3.83716345e-01
4.21319693e-01 -6.52880907e-01 -5.16814590e-01 5.89719415e-01
-2.87252069e-01 -1.28326869e+00 -3.40234548e-01 3.36111128e-01
7.60681391e-01 1.62952736e-01 -3.98018211e-01 -1.36916324e-01
1.41572729e-01 -1.04453290e+00 2.90637672e-01 6.39067590e-01
4.63724464e-01 -7.65527070e-01 6.49535656e-01 3.97051305e-01
-8.81740272e-01 4.95072544e-01 -5.95635474e-01 -2.68391967e-01
2.24718407e-01 8.29809487e-01 -3.42013270e-01 5.90648711e-01
7.34402776e-01 3.62476021e-01 -7.58718252e-02 9.50993121e-01
1.77095234e-01 5.14644980e-01 -4.34955359e-01 -8.54766726e-01
-2.12104172e-02 -8.05832505e-01 5.22281528e-01 9.36520696e-01
2.91370571e-01 4.72597361e-01 -7.86717087e-02 6.50935590e-01
7.29352117e-01 3.36736351e-01 -1.51036605e-01 -1.86270460e-01
3.20526928e-01 6.35979593e-01 -1.01333952e+00 -3.07946056e-01
1.63972959e-01 7.70134628e-01 -9.79813933e-02 2.25785390e-01
-5.37625432e-01 -1.86371967e-01 5.62521458e-01 2.94559658e-01
3.85192961e-01 -2.67998517e-01 1.31321400e-02 -7.26838946e-01
-2.26112425e-01 -9.37018335e-01 2.44411975e-01 -6.85404539e-01
-1.22661698e+00 -6.45185262e-02 2.80817389e-01 -4.95002180e-01
1.17357299e-01 -9.66049850e-01 -4.85173285e-01 1.23028982e+00
-9.09668803e-01 -7.09721804e-01 5.15996397e-01 -3.69145311e-02
-7.01295584e-02 2.10240439e-01 9.21608031e-01 -2.65957147e-01
-2.80080587e-01 1.12331882e-01 1.04956746e+00 -8.83848965e-01
6.91435754e-01 -1.05869710e+00 6.67011142e-01 4.26969379e-01
-7.72177398e-01 1.06247842e+00 1.21253538e+00 -1.41467667e+00
-1.65762377e+00 -5.30666828e-01 1.23781693e+00 -3.67109150e-01
3.04692894e-01 -2.36227632e-01 -8.40190172e-01 1.86483845e-01
3.42293173e-01 -6.25889361e-01 9.80856717e-01 -2.76666164e-01
1.56760573e-01 4.12195265e-01 -1.50457644e+00 4.65362728e-01
9.38761830e-01 -2.64616698e-01 -4.47176933e-01 7.71799088e-01
3.30468059e-01 -3.50289315e-01 -1.14801741e+00 2.89821297e-01
9.37106013e-01 -1.33805251e+00 1.15470946e+00 -6.17859960e-01
2.46757850e-01 -3.17508370e-01 -1.56864271e-01 -6.80752039e-01
-3.71597707e-01 -7.75131881e-01 4.60523553e-02 5.65593302e-01
2.59977847e-01 -9.15052474e-01 5.96438706e-01 2.17849284e-01
-2.00691134e-01 -9.77537155e-01 -1.30628419e+00 -7.45828032e-01
6.75182462e-01 1.79497618e-02 5.20062804e-01 5.53885996e-01
5.02041399e-01 1.85972974e-02 -2.37207502e-01 -2.91309297e-01
9.89321113e-01 3.45454067e-02 1.64583385e-01 -1.10657299e+00
-6.58624828e-01 -1.19189806e-01 -1.80978894e-01 -5.70335150e-01
-2.39327908e-01 -1.02155030e+00 -1.23153858e-01 -1.69028580e+00
8.28752756e-01 1.22295776e-02 -1.07210398e-01 1.62524655e-01
2.53559221e-02 7.83759579e-02 -1.77715153e-01 3.95800114e-01
-2.47411981e-01 2.62597769e-01 1.16934526e+00 3.35335165e-01
-8.58849939e-03 -3.19577694e-01 -3.78043130e-02 2.41403148e-01
1.21492207e+00 -5.22065699e-01 1.35886490e-01 8.04412782e-01
6.51173413e-01 -3.07955407e-03 3.35966408e-01 -6.70312762e-01
-1.33365974e-01 -3.81355584e-01 2.95874059e-01 -9.58272696e-01
1.68099940e-01 -2.80280948e-01 8.02850068e-01 1.52557766e+00
1.36357412e-01 2.57255852e-01 -5.30327810e-03 6.02272570e-01
5.14916122e-01 -7.14322209e-01 1.20379102e+00 -5.82634330e-01
-9.09913145e-03 -9.69583690e-02 -9.13698971e-01 -4.94468123e-01
1.06530583e+00 -5.14427781e-01 -3.65599394e-01 2.13706493e-01
-8.18350554e-01 -5.68233663e-03 1.07453477e+00 -4.36729610e-01
3.49974573e-01 -9.72320557e-01 -2.76436478e-01 9.61540714e-02
-2.75970906e-01 -5.54859579e-01 3.76634240e-01 1.32098651e+00
-1.24499786e+00 8.36158991e-01 -3.53645384e-01 -3.29154611e-01
-1.39813876e+00 4.83904034e-01 2.50361145e-01 -3.93768311e-01
-2.63575643e-01 5.25208414e-01 -3.39263052e-01 -3.98989439e-01
-4.39600497e-01 -2.00626612e-01 1.31299809e-01 -2.26500437e-01
4.06917453e-01 8.05186808e-01 7.25117847e-02 -5.93023837e-01
-7.05642402e-01 5.63316941e-01 7.80108646e-02 2.94238515e-03
1.41915309e+00 1.41405225e-01 -7.59244680e-01 4.25353467e-01
8.25804293e-01 -2.61352509e-01 -6.64242506e-01 2.87917525e-01
-6.07191445e-03 -2.99545545e-02 -4.86338139e-01 -9.50706899e-01
7.56325051e-02 7.22127438e-01 9.65016723e-01 3.19580249e-02
6.91402614e-01 -1.31123969e-02 1.02154386e+00 2.62870461e-01
6.22672021e-01 -1.30771685e+00 -5.79822481e-01 5.64785838e-01
4.79474247e-01 -6.33303106e-01 4.54094648e-01 -5.99540114e-01
-2.58831948e-01 1.28121603e+00 -3.25887017e-02 -4.49562557e-02
5.14123023e-01 1.11063190e-01 -5.59879243e-01 -1.79397419e-01
-7.52358377e-01 -7.93303028e-02 -1.64430484e-01 3.96900564e-01
8.08184206e-01 4.47720587e-01 -1.46564424e+00 2.89417028e-01
-2.01156139e-01 1.37574270e-01 4.43928361e-01 1.45287013e+00
-8.09617996e-01 -1.71881485e+00 -6.72061265e-01 5.20809412e-01
-2.89457798e-01 -4.08372015e-01 -7.85297096e-01 3.02199423e-01
1.07591860e-01 8.15978169e-01 -5.03125250e-01 6.23975992e-02
1.06961362e-01 7.14862168e-01 1.08943558e+00 -1.41291663e-01
-8.03391397e-01 1.29096925e-01 2.54237205e-01 -2.88295478e-01
-7.36419797e-01 -1.05791509e+00 -1.53042924e+00 -9.21180129e-01
-3.44022840e-01 8.45191181e-01 6.73323572e-01 7.62134731e-01
4.19006139e-01 6.86030313e-02 2.30766043e-01 -7.96322525e-01
-4.91449267e-01 -8.29269826e-01 -9.47178602e-01 -2.53862411e-01
-3.90793830e-02 -7.28475094e-01 -3.97618800e-01 -9.18638036e-02] | [4.734993934631348, 5.2795634269714355] |
707d2dfb-df64-49c4-936c-d15d624d6132 | instance-aware-hashing-for-multi-label-image | 1603.03234 | null | http://arxiv.org/abs/1603.03234v1 | http://arxiv.org/pdf/1603.03234v1.pdf | Instance-Aware Hashing for Multi-Label Image Retrieval | Similarity-preserving hashing is a commonly used method for nearest neighbour
search in large-scale image retrieval. For image retrieval, deep-networks-based
hashing methods are appealing since they can simultaneously learn effective
image representations and compact hash codes. This paper focuses on
deep-networks-based hashing for multi-label images, each of which may contain
objects of multiple categories. In most existing hashing methods, each image is
represented by one piece of hash code, which is referred to as semantic
hashing. This setting may be suboptimal for multi-label image retrieval. To
solve this problem, we propose a deep architecture that learns
\textbf{instance-aware} image representations for multi-label image data, which
are organized in multiple groups, with each group containing the features for
one category. The instance-aware representations not only bring advantages to
semantic hashing, but also can be used in category-aware hashing, in which an
image is represented by multiple pieces of hash codes and each piece of code
corresponds to a category. Extensive evaluations conducted on several benchmark
datasets demonstrate that, for both semantic hashing and category-aware
hashing, the proposed method shows substantial improvement over the
state-of-the-art supervised and unsupervised hashing methods. | ['Yunchao Wei', 'Xiangbo Shu', 'Shuicheng Yan', 'Pan Yan', 'Hanjiang Lai'] | 2016-03-10 | null | null | null | null | ['multi-label-image-retrieval'] | ['computer-vision'] | [-3.06339506e-02 -2.90051132e-01 -5.79333425e-01 -4.79475737e-01
-1.23527539e+00 -5.39674461e-01 3.02408785e-01 7.63320804e-01
-4.78440821e-01 3.60783756e-01 1.57232344e-01 3.21028262e-01
-1.89056650e-01 -8.84764850e-01 -6.51816607e-01 -1.13015604e+00
-5.86446561e-02 6.70671403e-01 1.26973033e-01 2.23837748e-01
3.62651825e-01 3.96274745e-01 -1.96277499e+00 3.62114966e-01
-7.90509433e-02 1.27079415e+00 7.25235566e-02 2.86536485e-01
3.59377749e-02 6.75366938e-01 -4.55145150e-01 -1.72694176e-01
8.14811215e-02 -1.13520220e-01 -8.31237793e-01 -3.72242749e-01
6.43949687e-01 -7.08339751e-01 -8.17432106e-01 1.01863074e+00
6.61056280e-01 1.56905144e-01 8.67105961e-01 -1.45618522e+00
-9.61171746e-01 5.32053113e-01 -5.90805173e-01 -1.77754108e-02
2.09131986e-01 -2.72339880e-01 1.45142102e+00 -9.27378416e-01
3.01862836e-01 1.44052696e+00 7.01967597e-01 4.19971585e-01
-1.01899254e+00 -7.98790336e-01 -4.77245837e-01 5.02409041e-01
-2.18732643e+00 -1.83495447e-01 5.17275691e-01 -1.38159260e-01
7.79081583e-01 -8.46591890e-02 4.80015785e-01 4.69427496e-01
1.93746492e-01 9.96792078e-01 7.32450783e-01 -1.64648458e-01
1.49918795e-01 4.20655534e-02 1.37378722e-01 7.66946018e-01
2.48690620e-01 -5.04941214e-03 -5.71782172e-01 -5.09725988e-01
5.96907556e-01 6.87357187e-01 -5.88338189e-02 -6.39542758e-01
-1.26482058e+00 1.52621162e+00 1.02991748e+00 3.05635422e-01
-2.89482564e-01 6.67385936e-01 8.50648165e-01 1.34081006e-01
2.79318616e-02 1.14054702e-01 3.86213548e-02 3.89829338e-01
-1.07378674e+00 5.07218599e-01 4.51756358e-01 1.01032794e+00
1.26896691e+00 -5.76298058e-01 -3.83732855e-01 1.08784211e+00
4.24708515e-01 6.49722695e-01 8.26427639e-01 -7.91031659e-01
8.78616869e-02 6.39603972e-01 -1.93953842e-01 -1.33860195e+00
-3.38063210e-01 2.53100321e-02 -1.12447405e+00 -5.22266686e-01
-8.92082676e-02 7.56209075e-01 -1.00481367e+00 1.75620127e+00
3.09992760e-01 1.92561358e-01 1.65334193e-03 8.90492857e-01
1.15331483e+00 8.79824758e-01 9.52426493e-02 1.32278815e-01
1.69833100e+00 -1.08415747e+00 -5.48594594e-01 9.46741700e-02
7.42568374e-01 -6.58342302e-01 6.66984081e-01 -8.83919373e-02
-8.18165421e-01 -5.72182715e-01 -9.06668246e-01 -3.69822115e-01
-6.84337080e-01 -2.65841305e-01 3.59672457e-01 3.45703572e-01
-1.32906079e+00 2.91976094e-01 -1.80548847e-01 -2.99352556e-01
4.07209039e-01 4.78280663e-01 -5.84526062e-01 -6.98440909e-01
-1.47611904e+00 4.95851487e-01 7.87742257e-01 -4.19307113e-01
-1.17093694e+00 -2.59036541e-01 -1.27237761e+00 4.04437274e-01
-1.86448112e-01 -3.50543231e-01 1.08428383e+00 -2.11457446e-01
-4.19706434e-01 1.12492514e+00 -1.25114217e-01 -4.72064674e-01
-3.63006711e-01 2.57793486e-01 -2.53594041e-01 8.34282100e-01
4.55765843e-01 1.29639268e+00 1.01564944e+00 -1.19827843e+00
-6.64780855e-01 -3.17252725e-01 6.22576661e-02 2.10153699e-01
-8.25069845e-01 2.70342268e-02 -3.60317200e-01 -6.68833733e-01
2.53499508e-01 -1.05929744e+00 3.40745710e-02 3.75064462e-01
-3.42561692e-01 -5.84790885e-01 1.02841556e+00 -1.77659109e-01
1.30066979e+00 -2.33680630e+00 7.92573299e-03 1.97466493e-01
2.69722819e-01 5.33180172e-03 -3.21336687e-01 5.91294825e-01
1.05283171e-01 5.69287986e-02 -1.70116618e-01 -6.10060751e-01
2.99520403e-01 4.11948651e-01 -3.12037498e-01 6.95484638e-01
-1.95009679e-01 1.06010640e+00 -7.98568726e-01 -9.45189834e-01
1.12681262e-01 8.07551801e-01 -3.61837983e-01 2.28925839e-01
2.77151376e-01 -3.04846793e-01 -3.19635689e-01 8.89885306e-01
6.85354829e-01 -6.38318837e-01 -2.06192896e-01 -3.86014521e-01
2.96423197e-01 -8.77820775e-02 -1.06045735e+00 1.80470121e+00
-3.32065642e-01 3.92128170e-01 -4.08689082e-01 -9.37759280e-01
8.01707923e-01 4.13873762e-01 4.69122708e-01 -7.44620740e-01
-3.40987816e-02 3.22059929e-01 -7.06244826e-01 -5.76253831e-02
7.48215735e-01 -1.72896713e-01 -4.61215973e-01 8.17330897e-01
1.90471992e-01 -2.72638369e-02 7.86975920e-02 2.77408898e-01
8.98839474e-01 -8.11132073e-01 3.47888649e-01 -9.70083252e-02
4.88402307e-01 -3.33907098e-01 3.54782403e-01 9.99729216e-01
-3.32518607e-01 8.24197114e-01 -1.20028354e-01 -8.01716387e-01
-1.14386761e+00 -9.58800793e-01 -4.17036057e-01 1.24213123e+00
6.56572700e-01 -5.13447344e-01 -3.91331464e-01 -4.99342024e-01
4.59740102e-01 -2.44862288e-01 -5.69351017e-01 -5.43232501e-01
-3.52436841e-01 -5.06622314e-01 7.32406974e-01 5.51406503e-01
5.27608633e-01 -1.26077795e+00 -5.44565678e-01 2.26290464e-01
-4.83962148e-01 -7.55858004e-01 -8.16786826e-01 3.25275272e-01
-5.61167002e-01 -1.04556835e+00 -1.14897048e+00 -1.40021050e+00
5.43542743e-01 9.05928791e-01 1.02551389e+00 5.48849881e-01
-6.37310445e-01 4.23257589e-01 -4.50818866e-01 1.73909456e-01
-1.38419852e-01 2.03991875e-01 1.62931949e-01 -6.69306610e-03
6.32668674e-01 -2.71656811e-01 -1.07497668e+00 4.73311126e-01
-1.44838858e+00 -6.28416657e-01 6.14611268e-01 1.10173655e+00
9.71762300e-01 2.83922732e-01 4.38151389e-01 -3.20839167e-01
2.42552713e-01 -6.13834262e-01 -1.88936427e-01 3.07017207e-01
-4.85243410e-01 9.63233858e-02 4.63448495e-01 -4.11122233e-01
-2.85048783e-02 1.38627276e-01 1.52480394e-01 -6.07037783e-01
-1.65299773e-01 5.86797476e-01 1.70115992e-01 -5.35417378e-01
3.63879502e-01 6.02880895e-01 9.92684439e-02 -2.62467504e-01
4.90227669e-01 9.33912694e-01 3.40476304e-01 -3.72540593e-01
7.78678298e-01 5.67184925e-01 9.95220914e-02 -5.01896620e-01
-8.41035962e-01 -1.10526633e+00 -5.84581733e-01 8.23659450e-02
9.36700284e-01 -1.36026573e+00 -7.68811941e-01 6.43286288e-01
-1.05226016e+00 2.26378053e-01 5.84326275e-02 2.22043052e-01
-6.36323035e-01 5.58097363e-01 -1.12932873e+00 -2.28022441e-01
-5.36905110e-01 -1.38180220e+00 1.74880278e+00 3.13631117e-01
9.12837535e-02 -9.33625996e-01 1.10018410e-01 2.57192731e-01
4.23245311e-01 -1.49606451e-01 1.17257142e+00 -6.77012086e-01
-5.55684686e-01 -6.12638175e-01 -5.80213249e-01 1.32572070e-01
-2.90865321e-02 -5.58748484e-01 -9.51724172e-01 -9.28636909e-01
-3.51180166e-01 -1.06716919e+00 9.85524535e-01 2.01215535e-01
1.46240532e+00 -5.89177012e-01 -3.95563781e-01 3.65407765e-01
1.75657332e+00 -5.93441874e-02 5.89780092e-01 4.04086828e-01
8.12682569e-01 3.01225632e-01 6.78800762e-01 6.96046412e-01
7.82283247e-01 9.74103808e-01 6.36540353e-01 -9.30786431e-02
1.08962275e-01 -2.51302838e-01 -1.33840710e-01 8.47002625e-01
8.59588385e-01 -1.94868773e-01 -8.24080825e-01 8.88582945e-01
-1.80153668e+00 -9.50154305e-01 4.09253031e-01 2.03367949e+00
9.45464313e-01 -3.81945997e-01 2.92109430e-01 2.22597718e-01
1.17678654e+00 4.05084759e-01 -5.43084443e-01 -1.14769809e-01
-4.12447192e-02 -5.06910728e-03 4.22334135e-01 -3.51456106e-02
-1.53769374e+00 7.02178180e-01 6.18057680e+00 1.14934123e+00
-8.85488808e-01 9.29952338e-02 4.54352409e-01 1.48649514e-01
-1.91930652e-01 -1.67602524e-01 -1.09612179e+00 5.58484197e-01
1.00657427e+00 2.06934046e-02 2.01757208e-01 1.05092120e+00
-6.66027784e-01 4.44790274e-02 -1.17582750e+00 1.38561141e+00
3.86605501e-01 -1.33586299e+00 6.50672555e-01 2.37760454e-01
6.24008715e-01 -1.32056132e-01 3.77972633e-01 3.66926998e-01
4.85992990e-02 -1.06719613e+00 6.26135111e-01 -4.30566519e-02
9.99067068e-01 -1.02722692e+00 9.38127875e-01 1.20513374e-02
-1.74603748e+00 -3.69119436e-01 -8.90675843e-01 5.24564981e-01
-1.22740656e-01 3.79801005e-01 -5.18221378e-01 3.06592733e-01
9.50162053e-01 9.32565510e-01 -5.30554235e-01 1.26882863e+00
2.74939150e-01 3.82559970e-02 -2.05424979e-01 1.30776986e-01
5.76571643e-01 3.64009142e-01 -1.88040420e-01 1.20230138e+00
4.46557909e-01 -6.08778223e-02 5.49863577e-01 4.98153597e-01
-4.10048425e-01 1.63806260e-01 -9.85160708e-01 -1.08918017e-02
1.03313243e+00 1.22591925e+00 -7.28861690e-01 -4.43446189e-01
-4.57152039e-01 8.93456876e-01 3.47695410e-01 -1.46450372e-02
-7.31045485e-01 -7.47177899e-01 7.61191249e-01 -3.09617251e-01
6.12708747e-01 9.76297036e-02 4.13585722e-01 -8.23669493e-01
-2.01394156e-01 -7.64246762e-01 8.86753559e-01 -5.96204638e-01
-1.28950047e+00 4.01333719e-01 -1.06959507e-01 -1.36064315e+00
-2.15047181e-01 -2.62129486e-01 -2.58936360e-03 4.41049218e-01
-1.81393456e+00 -1.36984706e+00 -3.48965228e-01 8.55664849e-01
2.32795089e-01 -2.03637928e-01 1.26463664e+00 6.64610207e-01
-9.33934376e-02 1.06615353e+00 6.16872847e-01 3.07369351e-01
9.76503730e-01 -1.02613997e+00 1.27079248e-01 1.22906879e-01
2.82417864e-01 7.17295468e-01 1.56191528e-01 -2.01087013e-01
-1.38975823e+00 -1.32357216e+00 8.52901995e-01 1.24793313e-02
5.76412082e-01 -2.73702741e-01 -1.07510006e+00 3.59746873e-01
-1.16643451e-01 5.05002916e-01 1.12004209e+00 -2.92357951e-01
-1.07181811e+00 -2.54725695e-01 -1.44423914e+00 6.15194514e-02
5.09499729e-01 -1.21544158e+00 -4.22798842e-01 5.80907583e-01
9.66530323e-01 -8.94318894e-02 -1.18495715e+00 1.83240190e-01
6.07335329e-01 -7.78747976e-01 1.37292016e+00 -1.13328747e-01
3.78467798e-01 -4.47392076e-01 -5.84619761e-01 -9.27852213e-01
-5.77277124e-01 1.65313900e-01 -2.38790363e-01 1.01018703e+00
-1.09139532e-01 -3.89624238e-01 5.30897737e-01 1.86710551e-01
1.84417501e-01 -6.51922405e-01 -1.11894941e+00 -8.66691232e-01
9.01332796e-02 3.72905731e-01 8.95192504e-01 9.21340168e-01
-2.32322350e-01 -1.07874401e-01 -5.26143491e-01 1.21336266e-01
9.88282084e-01 4.55820054e-01 3.45880389e-01 -1.32187140e+00
7.43766800e-02 -3.47914100e-01 -9.90770698e-01 -9.19055760e-01
4.21475261e-01 -1.03339815e+00 4.26892549e-01 -1.35348248e+00
9.11956906e-01 -5.73930860e-01 -8.50996554e-01 8.86123538e-01
3.19336094e-02 9.81232464e-01 2.03147143e-01 7.45529413e-01
-1.21151483e+00 5.83618522e-01 7.67575741e-01 -6.39507532e-01
5.03508747e-01 -4.40184861e-01 -6.07369304e-01 6.74850866e-02
4.40678477e-01 -8.41159523e-01 -2.59805143e-01 -3.27466249e-01
5.34185991e-02 1.20507963e-01 5.23806810e-01 -1.04940140e+00
5.52100539e-01 3.34951490e-01 1.33532792e-01 -8.12628984e-01
4.60445285e-01 -8.68605971e-01 5.53173991e-03 6.40865564e-01
-6.36065662e-01 1.20361298e-01 -1.19001061e-01 7.47976601e-01
-7.50469625e-01 -4.03335184e-01 1.08961439e+00 -2.43139967e-01
-8.70315015e-01 7.20365167e-01 -1.24426931e-01 -2.02946275e-01
8.84849072e-01 -1.16726890e-01 -4.38689947e-01 -4.68417525e-01
-3.05817038e-01 2.23478124e-01 6.47705197e-01 4.17627424e-01
1.01170564e+00 -2.06849837e+00 -3.74779373e-01 6.59050420e-02
8.62911522e-01 -8.64616483e-02 3.56526971e-01 1.82487160e-01
-3.30736011e-01 7.38081992e-01 -2.34980926e-01 -7.55466402e-01
-1.30773306e+00 9.94660556e-01 3.13371308e-02 -1.60151616e-01
-4.29287940e-01 9.44805861e-01 4.79003340e-01 -3.34010333e-01
5.22430182e-01 1.80452451e-01 -2.18425542e-01 2.74534643e-01
9.50088620e-01 -4.93658371e-02 -1.57388914e-02 -1.06200194e+00
-4.22673434e-01 1.05439067e+00 -6.00042760e-01 3.68678391e-01
1.10148954e+00 -2.77876377e-01 -6.13114536e-01 3.88638914e-01
2.09978819e+00 -1.05417895e+00 -6.33928895e-01 -6.83375597e-01
-7.63756037e-02 -4.53933120e-01 5.64907454e-02 -2.05890372e-01
-1.25232136e+00 8.37340713e-01 9.24507499e-01 4.02938873e-02
1.03297544e+00 2.61045277e-01 1.32784438e+00 7.02129781e-01
8.19189966e-01 -9.04084086e-01 5.19537747e-01 4.78356749e-01
5.12253821e-01 -1.50812685e+00 -1.74994975e-01 1.64959684e-01
-2.80746251e-01 9.50074196e-01 3.48190546e-01 7.21758232e-03
6.97841108e-01 -2.50327080e-01 -5.64733297e-02 -4.84337091e-01
-3.90140533e-01 -7.91630223e-02 2.56018132e-01 3.57151657e-01
1.13086730e-01 1.04070920e-02 4.50925250e-03 8.86815637e-02
2.39554346e-01 -3.73185128e-01 9.83230695e-02 1.12534070e+00
-8.24811161e-01 -1.00995302e+00 -4.50451195e-01 4.25795853e-01
-2.98970580e-01 -2.82177478e-01 3.53433043e-02 4.89292294e-01
-1.97890610e-01 7.34394908e-01 1.47222087e-01 -4.06203210e-01
-2.22822398e-01 -1.05474442e-01 1.96022630e-01 -6.94758475e-01
-4.58104044e-01 -2.80765831e-01 -6.87800586e-01 -5.31863213e-01
-3.64413768e-01 -4.12034512e-01 -1.38525152e+00 -3.87742639e-01
-2.93626219e-01 3.56076032e-01 5.30178607e-01 5.99803150e-01
2.97799796e-01 -2.50853658e-01 9.82451797e-01 -9.83717501e-01
-5.46616495e-01 -5.12561619e-01 -9.46458101e-01 7.15431988e-01
6.85846567e-01 -9.31649506e-01 -5.66895008e-01 -2.63655126e-01] | [11.349483489990234, 0.9395710825920105] |
700c29db-f701-4eaa-a1c3-7e7f42752dc5 | why-should-i-trust-you-bellman-the-bellman | 2201.12417 | null | https://arxiv.org/abs/2201.12417v2 | https://arxiv.org/pdf/2201.12417v2.pdf | Why Should I Trust You, Bellman? The Bellman Error is a Poor Replacement for Value Error | In this work, we study the use of the Bellman equation as a surrogate objective for value prediction accuracy. While the Bellman equation is uniquely solved by the true value function over all state-action pairs, we find that the Bellman error (the difference between both sides of the equation) is a poor proxy for the accuracy of the value function. In particular, we show that (1) due to cancellations from both sides of the Bellman equation, the magnitude of the Bellman error is only weakly related to the distance to the true value function, even when considering all state-action pairs, and (2) in the finite data regime, the Bellman equation can be satisfied exactly by infinitely many suboptimal solutions. This means that the Bellman error can be minimized without improving the accuracy of the value function. We demonstrate these phenomena through a series of propositions, illustrative toy examples, and empirical analysis in standard benchmark domains. | ['Shixiang Shane Gu', 'Ofir Nachum', 'Doina Precup', 'David Meger', 'Scott Fujimoto'] | 2022-01-28 | null | null | null | null | ['value-prediction'] | ['computer-code'] | [-1.20712988e-01 3.55072349e-01 -4.57233816e-01 -2.09537894e-01
-8.48531425e-01 -7.94201255e-01 3.48742247e-01 3.07888508e-01
-4.84971344e-01 1.19952559e+00 -1.31232545e-01 -3.68257582e-01
-4.66278881e-01 -4.71438169e-01 -5.63548148e-01 -8.24026942e-01
-5.73159009e-02 3.93657714e-01 1.06489502e-01 -2.70705044e-01
5.63338518e-01 4.62615550e-01 -1.07131040e+00 -2.70299464e-01
7.17240930e-01 1.53490543e+00 -4.35768545e-01 7.03109324e-01
1.61811724e-01 8.77708375e-01 -6.94504738e-01 -2.74550915e-01
5.18325984e-01 -6.73474908e-01 -8.68367970e-01 -2.74335474e-01
1.49433374e-01 -3.70873749e-01 -3.12475830e-01 1.33399880e+00
1.56112447e-01 2.31682047e-01 5.62362134e-01 -1.50702167e+00
-2.40901396e-01 2.63208956e-01 -1.95031792e-01 6.43664300e-02
3.58129740e-01 7.45089874e-02 1.40330863e+00 -3.09471637e-01
5.29637992e-01 7.09982574e-01 9.22235727e-01 3.79540622e-01
-1.34352374e+00 -1.69611424e-01 -7.67138526e-02 -1.01450652e-01
-1.54446077e+00 -4.54943269e-01 2.95163482e-01 -3.59669000e-01
7.75215566e-01 3.23637694e-01 6.66611135e-01 1.98095053e-01
5.64226270e-01 2.75254458e-01 9.25769389e-01 -2.06073463e-01
4.79497343e-01 2.77822644e-01 2.95265079e-01 6.32678866e-01
4.24799681e-01 4.00262535e-01 -1.10274360e-01 -2.57879138e-01
8.54966164e-01 -2.11761191e-01 -3.90006334e-01 -5.30023456e-01
-1.02440488e+00 6.73254132e-01 3.79151493e-01 3.45430851e-01
-3.46883863e-01 4.80993032e-01 9.46483314e-02 6.16995692e-01
5.79545051e-02 8.19417417e-01 -4.98423576e-01 -4.56096292e-01
-7.36191869e-01 6.17361307e-01 1.08426344e+00 5.45474350e-01
5.39275706e-01 -1.65048912e-01 -1.52103469e-01 2.05975890e-01
1.21249951e-01 5.06675780e-01 3.17214489e-01 -1.40640974e+00
2.61765748e-01 4.86526579e-01 9.27623749e-01 -6.75509393e-01
-5.56950808e-01 -5.81265807e-01 -4.28041518e-01 2.27000520e-01
1.43961358e+00 -4.61589664e-01 -3.84936333e-01 1.80786228e+00
1.65827855e-01 -1.12420581e-01 1.99617937e-01 9.59554791e-01
1.84078053e-01 6.29297793e-01 -3.76108736e-01 -6.39903486e-01
8.66385162e-01 -5.05438030e-01 -7.47775793e-01 -9.10841525e-02
8.17361176e-01 -5.39072156e-01 6.91144645e-01 2.88734317e-01
-1.26082861e+00 -1.42781883e-01 -1.06524694e+00 1.79318607e-01
9.28474590e-02 -4.69673038e-01 2.84079254e-01 4.85442549e-01
-8.83626759e-01 9.64434564e-01 -5.88977396e-01 -2.20990814e-02
-6.09168001e-02 4.78632569e-01 -3.30197424e-01 1.43706009e-01
-1.10736024e+00 1.27722120e+00 2.11186156e-01 9.02755558e-02
-6.19657993e-01 -6.99661076e-01 -4.23586488e-01 3.77623528e-01
5.21794736e-01 -1.61453709e-01 1.45924139e+00 -1.05132604e+00
-1.42836475e+00 3.99652541e-01 -6.94025606e-02 -5.73057890e-01
7.12732553e-01 5.67261837e-02 -9.51509401e-02 -9.91912484e-02
-2.56489575e-01 4.37121354e-02 4.90310818e-01 -1.12383950e+00
-4.50455278e-01 -3.36176485e-01 2.75600225e-01 -1.56851158e-01
3.28969330e-01 -3.88105363e-01 9.85526666e-02 -4.38957922e-02
3.02191794e-01 -9.53329802e-01 -3.16547692e-01 -6.69103637e-02
-1.69006616e-01 -7.75105432e-02 2.51338810e-01 -5.18188477e-01
1.38508213e+00 -2.10039067e+00 -2.03119591e-01 4.88264292e-01
3.38406600e-02 1.80409908e-01 1.20879733e-03 5.57884157e-01
-1.46132549e-02 9.95290205e-02 -2.66341269e-01 1.58010855e-01
1.58740669e-01 2.84791946e-01 -1.45815209e-01 8.61332297e-01
4.50745523e-02 1.03214240e+00 -9.22145128e-01 -2.35400751e-01
-2.74124295e-02 7.00144982e-03 -6.50539637e-01 2.11077612e-02
-1.33876175e-01 2.94398993e-01 -3.66449237e-01 2.41107177e-02
4.83839959e-01 -2.92015791e-01 4.93134111e-01 1.45506456e-01
-3.01871330e-01 4.36012387e-01 -1.38235104e+00 1.10182464e+00
-2.46440932e-01 6.49953365e-01 -2.51726415e-02 -1.27659750e+00
7.85330474e-01 3.29861104e-01 7.92523801e-01 -6.18394911e-01
3.42423409e-01 3.41581643e-01 3.12547833e-01 -2.29731247e-01
2.88727343e-01 -8.32997620e-01 -1.43332928e-01 4.28737730e-01
-2.04164922e-01 -2.17914060e-01 2.79806316e-01 -1.67824060e-01
1.04938078e+00 -1.12190180e-01 4.94745195e-01 -4.16215926e-01
3.81973863e-01 1.25176400e-01 6.42292559e-01 8.69006455e-01
-5.08340359e-01 3.95758569e-01 1.46383989e+00 -4.58571285e-01
-1.05741000e+00 -9.57334340e-01 -4.83788639e-01 3.92797023e-01
4.19381291e-01 -3.10406357e-01 -7.67428458e-01 -6.52381122e-01
3.12949657e-01 6.75107479e-01 -7.00338066e-01 -3.67428869e-01
-3.06173205e-01 -5.83668649e-01 4.12936896e-01 4.37802762e-01
3.68349433e-01 -4.28913265e-01 -6.66360497e-01 1.51989341e-01
1.06789237e-02 -1.01681840e+00 -4.32395041e-01 1.81536168e-01
-7.83878148e-01 -1.08698320e+00 -4.42905158e-01 -1.56771943e-01
3.51633221e-01 -2.02663556e-01 9.42321777e-01 1.79307327e-01
3.93193722e-01 1.41808510e-01 6.08495772e-02 -7.92427734e-02
-6.43736720e-01 -2.38081962e-01 -5.79757988e-02 -6.59943745e-02
-3.37561360e-03 -2.24011704e-01 -4.67643470e-01 5.40136158e-01
-4.00439441e-01 -6.03775978e-01 2.83252537e-01 8.75739992e-01
4.69460636e-01 2.52098769e-01 7.20769882e-01 -4.63750660e-01
5.02674580e-01 -3.29084784e-01 -1.32196653e+00 2.42824495e-01
-9.64690983e-01 5.79286993e-01 7.29820132e-01 -2.20274642e-01
-7.04600036e-01 -7.91390520e-03 8.68834779e-02 -1.11055039e-01
2.62876809e-01 6.03733420e-01 2.10428879e-01 -1.08175471e-01
3.83089989e-01 -2.49652416e-02 3.16384256e-01 -4.32914853e-01
1.56051770e-01 5.10723710e-01 6.48381650e-01 -5.74074388e-01
2.72664249e-01 2.17527822e-01 6.15402222e-01 -6.25339746e-01
-1.15043151e+00 -1.86866999e-01 -5.66707492e-01 -1.88660279e-01
6.00508809e-01 -3.55993241e-01 -1.42538631e+00 1.02567397e-01
-9.62214708e-01 -3.29732448e-01 -6.42764807e-01 5.26778817e-01
-7.15613961e-01 2.92447001e-01 -4.18401897e-01 -1.40138853e+00
1.67965159e-01 -1.06989741e+00 5.77317476e-01 1.41660422e-01
-2.99763560e-01 -1.09104335e+00 1.51333690e-01 -3.38278785e-02
1.87264383e-01 3.12466115e-01 9.56647575e-01 -5.51494837e-01
-2.53379494e-01 -8.77284467e-01 -2.43980363e-01 5.95413983e-01
-2.26217166e-01 -1.56357829e-02 -6.14376843e-01 -1.58362687e-01
1.84108555e-01 7.57463127e-02 5.26852310e-01 5.04831195e-01
5.95259190e-01 -4.69513834e-01 6.51029358e-03 2.05029801e-01
1.50064003e+00 3.19344848e-01 5.38422823e-01 1.02145322e-01
2.10257858e-01 4.03244376e-01 4.76938754e-01 7.07069039e-01
2.75022358e-01 7.44283795e-01 3.46985281e-01 3.81823629e-01
3.77723038e-01 -2.45515242e-01 4.46508825e-01 3.37938249e-01
-1.96677938e-01 8.70745257e-02 -8.29497337e-01 3.16392660e-01
-2.08908749e+00 -1.15134001e+00 -3.81021649e-01 2.92883635e+00
7.56036639e-01 3.83189291e-01 4.82416064e-01 1.94744036e-01
3.35999638e-01 -2.85364427e-02 -7.61765420e-01 -5.68977892e-01
2.02494219e-01 6.54268414e-02 9.65106964e-01 8.21797669e-01
-5.71967781e-01 3.78509194e-01 8.09603214e+00 4.51942116e-01
-9.39013302e-01 9.52179953e-02 3.28819424e-01 -2.54368857e-02
-1.16062015e-01 2.62795717e-01 -4.69506085e-01 5.95370412e-01
1.03855312e+00 -6.02462113e-01 5.77623308e-01 5.93726575e-01
1.60339192e-01 -4.68777686e-01 -1.19365942e+00 6.88917994e-01
-4.66675222e-01 -1.10920954e+00 -7.55570769e-01 2.15799570e-01
8.79250526e-01 -2.21017420e-01 -1.28340364e-01 8.13572779e-02
2.75547534e-01 -1.13086367e+00 7.87198365e-01 4.78771865e-01
5.71682632e-01 -8.63867939e-01 1.00391567e+00 4.42702323e-01
-8.58801961e-01 -3.36029083e-01 -2.06130534e-01 -5.08379340e-01
-1.78734213e-02 5.72231770e-01 -4.91430044e-01 3.12120557e-01
-5.68228289e-02 3.74143660e-01 6.43499047e-02 1.09442508e+00
-8.28864798e-03 2.59696245e-01 -5.79150140e-01 -1.53753370e-01
4.11787719e-01 -4.78108078e-01 2.84441233e-01 5.79596400e-01
5.16334176e-02 3.06166589e-01 -8.16120505e-02 9.54604864e-01
1.31689712e-01 -3.05914044e-01 -2.33304262e-01 -4.13306624e-01
4.21861678e-01 6.83047771e-01 -4.69412178e-01 -2.81260937e-01
-5.38000047e-01 4.41101879e-01 3.01788568e-01 5.54424644e-01
-7.13702977e-01 -4.48086143e-01 1.04131937e+00 7.09607527e-02
1.68872640e-01 -3.08372527e-01 -6.27782941e-01 -1.03847182e+00
9.29849446e-02 -5.45625448e-01 2.86701858e-01 -3.05445939e-01
-9.49042022e-01 7.10096285e-02 -1.16040505e-01 -1.07142437e+00
-5.05144119e-01 -7.04677701e-01 -3.25796425e-01 7.15534806e-01
-1.06998420e+00 4.34285216e-02 3.09015483e-01 2.58991867e-01
-2.99229175e-01 3.68254632e-01 6.11420631e-01 8.78334939e-02
-6.45557165e-01 5.58871388e-01 5.38950086e-01 5.83448932e-02
1.49679273e-01 -1.17947328e+00 1.00443363e-02 5.78109741e-01
-7.76313320e-02 4.71871346e-01 1.25055754e+00 -3.42378497e-01
-1.64320338e+00 -4.30049628e-01 7.47706115e-01 -5.94118476e-01
8.74918878e-01 1.50482252e-01 -7.65966654e-01 8.00690174e-01
-3.93484294e-01 2.11565271e-01 1.38047621e-01 1.56725064e-01
-1.76412731e-01 -9.53207463e-02 -1.30837691e+00 1.58767059e-01
5.18050611e-01 -4.59219217e-01 -3.19718450e-01 2.13255942e-01
2.99962610e-01 -3.20065260e-01 -1.12596989e+00 2.11536869e-01
1.06789374e+00 -1.12908697e+00 6.42090678e-01 -1.07914221e+00
3.84804606e-01 -2.39766657e-01 -5.37021995e-01 -1.10106814e+00
-4.84836474e-02 -7.71393776e-01 -1.57488286e-01 6.53816283e-01
5.06895304e-01 -8.49739611e-01 5.75364709e-01 1.22098768e+00
2.96145409e-01 -1.12475479e+00 -1.32412934e+00 -1.20017993e+00
4.61618245e-01 -3.50698888e-01 6.26862764e-01 7.85867214e-01
5.22631586e-01 2.47146398e-01 -3.39951545e-01 1.70519818e-02
4.87214983e-01 1.76386267e-01 4.45342124e-01 -1.06346834e+00
-4.99622494e-01 -6.53426647e-01 -5.40399253e-01 -1.37804496e+00
-3.31498198e-02 -5.85741520e-01 1.48611948e-01 -1.32547605e+00
5.07108681e-03 -5.76611340e-01 -3.60873520e-01 1.94338486e-01
1.22128278e-01 -1.17533080e-01 4.18729991e-01 7.24328682e-02
-4.37937081e-01 1.37973160e-01 1.40611923e+00 2.97155589e-01
-1.64767548e-01 1.79585353e-01 -5.33466876e-01 7.67775178e-01
6.47703409e-01 -5.95343709e-01 -4.24875431e-02 7.43860602e-02
5.53419411e-01 7.09129751e-01 4.76081133e-01 -7.35714436e-01
-4.92269360e-02 -5.27272463e-01 -4.08491828e-02 -1.48559257e-01
3.31566572e-01 -8.63490939e-01 1.43778801e-01 7.71212637e-01
-4.95301425e-01 -7.27984682e-02 -1.48835644e-01 4.28533643e-01
6.72145113e-02 -7.97219336e-01 1.05029798e+00 1.63506940e-01
-2.24571615e-01 -1.63270086e-01 -2.74441630e-01 2.90433794e-01
9.39507246e-01 8.20160657e-02 -4.03578877e-01 -6.98519409e-01
-5.79228163e-01 2.81097144e-01 5.86873353e-01 -1.94988266e-01
1.32504791e-01 -1.39061439e+00 -3.22990745e-01 2.34866709e-01
-2.05216944e-01 -3.71467322e-01 -1.12425812e-01 1.36373460e+00
-2.68959880e-01 6.53909385e-01 1.85582623e-01 -3.58467311e-01
-8.00268888e-01 5.86695254e-01 9.48618114e-01 -4.18764621e-01
-1.36897638e-01 5.17387688e-01 -4.99958172e-02 1.95233407e-03
9.74441618e-02 -4.76980209e-01 3.59945416e-01 -5.35605289e-02
4.95293468e-01 7.00012028e-01 -1.18707605e-01 -5.75779736e-01
-3.91608357e-01 5.40842831e-01 3.62142742e-01 -3.10516506e-01
9.13665056e-01 -6.07486591e-02 1.40816392e-02 6.50363147e-01
1.31379712e+00 -9.28575918e-02 -1.31601083e+00 3.66974883e-02
2.52141859e-02 -5.82883060e-01 1.55327499e-01 -9.00822282e-01
-8.74668181e-01 6.56337738e-01 4.20023024e-01 8.95181000e-01
7.74021924e-01 -1.86821654e-01 8.11368048e-01 4.34875816e-01
3.81226480e-01 -9.88058269e-01 -4.26808178e-01 4.90806341e-01
7.01704800e-01 -1.19847035e+00 1.00697331e-01 -6.96560964e-02
-6.43423557e-01 9.24449623e-01 1.18373238e-01 -3.54525864e-01
6.79283440e-01 1.42659917e-01 -9.17832274e-03 2.34198011e-02
-8.25901270e-01 -9.74498019e-02 1.97820425e-01 1.02118999e-01
3.51413310e-01 1.20995820e-01 -4.91656870e-01 6.00274563e-01
-2.98651308e-01 2.16103449e-01 5.28736353e-01 7.07628906e-01
-6.05503619e-01 -8.72376859e-01 -3.20374638e-01 4.45767432e-01
-4.17682618e-01 2.08356217e-01 -4.32841957e-01 8.75361264e-01
-3.07330042e-01 1.04295111e+00 2.74316043e-01 -1.73353732e-01
4.65606689e-01 1.86058253e-01 7.83901393e-01 -1.34967998e-01
-2.74153769e-01 -2.13423416e-01 1.74145937e-01 -7.11459696e-01
-1.14882961e-01 -5.75871408e-01 -1.39382827e+00 -8.07245493e-01
-4.48961765e-01 4.53182399e-01 5.21273017e-01 1.24129069e+00
1.06301708e-02 1.87831894e-01 6.83287144e-01 -3.16086650e-01
-1.33824241e+00 -6.71898901e-01 -9.15840268e-01 2.65022248e-01
7.40630925e-01 -5.68622351e-01 -6.04888737e-01 -4.38179374e-01] | [4.300815105438232, 2.4032363891601562] |
cd60a6b8-2aad-4fc3-b5f1-ed5f6d83de93 | online-unsupervised-video-object-segmentation | 2306.12048 | null | https://arxiv.org/abs/2306.12048v1 | https://arxiv.org/pdf/2306.12048v1.pdf | Online Unsupervised Video Object Segmentation via Contrastive Motion Clustering | Online unsupervised video object segmentation (UVOS) uses the previous frames as its input to automatically separate the primary object(s) from a streaming video without using any further manual annotation. A major challenge is that the model has no access to the future and must rely solely on the history, i.e., the segmentation mask is predicted from the current frame as soon as it is captured. In this work, a novel contrastive motion clustering algorithm with an optical flow as its input is proposed for the online UVOS by exploiting the common fate principle that visual elements tend to be perceived as a group if they possess the same motion pattern. We build a simple and effective auto-encoder to iteratively summarize non-learnable prototypical bases for the motion pattern, while the bases in turn help learn the representation of the embedding network. Further, a contrastive learning strategy based on a boundary prior is developed to improve foreground and background feature discrimination in the representation learning stage. The proposed algorithm can be optimized on arbitrarily-scale data i.e., frame, clip, dataset) and performed in an online fashion. Experiments on $\textit{DAVIS}_{\textit{16}}$, $\textit{FBMS}$, and $\textit{SegTrackV2}$ datasets show that the accuracy of our method surpasses the previous state-of-the-art (SoTA) online UVOS method by a margin of 0.8%, 2.9%, and 1.1%, respectively. Furthermore, by using an online deep subspace clustering to tackle the motion grouping, our method is able to achieve higher accuracy at $3\times$ faster inference time compared to SoTA online UVOS method, and making a good trade-off between effectiveness and efficiency. | ['Zhengguo Li', 'Zhong Liu', 'Xingming Wu', 'Weihai Chen', 'Lin Xi'] | 2023-06-21 | null | null | null | null | ['contrastive-learning', 'optical-flow-estimation', 'video-object-segmentation', 'video-semantic-segmentation', 'unsupervised-video-object-segmentation', 'contrastive-learning', 'clustering'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'methodology', 'methodology'] | [ 2.46677846e-01 -9.25680846e-02 -1.19411871e-01 -1.86820924e-01
-4.84607905e-01 -5.37706137e-01 2.45327175e-01 2.21763290e-02
-5.64066887e-01 2.49457881e-01 -2.97496587e-01 -3.81446742e-02
-1.25980705e-01 -6.20731711e-01 -7.67596960e-01 -9.17410493e-01
3.61141525e-02 3.47466826e-01 6.28192961e-01 2.14217111e-01
9.75979045e-02 3.47652376e-01 -1.58431911e+00 1.42662793e-01
7.35158026e-01 1.38867879e+00 5.41102171e-01 6.89000487e-01
-1.77489877e-01 9.57242489e-01 -3.37281018e-01 -2.78320640e-01
3.35822403e-01 -6.17217243e-01 -7.54895568e-01 5.27914584e-01
5.61383426e-01 -3.53306711e-01 -5.50787807e-01 1.04633880e+00
2.93874115e-01 3.62959713e-01 5.80737412e-01 -1.09319866e+00
-5.84812723e-02 3.57804060e-01 -6.14237785e-01 3.73210758e-01
-6.12854734e-02 1.82544321e-01 1.05998731e+00 -9.18318748e-01
7.88740635e-01 8.53547692e-01 3.15298676e-01 3.73791367e-01
-1.26322329e+00 -4.95493770e-01 5.30282617e-01 4.98696685e-01
-1.56675887e+00 -4.51026410e-01 1.02169049e+00 -6.99952245e-01
5.63395679e-01 2.68306404e-01 8.58254671e-01 7.42721736e-01
-2.55150497e-02 9.74706709e-01 7.21357286e-01 -2.74276853e-01
4.71123517e-01 -8.24791044e-02 1.47979438e-01 1.01277268e+00
-5.03929742e-02 -2.56209165e-01 -5.92023015e-01 1.41952187e-01
7.83822894e-01 6.40506670e-02 -2.48347372e-01 -4.48241621e-01
-1.07468617e+00 6.02234066e-01 2.76412219e-01 2.05638602e-01
-1.99927792e-01 2.18393758e-01 2.99354345e-01 -3.04103717e-02
3.76107901e-01 -6.08645715e-02 -3.95300448e-01 -1.47641256e-01
-1.27649486e+00 -1.65212192e-02 6.14981949e-01 8.51252377e-01
1.11855960e+00 1.64459482e-01 -6.58806041e-02 6.61983967e-01
3.23566556e-01 3.73006523e-01 3.50255787e-01 -1.34386742e+00
4.27345753e-01 6.43437326e-01 1.49018571e-01 -1.24126971e+00
-3.23301971e-01 -3.47481489e-01 -7.90467203e-01 1.09729394e-01
4.72854644e-01 -6.70760199e-02 -1.01147342e+00 1.68253708e+00
5.18041372e-01 5.93887806e-01 -1.14886381e-01 8.65518689e-01
4.91511881e-01 9.23052549e-01 -1.80591762e-01 -6.09587967e-01
1.14593244e+00 -1.05109334e+00 -6.30598307e-01 -2.38736287e-01
5.40839493e-01 -5.08085072e-01 7.79652178e-01 5.10363817e-01
-9.79484022e-01 -7.60490596e-01 -9.29656386e-01 3.69519964e-02
6.31548604e-03 2.98015654e-01 5.68106413e-01 4.60085213e-01
-8.79994392e-01 6.46269023e-01 -1.17244220e+00 -1.54899180e-01
5.64101994e-01 4.35487002e-01 -1.51295692e-01 2.88050380e-02
-6.25860631e-01 1.14482045e-01 5.74220598e-01 1.80511892e-01
-1.10158634e+00 -5.08470058e-01 -8.34347248e-01 3.37429047e-02
7.61723638e-01 -5.55216372e-01 7.09090292e-01 -1.23366928e+00
-1.58537209e+00 7.69992411e-01 -4.71411437e-01 -5.40938199e-01
6.57127082e-01 -2.67114520e-01 -1.03411324e-01 7.62887836e-01
9.88720916e-03 7.98832119e-01 1.13143742e+00 -1.16608846e+00
-8.80056977e-01 -3.26761544e-01 1.21745206e-01 1.22663356e-01
-4.80159432e-01 -3.44582461e-02 -1.08304286e+00 -7.43694186e-01
3.08377147e-01 -1.12093878e+00 -1.42812893e-01 2.60899335e-01
-1.94009379e-01 -1.66700885e-01 8.63765836e-01 -6.47111952e-01
1.48862386e+00 -2.31347156e+00 4.40912604e-01 1.13593630e-01
2.71546125e-01 3.96663994e-01 2.46889070e-01 1.04893938e-01
1.64144590e-01 2.70236400e-03 -5.02836049e-01 -3.59960377e-01
-1.99010715e-01 1.72901064e-01 -1.40650598e-02 6.66053832e-01
4.53579873e-02 5.07142961e-01 -7.93220758e-01 -8.38665783e-01
4.11612749e-01 2.63396502e-01 -6.91376805e-01 2.52731144e-01
-2.83531159e-01 4.54048306e-01 -2.81358957e-01 3.86123329e-01
5.36967874e-01 -3.51605326e-01 2.70570636e-01 -3.28674555e-01
-1.80076644e-01 -2.12167978e-01 -1.60364926e+00 1.89030623e+00
-1.21048115e-01 6.79843485e-01 1.67732686e-01 -1.31208694e+00
7.48754680e-01 2.09841892e-01 9.03594375e-01 -2.71499395e-01
2.05524862e-01 5.14763109e-02 -2.74836784e-03 -5.70011675e-01
1.88299984e-01 2.46999890e-01 9.67924669e-02 2.09730461e-01
2.42043704e-01 3.02410603e-01 4.90266204e-01 1.60128564e-01
9.60005939e-01 2.90659487e-01 -4.11794670e-02 -3.00835371e-01
7.84216583e-01 -7.29400516e-02 1.05326521e+00 6.40750110e-01
-2.90497810e-01 7.18935072e-01 4.49968904e-01 -3.55502903e-01
-7.99272418e-01 -9.17749226e-01 7.66885355e-02 9.12077367e-01
5.43978214e-01 -4.61480856e-01 -8.88402283e-01 -6.91745281e-01
-2.88650513e-01 3.94151717e-01 -4.83399481e-01 -3.63361426e-02
-7.75827825e-01 -5.70904374e-01 1.58146665e-01 5.98851085e-01
6.57338023e-01 -7.98115492e-01 -8.94837677e-01 2.78890610e-01
-2.99672931e-01 -1.34763825e+00 -5.26352823e-01 -8.61359462e-02
-7.90663958e-01 -9.76767540e-01 -6.05429053e-01 -7.83434272e-01
7.68665493e-01 3.52472454e-01 6.24838471e-01 3.62145789e-02
-2.62359589e-01 5.45862913e-01 -3.82426351e-01 1.00784777e-02
-1.02510922e-01 -1.07667714e-01 2.23950073e-02 7.50131667e-01
9.21876729e-02 -4.90976065e-01 -9.98982012e-01 3.86582375e-01
-9.73485112e-01 2.52341300e-01 2.64394611e-01 8.75241816e-01
8.43180656e-01 2.72275239e-01 2.00547770e-01 -7.68599272e-01
-4.06332791e-01 -3.08900237e-01 -8.29694390e-01 1.31155789e-01
-4.61992145e-01 -1.09417580e-01 7.37004280e-01 -4.27781254e-01
-9.47646320e-01 4.44646478e-01 1.20668232e-01 -9.57537413e-01
-1.17542461e-01 2.42803574e-01 -2.25319222e-01 1.87174961e-01
1.97609708e-01 4.39503819e-01 -3.10241785e-02 -4.78669077e-01
3.29771757e-01 4.40947980e-01 5.93325078e-01 -4.20157939e-01
7.78120339e-01 9.60273385e-01 -5.85527122e-02 -9.05308843e-01
-6.52759433e-01 -6.44369423e-01 -9.60726917e-01 -5.38558006e-01
1.16344690e+00 -1.02338529e+00 -6.64947987e-01 5.09798765e-01
-1.01475430e+00 -3.60327780e-01 -1.47891775e-01 5.20037949e-01
-6.72892511e-01 6.44952714e-01 -6.42392576e-01 -9.87542212e-01
-1.29172578e-01 -1.08955872e+00 8.09855998e-01 2.14810938e-01
-3.24332193e-02 -7.87099600e-01 -4.67445970e-01 6.15495443e-01
-2.26459950e-01 2.71803290e-01 6.21010125e-01 -3.79884183e-01
-9.91242886e-01 2.78731640e-02 -1.35234073e-01 5.48921347e-01
1.04178093e-01 2.78777897e-01 -8.14378262e-01 -3.93009514e-01
5.78699932e-02 3.20062190e-02 1.03047025e+00 4.36269194e-01
1.34953117e+00 -3.04648429e-01 -3.22784781e-01 7.31337607e-01
1.55941319e+00 4.68035311e-01 4.78944689e-01 6.03457913e-02
1.00273669e+00 6.17254972e-01 7.02817261e-01 6.65140152e-01
2.79725224e-01 6.56531513e-01 4.51888740e-01 1.66743085e-01
-1.39504611e-01 -8.66546556e-02 5.60997844e-01 9.84757245e-01
-1.02767386e-01 -3.20693761e-01 -7.89875448e-01 6.74587131e-01
-2.02512455e+00 -1.01274359e+00 -1.20994765e-02 2.28173971e+00
6.24204874e-01 1.68007195e-01 9.05692130e-02 2.96037763e-01
7.83496737e-01 4.13560867e-01 -5.89578331e-01 6.35464862e-02
-8.89370311e-03 8.98083001e-02 2.57665396e-01 4.83302951e-01
-1.37809801e+00 1.08268344e+00 4.26684284e+00 9.25515771e-01
-1.17737532e+00 8.89165178e-02 7.65230119e-01 -1.46838754e-01
9.74101797e-02 2.02286050e-01 -7.17471421e-01 7.50955999e-01
6.14573240e-01 8.42752978e-02 4.82001573e-01 6.78320467e-01
3.97447705e-01 -2.74467260e-01 -1.11829293e+00 1.22578323e+00
1.72222421e-01 -1.43216717e+00 -1.36199594e-01 -1.20141707e-01
6.56379163e-01 -2.42922947e-01 -2.96611711e-02 -2.66922191e-02
-2.13627219e-01 -6.03059649e-01 1.08357942e+00 5.34286559e-01
6.32412314e-01 -7.40414500e-01 4.48238045e-01 5.03692091e-01
-1.47279501e+00 -3.19460243e-01 -1.35755837e-01 4.28820550e-02
2.82892793e-01 5.44734299e-01 -3.94689500e-01 7.25246131e-01
8.84612441e-01 8.68859947e-01 -4.84601855e-01 8.06157112e-01
8.43001157e-03 7.73793876e-01 -3.64331782e-01 1.34072736e-01
2.22071290e-01 -3.87384444e-01 6.72073424e-01 1.14151549e+00
2.59954363e-01 2.56060302e-01 4.92665112e-01 6.84180200e-01
8.75477046e-02 1.06533475e-01 -1.76045522e-01 2.68213749e-02
2.91108757e-01 1.17314720e+00 -1.15358245e+00 -5.81628382e-01
-3.46838862e-01 1.22346246e+00 1.36330351e-01 4.29069698e-01
-1.01792824e+00 -2.55483925e-01 4.96076763e-01 5.34071475e-02
9.33460593e-01 -4.75271910e-01 3.20684095e-03 -1.33337796e+00
1.45839736e-01 -5.08774996e-01 4.07266557e-01 -4.66294944e-01
-8.45888555e-01 4.32520658e-01 -1.85426641e-02 -1.45388436e+00
-2.19900101e-01 -5.51787972e-01 -3.60721648e-01 2.69650161e-01
-1.16432428e+00 -8.54048669e-01 -3.64294529e-01 6.14707053e-01
7.58774996e-01 -1.48866298e-02 3.01843256e-01 4.41006392e-01
-9.23213780e-01 4.66208428e-01 2.56551534e-01 3.88139516e-01
3.32854122e-01 -1.01704538e+00 -1.99859619e-01 1.15966010e+00
4.60478634e-01 2.72789448e-01 4.93770599e-01 -5.24716437e-01
-1.51661265e+00 -1.24150670e+00 4.19488251e-01 -1.66022092e-01
6.07494593e-01 -5.35614073e-01 -8.79529834e-01 4.74671811e-01
2.32003219e-02 2.82882690e-01 5.89383006e-01 -3.69488269e-01
-7.32290894e-02 -5.34181595e-01 -8.27955782e-01 5.43919325e-01
1.32670021e+00 -3.34583819e-01 -3.32976431e-01 1.14206322e-01
6.94252014e-01 -3.48770499e-01 -8.55598927e-01 3.61245543e-01
4.10749495e-01 -1.03366959e+00 8.64648104e-01 -2.68504769e-01
3.40998173e-01 -6.45817876e-01 -2.31374338e-01 -6.52232587e-01
-3.13637614e-01 -9.15254533e-01 -4.87558514e-01 1.41076136e+00
4.64730151e-02 -3.52528930e-01 8.70665789e-01 4.30941582e-01
-1.47968248e-01 -8.61158073e-01 -1.08520341e+00 -7.34412730e-01
-3.18638057e-01 -6.03866935e-01 5.70249259e-02 8.82283628e-01
-4.73528445e-01 1.23097993e-01 -3.20005268e-01 4.49263990e-01
6.89515293e-01 3.39464396e-01 8.02491486e-01 -1.11173391e+00
-4.48162615e-01 -3.08931351e-01 -5.89033842e-01 -1.42968225e+00
1.44529909e-01 -9.00483191e-01 -4.52487096e-02 -1.25773406e+00
5.91995753e-02 -3.56533229e-01 -5.13565719e-01 2.96627641e-01
-2.02494919e-01 9.14093554e-02 5.18009603e-01 2.34685406e-01
-8.64379942e-01 5.63275814e-01 9.20025110e-01 -1.70400321e-01
-2.32981771e-01 -1.54889837e-01 -2.24165753e-01 9.01938677e-01
4.65022802e-01 -3.32045764e-01 -5.08728623e-01 -5.02675593e-01
-3.88112618e-03 1.80929050e-01 4.59838986e-01 -1.12318504e+00
3.60828847e-01 -1.13160409e-01 2.63280094e-01 -6.43079162e-01
4.69640136e-01 -8.86368752e-01 1.48152545e-01 4.51504230e-01
-8.28973129e-02 -2.01591760e-01 9.81846172e-03 1.01053512e+00
-1.67596862e-01 -3.24439943e-01 7.81893611e-01 -3.28492932e-02
-9.58489954e-01 6.05533779e-01 -4.48925704e-01 4.36939448e-02
1.17151010e+00 -4.25702840e-01 8.37700143e-02 -1.68507621e-01
-8.59722495e-01 3.42077732e-01 4.70920235e-01 1.95388630e-01
6.17345393e-01 -1.05121553e+00 -4.97800499e-01 5.05985953e-02
-5.36570773e-02 3.54814291e-01 6.10151827e-01 9.61579442e-01
-6.78658724e-01 -1.21848704e-02 1.01352900e-01 -1.02630556e+00
-1.20345700e+00 5.57439566e-01 1.99282676e-01 -7.19264969e-02
-7.93248594e-01 9.00624514e-01 3.79885465e-01 2.93473035e-01
1.97554350e-01 -1.96573228e-01 -1.52747855e-01 2.86909521e-01
4.64829594e-01 6.09136462e-01 -2.48522460e-01 -8.03109825e-01
-3.31626296e-01 7.55403340e-01 1.80229414e-02 -1.02341570e-01
1.24016237e+00 -3.06071490e-01 -1.29664347e-01 7.00529277e-01
1.39538181e+00 -1.20830767e-01 -1.76380551e+00 -2.51650751e-01
-1.09091759e-01 -6.15053415e-01 -9.13172364e-02 -1.67372555e-01
-1.37530315e+00 9.24527824e-01 7.73681223e-01 1.01318300e-01
1.22694147e+00 8.85265321e-02 8.23045492e-01 1.75286874e-01
3.81295443e-01 -1.35566926e+00 2.56000221e-01 2.97524929e-01
4.36494708e-01 -1.09099472e+00 6.27522171e-02 -5.40489197e-01
-5.35987854e-01 1.14113843e+00 4.67416525e-01 -1.77144349e-01
6.68184519e-01 -1.31216973e-01 -1.75296322e-01 -2.10986007e-02
-5.89089751e-01 -2.26121068e-01 2.18592361e-01 2.71740377e-01
-7.16142282e-02 -1.91786155e-01 -2.26136416e-01 5.03089726e-01
9.91772115e-02 -2.06542581e-01 4.32713956e-01 8.45411658e-01
-3.62390369e-01 -7.59899080e-01 -1.66505098e-01 2.60439873e-01
-5.73890805e-01 2.28509545e-01 -5.06251901e-02 5.57119608e-01
4.90690470e-01 8.80865157e-01 1.79514900e-01 -3.63772511e-01
-9.84156430e-02 8.09383094e-02 3.72920990e-01 -5.11836052e-01
-1.71743959e-01 4.87281293e-01 -1.95864767e-01 -7.67012596e-01
-7.09726155e-01 -8.99660707e-01 -1.48853612e+00 9.44371223e-02
-3.20300251e-01 -2.77682561e-02 2.80443728e-01 9.10995603e-01
4.42557901e-01 4.86829400e-01 8.86104286e-01 -1.00624502e+00
2.08868366e-02 -4.07522470e-01 -4.79277790e-01 5.97133458e-01
2.11734802e-01 -7.36198962e-01 -4.09910530e-01 6.56641245e-01] | [9.074275016784668, -0.17959581315517426] |
3238df93-5519-4a1b-8258-0f48d636bc6f | operator-valued-kernels-for-learning-from | 1510.08231 | null | http://arxiv.org/abs/1510.08231v3 | http://arxiv.org/pdf/1510.08231v3.pdf | Operator-valued Kernels for Learning from Functional Response Data | In this paper we consider the problems of supervised classification and
regression in the case where attributes and labels are functions: a data is
represented by a set of functions, and the label is also a function. We focus
on the use of reproducing kernel Hilbert space theory to learn from such
functional data. Basic concepts and properties of kernel-based learning are
extended to include the estimation of function-valued functions. In this
setting, the representer theorem is restated, a set of rigorously defined
infinite-dimensional operator-valued kernels that can be valuably applied when
the data are functions is described, and a learning algorithm for nonlinear
functional data analysis is introduced. The methodology is illustrated through
speech and audio signal processing experiments. | ['Stéphane Canu', 'Philippe Preux', 'Hachem Kadri', 'Julien Audiffren', 'Emmanuel Duflos', 'Alain Rakotomamonjy'] | 2015-10-28 | null | null | null | null | ['audio-signal-processing'] | ['audio'] | [ 2.16418386e-01 6.54082149e-02 -1.30216852e-01 -7.04213083e-01
-5.46744704e-01 -3.10532749e-01 3.09310675e-01 4.45001610e-02
-4.42620933e-01 6.71263576e-01 -1.29383147e-01 -1.38513237e-01
-4.72685695e-01 -4.36811209e-01 -3.69007498e-01 -9.67941821e-01
-6.45466864e-01 5.02390973e-02 -2.19866693e-01 -9.59141105e-02
1.17552482e-01 6.57598555e-01 -1.54489696e+00 7.16433227e-02
8.35505962e-01 1.27157760e+00 -2.92343438e-01 9.35062230e-01
1.86940148e-01 1.01398277e+00 -4.62473840e-01 1.48973996e-02
1.67263135e-01 -4.28878278e-01 -9.05123413e-01 3.64139199e-01
-2.32993811e-02 2.14497432e-01 -2.52052188e-01 1.04739416e+00
2.70678699e-01 5.69119513e-01 1.53892148e+00 -1.57204545e+00
-7.13346183e-01 1.72010735e-02 7.64704496e-02 1.69059709e-02
2.58129954e-01 -3.83139223e-01 9.02131200e-01 -1.20130002e+00
-8.55907202e-02 1.08392274e+00 8.66274476e-01 3.73626858e-01
-1.75725853e+00 -5.68326414e-02 -4.88823682e-01 2.46469915e-01
-1.35245311e+00 -4.05080408e-01 7.07214117e-01 -9.55307543e-01
5.85226893e-01 3.43631864e-01 5.02573371e-01 6.00496531e-01
2.57344902e-01 7.03149080e-01 1.06726396e+00 -7.41752148e-01
5.52754641e-01 4.33791906e-01 5.39540470e-01 9.60416675e-01
2.36420669e-02 2.27408662e-01 -1.67922407e-01 -5.18633008e-01
5.85709691e-01 -1.76280096e-01 -5.43474019e-01 -7.39005804e-01
-1.00851083e+00 1.36177552e+00 -4.85278070e-02 3.94866914e-01
-5.53353786e-01 -1.76567007e-02 6.39348924e-01 7.03108966e-01
8.27161372e-01 4.95034903e-01 -3.12079102e-01 2.05528438e-01
-6.46868765e-01 3.07868235e-02 1.25224507e+00 7.98040509e-01
7.66302466e-01 2.67867088e-01 5.78909926e-03 8.02579165e-01
1.12014852e-01 6.81069016e-01 3.63413781e-01 -9.15891111e-01
-1.68523505e-01 1.43465444e-01 3.51277798e-01 -6.49105132e-01
-5.30048132e-01 -5.01709357e-02 -5.52292526e-01 4.84643430e-01
6.21499956e-01 -6.69843778e-02 -4.48671609e-01 1.46331036e+00
1.52407989e-01 3.02061230e-01 2.42262751e-01 6.89646661e-01
3.58864427e-01 7.08714426e-01 -1.96150020e-01 -6.58076286e-01
8.03556204e-01 -3.62020850e-01 -9.65507686e-01 5.82867622e-01
8.82669747e-01 -5.46585858e-01 9.83383596e-01 5.21355689e-01
-9.42199171e-01 -5.24521112e-01 -1.06494629e+00 2.01838449e-01
-5.26044607e-01 1.29744440e-01 5.15060663e-01 5.86872220e-01
-1.05851245e+00 8.78553152e-01 -4.46684062e-01 -2.53358901e-01
5.63040115e-02 4.63785350e-01 -5.44916034e-01 2.88487613e-01
-1.17337310e+00 1.01922965e+00 4.72864747e-01 1.60717919e-01
-5.40197253e-01 -8.64873052e-01 -1.11390591e+00 -3.66497375e-02
-1.71943456e-01 -1.01377495e-01 1.38731503e+00 -1.19614911e+00
-1.33537948e+00 9.87677515e-01 9.59874988e-02 -2.86993116e-01
1.90718517e-01 1.59932703e-01 -8.50749075e-01 3.32938313e-01
4.01363522e-02 -2.82643199e-01 1.37384498e+00 -8.17845941e-01
-3.63662332e-01 -1.43558741e-01 -2.11967006e-01 -2.35366285e-01
-4.33386475e-01 -2.12736651e-02 5.91355503e-01 -5.20610988e-01
-3.71113010e-02 -7.06693232e-01 1.77948609e-01 1.40708178e-01
-1.91966712e-01 -4.71648753e-01 7.70801127e-01 -5.95931590e-01
1.09993851e+00 -2.46918631e+00 1.37038738e-01 4.96491820e-01
1.34164467e-02 -8.38690177e-02 1.20368801e-01 6.21045947e-01
-6.42374396e-01 -1.79864213e-01 -6.86071634e-01 1.56786665e-01
1.50170535e-01 1.29468337e-01 -3.66657645e-01 1.25048149e+00
2.79809743e-01 5.66532373e-01 -5.67280412e-01 -3.86987776e-01
3.22126031e-01 4.98402894e-01 -3.92640419e-02 3.38036716e-01
2.13731915e-01 1.80256590e-01 -6.14744365e-01 4.61728930e-01
4.48315918e-01 -1.40609294e-01 -3.38414997e-01 -1.87526494e-01
-1.38990715e-01 -2.53450602e-01 -1.15143573e+00 1.26132905e+00
-4.82485086e-01 8.40371311e-01 2.61496216e-01 -1.76435971e+00
1.01976371e+00 7.91185617e-01 7.68741012e-01 -1.00957043e-01
1.40217602e-01 3.06385279e-01 -1.05757313e-02 -8.62080038e-01
-2.96330869e-01 -5.08283198e-01 -1.20735265e-01 4.95281994e-01
8.43468532e-02 -1.22432552e-01 8.17107782e-03 -2.07349703e-01
1.02970326e+00 -1.08797744e-01 7.71100044e-01 -8.43734920e-01
1.01664555e+00 7.98877925e-02 1.23797677e-01 4.31023985e-01
-2.74526179e-01 1.55714363e-01 5.70608020e-01 -2.92207211e-01
-1.11404788e+00 -1.07754397e+00 -8.92123044e-01 1.20423198e+00
-4.51817185e-01 1.09109990e-01 -6.28621459e-01 -3.90481383e-01
5.15947163e-01 6.43359601e-01 -1.04077935e+00 -4.63949770e-01
-3.85314584e-01 -4.23157334e-01 5.67098200e-01 4.64413613e-01
6.41875640e-02 -1.01654208e+00 -3.63369972e-01 1.80407643e-01
3.84754628e-01 -6.31372333e-01 -6.00536227e-01 4.30500329e-01
-8.57561052e-01 -1.48941922e+00 -8.83734226e-01 -1.02860308e+00
6.42412543e-01 -2.69007593e-01 7.70149589e-01 -3.65056217e-01
-4.27375853e-01 1.07513964e+00 -1.90506175e-01 -3.69255453e-01
-7.12791324e-01 -5.61846733e-01 4.83566493e-01 6.48484349e-01
2.54545718e-01 -4.63212937e-01 -2.12555915e-01 3.30238223e-01
-8.32291603e-01 -5.48755825e-01 3.77790928e-01 1.14888692e+00
3.53399783e-01 7.27510303e-02 1.02633643e+00 -9.19894755e-01
8.16918671e-01 -5.17486274e-01 -5.00250161e-01 5.33849478e-01
-7.04725385e-01 2.99802125e-01 8.63512397e-01 -7.66108930e-01
-8.04325640e-01 2.06999049e-01 4.61734384e-01 -4.44922805e-01
-9.13081691e-02 5.78133523e-01 -2.55285680e-01 -4.95037854e-01
1.06377316e+00 3.53787869e-01 5.03514647e-01 -3.54939222e-01
3.78012896e-01 9.24389303e-01 4.26920801e-01 -7.18241751e-01
4.80258375e-01 3.42041016e-01 3.63204330e-01 -1.33746111e+00
-6.51846170e-01 -7.16023684e-01 -7.63513446e-01 -3.54438096e-01
6.55170441e-01 -3.92531902e-01 -1.03064752e+00 1.88632652e-01
-8.03735614e-01 -2.53903627e-01 -7.50892401e-01 9.95567679e-01
-1.40568471e+00 1.48783281e-01 -5.72409987e-01 -1.52287495e+00
-5.89727648e-02 -7.07107544e-01 5.68159282e-01 -2.27406591e-01
-3.90350893e-02 -1.65404475e+00 1.88036487e-01 -2.59456247e-01
3.10833603e-01 3.23479861e-01 1.16894698e+00 -1.15281093e+00
4.84733850e-01 -4.92437094e-01 -3.15297507e-02 8.93269241e-01
2.33984247e-01 -4.20489699e-01 -1.19452560e+00 -2.78765857e-01
5.78254938e-01 -4.31988537e-01 6.14736676e-01 5.51248491e-01
1.07418084e+00 -3.11383694e-01 4.27928641e-02 4.56288040e-01
1.27057135e+00 3.20063680e-01 -3.73945720e-02 -2.63514310e-01
3.45738888e-01 9.71012592e-01 4.94088024e-01 4.21450675e-01
-4.84405398e-01 3.54697227e-01 -8.36046487e-02 6.17078319e-02
5.16718209e-01 2.21253723e-01 2.31938019e-01 6.53725564e-01
6.56348541e-02 1.80954039e-01 -6.36821806e-01 2.77653247e-01
-2.05410099e+00 -9.39175904e-01 -3.42563659e-01 2.55846095e+00
7.94601440e-01 -3.32778186e-01 5.04420698e-01 5.62606275e-01
1.10867667e+00 -3.55866283e-01 -7.37828374e-01 -2.97972947e-01
-5.38527071e-02 5.29670238e-01 3.90278995e-01 6.59480035e-01
-1.33786118e+00 2.25365028e-01 7.72284651e+00 6.19641483e-01
-8.90282691e-01 4.68775295e-02 2.32525483e-01 5.02742887e-01
2.19170079e-01 -2.89280772e-01 -3.73340733e-02 2.62551248e-01
1.15866399e+00 -3.64240974e-01 4.75977540e-01 1.01189625e+00
2.18508601e-01 8.82508084e-02 -1.52305508e+00 1.18176389e+00
7.13918880e-02 -7.37562478e-01 -5.53940773e-01 7.32325017e-02
2.18996838e-01 -6.34682596e-01 1.46737128e-01 5.15241623e-01
-2.91836143e-01 -1.25419414e+00 3.77534747e-01 1.00099432e+00
9.17307913e-01 -8.49876285e-01 3.94323766e-01 3.14642459e-01
-1.07413423e+00 -1.85887173e-01 -5.42439401e-01 -2.06477717e-01
-3.06772739e-01 7.21061826e-01 -6.72333479e-01 2.46464014e-01
-3.26015241e-02 1.07047749e+00 3.04975975e-02 1.19495046e+00
8.77509043e-02 8.25394213e-01 -1.43912388e-02 -9.70075056e-02
-6.19693957e-02 -6.01046145e-01 3.29058498e-01 1.36296451e+00
2.63714492e-01 2.84472525e-01 2.94131339e-01 7.84044325e-01
3.62206489e-01 5.58894098e-01 -9.05246139e-01 -1.11022323e-01
1.56091034e-01 1.24475861e+00 -4.27044034e-01 -3.53071481e-01
-6.47678673e-01 1.09116066e+00 2.97381729e-01 7.06672907e-01
-5.20489573e-01 -6.28785849e-01 7.53158808e-01 7.57094994e-02
1.05929293e-01 6.27070740e-02 -8.43000785e-03 -1.20271468e+00
-1.36990443e-01 -1.83226466e-01 5.17687678e-01 -4.28484678e-01
-1.71279502e+00 2.13930253e-02 2.79794246e-01 -1.19958448e+00
-2.49380365e-01 -1.30014086e+00 -6.32628024e-01 1.08213770e+00
-1.00427604e+00 -6.28045678e-01 2.94333875e-01 1.00054884e+00
1.43341511e-01 -2.94815809e-01 1.09298325e+00 -5.46634616e-03
-2.06378162e-01 1.99432939e-01 5.23508489e-01 2.37255171e-01
3.99860740e-01 -1.79606581e+00 -3.15124005e-01 2.92514503e-01
1.48430511e-01 4.07622725e-01 7.96054006e-01 -2.97122896e-01
-1.39656878e+00 -8.94387305e-01 7.05570698e-01 -1.23848371e-01
9.99424875e-01 -4.54841882e-01 -1.20129645e+00 6.02779806e-01
-2.82232583e-01 6.66753292e-01 1.09046054e+00 5.50270230e-02
-2.92811453e-01 -9.19725969e-02 -1.33687365e+00 -1.62512317e-01
4.59143460e-01 -8.57396603e-01 -5.48822045e-01 7.06713319e-01
2.06192762e-01 1.23865649e-01 -1.27845919e+00 2.13529512e-01
5.82277119e-01 -7.30088890e-01 8.92951727e-01 -1.20478559e+00
-4.09120291e-01 -5.95754869e-02 -2.80506760e-01 -1.38798440e+00
-4.26610172e-01 -6.01497650e-01 -3.50865871e-01 5.90692580e-01
2.06284374e-01 -9.54933643e-01 2.84128338e-01 5.96023858e-01
-9.16418433e-02 -8.48183036e-01 -1.09169400e+00 -1.14888632e+00
1.26430303e-01 -4.21325207e-01 -9.90226045e-02 1.11851895e+00
6.50969267e-01 2.82308400e-01 -2.88959503e-01 -6.73954040e-02
7.17972457e-01 -8.93896520e-02 -1.19618103e-02 -1.58732891e+00
-3.82165104e-01 -3.54488611e-01 -9.51053321e-01 -5.07565558e-01
7.64288425e-01 -1.08824658e+00 1.10212103e-01 -9.39296901e-01
-2.60763466e-01 -3.16770375e-01 -3.79049927e-01 1.65139422e-01
2.90527672e-01 -8.74970555e-02 -3.10023665e-01 2.72423297e-01
-2.73743551e-03 4.98243809e-01 7.46848166e-01 6.07872680e-02
-5.49481809e-02 6.58813894e-01 -1.42268896e-01 7.04415619e-01
6.48028553e-01 -2.95771658e-01 -7.36564994e-01 6.71039820e-01
-4.34570760e-02 3.37197095e-01 4.76082653e-01 -8.20144892e-01
5.08019514e-02 -1.17459640e-01 4.39302713e-01 3.59871313e-02
4.30955648e-01 -9.39956248e-01 -2.08917797e-01 5.37197351e-01
-7.65549362e-01 -2.19078839e-01 -2.89296836e-01 7.33082891e-01
-1.82309449e-01 -6.75148785e-01 9.74284589e-01 3.25989306e-01
-5.92432082e-01 2.85441160e-01 -5.00986040e-01 2.77777493e-01
1.31190395e+00 -1.63185701e-01 2.63310581e-01 -4.66450721e-01
-1.26767361e+00 -2.88473725e-01 1.08358040e-01 -2.63764620e-01
7.83577323e-01 -1.55092466e+00 -7.90618122e-01 4.09736753e-01
3.70501637e-01 -7.42233098e-01 -1.00956678e-01 1.06458414e+00
-2.49775440e-01 4.78022009e-01 7.00807199e-02 -3.88906837e-01
-8.31095159e-01 7.64767945e-01 6.66668892e-01 4.18721974e-01
-5.74038506e-01 5.41243851e-01 3.34584296e-01 -3.34098518e-01
1.72761127e-01 -4.05555546e-01 -2.28895605e-01 -5.31113893e-02
5.99124312e-01 5.40470779e-01 -4.15543560e-03 -7.44201779e-01
-2.31957465e-01 2.66801268e-01 2.99718171e-01 -4.08655822e-01
9.66912329e-01 2.51150429e-01 -1.55646667e-01 1.35822761e+00
1.74204290e+00 -4.01664168e-01 -9.72577810e-01 -5.17332554e-01
3.01897615e-01 -1.50756106e-01 -1.01532914e-01 -4.13581520e-01
-5.06951332e-01 6.73855424e-01 6.60729766e-01 8.32951665e-01
1.05274761e+00 8.85059088e-02 1.55601293e-01 5.91702402e-01
1.12091221e-01 -1.17314303e+00 -1.58271715e-01 3.04481119e-01
9.73372877e-01 -1.13710785e+00 -1.81597874e-01 -4.82586354e-01
-4.84004408e-01 1.76576662e+00 -1.63690731e-01 -5.81138551e-01
1.33381283e+00 1.89991638e-01 -1.93541810e-01 1.18159913e-02
-4.16640103e-01 -3.24655056e-01 8.24470103e-01 8.94349515e-01
7.26349592e-01 2.42585301e-01 -3.95264596e-01 4.22098488e-01
2.21451055e-02 -1.64677039e-01 3.99758786e-01 8.83189380e-01
-6.81695163e-01 -7.29592383e-01 -4.18126285e-01 6.57235086e-01
-2.88846403e-01 1.30582362e-01 -4.20829989e-02 8.87140930e-01
-3.60428751e-01 9.04978693e-01 -7.51998350e-02 -1.46789342e-01
2.99957871e-01 5.93925714e-01 4.99228686e-01 -7.52282083e-01
-7.39110559e-02 -2.30695978e-01 -1.53298192e-02 -3.77480567e-01
-4.95440066e-01 -7.31117725e-01 -1.35107291e+00 2.38529909e-02
-3.63560379e-01 6.56963289e-01 8.22713494e-01 1.02880871e+00
-2.80606300e-01 2.06749782e-01 1.04368067e+00 -4.58406448e-01
-1.09982276e+00 -1.03885758e+00 -1.50613999e+00 2.44427666e-01
6.66800559e-01 -8.91339362e-01 -8.72887492e-01 1.80305287e-01] | [7.589210510253906, 4.087778568267822] |
06d5bde2-c85e-42a7-8632-62fcb6ac09cb | few-shot-multimodal-multitask-multilingual | 2303.12489 | null | https://arxiv.org/abs/2303.12489v1 | https://arxiv.org/pdf/2303.12489v1.pdf | Few-shot Multimodal Multitask Multilingual Learning | While few-shot learning as a transfer learning paradigm has gained significant traction for scenarios with limited data, it has primarily been explored in the context of building unimodal and unilingual models. Furthermore, a significant part of the existing literature in the domain of few-shot multitask learning perform in-context learning which requires manually generated prompts as the input, yielding varying outcomes depending on the level of manual prompt-engineering. In addition, in-context learning suffers from substantial computational, memory, and storage costs which eventually leads to high inference latency because it involves running all of the prompt's examples through the model every time a prediction is made. In contrast, methods based on the transfer learning via the fine-tuning paradigm avoid the aforementioned issues at a one-time cost of fine-tuning weights on a per-task basis. However, such methods lack exposure to few-shot multimodal multitask learning. In this paper, we propose few-shot learning for a multimodal multitask multilingual (FM3) setting by adapting pre-trained vision and language models using task-specific hypernetworks and contrastively fine-tuning them to enable few-shot learning. FM3's architecture combines the best of both worlds of in-context and fine-tuning based learning and consists of three major components: (i) multimodal contrastive fine-tuning to enable few-shot learning, (ii) hypernetwork task adaptation to perform multitask learning, and (iii) task-specific output heads to cater to a plethora of diverse tasks. FM3 learns the most prominent tasks in the vision and language domains along with their intersections, namely visual entailment (VE), visual question answering (VQA), and natural language understanding (NLU) tasks such as neural entity recognition (NER) and the GLUE benchmark including QNLI, MNLI, QQP, and SST-2. | ['Vinija Jain', 'Aman Chadha'] | 2023-02-19 | null | null | null | null | ['visual-entailment'] | ['reasoning'] | [ 2.94279784e-01 -7.86800459e-02 -1.02350675e-01 -3.18049431e-01
-1.16716528e+00 -5.26519060e-01 8.56493294e-01 5.65180853e-02
-7.81669140e-01 4.85833466e-01 -2.19391026e-02 -4.22906220e-01
-1.19965069e-01 -7.08735943e-01 -8.56668234e-01 -4.95969474e-01
3.84395778e-01 6.19982362e-01 2.70182639e-01 -4.44637060e-01
-1.44399211e-01 -5.77204339e-02 -1.67559004e+00 5.32299936e-01
9.35341775e-01 1.05881727e+00 4.03744012e-01 7.83897936e-01
-4.93819833e-01 7.41654038e-01 -4.75128651e-01 -7.02153921e-01
-9.87705365e-02 -2.54553407e-01 -8.59787881e-01 -9.92010534e-02
6.47142708e-01 -1.79114178e-01 -4.18302380e-02 8.93548608e-01
7.25874066e-01 4.69913423e-01 7.44588435e-01 -1.38368106e+00
-9.12169993e-01 2.71506429e-01 -5.19498348e-01 8.14869553e-02
1.64195225e-01 6.33143306e-01 1.01998079e+00 -1.23757613e+00
5.05762458e-01 1.22865796e+00 4.40111458e-01 7.57380128e-01
-1.28338957e+00 -4.73518521e-01 -5.09091914e-02 5.96599460e-01
-9.13466275e-01 -4.47337419e-01 4.93255824e-01 -4.66522664e-01
1.29553735e+00 -6.89401180e-02 2.62571841e-01 1.51154947e+00
1.69996414e-02 7.79368341e-01 1.11770511e+00 -6.03415608e-01
3.89391631e-01 1.51562378e-01 2.51497954e-01 6.89912558e-01
-4.05133277e-01 1.33800104e-01 -5.90717018e-01 4.04137634e-02
3.40634584e-01 7.31233209e-02 -1.16696343e-01 -3.56692135e-01
-1.16851270e+00 8.82422566e-01 3.35792005e-01 1.84965432e-01
-2.09588468e-01 -6.08129473e-03 6.68085217e-01 6.02892518e-01
2.45335266e-01 3.10162425e-01 -5.33886909e-01 -6.42829314e-02
-7.88527071e-01 -6.93605393e-02 7.12654829e-01 9.89023685e-01
1.16659200e+00 2.11495727e-01 -6.06158674e-01 1.07884789e+00
2.69878674e-02 4.63841051e-01 6.03070080e-01 -6.07806623e-01
6.86708510e-01 4.53059733e-01 -2.35407874e-01 -3.82230580e-01
-3.76271695e-01 5.51043823e-02 -7.56533444e-01 3.36118221e-01
3.99060726e-01 -4.79410708e-01 -1.29338765e+00 1.77737105e+00
3.01490456e-01 1.97417602e-01 2.89441884e-01 7.80474484e-01
1.24605727e+00 8.73078287e-01 5.37348390e-01 5.30590396e-03
1.54959786e+00 -1.20718837e+00 -5.44358671e-01 -4.40014005e-01
5.58607519e-01 -6.10485613e-01 1.67805457e+00 1.14702947e-01
-9.56315100e-01 -8.17463934e-01 -9.02490437e-01 -2.70469010e-01
-9.24142838e-01 -3.05118173e-01 3.31307620e-01 4.43245113e-01
-1.03161991e+00 1.17931813e-01 -3.15829724e-01 -6.39090419e-01
3.06115896e-01 1.41514063e-01 -4.17564929e-01 -3.64818394e-01
-1.44049704e+00 1.18927991e+00 4.71414745e-01 -2.23009154e-01
-9.45033312e-01 -8.99804473e-01 -1.18082678e+00 3.78801554e-01
7.82258928e-01 -9.30647671e-01 1.36234331e+00 -1.00343633e+00
-1.50836277e+00 7.96764910e-01 2.06351541e-02 -3.00528556e-01
3.38255078e-01 -3.61691974e-02 -4.19879079e-01 1.10813312e-01
-2.39494182e-02 9.16982532e-01 1.11275339e+00 -9.58365023e-01
-5.59727907e-01 -3.12103838e-01 2.53516138e-01 3.04870814e-01
-4.24583226e-01 -6.67853281e-02 -5.11993825e-01 -4.10844922e-01
-8.53790998e-01 -5.64748526e-01 5.59717864e-02 -2.56011933e-01
-1.21675722e-01 -4.53564763e-01 9.14322793e-01 -3.96220565e-01
8.94768715e-01 -2.15028071e+00 2.53709644e-01 -1.47444874e-01
9.12467912e-02 5.43758452e-01 -6.25295341e-01 5.19824147e-01
-6.09559156e-02 -1.88011408e-01 -2.37986311e-01 -3.51025701e-01
1.62577331e-01 1.88229680e-01 -1.76039457e-01 -9.40791816e-02
4.32186395e-01 1.43925714e+00 -9.31316137e-01 -5.52948296e-01
5.67734540e-01 4.58348840e-01 -4.13558364e-01 3.51228088e-01
-3.99943143e-01 2.19865605e-01 1.05965480e-01 6.91766858e-01
1.91346124e-01 -3.38177025e-01 5.62963299e-02 -3.98930192e-01
-2.31683515e-02 -4.77891892e-01 -1.00540519e+00 1.97220552e+00
-8.10867727e-01 5.16395211e-01 -1.21893302e-01 -1.00016260e+00
5.90005517e-01 5.13212502e-01 6.69203773e-02 -1.08381605e+00
6.66947514e-02 -5.56910003e-04 -1.26359433e-01 -8.33645761e-01
4.89556521e-01 -3.51115823e-01 -3.12103629e-01 4.20653015e-01
8.38593423e-01 2.08575174e-01 2.58460730e-01 3.27177405e-01
8.97532463e-01 2.14398876e-01 5.48761904e-01 2.27943465e-01
3.65390480e-01 3.75317000e-02 1.59598231e-01 7.53205061e-01
-3.62463921e-01 4.35541183e-01 3.00209939e-01 -2.74824202e-01
-9.35164750e-01 -1.20706904e+00 2.03908756e-01 1.92653024e+00
-3.89651768e-02 -1.74094424e-01 -6.50914550e-01 -7.00042963e-01
-5.20372279e-02 1.01797032e+00 -7.76771784e-01 -2.77528763e-01
-2.02639833e-01 -5.38607895e-01 4.68831986e-01 5.37244976e-01
2.75754631e-01 -1.38588524e+00 -9.04460609e-01 1.44762665e-01
-2.13186085e-01 -1.22370934e+00 -5.96560001e-01 4.98730361e-01
-4.58080113e-01 -1.19623697e+00 -1.00924158e+00 -7.63425648e-01
2.86252886e-01 2.47583553e-01 1.11553848e+00 -3.62073302e-01
-5.10862172e-01 8.21455300e-01 -3.02464306e-01 -4.48457003e-01
-1.77241877e-01 -3.99702303e-02 -1.46771684e-01 1.80616647e-01
6.04944706e-01 -3.48889768e-01 -2.97079027e-01 2.09543690e-01
-1.08457804e+00 -1.25048980e-01 8.26016009e-01 1.30205595e+00
4.99579459e-01 -5.73520601e-01 8.01588416e-01 -9.05484378e-01
7.55040407e-01 -6.09000206e-01 -3.60859364e-01 8.65961909e-01
-4.40809160e-01 -1.05391983e-02 4.80232179e-01 -7.05908537e-01
-1.29259264e+00 -4.75738980e-02 7.97696933e-02 -7.74830818e-01
-4.58530068e-01 5.26773453e-01 -1.27060875e-01 -6.54066280e-02
8.70467544e-01 1.63233399e-01 -1.10579550e-01 -2.39383429e-01
9.34911668e-01 5.72760880e-01 6.60500050e-01 -5.50500035e-01
5.87991178e-01 8.12929124e-02 -2.94439048e-01 -9.91241932e-01
-9.65694606e-01 -5.17440140e-01 -6.08457506e-01 -3.89545619e-01
1.24985933e+00 -8.09064090e-01 -7.32208729e-01 3.10507149e-01
-1.08235729e+00 -5.32273829e-01 -4.00673538e-01 2.48078763e-01
-6.47084713e-01 2.31578216e-01 -4.03263897e-01 -5.64679563e-01
-4.74381328e-01 -1.15886748e+00 8.55054080e-01 4.17237520e-01
-8.27295184e-02 -1.10532308e+00 3.29353288e-02 3.83061379e-01
5.61652541e-01 -3.28345299e-02 1.38972640e+00 -7.98896313e-01
-3.06537062e-01 1.39107585e-01 -4.76003408e-01 1.04641609e-01
-1.78067923e-01 -2.87183225e-01 -1.29542351e+00 -1.16015233e-01
-2.66196072e-01 -1.09567332e+00 9.13644016e-01 3.79062504e-01
7.79062748e-01 -3.97541150e-02 5.80580859e-03 5.91510355e-01
1.47711742e+00 1.96213089e-02 3.92506361e-01 2.37822726e-01
7.14387119e-01 7.58653760e-01 5.89653015e-01 1.34373799e-01
5.96850574e-01 6.29234195e-01 5.45850813e-01 5.86787658e-03
-2.68775433e-01 -1.25722557e-01 2.62895882e-01 5.45751274e-01
-4.75137644e-02 1.74890645e-03 -1.24258113e+00 6.23206854e-01
-2.01008081e+00 -1.08547294e+00 3.56666505e-01 2.07108760e+00
8.82139385e-01 -8.35885033e-02 1.77402750e-01 -3.73821318e-01
6.31554008e-01 2.46980339e-01 -6.16861999e-01 -5.59610009e-01
1.02066193e-02 2.28754073e-01 5.74002191e-02 2.94280648e-01
-1.07409561e+00 1.15787959e+00 5.18694115e+00 8.78914356e-01
-1.08343029e+00 5.52086234e-01 3.77158195e-01 -2.90935814e-01
-1.59796640e-01 -1.20333157e-01 -8.47863078e-01 1.79775506e-01
1.08515418e+00 4.29698899e-02 4.90653336e-01 7.66484439e-01
-4.02390927e-01 -1.03700891e-01 -1.20708203e+00 1.29382885e+00
4.02095377e-01 -1.18411708e+00 2.24975586e-01 -2.48085484e-01
7.36651480e-01 3.03770989e-01 1.77898094e-01 1.08638823e+00
1.16747871e-01 -1.01597452e+00 4.75447178e-01 5.70356548e-01
1.12041116e+00 -7.51495898e-01 4.30863261e-01 5.03625572e-01
-1.02698123e+00 -3.57490450e-01 -3.01994234e-01 2.13888660e-01
2.83063114e-01 1.86236173e-01 -6.42446399e-01 5.12136281e-01
7.23189235e-01 1.47913814e-01 -4.14074451e-01 9.12636161e-01
-1.14994839e-01 3.29204619e-01 1.85973242e-01 3.12239155e-02
5.62441885e-01 9.06449035e-02 4.03798431e-01 1.36279333e+00
1.53259888e-01 -6.81649819e-02 3.02991956e-01 7.80738115e-01
-2.47941822e-01 1.87410638e-01 -6.64983988e-01 -3.34286168e-02
2.88785160e-01 1.48265624e+00 -3.36368412e-01 -5.35772085e-01
-1.00145435e+00 9.36932683e-01 7.55504072e-01 6.58080459e-01
-7.38918483e-01 -5.84095240e-01 3.37811828e-01 -2.65168101e-01
4.96258259e-01 -7.05730021e-02 -1.08685598e-01 -1.08324027e+00
-2.96972126e-01 -8.39845240e-01 6.23172581e-01 -7.06499755e-01
-1.52391851e+00 5.95589042e-01 1.60278231e-02 -9.75681126e-01
-4.57687736e-01 -6.70611560e-01 -8.00410748e-01 9.06532407e-01
-1.93469441e+00 -1.46157277e+00 -3.96197528e-01 1.01036799e+00
9.91235614e-01 -3.93358141e-01 1.01342595e+00 3.46485227e-01
-4.86763090e-01 7.30910361e-01 -2.46873602e-01 -6.95378929e-02
1.22680914e+00 -1.34973729e+00 2.43796527e-01 4.32803333e-01
1.66860610e-01 2.94749975e-01 4.62110996e-01 -4.42551494e-01
-1.40721989e+00 -1.14639211e+00 9.10498559e-01 -4.46870089e-01
9.20117795e-01 -2.82422721e-01 -1.08263958e+00 6.69548512e-01
5.56659579e-01 1.51483521e-01 8.55302036e-01 3.06287915e-01
-6.40661120e-01 -7.46702999e-02 -8.16250861e-01 5.36442161e-01
5.77733219e-01 -8.51966321e-01 -8.79319310e-01 2.15586975e-01
6.94041491e-01 -8.37751031e-02 -8.05803359e-01 1.95618615e-01
4.79333997e-01 -7.15608180e-01 9.65656161e-01 -9.08090115e-01
5.40624797e-01 1.37409776e-01 -2.28566945e-01 -1.50671494e+00
-2.06281334e-01 -3.63622367e-01 -3.91298026e-01 1.19829679e+00
5.31017482e-01 -2.71804065e-01 2.31033325e-01 5.60773432e-01
-2.41288483e-01 -7.10564137e-01 -7.72657037e-01 -7.11059570e-01
-2.09871493e-02 -4.58797842e-01 2.53423780e-01 1.07014561e+00
-3.24447006e-02 9.42847371e-01 -5.58967710e-01 -2.07351923e-01
6.77191436e-01 3.67278606e-02 8.21113467e-01 -1.17379844e+00
-4.92625177e-01 -3.67965579e-01 9.13998708e-02 -6.05531037e-01
2.00860322e-01 -9.63239610e-01 2.52279133e-01 -1.55263841e+00
3.93349975e-01 9.05601382e-02 -4.40818250e-01 7.29835212e-01
-3.67596745e-01 7.31607974e-02 4.59930301e-01 5.52594513e-02
-9.32194173e-01 5.24002373e-01 1.14084136e+00 -2.61263251e-01
-1.93773627e-01 -2.33291224e-01 -3.70785147e-01 7.28995442e-01
4.30669636e-01 -1.37371764e-01 -4.86366719e-01 -5.74251533e-01
2.28919521e-01 1.64297357e-01 6.11593366e-01 -7.25887179e-01
5.67085385e-01 -8.89276117e-02 3.36241364e-01 -4.30555701e-01
5.80585480e-01 -7.53633440e-01 -5.13885796e-01 1.05312072e-01
-3.18385899e-01 -3.04524135e-03 2.60151029e-01 7.29254961e-01
-2.61365712e-01 -2.85305262e-01 8.17811310e-01 -2.85387576e-01
-1.54194188e+00 3.86273414e-01 -1.43566802e-01 4.07888591e-01
1.18026912e+00 -1.31633505e-01 -6.84995949e-01 -2.94816941e-01
-9.16026592e-01 5.13107717e-01 3.31139863e-02 7.09375978e-01
6.76733017e-01 -1.15969110e+00 -5.45458555e-01 8.70577991e-02
6.79143846e-01 -2.89475530e-01 7.63246059e-01 1.04831851e+00
2.81271040e-01 4.19557422e-01 -4.59862649e-01 -6.41997993e-01
-1.09593248e+00 7.96140492e-01 1.85958222e-01 -2.69062370e-01
-4.46079433e-01 8.35439801e-01 3.88006419e-01 -7.46810555e-01
4.70680743e-01 1.53705269e-01 -3.51883918e-01 5.33295631e-01
6.02381468e-01 3.28642279e-01 7.55852135e-03 -4.35701847e-01
-1.28257409e-01 4.96027529e-01 -1.65178761e-01 -1.58087224e-01
1.06193304e+00 -6.70592487e-02 3.82474184e-01 8.17283750e-01
1.05342352e+00 -7.34957755e-01 -1.35903847e+00 -6.33312106e-01
1.55280188e-01 7.74158463e-02 -1.39100835e-01 -1.18972600e+00
-6.47751033e-01 1.38277745e+00 5.74915588e-01 -1.46400869e-01
9.94048595e-01 2.16106609e-01 8.36897075e-01 6.44993544e-01
1.08843535e-01 -1.38521373e+00 4.28213924e-01 8.22213650e-01
6.40442908e-01 -1.69048548e+00 -6.60594225e-01 1.60390541e-01
-1.03476417e+00 1.06760645e+00 8.66262436e-01 3.12834144e-01
3.83491725e-01 6.68985024e-02 9.00815427e-02 -1.97452083e-01
-1.00324368e+00 -5.56867063e-01 6.34470940e-01 8.35442543e-01
2.17636347e-01 -1.20778039e-01 2.09414363e-01 7.02542841e-01
3.42794001e-01 -6.88703284e-02 -5.18833660e-03 8.21136057e-01
-5.86070180e-01 -7.93698311e-01 -1.60476074e-01 3.50841910e-01
-1.72980249e-01 -4.23157364e-01 -1.47843048e-01 8.16647112e-01
1.08842120e-01 8.53452981e-01 -3.48961279e-02 -1.85386717e-01
5.27719975e-01 6.64876699e-01 5.12058914e-01 -8.76701593e-01
-7.39421010e-01 -1.13261029e-01 -5.36086969e-03 -5.73212385e-01
-3.06093097e-01 -3.05235267e-01 -1.09409106e+00 5.39503656e-02
-8.96862894e-02 -1.96755618e-01 6.55529678e-01 1.29570770e+00
2.72004843e-01 6.73132300e-01 1.38609037e-01 -8.03106487e-01
-8.27244222e-01 -1.04710245e+00 -2.78942704e-01 4.49367046e-01
2.25780159e-01 -6.56396747e-01 -7.79124722e-02 -8.75612255e-03] | [10.625100135803223, 1.8218022584915161] |
3e82026c-9943-4f46-b722-7e8fa549ac5d | etat-de-lart-en-compression-multi-phrases | null | null | https://aclanthology.org/2021.jeptalnrecital-recital.6 | https://aclanthology.org/2021.jeptalnrecital-recital.6.pdf | Etat de l’art en compression multi-phrases pour la synthèse de documents (State-of-the-art of multi-sentence compression for document summarization) | La compression multi-phrases est utilisée dans différentes tâches de résumé (microblogs, opinions, réunions ou articles de presse). Leur objectif est de proposer une reformulation compressée et grammaticalement correcte des phrases sources tout en gardant les faits principaux. Dans cet article, nous présentons l’état de l’art de la compression multi-phrases en mettant en avant les différents corpus et outils à disposition. Nous axons notre analyse principalement sur la qualité grammaticale et informative plus que sur le taux de compression. | ['Kévin Espasa'] | null | null | null | null | jep-taln-recital-2021-6 | ['sentence-compression'] | ['natural-language-processing'] | [ 1.10067695e-01 -5.40107861e-02 8.85421559e-02 -2.45534733e-01
-6.22139215e-01 -7.47740805e-01 8.00063729e-01 1.17153072e+00
-5.81276596e-01 7.98546076e-01 6.43764734e-01 -6.55210391e-02
-3.02467868e-02 -1.12517691e+00 -9.69118893e-01 -3.29746842e-01
-9.15045366e-02 2.98535138e-01 8.80773962e-02 -5.09001315e-01
5.17855465e-01 2.94533223e-01 -1.18445086e+00 4.78512108e-01
5.34712017e-01 5.37189901e-01 1.93599075e-01 5.93788505e-01
-4.76542294e-01 6.03074551e-01 -6.40785158e-01 -8.65308404e-01
4.14388955e-01 -6.54891908e-01 -7.31395841e-01 -5.66280842e-01
3.94527733e-01 4.02284145e-01 -4.35438275e-01 1.10108280e+00
5.36313392e-02 -1.71322048e-01 1.30541992e+00 -7.41673768e-01
-9.25764918e-01 1.29896367e+00 -3.86295825e-01 1.23514557e+00
7.89829314e-01 -2.04877093e-01 1.30794740e+00 -6.80648088e-01
3.92323434e-01 9.23173845e-01 8.20160627e-01 3.00166219e-01
-6.17133260e-01 -3.12304765e-01 -2.77152121e-01 8.90580192e-03
-8.74353707e-01 -5.31798959e-01 -3.56430829e-01 -1.51361048e-01
9.28918421e-01 6.18665934e-01 5.36894381e-01 1.13456392e+00
8.12359512e-01 6.98024869e-01 1.07373452e+00 -4.64477479e-01
2.01379552e-01 4.30354863e-01 2.59305149e-01 5.80938041e-01
5.14596522e-01 -5.47273159e-02 -6.19990289e-01 -5.87859511e-01
4.15649056e-01 2.89346665e-01 -2.89469093e-01 6.73945427e-01
-8.87164235e-01 8.63308668e-01 2.73836136e-01 4.77730811e-01
-5.26140332e-01 2.04699695e-01 7.72622883e-01 4.16623861e-01
4.57538903e-01 5.48133194e-01 4.68974896e-02 -9.96791780e-01
-7.00744450e-01 5.09631038e-01 1.49544597e+00 1.25389576e+00
3.50132078e-01 -4.63661432e-01 5.56252003e-01 6.88642740e-01
3.16894680e-01 2.74220020e-01 3.11780006e-01 -2.76939690e-01
5.71426272e-01 6.56947047e-02 -4.79440182e-01 -8.68980527e-01
2.43726745e-01 -5.50371349e-01 -5.74902713e-01 -5.21446764e-01
-2.09099099e-01 4.73994389e-02 -9.79354158e-02 1.05979264e+00
-3.92943978e-01 -1.93726704e-01 2.74255183e-02 5.35226762e-01
1.03451982e-01 8.87957811e-01 4.08801824e-01 -7.24797964e-01
1.42232168e+00 -3.89670879e-01 -5.01274168e-01 2.26771031e-02
3.89552504e-01 -1.21061504e+00 6.06904507e-01 6.54924154e-01
-1.20218968e+00 -1.31144196e-01 -1.11253607e+00 -1.88703820e-01
-2.64092207e-01 -4.33914691e-01 -1.31025687e-01 5.70328057e-01
-7.55684316e-01 5.71589291e-01 -4.74972993e-01 -4.83139753e-01
-1.12832814e-01 4.98938300e-02 -1.23183802e-01 4.73440170e-01
-8.67149234e-01 1.14921200e+00 1.08219206e+00 -1.39282614e-01
-1.59902662e-01 -1.02389924e-01 -7.20089972e-01 5.34308732e-01
7.09375262e-01 -4.80529666e-01 1.60475409e+00 -6.99377120e-01
-1.15773988e+00 6.30370200e-01 -5.04611969e-01 -6.81177676e-01
5.19712329e-01 1.97168395e-01 -4.90106106e-01 3.92645746e-01
-2.57934004e-01 2.56734133e-01 3.83037627e-01 -8.92594576e-01
-1.23109639e+00 3.02969813e-01 2.47105211e-01 1.58655778e-01
-6.31441295e-01 1.03838466e-01 -1.40431210e-01 -1.35963786e+00
-5.10147365e-04 -1.00790834e+00 3.75471890e-01 4.91324440e-02
-2.75781780e-01 -1.17194459e-01 9.30503666e-01 -3.70252520e-01
1.44452643e+00 -2.00304079e+00 7.50373065e-01 -5.73095772e-03
5.42981505e-01 5.01861572e-01 -5.08534789e-01 8.95517588e-01
9.10958126e-02 5.69027185e-01 -6.25887036e-01 -2.75209546e-01
2.70204127e-01 5.43411449e-02 -2.96524942e-01 1.97654106e-02
-5.68520844e-01 8.36139679e-01 -9.48251784e-01 -5.91014683e-01
-3.19382012e-01 9.01193693e-02 -6.69432223e-01 -1.95293158e-01
-3.48257899e-01 -4.51616883e-01 -7.30906427e-01 1.29736677e-01
6.64406598e-01 -4.24643695e-01 -5.44351265e-02 -1.57628730e-01
-4.40203071e-01 6.09637618e-01 -9.72746313e-01 1.40312970e+00
-5.01137435e-01 3.22783351e-01 1.74984753e-01 -5.88999152e-01
1.33230627e+00 4.39182788e-01 6.51769817e-01 -2.34218523e-01
7.59763345e-02 7.71077454e-01 -5.64899087e-01 -5.64617142e-02
8.94035757e-01 -4.45829540e-01 2.69513339e-01 1.19339609e+00
-3.17118555e-01 -2.43802711e-01 1.07410049e+00 3.45703177e-02
1.40447056e+00 -3.18976343e-01 3.09010923e-01 -3.22698146e-01
6.63005412e-01 -1.11216381e-01 -2.74435282e-01 2.67559737e-01
4.67315838e-02 7.97959268e-02 5.57134390e-01 -1.96068048e-01
-1.80366075e+00 -6.85046673e-01 -1.24715656e-01 7.22279906e-01
-4.84191477e-01 -8.87595654e-01 -5.78676760e-01 -1.04914761e+00
1.47184163e-01 1.06219757e+00 -3.68355811e-01 4.70092557e-02
-3.97514641e-01 -8.51255774e-01 5.55030286e-01 4.78298992e-01
5.51197946e-01 -6.17330253e-01 -9.63839352e-01 3.85216296e-01
-5.61092854e-01 -9.61906850e-01 -1.98949695e-01 -1.33110672e-01
-7.06281304e-01 -6.50835037e-01 -6.98670149e-01 -2.92648692e-02
2.44900107e-01 3.31051260e-01 1.37028992e+00 -2.46218741e-01
-1.04547299e-01 2.33862117e-01 -1.10908449e+00 -6.17828846e-01
-1.24997926e+00 4.00060773e-01 -8.84687155e-02 -5.56614816e-01
6.77938163e-01 -6.69008076e-01 -6.13880932e-01 1.23058274e-01
-1.19568014e+00 -5.03670096e-01 8.60982001e-01 2.10886374e-01
3.15816194e-01 -5.62464632e-02 2.96088308e-01 -9.09178257e-01
1.50283492e+00 -8.86619508e-01 -9.50043127e-02 -1.70967951e-02
-5.12977183e-01 5.30267879e-02 1.20014071e+00 4.91449624e-01
-7.93146849e-01 -5.74452102e-01 -6.98972404e-01 1.61744822e-02
-3.35865259e-01 1.04166937e+00 4.29461420e-01 6.35469854e-01
9.98243690e-01 7.49215364e-01 -4.39458907e-01 -6.56545043e-01
2.26107806e-01 7.11422980e-01 2.32251525e-01 -7.32724249e-01
4.24913943e-01 2.24632412e-01 -3.94503146e-01 -4.61771429e-01
-2.03128457e-01 -3.41236621e-01 -6.59329355e-01 8.63801166e-02
7.30065286e-01 -4.71469611e-01 -2.40945816e-01 -1.84355319e-01
-1.36705148e+00 3.49254534e-02 -3.60107571e-01 7.42012322e-01
-4.45270985e-01 -6.42533274e-03 -1.07639968e+00 -3.38412136e-01
-3.02556455e-01 -4.21346128e-01 8.14762414e-01 -3.36881936e-01
-2.64179353e-02 -8.06045413e-01 7.80796587e-01 -2.13237256e-01
2.39709154e-01 -3.01981479e-01 9.53589320e-01 -2.96557575e-01
-9.38160896e-01 -2.05206528e-01 2.13578776e-01 3.51817042e-01
4.67177331e-01 1.35335416e-01 -2.55193621e-01 -3.52481931e-01
-3.09778959e-01 3.22547227e-01 8.48872960e-01 7.18269870e-02
9.54064012e-01 -7.39211798e-01 -5.57701997e-02 1.86750308e-01
2.06143594e+00 3.46346766e-01 3.41116399e-01 2.40128145e-01
-5.99153161e-01 5.17781019e-01 7.56629884e-01 1.33697599e-01
-8.09595883e-02 4.60153967e-01 1.12344421e-01 1.05374181e+00
7.86982894e-01 -1.16308264e-01 2.86850065e-01 1.69521403e+00
-3.90729100e-01 -8.66242707e-01 -9.03771341e-01 3.98916483e-01
-1.44650257e+00 -6.05593443e-01 -3.59656841e-01 1.85990965e+00
7.73436129e-01 4.46600586e-01 4.79087055e-01 2.69922651e-02
4.92908955e-01 6.74208283e-01 2.43055433e-01 -1.27296746e+00
-1.08712800e-01 2.09443584e-01 4.24800694e-01 4.49168652e-01
-4.66719568e-01 3.10841829e-01 6.44827414e+00 6.88745856e-01
-1.09929991e+00 -9.59223434e-02 5.06968856e-01 -1.35448650e-01
-7.85431564e-01 -3.95317793e-01 -7.68265963e-01 9.54185128e-01
1.80313230e+00 -7.66503632e-01 4.09789979e-01 4.58044857e-01
-7.00177848e-02 -1.03900567e-01 -1.06178558e+00 7.37779975e-01
4.44747090e-01 -1.76067448e+00 -2.80651152e-01 -3.09248984e-01
-1.15645528e-01 1.89698443e-01 2.69321501e-01 5.51894903e-01
1.59124926e-01 -9.41535413e-01 7.43964493e-01 3.51669669e-01
6.78569198e-01 -6.94749713e-01 -6.86967820e-02 3.57915998e-01
-7.25058496e-01 7.09210858e-02 -5.92146695e-01 1.81717217e-01
8.14466298e-01 2.94560373e-01 -5.09462953e-01 1.10933685e+00
7.01193437e-02 9.04277802e-01 -5.32517016e-01 1.25301504e+00
1.19916936e-02 8.93163204e-01 -8.30067933e-01 -7.55610466e-01
5.75231254e-01 -3.84872258e-01 7.90406406e-01 1.74982047e+00
7.45504320e-01 8.80545080e-01 -8.10134768e-01 7.75996327e-01
-1.34707332e-01 7.45317578e-01 -9.48852450e-02 -3.68782520e-01
2.62397230e-01 3.23684841e-01 -9.64181662e-01 -4.02912647e-02
-4.02369291e-01 7.15818226e-01 2.26453636e-02 -3.75157952e-01
-4.56758380e-01 -6.50828481e-01 -2.00794637e-03 -3.82132493e-02
7.61756953e-03 -8.03010225e-01 -3.75931710e-02 -1.19585657e+00
-2.60575414e-01 -1.03135681e+00 -4.70183976e-03 -4.96676296e-01
-9.46087539e-01 8.05944145e-01 8.58217001e-01 -7.83551872e-01
-7.92586565e-01 -3.04348558e-01 -2.11895496e-01 4.65929717e-01
-9.27654684e-01 -1.10058568e-01 5.78325748e-01 7.72228390e-02
7.32432902e-01 -1.66133404e-01 1.27880287e+00 7.24692643e-01
-6.87686563e-01 -8.55737890e-04 5.70367813e-01 -4.56761390e-01
5.54973543e-01 -1.24839580e+00 7.51122236e-01 5.81745803e-01
5.37939072e-01 8.11215878e-01 8.96537900e-01 -6.61206603e-01
-1.11358500e+00 -6.15746081e-01 1.11782539e+00 -1.88552335e-01
7.40799427e-01 -8.18394795e-02 -8.25257838e-01 1.45304906e+00
5.83266199e-01 -5.22936225e-01 3.15062642e-01 5.91155961e-02
-1.35087356e-01 2.67784949e-02 -7.26210594e-01 5.83073199e-01
9.59193483e-02 -3.14374924e-01 -5.86707413e-01 3.08769345e-01
4.41640973e-01 -7.47107387e-01 -1.05292928e+00 3.27246487e-01
-8.29235092e-02 -1.28207803e+00 3.56550753e-01 -1.50064826e-01
1.06772316e+00 1.55248687e-01 5.70197888e-02 -1.06010461e+00
4.39936340e-01 -5.79739094e-01 -5.69084823e-01 8.59714508e-01
2.36625075e-01 -2.22461939e-01 5.56764603e-01 -9.78731886e-02
7.86640719e-02 -4.58508581e-01 -1.18786192e+00 -1.79798603e-01
2.76200831e-01 -8.93321753e-01 5.78114569e-01 6.57333136e-01
-2.35265475e-02 4.53236818e-01 -1.06849983e-01 -3.85529995e-01
2.13271216e-01 -6.59240782e-02 2.80385971e-01 -5.38272262e-01
-5.13128042e-01 -2.82790810e-01 -3.32273126e-01 -1.33135128e+00
2.08170954e-02 -9.79310811e-01 -7.53513873e-01 -1.32430637e+00
3.72484028e-01 -9.45745930e-02 -2.84334868e-01 1.78500235e-01
2.24803403e-01 8.08158666e-02 9.31825757e-01 2.43636817e-01
-7.75831342e-01 4.08693910e-01 9.50134814e-01 1.94422901e-01
3.17653149e-01 1.21777140e-01 -3.28323245e-01 4.86335397e-01
4.97058600e-01 -7.66445160e-01 -3.05207193e-01 -4.02094513e-01
8.70076239e-01 -1.35368243e-01 -7.66010880e-02 -9.92628694e-01
5.28972089e-01 -1.02423161e-01 7.23422915e-02 -9.21657383e-01
8.45352888e-01 -7.33443618e-01 1.55184314e-01 1.64839074e-01
-6.35043204e-01 6.43656790e-01 4.83894646e-01 6.14833295e-01
-5.60237527e-01 -8.10348809e-01 7.49873638e-01 -2.64406592e-01
-7.57229269e-01 5.09863317e-01 -6.80294871e-01 4.96814221e-01
7.27322876e-01 -2.27457285e-01 -1.26632094e-01 -1.72719836e-01
-6.66993380e-01 -1.15755022e-01 5.07227600e-01 9.23871323e-02
7.04757929e-01 -8.19992423e-01 -1.65826321e+00 1.00642227e-01
-2.76257426e-01 -4.69776064e-01 6.13205552e-01 2.39033684e-01
-1.48738658e+00 7.72873223e-01 -4.21182334e-01 -8.07223856e-01
-1.79003906e+00 4.97442424e-01 1.20631874e-01 2.98632056e-01
-1.00553858e+00 1.09624374e+00 -1.06831992e+00 2.59284794e-01
-3.81319791e-01 -2.94971138e-01 2.82635599e-01 -2.31186077e-01
2.70179927e-01 8.29576612e-01 -1.48637190e-01 -4.14649248e-01
6.11350164e-02 3.82968187e-01 -4.37930197e-01 -7.11890042e-01
1.26842666e+00 -2.32447475e-01 -9.12073731e-01 4.30504382e-01
1.63026392e+00 7.07052469e-01 -1.12265444e+00 7.32187033e-01
-3.71648534e-03 -4.92567003e-01 -3.83790374e-01 -6.66880786e-01
-7.03905821e-01 4.52953011e-01 3.62918884e-01 5.99487662e-01
8.20323944e-01 -4.21175927e-01 9.76022482e-01 6.54928505e-01
5.94813466e-01 -9.61439610e-01 -3.74995917e-01 7.26345181e-01
1.03546929e+00 -8.33194494e-01 1.28855124e-01 -2.88258314e-01
-3.43109071e-01 1.54183233e+00 -4.92658198e-01 1.98861491e-02
1.04037511e+00 2.81287014e-01 -3.21438722e-02 -2.33696580e-01
-9.80126321e-01 3.33604842e-01 2.90301353e-01 3.76387574e-02
7.73812711e-01 -2.42435709e-01 -1.37297463e+00 3.51015367e-02
-8.66369903e-01 1.84131742e-01 8.52711380e-01 1.38345575e+00
-3.19668144e-01 -1.22403717e+00 -3.96216393e-01 3.46353531e-01
-9.68148530e-01 -3.22471976e-01 -5.31244099e-01 3.90855372e-01
3.68830785e-02 9.10752118e-01 7.30630457e-02 -4.29150045e-01
6.33962393e-01 -1.86368898e-01 1.56359658e-01 -8.73465002e-01
-1.21686220e+00 1.00462306e-02 7.37448037e-02 -2.03644127e-01
-9.01678264e-01 -4.52986777e-01 -4.05633509e-01 -1.20057857e+00
-2.19113722e-01 9.51003551e-01 7.36181080e-01 1.00232947e+00
1.11654282e-01 4.73844558e-01 1.18334852e-01 -3.73969674e-01
-5.39657354e-01 -8.89049947e-01 -1.02674222e+00 7.14070499e-02
3.99973929e-01 -7.00864419e-02 -2.07272634e-01 -1.63805440e-01] | [14.097638130187988, 13.307659149169922] |
7d2d1f13-9862-4618-adc2-281afd455cca | heavy-tails-in-sgd-and-compressibility-of | 2106.03795 | null | https://arxiv.org/abs/2106.03795v1 | https://arxiv.org/pdf/2106.03795v1.pdf | Heavy Tails in SGD and Compressibility of Overparametrized Neural Networks | Neural network compression techniques have become increasingly popular as they can drastically reduce the storage and computation requirements for very large networks. Recent empirical studies have illustrated that even simple pruning strategies can be surprisingly effective, and several theoretical studies have shown that compressible networks (in specific senses) should achieve a low generalization error. Yet, a theoretical characterization of the underlying cause that makes the networks amenable to such simple compression schemes is still missing. In this study, we address this fundamental question and reveal that the dynamics of the training algorithm has a key role in obtaining such compressible networks. Focusing our attention on stochastic gradient descent (SGD), our main contribution is to link compressibility to two recently established properties of SGD: (i) as the network size goes to infinity, the system can converge to a mean-field limit, where the network weights behave independently, (ii) for a large step-size/batch-size ratio, the SGD iterates can converge to a heavy-tailed stationary distribution. In the case where these two phenomena occur simultaneously, we prove that the networks are guaranteed to be '$\ell_p$-compressible', and the compression errors of different pruning techniques (magnitude, singular value, or node pruning) become arbitrarily small as the network size increases. We further prove generalization bounds adapted to our theoretical framework, which indeed confirm that the generalization error will be lower for more compressible networks. Our theory and numerical study on various neural networks show that large step-size/batch-size ratios introduce heavy-tails, which, in combination with overparametrization, result in compressibility. | ['Umut Şimşekli', 'Gaël Richard', 'Murat A. Erdogdu', 'Milad Sefidgaran', 'Melih Barsbey'] | 2021-06-07 | null | http://proceedings.neurips.cc/paper/2021/hash/f5c3dd7514bf620a1b85450d2ae374b1-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/f5c3dd7514bf620a1b85450d2ae374b1-Paper.pdf | neurips-2021-12 | ['neural-network-compression', 'neural-network-compression'] | ['methodology', 'miscellaneous'] | [ 1.75044045e-01 3.60032879e-02 -1.65035367e-01 -1.87066257e-01
-6.25153929e-02 -2.62327701e-01 1.40651867e-01 8.16377029e-02
-6.65153801e-01 8.95386755e-01 -3.67440641e-01 -5.12132823e-01
-5.99519193e-01 -7.11455464e-01 -9.42467391e-01 -1.01067054e+00
-4.92437959e-01 3.78311276e-01 1.81293637e-01 -1.62033781e-01
-3.32029536e-03 7.62855768e-01 -1.64276254e+00 -3.74140769e-01
7.26410210e-01 1.14510190e+00 2.39882410e-01 9.19601142e-01
1.00290708e-01 6.56743109e-01 -3.32713366e-01 -4.52827245e-01
7.43614376e-01 -3.87537658e-01 -7.16008306e-01 -2.63339996e-01
4.50670600e-01 -2.30426595e-01 -6.53133452e-01 1.59523523e+00
3.17634195e-01 5.17883301e-01 4.75134283e-01 -1.01811051e+00
-3.03684562e-01 8.40544462e-01 -3.13379407e-01 5.18396795e-01
-5.80183148e-01 -4.02785018e-02 8.31242085e-01 -3.70861262e-01
5.24511158e-01 8.86696756e-01 1.10525692e+00 7.22565174e-01
-1.36057711e+00 -6.35905325e-01 1.19746223e-01 -1.34150788e-01
-1.41304457e+00 -4.65683877e-01 7.17821360e-01 -6.87855408e-02
8.95575166e-01 2.89165497e-01 7.29650497e-01 4.86120731e-01
1.61856532e-01 4.52853262e-01 6.60360813e-01 -6.06478631e-01
4.64951336e-01 9.69586074e-02 2.19008237e-01 9.71150935e-01
7.67709732e-01 -7.57990703e-02 -2.70060956e-01 3.66764367e-02
9.37026024e-01 1.55646488e-01 -3.69519949e-01 -3.67199600e-01
-6.47639036e-01 8.86238813e-01 3.35894316e-01 4.83077407e-01
-2.53314197e-01 4.10470456e-01 5.14395058e-01 6.95800126e-01
4.76247311e-01 3.55510682e-01 -4.75432158e-01 -1.91305310e-01
-1.08557928e+00 3.45005274e-01 1.20102751e+00 8.53936076e-01
4.12495911e-01 2.59366035e-01 4.10490960e-01 9.17048931e-01
-1.68066293e-01 4.20354098e-01 7.12989151e-01 -1.21034825e+00
5.97949743e-01 2.32054502e-01 -1.86090320e-01 -1.16154373e+00
-3.60559165e-01 -8.49670589e-01 -1.46879625e+00 9.00666192e-02
6.69547677e-01 -2.38739878e-01 -5.47424555e-01 2.17736745e+00
8.11279193e-02 -1.44565359e-01 8.91085267e-02 6.04069054e-01
1.83418334e-01 3.83990198e-01 -1.30678102e-01 -5.27879715e-01
7.66085207e-01 -4.65441763e-01 -5.00720203e-01 -3.06917969e-02
8.71226609e-01 -2.04791665e-01 1.01368427e+00 3.23533148e-01
-1.57310641e+00 -2.02265680e-01 -1.15920639e+00 2.07113475e-01
-2.20631003e-01 -8.94335806e-02 6.38223946e-01 6.32918775e-01
-1.06774879e+00 1.47412539e+00 -1.15293717e+00 -1.33485168e-01
6.53866112e-01 5.74652791e-01 -2.46610969e-01 6.41354080e-03
-9.81370389e-01 6.59742892e-01 6.01949930e-01 1.41065866e-01
-5.26249468e-01 -6.90412402e-01 -4.74777907e-01 4.75299299e-01
5.35714179e-02 -6.62795246e-01 1.11161268e+00 -8.95955861e-01
-1.21424472e+00 5.70569158e-01 -3.20416950e-02 -1.02577448e+00
5.42628348e-01 1.73468634e-01 1.25526085e-01 5.16144753e-01
-3.47489089e-01 3.69647980e-01 7.26395905e-01 -6.96481287e-01
-4.41091061e-01 -4.78079349e-01 -5.54885529e-02 1.18961558e-01
-9.74351227e-01 -2.89043576e-01 -4.73639555e-02 -5.62590241e-01
2.38548309e-01 -8.05692434e-01 -2.91779250e-01 2.55124986e-01
-1.75419450e-01 -1.33834958e-01 5.48861980e-01 -2.92181492e-01
1.30077553e+00 -2.13583469e+00 9.01437700e-02 3.19083303e-01
5.80616653e-01 4.31704074e-01 7.90607482e-02 1.31030276e-01
-1.78524226e-01 3.62362862e-01 -3.37094605e-01 -3.92875880e-01
-5.80083579e-02 4.52717572e-01 -2.75726408e-01 6.92216754e-01
-2.06166640e-01 6.28317177e-01 -5.43015242e-01 -3.62535745e-01
-1.06538177e-01 4.52644706e-01 -7.77185857e-01 -3.30441356e-01
2.74686925e-02 -3.84192258e-01 -2.85496175e-01 1.06187440e-01
6.10091448e-01 -4.54384655e-01 1.22560471e-01 9.85009819e-02
2.21896425e-01 1.60348132e-01 -1.18024206e+00 1.01541162e+00
-3.83508593e-01 8.12531531e-01 4.29887176e-01 -1.63568389e+00
5.70467770e-01 1.84465617e-01 3.96280527e-01 -2.42394745e-01
3.06722075e-01 3.10929865e-01 6.95185363e-02 -1.93950370e-01
1.15910575e-01 -4.57067043e-01 4.28952754e-01 4.32765216e-01
1.06858440e-01 6.95122108e-02 5.78421891e-01 2.82076210e-01
1.17928517e+00 -6.26135468e-01 -5.00118174e-02 -4.71970499e-01
1.54785573e-01 -2.59332567e-01 4.13074911e-01 9.55530763e-01
-1.94759980e-01 2.72114396e-01 8.37597668e-01 -3.33948821e-01
-1.43907666e+00 -7.14227378e-01 -3.17742258e-01 9.35110629e-01
-8.87387991e-02 -1.83711663e-01 -1.09697843e+00 -5.68624921e-02
1.10742666e-01 3.21715176e-01 -6.96683884e-01 -4.15512502e-01
-5.41935802e-01 -9.82148170e-01 5.70479214e-01 4.60564643e-01
7.74177909e-01 -7.22456276e-01 -6.89710319e-01 1.05298400e-01
3.02254856e-01 -1.05286551e+00 -2.75548309e-01 5.82944751e-01
-1.67975092e+00 -9.63741124e-01 -8.75256658e-01 -6.74672365e-01
7.38585114e-01 2.04617769e-01 8.54447484e-01 4.65479553e-01
-1.51302606e-01 1.30466402e-01 9.96779725e-02 -1.58873424e-01
-4.23579067e-01 3.75155032e-01 4.09755439e-01 -3.91027838e-01
7.45438188e-02 -1.16791546e+00 -5.16050875e-01 6.01420775e-02
-1.09553242e+00 -6.39810562e-02 4.05898452e-01 7.44711041e-01
5.80535233e-01 6.52141750e-01 4.03959751e-01 -7.62990952e-01
9.35458541e-01 -2.34977424e-01 -8.52211714e-01 2.92010065e-02
-9.16551650e-01 3.37234735e-01 1.09979284e+00 -7.60857224e-01
-5.17672300e-01 -1.05227239e-01 -1.25762105e-01 -7.99237072e-01
9.32013765e-02 6.43045604e-01 2.47671902e-01 -3.35060924e-01
7.26540446e-01 4.32971239e-01 3.01242948e-01 -5.68104565e-01
2.83122137e-02 3.15201372e-01 4.55711842e-01 -5.08888483e-01
8.61350358e-01 5.18187523e-01 5.03759623e-01 -1.06800032e+00
-8.01067293e-01 -1.64567739e-01 -4.88033712e-01 1.02802187e-01
2.26508230e-01 -5.13059855e-01 -9.18191195e-01 2.64082491e-01
-8.10370147e-01 -4.41447943e-01 -7.86928296e-01 5.37973046e-01
-5.56393981e-01 3.89827639e-01 -1.02710354e+00 -1.00483239e+00
-4.80389446e-01 -7.53294766e-01 1.28261209e-01 2.24568695e-01
1.34543926e-01 -1.11121738e+00 -1.40717298e-01 -2.81450331e-01
5.94699502e-01 -8.14312026e-02 1.11212933e+00 -8.28256369e-01
-1.88234180e-01 -4.21003073e-01 -1.81624323e-01 6.74797475e-01
-4.99402076e-01 -1.19188815e-01 -4.83018696e-01 -3.17324489e-01
4.87833500e-01 -5.34620546e-02 8.55866611e-01 6.84078515e-01
1.33834934e+00 -8.14147949e-01 -1.19977795e-01 1.00285101e+00
1.50299227e+00 -9.53058079e-02 2.88819700e-01 7.50515535e-02
4.03131813e-01 4.08152252e-01 -1.99119031e-01 3.37618113e-01
-4.02904630e-01 1.31448016e-01 4.12586987e-01 3.77398372e-01
-2.86825877e-02 -1.54042676e-01 1.76644087e-01 1.27577984e+00
-3.63058209e-01 5.10693416e-02 -5.82743943e-01 4.07404482e-01
-1.58391273e+00 -1.13813281e+00 1.20730609e-01 2.38133144e+00
1.11862457e+00 4.90665883e-01 1.58117861e-01 4.76048887e-01
5.65037310e-01 4.27300073e-02 -8.00377965e-01 -3.87325287e-01
-2.73534656e-01 1.96466237e-01 8.85493577e-01 5.04173577e-01
-7.38229036e-01 5.15474260e-01 6.84412766e+00 1.04980659e+00
-1.10783005e+00 2.30973706e-01 6.88376546e-01 -5.63008249e-01
-9.83290076e-02 -3.97902459e-01 -9.98869658e-01 4.59417462e-01
1.29101086e+00 -4.37587947e-01 5.97216725e-01 1.17014599e+00
-2.59214360e-02 5.62455952e-02 -8.96072447e-01 1.09219408e+00
-2.12721080e-01 -1.52147913e+00 -1.20321490e-01 3.65667701e-01
6.93355799e-01 3.31398875e-01 9.05384570e-02 7.26176798e-02
3.24663848e-01 -9.72596347e-01 3.21290314e-01 1.88794315e-01
7.46483445e-01 -1.08379662e+00 5.85240722e-01 6.67292237e-01
-9.91665483e-01 -2.41391405e-01 -1.02309203e+00 -1.61738992e-01
-1.29676133e-01 9.94198978e-01 -2.90196806e-01 -5.55121489e-02
6.23551309e-01 4.29226071e-01 -3.57539803e-01 1.08562016e+00
2.73580104e-01 7.50651002e-01 -9.20662582e-01 -3.25430721e-01
1.14733316e-01 -4.47310627e-01 5.63387215e-01 1.04123473e+00
1.98951557e-01 1.67827055e-01 -4.09734845e-01 8.41056108e-01
-4.35635239e-01 -6.75173849e-02 -7.04487801e-01 -3.32412243e-01
4.03999299e-01 9.15609777e-01 -1.03570664e+00 -3.18218350e-01
-7.04806596e-02 6.16371155e-01 5.63394845e-01 3.55054379e-01
-3.80957276e-01 -7.48441219e-01 5.72668731e-01 1.31094515e-01
6.10709667e-01 -3.69800091e-01 -3.13955575e-01 -1.13715374e+00
2.37340406e-01 -5.27694762e-01 7.16137439e-02 -2.61943817e-01
-9.76318061e-01 4.50622320e-01 -1.42538734e-03 -6.89040661e-01
-2.13517487e-01 -6.71365499e-01 -4.32550728e-01 5.30922472e-01
-1.02898955e+00 -2.39043027e-01 2.49925509e-01 5.94967544e-01
3.31825644e-01 -1.43475875e-01 5.93546033e-01 4.92083132e-01
-7.32044816e-01 9.03549075e-01 6.90785110e-01 1.33810669e-01
-1.68484211e-01 -9.56042051e-01 -9.80472472e-03 9.28080618e-01
1.97306067e-01 9.01963413e-01 9.65522766e-01 -4.57081020e-01
-1.47246718e+00 -8.50447834e-01 8.65309119e-01 2.51225214e-02
7.57794619e-01 -2.67178863e-01 -1.08989906e+00 5.61808586e-01
-4.83333647e-01 2.64266044e-01 3.96805048e-01 1.98359922e-01
-1.44303367e-01 -3.23059350e-01 -1.21348894e+00 5.98542631e-01
1.22607756e+00 -4.43432271e-01 -2.86872208e-01 6.56861842e-01
5.77695131e-01 -2.39073917e-01 -9.08167601e-01 1.50703683e-01
4.88960832e-01 -9.91446257e-01 8.88344646e-01 -8.98963809e-01
6.31082714e-01 3.80859762e-01 -1.86792731e-01 -9.40400302e-01
5.97545989e-02 -9.17686045e-01 -6.78659320e-01 6.82660103e-01
2.40844220e-01 -8.17813396e-01 1.05175149e+00 8.06583941e-01
1.34253785e-01 -1.23341787e+00 -1.23318350e+00 -1.25457990e+00
4.33394969e-01 -4.37592566e-01 1.70034096e-01 6.50231659e-01
1.65654555e-01 1.83437578e-02 -7.82721117e-02 -1.92558631e-01
6.55816376e-01 -6.27800599e-02 1.25773266e-01 -1.38437891e+00
-4.87534732e-01 -9.21321213e-01 -4.70140368e-01 -1.40642333e+00
1.73630729e-01 -8.83730710e-01 -7.41841495e-02 -8.93555224e-01
1.73278287e-01 -6.71101689e-01 -2.43917450e-01 1.29027605e-01
2.85100311e-01 1.25730589e-01 1.87941477e-01 4.65930313e-01
-4.15321618e-01 4.66687739e-01 1.00824881e+00 2.18581498e-01
-2.77755439e-01 4.48237747e-01 -5.55823267e-01 9.59547579e-01
8.19914639e-01 -5.85955679e-01 -5.52525759e-01 -3.29594076e-01
5.48964500e-01 5.38729578e-02 1.92745060e-01 -1.20519459e+00
5.43531716e-01 1.58471182e-01 9.39783603e-02 -2.47305110e-01
2.25251406e-01 -8.09915960e-01 -2.11407691e-01 8.28568935e-01
-7.73275197e-01 5.67145646e-02 -1.15756681e-02 6.66642129e-01
8.95451978e-02 -8.62392604e-01 9.12722528e-01 -6.43559843e-02
2.77790934e-01 5.66571832e-01 -3.75587285e-01 3.46749067e-01
6.74099326e-01 -2.49132201e-01 1.11781828e-01 -5.27404666e-01
-7.15096533e-01 -8.68311450e-02 2.70136207e-01 -3.89436573e-01
4.29173708e-01 -1.10914230e+00 -2.49598131e-01 1.78687036e-01
-8.22263300e-01 2.01322466e-01 9.48607326e-02 1.00383914e+00
-5.78000247e-01 3.88828397e-01 1.00554749e-01 -3.96676660e-01
-1.28708768e+00 5.47218561e-01 5.81906438e-01 -2.20610708e-01
-8.84525180e-01 1.03571498e+00 -3.24805856e-01 -5.12087569e-02
7.00313389e-01 -4.27048236e-01 2.65463680e-01 -1.47676483e-01
5.31306148e-01 4.82477725e-01 1.05309971e-01 -1.71318173e-01
1.71294853e-01 4.34106648e-01 -2.52934545e-01 -2.36859825e-02
1.73146224e+00 1.10305194e-03 -1.56225835e-03 2.34775394e-01
1.58811283e+00 -3.56244951e-01 -1.34216642e+00 -3.13451260e-01
-3.06230366e-01 -1.27103359e-01 1.55620858e-01 4.42137197e-02
-1.57489645e+00 1.11674118e+00 4.70768899e-01 7.36604273e-01
1.09972060e+00 -2.93243695e-02 8.67231071e-01 9.91310358e-01
1.43989027e-01 -1.24757385e+00 -3.53284329e-01 6.60096765e-01
4.94606376e-01 -7.83334851e-01 1.71733260e-01 -2.08460748e-01
2.94307973e-02 1.21970057e+00 1.83506370e-01 -4.33988720e-01
8.98014128e-01 2.87113518e-01 -7.97086835e-01 -7.11315423e-02
-7.73108304e-01 2.68896639e-01 -1.45736441e-01 4.38313693e-01
9.90290046e-02 -2.98484236e-01 -2.23391891e-01 4.67933178e-01
-6.07519984e-01 1.02692440e-01 3.88634145e-01 7.41908491e-01
-7.60214269e-01 -5.90378344e-01 1.26776263e-01 9.14218724e-01
-6.72833264e-01 -1.88056707e-01 2.23478675e-02 8.74513090e-01
-3.57012868e-01 3.67576182e-01 2.03102246e-01 -1.44654155e-01
1.87750217e-02 1.67088062e-01 7.99661458e-01 -2.23784983e-01
-2.46611893e-01 -3.34347010e-01 -2.38644123e-01 -4.03715372e-01
-2.58495003e-01 -4.91307467e-01 -1.11852384e+00 -9.47130859e-01
-4.66567934e-01 2.23769218e-01 6.49505496e-01 1.06551242e+00
2.65309811e-01 2.79643387e-01 2.46747598e-01 -6.52648985e-01
-1.28338063e+00 -7.50634313e-01 -8.75026464e-01 2.69360125e-01
4.06291097e-01 -2.53554285e-01 -1.06260204e+00 -1.17821828e-03] | [8.06533145904541, 3.4932730197906494] |
f8299c1c-6902-4415-b70b-47162e6e09a4 | a-lightweight-domain-adversarial-neural | 2305.07446 | null | https://arxiv.org/abs/2305.07446v1 | https://arxiv.org/pdf/2305.07446v1.pdf | A Lightweight Domain Adversarial Neural Network Based on Knowledge Distillation for EEG-based Cross-subject Emotion Recognition | Individual differences of Electroencephalogram (EEG) could cause the domain shift which would significantly degrade the performance of cross-subject strategy. The domain adversarial neural networks (DANN), where the classification loss and domain loss jointly update the parameters of feature extractor, are adopted to deal with the domain shift. However, limited EEG data quantity and strong individual difference are challenges for the DANN with cumbersome feature extractor. In this work, we propose knowledge distillation (KD) based lightweight DANN to enhance cross-subject EEG-based emotion recognition. Specifically, the teacher model with strong context learning ability is utilized to learn complex temporal dynamics and spatial correlations of EEG, and robust lightweight student model is guided by the teacher model to learn more difficult domain-invariant features. In the feature-based KD framework, a transformer-based hierarchical temporalspatial learning model is served as the teacher model. The student model, which is composed of Bi-LSTM units, is a lightweight version of the teacher model. Hence, the student model could be supervised to mimic the robust feature representations of teacher model by leveraging complementary latent temporal features and spatial features. In the DANN-based cross-subject emotion recognition, we combine the obtained student model and a lightweight temporal-spatial feature interaction module as the feature extractor. And the feature aggregation is fed to the emotion classifier and domain classifier for domain-invariant feature learning. To verify the effectiveness of the proposed method, we conduct the subject-independent experiments on the public dataset DEAP with arousal and valence classification. The outstanding performance and t-SNE visualization of latent features verify the advantage and effectiveness of the proposed method. | ['Zhiqun Pan', 'Yiheng Tang', 'Jiapeng Zhang', 'Yongxiong Wang', 'Zhe Wang'] | 2023-05-12 | null | null | null | null | ['eeg', 'eeg'] | ['methodology', 'time-series'] | [-6.62903767e-03 -3.89853001e-01 2.24924341e-01 -3.33292991e-01
-6.66415811e-01 -5.30313253e-01 3.29437912e-01 -2.75065005e-01
-3.11587542e-01 8.59535098e-01 4.72666137e-02 2.43281171e-01
-5.09908855e-01 -3.68178755e-01 -6.20256424e-01 -1.19222355e+00
-3.56042534e-01 -3.65462214e-01 -3.97824273e-02 -1.92317113e-01
-8.81329998e-02 3.41094136e-01 -1.30611479e+00 3.27835858e-01
1.25859439e+00 1.54893017e+00 4.79833186e-02 2.45803781e-02
7.73512274e-02 3.35687280e-01 -6.49369836e-01 2.67676152e-02
7.47028440e-02 -5.19716799e-01 -2.84309030e-01 -1.86255425e-01
-1.67441651e-01 -5.10547347e-02 -4.31360036e-01 1.06432605e+00
9.65698779e-01 2.77812511e-01 6.84341073e-01 -1.60810673e+00
-5.89937627e-01 2.99232155e-01 -5.48337102e-01 4.30435687e-01
3.77979696e-01 2.23779008e-01 3.27898860e-01 -8.63761008e-01
5.80147654e-02 9.02714849e-01 5.08442223e-01 4.77106035e-01
-1.04515803e+00 -1.34563494e+00 4.13546264e-01 5.13459265e-01
-1.65292537e+00 -2.38740161e-01 1.37917268e+00 -3.58394414e-01
6.45601749e-01 1.27036765e-01 9.53943074e-01 1.53913414e+00
4.20824617e-01 7.55029500e-01 1.59389079e+00 1.26522288e-01
2.52616227e-01 4.53955770e-01 9.85788554e-02 3.65144372e-01
-3.81424934e-01 2.93162525e-01 -6.88238204e-01 -2.06683680e-01
5.69574177e-01 4.65335995e-01 -4.73345011e-01 -3.03146660e-01
-9.97361600e-01 4.13766652e-01 6.42053902e-01 3.29502940e-01
-5.48822880e-01 -2.14279518e-01 7.79249191e-01 5.13481855e-01
4.37886804e-01 2.96126276e-01 -6.66585326e-01 -1.95016697e-01
-8.85259330e-01 1.26666561e-01 3.64956647e-01 7.92403698e-01
4.30380225e-01 4.49414998e-01 -3.42621416e-01 6.17919207e-01
4.65695485e-02 4.63425696e-01 1.08455062e+00 -2.68117875e-01
2.03644216e-01 6.49815381e-01 -2.32649699e-01 -1.14972866e+00
-4.71113056e-01 -7.90570736e-01 -1.00774097e+00 -2.60893870e-02
-1.42637670e-01 -3.70467216e-01 -6.77142680e-01 2.17067885e+00
4.07256514e-01 6.55952632e-01 2.93113619e-01 9.08891439e-01
7.59951651e-01 6.07074201e-01 1.36370242e-01 -6.32623792e-01
1.46549082e+00 -4.68401581e-01 -8.65714490e-01 2.10363567e-01
3.39580536e-01 -3.30726266e-01 1.03074789e+00 5.04645228e-01
-9.14866626e-01 -6.14572823e-01 -1.30621469e+00 4.36904550e-01
-5.04456460e-01 -3.01041231e-02 4.94810998e-01 3.47717673e-01
-4.95270222e-01 5.26479542e-01 -8.10167432e-01 -1.94521099e-02
7.15802550e-01 6.04916215e-01 -3.72547477e-01 1.99091345e-01
-1.73949969e+00 6.87456548e-01 4.57647353e-01 2.42302582e-01
-1.05942428e+00 -8.90529573e-01 -6.49423599e-01 1.92524597e-01
8.22013393e-02 -4.13848013e-01 6.19690180e-01 -1.16102600e+00
-1.61332500e+00 3.59564692e-01 1.75965831e-01 -3.04434866e-01
3.45613629e-01 -4.54183333e-02 -7.48726428e-01 2.84436911e-01
-9.58020464e-02 3.19903463e-01 1.11445820e+00 -8.81518185e-01
-3.51119310e-01 -6.36298597e-01 -3.49093616e-01 3.95179838e-01
-7.60847092e-01 -7.29180276e-02 2.92664409e-01 -8.00580680e-01
-1.87854189e-02 -6.06550872e-01 2.95480430e-01 -2.62159109e-01
-1.25953808e-01 -1.58480912e-01 1.05497336e+00 -7.16569662e-01
1.09930956e+00 -2.53084207e+00 1.13344111e-01 3.02653193e-01
1.16607092e-01 3.09503395e-02 -4.59162891e-02 1.25114202e-01
-5.54181278e-01 -2.69670039e-01 -1.94970861e-01 2.15381667e-01
6.65669069e-02 -1.22949384e-01 -5.48622429e-01 5.45431912e-01
1.86234921e-01 8.20096016e-01 -7.29946375e-01 -2.82275736e-01
4.52992693e-03 5.04737854e-01 -2.89868206e-01 3.72710317e-01
1.29121676e-01 6.56236589e-01 -6.41061783e-01 4.52488542e-01
9.02356625e-01 2.50991166e-01 -2.18719989e-01 -5.26985824e-01
-2.67062057e-03 3.91202420e-02 -1.09781110e+00 1.88639736e+00
-3.95265341e-01 1.37910068e-01 3.15452442e-02 -1.16693938e+00
1.14026189e+00 6.24801457e-01 6.33924305e-01 -1.01412058e+00
1.96363479e-01 1.01502307e-01 5.85328452e-02 -7.00732470e-01
-2.46547475e-01 -2.52150029e-01 -2.22187728e-01 2.71287471e-01
2.83416927e-01 1.14419281e-01 -6.73391163e-01 -1.48109734e-01
1.00755274e+00 2.64732003e-01 6.78899512e-02 -4.72459167e-01
6.11131310e-01 -5.00758171e-01 8.69488358e-01 8.64430442e-02
-3.31090629e-01 -5.54787405e-02 4.05222565e-01 -2.64803499e-01
-4.07390416e-01 -1.21102321e+00 -3.98813099e-01 9.19729114e-01
2.92227358e-01 -8.29725992e-03 -5.99927306e-01 -7.88009465e-01
-1.56224847e-01 3.24606568e-01 -6.93879545e-01 -1.03265548e+00
-1.14705101e-01 -7.91528344e-01 6.58413053e-01 7.27735758e-01
8.20629597e-01 -1.07596362e+00 -5.44196904e-01 2.42445752e-01
9.48734656e-02 -7.43054569e-01 -4.93556410e-01 5.12327969e-01
-7.59495258e-01 -8.43216836e-01 -7.32134640e-01 -8.51468265e-01
3.97021919e-01 -2.55110830e-01 4.11060929e-01 -4.71260995e-01
-1.60864979e-01 3.28974158e-01 -3.22031051e-01 -4.10747260e-01
3.67278695e-01 -2.23012641e-01 7.17148423e-01 4.27451432e-01
4.24253821e-01 -1.44738305e+00 -9.01504338e-01 2.47523531e-01
-9.53243971e-01 -2.42741182e-01 4.67936218e-01 1.27396369e+00
4.02261108e-01 3.16265315e-01 1.30370307e+00 -3.15580040e-01
9.01933432e-01 -6.81073189e-01 -3.25941384e-01 1.04595974e-01
-5.01724362e-01 -1.37895197e-01 9.71156299e-01 -9.92705166e-01
-1.19648278e+00 -2.84774676e-02 2.45595556e-02 -8.19875538e-01
-6.11752383e-02 5.17122030e-01 -6.72151148e-01 -6.44580424e-02
5.35744011e-01 7.57516444e-01 -2.41542444e-01 -2.47301280e-01
9.03525203e-02 5.92504740e-01 5.33366621e-01 -7.60801494e-01
7.38806307e-01 1.20488703e-01 -2.14768887e-01 -4.67564136e-01
-4.36535388e-01 1.12506514e-02 -3.98181677e-01 7.09842443e-02
8.58126163e-01 -1.12252831e+00 -9.14186656e-01 6.37222528e-01
-1.06090140e+00 -1.10740595e-01 -1.75332487e-01 7.50973403e-01
-6.77081347e-01 -1.01898029e-01 -4.86191154e-01 -7.23680437e-01
-5.42237461e-01 -1.04661369e+00 6.22406721e-01 5.12970448e-01
1.67118907e-01 -8.42073917e-01 -3.39635238e-02 -1.40074700e-01
2.85328001e-01 4.46131557e-01 9.92203474e-01 -9.41997290e-01
-5.62485568e-02 -1.16124600e-01 -1.54355811e-02 4.70912516e-01
1.59009278e-01 -5.18757343e-01 -1.34797657e+00 -3.62635046e-01
6.07321799e-01 -4.65170979e-01 5.40263832e-01 1.42459944e-01
1.48298943e+00 -4.77854520e-01 -2.83846766e-01 8.81901920e-01
1.19677925e+00 6.52649343e-01 4.72202986e-01 1.06392838e-01
5.88962078e-01 3.66644055e-01 4.94206518e-01 5.97176075e-01
2.85966218e-01 5.19199431e-01 1.47714481e-01 -3.43742892e-02
4.41809952e-01 -3.40073556e-01 5.77626824e-01 1.14815056e+00
1.36220932e-01 3.56658012e-01 -5.92146516e-01 3.94347459e-01
-1.60363913e+00 -8.68508756e-01 5.75720131e-01 1.92740858e+00
9.46428776e-01 -1.98548883e-02 5.71945012e-02 1.89921275e-01
4.88081872e-01 -2.40819715e-02 -9.49423850e-01 -1.65311694e-01
-2.74735074e-02 2.95872211e-01 -1.63422957e-01 -2.33688220e-01
-8.99154186e-01 5.24154425e-01 4.60556459e+00 1.22831464e+00
-1.42991042e+00 2.48611823e-01 3.49941969e-01 -3.97458285e-01
-2.21564189e-01 -3.57644707e-01 -3.47506762e-01 9.18181658e-01
8.28027844e-01 -3.99077863e-01 5.84481835e-01 7.48113453e-01
3.66601944e-01 3.05007279e-01 -1.17344689e+00 1.26096451e+00
-4.64087687e-02 -7.03594506e-01 -2.40786910e-01 -9.96934623e-02
4.69874650e-01 -3.09466571e-01 4.39598173e-01 7.19680011e-01
-1.38436303e-01 -9.80008066e-01 3.70117694e-01 8.03196073e-01
8.73421907e-01 -1.08370900e+00 5.75304806e-01 3.25692534e-01
-1.24784017e+00 -3.62439990e-01 -3.22714031e-01 1.23979701e-02
-2.42673367e-01 3.97312224e-01 -8.11125115e-02 7.44451821e-01
9.78726149e-01 8.66137683e-01 -3.66449296e-01 7.69024193e-01
1.21252583e-02 5.25781572e-01 -2.37828538e-01 1.82643145e-01
1.48979753e-01 -1.96878821e-01 5.71223736e-01 1.03578746e+00
3.28611076e-01 3.10692787e-01 2.51047701e-01 9.31466162e-01
-7.08528459e-02 1.20142996e-01 -7.35704958e-01 1.30417675e-01
6.76928878e-01 1.30958045e+00 -1.91351518e-01 -1.75064772e-01
-1.83438107e-01 1.26167810e+00 2.36254200e-01 6.37118995e-01
-1.03502631e+00 -8.22374821e-01 6.70650065e-01 -4.66906756e-01
2.05474883e-01 1.37279347e-01 -6.74784482e-02 -1.26373589e+00
2.91691363e-01 -8.90137434e-01 4.09338802e-01 -7.96965420e-01
-1.62930191e+00 7.37229884e-01 3.20337117e-02 -1.51776826e+00
7.85104260e-02 -3.13121855e-01 -1.12231696e+00 1.12588274e+00
-1.49095213e+00 -1.12931824e+00 -3.63055140e-01 1.30158317e+00
8.47690850e-02 -2.71164447e-01 8.85202587e-01 3.56689453e-01
-7.70762086e-01 1.03938115e+00 2.74033487e-01 1.61614064e-02
7.76766837e-01 -8.71956527e-01 -6.23451114e-01 5.09876430e-01
-3.25327784e-01 7.54648805e-01 3.47884774e-01 -4.04934704e-01
-1.28807962e+00 -1.18648887e+00 3.29904370e-02 -5.48444726e-02
6.84286714e-01 -6.21971667e-01 -1.10528171e+00 4.33243990e-01
2.34932870e-01 2.32272163e-01 7.92274773e-01 -1.42200321e-01
-2.99544871e-01 -6.71905935e-01 -1.32716930e+00 4.61072475e-01
7.68094599e-01 -7.80010462e-01 -9.58878696e-01 1.20578706e-01
9.16433156e-01 -3.01918179e-01 -1.29865396e+00 4.52008903e-01
7.06445813e-01 -5.57055175e-01 8.09061646e-01 -7.54170895e-01
2.81505942e-01 -1.57478541e-01 -6.02483861e-02 -1.73259234e+00
-1.91852212e-01 -4.89315718e-01 -9.69177485e-02 1.52854609e+00
-5.37764765e-02 -9.23108637e-01 3.99764597e-01 4.93388861e-01
-2.79757142e-01 -1.07894635e+00 -1.18034792e+00 -9.52392936e-01
2.51758337e-01 -2.49211133e-01 8.85556698e-01 1.11193585e+00
5.36491275e-01 2.97150046e-01 -2.33924434e-01 1.27088428e-01
2.98948675e-01 1.20968580e-01 2.09391713e-01 -1.02816272e+00
-8.63927156e-02 -2.89322644e-01 -5.42056799e-01 -6.69520080e-01
4.34513777e-01 -1.03986561e+00 -1.78482652e-01 -7.86454141e-01
3.55371311e-02 -3.26086879e-01 -1.05127859e+00 4.43540186e-01
-1.56727016e-01 -8.54270011e-02 -1.10332854e-02 -5.03967889e-02
-5.01337051e-01 1.36017513e+00 1.25047767e+00 -2.53826708e-01
-5.93101859e-01 -8.40058550e-02 -5.91971636e-01 5.74928463e-01
8.18532526e-01 -5.42402267e-01 -7.50881493e-01 4.31996137e-02
-8.37915838e-02 2.52505809e-01 4.91907686e-01 -1.01797831e+00
3.99282068e-01 4.15613055e-02 9.48551595e-01 -2.88820118e-01
5.29358625e-01 -1.10661590e+00 -2.04597846e-01 3.56282443e-01
-2.67551363e-01 3.54254395e-02 5.66181958e-01 5.97625911e-01
-3.46687168e-01 3.37293148e-01 6.30189300e-01 2.97709554e-02
-4.11538363e-01 7.21802115e-01 -1.93583176e-01 1.42258137e-01
1.14030790e+00 -2.01099068e-01 -2.23053262e-01 -5.77923767e-02
-9.89520550e-01 4.63285148e-01 -7.52383620e-02 2.68529922e-01
8.46857488e-01 -1.83299458e+00 -4.00516182e-01 5.91735840e-01
2.44875431e-01 -2.38314614e-01 8.69409978e-01 1.21977293e+00
4.74009633e-01 9.31837186e-02 -6.06352806e-01 -4.65674311e-01
-8.50781560e-01 8.50172818e-01 5.65709233e-01 -1.04458846e-01
-5.63496947e-01 6.77321136e-01 4.81803715e-01 -3.53362769e-01
1.38306633e-01 -6.78059831e-02 -2.64250517e-01 3.11316580e-01
4.96435732e-01 1.67474627e-01 6.07802831e-02 -2.21027330e-01
-6.61101937e-01 3.72979581e-01 -6.55850917e-02 -1.13004915e-01
1.53459609e+00 -7.77168423e-02 -1.21058361e-03 4.97814864e-01
1.40015149e+00 -2.48338357e-01 -1.42509556e+00 -3.58484119e-01
-4.23795402e-01 -3.64562683e-02 1.08433746e-01 -1.00978458e+00
-1.04885864e+00 1.20801759e+00 1.12746084e+00 -1.94034114e-01
1.59861183e+00 -3.64782780e-01 8.09291959e-01 1.65615454e-01
2.55366683e-01 -1.02542186e+00 2.49496073e-01 1.83636814e-01
9.63924825e-01 -9.31868374e-01 -3.42433304e-01 2.95423150e-01
-6.95932150e-01 9.36121941e-01 9.31751132e-01 -3.16531271e-01
1.01801074e+00 2.60057390e-01 -8.18214044e-02 -2.86545813e-01
-8.07479501e-01 2.42399246e-01 3.94216388e-01 7.48092592e-01
-6.90087676e-02 9.15209670e-03 -1.97021142e-01 1.70713902e+00
-1.66675612e-01 1.00346729e-01 -5.84036931e-02 5.91399431e-01
-5.82562760e-02 -6.12672925e-01 -2.57653058e-01 5.05970597e-01
-2.00073212e-01 -1.59095347e-01 9.71855968e-02 5.33593357e-01
4.11566943e-01 6.68919504e-01 -9.61112157e-02 -9.81163502e-01
4.40999955e-01 5.22361100e-01 4.40203339e-01 -1.48841232e-01
-7.12516308e-01 2.32007541e-02 -6.24198437e-01 -7.45230675e-01
-1.25807464e-01 -5.48964143e-01 -1.41482615e+00 8.60217735e-02
-1.86454833e-01 3.33877921e-01 3.44097555e-01 1.10507548e+00
4.91841793e-01 8.54129195e-01 1.13650870e+00 -6.36775076e-01
-6.85586333e-01 -1.10266817e+00 -9.69404101e-01 4.35690165e-01
5.33540010e-01 -9.25947011e-01 -4.60769653e-01 -5.36994413e-02] | [13.176750183105469, 3.531377077102661] |
09bbe923-c185-45a0-8054-eebca5b73f34 | get-to-the-point-summarization-with-pointer | 1704.04368 | null | http://arxiv.org/abs/1704.04368v2 | http://arxiv.org/pdf/1704.04368v2.pdf | Get To The Point: Summarization with Pointer-Generator Networks | Neural sequence-to-sequence models have provided a viable new approach for
abstractive text summarization (meaning they are not restricted to simply
selecting and rearranging passages from the original text). However, these
models have two shortcomings: they are liable to reproduce factual details
inaccurately, and they tend to repeat themselves. In this work we propose a
novel architecture that augments the standard sequence-to-sequence attentional
model in two orthogonal ways. First, we use a hybrid pointer-generator network
that can copy words from the source text via pointing, which aids accurate
reproduction of information, while retaining the ability to produce novel words
through the generator. Second, we use coverage to keep track of what has been
summarized, which discourages repetition. We apply our model to the CNN / Daily
Mail summarization task, outperforming the current abstractive state-of-the-art
by at least 2 ROUGE points. | ['Abigail See', 'Peter J. Liu', 'Christopher D. Manning'] | 2017-04-14 | get-to-the-point-summarization-with-pointer-1 | https://aclanthology.org/P17-1099 | https://aclanthology.org/P17-1099.pdf | acl-2017-7 | ['extractive-document-summarization'] | ['natural-language-processing'] | [ 5.24017274e-01 2.51783609e-01 -1.05132312e-01 2.85552628e-02
-9.11489189e-01 -5.72877049e-01 7.62539268e-01 3.17991078e-01
-4.25496161e-01 9.87034678e-01 1.00968444e+00 -2.28658199e-01
3.00908238e-01 -6.42116189e-01 -7.33912647e-01 -2.47795820e-01
3.64288509e-01 5.60255945e-01 9.26634893e-02 -6.72816932e-01
8.36867034e-01 2.21104592e-01 -1.40239072e+00 5.11433661e-01
1.28822863e+00 1.75683469e-01 2.85103917e-01 8.51717710e-01
-4.49138284e-01 8.36486578e-01 -1.25111270e+00 -4.22855258e-01
-2.99093693e-01 -1.09673917e+00 -8.05741370e-01 -3.08506966e-01
7.96265543e-01 -6.14371359e-01 -4.75869417e-01 8.83114994e-01
8.07503462e-01 3.01070839e-01 7.50658274e-01 -7.75964200e-01
-1.27369225e+00 1.04467440e+00 -4.69851613e-01 3.67459089e-01
6.91964507e-01 2.08338618e-01 1.10176361e+00 -6.56957746e-01
7.37661481e-01 1.20213830e+00 5.90311706e-01 1.00753689e+00
-1.10986078e+00 -4.59660023e-01 1.64199084e-01 -1.05523355e-01
-8.86227369e-01 -8.62253070e-01 6.16115093e-01 -1.21064872e-01
1.50375879e+00 6.35601580e-01 6.18654907e-01 1.45891452e+00
4.32918638e-01 1.05765462e+00 3.03185403e-01 -5.39910853e-01
1.45657554e-01 -2.63566345e-01 1.91162810e-01 5.92741132e-01
4.33432192e-01 -1.64050967e-01 -6.89825594e-01 -1.35687709e-01
4.88908589e-01 -2.34664410e-01 -5.51437259e-01 1.19977616e-01
-1.24909437e+00 8.41742873e-01 3.78875941e-01 3.55676591e-01
-4.81357276e-01 2.49710783e-01 6.58770382e-01 1.64081007e-01
5.16412437e-01 1.04974198e+00 -7.44532943e-02 -2.99364358e-01
-1.46506798e+00 5.54307580e-01 9.15475190e-01 1.13631606e+00
3.25799912e-01 2.87819922e-01 -8.33735287e-01 6.64516151e-01
-1.19760655e-01 2.84305066e-01 9.98320043e-01 -7.80339122e-01
7.62794077e-01 4.08266485e-01 1.32572129e-01 -8.67813051e-01
-2.92974830e-01 -6.52318239e-01 -1.13901329e+00 -1.27341673e-01
-9.81680974e-02 -4.09251377e-02 -8.64437640e-01 1.60600841e+00
-5.27880371e-01 -1.12924248e-01 2.96757161e-03 4.77504432e-01
1.03850877e+00 8.73599827e-01 -1.17187820e-01 -2.33890206e-01
9.47504699e-01 -1.31598032e+00 -8.52144718e-01 -3.40007871e-01
5.69281399e-01 -5.60048103e-01 1.34021139e+00 9.69427675e-02
-1.58823240e+00 -5.71774006e-01 -1.10195410e+00 -3.16413701e-01
-1.85711130e-01 9.98638943e-02 2.63864219e-01 3.98738354e-01
-1.45305741e+00 8.93366516e-01 -5.87445617e-01 -4.63779479e-01
4.81819808e-01 -5.88191077e-02 -8.66858009e-03 3.12211692e-01
-1.16436470e+00 1.00708652e+00 6.99615657e-01 -1.62850633e-01
-4.62060988e-01 -7.64057219e-01 -7.72046566e-01 5.94120681e-01
1.88478261e-01 -1.37953115e+00 1.53985000e+00 -8.45460653e-01
-1.64017320e+00 4.97446358e-01 -3.79423648e-01 -7.65925109e-01
6.25625432e-01 -5.42754531e-01 -2.66259730e-01 9.21866745e-02
1.41201466e-01 8.04969549e-01 7.56496310e-01 -1.10685349e+00
-3.53090018e-01 1.84208862e-02 -1.77419052e-01 3.36938620e-01
-3.24323982e-01 -1.95928305e-01 -3.19227844e-01 -1.16281629e+00
-3.40450615e-01 -5.93427241e-01 -1.65320575e-01 -5.00320077e-01
-8.07972729e-01 -1.58017814e-01 3.54327798e-01 -8.26512098e-01
1.77634752e+00 -1.74555302e+00 3.21084499e-01 -4.35283005e-01
3.81940693e-01 5.42019248e-01 -4.84527409e-01 9.11196053e-01
7.49266967e-02 6.14312768e-01 -3.51305693e-01 -4.18639690e-01
-7.18209753e-03 -3.16156954e-01 -7.57892907e-01 -1.90370437e-02
2.51466453e-01 1.51050591e+00 -1.11049283e+00 -4.10946012e-01
-1.23496719e-01 3.71998660e-02 -6.78065538e-01 8.03775564e-02
-6.01233125e-01 1.51522383e-01 -2.59671092e-01 1.05270460e-01
3.23914915e-01 -1.82924986e-01 -1.81388438e-01 1.90547988e-01
-1.17504939e-01 7.97855198e-01 -5.23422897e-01 1.91352570e+00
-3.40149581e-01 8.11706781e-01 -4.05753314e-01 -5.42887330e-01
6.51277602e-01 2.55804658e-01 -4.80858833e-02 -7.04744339e-01
-5.79919340e-03 1.49847895e-01 -1.62100568e-01 -3.94879341e-01
1.35385811e+00 1.88082740e-01 -9.41792876e-02 8.30222845e-01
-3.05163115e-02 -1.91866234e-01 5.02096415e-01 6.04703367e-01
1.12269795e+00 1.23715214e-01 5.66576421e-01 -2.72148177e-02
3.39881629e-01 3.73279750e-02 1.46848559e-01 1.29930472e+00
1.59245610e-01 8.15666616e-01 5.16819060e-01 -3.74762379e-02
-1.26608980e+00 -8.64264309e-01 4.91542697e-01 1.06158364e+00
-1.54853642e-01 -5.55117846e-01 -8.56595576e-01 -7.74561822e-01
-2.55653620e-01 1.44095230e+00 -4.53352362e-01 -5.00279188e-01
-9.98476267e-01 -3.51483017e-01 9.27900910e-01 6.68465614e-01
3.55020285e-01 -1.38569105e+00 -6.22872531e-01 4.55021679e-01
-5.85073590e-01 -4.88527387e-01 -9.57286954e-01 -1.66648895e-01
-1.06441927e+00 -4.93689656e-01 -1.04462790e+00 -6.37256920e-01
4.66847658e-01 3.53098810e-01 1.52386045e+00 3.09149057e-01
9.09701139e-02 1.21826887e-01 -3.73699725e-01 -4.62997615e-01
-8.70912790e-01 7.84962058e-01 -3.27704728e-01 -5.46285748e-01
5.96838370e-02 -4.64051247e-01 -5.97120225e-01 -2.44128302e-01
-1.15060997e+00 2.50987113e-01 8.89169991e-01 9.08391297e-01
9.48802456e-02 -4.73961860e-01 1.18082321e+00 -1.07280874e+00
1.44381499e+00 -3.76022905e-01 1.11091405e-01 3.74198496e-01
-5.27547657e-01 2.58426160e-01 9.48684752e-01 -2.83815473e-01
-9.55402553e-01 -4.55708325e-01 -3.31263989e-01 -9.95781943e-02
-6.46375194e-02 5.69122553e-01 1.64256379e-01 4.82319981e-01
8.11100304e-01 6.94425762e-01 1.36260521e-02 -5.79519391e-01
8.08160663e-01 5.97602189e-01 7.44180441e-01 -2.50029296e-01
6.59191608e-01 9.94288772e-02 -4.40703273e-01 -8.09194446e-01
-6.91528380e-01 -2.72413343e-01 -6.13533735e-01 4.10230532e-02
5.16960144e-01 -6.08639061e-01 -4.51084793e-01 3.38684648e-01
-1.52078974e+00 -1.48274928e-01 -6.25008464e-01 9.12944879e-03
-6.08908772e-01 9.42439198e-01 -7.85072267e-01 -4.67351943e-01
-1.00672317e+00 -6.28192544e-01 1.05750024e+00 2.98406512e-01
-8.18293571e-01 -7.69009650e-01 1.36977509e-01 -6.79630339e-02
7.76653409e-01 4.53528017e-02 1.16431248e+00 -9.61909652e-01
-3.47730935e-01 -2.30154023e-01 -3.16552818e-02 2.45629102e-01
3.03894710e-02 -4.49281745e-03 -5.23028255e-01 -3.50904226e-01
-8.25149268e-02 -1.84695542e-01 1.40345359e+00 3.22408438e-01
1.17786014e+00 -9.34112132e-01 -3.29584748e-01 3.55459273e-01
9.27759409e-01 4.32506297e-03 8.93752217e-01 3.71061027e-01
5.03269970e-01 4.03570592e-01 8.71625841e-02 2.85778642e-01
2.29851469e-01 3.54840994e-01 2.74087161e-01 -4.63247299e-02
-5.36486864e-01 -7.40701795e-01 3.10450017e-01 1.17589581e+00
1.12197630e-01 -1.02404451e+00 -4.08728391e-01 5.54125249e-01
-1.87670779e+00 -1.34368360e+00 -1.37033194e-01 2.03845215e+00
9.74808991e-01 1.99784726e-01 1.37006408e-02 -2.73459703e-01
6.74074650e-01 5.16080618e-01 -5.30862868e-01 -6.03225946e-01
-2.38467529e-01 2.40521744e-01 3.35624009e-01 3.69966656e-01
-6.75830781e-01 1.12473309e+00 6.90529728e+00 8.37378323e-01
-1.00008428e+00 -2.19445258e-01 3.14076483e-01 -3.37332994e-01
-5.76091886e-01 -2.27950633e-01 -7.41437674e-01 6.27441227e-01
9.83715236e-01 -4.33023542e-01 3.74052435e-01 5.09796202e-01
1.95802674e-01 -3.79038379e-02 -1.19483936e+00 5.76837122e-01
6.75011516e-01 -1.77714229e+00 9.27676141e-01 -3.83995235e-01
8.44034851e-01 -2.43199378e-01 5.88795207e-02 4.53614891e-01
3.18318933e-01 -1.20199537e+00 8.43123913e-01 8.24347854e-01
7.66683578e-01 -6.10828340e-01 6.07925653e-01 5.93526304e-01
-5.32048106e-01 2.16187388e-01 -3.36345017e-01 1.24259582e-02
4.45730716e-01 3.28645140e-01 -8.92186821e-01 6.84757948e-01
1.43688381e-01 6.05943263e-01 -8.13000739e-01 1.01996076e+00
-2.23487347e-01 5.21888256e-01 1.24063402e-01 -4.48396742e-01
2.56712139e-01 9.35753956e-02 9.52328980e-01 1.54262352e+00
5.27564287e-01 -2.25280151e-01 -6.54309466e-02 9.81129467e-01
-5.65735996e-01 1.51126191e-01 -6.28963232e-01 -1.88931420e-01
5.09787261e-01 7.34357953e-01 -3.91513467e-01 -6.91095471e-01
-6.58791736e-02 1.43357730e+00 4.10480797e-01 4.61905152e-01
-5.48029244e-01 -8.27252150e-01 1.29772529e-01 -9.30304639e-03
3.19244713e-01 -1.87498480e-01 -4.55617189e-01 -1.30123770e+00
9.16474015e-02 -1.01245809e+00 1.07274227e-01 -9.85601008e-01
-1.13518155e+00 7.30105758e-01 -1.68449357e-01 -1.04784155e+00
-5.86096525e-01 6.93439157e-04 -8.77697170e-01 9.97008562e-01
-1.45147145e+00 -9.02486265e-01 -9.55449790e-02 1.51049271e-02
1.12224638e+00 -2.21386120e-01 8.24480593e-01 -5.17181978e-02
-4.09090459e-01 6.82465732e-01 3.65602314e-01 7.14648217e-02
8.48171949e-01 -1.21902335e+00 1.18284833e+00 9.30442572e-01
8.62617269e-02 1.05101633e+00 8.66059065e-01 -9.35052931e-01
-1.08181286e+00 -1.09195602e+00 1.28713810e+00 -5.55807769e-01
4.01444107e-01 -2.74627835e-01 -1.13530004e+00 8.57191205e-01
7.42556810e-01 -9.67960656e-01 4.93872404e-01 -9.88676250e-02
-2.11887851e-01 3.43602061e-01 -6.88073218e-01 1.20885134e+00
1.11884212e+00 -2.50835776e-01 -1.22846365e+00 2.28351444e-01
1.09138691e+00 -3.86881143e-01 -1.63869351e-01 3.73207219e-02
3.70110452e-01 -8.99747431e-01 7.83152938e-01 -8.24524939e-01
9.45437670e-01 -6.77716509e-02 4.16698098e-01 -1.79584205e+00
-5.98774791e-01 -9.79815483e-01 -5.10258198e-01 1.38928521e+00
4.92437780e-01 -6.39577448e-01 5.77521861e-01 5.01107648e-02
-6.23384237e-01 -5.49532413e-01 -6.89711809e-01 -7.86082447e-01
3.30594510e-01 1.05426088e-01 8.42071056e-01 6.39776886e-01
1.55429363e-01 9.48656619e-01 -3.99087220e-01 -6.31584108e-01
2.03951418e-01 -8.03403929e-02 6.12256408e-01 -9.84850466e-01
-1.43451363e-01 -1.12505102e+00 2.08189949e-01 -1.66029477e+00
2.35876501e-01 -1.11861479e+00 2.87455916e-01 -1.93795979e+00
2.83568531e-01 1.30922467e-01 4.35617492e-02 1.42598122e-01
-5.50247431e-01 -8.43201950e-02 2.12029472e-01 4.16875690e-01
-6.41077638e-01 9.19523180e-01 1.20870507e+00 -2.59715587e-01
-3.44847888e-01 1.87987257e-02 -1.21096146e+00 3.63146722e-01
9.25480366e-01 -2.87462145e-01 -2.12472200e-01 -7.34436512e-01
3.44876200e-01 1.47396311e-01 2.32352212e-01 -8.86074901e-01
4.22772795e-01 9.16162953e-02 3.81941140e-01 -9.72645521e-01
-1.73301809e-02 -3.36283855e-02 -3.27295661e-01 5.09233475e-01
-1.09632230e+00 4.67450291e-01 2.33932883e-01 7.67466903e-01
-4.14730655e-03 -6.46401227e-01 4.66867417e-01 -4.51769799e-01
-3.37523580e-01 -1.32531568e-01 -8.01730096e-01 3.04756939e-01
4.95172411e-01 -1.79581225e-01 -7.99398601e-01 -7.51912296e-01
-9.17442665e-02 1.70355752e-01 6.61788404e-01 5.07845938e-01
5.78424871e-01 -9.85093653e-01 -1.12045705e+00 -9.49294865e-03
1.83004327e-02 -1.37984961e-01 2.25135967e-01 3.80271912e-01
-6.73806369e-01 9.45536315e-01 -1.90990835e-01 -1.76553413e-01
-8.92797470e-01 7.19310641e-01 5.46450652e-02 -3.85467649e-01
-9.80991483e-01 6.57956302e-01 -1.03649059e-02 -2.75864303e-01
2.58012474e-01 -4.15578753e-01 -3.65103662e-01 1.03970818e-01
7.69619703e-01 5.67186475e-01 3.28054935e-01 -2.26285070e-01
1.26033008e-01 2.22655177e-01 -5.81392288e-01 -2.38437906e-01
1.07321620e+00 -2.41971299e-01 -1.84130549e-01 3.24351519e-01
9.91232693e-01 2.87555993e-01 -7.62763083e-01 -1.41112417e-01
1.58170432e-01 -1.63455039e-01 -2.95166701e-01 -1.06333780e+00
-3.92523617e-01 8.31660867e-01 -2.71819443e-01 3.47338736e-01
8.91188562e-01 -2.47414902e-01 1.24016166e+00 5.96397460e-01
-2.18903497e-01 -1.02411425e+00 2.59802043e-01 9.56405699e-01
1.22439647e+00 -7.44808495e-01 6.79018572e-02 4.04890301e-03
-7.64306605e-01 1.24940193e+00 6.02236748e-01 -9.91799533e-02
-2.80423522e-01 -9.39440802e-02 -2.84442514e-01 -5.55013008e-02
-9.94196236e-01 5.32605946e-02 2.59043753e-01 5.67618430e-01
7.49504328e-01 -2.80805826e-01 -5.65355837e-01 4.03123915e-01
-6.22294903e-01 -7.24391546e-03 9.21611965e-01 8.13016653e-01
-7.11943924e-01 -7.18215346e-01 -1.06508970e-01 7.51161575e-01
-4.52086836e-01 -6.45042598e-01 -7.25255549e-01 5.65660477e-01
-6.63595438e-01 9.04349864e-01 1.63428962e-01 -8.94279480e-02
4.48998809e-01 3.05290550e-01 3.18590611e-01 -8.94181430e-01
-9.38974440e-01 -1.09322600e-01 1.09612271e-01 -1.77998364e-01
1.28620565e-01 -4.63429898e-01 -9.72136080e-01 -4.45290178e-01
-2.69192815e-01 2.88798660e-01 4.14518863e-01 6.26741707e-01
7.50745893e-01 9.97755587e-01 2.88601696e-01 -9.75228250e-01
-1.03191423e+00 -1.30157590e+00 -1.37253478e-01 1.98423967e-01
5.54008842e-01 5.39803728e-02 -2.32368648e-01 8.61862488e-03] | [12.273171424865723, 9.292778015136719] |
e87dc4f9-60b5-4f40-bb25-d2d711673d09 | pushing-paraphrase-away-from-original | 2109.01862 | null | https://arxiv.org/abs/2109.01862v1 | https://arxiv.org/pdf/2109.01862v1.pdf | Pushing Paraphrase Away from Original Sentence: A Multi-Round Paraphrase Generation Approach | In recent years, neural paraphrase generation based on Seq2Seq has achieved superior performance, however, the generated paraphrase still has the problem of lack of diversity. In this paper, we focus on improving the diversity between the generated paraphrase and the original sentence, i.e., making generated paraphrase different from the original sentence as much as possible. We propose BTmPG (Back-Translation guided multi-round Paraphrase Generation), which leverages multi-round paraphrase generation to improve diversity and employs back-translation to preserve semantic information. We evaluate BTmPG on two benchmark datasets. Both automatic and human evaluation show BTmPG can improve the diversity of paraphrase while preserving the semantics of the original sentence. | ['Xiaojun Wan', 'Zhe Lin'] | 2021-09-04 | null | https://aclanthology.org/2021.findings-acl.135 | https://aclanthology.org/2021.findings-acl.135.pdf | findings-acl-2021-8 | ['paraphrase-generation', 'paraphrase-generation'] | ['computer-code', 'natural-language-processing'] | [ 3.30006093e-01 -4.09919471e-02 -2.59270728e-01 -4.10662442e-01
-9.32395339e-01 -5.95785379e-01 3.46044004e-01 7.78450444e-02
-1.97718114e-01 1.05012512e+00 9.45742071e-01 -1.31377533e-01
2.14589626e-01 -7.76251853e-01 -9.06629324e-01 -2.35753477e-01
8.25470924e-01 7.02843666e-02 6.76027918e-03 -4.77513939e-01
5.99163115e-01 -4.42835055e-02 -1.18600881e+00 9.25950050e-01
1.35312068e+00 4.00002420e-01 4.16182697e-01 3.17156613e-01
-2.82398462e-01 7.48327553e-01 -8.87150705e-01 -5.58138072e-01
3.25984806e-01 -1.15752351e+00 -1.03183019e+00 -3.12057406e-01
4.66956198e-01 -2.29107007e-01 -3.19534421e-01 1.19545829e+00
5.28499782e-01 1.17194533e-01 2.89769053e-01 -9.29647207e-01
-1.19953036e+00 8.32030475e-01 -4.59228665e-01 3.93708974e-01
6.69618189e-01 9.30222869e-02 1.18967402e+00 -9.58589792e-01
7.53809392e-01 1.28624392e+00 7.23211169e-01 7.22201109e-01
-1.34781051e+00 -6.75866961e-01 -1.64839104e-01 4.66200799e-01
-1.17555714e+00 -3.22127223e-01 7.29527116e-01 9.98765975e-03
1.12032497e+00 4.11089718e-01 6.16968572e-01 1.39514899e+00
6.34266376e-01 7.88191259e-01 8.99751127e-01 -4.09833968e-01
3.62038344e-01 -2.62286849e-02 1.13455310e-01 3.02716374e-01
2.07425341e-01 -1.44627795e-01 -7.30028749e-01 -1.24376297e-01
4.62305903e-01 3.55478898e-02 -3.31313878e-01 -2.00131759e-02
-1.17025948e+00 9.87185538e-01 6.24924660e-01 1.73893049e-01
-3.70974600e-01 -1.18562259e-01 6.81266308e-01 7.33673334e-01
4.56880510e-01 1.06176567e+00 -1.74626097e-01 -1.79785848e-01
-1.25435591e+00 5.25453150e-01 7.35319674e-01 9.95794237e-01
6.38630688e-01 -2.43422046e-01 -7.97027707e-01 1.26219881e+00
-2.45421633e-01 7.50925481e-01 1.13941956e+00 -1.02875364e+00
9.40439939e-01 6.58139586e-01 -3.57986949e-02 -8.93680274e-01
1.89559132e-01 -5.33042848e-01 -9.32359159e-01 -3.30316037e-01
-1.86551049e-01 -1.50077894e-01 -7.96095133e-01 1.78936589e+00
-1.92686975e-01 -9.98114571e-02 2.64252901e-01 8.82105589e-01
7.51464367e-01 9.05659676e-01 -2.65753269e-01 -2.47503426e-02
7.69726694e-01 -1.40089035e+00 -5.31937540e-01 -6.68470085e-01
5.93491733e-01 -6.78820014e-01 1.43606091e+00 -6.32558465e-02
-1.24081171e+00 -5.59645653e-01 -1.04934943e+00 -2.24191621e-02
1.35174006e-01 -1.80261165e-01 -8.38981420e-02 1.74247861e-01
-9.05341327e-01 8.12975049e-01 -4.28951353e-01 -4.08751905e-01
3.25692266e-01 -8.05722848e-02 -3.22968423e-01 -3.91146392e-01
-1.49503303e+00 1.02362669e+00 6.14424765e-01 -2.76175618e-01
-4.35951769e-01 -9.43888664e-01 -9.02623892e-01 2.88758308e-01
1.23867571e-01 -1.32700777e+00 1.39390290e+00 -1.26059055e+00
-1.61873794e+00 5.93312860e-01 -5.25706828e-01 -7.61372030e-01
3.61115336e-01 -2.53172845e-01 -1.87206686e-01 -4.38768715e-02
4.77628618e-01 6.13558233e-01 6.07290447e-01 -7.54105687e-01
-4.25449967e-01 -1.11646309e-01 -9.98216122e-02 4.85787272e-01
-2.96963573e-01 -1.16825603e-01 -1.42314896e-01 -1.05613387e+00
-3.18474323e-02 -9.31000352e-01 -2.19295010e-01 -4.49797899e-01
-5.85410953e-01 1.55723765e-01 4.48949367e-01 -8.32648456e-01
1.36638784e+00 -1.92997658e+00 5.62114716e-01 -2.30789766e-01
-1.45707011e-01 5.92715263e-01 -6.85186446e-01 9.15429354e-01
-9.52714309e-02 9.06159207e-02 -4.74407047e-01 -3.51168513e-01
-2.06210777e-01 2.37128988e-01 -6.87728465e-01 -2.34651074e-01
3.49260382e-02 1.44642818e+00 -1.23750830e+00 -2.78341681e-01
-2.00187415e-01 -1.15417138e-01 -7.75090754e-01 2.35079870e-01
-2.24368662e-01 2.05043882e-01 -2.76342422e-01 2.35402226e-01
6.43934548e-01 -1.95021436e-01 -1.27763912e-01 3.91022973e-02
2.08359748e-01 4.80415583e-01 -2.76798844e-01 2.10995221e+00
-8.93574178e-01 5.58137178e-01 -5.60331225e-01 -5.85813105e-01
9.89686728e-01 -9.00999531e-02 -2.42616087e-02 -1.13813913e+00
-2.80572563e-01 3.82213056e-01 -1.04878649e-01 -3.56015563e-01
7.28063524e-01 -5.37703514e-01 -2.03417808e-01 6.77749455e-01
-3.10080588e-01 -3.58418018e-01 2.94197738e-01 4.31317151e-01
1.09665060e+00 -9.34600607e-02 4.57419962e-01 -1.02572978e-01
4.58768636e-01 1.80557594e-01 6.24239326e-01 8.85245681e-01
4.92236996e-03 1.01894534e+00 3.66476387e-01 3.15971002e-02
-1.26598740e+00 -1.25047278e+00 4.76029903e-01 5.66059768e-01
2.38641396e-01 -4.71442103e-01 -8.49507034e-01 -7.75941312e-01
1.20894857e-01 1.26873517e+00 -5.09846985e-01 -8.63248706e-01
-7.99371600e-01 -4.84853148e-01 8.01442802e-01 5.93404889e-01
7.48284757e-01 -1.20458102e+00 -2.61019677e-01 2.41100416e-01
-7.21953452e-01 -8.89281631e-01 -1.01811624e+00 -3.92473370e-01
-9.44414556e-01 -5.74439883e-01 -1.01762474e+00 -7.89726138e-01
5.54878712e-01 6.46883130e-01 1.24003160e+00 -1.75124511e-01
1.42094657e-01 -5.11085689e-01 -6.90061033e-01 -1.02795884e-01
-8.43023419e-01 4.39402163e-01 -2.76913911e-01 -3.56304646e-01
2.75251627e-01 -7.08326280e-01 -5.93715608e-01 7.46132433e-02
-9.45962667e-01 4.21095282e-01 7.71919012e-01 1.09614646e+00
5.00942528e-01 -3.55107516e-01 1.15504062e+00 -7.43704498e-01
1.24123907e+00 -7.55528986e-01 9.65184122e-02 4.47803289e-01
-7.33246326e-01 2.33934253e-01 1.35206187e+00 -1.71865344e-01
-1.03399217e+00 -4.94887024e-01 -2.92701304e-01 -4.89020705e-01
2.24707767e-01 5.92305303e-01 1.46727562e-01 3.16429734e-01
7.08106756e-01 8.07925761e-01 1.97318867e-01 -4.83562708e-01
4.05125678e-01 9.25710022e-01 4.96366560e-01 -3.21512282e-01
4.84029859e-01 1.51320100e-01 -2.06068963e-01 -4.32246268e-01
-1.28250194e+00 -2.51699626e-01 -2.08336070e-01 3.17589015e-01
3.41887802e-01 -8.33284259e-01 1.72063246e-01 3.34713340e-01
-1.26475489e+00 -1.49439469e-01 -6.47450268e-01 8.76836404e-02
-4.37182605e-01 7.83978641e-01 -6.03269398e-01 2.51124464e-02
-9.97416079e-01 -8.70127380e-01 9.76601541e-01 1.52535856e-01
-5.31343579e-01 -7.46159196e-01 5.11260033e-01 4.34223622e-01
5.94279945e-01 -9.06068087e-02 9.57653940e-01 -5.54765522e-01
-2.63100147e-01 -1.17921978e-01 -3.23308051e-01 6.42506480e-01
4.82519627e-01 -5.55083394e-01 -2.31256098e-01 -3.69556993e-01
9.48234648e-02 -3.77992779e-01 1.01766503e+00 9.47043821e-02
8.13750505e-01 -7.01554775e-01 1.92589387e-02 6.94612563e-01
1.18504751e+00 -6.44369349e-02 7.39906013e-01 4.81635749e-01
5.68905711e-01 3.61626416e-01 7.13458717e-01 1.89669922e-01
2.23663718e-01 7.82531679e-01 -4.47549634e-02 2.80841619e-01
-4.72071826e-01 -8.61838698e-01 5.23961782e-01 1.05176353e+00
6.88973606e-01 -3.82249773e-01 -4.96042997e-01 6.93541706e-01
-2.11055470e+00 -1.19344246e+00 7.75811151e-02 2.07788968e+00
1.04653895e+00 -3.04486547e-02 9.19553265e-02 -1.88313618e-01
7.61298358e-01 3.15964788e-01 -7.82062531e-01 -7.26982117e-01
-3.73209804e-01 1.78396761e-01 1.71918869e-01 5.35923660e-01
-3.47939998e-01 1.08310068e+00 6.06960678e+00 9.81257200e-01
-1.08986413e+00 3.39650661e-02 3.66963983e-01 -6.06830299e-01
-7.42834926e-01 8.35953280e-02 -5.82616687e-01 1.01664650e+00
7.75485873e-01 -8.40886950e-01 6.10588849e-01 5.34540713e-01
5.14347553e-01 1.56378329e-01 -9.27315533e-01 8.26690078e-01
4.36432362e-01 -1.44950426e+00 7.65445232e-01 -3.98449004e-01
1.10125172e+00 -3.79951880e-03 -1.56245753e-01 4.53145355e-01
2.18508422e-01 -9.24450994e-01 5.94962835e-01 3.82712990e-01
6.09027743e-01 -9.10035372e-01 9.50065732e-01 5.38074613e-01
-7.66859770e-01 -1.02520689e-01 -5.24545610e-01 -2.05810461e-02
4.67514664e-01 7.20320880e-01 -8.09563935e-01 7.85923779e-01
2.35459298e-01 9.28487718e-01 -8.46998334e-01 9.54036891e-01
-4.52942520e-01 5.57551205e-01 1.58432677e-01 -2.89197534e-01
3.48133713e-01 -2.41371691e-01 8.01405609e-01 1.31664407e+00
5.67868054e-01 -2.38020867e-01 -2.30612028e-02 1.10504341e+00
-2.62716383e-01 2.25206420e-01 -7.01733291e-01 -3.78018394e-02
7.58757889e-01 6.48823977e-01 9.84369516e-02 -3.91455859e-01
-1.60796359e-01 1.76422453e+00 5.43051958e-01 2.70262241e-01
-7.65666246e-01 -8.75552356e-01 7.02355325e-01 -6.83773309e-02
1.84788704e-01 2.74671614e-01 -5.09035647e-01 -1.46171010e+00
3.04246694e-01 -1.01820076e+00 2.46623755e-01 -9.80327427e-01
-1.46560240e+00 7.23264337e-01 -1.17374361e-01 -1.10450137e+00
-5.38478374e-01 1.18920401e-01 -8.06121945e-01 1.14197505e+00
-1.20793533e+00 -1.10922241e+00 -2.29924813e-01 9.34932083e-02
1.12459660e+00 -1.62402049e-01 4.72600460e-01 4.98244874e-02
-5.19507289e-01 9.68508899e-01 3.89902085e-01 -1.78119346e-01
7.09786952e-01 -1.05033267e+00 1.00451791e+00 9.30314302e-01
-8.54277536e-02 9.71674919e-01 7.91475475e-01 -8.98201466e-01
-9.66818810e-01 -1.56831670e+00 1.33517599e+00 -1.33877903e-01
5.06450415e-01 -1.40923202e-01 -9.56015766e-01 4.89096612e-01
4.86724406e-01 -7.99659610e-01 3.02825242e-01 -1.48476183e-01
-5.30526698e-01 -8.23605582e-02 -1.10798836e+00 9.56343412e-01
1.28213525e+00 -6.45113170e-01 -1.09450662e+00 2.60194778e-01
1.10780764e+00 -2.78277963e-01 -4.21765447e-01 3.04103643e-01
4.60616171e-01 -1.11349881e+00 9.78532434e-01 -7.22468257e-01
1.15946805e+00 -3.62554565e-02 -1.80757672e-01 -2.04223514e+00
-5.48624456e-01 -4.98892754e-01 -7.16127157e-02 1.17534423e+00
3.18903327e-01 -6.21166348e-01 9.55640733e-01 1.18908077e-01
-2.96032578e-01 -8.90971363e-01 -8.38182330e-01 -1.29995298e+00
4.11595464e-01 3.51298988e-01 8.67107272e-01 7.62661755e-01
3.61950427e-01 8.24876070e-01 -5.35529137e-01 -4.91910994e-01
4.65679556e-01 6.75506711e-01 7.39651442e-01 -2.96657324e-01
-6.13438070e-01 -5.33239424e-01 -8.78084376e-02 -1.30730951e+00
5.02141297e-01 -1.28393686e+00 5.70470691e-02 -1.72182918e+00
7.68914044e-01 2.58788288e-01 -1.18547730e-01 2.15528294e-01
-6.01481616e-01 2.66133636e-01 3.77047151e-01 3.69132787e-01
-4.02574182e-01 1.01608562e+00 1.38240266e+00 -1.37770623e-01
-2.49393240e-01 -1.01430744e-01 -1.08559895e+00 1.89105585e-01
9.79695737e-01 -6.37580454e-01 -6.10241532e-01 -7.91588604e-01
1.71082526e-01 1.18424192e-01 2.30707556e-01 -8.24710369e-01
-2.78538130e-02 -3.32603902e-01 -1.90288320e-01 -4.10979241e-01
1.22952752e-01 -1.24958284e-01 1.42850071e-01 8.56459379e-01
-8.97282541e-01 4.53929424e-01 1.21177554e-01 6.21567369e-01
-4.99646157e-01 -6.22380495e-01 9.17802691e-01 -3.32352757e-01
-4.25058514e-01 -2.73743838e-01 -1.14493869e-01 5.38286924e-01
7.79863417e-01 -2.11873174e-01 -4.70823914e-01 -5.15263736e-01
-8.95862356e-02 3.00618440e-01 8.86294365e-01 6.10821664e-01
7.63910949e-01 -1.49046791e+00 -1.17333019e+00 9.89511758e-02
2.18066707e-01 -3.21335912e-01 2.85659254e-01 2.88719773e-01
-4.41750556e-01 6.19222045e-01 -2.83297747e-01 -5.66836633e-02
-1.38382232e+00 5.22180021e-01 3.06003511e-01 -4.17721003e-01
-6.51447296e-01 7.14729309e-01 1.53814152e-01 -6.21121705e-01
-2.98096001e-01 -1.70019031e-01 1.56916440e-01 -3.44810545e-01
5.51259935e-01 3.60703439e-01 7.72695318e-02 -2.98434734e-01
-1.45840362e-01 3.22476715e-01 -5.16807854e-01 -9.94005352e-02
1.05691731e+00 -2.81562477e-01 -1.23065077e-01 2.40827516e-01
1.46319759e+00 -2.51045618e-02 -7.37014771e-01 -3.33456516e-01
-1.27757028e-01 -8.13892543e-01 -2.65594631e-01 -1.06095767e+00
-8.79468799e-01 6.76179588e-01 -9.14032608e-02 -1.67362675e-01
1.10586524e+00 -2.19977796e-01 1.74460232e+00 6.63310349e-01
3.44896257e-01 -8.71635020e-01 4.04985756e-01 8.23633254e-01
1.19923532e+00 -8.88579428e-01 -1.22525580e-01 -2.49551326e-01
-9.82082963e-01 7.17007041e-01 7.15239525e-01 -3.07039738e-01
-1.44626141e-01 -3.83216262e-01 2.24060193e-03 1.39781445e-01
-8.75893414e-01 2.39752308e-01 1.62766561e-01 2.46100962e-01
2.70041972e-01 -1.35871962e-01 -6.51057243e-01 6.00706220e-01
-5.38105965e-01 2.96470344e-01 6.71955287e-01 9.44789767e-01
-5.51551938e-01 -1.25895619e+00 -7.53354356e-02 6.43768907e-01
-3.19815040e-01 -5.81197023e-01 -1.01276219e+00 1.20796427e-01
-3.14060539e-01 1.00119042e+00 -2.34455749e-01 -6.26135409e-01
5.51424682e-01 -1.19022623e-01 4.92583990e-01 -9.55213189e-01
-8.00339639e-01 -6.02754653e-01 2.25722194e-01 -4.61782455e-01
2.20042124e-01 -4.71852779e-01 -1.00953817e+00 -4.33229297e-01
-7.81257674e-02 4.47360754e-01 3.11935872e-01 9.26264226e-01
9.26638186e-01 2.99394071e-01 9.07012939e-01 -4.00840342e-01
-1.20814753e+00 -1.02259278e+00 -3.05885017e-01 9.25275385e-01
1.93637684e-01 -2.13273801e-02 -3.26362759e-01 -2.41916820e-01] | [11.735238075256348, 9.301515579223633] |
ce7c1b70-3dc6-47f7-addc-97c6d32b5996 | cross-project-software-vulnerability-1 | 2209.10406 | null | https://arxiv.org/abs/2209.10406v1 | https://arxiv.org/pdf/2209.10406v1.pdf | Cross Project Software Vulnerability Detection via Domain Adaptation and Max-Margin Principle | Software vulnerabilities (SVs) have become a common, serious and crucial concern due to the ubiquity of computer software. Many machine learning-based approaches have been proposed to solve the software vulnerability detection (SVD) problem. However, there are still two open and significant issues for SVD in terms of i) learning automatic representations to improve the predictive performance of SVD, and ii) tackling the scarcity of labeled vulnerabilities datasets that conventionally need laborious labeling effort by experts. In this paper, we propose a novel end-to-end approach to tackle these two crucial issues. We first exploit the automatic representation learning with deep domain adaptation for software vulnerability detection. We then propose a novel cross-domain kernel classifier leveraging the max-margin principle to significantly improve the transfer learning process of software vulnerabilities from labeled projects into unlabeled ones. The experimental results on real-world software datasets show the superiority of our proposed method over state-of-the-art baselines. In short, our method obtains a higher performance on F1-measure, the most important measure in SVD, from 1.83% to 6.25% compared to the second highest method in the used datasets. Our released source code samples are publicly available at https://github.com/vannguyennd/dam2p | ['Dinh Phung', 'Hung Nguyen', 'John Grundy', 'Chakkrit Tantithamthavorn', 'Trung Le', 'Van Nguyen'] | 2022-09-19 | cross-project-software-vulnerability | https://openreview.net/forum?id=f6R69En9_tH | https://openreview.net/pdf?id=f6R69En9_tH | null | ['vulnerability-detection'] | ['miscellaneous'] | [-1.48691133e-01 -7.64973834e-02 -6.13768250e-02 -2.00312063e-01
-1.02745569e+00 -8.46172750e-01 2.25898936e-01 1.24783970e-01
-1.71665862e-01 3.07813883e-01 -5.08330837e-02 -6.47078037e-01
2.44292207e-02 -6.77494049e-01 -5.49159110e-01 -4.93913144e-01
9.40254852e-02 -2.75616258e-01 4.91227776e-01 -6.68533891e-02
4.72851843e-01 2.00508222e-01 -1.23112273e+00 8.34659785e-02
1.08551908e+00 9.68773663e-01 5.50748892e-02 3.05286884e-01
-2.75670648e-01 6.71766460e-01 -4.04914975e-01 -7.62393832e-01
4.08165842e-01 1.10422755e-02 -6.75005913e-01 -4.83196527e-01
2.74764806e-01 -7.81418756e-02 -1.80226222e-01 1.51358044e+00
5.37205696e-01 -1.32354498e-01 6.52336419e-01 -1.18231428e+00
-1.09305036e+00 5.95727921e-01 -1.04659700e+00 3.65540028e-01
1.50301263e-01 7.22616389e-02 9.48683977e-01 -8.55955184e-01
4.41193640e-01 9.85987902e-01 7.78504729e-01 5.86112082e-01
-9.82508659e-01 -8.38651299e-01 1.45426750e-01 3.21718633e-01
-1.22631371e+00 -2.35137865e-01 9.23739374e-01 -9.57980990e-01
8.32974374e-01 -6.21028654e-02 -1.67052671e-01 1.13786328e+00
2.27307919e-02 4.57471371e-01 1.04593050e+00 -5.34541130e-01
9.51691493e-02 5.52381039e-01 5.34947634e-01 5.99781692e-01
2.31149107e-01 -1.01610728e-01 3.39514576e-02 -4.24838006e-01
4.97126251e-01 9.92555767e-02 -1.99016765e-01 -3.06272000e-01
-7.96831012e-01 8.66980016e-01 2.73044646e-01 2.44864404e-01
-2.43282005e-01 -2.63621360e-01 7.94698536e-01 3.93890828e-01
5.94267428e-01 2.67349869e-01 -7.11008906e-01 -1.37573555e-01
-6.10119700e-01 -2.84377160e-03 5.30511260e-01 6.74101532e-01
6.57964647e-01 -3.11210975e-02 1.66827559e-01 9.53489363e-01
4.41765308e-01 1.16346695e-01 5.27280867e-01 -3.25049818e-01
8.62242043e-01 8.28157127e-01 -3.58720720e-02 -1.24147177e+00
9.25372634e-03 -2.41961077e-01 -5.03574431e-01 2.70862877e-01
1.99691489e-01 -2.66640663e-01 -5.70764899e-01 1.51848626e+00
3.29188854e-01 5.74282110e-01 3.26844424e-01 5.87771595e-01
8.24038088e-01 5.24053097e-01 3.36180657e-01 5.41623496e-02
1.45042586e+00 -9.16726589e-01 -3.61252874e-01 -2.05184206e-01
6.50620580e-01 -9.51652586e-01 1.24976742e+00 3.48578334e-01
-3.61080676e-01 -4.25244868e-01 -1.09658742e+00 1.41456112e-01
-3.90637368e-01 3.36342335e-01 6.31618321e-01 9.71796393e-01
-8.12245309e-01 3.02334189e-01 -6.75150633e-01 -2.33577073e-01
5.45697093e-01 1.65417001e-01 -5.04478335e-01 -3.42569984e-02
-1.12957478e+00 5.82072973e-01 3.18137854e-01 -1.44665152e-01
-9.51992929e-01 -7.43014157e-01 -6.30212069e-01 6.56453744e-02
5.31629503e-01 1.43278345e-01 1.25673473e+00 -7.74740338e-01
-1.18828344e+00 8.41687560e-01 1.96089178e-01 -6.55037761e-02
2.91132659e-01 -4.79500502e-01 -4.98267353e-01 -6.20774664e-02
3.26007158e-02 -2.10914612e-01 5.90206921e-01 -1.15999293e+00
-3.76554996e-01 -3.28667223e-01 2.10428596e-01 -3.34517747e-01
-1.04668128e+00 3.52537602e-01 -2.65327036e-01 -7.16579378e-01
-1.60568178e-01 -7.10133195e-01 -1.28498748e-01 -3.03681046e-01
-3.32039654e-01 -4.16112572e-01 6.78020895e-01 -1.09581625e+00
1.53459513e+00 -2.30112267e+00 1.26884645e-02 -2.62591280e-02
1.59130856e-01 8.41267943e-01 -2.03469798e-01 4.92024690e-01
-4.34566200e-01 1.13358542e-01 -4.11826670e-01 -1.92550018e-01
-1.12278946e-02 -4.43826586e-01 -6.06211007e-01 3.57308745e-01
3.29074562e-01 4.33854729e-01 -7.79913843e-01 -3.23035508e-01
-1.45617977e-01 4.96969134e-01 -3.43891203e-01 4.47312385e-01
-8.09342042e-02 3.04474294e-01 -7.10056484e-01 7.45617867e-01
9.78978932e-01 -2.09272131e-02 2.52968613e-02 7.48265982e-02
-8.27785358e-02 1.39922082e-01 -1.28022873e+00 1.52662408e+00
-5.40966332e-01 3.14057678e-01 -1.81553409e-01 -1.13657379e+00
1.24436486e+00 3.99790943e-01 9.10244808e-02 -3.65323663e-01
4.95433025e-02 2.84630060e-01 -1.79832354e-01 -8.53515387e-01
1.27926379e-01 6.83148280e-02 -2.58281529e-01 4.21651661e-01
5.96180893e-02 4.93775576e-01 -1.22548878e-01 1.42071337e-01
1.24193704e+00 1.42682239e-01 2.77235478e-01 -2.73959786e-01
9.86621261e-01 -2.82140732e-01 8.64926219e-01 1.04144424e-01
-3.87443841e-01 3.21507573e-01 1.00804627e+00 -4.08917844e-01
-8.37457597e-01 -7.96056449e-01 -7.62883201e-02 8.69601488e-01
-1.09375916e-01 -2.37358734e-01 -9.61682439e-01 -1.23584890e+00
2.93849651e-02 6.10015988e-01 -5.52672625e-01 -2.18064010e-01
-4.18804258e-01 -7.69062102e-01 6.36106968e-01 6.79581344e-01
3.16954046e-01 -9.44115639e-01 -2.25386187e-01 -5.42450063e-02
-3.61967608e-02 -1.18109143e+00 -4.01221901e-01 -9.94982570e-02
-8.10193181e-01 -1.19148803e+00 -6.90015674e-01 -8.66053879e-01
7.30376065e-01 3.25406462e-01 7.91542649e-01 1.20941155e-01
-3.66410911e-01 1.72706172e-01 -6.50846064e-01 -2.28016898e-01
-2.76970267e-01 1.47719905e-01 1.21089118e-02 -2.55047455e-02
7.84285903e-01 -6.53667331e-01 -5.58078170e-01 2.04447329e-01
-8.08684230e-01 -5.59988379e-01 8.57105434e-01 5.74159205e-01
3.34595442e-01 2.00019374e-01 8.60442460e-01 -1.12032032e+00
8.00241768e-01 -1.00225163e+00 -9.39318419e-01 4.56314832e-01
-7.34171093e-01 3.94470990e-02 6.38755500e-01 -4.51767653e-01
-1.25422907e+00 1.67413160e-01 -6.00438938e-02 -4.31615442e-01
-1.71949491e-01 5.28778017e-01 -3.02937210e-01 -2.47038454e-01
8.52542698e-01 9.57867429e-02 -2.29565427e-01 -7.98670292e-01
3.22821170e-01 1.08448541e+00 2.93583691e-01 -8.78198445e-01
9.71610606e-01 1.65091362e-02 -3.49592030e-01 -4.55829054e-01
-5.51979005e-01 -5.33164442e-01 -5.94873905e-01 1.84263319e-01
8.13331783e-01 -9.42933381e-01 -4.84765470e-01 6.09266758e-01
-1.07653236e+00 6.96847290e-02 4.16668862e-01 2.18730330e-01
-4.15811054e-02 1.01977551e+00 -5.40235162e-01 -9.75907564e-01
-6.00728154e-01 -1.42770553e+00 6.56453669e-01 2.85354495e-01
2.52830833e-01 -8.07867348e-01 3.88660043e-01 4.42181140e-01
3.01863372e-01 4.38385814e-01 1.18190455e+00 -8.26046348e-01
-3.52273732e-01 -3.71615231e-01 -4.62696970e-01 7.15309024e-01
1.61850825e-02 9.13554877e-02 -1.11618149e+00 -3.00164282e-01
-3.00900489e-02 -2.60819048e-01 7.32994616e-01 -2.15006366e-01
1.07466578e+00 -8.70713219e-02 -2.52691686e-01 3.77600610e-01
1.76269138e+00 2.61133552e-01 5.87197959e-01 6.16583288e-01
6.60078406e-01 7.33205616e-01 7.69255221e-01 3.65758091e-01
4.31129664e-01 5.10836065e-01 5.58024347e-01 2.43727192e-01
2.07725465e-01 -1.75729454e-01 5.47568262e-01 1.01479709e+00
5.83999455e-02 8.93438756e-02 -1.39152074e+00 7.46451497e-01
-1.85052681e+00 -6.15952373e-01 -2.34966382e-01 2.39495420e+00
6.74006104e-01 1.59968197e-01 9.72807407e-02 1.31149665e-01
9.20160770e-01 -3.45462225e-02 -4.58678722e-01 -4.04346317e-01
3.37396353e-01 1.22087613e-01 2.41708398e-01 1.28746599e-01
-1.45741880e+00 9.07860458e-01 4.87523127e+00 7.72214115e-01
-1.02596509e+00 3.66835058e-01 3.66219968e-01 3.93367678e-01
-1.87300608e-01 1.00437917e-01 -7.78621137e-01 5.89177787e-01
9.20936704e-01 -1.85504228e-01 1.68196529e-01 1.34839284e+00
-2.60345936e-01 2.36677215e-01 -8.55107665e-01 8.01739931e-01
2.58303769e-02 -8.27329278e-01 -2.99818575e-01 -3.40777934e-02
7.55163133e-01 -1.94183975e-01 3.07818234e-01 5.46392739e-01
1.94594994e-01 -6.30508840e-01 4.58518952e-01 3.60961735e-01
6.63582504e-01 -9.28026557e-01 9.15953815e-01 2.96342045e-01
-1.30510294e+00 -3.93684119e-01 -6.88938260e-01 2.41749525e-01
-3.33396256e-01 8.13959181e-01 -5.96560657e-01 7.16685236e-01
7.81251013e-01 5.48805594e-01 -7.29582667e-01 1.02050149e+00
-5.26438653e-01 7.03638256e-01 3.38406742e-01 1.07221223e-01
2.99647469e-02 -7.15848058e-02 2.71261364e-01 1.16581738e+00
4.79634553e-01 6.68882160e-03 2.98926346e-02 8.19476843e-01
-1.26687437e-01 3.75505567e-01 -6.30428076e-01 -3.49910945e-01
5.32445848e-01 1.44980848e+00 -5.28873980e-01 3.83878425e-02
-8.51467490e-01 7.05729902e-01 5.83599925e-01 2.34956801e-01
-8.14413726e-01 -8.38522851e-01 8.12216520e-01 -7.75471851e-02
3.77255589e-01 -1.14279807e-01 -1.31076962e-01 -1.30521858e+00
5.08196712e-01 -1.04329777e+00 4.52377737e-01 -1.08573422e-01
-1.66587496e+00 8.88999581e-01 -1.31894350e-01 -1.45226777e+00
1.27768859e-01 -8.13572586e-01 -9.02442038e-01 1.04846632e+00
-1.74351621e+00 -1.18375802e+00 -3.24748814e-01 3.41577858e-01
3.95696133e-01 -6.13113701e-01 8.15341175e-01 6.70794487e-01
-9.40727472e-01 9.74592686e-01 3.07626426e-01 3.76785100e-01
9.02081728e-01 -1.14077890e+00 7.01227784e-01 1.28149199e+00
-2.20857069e-01 9.41934347e-01 2.85824627e-01 -7.76150227e-01
-1.18442011e+00 -9.82706726e-01 5.76284111e-01 -6.47365212e-01
8.62240434e-01 -3.37084562e-01 -1.21685231e+00 6.05881035e-01
7.69299939e-02 8.99769291e-02 9.07495379e-01 1.65927425e-01
-9.92505670e-01 -5.24069630e-02 -1.33102429e+00 2.59080261e-01
6.09274268e-01 -6.03093505e-01 -5.42126894e-01 2.06122532e-01
7.24440753e-01 -1.09565437e-01 -1.06955433e+00 2.47744575e-01
3.86099756e-01 -9.12295043e-01 9.59306240e-01 -6.10910177e-01
6.22564018e-01 -3.27839762e-01 -2.12191015e-01 -1.07229304e+00
-2.82963902e-01 -2.68115550e-01 -1.39079809e-01 1.78195083e+00
3.46214831e-01 -8.59707296e-01 6.49417341e-01 6.28179193e-01
1.40054420e-01 -8.19312871e-01 -6.84897184e-01 -8.85527074e-01
4.20359045e-01 -2.81117529e-01 6.36989653e-01 1.29265964e+00
2.02011019e-01 -3.83267812e-02 -2.24277303e-01 5.56201994e-01
9.04977441e-01 -7.05611259e-02 6.31919861e-01 -1.48399282e+00
-4.49922144e-01 -2.66292155e-01 -6.60097539e-01 -4.32832479e-01
3.87237579e-01 -8.63730073e-01 -1.53432399e-01 -1.17119896e+00
4.89765644e-01 -4.28287119e-01 -6.69200599e-01 5.63818038e-01
-5.94614744e-01 -1.91139311e-01 -2.86165383e-02 3.10297906e-01
-3.21793765e-01 2.28030041e-01 6.15627766e-01 -3.62681560e-02
2.92883068e-02 4.77370694e-02 -9.96793687e-01 6.43100619e-01
8.88327837e-01 -7.43239045e-01 -5.37756264e-01 -5.69345713e-01
1.62051246e-01 -6.14123121e-02 7.98988640e-02 -8.53812635e-01
-2.35625021e-02 -7.03544216e-03 2.57183295e-02 -5.49084321e-02
-1.93621010e-01 -6.27035737e-01 -1.70284614e-01 4.19528574e-01
-1.82504877e-01 1.31345481e-01 3.44406754e-01 6.50105894e-01
-1.37671396e-01 -7.11509705e-01 8.03816676e-01 -5.48723601e-02
-1.03437233e+00 1.88006639e-01 4.07791138e-02 1.21346176e-01
1.44336212e+00 2.20998213e-01 -5.75144589e-01 2.19241917e-01
-2.92450488e-01 7.09073693e-02 4.11078095e-01 7.91736066e-01
4.98652637e-01 -1.14073360e+00 -7.26672828e-01 2.57233270e-02
3.28263909e-01 -3.35434705e-01 5.26155114e-01 5.91998637e-01
-4.44477230e-01 3.03735524e-01 -3.16787094e-01 -1.26881301e-01
-1.36273289e+00 9.19694424e-01 -3.00852153e-02 -2.94291168e-01
-5.77185750e-01 1.10421968e+00 2.09289491e-01 -6.84970379e-01
3.66464555e-01 1.33996770e-01 -6.31270707e-01 -1.44019186e-01
6.88838482e-01 5.70348144e-01 1.92998350e-01 -5.94768882e-01
-5.88841081e-01 7.26194918e-01 -4.81840938e-01 2.83524096e-01
1.61232531e+00 3.28518420e-01 -2.48908684e-01 6.38797730e-02
1.30750775e+00 -1.34300692e-02 -1.09695494e+00 -4.30097491e-01
4.11701977e-01 -7.36859024e-01 2.50906814e-02 -8.15984368e-01
-1.20564783e+00 1.17529178e+00 9.14131641e-01 1.55503944e-01
1.08749759e+00 -1.51746094e-01 8.83979499e-01 2.64720410e-01
4.15292948e-01 -6.87566698e-01 1.73829541e-01 1.21768355e-01
7.24672556e-01 -1.41596532e+00 -1.12607576e-01 -5.48715115e-01
-7.16634870e-01 1.15980279e+00 9.45172787e-01 -1.91164136e-01
6.89022064e-01 2.70493090e-01 3.65751535e-01 -2.73736715e-02
-5.15137792e-01 2.37158343e-01 1.28864959e-01 7.97599018e-01
6.54620886e-01 -2.79377159e-02 -3.43260795e-01 1.06975102e+00
3.73785347e-01 -2.54185706e-01 5.77827573e-01 8.28305066e-01
-3.28886032e-01 -1.56362879e+00 -3.94517154e-01 2.68106729e-01
-8.63743067e-01 -5.00854328e-02 -3.24801207e-01 2.78779089e-01
4.59070690e-02 9.42651868e-01 -7.19755471e-01 -6.29343390e-01
4.89203870e-01 1.71079606e-01 6.59728795e-02 -7.72637546e-01
-6.41615808e-01 -3.09892595e-01 -2.29347914e-01 -3.57320607e-01
-5.30993901e-02 -8.02675784e-01 -9.16764259e-01 1.59963630e-02
-3.60908955e-01 2.08611861e-01 6.79055691e-01 5.33299983e-01
5.36172271e-01 5.11343300e-01 8.43037188e-01 -4.62667793e-01
-1.10523570e+00 -9.06164467e-01 -3.34096074e-01 3.60550761e-01
1.15457743e-01 -8.70017529e-01 -4.57506716e-01 7.16462657e-02] | [7.10014533996582, 7.7795891761779785] |
9bb992dc-369c-415e-b15e-8ac8b52fe4c0 | towards-adaptive-unknown-authentication-for | 2207.04494 | null | https://arxiv.org/abs/2207.04494v1 | https://arxiv.org/pdf/2207.04494v1.pdf | Towards Adaptive Unknown Authentication for Universal Domain Adaptation by Classifier Paradox | Universal domain adaptation (UniDA) is a general unsupervised domain adaptation setting, which addresses both domain and label shifts in adaptation. Its main challenge lies in how to identify target samples in unshared or unknown classes. Previous methods commonly strive to depict sample "confidence" along with a threshold for rejecting unknowns, and align feature distributions of shared classes across domains. However, it is still hard to pre-specify a "confidence" criterion and threshold which are adaptive to various real tasks, and a mis-prediction of unknowns further incurs misalignment of features in shared classes. In this paper, we propose a new UniDA method with adaptive Unknown Authentication by Classifier Paradox (UACP), considering that samples with paradoxical predictions are probably unknowns belonging to none of the source classes. In UACP, a composite classifier is jointly designed with two types of predictors. That is, a multi-class (MC) predictor classifies samples to one of the multiple source classes, while a binary one-vs-all (OVA) predictor further verifies the prediction by MC predictor. Samples with verification failure or paradox are identified as unknowns. Further, instead of feature alignment for shared classes, implicit domain alignment is conducted in output space such that samples across domains share the same decision boundary, though with feature discrepancy. Empirical results validate UACP under both open-set and universal UDA settings. | ['Songcan Chen', 'Yao Liu', 'Yunyun Wang'] | 2022-07-10 | null | null | null | null | ['universal-domain-adaptation'] | ['computer-vision'] | [ 5.33546388e-01 -2.59893417e-01 -3.56099725e-01 -5.99501610e-01
-7.92122006e-01 -8.70286822e-01 4.18833703e-01 2.56199270e-01
-1.07358791e-01 1.09338450e+00 -2.82992631e-01 -3.09028804e-01
-7.97369182e-02 -7.20912278e-01 -5.15038729e-01 -9.24710512e-01
3.02440643e-01 6.37543380e-01 2.83094674e-01 7.08748773e-02
-3.90551127e-02 2.66053438e-01 -1.54025602e+00 3.99294943e-01
9.08051312e-01 9.67539132e-01 -2.41844594e-01 4.43512708e-01
-6.10045902e-02 1.70065641e-01 -1.01268244e+00 -5.05401254e-01
4.46203411e-01 -6.87839031e-01 -6.59843028e-01 4.67006899e-02
5.24039567e-01 -2.97479540e-01 2.20662907e-01 1.22266412e+00
4.27315116e-01 -1.63052112e-01 1.06163168e+00 -1.88583016e+00
-6.76165104e-01 4.31202590e-01 -4.47492361e-01 2.87245004e-03
3.64770979e-01 1.46985501e-01 1.10413432e+00 -8.22097003e-01
4.36962694e-01 8.96727979e-01 6.57020092e-01 5.24257660e-01
-1.59869587e+00 -1.26804757e+00 3.20194572e-01 2.08760649e-01
-1.64749646e+00 -3.84283245e-01 5.61285973e-01 -6.22055531e-01
3.40223193e-01 4.30444300e-01 8.58282577e-03 1.30767274e+00
1.27394587e-01 4.92247105e-01 1.23319256e+00 -2.31641442e-01
6.35136187e-01 6.61053061e-01 4.15771425e-01 1.19772993e-01
4.26619112e-01 1.39863834e-01 -7.07002044e-01 -6.45193994e-01
2.69171864e-01 -1.04926955e-02 -2.58193791e-01 -7.27292895e-01
-1.08001184e+00 7.24632084e-01 -1.75443560e-01 -1.83985531e-01
-1.71956673e-01 -8.95635366e-01 2.78705835e-01 6.68367147e-01
2.13413373e-01 2.25360334e-01 -1.04725075e+00 2.38992259e-01
-6.15698040e-01 1.07941076e-01 8.73924553e-01 1.15932930e+00
1.14377677e+00 -1.53399006e-01 -1.86474815e-01 8.86052966e-01
9.60998237e-02 6.60839021e-01 7.09158182e-01 -3.84082615e-01
2.15416506e-01 7.42347658e-01 3.12592149e-01 -6.79386556e-01
-7.19657540e-02 -4.79598492e-01 -9.30658162e-01 3.75680506e-01
8.12169969e-01 -1.87996417e-01 -7.82031417e-01 1.96032560e+00
6.13689482e-01 5.84113777e-01 4.72342312e-01 9.68164563e-01
4.20597047e-01 2.13903904e-01 9.00159031e-02 -2.98215061e-01
1.48245633e+00 -3.40833992e-01 -3.87986660e-01 -1.92670465e-01
5.29067576e-01 -6.98287249e-01 1.00657964e+00 5.10247886e-01
-2.49942243e-01 -7.33023345e-01 -1.28560650e+00 4.45104063e-01
-5.09216428e-01 1.16202146e-01 1.20651193e-01 9.67991233e-01
-3.94611984e-01 2.65150160e-01 -2.19952330e-01 -3.40534449e-01
3.09622884e-01 6.35133147e-01 -5.48672557e-01 7.05804210e-03
-1.29495800e+00 5.35973310e-01 2.63232976e-01 -4.05917138e-01
-6.89904571e-01 -6.36608720e-01 -6.94021046e-01 -9.57438443e-03
2.47210026e-01 -4.48655874e-01 1.10731447e+00 -1.13806200e+00
-1.24273205e+00 8.91025841e-01 -3.51868391e-01 -4.05312091e-01
3.68925154e-01 2.09826902e-02 -1.10542834e+00 -4.36049193e-01
2.95417309e-01 3.41243297e-01 1.02438450e+00 -1.36703289e+00
-1.23540866e+00 -4.23499048e-01 -3.44970435e-01 4.74007102e-03
-4.00084049e-01 -2.42221624e-01 1.11295968e-01 -4.71976012e-01
2.13558331e-01 -6.91503465e-01 7.62250498e-02 3.71079370e-02
-4.86719996e-01 -3.33921701e-01 1.10349607e+00 -3.37139368e-01
1.23206270e+00 -2.37336230e+00 -2.80789167e-01 6.18113399e-01
2.21808776e-01 1.59789607e-01 -1.14132844e-01 -7.16917170e-03
-2.90894747e-01 -1.57884985e-01 -5.43879092e-01 3.38403322e-02
1.25274912e-01 4.74918604e-01 -8.41913998e-01 5.88845551e-01
4.99072224e-01 1.21899784e-01 -8.08598101e-01 -2.53511161e-01
-1.34294227e-01 -5.43108620e-02 -5.07307529e-01 3.92622411e-01
3.48050222e-02 4.52555835e-01 -2.34801486e-01 9.27675724e-01
1.17345357e+00 -3.77453089e-01 6.06985450e-01 -1.11603156e-01
2.45301336e-01 1.41258955e-01 -1.78457475e+00 8.42580974e-01
2.14423135e-01 1.77912533e-01 -1.37369847e-02 -1.05411255e+00
1.40873897e+00 3.09774298e-02 3.55936974e-01 -3.63817275e-01
-1.01624191e-01 4.16223943e-01 2.06500039e-01 2.20493022e-02
4.73489195e-01 -1.53304070e-01 -2.58336931e-01 4.46316779e-01
6.15779422e-02 2.69969672e-01 -2.60224164e-01 -1.11921519e-01
1.08082068e+00 1.05632022e-01 6.62581444e-01 -1.39337897e-01
8.04260910e-01 3.20528969e-02 1.07796907e+00 8.40118647e-01
-7.42272496e-01 8.01935077e-01 5.35522580e-01 -3.92199084e-02
-9.99459565e-01 -1.40804839e+00 -4.11556423e-01 1.21095169e+00
5.03304064e-01 -4.77814078e-02 -3.20471764e-01 -9.92850661e-01
4.90726471e-01 6.26760423e-01 -5.07344425e-01 -2.61348337e-01
-4.69829813e-02 -5.73612213e-01 6.58562064e-01 2.98881590e-01
1.98315546e-01 -5.64767838e-01 -1.41170681e-01 2.64466912e-01
1.19148202e-01 -9.06254649e-01 -4.81178701e-01 5.48455417e-01
-3.12201709e-01 -1.62165165e+00 -4.82334048e-01 -7.80446947e-01
8.02494586e-01 1.86915159e-01 7.93192744e-01 -1.94134831e-01
1.76152468e-01 9.30757225e-02 -3.93732935e-01 -3.10743332e-01
-5.46314418e-01 -1.12249628e-01 6.50095284e-01 4.84953314e-01
1.02603376e+00 -3.43610585e-01 -2.29165241e-01 9.35631335e-01
-7.55107403e-01 -4.62432206e-01 4.23098594e-01 1.08526325e+00
6.48972690e-01 5.90171963e-02 9.32888865e-01 -1.15513301e+00
6.60955906e-01 -1.03918815e+00 -4.41314697e-01 5.01527667e-01
-8.96323502e-01 -1.06948406e-01 8.15496206e-01 -9.66616154e-01
-7.03339636e-01 3.13350767e-01 1.55543700e-01 -4.96802658e-01
-5.73182166e-01 1.07022308e-01 -6.11119628e-01 2.73799509e-01
9.97967005e-01 3.60199988e-01 1.77327529e-01 -3.65207344e-01
-1.28179699e-01 9.94248688e-01 7.35626757e-01 -7.83214092e-01
1.01374936e+00 1.91052243e-01 -4.71774518e-01 -4.36124951e-01
-4.92935300e-01 -5.65320432e-01 -8.80270779e-01 2.20977843e-01
2.48978913e-01 -1.01981699e+00 -4.59564418e-01 6.09093428e-01
-7.22209096e-01 -1.51319355e-01 -3.55847657e-01 4.70330089e-01
-7.31039941e-02 6.67397141e-01 -1.71253225e-03 -6.99803233e-01
-1.55062139e-01 -1.04585838e+00 7.77831495e-01 4.02812511e-01
-6.36886835e-01 -6.09748542e-01 -1.34412184e-01 -1.05779201e-01
6.76256418e-02 -6.03817292e-02 8.49954784e-01 -1.53455567e+00
4.73844595e-02 -2.37327665e-01 -2.29577973e-01 5.29152930e-01
5.28435588e-01 1.01263905e-02 -1.17281842e+00 -4.79166716e-01
-1.68667987e-01 -1.73681825e-01 1.91984504e-01 6.74087852e-02
9.73727524e-01 -1.81641385e-01 -4.50587660e-01 3.79430950e-01
9.45967257e-01 4.52351034e-01 2.73901850e-01 2.77768373e-01
1.85674816e-01 4.66090262e-01 8.63163888e-01 7.96303570e-01
3.64305764e-01 5.58630109e-01 1.55146003e-01 9.89018232e-02
1.53537467e-01 -3.27107728e-01 4.73885655e-01 2.01510474e-01
7.58431613e-01 -3.29149723e-01 -7.71783173e-01 4.34107035e-01
-1.64140117e+00 -7.96387494e-01 -3.04238439e-01 2.70554662e+00
1.13029921e+00 1.60529181e-01 1.36607707e-01 4.42845821e-01
1.13069510e+00 -5.34592390e-01 -1.02773285e+00 -1.76284328e-01
-5.81322372e-01 2.48632669e-01 3.40282977e-01 4.11521196e-01
-1.17885756e+00 5.69797218e-01 5.61383247e+00 7.80032277e-01
-1.17894316e+00 -8.77797380e-02 6.96527421e-01 5.53035498e-01
-1.34534225e-01 1.86103731e-01 -1.23838520e+00 7.27201045e-01
7.00809956e-01 -3.47993761e-01 1.14038298e-02 1.03554082e+00
-3.62422228e-01 -8.14936385e-02 -1.44992518e+00 8.00496519e-01
-1.19726777e-01 -8.43134999e-01 -2.95118988e-02 8.50891843e-02
6.20782971e-01 -3.39184552e-01 8.19244087e-02 5.54862082e-01
6.72019958e-01 -5.04209042e-01 5.26560545e-01 1.76190674e-01
1.12589824e+00 -5.67785859e-01 8.26961279e-01 4.83514339e-01
-1.09754074e+00 -2.25801721e-01 -4.18477595e-01 -4.36226614e-02
-4.08354998e-01 4.93589222e-01 -1.15703058e+00 5.51350355e-01
6.49623871e-01 5.57083964e-01 -4.72342610e-01 9.58179653e-01
-5.27708754e-02 7.69539893e-01 -4.17492837e-01 2.71178007e-01
-3.68164897e-01 -2.90074255e-02 3.90017867e-01 9.51275110e-01
4.60857868e-01 1.02638431e-01 4.55656439e-01 5.59130788e-01
7.34211281e-02 9.16399509e-02 -4.06156093e-01 2.68470764e-01
1.02433002e+00 9.07159746e-01 -2.89580762e-01 -6.26059175e-01
-3.14115286e-01 1.01878178e+00 6.41606450e-02 4.14318651e-01
-7.96348572e-01 -3.57872963e-01 1.25427866e+00 -2.05809250e-03
2.59702444e-01 1.23518735e-01 -5.12143672e-01 -1.32092214e+00
-1.54348180e-01 -1.07496774e+00 1.06758189e+00 -3.66814256e-01
-2.05763221e+00 3.69753271e-01 -3.02501917e-01 -1.95115042e+00
-1.21203162e-01 -6.47296607e-01 -4.31293070e-01 1.17191541e+00
-1.50287366e+00 -8.90093923e-01 -2.25818828e-01 9.83490109e-01
2.50073075e-01 -4.48221356e-01 1.07353425e+00 3.09943020e-01
-5.65849483e-01 1.27063489e+00 3.85983467e-01 2.07499847e-01
1.43987882e+00 -1.01804650e+00 -8.40114132e-02 8.52009654e-01
-2.12008983e-01 7.08850801e-01 6.78275824e-01 -7.40091205e-01
-1.14231038e+00 -1.18993509e+00 9.29476380e-01 -4.09980595e-01
7.22249687e-01 -3.47425222e-01 -1.44158566e+00 6.07502341e-01
-2.66123354e-01 2.19764918e-01 1.18740523e+00 1.36601076e-01
-7.26267278e-01 -2.86633700e-01 -1.58801186e+00 2.71143973e-01
6.30558252e-01 -5.60085237e-01 -4.49853480e-01 1.71259537e-01
4.67384487e-01 -3.65540743e-01 -9.57232296e-01 4.12168920e-01
5.37474513e-01 -6.38984263e-01 8.31707597e-01 -7.75416434e-01
-1.70852803e-02 -8.15451324e-01 -4.76284236e-01 -1.17449236e+00
-3.53637189e-01 -2.80319512e-01 -5.52752381e-03 1.56350696e+00
3.99891376e-01 -1.13656938e+00 8.64848971e-01 6.82044744e-01
5.89065813e-02 -1.07030779e-01 -1.16750848e+00 -9.47330475e-01
6.87325299e-02 -3.30419153e-01 1.02580404e+00 1.56519604e+00
-2.60245614e-02 3.09646547e-01 -4.42742616e-01 8.38419378e-01
5.67733824e-01 1.58574224e-01 1.22700560e+00 -1.50724077e+00
-2.83765286e-01 -2.61803448e-01 -2.49773383e-01 -9.73221958e-01
8.71245861e-02 -7.82625854e-01 -6.72155805e-03 -5.02496004e-01
-1.55068226e-02 -8.78243923e-01 -6.56713426e-01 8.12967062e-01
-2.25033134e-01 7.55109265e-02 -2.46496275e-01 4.80455637e-01
-1.93352982e-01 2.90563047e-01 6.59381390e-01 -1.96970180e-01
-4.57363397e-01 2.99065173e-01 -8.58181477e-01 2.97283113e-01
8.64209354e-01 -4.94563073e-01 -2.40630448e-01 1.00473940e-01
-3.36500764e-01 -1.29788175e-01 4.27840859e-01 -1.08940470e+00
3.64089012e-01 -4.90036070e-01 7.15467095e-01 -6.31454647e-01
7.11039975e-02 -1.02529573e+00 3.58514339e-01 4.98950481e-01
-2.25125030e-01 -1.40232161e-01 8.69629309e-02 6.78860486e-01
-1.88264683e-01 7.99573287e-02 8.98169935e-01 4.84855145e-01
-1.00780368e+00 2.00705215e-01 -2.21589223e-01 -2.16511115e-01
1.15661013e+00 -5.83074868e-01 -5.21235943e-01 -9.41267796e-03
-6.84157670e-01 2.85315245e-01 4.41206127e-01 4.54159558e-01
5.55936396e-01 -1.24930418e+00 -8.31920922e-01 8.92669618e-01
5.57990193e-01 -1.12600230e-01 9.47897434e-02 3.98171902e-01
2.80653477e-01 -7.85565749e-02 -2.14399934e-01 -8.79605293e-01
-1.53958583e+00 3.62260401e-01 2.15276718e-01 -5.25867678e-02
-1.66265756e-01 8.07526827e-01 3.36429000e-01 -9.77889061e-01
3.44775058e-02 1.58517156e-02 -8.59017149e-02 1.40590474e-01
4.58096802e-01 2.23655347e-02 -4.67906594e-02 -6.03174388e-01
-5.24150789e-01 1.27883926e-01 -3.00787866e-01 3.21032465e-01
7.98744857e-01 -1.18272871e-01 1.75323665e-01 2.10096523e-01
8.83000910e-01 7.84810632e-02 -1.28999662e+00 -7.82815337e-01
2.66999472e-02 -5.40285110e-01 -7.01260269e-01 -1.04574513e+00
-5.56961119e-01 6.89332902e-01 9.89040375e-01 6.31934702e-02
1.22926116e+00 -2.28630170e-01 4.65095013e-01 1.30007759e-01
3.43747050e-01 -1.07662833e+00 -2.49869347e-01 5.96552610e-01
4.19591576e-01 -1.39970767e+00 -2.15173271e-02 -3.38867635e-01
-8.92966211e-01 1.04437339e+00 1.15558732e+00 2.45844141e-01
5.45766771e-01 1.75827429e-01 7.50548542e-02 4.81352985e-01
-7.13605285e-01 -1.10105528e-02 4.93101515e-02 1.05007780e+00
3.92187461e-02 3.46769959e-01 -1.99699089e-01 1.16079378e+00
-1.66271687e-01 -1.15018427e-01 3.45089912e-01 8.42091858e-01
-4.37120020e-01 -1.40473175e+00 -8.27583194e-01 5.39548576e-01
-9.66296569e-02 2.06317484e-01 -3.79079461e-01 5.69294274e-01
5.78727901e-01 9.11206901e-01 2.76559174e-01 -9.33588088e-01
3.63138139e-01 3.93757105e-01 -3.00015837e-01 -6.30612910e-01
-5.69943309e-01 -1.66012570e-01 -9.81981009e-02 1.23217337e-01
2.64198631e-02 -7.88880825e-01 -1.26336026e+00 -2.97628760e-01
-4.19664055e-01 2.96709269e-01 3.01994145e-01 8.43294561e-01
5.49567044e-01 -5.19881165e-03 9.85615253e-01 -9.07769129e-02
-9.64254498e-01 -6.90822244e-01 -7.94424295e-01 6.15072608e-01
4.09409106e-01 -8.57829213e-01 -4.48265404e-01 2.62285233e-01] | [10.32612419128418, 3.156432867050171] |
eb48409b-50a4-4d12-aee5-526bdc3b6347 | anti-unification-and-generalization-a-survey | 2302.00277 | null | https://arxiv.org/abs/2302.00277v5 | https://arxiv.org/pdf/2302.00277v5.pdf | Anti-unification and Generalization: A Survey | Anti-unification (AU) is a fundamental operation for generalization computation used for inductive inference. It is the dual operation to unification, an operation at the foundation of automated theorem proving. Interest in AU from the AI and related communities is growing, but without a systematic study of the concept nor surveys of existing work, investigations often resort to developing application-specific methods that existing approaches may cover. We provide the first survey of AU research and its applications and a general framework for categorizing existing and future developments. | ['Temur Kutsia', 'David M. Cerna'] | 2023-02-01 | null | null | null | null | ['automated-theorem-proving', 'automated-theorem-proving'] | ['miscellaneous', 'reasoning'] | [ 6.93448722e-01 4.96636271e-01 -6.54395223e-01 -9.97685343e-02
-3.55225243e-02 -7.17701256e-01 8.42741013e-01 9.56321135e-02
6.77891150e-02 1.18800092e+00 -3.71980667e-01 -1.21521187e+00
-2.63987660e-01 -1.21301138e+00 -7.32997954e-01 -3.54764193e-01
-2.15966359e-01 4.98521388e-01 9.46504623e-02 -6.05120718e-01
3.50959271e-01 3.91470790e-01 -2.27877808e+00 3.06840122e-01
1.03125465e+00 8.71732473e-01 -5.47079265e-01 7.93139696e-01
-4.08489376e-01 9.83070910e-01 -5.45714915e-01 -6.43232763e-01
2.17885599e-01 -5.30791163e-01 -1.47254038e+00 -3.89202476e-01
5.57817578e-01 -5.62430695e-02 -1.86686680e-01 1.45571351e+00
-6.87288269e-02 2.25730445e-02 5.81335366e-01 -1.99160361e+00
-6.10426605e-01 1.08814025e+00 4.58780117e-02 3.91656607e-01
1.01527071e+00 -5.28796434e-01 1.44342303e+00 -8.91898796e-02
7.14798391e-01 1.06836486e+00 6.98047280e-01 8.70284617e-01
-1.34816945e+00 -7.34169722e-01 -1.39619380e-01 6.82136595e-01
-1.60588753e+00 -3.55757415e-01 6.26129270e-01 -4.42769825e-01
1.01835167e+00 9.02735353e-01 4.74347860e-01 7.01580465e-01
4.70344543e-01 8.28916550e-01 1.06853366e+00 -1.00854182e+00
9.35854167e-02 3.67940068e-01 3.07599902e-01 8.07712555e-01
4.32670593e-01 1.85725585e-01 -4.95800555e-01 -3.70521903e-01
7.92150617e-01 -3.97877038e-01 -2.34936267e-01 -3.38428557e-01
-1.11665618e+00 8.51640761e-01 1.46307340e-02 7.69693375e-01
1.05798371e-01 5.13090156e-02 5.20441413e-01 9.66399550e-01
2.07381010e-01 6.75718367e-01 -4.63113040e-01 -7.91188627e-02
-8.79839242e-01 1.01400673e+00 1.13749266e+00 1.29725909e+00
6.02440119e-01 -1.07710533e-01 3.10041070e-01 -9.78383273e-02
1.30961640e-02 1.72873750e-01 1.76443189e-01 -1.07714570e+00
-1.10560572e-02 8.77846658e-01 -9.86019243e-03 -7.80457258e-01
-1.62452117e-01 -3.13260078e-01 -5.43707788e-01 -3.89969870e-02
2.14375868e-01 4.12608357e-03 -2.44832888e-01 1.81477606e+00
2.27915317e-01 4.89266396e-01 5.52234054e-01 2.67174721e-01
8.82433832e-01 3.09800655e-01 -1.10060729e-01 -5.35845816e-01
8.80627632e-01 -4.53340679e-01 -8.51767302e-01 1.21022448e-01
1.27674401e+00 -4.69429374e-01 3.32688123e-01 5.76212347e-01
-1.11804056e+00 -1.06501490e-01 -1.32165802e+00 -2.37988323e-01
-8.09075594e-01 -5.31059206e-01 1.58820415e+00 7.72642076e-01
-9.72889543e-01 4.88003403e-01 -3.30409825e-01 -2.70089447e-01
1.62595928e-01 5.95464647e-01 -3.50585133e-01 -2.96521336e-01
-1.80417693e+00 1.30564141e+00 7.28745759e-01 -7.61352777e-02
-3.89503807e-01 -1.00821626e+00 -1.21924162e+00 -2.88929582e-01
7.00477481e-01 -7.01255202e-01 1.41367733e+00 -6.38299584e-01
-1.25285637e+00 8.90556335e-01 -1.88825667e-01 -8.83986115e-01
2.45481417e-01 1.03692979e-01 -8.77672315e-01 -1.15925074e-01
2.78826207e-01 2.62118369e-01 2.88538128e-01 -1.04539001e+00
-1.05371046e+00 -4.02433604e-01 7.17264056e-01 -2.17122078e-01
1.45954400e-01 2.04960570e-01 4.65440959e-01 -3.48483235e-01
4.67186987e-01 -6.43155158e-01 -7.37079233e-02 -2.90452480e-01
-1.70535058e-01 -1.04266787e+00 8.26796830e-01 3.78970765e-02
1.66726005e+00 -1.63752317e+00 3.73644412e-01 4.47539359e-01
3.89098674e-01 -2.73182616e-02 5.62740386e-01 5.69123328e-01
-3.59537572e-01 3.08014721e-01 -2.05859676e-01 3.68294358e-01
4.28394258e-01 4.41545427e-01 -7.45058417e-01 3.51881474e-01
-2.93884594e-02 9.04734731e-01 -1.33483934e+00 -7.73649752e-01
4.28081542e-01 -4.04684842e-01 -6.80360496e-01 -1.18074216e-01
-6.50246620e-01 -1.63949937e-01 -4.99256134e-01 9.50976431e-01
4.79245484e-01 -2.05751192e-02 4.39704835e-01 6.80701882e-02
-3.58092993e-01 6.55241013e-01 -1.37887180e+00 1.59074521e+00
-2.31231675e-01 4.86251771e-01 -2.15131372e-01 -1.60003531e+00
7.00080574e-01 6.34712458e-01 3.45033109e-01 -1.76506847e-01
2.58502513e-01 4.96176898e-01 3.40863019e-01 -4.52254504e-01
3.44001234e-01 -4.93887097e-01 -3.51567477e-01 5.02964795e-01
2.34486803e-01 -4.63596612e-01 7.01074719e-01 3.79175842e-01
9.39747274e-01 2.84867704e-01 6.94262445e-01 -7.45448172e-01
1.08570659e+00 2.98743248e-01 4.63791639e-01 8.09610784e-01
1.19706683e-01 -3.18210870e-01 5.64962745e-01 -6.72722459e-01
-6.56291544e-01 -8.35795999e-01 -6.99132502e-01 1.01247394e+00
2.30136067e-01 -1.15807772e+00 -5.86672425e-01 -7.38865972e-01
1.95469603e-01 9.96523917e-01 -7.26440668e-01 -1.15618028e-01
-2.82959908e-01 -3.89377475e-01 9.44598258e-01 7.58703530e-01
4.39794213e-01 -7.95088530e-01 -5.10506153e-01 3.95840146e-02
-3.98321748e-01 -9.84218299e-01 7.07996011e-01 2.34526604e-01
-8.32008481e-01 -1.29270923e+00 3.09873372e-01 -5.68025887e-01
3.32661182e-01 -1.47090048e-01 1.25525022e+00 5.58519483e-01
-5.84429763e-02 2.68207014e-01 -3.11310261e-01 -9.58964288e-01
-7.71345079e-01 4.47701954e-04 3.02515656e-01 -6.79617226e-01
8.75176072e-01 -7.48031616e-01 4.42441344e-01 1.45815033e-02
-8.49746823e-01 -1.16941728e-01 5.89566492e-02 5.91894507e-01
4.00164936e-05 1.08057246e-01 4.99356091e-01 -1.13512290e+00
6.05207503e-01 -3.55272114e-01 -5.84939361e-01 5.58468819e-01
-7.00350404e-01 1.46269530e-01 4.29742634e-01 1.13976831e-02
-5.62322676e-01 -6.57136858e-01 1.11175187e-01 -3.64445732e-03
-2.49187097e-01 7.75131106e-01 -1.58997640e-01 -4.90764409e-01
8.07616889e-01 5.14131039e-02 -2.20315494e-02 9.72629935e-02
3.40779841e-01 5.76212347e-01 6.86041117e-01 -1.14584279e+00
7.14514792e-01 1.50012910e-01 6.51409984e-01 -7.56570399e-01
-1.03268731e+00 -2.29678228e-01 -4.85143721e-01 -3.96787189e-02
2.46255502e-01 -2.28115737e-01 -1.02085257e+00 -8.72165859e-02
-1.05557561e+00 -2.42150817e-02 -3.14433843e-01 3.34855825e-01
-6.77065194e-01 4.47537124e-01 -3.97278666e-02 -1.00181508e+00
-2.74930418e-01 -9.11947489e-01 4.54085022e-01 1.24651305e-01
-6.08347356e-01 -1.28463197e+00 -2.95706429e-02 2.67439961e-01
8.93198550e-02 3.10043782e-01 1.04624116e+00 -7.97746003e-01
-1.29701942e-01 -4.93714601e-01 8.42235163e-02 3.71741921e-01
-1.03848539e-02 2.54129022e-01 -8.84966016e-01 2.17067569e-01
2.98686735e-02 -3.31245273e-01 2.99823880e-01 6.90099448e-02
1.11974978e+00 -3.63820493e-01 -6.87321067e-01 1.81005016e-01
1.34511244e+00 -6.98874667e-02 7.37386823e-01 4.74440783e-01
9.14664753e-03 2.78586119e-01 6.85602605e-01 7.63413832e-02
3.58413488e-01 4.42461312e-01 2.57234871e-01 3.17212999e-01
2.46345192e-01 -5.56255784e-03 -2.37075407e-02 1.77500695e-01
-8.48454416e-01 4.25165892e-02 -1.08270347e+00 3.25063109e-01
-1.86583114e+00 -1.50563693e+00 -1.14911474e-01 2.04224443e+00
1.10088098e+00 4.67127115e-01 -5.77819794e-02 1.01458871e+00
7.14291513e-01 -3.91155899e-01 1.63124308e-01 -9.36008334e-01
-4.88007702e-02 8.00499380e-01 1.25383794e-01 6.29494250e-01
-8.16702008e-01 1.10800290e+00 8.13883495e+00 5.72908342e-01
-8.25871110e-01 -2.28984863e-01 -3.86137009e-01 5.59653044e-01
-4.96086925e-01 4.13349688e-01 -7.10039735e-01 -2.23067254e-01
7.77347982e-01 -5.75408936e-01 3.66149217e-01 9.51762438e-01
-6.77568972e-01 -1.99450478e-01 -1.83582485e+00 5.19714653e-01
3.51473600e-01 -1.62944496e+00 1.77947909e-01 -2.31090322e-01
7.64130533e-01 -4.12336707e-01 -3.96178991e-01 5.12965083e-01
1.28932104e-01 -8.38317573e-01 4.68061894e-01 1.38315663e-01
5.91927946e-01 -8.26635718e-01 1.05966377e+00 3.51994514e-01
-9.25057650e-01 -6.60268366e-02 -1.39862522e-01 -9.73879576e-01
-3.76018673e-01 6.86903670e-02 -6.10351503e-01 1.18422246e+00
3.11984032e-01 5.67097902e-01 -6.25611395e-02 6.07757449e-01
-6.22445464e-01 5.15316606e-01 -4.07018870e-01 -1.24474473e-01
1.27962455e-01 -1.84983671e-01 6.60659552e-01 1.17535126e+00
-2.66150802e-01 2.95486450e-01 9.82426107e-02 1.00992012e+00
1.24610886e-01 -3.09016168e-01 -9.18619156e-01 -8.97604302e-02
5.16386151e-01 7.83601820e-01 -2.60297835e-01 -7.83232689e-01
-4.18335170e-01 3.12236607e-01 1.29475184e-02 2.60417443e-03
-9.04376388e-01 -5.87157369e-01 5.63358247e-01 -8.58505294e-02
-8.40308368e-02 -3.36470604e-01 -5.39801598e-01 -1.25180435e+00
-1.39828861e-01 -1.17119384e+00 8.06282103e-01 -3.51885676e-01
-9.15901482e-01 2.11965024e-01 5.83476782e-01 -8.46758723e-01
-7.96609223e-01 -7.28644967e-01 -7.12130129e-01 7.15808868e-01
-1.38510215e+00 -1.18983245e+00 4.36375514e-02 4.99169856e-01
-8.72145146e-02 8.78029391e-02 1.54267359e+00 -4.71652932e-02
-3.03202033e-01 8.54753375e-01 -4.94186759e-01 -9.02979225e-02
3.15893054e-01 -1.20501447e+00 -3.02738175e-02 1.07621694e+00
2.98608951e-02 1.12848151e+00 1.24015403e+00 -4.05140936e-01
-1.75206816e+00 -4.11682576e-01 1.46275783e+00 -5.89506567e-01
1.16095221e+00 -1.84733942e-01 -7.01024950e-01 1.25379133e+00
-6.61269873e-02 1.44073572e-02 7.87231743e-01 7.65219748e-01
-5.25958478e-01 -6.08697161e-02 -1.31820953e+00 6.11821353e-01
1.16734505e+00 -5.75697064e-01 -1.34686303e+00 3.76778185e-01
4.90357578e-01 -5.48087716e-01 -1.32050455e+00 8.23997080e-01
6.50070488e-01 -7.18339801e-01 6.99938118e-01 -1.03666031e+00
5.12833357e-01 -4.72370356e-01 -2.82660753e-01 -4.89771187e-01
-2.03830495e-01 -9.66897905e-01 -5.89196086e-01 1.10834837e+00
6.21050715e-01 -9.29626763e-01 5.61201513e-01 9.14559484e-01
-3.03799421e-01 -7.16061592e-01 -9.21769023e-01 -7.38145769e-01
3.32579911e-01 -9.00054514e-01 6.92925513e-01 1.24582291e+00
1.12282634e+00 5.15130579e-01 3.42526555e-01 1.52441561e-01
5.80391049e-01 7.67019570e-01 8.89833152e-01 -1.53121066e+00
2.28264600e-01 -6.68096781e-01 -1.00397730e+00 -5.79721808e-01
5.62728286e-01 -1.30747056e+00 -2.90936053e-01 -1.39684653e+00
-1.50376931e-01 -4.56377923e-01 -1.49456903e-01 6.50992334e-01
1.70341864e-01 2.54022300e-01 -4.57460105e-01 -2.55511880e-01
-4.63370621e-01 7.56305382e-02 1.10508823e+00 -8.82930011e-02
4.83340882e-02 7.25800544e-02 -7.80881464e-01 7.48260856e-01
8.35380256e-01 -1.21342354e-02 -4.12535816e-01 1.07865885e-01
7.33667672e-01 -3.36305767e-01 3.52917224e-01 -1.15045941e+00
2.44280994e-01 -5.75125754e-01 -2.61864603e-01 -2.70556659e-01
-2.97693044e-01 -6.23572350e-01 -2.66287457e-02 4.50953454e-01
-2.07722217e-01 6.67669028e-02 2.74572730e-01 -3.96622382e-02
-8.65843147e-02 -6.39395773e-01 2.93707520e-01 -1.07041270e-01
-9.93747354e-01 3.51119637e-02 -3.12232077e-01 2.45336983e-02
1.06098688e+00 -2.34319896e-01 -6.88829497e-02 1.54214233e-01
-3.19207847e-01 2.84389943e-01 3.40167016e-01 2.05002159e-01
2.33212069e-01 -1.08490205e+00 -7.57516921e-01 2.78671496e-02
1.82436317e-01 1.74205080e-01 -2.77174264e-01 9.24111664e-01
-3.40405524e-01 7.81045854e-01 -2.59425044e-01 -2.61927187e-01
-1.08482742e+00 8.62554789e-01 3.03640604e-01 -9.68757272e-03
-5.21431029e-01 8.63975525e-01 -3.82153355e-02 -1.06950887e-01
1.87176436e-01 -4.44963813e-01 -7.33593553e-02 -5.49770951e-01
8.31194460e-01 4.42515820e-01 2.20081180e-01 -2.79212117e-01
-5.90434313e-01 2.08265528e-01 4.15148102e-02 8.21572095e-02
1.11068845e+00 3.35306972e-01 -8.48178029e-01 6.09534144e-01
5.70950627e-01 -3.73306811e-01 2.53059566e-01 -2.33341798e-01
-1.97853558e-02 -5.95130250e-02 2.31927723e-01 -4.33300972e-01
-2.18003854e-01 3.96586388e-01 -3.71687949e-01 7.45393097e-01
7.95312464e-01 1.77030548e-01 3.15147907e-01 7.88043678e-01
5.44327080e-01 -9.66493607e-01 -7.68248200e-01 7.00912118e-01
9.41783011e-01 -8.28875780e-01 4.86049205e-01 -9.26486969e-01
2.39134431e-02 1.19318128e+00 6.50707364e-01 -9.37816054e-02
5.53852141e-01 6.99937940e-01 -4.80605125e-01 -1.36176124e-01
-7.43034780e-01 -1.59734428e-01 3.55060130e-01 6.53442681e-01
8.59418929e-01 3.42041887e-02 -7.37460375e-01 5.80037057e-01
-7.69824862e-01 4.46638286e-01 4.07627463e-01 1.44168115e+00
-3.40737492e-01 -1.42235839e+00 -3.46635222e-01 4.72597271e-01
-4.44468707e-01 -2.57145315e-01 -5.98528028e-01 1.26815557e+00
5.66148102e-01 9.17200506e-01 -2.86669098e-02 -5.58632433e-01
-4.32135127e-02 3.08124274e-01 1.34857035e+00 -5.85065305e-01
-4.19858605e-01 -9.28762555e-01 4.17492568e-01 -3.33195835e-01
-8.94146025e-01 -7.23576546e-01 -1.35552382e+00 -9.27570224e-01
-8.95408630e-01 7.07403183e-01 3.58095974e-01 1.45815957e+00
-2.20814332e-01 3.67682308e-01 3.05818766e-01 -1.72826663e-01
-3.19173813e-01 -8.11999559e-01 -4.40385550e-01 -2.07231268e-02
1.91130310e-01 -1.00098014e+00 -4.61969167e-01 -3.53620830e-03] | [8.855131149291992, 6.92598819732666] |
b0bde91c-681a-4601-98d9-a340e30e4997 | encoding-based-saliency-detection-for-videos | null | null | http://openaccess.thecvf.com/content_cvpr_2015/html/Mauthner_Encoding_Based_Saliency_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Mauthner_Encoding_Based_Saliency_2015_CVPR_paper.pdf | Encoding Based Saliency Detection for Videos and Images | We present a novel video saliency detection method to support human activity recognition and weakly supervised training of activity detection algorithms. Recent research has emphasized the need for analyzing salient information in videos to minimize dataset bias or to supervise weakly labeled training of activity detectors. In contrast to previous methods we do not rely on training information given by either eye-gaze or annotation data, but propose a fully unsupervised algorithm to find salient regions within videos. In general, we enforce the Gestalt principle of figure-ground segregation for both appearance and motion cues. We introduce an encoding approach that allows for efficient computation of saliency by approximating joint feature distributions. We evaluate our approach on several datasets, including challenging scenarios with cluttered background and camera motion, as well as salient object detection in images. Overall, we demonstrate favorable performance compared to state-of-the-art methods in estimating both ground-truth eye-gaze and activity annotations. | ['Horst Bischof', 'Thomas Mauthner', 'Horst Possegger', 'Georg Waltner'] | 2015-06-01 | null | null | null | cvpr-2015-6 | ['video-saliency-detection'] | ['computer-vision'] | [ 6.65734231e-01 8.94613788e-02 -5.92325866e-01 -4.01054800e-01
-4.61514324e-01 -4.72923398e-01 5.58111846e-01 1.25294477e-01
-4.89168733e-01 7.26289749e-01 4.00202572e-01 1.75684333e-01
2.90529221e-01 -5.18006012e-02 -7.78980136e-01 -5.71161389e-01
-5.39027117e-02 -2.38535821e-01 6.65147603e-01 1.91589653e-01
6.13247812e-01 3.07006866e-01 -2.11605740e+00 2.91464388e-01
6.86814725e-01 8.12996268e-01 3.02232623e-01 7.49259949e-01
5.32119095e-01 1.42300379e+00 -3.88602406e-01 1.71094373e-01
5.17923646e-02 -7.76488781e-01 -7.61084020e-01 7.64484346e-01
8.88083994e-01 -4.10058439e-01 -1.90069243e-01 9.77782786e-01
1.43883899e-01 3.58968616e-01 5.79459429e-01 -1.40212059e+00
-2.97739148e-01 -8.95628482e-02 -8.59429836e-01 1.02395165e+00
7.86300480e-01 2.32054025e-01 8.82545114e-01 -1.03042436e+00
7.10370064e-01 7.98155725e-01 2.33789712e-01 6.41903996e-01
-1.31260586e+00 -6.51858598e-02 4.25105810e-01 3.63613009e-01
-1.25381887e+00 -8.70258570e-01 8.10504436e-01 -5.61701953e-01
7.49548197e-01 3.72772664e-01 8.40644896e-01 1.15006447e+00
-1.50357947e-01 1.34622145e+00 9.31638956e-01 -6.82422698e-01
3.86758417e-01 1.60264969e-01 9.48835313e-02 9.50899482e-01
3.96957576e-01 -6.02023210e-03 -1.19279921e+00 -1.99008942e-01
8.95114958e-01 7.66157135e-02 -5.19550681e-01 -8.29646289e-01
-1.50312626e+00 4.01934534e-01 1.69726774e-01 1.23617046e-01
-4.61356878e-01 2.15178996e-01 1.22839749e-01 -3.21409613e-01
6.45614564e-01 2.76834607e-01 -1.72216639e-01 -1.68267697e-01
-1.46245515e+00 2.29291290e-01 4.28746939e-01 9.06464994e-01
8.05726349e-01 4.90747718e-03 -5.25025904e-01 3.66633475e-01
3.58732998e-01 3.94793957e-01 4.11991566e-01 -1.37501287e+00
9.98153351e-03 5.11985898e-01 4.99893069e-01 -9.03274357e-01
-2.06549525e-01 -1.32091105e-01 -1.38657391e-01 2.96272844e-01
6.86741173e-01 9.38193351e-02 -7.94965982e-01 1.66622019e+00
4.46867049e-01 3.62919390e-01 -2.93900162e-01 1.17928255e+00
4.67365056e-01 1.43964678e-01 1.33212730e-01 -5.26113808e-01
1.27660394e+00 -1.19219530e+00 -8.76062393e-01 -4.21327710e-01
5.91692328e-01 -5.20212352e-01 1.05174088e+00 3.32604468e-01
-1.39525282e+00 -5.14442086e-01 -8.91187072e-01 -2.76499748e-01
9.48059652e-03 2.70815939e-01 5.87955236e-01 5.00902355e-01
-1.21153307e+00 3.44287783e-01 -9.16891754e-01 -6.13945901e-01
7.16186702e-01 1.78127840e-01 -2.54297554e-01 1.74256340e-01
-5.19911647e-01 7.97867835e-01 1.04948483e-01 -1.71737552e-01
-1.35769629e+00 -4.29814398e-01 -1.10804439e+00 -1.06497377e-01
5.85078001e-01 -6.27936840e-01 1.40089428e+00 -1.61551118e+00
-1.14671957e+00 1.25095630e+00 -8.44387949e-01 -6.07444584e-01
3.20093930e-01 -4.54526305e-01 -4.17722538e-02 6.69329941e-01
1.50582761e-01 1.00865662e+00 1.05940711e+00 -1.01195729e+00
-8.33741903e-01 -1.34188980e-01 1.27519980e-01 4.85905111e-01
-2.24210337e-01 4.96696264e-01 -3.21795672e-01 -6.33924842e-01
-8.38844925e-02 -7.47988522e-01 9.01933946e-03 3.72284234e-01
-1.67765751e-01 -2.61595637e-01 1.00059986e+00 -3.97661567e-01
1.18353188e+00 -2.03728771e+00 4.10697684e-02 -1.22202590e-01
3.64958465e-01 -4.29427400e-02 1.65417433e-01 -1.56771094e-01
1.04005553e-01 -3.65282834e-01 -2.36066714e-01 -4.12273794e-01
-2.97098309e-01 5.94830886e-02 -1.77088961e-01 1.02478385e+00
4.98404294e-01 8.87224972e-01 -1.29550493e+00 -8.83726001e-01
3.19202453e-01 2.34857649e-01 -6.84536219e-01 1.92889854e-01
-2.73188978e-01 5.99219680e-01 -2.36569002e-01 9.99269903e-01
2.28951082e-01 -5.46838880e-01 1.18744536e-03 1.37401298e-01
-2.77533472e-01 3.77591401e-01 -1.00174105e+00 1.93248034e+00
3.42350245e-01 1.19055164e+00 -6.36059884e-03 -8.71496677e-01
3.51224363e-01 1.60692751e-01 5.58032632e-01 -4.71152157e-01
1.34927318e-01 -1.75275505e-01 -1.58323631e-01 -4.65736985e-01
5.07910907e-01 3.34355891e-01 3.09382766e-01 6.13965750e-01
3.53890002e-01 3.26986045e-01 4.05375898e-01 2.85645217e-01
9.91585433e-01 6.25901461e-01 4.59666401e-01 -6.30603731e-01
4.88653004e-01 1.38468454e-02 4.63979483e-01 6.35025918e-01
-7.08299637e-01 7.43651927e-01 3.38289291e-01 -3.56467158e-01
-8.02651525e-01 -9.57342148e-01 1.12034030e-01 1.69934654e+00
5.08149266e-01 -3.90338719e-01 -1.19527030e+00 -8.00651073e-01
-5.09515524e-01 5.29276729e-01 -9.19792175e-01 6.09042719e-02
-3.97615254e-01 -4.37996864e-01 1.69851303e-01 5.16287565e-01
2.85230815e-01 -1.14756930e+00 -1.40674853e+00 -3.27067167e-01
-3.54393750e-01 -1.16969299e+00 -8.94449532e-01 6.66623339e-02
-8.48087966e-01 -1.39266670e+00 -6.92380726e-01 -7.57576406e-01
1.00480354e+00 7.94172525e-01 1.22612667e+00 1.46767139e-01
-3.91216248e-01 8.83929431e-01 -1.49805576e-01 -4.23988253e-01
1.80114120e-01 -2.28014305e-01 1.95102036e-01 3.04251730e-01
7.89177537e-01 -1.88589811e-01 -8.97790074e-01 5.11941493e-01
-7.51160026e-01 1.63191184e-01 3.57588142e-01 4.72315609e-01
6.31319225e-01 -5.46868920e-01 1.73471961e-02 -4.96983945e-01
1.98931828e-01 -2.79735833e-01 -5.50015628e-01 2.05851212e-01
-3.12300831e-01 -4.40203808e-02 -2.34266758e-01 -3.71703207e-01
-1.08095002e+00 4.79249716e-01 7.15549409e-01 -6.17272139e-01
-4.66971517e-01 -1.14431255e-01 1.27252564e-01 -1.62652880e-01
9.68775511e-01 2.92751431e-01 -6.90626726e-02 1.27009936e-02
2.28899777e-01 3.48890036e-01 7.19961584e-01 -3.01238060e-01
4.07157868e-01 9.41779912e-01 -1.35343164e-01 -1.09572864e+00
-1.18739414e+00 -8.83188605e-01 -1.00841117e+00 -5.55206954e-01
9.48311925e-01 -1.18565953e+00 -4.94344860e-01 1.31306916e-01
-1.12807906e+00 -3.68345827e-01 -3.82768899e-01 5.29048860e-01
-9.75788832e-01 5.60058951e-01 -2.12724581e-01 -1.11108577e+00
-1.10243715e-01 -8.24541509e-01 1.40025890e+00 3.47139508e-01
-5.69982708e-01 -9.19855952e-01 5.25503196e-02 3.25715899e-01
7.75937736e-02 4.09245908e-01 -7.13444352e-02 -3.07651997e-01
-8.96457314e-01 1.06522918e-01 -8.63595977e-02 1.81895763e-01
1.07830614e-01 -4.59643127e-03 -1.27957141e+00 -1.52235836e-01
2.85523199e-02 -3.92564207e-01 7.31787741e-01 8.02243471e-01
9.51378226e-01 -1.96499959e-01 -4.97429579e-01 4.13920730e-01
9.14218426e-01 -2.40160748e-01 5.34842968e-01 2.49143332e-01
5.34298003e-01 9.18316960e-01 9.67282951e-01 4.39432740e-01
2.51023889e-01 6.01242542e-01 3.65279526e-01 -2.10239142e-01
-9.33214054e-02 -2.47287080e-01 5.86877704e-01 -1.08535953e-01
-3.88374448e-01 2.95423670e-03 -6.76419616e-01 9.18459892e-01
-2.06415319e+00 -1.31845248e+00 -2.02624694e-01 2.32626104e+00
7.35430717e-01 6.37229905e-02 6.36317194e-01 9.70926359e-02
8.33160222e-01 1.66535869e-01 -4.79932755e-01 1.62364677e-01
-1.19348422e-01 -2.23875314e-01 4.64911252e-01 2.95336783e-01
-1.32293034e+00 8.40822518e-01 7.40810537e+00 2.68399954e-01
-7.78728545e-01 2.49723122e-01 5.60444534e-01 -7.05282032e-01
7.61437789e-02 1.04827350e-02 -6.68564677e-01 5.45318842e-01
6.89899206e-01 -4.89861481e-02 3.44759524e-01 9.67188835e-01
4.90187854e-01 -7.81418502e-01 -1.32631469e+00 1.19831777e+00
6.10857308e-01 -1.20072103e+00 -3.66410881e-01 3.64710428e-02
8.74719381e-01 4.56322245e-02 -5.58420643e-02 -3.13197345e-01
5.00369817e-02 -7.62661636e-01 1.00754225e+00 6.02472067e-01
5.25840521e-01 -3.07061166e-01 1.79076955e-01 3.37835997e-01
-9.66223419e-01 -2.29886491e-02 -7.86126554e-02 -3.72462988e-01
7.21707717e-02 2.20686823e-01 -6.44397378e-01 -1.58636570e-01
8.59222353e-01 1.13998115e+00 -7.66215205e-01 1.12773490e+00
-4.91737694e-01 5.45712829e-01 -1.69930369e-01 -8.52431729e-02
6.59360215e-02 1.08074568e-01 6.39566302e-01 1.23444784e+00
-1.49777278e-01 -2.33220458e-02 2.25576043e-01 8.14540803e-01
6.18344359e-02 -9.75459665e-02 -7.04860330e-01 4.01252173e-02
1.59628093e-01 1.14138424e+00 -9.50993896e-01 -5.47775328e-01
-5.38919449e-01 1.14501429e+00 2.54885733e-01 5.08614779e-01
-8.44130039e-01 -1.69955119e-02 5.41668713e-01 6.09419465e-01
4.21780080e-01 -2.64825791e-01 -8.53029639e-02 -1.41565752e+00
1.32466599e-01 -5.67599416e-01 4.05460775e-01 -1.11956990e+00
-5.22028804e-01 2.24465251e-01 1.61787868e-01 -1.28172231e+00
-3.91763151e-01 -3.65406781e-01 -6.91139400e-01 5.20271897e-01
-1.62236679e+00 -9.59475815e-01 -6.16634429e-01 7.77661502e-01
8.51561427e-01 1.66997120e-01 3.70425880e-01 -9.32919234e-02
-4.19705957e-01 1.19730048e-01 -3.44766229e-01 -1.62908919e-02
7.60778427e-01 -1.21331668e+00 5.40815257e-02 1.29780221e+00
5.72800279e-01 6.46824121e-01 9.75832582e-01 -4.02303100e-01
-1.07882667e+00 -8.65403712e-01 7.22954631e-01 -9.99654412e-01
4.05462295e-01 -5.21003962e-01 -7.21428335e-01 7.32047558e-01
5.32680273e-01 3.85437191e-01 7.79681683e-01 -5.81076965e-02
9.48797259e-03 3.04902464e-01 -8.83133531e-01 4.20535743e-01
1.34187388e+00 -5.78000367e-01 -8.16507280e-01 2.69544035e-01
1.92006379e-01 -3.38213712e-01 -1.16218232e-01 2.80348361e-01
4.89509165e-01 -1.13175559e+00 7.32866585e-01 -6.36119545e-01
2.29755864e-01 -6.98883951e-01 8.58560726e-02 -7.60124683e-01
-3.13358217e-01 -7.57062078e-01 -7.91153371e-01 6.69923961e-01
4.71703261e-02 7.41195381e-02 1.01277602e+00 5.66450775e-01
-3.45466956e-02 -4.07931179e-01 -5.40781617e-01 -3.46027344e-01
-9.86408353e-01 -1.59169525e-01 -2.45955944e-01 7.33018875e-01
2.86300957e-01 3.20932806e-01 -4.00955617e-01 3.19967791e-02
8.99070561e-01 7.12523535e-02 8.94552886e-01 -1.17148733e+00
-9.81319547e-02 -3.18570346e-01 -6.41366959e-01 -1.24473584e+00
2.22355366e-01 -2.83207953e-01 2.53264993e-01 -1.24910176e+00
4.95701224e-01 5.56504786e-01 -4.89941269e-01 4.28459883e-01
-2.43181124e-01 6.20989680e-01 -2.03135610e-01 3.05854052e-01
-1.47927153e+00 3.48709226e-01 1.06384039e+00 8.42842087e-02
-1.82348192e-01 -2.57717162e-01 -7.73443699e-01 9.00157690e-01
5.48695505e-01 -3.58327001e-01 -5.89053690e-01 -2.32914537e-01
1.77533105e-02 -3.59598100e-01 8.32324028e-01 -1.11907673e+00
4.32772368e-01 -4.15945530e-01 6.51533246e-01 -6.31331205e-01
1.92692742e-01 -4.52407092e-01 -5.44381976e-01 1.73112780e-01
-4.00063396e-01 -3.65396976e-01 1.40600860e-01 8.93394887e-01
-1.66795045e-01 -8.28179270e-02 8.66172433e-01 -1.64036155e-01
-1.03352284e+00 7.23842680e-02 -5.22418618e-01 9.75819901e-02
1.24638939e+00 -6.84947491e-01 -2.97244102e-01 -3.86354476e-01
-6.52752459e-01 9.13280472e-02 8.84195030e-01 3.86942863e-01
6.28564715e-01 -1.12817049e+00 -4.99291182e-01 3.23758274e-01
4.35186714e-01 -2.03273982e-01 8.95175636e-02 1.26363218e+00
-3.12151074e-01 4.01785135e-01 -3.32940191e-01 -1.01657724e+00
-1.44125366e+00 7.94234812e-01 2.64442831e-01 4.49727237e-01
-2.61056393e-01 7.08937168e-01 5.34507930e-01 5.84017456e-01
4.27254975e-01 -4.79110718e-01 -2.41783768e-01 -4.43282649e-02
1.01569557e+00 4.00086045e-01 -2.15204298e-01 -7.94565976e-01
-5.27078927e-01 3.09078008e-01 6.30787313e-02 -1.12630397e-01
9.12891746e-01 -4.46136653e-01 1.07519172e-01 5.71352005e-01
6.88717782e-01 -2.56491154e-02 -1.77792633e+00 -1.97598755e-01
2.07652211e-01 -8.08467567e-01 1.45818755e-01 -4.42676991e-01
-6.24740601e-01 7.54630804e-01 5.55286825e-01 1.60756901e-01
1.10132778e+00 2.62112051e-01 1.28664643e-01 2.30124086e-01
3.05523604e-01 -1.27356505e+00 5.14189780e-01 7.34066069e-02
5.72260380e-01 -1.52098203e+00 1.74686611e-01 -3.30908567e-01
-8.01348865e-01 7.04940796e-01 7.06115544e-01 -2.89541662e-01
2.48710632e-01 -6.63923398e-02 -1.79529756e-01 -3.81958246e-01
-8.48543286e-01 -7.02982545e-01 6.52258933e-01 6.85050726e-01
5.86411059e-01 -4.36163068e-01 -8.51983801e-02 3.95618416e-02
5.01778543e-01 3.07031512e-01 5.48550010e-01 1.29403710e+00
-8.10343206e-01 -4.34453577e-01 -3.60799849e-01 3.25278997e-01
-6.39310300e-01 -3.65508981e-02 -6.11931860e-01 4.65112299e-01
-4.11562510e-02 9.37502861e-01 3.68063301e-01 1.07668340e-01
2.20916253e-02 8.45550820e-02 8.26759517e-01 -8.47182989e-01
-6.76105097e-02 3.13928574e-01 1.78749822e-02 -8.69459033e-01
-1.28932035e+00 -9.86941576e-01 -9.11803722e-01 3.04983616e-01
-3.82599056e-01 7.20981583e-02 2.80115694e-01 9.26436603e-01
4.29405093e-01 3.11994940e-01 3.48597586e-01 -1.26505554e+00
1.73508570e-01 -1.01237154e+00 -6.27200305e-01 5.80095112e-01
6.17218792e-01 -1.07410872e+00 -3.37455392e-01 7.67437279e-01] | [8.623332977294922, 0.38948023319244385] |
16c71e55-e79f-40cc-a1e5-29704101a883 | from-zero-to-hero-human-in-the-loop-entity | null | null | https://aclanthology.org/2020.acl-main.624 | https://aclanthology.org/2020.acl-main.624.pdf | From Zero to Hero: Human-In-The-Loop Entity Linking in Low Resource Domains | Entity linking (EL) is concerned with disambiguating entity mentions in a text against knowledge bases (KB). It is crucial in a considerable number of fields like humanities, technical writing and biomedical sciences to enrich texts with semantics and discover more knowledge. The use of EL in such domains requires handling noisy texts, low resource settings and domain-specific KBs. Existing approaches are mostly inappropriate for this, as they depend on training data. However, in the above scenario, there exists hardly annotated data, and it needs to be created from scratch. We therefore present a novel domain-agnostic Human-In-The-Loop annotation approach: we use recommenders that suggest potential concepts and adaptive candidate ranking, thereby speeding up the overall annotation process and making it less tedious for users. We evaluate our ranking approach in a simulation on difficult texts and show that it greatly outperforms a strong baseline in ranking accuracy. In a user study, the annotation speed improves by 35{\%} compared to annotating without interactive support; users report that they strongly prefer our system. An open-source and ready-to-use implementation based on the text annotation platform INCEpTION (https://inception-project.github.io) is made available. | ['Jan-Christoph Klie', 'Iryna Gurevych', 'Richard Eckart de Castilho'] | 2020-07-01 | null | null | null | acl-2020-6 | ['text-annotation'] | ['natural-language-processing'] | [-2.56352097e-01 3.21123064e-01 -1.16627894e-01 -1.92207694e-01
-7.72858500e-01 -9.13205504e-01 4.63572800e-01 8.00229073e-01
-9.78694975e-01 1.03315318e+00 3.03567857e-01 -4.48904693e-01
-2.70897567e-01 -5.92662394e-01 -3.95584136e-01 -1.93548366e-01
1.42495468e-01 9.17620003e-01 7.69265831e-01 -5.67877650e-01
1.46942735e-01 7.26507679e-02 -1.32933640e+00 2.42001876e-01
9.64081049e-01 3.32808137e-01 3.99302155e-01 4.50991541e-01
-4.20291811e-01 4.81794417e-01 -4.60296124e-01 -8.01686943e-01
-7.79650509e-02 -6.70715272e-02 -1.08928263e+00 -5.20939410e-01
-2.75418931e-03 2.37220794e-01 6.85208812e-02 1.02740669e+00
7.68116534e-01 2.87201464e-01 4.15186018e-01 -7.61962116e-01
-4.94725436e-01 9.17543292e-01 -2.49003738e-01 2.15189621e-01
6.17098689e-01 -3.93574268e-01 1.19230092e+00 -1.05605578e+00
8.83287311e-01 8.26483846e-01 7.30816007e-01 5.15312374e-01
-1.07943952e+00 -5.05170166e-01 8.53895992e-02 1.08804695e-01
-1.52768970e+00 -3.78707945e-01 3.00924331e-01 -3.76325816e-01
1.03422761e+00 4.97790128e-01 2.41915688e-01 9.32970881e-01
-5.87206781e-01 7.00407386e-01 8.04638386e-01 -9.03947115e-01
2.74987102e-01 5.43900728e-01 3.57117712e-01 4.86176759e-01
5.78740180e-01 -5.28920174e-01 -3.65001857e-01 -4.84432667e-01
3.22029918e-01 -3.26730013e-01 -5.08634806e-01 -2.75093943e-01
-1.18228269e+00 5.93538761e-01 3.89559977e-02 7.07881331e-01
-3.52875590e-01 -2.16008604e-01 5.02485633e-01 3.10725480e-01
5.55252016e-01 8.92557502e-01 -8.19194019e-01 -2.11701304e-01
-9.43070173e-01 3.58851582e-01 1.29990077e+00 8.80108595e-01
5.45502126e-01 -9.08130169e-01 1.38324602e-02 1.11490285e+00
1.23190477e-01 2.73940682e-01 6.26513243e-01 -6.35750771e-01
4.45359290e-01 7.67386138e-01 6.09158754e-01 -8.86199474e-01
-6.88289106e-01 -3.10522079e-01 -4.46621478e-01 -3.42413485e-01
6.24928653e-01 -3.82762760e-01 -5.09937108e-01 1.50026405e+00
5.70200622e-01 1.93931833e-01 2.13388801e-01 8.83927822e-01
1.10273886e+00 3.74986529e-01 4.90219504e-01 -3.27125311e-01
1.75595474e+00 -6.39862657e-01 -8.89813662e-01 -1.60049975e-01
1.02089357e+00 -1.02663434e+00 1.11536145e+00 2.86399841e-01
-7.97410011e-01 -1.87443018e-01 -7.26498663e-01 -1.93451475e-02
-6.65069401e-01 2.09452182e-01 5.31094551e-01 6.56467617e-01
-8.84966373e-01 6.24072194e-01 -6.05863392e-01 -8.85549724e-01
4.66550663e-02 3.62560600e-01 -4.68586504e-01 1.17525160e-01
-1.71843970e+00 1.05983853e+00 7.78251886e-01 -1.84579015e-01
-6.57527009e-03 -6.17012620e-01 -6.33429348e-01 -2.08067689e-02
9.47229683e-01 -7.26233304e-01 1.35436118e+00 -6.35386586e-01
-1.17514467e+00 6.63563311e-01 1.34513248e-02 -2.79628307e-01
5.05911171e-01 -5.86343706e-01 -5.51393151e-01 -4.59096655e-02
1.97364777e-01 3.86897773e-01 1.48119614e-01 -9.72069561e-01
-7.37660885e-01 -1.14707567e-03 3.86044413e-01 3.17389429e-01
-6.13715827e-01 2.70359665e-01 -8.46167505e-01 -5.77753901e-01
-2.32327655e-01 -9.34451461e-01 -4.11500484e-01 -4.85131651e-01
-3.10678989e-01 -4.26353753e-01 5.21265507e-01 -7.07564056e-01
1.60470867e+00 -1.77639520e+00 -2.32054397e-01 3.37091029e-01
4.20621708e-02 6.07313693e-01 2.39022732e-01 7.61182129e-01
1.11211732e-01 2.99847096e-01 -1.31809376e-02 -3.46095972e-02
1.41905963e-01 6.30746037e-02 -5.21759801e-02 1.17954850e-01
-8.57146084e-02 7.56818891e-01 -1.36893201e+00 -7.80726373e-01
-6.92880824e-02 4.49010164e-01 -4.78623152e-01 -1.40311956e-01
-3.59876662e-01 3.23042095e-01 -7.83427417e-01 4.24777538e-01
2.13941470e-01 -4.42117333e-01 7.00775027e-01 -1.24429837e-01
-8.05789903e-02 4.59507495e-01 -1.57960451e+00 1.74267912e+00
-4.69909370e-01 3.19333404e-01 -3.16338688e-01 -8.27608824e-01
5.85072100e-01 6.98113322e-01 2.90220499e-01 -3.34011912e-01
6.13396168e-02 4.95366484e-01 -2.01284721e-01 -6.89490378e-01
7.57023394e-01 2.76834428e-01 -1.30847335e-01 4.73093361e-01
-6.02154136e-02 2.93396741e-01 5.84275723e-01 5.04980206e-01
1.28490877e+00 2.46044174e-01 6.84861541e-01 -4.10839826e-01
6.47172213e-01 1.57025263e-01 4.12576079e-01 6.01667404e-01
2.66528994e-01 8.85260850e-02 2.72580534e-01 -2.58186877e-01
-8.25266421e-01 -4.83268797e-01 -1.65629834e-01 1.30957580e+00
5.12667894e-02 -7.62454867e-01 -7.41265595e-01 -9.33037460e-01
-1.20863028e-01 8.36916447e-01 -4.57784027e-01 3.19594800e-01
-4.19663221e-01 -7.36383975e-01 5.84298313e-01 3.34653050e-01
1.65008977e-01 -1.11417496e+00 -5.34371793e-01 4.09583628e-01
-5.49003243e-01 -1.12788939e+00 -2.60706276e-01 1.74190223e-01
-5.60902774e-01 -1.04298294e+00 -6.11482561e-01 -6.32773340e-01
7.54327953e-01 -4.79149483e-02 1.30003119e+00 3.21566671e-01
-1.53598748e-02 3.49855840e-01 -7.67885089e-01 -4.90650982e-01
-3.02132308e-01 4.27970260e-01 7.38456473e-02 -4.81213242e-01
8.01934421e-01 -4.06066358e-01 -6.14717603e-01 4.33021814e-01
-7.98980713e-01 -7.27508292e-02 5.10377228e-01 8.06445837e-01
2.89211363e-01 6.55134544e-02 8.18765938e-01 -1.45583856e+00
8.15649033e-01 -5.94371438e-01 -5.22538185e-01 4.62606311e-01
-8.33272874e-01 1.96852684e-01 5.09272516e-01 -3.73003095e-01
-9.59402621e-01 1.76490694e-01 -3.81274313e-01 9.70183760e-02
-2.24607900e-01 9.01195467e-01 -1.61390737e-01 1.72077119e-01
1.06554711e+00 -2.49104634e-01 -5.94149113e-01 -7.17747629e-01
5.65514326e-01 8.52891982e-01 4.24349159e-01 -7.78362095e-01
6.70245707e-01 1.03576347e-01 -3.69142294e-01 -6.18621707e-01
-8.58853519e-01 -1.04237771e+00 -6.87153161e-01 6.35844050e-03
4.57878619e-01 -9.13878024e-01 -6.70354426e-01 -3.79153252e-01
-9.95695591e-01 -2.28619263e-01 -1.48745865e-01 4.78015512e-01
-5.31783886e-02 4.80648905e-01 -2.29534194e-01 -6.43553913e-01
-4.74089980e-01 -6.80898368e-01 8.57991755e-01 2.46303141e-01
-6.24022484e-01 -1.09653568e+00 2.37028092e-01 3.77293706e-01
2.93666393e-01 -1.03023713e-02 6.51103199e-01 -1.35424423e+00
-1.44027835e-02 -4.70755428e-01 2.45427117e-02 -1.15377642e-01
-4.02350873e-02 -2.15710461e-01 -9.29311097e-01 -1.14233181e-01
-6.41082585e-01 -1.90140218e-01 6.05563939e-01 -1.64103612e-01
9.39053893e-01 -3.46511185e-01 -7.11720765e-01 -1.42879009e-01
1.30800772e+00 -7.11622015e-02 4.72868264e-01 6.76428378e-01
4.87516403e-01 8.13894212e-01 8.62587154e-01 4.76775408e-01
4.95815486e-01 9.00249720e-01 -7.42198452e-02 -1.21193327e-01
2.72017896e-01 -1.30939364e-01 -5.73193692e-02 6.02474391e-01
-3.91212642e-01 -5.40267944e-01 -1.10415769e+00 8.24132979e-01
-2.10560632e+00 -9.40824747e-01 -2.46057808e-01 2.37443233e+00
1.34040928e+00 1.36821061e-01 1.51959851e-01 1.37766644e-01
7.12166488e-01 -5.42874157e-01 -1.90982044e-01 5.78613393e-02
1.65617213e-01 3.21085721e-01 4.81881022e-01 6.18034899e-01
-1.15201139e+00 1.15803528e+00 4.81863308e+00 1.02854455e+00
-8.00608337e-01 3.02855372e-01 3.34001303e-01 9.50926021e-02
-2.00583607e-01 1.17681831e-01 -1.02622807e+00 3.94554704e-01
1.04799950e+00 -3.75612080e-01 -4.70797904e-02 9.21858311e-01
1.64723307e-01 -8.06819126e-02 -8.51390958e-01 6.21140957e-01
-2.62073189e-01 -1.19754660e+00 -2.65722305e-01 -2.08528951e-01
3.38596433e-01 2.68908795e-02 -5.35813093e-01 4.55724210e-01
6.91952705e-01 -6.80026948e-01 4.49620724e-01 3.28288168e-01
6.33918345e-01 -5.78374803e-01 1.10910571e+00 4.92526770e-01
-1.05683529e+00 2.43863389e-01 -3.09095681e-01 3.63134444e-01
2.15681106e-01 8.76000702e-01 -1.14926910e+00 7.61148155e-01
7.86384583e-01 2.18972400e-01 -5.21003544e-01 1.00203168e+00
-3.93844604e-01 6.43669486e-01 -4.52646583e-01 -2.67581791e-01
-3.19608636e-02 2.74553627e-01 5.12754202e-01 1.79330719e+00
2.37708643e-01 4.43113416e-01 3.16438973e-01 2.57200032e-01
-2.79415846e-01 7.40284085e-01 -2.66666144e-01 7.36594107e-03
7.69657910e-01 1.65551269e+00 -1.06017172e+00 -4.37473565e-01
-4.09327239e-01 9.04388905e-01 2.81145960e-01 1.42012760e-01
-7.03055441e-01 -6.56792700e-01 1.99008912e-01 3.27689826e-01
2.93552667e-01 7.06477761e-02 6.54097199e-02 -1.11946523e+00
-6.26984686e-02 -7.85939872e-01 7.55515158e-01 -4.36620831e-01
-1.19607091e+00 6.50573552e-01 8.77911896e-02 -1.12267542e+00
-3.64181042e-01 -4.43840623e-01 -4.34211865e-02 7.11372912e-01
-1.51375985e+00 -9.08564925e-01 -2.50806600e-01 4.72408921e-01
2.61231720e-01 2.61141747e-01 1.00064588e+00 8.02091360e-01
-2.63801724e-01 6.13910258e-01 9.58021283e-02 2.61491835e-01
1.18282568e+00 -1.47994733e+00 3.77964586e-01 6.97204471e-01
3.76561016e-01 1.17470634e+00 9.53127265e-01 -8.66435349e-01
-9.57102716e-01 -9.74195480e-01 1.47884035e+00 -7.35655069e-01
8.17008436e-01 -3.57500851e-01 -1.09329736e+00 4.80517060e-01
2.78683662e-01 1.36086214e-02 1.03785586e+00 4.49490875e-01
-1.85073644e-01 2.82519847e-01 -1.01768625e+00 6.12202346e-01
1.03958380e+00 -3.22139800e-01 -7.96024084e-01 5.68174660e-01
4.90645856e-01 -5.58690786e-01 -9.85504925e-01 2.80604392e-01
4.50124532e-01 -4.38941002e-01 8.71603966e-01 -5.16698956e-01
-1.30956173e-01 -7.44242668e-01 1.60683066e-01 -1.13485336e+00
-2.40364581e-01 -6.92416430e-01 -1.53782815e-02 1.34984100e+00
9.23323035e-01 -5.54732203e-01 5.86001754e-01 8.00252318e-01
1.62518211e-02 -5.37787378e-01 -6.00602388e-01 -7.34656751e-01
-2.45607659e-01 -3.74833941e-01 5.06922901e-01 1.34685779e+00
5.68452597e-01 6.35173082e-01 -8.07186812e-02 3.77704352e-01
1.02345623e-01 -1.26330063e-01 6.29576623e-01 -1.56994581e+00
-4.05173481e-01 -1.79979041e-01 -1.38458535e-01 -7.48633444e-01
4.63565327e-02 -9.53072727e-01 4.19573747e-02 -1.69723368e+00
2.06793994e-01 -8.56970847e-01 -4.06414717e-01 9.38323975e-01
-5.68902910e-01 2.71230847e-01 -1.00315899e-01 4.31714624e-01
-1.09633100e+00 4.48322371e-02 6.22254491e-01 1.48937330e-01
-3.67252111e-01 -1.18491888e-01 -8.90076995e-01 7.70742774e-01
6.57284617e-01 -7.13607907e-01 -3.78927529e-01 -2.40007848e-01
7.57158995e-01 -2.38184512e-01 -2.18579471e-02 -8.21497500e-01
4.29296553e-01 4.19006348e-02 1.08479731e-01 -1.56435266e-01
-3.37960944e-02 -8.57877374e-01 2.62390584e-01 2.47851178e-01
-3.47347885e-01 -7.11556524e-02 2.69678265e-01 4.03874308e-01
-6.61011040e-03 -7.62484729e-01 3.25987011e-01 -2.24974364e-01
-8.59279692e-01 -6.22294843e-03 -1.97482705e-01 2.74435192e-01
9.17650878e-01 8.72924924e-02 -3.80903065e-01 -2.81808704e-01
-8.85326207e-01 2.75237948e-01 5.08702815e-01 2.65029192e-01
1.59963574e-02 -9.85978246e-01 -6.12591743e-01 -5.14361978e-01
4.66605008e-01 -9.11003202e-02 -1.00883573e-01 8.06633532e-01
-4.15207326e-01 4.75923777e-01 2.10971832e-01 -1.79408237e-01
-1.41038764e+00 4.07651007e-01 2.02055573e-02 -6.06305420e-01
-5.76498747e-01 7.89012372e-01 -1.32965267e-01 -5.49290299e-01
1.60693944e-01 -3.30005437e-02 -8.14482093e-01 3.11735153e-01
7.16159403e-01 2.05417410e-01 4.68838453e-01 -3.97353768e-01
-4.59840328e-01 1.66808575e-01 -2.35807106e-01 -2.18681067e-01
1.37278032e+00 -1.54261678e-01 -1.31271258e-02 2.64400154e-01
4.70351309e-01 5.12023449e-01 -4.61050510e-01 -4.52572674e-01
8.21470976e-01 -2.38257065e-01 9.93249007e-03 -1.07838285e+00
-5.16725063e-01 3.41285735e-01 3.42803895e-01 3.33229691e-01
9.63020384e-01 1.35072008e-01 5.96501648e-01 7.55406201e-01
4.30655748e-01 -1.27808189e+00 -2.98970670e-01 5.68033874e-01
5.57341337e-01 -1.22734523e+00 1.92659229e-01 -6.05313778e-01
-7.53275573e-01 1.09811521e+00 3.94139051e-01 3.96624088e-01
5.94467282e-01 1.90373614e-01 1.48888230e-01 -3.41939658e-01
-7.35247791e-01 -4.64221865e-01 4.17269319e-01 4.40629333e-01
1.02884150e+00 -1.59657821e-01 -8.59616160e-01 8.90701890e-01
-8.95697623e-02 2.28633270e-01 4.55618143e-01 1.08743954e+00
-4.27979052e-01 -1.60381937e+00 -1.77224353e-01 3.47328603e-01
-9.86628354e-01 -4.66830581e-01 -3.46765608e-01 8.17574739e-01
1.28905311e-01 9.25365031e-01 -3.77665311e-01 -1.11648917e-01
5.10370433e-01 2.55247831e-01 4.02320400e-02 -9.08333838e-01
-7.53765762e-01 4.76197675e-02 7.12919354e-01 -3.26425016e-01
-7.02423990e-01 -7.36781657e-01 -1.40888393e+00 -2.05600888e-01
-8.57993424e-01 7.75868118e-01 6.02079272e-01 9.10647929e-01
4.72236574e-01 1.70935765e-01 -3.14638168e-02 -3.61731946e-01
-7.40096867e-02 -1.08148861e+00 -2.38710761e-01 4.05302227e-01
-2.13693127e-01 -7.61133134e-01 2.71674395e-02 2.68573284e-01] | [9.451077461242676, 8.963886260986328] |
1eda0306-bb97-4978-9228-fea1209e06b4 | assessing-grammatical-correctness-in-language | null | null | https://aclanthology.org/2021.bea-1.15 | https://aclanthology.org/2021.bea-1.15.pdf | Assessing Grammatical Correctness in Language Learning | We present experiments on assessing the grammatical correctness of learners’ answers in a language-learning System (references to the System, and the links to the released data and code are withheld for anonymity). In particular, we explore the problem of detecting alternative-correct answers: when more than one inflected form of a lemma fits syntactically and semantically in a given context. We approach the problem with the methods for grammatical error detection (GED), since we hypothesize that models for detecting grammatical mistakes can assess the correctness of potential alternative answers in a learning setting. Due to the paucity of training data, we explore the ability of pre-trained BERT to detect grammatical errors and then fine-tune it using synthetic training data. In this work, we focus on errors in inflection. Our experiments show a. that pre-trained BERT performs worse at detecting grammatical irregularities for Russian than for English; b. that fine-tuned BERT yields promising results on assessing the correctness of grammatical exercises; and c. establish a new benchmark for Russian. To further investigate its performance, we compare fine-tuned BERT with one of the state-of-the-art models for GED (Bell et al., 2019) on our dataset and RULEC-GEC (Rozovskaya and Roth, 2019). We release the manually annotated learner dataset, used for testing, for general use. | ['Roman Yangarber', 'Anisia Katinskaia'] | null | null | null | null | eacl-bea-2021-4 | ['grammatical-error-detection'] | ['natural-language-processing'] | [-3.57478261e-01 3.43827516e-01 3.81923258e-01 -3.47251892e-01
-9.89017367e-01 -9.92599249e-01 1.50532514e-01 6.85615480e-01
-6.47738278e-01 9.55596685e-01 6.72625378e-02 -9.75397825e-01
2.66288016e-02 -8.76181841e-01 -9.07340646e-01 3.56550403e-02
1.21130005e-01 3.71463835e-01 3.04097325e-01 -4.28045869e-01
3.97926688e-01 2.64372051e-01 -1.62825608e+00 4.12312001e-01
1.51890039e+00 3.22894812e-01 -3.23190875e-02 8.84299517e-01
-3.14606458e-01 7.23990858e-01 -1.00444412e+00 -1.06699598e+00
-1.04420088e-01 -5.14699817e-01 -1.21230602e+00 -4.78769243e-01
9.10284400e-01 -7.06961304e-02 2.47198150e-01 1.26617038e+00
4.35674191e-01 -3.75258699e-02 6.50989950e-01 -9.42875624e-01
-7.95992911e-01 9.91247296e-01 1.16180718e-01 2.79574931e-01
8.02479267e-01 6.83214068e-02 1.08180153e+00 -1.00090909e+00
5.38166761e-01 1.12918079e+00 9.42026973e-01 8.83606672e-01
-1.11128008e+00 -5.34895241e-01 1.00996345e-01 1.54195711e-01
-1.15193605e+00 -3.90818745e-01 3.86728466e-01 -4.48441595e-01
8.69373441e-01 4.25507605e-01 4.69743013e-01 1.00995147e+00
-1.52932815e-02 7.78158069e-01 1.24180961e+00 -8.72253656e-01
5.09123988e-02 3.18218052e-01 5.60431719e-01 1.12265968e+00
4.38423604e-01 1.52621150e-01 -4.99016404e-01 1.04984678e-01
1.13667749e-01 -8.40206385e-01 -3.77107561e-01 1.97468832e-01
-7.46197879e-01 8.51044893e-01 3.07769310e-02 4.30157006e-01
3.25382918e-01 -8.55516642e-02 3.79355043e-01 1.07592225e+00
3.60903412e-01 7.72722542e-01 -7.49196172e-01 -3.95861059e-01
-6.77462697e-01 4.92665350e-01 1.06179523e+00 1.00380254e+00
4.15911973e-01 4.19391319e-02 5.08857183e-02 8.26269567e-01
4.83999044e-01 1.57516107e-01 8.14622462e-01 -6.23056293e-01
8.19239199e-01 6.79952085e-01 -1.46774068e-01 -6.05298936e-01
-4.17482883e-01 -3.29555869e-01 -2.26385623e-01 1.59068227e-01
1.14976585e+00 -7.51724094e-02 -4.61895138e-01 1.84909463e+00
1.99370995e-01 -1.55789390e-01 3.48488480e-01 3.19354385e-01
1.42328596e+00 4.18693870e-01 2.63742477e-01 -9.22729075e-02
1.10724807e+00 -5.49826205e-01 -5.65679014e-01 -5.65804495e-03
1.62158000e+00 -7.92300582e-01 1.68021274e+00 5.17210543e-01
-1.28482449e+00 -5.27585924e-01 -8.00574899e-01 -1.93508282e-01
-5.75886846e-01 1.34033367e-01 2.17496783e-01 1.11609495e+00
-1.00912869e+00 6.04164064e-01 -4.56302822e-01 -2.24827543e-01
-8.73124674e-02 8.57407600e-02 -4.40200806e-01 -1.54690385e-01
-1.32088768e+00 9.37557817e-01 3.62708449e-01 -2.14622483e-01
-4.06396419e-01 -6.06783152e-01 -1.22949243e+00 -7.80610293e-02
1.21691942e-01 -2.95743316e-01 1.47165203e+00 -8.38764548e-01
-1.34680700e+00 1.34965670e+00 -2.13622227e-02 -2.70877659e-01
7.27655649e-01 -1.27238244e-01 -4.01315719e-01 -3.13053280e-01
9.19706076e-02 3.49151820e-01 1.84165731e-01 -1.03725553e+00
-5.77234387e-01 -3.65772903e-01 2.02851683e-01 -9.88841802e-02
-1.64105713e-01 3.95825177e-01 1.67199269e-01 -7.75965393e-01
1.37954935e-01 -7.34255433e-01 1.91145018e-01 -6.13155782e-01
-2.75780737e-01 -8.93642485e-01 1.26226217e-01 -8.00983846e-01
1.76275980e+00 -1.96396410e+00 -2.44564950e-01 2.38060430e-01
8.59173611e-02 5.03268003e-01 -2.28996307e-01 2.34523699e-01
-3.11214715e-01 6.41123533e-01 -2.52242386e-01 -4.65308338e-01
2.29214504e-01 2.79690802e-01 -2.60974526e-01 2.42067814e-01
-2.44595073e-02 8.74472678e-01 -9.34897542e-01 -4.83817518e-01
-1.42285168e-01 -1.82907596e-01 -7.56298304e-01 2.56677240e-01
-5.76314442e-02 1.38298899e-01 -3.56048271e-02 6.09497011e-01
4.55608398e-01 3.84351313e-01 -2.31744777e-02 4.74643886e-01
-2.72391677e-01 9.72924113e-01 -1.50516093e+00 1.21347296e+00
-6.26641393e-01 3.81999463e-01 -2.50621229e-01 -6.76241934e-01
1.08291066e+00 3.30651939e-01 -3.68928939e-01 -5.87888479e-01
-6.27229586e-02 7.42488801e-01 2.26231039e-01 -5.60853541e-01
6.79396033e-01 2.98231281e-03 -3.08132917e-01 5.16418695e-01
3.42443377e-01 -3.29111814e-01 6.33111715e-01 2.65684694e-01
1.13604784e+00 4.96836677e-02 3.18509459e-01 -4.44481701e-01
7.66732872e-01 -1.26460195e-01 4.26727265e-01 8.40449691e-01
-2.04908133e-01 3.94210428e-01 5.19419670e-01 -2.97464728e-01
-5.37421286e-01 -1.02471364e+00 -2.35531121e-01 1.36776233e+00
-4.69135195e-01 -8.04865837e-01 -9.83190238e-01 -1.24354088e+00
4.51123118e-02 1.20677555e+00 -4.65516478e-01 -8.71792808e-02
-6.97419822e-01 -3.26955557e-01 8.70311379e-01 3.53639036e-01
1.68259293e-01 -1.31595540e+00 -3.87600422e-01 2.48972788e-01
-1.49060979e-01 -6.84434593e-01 -2.52666920e-01 3.72128487e-01
-6.86170936e-01 -1.37167060e+00 -4.87652607e-02 -1.02745438e+00
7.83506036e-01 -3.69274169e-01 1.68563020e+00 8.63859892e-01
9.95727926e-02 4.53603715e-01 -5.12040615e-01 -5.54460943e-01
-1.02372253e+00 1.43094867e-01 -4.87190261e-02 -7.93476939e-01
5.64770639e-01 -2.12025911e-01 1.43652946e-01 1.41763270e-01
-6.96933329e-01 -3.37832659e-01 -4.58303876e-02 7.35263228e-01
1.99915051e-01 -7.13554695e-02 7.21325636e-01 -1.43298149e+00
9.02057886e-01 -2.43383959e-01 -7.46829331e-01 3.83238375e-01
-6.59605682e-01 2.51287431e-01 9.16328788e-01 -3.34157534e-02
-7.07313716e-01 -2.85365045e-01 -7.69188762e-01 2.94214278e-01
-4.14135218e-01 6.31320179e-01 -3.33356172e-01 -2.68814683e-01
8.75976384e-01 -1.52590916e-01 -4.36729908e-01 -6.02845132e-01
2.09866300e-01 6.54183269e-01 5.47986090e-01 -1.19506645e+00
6.84843242e-01 -6.76180899e-01 -1.93813533e-01 -4.26595181e-01
-1.06690347e+00 -1.97677150e-01 -6.75269604e-01 -2.36365810e-01
3.64506453e-01 -8.51957738e-01 -6.61817729e-01 3.24838221e-01
-1.36051524e+00 -4.78710622e-01 -3.61345232e-01 3.40280563e-01
-5.10563016e-01 3.58113855e-01 -8.36694717e-01 -7.07133770e-01
-3.78742293e-02 -1.05214715e+00 5.80367744e-01 3.29285599e-02
-5.20678580e-01 -1.43539059e+00 3.22194695e-01 3.91786605e-01
4.67699543e-02 -3.31392810e-02 1.30643582e+00 -1.12521851e+00
-1.74119715e-02 1.58202037e-01 2.91171581e-01 4.75240111e-01
-3.19805950e-01 2.16583118e-01 -8.94923210e-01 -1.06498994e-01
-1.17430829e-01 -6.65458977e-01 7.82324493e-01 -1.44365266e-01
1.17661929e+00 -5.52753329e-01 3.88257682e-01 4.46261019e-01
1.12443292e+00 -2.61094183e-01 4.87720400e-01 5.71159661e-01
4.99307781e-01 1.02851415e+00 5.75537741e-01 -8.45889896e-02
6.73251212e-01 3.33286762e-01 2.18505010e-01 3.95091057e-01
-1.80341959e-01 -6.27452970e-01 6.09265924e-01 1.17481220e+00
1.73612714e-01 -4.15747941e-01 -1.11310005e+00 7.28017449e-01
-1.51336396e+00 -6.93437397e-01 -8.42478633e-01 2.38585854e+00
1.35952735e+00 2.66590625e-01 1.87238768e-01 4.62319613e-01
5.47001958e-01 -3.73614579e-01 3.21185380e-01 -1.15530121e+00
-1.84873268e-02 6.79176629e-01 1.11151323e-01 9.47713673e-01
-9.06789303e-01 1.17335212e+00 6.38915777e+00 6.80860996e-01
-7.50022411e-01 1.86166447e-02 5.01858652e-01 4.83427644e-01
-7.03500509e-01 -2.72591084e-01 -1.15445054e+00 4.72168624e-01
1.26315176e+00 -2.63497513e-02 2.32718930e-01 6.61520660e-01
-7.96208084e-02 -9.82522517e-02 -1.35248625e+00 6.14777029e-01
8.59544724e-02 -7.51716316e-01 -1.35611534e-01 -2.91840523e-01
6.49640560e-01 -3.25640738e-01 -3.15193571e-02 8.70456040e-01
7.42612422e-01 -1.19707274e+00 1.02652705e+00 1.86847523e-01
6.38856828e-01 -9.41024542e-01 8.57178748e-01 5.40933490e-01
-7.64605284e-01 9.75499004e-02 -4.80235308e-01 -3.09164762e-01
-4.69946802e-01 3.86558235e-01 -7.44015336e-01 3.75472724e-01
6.48072124e-01 3.59705716e-01 -1.31013083e+00 1.06830239e+00
-1.07282364e+00 1.15074933e+00 -3.18024933e-01 -3.06612611e-01
8.10909122e-02 -4.47403938e-02 3.40668917e-01 1.32581019e+00
4.63961124e-01 9.19151818e-04 2.80920099e-02 8.93631816e-01
-2.49954581e-01 5.47194004e-01 -5.93527198e-01 2.41088495e-01
5.63469946e-01 1.01471531e+00 -2.69439101e-01 -1.39755517e-01
-3.86615813e-01 6.53609395e-01 8.88440371e-01 1.07292403e-02
-4.20014024e-01 -4.54056591e-01 2.63093621e-01 1.02404684e-01
3.82266641e-02 -4.33730781e-02 -5.75151384e-01 -1.07345843e+00
5.76827005e-02 -1.29882872e+00 7.86515474e-01 -6.15211010e-01
-1.31662762e+00 4.61427301e-01 -2.49706805e-01 -8.86691272e-01
-3.38729411e-01 -9.70204353e-01 -8.32482874e-01 1.07301974e+00
-1.45125532e+00 -7.50794351e-01 -1.31150186e-01 5.04927635e-01
4.76311594e-01 -1.74775913e-01 1.02995121e+00 8.88626128e-02
-6.36772513e-01 1.21340644e+00 -8.04467425e-02 2.77247757e-01
9.17664647e-01 -1.99110448e+00 6.14223719e-01 1.05793858e+00
5.60473859e-01 6.45787239e-01 7.27427483e-01 -6.10864639e-01
-8.27349246e-01 -1.12049687e+00 1.72042608e+00 -9.00008559e-01
7.59274542e-01 -2.82297164e-01 -1.27766204e+00 7.60389507e-01
-4.27400842e-02 -1.12299107e-01 8.47006917e-01 4.39878970e-01
-3.96394879e-01 3.02712560e-01 -1.16948879e+00 4.76421714e-01
1.06255639e+00 -4.86054987e-01 -1.09497869e+00 3.06981891e-01
4.91673142e-01 -7.30417013e-01 -7.13287413e-01 2.32236877e-01
5.00940718e-02 -1.08889675e+00 3.43189329e-01 -1.05011201e+00
5.61644852e-01 -1.90283611e-01 3.18960845e-02 -1.71288788e+00
-3.29357311e-02 -5.67192197e-01 4.73140031e-02 1.62512541e+00
7.64321446e-01 -4.91821676e-01 6.45257533e-01 5.95548987e-01
-4.94109631e-01 -6.43316150e-01 -8.27847779e-01 -7.41377294e-01
7.92211950e-01 -6.63431346e-01 4.85864460e-01 1.10344827e+00
1.93126604e-01 2.53369153e-01 2.92777419e-01 1.33990765e-01
2.43021354e-01 -1.67924851e-01 8.00162375e-01 -1.28284144e+00
-3.38390052e-01 -4.39726055e-01 -2.38556072e-01 -8.20365846e-01
5.75974941e-01 -1.23920631e+00 6.16281852e-02 -1.18591309e+00
-3.32724780e-01 -6.36742055e-01 -3.82495634e-02 4.88470793e-01
-6.40117764e-01 2.36600235e-01 1.99000210e-01 -3.30552965e-01
-6.39699519e-01 2.19687194e-01 8.16691279e-01 2.33589932e-01
-2.94947829e-02 2.01956868e-01 -7.83907712e-01 8.66103172e-01
7.38994956e-01 -7.15551019e-01 4.61421534e-03 -4.56339240e-01
7.29256392e-01 -2.48515248e-01 2.74135590e-01 -8.79561245e-01
8.29083323e-02 1.48599535e-01 1.01972483e-01 -2.71901667e-01
-6.86992884e-01 -5.82660079e-01 -4.43649858e-01 4.56014097e-01
-6.04847729e-01 5.70066631e-01 2.92979240e-01 4.77503240e-02
-3.09370905e-01 -1.16789448e+00 6.46139085e-01 -2.20271572e-01
-4.42113191e-01 -1.53374121e-01 -5.08987069e-01 6.48765564e-01
6.75225675e-01 -1.54009774e-01 -4.48522747e-01 -2.26893112e-01
-5.79311073e-01 3.08041304e-01 4.13670540e-01 2.07527488e-01
4.47281420e-01 -1.06378436e+00 -9.39233363e-01 3.60042572e-01
3.27094555e-01 -8.40122700e-02 -2.19875827e-01 5.70684314e-01
-7.79556990e-01 2.82552630e-01 1.57062486e-01 -2.92169273e-01
-1.70392358e+00 2.12986216e-01 5.83459437e-01 -5.07385671e-01
-1.98315471e-01 1.16183293e+00 -2.24073574e-01 -1.18790579e+00
3.50939572e-01 -6.99015498e-01 -4.40467179e-01 -1.05505630e-01
4.92090762e-01 2.88935423e-01 5.81855476e-01 -4.38878655e-01
-9.61209685e-02 2.84617931e-01 -2.39964332e-02 1.74096242e-01
1.15060711e+00 3.18201557e-02 -1.15800709e-01 5.61021745e-01
8.47427964e-01 7.39862323e-01 -4.15471911e-01 -1.54604718e-01
3.90431523e-01 -3.30271393e-01 -2.76564956e-01 -1.14786565e+00
-5.73043764e-01 8.86712134e-01 2.45701924e-01 4.39369738e-01
7.20205724e-01 -7.92506188e-02 4.03444082e-01 5.12215793e-01
3.45875144e-01 -1.06230259e+00 -2.54499137e-01 1.18977988e+00
7.05679297e-01 -1.22843170e+00 -2.91362971e-01 -5.52242041e-01
-2.41442904e-01 1.44673502e+00 9.21504378e-01 1.85681149e-01
4.03802484e-01 1.90275356e-01 2.92760044e-01 -3.27820331e-02
-8.15767705e-01 -1.38316438e-01 3.54411721e-01 5.23241878e-01
1.05952930e+00 8.73245373e-02 -7.96317637e-01 9.45130825e-01
-1.10596716e+00 -4.60150331e-01 8.51140201e-01 7.41206467e-01
-5.18103063e-01 -1.27282798e+00 -4.65102822e-01 2.25945815e-01
-7.11535335e-01 -3.60593885e-01 -8.04514349e-01 1.00416481e+00
2.53156036e-01 1.09595168e+00 -1.42928272e-01 -1.19774148e-01
6.71664000e-01 4.88666385e-01 6.88643873e-01 -1.05907536e+00
-1.35997534e+00 -6.43679678e-01 5.20577371e-01 -3.18649620e-01
5.47391251e-02 -7.00527191e-01 -1.14455914e+00 -5.10332525e-01
-3.49851847e-01 6.26690507e-01 4.83022094e-01 1.10477245e+00
-1.40656725e-01 4.05180395e-01 2.91679114e-01 1.83469638e-01
-9.16691661e-01 -1.08005500e+00 -4.70993698e-01 4.75120276e-01
9.44368392e-02 -2.94997990e-01 -7.59806097e-01 -1.55566067e-01] | [10.958133697509766, 10.453315734863281] |
25b542b4-6373-4bee-85a2-90894fd44c24 | tryondiffusion-a-tale-of-two-unets-1 | 2306.08276 | null | https://arxiv.org/abs/2306.08276v1 | https://arxiv.org/pdf/2306.08276v1.pdf | TryOnDiffusion: A Tale of Two UNets | Given two images depicting a person and a garment worn by another person, our goal is to generate a visualization of how the garment might look on the input person. A key challenge is to synthesize a photorealistic detail-preserving visualization of the garment, while warping the garment to accommodate a significant body pose and shape change across the subjects. Previous methods either focus on garment detail preservation without effective pose and shape variation, or allow try-on with the desired shape and pose but lack garment details. In this paper, we propose a diffusion-based architecture that unifies two UNets (referred to as Parallel-UNet), which allows us to preserve garment details and warp the garment for significant pose and body change in a single network. The key ideas behind Parallel-UNet include: 1) garment is warped implicitly via a cross attention mechanism, 2) garment warp and person blend happen as part of a unified process as opposed to a sequence of two separate tasks. Experimental results indicate that TryOnDiffusion achieves state-of-the-art performance both qualitatively and quantitatively. | ['Ira Kemelmacher-Shlizerman', 'Mohammad Norouzi', 'Chitwan Saharia', 'William Chan', 'Fitsum Reda', 'Tyler Zhu', 'Dawei Yang', 'Luyang Zhu'] | 2023-06-14 | tryondiffusion-a-tale-of-two-unets | http://openaccess.thecvf.com//content/CVPR2023/html/Zhu_TryOnDiffusion_A_Tale_of_Two_UNets_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Zhu_TryOnDiffusion_A_Tale_of_Two_UNets_CVPR_2023_paper.pdf | cvpr-2023-1 | ['virtual-try-on'] | ['computer-vision'] | [ 1.10522516e-01 1.35148495e-01 2.50375122e-01 -1.92777321e-01
1.02875948e-01 -6.45026386e-01 7.23906398e-01 -4.81736541e-01
2.77306568e-02 3.98734808e-01 5.49880981e-01 1.40379995e-01
1.29915595e-01 -4.36728984e-01 -5.88054895e-01 -4.04675633e-01
-2.53873004e-04 3.41425955e-01 2.14731768e-02 -3.75724465e-01
-2.03074545e-01 6.78755343e-01 -1.17878163e+00 1.73100576e-01
5.07720530e-01 5.37830949e-01 -2.46777147e-01 7.48921335e-01
5.91497362e-01 3.44705343e-01 -9.01829541e-01 -6.07257783e-01
5.86912155e-01 -7.62059927e-01 -6.06788337e-01 5.68148255e-01
1.19298112e+00 -3.52372825e-01 -6.14517331e-01 8.31334710e-01
6.15632653e-01 4.11716223e-01 5.25209665e-01 -1.29579544e+00
-1.06229329e+00 4.24613506e-01 -9.17942882e-01 -4.11074460e-02
5.78340113e-01 5.33046365e-01 5.42992532e-01 -5.77039838e-01
8.05671811e-01 1.69452679e+00 4.78790879e-01 8.72151434e-01
-1.63504243e+00 -5.90665340e-01 5.58257163e-01 -2.62401581e-01
-1.35303307e+00 -2.43882924e-01 1.04844415e+00 -3.94193679e-01
6.04331911e-01 6.69703543e-01 1.39578521e+00 1.40204787e+00
3.73538911e-01 5.23062825e-01 1.09999657e+00 -4.63447981e-02
2.05535209e-03 -2.52294600e-01 -1.97795630e-01 6.23606086e-01
2.42353126e-01 1.55738488e-01 -6.54278994e-01 9.62306932e-02
1.47099459e+00 2.56304353e-01 -5.92682838e-01 -6.69407487e-01
-1.54701650e+00 4.46686327e-01 8.05376530e-01 2.17752382e-01
-4.31893349e-01 3.06096643e-01 3.09893787e-02 2.42254987e-01
5.21062672e-01 7.08623528e-01 2.76565760e-01 1.97484523e-01
-7.72474766e-01 6.30684376e-01 5.74326634e-01 8.48723590e-01
1.00319445e-01 2.34561175e-01 -4.91913676e-01 3.38080078e-01
3.16802204e-01 4.34170455e-01 9.93436202e-02 -7.67115653e-01
3.38706225e-01 7.02421725e-01 1.85340345e-01 -1.29624224e+00
-4.74917620e-01 -3.47829342e-01 -9.94379818e-01 6.98252499e-01
4.05210167e-01 -3.39825481e-01 -1.26895988e+00 1.86243498e+00
5.32730401e-01 9.25396457e-02 -2.83339918e-01 1.47496092e+00
7.70244837e-01 4.04109836e-01 1.81973934e-01 2.66721994e-01
1.62675416e+00 -1.08735192e+00 -8.93997192e-01 -6.30960703e-01
-2.71596789e-01 -5.88267505e-01 1.35346186e+00 1.87468737e-01
-1.37715721e+00 -7.37980247e-01 -1.24355745e+00 -4.11280990e-01
-1.07682407e-01 -1.70862570e-01 3.15936476e-01 4.63069677e-01
-1.01892459e+00 7.56595731e-01 -8.84827971e-01 -5.38218498e-01
2.21732721e-01 1.36471733e-01 -5.50778747e-01 1.33294091e-01
-8.62673402e-01 1.00890207e+00 2.36314144e-02 3.85793626e-01
-7.50804126e-01 -7.26790965e-01 -9.45276380e-01 -6.69079572e-02
2.92261690e-01 -1.36709261e+00 9.33613598e-01 -1.10261548e+00
-1.56912255e+00 1.08274055e+00 2.75942236e-01 -7.84012824e-02
1.12261569e+00 -4.53382581e-01 -3.43181759e-01 4.96273413e-02
-3.22896659e-01 7.11580634e-01 1.13739002e+00 -1.57722700e+00
1.00497171e-01 -6.57090545e-01 1.87037420e-02 7.16262937e-01
-6.35503307e-02 -9.70861018e-02 -6.58879340e-01 -1.23345113e+00
7.00881481e-02 -1.03594697e+00 -3.67012918e-02 5.31136632e-01
-7.81636536e-01 2.30234027e-01 1.12080657e+00 -9.93967533e-01
1.10103250e+00 -2.05688047e+00 7.83890426e-01 2.90230274e-01
7.29748964e-01 1.49363458e-01 -1.08714022e-01 4.79711294e-01
-4.35395867e-01 4.49047685e-02 -1.46972463e-01 -7.92951524e-01
-5.20780124e-02 8.82675946e-02 -1.27863616e-01 7.00622559e-01
3.01698595e-02 1.10220397e+00 -8.44216645e-01 -1.18339561e-01
2.55386889e-01 9.77676392e-01 -4.44924504e-01 3.16518903e-01
-5.40102134e-03 8.71821284e-01 -1.20166272e-01 6.11190021e-01
5.91883838e-01 -6.68638572e-02 2.10356921e-01 -6.81611478e-01
3.89644057e-02 -1.62260920e-01 -9.76628780e-01 1.86268866e+00
1.36418678e-02 5.26334345e-01 1.39480591e-01 -1.79546431e-01
8.93387020e-01 1.69272184e-01 1.28954649e-01 -4.49768007e-01
3.97831261e-01 -3.18910122e-01 5.72573543e-02 -3.52969497e-01
4.60940957e-01 -3.02433282e-01 -4.58867885e-02 4.03520375e-01
-2.74741232e-01 -3.08737457e-01 -2.43350387e-01 -2.95180455e-02
7.10218310e-01 3.43164116e-01 9.77380872e-02 -2.15049878e-01
-9.02136043e-02 -4.48759407e-01 7.84425735e-02 3.78441125e-01
-2.08713263e-01 1.09522343e+00 2.57457197e-01 -6.37637615e-01
-1.14465654e+00 -1.27076948e+00 5.21652341e-01 6.97424948e-01
5.31521738e-01 -2.66664505e-01 -8.08688223e-01 -5.63014388e-01
-1.46734426e-02 6.84859455e-01 -1.04319644e+00 -3.52179557e-01
-7.88137674e-01 -3.62872988e-01 3.04449946e-01 5.00664473e-01
8.06468189e-01 -1.02617288e+00 -1.07491469e+00 -2.69878268e-01
-9.14091095e-02 -7.66339481e-01 -1.27185142e+00 -5.96784949e-01
-5.84416866e-01 -7.71487832e-01 -1.28653657e+00 -5.89068174e-01
8.65009248e-01 2.32349530e-01 1.06796372e+00 2.62619779e-02
-4.29110199e-01 3.06198657e-01 -5.29041775e-02 -7.25554600e-02
-7.68497810e-02 -3.49531770e-01 2.17094839e-01 3.81923586e-01
-2.33736634e-01 -9.34083462e-01 -1.06311131e+00 3.41299355e-01
-6.91166043e-01 5.91848373e-01 4.70032126e-01 3.81598145e-01
3.92834544e-01 -2.24175051e-01 -1.48658916e-01 -4.68141764e-01
9.40427184e-01 1.70125570e-02 7.77185410e-02 2.28387550e-01
-1.90856829e-01 -3.39848548e-01 6.55139610e-02 -9.62946177e-01
-8.70992541e-01 -1.67589322e-01 1.88155085e-01 -9.81494069e-01
1.30624816e-01 -1.99094996e-01 -3.05466741e-01 -1.07249990e-01
6.35763288e-01 -1.22868873e-01 3.21844310e-01 -4.74042594e-01
6.65073991e-01 3.86510743e-03 8.96412194e-01 -4.18788105e-01
1.09929693e+00 6.25163674e-01 -5.87360747e-02 -5.34469426e-01
-4.71417218e-01 2.23987967e-01 -7.98538089e-01 -5.93019545e-01
1.14942002e+00 -5.90110958e-01 -1.02742290e+00 4.95481819e-01
-1.00495744e+00 -5.44689596e-01 -6.36103332e-01 1.98309928e-01
-2.93181658e-01 9.94128212e-02 -4.16472763e-01 -5.92480183e-01
-4.58072633e-01 -7.93465137e-01 1.12381685e+00 4.26663786e-01
-8.36096346e-01 -8.90690088e-01 7.66726583e-02 2.45344639e-01
3.32458228e-01 9.42042708e-01 6.00689471e-01 -1.66500673e-01
-2.59911984e-01 -2.45421559e-01 -1.14409946e-01 1.78021826e-02
4.86427516e-01 -1.56360701e-01 -6.95752501e-01 -6.62058532e-01
-1.78334221e-01 -4.12910478e-03 3.85058016e-01 2.05810130e-01
8.59127820e-01 -6.26454830e-01 -2.62843966e-01 7.35291719e-01
8.97481799e-01 5.28244451e-02 6.95097983e-01 -5.30760251e-02
1.11487472e+00 8.28059375e-01 8.38296339e-02 -6.71436451e-03
2.39133105e-01 8.61261725e-01 3.61279666e-01 -6.73222840e-01
-6.69240057e-01 -7.85323322e-01 2.05042765e-01 4.33962315e-01
-5.19827425e-01 -2.13621154e-01 -6.52426302e-01 3.66235584e-01
-1.77465069e+00 -8.86512458e-01 2.36822993e-01 2.20655298e+00
6.54286921e-01 -9.87807959e-02 3.89550209e-01 -1.84525087e-01
6.34739697e-01 3.93874526e-01 -9.23127532e-01 -3.03648293e-01
1.17642201e-01 -1.88507020e-01 6.50761351e-02 4.52900380e-01
-7.90707052e-01 6.47572458e-01 6.65356970e+00 1.46055073e-01
-1.17666459e+00 -2.71617532e-01 6.84699297e-01 -4.72547948e-01
-2.13014305e-01 -4.60569799e-01 -3.01687062e-01 2.13928282e-01
1.89606234e-01 -9.77430642e-02 4.79934484e-01 3.28672469e-01
1.51513338e-01 3.26627493e-01 -1.29571998e+00 1.10970330e+00
3.95044833e-01 -1.04795921e+00 2.68111497e-01 -4.31767255e-02
5.43171346e-01 -7.19245017e-01 3.93805087e-01 -1.21823691e-01
2.08138302e-01 -9.92908597e-01 1.13785744e+00 7.05619752e-01
1.06656253e+00 -5.28348088e-01 8.59479681e-02 -4.35273871e-02
-1.10392463e+00 2.00314581e-01 3.30371886e-01 -1.28023639e-01
3.77600193e-01 -3.14216688e-02 -2.82353014e-01 4.30794418e-01
8.49357307e-01 4.56329882e-01 -4.72558320e-01 8.07065010e-01
-3.98247480e-01 -7.81489015e-02 -5.12964875e-02 2.40875542e-01
-1.19366020e-01 -2.60902107e-01 1.02641320e+00 1.01676846e+00
1.69167712e-01 1.40129134e-01 2.93181568e-01 1.29228771e+00
-9.02961716e-02 -2.54700303e-01 -5.69091856e-01 -6.97903410e-02
2.22457260e-01 1.16141975e+00 -6.41602457e-01 -2.77524501e-01
-9.62347351e-03 1.47587979e+00 1.90714449e-02 6.36838078e-01
-7.72947550e-01 -2.08187953e-01 9.54630673e-01 5.80700576e-01
-6.06482998e-02 -3.42554331e-01 -3.06076705e-01 -1.10256934e+00
1.08685615e-02 -9.71166372e-01 1.69549704e-01 -1.20787978e+00
-1.41477847e+00 8.60409081e-01 1.89191833e-01 -1.21181202e+00
-1.32963419e-01 -9.75600407e-02 -8.03538918e-01 1.09420443e+00
-7.92332053e-01 -1.72389162e+00 -6.29844189e-01 6.99360371e-01
5.66947162e-01 3.70203644e-01 5.74099600e-01 7.48987272e-02
-6.70559108e-01 6.04019582e-01 -9.05731857e-01 -6.14975579e-02
6.60350919e-01 -1.17426968e+00 8.56943905e-01 1.01031649e+00
1.08254537e-01 9.13286448e-01 1.04084933e+00 -9.63862896e-01
-1.35177982e+00 -1.03114736e+00 5.56431949e-01 -5.99943280e-01
1.95005178e-01 -6.67095125e-01 -7.78020203e-01 1.05714464e+00
6.15165234e-01 -8.47377256e-02 3.90885234e-01 -6.39816448e-02
-3.78395498e-01 7.91162029e-02 -1.25879061e+00 1.30585003e+00
1.32668900e+00 -2.73925394e-01 -7.34740734e-01 1.09527491e-01
7.47265935e-01 -8.57029438e-01 -6.19943023e-01 9.96198654e-02
8.50271106e-01 -7.96631277e-01 1.27237487e+00 -6.01142764e-01
4.29477274e-01 -5.86150229e-01 2.88147807e-01 -1.65513337e+00
-7.07993746e-01 -1.11277997e+00 -3.89268279e-01 9.98748779e-01
-7.36566186e-02 -5.97894371e-01 5.45791388e-01 1.05772972e+00
5.88229746e-02 -5.36070228e-01 -6.59281671e-01 -6.16779625e-01
-2.01301441e-01 2.27684900e-01 6.56642497e-01 1.07825303e+00
-5.60535342e-02 3.08141232e-01 -8.77110779e-01 2.86035210e-01
5.78620911e-01 1.43757612e-01 8.74580204e-01 -9.15339530e-01
-3.60791385e-01 -2.84740031e-01 -2.99400330e-01 -1.01669061e+00
-5.30369341e-01 -4.99693245e-01 -7.52610117e-02 -1.46274829e+00
1.49186388e-01 1.44513801e-01 2.20018208e-01 7.69779801e-01
-1.71093836e-01 6.36302054e-01 6.98256135e-01 2.23401427e-01
6.02940470e-02 6.44176841e-01 1.91913164e+00 -2.57102102e-02
-4.25702810e-01 -1.91172153e-01 -8.73249829e-01 7.51811504e-01
4.42590594e-01 5.91096953e-02 -6.69591546e-01 -6.52881861e-01
-4.96957488e-02 7.59463161e-02 9.01937842e-01 -9.09010649e-01
9.07705575e-02 -1.89416334e-01 8.74712586e-01 -1.48469806e-01
8.08284819e-01 -8.10479701e-01 6.47867382e-01 5.96830904e-01
-1.61240250e-01 4.66987222e-01 2.80048281e-01 5.45218468e-01
2.11200699e-01 8.64377558e-01 7.55673647e-01 -8.92677978e-02
-3.76673758e-01 4.00248498e-01 7.24872341e-03 -3.17975432e-01
1.13386059e+00 -4.17342573e-01 -2.14370832e-01 -5.59709847e-01
-1.02062583e+00 2.55628198e-01 6.70130908e-01 8.81548524e-01
7.15095460e-01 -1.48027146e+00 -7.33086288e-01 6.00591779e-01
-2.28770837e-01 -3.72964926e-02 5.06748319e-01 5.07134974e-01
-5.00505924e-01 -2.03005850e-01 -5.67221820e-01 -4.38257605e-01
-1.52442265e+00 4.49438214e-01 6.78669035e-01 -2.37456635e-02
-1.24876320e+00 9.36635494e-01 5.37227333e-01 -1.98190838e-01
1.26949936e-01 -2.24475265e-01 -6.20663501e-02 -1.37787029e-01
6.86010242e-01 2.39314571e-01 -4.89084393e-01 -6.83136404e-01
-2.26557255e-01 8.94207478e-01 3.88551466e-02 -2.15116262e-01
1.17017770e+00 -3.22406501e-01 2.03132257e-01 3.20733279e-01
7.09903777e-01 2.46318765e-02 -1.79313791e+00 2.73018319e-04
-9.25770283e-01 -8.75161648e-01 -1.79508135e-01 -1.15183699e+00
-1.33098722e+00 6.68890953e-01 6.25557542e-01 4.85270564e-03
1.16080129e+00 -3.11668590e-02 6.87277198e-01 -3.21815372e-01
1.55020282e-01 -5.68919301e-01 4.04983252e-01 -2.23603278e-01
1.84129083e+00 -6.32470548e-01 1.10385090e-01 -5.29994845e-01
-8.39611530e-01 6.56734109e-01 8.91036391e-01 -3.69421929e-01
3.20766836e-01 1.51914656e-01 -1.54357469e-02 -5.53265452e-01
-4.29401964e-01 2.13484734e-01 9.74862754e-01 7.49865949e-01
2.27739662e-01 1.47050679e-01 5.42528694e-03 4.40206319e-01
-4.93079394e-01 -2.26657704e-01 -1.75275151e-02 8.41087282e-01
7.40698650e-02 -6.99966431e-01 -4.96033967e-01 -1.96692254e-02
3.09238639e-02 3.23015720e-01 -8.88568342e-01 1.00204277e+00
2.17933193e-01 6.20626330e-01 1.83212832e-01 -4.19176161e-01
9.35241461e-01 -1.40715912e-01 7.97147274e-01 -5.93284428e-01
-8.13858986e-01 1.85319722e-01 1.39897000e-02 -7.82731950e-01
-1.96890995e-01 -3.37683320e-01 -8.59894574e-01 -6.34905815e-01
3.37974012e-01 -5.13085902e-01 3.47238421e-01 6.39758825e-01
3.77936214e-01 9.19221818e-01 1.50829583e-01 -1.50759649e+00
-9.80600864e-02 -9.40399289e-01 -6.67499006e-01 8.05617332e-01
5.02886832e-01 -6.00990772e-01 -1.27102748e-01 1.36291191e-01] | [11.907524108886719, -0.8383541703224182] |
1ced93d2-6d30-4f9d-a603-85c2ed4af95b | 3d-geometric-salient-patterns-analysis-on-3d | 1906.07645 | null | https://arxiv.org/abs/1906.07645v1 | https://arxiv.org/pdf/1906.07645v1.pdf | 3D Geometric salient patterns analysis on 3D meshes | Pattern analysis is a wide domain that has wide applicability in many fields. In fact, texture analysis is one of those fields, since the texture is defined as a set of repetitive or quasi-repetitive patterns. Despite its importance in analyzing 3D meshes, geometric texture analysis is less studied by geometry processing community. This paper presents a new efficient approach for geometric texture analysis on 3D triangular meshes. The proposed method is a scale-aware approach that takes as input a 3D mesh and a user-scale. It provides, as a result, a similarity-based clustering of texels in meaningful classes. Experimental results of the proposed algorithm are presented for both real-world and synthetic meshes within various textures. Furthermore, the efficiency of the proposed approach was experimentally demonstrated under mesh simplification and noise addition on the mesh surface. In this paper, we present a practical application for semantic annotation of 3D geometric salient texels. | ['Jean-Marie Favreau', 'Fakhri Torkhani', 'Alice Othmani'] | 2019-06-18 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 5.72903097e-01 -1.16446115e-01 3.00056964e-01 -2.19559088e-01
-2.23580137e-01 -4.84591872e-01 4.72012818e-01 6.52536452e-01
-2.27152817e-02 2.81846225e-01 -2.86198556e-01 7.40048662e-02
-5.23953259e-01 -1.19333458e+00 -4.15770143e-01 -4.64632064e-01
-1.12997349e-02 8.37598741e-01 6.22396231e-01 -1.92663312e-01
4.78883833e-01 1.01358342e+00 -2.04655623e+00 1.86981261e-01
5.74631512e-01 1.00723314e+00 3.37207764e-01 3.87033463e-01
-3.51488203e-01 1.02847517e-01 -3.27582866e-01 -1.62286490e-01
2.52493739e-01 -3.34586442e-01 -5.98258078e-01 5.34936488e-01
4.81525868e-01 5.07532060e-01 4.05293643e-01 1.28567541e+00
2.73465246e-01 1.80984333e-01 8.53151977e-01 -7.59685159e-01
4.82847467e-02 1.59941778e-01 -8.07817042e-01 -4.65397425e-02
2.97407329e-01 -4.56159920e-01 5.86917877e-01 -1.07402837e+00
1.10142481e+00 1.42038929e+00 6.12047553e-01 -2.27110967e-01
-1.32254887e+00 -2.52779245e-01 -1.72623426e-01 1.09824039e-01
-1.68727887e+00 -6.85852990e-02 1.10756481e+00 -5.81621051e-01
5.60324013e-01 6.14735007e-01 7.17926681e-01 3.39587778e-01
3.54069263e-01 2.49330774e-01 1.48352313e+00 -6.89850867e-01
3.81625295e-01 1.07407965e-01 -1.00581937e-01 6.07490361e-01
2.28861183e-01 -4.52084720e-01 -1.13837406e-01 -2.14686677e-01
1.12769675e+00 -2.01043010e-01 6.14138460e-03 -8.71997178e-01
-1.14796793e+00 2.62122571e-01 7.69819890e-04 6.97332025e-01
-4.12896246e-01 -7.49105886e-02 5.08807719e-01 3.42809349e-01
1.05897689e+00 1.53717294e-01 -1.21603444e-01 -1.10063471e-01
-9.49644327e-01 3.22050095e-01 6.74126029e-01 1.09621882e+00
8.71967673e-01 -3.68486382e-02 2.47118101e-01 1.03810263e+00
1.88295767e-02 6.42522812e-01 -1.49250284e-01 -6.53527319e-01
8.96630362e-02 9.13072824e-01 2.37862696e-03 -1.81589174e+00
-4.26445514e-01 -1.35037407e-01 -9.27993774e-01 4.54226613e-01
4.24621254e-01 6.10755503e-01 -7.21037686e-01 1.21110070e+00
8.25394332e-01 6.91605881e-02 -4.64201957e-01 8.04998994e-01
6.58833563e-01 2.39867583e-01 -1.64179757e-01 -1.79996490e-01
1.49650276e+00 -3.14136595e-01 -7.96365678e-01 5.46268761e-01
2.07502827e-01 -1.50540292e+00 9.02623057e-01 4.84748393e-01
-1.04369342e+00 -5.16257405e-01 -6.46596849e-01 1.54919952e-01
-3.81737202e-01 2.13257536e-01 2.12480396e-01 5.41734815e-01
-8.48254025e-01 6.64503276e-01 -6.13285482e-01 -1.06436467e+00
1.80243284e-01 2.55236208e-01 -3.32907021e-01 2.55571067e-01
-5.13246715e-01 8.37741017e-01 2.33939961e-01 -7.95620605e-02
-3.65031995e-02 -3.51161450e-01 -5.59466600e-01 -8.93899277e-02
5.17454565e-01 -4.98704910e-01 7.90726066e-01 -8.71190429e-01
-1.43206227e+00 1.50918996e+00 -1.93446279e-01 -5.38797118e-02
6.76284313e-01 9.55811813e-02 -5.31578243e-01 2.03040183e-01
1.74956068e-01 -6.65495321e-02 1.02646375e+00 -1.37094235e+00
-5.58210313e-01 -4.36744124e-01 -2.35709548e-01 2.90432334e-01
1.67086750e-01 -1.32106170e-01 -5.17042220e-01 -1.04815710e+00
6.37186646e-01 -7.04916894e-01 -8.94599184e-02 3.90684642e-02
-2.78870136e-01 -2.37334564e-01 9.40275431e-01 -3.51687670e-01
1.21199501e+00 -2.32232475e+00 2.24691436e-01 9.68781531e-01
4.37237956e-02 -2.58477956e-01 3.42452794e-01 6.82472169e-01
6.22502044e-02 1.29610315e-01 -2.99620092e-01 -1.27853513e-01
4.66658846e-02 8.66055414e-02 -1.28626779e-01 6.33274972e-01
-5.54073639e-02 1.76589951e-01 -7.58073628e-01 -8.48522961e-01
7.14648843e-01 2.97858149e-01 -2.84388095e-01 -2.36757219e-01
-2.53713191e-01 5.36779463e-01 -7.26562977e-01 8.30948234e-01
9.34013188e-01 2.87538692e-02 2.87351489e-01 -5.08655906e-01
-5.23181915e-01 -4.93307561e-01 -1.61959708e+00 1.52338827e+00
-2.64051735e-01 3.70317221e-01 2.75217623e-01 -9.22492981e-01
1.44217002e+00 3.10296148e-01 7.29195774e-01 -5.23216367e-01
4.55879211e-01 7.00983644e-01 -3.22465807e-01 -4.74334180e-01
8.76756907e-01 -1.47959083e-01 -6.74164370e-02 3.25000942e-01
-5.22670865e-01 -6.52749658e-01 2.53431439e-01 -3.41875821e-01
6.93969369e-01 3.21129680e-01 6.14148021e-01 -7.90775359e-01
7.69375503e-01 3.36760730e-01 3.03428054e-01 4.06222552e-01
3.17838490e-01 5.91649234e-01 2.29805082e-01 -7.02407897e-01
-1.48680210e+00 -9.88175273e-01 -5.10076582e-01 4.60030258e-01
6.27953887e-01 -3.66016090e-01 -9.05288219e-01 -8.85049701e-02
1.80429175e-01 9.92753431e-02 -8.58216345e-01 5.79160035e-01
-6.78728104e-01 -3.45045179e-01 7.20902085e-02 -1.51502565e-01
5.59009612e-01 -9.74571705e-01 -7.60484219e-01 2.18593642e-01
1.10612335e-02 -8.95196855e-01 -8.96177292e-02 -4.02197719e-01
-1.08709288e+00 -1.32437050e+00 -5.75783551e-01 -1.03046799e+00
8.22105587e-01 3.15689802e-01 1.14772654e+00 2.74271220e-01
-3.79322261e-01 4.57516998e-01 -5.91554940e-01 -3.35796446e-01
-3.13499629e-01 3.22753675e-02 -4.49673235e-02 4.67440099e-01
4.18689139e-02 -7.04485714e-01 -2.92668313e-01 8.02002549e-01
-8.59120667e-01 2.19309106e-01 3.07507545e-01 3.91765177e-01
1.28263140e+00 4.08510596e-01 1.53691888e-01 -1.07074893e+00
5.33322453e-01 -2.57406920e-01 -6.46223843e-01 2.91665316e-01
3.18703279e-02 -3.24039310e-01 4.69944179e-01 -1.79523140e-01
-1.07153094e+00 2.39006672e-02 3.98513898e-02 -3.90805244e-01
-4.84735072e-01 4.48064715e-01 -1.55529454e-01 -3.40946823e-01
5.69611192e-01 2.78588794e-02 3.61441895e-02 -9.65407848e-01
1.11243628e-01 4.41179693e-01 3.66901964e-01 -8.73643100e-01
8.78533721e-01 8.45329106e-01 4.73857015e-01 -1.35341907e+00
-2.54820943e-01 -3.56999218e-01 -8.66240978e-01 -6.54432774e-01
7.98725367e-01 -3.61737460e-01 -5.44885218e-01 3.42644364e-01
-8.97378027e-01 1.70281623e-02 -3.56264502e-01 2.72921979e-01
-9.06657696e-01 4.70125288e-01 -6.74845278e-02 -7.34879017e-01
-2.86173578e-02 -1.03009939e+00 1.10071230e+00 4.05136459e-02
-5.50409496e-01 -1.08712375e+00 3.82230394e-02 -1.59000829e-02
2.19587326e-01 8.94664168e-01 1.20894337e+00 -1.18808515e-01
-4.81601685e-01 -1.72736183e-01 -1.59193948e-01 -1.98967129e-01
4.91329700e-01 1.69417143e-01 -6.79671645e-01 -5.51937595e-02
-1.29984170e-01 2.40699664e-01 1.84822813e-01 2.36381367e-01
1.23073411e+00 2.85507977e-01 -3.96154106e-01 3.74371976e-01
1.64657366e+00 2.15942159e-01 4.34224844e-01 4.29071754e-01
5.88066995e-01 6.90594316e-01 1.14274776e+00 5.43215513e-01
-5.80920950e-02 9.75930274e-01 2.94731647e-01 -2.52611130e-01
-1.42154053e-01 -5.47187589e-02 -4.43822086e-01 8.79343271e-01
-6.40009820e-01 -1.41507000e-01 -9.17469442e-01 3.47120464e-01
-1.86286199e+00 -9.62463856e-01 -6.65221989e-01 2.39589119e+00
3.76457393e-01 7.41657689e-02 -4.05920111e-02 5.26378036e-01
1.18825912e+00 -7.84443989e-02 -7.93681890e-02 -6.45160198e-01
-3.60328674e-01 6.14574611e-01 3.94813001e-01 5.50977647e-01
-1.13464391e+00 1.03397679e+00 5.29653597e+00 1.28210366e+00
-1.15889859e+00 -6.91921860e-02 3.87405396e-01 5.95252633e-01
-1.60913318e-01 -2.08228528e-01 -1.86626598e-01 2.50191659e-01
6.55694902e-02 -3.27692479e-01 1.02836959e-01 7.75561810e-01
5.21927655e-01 -5.41155696e-01 -6.35617554e-01 1.23196888e+00
-1.06717184e-01 -1.28451395e+00 2.83141732e-01 2.61425637e-02
8.51919830e-01 -6.42432272e-01 2.35950481e-02 -5.48855841e-01
-3.76354963e-01 -8.98249090e-01 1.05186582e+00 6.03372812e-01
1.11367762e+00 -8.95944536e-01 6.38168216e-01 3.47972359e-03
-1.73988879e+00 6.14374518e-01 -3.83216769e-01 -3.27375978e-02
1.79979071e-01 8.27791452e-01 -7.06211269e-01 8.88638675e-01
6.18221998e-01 8.49677980e-01 -4.35341954e-01 1.25454807e+00
2.03907013e-01 3.06066900e-01 -5.35027921e-01 -4.56968173e-02
1.86228007e-02 -6.83354855e-01 7.36999571e-01 1.32096577e+00
6.28689587e-01 1.20237119e-01 2.83595800e-01 6.23979747e-01
2.57279515e-01 7.93237031e-01 -9.19946909e-01 2.12810054e-01
5.73105276e-01 1.22063041e+00 -1.75177765e+00 -4.02142733e-01
2.20742896e-02 7.10729122e-01 -1.48974836e-01 1.03319041e-01
-5.24881780e-01 -3.78067046e-01 3.47708076e-01 5.23181736e-01
1.24240540e-01 -4.19625580e-01 -7.45400310e-01 -7.14949071e-01
7.87938386e-03 -5.79577506e-01 1.14263102e-01 -6.14766300e-01
-1.10589826e+00 4.87544894e-01 1.31971627e-01 -1.75186253e+00
4.60453808e-01 -5.09975076e-01 -2.37033755e-01 8.29019725e-01
-8.63422573e-01 -1.09985781e+00 -5.03424168e-01 5.37002325e-01
6.67565882e-01 7.63468444e-02 8.32936645e-01 4.13593471e-01
4.05906476e-02 -3.15181841e-03 3.20366621e-01 -3.67866516e-01
5.14730632e-01 -1.17395210e+00 1.89233974e-01 5.14562011e-01
-1.25765696e-01 4.00902301e-01 9.91267025e-01 -9.63299811e-01
-1.20205557e+00 -1.00275123e+00 9.80218947e-01 -1.02647811e-01
3.50570649e-01 -3.24283898e-01 -9.03811932e-01 2.87477791e-01
-1.95686007e-03 -2.57040143e-01 3.47025782e-01 -1.21748447e-01
1.90961882e-01 1.84446294e-02 -1.38848376e+00 6.25604391e-01
1.16925776e+00 -2.56204724e-01 -5.58629155e-01 1.51859969e-01
-1.18525840e-01 -6.82697892e-01 -1.23771632e+00 5.79109430e-01
5.18614829e-01 -1.09361851e+00 8.33377361e-01 1.72623307e-01
4.94838990e-02 -7.61610270e-01 -1.58105150e-01 -1.08005071e+00
-2.70905882e-01 -5.37324429e-01 5.13096690e-01 9.33459580e-01
-1.08761169e-01 -3.77375573e-01 7.32097626e-01 -2.05034375e-01
-1.99419335e-01 -6.22132123e-01 -9.73202109e-01 -7.22437561e-01
-4.53005880e-01 -5.09855747e-01 6.62957370e-01 1.17798543e+00
-2.75622070e-01 -1.04543023e-01 -2.34919056e-01 5.91222309e-02
8.02022994e-01 5.42298853e-01 9.16112483e-01 -1.72750783e+00
3.08157742e-01 -6.02198005e-01 -8.46924186e-01 -5.02265751e-01
-2.51282722e-01 -8.34290564e-01 -1.99251264e-01 -1.35771978e+00
-2.45967776e-01 -9.58108962e-01 1.61274493e-01 -8.69079009e-02
2.75094688e-01 5.06090820e-01 -1.15132727e-01 1.57175809e-01
-3.73084933e-01 2.82109469e-01 1.35335231e+00 8.80588070e-02
-1.12934582e-01 -1.06130712e-01 1.48367524e-01 8.60537827e-01
9.53804016e-01 -2.45635867e-01 -4.35564548e-01 -2.47613847e-01
2.50845611e-01 -2.50043482e-01 2.21939757e-01 -1.01331639e+00
-5.29316179e-02 -2.95319319e-01 1.78350255e-01 -7.34892070e-01
2.44043827e-01 -1.18274903e+00 8.85672271e-01 2.15239733e-01
2.65554309e-01 2.65880376e-01 1.28656641e-01 4.88886088e-01
-2.44961947e-01 6.86140805e-02 1.01273620e+00 -2.50327706e-01
-7.91820347e-01 1.36516511e-01 -4.53663290e-01 -1.55630007e-01
1.15781665e+00 -8.05629253e-01 2.26952240e-01 7.89560154e-02
-1.04448557e+00 -3.06699455e-01 1.20659149e+00 2.86526501e-01
6.62737668e-01 -1.45243144e+00 -3.25008810e-01 1.12696171e-01
2.12130666e-01 -9.61516704e-03 4.74073917e-01 8.64013076e-01
-1.19546628e+00 -3.21862623e-02 -5.10953128e-01 -8.92166317e-01
-1.59570086e+00 2.35081002e-01 1.58223361e-01 7.09493309e-02
-8.16381037e-01 2.81942517e-01 1.08841866e-01 -2.12533638e-01
5.45474812e-02 -5.03137529e-01 -3.16731513e-01 2.88356598e-02
1.94964353e-02 6.36082768e-01 3.12392890e-01 -9.12425935e-01
-2.38859937e-01 1.53557515e+00 6.12576604e-01 -3.74989994e-02
1.13213170e+00 -2.71128953e-01 -4.87017453e-01 8.18825424e-01
8.39714944e-01 4.21973467e-01 -5.46499431e-01 -2.53092080e-01
3.64779174e-01 -9.12527144e-01 -3.17752808e-01 -3.32038224e-01
-6.89461589e-01 5.98665833e-01 6.08073056e-01 4.74232465e-01
1.08516955e+00 -4.64479811e-02 3.20371360e-01 1.89863935e-01
8.32841635e-01 -1.37947631e+00 -3.63601685e-01 4.58619773e-01
1.04181480e+00 -7.54387677e-01 1.23206265e-01 -1.18494177e+00
-6.81600943e-02 1.36882436e+00 2.84721851e-01 -3.12279463e-01
8.99748445e-01 7.58319348e-02 -1.71125401e-02 -6.71555161e-01
-8.22673589e-02 -2.16976270e-01 2.81559110e-01 3.06978256e-01
4.95392829e-01 2.23755881e-01 -8.42441797e-01 -1.09801643e-01
-3.06253195e-01 -8.26624706e-02 2.66403198e-01 1.21123159e+00
-6.70417428e-01 -1.18523574e+00 -7.92680681e-01 4.09321249e-01
-3.15611452e-01 4.04006034e-01 -2.82529682e-01 9.52659130e-01
4.53489125e-01 6.77720726e-01 2.13184461e-01 -3.67814690e-01
7.52351761e-01 -2.12921605e-01 6.88185096e-01 -6.14472330e-01
-4.31924075e-01 2.57460088e-01 1.39864668e-01 -4.38962102e-01
-7.33296871e-01 -7.43642569e-01 -1.07271254e+00 -3.76123667e-01
-2.50152405e-03 2.90466070e-01 7.13705897e-01 5.55187583e-01
2.28669181e-01 2.53530055e-01 6.57926738e-01 -9.39178288e-01
3.83547485e-01 -7.71959364e-01 -9.84633267e-01 7.42494166e-01
-2.41645649e-01 -1.21064842e+00 -4.19013761e-02 3.35326016e-01] | [8.454460144042969, -2.599250555038452] |
6330fbc4-7010-49a4-ad2f-25a148cdcd60 | losparse-structured-compression-of-large | 2306.11222 | null | https://arxiv.org/abs/2306.11222v2 | https://arxiv.org/pdf/2306.11222v2.pdf | LoSparse: Structured Compression of Large Language Models based on Low-Rank and Sparse Approximation | Transformer models have achieved remarkable results in various natural language tasks, but they are often prohibitively large, requiring massive memories and computational resources. To reduce the size and complexity of these models, we propose LoSparse (Low-Rank and Sparse approximation), a novel model compression technique that approximates a weight matrix by the sum of a low-rank matrix and a sparse matrix. Our method combines the advantages of both low-rank approximations and pruning, while avoiding their limitations. Low-rank approximation compresses the coherent and expressive parts in neurons, while pruning removes the incoherent and non-expressive parts in neurons. Pruning enhances the diversity of low-rank approximations, and low-rank approximation prevents pruning from losing too many expressive neurons. We evaluate our method on natural language understanding, question answering, and natural language generation tasks. We show that it significantly outperforms existing compression methods. | ['Tuo Zhao', 'Weizhu Chen', 'Pengcheng He', 'Chen Liang', 'Qingru Zhang', 'Yifan Yu', 'Yixiao Li'] | 2023-06-20 | null | null | null | null | ['model-compression', 'text-generation', 'question-answering'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [ 3.48873645e-01 6.81763142e-02 -1.77006021e-01 -2.30690047e-01
-5.59307158e-01 -2.65204310e-01 4.13578600e-01 5.16946949e-02
-3.46160233e-01 6.18290961e-01 4.77099597e-01 -1.62275452e-02
-2.19638050e-01 -9.00002062e-01 -7.46691108e-01 -4.80076104e-01
3.74282151e-02 7.65599310e-01 4.00599360e-01 -1.04629606e-01
1.55066267e-01 3.25487643e-01 -2.03906274e+00 9.55169857e-01
9.05571878e-01 1.12503517e+00 4.30618256e-01 4.05936450e-01
-3.76673877e-01 1.08946717e+00 -3.43882293e-01 -2.73626536e-01
1.08241148e-01 -1.68448702e-01 -8.10688198e-01 -3.46459597e-01
5.37269652e-01 -2.63763934e-01 -7.42814481e-01 1.00849986e+00
1.68180317e-01 1.05453677e-01 3.57887149e-01 -1.11096907e+00
-4.50561434e-01 9.51665699e-01 -5.06695032e-01 1.99329615e-01
1.65407524e-01 -3.87404948e-01 1.01334774e+00 -9.37897861e-01
4.22570825e-01 1.34902227e+00 5.21780550e-01 4.74991709e-01
-1.35518730e+00 -8.14412057e-01 2.23786831e-01 1.13106705e-01
-1.48529959e+00 -7.69835711e-01 4.24759090e-01 -1.33669078e-01
1.41054964e+00 4.91344184e-01 7.48923302e-01 7.45393515e-01
-6.36953115e-02 9.86611962e-01 7.69674659e-01 -3.04588228e-01
2.83135682e-01 -2.46818841e-01 4.57076639e-01 1.09639943e+00
5.28227091e-01 -1.46963790e-01 -8.06861818e-01 -6.22235596e-01
6.86277092e-01 2.24142373e-01 -2.13296324e-01 -2.32674986e-01
-9.70250547e-01 6.34462535e-01 1.24267295e-01 2.79860407e-01
-3.44509095e-01 5.50907135e-01 3.61573011e-01 2.53404468e-01
1.11286469e-01 3.04969221e-01 -3.25278133e-01 -2.97701240e-01
-1.04510033e+00 3.36727798e-01 8.39736760e-01 1.17817938e+00
6.28977478e-01 1.32528588e-01 -1.42817795e-01 1.17531586e+00
-1.65395048e-02 4.00485039e-01 8.38865817e-01 -1.30386126e+00
4.30937856e-01 9.05738652e-01 -2.87858993e-01 -1.00433493e+00
-2.86106795e-01 -5.40369034e-01 -1.34579456e+00 -3.69385689e-01
-4.72490303e-02 4.17335957e-01 -7.08029926e-01 1.96975124e+00
2.01137662e-02 4.40569147e-02 -4.26942185e-02 4.27350312e-01
5.81488609e-01 8.57916892e-01 -5.91996163e-02 -3.69507194e-01
1.23235321e+00 -1.22112036e+00 -6.55097067e-01 -3.83772016e-01
7.82906473e-01 -4.36406642e-01 1.22697103e+00 5.58639169e-01
-1.60836256e+00 -3.15200597e-01 -8.67787600e-01 -4.74500358e-01
-1.12815388e-01 2.12181568e-01 8.89710367e-01 1.02858670e-01
-7.71286011e-01 6.06847107e-01 -7.92322159e-01 1.58032611e-01
5.12061059e-01 4.50594544e-01 -5.17753720e-01 -4.53576356e-01
-1.04122937e+00 5.11964381e-01 3.49794984e-01 -3.45915556e-01
-5.76832771e-01 -8.39570463e-01 -8.45772386e-01 5.74071109e-01
2.02756777e-01 -7.87599921e-01 1.36709321e+00 -4.87712860e-01
-9.73260343e-01 4.95729625e-01 -5.27458191e-01 -8.98942709e-01
-6.15522712e-02 -3.30044866e-01 -1.72141209e-01 2.10562617e-01
-2.68955350e-01 7.65593290e-01 7.00094461e-01 -8.06024015e-01
-6.13025248e-01 -3.25645268e-01 4.62646820e-02 8.89909714e-02
-8.95083785e-01 -3.49901408e-01 -6.98798716e-01 -7.84418821e-01
4.13412184e-01 -6.17811143e-01 -2.93225944e-01 7.07561746e-02
-8.81899670e-02 -1.57696903e-01 5.89178860e-01 -2.83951700e-01
1.84708595e+00 -2.19230938e+00 1.68658629e-01 1.61289528e-01
6.38095796e-01 1.45313546e-01 -5.31290233e-01 2.32589021e-01
1.27344981e-01 3.03170308e-02 -3.25594723e-01 -1.63684085e-01
6.66439580e-03 6.10341668e-01 -6.05740011e-01 -8.40657800e-02
-1.07137963e-01 7.05226898e-01 -8.94215941e-01 -5.43636501e-01
-2.16074079e-01 3.71445388e-01 -9.86270845e-01 -3.90995927e-02
-3.55106235e-01 -6.66274786e-01 -3.10337245e-01 5.33533633e-01
5.08362949e-01 -3.80706787e-01 2.64756512e-02 -4.27471757e-01
3.01610291e-01 5.76305687e-01 -1.14150703e+00 1.46816945e+00
-4.84117508e-01 3.31574738e-01 -1.92618091e-02 -8.80149305e-01
8.32730830e-01 -4.59234975e-03 3.84998828e-01 -8.64307046e-01
-1.01153225e-01 3.06796074e-01 -7.63024464e-02 -2.45184958e-01
5.32119036e-01 6.71292702e-03 -1.84987653e-02 5.96286058e-01
-7.67545775e-02 -5.16017713e-03 9.20792997e-01 6.43192470e-01
1.31216455e+00 -2.76105344e-01 2.31614053e-01 -2.69603699e-01
4.41542655e-01 -1.51753277e-01 6.24632955e-01 6.07100606e-01
3.33815128e-01 4.68053997e-01 5.95596850e-01 -6.25816762e-01
-9.32374001e-01 -1.05115521e+00 1.29079834e-01 1.30841458e+00
-2.56072432e-01 -1.11812007e+00 -7.75406420e-01 -2.32745737e-01
1.72352903e-02 7.58991838e-01 -3.88639182e-01 -4.84873474e-01
-6.10815406e-01 -6.02543592e-01 7.44043887e-01 7.64434636e-01
4.04876530e-01 -7.43602216e-01 -6.18683338e-01 1.85490862e-01
-4.27447915e-01 -1.02510846e+00 -4.97938156e-01 3.74915957e-01
-1.22707450e+00 -1.02894437e+00 -2.60373712e-01 -8.78520548e-01
9.27405596e-01 5.00534058e-01 1.34830391e+00 3.51756841e-01
-2.20479578e-01 -1.82648748e-01 -5.95030636e-02 -3.09486598e-01
-1.06515333e-01 8.18351060e-02 1.91265538e-01 -3.84004235e-01
4.77076083e-01 -8.07531536e-01 -5.69846183e-02 1.07743219e-01
-1.17923999e+00 4.55440551e-01 6.34689212e-01 9.72483039e-01
1.12886691e+00 2.11975157e-01 4.03438777e-01 -1.16250265e+00
7.45499790e-01 -2.41664663e-01 -6.22573614e-01 3.25831771e-01
-5.49883783e-01 4.62640762e-01 9.36709464e-01 -6.11135542e-01
-7.07764089e-01 2.14876235e-01 -1.87872183e-02 -6.09026372e-01
3.52743834e-01 5.81654072e-01 4.18924130e-02 1.90606534e-01
6.55591071e-01 4.92925614e-01 -1.29026875e-01 -7.11739957e-01
3.62757683e-01 4.53139573e-01 6.79479599e-01 -7.49530435e-01
5.95188558e-01 4.36686784e-01 3.56997959e-02 -8.04617822e-01
-1.03153253e+00 -3.08580190e-01 -1.43812001e-01 3.77074510e-01
8.63079652e-02 -1.05657697e+00 -2.60264009e-01 8.38325545e-02
-1.04694247e+00 -1.57703161e-01 -8.14782083e-01 3.98842037e-01
-5.45103252e-01 4.45458144e-01 -9.61477697e-01 -6.05971873e-01
-6.31920576e-01 -7.29333818e-01 7.83421099e-01 -7.76777565e-02
-3.94134432e-01 -2.27725282e-01 5.19827828e-02 4.44978029e-01
4.14655238e-01 -2.91118979e-01 1.66036117e+00 -6.20934725e-01
-5.66345334e-01 -3.49120736e-01 -3.05780858e-01 2.51782238e-01
-4.22703266e-01 -3.27165872e-01 -6.53689802e-01 -1.56170979e-01
4.38344479e-03 -7.01705456e-01 1.18562484e+00 1.12431571e-01
1.68825924e+00 -7.76713312e-01 -3.38518709e-01 6.51671231e-01
1.10501933e+00 -1.63108632e-01 5.88602901e-01 -8.28200281e-02
6.00051939e-01 4.08746034e-01 2.40468547e-01 5.72101951e-01
2.89253205e-01 1.87567085e-01 2.38984242e-01 1.31342426e-01
-7.93398544e-02 -4.06224340e-01 3.19094926e-01 1.54210699e+00
3.88634689e-02 -1.09765299e-01 -6.60269201e-01 6.22323275e-01
-2.06487274e+00 -1.04268503e+00 -5.01725115e-02 2.08388686e+00
1.09705436e+00 2.52165884e-01 -1.20824985e-01 3.81864190e-01
3.15224379e-01 1.38360009e-01 -5.34546852e-01 -5.39123893e-01
-3.59896749e-01 2.56233275e-01 7.82024339e-02 3.14267635e-01
-5.66076338e-01 8.82770360e-01 7.28964520e+00 1.10048258e+00
-4.24637884e-01 -9.79931932e-03 5.87463379e-01 -8.21131885e-01
-6.77819431e-01 -1.87693372e-01 -7.48632431e-01 8.40843096e-02
9.68334913e-01 -3.17524493e-01 8.23982954e-01 1.07624578e+00
-2.34716237e-01 5.58306910e-02 -1.30631554e+00 1.22274935e+00
2.38433182e-01 -1.63607550e+00 6.95668936e-01 -6.56066462e-02
8.50858927e-01 1.72278300e-01 -3.33013013e-02 5.05848706e-01
2.83514529e-01 -1.14241254e+00 6.18515074e-01 6.40698195e-01
6.34116948e-01 -9.57803011e-01 5.05755365e-01 6.99431658e-01
-1.26463175e+00 -3.36558461e-01 -9.69838560e-01 -2.16989949e-01
-3.28861959e-02 9.67174113e-01 1.49986729e-01 -2.53008336e-01
7.04993784e-01 4.11607832e-01 -3.46234113e-01 7.34372914e-01
1.46162927e-01 4.13848311e-01 -7.62730539e-01 1.46349832e-01
1.45528927e-01 -2.47382060e-01 2.30015144e-01 1.29365826e+00
3.71363670e-01 4.85972404e-01 6.21063560e-02 9.04673338e-01
-5.85487485e-01 -3.80788967e-02 -6.87815249e-01 -4.55088794e-01
7.66241789e-01 1.04766941e+00 -3.41030747e-01 -7.14062274e-01
-3.68573874e-01 5.79889119e-01 6.90814018e-01 3.38914171e-02
-5.90153158e-01 -3.94915938e-01 6.11528695e-01 2.93110400e-01
2.53471166e-01 -1.78656071e-01 -3.76881719e-01 -1.16765296e+00
1.04463547e-01 -1.24775743e+00 4.64019686e-01 -8.13025236e-01
-9.75497842e-01 6.47101283e-01 -1.28868334e-02 -8.51084292e-01
-3.51399422e-01 -3.42097491e-01 -3.66365425e-02 3.89304668e-01
-1.34589422e+00 -8.04582238e-01 -1.65246472e-01 5.79676032e-01
6.27215803e-01 -4.21750069e-01 1.03278923e+00 3.77082437e-01
-5.50294518e-01 6.48332298e-01 1.94020048e-01 -1.78484306e-01
1.21301562e-01 -8.22838783e-01 1.88716412e-01 5.40334880e-01
4.46888059e-01 9.29446459e-01 2.67034709e-01 -3.59024674e-01
-1.47240341e+00 -9.85492110e-01 1.41272390e+00 6.15178123e-02
5.39248168e-01 -3.57977062e-01 -9.88416255e-01 4.56381142e-01
-2.45728582e-01 -1.14859911e-02 9.51538622e-01 8.77507925e-02
-8.28962266e-01 -1.89907119e-01 -9.79283094e-01 8.82961631e-01
1.22816074e+00 -6.96401238e-01 -6.97333217e-01 5.20042598e-01
7.43529737e-01 -1.53382599e-01 -7.29044855e-01 2.66977072e-01
6.51184738e-01 -8.43483150e-01 1.03170502e+00 -6.80181205e-01
6.92480505e-01 -1.34287298e-01 -4.57050771e-01 -6.85698450e-01
-6.51609063e-01 -5.74533284e-01 -1.12777984e+00 8.18949103e-01
3.58570367e-01 -2.13333502e-01 1.05824137e+00 6.52392805e-01
-4.85508777e-02 -1.27824414e+00 -7.91973710e-01 -8.04161251e-01
-1.88410178e-01 -4.48528737e-01 7.63724267e-01 6.34570658e-01
1.77214965e-01 4.88134295e-01 -3.53375554e-01 -3.18368167e-01
5.74510515e-01 4.94469374e-01 5.31042993e-01 -1.30954063e+00
-5.06981134e-01 -4.65935528e-01 8.09900761e-02 -1.35011685e+00
1.16701625e-01 -1.08306289e+00 -1.65720716e-01 -1.49809456e+00
6.58806026e-01 -3.25666696e-01 -3.22545737e-01 8.22791040e-01
3.43153536e-01 3.12789530e-01 2.00207144e-01 3.92262965e-01
-7.89954782e-01 5.07069409e-01 9.39991713e-01 -3.61721277e-01
-5.87819181e-02 -4.71831232e-01 -7.72909522e-01 1.20208323e+00
5.93880594e-01 -7.55883753e-01 -6.80387735e-01 -8.45883310e-01
5.24176955e-01 -1.78859860e-01 -1.50653683e-02 -1.13053501e+00
4.47033018e-01 -2.32239086e-02 3.49374741e-01 -8.96457314e-01
4.34436440e-01 -7.63127685e-01 1.09171957e-01 7.95098126e-01
-5.94174266e-01 3.05504143e-01 1.43342182e-01 4.97848779e-01
-3.64861578e-01 -3.50290179e-01 8.79950523e-01 -3.02778333e-01
-4.31561857e-01 3.89117241e-01 -3.41244847e-01 1.85343698e-01
5.42165875e-01 -3.19158077e-01 -3.10295910e-01 -1.89561024e-01
-5.66108048e-01 1.95909247e-01 3.04020435e-01 2.55032897e-01
9.07424033e-01 -1.42945182e+00 -4.21537846e-01 5.13447583e-01
1.74951870e-02 2.09997505e-01 1.35076866e-01 5.56740284e-01
-6.24580801e-01 5.41876733e-01 -1.44638389e-01 -1.30197957e-01
-1.41850913e+00 6.30753517e-01 -6.78406432e-02 -6.35903418e-01
-5.33700526e-01 9.73518014e-01 2.91619658e-01 -2.65191764e-01
6.19266272e-01 -5.23129702e-01 -3.79039980e-02 -8.46854672e-02
8.88843536e-01 6.90375507e-01 -1.34884700e-01 -3.73091668e-01
-2.20943727e-02 3.43673497e-01 -4.25504327e-01 1.82569727e-01
1.54281342e+00 1.78408757e-01 -5.45939803e-01 2.48156592e-01
1.16441894e+00 -1.50860593e-01 -7.43397653e-01 -4.71955210e-01
-9.62101445e-02 -2.32076421e-01 2.80427486e-01 -4.55658734e-01
-1.16485035e+00 1.04765236e+00 1.26605153e-01 -1.32402882e-01
1.32236433e+00 -1.62862003e-01 1.41951656e+00 1.26445436e+00
4.96625900e-01 -1.29343557e+00 2.52375990e-01 9.06906068e-01
9.19697165e-01 -3.17444921e-01 2.83920974e-01 -4.37048733e-01
-2.68030196e-01 8.75213861e-01 6.65894270e-01 -2.32846320e-01
4.91176009e-01 7.13995039e-01 -6.19900286e-01 -8.05543587e-02
-1.43663633e+00 1.06593966e-01 1.60558224e-01 4.72054929e-01
2.70357043e-01 -1.17018268e-01 -3.12053084e-01 9.99133706e-01
-3.26654255e-01 1.85187906e-01 2.33251989e-01 8.53886247e-01
-7.59533823e-01 -1.05170894e+00 -1.13848925e-01 1.04176474e+00
-2.77665228e-01 -4.46628273e-01 -4.90571767e-01 3.89019608e-01
2.09695294e-01 4.94306058e-01 3.27891469e-01 -4.57253903e-01
2.62168825e-01 1.83039278e-01 5.12337387e-01 -6.32813871e-01
-5.65292895e-01 1.34126872e-01 3.58140767e-02 -9.35710490e-01
4.78761084e-02 -1.75290376e-01 -1.41625857e+00 -5.32436132e-01
-2.57263571e-01 1.16117842e-01 2.85318524e-01 6.90432370e-01
5.40561080e-01 3.45166564e-01 5.85621968e-02 -3.56119663e-01
-9.64735031e-01 -7.15405464e-01 -5.62876642e-01 4.06380296e-01
1.20610237e-01 -3.53929371e-01 -2.00835094e-01 7.62156919e-02] | [8.708120346069336, 3.599560260772705] |
58d9a36f-c336-4d68-9a08-16f845cfb4e0 | posenet-a-convolutional-network-for-real-time | 1505.07427 | null | http://arxiv.org/abs/1505.07427v4 | http://arxiv.org/pdf/1505.07427v4.pdf | PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization | We present a robust and real-time monocular six degree of freedom
relocalization system. Our system trains a convolutional neural network to
regress the 6-DOF camera pose from a single RGB image in an end-to-end manner
with no need of additional engineering or graph optimisation. The algorithm can
operate indoors and outdoors in real time, taking 5ms per frame to compute. It
obtains approximately 2m and 6 degree accuracy for large scale outdoor scenes
and 0.5m and 10 degree accuracy indoors. This is achieved using an efficient 23
layer deep convnet, demonstrating that convnets can be used to solve
complicated out of image plane regression problems. This was made possible by
leveraging transfer learning from large scale classification data. We show the
convnet localizes from high level features and is robust to difficult lighting,
motion blur and different camera intrinsics where point based SIFT registration
fails. Furthermore we show how the pose feature that is produced generalizes to
other scenes allowing us to regress pose with only a few dozen training
examples. PoseNet code, dataset and an online demonstration is available on our
project webpage, at http://mi.eng.cam.ac.uk/projects/relocalisation/ | ['Matthew Grimes', 'Alex Kendall', 'Roberto Cipolla'] | 2015-05-27 | posenet-a-convolutional-network-for-real-time-1 | http://openaccess.thecvf.com/content_iccv_2015/html/Kendall_PoseNet_A_Convolutional_ICCV_2015_paper.html | http://openaccess.thecvf.com/content_iccv_2015/papers/Kendall_PoseNet_A_Convolutional_ICCV_2015_paper.pdf | iccv-2015-12 | ['camera-relocalization'] | ['computer-vision'] | [ 3.46391797e-02 -4.75485027e-02 2.92872041e-01 -4.32091981e-01
-6.22471333e-01 -7.67459095e-01 5.29953420e-01 -3.68632257e-01
-7.06976712e-01 6.76342189e-01 -2.41416067e-01 -2.72859782e-01
-1.66674271e-01 -4.03127313e-01 -1.26449060e+00 -3.62488776e-01
-1.12052657e-01 4.05071080e-01 9.65424478e-02 -3.26336116e-01
2.66642451e-01 8.47457826e-01 -1.26056504e+00 -4.67418075e-01
1.26912430e-01 1.05828857e+00 1.92324474e-01 1.21209252e+00
7.44672656e-01 6.08979821e-01 -4.63491887e-01 1.54828236e-01
9.23623264e-01 2.20638379e-01 -4.20761257e-01 2.75234163e-01
1.28202140e+00 -5.18764794e-01 -6.10107780e-01 7.57646561e-01
6.06379867e-01 6.55674338e-02 1.25519812e-01 -1.15668416e+00
-3.10931891e-01 -5.68292320e-01 -2.84651577e-01 -1.18703172e-02
7.29095101e-01 3.52535903e-01 6.51495278e-01 -8.51689458e-01
6.93333387e-01 1.00374568e+00 1.11866808e+00 1.11582845e-01
-1.22609031e+00 -6.32762372e-01 -2.52241731e-01 -2.38575339e-01
-1.55869567e+00 -5.01913011e-01 6.37989581e-01 -3.69142205e-01
1.41298544e+00 1.15238748e-01 8.60663116e-01 7.68377483e-01
4.90376562e-01 1.56280816e-01 9.63334024e-01 -2.72577852e-01
-1.41668245e-01 -2.46270135e-01 -4.13722813e-01 9.33793128e-01
1.74913198e-01 2.32643306e-01 -4.72158372e-01 3.37733656e-01
1.33597529e+00 1.53553948e-01 -2.58613110e-01 -9.55390096e-01
-1.48070312e+00 6.51044488e-01 1.06782460e+00 -2.28068963e-01
-5.52430749e-02 7.98648834e-01 -9.79941934e-02 5.70458353e-01
2.61079103e-01 5.27046680e-01 -8.20415616e-01 -9.64529067e-02
-7.29400277e-01 1.39461920e-01 6.25157773e-01 1.36620486e+00
1.14421976e+00 8.21173415e-02 9.78798270e-01 2.85830528e-01
3.40981424e-01 7.02346861e-01 1.93226427e-01 -1.30197418e+00
3.01258624e-01 3.52151424e-01 2.59762108e-01 -1.12186313e+00
-7.57574737e-01 -2.03421757e-01 -6.65319741e-01 6.13174736e-01
3.27108443e-01 -2.38903388e-01 -9.91248429e-01 1.16862631e+00
3.37537587e-01 -1.20932525e-02 -1.81151643e-01 1.07457888e+00
5.24216235e-01 2.41272733e-01 -6.69181645e-01 3.93630117e-01
9.90670025e-01 -6.95492804e-01 -9.92754772e-02 -5.87266147e-01
6.01736546e-01 -1.07189810e+00 7.88569093e-01 4.96733159e-01
-9.19440746e-01 -6.54023528e-01 -1.45959330e+00 -4.20524150e-01
-3.98718536e-01 2.01554626e-01 7.54379928e-01 4.36320603e-01
-1.62207568e+00 6.85791731e-01 -1.03765702e+00 -6.56245410e-01
2.83550352e-01 1.02337849e+00 -8.85771513e-01 -1.54883593e-01
-6.92385495e-01 9.69595194e-01 1.47766784e-01 2.17653096e-01
-6.23209655e-01 -7.04865396e-01 -1.12948453e+00 -4.83774364e-01
1.98328316e-01 -9.68284130e-01 1.23803461e+00 -8.18232656e-01
-1.51174498e+00 7.81455636e-01 9.15756747e-02 -4.42258954e-01
5.51558256e-01 -6.12558365e-01 -1.71748158e-02 2.10003912e-01
1.20915778e-01 9.89766181e-01 8.72472405e-01 -8.46755564e-01
-4.95387793e-01 -4.86142755e-01 2.25921601e-01 4.13379699e-01
2.02923775e-01 -3.36081952e-01 -4.23906446e-01 -2.02498242e-01
5.41617393e-01 -1.39319491e+00 -4.18902785e-01 3.79172832e-01
-1.54852167e-01 2.81138539e-01 9.14735019e-01 -3.95991415e-01
4.32532094e-02 -1.94436121e+00 4.91065197e-02 6.91054612e-02
1.96482569e-01 -7.19050989e-02 -3.57087259e-03 4.08145458e-01
-2.17243239e-01 -2.96724737e-01 1.13207467e-01 -3.67351919e-01
-4.29550931e-02 1.60800651e-01 1.14723176e-01 1.28018272e+00
1.51694357e-01 8.78299713e-01 -8.14131379e-01 7.10333288e-02
8.06310177e-01 9.30144787e-01 -5.64958096e-01 1.89557925e-01
1.03118472e-01 7.38022029e-01 1.39072791e-01 7.28102088e-01
7.07457542e-01 -1.25854343e-01 -2.11486608e-01 -3.57313603e-01
-1.43553644e-01 3.28279912e-01 -1.40600324e+00 2.34896898e+00
-7.06855595e-01 1.26891017e+00 2.53211886e-01 -7.53425539e-01
1.01746631e+00 -2.82745436e-02 4.23804164e-01 -5.73953748e-01
2.55094886e-01 2.09354118e-01 -2.98477471e-01 1.96732208e-02
6.78466856e-01 -6.98709488e-03 -1.63630486e-01 -2.00453144e-03
2.75500327e-01 -8.24722826e-01 -2.03875959e-01 9.44943875e-02
1.09262085e+00 6.12220228e-01 2.32667103e-01 -3.06466639e-01
1.88189089e-01 1.45547032e-01 2.41260454e-01 3.63342702e-01
-3.14027220e-02 8.55658829e-01 -3.56820561e-02 -9.81420040e-01
-1.27045083e+00 -1.00152647e+00 -4.68492992e-02 5.29851139e-01
3.53365302e-01 -4.15477902e-01 -4.24118966e-01 -2.70907372e-01
9.79723111e-02 -1.00911736e-01 -2.93116182e-01 6.68697953e-02
-6.88752115e-01 -2.13610202e-01 2.42931008e-01 6.83268726e-01
7.57799029e-01 -5.44234216e-01 -8.94631505e-01 4.89127673e-02
2.72225559e-01 -1.30076563e+00 -2.95395970e-01 4.05803829e-01
-7.37593412e-01 -1.24470925e+00 -4.91882473e-01 -8.40113044e-01
8.19328249e-01 5.29748261e-01 1.09282172e+00 -4.93198000e-02
-5.68233490e-01 7.67874122e-01 1.01406887e-01 -1.18346117e-01
2.72718906e-01 3.53985913e-02 5.12287438e-01 -5.38513541e-01
2.60391593e-01 -8.83418739e-01 -8.29302132e-01 4.82926875e-01
-4.59348083e-01 2.27395212e-03 5.97072601e-01 6.59640372e-01
5.49578071e-01 -2.54518747e-01 -2.67152637e-01 -2.69069165e-01
3.38152982e-02 6.66163713e-02 -1.35444236e+00 -4.19058174e-01
-4.48311001e-01 -9.63772982e-02 5.16220570e-01 -1.65813282e-01
-3.51361692e-01 9.39573526e-01 6.25451803e-02 -5.01487017e-01
-4.98582512e-01 8.84839520e-02 1.63828477e-01 -8.98511231e-01
7.75136709e-01 -8.29238594e-02 3.97914536e-02 -2.03728989e-01
4.79263663e-01 3.70164990e-01 6.79331005e-01 -2.85216123e-01
1.12331057e+00 6.65230334e-01 4.57687497e-01 -1.11084378e+00
-5.56934297e-01 -6.67801082e-01 -1.26231873e+00 -2.06919193e-01
9.42090809e-01 -1.40450668e+00 -9.97137547e-01 3.58784050e-01
-1.16413486e+00 -5.91118038e-01 -3.35228778e-02 6.80185735e-01
-8.50666344e-01 7.94962198e-02 -3.49360883e-01 -3.43756080e-01
-1.07495815e-01 -1.21845829e+00 1.46274841e+00 1.22362033e-01
-1.82388648e-01 -9.18596447e-01 -4.38893884e-02 1.83468804e-01
2.74307668e-01 5.71411371e-01 -2.18043014e-01 1.45870790e-01
-1.23608267e+00 -5.54232001e-01 -2.31982470e-01 2.33727604e-01
4.32200544e-02 -1.31891921e-01 -8.36150408e-01 -7.31102765e-01
-6.09164089e-02 -5.50666451e-01 5.72191477e-01 2.68109053e-01
6.32773161e-01 -1.41595766e-01 -4.11071144e-02 1.41201174e+00
1.77666748e+00 -2.97501713e-01 5.63930750e-01 7.05285013e-01
1.23925769e+00 2.48868123e-01 5.14762044e-01 3.35366540e-02
3.86940718e-01 7.79455900e-01 7.75386691e-01 -3.45280886e-01
-2.15829834e-02 -1.15148433e-01 2.61593133e-01 5.64618111e-01
-1.79662123e-01 2.39803478e-01 -1.08791113e+00 3.46943527e-01
-1.58818257e+00 -3.70194644e-01 -2.71876365e-01 2.21954346e+00
3.26308519e-01 1.73550695e-01 -1.38911813e-01 -1.36892423e-02
2.28370443e-01 3.61975320e-02 -4.97879028e-01 -2.91938603e-01
8.91474411e-02 3.93350869e-01 1.43686330e+00 8.74511600e-01
-1.24495316e+00 1.13629055e+00 5.59833479e+00 1.49149984e-01
-1.20281589e+00 -2.80389607e-01 1.63994938e-01 -1.96603313e-01
3.20582002e-01 2.86036223e-01 -7.26860523e-01 -1.06185392e-01
9.08894777e-01 3.55791450e-01 6.80961370e-01 8.76138330e-01
2.79258471e-02 -2.34411895e-01 -1.10854495e+00 1.35910630e+00
1.27789527e-01 -1.35465991e+00 -5.54685056e-01 3.34131092e-01
7.24266946e-01 8.22388113e-01 -1.26626134e-01 -1.65356770e-01
4.56508845e-01 -1.16170418e+00 5.50805509e-01 3.90367866e-01
9.53451872e-01 -8.30879927e-01 4.60091978e-01 3.99974287e-01
-1.19002044e+00 1.27514094e-01 -5.58318257e-01 -4.29870665e-01
-8.15733746e-02 1.93592921e-01 -1.16140330e+00 4.79486793e-01
9.28409815e-01 1.03410232e+00 -7.69507349e-01 9.98850524e-01
-3.04754257e-01 -1.73275784e-01 -7.78347909e-01 -2.98019079e-03
2.59375453e-01 -1.85801193e-01 3.48096728e-01 9.83810484e-01
4.72782999e-01 -1.09819144e-01 1.67450204e-01 3.42855781e-01
-1.91014577e-02 -4.13757652e-01 -8.58864248e-01 5.11462450e-01
3.02977953e-02 1.41602552e+00 -6.71795547e-01 2.50098277e-02
-2.74084061e-01 1.22202551e+00 2.54045844e-01 3.42234612e-01
-5.98229766e-01 -7.00177908e-01 8.51516128e-01 2.50236660e-01
6.28611028e-01 -1.00209618e+00 -1.74007565e-01 -1.29973435e+00
1.25352755e-01 -6.26979530e-01 -2.84771770e-01 -1.29531837e+00
-7.27570057e-01 4.61016893e-01 -2.11674616e-01 -1.31197321e+00
-3.85968745e-01 -1.08397925e+00 -1.04124889e-01 8.22155178e-01
-1.45724773e+00 -1.17025745e+00 -7.73683906e-01 8.71033728e-01
3.75215590e-01 3.81104425e-02 8.08631301e-01 1.58047527e-01
9.13117826e-02 3.22121918e-01 -2.13174328e-01 2.91526020e-01
8.10436606e-01 -1.44361162e+00 7.93173790e-01 8.32175255e-01
1.83775440e-01 6.75309241e-01 7.22370207e-01 -2.71609634e-01
-1.97941232e+00 -1.00687146e+00 6.75753117e-01 -8.91222060e-01
7.04912305e-01 -7.05818117e-01 -2.94333041e-01 9.94926512e-01
2.98616409e-01 5.27133584e-01 6.94400817e-02 -1.16952606e-01
-3.39690745e-01 -3.45962882e-01 -1.03872347e+00 4.36609715e-01
1.14221334e+00 -5.98735631e-01 -2.69303530e-01 7.63680696e-01
7.85032809e-01 -1.10926771e+00 -8.67811084e-01 1.82310045e-01
5.38889468e-01 -1.05536282e+00 1.29672253e+00 -6.39294013e-02
1.22782148e-01 -5.46619177e-01 -3.93774122e-01 -1.13643742e+00
-2.67665267e-01 -1.01025248e+00 1.45140722e-01 5.45319080e-01
1.46765500e-01 -7.90647507e-01 9.03697550e-01 3.77246767e-01
-1.92973793e-01 -3.45123857e-01 -1.03099549e+00 -8.39761198e-01
-3.05510253e-01 -5.39154053e-01 2.87022799e-01 7.41211414e-01
-5.05317867e-01 4.53889072e-01 -4.12175208e-01 5.87937832e-01
7.61232913e-01 -5.70238642e-02 1.48347020e+00 -9.85816061e-01
-3.25994939e-01 3.77752297e-02 -1.20080662e+00 -1.41464591e+00
-6.97703063e-02 -5.86797595e-01 2.78597996e-02 -1.40372932e+00
-6.11724138e-01 -3.07881534e-01 9.88733396e-02 4.00457889e-01
4.81288999e-01 8.96547854e-01 1.61390647e-01 9.60177258e-02
-5.21582723e-01 1.43485263e-01 9.82883692e-01 1.47826016e-01
4.95453514e-02 -8.17117244e-02 -3.09581578e-01 8.68835270e-01
1.04841375e+00 -2.72177309e-01 -2.25683585e-01 -7.00262487e-01
4.48861361e-01 -1.53383702e-01 7.80682683e-01 -1.49117959e+00
4.74021971e-01 1.90250039e-01 9.74640131e-01 -6.19384289e-01
8.01330447e-01 -1.11995125e+00 1.65069342e-01 3.95245850e-01
1.47528529e-01 4.86664385e-01 3.89014095e-01 4.30395275e-01
9.38082486e-02 3.80846411e-01 6.29772365e-01 -4.12475884e-01
-8.67647946e-01 4.17292595e-01 8.27045366e-02 -2.78493881e-01
8.50816905e-01 -4.62191999e-01 -1.54625893e-01 -5.84747076e-01
-5.53624630e-01 -5.14911115e-02 1.01522982e+00 3.08598280e-01
7.08137393e-01 -1.38745260e+00 -3.64228427e-01 6.68353379e-01
9.25020799e-02 5.03272772e-01 -2.47148901e-01 8.28330576e-01
-1.39772284e+00 6.20444536e-01 -2.41883546e-01 -1.20231318e+00
-1.19223785e+00 2.77397662e-01 5.66963911e-01 3.69938970e-01
-7.23977923e-01 8.78914714e-01 -1.83831275e-01 -7.07462430e-01
1.81858048e-01 -4.35747743e-01 5.25132179e-01 -5.57103157e-01
1.84794933e-01 4.47699949e-02 2.81892270e-01 -7.64686286e-01
-5.73560357e-01 1.00639069e+00 2.84838587e-01 -1.72332063e-01
1.61862803e+00 -2.81745583e-01 -8.24272912e-03 1.45324260e-01
1.70986795e+00 -6.03078119e-02 -1.90831220e+00 1.83997765e-01
-3.34129184e-01 -7.44823158e-01 2.99790204e-01 -3.98889482e-01
-8.74678612e-01 8.39905500e-01 7.12805390e-01 -1.72025591e-01
1.00494576e+00 -1.51901320e-01 5.77688456e-01 8.41531813e-01
5.23605525e-01 -8.93319130e-01 -4.77698306e-03 7.90409386e-01
8.81976366e-01 -1.41050899e+00 4.76527393e-01 -1.14181489e-01
-6.22302480e-02 1.37157285e+00 5.05942702e-01 -8.89743984e-01
5.54716885e-01 2.80271053e-01 3.04399818e-01 -2.49212191e-01
-2.04555288e-01 2.58947890e-02 2.80895263e-01 8.36392403e-01
3.39979500e-01 -1.15883365e-01 4.21628326e-01 -5.84798276e-01
-6.40812874e-01 -2.04238549e-01 5.42626798e-01 1.15158629e+00
-2.72957206e-01 -8.96407783e-01 -4.19980586e-01 -7.35959858e-02
-1.62784189e-01 -6.43764660e-02 -4.13979650e-01 1.12143183e+00
6.54812828e-02 5.73924005e-01 1.33295074e-01 -5.69623291e-01
3.68598193e-01 -3.66290927e-01 8.85349691e-01 -4.25177455e-01
-3.81095469e-01 1.43354163e-01 -2.35030036e-02 -1.18168592e+00
-5.47955275e-01 -5.89631736e-01 -9.88312125e-01 -4.10994321e-01
-2.57698834e-01 -4.77634490e-01 1.19718301e+00 5.92870533e-01
4.15835410e-01 2.90337741e-01 5.08610487e-01 -1.63538480e+00
-1.11557037e-01 -6.83364689e-01 -2.60971248e-01 -6.09144755e-02
9.36590135e-01 -4.12084997e-01 -4.26718235e-01 1.21840388e-01] | [7.660409927368164, -2.2010600566864014] |
564378c3-8b76-481e-9b8b-5c4c084fba3f | knowledge-reasoning-via-jointly-modeling | 2301.02781 | null | https://arxiv.org/abs/2301.02781v1 | https://arxiv.org/pdf/2301.02781v1.pdf | Knowledge Reasoning via Jointly Modeling Knowledge Graphs and Soft Rules | Knowledge graphs (KGs) play a crucial role in many applications, such as question answering, but incompleteness is an urgent issue for their broad application. Much research in knowledge graph completion (KGC) has been performed to resolve this issue. The methods of KGC can be classified into two major categories: rule-based reasoning and embedding-based reasoning. The former has high accuracy and good interpretability, but a major challenge is to obtain effective rules on large-scale KGs. The latter has good efficiency and scalability, but it relies heavily on data richness and cannot fully use domain knowledge in the form of logical rules. We propose a novel method that injects rules and learns representations iteratively to take full advantage of rules and embeddings. Specifically, we model the conclusions of rule groundings as 0-1 variables and use a rule confidence regularizer to remove the uncertainty of the conclusions. The proposed approach has the following advantages: 1) It combines the benefits of both rules and knowledge graph embeddings (KGEs) and achieves a good balance between efficiency and scalability. 2) It uses an iterative method to continuously improve KGEs and remove incorrect rule conclusions. Evaluations on two public datasets show that our method outperforms the current state-of-the-art methods, improving performance by 2.7\% and 4.3\% in mean reciprocal rank (MRR). | ['Jun Zhao', 'Kang Liu', 'Shizhu He', 'Yinyu Lan'] | 2023-01-07 | null | null | null | null | ['knowledge-graph-embeddings', 'knowledge-graph-embeddings'] | ['graphs', 'methodology'] | [-1.02507509e-02 4.58372474e-01 -5.67911983e-01 -1.79305539e-01
-2.45979041e-01 -3.96843523e-01 5.16507149e-01 4.54079956e-01
-2.56652296e-01 7.29575396e-01 2.84456879e-01 -5.16351998e-01
-6.60579920e-01 -1.25626850e+00 -7.87334859e-01 -3.67772877e-01
7.76760057e-02 5.86243808e-01 5.91077328e-01 -2.80353487e-01
3.97713222e-02 3.04513365e-01 -1.54628491e+00 1.48867786e-01
1.29706085e+00 9.94868219e-01 -2.31656790e-01 8.47989023e-02
-4.87204105e-01 1.07122958e+00 -3.11564803e-01 -9.00675476e-01
7.20726028e-02 1.06515046e-02 -9.74548995e-01 -3.09267849e-01
-5.05528003e-02 -2.25270609e-03 -5.75315952e-01 1.10778165e+00
1.48653552e-01 1.45226136e-01 5.09888291e-01 -1.43917859e+00
-8.75529170e-01 8.22910666e-01 -3.56748134e-01 -1.64536342e-01
3.83865893e-01 -2.05405593e-01 1.30101633e+00 -8.48869860e-01
5.72498858e-01 1.28444409e+00 6.18654907e-01 3.66754353e-01
-1.09617889e+00 -5.98585665e-01 2.75705218e-01 6.92857265e-01
-1.64009106e+00 -2.27997765e-01 6.58873618e-01 -3.43996584e-01
9.65174019e-01 2.38472477e-01 4.99523938e-01 6.42862260e-01
-5.31763211e-02 7.05702603e-01 8.29241991e-01 -4.82133389e-01
5.49399376e-01 2.28149861e-01 2.62495726e-01 9.72669482e-01
8.03578138e-01 -2.01414570e-01 -5.11531770e-01 -4.20585811e-01
5.68933070e-01 1.33037701e-01 -4.28819656e-01 -4.37725812e-01
-1.00707233e+00 9.71397340e-01 4.66959000e-01 2.19371974e-01
-4.03218448e-01 1.35275155e-01 1.56687081e-01 2.30395153e-01
1.60911098e-01 4.92556036e-01 -5.45473516e-01 1.10934809e-01
-4.53620434e-01 3.30923408e-01 8.73305917e-01 8.27371001e-01
7.87143111e-01 -1.34491295e-01 -1.87529083e-02 8.49487305e-01
2.37049177e-01 4.29198295e-01 2.65709996e-01 -8.39977205e-01
5.98485947e-01 1.25904095e+00 9.31643099e-02 -1.48566401e+00
-1.98754176e-01 -2.08192840e-01 -8.80266905e-01 1.69185121e-02
1.25736177e-01 2.18487173e-01 -8.04714143e-01 1.64554012e+00
5.13202548e-01 1.51183590e-01 2.78412461e-01 5.69545567e-01
9.44482803e-01 3.52435559e-01 -7.38929659e-02 -6.94048256e-02
1.32886100e+00 -5.46312034e-01 -9.33550239e-01 -8.45319629e-02
4.58287269e-01 -2.15937927e-01 8.23254883e-01 3.52742940e-01
-7.29004562e-01 -2.27273092e-01 -1.17852056e+00 -1.67030826e-01
-7.87136018e-01 -1.51352927e-01 9.87024546e-01 5.15357196e-01
-8.18181336e-01 5.02930641e-01 -6.45030081e-01 -1.43335164e-01
4.52333540e-01 3.57748985e-01 -4.96210694e-01 -6.06661618e-01
-1.66926289e+00 8.02689433e-01 8.05244088e-01 1.45157427e-02
-3.15379202e-01 -8.04223657e-01 -1.04961848e+00 2.41960526e-01
1.14235520e+00 -7.24323511e-01 6.43562853e-01 -2.02525541e-01
-1.13012695e+00 2.91413128e-01 -9.30729732e-02 -3.48499447e-01
2.82022685e-01 -3.42308551e-01 -8.00896704e-01 6.52587861e-02
1.61410943e-01 2.20625326e-01 5.55800140e-01 -1.16116440e+00
-6.85996473e-01 -5.33423841e-01 3.43037397e-01 2.55464204e-02
-4.63429093e-01 -4.62622583e-01 -9.04986024e-01 -4.32982653e-01
2.12605730e-01 -7.01645613e-01 6.50540739e-02 -9.44445431e-02
-3.56790125e-01 -6.13007665e-01 5.59791744e-01 -6.65429831e-01
1.64959645e+00 -1.95000851e+00 2.17969045e-01 4.00938660e-01
5.35105050e-01 5.55638254e-01 2.05521345e-01 6.10169590e-01
1.43699750e-01 3.99145365e-01 -2.81035841e-01 1.94735348e-01
1.01169750e-01 5.99667013e-01 -3.79548997e-01 2.55015306e-02
2.11319208e-01 1.07616329e+00 -1.03391302e+00 -4.86904442e-01
1.38034388e-01 3.81851345e-01 -4.87377405e-01 -1.70449913e-01
-4.33702916e-01 -3.09452057e-01 -5.85693240e-01 6.10020041e-01
5.74612617e-01 -4.75295365e-01 7.94971228e-01 -3.12700182e-01
1.70844808e-01 4.17832211e-02 -1.54569173e+00 1.28886890e+00
-1.15225270e-01 -2.59215161e-02 -2.71997184e-01 -1.09236491e+00
9.01533842e-01 1.80270091e-01 2.61680365e-01 -4.61855888e-01
-1.90368131e-01 2.08515435e-01 -2.05572665e-01 -5.09287655e-01
3.69074583e-01 1.84660163e-02 -6.82684407e-02 8.99583176e-02
-8.52035545e-03 2.76668295e-02 3.24132591e-01 6.21566653e-01
1.33432984e+00 -1.39071688e-01 5.36401093e-01 -4.42214422e-02
5.90200663e-01 1.12104058e-01 1.12553167e+00 6.29507542e-01
2.39686936e-01 -1.40249520e-03 8.02069485e-01 -6.28280282e-01
-3.09666336e-01 -9.29412842e-01 2.62729257e-01 6.09462917e-01
1.95641503e-01 -8.46115828e-01 -1.88272402e-01 -1.06954050e+00
6.82154834e-01 9.57845867e-01 -7.70652056e-01 -3.33435178e-01
-2.26338878e-01 -4.82855618e-01 4.62833613e-01 7.54059613e-01
6.11303747e-01 -8.13059449e-01 1.22064399e-02 1.64203435e-01
-3.67862701e-01 -1.19841552e+00 -1.92988608e-02 -1.44828796e-01
-6.74339890e-01 -1.79805386e+00 7.45431334e-02 -4.80870038e-01
7.94360638e-01 2.46349156e-01 9.51550961e-01 2.33772188e-01
-7.42143914e-02 3.76832366e-01 -5.27817845e-01 -3.68268788e-01
-1.97610408e-01 -1.39995247e-01 2.27335349e-01 6.90672025e-02
5.81192374e-01 -5.30579269e-01 -3.05077225e-01 3.03317517e-01
-1.18519008e+00 -8.00794959e-02 6.71578109e-01 7.88335383e-01
7.62402594e-01 7.70259917e-01 7.08126664e-01 -1.38280272e+00
7.47670352e-01 -4.13769424e-01 -6.45700753e-01 7.54572034e-01
-1.27836883e+00 4.20038879e-01 7.63075888e-01 -7.54370466e-02
-1.02046323e+00 -2.02822059e-01 1.65125757e-01 -5.62883735e-01
3.33256185e-01 8.67039621e-01 -3.32556278e-01 3.34557593e-02
5.55574179e-01 6.54591694e-02 -3.44608463e-02 -4.35757309e-01
6.68775320e-01 4.49584126e-01 4.82080996e-01 -6.96382701e-01
9.36505139e-01 4.17367876e-01 1.77649092e-02 -5.42140305e-01
-8.91091466e-01 -3.55930120e-01 -2.47482300e-01 3.57333831e-02
5.06184995e-01 -6.77925587e-01 -8.97752762e-01 3.82458940e-02
-8.62481952e-01 1.05552487e-01 -2.53631115e-01 4.58864331e-01
-5.40456809e-02 6.04473591e-01 -3.09702307e-01 -6.20268881e-01
-2.35315040e-01 -6.56283379e-01 4.13623035e-01 1.90795600e-01
-6.75182119e-02 -9.11805093e-01 -1.74645796e-01 4.88813281e-01
2.65492588e-01 3.58483225e-01 1.50626469e+00 -7.23759413e-01
-5.42175949e-01 -2.68249005e-01 -3.24553996e-01 3.54913503e-01
1.97369635e-01 -1.99356273e-01 -4.93731767e-01 -5.00187725e-02
-3.89564395e-01 -2.71154523e-01 9.66133118e-01 -8.40445682e-02
1.16956222e+00 -3.99697810e-01 -4.10836995e-01 1.99878603e-01
1.52186024e+00 -8.25622305e-03 7.40576327e-01 9.78336111e-02
6.96132600e-01 5.14563203e-01 6.15153074e-01 3.66865575e-01
7.73704350e-01 5.25514066e-01 4.46478933e-01 3.54566336e-01
-1.53539851e-01 -7.01182425e-01 9.33809858e-03 8.25607657e-01
-2.87599862e-01 -3.37896720e-02 -1.01013112e+00 7.00144827e-01
-2.38672042e+00 -8.77324760e-01 -1.19135655e-01 2.23088980e+00
8.89975071e-01 2.50746071e-01 -2.60409087e-01 5.71772397e-01
7.46241212e-01 -6.30993247e-02 -6.15469158e-01 -2.23640710e-01
-6.64160401e-02 1.58308029e-01 3.68674338e-01 5.60364902e-01
-7.05360770e-01 9.67207134e-01 5.63166809e+00 7.74164736e-01
-5.61230600e-01 -1.85156062e-01 5.64863868e-02 2.17148378e-01
-5.85219920e-01 1.25756562e-01 -7.31631815e-01 3.29382390e-01
3.42308074e-01 -1.97396412e-01 5.25260389e-01 8.12752962e-01
-3.08188736e-01 -2.19986010e-02 -1.01969612e+00 9.20054674e-01
-5.77138737e-03 -1.52875113e+00 3.02803248e-01 5.88494241e-02
7.54699945e-01 -4.03693616e-01 -3.54773283e-01 7.61251330e-01
6.49972916e-01 -8.70942831e-01 3.40267181e-01 6.56911790e-01
7.63186514e-01 -8.93673658e-01 9.64427054e-01 2.13052884e-01
-1.14630628e+00 -1.97639853e-01 -4.15284157e-01 -1.22031659e-01
-1.63588867e-01 1.06202447e+00 -8.83015692e-01 1.18249118e+00
6.34098113e-01 5.90117395e-01 -5.48004329e-01 6.06210411e-01
-7.70950258e-01 5.50793469e-01 -2.15858459e-01 -2.93414220e-02
-1.43131316e-01 -1.17352098e-01 1.50000528e-01 7.32895255e-01
-4.97000925e-02 3.46671611e-01 1.66776940e-01 7.83780515e-01
-3.27227324e-01 -7.34785870e-02 -4.75013167e-01 -2.19328672e-01
7.57820725e-01 9.41575468e-01 -4.47873771e-01 -4.34713274e-01
-5.92490792e-01 6.34284973e-01 6.53405130e-01 3.45422179e-01
-6.84164882e-01 -7.28567719e-01 6.37127101e-01 -1.40958885e-03
5.99133015e-01 -2.20808592e-02 -3.75818200e-02 -1.14796662e+00
4.72478420e-01 -8.08748186e-01 8.23598087e-01 -4.62161958e-01
-1.26315975e+00 2.26359308e-01 5.25284782e-02 -5.72634637e-01
-4.61136103e-02 -6.68871403e-01 -2.61018842e-01 5.27421117e-01
-1.78760409e+00 -9.77488041e-01 -2.58105010e-01 7.06862569e-01
-7.78024346e-02 2.42548510e-02 9.53867555e-01 2.73204476e-01
-6.03627145e-01 5.41731954e-01 2.58926917e-02 1.60963863e-01
3.73816073e-01 -1.31843829e+00 1.11385368e-01 8.04405928e-01
3.85758318e-02 8.41258705e-01 3.08477461e-01 -9.59460557e-01
-1.79272640e+00 -1.24228156e+00 1.26859784e+00 -3.99389267e-01
5.96547306e-01 -1.38918787e-01 -1.04312348e+00 6.08683646e-01
-3.51740509e-01 2.52392709e-01 6.89358652e-01 7.16504514e-01
-8.64576340e-01 -4.77271676e-01 -1.14947784e+00 6.44170225e-01
1.14871883e+00 -3.96536887e-01 -1.00963056e+00 2.49132648e-01
8.10386658e-01 -8.91149566e-02 -9.98755276e-01 6.75436854e-01
4.22689676e-01 -8.09459388e-01 8.31759036e-01 -7.39325225e-01
2.15360820e-01 -6.74075127e-01 -2.27607399e-01 -1.07827902e+00
-5.72211981e-01 -4.83012438e-01 -7.10371375e-01 1.16498089e+00
5.00633001e-01 -9.72440779e-01 6.31391943e-01 8.28014672e-01
1.33761376e-01 -9.68896329e-01 -6.76129937e-01 -8.58101964e-01
-3.74638498e-01 -5.36269307e-01 9.33591783e-01 1.13512683e+00
2.86609381e-01 3.91298145e-01 -1.50159404e-01 4.98297960e-01
5.21518886e-01 4.05715168e-01 6.29159868e-01 -1.67277122e+00
-5.83602972e-02 -2.17944816e-01 -6.28963351e-01 -6.89041972e-01
1.49554446e-01 -1.12264061e+00 -5.01182616e-01 -2.09432817e+00
8.09736252e-02 -4.89272475e-01 -5.31659782e-01 1.06800616e+00
-3.36293340e-01 -3.38029474e-01 -5.92150763e-02 -1.31178731e-02
-7.92298436e-01 6.44543707e-01 8.29820693e-01 -1.99906006e-01
-8.75318944e-02 -3.44286799e-01 -1.11055624e+00 6.25627697e-01
7.29984164e-01 -3.97590190e-01 -7.10236549e-01 -3.33920538e-01
7.82794893e-01 -1.27403229e-01 3.46831232e-01 -7.36290395e-01
5.21741986e-01 -2.80913562e-01 7.09146038e-02 -3.48391861e-01
9.71498862e-02 -8.05137575e-01 2.71273673e-01 3.84035617e-01
-4.70843874e-02 -9.30526778e-02 1.80034824e-02 1.15152395e+00
-2.96834648e-01 -1.38751613e-02 2.35967562e-01 -7.58708082e-03
-9.31693077e-01 1.84616998e-01 1.37778550e-01 2.67226905e-01
8.81675243e-01 8.00869241e-02 -4.26576704e-01 -2.47407272e-01
-4.88435775e-01 5.89861393e-01 3.03963482e-01 6.52175844e-01
9.21261668e-01 -1.55141950e+00 -5.77248335e-01 8.03965181e-02
4.33253974e-01 1.49767160e-01 1.98211917e-03 6.63577378e-01
-2.30832383e-01 5.01387596e-01 3.51767153e-01 -1.07857503e-01
-8.42399716e-01 7.74084747e-01 -1.58860713e-01 -4.50704247e-01
-6.12558186e-01 5.74332118e-01 -1.38703868e-01 -5.18717527e-01
2.81483948e-01 -3.39558065e-01 -2.33899578e-01 -1.78906590e-01
6.47033453e-01 5.70761561e-01 2.27583185e-01 -5.69004640e-02
-5.63499033e-01 3.79782021e-01 -1.91612050e-01 3.64570588e-01
1.33287668e+00 3.30059171e-01 -2.86850959e-01 1.31166115e-01
7.50990331e-01 1.62915573e-01 -6.30635142e-01 -6.18310571e-01
6.02953732e-02 -5.89962661e-01 1.18256472e-01 -1.06290185e+00
-9.95789647e-01 3.75593603e-01 -1.39112562e-01 3.53720725e-01
9.08755422e-01 2.62399521e-02 5.67510068e-01 5.86640894e-01
3.92258584e-01 -1.23240983e+00 -1.81278452e-01 4.73527044e-01
8.88875544e-01 -1.04429281e+00 2.90041715e-01 -8.29124451e-01
-6.86043739e-01 9.08483803e-01 4.67080742e-01 2.24357441e-01
6.60278976e-01 -1.57571524e-01 -3.55412573e-01 -4.94409233e-01
-7.56984890e-01 -3.39495122e-01 2.99428433e-01 5.34003437e-01
-7.80466422e-02 2.49615937e-01 -4.71752077e-01 1.01562476e+00
1.17438965e-01 2.24277943e-01 7.04666302e-02 8.62970293e-01
-3.42290968e-01 -1.13590717e+00 -2.16186076e-01 6.38116241e-01
-2.24717051e-01 4.64103296e-02 -5.45702457e-01 9.32223201e-01
2.39551008e-01 1.24336243e+00 -6.52906299e-01 -6.55223429e-01
5.54758191e-01 3.33980262e-01 3.50762278e-01 -4.52513605e-01
-6.41846657e-02 -6.73583508e-01 2.64167935e-01 -5.72698295e-01
-2.86722630e-01 -3.27217817e-01 -1.43011653e+00 -4.57936078e-01
-4.59880292e-01 3.70670855e-01 4.94483531e-01 8.36513221e-01
8.69394779e-01 4.89092886e-01 4.01547492e-01 2.54896075e-01
-5.80243409e-01 -5.09478748e-01 -7.04606235e-01 5.40807307e-01
-1.72721639e-01 -1.00777793e+00 -3.82700026e-01 -2.56767899e-01] | [8.823484420776367, 7.823940277099609] |
ec3989f3-3029-4dce-aaee-8febdcfc965f | 191013296 | 1910.13296 | null | https://arxiv.org/abs/1910.13296v2 | https://arxiv.org/pdf/1910.13296v2.pdf | Improving sequence-to-sequence speech recognition training with on-the-fly data augmentation | Sequence-to-Sequence (S2S) models recently started to show state-of-the-art performance for automatic speech recognition (ASR). With these large and deep models overfitting remains the largest problem, outweighing performance improvements that can be obtained from better architectures. One solution to the overfitting problem is increasing the amount of available training data and the variety exhibited by the training data with the help of data augmentation. In this paper we examine the influence of three data augmentation methods on the performance of two S2S model architectures. One of the data augmentation method comes from literature, while two other methods are our own development - a time perturbation in the frequency domain and sub-sequence sampling. Our experiments on Switchboard and Fisher data show state-of-the-art performance for S2S models that are trained solely on the speech training data and do not use additional text data. | ['Jan Niehues', 'Thai-Son Nguyen', 'Sebastian Stueker', 'Alex Waibel'] | 2019-10-29 | null | null | null | null | ['sequence-to-sequence-speech-recognition'] | ['speech'] | [ 3.92642319e-01 1.84014648e-01 4.06792201e-02 -5.06664634e-01
-9.42093074e-01 -4.51760620e-01 8.32741559e-01 -9.60082784e-02
-4.78409648e-01 4.56352651e-01 5.04248738e-01 -7.58526504e-01
3.43585730e-01 -9.41083431e-02 -5.91272652e-01 -4.55747187e-01
1.38096929e-01 3.78465980e-01 2.42538974e-01 -6.28405094e-01
-1.00620769e-01 4.15009111e-01 -1.43627501e+00 3.04394722e-01
4.97053355e-01 9.08494294e-01 1.62163898e-01 9.62992609e-01
-2.70943075e-01 7.29597151e-01 -9.11862075e-01 1.14499196e-01
3.50326538e-01 -3.73388469e-01 -5.56036651e-01 8.34471658e-02
3.61380696e-01 -1.82307452e-01 -7.65091121e-01 6.71515167e-01
7.29212403e-01 3.35943311e-01 3.14183056e-01 -9.05627072e-01
-4.03507322e-01 5.82809269e-01 -1.40609294e-01 5.45315444e-01
1.99748531e-01 2.98122227e-01 6.26959503e-01 -8.90936553e-01
4.12342310e-01 1.38859606e+00 7.72448599e-01 6.68541014e-01
-1.32988858e+00 -5.30060291e-01 1.58523068e-01 5.96676487e-03
-1.16230953e+00 -1.18222880e+00 6.24708533e-01 -4.23641860e-01
1.56397784e+00 3.06382060e-01 4.77495402e-01 1.41395688e+00
-5.77756584e-01 9.37607944e-01 1.03330863e+00 -7.42140889e-01
3.39733213e-01 1.41189739e-01 1.69921383e-01 3.10697794e-01
-3.17086369e-01 4.90821004e-01 -6.58158481e-01 -1.70769498e-01
6.21065795e-01 -5.90762913e-01 -3.66750598e-01 1.25264525e-01
-9.08234000e-01 6.91696823e-01 -3.37839723e-02 5.14301956e-01
-1.95386723e-01 -6.27539726e-03 6.19996488e-01 4.86051589e-01
6.76189423e-01 4.91786957e-01 -9.24887300e-01 -6.36632919e-01
-1.27460051e+00 2.07964063e-01 6.42720819e-01 6.37014806e-01
2.41606623e-01 8.77914369e-01 -9.38724168e-03 1.09065449e+00
8.23656097e-02 3.97458404e-01 1.07654595e+00 -5.94301403e-01
5.54166198e-01 2.46621490e-01 -1.14249974e-01 -3.48462552e-01
-3.45575720e-01 -7.61602163e-01 -4.51266259e-01 9.89883021e-02
6.65694356e-01 -2.61287034e-01 -1.38975537e+00 1.85231113e+00
-1.40350580e-01 1.89206377e-02 1.16708934e-01 4.84742194e-01
5.79711854e-01 8.04728150e-01 3.02838106e-02 -2.15008989e-01
9.77724075e-01 -1.00192463e+00 -8.81270468e-01 -5.82275450e-01
9.57355678e-01 -7.71334291e-01 1.11726892e+00 3.08965743e-01
-9.76391613e-01 -7.09798217e-01 -1.12502038e+00 6.06036894e-02
-4.47648585e-01 1.88756138e-01 1.82906821e-01 7.97180235e-01
-1.14896894e+00 6.11660004e-01 -9.13815856e-01 -4.67124373e-01
1.56912237e-01 3.74895096e-01 -4.76420790e-01 1.42075017e-01
-1.28396893e+00 1.11571074e+00 5.20867109e-01 -2.11257666e-01
-6.00146234e-01 -8.07737291e-01 -6.61159158e-01 1.10282898e-01
3.34922343e-01 -1.83604077e-01 1.72668660e+00 -9.97627079e-01
-1.61986411e+00 6.48084879e-01 -1.58526212e-01 -7.11288095e-01
2.88677186e-01 -1.76822618e-01 -7.61425495e-01 -1.67593241e-01
-4.69028056e-01 5.25500000e-01 8.29639912e-01 -9.53020513e-01
-3.62713307e-01 -2.78923333e-01 -6.26503170e-01 -5.21635078e-02
-5.34675241e-01 1.63748682e-01 -7.44216815e-02 -9.71117020e-01
-7.58149996e-02 -9.87187088e-01 -1.98632509e-01 -5.92281699e-01
-2.67247975e-01 -1.93891898e-01 8.87187123e-01 -1.08818746e+00
1.44113326e+00 -2.39501214e+00 -2.17891056e-02 -5.52405380e-02
-3.70641530e-01 9.69954133e-01 -4.10355389e-01 6.97293162e-01
-4.14297312e-01 2.46583581e-01 -9.96257141e-02 -5.09584308e-01
-7.99002647e-02 2.04635620e-01 -3.43710840e-01 1.12673990e-01
4.35552925e-01 5.88408828e-01 -6.03832304e-01 1.24398395e-01
9.79226157e-02 4.46356744e-01 -4.04536694e-01 3.07134062e-01
-2.48049170e-01 4.17976141e-01 6.35908395e-02 8.84065703e-02
2.45744377e-01 6.11486062e-02 1.91045478e-02 1.19345486e-01
-8.16784352e-02 1.12185609e+00 -8.58261466e-01 1.58601367e+00
-3.90612900e-01 9.36211228e-01 5.37812561e-02 -1.01605642e+00
9.95280206e-01 7.48549521e-01 3.21433187e-01 -6.10618114e-01
2.53591649e-02 5.40156186e-01 5.96242845e-01 -4.55574363e-01
4.30593729e-01 -2.39263490e-01 5.22833705e-01 2.89954305e-01
2.77871639e-01 -1.30473986e-01 1.37222067e-01 -1.78087894e-02
1.23833752e+00 -1.20776100e-02 2.37618864e-01 -2.18912289e-01
3.54160339e-01 -1.05914563e-01 3.40092361e-01 6.86396956e-01
-5.28441695e-03 5.52298248e-01 2.72332221e-01 -3.27717751e-01
-1.51295733e+00 -6.43633425e-01 -2.53565982e-02 1.10288751e+00
-1.02386379e+00 -4.61213946e-01 -9.60736632e-01 -5.15642107e-01
-9.70071927e-02 9.71203685e-01 -4.25706029e-01 -2.20753670e-01
-7.17004955e-01 -6.10129058e-01 9.13001001e-01 7.08391964e-01
2.30083257e-01 -1.00416720e+00 -2.69914687e-01 5.43579280e-01
-1.83278888e-01 -1.35725951e+00 -4.75824326e-01 5.49796581e-01
-9.48499739e-01 -5.30320883e-01 -7.43009984e-01 -5.09547293e-01
3.25842738e-01 4.47364897e-02 9.43431616e-01 -3.35406745e-03
-2.55238619e-02 1.92980930e-01 -5.49213767e-01 -6.25626206e-01
-1.07537889e+00 3.76114845e-01 3.33689779e-01 -1.41795427e-01
3.82542491e-01 -5.76939881e-01 -2.07898557e-01 1.13468193e-01
-7.91083276e-01 -7.03789443e-02 6.70467556e-01 9.09723639e-01
6.99011832e-02 -1.06954433e-01 7.83066869e-01 -5.91883600e-01
5.50571740e-01 -2.96176881e-01 -3.41172606e-01 -1.43849254e-01
-4.82476026e-01 2.88191706e-01 5.66250265e-01 -7.06974506e-01
-8.66272211e-01 2.00456311e-03 -6.33267105e-01 -5.54743886e-01
-3.54571551e-01 5.71938932e-01 -5.29810041e-02 1.61765411e-01
8.11591089e-01 3.62524152e-01 3.55081677e-01 -8.43465745e-01
2.57353842e-01 9.75786448e-01 4.79830176e-01 -2.77619749e-01
5.37974477e-01 -8.92362148e-02 -4.05311286e-01 -1.49277103e+00
-7.10012019e-01 -3.94198805e-01 -4.42307502e-01 4.48602438e-02
4.69645381e-01 -7.97233999e-01 -1.02172613e-01 5.32320678e-01
-1.07957220e+00 -5.64955115e-01 -5.08615911e-01 5.38741708e-01
-2.97054946e-01 2.39079431e-01 -6.01215839e-01 -1.23803306e+00
-2.92198688e-01 -1.07276666e+00 8.12164724e-01 -2.56038576e-01
-3.98638338e-01 -7.74737477e-01 4.71560024e-02 4.71022993e-01
8.12268972e-01 -1.79609284e-01 9.32367086e-01 -1.45553720e+00
-1.90594777e-01 -4.28527921e-01 2.36659735e-01 8.80220652e-01
1.61093578e-01 -7.06188083e-02 -1.39494920e+00 -4.15057421e-01
3.38816673e-01 -2.15179726e-01 7.06704438e-01 3.36162865e-01
1.01701343e+00 -4.81367737e-01 -4.57159020e-02 3.37943852e-01
8.13918650e-01 5.33943355e-01 6.78272247e-01 3.50852400e-01
4.67433274e-01 6.77482009e-01 1.83941588e-01 1.42487839e-01
-4.99381796e-02 8.67745459e-01 -6.56473562e-02 -1.41446948e-01
-3.76652062e-01 -4.16255951e-01 5.44700563e-01 1.33581495e+00
2.55232096e-01 -3.53478551e-01 -1.12184060e+00 6.26971066e-01
-1.56649029e+00 -9.75101113e-01 -1.16656651e-03 2.21748042e+00
8.14259887e-01 5.24170578e-01 5.12792826e-01 6.33174479e-01
5.10990143e-01 1.39170527e-01 -4.88579750e-01 -4.34398681e-01
-1.41830519e-01 2.20752507e-01 3.21318626e-01 4.55579251e-01
-8.62463176e-01 8.17371726e-01 6.94065571e+00 1.02825236e+00
-1.28397465e+00 2.57878546e-02 7.57647455e-01 -2.61432409e-01
-9.05491337e-02 -1.96916997e-01 -7.83534110e-01 5.46784461e-01
1.70256257e+00 2.16032773e-01 5.57819724e-01 6.83753133e-01
4.31118011e-01 2.26060778e-01 -1.16022313e+00 8.06096435e-01
-5.84012233e-02 -1.14222336e+00 -1.39273927e-01 2.40240216e-01
4.94970649e-01 3.82045776e-01 5.42833097e-02 4.85371470e-01
6.88880533e-02 -1.10274374e+00 7.69320190e-01 5.12610413e-02
8.34963918e-01 -3.25688541e-01 6.70762897e-01 5.32320499e-01
-7.87337005e-01 -1.11607864e-01 1.90810010e-01 -7.65558705e-02
2.01215521e-01 3.10961545e-01 -1.20156014e+00 1.89050585e-01
3.30442399e-01 1.65408403e-01 -4.85478908e-01 7.33774781e-01
1.80856183e-01 1.19315684e+00 -5.98734677e-01 -8.54153633e-02
3.64250094e-01 1.37365088e-01 6.78710818e-01 1.45894122e+00
2.02590629e-01 4.18738872e-02 1.84955355e-02 3.62026006e-01
8.69709104e-02 -5.13676144e-02 -6.14367962e-01 -6.20830834e-01
6.53461576e-01 6.67499900e-01 -2.63253450e-01 -5.13265610e-01
-5.90766311e-01 4.85021830e-01 2.98266768e-01 4.27110910e-01
-2.98672020e-01 -2.49020770e-01 7.72205710e-01 4.88327146e-01
3.21876228e-01 -5.00315547e-01 -2.44808793e-01 -8.99767935e-01
1.08596880e-03 -1.37799454e+00 7.94711709e-02 -7.43637741e-01
-9.58248079e-01 8.20592463e-01 -1.36623874e-01 -8.27574193e-01
-7.15767741e-01 -6.94686174e-01 -3.53936732e-01 1.05393481e+00
-1.24460995e+00 -9.80010629e-01 2.28484839e-01 1.49673894e-01
9.08878744e-01 -6.62596643e-01 8.18095088e-01 3.73571604e-01
-3.36540818e-01 7.26829886e-01 2.19683930e-01 1.54023722e-01
4.30921644e-01 -1.10640836e+00 1.16134059e+00 9.70889807e-01
3.56804281e-01 5.92217267e-01 7.88698256e-01 -4.22235727e-01
-1.00921535e+00 -5.85421681e-01 8.67751241e-01 -4.58624274e-01
8.39809239e-01 -4.64844674e-01 -1.26451313e+00 6.61358356e-01
1.48432225e-01 -1.76997945e-01 6.29175127e-01 1.38270020e-01
-5.10162950e-01 7.81375766e-02 -7.57387996e-01 5.25556922e-01
8.52381527e-01 -8.60523224e-01 -6.75163329e-01 -2.78788079e-02
9.06778932e-01 -4.06818062e-01 -6.27744317e-01 4.18967873e-01
3.15998524e-01 -7.07343876e-01 7.45492339e-01 -9.59135056e-01
1.17398031e-01 -5.40627390e-02 -4.00800288e-01 -1.63569510e+00
-1.38481230e-01 -9.58120525e-01 -2.30172902e-01 1.24978948e+00
6.75602674e-01 -4.62191343e-01 7.12131679e-01 5.89843988e-01
-5.38769126e-01 -6.95705414e-01 -9.93372619e-01 -1.10273170e+00
1.17658809e-01 -6.17215514e-01 5.53267181e-01 8.89022171e-01
-7.62054995e-02 5.89788795e-01 -1.78089127e-01 -2.13616267e-01
8.83600637e-02 -7.18448818e-01 7.63831735e-01 -9.83591974e-01
-3.77408653e-01 -5.59095621e-01 -4.38939303e-01 -1.05816019e+00
1.33207208e-02 -6.74820542e-01 1.30974650e-01 -1.15447354e+00
-3.77854228e-01 -2.70839483e-01 -1.66660994e-01 6.47695303e-01
-1.48224503e-01 -2.35926956e-01 3.26217622e-01 -3.94176878e-02
1.97455823e-01 6.64776146e-01 6.99665725e-01 1.16831407e-01
-3.41288000e-01 7.41925016e-02 -4.05308932e-01 6.65935457e-01
9.53301787e-01 -2.24641174e-01 -5.17678201e-01 -4.75149333e-01
-1.72822863e-01 1.87538534e-01 -1.81362685e-03 -1.10092807e+00
2.09332053e-02 3.34026664e-01 1.37464792e-01 -3.67019266e-01
6.21289492e-01 -6.32390082e-01 1.16133122e-02 3.53211701e-01
-6.07517362e-01 7.61118233e-02 7.49063849e-01 3.95081729e-01
-2.12827846e-01 -1.91311672e-01 9.93454635e-01 -5.09230681e-02
-4.26021755e-01 -1.17507279e-01 -7.38557160e-01 2.46047929e-01
2.02398345e-01 -4.95085008e-02 -3.52555931e-01 -5.90697825e-01
-8.04575980e-01 -2.87731975e-01 1.02503940e-01 7.49423802e-01
2.28140295e-01 -1.24810147e+00 -8.25816214e-01 5.18676698e-01
1.80084677e-03 -4.50414687e-01 1.91816725e-02 5.99143505e-01
-1.17961466e-02 7.49590993e-01 1.23436287e-01 -4.43095833e-01
-1.18778062e+00 4.78414297e-01 3.69410276e-01 -1.11827143e-01
-5.29245257e-01 7.93118358e-01 -8.87569115e-02 -5.67165673e-01
4.72887546e-01 -3.64158124e-01 3.54897194e-02 -1.15580879e-01
4.39089984e-01 4.16074187e-01 5.49953103e-01 -5.33468008e-01
-2.80725926e-01 -3.18123065e-02 -3.20457876e-01 -5.61122298e-01
1.28739238e+00 2.60677159e-01 4.54807103e-01 7.50931144e-01
1.10894763e+00 -1.60078511e-01 -1.09236896e+00 -2.73318917e-01
1.62236512e-01 -2.63695508e-01 2.33221725e-01 -1.09031558e+00
-7.19581366e-01 1.01839948e+00 5.75717628e-01 6.69542730e-01
9.49273825e-01 -3.30018401e-01 8.84365082e-01 2.70681024e-01
-3.97079960e-02 -1.25700343e+00 -4.33059111e-02 7.24033415e-01
1.01953137e+00 -1.15835690e+00 -3.42741400e-01 -1.13663755e-01
-6.37052000e-01 1.05131125e+00 4.83451605e-01 1.65858209e-01
5.80507278e-01 5.16970038e-01 2.34868377e-01 1.58314437e-01
-9.78257716e-01 -2.36110017e-01 2.27879316e-01 5.60319960e-01
7.43902266e-01 -1.19726516e-01 1.12976204e-03 4.51554477e-01
-4.26602185e-01 -1.42883420e-01 4.04448539e-01 8.26159537e-01
-4.25579071e-01 -1.30142260e+00 -4.44996446e-01 6.23758674e-01
-5.54819524e-01 -2.95625985e-01 -5.14366329e-01 8.08796585e-01
-3.56568217e-01 1.20784009e+00 3.36993523e-02 -4.49657857e-01
5.05174696e-01 7.05314457e-01 3.40123057e-01 -6.28711879e-01
-6.92287266e-01 3.17975044e-01 5.08913040e-01 -2.25110799e-01
-5.29052503e-03 -8.01256001e-01 -1.02203238e+00 -2.20614806e-01
-5.07618010e-01 1.20530359e-01 1.05586863e+00 1.12678194e+00
4.56444621e-01 8.10190797e-01 4.00784373e-01 -8.14853489e-01
-9.64135945e-01 -1.55305433e+00 -4.17990923e-01 3.77274543e-01
7.84021795e-01 -3.23933274e-01 -5.25986493e-01 6.36556074e-02] | [14.477646827697754, 6.550546169281006] |
722d2a72-b421-417c-b32d-21e7ce221735 | are-character-level-translations-worth-the | 2302.14220 | null | https://arxiv.org/abs/2302.14220v2 | https://arxiv.org/pdf/2302.14220v2.pdf | Are Character-level Translations Worth the Wait? Comparing Character- and Subword-level Models for Machine Translation | Pretrained character-level language models were recently shown to be competitive with popular subword models across a range of NLP tasks. However, there has been little research on their effectiveness for neural machine translation (NMT). This work performs an extensive comparison across multiple languages and experimental conditions of state-of-the-art character- and subword-level pre-trained models (ByT5 and mT5, respectively) on NMT, showing the effectiveness of character-level modeling in translation, particularly in cases where training data is limited. In our analysis, we show how character models' performance gains are reflected in better translations of orthographically similar words and rare words. While evaluating the importance of source texts in driving model predictions, we highlight ByT5 word-level patterns suggesting an ability to modulate word and character-level information during the translation, providing insights into a potential weakness of character-level modeling. We conclude by assessing the efficiency tradeoff of character models, suggesting their usage in non-time-critical scenarios to boost translation quality. | ['Arianna Bisazza', 'Antonio Toral', 'Gabriele Sarti', 'Gertjan van Noord', 'Lukas Edman'] | 2023-02-28 | null | null | null | null | ['nmt'] | ['computer-code'] | [ 5.66560388e-01 -3.22323032e-02 -7.32424080e-01 -1.16804816e-01
-1.08564484e+00 -6.08998477e-01 9.36543226e-01 1.57198235e-01
-5.80240428e-01 7.67730474e-01 5.68515658e-01 -9.80628669e-01
2.67700613e-01 -5.35202503e-01 -1.04272258e+00 -1.15717463e-01
2.95898944e-01 7.79516757e-01 -4.60673809e-01 -5.72239935e-01
3.91743332e-01 3.65349829e-01 -9.33682144e-01 8.11501086e-01
1.16058731e+00 1.28990337e-01 3.37856799e-01 5.38956523e-01
-3.97577375e-01 1.48529559e-01 -5.96712530e-01 -7.23776102e-01
2.66871512e-01 -6.83849514e-01 -5.91111243e-01 -2.69338161e-01
6.33788228e-01 -5.37700057e-02 -1.24178462e-01 7.35051274e-01
5.92238724e-01 -2.88544804e-01 8.13900173e-01 -5.87688982e-01
-1.32916629e+00 9.94043171e-01 -3.68389726e-01 3.90554368e-01
2.44726405e-01 3.71811628e-01 1.30584717e+00 -1.28068292e+00
8.14790130e-01 1.43931413e+00 7.04835355e-01 6.88628554e-01
-1.44335341e+00 -4.04112667e-01 8.27657878e-02 1.11935280e-01
-9.63453412e-01 -7.21607327e-01 1.80360943e-01 -2.33515456e-01
1.82713580e+00 1.63505301e-01 5.30128956e-01 1.40348148e+00
7.14289784e-01 7.93930352e-01 1.35386181e+00 -1.02423596e+00
-1.83222190e-01 2.31882736e-01 -7.84355253e-02 3.09821039e-01
3.53481144e-01 1.57902718e-01 -8.96777034e-01 1.76001317e-03
6.05484962e-01 -5.41252971e-01 -1.94813777e-02 2.61094093e-01
-1.54523623e+00 7.82878458e-01 -3.32288258e-02 5.65738916e-01
-4.27167863e-01 1.29309952e-01 3.70334506e-01 5.67305207e-01
7.28323877e-01 8.20333838e-01 -6.51126683e-01 -4.28486079e-01
-1.23727489e+00 5.68217691e-03 6.46402538e-01 1.23725486e+00
4.50418800e-01 3.50796759e-01 -5.48618972e-01 1.11892366e+00
4.13446464e-02 6.81424856e-01 7.02414215e-01 -3.84747118e-01
9.46478069e-01 3.11602414e-01 4.02299799e-02 -3.13659877e-01
-4.79675420e-02 -6.60410285e-01 -3.77060056e-01 -4.56372410e-01
3.08177471e-01 -9.37740579e-02 -8.81045759e-01 1.91134226e+00
-3.96541089e-01 -3.67178380e-01 1.61975741e-01 5.47057271e-01
2.49060228e-01 1.06363392e+00 2.43459597e-01 -3.77855033e-01
1.30408299e+00 -1.18051934e+00 -5.95225811e-01 -7.19129741e-01
9.46156383e-01 -1.24143469e+00 1.46691549e+00 4.87089604e-02
-1.47976172e+00 -7.51600862e-01 -8.80127668e-01 -2.17597693e-01
-3.96678448e-01 3.48542839e-01 5.85397422e-01 8.14022779e-01
-1.13277793e+00 8.15306127e-01 -6.23846412e-01 -6.66732490e-01
1.77507684e-01 2.36185491e-01 -9.29355547e-02 -2.02458933e-01
-1.42337799e+00 1.60956442e+00 9.41136330e-02 -2.64677964e-02
-7.26779997e-01 -7.12789416e-01 -5.80108166e-01 -3.36668193e-02
-3.46689403e-01 -7.49173820e-01 1.26235235e+00 -9.90418077e-01
-1.52334332e+00 9.23600197e-01 -4.90665168e-01 -4.75488663e-01
4.83610421e-01 -2.93285757e-01 -5.90454996e-01 -2.96587348e-01
7.40917632e-03 9.44487572e-01 5.39954841e-01 -8.87302458e-01
-4.35068905e-01 -8.48623365e-02 -4.26208794e-01 3.53656381e-01
-5.64729273e-01 4.44994301e-01 -2.42856115e-01 -8.43206823e-01
-3.16405684e-01 -9.98856008e-01 -9.34478939e-02 -3.73297453e-01
-1.40743539e-01 -1.54261276e-01 2.07309742e-02 -9.02509451e-01
1.27164388e+00 -1.56662536e+00 2.35375270e-01 -1.98265105e-01
-5.40370643e-01 4.46196973e-01 -7.61032462e-01 1.09650111e+00
1.59293443e-01 5.88396788e-01 -1.74746439e-01 -4.10537153e-01
7.67978430e-02 1.85676232e-01 -4.13332552e-01 2.22232193e-01
6.82855725e-01 1.55179799e+00 -6.06549203e-01 -3.00503165e-01
-1.86217297e-02 4.74811703e-01 -3.87369245e-01 -1.44640937e-01
-4.52150792e-01 2.40371242e-01 4.20789681e-02 6.23183846e-01
2.83847421e-01 1.42767444e-01 2.52088070e-01 3.04756224e-01
-2.81227708e-01 1.04696369e+00 -1.93412542e-01 1.79624975e+00
-7.55246878e-01 8.95370960e-01 -3.16296637e-01 -5.28660119e-01
8.29111159e-01 3.31861258e-01 -1.05476148e-01 -1.08668041e+00
5.85793145e-02 7.25530922e-01 7.17419684e-01 -1.59454286e-01
8.55328202e-01 -4.10990506e-01 1.48794010e-01 6.92894876e-01
5.23692220e-02 -2.57550657e-01 3.11196983e-01 -4.08670828e-02
7.57793307e-01 4.79152083e-01 9.49857458e-02 -6.14444435e-01
9.38783363e-02 4.97799516e-01 1.88659400e-01 8.39270949e-01
1.24438234e-01 5.45803607e-01 5.88831529e-02 -4.97170798e-02
-1.66229236e+00 -6.52638555e-01 -2.03275934e-01 1.21722925e+00
-3.86227340e-01 -5.22770047e-01 -9.70016837e-01 -1.34027719e-01
-1.84114590e-01 1.33959579e+00 -5.28158844e-01 -3.23217183e-01
-1.13246107e+00 -1.02016234e+00 8.00528347e-01 4.27709430e-01
-1.42564580e-01 -1.07597160e+00 -3.26483995e-01 5.10435462e-01
-2.31374398e-01 -1.06825042e+00 -6.02901578e-01 2.24279091e-01
-1.36086512e+00 -1.91422939e-01 -9.50833738e-01 -9.05430496e-01
3.87649477e-01 3.26682001e-01 1.38398039e+00 -2.26125848e-02
-4.43293341e-02 -8.14094488e-03 -3.74502122e-01 -4.63571578e-01
-1.12163424e+00 5.43672502e-01 3.08482736e-01 -5.73013365e-01
6.65807664e-01 -4.61950898e-01 -2.48530716e-01 1.71005666e-01
-6.95791066e-01 2.58990049e-01 1.09322286e+00 9.09359634e-01
3.09398234e-01 -8.70881677e-01 5.05648196e-01 -8.24261427e-01
1.15054262e+00 -3.36633742e-01 -1.38984397e-01 4.50199783e-01
-1.05434525e+00 1.20368153e-01 7.64636338e-01 -7.60957837e-01
-7.93651581e-01 -5.83585322e-01 -1.80021122e-01 1.20152593e-01
2.06520975e-01 7.25682497e-01 1.33550465e-01 1.98299825e-01
8.39531422e-01 6.07410967e-01 2.23578401e-02 -5.06756246e-01
5.98513961e-01 6.20898724e-01 9.81210768e-02 -9.30651903e-01
7.49672174e-01 -1.42720833e-01 -3.52275014e-01 -7.31287003e-01
-4.38332200e-01 8.33141711e-03 -7.88567901e-01 1.95406064e-01
5.65456331e-01 -9.55321670e-01 1.77284807e-01 1.66367039e-01
-1.48079896e+00 -5.41290045e-01 -1.95258379e-01 5.90352118e-01
-6.12155080e-01 1.62420243e-01 -1.14020777e+00 -5.61533749e-01
-5.68561018e-01 -1.18739057e+00 9.71662343e-01 -1.18989617e-01
-6.05220914e-01 -1.10046363e+00 2.02854529e-01 4.35201287e-01
6.37583077e-01 -4.35755491e-01 1.67320943e+00 -7.03828037e-01
-4.25380081e-01 -5.90179153e-02 -1.20177485e-01 2.83600003e-01
-2.71547046e-02 -3.68022285e-02 -6.99668944e-01 -3.19569677e-01
-3.23773205e-01 -2.14314073e-01 6.43743217e-01 2.79923081e-01
1.18773714e-01 -3.43104273e-01 -1.31028667e-01 3.94488990e-01
1.22569227e+00 -6.78878948e-02 6.49058461e-01 5.03172278e-01
5.08848190e-01 6.56223476e-01 4.83297259e-01 -1.77259058e-01
1.03742965e-01 8.22433710e-01 -1.40964866e-01 -1.76292524e-01
-5.93627572e-01 -6.24074996e-01 9.34328377e-01 1.41104102e+00
7.03582019e-02 -5.00209391e-01 -9.50166583e-01 7.05198050e-01
-1.56511819e+00 -5.46156883e-01 -3.39648426e-01 1.99734116e+00
1.15927196e+00 3.83888423e-01 -4.78799418e-02 -4.84096140e-01
6.69317126e-01 -7.92008080e-03 -3.14946204e-01 -1.15126228e+00
-5.57686508e-01 4.84509021e-01 5.98951161e-01 5.94150782e-01
-2.47311339e-01 1.53358352e+00 7.63911676e+00 1.02877676e+00
-1.01641417e+00 3.01159829e-01 5.73387682e-01 -2.03894049e-01
-8.12359452e-01 1.11748308e-01 -9.82226431e-01 3.21347117e-01
1.55845654e+00 -1.86161831e-01 5.51902652e-01 4.87476677e-01
5.03663361e-01 2.36715540e-01 -1.45206392e+00 5.20847201e-01
1.63965434e-01 -1.47916913e+00 5.74096262e-01 2.70205110e-01
1.00348306e+00 3.59252036e-01 1.57217309e-01 5.52563071e-01
1.23525940e-01 -1.20659316e+00 9.86334145e-01 1.10018499e-01
9.45672989e-01 -6.08619750e-01 3.74756902e-01 3.89280081e-01
-5.65088093e-01 2.58906484e-01 -6.98111951e-01 -2.79198021e-01
2.95529634e-01 2.46194333e-01 -9.05913174e-01 3.45349938e-01
9.46747605e-03 4.83394057e-01 -5.87754667e-01 5.07442653e-01
-2.82152086e-01 1.10882533e+00 -3.09003778e-02 -2.86345154e-01
6.12270355e-01 -2.35493928e-01 5.18839955e-01 1.91514826e+00
5.56258023e-01 -2.67675906e-01 -2.98175514e-01 9.49583054e-01
-1.87611058e-01 6.83997691e-01 -5.26992798e-01 -5.55948555e-01
4.43549067e-01 8.04738522e-01 -4.93475914e-01 -2.87984937e-01
-5.01593053e-01 1.31973922e+00 4.85868692e-01 2.59181768e-01
-6.51802957e-01 7.78174251e-02 8.16897810e-01 1.43136561e-01
6.41233325e-02 -6.30929589e-01 -7.50504732e-01 -1.10480702e+00
1.18823551e-01 -1.41480505e+00 -1.18811637e-01 -7.41741538e-01
-1.14731431e+00 7.16371357e-01 -2.14878440e-01 -1.08931911e+00
-2.87299693e-01 -9.66680706e-01 -5.31668067e-01 1.41800594e+00
-1.51168382e+00 -1.34573734e+00 6.98159516e-01 2.78388280e-02
9.59590733e-01 -3.95232379e-01 9.93631423e-01 2.16815352e-01
-4.55597132e-01 1.00844121e+00 5.24312258e-01 -1.22638687e-01
8.69044363e-01 -8.15114796e-01 1.37339544e+00 1.02600145e+00
5.40085733e-01 1.17200017e+00 7.06404626e-01 -8.70229959e-01
-1.70259416e+00 -8.45683873e-01 1.64543128e+00 -7.98801303e-01
7.23541975e-01 -6.26754522e-01 -7.35755682e-01 4.49853897e-01
5.48308611e-01 -7.97128141e-01 7.35524476e-01 2.80852050e-01
-4.44739521e-01 4.27018493e-01 -7.09669828e-01 1.11644530e+00
1.20238769e+00 -7.08335638e-01 -6.87624514e-01 4.01000112e-01
9.73303556e-01 -1.01704590e-01 -7.31816828e-01 2.19763592e-01
5.93618095e-01 -4.56373900e-01 7.73393512e-01 -1.12758708e+00
9.55448270e-01 2.04461724e-01 -2.58431941e-01 -1.63337743e+00
-5.91505051e-01 -7.13964343e-01 4.22627479e-02 1.08693469e+00
1.24869990e+00 -3.89289320e-01 4.20705557e-01 2.94859201e-01
-4.80843902e-01 -8.57832491e-01 -8.75386894e-01 -1.07644248e+00
9.33047414e-01 -4.75077778e-01 3.87265921e-01 8.29643071e-01
1.01101629e-01 6.83027506e-01 -5.56065977e-01 -4.25227910e-01
2.73544323e-02 -1.06570460e-01 4.32428122e-01 -7.20404804e-01
-5.23980379e-01 -8.24356318e-01 6.89800605e-02 -1.14068449e+00
2.21658021e-01 -1.22569120e+00 -9.89453401e-03 -1.72155905e+00
3.00903469e-01 -2.44884595e-01 -9.84171182e-02 2.37782404e-01
-3.77753735e-01 4.00417626e-01 3.01668167e-01 5.69461346e-01
1.95880651e-01 4.71424967e-01 1.33388197e+00 -1.31672323e-01
-7.95729160e-02 -2.07050383e-01 -7.94972122e-01 5.49734198e-02
8.35178733e-01 -5.45881271e-01 -1.87529191e-01 -1.16081071e+00
3.61944497e-01 -2.94390976e-01 -2.68543005e-01 -5.72021008e-01
-1.12657905e-01 -3.69140774e-01 3.93268913e-01 -3.15697551e-01
3.26725334e-01 -3.57665956e-01 -1.68276116e-01 6.40345812e-01
-6.82641804e-01 6.64098442e-01 6.16761208e-01 1.10255174e-01
1.27772227e-01 -2.77480781e-01 5.77757299e-01 -3.33522499e-01
-3.33573520e-01 7.32615367e-02 -8.67878377e-01 1.26979709e-01
3.45740199e-01 -4.03658450e-01 -2.65712649e-01 -3.37630719e-01
-8.70170444e-02 -1.75876960e-01 6.89429045e-01 9.40469503e-01
2.55005449e-01 -1.26804960e+00 -1.27232802e+00 2.71344557e-02
3.39710444e-01 -9.16065276e-01 -2.48761669e-01 7.30357766e-01
-4.94500607e-01 9.42208469e-01 -3.27342480e-01 -3.94270927e-01
-9.16151762e-01 4.74865437e-01 1.41164809e-01 -2.65326589e-01
-3.59924167e-01 8.33905876e-01 -3.11608076e-01 -5.94944894e-01
-8.75764042e-02 -3.00196290e-01 3.58953744e-01 -9.11202431e-02
3.78271997e-01 2.19583318e-01 2.97109514e-01 -6.55507922e-01
-1.84713379e-01 3.31449121e-01 -4.22657937e-01 -6.30461633e-01
1.12585688e+00 -8.85401294e-02 -2.39394099e-01 5.21849394e-01
9.01585460e-01 1.47474855e-01 -6.86261296e-01 -3.03042918e-01
2.58441627e-01 -1.23043686e-01 -3.17008734e-01 -1.22994757e+00
-1.34069830e-01 1.36029196e+00 2.91126609e-01 -5.58281124e-01
6.19027734e-01 -3.35197151e-01 1.01605225e+00 4.34031516e-01
3.62325519e-01 -1.32037330e+00 -2.13058025e-01 9.59328711e-01
7.67894745e-01 -9.98286426e-01 -1.31512418e-01 -2.04165637e-01
-5.92262387e-01 1.12257218e+00 4.09220308e-01 1.00215115e-01
-9.28233042e-02 1.87285319e-01 3.23741436e-01 3.52704018e-01
-1.04014575e+00 1.14180483e-01 3.75873089e-01 4.70018715e-01
1.03689206e+00 2.56325275e-01 -9.04229343e-01 1.98617965e-01
-4.75805193e-01 -3.51615518e-01 3.72847736e-01 7.23440349e-01
-4.72449154e-01 -1.76136935e+00 -2.46330336e-01 3.82539839e-01
-5.09585083e-01 -1.20035136e+00 -7.83061266e-01 6.64737046e-01
-1.30721688e-01 1.01823580e+00 -1.22326598e-01 -4.00118142e-01
1.73313022e-01 6.32265031e-01 7.70966589e-01 -9.30493712e-01
-1.11190307e+00 1.86415091e-01 4.46054965e-01 -1.27705306e-01
1.09437898e-01 -8.07975531e-01 -6.74079478e-01 -4.43761557e-01
-2.70842105e-01 8.41668248e-02 9.22760069e-01 9.85172331e-01
5.24999499e-01 2.64466226e-01 1.54442966e-01 -6.58156037e-01
-7.82854259e-01 -1.38933182e+00 1.27709880e-01 1.49976403e-01
-1.12368278e-01 4.71440516e-02 -3.95068107e-03 -4.55946438e-02] | [11.545110702514648, 10.161425590515137] |
b1b30e7a-5f25-4d32-b21a-9e8c6e052e17 | variational-model-perturbation-for-source | 2210.10378 | null | https://arxiv.org/abs/2210.10378v1 | https://arxiv.org/pdf/2210.10378v1.pdf | Variational Model Perturbation for Source-Free Domain Adaptation | We aim for source-free domain adaptation, where the task is to deploy a model pre-trained on source domains to target domains. The challenges stem from the distribution shift from the source to the target domain, coupled with the unavailability of any source data and labeled target data for optimization. Rather than fine-tuning the model by updating the parameters, we propose to perturb the source model to achieve adaptation to target domains. We introduce perturbations into the model parameters by variational Bayesian inference in a probabilistic framework. By doing so, we can effectively adapt the model to the target domain while largely preserving the discriminative ability. Importantly, we demonstrate the theoretical connection to learning Bayesian neural networks, which proves the generalizability of the perturbed model to target domains. To enable more efficient optimization, we further employ a parameter sharing strategy, which substantially reduces the learnable parameters compared to a fully Bayesian neural network. Our model perturbation provides a new probabilistic way for domain adaptation which enables efficient adaptation to target domains while maximally preserving knowledge in source models. Experiments on several source-free benchmarks under three different evaluation settings verify the effectiveness of the proposed variational model perturbation for source-free domain adaptation. | ['Cees G. M. Snoek', 'Jingjing Li', 'XianTong Zhen', 'Mengmeng Jing'] | 2022-10-19 | null | null | null | null | ['source-free-domain-adaptation'] | ['computer-vision'] | [ 2.44075388e-01 5.53293154e-02 -3.27414155e-01 -4.77852583e-01
-1.11873674e+00 -8.64174306e-01 4.98234481e-01 -3.39835793e-01
-4.41086978e-01 8.10399115e-01 1.80089831e-01 1.13047846e-01
-1.44787222e-01 -7.30646908e-01 -9.21371698e-01 -7.86292255e-01
3.31881642e-01 7.23317325e-01 4.19669122e-01 -1.68050945e-01
-3.22415568e-02 2.80392170e-01 -9.90713954e-01 -7.78632239e-02
9.70689356e-01 7.91254222e-01 3.93134236e-01 3.34215403e-01
-9.35467482e-02 2.72303224e-01 -6.21254504e-01 -3.82011503e-01
3.71293575e-01 -1.28782034e-01 -5.14727712e-01 8.90507847e-02
3.68270099e-01 -3.59869331e-01 -3.58369887e-01 1.16831875e+00
3.75779510e-01 3.38666737e-01 9.93833363e-01 -1.13832521e+00
-9.20522809e-01 5.50960600e-01 -3.60969216e-01 -9.48097855e-02
-1.78898452e-03 -9.77169275e-02 9.42408860e-01 -7.64755785e-01
4.40168530e-01 1.44842899e+00 4.91923094e-01 7.74565279e-01
-1.44570291e+00 -8.10115993e-01 5.34939826e-01 1.16052106e-01
-1.45034552e+00 -5.08313477e-01 1.02472186e+00 -4.44175929e-01
4.75998133e-01 -2.52192467e-01 -1.73985660e-02 1.55914283e+00
-5.37542962e-02 6.57218754e-01 7.64788389e-01 -3.71368945e-01
5.52737355e-01 5.06651163e-01 7.73002431e-02 3.13210309e-01
1.97566047e-01 1.00095689e-01 -5.84483981e-01 -4.79962736e-01
7.23353148e-01 7.47331604e-02 -2.55385041e-01 -9.76750970e-01
-1.17230546e+00 8.09802949e-01 1.99793890e-01 -3.22987065e-02
-1.97350904e-01 6.46531284e-02 3.74646395e-01 6.40387684e-02
4.47963923e-01 2.89127588e-01 -8.22415650e-01 1.42212793e-01
-7.16639876e-01 3.75140309e-01 8.75090122e-01 1.18089128e+00
9.01922882e-01 5.36562838e-02 -3.96176040e-01 1.12843716e+00
5.42530298e-01 1.00534463e+00 4.42228794e-01 -1.11673069e+00
4.87958789e-01 3.56908500e-01 3.09958667e-01 -6.92762673e-01
6.62695803e-03 -4.93379563e-01 -7.70231664e-01 -1.73786748e-02
5.80899358e-01 -1.69203818e-01 -9.41894233e-01 2.28005433e+00
4.94797438e-01 1.85712770e-01 1.73253551e-01 6.45997941e-01
2.25737274e-01 6.60496891e-01 1.21828221e-01 2.97560357e-02
1.12125444e+00 -7.32939005e-01 -5.23998439e-01 -3.64592850e-01
3.91064674e-01 -4.60721642e-01 1.35996294e+00 2.25462243e-01
-6.73425555e-01 -4.89397168e-01 -1.05183005e+00 1.03527106e-01
-2.47534528e-01 9.57007613e-03 1.43978909e-01 6.24921203e-01
-6.56551540e-01 3.32080036e-01 -9.11200225e-01 -2.46944889e-01
5.66506565e-01 4.65119869e-01 -3.41342121e-01 -2.71998495e-01
-1.37804449e+00 7.09459484e-01 7.72543728e-01 -3.30257893e-01
-1.10098052e+00 -9.41749334e-01 -9.16801870e-01 1.18125662e-01
4.83097404e-01 -7.98683763e-01 1.30294073e+00 -8.71355116e-01
-1.98522270e+00 4.71467882e-01 -3.84336054e-01 -4.14987892e-01
3.54905099e-01 -1.76723614e-01 -2.58904338e-01 -4.34372835e-02
6.69211224e-02 5.69165885e-01 1.18515480e+00 -1.26790500e+00
-6.66939616e-01 -3.19841862e-01 8.92050713e-02 2.47376382e-01
-7.68167675e-01 -3.62792552e-01 -7.35280037e-01 -6.67843163e-01
-1.03912093e-01 -1.11467993e+00 -7.79832825e-02 -1.93340614e-01
-3.42970639e-01 -2.81742722e-01 8.04675102e-01 -5.66508710e-01
1.04982710e+00 -2.29805398e+00 3.07234228e-01 3.98728222e-01
2.48748422e-01 1.26619324e-01 -3.46433103e-01 4.57074977e-02
2.14544639e-01 -1.41062230e-01 -5.13020813e-01 -4.91654038e-01
3.30150634e-01 5.19139767e-01 -6.61045313e-01 3.51029932e-01
1.69409856e-01 7.96165466e-01 -7.57218480e-01 -4.70588952e-01
-6.63377286e-04 4.85494316e-01 -8.76183689e-01 4.03458744e-01
-4.44548398e-01 4.67822999e-01 -6.89375281e-01 3.79720986e-01
8.37921023e-01 -3.56083333e-01 2.57770777e-01 -1.02029100e-01
6.60304844e-01 2.79062390e-01 -1.25938010e+00 1.87174404e+00
-5.82767189e-01 1.05834126e-01 1.72229648e-01 -1.09029174e+00
9.33862746e-01 6.75659254e-02 2.77769566e-01 -3.65665108e-01
-3.05243820e-01 1.23942778e-01 -2.64991373e-01 -7.26295784e-02
2.58416593e-01 -2.40267202e-01 -3.65722090e-01 3.02530587e-01
3.16934496e-01 3.41395400e-02 -8.59034881e-02 1.99331269e-01
7.28975713e-01 4.49746162e-01 1.59304425e-01 -2.47273400e-01
4.17222649e-01 -6.44346327e-02 9.04557943e-01 8.56376052e-01
-2.43778467e-01 4.79476810e-01 3.38982970e-01 -2.45135427e-02
-9.34435785e-01 -1.56101704e+00 -2.35914260e-01 1.48558402e+00
1.13057196e-01 -1.06824927e-01 -7.49090433e-01 -1.04298556e+00
1.37901157e-01 8.41983557e-01 -6.31890178e-01 -4.96973276e-01
-5.68648517e-01 -8.15699399e-01 5.02232194e-01 4.60817456e-01
4.38831896e-01 -5.47738373e-01 -2.84977034e-02 2.65878379e-01
-3.04930389e-01 -1.15490174e+00 -8.14516783e-01 2.09304795e-01
-8.85328829e-01 -7.10565150e-01 -9.51210201e-01 -5.80323160e-01
5.58645070e-01 -5.68655357e-02 9.95007217e-01 -7.94835567e-01
4.32855666e-01 6.02962971e-01 -8.28401148e-02 -3.14044595e-01
-6.57079518e-01 4.75674421e-01 2.95824528e-01 -7.62262521e-03
5.36750495e-01 -7.32980132e-01 -3.45767021e-01 4.07047600e-01
-9.72884357e-01 -4.64820147e-01 5.82283854e-01 9.88035619e-01
6.26269639e-01 4.60025594e-02 6.68596029e-01 -1.05253243e+00
4.43330824e-01 -6.53056681e-01 -8.00293803e-01 3.41303110e-01
-6.26132429e-01 4.35766041e-01 7.43672431e-01 -8.52280855e-01
-1.41637123e+00 1.34358332e-01 7.66320303e-02 -5.69594145e-01
-2.62831986e-01 2.90312111e-01 -5.78134000e-01 1.52774319e-01
9.59080338e-01 3.20632190e-01 5.57509856e-03 -6.60438359e-01
6.00583613e-01 6.08718872e-01 5.45829475e-01 -1.05868566e+00
1.10242689e+00 5.08618414e-01 -1.95693269e-01 -4.48974729e-01
-9.70880747e-01 -2.67286271e-01 -7.36188650e-01 4.40888762e-01
5.49679339e-01 -1.10236108e+00 -3.40774387e-01 3.78791928e-01
-1.11620963e+00 -4.83851582e-01 -3.63937050e-01 4.15215582e-01
-5.47483504e-01 5.13903975e-01 -2.20209673e-01 -4.54469532e-01
-5.95399812e-02 -1.05412149e+00 1.05882406e+00 5.48192561e-02
-6.25860542e-02 -1.45920300e+00 3.31338942e-01 1.21010780e-01
5.03372967e-01 -1.34193301e-01 9.88105714e-01 -1.05767226e+00
-4.40209538e-01 -2.36009713e-02 -1.19028963e-01 6.54670238e-01
3.38954568e-01 -3.66993576e-01 -1.01440537e+00 -3.53876889e-01
5.68509521e-03 -2.54163831e-01 9.01742399e-01 5.25231659e-01
1.02624369e+00 -4.07503307e-01 -5.17374754e-01 7.76581407e-01
1.25782394e+00 -6.45044595e-02 2.56483823e-01 3.38071406e-01
8.53625417e-01 4.66175020e-01 4.41318840e-01 4.61278349e-01
5.50367594e-01 9.05893266e-01 1.39195889e-01 1.84213609e-01
-6.56474680e-02 -5.40948927e-01 6.68506205e-01 5.15327692e-01
5.48848093e-01 -9.95540693e-02 -9.41872239e-01 6.87234998e-01
-1.83285701e+00 -5.62189400e-01 5.53789854e-01 2.42663264e+00
1.19631147e+00 7.09708706e-02 2.32153699e-01 -5.47044635e-01
7.43515193e-01 -4.40431572e-02 -1.02365136e+00 1.59092575e-01
2.51025669e-02 7.22857378e-03 5.76750159e-01 6.33878946e-01
-1.07331121e+00 1.06128013e+00 6.77965927e+00 1.14117146e+00
-1.06374645e+00 3.39538187e-01 2.74824589e-01 -2.05632061e-01
-4.47296023e-01 5.22666797e-02 -1.17736280e+00 6.24716878e-01
1.00487983e+00 -3.66867125e-01 5.87378204e-01 1.26482058e+00
-3.47380154e-03 4.92576212e-01 -1.20065522e+00 7.08782971e-01
-1.16176546e-01 -9.83560443e-01 2.79736251e-01 6.56159669e-02
7.17255116e-01 1.13219969e-01 1.72784209e-01 6.44770682e-01
7.32557893e-01 -7.25597560e-01 6.38394237e-01 4.21113878e-01
8.04759324e-01 -7.32982159e-01 4.31222886e-01 4.34406042e-01
-8.10958982e-01 -6.96814954e-02 -8.17038536e-01 3.56358260e-01
1.50953960e-02 8.32514644e-01 -1.01166189e+00 3.53871912e-01
6.19540572e-01 7.57608294e-01 -3.49151582e-01 5.21056414e-01
-1.39227226e-01 6.92368209e-01 -5.47895372e-01 3.59881669e-01
-2.94032067e-01 -1.63798243e-01 7.88511872e-01 1.19690859e+00
3.29110295e-01 -3.99299085e-01 3.51247698e-01 1.23735559e+00
-2.15426326e-01 -1.39538541e-01 -6.59167707e-01 6.66255876e-02
9.60353315e-01 9.03281569e-01 -2.60641966e-02 -3.70782703e-01
-2.44786635e-01 1.03533006e+00 4.94447559e-01 9.08670187e-01
-9.42360580e-01 -3.33673149e-01 9.04863596e-01 -7.05921650e-02
4.60602582e-01 -1.72783941e-01 -2.27448002e-01 -1.44477153e+00
1.55833468e-01 -9.76219475e-01 3.74419123e-01 -2.19851449e-01
-1.70005190e+00 4.37146574e-01 4.46457505e-01 -1.08148324e+00
-4.21998382e-01 -5.70915997e-01 -2.71880299e-01 1.11304855e+00
-1.64223778e+00 -1.16371250e+00 -2.15122513e-02 9.25557852e-01
2.46648416e-01 -3.87741178e-01 7.54177451e-01 1.22591086e-01
-5.32500982e-01 1.04030371e+00 5.42111218e-01 -2.73191296e-02
1.20609748e+00 -1.37894773e+00 3.01971495e-01 8.51800978e-01
-7.89776519e-02 9.18886065e-01 6.46984518e-01 -4.50826108e-01
-1.15852642e+00 -1.34041572e+00 4.55414861e-01 -7.80123472e-01
5.57238519e-01 -5.70613742e-01 -1.17021859e+00 8.25483561e-01
-1.29074767e-01 -8.04165229e-02 7.74687529e-01 2.68126994e-01
-8.15523565e-01 -3.47071767e-01 -1.21136189e+00 5.24180412e-01
8.49351168e-01 -6.04704857e-01 -6.85155988e-01 2.46210933e-01
1.04141593e+00 -3.64616066e-01 -9.77684736e-01 2.92541146e-01
4.29908007e-01 -5.46697199e-01 1.25423729e+00 -6.82109773e-01
8.80578384e-02 -3.70841980e-01 -5.48666120e-01 -1.51541030e+00
-4.58792329e-01 -5.12234509e-01 -4.46094871e-01 1.51803899e+00
4.18259382e-01 -8.65688503e-01 7.75841475e-01 8.24767649e-01
1.03185080e-01 -2.45037600e-01 -1.04948425e+00 -1.08712804e+00
6.99128687e-01 -4.70959485e-01 6.59906447e-01 8.10568511e-01
-3.61344039e-01 2.88754970e-01 -4.03425485e-01 4.81049299e-01
9.52962637e-01 -7.47275427e-02 9.04080272e-01 -1.18794405e+00
-6.93399012e-01 -2.96599090e-01 3.33456621e-02 -1.44802940e+00
5.70839345e-01 -7.36728430e-01 1.65561080e-01 -1.13499963e+00
3.49228978e-01 -4.95698392e-01 -6.11425757e-01 5.17815709e-01
-4.65722620e-01 -1.45330310e-01 4.25947383e-02 5.19120276e-01
-5.59045196e-01 8.66598010e-01 1.04269314e+00 -1.56432584e-01
-3.71338099e-01 6.89655095e-02 -1.05590153e+00 6.18621171e-01
6.63341701e-01 -8.27509105e-01 -7.48566091e-01 -4.12652791e-01
-9.93763953e-02 -4.12609994e-01 2.73045778e-01 -6.69816136e-01
6.42959550e-02 -3.49138886e-01 3.21435452e-01 -6.06949031e-02
3.08858663e-01 -8.92840922e-01 -1.88597366e-01 1.44094964e-02
-4.06935513e-01 -5.65895379e-01 2.82303184e-01 9.68370676e-01
-1.00121185e-01 -1.61921725e-01 1.22775400e+00 2.03595266e-01
-7.37063408e-01 3.33927900e-01 -1.46347601e-02 3.16359371e-01
8.40389609e-01 3.61795276e-02 -1.81201249e-01 -3.12208235e-01
-7.45802939e-01 1.74579412e-01 6.90722167e-01 3.69042218e-01
3.03598195e-01 -1.48546970e+00 -7.01128662e-01 2.44245335e-01
3.28729928e-01 1.92420900e-01 2.57790953e-01 4.96040881e-01
1.29925966e-01 4.19187158e-01 -3.50179989e-03 -7.96207607e-01
-6.68683589e-01 6.29369080e-01 5.03526986e-01 -4.69657779e-01
-3.24917883e-01 7.71545708e-01 7.65962899e-01 -9.49842691e-01
2.98110962e-01 -1.72457621e-01 8.04131553e-02 -3.72327149e-01
3.61130297e-01 2.17758849e-01 -1.83765933e-01 -2.52680898e-01
-4.39560771e-01 5.26446700e-01 -3.29543978e-01 -3.01970512e-01
1.12011743e+00 -4.11628962e-01 2.17463538e-01 3.70439649e-01
1.17900681e+00 3.32664028e-02 -1.80566132e+00 -8.32749903e-01
-4.30309810e-02 -2.86147445e-01 -1.28542231e-02 -8.41737688e-01
-7.25626469e-01 8.14904749e-01 6.22610509e-01 -2.40110770e-01
1.09221399e+00 2.90821064e-02 6.87917948e-01 6.75455034e-01
3.37862968e-01 -1.12489402e+00 8.50330889e-02 6.09984577e-01
7.39040494e-01 -1.33805180e+00 -3.68827790e-01 -1.02910921e-01
-7.78672040e-01 8.34319532e-01 5.27940929e-01 -1.32310078e-01
7.52766013e-01 7.26036206e-02 -2.03246474e-02 2.31510788e-01
-6.71502650e-01 1.00012071e-01 4.63534415e-01 1.03396308e+00
6.68349266e-02 -5.75088598e-02 4.32049870e-01 9.44802940e-01
8.14932436e-02 3.92146409e-02 1.83838636e-01 4.84123260e-01
-3.68879199e-01 -1.48948348e+00 -5.15514731e-01 1.84592977e-01
-2.76014805e-01 -8.38409811e-02 -2.00426102e-01 7.47809410e-01
-1.72368601e-01 8.16577733e-01 -1.64452329e-01 -1.24889456e-01
3.74187678e-01 3.71333659e-01 3.22380781e-01 -7.56525278e-01
1.37854680e-01 1.25281975e-01 -3.27536941e-01 -4.41074133e-01
-7.78382272e-02 -6.20134592e-01 -9.97322083e-01 -1.67596951e-01
-8.20405111e-02 1.12278573e-01 4.66150343e-01 1.00369942e+00
7.89772034e-01 4.00108576e-01 5.03292561e-01 -6.23745441e-01
-1.26313519e+00 -9.76715207e-01 -6.73648536e-01 5.01703441e-01
4.27938133e-01 -9.29463327e-01 -4.13441986e-01 2.69400060e-01] | [10.334623336791992, 3.1874454021453857] |
efc42e50-74da-4f62-81f6-99b1ebcc9136 | on-designing-machine-learning-models-for | 1907.04846 | null | https://arxiv.org/abs/1907.04846v1 | https://arxiv.org/pdf/1907.04846v1.pdf | On Designing Machine Learning Models for Malicious Network Traffic Classification | Machine learning (ML) started to become widely deployed in cyber security settings for shortening the detection cycle of cyber attacks. To date, most ML-based systems are either proprietary or make specific choices of feature representations and machine learning models. The success of these techniques is difficult to assess as public benchmark datasets are currently unavailable. In this paper, we provide concrete guidelines and recommendations for using supervised ML in cyber security. As a case study, we consider the problem of botnet detection from network traffic data. Among our findings we highlight that: (1) feature representations should take into consideration attack characteristics; (2) ensemble models are well-suited to handle class imbalance; (3) the granularity of ground truth plays an important role in the success of these methods. | ['Tina Eliassi-Rad', 'Timothy Sakharaov', 'Alina Oprea', 'Talha Ongun', 'Simona Boboila'] | 2019-07-10 | null | null | null | null | ['traffic-classification'] | ['miscellaneous'] | [ 1.50553569e-01 -3.22110325e-01 -3.96074444e-01 -2.05137089e-01
-3.86517256e-01 -6.99668705e-01 6.58130944e-01 4.00086641e-01
-2.56983519e-01 5.80080330e-01 -1.04843117e-01 -9.51077700e-01
-4.61974353e-01 -8.06593359e-01 -2.46958911e-01 -4.96800542e-01
-2.39231855e-01 3.11413735e-01 1.07185416e-01 -1.80491313e-01
5.82690656e-01 8.92145574e-01 -1.36478138e+00 2.33050480e-01
5.05904019e-01 7.71784782e-01 -5.26409864e-01 7.61055589e-01
1.51300412e-02 9.66817439e-01 -1.07847846e+00 -4.66453999e-01
4.79281366e-01 -1.18402883e-01 -7.96643257e-01 -8.58649015e-02
9.14343074e-02 -4.51353282e-01 -3.03712487e-01 8.32812071e-01
3.06903601e-01 -2.26443201e-01 6.10641003e-01 -1.71204746e+00
-5.91145568e-02 4.17663634e-01 -4.34826136e-01 6.92817152e-01
7.93262944e-02 5.21428645e-01 1.14004672e+00 -2.88638204e-01
5.33614039e-01 1.06280375e+00 7.19970226e-01 3.50603849e-01
-1.31706464e+00 -7.92412460e-01 2.77784765e-01 2.94769943e-01
-9.18319821e-01 -3.02457362e-01 5.45955181e-01 -6.42200172e-01
1.21204293e+00 2.14703634e-01 4.35987294e-01 1.37511027e+00
3.48842323e-01 2.45259061e-01 1.31650865e+00 -3.39349717e-01
2.22024709e-01 3.83174658e-01 3.49178761e-01 2.32283249e-01
8.12524140e-01 3.05969685e-01 -4.20810543e-02 -6.96223617e-01
5.25901198e-01 2.12535709e-01 2.79153615e-01 9.23867673e-02
-7.08817899e-01 1.28299022e+00 2.21387073e-01 5.29723346e-01
-4.91549134e-01 6.27894700e-02 7.58705080e-01 5.75435758e-01
6.18190587e-01 8.42251837e-01 -5.26660025e-01 -2.70254046e-01
-6.70682192e-01 2.58191288e-01 9.14239824e-01 1.26535296e-01
5.88017344e-01 2.81630635e-01 3.56122583e-01 2.63575643e-01
9.21876878e-02 2.80418694e-01 1.19972363e-01 -6.85297430e-01
3.03782642e-01 6.59508526e-01 -7.25917295e-02 -1.22455168e+00
-3.35190952e-01 -4.73930359e-01 -3.40808362e-01 3.89715970e-01
5.98464191e-01 -3.43552113e-01 -4.85343575e-01 1.37478113e+00
1.09207556e-01 4.80464071e-01 -1.98641464e-01 3.00985247e-01
2.18235493e-01 3.57092112e-01 5.07528543e-01 -1.03990160e-01
9.64385509e-01 -5.45563661e-02 -4.41274613e-01 -2.65215099e-01
8.01940501e-01 -5.88750720e-01 6.51627600e-01 5.77549279e-01
-3.18431586e-01 -2.10120939e-02 -8.94235253e-01 8.86567652e-01
-8.27871799e-01 -3.28162402e-01 8.70846093e-01 1.30259931e+00
-5.12083590e-01 6.67633235e-01 -6.38262868e-01 -5.46355963e-01
3.04125428e-01 3.52441251e-01 -2.79958338e-01 6.94074333e-02
-1.10241699e+00 1.20045567e+00 3.31585854e-01 -2.94240832e-01
-8.43894780e-01 -5.03770828e-01 -3.36643994e-01 7.26790226e-04
4.37491685e-01 -2.25565881e-02 7.90787160e-01 -8.31176460e-01
-1.02227521e+00 6.30599082e-01 3.83465677e-01 -6.42139614e-01
1.68017611e-01 -6.38725013e-02 -6.93513572e-01 1.48803011e-01
-7.62338862e-02 -9.57319066e-02 8.97761941e-01 -1.33013141e+00
-7.39731133e-01 -2.92999923e-01 4.46014017e-01 -5.01108587e-01
-3.85738045e-01 5.97383380e-01 6.40829563e-01 -5.53109169e-01
-2.66937613e-01 -7.12656260e-01 -4.50887889e-01 -7.12297142e-01
-2.59615541e-01 -8.19351822e-02 1.33761215e+00 -6.90647840e-01
1.32373750e+00 -1.81339359e+00 -3.94938111e-01 5.33124745e-01
3.09195727e-01 7.30843365e-01 -2.79876795e-02 8.17979276e-01
-1.22720890e-01 6.96368933e-01 5.27429022e-02 4.35643308e-02
-2.49433890e-02 -2.32507028e-02 -7.64875114e-01 5.77362061e-01
4.80841607e-01 4.05085742e-01 -7.12042630e-01 -2.36319050e-01
5.41271627e-01 3.64767939e-01 -5.16739964e-01 1.27273366e-01
-8.13084245e-02 4.81864125e-01 -6.00490808e-01 9.60302114e-01
4.62838978e-01 -2.32577652e-01 3.66195917e-01 1.52708692e-02
-1.22157231e-01 5.10773420e-01 -1.02730989e+00 4.66031998e-01
-3.64205480e-01 6.65989876e-01 -2.10556891e-02 -9.54055369e-01
8.43503237e-01 2.60728598e-01 7.27009416e-01 -3.06557536e-01
4.42167997e-01 2.39203259e-01 4.15705502e-01 -3.61026675e-01
6.70781732e-02 1.29235044e-01 8.50015432e-02 1.04481113e+00
-3.31166461e-02 1.62768587e-02 1.22374289e-01 7.08497837e-02
1.49013317e+00 -4.56041038e-01 5.72474658e-01 3.71291637e-02
3.03236395e-01 2.17488483e-01 7.82799006e-01 7.60440469e-01
-4.65784311e-01 -4.21686359e-02 7.15755463e-01 -6.79679573e-01
-9.55608368e-01 -7.96969533e-01 -6.80801794e-02 1.03402984e+00
-3.50980282e-01 -5.01972258e-01 -8.42481852e-01 -1.15268445e+00
5.75156175e-02 7.59109795e-01 -4.03237224e-01 -1.10448860e-01
-5.93626559e-01 -1.02717519e+00 6.99142694e-01 2.57288873e-01
1.07171312e-01 -1.04183757e+00 -8.66333187e-01 2.10776731e-01
9.15196016e-02 -1.32921183e+00 2.76607931e-01 1.91902727e-01
-8.17997098e-01 -1.83017349e+00 4.42168206e-01 9.00019482e-02
4.80585784e-01 4.58545297e-01 9.42714512e-01 4.74319071e-01
-5.22735715e-01 5.82400322e-01 -7.46848583e-01 -7.46630490e-01
-6.19677186e-01 1.42310351e-01 2.43035764e-01 1.27437934e-02
6.68527722e-01 -6.61661446e-01 -1.40594348e-01 1.92044377e-01
-9.75591302e-01 -8.68547440e-01 7.23725736e-01 5.96722424e-01
-1.08570136e-01 4.98204887e-01 9.50718284e-01 -1.06298757e+00
9.68764782e-01 -9.30392921e-01 -5.14407456e-01 2.17416644e-01
-9.02059734e-01 -4.46207821e-01 5.69183528e-01 -5.05182326e-01
-7.23395050e-01 -4.90682125e-01 -7.41154179e-02 -3.03152502e-01
-5.20682931e-01 5.10442019e-01 3.02824192e-02 -3.92561436e-01
9.49772060e-01 -2.02845857e-01 1.95739925e-01 -4.01742578e-01
-6.80547357e-02 7.87630618e-01 -1.72982812e-01 -6.42402768e-01
1.04097235e+00 3.71114731e-01 -4.02071290e-02 -9.90654647e-01
-5.26051700e-01 -3.27187568e-01 -6.04784906e-01 -3.32095832e-01
3.98520797e-01 -4.35512781e-01 -6.60588861e-01 2.36546189e-01
-8.26693296e-01 -3.00087571e-01 -4.01039235e-02 5.94812632e-01
-1.65854841e-01 4.33820128e-01 -5.68842292e-01 -1.16901231e+00
-2.62024432e-01 -1.19410503e+00 3.59555870e-01 1.24750026e-01
-2.88754225e-01 -1.13220382e+00 1.79484010e-01 3.96759808e-01
6.70385242e-01 6.58075392e-01 1.03025711e+00 -1.39779699e+00
-3.18377644e-01 -6.99874878e-01 -1.36524230e-01 5.53184330e-01
2.08898529e-01 4.81274962e-01 -1.16253066e+00 -4.64969069e-01
-4.82803211e-02 -2.20704839e-01 5.64958334e-01 2.05015734e-01
9.47687507e-01 -4.44846362e-01 -2.46978760e-01 1.32470667e-01
1.28274715e+00 2.69783825e-01 3.89525294e-01 5.06500125e-01
3.46005857e-01 1.02499962e+00 5.73179781e-01 7.16969788e-01
4.58508171e-02 5.10701776e-01 6.45110190e-01 2.47456864e-01
1.53142542e-01 -2.17671767e-01 4.79695082e-01 3.84877175e-01
-3.27912383e-02 -4.17388976e-02 -1.14368534e+00 2.50622660e-01
-1.51599395e+00 -1.26945043e+00 5.39989099e-02 2.26924610e+00
2.21750990e-01 5.12509823e-01 6.14187479e-01 6.08168781e-01
7.07712412e-01 1.61556140e-01 -3.35664898e-01 -4.80943978e-01
9.64989811e-02 3.17950189e-01 5.89201689e-01 2.39968091e-01
-1.32579255e+00 7.73552656e-01 6.76028824e+00 6.76696360e-01
-1.14328122e+00 2.04689756e-01 6.19581938e-01 7.69822374e-02
1.02011658e-01 3.41294438e-01 -6.76000416e-01 3.79174590e-01
1.31782508e+00 4.72500585e-02 3.81339639e-01 8.10132980e-01
1.58569679e-01 2.63062656e-01 -6.51232600e-01 4.01907414e-01
-1.51380152e-01 -1.26297498e+00 -1.20054573e-01 5.17444670e-01
3.02166462e-01 2.61671990e-02 1.48812965e-01 3.63738269e-01
6.22358680e-01 -1.13483036e+00 3.26134562e-01 1.81672022e-01
4.27476823e-01 -8.90090048e-01 8.39321375e-01 4.05058771e-01
-8.08073580e-01 -6.02195323e-01 -2.26577953e-01 -2.53353745e-01
8.67533684e-02 7.35653877e-01 -1.09269822e+00 4.18380439e-01
5.86290598e-01 4.95632559e-01 -6.40291512e-01 1.05126989e+00
-2.18329862e-01 1.22044528e+00 -1.52676985e-01 4.39857831e-03
3.95207584e-01 6.58239871e-02 6.24731243e-01 1.15290916e+00
-2.60462742e-02 -7.91641921e-02 3.64596695e-01 5.41020632e-01
4.36508089e-01 -3.66704352e-02 -9.25954401e-01 -6.62976921e-01
7.62749672e-01 1.40542078e+00 -9.26034391e-01 1.31909996e-01
-4.07938153e-01 6.76778890e-03 7.18096867e-02 2.57946044e-01
-6.42025709e-01 -4.06810522e-01 1.26212335e+00 3.23000550e-01
-2.12159246e-01 -2.73413569e-01 -5.62616706e-01 -1.01096833e+00
-5.53350389e-01 -1.24895656e+00 6.16237283e-01 1.19800769e-01
-1.50268650e+00 6.76605880e-01 8.57779905e-02 -1.26202071e+00
-4.31130439e-01 -8.35325718e-01 -8.60252619e-01 4.87876832e-01
-1.11257410e+00 -1.07570410e+00 -1.65047478e-02 2.57997662e-01
1.67286575e-01 -5.85459411e-01 1.00723064e+00 1.68433264e-01
-8.84593070e-01 2.96985388e-01 -1.86792642e-01 2.34974638e-01
4.79516774e-01 -8.62225294e-01 5.14498174e-01 9.80279863e-01
2.15552956e-01 8.38733554e-01 8.50658536e-01 -9.07081962e-01
-1.13610840e+00 -8.71009588e-01 6.24374509e-01 -7.35870957e-01
9.61675525e-01 -2.76787162e-01 -7.27749765e-01 7.99021721e-01
-8.30265060e-02 -1.96425736e-01 1.11506557e+00 3.76564741e-01
-8.37494969e-01 -7.32264889e-04 -1.74303722e+00 3.28459144e-01
3.97348911e-01 -5.97894013e-01 -3.00841808e-01 4.41420048e-01
3.00601274e-01 2.83423752e-01 -8.93393338e-01 3.58013421e-01
4.06298816e-01 -1.04246211e+00 9.78057444e-01 -1.08371115e+00
-2.10640028e-01 -2.21835990e-02 -1.64597049e-01 -1.03596127e+00
-2.71090806e-01 -5.00417352e-01 -1.63874924e-01 1.21681690e+00
1.87326238e-01 -1.10319066e+00 7.93634117e-01 5.40302038e-01
3.77327353e-01 -6.76385343e-01 -8.48471820e-01 -7.52597749e-01
5.78505434e-02 -6.15223408e-01 7.24232018e-01 1.16175377e+00
5.01397848e-02 -1.41273990e-01 -4.18080121e-01 2.04163849e-01
7.93052733e-01 -2.35549182e-01 8.46872985e-01 -1.69158411e+00
-2.42730066e-01 -5.85964262e-01 -6.60974860e-01 1.62767425e-01
4.74459887e-01 -4.40336198e-01 -6.76397204e-01 -1.07540870e+00
4.93966565e-02 -7.20663905e-01 -3.43957692e-01 4.87108260e-01
6.49581328e-02 7.49807060e-02 2.60882437e-01 1.69257641e-01
-1.08039670e-01 -1.96815938e-01 3.01154554e-01 3.68654393e-02
5.36750406e-02 3.38489145e-01 -7.10648656e-01 7.77885258e-01
1.39825380e+00 -6.50134921e-01 -4.26050514e-01 1.19255826e-01
-2.21069865e-02 -2.71243095e-01 4.83750224e-01 -7.89510787e-01
-3.70226987e-02 -6.16158962e-01 2.24560201e-01 -2.11416706e-01
2.51279086e-01 -9.03189659e-01 5.46381325e-02 7.88926661e-01
-1.73434895e-02 2.94396222e-01 4.55747098e-02 4.88610715e-01
-7.20110163e-02 -3.63582224e-01 7.28208363e-01 -7.04196617e-02
-5.48506498e-01 3.02390397e-01 -6.70507610e-01 7.94994235e-02
1.19289958e+00 -2.07387313e-01 -5.99748492e-01 -4.45620179e-01
-2.73663431e-01 -3.03367764e-01 3.57557356e-01 4.04196680e-01
5.08226156e-01 -8.90900552e-01 -6.60144508e-01 1.30605042e-01
-1.08849168e-01 -8.97179782e-01 8.64680186e-02 7.94927180e-01
-4.59300309e-01 6.07103586e-01 -4.44913775e-01 -1.01452537e-01
-1.14421237e+00 4.39265341e-01 1.37949482e-01 -4.54617232e-01
-3.46263975e-01 4.71765965e-01 -5.89174688e-01 -4.84417975e-01
1.96097061e-01 2.74628043e-01 -3.66987586e-01 1.43545613e-01
6.35284901e-01 8.36388290e-01 1.71196107e-02 -6.86845779e-01
-5.81950843e-01 -1.51468322e-01 -2.93643981e-01 1.18666708e-01
1.40810084e+00 4.69585747e-01 -2.02347741e-01 1.30145758e-01
7.95274138e-01 -2.84393840e-02 -8.39056075e-01 5.91388978e-02
5.13226271e-01 -7.37041056e-01 -1.11887515e-01 -7.01209486e-01
-9.60781932e-01 8.53344798e-01 4.33162302e-01 7.58718550e-01
9.83992457e-01 -3.99693221e-01 4.56493914e-01 3.17293644e-01
6.35902584e-01 -1.09696007e+00 1.53752923e-01 4.80987251e-01
3.03474814e-01 -1.08659375e+00 4.35617939e-02 -3.81558537e-01
-5.57143688e-01 1.27543688e+00 5.15025377e-01 -3.45355421e-01
7.40188360e-01 2.77389884e-01 1.06858527e-02 -3.07934135e-01
-1.00166845e+00 -1.25488773e-01 -1.11536711e-01 9.08991992e-01
3.37437302e-01 2.97857076e-01 -2.87353694e-01 4.80636746e-01
4.67279330e-02 -4.68342990e-01 6.74136221e-01 1.06151795e+00
-6.20899498e-01 -1.44273353e+00 -5.89983582e-01 8.42105448e-01
-7.72535205e-01 1.78329691e-01 -8.18017304e-01 9.41251695e-01
-4.39099930e-02 1.57175422e+00 -3.66287231e-01 -9.33696866e-01
1.41680449e-01 1.33053541e-01 1.05382212e-01 -7.09284127e-01
-9.93726373e-01 -3.99697304e-01 2.28099927e-01 -6.61574543e-01
-5.76803349e-02 -8.27011883e-01 -6.43876433e-01 -7.10771620e-01
-3.36227506e-01 2.71096945e-01 8.48856509e-01 9.37644303e-01
4.38429296e-01 3.10947299e-01 8.62997234e-01 -6.71615779e-01
-1.04095662e+00 -1.05640399e+00 -5.69585025e-01 1.45449981e-01
-9.09064859e-02 -1.11747372e+00 -8.74338567e-01 -4.72252995e-01] | [5.3680100440979, 7.27393913269043] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.