Matches in SemOpenAlex for { <https://semopenalex.org/work/W4312856857> ?p ?o ?g. }
Showing items 1 to 85 of
85
with 100 items per page.
- W4312856857 abstract "Local Transformer-based classification models have recently achieved promising results with relatively low computational costs. However, the effect of aggregating spatial global information of local Transformer-based architecture is not clear. This work investigates the outcome of applying a global attention-based module named multi-resolution overlapped attention (MOA) in the local window-based transformer after each stage. The proposed MOA employs slightly larger and overlapped patches in the key to enable neighborhood pixel information transmission, which leads to significant performance gain. In addition, we thoroughly investigate the effect of the dimension of essential architecture components through extensive experiments and discover an optimum architecture design. Extensive experimental results CIFAR-10, CIFAR-100, and ImageNet-1K datasets demonstrate that the proposed approach outperforms previous vision Transformers with a comparatively fewer number of parameters. The source code and models are publicly available at: https://github.com/krushi1992/MOA-transformer" @default.
- W4312856857 created "2023-01-05" @default.
- W4312856857 creator A5014556254 @default.
- W4312856857 creator A5051094765 @default.
- W4312856857 creator A5056355057 @default.
- W4312856857 creator A5079024523 @default.
- W4312856857 date "2022-08-21" @default.
- W4312856857 modified "2023-10-11" @default.
- W4312856857 title "Aggregating Global Features into Local Vision Transformer" @default.
- W4312856857 cites W2108598243 @default.
- W4312856857 cites W2194775991 @default.
- W4312856857 cites W2565639579 @default.
- W4312856857 cites W2895340641 @default.
- W4312856857 cites W2955058313 @default.
- W4312856857 cites W2963091558 @default.
- W4312856857 cites W2964080601 @default.
- W4312856857 cites W2964137095 @default.
- W4312856857 cites W2982220924 @default.
- W4312856857 cites W2983446232 @default.
- W4312856857 cites W2992308087 @default.
- W4312856857 cites W3121523901 @default.
- W4312856857 cites W3129840128 @default.
- W4312856857 cites W3131500599 @default.
- W4312856857 cites W3138516171 @default.
- W4312856857 cites W3151130473 @default.
- W4312856857 cites W3160068080 @default.
- W4312856857 cites W3179869055 @default.
- W4312856857 cites W3181013887 @default.
- W4312856857 cites W3195814479 @default.
- W4312856857 cites W3210371470 @default.
- W4312856857 cites W3217736865 @default.
- W4312856857 cites W4200061105 @default.
- W4312856857 cites W4226455860 @default.
- W4312856857 doi "https://doi.org/10.1109/icpr56361.2022.9956379" @default.
- W4312856857 hasPublicationYear "2022" @default.
- W4312856857 type Work @default.
- W4312856857 citedByCount "10" @default.
- W4312856857 countsByYear W43128568572023 @default.
- W4312856857 crossrefType "proceedings-article" @default.
- W4312856857 hasAuthorship W4312856857A5014556254 @default.
- W4312856857 hasAuthorship W4312856857A5051094765 @default.
- W4312856857 hasAuthorship W4312856857A5056355057 @default.
- W4312856857 hasAuthorship W4312856857A5079024523 @default.
- W4312856857 hasBestOaLocation W43128568572 @default.
- W4312856857 hasConcept C119599485 @default.
- W4312856857 hasConcept C123657996 @default.
- W4312856857 hasConcept C127413603 @default.
- W4312856857 hasConcept C142362112 @default.
- W4312856857 hasConcept C153180895 @default.
- W4312856857 hasConcept C153349607 @default.
- W4312856857 hasConcept C154945302 @default.
- W4312856857 hasConcept C160633673 @default.
- W4312856857 hasConcept C165801399 @default.
- W4312856857 hasConcept C205372480 @default.
- W4312856857 hasConcept C41008148 @default.
- W4312856857 hasConcept C66322947 @default.
- W4312856857 hasConceptScore W4312856857C119599485 @default.
- W4312856857 hasConceptScore W4312856857C123657996 @default.
- W4312856857 hasConceptScore W4312856857C127413603 @default.
- W4312856857 hasConceptScore W4312856857C142362112 @default.
- W4312856857 hasConceptScore W4312856857C153180895 @default.
- W4312856857 hasConceptScore W4312856857C153349607 @default.
- W4312856857 hasConceptScore W4312856857C154945302 @default.
- W4312856857 hasConceptScore W4312856857C160633673 @default.
- W4312856857 hasConceptScore W4312856857C165801399 @default.
- W4312856857 hasConceptScore W4312856857C205372480 @default.
- W4312856857 hasConceptScore W4312856857C41008148 @default.
- W4312856857 hasConceptScore W4312856857C66322947 @default.
- W4312856857 hasLocation W43128568571 @default.
- W4312856857 hasLocation W43128568572 @default.
- W4312856857 hasOpenAccess W4312856857 @default.
- W4312856857 hasPrimaryLocation W43128568571 @default.
- W4312856857 hasRelatedWork W1965781815 @default.
- W4312856857 hasRelatedWork W2136485282 @default.
- W4312856857 hasRelatedWork W2149249189 @default.
- W4312856857 hasRelatedWork W2269705005 @default.
- W4312856857 hasRelatedWork W2542880803 @default.
- W4312856857 hasRelatedWork W2546871836 @default.
- W4312856857 hasRelatedWork W2547748020 @default.
- W4312856857 hasRelatedWork W2918100922 @default.
- W4312856857 hasRelatedWork W3043252291 @default.
- W4312856857 hasRelatedWork W3160284275 @default.
- W4312856857 isParatext "false" @default.
- W4312856857 isRetracted "false" @default.
- W4312856857 workType "article" @default.