Matches in SemOpenAlex for { <https://semopenalex.org/work/W4361762480> ?p ?o ?g. }
- W4361762480 endingPage "5630" @default.
- W4361762480 startingPage "5617" @default.
- W4361762480 abstract "Cross-domain text recognition is a very challenging task due to the domain drift problem. One solution is aligning feature distributions between domains through Unsupervised Domain Adaptation (UDA). All existing methods perform feature alignment based on the whole image or semantic character features. However, visual character features without contextual semantics also contain much valuable information, <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>e.g</i> ., stroke features of individual characters, which also benefits domain transfer. To this end, we propose a dual intra-Class Aggregation based unsupervised Domain Adaptation method (CADA) for text recognition, which aligns both visual and semantic character feature distributions. To our knowledge, CADA is the first to consider visual character features without contextual semantics in cross-domain text recognition tasks. Accordingly, a Single-head Self-Attention (SSA) mechanism is introduced for extracting visual character features. Thereafter, a dual intra-class aggregation strategy is designed, which performs class aggregations in both visual and semantic spaces. We test the proposed method on widely-used datasets by combining it with representative text recognition models with various decoding methods. Extensive experimental results demonstrate the superiority and generality. Moreover, there is no additional inference time introduced compared to the baselines." @default.
- W4361762480 created "2023-04-04" @default.
- W4361762480 creator A5000757861 @default.
- W4361762480 creator A5033973998 @default.
- W4361762480 creator A5052168662 @default.
- W4361762480 creator A5086235570 @default.
- W4361762480 date "2023-10-01" @default.
- W4361762480 modified "2023-10-15" @default.
- W4361762480 title "Unsupervised Domain Adaptation via Class Aggregation for Text Recognition" @default.
- W4361762480 cites W1971822075 @default.
- W4361762480 cites W1981283549 @default.
- W4361762480 cites W2008806374 @default.
- W4361762480 cites W2050995497 @default.
- W4361762480 cites W2099247484 @default.
- W4361762480 cites W2127141656 @default.
- W4361762480 cites W2144554289 @default.
- W4361762480 cites W2146835493 @default.
- W4361762480 cites W2152928267 @default.
- W4361762480 cites W2194187530 @default.
- W4361762480 cites W2343052201 @default.
- W4361762480 cites W2750938222 @default.
- W4361762480 cites W2767179670 @default.
- W4361762480 cites W2788069964 @default.
- W4361762480 cites W2791490164 @default.
- W4361762480 cites W2798681837 @default.
- W4361762480 cites W2810983211 @default.
- W4361762480 cites W2895281799 @default.
- W4361762480 cites W2959965583 @default.
- W4361762480 cites W2962968458 @default.
- W4361762480 cites W2963517393 @default.
- W4361762480 cites W2964288524 @default.
- W4361762480 cites W2976726839 @default.
- W4361762480 cites W2978036638 @default.
- W4361762480 cites W2987563462 @default.
- W4361762480 cites W2997864923 @default.
- W4361762480 cites W3004846386 @default.
- W4361762480 cites W3013224334 @default.
- W4361762480 cites W3035449864 @default.
- W4361762480 cites W3136420864 @default.
- W4361762480 cites W3163490880 @default.
- W4361762480 cites W3175618949 @default.
- W4361762480 cites W3175855397 @default.
- W4361762480 cites W3176512504 @default.
- W4361762480 cites W3177206657 @default.
- W4361762480 cites W3177684257 @default.
- W4361762480 cites W3181186176 @default.
- W4361762480 cites W3198290649 @default.
- W4361762480 cites W3198649044 @default.
- W4361762480 cites W3202415716 @default.
- W4361762480 cites W3202912918 @default.
- W4361762480 cites W3204778874 @default.
- W4361762480 cites W4207058148 @default.
- W4361762480 cites W4283819468 @default.
- W4361762480 cites W4308233807 @default.
- W4361762480 doi "https://doi.org/10.1109/tcsvt.2023.3262692" @default.
- W4361762480 hasPublicationYear "2023" @default.
- W4361762480 type Work @default.
- W4361762480 citedByCount "0" @default.
- W4361762480 crossrefType "journal-article" @default.
- W4361762480 hasAuthorship W4361762480A5000757861 @default.
- W4361762480 hasAuthorship W4361762480A5033973998 @default.
- W4361762480 hasAuthorship W4361762480A5052168662 @default.
- W4361762480 hasAuthorship W4361762480A5086235570 @default.
- W4361762480 hasConcept C134306372 @default.
- W4361762480 hasConcept C138885662 @default.
- W4361762480 hasConcept C153180895 @default.
- W4361762480 hasConcept C154945302 @default.
- W4361762480 hasConcept C15744967 @default.
- W4361762480 hasConcept C184337299 @default.
- W4361762480 hasConcept C199360897 @default.
- W4361762480 hasConcept C204321447 @default.
- W4361762480 hasConcept C2524010 @default.
- W4361762480 hasConcept C2776214188 @default.
- W4361762480 hasConcept C2776401178 @default.
- W4361762480 hasConcept C2777212361 @default.
- W4361762480 hasConcept C2780767217 @default.
- W4361762480 hasConcept C2780861071 @default.
- W4361762480 hasConcept C33923547 @default.
- W4361762480 hasConcept C36503486 @default.
- W4361762480 hasConcept C41008148 @default.
- W4361762480 hasConcept C41895202 @default.
- W4361762480 hasConcept C52622490 @default.
- W4361762480 hasConcept C542102704 @default.
- W4361762480 hasConceptScore W4361762480C134306372 @default.
- W4361762480 hasConceptScore W4361762480C138885662 @default.
- W4361762480 hasConceptScore W4361762480C153180895 @default.
- W4361762480 hasConceptScore W4361762480C154945302 @default.
- W4361762480 hasConceptScore W4361762480C15744967 @default.
- W4361762480 hasConceptScore W4361762480C184337299 @default.
- W4361762480 hasConceptScore W4361762480C199360897 @default.
- W4361762480 hasConceptScore W4361762480C204321447 @default.
- W4361762480 hasConceptScore W4361762480C2524010 @default.
- W4361762480 hasConceptScore W4361762480C2776214188 @default.
- W4361762480 hasConceptScore W4361762480C2776401178 @default.
- W4361762480 hasConceptScore W4361762480C2777212361 @default.
- W4361762480 hasConceptScore W4361762480C2780767217 @default.
- W4361762480 hasConceptScore W4361762480C2780861071 @default.
- W4361762480 hasConceptScore W4361762480C33923547 @default.