Matches in SemOpenAlex for { <https://semopenalex.org/work/W4379879209> ?p ?o ?g. }
- W4379879209 endingPage "110698" @default.
- W4379879209 startingPage "110698" @default.
- W4379879209 abstract "The purpose of one-shot semantic segmentation is to classify all pixels in a query image with a labeled support sample. Its current approaches divide data into two disjoint parts, a training set and a testing set. When the training and testing data are from different domains, the migration of the trained model to the testing data for prediction tends to perform less efficiently. Therefore, through the introduction of a distribution alignment module, a new dual-channel network (DCNet) for one-shot segmentation is proposed in this paper. The purpose of this introduction is to alleviate the performance degradation caused by distribution differences between the datasets. Furthermore, a soft pseudo-mask based prototype refinement module with Gumbel-Softmax is designed to enhance the expressive ability of prototypes and improve segmentation performance. Experiments are made on four datasets with large distribution differences, proving that our approach outperforms the compared methods by a large margin." @default.
- W4379879209 created "2023-06-09" @default.
- W4379879209 creator A5002229860 @default.
- W4379879209 creator A5058297364 @default.
- W4379879209 creator A5092116342 @default.
- W4379879209 date "2023-09-01" @default.
- W4379879209 modified "2023-09-27" @default.
- W4379879209 title "A dual-channel network for cross-domain one-shot semantic segmentation via adversarial learning" @default.
- W4379879209 cites W1903029394 @default.
- W4379879209 cites W2118246710 @default.
- W4379879209 cites W2140405352 @default.
- W4379879209 cites W2165698076 @default.
- W4379879209 cites W2194775991 @default.
- W4379879209 cites W2412782625 @default.
- W4379879209 cites W2765811365 @default.
- W4379879209 cites W2766938848 @default.
- W4379879209 cites W2807974043 @default.
- W4379879209 cites W2963107255 @default.
- W4379879209 cites W2963599420 @default.
- W4379879209 cites W2963881378 @default.
- W4379879209 cites W2990230185 @default.
- W4379879209 cites W2994528761 @default.
- W4379879209 cites W3033502887 @default.
- W4379879209 cites W3034562924 @default.
- W4379879209 cites W3034985049 @default.
- W4379879209 cites W3102785203 @default.
- W4379879209 cites W3103010481 @default.
- W4379879209 cites W3167559252 @default.
- W4379879209 cites W3176065502 @default.
- W4379879209 cites W3203637842 @default.
- W4379879209 cites W3204077273 @default.
- W4379879209 cites W3210969440 @default.
- W4379879209 cites W4214660208 @default.
- W4379879209 cites W4225575005 @default.
- W4379879209 cites W4225828097 @default.
- W4379879209 cites W4283832242 @default.
- W4379879209 cites W4287448664 @default.
- W4379879209 doi "https://doi.org/10.1016/j.knosys.2023.110698" @default.
- W4379879209 hasPublicationYear "2023" @default.
- W4379879209 type Work @default.
- W4379879209 citedByCount "0" @default.
- W4379879209 crossrefType "journal-article" @default.
- W4379879209 hasAuthorship W4379879209A5002229860 @default.
- W4379879209 hasAuthorship W4379879209A5058297364 @default.
- W4379879209 hasAuthorship W4379879209A5092116342 @default.
- W4379879209 hasConcept C105795698 @default.
- W4379879209 hasConcept C114614502 @default.
- W4379879209 hasConcept C119857082 @default.
- W4379879209 hasConcept C124101348 @default.
- W4379879209 hasConcept C124952713 @default.
- W4379879209 hasConcept C127162648 @default.
- W4379879209 hasConcept C137610916 @default.
- W4379879209 hasConcept C142362112 @default.
- W4379879209 hasConcept C147581598 @default.
- W4379879209 hasConcept C153180895 @default.
- W4379879209 hasConcept C154945302 @default.
- W4379879209 hasConcept C162324750 @default.
- W4379879209 hasConcept C16910744 @default.
- W4379879209 hasConcept C177264268 @default.
- W4379879209 hasConcept C187736073 @default.
- W4379879209 hasConcept C188441871 @default.
- W4379879209 hasConcept C199360897 @default.
- W4379879209 hasConcept C2780451532 @default.
- W4379879209 hasConcept C2780980858 @default.
- W4379879209 hasConcept C31258907 @default.
- W4379879209 hasConcept C33923547 @default.
- W4379879209 hasConcept C41008148 @default.
- W4379879209 hasConcept C45340560 @default.
- W4379879209 hasConcept C50644808 @default.
- W4379879209 hasConcept C58489278 @default.
- W4379879209 hasConcept C774472 @default.
- W4379879209 hasConcept C89600930 @default.
- W4379879209 hasConceptScore W4379879209C105795698 @default.
- W4379879209 hasConceptScore W4379879209C114614502 @default.
- W4379879209 hasConceptScore W4379879209C119857082 @default.
- W4379879209 hasConceptScore W4379879209C124101348 @default.
- W4379879209 hasConceptScore W4379879209C124952713 @default.
- W4379879209 hasConceptScore W4379879209C127162648 @default.
- W4379879209 hasConceptScore W4379879209C137610916 @default.
- W4379879209 hasConceptScore W4379879209C142362112 @default.
- W4379879209 hasConceptScore W4379879209C147581598 @default.
- W4379879209 hasConceptScore W4379879209C153180895 @default.
- W4379879209 hasConceptScore W4379879209C154945302 @default.
- W4379879209 hasConceptScore W4379879209C162324750 @default.
- W4379879209 hasConceptScore W4379879209C16910744 @default.
- W4379879209 hasConceptScore W4379879209C177264268 @default.
- W4379879209 hasConceptScore W4379879209C187736073 @default.
- W4379879209 hasConceptScore W4379879209C188441871 @default.
- W4379879209 hasConceptScore W4379879209C199360897 @default.
- W4379879209 hasConceptScore W4379879209C2780451532 @default.
- W4379879209 hasConceptScore W4379879209C2780980858 @default.
- W4379879209 hasConceptScore W4379879209C31258907 @default.
- W4379879209 hasConceptScore W4379879209C33923547 @default.
- W4379879209 hasConceptScore W4379879209C41008148 @default.
- W4379879209 hasConceptScore W4379879209C45340560 @default.
- W4379879209 hasConceptScore W4379879209C50644808 @default.
- W4379879209 hasConceptScore W4379879209C58489278 @default.
- W4379879209 hasConceptScore W4379879209C774472 @default.