Matches in SemOpenAlex for { <https://semopenalex.org/work/W4309675391> ?p ?o ?g. }
Showing items 1 to 92 of
92
with 100 items per page.
- W4309675391 abstract "emantic segmentation is an essential task in developing medical image diagnosis systems. However, building an annotated medical dataset is expensive. Thus, semi-supervised methods are significant in this circumstance. In semi-supervised learning, the quality of labels plays a crucial role in modelperformance. In this work, we present a new pseudo labeling strategy that enhances the quality of pseudo labels used for training student networks. We follow the multi-stage semi-supervised training approach, which trains a teacher model on a labeled dataset and then uses the trained teacher to render pseudo labels for student training. By doing so, the pseudo labels will be updated and more precise as training progress. The key difference between previous and our methods is that we update the teacher model during the student training process. So the quality of pseudo labels is improved during the student training process. We also propose a simple but effective strategy to enhance the quality of pseudo labels using a momentum model - a slow copy version of the original model during training. By applying the momentum model combined with re-rendering pseudo labels during student training, we achieved an average of 84.1% Dice Score on five datasets (i.e., Kvarsir, CVC-ClinicDB, ETIS-LaribPolypDB, CVC-ColonDB, and CVC-300) with only 20% of the dataset used as labeled data. Our results surpass common practice by 3% and even approach fully-supervised results on some datasets. Oursource code and pre-trained models are available at https://github.com/sun-asterisk-research/online_learning_ssl" @default.
- W4309675391 created "2022-11-29" @default.
- W4309675391 creator A5009640547 @default.
- W4309675391 creator A5025948040 @default.
- W4309675391 creator A5033754259 @default.
- W4309675391 creator A5045407393 @default.
- W4309675391 creator A5058037820 @default.
- W4309675391 creator A5064509843 @default.
- W4309675391 date "2022-10-19" @default.
- W4309675391 modified "2023-10-13" @default.
- W4309675391 title "Online pseudo labeling for polyp segmentation with momentum networks" @default.
- W4309675391 cites W1903029394 @default.
- W4309675391 cites W2979515228 @default.
- W4309675391 cites W2979805229 @default.
- W4309675391 cites W2985406498 @default.
- W4309675391 cites W2996290406 @default.
- W4309675391 cites W3035294798 @default.
- W4309675391 cites W3092344722 @default.
- W4309675391 cites W3175308890 @default.
- W4309675391 cites W3183021106 @default.
- W4309675391 cites W3202263958 @default.
- W4309675391 cites W3203497085 @default.
- W4309675391 cites W3204166336 @default.
- W4309675391 cites W4226329133 @default.
- W4309675391 doi "https://doi.org/10.1109/kse56063.2022.9953785" @default.
- W4309675391 hasPublicationYear "2022" @default.
- W4309675391 type Work @default.
- W4309675391 citedByCount "1" @default.
- W4309675391 countsByYear W43096753912023 @default.
- W4309675391 crossrefType "proceedings-article" @default.
- W4309675391 hasAuthorship W4309675391A5009640547 @default.
- W4309675391 hasAuthorship W4309675391A5025948040 @default.
- W4309675391 hasAuthorship W4309675391A5033754259 @default.
- W4309675391 hasAuthorship W4309675391A5045407393 @default.
- W4309675391 hasAuthorship W4309675391A5058037820 @default.
- W4309675391 hasAuthorship W4309675391A5064509843 @default.
- W4309675391 hasBestOaLocation W43096753912 @default.
- W4309675391 hasConcept C111472728 @default.
- W4309675391 hasConcept C111919701 @default.
- W4309675391 hasConcept C119857082 @default.
- W4309675391 hasConcept C136389625 @default.
- W4309675391 hasConcept C138885662 @default.
- W4309675391 hasConcept C154945302 @default.
- W4309675391 hasConcept C162324750 @default.
- W4309675391 hasConcept C177264268 @default.
- W4309675391 hasConcept C187736073 @default.
- W4309675391 hasConcept C199360897 @default.
- W4309675391 hasConcept C205711294 @default.
- W4309675391 hasConcept C2776145971 @default.
- W4309675391 hasConcept C2776760102 @default.
- W4309675391 hasConcept C2779530757 @default.
- W4309675391 hasConcept C2780451532 @default.
- W4309675391 hasConcept C41008148 @default.
- W4309675391 hasConcept C50644808 @default.
- W4309675391 hasConcept C89600930 @default.
- W4309675391 hasConcept C98045186 @default.
- W4309675391 hasConceptScore W4309675391C111472728 @default.
- W4309675391 hasConceptScore W4309675391C111919701 @default.
- W4309675391 hasConceptScore W4309675391C119857082 @default.
- W4309675391 hasConceptScore W4309675391C136389625 @default.
- W4309675391 hasConceptScore W4309675391C138885662 @default.
- W4309675391 hasConceptScore W4309675391C154945302 @default.
- W4309675391 hasConceptScore W4309675391C162324750 @default.
- W4309675391 hasConceptScore W4309675391C177264268 @default.
- W4309675391 hasConceptScore W4309675391C187736073 @default.
- W4309675391 hasConceptScore W4309675391C199360897 @default.
- W4309675391 hasConceptScore W4309675391C205711294 @default.
- W4309675391 hasConceptScore W4309675391C2776145971 @default.
- W4309675391 hasConceptScore W4309675391C2776760102 @default.
- W4309675391 hasConceptScore W4309675391C2779530757 @default.
- W4309675391 hasConceptScore W4309675391C2780451532 @default.
- W4309675391 hasConceptScore W4309675391C41008148 @default.
- W4309675391 hasConceptScore W4309675391C50644808 @default.
- W4309675391 hasConceptScore W4309675391C89600930 @default.
- W4309675391 hasConceptScore W4309675391C98045186 @default.
- W4309675391 hasLocation W43096753911 @default.
- W4309675391 hasLocation W43096753912 @default.
- W4309675391 hasOpenAccess W4309675391 @default.
- W4309675391 hasPrimaryLocation W43096753911 @default.
- W4309675391 hasRelatedWork W1614742790 @default.
- W4309675391 hasRelatedWork W2036486025 @default.
- W4309675391 hasRelatedWork W2131513867 @default.
- W4309675391 hasRelatedWork W2952937263 @default.
- W4309675391 hasRelatedWork W3170999895 @default.
- W4309675391 hasRelatedWork W34092691 @default.
- W4309675391 hasRelatedWork W4287117940 @default.
- W4309675391 hasRelatedWork W4300902524 @default.
- W4309675391 hasRelatedWork W4312414840 @default.
- W4309675391 hasRelatedWork W970023169 @default.
- W4309675391 isParatext "false" @default.
- W4309675391 isRetracted "false" @default.
- W4309675391 workType "article" @default.