Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387642396> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W4387642396 endingPage "206" @default.
- W4387642396 startingPage "196" @default.
- W4387642396 abstract "Federated learning (FL) enables multiple sites to collaboratively train powerful deep models without compromising data privacy and security. The statistical heterogeneity (e.g., non-IID data and domain shifts) is a primary obstacle in FL, impairing the generalization performance of the global model. Weakly supervised segmentation, which uses sparsely-grained (i.e., point-, bounding box-, scribble-, block-wise) supervision, is increasingly being paid attention to due to its great potential of reducing annotation costs. However, there may exist label heterogeneity, i.e., different annotation forms across sites. In this paper, we propose a novel personalized FL framework for medical image segmentation, named FedICRA, which uniformly leverages heterogeneous weak supervision via adaptIve Contrastive Representation and Aggregation. Concretely, to facilitate personalized modeling and to avoid confusion, a channel selection based site contrastive representation module is employed to adaptively cluster intra-site embeddings and separate inter-site ones. To effectively integrate the common knowledge from the global model with the unique knowledge from each local model, an adaptive aggregation module is applied for updating and initializing local models at the element level. Additionally, a weakly supervised objective function that leverages a multiscale tree energy loss and a gated CRF loss is employed to generate more precise pseudo-labels and further boost the segmentation performance. Through extensive experiments on two distinct medical image segmentation tasks of different modalities, the proposed FedICRA demonstrates overwhelming performance over other state-of-the-art personalized FL methods. Its performance even approaches that of fully supervised training on centralized data. Our code and data are available at https://github.com/llmir/FedICRA ." @default.
- W4387642396 created "2023-10-15" @default.
- W4387642396 creator A5001406512 @default.
- W4387642396 creator A5011095444 @default.
- W4387642396 creator A5037124096 @default.
- W4387642396 creator A5050401709 @default.
- W4387642396 creator A5089509548 @default.
- W4387642396 date "2023-10-15" @default.
- W4387642396 modified "2023-10-15" @default.
- W4387642396 title "Unifying and Personalizing Weakly-Supervised Federated Medical Image Segmentation via Adaptive Representation and Aggregation" @default.
- W4387642396 cites W1901129140 @default.
- W4387642396 cites W2034786340 @default.
- W4387642396 cites W2077371116 @default.
- W4387642396 cites W2768181417 @default.
- W4387642396 cites W2963198662 @default.
- W4387642396 cites W2979448322 @default.
- W4387642396 cites W2979637109 @default.
- W4387642396 cites W2995059005 @default.
- W4387642396 cites W3013935260 @default.
- W4387642396 cites W3111741353 @default.
- W4387642396 cites W3135073716 @default.
- W4387642396 cites W3164927745 @default.
- W4387642396 cites W3169044395 @default.
- W4387642396 cites W3176031707 @default.
- W4387642396 cites W3195545894 @default.
- W4387642396 cites W3203049650 @default.
- W4387642396 cites W4285229324 @default.
- W4387642396 cites W4293219563 @default.
- W4387642396 cites W4312974914 @default.
- W4387642396 cites W4313151123 @default.
- W4387642396 cites W4313887608 @default.
- W4387642396 cites W4377023536 @default.
- W4387642396 cites W4385338552 @default.
- W4387642396 doi "https://doi.org/10.1007/978-3-031-45673-2_20" @default.
- W4387642396 hasPublicationYear "2023" @default.
- W4387642396 type Work @default.
- W4387642396 citedByCount "0" @default.
- W4387642396 crossrefType "book-chapter" @default.
- W4387642396 hasAuthorship W4387642396A5001406512 @default.
- W4387642396 hasAuthorship W4387642396A5011095444 @default.
- W4387642396 hasAuthorship W4387642396A5037124096 @default.
- W4387642396 hasAuthorship W4387642396A5050401709 @default.
- W4387642396 hasAuthorship W4387642396A5089509548 @default.
- W4387642396 hasConcept C114466953 @default.
- W4387642396 hasConcept C115961682 @default.
- W4387642396 hasConcept C119857082 @default.
- W4387642396 hasConcept C147037132 @default.
- W4387642396 hasConcept C153180895 @default.
- W4387642396 hasConcept C154945302 @default.
- W4387642396 hasConcept C17744445 @default.
- W4387642396 hasConcept C199360897 @default.
- W4387642396 hasConcept C199539241 @default.
- W4387642396 hasConcept C2524010 @default.
- W4387642396 hasConcept C2776321320 @default.
- W4387642396 hasConcept C2776359362 @default.
- W4387642396 hasConcept C2777210771 @default.
- W4387642396 hasConcept C33923547 @default.
- W4387642396 hasConcept C41008148 @default.
- W4387642396 hasConcept C89600930 @default.
- W4387642396 hasConcept C94625758 @default.
- W4387642396 hasConceptScore W4387642396C114466953 @default.
- W4387642396 hasConceptScore W4387642396C115961682 @default.
- W4387642396 hasConceptScore W4387642396C119857082 @default.
- W4387642396 hasConceptScore W4387642396C147037132 @default.
- W4387642396 hasConceptScore W4387642396C153180895 @default.
- W4387642396 hasConceptScore W4387642396C154945302 @default.
- W4387642396 hasConceptScore W4387642396C17744445 @default.
- W4387642396 hasConceptScore W4387642396C199360897 @default.
- W4387642396 hasConceptScore W4387642396C199539241 @default.
- W4387642396 hasConceptScore W4387642396C2524010 @default.
- W4387642396 hasConceptScore W4387642396C2776321320 @default.
- W4387642396 hasConceptScore W4387642396C2776359362 @default.
- W4387642396 hasConceptScore W4387642396C2777210771 @default.
- W4387642396 hasConceptScore W4387642396C33923547 @default.
- W4387642396 hasConceptScore W4387642396C41008148 @default.
- W4387642396 hasConceptScore W4387642396C89600930 @default.
- W4387642396 hasConceptScore W4387642396C94625758 @default.
- W4387642396 hasLocation W43876423961 @default.
- W4387642396 hasOpenAccess W4387642396 @default.
- W4387642396 hasPrimaryLocation W43876423961 @default.
- W4387642396 hasRelatedWork W2776054962 @default.
- W4387642396 hasRelatedWork W2949303564 @default.
- W4387642396 hasRelatedWork W2963343840 @default.
- W4387642396 hasRelatedWork W2980046504 @default.
- W4387642396 hasRelatedWork W3039174484 @default.
- W4387642396 hasRelatedWork W3163786145 @default.
- W4387642396 hasRelatedWork W4287728368 @default.
- W4387642396 hasRelatedWork W4300812085 @default.
- W4387642396 hasRelatedWork W4353114450 @default.
- W4387642396 hasRelatedWork W4381053183 @default.
- W4387642396 isParatext "false" @default.
- W4387642396 isRetracted "false" @default.
- W4387642396 workType "book-chapter" @default.