Research map / Label-efficient and robust learning / Semi-supervised learning
Divide, Conquer, and Aggregate: Asymmetric Experts for Class-Imbalanced Semi-Supervised Medical Image Segmentation
Yajun Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add three asymmetric expert decoders (ablation) | V-Net baseline · Synapse 20% | Avg Dice: 47.92 -> 55.38 | Improved |
| add Log-Gap-based label-split strategy (ablation) | V-Net + 3 experts · Synapse 20% | Avg Dice: 55.38 -> 69.89 | Improved |
| add DFAM aggregation decoder (ablation) | DCA · Synapse 20% | Avg Dice: 69.89 -> 73.20 | Improved |
| use Log-Gap division instead of uniform division (ablation) | DCA · AMOS 5% | Avg Dice: 67.79 -> 69.90 | Improved |
| use Log-Gap division instead of uniform division (ablation) | DCA · AMOS 2% | Avg Dice: 55.53 -> 59.88 | Improved |
| use Log-Gap division instead of fixed-threshold division (ablation) | DCA · AMOS 2% | Avg Dice: 57.93 -> 59.88 | Improved |
| shallower head decoder and deeper tail decoder (V2) (ablation) | expert decoders · AMOS 5% | Avg Dice: 60.25 -> 67.33 | Improved |
| larger dilation rates (V3) (ablation) | expert decoders · AMOS 5% | Avg Dice: 60.25 -> 64.96 | Improved |
| combine V2 and V3 designs (ablation) | expert decoders · AMOS 5% | Avg Dice: 67.33 -> 69.90 | Improved |
| DCA vs best prior method | DCA · Synapse 20% | Avg Dice: 68.43 -> 73.20 | Improved |
| DCA vs best prior method | DCA · AMOS 5% | Avg Dice: 63.51 -> 69.90 | Improved |
| DCA vs SKCDF on surface distance | DCA · Synapse 20% | Avg ASD: 1.45 -> 1.78 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MagicNet — MagicNet: Semi-Supervised Multi-Organ Segmentation via Magic-Cube Partition and Recovery (CVPR 2023) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- A Semantic Knowledge Complementarity based Decoupling Framework for Semi-supervised Class-imbalanced Medical Image Segmentation (CVPR 2025 · similarity 0.91)
- Annotation Ambiguity Aware Semi-Supervised Medical Image Segmentation (CVPR 2025 · similarity 0.89)
- SGTC: Semantic-Guided Triplet Co-training for Sparsely Annotated Semi-Supervised Medical Image Segmentation (AAAI 2025 · similarity 0.88)
- Combinatorial CNN-Transformer Learning with Manifold Constraints for Semi-supervised Medical Image Segmentation (AAAI 2024 · similarity 0.88)
- DyCON: Dynamic Uncertainty-aware Consistency and Contrastive Learning for Semi-supervised Medical Image Segmentation (CVPR 2025 · similarity 0.88)
- Semi-supervised TEE Segmentation via Interacting with SAM Equipped with Noise-Resilient Prompting (AAAI 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
