Research map / Detection and segmentation / Open-set and few-shot detection
Dual Contrastive Learning with Anatomical Auxiliary Supervision for Few-Shot Medical Image Segmentation
Huisi Wu, Fangyan Xiao, Chongxin Liang
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 skip connections to sSENet (ablation) | sSENet · CHAOS-T2 | mean Dice (%): 47.45 -> 49.08 | Improved |
| add DCL module (ablation) | Baseline with skip connections · CHAOS-T2 | mean Dice (%): 49.08 -> 67.21 | Improved |
| add CIP module (ablation) | Baseline with skip connections · CHAOS-T2 | mean Dice (%): 49.08 -> 62.07 | Improved |
| add CIP on top of DCL (ablation) | Baseline+DCL · CHAOS-T2 | mean Dice (%): 67.21 -> 76.36 | Improved |
| add DCL on top of CIP (ablation) | Baseline+CIP · CHAOS-T2 | mean Dice (%): 62.07 -> 76.36 | Improved |
| use Felzenszwalb instead of SLIC superpixels (ablation) | AAS-DCL · CHAOS-T2 | mean Dice (%): 61.35 -> 76.36 | Improved |
| use Felzenszwalb instead of SEEDS superpixels (ablation) | AAS-DCL · CHAOS-T2 | mean Dice (%): 67.07 -> 76.36 | Improved |
| use 5 CIP iterations instead of 3 (ablation) | AAS-DCL · CHAOS-T2 | mean Dice (%): 72.58 -> 76.36 | Improved |
| use 7 CIP iterations instead of 5 (ablation) | AAS-DCL · CHAOS-T2 | mean Dice (%): 76.36 -> 76.12 | No clear effect |
| use 3 non-target slices instead of 1 (ablation) | AAS-DCL · CHAOS-T2 | mean Dice (%): 70.34 -> 76.36 | Improved |
| use 5 non-target slices instead of 3 (ablation) | AAS-DCL · CHAOS-T2 | mean Dice (%): 76.36 -> 74.80 | Hurt |
| AAS-DCL vs RP-Net | RP-Net · CHAOS-T2 | mean Dice (%): 75.81 -> 76.36 | Improved |
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.
- PANet — PANet: Few-Shot Image Semantic Segmentation With Prototype Alignment (ICCV 2019) · 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
- Rethinking Semi-Supervised Medical Image Segmentation: A Variance-Reduction Perspective (NeurIPS 2023 · similarity 0.88)
- Contrastive learning of global and local features for medical image segmentation with limited annotations (NeurIPS 2020 · similarity 0.88)
- Mine yOur owN Anatomy: Revisiting Medical Image Segmentation With Extremely Limited Labels (TPAMI 2024 · similarity 0.88)
- ConDSeg: A General Medical Image Segmentation Framework via Contrast-Driven Feature Enhancement (AAAI 2025 · similarity 0.85)
- SGTC: Semantic-Guided Triplet Co-training for Sparsely Annotated Semi-Supervised Medical Image Segmentation (AAAI 2025 · similarity 0.84)
- Training Like a Medical Resident: Context-Prior Learning Toward Universal Medical Image Segmentation (CVPR 2024 · similarity 0.84)
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.
