Research map / Detection and segmentation / Weakly supervised detection and segmentation
Self-supervision with Superpixels: Training Few-shot Medical Image Segmentation without Annotation
Cheng Ouyang, Carlo Biffi, Chen Chen, Turkay Kart, Huaqi Qiu, Daniel Rueckert
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 superpixel self-supervised training (no annotations) to PANet (ablation) | PANet · Abd-CT setting 2 | mean Dice: 29.43 -> 44.42 | Improved |
| add superpixel self-supervised training to PANet (ablation) | PANet · Abd-MRI setting 2 | mean Dice: 46.33 -> 54.85 | Improved |
| add adaptive local prototype pooling to SSL-PANet (ablation) | SSL-PANet · Abd-CT setting 2 | mean Dice: 44.42 -> 63.02 | Improved |
| add adaptive local prototype pooling to SSL-PANet (ablation) | SSL-PANet · Abd-MRI setting 2 | mean Dice: 54.85 -> 73.02 | Improved |
| add adaptive local prototype pooling to supervised PANet (ablation) | PANet · Card-MRI setting 1 | mean Dice: 42.96 -> 60.34 | Improved |
| add superpixel SSL to ALPNet (ablation) | ALPNet · Card-MRI setting 1 | mean Dice: 60.34 -> 76.90 | Improved |
| add SSL to ALPNet instead of supervised training (ablation) | ALPNet · Abd-CT setting 1 | mean Dice: 41.63 -> 73.35 | Improved |
| add intensity and geometric transforms between support and query (ablation) | SSL-ALPNet · Abd-CT setting 2 | mean Dice: 55.13 -> 63.02 | Improved |
| add intensity transform on top of geometric transform (ablation) | SSL-ALPNet · Abd-CT setting 2 liver | Dice: 78.93 -> 73.65 | Hurt |
| raise minimum pseudolabel size from 400 to 1600 px (ablation) | SSL-ALPNet · Abd-CT setting 2 | mean Dice: 63.02 -> 57.08 | Hurt |
| lower minimum pseudolabel size from 400 to 100 px (ablation) | SSL-ALPNet · Abd-CT setting 2 | mean Dice: 63.02 -> 55.49 | Hurt |
| initialize with COCO weights only, without training (ablation) | ALPNet-init · Abd-CT setting 2 | mean Dice: 20.90 vs 63.02 for SSL-ALPNet | Hurt |
Closest papers by meaning
- Semantic-Aware Superpixel for Weakly Supervised Semantic Segmentation (AAAI 2023 · similarity 0.87)
- Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training (AAAI 2024 · similarity 0.87)
- FAMNet: Frequency-aware Matching Network for Cross-domain Few-shot Medical Image Segmentation (AAAI 2025 · similarity 0.86)
- A Semantic Knowledge Complementarity based Decoupling Framework for Semi-supervised Class-imbalanced Medical Image Segmentation (CVPR 2025 · similarity 0.85)
- Enhancing SAM with Efficient Prompting and Preference Optimization for Semi-supervised Medical Image Segmentation (CVPR 2025 · similarity 0.85)
- Versatile Medical Image Segmentation Learned from Multi-Source Datasets via Model Self-Disambiguation (CVPR 2024 · similarity 0.85)
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.
