Research map / Detection and segmentation / Open-set and few-shot detection
Biomedical Image Segmentation via Representative Annotation
Hao Zheng, Lin Yang, Jianxu Chen, Jun Han, Yizhe Zhang, Peixian Liang, Zhuo Zhao, Chaoli Wang, Danny Z. Chen
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 |
|---|---|---|---|
| use ClsMC-RS instead of clustering-only selection (add max-cover) (ablation) | VAE-FEN RA, 30% annotation · GlaS Part B | F1 score: 0.838 -> 0.843 | Improved |
| use ClsMC-RS instead of max-cover-only selection (add clustering) (ablation) | VAE-FEN RA, 30% annotation · GlaS Part B | F1 score: 0.833 -> 0.843 | Improved |
| use ClsMC-RS instead of clustering-only selection (ablation) | VAE-FEN RA, 50% annotation · GlaS Part B | F1 score: 0.855 -> 0.862 | Improved |
| use ClsMC-RS instead of max-cover-only selection (ablation) | VAE-FEN RA, 50% annotation · GlaS Part A | Object Dice: 0.900 -> 0.897 | Hurt |
| use ClsMC-RS instead of clustering-only selection (ablation) | VAE-FEN RA · Fungus | F1 score: 0.9232 -> 0.9252 | Improved |
| use VAE-based feature extractor instead of AE (ablation) | RA at 30% annotation · GlaS Part B | F1 score: 0.810 -> 0.843 | Improved |
| use VAE-based feature extractor instead of DCGAN (ablation) | RA at 30% annotation · GlaS Part B | F1 score: 0.828 -> 0.843 | Improved |
| use VAE-based feature extractor instead of DCGAN (ablation) | RA at 50% annotation · GlaS Part A | Object Dice: 0.903 -> 0.897 | Hurt |
| representative annotation instead of uniform annotation of slices | Sparse-CliqueVoxNet · HVSMR 2016 | overall score: s2: -0.019 -> 0.019; s20: -0.04 -> -0.057 | Mixed |
| RA at 50% vs SA at 50% | VAE-RA · GlaS Part B | F1 score: 0.828 -> 0.862 | Improved |
| RA at 30% vs full-annotation baseline | VAE-RA · Fungus | F1 score: 0.9247 -> 0.9252 | No clear effect |
| CliqueVoxNet instead of DenseVoxNet | 3D heart segmentation · HVSMR 2016 | overall score: -0.161 -> 0.06 | Improved |
Closest papers by meaning
- An Annotation Sparsification Strategy for 3D Medical Image Segmentation via Representative Selection and Self-Training (AAAI 2020 · similarity 0.91)
- Integrating Deep Metric Learning with Coreset for Active Learning in 3D Segmentation (NeurIPS 2024 · similarity 0.89)
- Quantization of Fully Convolutional Networks for Accurate Biomedical Image Segmentation (CVPR 2018 · similarity 0.89)
- Diversified and Personalized Multi-rater Medical Image Segmentation (CVPR 2024 · similarity 0.85)
- Enhancing SAM with Efficient Prompting and Preference Optimization for Semi-supervised Medical Image Segmentation (CVPR 2025 · similarity 0.85)
- Show and Segment: Universal Medical Image Segmentation via In-Context Learning (CVPR 2025 · 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.
