Research map / Image generation / Fast diffusion sampling
Ambiguity-aware Truncated Flow Matching for Ambiguous Medical Image Segmentation
Fanding Li, Xiangyu Li, Xianghe Su, Xingyu Qiu, Suyu Dong, Wei Wang, Kuanquan Wang, Gongning Luo, Shuo Li
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 GTR to SFM (ATFM vs SFM) (ablation) | SFM · LIDC-IDRI | GED100: 0.176 -> 0.162 | Improved |
| add GTR to SFM (ATFM vs SFM) (ablation) | SFM · LIDC-IDRI | HM-IoU32: 0.631 -> 0.667 | Improved |
| add SFM flow matching to GTR (ATFM vs Act. GTR) (ablation) | Act. GTR · LIDC-IDRI | GED100: 0.230 -> 0.162 | Improved |
| add SFM flow matching to GTR (ATFM vs Act. GTR) (ablation) | Act. GTR · LIDC-IDRI | HM-IoU32: 0.550 -> 0.667 | Improved |
| add semantic consistency loss L_SF (ablation) | ATFM · LIDC-IDRI | GED100: 0.249 -> 0.162 | Improved |
| add semantic consistency loss L_SF (ablation) | ATFM · LIDC-IDRI | HM-IoU32: 0.597 -> 0.667 | Improved |
| add semantic consistency loss L_SF (ablation) | SFM · LIDC-IDRI | GED100: 0.185 -> 0.176 | Improved |
| add semantic consistency loss L_SF (ablation) | SFM · LIDC-IDRI | HM-IoU32: 0.624 -> 0.631 | Improved |
| use 25 inference steps in SFM rather than other counts (ablation) | ATFM · LIDC-IDRI | performance vs speed | Improved |
| tune alpha of L_SF between 0 and 5e-3 (ablation) | ATFM · LIDC-IDRI | performance | Mixed |
| use truncated 25-step ATFM instead of full diffusion CCDM | ATFM vs CCDM · LIDC-IDRI | inference time (s): 1100 -> 113 | Improved |
| ATFM vs runner-up on LIDC | ATFM · LIDC-IDRI | GED100: 0.183 -> 0.162 | Improved |
Closest papers by meaning
- MedFlowSeg: Flow Matching for Medical Image Segmentation with Frequency-Aware Attention (NeurIPS 2026 · similarity 0.87)
- Annotation Ambiguity Aware Semi-Supervised Medical Image Segmentation (CVPR 2025 · similarity 0.86)
- S2S2: Semantic Stacking for Robust Semantic Segmentation in Medical Imaging (AAAI 2025 · similarity 0.86)
- GapMatch: Bridging Instance and Model Perturbations for Enhanced Semi-Supervised Medical Image Segmentation (AAAI 2025 · similarity 0.84)
- Noise-Consistent Siamese-Diffusion for Medical Image Synthesis and Segmentation (CVPR 2025 · similarity 0.84)
- DyCON: Dynamic Uncertainty-aware Consistency and Contrastive Learning for Semi-supervised Medical Image Segmentation (CVPR 2025 · 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.
