Reproduce “DiffOSeg: Omni Medical Image Segmentation via Multi-Expert Collaboration Diffusion Model” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Han Zhang, Xiangde Luo, Yong Chen, Kang Li
DiffOSeg models plausible consensus masks and expert-specific segmentation styles from multi-rater medical images. In Stage I, categorical diffusion learns a distribution of masks by varying expert contributions; Stage II uses adaptive prompts to condition iterative mask generation on an expert identity. On LIDC-IDRI, its GED₃₀ is 0.0773, versus 0.1375 for D-Persona. On NPC-170, GED improves but threshold-aware Dice slightly declines versus D-Persona.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.