Reproduce “COIN: Confidence Score-Guided Distillation for Annotation-Free Cell Segmentation” 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
Sanghyun Jo, Seo Jin Lee, Seungwoo Lee, Seohyung Hong, Hyungseok Seo, Kyungsu Kim
COIN improves annotation-free cell instance segmentation by expanding candidate masks, filtering them through agreement with SAM, and training on accepted instances. It combines unsupervised pixel features with an existing model, then uses SAM as a consistency check. On MoNuSeg, SSA+COIN raises AJI from 0.259 to 0.580, but this test-set result does not prove every selected instance is correct.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.