Reproduce “Temporally-Aware Supervised Contrastive Learning for Polyp Counting in Colonoscopy” 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
Luca Parolari, Andrea Cherubini, Lamberto Ballan, Carlo Biffi
The method reunites fragmented polyp tracklets in colonoscopy videos by learning representations that weight same-polyp examples by temporal distance and clustering tracklets using visual similarity and temporal adjacency. On REAL-Colon, its fragmentation rate was 2.32 ± 1.80 versus TPC’s 5.10 ± 3.43, with similar reported false-positive rates. Evaluation used ground-truth-derived tracklets and was limited to REAL-Colon.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.