Reproduce “Improving SAM for Camouflaged Object Detection via Dual Stream Adapters” 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
Jiaming Liu, Linghe Kong, Guihai Chen
SAM-COD segments camouflaged objects from paired RGB and depth inputs processed in separate streams. Adapters refine both representations, while bidirectional distillation and mixed prompt embedding coordinate features before combining mask predictions. On COD10K with RGB-D input, SAM-COD reports a maximum F-measure of 0.873, versus 0.807 for DSAM; this cross-method comparison does not isolate the cause, and RGB-only performance is lower.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.