Reproduce “Depth-to-RGB: Repurposing a Frozen Depth Estimator for Geometry-Guided Compositing” 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, Chae Yeon Lim, Donghwan Lee, Sihyun Kim, Soo Ye Kim, Kyungsu Kim
D2R predicts the missing scene depth for object compositing from the background, reference, and mask, then holds it fixed while generating RGB. On 139 category-disjoint AnyInsertion++ cases, it lowers AbsRel from 0.119 to 0.067 and raises PSNR from 21.6 to 24.0 dB versus an RGB-only baseline. These gains reflect agreement with estimator-derived depth, not physical accuracy.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.