Reproduce “UA-Depth: Uncertainty-Aware Depth Restoration for Material-Agnostic Object Manipulation” 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

Rui Cao, Jia-Hui Pan, Yin Zhang, Qiwei Meng, Jiapeng Li, Gang Wang, Runsong Zhu, Wei Chen and 3 more

UA-Depth restores noisy or missing RGB-D depth by combining sensor measurements with a frozen Depth Anything V2 prior and recurrent uncertainty feedback that guides repeated corrections to visual and depth features. In zero-shot evaluation across eight datasets, it achieves a mean rank of 1.4, compared with 3.1 or worse for other methods; it ranks first on five, not all, datasets.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.