Reproduce “Next-Best-View Prediction for Active Stereo Cameras and Highly Reflective Objects” 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
Jun Yang, Steven L. Waslander
The authors present a next-best-view planner to complete missing depth on reflective objects using active stereo. It predicts whether projected light will yield measurements at candidate views, combining a reflectance model with CAD pose hypotheses. On Tube Fitting, it recovers 61.2% of initially missing pixels within 2 mm, outperforming simple baselines under a three-view cap; results are strongest on non-complex shapes.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.