Reproduce “Reflection-Robust Stereo Visual-Inertial Odometry via Kalman-Filtered Ground-Height Estimation” 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

Uihyun Hwang, Sungjae Shin, Dongjae Kim, Hyun Myung

The authors present a stereo visual-inertial odometry system that selectively rejects reflection-induced features on reflective floors. It estimates ground height from stereo features and inertial pose, then removes floor-region points inferred to lie below ground beyond a margin based on uncertainty. On E4, VINS-Fusion ATE RMSE falls from 0.722 to 0.335 m, but accuracy improves on six of eight sequences.

Summary by alphaXiv.

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