Reproduce “LHNav: Long-Horizon Humanoid Navigation via Multirate Coordination and Physics-Grounded Action Data” 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
Xueyang Wang, Zehong Shen, Dingkun Wei, Huaijin Pi, Changpeng Wang, Xing Zhu, Hujun Bao, Yujun Shen and 1 more
LHNav coordinates multi-destination humanoid navigation through separate task planning, arrival verification, and whole-body control. Its task manager updates visits asynchronously as a vision-language-action policy predicts motion tokens for a decoder to convert into joint targets. The system achieves 75.56% Task SR₁.₅ on 45 simulated tasks, requiring ordered visits within 1.5 meters horizontally of targets; this is not a real-world success rate.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.