Reproduce “OmniRisk: Omnidirectional Trajectory-Risk Learning for Agile Quadrotor Dynamic Avoidance” 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
Yifan He, Yang Liu, Wenhao Zhao, Hai Lin, Deping Zhang, Mingze Ma, Xin Zhou, Fei Gao and 3 more
OmniRisk trains trajectory-risk prediction offline and uses a learned planner to score 36 feasible quadrotor maneuvers onboard. Its fixed-candidate selection makes risk inference independent of dynamic-obstacle count, though measured planning latency exceeds FAPP’s in the reported benchmark. In simulation with six obstacles moving at 10 m/s, it achieved 84.0% success versus 8.0% for FAPP; this result is condition-specific.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.