SNAP: Semantic Navigation Among Pedestrians on arXiv
Our paper SNAP: Semantic Navigation Among Pedestrians is now available on arXiv.
SNAP is a reinforcement learning navigation policy that jointly reasons about semantic terrain and dynamic obstacles from a single semantically annotated bird’s-eye-view map, without explicit human-state estimation. It outperforms established crowd-navigation and semantic-aware baselines and transfers zero-shot to a legged robot. Joint work with Jakob Wolfram, Fan Yang, Cesar Cadena and Marco Hutter.