PanoDP: Learning Collision-Free Navigation with Panoramic Depth and Differentiable Physics
This work provides a more robust and efficient navigation solution for autonomous agents operating in complex, dynamic environments, which is a significant problem for robotics and autonomous systems.
This paper addresses collision-free navigation in cluttered environments by introducing PanoDP, a framework that integrates four-view panoramic depth perception with differentiable-physics-based training signals. PanoDP significantly improves collision-free and completion rates compared to single-view and non-physics-guided baselines across various settings, including high agent counts and obstacle densities.
Autonomous collision-free navigation in cluttered environments requires safe decision-making under partial observability with both static structure and dynamic obstacles. We present \textbf{PanoDP}, a communication-free learning framework that combines four-view panoramic depth perception with differentiable-physics-based training signals. PanoDP encodes panoramic depth using a lightweight CNN and optimizes policies with dense differentiable collision and motion-feasibility terms, improving training stability beyond sparse terminal collisions. We evaluate PanoDP on a controlled ring-to-center benchmark with systematic sweeps over agent count, obstacle density/layout, and dynamic behaviors, and further test out-of-distribution generalization in an external simulator (e.g., AirSim). Across settings, PanoDP increases collision-free and completion rates over single-view and non-physics-guided baselines under matched training budgets, and ablations (view masking, rotation augmentation) confirm the policy leverages 360-degree information. Code will be open source upon acceptance.