Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2603.07644 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908872008007680 |
|---|---|
| author | Zhong, Hao Chi, Pei Zhao, Jiang Yuan, Shenghai Gao, Xuyang Nguyen, Thien-Minh Xie, Lihua |
| author_facet | Zhong, Hao Chi, Pei Zhao, Jiang Yuan, Shenghai Gao, Xuyang Nguyen, Thien-Minh Xie, Lihua |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_07644 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | PanoDP: Learning Collision-Free Navigation with Panoramic Depth and Differentiable Physics Zhong, Hao Chi, Pei Zhao, Jiang Yuan, Shenghai Gao, Xuyang Nguyen, Thien-Minh Xie, Lihua Robotics 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. |
| title | PanoDP: Learning Collision-Free Navigation with Panoramic Depth and Differentiable Physics |
| topic | Robotics |
| url | https://arxiv.org/abs/2603.07644 |