Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2601.22965 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918534171328512 |
|---|---|
| author | Zhang, Runhua Hou, Junyi Cheng, Changxu Chen, Qiyi Wang, Tao Zhao, Wuyue |
| author_facet | Zhang, Runhua Hou, Junyi Cheng, Changxu Chen, Qiyi Wang, Tao Zhao, Wuyue |
| contents | Diffusion policies (DP) have demonstrated significant potential in visual navigation by capturing diverse multi-modal trajectory distributions. However, standard imitation learning (IL), which most DP methods rely on for training, often inherits sub-optimality and redundancy from expert demonstrations, thereby necessitating a computationally intensive "generate-then-filter" pipeline that relies on auxiliary selectors during inference. To address these challenges, we propose Self-Imitated Diffusion Policy (SIDP), a novel framework that learns improved planning by selectively imitating a set of trajectories sampled from itself. Specifically, SIDP introduces a reward-guided self-imitation mechanism that encourages the policy to consistently produce high-quality trajectories efficiently, rather than outputs of inconsistent quality, thereby reducing reliance on extensive sampling and post-filtering. During training, we employ a reward-driven curriculum learning paradigm to mitigate inefficient data utility, and goal-agnostic exploration for trajectory augmentation to improve planning robustness. Extensive evaluations on a comprehensive simulation benchmark show that SIDP significantly outperforms previous methods, with real-world experiments confirming its effectiveness across multiple robotic platforms. On Jetson Orin Nano, SIDP delivers a 2.5$\times$ faster inference than the baseline NavDP, i.e., 110ms VS 273ms, enabling efficient real-time deployment. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_22965 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Self-Imitated Diffusion Policy for Efficient and Robust Visual Navigation Zhang, Runhua Hou, Junyi Cheng, Changxu Chen, Qiyi Wang, Tao Zhao, Wuyue Robotics Diffusion policies (DP) have demonstrated significant potential in visual navigation by capturing diverse multi-modal trajectory distributions. However, standard imitation learning (IL), which most DP methods rely on for training, often inherits sub-optimality and redundancy from expert demonstrations, thereby necessitating a computationally intensive "generate-then-filter" pipeline that relies on auxiliary selectors during inference. To address these challenges, we propose Self-Imitated Diffusion Policy (SIDP), a novel framework that learns improved planning by selectively imitating a set of trajectories sampled from itself. Specifically, SIDP introduces a reward-guided self-imitation mechanism that encourages the policy to consistently produce high-quality trajectories efficiently, rather than outputs of inconsistent quality, thereby reducing reliance on extensive sampling and post-filtering. During training, we employ a reward-driven curriculum learning paradigm to mitigate inefficient data utility, and goal-agnostic exploration for trajectory augmentation to improve planning robustness. Extensive evaluations on a comprehensive simulation benchmark show that SIDP significantly outperforms previous methods, with real-world experiments confirming its effectiveness across multiple robotic platforms. On Jetson Orin Nano, SIDP delivers a 2.5$\times$ faster inference than the baseline NavDP, i.e., 110ms VS 273ms, enabling efficient real-time deployment. |
| title | Self-Imitated Diffusion Policy for Efficient and Robust Visual Navigation |
| topic | Robotics |
| url | https://arxiv.org/abs/2601.22965 |