arXiv 2510.241082026-07-09cs.ROcs.CV

基于轨迹评分器的零人类示范端到端自动驾驶

Zero-Human Demonstration End-to-end Autonomous Driving with Trajectory Scorer

Institute of Trustworthy Embodied AI（可信具身人工智能研究所）；Fudan University（复旦大学）；Shanghai Key Laboratory of Multimodal Embodied AI（上海多模态具身人工智能重点实验室）；NVIDIA；University of Michigan（密歇根大学）；East China Normal University（华东师范大学）

Zhenxin Li, Nadine Chang, Wenhao Yao, Xinglong Sun, Zi Wang, Maying Shen, Jingde Chen, Jingyu Song, Kailin Li, Zuxuan Wu, Shiyi Lan, Jose M. Alvarez

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

研究提出ZTRS，一种基于强化学习的端到端自动驾驶规划范式，仅依靠真实世界图像和规则奖励训练，无需人类示范。通过详尽策略优化，能更好推广到长尾驾驶场景，在相关数据集上展现出优于模仿学习方法的性能。

摘要：

人类示范被广泛视为端到端自动驾驶的基石，但在长尾和安全关键场景中稀缺。当前的端到端自动驾驶训练范式仍依赖人类示范，强化学习虽有潜力但多数现有方法仍隐含依赖。本文提出ZTRS，一种基于强化学习的端到端规划范式，仅用真实世界图像和基于规则的奖励训练，完全无需人类示范。通过提出的详尽策略优化，ZTRS能更好地推广到长尾驾驶场景，在相关数据集上优于模仿学习方法。

英文摘要：

Human demonstrations are widely considered the cornerstone of end-to-end (E2E) autonomous driving despite human demonstration's scarcity for long-tail and safety-critical scenarios. Nonetheless, current E2E autonomous driving (AD) training paradigms continue to rely on human demonstrations. Imitation learning (IL) requires human demonstrations for training, whereas reinforcement learning (RL) has emerged as a promising alternative to reduce this dependency. However, most existing RL methods for E2E AD still rely implicitly on human demonstrations. A pure rewards-based RL method can overcome the need for human demonstrations, but general RL policy gradient methods suffer from the cold-start problem. In this paper, we propose ZTRS (Zero-human demonstration end-to-end autonomous driving with TRajectory Scorer) - a complete RL-based E2E planning paradigm trained solely on real-world images and rule-based rewards, entirely without human demonstration. Through our proposed Exhaustive Policy Optimization (EPO), a policy gradient variant tailored for enumerable trajectory actions and dense supervision, ZTRS enables the model to generalize better to long-tail driving scenarios. We demonstrate this generalization through our SOTA performance against IL approaches on both long-tail Navhard and closed-loop HUGSIM datasets. Project page: https://zhenxinli.net/ZTRS/.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。