arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.14972cs.RO
收藏

从错误中学习:使用接管数据进行驾驶VLA的后训练

Learning from Mistakes: Post-Training for Driving VLA with Takeover Data

Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所)

Yinfeng Gao, Deqing Liu, Qichao Zhang, Yupeng Zheng, Haochen Tian, Guang Li, Hangjun Ye, Long Chen, Da-Wei Ding, Dongbin Zhao

AI总结:

本文提出TakeVLA框架,通过预接管语言监督和场景做梦方法,提升驾驶VLA的闭环性能与安全性,实验显示在Bench2Drive基准上优于SimLingo。

摘要:

当前端到端自动驾驶中的视觉-语言-动作(VLA)范式依赖于静态数据集的离线训练,使其易受分布偏移影响。最近的后训练方法通过增加高质量专家接管样本来缓解这一问题,但存在两个关键限制:监督仅限于接管时刻之后,导致策略安全余量有限,且被动偏好优化缺乏主动探索。本文提出TakeVLA,一种新型VLA后训练框架,通过两个互补创新克服这些不足。首先,引入预接管语言监督,使VLA能主动学习错误。通过明确教授模型在易出错情况下的行为,培养前瞻性思维,提前预见危险并显著扩大安全余量。其次,提出场景做梦,一种强化微调范式,在重构的接管场景中鼓励主动探索,超越单纯偏好拟合。实验表明,TakeVLA在Bench2Drive基准上实现了最先进的闭环性能,驾驶得分超越SimLingo 4.93分,安全余量提升11.76%。

英文摘要:

Current Vision-Language-Action (VLA) paradigms in end-to-end autonomous driving rely on offline training from static datasets, leaving them vulnerable to distribution shift. Recent post-training methods use takeover data to mitigate this by augmenting the dataset with high-quality expert takeover samples, yet they suffer from two key limitations: supervision restricted to the period after the takeover moments leads to policies with limited safety margins, and passive preference optimization lacks active exploration for optimal performance. In this paper, we propose TakeVLA, a novel VLA post-training framework that overcomes these shortcomings through two complementary innovations. First, we introduce pre-takeover language supervision, which allows the VLA to learn from mistakes proactively. By explicitly teaching the model about what to do in error-prone situations, we cultivate a precautionary mindset that anticipates hazards early and substantially enlarges safety margins. Second, we propose Scenario Dreaming, a reinforcement fine-tuning paradigm that operates in reconstruceted takeover scenarios, encouraging active exploration beyond mere preference fitting. Experiments on the Bench2Drive benchmark demonstrate that TakeVLA achieves state-of-the-art closed-loop performance, surpassing the strong VLA baseline SimLingo by 4.93 in driving score, with an enhanced safety margin as evidenced by an 11.76% increase in average TTC.

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。

↑