失敗から学ぶ：テイクオーバーデータを用いた運転VLAのポストトレーニング
Learning from Mistakes: Post-Training for Driving VLA with Takeover Data
自動運転のVLAモデルに対し、テイクオーバー（人間介入）データを活用した新しいポストトレーニング手法TakeVLAを提案。介入前の言語指示とシナリオ再構築による強化学習で安全マージンを拡大し、ベンチマークで最高性能を達成した。
詳しい要約
1. どんなもの？
2. 先行研究と比べてどこがすごい？
3. 技術・手法の肝は？
4. どうやって有効だと検証した？
5. 議論はある？
6. 次に読むべき論文は？
※ AIが要旨から生成した要約です。正確性は原文をご確認ください。
著者: Yinfeng Gao, Deqing Liu, Qichao Zhang, Yupeng Zheng, Haochen Tian, Guang Li, Hangjun Ye, Long Chen, Da-Wei Ding, Dongbin Zhao
分類: cs.RO
原文アブストラクト
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.
