用流匹配生成更安全多样的自动驾驶轨迹，支持实时约束控制
Beyond Imitation: Constraint-Aware Trajectory Generation with Flow Matching For End-to-End Autonomous Driving
- 基于约束流匹配生成轨迹，避免模式坍塌
- 在NavSim v2上取得51.31的EPDMS得分并获创新奖
- 可调节驾驶激进度，适合追求多样与安全的场景
端到端自动驾驶中的规划至关重要。现有模仿学习方法常因模式坍塌导致轨迹多样性不足，而生成方法又难以直接融入安全与物理约束，需额外优化阶段修正输出。为此，本文提出CATG框架，利用约束流匹配实现轨迹生成。该方法显式建模流匹配过程，天然缓解模式坍塌，并支持多种条件信号灵活引导。核心贡献在于将安全与运动学约束直接嵌入流匹配过程，确保生成轨迹满足关键规则。同时，通过参数化驾驶激进度作为生成控制信号，可精确调节轨迹风格。在NavSim v2挑战中，CATG获得第二名，EPDMS得分为51.31，并荣获创新奖。
原文摘要 · Abstract (English)
Planning is a critical component of end-to-end autonomous driving. However, prevailing imitation learning methods often suffer from mode collapse, failing to produce diverse trajectory hypotheses. Meanwhile, existing generative approaches struggle to incorporate crucial safety and physical constraints directly into the generative process, necessitating an additional optimization stage to refine their outputs. To address these limitations, we propose CATG, a novel planning framework that leverages Constrained Flow Matching. Concretely, CATG explicitly models the flow matching process, which inherently mitigates mode collapse and allows for flexible guidance from various conditioning signals. Our primary contribution is the novel imposition of explicit constraints directly within the flow matching process, ensuring that the generated trajectories adhere to vital safety and kinematic rules. Secondly, CATG parameterizes driving aggressiveness as a control signal during generation, enabling precise manipulation of trajectory style. Notably, on the NavSim v2 challenge, CATG achieved 2nd place with an EPDMS score of 51.31 and was honored with the Innovation Award.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。