arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.10285cs.RO
收藏

自适应时间步长流匹配用于自动驾驶运动规划

Adaptive Time Step Flow Matching for Autonomous Driving Motion Planning

HRI;Honda Research Institute(本田研究院);Northeastern University(东北大学);Princeton University(普林斯顿大学);University of Maryland(马里兰大学);Stanford University(斯坦福大学)

Ananya Trivedi, Anjian Li, Mohamed Elnoor, Yusuf Umut Ciftci, Avinash Singh, Jovin D'sa, Sangjae Bae, David Isele, Taskin Padir, Faizan M. Tariq

AI总结:

本文提出了一种基于条件流匹配的自适应时间步长框架,用于实时自动驾驶轨迹规划,通过在线调整推理步骤数和轨迹后处理提升性能。

摘要:

自动驾驶需要对周围交通的交互进行推理。主流方法是基于专家驾驶数据集的大规模模仿学习,旨在在多样化的现实场景中进行泛化。对于在线轨迹生成,此类方法必须以实时速率运行。扩散模型在推理时需要数百次去噪步骤,导致高延迟。一致性模型缓解了这一问题,但依赖于精心调制的噪声调度来捕捉自动驾驶中常见的多模态动作分布。调整调度通常需要昂贵的重新训练。为了解决这些限制,我们提出了一种基于条件流匹配的框架,该框架联合预测周围代理的未来运动,并实时规划自身的轨迹。我们训练了一个轻量级的方差估计器,该估计器在线上选择推理步骤的数量,从而在运行时间和模仿学习性能之间取得平衡。为进一步提升乘坐质量,我们引入了一个轨迹后处理步骤,将其视为一个凸二次规划问题,具有极小的计算开销。在Waymo Open Motion Dataset上训练的框架能够执行车道变更、巡航控制和无保护左转等操作,而无需针对特定场景进行调优。我们的方法在NVIDIA RTX 3070 GPU上保持20 Hz的更新速率,使其适合在线部署。与Transformer、扩散和一致性模型基线相比,我们实现了更平滑的轨迹和更好的动态约束遵守。实验视频和代码实现可在https://flow-matching-self-driving.github.io/上找到。

英文摘要:

Autonomous driving requires reasoning about interactions with surrounding traffic. A prevailing approach is large-scale imitation learning on expert driving datasets, aimed at generalizing across diverse real-world scenarios. For online trajectory generation, such methods must operate at real-time rates. Diffusion models require hundreds of denoising steps at inference, resulting in high latency. Consistency models mitigate this issue but rely on carefully tuned noise schedules to capture the multimodal action distributions common in autonomous driving. Adapting the schedule, typically requires expensive retraining. To address these limitations, we propose a framework based on conditional flow matching that jointly predicts future motions of surrounding agents and plans the ego trajectory in real time. We train a lightweight variance estimator that selects the number of inference steps online, removing the need for retraining to balance runtime and imitation learning performance. To further enhance ride quality, we introduce a trajectory post-processing step cast as a convex quadratic program, with negligible computational overhead. Trained on the Waymo Open Motion Dataset, the framework performs maneuvers such as lane changes, cruise control, and navigating unprotected left turns without requiring scenario-specific tuning. Our method maintains a 20 Hz update rate on an NVIDIA RTX 3070 GPU, making it suitable for online deployment. Compared to transformer, diffusion, and consistency model baselines, we achieve improved trajectory smoothness and better adherence to dynamic constraints. Experiment videos and code implementations can be found at https://flow-matching-self-driving.github.io/.

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。

↑