基于扩散模型的自动驾驶灵活引导规划
Diffusion-Based Planning for Autonomous Driving with Flexible Guidance
AI总结：
针对自动驾驶中类人驾驶行为难以实现的问题，提出基于 Transformer 的 Diffusion Planner，通过统一架构联合建模预测与规划任务，支持多模态行为与车辆协作，无需规则优化即可确保轨迹质量，在 nuPlan 及配送车辆数据集上实现了先进闭环性能与稳健迁移性。
摘要：
在复杂开放环境中实现类人驾驶行为是自动驾驶领域的关键挑战。当前基于学习的规划方法（如模仿学习）因适应性有限、难以学习人类规划中常见的复杂多模态行为，常难以平衡竞争目标且缺乏安全保障，还严重依赖预定义规则的 fallback 策略。本文提出一种新型基于 Transformer 的 Diffusion Planner 用于闭环规划，可有效建模多模态驾驶行为，无需规则优化即可确保轨迹质量。该模型支持在同一架构下联合建模预测与规划任务，实现车辆间的协作行为；此外，通过学习轨迹得分函数的梯度并采用灵活的分类器引导机制，Diffusion Planner 能有效实现安全且适应性强的规划行为。在大规模真实世界自动驾驶规划基准 nuPlan 及新收集的 200 小时配送车辆驾驶数据集上的评估表明，Diffusion Planner 取得了最先进的闭环性能，且在多种驾驶风格下具有稳健的迁移性。
英文摘要：
Achieving human-like driving behaviors in complex open-world environments is a critical challenge in autonomous driving. Contemporary learning-based planning approaches such as imitation learning methods often struggle to balance competing objectives and lack of safety assurance,due to limited adaptability and inadequacy in learning complex multi-modal behaviors commonly exhibited in human planning, not to mention their strong reliance on the fallback strategy with predefined rules. We propose a novel transformer-based Diffusion Planner for closed-loop planning, which can effectively model multi-modal driving behavior and ensure trajectory quality without any rule-based refinement. Our model supports joint modeling of both prediction and planning tasks under the same architecture, enabling cooperative behaviors between vehicles. Moreover, by learning the gradient of the trajectory score function and employing a flexible classifier guidance mechanism, Diffusion Planner effectively achieves safe and adaptable planning behaviors. Evaluations on the large-scale real-world autonomous planning benchmark nuPlan and our newly collected 200-hour delivery-vehicle driving dataset demonstrate that Diffusion Planner achieves state-of-the-art closed-loop performance with robust transferability in diverse driving styles.
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。
