Scholarlink

<p>In end-to-end autonomous driving, the motion prediction plays a pivotal role in ego-vehicle planning. However, existing methods often rely on globally aggregated motion features, overlooking that a few nearby agents largely determine planning decisions and represent the primary sources of risk. Insufficient attention to these critical local interactions can therefore obscure risks and compromise planning reliability. This work proposes FocalAD, a novel end-to-end autonomous driving framework that focuses on critical local neighbors and refines planning by enhancing local motion representations. Specifically, FocalAD comprises two core modules: the Ego-Local-Agents Interactor (ELAI) and the Focal-Local-Agents Loss (FLA Loss). ELAI constructs a graph-based ego-centric interaction representation that captures motion dynamics with local neighbors to enhance both ego planning and agent motion queries. FLA Loss increases the weights of decision-critical neighboring agents, guiding the model to prioritize those more relevant to planning. Extensive experiments show that FocalAD outperforms existing state-of-the-art methods on the open-loop nuScenes dataset and the closed-loop Bench2Drive benchmark. Notably, on the robustness-focused Adv-nuScenes dataset, FocalAD achieves even greater improvements, reducing the average collision rate by 41.9% compared to DiffusionDrive and by 15.6% compared to SparseDrive.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

FocalAD: Local Motion Planning for End-to-End Autonomous Driving

  • 期刊论文

  • 发表日期：

    2026年5月6日

查看全文
  • Bin Sun,
  • Boao Zhang,
  • Jiayi Lu,
  • Xinjie Feng,
  • Jiachen Shang,
  • Rui Cao,
  • Mengchao Zheng,
  • Chuanye Wang,
  • Shichun Yang,
  • Yaoguang Cao,
  • Ziying Song

摘要

In end-to-end autonomous driving, the motion prediction plays a pivotal role in ego-vehicle planning. However, existing methods often rely on globally aggregated motion features, overlooking that a few nearby agents largely determine planning decisions and represent the primary sources of risk. Insufficient attention to these critical local interactions can therefore obscure risks and compromise planning reliability. This work proposes FocalAD, a novel end-to-end autonomous driving framework that focuses on critical local neighbors and refines planning by enhancing local motion representations. Specifically, FocalAD comprises two core modules: the Ego-Local-Agents Interactor (ELAI) and the Focal-Local-Agents Loss (FLA Loss). ELAI constructs a graph-based ego-centric interaction representation that captures motion dynamics with local neighbors to enhance both ego planning and agent motion queries. FLA Loss increases the weights of decision-critical neighboring agents, guiding the model to prioritize those more relevant to planning. Extensive experiments show that FocalAD outperforms existing state-of-the-art methods on the open-loop nuScenes dataset and the closed-loop Bench2Drive benchmark. Notably, on the robustness-focused Adv-nuScenes dataset, FocalAD achieves even greater improvements, reducing the average collision rate by 41.9% compared to DiffusionDrive and by 15.6% compared to SparseDrive.