arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2403.20116cs.RO
收藏

LeGo-Drive:语言增强的面向目标的闭环端到端自动驾驶

LeGo-Drive: Language-enhanced Goal-oriented Closed-Loop End-to-End Autonomous Driving

The International Institute of Information Technology, Hyderabad(海得拉巴国际信息技术学院);University of Tartu, Estonia(爱沙尼亚塔尔图大学)

Pranjal Paul, Anant Garg, Tushar Choudhary, Arun Kumar Singh, K. Madhava Krishna

AI总结:

LeGo-Drive提出以语言命令估计目标位置作为中间表示,并通过端到端训练迭代优化目标与轨迹,在模拟环境中实现81%的目标到达成功率,验证了其在自动驾驶中的有效性。

摘要:

现有的视觉-语言模型(VLMs)基于其丰富的场景理解能力,要么估计长期轨迹航路点,要么估计一组控制动作,作为闭环规划的被动响应方案。然而,这些估计较为粗糙,并且受制于其“世界理解”,可能因感知误差而产生次优决策。在本文中,我们提出了LeGo-Drive,旨在通过在端到端设置中根据给定的语言命令估计一个目标位置作为中间表示来解决这一问题。估计的目标可能落在不理想的区域,例如在执行类似停车的命令时落在汽车顶部,从而导致规划不充分。因此,我们提出以端到端的方式训练该架构,从而实现对目标和轨迹的共同迭代优化。我们通过在多种模拟环境中进行的全面实验验证了所提方法的有效性。我们报告了标准自动驾驶指标的显著提升,目标到达成功率达到81%。我们进一步展示了LeGo-Drive在不同驾驶场景和语言输入下的通用性,凸显了其在自动驾驶车辆和智能交通系统中实际部署的潜力。

英文摘要:

Existing Vision-Language models (VLMs) estimate either long-term trajectory waypoints or a set of control actions as a reactive solution for closed-loop planning based on their rich scene comprehension. However, these estimations are coarse and are subjective to their "world understanding" which may generate sub-optimal decisions due to perception errors. In this paper, we introduce LeGo-Drive, which aims to address this issue by estimating a goal location based on the given language command as an intermediate representation in an end-to-end setting. The estimated goal might fall in a non-desirable region, like on top of a car for a parking-like command, leading to inadequate planning. Hence, we propose to train the architecture in an end-to-end manner, resulting in iterative refinement of both the goal and the trajectory collectively. We validate the effectiveness of our method through comprehensive experiments conducted in diverse simulated environments. We report significant improvements in standard autonomous driving metrics, with a goal reaching Success Rate of 81%. We further showcase the versatility of LeGo-Drive across different driving scenarios and linguistic inputs, underscoring its potential for practical deployment in autonomous vehicles and intelligent transportation systems.

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。

↑