端到端自动驾驶中的训练后处理

中文摘要 AI 辅助

将多模态输入直接映射到未来轨迹/操纵的端到端模型在自动驾驶中已成为日益突出的研究范式，包括视觉-语言-动作模型和轨迹生成规划器。与传统机器学习应用不同，自动驾驶车辆在安全关键且交互密集的环境中运行，传统的专家示范开环模仿不足以确保可靠性。小的执行错误会随时间累积，训练数据中恢复行为稀缺，逐点标签也无法捕捉安全和驾驶舒适性等长期目标。这些限制促使转向训练后技术，以进一步完善驾驶策略。本综述通过定义其范围并将现有文献按所使用的监督形式分为四个主要类别，对自动驾驶的训练后处理给出统一观点。针对每个类别，讨论其能力、局限性和开放挑战。旨在促进对这一新兴领域的系统理解，并激发未来关于可靠高效的自动驾驶训练后处理的研究。

英文摘要

End-to-end models that map multimodal inputs directly to future trajectories/maneuvers have emerged as an increasingly prominent research paradigm in autonomous driving. This class of models includes both Vision-Language-Action models and trajectory-generative planners. Unlike classic machine learning applications, autonomous vehicles operate in safety-critical and interaction-intensive environments where traditional open-loop imitation of expert demonstrations is not sufficient to ensure reliability. In particular, small execution errors can accumulate over time, while recovery behaviors are scarce in training data. In addition, long-horizon objectives such as safety and driving comfort are not captured by pointwise labels either. These limitations have motivated a shift toward post-training techniques, which further refine driving policies beyond pure imitation. This survey presents a unified view of post-training for autonomous driving by defining its scope and organizing the existing literature into four major families based on the form of supervision they use. For each family, we discuss its capabilities, limitations, and open challenges. We aim to facilitate a systematic understanding of this emerging area and stimulate future research on reliable and efficient post-training for autonomous driving.A collection of related papers is available at https://github.com/RYNing/Awesome-Post-Training-In-Autonomous-Driving-Papers.

发表机构

• Northeastern University（东北大学）
• Purdue University（普渡大学）

机构由 AI 辅助整理，请以论文原文为准。