用3D高斯表示构建真实世界数字孪生，实现端到端自动驾驶的闭环强化学习。
RAD: Training an End-to-End Driving Policy via Large-Scale 3DGS-based Reinforcement Learning
- 基于3DGS构建拟真数字世界，支持大规模试错训练
- 碰撞率比传统方法低3倍，闭环性能显著提升
- 融合模仿学习正则化，更贴近人类驾驶行为
现有端到端自动驾驶算法多采用模仿学习范式，面临因果混淆和开环缺陷。本文提出RAD，一种基于3DGS的闭环强化学习框架。通过3DGS技术构建逼真的物理世界数字孪生，使自动驾驶策略在大规模试错中探索状态空间，有效应对分布外场景。为提升安全性，设计专项奖励函数以引导策略应对安全事件并理解真实因果关系。同时引入模仿学习作为正则项，使策略更贴近人类驾驶行为。构建包含多样化未见3DGS环境的闭环评估基准。相比基于模仿学习的方法，RAD在多数闭环指标上表现更优，尤其碰撞率降低3倍。补充材料展示大量闭环实验结果。代码已开源（https://github.com/hustvl/RAD），推动后续研究。
原文摘要 · Abstract (English)
Existing end-to-end autonomous driving (AD) algorithms typically follow the Imitation Learning (IL) paradigm, which faces challenges such as causal confusion and an open-loop gap. In this work, we propose RAD, a 3DGS-based closed-loop Reinforcement Learning (RL) framework for end-to-end Autonomous Driving. By leveraging 3DGS techniques, we construct a photorealistic digital replica of the real physical world, enabling the AD policy to extensively explore the state space and learn to handle out-of-distribution scenarios through large-scale trial and error. To enhance safety, we design specialized rewards to guide the policy in effectively responding to safety-critical events and understanding real-world causal relationships. To better align with human driving behavior, we incorporate IL into RL training as a regularization term. We introduce a closed-loop evaluation benchmark consisting of diverse, previously unseen 3DGS environments. Compared to IL-based methods, RAD achieves stronger performance in most closed-loop metrics, particularly exhibiting a 3x lower collision rate. Abundant closed-loop results are presented in the supplementary material. Code is available at https://github.com/hustvl/RAD for facilitating future research.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。