홈 › Papers

RAD: Training an End-to-End Driving Policy via Large-Scale 3DGS-based Reinforcement Learning

2025-02-18 · Hao Gao, Shaoyu Chen, Bo Jiang, Bencheng Liao, Yiang Shi, Xiaoyang Guo, Yuechuan Pu, Haoran Yin, Xiangyu Li, Xinbang Zhang, Ying Zhang, Wenyu Liu, Qian Zhang, Xinggang Wang

Existing end-to-end autonomous driving (AD) algorithms typically follow the Imitation Learning (IL) paradigm, which faces challenges such as causal confusion and the open-loop gap. In this work, we establish a 3DGS-based closed-loop Reinforcement Learning (RL) training paradigm. 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 that guide the policy to effectively respond to safety-critical events and understand real-world causal relationships. For better alignment with human driving behavior, IL is incorporated 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, especially 3x lower collision rate. Abundant closed-loop results are presented at https://hgao-cv.github.io/RAD.

📄 PDF Abstract BibTeX arXiv:2502.13144

Code (0)

등록된 구현이 없습니다.

Tasks

3DGSAutonomous DrivingImitation LearningReinforcement Learning (RL)

Scaling Is All You Need: Autonomous Driving with JAX-Accelerated Reinforcement Learning

2023-12-23 · Moritz Harmel, Anubhav Paras, Andreas Pasternak, Nicholas Roy 외

Reinforcement learning has been demonstrated to outperform even the best humans in complex domains like video games. However, running reinforcement learning experiments on the required scale for autonomous driving is ext…

All Autonomous Driving reinforcement-learning Reinforcement Learning

CarPlanner: Consistent Auto-regressive Trajectory Planning for Large-scale Reinforcement Learning in Autonomous Driving

2025-02-27 · CVPR 2025 1 · Dongkun Zhang, Jiaming Liang, Ke Guo, Sha Lu 외

Trajectory planning is vital for autonomous driving, ensuring safe and efficient navigation in complex environments. While recent learning-based methods, particularly reinforcement learning (RL), have shown promise in sp…

Autonomous Driving Reinforcement Learning (RL)Trajectory Planning

Virtual to Real Reinforcement Learning for Autonomous Driving

2017-04-13 · Xinlei Pan, Yurong You, Ziyan Wang, Cewu Lu

Reinforcement learning is considered as a promising direction for driving policy learning. However, training autonomous driving vehicle with reinforcement learning in real environment involves non-affordable trial-and-er…

Autonomous Driving Domain Adaptation Image-to-Image Translation reinforcement-learning+5

Guided Online Distillation: Promoting Safe Reinforcement Learning by Offline Demonstration

2023-09-18 · Jinning Li, Xinyi Liu, Banghua Zhu, Jiantao Jiao 외

Safe Reinforcement Learning (RL) aims to find a policy that achieves high rewards while satisfying cost constraints. When learning from scratch, safe RL agents tend to be overly conservative, which impedes exploration an…

Autonomous Driving Decision Making reinforcement-learning Reinforcement Learning+2

End-to-End Urban Driving by Imitating a Reinforcement Learning Coach

2021-08-18 · ICCV 2021 10 · Zhejun Zhang, Alexander Liniger, Dengxin Dai, Fisher Yu 외

End-to-end approaches to autonomous driving commonly rely on expert demonstrations. Although humans are good drivers, they are not good coaches for end-to-end algorithms that demand dense on-policy supervision. On the co…

Autonomous Driving Imitation Learning reinforcement-learning Reinforcement Learning+1