CoIRL-AD: Collaborative-Competitive Imitation-Reinforcement Learning in Latent World Models for Autonomous Driving
By: Xiaoji Zheng , Ziyuan Yang , Yanhao Chen and more
Potential Business Impact:
Makes self-driving cars safer and better drivers.
End-to-end autonomous driving models trained solely with imitation learning (IL) often suffer from poor generalization. In contrast, reinforcement learning (RL) promotes exploration through reward maximization but faces challenges such as sample inefficiency and unstable convergence. A natural solution is to combine IL and RL. Moving beyond the conventional two-stage paradigm (IL pretraining followed by RL fine-tuning), we propose CoIRL-AD, a competitive dual-policy framework that enables IL and RL agents to interact during training. CoIRL-AD introduces a competition-based mechanism that facilitates knowledge exchange while preventing gradient conflicts. Experiments on the nuScenes dataset show an 18% reduction in collision rate compared to baselines, along with stronger generalization and improved performance on long-tail scenarios. Code is available at: https://github.com/SEU-zxj/CoIRL-AD.
Similar Papers
AD-R1: Closed-Loop Reinforcement Learning for End-to-End Autonomous Driving with Impartial World Models
CV and Pattern Recognition
Teaches self-driving cars to avoid crashes.
IN-RIL: Interleaved Reinforcement and Imitation Learning for Policy Fine-Tuning
Robotics
Teaches robots to learn faster from watching.
Unified Reinforcement and Imitation Learning for Vision-Language Models
CV and Pattern Recognition
Makes small AI understand pictures and words better.