CoIRL-AD: Collaborative–Competitive Imitation–Reinforcement Learning in Latent World Models for Autonomous Driving

Xiaoji Zheng, Ziyuan Yang, Yanhao Chen, Yuhang Peng, Yuanrong Tang, Gengyuan Liu, Bokui Chen*, Jiangtao Gong*

- Under Review

We are trying to conbine the advantage of imitation learning (converge fast) and reinforcement learning (trial and error) to enhance the generalization ability of autonomous driving models.

[ website ] [ paper ] [ code ] [ HuggingFace ]

Embodied Cognition Augmented End2End Autonomous Driving

Ling Niu, Xiaoji Zheng , Han Wang, Ziyuan Yang, Chen Zheng, Bokui Chen*, Jiangtao Gong*

NeurIPS 2025 - Accept

We collect expert/novide drivers' pair data, 1) align the EEG encoder and perception encoder via contrastive learning; 2) enhanced end-to-end AD models via integrating the cognition-aligned perception encoder.

[ paper ]

Large Language Models Powered Context-aware Motion Prediction

Xiaoji Zheng , Lixiu Wu, Zhijie Yan, Yuanrong Tang, Hao Zhao, Chen Zhong, Bokui Chen, Jiangtao Gong*

IROS 2024 - Accept

We first teach Large Language Model (GPT4-V) to understand Bird's Eye View like transportation context maps and output context information (intention, affordance, and scenario). Then, these context information will integrated into motion prediction pipeline to assists motion prediction.

[ website ] [ paper ] [ code ]

Extended VR: Exploring the Integration of VR Experiences and Real-world Engagement

Xiaoji Zheng , Shaojun Sun, Ying Cao, Jiatong Li, Ding Ding, Zhuying Li*

Designing Interactive Systems 2023 Companion (DIS'23 Companion) - Accept

In this paper, we explore whether the combination of virtual reality and ubiquitous computing can help overcome the limitation that Head-Mounted Displays based VR systems is often isolated from reality.

[ paper ]