Enhancing End-to-End Autonomous Driving with Latent World Model

Introduction

We propose a LAtent World model (LAW) for self-supervised learning that enhances the training of end-to-end autonomous driving.

Pipeline

Initailly, we develop an end-to-end driving framework to extract view latents and predict waypoints. Then, we predict the view latents of the next frame via the latent world model. The predicted view latent is supervised by the observed view latent of the next frame.

Citation

Please consider citing our work as follows if it is helpful.

@misc{li2024enhancing,
      title={Enhancing End-to-End Autonomous Driving with Latent World Model}, 
      author={Yingyan Li and Lue Fan and Jiawei He and Yuqi Wang and Yuntao Chen and Zhaoxiang Zhang and Tieniu Tan},
      year={2024},
      eprint={2406.08481},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}