World models have demonstrated significant promise for data synthesis in autonomous driving. However, existing methods predominantly concentrate on single-modality generation, typically focusing on either multi-camera video or LiDAR sequence synthesis. In this paper, we propose UniDriveDreamer, a single-stage unified multimodal world model for autonomous driving, which directly generates multimodal future observations without relying on intermediate representations or cascaded modules. Our framework introduces a LiDAR-specific variational autoencoder (VAE) designed to encode input LiDAR sequences, alongside a video VAE for multi-camera images. To ensure cross-modal compatibility and training stability, we propose Unified Latent Anchoring (ULA), which explicitly aligns the latent distributions of the two modalities. The aligned features are fused and processed by a diffusion transformer that jointly models their geometric correspondence and temporal evolution. Additionally, structured scene layout information is projected per modality as a conditioning signal to guide the synthesis. Extensive experiments demonstrate that UniDriveDreamer outperforms previous state-of-the-art methods in both video and LiDAR generation, while also yielding measurable improvements in downstream
@article{arxiv.2602.02002,
title = {UniDriveDreamer: A Single-Stage Multimodal World Model for Autonomous Driving},
author = {Guosheng Zhao and Yaozeng Wang and Xiaofeng Wang and Zheng Zhu and Tingdong Yu and Guan Huang and Yongchen Zai and Ji Jiao and Changliang Xue and Xiaole Wang and Zhen Yang and Futang Zhu and Xingang Wang},
journal= {arXiv preprint arXiv:2602.02002},
year = {2026}
}