Score: 1

DriveLaW:Unifying Planning and Video Generation in a Latent Driving World

Published: December 29, 2025 | arXiv ID: 2512.23421v1

By: Tianze Xia , Yongkang Li , Lijun Zhou and more

Potential Business Impact:

Helps self-driving cars plan safer, smarter routes.

Business Areas:
Autonomous Vehicles Transportation

World models have become crucial for autonomous driving, as they learn how scenarios evolve over time to address the long-tail challenges of the real world. However, current approaches relegate world models to limited roles: they operate within ostensibly unified architectures that still keep world prediction and motion planning as decoupled processes. To bridge this gap, we propose DriveLaW, a novel paradigm that unifies video generation and motion planning. By directly injecting the latent representation from its video generator into the planner, DriveLaW ensures inherent consistency between high-fidelity future generation and reliable trajectory planning. Specifically, DriveLaW consists of two core components: DriveLaW-Video, our powerful world model that generates high-fidelity forecasting with expressive latent representations, and DriveLaW-Act, a diffusion planner that generates consistent and reliable trajectories from the latent of DriveLaW-Video, with both components optimized by a three-stage progressive training strategy. The power of our unified paradigm is demonstrated by new state-of-the-art results across both tasks. DriveLaW not only advances video prediction significantly, surpassing best-performing work by 33.3% in FID and 1.8% in FVD, but also achieves a new record on the NAVSIM planning benchmark.

Country of Origin
🇨🇳 China

Page Count
17 pages

Category
Computer Science:
CV and Pattern Recognition