AutoAWG: Adverse Weather Generation with Adaptive Multi-Controls for Automotive Videos

TL;DR

AutoAWG is a novel framework for generating realistic adverse weather videos for autonomous driving, improving style fidelity, temporal consistency, and safety-critical target preservation, thereby enhancing perception robustness.

Contribution

It introduces a semantics-guided adaptive fusion, vanishing point-anchored temporal synthesis, and masked training to advance weather video generation for autonomous driving.

Findings

AutoAWG reduces FID by 50% and FVD by 16.1% without first-frame conditioning.

With first-frame conditioning, FID and FVD are further reduced by 8.7% and 7.2%.

AutoAWG outperforms prior methods in style fidelity, temporal consistency, and semantic-structural integrity.

Abstract

Perception robustness under adverse weather remains a critical challenge for autonomous driving, with the core bottleneck being the scarcity of real-world video data in adverse weather. Existing weather generation approaches struggle to balance visual quality and annotation reusability. We present AutoAWG, a controllable Adverse Weather video Generation framework for Autonomous driving. Our method employs a semantics-guided adaptive fusion of multiple controls to balance strong weather stylization with high-fidelity preservation of safety-critical targets; leverages a vanishing point-anchored temporal synthesis strategy to construct training sequences from static images, thereby reducing reliance on synthetic data; and adopts masked training to enhance long-horizon generation stability. On the nuScenes validation set, AutoAWG significantly outperforms prior state-of-the-art methods: without first-frame conditioning, FID and FVD are relatively reduced by 50.0% and 16.1%; with first-frame conditioning, they are further reduced by 8.7% and 7.2%, respectively. Extensive qualitative and quantitative results demonstrate advantages in style fidelity, temporal consistency, and semantic--structural integrity, underscoring the practical value of AutoAWG for improving downstream perception in autonomous driving. Our code is available at: https://github.com/higherhu/AutoAWG