Magnitude-Direction Decoupling for Fast Video Generation with Flow Matching Models
초록
Flow matching models for video generation achieve impressive performance but suffer from high computational overhead due to iterative denoising. In fact, the original model is not necessary for all denoising steps, allowing some steps to use lightweight alternatives for faster sampling. However, directly using caching or lightweight models can deviate from the original denoising trajectory, resulting in suboptimal performance. Through empirical analysis, we find that lightweight models can robustly capture the magnitude components of the original model's output, while caching provides reliable directional guidance. Building on this insight, we propose the Magnitude-Direction Decoupling (MDD) method, which adaptively employs a direction-calibrated lightweight model as a substitute for the original model to accelerate inference and effectively correct deviations in the denoising trajectory. Moreover, MDD further reduces inference costs by reusing magnitude information under classifier-free guidance (CFG). As a result, MDD offers a more reliable and lightweight solution to accelerate sampling. Experiments show that MDD outperforms existing acceleration methods, delivering promising speedups (e.g., up to 2.95x on Wan2.1) while preserving high visual fidelity and content richness.
저자 (8명)
- Haonan Xu — LinkedIn 검색
- Feiyang Chen — LinkedIn 검색
- Songkui Chen — LinkedIn 검색
- Hongpeng Pan — LinkedIn 검색
- Zhefeng Wang — LinkedIn 검색
- Xinyu Duan — LinkedIn 검색
- Baoxing Huai — LinkedIn 검색
- Yang Yang — LinkedIn 검색
저자 LinkedIn 변경 추적은 추후 자동화 예정입니다. 현재는 검색 링크를 제공합니다.