Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913002109796352 |
|---|---|
| author | Ge, Xingtong Zhang, Yi Huang, Yushi He, Dailan Wang, Xiahong Ma, Bingqi Song, Guanglu Liu, Yu Zhang, Jun |
| author_facet | Ge, Xingtong Zhang, Yi Huang, Yushi He, Dailan Wang, Xiahong Ma, Bingqi Song, Guanglu Liu, Yu Zhang, Jun |
| contents | Distilling video generation models to extremely low inference budgets (e.g., 2--4 NFEs) is crucial for real-time deployment, yet remains challenging. Trajectory-style consistency distillation often becomes conservative under complex video dynamics, yielding an over-smoothed appearance and weak motion. Distribution matching distillation (DMD) can recover sharp, mode-seeking samples, but its local training signals do not explicitly regularize how denoising updates compose across timesteps, making composed rollouts prone to drift. To overcome this challenge, we propose Self-Consistent Distribution Matching Distillation (SC-DMD), which explicitly regularizes the endpoint-consistent composition of consecutive denoising updates. For real-time autoregressive video generation, we further treat the KV cache as a quality parameterized condition and propose Cache-Distribution-Aware training. This training scheme applies SC-DMD over multi-step rollouts and introduces a cache-conditioned feature alignment objective that steers low-quality outputs toward high-quality references. Across extensive experiments on both non-autoregressive backbones (e.g., Wan~2.1) and autoregressive real-time paradigms (e.g., Self Forcing), our method, dubbed \textbf{Salt}, consistently improves low-NFE video generation quality while remaining compatible with diverse KV-cache memory mechanisms. Source code will be released at \href{https://github.com/XingtongGe/Salt}{https://github.com/XingtongGe/Salt}. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_03118 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video Generation Ge, Xingtong Zhang, Yi Huang, Yushi He, Dailan Wang, Xiahong Ma, Bingqi Song, Guanglu Liu, Yu Zhang, Jun Computer Vision and Pattern Recognition Image and Video Processing Distilling video generation models to extremely low inference budgets (e.g., 2--4 NFEs) is crucial for real-time deployment, yet remains challenging. Trajectory-style consistency distillation often becomes conservative under complex video dynamics, yielding an over-smoothed appearance and weak motion. Distribution matching distillation (DMD) can recover sharp, mode-seeking samples, but its local training signals do not explicitly regularize how denoising updates compose across timesteps, making composed rollouts prone to drift. To overcome this challenge, we propose Self-Consistent Distribution Matching Distillation (SC-DMD), which explicitly regularizes the endpoint-consistent composition of consecutive denoising updates. For real-time autoregressive video generation, we further treat the KV cache as a quality parameterized condition and propose Cache-Distribution-Aware training. This training scheme applies SC-DMD over multi-step rollouts and introduces a cache-conditioned feature alignment objective that steers low-quality outputs toward high-quality references. Across extensive experiments on both non-autoregressive backbones (e.g., Wan~2.1) and autoregressive real-time paradigms (e.g., Self Forcing), our method, dubbed \textbf{Salt}, consistently improves low-NFE video generation quality while remaining compatible with diverse KV-cache memory mechanisms. Source code will be released at \href{https://github.com/XingtongGe/Salt}{https://github.com/XingtongGe/Salt}. |
| title | Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video Generation |
| topic | Computer Vision and Pattern Recognition Image and Video Processing |
| url | https://arxiv.org/abs/2604.03118 |