面向运动的缓存机制用于高效的自回归视频生成
Motion-Aware Caching for Efficient Autoregressive Video Generation
AI总结：
本文提出MotionCache，通过利用帧间差异作为轻量级像素运动特征代理，优化自回归视频生成中的缓存重用策略，实现6.28倍和1.64倍的速度提升，同时保持生成质量。
摘要：
自回归视频生成范式在长视频合成中具有理论潜力，但其实际部署受序列迭代去噪计算负担的限制。尽管缓存重用策略可通过跳过冗余去噪步骤加速生成，但现有方法依赖于粗粒度的块级跳过，无法捕捉到细粒度的像素动态。本文通过理论分析将缓存错误与残差不稳定性联系起来，提出MotionCache，一种面向运动的缓存框架，利用帧间差异作为轻量级像素运动特征代理。MotionCache采用由粗到细的策略：初始的预热阶段建立语义一致性，随后进行运动加权的缓存重用，动态调整每个token的更新频率。在SkyReels-V2和MAGI-1等先进模型上的大量实验表明，MotionCache分别实现了6.28倍和1.64倍的速度提升，同时有效保持生成质量（VBench：1%下降和0.01%下降）。代码可在https://github.com/ywlq/MotionCache获取。
英文摘要：
Autoregressive video generation paradigms offer theoretical promise for long video synthesis, yet their practical deployment is hindered by the computational burden of sequential iterative denoising. While cache reuse strategies can accelerate generation by skipping redundant denoising steps, existing methods rely on coarse-grained chunk-level skipping that fails to capture fine-grained pixel dynamics. This oversight is critical: pixels with high motion require more denoising steps to prevent error accumulation, while static pixels tolerate aggressive skipping. We formalize this insight theoretically by linking cache errors to residual instability, and propose MotionCache, a motion-aware cache framework that exploits inter-frame differences as a lightweight proxy for pixel-level motion characteristics. MotionCache employs a coarse-to-fine strategy: an initial warm-up phase establishes semantic coherence, followed by motion-weighted cache reuse that dynamically adjusts update frequencies per token. Extensive experiments on state-of-the-art models like SkyReels-V2 and MAGI-1 demonstrate that MotionCache achieves significant speedups of $\textbf{6.28}\times$ and $\textbf{1.64}\times$ respectively, while effectively preserving generation quality (VBench: $1\%\downarrow$ and $0.01\%\downarrow$ respectively). The code is available at https://github.com/ywlq/MotionCache.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
