SeaCache：面向扩散模型加速的频谱演化感知缓存
SeaCache: Spectral-Evolution-Aware Cache for Accelerating Diffusion Models
对做扩散/流匹配采样加速的人，可直接把「频谱演化先验」当作缓存复用判据：用 SEA 滤波器滤掉高频噪声后再算相邻时间步特征距离，替代原始特征差，从而在不训练的情况下自适应决定哪些步可以跳过。
原文依据 a training-free cache schedule that bases reuse decisions on a spectrally aligned representation
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论扩散模型是视觉生成的强有力骨干网络，但其固有的顺序去噪过程导致推理缓慢。已有方法通过基于相邻时间步之间特征距离的缓存与中间输出复用来加速采样。然而，现有缓存策略通常依赖纠缠了内容与噪声的原始特征差异。这种设计忽略了频谱演化规律，即低频结构较早出现、高频细节较晚被细化。我们提出频谱演化感知缓存（SeaCache），一种免训练的缓存调度方案，将复用决策建立在频谱对齐的表示之上。通过理论与实证分析，我们推导出频谱演化感知（SEA）滤波器，在抑制噪声的同时保留与内容相关的成分。使用经 SEA 滤波的输入特征来估计冗余度，可得到随内容自适应、同时遵循底层扩散模型频谱先验的动态调度。在多种视觉生成模型与基线方法上的大量实验表明，SeaCache 实现了最优的延迟—质量权衡。
英文原文摘要 1161 字符
Diffusion models are a strong backbone for visual generation, but their inherently sequential denoising process leads to slow inference. Previous methods accelerate sampling by caching and reusing intermediate outputs based on feature distances between adjacent timesteps. However, existing caching strategies typically rely on raw feature differences that entangle content and noise. This design overlooks spectral evolution, where low-frequency structure appears early and high-frequency detail is refined later. We introduce Spectral-Evolution-Aware Cache (SeaCache), a training-free cache schedule that bases reuse decisions on a spectrally aligned representation. Through theoretical and empirical analysis, we derive a Spectral-Evolution-Aware (SEA) filter that preserves content-relevant components while suppressing noise. Employing SEA-filtered input features to estimate redundancy leads to dynamic schedules that adapt to content while respecting the spectral priors of the underlying diffusion model. Extensive experiments on diverse visual generative models and the baselines show that SeaCache achieves state-of-the-art latency-quality trade-offs.