arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.13901cs.LGeess.SPstat.ML
收藏

通过一种新颖的分析方法拓宽加速扩散模型的目标分布

Broadening Target Distributions for Accelerated Diffusion Models via a Novel Analysis Approach

The Ohio State University(俄亥俄州立大学);University of Kentucky(肯塔基大学)

Yuchen Liang, Peizhong Ju, Yingbin Liang, Ness Shroff

AI总结:

本文提出一种新的加速随机 DDPM 采样器,通过倾斜因子表示和 Tweedie 公式分析,将加速扩散模型的收敛保证推广到三类更广泛的目标分布,并改进维度依赖。

摘要:

加速扩散模型有望显著提高标准扩散过程的效率。理论上,这些模型已被证明比普通扩散模型的 $\\( \mathcal O(1/ε^2) \\)$ 收敛速率更快,其中 $ε$ 表示目标精度。然而,目前的理论研究仅对受限的目标分布类别建立了加速优势,例如沿整个采样路径施加平滑条件或具有有界支撑的分布。在这项工作中,我们通过一种新的加速随机 DDPM 采样器显著拓宽了目标分布类别。特别地,我们表明它对三个先前未考虑过的广泛分布类别实现了加速性能。我们的第一类仅依赖于对目标密度 $q_0$ 施加的平滑条件,这比现有的沿整个采样路径对所有 $q_t$ 施加的平滑条件要宽松得多。我们的第二类仅需要有限二阶矩条件,允许比现有有限支撑条件更广泛的目标分布类别。我们的第三类是高斯混合,对此我们的结果建立了首个加速保证。此外,在加速 DDPM 类型采样器中,我们针对有界支撑分布的结果显示了对数据维度 $d$ 的改进依赖。我们的分析引入了一种新技术,通过构造收敛误差的倾斜因子表示并利用 Tweedie 公式处理泰勒展开项来建立性能保证。这一新的分析框架可能具有独立的意义。

英文摘要:

Accelerated diffusion models hold the potential to significantly enhance the efficiency of standard diffusion processes. Theoretically, these models have been shown to achieve faster convergence rates than the standard $\mathcal O(1/ε^2)$ rate of vanilla diffusion models, where $ε$ denotes the target accuracy. However, current theoretical studies have established the acceleration advantage only for restrictive target distribution classes, such as those with smoothness conditions imposed along the entire sampling path or with bounded support. In this work, we significantly broaden the target distribution classes with a new accelerated stochastic DDPM sampler. In particular, we show that it achieves accelerated performance for three broad distribution classes not considered before. Our first class relies on the smoothness condition posed only to the target density $q_0$, which is far more relaxed than the existing smoothness conditions posed to all $q_t$ along the entire sampling path. Our second class requires only a finite second moment condition, allowing for a much wider class of target distributions than the existing finite-support condition. Our third class is Gaussian mixture, for which our result establishes the first acceleration guarantee. Moreover, among accelerated DDPM type samplers, our results specialized for bounded-support distributions show an improved dependency on the data dimension $d$. Our analysis introduces a novel technique for establishing performance guarantees via constructing a tilting factor representation of the convergence error and utilizing Tweedie's formula to handle Taylor expansion terms. This new analytical framework may be of independent interest.

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。

↑