arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2606.08309cs.LGcs.CV
收藏

分数函数所在之处:扩散的小波视角

Where the Score Lives: A Wavelet View of Diffusion

The Kempner Institute for the Study of Natural and Artificial Intelligence(肯普纳自然与人工智能研究所);Harvard University(哈佛大学)

Emma Finn, Binxu Wang, T. Anderson Keller, Demba E. Ba

AI总结:

提出基于二维正交小波基的分数函数参数化,通过数据分布矩分析揭示不同架构的归纳偏差,解释扩散模型中分数网络与数据分布的相互作用。

摘要:

基于分数的生成模型在过去十年中在生成多样化视觉上合理的图像方面取得了显著成功。在扩散建模中,包括CNN、U-Net和Transformer在内的多种架构被用作分数近似网络;然而,迄今为止,关于这些架构选择如何影响生成行为的了解相对较少。在这项工作中,为了提供对此领域的见解,我们提出了一种使用二维正交小波基展开的分数函数的解析可解参数化。特别地,我们根据数据分布的矩推导出可解释的最优分数函数。我们利用这种参数化提供了一种与架构无关的、基于矩的分析,揭示了数据分布的哪些属性对去噪最为重要。我们的分数机器足够灵活,可以部分模仿多种架构(包括U-Net和CNN)的相关归纳偏差,朝着理解不同分数架构为何表现出不同生成行为迈出了一步。由于我们的分数函数可以根据数据矩解析求解,我们可以开始理解数据分布如何与分数网络相互作用,从而产生我们在扩散模型中观察到的行为。

英文摘要:

Score-based generative models have had remarkable success over the last decade in generating a diverse set of visually plausible images. A variety of architectures including CNNs, U-Nets, and Transformers have been used as the score-approximation network in such diffusion modeling; however, to date, relatively little is known about how these architectural choices impact generative behavior. In this work, to provide insight into this area, we propose an analytically solvable parameterization of the score function using an expansion in a 2D orthogonal wavelet basis. In particular, we derive interpretable optimal score functions in terms of the moments of the data distribution. We use this parametrization to provide an architecture-agnostic, moment-based analysis that reveals which attributes of the data distribution tend to matter most for denoising. Our score machine is flexible enough to partially mimic the relevant inductive biases of multiple architectures, including U-Nets, and CNNs, taking a step towards understanding why different score architectures can exhibit distinct generative behavior. Since our score is solvable in terms of the moments of the data, we can begin to understand how the data distribution interacts with the score network to produce the behavior we observe in diffusion models.

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。

↑