arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.04468cs.LGeess.IVmath.PR
收藏

扩散模型的迭代重要性微调

Iterative Importance Fine-tuning of Diffusion Models

University College London(伦敦大学学院);University of Cambridge(剑桥大学);Xaira Therapeutics;ENS Paris - PSL & Inria Mokaplan(巴黎高等师范学院 - PSL & Inria Mokaplan)

Alexander Denker, Shreyas Padhy, Francisco Vargas, Johannes Hertrich

AI总结:

本文提出了一种基于自监督学习的扩散模型微调方法,通过迭代优化控制实现高效的条件采样,适用于图像生成和逆问题解决。

摘要:

扩散模型是生成建模的重要工具,在成像和蛋白质设计等应用中作为有效的先验。在应用扩散模型进行下游任务时,一个关键挑战是高效地从结果后验分布中采样,这可以通过Doob的h-变换来解决。本文介绍了一种自监督算法,用于通过学习最优控制来微调扩散模型,从而实现 amortised 条件采样。我们的方法通过使用基于路径的重要性权重重新采样合成数据集,迭代地优化控制。我们在类别条件采样、逆问题和文本到图像扩散模型的奖励微调中展示了该框架的有效性。

英文摘要:

Diffusion models are an important tool for generative modelling, serving as effective priors in applications such as imaging and protein design. A key challenge in applying diffusion models for downstream tasks is efficiently sampling from resulting posterior distributions, which can be addressed using Doob's $h$-transform. This work introduces a self-supervised algorithm for fine-tuning diffusion models by learning the optimal control, enabling amortised conditional sampling. Our method iteratively refines the control using a synthetic dataset resampled with path-based importance weights. We demonstrate the effectiveness of this framework on class-conditional sampling, inverse problems and reward fine-tuning for text-to-image 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。

↑