arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.02150cs.LG
收藏

流匹配的引导方法

On the Guidance of Flow Matching

Department of Artificial Intelligence, Westlake University, Hangzhou, China(人工智能系,西湖大学,杭州,中国);Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing, China(数学与系统科学研究院,中国科学院,北京,中国)

Ruiqi Feng, Chenglei Yu, Wenhao Deng, Peiyan Hu, Tailin Wu

AI总结:

本文提出首个通用流匹配引导框架,推导出免训练精确引导、训练损失及两类近似引导方法,并在合成数据、图像逆问题和离线强化学习中验证其有效性。

摘要:

流匹配在从图像生成到决策制定的各种生成任务中已展现出最先进的性能,其中能量引导(下文简称引导)至关重要。然而,流匹配的引导比其前身扩散模型的引导更为通用,因此与之有显著差异。因此,通用流匹配引导中的挑战在很大程度上仍未得到充分探索。在本文中,我们提出了通用流匹配引导的首个框架。基于该框架,我们推导出一系列可应用于通用流匹配的引导技术,包括一种新的免训练渐近精确引导、用于基于训练的引导的新型训练损失,以及两类涵盖经典梯度引导方法作为特例的近似引导。我们从理论上研究了这些不同方法,以提供在不同场景下选择合适方法的实用指南。在合成数据集、图像逆问题和离线强化学习上的实验证明了我们提出的引导方法的有效性,并验证了我们流匹配引导框架的正确性。复现实验的代码可在 https://github.com/AI4Science-WestlakeU/flow_guidance 找到。

英文摘要:

Flow matching has shown state-of-the-art performance in various generative tasks, ranging from image generation to decision-making, where generation under energy guidance (abbreviated as guidance in the following) is pivotal. However, the guidance of flow matching is more general than and thus substantially different from that of its predecessor, diffusion models. Therefore, the challenge in guidance for general flow matching remains largely underexplored. In this paper, we propose the first framework of general guidance for flow matching. From this framework, we derive a family of guidance techniques that can be applied to general flow matching. These include a new training-free asymptotically exact guidance, novel training losses for training-based guidance, and two classes of approximate guidance that cover classical gradient guidance methods as special cases. We theoretically investigate these different methods to give a practical guideline for choosing suitable methods in different scenarios. Experiments on synthetic datasets, image inverse problems, and offline reinforcement learning demonstrate the effectiveness of our proposed guidance methods and verify the correctness of our flow matching guidance framework. Code to reproduce the experiments can be found at https://github.com/AI4Science-WestlakeU/flow_guidance.

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。

↑