arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2505.04486cs.CVcs.AIcs.LG
收藏

使用潜在变量的高效流匹配

Efficient Flow Matching using Latent Variables

Argonne National Laboratory（阿贡国家实验室）；KTH Royal Institute of Technology（皇家理工学院）

Anirban Samaddar, Yixuan Sun, Viktor Nilsson, Sandeep Madireddy

阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出Latent-CFM方法，利用预训练深度潜在变量模型提取数据特征作为条件，提升流匹配模型的训练效率和生成质量，在图像和物理场生成任务中优于现有方法。

摘要：

流匹配模型在概率生成模型的图像生成任务中显示出巨大潜力。然而，文献中的大多数流匹配模型在从简单源分布（如标准高斯）学习流时，并未显式利用目标数据中的潜在聚类结构。这导致学习效率低下，尤其是对于许多通常位于低维流形中的高维真实世界数据集。为此，我们提出了 $\texttt{Latent-CFM}$，它通过使用预训练的深度潜在变量模型从数据中提取的特征作为条件，提供了高效的训练策略。通过对来自多模态分布的合成数据和广泛使用的图像基准数据集的实验，我们表明，$\texttt{Latent-CFM}$ 通过采用预训练的轻量级潜在变量模型，在显著减少训练和计算量的情况下，展现出比最先进的流匹配模型更好的生成质量。除了自然图像，我们还考虑了源自物理过程的空间场的生成建模。使用二维达西流数据集，我们证明了我们的方法比竞争方法生成更物理准确的样本。此外，通过潜在空间分析，我们证明了我们的方法可用于以潜在特征为条件的条件图像生成，这增加了生成过程的可解释性。

英文摘要：

Flow matching models have shown great potential in image generation tasks among probabilistic generative models. However, most flow matching models in the literature do not explicitly utilize the underlying clustering structure in the target data when learning the flow from a simple source distribution like the standard Gaussian. This leads to inefficient learning, especially for many high-dimensional real-world datasets, which often reside in a low-dimensional manifold. To this end, we present $\texttt{Latent-CFM}$, which provides efficient training strategies by conditioning on the features extracted from data using pretrained deep latent variable models. Through experiments on synthetic data from multi-modal distributions and widely used image benchmark datasets, we show that $\texttt{Latent-CFM}$ exhibits improved generation quality with significantly less training and computation than state-of-the-art flow matching models by adopting pretrained lightweight latent variable models. Beyond natural images, we consider generative modeling of spatial fields stemming from physical processes. Using a 2d Darcy flow dataset, we demonstrate that our approach generates more physically accurate samples than competing approaches. In addition, through latent space analysis, we demonstrate that our approach can be used for conditional image generation conditioned on latent features, which adds interpretability to the generation process.

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。

↑