Efficient Flow Matching using Latent Variables

[Submitted on 7 May 2025 (v1), last revised 13 Jun 2026 (this version, v4)]

View PDF HTML (experimental)

Abstract: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.

Submission history

From: Anirban Samaddar [view email]
[v1] Wed, 7 May 2025 14:59:23 UTC (12,961 KB)
[v2] Fri, 23 May 2025 18:49:37 UTC (18,073 KB)
[v3] Tue, 7 Oct 2025 18:10:05 UTC (17,127 KB)
[v4] Sat, 13 Jun 2026 21:56:42 UTC (17,126 KB)