跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.23134 · v1

Correcting Flows with Marginal Matching

通过边缘匹配校正流模型

首次提交 2026-11-18 01:53
版本更新 2026-11-18 01:53
雷达观测 2026-11-18 12:34
arXiv ↗PDF ↗

摘要

Flow matching models, ODE-based generative models, generate samples by gradually morphing a simple source distribution into a target distribution. However, these models still face challenges in optimally balancing computational efficiency with generation fidelity. While predictor-corrector schemes have been proven to help address this issue in diffusion models effectively, their application to flow matching models remains under-explored. In this work we present a novel training-based predictor-corrector framework for flow matching, which exploits the fact that the ODE can be easily reversed. Our key idea, termed Marginal Matching, is to correct samples at a particular time step in the forward ODE towards a matching marginal distribution of the backward ODE. This scheme proves to reduce a bound on the distance between the data distribution and the learned distribution, motivating two different implementations for the correction function. We show that our proposed method improves sample quality on CIFAR-10 and ImageNet-64, with minimal overhead in computation time.

AI 论文解读

基于摘要 · 非全文审阅

作者提出一种基于训练的流匹配预测-校正框架,利用前向 ODE 可逆性,在特定时间步将样本校正到反向 ODE 的匹配边缘分布,以提升生成质量。

正在读取账号与研究记录…

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-18 01:53 · 最新观测版本

Correcting Flows with Marginal Matching

Flow matching models, ODE-based generative models, generate samples by gradually morphing a simple source distribution into a target distribution. However, these models still face challenges in optimally balancing computational efficiency with generation fidelity. While predictor-corrector schemes have been proven to help address this issue in diffusion models effectively, their application to flow matching models remains under-explored. In this work we present a novel training-based predictor-corrector framework for flow matching, which exploits the fact that the ODE can be easily reversed. Our key idea, termed Marginal Matching, is to correct samples at a particular time step in the forward ODE towards a matching marginal distribution of the backward ODE. This scheme proves to reduce a bound on the distance between the data distribution and the learned distribution, motivating two different implementations for the correction function. We show that our proposed method improves sample quality on CIFAR-10 and ImageNet-64, with minimal overhead in computation time.

查看此版本解读 →查看此版本原文 ↗