MixFlow：混合源分布优化修正流模型

MixFlow: Mixed Source Distributions Improve Rectified Flows

April 10, 2026

作者: Nazir Nayal, Christopher Wewer, Jan Eric Lenssen

cs.AI

摘要

扩散模型及其变体（如整流流）能生成多样化且高质量的图像，但仍受限于其学习到的高度弯曲的生成路径所导致的缓慢迭代采样。先前研究表明，高曲率的重要成因在于源分布（标准高斯分布）与数据分布之间的独立性。本研究通过两项互补性贡献突破这一局限：首先提出κ-FC通用公式，通过将源分布与任意信号κ进行条件关联，使其更贴合数据分布，从而突破标准高斯假设；随后提出MixFlow训练策略，通过降低生成路径曲率显著提升采样效率。该策略在固定无条件分布与基于κ-FC的分布的线性混合上训练流模型，这种简单混合增强了源分布与数据分布的对齐性，在减少采样步数的同时提升生成质量，并大幅加速训练收敛。在固定采样预算下，本训练方法相比标准整流流将生成质量FID指标平均提升12%，较先前基线提升7%。代码详见：https://github.com/NazirNayal8/MixFlow

English

Diffusion models and their variations, such as rectified flows, generate diverse and high-quality images, but they are still hindered by slow iterative sampling caused by the highly curved generative paths they learn. An important cause of high curvature, as shown by previous work, is independence between the source distribution (standard Gaussian) and the data distribution. In this work, we tackle this limitation by two complementary contributions. First, we attempt to break away from the standard Gaussian assumption by introducing κ-FC, a general formulation that conditions the source distribution on an arbitrary signal κ that aligns it better with the data distribution. Then, we present MixFlow, a simple but effective training strategy that reduces the generative path curvatures and considerably improves sampling efficiency. MixFlow trains a flow model on linear mixtures of a fixed unconditional distribution and a κ-FC-based distribution. This simple mixture improves the alignment between the source and data, provides better generation quality with less required sampling steps, and accelerates the training convergence considerably. On average, our training procedure improves the generation quality by 12\% in FID compared to standard rectified flow and 7\% compared to previous baselines under a fixed sampling budget. Code available at: https://github.com/NazirNayal8/MixFlow{https://github.com/NazirNayal8/MixFlow}