高斯混合流匹配模型

摘要

扩散模型将去噪分布近似为高斯分布并预测其均值，而流匹配模型则将高斯均值重新参数化为流速度。然而，由于离散化误差，它们在少步采样中表现欠佳，并且在无分类器指导（CFG）下往往产生过度饱和的颜色。为应对这些局限，我们提出了一种新颖的高斯混合流匹配（GMFlow）模型：GMFlow不预测均值，而是预测动态高斯混合（GM）参数，以捕捉多模态流速度分布，该分布可通过KL散度损失进行学习。我们证明，GMFlow推广了先前的扩散和流匹配模型，后者使用L_2去噪损失学习单一高斯分布。在推理阶段，我们推导了GM-SDE/ODE求解器，利用解析去噪分布和速度场实现精确的少步采样。此外，我们引入了一种新颖的概率指导方案，缓解了CFG的过饱和问题，提升了图像生成质量。大量实验表明，GMFlow在生成质量上持续超越流匹配基线，在ImageNet 256×256数据集上仅用6个采样步骤即达到了0.942的精度。

English

Diffusion models approximate the denoising distribution as a Gaussian and predict its mean, whereas flow matching models reparameterize the Gaussian mean as flow velocity. However, they underperform in few-step sampling due to discretization error and tend to produce over-saturated colors under classifier-free guidance (CFG). To address these limitations, we propose a novel Gaussian mixture flow matching (GMFlow) model: instead of predicting the mean, GMFlow predicts dynamic Gaussian mixture (GM) parameters to capture a multi-modal flow velocity distribution, which can be learned with a KL divergence loss. We demonstrate that GMFlow generalizes previous diffusion and flow matching models where a single Gaussian is learned with an L_2 denoising loss. For inference, we derive GM-SDE/ODE solvers that leverage analytic denoising distributions and velocity fields for precise few-step sampling. Furthermore, we introduce a novel probabilistic guidance scheme that mitigates the over-saturation issues of CFG and improves image generation quality. Extensive experiments demonstrate that GMFlow consistently outperforms flow matching baselines in generation quality, achieving a Precision of 0.942 with only 6 sampling steps on ImageNet 256times256.

高斯混合流匹配模型

Gaussian Mixture Flow Matching Models

摘要

Summary

Support

Support