Mean Flows for One-step Generative Modeling
This work addresses the challenge of efficient high-quality image generation for AI applications, narrowing the gap between one-step and multi-step models, though it is incremental as it builds on existing flow matching methods.
The paper tackles the problem of one-step generative modeling by proposing MeanFlow, a framework that uses average velocity to characterize flow fields, achieving an FID of 3.43 on ImageNet 256x256 with a single function evaluation, significantly outperforming previous state-of-the-art one-step models.
We propose a principled and effective framework for one-step generative modeling. We introduce the notion of average velocity to characterize flow fields, in contrast to instantaneous velocity modeled by Flow Matching methods. A well-defined identity between average and instantaneous velocities is derived and used to guide neural network training. Our method, termed the MeanFlow model, is self-contained and requires no pre-training, distillation, or curriculum learning. MeanFlow demonstrates strong empirical performance: it achieves an FID of 3.43 with a single function evaluation (1-NFE) on ImageNet 256x256 trained from scratch, significantly outperforming previous state-of-the-art one-step diffusion/flow models. Our study substantially narrows the gap between one-step diffusion/flow models and their multi-step predecessors, and we hope it will motivate future research to revisit the foundations of these powerful models.