RMFlow: Refined Mean Flow by a Noise-Injection Step for Multimodal Generation
Yuhao Huang, Shih-Hsin Wang, Andrea L. Bertozzi, Bao Wang
Abstract
Mean flow (MeanFlow) enables efficient, high-fidelity image generation, yet its single-function evaluation (1-NFE) generation often cannot yield compelling results. We address this issue by introducing RMFlow, an efficient multimodal generative model that integrates a coarse 1-NFE MeanFlow transport with a subsequent tailored noise-injection refinement step. RMFlow approximates the average velocity of the flow path using a neural network trained with a new loss function that balances minimizing the Wasserstein distance between probability paths and maximizing sample likelihood. RMFlow achieves near state-of-the-art results on text-to-image, context-to-molecule, and time-series generation using only 1-NFE, at a computational cost comparable to the baseline MeanFlows.
BibTeX
@inproceedings{
huang2026rmflow,
title={{RMF}low: Refined Mean Flow by a Noise-Injection Step for Multimodal Generation},
author={Yuhao Huang and Shih-Hsin Wang and Andrea L. Bertozzi and Bao Wang},
booktitle={The Fourteenth International Conference on Learning Representations},
year={2026},
url={https://openreview.net/forum?id=p072J56yo4}
}