[2511.23342] Overcoming the Curvature Bottleneck in MeanFlow

Computer Science > Computer Vision and Pattern Recognition arXiv:2511.23342 (cs) [Submitted on 28 Nov 2025 (v1), last revised 29 Mar 2026 (this version, v3)] Title:Overcoming the Curvature Bottleneck in MeanFlow Authors:Xinxi Zhang, Shiwei Tan, Quang Nguyen, Quan Dao, Ligong Han, Xiaoxiao He, Tunyu Zhang, Chengzhi Mao, Dimitris Metaxas, Vladimir Pavlovic View a PDF of the paper titled Overcoming the Curvature Bottleneck in MeanFlow, by Xinxi Zhang and 9 other authors View PDF HTML (experimental) Abstract:MeanFlow offers a promising framework for one-step generative modeling by directly learning a mean-velocity field, bypassing expensive numerical integration. However, we find that the highly curved generative trajectories of existing models induce a noisy loss landscape, severely bottlenecking convergence and model quality. We leverage a fundamental geometric principle to overcome this: mean-velocity estimation is drastically simpler along straight paths. Building on this insight, we propose Rectified MeanFlow, a self-distillation approach that learns the mean-velocity field over a straightened velocity field, induced by rectified couplings from a pretrained model. To further promote linearity, we introduce a distance-based truncation heuristic that prunes residual high-curvature pairs. By smoothing the optimization landscape, our method achieves strong one-step generation performance. We improve the FID of baseline MeanFlow models from 30.9 to 8.6 under same training budg...