{"ID":2835601,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.23342","arxiv_id":"2511.23342","title":"Overcoming the Curvature Bottleneck in MeanFlow","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 budget, and outperform the recent 2-rectified flow++ by 33.4% in FID while running 26x faster. Our work suggests that the difficulty of one-step flow generation stems partially from the rugged optimization landscapes induced by curved trajectories. Code is available at https://github.com/Xinxi-Zhang/Re-MeanFlow.","short_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 qua...","url_abs":"https://arxiv.org/abs/2511.23342","url_pdf":"https://arxiv.org/pdf/2511.23342v3","authors":"[\"Xinxi Zhang\",\"Shiwei Tan\",\"Quang Nguyen\",\"Quan Dao\",\"Ligong Han\",\"Xiaoxiao He\",\"Tunyu Zhang\",\"Chengzhi Mao\",\"Dimitris Metaxas\",\"Vladimir Pavlovic\"]","published":"2025-11-28T16:50:08Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.AI\"]","methods":"[]","has_code":false,"code_links":[{"ID":606522,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2835601,"paper_url":"https://arxiv.org/abs/2511.23342","paper_title":"Overcoming the Curvature Bottleneck in MeanFlow","repo_url":"https://github.com/Xinxi-Zhang/Re-MeanFlow","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}