paper-with-me

Papers

Optimal Flow Matching: Learning Straight Trajectories in Just One Step

2024-03-19 · Nikita Kornilov, Petr Mokrov, Alexander Gasnikov, Alexander Korotin

Over the several recent years, there has been a boom in development of Flow Matching (FM) methods for generative modeling. One intriguing property pursued by the community is the ability to learn flows with straight trajectories which realize the Optimal Transport (OT) displacements. Straightness is crucial for the fast integration (inference) of the learned flow's paths. Unfortunately, most existing flow straightening methods are based on non-trivial iterative FM procedures which accumulate the error during training or exploit heuristics based on minibatch OT. To address these issues, we develop and theoretically justify the novel \textbf{Optimal Flow Matching} (OFM) approach which allows recovering the straight OT displacement for the quadratic transport in just one FM step. The main idea of our approach is the employment of vector field for FM which are parameterized by convex functions.

📄 PDF Abstract BibTeX arXiv:2403.13117

Code (1)

jhomanik/optimal-flow-matching 공식 구현 pytorch

Similar Papers 제목 키워드 기반

Block Flow: Learning Straight Flow on Data Blocks

2025-01-20 · Zibin Wang, Zhiyuan Ouyang, Xiangyun Zhang

Flow-matching models provide a powerful framework for various applications, offering efficient sampling and flexible probability path modeling. These models are characterized by flows with low curvature in learned genera…

DiversityImage Generation

Exploring Straighter Trajectories of Flow Matching with Diffusion Guidance

2023-11-28 · Siyu Xing, Jie Cao, Huaibo Huang, Xiao-Yu Zhang 외

Flow matching as a paradigm of generative model achieves notable success across various domains. However, existing methods use either multi-round training or knowledge within minibatches, posing challenges in finding a f…

Consistency Flow Matching: Defining Straight Flows with Velocity Consistency

2024-07-02 · Ling Yang, Zixiang Zhang, Zhilong Zhang, Xingchao Liu 외

Flow matching (FM) is a general framework for defining probability paths via Ordinary Differential Equations (ODEs) to transform between noise and data samples. Recent approaches attempt to straighten these flow trajecto…

Image Generation

Lagrangian Flow Matching: A Least-Action Framework for Principled Path Design

2026-05-14 · Shukai Du, Junzhe Zhang, Yiming Li arxiv

Flow matching trains a neural velocity field by regression against a target velocity associated with a prescribed probability path connecting a simple initial distribution to the data distribution. A central design choic…

Beyond Optimal Transport: Model-Aligned Coupling for Flow Matching

2025-05-29 · Yexiong Lin, Yu Yao, Tongliang Liu

Flow Matching (FM) is an effective framework for training a model to learn a vector field that transports samples from a source distribution to a target distribution. To train the model, early FM methods use random coupl…