Optimal Flow Matching: Learning Straight Trajectories in Just One Step

arXiv:2403.13117v1 Announce Type: cross
Abstract: Over the several recent years, there has been a boom in development of flow matching 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 fast integration of the learned flow's paths. Unfortunately, most existing flow straightening methods are based on non-trivial iterative procedures which accumulate the error during training or exploit heuristic minibatch …

abstract arxiv boom community cs.lg development flow generative generative modeling learn modeling property stat.ml transport type