Boosting Latent Diffusion with Flow Matching

arXiv:2312.07360v2 Announce Type: replace
Abstract: Recently, there has been tremendous progress in visual synthesis and the underlying generative models. Here, diffusion models (DMs) stand out particularly, but lately, flow matching (FM) has also garnered considerable interest. While DMs excel in providing diverse images, they suffer from long training and slow generation. With latent diffusion, these issues are only partially alleviated. Conversely, FM offers faster training and inference but exhibits less diversity in synthesis. We demonstrate that introducing FM between the …

abstract arxiv boosting cs.cv diffusion diffusion models diverse excel flow generative generative models images progress synthesis training type visual