Can I read 2025-Efficient Flow Matching Using Latent Variables on EtoBox?

2025-Efficient Flow Matching Using Latent Variables by Yingjian Wang is a document available to read on EtoBox.

What is 2025-Efficient Flow Matching Using Latent Variables about?

The document presents Latent-CFM, a novel flow matching framework that efficiently incorporates multi-modal data structures using pretrained deep latent variable models, significantly improving generation quality and reducing training time by up to 50%. Through experiments on synthetic and benchmark datasets, Latent-CFM demonstrates superior performance in generating physically accurate samples and allows for conditional image generation based on latent features. This approach addresses the limitations of e

Author

Yingjian Wang

Language

EN