Can I read Contrastive Flow Matching for Generative Models on EtoBox?

Contrastive Flow Matching for Generative Models by Akash Chandra Gupta is a document available to read on EtoBox.

What is Contrastive Flow Matching for Generative Models about?

Contrastive Flow Matching (∆FM) is introduced as an enhancement to flow matching for generative modeling, specifically addressing the limitations of conditional generation by enforcing uniqueness across different conditions. This method significantly improves training speed, reduces the number of de-noising steps, and lowers FID metrics compared to traditional flow matching models. Extensive experiments demonstrate that ∆FM yields more diverse and high-quality image outputs, making it a valuable addition to

Author

Akash Chandra Gupta

Language

EN