Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper elucidates the mathematical foundations of generative modeling techniques, specifically highlighting the interrelations between optimal transport, Schrödinger bridge, and flow matching. By clarifying these connections, the work provides a cohesive framework that enhances understanding of generative processes. The key result is the demonstration of how these diverse methodologies can be unified under a common theoretical umbrella, potentially leading to more robust generative models.
Unifying optimal transport with generative modeling techniques reveals deeper insights into their mathematical underpinnings and practical applications.
These notes recapitulate the high level mathematical principles behind different techniques for generative modeling. I show the connections between optimal transport and standard techniques such as Schr{ö}dinger bridge and flow matching.