Tutorial on Optimal Mass Transport Theory for Computer Vision

David Gu
SUNY Stony Brook

This tutorial covers the fundamental concepts, theorems and algorithms for optimal mass transport theory. Optimal mass transport map gives a measure-preserving mapping between general shapes, which has been broadly applied in many fields from economics to medical imaging. This tutorial covers its applications in vision, such as image registration, warping, surface matching, registration, tracking and so on.

Presentation (PDF File)

Back to Graduate Summer School: Computer Vision