Julie Delon

Julie Delon is a faculty member at Université Paris-Cité, France, recognized internationally for her expertise in optimal transport theory and its applications to Gaussian mixture models. In March 2024 she delivered an invited workshop presentation at Humboldt-University of Berlin on “Optimal Transport with Invariances between Gaussian Mixture Models,” outlining novel theoretical contributions and practical algorithms for computing Wasserstein and Gromov-Wasserstein distances between GMMs.

Her research spans the mathematical foundations of OT, including barycenters, geodesic spaces, and invariant transport plans, and extends to concrete applications such as texture synthesis, color transfer, and evaluation metrics for generative models (FID). The work integrates rigorous theory with efficient computational schemes, offering new tools for machine-learning practitioners and probabilistic modelers.