Agnes Desolneux

Agnes Desolneux holds dual roles as a Research Professor (CNRS Directrice de recherche) at the Centre Borelli and a Professor attached to the Department of Mathematics at ENS Paris-Saclay. Her affiliations include the Centre Borelli (ex-CMLA) and CNRS laboratories, focusing on advanced interdisciplinary research in mathematical imaging and stochastic methods.

Education and Background:

• PhD in Applied Mathematics (2000), ENS Cachan
• Habilitation à Diriger des Recherches (HDR) in Applied Mathematics (2010), Université Paris Descartes

Research Interests: Her work combines statistical analysis, stochastic processes, and mathematical modeling to address challenges in image analysis. Key areas include:

• Statistical methods (a contrario framework, maximum entropy models)
• Image restoration and texture synthesis
• Determinantal point processes (DPP) and optimal transport
• Geometric analysis of random fields (curvature, Euler characteristics)
• Perceptual modeling of Gestalt principles

Recent Contributions: Her research explores cutting-edge topics such as Gromov-Wasserstein distances in Gaussian mixture models, anisotropy analysis of random fields, and federated learning techniques. Her methodologies bridge theoretical probability and applied image processing.

Awards and Recognition: No specific awards listed, but her extensive publication record in top journals like SIAM Journal on Imaging Sciences and Annals of Applied Probability highlights her scholarly impact.

Grants and Collaborations: Active in projects like StoRI (Stochastic Recipes in Imaging) and the Master MVA program, emphasizing interdisciplinary training in mathematical data science. Her work often involves collaborations with institutions like the EDMH doctoral school and the Université Paris-Saclay.

Labs and Teams: Leads research within the Centre Borelli, a CNRS-ENS Paris-Saclay hub for mathematics and computer science, fostering innovation in imaging and stochastic modeling.