Guillaume Charpiat

Guillaume Charpiat is a Research Officer at INRIA Saclay and Head of the Data Science department at the LISN laboratory (Paris-Saclay University). He holds a PhD in computer vision from the École Polytechnique (2006), conducted under Olivier Faugeras and Renaud Keriven. His postdoctoral research at the Max Planck Institute for Biological Cybernetics focused on statistical learning. He joined the TAO/TAU team in 2015, specializing in machine learning and optimization. His research spans deep learning theory, neural networks, and applications in satellite imagery, genetics, fluid dynamics, and medical imaging.

Education: PhD in Computer Vision (École Polytechnique, 2006), Postdoc in Statistical Learning (Max Planck Institute). Teaching includes advanced deep learning courses at MVA/CentraleSupélec and AI master's programs. Research interests emphasize neural architecture design, generative models, and interdisciplinary applications. His work bridges machine learning with physics and biology, addressing challenges like dynamical systems modeling and population genetics inference.

Key contributions include neural architecture growth strategies, equivariant GNNs for materials science, and deep learning frameworks for genomic data. Advised over 20 students across PhD and master's levels, focusing on topics like neural networks, causal discovery, and generative models. Active in seminars and interdisciplinary collaborations, he co-leads the TAU team's research on machine learning and optimization.