Jean Feydy

Papers

Total Citations

H-Index

About

Jean Feydy is a leading researcher at the intersection of geometric data analysis, optimal transport, and machine learning, with a particular focus on shape matching and point cloud processing. Their most influential work, "Accurate Point Cloud Registration with Robust Optimal Transport" (2021, 26 citations), introduces a groundbreaking framework that leverages robust optimal transport solvers to dramatically improve the accuracy of both optimization-based and deep learning methods for 3D shape alignment. By demonstrating that modern OT algorithms can achieve superior registration performance at an affordable computational cost, Feydy has bridged a critical gap between theoretical optimal transport and practical geometric applications. This work has become essential reading for researchers in computer vision, medical imaging, and robotics, where precise point cloud registration is fundamental. Feydy's contributions extend beyond this paper, as they are widely recognized for developing efficient, differentiable OT solvers that have enabled new approaches in representation learning and domain adaptation. Their research exemplifies how mathematical rigor can drive practical advances, making complex geometric alignment problems both more accurate and computationally tractable for the broader scientific community.