Elie Chelly

Centre National de la Recherche Scientifique

Papers

Total Citations

H-Index

About

Elie Chelly is a researcher at the forefront of robotic manipulation, with a focus on accelerating grasp synthesis through artificial intelligence. His work primarily targets the critical bottleneck of interaction data in robotic learning, specifically for the complex task of six-degree-of-freedom (6-DoF) grasping. Chelly’s major contribution lies in demonstrating how Quality-Diversity (QD) algorithms can dramatically speed up the sampling of diverse, high-quality grasp poses, moving beyond traditional optimization methods that are often computationally prohibitive. His 2024 paper, "Speeding up 6-DoF Grasp Sampling with Quality-Diversity," has already garnered early citations, signaling its immediate impact on the field. By integrating QD with generative models, Chelly’s research paves the way for more efficient, generalizable robotic controllers that can learn from limited data. His work is particularly notable for bridging the gap between recent advances in AI—such as natural language-conditioned planning—and the practical, data-hungry reality of physical robot interaction. For students and researchers, Chelly’s approach offers a promising path toward scalable, data-efficient robotic learning.