Nicolas Perrin-Gilbert

Centre National de la Recherche Scientifique, Institut Systèmes Intelligents et de Robotique

Papers

Total Citations

H-Index

About

Nicolas Perrin-Gilbert is an emerging researcher working at the intersection of robotics, reinforcement learning, and evolutionary computation. His work addresses some of the most challenging problems in autonomous robot learning, particularly the difficulty of training agents to complete long-horizon tasks where rewards are sparse and infrequent — a significant bottleneck in real-world robotics deployment. His 2022 paper "Divide & Conquer Imitation Learning" tackles this challenge head-on by leveraging Imitation Learning within a Deep Reinforcement Learning framework, offering a promising strategy to bootstrap learning in otherwise intractable settings. Complementing this, his work on "Assessing Quality-Diversity Neuro-Evolution Algorithms Performance in Hard Exploration Problems" demonstrates a broader interest in bio-inspired approaches, evaluating how Quality-Diversity evolutionary methods — algorithms that seek collections of high-performing, diverse solutions — fare in complex exploration scenarios relevant to robotics and beyond. Though early in his citation trajectory, with his 2022 publications accumulating citations that reflect a growing community interest, Perrin-Gilbert's research speaks to fundamental open problems in AI and robotics. His dual focus on imitation-based and evolution-inspired learning positions him as a contributor bridging classical evolutionary computation with modern deep learning approaches to autonomous systems.