Papers
18
Total Citations
421
H-Index
7
About
Antonin Raffin is a leading researcher in robotics and reinforcement learning, best known for his foundational work in open-source robotics tools and state representation learning. He is the creator of PythonRobotics (95+ citations), a widely-used collection of robotics algorithms that has become an essential resource for beginners and practitioners in autonomous navigation. Raffin’s major contributions include decoupling feature extraction from policy learning, demonstrating how state representation learning can dramatically improve sample efficiency in goal-based robotics (24 citations). He also developed the S-RL Toolbox (20 citations), providing standardized environments and evaluation metrics that have shaped the field. As a key contributor to the Open X-Embodiment project (119 citations), Raffin helped pioneer large-scale, cross-embodiment robotic learning, enabling general-purpose models to transfer skills across diverse robot platforms. His research on generalized state-dependent exploration (9 citations) addresses the critical challenge of smooth, real-world robot motion in deep reinforcement learning. Raffin’s work has not only advanced the theoretical foundations of robot learning but also produced practical, widely-adopted tools that empower researchers and students worldwide to build more capable, sample-efficient robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2Open X-Embodiment: Robotic Learning Datasets and RT-X Models101 citations · 2023
- 3PythonRobotics: a Python code collection of robotics algorithms95 citations · 2018
- 4
- 5
- 6
- 7Smooth Exploration for Robotic Reinforcement Learning9 citations · 2020
- 8
- 9
- 10Unsupervised Learning of State Representations for Multiple Tasks5 citations · 2017