Papers
8
Total Citations
64
H-Index
4
About
Fabien Benureau is a leading researcher in developmental robotics, focusing on how robots can autonomously explore and learn from their environments through behavioral diversity and sensorimotor development. His major contributions include pioneering methods for generating behavioral diversity by reusing past exploration experiences, enabling robots to discover novel interaction possibilities in unknown settings. Benureau’s work on extended active inference has advanced goal-directed planning in both simulated and physical robots, demonstrating how variational recurrent neural networks can underpin teleological action generation. His research has garnered significant attention, with his most-cited paper on behavioral diversity generation accumulating 24 citations, while his goal-directed planning study has reached 15 citations. Notably, Benureau has explored the intersection of evolution and development, proposing morphological wobbling—where robot physical characteristics oscillate during learning—to enhance behavioral performance. His French-language thesis on robot self-exploration of sensorimotor spaces (10 citations) remains a foundational text in the field. Through innovative algorithms like Reuse for motor command transfer and diversity-driven selection in multi-armed bandits, Benureau has shaped how autonomous agents efficiently bootstrap learning across tasks, making him a key figure in lifelong robot learning and developmental evolution.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3L’auto-exploration des espaces sensorimoteurs chez les robots10 citations · 2015
- 4
- 5
- 6Autonomous reuse of motor exploration trajectories2 citations · 2013
- 7Reusing motor commands to learn object interaction2 citations · 2014
- 8Morphological Wobbling Can Help Robots Learn2 citations · 2022