Sethu Vijayakumar
University of Edinburgh, University of Southern California, Turing Institute, The Alan Turing Institute, RIKEN Center for Brain Science, Interaction Design (United Kingdom), Research Organization of Information and Systems, Tokyo Institute of Technology, Laboratoire d'Analyse et d'Architecture des Systèmes, Kannur University, MVJ College of Engineering, Centre de Robotique, Shenzhen Academy of Robotics
Papers
170
Total Citations
5,247
H-Index
35
About
Sethu Vijayakumar is a pioneering researcher at the intersection of robotics, machine learning, and human motor control, whose work has fundamentally shaped how intelligent robots learn and move. Best known for his contributions to reinforcement learning for high-dimensional robotic systems, his seminal papers — including "Natural Actor-Critic" (312 citations) and "Reinforcement Learning for Humanoid Robotics" (288 citations) — established foundational frameworks for enabling robots to autonomously acquire complex motor skills. His early investigations into inverse kinematics learning (282 citations) and scalable nonparametric statistical methods (238 citations) provided practical, real-time solutions for controlling humanoid robots with remarkable computational efficiency. Vijayakumar's research extends beyond pure robotics into prosthetics, exoskeletons, and human augmentation. His work on prosthesis control (195 citations) and human-in-the-loop exoskeleton optimization (96 citations) demonstrates a sustained commitment to translating robotic intelligence into tangible human benefit. His development of Crocoddyl, an open-source optimal control framework (257 citations), reflects his dedication to building community tools that accelerate the field. With research spanning variable stiffness actuators, optimal control, and human behavior modeling, Vijayakumar's career represents a remarkably cohesive vision: understanding and replicating the elegance of human movement through intelligent machines.
Research Focus
Key Achievements
Top Papers
- 1Natural Actor-Critic312 citations · 2005
- 2Reinforcement Learning for Humanoid Robotics288 citations · 2003
- 3Learning inverse kinematics282 citations · 2002
- 4Using humanoid robots to study human behavior266 citations · 2000
- 5
- 6Scalable Techniques from Nonparametric Statistics for Real Time Robot Learning238 citations · 2002
- 7
- 8
- 9
- 10