Shin Ishii
Nara Institute of Science and Technology, Centre for Research in Engineering Surface Technology, Advanced Telecommunications Research Institute International, Research Organization of Information and Systems, RIKEN Center for Brain Science, Kyoto University, Japan Science and Technology Agency
Papers
17
Total Citations
656
H-Index
9
About
Shin Ishii is a leading researcher at the intersection of reinforcement learning, robotics, and computational neuroscience. His work centers on developing adaptive control systems for bipedal locomotion, combining central pattern generators (CPGs) with reinforcement learning algorithms to create more natural and stable walking robots. Ishii's major contributions include pioneering the CPG-actor-critic method, which integrates biological locomotion principles with machine learning, enabling biped robots to achieve quasi-passive dynamic walking—a human-like gait requiring minimal energy. His most cited work, the "On-line EM Algorithm for the Normalized Gaussian Network" (280 citations), introduced a powerful framework for probabilistic function approximation. Across his career, Ishii has demonstrated how reinforcement learning can solve complex continuous control problems, from balancing the Acrobot to developing brain-computer interfaces for smart home environments. His research on robotic wheelchairs incorporating human velocity habituation models reflects his commitment to human-centered robotics. With multiple papers on biped locomotion accumulating hundreds of citations, Ishii has established himself as a key figure in bridging reinforcement learning theory with practical robotic applications.
Research Focus
Key Achievements
Top Papers
- 1On-line EM Algorithm for the Normalized Gaussian Network280 citations · 2000
- 2Reinforcement learning for a biped robot based on a CPG-actor-critic method118 citations · 2007
- 3Reinforcement learning for a CPG-driven biped robot66 citations · 2004
- 4Reinforcement Learning for Biped Locomotion35 citations · 2002
- 5
- 6
- 7Application of reinforcement learning to balancing of Acrobot19 citations · 2003
- 8
- 9
- 10Modeling of human velocity habituation for a robotic wheelchair9 citations · 2014