Katsushi Ikeuchi
Carnegie Mellon University, The University of Tokyo, Tokyo University of Information Sciences, Microsoft (United States), Tokyo University of Science, Digital Wave (United States), Induk University, Microsoft Research (United Kingdom), Advanced Institute of Industrial Technology, Microsoft Research Asia (China)
Papers
125
Total Citations
3,412
H-Index
31
About
Katsushi Ikeuchi is a pioneering robotics and computer vision researcher whose work has profoundly shaped how robots learn from human observation and demonstration. His research centers on two interconnected themes: robot programming by demonstration and humanoid robot motion imitation, with sustained contributions spanning over two decades. Ikeuchi is perhaps best known for developing the Assembly Plan from Observation (APO) paradigm, a landmark framework enabling robots to watch humans perform tasks, interpret their actions, and autonomously generate executable programs — a vision realized across a series of highly cited studies from the early 1990s accumulating hundreds of citations. His work on grasp recognition and temporal task segmentation laid critical groundwork for intelligent robotic manipulation. Equally influential is his research on humanoid imitation learning. Through innovative motion-capture analysis and task-model frameworks, Ikeuchi's team enabled biped humanoid robots to faithfully replicate complex human dance movements, addressing formidable challenges in whole-body motion generation and lower-limb dynamics — work that collectively attracted over 500 citations. His "Learning from Observation" paradigm, extended through hidden Markov models for continuous gesture recognition and multi-demonstration interaction extraction, represents a cohesive intellectual legacy bridging perception, cognition, and physical robot execution. Ikeuchi's research remains essential reading for anyone pursuing intuitive, human-centered robot programming.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5Imitating human dance motions through motion structure analysis128 citations · 2003
- 6
- 7
- 8
- 9
- 10