Shinzo Kitamura
Papers
14
Total Citations
144
H-Index
8
About
Shinzo Kitamura is a pioneering figure in robotics, whose work has fundamentally shaped the fields of bipedal locomotion, machine learning, and emergent computation. His most influential contributions center on the autonomous motion generation and control of legged robots. In his landmark 1990 paper, he introduced a hierarchical framework that models a biped robot's center of gravity as an inverted pendulum, using Hopfield-type neural networks to solve inverse kinematics—a concept that has garnered 23 citations and remains foundational. Kitamura further advanced the field by developing Q-Learning with Adaptive State Segmentation (QLASS) in 2002, a reinforcement learning algorithm that enables robots to adaptively segment their state space without prior knowledge, earning 22 citations. His innovative use of neuro-oscillators for trajectory generation in 2002 demonstrated stable periodic locomotion patterns, while his work on emergent design theory (1999) and collision avoidance using potential functions (1987) showcased his versatility. With over 130 total citations across his top papers, Kitamura’s research has provided essential tools for autonomous robot control, blending neural networks, reinforcement learning, and bio-inspired oscillators to solve complex real-world locomotion and planning challenges.
Research Focus
Key Achievements
Top Papers
- 1
- 2Autonomous trajectory generation of a biped locomotive robot22 citations · 1991
- 3Q-Learning with adaptive state segmentation (QLASS)22 citations · 2002
- 4
- 5A Design Method as Inverse Problems and Application of Emergent Computations11 citations · 2000
- 6
- 7
- 8An approach to the emergent design theory and applications8 citations · 1999
- 9
- 10