Papers
7
Total Citations
33
H-Index
3
About
Ryoichi Nakajo is a pioneering researcher in neurorobotics and developmental robotics, whose work bridges the gap between cognitive development and autonomous robotic learning. His primary research areas include sensorimotor learning, imitative learning, bimanual manipulation, and predictive learning with uncertainty estimation. Nakajo’s most impactful contribution is his model for viewpoint representation acquisition, enabling robots to learn imitation from their own sensory-motor experiences without requiring manual coordinate transformations—a foundational step for robots to recognize and replicate actions from diverse perspectives (9 citations). He further advanced this line of inquiry with a sequence-to-sequence deep learning framework for viewpoint transformation and action mapping (3 citations). In his highest-cited work (10 citations), Nakajo introduced a recurrent neural network with multiple attractors that allows robots to learn compound tasks by chaining shorter sensorimotor units, addressing the challenge of task complexity. His more recent work (2025) tackles coordinated bimanual manipulation using Transformers and action chunking, while his neurorobotics experiments on predictive learning with uncertainty estimation model how infants develop cognitive abilities through caregiver interactions. Nakajo’s research uniquely integrates cognitive science principles with cutting-edge machine learning, offering a principled path toward more flexible, human-like robots.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7