Masakazu Inoue
Papers
2
Total Citations
25
H-Index
2
About
Masakazu Inoue is a leading researcher in cognitive robotics and human-robot interaction, with a focus on enabling robots to understand and navigate human environments through hierarchical spatial concept formation. His major contributions lie in developing Bayesian generative models that integrate multimodal information—such as vision, position, and language—to allow robots to autonomously learn and abstract spatial concepts at varying levels of detail. In his most cited work (2018, 21 citations), Inoue introduced a method for hierarchical spatial concept formation that mirrors human cognitive flexibility, enabling robots to select appropriate abstractions based on context. Earlier, he pioneered the use of hierarchical Multimodal Latent Dirichlet Allocation (hMLDA) for place concept learning (2016, 4 citations), combining position and vision data to move beyond traditional SLAM approaches. His research has been partially supported by CREST, JST, highlighting its national significance. Inoue’s work is foundational for human support robots, bridging the gap between raw sensor data and human-like spatial reasoning, and continues to influence autonomous navigation and human-robot collaboration.
Research Focus
Key Achievements
Top Papers
- 1
- 2