Yoshiki Ando

Papers

Total Citations

H-Index

About

Yoshiki Ando is a pioneering researcher in developmental robotics and machine learning, specializing in how robots can autonomously form concepts and understand language through multimodal sensory experience. His work bridges artificial intelligence, cognitive science, and robotics, focusing on enabling machines to learn hierarchical and flexible representations of objects and actions. Ando’s major contributions include the development of novel probabilistic models—such as hierarchical latent Dirichlet allocation and infinite mixture models—that allow robots to integrate visual, auditory, haptic, and linguistic data from their environment. For instance, his 2013 paper on “Formation of hierarchical object concept” (27 citations) demonstrated how robots can build structured knowledge from raw sensorimotor interactions, moving beyond flat categorization. His 2015 work on “Concept formation by robots using an infinite mixture of models” (19 citations) further advanced this by enabling robots to discover and adapt concepts without predefined categories. These studies have been influential in the field of robot concept learning, providing a foundation for more sophisticated, human-like machine understanding. Ando’s research is notable for its interdisciplinary approach, combining Bayesian statistics with embodied cognition, and it continues to inspire new directions in autonomous learning systems.