Katsushi Miura

Papers

Total Citations

H-Index

About

Katsushi Miura is a pioneering researcher in developmental robotics, whose work explores how robots can acquire human-like vocalization and communication skills through social interaction. His primary research areas include vowel acquisition, imitation learning, and the correspondence problem—the challenge of mapping actions between differently structured bodies (robots vs. humans). Miura’s most influential contribution is the concept of “unconscious anchoring in maternal imitation,” introduced in his 2007 paper (69 citations), which demonstrates how a robot can learn vowel categories by leveraging a caregiver’s natural tendency to imitate, even without explicit segmentation of speech. He further advanced this line of inquiry with the “auto-mirroring bias” (AMB) model (2012, 23 citations), showing that infants—and by extension, robots—can acquire caregiver vowel categories even when imitation is infrequent, by finding correspondences through self-generated vocalizations. His 2008 work (6 citations) addresses the realistic scenario where caregivers do not always imitate, proposing clearer articulation strategies for vowel mapping. Miura’s research bridges cognitive science and robotics, offering elegant computational models that explain how social feedback drives early language development, with profound implications for building more natural human-robot interaction systems.