Max Robotics — Global Service Robot Directory + US Certification

About

Daiki Kimura is a researcher whose work sits at the intersection of robotics, machine learning, and human-computer interaction, with particular expertise in deep reinforcement learning, multimodal sensing, and cognitive health assessment. His contributions span over a decade, beginning with foundational work in ultra-fast online transfer learning for humanoid robots (2013), where he developed the STAR-SOINN framework to enable robots to incrementally learn from their environments in real time. He has since advanced robotic perception through multimodal object recognition for unknown objects and human motion prediction using Long Short-Term Memory networks, enabling more naturalistic human-robot interaction. A notable thread in Kimura's research is his effort to make deep learning more practical for real-world robotic deployment. His DAQN framework (2018) addresses the data-hungry nature of deep reinforcement learning by incorporating auto-encoding techniques to reduce training requirements. More recently, Kimura has pioneered the use of humanoid robots as screening tools for mild cognitive impairment, leveraging prosodic and acoustic conversational features — a timely contribution given the global rise in dementia cases. This work, among his most cited with 13 citations, reflects his commitment to applying intelligent robotics to pressing societal challenges, making his research both technically rigorous and meaningfully impactful.