About

Kazunori Komatani is a pioneering researcher at the intersection of robotics, machine learning, and human-robot interaction, with particular expertise in robot audition, neural network-based sensorimotor learning, and multimodal communication. His foundational contributions to robot audition — including the development of the HARK open-source software platform — have significantly advanced robots' ability to localize, separate, and recognize simultaneous speech in real-world acoustic environments, addressing one of the most challenging problems in embodied AI. Komatani's work on recurrent neural networks with parametric bias (RNNPB) has opened compelling pathways for robots to learn inter-modal mappings between language, motion, and visual perception through experience, enabling capabilities such as imitation learning and object dynamics prediction. His innovative research on musical robots — including systems that synchronize with live human musicians in real time — demonstrates a rare creative breadth that bridges technical rigor with human-centered applications. With papers accumulating citations across robotics, speech processing, and cognitive systems communities, Komatani's body of work, totaling over 387 citations across his top publications, reflects sustained and wide-reaching influence in building robots that perceive, communicate, and learn naturally within human environments.

Research Focus

Key Achievements

17
H-Index
74
Papers
964
Total Citations
13
Avg Citations/Paper
🏆 Most Cited Paper
Two-way translation of compound sentences and arm motions by recurrent neural networks
64 citations · 2007
📈 Most Prolific Year: 2007 (12 Papers)
🤝 Key Collaborators: 60
🏛 Institutions: Kyoto University, RIKEN Center for Brain Science, Kyoto College of Graduate Studies for Informatics, Kyoto University of Education, The University of Osaka, Nagoya University

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 34 days ago