About

Kazuhiro Nakadai is a pioneering researcher in the field of robot audition — the science of enabling robots to perceive, localize, and interpret sounds in real-world environments. His work sits at the rich intersection of signal processing, machine learning, and human-robot interaction, with a sustained focus on making robots capable listeners in complex, noisy settings. Nakadai's most significant contribution is the development of HARK (Honda Research Institute Japan Audition for Robots with Kyoto University), an open-source robot audition software system capable of simultaneously processing speech from multiple speakers. First presented in 2008 and expanded in 2010, HARK has become a foundational platform in the field, accumulating over 270 citations across related publications. His early work on active audition for humanoid robots (2000, 212 citations) established core principles for integrating sound source localization and speech recognition into embodied systems. Beyond stationary robots, Nakadai extended auditory scene analysis to dynamic and outdoor environments, including aerial drones equipped with microphone arrays. His more recent investigations into deep learning-based sound localization (2017, 129 citations) reflect his adaptability to emerging methodologies. With over 1,200 citations across his most influential works, Nakadai's research has fundamentally shaped how robots hear and understand the world around them.

Research Focus

Key Achievements

33
H-Index
199
Papers
4,103
Total Citations
21
Avg Citations/Paper
🏆 Most Cited Paper
Active Audition for Humanoid
212 citations · 2000
📈 Most Prolific Year: 2010 (23 Papers)
🤝 Key Collaborators: 197
🏛 Institutions: Honda (Japan), Tokyo Institute of Technology, Japan Science and Technology Agency, Waseda University, Keio University, Kyoto College of Graduate Studies for Informatics

Top Papers

  1. 1
    Active Audition for Humanoid
    212 citations · 2000
  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 · 1 days ago