Robot audition is a research field to realize listening capability using a robot’s own ears, which was proposed first in the world by Professor Okuno (Kyoto University, Kyoto, Japan) and Dr. Nakadai (Principal Researcher, HRI-JP). Aiming at the further evolution of robot audition technology, open source robot audition software calledHARK (HRI-JP Audition for Robots with Kyoto University) was released in 2008.
Automatic speech recognition with robot-embedded microphones is basically problematic due to environmental noise, and reverberation which are caused according to the distance between a microphone and a speaker. The all-new ASIMO uses more sophisticated robot audition technology than the original technology based on HARK so that now it can work properly beyond in-vitro environments. This technology was jointly improved by HRI-JP and Honda R&D.