Overview on 3D sound research at VASL, NoViC, KAIST

Abstract

We overview the research related with 3D sound technology done by Virtual Audio Systems Laboratory at KAIST. Modeling of Head-Related Transfer Function (HRTF), which is the core technology for synthesizing 3D sound, based on ellipsoidal head model or Principal Components Analysis (PCA) is dealt with. HRTF customization method based on subjective tuning of a few parameters is also proposed. Application of 3D sound technology to robots is also deal with in brief. Robot auditory system including speech and speaker recognition techniques and robot artificial ear is implemented to an actual robot platform, and the performance is verified from an experiment in a household environment.