Humanoid binaural sound tracking using kalman filtering and HRTFs

Abstract

Audio and visual perceptions are two crucial elements in the design of mobile humanoids. In unknown and uncontrolled environments, humanoids are supposed to navigate safely and to explore their surroundings autonomously. While robotic visual perception, e.g. stereo vision, has significantly evolved during the last few years, robotic audio perception, especially binaural audition, is still in its early stages. However, non-binaural sound source localization methods based on multiple microphone arrays, like multiple signal classification and estimation of Time-Delay Of Arrivals (TDOAs) between all microphone pairs, have been very actively explored [1], [2]. © Springer-Verlag London Limited 2007.