The impulse sound source tracking using Kalman filter and the cross-correlation

Abstract

In case of robot interacting with human, natural movement and action of the robot give the feeling of closeness to people. One of the basis technique of human-robot interaction is sound source tracking. In this paper, we propose a weighted Kalman filter using cross-correlation value based on TDOA method for sound source tracking. In the experiment, we use the footstep sound as an impulse sound. The experiment shows that the proposed method gives more accurate and natural azimuth results