Development of Sound Source Localization System using Explicit Adaptive Time Delay Estimation

Abstract

The problem of sound source localization is to determine the position of sound sources using the measurement of the acoustic signals received by microphones. To develop a good sound source localization system which is applicable to a mobile platform such as robots, a time delay estimator with low computational complexity and robustness to background noise or reverberations is necessary. In this paper, an explicit adaptive time delay estimation method for a sound source localization system is proposed. Proposed explicit adaptive time estimation algorithm employs direct adaptation of the delay parameter using a transform-based optimization technique, rather than a gradient-search-based optimization. This new adaptation method shows less computational burden, fast tracking performance for non-stationary signals, more robustness to reverberant environment. Performance of the proposed method is shown through computer simulations.