Performance Evaluation and Comparative Analysis for Widely Used Sound Source Localization Algorithms

Sound source localization (SSL) is technique that is used to process where the source and determine its direction using several SSL algorithms. Having an accurate, fast and computational efficient SSL algorithm, is extremely useful especially when designing hearing aid devices, that can guide the deaf person to the sound source direction, which is very important especially when encountering dangerous sound sources such as car horn. In this paper, a performance comparison of several SSL algorithms are conducted, bearing in mind the algorithms' accuracy, execution time and suitability for hearing aid devices. The conducted tests show that both the The MUltiple Signal Classification (MUSIC) and the steered response power (SRP) algorithms performed the nearly best in terms of accuracy especially under low signal-to-noise (SNR) conditions, the capability of identifying multiple audio sources, and the low processing delay, which make them best candidates to be deployed on hearing aid devices when compared with other SSL algorithms.