Computation saving in a SRP-PHAT sound source locator variant

The well-known Steered Response Power-Phase Transform (SRP-PHAT) speech source locator enjoys robustness and renders accurate results in high reverberation and medium noise situations. However, it suffers from high computation burden, which debars it from many applications. Various arrangements have been developed during the recent years to lessen the calculations and facilitate its implementation. In this paper, a new hybrid scheme is suggested and compared its performance and costs with several well-known contestant algorithms. The simulation results confirm that the algorithm performance is highly promising and noticeable.