Speaker Localization and Detection in Videoconferencing Environments Using a Modified SRP-PHAT Algorithm

Abstract

[EN] The Steered Response Power - Phase Transform (SRP-PHAT) algorithm has been shown to be one of the most robust sound source localization approaches operating in noisy and reverberant environments. However, its practical implementation is usually based on a costly fine grid-search procedure, making the computational cost of the method a real issue. In this paper, we introduce an effective strategy which performs a full exploration of the sampled space rather than computing the SRP at discrete spatial positions, increasing its robustness and allowing for a coarser spatial grid that reduces the computational cost required in a practical implementation. The modified SRP-PHAT functional has been successfully implemented in a real time speaker localization system for multiparticipant videoconferencing environments. Moreover, a localization-based speech-non speech frame discriminator is presented.