Analysis of Fast Localization Algorithms for Acoustical Environments

Abstract

Of the acoustical source localization algorithms that have been proposedin the literature, mostcan be dividedinto two separate groups. While some algorithmsare robust, they are computationally complex. Conversely, the fast algorithms cannot localize robustly in real world scenarios. Recently several new algorithms have been proposed that are robust and fairly fast, including Hierarchical Localization and Hybrid Localization. A method for optimal placement of candidate locations will be presented, which reduces computational complexity of Hierarchical Localization. The performance of Hierarchical Localization and Hybrid Localization will be compared in both simulated and real scenarios. It will be shown that Hybrid Localization is more robust than Hierarchical Localization and also computationally faster.