92 / 2021-03-30 10:55:52

A Modified SSA Functional for Real-Time Sound Source Localization

Microphone array,Sound source localization,Computational cost,Robustness

Special Sessions > Applications of machine learning in vibration and noise problems

A Modified SSA Functional for Real-Time Sound Source Localization

Linke zhang¹, Chang Liu ¹, Xiaohui Song ¹, Li Xia ¹ and Yongsheng Yu<sup>2,*

 1</sup> School of Energy and Power Engineering, Wuhan University of Technology, Wuhan, Hubei, China

 ² State Key Laboratory of Silicate Materials for Architectures, Wuhan University of Technology, Wuhan, Hubei, China

^*Corresponding author: yongshengyu@whut.edu.cn

Abstract: Accuracy and efficiency of sound source localization has always been an important issue in the field of acoustics. The Steered Response Power – Phase Transform（SRP-PHAT）algorithm has become the mainstream algorithm in the industry because of its high spatial resolution and robustness. The controlled sampling algorithm (SSA) based on the reciprocity of wave propagation has higher spatial resolution than SRP-PHAT algorithm, and its positioning accuracy has significant advantages under the disadvantages of small number of array elements and low SNR. However, its practical implementation is usually based on a costly fine grid-search procedure, making the computational cost of the algorithm a serious issue. This paper introduces an improved algorithm based on SSA algorithm, which can improve the efficiency of the algorithm by shrinking the scanning area. Since the SSA algorithm has only a few local extrema, the improved algorithm has a good convergence speed. Experimental results show that compared with the SSA algorithm, the proposed algorithm has almost the same positioning performance and robustness, as well as lower computational cost.



Keywords: Microphone array; Sound source localization; Computational cost; Robustness