Research on Speech Separation Algorithm Based on Microphone Line Array
LI Yong
ZHANG Erhua
TANG Zhenmin
Abstract：The importance of speech separation as a front-end module of speech recognition is self-evident.The multi-chan-nel speech separation algorithm can use the characteristics of speech between different channels to adjust the beam direction to the target direction to achieve the effect of separation.In this paper,based on the fixed beamforming method,the null steering function is added,and the weighting coefficient is constructed in the frequency domain to further remove the interference signal,and a multi-channel speech separation algorithm suitable for the microphone line array is studied.Experiments show that this method can extract the target speech from the two-person mixed speech,which obviously increases the intelligibility of the speech.
Keywords：microphone line arrayspeech separationbeamformingnull steeringweighting coefficient
Publication Date：2025-07-20
Online Publishing Date：2025-09-25(First online date of this platform, not the publication date of the document)
Pages：4( 1808-1811 )