A new microphone array speech enhancement method based on AR model

  • Authors:

  • Liyan Zhang;Fuliang Yin;Lijun Zhang

  • Affiliations:

  • School of Electronic Engineering, Dalian JiaoTong University, Dalian, China;School of Electronic and Infomation Engineering, Dalian University of Technology, Dalian, China;Grace International Group, Division R&D, Montreal, Canada

  • Venue:
  • LSMS/ICSEE'10 Proceedings of the 2010 international conference on Life system modeling and simulation and intelligent computing, and 2010 international conference on Intelligent computing for sustainable energy and environment: Part III
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper applies the single microphone speech enhancement method to the microphone array speech enhancement method, and proposes a new speech enhancement method based on autoregressive model (AR). First, for the input matrix, the method adopts the generalized cross correlation method based on onset signals to estimate the time delay. Then according to the time delay information, the method calculates the linear prediction coefficient of signals received by the microphone array by means of Levinson-Durbin algorithm, and then carries out AR model speech enhancement. Finally, the method combines the enhanced speech signals to one channel output signals. The simulation results show that the proposed method can eliminate the plus noise and improve the speech quality effectively.