Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (5): 144-146.DOI: 10.3778/j.issn.1002-8331.2009.05.042

• 数据库、信号与信息处理 • Previous Articles     Next Articles

Speech signal sparse decomposition based on matching pursuit algorithm

JING Ai-wen1,LIU Yun2,MA Yi-li3   

  1. 1.College of Science,Air Force Engineering University,Xi’an 710051,China
    2.the Missile Institute,Air Force Engineering University,Xi’an 710051,China
    3.Huanghe Science and Technology University,Zhengzhou 450006,China
  • Received:2008-01-09 Revised:2008-05-08 Online:2009-02-11 Published:2009-02-11
  • Contact: JING Ai-wen

基于MP算法的语音信号稀疏分解

井爱雯1,刘 云2,马轶丽3   

  1. 1.空军工程大学 理学院,西安 710051
    2.空军工程大学 导弹学院,西安 710051
    3.黄河科技学院,郑州 450006
  • 通讯作者: 井爱雯

Abstract: Speech signal sparse decomposition is a new method for decomposing audio data into a compact approximate representation.Speech signal sparse decomposition can be used in compression,denoising and recognition of speech signal.This paper researches speech signal sparse decomposition with matching pursuit algorithm.At last,the simulation experiments results show that speech signal sparse decomposition based on matching pursuit algorithm has better reconstructive precision and higher sparsity.

Key words: sparse decomposition, over-complete dictionary, matching pursuit algorithm

摘要: 语音信号稀疏分解是一种新的语音信号分解方法,可以将语音信号分解为很简洁的近似表达形式。在语音信号稀疏分解的基础上,可应用于语音处理的多个方面,如语音压缩、语音去噪和语音识别等。研究利用Matching Pursuit(MP)算法实现语音信号的稀疏分解,实验结果表明基于MP算法的语音信号稀疏分解具有较好的重建精度和较高的稀疏度。

关键词: 稀疏分解, 过完备库, 匹配追踪算法