计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (26): 74-75.

• 学术探讨 • 上一篇    下一篇

利用FFT实现语音信号稀疏分解的改进算法

刘 强,尹忠科,王建英   

  1. 西南交通大学 信息科学与技术工程学院,成都 610031
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-09-11 发布日期:2007-09-11
  • 通讯作者: 刘 强

Improvement of MP based voice signal sparse decomposition by using FFT

LIU Qiang,YIN Zhong-ke,WANG Jian-ying   

  1. School of Science & Technology of Information,Southwest Jiaotong University,Chengdu 610031,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-09-11 Published:2007-09-11
  • Contact: LIU Qiang

摘要: 研究基于Matching Pursuit(MP)方法实现的语音信号稀疏分解问题,通过对语音信号稀疏分解中使用的过完备原子库结构特性的分析,提出了一种改进的信号稀疏分解算法。该算法针对语音信号的特点,以FFT算法实现的稀疏分解为基础缩小了原子的搜索范围,从而不仅进一步提高分解速度,还能以更稀疏的形式表示语音信号。算法的有效性为实验结果所证实。

关键词: 信号处理, 稀疏表示, 稀疏分解, Matching Pursuit(MP), FFT

Abstract: After studying of Matching Pursuit(MP) based voice signal sparse decomposition,a new improvement of sparse decomposition algorithm is presented based on analysis of structure property of the over-complete atom dictionary used in voice signal sparse decomposition.According to the character of the voice signal,the new algorithm based on the FFT algorithm that reduces the range of the searching atoms,can not only heighten the speed of the decompositions,but also reconstruct the speech signal with less atoms.Finally the experimental results show that the performance of the proposed algorithm is good.

Key words: signal processing, sparse representation, sparse decomposition, Matching Pursuit(MP), FFT