计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (36): 40-42.

• 学术探讨 • 上一篇    下一篇

基于多频带谱减法的语音增强算法的研究

程 正1,赵鹤鸣2   

  1. 苏州大学 电子信息学院,江苏 苏州 215000
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-12-21 发布日期:2007-12-21
  • 通讯作者: 程 正

Speech enhancement based on spectral subtraction of multi-band scale

CHENG Zheng1,ZHAO He-ming2   

  1. College of Electronic,Suzhou University,Suzhou,Jiangsu 215000,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-12-21 Published:2007-12-21
  • Contact: CHENG Zheng

摘要: 提出了一种改进的基于多频带的语音增强算法。详细地分析了按照不同的线性比例、按照非线性的Bark频率和按照共振峰的一般分布规律三种方法划分频带后,算法性能的改善情况。采用板仓-斋藤距离、波形图、语谱图对算法进行客观测试,仿真实验表明:经过增强后的语音,得到了改善,在主观听觉上也提高了人耳的舒适度;特别是在非平稳环境下,比传统的谱减法有较大的改进。

关键词: 语音增强, 谱减法, 多频带, 板仓距离

Abstract: We propose an improved spectral subtraction method for reducing acoustic noise added to speech in noise environments like white noise and factory noise.We analyze the experimental result using three different methods to divide the bands,such as linear bands,bark bands,and formant orderliness bands proposed in the paper.Simulation shows a better quality in terms of Itakura-Saito distance,speech wave picture,spectral picture and perceptual evaluation of quality of the enhancement speech.

Key words: speech enhancement, spectral subtraction, multi-band, Itakura-Saito distance