计算机应用 ›› 2011, Vol. 31 ›› Issue (05): 1447-1449.DOI: 10.3724/SP.J.1087.2011.01447

• 典型应用 • 上一篇    下一篇

基于帧间相关性的语音活动检测方法

李宇1,2,郭雷勇1,2,谭洪舟2   

  1. 1.广东药学院 医药信息工程学院,广州 510006
    2.中山大学 信息科学与技术学院,广州 510275
  • 收稿日期:2010-10-22 修回日期:2010-12-20 发布日期:2011-05-01 出版日期:2011-05-01
  • 通讯作者: 李宇
  • 作者简介:李宇(1977-),男,广东梅州人,讲师,博士,主要研究方向:语音活动检测、医学信号处理;郭雷勇(1973-),男,湖南郴州人,讲师,博士,主要研究方向:语音活动检测、RFID技术;谭洪舟(1965-),男,重庆人,教授,博士生导师,主要研究方向:盲信号处理理论、音视频信号建模。
  • 基金资助:

    国家自然科学基金资助项目(60874060)。

Voice activity detection method based on inter-frame correlation

LI Yu1,2, GUO Lei-yong1,2, TAN Hong-zhou2   

  1. 1.College of Medical Information Engineering, Guangdong Pharmaceutical University, Guangzhou Guangdong 510006, China
    2.School of Information Science and Technology, Sun Yat-sen University, Guangzhou Guangdong 510275, China
  • Received:2010-10-22 Revised:2010-12-20 Online:2011-05-01 Published:2011-05-01
  • Contact: Li Yu

摘要: 为了提高统计模型似然比测试的语音活动检测(VAD)的检测性能,利用前后语音帧间存在的统计相关特性,提出一种改进VAD算法。通过前帧语音频谱分量对先验信噪比进行递归估计,然后利用前一帧的语音检测状态来设计判决阈值,建立了双阈值隐马尔可夫模型语音活动判决规则。实验表明,此帧间相关性VAD算法的检测指标值优于Sohn算法。

关键词: 语音活动检测, 统计模型, 相关性, 似然比测试, 先验信噪比, 阈值

Abstract: To enhance the detection performance of statistical model-based Voice Activity Detection (VAD) using likelihood ratio test, an improved VAD was proposed by utilizing the correlation between tandem speech frames. First a priori Signal-to-Noise Ratio (SNR) was estimated using recursive estimation method based on the result of the previous speech frame instead of the traditional decision-directed method. Secondly double thresholds were designed by depending on the previous frame's detention result. Finally a detection rule was presented based on two-state Hidden Markov Model (HMM) coupled with double thresholds. The experimental results show that the inter-frame correlation based VAD scheme gets better performance than the Sohn's VAD.

Key words: voice activity detection, statistical model, correlation, likelihood ratio test, a priori SNR, threshold