摘要
语音端点检测是语音处理分析过程中的一个重要组成部分,针对方差法在低信噪比下对语音信号端点检测准确率低的问题,本文提出一种子带频带方差法和谱减法相结合的检测算法,算法中采用多窗谱估计改进谱减法对河南方言语音进行降噪处理,并将得到的信号的子带频带方差进行端点检测.用Matlab进行仿真验证了此方法在端点检测中的准确性,该方法降低了河南方言语音处理的时间,为进一步提取河南方言语音识别特征参数提供了条件.
Speech endpoint detection is important part of speech processing. In order to improve correctness of endpoint detection method based on spectrum variance in the case of low signal to noise ratio （SNR）, this paper proposes a new speech endpoint detection algorithm based on combination subband variance with spectral subtraction, it reduces Henan dialect speech signal noise by using modified spectral subtraction for multi window spectral estimation and calculates its subband variance in endpoint detection. Matlab software is used to verify the accuracy of this method in the endpoint detection, the simulation results show that this algorithm can reduce the time of Henan dialect speech processing, and provide certain conditions for further extraction of Henan dialect speech recognition feature parameters.
出处
《西北师范大学学报（自然科学版）》
CAS
北大核心
2016年第5期55-59,共5页
Journal of Northwest Normal University(Natural Science)
基金
国家自然科学基金资助项目(61101232)
郑州市科技局科技发展计划项目(20140663)
郑州市嵌入式系统应用技术重点实验室建设项目(121PYFZX177)
郑州市教学质量工程资助项目(22LG201608)
关键词
语音端点检测
多窗谱
频带方差
谱减法
speech endpoint detection
multitaper method
band variance
spectral subtraction
