摘要
语音端点检测是语音信号处理过程中的一个重要步骤,其准确性对语音信号处理有直接影响。传统的双门限语音端点检测技术,在纯净语音或高信噪比的情况下,语音端点判断准确,但低信噪比的情况下,端点识别率很低,出错率较高。为了提高低信噪比条件下语音端点检测的识别率,在传统双门限语音端点检测的基础上融合了语音增强,通过Matlab仿真实验,取得了较高的语音端点检测准确率。
Speech endpoint detection is an important step in speech signal processing,whose accuracy has a direct impact on the speech signal processing.The traditional double threshold voice detection technology can determine voice endpoints more accurately,in the case of pure voice or a high signal to noise ratio（SNR）.However,in the case of low SNR,endpoint recognition rate is very low,and error rate is higher.To increase the recognition rate under the low SNR,speech enhancement is integrated based on double threshold voice endpoint detection.Through Matlab simulation experiments,a higher voice endpoint detection accuracy is achieved.
出处
《测控技术》
CSCD
2016年第11期33-35,共3页
Measurement & Control Technology
关键词
语音增强
双门限法
先验信噪比
端点检测
语音信号处理
speech enhancement
double threshold method
priori SNR（signal to noise ratio）
endpoint detection
speech signal processing
