摘要
为了提高语音信号的识别率,提出了一种改进的LPCC参数提取方法。该方法先对语音信号进行预加重、分帧加窗处理,然后进行小波分解,在此基础上提取LPCC参数,从而构成新向量作为每帧信号的特征参数。最后采用高斯混合模型(GMM)进行说话人语音识别,实验表明新特征参数取得了较好的识别率。
In order to improve the speech recognition rate, an improved LPCC parameter extraction method is proposed.First the pre-emphasis, frames and windows processing is conducted to speech signal in the method,then wavelet decomposition is used, the LPCC parameter is extracted on the basis,thus a new vector is formed as each frame signal characteristic parameter. Finally, the Gauss mixed model （GMM） is used for speaker speech recognition, and experiment shows that the new characteristic parameters obtaines better recognition rate.
出处
《电子设计工程》
2012年第6期29-30,33,共3页
Electronic Design Engineering
基金
宝鸡文理学院院级重点项目(ZK11127)
关键词
特征提取
小波变换
分解
LPCC参数
语音信号
feature extraction
wavelet transform
decomposition
LPCC parameter
speech signals
