基于EMD法的语音信号特征提取方法研究

Speech Signal Feature Extraction Method Research Based on EMD

为了提高语音信号的识别率,提出了一种基于经验模态分解(EMD)法的语音信号特征参数提取方法。该方法先对语音信号进行EMD分解,获得其内模函数;再进行FFT和DCT变换,得到特征分量,以此构成语音信号新特征参数。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, a speech signal feature parameter extraction method based on EMD method is proposed. First the speech signal is decomposited by EMD in the method, the intrinsic mode function is obtained, then FFT and DCT transform is conducted, and get the feature component , thereby forming a new speech signal feature parameters. Finally, the Gauss mixed model （ GMM ） is used for speaker speech recognition, and experiment shows that the new characteristic parameters obtaines better recognition rate.