作为说话人识别特征参量的MFCC的提取过程

说话人识别是人的个体特征识别中的一个重要分支,在实际生活中已得到广泛应用。而人的听觉系统是一个比较理想的说话人识别系统,MFCC(Mel倒谱系数)模拟了人的听觉特性,是符合人听觉特性的语音特征参量,在实际应用中取得了较高的识别率。文中通过一个卷积同态系统简单介绍了语音信号的倒谱分析方法,并通过对Mel频率刻度得到符合人听觉特性的Mel频率等效滤波器组,最后介绍了MFCC求取的一般过程和算法。

Process of Extracting the MFCC in Speaker Recognition

Speaker recognition is a process of automatical recognizing who is speaking on the basis of individual information included in speech signal.As a popular speech feature parameter,MFCCs(Mel-Frequency Cepstral Coefficients) are based on the human ears'non-linear frequency characteristic and perform a high recognition rate in practical application.By introducing the cepstral analysis method of speech waves and Mel-frequency scales,this paper presents some general arithmetic and processing steps to compute Mel-frequency cepstral coefficients.