基于改进的高斯混合模型的说话人识别的研究-计算机应用技术专业论文.docxVIP

ABSTRACTThis ABSTRACT This paper investigates not only the text—independent speaker recognition system based on the Gaussian mixture models(GMM)but also the improved Gaussian mixture models(GMM)．The main work in the paper is listed as below：1)．We establish a speech database which contains 30 speakers．21．We finish the MFCC feature extraction，and discuss some problems in the extracting process．3)．We introduce t驻orthogonal Gaussian mixture models and its application in the speaker recognition．Standard GMM assumes diagonal covariance matr!ces，and needs a large number of mixture components to obtai