基于GMM-SVM的说话人识别系统研究  

Research of speaker recognition system based on GMM-SVM

作　　者：赵立辉[1] 毛竹[2] 霍春宝[1] 杨红喆[3] 

机构地区：[1]辽宁工业大学电气工程学院,辽宁锦州121001 [2]中国电子科技集团光电研究所,辽宁锦州121001 [3]北京理工大学计算机学院,北京100083

出　　处：《工矿自动化》2014年第5期49-53,共5页Journal Of Mine Automation

基　　金：辽宁省科学技术基金项目(2007217003);辽宁省教育厅项目(L2012230)

摘　　要：针对同类语音数据的相似性和不同类数据具有不同几何距离的特点,提出了一种基于GMMSVM的说话人识别系统。该系统结合了GMM和SVM的优点,解决了GMM在语音数据较小时不能区分数据间的差异性及SVM在处理大量数据时识别率下降的问题;采用改进的K-Means算法实现模型参数初始化,提高了参数精度。试验结果表明,基于GMM-SVM的说话人识别系统较单独采用GMM或SVM的系统具有更好的识别率和鲁棒性。According to the similarity of speech data of the same type and characteristic that data of different types has different geometrical distance, the paper proposed a speaker recognition system based on SVM-GMM. The system combines advantages of GMM and SVM, solves problems that GMM cannot distinguish differences between the voice data while the data is small, and recognition rate of SVM drops while handling large amounts of data. The improved K-Means algorithm is used for initialization of model parameters to improve accuracy. The experiment results show that speaker recognition system based on SVM-GMM has better recognition rate and robustness than the system using GMM or SVM alone.

关 键 词：说话人识别 GMM SVM 识别率 

分 类 号：TD679[矿业工程—矿山机电]