基于SVM的多分类器融合算法在说话人识别中的应用  

SVM based multi-classifier applied in anti-noise speaker identification system

作　　者：王波[1] 徐毅琼[1] 李弼程[1] 

机构地区：[1]解放军信息工程大学信息工程学院,河南郑州450002

出　　处：《计算机工程与设计》2007年第12期2909-2910,2913,共3页Computer Engineering and Design

基　　金：解放军某部技术攀登基金项目(SP200503099)

摘　　要：多分类器协同合作克服了单个分类器识别效果不理想,适用范围较窄或对使用环境要求较高的不足,并为研制更高性能的分类器,提供了解决问题的另一个途径。提出了一种基于SVM的多分类器说话人识别系统,该系统的识别框架基于多分类器的协同工作。该多分类器系统采用了ANN(artificial neural networks)、GMM(gaussian mixed model)分类器和子带结构分类器,参数选取包括MFCC(mel frequency cepstrum coefficient)、LPCC(linear prediction cepstrum coefficient)。多分类器融合采用SVM方法。本系统在超短波信道(15 db)的实际应用中达到94%的识别率。Multi-classifier structure overcomes the shortcomings of single classifier, such as low recognition rate, narrow application field and critical demand of environment. A fusion method based on SVM is proposed. The involved classifiers include ANN, GMM, sub-band classifiers, etc. A practical speaker recognition system based on multi-classifier structure is presented. The input features of classifiers contain MFCC, LPCC. In practical application, the recognition rate of the system achieves 94 % in environment of super short wave （SNR 15 db）.

关 键 词：说话人识别 多分类器协同 子带结构分类器 分层识别 人工神经网络 混合高斯模型 支持向量机 

分 类 号：TP912[自动化与计算机技术]