节点文献
基于神经网络的说话人识别算法的研究与实验
Research and experiment of speaker identification algorithms based on artificial neural network
【摘要】 人工神经网络通过学习可以实现对输入向量的分类,也就是说,对于经过训练的神经网络,每输入一个矢量,人工神经网络输出一个该矢量所属类别的标号,神经网络的这种分类作用可以运用到说话人识别中。本文在介绍人工神经网络实现对输入向量分类原理的基础上,通过MATLAB实现了基于神经网络学习向量量化方法(LVQ)的说话人识别实验,取得了较为满意的结果。
【Abstract】 Artificial neural network(ANN)can be used as classifier through training.To make it specifically,the trained ANN can output the class label for an input vector,so the ANN can be applied into classifier system for speaker identification.In this paper,the principle of ANN classifier is introduced firstly,and then learning vector quantization(LVQ)of ANN is used in the speaker identification experiment with the assistance of Matlab,which satisfying results are obtained eventually.
【关键词】 说话人识别;
神经网络;
学习向量量化;
【Key words】 speaker identification; neural network; learning vector quantization;
【Key words】 speaker identification; neural network; learning vector quantization;
- 【文献出处】 电子测量技术 ,Electronic Measurement Technology , 编辑部邮箱 ,2007年11期
- 【分类号】TP183;TN912.34
- 【被引频次】10
- 【下载频次】165