基于BP神经网络的语音情感识别研究  

Speech Emotion Recognition Based on BP Neural Network

作　　者：徐照松[1] 元建[1] 

机构地区：[1]广西师范学院计算机与信息工程学院,广西南宁530023

出　　处：《软件导刊》2014年第4期11-13,共3页Software Guide

摘　　要：随着科技的迅速发展,人机交互越来越受到人们的重视,语音情感识别更是学术界研究的热点。将BP神经网络算法用于语音情感识别研究,并在汉语情感数据集上进行了相关实验,识别的准确率达到了91.5%,相较于SVM算法分类精度提高了5%。With the rapid development of technology ,human-computer interaction more and more suffer people’s attention . Research on speech emotion recognition is the focus of academic .In this article ,we use the BP neural network algorithm to research on speech emotion recognition and conducted experiments on chinese sentiment data sets ,recognition accuracy rate reached 91 .5 percent ,compared to the SVM accuracy is improved by 5% .

关 键 词：语音情感识别 BP神经网络 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]