文章基本信息

标题：Speech Emotion Recognition Using Combined Features of HMM & SVM Algorithm
本地全文：下载
作者：Aastha Joshi
期刊名称：International Journal of Advanced Research In Computer Science and Software Engineering
印刷版ISSN：2277-6451
电子版ISSN：2277-128X
出版年度：2013
卷号：3
期号：8
出版社：S.S. Mishra
摘要：Speech is an interactive interface medium as it is possible to express emotions and attitude through speech. In this paper, a hybrid of Hidden Markov Models (HMMs) and Support Vector Machines (SVM) has been proposed to classify four emotions viz. happy, angry, sad and aggressive. Combining advantage on capability to dynamic time warping of HMM and pattern recognition of SVM. HMMs, which export likelihood probabilities and optimal state sequences, have been used to model speech feature sequences i.e. our proposed system is trained using HMM algorithm for emotions considered, while SVM has been employed to make a decision i.e. for classification. The recognition result of the hybrid classification has been compared with the isolated SVM and the maximum recognition rates have reached 98.1% and 94.2% respectively
关键词：SER System; feature extraction; HMM algorithm; SVM classifier; Performance Parameters