基于SVM的汉语语音情感识别研究

Emotion recognition of Chinese speech based on SVM

随着信息技术的发展,对人机交互能力的要求不断提高,情感信息处理已成为提高人机交互能力的一个重要课题。本文提出了一种汉语语音情感分类方法,主要研究了4种基本的人类情感:高兴、愤怒、恐惧、悲伤。从汉语语音信号中提取了能量、基频、语速等特征,利用支持向量机方法识别,取得了43.7%的平均识别率。

Human-machine interface grows its importance in accordance with the development of information communication services which indicates that the role of emotion in human recognition is essential. In this paper, a chinese speech based on emotion classification method was presented. Four basic human emotions including happiness, anger,fear and sadness were investigated. The energy, pitch and speech rate were extracted from speech signal. The SVM was used to recognize the four basic human emotions, and we have achieved an average recognition rate of approximate 43.7 %.