【作者】 王青； 谢波； 陈根才；

【Author】 WangQing,Xie Bo,CHEN Gen-cai (College of Computer Science,ZheJiang University,HangZhou ZheJiang 310027,China)

【机构】 浙江大学计算机科学与技术学院；

【摘要】 本文讨论了基于神经网络的汉语语音情感识别。首先录制了400句汉语语音作为情感语音数据库,从语音信号中提取了能量、基频、语速等特征。然后利用神经网络识别汉语语音中的四类基本情感(愤怒、高兴、悲伤和害怕),取得了52.6%的平均识别率。更多还原

【Abstract】 In this paper, we discuss the emotion recognition of speech in mandarin using neural network. We have recorded a database containing emotional speech with over a 400 utterances. The energy, pitch and speech rate are extracted from speech signal. Neural network are used to recognize the four basic human emotions including anger, happiness, sadness and fear in speech, and we achieve an average recognition rate of approximately 52.6%.更多还原