摘要
文章对水语的七个单元音进行声学特性研究。首先利用线性预测编码LPC模型提取七个单元音的共振峰频率值,然后绘制声学元音图并进行分析,最后利用隐马尔可夫模型HMM对水语的七个单元进行识别,识别率为83.75%。实验结果表明:水语七个单元音的共振峰频率分布之间存在相关性,验证了水语七个单元音共振峰频率分布的正确性,为水语研究提供了有力依据。
In this paper,the acoustic characteristics of the seven purevowels in Sui language are studied.First,the linear predictive coding(LPC)model was used to extract the formant frequency values of the seven purevowels,and then the acoustic vowelschart was drawn and analyzed.Finally,the seven purevowelsof Sui were identified using the hidden markov model(HMM).The recognition rate was 83.75%.The experimental results show that correlation between the formant frequency distributions of the seven purevowels in Sui language.The correctness of the formant frequency distribution of the seven purevowels in Sui language is verified,which provides a strong basis for the study of Sui language.
作者
赵冬梅
王林
李学林
刘秋会
ZHAO Dongmei;WANG Lin;LI Xuelin;LIU Qiuhui(Guizhou Minzu University,Guiyang 550025,China)
出处
《现代信息科技》
2020年第4期17-19,22,共4页
Modern Information Technology
基金
贵州民族大学校级课题([2018]5773-QN02)。
