摘要
对语音识别进行了探讨,提出一种通过性别识别对连续隐马尔可夫模型(CHMM)分类的方法,在此基础上进行语音识别。首先,通过计算性别判定语音信号的Mel频率倒谱系数(MFCC)使用CHMM对说话人性别进行识别,然后再根据不同性别使用分类CHMM进行语音识别。最后通过实验验证了方法的有效性。
A method of speech recognition based on Continuous Hidden Markov Model(CHMM) that is classified by gender identification is introduced.First,gender is identified using CHMM through calculating the Mel Frequency Cepstral Coefficient (MFCC),and then,speech recognition is realized by CHMM classified based on gender.In the end,the affectivity of the method is tested by an experiment.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第21期187-189,共3页
Computer Engineering and Applications
基金
国家自然科学基金(the National Natural Science Foundation of China under Grant No.50407017)
安徽省教育厅自然科学基金项目(No.2004KJ058
No.2006KJ019A)。