摘要
针对现实中训练数据不足的特点,在说话人建模时采用高斯混合模型—通用背景模型(Gaussian MarkovModel-Uniform Background Model,GMM-UBM),主要从说话人识别模型的自适应方法和参数估计方法两个方面,研究如何提高说话人识别系统的识别率.在说话人识别模型自适应方面,改进传统的用最大后验概率MAP(Maximum A Posterior Probability)得到说话人模型的方法,将语音识别中的最大似然线性回归MLLR(MaximumLikelihood Linear Regression)和基于特征音(EigenVoice,EV)的自适应方法,应用到说话人识别模型自适应当中,并将其与MAP方法进行比较.
This thesis adopts GMM-UBM when model speaker recognition system considering of lacking data. In the aspect of adapting in speaker recognition system modeling and parameter estimating, attentions are put on researching in how to improve recognition rate. In the side of adapting in speaker recognition system modeling, we will ameliorate conventional MAP （Maximum A Posterior Probability） means to get speaker recognition model, apply MLLR （Maximum Likelihood Linear Regression） and EigenVoice adaptation ways which used in speech recognition into adapting in speaker recognition system modeling, and compare the results with MAP means.
出处
《计算机系统应用》
2013年第8期176-179,共4页
Computer Systems & Applications
