摘要
为提升身份验证系统的安全性,研究声纹识别技术在身份验证系统中的应用与优化。首先,提出一种基于声纹识别的身份验证系统框架。其次,针对声纹特征提取方法,重点研究基于梅尔频率倒谱系数(Mel Frequency Ceptral Coefficient,MFCC)的特征提取方法。再次,探讨基于贝叶斯估计的高斯混合模型(Gaussian Mixture Model,GMM)优化方法。最后,进行实验分析,评估识别率、准确率、召回率等性能指标,并与传统GMM方法进行比较。
To improve the security of identity verification systems,research the application and optimization of voiceprint recognition technology in identity verification systems.Firstly,a framework for identity verification system based on voiceprint recognition is proposed.Secondly,focusing on the feature extraction method of voiceprint,a feature extraction method based on Mel Frequency Cepstral Coefficients(MFCC)was studied.Once again,explore the optimization method of Gaussian Mixture Model(GMM)based on Bayesian estimation.Finally,conduct experimental analysis to evaluate performance indicators such as recognition rate,accuracy,and recall,and compare them with traditional GMM methods.
作者
陈丹丹
邓惠俊
CHEN Dandan;DENG Huijun(Hefei University of Economics,Hefei 230012,China)
出处
《电声技术》
2024年第3期18-20,共3页
Audio Engineering