首页 | 本学科首页   官方微博 | 高级检索  
     

基于改进MFCC语音特征参数的语音质量评估的研究
引用本文:陈明义,孙冬梅,何孝月.基于改进MFCC语音特征参数的语音质量评估的研究[J].电路与系统学报,2009,14(3).
作者姓名:陈明义  孙冬梅  何孝月
作者单位:中南大学信息科学与工程学院电子科学与技术系,湖南,长沙,410083
摘    要:.本文讨论了基于MFCC(Mel-Frequency Cepstral Coefficients)听觉感知模型的语音质量客观评估方法(EPESQ:Efficient Psychoacoustics Evaluation of Speech Quality)在网络语音质量客观评估中的应用,深入分析了EPESQ算法的基本原理,并对算法的MFCC语音特征参数进行了改进和性能分析.实验证明该算法在确保评测准确性的前提下,大大降低了算法的复杂度,具有较强的适应性,可靠性,实用性,完全可以应用于通信网络语音质量评估.

关 键 词:感知模型  语音质量  客观评估

Study on speech quality evaluation based on improved MFCC
CHEN Ming-yi,SUN Dong-mei,HE Xiao-yue.Study on speech quality evaluation based on improved MFCC[J].Journal of Circuits and Systems,2009,14(3).
Authors:CHEN Ming-yi  SUN Dong-mei  HE Xiao-yue
Affiliation:CHEN Ming-yi,SUN Dong-mei,HE Xiao-yue(College of Information Science& Engineering,Central South University,Changsha 410083,China)
Abstract:The paper discusses EPESQ algorithm based on MFCC which is used in quality evaluation of communication networks.The basic principle is studied extensively,MFCC is improved,and the performance of the algorithm is analyzed.Experimental results show that the new algorithm retains the comparable performance and reduces the complexity.It has a strong adaptability,reliability,practicality and can be applied in quality evaluation of communication networks.
Keywords:MFCC  EPESQ
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号