期刊文献+

基于MVQM的说话人识别的研究

Research on Speaker Recognition Based on MVQM
下载PDF
导出
摘要 提出了一种新的说话人识别方法。该方法综合了VQ和GMM的优点,通过用VQ误差尺度取代传统GMM的输出概率函数,减少了建模时对训练数据量的要求,提高了识别速度。实验结果证明了该方法的有效性。 A new approach of speaker recognition which combined the advantages of VQ and GMM is presented. By adopting VQ error scale instead of probability output of tradition GMM,data capacity is reduced during modeling and high recognition rate is gained. Experiment result show the proposed approach is effective on speaker recognition.
出处 《电声技术》 2006年第2期41-43,共3页 Audio Engineering
基金 教育部科学技术重点项目(03082) 国家自然基金(60472058)
关键词 说话人识别 矢量量化 混合高斯模型 混合矢量量化模型 speaker recognition VQ Gaussian Mixture Model (GMM) Mixture Vector Quantization Model(MVQM)
  • 相关文献

参考文献5

  • 1MARKOV K, NAKAGAWA S. Text-independent speaker recognition system using frame level likeli-hood processing[R]. Technical Report of IEICE,SP96-17, 1996 : 37-44.
  • 2TSENG B, SOONG F, ROSENBERG A. Continuous probabilistic acoustic Map for speaker recognition[J]. Proceedings of ICASSP'92, 1992,2:161-164.
  • 3REYNOLDS D A, ROSE R C. Robust text-independent speaker identification using gaussian mixture speaker models[J]. IEEE Trans. On Speech and Audio Processing, 1995,3 ( 1 ) :72-83.
  • 4BIMBOT F, MAGRIN-Chagnolleau I, MATHAN L. Second-order statistical measures for text -independent speaker identification[J]. Speech Communication, 1995,17(2) : 177-192.
  • 5MATSUI T, FURUI S. Likelihood normalization for speaker verification using a phoneme-and speaker-independent model[J]. Speech Communication,1995,17 (2)97-116.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部