作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2012, Vol. 38 ›› Issue (2): 184-185. doi: 10.3969/j.issn.1000-3428.2012.02.060

• 人工智能及识别技术 • 上一篇    下一篇

一种三层判决的说话人索引算法

陈雪芳 1,杨继臣 2   

  1. (1. 东莞理工学院计算机学院,广东 东莞 523808;2. 仲恺农业工程学院计算机科学与工程学院,广州 510225)
  • 收稿日期:2011-04-12 出版日期:2012-01-20 发布日期:2012-01-20
  • 作者简介:陈雪芳(1978-),女,讲师、硕士,主研方向:计算机视觉,多媒体信息处理,嵌入式系统设计;杨继臣,讲师、博士
  • 基金资助:
    东莞市2010年高等院校科研机构科技计划基金资助项目(201010814014)

Speaker Index Algorithm of Three-layer Criterion

CHEN Xue-fang 1, YANG Ji-chen 2   

  1. (1. School of Computer, Dongguan University of Technology, Dongguan 523808, China; 2. College of Computer Science and Engineering, Zhongkai University of Agriculture and Engineering, Guangzhou 510225, China)
  • Received:2011-04-12 Online:2012-01-20 Published:2012-01-20

摘要: 为提高说话人索引准确率,提出一种三层判决的说话人索引算法。第1层使用惩罚距离公式对说话人改变进行检测,第2层采用说话人模型自举法进行初次说话人辨认,第3层采用GMM说话人超级矢量进行判决,解决说话人模型自举法中产生的数据不匹配问题。实验结果表明,采用惩罚距离公式,与贝叶斯信息判决方法相比不需调整参数,与DISTBIC方法相比F1值提高2%,使用GMM说话人超级矢量,在说话人索引准确率和数量准确率方面分别提高8.95%、18.25%。

关键词: 三层判决, 说话人索引, 惩罚距离, 模型自举法, GMM说话人超级矢量

Abstract: To improve the precision of speaker index, a speaker indexing algorithm of three-layer criterion is proposed. In the first layer, penalty distance is proposed to judge whether speaker changes. In the second layer, speaker model bootstrapping is used to identify speaker first time. In the third layer, GMM Speaker Supervector(GMMSS) is used to identify speaker further in order to settle the problem of data mismatch in speaker model bootstrapping. Experimental results show that, it is no need to tune penalty factor compared to BIC and F1 can improve 2% compared to DISTBIC; speaker indexing accuracy can improve 8.95% and the accuracy on the number of speaker can improve 18.25% by using GMMSS in speaker identification.

Key words: three-layer criterion, speaker index, penalty distance, model bootstrapping method, GMM Speaker Supervector(GMMSS)

中图分类号: