基于特征均值距离的短语音段说话人聚类算法.pdfVIP

第 34 卷第 6 期 电 子 与 信 息 学 报 Vol.34No.6 2012 年 6 月 Journal of Electronics Information Technology Jun. 2012 基于特征均值距离的短语音段说话人聚类算法 李艳雄* 吴 永 贺前华 (华南理工大学电子与信息学院 广州 510640) 摘 要：该文提出一种基于特征均值距离的短语音段说话人聚类算法。首先，定义特征均值距离用来在特征层而不 是模型层刻画两个类之间的相似度；然后，迭代合并特征均值距离最小的两个类，直到任意两类之间的特征均值距 离的最小值大于一个自适应门限为止。采用取自两个语音数据库的短于 3 s 的语音段进行实验测试，结果表明：与 基于 AHC+BIC 的算法相比，F 度量值平均提高了 5%，运算速度约为以前算法的4.68 倍。 关键词：语音信号处理；说话人聚类；特征均值距离；短语音段 中图分类号：TN912.3 文献标识码： A 文章编号：1009-5896(2012)06-1404-04 DOI: 10.3724/SP.J.1146.2011.01139 Feature Mean Distance Based Speaker Clustering for Short Speech Segments Li Yan-xiong Wu Yong He Qian-hua (School of Electronic and Information Engineering, South China University of Technology, Guangzhou 510640, China) Abstract: An algorithm of speaker clustering is proposed based on Feature Mean Distance (FMD) for short speech segments. First, a distance measure, i.e. FMD, is introduced to represent the similarities between two clusters on the level of feature instead of the level of model. Then, two clusters with the minimum of FMDs are iteratively merged until the minimum of FMDs is larger than an adaptive threshold. Experimental results show average 5% improvements in F measure are obtained in comparison with the AHC+BIC based algorithm. In addition, the proposed algorithm is 4.68 times faster than the AHC+BIC based algorithm. Key words: Speech signal processing; Speaker clustering; Feature Mean Distance (FMD); Short speech segments 1 引言 [7] 只在大数据量(即较长的语音段)的情况下才有效 。 随着多媒体技术的发展，语音文档记录