期刊文献+

哼唱搜索中一种基于DTW的旋律相似度算法

Melody Similarity Algorithm Based on DTW Applied in Query by Humming System
下载PDF
导出
摘要 哼唱搜索由于其用户要求低,应用方便,易于普及的特点,是当前音乐搜索应用中的主流方式,而旋律的相似度计算是哼唱搜索中直接影响搜索排名结果的关键技术。动态时间规整作为度量时间序列相似性的经典方法,是旋律相似度计算的主要方法之一。根据哼唱搜索的特点,提出了一种基于DTW的旋律相似度算法,实验结果表明,该算法进一步提高了哼唱搜索结果的精准度。 Query by humming is the most popular method used in music research applications for its features of user-friendly and easy to popularize. Melody similarity calculation is the key technology in query by humming which would decide the result of retrieval directly. Dynamic time warping , as a classic method used for time based on sequence similarity measuring, is also one of the most important methods applied in melody similarity calculation. A melody similarity algorithm based on DTW is pro- vided. The experimental results show that great improvements have been achieved by this algorithm in accuracy of the result of query by humming.
出处 《电声技术》 2013年第4期36-38,43,共4页 Audio Engineering
关键词 哼唱搜索 旋律相似度计算 动态时间规整 query by humming melody similarity calculate Dynamic Time Warping (DTW)
  • 相关文献

参考文献7

  • 1UGMBBC. Android 4.1 Jelly Bean 增加哼歌搜索widget[ EB/OL]. [ 2012 - 07 - 01 ]. http://www, enbeta, eom/ articles/194894, htm.
  • 2XIAO Qingmei, TSUGE S, KITA K. Music retrieval method based on filter - bank feature and earth mover' s distance [ C ]//Seventh International Conference on Natural Compu- tation. [S. 1. ] :IEEE Press, 2011:1845 - 1849.
  • 3COHEN S,GUIBASM L. The earth mover's distance under transformation sets [ C ]//Proceedings of the Seventh IEEE International Conference on Computer Vision. [ S. 1. ]: IEEE Press, 1999 : 1076 - 1083.
  • 4苏昊,王民,李宝.一种改进的DTW语音识别系统[J].中国西部科技,2011,10(1):38-39. 被引量:6
  • 5RUBNER Y, TOMASI C, GUIBAS L J. A metric for distri- butions with applications to image databases[ C ]//Proceed-ings of the 1998 IEEE International Conference on Comput- er Vision . [ S. 1. ] : IEEE Press, 1998:59 - 66.
  • 6翁颖钧,朱仲英.基于动态时间弯曲的时序数据聚类算法的研究[J].计算机仿真,2004,21(3):37-40. 被引量:31
  • 7蒋嵘,李德毅.基于形态表示的时间序列相似性搜索[J].计算机研究与发展,2000,37(5):601-608. 被引量:34

二级参考文献15

  • 1息晓静,林坤辉,周昌乐,蔡骏.语音识别关键技术研究[J].计算机工程与应用,2006,42(11):66-69. 被引量:15
  • 2[1]G Das,K Lin,H Mannila,G Renganathan & P Smyth.Rule discovery form time series[C].Proceedings of the 4rd International Conference of Knowledge Discovery and Data Mining,AAAI Press:16-22.
  • 3[2]E Keogh & M Pazzani.An enhanced representation of time series which allows fast and accurate classification,clustering and relevance feedback[C].Proceedings of the 4rd International Conference of Knowledge Discovery and Data Mining,AAAI Press,1998:239-241.
  • 4[3]D Berndt & J Clifford.Using dynamic time warping to find patterns in time series[C].AAAI-94 Workshop on Knowledge Discovery in Databases(KDD-94),Seattle,Washington,1994.
  • 5[4]D T Pham and A B Chan.Control Chart Pattern Recognition using a New Type of Self Organizing Neural Network[C].Proc.Instn,Mech,Engrs.1998,212(1):115-127.
  • 6韩纪庆,张磊,郑铁然.语音信号处理[M].北京:清华大学出版社,2003.
  • 7HUANG Xuedong, A Acero, H W Hon. Spoken- Language - Processing[M].Prentice Hall, 2001.
  • 8Li D,Knowledge Based Syst,1998年,10期,431页
  • 9Li D,Proc Second Pacific-Asia Conf Knowledge Discovery & Data Mining.Melbourne,1998年,392页
  • 10Xia B B,Master dissertation,1997年

共引文献66

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