Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (21): 145-147.DOI: 10.3778/j.issn.1002-8331.2010.21.041

• 数据库、信号与信息处理 • Previous Articles     Next Articles

Improved algorithm of data preprocessing in audio fingerprint search

GUAN Yao-hua,SHEN Ling,WU Yun,ZHAO Yong   

  1. Key Lab of Integrated Microsystems,Shenzhen Graduate School,Peking University,Shenzhen,Guangdong 518055,China
  • Received:2009-01-08 Revised:2009-04-22 Online:2010-07-21 Published:2010-07-21
  • Contact: GUAN Yao-hua

音频指纹搜索中数据预处理的改进算法

关耀铧,申 凌,吴 云,赵 勇   

  1. 北京大学 深圳研究生院 集成微系统重点实验室,广东 深圳 518055
  • 通讯作者: 关耀铧

Abstract: Different Audio Fingerprint(AF) feature extraction algorithms require different AF searching algorithms.For some specific AF feature extraction algorithms(e.g.foosic algorithm),in data processing,two new AF searching algorithms are proposed and implemented:Principle Component Analysis(PCA) algorithm and sum of the corresponding data between different frames algorithm.The experimental results show that applying PCA algorithm,the accuracy is 94.98% while the search time is as low as 8.42%;applying sum algorithm,the accuracy is 95.92% while the search time is as low as 3.72%.

Key words: audio fingerprint search, Principle Component Analysis(PCA) algorithm, sum algorithm, data preprocessing

摘要: 不同的音频指纹提取算法需要不同的音频指纹搜索。针对某些特定(例如:foosic算法)的音频指纹提取算法,在数据预处理方面提出并实现了两种新的音频指纹搜索算法:PCA(Principle Component Analysis)主成分分析算法、不同帧之间相应的数据求和算法。实验结果表明:应用PCA算法,在搜索正确率为94.98%的情况下,搜索时间缩短为8.42%;应用求和算法,在搜索正确率为95.92%的情况下,搜索时间缩短为3.72%。

关键词: 音频指纹搜索, 主成分分析算法, 求和算法, 数据预处理

CLC Number: