A Music Retrieval Research and Implementation of New Methods

LI Peng^1,2, XIA Xiao-liang^2,3, ZHOU Ming-quan^2,3, LI Nan-shan^2,3

(1. School of Computer Science, Shanxi Normal University, Xi'an 710062; China; 
2. Virtual Reality Applications Research Center of Ministry of Education, Beijing Normal University, Beijing 100875, China;
3. College of Information Science and Technology, Beijing Normal University, Beijing 100875, China)

Received:2009-07-19 Revised:2010-01-04 Online:2010-06-28 Published:2010-05-14

Abstract:

A serial of improvements to implement music information retrieval is proposed, it extracts

the pitch with a improved YIN algorithm. A new “cabinet” post-treatment is put forward to

produce regular pitch sequence; a melody contour marker method and fuzzy matching

method is designed; a similarity measure system to complete the retrieval is defined as

well. Experiments show that the new process is an improvement of the traditional content-

based music retrieval.

Key words: YIN algorithm, "cabinet", post-treatment, melody contour, similarity measure

CLC Number:

LI Peng, XIA Xiao-liang, ZHOU Ming-quan, LI Nan-shan.

A Music Retrieval Research and Implementation of New Methods [J]. JOURNAL OF BEIJING UNIVERSITY OF POSTS AND TELECOM, 2010, 33(3): 62-66.

PDF