A Music Information Retrieval Algorithm using MP and SPM

Abstract

In the IT age where private music collections contain thousands of songs and commercial music catalogs consist of millions of songs, music recommender systems become more and more important especially for searching and browsing music catalogs. In this paper, retrieving the required information from acoustic music signal in an efficient way is considered. The approach is to segment those audio signals and determined similarities among songs, particularly, a piece of input music signal compared with storage music song’s signal into the database and then to retrieve the similar song. Representing the audio signal having sparse nature is accomplished by Matching Pursuit. In order to matching a candidate segment with the query segment, the audio signal similarity measure is performed by Spatial Pyramid Matching.