Music identification

Abstract

Music identification eases information anxiety by bypassing text-based input or search efforts of the user to retrieve its contents. As current efforts towards Music Information Retrieval have been directed towards system-centric designs, this project hopes to shed light and explore its user-centric design potential. It explores the science and mathematics towards direct recognition to a song in the database, and tempo and frequency adjustments made to identify the resampled query to a song in the database. The MATLAB software is used to craft the algorithms, cascade the two modules, collate their results and analyse them, and present them in the GUI application. Recommendations have also been proposed to move this project forwards.