Experiments with the Shazam music identification algorithm

The motivation of this study is to identify music without the original recording. The existing solutions tackle variations in some properties such as background sound and white noise, but the identification of samples containing large variations in key, tempo, ornamentation, and harmonization remains largely unsolved. This study takes an existing algorithm and uses an existing data set to explore the parameters required for successful identification, as well as variations in key. The findings show a simple way to identify and normalize the key of a sample. Future work will tackle tempo and ornamentation challenges.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Computer Science

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2019

Author dc:creator

• Xiao, Fangjian Flora

Contributors dc:contributor

• Fleck, Margaret M.

• music
• recognition

Statement dc:rights

• Copyright 2019 Fangjian Xiao

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/104731

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/104731