Masters Theses
Abstract
"The volume of multimedia data generated, processed, and consumed has been increasing with each passing day. In order to use multimedia data effectively, a system providing content-based queries and retrievals is essential. Music data has been one of the major components of multimedia data. Traditionally, searching for music files has been done through text input. This thesis proposes a system that is capable of retrieving matches from database of music files based on content-based queries.
A content-based query input is provided either as a music file or as a tune hummed into a microphone. The motivation for this research is to implement a system that would be capable of searching and retrieving matches based on melody comparison. The proposed system works for audio data in the MIDI format consisting of instrumental music. A set of MATLAB programs have been developed to achieve this. A MIDI file is given as input, from which the melody component is extracted. The extracted melody is represented in the form of character and numeric strings. The melody string for a given query is matched against other melody strings that have been extracted from songs in a database. A list of files containing melody similar to the query is presented based on the results of the string-matching program"--Abstract, page iii.
Advisor(s)
Subramanya, S. R.
Committee Member(s)
Sabharwal, Chaman
Kluczny, Raymond Michael
Department(s)
Computer Science
Degree Name
M.S. in Computer Science
Publisher
University of Missouri--Rolla
Publication Date
Summer 2003
Pagination
viii, 35 pages
Note about bibliography
Includes bibliographical references (pages 33-34).
Rights
© 2003 Sanjay Shankar, All rights reserved.
Document Type
Thesis - Restricted Access
File Type
text
Language
English
Subject Headings
MIDI (Standard)Computer sound processingInformation storage and retrieval systems
Thesis Number
T 8336
Print OCLC #
54887327
Recommended Citation
Shankar, Sanjay, "A content-based audio retrieval system using MIDI" (2003). Masters Theses. 4420.
https://scholarsmine.mst.edu/masters_theses/4420
Share My Thesis If you are the author of this work and would like to grant permission to make it openly accessible to all, please click the button above.
