Index of /~echew/papers/MIR2003

 Name                    Last modified      Size  Description
 Parent Directory                             -   
 MIR03_unal.pdf          13-Feb-2004 09:26   80K  
 MIR03journal_unal.pdf   29-Nov-2004 14:41  116K  
 reference_MIR03.txt     20-Feb-2004 22:55  483   
 reference_MMSJ05.txt    29-Nov-2004 17:11  494   

In this directory is the PDF file for a paper titled

"Creating Data Resources for Designing User-centric Front-ends for Query by Humming Systems"
by E. Unal, S.S. Narayanan, H.-H. Shih, E. Chew and C.-C. Jay Kuo

The paper was be presented at the 
5th ACM SIGMM International Workshop on Multimedia Information Retrieval (MIR 2003)
11th Annual ACM International Conference on Multimedia
Berkeley, California.  November 7, 2003.

The extended paper will appear in 
the ACM Multimedia Systems Journal in Spring 2005 in
a special issue of MIR'03 Workshop on 
Systems and Architectures for Multimedia Information Retrieval
Guest Editors: Chabane Djeraba, Nicu Sebe, and Michael S. Lew
Click on reference_MMSJ05.txt for the BibTeX reference.

The workshop paper was published in the
Multimedia Information Retrieval Workshop Proceedings
Click on reference_MIR03.txt for the BibTeX reference.

The project website is located at
http://sail.usc.edu/music

The workshop website is at 
http://carol.wins.uva.nl/~nicu/MIR

The conference website is at
http://turing.acm.org/sigs/sigmm/mm2003

The journal website is at
http://cairo.cs.uiuc.edu/mmsj.html

THE COMPLETE PAPERS, text with figures, can be viewed as a PDF document.
Click on MIR03journal_unal.pdf if you wish to view the journal paper in PDF format.
Click on MIR03_unal.pdf if you wish to view the workshop paper in PDF format.

--'--,--'--,--'--,--'--,--'--,--'--,--'--,--'--,--'--,--'--,--'--,--'--

"Creating Data Resources for Designing User-centric Front-ends for Query by Humming Systems"
by E. Unal, S.S. Narayanan, H.-H. Shih, E. Chew and C.-C. Jay Kuo

JOURNAL ABSTRACT: Advances in music retrieval research greatly depend
on appropriate database resources and their meaningful organization.
In this paper we describe data collection efforts related to the
design of query by humming (QBH) systems. We also provide a
statistical analysis for categorizing the collected data, especially
focusing on inter-subject variability issues. In total, 100 people
participated in our experiment resulting in around 2000 humming
samples drawn from a predefined melody list consisting of 22 different
well known music pieces, and over 500 samples of melodies that were
chosen spontaneously by our subjects. These data are being made
available for the research community. The data from each subject were
compared to the expected melody features, and an objective measure was
derived to quantify the statistical deviation from the baseline. The
results showed that the uncertainty in human humming varies depending
on the musical structure of the melodies and the musical background of
the subjects. Such details are important for designing robust QBH
systems.

WORKSHOP ABSTRACT: Advances in music retrieval research greatly depend
on appropriate database resources and their meaningful organization.
In this paper we describe the data collection efforts related to the
design of query by humming (QBH) systems.  We also provide a
statistical analysis for categorizing the collected data, especially
focussing on inter-subject variability issues.  In total, 100 people
participated in our experiment resulting in around 2000 humming
samples drawn from a predefined melody list consisting of 22 different
well known music pieces, and over 500 samples of melodies that were
chosen spontaneously by our subjects.  These data will be made
available for the research community.  The data from each subject were
sompared to the expected melody features, and an objective measure was
derived to quantify the statistical deviation from the baseline.  The
results showed that the uncertainty in the humming varies with respect
to the melodies' musical structure and subject's musical background.
Such details are important for designing robust QBH systems.