On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations

Bob L. Sturm, Pardis Noorzad

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

3080 Downloads (Pure)

Abstract

A recent system combining sparse representation classification (SRC)
and a perceptually-based acoustic feature (ATM)
\cite{Panagakis2009,Panagakis2009b,Panagakis2010c},
outperforms by a significant margin the state of the art in music genre recognition, e.g., \cite{Bergstra2006}.
With genre so difficult to define,
and seemingly based on factors more broad than acoustics,
this remarkable result motivates investigation into, among other things,
why it works and what it means for how humans organize music.
In this paper, we review the application of SRC and ATM to recognizing genre,
and attempt to reproduce the results of \cite{Panagakis2009}.
First, we find that classification results
are consistent for features extracted from different analyses.
Second, we find that SRC accuracy improves
when we pose the sparse representation problem
with inequality constraints.
Finally, we find that only when we reduce the number of classes by half
do we see the high accuracies reported in \cite{Panagakis2009}.
Original languageEnglish
Title of host publicationProceedings of the 9th International Symposium on Computer Music Modeling and Retrieval
Place of PublicationLondon
Publication date2012
Pages379-394
Publication statusPublished - 2012
EventComputer music modeling and retrieval - London, United Kingdom
Duration: 19 Jun 201222 Jun 2012

Conference

ConferenceComputer music modeling and retrieval
CountryUnited Kingdom
CityLondon
Period19/06/201222/06/2012

Fingerprint

Modulation
Automatic teller machines
Acoustics

Cite this

Sturm, B. L., & Noorzad, P. (2012). On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations. In Proceedings of the 9th International Symposium on Computer Music Modeling and Retrieval (pp. 379-394). London.
Sturm, Bob L. ; Noorzad, Pardis. / On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations. Proceedings of the 9th International Symposium on Computer Music Modeling and Retrieval. London, 2012. pp. 379-394
@inproceedings{a7c2e0cc5c51463fb71f26501f6f16aa,
title = "On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations",
abstract = "A recent system combining sparse representation classification (SRC) and a perceptually-based acoustic feature (ATM) \cite{Panagakis2009,Panagakis2009b,Panagakis2010c},outperforms by a significant margin the state of the art in music genre recognition, e.g., \cite{Bergstra2006}.With genre so difficult to define, and seemingly based on factors more broad than acoustics,this remarkable result motivates investigation into, among other things, why it works and what it means for how humans organize music.In this paper, we review the application of SRC and ATM to recognizing genre, and attempt to reproduce the results of \cite{Panagakis2009}.First, we find that classification results are consistent for features extracted from different analyses.Second, we find that SRC accuracy improveswhen we pose the sparse representation problem with inequality constraints.Finally, we find that only when we reduce the number of classes by halfdo we see the high accuracies reported in \cite{Panagakis2009}.",
author = "Sturm, {Bob L.} and Pardis Noorzad",
year = "2012",
language = "English",
pages = "379--394",
booktitle = "Proceedings of the 9th International Symposium on Computer Music Modeling and Retrieval",

}

Sturm, BL & Noorzad, P 2012, On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations. in Proceedings of the 9th International Symposium on Computer Music Modeling and Retrieval. London, pp. 379-394, Computer music modeling and retrieval, London, United Kingdom, 19/06/2012.

On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations. / Sturm, Bob L.; Noorzad, Pardis.

Proceedings of the 9th International Symposium on Computer Music Modeling and Retrieval. London, 2012. p. 379-394.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations

AU - Sturm, Bob L.

AU - Noorzad, Pardis

PY - 2012

Y1 - 2012

N2 - A recent system combining sparse representation classification (SRC) and a perceptually-based acoustic feature (ATM) \cite{Panagakis2009,Panagakis2009b,Panagakis2010c},outperforms by a significant margin the state of the art in music genre recognition, e.g., \cite{Bergstra2006}.With genre so difficult to define, and seemingly based on factors more broad than acoustics,this remarkable result motivates investigation into, among other things, why it works and what it means for how humans organize music.In this paper, we review the application of SRC and ATM to recognizing genre, and attempt to reproduce the results of \cite{Panagakis2009}.First, we find that classification results are consistent for features extracted from different analyses.Second, we find that SRC accuracy improveswhen we pose the sparse representation problem with inequality constraints.Finally, we find that only when we reduce the number of classes by halfdo we see the high accuracies reported in \cite{Panagakis2009}.

AB - A recent system combining sparse representation classification (SRC) and a perceptually-based acoustic feature (ATM) \cite{Panagakis2009,Panagakis2009b,Panagakis2010c},outperforms by a significant margin the state of the art in music genre recognition, e.g., \cite{Bergstra2006}.With genre so difficult to define, and seemingly based on factors more broad than acoustics,this remarkable result motivates investigation into, among other things, why it works and what it means for how humans organize music.In this paper, we review the application of SRC and ATM to recognizing genre, and attempt to reproduce the results of \cite{Panagakis2009}.First, we find that classification results are consistent for features extracted from different analyses.Second, we find that SRC accuracy improveswhen we pose the sparse representation problem with inequality constraints.Finally, we find that only when we reduce the number of classes by halfdo we see the high accuracies reported in \cite{Panagakis2009}.

M3 - Article in proceeding

SP - 379

EP - 394

BT - Proceedings of the 9th International Symposium on Computer Music Modeling and Retrieval

CY - London

ER -

Sturm BL, Noorzad P. On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations. In Proceedings of the 9th International Symposium on Computer Music Modeling and Retrieval. London. 2012. p. 379-394