The Tzanetakis music genre dataset: Its faults and the challenges they provide
Publikation: Forskning - peer review › Konferenceartikel i proceeding
Standard
The Tzanetakis music genre dataset: Its faults and the challenges they provide. / Sturm, Bob L..
I: Proc. Int. Society for Music Information Retrieval. 2012.Publikation: Forskning - peer review › Konferenceartikel i proceeding
Harvard
APA
CBE
MLA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - The Tzanetakis music genre dataset: Its faults and the challenges they provide
A1 - Sturm,Bob L.
AU - Sturm,Bob L.
PY - 2012/10/1/
Y1 - 2012/10/1/
N2 - Most research in automatic music genre recognition<br/>has used the dataset assembled by Tzanetakis et al. \cite{Tzanetakis2002,Tzanetakis2002b}.<br/>The integrity of this dataset, however, has never been analyzed.<br/>We catalog numerous serious problems in this dataset,<br/>including replications, mislabelings, versions, and data corruption.<br/>These problems affect the validity of all results derived from it;<br/>but they also present new challenges, <br/>especially now that researchers are using datasets so large<br/>that manual validation of their integrity is impossible.
AB - Most research in automatic music genre recognition<br/>has used the dataset assembled by Tzanetakis et al. \cite{Tzanetakis2002,Tzanetakis2002b}.<br/>The integrity of this dataset, however, has never been analyzed.<br/>We catalog numerous serious problems in this dataset,<br/>including replications, mislabelings, versions, and data corruption.<br/>These problems affect the validity of all results derived from it;<br/>but they also present new challenges, <br/>especially now that researchers are using datasets so large<br/>that manual validation of their integrity is impossible.
UR - http://ismir2012.ismir.net/
BT - Proc. Int. Society for Music Information Retrieval
T2 - Proc. Int. Society for Music Information Retrieval
ER -