¿El Caballo Viejo? Latin Genre Recognition with Deep Learning and Spectral Periodicity

Bob L. Sturm, Corey Mose Kereliuk, Jan Larsen

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

The “winning” system in the 2013 MIREX Latin Genre Classification Task was a deep neural network trained with simple features. An explanation for its winning performance has yet to be found. In previous work, we built similar systems using the BALLROOM music dataset, and found their performances to be greatly affected by slightly changing the tempo of the music of a test recording. In the MIREX task, however, systems are trained and tested using the Latin Music Dataset (LMD), which is 4.5 times larger than BALLROOM, and which does not seem to show as strong a relationship between tempo and label as BALLROOM. In this paper, we reproduce the “winning” deep learning system using LMD, and measure the effects of time dilation on its performance. We find that tempo changes of at most ±6 % greatly diminish and improve its performance. Interpreted with the low-level nature of the input features, this supports the conclusion that the system is exploiting some low-level absolute time characteristics to reproduce ground truth in LMD.
Original languageEnglish
Title of host publicationProceedings of the 5th International Conference Mathematics and Computation in Music, MCM 2015
EditorsTom Collins, David Meredith, Anja Volk
PublisherSpringer
Publication date2015
Pages335-346
ISBN (Print)978-3-319-20602-8
ISBN (Electronic)978-3-319-20603-5
DOIs
Publication statusPublished - 2015
Event5th International Conference on Mathematics and Computation in Music (MCM 2015) - London, United Kingdom
Duration: 22 Jun 201525 Jun 2015
Conference number: 5
http://mcm2015.qmul.ac.uk/

Conference

Conference5th International Conference on Mathematics and Computation in Music (MCM 2015)
Number5
CountryUnited Kingdom
CityLondon
Period22/06/201525/06/2015
Internet address
SeriesLecture Notes in Computer Science
Volume9110
ISSN0302-9743

Keywords

  • Machine music listening
  • Genre
  • Deep Learning
  • Evaluation

Cite this

Sturm, B. L., Kereliuk, C. M., & Larsen, J. (2015). ¿El Caballo Viejo? Latin Genre Recognition with Deep Learning and Spectral Periodicity. In T. Collins, D. Meredith, & A. Volk (Eds.), Proceedings of the 5th International Conference Mathematics and Computation in Music, MCM 2015 (pp. 335-346). Springer. Lecture Notes in Computer Science, Vol.. 9110 https://doi.org/10.1007/978-3-319-20603-5_34