AUTOMATIC CLASSIFICATION OF MIDI TRACKS

Alexandre Bernardo, Thibault Langlois

2008

Abstract

This paper presents a system for classifying MIDI tracks according to six predefined classes: Solo, Melody, Melody+Solo, Drums, Bass and Harmony. No metadata present in the MIDI file is used. The MIDI data (pitch of notes, onset time and note durations) are preprocessed in order to extract a set of features. These data sets are then used with several classifiers (Neural Networks, k-NN).

Paper Citation

in Harvard Style

Bernardo A. and Langlois T. (2008). AUTOMATIC CLASSIFICATION OF MIDI TRACKS . In Proceedings of the Tenth International Conference on Enterprise Information Systems - Volume 2: ICEIS, ISBN 978-989-8111-37-1, pages 539-543. DOI: 10.5220/0001708705390543

in Bibtex Style

@conference{iceis08,
author={Alexandre Bernardo and Thibault Langlois},
title={AUTOMATIC CLASSIFICATION OF MIDI TRACKS},
booktitle={Proceedings of the Tenth International Conference on Enterprise Information Systems - Volume 2: ICEIS,},
year={2008},
pages={539-543},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001708705390543},
isbn={978-989-8111-37-1},
}

in EndNote Style

TY - CONF
JO - Proceedings of the Tenth International Conference on Enterprise Information Systems - Volume 2: ICEIS,
TI - AUTOMATIC CLASSIFICATION OF MIDI TRACKS
SN - 978-989-8111-37-1
AU - Bernardo A.
AU - Langlois T.
PY - 2008
SP - 539
EP - 543
DO - 10.5220/0001708705390543