Publication result detail

Musical Form Retrieval

FEJFAR, J.; WEINLICHOVÁ, J.; ŠŤASTNÝ, J.

Original Title

Musical Form Retrieval

English Title

Musical Form Retrieval

Type

Paper in proceedings (conference paper)

Original Abstract

This paper describes the way how the Musical Form is retrieved from sound recording, using the Signal Processing and Artificial Neural Networks methods. It is important to choose appropriate signal characteristics, which creates feature vector describing the recording. Described recordings are clustered with Self-organizing maps. Limits of this approach are considered and possibilities of more complex structural recognition methods and semi-supervised learning utilisation are proposed.

English abstract

This paper describes the way how the Musical Form is retrieved from sound recording, using the Signal Processing and Artificial Neural Networks methods. It is important to choose appropriate signal characteristics, which creates feature vector describing the recording. Described recordings are clustered with Self-organizing maps. Limits of this approach are considered and possibilities of more complex structural recognition methods and semi-supervised learning utilisation are proposed.

Keywords

musical form, music information retrieval, self-organizing maps

Key words in English

musical form, music information retrieval, self-organizing maps

Authors

FEJFAR, J.; WEINLICHOVÁ, J.; ŠŤASTNÝ, J.

Released

23.06.2010

ISBN

978-80-214-4120-0

Book

Mendel 2010

Pages from

246

Pages to

251

Pages count

6

BibTex

@inproceedings{BUT34566,
  author="Jiří {Fejfar} and Jana {Weinlichová} and Jiří {Šťastný}",
  title="Musical Form Retrieval",
  booktitle="Mendel 2010",
  year="2010",
  number="16",
  pages="246--251",
  isbn="978-80-214-4120-0"
}