Detail publikace

Audio Data Classification by Means of New Algorithms

ŠŤASTNÝ, J. ŠKORPIL, V. FEJFAR, J.

Originální název

Audio Data Classification by Means of New Algorithms

Typ

článek ve sborníku ve WoS nebo Scopus

Jazyk

angličtina

Originální abstrakt

This paper describes classification of sound recordings based on their audio features. This is useful for querying large datasets, searching for recordings with some desired content. We use musical recordings as well as birdsongs recordings, which usually have rich structure and contain a lot of patterns suitable for classification. We present two different classification methods, one for musical recordings and one for birdsongs. These methods are compared and their differences are discussed. In case of musical recordings we use feature vectors describing the recording as a whole piece and we classify these feature vectors with the Self-organizing map and Learning Vector Quantization combination which represent a powerful algorithm using unlabeled as well as labeled data. In case of birdsongs we use feature vectors representing time frames of a recording.

Klíčová slova

sound processing, classification, semi-supervised learning, SOM, LVQ, HMM

Autoři

ŠŤASTNÝ, J.; ŠKORPIL, V.; FEJFAR, J.

Rok RIV

2013

Vydáno

2. 7. 2013

Nakladatel

TSP

Místo

Rome, Italy

ISBN

978-1-4799-0402-0

Kniha

Proceedings of the 36 th International Conference on Telecommunikations and Signal Processing (TSP 2013)

Číslo edice

1

Strany od

507

Strany do

511

Strany počet

5

BibTex

@inproceedings{BUT100910,
  author="Jiří {Šťastný} and Vladislav {Škorpil} and Jiří {Fejfar}",
  title="Audio Data Classification by Means of New Algorithms",
  booktitle="Proceedings of the 36 th International Conference on Telecommunikations and Signal Processing (TSP 2013)",
  year="2013",
  number="1",
  pages="507--511",
  publisher="TSP",
  address="Rome, Italy",
  isbn="978-1-4799-0402-0"
}