Publication Details

Multi-task Neural Networks For Speech Recognition

EGOROVA, E. Multi-task Neural Networks For Speech Recognition. Proceedings of the 20th Student Conference, EEICT 2014. Volume 2. Brno: Brno University of Technology, 2014. p. 24-26. ISBN: 978-80-214-4923-7.

Czech title

Víceúkolové trénování neuronových sítí pro rozpoznávání řeči

Type

conference paper

Language

English

Authors

Egorova Ekaterina, Ing., Ph.D.

URL

Keywords

Speech recognition, neural networks, deep neural networks, multi-task neural networks.

Annotation

The article covers experiments on TIMIT database exploring the possibility of using multitask neural networks for speech recognition. Multi-task neural networks are deep neural networks solving several different classification tasks simultaneously. The secondary tasks chosen for the experiments are gender, context, articulatory characteristics and a fusion of some of them. The experiments show that addition of such tasks can enhance the learning and improve recognition accuracy.

Published

2014

Pages

24–26

Proceedings

Proceedings of the 20th Student Conference, EEICT 2014

Series

Volume 2

ISBN

978-80-214-4923-7

Publisher

Brno University of Technology

Place

Brno

BibTeX

@inproceedings{BUT111586,
  author="Ekaterina {Egorova}",
  title="Multi-task Neural Networks For Speech Recognition",
  booktitle="Proceedings of the 20th Student Conference, EEICT 2014",
  year="2014",
  series="Volume 2",
  pages="24--26",
  publisher="Brno University of Technology",
  address="Brno",
  isbn="978-80-214-4923-7",
  url="http://www.feec.vutbr.cz/EEICT/wp-content/uploads/2014/04/2014-sbornik-mgr.pdf"
}