Publication detail

Manual and Semi-Automatic Approaches to Building a Multilingual Phoneme Set

EGOROVA, E. VESELÝ, K. KARAFIÁT, M. JANDA, M. ČERNOCKÝ, J.

Original Title

Manual and Semi-Automatic Approaches to Building a Multilingual Phoneme Set

Type

conference paper

Language

English

Original Abstract

This articles describes a comparison between manual and semi-automatic approaches to building a multilingual phoneme set. The two approaches were compared in cases of 1) a multilingual system with abundant data for all the languages, 2) multilingual systems excluding target language 3) multilingual systems with small amount of data for target languages. The work shows that careful choice of merging methods can help improve recognition of languages with no or little training data and reasonably reduce multilingual phoneme set without losing a lot of accuracy.

Keywords

multilingual speech recognition, phoneme set mapping, phoneme confusion matrix

Authors

EGOROVA, E.; VESELÝ, K.; KARAFIÁT, M.; JANDA, M.; ČERNOCKÝ, J.

RIV year

2013

Released

27. 5. 2013

Publisher

IEEE Signal Processing Society

Location

Vancouver

ISBN

978-1-4799-0355-9

Book

Proceedings of ICASSP 2013

Pages from

7324

Pages to

7328

Pages count

5

URL

BibTex

@inproceedings{BUT103490,
  author="Ekaterina {Egorova} and Karel {Veselý} and Martin {Karafiát} and Miloš {Janda} and Jan {Černocký}",
  title="Manual and Semi-Automatic Approaches to Building a Multilingual Phoneme Set",
  booktitle="Proceedings of ICASSP 2013",
  year="2013",
  pages="7324--7328",
  publisher="IEEE Signal Processing Society",
  address="Vancouver",
  isbn="978-1-4799-0355-9",
  url="https://www.fit.vut.cz/research/publication/10323/"
}