Recognition of Speech with Non-random Attributes

Czech title

Rozpoznavání řeči s nenáhodnými atributy

Type

conference paper

Language

English

Authors

Burget Lukáš, doc. Ing., Ph.D. (DCGM)
Černocký Jan, prof. Dr. Ing. (DCGM)

URL

Keywords

Speech recognition, Hidden Markov Models, HMM

Abstract

Most of current speech recognition systems are based on Hidden Markov Models assuming that speech features are sequence of stationary stochastic processes. However, there are certain speech attributes, such as background noise type or speaker voice color, that do not have stochastic character. This fact is often ignored, by designers of robust speaker independent recognition system. In this work, we investigate how the performance of a noisy speech recognition can be improved provided that we have prior knowledge about type and level of noise. Next, recognizer that is using separate models, each trained on a particular type and level of noise, is proposed for more appropriate modeling of speech.

Annotation

Most of current speech recognition systems are based on Hidden Markov Models assuming that speech features are sequence of stationary stochastic processes. However, there are certain speech attributes, such as background noise type or speaker voice color, that do not have stochastic character. This fact is often ignored, by designers of robust speaker independent recognition system. In this work, we investigate how the performance of a noisy speech recognition can be improved provided that we have prior knowledge about type and level of noise. Next, recognizer that is using separate models, each trained on a particular type and level of noise, is proposed for more appropriate modeling of speech.

Published

2003

Pages

6

Journal

Lecture Notes in Computer Science, vol. 2003, no. 09, ISSN 0302-9743

Proceedings

6th International Conference, TSD 2003 České Budějovice, Czech Republic, September 2003 Proceedings

ISBN

3-540-20024-X

Publisher

Springer Verlag

Place

České Budějovice

BibTeX

@inproceedings{BUT21496,
  author="Lukáš {Burget} and Jan {Černocký}",
  title="Recognition of Speech with Non-random Attributes",
  booktitle="6th International Conference, TSD 2003 České Budějovice, Czech Republic, September 2003 Proceedings",
  year="2003",
  journal="Lecture Notes in Computer Science",
  volume="2003",
  number="09",
  pages="6",
  publisher="Springer Verlag",
  address="České Budějovice",
  isbn="3-540-20024-X",
  issn="0302-9743",
  url="http://www.kiv.zcu.cz/events/tsd2003/, http://www.fit.vutbr.cz/~burget/phd_activities/burget_tsd03.pdf"
}