Speech recognition via phonetically-featured syllables

S. King; Paul Taylor; J. Frankel; K. Richmond

Speech recognition via phonetically-featured syllables

S. King, Paul Taylor, J. Frankel, K. Richmond

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We describe recent work on two new automatic speech recognition systems. The first part of this paper describes the components of a system based on phonological features (which we call EspressoA) in which the values of these features are estimated from the speech signal before being used as the basis for recognition. In the second part of the paper, another system (which we call EspressoB) is described in which articulatory parameters are used instead of phonological features and a linear dynamical system model is used to perform recognition from automatically estimated values of these articulatory parameters.

Original language	English
Title of host publication	PHONUS 5
Subtitle of host publication	Proceedings of the Workshop on "Phonetics and Phonology in ASR. Parameters and Features, and their Implications"
Editors	W. J. Barry, J Koreman
Place of Publication	Saarbruken
Publisher	Institute of Phonetics
Pages	15-34
Number of pages	20
Volume	5
Publication status	Published - 2000

Access to Document

http://www.coli.uni-saarland.de/groups/WB/Phonetics/page.php?id=phonus5

ROPA: SPEECH RECOGNITION
King, S.
EPSRC
1/10/99 → 31/01/02
Project: Research

Cite this

King, S., Taylor, P., Frankel, J., & Richmond, K. (2000). Speech recognition via phonetically-featured syllables. In W. J. Barry, & J. Koreman (Eds.), PHONUS 5 : Proceedings of the Workshop on "Phonetics and Phonology in ASR. Parameters and Features, and their Implications" (Vol. 5, pp. 15-34). Institute of Phonetics. http://www.coli.uni-saarland.de/groups/WB/Phonetics/page.php?id=phonus5

@inproceedings{e045621a8f074cb4926202907c584840,

title = "Speech recognition via phonetically-featured syllables",

abstract = "We describe recent work on two new automatic speech recognition systems. The first part of this paper describes the components of a system based on phonological features (which we call EspressoA) in which the values of these features are estimated from the speech signal before being used as the basis for recognition. In the second part of the paper, another system (which we call EspressoB) is described in which articulatory parameters are used instead of phonological features and a linear dynamical system model is used to perform recognition from automatically estimated values of these articulatory parameters.",

author = "S. King and Paul Taylor and J. Frankel and K. Richmond",

note = "No ISSN given",

year = "2000",

language = "English",

volume = "5",

pages = "15--34",

editor = "Barry, {W. J.} and J Koreman",

booktitle = "PHONUS 5",

publisher = "Institute of Phonetics",

}

King, S, Taylor, P, Frankel, J & Richmond, K 2000, Speech recognition via phonetically-featured syllables. in WJ Barry & J Koreman (eds), PHONUS 5 : Proceedings of the Workshop on "Phonetics and Phonology in ASR. Parameters and Features, and their Implications". vol. 5, Institute of Phonetics, Saarbruken, pp. 15-34. <http://www.coli.uni-saarland.de/groups/WB/Phonetics/page.php?id=phonus5>

Speech recognition via phonetically-featured syllables. / King, S.; Taylor, Paul; Frankel, J. et al.
PHONUS 5 : Proceedings of the Workshop on "Phonetics and Phonology in ASR. Parameters and Features, and their Implications". ed. / W. J. Barry; J Koreman. Vol. 5 Saarbruken: Institute of Phonetics, 2000. p. 15-34.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Speech recognition via phonetically-featured syllables

AU - King, S.

AU - Taylor, Paul

AU - Frankel, J.

AU - Richmond, K.

N1 - No ISSN given

PY - 2000

Y1 - 2000

N2 - We describe recent work on two new automatic speech recognition systems. The first part of this paper describes the components of a system based on phonological features (which we call EspressoA) in which the values of these features are estimated from the speech signal before being used as the basis for recognition. In the second part of the paper, another system (which we call EspressoB) is described in which articulatory parameters are used instead of phonological features and a linear dynamical system model is used to perform recognition from automatically estimated values of these articulatory parameters.

AB - We describe recent work on two new automatic speech recognition systems. The first part of this paper describes the components of a system based on phonological features (which we call EspressoA) in which the values of these features are estimated from the speech signal before being used as the basis for recognition. In the second part of the paper, another system (which we call EspressoB) is described in which articulatory parameters are used instead of phonological features and a linear dynamical system model is used to perform recognition from automatically estimated values of these articulatory parameters.

M3 - Conference contribution

VL - 5

SP - 15

EP - 34

BT - PHONUS 5

A2 - Barry, W. J.

A2 - Koreman, J

PB - Institute of Phonetics

CY - Saarbruken

ER -

Speech recognition via phonetically-featured syllables

Abstract

Access to Document

Fingerprint

Projects

ROPA: SPEECH RECOGNITION

Cite this