Garnishing a phonetic dictionary for ASR intake

Iben Nyholm Debess; Sandra Saxov Lamhauge; Peter Juel Henrichsen

Garnishing a phonetic dictionary for ASR intake

Iben Nyholm Debess, Sandra Saxov Lamhauge, Peter Juel Henrichsen

Føroyamálsdeildin - Faculty of Faroese Language and Literature

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

20 Downloads (Pure)

Abstract

We present a new method for preparing a lexical-phonetic database as a resource for acoustic model training. The research is an offshoot of the ongoing Project Ravnur (Speech Recognition for Faroese), but the method is language-independent. At NODALIDA 2019 we demonstrate the method (called SHARP) online, showing how a traditional lexical-phonetic dictionary (with a very rich phone inventory) is transformed into an ASR-friendly database (with reduced phonetics, preventing data sparseness). The mapping procedure is informed by a corpus of speech transcripts. We conclude with a discussion on the benefits of a well-thought-out BLARK design (Basic Language Resource Kit), making tools like SHARP possible.

Original language	English
Title of host publication	Proceedings of the 22nd Nordic Conference on Computational Linguistics
Place of Publication	Turku
Publisher	Linköping University Electronic Press
Pages	395-399
Number of pages	5
Volume	2019
Edition	September–October
Publication status	Published - 2019

Keywords

Phonetics
Databases

Access to Document

Debess, Heinrichsen, Lamhauge Garnishing a phonetic dictionary for ASR intake 2019Final published version, 258 KBLicence: CC BY

Cite this

@inbook{48e7d5c2cd7242bdafa4ca51d2fb82f9,

title = "Garnishing a phonetic dictionary for ASR intake",

abstract = "We present a new method for preparing a lexical-phonetic database as a resource for acoustic model training. The research is an offshoot of the ongoing Project Ravnur (Speech Recognition for Faroese), but the method is language-independent. At NODALIDA 2019 we demonstrate the method (called SHARP) online, showing how a traditional lexical-phonetic dictionary (with a very rich phone inventory) is transformed into an ASR-friendly database (with reduced phonetics, preventing data sparseness). The mapping procedure is informed by a corpus of speech transcripts. We conclude with a discussion on the benefits of a well-thought-out BLARK design (Basic Language Resource Kit), making tools like SHARP possible.",

keywords = "Phonetics, Databases",

author = "Debess, {Iben Nyholm} and Lamhauge, {Sandra Saxov} and {Juel Henrichsen}, Peter",

year = "2019",

language = "English",

volume = "2019",

pages = "395--399",

booktitle = "Proceedings of the 22nd Nordic Conference on Computational Linguistics",

publisher = "Link{\"o}ping University Electronic Press",

edition = "September–October",

}

Garnishing a phonetic dictionary for ASR intake. / Debess, Iben Nyholm ; Lamhauge, Sandra Saxov; Juel Henrichsen, Peter.
Proceedings of the 22nd Nordic Conference on Computational Linguistics. Vol. 2019 September–October. ed. Turku: Linköping University Electronic Press, 2019. p. 395-399.

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

TY - CHAP

T1 - Garnishing a phonetic dictionary for ASR intake

AU - Debess, Iben Nyholm

AU - Lamhauge, Sandra Saxov

AU - Juel Henrichsen, Peter

PY - 2019

Y1 - 2019

N2 - We present a new method for preparing a lexical-phonetic database as a resource for acoustic model training. The research is an offshoot of the ongoing Project Ravnur (Speech Recognition for Faroese), but the method is language-independent. At NODALIDA 2019 we demonstrate the method (called SHARP) online, showing how a traditional lexical-phonetic dictionary (with a very rich phone inventory) is transformed into an ASR-friendly database (with reduced phonetics, preventing data sparseness). The mapping procedure is informed by a corpus of speech transcripts. We conclude with a discussion on the benefits of a well-thought-out BLARK design (Basic Language Resource Kit), making tools like SHARP possible.

AB - We present a new method for preparing a lexical-phonetic database as a resource for acoustic model training. The research is an offshoot of the ongoing Project Ravnur (Speech Recognition for Faroese), but the method is language-independent. At NODALIDA 2019 we demonstrate the method (called SHARP) online, showing how a traditional lexical-phonetic dictionary (with a very rich phone inventory) is transformed into an ASR-friendly database (with reduced phonetics, preventing data sparseness). The mapping procedure is informed by a corpus of speech transcripts. We conclude with a discussion on the benefits of a well-thought-out BLARK design (Basic Language Resource Kit), making tools like SHARP possible.

KW - Phonetics

KW - Databases

UR - https://aclanthology.org/W19-6147/

M3 - Chapter

VL - 2019

SP - 395

EP - 399

BT - Proceedings of the 22nd Nordic Conference on Computational Linguistics

PB - Linköping University Electronic Press

CY - Turku

ER -

Garnishing a phonetic dictionary for ASR intake

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this