Modeling and recognition of phonetic and prosodic factors for improvements to acoustic speech recognition models

Sarah Borys; Mark Hasegawa-Johnson; Jennifer Cole; Aaron Cohen

Modeling and recognition of phonetic and prosodic factors for improvements to acoustic speech recognition models

Sarah Borys, Mark Hasegawa-Johnson, Jennifer Cole, Aaron Cohen

Linguistics

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

Abstract

This paper examines the usefulness of including prosodic and phonetic context information in the phoneme model of a speech recognizer. This is done by creating a series of prosodic and phonetic models and then comparing the mutual information between the observations and each possible context variable. Prosodic variables show improvement less often than phone context variables, however, prosodic variables generally show a larger increase in mutual information. A recognizer with allophones defined using the maximum mutual information prosodic and phonetic variables outperforms a recognizer with allophones defined exclusively using phonetic variables.

Original language	English (US)
Title of host publication	8th International Conference on Spoken Language Processing, ICSLP 2004
Publisher	International Speech Communication Association
Pages	3013-3016
Number of pages	4
State	Published - Jan 1 2004
Event	8th International Conference on Spoken Language Processing, ICSLP 2004 - Jeju, Jeju Island, Korea, Republic of Duration: Oct 4 2004 → Oct 8 2004

Other

Other	8th International Conference on Spoken Language Processing, ICSLP 2004
Country/Territory	Korea, Republic of
City	Jeju, Jeju Island
Period	10/4/04 → 10/8/04

ASJC Scopus subject areas

Language and Linguistics
Linguistics and Language

Cite this

@inproceedings{500243d47c8d465aba018d844d5c9e39,

title = "Modeling and recognition of phonetic and prosodic factors for improvements to acoustic speech recognition models",

abstract = "This paper examines the usefulness of including prosodic and phonetic context information in the phoneme model of a speech recognizer. This is done by creating a series of prosodic and phonetic models and then comparing the mutual information between the observations and each possible context variable. Prosodic variables show improvement less often than phone context variables, however, prosodic variables generally show a larger increase in mutual information. A recognizer with allophones defined using the maximum mutual information prosodic and phonetic variables outperforms a recognizer with allophones defined exclusively using phonetic variables.",

author = "Sarah Borys and Mark Hasegawa-Johnson and Jennifer Cole and Aaron Cohen",

year = "2004",

month = jan,

day = "1",

language = "English (US)",

pages = "3013--3016",

booktitle = "8th International Conference on Spoken Language Processing, ICSLP 2004",

publisher = "International Speech Communication Association",

note = "8th International Conference on Spoken Language Processing, ICSLP 2004 ; Conference date: 04-10-2004 Through 08-10-2004",

}

Borys, S, Hasegawa-Johnson, M, Cole, J & Cohen, A 2004, Modeling and recognition of phonetic and prosodic factors for improvements to acoustic speech recognition models. in 8th International Conference on Spoken Language Processing, ICSLP 2004. International Speech Communication Association, pp. 3013-3016, 8th International Conference on Spoken Language Processing, ICSLP 2004, Jeju, Jeju Island, Korea, Republic of, 10/4/04.

Modeling and recognition of phonetic and prosodic factors for improvements to acoustic speech recognition models. / Borys, Sarah; Hasegawa-Johnson, Mark; Cole, Jennifer et al.
8th International Conference on Spoken Language Processing, ICSLP 2004. International Speech Communication Association, 2004. p. 3013-3016.

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

TY - GEN

T1 - Modeling and recognition of phonetic and prosodic factors for improvements to acoustic speech recognition models

AU - Borys, Sarah

AU - Hasegawa-Johnson, Mark

AU - Cole, Jennifer

AU - Cohen, Aaron

PY - 2004/1/1

Y1 - 2004/1/1

N2 - This paper examines the usefulness of including prosodic and phonetic context information in the phoneme model of a speech recognizer. This is done by creating a series of prosodic and phonetic models and then comparing the mutual information between the observations and each possible context variable. Prosodic variables show improvement less often than phone context variables, however, prosodic variables generally show a larger increase in mutual information. A recognizer with allophones defined using the maximum mutual information prosodic and phonetic variables outperforms a recognizer with allophones defined exclusively using phonetic variables.

AB - This paper examines the usefulness of including prosodic and phonetic context information in the phoneme model of a speech recognizer. This is done by creating a series of prosodic and phonetic models and then comparing the mutual information between the observations and each possible context variable. Prosodic variables show improvement less often than phone context variables, however, prosodic variables generally show a larger increase in mutual information. A recognizer with allophones defined using the maximum mutual information prosodic and phonetic variables outperforms a recognizer with allophones defined exclusively using phonetic variables.

UR - http://www.scopus.com/inward/record.url?scp=85009115699&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85009115699&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85009115699

SP - 3013

EP - 3016

BT - 8th International Conference on Spoken Language Processing, ICSLP 2004

PB - International Speech Communication Association

T2 - 8th International Conference on Spoken Language Processing, ICSLP 2004

Y2 - 4 October 2004 through 8 October 2004

ER -

Modeling and recognition of phonetic and prosodic factors for improvements to acoustic speech recognition models

Abstract

Other

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this