Using clinician annotations to improve automatic speech recognition of stuttered speech

Peter Heeman, Rebecca Lunsford, Andy McMillin, J. Scott Yaruss

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

In treating people who stutter, clinicians often have their clients read a story in order to determine their stuttering frequency. As the client is speaking, the clinician annotates each disfluency. For further analysis of the client's speech, it is useful to have a word transcription of what was said. However, as these are realtime annotations, they are not always correct, and they usually lag where the actual disfluency occurred. We have built a tool that rescores a word lattice taking into account the clinician's annotations. In the paper, we describe how we incorporate the clinician's annotations, and the improvement over a baseline version. This approach of leveraging clinician annotations can be used for other clinical tasks where a word transcription is useful for further or richer analysis.

Original languageEnglish (US)
Pages (from-to)2651-2655
Number of pages5
JournalUnknown Journal
Volume08-12-September-2016
DOIs
StatePublished - 2016

Fingerprint

Stuttering
Automatic Speech Recognition
Transcription
Speech recognition
Annotation
Baseline
Real-time
Speech
Clinicians

Keywords

  • Automatic speech recognition
  • Disfluency counts
  • Stuttering
  • User-interface

ASJC Scopus subject areas

  • Language and Linguistics
  • Human-Computer Interaction
  • Signal Processing
  • Software
  • Modeling and Simulation

Cite this

Using clinician annotations to improve automatic speech recognition of stuttered speech. / Heeman, Peter; Lunsford, Rebecca; McMillin, Andy; Yaruss, J. Scott.

In: Unknown Journal, Vol. 08-12-September-2016, 2016, p. 2651-2655.

Research output: Contribution to journalArticle

Heeman, Peter ; Lunsford, Rebecca ; McMillin, Andy ; Yaruss, J. Scott. / Using clinician annotations to improve automatic speech recognition of stuttered speech. In: Unknown Journal. 2016 ; Vol. 08-12-September-2016. pp. 2651-2655.
@article{f50262e6c37148da8424c111e2f970a8,
title = "Using clinician annotations to improve automatic speech recognition of stuttered speech",
abstract = "In treating people who stutter, clinicians often have their clients read a story in order to determine their stuttering frequency. As the client is speaking, the clinician annotates each disfluency. For further analysis of the client's speech, it is useful to have a word transcription of what was said. However, as these are realtime annotations, they are not always correct, and they usually lag where the actual disfluency occurred. We have built a tool that rescores a word lattice taking into account the clinician's annotations. In the paper, we describe how we incorporate the clinician's annotations, and the improvement over a baseline version. This approach of leveraging clinician annotations can be used for other clinical tasks where a word transcription is useful for further or richer analysis.",
keywords = "Automatic speech recognition, Disfluency counts, Stuttering, User-interface",
author = "Peter Heeman and Rebecca Lunsford and Andy McMillin and Yaruss, {J. Scott}",
year = "2016",
doi = "10.21437/Interspeech.2016-1388",
language = "English (US)",
volume = "08-12-September-2016",
pages = "2651--2655",
journal = "Indian Journal of Rheumatology",
issn = "0973-3698",
publisher = "Elsevier (Singapore) Pte Ltd",

}

TY - JOUR

T1 - Using clinician annotations to improve automatic speech recognition of stuttered speech

AU - Heeman, Peter

AU - Lunsford, Rebecca

AU - McMillin, Andy

AU - Yaruss, J. Scott

PY - 2016

Y1 - 2016

N2 - In treating people who stutter, clinicians often have their clients read a story in order to determine their stuttering frequency. As the client is speaking, the clinician annotates each disfluency. For further analysis of the client's speech, it is useful to have a word transcription of what was said. However, as these are realtime annotations, they are not always correct, and they usually lag where the actual disfluency occurred. We have built a tool that rescores a word lattice taking into account the clinician's annotations. In the paper, we describe how we incorporate the clinician's annotations, and the improvement over a baseline version. This approach of leveraging clinician annotations can be used for other clinical tasks where a word transcription is useful for further or richer analysis.

AB - In treating people who stutter, clinicians often have their clients read a story in order to determine their stuttering frequency. As the client is speaking, the clinician annotates each disfluency. For further analysis of the client's speech, it is useful to have a word transcription of what was said. However, as these are realtime annotations, they are not always correct, and they usually lag where the actual disfluency occurred. We have built a tool that rescores a word lattice taking into account the clinician's annotations. In the paper, we describe how we incorporate the clinician's annotations, and the improvement over a baseline version. This approach of leveraging clinician annotations can be used for other clinical tasks where a word transcription is useful for further or richer analysis.

KW - Automatic speech recognition

KW - Disfluency counts

KW - Stuttering

KW - User-interface

UR - http://www.scopus.com/inward/record.url?scp=84994210599&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84994210599&partnerID=8YFLogxK

U2 - 10.21437/Interspeech.2016-1388

DO - 10.21437/Interspeech.2016-1388

M3 - Article

AN - SCOPUS:84994210599

VL - 08-12-September-2016

SP - 2651

EP - 2655

JO - Indian Journal of Rheumatology

JF - Indian Journal of Rheumatology

SN - 0973-3698

ER -