Automatic analysis of pronunciations for children with speech sound disorders

Shiran Dudy, Steven Bedrick, Meysam Asgari, Alexander Kain

Research output: Contribution to journalArticle

Abstract

Computer-Assisted Pronunciation Training (CAPT) systems aim to help a child learn the correct pronunciations of words. However, while there are many online commercial CAPT apps, there is no consensus among Speech Language Therapists (SLPs) or non-professionals about which CAPT systems, if any, work well. The prevailing assumption is that practicing with such programs is less reliable and thus does not provide the feedback necessary to allow children to improve their performance. The most common method for assessing pronunciation performance is the Goodness of Pronunciation (GOP) technique. Our paper proposes two new GOP techniques. We have found that pronunciation models that use explicit knowledge about error pronunciation patterns can lead to more accurate classification whether a phoneme was correctly pronounced or not. We evaluate the proposed pronunciation assessment methods against a baseline state of the art GOP approach, and show that the proposed techniques lead to classification performance that is more similar to that of a human expert.

LanguageEnglish (US)
Pages62-84
Number of pages23
JournalComputer Speech and Language
Volume50
DOIs
StatePublished - Jul 1 2018

Fingerprint

Disorder
Acoustic waves
Application programs
Baseline
Feedback
Necessary
Evaluate
Children
Speech
Training
Sound
Model
Knowledge
Language
Human

Keywords

  • Diagnostic tools
  • Educational software
  • Goodness of Pronunciation
  • Speech disorders
  • Speech recognition
  • Support Vector Machine

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Human-Computer Interaction

Cite this

Automatic analysis of pronunciations for children with speech sound disorders. / Dudy, Shiran; Bedrick, Steven; Asgari, Meysam; Kain, Alexander.

In: Computer Speech and Language, Vol. 50, 01.07.2018, p. 62-84.

Research output: Contribution to journalArticle

@article{dfbce3b3c3544b01ade38288f265995b,
title = "Automatic analysis of pronunciations for children with speech sound disorders",
abstract = "Computer-Assisted Pronunciation Training (CAPT) systems aim to help a child learn the correct pronunciations of words. However, while there are many online commercial CAPT apps, there is no consensus among Speech Language Therapists (SLPs) or non-professionals about which CAPT systems, if any, work well. The prevailing assumption is that practicing with such programs is less reliable and thus does not provide the feedback necessary to allow children to improve their performance. The most common method for assessing pronunciation performance is the Goodness of Pronunciation (GOP) technique. Our paper proposes two new GOP techniques. We have found that pronunciation models that use explicit knowledge about error pronunciation patterns can lead to more accurate classification whether a phoneme was correctly pronounced or not. We evaluate the proposed pronunciation assessment methods against a baseline state of the art GOP approach, and show that the proposed techniques lead to classification performance that is more similar to that of a human expert.",
keywords = "Diagnostic tools, Educational software, Goodness of Pronunciation, Speech disorders, Speech recognition, Support Vector Machine",
author = "Shiran Dudy and Steven Bedrick and Meysam Asgari and Alexander Kain",
year = "2018",
month = "7",
day = "1",
doi = "10.1016/j.csl.2017.12.006",
language = "English (US)",
volume = "50",
pages = "62--84",
journal = "Computer Speech and Language",
issn = "0885-2308",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Automatic analysis of pronunciations for children with speech sound disorders

AU - Dudy, Shiran

AU - Bedrick, Steven

AU - Asgari, Meysam

AU - Kain, Alexander

PY - 2018/7/1

Y1 - 2018/7/1

N2 - Computer-Assisted Pronunciation Training (CAPT) systems aim to help a child learn the correct pronunciations of words. However, while there are many online commercial CAPT apps, there is no consensus among Speech Language Therapists (SLPs) or non-professionals about which CAPT systems, if any, work well. The prevailing assumption is that practicing with such programs is less reliable and thus does not provide the feedback necessary to allow children to improve their performance. The most common method for assessing pronunciation performance is the Goodness of Pronunciation (GOP) technique. Our paper proposes two new GOP techniques. We have found that pronunciation models that use explicit knowledge about error pronunciation patterns can lead to more accurate classification whether a phoneme was correctly pronounced or not. We evaluate the proposed pronunciation assessment methods against a baseline state of the art GOP approach, and show that the proposed techniques lead to classification performance that is more similar to that of a human expert.

AB - Computer-Assisted Pronunciation Training (CAPT) systems aim to help a child learn the correct pronunciations of words. However, while there are many online commercial CAPT apps, there is no consensus among Speech Language Therapists (SLPs) or non-professionals about which CAPT systems, if any, work well. The prevailing assumption is that practicing with such programs is less reliable and thus does not provide the feedback necessary to allow children to improve their performance. The most common method for assessing pronunciation performance is the Goodness of Pronunciation (GOP) technique. Our paper proposes two new GOP techniques. We have found that pronunciation models that use explicit knowledge about error pronunciation patterns can lead to more accurate classification whether a phoneme was correctly pronounced or not. We evaluate the proposed pronunciation assessment methods against a baseline state of the art GOP approach, and show that the proposed techniques lead to classification performance that is more similar to that of a human expert.

KW - Diagnostic tools

KW - Educational software

KW - Goodness of Pronunciation

KW - Speech disorders

KW - Speech recognition

KW - Support Vector Machine

UR - http://www.scopus.com/inward/record.url?scp=85040252032&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85040252032&partnerID=8YFLogxK

U2 - 10.1016/j.csl.2017.12.006

DO - 10.1016/j.csl.2017.12.006

M3 - Article

VL - 50

SP - 62

EP - 84

JO - Computer Speech and Language

T2 - Computer Speech and Language

JF - Computer Speech and Language

SN - 0885-2308

ER -