Automatic Screening of Children with Speech Sound Disorders Using Paralinguistic Features

Mostafa Shahin, Beena Ahmed, Daniel V. Smith, Andreas Duenser, Julien Epps

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Subjective screening of children with speech disorders is costly, time consuming and infeasible due to the limited availability of Speech and Language Pathologists (SLPs). Therefore, there is an increasing interest in automatic speech analysis of children with speech disorders as it can offer a practical alternative to human assessment. Paralinguistic features are a set of low-level descriptors commonly used in speech emotion recognition. However, they have not yet been examined with childhood speech sound disorders such as, apraxia-of-speech and phonological and articulation disorders. In this paper, we investigated the effectiveness of paralinguistic features in discriminating between typically developing children and those who suffer from different types of speech sound disorders. Two types of standard paralinguistic features were explored, the Geneva Minimalistic Acoustic Parameter Set (GeMAPS) and its extended version, (eGeMAPS) feature sets. We applied feature selection to find the most discriminant set of features and employed binary classification using a support vector machine (SVM) to discriminate between the two groups. The method was tested on a recently-released public speech corpus collected from typically developing children and children with various types of speech sound disorders. The system achieved segment-level and subject-level unweighted average recall (UAR) of around 78% and 87% respectively.

Original languageEnglish
Title of host publication2019 IEEE 29th International Workshop on Machine Learning for Signal Processing, MLSP 2019
PublisherIEEE Computer Society
ISBN (Electronic)9781728108247
DOIs
Publication statusPublished - Oct 2019
Event29th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2019 - Pittsburgh, United States
Duration: 13 Oct 201916 Oct 2019

Publication series

NameIEEE International Workshop on Machine Learning for Signal Processing, MLSP
Volume2019-October
ISSN (Print)2161-0363
ISSN (Electronic)2161-0371

Conference

Conference29th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2019
CountryUnited States
CityPittsburgh
Period13/10/1916/10/19

Fingerprint

Screening
Acoustic waves
Speech analysis
Speech recognition
Support vector machines
Feature extraction
Acoustics
Availability

Keywords

  • paralinguistic features
  • speech sound disorders
  • speech therapy

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Signal Processing

Cite this

Shahin, M., Ahmed, B., Smith, D. V., Duenser, A., & Epps, J. (2019). Automatic Screening of Children with Speech Sound Disorders Using Paralinguistic Features. In 2019 IEEE 29th International Workshop on Machine Learning for Signal Processing, MLSP 2019 [8918725] (IEEE International Workshop on Machine Learning for Signal Processing, MLSP; Vol. 2019-October). IEEE Computer Society. https://doi.org/10.1109/MLSP.2019.8918725

Automatic Screening of Children with Speech Sound Disorders Using Paralinguistic Features. / Shahin, Mostafa; Ahmed, Beena; Smith, Daniel V.; Duenser, Andreas; Epps, Julien.

2019 IEEE 29th International Workshop on Machine Learning for Signal Processing, MLSP 2019. IEEE Computer Society, 2019. 8918725 (IEEE International Workshop on Machine Learning for Signal Processing, MLSP; Vol. 2019-October).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Shahin, M, Ahmed, B, Smith, DV, Duenser, A & Epps, J 2019, Automatic Screening of Children with Speech Sound Disorders Using Paralinguistic Features. in 2019 IEEE 29th International Workshop on Machine Learning for Signal Processing, MLSP 2019., 8918725, IEEE International Workshop on Machine Learning for Signal Processing, MLSP, vol. 2019-October, IEEE Computer Society, 29th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2019, Pittsburgh, United States, 13/10/19. https://doi.org/10.1109/MLSP.2019.8918725
Shahin M, Ahmed B, Smith DV, Duenser A, Epps J. Automatic Screening of Children with Speech Sound Disorders Using Paralinguistic Features. In 2019 IEEE 29th International Workshop on Machine Learning for Signal Processing, MLSP 2019. IEEE Computer Society. 2019. 8918725. (IEEE International Workshop on Machine Learning for Signal Processing, MLSP). https://doi.org/10.1109/MLSP.2019.8918725
Shahin, Mostafa ; Ahmed, Beena ; Smith, Daniel V. ; Duenser, Andreas ; Epps, Julien. / Automatic Screening of Children with Speech Sound Disorders Using Paralinguistic Features. 2019 IEEE 29th International Workshop on Machine Learning for Signal Processing, MLSP 2019. IEEE Computer Society, 2019. (IEEE International Workshop on Machine Learning for Signal Processing, MLSP).
@inproceedings{cc9dd96ed83245439021ad2afdb98f7d,
title = "Automatic Screening of Children with Speech Sound Disorders Using Paralinguistic Features",
abstract = "Subjective screening of children with speech disorders is costly, time consuming and infeasible due to the limited availability of Speech and Language Pathologists (SLPs). Therefore, there is an increasing interest in automatic speech analysis of children with speech disorders as it can offer a practical alternative to human assessment. Paralinguistic features are a set of low-level descriptors commonly used in speech emotion recognition. However, they have not yet been examined with childhood speech sound disorders such as, apraxia-of-speech and phonological and articulation disorders. In this paper, we investigated the effectiveness of paralinguistic features in discriminating between typically developing children and those who suffer from different types of speech sound disorders. Two types of standard paralinguistic features were explored, the Geneva Minimalistic Acoustic Parameter Set (GeMAPS) and its extended version, (eGeMAPS) feature sets. We applied feature selection to find the most discriminant set of features and employed binary classification using a support vector machine (SVM) to discriminate between the two groups. The method was tested on a recently-released public speech corpus collected from typically developing children and children with various types of speech sound disorders. The system achieved segment-level and subject-level unweighted average recall (UAR) of around 78{\%} and 87{\%} respectively.",
keywords = "paralinguistic features, speech sound disorders, speech therapy",
author = "Mostafa Shahin and Beena Ahmed and Smith, {Daniel V.} and Andreas Duenser and Julien Epps",
year = "2019",
month = "10",
doi = "10.1109/MLSP.2019.8918725",
language = "English",
series = "IEEE International Workshop on Machine Learning for Signal Processing, MLSP",
publisher = "IEEE Computer Society",
booktitle = "2019 IEEE 29th International Workshop on Machine Learning for Signal Processing, MLSP 2019",

}

TY - GEN

T1 - Automatic Screening of Children with Speech Sound Disorders Using Paralinguistic Features

AU - Shahin, Mostafa

AU - Ahmed, Beena

AU - Smith, Daniel V.

AU - Duenser, Andreas

AU - Epps, Julien

PY - 2019/10

Y1 - 2019/10

N2 - Subjective screening of children with speech disorders is costly, time consuming and infeasible due to the limited availability of Speech and Language Pathologists (SLPs). Therefore, there is an increasing interest in automatic speech analysis of children with speech disorders as it can offer a practical alternative to human assessment. Paralinguistic features are a set of low-level descriptors commonly used in speech emotion recognition. However, they have not yet been examined with childhood speech sound disorders such as, apraxia-of-speech and phonological and articulation disorders. In this paper, we investigated the effectiveness of paralinguistic features in discriminating between typically developing children and those who suffer from different types of speech sound disorders. Two types of standard paralinguistic features were explored, the Geneva Minimalistic Acoustic Parameter Set (GeMAPS) and its extended version, (eGeMAPS) feature sets. We applied feature selection to find the most discriminant set of features and employed binary classification using a support vector machine (SVM) to discriminate between the two groups. The method was tested on a recently-released public speech corpus collected from typically developing children and children with various types of speech sound disorders. The system achieved segment-level and subject-level unweighted average recall (UAR) of around 78% and 87% respectively.

AB - Subjective screening of children with speech disorders is costly, time consuming and infeasible due to the limited availability of Speech and Language Pathologists (SLPs). Therefore, there is an increasing interest in automatic speech analysis of children with speech disorders as it can offer a practical alternative to human assessment. Paralinguistic features are a set of low-level descriptors commonly used in speech emotion recognition. However, they have not yet been examined with childhood speech sound disorders such as, apraxia-of-speech and phonological and articulation disorders. In this paper, we investigated the effectiveness of paralinguistic features in discriminating between typically developing children and those who suffer from different types of speech sound disorders. Two types of standard paralinguistic features were explored, the Geneva Minimalistic Acoustic Parameter Set (GeMAPS) and its extended version, (eGeMAPS) feature sets. We applied feature selection to find the most discriminant set of features and employed binary classification using a support vector machine (SVM) to discriminate between the two groups. The method was tested on a recently-released public speech corpus collected from typically developing children and children with various types of speech sound disorders. The system achieved segment-level and subject-level unweighted average recall (UAR) of around 78% and 87% respectively.

KW - paralinguistic features

KW - speech sound disorders

KW - speech therapy

UR - http://www.scopus.com/inward/record.url?scp=85077699531&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85077699531&partnerID=8YFLogxK

U2 - 10.1109/MLSP.2019.8918725

DO - 10.1109/MLSP.2019.8918725

M3 - Conference contribution

AN - SCOPUS:85077699531

T3 - IEEE International Workshop on Machine Learning for Signal Processing, MLSP

BT - 2019 IEEE 29th International Workshop on Machine Learning for Signal Processing, MLSP 2019

PB - IEEE Computer Society

ER -