Automatically measuring L2 speech fluency without the need of ASR

A proof-of-concept study with Japanese learners of French

Lionel Fontan, Maxime Le Coz, Sylvain Matthieu Julien Detey

Research output: Contribution to journalConference article

1 Citation (Scopus)

Abstract

This research work investigates the possibility of using automatic acoustic measures to assess speech fluency in the context of second language (L2) acquisition. To this end, three experts rated speech recordings of Japanese learners of French who were instructed to read aloud a 21-sentence-long text. A Forward-Backward Divergence Segmentation (FBDS) algorithm was used to segment speech recordings (sentences) into acoustically homogeneous units at a subphonemic scale. The FBDS processing results were used - along with more classic measures such as raw percentage of speech and length/standard deviation of silent pauses - to estimate speech rate and regularity of speech rate, while a formant tracking algorithm was used to estimate speech fluidity (i.e., quality of coarticulation). A step-by-step multiple linear regression was finally computed to predict the experts' mean fluency ratings. Results show that FBDS-derived measures, raw percentage of speech, and standard deviation of the first formant curve derivative can be combined together to calculate accurate estimates of speakers' fluency scores (R = .92; P < .001). As only low-level signal features were used in the study, the method could also be relevant for the assessment of speakers of other target languages, as well as for the assessment of disordered speech.

Original languageEnglish
Pages (from-to)2544-2548
Number of pages5
JournalProceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
Volume2018-September
DOIs
Publication statusPublished - 2018 Jan 1
Event19th Annual Conference of the International Speech Communication, INTERSPEECH 2018 - Hyderabad, India
Duration: 2018 Sep 22018 Sep 6

Fingerprint

Divergence
Segmentation
Standard deviation
Percentage
Estimate
Speech
Concepts
Fluency
Multiple Linear Regression
Fluidity
Linear regression
Acoustics
Regularity
Derivatives
Calculate
Derivative
Predict
Curve
Target
Unit

Keywords

  • Automatic measures
  • French
  • Japanese
  • Learner corpus
  • Second language acquisition
  • Speech fluency

ASJC Scopus subject areas

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

Cite this

@article{dc3bf323913f4229829cbd083e305797,
title = "Automatically measuring L2 speech fluency without the need of ASR: A proof-of-concept study with Japanese learners of French",
abstract = "This research work investigates the possibility of using automatic acoustic measures to assess speech fluency in the context of second language (L2) acquisition. To this end, three experts rated speech recordings of Japanese learners of French who were instructed to read aloud a 21-sentence-long text. A Forward-Backward Divergence Segmentation (FBDS) algorithm was used to segment speech recordings (sentences) into acoustically homogeneous units at a subphonemic scale. The FBDS processing results were used - along with more classic measures such as raw percentage of speech and length/standard deviation of silent pauses - to estimate speech rate and regularity of speech rate, while a formant tracking algorithm was used to estimate speech fluidity (i.e., quality of coarticulation). A step-by-step multiple linear regression was finally computed to predict the experts' mean fluency ratings. Results show that FBDS-derived measures, raw percentage of speech, and standard deviation of the first formant curve derivative can be combined together to calculate accurate estimates of speakers' fluency scores (R = .92; P < .001). As only low-level signal features were used in the study, the method could also be relevant for the assessment of speakers of other target languages, as well as for the assessment of disordered speech.",
keywords = "Automatic measures, French, Japanese, Learner corpus, Second language acquisition, Speech fluency",
author = "Lionel Fontan and {Le Coz}, Maxime and Detey, {Sylvain Matthieu Julien}",
year = "2018",
month = "1",
day = "1",
doi = "10.21437/Interspeech.2018-1336",
language = "English",
volume = "2018-September",
pages = "2544--2548",
journal = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",
issn = "2308-457X",

}

TY - JOUR

T1 - Automatically measuring L2 speech fluency without the need of ASR

T2 - A proof-of-concept study with Japanese learners of French

AU - Fontan, Lionel

AU - Le Coz, Maxime

AU - Detey, Sylvain Matthieu Julien

PY - 2018/1/1

Y1 - 2018/1/1

N2 - This research work investigates the possibility of using automatic acoustic measures to assess speech fluency in the context of second language (L2) acquisition. To this end, three experts rated speech recordings of Japanese learners of French who were instructed to read aloud a 21-sentence-long text. A Forward-Backward Divergence Segmentation (FBDS) algorithm was used to segment speech recordings (sentences) into acoustically homogeneous units at a subphonemic scale. The FBDS processing results were used - along with more classic measures such as raw percentage of speech and length/standard deviation of silent pauses - to estimate speech rate and regularity of speech rate, while a formant tracking algorithm was used to estimate speech fluidity (i.e., quality of coarticulation). A step-by-step multiple linear regression was finally computed to predict the experts' mean fluency ratings. Results show that FBDS-derived measures, raw percentage of speech, and standard deviation of the first formant curve derivative can be combined together to calculate accurate estimates of speakers' fluency scores (R = .92; P < .001). As only low-level signal features were used in the study, the method could also be relevant for the assessment of speakers of other target languages, as well as for the assessment of disordered speech.

AB - This research work investigates the possibility of using automatic acoustic measures to assess speech fluency in the context of second language (L2) acquisition. To this end, three experts rated speech recordings of Japanese learners of French who were instructed to read aloud a 21-sentence-long text. A Forward-Backward Divergence Segmentation (FBDS) algorithm was used to segment speech recordings (sentences) into acoustically homogeneous units at a subphonemic scale. The FBDS processing results were used - along with more classic measures such as raw percentage of speech and length/standard deviation of silent pauses - to estimate speech rate and regularity of speech rate, while a formant tracking algorithm was used to estimate speech fluidity (i.e., quality of coarticulation). A step-by-step multiple linear regression was finally computed to predict the experts' mean fluency ratings. Results show that FBDS-derived measures, raw percentage of speech, and standard deviation of the first formant curve derivative can be combined together to calculate accurate estimates of speakers' fluency scores (R = .92; P < .001). As only low-level signal features were used in the study, the method could also be relevant for the assessment of speakers of other target languages, as well as for the assessment of disordered speech.

KW - Automatic measures

KW - French

KW - Japanese

KW - Learner corpus

KW - Second language acquisition

KW - Speech fluency

UR - http://www.scopus.com/inward/record.url?scp=85054984906&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054984906&partnerID=8YFLogxK

U2 - 10.21437/Interspeech.2018-1336

DO - 10.21437/Interspeech.2018-1336

M3 - Conference article

VL - 2018-September

SP - 2544

EP - 2548

JO - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

JF - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

SN - 2308-457X

ER -