Improving Text-Independent Forced Alignment to Support Speech-Language Pathologists with Phonetic Transcription
Sensors · 2023
DOI: 10.3390/s23249650 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, 2023. Verfasst von Ying Li und 6 weiteren Autor:innen. Beteiligte Einrichtung: School of EECMS, Curtin University, Bentley, WA 6102, Australia. Hauptthemen laut MEDLINE: Voice, Speech Perception. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: The Department of Health WA, The Prompt Institute, PROMPT Institute Research.
Abstract (Originalsprache)
Problem: Phonetic transcription is crucial in diagnosing speech sound disorders (SSDs) but is susceptible to transcriber experience and perceptual bias. Current forced alignment (FA) tools, which annotate audio files to determine spoken content and its placement, often require manual transcription, limiting their effectiveness.
Method
We introduce a novel, text-independent forced alignment model that autonomously recognises individual phonemes and their boundaries, addressing these limitations. Our approach leverages an advanced, pre-trained wav2vec 2.0 model to segment speech into tokens and recognise them automatically. To accurately identify phoneme boundaries, we utilise an unsupervised segmentation tool, UnsupSeg. Labelling of segments employs nearest-neighbour classification with wav2vec 2.0 labels, before connectionist temporal classification (CTC) collapse, determining class labels based on maximum overlap. Additional post-processing, including overfitting cleaning and voice activity detection, is implemented to enhance segmentation.
Results
We benchmarked our model against existing methods using the TIMIT dataset for normal speakers and, for the first time, evaluated its performance on the TORGO dataset containing SSD speakers. Our model demonstrated competitive performance, achieving a harmonic mean score of 76.88% on TIMIT and 70.31% on TORGO.
Implications
This research presents a significant advancement in the assessment and diagnosis of SSDs, offering a more objective and less biased approach than traditional methods. Our model’s effectiveness, particularly with SSD speakers, opens new avenues for research and clinical application in speech pathology.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of EECMS, Curtin University, Bentley, WA 6102, Australia.
- School of Allied Health, Curtin University, Bentley, WA 6102, Australia.
ORCID-Profile: Ying Li, Bryce Johannas Wohlan, Duc-Son Pham, Kit Yan Chan, Roslyn Ward, Neville Hennessey, Tele Tan
Förderung
- The Department of Health WA
- The Prompt Institute
- PROMPT Institute Research
- WA Near Miss Award
- Department of Health WA and administered through the Future Health Research and Innovation (FHRI) Fund
Zitieren
APA
Li, Y., Wohlan, B. J., Pham, D. S., Chan, K. Y., Ward, R., Hennessey, N., & Tan, T. (2023). Improving Text-Independent Forced Alignment to Support Speech-Language Pathologists with Phonetic Transcription. Sensors. https://doi.org/10.3390/s23249650
Vancouver
Li Y, Wohlan BJ, Pham DS, Chan KY, Ward R, Hennessey N, et al. Improving Text-Independent Forced Alignment to Support Speech-Language Pathologists with Phonetic Transcription. Sensors. 2023. doi:10.3390/s23249650
BibTeX
@article{Li2023,
author = {Li, Ying and Wohlan, Bryce Johannas and Pham, Duc-Son and Chan, Kit Yan and Ward, Roslyn and Hennessey, Neville and Tan, Tele},
title = {Improving Text-Independent Forced Alignment to Support Speech-Language Pathologists with Phonetic Transcription},
journal = {Sensors},
year = {2023},
doi = {10.3390/s23249650}
}
EUVIMED