Direct articulatory observation reveals phoneme recognition performance characteristics of a self-supervised speech model
JASA Express Letters · 2024
DOI: 10.1121/10.0034430 ↗Kurz gefasst
Fachartikel in der Zeitschrift JASA Express Letters, 2024. Verfasst von Xuan Shi und 8 weiteren Autor:innen. Beteiligte Einrichtung: Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, California 90089 USA. Hauptthemen laut MEDLINE: Phonetics. Die Arbeit nennt 25 Literaturangaben. Gefördert von: National Science Foundation.
Abstract (Originalsprache)
Variability in speech pronunciation is widely observed across different linguistic backgrounds, which impacts modern automatic speech recognition performance. Here, we evaluate the performance of a self-supervised speech model in phoneme recognition using direct articulatory evidence. Findings indicate significant differences in phoneme recognition, especially in front vowels, between American English and Indian English speakers. To gain a deeper understanding of these differences, we conduct real-time MRI-based articulatory analysis, revealing distinct velar region patterns during the production of specific front vowels. This underscores the need to deepen the scientific understanding of self-supervised speech model variances to advance robust and inclusive speech technology.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, California 90089 USA.
- Department of Linguistics, University of Southern California, Los Angeles, California 90089 USAxuanshi@usc.edu, tiantiaf@usc.edu, kevinyhu@usc.edu, skadiri@usc.edu, jihwan@usc.edu, yijinglu@usc.edu, yubinzha@usc.edu, lo…
ORCID-Profile: Xuan Shi, Tiantian Feng
Förderung
- National Science Foundation
Zitieren
APA
Shi, X., Feng, T., Huang, K., Kadiri, S. R., Lee, J., Lu, Y., Zhang, Y., Goldstein, L., & Narayanan, S. (2024). Direct articulatory observation reveals phoneme recognition performance characteristics of a self-supervised speech model. JASA Express Letters. https://doi.org/10.1121/10.0034430
Vancouver
Shi X, Feng T, Huang K, Kadiri SR, Lee J, Lu Y, et al. Direct articulatory observation reveals phoneme recognition performance characteristics of a self-supervised speech model. JASA Express Letters. 2024. doi:10.1121/10.0034430
BibTeX
@article{Shi2024,
author = {Shi, Xuan and Feng, Tiantian and Huang, Kevin and Kadiri, Sudarsana Reddy and Lee, Jihwan and Lu, Yijing and Zhang, Yubin and Goldstein, Louis and Narayanan, Shrikanth},
title = {Direct articulatory observation reveals phoneme recognition performance characteristics of a self-supervised speech model},
journal = {JASA Express Letters},
year = {2024},
doi = {10.1121/10.0034430}
}
EUVIMED