Beyond Orthography: Automatic Recovery of Short Vowels and Dialectal Sounds in Arabic

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Kheir, Yassine El, Mubarak, Hamdy, Ali, Ahmed, Chowdhury, Shammur Absar
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911977417211904
author Kheir, Yassine El
Mubarak, Hamdy
Ali, Ahmed
Chowdhury, Shammur Absar
author_facet Kheir, Yassine El
Mubarak, Hamdy
Ali, Ahmed
Chowdhury, Shammur Absar
contents This paper presents a novel Dialectal Sound and Vowelization Recovery framework, designed to recognize borrowed and dialectal sounds within phonologically diverse and dialect-rich languages, that extends beyond its standard orthographic sound sets. The proposed framework utilized a quantized sequence of input with(out) continuous pretrained self-supervised representation. We show the efficacy of the pipeline using limited data for Arabic, a dialect-rich language containing more than 22 major dialects. Phonetically correct transcribed speech resources for dialectal Arabic are scarce. Therefore, we introduce ArabVoice15, a first-of-its-kind, curated test set featuring 5 hours of dialectal speech across 15 Arab countries, with phonetically accurate transcriptions, including borrowed and dialect-specific sounds. We described in detail the annotation guideline along with the analysis of the dialectal confusion pairs. Our extensive evaluation includes both subjective -- human perception tests and objective measures. Our empirical results, reported with three test sets, show that with only one and half hours of training data, our model improve character error rate by ~ 7\% in ArabVoice15 compared to the baseline.
format Preprint
id arxiv_https___arxiv_org_abs_2408_02430
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Beyond Orthography: Automatic Recovery of Short Vowels and Dialectal Sounds in Arabic
Kheir, Yassine El
Mubarak, Hamdy
Ali, Ahmed
Chowdhury, Shammur Absar
Audio and Speech Processing
This paper presents a novel Dialectal Sound and Vowelization Recovery framework, designed to recognize borrowed and dialectal sounds within phonologically diverse and dialect-rich languages, that extends beyond its standard orthographic sound sets. The proposed framework utilized a quantized sequence of input with(out) continuous pretrained self-supervised representation. We show the efficacy of the pipeline using limited data for Arabic, a dialect-rich language containing more than 22 major dialects. Phonetically correct transcribed speech resources for dialectal Arabic are scarce. Therefore, we introduce ArabVoice15, a first-of-its-kind, curated test set featuring 5 hours of dialectal speech across 15 Arab countries, with phonetically accurate transcriptions, including borrowed and dialect-specific sounds. We described in detail the annotation guideline along with the analysis of the dialectal confusion pairs. Our extensive evaluation includes both subjective -- human perception tests and objective measures. Our empirical results, reported with three test sets, show that with only one and half hours of training data, our model improve character error rate by ~ 7\% in ArabVoice15 compared to the baseline.
title Beyond Orthography: Automatic Recovery of Short Vowels and Dialectal Sounds in Arabic
topic Audio and Speech Processing
url https://arxiv.org/abs/2408.02430