analiza aiqbench
This paper addresses a critical gap in Arabic Automatic Speech Recognition (ASR) by introducing AraYoungVoices, a specialized corpus covering children and adolescents across native and second-language contexts. The study demonstrates that current models struggle significantly with younger L2 speakers and provides empirical evidence that age-specific fine-tuning strategies outperform joint training for demographic balance. By focusing on the unique challenges of non-adult speech patterns, this work directly advances the inclusivity and robustness of Arabic language AI systems.
AraYoungVoices: A Diverse L1/L2 Corpus of Arabic Child and Adolescent Speech
arXiv:2610.05044v1 Announce Type: cross Abstract: State-of-the-art ASR systems primarily target native adult speech, leading to substantial performance gaps for children, adolescents, and L2 speakers. We introduce AraYoungVoices, a 151.72-hour Arabic read-speech corpus from 286 speakers aged 7--18, comprising AraKids (7--12) and AraTeens (13--18). The corpus includes 146 native Arabic (L1) and 140 second-language (L2) speakers, with native speakers spanning Egyptian, Gulf, Levantine, and North African dialectal backgrounds and L2 speakers representing diverse linguistic backgrounds across the Americas, Asia, Africa, and Europe. We benchmark four pretrained ASR models under zero-shot and fine-tuned settings using unseen-speaker-$\&$-unseen-prompt (USUP) and unseen-speaker-$\&$-seen-prompt (USSP) evaluations. Results show that L2 speech remains substantially more challenging than L1 speech, with the largest errors observed mainly for younger L2 speakers. Age-specific fine-tuning improves the matched age group, while joint fine-tuning provides a stronger balance across populations. ASR hypotheses are also consistently closer to the standard reading prompt than to the verbatim transcription, particularly for L2 speech, suggesting partial normalization of reading deviations.
Abstract:State-of-the-art ASR systems primarily target native adult speech, leading to substantial performance gaps for children, adolescents, and L2 speakers. We introduce AraYoungVoices, a 151.72-hour Arabic read-speech corpus from 286 speakers aged 7--18, comprising AraKids (7--12) and AraTeens (13--18). The corpus includes 146 native Arabic (L1) and 140 second-language (L2) speakers, with native speakers spanning Egyptian, Gulf, Levantine, and North African dialectal backgrounds and L2 speakers representing diverse linguistic backgrounds across the Americas, Asia, Africa, and Europe. We benchmark four pretrained ASR models under zero-shot and fine-tuned settings using unseen-speaker-$\&$-unseen-prompt (USUP) and unseen-speaker-$\&$-seen-prompt (USSP) evaluations. Results show that L2 speech remains substantially more challenging than L1 speech, with the largest errors observed mainly for younger L2 speakers. Age-specific fine-tuning improves the matched age group, while joint fine-tuning provides a stronger balance across populations. ASR hypotheses are also consistently closer to the standard reading prompt than to the verbatim transcription, particularly for L2 speech, suggesting partial normalization of reading deviations.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Sound (cs.SD)
arXiv:2610.05044 [cs.CL]
(or arXiv:2610.05044v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2610.05044 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 258 cuvinte
Subiecte conexe
Nuha-Speech: Building General-Purpose Arabic Speech-LLMs
arXiv · 11 sept. 2026
Gumbel-BEARD: Automatic Layer Selection for Self-Supervised Adaptation of Whisper in Low-Resource Domains
arXiv · 11 iun. 2026
From Exposure to Expectation: Frequency, Surprisal, and Language Across Development in Spanish
arXiv · 25 aug. 2026
BRICS Draws a New Global Line: Full Text of New Delhi Declaration
slguardian.org · 12 sept. 2026