Arabic Little STT: Arabic Children Speech Recognition Dataset

TL;DR

This paper introduces Arabic Little STT, a new dataset of Levantine Arabic children's speech, and evaluates the performance of state-of-the-art ASR models, revealing significant challenges and the need for dedicated child speech data.

Contribution

The creation of the first Levantine Arabic child speech dataset and systematic assessment of Whisper ASR models on this data.

Findings

Whisper models perform poorly on child speech with high WER.

Significant performance gap between child and adult speech recognition.

Highlighting the need for child-specific speech datasets in Arabic.

Abstract

The performance of Artificial Intelligence (AI) systems fundamentally depends on high-quality training data. However, low-resource languages like Arabic suffer from severe data scarcity. Moreover, the absence of child-specific speech corpora is an essential gap that poses significant challenges. To address this gap, we present our created dataset, Arabic Little STT, a dataset of Levantine Arabic child speech recorded in classrooms, containing 355 utterances from 288 children (ages 6 - 13). We further conduct a systematic assessment of Whisper, a state-of-the-art automatic speech recognition (ASR) model, on this dataset and compare its performance with adult Arabic benchmarks. Our evaluation across eight Whisper variants reveals that even the best-performing model (Large_v3) struggles significantly, achieving a 0.66 word error rate (WER) on child speech, starkly contrasting with its sub 0.20 WER on adult datasets. These results align with other research on English speech. Results highlight the critical need for dedicated child speech benchmarks and inclusive training data in ASR development. Emphasizing that such data must be governed by strict ethical and privacy frameworks to protect sensitive child information. We hope that this study provides an initial step for future work on equitable speech technologies for Arabic-speaking children. We hope that our publicly available dataset enrich the children's demographic representation in ASR datasets.