Conventional research on speech recognition modeling relies on the canonical form for most low-resource languages while automatic speech recognition (ASR) for regional dialects is treated as a fine-tuning task. To investigate the effects of dialectal variations on ASR we develop a 78-hour annotated Bengali Speech-to-Text (STT) corpus named Ben-10. Investigation from linguistic and data-driven perspectives shows that speech foundation models struggle heavily in regional dialect ASR, both in zero-shot and fine-tuned settings. We observe that all deep learning methods struggle to model speech data under dialectal variations but dialect specific model training alleviates the issue. Our dataset also serves as a out of-distribution (OOD) resource for ASR modeling under constrained resources in ASR algorithms. The dataset and code developed for this project are publicly available
@article{arxiv.2510.23252,
title = {Are ASR foundation models generalized enough to capture features of regional dialects for low-resource languages?},
author = {Tawsif Tashwar Dipto and Azmol Hossain and Rubayet Sabbir Faruque and Md. Rezuwan Hassan and Kanij Fatema and Tanmoy Shome and Ruwad Naswan and Md. Foriduzzaman Zihad and Mohaymen Ul Anam and Nazia Tasnim and Hasan Mahmud and Md Kamrul Hasan and Md. Mehedi Hasan Shawon and Farig Sadeque and Tahsin Reasat},
journal= {arXiv preprint arXiv:2510.23252},
year = {2025}
}
Comments
The manuscript has to be withdrawn to address an authorship and intellectual property clarification