Efficient Dialect-Aware Modeling and Conditioning for Low-Resource Taiwanese Hakka Speech Processing

Peng An-Ci, Kuan-Tang Huang, Tien-Hong Lo, Hung-Shin Lee, Hsin-Min Wang, Berlin Chen


Abstract
Taiwanese Hakka is a low-resource, endangered language that poses significant challenges for automatic speech recognition (ASR), including high dialectal variability and the presence of two distinct writing systems (Hanzi and Pinyin). Traditional ASR models often encounter difficulties in this context, as they tend to conflate essential linguistic content with dialect-specific variations across both phonological and lexical dimensions. To address these challenges, we propose a unified framework grounded in the Recurrent Neural Network Transducers (RNN-T). Central to our approach is the introduction of dialect-aware modeling strategies designed to disentangle dialectal ”style” from linguistic ”content”, which enhances the model’s capacity to learn robust and generalized representations. Additionally, the framework employs parameter-efficient prediction networks to concurrently model ASR (Hanzi and Pinyin). We demonstrate that these tasks create a powerful synergy, wherein the cross-script objective serves as a mutual regularizer to improve the primary ASR tasks. Experiments conducted on the HAT corpus reveal that our model achieves 57.00% and 40.41% relative error rate reduction on Hanzi and Pinyin ASR, respectively. To our knowledge, this is the first systematic investigation into the impact of Hakka dialectal variations on ASR and the first single model capable of jointly addressing these tasks.
Anthology ID:
2026.lrec-1.491
Volume:
Proceedings of the Fifteenth Language Resources and Evaluation Conference
Month:
May
Year:
2026
Address:
Palma de Mallorca, Spain
Editors:
Stelios Piperidis, Núria Bel, Henk van den Heuvel, Nancy Ide, Simon Krek, Antonio Toral
Venue:
LREC
SIG:
Publisher:
ELRA Language Resource Association
Note:
Pages:
6206–6215
Language:
External URL:
https://lrec.elra.info/lrec2026-main-491
DOI:
10.63317/36umc2hxdiw5
Bibkey:
Cite (ACL):
Peng An-Ci, Kuan-Tang Huang, Tien-Hong Lo, Hung-Shin Lee, Hsin-Min Wang, and Berlin Chen. 2026. Efficient Dialect-Aware Modeling and Conditioning for Low-Resource Taiwanese Hakka Speech Processing. In Proceedings of the Fifteenth Language Resources and Evaluation Conference, pages 6206–6215, Palma de Mallorca, Spain. ELRA Language Resource Association.
Cite (Informal):
Efficient Dialect-Aware Modeling and Conditioning for Low-Resource Taiwanese Hakka Speech Processing (An-Ci et al., LREC 2026)
Copy Citation: