Improving Low-resource ASR Using Bilingual Fine-tuning with Language Identification: A Cross-linguistic Evaluation

Reihaneh Amooie, Yun Hao, Wietse de Vries, Jelske Dijkstra, Matt Coler, Martijn Wieling


Abstract
This study explores how bilingual fine-tuning affects automatic speech recognition (ASR) in low-resource languages. We evaluate this method across nine linguistically and geographically diverse language pairs, covering a range of language families and writing systems. To distinguish the two languages, during training, we pre-pend each input text with a language identification token. At inference, the model jointly predicts both the language and transcription from the speech input alone. As texts for which the language is incorrectly determined show low ASR performance, we also conduct a follow-up experiment in which the language identification token is provided both during training and inference. Our results show that bilingual fine-tuning can be beneficial when language identification accuracy is high, and that in cases where language identification performance is low, including the language identification token at inference helps to improve ASR performance.
Anthology ID:
2026.speakable-1.15
Volume:
Proceedings of Speech Language Models in Low-Resource Settings: Performance, Evaluation, and Bias Analysis (SPEAKABLE) @ LREC 2026
Month:
May
Year:
2026
Address:
Palma, Mallorca (Spain)
Editors:
Nina Hosseini-Kivanani, Alessio Brutti, Marco Matassoni, Sandipana Dowerah, Davide Liga, Christoph Schommer
Venues:
SPEAKABLE | WS
SIG:
Publisher:
ELRA Language Resources Association (ELRA)
Note:
Pages:
132–138
Language:
External URL:
https://lrec.elra.info/lrec2026-ws-speakable-15
DOI:
10.63317/39snpbvopu74
Bibkey:
Cite (ACL):
Reihaneh Amooie, Yun Hao, Wietse de Vries, Jelske Dijkstra, Matt Coler, and Martijn Wieling. 2026. Improving Low-resource ASR Using Bilingual Fine-tuning with Language Identification: A Cross-linguistic Evaluation. In Proceedings of Speech Language Models in Low-Resource Settings: Performance, Evaluation, and Bias Analysis (SPEAKABLE) @ LREC 2026, pages 132–138, Palma, Mallorca (Spain). ELRA Language Resources Association (ELRA).
Cite (Informal):
Improving Low-resource ASR Using Bilingual Fine-tuning with Language Identification: A Cross-linguistic Evaluation (Amooie et al., SPEAKABLE 2026)
Copy Citation: