From Sinhala to Dhivehi: Cross-Lingual Transfer Learning for Low-Resource Speech Recognition
arXiv:2607.06289v1 Announce Type: new Abstract: Dhivehi, the national language of the Maldives, is currently under-resourced for automatic speech recognition (ASR) and other NLP tasks. This study investigates whether cross-lingual transfer learning from Sinhala, a linguistically related, relatively well-resourced Insular Indo-Aryan language, can improve Dhivehi ASR. We conduct seventeen experiments across five transfer learning paradigms: Dhivehi-only baselines, sequential fine-tuning, multilingual fine-tuning, continual pre-training, and a control using Turkish as an unrelated language. The s
The increasing maturity of cross-lingual transfer learning techniques and the growing global demand for AI applications in diverse linguistic contexts make this a timely development.
This research demonstrates a viable path for developing AI capabilities for low-resource languages, reducing digital inequality and enabling local AI development beyond major linguistic blocs.
Previously unfeasible AI applications for certain languages become potentially viable, expanding the global reach and utility of AI technologies.
- · Maldives government
- · Dhivehi language speakers
- · AI researchers in low-resource contexts
- · Companies seeking to expand AI services globally
- · Monolingual AI developers
- · Languages without related, well-resourced languages
Improved Dhivehi automatic speech recognition (ASR) systems will become possible.
Enhanced digital services and accessibility for Dhivehi speakers could emerge, fostering local innovation.
This methodology could be replicated for hundreds of other low-resource languages, driving a more linguistically diverse AI landscape.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.CL