Bilingual fine-tuning with language ID tokens improves low-resource ASR when identification accuracy is high, and providing the token at inference helps when accuracy is low.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Improving low-resource ASR using bilingual fine-tuning with language identification: a cross-linguistic evaluation
Bilingual fine-tuning with language ID tokens improves low-resource ASR when identification accuracy is high, and providing the token at inference helps when accuracy is low.