A hybrid HMM system using self-supervised features and extra text/audio data beats fine-tuned Whisper by 32 percent relative WER on Scottish Gaelic broadcast speech.
We thank BBC Alba for providing the data and Cailean Gordan, Al- ison Diack and Fearchar MacIllFhinnein for transcribing train- ing and testing data
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
A Practitioner's Guide to Building ASR Models for Low-Resource Languages: A Case Study on Scottish Gaelic
A hybrid HMM system using self-supervised features and extra text/audio data beats fine-tuned Whisper by 32 percent relative WER on Scottish Gaelic broadcast speech.