Merging multiple fine-tuned Whisper models reduces word error rate on dysarthric speech by 12-16% relative to standard fine-tuning, with gains on long audio and low-data settings.
Our findings demonstrate that model merg- ing, particularly selective merging across multiple trajectories, significantly improves WER compared to traditional fine-tuning methods
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
eess.AS 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Robust fine-tuning of speech recognition models via model merging: application to disordered speech
Merging multiple fine-tuned Whisper models reduces word error rate on dysarthric speech by 12-16% relative to standard fine-tuning, with gains on long audio and low-data settings.