A tone-conditioned curriculum framework with gated adapters achieves 28.41% average WER on six Southern Bantu languages, showing architecture-specific performance differences between W2V-BERT and Whisper.
Scaling speech technology to 1,000+ languages
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CL 2years
2026 2representative citing papers
citing papers explorer
-
Tone-Conditioned Curriculum Learning for Low-Resource Bantu Speech Recognition
A tone-conditioned curriculum framework with gated adapters achieves 28.41% average WER on six Southern Bantu languages, showing architecture-specific performance differences between W2V-BERT and Whisper.