A Speech Corpus for Mizo Automatic Speech Recognition: Whisper and SraVaani 1.0 Fine-Tuning with Morphology-Aware Evaluation
Advancing ASR performance for the Mizo language through fine-tuned whisper models and morphology-aware metrics.
This study reports the development of an Automatic Speech Recognition (ASR) system in Mizo, a low-resource language. The development included collecting 17.62 hours of speech data, curating it, and fine-tuning the Mizo ASR system with three Whisper multilingual models and with th...
ArXiv CS.CL
Source
Generated insight by OnAirToday AI
This research bridges the digital divide for low-resource languages by demonstrating how fine-tuning large-scale models can achieve high accuracy in Mizo. It proves that morphology-aware evaluation is critical for accurately measuring performance in complex linguistic structures beyond simple word error rates.
Comments
Keep comments relevant and respectful. No links, advertising, self-promotion, or spam.