Tag
This study develops an Automatic Speech Recognition system for Mizo, a low-resource language, by fine-tuning Whisper and SraVaani 1.0 models, achieving a morphology-aware WER of 7.22% with Whisper-large-v3.