Multilingual DistilWhisper: Efficient Distillation of Multi-task Speech Models via Language-Specific Experts

arXiv:2311.01070v3 Announce Type: replace
Abstract: Whisper is a multitask and multilingual speech model covering 99 languages. It yields commendable automatic speech recognition (ASR) results in a subset of its covered languages, but the model still underperforms on a non-negligible number of under-represented languages, a problem exacerbated in smaller model versions. In this work, we propose DistilWhisper, an approach able to bridge the performance gap in ASR for these languages while retaining the advantages of multitask and multilingual capabilities. Our approach …

abstract arxiv asr automatic speech recognition cs.cl cs.sd distillation eess.as experts language languages multilingual recognition results speech speech recognition type via whisper