Towards scalable efficient on-device ASR with transfer learning
It addresses scalable on-device ASR for low-resource languages, but the approach is incremental as it builds on existing transfer learning methods.
This study tackled the problem of improving low-resource monolingual automatic speech recognition (ASR) by investigating transfer learning, finding that RNNT-loss pretraining with MinWER fine-tuning reduces Word Error Rates by up to 42.8% compared to baselines, with out-of-domain pretraining yielding 28% higher reductions than in-domain.
Multilingual pretraining for transfer learning significantly boosts the robustness of low-resource monolingual ASR models. This study systematically investigates three main aspects: (a) the impact of transfer learning on model performance during initial training or fine-tuning, (b) the influence of transfer learning across dataset domains and languages, and (c) the effect on rare-word recognition compared to non-rare words. Our finding suggests that RNNT-loss pretraining, followed by monolingual fine-tuning with Minimum Word Error Rate (MinWER) loss, consistently reduces Word Error Rates (WER) across languages like Italian and French. WER Reductions (WERR) reach 36.2% and 42.8% compared to monolingual baselines for MLS and in-house datasets. Out-of-domain pretraining leads to 28% higher WERR than in-domain pretraining. Both rare and non-rare words benefit, with rare words showing greater improvements with out-of-domain pretraining, and non-rare words with in-domain pretraining.