Improving Rare Word Recognition with LM-aware MWER Training. (arXiv:2204.07553v2 [cs.CL] UPDATED)

Language models (LMs) significantly improve the recognition accuracy of
end-to-end (E2E) models on words rarely seen during training, when used in
either the shallow fusion or the rescoring setups. In this work, we introduce
LMs in the learning of hybrid autoregressive transducer (HAT) models in the
discriminative training framework, to mitigate the training versus inference
gap regarding the use of LMs. For the shallow fusion setup, we use LMs during
both hypotheses generation and loss computation, and the LM-aware MWER-trained …

arxiv training