Context-Aware Transformer Transducer for Speech Recognition
This work addresses the challenge of rare word recognition in speech recognition systems, which is an incremental improvement for applications requiring high accuracy on specialized vocabularies.
The paper tackles the problem of recognizing uncommon words in end-to-end automatic speech recognition by incorporating personalized/contextual information at inference, resulting in a 24.2% improvement in word error rate over a baseline transformer transducer and 19.4% over an existing deep contextual model.
End-to-end (E2E) automatic speech recognition (ASR) systems often have difficulty recognizing uncommon words, that appear infrequently in the training data. One promising method, to improve the recognition accuracy on such rare words, is to latch onto personalized/contextual information at inference. In this work, we present a novel context-aware transformer transducer (CATT) network that improves the state-of-the-art transformer-based ASR system by taking advantage of such contextual signals. Specifically, we propose a multi-head attention-based context-biasing network, which is jointly trained with the rest of the ASR sub-networks. We explore different techniques to encode contextual data and to create the final attention context vectors. We also leverage both BLSTM and pretrained BERT based models to encode contextual data and guide the network training. Using an in-house far-field dataset, we show that CATT, using a BERT based context encoder, improves the word error rate of the baseline transformer transducer and outperforms an existing deep contextual model by 24.2% and 19.4% respectively.