Two-Pass End-to-End Speech Recognition

Tara N. Sainath, Ruoming Pang, David Rybach, Yanzhang He, Rohit Prabhavalkar, Wei Li, Mirkó Visontai, Qiao Liang, Trevor Strohman, Yonghui Wu, Ian McGraw, Chung-Cheng Chiu


The requirements for many applications of state-of-the-art speech recognition systems include not only low word error rate (WER) but also low latency. Specifically, for many use-cases, the system must be able to decode utterances in a streaming fashion and faster than real-time. Recently, a streaming recurrent neural network transducer (RNN-T) end-to-end (E2E) model has shown to be a good candidate for on-device speech recognition, with improved WER and latency metrics compared to conventional on-device models [1]. However, this model still lags behind a large state-of-the-art conventional model in quality [2]. On the other hand, a non-streaming E2E Listen, Attend and Spell (LAS) model has shown comparable quality to large conventional models [3]. This work aims to bring the quality of an E2E streaming model closer to that of a conventional system by incorporating a LAS network as a second-pass component, while still abiding by latency constraints. Our proposed two-pass model achieves a 17%–22% relative reduction in WER compared to RNN-T alone and increases latency by a small fraction over RNN-T.


 DOI: 10.21437/Interspeech.2019-1341

Cite as: Sainath, T.N., Pang, R., Rybach, D., He, Y., Prabhavalkar, R., Li, W., Visontai, M., Liang, Q., Strohman, T., Wu, Y., McGraw, I., Chiu, C. (2019) Two-Pass End-to-End Speech Recognition. Proc. Interspeech 2019, 2773-2777, DOI: 10.21437/Interspeech.2019-1341.


@inproceedings{Sainath2019,
  author={Tara N. Sainath and Ruoming Pang and David Rybach and Yanzhang He and Rohit Prabhavalkar and Wei Li and Mirkó Visontai and Qiao Liang and Trevor Strohman and Yonghui Wu and Ian McGraw and Chung-Cheng Chiu},
  title={{Two-Pass End-to-End Speech Recognition}},
  year=2019,
  booktitle={Proc. Interspeech 2019},
  pages={2773--2777},
  doi={10.21437/Interspeech.2019-1341},
  url={http://dx.doi.org/10.21437/Interspeech.2019-1341}
}