End-to-end (E2E) models, which directly predict output character sequences given input speech, are good candidates for on-device speech recognition. E2E models, however, present numerous challenges: In order to be truly useful, such models must decode speech utterances in a streaming fashion, in real time; they must be robust to the long tail of use cases; they must be able to leverage user-specific context (e.g., contact lists); and above all, they must be extremely accurate. In this work, we describe our efforts at building an E2E speech recognizer using a recurrent neural network transducer. In experimental evaluations, we find that the proposed approach can outperform a conventional CTC-based model in terms of both latency and accuracy in a number of evaluation categories.
@article{arxiv.1811.06621,
title = {Streaming End-to-end Speech Recognition For Mobile Devices},
author = {Yanzhang He and Tara N. Sainath and Rohit Prabhavalkar and Ian McGraw and Raziel Alvarez and Ding Zhao and David Rybach and Anjuli Kannan and Yonghui Wu and Ruoming Pang and Qiao Liang and Deepti Bhatia and Yuan Shangguan and Bo Li and Golan Pundak and Khe Chai Sim and Tom Bagby and Shuo-yiin Chang and Kanishka Rao and Alexander Gruenstein},
journal= {arXiv preprint arXiv:1811.06621},
year = {2018}
}