Neural language models (NLM) have been shown to outperform conventional n-gram language models by a substantial margin in Automatic Speech Recognition (ASR) and other tasks. There are, however, a number of challenges that need to be addressed for an NLM to be used in a practical large-scale ASR system. In this paper, we present solutions to some of the challenges, including training NLM from heterogenous corpora, limiting latency impact and handling personalized bias in the second-pass rescorer. Overall, we show that we can achieve a 6.2% relative WER reduction using neural LM in a second-pass n-best rescoring framework with a minimal increase in latency.
@article{arxiv.1907.01677,
title = {Scalable Multi Corpora Neural Language Models for ASR},
author = {Anirudh Raju and Denis Filimonov and Gautam Tiwari and Guitang Lan and Ariya Rastrow},
journal= {arXiv preprint arXiv:1907.01677},
year = {2019}
}