We propose Neural-FST Class Language Model (NFCLM) for end-to-end speech recognition, a novel method that combines neural network language models (NNLMs) and finite state transducers (FSTs) in a mathematically consistent framework. Our method utilizes a background NNLM which models generic background text together with a collection of domain-specific entities modeled as individual FSTs. Each output token is generated by a mixture of these components; the mixture weights are estimated with a separately trained neural decider. We show that NFCLM significantly outperforms NNLM by 15.8% relative in terms of Word Error Rate. NFCLM achieves similar performance as traditional NNLM and FST shallow fusion while being less prone to overbiasing and 12 times more compact, making it more suitable for on-device usage.
@article{arxiv.2201.11867,
title = {Neural-FST Class Language Model for End-to-End Speech Recognition},
author = {Antoine Bruguier and Duc Le and Rohit Prabhavalkar and Dangna Li and Zhe Liu and Bo Wang and Eun Chang and Fuchun Peng and Ozlem Kalinli and Michael L. Seltzer},
journal= {arXiv preprint arXiv:2201.11867},
year = {2022}
}