Lattice Rescoring Strategies for Long Short Term Memory Language Models in Speech Recognition

TL;DR

This paper evaluates lattice rescoring algorithms using LSTM language models to improve speech recognition accuracy, demonstrating an 8% relative reduction in word error rate on a YouTube speech dataset.

Contribution

It introduces new variants of lattice rescoring algorithms and evaluates their effectiveness with LSTM LMs in speech recognition.

Findings

LSTM LMs outperform N-gram LMs in speech recognition.

Lattice rescoring with LSTMLMs reduces WER by 8%.

New rescoring variants show promising results.

Abstract

Recurrent neural network (RNN) language models (LMs) and Long Short Term Memory (LSTM) LMs, a variant of RNN LMs, have been shown to outperform traditional N-gram LMs on speech recognition tasks. However, these models are computationally more expensive than N-gram LMs for decoding, and thus, challenging to integrate into speech recognizers. Recent research has proposed the use of lattice-rescoring algorithms using RNNLMs and LSTMLMs as an efficient strategy to integrate these models into a speech recognition system. In this paper, we evaluate existing lattice rescoring algorithms along with new variants on a YouTube speech recognition task. Lattice rescoring using LSTMLMs reduces the word error rate (WER) for this task by 8\% relative to the WER obtained using an N-gram LM.