Phonetic and Graphemic Systems for Multi-Genre Broadcast Transcription

TL;DR

This study compares phonetic and graphemic acoustic models for English broadcast transcription, showing that combining both models consistently improves performance, especially with advanced system combination techniques.

Contribution

It demonstrates that deep-learning based acoustic models can effectively use graphemic representations, and combining phonetic and graphemic models enhances transcription accuracy.

Findings

Long-span temporal context reduces performance gap

System combination yields consistent improvements

Complex combination methods further minimize differences

Abstract

State-of-the-art English automatic speech recognition systems typically use phonetic rather than graphemic lexicons. Graphemic systems are known to perform less well for English as the mapping from the written form to the spoken form is complicated. However, in recent years the representational power of deep-learning based acoustic models has improved, raising interest in graphemic acoustic models for English, due to the simplicity of generating the lexicon. In this paper, phonetic and graphemic models are compared for an English Multi-Genre Broadcast transcription task. A range of acoustic models based on lattice-free MMI training are constructed using phonetic and graphemic lexicons. For this task, it is found that having a long-span temporal history reduces the difference in performance between the two forms of models. In addition, system combination is examined, using parameter smoothing and hypothesis combination. As the combination approaches become more complicated the difference between the phonetic and graphemic systems further decreases. Finally, for all configurations examined the combination of phonetic and graphemic systems yields consistent gains.