Improving Multilingual ASR in the Wild Using Simple N-best Re-ranking

TL;DR

This paper introduces a straightforward N-best re-ranking method that enhances multilingual ASR performance by integrating external language identification and language models, significantly reducing error rates in practical scenarios.

Contribution

It proposes a simple re-ranking approach that improves multilingual ASR accuracy by effectively utilizing external language identification and language models.

Findings

Language identification accuracy improved by up to 8.7%

Word error rates reduced by up to 3.3%

Effective across multiple acoustic models

Abstract

Multilingual Automatic Speech Recognition (ASR) models are typically evaluated in a setting where the ground-truth language of the speech utterance is known, however, this is often not the case for most practical settings. Automatic Spoken Language Identification (SLID) models are not perfect and misclassifications have a substantial impact on the final ASR accuracy. In this paper, we present a simple and effective N-best re-ranking approach to improve multilingual ASR accuracy for several prominent acoustic models by employing external features such as language models and text-based language identification models. Our results on FLEURS using the MMS and Whisper models show spoken language identification accuracy improvements of 8.7% and 6.1%, respectively and word error rates which are 3.3% and 2.0% lower on these benchmarks.