Improving Massively Multilingual ASR With Auxiliary CTC Objectives

TL;DR

This paper enhances multilingual ASR performance across 102 languages by conditioning models on language identity using auxiliary CTC objectives, leading to state-of-the-art results on the FLEURS benchmark.

Contribution

It introduces a novel technique of conditioning on language identity with auxiliary CTC tasks to improve multilingual speech recognition accuracy.

Findings

Achieved a 28.4% relative CER reduction on FLEURS.

Demonstrated effectiveness of auxiliary CTC objectives over standard models.

Provided reproducible recipes and models for further research.

Abstract

Multilingual Automatic Speech Recognition (ASR) models have extended the usability of speech technologies to a wide variety of languages. With how many languages these models have to handle, however, a key to understanding their imbalanced performance across different languages is to examine if the model actually knows which language it should transcribe. In this paper, we introduce our work on improving performance on FLEURS, a 102-language open ASR benchmark, by conditioning the entire model on language identity (LID). We investigate techniques inspired from recent Connectionist Temporal Classification (CTC) studies to help the model handle the large number of languages, conditioning on the LID predictions of auxiliary tasks. Our experimental results demonstrate the effectiveness of our technique over standard CTC/Attention-based hybrid models. Furthermore, our state-of-the-art systems using self-supervised models with the Conformer architecture improve over the results of prior work on FLEURS by a relative 28.4% CER. Trained models and reproducible recipes are available at https://github.com/espnet/espnet/tree/master/egs2/fleurs/asr1 .