Towards Language-Universal End-to-End Speech Recognition

TL;DR

This paper introduces a universal multilingual speech recognition system using shared character sets and language-specific gating, outperforming monolingual and multi-task models across multiple languages.

Contribution

The work proposes a novel end-to-end multilingual speech recognition model with a universal character set and gating mechanism, enabling recognition of multiple languages in a single system.

Findings

Outperforms monolingual systems on the Microsoft Cortana task

Effective in code-switching scenarios

Can initialize monolingual recognizers

Abstract

Building speech recognizers in multiple languages typically involves replicating a monolingual training recipe for each language, or utilizing a multi-task learning approach where models for different languages have separate output labels but share some internal parameters. In this work, we exploit recent progress in end-to-end speech recognition to create a single multilingual speech recognition system capable of recognizing any of the languages seen in training. To do so, we propose the use of a universal character set that is shared among all languages. We also create a language-specific gating mechanism within the network that can modulate the network's internal representations in a language-specific way. We evaluate our proposed approach on the Microsoft Cortana task across three languages and show that our system outperforms both the individual monolingual systems and systems built with a multi-task learning approach. We also show that this model can be used to initialize a monolingual speech recognizer, and can be used to create a bilingual model for use in code-switching scenarios.