TG-ASR: Translation-Guided Learning with Parallel Gated Cross Attention for Low-Resource Automatic Speech Recognition

TL;DR

This paper introduces TG-ASR, a translation-guided learning framework with a novel cross-attention mechanism, to improve low-resource Taiwanese Hokkien speech recognition by leveraging multilingual translation embeddings.

Contribution

It proposes the PGCA mechanism for integrating auxiliary language embeddings into ASR, and provides a new Taiwanese Hokkien speech corpus for low-resource language research.

Findings

Achieved 14.77% relative reduction in character error rate.

Demonstrated effective cross-linguistic semantic guidance.

Identified optimal auxiliary languages for ASR enhancement.

Abstract

Low-resource automatic speech recognition (ASR) continues to pose significant challenges, primarily due to the limited availability of transcribed data for numerous languages. While a wealth of spoken content is accessible in television dramas and online videos, Taiwanese Hokkien exemplifies this issue, with transcriptions often being scarce and the majority of available subtitles provided only in Mandarin. To address this deficiency, we introduce TG-ASR for Taiwanese Hokkien drama speech recognition, a translation-guided ASR framework that utilizes multilingual translation embeddings to enhance recognition performance in low-resource environments. The framework is centered around the parallel gated cross-attention (PGCA) mechanism, which adaptively integrates embeddings from various auxiliary languages into the ASR decoder. This mechanism facilitates robust cross-linguistic semantic guidance while ensuring stable optimization and minimizing interference between languages. To support ongoing research initiatives, we present YT-THDC, a 30-hour corpus of Taiwanese Hokkien drama speech with aligned Mandarin subtitles and manually verified Taiwanese Hokkien transcriptions. Comprehensive experiments and analyses identify the auxiliary languages that most effectively enhance ASR performance, achieving a 14.77% relative reduction in character error rate and demonstrating the efficacy of translation-guided learning for underrepresented languages in practical applications.