Multilingual Stutter Event Detection for English, German, and Mandarin Speech

TL;DR

This study develops a multilingual, multi-corpus stuttering detection system that effectively identifies stuttering across English, German, and Mandarin, highlighting cross-linguistic consistency.

Contribution

The paper introduces a novel multilingual detection system trained on diverse corpora, demonstrating improved generalizability and robustness across languages.

Findings

Multilingual training achieves comparable or better performance than previous systems.

The system effectively detects stuttering across three different languages.

Results support the idea of language-independent characteristics of stuttering.

Abstract

This paper presents a multi-label stuttering detection system trained on multi-corpus, multilingual data in English, German, and Mandarin.By leveraging annotated stuttering data from three languages and four corpora, the model captures language-independent characteristics of stuttering, enabling robust detection across linguistic contexts. Experimental results demonstrate that multilingual training achieves performance comparable to and, in some cases, even exceeds that of previous systems. These findings suggest that stuttering exhibits cross-linguistic consistency, which supports the development of language-agnostic detection systems. Our work demonstrates the feasibility and advantages of using multilingual data to improve generalizability and reliability in automated stuttering detection.