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

arXiv cs.CL / 3/31/2026

💬 OpinionIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper proposes a multi-label stutter event detection system trained on annotated data from English, German, and Mandarin using four corpora.
By learning from multilingual, multi-corpus examples, the model aims to capture language-independent characteristics of stuttering for more robust cross-linguistic performance.
Experiments show multilingual training reaches performance comparable to earlier approaches and can outperform them in some cases.
The authors interpret results as evidence that stuttering has cross-linguistic consistency, supporting the feasibility of language-agnostic automated detection.
Overall, the study demonstrates that leveraging multilingual data can improve generalizability and reliability for stuttering detection systems.

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.