2024 INTERSPEECH INTERSPEECH 2024

Ethnolinguistic Identification of Vietnamese-German Heritage Speech

Abstract

We examined whether listeners can identify the heritage of speakers based on their voices. In Experiment 1, native German listeners with and without an Asian heritage listened to sentences that had either been produced by native German speakers with a Vietnamese heritage or by Germans without any Asian heritage. Asian heritage listeners performed more accurately when speakers matched their heritage background, but they could identify German speakers with no Asian heritage with above chance accuracy. Listeners with no Asian heritage performed more accurately when the speakers had no Vietnamese background than when they did, but they could not identify speakers with a Vietnamese heritage with above chance accuracy. In Experiment 2, the pattern for German listeners with no Asian heritage was replicated with monotonized stimuli. Thus, matching heritage background facilitates ethnolinguistic identification for Vietnamese-German heritage speech, which persists even when pitch is absent.

🧭 Keyword Pioneer — ethnolinguistic identity
🐝 Cross-Pollinator — Artificial Intelligence, Computer Science, Computer Vision, Data Science & Analytics, Deep Learning, Healthcare & Medicine, Interdisciplinary, Knowledge & Reasoning, Machine Learning, Mathematics & Optimization, Natural Language Processing, Reinforcement Learning, Security & Privacy, Speech & Audio