Abstract
Regional accent cues can be captured under matched conditions, but it remains unclear whether they persist between read and spontaneous speech. We study RVG1, with 500 German speakers from nine regions, comparing ten speech representations on regional classification and continuous geolocation under matched conditions and speaker-independent read--spontaneous transfer. Whisper performs best under matched conditions, reaching 0.489 nine-way UAR and 148 km median geolocation error, but drops to 0.11/0.18 UAR across transfer directions and 363 km geolocation error. Self-supervised models show a similar degradation, whereas speaker embeddings are less discriminative in-domain but more robust under transfer. This contrast is consistent across classification and geolocation. Across representations, robustness is associated with how little a representation shifts between styles (style-invariance), for which crossstyle speaker retrieval is an interpretable proxy. Age, sex, sentence-overlap, and duration controls do not account for the gap, although channel characteristics contribute. These results show that strong matched-condition performance does not indicate robust regional information.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Perez-Toro, P. A., Arias-Vergara, T., Schwarz, A., Hernandez, A., Horr, A., Kristen, C., & Maier, A. (2026). Do Speech Representations Preserve Regional Accent Across Read and Spontaneous Speech? https://omanscience.com/en/articles/do-speech-representations-preserve-regional-accent-across-read-and-spontaneous-speech
MLA 9
Perez-Toro, Paula A., et al. "Do Speech Representations Preserve Regional Accent Across Read and Spontaneous Speech?" https://omanscience.com/en/articles/do-speech-representations-preserve-regional-accent-across-read-and-spontaneous-speech.
Chicago (author–date)
Perez-Toro, Paula A., Tomas Arias-Vergara, Annette Schwarz, Abner Hernandez, Andreas Horr, Cornelia Kristen, and Andreas Maier. 2026. "Do Speech Representations Preserve Regional Accent Across Read and Spontaneous Speech?" https://omanscience.com/en/articles/do-speech-representations-preserve-regional-accent-across-read-and-spontaneous-speech.
Harvard
Perez-Toro, P. A., Arias-Vergara, T., Schwarz, A., Hernandez, A., Horr, A., Kristen, C. and Maier, A. (2026) 'Do Speech Representations Preserve Regional Accent Across Read and Spontaneous Speech?', Available at: https://omanscience.com/en/articles/do-speech-representations-preserve-regional-accent-across-read-and-spontaneous-speech.
Vancouver
Perez-Toro PA, Arias-Vergara T, Schwarz A, Hernandez A, Horr A, Kristen C, et al. Do Speech Representations Preserve Regional Accent Across Read and Spontaneous Speech? https://omanscience.com/en/articles/do-speech-representations-preserve-regional-accent-across-read-and-spontaneous-speech
IEEE
P. A. Perez-Toro, T. Arias-Vergara, A. Schwarz, A. Hernandez, A. Horr, C. Kristen, and A. Maier, "Do Speech Representations Preserve Regional Accent Across Read and Spontaneous Speech?," https://omanscience.com/en/articles/do-speech-representations-preserve-regional-accent-across-read-and-spontaneous-speech.