Bayesian Learning for Domain-Invariant Speaker Verification and Anti-Spoofing

Bayesian Learning for Domain-Invariant Speaker Verification and Anti-Spoofing

Stať ve sborníku v databázi WoS či Scopus

The performance of automatic speaker verification (ASV) and anti-spoofing drops seriously under real-world domain mismatch conditions. The relaxed instance frequency-wise normalization (RFN), which normalizes the frequency components based on the feature statistics along the time and channel axes, is a promising approach to reducing the domain dependence in the feature maps of a speaker embedding network. We advocate that the different frequencies should receive different weights and that the weights' uncertainty due to domain shift should be accounted for. To these ends, we propose leveraging variational inference to model the posterior distribution of the weights, which results in Bayesian weighted RFN (BWRFN). This approach overcomes the limitations of fixed-weight RFN, making it more effective under domain mismatch conditions. Extensive experiments on cross-dataset ASV, cross-TTS anti-spoofing, and spoofing-robust ASV show that BWRFN is significantly better than WRFN and RFN.

The performance of automatic speaker verification (ASV) and anti-spoofing drops seriously under real-world domain mismatch conditions. The relaxed instance frequency-wise normalization (RFN), which normalizes the frequency components based on the feature statistics along the time and channel axes, is a promising approach to reducing the domain dependence in the feature maps of a speaker embedding network. We advocate that the different frequencies should receive different weights and that the weights' uncertainty due to domain shift should be accounted for. To these ends, we propose leveraging variational inference to model the posterior distribution of the weights, which results in Bayesian weighted RFN (BWRFN). This approach overcomes the limitations of fixed-weight RFN, making it more effective under domain mismatch conditions. Extensive experiments on cross-dataset ASV, cross-TTS anti-spoofing, and spoofing-robust ASV show that BWRFN is significantly better than WRFN and RFN.

anti-spoofing | Bayesian learning | domain generalization | speaker verification

anti-spoofing | Bayesian learning | domain generalization | speaker verification

LI, J.; MAK, M.; ROHDIN, J.; LEE, K.; HERMANSKY, H.

2026

17.08.2025

International Speech Communication Association

Rotterdam

Proceedings of the Annual Conference of the International Speech Communication Association Interspeech

Interspeech

Nizozemsko

1123

1127

5

https://www.isca-archive.org/interspeech_2025/li25h_interspeech.pdf

Jin Li_2025h_interspeech