This paper describes the BUT submitted systems for the ASVspoof 5 challenge, along with analyses. For the conventional deepfake detection task, we use ResNet18 and self-supervised models for the closed and open conditions, respectively. In addition, we analyze and visualize different combinations of speaker information and spoofing information as label schemes for training. For spoofing-robust automatic speaker verification (SASV), we introduce effective priors and propose using logistic regression to jointly train affine transformations of the countermeasure scores and the automatic speaker verification scores in such a way that the SASV LLR is optimized.
@article{arxiv.2408.11152,
title = {BUT Systems and Analyses for the ASVspoof 5 Challenge},
author = {Johan Rohdin and Lin Zhang and Oldřich Plchot and Vojtěch Staněk and David Mihola and Junyi Peng and Themos Stafylakis and Dmitriy Beveraki and Anna Silnova and Jan Brukner and Lukáš Burget},
journal= {arXiv preprint arXiv:2408.11152},
year = {2024}
}