Communication Dans Un Congrès Année : 2024

Malacopula: adversarial automatic speaker verification attacks using a neural-based generalised Hammerstein model

Résumé

We present Malacopula, a neural-based generalised Hammerstein model designed to introduce adversarial perturbations to spoofed speech utterances so that they better deceive automatic speaker verification (ASV) systems. Using non-linear processes to modify speech utterances, Malacopula enhances the effectiveness of spoofing attacks. The model comprises parallel branches of polynomial functions followed by linear timeinvariant filters. The adversarial optimisation procedure acts to minimise the cosine distance between speaker embeddings extracted from spoofed and bona fide utterances. Experiments, performed using three recent ASV systems and the ASVspoof 2019 dataset, show that Malacopula increases vulnerabilities by a substantial margin. However, speech quality is reduced and attacks can be detected effectively under controlled conditions. The findings emphasise the need to identify new vulnerabilities and design defences to protect ASV systems from adversarial attacks in the wild.

Fichier principal
Vignette du fichier
publi-7828.pdf (1.54 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05299794 , version 1 (06-10-2025)

Licence

Identifiants

Citer

Massimiliano Todisco, Michele Panariello, Xin Wang, Héctor Delgado, Kong Aik Lee, et al.. Malacopula: adversarial automatic speaker verification attacks using a neural-based generalised Hammerstein model. ASVspoof Workshop 2024, Automatic Speaker Verification Spoofing and Countermeasures Challenge, ISCA, Aug 2024, Kos Island, Greece. pp.94-100, ⟨10.21437/asvspoof.2024-14⟩. ⟨hal-05299794⟩
232 Consultations
162 Téléchargements

Altmetric

Partager

  • More