You Are What You Say: Exploiting Linguistic Content For Voiceprivacy Attacks

Abstract

Speaker anonymization systems hide the identity of speakers while preserving other information such as linguistic content and emotions. To evaluate their privacy benefits, attacks in the form of automatic speaker verification (ASV) systems are employed. In this study, we assess the impact of intra-speaker linguistic content similarity in the attacker training and evaluation datasets, by adapting BERT, a language model, as an ASV system. On the VoicePrivacy Attacker Challenge datasets, our method achieves a mean equal error rate (EER) of 35%, with certain speakers attaining EERs as low as 2%, based solely on the textual content of their utterances. Our explainability study reveals that the system decisions are linked to semantically similar keywords within utterances, stemming from how LibriSpeech is curated. Our study suggests reworking the VoicePrivacy datasets to ensure a fair and unbiased evaluation and challenge the reliance on global EER for privacy evaluations.

Stats

• citations2
• S2 citations—
• github stars0
• HF likes0
• heat score3.58
• arxiv keygaznepoglu2025you

Related papers

• Ghostvec: A New Threat To Speaker Privacy Of End-to-end Speech Recognition System (2023)0.00
• Attacking Voice Anonymization Systems With Augmented Feature And Speaker Identity Difference (2024)6.34
• Asynchronous Voice Anonymization Using Adversarial Perturbation On Speaker Embedding (2024)7.16
• Privacy-utility Balanced Voice De-identification Using Adversarial Examples (2022)0.00
• Language-independent Speaker Anonymization Approach Using Self-supervised Pre-trained Models (2022)9.92
• Impact Of Phonetics On Speaker Identity In Adversarial Voice Attack (2025)0.00
• A Speech Representation Anonymization Framework Via Selective Noise Perturbation (2022)6.34
• Adversarial Speech For Voice Privacy Protection From Personalized Speech Generation (2024)8.09