Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2506.09521 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918514043912192 |
|---|---|
| author | Gaznepoglu, Ünal Ege Leschanowsky, Anna Aloradi, Ahmad Singh, Prachi Tenbrinck, Daniel Habets, Emanuël A. P. Peters, Nils |
| author_facet | Gaznepoglu, Ünal Ege Leschanowsky, Anna Aloradi, Ahmad Singh, Prachi Tenbrinck, Daniel Habets, Emanuël A. P. Peters, Nils |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_09521 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | You Are What You Say: Exploiting Linguistic Content for VoicePrivacy Attacks Gaznepoglu, Ünal Ege Leschanowsky, Anna Aloradi, Ahmad Singh, Prachi Tenbrinck, Daniel Habets, Emanuël A. P. Peters, Nils Audio and Speech Processing Computation and Language 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. |
| title | You Are What You Say: Exploiting Linguistic Content for VoicePrivacy Attacks |
| topic | Audio and Speech Processing Computation and Language |
| url | https://arxiv.org/abs/2506.09521 |