Metraipolis › archive › dc_6dc2cd14b600
Towards Zero-Shot Attribution of Synthetic Speech via Audio-Text Contrastive Retrieval — arXiv:2609.21581
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.21581
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-10T03:12:15.477Z
- SHA-256 of the archived text
27bbcdfe368c17faa14dcd5a797471a614015f668200bfdfdfa76f8fd500a924- Stable URL of this page
https://metraipolis.com/documento/dc_6dc2cd14b600
How to cite
Towards Zero-Shot Attribution of Synthetic Speech via Audio-Text Contrastive Retrieval — arXiv:2609.21581 — arXiv (en), https://arxiv.org/abs/2609.21581 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6dc2cd14b600 (Metraipolis archive dc_6dc2cd14b600, sha256 27bbcdfe368c17faa14dcd5a797471a614015f668200bfdfdfa76f8fd500a924)
BibTeX
@misc{dc6dc2cd14b600,
title = {Towards Zero-Shot Attribution of Synthetic Speech via Audio-Text Contrastive Retrieval — arXiv:2609.21581},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.21581},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6dc2cd14b600; sha256 27bbcdfe368c17faa14dcd5a797471a614015f668200bfdfdfa76f8fd500a924},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21581. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Audio deepfake forensics is moving beyond a simple real-or-fake verdict toward attribution: which system generated the audio clip? Most source-attribution methods cast this as closed-set classification, so they cannot name a generator that was absent from training, a gap that widens with every newly released text-to-speech (TTS) system. We instead frame attribution as cross-modal retrieval: each generator is described in natural language, and a clip is attributed by retrieving the description closest to it in a shared audio-text embedding space. Adding a new system then takes nothing more than writing its description, with no retraining and no new classifier head. Our model couples a frozen Wav2Vec2-BERT audio encoder with a frozen E5 text encoder and aligns them through small trainable projection heads, using a contrastive objective that combines a cross-modal supervised-contrastive loss with intra-modal terms. We evaluate on MLAAD v9 (140 TTS models, 51 languages) under 10-fold leave-models-out cross-validation. For generators it has never encountered before, the model reaches a model-level mean reciprocal rank (MRR) of 58.4%. Even when the correct model is not identified, the audio clip is often matched to systems that share the true generator's vocoder, acoustic model, or architecture. Because the same embedding space also answers natural-language attribute queries, one set of descriptions covers both open-set attribution and attribute-level forensic profiling. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in eess.AS). Categorie arXiv: Audio and Speech Processing (eess.AS). Pagina del lavoro: https://arxiv.org/abs/2609.21581 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-10T03:01:34.413Z dalla pagina https://arxiv.org/catchup/eess.AS/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).