Metraipolis › archive › dc_cb9fc1bbe47f

Language Orthogonalization for Zero-Shot Cross-Lingual Audio Deepfake Detection — arXiv:2609.16458

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.16458
License (as archived)
CC0-1.0
Archived
2026-10-15T19:23:26.967Z
SHA-256 of the archived text
02b823be5e1f74c9b5464dbe0b24e4e8b6ebef23dd8003664283b7b8530f19cf
Stable URL of this page
https://metraipolis.com/documento/dc_cb9fc1bbe47f

How to cite

Language Orthogonalization for Zero-Shot Cross-Lingual Audio Deepfake Detection — arXiv:2609.16458 — arXiv (en), https://arxiv.org/abs/2609.16458 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_cb9fc1bbe47f (Metraipolis archive dc_cb9fc1bbe47f, sha256 02b823be5e1f74c9b5464dbe0b24e4e8b6ebef23dd8003664283b7b8530f19cf)
BibTeX
@misc{dccb9fc1bbe47f,
  title = {Language Orthogonalization for Zero-Shot Cross-Lingual Audio Deepfake Detection — arXiv:2609.16458},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.16458},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_cb9fc1bbe47f; sha256 02b823be5e1f74c9b5464dbe0b24e4e8b6ebef23dd8003664283b7b8530f19cf},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.16458. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Audio deepfake detectors need to transfer to languages absent from training, as multilingual speech synthesis outpaces labeled anti-spoofing resources. While detectors increasingly rely on self-supervised speech models (S3Ms), these backbones encode language-dependent structure that confounds spoof cues. We address this confound through language orthogonalization, a target-free ridge map that removes S3M variation projected onto continuous language-identification (LID) embeddings. Across six languages, six S3M backbones, and all Leave-N-Out settings, it consistently reduces EER across unseen languages. Cross-lingual EER correlates with LID-space distance, where orthogonalization yields larger gains for more distant transfers.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-16 in cs.CL).
Categorie arXiv: Audio and Speech Processing (eess.AS) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.16458 — 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 2026-10-15T18:32:32.900Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-16?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).