Metraipolis › archive › dc_4df096fcac87

VeriSpoof: Beyond Real or Fake with One Audio LLM — arXiv:2611.04634

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.04634
License (as archived)
CC0-1.0
Archived
2026-11-04T18:52:41.509Z
SHA-256 of the archived text
9b3ff2cdc1d2c3c45aad5b984c7fcaff149d766ce6f7f90b0aebeec9ddfefc58
Stable URL of this page
https://metraipolis.com/documento/dc_4df096fcac87

How to cite

VeriSpoof: Beyond Real or Fake with One Audio LLM — arXiv:2611.04634 — arXiv (en), https://arxiv.org/abs/2611.04634 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4df096fcac87 (Metraipolis archive dc_4df096fcac87, sha256 9b3ff2cdc1d2c3c45aad5b984c7fcaff149d766ce6f7f90b0aebeec9ddfefc58)
BibTeX
@misc{dc4df096fcac87,
  title = {VeriSpoof: Beyond Real or Fake with One Audio LLM — arXiv:2611.04634},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04634},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4df096fcac87; sha256 9b3ff2cdc1d2c3c45aad5b984c7fcaff149d766ce6f7f90b0aebeec9ddfefc58},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04634. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A speech deepfake detector estimates whether a recording is synthetic, but a forensic analyst also asks which system produced it, whether recordings share a source, and what evidence supports each conclusion. We present VeriSpoof, an audio language model that unifies five tasks with shared weights: deepfake detection, generator attribution (generator, acoustic model and vocoder), recording-condition analysis, synthesized-span localisation, and source verification. Each decision is supported by grounded rationales whose statements can be independently verified. Answers are scored by sentence likelihood, allowing detection and source verification to be thresholded without separate heads. Rationale statements can be checked against waveform measurements, span labels, or generator metadata. Counterfactual examples train decisions to follow the stated evidence. Across the 14 Speech-DF-Arena test sets, VeriSpoof-Plus achieves 2.14% pooled EER, ranking third of 28 systems and first among open-weight systems. VeriSpoof-Base, trained only on official training splits of eight corpora, achieves 4.43% EER, less than half that of the next open-weight detector. With the same weights, VeriSpoof-Plus verifies whether recordings from generators unseen in training share a source at 6.2% EER. Controlled studies show that the audio representation governs detection, the LLM governs recording comparison, and source supervision is necessary for neural-codec fakes. Statement-checkable rationales catch partial fakes missed by one-step answers (99.6% vs. 78.4% on PartialEdit) and determine decisions under edits unseen in training (99.5%). Training updates only 29.7M of 2.85B parameters, fitting a single 24 GB GPU with gradient checkpointing. We release both models, inference code, a demo, and VeriSpoof-Bench, the evaluation suite covering these questions.

Tema degli esploratori: fiducia digitale e crittografia (elenco arXiv cs.CR, nuovi lavori del giorno).
Categorie arXiv: Audio and Speech Processing (eess.AS) ; Cryptography and Security (cs.CR); Sound (cs.SD).
Pagina del lavoro: https://arxiv.org/abs/2611.04634 — 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 2026-11-04T18:35:07.141Z dalla pagina https://arxiv.org/list/cs.CR/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).