Metraipolis › archive › dc_79dfd5f78d14

Cross-modal Translation via Conditional Latent Denoising for Video Deepfake Detection — arXiv:2609.33394

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.33394
License (as archived): CC0-1.0
Archived: 2026-10-07T20:56:45.006Z
SHA-256 of the archived text: 86a60708d325771cfade0649ab0476695b1928b7c0164e8864b436968913eabf
Stable URL of this page: https://metraipolis.com/documento/dc_79dfd5f78d14

How to cite

Cross-modal Translation via Conditional Latent Denoising for Video Deepfake Detection — arXiv:2609.33394 — arXiv (en), https://arxiv.org/abs/2609.33394 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_79dfd5f78d14 (Metraipolis archive dc_79dfd5f78d14, sha256 86a60708d325771cfade0649ab0476695b1928b7c0164e8864b436968913eabf)

BibTeX

@misc{dc79dfd5f78d14,
  title = {Cross-modal Translation via Conditional Latent Denoising for Video Deepfake Detection — arXiv:2609.33394},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.33394},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_79dfd5f78d14; sha256 86a60708d325771cfade0649ab0476695b1928b7c0164e8864b436968913eabf},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.33394. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

The growing threat of video deepfakes necessitates multimodal detection. Beyond serving as independent indicators of authenticity, audio and visual signals have intrinsic dependencies that also provide an essential criterion for detection. Previous methods often overlook the cross-modal correspondences, hindering information transfer between domains and leaving crucial detection cues unexplored. To address this challenge, we propose a framework called Cross-modal Translation via Conditional Latent Denoising (CTCLD) for video deepfake detection. It connects the distinct distributions of heterogeneous modalities in latent spaces, enabling smooth cross-domain information transfer to improve detection performance. We first establish a Bayesian foundation by decomposing the audio-visual joint distribution. Subsequently, CTCLD translates both modalities via bidirectional latent denoising conditioned on each other, effectively capturing subtle inconsistencies in the manipulated signals. Experimental results demonstrate that the proposed CTCLD enables comprehensive domain alignment, resulting in a robust video deepfake detection approach with competitive performance.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.33394 — 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 2026-10-07T20:29:01.239Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).