Metraipolis › archive › dc_f66e2228f643
Harnessing CLIP and DINO: An Uncertainty-Aware Cascaded Fusion Network for Generalizable Deepfake Image Detection — arXiv:2609.07670
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.07670
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T15:11:56.358Z
- SHA-256 of the archived text
78478e0ac6ce86c96bd623c120324eca748319b8075f623c960cc48cf470e318- Stable URL of this page
https://metraipolis.com/documento/dc_f66e2228f643
How to cite
Harnessing CLIP and DINO: An Uncertainty-Aware Cascaded Fusion Network for Generalizable Deepfake Image Detection — arXiv:2609.07670 — arXiv (en), https://arxiv.org/abs/2609.07670 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f66e2228f643 (Metraipolis archive dc_f66e2228f643, sha256 78478e0ac6ce86c96bd623c120324eca748319b8075f623c960cc48cf470e318)
BibTeX
@misc{dcf66e2228f643,
title = {Harnessing CLIP and DINO: An Uncertainty-Aware Cascaded Fusion Network for Generalizable Deepfake Image Detection — arXiv:2609.07670},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.07670},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f66e2228f643; sha256 78478e0ac6ce86c96bd623c120324eca748319b8075f623c960cc48cf470e318},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.07670. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
The growing realism and accessibility of manipulated and generated faces threaten the trustworthiness of digital media. To detect such forgeries, deepfake detectors based on vision foundation models have shown promising performance, but they typically rely on a single pretrained representation and are prone to overfitting to particular training distributions. To improve generalization to unseen forgeries, we propose UCF-Net, an uncertainty-aware cascaded fusion network that harnesses CLIP's language-aligned semantic priors and DINO's self-supervised visual-structure priors. UCF-Net extracts hierarchical features across Transformer depths, uses layer-wise expert aggregation to adaptively combine each encoder's multi-level cues, and performs weighted fusion of the resulting representations based on entropy-derived uncertainty. We further consolidate public deepfake datasets into a unified benchmark of approximately 4M images and construct a separate cross-generator evaluation set with over 8K face images from eight recent generators. On the unified benchmark, UCF-Net achieves the best mean AUC among the evaluated methods in both in-domain and cross-domain evaluations. On the cross-generator set, it adapts effectively with limited target-domain data, although zero-shot transfer remains challenging. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-09 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.07670 — 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 2026-10-16T14:49:57.243Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).