Metraipolis › archive › dc_eece87b73011
GAP-SAM: A Global Artifact Prior for Generalizable AI-Generated Image Manipulation Localization — arXiv:2608.20929
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.20929
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T10:53:05.736Z
- SHA-256 of the archived text
a328014e259328b9b35c8ec61de667294d12e7df655e336ae71a2704bcb07919- Stable URL of this page
https://metraipolis.com/documento/dc_eece87b73011
How to cite
GAP-SAM: A Global Artifact Prior for Generalizable AI-Generated Image Manipulation Localization — arXiv:2608.20929 — arXiv (en), https://arxiv.org/abs/2608.20929 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eece87b73011 (Metraipolis archive dc_eece87b73011, sha256 a328014e259328b9b35c8ec61de667294d12e7df655e336ae71a2704bcb07919)
BibTeX
@misc{dceece87b73011,
title = {GAP-SAM: A Global Artifact Prior for Generalizable AI-Generated Image Manipulation Localization — arXiv:2608.20929},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.20929},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eece87b73011; sha256 a328014e259328b9b35c8ec61de667294d12e7df655e336ae71a2704bcb07919},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.20929. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
AI-generated image manipulation localization identifies edited pixels, but its OOD performance lags behind image-level detection partly because pixel supervision entangles forensic evidence with dataset-specific mask geometry and semantic boundaries. Extending image-level distribution alignment to localization, we construct COCO-ControlNet with source-image Canny edges and depth maps to align semantics and geometry, improving OOD performance across multiple localizers. Yet tighter Mask-VAE Reconstruction Alignment (Mask-VAE) underperforms COCO-ControlNet, showing that VAE reconstruction artifacts transfer poorly to local diffusion-inpainting artifacts. We also identify \emph{boundary adhesion}, where fine-tuned segmentation models snap predictions to semantic object contours rather than true manipulation boundaries. These findings motivate GAP-SAM, which encodes an image and its frozen VAE reconstruction into a global artifact token and injects it into SAM3's feature pyramid via zero-gated FiLM before pixel decoding. Without prescribing a spatial region, this token modulates dense decoding to preserve localization while suppressing semantic-boundary shortcuts. Across six datasets, GAP-SAM averages 79.8 Pixel-F1, outperforming the strongest prior method by 12.6 points. It also performs best at every tested severity of JPEG compression, Gaussian blur, and resizing.
Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-24 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.20929 — 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 2026-10-18T10:48:13.579Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).