Metraipolis › archive › dc_d22d4558bbaf
Semantic Watermarking for Malicious Image Manipulation Detection — arXiv:2609.39623
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.39623
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T14:46:17.018Z
- SHA-256 of the archived text
c9e3c30b3cb57a0d9e649f324992180ef06e1d4f648684b4c8dcca3dfc5e37c6- Stable URL of this page
https://metraipolis.com/documento/dc_d22d4558bbaf
How to cite
Semantic Watermarking for Malicious Image Manipulation Detection — arXiv:2609.39623 — arXiv (en), https://arxiv.org/abs/2609.39623 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d22d4558bbaf (Metraipolis archive dc_d22d4558bbaf, sha256 c9e3c30b3cb57a0d9e649f324992180ef06e1d4f648684b4c8dcca3dfc5e37c6)
BibTeX
@misc{dcd22d4558bbaf,
title = {Semantic Watermarking for Malicious Image Manipulation Detection — arXiv:2609.39623},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.39623},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d22d4558bbaf; sha256 c9e3c30b3cb57a0d9e649f324992180ef06e1d4f648684b4c8dcca3dfc5e37c6},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39623. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
The proliferation of high-fidelity generative editing models has made it possible to inject violent or sexual content into otherwise ordinary images while preserving visual plausibility, with concrete consequences for public discourse and vulnerable populations. We propose a robust semantic watermarking framework that reframes the watermark as a recoverable semantic reference rather than an opaque identifier. Our framework combines a $\beta$-VAE-based binary watermark (CLIP-VAE) with explicit channel-aware training---random bit-flip noise is injected during training so that the decoder learns graceful degradation under the noisy watermarking channel. As a downstream application, a lightweight module SDA-Net uses the recovered semantic embedding to expose not only whether but in which semantic direction an image has been altered. In a 5-way comparison against representative binary hashing baselines (SimHash, ITQ, HashNet, and their robust-MLP variants), CLIP-VAE achieves the highest reconstruction cosine similarity to the original CLIP embedding under realistic InstructPix2Pix bit-error rates, and uniquely supports direction-of-drift detection---a forensic complement to existing content-moderation pipelines. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-01 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2609.39623 — 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 2026-10-07T14:19:00.777Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).