Editing Away the Evidence: Diffusion-Based Image Manipulation and the Failure Modes of Robust Watermarking

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Qi, Qian, Tang, Jiangyun, Lee, Jim, Davis, Emily, Carter, Finn
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912964822433792
author Qi, Qian
Tang, Jiangyun
Lee, Jim
Davis, Emily
Carter, Finn
author_facet Qi, Qian
Tang, Jiangyun
Lee, Jim
Davis, Emily
Carter, Finn
contents Robust invisible watermarks are widely used to support copyright protection, content provenance, and accountability by embedding hidden signals designed to survive common post-processing operations. However, diffusion-based image editing introduces a fundamentally different class of transformations: it injects noise and reconstructs images through a powerful generative prior, often altering semantic content while preserving photorealism. In this paper, we provide a unified theoretical and empirical analysis showing that non-adversarial diffusion editing can unintentionally degrade or remove robust watermarks. We model diffusion editing as a stochastic transformation that progressively contracts off-manifold perturbations, causing the low-amplitude signals used by many watermarking schemes to decay. Our analysis derives bounds on watermark signal-to-noise ratio and mutual information along diffusion trajectories, yielding conditions under which reliable recovery becomes information-theoretically impossible. We further evaluate representative watermarking systems under a range of diffusion-based editing scenarios and strengths. The results indicate that even routine semantic edits can significantly reduce watermark recoverability. Finally, we discuss the implications for content provenance and outline principles for designing watermarking approaches that remain robust under generative image editing.
format Preprint
id arxiv_https___arxiv_org_abs_2603_12949
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Editing Away the Evidence: Diffusion-Based Image Manipulation and the Failure Modes of Robust Watermarking
Qi, Qian
Tang, Jiangyun
Lee, Jim
Davis, Emily
Carter, Finn
Image and Video Processing
Cryptography and Security
Multimedia
Robust invisible watermarks are widely used to support copyright protection, content provenance, and accountability by embedding hidden signals designed to survive common post-processing operations. However, diffusion-based image editing introduces a fundamentally different class of transformations: it injects noise and reconstructs images through a powerful generative prior, often altering semantic content while preserving photorealism. In this paper, we provide a unified theoretical and empirical analysis showing that non-adversarial diffusion editing can unintentionally degrade or remove robust watermarks. We model diffusion editing as a stochastic transformation that progressively contracts off-manifold perturbations, causing the low-amplitude signals used by many watermarking schemes to decay. Our analysis derives bounds on watermark signal-to-noise ratio and mutual information along diffusion trajectories, yielding conditions under which reliable recovery becomes information-theoretically impossible. We further evaluate representative watermarking systems under a range of diffusion-based editing scenarios and strengths. The results indicate that even routine semantic edits can significantly reduce watermark recoverability. Finally, we discuss the implications for content provenance and outline principles for designing watermarking approaches that remain robust under generative image editing.
title Editing Away the Evidence: Diffusion-Based Image Manipulation and the Failure Modes of Robust Watermarking
topic Image and Video Processing
Cryptography and Security
Multimedia
url https://arxiv.org/abs/2603.12949