Breaking Watermarks in the Frequency Domain: A Modulated Diffusion Attack Framework

TL;DR

This paper introduces FMDiffWA, a frequency-domain diffusion-based framework that effectively neutralizes watermarks in images while maintaining high visual quality, advancing watermark attack techniques.

Contribution

It proposes a novel frequency-domain modulation method integrated into diffusion models, improving attack effectiveness and generalization over existing watermark attack approaches.

Findings

FMDiffWA achieves higher visual fidelity than existing watermark attacks.

The framework effectively neutralizes watermarks across diverse schemes.

Enhanced attack efficacy with preserved image quality.

Abstract

Digital image watermarking has advanced rapidly for copyright protection of generative AI, yet the comparatively limited progress in watermark attack techniques has broken the attack-defense balance and hindered further advances in the field. In this paper, we propose FMDiffWA, a frequency-domain modulated diffusion framework for watermark attacks. Specifically, we introduce a frequency-domain watermark modulation (FWM) module and incorporate it into the sampling stages both the forward and reverse diffusion processes. This mechanism enables selective modulation of watermark-related frequency components, thereby allowing FMDiffWA to effectively neutralize the invisible watermark signals while preserving the perceptual quality of the attacked watermarked images. To achieve a better trade-off between attack efficacy and visual fidelity, we reformulate the training strategy of conventional diffusion models by augmenting the canonical noise estimation objective with an auxiliary refinement constraint. Comprehensive experiments demonstrate that FMDiffWA achieves superior visual fidelity compared to existing watermark attacks, while exhibiting strong generalization across diverse watermarking schemes.