Breaking Watermarks in the Frequency Domain: A Modulated Diffusion Attack Framework
For researchers and practitioners in digital watermarking and generative AI security, this work addresses the imbalance between watermarking and attack techniques, though it is incremental as it applies diffusion models to a known problem.
The paper proposes FMDiffWA, a frequency-domain modulated diffusion framework for attacking digital watermarks, achieving superior visual fidelity and strong generalization across diverse watermarking schemes.
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.