FDDWAN: A Frequency-Decoupled Diffusion Network for Watermarking Attack

2026-07-30

AI Summary

Existing methods for invisible watermark removal struggle to effectively suppress watermarks while preserving visual image quality. This work proposes a coarse-to-fine frequency-decoupled diffusion network that innovatively integrates wavelet-based frequency decomposition with diffusion modeling. Specifically, it first applies targeted attack strategies to high- and low-frequency subbands in the wavelet domain, then employs a frequency-aware residual diffusion module to selectively refine only the watermark-related residuals. Evaluated on CelebA and ImageNet, the proposed method consistently outperforms both conventional and learning-based attack approaches across four mainstream watermarking schemes, achieving a superior trade-off between watermark removal efficacy and content fidelity.

Abstract

Existing invisible watermark removal methods often struggle to accurately capture the watermark-bearing features, leading to an unfavorable trade-off between watermark suppression and perceptual fidelity. In this paper, we propose the Frequency-Decoupled Diffusion Watermark Attack Network (FDDWAN), a coarse-to-fine framework that performs watermark removal through wavelet-domain decomposition and residual diffusion refinement. In the initial stage, the Wavelet-based Frequency-domain Preliminary Attack Module (WFPAM) decomposes the watermarked image into low- and high-frequency subbands and applies frequency-specific attack strategies tailored to their respective contributions to watermark robustness and perceptual quality. In the next stage, the Frequency-domain Residual Diffusion Attack Module (FRDAM) separately models the residual distributions between the preliminarily attacked outputs and the corresponding watermark-free references during training. Rather than reconstructing the entire image, FRDAM selectively refines frequency-domain residuals, directing the diffusion process toward the remaining watermark related discrepancies while minimizing modifications to image content. Extensive experiments on CelebA and ImageNet across four representative watermarking schemes demonstrate that FDDWAN achieves a more favorable trade-off between watermark removal effectiveness and visual fidelity than conventional and learning-based attack methods.

Problem

• watermark removal
• perceptual fidelity
• invisible watermarking
• frequency-domain features
• watermark suppression

Innovation

• frequency-decoupled
• diffusion model
• watermark removal
• wavelet decomposition
• residual refinement