Saliency-Aware Diffusion Reconstruction for Effective Invisible Watermark Removal

2025-04-17

AI Summary

Current robust invisible watermarking schemes pose significant challenges for efficient removal. To address this, we propose a saliency-aware diffusion reconstruction framework. Our method operates in the latent space and introduces saliency-guided adaptive noise injection and region-wise perturbation to selectively disrupt watermark structures while preserving semantic integrity of the host content. We pioneer a saliency-driven diffusion inversion reconstruction paradigm, theoretically ensuring reconstruction stability and enabling watermark-strength-adaptive noise scheduling. Extensive experiments on multiple state-of-the-art (SOTA) watermarking methods demonstrate that our approach achieves superior performance across key metrics—PSNR, SSIM, and watermark residual rate—significantly outperforming existing denoising-based watermark removal techniques and establishing a new benchmark.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Natural Language Processing: Safety and Robustness

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

As digital content becomes increasingly ubiquitous, the need for robust watermark removal techniques has grown due to the inadequacy of existing embedding techniques, which lack robustness. This paper introduces a novel Saliency-Aware Diffusion Reconstruction (SADRE) framework for watermark elimination on the web, combining adaptive noise injection, region-specific perturbations, and advanced diffusion-based reconstruction. SADRE disrupts embedded watermarks by injecting targeted noise into latent representations guided by saliency masks although preserving essential image features. A reverse diffusion process ensures high-fidelity image restoration, leveraging adaptive noise levels determined by watermark strength. Our framework is theoretically grounded with stability guarantees and achieves robust watermark removal across diverse scenarios. Empirical evaluations on state-of-the-art (SOTA) watermarking techniques demonstrate SADRE's superiority in balancing watermark disruption and image quality. SADRE sets a new benchmark for watermark elimination, offering a flexible and reliable solution for real-world web content. Code is available on~href{https://github.com/inzamamulDU/SADRE}{ extbf{https://github.com/inzamamulDU/SADRE}}.

Problem

• Robust removal of invisible watermarks from digital content
• Preserving essential image features during watermark elimination
• Balancing watermark disruption and high-fidelity image restoration

Innovation

• Saliency-aware diffusion reconstruction for watermark removal
• Adaptive noise injection guided by saliency masks
• Reverse diffusion process for high-fidelity restoration