GaussMarker: Robust Dual-Domain Watermark for Diffusion Models

2025-06-13

AI Summary

To address the insufficient robustness of single-domain watermarking in diffusion model-based image copyright protection, this paper proposes the first spatial-frequency dual-domain collaborative watermarking framework. Methodologically: (1) a dual-domain pipelined injector synchronously embeds watermarks into both spatial and frequency domains during image generation; (2) a model-agnostic, learnable Gaussian noise restorer (GNR) is introduced to enhance watermark recovery after tampering; and (3) a cross-domain detection score fusion strategy is designed to improve discriminative reliability. Evaluated across three Stable Diffusion variants, the method achieves state-of-the-art performance under eight common image distortions and four advanced adversarial attacks. It significantly improves recall while reducing false positive rates. This work establishes a robust, generalizable paradigm for copyright protection of generative images.

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: Vertical and domain-specific search
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

As Diffusion Models (DM) generate increasingly realistic images, related issues such as copyright and misuse have become a growing concern. Watermarking is one of the promising solutions. Existing methods inject the watermark into the single-domain of initial Gaussian noise for generation, which suffers from unsatisfactory robustness. This paper presents the first dual-domain DM watermarking approach using a pipelined injector to consistently embed watermarks in both the spatial and frequency domains. To further boost robustness against certain image manipulations and advanced attacks, we introduce a model-independent learnable Gaussian Noise Restorer (GNR) to refine Gaussian noise extracted from manipulated images and enhance detection robustness by integrating the detection scores of both watermarks. GaussMarker efficiently achieves state-of-the-art performance under eight image distortions and four advanced attacks across three versions of Stable Diffusion with better recall and lower false positive rates, as preferred in real applications.

Problem

• Enhances watermark robustness in diffusion models
• Embeds watermarks in spatial and frequency domains
• Improves detection against image manipulations and attacks

Innovation

• Dual-domain watermarking in spatial and frequency domains
• Model-independent Gaussian Noise Restorer for robustness
• Pipelined injector for consistent watermark embedding