The advancement of artificial intelligence generated content (AIGC) has created a pressing need for robust image watermarking that can withstand both conventional signal processing and novel semantic editing attacks. Current deep learning-based methods rely on training with hand-crafted noise simulation layers, which inherently limit their generalization to unforeseen distortions. In this work, we propose $\textbf{SimuFreeMark}$, a noise-$\underline{\text{simu}}$lation-$\underline{\text{free}}$ water$\underline{\text{mark}}$ing framework that circumvents this limitation by exploiting the inherent stability of image low-frequency components. We first systematically establish that low-frequency components exhibit significant robustness against a wide range of attacks. Building on this foundation, SimuFreeMark embeds watermarks directly into the deep feature space of the low-frequency components, leveraging a pre-trained variational autoencoder (VAE) to bind the watermark with structurally stable image representations. This design completely eliminates the need for noise simulation during training. Extensive experiments demonstrate that SimuFreeMark outperforms state-of-the-art methods across a wide range of conventional and semantic attacks, while maintaining superior visual quality.
JIGMARK: A Black-Box Approach for Enhancing Image Watermarks against Diffusion Model Edits
Northeastern University(东北大学);Virginia Tech(弗吉尼亚理工大学);Netflix(网飞公司);Eyeline Studios(Eyeline工作室);University of California, Los Angeles(加州大学洛杉矶分校);Princeton University(普林斯顿大学)
OmniGuard: Hybrid Manipulation Localization via Augmented Versatile Deep Image Watermarking
School of Electronic and Computer Engineering, Peking University(北京大学电子与计算机工程学院);Peking University Shenzhen Graduate School-Rabbitpre AIGC Joint Research Laboratory(北京大学深圳研究生院Rabbitpre AIGC联合研究实验室);School of Arts, Peking University(北京大学艺术学院)
TAG-WM: Tamper-Aware Generative Image Watermarking via Diffusion Inversion Sensitivity
Anhui Province Key Laboratory of Digital Security, University of Science and Technology of China(安徽省数字安全重点实验室,中国科学技术大学);School of Computing, National University of Singapore(computing学院,新加坡国立大学)