MaXsive: High-Capacity and Robust Training-Free Generative Image Watermarking in Diffusion Models
This addresses copyright protection and content generation issues in diffusion models, offering an incremental improvement over prior methods.
The authors tackled the vulnerability of training-free diffusion watermarking to rotation, scaling, and translation attacks by proposing MaXsive, which uses an X-shape template to enhance robustness without reducing capacity, achieving verified effectiveness on benchmarks.
The great success of the diffusion model in image synthesis led to the release of gigantic commercial models, raising the issue of copyright protection and inappropriate content generation. Training-free diffusion watermarking provides a low-cost solution for these issues. However, the prior works remain vulnerable to rotation, scaling, and translation (RST) attacks. Although some methods employ meticulously designed patterns to mitigate this issue, they often reduce watermark capacity, which can result in identity (ID) collusion. To address these problems, we propose MaXsive, a training-free diffusion model generative watermarking technique that has high capacity and robustness. MaXsive best utilizes the initial noise to watermark the diffusion model. Moreover, instead of using a meticulously repetitive ring pattern, we propose injecting the X-shape template to recover the RST distortions. This design significantly increases robustness without losing any capacity, making ID collusion less likely to happen. The effectiveness of MaXsive has been verified on two well-known watermarking benchmarks under the scenarios of verification and identification.