MaXsive: High-Capacity and Robust Training-Free Generative Image Watermarking in Diffusion Models

Open in new window