What is Hope?
Hope:Re is a free, open-source tool built to protect your artwork from being scraped, trained, or copied by generative AI without your permission.
In a nutshell, it adds invisible changes (adversarial perturbations) and a permanent hidden watermark (using blind_watermark) to your images. This shields your unique style and proves your ownership, even if the image is cropped or compressed.
How does it work?
Hope:Re integrates algorithms from The Glaze Project (Glaze, Nightshade, and Noise) to overlay an invisible protection layer.
This perturbation neutralizes img2img styles, preventing generative AI models from successfully replicating or learning your art style.
Courtesy of Haruyu Sato (opens in a new tab)
How does it work?
Hope:Re embeds an invisible signature deep within the image's frequency spectrum (powered by blind_watermark).
Unlike standard watermarks, this hidden mark is indestructible—remaining fully extractable even after cropping, compression, or screenshotting.
Courtesy of Haruyu Sato (opens in a new tab)
