New Software Stable Attribution Can Attribute AI Art to Artists Work
Earlier this year we reported that the makers of Stable Diffusion and Midjourney are being sued for stealing artist's work. Now a new tool is available to further help artists in their battle with the developers behind text-to-image AI software.
The tool, named Stable Attribution, can identify which images were used to train the AI software. This can enable artists to be attributed for their work so they can refuse their work from being used in further training models or collaborate with the developers on ethical terms.
The makers of Stable Attribution say, Artists rely on attribution of their work for recognition and income, and A.I. models need human created images to function. But the training data for many popular A.I. image generators was scraped from the web, in ways the creators didn't intend or consent to. In the process, attribution to the original creators was lost."
Stable Attribution Can Help Artists Earn From AI ArtThey argue that, By crediting artists and sharing revenue with creators according to attribution, A.I. driven apps could let artists passively earn income every time a model generates an image. By highlighting the artists who influenced the A.I. generated images, anyone can easily discover new creators whose work they love, just by typing what they want to see."
The tool works by decoding an AI created image and comparing it with images from which the model was trained with. The training model of Stable Diffusion is publicly available. This means that after the model is indexed most original images can be found.
The tool is not perfect but the makers, Chroma, are working on improving it. Also, the images in the training set do not include attribution to the artist that created the image, only the image. If you want to find the artist behind the art, you need to do your own research.
Stable Attribution want to make sure all artists are accredited for their art. If you find out who created the image you can share a link with them to get them credit.
This tool will be happily welcomed by artists who are concerned about how text-to-image AI could hurt their income. It could also help artists who want to make sure their art is not being used to train AI models and lawyers fight cases against AI Art developers.
The post New Software Stable Attribution Can Attribute AI Art to Artists Work appeared first on The Tech Report.