This stable diffusion fork can generate mosaic images

0
28


The full version of the Stable Diffusion model can generate images with artificial intelligence directly on your computer. There is now a modified version that is perfect for creating repeating patterns.

“Tileable Stable Diffusion”, developed by Thomas Moore, is a modified version of the same Stable Diffusion model that can be used to generate amazing AI art and realistic images. The main difference is that this fork (that is, software that has been separated from another code base) is designed to generate images that can be used in continuous repeating patterns.

Generating an image for "glass window with a decorative pattern"
Generating an image with Tileable Stable Diffusion
The generated pattern is repeated three times vertically.
The generated pattern is repeated three times horizontally

Perhaps the best use for this is to create textures for objects in video games, especially since the size of the grid can be customized to match the requirements of a specific project or game engine. Patterns can also be useful for device wallpapers, website backgrounds (especially if you’re going for a ’90s aesthetic), pattern fills in Photoshop, or even wallpaper for an actual room.

According to Moore, the modified model “uses a circular convolution, so the model sees the image as if all parallel edges are connected. Like a tube, but in both directions to form a torus.” The result is that each pattern has no discernible edge when repeated in a tiled pattern, as in-game textures are supposed to work.

You can test the model in your web browser to generate images, powered by Nvidia T4 graphics cards on remote servers, or you can run it locally on your PC with code from the GitHub repository.

Font: @ReplicateHQ