AI, QR Codes & art

Exploration

This model deliver interesting results and I wanted to give a few tries. It’s primarily designed to generate creative QR codes, but it can also be used to render illusions. The model in question is a Stable Diffusion model integrated with ControlNet, a neural network structure to control diffusion models by adding extra conditions, for example an image input.

I used my avatar image (as displayed in the header) to test it out. The initial step in my process involved generating a simpler (sort of black and white) render of the picture. Subsequently, this render was used as the input for the image parameter. I deep dive into ControlNet & Stable Diffusion sounds like an interesting idea for a next project.

First step - Image input

Results