Oliver Roeger

I'm a graphic designer and instructor interested in developing new visual languages using AI.

roegerdesign.com/ai /

All results based on the same prompt and method. In this case it's an image to image generation. The original was a rather abstract image generated in Midjourney. Then the image to image generation was done in the Wombo app using a prompt describing a person. Interesting here to observe that the letters “AI” are sometimes appearing in the background and sometimes are the person itself.

Merging typographic form with architecture

More pattern and prompt combos done together with Stable Diffusion (patterns) and Midjourney (final).

Typographic experiments using Control Net in Stable Diffusion.

Combining graphic patterns and portrait prompts in Midjourney.