2024 : i’m sick of drawing.

This project started with digital sketches. Using image altering programs like Runawayml, Magnificia, and Midjourney I processed the drawings looking for the widest variety of outcomes. Taking the outcomes drawing overtop, running them through other programs sometimes with clean ups in photoshop .Outputs would vary from super clean to really rough. A rough outputs with a areas that would draw attention, I would process multiple times to build a strong comp around that vocal point.

Phase 2 of the project was to find a way to make the computer make images like my drawings. Using Controlnet , Stable diffusion I trained an ai. model with over a hundred drawings, (youtube is your friend) which would mimic characteristics from all the images uploaded. With the newly trained model I then created images set to the prompt and parameters given. Example: please make me a image of “dragon ball z in the style of askew” and in seconds, the program would give me the following which I could rerun an infinite amount of times with an infinite amount of variations.

Attached are some of my favorite outcomes.

Thank you Akira Toriyama