Stable Diffusion Portraits

In 2022 I trained an AI model for Stable Diffusion using images of my face and body. I had to use images with different facial expressions and body positions for the model to have sufficiently broad enough information for the text prompts to be effective. The below images are some of the more successful outcomes, after being upscaled.

I used this as a small project to learn more about AI as I’d seen some very interesting posts online where people were using images from Stable Diffusion as animation frames to create video. I am still very intrgued by this and I’m open to exploring it further in the future. I did plan on training more models based on images of my previous artworks as a means of finding inspiration. However, because computer graphics is such a broad arena, and I didn’t find much reward in typing text prompts into an AI, I decided to focus my attention on softwares like Houdini, TouchDesigner and learning more about shader programming. This is why I didn’t pursue any more projects using Stable Diffusion.