This self portrait was created using the VQGAN+CLIP text-to-image model.
With a base video for tracking, the AI was given the prompt “Painting by Piet Mondrian”
An exploration using GANs to replicate specific painters’ styles over a base 3D animation
