Photoshop has introduced a new AI feature called Generative Fill. This next image was NOT created on a mobile device. I thought I’d show you anyway, because I’m sure it will eventually become part of mobile imaging.
I started with the above image, except without the added stylization of DeepArtEffects. I added the additional background on the right, then the cactus, then the sleeping coyote. The cactus was disappointing, because it picked up too much of the cartoonish look of the original, and the lighting was much brighter than this result.
MidJourney has some troubles with creating portraits using the user’s face, even with a seed image. You can train Stsble Diffusion, but not MidJourney or BlueWillow. But there’s a new app on Discord called Insight Fsce Swap, which allows you to swap your choice of face into created portraits. I swapped my face into a created image of a bartender, then got the watercolor style using BeCasso.