Comments
Loading Dream Comments...
You must be logged in to write a comment - Log In
My only comment related to the title: the more complex image one wants to get from AI, the more problems show up on the image. A beautiful woman, landscape, animal, building, street, robot, flower: they all go well. But if you want some scene that is out of "ordinary", things get messed up every soon.
For example, I tried to make a simple picture of "somersault" (which even shouldn't be such a complex concept), and the outcomes were crazy! Except AI Vision, every other model showed torturous body twists that are pain to look at. Even Flux that is amazingly good at many things, suddenly presented me a girl with three legs.