I have a picture of a superhero I made using AI and my client loves it, however I've never been able to recreate the same character in different situations, evertime I feed AI the image and a prompt, the superhero changes completely and is unrecognizable. Can your software overcome this hurdle?