Reply to thread

Message: <blockquote data-quote="Jfdlsjfd" data-source="post: 9148400" data-attributes="member: 42856">It's difficult to answer generally because it depends on the training of the model. An observer that is asked to describe a picture of a man will mention if he has a beard (most of the time), but will stay silent if he has no beard, instead of saying this guy (or this woman) is beardless. Therefore beardless or glabrous will be learnt, but as a far less useful descriptor of an image and there might not be enough occurrences (in some models) to guarantee it's correctly associated with a lack of facial hair.So your wizard is bound to have a beard, because a wizard is like Merlin, Gandalf or Dumbledore, and it's difficult to make him drop the beard during the generation.Most AI image can be made better with a slight amount of touching up with either another AI tool to redraw part of the image or more classical digital graphic editors. Sometimes it's quicker to just edit the part you don't like instead of fighting the prompt-reading engine.</blockquote>