Hi,
I am creating a furry art comic and I need the character in the image in different poses but same characteristics (face/body etc) the exact same. Is there any way to do this because if I copy the prompts and seed from the image and paste it, it doesn’t even generate the same image, most things change.
If anyone can help, i’d appreciate it.
You must log in or register to comment.
That is very very hard. ways:
1: Yes even the same seed and prompt has variation, tho sometimes you can alter a word in the prompt and get the same character in a different scene.
2: You can take the image in to an ai movie service and make movies of the character in that image doing different things then turn them back in to images.
3: The most achievable method, according to me, is what the covergirl of Beautiful People is. She is a mix of two popular characters and a profession. Since the ai knows the characters, it mixes them relatively similarly each time. These are all the BP covergirl and you can see they are fairly similar. I could gen a bunch and have it look like 1 person. Yet she is unique. This is the path I suggest.
beautifulpeople covergirl
It is even more obvious with Taylor Swift pics or any famous person or character. The ai can gen the same person in diff poses. So to have this benefit for a unique character, all you have to do is mix known characters and keep the recipe to yourself.
taylorswift
