You must log in or register to comment.
The deepfakes would contain the prompt image provided by the creator. They did not create a whole new approximation of their face as the entire pool it can pull on for that specific part is a single or group of images provided by the prompter.
yeah idk why they said that, it’s objectively wrong.