Model to generate unmasking sequences like in spy movies.
Description
FAQ
Comments (4)
Very cool idea. Could you share your prompts? I know about the trigger word, but a lot of these videos are the very opposite of "unmasking". I want to see how you control them differently.
Also, I think this would be really good as I2V, so we could start with a real specific face, instead of an obvious mask. Similar to your videos showing the people putting on masks, where both look real.
hi, thanks for your comment. here an example of prompt: "She brings her hands up to her face and slowly removes a flesh like mask and a long hair wig, revealing her true identity. The mask has a neutral expression and is designed to cover her entire face. As she removes the mask, her eyes widen in surprise or recognition, and her mouth smile." If you repalce "remove" by "put on" it work on opposite way.
Ok now I'm paranoid my AI girlfriend isn't real.
I guess I need therapy now