The model takes in a photo (which uses as main image and pose/control image) and generates a new image based on the prompts provided while maintaining the person’s identity and specified control nets. The model is made for a specific use case which is to change the style / vibe of portrait-close-up-medium-shot type of images as fast as possible while keeping most of the image structure as it is.
ahmdyassr
/
instant-id-controlnet-lcm
Instant ID with ControlNet and LCM