This model generates images similar to your own artwork. Generated images follow style as well as content of input image very well. One notable difference from other models is that, in the prompt, you should describe input image, rather than what you would like to generate. Some examples:

Input Image(The street by Balthus)

Prompt: A street scene


Input Image(Joan by Eyvind Earle)

Prompt: A dreamy girl

Input image has to be square, here we added some black padding to make out image square. This black padding can also be seen in generations.


Input Image

Prompt: Luffy, One Piece the Anime Luffy