You can do using img2img method based on top of SD. Another option is to create an embedding of a concept based on textual inversion and then use the embedding

You can do using img2img method based on top of SD. Another option is to create an embedding of a concept based on textual inversion and then use the embedding to guide SD generation. Both methods are possible using this: https://github.com/hlky/stable-diffusion-webui https://github.com/hlky/stable-diffusion-webui