Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗
Who is Crimson? Do you recognise her? You don't. She's a creation of Stable Diffusion and is here for your use in your prompts.
Model source
Source description
Who is Crimson? Do you recognise her? You don't. She's a creation of Stable Diffusion and is here for your use in your prompts.
I recommend you use ADetailer to show it at its best.
Any buzz tips or ratings appreciated!
I've been using the LastName process for a while for my nobodies, mainly as it's very quick. However with this I went back to my own method, which involves creating a face I like, selecting a set of cropped images from a real person, then faceapping the face onto all the images, and training off that.
I'm trying to do it again, to try and be able to do it more quickly, as it's a method with some real strengths - the big one is that it just gives slightly better looking characters, and seems to be better at getting details outside the normal training set. For example, with Aiko my aim was to create a character with brightly dyed hair, and it just didn't work. However with Crimson I started with a woman with bright red hair, and I end up with a woman with bright red hair. Not as bright as the files, but still plenty bright,.
Sources
1 sourceModel artifacts
1 artifactSource excerpts
1 excerptSource context: 454 downloads · Type TextualInversion · Base model SD 1.5
The main downside (except having to manually do each faceswap) is that it tends to make the characters prettier. When most Stable Diffusion models already does that, that's a lot of glamour for a Nobody.
