"Use it or lose it"
Please don't sue me Johnson & Johnson, I'm trying to do it honor.
See below for more, but this is an EXPERIMENT and may not produce high quality results.
Trigger Words:
rogaine - main trigger word
over ground - hair on the ground. It might look like you are in a barber shop or just add something akin to shag carpeting
Hair Styles:
bald
beard
curly
dusting
fluffy
fuzzy
hairspray
hairy
messy
mustache
shaggy
straight
stringy
tufts
wavy
whispy
Hair color:
black
blonde
brunette
cream
gray
redhead
white
What is this for:
To add hair and give you control of the style
Why did I make this:
I did not see a 'style' lora for hair, head/body/other, and now I know why.
I wanted to finish a project instead of giving up. I guess it's not done yet?
The goal of the lora was HAIR, as I noticed that the granular detail of getting hair required a mountain of effort, and to defy the constant clamp of where hair is. If you wanted to make a bigfoot or just some peach fuzz, or even add it to items that hair does not belong, this lora was supposed to help.
But really, I don't like it.
It just does not work well enough in my opinion.
What a mess. I have been working on a few projects and started this as a side project, falling down the rabbit hole. This was an experiment in making a dataset and using CIVITAI's LORA tool.
I published this because CIVITAI says it is going to go away if I don't, but I'm not happy with the results at all. Maybe I can do a second pass...
Gimme your harshest feedback or sick burn or funniest quip, or advice. I've looked at a few but I'd love to hear from you.
Description
FAQ
Comments (5)
Maybe s/he's born with it, maybe it's just a lora!
I like the concept and flexibility of this. There's quite a few hairstyles/hair behaviors I haven't seen achievable in SD yet produced here. It does seem to lend itself to an overabundance of curly hair, I had to drop that in negatives often.
Haven't seen too much other impact on composition or body yet, but I'll keep trying it. Sticking to mid-weights on the lora and just weighting tokens as needed. So far it just seems weak, not bad.
Spoke too soon! Seems like 'cream' tries to impose white/light colored objects or clothing in frame, and 'messy' will sometimes put hair where it doesn't belong (mostly on clothes).
@ptdurnvcoedqegsdnw Thank you! That is inspiring. I'll revisit my images and make sure I have a better label of messy, cream, and curly and more image label cohesion.
I love the label. I made these images for this post and I felt like I was looking at a 70's hair catalogue from the salon.
@l0ud_ninja I noticed some of the hair-transference on shaggy and fluffy as well. YMMV, seems like the more volumetric the hairstyle the more likely it wanted the same hairy volume to show up in clothes or hair accessories.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.