Support my work on Patreon and Ko-Fi and get access to tutorials and exclusive models.
Rate and leave a like if you enjoyed it, and follow for new models!
LoRA trained to generate infirmary backgrounds, together with characters in different positions and poses, or just the background itself.
Trigger: infirmary
Optional useful tags: curtains, hospital bed, etc. Be creative with your prompts and post them here!
Description
FAQ
Comments (16)
you don't even know how long have i been waiting for this
haha - hope you enjoy it!
Awwwwwsome!!!
I am probably asking too much, but is it possible to also make operating room?
thanks!
I can look into it but it will depend on whether there's good enough data out there to train the model
Of all the weird fetishes I have this is the most difficult to explain...
Nurses are probably the most vanilla of fetishes tho.
At this point, I wonder why nobody proposed on Civit AI a gynec*logist studio, with the classic examination chair... it would be sooo usefull to ~really~ test all news checkpoints and LORAs !
I second that, but no need to censor.
@Taco360 @settima_ai , I tried to make one but didn't work well. there's not much good data out there.
@psoft im sure you can find some on imagefap, ehentai, deviantart, pixiv, rule34.xxx, etc.
@Taco360 as for the chair itself, is easy:
gynecologic chair
For specific imageg psoft is true: is very difficult to find because they are very confused and mixed with bondage and other stuff... For example, on rule34 there is the tag:
https://rule34.xxx/index.php?page=post&s=list&tags=gynecological_chair+
and also the related tag:
https://rule34.xxx/index.php?page=post&s=list&tags=speculum
...but, you see, finding something usable is a pain.
But I remenber this "theme" was strong in Japan, so a lot of AV movie should be easy to find, for example, search "gynecologic examination erotic" with google images...
@settima_ai @Taco360 I gathered data from several sources, including some of those. the problem is most isn't really useful to represent the concept, or are low quality, and using realistic data is usually not an option as it usually impacts style too much.
But I'll see what I can do
@psoft I suggest starting not representig the concept but simply the object.
@settima_ai based on my experience, and my tests so far, if there's not sufficient data that is more uniform, then what the training will result on is (roughly): the character's pose, and that the character is on some form of chair. to get a replica of a gyn chair I'd need many images of the similar type of chair with different characters. training SD means training what it sees more of. there's no way to select training concept or object. that comes down to proper dataset preparation
@psoft News: now there is! I've tested it and is quite good:
https://civitai.com/models/141441/woman-on-gynochair?modelVersionId=156898
Totally, but also some tools. I have tried to make a sample/specimen cup with no success...
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.