AI [AI Art] - Show Us Your AI Skill *NO TEENS*

[FoolsBeGone]

Slideshow
Using only image to image and Realiticvison with no adetailer I wondered how many recognizable faces I could get.

View attachment 2894219

It gave me an idea for a universal Lora project where you use daz3d models in various poses to suit the lora (Adult themed if you so desire) and have the same pose with different base skin colors and breast sizes.

Then do an image to image on the appropriate skin color and breast size images.

Add some real photos of the person your trying to train and see what result pops out.

Ok I recognized

Spoiler: Identities

You don't have permission to view the spoiler content. Log in or register now.

the 3rd face is familiar but I can't place her the only one I didn't recognize was the second one.

 

[Sharinel]

Back to the 70's

View attachment 2898508

There's more than a hint of GTA 5 loading screen about this /thumbsup

 

[Sharinel]

If you haven't tried the new version of Juggernaut XL, I quite like it





Still not quite got it for 'ze bewbz' but not far off now

 

[felldude]

Ok I recognized

Spoiler: Identities

You don't have permission to view the spoiler content. Log in or register now.

the 3rd face is familiar but I can't place her the only one I didn't recognize was the second one.

So it was Kara Delvine, Elizabeth Olsen, Zendaya and I think angilina jolie I don't remember anymore lol

 

[Dagg0th]

Still trying to figure out Regional prompter extension, plus

You must be registered to see the links

&

You must be registered to see the links

Loras

 

[Artiour]

something random

 

[FoolsBeGone]

something random
View attachment 2900296

I don't think it was intentional but the top girl really reminds me of Susan Dey back when she was on the Partridge Family. God that girl was effin gorgeous!!!

 

[Artiour]

I don't think it was intentional but the top girl really reminds me of Susan Dey back when she was on the Partridge Family. God that girl was effin gorgeous!!!

hard for me to see it because she already is someone, maybe they look a bit alike, her name is Skyla Novea

 

[felldude]

I don't think it was intentional but the top girl really reminds me of Susan Dey back when she was on the Partridge Family. God that girl was effin gorgeous!!!

Using just the text prompt (Susan Dey:1.2) got me a few images with realistic vision merge....But I think its mixing adult and young images to make a face that is eh



Kinda looks generic on an image to image. Some of that could be the hairstyle not matching



Compared to a person that likely has way more training data in the checkpoint

 

[Artiour]

has anyone created anything good on civitai yet
my best result so far

I can't tell if it's good yet since my favorite models aren't available

 

[felldude]

has anyone created anything good on civitai yet
my best result so far
View attachment 2900434 View attachment 2900435 View attachment 2900436 View attachment 2900437
I can't tell if it's good yet since my favorite models aren't available

It looks a bit like it is merging anime with realistic, the face looks a bit flat but the comp is good.
I wasn't aware civitai had an option to generate.

The only online checkpoint I would be curious to look at is DreamUp but I doubt that deviant is providing it, and the month I had core I didn't try it.

 

[Artiour]

It looks a bit like it is merging anime with realistic, the face looks a bit flat but the comp is good.
I wasn't aware civitai had an option to generate.

The only online checkpoint I would be curious to look at is DreamUp but I doubt that deviant is providing it, and the month I had core I didn't try it.

civitai just provided the option to generate 3 or 4 days ago
Do you mean DreamUp from deviant art, that is the worst AI I've ever used, very limited prompt size and only sd 1.5 no extra models no LoRA no nothing
I'll give another month since I have fragment I have nothing to use them for

 

[felldude]

civitai just provided the option to generate 3 or 4 days ago
Do you mean DreamUp from deviant art, that is the worst AI I've ever used, very limited prompt size and only sd 1.5 no extra models no LoRA no nothing
I'll give another month since I have fragment I have nothing to use them for

Thank you, yeah that is a lack luster UI



I was expecting a little more from them, I'm sure if I knew nothing about local image generation it would be cool for a bit....Im glad it wasn't my first experience with AI

 

[FoolsBeGone]

I tried another AI generator today (

You must be registered to see the links

) I think the pictures came out pretty good
View attachment 863523c751721e98c5bc73c9c96eee5d_2292_4096.webp View attachment dc174cf1553d12d68efdc54fe5bc87a2_2728_4096.webp

 

[ManlyManperorOfMankind]

A quick Cortana picture i did in SD. Any tips for a newbie ? I somehow cant make inpaint and get a seamless transition with the backround

 

[FoolsBeGone]

Just some nature based pictures I did

 

[PandaRepublic]

Cecilia Lion ​

 

[felldude]

View attachment 2901342
A quick Cortana picture i did in SD. Any tips for a newbie ? I somehow cant make inpaint and get a seamless transition with the backround

Have you tried cycling through the different samplers at the same rate. Simple/Euler will paint a lot more evenly then Karras/Euler but Karras is good for skin detail

 

[Dagg0th]

Working on a new Lora

(I'll fix the hand later.. it's late here ZZzz)

 

[FoolsBeGone]

Put that burning cat-alope in the dungeon and dragons movie right next to the owlbear

Damn now I want to try and bring some D&D monsters to photoreal....

Put that burning cat-alope in the dungeon and dragons movie right next to the owlbear

Damn now I want to try and bring some D&D monsters to photoreal....

You mean like this cute lil guy?