Photorealistic AI Generated Images

Zen10101

Well-Known Member
Aug 20, 2008
866
1,173
Retro models

00004-883928964.png00003-1600637577.png00002-2592160957.png
 

Zen10101

Well-Known Member
Aug 20, 2008
866
1,173
AI vs real (Hong Kong actress Sophia Yip). Very pleased with the results, all from just a sample of 5 images. What a time to be alive. Let the fear of consequences come later.

sy.jpg
 

_2old_

Panties, Panties, Panties!
Jun 29, 2010
428
448
Back to my favourite subject:
 

Attachments

  • 18110126.jpeg
    18110126.jpeg
    156.3 KB · Views: 14
  • 18110129.jpeg
    18110129.jpeg
    148.8 KB · Views: 14
  • 18110127.jpeg
    18110127.jpeg
    116.5 KB · Views: 15
  • 18110124.jpeg
    18110124.jpeg
    159.3 KB · Views: 14
  • 18110128.jpeg
    18110128.jpeg
    158.3 KB · Views: 15
  • Love
Reactions: jcjc and Casshern2

_2old_

Panties, Panties, Panties!
Jun 29, 2010
428
448
Project Lovepop is still alive and kicking!

New Checkpoint, cleaned and improved Lora: Maid.jpg

Still a work in progress and still far too addictive ;)

(Disclaimer: This model is in no way intended to replicate anyone specific).
 
Last edited:

zombierambo

高鼻子外国人
Feb 8, 2022
64
42
Update on Project Lovepop for anyone who may be interested. I've started merging my lora into my own tweaked SD model. The results so far are promising. No prompts were needed to create these images. Still a work in progress, still too damned addictive..
How does generating them without prompts work?
 

gregorsamsa

Well-Known Member
Apr 6, 2024
356
252
Update on Project Lovepop for anyone who may be interested. I've started merging my lora into my own tweaked SD model. The results so far are promising. No prompts were needed to create these images. Still a work in progress, still too damned addictive..
_
View attachment 3584013View attachment 3584015View attachment 3584014View attachment 3584016
The rightmost one is like a perfect mix of Asian and Caucasian (Japanese with Latin, maybe French and even some Slavic influences)
 

_2old_

Panties, Panties, Panties!
Jun 29, 2010
428
448
Still somewhat less than 'photoreal', and hands are still an issue for AI,
plp-p1.jpeg
How does generating them without prompts work?
Basically a render with no prompts appears to generate an image based solely on the data used to create\\train the Model. I'm using a modified SD Pony Model which I have embedded my lora into. Using a blank prompt gives me an idea of how successful the merge has been.

Whether this works with online generation or not, I don't know, but it definitely works with EasyDiffusion offline\local.

The examples below: DuchaitenPonyReal, FantasmPonyXL & PLP_10_v2, (named after their respective SD Models), were all generated without a prompt to get a feel for what the Models unadulterated output will look like.
 

Attachments

  • DuchaitenPonyReal.jpeg
    DuchaitenPonyReal.jpeg
    28 KB · Views: 15
  • FantasmPonyXL.jpeg
    FantasmPonyXL.jpeg
    38.4 KB · Views: 14
  • PLP_10_v2.jpeg
    PLP_10_v2.jpeg
    24.7 KB · Views: 18
Last edited:

_2old_

Panties, Panties, Panties!
Jun 29, 2010
428
448
The rightmost one is like a perfect mix of Asian and Caucasian (Japanese with Latin, maybe French and even some Slavic influences)
This model is mainly based on Jav\Idol, but there are a few Eastern European influences involved along the way ;)
 
  • Like
Reactions: gregorsamsa