Milovana.com

Your first displaced fingers tease is there!

Blonde Snakes & Ladders - AI Generated Edition

https://milovana.com/webteases/showtease.php?id=62679

"Maid Annie will take care of your house and you. All arts are generated by AI just for purpose of this tease. Roll the dice and climb the ladder. Multiple endings. Vanilla content."

I read about concepts of creating webtease with just AI generated art and it is there! I think this is the first tease like this on this website - prove me if I'm wrong.

Yes, some pictures has minor artifacts, but I think overall it turned out pretty good. It contains around 100 pictures at takes 5-20 minutes to complete. I hope you guys will have fun with this one.
This is my first webtease so please let me know about anything I should be aware of.

Love
Hevelius

That was a lot of fun, particularly since it was all AI generated. I'd love to know the model and prompts you used and if it was txt2img or img2img. I've been collecting images to try to put something together but it's pretty time consuming to weed out the garbage pictures to get the few good ones.

Great job as a concept tease. Can't wait to see more!

This is gold! What version of stable diffusion did you use, and what prompts?

I'm amazed at the quality of the artwork generated, and I like your taste.
I need to learn more about this stable diffusion thing.

I haven't played all the way through yet, but I think the game could use a little more instruction to make it more fun, but overall I love what you've done here! Even just basic instructions like a number of strokes to do, or x fast strokes, or x slow strokes etc would add a lot to it for me.

Great stuff!

PG

Hello

I see there is some interest in the topic so I will try to explain briefly how I achieved this. The topic is a bit complex and if you are not a technical person it could be tedious to do.
Disclaimer: I am not using any external website, everything is running on my local PC, so if you have low-tier laptop it will be time and resource consuming.

First step is to download stable diffusion web-ui. Many people use this one, so you will be able to find tutorials how to open and use it. I know the easiest way to play around is when you have Nvidia GPU, but in case of different manufacturers there are possible workarounds. https://github.com/AUTOMATIC1111/stable-diffusion-webui

Then you need a model, for this I recommend website https://civitai.com . You download it and paste to models/Stable-diffusion folder in stable-diffusion-webui. I was using different ones and results may vary. It took me few weeks to tweak and learn parameters and still I feel like a newbie, but this is a good starting point for any of you.

Then you just start webui-user.bat or .sh and you should be ready to go. The application is running on local ip, you need to copy it to the browser when you see the ip address (with port) in the console.

Tip: playing with negative prompts is as much important as with positive ones! You need to specify both positive and negative. I was using text-2img, but you can also use GUI to try playing around with different methods.

Here I share with you some examples, but you can also find similar ones in civit.ai models.

Prompts
positive 1:

Spoiler: show

exposed breasts, (photorealism:1.1), wearing robe, innocent, ((high angle POV:1.4)), smiling, scattering, detailed skin, incredibly detailed, cinematic, super realistic,, ((full body portrait)), kneeling, extremely detailed CG unity 8k wallpaper), (wet skin), (hyperdetailed), a portrait of a pole, undercut hair, blonde, maid, freckles, apron, amazing body, pronounced feminine feature, legwear suspenders, trending on ArtStation, trending on CGSociety

positive 2:

Spoiler: show

nude, most beautiful woman, undercut hair, blonde, lingerie, ((nsfw)), spread legs, (orgasm), (yawn), ((holding bra)), (hands on breasts), (photorealism:1.1), epic, (low angle POV:1.4), (look at the camera:1.7), deailed skin, super realistic, ((full body portrait)), extremely detailed CG unity 8k wallpaper, a portrait of a pole, , freckles, amazing body, pronounced feminine feature

positive 3:

Spoiler: show

((extremely detailed CG unity 8k wallpaper)), ((sexual pose)), ((ass)), (undercut hair), blonde, (bending), super realistic, 8k, (hyperrealism), detailed skin, living room, small breasts, (oiled skin), curvy, nsfw, nude, a portrait of a pole, freckles, pronounced feminine feature

negative 1:

Spoiler: show

deformed, bad anatomy, disfigured, poorly drawn face, mutation, mutated, extra limb, ugly, disgusting, poorly drawn hands, missing limb, floating limbs, disconnected limbs, malformed hands, blurry, ((((mutated hands and fingers)))), watermark, watermarked, oversaturated, censored, distorted hands, amputation, missing hands, obese, doubled face, double hands, asian, b&w, black and white, sepia

negative 2:

Spoiler: show

(censored), cartoon, 3d, ((disfigured)), ((bad art)), ((deformed)),((extra limbs)),((close up)),((b&w)), wierd colors, blurry, (((duplicate))), ((morbid)), ((mutilated)), [out of frame], extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), cloned face, (((disfigured))), out of frame, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck))), Photoshop, video game, ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, mutation, mutated, extra limbs, extra legs, extra arms, disfigured, deformed, cross-eye, body out of frame, blurry, bad art, bad anatomy, 3d render

Love,
Hevelius

Wow, I'm amazed at how good this art is! Game is simple, but effective! We should train our own text generation network for Milovana, that generates text for teases! Combine with such image generators and bam! Infinite webteases!

0088_ wrote: Mon Feb 13, 2023 8:04 pm Wow, I'm amazed at how good this art is! Game is simple, but effective! We should train our own text generation network for Milovana, that generates text for teases! Combine with such image generators and bam! Infinite webteases!

Wouldn't that be awesome! Give it a scenario and let it rip... hmmmmmm

I don't have a computer with a GPU (only an older laptop) so probably can't run stable diffusion, but definitely going to check it out.

PG

This is super impressive. Great job getting some really good looking images! Can't wait to see more like this

i tinkered around a bit with AI a few months ago and didnt see a useful tease coming out of that work at that time so i waited for technology to advance. i will give civitai a try, maybe i get it done this time. thanks anyway for the pioneer work