I have been using stable diffusion quite a bit and have become quite adept at creating pornographic images for personal pleasure. Mostly portrait still images of a naked person stnading, but with the advent of more tools like controlnet and whatnot. I have been thinking it is very possible to create a backend system where people can submit character descriptions and using automation scripts I could make a dynamic image generation system for games. Imagine a game like EVIL, which is a qsp game where you can essentially collect wives and sexual partners, games like that could benefit greatly from this, I could make a system to automatically generate a interactable npc's images in various forms of attire, hair colour, skin colour, body types, face types. This could add significant depth to various html games as well, you could have very heavy character modification, eventually download embeddings or loras for certain clothing attires or aesthetic looks. This could make games like EVIL so much better. I can share some code with you guys but just wanted to hear about what you guys think about such a system before I start working on it. Would someone be interested in something like this ?
Stable diffusion batch character backend system
- Thread starter Deleted member 5811045
- Start date
- Apr 16, 2023
- 27
- 163
Its definitely the future. In all my experimentation, stable diffusion just cant produce good results for actual P in V images, so I eventually gave up on trying. But its already really good at making pretty girls. You can even make decent looking videos of girls dancing or whatever. Perfect if you're interested in making more of a soft core game.
I will make an application to have a system from which stable diffusion characters are created and registered to a local database. To enable the creation of recurrent still images. I think having some code for future p and v images is good, I think everyone can agree it's only a matter of time. Aslo could you please elaborate on the vids of girls dancing, I am not aware of this.
It's just a matter of time, if we're not already there at the time of this post, where stable diffusion is going to take over adult visual novel content creation entirely. Yeah, I want to know more about controlnet and how you can specify poses as well as how seeds work for model consistency. I'd like more discussion concerning stable diffusion and AI art in general by you guys. I'm just an outsider trying to learn.
Also, I'm utterly bored by DAZ models now that I've seen the potential of AI art. I can't not believe that there are like 10 or 20 secret AI projects about to be revealed here on this site soon.
Look at this
promt = "Sexy blonde with tanned skin and perfect body"
neg prompt="animation, ugly, fat, chubby, overweight, asian, black and white"
That is from a basic dazmodel, I know hands are fucked but there has to be ways to fix that fosho
Right! So DAZ at this point could be used to create scenes and models general appearance as well as positioning. And then AI could create based off of that. As opposed to relying on DAZ alone. I would imagine the time to create a VN would be drastically reduced as well as a drastic increase in overall quality.
And yeah hands can be fixed manually but it has to be done after the initial render using various AI tools. The girl in your example is only semi-realistic. Full on realistic models can be created so the potential is just scary at this point.
Last edited: