I may be mistaken but it look more like someone used a pretrained image generator and retrained it with child pics.
The following site for exemple teach you to build and train a image generator you just have to change the kind of pictures it is trained on to make it malevolent. https://www.assemblyai.com/blog/minimagen-build-your-own-imagen-text-to-image-model/
You’re right, it does work that way - it’s why ‘a photo of an astronaut riding a horse’ is the standard demo for SD, to show that it can create things it wasn’t trained on by remixing and extrapolating elements. Even without that though, it can do things like turn a cartoon image into a realistic one (or vise versa) with img2img without necessarily needing to know what the content is at all.
Also, it’s possible to recursively train models - create a rough model, use its output as training data for a more refined model, rinse and repeat. I’ve found it works well for getting a strong and consistent face LoRA, but I imagine the same method could be used to create any sort of model without using real photos.
Initial training picture may come from abused child’s. And we can theorize that they will need more to keep training the AI.
deleted by creator
I may be mistaken but it look more like someone used a pretrained image generator and retrained it with child pics. The following site for exemple teach you to build and train a image generator you just have to change the kind of pictures it is trained on to make it malevolent. https://www.assemblyai.com/blog/minimagen-build-your-own-imagen-text-to-image-model/
deleted by creator
You’re right, it does work that way - it’s why ‘a photo of an astronaut riding a horse’ is the standard demo for SD, to show that it can create things it wasn’t trained on by remixing and extrapolating elements. Even without that though, it can do things like turn a cartoon image into a realistic one (or vise versa) with img2img without necessarily needing to know what the content is at all.
Also, it’s possible to recursively train models - create a rough model, use its output as training data for a more refined model, rinse and repeat. I’ve found it works well for getting a strong and consistent face LoRA, but I imagine the same method could be used to create any sort of model without using real photos.
Thanks, learnt a thing.
deleted by creator
deleted by creator