Anyone else trying to make consistent characters with AI adult image tools?
I'm finding it difficult to make consistent characters with AI adult image generators. Does anyone have a process for refining the prompts to keep the characters recognisable? Do you describe every specific detail each time or just some key features? Or, do the characters always drift significantly between generations no matter what you do? I want to create a particular aesthetic using AI models, but I am wondering if I am barking up the wrong tree.
13 Views
I’ve been experimenting in a similar way, mostly just comparing different tools side by side to see how they react to identical inputs. It’s interesting because some platforms feel almost predictable, while others seem to change behavior depending on factors you can’t really see. When I’m testing variations, this is what I use https://undress.cc/edit just to add another reference point in the mix. I don’t treat it as a main solution, more like one of several tools I rotate through when checking how outputs differ. What helped me is running multiple versions and comparing them instead of trusting the first result, because that’s where you start noticing patterns in how each system behaves.