After our earlier Minecraft-server project, Star Dream Studio announced an anime-style image-generation experiment based on Stable Diffusion. I handled this project and shared some results through our Pixiv profile.
These are AI-generated examples, rather than photographs or a claim of hand-drawn production. The original announcement described model training, but did not document the dataset, licensing, or evaluation, so this post does not make additional claims about those details.
At the time, we did not plan an open-source release. A public generation interface with a limited free allowance was something we were considering, not a service already promised or launched by this article. The portfolio is the place to see the work that was actually shared.
Adapted from the original Chinese article, published on January 20, 2025.
Comments NOTHING