2/11
3/11
4/11
5/11
6/11
7/11
8/11
9/11
10/11
11/11
welemHello everyone! I'm doing a bit of an experiment here. Having seen a lot of AI-generated images, I decided to try generating images from Koikatsu characters using Stable Diffusion. It's relatively simple to start with Automatic1111, and I did a quick LORA model training to get the faces more accurate.
This approach may be good to get specific camera angles, proportions, positions, and costumes (if you know how to use Koikatsu) compared to a text-to-image prompt, although you'll need to manage the denoising effect.
Koikatsu Cards:
Minami Fuyuki (冬木 美波) --> illust/116162697
Shushu Suruga (駿河 朱々) --> illust/117523214
Yumiella Dolkness (ユミエラ・ドルクネス) --> illust/116513433
Katsumi Shinohara (篠原 香津美) --> illust/116307882
Yukari Sugisawa (杉沢 由香里) --> illust/115946369
Pai Kawahigashi (河東 ぱい) --> illust/117940088
2024-04-28 19:14:02 +0000