So I trained Stable Diffusion with my own images, many where from the JARG works from all over the years, so that's what I tried to generate first.
Many attempts failed, but there are some like this one where I am impressed at how it is able to capture the character's look and proportions. It's funny how it recognizes the usual distress expression and blushing, or even a shy pose which I wasn´t able to get before.
I'll do a bigger training with more of my images to see how it goes.
UPDATE: I updated this one with Spanish dialogues just becuse it was something that I wanted to do back then but never did. XD
Meanbeard
2023-03-29 00:52:24 +0000 UTCAndrew Whelan
2023-03-28 16:21:40 +0000 UTC