Type | |
Stats | 1,436 |
Reviews | (284) |
Published | Jul 4, 2023 |
Base Model | |
Hash | AutoV2 CE6ADEF408 |
Big thanks to cyan8 for helping with the dataset and testing for this LoRA!
Based on Soejima Shigenori's art!
They're known for their work on the Persona Series and Catherine!
There were already some releases related to Soejima's style, but they were mostly based on his flat/stylized illustrations (Promotional art, cut-ins, dialogue portraits etc.), this LoRA is mostly based on his fully rendered (2.5D-/3D-ish, with proper shading etc.) images!
If you however want exactly that flat style you can find the LoRAs I mentioned here:
Remember to support them if you enjoy their art!
Weight:0.7-0.9
Triggerwords: None.
Notes:
The eyes are sadly often unclear, I tried fixing this multiple times but it's still a bit gambly. Adding looking at viewer to the prompt helps slightly. Close-ups in general tend to be more detailed and so the eyes get less scuffed, so adding tags like portrait or upper body is helpful as well.
It used to inject random reds into images a lot, while it does it way less now it can still happen.
I wish we got a Persona 1 and 2 port instead of a 3rd Persona 5 spinoff.
Feel free to share some of the pictures you create with me and post them here!