The clothing in this version leans towards cheongsam and has good generalization, making it suitable for many large models. What's even more surprising is that the composition and atmosphere of the image differ depending on the different base models. You can try combining it with various cyber elements, as well as keywords related to technology-related vehicles and props for a more immersive experience.
For clothing color, you can also try pairing two colors together, such as blue/pink, black/red, blue/white, or red/green, which all make for great combinations.
As for movements, currently standing and sitting poses are more easily triggered, but it depends on whether the animation is compatible and generalized enough with the model.
The example images are all produced directly from the model：
AWportrait https://civitai.com/models/61170/awportrait,
lyriel https://civitai.com/models/22922/lyriel
Guofeng3 https://civitai.com/models/10415/3-guofeng3
majicMIX reverie https://civitai.com/models/65055/majicmix-reverie
I hope that everyone will try it out and welcome any feedback or suggestion!
