Stonelax again,

This is a ControlNet (especially openpose) + IPAdapter + Img2img style transfer workflow.   Since openpose is only achieved through InstantX's ControlNet Union, and IPA is only achieved through Xlab, it is a workflow that inlcudes solutions from both parties.  Things should ge much easier once one of them support both features.

It consists of 3 main sections:

1. ControlNet Union Pro for character pose control. (you can swap it out for Canny, Depth or other control types if you want)
2. IPAdapter for style transfer. (To be honest, the current IPAdapter isn’t very powerful yet, at least not for style transfer.)
3. Img2Img to further enhance style transfer effect, (it does a good job to ensure that the lighting and color tones of the image are relatively consistent.)

Unfortunately similar to my other Flux style transfer workflow( https://openart.ai/workflows/stonelax/100-flux-native-controlnet-ipadapter-style-transfer/u55OSJqYClitcKGAqQ8B ), the current Flux IPA is not very powerful. So you’ll need to adjust the parameters according to your specific needs:

1. ControlNet Strength: I tend to keep it around 0.2 - 0.3.  The Controlnet Union Pro works differently than Xlab's.  For Openpose, anything above 0.3 seems to control the image so strongly that you can hardly change its style.  Going below 0.15 however leads to a different image composition.
2. IPA Strength: Doesn’t need to be very high. In theory, the higher the IPA strength, the more style transfer you get, but in practice, it affects image quality and the transfer effect isn’t great. So, I suggest keeping it below 0.5.
3. Img2Img weight:  Img2Img contributes more to overall image quality and the similarity in lighting and color compared to IPA. I set it around 0.1. If you go up to 0.2, the image composition might begin shifting outta control.
4. Prompt: This is also crucial. Since you can’t set IPA and Img2Img weights too high (doing so would increase noise and reduce detail), matching the right prompt, especially a style prompt, can effectively help produce better images.