Members will soon have access to IP-Mix, an IP-Adapter/Controller/LoRA tool for using your images as one-image LoRAs to create new looks that are a mix of both, plus guidance.
Here, the settings are creating a new face from a very simple prompt, with no controlnets enabled...
While here, we've turned on a depth map, and are using the center of image 1 and the attention area of image 2 to create a re-composed image that contains portions of the look of the inputs...
And finally, here's an example using the same settings, but enabling a second controlnet (anime lineart) in addition to the depth map:
As always, many options are switch controlled for ease of operation and choice of options. As with all of AegisFlow suite, these switches are always yellow to make them simple to identify. (notes are universally dark blue; workflow-to-workflow image connectors are bright blue, etc).
I can also be linked to other modules and accepts input from other workflows. Here, I've inserted the MaskFX template to do some postprocessing.
Interested in helping decide the future of AegisFlow? Interested in supporting my continuing work to become a professional Stable Diffusion suite? Join us at http://majorstudio.gumroad.com to join Major Studio, get AegisFlow DIY (main workflows only, not templates and supplementals) or the paid installer version with a growing library of modules (easy to install, updated by me) , or support our efforts with a small monthly donation ($1 and $5 tiers).
Thanks!