This is a simple img2img for portraits and other images.What you can tweak :<ul><li>Inputs ; Steps. batch, seed, and so on...</li><li>Tweak the K Sampler</li><li>Play around with the weight</li><li>Do something with the background mask?</li><li>Add a face detailer or a face swap at the end of generation</li></ul>You will need to install properly : <a target="_blank" rel="ugc" href="https://github.com/cubiq/ComfyUI_IPAdapter_plus">https://github.com/cubiq/ComfyUI_IPAdapter_plus</a>You also need to have the proper CLIP VISION bin file for SDXL otherwise this will result in errors at generation.You will need the SDXL version of OpenPose and Depth which can be found here or elsewhere : <a target="_blank" rel="ugc" href="https://github.com/lllyasviel/stable-diffusion-webui-forge/wiki/ControlNet-model-download">https://github.com/lllyasviel/stable-diffusion-webui-forge/wiki/ControlNet-model-download</a>

Simple SDXL img2img workflow

<ul><li>Working as intended</li></ul>

Files with SHA256: 1c388f22763ad396d4643c7a77c96dcbde234e0f6c347df522ef264a9d9c7f17