Auto clothes segment detection, Inpaint mask and restore everything that was not masked after VAE DecodeThe extended version comes with multiple refiners,A variant of my<a target="_blank" rel="ugc" href="https://civarchive.com/models/557147/sd3-xl-supir-upscale-workflow"> SD3 XL SUPIR Upscale Workflow</a>(SDXL Supir, SD3 face details, SD3 Eye details, SDXL breast refine, SD3 fine details upscale, then restore SDXL breast refine using mask after SD3 fine details Upscale)IpAdapter OptionControlNet Option <a target="_blank" rel="ugc" href="https://huggingface.co/xinsir/controlnet-openpose-sdxl-1.0">https://huggingface.co/xinsir/controlnet-openpose-sdxl-1.0</a>The extended part of the workflow can make things looks better but It won't fix a bad first pass diffusion. Inpaint models are not perfect, but this workflow can be a good starting point.I recommend loading images around ≈1kpx resolution,Set higher Supir Control Scale End if it creates undesired details or hallucinate0.94 Is my default value, 0.96/0.98 should be less detailed but with lower hallucinations chances

Files with SHA256: c03de1336ab0ffacf1ab7e6ed60f68b037aea9a4f35828c9108e3f6c7bc6a4a4