With this workflow for ComfyUi you can modify clothes on man and woman with different style. Input image use MaskEditor and wait for output image at full resolution.
Images used for examples:
https://civarchive.com/images/6340272
https://civarchive.com/images/6373921
https://civarchive.com/images/6389338
Work based on
Inpaint at Full Resolution
Text2img+Lora+Hires+ADx2+Upscale+Style filter
Description
Base version
FAQ
Comments (14)
there's nothing in your workflow. no nodes, no anything.
Useless workflow. The output picture is exactly the same as the input.
u need Right click -> Open in MaskEditor，then it works
@luiousnewpome398 Hmm.... marking over the clothes, right? mine still isn't doing anything.
@yangxiaolong I wanna know this work flow generate a what
I keep getting stuck on the modal loader. Some error I don't understand
Hi friends, in addition to the necessary controlnet model, I also downloaded two files when running:
1.yolox_l.torchscript.pt
2.breastsizeslideroffset.safetensors
These two models are placed in my working directory:
1.ComfyUI\custom_nodes\comfyui_controlnet_aux\ckpts\hr16\yolox-onnx
2.ComfyUI\models\loras
Note: The yolo model is automatically downloaded. You need to download breastsizeslideroffset.safetensors at https://civitai.com/models/131864/breast-size-slider.
If you don't have breastsizeslideroffset.safetensors, you will be stuck at the LoRA Stacker node
do not try it!!!!!gpu, maybe will broken,unkown err....
Error occurred when executing DWPreprocessor:
could not broadcast input array from shape (640,426,4) into shape (640,426,3)
I got this error when using same image as you and clicking queue prompt (i'm a total beginner)
Add an "Images to RGB" node (can just copy paste from right nearby) before the "DWPose Estimator" node.
This is working really good! Any ideas on how to get the skin tone matching better to the original skin?
It simply does not work.
Thank you for your efforts.