Does fooocus support control net?

[martinenkoEduard]

Does fooocus support control net?

[Crocodilian12]

That's what the image prompts are, ControlNets. The basic image prompt is a tweaked version of IP Adaptor; PyraCanny and CPDS are both ControlNets . . . I'm not sure about the Faceswap tools.

[ckotty-alex]

That's what the image prompts are, ControlNets. The basic image prompt is a tweaked version of IP Adaptor; PyraCanny and CPDS are both ControlNets . . . I'm not sure about the Faceswap tools.

faceswap is faceid ipadapter

[ckotty-alex]

by the way, openpose works too in Fooocus. I've changed control-lora-canny to controlnet union, turn on "skip preprocessors" , then insert openpose images to image promp "canny" and it's really works

[Yogesh-DevHub]

Thanks for mentioning, was using comfy UI for that. Will try in fooocus.

[cubiwan]

I have tested using openpose images and it works, no code changes are necessary:

Download diffusion_pytorch_model.safetensors

Save in controlnet folder

Rename to control-lora-canny-rank128.safetensors

Launch Fooocus

Advanced tab, Control tab, check Skip Preprocessors

Image Prompt tab with an openpose skeleton image (512x512 works well)

Select PyraCanny option

Select Aspect Ratio 1:1 (with another the generated image appears cut off)

Write a prompt and generate image.

[briziomusic]

I have tested using openpose images and it works, no code changes are necessary:

Download diffusion_pytorch_model.safetensors

Save in controlnet folder

Rename to control-lora-canny-rank128.safetensors

Launch Fooocus

Advanced tab, Control tab, check Skip Preprocessors

Image Prompt tab with an openpose skeleton image (512x512 works well)

Select PyraCanny option

Select Aspect Ratio 1:1 (with another the generated image appears cut off)

Write a prompt and generate image.

Sorry, could you please guide me? I'm not able to reproduce.

is this image uploaded to 'pyracanny' okay (512x512)?

result

controlnet folder

the file was downloaded from
https://huggingface.co/InstantX/InstantID/blob/main/ControlNetModel/diffusion_pytorch_model.safetensors
and renamed to control-lora-canny-rank128.safetensors

what am I doing wrong?

[Yogesh-DevHub]

Get this one https://huggingface.co/xinsir/controlnet-union-sdxl-1.0/tree/main

it will work on default stop-at and weight, try keeping aspect ratio same as the uploaded image, it has an impact on result

[briziomusic]

Get this one https://huggingface.co/xinsir/controlnet-union-sdxl-1.0/tree/main

it will work on default stop-at and weight, try keeping aspect ratio same as the uploaded image, it has an impact on result

Thank you. It's working!