Comfyui is absolutely not beginner friendly, I would encourage entry to the rabbit hole with
You must be registered to see the links
or
You must be registered to see the links
's UI.
Here's a couple of examples of workflows:
View attachment 3364478
View attachment 3364479
The first is my own custom workflow with multiple stages of LORA injection and hires passes, using two separate models loaded for different passes. Features two stages of facial restoration, and one stage of hand restoration.
The second workflow is an absolute monstrosity called 'nudify' made by a community member. For larger images it can take my 3080ti upwards of 10 minutes to spit out one image that sucks, then fine tune, and start again.
I'd love to replicate some of it's features for detecting body features but it's a little bit beyond me.
My attempts at demosaic with stable diffusion have been mostly failures and not worth sharing lol.