ControlNet is awesome. If you need a specific pose you can either spend a weekend creative-writing and filling up drives with crappy LoRAs, or you can just use a reference. A smart reference that you can weight instead of denoising one into oblivion.

This is a super simple, quick and dirty depth map generator. Upload single images or do it like a batcher. Makes no difference to me. Also threw in a template group that you can copy over to a t2i workflow in case you're new to controlnet. Just plug it in per the instructions in the notes and you can use depth maps (or skeleton poses) in your existing setups. It's very handy so I thought you should have it. You have ten of these already? Ok fine, here's a crappy version of one of those. It will probably work best if you download a depth_anything model. Running any workflow without a model is inadvisable, actually. I've tried.