[Feature]: About HANDS..

[RGX650]

Feature description

In a real life scenario, controlling and customizing handpose is quite challenging besides being able to generate it consistently.

I mean, lets redefine the very first starting points: it can be only by 2 possibilities: -reference image -pose data input quality, interpretability, consitency BUT ALSO ability to be really controlled, finetuned or adjusted in time!

PRE ANNOTATOR / GUIDED REFINER ?

PRE ANNOTATOR: I am wondering, would it be relevant, possible to consider a "pre-annotator" that would snapshot a UltraLeap Leapmotion controller pose. That "pre-annotator" would then be used to guide/finetune any exising pose related annotators & controlnets: (handrefiner, openpose hand, dwpose, densepose, "sided mediapipe R/L hand(futur?)

GUIDED REFINER: Offical pytorch implementation of "HandOccNet: Occlusion-Robust 3D Hand Mesh Estimation Network", CVPR 2022. https://github.com/namepllet/HandOccNet.git

https://github.com/Fannovel16/comfyui_controlnet_aux/issues/205

Version Platform Description

18:45:30-774495 INFO Python 3.10.11 on Windows 18:45:31-203817 INFO Version: app=sd.next updated=2023-12-30 hash=ab7b78cc url=https://github.com/vladmandic/automatic.git/tree/master 18:45:31-774079 INFO Platform: arch=AMD64 cpu=AMD64 Family 25 Model 97 Stepping 2, AuthenticAMD system=Windows release=Windows-10-10.0.22631-SP0 python=3.10.11

[eadnams22]

Controlnet with open pose and open pose editor does this.

beyond that, this should probably be an extension, not baked in.

[vladmandic]

closing based on the feedback provided.