search

lmb-freiburg / hand3d

Network estimating 3D Handpose from single color images
GNU General Public License v2.0
802 stars 252 forks source link

Detecting both hands #37

Open HelloEgg opened 4 years ago

HelloEgg commented 4 years ago

Hi. Thanks for the great work! Can this network detect both hands at a time? I couldn't find a code that can pose estimate both hands simultaneously from run.py. So I manually masked one side of image if both hands are present. Would this be only option for detecting 2 hands? or would there be a more convenient way? Thankyou

QuYJLeo commented 4 years ago

Hi. Thanks for the great work! Can this network detect both hands at a time? I couldn't find a code that can pose estimate both hands simultaneously from run.py. So I manually masked one side of image if both hands are present. Would this be only option for detecting 2 hands? or would there be a more convenient way? Thankyou

Hi0, Have you found a way to detect two hands? And How？

HelloEgg commented 4 years ago

@QuYJLeo After trying various... methods, two methods seemed to work :

  1. Manually masking either side. Detecting one side of hand and masking other half of frame... very primitive method but somehow it did work. Note that naive masking may lose channel value which may cause the network to confuse correct hand position.
  2. Use Mano model (which is different project) I shifted to using MANO model (https://mano.is.tue.mpg.de/) which can detect both hands (I'm not sure if it offered both hands in github repo. I may have wrote my own code to port parameters. You may also need Openpose 2D pose estimation result)