search

Jeff-sjtu / HybrIK

Official code of "HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimation", CVPR 2021
MIT License
1.17k stars 142 forks source link

question about camera transformation #87

Open afiretony opened 2 years ago

afiretony commented 2 years ago

Hi, thanks for the great work? Could you briefly explain how can we get camera intrinsic and extrinsic parameters from the predicted results? I checked "cam_root" and "transl" from pose_out but not sure what to do with them. Thanks!

Jeff-sjtu commented 1 year ago

Hi @afiretony, For the model with camera prediction, cam_root is the root coordinate of the human body. The intrinsic parameters we used here are: focal=1000 and image center as the principal point.

Tengoles commented 1 year ago

@Jeff-sjtu the coordinates from cam_root are relative to the position of the camera in transl or cam_trans?