search

microsoft / voxelpose-pytorch

Official implementation of "VoxelPose: Towards Multi-Camera 3D Human Pose Estimation in Wild Environment"
MIT License
486 stars 92 forks source link

"pose3d[:, 0:3] = pose3d[:, 0:3].dot(M)" what means about M #11

Closed special-yang closed 3 years ago

special-yang commented 3 years ago

"pose3d[:, 0:3] = pose3d[:, 0:3].dot(M)" in https://github.com/microsoft/voxelpose-pytorch/blob/main/lib/dataset/panoptic.py line 160 why dot M?

special-yang commented 3 years ago

"pose3d[:, 0:3] = pose3d[:, 0:3].dot(M)" in https://github.com/microsoft/voxelpose-pytorch/blob/main/lib/dataset/panoptic.py line 160 why dot M?

CHUNYUWANG commented 3 years ago

This is to change the coordinate system of poses in the Panoptic dataset such that we can use our projection code directly.

special-yang commented 3 years ago

This is to change the coordinate system of poses in the Panoptic dataset such that we can use our projection code directly.

get ， to change the Panoptic coordinate system to math coordinate system(grid)

ZichengDuan commented 3 years ago

@special-yang @CHUNYUWANG Hi, I also have this confusion. According to panoptic toolbox guidance, the 3d coordinates are denoted as [x1, y1, z1], while the transformation matrix M will transfer [x1, y1, z1] into [x1, z1, -y1]. I still don't get the meaning of this. Could you please share your insights in this translation matrix? Many thanks.