The model consists of the followwing steps:
1.Keypoint detection
2.Inverse IK
3.Motion propagation
4.Mesh refinement
Detailed documentation will be updated.
If you find our code or paper helps, please consider citing:
@article{jnchen23ws3dpt,
title={Weakly-supervised 3D Pose Transfer with Keypoints},
author={Jinnan Chen and Chen Li Gim Hee Lee},
year={2023}
}