search

3dpose / GnTCN

Graph and Temporal Convolutional Networks for 3D Multi-person Pose Estimation in Monocular Videos
MIT License
96 stars 13 forks source link

About Network Architecture #1

Closed ZJJTSL closed 3 years ago

ZJJTSL commented 3 years ago

i wonder if u can share the details of network architecture of gcns and tcns, it's hard to understand the implementation according to the codes.

3dpose commented 3 years ago

Please refer to the Graph Convolutional Layers and Temporal Convolutional Network sections in Supplementary Material for the details of network architecture of gcns and tcns, see page 10 in the arXiv paper.