search

garyzhao / SemGCN

The Pytorch implementation for "Semantic Graph Convolutional Networks for 3D Human Pose Regression" (CVPR 2019).
https://arxiv.org/abs/1904.03345
Apache License 2.0
465 stars 78 forks source link

Stacked Hourglass detections #14

Open ghost opened 4 years ago

ghost commented 4 years ago

Hello!You mention in your paper that you pre-train a network to realize 2D pose estimation, but it seems that this part of network was not given in your code, and the 2D estimated result was directly used. Could you please provide the code for the python version of this network? Thanks.

garyzhao commented 4 years ago

Hi @luhengjie

Thanks for your interest in our work.

Currently, we do not have the plan to release the 2D detection part since we focus on 3D lifting in this repo.

In this repo, we directly use the 2D results provided by https://github.com/una-dinosauria/3d-pose-baseline. Please refer to their repo for how to get the 2D network.

Some good alternatives may be: https://github.com/bearpaw/pytorch-pose https://github.com/JimmySuen/integral-human-pose

Best, Long

ghost commented 4 years ago

Thank you for your reply!