Double-chain Constraints for 3D Human Pose Estimation in Images and Videos

Double-chain Constraints for 3D Human Pose Estimation in Images and Videos,
Hongbo Kang, Yong Wang, Mengyuan Liu, Doudou Wu, Peng Liu, Wenming Yang
arXiv, 2023

Results on Human3.6M

Protocol 1 (mean per-joint position error) when 2D keypoints detected by CPN, HRNet and the ground truth of 2D poses.

Method	2D Pose	MPJPE
DC-GCT	GT	32.4 mm
DC-GCT	CPN	48.4 mm
DC-GCT (w/refine)	CPN	47.4 mm
DC-GCT	HRNet	47.2 mm
DC-GCT (w/refine)	HRNet	46.1 mm

Dependencies

Python 3.7+
PyTorch >= 1.10.0

pip install -r requirement.txt

Dataset setup

Please download the dataset here and refer to VideoPose3D to set up the Human3.6M dataset ('./dataset' directory).

${POSE_ROOT}/
|-- dataset
|   |-- data_3d_h36m.npz
|   |-- data_2d_h36m_gt.npz
|   |-- data_2d_h36m_cpn_ft_h36m_dbb.npz

Download pretrained model

The pretrained model is here, please download it and put it in the './ckpt/pretrained' directory.

Test the model

To test on Human3.6M on single frame, run:

python main.py --reload --previous_dir "ckpt/pretrained"

Train the model

To train on Human3.6M with single frame, run:

python main.py --train -n 'name'

Demo

To begin, download the YOLOv3 and HRNet pretrained models here and put it in the './demo/lib/checkpoint' directory. Next, download the pretrained model and put it in the './ckpt/pretrained' directory. Lastly, Put your own images in the './demo/figure', and run:

python demo/vis.py

Citation

If you find our work useful in your research, please consider citing:

@article{kang2023double,
title={Double-chain Constraints for 3D Human Pose Estimation in Images and Videos},
author={Kang, Hongbo and Wang, Yong and Liu, Mengyuan and Wu, Doudou and Liu, Peng and Yang, Wenming},
journal={arXiv preprint arXiv:2308.05298},
year={2023}
}

Acknowledgement

Our code is extended from the following repositories. We thank the authors for releasing the codes.

Licence

MIT

KHB1698/DC-GCT