/HRNet_Pose_Estimation_TensorFlow2

A tensorflow2 implementation of HRNet for human pose estimation.

Primary LanguagePythonMIT LicenseMIT

HRNet_Pose_Estimation_TensorFlow2

A tensorflow2 implementation of HRNet for human pose estimation.

Requirements:

  • Python == 3.7
  • TensorFlow == 2.1.0
  • numpy == 1.17.0
  • opencv-python == 4.1.0

Usage

Prepare dataset

  1. Download the COCO2017 dataset.
  2. Unzip the train2017.zip, annotations_trainval2017.zip and place them in the 'dataset' folder, make sure the directory is like this :
|——dataset
    |——COCO
        |——2017
            |——annotations
            |——train2017

Train on COCO2017

  1. Open the file "./configuration/base_config.py", change the parameters according to your needs.
  2. Run "write_coco_to_txt.py" to generate coco annotation files.
  3. Run "train.py" to train on coco2017 dataset.

Test

  1. Prepare the test pictures and make sure that the TEST_PICTURES_DIRS in "./configuration/base_config.py" is correct.
  2. Run "test.py" to test on your test pictures.

Acknowledgments

  1. https://github.com/leoxiaobin/deep-high-resolution-net.pytorch

Citation

@inproceedings{sun2019deep,
  title={Deep High-Resolution Representation Learning for Human Pose Estimation},
  author={Sun, Ke and Xiao, Bin and Liu, Dong and Wang, Jingdong},
  booktitle={CVPR},
  year={2019}
}

@inproceedings{xiao2018simple,
    author={Xiao, Bin and Wu, Haiping and Wei, Yichen},
    title={Simple Baselines for Human Pose Estimation and Tracking},
    booktitle = {European Conference on Computer Vision (ECCV)},
    year = {2018}
}

@article{WangSCJDZLMTWLX19,
  title={Deep High-Resolution Representation Learning for Visual Recognition},
  author={Jingdong Wang and Ke Sun and Tianheng Cheng and 
          Borui Jiang and Chaorui Deng and Yang Zhao and Dong Liu and Yadong Mu and 
          Mingkui Tan and Xinggang Wang and Wenyu Liu and Bin Xiao},
  journal   = {CoRR},
  volume    = {abs/1908.07919},
  year={2019}
}

References

  1. Paper: Deep High-Resolution Representation Learning for Human Pose Estimation