pytorch_Realtime_Multi-Person_Pose_Estimation

This is a pytorch version of Realtime_Multi-Person_Pose_Estimation, origin code is here https://github.com/ZheC/Realtime_Multi-Person_Pose_Estimation

Introduction

Code repo for reproducing 2017 CVPR Oral paper using pytorch.

Results

Contents

1. Testing

2. Training

Require

1. Pytorch

2. Caffe is required if you want convert caffe model to a pytorch model.

Testing

• cd model; sh get_model.sh to download caffe model or download converted pytorch model(https://www.dropbox.com/s/ae071mfm2qoyc8v/pose_model.pth?dl=0).

• cd caffe_to_pytorch; python convert.py to convert a trained caffe model to pytorch model. The converted model have relative error less than 1e-6, and will be located in ./model after convert.

• python picture_demo.py to run the picture demo.

• python web_demo.py to run the web demo.

Training

• cd training; bash getData.sh to obtain the COCO images in dataset/COCO/images/, keypoints annotations in dataset/COCO/annotations/ and COCO official toolbox in `dataset/COCO/coco/ .

• cd training/dataset/COCO/coco/PythonAPI; sudo python setup.py install to install pycocotools .

Related repository

• CVPR'16, Convolutional Pose Machines.

• CVPR'17, Realtime Multi-Person Pose Estimation.

Network Architecture

• testing architecture 

• training architecture 

Citation

Please cite the paper in your publications if it helps your research:

@InProceedings{cao2017realtime,
  title = {Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields},
  author = {Zhe Cao and Tomas Simon and Shih-En Wei and Yaser Sheikh},
  booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
  year = {2017}
  }