# CVPR2020-OOH

**Repository Path**: AI52CV/CVPR2020-OOH

## Basic Information

- **Project Name**: CVPR2020-OOH
- **Description**: This is the repository of the implemantation of cvpr 2020. 
- **Primary Language**: Python
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 1
- **Forks**: 6
- **Created**: 2021-04-06
- **Last Updated**: 2021-04-06

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Object-Occluded Human Shape and Pose Estimation from a Single Color Image (CVPR2020 Oral)

## Introduction
This is the offical implentation of 'Object-Occluded Human Shape and Pose Estimation from a Single Color Image' (CVPR2020 Oral). [[Project Page](https://www.yangangwang.com/papers/ZHANG-OOH-2020-03.html)]

![figure](https://images.gitee.com/uploads/images/2020/0601/202625_9aa5f92b_5505182.png "pipeline1.png")

## Environment
This code is based on Python 3.7, CUDA 10.0, cuDNN 7.6 on Windows10. 

## Getting Started
Clone the repo:
```
git clone https://gitee.com/seuvcl/CVPR2020-OOH.git
```
Install the requirements using conda:
```
conda create -n occlusion python=3.7
conda install pytorch==1.1.0 torchvision==0.3.0 cudatoolkit=10.0 -c pytorch
pip install -r requirements.txt
```

## Running the Demo
First, you need download the trained model from [here](https://seueducn1-my.sharepoint.com/personal/yangangwang_seu_edu_cn/_layouts/15/onedrive.aspx?originalPath=aHR0cHM6Ly9zZXVlZHVjbjEtbXkuc2hhcmVwb2ludC5jb20vOmY6L2cvcGVyc29uYWwveWFuZ2FuZ3dhbmdfc2V1X2VkdV9jbi9FczZrRnZHTzJXMUdxRDJnbzVQQUF5MEItZlcxZUprVWxYQzVCTHA5dGd1VGpnP3J0aW1lPVNLenl2c1ZGMkVn&id=%2Fpersonal%2Fyangangwang%5Fseu%5Fedu%5Fcn%2FDocuments%2F3DOH50K%2Fmodel) and put it in ```trained_model``` folder. Then, download the official [SMPL](http://smplify.is.tue.mpg.de/) model and [preprocess](https://github.com/CalciferZh/SMPL) the neutral model to support pytorch implementation. Finally, put the generated ```model.pkl``` in ```data``` folder.

To test on your own image, you can edit the ```cfg_files\demo.yaml``` and run:
```
python demo.py --config cfg_files\demo.yaml
```

## Dataset
We provide the [3DOH50K dataset](https://pan.baidu.com/s/1j1pYGCoPjlbpIBt0Nn1l_g) (<font color=red>extraction code</font> **\[hb1d\]**), which is the first real 3D human dataset for the problem of human reconstruction and pose estimation in occlusion scenarios.
Visualizing 3DOH50K:
```
python utils/visualize.py --base_dir /PATH/TO/THE/3DOH50K
```
## Citation
```
@inproceedings{ooh20,
  title = {Object-Occluded Human Shape and Pose Estimation from a Single Color Image},
  author = {Tianshu, Zhang and Buzhen, Huang and Yangang, Wang},
  booktitle = {Proceedings IEEE Conf. on Computer Vision and Pattern Recognition (CVPR)},
  year = {2020}
}
```

## References
PyTorch implementation of SMPL model is from [CalciferZh](https://github.com/CalciferZh/SMPL).