Crowd Counting via Cross-stage Refinement Networks

Paper

This is a Tensorflow implementation of TIP 2020 paper "Crowd Counting via Cross-stage Refinement Networks".
Yongtuo Liu, Qiang Wen, Haoxin Chen, Wenxi Liu, Jing Qin, Guoqiang Han, and Shengfeng He*

Requirement

• Python 3.5
• Tensorflow 1.9.0
• Matplotlib 2.0.2
• h5py 2.8.0
• scipy 1.1.0
• Ubuntu 16.04 LTS

Dataset

The ShanghaiTec dataset can be downloaded from Baidu Disk: link.
Other counting-related datasets can be found in link.

Training

python train.py

Testing

python test.py

Citation

@article{liu2020crowd,
title={Crowd Counting via Cross-stage Refinement Networks},
author={Liu, Yongtuo and Wen, Qiang and Chen, Haoxin and Liu, Wenxi and Qin, Jing and Han, Guoqiang and He, Shengfeng},
journal={IEEE Transactions on Image Processing},
year={2020}
}

Contact

For any questions with respect to this paper and code, please feel free to contact Yongtuo Liu.