aditya-vora / crowd_counting_tensorflow

Code of CVPR 2016 work of Single Image Crowd Counting using Multi-Column CNN

Single Image crowd counting using Multi-Column CNN.

This is unofficial and very simple implementation of the CVPR 2016 paper on crowd counting "Single-image crowd counting via multi-column convolutional neural network". Full code is developed on Tensorflow platform.

Installation

1. Install Tensorflow
2. Clone this repository

git clone https://github.com/aditya-vora/crowd_counting_tensorflow.git

3. Download the dataset and keep it in the $ROOT/data folder

4. The dataset can be downloaded from:

   Dropbox: https://www.dropbox.com/s/fipgjqxl7uj8hd5/ShanghaiTech.zip?dl=0

   Baidu Disk: http://pan.baidu.com/s/1nuAYslz

5. Train the model using $ROOT/train.py

6. Test the model

Note: More details about the results and some other code files for data parsing will be updated soon.

Citations:

[1] Zhang, Yingying, et al. "Single-image crowd counting via multi-column convolutional neural network." Proceedings of the IEEE conference on computer vision and pattern recognition. 2016.