Skip to content

Latest commit

 

History

History
33 lines (33 loc) · 1.75 KB

README.md

File metadata and controls

33 lines (33 loc) · 1.75 KB
Raw

Context-Aware_Crowd_Counting-pytorch

This is an simple and clean unoffical implemention of CVPR 2019 paper "Context-Aware Crowd Counting".

Installation

 1. Install pytorch 1.0.0 later and python 3.6 later
 2. Install visdom

pip install visdom

 3. Install tqdm

pip install tqdm

 4. Clone this repository

git clone https://github.com/CommissarMa/Context-Aware_Crowd_Counting-pytorch.git

We'll call the directory that you cloned Context-Aware_Crowd_Counting-pytorch as ROOT.

Data Setup

 1. Download ShanghaiTech Dataset from Dropbox: link or Baidu Disk: link
 2. Put ShanghaiTech Dataset in ROOT and use "data_preparation/k_nearest_gaussian_kernel.py" to generate ground truth density-map. (Mind that you need modify the root_path in the main function of "data_preparation/k_nearest_gaussian_kernel.py")

Training

 1. Modify the root path in "train.py" according to your dataset position.
 2. In command line:

python -m visdom.server

 3. Run train.py

Testing

 1. Modify the root path in "test.py" according to your dataset position.
 2. Run test.py for calculate MAE of test images or just show an estimated density-map.

Other notes

we got the comparable MAE at the 353 epoch BaiduDisk download with Extraction code: yfwb or Dropbox Link which is reported in paper. Thanks for the author's(Weizhe Liu) response by email. His mainpage is link.