A crowd counting application based on the CSRNet benchmark
ShanghaiTech Dataset: Google Drive
Recommended environment: Anaconda.
Python: 3.7
PyTorch: 1.7.0
CUDA: 10.1
Follow the main.ipynb to generate the ground truth.
Note: you need to generate your own json file
Run command python train.py train.json val.json 0 0 to start training process.
Follow the val.ipynb to try the validation.