It's a yolov3 based project. This project fullly based on https://github.com/ultralytics/yolov3. About how to train/infer , you can go to https://github.com/ultralytics/yolov3 for referance.
The repo contains inference and training code for YOLOv3 in PyTorch. The code works on Linux, MacOS and Windows.It trained on wearing_helmet_or_not dataset.It is now private.But you can use the traind weights in the weights dictionary.
Well,I think it's meaningless to run it on the laptop or desktop.When deployed to the cloud, it well be productive.
So, there is a server.py also a client.py.
Python 3.7 or later with the following pip3 install -U -r requirements.txt packages:
waitress face_recognition flask requests numpy opencv-python torch >= 1.2 matplotlib pycocotools tqdm tb-nightly future Pillow
download weights: https://pan.baidu.com/s/1QwqA3N92BgoI57Y6DOeQ0w Code：Y52j
git clone https://github.com/ralph0813/Who_wants_to_die.git
pip3 install -U -r requirements.txt
download weights baidu: https://pan.baidu.com/s/1QwqA3N92BgoI57Y6DOeQ0w Code：Y52j
Google Drive :https://drive.google.com/open?id=1s0eLELJNOxsuU7B_0zPmb3UhSj_HD9hq
Copy the .weights to weights forder.
cd Who_wants_to_die
python3 server.py
Another terminal:
python3 client.py --file data/samples/timg.jpeg or python3 client.py --file /path/to/your/picture/
If you wants to add other faces,find a Positive face photo and put it into data/known_faces/ after restart server,it will be registered.