face-mask-detection

Hello welcome to my face mask detection

its a python model which can detect faces with FaceBoxes : https://github.com/zisianw/FaceBoxes.PyTorch and classify if the face has a mask or not

Requirements

Ubuntu " It's only tested on Ubuntu, so it may not work on Windows "

GPU : Any GPU that is works with PyTorch with atleast 4GP vram

numpy : https://numpy.org/

PyTorch : https://pytorch.org/

torchvision : https://pypi.org/project/torchvision/

openCV : https://pypi.org/project/opencv-python/

Accuracy

data set : https://www.kaggle.com/ashishjangra27/face-mask-12k-images-dataset

train set accuracy is : 100%

test set accuracy is : 99.8%

val set accuracy is : 100%

to download my model : https://drive.google.com/file/d/15JenLZWt4qh24PbfaNgGaiCg0pZ16dWt/view?usp=sharing

Usage

• Clone the Repo :

  $ git clone https://github.com/AbdallahOmarAhmed/face-mask-detection
  

• Compile the nms :

  $ bash ./make.sh
  

• Download the data set and put it in your project dir :

  https://www.kaggle.com/ashishjangra27/face-mask-12k-images-dataset
  

• Run the train file to start training :

  $ python3 train_model.py
  

• OR : download my model and put it in the weights folder

• Run the test file to Calculate accuracy :

  $ python3 test_model.py
  

• Add a video to test then run the model :

  $ python3 make_video.py
  

• Convert the models to onnx :

  $ python3 make_onnx_copy.py
  

Note : the name of the video that you want to test must be ' test.mp4 ' and the output video name will be ' outpy.avi '