This repository contains the code to build a Convolutional neural network (RestNet50) to detect if a person is correctly wearing a mask (CWM), incorrectly wearing a mask (IWM) or not wearing a mask (NWM)

In deep learning, a Convolutional Neural Network (CNN or ConvNet) is a class of deep neural networks, most commonly applied to analyzing visual imagery. A CNN usually have multiple convolutional layers to extrac useful information from images.
A Residual Neural Network (ResNet) is an artificial neural network (ANN) of a kind that builds on constructs known from pyramidal cells in the cerebral cortex. Residual neural networks do this by utilizing skip connections, or shortcuts to jump over some layers. Typical ResNet models are implemented with double- or triple- layer skips that contain nonlinearities (ReLU) and batch normalization in between.

All the necessary modules are listed in requirements.txt. Create a virtual environment and install the modules

pip3 install -r requirements.txt

MAKE SURE YOUR CPU SUPPORTS AVX INSTRUCTIONS THAT TENSORFLOW 2.3.1 USES IF YOU WANT TO RUN IT ON CPU

The data was collected from various source

• scraper
• Google Drive - see here for more info
• MaskedFace-Net

ResNet50 is a residual neural network that is 50 layers deep. For this specific project the ResNet50 has additional few layers along with a dropout layer to address overfitting. The input of the network is of size 224x224x3 and output is a dense layer of dimension 3x1 to predict between CWM, IWM and NWM. The network has a total of 23,850,371 parameters, of which 23,797,251 are trainable and 53,120 are fixed or non-trainable.
The visual plot of the model architecture is shown here. After runnig it for 10 epochs the model got 98% accuracy on the test data

To clone the repo

git clone https://github.com/boop34/face-mask-detection.git

To get the model.h5 you need to install git-lfs. Please refer to the link to learn more.
To get the face detection model cfg and weights refer to here.
To set up the server, run app.py

./app.py

or

python app.py

To use the client run detect_face_mask.py with appropriate arguments, example

./detect_face_mask.py -i sample.jpg

and to learn more about the arguments run detect-face-mask.py without any arguments

./detect_face_mask.py

or with -h or --help option

./detect_face_ask.py --help

To build a docker image of the server run the following command

docker build -t face_mask_detection .

Now to make a container run the following command to access the server on the default flask port 5000

docker run -p 5000:5000 face_mask_detection

otherwise to acces it on different port or to create multiple containers (for example: port 80, 81 and so on), make the adjustments in the detect_face_mask.py as needed

docker run -p 80:5000 face_mask_detection
docker run -p 81:5000 face_mask_detection

there are various other things that can be done with docker, check their official page for more information.

Input Image	Output Image