Real-time face mask detection with alert system

Real-time mask detection and prediction are the main objectives of research using deep learning, which was established in two stages; the first stage was training a model using a pre-trained MobileNetV2 network for image classification. Our training data was 1785 images divided into two categories (mask without mask). In the second stage, we use the OpenCV library to detect faces in real-time video streams using the 'Single Shot Multi-box Detector' (SSD) detector and the ResNet- 10 architecture. Then we utilize these results as inputs to our model to determine whether or not someone is wearing a mask. The F1 score was 100 percent, while the model accuracy was 99.72 percent. ‎Training data was obtained from various sources, and all downloaded from the internet.