This is a simple image classification project trained on the top of Keras/Tensorflow API with MobileNetV2 deep neural network architecture having weights considered as pre-trained 'imagenet' weights. The trained model (mask-detector-model.h5) takes the real-time video from webcam as an input and predicts if the face landmarks in Region of Interest (ROI) is 'Mask' or 'No Mask' with real-time on screen accuracy.