in this project, we will train mask detection model based on 12k training pictures from kaggle dateset in this url: https://www.kaggle.com/ashishjangra27/face-mask-12k-images-dataset
Perform data augmantation of the 10k traing data set to over come over fitting as known that data augmentation is one of the solution of overfitting.
Using Pre-trained model " Mobile net v2" in https://www.tensorflow.org/api_docs/python/tf/keras/applications/mobilenet_v2/MobileNetV2 to tranfer learning to get better accuracy.
getting accuracy for validation set over 99%
Testing the model using real images with different faces in the same image First we detect faces using haarcascade_frontalface_default and then pass every detected face to our trained model to identify if the person wearing mask or not
