Detecting face mask with CNN architecture and transfer learning with Imagnet
Tried different cnn architectures(with and without data augmentation) and uploaded the one resulting in higher accuracy (keeping bias and variance in mind) I found transfer learning working better in this case.
PS: I have created a folder(called images) which would contain with_mask and without_mask folders.
Dataset borrowed from : https://github.com/chandrikadeb7/Face-Mask-Detection (Thank you :) )