Built a robust model for mask detection using transfer learning(VGG19) and HaarCascade. Successfully achieved an accuracy of 98%. Haarcascade classifier was used to detect the faces and get the bounding boxes across face. Which was further feed into model to classify into mask and not mask labels.
dataset link:-https://drive.google.com/file/d/1mdGtZRqQWR5ivfpCgjx2rTTSXcd-2UNB/view?usp=sharing Published Paper:- Patel, Chintan, et al. "Mask detection as Covid combat using Transfer learning and Haar Cascade." (2021).