Deep Learning Based FPN and MT-CNN Face Mask Detection System

In this study, we provide a novel face mask identification method that utilizes FPN and MT-CNN models trained using deep learning. Images of persons with and without face masks were used to test the proposed models. The FPN model got 97.3% accuracy while the MT-CNN model got 98.1% accuracy. Both models performed better than previous methods of face mask detection. However, the MT-CNN model beat the FPN model in both accuracy and efficiency, possibly because to its capacity to recognize face landmarks and angles, whereas the FPN model demonstrated computational efficiency. In particular, the work emphasizes the promise of deep learning-based models for face mask recognition and hints that such models might offer a dependable and effective way of ensuring compliance with face mask laws in public places. Optimizing the models' performance, assessing the models in a variety of real-world circumstances, delving into the ethical implications, and creating user-friendly and accessible face mask detection systems are all potential future paths for study.