Real Time Multi-Scale Facial Mask Detection and Classification Using Deep Transfer Learning Techniques
Ssvr Kumar Addagarla
Abstract
In the era of deep learning, object detection plays an influential role for many industries Detecting minute things are very much essential without human intervention especially at large scale industries In this paper we have proposed multiple approaches for Multi-scale facial mask real time detection and classification for the hospital industry, crowd surveillance in the streets and malls are more useful in this COVID-19 Pandemic Situation In our approach we have implemented two different detection models which are FMY3 using Yolov3 Algorithm and FMNMobile using NASNetMobile and Resnet_SSD300 Algorithms and used two different face mask dataset with 680 and 1400 images respectively We have analyzed both the models by computing various probabilistic accuracy measures and achieved the 34% Mean Average Precision (mAP) and 91 7% Recall rate on FMY3 Model and achieved the 98% and 99% of accuracy and recall rate on FMNMobile Model Finally we have shown results of various face mask detections from both the models © 2020, World Academy of Research in Science and Engineering All rights reserved