Open AccessFace Mask Detection System using Mobilenetv2
Author(s) -
Mandeep Kumar Arora,
Sarthak Garg,
A. Srivani
Publication year - 2021
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.d2404.0410421
Subject(s) - computer science , convolutional neural network , artificial intelligence , face (sociological concept) , process (computing) , identification (biology) , face detection , feature (linguistics) , task (project management) , computer vision , feature extraction , facial recognition system , computer security , engineering , social science , linguistics , philosophy , botany , systems engineering , sociology , biology , operating system
In this pandemic, it is getting more and more difficultto keep a track of people who are wearing masks regularly or not.It cannot solely depend on human efforts to take care of this taskand therefore there is a need to develop software that canautomatically detect whether any given person is wearing a maskor not. Face Detection has evolved as a really popular problem inimage processing and computer vision. Many new algorithms arebeing devised using convolutional architectures to form thealgorithm as accurately as possible. These convolutionalarchitectures have made it possible to extract even the pixeldetails. Training is performed through Fully ConvolutionalNeural Networks to semantically segment out the faces present inthat image. Feature detection and feature extraction techniqueshelp us identify whether a person is wearing a mask or not. Theface mask detector will use a dataset of morphed masked images.Therefore, the created model will be accurate and it will also becomputationally efficient and easily deployable in embeddedsystems since the MobileNetV2 architecture will be incorporated(Raspberry Pi, Google Coral, etc.). This framework can also beused in real-time applications that, due to the outbreak ofCovid-19, require face-mask detection for safety purposes. Thisproject can be merged with embedded application systems atairports, train stations, workplaces, schools, and public places toensure compliance with the guidelines for public safety. Theabove topic is very prominent in recent times as the identificationprocess will not only help us classify individuals but also willreduce the workforce required to do the same exponentially.