Your browser doesn't support javascript.
Artificial Intelligence-Powered Contactless Face Recognition Technique for Internet of Things Access for Smart Mobility
Wireless Communications & Mobile Computing (Online) ; 2022, 2022.
Article in English | ProQuest Central | ID: covidwho-2053444
ABSTRACT
A contactless system became necessary for smart mobility during the COVID-19 pandemic. There are many touchpoints in private and public areas where contact is essential, such as intelligent transportation systems for vaccine carriers, patient ambulances, elevators, metros, buses, hospitals, and banks. A secured contactless device reduces the chances of COVID-19 infection spread. Several devices use smart cards, fingerprint identification, or code-based access. Most of these devices require some form of touch. The cost of such devices varies, depending on their capability and intended use. Sensors developed by using artificial intelligence (AI) to provide secured access are an emerging area. This paper presents an AI-powered contactless face recognition system. The solution has the Internet of Things (IoT) enabled access system. To identify a person, it uses AI assistance for face recognition with the help of Python Dlib’s facial recognition network. Dlib offers a wide range of functionality across several machine learning sectors and is open-source. The Arduino Uno (ATmega328P) and STK500 protocol has been used for communication to testify and validate the performance of the proposed technique. The objective is to detect and recognize faces by the proposed contactless approach. The obtained result shows 92% accuracy, 94% sensitivity, 96% precision and FRR 6% for face detection. There is a significant improvement in FRR in our work compared to the published 27.27%. The implemented solution in this paper provides accurate and secure contactless access to conventional, readily available techniques in public health safety.
Keywords

Full text: Available Collection: Databases of international organizations Database: ProQuest Central Language: English Journal: Wireless Communications & Mobile Computing (Online) Year: 2022 Document Type: Article

Similar

MEDLINE

...
LILACS

LIS


Full text: Available Collection: Databases of international organizations Database: ProQuest Central Language: English Journal: Wireless Communications & Mobile Computing (Online) Year: 2022 Document Type: Article