An AI-based Mask-Wearing Status Recognition and Person Identification System
2021 International Symposium on Advanced Technologies and Applications in the Internet of Things, ATAIT 2021
; 3131:20-27, 2021.
Article
Dans Anglais
| Scopus | ID: covidwho-1843201
ABSTRACT
In the tough COVID-19 pandemic, wearing a mask in daily life becomes an important habit. However, sometimes people forget to wear a mask or wear a mask incorrectly in careless. Hence, alarming the problem of protecting ourselves from COVID-19 becomes a key challenge. Unfortunately, at home security or oce security systems, wearing a mask lets the person identication lost its function. Hence, masked-person identication becomes an essential issue. This paper proposes an AI-based mask-wearing status recognition and person identication system for solving the above problems. The system consists of three stages, face detection based on MTCNN, mask-wearing status recognition, and person identication using MobileNetV2. Masked-person identication is one of the functions of the proposed system. The experimental results show that the face detector reaches almost 100% accuracy among 3000 images. The mask-wearing status recognition has a 96.1% test accuracy in 300 test images, and person identication achieves a 98% recognition rate. In summary, the eectiveness of the proposed system is proved by the high accuracy recognition rate. © 2021 CEUR-WS. All rights reserved.
Collection:
Bases de données des oragnisations internationales
Base de données:
Scopus
langue:
Anglais
Revue:
2021 International Symposium on Advanced Technologies and Applications in the Internet of Things, ATAIT 2021
Année:
2021
Type de document:
Article
Documents relatifs à ce sujet
MEDLINE
...
LILACS
LIS