Reflecting on Public Behavior With Artificial Intelligence-assisted Detection of Face Mask Wearing During the COVID-19 Pandemic

Abstract

\n Background: COVID-19 has created health and socioeconomic damage worldwide, and face masks are a low-cost but effective method of preventing transmission of this disease. Artificial intelligence (AI)-assisted systems can come into play to help visualize the public’s awareness of mask wearing and gain a better picture of whether there is adequate practice of protection during the outbreak. We reported the rate of face mask wearing by the general public using the artificial intelligence-assisted face mask detector, AiMASK.Methods: This cross-sectional study was conducted between January 23 and April 22, 2021 in over 32 districts in Bangkok, Thailand. After the introduction of AiMASK, development and internal validation were performed, and average accuracy of 97.8% was found. Data were classified into a protected group (correct face mask wearing) and an unprotected group (incorrect or non-mask wearing). We analyzed the association between factors affecting the unprotected group using univariate logistic regression analysis.Results: No significant difference was found between results from human graders and those of AiMASK using two proportion Z test (p=0.74). AiMASK detected a total of 1,124,524 people, the majority of whom were in the protected group (95.98%). The unprotected group consisted of 2.06% who practised incorrect mask-wearing, and the other 1.96% were those who did not wear masks. A moderate negative correlation was found between the number of COVID-19 patients and the proportion of unprotected people (r= -0.507, p<0.001). People were 1.15 times more likely to be in the unprotected group during the holidays and in the evening than on working days and in the morning (OR=1.15, 95% CI 1.13-1.17, p<0.001). Districts in the city center were 1.31 times more likely to have higher proportions of unprotected individuals than suburban districts (OR=1.31, 95% CI 1.28-1.34, p<0.001). Conclusions: AiMASK was as effective as human graders in detecting face mask wearing. The prevailing number of COVID-19 infections affected people’s mask-wearing behavior, and half of the unprotected group were those who wore masks incorrectly. Public policies should communicate the importance of wearing masks consistently throughout the day and during holidays as well as providing instructions for effective mask wearing to prevent virus transmission.