Abstract
This article reveals how artificial intelligence and big data analytics help the non-pharmacological disease control measures. Several cutting-edge technologies are illustrated in terms of the system architecture, the data workflows, and the machine learning/deep learning models. This article will also investigate a comprehensive social control system that is designed for disease control measures by integrating the above mentioned technologies. For each component of the system, real-world applications will be represented in the form of examining the capability of the proposed models. The proposed system can detect whether people are keeping social distancing and wearing a facial mask in public spaces, along with measuring the mobility assessment, which can be applied to screen the stay-at-home orders using big data and visual mining. A fine-tuned CNN-based network will be applied for monitoring the social distancing, while the face mask detection module is trained by fine-tuning the MobileNet architecture.
TopIntroduction
During the COVID-19 pandemic, a variety of non-pharmacological disease control measures, including travel restrictions, lockdowns, stay-at-home orders, wearing-a-mask policies, and social distancing regulations, have been implemented by governments and lawmakers. Issuing travel restrictions or lockdown policies can help to mitigate the spread of COVID-19 by reducing human mobility and decreasing the probability of contact (Grépin et al., 2021). It is more important to keep social distancing now claims than ever before, since it is one of the best ways to prevent the spread of the disease except wearing face masks (Milne & Xie, 2020). Wearing masks can significantly reduce the amount of coronavirus transmitted by droplets and aerosols (Eikenberry et al., 2020). Almost all countries now carry out such non-pharmacological disease control policies as mandatory strategies. Based on the proposed requirements by the WHO, the minimum distance between individuals must be kept at least 6 feet to achieve a safe social distancing among the people. The medical researchers have pointed out that individuals with mild or no symptoms may also be carriers of the novel coronavirus (Wang et al., 2020), therefore it is important to require all people to maintain controlled behaviors and to keep social distancing. However, it may be a challenging task to monitor the amount of infection spread and the efficiency of the constraints.
Since the end of 2019, the lives of people all around the world have been drastically affected by the COVID-19 pandemic. The world economy has been in a depression due to a loss of jobs, while face-to-face communication has been restricted to control the infection rate (Feyisa, 2020). Although it has been more than 18 months since the global outbreak, medical researchers are still unable to confirm the end of the pandemic. Despite the effectiveness of the new vaccines by some degrees, the Delta variant leads the significant uncertainty, which makes the situation moving towards the undesired detection. Due to such the circumstance, research communities have pointed out that society may go through a long period of abnormality. Governments have to continue to enforce mask-wearing, social distancing, and quarantines. Many changes for society, including online education, mandatory facial masks, and the vast majority of people working from home, have been made in the new normal, and perhaps continue for a long time (Odusanya et al., 2020).
Cutting-edge technologies, such as machine learning, deep learning, computer vision, and big data analytics, can be applied to implement efficient non-pharmacological disease control measures (Lakhani et al., 2020). Motivated by the current demand in disease control for Covid-19, this chapter is designed to investigate how such technologies work to solve a broad range of real-world problems, such as tracking and visualizing stay-at-home measurements, monitoring social distancing regulations, and detecting face-mask-wearing practices. The objectives of this chapter are listed as follows:
- •
illustrating how big data helps in measuring human mobility to track stay-at-home measures with information retrieval and visual mining.
- •
investigating how deep learning and computer vision work for implementing a social distancing monitor using a pre-trained detector, i.e. YOLO v3.
- •
examining how to detect whether individuals wear masks or not using deep learning approaches, along with a real-world testing process.
Key Terms in this Chapter
Convolutional Neural Network: A typical deep learning model that is commonly used to image classification, object detection, natural language procession, and predictive analysis. Such a network structure is a regularized version of fully connected networks, which belong to the class of artificial neural network.
Non-Pharmacological Disease Control: A set of actions, apart from medicine and vaccination, that communities can slow down the spread of a disease, a.k.a. non-pharmaceutical interventions (NPIs).
Face Mask Detection: A computer-based monitor that detects whether individuals are wearing a mask or not.
Social Distancing Monitor: A technology that is designed to warn individuals when they get too close to each other, particularly relying on communications or contacts in short distances.
Human Mobility: A measurement that describes how people move within a network or system through tracking and analyzing human behavior patterns demographically and geographically over time.
Deep Learning: A broad family of machine learning models based on neural networks. Typical deep learning models are deep neural networks, convolutional neural networks, recurrent neural networks, deep belief networks, and deep reinforcement learning.
Computer Vision: An automation technology that makes computers to gain high-level understanding from images and videos throughout acquiring, processing, analyzing, and recognizing digital data by transforming visual images into numerical or symbolic information.