New normal: cooperative paradigm for COVID-19 timely detection and containment using Internet of things and deep learning

Abstract

The spread of the novel coronavirus (COVID-19) has caused trillions of dollars of damages to the governments and health authorities by affecting the global economies. It is essential to identify, track and trace COVID-19 spread at its earliest detection. Timely action can not only reduce further spread but also help in providing an efficient medical response. Existing schemes rely on volunteer participation, and/or mobile traceability, which leads to delays in containing the spread. There is a need for an autonomous, connected, and centralized paradigm that can identify, trace and inform connected personals. We propose a novel connected Internet of Things (IoT) based paradigm using convolution neural networks (CNN), smart wearable, and connected E-Health to help governments resume normal life again. Our autonomous scheme provides three-level detection: inter-object distance for social distancing violations using CNN, area-based monitoring of active smartphone users and their current state of illness using connected E-Health, and smart wearable. Our exhaustive performance evaluation identifies that the proposed scheme with CNN YOLOv3 achieves up to 90% mean average precision in detecting social distancing violations, as compared to existing YOLOv2 achieving 70% and faster R-CNN with 76%. Our evaluation also identifies that wearing protective gear reduces spread by 50%, and timely actions in the first hour can help avoid three times COVID-19 exposure.