“…Besides, recent improvements in ML and DL allow object detection techniques to be quite efficient, which has benefited researchers to measure and monitor SD among pedestrians in public areas by analyzing recorded videos from fixed surveillance (e.g., CCTV cameras) or drone-based surveillance ( Elharrouss et al, 2021 , Haq et al, 2022 ). Typically, vision-based IoT systems already installed in public areas can be augmented with the people detection capability, which is a sub-task of the generic object detection process ( Gaisie et al, 2022 , Manzira et al, 2022 ). Moving on, adequate measures can be then initiated to measure the physical distances between detected pedestrians.…”