“…Here, author uses a deep neural network to identify the behaviors of a lot of persons in a video frame. Due to the high-quality video process [5,24], data travel time across the network, and the time it takes to access and present the feed on a screen, security cameras might lag or experience a delay [4,22]. However, to mitigate such a problem, we proposed a new architecture for our system that detects suspicious activity based on video footage in real-time and provides the most e cient and reliable performance.…”