Emergency Activity Tracking: An Intelligent System for Monitoring Human Activity

In emergency scenarios like kidnappings or terrorist attacks, retracing individuals' movements within a building or location is crucial for identifying perpetrators or victims. Traditional methods relying on eyewitness accounts and physical evidence prove unreliable and time-consuming. This paper introduces a novel system for automated activity monitoring and analysis, facilitating comprehensive logging of individuals and their actions into a database. Leveraging facial recognition technology and the Kinetics 700 dataset with the MMaction library in Python, the system identifies authorized personnel and recognizes human activities. Such advancements promise swift and accurate response measures during crises, reducing reliance on manual processes. The results of this study span several fields, including law enforcement, facility management, and urban planning, where the ability to track human behavior in real-time is crucial for ensuring public safety.