Automated Human Tracking: An AI based Modular Approach

This paper has proposed a novel end-to-end automated system for the task of human tracking. Based on the research carried out, a modular approach has been provided for implementing such a system. The proposed solution comprises of Deep-learning (specifically Face recognition), Database, Path Finding, UI and Bio-mapping concepts. The primary source of data assumed are CCTV Footage. The dataset used for our experiments were collected by comprising a single image and few videos of a several people. This research work attempts to provide the recent location history and geographical location of a person given just a “single” image of that person. Given how important human tracking is to law-enforcement agencies, businesses and healthcare systems, it is important to improve the efficiency of the current tracking system which heavily relies on manual intervention. Thus, this research work proposes an End-to-End automated human tracking system ’God’s Eye’ as a tool, which can help in providing rapid and accurate results by reducing human involvement. God’s Eye aims to be a robust, low-cost, hardware-independent and seamless solution for human tracking in the present times. Implementing such a system at scale can be widely beneficial to the society and it is considered as a step in the direction of building a smarter & safer cities.