Social Distancing Detection

Abstract

Social distancing is critical for many diseases including COVID-19. Researchers focused on getting a robust model for social distancing violation using Artificial Intelligence and they have made remarkable progress in achieving the robust cascaded algorithms for finding the violations. The challenge which still inconvenience the application of such algorithms is the scalability and practical deployment of such models for business use cases. Since the algorithms are universal it becomes very much generic so tweaking it for a particular business use case has not been successful so far by using only the software as in many cases the environment, identity and angle of camera calibration varies. This project is designed to help researchers and scientists using Artificial Intelligence to make a complete software ready for production and deployment in factories and warehouses based on the results achieved through various experiments. Furthermore, this project would help to store the identity and track the people who violated the social distancing over the period and generate alerts.