COSMOS Testbed Response to COVID-19 Pandemic
Video
Team Information
Team Members
Mahshid Ghasemi, PhD Candidate, School of Engineering and Applied Sciences (SEAS), Columbia Engineering
Faculty Advisors:
Gil Zussman, Professor, Department of Electrical Engineering, School of Engineering and Applied Sciences (SEAS), Columbia Engineering
Javad Ghaderi, Associate Professor, Department of Electrical Engineering, School of Engineering and Applied Sciences (SEAS), Columbia Engineering
Zoran Kostic, Professor of Professional Practice, Department of Electrical Engineering, School of Engineering and Applied Sciences (SEAS), Columbia Engineering
Abstract
Smart city intersections will be at the core of an AI-powered traffic management system for crowded metropolises. COSMOS deploys a variety of infrastructure sensors, including street-level and bird’s eye cameras, whose data will be aggregated by the servers. The servers will run real-time algorithms to monitor and manage traffic. COSMOS technologies can also help us combat the coronavirus pandemic, better understand the impacts of social distancing protocols on people's daily life, and determine how well they follow the unprecedented rules. We utilized these infrastructures to design a fully automated multi-stage social distancing analyzer pipeline that monitors the distance between pedestrians and decides whether or not they are maintaining a proper distance. This pipeline removes all the distortions caused by the cameras and converts pixel distance to on-ground distance with high accuracy (less than 10 cm error). The pipeline is also capable of detecting groups of people walking together and exclude them from social distancing violation. We applied this pipeline on videos recorded from COSMOS pilot site and analyzed how social distancing protocols are impacting people’s social life. The results show that after COVID-19 crisis, only 10-23 % of people do not comply with the social distancing rules, also only around 10 % of pedestrians tend to walk as a group. This work is an initial step toward overcoming the challenges of potential deployment of autonomous vehicles, including a large number of vehicles moving at various speeds, obstructions which are opaque to in-vehicle sensors, and chaotic behavior of pedestrians.
Contact this Team
Team Contact: Mahshid Ghasemi (use form to send email)
