In 2017, ‘citizen scientists’ can help professionals with everything from searching for planets and stars to preventing traffic deaths by teaching computers to predict high risk road environments. The public can now help the city of Bellevue, the University of Washington, and Microsoft, who have teamed up on a new technology development project called Video Analytics Towards Vision Zero.
Vision Zero is an international movement which seeks to reduce traffic fatalities to zero by improving infrastructure so that collisions will not result in death or injury. With the Video Analytics Towards Vision Zero project, the public private partnership is looking to create a tool that will enable cities to detect and eliminate problem spots in their communities before they occur.
According to Microsoft Research Scientist Victor Bahl, “Computer vision algorithms applied to video feeds from traffic cameras have a huge potential of improving traffic flow and reducing traffic crashes and fatalities.” And here is where you come in; researchers are turning to crowdsourcing for help convert raw video into a powerful dataset which can be used to ‘teach’ computers to recognize vehicles, pedestrians and bicyclists.
Volunteers analyze existing intersection traffic camera footage in an effort to pinpoint ‘close calls’, which are more frequent than crashes and thus more useful in detecting systemic safety issues. The more video the project amasses, the more effectively researchers can utilize video analytics and machine learning systems together to make roads safer.
Along with a robust Bicycle and Pedestrian Planning program, Bellevue’s Smart Transportation commitment includes the goal of reducing collisions, and improving pedestrian safety is high priority for the community. Bellevue Mayor John Stokes stresses the importance of this work, “Everyone who participates in the Video Analytics project can make a small contribution that could add up to a big impact in keeping our walkers, bicyclists and drivers safe.”
To contribute to the project, follow this link for more information.