Prediction through Description

Traffic, and more generally, mobility are important topics for densely populated cities around the world. Traffic congestion causes negative social and environmental effects, including air pollution. Improved traffic prediction is of great social and environmental value. Superior predictive models will allow capturing the simple implicit rules underlying complex traffic systems and empower the planning of smart cities and efficient road networks.

Our Traffic4cast competition challenges participants to understand complex traffic systems to predict their future states. Our traffic map movies encode the static (road maps) and dynamic (traffic volume and speed) data split into spatial cells. The dynamic data are derived from the GPS trajectories of a large fleet of probe vehicles, made available by HERE Technologies. The unique industrial-scale dataset provides a high resolution privacy preserving view of urban traffic. The Traffic4cast competition series is presented as a part of the NeurIPS conference.