V-DS logo Visual Data Science Lab FGV EMAp · Rio de Janeiro

Urban Mobility

How people and vehicles move through cities, from taxi trips to intersection cameras

Urban Mobility develops visual systems to analyze how people and vehicles move through cities. The line began with the analysis of New York City taxi trips (TaxiVis) and vector-field models of traffic dynamics, continued with the exploration of traffic jams (JamVis), and now focuses on sensing urban intersections directly: UrbanClipAtlas retrieves events and scenes from long street-level videos with vision-language models, and our 3D reconstruction work recovers pedestrian pose, trajectories, and walking speed from roadside cameras to support traffic-safety assessment. This work is developed in collaboration with the VIDA group at New York University.

Urban Mobility

Researchers

In collaboration with

Publications