AI for Public Spaces

Public spaces are the backbone of urban life. They shape how people move, meet, and interact, yet we know surprisingly little about their use across cultures and over time. This project applies computer vision to large-scale video and imagery to quantify patterns such as walking speed, group formation, and social mixing.

Students will help build and test computer vision models, then use the outputs to compare public space usage across cities worldwide. Possible directions include cross-cultural comparisons, long-term change (e.g., shifts in street use by age and gender), and the design of new data collection pipelines. The goal is to build a global dataset to understand how public spaces are used, so we can better plan and design them.