SENSING THE CITY

Leveraging geotagged social media posts and street view imagery to model urban
streetscapes using deep neural networks

 

 

This research project represents a prototype, utilizing the latest technologies in AI and digital twins to inform and enhance decision-making processes. The core of this study involves capturing and analyzing public sentiments from social media data, coupled with urban context, through deep learning-based image analysis. This dual approach allows us to create a comprehensive picture of the city’s dynamics, both in terms of its physical structure and the emotional landscape of its inhabitants.

 

The culmination of this project is a real-time visualization tool that portrays sentiment rankings within an urban environment. This tool, which forms part of a dynamic 3D digital twin model of the city, serves as a crucial decision-making aid for urban planners and authorities. By providing a direct, intuitive visualization of public sentiment, it enhances the user experience and enables authorities to make more informed, data-driven decisions. This approach represents a leap forward in creating cities that are not only efficient and sustainable but also responsive to the needs and emotions of their residents.

 

INFO

Aman, J., Matisziw, T. C., Kim, J. B., & Dan, L. (2022). Sensing Urban Streetscape: Leveraging geotagged social media posts and street view imagery to model urban streetscapes using deep neural networks. POST-CARBON, Proceedings of the 27th International Conference of the Association for ComputerAided Architectural Design Research in Asia (CAADRIA) 2022, 1, 595–604.

 

Public Talk link
Proceeding link
Github Codelink

To what extent does civil sentiment may get influenced by characteristics of the urban environment and how these relationships might change over time?

 

Framework Development

Taking our sentiment and urban context data, we then brought it into the realm of a digital twin. This is a dynamic 3D model of downtown Columbia, where we directly mapped sentiments onto this digital landscape. The result is an interactive, real-time visualization of sentiment rankings, which acts as a living, breathing digital twin of our city for more intuitive urban planning and management.

Steps of the Framework

Implementation of the framework for a research question>>>