CASA Research Measures Urban Inequalities using Deep Learning

Dr Esra Suel’s research aims to enhance our ability to make quantitative measurements at high spatio-temporal resolution by using emerging sources of imagery data. The research develops capabilities that augment the temporal resolution of our measurements – an improvement which is crucial for informing policymakers, ensuring that timely data is available to track whether interventions result in expected outcomes in the built environment. 

Currently, the capacity to monitor changes at a granular spatial and temporal level, even in data-rich urban areas, is limited. Traditional data sources such as census are limited to measurements every 10 years, and smaller scale surveys suffer from limited sample sizes, thereby constraining our capacity for city-scale measurements and tracking. Emerging sources of large-scale data, including images captured by street-level cameras and satellites, potentially offer information on urban change. They also capture information on a unique point of view – enabling the development of novel exposure metrics, which can then be used to study links of built environment to health outcomes. 

The research received funding from MRC and Wellcome Trust. It is also receiving ongoing project funding from the National Institutes of Health (NIH) to explore links with image derived features of the built environment and health outcomes, and environmental inequalities and how they have changed over the past 10 years in the US. Further funding has also been received from the Swiss Data Science Centre for work to investigate change detection in Sub-Saharan Africa as well as within urban streets. 

More information

• Visit Dr Esra Suel’s profile
• Explore more research at CASA

Recent papers from Dr Esra Suel

• Do poverty and wealth look the same the world over? A comparative study of 12 cities from five high-income countries using street images
• Self-supervised learning unveils change in urban housing from street-level images

Image: Changes in London's housing between 2008 and 2021, as captured by self-supervised learning using street-level images.