Perceiving exposure segregation with open urban imagery

2026-06-19

AI Summary

This study investigates how the built environment influences daily social interactions across income groups through its physical characteristics, thereby driving social exposure segregation. To this end, we propose the VISAGE framework, which uniquely integrates interdisciplinary sociological theory into an interpretable multimodal vision model. Leveraging open-source satellite and street-view imagery, we construct a visual codebook to identify key physical moderators—such as defensive architecture and single-use zoning—that shape social dynamics. Empirical analysis demonstrates that built environment features significantly explain mobility-derived patterns of social segregation (Pearson r = 0.770). Furthermore, the findings reveal that mixed-use infrastructure fosters greater social interaction, while inclusive housing policies exhibit a strong visual association with heightened socioeconomic mixing.

Abstract

Socioeconomic exposure segregation -- the lack of daily interaction between income groups -- erodes social capital and entrenches inequality, yet the specific physical features that drive these behavioral restrictions remain poorly understood. Prior research has quantified where segregation occurs using mobility data, but has not identified how the built environment facilitates or inhibits these interactions. Here we introduce VISAGE, a large multi-modal model-enabled framework that perceives exposure segregation directly from open satellite and street-level imagery across 10,030 communities in 31 U.S. cities. Moving beyond black-box correlations, we operationalize cross-disciplinary sociological theory into an interpretable visual codebook to detect physical regulators of social mixing. We find that the built environment encodes a legible grammar of segregation: "defensible" architectural forms (e.g., fences, gated enclosures) and monofunctional zoning systematically predict higher social isolation, whereas mixed-use infrastructure fosters interaction, explaining substantial variance in mobility-derived segregation patterns (Pearson $r=0.770$). Crucially, we show that inclusionary housing policies manifest in distinct visual signatures associated with higher mixing, suggesting that policy interventions successfully alter the physical landscape to encourage diversity. Our findings offer a scalable pathway to decipher the social production of space, providing a mechanism-based lens to understand how the built environment shapes social behavior.

Problem

• exposure segregation
• built environment
• social interaction
• urban imagery
• socioeconomic inequality

Innovation

• VISAGE
• exposure segregation
• built environment
• visual codebook
• multi-modal modeling