Urban Boundaries, Social Barriers: A Benchmark and Vision-Centric Framework for Mapping Gated Communities and Equity Implications

Minwei Zhao, Weiming Zhang, Jiawang Du, Qiming Liu, Weiming Zhuang, Pei Nie, Cai Wu
September, 2026
Cite arXiv Conference Code and data

Abstract

This paper introduces a benchmark of residential compounds in the Greater Bay Area and a vision-centred multimodal approach to identifying gated communities. Combining imagery, text, and structured attributes enables metropolitan-scale analysis of enclosure, green space, and pedestrian connectivity.

Type
Conference paper
Publication
European Conference on Computer Vision (ECCV 2026)

Accepted at ECCV 2026. The camera-ready version was posted to arXiv on 3 September 2026.

Cai Wu
Cai Wu
Assistant Professor

I study computational urban morphology, GeoAI, and data-driven urban design, developing reproducible methods for understanding cities across scales.