Liu Liu
Liu Liu is a Ph.D. candidate in Urban Science at MIT’s Department of Urban Studies and Planning, where he works with Professor Andres Sevtsuk at the City Form Lab.
Sidewalks are where urban life actually happens. They are stages for social mixing, corridors for low-carbon mobility, and environments that directly shape public wellbeing. Yet, compared to roadways, sidewalks remain vastly undermeasured. Liu’s research fills this gap by treating sidewalks as social ecosystems rather than mere infrastructure. Using street-level imagery and vision-language models, he develops scalable frameworks to detect not just where people walk, but how they move together and interact across hundreds of American cities.
Beyond his dissertation, Liu has applied these multimodal methods across diverse contexts. As a PhD Applied Scientist Intern at Amazon AGI, he designed large-scale benchmarks for audio and reasoning evaluation. As a Visiting Researcher at UCLA’s Vision and Autonomy Intelligence Lab, he contributed to sidewalk perception systems for autonomous delivery robots. Before MIT, he founded CitoryTech, an urban AI studio that led over thirty projects for clients including the World Bank, Daimler, and various municipal governments.
His work has appeared in AAAI, Nature Cities, Cities, and ECCV, with over 1,500 citations. His research was exhibited at the 2025 Venice Architecture Biennale, and he has consulted for organizations including the World Bank and Tencent.
For more details on my research and publications, please refer to my CV and website