Using computer vision to improve sidewalk maps

NYU Center for Data Science
2 min readApr 12, 2023


A tool developed by a team of researchers including CDS Joint Faculty Claudio Silva is expanding pedestrian infrastructure

CDS Joint Faculty, Claudio Silva

While cities are increasingly promoting pedestrian-accessible spaces, walkers are often left to figure out routes without reliable information on hand. This may soon change as a team of researchers has developed an open-source tool using aerial imagery and image recognition to map out sidewalks, crosswalks, and walkways.

The tool TILE2NET was presented in the paper “Mapping the Walk: A Scalable Computer Vision Approach for Generating Sidewalk Network Datasets from Aerial Imagery,” published in the April 2023 edition of the journal Computers, Environment, and Urban Systems. The project will help urban planners and policymakers expand pedestrian infrastructure and identify areas of improvement.

Among the team of researchers who authored the paper is CDS Joint Faculty, Claudio Silva. He is also a Professor of Computer Science and Engineering at NYU Tandon School of Engineering. His co-authors include Postdoctoral Associate at the MIT Department of Urban Studies and Planning Maryam Hosseini, Associate Professor of Urban Science and Planning and Head of the City Design and Development Group at MIT Andres Sevtsuk, Assistant Professor of Computer Science at the University of Illinois at Chicago Fabio Miranda, and Professor of Computer Science at the University of Sao Paulo Roberto M. Cesar Jr.

The team trained the tool on 20,000 aerial images collected from cities with comprehensive pedestrian maps — Cambridge, MA, Washington DC, and New York City — and can be readily applied to any metro area where the city’s aerial imagery is accessible. The project will improve information about the location and connectivity of urban sidewalks, as many urban spaces have networks of incomplete walkways. The data also fills in a gap that has held back the development of useful apps for pedestrians, wheelchair users, street vendors, and other groups of sidewalk users.

Infrastructure for pedestrians and cyclists has been expanding more than decades in the past. A major driver of the growing trend is lowering planet-warming emissions from the transportation sector due to climate change. “When cities talk about cutting carbon emissions, there’s no other way to make a big dent than to address transportation,” said Sevtsuk for MIT news. “The whole world of urban data for public transit and pedestrians and bicycles is really far behind [vehicle data] in quality. Analyzing how cities can be operational without a car requires this kind of data.”

The authors write the tool will also “contribute to over-due efforts to equalize data opportunities for pedestrians” particularly in cities lacking resources for more conventional mapping methods.

By Meryl Phair



NYU Center for Data Science

Official account of the Center for Data Science at NYU, home of the Undergraduate, Master’s, and Ph.D. programs in Data Science.