A map that reveals patterns of arrangement of buildings

A dataset containing 125,192,184 computer generated building footprints in all 50 US states is the source for a New York Times’s map of every building in America.

Published on 12th October, this map represents every building in the US by a black speck, reflecting the built legacy of the United States.

The dataset was publicly released by Microsoft early this year. The company’s computer engineers trained a neural network to analyze satellite imagery and then to trace the shapes of buildings across the country.

DNN architecture: The network foundation is ResNet34. The model is fully-convolutional, meaning that the model can be applied to an image of any size (constrained by GPU memory, 4096×4096 in the case).

The map reveals patterns in the arrangements of buildings. Traditional road maps highlight streets and highways; here they show up as a linear absence. As a result, “… you can read history in the transition from curving, paved-over cow paths in old downtowns to suburban sprawl; you can detect signals of wealth and poverty, sometimes almost next door to each other.”.

In the south of New Orleans, it’s possible to notice the layout of buildings along a narrow spit of land on either side of a Louisiana bayou, which may reflect the imprint of the region’s history under France: “… “long lot” development, which stretched skinny holdings laterally away from important waterways. Geography shapes settlement, but culture does, as well.”
Buildings along Louisiana bayou

A Computer Vision and ML approach to understand urban changes

By comparing 1.6 million pairs of photos taken seven years apart, researchers from MIT’s Collective Learning Group now used a new computer vision system to quantify the physical improvement or deterioration of neighborhoods in five American cities, in an attempt to identify factors that predict urban change.

A large positive Streetchange value is typically indicative of major new construction (top row). A large negative Streetchange value is typically indicative of abandoned or demolished housing (bottom row).

The project is called Streetchange. An article introducing the article can be found here.

Reference:Naik, Nikhil, Scott Duke Kominers, Ramesh Raskar, Edward L. Glaeser, and César A. Hidalgo. “Computer Vision Uncovers Predictors of Physical Urban Change.” Proceedings of the National Academy of Sciences 114, no. 29 (July 18, 2017): 7571–76. doi:10.1073/pnas.1619003114.