The New York Times uses neural network to map every building in the U.S.

            <img src="https://archinect.imgix.net/uploads/bd/bdd5c5ca738fe43a8d46e606697deeac.jpg?fit=crop&auto=compress%2Cformat&w=1200" border="0" />The&nbsp;<em>New York Times</em> has <a href="https://www.nytimes.com/interactive/2018/10/12/us/map-of-every-building-in-the-united-states.html" rel="nofollow" >made a map</a> of every building in the&nbsp;United States. Using a <a href="https://archinect.com/news/tag/791676/neural-networks" rel="nofollow" >neural network</a> to analyze satellite imagery, the team's program then traced the shape of buildings across the country. Users can enter a city, zip code, or address, and explore these areas in detail.&nbsp;
It's pretty fun to play with, but the tool also helps illuminate the underlying patterns of cities and places that make them feel the way they do. Looking at these different patterns, we can detect the imprints of geology, the effects of suburbanization, the intentional designs of cities, and the traces of culture embedded within. By looking at the physical infrastructure, the maps tell us so much more about people’s connections, stories, and experiences relating to a specific environment.  Take a look here.

Leave a Reply