You can see the subtle changes when a neighborhood is on its way up—streets get cleaner, building facades improve, new businesses start moving in. Across an entire city, however, it’s harder to track such changes, to understand in real time which neighborhoods are improving and which may need an extra boost in development.
A new tool by Harvard and MIT researchers promises to do just that, with the help of machine learning and the biggest library of urban images on the Internet: Google Street View. By using digital tools, researchers were able to assess long-held beliefs about urban improvement and offer new methods to city planners and real estate developers looking to identify areas in need of improvement.
The collaborators reveal their findings in a May 2017 paper in the Proceedings of the National Academy of Sciences, Computer Vision Uncovers Predictors of Physical Urban Change. The paper was written by Nikhil Naik, a Prize Fellow at Harvard University; Scott Duke Kominers, the Harvard Business School MBA Class of 1960 Associate Professor; Edward L. Glaeser, the Fred and Eleanor Glimp Professor of Economics at Harvard University; and MIT Media Lab professors César A. Hidalgo and Ramesh Raskar.
Read original article

