What makes Paris stand out? It’s not just the Eiffel Tower as the devil is in the detail according to researchers. They have developed a program that helps bring to light “minor” visual differences in a city’s winning sights and architecture.
Researchers at Carnegie Mellon University and the INRIA laboratory at the Ecole Normale Supérieure in Paris have created visual data mining software that can detect “minor details”, such as street signs, streetlamps and balcony railings that give Paris part of its charm. Other cities can also be scrutinized in a similar way.
The software has already analyzed more than 250 million visual elements obtained from 40,000 Google Street View images of Paris, London, New York and Barcelona to finally get the ones that actually distinguish one city from the other.
Among the highlights unique to each city were cast-iron balconies in Paris, fire escapes in New York City and bay windows in San Francisco.
The discovered “star elements” can be useful for a variety of tasks, from mapping architectural patterns and influences across several cities, to finding the characteristic elements of different continents.