Image recognition software is used for all sorts of things, from tagging people in photos to security surveillance to identifying species. Now, researchers are attempting to tweak those algorithms to recognize hipsters, goths and other "social tribes." The Financial Express explains the motivations behind designing such a platform:
An algorithm able to identify people's urban tribes would have a wide range of applications, from generating more relevant search results and ads, to allowing social networks to provide better recommendations and content.
Researchers designed the model to recognize people who identify themselves as belonging to biker, country, goth, heavy metal, hip hop, hipster, raver and surfer subcultures. The computer analyzes images by breaking humans down into six different sections and categorizing attributes such as haircut, makeup, accessories, tattoos and clothing. It also takes color and texture into account.
So far, the team has only achieved 48 percent success on initial trials with hundreds of images. But they're already finding ways to improve the algorithm's accuracy, such as analyzing photos of groups of friends rather than individuals. Although preliminary, these efforts hint at a future in which personalized ad campaigns know whether to flash a pair of black platform boots or a single-speed bike when you walk by.
More from Smithsonian.com: