InSight – a Google Glass app that recognizes people by what they're wearing
By Chris Wood
March 8, 2013
A new system, known as InSight, aims to provide something a little more ambitious than facial recognition. The technology, which is part funded by Google, will work between Glass and a smartphone app and aims to let users spot their friends in a crowd based purely on what they're wearing.
The system is being developed by a team at Duke University in Durham, North Carolina and aims to extend augmented reality to humans. It works by creating a “fashion fingerprint” of a person's outfit, taking defining elements such as clothes, jewellery and glasses into account. This fingerprint is created by a smartphone app that takes pictures of the user during general use, producing a file detailing the spatial distribution of colors, textures and patterns of the user.
A user's fingerprint changes every time that they change clothes, assuaging obvious privacy concerns. InSight developer Srihari Nelakuditi discusses this, stating that “A person's visual fingerprint is only temporary, say for a day or an evening.”
Early testing for the technology has already taken place, with the team successfully identifying its target an impressive 93 percent of the time, even when the subject's back is turned.
The team has also suggested a number of less obvious uses for the tech, such as the ability to display a public message saying “looking to share a cab” above your head at an airport. The system could also work as an aid for those who suffer from a neurological condition known a face blindness.