Cannot spot your friends in the market? Here’s Google glass for you. It will spot their faces through a human recognition system, even if their faces are not clear. Designed for Google’s upcoming gadget, Glass headset, it recognizes people by the clothes they are wearing and you can see the name written on the headset’s video.
The system called InSight, which is partly funded by Google, will help you find friends and also help to get yourself spotted in bustling market places viz., shopping centres, sports stadiums and airports. Face recognition systems cannot be used for this, says InSight developer Srihari Nelakuditi at the University of South Carolina, Columbia, it is unlikely someone in a crowd will be looking straight at a headset’s camera. So Nelakuditi in collaboration with Romit Roy Choudhury and colleagues at Duke University in Durham, North Carolina, developed a recognition system based on a ‘fashion fingerprint’ of a person’s outfit, from their clothes to their jewellery, badges and glasses.
A smartphone app constructs this fingerprint by clicking a series of photographs of the user as they are reading web pages, emails or tweets. After this, a file called a spatiogram is created. The spatiogram also captures the spatial distribution of colors, textures and patterns (vertical or horizontal stripes, for instance) of clothes that the user is wearing. The analysis of this color-texture-pattern combination makes it easier to spot someone over long distances or at odd viewing angles.
The unique selling point of Google Glass is that it protects privacy because the fingerprint changes every time you change your clothes. This way, you can be anonymous whenever you wish. ‘A person’s visual fingerprint is only temporary, say for a day or an evening,’ says Nelakuditi.
In early tests, 15 volunteers were used and it was found that the team identified people 93 per cent of the time, even when they had their backs to the headset user. Matching data from the phone’s motion sensor with motion in the Glass field of view will boost accuracy. ‘There are a lot of personal characteristics that make us unique,’ says Mark Nixon, a biometrics specialist at the University of Southampton, UK. ‘Clothing and movement are highly related to gait – and gait has been shown to be unique.’ The system can be used by someone who wants to attract attention to themselves and their CV at a job fair, or outside a stadium where they are selling a spare ticket, the team says.
It could also help people with face blindness, a neurological disorder that makes it impossible to recognize others.