Aura Vision Labs demonstrates retail tracking system

Aura Vision Labs, a startup based in the University of Southampton’s Future Worlds incubator, has come up with a solution that analyses CCTV footage to track people in real-time and detect the gender, age and clothing style of every person in a crowd. Its cloud-based platform was first demonstrated to a global audience at last month’s Consumer Electronics Show (CES) in Las Vegas.

Machine learning specialists Daniel Martinho-Corbishley and Jaime Lomeli are building Aura Vision Labs on PhD research undertaken on the South Coast of England. “There are over 245 million professionally installed video surveillance cameras in operation globally, but only one per cent of this footage is currently being analysed,” Martinho-Corbishley explains. “Our technology harnesses this fantastic source of video data, working with low-cost cameras and existing security systems to provide retailers with a simple, cost-effective way to gain more accurate, in-depth and real-time information than has ever been possible before.”

Loyalty cards have for many years been the primary solution for analysing customer spending habits in-store, he adds, but invasive customer signups and waning popularity means many purchases are not tracked. “With this novel technology, retail marketers could go beyond simply counting customers,” Martinho-Corbishley comments. “Now they can capture 100% of in-store visitor demographics, dwell spots, walk-ins and walk-bys; this enables businesses to improve conversion rates, act on missed opportunities and target specific customer segments in unprecedented detail.”