A new Google Glass service can detect the gender, age and emotions of people's faces as they appear in photos or video taken by Google's smart glasses. But the service avoids Google's recent ban on facial recognition technology by not connecting faces to personal identities.
The "ReKognition" service launched on June 12 by offering the tools to build apps for reading faces in photos without compromising individual privacy. Such a move suggests that Google's decision to ban facial recognition for Google Glass represents only a speed bump on the highway toward a future world where mobile devices can automatically detect everyone and everything in view.
"We believe the ban on facial recognition is going to be temporary," said Ning Xu, CEO of Orbeus Inc. "The simple reason is that the technology is useful."
But Orbeus launched its ReKognition service with greater ambitions than just enabling facial recognition. The San Francisco-based company has built software capable of recognizing faces, objects and background scenery — a step toward enabling computers and smart devices to better interpret the world as seen through photos and video.
"We want to build a one-stop shop solution for developers to build into their apps that can understand the data behind pictures and videos," Xu told TechNewsDaily.
The "computer vision" technology offered by Orbeus may prove indispensable for humans trying to make sense of a world saturated by multimedia. People already upload 350 million photos to Facebook every day and 100 hours of video to YouTube every minute. The popularity of smartphones and new devices such as smart glasses have led to an explosion of online media content by putting cameras in everybody's hands.
Some facial recognition or detection capabilities already exist in popular online services such as Facebook and Google's Picasa Web photo albums. Orbeus offers similar services for a wide range of clients, including an online photo album provider that automatically sorts pictures into personalized albums according to people's faces.
One Asian dating site uses Orbeus' services to screen the faces of its users according to age, gender and even beauty. Another client creates ad displays that automatically profile the faces of shoppers to target them "Minority Report" style. (For instance, a 25-year-old female might be presented with a cosmetics ad.)
Sophisticated computer vision algorithms have already existed for about a decade, Xu said. But the rise of "cloud computing" — computer services offered remotely through an online connection — has allowed companies such as Orbeus to offer computer vision technologies to just about anybody with a mobile device connected to the Internet.
"Today with the development of cloud computing it's very economical," Xu explained. "It's getting cheaper and cheaper to deploy computer vision algorithms."
Such trends mean that ordinary people can expect to soon get their hands on facial recognition technology through mobile devices and wearable gadgets — as long as the privacy concerns don't outweigh the expected benefits.
"We believe the technology can be used without revealing any personal identity information," Xu said. "If there are any Google Glass killer apps, computer vision tech has to be in it to make it a completely new user experience."
You can follow TechNewsDaily Senior Writer Jeremy Hsu on Twitter. Follow TechNewsDaily on Twitter, or on.