Mobile security: A mother lode of new tools

You may know your smartphone, but pretty soon the question will be: How well does your smartphone know you?

1 2 3 4 Page 3
Page 3 of 4

The sensors have anti-spoofing features that look for the characteristics of real skin so as not to be fooled by, say, latex, he adds.

Vendors of other types of biometric systems are also jockeying for position in the mobile market. Their products verify identity using other characteristics of the human body, such as the whites of the eyes, which, according to Reza Derakhshani, chief scientist at startup EyeVerify, "are the only parts of body with exposed and easily seen vasculature."

Kansas City-based EyeVerify has patented authentication technology that requires a user to hold a smartphone 6 to 10 inches from his face to capture images of his eyes and the unique patterns of the veins. Derakhshani says the technology is as accurate as fingerprinting. Various means of detecting "live-ness" -- such as evaluating the movement, curvature and reflective properties of the eyeball -- prevent spoofing, he says. The technology can identify people even if they're suffering from allergies, colds or hangovers, he adds, because, while the vessels become engorged under those conditions, their patterns don't change.

Another part of the eye that's used for biometric identification is the iris, which controls the size of the pupil. Each iris has 240 unique characteristics that can be mapped for authentication, producing a false rejection rate of 1 in 2.25 trillion, says Anthony Antolino, head of marketing at EyeLock, a New York-based vendor that developed a handheld iris recognition system called Myris.

The drawback of iris authentication is that it requires infrared light to capture the details of the iris, and that means iris recognition tools need a special light source and a special camera. EyeLock's device, initially aimed at the desktop environment, costs less than $300, says Antolino, adding that he expects to see the technology built into mobile devices in the near future

Voice recognition technology is another option. One vendor in this niche of the biometrics market is Madrid-based Agnitio. With Agnitio's technology, the user says a passphrase three times, and the software learns his identity by analyzing his vocal anatomy, explains Agnitio vice president Mike Goldgof. The technology is now in its fourth generation, and its false acceptance and rejection rates are both about 1 in 1,000 -- though it may not work in noisy environments, he says.

Attempts to spoof the system with a recording of a user's voice can be detected 97% of the time, because the vocal range of a recording is typically narrower than that of the original voice, according to Goldgof. Live impersonations are even easier to detect -- "people are not that good," he says. And the system can recognize users even if they have colds, as long as they can speak above a whisper.

The other major biometric ID option is facial recognition technology. But Goode says these systems can be spoofed with pictures. And Alikhani adds that he's seen them "do well in environments that were well lit, and fail miserably in darkness."

1 2 3 4 Page 3
Page 3 of 4
It’s time to break the ChatGPT habit
Shop Tech Products at Amazon