By Oppenheimer Funds. In recent years, technologies guided by data and information flow, and coupled with enhanced forms of our senses, are providing new ways of understanding ourselves. Emerging biometric and identity science technologies are using data to create new methods of analyzing our physical bodies and behavioral patterns, leading to novel ways of understanding ourselves and cracking problems once thought unsolvable. http://paidpost.nytimes.com/oppenheimerfunds/from-man-to-machine.html?tbs_nyt=2016-september-nytnative_hpstory-oppenheimerfunds-0930-1030?module=PaidPostDriver®ion=PaidPostMOTH&pgType=Homepage&action=click It’s also leading to new opportunities in myriad fields that previously had never imagined using such technology.
There are more than 17,000 tactile units in the hand alone, and that’s just a small portion of the sensory receptors that make up our sense of touch. Believed to be the first sense fetuses develop, touch is so important as a means of communication, establishing intimate bonds and orienting ourselves in the world, it can actually shape chemical development in the brain and have lasting physical and social effects.
Just as touch is vital for identifying our surroundings and the people in them, it’s also a key way for men and machines identify us. Fingerprint matching focuses on the minutia — the shape, endings and bifurcations of the ridges — which are so distinctive, not even identical twins share the same patterns. This extraordinary individuation has made fingerprinting (and fingerprint readers) a hallmark of law enforcement security and identification protocols, with technologies being used in areas spanning from artwork authentication and financial transactions at billion-dollar institutions, all the way down to elementary schools using fingerprint payment systems for students to buy lunch.
Anil K. Jain, a computer science and engineering professor at Michigan State University, and his Ph.D. student, Sunpreet Arora, made the world’s first 3D printed fingerprint two years ago. He says that the ability to create physical models of prints will help improve accuracy of fingerprint scanners. Hybrid or multi-modal identification technologies that combine traditional fingerprinting with other non-biometric and biometric identification methods are being adopted to improve accuracy and make the biometric systems more secure.
We have “a multitude of sensors (and) each one is capturing some incomplete and noisy information,” Jain says. “It’s the fusion of all this multisensory information which gives a more sure way of knowing who this person is.”
The ability to find friends, build communities and avoid foes is crucial to the survival of many species. So it’s not surprising that human babies recognize faces almost from the second they’re born. Research shows that infants mere minutes old can recognize facial patterns and, within a couple days, can pick out their mothers faces among a sea of others. By six months, children are so adept at facial recognition, they can identify individuals within an animal species more accurately than adults can.
Using some techniques that mimic how cells in the visual cortex process information, facial recognition algorithms are catching up to their human counterparts. Widespread in the biometrics world, facial recognition programs are used by an enormous array of organizations, from law enforcement agencies searching for suspects to bars and restaurants identifying loyal customers to mining companies detecting whether machine operators are alert or drowsy.
Facial recognition is endemic in next generation security protocols, and accuracy is improving as researchers gain access to broader data sets that can train algorithms for variations in poses, illumination, facial expressions, and image quality, says Damon L. Woodard, an associate professor in electrical and computer engineering and a faculty member at the Florida Institute for Cybersecurity (FICS) Research at the University of Florida.
“In the next five years, I think more emphasis is going to be on the capability of doing biometric recognition under less and less constrained scenarios,” he says.
Crucial for nonverbal communication and interpersonal bonding, eyes have evolved to be noticeable and emotive. Designed to visually stand out — humans are one of the only primates with high contrast between the iris and the white part of the eye called the sclera — eyes are rich information portals that babies instinctually seek out. One study published in Proceedings of the National Academy of Sciences found that newborns as young as two days old show preference for people who gaze directly into their eyes rather than away. There are chemical reasons behind that. Sustained eye contact releases oxytocin, a hormone associated with social bonding, which may explain the fuzzy feelings that come when gazing deep into a lover’s (or dog’s) eyes.
Iris scanning has an advantage over other biometric identification methods because irises don’t significantly change with age.
Eyes are valuable data hotspots for identification and security systems, too, in part because irises are far more distinctive than other biomarkers. Compared to 10-point DNA analysis, where the chance of two unrelated people having a match on 10 genetic markers and thus yielding a false identification is about one out of every billion (109), the chances of two people having identical irises is about drastically smaller (1078). Iris scanners have another major advantage over other biometric identification systems because, unlike faces and voices, irises don’t significantly change with age.