this article first appeared on dignari.com
Biometric technology is sort of like the NBA Slam Dunk contest. Just when you think all measures of creativity and innovation have been exhausted, they surprise you again.
I sort of felt that way at this week’s ConnectID conference in Washington, D.C. I’ve been working in the biometrics field in some capacity over the last 20 years and there are times when industry innovation seems dormant. And then there are peaks in the industry where biometrics are blazing new frontiers and the excitement is palpable.
Even if this year hasn’t completely been on fire, there are still plenty of advancements happening.
Here are 6 trends from this year’s ConnectID conference that you should be aware of.
Facial biometrics have been around for a number of years but they have recently gained significant momentum due mostly to their ease of capture, improved and commoditized high-quality cameras (think of your phone), and the fact people’s faces are everywhere.
But, just like an A-List celebrity walking the red carpet, they may not fully deserve all the adulation they are receiving. Sure, face has improved dramatically from when I first started working with it in the late 1990’s but there are still inherent issues with the technology: lighting, pose, complex backgrounds, interoperability, etc. These challenges are certainly not insurmountable but they also can’t be fully ignored.
Even with that said, there is a tremendous amount of upside for face as the biometric of choice for both public and private sector solutions. As face-in-video, selfie solutions, and enhanced image processing advances, expect even more facial systems hitting the street.
Let’s face it. Biometric device design has traditionally been, boring, to put it kindly. Boxy, gray, metal, oversized, limited user feedback, and generally no “sexiness”. Fortunately, this has been changing over the last year or so.
Design is now being baked into the device as much as functionality. This is perhaps due to the increased acceptance of biometrics from the public and the desire to appeal to consumers across industries. As biometrics become more commonplace in people’s lives, marketing, user experience, and design gain importance.
There were a number of devices that stood out on the floor. Tascent’s Insight One face/iris unit looks like something straight out of Cupertino. NEC’s second generation NeoFace Express camera provides phenomenal features while still catering nicely to the end user. Even the grandfather of biometrics, fingerprints, received a makeover thanks to Jenetrics. Their TFT-based fingerprint device elegantly displays user feedback just under where the prints are captured. This approach is definitely bringing sexy back to the tenprint.
Just as facial biometrics are gaining steam, so too are the use of selfies as the probe image. It makes sense that systems integrators and biometric vendors would start to leverage the camera most of us have in our pocket.
Last year Mastercard introduced selfie pay and at this year’s conference, the state of Alabama showed how the selfie plays a vital role in tax refund fraud prevention. As our phone cameras continue to evolve and get better, expect for more face-based operations to occur from our phones.
Just when you thought you nailed biometrics you find out it isn’t enough. As is the case with any multi-layered security approach, biometrics should only play a part in your overall scheme.
The industry is fully embracing the idea of multi-modal biometrics and layering in contextual information on top of that to make each transaction that much more secure. Gone are the days where a system would grant access solely based on passing a single biometric check.
As attack vectors increase and biometrics become ubiquitous, layered security becomes even more important. Yes, biometrics can be wonderful, but they should not be the sole determinant of authentication and authorization.
It used to be that you would buy a biometric algorithm and you were done. Not anymore. Algorithms are constantly improving and most likely doing so on their own. As computer systems become more complex and the dynamism of cloud computing is adopted, algorithms are leveraging powerful platforms to constantly learn and improve.
Deep learning is more than just a buzzword in the biometrics field. Numerous vendors are starting to leverage the technique to continue to hone their algorithms to ensure the best possible match over time.
The ConnectID conference is billed as “an exploration of physical and digital identity in the 21st century”. Biometrics is, of course, a large part of identity exploration.
As part of that journey, we are seeing a significant shift in biometric use cases that drive away from the traditional security role. The technology is now being used for convenience, such as the recent announcement by Google that the Google Home device would be able to recognize up to 6 different unique users. Biometrics are also being used for safety, such as vehicle systems that can detect drowsy drivers.
It would have been nice to see some of these novel ways in which biometrics are being used as part of the ConnectID conference. However, with the conference being held in Washington, D.C. and with a heavy federal presence, there’s no wonder it focused mainly on government implementations.
Beyond the trends listed above, there is also a continuance of some of the other well-known recent advancements in today’s biometrics, including the desire for contactless, on-the-move, and non-intrusive capture as well as cloud-based matching.
There’s no doubt that the industry is seeing significant growth and adoption beyond typical clients and operational models. This in turn is forcing vendors to meet the demand and move beyond the traditional in order to deliver the remarkable. These are indeed exciting times for biometrics and the quest for the next slam dunk solution is well underway.