Facial Recognition Softwares Starting To Raise Eyebrows
April 2, 2019
The IoT Agenda of Getting to Know You
Royal Caribbean operates passenger ships, but a slow boarding process prevented its customers from cruising through check-in. Two hour-long security lines in the heat and humidity are not consistent with luxury travel.
The famous ocean-liner company did something about the disconnect by adopting biometric facial recognition. Passengers can upload a photo, along with their passports, from home. Check-in is a cool breeze as they’re recognized by a camera when they board. The technology was first rolled out in Ft. Lauderdale, FL, and Galveston, TX has just come online. Their senior digital VP is realizing his objective for passengers: “From the car to the bar in less than 10 minutes.”
Everybody is a Star
Previously, only the George Clooneys of the world were waved through lines. Now, we’re all potential celebrities in our own right, instantly recognizable for better or worse. While facial recognition is creating happier passengers, its reliability makes a compelling case for its security applications across private and government sectors.
A publication by the National Institute of Standards and Technology reported last year on an evaluation of the accuracy of algorithms in facial recognition:
The major result of the evaluation is that massive gains in accuracy have been achieved in the last five years (2013- 2018) and these far exceed improvements made in the prior period (2010-2013). While the industry gains are broad – at least 28 developers’ algorithms now outperform the most accurate algorithm from late 2013 – there remains a wide range of capabilities. With good quality portrait photos, the most accurate algorithms will find matching entries, when present, in galleries containing 12 million individuals, with error rates below 0.2%.
On Day #3 of its roll-out of a “cutting-edge facial comparison biometric system”, U.S. Customs and Border Protection intercepted an impostor at Dulles International Airport. A man from the Republic of Congo was traveling with an authentic French passport that did not belong to him. A face scan detected that the man’s face didn’t match his passport. He was detained and investigated before being allowed to leave the U.S.
If you’re responsible for security for a small to medium-sized company, or across multiple business channels in the corporate or government sectors, you’re facing growing challenges. Whether you work in loss mitigation, law enforcement, entertainment, hospitality, education (basically anywhere there are people and other assets), adopting face recognition technology (FRT) can provide an invaluable layer of protection.
Does Our Face Look Familiar?
ClearBlade’s Enterprise IoT Platform is recognizable as the only platform that fully deploys the entire software stack at the edge. Depending upon your security niche, you may be saying,
“So, what is this business about being at the edge, and what does it have to do with facial recognition?”
There are tangible performance (speed), and reliability differences between devices like cameras and biometric face recognition software that rely on data being sent back-and-forth between centralized servers or to and from a cloud platform.
If you’re a law enforcement professional out on a call, you can’t afford for your “app” not to work or to be operating at a snail’s pace. Neither can you rely on sent or received images of poor quality. The same is pretty much true regardless of your industry even though the “bad guys or gals” you’re filtering may range from shoplifters to those with an outstanding traffic warrant, to those who pose an imminent threat to life.
Keeping it Local
By moving computing to the edge, i.e. inside the device itself, latency is virtually eliminated along with potential failure due to the loss of a network connection, and quality is enhanced. Data is also more secure the closer it stays to “home”.
What if you’re in rural Texas, and there’s no connection? With edge, some systems are capable of operating offline.
Seven Hackers Hacking – One Safe iPhone
An example of edge computing and face recognition is the iPhone X where face scans using Apple’s sophisticated technology are stored on your phone’s “Secure Enclave” and never sent to the cloud. Apple has stated that “The probability that a random person in the population could look at your iPhone or iPad Pro and unlock it using Face ID is approximately 1 in 1,000,000 with a single enrolled appearance.” But, a group of people had fun trying.
Virtually no one except Apple has a budget of their size. And, you don’t have to. There are solutions at price points affordable for many companies, as the stuff of James Bond movies becomes a reality.
You may be tasked with security in retail, like convenience stores. Using FRT to identify bad actors when they enter a store can have obvious benefits. Conversely, you can use face recognition to tailor shopping experiences for your best customers in fashion apparel, accessories, groceries, cars and more.
Got an Evil Twin?
All bets are off as identical twins can stump FRT. But, a good system won’t mistake an armadillo in your swimming pool for a skinny-dipping human, or the neighbor’s dog for a package thief. (Unless he steals your box from Amazon).
If you’re interested in face recognition that works for you, we can help. We tailor appropriate solutions for your business model and existing technical infrastructure. That means working at the edge, in the (IoT) cloud platform, and more. The best solutions are typically not “either or”. The cloud is useful to aggregate lots of data from multiple sources or databases, and in performing analytics on large amounts of data.
Talk to us about IoT infrastructure, security, and application enablement platforms. We help you with your business today while planning for its growth with scalable solutions.