Retail Executive

JAN-FEB 2018

Retail Executive is the trusted advisor to top retail executives from the industry’s most profitable retailers. We help retail executives succeed in their job role and grow their business via exclusive, actionable, peer-driven content.

Issue link:

Contents of this Issue


Page 21 of 43

TE CH KINKS Williams explains that while facial recognition tech- nology is the wave of the future, challenges exist with the technology. "First of all, determining the thresh- old number that really identifies a match is an issue," she explains. " What is that number that truly identi- fies a match? For example, when comparing photos of my sister and me, we have an 82 percent match. In for developers interested in facial recognition. Google has a similar offering. "These tools allow developers to quickly build and integrate computer vision and facial recognition technology into their applications," Wil- liams explains. "There are several cloud companies that offer services that allow you to plug facial recognition into your existing systems. So really it's a combination of internal development teams using cloud services al- ready on the market." R emember a f e w y e ar s a go , when the indus tr y w a s up in arms o v er whe ther c onsumer s w ould f ind per sonaliz a tion s tr a tegie s and ta c tics cr eep y? W ell, her e w e ar e , jus t a f e w y e ar s la ter with cus tomer s e xpec ting their shopping e xperienc e s to be per sonaliz ed. P arla y tha t thinking to f a cial r ec ognition. ENTER APPLE Facial recognition for retailers is one thing. But con- sider, for example, that Apple's new iPhone X, which includes Face ID (Apple's name for the technology), replaces Touch ID for unlocking the device and for mo- bile payment authentication. Just as facial recognition itself isn't new, smartphones with facial recognition software aren't new, as Samsung already uses early ver- sions of the technology. But, according to Apple, with Face ID, the chance a random person can unlock your phone using their face is 1 in a million (it's 1 in 50,000 for Touch ID). Plenty has been written about smart- phone facial recognition technology — but here's why it affects retail executives. Remember a few years ago, when the industry was up in arms over whether con- sumers would find personalization strategies and tac- tics creepy? Well, here we are, just a few years later with customers expecting their shopping experiences to be personalized. Parlay that thinking to facial rec- ognition. Yes, security and privacy concerns abound, but now that facial recognition will be in consumers' pockets, widespread acceptance is likely, and mobile payment is likely to experience adoption growth. "As a result of facial recognition and, now, Apple's Face ID, I definitely see mobile payment becoming the preferred payment option," Williams says. "The best way to en- hance security is through facial recognition — it's going to be the standard very soon." It doesn't matter whether you yourself are an Apple enthusiast. Some of your cus- tomers are. And, as facial recognition technology pro- liferates in consumers' lives via their phones, over time, they'll come to expect such computer vision technology from the retailers they frequent. that case, 82 percent seems like a high percentage, but in this case, it's too low because we're comparing photos of family members." And, so, the challenge remains — what must the match percentage be to be considered correct and secure? She states that image quality is also very important when matching a person, as image quality and poor light- ing can impact how well the matching algorithm works. "Next, consider the angle from which the picture was tak- en, as that can also impact the face matching algorithm on the back end," she explains. "It's very important to have multiple angles of a person — the front angle, the left an- gle, the right angle, etc., because the more image data you have on a person, the better the face matching algorithm will work. Some people may not realize that every time you upload a photo to Facebook, you're making Facebook smarter at identifying each person. Facebook uses facial recognition when it begins automatically tagging you and your friends when you upload a picture. Facebook has a 98 percent match percentage, which is due to the data it has access to, such as what people look like from different an- gles and in different settings. In short, the more data you have, the smarter the computer program becomes." Senior leadership teams making decisions about facial recognition technology have a lot to consider. But, per Williams, if retail executives are not considering comput- er vision and facial recognition as a part of their overall business strategy, they're already behind. Other compa- nies such as Amazon, MasterCard, and Delta, for exam- ple, are already experimenting with and actually using facial recognition in a production system. If you're not already thinking about facial recognition technology as part of your overall business strategy, it's time to start. R RETAILEXECUTIVE.COM JANUARY/FEBRUARY 2018 22 By E. Harris THE FUTURE IS FACIAL RECOGNITION EXECUTIVE Exclusive retail feature

Articles in this issue

Links on this page

Archives of this issue

view archives of Retail Executive - JAN-FEB 2018