Facial recognition for online stores: Q&A with Hoyos Labs CEO Hector Hoyos
The Dutch online pharmacy Koopjesdrogisterij has become the first online store to use facial recognition or fingerprint authentication for purchases.
The website’s new payment recognition method, which is based on Hoyos Labs’ 1U platform, replaces the use of credit cards or additional banking devices.
The 1U app provides the frontend customer experience for HoyosID platform’s identity authentication process, allowing Koopjesdrogisterij customers to take a selfie or conduct a four-finger scan to authorize purchases.
After downloading the app, users will be prompted to enter their credit card information, once only, and enroll their face and fingerprints. After this initial registration, users can simply authenticate their biometrics in place of entering usernames/passwords/credit card details.
Using the smartphone’s camera, 1U scans the customer’s face or fingerprints for a full range of unique features to authenticate their identity.
The four-finger scan is established using Hoyos Labs’ new biometric four-finger (4F) technology, which captures four fingerprints on a smartphone simultaneously.
BiometricUpdate.com had the opportunity to interview Hoyos Labs CEO and founder, Hector Hoyos, in which he discussed how the 1U and 4F platforms work, how Koopjesdrogisterij is using these technologies to authenticate its customers, and plans to implement these solutions in other online stores.
How did Hoyos Labs get involved with the Dutch online store Koopjesdrogisterij?
Hector Hoyos: Our local commercial representative in the Netherlands was contacted and they asked us if we would be interested in working with Koopjesdrogisterij to do this and we said, “Of course”.
How does the 1U facial recognition app work?
We provided two things. We provided facial recognition and we provided fingerprint authentication technology. 1U is not just an app, it’s a platform. At the heart of the platform is the Biometrics Open Protocol Standard (BOPS). In early 2014, IEEE came to Hoyos Labs and asked if we would be willing to contribute the work we had done with the Biometrics Open Protocol at that point because they believed that it needed to become the standard framework to enable the implementation of end-to-end biometric authentication platforms whether it was our devices or anyone else’s. So we contributed to them but we kept the IP, much in the same way that companies like Qualcomm have done with Wi-Fi. BOPS can link to any biometrics and link to any biometrics from any vendor so it’s not only us. We, in fact, have deployment in other companies with biometrics from other vendors. So at that point we implemented our face SDK and our 4FingerID (4F) technology.
At the end of the day, one of the things that Koopjesdrogisterij realized, along with many other companies, is that you just can’t give customers one biometric option. You have to have multiple or hybrid biometrics whether the customer wants to use face or fingerprint or eventually iris. It’s not the same thing for me to buy $10 worth of meds than for me to go and buy $2,000 worth of cancer drugs. There has to be a higher security threshold for the $2,000 purchase. And that’s when you require the combination of both face and 4F. If the user determines that they want to use face, but then they make an extra $2,000 purchase, the backend — BOPS — has the capability to automatically change that parameter and require a combination of both face and 4F.
Can you explain the liveness detection feature that is built into the technology?
Liveness detection on face is a traditional liveness detection. You have active liveness and passive liveness. Active liveness is when you’re asked to smile, or you raise your eyebrows, or you shake your head up and down, or you nod your head left and right — that’s active liveness detection. 4F has passive liveness detection. Now, the passive liveness detection is much much more secure and convenient; it’s transparent, the user is not asked to do anything. And it is highly secure because it has the benefit that you capture it with your rear camera on your mobile device and those cameras have LED. So we create a multispectral image. We don’t just flicker the flash once. Several micro-flashes are generated, and all this happens within a millisecond. This helps us do several types of analysis including depth analysis so that we know that it is a 3D image and that it’s not a video or photograph. It also allows us to look at blood pressure.
Is Hoyos Labs currently working with any other online stores to implement this technology?
We’re working with 27 online stores in 15 different countries right now. We expect to be in a significant number of online stores by the end of the first quarter of next year.
Are any of these online stores based in the U.S.?
No, they’re all overseas. We’re noticing that in the U.S., companies are afraid to compete against PayPal. So in the U.S. we find that the process is much more complex. The online stores in the U.S. are a lot more fearful of the established payment companies. If you go to Europe, if you go to Asia, if you go to Latin America, that’s not the case. They’re all totally the opposite. They’re much more aggressive, they want to put these solutions out there. It just takes a bit longer because we have to translate to the different languages to localize our technology.