Facing the Future with Multi-modal Biometrics

Feb 14, 2017

By Anthony Gioeli


Many of us remember those stretch-of-the-imagination science-fiction movies and TV shows in which people were screened by facial and iris scans to access otherwise-off-limits locations.  While these futuristic depictions of technology have found success in some government and commercial applications, its use for consumers is still in its infancy.

By contrast, fingerprint authentication has been readily adopted on consumer flagship smartphones and laptop PCs. And the attach rates on mainstream and budget models are rapidly increasing due to a) today’s wider variety and lower costs for sensors, b) the convenience of a password-free user experience, and c) the rise in mobile payments globally.

What’s next is the combination of multiple biometrics to offer even greater convenience and security. Two or more modes of biometric authentication is a critical path forward for retaining a positive human interface – and for maintaining highly secure data and transactions.

Multi-factor authentication can already be found on a few devices but the user experience has not been perfected. Iris scanning offers excellent results but is slow and requires additional hardware costs. Voice recognition as a biometric is improving but not yet ready for mobile device use in noisy environments. Facial recognition is proven and reliable, and the necessary hardware is already on the device, whether smartphone or PC. This makes facial plus fingerprint recognition the best first step forward for multi-factor authentication.

A critical element in piecing together the solution is the multi-modal recognition algorithm. A smartphone or laptop designed with Synaptics’ multi-factor fusion engine software, for example, can authenticate its user through a fingerprint, a facial scan by the device’s camera or, depending on the required security needs, both. That user might normally use the phone’s fingerprint reader to authorize a low-value transaction, like buying a cup of coffee, but occasionally opt for facial authentication when the environment dictates, such as while wearing ski gloves. However, for a transaction of a more-sensitive nature – such as a high-value purchase, accessing a personal record, or transferring money – the device can be configured to require fingerprint authentication and facial recognition to complete it.

Plural IDTM, the Synaptics multi-factor fusion engine software, can manage any biometric including voice and iris, but with cameras already integrated on devices, the initial release is focused on facial recognition. What this means for smartphone, tablet and notebook PC makers is a path to enhancing and differentiating their products through an additional, native layer of biometric security. Those device makers have the flexibility to customize deployment of the fusion engine. Security policy setting examples include: fingerprint and face; fingerprint and optionally face; fingerprint or face.

How this new fusion engine functions, and the unprecedented level of security it delivers, is one more testament to Synaptics innovation.  Taking input from fingerprint sensing and facial recognition, the fusion engine combines their raw “scores” to verify – or reject – the user.  Each score must meet a minimum authentication threshold before authentication is achieved, and their combined score raises accuracy.

The fusion engine’s security features don’t end there.  Devices can be set to time-out, if, for example, a fingerprint is verified yet the required follow-on facial verification isn’t completed within, say, 20 seconds, the entire authentication process is denied. Additionally, both fingerprint and facial anti-spoofing tools are sensitive enough to factor in fake fingerprints and the users’ head movements or blinking eyes.

The end goal is persistent authentication allowing for effortless trust between humans and devices. Continual multi-modal biometric security, whether by touching or looking at a device, can counterbalance trust scores and the ongoing activity, or lack thereof, drives the ability to authorize user expectations ranging to mundane texting to sensitive banking.

For users, it provides exceptional ease of use, convenience and security in accessing and making transactions through their intelligent devices. They, too, may be able to choose which of the authentication modes are used, individually or in combination. Additionally, IT managers can set, according to company policies, the biometrics-based authentication required for device log-in and network access.

We envision Plural ID as central to next-generation user authentication and device security – and the reasons are compelling. The number of annual transactions made via mobile devices is expected to nearly double by 2019 – to more than one trillion! You can just imagine, then, the countless instances in which users will welcome having a choice among authentication methods, along with the peace of mind from knowing transactions are all the more secure through multi-modal verification.

Additionally, Synaptics’ Plural ID multi-factor fusion engine is backed by SentryPoint™ technology, offering OEMs a wide-range of unique and highly secure authentication features including Quantum Matcher™ fingerprint matching engine, Match-in-Sensor architecture, PurePrint™ anti-spoof technology, and SecureLink™ 256-bit AES with TLS 1.2 encryption. PurePrint examines fingerprint images using unique artificial intelligence technology to distinguish between fake and actual fingers.

And that should confirm that the future is now!