trinamiX’s new solution adds skin detection capabilities to biometric authentication.
I’ve been following trinamiX ever since I first met with the company back in 2018 at Mobile World Congress Barcelona. Back then, it was a somewhat newly formed subsidiary of BASF, spun-off in 2015. The company refers to itself as an innovative sensing technology company, and trinamiX’s Founder and Managing Director, Dr. Ingmar Bruder, is himself a physicist with a background in organic solar cells. The company’s four key areas of expertise are IR detectors, NIR Spectroscopy, 3D imaging and Distance Measurement—all of which are relevant to today’s technology industry needs. About 1/3 of the company’s employees (34 of 104) are doctorate holders, which likely gives the company a leg up in finding solutions to industry problems.
Challenges of facial recognition
Recently, around what was supposed to be the Mobile World Congress Barcelona timeframe (the event was canceled due to coronavirus concerns), the company made a series of announcements and did a press tour to show off its 3D sensing capabilities. 3D sensing is one of the key businesses within trinamiX that the company is seeking to grow. By the looks of it, the company has a pretty compelling solution. The primary use case for 3D sensing today is the ability to do perform facial recognition for biometric authentication. Currently, many facial recognition solutions either only take a 2D image of the face (the least secure solution), or they combine a 3D depth map with a 2D image to ensure that it isn’t just a photo of someone’s face. However, with this approach, face masks could be used to replicate someone’s facial structure. TrinamiX has a solution that deals with these potential risks.
trinamiX adds another factor
The company’s unique beam profile analysis solution adds a third aspect to the biometric authentication equation—live skin detection—adding yet another layer of security. This live skin detection algorithm is an extension of the company’s beam profile analysis algorithm, which enables it to tell the different physical properties of materials and the different ways they reflect the IR waves. This enables trinamiX to determine if a face is made of actual skin (and not a mask). This is not just a hardware solution, but rather a complete solution that runs a set of proprietary patented algorithms on an SoC. Part of trinamiX’s announcement (originally planned for MWC) was a partnership with Qualcomm to run the algorithms quickly and efficiently on Qualcomm’s Hexagon processor inside the Snapdragon SoC. In addition, trinamiX joined Qualcomm’s Software Accelerator Program, which should give smartphone OEMs easier access to trinamiX’s algorithms and enable them to more easily integrate them into their devices.
However, trinamiX’s solution is not just a smartphone application; it can also be used in laptops. More and more manufacturers are adopting facial sign-on with features like Windows Hello. This technology could even be used for securing military-grade access control systems, since it requires a live person to be present. Additionally, this same material detection technology could be used to determine the difference between fingers, hands and objects behind them. So, by combining the 3D data from the infrared sensor and the material classification, an AR or VR headset could use an outward-facing infrared sensor to do hand-tracking (an increasingly sought-after feature in XR). Qualcomm seems to believe that this feature is important—it integrated an outward-facing infrared sensor into its XR2 reference design specifically for hand-tracking.
From what I’ve seen, I believe that trinamiX’s 3D sensing technology with live skin detection is a very promising solution. I believe that the applications are probably broader than the company itself even realizes. Also, since the algorithms don’t require any additional hardware from OEMs and already work with Qualcomm’s SoCs, this technology is fairly easy to implement. Many smartphone OEMs already have limited board space and rock bottom BOM costs—if you can add features to a phone or a PC without, in some cases, adding any hardware, that’s already a win in my book. We didn’t get a concrete date for when this technology will be in smartphones, but key partners should be receiving development kits shortly. This means there’s a chance we could see it as soon as next year.
Disclosure: Moor Insights & Strategy, like all research and analyst firms, provides or has provided research, analysis, advising, and/or consulting to many high-tech companies in the industry. The author does not hold any equity positions with any companies cited in this column.