Wednesday, March 4, 2020
A 3D imaging and infrared detection sensor company, TrinamiX, has revealed a technology that helps overcome facial recognition cheats. The company is working with Qualcomm Technologies to implement it on to both high-end and low-end smartphones.
Trinamix was established in 2015 as a wholly-owned subsidiary of chemical company BASF, though it operates independently. The company said it can classify and identify materials based on their physical properties using its patented beam profile analysis technology, which enables the detection of live skin. That capability is intended to prevent users fooling smartphones into unlocking via a mask, a high-resolution print or even a 3D rendering of an owner’s facial features.
Its 3D sensing systems determine distance from beam profile properties and work with wavelengths from visible to infrared. It extracts three data streams from a single camera system: a 2D IR image, a 3D depth map and material classification. While its latest announcement targets enhanced security for smartphones and consumer products, TrinamiX said its technology is already used in bin picking (in factory and industrial environments), robot vision, and object recognition.
In a recent briefing with EE Times in London, the managing director and founder of TrinamiX, Ingmar Bruder, told us, “This is a completely new secure solution that works with facial recognition technology across all platforms and industries, on both Android and Windows.”
He explained the background. “There are currently basically two ways of capturing depth: time-of-flight (ToF), and stereo, including structured light. We have figured out a third way, called beam profile analysis, for which we have several patents and own all the intellectual property. Our technology is based on back-scattering properties, creating a dense depth map using 40,000 dots for which we can measure the distance for each.” He added, “Our technology is based on three components: a NIR camera module, flood illumination, and a light projector. This can sit on top of any facial recognition capability.”
As part of its announcement, TrinamiX said it has joined the Qualcomm Software Accelerator Program to enable its patented technology for “live skin” detection as the new gold-standard for facial recognition across mobile devices. Through this collaboration, OEM customers can use the TrinamiX beam profile analysis technology to extract the three data streams from a single camera system. Material classification is provided using TrinamiX proprietary algorithms running on the Qualcomm Hexagon processor on Qualcomm Snapdragon mobile platforms.
The company said this ability to “sense” live skin is a new approach to security which can now be combined with any standard third-party facial-recognition algorithms. This enables secure and robust facial recognition in devices ranging from smartphones to laptops and up to securing access control systems.
Depth measurement through a monocular system
The ability to classify and identify materials based on their physical properties is a result of TrinamiX research and development on sensor systems while operating as a wholly owned subsidiary of BASF. In 2011, its scientists working within BASF made an unexpected discovery while characterizing under-development opto-electronic materials. The discovery, at first considered a bug, turned out to be an eye-opening feature that enabled instantaneous depth measurement through a monocular system. As TrinamiX evolved, it placed great emphasis on not only foundational research into both chemistry and physics, but also into algorithmic design for 3D imaging, cloud-based spectroscopic services and fiber optic-based distance measurement across multiple industries.
In 3D imaging, the beam profile analysis technology turns a standard camera and light projector into an imaging system capable of sensing through a single camera. Beam profile analysis adds a dimension of reliability to both object- and user-recognition by extracting three data streams from one system: a 2D image, a 3D depth map and material classification capability. The analysis is done with proprietary algorithms that run on a multitude of different platforms, such as the CPU or GPU of a PC or the DSP of an embedded compute unit.
In the bin-picking application, a robot can train itself to identify a box or a container, based on the technology’s capability in identifying the material — hence it can define the boundaries of a box or container on the fly.
“You are training the system based on the actual materials. In robot picking, you don’t need to pre-define the boundaries, as the system can look for the material and the boundaries. It can also sort materials, for example in recycling,” Bruder said.
In addition to industrial applications, the company is looking to enter the consumer market on existing hardware and software, he said.
“One of the first products we are likely to see based on this technology is robot vacuum cleaners,” he said. “This is in evaluation at the moment, but products are likely on the market in 2021. Meanwhile, in consumer products, we will see both high-end and low-end smartphones using beam profile analysis in early 2021.”
The company has developed an entire suite of different sensor technologies. With a team of over 100 experts across different fields ranging from algorithmic design to chemistry and optics, TrinamiX customers and partners include Qualcomm, Siemens, and BASF.
Manvinder Singh, vice president, product management, Qualcomm Technologies, Inc., said, “We are very excited to be working with TrinamiX and look forward to TrinamiX getting this novel technology into the hands of customers at the earliest possible opportunity.”
Bruder added, “TrinamiX has distinguished itself by providing the world’s most innovative 3D imaging solution using our beam profile analysis technology, and we are thrilled to be working with Qualcomm Technologies, the world’s leading wireless technology innovator. Using our patented 3D imaging technology, the solution will enable mobile devices powered by the Hexagon Processor inside Snapdragon mobile platforms to achieve a previously unattainable goal — the ability to sense live skin as part of a secure facial recognition.”
Copyright © 2019 CST, Inc. All Rights Reserved