Interview with Joe Rosenkrantz, CEO, Airborne Biometrics Group, Inc.

Companies referenced in this item: Airborne Biometrics Group, Cognitec Systems, FaceFirst

FB: Could you please tell us about the background of your company?

Airborne: Airborne Biometrics was founded jointly with myself and Airborne Technologies, here in Camarillo, California. Airborne Technologies is a 30 year old aerospace company that is an OEM for Lockheed in the manufacturing and distribution of major military aircraft parts. Due to Airborne Technologies’ government affiliations, certifications, credibility, etc, we felt that it was important to align ourselves with a company that was well-backed and well established.

FB: You launched the FaceFirst product at ASIS last year, would you please describe this product for us?

Airborne: FaceFirst is a biometric platform, with capability of harnessing multiple biometric technologies. The reason that we named the company FaceFirst revolved around the fact that we chose the facial biometric to be our first implementation of biometrics, due to the vast market opportunity.

In 2007 we found that facial biometrics were relegated to military use and university labs and not nearly mature enough to be used in commercial applications. The near real-time face recognition that you would see on T.V., shows such as Vegas and 24, were simply a fantasy. Yet, the National Institute for Standards and Technology had been tracking facial recognition algorithms since 1993, and it so happens that in 2006, facial recognition algorithms had actually surpassed human recognition. We saw an opportunity in the market to create a turn-key platform that would allow us to offer the best commercially available face recognition algorithms. By creating a platform, rather than an algorithm-specific piece of software, we are able to quickly adapt our software to take advantage of the most cutting-edge capabilities available.

 FB: Which facial algorithm do you currently use?

Airborne: Currently, we are focusing on the Cognitec algorithm. Cognitec is considered top rated by the National Institute for Standards in Technologies and we have found it extremely accurate.

 FB: Very good; we too are very familiar with them. What are the targeted applications for FaceFirst?

Airborne: Well, with FaceFirst there are three main applications that we’ve created. The first is the live video component; and that allows us to accurately track people at 20 plus frames per second, which permits us to detect and identify every face on every frame of video, simultaneously.

The FaceFirst system was built in a cloud-architecture so that it can be run either at a single site, or within a corporate or internet cloud. This way, it enables us to remotely host the main matching and encoding engines, and from there to, essentially, transform a high definition video stream down to just a series of the best representation of each face that passes by cameras. This means that we can provide wide-area networking capabilities, something that previously was not possible using high definition video because of the high amount of bandwidth required per camera.

The second module is the mobile face recognition capability. We’ve engineered the software to run on most of the commodity mobile phone platforms that are available today, such as Apple, Motorola, Android, Blackberry, and Nokia. We’re able to use the phone as a dumb device to acquire the image of the person and securely log into our system whereupon the system then transmits back all the possible matches for the person you’ve identified, as well as providing all the biographical information at the same time. Therefore, the system is not only capable of identifying people that you’ve pointed the camera at, but it’s also capable of notifying any person who has subscribed to an alert that this person has been detected and identified.

Now, as to the third module, this is our mobile SLR capability. We’ve engineered a capability that allows us to enable any of over 1,000 digital camera models to enable a person to photograph either a single person, or even an entire crowd, and that image will then be wirelessly transmitted back to our matching engine in the cloud, where it then can be used to detect, crop and identify every single face on that image, no matter how many faces are on that image. This technology was specifically created for the counter intelligence community, as it allows them to photograph special events or people who they are attempting to track. This mobile SLR capability provides them with not only the match, but also all the biographical information concerning the person they’ve photographed. That information is provided through either an e-mail or an SMS alert sent directly to their mobile device or their e-mail.

FB: So, was FaceFirst designed predominantly for the counter intelligence community? I ask since I see a lot of commercial applications where this could be utilized.

Airborne: No, not just for counter intelligence; we also have applications for Corporate America as well. We enabled customers who are using the system in a corporate environment to manage their lobbies, for purposes of identifying people who are on a watch-list that might enter secure buildings.

FaceFirst also provides an application for photographers in the entertainment industry. That application allows the photographer to snap photos of a great number of people—for example, at a theme park, or cruise ship, and our application allows the system to automatically cluster together the pictures of each person who is on that ship, or at that park that day, or any such venue. This creates an opportunity for the person to shop for photographs by using their face. For example, when they approach a kiosk the system identifies them and shows all the pictures that are associated with them.

FB: I agree, that’s a great application.

Airborne: Yes, so we’re in the entertainment industry, as well as transportation and airlines. The beauty is that we are able to penetrate many different industries.

FB: Now, what comes to mind for me, being centred in Toronto as I am, and we just had the G20 Summit here and there was a fair bit of vandalism by hooligans who were known by the law enforcement community; could your product have been used by the police to assist them in picking out potential troublemakers from a crowd that had gathered?

Airborne: Yes indeed; our application allows law enforcement to quickly set up the system as a mobile deployment, placing cameras in key areas or at pinch-points, such as metal detectors and magnetometers, and then as people are entering an event every face will automatically be scanned and if needed, will create an alert to the folks either on or off site who might be interested in the identification of that particular person. They are able to do that using two-dimensional picture data bases that they already have.

FB: Fantastic; I’m almost sure we could have used your system here in Toronto this past month, and it may have eliminated some of the troublesome issues we had here.

Airborne: No doubt; and there are stories that I could tell you, —but we absolutely have those kinds of applications that are specific to those kinds of events.

FB: Now, with this new technology, what are some of the challenges that you face in implementing it? For example—privacy issues, do those tend to crop up to be somewhat of a challenge?

Airborne: Privacy interests and issues are a major challenge that we face on a daily basis here. This is especially true of the E.U. and the United Kingdom where there are very specific privacy concerns associated with photographing and retaining pictures of people, as well as enrolling people’s faces into a biometric face recognition data-base without their express permission. So yes, this does continue to be a challenge and therefore we continue to work with various legal systems to be able to permit our systems to operate in such areas.

FB: What can we expect to see next from your company?

Airborne: I believe that the future for us lies within multiple biometrics; the meshing of multiple biometrics to create a system that is not only more accurate but also that will move closer to 100% automated.

FB: Thank you very much for filling us in. The product is truly exciting and we look forward to hearing more about it as your company progresses.

Airborne: Thank you Peter and we look forward to speaking again in the near future.
 

About Facefirst

FaceFirst provides a fully automated, user friendly, turnkey mobile and live-video surveillance facial recognition system which generates automated alerts whenever a face match above a user defined probability is reached. FaceFirst has a completely open and scalable, system architecture. FaceFirst technology excels in low resolution environments enabling real-world performance. Products and services: Facial recognition, Justice & Law Enforcement, Physical Access, Mobile Biometrics and Financial Transactional.  http://www.facefirst.com/