Is your face telling me about your past and plans?
Mind if I check that?
In the sci-fi movie “Minority Report”, technology has evolved to assist the Government as well as private corporations constantly monitor citizen’s activities. Through a vast network of surveillance cameras supported by biometric technology, citizens are provided with targeted advertising, personalized services and, for some, predictive and preemptive law enforcement actions to thwart crimes not yet committed. For privacy rights groups, the idea of a society subject to the unrelenting monitoring through technology is their greatest fear, but for law enforcement and homeland security professionals involved executive protection and controlling access to secure sites, the use of biometric technologies represents a great opportunity.
Biometric technology refers to any technology that utilizes any aspect of an individual’s body to positively identify or catalog the activities of a person. Examples of biometrics include gait recognition, fingerprint and palm print recognition, facial recognition, iris recognition (retina scanning) and DNA identification. One of the most commonly discussed biometric technologies is facial recognition, which is generally categorized into 2 categories: verification and identification. Verification systems query captured images against a defined database of stored 2D images to confirm an identity; employee access into a work area for example. Identification systems query a captured image against a much larger database, or multiple databases, to identify unknown individuals. Both categories rely on geometry-based algorithms that capture the geometric location of an individual’s facial features for use as “anchor points”. Distances and angles between these “anchor points” are measured and cataloged to create a unique face “print”. As in any technology, many factors have been shown to affect a facial recognition system’s accuracy. Some of these factors, including the illumination of the subject (shadowing and shine), pose, time delay (query speed) of the system, occlusions, size of the reference data set and age progression of both the captured images and data set images, have proven to be difficult to overcome when a system is deployed outside of a laboratory setting.
Recent technology research have indicated that a system’s accuracy can be greatly increased through the use of “image averaging”. This process creates one 2D “compilation” image by combining multiple photographs of an individual to be used as a reference in the automated facial recognition database. Alternatively, the Carnegie Mellon University — Software Engineering Institute’s CyLab (CMU-SEI) has made great strides through the innovative development of a facial recognition system that produces extremely lifelike 3D images from a 2D photograph. Reportedly, this system provides greater accuracy because the system can capture a less than optimal photograph and accurately search the target data set. Early testing of this system has also indicated that the use of a 3D image makes the system less susceptible to false or inaccurate queries due to differences in facial hair, glasses or expression on the test subject or captured image.
Because of the constantly changing threat posed from terrorist or attacker planning, law enforcement agencies with the mission to provide executive security for high-risk individuals that may be targeted must consistently identify and deploy the latest technology. One of the most important aspects of executive protection is maintaining situational awareness of what is occurring in the vicinity of protected person and denying a potential attacker access to the target. Reportedly, some protective agencies are reviewing technology that monitors all venue access points with biometric supported cameras that automatically check the facial images and other biometrics of attendees against databases of violent criminals, terrorists or people known to have threatened or stalked executives.
But what happens when the executive leaves the secured facility and goes mobile? Imagine deploying mobile facial recognition systems at large outdoor venues such as the sporting events, parades, and other mass gatherings that make attractive terror targets? The technique, often referred to as BOSS (Biometric Optical Scanning System) allows security teams to rapidly scan a large crowd of people and match those images against a watch list data set. The technique would greatly assist security teams in ensuring the executive is secure while outside the secure facility and exposed to large crowds which were previously impossible to screen effectively.
As can be expected when discussing any use of technology that could be used to monitor citizens, the proliferation of biometric technology has resulted in widespread fear of government abuses among some groups. The ACLU has challenged the use of facial recognition and has warned that the technology and our rights will be abused because the government can’t be trusted. Their argument centers on the lack of transparency of the technology and its surreptitious deployment that decreases the public’s ability to accurately monitor the government’s usage. Additionally, they warn that should the images or technology be housed on an insecure system or operate across the Internet, the possibility exists that the data could be stolen or corrupted by skilled hackers. Finally, security services must be cognizant of the privacy considerations involved in capturing images of the public who are merely walking outside of the protected venue. A question that has yet to be definitively answered is: if a citizen is walking in public, is their facial image protected from government scrutiny?
Interestingly, the BOSS type technology has received the greatest support and deployment from the private industry, mainly because it has been deployed within the retail sales field and has been publicized as enhancing the shopping experience. Reportedly, the use of this technology allows retailers to custom tailor the shopping experience to the individual customer and to identify advertising specific to each shopper’s needs. As described, the lack of concern by the public regarding the deployment of this technology by the private sector, which is not subject to oversight as is the government, is interesting. For consideration, at some point in the future, it may be possible that a shopper will not be provided the opportunity to view advertisements or information regarding an item they don’t usually purchase. Instead, their mobile devices will be inundated with targeted advertisements and promotions specifically tailored to their purchasing history. At that point, how can anyone feel confident that his or her shopping experience or knowledge hasn’t been controlled to a point that free will has ceased?
Although the use of facial recognition and other biometric technology will undoubtedly continue to develop, the value of this technology for law enforcement is undeniable. Should that level of cataloging by the private industry scare us more than a reviewable governmental or public safety agency? The public needs to carefully consider the implications and, as in all security discussions, decide how much safety we desire and how much privacy we are willing to give up.
