Biometrics is generally known as a technology powered identification / authentication approach that uses a subject’s inherent characteristics to recognize them. However, a lesser known but powerful and widely deployed application of biometric technology is surveillance, which leverages biometric cues of target population to spot the subject(s) on surveillance.
This article sheds light on use of biometric technology in mass surveillance applications and restrictions on such programs that are taking place globally due to privacy concerns.
What is a biometric surveillance system?
A biometric surveillance system is a surveillance system that uses biometric technology to identify a subject on surveillance. A typical example of use of biometrics for surveillance is facial recognition, which is widely used in conjunction with CCTV based surveillance.
Why biometric surveillance is required in the first place?
Surveillance systems can be installed privately for securing home, a facility and or in an organization to implement physical security. They are prominently used by law enforcement and intelligence agencies to watch over and spot subjects on surveillance to ensure safety and security. Surveillance cameras have been in use for a long time to watch over public places and sensitive areas in real time as well as to record the incidents taking places in those areas.
Use of biometrics in surveillance applications can save a lot of effort that authorities have to put for human intervened monitoring. Leveraging biometric recognition technology with surveillance systems can automate the entire process. For example, use of face and/or gait recognition in video surveillance can automating the process of identification. Face recognition systems, which are specially designed for surveillance purposes, can process video feed in real time. If a subject on surveillance is found on the feed, the recognition system can spot the subject.
Surveillance, when done on a large target population (mass surveillance) can be a complex endeavour. Surveillance requires continuous monitoring and with traditional methods (which requires human intervention for monitoring), efficient surveillance can be a challenge. Mass surveillance applications generate a lot of data and without the proper automation in place, processing it can be very time consuming. Biometric surveillance enables law enforcement and intelligence agencies to automate identification of subjects on surveillance.
Facial recognition and surveillance applications
With biometrics, traditional methods of surveillance become more efficient and produce more precise results. Law enforcement biometric surveillance systems aim to identify individuals in challenging situations like crowd, people walking in large groups, etc. Ability to identify individuals from a distance is a must for these systems.
This ability can be implemented using biometric technologies like face recognition, gait recognition, iris recognition and many other behavioral biometrics technologies. Face recognition remains the top choice for surveillance applications due to its efficiency, cost effectiveness and ability to integrate with existing videos surveillance systems.
Face recognition has been widely deployed with video / CCTV surveillance. Face biometrics offers distinctive advantages that makes is the most suitable biometric modality for surveillance applications.
Except some cases (e.g. twins, people with facial resemblance, etc.), each individual has a unique face and set of facial features, which forms the basis of biometric face recognition. Face recognition is already in use for a variety of identification, authentication and access control applications. For law enforcement biometric surveillance systems; it offers some unique benefits, which are not offered by any other biometric modality.
To be used in mass surveillance, a biometric modality should be able to recognize individual from a distance, which is now possible with high resolution surveillance cameras. It has now become possible to capture facial details of people even in challenging conditions. Highly efficient face recognition algorithms specially designed for law enforcement applications can extract facial features of a subject even from imperfectly captured images.
Notable examples of biometric surveillance systems
Total information awareness (TIA)
Total information awareness (TIA) was a short lived mass surveillance program designed to predict criminals on the basis of detailed information gathering.
After a short run, the program was renamed as the Terrorism Information Awareness. The program was meant to search and collect all sorts of personal information to identify terrorists around the globe. TIA comprised of several components and it was able to collect biometric information like fingerprints, gait, face and iris data along with several other non-biometric cues.
Operation virtual shield
Operation Virtual Shield is a biometric surveillance program running in the U.S. city of Chicago. This program was initially deployed as camera surveillance project, facial recognition ability was incorporated later to enhance the efficiency of the system.
Skynet – China’s countrywide CCTV surveillance
China is in the process of deploying a countrywide network of surveillance cameras to watch over its people and spot subjects on surveillance. It is estimated that the country has already deployed more than 200 million surveillance cameras across its major cities, which is expected to reach 626 million by 2020.
Biometric surveillance and the right to privacy
Increasing awareness about global surveillance programs and use of biometrics in these projects have not only enraged people but also raised privacy concerns around the world. Disclosures of classified United States government surveillance programs by whistleblowers and widespread coverage of these revelations have also made people express their disagreement about such programs.
Face recognition is arguably the most used biometric modality for watching over the people. Privacy concerns over face biometrics in social media apps and their extensive media coverage have also increased awareness how it invades privacy. Surveys suggest that people do not like and want to be watched over, which is why concepts like anonymity on the web and Right to be Forgotten are being included in the new data privacy laws.
Many international bodies and privacy advocates have expressed concerns about the severity of surveillance programs around the world.
International Civil Liberties Monitoring Group (ICLMG), in collaboration with civil liberties groups and trade union partners, issued a report on February 10, 2010, denouncing abusive border control measures and human rights violations of travelers. The report, which is based on research and testimony collected through this website, sheds light on the true consequences of “enhanced” border control measures, no-fly lists and other government monitoring lists on people.
The report raised serious concerns about the upcoming implementation of the new US Secure Flight regulations or any similar program that the Department of Public Safety is developing in the privacy of parliamentarians.
The report also describes all of the extremely complex traveller safety programs that use new technologies that only a few years ago were science fiction. These programs allow governments to establish the risk rating of travelers, as well as to collect and store more and more personal information.
Global restrictions on facial biometric surveillance
After long and unceasing efforts by lobbyists, activists and privacy advocates, the government authorities have started to acknowledge the real concerns behind surveillance using facial biometrics. This has resulted in prohibition on the use of face recognition technology for surveillance in several regions and cities around the globe.
In May 2019, San Francisco became the first city in the United States to ban the use of facial recognition by the local agencies. The 8-to-1 vote by the city’s Board of Supervisors will forbid public agencies from using the artificial-intelligence software to find the identity of someone based on a video clip or photograph.
In June 2019, Somerville City Council voted to ban the use of facial recognition technology in police investigations and municipal surveillance programs, taking an aggressive stance against the practice amid a national debate over online privacy. With this ban, it became the second city after San Francisco to ban the use of facial recognition.
In July 2019, the U.S. city of Oakland also banned the use of facial recognition. The city passed an ordinance that prohibits the use of the technology on the grounds that it is often inaccurate, potentially invasive and lacks standards.
Experts and policy advisors suggests that this is just the beginning and that more regions will follow the lead.
New biometric surveillance tech is underway
Face biometrics’ invasive nature has led to the ban on the use of this technology in many regions and many more are expected to follow. On the other hand, use of biometrics for surveillance has become a necessity due to increasing criminal and terrorist threats. Ban on face recognition has pushed experts to look for other ways of positive biometric identification of criminals, terrorists and other subjects on surveillance.
Fortunately, biometrics offers a large portfolio of modalities and there are many other human physiological and behavioral traits that can be used to surveil subjects.
Face biometrics can identify a subject with the use of surveillance cameras, but there is one more biometric modality that can do the same job by analysing a video feed: Gait Recognition. Gait recognition is a lesser known but a powerful biometric modality that can patch the gap created by global ban on face biometrics surveillance.
A gait recognition system is a biometric recognition system, which can identify a subject with his/her manner of walking. These systems are designed to map human gait pattern and identify individuals on the basis of the mapped pattern. While face recognition systems depend on stills from a surveillance video feed to spot a subject, gait recognition systems can also make use of sensor data to identify a subject.
In surveillance applications, putting on-body sensors on a subject is not possible; however, sensor data captured by phones, wearables or other smart devices carried by the subject can be leveraged to enhance confidence in identification. Experts are also experimenting with radars for gait recognition, which is expected to return better accuracy than camera based systems.
Future attribute screening technology (FAST)
Future Attribute Screening Technology (FAST) is a program launched by the Department of Homeland Security to spot a suspect that may have immediate plans to commit crime or cause disruption. This system screens people for psychological and physiological indicators like pulse rate, skin temperature, breathing pattern, body movement, facial expression, pupil dilation, etc. It also analyses several other psycho physiological and behavioral attributes to get a hint of malicious intents of a hostile subject.
FAST can be particularly useful in watching over places like airports, borders and areas where disruptive activities can be carried out by an identified subject.
ECG biometrics relies on small electrical changes that take place during the heartbeat of a person, which are considered unique to an individual. ECG (Electrocardiography) biometrics is being seen as the potential tool for biometric surveillance with research and prototyping of the system is already underway.
United States Department of Defense has developed a device named “Jetson”, which can identify a person with his/her heartbeat. It uses laser vibrometry to detect movements on the surface of the skin caused by a person’s pulse. This device can pick a subject from a distance up to 200 meters and can detect heartbeat through typical clothing.
Identification using radio reflections
Radio waves are omnipresent. They are emitted by wireless communication equipment such as cellular antenna, Wi-Fi devices, GPS satellites and even by distant stars and galaxies. Radio waves are all around us. Human body absorbs some amount of these waves and also reflects some part of them.
A study conducted in 2018 at the University of California by a group of researchers suggests that each individual’s body reflects these radio waves uniquely and this unique pattern of reflection can be measured to identify that individual, once his/her identity is established.
Behavioral biometrics using affective computing can analyse facial expressions, rate of speech, posture, tone and pitch of their voice, and other behavioral traits to identify a subject from a distance.
Facial thermographs can help law enforcement officers to spot a suspect. When a suspect is about to commit a crime or cause disruption, his/her facial thermograph changes due to increased flood flow due to stress, nervousness, etc. Facial thermography is done using infrared camera. These thermograms can be processed through the software specially developed to detect changes to spot a suspect.
Global surveillance has reached its tipping point in terms of invasion of privacy. From personal devices to public places every movement of people can now be tracked.
After prolonged efforts of privacy advocates, government authorities are finally stepping up to ban facial recognition. However, ban on facial recognition does not mean that days of biometric surveillance are numbered.
Biometric surveillance experts have been working on new means of biometric surveillance already and have even showcased biometric surveillance with gait, behavioral biometrics, psychological / physiological indicators, electrocardiography and more.