The definition of facial recognition is a piece of technology that is capable of matching a human face from an image taken from a digital picture or video frame against an identity stored on a database of faces. Facial recognition is used around the world for a wide range of applications, from unlocking smartphones to identifying potential criminals.
Facial recognition system is a sophisticated way to verify or ascertain someone’s identity using an algorithm that processes a digital image or video frame. It picks out distinguishing features of someone’s face shown in an image and matches these to the faces already logged within a database. It is growing in popularity, with new uses for it being developed all the time. Digital images and video stills are becoming clearer and easier to pick out distinct people and faces, while the matching software and algorithms are benefitting from increased data sources and accuracy.
The use of facial recognition and data is also increasingly becoming a vital part of commercial identification, helping to target individuals and personalize sales and marketing messages. Facial recognition is also increasingly used in automatic image indexing, human-technology interactions, and video surveillance systems. Technological advances are constantly happening, with significant upgrades to data quality.
The facial recognition process works via software that searches for faces on the image being analyzed. After a face has been detected in an image by the facial recognition solution, it is analyzed by its own computer-generated filters. The person’s features are measured and pinpointed by the facial recognition system to work out if they match any of the faces kept on a database.
Everybody’s face is slightly different and distinguishing features, which are known in facial recognition terms as ‘nodal points’ can be identified and analyzed one by one to narrow down the search. The visual data from the digital image is turned into numerical expressions that enable the system to determine similarities with facial images already on file. These expressions are collectively known as the ‘faceprint’. Information around the person’s gender, age, and ethnicity can also help the facial recognition software to make a match.
Once the analysis is complete, a match can be found and identity confirmed. The process employs ‘deep learning and artificial neural networks to run the data and analyze the findings scientifically. This helps to produce as accurate a result as possible.
In the 1960s, work was already being done on computer applications designed to measure and pinpoint facial features with a view to identifying them from a collection of faces on a database. Facial recognition as a concept, however, was formally pioneered in 1964 by computer scientists and mathematicians, Woody Bledsoe, Charles Bisson, and Helen Chan Wolf. They used manual mapping to explore the opportunities at first, however much of their initial work was never made public, due to the secret nature of its funding from an unnamed intelligence agency.
By the 1970s, facial recognition accuracy was coming on in leaps and bounds with aspects such as hair color, lip shape, and other key identifying areas now included in the process. Linear algebra was used more and more in the decades that followed to hone the system even more finely. A key milestone came in 2001 when law enforcement in the US used facial recognition for the very first time to help identify Super Bowl spectators potentially causing disruption or harm. In 2011, facial recognition software played a large part in identifying Osama Bin Laden.
As with any software and IT, facial recognition is increasing in sophistication and accuracy all the time. As more faces are added to the source databases as the software is used, the system has a larger bank of potential matches to search through. Results tend to be more accurate when the system is analyzing images containing a single face, rather than one that must be picked out from a crowd. It also works best when the person is looking head-on at the camera, with their body and faces straight.
However, algorithms are increasing in sensitivity and sophistication, allowing images taken at more of an angle to still be accurately analyzed and identified. When answering the question, what data does facial recognition use? It is clear that the answer is, increasingly, every single aspect of a person’s face, whichever angle they are positioned at.
Nist executes FRVT (Face Recognition Vendor Test) tests between face recognition vendors and these test results give the accuracy of the facial recognition algorithm.
Facial recognition systems are increasingly common in many, if not all aspects of modern-day life. If you own a smartphone or use social media, chances are that you have the option to turn on facial recognition to sign you in to the system without having to enter a password or code. Other examples of what facial recognition is used for include the following:
Facial recognition enables a mobile phone or smart tablet’s authorized owner to open the device using just their camera to show their face to the system for analysis. This process is quick and easy, making what is already a convenient tool for daily life even easier to access and employ.
Criminals have long been identified by images captured on CCTV and other video surveillance devices. Witness descriptions have also enabled E-FIT (Electronic Facial Identification Technique) mock-ups and artists’ sketches to be developed to help track down people of interest. Facial recognition takes these techniques one stage further, linking digital images to databases of potential matches. This is called preemptive public security.
Facial recognition can help to speed up cumbersome check-in and check-out processes in busy airports. It is easy to match digital images with people’s passport photos for immigration and visa purposes. Additionally, facial recognition in airports enhances overall security as the system can quickly spot people who could be intent of doing harm or who are officially identified as missing persons.
Facial recognition can help enormously in searching for missing people. Images captured by video surveillance cameras, CCTV, etc can pick out individuals using sophisticated algorithms that match distinct features to photos on a database. Such searches can happen on a global scale, rather than being limited to a missing person’s local area or most recent sighting.
Shopping centers, marketplaces, boutiques, and high streets can all benefit from having surveillance cameras linked to facial recognition software to quickly identify any would-be criminals. This can help reduce theft, violence, and abuse, making shops safer and more pleasant for everyone.
Banking comes with a very serious requirement for identity security and privacy. Facial recognition can ensure that only the authorized person or their approved representatives can access sensitive banking data or withdraw or transfer money. It can help do away with the need for inputting lengthy passwords, memorable data, and other confidential info, which can reduce queues and make banking a faster, more convenient process.
Medical records, hospital treatments, and other sensitive data around people’s health must be kept confidential, yet also be accessible to the professionals and patients who need it. Facial recognition software helps keep details private and secure. It also helps prevent mistaken identity from happening, which can lead to distressing mistakes with medication and diagnoses.
Facial Recognition also helps to build easy access control systems and visitor management systems for new smart building structures. Facial recognition helps to access offices and flats without using keys, cards, or even fingerprints. Facial recognition access control systems are able to do this with current technology.
Today, facial recognition is present in more and more areas of our lives, bringing us enhanced security, personalization, and convenience. This includes logging into smartphones and personal devices simply by pointing the camera at the user’s face to prompt the system to unlock access. 2017 saw Apple unveil its Face ID technology for the iPhone X. This is now an ongoing key part of iPhone technology.
Other firms that use the technology include British Airways, which employs facial recognition software to enable passengers to board flights from the US without having to present their passport or boarding pass. Facial recognition technology is also installed inside the security screening area at Heathrow Airport’s Terminal 5 to enhance security and speed up identity checks.
The arrival of COVID-19 changed the way we interacted in many ways, large and small. One significant change has been around the use of masks in public areas. Obviously, when people wear something that covers much of their face, this will have an impact on the efficacy of facial recognition analysis. Companies are already working to mitigate against this by reworking algorithms to take masks into account. Many facial recognition systems are already starting to adapt, resulting in pretty high levels of accuracy in some cases.
There are many benefits to facial recognition software. A key area is improved security, both in public areas and for companies and organizations, such as banks, schools, prisons, and airports. It can help law enforcers to identify people of interest more quickly and work out their movements to track them down and prevent them from causing any harm.
It can also speed up identity checks in airports, at borders, and for personal admin tasks, such as banking or entering and exiting workplaces and other buildings. It also works very well with social media, providing an alternative, non-invasive method of accessing accounts rather than inputting passwords or codes. Retailers and marketing professionals can also use the technology to tailor advertisements and commercial messages to customers more precisely.
As with any technological system, some people have concerns over facial recognitions, citing concerns about issues such as privacy and the right to go about one’s business undisturbed. Privacy laws are tightening up to keep pace with innovations, such as social media and remote communications. Ensuring consent and full transparency over what data is going to be used for is extremely important. Facial recognition software must adhere to these laws and ensure that the data being captured is used and stored correctly and ethically to protect people’s identities.
Another concern is around the potential misuse of data. With the best will in the world, misidentification is always a possibility, as is the risk of a data breach. Again, facial recognition practitioners must ensure that their protocols and security measures in this area are as tight as possible to protect people from potential misidentification or accidental exposure of personal data. Algorithms must be updated to eliminate any outdated systems that could cause unintended bias, for example when identifying women, older people, or people from different ethnicities
The law surrounding facial recognition software and how it can be used varies from country to country. It is worth checking what the law is in your own country before proceeding with a facial recognition project of any kind. Additionally, research around privacy laws, data usage, and storage and identity research guidelines will give you a clearer picture of what is and is not allowed. Legally, the general public cannot appeal against the installation of a security camera per se. However, they should be informed and consulted wherever possible. Any collection and use of personal data, including facial recognition data, is protected by law. Generally, the benefits of installing the technology must clearly and significantly outweigh any potential public distrust or concern.