Published: Apr 21, 2020

Facial recognition technology has dominated discussions in technology circles for some time now. Faced with increased surveillance in public spaces, it has become imperative for stakeholders to have some input on future deployments of these novel technologies. More importantly, the general public should have some degree of understanding of facial recognition and how it’s being used today.

Facial recognition explained

Facial recognition is a term used to refer to technologies used to analyze and recognize faces from video recordings and still images. Advancements in image processing and AI have enabled today’s computer to read even the subtlest details in the human face like eyelashes to differentiate people. Facial recognition technology relies on biometric identifiers that are unique from person to person.

Facial recognition has been around for years. However, in the past, it would take a long time and several passes for a computer to read, process, and recognize these biometric identifiers accurately. The emergence of faster, more robust computing resources has enabled the batch processing of video frames and still images for biometric identification. It is with these developments that governments and private entities are now starting to use facial recognition on a massive scale giving rise to facial recognition controversies.

Components of a facial recognition system

Most of the facial recognition systems in use today are made of four inter-dependent components, namely:

Capture. A capture device is used to record facial data (video frames or audio). In most cases, any camera can be used to capture images or video. However, newer cameras have come to market explicitly made for capturing essential facial features that can be processed.

Extraction. After capture, an extraction system is used to get details needed by facial processing and recognition systems. Data extracted can range from facial patterns, skin textures, facial features, among others.

Comparison. Most facial recognition systems use already existing data to identify a face. For instance, some governments use images in their identification databases to identify faces transmitted to their facial recognition system.

Decision Making. Here, an algorithm is applied to the images to decide whether it’s a match or not. For instance, some algorithms use face maps and the relative positioning of facial features to identify similarities between existing and new facial data. Depending on the system in use, a verification process will follow to eliminate cases of mistaken identity.

Facial recognition opponents have always argued against it using the effectiveness or lack thereof of the verification phase. For instance, wrong identification could lead to wrongful arrests or death, where facial recognition is used in law enforcement.

How and where is facial recognition being used?

Facial recognition technology has improved so much that it is being used in practically every industry. Some of the areas where it’s being used with success include:

Private security. Facial recognition is being used for authentication purposes in sensitive business establishments such as data centers, vaults, and private laboratories. Here companies can use the technology to authenticate incoming employees, tracking them, and restricting access to certain areas.

They can also use the technology to identify unauthorized access to company facilities using a closed facial detection and comparison system. However, concerns and even lawsuits continue to feature in the use of facial recognition technology in the private sector.

Social Media. Social media companies such as Facebook, VK, and WeChat have been using facial recognition systems in their platforms for a while. For instance, Facebook’s automatic face tagging feature is a shining example of facial recognition in the mainstream.

Government Agencies. Over 70 countries have either started using facial recognition or are already using it for various purposes. For instance, facial recognition is heavily relied upon in countries like China and Russia to identify and track people with a criminal record as they move around.

That said, the use of facial recognition by government agencies has attracted a lot of bad press in the recent past. Experts are just now starting to question the amount of tracking being carried out by government agencies and how ethical it is as far as privacy is concerned.

The bottom line

While facial recognition has shown its superior capabilities, several problematic areas may need to be addressed across the board as the technology evolves. More importantly, those using it should put user privacy at the forefront going forward.