Facial recognition is inevitable in our everyday lives. Whether we are unlocking our phone or entering a country, our faces are being scanned regularly in order to confirm our identities. Advancement in technology has enabled computers to analyze patterns with the unique contours of our faces and make comparisons to other faces in a database.
How does facial recognition work?
Facial recognition technology gives us the capability of verifying an individual’s identity from a digital image or a video source. There are three main components in a facial recognition system: a video or image capture device, the required hardware and software to analyze the image, and a database of images to compare the given videos and images. The accuracy of such a system is very much dependent on the quality of all three components.
In its simplest form, facial recognition technology works in four steps: detection, analysis, data conversion, and comparison.
- Detection
Basically, the algorithm detects the presence of a face in the image or video from the geometry. Back then, facial recognition technology relied on 2-dimensional technology and scanning a static image was the only way to get accurate data. With improvements along the way, now an image can be detected from a static image or a live video feed. The software can now analyze multiple images in a frame and pick out an individual face. The detection process works best when an individual is facing the camera. Time over time, the systems are more capable to handle more significant variations in camera angle.
- Analysis
Once the system has identified a face, it analyzes the features of the faces. Every human face has 80 facial landmarks or nodal points. These include the distance between the eyes, the depth of eye sockets, jawline length, cheekbone shape, the nose’s width, and other distinct characteristics. When taken together, the measurements of these facial landmarks create a unique profile of the face.
- Data Conversion
Next, the analysis is converted into what is referred to as a digital faceprint. Unlike a fingerprint, the faceprint is not something we can look at because the data is converted into computer code or mathematica formula. However, like a fingerprint, each face has a unique code to differentiate it from any other face.
- Comparison
The code for the unique faceprint is then digitally compared against a database of other faceprints. Without this comparison stage, there is not much value to the process. This means that the more the database grows, the more useful facial recognition technology becomes. The FBI currently has over 641 million photos in its database. They come from various sources such as state DMV records. With more photos becoming digitized and more images being shared on social media, the size of the database will continue to grow. When the faceprint matches an image in the database, the result is returned. In most cases, due to the limits of current technology, the images are then compared by human eyes to verify the result.
There, you are recognized after all 4 steps are completed, usually in a matter of seconds.
Applications of Facial Recognition Technology
The potential for the use of facial recognition technology is seemingly unlimited. While we are already seeing it used a lot in our daily lives, new applications are continually developed to make our lives easier, safer, and more convenient.
- Device Security
The most obvious application would have to be this one. Paired with attention tracking, you unlock your iPhone with just a glance. Facial recognition can serve as another layer of protection to individual applications.
- Payments
We have quickly shifted from a monetary system based on cash and checks to credit cards and mobile phones. Believe it or not, we are now able to complete a transaction with nothing more than our face. Payment systems use facial recognition technology to identify customers whose images are linked to their payment information. Amazon Go is an example. You simply walk out the stores, and you will receive a receipt shortly. There are now 21 of them located in New York, Chicago, San Francisco and Chicago.
- Missing Persons
As the database of facial recognition grows and live scanning technology becomes more advanced, this technology can help find missing children and adults. Cameras in airports and other public spaces can scan faces and alert authorities when a faceprint matches people who are missing or have been abducted.
- Building Security
Many secure facilities have made a move from keys to RFID tags and even fingerprint readers. Facial recognition technology could make it even easier to control and log access to buildings and sensitive areas. All it takes is to look at the door, you’ll have access to your safe home, guarded by your facial recognition system.
- Law Enforcement
Law enforcement was one of the first applications for facial recognition technology. As technology has advanced, it has moved from a way to identify suspects to more complex tasks. Technology is in the works to allow police to scan faces in real-time to provide contextual data on individuals during traffic stops and other interactions.
The Risks of Facial Recognition Technology
Come with the endless possibilities are the risks of this somewhat untested technology.
Think about it, you basically have all your personal information, from name, age, gender to address, bank number and a lot more, written on your face. You are under surveillance anytime, anywhere with all these cameras and who knows who is watching.
IBM, one of world’s leading technological company, announced a halt to the development of facial recognition technology and its related products. Arvind Krishna, CEO of IBM, believed the technology could be abused, with examples as ‘mass surveillance, racial profiling and violation of basic human rights and freedoms’.
Privacy advocates worry about the government’s ability to easily track and monitor citizens without permission or even court order. Law enforcement officials can access extensive data, including place of residence, employment status and criminal records, just from a captured image of your face. Worse still, if any unwanted third parties hacked into the system, they could take advantages with your personal information without authorization. For instance, aggressive, predatory marketers could track our movements and create profiles of our behaviors in order to do targeted advertising.
Facial recognition will be more inevitable as it continues to develop. When will there be a law or a guideline to ensure our privacy and safety? This might not be up to us to control. Still, there is one thing we can do: as we enjoy the convenience, it’s high time we learn to protect ourselves.