Author: Katrina Rousseau
Introduction
Long gone are the days of passports and IDs, a new technology emerges onto the scene -- facial recognition software. Facial recognition has shown to be a profitable industry, growing from $4 billion in 2017 to an expected $7.7 billion in 2022, and only continuing to rocket from there. Many are already familiar with this technology, despite it being only recently curated and released. Facial recognition is everywhere. In fact, you probably have a facial recognition device right in your hand! Today, newer phones and laptops come equipped with this software instead of the usual entered PIN or password. Why exactly has facial recognition become so popular? Some reasons may include the technology being unobtrusive and being already implemented within our devices. Facial recognition is also very versatile, being able to be used for security to commercial purposes, such as surveillance and marketing. However, does this convenience come at a price?
What exactly is facial recognition?
There are many types and variations of facial recognition software. Generally, however, facial recognition is software that utilizes biometrics (facial features, eye color, etc.) to map out a face from photos or videos, comparing that map with databases containing names and other identifying information.
Facial recognition technology is composed of a few main steps. The first component is face detection. The software must first detect that there is a face to begin analyzing it more in-depth. The image then enters a pre-processing phase in which the data is normalized, meaning that the photo changes to match the format of the other photos in the database. This could mean altering the resolution, zoom, lighting, or orientation of the photo. The software then records the geometry of the face, detailing key features that make your face unique, including factors such as the distance between your features, the shape of your features, etc. This step is where the software analyzes your facial features and turns them into data! Each nodal point is recorded into the database, forming your very own faceprint. Lastly, the software then compares your data to a database of other facial codes. If a match is found, it then returns the relevant information associated with the data, such as a name and address. Although this is a very simplified version of the process, most facial recognition software doesn’t stray too far from this model.
History of Facial Recognition
Though the popularity of the industry has surged in recent times, the concept of facial recognition software is more than 50 years old! One of the oldest recorded instances of the software being used dates back to 1964, when a research team led by Woodrow W Bledsoe ran experiments with early computers intended to test whether or not the computers could recognize human faces. The research process included the construction of a database in which computers were fed coordinates on a human face -- which they then used to identify the identity of a person in the photograph, based on the measurements that they had been fed. The venture ultimately failed, Bledsoe remarking, “This recognition problem is made difficult by the great variability in head rotation and tilt, lighting intensity and angle, facial expression, aging, etc.” While the first attempts were unsuccessful, in 1996, researcher Peter Hart and his team picked the project back up at the Stanford Research Institute and were able to complete the same tasks that Bledsoe had previously failed to do, exclaiming, “It really worked!”
Real-World Examples
While you may readily recognize Apple’s iPhones as having facial recognition, did you know that your face is being tracked as you travel? British Airways began to permanently adopt facial recognition in their flight process, streamlining the process to remove the need for passports or boarding passes, allowing them to board almost 240 customers in 10 minutes! Some automakers are even installing facial recognition software in their cars to prevent crashes and make sure that you are paying attention to the road at all times. Even your banking and insurance may be getting in on the facial recognition train, HSBC and Chase utilize Apple’s facial recognition feature to allow their users to log into their banking apps. Some are even allowing a photo of your face in place of a signature! The trust in facial recognition programs is evident through how widely used they are within these businesses, boasting their facial recognition features as giving added security, being replacements for, or additions to a PIN or password.
Too far? Privacy Issues
The main qualm that people have with machine learning of facial recognition is the issue of privacy. Facial recognition is increasingly being used to combat crime, law enforcement agencies see the value in a device that can analyze a face and match it to a name and address. The US government used facial recognition software to identify terrorist leader Osama Bin Laden after the 9/11 attacks. The use of facial recognition by large corporations and law enforcement bring up concerns over the accuracy of the software, as well as additional issues such as racial profiling and protester identification.
To combat the fear of exposure by facial recognition technology, many big companies such as Google and Facebook place limitations on the technology’s use by law enforcement. Facebook had stated that they do not sell their facial recognition data to third parties, the information is restricted to their site only. Facebook isn’t completely in the clear, though. Just last year they had to scrap their facial recognition program, deleting over 1 billion faces of its users due to years of concern over the technology and even had a lawsuit filed against them. It’s becoming even more difficult to trust corporations’ statements in what they say that they do with their data. In 2020, it was revealed that Microsoft had been working with the Drug Enforcement Administration, trying to sell them the technology, despite publicly stating that they would not work with the police. Internet users continue to push for clearer regulations and harsher punishments in case of violation of facial recognition regulations to keep internet privacy intact.
About The Author: Katrina Rousseau
Katrina is a junior at LB Polytechnic High School with a passion for technology. She aims to pursue a career in STEM and hopes that everyone is able to cultivate an interest in technology!
Works Cited
“AI Facial Recognition Technology Overview 2021.” RecFaces, 30 Jan. 2021, recfaces.com/articles/ai-facial-recognition.
Airways, British. “TRAVEL of the FUTURE? IT’S HAPPENING NOW, with 250,000 BRITISH AIRWAYS CUSTOMERS USING BIOMETRIC TECHNOLOGY on INTERNATIONAL FLIGHTS.” Mediacentre.britishairways.com, 27 Apr. 2010, mediacentre.britishairways.com/pressrelease/details/86/0/11003#:~:text=At%20Orlando%2C%20British%20Airways%20is. Accessed 26 Feb. 2022.
Hill, Kashmir, and Ryan Mac. “Facebook, Citing Societal Concerns, Plans to Shut down Facial Recognition System.” The New York Times, 2 Nov. 2021, www.nytimes.com/2021/11/02/technology/facebook-facial-recognition.html.
Norman, Jeremy. “Woodrow Bledsoe Originates of Automated Facial Recognition : History of Information.” Www.historyofinformation.com, www.historyofinformation.com/detail.php?entryid=2495.
Symanovich, Steve. “What Is Facial Recognition? How Facial Recognition Works.” Norton.com, 20 Aug. 2021, us.norton.com/internetsecurity-iot-how-facial-recognition-software-works.html.
Taylor, Christie. “Black Lives Matter Protests Shine Light on Facial Recognition Problems.” Science Friday, 19 June 2020, www.sciencefriday.com/segments/ai-equity/.