Data collection, especially biometric data collection, is a charged subject on the internet and in law.
Facial recognition is already present in consumer technology: smartphones can be unlocked by simply showing your face, and media apps like Facebook and Google Photos can detect who’s in a photo. The next frontier for facial detection is surveillance — and it’s already available for enterprise-level security operations and some home cameras.
In early October of 2020, at the ISC West Security Conference in Las Vegas, RealNet Networks, a digital media software and services corporation, launched SAFR for Security, a facial recognition platform for live video. The software boasts a 99.87% accuracy rate for Labeled Faces in the Wild and a 99.85% accuracy rate for masked faces. According to the National Institute of Standards and Technology test results, SAFR is the fastest facial-recognition algorithm to date, reaching 99.9% true positive accuracy faster than any other.
SAFR for Security, along with its video management system, leverages facial recognition to identify individuals of interest and track their movements automatically. It’s designed for large enterprises — hospitals, stadiums, corporate campuses, airports, and similar places — where security personnel must actively monitor large groups and spaces.
With SAFR, security teams can opt to receive notifications if specific individuals are detected on a video feed. They can also search past footage for specific people in the event they need to conduct forensic analysis.
What Does Facial Recognition Mean for Home Security?
Facial recognition technology works by mapping the geometry of your facial features, like the length from your chin to your forehead, based on a photo or video. This is known as a “faceprint,” similar to a fingerprint. The information is then stored and compared to millions of other faces in a database.
The security applications for facial recognition are clear. A surveillance system equipped with powerful facial technology empowers a security team to respond to suspicious, threatening or banned individuals immediately. A robust video management system like SAFR can also put together the pieces when analyzing a crime or security breach after the fact.
When it comes to home security, facial recognition technology in security cameras can be used to create a personal database of people who regularly visit your home. This way, your camera can detect if a person should be on your property or not.
Using facial recognition for home security can optimize both safety and convenience. In fact, it’s already being put to use in several products, like a camera or digital doorbell. Video doorbells like the Nest Hello (in conjunction with a Nest Aware subscription service) can be programmed to open for certain people automatically. Cameras — such as the Tend Insights Linx and Nest Cam IQ — have similar facial recognition abilities that allow you to track and check on familiar faces.
Once artificial intelligence features like facial detection, movement detection and event detection become reliable and widespread enough, it could make professionally monitored home security systems and cameras virtually obsolete.
What About Errors and Privacy?
While facial recognition technology has already proven useful, it has met a mixed reaction. For one, it has been unreliable. A large-scale experiment conducted by New York City’s Metropolitan Transportation Authority in 2019 was an evident failure. The current batch of home devices can be tricked or make mistakes (maybe they won’t recognize Grandma with her glasses off or the babysitter’s haircut).
Perhaps more importantly, they raise questions about an individual’s right to privacy. For instance, the Shelby American Car Museum in Las Vegas has already put SAFR technology to work. Richard Sparkman, the museum’s technology director, praised its security applications and its potential as a customer data aggregator. “SAFR for Security makes it simple to maintain higher security in public and restricted areas in our facility and helps us understand who is moving through our museum by age, gender, and time of day — allowing us to better tailor our museum experience,” he said in a press release. But it’s unclear whether visitors are aware of being monitored or consent to their personal data being tracked this closely.
As pointed out by Molly Price of CNET, facial recognition also introduces ethical dilemmas to the home. Data collection, especially biometric data collection, is a charged subject on the internet and in law. Many argue that consumers and private citizens have a right to decide how their personal data may be used and stored. By extension, facial recognition in public and private security cameras (which store data on the cloud or a private server) would represent a violation of that right.
Facial recognition represents just a fraction of the artificial intelligence technology that is poised to enter everyday life. In our homes and in public, we will need to answer the question: when should safety and convenience take priority over privacy and individual freedom?
Source: RealNetworks, Inc.