Facial recognition technology (FRT) is a hotbed of controversy, and here are ten fascinating facts that illustrate its power, pitfalls, and surprising applications.
Fascinating Facts About Facial Recognition
10 The Race Question
The United States grapples with entrenched racial disparities in traffic stops, stop‑and‑frisk encounters, and arrests. African Americans face arrest rates twice those of other groups and are surveilled nearly three times as often.
Studies show facial‑recognition software struggles with this demographic. A 2012 analysis of mug‑shots from Pinellas County, Florida found the algorithms were five to ten percent less accurate at identifying Black individuals compared with white ones. One of the vendors, Cognitec, already supplies law‑enforcement agencies in Maryland, Pennsylvania, California and elsewhere. As Rep. Elijah Cummings warned, “If you’re black, you’re more likely to be affected by this technology, and the technology is more likely to be wrong.”
9 Vulnerability
Apple claims the iPhone X Face ID can’t be fooled, yet just ten days after launch a Vietnamese security firm, Bkav, demonstrated a 3‑D printed mask that bypassed the system for roughly $150. Apple maintains the feat is impossible to replicate. Wired magazine hired Hollywood makeup artists to try, but they failed.
More modest tricks—scarves, hats, sunglasses, or even face paint—have tripped the technology, according to Carnegie Mellon research. Cyber‑security expert Stu Sjouwerman notes that while PINs can be changed, a face and fingerprints are permanent.
8 Apple’s Giant Investment
On December 13, Apple poured $390 million into Finisar, the maker of the lasers that power the iPhone X’s TrueDepth camera, Face ID, Portrait mode, Animoji, and AirPods proximity sensing. Apple will order ten times its quarterly production.
The infusion sent Finisar’s stock up over 30 % while rival Lumentum fell about 10 %. The cash came from Apple’s $1 billion Advanced Manufacturing Fund, aimed at boosting U.S. suppliers and job creation. Finisar, founded in 1988 with 14 000 employees, had seen its share price tumble 40 % earlier that year.
7 Facial Gaydar
In September 2017, Stanford researchers unveiled software that could infer sexual orientation from facial structure. The algorithm correctly identified gay men 81 % of the time, rising to 91 % when five photos per person were used. For lesbian women the success rates were 71 % and 83 % respectively. Human judges performed worse, at 61 % for men and 54 % for women.
The study noted that gay men tended to have narrower jaws and longer noses, while lesbians displayed larger jaws—potentially reflecting prenatal hormone exposure.
6 The Future Of Medicine
Researchers at Macquarie University in Australia have trained a facial‑recognition model to estimate health metrics such as BMI, body fat, and blood pressure from face shape. Lead scientist Dr. Ian Stephen says the face holds perceptible clues to physiological health.
In experiments, participants altered their digital faces to appear slimmer, with lower BMI and blood pressure, indicating humans also read health signals from faces.
5 China’s Push For Dominance
China is building the world’s largest camera‑surveillance network. With 140 million CCTV cameras already active, the plan is to add another 400 million over three years, many equipped with facial‑recognition software.
Chinese citizens can pay for coffee, enter attractions, or withdraw cash using just their face. Some cities even publicly shame jaywalkers by flashing their faces on displays. Shanghai announced “intelligent” subway ticketing that combines voice and facial recognition, requiring users to submit a photo for verification.
4 Facial Recognition Gold Rush
In November 2017, the U.S. Department of Homeland Security convened a competition in Menlo Park, pitting private firms against each other to develop facial‑recognition tools for border crossings. The goal: identify occupants of a moving vehicle (≈40 km/h) in light rain, tolerating an initial 70 % error rate.
In January, an executive order accelerated biometric security at borders. While the push promises faster processing, privacy advocates warn of potential abuses. Georgetown’s Harrison Rudolph notes many of these programs aren’t ready for deployment.
3 Fowl Face
Chinese fintech firm ZhongAn Online launched the “GoGo Chicken” project, aiming to catalog 23 million chickens over three years using facial‑recognition technology. The system will monitor flock health in real time, targeting urban consumers who value “organic” or “free‑range” labels.
CEO Chen Wei sees the data supporting farm‑based tourism and lowering credit‑assessment costs. Google has also rolled out pet‑recognition features in Google Photos for dogs and cats.
2 The Faces Of Facebook
In 2015, a class‑action lawsuit in Illinois accused Facebook of harvesting biometric data without consent, violating state law. With no federal ban on biometric data sales, Facebook’s ad‑driven model thrives on such information.
Beyond facial recognition, the company has experimented with identifying people from blurry images using posture, clothing, and body shape, achieving 83 % accuracy. Recently, Facebook filed a patent for technology that would let retailers tailor customer service by reading facial expressions and social‑media activity.
1 Stalker’s Paradise
In March 2016, Russian developers unveiled FindFace, an app that could match strangers in a crowd with about 70 % reliability by comparing uploaded photos to a database of 200 million profiles from a Russian social network. Creators Artem Kukharenko and Alexander Kabakov note the tool can also be used to locate celebrities or even ex‑partners, generating a list of similar‑looking faces.
By September 2017, Moscow equipped 5,000 CCTV cameras with FindFace, automatically scanning live footage for wanted individuals and reporting six arrests in the first two months.