The following is a guest post by Mohammed Murad, Vice President, Global Sales and Business Development, Iris ID
Two of 2020’s biggest news stories – the COVID-19 coronavirus pandemic and protests seeking racial equality – have affected businesses worldwide, with the biometric industry being no exception. How long the impacts will last is, at best, a guess. But let’s take a current look at how biometrics’ most-used technologies, fingerprint, facial, and iris recognition, are faring in this new era.
According to a leading industry analyst, the virus sounded a “death knell” for contact-based, the single finger readers commonly used in office buildings, airports, ATMs and many other locations. The U.S. Centers for Disease Control and Prevention said the disease is spread overwhelmingly through close person-to-person contact. Yet the agency can’t rule out people becoming infected after touching a surface with the coronavirus on it, then touching their mouth, nose or eyes. Even if surface transmission is relatively rare, few people will risk their health when alternative technologies exist.
What’s needed is a contactless solution capable of matching fingerprints with the millions of images stored in public and private databases. Several contactless fingerprint readers are now available. However, they have mixed results with accuracy. A 2019 U.S. study conducted by the U.S. National Institute of Standards and Technology (NIST) compared prints, scanned without contact, to database images. Accuracy rates using any single finger were in the 60 percent to 70 percent range – far too low for use in security. Multiple finger combinations topped 99 percent accuracy. Manufacturers are working to improve the reliability of contactless single finger devices. Another major obstacle to implementation of the new fingerprint technology is the cost of the contactless readers, which can sell for nearly three times the cost of comparable contact devices.
While facial recognition is contactless, this year’s events highlighted problems with the technology. Healthcare workers, first responders and the general public now wear personal protective equipment (PPE) to help protect themselves from COVID-19 infection. The masks, goggles and medical caps cover much of a person’s face and reduce accuracy.
Recent mass protests pointed out more troubling problems for facial recognition. Another 2019 NIST study found facial recognition systems misidentified people of color up to 100 times more often than middle-aged white men. The study tested systems with more than 18 million photos of 8 million people in databases run by the U.S. State Department, FBI and Department of Homeland Security.
Many law enforcement organizations now use facial recognition to identify people attending civil rights protests. Authorities comb through driver’s license, passport and other government-issued ID photos to gain demographic data about people of interest. Some agencies turn to a private database, Clearview AI, which claims a database of more than 3 billion photos taken from social media and video streaming sites.
Civil libertarians have objected strongly to the use of the technology. In response, U.S. cities from Massachusetts to California have passed laws banning their local agencies from using facial recognition. There are no federal guidelines for using the technology, although the U.S. House of Representatives recently introduced legislation placing limits on the use of police bodycam video. The proposed law makes no mention of photos or videos from private sources.
The backlash has spread to facial recognition system manufacturers. This month, IBM said it would no longer develop or sell facial recognition software. In a statement, the company said it opposed any technology used for “mass surveillance, racial profiling, violations of basic human rights and freedoms.” Both Amazon and Microsoft also stopped sales of their facial software to police departments while awaiting national standards. However, other facial recognition providers said they intend to continue selling their products, with one adding the technology will enable end-users to combat racism by “fairly and effectively conduct[ing] investigations for social justice.”
Iris recognition, like facial technology, is contactless. Iris readers are proven effective in office buildings, airports, correctional institutions, border crossings, and ATMs. It’s highly accurate and also can identify children since the iris pattern forms at birth.
Iris readers aren’t affected by people wearing PPE, including goggles, nor are they by a person’s race, gender or age. Most readers require a live iris for identification and verification purposes, eliminating the ability to ID a person using a photograph. Also, iris recognition is an opt-in technology that helps to protect a person’s privacy.
Although India has more than 1 billion iris templates stored for its national ID program, worldwide iris databases are much smaller in comparison with fingerprint and facial systems. That limits the ability of law enforcement to identify many people. Yet, the current environment is forcing organization to use non-contact technology that can handle identity and protect privacy.
How will the top three biometrics be used in the future? Undoubtedly, there will be a role for each technology. Contactless fingerprint readers will become commonplace. Regulations will curb facial recognition abuses. And the number of iris templates stored in public and private databases will continue to grow.
Eventually, it won’t be surprising to see two, and maybe all three, technologies used simultaneously at mission-critical locations. However, it’s certain the events of 2020 and their effect on the biometric industry will not soon be forgotten.