Stanford prof looks to lower false rejections, increase speed of Aadhaar
Stanford University business professor Lawrence Wein has published new research outlining a system for improving accuracy in India’s Aadhaar program, which has so far collected fingerprint and iris biometric information from 500 million people.
According to the research paper, Analyzing Personalized Policies for Online Biometric Verification, there is a tradeoff in the Aadhaar program between accuracy and speed, based on the way fingerprint and iris data is captured and then verified.
When a person first enrolls for an Aadhaar number, scanners take image data for all 10 fingers and both irises. This data is sent to a central database, which is used to verify biometrics when people show up at local offices to receive benefits or other government transactions.
The problem, Wein’s paper argues, is that comparisons can be complicated, as the scanning equipment used for initial enrollments is more sophisticated than it is at local government offices, which sets the stage for a lot of false rejections. There is also an issue of speed, seeing as the system is supposed to perform 1 million verifications per hour.
This isn’t a problem the UIDAI isn’t aware of, as the government agency has previously tried to speed up the process by comparing fewer fingers, or also by using only an applicant’s ‘best’ fingerprint.
According to an announcement from the Stanford Business School, Wein and his team’s solution is to focus on a particular subset of each person’s fingerprints and iris images that are the easiest to compare to those originally scanned. For many people, a check of one or two fingerprints was enough for accurate identity verification.
“For about 37 percent of people, it’s necessary to compare just the irises. And for a very small number of people, it’s necessary to compare both irises and some fingerprints,” The Stanford School of Business reported.
“By spending a small amount of time on most people, and more time on a minority of others, the researchers found they could keep the average verification time to just 37 seconds. That’s a bit longer than it takes to just compare one finger, but the rate of false rejections is about 200,000 times lower.”