Winners announced in NIST Challenge for contactless fingerprint biometrics
Phase 2 results in the Mobile Fingerprinting Innovation Technology (mFIT) Challenge hosted by the U.S. National Institute of Standards and Technology (NIST) have been partially revealed, and experts say the evaluation shows touchless fingerprint biometrics are now ready for practical deployments.
Phase 1 winners, announced in June, 2021, were the BlueBible Team, ENGR Dynamics, Idemia, Identy, The Slapshot SDK Team (which consists of Integrated Biometrics and Sciometrics), and Telos ID. Each received $8,500 in prize money. Two additional teams also participated in Phase 2, Tech5 and a team including T3K.
First place overall in the mFIT Challenge Phase 2 is a tie between Idemia and Telos, which also tied for the ‘First Responder Choice Award.’ The Slapshot SDK Team tied with Identy and ‘T3K & Teeltech’ for third place. The Slapshot Team’s technology delivered a matching accuracy rate of 95 percent or better in Phase 2 evaluation, tops in the challenge, however, Integrated Biometrics Executive Vice President Tom Buss told Biometric Update in an interview.
The challenge is held in collaboration with first-responders telecommunications network FirstNet, and is intended to establish the effectiveness of technologies for contactless mobile fingerprint capture, transmission over FirstNet and matching to legacy databases populated with reference templates collected on contact-based biometric scanners.
This will allow police to identify persons of interest without first transporting them to a station, saving time and money for the force, while making the process less time-consuming and inconvenient for the subject. Early identification of a suspect could also have significant benefits to public or officer safety.
Buss explains that NIST took a database of 2 million reference templates and added fingerprint biometrics from 100 volunteers collected on contact-based scanners to match with the ABIS.
The NIST division based in Colorado works with the Federal Communications Commission, and ran the mFIT Challenge to encourage the development of technological innovations to support public safety workers, like police and other first responders. The industry was challenged to come up with a method for slap (four-finger) prints collection with a smartphone camera that would result in processable images, Buss explains.
“You’d like to have some kind of a tool that’s not difficult to carry or a separate piece of gear you have to plug in.”
NIST measured the matching accuracy of the entries, but also “a bunch of other parameters that have to do with ease of use, look and feel,” Buss says. While these items are important to overall performance, they were not weight as heavily, and IB was focused on biometric matching results, Buss says, as the applications will be further developed and polished by system integrators for production rollouts.
The integrators for mobile apps to capture contactless fingerprints are the same group as already serve the law enforcement community with the biometric technologies used in police stations, according to Buss.
Team Slapshot’s approach
The Slapshot SDK captures images within 2 percent of the 500 pixels per inch of the templates in the legacy database.
“If the images that are presented to the ABIS are not close to 500 ppi resolution, the accuracy falls off rather quickly,” Buss notes.
The Sciometrics software included in the SDK controls the accuracy of the resolution at which all fingers are captured, with a high level of focus delivered by a series of images, with the optimal image for each finger selected to create the probe template.
CEO Shawn O’Rourke says IB has a patent portfolio for touchless technology that is equally robust to its contact-based IP portfolio.
Meanwhile, standards are being created for contactless fingerprint capture, and will be in part informed by the mFIT results.
Sciometrics President Mark Walch characterizes the results of the mFIT Challenge as showing that touchless fingerprint collection is “ready for prime time,” which he expects to spur action from NIST to support the creation of those standards.
Buss says that IB is planning to leverage its leadership in the contactless identification space to address the demand in identity verification with a future software release. This will involve some different considerations, as for instance the quality of phone cameras involved varies more widely.
“We’re going to win this contactless market just like we won the contact market when it comes to fingerprints,” assesses IB EVP of Global Sales and Marketing Dave Gerulski.
accuracy | biometric matching | biometric testing | biometrics | biometrics research | contactless | fingerprint biometrics | Integrated Biometrics | mFIT Challenge | NIST | Sciometrics | Slapshot SDK