Crossmatch to Continue On to Phase Two of IARPA’s Thor R&D Program

“Phase 2 of Thor will see a higher concentration of software work, building upon the prototype scanners developed in Phase 1. The work includes developing PAD algorithms based on biometric features, Artificial Intelligence (AI) and machine learning that work for both known and unknown presentation attacks.” – Bill McClurg, SVP & CTO, Crossmatch

Crossmatch to Continue On to Phase Two of IARPA's Thor R&D Program

The first phase of the Intelligence Advanced Research Projects Activity’s (IARPA) “Thor” program has concluded, and Crossmatch has announced that it has been selected to continue on to Phase Two.

First announced last year, the Thor program is aimed at developing Presentation Attack Detection technologies to thwart spoofing attempts. It’s part of a broader program, codenamed “Odin”, that involves not only IARPA but also the National Institute of Standards and Technology (NIST) and the John Hopkins University Applied Physics Lab, with the latter organizations having committed to helping to evaluate the technologies developed.

The program reflects a broader, intensifying concern over presentation attacks – commonly known as spoofing – against biometric technologies, with the FIDO Alliance having announced a biometrics certification program including an important PAD component earlier this year. In pursuing its own solutions, IARPA, a part of the Office of the Director of National Intelligence, is demonstrating the government’s interest in such technology.

In announcing its participation in the next phase of the Thor program, Crossmatch asserted that it was one of three teams selected to participate, and that it “is the only Prime from the biometric industry,” according to a statement. Elaborating further on the work itself, Crossmatch SVP and Chief Technology Officer Bill McClurg indicated that his firm has been working on fingerprint and iris biometrics, and said this new phase of the program “will see a higher concentration of software work, building upon the prototype scanners developed in Phase 1.” McClurg added that Phase Two will include the development of PAD algorithms, artificial intelligence, and machine learning technologies.

Crossmatch also explained that the third and final phase of the Thor program is expected to conclude in early 2021.

September 20, 2018 – by Alex Perala