IARPA challenge aims to improve face recognition

Jul 12, 2018 | Stephen Mayhew

The Intelligence Advanced Research Projects Activity (IARPA) in partnership with the National Institute of Standards and Technology (NIST) is holding a facial recognition challenge with a total prize purse of $70,000. The Fusion of Face Recognition Algorithms (FOFRA) prize challenge looks to improve recognition of face images through fusing the outputs of multiple algorithms applied to the same input imagery.

“Face recognition has made significant advances in recent years and the fusion of more than one algorithm has shown to improve its accuracy. IARPA is in a unique position to provide data for researchers to develop fusion methods that should further improve performance,” explains IARPA program manager Chris Boehnen. “These fusion schemes could allow users to leverage multiple algorithms without overhauling or re-designing the core technologies.”

With this challenge IARPA is inviting the broader research community of industry and academia, both domestic and international, to participate in a simple and non-contractual way. The registration period officially runs until August 6, 2018. Participants will submit their fusion algorithms for evaluation to NIST. Solvers who are eligible to win a prize and with the most accurate and complete solutions will be eligible to win cash prizes from a total prize purse of $70,000. Winners will be announced in October 2018.

The technology IARPA hopes to generate through the contest would have across-the-board use in face recognition, though it wouldn’t necessarily apply to every scenario. “The improvements sought within FOFRA would be relevant to watch list style recognition typical for a terrorism use case and to authentication verification scenarios such as securing a mobile device,” Boehnen told Government CIO Media.