January 13, 2017 -
The U.S. National Institute of Standards and Technology (NIST) will launch a new, ongoing “Face Recognition Vendor Test”, beginning in February 2017.
The test will be aimed at evaluating the face recognition performance on cooperative images, as collected in civil and criminal identity management applications. NIST also notes that the test will additionally evaluate accuracy on more difficult images, including in-the-wild and photojournalism images and non-cooperative surveillance stills.
The test will result in performance reports that will include measurements of accuracy, speed, storage and memory consumption, and resilience. NIST states that it will report the dependence of performance on the properties of the images and the subjects. In its initial form, the test will only have one assessment track, specifically designed for face verification.
Under the new test, developers may submit algorithms to NIST whenever they are ready, but no more frequently than 120 days. The algorithms will be evaluated rapidly, on a first-come, first-serve basis.
After the verification evaluation has stabilized based on the test, NIST will introduce evaluations of open-set, one-to-many identification accuracy, face detection accuracy, and also age and sex estimation. Additionally, NIST has stated that it will examine other types of evaluations.
Face Recognition Vendor Tests, conducted by NIST, provide independent government evaluations of commercially available and prototype face recognition technologies. These evaluations are designed to provide U.S. government and law enforcement agencies with information to assist them in determining where and how facial recognition technology can best be deployed. In addition, test results help identify future research directions for the face recognition community.
More information about the ongoing vendor test process can be obtained at the NIST Web site.