The National Institute for Standards and Technology (NIST) has created two standard datasets of images to test facial recognition technologies: 1) “In the wild” (LFW) images that are more typical of images captured in video surveillance, and 2) constrained (IJB-8) images that are based on celebrity photos from media outlets. This video explains how BriefCam ensures high accuracy of results with its “in the wild” Face Recognition technology by creating several diverse benchmarking and training datasets, then manually annotating them with ground truth data sets, then testing them for each BriefCam product release.