Abstract
Face recognition is an interdisciplinary field with elements from pattern recognition, computer vision, computer graphics, psychology, evaluation methods, and statistics.  Research in all these elements  is important to advancing automatic face recognition, and they have been included in the Face Recognition Grand Challenge (FRGC). Since the Face Recognition Vendor Test (FRVT) 2002, new techniques and algorithms have been proposed to improve automatic face recognition performance.  Among these techniques are high resolution still images, multiple still images, and three-dimensional face scans.  One of the goals of the Face Recognition Grand Challenge (FRGC) is to promote the development and assess the potential of these new face recognition techniques.  In this talk I will review FRGC and progress on face recognition since FRVT 2002. This includes progress made on face recognition from high resolution still images, multiple still images, and three-dimensional face scans.  Part of the FRGC project is to directly compare human and machine performance.  I will present new work that shows machines are capable of out performing humans.