Abstract:
Over the next decade automated face recognition is expected to make the leap from a niche application to a fairly ubiquitous function in our daily lives. The impetus for this rise is progress in what has been termed "unconstrained face recognition". Unconstrained face recognition is the science of automatically inferring facial identity from images and videos that contain arbitrary viewpoints and illuminations, as well as facial occlusions and expression variations. Currently, without imposing any constraint on such dimensions, the process of automated face detection is quite difficult, let alone automated face recognition. In this talk a brief intersection will be provided across the technical challenges limiting unconstrained face recognition, and the applications that are driving progress in this field.

Biography:
Brendan Klare is the co-founder and CEO of Rank One Computing, a small technology startup focus on software for face recognition applications. Brendan received his PhD in Computer Science from the PRIP lab at Michigan State University. Brendan has since supported several federal agencies in their use of face recognition technologies, to include IARPA, the FBI, and the Dept. of State.

Host:
Dr. Anil K. Jain