Using virtual heads for person identification: An empirical study comparing photographs to photogrammetrically-generated models
The purpose of this study was to examine the effectiveness of virtual heads (i.e., three-dimensional models of human heads and faces). Our goal was to test these virtual head models as functional substitutes for photographs of humans as well as for live humans during eyewitness lineups and other processes relating to person identification. We tested the effectiveness of virtual heads by taking photographs of people and then using 3DMeNow software by bioVirtual to build three-dimensional models that resembled the photographs. We tested to see how easily experimental subjects would recognize images of the three-dimensional models (compared to photographs) after being trained on short video clips of people. The goal of this study was to examine subjects' recognition of virtual faces and to compare this performance to recognition of real faces. In the following sections, we discuss relevant previous research, present the methods and results of the current study, and finally point to directions for future work.