Performance of Geometrix ActiveID^TM 3D Face Recognition Engine on the FRGC Data

We present and analyze the performance of the Geometrix ActiveIDTM Biometric Identity System on the FRGC data from Experiment 3, consisting of 4007 3D faces. The performance is presented and analyzed for different facial expression categories. We first present the Geometrix FaceVision 3D face recognition technology, then describe how shape and texture are fused. The effect of fusion on error rates is discussed, and the remaining errors are analyzed. We conclude with lessons learned, general findings, and what we expect in the near future.