In this paper, we present a novel representation of the human face for esti
mating the orientation of the human head in a two dimensional intensity ima
ge. The method combines the use of the much familiar eigenvalue based dissi
milarity measure with image based rendering. There are two main components
of the algorithm described here: the offline hierarchical image database ge
neration and organization, and the online pose estimation stage. The synthe
tic images of the subject's face are automatically generated offline, for a
large set of pose parameter values, using an affine coordinate based image
reprojection technique. The resulting database is formally called as the I
BR (or image based rendered) database. This is followed by the hierarchical
organization of the database, which is driven by the eigenvalue based diss
imilarity measure between any two synthetic image pair. This hierarchically
organized database is a detailed, yet structured, representation of the su
bject's face. During the pose estimation of a subject in an image, the eige
nvalue based measure is invoked again to search the synthetic (IBR) image c
losest to the real image. This approach provides a relatively easy first st
ep to narrow down the search space for complex feature detection and tracki
ng algorithms in potential applications like virtual reality and video-tele
conferencing applications. (C) 2001 Pattern Recognition Society. Published
by Elsevier Science Ltd. All rights reserved.