The Community for Technology Leaders
Proceedings of Fifth IEEE International Conference on Automatic Face Gesture Recognition (2002)
Washinton D.C.
May 20, 2002 to May 21, 2002
ISBN: 0-7695-1602-5
pp: 0107
Stan Z. Li , Microsoft Research Asia
HongJiang Zhang , Microsoft Research Asia
XianHuan Peng , Peking University
XinWen Hou , Peking University
QianSheng Cheng , Peking University
ABSTRACT
Independent subspace analysis (ISA) is able to learn view-subspaces unsupervisedly from (view-unlabled) multi-view face examples \cite{Li-ISA-ICCV-01}. In this paper, we explain underlying reasons for the emergent formation of ISA view-subspaces. Based on the analysis, we present a supervised method for more effective learning of view-subspace, assuming that view-labeled face examples are available. The models thus learned give more accurate pose estimation than obtained with the unsupervised ISA.
INDEX TERMS
Face pose estimation, multi-view, independent subspace analysis, supervised learning
CITATION

X. Peng, S. Z. Li, H. Zhang, X. Hou and Q. Cheng, "Multi-View Face Pose Estimation Based on Supervised ISA Learning," Proceedings of Fifth IEEE International Conference on Automatic Face Gesture Recognition(FG), Washinton D.C., 2002, pp. 0107.
doi:10.1109/AFGR.2002.1004140
88 ms
(Ver 3.3 (11022016))