Abstract
Identifiying human faces with only a single face model per person is a difficult task, because the input face image may vary with position, expression and pose. This paper describes a flexible face identification system based on Gabor wavelet representation and flexible neural network matching. The face is represented by a hexagonal graph of neurons where each node contains local feature information extracted by Gabor wavelet transform at the corresponding position. An innovative flexible neural network matching is employed to finding out the exact correspondence of local features between the model and the input image based on the local feature similarity and neighborhood grouping neurons. The matching process is evaluated by competition rule based on the correlation of neuron activation in the model layers and input layer. Experiments with face images that include the variations of rotation-in-plane, rotation-in-depth, and the change of facial expression are also presented.
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
D.J. Beymer. Face recognition under varying pose. In Proc. IEEE conf. on Comp. Vision and Patt. Recog., 756–761, 1994.
D.J.Beymer. Face recognition from one example view. A.I. Memo No.1536, Artificial Intelligence Laboratory, Massachusetts Institute of Technology, 1995
I. Her. A symmetrical coordinate frame on the hexagonal grid for computer graphics and vision. Trans. of the ASME, Jour. of Mechanical Design, 155:447–449, 1993.
I. Her. Geometric transformation on the hexagonal grid. IEEE Trans. on Image Processing, 4(9):1213–1222, 1995.
W. Konen and C. von der Malsburg. Learning to generalize from single examples in the dynamic link architecture. Neural Computation, 5:719–735, 1993.
W. Konen, T. Maurer and C. von der Malsburg. A fast dynamic link matching algorithm for invariant pattern recognition. Neural Network, 7:1019–1030, 1994.
M. Lades et.al. Distortion invariant object recognition in the dynamic link architecture. IEEE Trans. on Computers, 42(3):300–311, 1993.
T. Maurer and von der Malsburg. Single-view based recognition of faces rotated in depth. In Proc. of the Intl. Wrkshp on Auto. Face-and Gest. Recog., 248–253, 1995.
A. Pentland, B. Moghaddam, T. Starner. View-based and modular eigenspaces for face recognition. In Proc. of the IEEE Conf. on Comp. Vis. and Patt. Recog., 84–91, 1994.
T. Poggio and D. Beymer. Learning networks for face analysis and synthesis. In Proc. of the Intl. Wrkshp on Auto. Face-and Gesture Recog., 160–165, 1995.
L. Wiskott, J.M. Fellous, N. Kruger, von der Malsburg. Face Recognition and Gender Determination. In Proc. Intl. Wrkshp on Auto. Face and Gest. Recog., 92–97, 1995.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Pramadihanto, D., Iwai, Y., Yachida, M., Wu, H. (1997). Identifying faces under varying pose using a single example view. In: Chin, R., Pong, TC. (eds) Computer Vision — ACCV'98. ACCV 1998. Lecture Notes in Computer Science, vol 1352. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-63931-4_273
Download citation
DOI: https://doi.org/10.1007/3-540-63931-4_273
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63931-2
Online ISBN: 978-3-540-69670-4
eBook Packages: Springer Book Archive