Gender Classification on Real-Life Faces - Advanced Concepts for Intelligent Vision Systems - page 329

much less than those of SVMs using raw pixels or standard LBP. As observed in
Table 2, the boosted LBP based SVM also produces the smallest standard vari-
ation, thus more robust than other methods. We see in Table 2 there is notable
bias towards males in all experiments. This is also observed in existing studies
[10]. Finally we show in Fig. 8 some examples of mis-classification, some of which
could be due to pose variations, occlusion (e.g., glasses), and facial expressions.
4 Conclusions
In this paper, we investigate gender recognition from faces acquired in uncon-
strained conditions. Extensive experiments have been conducted on the LFW
database. We adopted Adaboost to learn the discriminative LBP features, and
SVM with boosted LBP features achieves the accuracy of 94.44% on this dicult
database.
References
1. Golomb, B.A., Lawrence, D.T., Sejnowski, T.J.: Sexnet: A neural network identifies
sex from human faces. In: Advances in Neural Information Processing Systems,
NIPS (1991)
2. Brunelli, R., Poggio, T.: Hyperbf networks for gender classification. In: DRAPA
Image Understanding Workshop (1992)
3. Yang, Z., Li, M., Ai, H.: An experimental study on automatic face gender classifica-
tion. In: International Conference on Pattern Recognition (ICPR), pp. 1099-1102
(2006)
4. Hadid, A., Pietikainen, M.: Combining appearance and motion for face and gender
recognition from videos. Pattern Recognition 42(11), 2818-2827 (2009)
5. Moghaddam, B., Yang, M.: Learning gender with support faces. IEEE Transactions
on Pattern Analysis and Machine Intelligence 24(5), 707-711 (2002)
6. BenAbdelkader, C., Grin, P.: A local region-based approach to gender classifica-
tion from face images. In: Computer Vision and Pattern Recognition Workshop,
pp. 52-52 (2005)
7. Lapedriza, A., Marin-Jimenez, M.J., Vitria, J.: Gender recognition in non con-
trolled environments. In: International Conference on Pattern Recognition (ICPR),
pp. 834-837 (2006)
8. Baluja, S., Rowley, H.A.: Boosting set identification performance. International
Journal of Computer Vision (IJCV) 71(1), 111-119 (2007)
9. Makinen, E., Raisamo, R.: Evaluation of gender classification methods with auto-
matically detected and aligned faces. IEEE Transactions on Pattern Analysis and
Machine Intelligence 30(3), 541-547 (2008)
10. Shakhnarovich, G., Viola, P.A., Moghaddam, B.: A unified learning framework for
real time face detection and classification. In: IEEE International Conference on
Automatic Face & Gesture Recognition (FG 2002), pp. 14-21 (2002)
11. Gao, W., Ai, H.: Face gender classification on consumer images in a multiethnic
environment. In: International Conference on Biometrics (ICB), pp. 169-178 (2009)
12. Huang, G.B., Ramesh, M., Berg, T., Learned-Miller, E.: Labeled faces in the wild:
A database for studying face recognition in unconstrained environments. Tech.
Rep. 07-49, University of Massachusetts, Amherst (October 2007)