On Training Cascade Face Detectors

In this paper we present two improvements over Viola and Jones [1] training scheme for face detection. The first is 300-fold speed improvement over the training method presented by Viola and Jones [1] with a modest increase in execution time. The second is a principled method for determining a cascade classifier of optimal speed. We present some preliminary results of our methods.

[1]  Omprakash K. Gupta,et al.  Branch and Bound Experiments in Convex Nonlinear Integer Programming , 1985 .

[2]  Ignacio E. Grossmann,et al.  An outer-approximation algorithm for a class of mixed-integer nonlinear programs , 1986, Math. Program..

[3]  Yoav Freund,et al.  Experiments with a New Boosting Algorithm , 1996, ICML.

[4]  Takeo Kanade,et al.  Neural Network-Based Face Detection , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.