In this talk I will outline a relatively new method of real time face detection with application to an unencumbered virtual reality system. The method uses a learning algorithm called Cascaded AdaBoost, and a very large set of simple features. Further I will discuss a method for producing "optimal" speed face detectors and a method for greatly improving the training time of these classifiers.
Last modified: Friday, 16-Jul-2004 15:37:32 NZST
This page is maintained by Caroline Wills.