Face Tracking in Video

In a face tracking application, Boosting and cascading detectors have gained great popularity due to the their efficiency in selecting features for faces. However these detectors suffer from high miss-classification rates. In addition they depend on the orientation of the face. They require that the face be in a full frontal position. If there is any deviations from this position (the face is tilted by 45 degrees, or the face turns to a profile position) the AdaBoost fails to detect a face.