Hi,
I am trying to find face points from the captured camera image. I have seen into some of the samples and blogs like MUCT (http://www.milbo.org/muct/) and also read about Active Appearance Models and Active Shape Models. In all the documentation and blogs, it is mentioned that we need to annotate some of the training images with landmark points of mouth and eyes for PCA analysis.
But can we find landmark points directly from the image.