Mastering Computer Vision with TensorFlow 2.x

Attentional cascade classifiers

Each of the strong classifiers previously described forms a cascade where each weak classifier represents one stage to quickly remove the negative subwindow and retain the positive subwindow. A positive response from the first classifier implies that the region of the face (for example, the eye region) has been detected and then the algorithm moves on to the next feature (for example, the nose region) to trigger the evaluation of a second classifier, and so on. A negative outcome at any point leads to the immediate rejection of the stage. The following image illustrates this point:

This image shows that the ...