We saw detection of various objects in the previous chapter. The human face is much more complicated than just a regular color object, for example. More complex detectors share in common such things like the usage of brightness information and the patterns that this information forms. First of all, we need to see how face recognition is done. Without that, the tracking process will be quite difficult to understand. Actually, in most cases, the face recognition part is just the first step of face tracking algorithms.
We start with the JSFeat project. This awesome library provides a functionality to detect a face in two ways. Both have many applications in the real world. We will see how both of them work from the inside and discuss the API provided by JSFeat.