NEWS

0.1 is officially released. Work on 0.2 already started!

New build file for examples. They are automatically built if -DBUILD_EXAMPLES=1 set when 
running cmake.

The clustering in face detection works *perfect* now. All faces are detected in my database.
A "genuineness" has been assigned to each face, which is the number of duplicate detections that face had.
We can remove faces that have less than a fixed number of duplicates.
This threshold has been kept tunable for future optimizations, though the need for tweaking this is unlikely.
From now on, it might be okay if no code is written for improving face detection accuracy, I think tuning these parameters is enough.

Now, Face Detection works for every non-microscopic frontal face, period.
Detecting non-frontal faces is now a trivial task of adding one more line :
cascadeSet->addCascade("profile-blah-blah.xml");

We have face detection with Haar features to work. I think this is 
indeed a milestone for the project and should be celebrated. Well
done to everyone that got involved this far.

Next will be refining Face Recognition :) 

Eigenfaces should be working now. Althought this needs testing. Don't
know how good it is. But it works :)