Before the new release of Bob we should remove traces of evaluate.py in the documentation and in the baselines script. Furthermore, we should test this baseline script.

Who is currently working with speaker recognition here? Maybe @amohammadi, @ssarfjoo ?

Related to bob#255