Distinguishing features in f0:
As an example, let’s look at a maximally
simple classifier on f0 data:
If we just look at average and slope:
how well do average and slope distinguish
between the four utterance types?
are there dialect differences?