Computation of segment Likelihood functions using PPT
In (14.2), the PDFs
are the segment PDFs
conditioned on knowing the segment subclass (the
segment subclasses are assumed to be fixed for each
segmentation ). These are obtained using the PPT (equation 2.2)
and a feature transformation specific to subclass
on a segment of length base segments:

(14.4) 
where
is the feature used
for a segment of size and subclass ,
is the PDF of that feature assuming subclass ,
and
is the PDF of that feature
under the reference hypothesis.
This calculation has been described for a range of different
features (See chapters 4
through 10).
The MRHMM requires a large amount of
frontend processing to calculate all the necessary features
and PDFs for all segments (seen in Figure 14.1 ``Available segments")
assuming each possible subclass.
Baggenstoss
20170519