HI, I am trying to do a Speech recognition testing .

Now, I can get a word lattice using HVITE , tri-phone based model.

I was wondering , how could I generate a phone lattice from a word lattice based on pronunciation dictionary?

Does anybody can help? Many thanks in advance!

--- (Edited on 5/16/2013 2:59 am [GMT-0500] by hyguo91) ---