I'm working in a project of speech recognition, I 'm in the clustering step, I segmented
the phonemes into equal duration segments, then I started the binary splitting algorithm, then
the k-means but the resulting clusters are groups of phonemes too far from each others.
   
  Have I to do some thing like an alignment on the phonemes before applying the binary
splitting?
   
  if yes, how to do so?
	  please give me a help in that problem
	

(You need to be a member of speech-recognition -- send a blank email to speech-recognition-subscribe@yahoogroups.com )