How to build HMM model for continuous speech recognition?

1 vue (au cours des 30 derniers jours)
Ruchi kansara
Ruchi kansara le 11 Mar 2012
Commenté : RITESH le 23 Fév 2014
Hey! i m working on continuous speech recognition. i read everything on Hidden Markov Model(HMM), i m not getting how to implement it in Matlab. Can any 1 help me out on how to build model in matlab? i m also not getting what shud be hmm states and hmm observation for my model.. Thanx in advance
  1 commentaire
RITESH
RITESH le 23 Fév 2014
Are you done with your HMM based Speech Recognizer? of so, can you share the file with me, i mean the source code for reference. I have no idea where to start and how to develop it from scratch.Also may i know how much time it may involve?

Connectez-vous pour commenter.

Réponse acceptée

777
777 le 11 Mar 2012
hmm based speech recognition system are really difficult to built.i worked hard 4 months but could not get it.then i changed it to word recognition system based on psd and filter fingerprints.i dont want to demoralize.
for any other help just google ,you may get some help from there.here you will get help only if you have some basic idea already .
  3 commentaires
Ruchi kansara
Ruchi kansara le 12 Mar 2012
i want to do word spotting in continuous speech, b4 i tried dtw algorithm but with constraint that input speech shud have reasonable pauses in between each word.. that's y i switched 2 hmm.. i read all about hmm but confused what shud be hmm states.. i got idea that v have 2 take hmm states as vocal tract shapes and each state comprising of phonemes as observations.. but how to identify dese for a particular word.. do help me for identifying hmm states and its observations.
RITESH
RITESH le 23 Fév 2014
Are you done with your HMM based Speech Recognizer? of so, can you share the file with me, i mean the source code for reference. I have no idea where to start and how to develop it from scratch.Also may i know how much time it may involve?

Connectez-vous pour commenter.

Plus de réponses (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by