Training Hidden Markov Models

I am trying to create a sound recognition software using Hidden Markov Models. I am having difficulties training the HMM models. I have created a set of Mel-frequencies coefficients for each of my frames. I also know the which sound is associated with which MFCC. I thought I would use the function "hmmestimate" to create my transition and emission estimates. However, this function assumes the input (seq) is a unique finite state. I am using my MFCCs as the input. I know there is something I am missing, but I can't figure it out.

Thanks for any help Brian

Answer 1

[engstudent]

hi i am working in speech recognition using hmm and i need to know what is the step to build hmm and what is the reguired as i/p ,o/p,and to train hmm in matlab ???????plz help me .....................

Answer 2

[imene s]

i need source code of application of speech recognition using hmm,please help me .