I would like to create a word model (Dictation mode) using sphinx train ...
whether any changes in config file... whether HMM states varies or fixed (3 or
5) for all words ( long and short words)
Sorry, it's hard to understand your question. You probably want to spend more
time to formulate it properly.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
ok....
I would like to create a word model(1) using sphinx train
1. whether any changes is required in training config file
2. whether HMM states varies or fixed for long and short words
3. if I train the same words with triphone (2) .which model 1 or 2 gives better accuracy
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I would like to create a word model (Dictation mode) using sphinx train ...
whether any changes in config file...
whether HMM states varies or fixed (3 or 5) for all words ( long and short
words)
Sorry, it's hard to understand your question. You probably want to spend more
time to formulate it properly.
ok....
I would like to create a word model(1) using sphinx train
1. whether any changes is required in training config file
2. whether HMM states varies or fixed for long and short words
3. if I train the same words with triphone (2) .which model 1 or 2 gives better accuracy
Do you mean you want to use word-based phoneset?
Yes, you need to use another phoneset
In sphinxtrain number of states per phone is fixed
Triphone model gives better accuracy. See also http://cmusphinx.sourceforge.n
et/wiki/tutorialam#data_preparation