Dear sir,
i have an idea for improve the accuracy when the speaker switch context
between 2 or more domain.
for each domain, i create one language model. At runtime, the posterior
probability of input speech was calculated for each language model separated.
Compare two result and get higher probability result.
that still be an idea and i need to evalute it. Could i use pocketsphinx to do
that ? is it possible ?
thank in advance.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Dear sir,
i have an idea for improve the accuracy when the speaker switch context
between 2 or more domain.
for each domain, i create one language model. At runtime, the posterior
probability of input speech was calculated for each language model separated.
Compare two result and get higher probability result.
that still be an idea and i need to evalute it. Could i use pocketsphinx to do
that ? is it possible ?
thank in advance.
Yes, see ps_get_prob function call.