Since Burmese is not currently a supported language for Android's built-in speech recognition, PocketSphinx may be your best bet. Unfortunately, I do not believe that there is an existing language model for Burmese, so you will need to create a regular Burmese language model, which requires a large text text and linguistic experience.
CMUSphinx - Language Building Model
... in addition, you will need to train an acoustic model that requires a lot of recorded sound. For general dictation (i.e. Speech for general-purpose text that many different people can use), Sphinx project people recommend that you have 50 hours of recording from 200 different speakers.
CMUSphinx - Acoustic Training Model for CMUSphinx
I had good experience using PocketSphinx on Android, but only for a small language model (i.e. for command and control application types) and only in English. This can be done for the Burmese, but it will take a lot of work to create and train the models mentioned above.
source share