Menu

Voice Recognition with Dinamic Grammar File?

Help
2015-07-02
2015-07-03
  • Juan Sebastian Marquerie

    Hello and thanks for reading!
    I am making a text to speech program with the Pocketsphinx library in C, in my Raspberry Pi model 2.
    The program runs 2 recognitors: 1, for the keyword detection; and other for the command itself.
    This has worked pretty well with very good results, but now i am facing a the problem of adding parameters to the command part. This would be sholved by adding a 3 recognitor, but i was thinking if it would be a way to "set" the grammar file to another. Something like this:
    The JSGF File:

    ->grammar 1
    grammar 2

    Now is setted into grammar 1, but now it recognice the command: "configure alarm":
    grammar 1
    ->grammar 2

    So the grammar 2, that contains the structure for number recognition it would be implemented.

    Has enyone tried to made something like this? There is some kind of function that I overlooked?
    Sorry for the bother and thank you!
    (P.S: Sorry if i made some mistakes in the spelling part, my english is not perfect...)

     
    • Nickolay V. Shmyrev

      You can add two differently names searches with ps_set_jsgf_file and another one keyword search with ps_set_kws. You can switch between three in any order with ps_set_search.

       
  • Juan Sebastian Marquerie

    Ok Thank a lot!!

     

Log in to post a comment.