Changing Language Model doesn't work the second time.

Home › Forums › OpenEars › Changing Language Model doesn't work the second time.

Viewing 3 posts - 1 through 3 (of 3 total)
Advertisement: “RuleORama is an OpenEars™ plugin that lets you create rules-based grammars for fixed phrase recognition, fast enough for RapidEars!”

  • Author
    Posts
  • December 18, 2011 at 12:29 am #8250
    aerialcombat
    Participant

    I’m creating language model files dynamically.

    The first “StartListeningWith…” (Model A) works great with no problem.

    When I “ChangeLanguageModelToFile” (Model B), It changes without a problem also.

    But, when I try to change the Language Model for the second time (to Model C) it starts having recognition problems. It keeps getting empty hypothesis back.

    I’m wondering, is it designed this way, or is it a programming error on my part?

    I tried suspending/resuming while changing language models, etc. I couldn’t get it to work the second time changing around. I’m not using threaded model. If it worked the first time, I don’t see why second time should be a problem.

    Please help. Thank you.

    December 18, 2011 at 12:50 am #8251
    aerialcombat
    Participant

    I seemed to have answered my own question. I had to change LM File names between language models. Thanks.

    December 20, 2011 at 11:42 pm #8254
    Isaac Squires
    Participant

    Ah! I was having the same problem, I’ll try that out.

  • Author
    Posts
Viewing 3 posts - 1 through 3 (of 3 total)
  • You must be logged in to reply to this topic.