one additional follow-up NBest hypothesis

Home Forums OpenEars one additional follow-up NBest hypothesis

Viewing 4 posts - 1 through 4 (of 4 total)

  • Author
    Posts
  • #1032912
    zhoul22
    Participant

    Hi Halle,

    Thanks for responding to my forum question regarding to n-best hypothesis previously. I fully understand you might not be able to have time to write sample code. However, I would highly appreciate I could receive a little bit further guidance.

    I did set the OEPocketsphinxController.sharedInstance().reterunNbest to be true and nBestNumber property to be 2.

    The property is set right after the sharedinstance is set active(true) in viewcontroller. I verified this by setting verbosePocketSphinx to be true and I could receive extensive logging.

    I did my best to review the documentations and double-checked the OEEventsObserver delegate and callbacks portion. It seems to be me that the sampleswift code has all set up properly (headers etc).

    Therefore, I added the following delegate method of OEEventsObserver in the viewController. However, the output window report nothings when app is running.

    func pocketsphinxDidReceiveNBestHypothesisArray(_ hypothesisArray: [Any]!) {
    print(“Local callback test)”)
    }

    Thanks a lot for your support.

    #1032913
    Halle Winkler
    Politepix

    I would first start with the unmodified sample app from the distribution, and make sure basic recognition works for you for some utterance which is included in the language model, i.e. ruling out that the issue isn’t with recognition in general rather than nbest specifically. Then turn on nbest and set a valid nbest number, and turn on null hypotheses so it is also called even if there isn’t a match with the language model, and see if that helps. Print the actual array in the callback so you can see if null hyps are being returned.

    The most powerful tool you can apply to your own troubleshooting is turning on all the applicable forms of logging documented in the docs and headers – these will show you what the engine is perceiving and whether there is a difference between that and what is ending up on the callback, as well as telling you if your nbest settings are passing validation.

    Good luck!

    #1032914
    zhoul22
    Participant

    Thank you for the prompted response. I followed your instruction (in first paragraph) and it did work!

    #1032916
    Halle Winkler
    Politepix

    Fantastic!

Viewing 4 posts - 1 through 4 (of 4 total)
  • The topic ‘one additional follow-up NBest hypothesis’ is closed to new replies.