Menu

Recognize only one word... from .wav

Help
Waldemar
2020-05-06
2020-05-06
  • Waldemar

    Waldemar - 2020-05-06

    Hi!
    I have 3 hours .wav file, with a lot of speakers talking hebrew.. I need to count the word "Yes" that was said by one specific person ( maybe use vosk? for speaker diarization ), all other words, sounds need to be ignored.
    So, beside the steps of installation, that I concider to be the following ( correct me if I'm wrong ):
    Ubuntu 2020:
    Install CMUSphinx https://cmusphinx.github.io/wiki/tutorialpocketsphinx/ as it described here
    Install python package pocketsphinx and all dependencies(pyaudio and so on and so forth)


    Train acoustic model for this hebrew word
    (for one word it sound reasonable, but by the way is there an acoustic model phonetic dictionary and all the things that pocketsphinx need for hebrew?):
    I need 3 files for python script: .lm.bin file with language model?
    .dict file?
    and hmm file?
    What is the proper way to that?
    Please refer me to tutorial if it exist...
    Thank You for Your work there at CMU...

     
    • Nickolay V. Shmyrev

      For such short words like "yes" it is impossible to do keyword spotting because false alarm rate is too high. You have to impelment full LVCSR recognizer with speaker separation and search in the results.

      cmusphinx tutorial is here https://cmusphinx.github.io/wiki/tutorial

       
      • Waldemar

        Waldemar - 2020-05-06

        Thank's a lot for Your answer, Nickolay.

         

Log in to post a comment.

Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.