Menu

Training Acoustic Models

Help
Leimiaoren
2015-11-08
2015-11-08
  • Leimiaoren

    Leimiaoren - 2015-11-08

    I have been trying to develop a speech to text application so fR I have all the necessary data that I will be needing.

    1) WAV files already segmented 5-30seconds each
    2) Transcript file of the WAV files
    3) phonetic transcription
    4) phoneset
    5) language model (not yet in arpa just a text file)

    I have already installed active perl and python, downloaded sphinxtrain and sphinxbase but can someone enlighten me what does run "configure" "make" and "makefile" means in the website of cmusphinx. I want to start training my acoustic mode but I dont know how or where do I run these files. Someone please help me. in need of help

     
    • Nickolay V. Shmyrev

      configure is used for Linux build. We recommend to use Linux to train models.

      You can find information here:

      http://www.codecoffee.com/tipsforlinux/articles/27.html

      On Windows you can download prebuilt binaries from our downloads.

       
  • Leimiaoren

    Leimiaoren - 2015-11-08

    i will be using windows. Where can i download this prebuilt binaries?

     

Log in to post a comment.