Menu

CMU SLM help

Help
SKR
2016-10-28
2016-11-07
  • SKR

    SKR - 2016-10-28

    Hello,

    I am working with a speech recognition system using CMU Sphinx with SLM language modelling toolkit.

    I am stuck at LM creation at this step: ./binlmcreation -binarya.binlm -arpa a.arpa
    with an error....

    Please help me with this error.
    Reading binary language model from Malayalam.binlm...get_vocab_fm_ht ERROR: the hash table does not contain wordid 1081.

    Thanks in advance....

     
    • Nickolay V. Shmyrev

      Use modern toolkits - mitlm, srilm.

       
  • SKR

    SKR - 2016-10-31

    Okay,, Thank u,,

     
  • SKR

    SKR - 2016-11-07

    Excuse me Sir,,

    Now I am in step of Acoustic modelling for speech recognition using sphinx,.. I got several warnings like this:

    *
    WARNING: This word: ചോറു was in the transcript file, but is not in the dictionary ( ചോറു കുറവാകയാൽ ). Do cases match? *

    *
    WARNING: This phone (f) occurs in the phonelist (/home/adspr/sphinx/Malayalam/etc/Malayalam.phone), but not in any word in the transcription (/home/adspr/sphinx/Malayalam/etc/Malayalamtrain.transcription) *

    Why this?? How can I solve this??

    Its a Malayalam Speech recognition system.
    Sir, How can I create the dictionary?? Now I created it with the help of my vocab file in LM... It doesn't contain some words in my wave (trian) transcription file...
    Do my dictionary contain all words in wave trainscription file?

    Please give me some idea regarding dictionary file...

    Thanks in Advance....

     
    • Nickolay V. Shmyrev

      To ask a question on new topic start a new thread. Do not post to the same topic.

       

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.