There is no voice recognition at all

Eyal
2011-09-08
2013-01-13
  • Eyal
    Eyal
    2011-09-08

    **Hi All,

    When I try to activate one of the scenarios (for example: Firefox), I've got the follow popup error, I quote:**

    The recognition reported the following error:
    The recognition could not be started because your model contains words that consists of sounds that are not covered by your acoustic model.

    You need to either remove those words, transcribe them differently or train them.

    Warning: The latter will not work if you are using static base models!

    This could also be a sign of a base model that uses a different phoneme set than your scenario vocabulary.

    The following words are affected (list may not be complete):
    Input, Browser

    The following phonemes are affected (list may not be complete):
    aw-z+er, ih-n+p, n-p+uh, r-aw+z,

    STAT: include config: /home/eyal/.kde/share/apps/simond/models/test/active/julius.jconf
    STAT: jconf successfully finalized
    STAT: *** loading AM00 _default
    Stat: init_phmm: Reading in HMM definition
    Stat: rdhmmdef: ascii format HMM definition
    Stat: rdhmmdef: limit check passed
    Stat: check_hmm_restriction: an HMM with several arcs from initial state found: "sp"
    Stat: rdhmmdef: this HMM requires multipath handling at decoding
    Stat: init_phmm: defined HMMs: 4382
    Stat: init_phmm: loading ascii hmmlist
    Stat: init_phmm: logical names: 6075 in HMMList
    Stat: init_phmm: base phones: 44 used in logical
    Stat: init_phmm: finished reading HMM definitions
    STAT: making pseudo bi/mono-phone for IW-triphone
    Stat: hmm_lookup: 1026 pseudo phones are added to logical HMM list
    STAT: *** AM00 _default loaded
    STAT: *** loading LM00 _default
    STAT: reading  and …
    Error: voca_load_htkdict: line 4: triphone "ih-n+p" not found
    Error: voca_load_htkdict: line 4: triphone "n-p+uh" not found
    Error: voca_load_htkdict: the line content was: 2  ih n p uh t
    Error: voca_load_htkdict: line 24: triphone "r-aw+z" not found
    Error: voca_load_htkdict: line 24: triphone "aw-z+er" not found
    Error: voca_load_htkdict: the line content was: 10  b r aw z er
    Error: voca_load_htkdict: begin missing phones
    Error: voca_load_htkdict: aw-z+er
    Error: voca_load_htkdict: ih-n+p
    Error: voca_load_htkdict: n-p+uh
    Error: voca_load_htkdict: r-aw+z
    Error: voca_load_htkdict: end missing phones
    Error: init_voca: error in reading /home/eyal/.kde/share/apps/simond/models/test/active/model.dict: 2 words failed out of 33 words
    ERROR: failed to read dictionary "/home/eyal/.kde/share/apps/simond/models/test/active/model.dict"
    ERROR: m_fusion: some error occured in reading grammars
    ERROR: j_create_instance_from_jconf: error in loading model

    I solved the error by copy as you can see:
    cp  ./Julius_AcousticModels_16kHz-16bit_MFCC_O_D_\(0_1_1-build726\)/sample.dict /home/eyal/.kde/share/apps/simond/models/test/active/model.dict

    now I can activate simon but,
    There is no voice recognition at all - Is someone can help me to run simon correctly

    PS. I did not added / deleted / updated or what ever any word. all of the words is defaults!

    Best Regards,
    Eyal

     
  • Peter Grasch
    Peter Grasch
    2011-09-08

    Hi Eyal!

    Your "fix" to the previous problem, destroyed (acually overwrote) your dictionary. That's why you don't get any recognition…
    To restore your dictionary simply change *something* (add a new scenario or something like that) and it will be overwritten again.

    As for your initial problem: I think you may be using a very old version of the Voxforge acoustic model.

    Please try again with a current version:
    http://www.repository.voxforge1.org/downloads/Nightly_Builds/current/

    Best regards,
    Peter

     
  • Eyal
    Eyal
    2011-09-09

    **Hi Peter,

    The above Error  - Fixed.
    You are totally right, my acoustic model was too old.**

    Fixed description:
    I download the file
    HTK_AcousticModel-2010-12-16_16kHz_16bit_MFCC_O_D.tgz from above link, 
    I opened and directed simon again:  configur simon -> 
           model setting: general model
           Base definition: all files directed to the new downloaded acoustic model.
    I chose a new scenario and … no error occur.

    But still There is no voice recognition at all

    Thank you,
    Eyal

     
  • Eyal
    Eyal
    2011-09-14

    There is no voice recognition at all
    FIXED

    The way:
    I have linux 11.04 with gnume.
    I installed packages kubuntu-desktop and kubuntu-default-settings
    thats all!

    Thanks all of of you and specially to Mr. Peter.
    Eyal