[SRILM User List] A question about "-lm" and "-output-ctm" when using lattice-tool

[Chen, Lei]

Hi,

  I generated a lattice file, htk.slf, using a LM (lm1) in HTK. This LM
has also been converted to SRI-LM format (lm2).  
  Then, I tried using lattice-tool to decode the lattice.

  W/o using "-lm", the following command generates the lm1.ctm with an
WER identical to the WER obtained in HTK.
  "lattice-tool -read-htk -in-lattice htk.slf -viterbi-decode
-output-ctm > lm1.ctm"

  However, when using " -lm lm2", the obtained lm2.ctm has a much worse
WER compared to HTK's result.
  "lattice-tool -lm lm2  -no-expansion  -read-htk -in-lattice htk.slf
-viterbi-decode -output-ctm > lm2.ctm"
  
 For my understanding, since many parameters (e.g., LM weight and word
penalty) were stored at the header of lattice file, we don't need to
specify them in the command line.

  I have studied manual of lattice-tool for a while but cannot solve
this problem. Will you please help on this.

 Thanks very much.

Lei




  


    


--------------------------------------------------
This e-mail and any files transmitted with it may contain privileged or confidential information.
It is solely for use by the individual for whom it is intended, even if addressed incorrectly.
If you received this e-mail in error, please notify the sender; do not disclose, copy, distribute,
or take any action in reliance on the contents of this information; and delete it from
your system. Any other use of this e-mail is prohibited.

Thank you for your compliance.
--------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.speech.sri.com/pipermail/srilm-user/attachments/20100414/26da0a9f/attachment.html>