Saturday, February 11, 2006

Start updating my thesis again

I start working again after my lazy time for a month.
Here are the works that I have done so far:
  • I already got the result from CART (which was run on login server since X'mas). So I wrote the program to extract the pronunciations for each names from CART's output format.
  • Then I used the python program that I wrote last year to make the wave file for subjective evaluation.
  • We have 4 methods to compare: multistrategy PbA, CMU, Table lookup II, and CART.
  • We randomly chose 150 names from each method that pronounce different from the other methods. That means it come up with 600 names in total.

Here are the next things that I need to do:

  • Find the edit distances for each names (only the ones that different from the others)
  • Rewrite program for multistartegy PbA to run with BEEP lexicon.
  • Prepare for subject evaluation: sheet, wmp file, email, and keys for sound proof room.

0 Comments:

Post a Comment

<< Home