hmm-persian-ocr's Introduction

Kian Paimani aka `kianenigma`

Wisdom

 _________________________________________
/ October.                                \
|                                         |
| This is one of the peculiarly dangerous |
| months to speculate in stocks in.       |
|                                         |
| The others are July, January,           |
| September, April, November, May, March, |
| June, December, August, and February.   |
|                                         |
| -- Mark Twain, "Pudd'nhead Wilson's     |
\ Calendar"                               /
 -----------------------------------------
        \   ^__^
         \  (oo)\_______
            (__)\       )\/\
                ||----w |
                ||     ||

Stats

hmm-persian-ocr's People

Contributors

Stargazers

Watchers

hmm-persian-ocr's Issues

I have run the test example, but I got a quite different result..

I think there must be something wrong but I don't know where it is. I have followed your readme instructions to install all the dependencies. Then I ran the test example and get the following result:

====================== HTK Results Analysis =======================
  Date: Mon Jan  9 16:17:27 2017
  Ref : samplesRef.mlf
  Rec : res0.mlf
------------------------ Overall Results --------------------------
SENT: %Correct=3.00 [H=3, S=97, N=100]
WORD: %Corr=3.00, Acc=3.00 [H=3, D=0, S=97, I=0, N=100]
===================================================================

I can't figure out why the accuracy would be so low. And I notice that you mentioned in the Usage example that test/text_repo contains a set of 1000 persian word. But I find the folder I download contains 3315 lines of persian word. Does that have something to do with the low accuracy?

Recommend Projects