As far as I know, there is no tool that can merge finished training.
IMO for LSMT it should work as incremental training e.g. you used the
output of fine-tuning as input for the next fine-tuning... (I never tried
it ;-) )
Zdenko
so 22. 10. 2022 o 9:36 Umanda Dikwatta napísal(a):
>
> LSTM
> On
LSTM
On Saturday, October 22, 2022 at 1:00:47 PM UTC+5:30 zdenop wrote:
> Legacy or LSTM?
>
> Zdenko
>
>
> so 22. 10. 2022 o 8:09 Umanda Dikwatta napísal(a):
>
>> Hello,
>>
>> I have several traineddata files for different fonts for the same
>> language. How to combine these files to get one tr
Legacy or LSTM?
Zdenko
so 22. 10. 2022 o 8:09 Umanda Dikwatta napísal(a):
> Hello,
>
> I have several traineddata files for different fonts for the same
> language. How to combine these files to get one traineddata file ?
>
> Best Regards
>
> --
> You received this message because you are subs
Hello,
I have several traineddata files for different fonts for the same language.
How to combine these files to get one traineddata file ?
Best Regards
--
You received this message because you are subscribed to the Google Groups
"tesseract-ocr" group.
To unsubscribe from this group and stop
Hi ,
I have two trained data files one is actual one eng.traineddata and second
one is newly created one eng1.traineddata . As you said am doing like
this
C:\Program Files (x86)\Tesseract-OCR>tesseract -l eng+eng1
for me it is not working..kindly help me out here
On Wednesday, 11 Februa
Thanks!
在 2015年2月11日星期三 UTC+8上午11:23:58，shree写道：
>
> You cannot combine two traineddata filesets
> but you can give two traineddata sets for recognition using -l option (for
> languages)
>
> examples
>
> -l eng+myeng
> -l eng+jpn
> -l jpn+eng
> -l hin+san
> -l eng+tam
>
>
> ShreeDevi
> __
You cannot combine two traineddata filesets
but you can give two traineddata sets for recognition using -l option (for
languages)
examples
-l eng+myeng
-l eng+jpn
-l jpn+eng
-l hin+san
-l eng+tam
ShreeDevi
भजन - कीर्तन - आरती @ http:/
Tesseract Version: 3.02
Platform: Ubuntu 14.04 LTS (GNU/Linux 3.13.0-24-generic x86_64)
Tesseract not recognizing few characters in image files.
I did correct improper tesseract recognition for those characters through
learning and have a traineddata file generated. However, I dont want to r
8 matches
Mail list logo
