[Project-ideas] Queries on Indic OCR [was:Re: (no subject)]

[Sankarshan Mukhopadhyay]

On Fri, Mar 30, 2012 at 7:47 PM, Priyanshu Sharma
<priyanshu.iitropar at yahoo.com> wrote:
> hi
> Myself Priyanshu sharma from iit ropar. 3rd year computer student.
> i am interested in following project:-
>                        ""Improve the accuracy of OCR tools for Bengali
> language to 98%""
>
>
> @. I wanted to know what are the key points/technology you are looking for
> this project.

I would have thought that the problem statement makes it adequately
clear - the accuracy and confidence of the OCR tools used upon
digitized Bengali texts need to have a confidence level of around 98%

> @. I have read few archive and found that there is one more student applying
> for this project and he has already submitted his proposal.Should i apply
> for this project,because i think that his proposal is good.and he has
> already done so much.Is there something missing in his proposal which i can
> do/improve and thus get selected for GSoC.

That is an exercise for you.

> @. Also tell me that how selection procedure is done.and what is the
> priority of this proposal.
> if priority is low then i should in fact move to some other organization as
> i am quite late now.
> please suggest me.

Once the gates for proposal submissions close, the mentors and
reviewers would be working closely with the candidates to iron out the
remaining wrinkles and, the proposals would be scored on feasibility
of implementation. The organization holds this proposal in some
importance, so the priority is reasonably high.

/sankarshan

ps: The next time, please have a better subject line. It helps when
scanning through the emails and deciding which one requires an
immediate response.


-- 
sankarshan mukhopadhyay
<http://sankarshan.randomink.org/blog/>