digiKam › digikam-devel

Re: Request to be mentored for digikam face-recognition

Classic

List

Threaded

7 messages Options

kunal ghosh

Re: Request to be mentored for digikam face-recognition

hi,
i have been pondering over other prospective digikam based projects for SoK . Related to image recognition in some way or the other :)

one would be to integrate KROSS plugin into digikam
and the other would be to have full scripting support in digikam
like "for all the the photos in X album search for X person" or "when new images are loaded into digikam, do <this> <this> and <this>"
3rd use case for scripting : "Search for all images where person A and B were together" OR "search for the images where person
A was at place <X> " the last idea would use object detection.

Another Idea: duplicate photo detection, to save space

another idea: search images by color
another idea: geo tag images based on monuments in images, location specific features etc.

On Sun, May 2, 2010 at 9:04 AM, kunal ghosh <[hidden email]> wrote:

sure michael,
I will look into other parts of digikam/kipi-plugins as well.

regards,

On Sun, May 2, 2010 at 1:22 AM, Michael G. Hansen <[hidden email]> wrote:

Hi Kunal,

On 05/01/2010 08:24 AM, kunal ghosh wrote:

hi michael, i am very keen to participate in SoK. Could i take up digikam
itself
and ask help from the community (digikam-devel) or KDE-devel instead of
being mentored.

I think you need an assigned mentor to be able to complete SoK. If we can find one, that should work, but ...

I could take up integrating KROSS framework into digikam and then use the
same to implement my face-recognition code.

... it would not be good to have two people working on implementing the same feature in a different way, only to discard one in the end. So if you want would like to participate in digikam development as part of SoK on some other part, we will continue to try to find a mentor. Please let us know whether you could imagine to work on a different part of digikam/kipi-plugins.

Michael

I would submit progress reports from time to time.
Looking for a positive reply.

regards,

On Sat, May 1, 2010 at 12:35 AM, Michael G. Hansen<[hidden email]> wrote:

Hi Kunal,

I talked to possible mentors in the digikam team and we do not currently
have the resources to mentor another student. Maybe you would like to work
on a different project, for another application, during SoK?

@possible mentors: Is there somebody who still has a project for which he
would need a student?

Michael

_______________________________________________
Kde-soc-mentor mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/kde-soc-mentor

--
Michael Hansen - [hidden email]
http://www.pfna.de/

--
regards
-------
Kunal Ghosh
Dept of Computer Sc. & Engineering.
Sir MVIT
Bangalore,India

Quote:
"Ignorance is not a sin, the persistence of ignorance is"
--
"If you find a task difficult today, you'll find it difficult 10yrs later too !"
-----
"Failing to Plan is Planning to Fail"

Blog:kunalghosh.wordpress.com
Website:www.kunalghosh.net46.net
V-card:http://tinyurl.com/86qjyk

--
regards
-------
Kunal Ghosh
Dept of Computer Sc. & Engineering.
Sir MVIT
Bangalore,India

Quote:
"Ignorance is not a sin, the persistence of ignorance is"
--
"If you find a task difficult today, you'll find it difficult 10yrs later too !"
-----
"Failing to Plan is Planning to Fail"

Blog:kunalghosh.wordpress.com
Website:www.kunalghosh.net46.net
V-card:http://tinyurl.com/86qjyk

_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

Arnd Baecker

Re: Request to be mentored for digikam face-recognition

Hi Kunal,

On Sun, 2 May 2010, kunal ghosh wrote:

> hi,
> i have been pondering over other prospective digikam based projects for SoK
> . Related to image recognition in some way or the other :)
>
> one would be to integrate KROSS plugin into digikam
> and the other would be to have full scripting support in digikam
> like "for all the the photos in X album search for X person" or "when new
> images are loaded into digikam, do <this> <this> and <this>"
> 3rd use case for scripting : "Search for all images where person A and B
> were together" OR "search for the images where person
> A was at place <X> " the last idea would use object detection.

To me KROSS seems indeed very interesting with many possible applications.
Have a look at
http://bugs.kde.org/show_bug.cgi?id=146866
where this has been discussed, though I don't know about
the present status of KROSS itself.

Best, Arnd
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

Marcel Wiesweg

Re: Request to be mentored for digikam face-recognition

In reply to this post by kunal ghosh

Scripting in digikam is indeed an interesting idea.
The question is which scripting technology is to be chosen:
Is KROSS the best candidate? Is it under active development?
Or is maybe QtScript the better solution? How does plasma handle its bindings,
what does Amarok use?
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

kunal ghosh

Re: Request to be mentored for digikam face-recognition

In reply to this post by Arnd Baecker

Hi arnd,
Since the inclusion of kross into kdelibs (KDE4) i donot see any development ( Or i could not find the url etc where new
development is recorded).But its inclusion in Kdelibs and so many kde projects using it is quite reassuring.

So, how would you like the below mentioned ideas ?
Is the proposed task sufficient for SoK or should i also write a plugin (using kross) for digikam after integrating kross.

regards.

On Sun, May 2, 2010 at 12:03 PM, Arnd Baecker <[hidden email]> wrote:

Hi Kunal,

On Sun, 2 May 2010, kunal ghosh wrote:

> hi,
> i have been pondering over other prospective digikam based projects for SoK
> . Related to image recognition in some way or the other :)
>
> one would be to integrate KROSS plugin into digikam
> and the other would be to have full scripting support in digikam
> like "for all the the photos in X album search for X person" or "when new
> images are loaded into digikam, do <this> <this> and <this>"
> 3rd use case for scripting : "Search for all images where person A and B
> were together" OR "search for the images where person
> A was at place <X> " the last idea would use object detection.

To me KROSS seems indeed very interesting with many possible applications.
Have a look at
http://bugs.kde.org/show_bug.cgi?id=146866
where this has been discussed, though I don't know about
the present status of KROSS itself.

Best, Arnd
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

kunal ghosh

Re: Request to be mentored for digikam face-recognition

In reply to this post by Marcel Wiesweg

On Sun, May 2, 2010 at 7:11 PM, Marcel Wiesweg <[hidden email]> wrote:

> hi,
> i have been pondering over other prospective digikam based projects for SoK
> . Related to image recognition in some way or the other :)
>
> one would be to integrate KROSS plugin into digikam
> and the other would be to have full scripting support in digikam
> like "for all the the photos in X album search for X person" or "when new
> images are loaded into digikam, do <this> <this> and <this>"
> 3rd use case for scripting : "Search for all images where person A and B
> were together" OR "search for the images where person
> A was at place <X> " the last idea would use object detection.

Scripting in digikam is indeed an interesting idea.
The question is which scripting technology is to be chosen:

Is KROSS the best candidate? Is it under active development?

That is a problem, after extensive search i have come to a conclusion that, unfortunately the Kross framework is not
under development anymore.

Or is maybe QtScript the better solution? How does plasma handle its bindings,

QtScript is Javascript based. IMHO python is a better scripting language (Python is beginner friendly implies more new script
writers). That's why i proposed KROSS+Python.

Also PythonQt (not pyQt) looks like a good scripting framework.
http://doc.trolltech.com/qq/qq23-pythonqt.html
http://sourceforge.net/projects/pythonqt/

Plasma allows for Python as a language to write plasmoids too.
http://techbase.kde.org/Getting_Started/Build/KDE4/Python_Support

what does Amarok use?

Amarok uses QtScript for scripting.

_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

Gilles Caulier-4

Re: Request to be mentored for digikam face-recognition

2010/5/2 kunal ghosh <[hidden email]>:

>
>
> On Sun, May 2, 2010 at 7:11 PM, Marcel Wiesweg <[hidden email]>
> wrote:
>>
>> > hi,
>> > i have been pondering over other prospective digikam based projects for
>> > SoK
>> > . Related to image recognition in some way or the other :)
>> >
>> > one would be to integrate KROSS plugin into digikam
>> > and the other would be to have full scripting support in digikam
>> > like "for all the the photos in X album search for X person" or "when
>> > new
>> > images are loaded into digikam, do <this> <this> and <this>"
>> > 3rd use case for scripting : "Search for all images where person A and B
>> > were together" OR "search for the images where person
>> > A was at place <X> " the last idea would use object detection.
>>
>> Scripting in digikam is indeed an interesting idea.
>> The question is which scripting technology is to be chosen:
>>
>> Is KROSS the best candidate? Is it under active development?
>
>
> That is a problem, after extensive search i have come to a conclusion that,
> unfortunately the Kross framework is not
> under development anymore.
>
>> Or is maybe QtScript the better solution? How does plasma handle its
>> bindings,
>
> QtScript is Javascript based. IMHO python is a better scripting language
> (Python is beginner friendly implies more new script
> writers). That's why i proposed KROSS+Python.
>
> Also PythonQt (not pyQt) looks like a good scripting framework.
> http://doc.trolltech.com/qq/qq23-pythonqt.html
> http://sourceforge.net/projects/pythonqt/
>
> Plasma allows for Python as a language to write plasmoids too.
> http://techbase.kde.org/Getting_Started/Build/KDE4/Python_Support
>
>>
>> what does Amarok use?
>
> Amarok uses QtScript for scripting.
>

As digiKam is Amarok like for Photo and video, i prefer to see
QtScript used in digiKam. Also, as Kross is not maintained anymore,
it's a non sense to use it...

Note : i can try to mentor kunal...

Gilles Caulier
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

kunal ghosh

Re: Request to be mentored for digikam face-recognition

As digiKam is Amarok like for Photo and video, i prefer to see
QtScript used in digiKam. Also, as Kross is not maintained anymore,
it's a non sense to use it...

QtScript looks quite good. Lets go ahead and integrate it in digikam.
A nice tutorial on QtScript.
http://www.youtube.com/watch?v=pMYFY8N7bUM

Note : i can try to mentor kunal...

It would be great !
when can we meet on IRC (UTC) ? to discuss the QtScript integration in digikam.
my irc nick is gancient

Gilles Caulier

_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel