[Bug 295334] New: Add some way to tag a face result as "ignore this one"

https://bugs.kde.org/show_bug.cgi?id=295334

            Bug ID: 295334
          Severity: wishlist
           Version: unspecified
          Priority: NOR
          Assignee: [hidden email]
           Summary: Add some way to tag a face result as "ignore this one"
    Classification: Unclassified
                OS: Linux
          Reporter: [hidden email]
          Hardware: Compiled Sources
            Status: UNCONFIRMED
         Component: Face Management
           Product: digikam

I regularly relaunch a full face scan on my library (in order to ensure that
all my pictures have been correctly searched, which is not always the case),
but my problem is that the face detection plugin detects quite often "things"
that may look like faces (clouds, fireplace, drawings on clothes or on
walls...), as well as faces of unkwown people that happen to appear on my
pictures.
A truly very good option would be, in those cases, to have (instead of the
"remove tag button") an option to ignore the result, so that it would not be
proposed on the next scan.
Currently, for that purpose, I create two tags : "Nobody" and "Other people",
the first one being for misdetections, and the second one for the people I
don't want to name.

A solution would be to record the tag information in the database, but not in
the image (since it would then appear as a tag in any other software I may
use), and to filter the result based on that information.

Thanks if this function appears someday!

Frédéric.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Marcel Wiesweg]

https://bugs.kde.org/show_bug.cgi?id=295334

Marcel Wiesweg <[hidden email]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[hidden email]

--- Comment #1 from Marcel Wiesweg <[hidden email]> ---
I am tempted to say "Why not use the option to rescan only unscanned pictures"
but I see some sense in that wish assuming a new algorithm became available, or
you changed search settings.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

https://bugs.kde.org/show_bug.cgi?id=295334

--- Comment #2 from Frederic Grelot <[hidden email]> ---
Yep, that is one case to consider (I don't know is the algorithm still evolves,
but I think that I see some new faces from time to time), together with the
case where, due to some bugs (like https://bugs.kde.org/show_bug.cgi?id=281792
for example), some faces are lost.
Since I still regularly experience bugs with face recognition (and you must
notice it : everytime I decide to tag my new photos in my collection, few bugs
arise...), I still prefer to run the full search occasionnaly...

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=295334

Gilles Caulier <[hidden email]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[hidden email]

--- Comment #3 from Gilles Caulier <[hidden email]> ---
On the top right corner of face thumb, there is a button to reject face with
the tooltip "This is not a face, click to reject it".

This button do not solve your problem ?

Please test with last digiKam 4.2.0.

Thanks in advance

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=295334

--- Comment #4 from Gilles Caulier <[hidden email]> ---
Marcel,

"Face Rejection overlay" is not the purpose to "Ignore this face" feature ?

Gilles

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Marcel Wiesweg]

https://bugs.kde.org/show_bug.cgi?id=295334

--- Comment #5 from Marcel Wiesweg <[hidden email]> ---
Well it deletes the face, I believe the post is about storing the information
that this face shall be ignored when rescanning in the future. Which is not
implemented.

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[Gilles Caulier-4]

https://bugs.kde.org/show_bug.cgi?id=295334

Gilles Caulier <[hidden email]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Version|unspecified                 |3.5.0
            Summary|Add some way to tag a face  |Add some way to tag a face
                   |result as "ignore this one" |result as "ignore this one"
                   |                            |in database for future
                   |                            |scans

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=295334

Jens B. Benecke <[hidden email]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jens-bugs.kde.org@spamfreem
                   |                            |ail.de

--- Comment #6 from Jens B. Benecke <[hidden email]> ---
+1 from my side too. I want to be able to ignore (for all future face scans)

- faces that I do not want to be recognized (eg. complete strangers which only
appear on one photo anyway)
- stuff that is not a face or animal faces that are accidentally recognized.

I also would like to be able to search pictures that have these flags set (e.g.
"images with ignored faces").

Thank you!

--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Digikam-devel mailing list
[hidden email]
https://mail.kde.org/mailman/listinfo/digikam-devel

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=295334

[hidden email] changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[hidden email]

--- Comment #7 from [hidden email] ---
I'm willing to work on this issue given minimal direction from core developers.

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=295334

--- Comment #8 from [hidden email] ---
To implement this feature, I think that "ignored" faces will have to be
retained as Identities in the `Identities` table. To mark as ignored, we could
either use the `IdentityAttributes` or possibly use the `Identities.type`.
However `Identities.type` is completely unused at the moment, and is not mapped
through the `Identity` class.

Given the current code, and the user story of wanting to "search for images
with ignored faces", I think using an attribute (in the `IdentityAttributes`
table) would be the preferred way of persisting ignored status.  Any
disagreement or alternate ideas before I begin implementing?

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=295334

--- Comment #9 from [hidden email] ---
yes, IdentityAttributes sound like a good candidate to implement this feature.
No objection from me.

Gilles Caulier

--
You are receiving this mail because:
You are the assignee for the bug.

[bugzilla_noreply]

https://bugs.kde.org/show_bug.cgi?id=295334

[hidden email] changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|Faces-Management            |Database-Faces

--
You are receiving this mail because:
You are the assignee for the bug.