[digikam] [Bug 263400] Newly detected faces should be stored in a dedicated virtual album for checking

bugzilla_noreply at kde.org bugzilla_noreply at kde.org
Mon Jan 16 19:02:01 GMT 2017


caulier.gilles at gmail.com changed:

           What    |Removed                     |Added
   Version Fixed In|                            |5.5.0
         Resolution|DUPLICATE                   |FIXED

--- Comment #4 from caulier.gilles at gmail.com ---
Git commit 7a7c29c0b8b36bd30a190a6cb4dc5f43960034a8 by Mario Frank.
Committed on 16/01/2017 at 18:24.
Pushed by mfrank into branch 'master'.

I introduced the new special tag Unconfirmed which is automatically created if
not present. In detection phase, faces are marked as unknown as usual.
In recognition phase, the faces are marked with the recognised face tag, as
usual. But additionally, they are connected as autodetected person to the
Unconfirmed tag. In this property, I store the id of the person as which the
face was recognised and also the
face region.

With small adoptions in the image lister all recognised faces are now shown
both in the face tag which was recognised (person A) and the Unconfirmed tag
(with the suggested tag being person A).

Confirming the person can now be done in both tags. And the confirmed identity
is set as expected. Confirming the face also removes the face from unconfirmed
Rejecting the face also removes the face from both the recognised tag and also
from the Unconfirmed tag. Rescanning for faces (detection) also removes the
to the Unconfirmed tag if unconfirmed results should be thrown away.
FIXED-IN: 5.5.0

M  +3    -1    NEWS
M  +10   -0    libs/database/coredb/coredbconstants.cpp
M  +2    -0    libs/database/coredb/coredbconstants.h
M  +21   -3    libs/database/item/imagelister.cpp
M  +25   -0    libs/database/tags/facetags.cpp
M  +3    -0    libs/database/tags/facetags.h
M  +15   -1    libs/database/tags/facetagseditor.cpp
M  +12   -0    libs/database/tags/facetagsiface.cpp
M  +5    -0    libs/database/tags/facetagsiface.h


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

More information about the Digikam-devel mailing list