[digiKam-users] status of the face detecting?

Christoph Knauer christoph.knauer at web.de
Sat Jun 13 19:18:32 BST 2020


Hi,

I think, it's more important to move the release-date a little, so that it 
could be possible to get it with a useable face detecting.

Greetz

Am Montag, 8. Juni 2020, 12:39:08 CEST schrieb Gilles Caulier:
> Hi,
> 
> Sorry for the delay, but we are busy with students working while
> Google Summer of Code this year, about...
> 
> ...Face Engine (DNN) and Face Workflow (GUI)...
> 
> The students work and progress well. There will be a blog about the
> project advancement soon. One has just posted this one :
> 
> https://kart1kx.wordpress.com/2020/06/05/reordering-the-people-sidebar-in-di
> gikam/
> 
> and so yes, to respond to your original message, dysfunctions still
> here in current 7.0.0. This is why we have 2 students this year
> working on faces management.
> 
> So the optimisations and improvements will appear after this summer,
> and after we backport students branch the master code.
> 
> Now, I open an important question for everyone in this room (students,
> end users, developers) :
> 
> "It's judicious to release digiKam 7.0.0 following the planned date
> before this summer, or we need to wait a little bit after this summer
> to integrate all faces management fixes ???"
> 
> I'm waiting for constructive responses.
> 
> Thanks in advance
> 
> Gilles Caulier
> 
> Le jeu. 4 juin 2020 à 13:47, Tóth Csaba <ignis at domen.hu> a écrit :
> > Hello!
> > 
> > I'm using the Build date: May 22 2020 (target: RelWithDebInfo) version
> > on windows 10.
> > 
> > In the 7.0 Beta1 (?) 2(?) (in January) the new face detecting (with deep
> > learning, or other soft computing) debut. At that point in the "people"
> > tab pane, was new options for face detecting: rebuild, learning
> > something related for the learning algorithm, but in the new versions it
> > is missing.
> > 
> > And I read days before about the improvements to the faceengine.
> > 
> > So what is under the hood? Its the "new" one with the deep-learning, or
> > the old one what is was in the 6.x?
> > 
> > (what is now working is new very useful: many face not detected, the
> > recognize is suggest wrong faces example with glasses for people without
> > glasses)
> > 
> > Thanx
> > 
> > Csaba






More information about the Digikam-users mailing list