[digiKam-users] Face recognition - to YOLO or not

Gilles Caulier caulier.gilles at gmail.com
Mon Mar 21 12:13:14 GMT 2022


Hi,

No Yolo DNN face data model is better, but from a performance viewpoint, it:

- use more memory as the model is bigger.
- take more time to process as it detects more faces especially in
difficult shot conditions.

So, it depends of use cases.

Best

Gilles Caulier

Le lun. 21 mars 2022 à 13:02, daj omu <dajomu1 at gmail.com> a écrit :

> Hi,
>
> There is an option to scan faces using YOLO, but should I use that or not?
> I see around the web that this good, but is YOLO the better option? I
> assume that the default options are the ones that will give me the best
> result, correct?
>
>
> Dajomu
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/digikam-users/attachments/20220321/a37ab4d7/attachment.htm>


More information about the Digikam-users mailing list