https://bugs.kde.org/show_bug.cgi?id=415557
--- Comment #5 from Alexandre Belz <alexandre.b...@live.fr> --- Thank you Maik for this explanation. There is something i don't understand, then : as the User Manual explains, "In case of unsatisfying results it might be helpful to use Clear and rebuild all training data. One reason can be that there are too many face tags assigned to a person which shows this person in a way that doesn't really help the search algorithm, e.g. with sunglasses, blurred, unusual colors, carnival make up, dark shaded areas in the face, baby/kid/adult photographs mixed..." ... i then expect the possibility to reject some folders (odd faces) and rebuild the DNN database on recent pictures only (the selected folders). Which is not possible currently. If we rebuild the learning DNN database on the whole collection, we'll still re-include old/misleading pictures... Don't you think so ? -- You are receiving this mail because: You are watching all bug changes.
