<br><br><div class="gmail_quote">On Fri, Nov 5, 2010 at 7:55 PM, Marcel Wiesweg <span dir="ltr"><<a href="mailto:marcel.wiesweg@gmx.de">marcel.wiesweg@gmx.de</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im"><br>
> I think it really is the first tag I added.<br>
><br>
> And those faces confused are already quite well trained, i.e. > 50 pitures<br>
> confirmed for each.<br>
<br>
</div>I'll test face recognition. Maybe it simply still doesnt work.<br>
<div class="im"><br>
<br></div></blockquote><div><br></div><div>I recall that there was a bug in libface, the very first training has to be done with more than one tag IDs. That is, in the very first training of the database, all the faces in the training vector should not have the same ID. This is an issue with the very nature of the algorithm. </div>
<div>Two ways to fix this : </div><div><b><br></b></div><div><b>Patching libface</b></div><div>I might take a look into how to fix this soon, most likely it will be fixed by appending a dummy face with a special ID (one that will be never assigned by digikam normally).</div>
<div><br></div><div><b>Workaround in digiKam</b></div><div>I haven't looked in the new code about how things are done, but the faces could be given for training in bunches of 10 or 20 at a time. </div><div><br></div><div>
The second way is rather hackish however, and I'd look at patching libface as a solution to this.</div><div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im">><br>
> This gets me to another point. Would it be possible to not only have<br>
> unknown pictures but groups of unknown, i.e. digikam should be able to<br>
> guess which faces might look the same although it does not know their<br>
> name.<br>
<br>
</div>I have already thought of this. No idea if face recognition technology<br>
availalbe to us is ready for this.<br>
<div class="im"><br></div></blockquote><div><br></div><div>About grouping within the "Unknown" set:</div><div>This, although possible in theory, is *very* computationally expensive. It requires a clustering algorithm to condense the faces into groups, however, in the clustering, the computation of the "distance" between faces in the face space is computationally. More so because the dimension of the face space changes with the amount of faces, plus the number of unknown faces is supposed to change so frequently that a re-clustering every now and then would be required. AFAIK, none of the 'competition' do this.</div>
<div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="im">
><br>
> And also, what happens if a user scans the whole collection and ends up<br>
> with thousands of pictures in unknown? He would have to assign tags to<br>
> each of them and digikam would not try automatically to guess faces again<br>
> after it got some training. So the user would have to know about the<br>
> "recognise again" feature in the scanning GUI. I think there should be<br>
> some smartness in digikam that at least asks after x pictures training for<br>
> a tag whether it should try to find all pictures for that person.<br>
<br>
</div>Yes, interesting idea.<br></blockquote><div><br></div><div>I support this idea too.</div><div> </div></div><br>-- <br>Adi<br>