PS: Google for H-Eigenfaces (aka Hybrid-Eigenfaces). It is a nifty tweak to the original method. A professor in my university just showed me a paper that proposes this method, and it greatly solves the problem of pose variation.<br>
<br><div class="gmail_quote">On Sun, Apr 18, 2010 at 10:20 AM, Aditya Bhatt <span dir="ltr"><<a href="mailto:adityabhatt1991@gmail.com">adityabhatt1991@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Hi Kunal,<br> <br><div class="gmail_quote"><div class="im"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="gmail_quote"><div> Actually, there is a small interpretation error, the conditions regarding publishable usage is only for the adopters (ie. NVIDIA or ATI etc <a href="http://www.khronos.org/members/conformant/" target="_blank">http://www.khronos.org/members/conformant/</a> ) who come up with OpenCL compliant drivers / API implementation. As mentioned in the link "<a href="http://www.khronos.org/adopters" target="_blank">main adopters page</a>" we as implementers should not have any problems. (Quoting from the site:<a href="http://www.khronos.org/adopters" target="_blank">http://www.khronos.org/adopters</a>)<br>
<br>"<i><b> Implementers</b> - for no cost or license fees
you may:</i>
<ul><li><i>Create and deliver a product using the publicly released
specifications and technologies</i></li><li><i>But you may *not* claim that it is "compliant" unless they enter
and pass conformance testing</i></li><li><i>And if the product is a software or hardware engine, you may not
advertise it using the Khronos API technology logos or trademarks</i> "<br></li></ul>Regarding the second and the third point , that's ok from an OpenSource project perspective . As it provides us, both the freedoms<br>
of free speech & free beer , but asks us not to publicize it :) (without conformance testing).</div></div></blockquote><div><br></div></div><div>Ok then. But what I'm not sure about is : if we incorporate this in libface and consequently digiKam, doesn't that amount to publicizing it? If not, then that's great with me. But please ask the others about their opinions too - we'd want as less dependencies as possible.</div>
<div class="im">
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="gmail_quote"><div>Actually, freely available implementations of EBGM are available (as i had pointed out , in a reply to Marcel's mail sometime back the implementations can be found at <a href="http://malic.sourceforge.net/" target="_blank">http://malic.sourceforge.net/</a>
and also at <a href="http://www.cs.colostate.edu/evalfacerec/algorithms5.html" target="_blank">http://www.cs.colostate.edu/evalfacerec/algorithms5.html</a> ). Also i would like to interact with you'll on IRC,on which IRC do you'll (Alex and you) usually meetup.<br>
<br></div></div></blockquote><div><br></div></div><div>Yes, the CSU project was also my first choice for EBGM.</div><div class="im"><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="gmail_quote"><div>Now let us look at <i>how many</i> training images we would need (of a single person) for satisfactory recognition results.<br>
<br>Assuming a person looking straight at the camera, a 1 degree variation in pose from 0 degree ( face towards left) to 180 degree (face towards right) would result in 180 images. <br><br>Now if the person looks upwards 1 degree (again we would have 180 images from left to right) and if we keep varying poses we would get approximately 180 x 180 images of the same person.<br>
<br>Also for each successive pose varied image added to the training set the training time would increase exponentially.<br><br></div></div></blockquote><div><br></div></div><div>Actually, no - we can assume about 10-20 degrees tolerance. What I think is, ~4 to 5 representative images per person should be okay for a photo management suite. What is required for a person is : A frontal face, a profile face, and a sideways face. And then more faces can be added on the fly as more variations are encountered. This way, we get a denser sampling over pose as time proceeds.</div>
<div>And as I already said, the average number of friends/acquaintances that people would tag in photos is not so large as to noticeably slow down the re-training.</div><div>And compared to eigenfaces, the PCA + LDA approach is, to some extent, able to accomodate pose variations (20-30 degrees) and therefore fisherfaces should be fine for digiKam if the above method is followed.</div>
<div class="im">
<div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="gmail_quote"><div>I would love to add the implementation to libface. But looking at the similarity between Eigenfaces and Fisherfaces IMHO the<br>
effort should be to get as many different algorithms implemented as possible.<br><br></div></div></blockquote><div><br></div></div><div>Very true. But for now, the priority is to get fisherfaces up-and-running. It's sort of like an insurance of sorts - fisherfaces should be quite easy to implement and although less accurate than EBGM, it is okay enough to be incorporated into digiKam. So we'll definitely have at least one working algorithm. Meanwhile, we might want EBGM at some point, so jump in :)</div>
<div><br></div><div>And I and Alex (and the rest of the digiKam team, for that matter) doesn't communicate much via IRC, we use the ML more.</div><div>Get yourself subscribed to the libface ML. As for me, I usually idle in #digikam, #kde, and #kde-in. I think I've talked to you before on #kde-in :)</div>
<div><br></div><div>Cheers,</div></div><div><div></div><div class="h5">-- <br>Aditya Bhatt<br>Blog : <a href="http://adityabhatt.wordpress.com" target="_blank">http://adityabhatt.wordpress.com</a><br>Face Recognition Library : <a href="http://libface.sourceforge.net" target="_blank">http://libface.sourceforge.net</a><br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br>Aditya Bhatt<br>Blog : <a href="http://adityabhatt.wordpress.com">http://adityabhatt.wordpress.com</a><br>Face Recognition Library : <a href="http://libface.sourceforge.net">http://libface.sourceforge.net</a><br>