News:

That's okay, I know how to turn my washing machine into a centrifuge if need be.

Main Menu

Melting two faces with MorphMan

Started by bugmenоt, January 20, 2010, 04:52:49 PM

Previous topic - Next topic

bugmenоt

I AM NOT JUST GONNA KICK YOUR ASS, AFTER THAT I'M GONNA OUTSMART YOU USING MY NERD GLASSES
                                                      |


bugmenоt


Sir Squid Diddimus




Rococo Modem Basilisk

It would be nice if this could be automated... Anyone have facial recognition code? I'm willing to dedicate a weekend on my server to melding together all possible combinations of two pictures from the Burns' spagbook rip.


I am not "full of hate" as if I were some passive container. I am a generator of hate, and my rage is a renewable resource, like sunshine.

bugmenоt

I don't think that MorphMan uses face recognition. Every spot is set manually by clicking for example on both left eyes. Then, I think, both pictures are stretched to the average point and one picture is made 50% transparent and layered above the other one.

For homemade face recognition look at this: http://pages.cpsc.ucalgary.ca/~hanlen/vision/facelinks.html

Rococo Modem Basilisk

I mean, with facial recognition no one would need to click anything, which means that it could go through the whole archive automagically without human intervention -- which, for large data sets, is generally the best idea anyway.


I am not "full of hate" as if I were some passive container. I am a generator of hate, and my rage is a renewable resource, like sunshine.