Zarxrax wrote:Well, i guess that might work for people who are kind of artistically inclined, or have good memories of the amvs...
But me... I doubt I could go much farther than simple stick figures :p
Well, the idea goes something like:
How many things have you seen but can't remember any identifying details about, beyond a rough recollection of some visual content here and there?
I guess the search algorithm could be further relaxed to allow for permutations of key frames, although that could be murderous if you did a straightforward extension of imgSeek's matching algorithms.
Even if it turns out to not be useful, it's at least something new, and not the old, tired text-lookup systems that we've had to bend ourselves around for years. So it's still interesting in that sense. :P
