Given the insane progress in speech\voice recognition in audio in the last 2-3 years, I have been wondering how soon we are going to see the same accuracy levels for object rec in video. Lots and lots of new apps on the horizon...