3 ms·
There are several future inflection points for facial recognition software. First, three dimensional modelling, initially just based on stills (e.g. from a sing
by InclinedPlane 15y ago
There are several future inflection points for facial recognition software. First, three dimensional modelling, initially just based on stills (e.g. from a single stereo image pair or laser scan or such-like) but later able to process video and extract a facial model having some canonical expression, making matching easier. That alone is a bit scary once you combine it with the future utter ubiquity of massive processing power, imaging sensors, and communications. Second, a far more sophisticated facial/body recognition technology based on fundamental kinematics. You won't be able to fool such systems by wearing a disguise alone.
- kaybe 15y agoThe question is, what will be enough? More than you'd need to fool an average person?
- InclinedPlane 15y agoI don't know, but we can infer based on the capabilities of humans. It's at least possible to recognize someone from the way they walk and other characteristics, even after they've aged, put on or lost weight, etc. as people do that sort of thing all the time. Beyond that I don't know that the problem has been studied well enough to know the limits.