Tuesday, November 29, 2011

(Facial Recognition vs Human) & (Facial Recognition + Human)

Terrorists picked out in a crowd by cameras that can beat human eye (This is London)
Mr Bada was given five minutes to memorise 30 faces before six of the subjects mingled with crowds at Charing Cross railway station and walked past a fixed point.

The Met expert was able to pick out two while the Face Alert camera spotted four. Tim Noest, the managing director of Lodge Service Intelligence which makes Face Alert, said two of the faces were obscured by crowds.
Humans are awesome at ID management among people they know. The processes people use to identify people with high confidence are extremely complex and may take into account gender, age, gait, time, posture, scent, sound, weight, location and countless other details processed simultaneously and without necessarily involving a lot of conscious effort. People, however, aren't very good at identity management among large numbers of people they don't know.

In biometrics, software takes in a mere fraction of the information people use and it doesn't make any inference about it. It treats the problem in a way that closely resembles Nikola Tesla's famous critique of Thomas Edison.
“If Edison had a needle to find in a haystack, he would proceed at once with the diligence of the bee to examine straw after straw until he found the object of his search.”
When dealing with people we don't know, humans are relegated to the needle-in-the-haystack process. Even if you believe that computers aren't very good at recognizing people this way, they're better at it than people are.

When biometric software is used to sort a large list by the probability of a match, then present the list to a human such as Mr. Bada, the results can be impressive indeed.


h/t @Allevate