Karl Malden says no
Lets do a quick test on Karl Malden using Photoshop's distance measure tool
Middle age Karl:
I make the gap 49 pixels between centre of pupils, and 34 from the intersection of the line joining the center pupils to the low point of the nose.
Gap between pupils 73 pixels, bottom of nose to intersect line 52
So to get that pasts we'd need say +- 2% tolerance, about 1.4 to 1.48
So I reckon the extremes are say 0.75 to 2, and there are 6 billion people,
as a rule of thumb even if they were all evenly spread it would be 344 million people with that same facial biometric.
You know, that's not very good to say the least. The problem with facial recognition is that it's comparing fleshy points that change over time. The bone even changes, but the flesh changes much more.
Karl nose is fleshy of course which is why I chose him to illustrate the problem, although it fails all over the place for the same reason.
Now lucky for the facial recognition people that they're saying yes or no, 50-50 chance of getting right.
And if they get it wrong and get a false positive?... Well who will know? Nobody checks the machine. And if they simply say yes nearly always then it's a pat on the back and people think it works. Sort of the astrology of the sciences.... "you will have a good day" and dumb people believe it.
But you do have to wonder about the competence of Jacqui. Only a fooking idiot is an early adopter that doesn't do even the quickest of tests.
Even if she stopped and used her brain (sic) on this for a fraction of a second:
Any terrorist can simply try this machine to see if it accept him as the passport holder in a friendly airport till he finds a passport that matches, then he knows he can get through in unfriendly ones. Even if terrorists weren't entering with all the proper paperwork (or live here, our terrorists lived in leeds), you just weakened the security.
Jacqui just resign already. You're not fit to be in the position you occupy.