UK Police Plan To Deploy 'Staggeringly Inaccurate' Facial Recognition in London (independent.co.uk)
An anonymous reader quotes the Independent:
Millions of people face the prospect of being scanned by police facial recognition technology that has sparked human rights concerns. The controversial software, which officers use to identify suspects, has been found to be "staggeringly inaccurate", while campaigners have branded its use a violation of privacy. But Britain's largest police force is set to expand a trial across six locations in London over the coming months.
Police leaders claimed officers make the decision to act on potential matches with police records and images that do not spark an alert are immediately deleted. But last month The Independent revealed the Metropolitan Police's software was returning "false positives" -- images of people who were not on a police database -- in 98 percent of alerts... Detective Superintendent Bernie Galopin, the lead on facial recognition for London's Metropolitan Police, said the operation was targeting wanted suspects to help reduce violent crime and make the area safer. "It allows us to deal with persons that are wanted by police where traditional methods may have failed," he told The Independent, after statistics showed police were failing to solve 63 per cent of knife crimes committed against under-25s....
Det Supt Galopin said the Met was assessing how effective facial recognition was at tackling different challenges in British policing, which is currently being stretched by budget cuts, falling officer numbers, rising demand and the terror threat.
A policy officer from the National Council for Civil Liberties called the technology "lawless," adding "the use of this technology in a public place is not compatible with privacy, and has a chilling effect on society."
But a Home Office minister said the technology was vital for protecting people from terrorism, though "we must ensure that privacy is respected. This strategy makes clear that we will grasp the opportunities that technology brings while remaining committed to strengthening safeguards."
Police leaders claimed officers make the decision to act on potential matches with police records and images that do not spark an alert are immediately deleted. But last month The Independent revealed the Metropolitan Police's software was returning "false positives" -- images of people who were not on a police database -- in 98 percent of alerts... Detective Superintendent Bernie Galopin, the lead on facial recognition for London's Metropolitan Police, said the operation was targeting wanted suspects to help reduce violent crime and make the area safer. "It allows us to deal with persons that are wanted by police where traditional methods may have failed," he told The Independent, after statistics showed police were failing to solve 63 per cent of knife crimes committed against under-25s....
Det Supt Galopin said the Met was assessing how effective facial recognition was at tackling different challenges in British policing, which is currently being stretched by budget cuts, falling officer numbers, rising demand and the terror threat.
A policy officer from the National Council for Civil Liberties called the technology "lawless," adding "the use of this technology in a public place is not compatible with privacy, and has a chilling effect on society."
But a Home Office minister said the technology was vital for protecting people from terrorism, though "we must ensure that privacy is respected. This strategy makes clear that we will grasp the opportunities that technology brings while remaining committed to strengthening safeguards."
Now they can go after more efficiently anyone who says something unapproved on Twitter!
(e.g. someone reporting on pedophile gangs)
yay technology. UK is screwed.
Even antivirus testers deduce marks for false positives.
For example in the latest AV Comparatives test Symantec got dropped a grade for having 90 false positives out of 20,000.
That's 0.45%
Whatever happened Blackstone's "It is better that ten guilty persons escape than that one innocent suffer"
I think, therefore you are.
If enough people start wearing anti-surveillance clothing:
https://www.theguardian.com/te...
it just might reduce the success rate below "justifiable cost"
They sentenced me to twenty years of boredom
CCTV and images of people goes back to the days of the ring of steel https://en.wikipedia.org/wiki/... :)
All driver and passenger face in and out of a city in real time. Every license plate. Thats was with early 1990's tech.
Talking collected on every generation of "encrypted" cell phone. Voice prints, wifi. It all got collected on. Faces linked to all passports, all photo ID in use in real time.
Smile for every flight into/out of the UK, all the ferry routes. Truck stops. Ports. Rail. Webcam use. All internet use. That decade of international VoIP calls
Voice print to face, face to voice print.
The part that is so difficult is the need for parallel construction.
Domestic spying is now "Benign Information Gathering"
... one again.
How come that we in the West, who for half a century fought dictatorships with population surveillance and control, now willingly and without resistance walk into an unprecedented surveillance society?
The reasons cited ("budget cuts, falling officer numbers, rising demand and the terror threat") are not exactly new. Why is public face recognition, fingerprint recording and opening and reading mail acceptable nowadays?
Was the Ted Kazinsky correct in his prediction of the control society, that our modern society requires ultimate control of its citizens to function in an (post-)industrial setting and that our freedoms therefore must be taken away from us?
Several members of the astrology squad died in mysterious circumstances and the rest quit.
You know how it is in big bureaucracies - they had to use the budget for something.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
RIP GB
The guy who found so long ago all died, or are all but dead. The average MP and police chief in hierarchy is now 50, means they are the older children or grand children of those who fought, and were born in a time of boon. Those who have security and comfort are trying to keep it using method their parent would have blushed or cursed to. Frankly, this is the same type of people which deride millenial when at the same time 50 years ago they had comparatively cheap housing and a great potential. They are the same one predominentely wanting an empire they knew only in their stories. They are also the one wanting demographically brexit. that should tell you all. They are looking inward and going to the fetal position.
C. Sagan : A demon haunted world:
http://www.amazon.com/gp/product/0345409469/
visit randi.org
I have not worked on face recognition, but I have worked on text recognition, which has some of the same problems. You have several recognition algorithms that work. You can make some super-algorithm by polling all of the better ones. Oddly, the performance of the super-algorithm is measurably worse if you take away the worst one out of ten. It turns out that diversity is an important part of recognition. Sometimes the errors are trivial - I heard from someone who's lab had a face location system that had been trained on internet pictures of faces and cats, and could identify everyone in the lab except for one guy with a beard who was always a cat.
Let us compare this with our baseline, which is what we do. We recognise people but we come with a lot of faults and biases. We are able to see small differences in people like us, and are less able to distinguish people who are different. We can have our memories altered after the event. We cannot even run proper tests the way we can with computer recognition systems. People usually know when they are being tested. So, the baseline is pretty poor, too.
We can fix this. I see no reason why computer recognition systems should not be superior to humans. But this software is going to need training and development, and this process should be open if we are to be sure it does not have any built-in biases. Software does not spring fully-formed from the brow of Alan Turing, and to expect it to work from day one is unrealistic. To rely on a boxed solution from some private contractor is also unrealistic, and also dangerous. If we automatically identify any development in recognition systems as Orwellian evil, then I am afraid that's what we are going to get.
This technology might be justifiable if the UK were serious about fighting terrorism, but they're not really serious. At best this is just security theater. At worst it's part of an ever-growing police state. Terrorism is just an excuse. They'll ignore it until it's impossible to stop the totalitarianism.
Seems like it could become a thing for there to be contests to trigger a false positive.
Some drink at the fountain of knowledge. Others just gargle.
They can't shake off their authoritarian,Big Brother tendencies, but they do attempt to implement them poorly.
Just like dope sniffing dogs and broken tail lights.
Have gnu, will travel.
Especially violent crimes like murder and armed robbery have been going down since the mid 90s.
Just checked that assertion. It is not true. What actually happened is it skyrocketed after 1995, peaked, then began to fall. Reaching levels similar to 1995 levels around the 2010 time frame.
You could easily find that out by using Google and taking a look at the statistics
Easy for you to say, but per my brief, which denotes easy, search a vast majority of the crime statistics data seemingly does not exist after 2010ish time frame.
Overall I'm not implying the poster you are replying to is correct and you are incorrect, I just don't see any solid evidence your assertions are correct.
Caution: Contents under pressure
That's just GOP propaganda. The actual figures reveal that crime has been going down world over steadily for a few decades now. There have been local upticks in time, like (possibly) a current one, but the long-term trend is indisputable. But, do not believe me, and do not look up yourself the figures; just keep eating the Fox News garbage.
That particular A/C is full of shit, and probably has never been to London.
That whole rant reads like it came from some Mid-Western US basement somewhere, but he is correct about the CCTV quality.
He is drawing the wrong conclusion of course. If he had ever been to London he would know that no-one worries about CCTV because they all know how shit it is.
The system isn't judge jury and executioner. The system is just trying to reduce the search space for good old fashioned policing. False positives don't mean that it will only identify 2% of cases, it means from the cases its identified 2% are correct.
If you're looking at a million people, and you're looking for 1 person, having a search space of 49 to go through is an incredible win for policing. Now tracking implications on the other hand are quite severe, but then that is actually helped for privacy advocates by a system which may confuse you with 49 other people.
... an officer looks at the match and dismisses it.
Where exactly is the harm in that?
Sure one might discuss if CCTVs everywhere and facial recognition are desirable, but the fraction of "false positives" is no valid argument in that discussion, since all positives are checked a second time by a human. The automated system serves as a first filter, meaning less comparisons by humans are necessary so the whole system becomes more efficient. A more interesting number are false negatives, but about those we know nothing.
To all those who now are opposed to the software because it generates so many false positives: Ask yourself if you'd really prefer an improved software that has a much higher accuracy. If not, then the high number of false positives is not really an argument in favor of what you want.
"By the way if anyone here is in advertising or marketing... kill yourself." -- Bill Hicks
Have you noticed how easy it is to bias the reader just by the order you say things in? Take this quote from the summary.
A policy officer from the National Council for Civil Liberties called the technology "lawless," adding "the use of this technology in a public place is not compatible with privacy, and has a chilling effect on society."
But a Home Office minister said the technology was vital for protecting people from terrorism, though "we must ensure that privacy is respected. This strategy makes clear that we will grasp the opportunities that technology brings while remaining committed to strengthening safeguards."
Now let's reverse the two paragraphs and move the word "but", which tells us the second quote rebuts the first one.
A Home Office minister said the technology was vital for protecting people from terrorism, though "we must ensure that privacy is respected. This strategy makes clear that we will grasp the opportunities that technology brings while remaining committed to strengthening safeguards.
But a policy officer from the National Council for Civil Liberties called the technology "lawless," adding "the use of this technology in a public place is not compatible with privacy, and has a chilling effect on society."
Funny how that leaves you thinking something totally different. Now let's try to write it in a more evenhanded way that doesn't tell the reader who is right.
Commenters disagreed on the legality and privacy implications of the technology. A policy officer from the National Council for Civil Liberties called the technology "lawless," adding "the use of this technology in a public place is not compatible with privacy, and has a chilling effect on society."
On the other hand, a Home Office minister said the technology was vital for protecting people from terrorism, though "we must ensure that privacy is respected. This strategy makes clear that we will grasp the opportunities that technology brings while remaining committed to strengthening safeguards."
"I'm too busy to research this and form an educated opinion, but I do have time to tell everyone my uninformed opinion."
Of course, a terrorist would never think of wearing a disguise. No, never.