According to an article in Muckrock, the Hennepin County Sheriff’s Office (HCSO) doesn’t want the public to know about their ‘secret’ facial recognition cameras.
Police in Minneapolis are going to great lengths, to keep the public from discovering how they’re being spied on.
Last weekend, Tony Webster was notified that the HCSO was asking the Supreme Court to postpone the release of their facial recognition documents.
The Star Tribune revealed that the HCSO has been secretly using facial recognition cameras since 2013.
Is DHS and law enforcement using the 2018 Superbowl as an excuse to create a ‘secret’ facial recognition system?
According to another article written by Webster, Minnesota law enforcement are using Cognitec and Securonet to spy on everyone.
In a 2015 email, an HCSO employee wrote, “…we need to hook up Securonet with Cognitec and their VideoScan product.”
FaceVAC
Cognitec’s FaceVACS system specializes in identifying people in real-time.
“FaceVACS-VideoScan detects and identifies persons of interest in real-time while computing demographic and behavioral data…” FaceVACS can also detect frequent visitors.
Earlier this year, I warned everyone that facial recognition cameras do more than just identify people, they detect a person’s gender, age, ethnicity and much more.
Securonet let’s police use CCTV cameras to spy on everyone. Last year, I wrote a four-part series warning everyone that police are using private CCTV cameras to spy on the public.
Securonet
The above admission by Securonet should dispel any doubts about police using private CCTV cameras to spy on the public. (Law enforcement also uses GenICam software to spy on the public.)
Police deny FOI requests, for words used in the English Dictionary
For months, the HCSO has denied Webster’s FOI requests, claiming it was too burdensome and more.
But I’ve never, seen such a brazen and disgusting argument by law enforcement to obfuscate government spying.
The HCSO has decided to argue that because Webster asked for a keyword search, the request wasn’t valid because they don’t store emails by “words in the English dictionary.” “Which is absurd right?” says Webster.
Fortunately, a judge ruled that law enforcement’s ‘English dictionary’ argument was a play on words, meant to discourage the public.
DHS has succeeded in creating a ‘secret national police force’ who use the courts, facial recognition cameras, Stingray’s, license plate readers etc., to spy on everyone.
For more information about police using CCTV cameras to spy on the public read Cognitec’s Whitepaper, “The Advantages of Industrial Cameras for Face Recognition Applications.”
Cognitec reveals, how retail stores are using ‘industrial’ facial recognition cameras.
http://massprivatei.blogspot.com/2017/05/police-public-cant-use-words-in-english.html
*** The Star Tribune revealed that the HCSO has been secretly using facial recognition cameras since 2013. ***
The article at the link in the above quote claims that the software those pigs are using only works on still images, and they say it takes about 30 minutes to get a match. Also, the matches aren’t always correct.
They might be better off just posting the suspect’s photo on the news or the Web.
*** “FaceVACS-VideoScan detects and identifies persons of interest in real-time while computing demographic and behavioral data…” FaceVACS can also detect frequent visitors. ***
I read the technical details provided on their website in the following PDF:
http://www.cognitec.com/files/layout/downloads/FaceVACS-algorithm-performance-B10.pdf
Note how their algorithm works very well for frontal mugshots, but accuracy gets badly degraded under less-then-ideal conditions. Even for a database of only ~1000 people, they get a lot of misses and false positives. The accuracy of these systems goes WAY down as the number of people in the database goes up. The company is not going to emphasize facts like this its sales pitch to the pigs.
Also note that their system only looks for certain people in real-time. Computationally speaking, that’s much easier to do than IDing everyone it sees in real-time.
So, as I’ve said before: FR doesn’t seem to be there yet. It does work to a degree, but it’s nothing like what you see on shows like CSI. If the surveillance state were all-seeing and all-knowing, I’d expect murder clearance rates in the cities to be a whole lot better than they are.
Still, FR undoubtedly increases the threat to privacy and public anonymity. I wish it had never been invented; however, it’s not going to be uninvented. The good news is that there are ways to degrade its accuracy (sunglasses, hats), and ways to defeat it completely when needed (cover your head and face, especially around the eyes).
Some friendly advice, don’t rely on a corporate document to reveal whether they can identify people in real-time.
Numerous documents exist that state, they have the ability to identify people in near real-time. Facial recognition cameras also have the ability to flag individuals.
Don’t forget the police are using a database of drivers license and ID’s to match a person’s face to.
In the not too distant future we should expect every camera to be equipped with facial recognition.
If you value your privacy hide your face.
*** Some friendly advice, don’t rely on a corporate document to reveal whether they can identify people in real-time. ***
I agree, but if they’re going to be less than truthful in their literature, I’d expect them to **exaggerate** the capabilities of their software. There are multiple companies trying to sell this software to the pigs, and they’re in competition with each other, so it wouldn’t make sense for them to claim their software is less effective than it actually is.
*** Numerous documents exist that state, they have the ability to identify people in near real-time. Facial recognition cameras also have the ability to flag individuals. ***
Yes, but again, it depends on lighting, camera angle, and other variables, including the biggest variable of all: database size. As the number of people in the database goes up, accuracy goes down. So, a system that is 92% accurate when tested on 1000 people may only be 60% accurate when used in a city of a million people.
Ultimately it’s a matter of probability as to whether one of these cameras correctly identifies someone or not. The better systems simply increase the rate of true matches and have a lower rate of false matches.
For frontal mugshot-type photos, accuracy is typically very high. But most surveillance cameras are elevated and looking down at you, making them less effective for FR (especially if you keep your head down, wear a hat with a wide brim, or wear a hood).
*** In the not too distant future we should expect every camera to be equipped with facial recognition. ***
For that to work, every camera would have to be networked to a massive database, requiring massive bandwidth and computing power. You might be surprised at the technical challenges involved in a project like that. Doing it on a camera network in a limited region, like an airport or section of inner city, may be possible. But the database is probably limited.
These challenges would explain why they can’t seem to solve more murders in cities like Chicago, which has been covered by a Big Brother camera network for years.
*** If you value your privacy hide your face. ***
Absolutely, and that would be true even if there were no automated FR. Simply getting photographed is enough to prove you were somewhere — people can be easily identified after the fact via crowdsourcing.
As long as “catching a criminal” (regardless of guilt or innocence) brings in the moolah–it’s all about the money, right?–then they won’t care if this “facial recognition software” works or not!
Back to growing my beard down to my huge belt buckle ..that way i can wrap it around my face to look like 100X more harry than Grizzly Adams .. even over my eyes , than i put on my sunglasses
or start doing face paining every day as you go out , with a different camo job on yer face
or dammit , I’m ugly as sin , so I’ll just shave my ass and teach myself to walk backwards , there .. fixed it
“Fortunately, a judge ruled that law enforcement’s ‘English dictionary’ argument was a play on words, meant to discourage the public.”
More than they already are?