This time the Israeli’s are watching you at the mall, not just sporting events like the Superbowl.
Some of you locals may remember the last time the Superbowl was in Tampa. There was a widely publicized effort to digitally capture every single face that came into the Superbowl.
In the guise of protecting the nation from terrorists Tampa authorities went beyond the law and beyond the constitution. They hired a foreign company to set up camera’s use facial recognition software and run background checks on every single man (or persons?) that entered the super bowl. No approval from the voters and in fact, the program was kept secret until days before the SuperBowl. After it’s release it was touted as the end all to security. Maybe the common man felt safer, but those of us in the know, were worried. Really worried. Well we’re worried again, because they’re at it again.
In the end the terrorists didn’t try to mess with the SuperBowl, lucky us. The software failed miserably. In fact it failed so badly that they wouldn’t release the number of false positives. It’s impossible to say how many people on the terror watch lists it missed. While they didn’t find any terrorists, they did say they found a couple probation violators (People behind in their bills – how dangerous – protect us from them!)
I did however, dig up some interesting research about facial recognition technology, as it used to be called. They’re now trying to rebrand the same technology as “Audience measurement” technology. Either way you say it, it doesn’t work properly, is run in secret by foreign companies and just shouldn’t be used, maybe ever.
Here’s the best case error rate: 1 in 250. That’s the best rate under controlled circumstances. Best circumstances would be something akin having a database of three faces and showing it one of the ones that was in the database, in exactly the pose, distance from camera, facial expression, picture quality etc. It was still wrong 1 out of every 250 tries. Real world testing results put the error rate, at minimum, as 1 of every three people triggers a false positive and nearly a 100% rate of false negatives (“terrorists” that should have been recognized, but were not), have never been released but some results is the effectiveness is around 0.
And there are companies out there that want to use this stuff to secure airports!
Even if it worked exactly perfectly it’s still quite ripe for abuse. In fact, it would be quite hard to use it without sharing the results of people travels, personal info and habits etc to private companies and governments not our own. (Link)
I’ve also learned that even though there is nothing positive (that’s true anyway) to say about this technology, our formally republican government was in love with it. Here’s an example:Link. Please note that most if not all of the “facts” in the article are mostly just BS (sorry no other way to put it, they jus lyin!)
Then there are the legal issues. the couts have ruled numerous times that this is a serch and you need a warrant or probable cause to search someone this way. No, walking, driving going to football games, are NOT probable cause.
There’s so much research out there, there’s so much info pointing these different companies making this crap and trying to rip people off, violate their privacy (one CEO even tried to say it didn’t store the faces (that may be, but I bet it keeps a list of people it thinks are matches and that -I read thier terms – they give away freely to government and other businesses).
I can’t write anymore. This is truly ridiculous.