Speaking of neural nets and image recognition, my former employers landed a contract with "various law enforcement agencies" although the project manager I spoke to said primarily the FBI and CIA - to develop neural network hardware and software to do face recognition, to be used to identify criminals at the airports. Indeed, I recall being photographed twice when I hired on - normal lighting for my regular id, and low lighting. At the time, I wondered why but didn't think anything of it. The project manager said they were amassing a database of pictures to use as training sets. Pretty interesting...
Speaking of neural nets and image recognition, my former employers landed a contract with "various law enforcement agencies" although the project manager I spoke to said primarily the FBI and CIA - to develop neural network hardware and software to do face recognition, to be used to identify criminals at the airports.
Indeed, I recall being photographed twice when I hired on - normal lighting for my regular id, and low lighting. At the time, I wondered why but didn't think anything of it. The project manager said they were amassing a database of pictures to use as training sets.
I think you mean "Spooking of neural nets and image recognition..." This technology is supposedly being deployed at Customs offices and especially at airports. Long hair may come back, as I hear that the _ear_ is one of the main identifying points...ears are fairly characteristic of individuals (not as much as fingerprints, but enough for a rough sort) and edge-tracing algorithms can be run quickly. So, keep those ears covered! Of course, only criminals fear such security monitoring. Remember, the policement is your _friend_. -Tim May --
Are the pictures taken by ATMs good enough for automated recognition? Nick Szabo szabo@netcom.com
On Thu, 12 Aug 93 18:02:06 PDT, szabo@netcom.com (Nick Szabo) said:
Are the pictures taken by ATMs good enough for automated recognition? Nick Szabo szabo@netcom.com
Do atm's usually have real cameras? I always figured most of them were fakes. Mike
participants (4)
-
Mike Rose
-
remail@tamsun.tamu.edu
-
szabo@netcom.com
-
tcmay@netcom.com