QUOTES FROM FACIAL REC EXPERTS
"We've used a face recognition system for access to our laboratory for three or four years, and there are still times it doesn't recognize me as me," says Jim Wayman, director of the National Biometric Test Center at San Jose State University. "I was once locked out of the lab simply because a light bulb had burned out -- and I had left the spare bulbs inside the lab."
"A test last year by the Defense Department's counterdrug technology office found that even under carefully controlled conditions, the best commercial systems had a false acceptance rate -- the failure to detect someone they were supposed to catch -- of 10 percent. Even worse, they had a false detection rate -- incorrectly identifying an innocent person as someone on the wanted list -- of more than 30 percent."
"This technology is too immature for these types of applications," said James Wayman, a research director at San Jose State University. "There's no evidence from the government tests that a system like the one in Ybor City would work as anything other than a deterrent," Wayman said.
"It's very, very difficult to pick a face out of a crowd that you can compare to a database," added John Woodward Jr., a senior analyst with the RAND policy research institute.
``You could expect a surveillance system using biometrics to capture a very very small percentage of known criminals in a given database,'' said Samir Nanavati, a partner at New York-based consulting firm International Biometric Group.
"the enormous percentage of false matches will condition security workers to assume that all positive matches are mistaken. The great cost of implementing and maintaining the face recognition systems will have gone to waste. The fact is, spotting terrorists in a crowd is a needle-in-a-haystack problem, and automatic face recognition is not a needle-in-a-haystack-quality technology."
WSJ Article
But academics and other experts in the facial-recognition field have a sharply less-optimistic assessment. In a typical airport situation, "I don't think the systems will do the job reliably," said Takeo Kanade, a professor at Carnegie-Mellon University who has been studying computer facial recognition since the 1970s. |
