Australian facial recognition system could place well in NIST tests, official says
Australia’s Department of Home Affairs is using multiple algorithms to train its facial recognition systems, and believes those algorithms are among the top three most effective by NIST standards, ZDNet reports. One of the benefits of the approach is a reduction in discrepancies between the system’s accuracy rates for people of different ethnicities.
“We tune our algorithms all the time against a wide-ranging algorithm, we are very fortunate in Australia that we are a multicultural society, we get people coming in and out of the country from various nationalities,” ZDNet reports Home Affairs Acting Deputy Secretary of Intelligence and Capability Joe Franzi told Senate Estimates.
He said that based on working with NIST, the algorithms used by Home Affairs are in the top three.
Franzi also said that the high error rate of the system used by South Wales police to scan the crowd at a UEFA Champions League match are a result of the incomplete development of the technology for mass-surveillance applications.
“They perform really well when you go to the conference … and it works really well in a controlled environment, but once you actually move into a real-world environment … then you get all the issues around different lighting, angles of faces, where are cameras, do people have hats on, glasses on, [hoodies] — a whole range of things.”
Home Affairs has also revealed that it has spent AU$5.5 million (US$4.15 million) of its AU$10 million ($7.5 million) budget for integrating the IT systems being consolidated under the country’s division of responsibilities.
The extent of the participation in the facial recognition program by Australia’s territories is uncertain after the Australian Capital Territory and Victoria objected to its increasing scope, saying it may clash with territorial laws.