AI can recognise your face and your gait; how will it be used?
High-tech surveillance systems pose a unique dilemma for governments and their citizens. On one hand, they can be a very effective tool in combating crime and terrorism. On the other, they can be used to monitor ordinary people who are minding their own business. As technologies improve, the level of surveillance governments can exercise becomes staggering, and biometric artificial intelligence is driving it all.
Australia is developing a facial image recognition system called the “National Facial Biometric Matching Capability.” Using photos collected from various state and federal government sources — driver licences, passports, visa applications, etc. — the artificial intelligence can instantly compare faces caught on surveillance methods such as closed-circuit cameras.
New South Wales has allocated $52.6 million to the programme over four years, with NSW Minister for Counter-Terrorism David Elliott saying “the capability” would enable authorities “to quickly identify a person of interest to help keep the community safe.”
The face verification service, which matches a known individual’s photo with photos from government records, is already up and running. Still to come is the face identification service, which will use an image of an unknown face and scan through government records in an attempt to identify the person. That phase is expected to be operational before the end of 2018.
Of course, people who are up to no good frequently cover their faces. China has an answer for that. It’s AI that picks up on something people can’t mask: their gait. Already in use by Beijing and Shanghai police, “gait recognition” software can identify people by how they walk.
Huang Yongzhen, CEO of Watrix, said his company’s AI can identify individuals at a distance up to 50 metres, even if their backs are turned.
“You don’t need people’s cooperation for us to be able to recognize their identity,” Huang told the Associated Press. “Gait analysis can’t be fooled by simply limping, walking with splayed feet or hunching over, because we’re analyzing all the features of an entire body.”
Whereas facial recognition software requires just one image to identify someone, gait recognition needs a series of images, making it much more complex. Watrix’s software needs about 10 minutes to process an hour of video, using silhouettes to create a model of how an individual walks.
Watrix has raised nearly $20 million to develop and sell its gait recognition technology, and Huang left academia in 2016 to capitalise on what he sees as its commercial applications. The Watrix website touts intelligent guest welcoming and security for events and conferences, and Huang said the biometric AI can identify people in distress. At 94 per cent accuracy, the AI is good enough for commercial use, Huang said.
Both facial and gait recognition are mostly of interest to governments for their ability to identify criminals and terrorists.
“Identity crime causes substantial harm to the economy and individuals each year and is a key enabler of terrorism and serious and organised crime,” said Andrew Rice, assistant secretary in Australia’s home affairs department.
Of course, not everyone caught on camera in public is committing a crime, and opponents worry about governments using the biometric technology as an Orwellian tool of mass surveillance and an invasion of privacy.
“It’s hard to believe that it won’t lead to pressure, in the not too distant future, for this capability to be used in many contexts, and for many reasons,” Lesley Lynch, deputy president of the NSW Council for Civil Liberties, told the Guardian. “This brings with it a real threat to anonymity. But the more concerning dimension is the attendant chilling effect on freedoms of political discussion, the right to protest and the right to dissent. We think these potential implications should be of concern to us all.”
While NSW officials say “the capability” won’t be used for mass surveillance and comes with robust privacy safeguards to prevent private corporations from accessing the system, the Chinese government is making no such promises in regard to gait recognition. Officials in Xinjiang, where there is a substantial and scrutinized Muslim population, are interested in the technology.
As Watrix has demonstrated, there is money to be made in AI, and information is always valuable.
As Chinese columnist Shi Shusi told the AP, “using biometric recognition to maintain social stability and manage society is an unstoppable trend. It’s great business.”