Scientists from the Department of Defence Science and Technology (DST) have developed a new facial recognition algorithm, significantly improving the AI’s detection capabilities in adverse environments, the Defence Department has revealed.
Research conducted by the DST’s biometrics team sought a new method to enhance facial recognition algorithms for use in “adverse environments”, such as poorly lit alleyways and distances over 250 metres.
The algorithm can be also be tweaked interactively to adjust for changing environmental conditions.
“As the user moves the sliders the output will be updated in real-time, allowing a tailoring of the algorithm to get the best results for a particular environment,” the DST revealed in a statement.
Contributing researcher Dmitri Kamenetsky said the new detection algorithm could offer a capability boost for ‘stand-off’ – otherwise known as ‘non-contact’ – surveillance systems, typically deployed around airports and high-security checkpoints.
The DST research team presented their results at last year’s Digital Image Computing: Techniques and Applications (DICTA) conference, demonstrating the improvements in recognition and face matching delivered by the enhanced algorithm.
“We’ve released a description of the algorithm, allowing other researchers to implement it and make further improvements. Interestingly, most of the research presented at DICTA was using deep learning in some way, ours is just a relatively simple yet effective mathematical approach,” Kamenetsky said.
Meanwhile, NSW’s Minister for Transport Andrew Constance flagged the possibility of deploying facial detection technologies on the state’s public transport network to support a “frictionless transport payments” system – one which may complement, or potentially replace, the existing card-based payments system.
Speaking at the Sydney Institute last Tuesday, Constance said the possibility of Sydney commuters paying for their train or bus fares with just their face could become reality “in the not too distant future”.
“No more gate barriers,” he said. “Just a smooth journey.”
The envisaged ‘barrier-free’ fares system, which the Transport Minister said would be operated on an ‘opt-in’ basis, would scan commuters’ faces and automatically charge their fare to a linked Opal account.
Constance said the biometric payments system would be comparable to Amazon Go’s ‘Just Walk Out’ capability deployed in its new physical stores, which allow customers to pick up goods as they shop and simply walk out of the store, being tracked and charged as they leave.
However, it remains to be seen whether a face-tracking fares system would be readily accepted by NSW commuters, a vast majority of whom remain content with the current card-based arrangement.
Upwards of 90 per cent of surveyed commuters said were satisfied with the ease of using an Opal card on the train network, according to the most recent customer satisfaction survey by Transport for NSW.
Curiously, legislators for the City of San Francisco, long considered the heart of the tech revolution, last May enacted a ban on the use of facial recognition software by government authorities, including police, public transport service and other government agencies, with concerns the technology remains unreliable, is liable to abuse, and may infringe on people’s privacy and liberty.
The ban makes San Francisco the first major US city to block facial recognition technologies, used increasingly by authorities to identify and track alleged criminals.
Airport ‘smartgate’ upgrade on hiatus
iTnews has revealed the Federal Government’s anticipated rollout of new facial biometrics-ready ‘smartgates’ across Australia’s international airports has stalled after the Department of Home Affairs (DHS) opted to pause its only operational trial of the system while it undertakes a review.
The next-generation smartgates, which will eventually allow travellers to pass through without producing a passport, are key to the department’s plan to process 90 per cent of international travellers through automated means by the end of 2020, iTnews reports.
The new smartgates will replace the existing paper-based incoming passenger card and exit marshalling process for passengers entering Australia.
The only fully operational trial of the facial recognition technology-backed smartgates – dubbed ‘face on the move’ – which had been conducted in Canberra airport until this latest pause, correctly matched 94 per cent of the 2200 travellers that took part.