Australians are being urged to send in sensor camera footage to train artificial intelligence to recognise Australian native animals, in order to track how species are recovering after the bushfires.
Subscribe now for unlimited access.
or signup to continue reading
WWF-Australia and Conservation International will install more than 600 cameras in areas hit by the Black Summer bushfires, including the Blue Mountains, East Gippsland, and Kangaroo Island.
Images captured will then be analysed by Wildlife Insights, which uses artificial intelligence and machine learning technology developed by Google to identify species.
The platform can assess 3.6 million photos an hour. By comparison, human researchers can only assess 300 to 1000 photos an hour.
The project dubbed "An Eye on Recovery", has been supported through a $1 million USD grant from Google's philanthropic arm.
WWF-Australia's Head of Healthy Land and Seascapes Darren Grover said the program put would "hundreds of pairs of eyes" in fire-afflicted landscapes.
"This will give us a better understanding of what animals have survived and where we should target our recovery actions," Mr Glover said.
The cameras have already captured a glimpse of the elusive Kangaroo Island Dunnart. The species was already listed as endangered before than 90 per cent of its habitat was lost in the fires.
But while the technology has been trained to recognise more than 700 species, it is still learning to distinguish Australian natives.
WWF-Australia called on Australians with sensor cameras to submit their footage to help.
"Like humans, AI models get better at recognising and identifying animals if they can look at hundreds or thousands of images," Mr Grover said.
"Whether you're a researcher with a suite of images sitting in a data storage system or a hobbyist with a sensor camera in your garden to monitor local wildlife, everyone can play a part in this project."
Citizen scientists have played a key role in the wake of last summer's bushfires.
The Citizen Science Bushfire Project Finder was set up after the bushfire science roundtable in January.
One of the key projects was the Kangaroo Island dunnart survey.
Around 550 people volunteered to sift through almost 25,000 images of native animals taken by cameras deployed across the western end of the Kangaroo Island to search for mouse-like dunnart.
READ MORE:
The CSIRO said in June the survey set a record for the largest mobilisation of citizen scientists.
Volunteers processed an average of 350 photos per day, freeing up researchers to carry out ground surveys and feral animal control.
"There is no substitute for the power of people, and science is at its most beneficial when it's in the hands of everyday Australians, using it to solve challenges that matter the most to them," CSIRO Chief Executive Dr Larry Marshall said at the time.
"This summer's awful bushfire season devastated many communities, but citizen science is one way that everyone can help in the recovery and contribute to Australia's future resilience."
