Google machine learning to save sea cows
Google machine learning to save sea cows Getty Images/Guillaume Souvant

Can Google's machine learning technology help sea mammals such as sea cows that are under the threat of extinction? Scientists believe it can.

Amanda Hodgson of Murdoch University and her team, along with computer scientist at Queensland University of Technology Frederic Maire, have created a detector using Google's TensorFlow platform that can automatically find sea cows in aerial photos of oceans taken from drones.

TensorFlow, Google's second generation machine-learning system was launched last year. It is developed by the researchers at Google Brain Team within Google's Machine Intelligence research organisation. TensorFlow is twice as fast as its predecessor DistBelief.

This technology is very much like the image recognition tool that lets you search Google Photos for particular things such as dog species etc.

Josh Gordon, developer advocate at Google noted that scientists had to spend a long time peering out of planes for sea cow census, which was an expensive and hazardous process. Meanwhile, Hodgson came up with an idea to use drones to take aerial photography of oceans.

Hodgson had clicked as many as 45,000 ocean photos using drones and had to undertake the task of manually detecting sea cows in those photos. But Google's TensorFlow can perform the task of finding the endangered species from the photos.

Scientists said that an early version of the detector helped them find 80% of the sea cows that they had found manually in photos. Now, they expect to improve the performance of the detector and use the technology to detect other sea mammals such as humpback whales and certain dolphins.

"Eventually if they're able to track these threatened populations on a large scale, conservationists have a much better shot at knowing how they're impacted by human activities, and where it's most urgent we protect their habitats. In a small way, machine learning might help save the humble sea cow," Gordon noted.