Artificial Intelligence and the Dugong
The goal of Laura Mannocci’s work in New Caledonia is to better protect marine megafauna using social media and artificial intelligence. It’s a project that involves taking to the skies… in a microlight.
Dugongs, rays, sharks, turtles… these large representatives of what is known as marine megafauna are particularly vulnerable. “These are animals that reproduce late in life and have few offspring, making them fragile species that require special protection,” explains Laura Mannocci, a marine biologist at the Marbec laboratory. Fewer than 1,000 dugongs, these large marine herbivores, are believed to still inhabit the waters of New Caledonia.
To better protect these majestic animals, we must first improve our ability to track them. This is the task that researchers from Marbec, Entropie, and Lirmm have set out to tackle, deploying an innovative method that utilizes artificial intelligence— “tools that ecologists are increasingly adopting,” notes Laura Mannocci.
Aerial survey
Whereas surveys are traditionally conducted below sea level, Laura Mannocci chose to take to the skies. More precisely, 50 to 60 meters above sea level—the altitude at which her microlight flew over a section of New Caledonia’s western lagoon.“These waters are very shallow—two meters at most—so you can see marine animals even from this altitude,”explains Laura Mannocci. To make sure she didn’t miss a thing, the researcher mounted a camera beneath the ultralight aircraft to capture the ballet of the rays and the passage of the turtles.
Filming is one thing, but identifying what’s in the footage is quite another… And that’s where algorithms come into play.“We used what’s known as deep learning —artificial intelligence applied to pattern recognition,” explains the biologist. The researchers thus developed an“intelligent”program designed to analyze the footage captured during the flight over the lagoon and identify, for example, a school of rays or a dugong hidden next to a coral head.
Social media
To identify these animals, the“self-learning”recognition algorithm must first have seen many of them.“We needed images taken from multiple angles and in various habitats,” explains the biologist. How do you collect so many images of an elusive animal like a shark or a rare one like a dugong, for example? To overcome this obstacle, the scientists turned to… social media.“The dugong is such an iconic animal that we figured that when people had the chance to film it, they probably posted the images on social media,” explains Laura Mannocci.
After searching Facebook, Instagram, and YouTube, the scientists identified 20 videos of dugongs, which they used to “train” the algorithm to recognize the animal. The result: thanks to deep learning, the program is capable of detecting and identifying 80% ofthe animals in the videos.“This method thus offers a new, effective way to count and map dugongs as well as other charismatic marine species with the goal of better protecting them. These surveys could lead to recommendations for designating certain areas as strict reserves if necessary,” explains Laura Mannocci.
