Facial recognition can help conserve seals, scientists say | National

FREEPORT, Maine (AP) – Most face recognition technology is associated with uses such as surveillance and human face verification, but scientists believe they have discovered a new use for it – seal keeping.

Researchers at Colgate University have created SealNet, a database of seals created by photographing dozens of bays in Maine’s Casco Bay. The team found the accuracy of the device to identify marine mammals is almost 100%, which is not a small achievement in an ecosystem with thousands of seals.

Researchers are working on expanding their database to make it available to other scientists, said Krista Ingram, a professor of biology at Colgate and a member of the team. Expanding the database to include rare species, such as the Mediterranean and Hawaiian monk seals, could help inform conservation efforts to save them, she said.

Ingram said the catalog of seals and the use of learning machines to identify them could also help scientists gain a better idea of ​​where the seabed is located.

“Understanding their dispersal, understanding their model really helps to inform any conservation efforts for the coast,” she said. “For mobile marine mammals that move around a lot and find it difficult to take pictures in the water, we need to be able to identify individuals.”

Also Read :  How AI and ML are Revolutionizing the Logistics Segment

SealNet is designed to automatically detect faces in cropped images and recognize them based on facial features such as eyes and nose, just like humans. A similar tool called PrimNet, used for worms, has been used on seals in the past, but SealNet runs it better than Colgate researchers, Colgate researchers said.

Colgate published its findings in April in the journal Ecology and Evolution. They processed more than 1,700 images of more than 400 individual seals, the newspaper said.

“The ease and wealth of image data that can be processed using SealNet contributes to an important tool for studying the environment and behavior of marine mammals in the evolving field of conservation technology,” the paper said.

Harbor seals are a conservation success story in the United States, animals were once awarded in New England, where they were widely viewed by fishermen as insects in the 19th and early 20th centuries. But the Marine Mammal Protection Act, which reached its 50th anniversary in October, extended new protections to them and the population began to recover.

Also Read :  The Most Important Skills Technology Students Should Learn

Seals and other marine mammals have long been studied using satellite tracking devices. Jason Holmberg, chief executive of Wild Me, an Oregon-based company that works to bring machine learning to biologists, says using artificial intelligence to study them is one way to bring conservation to them. In the 21st century. Wild Me is forming a potential partnership with SealNet.

“This is a change and upgrade of ‘Big Brother’ style technology to a more conservative goal,” Holmberg said.

Harbor seals are now plentiful in New England waters, where they drag on rocks and enjoy boating, seals and beach trips. However, other seal species are still endangered. The Mediterranean monk seal is thought to be the world’s most endangered seal, with only a few hundred animals left.

Michelle Berger, a fellow scientist at the Shaw Institute in Maine who did not participate in the SealNet research, said the use of face recognition could provide more valuable data.

“Once the system was perfected, I was able to photograph a lot of interesting ecological applications for it,” Berger said. “If they can get to know the stamps and recognize them from year to year, that will give us a lot of information about the movement, how they move from place to place.”

Also Read :  Watch Google's ping pong robot pull off a 340-hit rally • TechCrunch

Colgate researchers are also working with FruitPunch, a Dutch artificial intelligence company, to improve some aspects of SealNet to encourage wider use. “FruitPunch is hiring dozens of scientists around the world to work on a challenge to streamline SealNet workflow,” said Tjomme Dooper, FruitPunch’s head of partnerships and growth.

Improving the automation of face recognition technology could make SealNet even more useful for scientists, Dooper said. It will open up new opportunities to study animals and help protect them, he said.

“What it does is help biologists study the behavior of sealed animals and population dynamics,” Dooper said. “Port seals are an important type of indicator for the ecosystems around them.”


Follow Patrick Whittle on Twitter: @pxwhittle


The Associated Press Climate and Environmental Coverage is supported by a number of private funds. See more about AP’s climate initiative here. AP is solely responsible for all content.

Copyright 2022 Associated Press. All rights reserved. This material may not be published, rewritten or redistributed without permission.

Source

Leave a Reply

Your email address will not be published.