Scientists say facial recognition could help preserve seals

Scientists say facial recognition could help preserve seals

By Patrick Whittle, Related Press

Freeport, Maine (AP) — Facial recognition know-how is usually related to makes use of like monitoring and authenticating human faces, however scientists suppose they’ve discovered a brand new use for it — seal seals.

A analysis group at Colgate College has developed SealNet, a database of seal faces that was generated by taking images of dozens of harbor seals in Casco Bay, Maine. The group discovered that the instrument’s accuracy in figuring out marine mammals is near 100%, which isn’t any small feat in an ecosystem residence to 1000’s of seals.

Christa Ingram, a professor of biology at Colgate and a member of the group, mentioned the researchers are working to increase their database to make it accessible to different scientists. She mentioned increasing the database to incorporate uncommon species such because the Mediterranean monk seal and Hawaiian seal may assist conservation efforts to avoid wasting these species.

Ingram mentioned cataloging seal faces and utilizing machine studying to determine them may additionally assist scientists get a greater concept of ​​the place seals are within the ocean.

Political cartoon

“Understanding their unfold and understanding their patterns actually helps inform any coastal conservation efforts,” she mentioned. “For cellular marine mammals that transfer so much and are troublesome to {photograph} within the water, we’d like to have the ability to determine people.”

SealNet is designed to robotically detect a face in a photograph, crop it, and acknowledge it primarily based on facial patterns corresponding to eyes and nostril form, simply as it will for a human. Colgate researchers mentioned the same instrument known as PrimNet supposed to be used on primates had beforehand been used on seals, however was outperformed by SealNet.

The Colgate group revealed their findings in April within the scientific journal Ecology and Evolution. The newspaper mentioned they processed greater than 1,700 photographs of greater than 400 particular person seals.

The paper acknowledged, “The benefit and wealth of picture knowledge that may be processed utilizing SealNet software program contributes to an important instrument for ecological and behavioral research of marine mammals within the growth of conservation know-how.”

Harbor seals are a conservation success story in the USA. The animals have been as soon as topic to bounty in New England, the place they have been extensively considered as pests by hunters within the nineteenth and early twentieth centuries. However the Marine Mammal Safety Act, which turned 50 in October, has expanded new protections for them — and populations are beginning to rebound.

Seals and different marine mammals have lengthy been studied utilizing satellite tv for pc trackers. Utilizing AI to check them is a solution to convey conservation into the twenty first century, mentioned Jason Holmberg, CEO of Wild Me, an Oregon-based firm that gives machine studying to biologists. Wild Me is growing a possible partnership with SealNet.

“This can be a transformation and upliftment from a ‘Massive Brother’ model to a really charitable function of conservation model,” mentioned Holmberg.

Harbor seals are actually pretty plentiful in New England waters, hopping on the rocks and delighting sea seal-watching excursions and beach-goers. Different seal species, nonetheless, are nonetheless in peril. The Mediterranean seal is believed to be probably the most endangered seal on the earth with only some hundred animals left.

Utilizing facial recognition may present extra priceless knowledge, mentioned Michelle Berger, an affiliate scientist on the Shaw Institute in Maine who was not concerned within the SealNet analysis.

“As soon as the system is full, I can think about plenty of fascinating environmental purposes for it,” Berger mentioned. “If they’ll determine the seals, and determine them from 12 months to 12 months, that can give us plenty of details about the motion, and the way far they journey from one web site to the subsequent.”

The Colgate researchers are additionally working with FruitPunch, a Dutch synthetic intelligence firm, to enhance some elements of SealNet to encourage wider use. FruitPunch is bringing in a couple of dozen scientists all over the world to work on a problem to streamline SealNet’s workflow, mentioned Tjomme Dooper, FruitPunch’s head of partnerships and progress.

Dubier mentioned that enhancing the automation of facial recognition know-how may make SealNet extra helpful to extra scientists. He mentioned that may open up new alternatives to check the animals and assist shield them.

“What this does is assist biologists examine seal conduct, in addition to aggregation dynamics,” Duber mentioned. “Port seals are vital indicators of species for the ecosystem round them.”

Comply with Patrick Whittle on Twitter: @pxwhittle

The Related Press’s local weather and environmental protection receives help from a number of personal foundations. Study extra in regards to the AP’s local weather initiative right here. AP is solely chargeable for all content material.

Copyright 2022 News agency. All rights reserved. This materials will not be revealed, broadcast, rewritten, or redistributed.

#Scientists #facial #recognition #protect #seals

Leave a Comment

Your email address will not be published. Required fields are marked *