seal
Credit: CC0 Public Domain

Scientists believe that facial recognition technology can be used to save seals.

SealNet is a database of seal faces created by taking pictures of dozens of seals in Maine. The tool's accuracy in identifying marine mammals is close to 100%, which is no small achievement in an environment where thousands of seals live.

Krista Ingram is a biology professor at Colgate and a team member. The Mediterranean monk seal and Hawaiian monk seal are examples of rare species that could be included in the database.

Scientists can get a better idea of where in the ocean seals are with the help of cataloging and machine learning.

Understanding their dispersal helps inform any efforts to conserve the coast. Mobile marine mammals are hard to photograph in the water and we need to be able to identify individuals.

SealNet is designed to automatically detect the face in a picture, crop it and recognize it based on facial patterns, like a human would. The Colgate researchers said that SealNet was better than a tool called PrimNet that had been used on seals.

The Colgate team's findings were published in the journal. The paper said they processed more than 1,700 images.

The paper states that the "ease and wealth of image data that can be processed using SealNet software contributes a vital tool for ecological and behavioral studies of marine mammals."

In the 19th and early 20th century harbor seals were seen as pests in New England and were subject to bounties. The Marine Mammal Protection Act extendedTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkiaTrademarkia

Satellite tracking has been used to study seals and other mammals. The executive director of Wild Me, an Oregon-based company that works to bring machine learning to biologists, said that artificial intelligence can be used to study animals. Wild Me is in talks with SealNet.

"This is a shift and a lift of 'big brother' style technology to a very benevolentConservation-style goal."

There are a lot of harbor seals in New England waters where they haul out on rocks. There are other seal species that are in danger. The Mediterranean monk seal is believed to be the most threatened seal in the world.

An associate scientist at the Shaw Institute in Maine who was not involved in the SealNet research said that facial recognition could provide more valuable data.

Berger said that he can see lots of interesting ecological applications for the system. It would give us a lot of information about movement if they could recognize seals from year to year.

Colgate researchers are working with FruitPunch to improve some aspects of SealNet to encourage wider use. FruitPunch is getting a few dozen scientists around the world to work on a challenge to streamline SealNet's workflows.

Dooper said that SealNet could be more useful to more scientists. He said that it would allow for new opportunities to study the animals.

It helps the biologists study the behavior of seals and population dynamics. TheHarbor seals are an indicator species.

SEALNET: Facial recognition software for ecological studies of harbor seals is the topic of a paper by the authors. It's called ece 3.8851.

Journal information: Ecology and Evolution

The Associated Press was published in 1992. All rights belong to the person. The material may not be published, broadcast, or redistributed.