A team of New York scientists have discovered a new, life-saving use for facial recognition technology: saving seals. A team of researchers from Colgate University have recently completed a project called SealNet, ABC News reports. SealNet is a collection of photographs featuring the faces dozens of harbor seals native to Casco Bay. The team found that the tool was nearly 100% accurate in identifying marine mammals. This is a remarkable achievement considering there are thousands of seals in the ecosystem.
Krista Ingram is a biology professor at Colgate and team member. She said that the researchers are expanding their database to make it available to other scientists. They’ve added unusual species like the Mediterranean monk seal and Hawaiian monk seal to the database. This could help people understand how best to protect those animals, she said.
According to Ingram, by cataloguing seal faces and using machine learning to identify them, scientists can get a better understanding of where seals are found in the ocean. “Understanding their patterns really helps inform any conservation efforts for the coast,” Ingram explained. “For mobile marine mammals that move around a lot and are hard to photograph in the water, we need to be able to identify individuals.”
The program uses over 1,700 images of seals
The team processed more than 1,700 images of over 400 individual seals for their research paper. According to the paper, SealNet is revolutionary for wildlife conservation because of how much data it can process.
The Marine Mammal Protection Act, which turned 50 in October, granted Harbor seals new protections. It stopped the hunting of them– causing their populations to increase rapidly.
AI has been used to study seals and other marine mammals for quite some time. This is according to Jason Holmberg, executive director of Wild Me. Wild Me is an Oregon-based company that works to bring machine learning to biologists and is currently in talks with SealNet about a possible partnership. “This is a shift and a lift of ‘big brother’ style technology to a very benevolent conservation-style goal,” Holmberg explained.
Though harbor seals are nowpopulating New England waters, 16 other species remain in jeopardy. The particular danger to the world’s most endangered seal, the Mediterranean monk, is lurking with only a few hundred animals remaining.
The use of facial recognition could provide more valuable data, said Michelle Berger, an associate scientist at the Shaw Institute in Maine. “If they could recognize seals, and recognize them from year to year, that would give us lots of information about movement, how much they move from site to site.”
Besides Colgate, the Dutch AI company FruitPunch is also working to further improve SealNet’s efficiency so that more people will begin to use it. To do this, they have gathered a team of around two dozen scientists from all over the world. They are now focusing on streamlining SealNet’s current workflow.