Play Live Radio
Next Up:
0:00
0:00
0:00 0:00
Available On Air Stations

A computer program designed to sort mice squeaks is also finding whales in the deep

The green boxes show portions of the audio spectrogram that artificial intelligence has identified as marine mammal calls.
Ocean Science Analytics
The green boxes show portions of the audio spectrogram that artificial intelligence has identified as marine mammal calls.

Updated May 31, 2022 at 12:15 PM ET

Deep Squeak is the name of an artificial intelligence program that was designed to detect the high-frequency "squeaks" mice and rats make when they are stressed.

But a new application of the technology is putting a much bigger emphasis on the "deep": It's being used to search for whales and other marine mammals in a ocean environments.

If that seems like a classic case of mislabeling, blame marine ecologist Elizabeth Ferguson and her company Ocean Science Analytics, which leads the project.

One of the company's lines of work is helping people building offshore wind farms track the impact of their projects on marine mammals, to make sure they aren't being harmed.

"Any kind of operations that happen in the ocean require there be some monitoring or mitigation," Ferguson says.

You could just go out in a boat and look for whales and dolphins in the area of interest, but she says that doesn't always give you an accurate count: "Some species are difficult to see at the surface or they spent a long time at depth."

Teaching a computer to spot squeaks

She found a different solution in the work of Kevin Coffey, a behavioral neuroscientist at the University of Washington who studies the calls rats and mice make when they're stressed. Those call are different from the sounds they make when they're not stressed.

On his longer-term projects, someone in his lab often got stuck listening to many hours of audio to identify the rodent calls. He and his colleagues at the University of Washington thought they could turn to artificial intelligence to ease that burden.

"You take the audio signal you turn it into an image and then you can you can see the calls by eye," Coffey says. And computers have gotten very good at analyzing and identifying images using what's called deep learning.

Coffey created a program that was good at classifying the visual representations of the mouse calls as stressed or non-stressed, and called it Deep Squeak.

Searching for undersea songs

Elizabeth Ferguson heard about the program and figured that what works for mice in cages could be modified to work with marine mammals in the ocean.

She shows the results of using her modified version of Deep Squeak on about two and a half hours of audio recorded within a couple of miles of the Oregon coast. The program has drawn a green box around anything it thinks looks like a marine mammal sound.

"You can see that there's definitely a wide range of calls and a high degree of variability in those calls But it's still done a pretty good job of detecting them," Ferguson says.

What's in a name?

But really: Is Deep Squeak the name you want to use for a program that detects whale calls?

"No we're going to change it," Ferguson says with a laugh. "So we're going to be calling at 'Deep Waves.' "

I told her I didn't think that had the same panache.

"Should we find something better? Have any suggestions?"

So far, I haven't. But if you have an idea, drop me a line. Jpalca@npr.org. I'll pass it along.

Copyright 2023 NPR. To see more, visit https://www.npr.org.

Joe Palca is a science correspondent for NPR. Since joining NPR in 1992, Palca has covered a range of science topics — everything from biomedical research to astronomy. He is currently focused on the eponymous series, "Joe's Big Idea." Stories in the series explore the minds and motivations of scientists and inventors. Palca is also the founder of NPR Scicommers – A science communication collective.