Flash News

E-TJERA

Will artificial intelligence help us talk to animals?

Will artificial intelligence help us talk to animals?

Nature is noisy. If you sit on a rock in the forest, on a mountain, or in a field, you will usually hear something chirping.

There has been a boom in our understanding of animal communication in recent years, and we owe it in part to artificial intelligence (AI).

AI allows researchers to analyze vast amounts of audio data of animal calls in seconds, which would take humans decades to analyze.

There are hundreds of AI tools to analyze the vocalizations of different species.

Kevin Coffey, a neuroscientist at the University of Washington, US, helped build DeepSqueak, a machine learning tool that decodes rodent chatter.

DeepSqueak extracts rodent calls from raw audio data, compares them to calls with similar characteristics, and provides insights into animal behavior.

"Mice use ultrasonic vocalizations (USVs). The high-pitched 50 kilohertz (kHz) calls have been described as similar to a laugh, but there are many types of these made in different positive situations, such as play, mating, or even during a drug overdose," Coffey said.

Rats also have 22 kHz calls that are used in negative situations, such as when they are in pain or vomiting. Coffey uses those frequencies to indicate when an experiment is making his lab rats feel bad.

Humans cannot hear these calls as they are outside the frequency range that humans can hear, but DeepSqueak and other tools can help us decipher them.

Since its introduction in 2018, DeepSqueak has been used to study rodent social behavior, drug use, autism, and more. It has also been modified for use with many other species, such as dolphins, monkeys, and birds.

Latest news