From moo-d swings to purr-sonality: AI cracks the code of animal emotions with 89pc accuracy, says Danish study

COPENHAGEN, Feb 25 — Every day, when looking at our dogs or cats, we can easily perceive signs of joy, anger or sadness in these animals. Yet this impression is largely influenced by our own experience. The study of animal emotions is a complex subject that continues to pose many challenges to science. However, a recent Danish study, published in the journal, iScience, suggests that artificial intelligence could play a key role in better deciphering them.

Using artificial intelligence, researchers at the University of Copenhagen have succeeded in identifying and distinguishing between positive and negative emotions in seven species of ungulates, including cows, pigs and wild boars. To achieve this, they analysed thousands of vocalisations recorded in various emotional contexts.

Their study reveals several key indicators for differentiating positive from negative emotion, including sound duration, acoustic energy distribution, fundamental frequency and amplitude modulation. More surprisingly, these vocal patterns appear to be common to several species, suggesting that expressing emotions through the voice is a mechanism deeply rooted in evolution.

The team developed a machine learning model with an impressive accuracy rate of 89.49 per cent. According to the researchers, this is the first tool capable of analysing the emotional states of different animal species based solely on their vocalisations.

From cows to pigs and chickens

In the future, this breakthrough could have far-reaching implications, transforming the way we interact with farm animals and certain endangered wild species. In particular, it could pave the way for real-time animal welfare monitoring systems. “Understanding how animals express emotions can help us improve their well-being. If we can detect stress or discomfort early, we can intervene before it escalates. Equally important, we could also promote positive emotions. This would be a game-changer for animal welfare,” explains Élodie F. Briefer, Associate Professor at the Department of Biology and last author of the study, quoted in a news release.

With this in mind, the researchers have chosen to make their database public to enable other scientists to further explore the potential of AI in understanding animal emotions. “We want this to be a resource for other scientists. By making the data open access, we hope to accelerate research into how AI can help us better understand animals and improve their welfare,” adds Élodie F. Briefer.

This technological breakthrough marks another step towards a future where artificial intelligence’s decoding of animal emotions could revolutionise breeding and conservation practices. This field is attracting growing interest in the scientific world. In 2023, researchers at the University of Tokyo developed an AI capable of identifying the emotional state of hens by analysing their clucking. Prior to this, France’s National Research Institute for Agriculture, Food and Environment, in collaboration with the ETH Zurich and the University of Copenhagen, developed a system for interpreting pigs’ grunts.

While these technologies still need to be perfected, they could profoundly alter our relationship with animals, and help us to take better account of their emotions in day-to-day life. — ETX Studio