They have designed an algorithm from 7,000 recordings that can decode whether a pig is happy, sad, or stressed
We love them for their hams, but pigs aren’t just a source of legged pork. They are mammals with emotions, and perhaps if we understood them better we could return the favor and improve their lives before they end up on the plate. With this goal in mind, an international team of scientists has succeeded, for the first time, in translating the grunts of pigs into real emotions. To do this, they have used more than 7,000 acoustic recordings collected throughout the animals’ lives, so that they reliably know when they express happiness and when they are scared or stressed. The findings have just been published in ‘Scientific Reports.
The researchers, led by the University of Copenhagen, ETH Zurich, and France’s National Research Institute for Agriculture, Food and the Environment (INRAE), recorded the grunting of pigs intended for human consumption, from birth to death. , in a wide variety of situations.
Some are pleasant, such as nursing piglets, snuggling with litter members, or running free, while others are negative, such as fighting, castration, separation from the group, or waiting for slaughter.
Additionally, the researchers created several simulated scenarios for the pigs, designed to evoke more nuanced emotions. These included an arena with toys or food and an arena without any stimulation. They also placed new and unfamiliar objects in the experiment for the pigs to interact with. Along the way, the animals’ calls, behavior, and heart rate were monitored and recorded where possible.
The scientists then analyzed the thousands of audio recordings from more than 400 pigs to see if there was a pattern to sounds based on emotions and if they could distinguish between positive and negative situations and emotions. They collected more high-frequency calls (such as yelling and shrieking) in negative situations. And much shorter, low-frequency calls (grunts) occurred in positive situations.
‘By training an algorithm to recognize these sounds, we can classify 92% of the calls into the correct emotion,’ explains Elodie Briefer from the Department of Biology at the University of Copenhagen and co-leader of the study. Getting them translated is “an important step towards improving animal welfare for livestock,” she says.
Animal welfare
As the authors explain, the study of animal emotions is a relatively new field that has emerged in the last 20 years. Today, it is widely accepted that the mental health of cattle is important to their general well-being. However, current animal welfare focuses primarily on the physical health of livestock. In fact, there are several systems that can automatically monitor the physical health of an animal for a farmer.
Analogous systems for monitoring the mental health of animals have not yet been developed. The researchers hope their algorithm could pave the way for a new platform for farmers to monitor the psychological well-being of their animals. Without having to lay them down on a couch.
“We have trained the algorithm to decode pig grunting. Now, we need someone who wants to develop the algorithm into an application that farmers can use to improve the welfare of their animals,” says Briefer. Moreover, in his opinion, with enough data to train the algorithm, the method could also be used to better understand the emotions of other mammals.
