An algorithm built by European researchers could help farmers “speak pig” to improve the animals’ welfare.
At any given moment, there are as many as 12,500 Duroc hogs snorting around the barnyards of Imani Farms, a pig farm in southwestern Ontario.
The farm’s pens are a cacophony of squeals, screams, barks and grunts, with each sound telegraphing a different feeling or need. Pigs are expressive animals with a wide range of vocalizations, according to Stewart Skinner, 38, a co-owner of the farm. Interpreting their calls can occasionally stump even experienced farmers.
“I have often joked that this job would be far easier if we could speak pig,” Mr. Skinner said.
Decoding the emotions behind those oinks could soon become a little easier. Researchers in Europe have created an algorithm that assesses pigs’ emotional states based on the sound the animals make.
“Animal welfare is nowadays widely accepted to be based not only on the physical health of animals, but also their mental health,” said Elodie Briefer, an associate professor of biology at the University of Copenhagen and an author of the study published this week in the journal Scientific Reports. The sooner a farmer can discern whether an animal is pleased or distressed, the faster any issues in the animal’s environment that may affect its health can be addressed.
Pigs are among the more voluble of domestic animals, producing a wider range of sounds more frequently than relatively taciturn goats, sheep and cows. To crack the code of pig communication, scientists in five research labs across Europe used hand-held microphones to gather roughly 7,400 distinct calls from 411 individual pigs. The calls were recorded during all types of situations in the life span of a pig, from birth to the slaughterhouse.
Researchers then assigned each sound a positive or negative emotional value based on what the paper calls “intuitive inference.” In other words, researchers made an educated guess about how the pig likely felt about the event at which the sound was recorded (i.e. feeding, good; castration, bad).
Upon first listen, most people tend to do slightly better than chance at guessing a pig’s feelings based on its sound alone. Listen closely to enough pig calls, though, and patterns emerge.
Grunts associated with positive emotions — the sounds pigs make when feeding, running or reuniting with their mothers or littermates after a separation — tend to be shorter, and have a one-note consistency in tone.
Unsurprisingly, an unhappy pig sounds awful. Situations that produced cries of distress included being inadvertently crushed by a mother sow (a common peril for piglets), awaiting slaughter, hunger, fights and the unwelcome surprise of strange people or objects in their pens. The screams, squeals and barks recorded from animals experiencing fear or pain are both longer in duration and more variable in tone than the sounds of contentment.
When taught to listen for these simple distinctions, humans do a better job of accurately interpreting an animal’s emotional state, Dr. Briefer said. But artificial intelligence performed best of all. The researchers’ algorithm, designed by co-author Ciara Sypherd, correctly identified the animal’s emotion as positive or negative 92 percent of the time.
The study is the product of SoundWel, a project sponsored by the European Union to improve animal health and welfare. Researchers with the project are now looking to partner with an engineer who can incorporate their data into an app or other tool that farmers could use to interpret their animals’ calls, and emotional state, in real time, Dr. Briefer said.
Understanding animals’ emotions has practical and legal consequences. Animal sentience laws like the one currently before Britain’s parliament assert that animals are capable of thought and feeling, and that the government must take their welfare into account when making policies that might affect them. The European Union recognized animal sentience in 2009.
A cost-effective and user-friendly tool for decoding pig grunts could be a valuable asset on a farm, Mr. Skinner said.
“The ability to recognize problems early is the largest determining factor in success of treatment,” Mr. Skinner said. “Any tool that is adaptable to barn settings that would increase the understanding of what the individual animals are feeling would have value.”
#News | https://sciencespies.com/news/when-pigs-cry-tool-decodes-the-emotional-lives-of-swine/
No comments:
Post a Comment