Artificial intelligence has learned to read the emotions of animals
Scientists at the Wageningen University in the Netherlands have created an artificial intelligence-based computer system that can recognize the emotions of animals. In particular, it determines how happy they are. The authors hope that the system will improve the living conditions of livestock on agricultural farms, according to bioRxiv.
Several thousand photos of cows and pigs from six farms around the world have been uploaded to a network called WUR Wolf. After training, she gave an accurate result 85% of the time.
The scientists explained that the system analyzes 13 signs in the animal's “facial expressions”, including eye and ear movements. This data is then compared with nine emotional states: including aggressiveness, calmness, or neutrality of the animal.
So, if a cow keeps her ears upright, she is most likely horny. If the ears are pointing forward, this is a sign of negative emotions, such as frustration.
Rapid ear movements in pigs indicate stress. If the ears hang or curl in the direction of the eyes – the animal is calm. One of the pigs during the experiment was described as “alert and neutral”.
The scientists noted that it will take several more years before their system is ready for large-scale use on farms. But after its implementation, farmers will be able to constantly monitor the condition of livestock and provide them with better conditions.
“There is a need to move from simply eliminating negative emotional states to providing positive states such as playful behavior,” they said.
Earlier it was reported that artificial intelligence has learned to guess the name of a person from a photograph. The technology began to be developed in 2018.