Decoding pig grunts using Artificial Intelligence: A move towards animals’ mental wellbeing

Can an oink mean “I’m hungry”? Perhaps so!

A team of researchers recently published a study¹ that classifies 7,414 vocal expressions from 400 pigs, using AI, into positive and negative emotions. It gives the world an opportunity to enter the world of oinks and squeals, and understand the psychological behavior of pigs. This can help farmers monitor the mental health of animals by tracking their behavior.

The researchers, co-led by Dr Elodie F Briefer, an Associate Professor at the University of Copenhagen, Denmark, and an animal communications expert, trained a neural network to identify and classify pig calls as being positive or negative. They used sounds pertaining to different stages of a pig’s life cycle, from birth till death, to build the algorithm. For the study, they collected audio samples of pigs from various setups such as farms, commercial settings, and mock setups with toys, food, and random things to explore and play around with. Observing pig calls in different situations and scenarios helped them understand varied and nuanced emotions.

What comes across as mere grunts and snuffles is actually an entire book of emotional expressions, the study results indicated. The algorithm differentiated the calls that represented happy moments (such as getting close to other pigs or reuniting with families) from negative situations (such as fighting, or getting castrated).

Researchers associated high-pitched sounds with fear and anxiety, and low-pitched sounds with various situations across the scale. “In the positive situations, the calls are far shorter, with minor fluctuations in amplitude,”² said Dr Briefer.

With more recordings and advanced research on the subject, the researchers hope to dwell deeper into animal psychology and focus on the mental wellbeing of animals. Dr Briefer emphasized the need for the algorithm to be turned into an app that will help farmers identify the animals’ needs and work on them.

References:

1. Briefer, E.F., Sypherd, C.C.R., Linhart, P. et al. Classification of pig calls produced from birth to slaughter according to their emotional valence and context of production. Sci Rep 12, 3409 (2022). https://doi.org/10.1038/s41598-022-07174-8

2. Sample, I. Grunt of the litter: scientists use AI to decode pig calls. https://www.theguardian.com/science/2022/mar/07/scientists-use-ai-to-decode-pig-calls

Related articles:

Injured plants send chemical signals to neighboring plants

The role of regulatory bodies governing animal research

What do institutional animal ethics boards do?

How is artificial intelligence propelling science and research?