It is possible to identify pig screams from other sounds using monitoring equipment.
Pig screams can be clearly defined – knowing when a pig screams, it is easier to learn about their current state of health and welfare.
That was the conclusion of a group of researchers from Belgium and Italy, which was published in the scientific journal PLOS One. They defined screams to contain the following features discerning them from other sounds: a formant structure, adequate power, high frequency content, sufficient variability and duration.
Two properties of lesser importance were the high frequency content and the variability of the signal. They added that it was not necessary for a scream to have all these properties.
Pig vocalisations – what are they?
The researchers wrote in their abstract that pig vocalisations convey information about their current state of health and welfare. Continuously monitoring these vocalisations can provide useful information for pig producers, they said. For instance, pig screams can indicate stressful situations.
They added that when monitoring screams, other sounds can interfere with scream detection. Therefore, they concluded, identifying screams from other sounds would be essential and they set out to understand which sound features define a scream.
Method to detect screams
A method to detect screams based on sound features was developed, with physical meaning and explicit rules. The researchers used 7 hours of labelled data from 24 pigs. The developed detection method attained 72% sensitivity, 91% specificity and 83% precision.
Interested in listening to the pig screams that were analysed? Search in the Animal Sound Archive for 'Sus Scrofa Domestica'.
The research, called 'Discerning Pig Screams in Production Environments, was published in 2015. It was authored by J. Vandermeulen, C. Bahr, M. Kashiha, T.A. Niewold and D. Berckmans, KU Leuven, Belgium; E. Tullo, I. Fontana and M. Guarino, The University of Milan, Italy; S. Ott and C.P.H. Moons, Ghent University, Belgium; F.A.M. Tuyttens, Institute for Agricultural and Fisheries Research (ILVO), Belgium.