14 DAY TRIAL // Learning the dog "speech" can seem impossible, but a new research published in the journal Animal Cognition could help us: a software that translates dog "language" to us. The software has been developed by Csaba Molnar from Eoetvoes Lorand University in Hungary.
The program can assign dog barks to different situations and even recognize barks from various dogs, tasks extremely difficult for most of us.
The software checked over 6000 barks from 14 Hungarian sheepdogs (Mudi breed) in 6 contexts: 'stranger', 'fight', 'walk', 'alone', 'ball' and 'play'. The barks' recordings were digitalized by a computer, and their sounds analyzed, coded and classified.
The first trial assessed the software's ability to classify barks into various contexts: it was right in 43 % of situations. The software worked best for 'fight' and 'stranger' situation, and worst for 'play' barks. Various motivational states of dogs in aggressive, friendly or submissive behavior may come with acoustically very distinct barks.
"After analyzing digital versions of the barks, overall the computer program correctly identified the kinds of barks the dogs made 43 % of the time - about the same as humans' 40 %. The software identified 'walk' and 'ball' barks better than people, although people identified 'play' and 'alone' barks better than the software," Moln?r said.
A second trial determined the software's ability to recognize barks of individual dogs: it worked in 52% of cases. Humans cannot make this difference, but it clearly appears there are sonic differences in barks of various individual dogs.
"The use of advanced machine learning algorithms to classify and analyze animal sounds opens new perspectives for the understanding of animal communication?The promising results obtained strongly suggest that advanced machine learning approaches deserve to be considered as a new relevant tool for ethology," the authors wrote.