Method for classifying dog vocalizations as a marker of anxious or aggressive behavior

УДК 616.995.1]639.1.091
DOI 10.33861/2071-8020-2022-3-35-37

Avilov A.V., Fomina A.S., Krikunova A.A., Ermakov A.M.

Summary. Dog barking is characterized by a wide range of acoustic parameters -frequency, tonality and rhythm. The acoustic parameters of barking depend on the context and can be used as a reflection of the internal state of the dog and its communication with the humans. It is assumed that rapid intermodal matching of vocalization frequency and body size is one of the foundations of successful intragroup interactions. The article describes a method developed by the authors for classifying dog vocalizations as a marker of anxiety or aggressive behavior, and also proposes a method for classifying dog vocalizations using the duration of vocalizations over a certain period of time. The authors describe an algorithm for registering the type of animal activity during vocalization, also proposed a block diagram of a device for dog vocalizations registration for integrating the proposed methodology into the ‘Smart Home’ system. Audio recordings of dog vocalizations and their spectrograms were used for the study. An increase in signal intensity during vocalization was recorded using a noise sensor or a microphone placed on the dog’s collar, and the amplitude and duration of the signal were recorded using a microcontroller that processes and classifies the animal’s alarm event with data transfer to a remote server and the owner’s Telegram bot. The results of the study conducted by the authors will help develop a method for determining the state of the animal using the acoustic method, determine the vocalization parameters of dogs under various environmental conditions, and understand how barking performs different functions in different contexts.

Keywords: dog, vocalization, barking, whining, marker, anxiety, aggressive behavior, noise sensor, signal amplitude, spectrogram.

References:

1. Avilov A.V. Авилов А.В. Issledovanie metodiki identifikatsii zabolevaniy golosovogo apparata cheloveka [Study of methods for identifying diseases of the human vocal apparatus]. - Actuale biotechnology. - 2015 (2 (13). - pp. 26-28.

2. Volodin I.A., Volodina E.V., Filatova O.A. Nelineynye fenomeny, opredelyayushchie vysokuyu strukturnuyu izmenchivost skuleniy domashney sobaki Canis familiaris (Carnivora, Canidae) [Nonlinear phenomena determining the high structural variability of whining in the domestic dog Canis familiaris (Carnivora, Canidae)]. - 2007.

3. Chulkina M.M., Volodin I.A., Volodina E.V. Individualnaya, polovaya i mezhporodnaya izmenchivost laya domashney sobaki, Canis familiaris (Carnivora, Canidae) [Individual, sexual and interbreed variability in the barking of the domestic dog, Canis familiaris (Carnivora, Canidae)]. - Zool. zhurn. - 2006 (85 (4). - pp. 544-555.

4-24. Vide supra.

Author affiliation:

Avilov Aleksey V., docent of the Department of Instrumentation and Biomedical Engineering of the Don State Technical University; 1, Gagarina sq., Rostov-on-Don, 344000; phone: 8-903-4363602; e-mail: znakav@mail.ru.

Fomina Anna S., docent of the Department of Biology and General Pathology of the Don State Technical University; 1, Gagarina sq., Rostov-on-Don, 344000; phone: 8-908-1742627; e-mail: a_bogun@mail.ru.

Krikunova Anastasia A., student of the Don State Technical University; 1, Gagarina sq., Rostov-on-Don, 344000; phone: 8-918-5342234; e-mail: lcd.05@mail.ru.

Responsible for correspondence with the editorial board: Ermakov Aleksey M., D.Sc. in Biology, professor, Dean of the Faculty of Bioengineering and Veterinary Medicine of the Don State Technical University; 1, Gagarina sq., Rostov-on-Don, 344000; phone: 8-928-2143344; e-mail: amermakov@ya.ru.