Head of Laboratory

Anna Sergeevna Fomina

PhD in Biological/biomedical sciences, associate professor
Publications
18
Citations
11
h-index
1
Authorization required.

The main activity of the laboratory is aimed at studying the behavior of small domestic and farm animals, including in the "owner-pet" dyad. The laboratory uses both classical methods of zoopsychology and ethology, as well as modern technologies related to the use of neural networks for the analysis and interpretation of behavior markers. The welfare of companion animals in the context of interaction with humans is being investigated.

  1. Observation
  2. Neural networks
  3. Thermography
  4. Electrocardiography (ECG)
  5. This is a program
  6. Survey
Anna Fomina 🥼 🤝
Head of Laboratory
Anastasiya Krikunova
Junior researcher

Research directions

An integrated approach to identifying problematic behavior of companion dogs using artificial intelligence technology

+
During the first year, methods and test samples were selected and tested to identify markers of problematic dog behavior using behavioral, physiological indicators, and artificial intelligence technology. The object of the study was the behavior of companion dogs due to an increased level of anxiety. The novelty lies in the development of a comprehensive method that makes it possible, based on measurable metrics of behavioral and physiological indicators using neural networks, to differentiate dogs into calm, anxious and excitable. This is the first time such a study has been conducted in the Russian Federation. Promo video of the project: https://disk .yandex.ru/i/o6jn-Qz4ky9WUA

Prototype of the Dog Nanny contactless activity monitoring system

+
The object of the development is a prototype contactless device for determining the anxiety level of companion dogs. The aim of the work was to develop a prototype of a contactless device for determining the anxious behavior of dogs based on the analysis of vocalization intervals and video recording of motor activity. Vocalizations were analyzed using a microcontroller, Audacity software package, and a Behringer XM8500 Behringer XM8500 microphone with a sampling frequency of 44 kHz and a wide frequency bandwidth from 50 HZ to 15 kHz. The Arduino Software (IDE) programming environment was used to write the program and firmware of the microcontroller. Data about the voice event with a note about the alarm level is transmitted to the remote server and the Telegram bot. As a result of the work carried out, a structural diagram of the device for recording and classifying dog vocalizations has been developed. A prototype system has been developed to identify dogs in photos and videos in real and delayed time. The device allows, based on the calculation and analysis of vocalization intervals, to determine and send information about dog anxiety, data on animal activity and ambient temperature to the Telegram bot. The field of application of the results: cynology, zoopsychology, veterinary medicine, engineering development of portable devices, tracking systems. Implementation recommendations: The device can be used to determine the anxiety level of companion dogs, service dogs, and dogs in shelters. The significance of this development is determined by the creation of a prototype of a software and hardware complex that provides an objective analysis of information about changes in dog behavior in real time. The use of the complex makes it possible to evaluate the behavior of dogs and creates the opportunity to obtain new data on the likelihood of violations of dog behavior caused by a high level of anxiety. The complex can act as a prototype for creating systems for tracking and identifying the behavior of other animal species.

A system for evaluating the motor activity of chickens as a predictor of productivity based on artificial intelligence technology

+
The aim of the project is to create an information system (IS) for automatic analysis of poultry welfare using computer vision methods. The created IP is aimed at assessing the correlation of indicators of production efficiency and motor activity of poultry at rest and in response to the action of a disturbing factor. The solved need of the agroindustrial complex is a routine multidimensional assessment of the well-being of poultry and forecasting productivity based on data on motor activity. The analysis of video images of poultry using IP is a way to non-contact assessment of the comfort of housing conditions and signs of maladaptation of farm animals and poultry. The implementation of the developed IP will create an opportunity to improve poultry management, advance forecasting and prevent a decrease in productivity. The final result of the IS work is the possibility of adaptive management of the welfare of farm animals and poultry based on the results of an objective non-contact assessment of motor activity.

Publications and patents

Found 

Partners

Lab address

пл. Гагарина, 1, корп. 7, Ростов-на-Дону
Authorization required.