Dog pain detection with computer vision

This research project aims to automatically recognize pain in dogs and is based on previous research by Zhu et al. (2022), which is one of the first works on machine learning based pain estimation of dogs. While acute pain is generally easily discernible by owners due to the expressiveness of acute pain signals, dogs suffering from chronic pain are great at masking pain signals since they must deal with it daily. Especially for dogs in chronic pain, it can be fruitful to perform automatic pain detection to discover hidden pain and facilitate proper veterinary treatment. This research will extend the findings of Zhu et al. (2022), by integrating behavior prediction into the streamline to aid in the pain detection, replacing older networks and adding more explainability to the results such that the veterinarians can pinpoint the source of the dog’s pain.

Project leader

Lisanne Koetsier, MSc student

Academic supervisors

Prof. dr. A.A. (Albert) Salah, Prof. dr. R.C. (Remco) Veltkamp

Other partners

Ankara University

Relevant publications

Zhu, H., Salgırlı, Y., Can, P., Atılgan, D., & Salah, A. A. (2023). Video-based estimation of pain indicators in dogs. In 2023 11th International Conference on Affective Computing and Intelligent Interaction (ACII) (pp. 1-8). IEEE.