Say Hello to the New Digital DR DOLITTLE: AI that Knows How to Read Cats' Emotions!
Marcelo Feighelstein#, Anna Zamansky and Ilan Shimshoni
Department of Information Systems
PhD Grant 2021
Cats have historically been known to hide their pain and make it difficult to detect.
A groundbreaking new study from the University of Haifa in Israel, in collaboration with researchers from universities from Sao Pablo in Brazil and Lincoln in the United Kingdom, has successfully developed artificial intelligence models capable of identifying pain in cats. This development could revolutionize the way pet owners care for their furry friends.
In the study, led by Marcelo Feighelstein, PhD student under the supervision of Prof. Anna Zamansky from Tech4Animal Lab and Prof. Ilan Shimshoni from Data Science Research Center of University of Haifa which was published in Nature Science Reports, 29 short-hair British cats were brought in to be sterilized. Researchers analyzed 48 specific points on the cats’ faces, including their eyes, ears, and mustaches, using and also the whole cat face image using two AI models. The results were impressive, with the AI models able to identify pain in cats with over 70% accuracy. The researchers found that the areas around the mouth and eyes were the most significant in detecting pain.
The Tech4Animals Lab has the ultimate goal of developing a digital DR DOLITTLE – AI models that can read animals’ emotions based on their facial expressions and behavior. The first step towards achieving this goal is the app they are developing that will allow cat owners to take a picture of their pet and check if they are in pain, without physically touching or holding them.
This research demonstrates the power of artificial intelligence to improve communication between humans and their animal neighbors on the planet. The new digital DR DOLITTLE has the potential to greatly improve the lives of both pets and their owners, making it easier to identify and alleviate pain in cats.