Predicting Dog Emotions Based on Posture Analysis Using DeepLabCut

This paper describes an emotion recognition system for dogs automatically identifying the emotions anger, fear, happiness, and relaxation. It is based on a previously trained machine learning model, which uses automatic pose estimation to differentiate emotional states of canines. Towards that goal, we have compiled picture library with full body dog pictures featuring 400 images 100 samples each “Anger”, “Fear”, “Happiness” “Relaxation”. A new keypoint detection model was built using framework DeepLabCut animal detector training. The newly learned from total 13,809 annotated possesses capability estimate coordinates 24 different part keypoints. Our application able determine dog’s state visually accuracy between 60% 70%, exceeding human recognize emotions.