Exploring Canine Emotions: A Transfer Learning and 3DCNN-Based Study for Small Databases

This study aims to categorize the following dogs' primary emotional states: anger, fear, and happiness. We used transfer learning (TL) and advanced 2D/3D Convolutional Neural Networks (CNNs) to identify these emotions from two small datasets: static images and video clips. The image data set, annotated by two veterinarian experts, covers various canine emotional expressions in various contexts, which forms the basis for training the dog emotion recognition transfer learning neural network algorithm (DER-TLNN). DER-TLNN has an accuracy rate of 89 % for emotion classification based on static images (400 images). In addition, DER-TLNN is the basis for the Video Dog Emotion Recognition Network (VDERNet) algorithm, which is designed to categorize video clips that depict dogs' emotional states. Trained on expert-annotated video data sourced online (39 videos), VDERNet achieves an accuracy of 93 %, highlighting AI's potential in understanding canine emotions in dynamic contexts with non-invasive techniques. These findings demonstrate the efficacy of TL and 3D CNNs in animal behavior analysis and contribute to a better understanding of canine emotions. The research provides promising applications to improve humandog interactions and canine welfare, showcasing the immense potential of AI in animal behavior studies. The achievement of notable results with small datasets, along with data augmentation and regularization techniques, underscores the potential impact of AI in animal behavior studies.