Emotion Recognition in Horses with Convolutional Neural Networks

Creating intelligent systems capable of recognizing emotions is a difficult task, especially when looking at in animals. This paper describes the process designing “proof concept” system to recognize horses. formed by two elements, detector and model. The fast region-based convolutional neural network that detects horses an image. model predicts those These elements were trained with multiple images until they achieved high accuracy their tasks. In total, 400 collected labeled train both while 40 used test system. Once components validated, combined into testable would detect equine based on established behavioral ethograms indicating emotional affect through head, neck, ear, muzzle, eye position. showed 80% validation set 65% set, demonstrating it possible predict animals using autonomous systems. Such has applications including further studies growing field animal as well veterinary determine physical welfare or other livestock.