OkeyDoggy3D: A Mobile Application for Recognizing Stress-Related Behaviors in Companion Dogs Based on Three-Dimensional Pose Estimation through Deep Learning

Dogs often express their stress through physical motions that can be recognized by owners. We propose a mobile application analyzes companion dog’s behavior and three-dimensional poses via deep learning. As existing research on pose estimation has focused humans, obtaining large dataset comprising images showing animal joint locations is challenge. Nevertheless, we generated such used it to train an AI model. Furthermore, analyzed circling behavior, which associated with in dogs. To this end, the VideoPose3D model estimate 3D of dogs from 2D technique derived DeepLabCut developed app provides analytical information stress-related behaviors, as well walking isolation times, Finally, interviewed five certified experts evaluate validity applicability app.