Artificial Intelligence in Fitness: Pose Estimation and Movement Correction

In this study, we focus on recent developments in the use of artificial intelligence (AI) and deep learning for fitness movement assessment and human pose estimation. The incorporation of AI techniques has become important in fitness tracking; therefore, methods such as convolutional neural networks and pose estimation models, including PoseNet and ConvNeXt, are increasingly utilized. Identified article topics are based on 2D and 3D pose estimation methods, real-time feedback technology, and augmented reality implementation plans. We also state that these technologies can be used to correct exercise postures, prevent injuries, and, generally, augment all fitness-related exercises because of accurate real-time feedback. The survey is rounded off with a focus on mobility and issues concerning it, while also extending the reach towards mobile-based applications.