Optical Motion Capture

Optical motion capture (MoCap) aims to accurately record and reconstruct human movement, traditionally using marker-based systems that require extensive manual labeling and correction. Current research focuses on developing robust, automated methods using machine learning, including graph neural networks and generative adversarial networks (GANs), to address challenges like noisy data, occlusions, and the need for fewer sensors. These advancements are improving the accuracy and efficiency of MoCap, enabling applications in sports performance analysis, animation, and human-computer interaction, while also exploring alternative sensing modalities like radio frequency signals.