| Application of AI in Volleyball Biomechanics: A Narrative Review |
| Paper ID : 1237-SPORTCONGRESS |
| Authors |
|
Ali Rasouli Pour Khameneh *1, Amirhossein Teymouri2 1Department of Sport Biomechanics, Faculty of Sport Science and Health, University of Tehran, Tehran, Iran 2Department of Sport injuries & Biomechanics, Faculty of Sport Sciences & Health, University of Tehran |
| Abstract |
| Artificial intelligence (AI) is rapidly transforming the field of sports biomechanics, offering new possibilities for performance analysis, injury prevention, and real-time feedback. This narrative review synthesizes current evidence on the application of AI in volleyball biomechanics, focusing on markerless motion capture, wearable sensor analytics, and machine learning–based performance evaluation. Traditional laboratory-based systems provide precise kinematic data but are limited by high cost and low ecological validity. Recent advances in deep learning pose estimation, such as OpenPose and HRNet, allow accurate three-dimensional (3D) reconstruction of volleyball movements using standard video footage, with reported mean absolute errors below 30 mm under controlled conditions. These tools enable the quantification of complex skills such as serves, spikes, and blocks in training and competition contexts. Complementarily, inertial measurement units (IMUs) coupled with supervised machine learning algorithms have demonstrated high accuracy in classifying jump types and quantifying training load, making continuous monitoring feasible. AI-driven video and sensor fusion systems are emerging for action recognition, automatic event tagging, and early detection of movement patterns linked to overuse or acute injuries. Despite these advances, significant challenges remain regarding data quality, occlusion handling, model generalization, and limited annotated datasets. Integrating physics-informed AI models, hybrid camera–IMU pipelines, and standardized validation frameworks will be essential for reliable application in the field. Ethical considerations, including data privacy and transparent model interpretation, are also critical for widespread adoption. Overall, AI holds strong potential to democratize biomechanical analysis in volleyball, translating complex motion data into actionable insights for coaches, clinicians, and athletes. Continued multidisciplinary collaboration will be key to achieving robust, interpretable, and ethically responsible AI solutions for high-performance sport. |
| Keywords |
| volleyball biomechanics, artificial intelligence, machine learning, markerless motion capture, wearable sensors, pose estimation, injury prevention |
| Status: Abstract Accepted (Poster Presentation) |
