Related Experiment Video
Updated: Aug 5, 2026
08:16
Movement Retraining using Real-time Feedback of Performance
Published on: January 17, 2013
From Optical to AI-Driven Markerless Motion Capture in Motor Learning and Rehabilitation
Panagiotis Georganakis1, Konstantinos Spinthiropoulos1, Konstantinos Panitsidis1
1Department of Management Science and Technology, School of Economics, University of Western Macedonia, 50100 Kozani, Greece.
Bioengineering (Basel, Switzerland)
|July 28, 2026
Summary
Artificial intelligence-based markerless motion capture (MMC) offers a cost-effective, accessible alternative to traditional biomechanical analysis. This technology democratizes movement science for clinical rehabilitation and patient monitoring.
Area of Science:
- Biomechanical analysis
- Movement science
- Clinical rehabilitation
Background:
- Traditional marker-based motion capture is limited by high costs and physical constraints.
- These limitations hinder widespread clinical adoption and accessibility in movement science.
Purpose of the Study:
- To review the advancements in artificial intelligence (AI)-based markerless motion capture (MMC) for clinical rehabilitation.
- To evaluate AI-MMC as a democratizing approach for movement analysis.
Main Methods:
- The review traces the evolution of MMC from geometric visual hulls to deep learning models.
- Focus on YOLO-based 2D detection and spatio-temporal transformers for 3D pose estimation.
Main Results:
- Multi-camera AI-MMC achieves research-grade accuracy (16-34 mm MPJPE).
- Monocular systems show high sensitivity (82-88%) for monitoring geriatric fall risk and stroke recovery.
Conclusions:
- AI-based MMC redefines movement analysis, making it accessible beyond laboratory settings.
- Real-time signal refinement enables objective, ecologically valid assessments in community healthcare.
