Traditional biomechanical analysis is constrained by high capital costs and the physical limitations imposed by markers, posing significant barriers to clinical adoption. This review evaluates the emergence of artificial intelligence (AI)-based markerless motion capture (MMC) as a transformative approach for democratizing movement science in clinical rehabilitation. The discussion outlines the progression from legacy geometric visual hulls to advanced deep learning architectures, with particular focus on YOLO-based two-dimensional detection and spatio-temporal transformer models for three-dimensional pose estimation. Evidence indicates that multi-camera MMC frameworks achieve research-grade positional accuracy (16–34 mm Mean Per-Joint Position Error—MPJPE), while monocular systems provide sufficient sensitivity (82–88%) for longitudinal monitoring of geriatric fall risk and stroke recovery. While challenges persist in achieving precise axial rotation measurement, integrating real-time signal refinement enables objective and ecologically valid assessments in community-based healthcare settings. This technological advancement redefines movement analysis, shifting it from a laboratory-bound procedure to a widely accessible and interoperable diagnostic tool.

The rapid evolution of deep learning and computer vision has transformed markerless motion capture (MoCap) from specialized biomechanical tools into a key technology for real-time interactive systems, video game engines, and virtual production. This review analyzes state-of-the-art markerless pipelines and focuses on integrating 3D human pose estimation (HPE) into standard graphics engines. We evaluate the shift from single-view regression to temporal convolutional architectures. Our review emphasizes the trade-offs between spatial precision and algorithmic overhead for interactive use cases. Central to our analysis is the spatio-temporal refinement of noisy skeletal data. We assess adaptive smoothing techniques, such as the 1-Euro filter and predictive neural filtering, designed to minimize jitter and motion-to-photon latency. Findings indicate that modern markerless systems approach the standards of marker-based tools. However, maintaining temporal consistency under strict real-time budgets remains a fundamental challenge for human-display feedback loops. We examine industry-standard engines and their role in permitting seamless character retargeting throughout diverse applications, including clinical monitoring and sports coaching. Ultimately, this work synthesizes methodologies to optimize the responsiveness of AI-based MoCap systems within contemporary real-time software architectures.

The rapid evolution of deep learning and computer vision has moved markerless mo-tion capture (MoCap) from specialized biomechanical tools into a key technology for real-time interactive systems, video game engines, and virtual production. This review analyzes state-of-the-art markerless pipelines and focuses on integrating 3D human pose estimation (HPE) into standard graphics engines. We evaluate the shift from sin-gle-view regression to temporal convolutional architectures. Our review emphasizes the trade-offs between spatial precision and algorithmic overhead for interactive use cases. Central to our analysis is the spatio-temporal refinement of noisy skeletal data. We assess adaptive smoothing techniques, such as the 1-Euro filter and predictive neural filtering, designed to minimize jitter and motion-to-photon latency. Findings indicate that modern markerless systems approach the standards of marker-based tools. However, maintaining temporal consistency under strict real-time budgets re-mains a fundamental challenge for human-display feedback loops. We examine indus-try-standard engines and their role in permitting seamless character retargeting throughout diverse applications, including clinical monitoring and sports coaching. Ultimately, this work synthesizes methodologies to optimize the responsiveness of AI-based MoCap systems within contemporary real-time software architectures.