motion capture analysis

Conventional optical motion capture (MoCap) and inertial measurement unit (IMU)-based systems suffer from marker dependency, hardware complexity, and requirement for specialized personnel and controlled environments—limiting their clinical utility for low-speed movement analysis. Method: This study proposes an end-to-end, video-driven kinematic parameter extraction pipeline leveraging deep learning–based 2D/3D pose estimation and SMPL mesh recovery to enable markerless visual motion capture. Contribution/Results: Validated in real-world clinical settings, the method achieves joint angle errors under 5°—meeting clinical tolerance thresholds—while reducing modeling and acquisition time by 90%. It eliminates the need for markers, dedicated facilities, or expert operators. To our knowledge, this is the first demonstration of markerless motion capture achieving a balanced trade-off among accuracy, portability, and usability in authentic clinical practice, thereby establishing a clinically deployable framework for bedside motor function assessment.

This study addresses the challenge of performing accurate, lab-free 3D human kinematic assessment during Activities of Daily Living (ADL) for telemedicine, sports science, and rehabilitation. We systematically benchmark monocular video-based and IMU-based approaches using state-of-the-art models—including MotionAGFormer, MotionBERT, MMPose (2D-to-3D), and NVIDIA BodyTrack—and unify evaluation via OpenSim inverse dynamics and Human3.6M joint-angle metrics. Results demonstrate that MotionAGFormer achieves the highest accuracy (RMSE = 9.27° ± 4.80°, MAE = 7.86° ± 4.18°, correlation r = 0.86 ± 0.15, R² = 0.67 ± 0.28), confirming the clinical feasibility of monocular video in real-world settings. We introduce a new benchmark for in-the-wild human motion capture, explicitly characterizing the trade-offs among accuracy, cost, and deployment practicality between video and IMU modalities. This work provides empirical validation and methodological guidance for scalable, low-cost remote motion monitoring.

To address the challenges of real-time quantification of workers’ limb motion and low accuracy in micro-task-level fatigue detection in industrial settings, this paper proposes a video-based analytical method grounded in multivariate statistical process control (SPC). First, joint keypoints are extracted from video sequences using HRNet or OpenPose, followed by construction of multidimensional motion feature vectors. Innovatively, Hotelling’s T² control chart is introduced to dynamically model and monitor motion intensity, triggering fatigue alerts when the T² statistic exceeds a predefined threshold. This work represents the first application of multivariate SPC charts to human motion quantification, significantly enhancing discrimination of subtle motion differences at the micro-task level—achieving approximately 35% higher correlation than conventional macro-task analysis. Validation in real industrial environments demonstrates millisecond-level system responsiveness and high-precision motion quantification, enabling fine-grained human factors ergonomics assessment and proactive intervention.

High-quality, multimodal motion data are critically needed for physical therapy and gait analysis, yet existing datasets suffer from high acquisition costs and poor generalizability. To address this, we introduce the first open-source, multimodal dataset specifically designed for rehabilitation assessment. It comprises synchronized inertial measurement unit (IMU) data (9 channels) and optical motion capture data (68 markers) from 19 participants performing 12 standardized rehabilitation and gait tasks. Our key contributions include: (i) millisecond-level temporal synchronization between IMU and optical data; (ii) IMU orientation calibration within a standardized anatomical coordinate system; (iii) subject-specific OpenSim model–driven inverse kinematics outputs; and (iv) comprehensive temporal annotations and clinical expert ratings for all movements. We publicly release preprocessing code, validation tools, and an interactive visualization platform. This resource significantly enhances reproducibility and generalizability in movement quality assessment, temporal segmentation, and biomechanical modeling.

Clinical movement analysis lacks high-quality, markerless biomechanical datasets and general-purpose models tailored for rehabilitation medicine. Method: We introduce BioMotionLM, the first multimodal foundation model for rehabilitation—(1) constructing a 30+ hour, cross-population biomechanical trajectory dataset (including diverse motor disorders) with trajectory tokenization; (2) designing a multimodal Transformer architecture that enables end-to-end alignment between biomechanical trajectories and clinical semantic queries; and (3) releasing a large-scale clinical movement question-answering dataset and performing instruction tuning. Contribution/Results: BioMotionLM significantly outperforms unimodal baselines across five clinically relevant tasks—activity recognition, motor disorder detection, diagnostic inference, clinical scale scoring, and gait quantification—demonstrating strong generalizability, interpretability, and clinical readiness. It establishes a unified foundation model framework for rehabilitation movement analysis.

Chronic pain often leads to diminished functional capacity, yet objective and convenient methods for quantifying patients’ motor function in real-world settings remain scarce. This study proposes a computer vision approach leveraging monocular smartphone videos and deep learning–driven 3D human pose estimation to remotely extract clinically relevant kinematic biomarkers in home environments without specialized equipment. Through systematic bias correction and individualized leave-one-subject-out calibration, the method demonstrates high agreement with gold-standard optical motion capture in laboratory validation (r > 0.85) and exhibits excellent test–retest reliability (r > 0.86) as well as significant sensitivity to group differences in patients with fibromyalgia and sciatica. This work represents the first demonstration of high-precision, scalable 3D motor function assessment in unsupervised home settings.

This work addresses the challenge of uncovering intrinsic patterns underlying behavioral coordination and dynamic change in high-dimensional, noisy, and temporally complex human pose data. To this end, the authors propose a general-purpose analytical framework that integrates systematic preprocessing, flexible dimensionality reduction—accommodating both linear and nonlinear methods—and temporal recurrence analysis. The framework is designed to handle diverse pose data modalities, including facial or full-body, 2D or 3D, and single- or multi-person configurations. It enables unified modeling of pose dynamics across varied experimental contexts and demonstrates strong effectiveness and generalizability in extracting theoretically interpretable movement patterns, as validated through three empirical case studies.

This study addresses the clinical need for a convenient and low-cost method to quantify joint angles. The authors propose a novel approach that directly maps segment rotation matrices—output by arbitrary parametric human body models such as GEM-X or SAM 3D Body—to clinically relevant joint angles, without requiring inverse kinematics, musculoskeletal modeling, subject-specific calibration, height measurements, camera parameters, or individualized modeling. Relying solely on a compact calibration table, the method is computationally efficient, enables real-time processing of monocular video, and demonstrates cross-model generalizability. Evaluated on the OpenCap LabValidation dataset, it achieves a mean absolute error of 4.50 degrees, matching the performance of OpenCap Monocular.

Current AI-generated videos lack systematic evaluation regarding the realism and biomechanical plausibility of human motion. To address this gap, this work proposes HumanScore, a novel multidimensional evaluation framework specifically designed for human actions, introducing six interpretable metrics that assess kinematic plausibility, temporal stability, and biomechanical consistency, among other dimensions. Leveraging a carefully curated set of motion prompts and integrating kinematic modeling, temporal analysis, and biomechanical constraints, the framework quantitatively evaluates 13 state-of-the-art video generation models. The results not only yield a robust ranking of models based on motion quality but also uncover a significant discrepancy between visual fidelity and physical plausibility, while identifying common failure modes such as temporal jittering and anatomically implausible poses.

Existing markerless hand motion capture methods suffer from low accuracy under complex dexterous motions, are highly susceptible to occlusions, and often fail to satisfy biomechanical constraints. To address these limitations, this work proposes an end-to-end, multi-view markerless hand motion capture approach that integrates a differentiable biomechanical hand model into a gradient-based optimization pipeline, jointly refining pose and shape parameters directly from multi-view video inputs. Evaluated on an 8-camera system, the method demonstrates strong robustness to occlusions and intricate hand gestures, achieving a 100% reconstruction success rate across 121 test sequences—significantly outperforming conventional two-stage methods, which attain only 85%. The reconstructed hand motions exhibit superior physiological plausibility, adhering closely to realistic biomechanical behavior.