Gait is a behavioral biometric trait that enables non-invasive person recognition based on individual walking patterns. Camera-based gait acquisition is convenient, but silhouette sequences often contain substantial motion-irrelevant appearance information, such as body shape, clothing, and carried objects. To address this problem, a multi-scale time series differencer is proposed to acquire tensor difference data between adjacent frames, so as to extract dynamic feature information in motion gait image sequences. Experiments on the CASIA-B dataset show that the proposed method achieves Rank-1 accuracies of 97.7%, 94.6%, and 80.0% under NM, BG, and CL conditions, respectively. Ablation results further demonstrate that MTDU improves the mean accuracy from 84.7% to 90.8% compared with single-scale temporal differencing. The multi-scale time series differencer shows potential for fields including sports motion gait detection and recognition, surveillance security motion gait identity authentication, and medical motion gait recovery assessment for sports injuries, demonstrating practical application value.

Physiological indicators are increasingly employed in sports training. However, studies on surface electromyography (sEMG) primarily focus on the analysis of isometric contraction. Research on sEMG related to isotonic contraction, which is more relevant to athletic performance, remains relatively limited. This paper examines the changes in the isotonic contraction performance of the male upper arm muscles resulting from long-term basketball training using the sEMG metrics. We recruited basketball physical education (B-PE) and non-PE majors to conduct a controlled isotonic contraction experiment to collect and analyze sEMG signals. The sample entropy event detection method was utilized to extract the epochs of active segments of data. Subsequently, statistical analysis methods were applied to extract the key sEMG time domain (TD) and frequency domain (FD) features of isotonic contraction that can differentiate between professional and amateur athletes. Machine learning methods were employed to perform ten-fold cross-validation and repeated experiments to verify the effectiveness of the features across the different groups. This paper investigates the key features and channels of interest for categorizing male participants from non-PE and B-PE backgrounds. The experimental results show that the F12B feature group consistently achieved an accuracy of between 80% and 90% with the SVM2 model, balancing both accuracy and efficiency, which can serve as evaluation indices for isotonic contraction performance of upper limb muscles during basketball training. This has practical significance for monitoring isotonic sEMG features in sports and training, as well as for providing individualized training regimens.

This paper investigates the adaptive control problem for a class of Euler-Lagrangian (EL) systems with uncertain parameters and external disturbances. While the EL system is one of the most classical system models, the dynamics of exoskeleton motion rehabilitation robots (EMRR) can be modeled as an EL system by the Lagrangian approach. First, in the absence of uncertainties, it is shown that an effective controller can be designed easily such that the system state tracks the reference signal. Second, for the case where uncertainties exist, a neural network (NN) is introduced to approximate the uncertain terms, and a novel adaptive controller design strategy is proposed to improve the resilience against disturbance by taking H _∞ performance into account. Moreover, based on the constructed adaptive controller, an effective stability analysis technique is also proposed for the cases where intermittent measurements are collected. Finally, an EMRR system example is provided to demonstrate the effectiveness of the proposed method.

Introduction: Drawing on the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) and technological affordances theory, this study constructs a theoretical model to investigate key factors influencing adolescents' digital sports participation and its underlying mechanisms. With virtual-real integration as the core mediating variable, we examine how five antecedents—performance expectancy, effort expectancy, social influence, facilitating conditions, and hedonic motivation—affect active and deep sports participation through this mediator. Methods: Data were collected from 417 adolescent participants using validated scales. Structural equation modeling was employed to test the proposed relationships, including direct and mediating effects. Results: The results confirm that virtual-real integration serves as a critical behavioral mechanism linking technology affordances to digital sports participation. The UTAUT2 core variables demonstrate universal explanatory power, though their influence pathways diverge. Specifically, performance expectancy's effect on deep participation is fully mediated by virtual-real integration; facilitating conditions has no direct effect on either participation behavior, with its influence entirely mediated through virtual-real integration. The model explains 50.7%, 51.5%, and 43.1% of the variance in virtual-real integration, active participation, and deep participation, respectively, indicating strong explanatory power. Discussion: By introducing virtual-real integration as a behavioral construct, this study extends the theoretical boundaries of UTAUT2, providing explanations and empirical evidence for the transformation process from technology acceptance to deep sports participation among adolescents. The findings offer insights for designers, educators, and policymakers: prioritize features that foster virtual-real integration, leverage community influence to establish blended online-offline sports communities, and improve accessibility conditions to support integrated behaviors, thereby effectively promoting technology-integrated sports activities among adolescents.

Postural control in curling is expressed during a dynamic, sport-specific delivery on a single sliding leg, yet it is still assessed almost exclusively with static single-leg force-plate protocols that poorly reflect on-ice demands. This study developed and validated a wearable multi-sensor system that fuses trunk and pelvis inertial measurement units (IMUs), instrumented plantar-pressure insoles and core-muscle surface electromyography (sEMG) to quantify dynamic postural stability during the simulated curling delivery in 110 young adult curlers (age range 18–24 years). A four-phase segmentation of the delivery was implemented, and a compact 28-feature set was extracted and combined through a complementary Kalman fusion scheme into a single Dynamic Stability Index (DSI). Validated against a reference force plate during static single-leg standing, the IMU-derived sway path showed a small bias (~0.4 cm) with 95% limits of agreement of approximately +4.4 to −3.6 cm, and between-day reliability of the fused DSI was excellent (ICC(3,1) = 0.90, 95% CI [0.82, 0.95]). A random-forest classifier discriminated high- from low-stability deliveries with 91.8% accuracy (95% CI [87.2%, 95.1%]; AUC = 0.92, 95% CI [0.87, 0.96]), outperforming any single-sensor or static force-plate baseline. Sensor-derived dynamic metrics correlated with external-oblique strength, trunk proprioceptive error and motor self-efficacy (Spearman ρ ranging from −0.55 to 0.51, all p < 0.001 after Holm–Bonferroni correction), and dynamic instability increased monotonically across descending self-efficacy tertiles. The proposed system provides a portable, context-relevant and interpretable tool for balance profiling in young curlers, and demonstrates that multi-sensor fusion substantially improves measurement fidelity over conventional single-modality assessment.