Jul 2026· Big Data and Cognitive Computing· Vol 10, pp. 227· 0 citations· 42 references
TL;DR
This paper proposes a data-driven FFH detection method that integrates multiple complementary features into a unified score-based model, achieving a practical balance between detection sensitivity, false-positive suppression, computational efficiency, and real-time feasibility.
Abstract
Fall-from-height (FFH) detection is a critical component in wearable safety systems, particularly in environments where high-intensity movements can lead to frequent false positives. Conventional approaches based on simple thresholding of acceleration signals often fail to reliably distinguish FFH events from non-fall activities due to overlapping signal characteristics. This paper proposes a data-driven FFH detection method that integrates multiple complementary features into a unified score-based model. The proposed approach first performs structured peak detection to extract candidate impact events while significantly reducing the number of samples requiring further processing. Each candidate is then evaluated using pre-peak structure, post-impact stability, and pressure variation, which respectively capture structural, temporal, and physical characteristics of FFH events. Based on statistical analysis, feature-wise score contributions are designed to reflect their discriminative strength, and the final FFH decision is performed using an additive scoring mechanism. This formulation enables flexible handling of ambiguous cases while preserving strong FFH characteristics. Experimental results demonstrate that the proposed method maintains 100% recall at the selected decision threshold while significantly reducing false positives from non-FFH activities. In addition, the peak detection stage reduces more than 99% of raw samples, enabling efficient on-device processing suitable for wearable systems. The proposed method also includes quantitative analysis of latency characteristics. Although FFH inference latency is influenced by asynchronous pressure sensing, the delay remains bounded and predictable, and most detections are completed within a practical time range for real-time wearable safety applications. Overall, the proposed method achieves a practical balance between detection sensitivity, false-positive suppression, computational efficiency, and real-time feasibility, demonstrating its applicability to wearable safety systems.
Falling detection is vital for elderly care and intelligent surveillance; however, prevailing vision-based approaches predominantly frame it as static pose classification or discrete temporal pattern matching, fundamentally overlooking the instability dynamics of the human support system. This paper proposes a physics-informed falling detection framework that recasts falling as a stability-loss event in a coupled dynamical system. We introduce a novel dual-LTC architecture comprising a Center-of-Mass (CoM) subsystem and a Base-of-Support (BoS) subsystem, both instantiated as Liquid Time-Constant (LTC) neural networks to continuously model inertial trajectory evolution and ground-contact adjustment through adaptive time constants, Physical interpretability of falling motion. A learnable coupling module emulates physical interaction between the two subsystems, while a Stability Manifold classifier operates in the joint latent space to detect boundary crossing via Lyapunov-inspired stability metrics. Complementary counterfactual trajectory projection and Time-to-Collision (TTC) estimation further enable irreversibility assessment and early warning. The architecture is designed to support a three-state prediction paradigm (Normal, Falling, Fallen); in this preliminary study, we validate the core stability discrimination capability on a two-class dataset (Normal vs. Falling), leaving the full three-state temporal transition to future work. Unlike conventional CNN--RNN pipelines, the proposed formulation encodes continuous-time mechanical inertia, yielding a sub-50K-parameter network capable of real-time inference on resource-constrained edge devices. Extensive experiments demonstrate competitive accuracy with superior physical interpretability, validating its efficacy for low-compute visual fall detection.
Wenjun Xia, Zhicheng Peng, Haopeng Li et al.· 0 citations
Accurate real-time estimation of human activity intensity is essential for diverse applications such as health monitoring, ergonomics, sports science, and adaptive building management. However, existing methods often depend on intrusive wearable sensors, discrete activity classifications, or extensive training datasets, which compromise their practicality and generalizability. To address these gaps, we propose a novel activity intensity score (AIS) framework that provides a nonintrusive and continuous measure of activity intensity by analyzing video data. The proposed method applies pose estimation to video data to extract body landmarks, which are then used to compute kinematic parameters including the angular velocity, angular acceleration, range of motion, peak speed, movement frequency, and rotational energy across defined kinematic chains (e.g., arms, legs, torso). These parameters are then normalized and combined through an optimized weighted summation to produce a continuous activity intensity metric. Experimental validation was conducted with 20 participants performing various activities from low, moderate, and high intensity. Results demonstrated strong correlations between AIS scores and both activity intensity levels (Spearman’s
ρ
=
0.943
,
p
<
0.001
) and participants perceived exertion ratings (Pearson’s
r
=
0.923
,
p
<
0.001
). Statistical comparisons demonstrated that the AIS values effectively discriminate among these three intensity categories (Spearman’s
ρ
=
0.943
) and significant group differences confirmed by ANOVA (
p
<
0.001
). Moreover, the AIS exhibited a strong correlation (Pearson’s
r
=
0.923
) with self-reported exertion (Borg rating of perceived exertion), indicating consistency with participants’ subjective perceptions. This occupant-invariant and domain-independent method enables robust, real-time measurement of movement intensity for applications ranging from healthcare and workplace ergonomics to sports analytics and adaptive HVAC control.
Moein Younesi Heravi, Inbae Jeong, Youjin Jang· Journal of computing in civi...· 0 citations
The findings show that CNN-based architectures dominate algorithm choice, edge devices dominate deployment platforms, and optimization remains central to real-time inference on constrained hardware.
Mahammad Nabizade, Réda Yahiaoui, Isabelle Lajoie et al.· Italian National Conference...· 0 citations
Falls among older adults can lead to severe injury when assistance is delayed. Vision-based monitoring offers a contactless means of detecting falls, but deploying such systems in indoor environments raises important privacy and implementation challenges. This study presents a privacy-aware fall detection framework that combines pose-derived biomechanical features, wearable inertial sensing, and deployment-aware computation within an edge–fog–cloud architecture. The proposed approach is evaluated on three public datasets (UR Fall, UP-Fall, and Le2i). Static biomechanical features achieve an F1-score of 99.95% on UR Fall and above 91% on Le2i, whereas performance on the more challenging UP-Fall dataset reaches approximately 80% F1-score. The effect of temporal aggregation is found to depend on dataset characteristics, with larger temporal windows improving performance in some cases while providing little benefit in others. Cross-dataset evaluation reveals substantial performance degradation under dataset shift, indicating limited generalization across different recording conditions and participant populations. To investigate this behaviour, feature distributions are analysed using Jensen–Shannon divergence and Wasserstein distance, and robustness-oriented feature selection improves transfer performance compared with using the complete temporal feature set. Multimodal fusion of pose-based predictions with inertial sensing further improves recall and F1-score on the UP-Fall dataset. End-to-end deployment profiling on a Raspberry Pi 5 demonstrates that the complete inference pipeline, including pose estimation, biomechanical feature extraction, temporal feature aggregation, and classification, operates in real time on resource-constrained hardware. Deployment analysis using iFogSim further shows that privacy-aware placement strategies substantially reduce network communication while providing favourable trade-offs among latency, privacy, and resource utilization compared with cloud-centric execution. Overall, the study demonstrates that combining robust biomechanical features with multimodal sensing and deployment-aware resource management provides an effective foundation for practical privacy-preserving fall detection.
Sona Mundody, R. R. Guddeti· IEEE Access· 0 citations
Accidental falls among the elderly demand highly reliable detection systems; however, existing solutions based on wearables, cameras, or WiFi often suffer from environmental interference, privacy concerns, and poor performance in detecting slow-onset falls (e.g., fainting). Despite these challenges, millimeter-wave (mmWave) radar-based methods have emerged as an attractive alternative. In this article, we propose an effective and efficient fall detection method, mmFallBbox, which leverages mmWave radar to track 3-D human bounding box dynamics. By exploiting the correlation between fall states and bounding box evolution, our approach effectively distinguishes between normal activities and slow fall states, such as fainting or medical conditions, which traditional systems struggle to detect. To evaluate the performance of mmFallBbox, we collected a large-scale fall detection dataset consisting of 60 h of radar data synchronized with video annotations. This dataset will be made publicly available to the research community for further development. We achieved an $F1$ score of 0.977 on our dataset and achieved state-of-the-art performance with limited computational complexity. Moreover, extensive experimental results show significant improvements in detecting slow-onset falls and providing explainable outputs, offering a promising solution for real-world fall detection applications.
Wenxuan Li, Dongheng Zhang, Jianwen Tong et al.· IEEE Transactions on Radar S...· 0 citations
Photoplethysmography (PPG) is an effective noninvasive approach to detect drowsiness and cognitive impairment. Given its compatibility with wearable and edge devices, this method is increasingly adopted in environments requiring continuous monitoring, such as driving and high-intensity workloads. However, existing studies lack a systematic evaluation of individual feature domain contributions and do not adequately address the feasibility of deployment on resource-constrained hardware. This study proposes a multi-domain feature extraction pipeline comprising four domains (time, frequency, wavelet, and nonlinear), coupled with a Random Forest classifier for drowsiness detection. A comprehensive ablation study on 15 feature combinations is conducted to analyze the trade-off between classification performance and computational cost. Experimental results indicate that the 3-domain combination (time, wavelet, and nonlinear) achieves the highest F1-score (0.601), while the full 4-domain configuration yields a higher recall rate (0.775), making it more suitable for safety-critical applications. In particular, computational cost analysis indicates that the feature extraction phase constitutes the primary computational bottleneck of the system. Specifically, extracting nonlinear features alone consumes 245.2 ms, approximately 500 times more than the model’s inference time (0.5 ms with a memory footprint of 0.3 MB). These findings highlight the critical role of feature domain selection in practical deployment. Consequently, this study provides an empirical foundation for designing reliable and resource-optimized drowsiness detection systems on edge hardware platforms.
Vu Trinh Tu, Thang Manh Hoang, Huy Duy Nguyen· IEEE International Conferenc...· 0 citations