Search Results (7,582)

Reliable state identification is essential for intelligent flotation control because recovery, concentrate grade, entrainment, and mineral losses are only partially observable online. This systematic review examines field instrumentation, online analyzers, froth imaging, temporal synchronization, machine-vision methods, multimodal soft sensing, and the engineering requirements that determine whether a predictive model can operate as an industrial sensor. Scopus and Web of Science publications from 2021 to June 2026 were screened using a PRISMA-based protocol. The systematic evidence base includes 98 peer-reviewed technical studies published between 2021 and June 2026, and two PRISMA methodological publications are used to ensure the methodology for presenting the review. Additional methodological and contextual sources cited outside the systematic body of evidence are not included in the number of studies reflected in PRISMA. The evidence shows that machine vision is the most mature non-contact sensing approach, supporting bubble-size measurement, froth-velocity estimation, operating-state recognition, grade prediction, and visual monitoring. Current research is shifting from handcrafted descriptors toward convolutional, transformer, self-supervised, graph-based, temporal, and multimodal models. However, predictive accuracy alone does not demonstrate industrial readiness when camera geometry, illumination, contamination, delay compensation, temporal leakage, domain shift, uncertainty, inference latency, and SCADA/PLC integration are not evaluated. A five-dimensional Sensor Readiness Index is proposed to assess metrological validity, temporal integrity, validation rigor, operational robustness, and automation integration. The review defines the principal requirements for reliable industrial deployment of flotation sensing systems. Full article

Precise signal extraction and geophysical interpretation of Global Navigation Satellite System (GNSS) coordinate time series constitute the core foundation for establishing the International Terrestrial Reference Frame (ITRF) and inverting surface mass redistribution. This paper reviews major advances in this field across four dimensions: model framework, signal characteristics, extraction methods, and geophysical mechanisms. Key findings include: (1) Maximum Likelihood Estimation (MLE) has become the recognized standard for linear trend extraction: by jointly estimating deformation parameters and the noise covariance, it corrects the up to 5–10-fold underestimation of velocity uncertainty that arises in conventional least-squares analyses when colored noise is present but a white-noise covariance is assumed; (2) in CMONOC benchmark tests reported by Wu et al., Variational Mode Decomposition (VMD) achieves an average 69.8% residual RMS reduction at 97.9% of stations—results that are promising but not yet independently replicated on other networks—while the self-supervised model GNSS-FM (currently an unreviewed preprint) represents an emerging intelligent analysis paradigm; (3) the combined effect of atmospheric, non-tidal ocean and hydrological loading explains about 42% of the residual power of the annual vertical signal globally after pole tide correction and reduces the weighted mean vertical annual amplitude from 4.19 mm to 3.19 mm, while for horizontal components the fraction explained by current loading models is much smaller (amplitude ratio, explained variance and RMS/WRMS reduction are distinct metrics and are not directly interchangeable). This paper further highlights that the annual period was reported to fluctuate between 363 and 367 days at the ten CMONOC stations analyzed by Li et al.—a time variability that, if general, challenges fixed-frequency signal separation methods, although apparent period changes may also arise from amplitude/phase modulation, spectral leakage, finite-record effects, colored noise or data gaps—and it identifies thermoelastic deformation (TED) as a long-neglected but potentially quantifiable component, based on a recently released preprint dataset that has not yet undergone peer review. Finally, key research prospects are outlined, including physics-informed fusion methods, self-supervised foundation models, and a unified multi-source inversion framework. Full article

Background/Objectives: Wearable powered gait orthoses are prescribed as assistive equipment, on the premise that torque at an impaired joint improves walking while the device is worn; the device also adds mass. Where a hip module and a knee–ankle–foot module can be worn together, the combined configuration places the mass of both on one limb, and whether its walking cost exceeds that of either module alone has not been tested under a common protocol. Methods: We analysed individual participant data from two previously published prospective studies, neither of which compared configurations. Forty-three adults with central neurological gait impairment, enrolled during 2024, were trained over six supervised sessions with a bilateral hip orthosis (HO, n = 13), a unilateral knee–ankle–foot orthosis (KAFO, n = 15), or both worn together (HKAFO, n = 15); the primary 6MWT analysis was performed on the complete-case subsample (n = 41, with HO n = 12, KAFO n = 14, HKAFO n = 15). The assisted–unassisted differential was computed per participant as (device-assisted − unassisted)/unassisted × 100; negative values indicate a shorter distance with the device. Results: No statistically significant baseline differences were detected across the three groups. The six-minute walk differential was larger for the HKAFO than for either single-region configuration at baseline (−24.9% versus −6.0% and −7.6%; Kruskal–Wallis p = 0.008) and after six sessions (−17.1% versus −5.9% and −4.3%; p = 0.012), whereas the two single-region configurations were indistinguishable at both assessments (p = 0.397 and 0.916). Short-distance gait speed did not differ between configurations (p = 0.290 and 0.457). Conclusions: The walking cost of the combined configuration was associated with a larger assisted–unassisted differential than that of either constituent module. We propose that this reflects the mass carried on a single limb rather than a shortfall in assistance, a reading these data support but cannot establish in the absence of kinematic or metabolic measurements. These findings should be regarded as hypothesis-generating and require prospective replication in a larger, adequately powered cohort. Full article

Background/Objectives: Identifying accessible and effective exercise interventions is essential for improving health and physical fitness in individuals with overweight or obesity. Whole-body high-intensity interval training (WB-HIIT) is a time-efficient approach that requires minimal equipment and can be adapted to a wide range of settings and fitness levels. Therefore, this study aimed to investigate the effects of WB-HIIT on cardiorespiratory and musculoskeletal fitness, body composition, and metabolic health markers. Methods: Thirty-six healthy adults with overweight were assigned to either a WB-HIIT group (n = 19; 35 ± 9 years; BMI 30.2 ± 3.1 kg/m²) or a control group (n = 17; 33 ± 11 years; BMI 30.3 ± 3.3 kg/m²). Participants completed pre- and post-intervention assessments including dual-energy X-ray absorptiometry (DXA), blood sampling, maximal cardiopulmonary exercise testing, and measures of muscular strength (3 RM) and functional endurance (1 min sit-to-stand and arm curl). The 10-week supervised WB-HIIT (2–3 sessions per week) included a warm-up, 3–4 sets of 10–12 all-out whole-body exercises (30–40 s) interspersed with 20–30 s recovery, and a cool-down. Results: The control group showed no significant changes in any of the assessed parameters after 10 weeks, except for a modest increase in body fat mass. The WB-HIIT group showed significant improvements in cardiorespiratory fitness, evidenced by increases in VO₂peak (p < 0.05), maximal treadmill incline (p < 0.01), and first ventilatory threshold parameters (p < 0.05). Maximal strength increased significantly in both the leg and bench press (p < 0.001), accompanied by a significant improvement in upper-limb functional endurance (arm curl, p < 0.05). No significant changes were observed in body weight or BMI, but small reductions were found in body fat percentage, trunk fat mass, and waist circumference (p < 0.05). Except for a modest decrease in glycated hemoglobin (p < 0.05), metabolic blood parameters remained unchanged and were already within the normal range at baseline. Conclusions: A 10-week WB-HIIT intervention improved cardiorespiratory fitness, muscular strength, and upper-limb functional endurance, and elicited a modest reduction in trunk fat mass in individuals with overweight or obesity. Full article

Infrared and visible images provide complementary cues for maritime ship detection, but practical dual-sensor systems often produce image pairs that are not strictly registered. Directly fusing such unregistered pairs may introduce ghosting artifacts, blurred target boundaries, and feature conflicts, especially over weak-texture sea surfaces where reliable correspondence cues are sparse. To address this problem, we propose UA-FusionDet, an unregistered-aware infrared-visible fusion framework with cross-modal feature alignment for maritime ship detection. The proposed framework extracts visible and infrared features with a dual-branch encoder, aligns the visible feature to the infrared reference through cross-modal deformable feature alignment, and suppresses unstable background offsets using sea-surface saliency guidance. Wavelet-guided complementary fusion then decomposes the aligned features into low- and high-frequency sub-bands, enabling frequency-aware fusion of infrared thermal saliency and visible structural details before feeding the fused representation to both a lightweight reconstruction decoder and a ship detection head. The reconstruction decoder provides auxiliary image-level regularization, while the detection branch supervises the task-oriented fused representation with ship bounding-box annotations. UA-FusionDet does not require registration ground truth or fused-image ground truth during training, making it suitable for realistic maritime monitoring scenarios with imperfectly aligned visible and infrared sensors. Experiments on 3132 unregistered visible–LWIR maritime image pairs show that UA-FusionDet achieves a precision of 0.904, a recall of 0.866, an ${\mathrm{mAP}}_{50}$ of 0.912, and an ${\mathrm{mAP}}_{50–95}$ of 0.566, exceeding the strongest competing method by 2.5 and 2.4 percentage points on the two mAP metrics, respectively, while maintaining an inference speed of 52.6 FPS. These results demonstrate that the proposed alignment and fusion framework improves detection accuracy under cross-modal misregistration while retaining practical inference efficiency. Full article

Background/Objectives: Several studies have reported a high prevalence of overweight and obesity among children and adults, with relevant implications for long-term health. Digital platforms and social media may support nutrition education when their content is developed and supervised by health professionals. This study evaluated whether dietary habits, lifestyle indicators, body mass index-related measures, and Mediterranean diet adherence differed between baseline and 210-day assessments among participants in a family- and school-mediated multimedia educational program aimed at promoting healthier lifestyles and greater adherence to the Mediterranean diet. Methods: This was an uncontrolled repeated cross-sectional before-and-after evaluation based on online surveys administered at baseline and after 210 days to respondent samples from the same project population. The multimedia educational program was delivered through a dedicated website and social media channels. The analysis included 329 children/adolescents at both assessment waves and 119 and 122 adult female respondents at baseline and 210 days, respectively. Since only 11 adult males responded, analyses of adults were restricted to women. Additionally, Mediterranean diet adherence was assessed using a modified KIDMED-derived score in children/adolescents and the PREDIMED score in adults. Data from the two assessment waves were compared at the group level, rather than as paired longitudinal observations. Results: Among children/adolescents, higher physical activity, lower screen time, and more favorable dietary indicators were observed at 210 days than at baseline. Moreover, daily fruit intake, daily vegetable intake, regular fish consumption, and the modified KIDMED score were higher at 210 days. Good adherence to the Mediterranean diet increased from 46.8% at baseline to 61.4% at 210 days. Among adult female respondents, the median body mass index was lower at 210 days than at baseline, and favorable differences were observed for vegetable, fish/seafood, legume, and nut intake. The median PREDIMED score increased from 7.0 [6.0–8.0] at baseline to 9.0 [7.0–10.0] at 210 days, and good Mediterranean diet adherence increased from 10.9% to 35.2%. Conclusions: Among children/adolescents, physical activity, screen time, and dietary indicators showed more favorable distributions in the 210-day respondent group than in the baseline respondent group. Daily fruit intake, daily vegetable intake, regular fish consumption, and the modified KIDMED-derived score were higher at 210 days, and good Mediterranean diet adherence was more frequent at 210 days than at baseline (61.4% vs. 46.8%). Among adult female respondents, the median body mass index was lower in the 210-day respondent group than in the baseline respondent group, and more favorable distributions were observed for vegetable, fish/seafood, legume, and nut intake. The median PREDIMED score was higher at 210 days than at baseline: 9.0 [7.0–10.0] versus 7.0 [6.0–8.0], respectively. Good Mediterranean diet adherence was also more frequent at 210 days (35.2% vs. 10.9%). Full article

In unmanned aerial vehicle inspection of transmission lines, insulators often exhibit arbitrary orientations and elongated shapes and are frequently embedded in complex backgrounds. Horizontal bounding boxes tend to include substantial redundant regions. Meanwhile, existing oriented object detection methods still suffer from angular discontinuities at periodic boundaries, insufficient geometric supervision for rotated bounding boxes, and difficulties in lightweight deployment. To address these issues, this paper proposes a lightweight oriented object detection model, termed R-YOLOv8-PSGH, which integrates dual-frequency phase-shift encoding and Gaussian geometric supervision. Based on a lightweight R-YOLOv8 architecture, a rotated detection head is developed to decouple the predictions of object categories, bounding-box locations, and orientation angles. To improve the periodic continuity of angle representations and strengthen the geometric constraints on rotated bounding boxes, a dual-frequency phase-shift angle encoding strategy and a Gaussian geometric localization loss are designed. Specifically, the complementary relationship between periodic signals with periods of 180°and 90° is exploited to map orientation angles into continuous phase responses, thereby improving the stability of orientation prediction. Moreover, the spatial structure of each rotated bounding box is modeled as a two-dimensional Gaussian distribution, and overlap consistency, center distance, and shape discrepancy are jointly optimized. In this manner, the orientation representation and bounding-box-level geometric supervision are collaboratively enhanced. Experimental results demonstrate that the proposed method improves the detection accuracy and localization stability of rotated objects while maintaining favorable lightweight deployment capability, providing a new solution for lightweight object detection in complex scenarios. Full article

Earthquakes remain among the most destructive natural hazards, necessitating reliable monitoring and early warning systems for effective risk mitigation. Recent advances in machine learning (ML) and deep learning (DL) have significantly improved seismic signal analysis, enabling more accurate event detection, phase picking, classification, and magnitude estimation. This study presents a systematic review of ML- and DL-based approaches for earthquake monitoring, with particular emphasis on Distributed Acoustic Sensing (DAS) as an emerging technology for high-resolution, real-time seismic observation. Following the PRISMA 2020 guidelines, a systematic literature search was conducted across Scopus, Web of Science, IEEE Xplore, and Google Scholar, yielding 252,223 initial records. After applying the predefined publication period, removing duplicate records, conducting relevance screening, and performing eligibility assessment, 138 peer-reviewed studies published between 2021 and 2025 were retained for detailed analysis and synthesis. The review reveals a significant transition from conventional signal-processing techniques to advanced artificial intelligence-based approaches, including Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) networks, Bidirectional Long Short-Term Memory (BiLSTM) networks, Transformer-based architectures, hybrid models, and Bayesian learning methods for uncertainty quantification. The findings further demonstrate that Distributed Acoustic Sensing (DAS) has emerged as a transformative sensing technology because of its dense spatial coverage, high spatial resolution, and continuous monitoring capability. However, several challenges remain, including the lack of standardized datasets, limited model generalization across diverse geological settings, insufficient model interpretability, high computational complexity, and the limited integration of uncertainty-aware approaches for real-time seismic monitoring. This review identifies these critical research gaps and highlights promising future research directions, including multimodal data fusion, interpretable artificial intelligence, physics-informed learning, self-supervised learning, and robust uncertainty quantification for next-generation intelligent seismic monitoring systems. Unlike previous review studies that primarily focus on individual machine learning techniques or conventional seismic monitoring, this review provides a comprehensive and systematic synthesis of recent advances in machine learning, deep learning, and Distributed Acoustic Sensing (DAS), identifies current research gaps, and offers practical recommendations to guide future research on intelligent earthquake monitoring systems. Full article

Accurate satellite component segmentation plays a fundamental role in numerous on-orbit perception tasks, including spacecraft pose estimation, autonomous robotic servicing, and space situational awareness. However, existing segmentation methods still face three major challenges: the scarcity of fine-grained annotations, the difficulty of preserving structural continuity under large pose variations, and the severe structural imbalance caused by extremely small and slender components such as antennas. To address these issues, this paper proposes LRSRNet, a low-resource spatial reconstruction network for satellite component segmentation. First, a LoRA-adapted DINOv3 foundation model is employed to efficiently transfer large-scale visual priors to the satellite domain, enabling robust structural representation under limited supervision. To exploit the complementary information encoded at different semantic levels, multi-level transformer features are uniformly aggregated for structural representation learning. Subsequently, a Converse2D-based spatial reconstruction decoder progressively restores the spatial continuity of satellite components, facilitating the recovery of fine structural details lost during hierarchical feature encoding. Furthermore, a structure-aware optimization strategy is introduced by jointly considering the geometric characteristics and category imbalance of satellite components during training, thereby improving the learning of geometrically fragile structures without sacrificing overall segmentation performance. Experimental results on a public satellite component segmentation benchmark demonstrate the effectiveness of the proposed method. LRSRNet achieves an mIoU of 77.10% on foreground categories and 82.45% over all categories. Ablation studies and visualization results further validate the contributions of the proposed representation adaptation, spatial reconstruction, and structure-aware optimization strategies. Full article

Semi-supervised fine-grained classification suffers from accumulating pseudo-label errors among potentially confusable classes when labeled data are scarce, yet negative-rule information that excludes classes at the attribute level remains unused. We propose AW-RS-ABL, a semi-supervised abductive learning method that fuses intra-class and inter-class attribute importance within rough sets and couples positive-rule confidence matching with negative-rule candidate-set reduction for pseudo-label correction. Samples, attributes, and labels form a decision information system $S=(U,A,F,d)$. Inter-class discriminative power is measured by positive-region dependency ${\gamma }_A\left(d\right)$ and intra-class stability by class-conditional entropy $H(a\mid d)$; the two are normalized and linearly fused into an attribute weight that selects a key attribute subset $A^{*}$. Weighted positive rules and high-confidence negative rules are built on $A^{*}$: negative rules shrink the candidate class set, and positive rules complete the correction via confidence matching. Rule consistency checking is embedded in the semi-supervised pseudo-label loop so that rule verification and perceptual model prediction co-evolve across training iterations. On six datasets spanning SAR target recognition, medical diagnosis, and agricultural classification at a reported 10% label ratio, AW-RS-ABL has the highest reported mean top-1 accuracy in each comparison, with displayed mean differences of 1.36–2.58 percentage points relative to FixMatch-Attr. Exact paired Wilcoxon tests with Holm correction support the differences for five datasets, whereas the difference for Dermatology is not significant. On the imbalanced Nursery test set, a representative seed yields 87.34% accuracy but 73.54% macro-F1 and 18.18% recall for the minority class. The reported formulation relies on conditional attributes, and its results remain conditional on their representation and binarization. Full article

Background/Objectives: Artificial intelligence tools have emerged as promising methodological support for systematic reviews and health technology assessment (HTA). Smart infusion pump interoperability represents a relevant case study due to its implications for medication safety, nursing workflow, and hospital quality improvement. The aim was to evaluate the performance of artificial intelligence as a methodological support tool across a systematic review, using the evidence synthesis on smart infusion pump–electronic health record interoperability as a case study. Methods: A systematic review following PRISMA 2020 guidelines was conducted. Searches were performed in MEDLINE, Embase, and Cochrane Library databases. AI-assisted tools (ChatGPT GPT-4 and Open Science Reviewer) were incorporated into data extraction, reporting appraisal based on STROBE criteria, and exploratory identification of methodological limitations under strict human supervision. Concordance between AI-assisted and manual extraction was evaluated descriptively. Results: Overall concordance between AI-assisted and manual data extraction was 82.5% (99/120) across assessed variables. Agreement was highest for structured variables, including study design (10/10; 100.0%), study identification variables (19/20; 95.0%), and participant characteristics (36/40; 90.0%). Agreement was lower for study content variables (18/30; 60.0%) and methodological appraisal (16/20; 80.0%). Among the 21 discrepancies, misclassification errors were most common (13/21; 61.9%), followed by omissions (3/21; 14.3%), incomplete data (3/21; 14.3%), and hallucinations (2/21; 9.5%). AI-assisted identification of methodological limitations showed substantial descriptive agreement with human assessments but demonstrated limited capacity for judgmental interpretations. Conclusions: Artificial intelligence demonstrated utility for structured review tasks such as data extraction and reporting appraisal, but showed limitations in tasks requiring interpretative and methodological judgement. Human oversight therefore remains essential throughout the review process. These findings derive from a single case study and should not be generalised beyond the evaluated context and AI tools. Full article

Interest changes complicate sequential recommendation when interaction histories are combined with item content. We evaluate MM-DRLSR on four public Amazon and Yelp benchmarks. The model integrates category-overlap drift supervision, history-derived drift representations, lightweight identifier–image–text fusion, candidate-conditioned scoring, and an offline actor–critic surrogate objective trained by logged-context replay. The observed next item is used only to construct training labels and rewards; inference ranks candidates from the observed history and candidate content. In the reported five-run summaries under a common leave-one-out protocol, MM-DRLSR attains the highest mean Recall and NDCG among the evaluated methods, with consistent advantages of 0.12–0.22 percentage points over the strongest contemporary multimodal baselines (relative gains of about 1.1–3.6%) that reach Holm-adjusted significance on three of four metrics against Harnessing MLLMs and on Recall@20 against DMESR. The practical value of the method lies in reaching, and in several comparisons, significantly exceeding, the accuracy of heavyweight multimodal-LLM-style approaches with a lightweight architecture whose inference overhead is only about 20% above IDURL. Ablation, sensitivity, and observed interest-shift summaries further describe the contributions of multimodal fusion and offline policy adaptation. The reported results indicate competitive public-data sequential ranking under the stated protocol, together with a reproducible and inference-safe evaluation design. Full article

Point cloud semantic segmentation is fundamental for 3D scene understanding and has been widely used in autonomous driving and infrastructure inspection applications. However, its performance is often limited by insufficient representation of local geometric structures and the high cost of point-wise annotation. To address these issues, this paper proposes GeoFuse-AL, a label-efficient segmentation framework that integrates a geometry-guided point–voxel network with multi-cue active learning. The base model, GeoFuseNet, builds on a hybrid point–voxel backbone and incorporates a Local Geometry Prototype Attention module to enhance object boundaries, fine-grained structures, and local geometric patterns. An Adaptive Channel Fusion module is further designed to improve feature interaction between point-level details and voxel-level context. To reduce annotation dependence, a Multi-Cue Diversity Active Sampling strategy combines prediction uncertainty, color-gradient variation, geometric curvature, and feature-space clustering to select informative and diverse samples. Experiments on S3DIS and SemanticKITTI demonstrate that the proposed model achieves mIoU scores of 63.3% and 61.7%, respectively, outperforming several representative methods. Under limited annotation settings, the proposed strategy reaches 99.2% of fully supervised performance with only 15% labeled data on S3DIS and 97.6% with only 5% labeled data on SemanticKITTI. These results demonstrate that GeoFuse-AL improves segmentation accuracy while substantially reducing annotation requirements. Full article

Remaining useful life (RUL) prediction for offshore wind turbines is important for predictive maintenance. However, its practical use is limited by the lack of reliable RUL labels, the small number of fault events, and the complex degradation links among turbine components. Most existing data-driven methods require labeled failure data and cannot fully capture how the degradation of different components is related. To address these limitations, a Dual-Pathway Degradation-Aware Network for label-free RUL prediction is proposed. The proposed method first constructs component-level health indicators (HIs) from SCADA data in a self-supervised manner and generates pseudo-RUL labels through a linear countdown strategy. It then jointly models the global degradation evolution and cross-component interactions using a global temporal branch and a fully connected graph branch, where residual aggregation is adopted to fuse node representations for RUL estimation. Experiments are conducted on the German North Sea Farm B dataset under a leave-one-out cross-validation (LOOCV) protocol. From the obtained results, the proposed method achieves an MAE of 111.3 h and an RMSE of 122.6 h, reducing the average MAE by approximately 19% compared with the strongest baseline. The proposed method achieves the best performance on three of the five turbines with fault events. Furthermore, the proposed method demonstrates superior capability in capturing component interaction patterns and provides an effective label-free solution for offshore wind turbine RUL prediction under limited fault data. Full article

Assessing visual fatigue is crucial in eye-controlled interaction. Traditional methods are either overly subjective or rely on highly invasive, costly equipment and complex procedures that require expert supervision. This study proposes a machine-learning-based approach for visual fatigue assessment. Data collection employs non-intrusive, easily monitored eye-tracking to capture ocular eye movement data and task performance data, while subjective questionnaires label fatigue states. For feature selection, participant-level Wilcoxon signed-rank tests with Benjamini–Hochberg FDR correction were used to identify fatigue-related indicators, and a redundancy-removal step based on Spearman correlation yielded a final set of six non-redundant features. For the assessment method, we introduced a gradient boosting decision tree–logistic regression (GBDT-LR) model whose hyperparameters are optimized via Bayesian optimization. All models were evaluated under a unified 5-fold stratified cross-validation framework with within-fold standardization and nested hyperparameter tuning. Results indicate that this model can effectively predict the state of visual fatigue. Compared with the performance of five other models—gradient boosting decision tree (GBDT), logistic regression (LR), support vector machine (SVM), random forest (RF), and RF-SVM—the proposed GBDT-LR model achieved an assessment accuracy of 89.79%, demonstrating strong predictive performance. This study provides an effective method for predicting visual fatigue in eye-controlled interaction, laying a research foundation for optimizing the user experience of eye-controlled interaction and promoting the sustainable development of eye-control technology. Full article