Search Results (83,362)

Monocular 3D multi-object tracking (3D MOT) remains challenging because it is hard to model how objects move over time and to keep correct identities without explicit depth information. In this context, we introduce D3SSTrack, a novel tracking-by-detection framework that integrates Mamba state-space modeling into the 3D tracking pipeline. At its core is the Solid State Track (SST) block, which extends the original Mamba block with dropout regularization and an additional projection layer to improve feature integration before temporal fusion. This design enables efficient modeling of long-range temporal dependencies while maintaining real-time performance at 38 FPS on a single GPU. The proposed tracker combines structured sequence modeling with effective temporal association, improving robustness against occlusions and abrupt motion changes. On the KITTI benchmark, D3SSTrack achieves the best sAMOTA (97.12%) and AMOTA (49.95%) among recent monocular 3D MOT methods, outperforming the best model S3MOT by 0.16% and 0.22%, respectively. Our results highlight the potential of state space-based architectures for real-world monocular 3D MOT applications. Full article

Background: Internal quality control (IQC) data offers continuous insight into analytical performance under routine conditions. This study evaluated IQC practices and long-term analytical imprecision (CV_a) across primary healthcare laboratories to derive analyte-specific reference change values (RCVs) for non-communicable disease (NCD) monitoring. Methods: A 22-month retrospective analysis of IQC data was conducted across 29 primary healthcare laboratories using 32 analytical units (Beckman Coulter AU480) in Malaysian primary healthcare. Six analytes were assessed: glucose, creatinine, total cholesterol, triglycerides, HDL cholesterol, and ALT. CV_a was estimated using median and 90th percentile (P90) coefficients of variation across two concentration levels. RCVs were calculated at 95% probability (Z = 1.96) by integrating observed CV_a with within-subject biological variation (CV_i) from EFLM databases. Results: IQC testing was highly standardized (median: 20 measurements/month). Long-term data showed stable, concentration-dependent imprecision. Median CV_a was lowest for glucose and lipids (1.7–1.9%) but higher for ALT (3.79%) and creatinine (3.52%) at Level 1. Derived RCV ranged from 14% (glucose) to 55.1% (triglycerides), with CV_i being the dominant contributor to RCV magnitude for most analytes. Conclusions: Long-term routine IQC data provide an analytically realistic foundation for deriving RCV in primary healthcare by reflecting real-world performance. Applying these RCV supports evidence-based interpretation of serial results, enhancing NCD monitoring by distinguishing true physiological change from analytical and biological noise. Full article

Background/Objectives: Sentinel lymph node biopsy is central to axillary staging in breast cancer, but conventional mapping often relies on radioisotopes and/or blue dye. Indocyanine green fluorescence has emerged as an alternative, although evidence for its use as a sole tracer in routine practice remains limited. This study evaluated the technical feasibility, lymph node yield, nodal metastasis detection, and short-term clinical outcomes of indocyanine green used as the only tracer for axillary staging in a consecutive single-centre cohort. Methods: This retrospective observational cohort study included 260 patients with histologically confirmed breast cancer who underwent axillary surgery at University Hospital Kaspela between 2024 and 2025 under an institutional protocol using indocyanine green as the sole tracer. Indocyanine green-guided mapping was attempted in all patients. For node-focused statistical analyses, a predefined complete-case–cohort of 230 patients was used. Descriptive analyses assessed axillary procedure distribution, lymph node yield, nodal metastasis, and postoperative outcomes. Exploratory multivariable logistic regression was performed to evaluate predictors of nodal metastasis. Results: Mapping was successful in 259/260 patients (99.6%). In the complete-case–cohort, sentinel lymph node biopsy was performed in 166/230 patients (72.2%), targeted axillary dissection in 4/230 (1.7%), and axillary lymph node dissection in 60/230 (26.1%). Median overall lymph node yield was 4 (IQR 3–7), but this pooled value reflected heterogeneous axillary procedures and should not be interpreted as sentinel node yield alone. In the clinically node-negative upfront SLNB subgroup, median lymph node yield was 4 (IQR 2.75–5), and nodal metastasis was identified in 22/112 patients (19.6%). Overall, nodal metastasis was identified in 58/230 patients (25.2%), with a median of 2 metastatic nodes (IQR 1–3) among nodal-positive cases. Reoperation for axillary lymph node dissection occurred in 14/230 patients (6.1%). In exploratory multivariable analysis, suspicious biopsied-positive nodes (OR 12.85, 95% CI 3.98–41.52), suspicious non-biopsied nodes (OR 15.58, 95% CI 3.44–70.59), and neoadjuvant therapy (OR 0.31, 95% CI 0.11–0.87) were associated with nodal metastasis; these findings should be interpreted cautiously given the expected clinical relationship between preoperative nodal suspicion and nodal positivity, and the limited number of nodal-positive events. Conclusions: Indocyanine green used as a sole tracer demonstrated high technical feasibility within a heterogeneous real-world axillary staging workflow in this single-centre cohort. These findings should be interpreted as implementation-focused feasibility data rather than formal diagnostic validation, given the retrospective design, heterogeneous case mix, and absence of an internal comparator. Full article

Traffic accidents remain a major challenge for intelligent transportation systems, requiring reliable and real-time detection under complex visual conditions. This study aims to investigate how backbone-level architectural modifications affect traffic accident detection performance in video-based scenarios. A dataset of 250 accident videos was curated from a public traffic surveillance source. This resulted in approximately 3000 manually annotated frames covering diverse accident conditions such as motion blur, occlusion, and illumination variation. To improve detection performance, we introduce Cross Stage Partial (CSP)-based feature partitioning and extend Efficient Layer Aggregation Network (ELAN) structures within the YOLOv9 backbone. Experimental evaluation demonstrates that the CSP-enhanced YOLOv9-t model achieves the best performance among all tested variants, improving mAP50 from 0.35 to 0.50 (approximately 42.8% relative improvement) compared to the baseline YOLOv9-t model, while maintaining real-time inference speed. The results further reveal that CSP improves localization precision, whereas ELAN enhances recall, highlighting complementary behaviors of backbone-level modifications in traffic accident detection tasks. These findings provide insights into how targeted architectural refinements can improve detection robustness in challenging real-world traffic scenarios. Full article

Background: Immunotherapy has become an integral part of systemic treatment for cervical cancer (CC). This study assessed the safety profile of cemiplimab and the association between immune-related adverse events (irAEs) and treatment outcomes in patients with persistent, recurrent or metastatic CC. Methods: This ambispective, multicenter, real-world cohort study included 101 patients treated in 13 reference oncology centers as part of the PCCIG-01 study. We evaluated the frequency and severity of irAEs and their association with progression-free survival (PFS) and overall survival (OS). Survival outcomes were analyzed using the Kaplan–Meier method and Cox proportional hazards models, with p < 0.05 considered statistically significant. Results: After a median follow-up of 7.5 months, adverse events occurred in 45 patients (44.6%) and were mostly grade (G) 1–2. IrAEs were observed in 34 patients (33.7%). Endocrine toxicities predominated (n = 24, 58.5% of irAEs), followed by hepatic (n = 5, 12.2%) and gastrointestinal events (n = 4, 9.8%). G3 irAEs occurred in 8 patients (7.9%). Median PFS was 3.9 months (95% CI 2.9–5.6) in patients without irAEs and 10.9 months (95% CI 5.7–16.3) in those with irAEs (p = 0.03). Median OS was 15.3 months (95% CI 8.6–25.9) in patients without irAEs and was not reached in those with irAEs (95% CI 11.6-NR; p = 0.11). The development of irAEs was associated with a 54% reduction in the risk of progression (HR 0.46, 95% CI 0.27–0.80), with no statistically significant impact on OS. Conclusions: In exploratory analyses, the occurrence of irAEs was associated with improved PFS in cemiplimab-treated patients with persistent, recurrent or metastatic CC. Cemiplimab showed a manageable safety profile, with most toxicities being G1–G2. Full article

Robotic waste sorting presents significant challenges, including object variability, cluttered environments, and the predominant reliance on deep learning and traditional computer vision techniques, which typically demand extensive datasets and task-specific training. This paper introduces a robotic waste sorting system that integrates the Gemini Vision–Language–Action (VLA) model with a KUKA LBR iiwa collaborative robot and an RGB-D camera. Our approach leverages the advanced reasoning capabilities of large, pre-trained VLA models to perform waste sorting, without requiring explicit training or dataset collection. Key contributions include the development of effective prompt engineering strategies for waste object identification, the assessment of the VLA’s performance in terms of inference time and accuracy, and the development of different grasping strategies for operation in cluttered scenarios. Our experimental tests demonstrated that the system’s inference time is between 2 and 4 s, which is suitable for collaborative robotic applications, and the system achieved a high overall classification accuracy of 89.64%. Crucially, we demonstrated that integration of RGB-D sensing enhanced the model’s ability to perceive object heights, resolve occlusions, and make informed grasping decisions in realistic, three-dimensional settings. We further validated multiple real-world grasping strategies, demonstrating tradeoffs between system efficiency and safety in heavily cluttered scenarios. This work establishes a practical and adaptable framework for deploying VLA-driven intelligence on commercial robotic platforms, highlighting the potential of VLAs for complex manipulation tasks beyond waste sorting. Full article

Recent advancements in Unmanned Aerial Vehicle (UAV) physics simulators, flight control firmware, and network virtualization have been substantial. However, operating these systems independently fails to capture the complex dynamics of real-world multi-UAV networks, thereby compromising simulation reliability. To address this, we propose the Multi-UAV Network-in-the-Loop Simulation (MUNILS) platform, which seamlessly integrates the Gazebo physics engine, the PX4 flight controller, and the ns-3 network simulator via Robot Operating System 2 (ROS2) middleware. Specifically, MUNILS leverages Micro eXtremely Resource Constrained Environments–Data Distribution Service (XRCE-DDS) for high-speed data bridging and employs Linux network namespaces to enforce traffic isolation and routing exclusively through ns-3. Crucially, we introduce a precise cross-layer time synchronization mechanism spanning the physical, control, and network domains to resolve inherent clock discrepancies among these heterogeneous simulators. Experimental evaluations confirm that MUNILS achieves strict traffic isolation, scalable closed-loop flight control, and highly accurate time synchronization across all integrated modules (Gazebo, ns-3, ROS2, and PX4) without cumulative clock drift, thereby providing a highly reliable verification environment for large-scale swarm operations on a single machine. Full article

Todos Santos Bay, Mexico, features several wave-focusing areas driven by its complex bathymetry, making it an ideal real-world test case for wave energy converter (WEC) park optimization. This study quantifies the influence of submarine cable costs and bathymetry-dependent mooring costs on the proposed park layout (hereafter the star-layout) and the levelized cost of energy (LCOE) of a 10-device WEC park, using a multi-state operational wave climatology of $N=179$ representative sea states from a 2008–2018 SNL-SWAN hindcast (covering $97.20\%$ of the annual time). A binary genetic algorithm combined with K-means clustering analysis was used to minimize LCOE under three cost scenarios: baseline, cable-only, and cable plus bathymetry-dependent mooring. Both infrastructure cost components contribute substantially: cable costs add $52.2\%$ to the baseline LCOE, and bathymetry-dependent mooring costs add a further $16.0\%$ at this site, with cable approximately three times more impactful. These quantitative magnitudes are conditioned on the moderate depth-gradient setting of Todos Santos Bay; the qualitative cost-component hierarchy is expected to generalize, but the relative weights will depend on the bathymetric and wave-climate characteristics of each candidate site. The mooring contribution is nontrivial both economically and spatially (the centroid of the park shifts by approximately 151 m between the cable-only and cable-plus-depth scenarios). K-means clustering identified 2–4 layout families per scenario (K $=4\to 3\to 2$ as cost components are added), indicating that infrastructure constraints reduce the viable solution space. These results support the central hypothesis of this work: WEC park optimization studies that adopt flat-bathymetry simplifications, the prevailing assumption in much of the prior literature, risk substantial underestimation of LCOE at sites with nontrivial depth variation. We recommend that bathymetry-dependent mooring costs be included alongside cable costs in any early-stage techno-economic assessment of WEC parks at sites with complex bathymetry. Full article

Background/Objectives: Non-celiac gluten sensitivity (NCGS) is a syndrome characterized by intestinal and extraintestinal symptoms triggered by gluten ingestion. Although anti-gliadin IgG antibody (AGA IgG) has been proposed as a potential biomarker for NCGS, its sensitivity and specificity in real-world clinical settings remain unclear. This study aimed to evaluate the clinical utility of AGA IgG in NCGS and to characterize its clinical features, including psychological distress and physical quality of life (QOL), in patients with clinically suspected NCGS attending a specialized outpatient unit in Japan, where patients reported symptoms related to the ingestion of gluten-containing grains (primarily wheat). Methods: We evaluated plasma AGA IgG levels in 45 patients with suspected NCGS based on clinical presentation and in 83 age- and sex-matched healthy controls. Plasma AGA IgG was measured using ELISA. Clinical symptoms and QOL were assessed using validated scales, including the 36-Item Short Form Health Survey (SF-36), Patient Health Questionnaire (PHQ-9 and PHQ-15), Generalized Anxiety Disorder-7 (GAD-7), and the Japanese version of the Irritable Bowel Syndrome Quality of Life measure (IBS-QOL-J). Results: The AGA IgG positivity rate was significantly higher in the suspected NCGS group (33.3%) than in the control group (13.3%; p < 0.01). Using clinical suspicion as the reference, the sensitivity and specificity of AGA IgG were 33.3% and 86.7%, respectively. Patients with suspected NCGS exhibited significantly lower physical and mental QOL and higher scores for depressive, anxiety, and somatic symptoms compared to controls. No significant clinical differences were found between AGA IgG-positive and IgG-negative individuals within the suspected NCGS group. Conclusions: AGA IgG demonstrated a specificity of 86.7% and a sensitivity of 33.3% for suspected NCGS, indicating its limited utility as a standalone biomarker. These findings suggest that suspected NCGS involves significant somatic and psychological burdens regardless of serological status. Future studies should explore whether a multi-marker panel could improve the identification of “True NCGS” in diverse clinical populations. Full article

Acoustic indices are popular tools for rapid biodiversity assessment using passive acoustic monitoring recordings, yet anthropogenic sounds in human activity areas compromise their robustness. In this paper, we focus on the typical urban–rural soundscape, where anthropogenic noise mainly originates from a narrow angular sector far from the monitoring device. We propose a denoising preprocessing algorithm with two microphone sensors for the robust application of existing acoustic indices. Our algorithm first develops an adaptive multi-tap null-steering beamformer based on a back-to-back first-order differential microphone array, which increases the system degrees of freedom to enhance the broadband interference cancellation capability. Building on this, a parallel bank of mutually orthogonal null-steering beamformers is proposed, each forming deep nulls toward directional interference-concentrated bands and generating diverse responses to the target signal. Finally, a signal compensation mechanism is applied to the beamformers’ outputs, mitigating the signal self-cancellation effects from these unconstrained adaptive beamformers prior to index calculation. The proposed preprocessing method is evaluated using the frequency-dependent acoustic diversity index as a representative of acoustic indices. Experiment results on both simulation and real-world recordings show that the proposed method generates high-fidelity acoustic information for subsequent acoustic index calculation over a much wider signal-to-interference-plus-noise ratio (SINR) range in urban–rural soundscapes characterized by directional anthropogenic interference. Full article

With the large-scale integration of offshore wind farms (OWFs), harmonic issues caused by the interaction between high-frequency switching of converters and complex network impedances pose severe challenges to power quality. Traditional harmonic monitoring heavily relies on post-event fixed-threshold alarm mechanisms, which struggle to achieve early warning during the low-distortion sub-health operation stage and lack the capability for multi-dimensional tracing of harmonic degradation sources. To address these limitations, this paper proposes a deep warning network for grid harmonics combining multi-step regression and multi-dimensional tracing within a unified multi-task learning (MTL) architecture. First, a deep shared feature encoder, integrating a bi-directional long short-term memory (Bi-LSTM) network with a multi-head self-attention (MHSA) mechanism, is utilized to extract high-order temporal coupling features between meteorological evolution and multi-node electrical states. Subsequently, the main task branch executes a k-step-ahead multivariate time-series regression to accurately predict the evolution trend of total harmonic distortion (THD) at both the point of common coupling (PCC) and the turbine terminal. Simultaneously, the auxiliary task branch performs multi-label micro-state classification based on relative degradation thresholds, achieving fine-grained multi-dimensional tracing covering spatial nodes, electrical attributes, and their joint micro-states. Experimental results on real-world OWF operational data demonstrate that through the joint optimization of regression and tracing tasks, the proposed MultiDimKStepMTL model significantly improves time-series prediction accuracy, achieving a $10.3\%$ relative improvement over single-task baselines, while substantially reducing computational overhead. This research successfully advances grid harmonic monitoring from passive response to proactive micro-state early warning, providing a solid, highly interpretable data-driven foundation for active filter control of offshore wind clusters. Full article

Handover optimization is essential for seamless connectivity in 5G and beyond networks. Existing approaches present fundamental challenges of centralized solutions achieving coordination and accuracy but creating privacy risks under the General Data Protection Regulation (GDPR), while distributed privacy-preserving approaches protect user data but lack the network-wide visibility necessary for optimal mobility decisions. This systematic review synthesizes 49 peer-reviewed studies published between 2010 and 2025, identified through a PRISMA-compliant search across IEEE Xplore, ScienceDirect, SpringerLink, MDPI, ACM Digital Library, and Google Scholar. Eligible studies addressed cellular handover or mobility management using traditional signal-based, Machine Learning, Federated Learning, Software-Defined Networking strategies, and reported quantitative performance metrics. A structured quality assessment evaluated methodological rigor, dataset validation, benchmarking practices, handover-specific metrics, and scalability. Synthesis evidence shows that existing approaches do not simultaneously satisfy critical requirements for next-generation mobility management of accuracy, privacy, scalability, and real-time network-wide coordination. Machine learning achieves high accuracy (up to 97%) but depends on centralized data; Reinforcement Learning supports real-time adaptation but incurs high computational costs; federated learning preserve privacy but suffers from limited global coordination; and software-defined networking enables centralized control but requires continuous transmission of raw data. Evidence quality is further limited to simulation-based assessments and limited real-world datasets. Overall, the reviews identify a clear evolution from reactive threshold-based methods towards proactive prediction and highlights the need for unified, privacy-preserving and globally coordinated handover frameworks. The findings point toward integrating federated learning with Software-Defined Mobile Networking as promising architectural direction for 6G mobility management. Full article

Rapid localisation of trapped victims after urban disasters is essential but challenging because Bluetooth Low Energy (BLE) beacons are intermittent, radio propagation is obstructed by rubble, UAVs are energy-constrained, and real-world multi-UAV training is impractical in high-risk search-and-rescue (SAR) environments. This study formulates post-disaster victim localisation as a cooperative Dec-POMDP and adapts a model-aided federated multi-agent reinforcement learning framework based on FedQMIX. The proposed pipeline combines a lightweight LoS/NLoS surrogate channel model, PSO-based victim-position estimation, return-to-base and map-feasibility safety checks, an SAR-aligned shaped reward, and a leakage-free centralised training state based on estimated rather than ground-truth victim locations. Each UAV trains locally inside a learned digital-twin simulator and periodically shares only QMIX network parameters, avoiding the exchange of raw trajectories or RSSI logs. The framework is evaluated on two synthetic post-earthquake urban maps representing a compact return-to-base scenario and a larger reach-to-destination scenario. Across five independent seeds per method and map, Model-Aided FedQMIX achieves the highest and most stable victim-localisation performance, with the clearest advantage observed in the larger long-horizon scenario. Additional diagnostic tests examine reward-weight sensitivity, RF channel-shift robustness, BLE/smartphone hardware heterogeneity, non-IID client-data variation, and partial-client FedAvg under missing client updates. The results indicate that combining model-aided localisation cues, decentralised value factorisation, SAR-aligned objective design, and federated parameter sharing can improve the robustness of UAV-based victim-localisation policies. The framework also clarifies deployment considerations for federated SAR coordination, including communication payload, privacy boundaries, heterogeneous client experience, device variability, and intermittent connectivity. This study remains simulation-based, and future validation with real UAVs, BLE devices, and rubble-inspired testbeds is required before operational deployment. Full article

Cascade hydropower scheduling is a high-dimensional, tightly constrained multi-objective optimization problem in which classical genetic and evolutionary algorithms struggle to find feasible solutions. Under random initialization, algorithms such as the Non-dominated Sorting Genetic Algorithm II (NSGA-II), the Non-dominated Sorting Genetic Algorithm III (NSGA-III), and the Constrained Two-Archive Evolutionary Algorithm (C-TAEA) rarely produce any feasible solution when the feasible region occupies a vanishingly small fraction of the search space. This paper presents a three-phase framework that combines physics-guided deep learning with evolutionary computation to accelerate both NSGA-II and NSGA-III. The method trains a Physics-Embedded Temporal Convolutional Network (PeTCN) as a differentiable surrogate model that explicitly incorporates physical constraints, applies gradient-based inverse optimization to obtain a feasible or near-feasible solution of high quality, and warm-starts NSGA-II or NSGA-III with that solution for efficient Pareto front exploration. Experiments on a real-world six-station cascade system show that, under a 1500 s fixed-time budget across 20 independent runs, Boosted NSGA-II and Boosted NSGA-III both find feasible solutions in all runs. Boosted NSGA-II and Boosted NSGA-III both reach the first feasible solution within roughly 50–60 generations of Phase 3 search on average, whereas standard NSGA-II produces no feasible run within the same budget and standard NSGA-III requires thousands of generations among its successful runs. The mean final hypervolume reaches $43.84\times {10}^6$ for Boosted NSGA-II and $46.52\times {10}^6$ for Boosted NSGA-III, and both boosted algorithms reach a target hypervolume of $35.00\times {10}^6$ in all 10 target-hypervolume runs. These results demonstrate that coupling physics-embedded surrogates with gradient-based initialization is an effective strategy for constrained multi-objective problems in which feasible solutions are extremely sparse. Full article

Cairo’s Islamic architectural heritage represents one of the world’s most significant concentrations of pre-industrial environmental ingenuity. For over a millennium, an integrated suite of passive climate-control systems—the Mašrabiya latticework screen, the open courtyard (Ṣaḥn), the wind-scoop (Malqaf), and stalactite vaulting (Muqarnas)—has moderated temperature, humidity, and airflow with remarkable effectiveness. Today, these inherited solutions are under unprecedented stress from urban densification, chronic particulate pollution, climate-driven temperature rise, and growing visitor footfall. This study investigates indoor environmental quality (IEQ) in six Fatimid- and Mamlūk-era buildings in Historic Cairo through the integrated IQAD-IAH framework, combining IoT field monitoring (January–December 2023) of temperature, relative humidity, CO₂, and PM2.5 with CNN-based deterioration image analysis and Random Forest predictive modeling. Results document critical summer thermal buffering failures reaching 28% of occupied hours above the ASHRAE 55 adaptive comfort limit; hygrothermal stress cycles exceeding the EN 15757 ±10% RH safe threshold for up to 38% of annual hours; and PM2.5 courtyard concentrations of 40–61 µg/m³ under normal conditions, surging to 180–320 µg/m³ during Ḫamāsῑn-seasonal wind events. Machine-learning projections indicate all three principal passive elements will cross the critical deterioration threshold of 70/100 under RCP 8.5 before 2050. A precautionary intervention window is identified between 2025 and 2032. Evidence-based management recommendations compatible with UNESCO World Heritage obligations are presented. Full article