Search Results (75)

Radar-based human activity recognition (HAR) has emerged as a promising alternative to vision-based systems, as it can operate under poor lighting, occlusions, and privacy-sensitive scenarios. However, existing radar HAR methods often suffer from limited feature extraction capability due to noise, signal attenuation, and the challenge of capturing both global motion patterns and local micro-Doppler dynamics simultaneously. To address these issues, this paper proposes RepHARNet, a novel HAR network built on structural reparameterization for radar micro-Doppler signals. Specifically, RepHARNet decomposes the weight matrices into global and local components through a reparameterization strategy, enabling the network to simultaneously capture coarse-grained inter-channel dependencies via a Global Perceptron and fine-grained intra-channel spatial correlations via a Channel Perceptron. Furthermore, a parameter-efficient share-set mechanism is integrated into the Channel Perceptron to substantially reduce the computational overhead while maintaining the representational capacity. Extensive experiments on the public IMG848 dataset demonstrate that all four RepHARNet variants achieve top-1 accuracies above 93.00%, among which RepHARNet-large achieves the highest at 94.86%, significantly outperforming existing mainstream methods. Additional evaluations on the Ci4R dataset further verify the robustness and effectiveness of RepHARNet, where all variants achieve competitive accuracies above 90.00%. The results verify the effectiveness and superiority of RepHARNet in radar-based HAR. Full article

Sensor-based human activity recognition (HAR) plays a fundamental role in healthcare monitoring, sports analytics, and ambient-assisted living. Although deep learning has substantially advanced HAR performance, two practical issues still limit its real-world deployment: (i) the distribution shift caused by changes in users or sensor placements can degrade generalization, and (ii) the quadratic $\mathcal{O}\left(L^2\right)$ complexity of standard self-attention hinders efficient long-sequence modeling on resource-constrained wearable devices. To address these issues, we propose DSHformer, which is an accuracy-oriented HAR framework that combines compact channel–temporal encoding with locality-sensitive hashing (LSH)-based attention. Specifically, DSHformer (i) employs a low-parameter patch-based graph-attention encoder to jointly model latent relationships among sensor channel–temporal dynamics; (ii) introduces a trainable prototype pool together with a multi-layer decomposition network to improve intra-class compactness and inter-class separability via prototype alignment; and (iii) introduces a decomposition-stable LSH-based attention mechanism tailored for HAR, whose core design couples prototype-guided feature decomposition with locality-sensitive hashing to ensure that semantically related tokens remain consistently grouped in the same hash bucket even after decomposition-induced attenuation. The mechanism thereby operates at $\mathcal{O}(LlogL)$ attention complexity on longer sensor sequences. Extensive experiments on five public benchmarks (WISDM, UCI-HAR, PAMAP2, Opportunity, and UniMiB-SHAR) show that DSHformer achieves accuracies of 98.6%, 93.7%, 98.4%, 88.5%, and 96.6%, respectively, achieving competitive or superior performance compared with both Transformer variants and HAR-specific baselines under the adopted benchmark protocols. Ablation studies further confirm the complementary contribution of each component. Full article

Effectively integrating graph topology and node attributes, while assigning nodes with both semantic similarity and structural closeness to the same community, remains a key challenge in attributed graph community detection. To address this challenge, this study proposes TVHCL-CD, a two-view hierarchical contrastive learning-driven method for community detection. The proposed method constructs an attribute view and a modularity view from the node attribute matrix and the modularity matrix, respectively, to model attribute semantics and high-order community structure priors. Structure-aware two-view representations are then learned in parallel through dual-view graph attention encoders incorporating multi-order neighborhood priors. Furthermore, a structure-enhanced Graph Transformer fusion module is designed to achieve node-level adaptive fusion of the two-view representations by introducing a learnable adjacency bias into global self-attention and a view-aware gating mechanism into the feed-forward network. To align the optimization objective with community semantics, a hierarchical contrastive learning strategy is further developed. Specifically, view-level consistency contrastive learning constructs modularity-guided augmented views to improve representation robustness, while community-level semantic contrastive learning incorporates partial ground-truth labels to enhance intra-community compactness and inter-community separation. Finally, clustering is performed on the fused representations to obtain community partitions. Experimental results on eight real-world attributed graphs and the generated tree-like attributed graph Tree-2500 indicate that TVHCL-CD achieves competitive performance under the semi-supervised transductive setting, while ablation results support the contributions of its main components. Full article

This study develops a game-theoretic framework to analyze collaborative advertising decisions between manufacturers and retailers in seasonal product supply chains characterized by competitive–cooperative channel relationships. We formulate a mathematical programming model to jointly optimize advertising efforts, the manufacturer’s advertising cost-sharing rate, order quantities, and inventory decisions across distinct channel configurations—including a single manufacturer–retailer dyad and a competitive multi-channel market. Numerical experiments and sensitivity analyses are conducted to investigate how key structural parameters—particularly demand elasticity and channel power asymmetry—influence overall system performance and equilibrium decision outcomes. Results indicate that well-designed collaborative advertising mechanisms enhance total channel profitability and, under specific conditions, yield Pareto-improving outcomes for both parties. This study makes three primary contributions: (i) it integrates inter-firm competition with intra-channel cooperation within a unified strategic framework; (ii) it jointly coordinates advertising and inventory decisions—two critical operational levers—rather than treating them in isolation; and (iii) it embeds financial arrangements (e.g., cost sharing) endogenously into the analytical model, thereby offering a novel, theoretically grounded, and practically implementable decision-support framework for distribution systems operating in complex, dynamic market environments. Full article

Time series anomaly detection aims to identify deviations from the normal distribution of temporal data. Reconstruction error is a natural and practical anomaly criterion, and using reconstruction error as an anomaly criterion is a well-established and practical paradigm. However, existing reconstruction-based methods often fail to capture complex structures in high-dimensional time series data and typically lack in-depth analysis of periodicity, limiting detection accuracy. To address these challenges, we propose ReDTF-AD, a novel approach that integrates reconstruction error with time–frequency fusion. Specifically, the input series is decomposed into seasonal and trend components. For seasonal components, we designed a time–frequency fusion block (TFFB) to enhance frequency-domain features while preserving residual time-domain information, ultimately achieving the fusion of time–frequency information. The Top-k transformation converts 1D sequences into 2D representations based on the periodicity of time series data, enabling deeper analysis of intra-period and inter-week variations through our newly proposed Split Concat Block (SCBlock). For the trend term, a linear module captures long-term patterns. In the unsupervised time series anomaly detection experiments based on reconstruction error, ReDTF-AD shows competitive performance on three public benchmark datasets. Full article

Motor imagery electroencephalography (MI-EEG) decoding remains challenging in cross-subject scenarios due to pronounced inter-subject variability and signal non-stationarity, which often lead to performance degradation on unseen subjects. Existing prototype-based and domain adaptation methods typically rely on global feature alignment or single-level class representation, limiting their ability to capture both channel-wise spatial variability and high-level semantic structure. To address these limitations, we propose a dual-stage prototype representation framework for cross-subject MI-EEG decoding. The framework models spatial and semantic variability in a hierarchical manner by introducing channel prototypes and feature prototypes, enabling more consistent representations across subjects. Furthermore, a prototype-guided pairwise similarity learning strategy is employed to enhance intra-class compactness and inter-class separability in the embedding space. To mitigate cross-subject distribution shifts, we integrate a lightweight statistical perturbation method (StyleMix) with Wasserstein-based domain alignment, helping reduce subject-specific distribution variations. Experiments on the BCI Competition IV 2a and 2b datasets show that the proposed method achieves competitive performance under the evaluated target-assisted few-shot setting, reaching average accuracies of 79.12% and 87.31%, respectively, and improving over the strongest baseline by up to 2.99 percentage points. Full article

Fatal intraspecific aggression in brown bears (Ursus arctos) remains poorly documented, yet elucidating its dynamics is critical in order to understand species’ physiology, informing management strategies, and advancing wildlife forensic science applications, which are useful in cases where a natural or illegal cause of death needs to be discerned. In this study, we reported four confirmed cases of lethal aggression (two yearlings and two adults) in the Italian Alps. Comprehensive autopsies were performed to characterize lesion patterns and infer the aggressor identity. Claw-induced lacerations, bite marks and the aspect of hemorrhages suggested the attack sequence. Aggressor identity was investigated by using forensic odontology through inter-canine distance (I-CD) and genetic analysis of peri-lesional saliva. I-CD allowed us to plausibly hypothesize the aggressor’s species and, in the cases where it was possible, to classify the sex and/or age group of the aggressors. While genetic analysis allowed the identification of the four brown bear victims, it did not provide informative results on the aggressors. The cause and manner of death were coded according to international criteria (International Classification of Diseases 11th Revision [ICD-11], WHO). Adult fatalities, supported by gastric content analysis, reflect trophic competition regardless of the mating context and highlight the role of anthropogenic food sources in conflict emergence. These findings underscore the value of integrated approaches in wildlife investigations and provide new insights into ecophysiological factors driving lethal intraspecific aggression. Full article

Background/Objectives: Legionella micdadei is a clinically significant species within the Legionella genus, requiring accurate detection methods, surveillance, and precise clinical diagnosis. Our objective was to develop a sensitive polymerase chain reaction (PCR) assay specific for L. micdadei to detect its presence in environmental specimens. Methods: We targeted the 23S–5S intergenic spacer region, which can differentiate Legionella spp. We tested the detection of L. micdadei with 20 strains and determined the limit of detection with 2 strains. We verified assay specificity with 17 strains of other Legionella spp., 62 strains of other bacterial and fungal genera, and three human DNA specimens. We evaluated intra- and inter-run precision. We tested 15 environmental specimens (water, swabs of water faucets, mulch, and soil) by PCR. Results: The PCR assay demonstrated 100% analytical specificity (no cross-reactivity with non-targeted species), 100% inclusivity (detection of all L. micdadei strains), and high precision, with a coefficient of variation ≤ 2% across replicates. The limit of detection was estimated at 5 genomic DNA copies per reaction. We detected L. micdadei in environmental specimens. Conclusions: This PCR assay enables accurate detection of L. micdadei and is not subject to competition with other Legionella spp., thereby addressing limitations of current broad-spectrum Legionella approaches. The evaluation supports its application in environmental detection for surveillance. Full article

Intra-specific variation in functional traits and their inter-relationships reflect how plants allocate resources, adapt, and evolve in response to environmental changes. This study investigated eight functional traits—leaf area (LA), specific leaf area (SLA), leaf dry matter content (LDMC), chlorophyll content (CHL), leaf nitrogen content (LNC), leaf phosphorus content (LPC), twig tissue density (TTD), and wood density (WD)—in Cunninghamia lanceolata plantations of three stand ages (15, 30, and 50 years), using a space-for-time substitution approach. We examined differences in trait values, intra-specific variation, and trait correlations across forest ages and diameter classes. The results showed that (1) Functional traits exhibited varying degrees of intra-specific variation, with LA having the highest coefficient of variation (21.66%) and LPC is lowest (9.31%). (2) Forest age had a stronger influence on trait variation than diameter class, with all traits differing significantly across ages, while only WD varied significantly among diameter classes. (3) PC1 (25.5%) and PC2 (19.4%) together explained approximately 44.9% of the total variation, with PC1 primarily reflecting functional trait changes driven by forest age. PCA results showed that LA and CHL tended to exhibit higher values in young forests, whereas SLA, LDMC, LPC, and LNC had relatively higher values in mature forests. This pattern suggests a shift in functional trait expression from resource acquisition to resource conservation strategies with increasing forest age. (4) Significant positive correlations between LNC and LPC, and negative correlations between SLA and LDMC, were observed in most groups, except in large-diameter trees at the over-mature stage. C. lanceolata adjusts trait combinations to enhance fitness across developmental stages. Juvenile trees adopt traits favoring efficient light and nutrient use to support rapid growth and competition. Middle-aged trees prioritize balanced water and nutrient use to maintain productivity and resist disturbances. Mature trees focus on sustained resource use and offspring protection to support ecosystem stability and regeneration. These findings reveal age-specific adaptive strategies and provide insights into the coordination and trade-offs among traits in response to environmental conditions. Full article

Due to the severe hazard of aflatoxins (AFs) to humans, it is of great significance to detect the key aflatoxins, aflatoxin B₁ (AFB₁) and aflatoxin G₁ (AFG₁), in food and feed in simple, rapid, and semi-quantitative ways. The hybridoma clone 3A1 was prepared in this study, and anti-AFB₁ monoclonal antibody (mAb) with high specificity and affinity (9.38 × 10⁸ L/mol) from 3A1 was purified. The indirect competitive enzyme-linked immunosorbent assay (ic-ELISA) demonstrated that the linear detection range for AFB₁ was 0.029–1.526 ng/mL with a limits of determination (LOD) of 0.023 ng/mL. A latex microsphere-based immunochromatographic test strip (LM-ICTS) was constructed based on 3A1, which showed that the strip could detect AFB₁ (LOD: lower than 1.79 ng/mL) and AFG₁ (LOD: lower than 8.08 ng/mL), and the linear detection ranges for AFB₁ and AFG₁ are 1.79–48.46 ng/mL and 8.08–107.40 ng/mL, respectively. The average recoveries of intra-assay and inter-assay for peanuts were (98.4 ± 4.7)% and (92.6 ± 7.6)%, and the average coefficient of variation (CVs) were 4.38% and 8.15%, respectively. For sunflower seeds, the intra-assay and inter-assay recoveries were (94.4 ± 7.2)% and (89.2 ± 4.3)%, and the average CVs were 6.6% and 4.9%, respectively. In summary, the developed LM-ICTS exhibited excellent sensitivity and specificity, which provided a rapidly stable on-site detection choice for AFB₁ and AFG₁ to contaminated agricultural samples, including grain and feed. Full article

Peste des petits ruminants virus (PPRV) has been classified into four lineages based on the nucleocapsid and fusion genes, with lineage IV strains being the most widely distributed. In Africa, recent epidemiological data revealed that PPRV lineage IV is increasingly displacing other lineages in prevalence, suggesting a competitive advantage in viral transmission and adaptability. Moreover, a lineage IV strain was the only confirmed strain in Europe and Asia. In this study, a one-step Taqman quantitative real-time reverse transcription polymerase chain reaction (RT-qPCR) assay for lineage IV PPRV was established by targeting the hemagglutinin (H) gene. The results indicated that this method could detect approximately six copies of PPRV RNA, indicating high sensitivity. No cross-reactions with related viruses or other lineages of PPRV were observed. The results of a repeatability test indicated that the coefficient of variation values were low in both the inter-assay and intra-assay experimental groups. Detection of field samples indicated that all positive samples could be detected successfully using the developed method. This RT-qPCR assay provides a valuable tool to facilitate targeted surveillance and rapid differential diagnosis in regions with active circulation of PPRV lineage IV, enabling timely epidemiological investigations and strain-specific identification. Full article

Living mulch intercropping systems are considered as nature-based solutions with a low environmental footprint for managing weeds, improving biodiversity and agroecosystem sustainability. In drylands, however, they may increase intra/inter-specific competition for water, reducing crop productivity. We tested conservation tillage (TLG) carob plots with and without irrigation (TLG_irr; TLG_dry) vs. rainfed intercropping systems of carob and (i) thyme (Thymbra capitata; T-System) or (ii) clover (Trifolium squarrosum; C-System), strategically planted on the south (sun)-exposed soil side (SES) of carobs, to reduce soil temperature/evaporation. Carob water relations, productivity and environmental footprints were examined for three years under semi-arid, low weed-competition (Skarinou-SKR) and arid high weed-competition (Vrysoules-VRY) conditions in Cyprus. Carob yield efficiency (kg/m³) in SKR, was >27% higher for the T-System (p < 0.05; SES cover ca. 85%; year-3), matching a higher leaf water content (p < 0.001) compared to TLG_dry. The T-System reached 28% and 56% of TLG_irr yields during very dry and normal rainfall years; TLG_dry yields approached zero. For VRY, no negative impacts on carob leaf water, at 25% SES cover, were found. SKR’s C-System improved leaf water content (p < 0.05) for only one year. The T-System also outperformed TLG_irr and TLG_dry in terms of reducing irrigation needs and energy consumption, breaking new grounds for dryland agroforestry. Full article

With the increase in greenwashing, corporate greenwashing governance has become a crucial component of urban environmental management. Industrial clusters are a key form of urban economic organization, yet the mechanisms through which they affect corporate greenwashing remain unclear. This study examines how different types of industrial agglomeration influence corporate greenwashing using a sample of Chinese A-share listed companies. The key findings include the following: (1) Specialized agglomeration inhibits corporate greenwashing through a “supervision” effect generated by intra-industry competition, while diversified agglomeration exacerbates greenwashing via a “collusion” effect arising from inter-industry cooperation. (2) The inhibitory role of specialized agglomeration is amplified under conditions of low public and strong government environmental concern, while the promotional effect of diversified agglomeration becomes more pronounced in contexts of high public and weak government environmental concern. Government environmental concern can be categorized into “general” and “specific” types, with the former being more effective in governing corporate greenwashing. (3) Specialized agglomeration demonstrates superior efficacy in curbing greenwashing among firms with green innovations related to energy-saving, alternative energy production, waste management, and transportation, while diversified agglomeration intensifies greenwashing tendencies in firms without green innovations. (4) Collusive greenwashing under diversified agglomeration yields short-term firm value gains but incurs hidden costs, including elevated operational risks and declining profit margins. This research provides critical insights for promoting corporate green transition and fostering zero-carbon industrial clusters. Full article

A competition experiment between Vachellia gerrardii and invasive Nicotiana glauca Graham was conducted to assess the impact of Arbuscular Mycorrhizal Fungi (AMF) symbiosis on the inter and intraspecific competition between the two species. Seedlings were established under mono and mixed plantations with different species proportions (3:1, 2:2, 1:3) and plant densities (1, 2, 3, and 4 plants/pot) for mixed and mono planting respectively, with and without AMF. The vegetative growth parameters (height, leaf area and number, total dry weight/plant, relative yield, relative yield total), roots characteristics (length, surface area, volume, tips number), competitive interaction (aggressivity), and physiological traits (chlorophyll a, chlorophyll b, photosynthesis, stomatal conductance) were measured to evaluate plant responses to AMF symbiosis and competition. The results revealed that AMF symbiosis significantly enhanced the vegetative parameters (leaf area, height, and total dry weight) in both species under mono and mixed plantations compared to plants without AMF. Under AMF treatment, in the interspecific competition, most vegetative and root parameters of N. glauca were higher than V. gerrardii. At inoculant and species proportions, the relative yield of N. glauca exceeded that for V. gerrardii; however, N. glauca was more aggressive towards V. gerrardii. N. glauca root indices were higher than V. gerrardii under inter and intraspecific competition. Simultaneously, for both species, in monoculture plantations, most parameters decreased as plant density increased, wherein the decrease was higher for plants grown without AMF. Photosynthesis increased in AMF treatment, particularly for N. glauca. In conclusion, AMF promoted the growth of invasive N. glauca more than native V. gerrardii, particularly in terms of the root system. Our results provide a critical perspective that the AMF has the potential to contribute and facilitate the invasion of N. glauca, as well as support it with a competitive advantage over V. gerrardii, thus highlighting its potential role in shaping plant–plant interaction in invaded habitats. Full article

Camels, with the ability to survive under drought and chronic hunger, developed exceptional efficient lipid reserves and energy substance metabolic characteristics. Fibroblast growth factor (FGF) 21 is a hormone that regulates important metabolic pathways and energy homeostasis. However, the absence of a specific detection method for camel FGF21 impacts research on camels’ metabolic regulation. This study established a direct competition ELISA assay for detecting camel FGF21. Camel FGF21 antigen was expressed and purified through prokaryotic expression system. Polyclonal antibody was produced and purified via immunizing guinea pigs and affinity chromatography assay. Biotin-labeled FGF21 was synthesized artificially as the competitive antigen. After the determination of optimal conditions, including the working concentrations of the antibody and antigen, blocking solution, dilution buffer, and the competition reaction time, the standard curve with a typical “S” shape was generated using GraphPad Prism. The regression equation was Y = 0.1111 + (X^−0.7894) × (2.162 − 0.1111)/(X^−0.7894 + 15.76^−0.7894), with the IC₅₀ 15.59 ng/mL, the limit of detection (LOD) 0.024 ng/mL, the limit of quantification (LOQ) 1.861 ng/mL, and the linear range IC₂₀~IC₈₀ 2.0~119.22 ng/mL. The verification test showed that the recovery rate ranged from 91.34% to 98.9%, and the coefficients of variation for the intra- and inter-plate both were less than 10%, indicating that the ELISA method had high accuracy, good repeatability, and high stability. In addition, this ELISA method had the potential to detect FGF21 secretion levels in other species such as mouse, human, and pig. This study provided a rapid quantitative tool for conducting research on the FGF21 factor in camels. Full article