Search Results (157)

Camouflaged object detection (COD) represents one of the most challenging tasks in computer vision, requiring sophisticated approaches to accurately extract objects that seamlessly blend within visually similar backgrounds. While contemporary techniques demonstrate promising detection performance, they predominantly suffer from computational complexity and resource requirements that severely limit their deployment in real-time applications, particularly on mobile devices and edge computing platforms. To address these limitations, we propose LiteCOD, an efficient lightweight framework that integrates local and global perceptions through holistic feature fusion and specially designed efficient attention mechanisms. Our approach achieves superior detection accuracy while maintaining computational efficiency essential for practical deployment, with enhanced feature propagation and minimal computational overhead. Extensive experiments validate LiteCOD’s effectiveness, demonstrating that it surpasses existing lightweight methods with average improvements of 7.55% in the F-measure and 8.08% overall performance gain across three benchmark datasets. Our results indicate that our framework consistently outperforms 20 state-of-the-art methods across quantitative metrics, computational efficiency, and overall performance while achieving real-time inference capabilities with a significantly reduced parameter count of 5.15M parameters. LiteCOD establishes a practical solution bridging the gap between detection accuracy and deployment feasibility in resource-constrained environments. Full article

Biomimetic design, derived from the study of biological systems, has emerged as a pivotal methodology in contemporary art and design. By systematically integrating the morphological traits, structural principles, and functional mechanisms of living organisms into design thinking, it provides both a novel theoretical perspective and methodological support for modern design practice. This design philosophy draws abundant inspiration from nature’s aesthetics and achieves a profound fusion of organic form and artistic expression. This study systematically traces the theoretical evolution of biomimetic design—from its early phase of direct form-mimicry to today’s holistic, systems-based approach—and clarifies its interdisciplinary logic and developmental trajectory. We examine its applications in public installations, product development, architecture, and fashion. Through a structured analysis of plant-inspired, animal-inspired, and ecosystem-inspired strategies—linked with the aesthetic demands and cultural contexts of design—this study uncovers the underlying mechanisms by which biological models drive innovation. The findings demonstrate that, by organically combining form simulation, function optimization, and ecological awareness, biomimetic design not only elevates the aesthetic value, visual impact, and emotional resonance of design works but also amplifies their social role and cultural significance. Moreover, its interdisciplinary potential in materials innovation, technological integration, and environmental sustainability highlights unique pathways for addressing complex contemporary challenges. This study adopts a methodology that blends case-study analysis and theoretical interpretation. Through an in-depth examination of exemplar projects, it validates that biomimetic design not only achieves a seamless unity of function and form but also offers a robust theoretical framework and practical strategies for sustainable design implementation. These insights advance both the theoretical depth and practical innovation of the design discipline. Full article

Accurate classification of cardiac arrhythmias remains a crucial task in biomedical signal processing. This study proposes a hybrid deep learning approach called 1D CNN-eGRU that integrates one-dimensional convolutional neural network models (1D CNN) and a gated recurrent unit (GRU) architecture with an attention mechanism for the precise classification of cardiac arrhythmias based on ECG Lead II signals. To enhance the classification of cardiac arrhythmias, we also address data imbalances in the MIT-BIH arrhythmia dataset by introducing a hybrid data balancing method that blends resampling and class-weight learning. Additionally, we apply Sig-LIME, a refined variant of LIME tailored for signal datasets, to provide comprehensive insights into model decisions. The suggested hybrid 1D CNN-eGRU approach, a fusion of 1D CNN-GRU along with an attention mechanism, is designed to acquire intricate temporal and spatial dependencies in ECG signals. It aims to distinguish between four distinct arrhythmia classes from the MIT-BIH dataset, addressing a significant challenge in medical diagnostics. Demonstrating strong performance, the proposed hybrid 1D CNN-eGRU model achieves an overall accuracy of 0.99, sensitivity of 0.93, and specificity of 0.99. Per-class evaluation shows precision ranging from 0.80 to 1.00, sensitivity from 0.83 to 0.99, and F1-scores between 0.82 and 0.99 across four arrhythmia types (normal, supraventricular, ventricular, and fusion). The model also attains an AUC of 1.00 on average, with a final test loss of 0.07. These results not only demonstrate the model’s effectiveness in arrhythmia classification but also underscore the added value of interpretability enabled through the use of the Sig-LIME technique. Full article

Facial emotion recognition has become increasingly important in healthcare, where understanding delicate cues like pain, discomfort, or unconsciousness can support more timely and responsive care. Yet, recognizing facial expressions in real-world settings remains challenging due to varying lighting, facial occlusions, and hardware limitations in clinical environments. To address this, we propose TriViT-Lite, a lightweight yet powerful model that blends three complementary components: MobileNet, for capturing fine-grained local features efficiently; Vision Transformers (ViT), for modeling global facial patterns; and handcrafted texture descriptors, such as Local Binary Patterns (LBP) and Histograms of Oriented Gradients (HOG), for added robustness. These multi-scale features are brought together through a texture-aware cross-attention fusion mechanism that helps the model focus on the most relevant facial regions dynamically. TriViT-Lite is evaluated on both benchmark datasets (FER2013, AffectNet) and a custom healthcare-oriented dataset covering seven critical emotional states, including pain and unconsciousness. It achieves a competitive accuracy of 91.8% on FER2013 and of 87.5% on the custom dataset while maintaining real-time performance (~15 FPS) on resource-constrained edge devices. Our results show that TriViT-Lite offers a practical and accurate solution for real-time emotion recognition, particularly in healthcare settings. It strikes a balance between performance, interpretability, and efficiency, making it a strong candidate for machine-learning-driven pattern recognition in patient-monitoring applications. Full article

To address the challenges of model uncertainty, strong nonlinearities, and controller tuning in high-precision trajectory tracking for hydraulic servo systems, this paper proposes a hierarchical GA-PID-MPC fusion strategy. The architecture integrates three functional layers: a Genetic Algorithm (GA) for online parameter optimization, a Model Predictive Controller (MPC) for future-oriented planning, and a Proportional–Integral–Derivative (PID) controller for fast feedback correction. These modules are dynamically coordinated through an adaptive cost-aware blending mechanism based on real-time performance evaluation. The MPC module operates on a linearized state–space model and performs receding-horizon control with weights and horizon length $\theta =[q,r,T_p]$ tuned by GA. In parallel, the PID controller is enhanced with online gain projection to mitigate nonlinear effects. The blending coefficient $\sigma \left(t\right)$ is adaptively updated to balance predictive accuracy and real-time responsiveness, forming a robust single-loop controller. Rigorous theoretical analysis establishes global input-to-state stability and $H_{\infty }$ performance under average dwell-time constraints. Full article

Ubiquitous blended learning, leveraging mobile devices, has democratized education by enabling autonomous and readily accessible knowledge acquisition. However, its reliance on traditional interfaces often limits learner immersion and meaningful interaction. The emergence of the wearable metaverse offers a compelling solution, promising enhanced multisensory experiences and adaptable learning environments that transcend the constraints of conventional ubiquitous learning. This research proposes a novel framework for ubiquitous blended learning in the wearable metaverse, aiming to address critical challenges, such as multi-source data fusion, effective human–computer collaboration, and efficient rendering on resource-constrained wearable devices, through the integration of embodied interaction and multi-agent collaboration. This framework leverages a real-time multi-modal data analysis architecture, powered by the MobileNetV4 and xLSTM neural networks, to facilitate the dynamic understanding of the learner’s context and environment. Furthermore, we introduced a multi-agent interaction model, utilizing CrewAI and spatio-temporal graph neural networks, to orchestrate collaborative learning experiences and provide personalized guidance. Finally, we incorporated lightweight SLAM algorithms, augmented using visual perception techniques, to enable accurate spatial awareness and seamless navigation within the metaverse environment. This innovative framework aims to create immersive, scalable, and cost-effective learning spaces within the wearable metaverse. Full article

Low-latency and robust beamforming are vital for sustaining signal quality and spectral efficiency in emerging high-mobility 5G and future 6G wireless networks. Conventional beam management approaches, which rely on periodic Channel State Information feedback and static codebooks, as outlined in 3GPP standards, are insufficient in rapidly varying propagation environments. In this work, we propose a Dominance-Enforced Adaptive Clustered Sliding Window Regression (DE-ACSW-R) framework for predictive beamforming in O-RAN Split 7-2x architectures. DE-ACSW-R leverages a sliding window of recent angle of arrival (AoA) estimates, applying in-window change-point detection to segment user trajectories and performing both Linear Regression (LR) and curvature-adaptive Support Vector Regression (SVR) for short-term and non-linear prediction. A confidence-weighted fusion mechanism adaptively blends LR and SVR outputs, incorporating robust outlier detection and a dominance-enforced selection regime to address strong disagreements. The Open Radio Unit (O-RU) autonomously triggers localised MUSIC scans when prediction confidence degrades, minimising unnecessary full-spectrum searches and saving delay. Simulation results demonstrate that the proposed DE-ACSW-R approach significantly enhances AoA tracking accuracy, beamforming gain, and adaptability under realistic high-mobility conditions, surpassing conventional LR/SVR baselines. This AI-native modular pipeline aligns with O-RAN architectural principles, enabling scalable and real-time beam management for next-generation wireless deployments. Full article

Face recognition systems often falter when deployed in uncontrolled settings, grappling with low light, unexpected occlusions, motion blur, and the degradation of sensor signals. Most contemporary algorithms chase raw accuracy yet overlook the pragmatic need for uncertainty estimation and multispectral reasoning rolled into a single framework. This study introduces HUE-Net—a Human-centric, Uncertainty-aware, Event-fused Network—designed specifically to thrive under severe environmental stress. HUE-Net marries the visible RGB band with near-infrared (NIR) imagery and high-temporal-event data through an early-fusion pipeline, proven more responsive than serial approaches. A custom hybrid backbone that couples convolutional networks with transformers keeps the model nimble enough for edge devices. Central to the architecture is the perturbed multi-branch variational module, which distills probabilistic identity embeddings while delivering calibrated confidence scores. Complementing this, an Adaptive Spectral Attention mechanism dynamically reweights each stream to amplify the most reliable facial features in real time. Unlike previous efforts that compartmentalize uncertainty handling, spectral blending, or computational thrift, HUE-Net unites all three in a lightweight package. Benchmarks on the IJB-C and N-SpectralFace datasets illustrate that the system not only secures state-of-the-art accuracy but also exhibits unmatched spectral robustness and reliable probability calibration. The results indicate that HUE-Net is well-positioned for forensic missions and humanitarian scenarios where trustworthy identification cannot be deferred. Full article

Titanium alloys, especially Ti6Al4V, are widely used in biomedical implants due to their biocompatibility and mechanical strength. However, their high elastic modulus (>100 GPa), compared to that of human bone (10–30 GPa), often causes stress shielding, reducing implant lifespan. To address this, titanium alloys with lower elastic modulus are under development. In this study, Ti-based multi-element alloy with 16 wt.% Nb samples were fabricated using laser powder bed fusion (L-PBF) from a premixed powder blend of Ti6Al4V and Nb-Hf-Ti. Processing high-melting Nb-based alloys via L-PBF poses challenges, which were mitigated through optimized parameters, including a maximum laser power of 100 W. Eleven parameter sets were employed to evaluate printability, microstructure, and mechanical properties. Microstructural analysis revealed Widmanstätten structures composed of α and β phases, along with isolated spherical pores. Reduced hatch spacing and slower laser speed led to increased hardness. The highest hardness (~43 HRC) was observed at the highest energy density (266 J/mm³), while the lowest (~28 HRC) corresponded to 44 J/mm³. Elastic modulus values ranged from 30 to 35 GPa, closely matching that of bone. These results demonstrate the potential of the developed Ti-based alloy containing 16 wt.% Nb as a promising candidate for load-bearing biomedical implants. Full article

Bringing together the Internet of Things (IoT), LLMs, and Federated Learning (FL) offers exciting possibilities, creating a synergy to build smarter, privacy-preserving distributed systems. This review explores the merging of these technologies, particularly within edge computing environments. We examine current architectures and practical methods enabling this fusion, such as efficient low-rank adaptation (LoRA) for fine-tuning large models and memory-efficient Split Federated Learning (SFL) for collaborative edge training. However, this integration faces significant hurdles: the resource limitations of IoT devices, unreliable network communication, data heterogeneity, diverse security threats, fairness considerations, and regulatory demands. While other surveys cover pairwise combinations, this review distinctively analyzes the three-way synergy, highlighting how IoT, LLMs, and FL working in concert unlock capabilities unattainable otherwise. Our analysis compares various strategies proposed to tackle these issues (e.g., federated vs. centralized, SFL vs. standard FL, DP vs. cryptographic privacy), outlining their practical trade-offs. We showcase real-world progress and potential applications in domains like Industrial IoT and smart cities, considering both opportunities and limitations. Finally, this review identifies critical open questions and promising future research paths, including ultra-lightweight models, robust algorithms for heterogeneity, machine unlearning, standardized benchmarks, novel FL paradigms, and next-generation security. Addressing these areas is essential for responsibly harnessing this powerful technological blend. Full article

Rail surface defects pose significant risks to the operational efficiency and safety of industrial equipment. Traditional visual defect detection methods typically rely on high-quality RGB images; however, they struggle in low-light conditions due to small, low-contrast defects that blend into complex backgrounds. Therefore, this paper proposes a novel defect segmentation method leveraging a dual-stream feature fusion network that combines polarization images with DeepLabV3+. The approach utilizes the pruned MobileNetV3 as the backbone network, incorporating a coordinate attention mechanism for feature extraction. This reduces the number of model parameters and enhances computational efficiency. The dual-stream module implements cascade and addition strategies to effectively merge shallow and deep features from both the original and polarization images. This enhances the detection of low-contrast defects in complex backgrounds. Furthermore, the CBAM is integrated into the decoding area to refine feature fusion and mitigate the issue of missing small-target defects. Experimental results demonstrate that the enhanced DeepLabV3+ model outperforms existing models such as U-Net, PSPNet, and the original DeepLabV3+ in terms of MIoU and MPA metrics, achieving 73.00% and 80.59%, respectively. The comprehensive detection accuracy reaches 97.82%, meeting the demanding requirements for effective rail surface defect detection. Full article

Sustainable polymers have attracted interest due to their ability to biodegrade under specific conditions in soil, compost, and the marine environment; however, they have comparatively lower mechanical properties, limiting their widespread use. This study explores the effect of incorporating waste soy biomass into sustainable polymers (including biodegradable and biobased) on the thermal and mechanical properties of the resultant blends. The dispersion of the waste soy biomass in the polymer matrix is also investigated in relation to particle size (17 µm vs. 1000 µm). Fine waste soy biomass did not significantly affect the melting temperature of the polymers (polyhydroxyalkanoates, polybutylene adipate terephthalate, polybutylene adipate terephthalate/poly(lactic) acid, and biobased linear low-density polyethylene) used in this study, but their enthalpy of fusion decreased after soy was melt-blended with the polymers. The tensile modulus of the polymers filled with fine waste soy biomass powder (17 µm) was enhanced when melt-blended as compared to unfilled polymers. Additionally, it was found that fine waste soy powder (17 µm) increased the tensile modulus of the polymer blends without significantly affecting processability, while coarse waste soy meal (1000 µm) generally reduced elongation at break due to poor dispersion and stress concentration; however, this effect was less pronounced in PHA blends, where improved compatibility was observed. Full article

Remaining useful life (RUL) prediction of industrial equipment is critical for achieving safe operations and optimizing predictive maintenance. To tackle the limitations of poor interpretability, inaccurate predictions, and high computational cost in complex system degradation modeling, this paper proposes SiMBA-PINN, a novel fusion framework that synergizes Physics-Informed Neural Network (PINN) with an enhanced state-space model (SiMBA). The framework achieves dynamic fusion of data-driven features and physical laws through a two-branch synergistic mechanism: the temporal modeling branch combines selective state-space SiMBA with Einstein Fast Fourier Transform (EinFFT)-based spectral mixing to efficiently capture cross-sensor temporal dependencies and degradation trends, while the physics-constraint branch embeds automatically differentiable partial differential equation residuals derived from domain-specific degradation mechanisms, enforcing physical consistency through deep hidden physics modeling. Here, the EinFFT-based spectral mixing leverages frequency-domain interactions to effectively blend the spectral components of multivariate time-series data, thereby enhancing the modeling of cross-sensor dependencies. Meanwhile, deep hidden physics modeling integrates physics-informed partial differential equation (PDE) residuals through differentiable operators, aligning the learned representations with domain-specific dynamics via a constraint-driven loss design. Experimental results from the C-MAPSS dataset confirm that the proposed model significantly outperforms PINN-, Mamba- and attention mechanism-based models, achieving State-of-the-Art RMSE on the most challenging FD004 subset. This physics-aware framework achieves deployable and interpretable RUL prediction by balancing accuracy with linear-time complexity. Full article

This study examines the effectiveness of combining semantic intelligence drawn from large language models (LLMs) such as ChatGPT-4o with traditional machine-learning (ML) algorithms to develop predictive portfolio strategies for NASDAQ-100 stocks over the 2020–2025 period. Three different predictive frameworks––fundamental, technical, and entropy-based––are tested through examination of novel combinations of ML- and LLM-derived semantic metrics. The empirical results reveal a considerable divergence in optimal blending methods across the methodologies; namely, the technical methodology exhibits the best performance when using only ML predictions, with around 1978% cumulative returns with monthly rebalancing. In contrast, the fundamental methodology achieves its full potential when it is based primarily on LLM-derived semantic insights. The Entropy methodology is improved by a balanced combination of both semantic and ML signals, thus highlighting the potential of LLMs to improve predictive power by offering interpretative context for complex market interactions. These findings highlight the strategic importance of tailoring the semantic–algorithmic fusion to suit the nature of the predictive data and the investment horizon, with significant implications for portfolio management and future research in financial modeling. Full article

So far, synthetic biology approaches for the construction of artificial microorganisms have fostered the transformation of acceptor cells with genomes from donor cells. However, this strategy seems to be limited to closely related bacterial species only, due to the need for a “fit” between donor and acceptor proteomes and structures. “Fitting” of cellular regulation of metabolite fluxes and turnover between donor and acceptor cells, i.e. cybernetic heredity, may be even more difficult to achieve. The bacterial transformation experiment design 1.0, as introduced by Frederick Griffith almost one century ago, may support integration of DNA, macromolecular, topological, cybernetic and cellular heredity: (i) attenuation of donor Pneumococci of (S) serotype fosters release of DNA, and hypothetically of non-DNA structures compatible with subsequent transfer to and transformation of acceptor Pneumococci from (R) to (S) serotype; (ii) use of intact donor cells rather than of subcellular or purified fractions may guarantee maximal diversity of the structural and cybernetic matter and information transferred; (iii) “Blending” or mixing and fusion of donor and acceptor Pneumococci may occur under accompanying transfer of metabolites and regulatory circuits. A Griffith transformation experiment design 2.0 is suggested, which may enable efficient exchange of DNA as well as non-DNA structural and cybernetic matter and information, leading to unicellular hybrid microorganisms with large morphological/metabolic phenotypic differences and major features compared to predeceding cells. The prerequisites of horizontal gene and somatic cell nuclear transfer, the molecular mechanism of transformation, the machineries for the biogenesis of bacterial cytoskeleton, micelle-like complexes and membrane landscapes are briefly reviewed on the basis of underlying conceptions, ranging from Darwin’s “gemmules” to “stirps”, cytoplasmic and “plasmon” inheritance, “rhizene agency”, “communicology”, “transdisciplinary membranology” to up to Kirschner’s “facilitated variation”. Full article