Search Results (199)

With the rapid advancement of remote sensing technologies, the joint classification of hyperspectral image (HSI) and LiDAR data has become a key research focus in the field. To address the impact of inherent uncertainties in hyperspectral images on classification—such as the “same spectrum, different materials” and “same material, different spectra” phenomena, as well as the complexity of spectral features. Furthermore, existing multimodal fusion approaches often fail to fully leverage the complementary advantages of hyperspectral and LiDAR data. We propose a fuzzy-enhanced multi-scale cross-modal fusion network (FE-MCFN) designed to achieve joint classification of hyperspectral and LiDAR data. The FE-MCFN enhances convolutional neural networks through the application of fuzzy theory and effectively integrates global contextual information via a cross-modal attention mechanism. The fuzzy learning module utilizes a Gaussian membership function to assign weights to features, thereby adeptly capturing uncertainties and subtle distinctions within the data. To maximize the complementary advantages of multimodal data, a fuzzy fusion module is designed, which is grounded in fuzzy rules and integrates multimodal features across various scales while taking into account both local features and global information, ultimately enhancing the model’s classification performance. Experimental results obtained from the Houston2013, Trento, and MUUFL datasets demonstrate that the proposed method outperforms current state-of-the-art classification techniques, thereby validating its effectiveness and applicability across diverse scenarios. Full article

The escalating environmental noise pollution along urban rail transit corridors, exacerbated by rapid urbanization, necessitates innovative and efficient noise control measures. A comprehensive investigation was conducted that utilized field measurements of train passing-by noise to establish a statistical energy analysis model for evaluating the acoustic performance of both vertical (VB) and fully enclosed (FB) barrier configurations. The study incorporated Maa’s theory of micro-perforated plate (MPP) parameter optimization and developed a neural network surrogate model focused on insertion loss maximization for barrier geometric design. Key findings revealed significant barrier-induced near-track noise amplification, with peak effects observed at the point located 1 m from the barrier and 2 m above the rail. Frequency-dependent analysis demonstrated a characteristic rise-and-fall reflection pattern, showing maximum amplifications of 1.47 dB for VB and 4.13 dB for FB within the 400–2000 Hz range. The implementation of optimized MPPs was found to effectively eliminate the near-field noise amplification effects, achieving sound pressure level reductions of 4–8 dB at acoustically sensitive locations. Furthermore, the high-precision surrogate model (R² = 0.9094, MSE = 0.8711) facilitated optimal geometric design solutions. The synergistic combination of MPP absorption characteristics and geometric optimization resulted in substantially enhanced barrier performance, offering practical solutions for urban rail noise mitigation strategies. Full article

We present a formal framework for modeling neural network dynamics using Category Theory, specifically through Markov categories. In this setting, neural states are represented as objects and state transitions as Markov kernels, i.e., morphisms in the category. This categorical perspective offers an algebraic alternative to traditional approaches based on stochastic differential equations, enabling a rigorous and structured approach to studying neural dynamics as a stochastic process with topological insights. By abstracting neural states as submeasurable spaces and transitions as kernels, our framework bridges biological complexity with formal mathematical structure, providing a foundation for analyzing emergent behavior. As part of this approach, we incorporate concepts from Interacting Particle Systems and employ mean-field approximations to construct Markov kernels, which are then used to simulate neural dynamics via the Ising model. Our simulations reveal a shift from unimodal to multimodal transition distributions near critical temperatures, reinforcing the connection between emergent behavior and abrupt changes in system dynamics. Full article

Rolling bearings are widely used in rotating machinery, and their health status is crucial for the safe operation of the equipment. The research on relevant fault diagnosis algorithms is a hot topic in the field. As a leading deep learning paradigm, Convolutional Neural Networks (CNNs) have demonstrated remarkable effectiveness in bearing fault diagnosis. However, conventional CNNs encounter significant limitations in accurately identifying and classifying early-stage bearing faults, primarily due to two challenges: (1) the diagnostic accuracy is highly susceptible to variations in the input signal length and segmentation strategies and (2) incipient faults are characterized by extremely low signal-to-noise ratios (SNRs), which obscure fault signatures. To address these challenges, we propose a Waveform Intersection-CNN (WI-CNN)-based intelligent diagnosis method for early faults. This approach integrates Gramian Angular Field theory to construct high-resolution fault signatures, enabling the CNN-based diagnosis of incipient bearing faults. Validation using the Case Western Reserve University dataset demonstrates an average diagnostic accuracy exceeding 98%. Furthermore, we established a custom test platform to develop a hybrid diagnosis strategy for 10 distinct fault types. Comparative studies against two conventional CNN diagnostic methods confirm that our approach delivers superior diagnostic precision, a faster iteration speed, and enhanced algorithmic robustness. The empirical findings demonstrate that the model achieves an accuracy of 99.67% during training and 98.167% in the testing phase. Crucially, the proposed method offers exceptional simplicity, computational efficiency, and practical applicability, facilitating its widespread implementation. Full article

The study of dynamical processes on complex networks constitutes a foundational domain bridging applied mathematics, statistical physics, systems theory, and data science. Temporal evolution, not static topology, determines the controllability, stability, and inference limits of real-world systems, from epidemics and neural circuits to power grids and social media. However, the methodological landscape remains fragmented, with distinct communities advancing separate formalisms for spreading, control, inference, and design. This review presents a unifying six-pillar framework for the analysis of network dynamics: (i) spectral and structural foundations; (ii) deterministic mean-field reductions; (iii) control and observability theory; (iv) adaptive and temporal networks; (v) probabilistic inference and belief propagation; (vi) multilayer and interdependent systems. Within each pillar, we delineate conceptual motivations, canonical models, analytical methodologies, and open challenges. Our corpus, selected via a PRISMA-guided screening of 134 mathematically substantive works (1997–2024), is organized to emphasize internal logic and cross-pillar connectivity. By mapping the field onto a coherent methodological spine, this survey aims to equip theorists and practitioners with a transferable toolkit for interpreting, designing, and controlling dynamic behavior on networks. Full article

Traditional permanent-magnet synchronous motor (PMSM) models assume constant inductance parameters in the dq frame, attributing torque ripple solely to local non-sinusoidal disturbances while neglecting nonlinear effects like iron saturation, flux linkage spatial harmonics, and inter-axis mutual coupling. These simplifications limit such models to predicting average torque but fail to capture harmonic components. To overcome these limitations, this study develops a nonlinear PMSM model using differential inductance theory and constructs a physics-informed neural network (PINN) surrogate trained on finite-element data. The proposed hybrid framework demonstrates high-fidelity torque prediction, validated against finite-element simulations, and provides insights into harmonic generation mechanisms under saturation and spatial field distortions. Full article

In this work, we aimed to determine the nonlinear disturbance caused by cascaded coupling rigid–flexible deformation and friction in a direct-driven servo electric cylinder terminal positioning system (DDSEC-TPS) during feed motion of an intermittent, reciprocating, and time-varying load. For this purpose, a cascaded coupling dynamic error model of DDSEC-TPS was established based on the position–pose error model of the parallel motion platform and the rotor field-oriented vector transform. Then, a model to observe the dynamic error of the DDSEC-TPS was established using the improved beetle antennae search algorithm backpropagation neural network (IBAS-BPNN) prediction model according to the rigid–flexible deformation error theory of feed motion, and the observed dynamic error was compensated for in the vector control strategy of the DDSEC-TPS. The length and error prediction models were trained and validated using opposite and mixed datasets tested on the experimental platform, to observe dynamic errors and evaluate and optimize the prediction models. The experimental results show that dynamic error compensation can improve the position tracking accuracy of the DDSEC-TPS and the position–pose performance of the parallel motion platform. This study is of great significance for improving the consistency of following multiple DDSEC-TPSs and the position–pose accuracy of parallel motion platforms. Full article

Terahertz metamaterial dual-band absorbers are used for multi-target detection and high-sensitivity sensing in complex environments by enhancing information that reflects differences in the measured substances. Traditional design processes are complex and time-consuming. Machine learning-based methods, such as neural networks and deep learning, require a large number of simulations to gather training samples. Existing design methods based on single-objective optimization often result in uneven multi-objective optimization, which restricts practical applications. In this study, we developed a metamaterial absorber featuring a circular split-ring resonator with four gaps nested in a “卍” structure and used the Multi-Objective Firefly Algorithm based on Multiple Cooperative Strategies to achieve fast optimization of the absorber’s structural parameters. A comparison revealed that our approach requires fewer iterations than the Multi-Objective Particle Swarm Optimization and reduces design time by nearly half. The absorber designed using this method exhibited two resonant peaks at 0.607 THz and 0.936 THz, with absorptivity exceeding 99%, indicating near-perfect absorption and quality factors of 31.42 and 30.08, respectively. Additionally, we validated the absorber’s wave-absorbing mechanism by applying impedance-matching theory. Finally, we elucidated the resonance-peak formation mechanism of the absorber based on the surface current and electric-field distribution at the resonance frequencies. These results confirmed that the proposed dual-band metamaterial absorber design is efficient, representing a significant step toward the development of metamaterial devices. Full article

This study addresses the heterogeneous formation control problem for cooperative unmanned aerial vehicles (UAVs) and unmanned surface vehicles (USVs) operating under input quantization constraints. A unified mathematical framework is developed to harmonize the distinct dynamic models of UAVs and USVs in the horizontal plane. The proposed control architecture adopts a hierarchical design, decomposing the system into kinematic and dynamic subsystems. At the kinematic level, an artificial potential field method is implemented to ensure collision avoidance between vehicles and obstacles. The dynamic subsystem incorporates neural network-based estimation to compensate for system uncertainties and unknown parameters. To address communication constraints, a linear quantization model is introduced for control input processing. Additionally, adaptive control laws are formulated in the vertical plane to achieve precise altitude tracking. The overall system stability is rigorously analyzed using input-to-state stability theory. Finally, numerical simulations demonstrate the effectiveness of the proposed control strategy in achieving coordinated formation control. Full article

Gait is considered a valuable biometric feature, and it is essential for uncovering the latent information embedded within gait patterns. Gait recognition methods are expected to serve as significant components in numerous applications. However, existing gait recognition methods exhibit limitations in complex scenarios. To address these, we construct a dual-Kinect V2 system that focuses more on gait skeleton joint data and related acoustic signals. This setup lays a solid foundation for subsequent methods and updating strategies. The core framework consists of enhanced ensemble learning methods and Dempster–Shafer Evidence Theory (D-SET). Our recognition methods serve as the foundation, and the decision support mechanism is used to evaluate the compatibility of various modules within our system. On this basis, our main contributions are as follows: (1) an improved gait skeleton joint AdaBoost recognition method based on Circle Chaotic Mapping and Gramian Angular Field (GAF) representations; (2) a data-adaptive gait-related acoustic signal AdaBoost recognition method based on GAF and a Parallel Convolutional Neural Network (PCNN); and (3) an amalgamation of the Triangulation Topology Aggregation Optimizer (TTAO) and D-SET, providing a robust and innovative decision support mechanism. These collaborations improve the overall recognition accuracy and demonstrate their considerable application values. Full article

Mirror neurons (MNs), a set of neurons that are activated during the processes of observation and execution of actions, have drawn significant attention in the research of neurodegenerative and psychological disorders. Research in the field of Autism Spectrum Disorder (ASD) and schizophrenia demonstrates evidence in favour of common underlying neural mechanisms underlying the two conditions, especially with respect to mu rhythm suppression, a proxy for MN activation and socio-cognitive impairments. This paper aims to review the most recent studies on the neurological underpinnings of social cognition deficits and cognitive discrepancies shared by ASD and schizophrenia, as detected by measuring the functionality and activation of the mirror neuron system. The findings of the review reveal a lack of consensus with respect to the validity of the “broken mirror” theory. The review also shows that further research is warranted to shed light on the implications of mirror neuron dysfunction in neuropsychiatric conditions and assist the development of technological interventions and treatments. Full article

Video emotion recognition (VER), situated at the convergence of affective computing and computer vision, aims to predict the primary emotion evoked in most viewers through video content, with extensive applications in video recommendation, human–computer interaction, and intelligent education. This paper commences with an analysis of the psychological models that constitute the foundation of VER theory. The paper further elaborates on datasets and evaluation metrics commonly utilized in VER. Then, the paper reviews VER algorithms according to their categories, and compares and analyzes the experimental results of classic methods on four datasets. Based on a comprehensive analysis and investigations, the paper identifies the prevailing challenges currently faced in the VER field, including gaps between emotional representations and labels, large-scale and high-quality VER datasets, and the efficient integration of multiple modalities. Furthermore, this study proposes potential research directions to address these challenges, e.g., advanced neural network architectures, efficient multimodal fusion strategies, high-quality emotional representation, and robust active learning strategies. Full article

During the COVID-19 pandemic, social media platforms emerged as both vital information sources and conduits for the rapid spread of propaganda and misinformation. However, existing studies often rely on single-label classification, lack contextual sensitivity, or use models that struggle to effectively capture nuanced propaganda cues across multiple categories. These limitations hinder the development of robust, generalizable detection systems in dynamic online environments. In this study, we propose a novel deep learning (DL) framework grounded in fine-tuning the RoBERTa model for a multi-label, multi-class (ML-MC) classification task, selecting RoBERTa due to its strong contextual representation capabilities and demonstrated superiority in complex NLP tasks. Our approach is rigorously benchmarked against traditional and neural methods, including, TF-IDF with n-grams, Conditional Random Fields (CRFs), and long short-term memory (LSTM) networks. While LSTM models show strong performance in capturing sequential patterns, our RoBERTa-based model achieves the highest overall accuracy at 88%, outperforming state-of-the-art baselines. Framed within the diffusion of innovations theory, the proposed model offers clear relative advantages—including accuracy, scalability, and contextual adaptability—that support its early adoption by Information Systems researchers and practitioners. This study not only contributes a high-performing detection model but also delivers methodological and theoretical insights for combating propaganda in digital discourse, enhancing resilience in online information ecosystems. Full article

In communication, people adjust their information expression based on the audience’s attitude toward a topic, which is known as the audience-tuning effect. This effect also leads individuals to develop memory biases favoring the audience’s attitude, a process termed the “saying-is-believing” (SIB) effect. This study validates the SIB effect using a classical paradigm based on shared reality theory. Additionally, it explores the impact of different communication modalities on the SIB effect, considering the information dissemination context in the internet era and the unique characteristic of “visual anonymity” in online communication compared to offline communication. A two-factor mixed experimental design with 2 (audience’s attitude: positive, negative) × 2 (communication modality: online, offline) was employed. The following results were found: (1) The SIB effect exists, meaning that people adjust their descriptions and recalls based on the audience’s attitude. (2) Communication modality and the audience’s attitude interactively influence the SIB effect, with a greater deviation in description and recall valence when the audience’s attitude is negative (positive) in online (offline) compared to offline (online) communication. In summary, online communication is more likely to generate negative information than offline communication. This study enriches and expands the research field of the SIB effect, filling the gap in cross-media comparisons within this field. Moreover, it further enhances individuals’ understanding of online and offline communication modalities, which has certain guiding significance for enhancing work and learning effectiveness, improving the internet environment, and supporting enterprise management. Future research can further subdivide communication modalities, improve the classical paradigm to make it more practical, and incorporate neural technologies to delve deeper into the influencing factors and underlying mechanisms of the SIB effect. Full article

The properties of wood change after heat treatment, affecting its applications. Glossiness, a key aesthetic property, is of great significance in fields like furniture. Precise prediction can optimize the process and improve product quality. Although the traditional back propagation neural network (BPNN) has been applied in the field of wood properties, it still has issues such as poor prediction accuracy. This study proposes an improved whale optimization algorithm (IWOA) to optimize BPNN, constructing an IWOA-BPNN model for predicting the glossiness of heat-treated wood. IWOA uses chaos theory and tent chaos mapping to accelerate convergence, combines with the sine cosine algorithm to enhance optimization, and adopts an adaptive inertia weight to balance search and exploitation. A dataset containing 216 data entries from four different wood species was collected. Through model comparison, the IWOA-BPNN model showed significant advantages. Compared with the traditional BPNN model, the mean absolute error (MAE) value decreased by 66.02%, the mean absolute percentage error (MAPE) value decreased by 64.21%, the root mean square error (RMSE) value decreased by 69.60%, and the R² value increased by 12.87%. This model provides an efficient method for optimizing wood heat treatment processes and promotes the development of the wood industry. Full article