Search Results (644)

Accurate extraction of water bodies in remote sensing images is crucial for natural disaster prediction, aquatic ecosystem monitoring, and resource management. However, most existing deep-learning-based methods primarily operate in the raw pixel space of images and fail to leverage the frequency characteristics of remote sensing images, resulting in an inability to fully exploit the representational power of deep models when predicting mask images. This paper proposes a Frequency-Band Decoupling Detail Enhancement Network (FDEN) to achieve high-precision water body extraction. The FDEN begins with an initial decoupling and enhancement stage for frequency information. Based on this multi-frequency representation, we further propose a Multi-Band Detail-Aware Module (MDAM), designed to adaptively enhance salient structural cues for water bodies across frequency bands while effectively suppressing irrelevant or noisy components. Extensive experiments demonstrate that the FDEN model outperforms state-of-the-art methods in terms of its segmentation accuracy and robustness. Full article

Knee osteoarthritis (KOA) is a most prevalent chronic muscoloskeletal disorder causing pain and functional impairment. Accurate predictions of KOA evolution are important for early interventions and preventive treatment planning. In this paper, we propose a novel dynamic hypergraph-based risk model (DyHRM) which integrates the encoder–decoder (ED) architecture with hypergraph convolutional neural networks (HGCNs). The risk model is used to generate longitudinal forecasts of KOA incidence and progression based on the knee evolution at a historical stage. DyHRM comprises two main parts, namely the dynamic hypergraph gated recurrent unit (DyHGRU) and the multi-view HGCN (MHGCN) networks. The ED-based DyHGRU follows the sequence-to-sequence learning approach. The encoder first transforms a knee sequence at the historical stage into a sequence of hidden states in a latent space. The Attention-based Context Transformer (ACT) is designed to identify important temporal trends in the encoder’s state sequence, while the decoder is used to generate sequences of KOA progression, at the prediction stage. MHGCN conducts multi-view spatial HGCN convolutions of the original knee data at each step of the historic stage. The aim is to acquire more comprehensive feature representations of nodes by exploiting different hyperedges (views), including the global shape descriptors of the cartilage volume, the injury history, and the demographic risk factors. In addition to DyHRM, we also propose the HyGraphSMOTE method to confront the inherent class imbalance problem in KOA datasets, between the knee progressors (minority) and non-progressors (majority). Embedded in MHGCN, the HyGraphSMOTE algorithm tackles data balancing in a systematic way, by generating new synthetic node sequences of the minority class via interpolation. Extensive experiments are conducted using the Osteoarthritis Initiative (OAI) cohort to validate the accuracy of longitudinal predictions acquired by DyHRM under different definition criteria of KOA incidence and progression. The basic finding of the experiments is that the larger the historic depth, the higher the accuracy of the obtained forecasts ahead. Comparative results demonstrate the efficacy of DyHRM against other state-of-the-art methods in this field. Full article

This study addresses the nonlinear estimation problem in the strapdown inertial navigation system (SINS) and Doppler velocity log (DVL) integrated navigation by proposing an improved filtering algorithm based on $S{E}_2(3)$ Lie group state representation. A dynamic model satisfying the group affine condition is established to systematically construct both left-invariant and right-invariant error state spaces, upon which two nonlinear filtering approaches are developed. Although the fast unscented transformation method is not novel by itself, its first integration with the $S{E}_2(3)$ Lie group model for SINS/DVL integrated navigation represents a significant advancement. Experimental results demonstrate that under large misalignment angles, the proposed method achieves slightly lower attitude errors compared to linear approaches, while also reducing position estimation errors during dynamic maneuvers. The 12,000 s endurance test confirms the algorithm’s stable long-term performance. Compared with conventional unscented Kalman filter methods, the proposed approach not only reduces computation time by 90% but also achieves real-time processing capability on embedded platforms through optimized sampling strategies and hierarchical state propagation mechanisms. These innovations provide an underwater navigation solution that combines theoretical rigor with engineering practicality, effectively overcoming the computational efficiency and dynamic adaptability limitations of traditional nonlinear filtering methods. Full article

The exponential growth of multi-modal data in the real world poses significant challenges to efficient retrieval, and traditional single-modal methods are no longer suitable for the growth of multi-modal data. To address this issue, hashing retrieval methods play an important role in cross-modal retrieval tasks when referring to a large amount of multi-modal data. However, effectively embedding multi-modal data into a common low-dimensional Hamming space remains challenging. A critical issue is that feature redundancies in existing methods lead to suboptimal hash codes, severely degrading retrieval performance; yet, selecting optimal features remains an open problem in deep cross-modal hashing. In this paper, we propose an end-to-end approach, named Robust Supervised Deep Discrete Hashing (RSDDH), which can accomplish feature learning and hashing learning simultaneously. RSDDH has a hybrid deep architecture consisting of a convolutional neural network and a multilayer perceptron adaptively learning modality-specific representations. Moreover, it utilizes a non-redundant feature selection strategy to select optimal features for generating discriminative hash codes. Furthermore, it employs a direct discrete hashing scheme (SVDDH) to solve the binary constraint optimization problem without relaxation, fully preserving the intrinsic properties of hash codes. Additionally, RSDDH employs inter-modal and intra-modal consistency preservation strategies to reduce the gap between modalities and improve the discriminability of learned Hamming space. Extensive experiments on four benchmark datasets demonstrate that RSDDH significantly outperforms state-of-the-art cross-modal hashing methods. Full article

In recent years, electrocardiogram (ECG) biometrics has received extensive attention and achieved a series of exciting results. In order to achieve optimal ECG biometric recognition, it is crucial to effectively process the original ECG signals. However, most existing methods only focus on extracting features from one-dimensional time series, limiting the discriminability of individual identification to some extent. To overcome this limitation, we propose a novel framework that integrates dual-level features, i.e., 1D (time series) and 2D (relative position matrix) representations, through collaborative embedding, dimensional attention weight learning, and projection matrix learning. Specifically, we leverage collective matrix factorization to learn the shared latent representations by embedding dual-level features to fully mine these two kinds of features and preserve as much information as possible. To further enhance the discrimination of learned representations, we preserve the diverse information for different dimensions of the latent representations by means of dimensional attention weight learning. In addition, the learned projection matrix simultaneously facilitates the integration of dual-level features and enables the transformation of out-of-sample queries into the discriminative latent representation space. Furthermore, we propose an effective and efficient optimization algorithm to minimize the overall objective loss. To evaluate the effectiveness of our learned latent representations, we conducted experiments on two benchmark datasets, and our experimental results show that our method can outperform state-of-the-art methods. Full article

Functional near-infrared spectroscopy (fNIRS) offers a safe and portable signal source for brain–computer interface (BCI) applications, particularly in motor imagery (MI) decoding. However, its low sampling rate and hemodynamic delay pose challenges for temporal modeling and dynamic brain network analysis. To address these limitations in temporal dynamics, static graph modeling, and feature fusion interpretability, we propose TopoTempNet, an innovative topology-enhanced temporal network for biomedical signal decoding. TopoTempNet integrates multi-level graph features with temporal modeling through three key innovations: (1) multi-level topological feature construction using local and global functional connectivity metrics (e.g., connection strength, density, global efficiency); (2) a graph-modulated attention mechanism combining Transformer and Bi-LSTM to dynamically model key connections; and (3) a multimodal fusion strategy uniting raw signals, graph structures, and temporal representations into a high-dimensional discriminative space. Evaluated on three public fNIRS datasets (MA, WG, UFFT), TopoTempNet achieves superior accuracy (up to 90.04% ± 3.53%) and Kappa scores compared to state-of-the-art models. The ROC curves and t-SNE visualizations confirm its excellent feature discrimination and structural clarity. Furthermore, the statistical analysis of graph features reveals the model’s ability to capture task-specific functional connectivity patterns, enhancing the interpretability of decoding outcomes. TopoTempNet provides a novel pathway for building interpretable and high-performance BCI systems based on fNIRS. Full article

Detecting financial fraud is a critical aspect of modern intelligent financial systems. Despite the advances brought by deep learning in predictive accuracy, challenges persist—particularly in capturing complex, high-dimensional nonlinear features. This study introduces a novel hybrid quantum recurrent neural network for fraud detection (HQRNN-FD). The model utilizes variational quantum circuits (VQCs) incorporating angle encoding, data reuploading, and hierarchical entanglement to project transaction features into quantum state spaces, thereby facilitating quantum-enhanced feature extraction. For sequential analysis, the model integrates a recurrent neural network (RNN) with a self-attention mechanism to effectively capture temporal dependencies and uncover latent fraudulent patterns. To mitigate class imbalance, the synthetic minority over-sampling technique (SMOTE) is employed during preprocessing, enhancing both class representation and model generalizability. Experimental evaluations reveal that HQRNN-FD attains an accuracy of 0.972 on publicly available fraud detection datasets, outperforming conventional models by 2.4%. In addition, the framework exhibits robustness against quantum noise and improved predictive performance with increasing qubit numbers, validating its efficacy and scalability for imbalanced financial classification tasks. Full article

As human–robot shared service environments become increasingly common, autonomous navigation in narrow space environments (NSEs), such as indoor corridors and crosswalks, becomes challenging. Mobile robots must go beyond reactive collision avoidance and interpret surrounding risks to proactively select safer routes in dynamic and spatially constrained environments. This study proposes a deep reinforcement learning (DRL)-based navigation framework that enables mobile robots to interact with pedestrians while identifying and traversing open and safe spaces. The framework fuses 3D LiDAR and RGB camera data to recognize individual pedestrians and estimate their position and velocity in real time. Based on this, a human-aware occupancy map (HAOM) is constructed, combining both static obstacles and dynamic risk zones, and used as the input state for DRL. To promote proactive and safe navigation behaviors, we design a state representation and reward structure that guide the robot toward less risky areas, overcoming the limitations of traditional approaches. The proposed method is validated through a series of simulation experiments, including straight, L-shaped, and cross-shaped layouts, designed to reflect typical narrow space environments. Various dynamic obstacle scenarios were incorporated during both training and evaluation. The results demonstrate that the proposed approach significantly improves navigation success rates and reduces collision incidents compared to conventional navigation planners across diverse NSE conditions. Full article

Industrial Control Systems (ICSs) are fundamental to critical infrastructure, yet they face increasing cybersecurity threats, particularly data integrity attacks like replay and data forgery attacks. Traditional IT-centric security measures are often inadequate for the Operational Technology (OT) environment due to stringent real-time and reliability requirements. This paper proposes an endogenous security defense mechanism based on the Luenberger observer and residual analysis. By embedding a mathematical model of the physical process into the control system, this approach enables real-time state estimation and anomaly detection. We model the ICS using a linear state-space representation and design a Luenberger observer to generate a residual signal, which is the difference between the actual sensor measurements and the observer’s predictions. Under normal conditions, this residual is minimal, but it deviates significantly during a replay attack. We formalize the system model, observer design, and attack detection algorithm. The effectiveness of the proposed method is validated through a simulation of an ICS under a replay attack. The results demonstrate that the residual-based approach can detect the attack promptly and effectively, providing a lightweight yet robust solution for enhancing ICS security. Full article

This study proposes a novel framework, Mamba-DQN, which integrates the state space-based time-series encoder Mamba-SSM into the Deep Q-Network (DQN) architecture to improve reinforcement learning performance in dynamic environments. Conventional reinforcement learning models primarily rely on instantaneous state information, limiting their ability to effectively capture temporal dependencies. To address this limitation, the proposed Mamba-DQN generates latent representations that summarize temporal information from state sequences and utilizes them for both Q-value estimation and Prioritized Experience Replay (PER), thereby enhancing the adaptability of policy learning and improving sample efficiency. The Mamba-SSM offers linear computational complexity and is optimized for parallel processing, enabling real-time learning and policy updates even in environments characterized by high state transition rates. The effectiveness of the proposed framework was validated through experiments conducted in environments with strong temporal dependencies and sparse rewards. Experimental results demonstrate that Mamba-DQN achieves superior stability and efficiency in policy learning compared to conventional DQN, LSTM-DQN, and Transformer-DQN models. Full article

To address the challenges of missed and incorrect segmentation in cabbage head detection under complex field conditions using UAV-based low-altitude remote sensing, this study proposes U-MoEMamba, an innovative dynamic state-space framework with a mixture-of-experts (MoE) collaborative segmentation network. The network constructs a dynamic multi-scale expert architecture, integrating three expert paradigms—multi-scale convolution, attention mechanisms, and Mamba pathways—for efficient and accurate segmentation. First, we design the MambaMoEFusion module, a collaborative expert fusion block that employs a lightweight gating network to dynamically integrate outputs from different experts, enabling adaptive selection and optimal feature aggregation. Second, we propose an MSCrossDualAttention module as an attention expert branch, leveraging a dual-path interactive attention mechanism to jointly extract shallow details and deep semantic information, effectively capturing the contextual features of cabbages. Third, the VSSBlock is incorporated as an expert pathway to model long-range dependencies via visual state-space representation. Evaluation on datasets of different cabbage growth stages shows that U-MoEMamba achieves an mIoU of 89.51% on the early-heading dataset, outperforming SegMamba and EfficientPyramidMamba by 3.91% and 1.4%, respectively. On the compact heading dataset, it reaches 91.88%, with improvements of 2.41% and 1.65%. This study provides a novel paradigm for intelligent monitoring of open-field crops. Full article

Current agentic AI frameworks such as LangGraph and AutoGen simulate autonomy via sequential prompt chaining but lack true multi-agent coordination architectures. These systems conflate semantic reasoning with orchestration, requiring LLMs at every coordination step and limiting scalability. By contrast, TB-CSPN (Topic-Based Communication Space Petri Net) is a hybrid formal architecture that fundamentally separates semantic processing from coordination logic. Unlike traditional Petri net applications, where the entire system state is encoded within the network structure, TB-CSPN uses Petri nets exclusively for coordination workflow modeling, letting communication and interaction between agents drive semantically rich, topic-based representations. At the same time, unlike first-generation agentic frameworks, here LLMs are confined to topic extraction, with business logic coordination implemented by structured token communication. This hybrid architectural separation preserves human strategic oversight (as supervisors) while delegating consultant and worker roles to LLMs and specialized AI agents, avoiding the state-space explosion typical of monolithic formal systems. Our empirical evaluation shows that TB-CSPN achieves 62.5% faster processing, 66.7% fewer LLM API calls, and 167% higher throughput compared to LangGraph-style orchestration, without sacrificing reliability. Scaling experiments with 10–100 agents reveal sub-linear memory growth (10× efficiency improvement), directly contradicting traditional Petri Net scalability concerns through our semantic-coordination-based architectural separation. These performance gains arise from the hybrid design, where coordination patterns remain constant while semantic spaces scale independently. TB-CSPN demonstrates that efficient agentic AI emerges not by over-relying on modern AI components but by embedding them strategically within a hybrid architecture that combines formal coordination guarantees with semantic flexibility. Our implementation and evaluation methodology are openly available, inviting community validation and extension of these principles. Full article

Rolling bearings are prone to failure, meaning that research on intelligent fault diagnosis is crucial in relation to this key transmission component in rotating machinery. The application of deep learning (DL) has significantly advanced the development of intelligent fault diagnosis. This paper proposes a novel method for rolling bearing fault diagnosis based on the fractional constant Q non-stationary Gabor transform (FCO-NSGT) and VMamba-Conv. Firstly, a rolling bearing fault experimental platform is established and the vibration signals of rolling bearings under various working conditions are collected using an acceleration sensor. Secondly, a kurtosis-to-entropy ratio (KER) method and the rotational kernel function of the fractional Fourier transform (FRFT) are proposed and applied to the original CO-NSGT to overcome the limitations of the original CO-NSGT, such as the unsatisfactory time–frequency representation due to manual parameter setting and the energy dispersion problem of frequency-modulated signals that vary with time. A lightweight fault diagnosis model, VMamba-Conv, is proposed, which is a restructured version of VMamba. It integrates an efficient selective scanning mechanism, a state space model, and a convolutional network based on SimAX into a dual-branch architecture and uses inverted residual blocks to achieve a lightweight design while maintaining strong feature extraction capabilities. Finally, the time–frequency graph is inputted into VMamba-Conv to diagnose rolling bearing faults. This approach reduces the number of parameters, as well as the computational complexity, while ensuring high accuracy and excellent noise resistance. The results show that the proposed method has excellent fault diagnosis capabilities, with an average accuracy of 99.81%. By comparing the Adjusted Rand Index, Normalized Mutual Information, F1 Score, and accuracy, it is concluded that the proposed method outperforms other comparison methods, demonstrating its effectiveness and superiority. Full article

Semiclassical (SC) approximations for various representations of a quantum state are constructed on a single (Lagrangian) surface in the phase space but such surface is not available for chaotic systems. An analogous evolution surface underlies SC representations of the evolution operator, albeit in a doubled phase space. Here, it is shown that corresponding to the Fourier transform on a unitary operator, represented as a Green function or spectral Wigner function, a Legendre transform generates a resolvent surface as the classical basis for SC representations of the resolvent operator in the double-phase space, independently of the integrable or chaotic nature of the system. This surface coincides with derivatives of action functions (or generating functions) depending on the choice of appropriate coordinates, and its growth departs from the energy shell following trajectories in the double-phase space. In an initial study of the resolvent surface based on its caustics, its complex nature is revealed to be analogous to a multidimensional sponge. Resummation of the trace of the resolvent in terms of linear combinations of periodic orbits, known as pseudo orbits or composite orbits, provides a cutoff to the SC sum at the Heisenberg time. Here, it is shown that the corresponding actions for higher times can be approximately included within true secondary periodic orbits, in which heteroclinic orbits join multiple windings of relatively short periodic orbits into larger circuits. Full article

Skeleton-based action recognition is an important subject in deep learning. Graph Convolutional Networks (GCNs) have demonstrated strong performance by modeling the human skeleton as a natural topological graph, representing the connections between joints. However, most existing methods rely on non-adaptive topologies or insufficiently expressive representations. To address these limitations, we propose a Multi-view Topology Refinement Graph Convolutional Network (MTR-GCN), which is efficient, lightweight, and delivers high performance. Specifically: (1) We propose a new spatial topology modeling approach that incorporates two views. A dynamic view fuses joint information from dual streams in a pairwise manner, while a static view encodes the shortest static paths between joints, preserving the original connectivity relationships. (2) We propose a new MultiScale Temporal Convolutional Network (MSTC), which is efficient and lightweight. (3) Furthermore, we introduce a new temporal topology strategy by modeling temporal frames as a graph, which strengthens the extraction of temporal features. By modeling the human skeleton as both a spatial and a temporal graph, we reveal a topological symmetry between space and time within the unified spatio-temporal framework. The proposed model achieves state-of-the-art performance on several benchmark datasets, including NTU RGB + D (XSub: 92.8%, XView: 96.8%), NTU RGB + D 120 (XSub: 89.6%, XSet: 90.8%), and NW-UCLA (95.7%), demonstrating the effectiveness of our GCN module, TCN module, and overall architecture. Full article