Search Results (164)

Anti-UAV multi-object tracking in remote sensing videos is challenging because UAV targets are small, weakly textured, and often affected by cluttered backgrounds, abrupt motion, occlusion, and intermittent visibility. To address these challenges, we formulate anti-UAV multi-object tracking as a context-aware identity prediction task, in which target identities and locations are inferred from historical trajectory priors instead of current-frame observations alone. Under this formulation, we propose a dual-track parallel tracking framework. The adaptive identity disambiguation (AID) module combines motion cues with appearance features according to their estimated reliability, improving short-term association when visual evidence is weak. In parallel, the motion-evolution temporal memory (METM) module models trajectory dynamics using motion anomaly detection and time-decayed memory, enabling spatiotemporal recovery after occlusion, temporary disappearance, or abrupt motion. The outputs of the two branches are integrated by a unified identity decision layer to produce stable tracking results. Experiments are conducted on the public 4th Anti-UAV Benchmark Track-3 and our newly constructed Anti-UAV Multi-Object Tracking dataset, AU-MOT. On the 4th Anti-UAV Benchmark Track-3, our method achieves 63.6% HOTA and 64.1% IDF1, outperforming the strongest competing method by 3.5% and 3.9%, respectively, while reducing identity switches and track fragments by 20.8% and 23.8%. On AU-MOT, it achieves 67.2% HOTA and 67.8% IDF1, with 20.2% fewer identity switches and 22.3% fewer track fragments. These results demonstrate its effectiveness under long-range observation, weak target appearance, cluttered backgrounds, abrupt motion, and intermittent target visibility. Full article

GNSS array receivers suffer tracking degradation under array nonidealities such as element-position perturbations, channel amplitude/phase errors, and slowly varying manifold mismatch. Conventional blind anti-jamming suppresses interference, but adaptive weight fluctuations can propagate into the correlator domain, increasing cross-branch correlation, causing Early/Late metric imbalance, and reducing Prompt phase consistency. Existing noncoherent combining methods mainly convert multi-branch correlator outputs into scalar energy metrics for code tracking, leaving the carrier loop’s complex Prompt input insufficiently constrained. To address this problem, we propose a blind adaptive joint code–carrier channel-combining method for nonideal arrays. After first-stage anti-jamming, the method estimates an Early/Late correlator-domain covariance matrix and reuses it as a shared statistical constraint. In the code loop, this matrix drives whitened noncoherent energy combining with closed-loop gain normalization to stabilize the DLL discriminator scale. In the carrier loop, it is combined with a Prompt-derived coherent direction to form a covariance-constrained PLL complex input. Simulations under wideband interference, static array errors, and dynamic mismatch show that the proposed J-WNCC reduces both code-phase error and carrier-phase jitter, improving joint tracking robustness in nonideal array environments. Ablation results further reveal a dominant-effect separation: DLL gain normalization mainly calibrates the whitened code-discriminator scale, whereas coherent Prompt combining mainly reconstructs the complex PLL input. Full article

Selective laser melting (SLM) equipment operates as a complex cyber–physical system, wherein strong implicit coupling among internal subsystems presents significant challenges for condition monitoring and fault diagnosis. Existing deep learning methods often suffer from feature submersion when processing multi-source heterogeneous data and lack the capability for system-level topological causal inference. To address these issues, we propose a multisystem coupling modeling and anomaly traceability method based on a decoupled graph attention network (ST-DBGAE). Independent local spatiotemporal feature alignment modules are constructed to map heterogeneous sensory data into a unified latent space. This eliminates dimensional discrepancies while strictly maintaining the feature independence of underlying hardware subsystems, such as optical and gas circuits. A dynamic graph attention mechanism with sparse priors is subsequently introduced to adaptively capture time-varying coupling weights triggered by implicit interactions (e.g., thermal fluids), bypassing the need for predefined rigid physical connections. Furthermore, a dual-branch two-stage decoupled optimization architecture is designed. By blocking the cross-interference of global backpropagation, this architecture outputs a continuous equipment health index (HI) based on reconstruction errors and employs a topological difference matrix inference mechanism to reversely anchor the root-cause nodes responsible for cross-system cascading degradation. Experimental results based on over 310,000 real operational monitoring records from industrial SLM equipment demonstrate that the proposed model achieves a comprehensive diagnostic Macro-F1 score of 96.5% across eight operating states. The single-class detection rates (ACCs) of specific underlying anomalies are significantly improved. This method not only enables high-precision equipment health warnings but also provides a physically interpretable microscopic fault propagation mapping for predictive maintenance. Full article

MD-Hill-SPN is the first Hill-based construction to combine a multi-tier diffusion mix layer, a memory-hard KDF, and a simultaneous multi-metric empirical evaluation. Two independent runs of the full metric suite yield: (a) full plaintext avalanche from round 1 (mean 63.97–64.67 of 128 bits, ideal 64); (b) the differential-probability sampling floor of 2 × 10⁻⁵ reached at round 4 (50,000 of 50,000 output differences distinct, both sessions); (c) algebraic-degree lower-bound saturation at the maximum observable value from round 1; (d) linear-bias indistinguishable from random (combined exceedance 4.40%, below the 4.55% noise floor); and (e) branch numbers at the Singleton (MDS) bound for every tier (B = 5 for 4 × 4, B = 9 for 8 × 8, B = 17 for 16 × 16), computed exhaustively over weight-1 inputs. MD-Hill-SPN therefore moves beyond theoretical construction to a construction that passes a defined empirical evaluation suite: avalanche, differential sampling, linear-bias probing, algebraic-degree lower bounds, and MDS branch numbers under single-key, known-plaintext conditions with fixed parameters, an evaluation no prior Hill cipher variant has reported in full. Full article

Remote sensing image change detection is widely used in urban expansion analysis, land-use monitoring, and disaster assessment. Nevertheless, it still faces significant challenges due to pseudo-change interference in high-resolution imagery, the large-scale variation in small changed objects, and the need for lightweight models in real-world engineering applications. To address these issues, this paper proposes LDST-ChangeNet, a lightweight dual spatiotemporal attention network for change detection. The network adopts a Siamese EfficientNet-B1 as its dual-branch encoder and employs a differential bi-temporal feature fusion strategy (Diff) to explicitly model temporal discrepancies, enabling efficient feature extraction while significantly reducing model complexity. A Position Attention Module (PAM) is introduced at the encoder bottleneck to suppress pseudo changes caused by non-structural factors. Meanwhile, a lightweight Pyramid Pooling Module (PPM-lite) is incorporated at the entrance of the deepest decoder features to enhance multi-scale contextual representation. Furthermore, a Boundary Attention Module (BAM) is applied in the decoder output stage to improve boundary delineation and small-object change detection. Experimental results on the LEVIR-CD and WHU-CD datasets show that LDST-ChangeNet outperforms other state-of-the-art methods, achieving F1-scores of 90.67% and 91.08%, respectively. The model maintains a lightweight design, requiring only 11.72 M parameters and 10.03 GFLOPs on LEVIR-CD, and 11.77 M parameters and 9.12 GFLOPs on WHU-CD. Full article

To improve the accuracy and stability of short-term traffic flow prediction in complex road networks and address the limitations of existing models in modeling spatiotemporal dependencies, this paper proposes a traffic flow prediction model based on a Hypergraph Spatio-Temporal Interaction Network (HGSTIN) in the context of intelligent transportation systems. The study constructs a multi-dimensional traffic pattern input tensor by integrating three temporal scales—proximity, intra-day, and intra-week—while taking traffic flow as the prediction target and introducing average speed and lane occupancy as auxiliary features. In terms of temporal modeling, a Transformer architecture integrated with a Dynamic Tanh (DyT) mechanism is adopted to capture multi-period variations. For spatial modeling, a neighborhood hypergraph and a DTW-based semantic hypergraph are combined to enhance the representation of local and global through spatial self-attention and hypergraph neural network branches, and an adaptive feature fusion module is designed to perform adaptively weighted fusion of the outputs from the two branches. In terms of loss function design, a temporal gradient consistency loss function is proposed to enhance the robustness of predictions. Experimental results on the PEMS04 and PEMS08 datasets show that the proposed model achieves average improvements of approximately 5.15%, 1.76%, and 3.88% in MAE, RMSE, and MAPE, respectively, compared to the second-best baseline model. The model exhibits the smallest performance degradation in multi-step prediction scenarios, and the effectiveness of each module is validated through ablation studies. The findings demonstrate that HGSTIN can effectively capture the dynamic spatiotemporal characteristics of complex traffic scenarios, thereby providing high-precision prediction support for intelligent transportation systems. Full article

Controllable symbolic music generation must preserve a reference melody while remaining responsive to style prompts. Existing hierarchical diffusion systems typically reuse a shared condition vector across harmony, rhythm, and timbre stages, which can entangle stylistic factors and weaken melody preservation. We present HCDMG++, a hierarchical diffusion framework that addresses these two limitations through stage-aware style routing and differentiable melody regularization. The routing module uses a residual multi-layer perceptron (MLP) with zero-initialized scalar gates to project text-derived style embeddings into harmony-, rhythm-, and timbre-specific subspaces, whereas the regularization branch aligns soft pitch histograms and contour trajectories with the conditioning melody during training without breaking the differentiable computation graph. We evaluate the integrated system on a 384-sample benchmark covering four melodies, eight styles, four random seeds, and three denoising budgets, supplemented by a matched legacy-compatible reference and inference-time component ablation that contrasts legacy behavior, silenced gates, an automated uniform gamma routing sweep, and the full forward pass. HCDMG++ produces valid four-track outputs in all 384 runs, reaches a peak pitch histogram similarity score of 0.508 under a 64-step budget, and improves pitch histogram alignment over Legacy-HCDMG by roughly two orders of magnitude on the matched slice, while attaining a positive Fisher-style style separability score where the legacy benchmark is too sparse to support one. These results indicate that stage-specific conditioning and differentiable structural guidance jointly improve controllability in symbolic music diffusion, while also exposing the remaining limitations in long-form generalization and perceptual validation, which motivate the future work outlined at the end of this paper. Full article

Cracks are among the most common pavement distresses, and their timely detection is crucial for road maintenance. Existing methods struggle to completely capture elongated and irregular cracks, often resulting in fragmented detection outputs, which leads to the inaccurate assessment of crack length and affects the reliability of pavement condition evaluation. To address this issue, this paper proposes a dual-path crack segmentation network that integrates CNN and Transformers. The CNN branch incorporates a dynamic multi-branch convolution module to enhance the directional perception and structural modeling of elongated cracks. The Transformer branch employs a lightweight DCNv4 module to replace traditional self-attention mechanisms, effectively capturing long-range dependencies while reducing computational complexity. A multi-path fusion module is designed to achieve the collaborative enhancement of dual-path features, improving the semantic representation of continuous crack regions. Additionally, a combined loss function of BCE and Dice is adopted to alleviate the severe class imbalance between crack and background pixels, further improving the completeness of crack segmentation. Experiments on four datasets, including CFD, DeepCrack537, Gaps384, and Crack500, demonstrate that the proposed model outperforms all compared methods in terms of F-score and mIoU. Ablation studies further validate the effectiveness of the dual-path architecture and its key modules in improving performance. Furthermore, in field validation on real road scenarios, the pavement condition index (PCI) calculated based on the proposed method shows an average deviation of only 0.81 compared to manually interpreted ground truth, demonstrating the practical value of continuous crack detection for pavement maintenance assessment. Full article

Building change detection from bi-temporal remote-sensing imagery underpins urban planning, infrastructure monitoring, and disaster assessment. Existing deep-learning methods achieve high accuracy but rely on large parameter counts, while pixel-level supervision provides limited boundary guidance. We propose POCA-lite, a lightweight encoder–decoder with an inference-coupled geometry branch: three geometric prediction heads—distance transform, boundary, and center heatmap—whose outputs are fused back into the decoder via a feedback pathway active at both training and inference. On the LEVIR-CD benchmark under a unified retraining protocol, multi-seed evaluation shows that POCA-lite matches SNUNet in mean F1 while using 47% fewer parameters and 53% fewer FLOPs. Boundary F1 improves by 9.22 pp over the no-geometry baseline. Decomposition ablations reveal two complementary improvement sources: geometric supervision alone recovers 85% of the total gain, while the feedback fusion pathway recovers 92%; their combination achieves the full result. Geometry-aware targets outperform a generic multitask control. Cross-architecture transfer to SNUNet yields +1.06 pp F1. However, cross-dataset evaluation on WHU-CD shows that the method underperforms SNUNet on dense urban morphology, and zero-shot cross-dataset transfer is not established. These results indicate that inference-coupled geometric supervision is effective for lightweight, boundary-sensitive change detection on domains with well-separated building morphology, but its applicability is scope-bounded. Full article

As a critical component of ice-snow tourism, land cover classification for ski resorts is crucial to ice-snow resource management. However, there is currently a scarcity of datasets and methods capable of high-precision mapping for such fine-grained scenarios. Although Transformers with long-sequence interactions and convolutional neural networks (CNNs) have emerged as mainstream solutions, their performance remains limited on high-resolution remote sensing data characterized by small datasets and high heterogeneity. Targeting land cover classification in ski resort areas, this study proposes a triple-branch segmentation framework integrating CNNs and Transformers to extract global, detail and boundary features (GDBNet), and constructs the first high-resolution ski resort land cover dataset with a resolution of 0.75 m using JiLin-1 satellite constellation (LULC_SKI). The framework employs a backbone combining SegFormer with dual CNN branches. SegFormer captures global semantic context, while dual ResNet-18 branches extract local semantics and edge details respectively. The neck integrates two specialized feature interaction modules, the proposed Pixel-Guided Feature Attention (PG-AFM) and Boundary-Guided Feature Attention (BG-AFM), which synergistically fuse these heterogeneous feature representations for enhanced multi-scale modeling. For the segmentation head, a multi-task learning approach supervises both semantic and edge outputs. LULC_SKI covers seven representative ski resorts in Jilin Province, China, comprising 10,000 multi-seasonal images annotated with six land cover classes, including roads, vegetation, built-up areas, ski runs, water bodies, and cropland. Experiments demonstrate GDBNet achieves 85.44% mIoU and 91.84% mF1 on LULC_SKI, outperforming other advanced models with particularly significant improvements for linear objects like roads and ski runs. Extensive experimental comparisons show that GDBNet delivers consistently excellent performance on both the iSAID and LoveDA datasets, underscoring the superiority of our proposed method. Ablation studies validate the effectiveness of the triple-branch architecture, attention modules, and multi-task supervision. This work proposes a modular framework for land cover classification in complex ski resort scenarios. Full article

This study presents a unified semantic segmentation framework for UAV-based multispectral crop stress mapping, focusing on the integration of water stress and rust disease conditions within a common label space. Unlike conventional approaches that address individual stress factors independently, the proposed framework harmonizes heterogeneous datasets with different annotation schemes into a single multi-class segmentation problem. To achieve this, UAV multispectral orthomosaics are processed using a patch-based strategy and a multi-head UNet++ architecture incorporating segmentation, edge-aware, and Signed Distance Transform (SDT) branches. In addition, a physics-informed output-space refinement module based on fractional partial differential equations (FPDE) is introduced to enhance spatial coherence and boundary preservation in the predicted maps. Experimental results demonstrate the effectiveness of the proposed framework within the evaluated dataset setting, particularly in terms of boundary delineation, spatial consistency, and minority-class detection. The study highlights the feasibility of integrating heterogeneous stress conditions into a unified segmentation framework and provides a foundation for future research on scalable multi-source agricultural monitoring systems. Full article

The nutation face gear transmission combines the high contact ratio of face gears with the large reduction ratio of nutation drives, making it a promising transmission solution. However, the natural characteristics of the dual-path system under multi-stiffness coupling remain insufficiently understood. A 22-degree-of-freedom bending–torsional–axial coupled dynamic model is established using the lumped parameter method, incorporating axial vibration of the face gears and torsional deformation of the input shaft. By solving the undamped free vibration equations, the natural frequencies, mode shapes and stiffness effects are obtained, and the key stiffness parameters governing lower-order modal behavior are identified. The results indicate that no rigid body modes exist, while structural symmetry leads to repeated frequencies in specific modes. The lower-order modes are dominated by torsional vibration of the input shaft. The intermediate modes are characterized by the coupled vibration of the local branch and the output face gear. The higher-order modes are dominated by the vibration of the nutation face gear and exhibit coupled characteristics. The results identify input-shaft torsional stiffness as a key parameter for lower-order modal tuning, while support stiffness mainly affects local modes and exhibits saturation behavior. This study provides theoretical guidance for resonance avoidance and load-sharing optimization. Full article

We introduce a quantum multiplexer (GHZ MUX) architecture that enables deterministic routing of an unknown qubit from a single sender to one of $2^n$ receivers using only local tripartite Greenberger–Horne–Zeilinger (GHZ) states arranged in a binary tree. At each level of the hierarchy, a Bell-basis measurement and classical feed-forward propagate the encoded quantum information along a selected branch while maintaining the appropriate Pauli correction frame. Unlike quantum routing architectures that rely on globally entangled multipartite states, the proposed design composes small GHZ clusters into a modular teleportation hierarchy that requires only local entanglement generation and coherence. This structure achieves full input–output connectivity while preserving deterministic routing control and experimental feasibility for near-term small-scale quantum networks. Beyond routing functionality, we show that the same GHZ-tree structure naturally supports hidden-destination communication. We formalize this extension as the Hidden-Secret GHZ-Tree Routing (HS-GTR) protocol, in which the final receiver remains unknown to external observers and the transmitted quantum state may optionally be protected by a quantum one-time pad. This construction demonstrates that hierarchical GHZ routing can serve not only as a quantum switching architecture but also as a building block for privacy-preserving communication and multi-receiver key establishment in distributed quantum networks. Full article

Reconfigurable intelligent surface sensors (RISs) have emerged as a promising technology for enhancing wireless indoor localization by intelligently controlling signal propagation; however, extracting reliable localization fingerprints from RIS-assisted signals remains challenging due to multipath fading, environmental noise, and nonlinear spatial–temporal channel dynamics. To address this, we propose an Adaptive Dual-Reinforcement Learning-Hybrid Spatial–Temporal Network (ADRL-HSTNet) for RIS-assisted indoor localization. The framework utilizes dual-channel RSSI and phase measurements, followed by noise filtering, normalization, and sliding-window segmentation prior to feature extraction. It then constructs enhanced representations through handcrafted feature extraction and multi-branch processing, including patch-based features, wavelet-domain representations, statistical descriptors, and multi-level segmentation masks. These heterogeneous inputs are encoded using lightweight transformer-based encoders to capture multiscale dependencies. A first reinforcement learning selector adaptively weights the most informative feature branches to produce a fused representation, which is further processed by spatial and temporal transformer modules. Their outputs are adaptively combined via a second reinforcement learning selector to obtain robust localization embedding. The model jointly performs classification, coordinate regression, and uncertainty estimation end-to-end. Experimental results across multiple RIS configurations outperformed the KAN, LSTM-KAN, and RHL-Net (compared against the proposed ADRL-HSTNet) baselines, achieving accuracies of 83.33%, 75.22%, 93.33%, and 88.89%, confirming the effectiveness of the proposed approach. Full article

Deep learning has substantially advanced the automated classification and segmentation of breast ultrasound images. However, many existing methods do not fully exploit task correlations, which weakens information exchange and limits the delineation of fine structures. In addition, commonly used loss functions often fail to balance classification and segmentation objectives effectively. To address these issues, we propose N-Unet, a multi-task learning framework that combines adaptive optimization with feature-enhancement modules. Specifically, the Adaptive Multi-Task Loss (AMTL) dynamically balances the two task objectives to promote stable joint learning. The Adaptive Feature Fusion (AFF) and Cross-Level Attention Enhancement (CLAE) modules improve feature representation through multi-scale integration and semantic refinement. The Conditional Segmentation Boosting (CSB) module further refines segmentation outputs according to the classification result, improving inference-stage consistency. Together, these components form a unified multi-task framework with a shared encoder, a segmentation branch, and an integrated classification branch whose output further supports segmentation-consistency refinement. Experiments on the BUSI and BUS-UCLM datasets demonstrate the superiority of N-Unet. The model achieves classification accuracies of 96.54% on BUSI and 95.83% on BUS-UCLM, with corresponding Dice scores of 80.70% and 92.16%. It reaches this performance with only 8.95 M parameters and 14.74 GFLOPs, showing a favorable performance-efficiency trade-off. These results confirm the effectiveness of N-Unet and its robustness across the two BUS datasets studied here, supporting its potential for practical breast nodule assessment, while broader external generalization remains to be validated. Full article