Search Results (4,484)

Road target detection is a crucial aspect of current research in automotive advanced driver assistance systems and intelligent transportation systems, where accuracy, speed, and lightweight design are key considerations. Compared to various sensors employed in driving assistance systems, millimeter-wave radar offers advantages such as all-weather operation, low hardware cost, strong penetration capability, and the ability to extract rich spatial information about targets. This paper tackles the challenges posed by the characteristics of Range-Angle map data from 77 GHz Frequency-Modulated Continuous Wave radar—namely, non-visible light imagery, abstract representation, rich fine details, and overlapping features. To this end, this paper proposes MMPFNet, a lightweight model based on the hypergraph mechanism with attention enhancement, as an extension of YOLOv13. First, an M-DSC3k2 module is proposed based on the hypergraph mechanism to enhance attention toward small targets. Second, a detection head with a double-bottleneck inverted MBConv-block structure is designed to improve the model’s accuracy and generalization capability. Third, a lightweight PPLConv module is customized to transform the backbone network, enhancing the model’s lightweight design while slightly reducing its accuracy. Considering the differences from traditional visible light datasets, the Focus Expansion-IoU loss function is introduced into the model to focus attention on different regression samples. The MMPFNet model achieves significant improvements in detecting common road targets such as pedestrians, bicycles, cars, and trucks on the Frequency-Modulated Continuous Wave radar Range-Angle dataset compared to the baseline YOLOv13n model: mAP50-95 increases by 16%, precision improves by 6%, and recall rises by 8.7%. MMPFNet is also evaluated on other non-visible light datasets such as CRUW-ONRD and soundprint datasets. Compared to commonly used detection models like FCOS and RetinaNet, MMPFNet achieves significant performance gains, attaining state-of-the-art results. Full article

Cloud detection in remote sensing imagery is a research hotspot in the field of image processing, and accurately detecting and segmenting cloud regions is crucial for improving the utilization efficiency of remote sensing data. However, standard convolutional neural networks face limitations in modeling the complex spatial structures of clouds. To address these challenges, this paper proposes a cloud detection method based on DDS-DeeplabV3+. First, a lightweight design of the Xception network is adopted to control model complexity, and part of its standard convolutional layers are replaced with Deformable Convolutional Networks (DCN), which enhances the capability of the model to capture geometric features of irregular cloud formations. Second, a Dual-Branch Collaborative Mechanism (DCM) that integrates global context modeling with local detail perception is designed to reconstruct the Atrous Spatial Pyramid Pooling (ASPP) module, thereby improving performance in handling complex scenes and fine boundary delineation. Finally, the SimAM (Simple, Parameter-Free Attention Module) is incorporated into the decoder module, enhancing thin cloud detection capability. Experimental results on the Landsat-8 and GF-1 datasets show that the proposed model achieves Mean Intersection over Union (MIoU) values of 92.61% and 94.04%, respectively, outperforming other comparative methods and demonstrating its superior performance in cloud detection tasks. Full article

Background: Alzheimer disease (AD) is a neurodegenerative condition that progressively develops structural changes in the brain, resulting in different stages of severity, which makes accurate multiclass classification from magnetic resonance imaging (MRI) challenging. Despite promising outcomes of deep learning models, a great number of current methods disregard disease progression, suffer from evaluation leakage, or lack interpretability. Objectives: This paper introduces DeepAttentionADNet, a lightweight hybrid CNN–Transformer framework designed for multiclass staging of Alzheimer’s disease using MRI images. Methods: The proposed model integrates convolutional feature extraction with transformer-based global context modeling. To capture the ordered nature of disease severity, a progression-aware ordinal learning objective is proposed. Moreover, consistency regularization is utilized to enhance robustness by imposing consistent prediction with spatial perturbation. A leakage-free k-fold cross-validation protocol is adopted, in which data splitting is performed prior to augmentation. Also, to promote interpretability, token-level importance maps based on transformer embeddings are utilized to visualize spatial regions that were used to make classification decisions. Results: The experimental findings on a multiclass MRI dataset of Alzheimer disease demonstrate consistent and high performance across cross-validation folds (mean F1-score (0.991 ± 0.003) and AUROC (0.9998 ± 0.0002)), without losing transparency and progress awareness. Conclusions: The proposed framework provided a robust and interpretable method of Alzheimer disease severity classification using MRI. Full article

To develop and design new alloy materials with lightweight and superior comprehensive performance. In this study, a MgAlTiVFeCo lightweight high-entropy alloy (LW-HEA) was fabricated via mechanical alloying and spark plasma sintering (SPS) to investigate the effects of sintering temperature on its phase structure, microstructure, densification, microhardness, high-temperature oxidation resistance, and corrosion resistance. The results indicate that the ball-milled MgAlTiVFeCo LW-HEA formed a simple solid solution phase with a BCC structure. After spark plasma sintering, the phase structure of the alloy changed, maintaining the BCC phase as the primary phase while accompanying the precipitation of secondary phases. When the sintering temperature reached 1000 °C, the alloy achieved a densification of 96.7% and a microhardness of 1235.5 HV. Its hardness value is comparable to the typical range of cemented carbides, demonstrating outstanding mechanical properties. The oxidation kinetics of MgAlTiVFeCo high-entropy alloys sintered at different temperatures at 900 °C follow a parabolic law, which is diffusion-controlled and can be divided into two stages: rapid growth and slow stabilization. At a sintering temperature of 1000 °C, the fitted oxidation rate constants, k_p1 (0–25 h) and k_p2 (25–60 h), are 3.76 × 10⁻² mg²·cm⁻⁴·s⁻¹ and 1.10 × 10⁻¹ mg²·cm⁻⁴·s⁻¹, respectively, outperforming those of alloys sintered at other temperatures. In a 3.5 wt% NaCl solution, the corrosion resistance of the alloy improves with increasing sintering temperature. Compared to alloys sintered at medium-to-low temperatures (850–950 °C), the alloy sintered at a high temperature (1000 °C) exhibits a more positive corrosion potential (−0.438 V) and a lower corrosion current density (1.07 × 10⁻⁶ A·cm⁻²), indicating excellent corrosion resistance. It is evident that 1000 °C is the optimal sintering temperature, and the MgAlTiVFeCo LW-HEA demonstrates superior comprehensive properties. Full article

Natural fibre-reinforced composites (NFCs) have attracted attention as sustainable alternatives to synthetic fibre composites. However, their hydrophilic nature and susceptibility to moisture absorption, especially in combination with process-related defects, can compromise long-term performance. This study critically examines the effects of hydrophobic fumed silica, incorporated into an epoxy matrix, on the processing, moisture uptake, and mechanical properties of flax/epoxy laminates produced via resin transfer moulding (RTM). Epoxy systems containing 0–5 wt% silica were characterised in terms of particle dispersion, rheological properties, thermal behaviour, and water absorption. Corresponding laminates were analysed for void content, Fickian diffusion behaviour, and tensile performance in dry and saturated states. Despite its hydrophobic surface treatment, silica increased resin water uptake and, at 5 wt%, led to a substantial rise in viscosity, poor fibre impregnation, and increased porosity. The resulting laminates exhibited faster and higher moisture uptake and significantly reduced wet mechanical properties, especially for highly filled systems. While thermal stability improved slightly, the overall findings revealed that the chosen silica-based matrix modification led to clear trade-offs and processing limitations under RTM conditions. This study highlights the importance of assessing such limitations early in the design process and demonstrates that the selected silica type is not a viable strategy for improving moisture resistance in NFCs. Full article

Hypersonic vehicles impose stringent requirements on lightweight structures to maintain mechanical integrity under extreme thermal environments. Bismaleimide (BMI)-based carbon fiber-reinforced polymer (CFRP) composites, featuring a high glass transition temperature and excellent thermal stability, are regarded as promising candidates for such applications. However, the high curing temperature and narrow processing window of BMI resins make it challenging to manufacture stiffened-shell structures with low defect levels and high fiber volume fractions. In this study, an integrated manufacturing route—hot-melt prepregging–filament winding–matched-metal mold forming—is proposed, and the key processing parameters are optimized via single-factor experiments and the Box–Behnken response surface methodology. The tensile strength of the laminate is selected as the response variable to evaluate the effects of the compression displacement (A), thermal consolidation/bonding temperature (B), heating rate (C), and cooling rate (D). The results reveal a unimodal dependence of the tensile strength on each parameter, with the significance ranking B > D > A > C; moreover, the A–B and A–D interactions are significant (p < 0.01). The established quadratic regression model exhibits good agreement with experimental data (R² = 0.974; R²_adj = 0.949). The predicted optimum conditions are A = 0.07 mm, B = 114.93 °C, C = 1.35 °C·min⁻¹, and D = 4.58 °C·min⁻¹, corresponding to a predicted tensile strength of approximately 2287 MPa. Validation experiments yielded 2291 MPa, in excellent agreement with the prediction. Microstructural observations indicate tight interlaminar bonding and a pronounced reduction in voids under the optimized conditions. Applying the optimized process to fabricate stiffened-shell demonstrators achieves a fiber volume fraction of >60% and a void content of <1%. This work provides a quantitatively defined processing window and parameter optimization basis for the high-quality manufacturing of BMI-CFRP stiffened-shell structures, with significant engineering relevance. Full article

Context: Code recommendation plays an important role in improving programming efficiency and software quality. Existing approaches mainly focus on method- or API-level recommendations, which limits their effectiveness to local code contexts. From a multi-stage recommendation perspective, class-level code recommendation aims to efficiently narrow a large candidate code space while preserving essential structural information. Objective: This paper proposes RioCC, a class-level code recommendation framework that leverages deep forest-based code clone detection to progressively reduce the candidate space and improve recommendation efficiency in large-scale code spaces. Method: RioCC models the recommendation process as a coarse-to-fine candidate reduction procedure. In the coarse-grained stage, a quick search-based filtering module performs rapid candidate screening and initial similarity estimation, effectively pruning irrelevant candidates and narrowing the search space. In the fine-grained stage, a deep forest-based analysis with cascade learning and multi-grained scanning captures context- and structure-aware representations of class-level code fragments, enabling accurate similarity assessment and recommendation. This two-stage design explicitly separates coarse candidate filtering from detailed semantic matching to balance efficiency and accuracy. Results: Experiments on a large-scale dataset containing 192,000 clone pairs from BigCloneBench and a collected code pool show that RioCC consistently outperforms state-of-the-art methods, including CCLearner, Oreo, and RSharer, across four types of code clones, while significantly accelerating the recommendation process with comparable detection accuracy. Conclusions: By explicitly formulating class-level code recommendation as a staged retrieval and refinement problem, RioCC provides an efficient and scalable solution for large-scale code recommendation and demonstrates the practical value of integrating lightweight filtering with deep forest-based learning. Full article

Aluminum–lithium (Al-Li) alloys are extensively employed in aerospace and space structures because of their low density, high specific stiffness, and excellent fatigue resistance. However, welding of these alloys remains challenging, since the joints typically exhibit unique microstructural features, including equiaxed grain zones (EQZ) along the fusion boundary and coarse columnar grains in the fusion zone, which degrade mechanical performance and increase susceptibility to cracking. This review provides an overview of the generational evolution of Al-Li alloys and their associated weldability, highlights the advantages and limitations of major welding processes, such as laser, arc, and hybrid techniques, and systematically examines the formation mechanisms of EQZ, columnar grains, and equiaxed grain bands. Various strategies for microstructural control are compared, including filler design, pulsed current, and external-field-assisted welding. Special attention is given to grain refinement achieved through heterogeneous nucleation, dendrite fragmentation, and columnar-to-equiaxed transition. Finally, prospects for advanced microstructural control strategies are discussed, with the goal of achieving high-quality welds for next-generation lightweight structural applications. Full article

Wampee (Clausena lansium) harvesting currently relies heavily on manual labor, but automation is significantly hindered by clustered fruit growth patterns, small fruit sizes, and complex orchard backgrounds, which make accurate detection highly challenging. This study proposes Wampee-YOLO, a lightweight and high-precision model based on the YOLO11n architecture, specifically designed for real-time wampee detection in natural orchard environments. The proposed model integrates several architectural enhancements: the RFEMAConv module for expanded receptive fields, an AIFI module for improved small target interaction, and a C2PSA-MSCADYT structure to boost multi-scale adaptability. Additionally, a Triplet Attention mechanism strengthens multi-dimensional feature representation, while an AFPN-Pro2345 neck structure optimizes cross-scale feature fusion. Experimental results demonstrate that Wampee-YOLO achieves an mAP₅₀ of 90.3%, a precision of 92.1%, and F1 score of 87%. This represents a significant 3.4% mAP₅₀ improvement over the YOLO11n baseline, with a slight increase to 3.28 M parameters. Ablation studies further confirm that the AFPN-Pro2345 module provides the most substantial performance gain, increasing mAP₅₀ by 2.4%. The model effectively balances computational efficiency with detection accuracy. These findings indicate that Wampee-YOLO offers a robust and efficient visual detection solution suitable for deployment on resource-constrained edge devices in smart orchard applications. Full article

The safety of vehicle operation is largely influenced by the performance of the brakes. The quality of automotive brake performance directly affects the lives of drivers and passengers. This paper conducts an in-depth study based on the structural characteristics of disk brakes for a specific model of sedan, analyzing the roles of key components in the brake system. Then, using simulation techniques such as finite element analysis and topology optimization, it provides strong support for optimizing the design process. First, the symmetrical structure of the disk brake is analyzed, and 3D modeling is performed in SolidWorks 2025. Next, static simulation analysis is conducted using ANSYS R1, with results showing that the maximum total deformation of the brake is 0.038 mm (not strain), and the maximum stress is 155.78 MPa, which meets the requirements for emergency braking. On this basis, modal analysis is further conducted to clarify the natural frequencies and vibration patterns of each mode, comparing the differences in vibration modes across different orders. Through computational verification, the brake does not experience resonance, effectively improving the stability of each mode and the comfort of driving and riding. Finally, the variable-density method enabled 10.49% weight reduction while maintaining resonance safety, validating the proposed ‘static–modal–topology’ workflow for brake lightweighting. Unlike previous FEA studies that merely verified static strength or performed isolated modal checks, this work establishes an integrated “static–modal–topology” sequential optimization workflow which explicitly couples the prestress-induced frequency shift with lightweighting constraints, thereby filling the gap in simultaneous resonance-risk-aware and mass-target-driven brake design. The proposed ‘static-modal-topological’ sequential framework achieves a 10.49% weight reduction rate, representing a 26.4% improvement over the 8.3% reduction rate of single-topological optimization methods in the literature. Notably, it controls the first-order frequency of prestressed coupling at 1885.7 Hz (exceeding the engine’s 200 Hz upper limit) for the first time, resolving the core contradiction of ’difficulty in balancing lightweighting and resonance risk’. Full article

Object detection in UAV aerial imagery presents significant challenges, including large-scale variations, complex background interference, object occlusion, and a high density of small targets. These factors restrict the generalization and localization capabilities of existing detectors. To address these issues, we propose YOLO-DMA, an efficient detection framework for aerial images. The framework incorporates three key improvements. First, we designed a Hierarchical Deformable Block (HDB), which uses adaptive sampling grids and a progressive multi-branch structure to capture features of irregular objects while preserving network depth, enabling richer hierarchical feature representation. Second, we proposed a Dual-Path Linear-complexity Perception (DPLP) module. One path employs a linear-complexity attention mechanism to model the global context efficiently, while the other utilizes lightweight convolutions to extract local details. This design effectively fuses shallow details with mid-level semantics, improving detection and localization accuracy. Third, we adopted the Wise-IoU v3 loss function, which dynamically adjusts optimization objectives, suppressing harmful gradients from low-quality samples and emphasizing small objects during training. Comprehensive experiments on the VisDrone dataset show that YOLO-DMA achieves 42.8% mAP50 and 25.7% mAP50:95. These correspond to improvements of 4.8% and 3.1% over YOLOv10. Experimental results demonstrate the effectiveness and practicality of the proposed framework. Full article

Austenitic Fe-Mn-Al-C lightweight steels have attracted considerable interest for automotive applications due to their exceptional specific strength, where κ-carbides precipitation critically influences mechanical properties. This review systematically examines the crystal structure, classification, and precipitation kinetics of κ-carbides, emphasizing their spatial distribution-dependent effects: coarse κ-carbides at austenite grain boundaries induce embrittlement and degrade toughness, while nanoscale κ’-carbides within grains enhance strength and ductility through dislocation interactions (e.g., Orowan bypassing and shearing), activating deformation mechanisms such as Dynamic Slip Band Refinement (DSBR), Shear Band-Induced Plasticity (SIP), and Microband-Induced Plasticity (MBIP). Thermodynamic calculations guide alloy design to ensure a single-phase austenite structure at the typical hot-rolling finishing temperature (~900 °C), avoiding harmful phases while promoting beneficial precipitates. Mn suppresses κ-carbide formation, whereas Al and C act as promoters, with intragranular κ’-carbides favoring higher Al/C concentrations (e.g., >6.2% Al and >1.0% C). Heat treatment parameters critically influence κ-carbide distribution, where rapid cooling (e.g., water quenching) suppresses κ-carbides, and subsequent aging (500–700 °C) enables homogeneous precipitation of κ’-carbides. Pre-deformation prior to annealing further accelerates κ-carbide nucleation by introducing crystal defects. Optimal performance requires integrated composition-processing-microstructure optimization to achieve a nnanoscaleκ’-carbide-strengthened austenite matrix through controlled composition and thermo-mechanical processing to achieve an optimal strength-ductility balance. Full article

Accurate and real-time glottis localization is critical for ensuring intraoperative oxygenation and patient safety during nasotracheal intubation. However, representative foundation models exemplified by the Segment Anything Model exhibit notable limitations in medical applications, stemming from their rigid attention mechanisms, feature space misalignment, and insufficient generalization to complex glottal anatomies. To address these challenges, we propose Glottis-SAM, a lightweight and task-adaptive segmentation framework that integrates dynamic representation learning with multi-prior contextual modeling. Specifically, we introduce a hierarchical low-rank adaptation strategy that enables efficient fine-tuning of visual foundation models by preserving geometric priors while significantly reducing computational overhead. To further enhance semantic fusion and generalization, we design a feature aggregation module with dual-path dynamic feature pyramids, which enables complementary optimization from local textures to global semantic structures under varying anatomical conditions. Extensive experiments on three diverse datasets demonstrate that Glottis-SAM achieves state-of-the-art segmentation accuracy with 72.6% mDice, a compact 55.2 MB model size, and 44.3 FPS inference speed on clinical data. These results highlight the model’s robustness, efficiency, and potential for deployment in visual guidance systems for nasotracheal intubation. Full article

Hydroforming has become an effective manufacturing technique for asymmetric corrugated thin-walled tubular components in lightweight automotive structures, owing to its capability to integrally form complex geometries. In this study, a finite-element model of the hydroforming process for 316L stainless-steel asymmetric corrugated thin-walled tubes was established, and three representative internal pressure loading paths—pulsating, linear, and stepped—were investigated using the DYNAFORM/LS-DYNA platform. The effects of different loading paths on material flow behavior, strain evolution, and forming quality, particularly wall-thickness distribution, were systematically compared. Among the three loading strategies, the pulsating pressure path exhibited the most balanced forming performance for asymmetric thin-walled tubes in terms of overall forming quality and wall-thickness control, although limited forming stability was observed in the initial pulsation scheme. To address this limitation, a dual-window orthogonal pulsation strategy was employed to optimize the initial pulsating loading path and further enhance its forming performance. The optimized pulsating curve completely eliminated the wrinkling tendency in the corrugated regions and reduced the maximum wall-thickness thinning ratio from 21.8% to 19.6%. Furthermore, the numerical simulation results show good agreement with experimental observations, with both the average wall-thickness deviation and the minimum wall-thickness error calculated using the interpolation method remaining within 2%. These results confirm the effectiveness of the optimized pulsating loading path for the hydroforming process design of asymmetric corrugated thin-walled tubes. Full article

Synthetic Aperture Radar (SAR) is a powerful observation system capable of delivering high-resolution imagery under variable sea conditions to support target detection and tracking, such as for ships. However, conventional optical target detection models are typically engineered for complex optical imagery, leading to limitations in accuracy and high computational resource consumption when directly applied to SAR imagery. To address this, this paper proposes a lightweight shape-aware and direction-weighted algorithm for SAR ship detection, SADW-Det. First, a lightweight streamlined backbone network, LSFP-NET, is redesigned based on the YOLOX architecture. This achieves reduced parameter counts and computational burden by incorporating depthwise separable convolutions and factorized convolutions. Concurrently, a parallel fusion module is designed, leveraging multiple small-kernel depthwise separable convolutions to extract features in parallel. This approach maintains accuracy while achieving lightweight processing. Furthermore, addressing the differences between SAR imagery and other imaging modalities, a direction-weighted attention was devised. This enhances model performance with minimal computational overhead by incorporating positional information while preserving channel data. Experimental results demonstrate superior detection accuracy compared to existing methods on three representative SAR datasets, SSDD, HRSID and DSSDD, while achieving reduced parameter counts and computational complexity, indicating strong application potential and laying the foundation for cross-modal applications. Full article