Search Results (42)

Accurate detection and reliable grasping of small bolts are essential for intelligent manufacturing and automated assembly. However, this remains a challenge due to the small size, slender geometry, and metallic reflective surfaces of bolts. In this paper, we propose a vision-guided robotic bolt handling framework that integrates lightweight object detection, optimization-oriented grasp execution, and collision-aware trajectory planning. The lightweight YOLOv8n-BoltLite detector, improved with E-C2f, LCA, SA-PAN, and WD-IoU loss, enhances localization accuracy and feature representation for small and slender bolts. A robotic grasping framework is designed to transform detection results into executable robotic actions through 3D pose estimation, mid-shank grasp point generation, and optimization-oriented execution formulation. Additionally, a five-segment trajectory planning strategy ensures safe and efficient robot motion. Experimental results show that YOLOv8n-BoltLite achieves a five-run average mAP of 99.64 ± 0.05% with 198 FPS, and 3.02 M parameters. On an additional challenging external test set involving illumination variation, clutter, partial occlusion, reflection, and clustered bolts, the proposed detector achieves 94.62 ± 0.18%, outperforming recent lightweight detectors under the same training protocol. Robotic experiments involving 1000 controlled grasping trials and 300 multi-target grasping attempts demonstrate a controlled-condition success rate of 97.0% and improved target-selection reliability in multi-bolt scenes. These results suggest that the proposed framework offers a practical and efficient solution for automated bolt handling in intelligent manufacturing environments. Full article

Linear assets (LAs), such as conveyor systems, road networks, pipelines, and power transmission lines, are typically maintained through localized, segment-based interventions. While such approaches effectively address spatially heterogeneous degradation, they often neglect the system-level consequences of repeated local actions. In particular, improvements in segment condition may be accompanied by increased structural complexity, leading to reduced reliability and higher lifecycle costs. This paper proposes a unified engineering framework that integrates segment-level condition assessment with system-level structural effects. The framework is based on a dual representation of asset condition, distinguishing between material state (MS) and structural state (SS), which correspond to material aging (MA) and structural aging (SA), respectively. A key contribution is the introduction of the fragmentation penalty (FP), capturing the negative impact of increasing segmentation and interface density on system performance. The framework incorporates multi-threshold decision logic, enabling differentiation between operational, refurbishment, and replacement regimes, and interprets maintenance actions as transformations affecting both condition and structure. A formal model is developed to represent the asset as a dynamic system of segments and interfaces. It provides a basis for future empirical calibration and structure-aware optimization. Although the model is developed using conveyor belt loops as a reference case, its broader relevance is discussed for other classes of linear assets with repeated local intervention and evolving structural heterogeneity. A simple worked example is included to demonstrate the operational meaning of the proposed fragmentation-aware perspective. The results show that maintenance decisions may change when structural side effects are considered together with local condition improvement, and they provide a basis for future empirical calibration and structure-aware optimization of maintenance strategies. Full article

Hypnosis has traditionally been conceptualized as a clinical technique for reducing physiological symptoms (e.g., pain, nausea) and psychological symptoms (e.g., anxiety, intrusive thoughts), yet emerging neuroscientific evidence suggests it operates through the fundamental mechanisms of emotional regulation and self-integration. This integrative review synthesizes research on clinical hypnosis from cognitive neuroscience, affective science, and clinical practice to examine how hypnotic phenomena modulate large-scale brain networks—particularly the default mode network (DMN), executive control network (ECN), and salience network (SaN)—to reorganize emotional experience and self-referential processing. We propose a formal mechanistic model in which hypnotic induction produces heightened experiential plasticity through coordinated network reconfiguration, enabling adaptive emotion regulation and reduced dissociative fragmentation. Central to this framework is the construct of Fundamental Peace (FP), operationalized as a dynamic neuro-experiential state characterized by: (1) flexible attentional control without effortful suppression; (2) emotional coherence across self-states; (3) reduced self-referential rigidity; (4) compassionate self-awareness. Unlike equanimity (affective neutrality) or well-being (positive evaluation), Fundamental Peace represents integrated regulatory capacity under changing conditions. Key findings from neuroimaging studies demonstrate that hypnotic states consistently reduce DMN activity, enhance ECN-SaN coupling, and modulate connectivity patterns associated with self-referential processing. Meta-analytic evidence from 85 controlled experimental trials shows robust pain reduction effects, while clinical studies document improvements in trauma-related dissociation and emotional dysregulation. We critically evaluate this framework against alternative theories (dissociated control, cold control, predictive processing, social-cognitive models), specify testable predictions, and assess evidence quality across neuroimaging and clinical domains. Implications for trauma treatment, clinical implementation, and future research integrating causal inference methods are discussed, alongside ethical and cultural considerations. Full article

We demonstrate a monolithically integrated photonic chip that combines an optical spiking neuron with a tunable synaptic element. The spiking neuron is realized using a quantum-well Fabry–Perot laser integrated with a saturable absorber (FP-SA), while a semiconductor optical amplifier (SOA) functions as a photonic synapse. Two photonic-crystal (PC) mirrors define the laser cavity and enable effective modulation of the synaptic weight. Experimental results further confirm the capability of the SOA for continuous and controllable synaptic weight tuning. This work represents an important step toward scalable on-chip photonic spiking neural networks. Full article

Tracking small, fast-moving unmanned aerial vehicles (UAVs) in thermal infrared (TIR) imagery is a significant challenge due to low-resolution targets, Dynamic Background Clutter, and frequent occlusions. To address this, we introduce MemLoTrack, a novel onestream Vision Transformer tracker that integrates a memory mechanism into a parameterefficient LoRA framework. MemLoTrack enhances a baseline tracker (LoRAT) with two key components: (i) a gated First-In, First-Out (FIFO) memory bank (MB) for temporal context aggregation and (ii) a lightweight Memory Attention Layer (MAL) for effective information retrieval. A key component of our method is a selective memory update policy, which commits a frame to the memory bank only when it satisfies both a classification confidence threshold ($\tau$) and a Kalman filter-based motion consistency check. This gating mechanism robustly prevents memory contamination due to distractors, occlusions, and reappearance events. Our training is highly efficient, updating only the LoRA adapters, MAL, and prediction head while the pretrained DINOv2 backbone remains frozen. Evaluated on the challenging Anti-UAV410 benchmark, MemLoTrack (L_mem = 7, $\tau$ = 0.8) achieves an AUC of 63.6 and a State Accuracy (SA) of 64.0, representing a significant improvement over the LoRAT baseline by +1.4 AUC and +1.5 SA. Compared to the state-of-the-art method FocusTrack, MemLoTrack demonstrates superior robustness with higher AUC (63.6 vs. 62.8) and SA (64.0 vs. 63.9), while trading lower precision (P/P-Norm) scores. Furthermore, MemLoTrack operates at 153 FPS on a single RTX 4070 Ti SUPER, demonstrating that parameter-efficient fine-tuning with a selective memory mechanism is a powerful and deployable strategy for real-time Anti-UAV tracking in demanding TIR environments. Full article

This paper presents a Doppler lidar system based on a mode-locked semiconductor laser (ML-SL) source. The ML-SL consists of two sections: a Fabry–Pérot (F-P) cavity and a saturable absorber (SA) region. The system utilizes multiple phase-correlated modes of the optical frequency comb to acquire multiple Doppler shift signals; through cross-referencing of these signals, the robustness of the velocimetry system is enhanced. Experimental validation of precise velocity measurements for moving objects was conducted within the speed range of 0.005 m/s to 0.5 m/s. For target speeds of 0.563 m/s and 0.00563 m/s, the maximum and minimum absolute errors were 0.00064 m/s and 0.00003 m/s, respectively, with relative errors consistently below 1%. Comparative experiments demonstrated that utilizing multiple comb teeth reduces the maximum absolute error from 0.001286 m/s (observed when using a single tooth) to 0.000833 m/s. Furthermore, the velocity resolution of the system was analyzed: a frequency resolution of 30 Hz corresponds to a velocity resolution of 0.1117 m/s, while improving the frequency resolution to 1 Hz yields a velocity resolution of 0.0037 m/s. Full article

3D object detection is a fundamental task in autonomous driving. In recent years, voxel-based methods have demonstrated significant advantages in reducing computational complexity and memory consumption when processing large-scale point cloud data. A representative method, Voxel-RCNN, introduces Region of Interest (RoI) pooling on voxel features, successfully bridging the gap between voxel and point cloud representations for enhanced 3D object detection. However, its robustness deteriorates when detecting distant objects or in the presence of noisy points (e.g., traffic signs and trees). To address this limitation, we propose an enhanced approach named Self-Attention Voxel-RCNN (SA-VoxelRCNN). Our method integrates two complementary attention mechanisms into the feature extraction phase. First, a full self-attention (FSA) module improves global context modeling across all voxel features. Second, a deformable self-attention (DSA) module enables adaptive sampling of representative feature subsets at strategically selected positions. After extracting contextual features through attention mechanisms, these features are fused with spatial features from the base algorithm to form enhanced feature representations, which are subsequently input into the region proposal network (RPN) to generate high-quality 3D bounding boxes. Experimental results on the KITTI test set demonstrate that SA-VoxelRCNN achieves consistent improvements in challenging scenarios, with gains of 2.49 and 1.87 percentage points at Moderate and Hard difficulty levels, respectively, while maintaining real-time performance at 22.3 FPS. This approach effectively balances local geometric details with global contextual information, providing a robust detection solution for autonomous driving applications. Full article

With the rapid expansion of the aquaculture scale, the environmental pollution caused by the accumulation of fish pond sediment (FPS) has become increasingly prominent, making it urgent to establish sustainable resource utilization solutions. This study investigates the potential of using FPS as a soil amendment to improve saline–alkali soil (SAS) quality and enhance vegetable growth, while also quantifying ecological benefits through Gross Ecosystem Product (GEP) accounting. A pot experiment was conducted to evaluate the effects of different FPS mass percentages (0%, 20%, 40%, 80%, and 100%) on the growth of three vegetables (water spinach, lettuce, and chili) and soil quality. The results demonstrated that FPS addition at ≥40% significantly improves SAS quality, reducing the pH and salinity (p < 0.05), while enhancing organic matter, nutrient availability, and microbial activity. Among the treatments, 80% FPS maximized vegetable yields, with water spinach achieving the highest edible biomass (37.32 g). Compared to the control, nutritional quality under ≥80% FPS treatment showed substantial increases: vitamin C (133.33–307.03%), soluble sugars (49.97–73.53%), and protein (26.14–48.08%). An economic analysis revealed that 80% FPS with water spinach cultivation generated peak ecological benefits (274,951 CNY·ha⁻¹; 185% above control). These findings provide a scientific basis and effective model for the resource utilization of FPS and the improvement of saline–alkali soil, offering significant implications for the sustainable development of agriculture and environmental protection. Full article

The distribution of the health workforce affects the availability of health service delivery to the public. In practice, the demographic and geographic maldistribution of the health workforce is a long-standing national crisis. In this study, we present an approach based on Functional Principal Component Analysis (FPCA) of data to identify patterns in the availability of health workers across different regions of Kazakhstan in order to forecast their needs up to 2033. FPCA was applied to the data to reduce dimensionality and capture common patterns across regions. To evaluate the forecasting performance of the model, we employed rolling origin cross-validation with an expanding window. The resulting scores were forecasted one year ahead using Autoregressive Integrated Moving Average (ARIMA) and Long Short-Term Memory (LSTM) methods. LSTM showed higher accuracy compared to ARIMA. The use of the FPCA method allowed us to identify national and regional trends in the dynamics of the number of doctors. We identified regions with different growth rates, highlighting where the most and least intensive growth is taking place. Based on the FPSA, we have predicted the need for doctors in each region in the period up to 2033. Our results show that the FPCA can serve as a significant tool for analyzing the situation relating to human resources in healthcare and be used for an approximate assessment of future needs for medical personnel. Full article

Phenotypic traits and phenotypic extraction at the seedling stage of oilseed rape play a crucial role in assessing oilseed rape growth, breeding new varieties and estimating yield. Manual phenotyping not only consumes a lot of labor and time costs, but even the measurement process can cause structural damage to oilseed rape plants. Existing crop phenotype acquisition methods have limitations in terms of throughput and accuracy, which are difficult to meet the demands of phenotype analysis. We propose an oilseed rape segmentation and phenotyping measurement method based on 3D Gaussian splatting with improved PointNet++. The CKG-PointNet++ network is designed to integrate CGLU and FastKAN convolutional modules in the SA layer, and introduce MogaBlock and a self-attention mechanism in the FP layer to enhance local and global feature extraction. Experiments show that the method achieves a 97.70% overall accuracy (OA) and 96.01% mean intersection over union (mIoU) on the oilseed rape point cloud segmentation task. The extracted phenotypic parameters were highly correlated with manual measurements, with leaf length and width, leaf area and leaf inclination R² of 0.9843, 0.9632, 0.9806 and 0.8890, and RMSE of 0.1621 cm, 0.1546 cm, 0.6892 cm² and 2.1144°, respectively. This technique provides a feasible solution for high-throughput and rapid measurement of seedling phenotypes in oilseed rape. Full article

Object detection in complex road scenes is vital for autonomous driving, facing challenges such as object occlusion, small target sizes, and irregularly shaped targets. To address these issues, this paper introduces DGSS-YOLOv8s, a model designed to enhance detection accuracy and high-FPS performance within the You Only Look Once version 8 small (YOLOv8s) framework. The key innovation lies in the synergistic integration of several architectural enhancements: the DCNv3_LKA_C2f module, leveraging Deformable Convolution v3 (DCNv3) and Large Kernel Attention (LKA) for better the capture of complex object shapes; an Optimized Feature Pyramid Network structure (Optimized-GFPN) for improved multi-scale feature fusion; the Detect_SA module, incorporating spatial Self-Attention (SA) at the detection head for broader context awareness; and an Inner-Shape Intersection over Union (IoU) loss function to improve bounding box regression accuracy. These components collectively target the aforementioned challenges in road environments. Evaluations on the Berkeley DeepDrive 100K (BDD100K) and Karlsruhe Institute of Technology and Toyota Technological Institute (KITTI) datasets demonstrate the model’s effectiveness. Compared to baseline YOLOv8s, DGSS-YOLOv8s achieves mean Average Precision (mAP)@50 improvements of 2.4% (BDD100K) and 4.6% (KITTI). Significant gains were observed for challenging categories, notably 87.3% mAP@50 for cyclists on KITTI, and small object detection (AP-small) improved by up to 9.7% on KITTI. Crucially, DGSS-YOLOv8s achieved high processing speeds suitable for autonomous driving, operating at 103.1 FPS (BDD100K) and 102.5 FPS (KITTI) on an NVIDIA GeForce RTX 4090 GPU. These results highlight that DGSS-YOLOv8s effectively balances enhanced detection accuracy for complex scenarios with high processing speed, demonstrating its potential for demanding autonomous driving applications. Full article

Rhododendron L., a renowned ornamental species and one of the ten famous flowers in China, is highly regarded for its aesthetic value and extensive applications in landscaping. However, its growth and quality are significantly compromised by drought stress, particularly in regions with dry conditions. To elucidate the drought response mechanisms of Rhododendron, two cultivars, ‘SaKeSiZhiXing’ (SKSZX) and ‘TuRuiMeiGui’ (TRMG), were subjected to natural drought stress, and changes in chlorophyll fluorescence and transcriptomic profiles were examined at 0 days (d), 4 d, and 8 d of drought exposure. An OJIP fluorescence transient (O-J-I-P) analysis revealed a progressive decline in the F_P parameter and an increase in the F_J parameter as drought stress intensified. Additionally, a delayed fluorescence (DF) analysis showed a gradual reduction in the I₁ and I₂ values within the induction and decay curves under prolonged drought conditions. The 820 nm curve indicated the deactivation of a transient phase characterized by a rapid decline, followed by a slow recovery in the modulated reflection (MR) signal. A transcriptomic analysis of leaves identified 24,352, 18,688, and 32,261 differentially expressed genes (DEGs) in SKSZX at 0 d, 4 d, and 8 d of drought treatment, respectively. In contrast, TRMG exhibited more pronounced and earlier drought-induced alterations. These DEGs were primarily enriched in pathways related to phenylpropanoid biosynthesis, plant hormone signaling, photosynthesis, and photosynthesis-antenna proteins. Additionally, 565 transcription factors (TFs) were identified, including bHLH, WRKY, bZIP, MYB-related, MYB, C2H2, and HSF families. The drought-induced changes in TRMG were more substantial and occurred earlier compared to SKSZX, with a greater impairment in the electron transfer capacity at both the donor and acceptor sides of photosystem II (PSII). This study provides valuable insights into the molecular mechanisms underlying drought tolerance in Rhododendron and offers a foundation for molecular breeding strategies aimed at enhancing drought resistance in future cultivars. Full article

Fusarium pseudograminearum (Fp) and soil salinity are two types of stress that interact in complex ways, potentially leading to more severe consequences on wheat growth and productivity. However, little is known about the colonization efficiency and the signal pathways of the beneficial Trichoderma longibrachiatum TG1 (TG1) in controlling wheat Fusarium crown rot caused by Fp, and enhancing wheat seedling growth under combined salinity and Fp stresses. Therefore, the present study aims to determine the colonization, phytohormone profile, and signaling pathway in TG1-treated wheat seedlings under salinity and Fp stresses. In a dual culture assay, TG1 exhibited a mycoparasitic effect on Fp growth by coiling, conidial attachment, and parasitism observed under fluorescent microscopy. In addition, TG1 colonized the outermost layers of the wheat seedling roots with biomass consisting of conidia and hyphae. Under 100 mM NaCl stress, the combined TG1+Fp-treated seedlings recorded a control efficacy of 47.01% for the wheat crown rot disease compared with Fp-alone-treated seedlings. The contents of indole-3-acetic acid (IAA), gibberellic acid (GA₃), abscisic acid (ABA) and jasmonic acid (JA) significantly increased by 72.16%, 86.91%, 20.04%, and 50.40%, respectively, in the combined TG1+Fp treatments, whereas the ethylene (ET) content decreased by 39.07% compared with Fp alone at day 14; and 5.07 and 2.78-fold increases in the expression of salicylic acid (SA) signaling pathway genes, such as pathogenesis-related protein 1 (PR1) and isochorismate synthase 1 (ICS1) genes were recorded respectively, in the combined TG1+Fp-treated seedlings compared with the control at day 14. Full article

Tuberculosis (TB) remains a leading cause of mortality, with drug resistance highlighting the need for new vaccine targets. Peptidyl-prolyl isomerase A (PpiA), a conserved Mycobacterium tuberculosis (Mtb) protein, plays a role in bacterial stress adaptation and immune evasion, making it a potential target for immunotherapy. This study uses computational methods to assess PpiA’s antigenicity, structural integrity, and immunogenic potential. The PpiA sequence was retrieved from NCBI and analyzed for antigenicity and allergenicity using VaxiJen, AllerTOP, and AllergenFP. Physicochemical properties were evaluated using ProtParam, and structural models were generated through PSIPRED and SWISS-MODEL. Structural validation was performed with MolProbity, QMEANDisCo, and ProSA-Web. B-cell epitopes were predicted using BepiPred 2.0 and IEDB, while T-cell epitopes were mapped via IEDB’s MHC-I and MHC-II tools. Epitope conservation across Mtb strains was confirmed using ConSurf. Results indicate PpiA is highly antigenic, non-allergenic, and stable, with several immunogenic epitopes identified for both B- and T-cells. This study supports PpiA as a promising immunogenic target for TB vaccine development. Full article

Navigating unstructured roads in vineyards with weak satellite signals presents significant challenges for robotic systems. This research introduces DoubleNet, an innovative deep-learning model designed to generate navigation lines for such conditions. To improve the model’s ability to extract image features, DoubleNet incorporates several key innovations, such as a unique multi-head self-attention mechanism (Fused-MHSA), a modified activation function (SA-GELU), and a specialized operation block (DNBLK). Based on them, DoubleNet is structured as an encoder–decoder network that includes two parallel subnetworks: one dedicated to processing 2D feature maps and the other focused on 1D tensors. These subnetworks interact through two feature fusion networks, which operate in both the encoder and decoder stages, facilitating a more integrated feature extraction process. Additionally, we utilized a specially annotated dataset comprising images fused with RGB and mask, with five navigation points marked to enhance the accuracy of point localization. As a result of these innovations, DoubleNet achieves a remarkable 95.75% percentage of correct key points (PCK) and operates at 71.16 FPS on our dataset, with a combined performance that outperformed several well-known key point detection algorithms. DoubleNet demonstrates strong potential as a competitive solution for generating effective navigation routes for robots operating in vineyards with unstructured roads. Full article