Search Results (402)

Lunar optical and synthetic aperture radar (SAR) imagery provide complementary information for characterizing the lunar surface. However, their joint use remains challenging because of substantial cross-modality differences and severe illumination constraints, particularly in polar regions. To address this challenge, we propose LS2ODiff (Lunar SAR-to-Optical Diffusion), a diffusion-based framework designed for SAR-to-optical image translation in lunar environments. LS2ODiff uses SAR observations as conditional guidance in the diffusion process and incorporates a partial-convolution strategy into the U-Net backbone to handle irregular invalid regions. In addition, self-attention modules are incorporated into the downsampling stages of the U-Net to model long-range spatial dependencies and enhance global structural consistency in complex lunar terrain. We further construct a dedicated paired dataset of the lunar south polar region by registering Chandrayaan-II DFSAR data with Lunar Reconnaissance Orbiter (LRO) Narrow-Angle Camera (NAC) imagery. Comparative experiments against Pix2Pix, CycleGAN, SynDiff, and ConDiff demonstrate that LS2ODiff achieves better visual fidelity and quantitative performance in terms of peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), Fréchet inception distance (FID), and learned perceptual image patch similarity (LPIPS). These results demonstrate the potential of diffusion models for high-fidelity lunar image translation, offering new opportunities for polar terrain interpretation and future exploration missions. Full article

Accurate individual tree crown (ITC) segmentation from unmanned aerial vehicle (UAV) imagery is important for fine-scale forest inventory, plantation management, and ecological monitoring. However, delineating ITCs in dense plantation environments remains difficult because crowns are strongly adjacent, canopy structures are highly homogeneous, and crown boundaries are often blurred, making it hard for existing methods to preserve both regional integrity and boundary continuity. This study proposes the Perceptual Segment-Anything Model with Multi-head Cross-Parallel Attention (Per-SAM-MCPA), a lightweight and effective framework for fine-grained ITC segmentation in dense plantation scenes. Based on a compact ResNet-50 backbone, the framework integrates perceptual target-aware representation, multi-scale detail enhancement, global contextual modeling, and semantic-boundary collaborative refinement to improve crown discrimination and structural consistency. A perceptual relation module is used to strengthen pixel-level semantic dependency modeling, and a Multi-head Cross-Parallel Attention (MCPA) mechanism is designed to capture long-range contextual interactions through orthogonally decomposed spatial attention, improving global geometric consistency with limited computational overhead. A Composite Constraint Loss (CCL) that combines a weighted cross-entropy loss, a structural similarity loss, and a boundary term based on Hausdorff distance is introduced to jointly optimize region-level segmentation quality and boundary fidelity. Experiments on the Catalpa bungei UAV dataset show that the proposed method achieves an intersection over union (IoU) of 87.3% and an F1-score of 91.0%, outperforming representative baseline methods such as SAM and Mask R-CNN while maintaining an inference speed of 35.7 FPS on a single GPU. These results indicate that Per-SAM-MCPA offers an accurate, efficient, and practical solution for ITC segmentation in dense plantation environments. Full article

Background: Visio-spatial skills (VSS) are essential perceptual-cognitive skills that enable athletes to process visual information, interpret spatial relationships, and execute appropriate motor responses in dynamic sporting environments. In combat sports, athletes must rapidly anticipate and react to an opponent’s actions, making well-developed VSS crucial for optimal performance. Although boxing, karate, and taekwondo share similar competitive characteristics, each discipline presents distinct technical and perceptual demands that may influence the development of specific VSS profiles. This study aimed to investigate whether significant differences exist in VSS profiles among boxing, karate, and taekwondo athletes. Methods: A comparative cross-sectional design was used involving 150 amateur combat sport athletes, 50 boxers, 50 karate athletes, and 50 taekwondo athletes. Participants were assessed using a VSS test battery measuring six variables: accommodation facility (AF), saccadic eye movement (SEM), speed of recognition (SR), (HEC), peripheral awareness (PA), and visual memory (VM). Data was analyzed using one-way ANOVA with η², ω², and Cohen’s f effect sizes, and principal component analysis (PCA). Results: One-way ANOVA revealed statistically significant differences in five of six VSS (all p < 0.001). PA produced the largest sport-specific differentiation (η² = 0.457, Cohen’s f = 0.918), followed by HEC (η² = 0.273, f = 0.612), SR (η² = 0.224, f = 0.537), and SEM (η² = 0.180, f = 0.468). AF yielded a significant moderate effect (η² = 0.108, f = 0.347). VM was the sole non-significant variable (F (2.147) = 0.74, p = 0.479, ω² = 0.000), suggesting domain-general encoding processes insensitive to discipline-specific training at this developmental level. Boxing athletes achieved the highest scores in SEM, SR, and PA, while karate athletes led in AF and HEC. PCA revealed a single dominant component (PC1 = 93.91% of variance), confirming that VSS function as a highly integrated perceptual-motor construct rather than independent sub-skills. Conclusions: Visio-spatial skills in combat sports are governed by a dominant integrated factor, with discipline-specific variations reflecting unique performance requirements. Visio-spatial skills in combat sport athletes are highly interdependent and largely governed by a single perceptual-motor construct, with discipline-specific profiles observed across boxing, karate, and taekwondo. The findings support the integration of sport-specific, ecologically valid visual training programs targeting key perceptual-cognitive skills, alongside routine assessment to inform athlete development and performance optimization. Full article

In recent years, animation-based systems for human-computer interaction have attracted increasing attention. This work proposes a hybrid framework that combines mathematical modeling and bio-inspired optimization algorithms to generate motion sequences from skeletal data. The framework takes as input a complete skeletal sequence corresponding to a given action and optimizes both the number of key poses and the parameters of a homotopy-based formulation to generate transitions between consecutive poses. A homotopy-based approach is used to compute transitions between selected key poses. The homotopy parameter $\lambda$ serves as an indicator of the completeness of the transition between pairs of key poses. Four nature-inspired optimization algorithms: Genetic Algorithm, Micro Genetic Algorithm, Particle Swarm Optimization, and Ant Colony Optimization were evaluated to determine the number of key poses and homotopy parameters that enable feasible motion generation. Dynamic Time Warping (DTW) is used as an external metric to assess the similarity between generated and reference sequences. It is important to note that Dynamic Time Warping (DTW) should be considered as a sequence similarity measure, as it does not explicitly evaluate perceptual realism or biomechanical plausibility. The framework was evaluated on 18 action sequences, demonstrating its ability to generate feasible motion transitions in 16 of the 18 evaluated actions when using PSO and MicroGA. For each pair of key poses, a fixed number of intermediate frames is generated to provide a uniform temporal discretization of the motion. The results suggest that homotopy-based methods provide a feasible approach for animation-based interaction systems. Full article

Background/Objectives: Temporal predictability may sharpen our ability to distinguish similar sounds, but whether this relies on attention is unclear. This study examined how temporal structure influences frequency discrimination. Methods: Thirty-six adults completed active (attend) and passive (ignore) listening tasks across three paradigms that varied in temporal structure: oddball (isolated deviants), two-tone frequency discrimination paradigm (pairs comparison), and local irregularity of the local/global paradigm (five-tone sequences, bundles). Stimuli varied in difficulty via small or large frequency deviations. Behavioral responses and subjective ratings were collected during active and passive listening. EEG was recorded to assess mismatch negativity (MMN) (either early MMN (eMMN) or mismatch response (MMR)) and P300 event-related potentials. Results: Under active listening, temporal predictability significantly improved performance, but only for difficult discriminations. The local-irregularity condition yielded higher hit rates and greater perceptual sensitivity (d’) than the other paradigms. This benefit was accompanied by enhanced P300, yet participants rated the conditions as equally difficult, indicating no metacognitive awareness. Under passive listening, predictability helped only for easy stimuli, marked by a larger MMR. No reliable change-detection response occurred for difficult sounds when attention was diverted. Conclusions: These findings suggest that the combination of temporal predictability and repeated standard presentation in the local irregularity paradigm can improve frequency discrimination under challenging, attended conditions, with some evidence for partial dissociation between objective performance and subjective awareness. However, substantial individual variability and cross-paradigm confounds caution against strong causal claims. These results are broadly consistent with predictive coding frameworks but require replication with counterbalanced designs and larger deviant trial counts. Full article

Tunnels, soundproof screens and other vertical roadside traffic facilities play an important role in isolating the driving environment, maintaining driving safety, and reducing driving noise. As the usage time increases, these facade traffic buildings become polluted and cause traffic safety problems. Obstacles on three-dimensional walls of different shapes, colors, and sizes are the most challenging problem in intelligent cleaning environment perception. This paper proposes an obstacle segmentation method based on a visual language model to overcome these problems. Firstly, in the constructed experimental environment, a visual–language obstacle dataset is collected, named the Road-side General Obstacles Dataset (RGOD), and the collected dataset is labeled with both a segmentation mask and a language description. These preprocessing results are used as the training input of the perception model to obtain the foreground and background separation results. Secondly, a VLM-GOS model was proposed to segmentation special-shaped obstacles, which emphasizes the distinction between background and foreground targets. Finally, the general obstacle is segmented by a vision–language model with a similar loss function, and evaluated with different metrics. Experimental results show that compared with models such as MaskFormer, SegFormer, and ASD-Net, this method improves the model’s perceptual ability and increases accuracy by 3%. More importantly, the model is more interpretable. Full article

This study analysed the potential influence of crew size on performance (stroke rate, strokes/min; distance travelled, m/min; and average power, W), physiological responses (post-exercise heart rate and heart rate measured three minutes after exercise) and perceptual responses (Borg scale). A total of 136 adolescent athletes (100 males and 36 females; mean age = 15.79 ± 1.14 years) performed four three-minute maximal-effort trials on a rowing ergometer across four conditions: individual trials (C1), two-person crews (C2), four-person crews (C3), and eight-person crews (C4). Results showed a significant increase in stroke rate (strokes/min) in both sexes as crew size increased (C1 33.16 ± 2.54 vs. C4 34.19 ± 2.21 strokes/min; C1–C4 p = 0.01; C2–C4 p = 0.003). Men reported greater perceived exertion in C1 compared with C4 (Borg 7.80 ± 0.79 vs. 7.46 ± 0.74; p = 0.032), despite no associated changes in performance (863.88 ± 45.10 vs. 863.26 ± 47.63 m/min) or average power (311.71 ± 46.43 vs. 311.44 ± 50.43 W), whereas no differences in perceived exertion were observed in women (Borg 7.59 ± 0.84 vs. 7.56 ± 0.76). Cardiovascular responses were similar across sexes and experimental conditions. In summary, these preliminary findings could point toward the existence of sex-differentiated patterns. The data appear to suggest a more pronounced tendency toward the ‘crew-size effect’ among the men in the sample, whereas an inclination toward maintaining individual responsibility is observed in the women. Full article

3D scene reconstruction serves as a fundamental technology with widespread applications in virtual reality, structural inspection, and robotic systems. While recent advances in 3D Gaussian Splatting have significantly enhanced scene reconstruction capabilities, the performance of such methods remains suboptimal when applied to highly reflective environments. To overcome this limitation, we introduce ARS-GS, a novel framework that integrates Anisotropic Spherical Gaussian reflection modeling and spherical harmonics diffuse approximation into a physically based rendering pipeline. This architecture incorporates a skip connection between the Anisotropic Spherical Gaussian module and the Gaussian primitives, effectively preserving surface details while maintaining computational efficiency. Comprehensive experimental evaluations validate the efficacy of ARS-GS across multiple datasets. Specifically, our method establishes new state-of-the-art quantitative benchmarks, achieving a peak signal-to-noise ratio of 38.30 and a structural similarity index measure of 0.997 on the neural radiance fields synthetic dataset, alongside a peak signal-to-noise ratio of 46.31 on the Gloss Blender dataset. Furthermore, on the challenging reflective neural radiance fields real-world dataset, our approach secures the highest peak signal-to-noise ratio scores, highlighted by a metric of 26.26 on the Sedan scene. The proposed method also substantially reduces perceptual errors, yielding a learned perceptual image patch similarity as low as 0.204, thereby consistently outperforming existing techniques in the reconstruction of highly specular surfaces with superior geometric fidelity. Full article

Background: Small-sided games (SSG) and running-based high-intensity interval training (HIIT) are commonly used in soccer conditioning to improve aerobic fitness and performance. Although both modalities induce high cardiovascular stress, their acute neuromuscular, perceptual, and cognitive responses remain incompletely understood when examined within the same cohort. This study compared the acute physical, psychophysiological, and cognitive responses to SSG and Tabata-type HIIT in amateur male soccer players. Methods: Thirty-two male amateur players (n = 32; age: 20.53 ± 1.65 years) completed a counterbalanced within-subject crossover design. Participants performed a 4v4 SSG protocol and a running-based Tabata-HIIT protocol (8 × 20 s, 10 s recovery) on separate days (48 h apart). Countermovement jump (CMJ), squat jump (SJ), 20-m sprint, agility t-test, heart rate, perceived exertion (Borg CR-10), mental effort, and cognitive performance (d2 test) were assessed pre- and post-exercise. Parametric variables were analyzed using 2 × 2 repeated-measures ANOVA (time × protocol; η²p), and non-parametric data were analyzed using Friedman and Wilcoxon tests (r) (p < 0.05). Results: Both protocols elicited similar cardiovascular responses (~90% HR_max). A significant protocol × time interaction was observed for CMJ (p < 0.001), showing a decline after Tabata-HIIT, whereas performance was maintained after SSG. No inter-protocol differences were found for SJ, sprint, or agility. Perceived exertion and mental effort during recovery were higher following Tabata-HIIT (p < 0.05). Cognitive performance improved after both protocols (p < 0.001), with no between-protocol differences. Conclusions: Despite comparable cardiovascular load, Tabata-HIIT was associated with greater acute neuromuscular and perceptual strain, whereas SSG preserved neuromuscular performance. Perceptual and mental responses may therefore differ despite similar physiological intensity, which may inform soccer training prescription. Full article

As autonomous driving technology advances, the deployment of autonomous vehicles in urban environments is rapidly increasing. Lens flare—an often overlooked optical artifact in object detection research—can lead to increased false positives or missed detections, particularly in the challenging conditions inherent to autonomous driving. Current mitigation methods are often ill-suited for real-time implementation. This work proposes a solution to alleviate the adverse effects of lens flare by utilizing a lightweight lens flare perception network, eliminating the need for additional hardware or complex image pre-processing. Specifically, we propose a reference-free model utilizing a ResNet18 backbone integrated with a lightweight Multi-Layer Perceptron (MLP) to extract and leverage lens flare information. This model is developed via a teacher–student framework, which was distilled from an end-to-end reference-based model optimized using the Learned Perceptual Image Patch Similarity (LPIPS) metric. Our experiments demonstrate that incorporating lens flare information significantly enhances the performance of the baseline object detection network, outperforming previous mitigation methods by a substantial margin. The proposed method can be seamlessly integrated into existing object detectors and requires only an efficient training process, facilitating its deployment in practical autonomous driving tasks. Full article

Obtaining multiple MRI contrasts for each patient prolongs scan acquisition time, increases healthcare costs, and may not always be feasible due to patient specific constraints. Deep learning-based MRI contrast synthesis offers a potential solution, yet most existing approaches are evaluated on preprocessed public benchmarks that do not reflect real-world clinical variability. In this study, we propose a fusion U-Net transformer framework for bidirectional T1-weighted ↔ T2-weighted brain MRI synthesis trained and evaluated exclusively on retrospectively acquired clinical data. The proposed architecture integrates multiscale convolutional feature extraction with axial attention mechanisms and a transformer bottleneck for efficient global context modeling. A fusion refinement block is incorporated to mitigate skip connection artifacts. An adversarial training strategy with the least squares GAN objective and a hybrid loss combining L1 reconstruction and structural similarity (SSIM) is employed to promote both pixel-level accuracy and perceptual fidelity. The model is evaluated using SSIM and PSNR metrics alongside qualitative expert assessment conducted by two board-certified radiologists. For both synthesis directions, the framework achieves competitive quantitative performance against baseline models under the challenging conditions of clinical data. Expert evaluation confirms high anatomical fidelity and clinically acceptable image quality across both synthesis directions. These results indicate that the proposed framework represents a promising approach for multi-contrast MRI synthesis in clinically heterogeneous data environments. Full article

In this study, a generalized and advanced form of the near soft set theory (NST) framework is proposed for information aggregation (IA) processes. The primary motivation of the study is to address the lack of similarity-based uncertainty modeling in the literature by integrating the parametric structure of soft sets with the similarity-oriented structure of nearness approximation spaces. Within this framework, the AND-product and OR-product operations are introduced as the main methodological tools, and their algebraic structures are analyzed in detail. It is mathematically demonstrated that these operations satisfy fundamental properties such as idempotency, absorption, distributivity, and De Morgan identities. The principal original contribution of the study is the development of a novel Uni–Int-based decision-making mechanism that enables the systematic distinction between strong and acceptable alternatives. In addition, the boundary frequency indicator ($b_r$), which quantitatively evaluates the reliability of objects under perceptual uncertainty and is introduced for the first time in the literature, is proposed. The applicability of the proposed model is demonstrated through a real-estate selection problem, and a sensitivity analysis is conducted to reveal the determining effect of the nearness parameter r on decision granularity. The obtained findings indicate that the proposed NST framework provides a more flexible, more discriminative, and structurally robust decision-support model than classical approaches, particularly for similarity-based IA problems. Full article

The rapid growth in large-scale image repositories over the past few years has made exact and near-duplicate images increasingly common, creating substantial redundancy that wastes storage resources and reduces retrieval efficiency in practical systems. Even though perceptual hashing and deep learning are promising deduplication strategies, the lack of standardized benchmarks complicates direct comparison. In this study, we conduct a unified, controlled evaluation of five commonly used methods, including four classical perceptual hashes (AHash, DHash, PHash, and WHash) and a CNN-based embedding model. We evaluate all methods on the UKBench and Amazon Berkeley Objects datasets using identical preprocessing, thresholds, and metrics, which include exact duplicates, near-duplicates, and geometrically transformed duplicates. Our experiments highlight a clear trade-off between speed and robustness. Hashing methods are computationally efficient and effective for exact matches, but perform poorly on near-duplicates and under geometric transformations, whereas the CNN model is significantly more robust across all duplicate types, but comes at a high computational cost. Based on these results, we outline practical recommendations for selecting deduplication strategies in large-scale applications. In addition, our evaluation setup serves as a reproducible baseline for future research in image similarity and large-scale deduplication. Full article

Microwave imaging is emerging as an alternative to conventional medical diagnostic techniques. Traditional analytical and numerical methods fail to adequately address these fundamental challenges: they often rely on strict linear approximations or simplified physical models, leading to low reconstruction accuracy, poor robustness, and limited generalization ability in complex clinical scenarios. As a result, they cannot meet the high-precision requirements of practical stroke microwave imaging. To further improve the accuracy of microwave imaging algorithms in recognizing stroke regions and solving the backscattering problem, this study employs a combination of methods with deep learning. It presents the Multi-Scale Attention Network (MSA-Net) for microwave imaging. The network is based on the EGE-UNet network structure with improved multi-axis Hadamard attention, incorporating null-space pyramid pooling and introducing a deep supervisory mechanism to improve the network performance further. To combine microwave imaging with deep learning, firstly, a large amount of microwave data need to be simulated with HFSS, in which the simulation model is a human brain stroke model constructed by an HFSS simulation system. Secondly, the microwave data obtained from the simulation are converted into a tensor format. Then, the tensor data are input into the MSA-Net neural network, which generates a binary mask image that can be used to detect the size and location of the stroke. This study also prompts the model to converge faster by sparsifying the microwave data to improve training efficiency. The method has been tested using simulation data, and based on the comparison experiments with other networks, MSA-Net is more accurate in detecting the location and the bleed size. The experimental results show that the proposed method is superior for stroke imaging. The experimental results show that the proposed model achieves a 1.08 improvement in peak signal-to-noise ratio and a 0.017 reduction in learned perceptual image block similarity, fully validating the effectiveness of the structural optimization strategy proposed in this paper. Full article

Low-count positron emission tomography (PET) is inherently affected by Poisson-dominated noise, which degrades image contrast, structural delineation, and quantitative reliability. This study systematically evaluated residual learning-based deep neural networks to investigate the influence of residual block depth on PET image restoration performance under low-count conditions. We employed a physically controlled striatum phantom, fabricated using 3D printing technology, to ensure reproducible acquisition conditions and controlled physical variability. PET images were acquired using a clinical PET/computed tomography (CT) system with list-mode acquisition. Low-count images reconstructed from short-duration acquisition were paired with high-count reference images reconstructed from extended acquisitions. We compared conventional filtering techniques, including median, Wiener, and modified median Wiener filters, with residual network (ResNet)-based models incorporating 8, 16, and 32 residual blocks. Image quality was quantitatively assessed using contrast-to-noise ratio (CNR), coefficient of variation (COV), line profile analysis, universal quality index (UQI), and perceptual image patch similarity (LPIPS). The results demonstrated that ResNet-based restorations substantially outperformed conventional filtering techniques in contrast recovery, signal stability, and structural preservation. The ResNet-16 model achieved the most balanced performance, yielding the highest CNR (9.02) and lowest COV (0.105), while also demonstrating superior structural and perceptual similarity, as indicated by UQI (0.9224) and LPIPS (0.0174), relative to the high-count reference images. Deeper network configurations exhibited diminishing returns and reduced structural consistencies. These findings indicate that an intermediate residual block depth is optimal for low-count PET image restoration and highlight the importance of architectural optimization in deep learning-based PET image enhancement with phantom-based evaluation frameworks. Full article