Search Results (1,523)

This study investigated neural oscillatory dynamics underlying visual word recognition in university students with dyslexia using a portable brain–computer interface (BCI) EEG system. The sample included university students with dyslexia (N = 12) and matched controls (N = 14) who completed auditory discrimination and visual word recognition tasks, with and without musical accompaniment. Through these experimental conditions, the researchers assessed (a) the cortical activation across frequency bands, (b) the modulatory effect of background music, and (c) the relationship between emotional states and brain activity. Results revealed significant group differences in oscillatory patterns, with reduced β- and γ-band activity in the left occipito-temporal cortex among participants with dyslexia, confirming disrupted temporal coordination in posterior reading networks. Compensatory right-hemisphere activation was observed, particularly under musical conditions, accompanied by increased α-band power and reduced δ activity, indicating enhanced attentional engagement and reduced cognitive fatigue. Emotional assessment using the DASS-21 revealed higher stress and anxiety scores in the dyslexic group, suggesting that affective factors may modulate oscillatory dynamics. The presence of background music appeared to attenuate these effects, supporting improved emotional regulation and cognitive focus. These findings demonstrate that dyslexia reflects a distributed disruption in neural synchrony and cross-frequency coupling, influenced by both cognitive and affective mechanisms. The integration of portable EEG technology with rhythmic auditory stimulation offers new insights into the neurophysiological and emotional aspects of dyslexia, highlighting the potential of rhythm- and music-based approaches for both diagnostic and therapeutic applications. Full article

γ-photon tomography, which leverages the high penetration and electrical neutrality of high-energy γ-photons, offers a promising non-contact approach for industrial flow field monitoring. However, γ-photon flow-field images are inherently grayscale and exhibit probabilistic statistical imaging characteristics, leading to color banding artifacts when processed by mainstream colorization algorithms like DeOldify, which compromise structural continuity and visual consistency. To address this issue, this paper proposes a Structure Enhancement Colorization Network (SECN) model for γ-photon flow-field image colorization. A U-Net + GAN framework is employed, with ResNet101 as the generator backbone. It integrates structure-aware enhancement and multi-scale attention modules, while the discriminator incorporates enhanced blocks for improved boundary and texture discrimination. By adaptively fusing global–local features across channel and spatial dimensions, the SECN model effectively suppresses color banding artifacts and enhances structural consistency. To validate the effectiveness of the proposed algorithm, two CFD-simulated γ-photon flow-field image colorization scenarios—namely a large-scale vortex wake and a horizontal wake—are used as evaluation targets. In terms of image quality metrics, the proposed colorization algorithm achieves PSNR, SSIM, FID, and MAE values of 32.5831, 0.8612, 17.8514, and 0.0191, respectively, corresponding to improvements over DeOldify of 4.54%, 2.82%, 5.18%, and 11.16%. When considering information entropy, the proposed colorization algorithm achieves an average entropy value of 4.0257, marking a 4.44% increase compared to DeOldify’s 3.8543, demonstrating superior information preservation and reduced uncertainty in reconstructing complex probabilistic structures. Furthermore, from the perspective of parameter inversion, the temperature inversion MAPE is 7.60%, which is a significant reduction of 18.42% compared to that of DeOldify. Full article

Computational models of the MLG1 neurons in crab Neohelice granulata have been developed to detect and spatially localize looming stimuli. However, existing models suffer from significant performance degradation in dim scenarios, primarily due to visual signal corruption from stochastic noise such as photon shot noise. To address this challenge, we propose a computational framework that embeds Daubechies wavelet directly into ON/OFF visual pathways. The ON/OFF mechanism separates the input signals in parallel based on luminance changes to capture dynamic differences between target and background. Embedding Daubechies wavelet enables multi-scale frequency decomposition, allowing the model to suppress high-frequency noise while enhancing low-frequency looming trends. This process extracts low-frequency components and high-frequency details, providing the MLG1 neuron with more discriminative feature inputs. Experimental results demonstrate that the model achieves reliable looming spatial localization under extremely low contrast conditions, offering a robust methodology for bionic vision in extreme dim light environments. Full article

Background/Objectives: This study aimed to develop a nomogram-based prediction tool to estimate visual field (VF) progression in patients with bilateral myopic normal-tension glaucoma (mNTG) by integrating key structural and vascular parameters. Methods: This retrospective cohort study included 150 eyes from 75 treatment-naïve patients with mNTG. All subjects were followed for at least five years with at least six reliable VF examinations. Key structural features, including the lamina cribrosa steepness index (LCSI) via enhanced-depth imaging optical coherence tomography (OCT) and choroidal microvascular dropout (cMvD) via OCT angiography (OCTA), were evaluated. VF progression was determined by event-based glaucoma progression analysis (GPA). To construct the predictive nomogram, clustered logistic regression with forward selection and 1000 bootstrap iterations was used to identify independent predictors. Results: Of the 150 eyes, 58 (38.7%) exhibited VF progression. Multivariable analysis identified steeper LCSI and the presence of parapapillary cMvD at baseline as significant independent predictors of progression. The resulting nomogram demonstrated excellent predictive accuracy, with an AUC of 0.922 and a C-index of approximately 0.92, indicating strong discriminative ability. Conclusions: This nomogram, incorporating structural (LCSI) and vascular (cMvD) markers, may offer a useful individualized tool for predicting VF progression in mNTG. This tool could assist in the early identification of high-risk patients and supports personalized treatment planning to optimize long-term visual outcomes. Full article

Dragon fruit (Hylocereus spp.) is an emerging crop in the tropics and subtropics, but its production is increasingly threatened by diseases that reduce yield and profitability. Early diagnosis of these diseases is crucial for timely intervention, yet visual symptoms often appear only after significant infection has occurred. The study aims to evaluate how optical spectral reflectance can detect dragon fruit diseases and identify the most responsive spectral regions. In this study, six major dragon fruit stem diseases: Neoscytalidium stem canker, stem sunburn, anthracnose, Botryosphaeria stem canker, Bipolaris stem rot, and bacterial soft rot were characterized by the goal of identifying unique spectral signatures for early detection and differentiation of each disease. Seventy-two potted dragon fruit plants of three distinct species were grown under four organic vermicompost treatments (0, 5, 10, 20 tons/acre) in both open-field and high-tunnel conditions together, in a randomized complete block design. A handheld spectroradiometer (350–2500 nm) was used to collect reflectance from the diseased and healthy cladodes (stem segment). Various spectral vegetative indices were computed to identify disease-specific features. The results revealed distinct spectral features for each disease. Infected cladodes consistently exhibited higher reflectance especially in the visible region (400–700 nm) and the near-infrared region (900–2500 nm) of the spectrum than healthy cladodes. The Normalized Difference Vegetative Index (NDVI), Green Normalized Difference Vegetative Index (GNDVI), and Spectral Ratio (SR) spectral indices were significantly higher in healthy plants than in diseased ones, reflecting higher chlorophyll concentration and plant biomass. Conversely, the 1110/810 ratio was lower in healthy plants than in diseased plants, suggesting a more compact internal plant structure. Statistical analysis revealed highly significant differences (p < 0.00001) between healthy and diseased spectra in the Red, Green and NIR regions. Linear Discriminant Analysis(LDA) achieved the highest classification accuracy (OA = 0.642, κ = 0.488), though performance was limited for minority classes. These findings demonstrate that targeted spectral sensing can identify dragon fruit diseases before obvious symptoms emerge. By pinpointing disease-specific spectral indices, our study paves the way for early-warning tools such as targeted multispectral sensors or drone-based imaging that would enable growers to intervene sooner and limit losses. These results highlight the potential for development of UAV-based or portable spectral sensors for large-scale, near real-time disease monitoring in dragon fruit production. Full article

Metallic Shear Plate Dampers (MSPDs) are essential components in passive vibration control systems and require rapid post-earthquake inspection to assess damage and determine replacement needs. Traditional visual inspection methods suffer from low efficiency and limited ability to detect concealed damage. This study proposes a novel MSPD damage detection method based on active sensing and the k-nearest neighbor (KNN) algorithm, featuring high accuracy, efficiency, and low cost. Quasi-static tests were conducted to simulate various damage states. Sweep-frequency excitation was applied using a charge amplifier, and piezoelectric sensors were employed to generate and receive stress wave signals corresponding to different damage conditions. The acquired signals were processed using wavelet packet transform (WPT) and energy spectrum analysis to extract discriminative time–frequency features, which were used to train and validate the KNN model. Results show that the model achieved a validation accuracy of 98.9% using all valid data and 98.1% using a single excitation-sensing channel. When tested on an MSPD with a similar overall structure but lacking stiffeners, the model achieved an accuracy of 92.6% in distinguishing between healthy and damaged states. This indicates that the proposed method has good robustness and practical potential for MSPDs with similar damage evolution and failure modes despite certain structural variations. Full article

Image classification in low-data regimes remains a challenging problem, particularly in stylized visual domains where intra-class similarity and inter-class feature overlap limit discriminative capacity. This study presents a systematic evaluation of regularization and transfer learning strategies for multi-class comic character recognition under constrained data conditions. Four convolutional architectures are compared: (i) a baseline CNN trained from scratch, (ii) a regularized CNN incorporating data augmentation, dropout, and early stopping, (iii) a pretrained ResNet-50 used as a fixed feature extractor, and (iv) a partially fine-tuned ResNet-50 with selective layer unfreezing. Experiments are conducted on a custom four-class dataset exhibiting moderate class imbalance, evaluated using both a fixed 70/20/10 split and 5-fold cross-validation to assess generalization stability. Results indicate that shallow CNN architectures suffer from substantial overfitting, even when regularization is applied, whereas transfer learning significantly improves macro-averaged F1-score and out-of-distribution detection performance. Cross-validated results, the primary basis for inference given the dataset scale, show that both ResNet-50 strategies achieve equivalent mean accuracy of 95.0% (SD: ±0.4% for feature extraction, ±0.8% for fine-tuning; paired t = 0.00, p = 1.000), while shallow CNN architectures reach only 81–87%. Under a single fixed 70/20/10 partition (n = 69 test samples, 95% CI: ±9–12%), fine-tuning nominally reaches 98.5%; crucially, cross-validation deflates this figure to parity with feature extraction, confirming it reflects favorable partitioning rather than genuine architectural superiority. The primary finding is therefore that frozen ResNet-50 feature extraction is the recommended strategy: it matches fine-tuning in cross-validated generalization while requiring 15× fewer trainable parameters and exhibiting lower fold-to-fold variance. The findings demonstrate that pretrained deep residual representations transfer effectively to stylized comic imagery and that evaluation protocol selection critically impacts perceived performance in small datasets. These results provide practical guidelines for robust model selection in domain-specific, limited-data image classification tasks. Full article

Cognitive or visual distraction caused by drivers using mobile phones, operating the central console, or conversing with passengers while driving is a significant contributing factor to road traffic accidents. Aiming to solve the problem that existing driving behavior monitoring systems exhibit insufficient recognition accuracy and low real-time detection performance in complex driving environments, this study proposes a dual-flow driver distraction detection model based on Sobel edge detection (DFSED-Model). The model is designed with a collaborative learning framework: the first flow adopts a lightweight pre-trained backbone network to achieve efficient semantic feature extraction. The second flow utilizes Sobel edge detection to extract the driver’s driving contours and enhances the model’s spatial sensitivity to driving movements and hand movements. Through the feature learning process of the first-flow-guided auxiliary branch, collaborative optimization of knowledge transfer and attention focusing is realized, thereby improving the model’s convergence speed and discriminative performance. The proposed model is evaluated on three widely used public datasets: the State Farm Distracted Driver Detection (SFD) dataset, the 100-Driver dataset, and the American University in Cairo Distracted Driver Dataset (AUCDD-V1). Under the premise of maintaining low computational overhead, the accuracy of the DFSED-Model reaches 99.87%, 99.86%, and 95.71%, respectively, which is significantly superior to that of many mainstream models. The results demonstrate that the proposed method achieves a favorable balance between accuracy, parameter count, and efficiency, and possesses strong practical value and deployment potential. Full article

Recent advances in vision-language cross-modal learning have substantially improved the performance of video temporal grounding. However, most existing methods directly associate global video features with sentence-level features, overlooking the fact that textual semantics usually correspond to only limited spatio-temporal regions within a video. This limitation often leads to unstable alignment in complex scenarios involving intertwined events and diverse actions. In essence, accurate video temporal grounding requires the joint modeling of fine-grained spatial semantics and heterogeneous temporal event structures. Motivated by this observation, we propose a hierarchical prototype alignment approach that models cross-modal correspondence between video and text through structured intermediate prototype representations. Specifically, the alignment process is decomposed into two complementary stages: object-phrase alignment and event-sentence alignment. In the object-phrase alignment stage, discriminative local visual regions and informative textual words are aggregated to construct object and phrase prototypes, thereby enhancing fine-grained spatial correspondence at the level of entities and localized actions. In the event-sentence alignment stage, object prototypes are further integrated along the temporal dimension to form event prototypes that represent continuous action units, enabling effective alignment with sentence-level semantics and facilitating the modeling of diverse temporal event structures. On this basis, we further directly inject cross-modal alignment information into candidate moment aggregation. This design allows candidate moment representations to emphasize query-relevant temporal regions. Extensive experiments on Charades-STA, ActivityNet Captions, and TACoS demonstrate that the proposed method outperforms existing approaches, validating the effectiveness of hierarchical prototype alignment for improving both cross-modal alignment quality and temporal grounding accuracy. Full article

Accurate and early identification of ocular diseases is essential to prevent vision impairment and enable timely medical intervention. In routine clinical practice, ophthalmologists rely on a structured diagnostic workflow that incorporates multiple imaging modalities to manually assess and diagnose ocular diseases. However, interpreting each modality requires significant clinical experience and can be time-consuming. These limitations can be effectively addressed through the application of AI (Artificial intelligence)-driven multimodal fusion techniques. In this study, we conducted an empirical investigation to assess the impact of different fusion strategies—including early, intermediate, and late fusion—on diagnostic performance, training requirements, and interpretability. The proposed methodology was evaluated using three publicly available datasets: FFA-Fundus (Fundus fluorescein angiography), GAMMA (Glaucoma Analysis and Multi-Modal Assessment), and OLIVES (Ophthalmic Labels to Investigate Visual Eye Semantics). Experimental results demonstrate that multimodal feature fusion improves disease detection performance. Although fused models typically required an increase in training parameters compared to single-modality models, they provided interpretability on par with that of individual single-modal networks. However, inference time increased by approximately 50% for multimodal architectures. These findings underscore the value of integrating diverse ophthalmic imaging modalities to enhance diagnostic accuracy in automated disease detection systems. At the same time, the results highlight that unimodal models containing highly discriminative features can also perform competitively, particularly when a single modality is sufficient for disease identification. Multimodal fusion provides the greatest benefit in scenarios where complementary information across modalities contributes distinct and non-redundant features. Furthermore, fusing all available modalities may not be optimal due to increased computational cost and reduced inference efficiency; thus, selective modality integration and lightweight fusion strategies are essential to balance accuracy, interpretability, and efficiency in clinical deployment. Full article

Therapist shortages make home-based rehabilitation an essential component of post-stroke care, yet patients often exhibit reduced adherence when functional gains are difficult to quantify and interpret. This study presents a cloud-enabled assessment framework centered on a dynamic reaching task for upper-limb rehabilitation in individuals with mild stroke. The proposed system combines wearable sensing and Internet of Things (IoT) connectivity to stream kinematic data to the cloud for near real-time analysis, and integrates a force-feedback rehabilitation robot to deliver motion guidance during training. The pipeline proceeds in three stages. First, smoothness-related kinematic descriptors are extracted and fed into a deep multi-class classifier to discriminate the affected side (left, right, or healthy). Second, movement quality is modeled using a Gaussian Mixture Model (GMM) trained on IoT-acquired trajectories to quantify performance via probabilistic similarity. Third, a calibrated scoring function transforms GMM log-likelihood into a normalized 0–1 quality index, producing visual reports that support interpretable feedback for patients and therapists. The framework is validated using motion data collected from stroke patients at Taipei Veterans General Hospital. Experimental results demonstrate that the neural network multi-classifier achieved an F1-score of 0.95. Incorporating robot-derived interaction signals further improved classification performance by approximately 5%. For movement quality assessment, the derived scores showed a significant positive correlation (Pearson correlation = 0.632, p = 0.02) with therapist-defined gold reference standards for right-affected patients. Additionally, integrating robot force-feedback signals and AIoT-enabled dynamic streams improved score accuracy by 8% and score responsiveness by 10%. These quantitative outcomes substantiate the efficacy of combining IoT-driven sensing and robot-assisted training for objective, interpretable, and remotely deployable motor assessment. Full article

Human pose estimation (HPE) aims to localize human keypoints from visual inputs, which faces persistent challenges in balancing high accuracy with computational efficiency in resource constrained and real-time scenarios. To address these challenges, we propose a lightweight method named CDO-POSE based on an improved YOLOv11. Specifically, we first introduce the Context Anchor Attention (CAA) module, which is composed of three convolutional layers and two bottleneck modules to enhance feature representation while maintaining parameter efficiency. Building on this, to address the limited precision of traditional nearest-neighbor upsampling, we incorporate the Dynamic Sampling (DySample) method, which adaptively adjusts the sampling strategy according to feature importance, thereby improving upsampling accuracy. Furthermore, to align the training objective more closely with the goal of precise pose estimation, we employ the Object Keypoint Similarity Loss (OKS-Loss), which provides a more discriminative evaluation of keypoint localization errors. The experiments on MS COCO2017 and CrowdPose datasets demonstrate that our model achieves almost the same accuracy as YOLOv11s-pose with significantly fewer parameters. Moreover, the model achieves 39.79 FPS and 29.23 FPS for inference at 480p and 720p, respectively, on the NVIDIA Jetson Orin Nano, suggesting that it is suitable for real-time deployment on edge devices. Full article

Remote sensing images are highly susceptible to spatially non-uniform haze under complex atmospheric conditions, leading to contrast degradation and structural detail loss. Moreover, remote sensing scenes usually exhibit complex spatial structures, highly uneven haze distribution, and significant statistical variability, which further increases the difficulty of haze removal. To address this issue, we revisit the haze degradation mechanism of remote sensing imagery and propose a dynamic histogram-sorted Transformer dehazing method from the perspectives of statistical distribution modeling and region-adaptive restoration. Specifically, a Histogram-Sorted Adaptive Attention is designed to map spatial features into the statistical distribution domain through a dynamic histogram sorting mechanism, enabling explicit discrimination and precise modeling of regions with different haze densities. Meanwhile, a Perception-Adaptive Feed-Forward Network is constructed, which incorporates a stable routing-based mixture-of-experts mechanism to adaptively select restoration strategies according to local texture characteristics and global haze density, thereby significantly enhancing the adaptability of the model in complex remote sensing scenarios. Extensive experimental results demonstrate that the proposed method achieves superior performance over existing approaches across multiple remote sensing benchmark datasets, effectively improving both visual quality and robustness of remote sensing imagery. Full article

This study addresses the automatic classification of sugar beet seeds according to their spraying levels using RGB images, aiming to enable a fast, practical, and non-destructive early warning system without chemical analysis. A dataset of 16,519 seed images acquired under controlled lighting conditions was used to evaluate YOLOv8-CLS and YOLO11-CLS architectures, including the n, s, m, l, and x scale variants within the Ultralytics framework. All experiments were conducted using a 10-fold cross-validation strategy, with models trained under different batch size and learning rate configurations. The results indicate that both architectures achieve reliable performance, with accuracy values ranging from approximately 78–83% for YOLOv8-CLS and 80–82% for YOLO11-CLS models. ROC-AUC scores consistently above 0.94 demonstrate strong inter-class discrimination. Misclassification analysis shows that errors mainly occur between visually similar intermediate treatment levels, particularly 25% and 50%. Despite this challenge, low log-loss values and balanced precision–recall profiles indicate stable decision behavior. Overall, the findings confirm that sugar beet seed treatment levels can be effectively distinguished using only RGB imagery, providing a potentially low-cost and scalable approach for early warning and quality control in seed treatment processes. Full article

Background: Parkinson’s disease (PD) is characterized by progressive degeneration of dopaminergic neurons in the substantia nigra pars compacta, accompanied by neuromelanin loss. Neuromelanin-sensitive magnetic resonance imaging (NM-MRI) enables in vivo visualization of these changes; however, its diagnostic and clinical utility remains incompletely defined. This study evaluated the feasibility, reliability, and biological sensitivity of semi-automated NM-MRI–based substantia nigra volumetry in PD. Methods: In this prospective case–control study, 50 participants (25 PD patients and 25 healthy controls) underwent 3T NM-sensitive MRI using a high-resolution T1-weighted spin-echo sequence. Semi-automated segmentation of hyperintense substantia nigra regions was performed using Mango v3.5.1, with intracranial volume normalization derived from FreeSurfer v7.3. Four participants were excluded due to motion artifacts, yielding a final cohort of 46 subjects. Clinical assessment included the Unified Parkinson’s Disease Rating Scale (UPDRS) Part III and Hoehn and Yahr (H&Y) staging. Group comparisons, receiver operating characteristic (ROC) analysis, and reliability testing using intraclass correlation coefficients (ICC) were performed. Results: Corrected substantia nigra volume was significantly reduced in PD patients compared with controls (18% reduction; p = 0.039, Mann–Whitney U test). Semi-automated measurements demonstrated excellent agreement with manual segmentation (ICC = 0.945). ROC analysis showed moderate discriminative performance for corrected volume (AUC = 0.700; sensitivity 68.4%, specificity 74.1%). No significant correlation was observed between corrected substantia nigra volume and UPDRS-III motor scores, while a trend toward lower SNc volume was observed with advancing H&Y stage. Conclusions: Semi-automated NM-MRI volumetry detects biologically meaningful substantia nigra volume loss in early-stage Parkinson’s disease with high measurement reliability. However, diagnostic performance was moderate and insufficient for standalone clinical diagnosis or motor severity prediction. These findings support the role of NM-MRI as a complementary imaging marker within multimodal diagnostic and research frameworks rather than as an independent diagnostic tool. Full article