Sign in to use this feature.

Years

Between: -

Subjects

remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline

Journals

remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline

Article Types

Countries / Regions

remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline
remove_circle_outline

Search Results (929)

Search Parameters:
Keywords = visual graph

Order results
Result details
Results per page
Select all
Export citation of selected articles as:
21 pages, 2678 KB  
Article
TopoTempNet: A High-Accuracy and Interpretable Decoding Method for fNIRS-Based Motor Imagery
by Qiulei Han, Hongbiao Ye, Yan Sun, Ze Song, Jian Zhao, Lijuan Shi and Zhejun Kuang
Sensors 2025, 25(17), 5337; https://doi.org/10.3390/s25175337 - 28 Aug 2025
Abstract
Functional near-infrared spectroscopy (fNIRS) offers a safe and portable signal source for brain–computer interface (BCI) applications, particularly in motor imagery (MI) decoding. However, its low sampling rate and hemodynamic delay pose challenges for temporal modeling and dynamic brain network analysis. To address these [...] Read more.
Functional near-infrared spectroscopy (fNIRS) offers a safe and portable signal source for brain–computer interface (BCI) applications, particularly in motor imagery (MI) decoding. However, its low sampling rate and hemodynamic delay pose challenges for temporal modeling and dynamic brain network analysis. To address these limitations in temporal dynamics, static graph modeling, and feature fusion interpretability, we propose TopoTempNet, an innovative topology-enhanced temporal network for biomedical signal decoding. TopoTempNet integrates multi-level graph features with temporal modeling through three key innovations: (1) multi-level topological feature construction using local and global functional connectivity metrics (e.g., connection strength, density, global efficiency); (2) a graph-modulated attention mechanism combining Transformer and Bi-LSTM to dynamically model key connections; and (3) a multimodal fusion strategy uniting raw signals, graph structures, and temporal representations into a high-dimensional discriminative space. Evaluated on three public fNIRS datasets (MA, WG, UFFT), TopoTempNet achieves superior accuracy (up to 90.04% ± 3.53%) and Kappa scores compared to state-of-the-art models. The ROC curves and t-SNE visualizations confirm its excellent feature discrimination and structural clarity. Furthermore, the statistical analysis of graph features reveals the model’s ability to capture task-specific functional connectivity patterns, enhancing the interpretability of decoding outcomes. TopoTempNet provides a novel pathway for building interpretable and high-performance BCI systems based on fNIRS. Full article
(This article belongs to the Special Issue (Bio)sensors for Physiological Monitoring)
Show Figures

Figure 1

23 pages, 3314 KB  
Article
Optimization of Manifold Learning Using Differential Geometry for 3D Reconstruction in Computer Vision
by Yawen Wang
Mathematics 2025, 13(17), 2771; https://doi.org/10.3390/math13172771 - 28 Aug 2025
Abstract
Manifold learning is a significant computer vision task used to describe high-dimensional visual data in lower-dimensional manifolds without sacrificing the intrinsic structural properties required for 3D reconstruction. Isomap, Locally Linear Embedding (LLE), Laplacian Eigenmaps, and t-SNE are helpful in data topology preservation but [...] Read more.
Manifold learning is a significant computer vision task used to describe high-dimensional visual data in lower-dimensional manifolds without sacrificing the intrinsic structural properties required for 3D reconstruction. Isomap, Locally Linear Embedding (LLE), Laplacian Eigenmaps, and t-SNE are helpful in data topology preservation but are typically indifferent to the intrinsic differential geometric characteristics of the manifolds, thus leading to deformation of spatial relations and reconstruction accuracy loss. This research proposes an Optimization of Manifold Learning using Differential Geometry Framework (OML-DGF) to overcome the drawbacks of current manifold learning techniques in 3D reconstruction. The framework employs intrinsic geometric properties—like curvature preservation, geodesic coherence, and local–global structure correspondence—to produce structurally correct and topologically consistent low-dimensional embeddings. The model utilizes a Riemannian metric-based neighborhood graph, approximations of geodesic distances with shortest path algorithms, and curvature-sensitive embedding from second-order derivatives in local tangent spaces. A curvature-regularized objective function is derived to steer the embedding toward facilitating improved geometric coherence. Principal Component Analysis (PCA) reduces initial dimensionality and modifies LLE with curvature weighting. Experiments on the ModelNet40 dataset show an impressive improvement in reconstruction quality, with accuracy gains of up to 17% and better structure preservation than traditional methods. These findings confirm the advantage of employing intrinsic geometry as an embedding to improve the accuracy of 3D reconstruction. The suggested approach is computationally light and scalable and can be utilized in real-time contexts such as robotic navigation, medical image diagnosis, digital heritage reconstruction, and augmented/virtual reality systems in which strong 3D modeling is a critical need. Full article
Show Figures

Figure 1

26 pages, 7962 KB  
Article
IntegraPSG: Integrating LLM Guidance with Multimodal Feature Fusion for Single-Stage Panoptic Scene Graph Generation
by Yishuang Zhao, Qiang Zhang, Xueying Sun and Guanchen Liu
Electronics 2025, 14(17), 3428; https://doi.org/10.3390/electronics14173428 - 28 Aug 2025
Abstract
Panoptic scene graph generation (PSG) aims to simultaneously segment both foreground objects and background regions while predicting object relations for fine-grained scene modeling. Despite significant progress in panoptic scene understanding, current PSG methods face challenging problems: relation prediction often only relies on visual [...] Read more.
Panoptic scene graph generation (PSG) aims to simultaneously segment both foreground objects and background regions while predicting object relations for fine-grained scene modeling. Despite significant progress in panoptic scene understanding, current PSG methods face challenging problems: relation prediction often only relies on visual representations and is hindered by imbalanced relation category distributions. Accordingly, we propose IntegraPSG, a single-stage framework that integrates large language model (LLM) guidance with multimodal feature fusion. IntegraPSG introduces a multimodal sparse relation prediction network that efficiently integrates visual, linguistic, and depth cues to identify subject–object pairs most likely to form relations, enhancing the screening of subject–object pairs and filtering dense candidates into sparse, effective pairs. To alleviate the long-tail distribution problem of relations, we design a language-guided multimodal relation decoder where LLM is utilized to generate language descriptions for relation triplets, which are cross-modally attended with vision pair features. This design enables more accurate relation predictions for sparse subject–object pairs and effectively improves discriminative capability for rare relations. Experimental results show that IntegraPSG achieves steady and strong performance on the PSG dataset, especially with the R@100, mR@100, and mean reaching 38.7%, 28.6%, and 30.0%, respectively, indicating strong overall results and supporting the validity of the proposed method. Full article
Show Figures

Figure 1

24 pages, 1747 KB  
Article
HortiVQA-PP: Multitask Framework for Pest Segmentation and Visual Question Answering in Horticulture
by Zhongxu Li, Chenxi Du, Shengrong Li, Yaqi Jiang, Linwan Zhang, Changhao Ju, Fansen Yue and Min Dong
Horticulturae 2025, 11(9), 1009; https://doi.org/10.3390/horticulturae11091009 - 25 Aug 2025
Viewed by 274
Abstract
A multimodal interactive system, HortiVQA-PP, is proposed for horticultural scenarios, with the aim of achieving precise identification of pests and their natural predators, modeling ecological co-occurrence relationships, and providing intelligent question-answering services tailored to agricultural users. The system integrates three core modules: semantic [...] Read more.
A multimodal interactive system, HortiVQA-PP, is proposed for horticultural scenarios, with the aim of achieving precise identification of pests and their natural predators, modeling ecological co-occurrence relationships, and providing intelligent question-answering services tailored to agricultural users. The system integrates three core modules: semantic segmentation, pest–predator co-occurrence detection, and knowledge-enhanced visual question answering. A multimodal dataset comprising 30 pest categories and 10 predator categories has been constructed, encompassing annotated images and corresponding question–answer pairs. In the semantic segmentation task, HortiVQA-PP outperformed existing models across all five evaluation metrics, achieving a precision of 89.6%, recall of 85.2%, F1-score of 87.3%, mAP@50 of 82.4%, and IoU of 75.1%, representing an average improvement of approximately 4.1% over the Segment Anything model. For the pest–predator co-occurrence matching task, the model attained a multi-label accuracy of 83.5%, a reduced Hamming Loss of 0.063, and a macro-F1 score of 79.4%, significantly surpassing methods such as ASL and ML-GCN, thereby demonstrating robust structural modeling capability. In the visual question answering task, the incorporation of a horticulture-specific knowledge graph enhanced the model’s reasoning ability. The system achieved 48.7% in BLEU-4, 54.8% in ROUGE-L, 43.3% in METEOR, 36.9% in exact match (EM), and a GPT expert score of 4.5, outperforming mainstream models including BLIP-2, Flamingo, and MiniGPT-4 across all metrics. Experimental results indicate that HortiVQA-PP exhibits strong recognition and interaction capabilities in complex pest scenarios, offering a high-precision, interpretable, and widely applicable artificial intelligence solution for digital horticulture. Full article
Show Figures

Figure 1

13 pages, 2180 KB  
Article
Research on Knowledge Graph Construction and Application for Online Emergency Load Transfer in Power Systems
by Nan Lou, Shiqi Liu, Rong Yan, Ruiqi Si, Wanya Yu, Ke Wang, Zhantao Fan, Zhengbo Shan, Hongxuan Zhang, Xinyue Yu, Dawei Wang and Jun Zhang
Electronics 2025, 14(17), 3370; https://doi.org/10.3390/electronics14173370 - 25 Aug 2025
Viewed by 141
Abstract
Efficient emergency load transfer is crucial for ensuring the power system’s safe operation and reliable power supply. However, traditional load transfer methods that rely on human experience have limitations, such as slow response times and low efficiency, which make it difficult to address [...] Read more.
Efficient emergency load transfer is crucial for ensuring the power system’s safe operation and reliable power supply. However, traditional load transfer methods that rely on human experience have limitations, such as slow response times and low efficiency, which make it difficult to address complex and diverse fault scenarios effectively. Therefore, this paper proposes an emergency load transfer method based on knowledge graphs to achieve intelligent management and efficient retrieval of emergency knowledge. Firstly, a named entity recognition model based on ERNIE-BiGRU-CRF is constructed to automatically extract key entities and relationships from the load transfer plan texts, obtaining information such as fault names, fault causes, and operation steps. Secondly, a power system emergency load transfer knowledge graph is constructed based on the extracted structured knowledge, which is efficiently stored using a graph database and enables the visualization and interactive query of knowledge. Finally, real power system fault cases prove that the proposed method can effectively improve the retrieval efficiency of fault knowledge and provide intelligent support for online emergency load transfer decisions. Full article
Show Figures

Figure 1

18 pages, 1212 KB  
Article
Part-Wise Graph Fourier Learning for Skeleton-Based Continuous Sign Language Recognition
by Dong Wei, Hongxiang Hu and Gang-Feng Ma
J. Imaging 2025, 11(8), 286; https://doi.org/10.3390/jimaging11080286 - 21 Aug 2025
Viewed by 329
Abstract
Sign language is a visual language articulated through body movements. Existing approaches predominantly leverage RGB inputs, incurring substantial computational overhead and remaining susceptible to interference from foreground and background noise. A second fundamental challenge lies in accurately modeling the nonlinear temporal dynamics and [...] Read more.
Sign language is a visual language articulated through body movements. Existing approaches predominantly leverage RGB inputs, incurring substantial computational overhead and remaining susceptible to interference from foreground and background noise. A second fundamental challenge lies in accurately modeling the nonlinear temporal dynamics and inherent asynchrony across body parts that characterize sign language sequences. To address these challenges, we propose a novel part-wise graph Fourier learning method for skeleton-based continuous sign language recognition (PGF-SLR), which uniformly models the spatiotemporal relations of multiple body parts in a globally ordered yet locally unordered manner. Specifically, different parts within different time steps are treated as nodes, while the frequency domain attention between parts is treated as edges to construct a part-level Fourier fully connected graph. This enables the graph Fourier learning module to jointly capture spatiotemporal dependencies in the frequency domain, while our adaptive frequency enhancement method further amplifies discriminative action features in a lightweight and robust fashion. Finally, a dual-branch action learning module featuring an auxiliary action prediction branch to assist the recognition branch is designed to enhance the understanding of sign language. Our experimental results show that the proposed PGF-SLR achieved relative improvements of 3.31%/3.70% and 2.81%/7.33% compared to SOTA methods on the dev/test sets of the PHOENIX14 and PHOENIX14-T datasets. It also demonstrated highly competitive recognition performance on the CSL-Daily dataset, showcasing strong generalization while reducing computational costs in both offline and online settings. Full article
(This article belongs to the Special Issue Advances in Machine Learning for Computer Vision Applications)
Show Figures

Figure 1

20 pages, 1818 KB  
Article
Image Captioning Model Based on Multi-Step Cross-Attention Cross-Modal Alignment and External Commonsense Knowledge Augmentation
by Liang Wang, Meiqing Jiao, Zhihai Li, Mengxue Zhang, Haiyan Wei, Yuru Ma, Honghui An, Jiaqi Lin and Jun Wang
Electronics 2025, 14(16), 3325; https://doi.org/10.3390/electronics14163325 - 21 Aug 2025
Viewed by 413
Abstract
To address the semantic mismatch between limited textual descriptions in image captioning training datasets and the multi-semantic nature of images, as well as the underutilized external commonsense knowledge, this article proposes a novel image captioning model based on multi-step cross-attention cross-modal alignment and [...] Read more.
To address the semantic mismatch between limited textual descriptions in image captioning training datasets and the multi-semantic nature of images, as well as the underutilized external commonsense knowledge, this article proposes a novel image captioning model based on multi-step cross-attention cross-modal alignment and external commonsense knowledge enhancement. The model employs a backbone architecture comprising CLIP’s ViT visual encoder, Faster R-CNN, BERT text encoder, and GPT-2 text decoder. It incorporates two core mechanisms: a multi-step cross-attention mechanism that iteratively aligns image and text features across multiple rounds, progressively enhancing inter-modal semantic consistency for more accurate cross-modal representation fusion. Moreover, the model employs Faster R-CNN to extract region-based object features. These features are mapped to corresponding entities within the dataset through entity probability calculation and entity linking. External commonsense knowledge associated with these entities is then retrieved from the ConceptNet knowledge graph, followed by knowledge embedding via TransE and multi-hop reasoning. Finally, the fused multimodal features are fed into the GPT-2 decoder to steer caption generation, enhancing the lexical richness, factual accuracy, and cognitive plausibility of the generated descriptions. In the experiments, the model achieves CIDEr scores of 142.6 on MSCOCO and 78.4 on Flickr30k. Ablations confirm both modules enhance caption quality. Full article
Show Figures

Figure 1

21 pages, 1242 KB  
Article
Smart Monitoring and Management of Local Electricity Systems with Renewable Energy Sources
by Olexandr Kyrylenko, Serhii Denysiuk, Halyna Bielokha, Artur Dyczko, Beniamin Stecuła and Yuliya Pazynich
Energies 2025, 18(16), 4434; https://doi.org/10.3390/en18164434 - 20 Aug 2025
Viewed by 431
Abstract
Smart monitoring of local electricity systems (LESs) with sources based on renewable energy resources (RESs) from the point of view of the requirements of the functions of an intelligent system are hardware and software systems that can solve the tasks of both analysis [...] Read more.
Smart monitoring of local electricity systems (LESs) with sources based on renewable energy resources (RESs) from the point of view of the requirements of the functions of an intelligent system are hardware and software systems that can solve the tasks of both analysis (optimization) and synthesis (design, planning, control). The article considers the following: a functional scheme of smart monitoring of LESs, describing its main components and scope of application; an assessment of the state of the processes and the state of the equipment of generators and loads; dynamic pricing and a dynamic assessment of the state of use of primary fuel and/or current costs of generators; economic efficiency of generator operation and loads; an assessment of environmental acceptability, in particular, the volume of CO2 emissions; provides demand-side management, managing maximum energy consumption; a forecast of system development; an assessment of mutual flows of electricity; system resistance to disturbances; a forecast of metrological indicators, potential opportunities for generating RESs (wind power plants, solar power plants, etc.); an assessment of current costs; the state of electromagnetic compatibility of system elements and operation of electricity storage devices; and ensures work on local electricity markets. The application of smart monitoring in the formation of tariffs on local energy markets for transactive energy systems is shown by conducting a combined comprehensive assessment of the energy produced by each individual power source with graphs of the dependence of costs on the generated power. Algorithms for the comprehensive assessment of the cost of electricity production in a transactive system for calculating planned costs are developed, and the calculation of the cost of production per 1 kW is also presented. A visualization of the results of applying this algorithm is presented. Full article
(This article belongs to the Section A: Sustainable Energy)
Show Figures

Figure 1

26 pages, 1563 KB  
Article
EPCDescriptor: A Multi-Attribute Visual Network Modeling of Housing Energy Performance
by Hafiz Muhammad Shakeel, Shamaila Iram, Hafiz Muhammad Athar Farid, Richard Hill and Hassam ur Rehman
Buildings 2025, 15(16), 2929; https://doi.org/10.3390/buildings15162929 - 18 Aug 2025
Viewed by 291
Abstract
Conventional methods of studying houses’ Energy Performance Certificates (EPCs) typically fail to investigate the impact of interrelated contextual elements instead fixating exclusively on the specific attributes of individual houses. This study presents a new method that combines network graph analytics (NGA) with interactive [...] Read more.
Conventional methods of studying houses’ Energy Performance Certificates (EPCs) typically fail to investigate the impact of interrelated contextual elements instead fixating exclusively on the specific attributes of individual houses. This study presents a new method that combines network graph analytics (NGA) with interactive visual analytics to investigate hidden linkages at the individual house level. Our proposed platform collects and analyses data related to housing attributes, creates a network based on the links between these attributes, and employs sophisticated graph algorithms to provide visual representations. Users have the ability to dynamically choose postcodes, metrics, and attributes, which, in turn, generate layouts of networks that provide valuable insights. The visualisation utilises colour gradients and node metrics to improve the comprehensibility of energy performance areas. The platform enables homeowners and stakeholders to comprehend the interrelationships between aspects such as neighbouring housing features, and house infrastructure. The results prove the efficacy of the strategy by giving a collection of case studies that encompass various Energy Performance Certificates (EPCs) ranging from A to G. Each case study demonstrates the evolution of network architectures and visual assessments, showcasing the energy performance linked to certain EPC ratings. The platform offers a user-friendly interface for stakeholders to investigate and understand attribute relationships. Full article
(This article belongs to the Collection Sustainable Buildings in the Built Environment)
Show Figures

Figure 1

27 pages, 5515 KB  
Article
Optimizing Multi-Camera Mobile Mapping Systems with Pose Graph and Feature-Based Approaches
by Ahmad El-Alailyi, Luca Morelli, Paweł Trybała, Francesco Fassi and Fabio Remondino
Remote Sens. 2025, 17(16), 2810; https://doi.org/10.3390/rs17162810 - 13 Aug 2025
Viewed by 443
Abstract
Multi-camera Visual Simultaneous Localization and Mapping (V-SLAM) increases spatial coverage through multi-view image streams, improving localization accuracy and reducing data acquisition time. Despite its speed and generally robustness, V-SLAM often struggles to achieve precise camera poses necessary for accurate 3D reconstruction, especially in [...] Read more.
Multi-camera Visual Simultaneous Localization and Mapping (V-SLAM) increases spatial coverage through multi-view image streams, improving localization accuracy and reducing data acquisition time. Despite its speed and generally robustness, V-SLAM often struggles to achieve precise camera poses necessary for accurate 3D reconstruction, especially in complex environments. This study introduces two novel multi-camera optimization methods to enhance pose accuracy, reduce drift, and ensure loop closures. These methods refine multi-camera V-SLAM outputs within existing frameworks and are evaluated in two configurations: (1) multiple independent stereo V-SLAM instances operating on separate camera pairs; and (2) multi-view odometry processing all camera streams simultaneously. The proposed optimizations include (1) a multi-view feature-based optimization that integrates V-SLAM poses with rigid inter-camera constraints and bundle adjustment; and (2) a multi-camera pose graph optimization that fuses multiple trajectories using relative pose constraints and robust noise models. Validation is conducted through two complex 3D surveys using the ATOM-ANT3D multi-camera fisheye mobile mapping system. Results demonstrate survey-grade accuracy comparable to traditional photogrammetry, with reduced computational time, advancing toward near real-time 3D mapping of challenging environments. Full article
Show Figures

Graphical abstract

24 pages, 2268 KB  
Review
Raman Spectroscopy as a Tool for Early Identification of Tan Spot Disease and Assessment of Fungicide Response in Wheat
by Ioannis Vagelas
Agronomy 2025, 15(8), 1952; https://doi.org/10.3390/agronomy15081952 - 13 Aug 2025
Viewed by 386
Abstract
Tan spot disease, caused by Pyrenophora tritici-repentis, poses a significant threat to wheat production worldwide. Early detection and precise fungicide application are essential for effective disease management. This study explores the potential of Raman spectroscopy—specifically surface-enhanced Raman spectroscopy (SERS) and coherent anti-Stokes [...] Read more.
Tan spot disease, caused by Pyrenophora tritici-repentis, poses a significant threat to wheat production worldwide. Early detection and precise fungicide application are essential for effective disease management. This study explores the potential of Raman spectroscopy—specifically surface-enhanced Raman spectroscopy (SERS) and coherent anti-Stokes Raman scattering (CARS)—as non-invasive tools for identifying fungal infection and assessing wheat’s biochemical response to propiconazole treatment. The methodology is entirely theoretical; no laboratory experiments were conducted. Instead, all spectral graphs and figures were generated through a collaborative process between the author and Microsoft Copilot, which served as a rendering tool. These AI-assisted visualizations simulate Raman responses based on known molecular interactions and literature data. The results demonstrate the conceptual feasibility of Raman-based diagnostics for precision agriculture, offering a sustainable approach to disease monitoring and fungicide management. Full article
Show Figures

Figure 1

17 pages, 1534 KB  
Article
TSGformer: A Unified Temporal–Spatial Graph Transformer with Adaptive Cross-Scale Modeling for Multivariate Time Series
by Yan Chen, Cheng Li and Xiaoli Zhao
Systems 2025, 13(8), 688; https://doi.org/10.3390/systems13080688 - 12 Aug 2025
Viewed by 395
Abstract
Multivariate time series forecasting requires modeling complex and evolving spatio-temporal dependencies as well as frequency-domain patterns; however, the existing Transformer-based approaches often struggle to effectively capture dynamic inter-series correlations and disentangle relevant spectral components, leading to limited forecasting accuracy and robustness under non-stationary [...] Read more.
Multivariate time series forecasting requires modeling complex and evolving spatio-temporal dependencies as well as frequency-domain patterns; however, the existing Transformer-based approaches often struggle to effectively capture dynamic inter-series correlations and disentangle relevant spectral components, leading to limited forecasting accuracy and robustness under non-stationary conditions. To address these challenges, we propose TSGformer, a Transformer-based architecture that integrates multi-scale adaptive graph learning, adaptive spectral decomposition, and cross-scale interactive fusion modules to jointly model temporal, spatial, and spectral dynamics in multivariate time series data. Specifically, TSGformer constructs dynamic graphs at multiple temporal scales to adaptively learn evolving inter-variable relationships, applies an adaptive spectral enhancement module to emphasize critical frequency components while suppressing noise, and employs interactive convolution blocks to fuse multi-domain features effectively. Extensive experiments across eight benchmark datasets show that TSGformer achieves the best results on five datasets, with an MSE of 0.354 on Exchange, improving upon the best baselisnes by 2.4%. Ablation studies further verify the effectiveness of each proposed component, and visualization analyses reveal that TSGformer captures meaningful dynamic correlations aligned with real-world patterns. Full article
Show Figures

Figure 1

27 pages, 490 KB  
Article
Dynamic Asymmetric Attention for Enhanced Reasoning and Interpretability in LLMs
by Feng Wen, Xiaoming Lu, Haikun Yu, Chunyang Lu, Huijie Li and Xiayang Shi
Symmetry 2025, 17(8), 1303; https://doi.org/10.3390/sym17081303 - 12 Aug 2025
Viewed by 487
Abstract
The remarkable success of autoregressive Large Language Models (LLMs) is predicated on the causal attention mechanism, which enforces a static and rigid form of informational asymmetry by permitting each token to attend only to its predecessors. While effective for sequential generation, this hard-coded [...] Read more.
The remarkable success of autoregressive Large Language Models (LLMs) is predicated on the causal attention mechanism, which enforces a static and rigid form of informational asymmetry by permitting each token to attend only to its predecessors. While effective for sequential generation, this hard-coded unidirectional constraint fails to capture the more complex, dynamic, and nonlinear dependencies inherent in sophisticated reasoning, logical inference, and discourse. In this paper, we challenge this paradigm by introducing Dynamic Asymmetric Attention (DAA), a novel mechanism that replaces the static causal mask with a learnable context-aware guidance module. DAA dynamically generates a continuous-valued attention bias for each query–key pair, effectively learning a “soft” information flow policy that guides rather than merely restricts the model’s focus. Trained end-to-end, our DAA-augmented models demonstrate significant performance gains on a suite of benchmarks, including improvements in perplexity on language modeling and notable accuracy boosts on complex reasoning tasks such as code generation (HumanEval) and mathematical problem-solving (GSM8k). Crucially, DAA provides a new lens for model interpretability. By visualizing the learned asymmetric attention patterns, it is possible to uncover the implicit information flow graphs that the model constructs during inference. These visualizations reveal how the model dynamically prioritizes evidence and forges directed logical links in chain-of-thought reasoning, making its decision-making process more transparent. Our work demonstrates that transitioning from a static hard-wired asymmetry to a learned and dynamic one not only enhances model performance but also paves the way for a new class of more capable and profoundly more explainable LLMs. Full article
Show Figures

Figure 1

15 pages, 2953 KB  
Article
More than Just Figures: Structural and Visual Complexity in Soil Science Articles
by Agnieszka Wnuk and Dariusz Gozdowski
Appl. Sci. 2025, 15(15), 8724; https://doi.org/10.3390/app15158724 - 7 Aug 2025
Viewed by 288
Abstract
The structure of a scientific article is crucial for clearly conveying research findings. Modern scientific publications combine text with various elements—such as tables, graphs, images, diagrams, and maps—that support the narrative and aid data interpretation. Understanding how these components influence a publication’s reception [...] Read more.
The structure of a scientific article is crucial for clearly conveying research findings. Modern scientific publications combine text with various elements—such as tables, graphs, images, diagrams, and maps—that support the narrative and aid data interpretation. Understanding how these components influence a publication’s reception and scientific impact is essential. This study analyzes differences among 15 soil science journals (indexed in the Web of Science) in terms of visual elements, tables, number of authors, and article length. The journals had a 5-year Impact Factor (2023) ranging from 0.9 (Soil and Environment) to 10.4 (Soil Biology and Biochemistry). The Kruskal–Wallis test and Bonferroni-adjusted Dunn’s post hoc tests revealed statistically significant differences across all variables (p < 0.05). The relationships were further assessed using Pearson’s correlation, based on the median number of authors and article length, as well as the percentage of articles that include at least one element of a given type (e.g., table, graph, image, diagram, or map). Key findings show that journals with a higher impact factor tend to publish articles with more authors (r = 0.62, p = 0.014), use diagrams more frequently (r = 0.69, p = 0.004), and include fewer tables (r = –0.85, p < 0.001). These results suggest that journals with a higher 5-year IF tend to include articles with a greater number of authors and a higher frequency of diagram use, while relying less on tables. Full article
(This article belongs to the Section Agricultural Science and Technology)
Show Figures

Figure 1

29 pages, 16016 KB  
Article
An Eye Movement Monitoring Tool: Towards a Non-Invasive Device for Amblyopia Treatment
by Juan Camilo Castro-Rizo, Juan Pablo Moreno-Garzón, Carlos Arturo Narváez Delgado, Nicolas Valencia-Jimenéz, Javier Ferney Castillo García and Alvaro Alexander Ocampo-Gonzalez
Sensors 2025, 25(15), 4823; https://doi.org/10.3390/s25154823 - 6 Aug 2025
Viewed by 480
Abstract
Amblyopia, commonly affecting children aged 0–6 years, results from disrupted visual processing during early development and often leads to reduced visual acuity in one eye. This study presents the development and preliminary usability assessment of a non-invasive ocular monitoring device designed to support [...] Read more.
Amblyopia, commonly affecting children aged 0–6 years, results from disrupted visual processing during early development and often leads to reduced visual acuity in one eye. This study presents the development and preliminary usability assessment of a non-invasive ocular monitoring device designed to support oculomotor engagement and therapy adherence in amblyopia management. The system incorporates an interactive maze-navigation task controlled via gaze direction, implemented during monocular and binocular sessions. The device tracks lateral and anteroposterior eye movements and generates visual reports, including displacement metrics and elliptical movement graphs. Usability testing was conducted with a non-probabilistic adult sample (n = 15), including individuals with and without amblyopia. The System Usability Scale (SUS) yielded an average score of 75, indicating good usability. Preliminary tests with two adults diagnosed with amblyopia suggested increased eye displacement during monocular sessions, potentially reflecting enhanced engagement rather than direct therapeutic improvement. This feasibility study demonstrates the device’s potential as a supportive, gaze-controlled platform for visual engagement monitoring in amblyopia rehabilitation. Future clinical studies involving pediatric populations and integration of visual stimuli modulation are recommended to evaluate therapeutic efficacy and adaptability for early intervention. Full article
(This article belongs to the Section Biomedical Sensors)
Show Figures

Graphical abstract

Back to TopTop