Search Results (40,268)

This study investigates model-based emergency strategies for damage prevention following cable breaks in six-degree-of-freedom cable-driven parallel robots. By relaxing real-time constraints, the empirical performance limit of existing and novel recovery approaches is assessed through comprehensive simulations of the SEGESTA prototype. While extending prediction horizons improves recovery, it induces excessive computational loads. A hybrid approach combining State-of-the-Art approaches emerges as the optimal compromise, preserving high success rates while reducing calculation time and enabling active obstacle avoidance. Furthermore, a time-to-crash analysis after cable breaks demonstrates that the physical CDPR design fundamentally influences post-failure dynamics and could lead to narrow time windows for recovery approaches. Consequently, model-based emergency strategies are approaching their limits. Future research must overcome this by synergizing data-driven control paradigms with structurally fault-tolerant robot designs to ensure robust, real-time damage prevention after cable breaks. Full article

The rapid growth of Generative Adversarial Networks (GANs) has made synthetic media incredibly realistic and led to the emergence of deepfakes that can harm cybersecurity, trust in the digital world, and information integrity. The current state of the art in deepfake detection faces several issues, such as limited generalization across several manipulation methods and compressed media, and poor spatiotemporal learning capabilities. To solve these problems, this research develops a novel method that incorporates a robust spatiotemporal learning model that uses an Extreme Inception Network (XceptionNet) architecture to extract spatial features, a Bidirectional Long Short-Term Memory (BiLSTM) network for temporal sequence learning, and a temporal attention mechanism for adaptive frame analysis. The proposed solution was tested on a popular dataset, FaceForensics++, which includes multiple types of GAN-based manipulation, such as DeepFakes, Face2Face, FaceSwap, FaceShifter, and NeuralTextures. The results show improved accuracy in both binary and multiclass classification, along with Explainable Artificial Intelligence tools such as Gradient-weighted Class Activation Mapping (Grad-CAM) and SHapley Additive exPlanations (SHAP). Full article

Accurate age estimation and gender classification from facial images remain challenging tasks owing to substantial variations in facial appearance caused by pose, illumination, expression, occlusion, image quality, and age-related morphological changes. Although transformer-based architectures have recently demonstrated remarkable performance in facial analysis, effectively exploiting their complementary representation capabilities within a unified multi-task framework remains an open research problem. To address this limitation, this study proposes a transformer-based Multi-task Learning (MTL) framework that jointly performs age estimation and gender classification using a shared feature extraction backbone and task-specific prediction heads. Three state-of-the-art vision transformers, namely Vision Transformer (ViT), Shifted Window Transformer (Swin Transformer), and Data-efficient Image Transformer (DeiT), were independently implemented within the proposed MTL architecture to learn both global contextual information and local facial characteristics. Furthermore, two ensemble learning strategies, namely standard ensemble and Task-aware Weighted Ensemble (TAWE), were developed to exploit the complementary strengths of the individual transformer models. Experiments were conducted on the publicly available UTKFace dataset comprising 15,503 facial images, which were divided into 10,385 training, 1833 validation, and 3285 test samples, while gender balancing was applied during model training to alleviate class imbalance. Age estimation performance was evaluated using Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE), whereas gender classification performance was assessed using Accuracy and F1-score. Experimental results demonstrate that the proposed TAWE achieved the best point estimates among the evaluated ensemble strategies, obtaining an MAE of 4.8905, an RMSE of 6.4674, a gender classification accuracy of 94.86%, and an F1-score of 95.17% for the dataset-defined positive class. Compared with equal-weight averaging, TAWE provided a statistically significant improvement in gender classification accuracy, while the improvement in age estimation was modest and not statistically significant. Full article

Although the central panel of The Garden of Earthly Delights is perceived as a coherent visual whole, it lacks an obvious compositional center and is commonly interpreted as a polycentric image. This study argues that its visual organization is nevertheless governed by a reconstructable functional hierarchy. To investigate this structure, the study develops a functional method based on the analysis of compositional operators and the viewer’s perceptual route. The method is first applied to selected fifteenth-century paintings and then used to examine the central panel of Bosch’s triptych. The analysis demonstrates that Bosch preserves the traditional principle of compositional hierarchy while transforming its spatial organization: instead of unfolding along the vertical axis, the hierarchy is reconfigured through pictorial depth. The study therefore offers a new explanation of the compositional coherence of the central panel. It also extends the concept of visual engineering by treating compositional hierarchy as both a mechanism of visual organization and an independent object of art-historical analysis. Full article

This essay examines and compares Watteau’s approach to crafting satyrs and the human males that resemble them both in body and, especially, in their impulsive, animalistic actions. The demi-gods known as satyrs were a familiar presence in European art and literature extending from classical antiquity; their close association with fertility and lewdness evidently stemmed directly from their half-goat, half-human corporality. Al-though as an artist specializing in pastoral scenes of fashionable and theatrical sociability, Watteau did not have many occasions to paint mythological subjects, satyrs occupy a notable place among his drawings. These include both studies after other artists and an unusual group of drawings from the nude male model. I compare Watteau’s treatment of the satyr with certain aggressive males in his fêtes galantes, arguing that these satyric characters destabilize the humanism of his gatherings. Watteau’s improvisatory approach to composition moreover fostered an unusually egalitarian treatment of human and nonhuman animals. Full article

Automated analysis of fundus images is a promising technique for efficient, scalable retinal screening. In real-time screening, binary classification provides a straightforward first-line solution by differentiating normal and abnormal images, enabling prioritization of potentially abnormal cases for subsequent medical and disease-specific analysis. This study proposes RetinoMamba-AnatomyQuery, a lightweight model that integrates Vision Mamba-based global contextual representations, EfficientNet-B3-derived local features, anatomy–lesion query refinement, and anatomy-aware feature fusion. The Ocular Disease Intelligent Recognition (ODIR)-2019 dataset and Brazilian Multilabel Ophthalmological Dataset (BRSET) are harmonized to construct the internal cohort, yielding 10,842 eligible unique patients following expert-assisted quality control and data curation. The proposed model achieves accuracy of 97.93%, recall of 97.77%, specificity of 98.07%, precision of 97.87%, and an F1-score of 97.82%, outperforming state-of-the-art models. Independent evaluation on the external dataset (Retinal Fundus Multi-disease Image Dataset (RFMiD)) reports an accuracy of 96.21% and an F1-score of 95.99%, demonstrating sustained performance across an independent data distribution. Computational evaluation demonstrates a controlled footprint of 23.9 million parameters and 8.3 giga floating-point operations, supporting computationally efficient screening. Overall, the findings demonstrate the effectiveness of anatomy-conditioned global–local feature integration for reliable normal/abnormal fundus discrimination. Full article

Large-scale image databases are continuously being accumulated across diverse information systems, and JPEG is the de facto standard for storing full-color images. We consider the problem of reducing the storage expenses of large volumes of photorealistic JPEG images with controlled distortion levels and without significant degradation in subsequent classification accuracy of convolutional neural networks. To address this problem, we apply the Discrete Atomic Compression (DAC) algorithm, which is based on atomic functions. Prior work has shown that this method combines compression with image content protection in a way that is relevant to modern secure image processing. For comparison, we also include a Progressive DCT-based Coder (PDCTC). Experiments are conducted on the ImageNet validation set (50,000 images across 1000 classes). Compression performance is assessed using MAD, RMS, PSNR, and compression ratio (CR). The compressed images are then classified by 28 pre-trained CNN classifiers spanning eight architectural families from established (VGG, ResNet) to state-of-the-art (EfficientNetV2) models. Classification is evaluated using Top-1 and Top-5 accuracy, among other metrics. DAC achieves a mean compression ratio of 1.93 relative to JPEG and PDCTC achieves 1.58. Both methods provide high reconstruction quality (mean PSNR > 36 dB). Across all classifiers, the Top-1 accuracy drop remains below 1.73 percentage points for DAC and 1.39 for PDCTC. Thus, DAC is promising for large-scale storage of JPEG images, offering stronger compression at the cost of a marginally larger classification accuracy drop compared to PDCTC. Full article

This study examined the relationships among university students’ social-emotional learning (SEL) skills, emotional intelligence, emotion regulation, and learning motivation, with particular attention to a proposed serial indirect relationship through emotional intelligence and emotion regulation. A cross-sectional correlational design was employed with 372 students from the Faculties of Education and Arts and Sciences at a university in Türkiye, selected through stratified random cluster sampling. Data were analyzed using observed-variable path analysis in AMOS; the specific serial indirect effect in the original theoretical model was additionally estimated with PROCESS using bootstrap confidence intervals. SEL skills were positively related to learning motivation and emotional intelligence, emotional intelligence was positively related to emotion regulation, and emotion regulation was positively related to learning motivation. However, the direct relationships between SEL skills and emotion regulation and between emotional intelligence and learning motivation were not statistically significant. In the original theoretical model, the specific serial indirect relationship from SEL skills to learning motivation through emotional intelligence and emotion regulation was small and not statistically significant (β = 0.036, 95% CI [−0.0003, 0.0750]); therefore, the hypothesized serial indirect relationship was not supported. A post hoc parsimonious model excluding the two nonsignificant direct paths revealed a small but statistically significant serial indirect relationship (β = 0.043, 95% CI [0.005, 0.082]). This exploratory pattern should be interpreted cautiously and requires replication through longitudinal or experimental research. The findings may also inform curriculum developers by identifying social-emotional dimensions that warrant consideration in needs analysis and future curriculum-design and intervention studies in higher education, particularly in teacher education. Full article

This research provides deep insights into the capabilities of encoder models, offering a unique, controlled benchmark of over 20 state-of-the-art models. The aim is to analyze the viability of a potentially advantageous recommender architecture that capitalizes on the progress made in natural language processing (NLP) without the drawbacks of more advanced deep learning models or generative decoder-only large language models. We present a zero-shot recommender architecture used in a training-free setting. In this setting, item embeddings are aggregated into weighted user profiles and scored by cosine similarity against the full candidate pool. This approach makes use of state-of-the-art encoder language models in a frozen, inference-only, context-free state. We employed a temporal evaluation designed to simulate the cold-start problem, ordering user interactions over time and progressively revealing later interactions. We also used three contrasting datasets with specific challenges, as well as baseline methods of similar complexity and resource demands, such as collaborative filtering or content-based TF-IDF methods. For our recommendation task, we also compared the local large language model with the encoder models in a limited experiment. Our results suggest that frozen embedding-based recommendation could be a viable option, particularly when considering the cost/performance ratio. However, model and method selection should be evaluated against domain-specific tasks and protocols, rather than being based on general intuitions about scaling, general-purpose benchmark scores, or even leave-one-out validation methods. Full article

Process-based arts interventions—engaging participants in active embodied creation rather than analysis of completed works—produce more substantive cognitive, emotional, and behavioural change than product-based methods, yet present a significant accessibility challenge: non-artists asked to engage in such practices often respond with resistance, discomfort, and self-doubt, particularly in higher education where embodied learning differs from assumptions about legitimate study formats. This article introduces ARTFUL—a six-stage framework informed by neuroscience research on physiological readiness for learning, cognitive scaffolding, and embodied cognition—which addresses this barrier through progressive pairing of cognitive understanding with embodied experience. The framework was developed iteratively in an exploratory qualitative case study format: earlier course editions employed a three-stage format and the resistance observed motivated the extension of the format. Qualitative data from 24 management students (pre-programme surveys, post-programme semi-structured interviews, creative diaries and reflective portfolios) were analysed thematically. Three self-reported domains of cognitive change emerged: enhanced attention and concentration, increased openness to novelty, and reduced cognitive rigidity. Students moved from initially dismissing embodied methods as inappropriate for higher education to actively integrating them into professional and academic contexts. These findings suggest that systematic scaffolding may support a self-reported shift from initial resistance towards perceived cognitive flexibility, offering a transferable approach for designing accessible process-based arts interventions for non-artist learners. Full article

This essay examines the Italian Mannerist artist Giorgio Vasari’s mythological depictions of Minerva, Venus, and the Ephesian Diana as a coherent allegorical program rooted in Florentine Neoplatonism. Through detailed visual and textual analysis of The Forge of Vulcan (1564, Gallerie degli Uffizi, Florence), The Toilet of Venus (1558, Staatsgalerie, Stuttgart), and Diana of Ephesus (1548, Casa Vasari, Arezzo), it shows that Vasari shifted away from narrative myth toward philosophical emblematic abstraction. These figures symbolize Mind, Beauty, and Nature, aligning with the Renaissance Neoplatonic philosophy of Marsilio Ficino (1433–1499) and his Ficinian hierarchy, in which the soul ascends from material diversity through aesthetic contemplation to rational unity. By integrating readings of Ficino’s writings with insights from Vasari’s 1568 Le Vite (Le Vite de’ più eccellenti pittori, scultori e architettori), this study situates Vasari within a Warburg-inspired iconological framework and proposes that the Mannerist style itself functions as a form of metaphysical rhetoric. Vasari’s reputation has long rested on two achievements: his architectural and pictorial contributions to the Medici court, and his authorship of Le Vite. His mythological imagery has seldom been examined as part of a sustained philosophical program. Rather than depicting specific mythological stories, Vasari condensed these goddesses into symbolic ideas. The resulting hierarchy—Nature (Diana), Beauty (Venus), and Reason (Minerva)—reflects the Ficinian ascent of the soul. In this way, Vasari engaged in a distinct sixteenth-century fusion between classical imagery and metaphysical meaning. Vasari’s Diana, Venus, and Minerva allude to a structured ascent from Nature through Beauty to Reason. Through Mannerist stylization and Renaissance Neoplatonic metaphysics, he transformed mythology into philosophical allegory. Vasari’s art visually enacts Ficino’s ascent (anados). The viewer encounters generative multiplicity, contemplative beauty, and sovereign intellect—a unique ordered ladder of being, depicted in paint. Full article

Previous studies propose methodologies to transform any European curriculum into STEAM methodologies, where, based on curricular analysis, the areas of opportunity most likely to form a STEAM project with a high number of connections can be determined. Recent updates of the STEAM model toward a humanistic framework, STEAM∀H (STEAM for all Humanity), allow us to state the objective of this article: to expand these methodologies by connecting them with the SDGs addressed by each thematic area in the curriculum, where the “H” concept (for all humanity) is determined by the SDGs tackled within the designed project. The methodology proposed in this article consists of 10 phases to analyze the connection with the SDGs. To guarantee the validity of the analysis, the procedure includes an item-coding phase, the evaluation of inter-rater reliability, and the construction of adjacency matrices for network analysis. As a case study to test methodological feasibility, this article analyzes the LOMLOE Primary Education curriculum in the Valencian Community to identify the potential for designing interdisciplinary projects under the STEAM∀H approach and the Sustainable Development Goals (SDGs). Although LOMLOE prioritizes the STEM model—excluding the arts and humanities—the analysis of 546 items from Science, Mathematics, and Art Education demonstrates that the regulations allow for their integration. After refining the sample into 32 thematic areas, 12 highly interconnected conceptual areas were identified. These conceptual areas are configured as potential ideal generative topics for Project-Based Learning (PBL) linked to the 2030 Agenda, which could foster the development of creativity and social commitment in students as a pedagogical implication. Full article

Unresectable locally advanced non-small cell lung cancer (LA-NSCLC) remains a therapeutic challenge, despite the paradigm shift established by the PACIFIC trial (concurrent chemoradiation followed by Durvalumab consolidation). A substantial number of patients still experience disease progression and significant toxicity. This narrative review explores innovative radiation therapy (RT) strategies and their integration with systemic therapy. Key innovations analyzed include: 1. Systemic Therapy Intensification: Induction chemo-immunotherapy or novel multi-agent consolidation regimens show promise, but managing increased toxicity remains critical. The integration of DNA damage response inhibitors with RT is also being actively explored to maximize radiosensitization. 2. RT Dose and Volume Modulation: Recent data suggests that dose escalation via highly conformal hypofractionation or stereotactic ablative body radiotherapy (SABR) may enhance efficacy, particularly when combined with immunotherapy. Conversely, risk-adapted de-escalation of dose or volumes is studied to reduce immunosuppression and treatment-related toxicity, allowing more patients, including the elderly and frail, to complete consolidation. 3. Organ-at-Risk (OAR) Sparing: Advanced techniques like intensity-modulated radiation therapy (IMRT), adaptive RT (ART), and proton beam therapy (PBT) are crucial for minimizing dose to OARs and new substructures, such as specific cardiac subregions and lymphoid tissues. This OAR sparing is strongly correlated with improved OS and reduced immunosuppression. In conclusion, this narrative review focuses on the future of LA-NSCLC treatment, which is moving toward personalized, precision RT, embracing both intensification and thoughtful de-escalation, supported by synergistic systemic agents, to maximize curative potential and minimize toxicity. Full article

Islatravir is a novel nucleoside reverse transcriptase translocation inhibitor recently approved for the treatment of HIV-1 infection in virologically suppressed PLWH. While we previously documented high-level phenotypic resistance to islatravir (ISL) in HIV-1 subtype C variants harboring combinations of type 2 thymidine analogue mutations (TAM2) along with the M184V resistance-associated mutation, the primary contributor to the observed phenotype remains unclear. This report dissects the relative contribution of M184V to ISL resistance in the context of TAM2-containing variants through systematic reversion analysis. We show that reverting mutant 184V to wildtype M184 resulted in substantial sensitization to ISL, with fold-change reductions ranging from 8.1 to 13.4-fold across three TAM2 genetic backgrounds. Contrary to the well-documented antagonism between M184V and TAMs in zidovudine resistance, these two resistance pathways exhibit cooperative rather than antagonistic interactions in the context of ISL resistance. These findings provide critical evidence for refinement of genotypic resistance interpretation algorithms for ISL and have direct implications for the future clinical management of ISL-based antiretroviral therapy in people living with HIV. Importantly, elucidating the mechanistic basis of M184V-TAM2 cooperation would establish a more predictive framework for understanding how complex resistance mutation patterns impact ISL susceptibility and guide treatment decisions in virologically experienced individuals. Considering the diminished use of TAM2-seletive antiretroviral drugs in contemporary ART regimens, islatravir is likely to be an effective treatment option. Full article

Additive Manufacturing for Electronics (AME) techniques are currently becoming more widespread as they allow for the rapid production of electronic devices with confined costs. In this context, AME represents a relatively novel fabrication technique when used for power electronics applications. The research presented in this paper is focused on the investigation of the feasibility of inkjet-based AME for the fabrication of structures and components for power electronics applications. Seven circuit models for a power module were developed and printed with a multimaterial 3D printer. The models were then analyzed to optimize their design in terms of dimensional accuracy. Furthermore, the electrical conductivity and the magnetic field of one of the devices were examined with numerical simulation tools to verify its effectiveness and efficiency when compared to factory-built power modules. The resulting devices could be considered for coupling and improving the functionality of a module built using non-AME technology, showing that custom-built electronic devices of this kind can be a valid upgrade to the current state of the art. Thus, AME-fabricated devices might be considered as potential enablers for next-generation wide-bandgap (WBG) devices. Full article