Search Results (373)

Research on wine perception shows that extrinsic information can alter reported sensory perceptions. This study adopts a repeated measures design to quantify how label information affects perceived quality relative to blind tasting. Forty-five adult tasters evaluated eight PDO Málaga wines in two tasting contexts: blind and label-informed conditions. Participants rated the perceived quality of aroma, colour, and flavour using 5-point scales, capturing evaluative judgments rather than intensity. Mean global sensory scores and attribute-level evaluations were analysed to assess context-related effects. The results reveal a robust and consistent effect of informational context: label disclosure systematically increased perceived quality ratings compared with blind tasting. Difference-based analysis (informed − blind) indicates that this effect varies across wines and attributes, acting primarily as an amplification mechanism rather than fundamentally altering relative evaluations. At the same time, colour wines were clearly differentiated under blind conditions, confirming the presence of intrinsic sensory variation independent of contextual cues. Sensory attributes showed strong correlations (r > 0.95), indicating an integrated perceptual response, but also suggesting limited discriminant capacity among attributes and a tendency toward global evaluative judgments. No statistically significant differences were observed between consumers and sommeliers, although subgroup comparisons should be interpreted with caution given the limited size of the expert sample. Overall, the findings are consistent with expectation-driven perception, in which labelling cues correlate (r > 0.95), indicating an integrated perceptual response; however, this also suggests limited discriminant capacity between attributes and a tendency toward global evaluative judgments. No statistically robust differences were observed between consumers and sommeliers, although subgroup comparisons should be interpreted with caution due to the limited size of the expert sample. The study suggests the relevance of aligning sensory profiles with extrinsic communication strategies in wine marketing and PDO contexts, while acknowledging limitations in sample composition, measurement scales, and the use of self-reported sensory ratings. Full article

Automated fracture detection remains challenging because fracture regions often exhibit low contrast, blurred boundaries, large scale variations, and substantial morphological diversity. Although deep learning-based detectors show promising performance, they often suffer from limited category coverage and insufficient multi-scale feature representation. To address these issues, we propose LiteFracNet, a lightweight framework for accurate and efficient fracture detection. C3-CFormer enhances feature representation via residual aggregation, gated dynamic modeling, and long-range dependency extraction. C2Mona improves fine-grained fracture perception using multi-scale convolution and feature separation–reconstruction. The OmniKernel Fusion Pyramid Network (OFPN) promotes cross-level feature interaction and multi-scale information propagation, improving detection of subtle fractures. The Fusion-Enhanced Detection Head (FED-Head) employs channel alignment and shared convolutions to unify multi-scale features, reducing redundancy in conventional multi-branch heads. On the HBFMID, LiteFracNet achieves a mAP50 of 93.53% with 2.41 M parameters, 8.1 GFLOPs, a 4.9 MB model size, and 87.1 FPS inference speed. It also achieves a mAP50 of 61.43% on the pediatric GRAZPEDWRI-DX dataset, demonstrating competitive cross-dataset adaptability. These results indicate that LiteFracNet has the potential to efficiently assist computers in detecting fractures. Full article

This study presents a modular AI-assisted workflow for converting single 2D interior images into textured 3D assets and for evaluating those assets when ground-truth 3D meshes are unavailable. The proposed pipeline combines object detection, instance isolation, monocular-depth estimation, image-to-3D generation, texture synthesis, mesh export, and cloud-based execution to support early-stage interior-design and real-estate visualization tasks. A reference-free validation protocol is introduced, based on rendered multi-view comparisons, silhouette Intersection-over-Union, automated captioning, and multimodal embedding similarity, and is complemented by a composite validation framework that benchmarks reconstructed scenes against 200 panoramic indoor scenes from the Structured3D dataset using Hungarian-matched placement, size, recall, and relative-distance metrics. The workflow was implemented and tested using contemporary computer-vision and generative 3D components, with Hunyuan3D 2.0 used as the main reconstruction model. Proof-of-concept experiments on a representative corpus of 178 synthetically generated single-object images spanning a range of interior furniture categories show comparable silhouette IoU for textured and non-textured outputs and indicate that texture-preserving renderings improve visual and semantic similarity scores across CLIP-based evaluations. The 200-scene dataset evaluation reveals stable spatial localization (placement error ≈ 1.18 m, relative-distance error ≈ 0.54 m) alongside systematic over-prediction and size-calibration errors. Beyond the applied pipeline, the study contributes a reference-free, ground-truth-free protocol for 3D-asset evaluation and a first quantified account of where object-centric single-image reconstruction is reliable—spatial placement—and where it is not—object scale and spurious detection—at interior-scene scale. The results demonstrate the feasibility of integrating perception, 3D reconstruction, semantic assessment, and scalable deployment into a single applied pipeline, while remaining proof-of-concept and requiring extension to larger object and scene corpora, baselines, real-photograph evaluation, and human-centered assessment before broad claims about general interior-scene reconstruction can be made. Full article

Background/Objectives: Identifying pregnant women with heightened emotional vulnerability remains a challenge in antenatal care. A positive screening result is a comparatively infrequent event in clinical samples of moderate size, which complicates the statistical modelling of associated factors. This study aimed to characterise the psychosocial profile associated with a positive antenatal depression screen, integrating attachment style, perception of motherhood, previous emotional history and sleep problems. Methods: A total of 140 women in the third trimester of pregnancy were assessed cross-sectionally using the Edinburgh Postnatal Depression Scale (EPDS), with a threshold of 14 or above defining a positive screen; the Adult Attachment Scale; and single-item psychosocial indicators. Associations were examined using the chi-square test, the Fisher–Freeman–Halton exact test, and Firth penalized logistic regression, chosen to correct the separation arising from the small number of events. Results: The screen was positive in 15.0% of participants. Negative perception of motherhood, anxious–ambivalent attachment, previous emotional history and sleep disturbances were all significantly associated with depressive symptoms; in the multivariable model, only negative perception of motherhood remained independently associated with a positive screen. Conclusions: Structured psychosocial assessment may usefully complement standard EPDS screening in the early identification of antenatal emotional vulnerability. Full article

Artificial intelligence architecture is no longer adequately described by model size alone. Dense Transformers remain the reference architecture for language and multimodal reasoning, but production systems increasingly combine conditional computation, retrieval, memory, tools, verifiers, edge-cloud routing, observability and governance. This review makes three engineering claims. First, sparse Mixture-of-Experts models are currently the clearest capacity-scaling pattern, because they decouple total parameters from active per-token computation, although routing imbalance and distributed communication remain hard constraints. Second, state-space, recurrent and linear attention hybrids are best interpreted as attention-budgeting architectures: they reduce KV-cache and long-context costs, but do not yet displace dense attention in every reasoning regime. Third, JEPA-style latent world models change the learning objective from surface-token or pixel prediction to representation prediction, which is strategically important for perception and planning but still not a drop-in replacement for general language interfaces. To make the maturity claims auditable, this review uses a PRISMA-inspired search protocol, an explicit technology readiness rubric, quantitative comparison tables, hardware and memory-bandwidth analysis, deployment and reproducibility categories, and failure cases for RAG and agents. The main conclusion is that the optimal architecture is task- and constraint-dependent: small dense or hybrid models are often preferred for real-time edge inference, RAG and graph memory for changing enterprise knowledge, frontier dense or sparse models for difficult reasoning, and agentic workflows only when tool permissions, rollback, provenance and human oversight are engineered as first-class components. Full article

In schools located in the Colombian Amazon, there is still limited empirical evidence on how students connect pro-environmental values with emotions and environmental peace culture. In this article, environmental peace culture is understood as a way of learning to care for nature, live with it, and assume shared responsibility for its protection. The study investigated “Sembrando Paz Ambiental”, a place-based pedagogical proposal implemented among secondary school students through experiential, playful, and participatory activities. A convergent mixed-methods design was employed, integrating a quasi-experimental pretest–posttest component with a qualitative analysis of students’ perceptions and pedagogical productions. The quantitative component included 66 paired cases assessed through a 24-item Likert scale, with data analyzed using the Wilcoxon signed-rank test. The qualitative component included open-ended responses, drawings, and written productions analyzed through thematic coding and co-occurrence analysis in ATLAS.ti. The findings revealed statistically significant increases across all pro-environmental values, with the hedonic value showing the largest effect size. Qualitative findings associated environmental peace culture with emotional well-being, cooperation, care for nature, and territorial belonging. The study contributes to the literature by showing that the mobilization of pro-environmental values depends not only on normative conservation discourses, but also on pedagogical experiences that connect emotion, territory, and environmental care. Full article

Multimodal object detection leveraging RGB and infrared imagery has become essential for robust all-weather perception in unmanned aerial vehicle (UAV) applications. However, existing methods still struggle with effective cross-modal feature fusion, spatial misalignment between modalities, and scale variation of objects in aerial views. In this paper, we propose CMGFDet, a Cross-Modal Gated Fusion Network with Multi-Receptive Field Aggregation designed for RGB–infrared aerial object detection. Our framework introduces three coordinated modules: (1) a Cross-Modal Feature Fusion Network (CMFFN) that employs a gated attention mechanism to selectively aggregate complementary information from both modalities during encoding; (2) a Global–Local Attention Module (GLAM) that performs hierarchical cross-modal feature alignment by jointly modelling global channel statistics and local spatial correlations in the decoder; and (3) a Multi-Receptive Field Aggregation Network (MRFAN) that captures multi-scale contextual information through parallel depthwise convolutions with diverse kernel sizes. Additionally, we incorporate a deep supervision strategy and a composite loss function to enhance training efficiency. Extensive experiments on four public benchmarks (DroneVehicle, RGBTDronePerson, VEDAI, and VTUAV) show that CMGFDet improves the previous best mAP@0.5 by 1.6%, 2.2%, 1.9%, and 2.2%, respectively. The implementation code will be released upon acceptance to support reproducibility. Full article

Precise prediction of voltage degradation is critical for the prognostics and health management of proton exchange membrane fuel cells (PEMFCs). The performance degradation of PEMFCs is governed by a complex coupling of multiple physicochemical mechanisms, including catalyst layer and proton exchange membrane degradation. Crucially, these internal degradation processes evolve across highly heterogeneous time scales, ranging from transient high-frequency fluctuations to low-frequency and long-term irreversible performance fade. Conventional predictive models, which typically rely on single-scale architectures or fixed receptive fields, are inherently ill-equipped to simultaneously decouple and capture these cross-scale temporal dynamics. To tackle this challenge, this paper innovatively proposes a multi-scale deep learning framework that integrates a multi-scale degradation trend perception module, a long short-term memory (LSTM)-based encoder–decoder architecture, and a multi-head attention mechanism. One-dimensional convolutional layers with different kernel sizes are employed to simultaneously extract local temporal features at multiple granularities, followed by the LSTM encoder–decoder to model long-range temporal dependencies, while the cross-attention mechanism dynamically allocates attention across the encoded context at each autoregressive decoding step. Experimental outcomes indicate that the proposed model realizes excellent predictive accuracy across five evaluation indices in comparison with standard baselines. In particular, the mean absolute percentage error (MAPE) reaches 1.6696%, and the maximum absolute percentage error (Max-APE) is strictly bounded within 5%, substantiating the reliability of the proposed framework for high precision and long-horizon health prognostics for PEMFCs. Full article

Algorithm-aversion research predicts that consumers prefer human curators to algorithmic ones in subjective decision domains, including taste-based hedonic recommendation. Drawing on the algorithmic-symbiosis paradigm and on assortment-perception theory, this paper identifies a boundary condition on that prediction: disclosing that recommendations are AI-curated rather than human-curated lifts purchase intention in hedonic e-commerce but not in utilitarian e-commerce. The mechanism is a search-side option-breadth inference—the consumer’s attribution about the size of the option pool the curator considered upstream—which is diagnostic in preference-formative consumption categories where consumers build, rather than match, a preference. Three online experiments deployed through a Chinese consumer panel test the framework. Study 1 ($N=228$) finds the predicted Disclosure × Product-type interaction (${\eta }_p^2=0.025$) with the AI-versus-human lift confined to the hedonic cell ($d=0.65$). Study 2 ($N=257$) isolates the option-breadth pathway against trust and competence as competing mediators. Study 3 ($N=519$) extends the design to a second hedonic category, decomposes option breadth into search-side and display-side subdimensions through an eight-item bi-factor scale, tests mentalizing alongside option breadth as a competing mediator, and brings the moderated-mediation test by consumer AI familiarity to conventional statistical power (index of moderated mediation $=+0.065$, 95% CI $[+0.014,+0.118]$). Implications for interactive-marketing practice and algorithmic-disclosure regulation are discussed. Full article

The integration of AI-generated content into higher education has intensified interest in how students form and calibrate trust toward algorithmic outputs and whether pedagogical relationships can serve as conduits for that trust. Grounded in Trust Transfer Theory, this between-subjects randomized experimental study (N = 320) investigated whether teacher endorsement shapes students’ perceptions of AI-generated educational content across four dimensions: Perceived AI Competence, Academic Integrity, Perceived Human-Mediated Reliability, and Behavioral Intention to adopt. Participants from Saudi Arabian universities were randomly assigned to an endorsed or non-endorsed vignette condition and responded to a validated 13-item Trust and Acceptance Scale. Independent-samples t-tests confirmed statistically significant differences across all four dimensions in favor of the endorsed condition, with effect sizes ranging from small to large, and findings remained robust after controlling for gender via ANCOVA. Within-condition regression analyses further established Perceived Human-Mediated Reliability as a structurally stable positive predictor of trust outcomes in both conditions, with predictive power consistently amplified under endorsement. Postgraduate students placed greater emphasis on human oversight, while no disciplinary differences emerged, confirming the cross-disciplinary universality of the trust transfer mechanism. These findings are consistent with positioning the teacher as a trust guarantor in AI-mediated learning environments and carry direct implications for pedagogical design and institutional AI governance. Full article

Accurate detection of grain pests is important for ensuring food security. The existing detection algorithms still face challenges such as low precision, high false positives, and missed detection when identifying storage pests due to the small size, complex backgrounds, and limited data availability. For this reason, this paper proposes a solution jointly driven by generative AI and analytical AI, named REGAN-BS-YOLOv8. Specifically, on the one hand, an image-generative AI method of Real_ESRGAN and the mosaic data augmentation method is introduced to improve the quality, resolution and number of grain pest images. On the other hand, a novel recognition algorithm is proposed by introducing Swin Transformer and BiFPN to YOLOv8n to obtain better perception ability, positioning accuracy, detail retention and edge clarity, improving the recognition accuracy, robustness, and generalization of stored-grain pest and the anti-interference capability in complex environments. In addition, a new dataset including 5818 images of nine types of grain pests at various scales and under diverse backgrounds was collected in this study. The experimental results show that the mAP@0.5, Precision and Recall are 97.7%, 96.2% and 95.4% respectively using the method proposed in this study, outperforming the other nine models. The performance of the model we proposed is excellent for different types of storage pests under different backgrounds, and the model identifies grain pests of different sizes in different environments with high precision and provides an important theoretical basis for the management of grain silos. Full article

Background: Systematic among patients with bilateral vocal fold paralysis (BVFP), the most clinically significant symptom is dyspnea, which often requires surgical intervention. One of the most common causes of BVFP is iatrogenic injury during thyroidectomy. Some patients report not only day time breathing difficulties but also a deterioration in sleep quality since the onset of BVFP. Obstructive sleep apnea (OSA) is a common sleep-related breathing disorder caused by recurrent upper airway obstruction during sleep. Although it is usually associated with pharyngeal collapse, fixed laryngeal obstruction has also been suggested as a potential contributor to sleep-disordered breathing. Therefore, we aimed to assess the impact of surgical treatment of OSA in patients with BVFP. Methods: Between 2022 and 2025, 18 patients with BVFP were screened. Patients who met diagnostic criteria for OSA based on preoperative polysomnography (PSG) and the Epworth Sleepiness Scale (ESS) AHI ≥ 15 events/h or AHI ≥ 5 events/h with associated symptoms were included in the study. Five female patients met the criteria and underwent arytenoidectomy combined with posterior cordectomy. A follow-up PSG and re-evaluation of ESS were performed 6–12 months after surgery, and patients were asked about their subjective perception of sleep quality improvement. Results: An improvement in sleep quality, decrease in the median ESS score and respiratory parameters showed an improvement, although this was not statistically significant due to the small sample size (n = 5). Conclusions: These results indicate a positive postoperative trend in sleep quality among patients with BVFP however, given the limited sample size, further studies are required to confirm these observations. Full article

Automated detection of lunar impact craters from digital elevation model (DEM) data is important for lunar geological analysis, landing-site selection, and crater catalog updating. However, this task remains challenging because lunar craters exhibit large scale variations, weak or degraded rims, ambiguous boundaries, and complex topographic backgrounds. In addition, large-scale lunar remote sensing applications require detection models to achieve a reasonable balance among accuracy, model complexity, and inference efficiency. To address these challenges, this study proposes FPW-YOLO11n, a frequency-perception crater detection method developed based on YOLO11n. First, a Frequency-Directional Attention Module (FDA-Module) is introduced into the shallow stage of the backbone. This module combines frequency-aware channel attention and direction-aware spatial attention to enhance the representation of crater rim structures, elevation variations, and directional topographic cues in DEM data. Second, a C2PSA-LRSA module is designed by embedding Local Region Self-Attention into the C2PSA framework, thereby improving local contextual feature interaction while reducing the excessive cost associated with global self-attention. Third, Inner-WIoU is adopted to replace the original CIoU loss in YOLO11n. By combining the auxiliary-box mechanism of Inner-IoU with the sample-quality-aware weighting strategy of WIoU, Inner-WIoU provides a more flexible bounding-box regression objective for craters with weak rims, scale variations, and uncertain boundaries. A DEM-based lunar crater dataset was constructed from the Moon LRO LOLA–SELENE Kaguya TC DEM Merge 60N60S 59m product and the Robbins lunar crater catalog, covering the non-polar region from 60° S to 60° N and containing 4760 image tiles. Under the random data-splitting strategy, FPW-YOLO11n achieves 78.3% Precision, 66.2% Recall, 75.1% mAP@0.5, and 50.2% mAP@0.5:0.95, outperforming the YOLO11n baseline by 1.2, 2.0, 1.6, and 4.0 percentage points, respectively. Additional experiments based on geographically disjoint data splitting further show that the proposed method consistently performs better than YOLO11n on DEM data, indicating that the proposed structural improvements remain effective under a more rigorous spatially independent evaluation setting. Although the computational cost increases from 6.3 to 24.0 GFLOPs, FPW-YOLO11n maintains a compact parameter size of 2.59 M and a high inference speed, demonstrating an improved accuracy–efficiency trade-off for lunar crater detection from DEM data. Full article

As unmanned aerial vehicles (UAVs) become central to traffic inspection, urban security, and emergency response, UAV-based environmental perception requires both high accuracy and real-time efficiency. However, UAV imagery remains challenging due to three primary factors: detail loss, where small targets occupy minimal pixels and weak edges are diluted by downsampling; ineffective cross-scale fusion, where semantic gaps between shallow and deep features lead to scale misalignment and small-object suppression; and environmental interference, where clutter, occlusion, and dense layouts cause localization drift. To address these challenges, we propose an optimized efficient detector built upon the YOLOv8s framework, incorporating multi-scale feature enhancement and saliency-guided cross-layer fusion. Specifically, we integrate RFCAConv and RGCSP modules into the backbone to strengthen local detail and spatial structure modeling. Furthermore, we design a Multi-Scale Adaptive Fusion Module (MSAFM) to align deep and shallow cues through dual-pooling and adaptive channel recalibration. To handle complex backgrounds, a Saliency-Guided Contextual Attention Module (CASM) is introduced to emphasize target regions, alongside a dynamic detection head for adaptive feature modulation. Evaluated on the VisDrone2019 dataset, our method achieves 48.3% mAP@0.5 and 29.0% mAP@[0.5:0.95], outperforming YOLOv8s by 10.2 and 6.3 points, respectively, while keeping the model compact with 7.2M parameters and a 14.4 MB model size. Full article

Background: Many individuals with traumatic brain injuries (TBIs) exhibit oculomotor dysfunctions that impact their daily functioning. As current clinical screening tools are limited, we have created and pilot-tested the Rehabilitation Oculomotor Screening Evaluation (ROSE) previously in a small sample of people with acquired brain injuries and neurotypical participants. The current study aims to validate ROSE in persons with TBI, focusing on mild TBI (mTBI). Methods: Participants with TBI (n = 25) completed different clinical scales, including ROSE, Sensory Organization Test (SOT) for standing balance, Reintegration to Normal Living Index (RNLI), Timed Up and Go (TUG) for mobility, and a visual analogue scale for the subjective perception of visual vertigo. Neurotypical individuals (n = 24) who were age- and sex-matched completed only ROSE. Results: The group with mTBI (n = 18) had significantly higher ROSE scores compared to the neurotypical group, with a large effect size. Significant correlation was found between ROSE and RNLI scores, but not with other clinical outcomes. Conclusions: Significant between-group difference in ROSE scores and their association with RNLI scores suggest that ROSE is a valid tool in detecting oculomotor dysfunction in TBI. Future studies should continue the validation of ROSE in other TBI and neurologic populations and in larger sample sizes. Full article