Search Results (1,784)

Gaze estimation plays a crucial role in human-computer interaction and behavior analysis. However, in dynamic scenes, rigid head movements and rapid gaze shifts pose significant challenges to accurate gaze prediction. Most existing methods either process single-frame images independently or rely on long video sequences, making it difficult to simultaneously achieve strong performance and high computational efficiency. To address this issue, we propose DGAGaze, a gaze estimation framework based on a difference-driven spatiotemporal attention mechanism. This framework uses a geometry-aware temporal alignment module to mitigate interference from rigid head movements, compensating for them through pose estimation and affine feature warping, thereby achieving explicit decoupling between global head motion and local eye motion. Based on the aligned features, inter-frame differences are used to adjust spatial and channel attention weights, enhancing motion-sensitive representations without introducing an additional temporal modeling layer. Extensive experiments on the EyeDiap and Gaze360 datasets demonstrate the effectiveness of the proposed approach. DGAGaze achieves improved gaze estimation accuracy while maintaining a lightweight architecture based on a ResNet-18 backbone, outperforming existing state-of-the-art methods. Full article

Background: In multicultural healthcare systems such as the Italian one, migrant children may experience dental care as particularly stressful because linguistic and cultural barriers can limit communication, emotional expression, and understanding of the clinical setting. Aim: Understanding the emotional experience of migrant children during dental visits is essential for improving clinical management in pediatric dentistry and orthodontics within multicultural contexts. Because linguistic barriers often limit verbal communication, this study aimed to explore children’s mental representations, emotional states, and perceptions of the dental environment through drawing and to evaluate the clinical implications for communication and therapeutic collaboration. Methods: This qualitative study was conducted in Italy between 2016 and 2025 and analyzed 50 drawings produced by 50 foreign-born migrant children aged 6–13 years, recruited through an educational cooperative in Piacenza. Most participants originated from developing countries and had limited proficiency in Italian, frequently showing a marked “experience gap” in drawing ability that interfered with normative developmental stages described by Lowenfeld. The analysis focused on spatial organization, line quality, color use, posture, interpersonal distance, and representation of the clinical environment, integrating graphic competence assessment with emotional interpretation. Results: Younger children commonly depicted rigid lines, essential settings, and oversized dental unit lamps, whereas older children increasingly represented threatening or disproportionate instruments, aggressive dentists, and omission of the patient figure. Around age 10, drawings became more detailed and colorful, although symbols of closure, such as locked doors, persisted. In adolescents, representations polarized between rich, coherent scenes and extremely essential drawings dominated by fear, rigidity, minimal environments, and symbols of constraint. The findings suggest that drawing may represent a valuable non-verbal clinical and communicative resource for exploring migrant children’s emotional experience of dental care and for identifying signs of anxiety and vulnerability that may not emerge through verbal interaction alone. Conclusions: These findings support the value of a culturally sensitive dental approach integrating drawing, visual aids, multilingual educational materials, and play-based strategies to reduce communication barriers and improve cooperation in migrant children receiving pediatric dental and orthodontic care. Full article

Historic streets, as living heritage environments, preserve everyday cultural practices while facing increasing digital mediation in tourism and daily life. This study examines how a digital sense of place is constructed online in the Ming–Qing Old Street of Songyang, China. User-generated text and image data were collected primarily from Weibo, supplemented by user reviews from major travel platforms, including Dianping, Fliggy, Mafengwo, and Ctrip, and analysed through a multimodal framework. BERTopic was applied to identify thematic narratives in textual content, and ResNet-50 was used to classify visual scene elements in shared images, enabling an integrated interpretation of textual and visual representations. The results reveal four dominant dimensions of digital place perception: local food culture, living handicrafts, historic spatial fabric, and everyday atmosphere. Textual narratives emphasise emotional attachment and experiential interpretation, while visual representations highlight photogenic, performative, and shareable street scenes. The integration of these modalities forms a layered digital sense of place grounded in cultural continuity and daily life. The study demonstrates the value of multimodal social media analysis in understanding how living heritage streets are digitally represented and perceived, offering implications for sustainable heritage conservation, community-centred revitalisation, and data-informed cultural tourism management. Full article

Industrial heritage in resource-depleted mining towns faces the dual challenge of physical decay and social severance. To achieve sustainable urban revitalization, adaptive reuse strategies must align with local collective memory and emerging experiential consumption trends. Adopting a Scene Theory perspective, this study constructs a multi-level analytical framework using Meitanba Town (Hunan, China) and its power plant as a case study. A mixed-methods approach was employed, combining semantic network analysis of 1582 online user comments with 61 offline questionnaires distributed to local residents to quantitatively diagnose current scene elements, functions, and features. The quantitative results reveal a significant imbalance: while “Functional Media” achieved the highest comprehensive score (10.0) due to strong historical recognition, “Diverse Groups” scored the lowest (3.4), indicating a lack of social inclusivity. Specifically, residents expressed the highest demand for sports facilities (31.2%) and cultural spaces (23.7%), identifying the main workshop (26.4%) and chimney as core carriers of industrial identity. Responding to these findings, the paper proposes three targeted strategies: (1) Activate: creating open-access recreation scenes to satisfy urgent sports demands; (2) Link: constructing immersive cultural scenes to narrate the “coal–electricity–life” history; and (3) Enhance: developing industry-powered commercial scenes to avoid homogenization. This study enriches the localized application of Scene Theory and provides a data-driven, context-adjustable analytical and strategic model that can inform the sustainable renewal of mining towns globally, with its specific implementation requiring adaptation to local social, economic, and cultural characteristics. Full article

Extracting road networks from Synthetic Aperture Radar (SAR) data represents a core challenge in remote sensing scene analysis, particularly for applications in traffic monitoring and emergency management. The task is complicated by several inherent limitations: speckle noise degrades image quality, geometric distortions arise from the side-looking acquisition geometry, and roads often exhibit weak radiometric separation from surrounding terrain. Traditional processing pipelines and recent single-branch deep learning frameworks have shown insufficient performance when global contextual reasoning and fine-scale spatial detail must both be addressed. This work presents DINOv3-PEFT, a parameter-efficient dual-encoder network designed specifically for SAR road segmentation. The architecture employs two complementary processing streams tailored to SAR characteristics: one stream utilizes adapter-based fine-tuning applied to pre-trained DINOv3 weights (kept frozen), which captures long-distance spatial relationships crucial for maintaining network connectivity despite speckle corruption. The second stream, based on convolutional operations, focuses on extracting localized geometric features that preserve the narrow, elongated structure and sharp boundaries typical of road infrastructure. Feature fusion occurs through the Topological-Geometric Feature Integration (TGFI) Module, which synthesizes multi-scale representations hierarchically. This mechanism proves effective at bridging fragmented road segments and recovering geometric accuracy in scenarios with heavy shadow casting or signal interference. Performance evaluation on the GF-3 satellite dataset across four spatial resolutions (1 m, 3 m, 5 m, and 10 m) demonstrates the proposed method achieves an 82.61% F1-score, a 76.51% IoU, and a 98.08% overall accuracy, all averaged across the four resolutions. When benchmarked against six state-of-the-art methods, DINOv3-PEFT demonstrates substantial improvements in road class segmentation quality and topological connectivity preservation, supporting its robustness for operational SAR road mapping tasks. Full article

The term asphyxia refers to a disruption in brain function due to rapid and persistent cerebral hypoxia or anoxia as a consequence of accidental or non-accidental injury. Considering the different mechanisms that may determine asphyxiation, such injuries can be referred to different categories: strangulation (death by hanging, ligature or manual strangulation), suffocation (smothering, choking, confined spaces and vitiated atmosphere), mechanical asphyxia (positional and traumatic asphyxia) and drowning (submersion or immersion in liquid). In both human and veterinary forensic practice, fatal asphyxia is considered among the most diagnostically challenging categories of sudden death, as it often produces only subtle and non-pathognomonic macroscopic signs, which can be easily covered by post-mortem alterations. Therefore, a wide range of information is often needed for the diagnosis of asphyxiation, including medical history, crime scene analysis, testimonies and physical evidence, along with the macroscopic and histological findings. The following review addresses the main lesions, ancillary tests and diagnostic issues associated with non-drowning asphyxia in veterinary forensic pathology. Full article

Rapid post-disaster building damage assessment requires recognizing explicit structural failures and interpreting implicit situational cues in visually complex scenes. Whereas conventional automated methods are often confined to detecting explicit damage patterns, human perception naturally integrates both types of information into a holistic risk judgment. This study presents an exploratory investigation into the neural signatures underlying this integrated judgment process using electroencephalography. A modified paradigm was employed to probe the cognitive dynamics of risk evaluation in participants with civil engineering backgrounds. Although participants were instructed only to identify damaged buildings without explicit severity grading, event-related potential analysis revealed systematic, graded neural responses that scaled with damage severity. This suggests that the brain encodes damage-related information not as a binary state but as a continuous spectrum of perceived risk, implicitly processing severity, even in the absence of explicit instructions. Furthermore, single-trial analysis demonstrated that time-domain features contain robust discriminative information, verifying the feasibility of decoding these latent judgments from brain activity. These findings provide a physiological basis for developing future cognition-informed algorithms and human-in-the-loop frameworks, bridging the semantic gap to enhance the reliability of automated disaster assessment. Full article

Infrared UAV small target detection is critical in areas such as military reconnaissance, disaster monitoring, and border patrol. However, it faces challenges due to the small size of targets, weak texture, and complex backgrounds in infrared images. Existing deep learning-based object detection models often learn spurious correlations between targets and their backgrounds. This leads to poor generalization and higher rates of false positives and missed detections in complex scenes. To overcome feature bias and improve performance, this paper proposes an enhanced detection framework based on causal reasoning. The framework builds on the advanced CSFPR-RTDETR detector. Guided by the principles of structural causal models, it explicitly separates causal and non-causal features in the feature space. Feature debiasing is achieved through a three-path approach. First, a causal data augmentation module is introduced. It applies frequency perturbations drawn from a Gaussian distribution to non-causal features. This strengthens the model’s robustness against mixed disturbances. Second, a counterfactual reasoning module is integrated into the backbone network. This module generates counterfactual samples to intervene in the feature distribution, helping the model identify and utilize causal features more effectively. Third, a causal attention mechanism module is added to the encoder. By distinguishing and weighting causal and non-causal features, it guides the model to focus on features that are essential for detecting targets. Experiments on the HIT-UAV public dataset show that the proposed framework improves mAP@50 by 5.6% and mAP@50:95 by 1.8%. Visualization analysis further confirms that the framework enhances feature discrimination and overall detection performance. Full article

Side-scan sonar (SSS) serves as the primary perceptual instrument for Autonomous Underwater Vehicles (AUVs) in large-scale marine search and rescue (SAR) operations. However, the detection of critical targets is frequently hindered by severe hydro-acoustic noise, the spatial discontinuity of wreckage, and the weak visual signatures of small targets. To surmount these challenges, this paper presents WPG-DetNet. First, we introduce a Wavelet-Embedded Residual Backbone (WERB) to reconstruct the conventional downsampling paradigm. By substituting standard pooling with the Discrete Wavelet Transform (DWT), this architecture explicitly disentangles high-frequency noise from structural information in the frequency domain, thereby achieving the adaptive preservation of edge fidelity for large human-made targets while filtering out speckle interference. Then, addressing the distinct challenge of discontinuous aircraft wreckage, the framework further incorporates a Debris Graph Reasoning Module (D-GRM). This module models scattered fragments as nodes in a topological graph to capture long-range semantic dependencies, transforming isolated instance recognition into context-aware scene understanding. Finally, to bridge the gap between AI and underwater physics, we design a Shadow-Aided Decoupling Head (SADH) equipped with a physics-informed geometric loss. By enforcing mathematical consistency between target height and acoustic shadow length, this mechanism establishes a rigorous discriminative criterion capable of distinguishing weak-echo human bodies from seabed rocks based on shadow geometry. Experiments on the SCTD dataset demonstrate that WPG-DetNet achieves a mean Average Precision ($mA{P}_{50}$) of 97.5% and a Recall of 96.9%. Quantitative analysis reveals that our framework outperforms the classic Faster R-CNN by a margin of 12.8% in $mA{P}_{50}$ and surpasses the Transformer-based RT-DETR-R18 by 5.6% in high-precision localization metrics ($mA{P}_{50:95}$). Simultaneously, WPG-DetNet maintains superior efficiency with an inference speed of 62.5 FPS and a lightweight parameter count of 16.8 M, striking an optimal balance between robust perception and the real-time constraints of AUV operations. Full article

Gaze analysis often relies on hypothesised, subjectively defined regions of interest (ROIs) or heatmaps: ROIs enable condition comparisons but reduce objectivity and exploration; while heatmaps avoid this, they require many pixel-wise comparisons, making differences hard to detect. Here, we propose an advanced data-driven approach for analysing gaze behaviour. We use DNNs (adapted versions of AlexNet) to classify conditions from gaze patterns, paired with reverse correlation to show where and how gaze differs between conditions. We test our approach on data from an experiment investigating the effects of object-specific sounds (e.g., church bell ringing) on gaze allocation. ROI-based analysis shows a significant difference between conditions (congruent sound, no sound, phase-scrambled sound and pink noise), with more gaze allocation on sound-associated objects in the congruent sound condition. However, as expected, significance depends on the definition of the ROIs. Heatmaps show some unclear qualitative differences, but none are significant after correcting for pixelwise comparisons. We showed that, for some scenes, the DNNs could classify the task based on individual fixations with accuracy significantly higher than chance. Our approach shows that sound can alter gaze allocation, revealing task-specific, non-trivial strategies: fixations are not always drawn to the sound source but shift away from salient features, sometimes falling between salient features and the sound source. Crucially, such fixation strategies could not be revealed using a traditional hypothesis-driven approach. Overall, the method is objective, data-driven, and enables clear comparisons of conditions. Full article

Accurate detection of small objects in video analytics is limited by low pixel resolution and insufficient visual cues. While software-based enhancements often fail to recover missing details, Pan–Tilt–Zoom (PTZ) cameras can physically increase spatial resolution through optical zoom. However, mechanical latency and configuration complexity hinder their real-time applicability. We propose ZoomPatch, a real-time video analytics framework tailored for small object detection. ZoomPatch actively schedules PTZ adjustments to capture optically enhanced subframes of regions of interest (ROIs) and fuses inference results back to the global reference frame. Specifically, it introduces a dynamic Cycle Length Proposer to adapt analysis cycles based on scene motion, and a Mixed Integer Linear Programming (MILP)-based Configuration Decider to determine the optimal sequence of pan, tilt, and zoom adjustments under time budget constraints. Simulation-based experimental evaluations across diverse workloads demonstrate that ZoomPatch significantly outperforms fixed-perspective, super-resolution (SR), and greedy baselines. Notably, in the detection task using YOLOv10, ZoomPatch improves the F1-score from 0.33 to 0.47 (a 42% increase) compared to the fixed-perspective baseline. Furthermore, ZoomPatch yields performance gains of 30% and 7% over the SR baseline (0.36) and the greedy baseline (0.44). Full article

Terrestrial laser scanning (TLS) provides dense point clouds for landslide monitoring, yet occlusion, heterogeneous point density, and seasonal vegetation introduce noise and unstable deformation boundaries in multi-temporal change detection. To overcome the limitations of the multiscale model-to-model cloud comparison (M3C2) method under dominant downslope tangential motion and vegetation disturbance, we propose a block-wise ICP method to retrieve 3D displacement vectors. The scene is partitioned into local sub-blocks; rigid registration is performed within each sub-block, and the estimated translation is assigned to the sub-block center. A two-stage matching and quality control procedure removes under-constrained sub-blocks, enabling the direct retrieval of 3D displacement vectors and interpretable boundaries. Applied to the Longxigou landslide in Wenchuan using RIEGL VZ-2000i surveys on 1 November 2023 and 23 May 2024, the proposed method produces a more continuous displacement field and clearer boundaries than M3C2. For a tower target, manual measurements indicate a displacement of 0.41–0.63 m; our estimates are within 0.33–0.40 m, whereas M3C2 mostly falls between −0.25 and 0.25 m. In a seasonal vegetation change scene, we detect a canopy envelope expansion of approximately 0.20–0.40 m, while M3C2 shows scattered canopy responses that hinder boundary interpretation. A sensitivity analysis indicates a block-scale trade-off between boundary stability and peak preservation, motivating adaptive multi-scale blocking and uncertainty quantification. Full article

The large-scale integration of renewable energy sources and power-electronic-interfaced devices has significantly weakened transient support capability and disturbance tolerance, posing new challenges to the secure and stable operation of modern power systems. Conventional stability analysis methods suffer from high computational burden, long execution time, and limited adaptability to diverse operating scenarios. The rapid development of artificial intelligence (AI) provides effective technical support for fast and accurate assessment of power-system security and stability. This paper presents a comprehensive review of AI-based methods and the interpretability for transient stability assessment (TSA) in modern power systems. First, an intelligent TSA framework is introduced, consisting of three key stages: sample construction and enhancement, intelligent algorithms and learning mechanisms, and model training and interpretability. Subsequently, existing methods for data augmentation, intelligent algorithms, learning mechanisms, and interpretability analysis are systematically reviewed, and the corresponding application scene, technical superiority and limitations are discussed. Finally, from a knowledge–data fusion perspective, four representative integration paradigms combining mechanism-based models and data-driven approaches are summarized, and the application prospects in power-system stability analysis are discussed. Full article

Generative AI enables rapid visualization of sustainable urban design scenarios, yet the question of whether these outputs encode sustainability as operable spatial logic, rather than merely depicting it as a visual impression, remains underexplored. This study proposes a two-level assessment framework that scores the same sustainability dimensions at both the visual-representation level and the spatial-logic level, treating the systematic decoupling between the two as a form of visual greenwashing: system-induced representational distortion rather than deliberate misrepresentation. Using AI-workflow reports from two site-based urban design studios (47 students, 12 teams, 36 coded scenes), the framework integrates rubric-based scoring with qualitative process tracing of breakdown–repair logs. Results show that image-level scores consistently outperform logic-level scores across all five dimensions, with the gap most severe in mobility hierarchy and walkability and smallest in green/blue infrastructure. Case analysis reveals that breakdowns arise from failures in program encoding, urban-scale coherence, functional-boundary demarcation, and relational-condition matching, and that students deploy multi-stage repair pipelines, including prompt restructuring, tool switching, reference injection, and external-source compositing, to re-inject collapsed spatial logic. These findings reframe AI-assisted urban design as repair-centered workmanship rather than automated production. The study proposes three guardrails to prevent visual sustainability from substituting for spatial-logic sustainability: image–logic paired submission, design audit trail formalization, and gap-based red-flag review. Full article

Traditional concert-hall evaluations primarily rely on ISO 3382-1 scalar parameters (e.g., C₅₀ and C₈₀), which summarize temporal energy behavior but provide limited insight into the directional composition of early reflections, particularly in geometrically shadowed seating zones. This paper presents a first-order Ambisonics (FOA)-based 3D acoustic sensing framework to diagnose under-balcony directional imbalance, with emphasis on early vertical-reflection deficiency. Scene-based FOA impulse responses (WXYZ) were measured at 11 audience positions (P1–P11) in the National Concert Hall (Taipei) and analyzed using intensity-based direction-of-arrival (DoA) proxies, axis-resolved directional energy build-up, and a distributional descriptor based on directional spatial entropy. Results are presented at three representative frequencies (125 Hz, 1 kHz, and 4 kHz) and analyzed within full (0–200 ms), early (0–80 ms), and late (80–200 ms) windows. While the magnitude proxy $\mid p^{meas}\left(f\right)\mid$ exhibits strong seat-to-seat variability and does not support a uniform attenuation assumption under the balcony, direction-resolved metrics reveal a consistent under-balcony signature. Specifically, the early–late vertical energy discrepancy $\Delta R_z={\displaystyle R_z^{early}}-{\displaystyle R_z^{late}}$ is persistently negative at under-balcony positions (P7–P11) across all three frequencies, indicating a selective reduction in early vertical contribution relative to the late field. Directional entropy analysis further shows predominantly negative $\Delta H_n={\displaystyle H_n^{early}}-{\displaystyle H_n^{late}}$, with more negative values in the under-balcony group, consistent with stronger early directional constraint in shadowed seats. Spatial trend maps are provided via Gaussian RBF interpolation within the audience domain for visualization only. The proposed FOA-based diagnostic framework provides a practical and physically interpretable approach to identify direction-specific early-reflection deficits that remain masked in conventional scalar evaluations, supporting mechanism-oriented assessment and targeted intervention in geometrically constrained listening areas. Full article