Search Results (7,396)

Deep learning approaches for egocentric video understanding often lack a principled theoretical treatment of stability, particularly when dealing with the sparse, noisy, and temporally ambiguous observations characteristic of first-person imaging. In this work, we frame egocentric video question answering not merely as a classification task, but as an ill-posed inverse problem aimed at reconstructing latent semantic intent from stochastically perturbed visual signals. To address the instability inherent in standard dual-encoder architectures, we present a framework with a mathematical interpretation that incorporates gated cross-modal interaction within the transformer backbone. Formally, the video-side update analyzed in this work is defined as a learnable convex combination of unimodal feature representations and cross-modal attention residuals; the full implementation applies analogous gated cross-modal updates bidirectionally. From a regularization perspective, the gating mechanism can be interpreted as an adaptive parameter that balances data fidelity against language-conditioned structural constraints during feature reconstruction. We provide the Bounded Update Property (Lemma 1) and an analytical layer-wise sensitivity bound and empirically demonstrate that the proposed framework achieves measurable improvements in both accuracy and stability on the EgoTaskQA and MSR-VTT benchmarks. On EgoTaskQA, our model improves accuracy from 27.0% to 31.7% (+4.7 pp) and reduces the accuracy drop under 50% frame drop from 3.93 pp to 0.94 pp. On MSR-VTT, our model improves accuracy by 13.0 pp over the dual-encoder baseline. Under severe perturbation (50% frame drop) on MSR-VTT, our model retains 97.7% of its clean performance, whereas the baseline exhibits near-zero drop accompanied by majority-class behavior. These results provide empirical evidence that the proposed interaction induces stable behavior under perturbations in an ill-posed multimodal inference setting, mitigating sensitivity to sampling variability while preserving query-relevant temporal structure. Furthermore, an entropy-based analysis indicates that the gating mechanism prevents excessive diffusion of attention, promoting coherent temporal reasoning. Overall, this work offers a mathematically informed perspective on designing interaction mechanisms for stable multimodal systems, with a focus on robust reasoning under temporal ambiguity. Full article

The resistance of pathogenic bacteria to antimicrobial agents is currently one of the most significant health challenges. Polymers and nano-polymer composites with antimicrobial properties are widely used, particularly in hospitals, biocompatible implants, and the medical device industry. Syzygium aromaticum (clove) contains several bioactive compounds, including potent antioxidants and antimicrobials, which confer antioxidant, antibacterial, and antiseptic properties. For this purpose, polyvinyl alcohol (PVA) films were produced at three different concentrations using a direct integration method and doped with clove essential oil. The spectral, structural, and thermal properties of the produced films were analyzed, and their antibacterial activity against Klebsiella pneumoniae was tested. Fourier Transform Infrared Spectroscopy (FTIR) results confirm that the structural integrity of the PVA matrix is preserved and that the essential oil is physically trapped within the polymer network. Overall, the Differential Scanning Calorimetry (DSC) results confirm that Syzygium aromaticum essential oil (SAEO) acts as an effective plasticizer in PVA films, significantly modifying the glass transition behavior and enhancing polymer chain mobility in a concentration-dependent manner. The Dynamic Mechanical Analysis (DMA) results, supported by DSC analysis, clearly demonstrate that SAEO acts as an effective plasticizing agent in PVA films by increasing molecular mobility, lowering the glass transition temperature (Tg), and promoting thermally induced deformation. The concentration-dependent increase in the diameter of the inhibition zone of essential-oil-added films showed that their antibacterial efficacy increased as the S. aromaticum essential oil content increased (0.5%, 0.75%, and 1.0%). Additionally, molecular docking was performed to examine interactions between selected virulence proteins of K. pneumoniae and the main components of clove essential oil. As a result, S. aromaticum essential oil conferred antibacterial properties to the polyvinyl alcohol films without significantly altering their transparency and thermal properties. Full article

This paper presents a new continuous data model, the Unit Arcsine–Exponential distribution (UASED), a flexible data model on the unit interval. It is built up by an exponential-based arcsine-type transformation to allow it to represent a very wide range of shapes that can be used to model proportions and rates. A number of basic properties are obtained, such as closed-form formulas of the quantile function, moments, and entropy measures. Maximum likelihood and maximum product of spacings methods are developed to estimate parameters, and their performance is determined by Monte Carlo simulation, which shows that these methods can reasonably estimate the parameters and be stable over a variety of different parameter settings. To demonstrate that a model is practically useful, an application to real-world data on the reliability of devices in terms of failure time is discussed. The findings indicate that the UASED is a good fit to the data, in the sense that it is effective in terms of skewness and tail behavior and compares well or competes favorably with current unit distributions. All in all, the suggested model is a sparse alternative to model bounded data with sound inferential characteristics and high practical utility. Full article

Foaming technology can effectively reduce the viscosity of polymer-modified asphalt and significantly decrease energy consumption during pavement construction, making it an effective approach for achieving low-carbon pavement construction and maintenance. However, mechanically foamed asphalt relies on specialized equipment and requires strict parameter control. Although water-based foaming methods using zeolites or ethanol can alleviate these issues to some extent, they still present disadvantages such as significant variability in foaming performance and potential risks during transportation and construction. Therefore, this study investigates the feasibility of using crystalline hydrates with high water of crystallization for micro-foamed asphalt. Three types of micro-foamed SBS-modified asphalt (MFPA) were prepared using hydrates with different contents of water of crystallization. Physical property tests, foaming characteristic parameters, viscosity–temperature analysis, Fourier transform infrared spectroscopy (FTIR), adhesion tensile tests, scanning electron microscopy (SEM), and fluorescence microscopy were conducted to evaluate their effects on the physical and chemical properties, viscosity reduction performance, adhesion, and compatibility of SBS-modified asphalt. Furthermore, dynamic shear rheometer (DSR) tests, bending beam rheometer (BBR) tests, fatigue life modeling, and morphological analysis were employed to investigate the rheological properties, fatigue life, and bubble evolution behavior of the MFPA system. The results indicate that utilizing the thermal decomposition characteristics of crystalline hydrates with high water of crystallization (Na₂SO₄·10H₂O, Na₂HPO₄·12H₂O, and Na₂CO₃·10H₂O) to release H₂O and CO₂ in SBS-modified asphalt for micro-foaming is a short-term reversible physical viscosity reduction process. The maximum expansion ratio (ERmax) of MFPA reaches 8–10, the half-life (HL) remains stable at approximately 180 s, and the foaming index (FI) peak is about 1160. The construction temperature can be reduced by 10–15%, and the viscosity reduction effect remains stable within 60 min. Compared with unfoamed SBS-modified asphalt, the compatibility, rutting resistance, and fatigue life of MFPA increase by approximately 65%, 32%, and 30%, respectively, while the low-temperature performance decreases by 18%. Under the same short-term and long-term aging conditions, MFPA exhibits better aging resistance. Specifically, its rutting resistance increases by 37%, and fatigue resistance improves by 30% compared with aged SBS-modified asphalt, while the low-temperature performance remains essentially unchanged. Full article

The rapid growth of electric vehicles (EVs) is reshaping transport systems and accelerating the sustainable digital transformation of smart mobility. EV battery-swapping, delivered through platform-based, data-driven service networks, offers a low-carbon alternative to conventional refueling and plug-in charging by shortening replenishment time and enabling centralized battery management. However, the behavioral mechanisms driving user adoption of this digitally enabled infrastructure remain insufficiently understood. This study develops a socio-technical system (STS) model in which social and technical drivers influence users’ intention to adopt EV battery-swapping services via the dual mediation of perceived trust and perceived risk. Using a three-stage mixed-methods design that combines a PRISMA-based literature review, expert interviews with user-journey mapping, and a large-scale user survey, the study identifies six social and technical antecedents of EV battery-swapping adoption. Based on 565 valid responses from EV users in the Beijing–Tianjin–Hebei region, partial least squares structural equation modeling and multi-group analysis are employed to test the proposed framework. The results show that all six antecedents significantly affect perceived trust and perceived risk, which in turn mediate their impacts on adoption intention, with notable heterogeneity across income and usage-frequency groups. The findings provide a mechanism-based extension of STS theory for digitally mediated battery-swapping infrastructure by showing how socio-technical conditions shape adoption via trust and risk, and they offer actionable implications for operators and policymakers to build secure, user-centered swapping services within intelligent transport systems. Full article

Hyperspectral unmixing (HU) is fundamental for conducting quantitative analyses in remote sensing, yet existing methods face a persistent tradeoff between model performance and physical interpretability. Although deep learning models achieve superior performance, even “gray-box” models that incorporate physical constraints still suffer from an intrinsically opaque decision-making process, which hinders their trustworthiness in critical applications. To address this challenge, this paper introduces a theory-guided unmixing framework aimed at enhancing mechanistic interpretability called the sparse and subspace-attentive transformer unmixing network (SSTU-Net). Unlike heuristic architectures, SSTU-Net is rigorously derived from the first principles of sparse rate reduction (SRR) theory. Its core modules—the multi-head subspace self-attention (MSSA) and the iterative shrinkage-thresholding algorithm (ISTA)—directly implement the essential mathematical steps of information compression and sparsification within the SRR theory, respectively. Extensive experiments on both synthetic and real hyperspectral datasets demonstrate that SSTU-Net achieves competitive performance compared to representative state-of-the-art methods—including advanced autoencoder-based networks (e.g., CyCU-Net and DAAN) and recent transformer-based unmixing architectures (e.g., DeepTrans and MAT-Net)—while strictly adhering to theoretically predicted evolutionary trajectories. More importantly, a series of specifically designed structural interpretability validation experiments mechanistically confirm the theoretically predicted behaviors, such as layer-wise information compression, feature sparsification, and subspace orthogonalization. These results reveal the internal working mechanisms of SSTU-Net, validating the feasibility and significant potential of our principled theory-guided framework for developing high-performance and trustworthy intelligent models in remote sensing. Full article

Neuroinflammation is a central hallmark of numerous neurological disorders, including Alzheimer’s disease, Parkinson’s disease, traumatic brain injury, and spinal cord damage. Its persistent and dysregulated nature not only accelerates neuronal loss but also impedes endogenous repair, posing a major challenge for effective therapeutic intervention. Recent advances in nanobiotechnology have opened transformative opportunities to modulate neuroinflammation with unprecedented precision while simultaneously supporting neural regeneration. This review highlights emerging nanomaterial-based strategies including lipid-based, polymeric, inorganic nanoparticles designed to traverse the blood–brain barrier (BBB), deliver anti-inflammatory agents, modulate immune cell behavior, and attenuate glial activation. Extending beyond nanoparticle-based delivery systems, recent advances also emphasize the integration of nanomaterials into biomimetic architectures to provide structural and functional cues for neural repair. We further summarize how these functional nanostructured scaffolds, such as extracellular matrix (ECM) mimetic, nanofibrous and conductive hydrogels, are being leveraged in neural tissue engineering to direct stem cell fate, promote axonal outgrowth, and rebuild damaged neuroarchitectures. Moreover, pharmacokinetics, biodistribution, safety, clinical trials, regulatory considerations and limitations of nanotherapeutics in neurodegenerative diseases are discussed. By outlining the current progress, mechanistic insights, and translational challenges, this review underscores the potential of nanobiotechnology-enabled therapeutics to revolutionize the treatment of neuroinflammatory conditions and advance next-generation neural repair technologies. Full article

In this paper, we propose an analytic deformation acting on probability measures, designed to encompass and extend two fundamental operators in free probability: the $(a,b)$- and the $T_c$-deformations. This unified operator, indicated by ${\mathcal{X}}_{(a,b,c)}$, is introduced through a functional relation for the Cauchy–Stieltjes transform. We have ${\mathcal{X}}_{(a,b,0)}={\tilde{U}}^{(a,b)}$ and ${\mathcal{X}}_{(1,1,c)}=T_c$. We examine the structural properties of this transformation within the setting of Cauchy–Stieltjes kernel (CSK) families, with special emphasis on the behavior of the associated variance functions (VFs). An explicit formula for the VF corresponding to measure deformed by ${\mathcal{X}}_{(a,b,c)}$ is established. This result allows us to demonstrate a key invariance property: the free Meixner class of probability measures remains stable under the ${\mathcal{X}}_{(a,b,c)}$-transformation. Furthermore, a novel characterization of the semicircle law is obtained through the action of ${\mathcal{X}}_{(a,1,c)}$, highlighting the role of symmetry in the deformation and preservation of free-probabilistic distributions. Full article

With the continuous elevation of aviation safety standards, accurate monitoring of ground support vehicles in aircraft stand areas has become a critical task for enhancing overall aircraft stand operational safety. Given the limitations of existing surface movement radar and multi-camera surveillance systems in terms of cost, deployment complexity, and coverage, this paper proposes a lightweight vision-based framework for vehicle state perception and spatiotemporal behavior analysis oriented toward aircraft stand safety. Leveraging existing fixed monocular monitoring resources in the stand area, the framework first establishes a precise mapping from image pixel coordinates to the physical plane through self-calibration and homography transformation utilizing scene line features, thereby achieving unified spatial measurement of vehicle targets. Subsequently, it integrates an improved lightweight YOLO detector (incorporating Ghost modules and CBAM for noise suppression) with the ByteTrack tracking algorithm to enable stable extraction of vehicle trajectories under complex occlusion conditions. Finally, by combining functional zone division within the stand, a semantic map is constructed, and a behavior analysis method based on a spatiotemporal finite state machine is proposed. This method performs joint reasoning by fusing multi-dimensional constraints including position, zone, and time, enabling automatic detection of abnormal behaviors such as “intrusion into restricted areas” and “abnormal stop.” Quantitative evaluations demonstrate the framework’s efficacy: it achieves an average physical localization error (RMSE) of 0.32 m, and the improved detection model reaches an accuracy (mAP@50) of 90.4% for ground support vehicles. In tests simulating typical violation scenarios, the system achieved high recall (96.0%) and precision (95.8%) rates in detecting ‘area intrusion’ and ‘abnormal stop’ violations, respectively. These results, achieved using only existing surveillance cameras, validate its potential as a cost-effective and easily deployable tool to augment existing safety monitoring systems for airport ground operations. Full article

The valorization of agro-industrial byproducts as sources of functional polysaccharides is a promising strategy for developing sustainable materials. In this study, cellulose was extracted and purified from rice husk and apple pomace through sequential alkaline and bleaching treatments. Then it was chemically modified via TEMPO-mediated oxidation to obtain cellulose nanofibers (TOCNFs) with cellulose yields ranging from 23.8 to 32.4% for rice husk and 9.3–13.8% for apple pomace. Owing to its higher recovery and structural regularity, rice husk was selected for surface modification with 3-aminopropyltriethoxysilane (APTES). The resulting TOCNFs exhibited an average width of 8 nm and a carboxyl content of 0.48 mmol g⁻¹. Fourier-transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), and nitrogen determination (1.72 mg g⁻¹) confirmed the presence of aminosilane functionalities. APTES-modified TOCNFs were incorporated as active components to develop hybrid poly(vinyl acetate) (PVA) adhesives synthesized via in situ heterogeneous water-based polymerization. The influence of TOCNF surface chemistry and sodium dodecyl sulfate (SDS) on latex particle size, rheological behavior, and adhesive performance was systematically investigated. Latex particle size increased from 193 nm (PVA-SDS) to 625 nm with TOCNF-APTES and decreased to 247 nm upon SDS addition. Rheological analysis revealed pronounced shear-thinning behavior associated with the formation of percolated nanofibrillar networks, with low-shear viscosity increasing up to 477 Pa·s for TOCNF–APTES and decreasing to 370 Pa·s with SDS. Lap-shear testing (ASTM D905) showed substantial improvements in adhesive strength, reaching up to 250 kPa compared to PVA-SDS. These results demonstrate that surface-modified CNFs act not only as mechanical reinforcements but also as interfacially active components governing polymerization behavior, rheology, and adhesive performance. This exploratory study provides a proof-of-concept for the development of sustainable wood adhesives from agro-industrial byproducts. Full article

Conventional approaches to Intangible Cultural Heritage (ICH) preservation, such as static documentation and superficial commercialization, frequently undermine its vitality by reifying it as a fixed artifact detached from its evolving socio-cultural context. This study challenges this object-centric paradigm by proposing an ecosystem-centric framework that reconceptualizes ICH as a dynamic, self-organizing cultural ecosystem. Our framework integrates Complex Adaptive Systems (CAS) theory to provide a macro-level ecological perspective, with Emotional Design theory offering a micro-level mechanism for fostering public engagement. We theoretically instantiate this framework through the Yangliuqing Narrative Ecosystem, a design case applied to Tianjin Yangliuqing New Year Paintings. This system combines tangible, modular cultural gene carriers with a digital co-creation platform that guides users through visceral, behavioral, and reflective levels of engagement, aiming to transform them from passive consumers into active co-creators. This process is designed to cultivate a community of practice that drives the heritage’s adaptive evolution. The study contributes a novel theoretical framework and a transferable design methodology, presenting a robust model for reactivating the intrinsic vitality of cultural traditions in the digital age. Full article

Next Point-of-Interest (POI) recommendations are pivotal for enhancing location-based services; however, accurate prediction remains challenging due to the complex interplay between dynamic user preferences and spatiotemporal constraints. Existing graph-sequence hybrids often fail to unify these dimensions, typically treating temporal contexts as disjoint features or neglecting implicit collaborative signals within sparse user trajectories. This fragmentation limits the ability to capture high-order dependencies in user mobility. To address these challenges, we propose UPTRec, a unified framework that synergizes social, spatial, and temporal reasoning. UPTRec constructs a TF-IDF-weighted user similarity graph to recover latent social connections and a flow-based POI-transition graph to encode sequential mobility patterns. These structural priors are fused with fine-grained temporal-category embeddings (utilizing Time2Vec and periodic encoding) via a multi-layer Transformer encoder to comprehensively capture user behavior. Extensive experiments on three real-world datasets (NYC, TKY, and CA) demonstrate that UPTRec achieves state-of-the-art performance among the compared baselines under the same experimental settings. On the NYC dataset, UPTRec yields a Top-1 Accuracy of 25.76% and a Mean Reciprocal Rank (MRR) of 0.3879, representing a relative improvement of 5.8% and 7.1% over the strongest baseline (GETNext). These results validate the efficacy of jointly modeling collaborative and spatiotemporal dependencies. Full article

The primary objective of this study is to investigate the effect and mechanism of replacing limestone powder with red mud as a filler on asphalt aging resistance. The microstructure and porosity characteristics of limestone powder, Bayer process red mud, and sintered red mud were analyzed. Asphalt mastics were then prepared using these fillers. The effect of red mud on the aging resistance of asphalt was evaluated by comparing the conventional physical properties, rheological behavior, and functional groups of the asphalt mastics before and after aging. Fourier transform infrared spectroscopy (FTIR), gel permeation chromatography (GPC), and ultraviolet-visible spectroscopy (UV-Vis) were further employed to elucidate the underlying anti-aging mechanisms. The results indicate that the asphalt mastic containing 4% sintered red mud exhibits the strongest resistance to both thermo-oxidative and UV aging. It shows the lowest increments in softening point, viscosity aging index, and complex modulus aging index, with performance comparable to a commercial anti-aging agent. FTIR and GPC analyses reveal that sintered red mud selectively adsorbs light asphalt components, thereby inhibiting their conversion into heavier fractions during thermo-oxidative aging. UV-vis analysis demonstrates that sintered red mud provides effective UV shielding within the asphalt mastic, substantially mitigating UV-induced damage. In summary, the incorporation of 4% sintered red mud can significantly delay both thermo-oxidative and UV aging processes in asphalt mastics, thereby effectively enhancing the aging resistance of asphalt pavement. Full article

Titanium dioxide nanoparticles (TiO₂ NPs) are incorporated into automotive coatings to enhance durability, corrosion, UV resistance, and, in some formulations, photocatalytic self-cleaning. While the toxicology of pristine TiO₂ is well studied, the behavior of TiO₂ NPs embedded in polymer matrices and subjected to real-world aging, maintenance, and removal remains poorly characterized. This narrative review synthesizes 24 publications spanning the lifecycle of TiO₂ nano-enabled automotive coatings, from synthesis and formulation through application, in-service weathering, repair, refinishing, and end-of-life environmental fate. Upstream properties, such as coating functionality and performance, have been examined as determinants of later-life release, exposure, and fate. Across studies, dispersion state, interfacial compatibility, and surface modification—together with transformations such as agglomeration, photocatalysis, weathering, and eco-corona formation—shape particle stability, release, exposure relevance, and toxicological risk. Evidence indicates that sanding and accelerated weathering predominantly generate matrix-associated, polymer-fragment-dominated aerosols rather than pristine TiO₂ NPs, while NP-specific exposure measurements during spray application remain limited. Hazard data suggest matrix embedding may attenuate, but does not eliminate, biological responses relative to pure particles. Wastewater treatment plants and biosolids have been shown to act as sinks with potential for soil accumulation following sludge application. Regulatory frameworks rarely account for aging, transformation, and release, stressing the need for synchronized testing of aged materials and nano-specific exposure metrics to support safer-by-design coatings and risk governance. Full article

Green innovation is a crucial aspect of an enterprise’s core competitiveness and long-term sustainable development, garnering significant attention from both academic scholars and industry practitioners. However, while existing research has primarily focused on green innovation at the organizational level, the mechanisms driving green innovation behaviors at the individual level have not been thoroughly explored in the literature. This study is grounded in the classic Job Demands–Resources (JD-R) theoretical framework and highlights the interplay between job demands (such as environmental ethics and corporate environmental strategies) and job resources (such as green human resource management practices and green transformational leadership). It also integrates individual-level characteristics, specifically green mindfulness and connectedness to nature, to construct a multidimensional interactive model aimed at uncovering the complex mechanisms driving employees’ green innovation. To achieve this, the study employs fuzzy-set qualitative comparative analysis (fsQCA). The findings suggest that no single condition is necessary for employee green innovation. However, connectedness to nature consistently appears across all core configurations, indicating a prominent “enabling” effect. This suggests that employee green innovation is an active and proactive form of environmentally responsible behavior, largely driven by individuals’ emotional affinity with nature. Additionally, connectedness to nature serves as a foundational source of intrinsic motivation for environmental awareness and acts as a catalyst across multiple pathways. Configurational analysis reveals an equifinal pattern, identifying three distinct motivational pathways: (1) Self-motivation Combined with Resource Support; (2) Self-motivation Combined with Job Demands; and (3) Triple Interaction of Demand, Resources, and Individuals. This study possesses both theoretical and practical significance in systematically examining green innovation behaviors at the individual level. Full article