Search Results (3,907)

The positional arrangement of BN and CC units in B₆C₆N₆ cyclic nanorings profoundly influences their stability, electronic structure, optical response, and reactivity. Here, we comparatively investigate eight positional isomers (C1–C8) using DFT and TD-DFT calculations. Among C1–C8, C1 is the most stable, and C8 is the most unstable in the range of 200–1000 K. Their relative stability is governed by B-N charge separation, homonuclear B-B and N-N defects (charge repulsion), and bond-angle distortion (ring tension). The HOMO–LUMO gaps range from 4.40 eV (C3) to 8.45 eV (C2), indicating distinct kinetic stability. Aromaticity analysis reveals that all isomers are nonaromatic. In the gas phase, the lowest-energy absorption bands of C1 and C3 are located at about 429 nm and 606 nm, respectively. Due to different transition mechanisms, namely locally excited (LE) for the former and charge-transfer (CT) for the latter, solvent polarity has dramatically different influence on these two absorption bands. Compared to their positions in the gas phase, these absorption bands are blue-shifted about 20 nm and 220 nm in water, respectively. Reactivity analysis identifies the B-B bond in C7 as the strongest electrophilic site (LEAE = −2.93 eV), with the surrounding framework serving as nucleophilic domains, endowing C7 with the strongest bifunctional reactivity. This work establishes a comprehensive structure–property map for B₆C₆N₆ isomers, providing guidance for designing BCN-based nanorings for catalysis, molecular recognition, and optoelectronics. Full article

Camouflaged object detection (COD) remains a challenging task because objects blend into the background, exhibiting low contrast, incomplete edges, and highly similar appearances. Recent deep learning methods have improved detection performance, but most rely solely on visual features and lack semantic-level reasoning to distinguish concealed objects. To address both performance and deployment scalability, we use an edge–cloud network paradigm in which lightweight visual processing operates on an internet device while semantic reasoning is handled remotely, enabling real-time COD in internet-scale applications such as wildlife monitoring, perimeter surveillance, and UAV-based sensing. We implement this in SPINet, a vision–language-driven hybrid COD framework that integrates a multi-stage BiRefNet-Large decoder with BLIP-Large semantic comprehension. Our Multi-Stage Semantic Prior Injection (MS-SPI) module extracts and injects three complementary semantic representations for global context, region-level features, and spatial attention maps into three decoder stages (Stages 3, 4, and 5) of BiRefNet, enabling hierarchical semantic guidance at multiple scales. Experiments on three benchmark datasets (COD10K, CAMO, and NC4K) demonstrate that SPINet achieves consistent improvements over the BiRefNet-Large visual-only baseline across all benchmarks. SPINet attains $S_{\alpha }=0.921$ on COD10K, $0.859$ on CAMO, and $0.893$ on NC4K, outperforming BiRefNet-Large by $+0.9\%$, $+1.7\%$, and $+3.5\%$ in structure measure, respectively, with MAE reductions of $7\%$, $29\%$, and $22\%$. These results show that frozen semantic priors provide robust and transferable guidance for COD with negligible additional parameters. Full article

The Collatz map is a simple deterministic transformation whose orbit structure remains highly nontrivial. A recent direction-phase decomposition partitions each orbit into upward and downward steps, and numerical observations indicate that the number of upward phases, $N_{\uparrow }$, follows an approximate Gamma distribution. In this work, we provide a mechanistic explanation for this statistical regularity by modeling the occurrence of upward phases in the odd-compressed, or Syracuse, version of the Collatz map as a homogeneous Poisson process. From the mean-field logarithmic balance and the geometric distribution of 2-adic valuations, we derive closed-form expressions for the Gamma parameters: the scale parameter $\theta =2/{(2-{\log }_{2}3)}^2\approx 11.61$ is constant, whereas the shape parameter K grows logarithmically with the maximal initial value $X_0=2L+1$. We also analyze the closure conditions for periodic orbits, showing that nontrivial cycles are severely constrained, which supports the plausibility of the statistical framework. Numerical validation for L ranging from ${10}^5$ to ${10}^{15}$ confirms the theory with relative errors below $3\%$, and a bias-corrected mean estimate reduces the error to ${10}^{-3}$–${10}^{-2}\%$. These results establish a quantitative link between the arithmetic properties of the Collatz map and Gamma-type statistics, and suggest possible extensions to generalized Collatz-type problems. Full article

Images may evoke different emotional responses depending on their content, style, and viewer interpretation. This is particularly important when evaluating works of art, architectural designs, or interior design choices. Visual emotion recognition (VER) models are commonly evaluated using supervised classification metrics such as accuracy, precision, recall, and macro-F1, together with general uncertainty indicators such as entropy, maximum softmax probability (MSP), and top-1/top-2 margin. However, these measures do not indicate whether the model’s strongest competing-emotion predictions remain within the same sentiment group or cross the positive–negative sentiment boundary. This paper proposes the top-2 cross-sentiment rate (CrossRate), a label-free measure for analyzing sentiment-level ambiguity in VER models. CrossRate measures the proportion of samples for which the top-1 and top-2 predicted emotion classes belong to opposite sentiment groups. The measure is evaluated on VER datasets using both standard classification metrics and uncertainty indicators. Experiments on EmoSet-118K show that varying the model’s parameters reduces CrossRate from ($22.15\pm 0.45)\%$ to ($7.81\pm 0.62)\%$ and increases accuracy from ($79.13\pm 0.16)\%$ to ($80.10\pm 0.15)\%$. These changes are not fully reflected by entropy, MSP, or margin, indicating that CrossRate captures a complementary aspect of sentiment-level prediction behavior. The WikiArt case study further demonstrates that CrossRate can be applied when ground-truth emotion labels are unavailable. The proposed measure is applicable to any VER model whose predicted emotion classes can be mapped into positive and negative sentiment groups. The application of CrossRate is illustrated by its use in estimating the emotions of artworks. It offers even non-art experts the opportunity to form an opinion. Full article

Hydraulic conductivity (K) is essential for understanding groundwater movement in shallow alluvial aquifers. Upstream of the Siret–Moldova confluence, Romania, this parameter is poorly constrained, because available K values come from only 34 wells reported in existing hydrogeological documentation. This study provides the first directional geostatistical characterization of K in this shallow alluvial aquifer and evaluates the spatial reliability of the mapped zones. K values ranged from 9.96 to 171.73 m/day. Because the raw data were positively skewed, the values were log-transformed before geostatistical modelling. Spatial continuity and anisotropy were examined using omnidirectional and directional semivariograms. Three theoretical models were tested—Spherical, Gaussian and Exponential—and their performance was compared using cross-validation statistics and prediction standard-error maps. The Gaussian and Spherical models showed comparable performance; however, the Gaussian model was retained because it produced marginally lower errors and a smoother spatial pattern (ME = 0.0055, RMSE = 0.4870, RMSSE = 0.8697, R² = 0.6712). The back-transformed K map shows three main zones: lower values in the northern and eastern sectors, intermediate values in the central sector, and the highest values toward the west. This pattern is consistent with coarser Moldova River alluvial deposits in the west and more heterogeneous lithological conditions in the central area. The prediction standard-error maps indicate the areas where the interpolated hydraulic conductivity (K) is more reliably constrained. The resulting map provides a preliminary spatial representation of the main hydraulic-conductivity zones and indicates where additional K measurements are needed before detailed groundwater-flow modelling or local groundwater-resource assessment. Full article

Infrared multi-class small-target detection is challenging because targets occupy few pixels, exhibit weak texture, and are easily confused with background clutter. We present RADFRNet, a YOLOv11-n-based detector designed to address two forms of information degradation: detail loss in the backbone and semantic–spatial mismatch during cross-level feature fusion. First, the previously proposed DEConv operator is embedded into selected C3K2 stages to form C3DEConv; the contribution lies in its C3K2-compatible, detector-oriented integration rather than in a new differential-convolution formulation. Second, an Adaptive Feature Recalibration (AFRE) block constructed from three Recalibration Attention Units performs bidirectional interaction between shallow spatial details and deep semantic features. We also construct four-class bounding-box annotations for BIT-SIRST. RADFRNet achieves mAP@0.5 scores of 93.2% and 65.3% on BIT-SIRST and FLIR-ADAS-v2, improving YOLOv11-n by 4.4 and 8.5 percentage points, respectively. Under the same original $640\times 640$ GPU inference setup, the per-image latency increases from 3.3 to 6.7 ms on BIT-SIRST and from 3.0 to 7.8 ms on FLIR-ADAS-v2, corresponding to nominal throughputs of approximately 149 and 128 FPS for RADFRNet. The reported model-complexity values are 8.2 M and 8.6 M, respectively. These results show that RADFRNet retains high-rate GPU inference, although the accuracy gains are obtained at a clear computational cost; the model is therefore positioned as an accuracy-oriented detector rather than a latency-neutral replacement for YOLOv11-n. Full article

Contact-induced mistuning at the blade–disk interface is a major source of variability in the dynamic response of bladed disks, yet its identification remains challenging due to the intrinsic complexity of the contact phenomenon. Contact mistuning in this framework is defined as the variation of the contact topology at the blade–disk interface, i.e., the spatial distribution of contacting and non-contacting regions governing the mechanical interaction between components. This work proposes a data-driven framework for the identification of contact mistuning based on a reduced set of experimentally measured maps. A statistical representation of the contact space is constructed using Principal Component Analysis. This representation is subsequently exploited to generate physically consistent synthetic contact patterns. These are combined with blade intrinsic mistuning parameters to build a dataset linking contact conditions and blade variability to the natural frequencies of a disk–one-blade assembly. A k-nearest neighbors algorithm is employed for inverse identification using a subset of measured natural frequencies. The identification relies on modes selected through sensitivity analysis, and robustness is improved using a weighted distance metrics. The methodology is validated using both a reduced-order model based digital twin and experimental data. The results show accurate identification of contact patterns, confirmed through comparison of modal properties of the full mistuned assembly. Therefore, the proposed framework provides a practical tool for the experimental identification of contact mistuning in bladed disks. Full article

Infrared small target detection (IRSTD) supports remote-sensing surveillance and early warning, but infrared targets often occupy only a few pixels, exhibit weak appearance cues, and resemble thermal clutter. Vision-language detectors such as Grounding DINO provide a flexible prompt-driven detection interface, but direct transfer to IRSTD faces three mismatches: a single prompt cannot describe target appearance diversity, generic decoder queries are not calibrated for weak scale-sensitive targets, and IoU-style localization losses are unstable for few-pixel boxes. We propose PromptScaleDINO as a lightweight adaptation framework for Grounding DINO. It introduces an Anchor Prompt Bank (APB) to enrich the text input while keeping supervision connected to a stable anchor phrase, a Scale-Aware Query Refinement Network (SQRN) to refine selected queries according to predicted box scale and semantic confidence, and a geometry-aware localization objective that combines Normalized Wasserstein Distance (NWD) with Generalized Intersection over Union (GIoU). Under a unified detection protocol, PromptScaleDINO achieves its largest gains on the challenging real-scene IRSTD-1k benchmark while maintaining near-ceiling performance on SIRST and NUDT-SIRST. On IRSTD-1k, it reaches 88.32% F1 and 87.40% mAP@0.5, improving the baseline by 2.56 and 4.00 percentage points, respectively. These gains mainly reflect improved weak-target recovery at comparable precision; accordingly, we position the framework as a target-sensitivity and localization adaptation rather than an explicit false-alarm suppression method. Full article

Ukraine’s parliament is considering a monthly ownership levy on battery-electric vehicles (BEVs) of up to UAH 4000 as a road-fund replacement following the reinstatement of 20% value-added tax on BEV imports from January 2026, a fiscal reversal of the sustained incentive framework (zero customs duty since 2015; full VAT and excise relief from 2018) that has underpinned a fleet of approximately 246,000 registered BEVs in a wartime economy with acute petroleum import dependence. A five-year total-cost-of-ownership (TCO) model calibrated to Ukrainian market data (April 2026) computes the breakeven monthly levy across three charging scenarios and two energy-price assumptions; a structured comparative policy analysis maps the instrument against cross-jurisdictional adoption-suppression evidence; and a legal doctrinal analysis applies the fair-balance test under Article 1 of Protocol No. 1 to the European Convention on Human Rights. The breakeven levy (L*) ranges from UAH 931 to UAH 1626 per month under baseline conditions and from UAH 63 to UAH 862 under adverse energy-price assumptions, placing the proposed UAH 4000 rate at 2.5 to 4.3 times the threshold. Cross-jurisdictional evidence positions Ukraine in the high adoption-suppression zone for flat ownership charges. The legal analysis indicates that the incentive programme satisfies ECHR legitimate-expectation criteria and that the proposed levy is likely to fail both the proportionality and transitional adequacy limbs of the fair-balance test. The preferred alternative combines a kWh surcharge on public charging with a phased per-kilometre road-user charge and three-year grandfathering for existing owners. Full article

AA6061-T6 undergoes work hardening, dynamic recovery, and progressive flow softening during hot torsion, yet a systematic single-campaign dataset with quantified experimental uncertainty is absent from the literature. Gleeble hot torsion tests were conducted at eleven conditions from 250 to 450 °C and 0.91 to 9.07 s⁻¹. With the stress multiplier fixed a priori at α = 0.045 MPa⁻¹ from compression literature on this alloy, a two-stage calibration determined the remaining Garofalo–Arrhenius constants: the temperature-slope stage anchors Q = 151.1 kJ mol⁻¹ (consistent with Al lattice self-diffusion), and a global Zener–Hollomon regression conditional on Q yields n = 1.371 and A = 3.51 × 10¹⁰ s⁻¹; a fully simultaneous three-parameter fit is shown to be practically unidentifiable on the three-level matrix. Training AARE = 15.5% (R = 0.908); leave-one-out cross-validation gives AARE = 23.0%, bounding the predictive uncertainty. The Prasad instability map identifies 400–450 °C at 0.91–2.72 s⁻¹ as the optimal hot-forming window; flow instability is predicted at 350 °C (outright at 2.72 and 9.07 s⁻¹, with the 0.91 s⁻¹ condition at the map boundary), and macroscopic fracture was observed in all three specimens tested there. Optical microscopy in specimen T1 (ε_eq = 3.69) shows elongated subgrains at the gauge center and fine-grained zones near the fracture surface consistent with localized geometric dynamic recrystallization. The activation energy, smooth post-peak softening, and subgrain wall morphology identify dynamic recovery as the likely dominant restoration mechanism. Adiabatic heating and a 24% peak-stress repeatability scatter at the single repeated condition (450 °C, 9.07 s⁻¹) are quantified and propagated into the constitutive model uncertainty bounds. Because the training-to-cross-validation error gap (15.5% versus 23.0% AARE) reflects the limited three-level strain-rate matrix, the calibrated equation is recommended for interpolation within the tested window of 300 to 450 °C and 0.91 to 9.07 s⁻¹ (noting that 300 °C was tested only at 0.91 s⁻¹, so higher-rate predictions at that temperature are extrapolations) and for forming-window identification, not for extrapolation beyond this domain without additional validation data. Full article

In existing research on non-contact damage by high-power microwave (HPM) systems to electronic systems such as seekers, challenges include high risk, high cost, and a scarcity of HPM-specific damage assessment models. To address this, this paper proposes a multi-port HPM damage level assessment model based on interval-valued intuitionistic fuzzy (IVIF) sets. Firstly, electromagnetic modeling and field-circuit coupling simulation of multi-target structures provide input data from field-circuit simulations. Secondly, a multi-indicator fuzzy matrix reflecting system damage uncertainty is formed using interval fuzzy membership functions. Thirdly, an information entropy-driven dynamic weighting mechanism weights each indicator’s fuzzy distribution characteristics, and hybrid weights are constructed by dynamic and fixed weights to optimize parameter range rationality. Finally, using hybrid weights and the exponential damage mapping function, we achieve probabilistic fusion of multiple parameters and determine damage levels from the probability results. Assessment shows that under the same incident field strength of 25 kV/m, electromagnetic simulation software-based simulation assigns a composite damage probability of 0.85 to L-band Port 5 and 0.096 to X-band Port 6, highlighting the contrast between the dominant L-band hotspot (Port 5) and the low X-band response at Port 6 under identical irradiation. Full article

UAV infrared small-object detection is challenging because targets occupy few pixels, provide weak thermal contrast, and are easily confused with cluttered backgrounds. Although convolutional and Transformer-based detectors improve local representation and global context modeling, their predominantly spatial processing pipelines do not explicitly prevent fragile target responses from being attenuated during early encoding and multi-scale aggregation. To address this gap, we propose SFCD-Det, a spatial–frequency collaborative detector organized around a progressive preservation–purification–coordination methodology. A Wavelet-Transform Stem preserves low-frequency structures and localized high-frequency details before backbone encoding. A Feature Purification Layer regulates adjacent-scale interactions and suppresses clutter-dominated responses before aggregation, while a Spatial–Frequency Coordinated Feature Pyramid reconstructs multi-scale features using shallow spatial anchors, purified intermediate responses, and deep spatial–frequency priors. Experiments on four benchmarks show that SFCD-Det consistently outperforms the DEIM-N baseline. It achieves 62.6% mAP and 95.1% mAP₅₀ on HIT-UAV, 41.4% and 86.4% on IRSTD-1K, 16.8% and 46.8% on RGBTDronePerson, and 34.1% and 88.0% on USOD, respectively. These results demonstrate that SFCD-Det strengthens weak-target representation in UAV infrared imagery. Its additional gains on USOD suggest that the frequency mechanism may also benefit weak and spatially localized responses in visible imagery under low illumination or shadow. Full article

Electrohydrodynamic emission from a sessile droplet depends on a coupled balance among electric traction, capillarity, gravity, charge transport and liquid motion. The present work addresses only the electrostatic-loading part of that problem. Verification-backed axisymmetric and three-dimensional Laplace solvers are used to map how parallel-plate, on-axis-pin, off-axis-pin and bipolar double-pin electrodes redistribute the normal electric field over a prescribed conducting water-droplet interface. The primary response is the dimensionless electric capillary number, Ca_E = ε₀E_n²R_v/γ. To compare geometries on a common voltage scale, V₁ is defined as the applied voltage at which the peak prescribed-interface loading reaches Ca_E = 1. V₁ is a normalisation voltage and not a jetting or stability threshold. The axisymmetric solver reproduces the exact conducting-hemisphere solution to within 0.07% at the finest grid. The three-dimensional finite-difference results are mesh-assessed, and their normalised surface-field topology is cross-checked against an independently implemented Galerkin finite-element model. At 4 kV, the finite parallel-plate cell produces an apex enhancement of 3.24 relative to V/H. Replacing the plate with an on-axis 1 mm pin reduces the apex field by 39.5%, which corresponds to a 63% reduction in Ca_E, and increases V₁ from approximately 5.0 to 8.2 kV. Lateral pin displacement moves the surface-field maximum away from the apex and produces a broad nominal plateau near d = 5–7 mm, although the sub-grid steering distance remains sensitive to mesh and extraction settings. The bipolar double-pin configuration produces two symmetric surface-field maxima together with a near-null at the apex. This topology, but not its absolute magnitude, is reproduced by the finite-element cross-check. A Gaussian-process model interpolates the one-dimensional offset family accurately under leave-one-offset-out validation (R² = 0.999). Four Bayesian-optimisation trials locate the broad steering plateau but show no visible evaluation-count advantage over random sampling in this one-dimensional test. A three-mesh study gives a reported field-magnitude mesh-sensitivity estimate of approximately 6.2% at the finest grid (rising to about 9.5% at the h = 0.20 mm production mesh) for the representative three-dimensional case, and an indicative combined-uncertainty band of approximately 10% is shown for V₁ in the exploratory trade-off plot. Illustrative Young–Laplace profiles at contact angles of 70° to 110° preserve the comparative pin-versus-plate field reduction, whereas V₁ varies by up to approximately 50%. A simplified Peek-law screening estimate places corona inception (the pin being cathodic) in the approximate range of 4.8–10 kV, which is comparable to the on-axis-pin V₁, so gas discharge may intervene before large electrocapillary loading is reached in ambient air. The results establish electrode geometry as a controllable electrostatic-loading parameter while explicitly deferring coupled stability analysis and experimental validation. By resolving this loading on a single exact-solution-verified basis, the study quantifies electrode geometry as a control parameter that idealised enhancement factors and the nominal gap field cannot capture and provides a verified fixed-interface reference state for subsequent coupled electrohydrodynamic modelling. Full article

This paper investigates the enumeration of orientable vertex-transitive embeddings of complete graphs ${\mathsf{K}}_{p^2}$, where $p\ge 5$ is a prime. Building on the classification framework established by Li, we derive a precise count of these embeddings up to non-isomorphism. Furthermore, we provide explicit examples to demonstrate the realizability of the maps corresponding to our enumeration results. Full article

For surface crack detection in in-service concrete buildings, this study proposes DCFYOLO, a lightweight crack instance segmentation algorithm improved from YOLO11n-seg, which enhances feature representation and multi-scale contextual fusion to improve crack detection and segmentation performance in complex scenarios. Frequency Channel Attention (FCA) is introduced into the C3k2 units of the backbone network, where multi-scale pooling extracts multi-spectral information to enhance channel-level feature discrimination. Context Anchor Attention (CAA) is added to the PAN-FPN structure in the neck to address multi-scale feature fusion and long-range context modeling in complex scenes. A Dynamic Snake Convolution with Efficient Channel Attention (DSECA) module is constructed by combining dynamic snake convolution and efficient channel attention. Through a serial design, this module provides the multi-scale feature fusion process with both geometrically adaptive sampling and channel-discriminative optimization. Experimental results on the DeepCrack dataset show that DCFYOLO achieves 73.2% Box mAP@0.5 and 67.4% Mask mAP@0.5 with 2.47 M parameters, improving the baseline YOLO11n-seg by 2.8 and 2.7 percentage points, respectively. Ablation experiments verify the independent contribution of each improved module. The algorithm demonstrates a balance between segmentation accuracy and inference efficiency under lightweight constraints on the DeepCrack dataset, and can provide a reference for research on concrete surface crack instance segmentation. Full article