Search Results (19)

Public-sector sustainability and climate reporting increasingly address environmental exposure, governance, and financial effects, yet existing frameworks do not adequately disclose preparedness for low-probability, high-impact systemic risks whose probabilities, timing, thresholds, and transmission channels remain deeply uncertain. This article develops a Public-Sector Resilience Reporting Standard (PSRRS) as a pre-standard architecture for government preparedness disclosure. The design has three bounded objectives: diagnose cross-framework disclosure gaps, translate these gaps into a theoretically grounded capability-to-disclosure architecture, and demonstrate its analytical use through an illustrative Florida application and two hazard-neutral stress tests. The documentary corpus includes international sustainability and public-sector reporting standards, ISO and UNDRR resilience and continuity instruments, three Florida resilience documents, and peer-reviewed literature on resilience governance, decision-making under deep uncertainty, critical infrastructure interdependency, catastrophic uncertainty, climate-risk disclosure, public finance, climate-risk pricing, local-government credit risk, investor attention, and ransomware service disruption. A structured interpretive coding protocol classifies each framework as explicit, partial, or not explicit across nine disclosure dimensions; a codebook appendix identifies the assessment criteria, the a priori and inductively refined dimensions, and the validation boundaries. Florida is not treated as a basis for statistical or jurisdictional generalization. Instead, it illustrates how a comparatively developed resilience architecture may disclose statutory continuity, critical-asset data, project ranking, and output metrics while leaving systemic dependencies, adaptive triggers, long-horizon fiscal exposure, residual service risk, distributional effects, and assurance mechanisms insufficiently visible in the reviewed reporting corpus. AMOC and case-grounded cyber-fiscal stress tests show how the PSRRS shifts reporting from hazard inventories and funded projects toward auditable evidence of institutional capacity, adaptive readiness, and public-value protection. The article specifies mandatory, recommended, and optional clauses, evidence requirements, indicator examples, a disclosure index, a sample report structure, and a three-tier pilot conformity model. The contribution is conceptual and operational, but not yet a validated formal standard; cross-jurisdictional piloting, inter-rater coding, cost testing, assurance testing, and stakeholder consultation are identified as the next stage of standardization. Full article

Cross-dataset pothole segmentation is difficult because illumination, camera characteristics, asphalt texture, and boundary quality vary between training and deployment environments. This study proposes PotholeBAF-Net, an EfficientNet-B3 encoder–decoder network comprising a Low-Rank Transformer Bridge, a Boundary-Aware Fusion (BAF) decoder, and a lightweight Boundary Refinement Head. The bridge uses standard multi-head attention whose pre-softmax per-head affinity matrix is rank-bounded by the query/key dimension; BAF then regulates each encoder skip using a learned soft boundary prior, channel–spatial attention, and spatial–channel gated interpolation with decoder semantics. The model was trained and validated on the Large Public Dataset and evaluated without target-domain training or calibration on the independently collected 188-image Collected Taiz Dataset. Across three independent seeds, PotholeBAF-Net obtained external-test precision of $0.8562\pm 0.1040$, recall of $0.4854\pm 0.0716$, F1/Dice of $0.6196\pm 0.0395$, IoU of $0.4164\pm 0.0365$, and MCC of $0.5802\pm 0.0198$. Removing the boundary prior, Transformer bridge, or refinement head reduced mean IoU by 0.0763, 0.0685, and 0.0675, respectively. The network contains 12.463 M parameters and requires 4.779 GMACs. These findings support controlled skip transfer as a useful design for target-free pothole segmentation, while the remaining seed variability and recall-dominated failures show that shadows, weak boundaries, and visually ambiguous pavement remain unresolved challenges. Full article

Colorectal polyp segmentation requires accurate boundary delineation, but adapting large vision foundation models to endoscopic images can be computationally expensive. This study investigates whether the Segment Anything Model (SAM) can be specialized for polyp segmentation by updating only a small attention subspace. We propose PolypSAM-Lite, which freezes the SAM vision backbone and applies low-rank reparameterization only to the fused Query–Key–Value attention projections. The model uses bounding-box prompts, binary cross-entropy plus Dice loss, AdamW optimization, and evaluation on Kvasir-SEG with external testing on CVC-ClinicDB and ETIS-LaribPolypDB. PolypSAM-Lite updates 4.2 million parameters and achieves a Dice Similarity Coefficient of 0.9507 on Kvasir-SEG, compared with 0.8804 for zero-shot SAM. External Dice scores are 0.9271 on CVC-ClinicDB and 0.9198 on ETIS-LaribPolypDB, indicating cross-dataset generalization under domain shift. These results suggest that QKV-restricted low-rank attention reparameterization can provide an efficient and effective strategy for adapting SAM to colorectal polyp segmentation without full-backbone fine-tuning. Full article

Large language models (LLMs) built upon the Transformer architecture have achieved remarkable performance in natural language understanding, text generation and logical reasoning, while their internal working mechanisms remain poorly interpreted. This paper establishes a systematic mathematical analysis framework tailored for decoder-only Transformer LLMs, based on linear algebra, tensor analysis, probability theory, information theory, optimization dynamics and geometric deep learning. We conduct rigorous mathematical modeling and theoretical deduction on core modules including word embedding, position encoding, self-attention, feed-forward networks, training optimization and generalization reasoning, and explore the mathematical nature of semantic representation, contextual correlation, knowledge storage and logical inference within models. In this paper, we strictly distinguish between classic established Transformer theories and our original mathematical derivations and conclusions. Distinct from existing fragmented theoretical studies, this work presents six targeted novel contributions beyond conventional Transformer theories: (1) we construct the first full-process unified mathematical framework covering all core modules and the entire lifecycle of Transformer-based LLMs; (2) we provide strict mathematical proof to verify that single-head self-attention is essentially a kernel weighted average operation in reproducing kernel Hilbert space and derive the low-rank and sparse properties of attention weights; (3) we establish a high-dimensional non-convex optimization dynamics model for pre-training and mathematically prove that model training converges to flat local minima; (4) we derive a tighter upper bound of generalization error and quantify the quantitative relationship among model parameters, sequence length, training data scale and generalization performance; (5) we characterize the latent space as a low-curvature smooth Riemannian manifold and model logical reasoning as geometric transformation on this manifold; (6) we design multi-group controlled experiments on mainstream datasets to quantitatively validate all above theoretical conclusions. This paper further summarizes the inherent mathematical limitations of current Transformer LLMs and proposes feasible theoretical optimization paths, referring to state-of-the-art research published from 2021 to 2026. The outcomes of this research can provide solid mathematical theoretical support for improving model interpretability, optimizing network structures and boosting practical performance, and facilitate the transition of LLM research from empirical engineering practice to theory-driven development. Full article

Intelligent autonomous driving systems must not only predict the appropriate driving manoeuvre but also provide human-interpretable evidence that justifies the decision. However, existing methods typically address these objectives separately, leading to three practical limitations: multi-stage perception-to-language pipelines can propagate upstream perception errors into downstream explanations; post hoc saliency methods often produce pixel-level highlights that are difficult to interpret semantically; and decoupled decision and explanation modules cannot guarantee that the explanation reflects the same scene evidence used for behaviour prediction. In this paper, we propose a unified framework that jointly performs vehicle behaviour prediction and human-centric interpretation from a shared visual backbone. Specifically, a hierarchical Swin Transformer encodes the driving scene into a sequence of spatial tokens, which are processed by two complementary branches. The first branch, termed the Object Selection Module (OSM), learns a compact scene-level semantic representation through query-guided cross-attention, while the second branch extracts a small set of class-agnostic object-centric tokens without requiring bounding-box or segmentation supervision. These two representations are subsequently integrated by a Semantic–Object Fusion (SOF) module based on scaled dot-product attention, residual connections, and a feed-forward network. The behaviour prediction head operates on the fused representation, whereas the interpretation head leverages the semantic representation through a skip connection to preserve decision-relevant context. For surround-view perception, learnable per-camera embeddings are introduced to maintain viewpoint identity with negligible additional parameter cost. Furthermore, a compact language model fine-tuned via Low-Rank Adaptation (LoRA) generates fluent, label-conditioned natural-language justifications. Extensive experiments on two public benchmarks, BDD-OIA and nu-AD, demonstrate that the proposed framework consistently delivers superior performance and provides effective, human-readable interpretations of driving decisions. Full article

No-reference image quality assessment (NR-IQA) models achieve high correlation with human mean opinion scores (MOS) on clean benchmarks, yet recent work shows they can be highly vulnerable to small adversarial perturbations that severely degrade ranking consistency, including in black-box settings. We introduce the Spectral Robustness Mixer (SRM), a lightweight neck inserted between an NR-IQA backbone and regression head, designed to reduce adversarial sensitivity without changing the dataset, label format, or target metric. SRM couples (i) deep-to-shallow cross-scale fusion via a Nyström low-rank attention surrogate, (ii) ridge-conditioned landmark kernels with ridge regularization, solved via numerically stable small-matrix factorization (SVD/LU) to improve conditioning, and (iii) variance-aware entropy-regularized fusion gates with a bounded gain cap to limit gradient amplification. We evaluate SRM on TID2013 and KonIQ-10k under a white-box ${\mathcal{l}}_{\infty }/{\mathcal{l}}_2$ attack ensemble that includes per-image regression objectives and a correlation-aware pairwise inversion objective (a ranking-inspired surrogate for correlation inversion), with expectation-over-transformation (EOT) and anti-gradient masking checks. At $ϵ=4/255$ (${\mathcal{l}}_{\infty }$), SRM improves worst-case robust Spearman’s rank-order correlation coefficient (SROCC; defined as the minimum over our fixed attack ensemble) by an absolute $0.06$–$0.08$ SROCC points (i.e., correlation-coefficient units, not percentage gain) across datasets/backbones, while keeping clean SROCC within $0.00$–$0.01$ of the baseline. We observe similar trends for Pearson linear correlation coefficient (PLCC). Full article

The rapid development of Convolutional Neural Networks (CNNs) has markedly boosted the performance of object detection in remote sensing. Nevertheless, tiny objects typically account for an extremely small fraction of the total area in remote sensing images, rendering existing IoU-based or area-based evaluation metrics highly sensitive to minor pixel deviations. Meanwhile, classic detection models face inherent bottlenecks in efficiently mining discriminative features for tiny objects, leaving the task of tiny object detection in remote sensing images as an ongoing challenge in this field. To alleviate these issues, this paper proposes a tiny object detection method based on Normalized Gaussian Label Assignment and Multi-scale Hybrid Attention. Firstly, 2D Gaussian modeling is performed on the feature receptive field and the actual bounding box, using Normalized Bhattacharyya Distance for precise similarity measurement. Furthermore, a candidate sample quality ranking mechanism is constructed to select high-quality positive samples. Finally, a Multi-scale Hybrid Attention module is designed to enhance the discriminative feature extraction of tiny objects. The proposed method achieves 25.7% and 27.9% AP on the AI-TOD-v2 and VisDrone2019 datasets, respectively, significantly improving the detection capability of tiny objects in complex remote sensing scenarios. Full article

Head and neck squamous cell carcinoma (HNSCC) refers to the malignancy of squamous cells in the head and neck region. Ranked as the seventh most common cancer worldwide, HNSCC has a very low survival rate, highlighting the importance of finding therapeutic targets for the disease. Integrins are cell surface receptors that play a crucial role in mediating cellular interactions with the extracellular matrix (ECM). Within this protein family, Integrin αV (ITGAV) has received attention for its important functional role in cancer progression. In this study, we first demonstrated the upregulation of ITGAV expression in HNSCC, with higher ITGAV expression levels correlating with significantly lower overall survival, based on TCGA (the Cancer Genome Atlas) and GEO datasets. Subsequent in vitro analyses revealed an overexpression of ITGAV in highly invasive HNSCC cell lines UM1 and UMSCC-5 in comparison to low invasive HNSCC cell lines UM2 and UMSCC-6. In addition, knockdown of ITGAV significantly inhibited the migration, invasion, viability, and colony formation of HNSCC cells. In addition, chromatin immunoprecipitation (ChIP) assays indicated that SOX11 bound to the promoter of ITGAV gene, and SOX11 knockdown resulted in decreased ITGAV expression in HNSCC cells. In conclusion, our studies suggest that ITGAV promotes the progression of HNSCC cells and may be regulated by SOX11 in HNSCC cells. Full article

Forest fires rank among the costliest and deadliest natural disasters globally. Identifying the smoke generated by forest fires is pivotal in facilitating the prompt suppression of developing fires. Nevertheless, succeeding techniques for detecting forest fire smoke encounter persistent issues, including a slow identification rate, suboptimal accuracy in detection, and challenges in distinguishing smoke originating from small sources. This study presents an enhanced YOLOv8 model customized to the context of unmanned aerial vehicle (UAV) images to address the challenges above and attain heightened precision in detection accuracy. Firstly, the research incorporates Wise-IoU (WIoU) v3 as a regression loss for bounding boxes, supplemented by a reasonable gradient allocation strategy that prioritizes samples of common quality. This strategic approach enhances the model’s capacity for precise localization. Secondly, the conventional convolutional process within the intermediate neck layer is substituted with the Ghost Shuffle Convolution mechanism. This strategic substitution reduces model parameters and expedites the convergence rate. Thirdly, recognizing the challenge of inadequately capturing salient features of forest fire smoke within intricate wooded settings, this study introduces the BiFormer attention mechanism. This mechanism strategically directs the model’s attention towards the feature intricacies of forest fire smoke, simultaneously suppressing the influence of irrelevant, non-target background information. The obtained experimental findings highlight the enhanced YOLOv8 model’s effectiveness in smoke detection, proving an average precision (AP) of 79.4%, signifying a notable 3.3% enhancement over the baseline. The model’s performance extends to average precision small (APS) and average precision large (APL), registering robust values of 71.3% and 92.6%, respectively. Full article

Recently, intelligent reflecting surfaces (IRSs) have drawn huge attention as a promising solution for 6G networks to enhance diverse performance metrics in a cost-effective way. For massive connectivity toward a higher spectral efficiency, we address an intelligent reflecting surface (IRS) to an uplink nonorthogonal multiple access (NOMA) network supported by a multiantenna receiver. We maximize the sum rate of the IRS-aided NOMA network by optimizing the IRS reflection pattern under unit modulus and practical reflection. For a moderate-sized IRS, we obtain an upper bound on the optimal sum rate by solving a determinant maximization (max-det) problem after rank relaxation, which also leads to a feasible solution through Gaussian randomization. For a large number of IRS elements, we apply the iterative algorithms relying on the gradient, such as Broyden–Fletcher–Goldfarb–Shanno (BFGS) and limited-memory BFGS algorithms for which the gradient of the sum rate is derived in a computationally efficient form. The results show that the max-det approach provides a near-optimal performance under unit modulus reflection, while the gradient-based iterative algorithms exhibit merits in performance and complexity for a large-sized IRS with practical reflection. Full article

Back in the 1930s, the aphorism “publish or perish” first appeared in an academic context. Today, this phrase is becoming a harsh reality in several academic environments, and scientists are giving increasing attention to publishing and disseminating their scientific work. Soil erosion modelers make no exception. With the introduction of the bibliometric field, the evaluation of the impact of a piece of scientific work becomes more articulated. The post-publication impact of the research became an important aspect too. In this study, we analyse the outreach and the impact of the literature on soil erosion modelling using the altmetric database, i.e., Altmetric. In our analysis, we use only a small fraction (around 15%) of Global Applications of Soil Erosion Modelling Tracker (GASEMT) papers because only 257 papers out of 1697 had an Altmetric Score (AS) larger than 0. We observed that media and policy documents mentioned more frequently literature dealing with global-scale assessments and future projection studies than local-scale ones. Papers that are frequently cited by researchers do not necessarily also yield high media and policy outreach. The GASEMT papers that had an AS larger than 0 were, on average, mentioned by one policy document and five Twitter users and had 100 Mendeley readers. Only around 5% and 9% of papers with AS > 0 appeared in news articles and blogs, respectively. However, this percentage was around 45% for Twitter and policy mentions. The top GASEMT paper’s upper bound was around 1 million Twitter followers, while this number was around 10,000 for the 10th ranked GASEMT paper. The exponentially increasing trend for erosion modelling papers having an AS has been confirmed, as during the last 3 years (2014–2017), we estimated that the number of entries had doubled compared to 2011–2014 and quadrupled if we compare it with 2008–2011. Full article

Anticipating pedestrian crossing behavior in urban scenarios is a challenging task for autonomous vehicles. Early this year, a benchmark comprising JAAD and PIE datasets have been released. In the benchmark, several state-of-the-art methods have been ranked. However, most of the ranked temporal models rely on recurrent architectures. In our case, we propose, as far as we are concerned, the first self-attention alternative, based on transformer architecture, which has had enormous success in natural language processing (NLP) and recently in computer vision. Our architecture is composed of various branches which fuse video and kinematic data. The video branch is based on two possible architectures: RubiksNet and TimeSformer. The kinematic branch is based on different configurations of transformer encoder. Several experiments have been performed mainly focusing on pre-processing input data, highlighting problems with two kinematic data sources: pose keypoints and ego-vehicle speed. Our proposed model results are comparable to PCPA, the best performing model in the benchmark reaching an F1 Score of nearly $0.78$ against $0.77$. Furthermore, by using only bounding box coordinates and image data, our model surpasses PCPA by a larger margin ($F1=0.75$ vs. $F1=0.72$). Our model has proven to be a valid alternative to recurrent architectures, providing advantages such as parallelization and whole sequence processing, learning relationships between samples not possible with recurrent architectures. Full article

Sigma-1 receptor (S1R) is an intracellular, multi-functional, ligand operated protein that also acts as a chaperone. It is considered as a pluripotent drug target in several pathologies. The publication of agonist and antagonist bound receptor structures has paved the way for receptor-based in silico drug design. However, recent studies on this subject payed no attention to the structural differences of agonist and antagonist binding. In this work, we have developed a new ensemble docking-based virtual screening protocol utilizing both agonist and antagonist bound S1R structures. This protocol was used to screen our in-house compound library. The S1R binding affinities of the 40 highest ranked compounds were measured in competitive radioligand binding assays and the sigma-2 receptor (S2R) affinities of the best S1R binders were also determined. This way three novel high affinity S1R ligands were identified and one of them exhibited a notable S1R/S2R selectivity. Full article

We consider a gauge theory of vector fields in $3D$ Minkowski space. At the free level, the dynamical variables are subjected to the extended Chern–Simons (ECS) equations with higher derivatives. If the color index takes n values, the third-order model admits a $2n$-parameter series of second-rank conserved tensors, which includes the canonical energy–momentum. Even though the canonical energy is unbounded, the other representatives in the series have a bounded from below the 00-component. The theory admits consistent self-interactions with the Yang–Mills gauge symmetry. The Lagrangian couplings preserve the energy–momentum tensor that is unbounded from below, and they do not lead to a stable non-linear theory. The non-Lagrangian couplings are consistent with the existence of a conserved tensor with a 00-component bounded from below. These models are stable at the non-linear level. The dynamics of interacting theory admit a constraint Hamiltonian form. The Hamiltonian density is given by the 00-component of the conserved tensor. In the case of stable interactions, the Poisson bracket and Hamiltonian do not follow from the canonical Ostrogradski construction. Particular attention is paid to the “triply massless” ECS theory, which demonstrates instability even at the free level. It is shown that the introduction of extra scalar field, serving as Higgs, can stabilize the dynamics in the vicinity of the local minimum of energy. The equations of motion of the stable model are non-Lagrangian, but they admit the Hamiltonian form of dynamics with a Hamiltonian that is bounded from below. Full article

Coal combustion is an anthropogenic source of mercury (Hg) emissions to the atmosphere. The strong toxicity and bioaccumulation potential have prompted attention to the control of mercury emissions. Pyrolysis has been regarded as an efficient Hg removal technology before coal combustion and other utilization processes. In this work, the Hg speciation in coal and its thermal stability were investigated by combined sequential chemical extraction and temperature programmed decomposition methods; the effect of coal rank on Hg speciation distribution and Hg release characteristics were clarified based on the weight loss of coal; the amount of Hg released; and the emission of sulfur-containing gases during coal pyrolysis. Five species of mercury were determined in this study: exchangeable Hg (F1), carbonate + sulfate + oxide bound Hg (F2), silicate + aluminosilicate bound Hg (F3), sulfide bound Hg (F4), and residual Hg (F5), which are quite distinct in different rank coals. Generally, Hg enriched in carbonates, sulfates, and oxides might migrate to sulfides with the transformation of minerals during the coalification process. The order of thermal stability of different Hg speciation in coal is F1 < F5 < F2 < F4 < F3. Meanwhile, the release of Hg is accompanied with sulfur gases during coal pyrolysis, which is heavily dependent on the coal rank. Full article