Search Results (1,020)

Infrared small target detection (IRSTD) is a challenging task in intelligent infrared sensing and electronic imaging systems, because dim targets often occupy only a few pixels and are easily disturbed by clutter, noise, and low-contrast background structures. A practical detector should preserve pixel-level target cues while suppressing target-like false responses. This paper proposes a Pyramid Target Perception Network (PTPN) for single-frame pixel-level IRSTD. The network integrates three complementary components: an Efficient Context Modeling (ECM) encoder employing 7 × 7 depthwise separable convolution for lightweight contextual feature extraction, a multi-scale target cross-attention (MTCA) module for hierarchical feature interaction, and a small-target feature pyramid network (STFPN) for target-preserving multi-scale aggregation. In addition, a physics-constrained loss (PCL) is introduced during training to regularize predictions according to infrared imaging characteristics, including point spread consistency, target-region relative intensity consistency, and signal-to-noise-ratio-aware separability. Experiments on IRSTD-1k, NUAA-SIRST, and NUDT-SIRST demonstrate that PTPN achieves $IoU$ scores of 71.87%, 79.56%, and 86.47%, respectively, with 4.55M parameters, 4.96G FLOPs at an input resolution of 256 × 256, and an inference speed of 45.0 FPS. Although PTPN achieves competitive overall performance, it does not attain the highest $IoU$ on NUDT-SIRST, indicating that pixel-level target-region estimation under complex scenes remains an area for further improvement. Overall, PTPN provides an effective balance between target localization, false-alarm suppression, and computational efficiency, supporting its potential application in AI-driven infrared image processing and intelligent electronic sensing systems. Full article

Confronted with severe global challenges, including water scarcity, excessive use of chemical fertilizers and pesticides, and heavy metal contamination in soils, conventional agricultural technologies exhibit marked limitations in integrated water–fertilizer management and non-point source pollution control. Leveraging their excellent water retention capacity, intelligent sustained-release properties, and environmental responsiveness, hydrogels offer innovative solutions to advance sustainable agricultural development. This review comprehensively outlines the fundamental types, crosslinking mechanisms, and key functional properties of hydrogels, with a focused discussion on their agricultural deployment as high-efficiency soil conditioners, fertilizer vectors, and pesticide carriers; it deciphers the microscopic water-holding mechanisms under the tristate water model, delineates the divergent water-uptake and retention behaviors between ionic and non-ionic hydrogels, and clarifies the cyclic water-holding and release mechanisms of hydrogels during soil amelioration. Thise paper further synthesizes hydrogel-enabled environmental remediation applications, in which heavy metals and pesticide residues in soils and aquatic systems are removed via functional-group coordination adsorption or photocatalytic degradation; concurrently, hydrogels have been shown to activate plant systemic immunity through calcium-signaling pathways, thereby inducing broad-spectrum antiviral defense responses. Moreover, hydrogels can be integrated into precision agriculture frameworks to enable real-time monitoring of crop physiological status and to support targeted irrigation and fertilization management. This work also evaluates the role of hydrogels in promoting seed germination, root system development, crop metabolic regulation, and stress resilience, while introducing tailored application strategies across distinct plant growth stages. Their documented economic advantages include water conservation, enhanced crop yields, reduced dependence on synthetic fertilizers, and lower labor costs. Nevertheless, the large-scale implementation of hydrogels continues to face multifaceted challenges—particularly poor degradability and latent ecological risks, as conventional polyacrylamide (PAM)-based gels resist soil mineralization and retain potentially neurotoxic monomers, leaving a critical gap in multi-annual field data concerning their non-target interference with native soil aggregate evolution, pore distribution, and rhizospheric carbon–nitrogen footprints. Mechanistically, many hydrogels with tensile strengths below 1 MPa are highly susceptible to three-dimensional network collapse under high-salinity osmotic shock and tillage mechanical stress, exhibiting a precipitous drop in water retention after more than three wet–dry cycles due to deficient long-term structural stability. Compounding these technical gaps are elevated production costs and low farmer adoption, driven by the absence of texture-specific performance thresholds—such as an available water increment ≥ 40% for sandy soils—and the lack of established life-cycle cost models and farmer incentive mechanisms for bio-based hydrogels. Moving forward, hydrogel technology should pivot toward materials innovation and cost-reduction engineering to broaden its applicability, employ ≥3-year, multi-habitat regional trials to delineate ecological benefit–risk boundaries, and ultimately position hydrogels as pivotal enablers of sustainable, green agricultural paradigms. Full article

Unmanned aerial vehicle (UAV) imagery supports intelligent surveillance, environmental monitoring, traffic management, and infrastructure inspection. Yet aerial detection is difficult when objects are small, crowded, and observed at markedly different scales. Background clutter and illumination changes further weaken target cues and impair localization. We therefore propose HD-YOLO, a lightweight multiscale detector for small objects in UAV imagery. Its Multi-Dilation Shared Convolution Kernel (DSCK) extracts local texture and contextual information with shared dilated kernels. The Hybrid Dilated Bidirectional Feature Pyramid Network (HDFPN) reconstructs global and local cues before bidirectional aggregation, enabling high-resolution evidence to reach the prediction layers. The Efficient and Slim Head (ES-Head) combines shared operations with differential convolution to reduce cost and strengthen boundary-sensitive features. A joint ShapeIoU and Normalized Wasserstein Distance loss improves regression for small, irregular objects. Together, these components reduce missed detections in dense, cluttered scenes without relying on large model capacity. On VisDrone2019, HD-YOLO improves precision, recall, mAP50, and mAP50:95 over YOLOv8n by 6.9%, 7.2%, 8.2%, and 5.2%, respectively, while reducing parameters from 3.0 M to 0.9 M. Evaluations on TinyPerson and HIT-UAV also support its utility for tiny pedestrians and infrared aerial targets. HD-YOLO therefore improves small-object detection with a compact parameter footprint, while direct hardware benchmarks remain necessary to establish deployment efficiency. Full article

Maize diseases affecting leaves, stalks, and ears can substantially reduce yield and quality; therefore, rapid and accurate recognition in complex field environments is important for intelligent agricultural monitoring. To address the large-scale variation, weak fine-grained texture, and strong background interference associated with multi-part maize diseases, this study proposes YOLOv11-MPD (YOLOv11 for Maize Multi-Part Disease Detection), a maize disease detection algorithm based on YOLOv11n. The method jointly improves spatial position awareness, shallow detail preservation, local-context modeling, key semantic-region enhancement, and lightweight detection-head reconstruction. RFCAConv, C3k2_RFCAConv, and Detect_LSDECD are introduced into the baseline network to strengthen directional texture modeling, multi-scale feature aggregation, and detection-head feature representation. FG-RFCAConv, HGD-C3k2, LCA-C3k2, and GRN-BiAttn are further designed for high-frequency differential gated detail compensation, P3 high-resolution detail enhancement, local-context fusion, and global-response-normalized attention regulation, respectively. Experimental results show that YOLOv11-MPD achieves Precision, Recall, mAP50, and mAP50-95 of 72.3%, 72.8%, 79.5%, and 50.2%, improving YOLOv11n by 2.4, 2.5, 2.9, and 2.4 percentage points, respectively, while reducing parameters from 2.6 M to 2.4 M. These results indicate that, within the scope of the dataset used in this study, YOLOv11-MPD improves multi-part maize disease detection under complex field conditions. However, the current conclusions are limited to the constructed dataset, and further validation using larger multi-region, multi-season, and multi-device datasets is required to evaluate its broader generalization ability. Full article

To reduce aggregate gradation counting errors caused by over-segmentation and under-segmentation of stacked and clustered aggregates with mixed types and diverse spatial distributions in hydraulic concrete, this study proposes BC-GECO2, a coarse and fine aggregate segmentation and counting method. Firstly, a BAHiera feature extraction network is designed to extract multi-scale deep features through edge-aware attention. In addition, a DFG-Edge module is developed to enhance the boundary features of densely distributed aggregates by integrating wavelet transform with a gated fusion mechanism, thereby alleviating the loss of small aggregate features during downsampling. Secondly, a CSFM-GFFCA module is constructed, in which a dual-branch structure is employed to adaptively fuse adjacent-scale features, strengthen the edge responses of densely distributed small aggregates, and enhance cross-layer feature interaction. Finally, a joint optimization function combining Focal loss and counting loss is established to guide the model toward hard-to-classify pixels, especially boundary pixels, thereby improving segmentation integrity and counting accuracy. Experiments conducted on an aggregate dataset collected from practical construction sites show that, compared with the baseline GECO2 model, the proposed method improves the average segmentation IoU, Dice, and BIoU by 2.92%, 5.04%, and 2.83%, respectively, while reducing the average counting MAE and RMSE by 6.92 and 15.65, respectively. Moreover, BC-GECO2 exhibits superior robustness and generalization capability under different stacking densities and blurred-boundary scenarios, providing technical support for the intelligent development of rapid concrete gradation detection. Full article

Large language models (LLMs) are increasingly employed in scholarly peer review, yet their suitability as autonomous evaluators remains uncertain. Using the ICLR 2025 review process, this study compares 2401 human reviews with 7203 reviews produced in separate, context-isolated API runs using Claude Sonnet 4.5, GPT-5.2 Thinking, and Gemini 3 Pro Preview across decision agreement, review-text characteristics, inter-model consistency, and human–AI aggregation. Raw LLM scores showed systematic leniency and score compression. A 0.1-point grid search identified thresholds of 6.2, 6.3, and 6.7 for Claude, GPT, and Gemini, respectively; repeated stratified cross-validation reproduced these thresholds. When applied without retuning to a stratified balanced sample of 300 ICLR 2024 papers, decision-agreement accuracy was 0.927, 0.913, and 0.930. Independent human coding of research type and primary field showed substantial pre-adjudication agreement (Cohen’s kappa = 0.774 and 0.714), and the recalculated analyses did not support H3. Review-text indicators showed similar structural completeness across sources but uneven critical-section length; these descriptive measures do not establish review quality. Human-containing aggregation rules showed higher agreement with conference decisions than corresponding AI-only rules, without establishing independent review quality or causal complementarity. A textual-overlap check found very low exact eight-gram containment, and manual inspection of the highest-similarity 1% found shared manuscript content or domain terminology rather than reviewer-specific evaluative language; possible prior exposure nevertheless could not be excluded. Full article

Real estate investment decisions are shaped by a complex environment of macroeconomic, demographic, and structural factors, where traditional linear assessment methods often fail to capture nonlinear relationships influencing aggregate housing market performance. The aim of this article is to develop a data-driven artificial intelligence framework for assessing how macroeconomic and structural conditions influence aggregate housing market performance and for providing a conceptual basis for evaluating real estate investment strategies under different economic contexts. The study uses machine learning algorithms that allow for modeling complex relationships between investment return indicators and key macroeconomic factors, such as economic growth rates, price dynamics, population concentration, and long-term structural changes. Unlike traditional econometric methods, the proposed approach identifies nonlinear and regime-dependent relationships between macroeconomic conditions and housing market performance, providing insights that can support the interpretation of different investment strategies. The results show that the factors determining investment returns are not universal, and their significance depends on the broader economic regime and market structure. This allows us to examine how changing macroeconomic conditions influence aggregate housing market performance and to discuss the potential implications for different investment strategies. The study contributes by proposing an artificial intelligence-based methodological framework that combines predictive modelling with explainable AI to support the analysis of macroeconomic influences on housing markets and to inform strategic real estate investment decision-making within complex socioeconomic systems. Full article

Accurately forecasting waypoint-level tropical cyclone (TC) intensity, defined as the local wind speed at specific maritime route waypoints under TC influence, is crucial for navigation safety and voyage planning. Conventional studies mainly focus on the central intensity of TC systems and underutilize the complementary value of multimodal meteorological data with inconsistent sampling intervals. To address these challenges, this study proposes a multimodal-augmented conditional diffusion model (MADiff) for waypoint-level TC intensity prediction. To exploit the potential of multimodal inputs, we first design a temporal-adaptive dynamic convolution module (TDConv) to capture multi-timescale features, mitigating multimodal sampling discrepancies without rigid temporal alignment. Second, we develop a discriminative cross-fusion module (DisCF) to aggregate multi-timescale features across diverse modalities, quantifying multimodal heterogeneity and integrating valuable modality-specific features while suppressing noise interference. Fused features are fed into a diffusion model with physics-informed regularization to generate final intensity forecasts. Extensive experiments on four Western North Pacific datasets show that MADiff achieves average MAE and RMSE values of 2.08 kt and 2.37 kt, respectively, for 12 h intensity forecasting. Compared with the state-of-the-art baseline (TC-Clouds-DP), MADiff yields substantial performance improvements, reducing MAE by 16.3% and RMSE by 10.6% on average. This study provides an effective framework for fine-grained TC intensity forecasting, offering valuable insights for extreme marine weather early warning and intelligent navigation decision-making. Full article

Recreational coastal waters are increasingly affected by urbanization, maritime activities, and tourism, creating a need for predictive tools that support proactive water quality management. This study proposes an explainable machine learning framework for forecasting near-future changes in the Tourism-Oriented Seawater Quality Index (SeaWQI-T) using high-frequency seawater monitoring data collected in the Gulf of Vlorë, Albania. A summer monitoring campaign (June–August 2025) produced 102,988 physicochemical observations from six monitoring stations using an unmanned surface vehicle equipped with a Horiba U53 multiparameter sonde. Following quality control and temporal aggregation, the data were used to formulate a multiclass forecasting problem (decrease, stable, or increase), and Logistic Regression, Random Forest, and Extreme Gradient Boosting (XGBoost) models were evaluated across multiple forecasting horizons. XGBoost achieved the best validation performance, while Random Forest demonstrated superior generalization on the independent test dataset and provided the most stable explainability results. SHapley Additive exPlanations (SHAP) identified SeaWQI-T dynamics, turbidity, dissolved oxygen, and oxidation–reduction potential as the most influential predictors. The proposed framework demonstrates that integrating explainable machine learning with autonomous high-frequency monitoring can provide accurate, interpretable forecasts to support intelligent coastal recreation management and sustainable tourism planning. Full article

Recent progress in intelligent, adaptive, and security-aware Software-Defined Wireless Sensor Networks (SDWSNs) is driven by the integration of Machine Learning (ML) with Software-Defined Networking (SDN) and Wireless Sensor Networks (WSNs). This systematic survey analyzes 46 peer-reviewed articles published between 2024 and 2025, providing a problem-oriented synthesis of ML-SDWSN research. Emphasizing security, routing, and performance optimization, with a particular focus on deployment architectures, the survey identifies three major trends: increased adoption of ensemble and Reinforcement Learning (RL) methods for security and adaptive control; broader implementation of edge-based ML to minimize inference latency; and greater emphasis on privacy-preserving techniques, especially Federated Learning (FL). The survey presents a structured taxonomy encompassing seven thematic areas: Distributed Denial-of-Service (DDoS) mitigation, Intrusion Detection Systems (IDSs), routing optimization, Quality of Service (QoS) management, privacy preservation, data integrity, and network-efficiency optimization. Findings are synthesized from over 120 experimental configurations reported in the literature. Due to substantial differences among the reviewed studies in terms of datasets, network topologies, hardware platforms, measurement definitions, and validation methodologies, the reported values are presented as descriptive cross-study aggregates rather than direct comparative benchmarks or formal effect-size estimates. Within these constraints, the survey identifies recurring trade-offs among accuracy, latency, scalability, and privacy. It provides evidence-based design considerations for researchers and practitioners. The survey also highlights eight critical research gaps, including limited multi-dataset validation, a lack of real-world deployments, insufficient scalability analysis, and the need for rigorous evaluation of RL-based SDWSN control. Full article

Cross-organization cyber defense must reconcile collaborative learning with privacy and adversarial robustness, yet standard federated learning ships full gradient tensors, leaking sensitive posture and inviting Byzantine manipulation. We present FedMARL-LTI, a federated multi-agent reinforcement learning framework whose architecture answers both pressures with a single decision: each organization’s threat intelligence is shared only as a differentially private 768-dimensional semantic embedding, never as raw data. In the evaluated system, a Weight-DP-protected model-weight delta is also exchanged through the federated aggregator (the semantic abstraction embedding is a parallel channel); the privacy guarantee below is stated for the semantic abstraction channel, and an embeddings-only architecture—which the guarantee enables—is the design this points toward. The contribution is fourfold. (1) Semantic Abstraction (SA) channel: per organization, each round, the local gradient is summarized by an LLM, projected to a 768-dim embedding, L2-clipped, and Gaussian-noised before any numeric quantity leaves the host. The bottleneck reduces the aggregate noise magnitude—the expected L2 norm of the DP noise vector—from $O\left(\sqrt{d_{\mathrm{model}}}\right)$ to $O\left(\sqrt{m}\right)$ with $m=768\ll d_{\mathrm{model}}\approx 3\times {10}^5$. (2) Formal privacy analysis: the SA + DP cascade satisfies $(\epsilon ,\delta )$-DP and bounds per-round mutual information leakage by $\min \{T_{\mathrm{tok}}{\log }_{2}V,m/2{\log }_2(1+C^2/\left(m{\sigma }^2\right)\left)\right\}$, with Rényi composition over $T$ federation rounds. Scope of the guarantee: this bound certifies (i) the semantic-abstraction channel. It does not by itself cover (ii) the weight-aggregation channel, whose Weight-DP protection is analyzed separately, nor (iii) the whole deployed system, which is the composition of the two. We therefore state the ≈1.4-bit/MI bound as a per-round guarantee on information leaving the organization through the SA channel not over every byte the system emits; an embeddings-only configuration—which this bound enables—closes the gap to a whole-system guarantee. (3) Byzantine-resilient ClippedClustering aggregator combining L2 clipping with cosine-similarity clustering. (4) Hierarchical MARL policy with threat-profile-aware LLM-IRR reward shaping, wired end-to-end and disclosed honestly (the evaluated system uses a deterministic Johnson–Lindenstrauss projection in place of the LLM call for reproducibility; the architecture is thus LLM-compatible rather than dependent on a specific model, and a full LLM deployment is the planned extension). We evaluate on CybORG CAGE-4 with $n=5$ organizations, 30 federation rounds × 5 episodes × 100 steps per round. Releasing the SA channel in parallel shows no statistically detectable reward cost at N = 5 vs. the no-privacy baseline; this is measured at reward-shaping coefficient β = 0, so it establishes that the private semantic release does not disturb weight-channel training rather than that semantic sharing improves defense: SA-only $\Delta$reward = $+4.58$ ($t=+1.37$, NS), dual SA + Weight-DP $\Delta$reward = $+4.31$ ($t=+1.30$, NS), all $N=5$ seeds, all $\left|t\right|<1.4$. A controlled signal/noise probe confirms a $19.58\times$ improvement of SA over Weight-DP at a fixed DP budget—matching the predicted $\sqrt{d/m}\approx 19.8$. Under Byzantine sign_flip at 30% ($N=15$), ClippedClustering is directionally strongest ($F_1=0.025$ vs. FedAvg $0.020$, Krum $0.016$) but the edge is not statistically significant (CC vs. Krum $t=+1.59$, $p=0.15$, $d=+0.58$; the earlier $N=5$ “$3.4\times$” gap was small-sample optimism); its Byzantine behavior is on the harsher random_noise attack. Under a corrected implementation, the undefended baselines do not diverge or collapse; the earlier reading (Krum $0.002$, ClippedClustering $0.020$) was a noise-injection artifact and is withdrawn; ClippedClustering is now directionally best on F1 but not significantly, and trails Krum on reward (superseded Cohen’s $d=+3.77$). The cooperative-PPO family (MAPPO, IPPO) outperforms value/actor-critic (QMIX, MADDPG) by $\approx 20$ reward units, $p<0.001$. All host-level $F_1$ values stay below $0.05$ at the 15K-step training horizon used here; the relative claims of the paper (no detectable privacy reward cost, ClippedClustering’s competitive (not decisive) Byzantine behavior on the harsher attacks, cooperative-PPO dominance) are unaffected by this scope. A 200K-step long-horizon replication lifts $F_1$ above the 15K plateau (to $\approx 0.044$, $N=5$)—confirming that horizon, not the privacy/Byzantine machinery, gates absolute accuracy—but a finer 60-checkpoint run shows the climb is volatile and non-monotonic and does not reach deployment-grade, an honest stability-not-compute limitation. FedMARL-LTI is therefore presented as a proof-of-concept for the relative privacy and robustness trade-offs it isolates, not as an operationally deployable cyber defense system. We release all 141 raw run JSON outputs (Phases 1–3, the L4 backend comparison, and the algorithm/aggregator baselines), the figures, and analysis scripts for replication. Full article

Patients with implanted cardiac devices are a rapidly growing imaging population, and electrocardiogram-gated cardiac computed tomography (CT) is increasingly used to characterize device geometry, multi-device relationships, and dynamic behavior across the cardiac cycle. Cinematic rendering (CR) is a photorealistic three-dimensional (3D) visualization technique for cardiac CT whose established contribution in this population is communicative: it conveys 3D device geometry and material distinctions within a single rendered volume. We describe a demonstrative case series extending CR across the cardiac cycle—time-resolved “4D cine” CR—to depict dynamic device behavior and time-resolved multi-device interaction in a single volume; this is an illustrative technical experience rather than a systematic evaluation of diagnostic performance. Illustrative examples include an EVOQUE transcatheter tricuspid valve rendered together with concurrent surgical mitral and transcatheter aortic valves, a left atrial appendage occlusion device, a normally positioned Impella catheter, and a HeartMate 3 left ventricular assist device (LVAD). Across cases, 4D cine CR feasibility scaled inversely with metallic burden—the aggregate volume and radiodensity of metallic device components within the scan field—with renderings informative for low-metal nitinol and catheter devices but substantially degraded by streak artifact in high-metal LVAD housings. This relationship was observed qualitatively in a small selected series and is offered as an initial observation rather than an established characteristic of the technique. We discuss current limitations and emerging directions such as photon-counting detector CT, metal artifact reduction, and artificial-intelligence-assisted post-processing that may extend 4D cine CR in this population. Full article

Ship detection over long distances is crucial for the visual perception of intelligent ships. AI techniques, particularly machine learning and deep learning, have achieved a series of breakthroughs in this field. However, due to the limited pixels of ships over long distances, such objects often suffer from weak feature representation and are susceptible to interference in complex environments. To address these challenges, this paper proposes an improved architecture named Ship-DiffDet, based on DiffusionDet. First, we redesign the backbone feature extraction network and propose IDC-Net, which utilizes inception depthwise convolution to enhance feature extraction efficiency while reducing computational complexity. Second, to tackle the difficulty of effectively extracting features from small objects, we design a Hybrid Pooling Attention-enhanced Feature Pyramid Network. By incorporating a hybrid pooling attention mechanism, it strengthens multi-scale feature fusion, thereby improving the performance of the detection heads. Furthermore, we introduce a multi-order gated aggregation mechanism into the dynamic detection head to optimize dynamic convolution and further compress the model’s parameter count. Experimental results demonstrate our method achieves an effective balance between detection accuracy and computational efficiency. On our custom-built small-object ship dataset, the proposed method improves AP₅₀ by 1.7% over the baseline while reducing the parameter and FLOPs counts by 48.8% and 22%, respectively. Full article

Background: Skin cancer encompasses a diverse range of malignancies and remains a significant global health challenge. Accurate machine-learning-assisted diagnosis can substantially improve patient outcomes through early detection and timely clinical intervention. Federated Learning (FL) enables privacy-preserving collaborative model training across multiple healthcare institutions while ensuring that sensitive patient data remain decentralized. However, deploying advanced architectures such as Vision Transformers (ViTs) in clinical environments is challenging due to the high computational demands of self-attention mechanisms. Methods: This work proposes FedSwin-LHTP, an efficient federated learning framework for skin lesion classification that integrates a Swin Transformer backbone with a Lightweight Hessian-Inspired Token Pruning (LHTP) mechanism. LHTP estimates token importance using a second-order Taylor approximation around converged local model parameters to identify less informative patch tokens, enabling the early pruning of redundant representations without explicitly computing the Hessian matrix. Furthermore, the framework incorporates the FedProx optimization objective to mitigate client drift under heterogeneous non-IID data distributions. The proposed framework is evaluated on the HAM10000 and ISIC datasets under realistic non-IID federated settings. Results: Experimental results demonstrate stable convergence, effective knowledge aggregation, and robust diagnostic discrimination across distributed clients. By adaptively pruning approximately 60% of Stage-1 tokens, the proposed framework substantially reduces the computational burden of local transformer processing while maintaining high multiclass classification performance, achieving an accuracy of up to 96.1% on the evaluated datasets. Conclusions: These results highlight the potential of FedSwin-LHTP as a practical, privacy-preserving, and resource-efficient solution for collaborative healthcare intelligence. Full article

Wireless sensing networks increasingly extend into obstacle-prone deployments, where physical blockage degrades reliability and open propagation exposes transmission activity. Intelligent reflecting surfaces (IRSs) establish programmable paths around obstacles while passive elements remain constrained by severe cascaded attenuation. To address the tradeoff between reliability and covertness, we propose an unmanned aerial vehicle (UAV) -assisted active-IRS architecture under probabilistic line-of-sight and non-line-of-sight propagation conditions that accounts for direct leakage from the transmitter to the warden together with residual jammer cancellation and always-on IRS circuit noise under a finite output power budget. Furthermore, bidirectional Kullback–Leibler analysis identifies the reverse divergence as the tighter restriction and converts the covertness requirement into conservative gain bounds under warden location uncertainty and relative phase uncertainty conditions between the direct and aggregate reflected fields. Subsequently, closed-form phase control for calibrated equal-gain elements and gain monotonicity reduce the joint design to an exhaustive search over the prescribed placement grid. The numerical results demonstrate a SINR advantage over passive reflection and single-element relaying across the evaluated settings. The finite-array and hardware analyses show that gain back-off enforces a prescribed covert-outage limit while direct leakage and residual self-interference remain explicitly controlled. Overall, the framework provides a transparent basis for reliable covert sensing through UAV-assisted active reflection. Full article