Search Results (623)

Wideband power amplifiers (PAs) operating with high-order modulation signals exhibit strong nonlinear distortion and dynamic memory effects, making real-time digital predistortion (DPD) increasingly challenging under strict computational constraints. This work proposes a functionally decoupled neural DPD architecture, termed EA-CCF-AttGRU, which explicitly separates instantaneous nonlinear feature representation from temporal memory compensation within a unified end-to-end framework. Instead of introducing envelope features, cross-channel fusion, and recurrent attention as isolated modules, the proposed architecture assigns different compensation functions to dedicated components: envelope-assisted augmentation and point-wise cross-channel fusion enhance instantaneous nonlinear representation, while attention-guided recurrent modeling captures dynamic memory effects. A global linear bypass further reduces the burden of nonlinear compensation by preserving the linear transformation. Experimental results under a 160 MHz 1024-ary quadrature amplitude modulation (1024-QAM) baseband excitation with a 10.38 dB peak-to-average power ratio (PAPR) demonstrate that the proposed method achieves an adjacent channel leakage ratio (ACLR) of −65.91 dBc, a normalized mean square error (NMSE) of −57.84 dB, and an error vector magnitude (EVM) of 0.07% with only 6009 trainable parameters. The proposed architecture achieves an effective complexity–performance trade-off for wideband DPD applications and provides potential for future hardware-oriented implementation and synthesis validation. Full article

Traffic sign recognition (TSR) sits on the critical path of advanced driver assistance and autonomous driving, yet deployed classifiers fail in two qualitatively distinct regimes that prior work has largely defended against in isolation worst case adversarial manipulation from imperceptible digital perturbations to physically realizable stickers, patches, and outline-conforming edge attacks and average case environmental degradation such as fog, glare, motion blur, fading, and occlusion. We observe that, despite their differing origins, both regimes leave the same observable signature on the over specified structure of a sign whose class is redundantly encoded by the silhouette, color scheme, and central pictogram: a spatially localized disagreement among otherwise independent cues, scored against a small, enumerable grammar of physically valid attribute tuples. Recasting robustness as detection of this signature rather than defense against any single threat, we propose SAFER-Sign, which integrates four components that each repair a documented failure mode of prior approaches: (i) class conditionally decorrelated shape, color, and glyph encoders that render over the specification genuine rather than nominal; (ii) evidential per channel uncertainty that lets a degraded cue abstain instead of voting confidently wrong; (iii) a soft, factorized, confidence gated sign grammar prior that rewards jointly consistent tuples without becoming a single attribute attack surface; and (iv) a jointly trained spatial reliability gate anchored to a parameter-free cross-channel disagreement signal, so it cannot be suppressed like a decoupled front end. Taken together, these components mean that a successful adaptive attack in our evaluated settings had to jointly address class evidence, cross channel consistency, grammar compatibility, reliability gating, and abstention. This raises the number of coupled attack objectives, but we emphasize that it does not guarantee that all three channels must be corrupted. Full article

Visible-infrared (RGB-IR) object detection leverages multimodal information to ensure reliable perception in complex environments. However, dynamic scenes pose significant challenges due to the frequent inconsistency between scene-level modality contributions and local spatial reliability. Furthermore, standard feature extraction progressively attenuates boundary-sensitive structural cues, and unified fusion strategies often fail to capture spatially varying cross-modal complementarity. To overcome these limitations, we propose a Boundary-Guided Dual-Perspective Cross-Modal Fusion Network (BDPNet) to explicitly preserve shallow geometric structures and decouple deep semantic fusion into macroscopic and microscopic perspectives. Specifically, a Geometric Boundary Enhancement Module (GBEM) embeds Sobel-based high-frequency priors into shallow dual-modal features via residual spatial modulation, preventing the loss of crucial localization cues during downsampling. In the deep semantic space, a Hybrid Dual-Perspective Adaptive Fusion Module (HDAM) employs an illumination-aware branch for global modality weighting and a spatial confidence-driven branch for local cross-modal rectification. A spatial gating mechanism then dynamically reconciles these macro-environmental and micro-signal features. Extensive experiments on M3FD, LLVIP, and DroneVehicle demonstrate the effectiveness of BDPNet. Compared with state-of-the-art methods, BDPNet improves mAP_50–95 by 0.8% and 1.0% on M3FD and LLVIP, respectively, and improves mAP₅₀ by 0.6% on DroneVehicle, while using substantially fewer parameters and lower computational cost. Full article

In static acrobatic gymnastics pyramids, maintaining stability in a multi-person system is critical, yet unconstrained floor baselines and base-to-top pairing mechanics remain poorly quantified. This study evaluated postural stability in shoulder-stand pyramids, examining the interplay of two specific top-athlete variations (lighter Pyramid T1 vs. heavier Pyramid T2) and base-athlete stances (parallel vs. tandem). Five elite base-athletes were monitored while supporting two top-athlete variations, a lighter (Pyramid T1) and a heavier (Pyramid T2) during a standard static shoulder-stand pyramid across parallel and tandem foot placement configurations (three trials in each Pyramid variation). Unconstrained free floor-standing baselines were also recorded for base-athletes and for tops prior to and following pyramid trials. The root mean square (RMS) of the resultant 3D free acceleration (Xsens MTw Awinda inertial sensors sampling at 100 Hz, Xsens MT Manager version 4.6.5 software) positioned at the bases’ and the tops’ shanks (right and left) was used to assess postural stability. A five-point median filter followed by a zero-phase, 2nd-order forward and reverse Butterworth low-pass filter (yielding an effective 4th-order response at 5 Hz and 10 Hz cutoffs) was applied to all signals (MATLAB R2025b). Stance-envelope dimensions were calculated from rectangular boundaries fitted to the base’s foot outlines. Non-parametric Spearman rank correlations ($\rho$) and parametric correlations ($r,R^2$) were used to test the interbase-consistency and base-to-top coupling. Two-way repeated measures ANOVAs (pyramid x stance configurations) were applied with primary analytical emphasis placed on descriptive effect sizes alongside exact $p$-values (SPSS v30, p < 0.05). The acrobatic tandem stance expanded the parallel stance-envelope area by $148.3\%$ and its width by $24.7\%$. When base-athletes transitioned from free standing to pyramids there was a substantial acceleration RMS increase (Pyramid T1: +24.3% to 83.6%, Pyramid T2: 47.2% to 166.2%). Supporting the heavier top-athlete (Pyramid T2) significantly increased the bases’ resultant acceleration RMS by $+37\%$ to $+48\%$ across both filter cutoff thresholds ($p<0.05$). Furthermore, top athletes exhibited differential behaviors: while Top 2 displayed lower acceleration RMS than Top 1, she experienced a greater acceleration surge when in Pyramid ($+238.4\% \mathrm{to} +266.2\%$). The bases’ and the tops’ acceleration profiles did not exhibit parallel responses, indicating decoupled rather than mirrored stability adjustments. Furthermore, when the acceleration RMS was normalized to stance-envelope dimensions, significant pyramid x stance interaction ($p<0.05$) was observed in the anteroposterior but not the mediolateral acceleration. Base-athlete stability varies significantly across top-athlete variations and stance geometries. Evaluating unconstrained floor baselines alongside spatial stance boundaries is essential for capturing structural loading dynamics in multi-person athletic tasks. Full article

Graph Neural Networks (GNNs) have become a foundational tool for e-commerce recommendation systems, yet they consistently fail in zero-shot cold-start scenarios where new products enter the market without prior interactions. In this paper, we diagnose this failure as a structural vulnerability rather than a simple data sparsity issue. We introduce the concept of asymmetric learning, demonstrating that in severely imbalanced bipartite graphs, minority-type nodes (products) become disproportionately reliant on topological signals. By evaluating this phenomenon alongside a relatively balanced control dataset, we confirm that this performance collapse is a byproduct of the data structure rather than architectural design. To mitigate this limitation, we propose Similarity-Driven History Augmentation (SHA), a data-centric approach that assigns synthetic interaction histories to cold-start products by matching them with semantically similar established donors. To prevent these synthetic signals from degrading the representations of established nodes, we further introduce a decoupled hybrid framework alongside an enhanced SHA strategy that selectively filters active customers. Comprehensive evaluations across multiple real-world e-commerce datasets, including DataCo and Amazon Gift Cards, confirm the effectiveness and stability of our approach across different GNN architectures. Full article

This study investigates the structural characteristics, annual trends, and qualitative rationales underlying modified audit opinions issued within the independent audit reports of Turkish banks between 2022 and 2025. Utilizing a formal content analysis of 248 bank-year independent audit reports based on a complete census approach, the study evaluates the incidence and persistence of modifications under ISA 705 and IAS 37. The empirical findings reveal a recurring pattern of qualified opinions associated with unmandated discretionary provisioning, with all qualified opinions in the sample issued by Big Four audit firms and recurring across multiple reporting periods. Interpreted through institutional decoupling and signaling perspectives, the findings are consistent with the possibility that these qualifications may reflect interactions between accounting requirements, regulatory conditions, and provisioning practices rather than providing direct evidence of financial distress or managerial intent. The study contributes to the auditing and banking literature by examining the accounting bases and longitudinal patterns underlying modified audit opinions and by discussing their implications within an institutional framework relevant to emerging economies. Full article

Post-activation performance enhancement (PAPE) is often attributed to elevated neural drive, yet neuromuscular and mechanical responses are rarely tracked concurrently. This study examined whether the two scale proportionally over the 1–4 min following a conditioning activity, with a maximal countermovement jump (CMJ) at each minute. Twenty-two male youth soccer players performed a baseline CMJ, a conditioning activity, then a CMJ at each minute to 4 min, with bilateral vastus lateralis and semitendinosus EMG (fixed 300 ms pre-take-off window) and jump height recorded simultaneously. Jump height varied across time, owing to a transient decrease at 1 min (−1.41 cm); from 2 min onward, it was equivalent to baseline (TOST, ±1.5 cm, all p ≤ 0.005). The composite EMG index instead rose almost exclusively at 4 min (d_z = 3.22), as did each channel individually, producing a signal × time interaction (F(3, 63) = 108.85, p < 0.001) and a decoupling index of +120.4%; the within-participant association between the two measures was weak (rrm = 0.11). Surface EMG amplitude can therefore rise sharply where mechanical output is equivalent to baseline within ±1.5 cm, and cannot be assumed to index a proportional mechanical gain. With no control condition, the observed time course reflects the conditioning activity together with the intervening maximal jumps. Full article

Plain river networks are weakly flushed and frequently regulated, which can decouple instantaneous water quality from biological responses. We integrated physicochemical variables, phytoplankton, zooplankton, benthic macroinvertebrates and fish environmental DNA (eDNA) to diagnose ecological condition in the Lixiahe plain river network, China. Regional water quality was assessed at 62 sites, and coupled analyses used 36 sites with complete multi-trophic data. Moderate pollution accounted for 62.90% of classified sites, with total nitrogen as the principal pressure. After Benjamini–Hochberg correction, seven independent associations with primary biological indicators remained significant, mainly linking TOC and TDS with planktonic metrics; the TOC-MEII correlation was treated separately as a non-independent composite association. Mean fish eDNA observed OTU richness declined to 22.33 at the three severely polluted matched sites, although grade-wise contrasts were descriptive because both extreme groups contained only three sites. Adjusted R² values for the four RDA models ranged from 9.8% to 11.4%; all overall fixed-seed permutation tests were significant (p = 0.0002–0.0298), indicating a detectable but limited water-quality signal. The study-specific MEII ranged from 29.50 to 67.42 (mean 48.04). These results show asynchronous trophic responses and support MEII as a preliminary within-survey screening tool that should be interpreted with its component sub-indices rather than as a calibrated reference-condition index. Full article

In recent years, the use of UAVs has become increasingly widespread, and the public safety risks posed by unauthorized UAV flights have become increasingly prominent, creating an urgent need for effective detection and identification of UAV targets. However, such targets are small in size, have low contrast, and exhibit an extremely low signal-to-noise ratio; conventional detection methods generally suffer from insufficient feature discrimination, missed detections, and false alarms in complex backgrounds. To address these challenges, this paper proposes a Frequency-Guided Cross-scale Refinement Network (FGCR-Net). Based on an encoder-decoder architecture, this network achieves end-to-end collaborative optimization through cross-layer feature fusion, side-channel prediction refinement, and frequency-domain background suppression. First, a multi-path selective cross-layer fusion module (SCFM) is designed. This module employs coordinated modeling via both channel and spatial paths, supplemented by adaptive weighting with learnable coefficients, to perform differentiated selective fusion of the encoder’s fine-grained features and the decoder’s semantic features, thereby bridging the semantic gap at jump connections; Second, we designed a Cross-Scale Adaptive Fusion Enhancement Attention Module (CAFEM), which cascades multi-receptive-field hollow convolutions, strip pooling, and a bidirectional semantic guidance mechanism to perform cross-scale refinement on the side outputs of each decoder layer, thereby alleviating the issues of blurred boundaries and false alarms caused by inconsistent quality of multi-scale prediction maps and insufficient cross-layer consistency; finally, we design a Frequency-Guided Semantic Enhancement Module (FGSEM), which uses the Fast Fourier Transform (FFT) to decouple encoder features into the frequency domain. By leveraging low-frequency energy to predict the background confidence map and applying spatially selective suppression to high-frequency components, this module distinguishes, from a frequency-domain perspective, the high-frequency responses of complex backgrounds and targets that are highly similar in the spatial domain. Experiments on MSDS-UAV, a self-built multi-scenario UAV dataset for small targets, demonstrate that our method consistently outperforms existing state-of-the-art methods across multiple performance metrics, with Pixel Accuracy, Mean Intersection over Union, and Probability of Detection reaching 92.76%, 70.91%, and 92.69%, respectively; Compared to the baseline model, these three metrics improved by 1.90, 3.20, and 3.76 percentage points, respectively, fully validating the effectiveness and superiority of the proposed method. Full article

Generalized category discovery (GCD) aims to classify known categories while discovering novel ones in unlabeled data, yet existing methods lack mechanisms to correct boundary-ambiguous samples that receive noisy pseudo-labels, as they primarily rely on visual feature learning without external semantic guidance. Vision-language models (VLMs) offer a natural source of cross-modal semantic correction. However, applying VLM-guided contrastive signals directly within the GCD training loop proves counterproductive because the locally-oriented InfoNCE loss conflicts geometrically with the globally oriented K-means objective in the shared backbone space. We identify the root cause as a dual resource allocation problem: the VLM-derived signal must be allocated to the correct feature subspace to avoid geometric conflict with K-means clustering (space allocation), and the limited VLM inference budget must be allocated to the correct samples to maximize discriminative return (budget allocation). These two decisions are coupled; failure on either renders the other ineffective. To resolve this, we propose DualGLEAN, a framework that addresses the dual allocation challenge through two coupled mechanisms: decoupled contrastive alignment (DCA), which routes the VLM-guided neighbor contrastive loss to a dedicated projector space while preserving the backbone space for global clustering, and compound uncertainty querying (CUQ), a three-stage filtering metric that jointly evaluates predictive entropy, boundary proximity, and local label inconsistency to direct VLM queries exclusively to truly boundary-critical samples. Extensive experiments on the AID and RSSDIVCS datasets demonstrate that DualGLEAN achieves strong performance, improves four diverse GCD baselines as a plug-in module, generalizes across seven VLM backbones, introduces zero additional trainable parameters to the base GCD network, and incurs a total VLM API cost of only CNY 2.45 per full training run on the AID dataset under the default search-scope configuration, with the cost scaling linearly with the query budget. Full article

Flexible and stretchable electronics inevitably undergo stretching, compression, bending and torsion when conformally attached to skin, soft tissues and dynamic organs. While deformation-induced electrical variations act as target signals for motion sensors, they cause resistance/impedance drift, baseline shift and sensitivity degradation in physiological electrodes, temperature/chemical sensors, interconnects and stimulation devices, leading to motion artifacts and reduced long-term reliability. Hydrogels are pivotal materials for soft bioelectronic interfaces owing to their high water content, low modulus, tissue compatibility and ionic conductivity. However, their conductive networks are susceptible to structural reconstruction under deformation, dehydration, swelling and cyclic fatigue, meaning that stretchability is by no means equivalent to strain insensitivity. This review focuses on stable resistance/impedance and functional output within a specified strain window, this paper reviews three representative material systems, liquid metal (LM)-based composite hydrogels, conductive polymer/elastic network composite hydrogels, and hydrogen-bonded isotropic architectures. It further summarizes three design strategies—geometric and functional compensation, mechanical decoupling and strain isolation, and interfacial engineering for conductive network stabilization—and discusses their applications in wearable epidermal and implantable bioelectronics. Finally, unified evaluation metrics for strain insensitivity are proposed, with future directions covering high-conductivity–low-modulus synergy, long-term water/ionic stability, robust soft-hard interfaces, multiaxial deformation tolerance and scalable manufacturability. Full article

To address severe transmission signal attenuation caused by inter-phase electromagnetic coupling in three-phase cross-bonded cables, which significantly limits the identification of insulation defects and fault types, this paper proposes a broadband impedance spectroscopy differential-frequency decoupling and defect diagnosis method based on three-phase high-frequency, high-voltage coordinated excitation. First, a decoupled transmission line model for three-phase cross-bonded cables is established to systematically analyze how the degree of insulation damage and fault equivalent impedance affect the frequency domain response at the test terminal. Three core diagnostic parameters, namely the fundamental resonant peak amplitude variation ratio, the resonant period change rate, and the initial phase offset, are extracted to construct a multi-dimensional collaborative diagnosis model for accurate defect and fault identification. Second, a high-voltage topology based on proportional optical isolation driving is designed to overcome the gain and bandwidth trade-off of conventional power amplifiers, effectively mitigating excessive high-frequency signal attenuation and low signal-to-noise ratios in long-distance transmission lines. Using this topology, a three-phase differential-frequency coordinated excitation test platform is developed to ensure the accurate, synchronous injection of isolated three-phase differential-frequency sinusoidal waveforms into the cable sending end. Combined with broadband impedance spectra acquired by the test system, the proposed diagnostic model enables the reliable identification of defect and fault types in three-phase cross-bonded cables, offering practical significance for the safe and stable operation of power transmission systems. Full article

Reinforced concrete structures are critical components of modern infrastructure, and the detection of internal damage within these structures has long been an important research topic. These damages can be detected using the vibration response of the structures, and laser Doppler vibrometry (LDV) on a moving platform offers an efficient, long-range sensing approach for structural vibration monitoring. However, extending LDV-based damage detection from static to mobile measurement requires addressing the effects of measurement signal frequency shift, platform vibrations, and speckle noise. To address these issues, this paper proposes a mobile measurement damage detection framework that integrates multi-source information with artificial intelligence algorithms. First, theoretical derivation and numerical simulation demonstrate that the vibration frequency shift induced by moving speed is negligible, proving that mobile and static measurement signals are similar in both time and frequency domains. Then, a multi-sensor data processing framework is used to decouple the platform vibration and suppress speckle noise. Finally, a spatial-aware CNN network is employed to achieve damage detection under mobile measurement. The results reveal that the vibration signals for large-scale voids were effectively recovered, whereas signals for small-scale voids and healthy regions were only partially recovered. Voids with a tested size of 0.4 m and larger were successfully identified under the experimental conditions. The results demonstrate the feasibility of extending static LDV-based void detection to mobile measurement, providing a theoretical and technical basis for efficient, non-contact mobile inspection of infrastructure. Full article

Small cargo unmanned aerial vehicles (UAVs) operating in narrow confined spaces suffer instantaneous collision impacts that induce gyroscope heteroscedastic noise and abrupt angular velocity derivatives, triggering severe dynamic attitude errors in traditional strapdown inertial navigation systems (SINS). Existing algorithms separate angular velocity fitting and noise suppression, adopt unified three-axis weighting, and lack adaptive segmentation for collision disturbances, limiting navigation accuracy without raising computational costs. This paper proposes an integrated heteroscedastic decoupling algorithm for UAV SINS under collision interference. Hermite orthogonal polynomials are utilized to fit non-stationary angular velocity with derivative matching constraints, an optimized single-pass CUSUM detector with steady-state residual compensation is proposed to identify collision-induced variance change points. An axis-differentiated weighting strategy is developed to suppress heteroscedastic noise. Recursive least-squares is adopted to lower online computation overhead. Multi-condition coning motion simulations show Hermite polynomials achieve the lowest attitude RMSE under steady flight; the improved CUSUM detector delivers shorter detection delay, fewer false alarms, and lighter computation than mainstream detection methods, and segmented differentiated weighting eliminates collision-induced noise distortion at the raw measurement stage. The proposed algorithm unifies signal fitting and noise correction with minimal computational overhead, effectively mitigating dynamic errors for lightweight airborne navigation hardware and offering a high-precision front-end preprocessing solution for cargo UAVs operating in cluttered obstacle environments. The proposed algorithm is positioned as a gyro-only front-end preprocessing module; accelerometer-related error compensation and full multi-sensor back-end integration are addressed in ongoing work. Full article

Conventional image acquisition requires an electrically powered image sensor to be placed directly behind the camera lens, constraining camera placement. To overcome this issue, we introduce M3-RGB as an incoherent-light fiber imaging system in which a multicore, multimode optical fiber passively relays lens images to a remotely located image sensor. Unlike conventional approaches, M3-RGB is designed to operate directly on incoherent light and requires no electrical power or active components at the sensing interface. Because propagation through the fiber yields spatially scrambled patterns, a neural network is used to reconstruct the original scene by exploiting the spatial locality preserved by the multicore structure. In a controlled optical bench setup, where a liquid crystal display monitor displays road-scene images, we construct a paired dataset of scrambled and ground-truth images and quantitatively evaluate reconstruction performance across different fiber core counts, fiber lengths, and calibration settings, utilizing the peak signal-to-noise ratio and structural similarity index measure as performance metrics. By decoupling imaging electronics from the sensing point, this passive remote image relay approach may expand sensor placement options for potential applications such as all-around perception for mobile robots and autonomous vehicles, surveillance, and inspection in confined spaces. Evaluations in real outdoor environments constitute future work. Full article