Search Results (2,538)

The per-switch condition monitoring of insulated-gate bipolar transistors (IGBTs) underpins condition-based inverter maintenance, yet existing approaches collapse the six-switch state space into one system-level label, leaving individual devices unresolved. This paper presents a Continuous Wavelet Transform Multi-Task Network (CWT-MTNet), a lightweight convolutional network that predicts the four-class aging state (Healthy, Mild, Moderate, or Severe) of all six switches from one CWT-based RGB image whose channels carry the scalograms of three symmetrical-component deviation signals. Five Depthwise Separable Convolution (DS-Conv) blocks and six classification heads give 172 K parameters in a 0.7 MB footprint—a 21-fold reduction over MobileNet-v2—within the on-chip memory of an embedded inverter controller, a footprint comparison rather than a demonstrated implementation. Trained from scratch on 15,625 samples spanning all $5^6$ state combinations, it attains 98.15 ± 0.33% accuracy and 95.81 ± 0.82% Mild recall over five seeds: within 0.84 percentage points of MobileNet-v2 in accuracy, statistically indistinguishable in Mild recall, at one twenty-first of the parameters. Scratch training outperforms ImageNet pretraining, indicating a domain mismatch with CWT scalograms. Multi-task gradient conflict costs ResNet-18 42.1 points of Mild recall under six-head operation; DS-Conv architectures change by at most 3.3. Grad-CAM attributes this to distributed time–frequency attention; Mild-to-Healthy confusion is the dominant safety risk. A representation ablation shows the CWT recovers 2.4 of the 4.0 points lost by discarding phase and makes per-switch accuracy twelve times more uniform, without being necessary here. Independent sensor noise at 0.5% of the phase RMS exceeds the aging signature sixfold and defeats every encoding examined, so the pipeline requires coherent averaging at the acquisition front end. Full article

CLIP-based anomaly detectors have markedly advanced training-free and zero-shot industrial anomaly detection and localization, yet their predictions remain dominated by patch-wise vision–language similarity or anomaly-aware feature scoring. This formulation is intrinsically limited for logical anomalies, in which every visible part can appear locally normal while its count, position, arrangement, or co-occurrence violates a normal configuration. We introduce a training-free relation prototype re-scoring module that reuses the semantic–spatial relations already encoded by the visual transformer. Because patch tokens contain positional embeddings, the self-attention graph is position-aware as well as content-dependent; we use it as a message-passing operator over detector-specific patch features. Normal images define category-wise, spatially indexed relation prototypes, and each test patch is scored by the Euclidean deviation of its attention-aggregated feature from the corresponding normal prototype. The resulting map supports both dense localization and map-derived image detection. Although we also report the feature-relation map alone to analyze its intrinsic behavior, the final method fuses this map with the original anomaly map so that relation-sensitive evidence is added without discarding the baseline detector’s local appearance cues. We develop the method on AnomalyCLIP and verify its generality on WinCLIP and AA-CLIP. On the logical split of MVTec LOCO AD, the proposed fusion improves AnomalyCLIP from 53.9 to 74.4 pixel AUROC and from 28.7 to 52.7 pixel AUPRO, while image AUROC rises from 50.5 to 71.7. Similar improvements are observed for WinCLIP and AA-CLIP. On CAD-SD, the proposed fusion reaches 92.0 and 93.8 image AUROC for AnomalyCLIP and WinCLIP, respectively, on co-occurrence anomalies. Experiments on the structural split of MVTec LOCO AD and the MVTec AD benchmark show why the baseline map must be retained: fusion preserves substantially more local-defect evidence than relation-only scoring, although its benefit remains detector- and category-dependent. These results identify semantic–spatial relation deviation as a missing cue in CLIP-based logical anomaly detection and localization, without requiring explicit component models, symbolic rules, additional training, or modification of the baseline architecture. Full article

Adversarial patch attacks pose a tangible physical-world threat to traffic sign recognition in autonomous driving systems. Current state-of-the-art defenses based on image reconstruction require dual-model deployment and add per-frame inference latency, making them impractical for resource-constrained embedded platforms such as mass-produced ADAS systems and aftermarket dashcams. This paper proposes a two-component defense that separates detection from mitigation. PatchGuard-Freq leverages frequency-domain analysis to detect attacked images with high accuracy, while adversarial fine-tuning enables the detector to recover most of its detection performance under attack without adding any inference module or architectural change. Experimental results show that the approach is effective in the latency regime targeted by this study: on the evaluated stop-sign and nine-class LISA benchmarks, under the evaluated detection placement (a fresh $110\times 110$ patch at a uniformly random position), the detector achieves AUC = 1.0 with zero false positives across all four dataset configurations and all combined frequency variants, while the frequency-only arm of the ablation separates the data only where the patch covers almost the entire input, and adversarial fine-tuning recovers attacked mAP@0.5 on the COCO stop-sign test set from 19.4% to 68.4% on the $110\times 110$-patch configuration while degrading clean mAP by only about 0.7 points. We further show that PatchGuard-Freq is vulnerable to adaptive (defense-aware) attacks and introduce defense-aware training that restores detection of such attacks to 100% without increasing false alarms. This work characterizes the complementary relationship between reconstruction-based and robustness-based paradigms in the accuracy–efficiency design space under the evaluated conditions: the former suits compute-unconstrained scenarios while the latter serves latency-constrained deployments. All quantitative results were obtained on desktop-grade GPUs; embedded-platform latency, memory, and energy were not measured, and the embedded discussion is limited to hardware-independent parameter and FLOP counts. Full article

Background/Objectives: Alexithymia may constrain emotion regulation, but its relevance may differ across personality configurations. This study examined whether narcissistic Admiration and Rivalry are embedded in distinct early maladaptive schema and alexithymia profiles. Methods: A non-clinical adult sample (N = 311) completed the Narcissistic Admiration and Rivalry Questionnaire, the Young Schema Questionnaire–Short Form 3, and the Toronto Alexithymia Scale–20. Analyses examined 18 schemas and three alexithymia components—Difficulty Identifying Feelings (DIF), Difficulty Describing Feelings (DDF), and Externally Oriented Thinking (EOT)—using zero-order correlations, within-domain regressions, hierarchical regressions, and structural equation models. Results: Admiration was associated mainly with Approval/Recognition Seeking and Unrelenting Standards and with lower Social Isolation/Alienation, Defectiveness/Shame, Failure, Subjugation, Emotional Inhibition, and Insufficient Self-Control. Rivalry showed a threat-related profile involving Defectiveness/Shame, Subjugation, Emotional Inhibition, Negativity/Pessimism, and Insufficient Self-Control. Entitlement/Grandiosity was positively associated with both dimensions and did not reliably differentiate them. Alexithymia was essentially unrelated to Admiration and did not improve its schema model. In contrast, adding DIF, DDF, and EOT increased explained variance in Rivalry by 5.5%, a small-to-moderate increment; DIF and EOT made independent contributions, whereas DDF did not. A parsimonious Rivalry structural model reproduced these associations but showed mixed global fit. Conclusions: These cross-sectional findings indicate a modest, dimension-specific contribution of alexithymic processing to Rivalry rather than a general association with narcissistic self-regulation. The EOT findings require particular caution because of the subscale’s limited reliability. Longitudinal and experimental studies are needed to test the proposed vulnerability–stress process. Full article

Integrating Wire Arc Additive Manufacturing (WAAM) into the Selective Paste Intrusion (SPI) process enables the fully additive fabrication of reinforced concrete structures with complex geometries. Previous investigations have demonstrated that the thermal impact of the WAAM process can adversely affect the SPI process. Thus, dedicated cooling strategies are required. One proposed approach increases the vertical distance between the welding point and the particle bed by introducing a defined vertical protrusion of the reinforcement bar. This configuration may give rise to shadowing effects, here understood as a process-induced disturbance of material deposition in the vicinity of the protruding bar. Two distinct manifestations are considered in parallel. The first is a geometrically projected shadowed region within the particle bed, depending on bar diameter and inclination. The second is a layer-wise modification of the contact zone along the lower half of the bar surface within the bond length, largely independent of inclination. To isolate the geometric component from thermal effects, the present study focuses on controlled reinforcement configurations with constant vertical protrusion. Two hypotheses are tested: bond decreases with increasing bar diameter (H1), and, at constant diameter, with decreasing inclination angle (H2), the latter being the signature of the projected shadow. To assess these effects, reinforcement bars with a constant vertical protrusion of 40 mm and varying inclination angles were embedded into the particle bed, and concrete specimens were produced above them using the SPI process. Bar diameters of 8 $\mathrm{m}$$\mathrm{m}$, 16 $\mathrm{m}$$\mathrm{m}$, and 25 $\mathrm{m}$$\mathrm{m}$ and inclination angles from 0° to 90° in 15° increments were investigated systematically. Bond strength was determined using push-through tests derived from RILEM RC6, and the bond response was evaluated against both a quantitative measure of the projected shadowed area and a process-based indicator of the affected contact zone. The bar diameter dominates the bond response, most pronounced at the developed-interlock and capacity levels. The inclination angle produces no monotonic trend from 0° to 90°, and individual angle contrasts remain largely within the experimental scatter. The projected shadowed area cannot consistently explain the observed behaviour and is at most a secondary factor, whereas the layer-wise contact-zone disturbance along the lower bar surface is the most probable interpretation of the data. The findings identify shadowing as a boundary condition for reinforcement integration in SPI: the observed bond reduction at the developed-interlock and capacity levels is attributed to the layer-wise contact-zone disturbance rather than to the projected shadowed area, an attribution that remains a hypothesis until the contact zone has been verified directly. Full article

Background: Metamorphic malware is among the most persistent adversarial challenges in cybersecurity: it rewrites its own instruction stream on every propagation, preserving functional semantics while presenting a syntactically distinct binary that defeats signature-based and many learning-based detectors. Methods: We propose MetaGNN-Sec, a graph-augmented neural framework that detects metamorphic malware from program structure rather than surface bytes. The framework composes four components, each addressing a distinct facet of the problem: (i) control-flow graph (CFG) extraction with semantic opcode embeddings; (ii) a heterogeneous graph neural network (hGNN) operating over program-dependence graphs that capture mutation-stable control- and data-flow invariants; (iii) an adversarial training loop derived from the Wasserstein generative adversarial network (WGAN) that hardens the classifier against adaptive evasion mutations; and (iv) a quantum-kernel anomaly layer implemented in PennyLane for separation of heavily obfuscated outliers in a high-dimensional feature space. Results: Experiments are conducted on two public corpora—VirusShare 2024 and a SOREL-20M subset—comprising 200,175 binary samples in total (155,175 malware and 45,000 benign), in agreement with the corpus totals reported in Datasets Section of this paper. MetaGNN-Sec achieves a detection rate of 97.83%, a false-positive rate of 0.41%, and an $F_1$ score of 0.978 on held-out metamorphic families, exceeding the next-best baseline (MalConv+) by 4.6 percentage points on clean data and degrading by only 5.4 points under adaptive adversarial evasion (versus 17–31 points for the baselines). The quantum-kernel module contributes a further 1.2 pp reduction in false-negative rate, concentrated on the most heavily mutated families. Conclusions: The framework provides a heterogeneous PDG representation with a conditional score-shift bound under graph-edit-bounded mutations, a WGAN hardening loop that delivers measurable adversarial robustness, a quantum-kernel pre-filter with an explicit cost/benefit characterization, and a reproducible, near-real-time pipeline suitable for enterprise endpoint deployment. Full article

Log anomaly detection requires models that capture long-range event dependencies without the quadratic sequence-length cost of self-attention. We present LogMamba, a reconstruction-based model that combines a bidirectional selective state-space branch with a Multi-Scale Frequency Learner (MSFL). The sequence branch models ordered semantic dependencies, whereas the MSFL applies a Fast Fourier Transform to the sequence-position axis and processes multiple spectral bands with independent multilayer perceptrons. An adaptive gate integrates both representations before semantic-embedding reconstruction. Because the position-frequency branch uses FFT operations, the complete block has $\mathcal{O}\left(L\log L\right)$ complexity with respect to a sequence of length L. On HDFS, BGL, and Thunderbird, LogMamba obtains mean F1-scores of 0.837 ± 0.006, 0.987 ± 0.002, and 0.983 ± 0.003, respectively. The HDFS score is 6.3% higher than the strongest baseline reported in our comparison, while recall reaches 0.996 on BGL and 0.998 on Thunderbird. Component ablations indicate that the sequence and position-frequency branches contribute differently across datasets, supporting their complementary use for log-sequence reconstruction. Full article

The soil-biosphere nexus is a critical component embedded in the concepts of soil security and ecosystem services (ES) and is directly linked to several global challenges identified by the United Nations (UN). Although soil security has been proposed as a policy framework, a major problem is that there are no standard procedures to assess the five dimensions of soil security: capability, condition, capital, connectivity, and codification in relation to the soil-biosphere nexus. This study proposes a land cover change matrix, disaggregated by soil type (“biophysical soil security matrix”), as a tool to evaluate the biophysical soil security continuum and its temporary changes, integrated with ES valuation. The matrix was tested using the state of Minnesota (MN) as a case study. Although the dominant soil orders in MN possess high natural capability, widespread human-caused land degradation has dramatically lowered their actual biophysical condition and has fueled massive soil decarbonization. Historic land degradation in MN due to human activity totaled 98,516 km² through 2024, with nearly 480 km² of anthropogenically degraded land created between 2001 and 2024. Based on carbon emissions alone, we estimate that historic land degradation in MN has resulted in total social costs of nearly $50B (U.S. dollars, B = billion = 10⁹) through 2024, with about 10% of this total social cost being realized between 2001 and 2024. Trends like this threaten soil security by directly eroding the soil’s capacity to sustain ES. Based on the analyses and results, this study recommends conducting both soil-centric and human-centric analyses of soil security to ensure the sustainable use of soil. Full article

Human tracking plays a crucial role in video surveillance systems. However, tracking humans in surveillance videos remains challenging because targets are often captured at long distances, occupy only a small number of pixels, and exhibit substantial scale variations. These challenges require not only accurate detection of small, low-texture targets but also fast and robust data association for multi-object tracking. We propose an enhanced human-tracking method that integrates improved object detection with complementary appearance and motion cues. Specifically, we improve YOLOv12 by incorporating large-kernel deformable attention, dynamic convolution, and phantom convolution. These components enhance the detector’s ability to perceive small-target features under complex backgrounds and occlusion while reducing its computational cost. The OSNet appearance embeddings incorporated into the EMA update framework construct a robust trajectory-level temporal appearance representation, which improves the discrimination between different individuals in the tracking stage. Extensive experiments demonstrate that the proposed method achieves more accurate identity association and more stable target trajectories than state-of-the-art tracking methods, including StrongSORT and ByteTrack. Full article

Lithium-ion batteries have become essential for electric vehicles and energy storage systems; however, safety risks related to thermal runaway have emerged as a critical concern. Conventional monitoring methods based on voltage, current, and temperature cannot detect internal faults at an early stage. To address this limitation, proactive diagnostic methods employing electrochemical impedance spectroscopy (EIS) have been extensively investigated. This study presents an excitation-current generation method for EIS measurement based on a variable-inductance bidirectional Ćuk converter. The proposed circuit operates using the energy stored in the battery system, eliminating the need for an external auxiliary power source for excitation energy, while generating sinusoidal excitation currents over a wide frequency range. A small-signal analysis incorporating a first-order battery equivalent circuit model and the parasitic elements of the Ćuk converter was conducted to evaluate system stability and the effects of key design parameters. The experimental results confirm sinusoidal excitation-current generation over the frequency range of 0.1 Hz to 3 kHz. Under a representative test condition, the prototype generates a sinusoidal excitation current with a 2 A peak-to-peak AC component superimposed on a 3 A DC component. These findings verify the feasibility of the proposed method as a dedicated excitation-current generator for embedded EIS measurements. Full article

This paper raises a methodological question: How can we assess machine-made Old English when there is no parallel reference text and the standard metrics do not fit the task? We propose a pipeline with five measuring layers plus two compliance components, including lexical attestation with form linking, frequency-profile diagnostics, character-level comparison, word-embedding geometry under a verified mapping and dependency parsing, with bootstrap confidence intervals around the main contrasts. We apply the pipeline to a machine-made version of Gregory’s Dialogues: 4217 sentences, one for each sentence of the Old English original, generated under hard constraints. The unattested residue is two word types and 0.003% of tokens. The frequency profile diverges from the original by 0.008, less than the original diverges from the background corpus. At character level, in embedding space and in parsed syntax, the generated text stands at the same distance from the Dictionary of Old English Corpus as the original itself does. We propose an overall metric G, the geometric mean of seven bounded components, which scores the text at 0.991 with a confidence interval of [0.991, 0.992]. Two blind detection experiments with expert judges place the index externally: roughly two-thirds of generated sentences pass as authentic to specialists, so the divergence the pipeline measures is real at corpus scale but not available to sentence-by-sentence reading. The main contribution is a reusable evaluation method for historical language generation, together with a single interpretable score that subsumes the partial metrics without hiding them. Full article

Artificial intelligence (AI) is increasingly embedded in organizational decision-making, requiring employees not only to use AI-generated recommendations but also to evaluate their quality and determine when reliance is appropriate. Although established research examines behavioral reliance on algorithmic and AI advice, fewer studies have approached performance under AI advice as an individual-differences assessment problem integrating psychometric structure, cognitive correlates, process indicators, and criterion-related evidence. This study developed and initially validated a CHC-informed, performance-based assessment of cognitive performance under AI advice using 24 organizational decision scenarios. The assessment was designed around three closely related content/performance dimensions—AI error detection, evidence integration, and cognitive control and adaptive reliance—while also capturing confidence, response time, and reliance behavior. The validation sample comprised 780 employed adults in Türkiye. Psychometric analyses included confirmatory factor analysis, multidimensional item response theory, response-time analyses, scenario-level logistic regression, measurement invariance, differential item functioning, and internal cross-validation. Results indicated a dominant general cognitive-performance component together with additional structure corresponding to the three theoretically specified dimensions. Assessment performance was positively associated with established cognitive measures, including ICAR-16 reasoning performance, working memory, processing speed, and attentional control, whereas associations with AI-related self-reports were generally weaker. Dimension-aligned analyses supported the expected associations of ICAR-16 with AI error detection and working memory with evidence integration. Performance was also moderately associated with concurrently assessed organizational decision quality (r = 0.408) and explained additional variance in this criterion beyond demographic and work characteristics, AI experience, conventional cognitive-performance measures, and AI-related self-reports (ΔR² = 0.103, p < .001). In contrast, several hypothesized scenario-specific associations involving AI confidence, time pressure, interruptions, and resistance to confidently inaccurate advice were not supported. Overall, the findings provide initial evidence for a performance-based approach to assessing how employees evaluate and respond to AI advice, while indicating that the proposed scenario-specific mechanisms and group-comparability findings require further replication before consequential applications are considered. Full article

Multi-label node classification requires capturing complex label-co-occurrence patterns, which are critical for accurate predictions. However, traditional methods often treat label prediction as an independent task, overlooking the semantic relationships and dependencies between labels. This limitation hinders the ability to model intricate label-co-occurrence information. To address this issue, we propose the Mutual Optimization of Label-Prediction and Node-Representations (MOLN) framework. MOLN integrates two interrelated components: Semantic-Aware Contrastive Learning (SACL) and the Label-Interaction-Enhancement Network (LIE). SACL operates in the node-representation space, aligning embeddings with label-co-occurrence patterns to capture meaningful semantic relationships. LIE, on the other hand, works in the label-prediction space, explicitly modeling mutual dependencies among labels to refine predictions. The natural synergy between these components ensures mutual optimization: SACL enriches node features for prediction, while LIE improves predictions to guide better semantic alignment in SACL. By bridging the representation and prediction spaces, MOLN effectively addresses the label-co-occurrence problem, achieving state-of-the-art performance on benchmark datasets with significant improvements. Full article

Conversion flight turns the aerodynamic versatility of a tiltrotor into a demanding propulsion-control problem. As the nacelles rotate and vertical load transfers from the proprotors to the wing, the two turboshaft engines must follow a rapidly changing shaft-power demand while respecting fuel command magnitude and rate limits, compressor-pressure limits and turbine temperature limits. A control-oriented conversion model is coupled to a component-level turboshaft engine through a long short-term memory (LSTM) dynamic surrogate embedded in a constrained receding-horizon controller. The aircraft model resolves wing force balance, blade-element/momentum rotor loads, forward acceleration, nacelle actuation and accessory power. The LSTM predicts six engine outputs from flight conditions, fuel command and previous-step spool speeds. Training and evaluation use 537 converged component model cases divided by complete simulation cases into 375 training, 80 validation and 82 held-out test cases. Against parameter-matched multilayer perceptron, temporal convolutional network and gated recurrent unit baselines, the LSTM gives the lowest power root-mean-square error (20.15 kW before online output correction). Its corrected 20–320-step forecasts outperform a linear autoregressive model and zero-order hold prediction, although the linear model remains slightly better at one step. In direct component-level closed-loop simulation, LSTM engine-surrogate nonlinear model predictive control (NMPC) reduces power RMSE from 29.80 to 16.23 kW relative to linear MPC and from 34.94 to 16.23 kW relative to PI control while reducing cumulative fuel command variation by 46.5% relative to linear MPC. Turbine temperature and compressor-pressure margins remain 119.0 K and 85.4 kPa, respectively. Mean optimization time is 29.6 ms for a 1.92 s update interval. The resulting framework connects conversion flight aerodynamic loading, multi-step engine prediction and constrained power control in a reproducible numerical validation chain. Full article

Why does the same host economy see foreign investors co-build vocational training institutions in some industries but not others, even when FDI volumes are comparable? We argue that part of the answer lies in the governance mode of the global value chain (GVC) an investor is embedded in, alongside the liability-of-foreignness logic that dominates the co-creation literature. Relational and captive governance (high transaction complexity, low codifiability of required capabilities, and supplier competence that cannot be bought off the shelf) create a mutual dependence that can make joint institutional investment rational for both firms and the host state, whereas modular and market governance do not. Morocco’s aerospace and automotive value chains sit at the relational/captive end of this spectrum; its textile and agro-processing value chains sit closer to modular/market governance. We examine the argument in two stages. First, using a national-level 2SLS/DOLS/FMOLS estimation on 44 annual observations (1977–2020), we find FDI inflows positively associated with secondary-school enrollment nationally (β = 7.18 USD billions, p < 0.001 under 2SLS, corroborated by DOLS but not by FMOLS), though the supporting evidence is not uniform across estimators and diagnostic tests, and a national aggregate cannot, by itself, explain sector-by-sector variation. Second, we contrast the aerospace/automotive and textile/agro-processing value chains directly: the relational/captive chains show co-designed curricula, co-funded institutes, and co-governed placement systems, while the modular/market chains show comparable FDI intensity but no comparable institutional response. This sectoral contrast, documented in greater depth for aerospace and automotive than for the comparison sectors, is consistent with GVC governance mode, rather than FDI volume or liability of foreignness alone, playing a role in whether institutional co-creation occurs, though the evidence here is suggestive rather than conclusive. We report the national-level estimation transparently, including a set of diagnostic limitations (cointegration-rank and integration-order ambiguity across Johansen, Gregory–Hansen, and ARDL bounds tests; an instrument-validity caveat that persists even after removing individual instruments; a digital-infrastructure composite missing its fixed-broadband component; and an estimator-sensitive FDI coefficient that DOLS corroborates but FMOLS does not), which qualify the macro evidence and should be read alongside, rather than in place of, the sectoral comparison. Full article