Scene text image super-resolution must preserve character identity and glyph geometry, including approximate symmetry where present and asymmetric details that distinguish characters. Erroneous optical character recognition (OCR) priors can distort characters during generative reconstruction. We combine constrained large language model (LLM) correction and Mamba-based visual sequence modeling in a conditional residual diffusion method. The LLM proposes replacements within OCR candidate and edit constraints; deterministic validation rejects invalid suggestions and retains the original transcription. Corrected text, position, and low-resolution structural priors are fused by sample-level reliability gating and injected through cross-attention. A bottleneck Mamba block models horizontal visual dependencies. Under separately trained TextZoom and Real-CE protocols, the method achieves peak signal-to-noise ratios of 25.06 and 24.38 dB and exact-match accuracies of 83.4% and 50.2%, respectively. Five-seed paired ablations show that constrained OCR correction improves full-model reconstruction by 0.07 dB in PSNR and 0.6 percentage points in exact-match accuracy on both TextZoom and Real-CE, with the corresponding 95% confidence intervals excluding zero. Compared with the tested lightweight Transformer, Mamba reduces reconstruction latency by 7.82 ms with comparable quality. Independent recognition evaluation supports improved readability. The results support constrained correction and complementary visual conditioning.
During symmetrical faults in photovoltaic (PV)-integrated distribution networks, induction motor (IM) deceleration elevates reactive-power demand and delays post-fault voltage recovery, whereas conventional voltage-driven support can withdraw prematurely upon initial voltage rebound. To resolve this mismatch, this paper proposes a coordinated transient voltage support (CTVS) strategy that explicitly incorporates IM electromechanical dynamics. A coupled differential-algebraic equation (DAE) model integrating third-order IM dynamics, distributed PV (DPV) DC-link dynamics, and network power balance is formulated, from which a trajectory-local closed-loop Q–V sensitivity is derived to characterize location-dependent support effectiveness. The CTVS controller coordinates critical-bus voltage magnitude, voltage-decrease rate, and rotor slip through fixed spatial participation coefficients under local converter current limits and DC-link voltage feedback. Case studies on a modified IEEE 33-bus feeder under Bus-32 three-phase faults demonstrate that, for the representative 140 ms fault, CTVS suppresses post-fault secondary voltage depression and reduces the sustained recovery delay to 23.85 ms, compared with 342.30 ms for local low-voltage ride-through and 390.85 ms for adaptive Q–V droop control. Matched ablations identify slip-conditioned support persistence as the principal contributor to sustained voltage recovery, whereas the fixed spatial coefficients provide a local allocation advantage under unsaturated conditions rather than a separate large-disturbance recovery benefit under the tested conditions.
Video anomaly detection is critical for information security in IoT-based intelligent surveillance, where heterogeneous sensors continuously measure the physical environment and anomalies manifest as deviations from learned normal behavior. Existing methods, however, still struggle to jointly model long-range spatio-temporal dependencies and to fuse heterogeneous multimodal measurements. Deploying such detectors on resource-constrained edge devices remains a further practical barrier. To address these challenges, we propose E-STG-DiffNet, an edge–cloud collaborative multimodal spatio-temporal graph diffusion network that formulates secure IoT surveillance as a network measurement problem. The system employs a Video Swin Transformer to extract multi-scale visual features with linear complexity. A dynamic adaptive spatio-temporal graph convolutional network (ST-GCN) with trainable node embeddings then fuses the heterogeneous sensor measurements in non-Euclidean space, and a masked latent diffusion model (MLDM) combines masked context reconstruction with stochastic denoising for anomaly scoring. A confidence-aware edge–cloud collaborative inference mechanism, supported by knowledge distillation, enables low-latency deployment. On UCSD Ped2, CUHK Avenue, and ShanghaiTech, E-STG-DiffNet reaches frame-level micro-AUC scores of 98.1%, 90.4%, and 78.3%, respectively, competitive with the strongest recent methods on these benchmarks. All results are averaged over three random seeds. Ablation studies confirm that each core component contributes to the overall performance. In the efficiency analysis, the system sustains 26.3 FPS with a mean end-to-end latency of 43 ms (99th percentile: 88 ms), while only about 10% of frames are offloaded to the cloud as compressed latents, about 0.03% of the raw-video uplink traffic. Because the public benchmarks used in our experiments do not contain real non-visual sensor streams, the temperature, vibration, and acoustic modalities are simulated following an established protocol. The reported results therefore primarily validate the video anomaly detection capability of the framework, and validation on real multimodal sensor networks is left to future work.
In this study, the mechanical behavior of continuous-fiber-reinforced rubber composites is investigated experimentally and numerically. Uniaxial tensile tests were performed on the rubber matrix, individual reinforcement fibers, and composite laminates with different fiber orientations and layer configurations. The experimental results were used to calibrate and validate a finite element model developed in ANSYS MAPDL (release 2026 R1), in which the rubber matrix was described using an incompressible Mooney–Rivlin hyperelastic formulation, while the reinforcement was represented by embedded discrete elements with a linear-elastic or a nonlinear elastic cord model. The composite simulations capture the main influence of fiber orientation and laminate structure on the global stress–strain response. The 0∘/90∘ configurations exhibit higher stiffness because fibers aligned with the loading direction transfer load directly, whereas the 45∘/−45∘ configurations show a more compliant response governed by matrix deformation and fiber reorientation. At higher deformation levels, the measured softening was not captured by the model. These differences may be attributed to damage mechanisms such as matrix degradation and fiber–matrix debonding, which are not included in the current formulation. The proposed modeling approach provides a basis for assessing the global mechanical response of such composites and for future extensions that include interfacial damage and fatigue.
Advanced persistent threats are complex, stealthy, and long-lasting attacks that can seriously compromise networked systems. System logs provide critical information for APT detection, and provenance graph analysis is commonly used for log anomaly detection. However, existing graph-based methods build provenance graphs from a single perspective, lacking adequate integration of structural and semantic information among entities. To address this issue, we propose a log anomaly detection method based on dual graph modeling and mutual knowledge distillation. The structural branch captures event interactions among entities, while the semantic branch captures associations implied by node text attributes. Relational graph convolution is applied for representation learning in both views. During training, a mutual knowledge distillation framework enables the two branches to serve as teachers and students for each other, facilitating bidirectional knowledge transfer. A reliability estimator based on noise-perturbation invariance assigns higher weights to samples with more trustworthy teacher outputs. Experiments on three public datasets demonstrate that the proposed method achieves statistically significant improvements in Recall on Wget and Cadets, and in Precision on StreamSpot, while maintaining F1-scores comparable to MAGIC across several datasets. Threshold-independent ROC-AUC evaluation further shows that the observed advantages on Wget and Cadets are also reflected in anomaly ranking performance rather than being solely attributable to threshold selection. Ablation experiments validate the contributions of dual-graph modeling, mutual distillation, and the reliability estimation mechanism.
