Search Results (1,631)

In past decades, the macroeconomic stability of India has been tested repeatedly by major global disruptions, including oil price shocks, the 2008 global financial crisis and the COVID-19 pandemic. Analysing how macroeconomic variables respond to these shocks is essential for evaluating external vulnerability and policy resilience in emerging economies. Our study provides a comprehensive empirical investigation of the dynamic responses of wholesale price inflation, industrial output, oil prices and exchange rates in India by employing monthly data from January 1993 to December 2024. To examine long-run equilibrium relationships along with short-run adjustment dynamics, the present study employs co-integration analysis within a Vector Error Correction Model (VECM) framework. Further, we applied impulse response functions and forecast error variance decomposition to track volatility spillover mechanisms. Quantile regression and ARCH–GARCH models were further estimated to account for distributional heterogeneity and time-varying volatility. The findings of our study suggested stable long-run linkages among the selected variables, where oil price shocks emerged as a key external source of macroeconomic fluctuations. Short-run dynamics suggested that shocks in oil prices are transmitted primarily through inflation and exchange rate channels and then affect industrial output. Distributional estimates revealed the effects were stronger during stress periods, indicating tail risks that were not captured by the mean-based models. Lastly, volatility analysis confirmed persistent clustering, especially during phases of crisis. Overall, the findings suggest that India’s macroeconomic system remains externally sensitive, with adjustment mechanisms that operate gradually but come under strain during global disruptions. These results underscore the importance of energy risk management and crisis-responsive macroeconomic stabilisation policies. Full article

High dynamic range (HDR) image reconstruction from a single low dynamic range (LDR) input remains an important problem for computational photography, particularly when practical deployment on consumer-grade hardware is considered. With the increasing availability of hardware supporting HDR, public demand for capturing and viewing HDR images has grown significantly. Recent research has explored deep learning-based approaches to reconstruct HDR images from low dynamic range (LDR) inputs by extracting regional pixel features or leveraging the camera response function (CRF) for model training. Many of these approaches employ Convolutional Neural Network (CNN) architectures and utilize skip connections to preserve learned information. Nevertheless, the configuration-level effects of data augmentation in HDR reconstruction remain insufficiently discussed. Existing CNN-based approaches, such as HDRCNN, HDRUNet, and ExpandNet, have demonstrated promising reconstruction ability, but they may involve a heavy backbone architecture, a long training time, or a limited discussion of how preprocessing configurations affect reconstruction performance. This study presents an engineering-oriented HDR reconstruction framework derived from HDRCNN, focusing on practical efficiency, structural fidelity, and training feasibility. The proposed framework introduces three modifications: (1) a configuration-level comparison of composite data augmentation settings, including unsharp masking, denoising, Gaussian blur, and brightness–contrast adjustment; (2) the replacement of the original VGG16 backbone with a ResNet50-based encoder enhanced with attention blocks and squeeze-and-excitation (SE) blocks for improved multi-scale feature extraction and channel-wise recalibration; and (3) the integration of mixed-precision training with cosine annealing learning-rate scheduling to reduce computational cost. Experimental results on the SI-HDR dataset show that the best composite augmentation configuration improves PSNR from 19.05 dB to 22.10 dB and SSIM from 0.6444 to 0.7714 without increasing the training time. Compared with the original VGG16-based HDRCNN setting, the ResNet50-based model reduces training time while improving SSIM from 0.2705 to 0.8512. Under the adopted comparison protocol, the proposed model achieves the shortest training time and slightly higher PSNR than HDRUNet, while HDRUNet retains a higher SSIM. This indicates a trade-off among pixel-wise fidelity, structural similarity, and computational efficiency. The current evaluation is limited by a small test setting, composite rather than operation-level augmentation analysis, and the use of PSNR and SSIM only; therefore, future work should include full benchmark evaluation, additional perceptual/HDR-specific metrics, and controlled component-level ablation studies. Full article

Blind image watermarking models such as HiDDeN have laid an important foundation for end-to-end watermarking. Nevertheless, they still suffer from three major limitations: single-scale feature extraction, fixed fusion weights, and slow training convergence. To address these issues, this paper proposes an adaptive multi-scale watermarking algorithm based on collaborative dual-attention mechanisms. The algorithm designs an adaptive multi-scale feature fusion module (MA-FFM) with a dynamic gating network in the encoder, which flexibly combines local multi-scale textures with global contextual information, overcoming the limitation of fixed fusion weights. In the decoder, a multi-level channel attention module is embedded to strengthen the extraction of watermark signals. The two attention modules work synergistically: the encoder focuses on adaptive feature fusion while the decoder leverages channel attention to selectively enhance watermark-related features, forming a dual-attention synergy that balances robustness and imperceptibility. Moreover, the dynamic gating network adaptively adjusts the contribution of local versus global features via learnable weights, whose evolution from approximately 0.51 to about 0.89 improves model interpretability. Experiments are conducted on the COCO 2017 dataset. Compared with HiDDeN, the proposed algorithm reduces the bit error rate (BER) from 0.1696 to 0.1538 under no attack with a relative reduction of 9.3%, increases PSNR by 0.61 dB, and improves SSIM from 0.9058 to 0.9077. Under various attacks—including JPEG compression, Gaussian noise, salt-and-pepper noise, and brightness/contrast adjustments—the BER remains consistently lower than that of HiDDeN. Ablation studies confirm the effectiveness of each module. Overall, the proposed algorithm preserves visual quality, improves the accuracy of watermark embedding and extraction, and exhibits strong generalization robustness against common image distortions. Full article

A reinforcement-learning-based wall-modeled large-eddy simulation (RL-WMLES) framework is proposed to improve the physical consistency of near-wall turbulence predictions. In this approach, a reinforcement learning agent is coupled with the WMLES solver to dynamically adjust a compensating stress term, with the objective of enforcing agreement between the LES solution and the law of the wall. The agent is trained using the proximal policy optimization (PPO) algorithm, where the state is defined as the discrepancy between the near-wall LES velocity and the wall-model prediction, and the action corresponds to modifying a parameterized support viscosity distribution. The proposed method is implemented within a high-performance CFD solver and trained on turbulent channel flow. Numerical results demonstrate that the trained agent effectively reduces the log-layer mismatch and significantly improves the accuracy of near-wall velocity predictions. Furthermore, the RL-WMLES framework exhibits a degree of generalization capability: the trained agent performs robustly with varying levels of numerical dissipation and Reynolds numbers. By introducing a simple interpolation strategy, the same agent can be successfully applied to configurations with different matching locations. Overall, the RL-WMLES framework provides a flexible and data-driven approach for enforcing physical constraints in turbulence modeling. The method shows strong potential for extension to more complex flows. Full article

Continuous monitoring of lower-limb kinematics in natural environments is essential for gait analysis and rehabilitation but remains challenging due to the limitations of optical systems and the inaccuracy of sparse inertial sensor methods. To address this, we propose a high-precision, minimalist wearable system utilizing only three inertial measurement units placed on the pelvis and shanks. In the data preprocessing stage, engineering modifications are made based on the traditional gradient descent algorithm to implement adaptive channel adjustment on the acceleration and magnetic data of a single IMU, aiming to alleviate the impact of motion acceleration and external magnetic interference on the temporal feature manifold. Subsequently, a pure Transformer neural network is utilized to capture long-range temporal dependencies, reconstructing full lower-limb kinematics without relying on rigid biomechanical assumptions. The model was optimized and deployed on an STM32N647 microcontroller to achieve real-time edge inference with a low latency of approximately 17 ms. Experimental results demonstrate that the proposed method achieves a mean absolute error of 2.41° for level walking, significantly outperforming traditional constrained Kalman filter approaches. Furthermore, it maintains high tracking robustness during complex nonlinear movements such as squatting and lunging. In conclusion, this edge-computing-enabled framework provides an accurate, comfortable, and real-time solution for unconstrained human motion capture in daily scenarios. Full article

Coastal reclamation reshapes tidal hydrodynamics in semi-enclosed bays by removing intertidal storage, modifying channel conveyance, and redistributing tidal exchange among connected sub-regions. This study quantifies the multi-decadal cumulative impacts of reclamation on tidal currents and tidal prism in Yueqing Bay, China, using shoreline and bathymetric reconstructions for 1978, 2002, 2013, and 2020; hydrological observations; and a two-dimensional MIKE21 FM tidal hydrodynamic model. Characteristic cross-sections were used to estimate bay-wide and sub-regional tidal prisms, and representative stations were used to diagnose current-speed responses. The bay-wide tidal prism decreased from 15.235 × 10⁸ m³ in 1978 to 12.316 × 10⁸ m³ in 2020, corresponding to a reduction of 2.919 × 10⁸ m³ (19.16%). The strongest loss occurred during 1978–2002, when large-scale reclamation and closure of the Xuanmen Channel removed tidal storage and redirected flow into the remaining main-channel system. Although reclamation intensity weakened after 2013, mean current speed still changed by −0.050 to 0.033 m/s and sub-regional tidal-prism shares continued to adjust, indicating delayed hydrodynamic reorganization rather than immediate stabilization. These results show that reclamation impacts cannot be explained by reclaimed area alone; they depend on project timing, spatial layout, and the connectivity with key tidal pathways. The findings support staged assessment and pathway-sensitive shoreline management in reclaimed semi-enclosed bays. Full article

Accurate surface defect classification is a critical requirement in structural health monitoring and infrastructure inspection, where defects, including cracks, spalling, delamination and noncrack regions, often appear with low-contrast and complex background textures. Motivated by the need for a robust and discriminative framework that can enhance defect visibility and focus learning on damage-critical regions, this research proposes a novel damage-aware DenseNet-201 (DA-DenseNet-201) model for surface defect classification. As a critical novelty, a damage-aware adaptive contrast-limited adaptive histogram equalisation (DAC) filtering strategy is introduced as a preprocessing stage. The proposed DAC filter dynamically adjusts contrast enhancement parameters based on damage indicators, selectively amplifying crack edges and defect textures while preserving healthy surface regions and suppressing noise. Building on this method, enhanced images are processed using a pretrained DenseNet-201 backbone, retaining the benefits of dense feature propagation and efficient gradient flow. To strengthen the discriminative learning of DA-DenseNet-201 further, an attention refinement block is integrated into the network, combining channel attention to emphasise defect-relevant feature responses and spatial attention to localise damage regions accurately. In addition, a multiscale feature fusion mechanism aggregates feature maps from multiple dense blocks to capture fine-grained crack patterns, texture-level degradation and high-level semantic damage information. Extensive experiments conducted on surface defect datasets demonstrate its effectiveness, achieving a superior classification accuracy of 98.93%, along with notable improvements in sensitivity, specificity and the intersection over union compared with state-of-the-art models. These results confirm that the proposed DA-DenseNet-201 provides a reliable and high-performance solution for automated surface defect classification. Full article

Wetland restoration plays a crucial role in enhancing hydrologic resilience amidst the challenges posed by climate change and evolving land uses. The historical reduction in beaver populations due to the fur trade and alterations to riparian zones have compromised the ecological stability of many landscapes. Presently beaver populations are increasing as there are now protections in place for them. In response, Beaver Dam Analogs (BDAs) have emerged as an effective restoration strategy, particularly in regions where natural beaver activity is limited due to inadequate habitat conditions. BDAs are a human-made structure that mimics the function and form of natural beaver dams. This paper focuses on a restoration project within the Fish Creek area between the year 2019 and 2021, which is a part of the Weber River watershed in northern Utah, where BDAs were installed to rehabilitate a degraded wetland and rectify an incised channel network. Over the initial two years following the installation (2019–2021), significant ecological transformations were observed. Notably, there was an increase in the areal coverage of sediments that sizes ranged from 1 to 256 mm within the stream channel, alongside a corresponding decrease in coarser substrates. These changes facilitated a reduced channel slope, indicating substantial sediment deposition above the installed BDAs. Concurrently, there was an expansion in riparian vegetation along an approximate stretch of 40 m, primarily grasses, reflecting an adjustment in habitat conditions favorable to riparian recovery. The preliminary outcomes from this study contribute to a broader understanding of the dynamics involved in BDA-driven restoration efforts in semiarid regions like the western United States, highlighting the potential shifts in riparian habitats prompted by such interventions. Full article

Small object detection under adverse weather remains challenging due to weather-induced domain shifts and sparse visual cues of small targets. In contrast to R-YOLO/QTNet and conventional UDA methods, which mainly rely on weather-specific restoration/enhancement or global feature/magnitude alignment, SABDR explicitly targets cross-weather small object adaptation through bidirectional domain translation, degradation-aware receptive-field modeling, feature-statistics modulation, and style-direction alignment. Specifically, the Bidirectional Dynamic Domain Adaptation Network, termed BiDDC-Net, translates between source and target domains and dynamically adjusts receptive fields according to weather severity. The Style-Aware Domain Adaptation Module, termed AIFI-DA, enhances discriminative small-object channels using feature statistics. SDA is further used as a complementary training-time regularizer to encourage style-direction consistency without directly matching feature magnitudes. Experiments are conducted on Cityscapes→Foggy Cityscapes and MOT-Fly→Foggy/Rainy/Snowy MOT-Fly, including newly added rainy and snowy MOT-Fly settings, with both YOLOv5s and YOLO26 evaluated on all MOT-Fly weather conditions. SABDR achieves 47.7 ${\mathrm{mAP}}_{50}$ on Cityscapes→Foggy Cityscapes, and obtains 96.0%/96.8%, 66.7%/77.1%, and 95.0%/95.6% ${\mathrm{mAP}}_{50}$ on Foggy, Rainy, and Snowy MOT-Fly with YOLOv5s/YOLO26, respectively. The improvements on MOT-Fly are reported under a fixed single-seed setting and should therefore be interpreted as single-run empirical gains rather than statistically validated improvements. These results demonstrate its effectiveness under the evaluated fog/rain/snow cross-weather small object detection settings. Full article

Against the backdrop of global energy transition and multiple uncertainties, enhancing energy resilience has become a core priority for China’s pursuit of secure and sustainable development. Using Chinese provincial panel data from 2011 to 2019, this study applies a two-way fixed effects model, mediation effect tests, and interaction term analysis to empirically investigate the relationship between industrial structure, green finance, and energy resilience. The main findings are as follows. First, the increases in gross regional product (GRP) and the added value of the secondary and tertiary sectors significantly enhance energy resilience. Second, heterogeneity analysis indicates that in regions with a high level of green finance, both GRP and the secondary sector’s added value exhibit stronger positive effects on energy resilience, whereas in regions with lower levels of green finance, the tertiary sector’s added value contributes more significantly to energy resilience improvement. In areas with high coal dependency, the secondary sector’s added value shows a significantly positive effect on energy resilience. Increases in industrial and construction industry added value significantly enhance energy resilience, suggesting that the expansion of the secondary industry contributes positively to the stability and resilience of the energy system. Third, the mechanism analysis shows that green finance contributes to energy resilience partly through the optimization of the energy consumption structure. Specifically, by effectively curbing coal consumption and, to a lesser extent, fuel oil production, green finance reduces the structural dependence of the economy on high-carbon energy. By contrast, channels such as electricity generation yield weaker and less robust evidence. These findings suggest that energy resilience is fundamentally shaped by the interplay of industrial structure, financial intermediation, and energy structure adjustment. Therefore, policy should shift from single instruments to integrated governance, synergizing industrial policy, green finance, and energy optimization to bolster energy resilience. Full article

This study examines the dynamic relationship between global oil prices, fiscal policy, and non-oil economic activity in Saudi Arabia. It provides an integrated analysis of how oil price shocks are transmitted to domestic macroeconomic conditions and examines the role of fiscal policy in shaping economic adjustment and diversification dynamics under Vision 2030. Using quarterly data from 2010 Q1 to 2025 Q3, a Vector Error Correction Model is applied to analyze long-run equilibrium relationships and short-run dynamics among oil prices, consumer prices, government consumption expenditure, and non-oil GDP. The findings reveal a dominant role for fiscal policy within the estimated transmission framework, with oil price increases significantly stimulating government spending. Although fiscal expansion supports non-oil output, its effects remain modest and short-lived, reflecting relatively weak fiscal multipliers. Oil price shocks influence inflation mainly through indirect fiscal channels. While the variables exhibit long-run equilibrium relationships, the response of non-oil GDP to oil price shocks remains indirect, moderate, and transitory, suggesting gradual but incomplete progress toward economic diversification. The study highlights the importance of strengthening countercyclical fiscal frameworks, improving expenditure efficiency, and sustaining structural reforms to enhance macroeconomic stability and reduce vulnerability to oil price volatility. Full article

Current terahertz (THz) metasurfaces are often constrained by fixed operational states, lacking the flexibility to switch dynamically between transmission and reflection modes. To address this limitation, we propose a tunable coded metasurface based on the photo-adjustable conductivity of silicon, enabling seamless mode switching and versatile wavefront manipulation. By leveraging the photo-induced dielectric-to-metallic transition, the device functions as a high-efficiency transmission-type polarization converter under zero pump fluence, transforming incident X-polarized waves into Y-polarized waves across a broad frequency range of 0.85–1.5 THz, with a polarization conversion ratio (PCR) exceeding 99%. Upon excitation by 800 nm near-infrared laser pulses, the metasurface transitions to reflection mode, where it simultaneously achieves linear polarization conversion and generates dual-channel orbital angular momentum (OAM) beams through a phase-coding strategy integrated with Fourier convolution. Furthermore, by employing the Gerchberg–Saxton (GS) algorithm to optimize the phase profile, holographic reconstruction is realized in the far field. This design integrates diverse manipulation capabilities into a single, dynamically controllable platform, offering a promising technological approach for THz information processing and integrated photonic systems. Full article

To address modality inconsistency, insufficient intra-modal affective representation, and the limited adaptability of conventional fusion strategies in multimodal sentiment analysis, this study proposes ECA-CMF Net, an Efficient Channel Attention-enhanced Conditional Modulation Fusion network. The framework integrates unified indexing-based preprocessing, heterogeneous feature extraction with ECA, and Conditional Modulation Fusion to improve multimodal representation learning and sentiment classification. Specifically, sample-level alignment and modality-specific standardisation are first applied to textual, visual, and acoustic inputs to reduce distribution shifts and noise interference. Then, heterogeneous encoders extract modality-specific features, while ECA adaptively recalibrates sentiment-relevant channels and suppresses redundant information. Finally, the CMF mechanism generates modulation parameters from joint multimodal context to scale and shift modality features, enabling dynamic cross-modal interaction and contribution adjustment. Experiments on CMU-MOSI and CMU-MOSEI show that ECA-CMF Net achieves ACC/F1 scores of 0.8874/0.8870 and 0.7089/0.7008, respectively. Compared with the strongest reproduced baselines, it improves ACC/F1 by 3.40/3.38 percentage points on CMU-MOSI and 1.53/1.85 percentage points on CMU-MOSEI, demonstrating improved multimodal collaboration, adaptive fusion, and robustness. Full article

Vehicle sales constitute an important component of household consumption and a key transmission channel of macro-financial conditions in South Africa. This study investigates the macroeconomic determinants of vehicle sales by examining the roles of economic activity, interest rates, and inflation over the period 2000Q1 to 2025Q4. Using quarterly data, the analysis employs the autoregressive distributed lag (ARDL) bounds testing approach to estimate both long-run and short-run relationships, complemented by an error correction model and Granger causality analysis. The results confirm the existence of a stable long-run cointegrating relationship among the variables. In the long run, vehicle sales respond positively to economic growth, while inflation and interest rates are associated with reduced demand. Short-run dynamics indicate that vehicle sales respond positively to economic growth, and negatively to interest rates and inflation, reflecting affordability and credit constraints, alongside rapid adjustment to macroeconomic shocks. The Granger causality results suggest that vehicle demand is largely driven by macro-financial conditions rather than exerting feedback effects on them. Overall, the findings highlight the sensitivity of South Africa’s automotive sector to macroeconomic stability and underscore the importance of prudent monetary policy and price stability in sustaining durable goods demand. Full article

Background: Obstructive sleep apnea (OSA) is a common sleep-related breathing disorder characterized by recurrent episodes of intermittent hypoxia and systemic inflammation. Pannexin-1 (Panx1) is a transmembrane channel involved in ATP release and purinergic signaling and has been implicated in hypoxia-related inflammatory responses. However, the clinical relevance of circulating Panx1 levels in patients with OSA remains poorly understood. This study aimed to evaluate serum Panx1 concentrations in patients with OSA and to investigate their association with nocturnal hypoxemic burden. Methods: In this cross-sectional study, 40 patients with obstructive sleep apnea (OSA) and 40 control subjects underwent overnight polysomnography for diagnostic evaluation. Serum Panx1 concentrations were measured using an enzyme-linked immunosorbent assay (ELISA). Logistic regression models were constructed to evaluate the association between Panx1 and OSA status while adjusting for clinical covariates. In addition, a propensity score–matched sensitivity analysis based on age, sex, and body mass index was performed to further assess potential confounding. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic performance of Panx1 alone and in combination with clinical variables. Results: Serum Panx1 levels were significantly lower in patients with OSA than in controls (4.27 ± 2.66 vs. 6.24 ± 4.75 ng/mL, p = 0.013). Although Panx1 was not an independent predictor of OSA after adjustment for age, sex, and body mass index, its integration with clinical variables significantly improved diagnostic discrimination. The area under the receiver operating characteristic curve increased from 0.662 for Panx1 alone to 0.858 in the fully adjusted model. Sensitivity analyses attenuated the observed association after matching for major baseline characteristics, suggesting a potential contribution of demographic and anthropometric factors. In addition, Panx1 concentrations were inversely correlated with markers of nocturnal hypoxemic burden, particularly the cumulative time spent with oxygen saturation below 90% (T90). Conclusions: Lower serum Panx1 concentrations were associated with OSA status and nocturnal hypoxemic burden. While Panx1 alone demonstrated modest discriminatory ability, its integration with established clinical factors improved diagnostic performance. These findings suggest that Panx1 may represent a biologically plausible adjunct biomarker reflecting hypoxic burden and may contribute to multi-parameter approaches for OSA risk assessment; however, further validation in larger matched cohorts is warranted. Full article