Search Results (3,343)

The rapid proliferation of image editing tools poses unprecedented challenges to information sustainability and social trust, as malicious digital forgeries can easily contaminate public discourse, news reporting, and legal forensics. Advanced image editing techniques make image tampering increasingly difficult for the naked eye to recognize, which requires highly accurate methods for detecting and localizing image tampering. In this paper, an end-to-end network model named AEM-Net is proposed. AEM-Net combines RGB and SRM features to enhance the model’s sensitivity to image details and potentially tampered regions through multi-scale feature extraction and fusion. AEM-Net consists of the HRNet-based Multiscale Feature Extraction Module and the Context-Aggregated Pyramid Localization Module (CAPLM). The multi-scale feature extraction module utilizes the Attentional Perceptual Feature Fusion Module to adaptively focus on the anomalous regions. In contrast, the CAPLM utilizes the Expanded Convolutional Feedback Enhancement Module to effectively exploit contextual feature information for achieving pixel-level localization of tampered regions. Experimental results on public benchmark datasets demonstrate that AEM-Net achieves superior performance compared with existing state-of-the-art methods. In particular, AEM-Net achieves an AUC/F1 score of 95.36%/67.19% on CasiaV1, 93.25%/79.75% on Coverage, and 87.36%/66.24% on NIST16, while requiring only 0.09 s to process a single image, demonstrating both high localization accuracy and computational efficiency. Full article

Skeleton-based human action recognition (HAR) requires models that preserve the local kinematic structure of the human body while capturing long-range spatiotemporal dependencies under noisy or incomplete joint observations. Traditional Graph Convolutional Networks (GCNs) provide topology-aligned inductive bias but are often limited by local information aggregation from neighboring joints. In contrast, attention-based mechanisms capture global interactions, yet they may attend to spurious correlations when skeletal constraints are weakly enforced. This paper proposes Differential Hyperedge Attention-enhanced GCN (DHA-eGCN), a hybrid architecture that couples structure-aware Differential Hyperedge Attention with multi-scale temporal convolution for spatiotemporal skeleton sequence processing. DHA injects skeletal structure into attention via hop-distance relative positional encoding and hyperedge context tokens generated via joint-to-part pooling. It further employs differential attention to suppress shared noisy correlations and enhance interaction selectivity. To strengthen spatial grounding, an explicit GCN branch is added under partial- or full-depth configurations, where the first four or all ten layers are applied with graph convolutions. The model further employs an ensemble strategy that combines predictions from multiple complementary model instances. Our experiments on NTU RGB+D 60 under the X-Sub and X-View protocols, NTU RGB+D 120 under the X-Sub and X-Set protocols, and Northwestern-UCLA demonstrate that DHA-eGCN consistently outperforms or remains competitive with strong graph-based, transformer-based, and hybrid state-of-the-art methods based on the same four-stream architecture. The best configuration achieves 93.7% and 97.0% on NTU RGB+D 60 X-Sub and X-View, respectively; 90.9% and 91.9% on NTU RGB+D 120 X-Sub and X-Set, respectively; and 97.6% on Northwestern-UCLA. Full article

Face super-resolution (FSR) aims to reconstruct high-quality high-resolution face images from low-resolution inputs. Although CNN–Transformer hybrid models have shown promising performance by jointly modeling local textures and global dependencies, their large parameter sizes and high computational costs hinder practical deployment in resource-constrained scenarios such as mobile devices and embedded systems. Meanwhile, existing lightweight SR models usually reduce complexity by simplifying network depth, channel dimensions, or convolutional operations, which may weaken feature representation capability and lead to insufficient recovery of fine facial structures. To address these issues, this paper proposes HCTIUNet, a lightweight CNN–Transformer hybrid network based on an inverted U-shaped architecture. Specifically, the proposed network integrates lightweight CNN branches for local facial texture extraction and Transformer branches for global dependency modeling, while introducing a multi-scale feature interaction strategy and a global feature refinement module to enhance facial structural details. Experimental results on the FFHQ, CelebA, and Helen datasets demonstrate that HCTIUNet achieves competitive performance under the ×8 face super-resolution setting, obtaining PSNR/SSIM/LPIPS values of 27.55 dB/0.765/0.225, 27.63 dB/0.761/0.212, and 27.53 dB/0.777/0.213, respectively. Moreover, HCTIUNet contains 10.5 M parameters, requires 9.9 G FLOPs, and achieves an inference time of 0.021 s. These results indicate that the proposed method achieves a favorable trade-off between reconstruction accuracy, perceptual quality, and computational efficiency, making it suitable for efficient face super-resolution applications. Full article

Wave overtopping is one of the most complex coastal hazards to characterize in field conditions due to its high non-linearity and the interaction between unsteady hydrodynamics and wave–structure processes. To get insights into the underlying occurrence and persistence of overtopping, this study proposes an integration of numerical and data-driven models. Multi-month field observations made at a breakwater are used to investigate the hydro-meteorological parameters causing overtopping initiation and persistence. High-frequency video-derived overtopping detections are combined with coupled ADCIRC–UnSWAN (ADvanced CIRCulation–Unstructured Simulating WAves Nearshore) hindcasts to construct near-structure hydro-meteorological conditions. The results reveal a clear dynamical asymmetry showing that overtopping initiation corresponds to exceedance of crest elevation at individual wave-scale associated with elevated wave height, water level, wave steepness, and wind characteristics, whereas overtopping persistence depends on short-term temporal effects associated with wave energy, direction, and sustained water levels. Gradient-boosted decision trees, temporal convolutional networks, and Transformer models are employed, demonstrating that persistence cannot be inferred from instantaneous sea-states alone, indicating a separation of timescales between triggering and sustained overtopping dynamics. These findings provide field-scale evidence of distinct hydrodynamic regimes governing overtopping processes, highlighting the importance of temporal characteristics for understanding overtopping dynamics and developing predictive coastal hazard frameworks. Full article

Edge vision systems require convolutional neural networks (CNNs) that preserve recognition accuracy under strict storage, computation, and latency constraints. Although ResNet18 is a compact residual backbone, direct deployment on resource-constrained devices remains costly, whereas simple channel reduction weakens representation capacity. This study aims to build a deployable ResNet18-based classifier that reduces model complexity while recovering the accuracy lost during compression. We propose a lightweight framework that combines global channel scaling, a re-parameterized attention residual block, and teacher–student knowledge distillation. The proposed block uses multi-branch convolution and squeeze-and-excitation attention during training, then folds the linear branches into a single 3-by-3 convolution for inference. Experiments on CIFAR-100 show that the final model reduces parameters from 11.220 M to 2.841 M, retains comparable Top-1 accuracy (0.7579 vs. 0.7606), improves Top-5 accuracy (0.9340 vs. 0.9253), and reduces graphics processing unit (GPU) batch inference latency from 3.279 ms to 2.161 ms. Deployment on PYNQ-Z2 verifies the complete camera-based CPU-side inference workflow, with an average end-to-end latency of 421.467 ms/frame. The results indicate that residual topology preservation, re-parameterized feature enhancement, and distillation form a practical route for edge-oriented lightweight CNN deployment. Full article

Stock index price series are composed of superimposed multi-frequency components, including long-term trends, cyclical fluctuations, and stochastic noise. Effectively decoupling these heterogeneous components and modeling them separately is key to improving forecasting accuracy. Existing methods under the “decomposition–prediction” paradigm mostly employ fixed-scale decomposition, and the forecasting models are not specifically adapted to the non-stationary and high-noise characteristics of financial data, resulting in limitations in adaptivity and local dynamic capture. This paper proposes a frequency-aware adaptive multi-scale decomposition Transformer hybrid model (FAMS-Transformer). At the decomposition level, the fast Fourier transform is used to dynamically identify dominant cycles, thereby adaptively decoupling trends and fluctuations, overcoming the limitations of fixed-scale decomposition. At the forecasting level, a lightweight depthwise separable convolution is embedded between the self-attention and feedforward network of the Transformer encoder, enhancing the model’s ability to capture local temporal dynamics and achieving collaborative modeling of global dependencies and local information. Comparative experiments with 15 baseline models including LSTM, Transformer, TimesNet, and FreTS on three representative Chinese market indices—Shanghai Composite Index, Shenzhen Component Index, and Small and Medium Enterprises 100 Index—across four prediction horizons from one step to 15 steps demonstrate that FAMS-Transformer achieves the best forecasting accuracy in all scenarios. The coefficient of determination for 15-step prediction remains stably between 0.730 and 0.928. Moreover, the model still performs well on the S & P 500 dataset. Ablation studies and significance tests further validate the effectiveness of each core module and the statistical significance of the performance improvements. Full article

Small object detection in remote sensing imagery is a fundamental task for visual information extraction, yet it remains challenging due to extremely small target scales, complex backgrounds, and the loss of discriminative feature information caused by repeated downsampling. To address these issues, this paper proposes a Wavelet-Convolution and Fine-Grained Preservation Network (WCFPNet) based on YOLOv8n. Specifically, a Wavelet-Convolution Module (WCM) is introduced into the backbone to decompose feature maps into low- and high-frequency sub-bands, thereby enhancing structural feature modeling and preserving subtle target details. To compensate for the weakened fine-grained information after repeated downsampling, an Enhanced Spatial Pyramid Pooling-Fast (ESPPF) module is embedded at the end of the backbone to strengthen multi-scale contextual aggregation. In addition, an Enhanced Feature Pyramid Network (EFPN) is designed in the neck to facilitate the propagation of shallow and intermediate fine-grained features to high-level semantic features through cross-level fusion and the Convolutional Block Attention Module (CBAM). Experiments on the NWPU VHR-10 dataset show that WCFPNet achieves 0.879 mAP@0.5 and 0.515 mAP@0.5:0.95, outperforming YOLOv8n by 1.7 and 2.5 percentage points, respectively. Moreover, the proposed WCFPNet achieves a competitive performance compared with several representative detectors while maintaining moderate model complexity. These results demonstrate the effectiveness of WCFPNet in challenging remote sensing scenes characterized by complex backgrounds, dense object distributions, and weak textures. Full article

Accurate electricity price forecasting (EPF) is essential for market participants to optimize trading strategies and for power systems to accommodate the increasing penetration of volatile renewable energy sources. However, electricity price series are characterized by strong nonlinearity, high volatility, and significant structural breaks, which pose substantial challenges to conventional forecasting models. Although numerous hybrid deep learning models have been proposed for EPF, most existing approaches either overlook structural breaks or treat them as outliers rather than as signals of regime shifts, often resulting in systematic forecasting degradation when market conditions change abruptly. To address this issue, this study proposes COCAL-TTL, a novel multi-stage structural break-aware forecasting framework that integrates regime-adaptive data partitioning with a functionally differentiated hybrid deep learning architecture. First, a joint detection scheme combining the Iterated Cumulative Sum of Squares (ICSS) algorithm and the Chow test is employed to partition Spanish electricity market data from 2014 to 2023 into distinct regimes. Within each regime, CEEMDAN is applied to extract multi-scale features, which are subsequently reconstructed into trend, periodic, and random components based on an independent sample t-test and Fast Fourier Transform (FFT). The CNN-SE Attention-LSTM (CAL) model, with hyperparameters optimized by the Osprey Optimization Algorithm (OOA), serves as the primary forecasting engine. In addition, a dedicated heterogeneous error correction module, namely TTL, is introduced, in which Temporal Convolutional Network, Transformer, and LSTM are designed to capture local transients, long-range dependencies, and transitional dynamics in the residual series, respectively. Empirical results demonstrate that compared with the Naive benchmark, COCAL-TTL achieves percentage MAPE improvements of 58.48% and 48.97% in low- and high-volatility regimes, respectively. These findings indicate that the proposed structural break-aware framework provides a robust data-driven solution for EPF under heterogeneous market conditions and offers technical support for stable electricity market operation in the context of renewable energy integration. Full article

Quantum convolutional neural networks (QCNNs) are emerging as promising models in quantum machine learning, particularly for image classification and computer vision tasks. Recent developments include hybrid classical–quantum architectures, advanced quantum encoding methods, and novel circuit designs that improve data processing on Noisy Intermediate-Scale Quantum (NISQ) devices. However, practical implementation remains challenging due to circuit complexity, gate count, qubit connectivity, and hardware noise, which limit scalability and performance. Consequently, quantum circuit optimization has become essential for reducing resource requirements and improving classification accuracy. This study presents a systematic literature review of 40 research papers published between 2014 and 2025. The review covers QCNNs together with closely related quantum neural network (QNN) models and quantum circuit optimization studies, since circuit-optimization techniques are frequently developed for QNNs more broadly rather than for QCNN architectures in isolation. Within this scope, it examines network architectures, encoding strategies, application domains, and optimization techniques, with particular attention to heuristic and metaheuristic approaches such as genetic algorithms and evolutionary strategies. The findings highlight growing trends in hybrid quantum–classical integration, the widespread adoption of metaheuristic optimization, and the importance of multi-objective frameworks adapted to quantum hardware constraints. Finally, the review identifies key research gaps and future directions for practical QCNN deployment on near-term quantum devices. Full article

To address the issues of severe noise interference, limited classification capability of linear classifiers, and difficulty in adaptively optimizing classifier parameters in rolling bearing fault diagnosis, this paper proposes a hybrid diagnostic model integrating the multi−scale convolutional neural network and rime ice optimization algorithm optimized kernel extreme learning machine. The method first employs the synchrosqueezed wavelet transform to convert raw vibration signals into high−resolution time−frequency images, effectively enhancing the visualization of fault impact features. Then, the multi−scale convolutional neural network is used to extract preliminary features from the time−frequency images, and the kernel extreme learning machine is introduced to replace the Softmax linear classifier in traditional convolutional neural networks, thereby constructing a nonlinear decision boundary to more effectively separate complex fault patterns. Finally, the rime algorithm is introduced to optimize the regularization coefficient and kernel parameters of the kernel extreme learning machine, enabling the kernel extreme learning machine to perform fault classification with an optimal nonlinear decision boundary. Experimental results on the bearing datasets from Huazhong University of Science and Technology and Case Western Reserve University show that the proposed method achieves classification accuracies of 99.75% and 99.83%, respectively, outperforming several comparison models. Furthermore, noise robustness experiments demonstrate that the proposed model maintains an accuracy of approximately 90% under low signal−to−noise ratio (SNR) conditions, outperforming all comparison models and demonstrating high classification accuracy under strong noise. Full article

The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) has collected tens of millions of spectra, providing an unprecedented resource for large-scale spectroscopic studies. Efficient retrieval techniques are therefore essential for exploring such massive datasets. Existing approaches often rely on predefined templates or manually labeled training samples, which can limit their applicability in large and diverse spectral archives. In this work, we present a general similarity-retrieval framework that combines self-supervised contrastive learning based on a convolutional neural network with Facebook AI Similarity Search (FAISS) for efficient large-scale spectral retrieval. The framework learns spectral representations directly from unlabeled data and enables flexible retrieval from user-defined wavelength regions based on feature similarity. We evaluate the framework on several stellar populations in LAMOST DR8. For late-type M8-star retrieval, 90.5% of the top 1000 retrieved spectra are later than M6. For M0–M5 giants, the mean retrieval accuracy across six subtypes reaches 94.8%. Using a C-H star spectrum as the query spectrum, 90.8% of the top 1000 retrieved candidates are classified as carbon stars by the LAMOST pipeline. Cross-matching with SIMBAD further confirms 255 C-H stars and 47 C-R stars among the retrieved candidates. These results demonstrate that the proposed framework can efficiently identify spectrally similar objects across large spectroscopic databases and can serve as a useful tool for searching for rare or spectrally distinctive stellar populations. Full article

Early depression screening is important for timely intervention, and electroencephalography (EEG) offers an objective and potentially portable sensing modality for computer-aided assessment. However, recognition from fixed three-channel frontal EEG remains difficult because of limited spatial information and incomplete modeling of temporal–frequency characteristics and temporal dependencies. This study proposes DEP-TFDualNet for acquisition-constrained frontal resting-state EEG. The framework integrates multi-scale convolution, dual-domain channel attention, temporal modeling derived from the independent recurrent neural network (IndRNN) architecture, and decision-stage fusion of deep representations with low-order statistical descriptors through a Kolmogorov–Arnold Network (KAN)-based nonlinear projection layer. Experiments were conducted on the publicly available three-channel frontal EEG subset of the MODMA dataset. After additional quality control, 48 subjects were retained (22 patients with major depressive disorder, 26 healthy controls). Under subject-wise stratified five-fold cross-validation, DEP-TFDualNet achieved 85.42% accuracy, 85.26% macro-F1, 81.82% sensitivity, 88.46% specificity, an AUC of 0.82, and a Brier score of 0.121. It achieved the best threshold-based subject-level performance and the lowest Brier score among the evaluated models. These results provide preliminary evidence that simplified frontal EEG sensing may support depression recognition in acquisition-constrained settings, although larger and external validation is still required. Full article

Aero-engine predictive maintenance is challenged by heterogeneous operating conditions, complex degradation patterns, and the need for interpretable maintenance alerts rather than solely numerical life estimates. This study investigates a condition-aware data-driven framework for jet engine health assessment using the NASA C-MAPSS dataset, which contains four benchmark subsets (FD001–FD004) with different operating conditions and fault modes. Instead of formulating the task as conventional remaining useful life regression, this study reformulates degradation assessment as a three-class health state classification problem, including Normal, Warning, and Fault. A unified preprocessing pipeline is developed, incorporating condition-wise normalization, first-order differential feature construction, and per-unit sliding window segmentation to reduce operating-condition bias, capture degradation dynamics, and prevent data leakage. Five representative models are evaluated under the same framework, including XGBoost, LightGBM, Random Forest, a context-aware multi-scale temporal attention convolutional neural network, and a bidirectional long short-term memory network. The results show that the proposed framework achieves consistently high classification accuracy across all four subsets, with the best results of 0.9841 on FD001, 0.9764 on FD002, 0.9891 on FD003, and 0.9832 on FD004. In addition, Bi-LSTM outperforms MSTA-CNN on all subsets, for example improving accuracy from 0.9614 to 0.9747 on FD002 and from 0.9773 to 0.9806 on FD004, which is consistent with the importance of long-term temporal dependency modeling for this task. These findings suggest that the proposed framework provides an effective and maintenance-decision-aligned solution for C-MAPSS-based health monitoring, where the three-class alert output offers clearer operational meaning than a single numerical life estimate. Full article

Signals generated during tool wear are nonlinear, non-stationary, and easily affected by machining noise, which makes reliable tool condition monitoring difficult in intelligent manufacturing. To address this issue, this study proposes a tool wear degree classification framework, FCTrans-CA, that fuses time-domain and frequency-domain information through a lightweight cross-attention (CA) bridge. Fast Fourier transform (FFT) is first used to obtain frequency-domain representations. The raw time-domain signals are processed by a multi-scale one-dimensional convolutional neural network (MS-CNN) to extract temporal wear features, while the FFT-derived representations provide complementary spectral cues. These two feature streams are fused by an asymmetric CA module in which frequency-domain features guide the selection of wear-sensitive temporal features. K-means clustering is used to divide the measured flank wear (VB) trajectory of each tool into initial-, normal-, and severe-wear stages, thereby reducing subjectivity in label generation. Experiments on the PHM2010 milling dataset show that FCTrans-CA achieves 99.43% classification accuracy on 40,648 test samples. The results indicate that cross-domain feature interaction improves the separability of wear states and provides a reproducible data-driven route for tool wear monitoring. Full article

Single-lead electrocardiogram (ECG) is widely used in wearable devices for atrial fibrillation (AF) screening. Nevertheless, subtle pathological characteristics like P-waves and f-waves in practical signals are vulnerable to noise contamination. Meanwhile, the scarcity of high-quality annotated abnormal data instances leads to severe class imbalance. To mitigate these issues, we present an end-to-end framework designed for arrhythmia diagnosis using single-lead ECG signals, which integrates quality-aware data augmentation with a Peak-Enhanced attention mechanism. First, to mitigate the problem of data imbalance, a Quality-Aware Generative Adversarial Network (QA-GAN) is designed. This network integrates a signal quality evaluation module based on signal kurtosis, together with a dynamic soft-label training scheme, guiding the generator to prioritize learning high-quality morphological features, thereby synthesizing high-fidelity minority class samples. Second, to accurately capture subtle pathological features in electrocardiograms, a Peak-Enhanced Attention Convolutional Network (PEAC-Net) classification model is proposed. This model incorporates a Peak-Enhanced Attention (PE-Att) module, which employs learnable derivative convolutional kernels to precisely identify the transition points in the ECG signal. Furthermore, by integrating one-dimensional multi-scale dilated convolution (DSGC1D) with bidirectional LSTM, the model achieves effective capturing of both fine-grained local morphological features and long-range global rhythm patterns. Experimental results on the PhysioNet 2017 dataset indicate that the presented model attains an accuracy of 0.902 and a macro-F1 score of 0.880, respectively, outperforming other state-of-the-art models and also exhibiting robust data adaptability on the MIT-BIH dataset. Full article