Search Results (892)

Accurate forecasting of oil well production via field monitoring systems is significantly restricted by a structural conflict in modeling, where temporal dependency learning and nonlinear feature representation are closely coupled. Such coupling forces a trade-off between capturing long-term temporal dependencies and retaining sensitivity to short-term sensor fluctuations, while amplified local sensitivity easily increases noise interference and weakens model robustness under complex non-stationary sensor dynamics. To solve this problem, this study proposes a novel sensor-driven hybrid framework named Temporal Augmented Residual Network (TAR-Net), which adopts a decoupled paradigm to separate global temporal modeling and local fluctuation compensation explicitly. A multi-scale dilated Temporal Convolutional Network (TCN) extracts long-range temporal patterns from multi-source sensor data, and a LightGBM-based residual module conducts targeted error correction. Meanwhile, multi-scale temporal features and adaptive multi-fidelity Bayesian optimization are applied to enhance model adaptability. Validated on real sensor data from the Volve oilfield, TAR-Net surpasses 13 benchmark models with an R² of 0.9832 and a MAPE of 7.8%. Residual and trajectory analyses verify its balance between global trend consistency and local fluctuation sensitivity. This framework offers a robust sensor-aware solution for complex multi-scale temporal modeling in industrial production systems. Full article

This study investigates whether CNN-based front-end feature extraction improves the predictive performance of deep learning models applied to 1 min intraday CSI 300 index data. Three baseline sequence models, LSTM, GRU, and TCN, are compared against their CNN hybrid and dual-branch fusion variants across five input window sizes, with all comparisons using identical back-end configurations. A total of 45 model configurations are trained and evaluated across 20 independent runs, with performance assessed on four metrics (MAE, RMSE, Directional Accuracy, and Information Coefficient) and statistical significance evaluated by paired t-tests. After standardisation, adding a CNN front-end does not consistently improve performance over the raw baseline and reduces IC for LSTM- and GRU-based models in many cases (e.g., IC of 0.0187 vs. 0.1031 for CNN-LSTM vs. LSTM at $W=1$), suggesting that standardised recurrent models can extract useful patterns directly from the raw sequence without CNN preprocessing. The dual-branch fusion architecture, which retains both the raw and CNN-compressed sequence branches, consistently outperforms the pure CNN hybrid on MAE, RMSE, and IC for LSTM- and GRU-based models (e.g., LSTMDualBranchFusion achieves statistically significant MAE reductions over CNN-LSTM at $W=1$, $W=2$, $W=4$, and $W=5$), indicating that the raw sequence carries complementary predictive information that the CNN front-end discards. TCN-based models produce near-zero or negative IC values regardless of architecture variant, suggesting a possible limitation of dilated convolutional architectures for return rank-ordering on this dataset and sample period. These findings are consistent across all five window sizes examined. Full article

The widespread deployment of unmanned aerial vehicles has raised growing privacy and security concerns, creating an urgent need for effective drone detection and identification. To address these challenges, this paper proposes a semantic learning framework termed Drone Time–Frequency-Semantic Network (DroneTFS-SemNet) for drone recognition. The main contributions are as follows. A Temporal Enhancement module is designed to improve model adaptability and robustness through radio frequency (RF) signal noise simulation, feature pre-extraction, learnable frequency-band partitioning, and an attention mechanism. A Cross-Channel Recalibration Module is introduced to promote cross-channel interaction of RF features via dilated convolutions and Channel Shuffle, thereby enhancing the model’s learning capability. Third, Feature Fusion Attention is employed to weight, align, and fuse RF features from different levels, thereby improving the model’s ability to represent differences among drone categories and operation modes. A semantic construction and decision module is further developed to extract high-level semantic features and accomplish drone detection, classification, and recognition. Experimental results on the DroneRF dataset show that DroneTFS-SemNet achieves a drone classification accuracy of 99.73% and an operation mode recognition accuracy of 99.65%, outperforming multiple comparison methods. These results demonstrate that the proposed model has strong recognition capability on this public dataset and under controlled experimental conditions, providing an effective method for RF-based drone recognition research. Full article

Transparent object segmentation plays a critical role in indoor and outdoor scene understanding, particularly driven by the rapid advancements in autonomous driving and robotics. However, this task presents significant challenges due to the lack of distinct texture and chromatic features in transparent objects, causing their appearance to blend into the background. Existing methods face inherent architectural limitations: CNNs are restricted by limited receptive fields, while Transformer-based methods may inadvertently suppress the weak feature details of transparent surfaces due to the inherent low-pass filtering property of self-attention mechanisms, treating them as background noise. Consequently, these approaches struggle to consistently segment transparent objects across diverse scales, failing to preserve both fine details and large-scale structures. To address these limitations, we propose the Global Context-Aware Transformer (GCA-Trans). Specifically, we design a Multi-scale Context Mining (MCM) module that leverages parallel dilated convolutions with varying receptive fields to simultaneously extract features at multiple scales. This design allows the model to capture and fuse fine-grained local details (e.g., edges and textures) with coarse-grained global spatial context (e.g., overall object shapes), ensuring robust segmentation performance for transparent objects of varying scales. Extensive experiments on four benchmark datasets demonstrate that GCA-Trans sets a new state of the art, achieving significant improvements of 2.53% mIoU on Trans10K-v2, 2.1% IoU on RGB-D GSD, 2.2% IoU on GDD, and 1.9% IoU on GSD, validating the effectiveness and robustness of our approach. Full article

Underground coal mine environments commonly suffer from insufficient illumination, high dust concentrations, and cluttered backgrounds, which substantially degrade the accuracy of conventional object detection algorithms. To address these issues, this paper proposes LLE-YOLO, a detection network built upon YOLOv11n. At the input stage, an Adaptive Low-Light Enhancement Module (ALEM) is introduced, which integrates Retinex decomposition, Contrast-Limited Adaptive Histogram Equalization (CLAHE), and brightness-dependent Gamma mapping to dynamically select the optimal enhancement strategy according to the global luminance. Furthermore, a Degradation-Aware Efficient Multi-Scale Attention (DEMA) module is proposed, which incorporates Contrast-Aware Modulation (CAM), an asymmetric dilated convolution group, and a Degradation-aware Spatial Gate (DSG) into the EMA channel-grouping and cross-spatial learning framework, thereby strengthening multi-scale personnel detection while keeping the parameter count tractable. On the publicly available DsDPM66 dataset, which covers 66 coal mine sites and 105,096 annotated images, LLE-YOLO achieves an mAP@0.5 of 83.7%, representing gains of 8.1 percentage points over YOLOv11n and 5.2 percentage points over the GCB-YOLOv11 baseline, while the recall increases from 71.2% to 78.2%. Under extremely dark scenarios (<30 lux), the mAP@0.5 is further improved by 15.3 percentage points. Ablation studies and Grad-CAM visualizations confirm the contribution of each module, offering a practical engineering reference for intelligent underground monitoring systems. Full article

Railway turnout systems are important components of railway signaling infrastructure, and timely fault diagnosis is essential for ensuring operational safety and maintenance efficiency. In practical applications, turnout fault diagnosis based on power data is often challenged by limited fault samples and severe class imbalance. To address these issues, this paper proposes a DLeNN-Attention-based fault diagnosis method for railway turnout power data, where DLeNN-Attention denotes Dilated-LeNet5-Attention. First, the original power sequences are standardized to a unified length through truncation, zero-padding, and normalization. Then, a hybrid data augmentation strategy combining the Synthetic Minority Over-sampling Technique (SMOTE) and a generative adversarial network (GAN) is adopted to enrich minority fault samples and alleviate class imbalance. Based on the augmented data, a DLeNN-Attention model is designed by integrating dilated convolution with the Convolutional Block Attention Module (CBAM), so as to capture richer temporal characteristics and enhance discriminative fault-related information. In this way, the proposed method can effectively learn representative features from turnout power data and improve fault classification performance. Experimental results on S700K turnout power data demonstrate that the proposed method achieves better diagnosis performance than several baseline models. The results indicate that the proposed method is effective for turnout fault diagnosis under limited-data conditions and shows promising application potential in intelligent health monitoring of railway turnout systems. Full article

As a critical component of sustainable energy systems, the efficient maintenance of photovoltaic (PV) panels is essential. While deep learning is an important approach for PV panel defect detection, the high complexity of existing models and their substantial computational demand make deployment on edge platforms difficult. This paper studies an acceleration method for photovoltaic panel defect detection on the Zynq-7020 heterogeneous platform. We design DARE-YOLO, a lightweight network for photovoltaic panel defect detection, together with a Zynq-based accelerator. In DARE-YOLO, we introduce RepConv and a lightweight single-path backbone to reduce the memory bandwidth overhead caused by multi-branch structures. We further design a Dilated Context Block (DCB) and a Dual-scale Decoupled Head (DDH), which effectively improve the detection accuracy of DARE-YOLO. On the Zynq platform, we develop the accelerator through a mixed fixed-point quantization strategy, a custom convolution IP core, and pipeline unrolling. These optimizations reduce data access latency, improve computational parallelism, and increase computational throughput. Experimental results show that DARE-YOLO achieves 93.84% mAP@0.5 with only 6.4 M parameters. The accelerator has a total on-board power consumption of only 1.95 W, while delivering a throughput of 37.5 GOPS, an energy efficiency of 19.23 GOPS/W. The image inference latency is 661.3 ms. This low-power, high-efficiency co-design paradigm ensures the long-term reliability of renewable energy facilities. Full article

To overcome the limitations of existing UAV object detection algorithms—particularly missed detections, false alarms, and the progressive loss of fine-grained features for small objects—this paper proposes SDP-YOLOv8, a lightweight and parameter-efficient enhancement of YOLOv8. The design aims to improve small-object detection accuracy while maintaining a lightweight architecture suitable for deployment on memory-constrained UAV platforms. Four lightweight-oriented modifications are introduced: (1) SCFS, which combines SPD-Conv for low-information-loss downsampling with a C2f block and SimAM attention; (2) DCSPPF, expanding the receptive field via parallel dilated convolutions; (3) a GhostConv-infused Patch Merging upsampling layer for local context enhancement; and (4) an extra small-scale detection head to preserve fine details. On VisDrone2019, experimental results show that SDP-YOLOv8 improved mAP@0.5 by 3.90% and mAP@0.5:0.95 by 2.60%, with a 14.4% reduction in parameters. The model maintains real-time performance (53.5 FPS on an RTX 3090 at FP32 with batch size 1, 38.7 FPS on a Jetson Orin Nano with TensorRT FP16 at batch size 1) and offers a favorable trade-off between detection accuracy, parameter efficiency, and memory footprint, making it a potential candidate for onboard deployment on resource-limited UAVs in aerial monitoring scenarios, pending further validation on diverse datasets and hardware platforms. Full article

Object detection in Unmanned Aerial Vehicle (UAV) imagery faces significant challenges due to the unique aerial perspective. A major bottleneck is the weak feature representation of small objects, which limits both detection accuracy and computational efficiency. To address this issue, we propose a Lightweight Detail-Sensitive Network (LDSNet). Specifically, LDSNet consists of three key components: (1) Lightweight Detail-Sensitive Downsampling (LDSDown), which combines anti-aliasing smoothing with dual-path feature extraction to preserve the spatial details of small objects during downsampling; (2) Shared Recursive Dilated Convolution (SRDC), which uses weight-shared multi-rate dilated convolutions to capture multi-scale context and enlarge the receptive field without introducing extra parameters; and (3) Deeply Decoupled Grouped Head (DGHead), which employs high-ratio grouped convolutions to significantly reduce the computational cost of processing high-resolution inputs. Extensive experiments on the VisDrone2019 and HIT-UAV datasets demonstrate that LDSNet achieves an excellent trade-off between accuracy and efficiency. Compared to the YOLOv11n baseline, LDSNet reduces parameters by 84.6% (from 2.6 M to 0.4 M) and FLOPs by 29.2% (from 6.5 G to 4.6 G), while improving mAP₅₀ by 2.2% on VisDrone2019 and achieving 94.5% on HIT-UAV. Full article

Short-term precipitation forecasting is an important research direction in meteorological studies, holding significant implications for disaster prevention and mitigation, urban flood drainage, and agricultural meteorological management. Existing deep learning models have achieved favourable results in modeling local features, yet they generally suffer from insufficient sensitivity to heavy precipitation areas, limitations in modeling temporal dependencies, and gradient instability issues. To address these limitations, we propose a novel spatiotemporal dual-branch neural network (ST-DualNet) for short-term precipitation forecasting based on radar echo maps. The network comprises a temporal branch (based on an enhanced ST-DConvLSTM) and a spatial branch (based on dilated convolutions and Transformer), respectively capturing the dynamic evolution and spatial structural features of precipitation. The two branches are integrated through the CBAM attention module and 3D convolution layer to achieve cross-branch feature fusion and prediction output. Experimental results demonstrate that ST-DualNet outperforms multiple mainstream models on the KNMI radar precipitation dataset, especially in heavy precipitation forecasting, providing an effective new framework for short-term precipitation forecasting. Full article

Farmland resources are fundamental to human survival and play a vital role in ensuring global food security. However, farmland boundary detection remains a significant technical challenge due to the low proportion of boundary pixels, multi-scale variations, and weak boundary continuity. To address these issues, this study proposes DCFENet, a dual-branch collaborative feature enhancement network. Specifically, a multi-scale feature fusion attention module TA-ASPP (Task-Aware Atrous Spatial Pyramid Pooling) is designed, which effectively enhances the network’s perception of farmland boundary features by integrating multi-scale dilated convolutions with skeleton-aware attention. In addition, a dual-branch decoding structure is proposed to enhance boundary localization and global topology modeling through boundary-aware gating and cross-branch feature fusion, thereby improving the boundary continuity. Furthermore, a collaborative constraint mechanism is proposed for dual-branch decoding, which supervises the two decoders using boundary loss and skeleton loss, thereby enhancing structural consistency and topology preservation. Experimental results demonstrate that DCFENet achieves precision, recall, and boundary IoU of 74.5%, 68.1%, and 77.4%, respectively, representing an improvement of 26.8%, 36.3%, and 13.2% compared with ResNet18_UNet. It also outperforms mainstream methods such as UNet, EdgeNAT, and EDTER. In terms of computational efficiency, DCFENet contains 26.43 M parameters and 37.43 G FLOPs, with a memory usage of 1.03 GB and an inference speed of 97.97 FPS, achieving a good balance between accuracy and efficiency. The results demonstrate the efficiency and accuracy of DCFENet in extracting farmland boundaries from high-resolution remote sensing images, providing technical support for farmland management and the advancement of precision and digital agriculture. Full article

Low-light enhancement technologies are of great significance for visual driver assistance applications and autonomous driving systems. Infrared vision can improve nighttime visibility but also faces challenges of low resolution and lack of color information. This paper presents a unified framework for RGB-guided infrared super-resolution and infrared-visible fusion that achieves high-resolution output under limited computational resources. Our approach employs a U-Net architecture with novel triple-grouped window attention (TGWA) encoding that captures global dependencies through grouped attention while reducing computational overhead, and adaptive multi-dilated convolutional (AMDC) decoding that adaptively selects optimal dilation rates using mixture-of-experts-inspired routing. Experiments on multiple datasets achieve competitive super-resolution and fusion results with minimal computational complexity, while real-world downstream object detection validation confirms robust performance in challenging nighttime scenarios. Quantitatively, the proposed method achieves 28.744 dB/0.872 SSIM on PBVS24 and 31.424 dB/0.882 SSIM on HDRT-Night for 8× infrared super-resolution, reaches competitive fusion quality on both MSRS and HDRT-Night, and attains 69.4% mAP@0.5 in downstream object detection on FLIR_aligned, while requiring only 1.12 M parameters and 85.44 G FLOPs. This work provides new possibilities for seeing clearly in the dark. Full article

This paper proposes a structure-constrained deep reconstruction framework for compressed sensing in shift-invariant spaces (SISs). The reconstruction is formulated as an inverse operator estimation problem derived from the matrix factorization $\mathbf{H}\left(\omega \right)=\mathbf{W}\left(\omega \right)\mathbf{A}$ and approximated using a hybrid CNN–Transformer architecture. Residual dilated convolutions capture localized signal structures, while the Transformer module models global frequency-domain dependencies. A variational inference-inspired regularization mechanism is incorporated to implicitly learn sparsity-aware priors. Experiments on both synthetic SIS signals and real-world ECG data demonstrate consistent improvements over classical optimization-based algorithms (ISTA, OMP) and a deep unfolding baseline (ISTA-Net+). At a 30% sampling rate, the proposed method achieves a PSNR of 35.46 dB. The feed-forward design eliminates iterative reconstruction, achieving a GPU inference time of 0.85 ms per signal. Full article

The scale features of rock art exhibit significant diversity and graduality. Among the existing semantic segmentation methods for rock art, although some models have taken note of the scale differences in rock art patterns and the complexity of directional features, and proposed targeted improvement strategies, most of these methods view scale adaptation and directional representation as unconnected problems. They fail to model the intrinsic correlation between the scale adaptation and directional representation, and particularly overlook the restrictive effect of scale accuracy on the extraction of directional features. This ultimately leads to the problem of “spatial representation misalignment” in the semantic segmentation of rock art. To address the above problems, this paper proposes a Dynamic Fine-tuning Rotation Network (DFTR-Net), which aims to solve the problems of imprecise scale feature extraction and directional misalignment for rock art patterns with arbitrary orientations. The network consists of a dynamic selective convolution structure and a shapeaware spatial feature extraction module. Specifically, the dynamic selective convolution dynamically adjusts the coverage range of the receptive field through inter-layer feature aggregation. It uses stacked small dilated convolution kernels to replace large convolution kernels with the same receptive field for extracting the neighborhood details of patterns. Then, by combining with feature aggregation, it constructs spatial feature differences and realizes intra-layer dynamic weighted fusion, thereby achieving accurate scale feature extraction. After obtaining fine-grained scale features, the shape-aware module first corrects the initial segmentation candidate regions of the patterns to generate directional guide boxes. Subsequently, it drives the rotational sampling of convolution kernels based on the angles of the guide boxes, forming region-constrained deformable convolutions that adapt to the shape of the patterns. These convolution kernels obtain strong supervision based on pixel-level annotations, which enhances the sensitivity to the directional features of the patterns and effectively alleviates the problem of directional misalignment. Extensive experiments show that DFTR-Net can achieve higher performance on the 3D-pitoti and Petroglyph Annotation datasets compared with the existing methods. Full article

Robust brain tumor classification from magnetic resonance imaging (MRI) remains challenging due to complex structural heterogeneity and subtle inter-class variability. Beyond predictive accuracy, conventional convolutional neural networks predominantly rely on texture-dominant features and fixed receptive fields, which may limit the extraction of clinically meaningful structural information. This study proposes a morphology-aware multi-scale deep representation learning framework that embeds morphological inductive bias directly within hierarchical feature extraction. The proposed architecture synergistically integrates trainable morphological operations with multi-scale convolutional feature learning inside a unified residual framework, supported by an in-block morphological refinement mechanism and a morphology-aware downsampling module. Unlike prior approaches that treat morphological operators as preprocessing or auxiliary branches, the proposed design incorporates differentiable dilation and erosion into the core feature hierarchy to guide structure-aware representation formation. The model was evaluated using five-fold cross-validation and an independent test set, achieving an overall test accuracy of 99.31% with consistently high macro-averaged precision, recall, F1-score, and AUC values. Grad-CAM analysis further demonstrates that the learned representations emphasize clinically relevant tumor regions, supporting interpretable structural knowledge extraction. Ablation studies confirm that performance improvements arise from the synergistic integration of multi-scale learning and morphology-aware refinement. Overall, embedding structural inductive bias within multi-scale deep representation learning enhances robustness, stability, and interpretable knowledge extraction for brain tumor MRI analysis. Full article