Search Results (197)

It is a challenge to identify food waste sources in all-weather industrial environments, as variable lighting conditions can compromise the effectiveness of visual recognition models. This study proposes and validates a robust, interpretable, and adaptive multimodal logic fusion method in which sensor dominance is dynamically assigned based on real-time illuminance intensity. The method comprises two foundational components: (1) a lightweight MobileNetV3 + EMA model for image recognition; and (2) an audio model employing Fast Fourier Transform (FFT) for feature extraction and Support Vector Machine (SVM) for classification. The key contribution of this system lies in its environment-aware conditional logic. The image model MobileNetV3 + EMA achieves an accuracy of 99.46% within the optimal brightness range (120–240 cd m⁻²), significantly outperforming the audio model. However, its performance degrades significantly outside the optimal range, while the audio model maintains an illumination-independent accuracy of 0.80, a recall of 0.78, and an F1 score of 0.80. When light intensity falls below the threshold of 84 cd m⁻², the audio recognition results take precedence. This strategy ensures robust classification accuracy under variable environmental conditions, preventing model failure. Validated on an independent test set, the fusion method achieves an overall accuracy of 90.25%, providing an interpretable and resilient solution for real-world industrial deployment. Full article

The grid-forming static var generator (SVG) is a key device that supports the stable operation of power grids with a high penetration of renewable energy. The cooling efficiency of its forced water-cooling system directly determines the reliability of the entire unit. However, existing wired monitoring methods suffer from complex cabling and limited capacity to provide a full perception of the water-cooling condition. To address these limitations, this study develops a wireless monitoring system based on multi-source information fusion for real-time evaluation of cooling efficiency and early fault warning. A heterogeneous wireless sensor network was designed and implemented by deploying liquid-level, vibration, sound, and infrared sensors at critical locations of the SVG water-cooling system. These nodes work collaboratively to collect multi-physical field data—thermal, acoustic, vibrational, and visual information—in an integrated manner. The system adopts a hybrid Wireless Fidelity/Bluetooth (Wi-Fi/Bluetooth) networking scheme with electromagnetic interference-resistant design to ensure reliable data transmission in the complex environment of converter valve halls. To achieve precise and robust diagnosis, a three-layer hierarchical weighted fusion framework was established, consisting of individual sensor feature extraction and preliminary analysis, feature-level weighted fusion, and final fault classification. Experimental validation indicates that the proposed system achieves highly reliable data transmission with a packet loss rate below 1.5%. Compared with single-sensor monitoring, the multi-source fusion approach improves the diagnostic accuracy for pump bearing wear, pipeline micro-leakage, and radiator blockage to 98.2% and effectively distinguishes fault causes and degradation tendencies of cooling efficiency. Overall, the developed wireless monitoring system overcomes the limitations of traditional wired approaches and, by leveraging multi-source fusion technology, enables a comprehensive assessment of cooling efficiency and intelligent fault diagnosis. This advancement significantly enhances the precision and reliability of SVG operation and maintenance, providing an effective solution to ensure the safe and stable operation of both grid-forming SVG units and the broader power grid. Full article

Water quality monitoring faces increasing challenges due to accelerating industrialization and urbanization, demanding accurate, real-time, and reliable prediction technologies. This study presents a novel ensemble learning framework integrating Gaussian Process Regression, Support Vector Regression, and Random Forest algorithms for high-precision water quality pH prediction. The research utilized a comprehensive spatiotemporal dataset, comprising 11 water quality parameters from 37 monitoring stations across Georgia, USA, spanning 705 days from January 2016 to January 2018. The ensemble model employed a dynamic weight allocation strategy based on cross-validation error performance, assigning optimal weights of 34.27% to Random Forest, 33.26% to Support Vector Regression, and 32.47% to Gaussian Process Regression. The integrated approach achieved superior predictive performance, with a mean absolute error of 0.0062 and coefficient of determination of 0.8533, outperforming individual base learners across multiple evaluation metrics. Statistical significance testing using Wilcoxon signed-rank tests with a Bonferroni correction confirmed that the ensemble significantly outperforms all individual models (p < 0.001). Comparison with state-of-the-art models (LightGBM, XGBoost, TabNet) demonstrated competitive or superior ensemble performance. Comprehensive ablation experiments revealed that Random Forest removal causes the largest performance degradation (+4.43% MAE increase). Feature importance analysis revealed the dissolved oxygen maximum and conductance mean as the most influential predictors, contributing 22.1% and 17.5%, respectively. Cross-validation results demonstrated robust model stability with a mean absolute error of 0.0053 ± 0.0002, while bootstrap confidence intervals confirmed narrow uncertainty bounds of 0.0060 to 0.0066. Spatiotemporal analysis identified station-specific performance variations ranging from 0.0036 to 0.0150 MAE. High-error stations (12, 29, 33) were analyzed to distinguish characteristics, including higher pH variability and potential upstream pollution influences. An integrated software platform was developed featuring intuitive interface, real-time prediction, and comprehensive visualization tools for environmental monitoring applications. Full article

Cooperative task allocation is one of the critical enablers for multi-Autonomous Underwater Vehicle (AUV) missions, but existing approaches often assume reliable communication that rarely holds in real underwater acoustic environments. We study here the performance and robustness of the Consensus-Based Bundle Algorithm (CBBA) for multi-AUV task allocation under realistically degraded underwater communication conditions with dynamically appearing tasks. An integrated simulation framework that incorporates a Dubins-based kinematic model with minimum turning radius constraints, a configurable underwater acoustic communication model (range, delay, packet loss, and bandwidth), and a full implementation of improved CBBA with new features, complemented by 3D trajectory and network-topology visualization. We define five communication regimes, from ideal fully connected networks to severe conditions with short range and high packet loss. Within these regimes, we assess CBBA based on task allocation quality (total bundle value and task completion rate), convergence behavior (iterations and convergence rate), and communication efficiency (message delivery rate, average delay, and network connectivity), with additional metrics on the number of conflicts during dynamic task reallocation. Our simulation results indicate that CBBA maintains performance close to the optimum when the conditions are good and moderate but degrades significantly when connectivity becomes intermittent. We then introduce a local-communication-based conflict resolution strategy in the face of frequent task conflicts under very poor conditions: neighborhood-limited information exchange, negotiation within task areas, and decentralized local decisions. The proposed conflict resolution strategy significantly reduces the occurrence of conflicts and improves task completion under stringent communication constraints. This provides practical design insights for deploying multi-AUV systems under weak underwater acoustic networks. Full article

Manual pill identification is often inefficient and error-prone due to the large variety of medications and frequent visual similarity among pills, leading to misuse or dispensing errors. These challenges are exacerbated when pill imprints are engraved, curved, or irregularly arranged, conditions under which conventional optical character recognition (OCR)-based methods degrade significantly. To address this problem, we propose GO-PILL, a geometry-aware OCR pipeline for robust pill imprint recognition. The framework extracts text centerlines and imprint regions using the TextSnake algorithm. During imprint refinement, background noise is suppressed and contrast is enhanced to improve the visibility of embossed and debossed imprints. The imprint localization and alignment stage then rectifies curved or obliquely oriented text into a linear representation, producing geometrically normalized inputs suitable for OCR decoding. The refined imprints are processed by a multimodal OCR module that integrates a non-autoregressive language–vision fusion architecture for accurate character-level recognition. Experiments on a pill image dataset from the U.S. National Library of Medicine show that GO-PILL achieves an F1-score of 81.83% under set-based evaluation and a Top-10 pill identification accuracy of 76.52% in a simulated clinical scenario. GO-PILL consistently outperforms existing methods under challenging imprint conditions, demonstrating strong robustness and practical feasibility. Full article

Unmanned Aerial Vehicle (UAV)-captured images are easily affected by various degradations such as motion blur, noise, low illumination, haze, and raindrops in complex environments, and these degradations exhibit significant differences in the frequency domain. Existing all-in-one models typically operate in the spatial domain, making it difficult to effectively distinguish different degradation types, which leads to degradation interference and suboptimal restoration quality. To address these challenges, this paper proposes a degradation-aware prompt state space model for unified UAV image restoration. Specifically, we design a Prompt-Guided Mamba Block (PGMB) that injects dynamic degradation prompts into state space modeling, achieving differentiated global structure modeling. Meanwhile, we introduce an Adaptive Frequency Prompt Block (AFPB) that explicitly perceives frequency characteristics of diverse degradations and dynamically guides the reconstruction process via frequency-domain prompts, enabling collaborative restoration in both spatial and frequency domains. Extensive experiments demonstrate that the proposed method outperforms existing all-in-one approaches across multiple typical degradation tasks, significantly enhancing visual quality and robustness of UAV imagery. Full article

Online visual object tracking is a critical component of remote sensing-based aerial vehicle physical tracking, enabling applications such as environmental monitoring, target surveillance, and disaster response. In real-world remote sensing scenarios, the inherent processing delay of tracking algorithms results in the tracker’s output lagging behind the actual state of the observed scene. This latency not only degrades the accuracy of visual tracking in dynamic remote sensing environments but also impairs the reliability of UAV physical tracking control systems. Although predictive trackers have shown promise in mitigating latency impacts by forecasting target future states, existing methods face two key challenges in remote sensing applications: weak correlation between trackers and predictors, where predictions rely solely on motion information without leveraging rich remote sensing visual features; and inadequate modeling of continuous historical memory from discrete remote sensing data, limiting adaptability to complex spatiotemporal changes. To address these issues, we propose TPC-Tracker, a Tracker-Predictor Correlation Framework tailored for latency compensation in remote sensing-based aerial tracking. A Visual Motion Decoder (VMD) is designed to fuse high-dimensional visual features from remote sensing imagery with motion information, strengthening the tracker-predictor connection. Additionally, the Visual Memory Module (VMM) and Motion Memory Module (M3) model discrete historical remote sensing data into continuous spatiotemporal memory, enhancing predictive robustness. Compared with state-of-the-art predictive trackers, TPC-Tracker reduces the Mean Squared Error (MSE) by up to 38.95% in remote sensing-oriented physical tracking simulations. Deployed on VTOL drones, it achieves stable tracking of remote sensing targets at 80 m altitude and 20 m/s speed. Extensive experiments on public UAV remote sensing datasets and real-world remote sensing tasks validate the framework’s superiority in handling latency-induced challenges in aerial remote sensing scenarios. Full article

Real-time visual inference on resource-constrained hardware remains a core challenge for edge computing and embedded artificial intelligence systems. Recent deep learning architectures, particularly Vision Transformers (ViTs) and Detection Transformers (DETRs), achieve high detection accuracy but impose substantial computational and memory demands that limit their deployment on low-power edge platforms such as NVIDIA Jetson and Raspberry Pi devices. This paper presents a systematic review of model compression and optimization strategies—specifically quantization, pruning, and knowledge distillation—applied to lightweight object detection architectures for edge deployment. Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, peer-reviewed studies were analyzed from Scopus, IEEE Xplore, and ScienceDirect to examine the evolution of efficient detectors from convolutional neural networks to transformer-based models. The synthesis highlights a growing focus on real-time transformer variants, including Real-Time DETR (RT-DETR) and low-bit quantized approaches such as Q-DETR, alongside optimized YOLO-based architectures. While quantization enables substantial theoretical acceleration (e.g., up to 16× operation reduction), aggressive low-bit precision introduces accuracy degradation, particularly in transformer attention mechanisms, highlighting a critical efficiency-accuracy tradeoff. The review further shows that Quantization-Aware Training (QAT) consistently outperforms Post-Training Quantization (PTQ) in preserving performance under low-precision constraints. Finally, this review identifies critical open research challenges, emphasizing the efficiency–accuracy tradeoff and the high computational demands imposed by Transformer architectures. Future directions are proposed, including hardware-aware optimization, robustness to imbalanced datasets, and multimodal sensing integration, to ensure reliable real-time inference in practical agricultural edge computing environments. Full article

Symmetry and asymmetry are foundational design principles in artificial intelligence, defining the balance between invariance and adaptability in multimodal learning systems. In audio-visual speaker verification, where speech and lip-motion features are jointly modeled to determine whether two utterances belong to the same individual, these principles govern both fairness and discriminative power. In this work, we analyze how symmetry and asymmetry emerge within a gated-fusion architecture that integrates Time-Delay Neural Networks and Bidirectional Long Short-Term Memory encoders for speech, ResNet-based visual lip encoders, and a shared Conformer-based temporal backbone. Structural symmetry is preserved through weight-sharing across paired utterances and symmetric cosine-based scoring, ensuring verification consistency regardless of input order. In contrast, asymmetry is intentionally introduced through modality-dependent temporal encoding, multi-head attention pooling, and a learnable gating mechanism that dynamically re-weights the contribution of audio and visual streams at each timestep. This controlled asymmetry allows the model to rely on visual cues when speech is noisy, and conversely on speech when lip visibility is degraded, yielding adaptive robustness under cross-modal degradation. Experimental results demonstrate that combining symmetric embedding space design with adaptive asymmetric fusion significantly improves generalization, reducing Equal Error Rate (EER) to 3.419% on VoxCeleb-2 test dataset without sacrificing interpretability. The findings show that symmetry ensures stable and fair decision-making, while learnable asymmetry enables modality awareness together forming a principled foundation for next-generation audio-visual speaker verification systems. Full article

Ghost imaging (GI) offers a robust framework for remote sensing under degraded visibility conditions. However, atmospheric scattering in phenomena such as fog introduces significant noise and signal attenuation, thereby limiting its efficacy. Inspired by the selective attention mechanisms of biological visual systems, this study introduces a novel deep learning (DL) architecture that embeds a self-attention mechanism to enhance GI reconstruction in foggy environments. The proposed approach mimics neural processes by modeling both local and global dependencies within one-dimensional bucket measurements, enabling superior recovery of image details and structural coherence even at reduced sampling rates. Extensive simulations on the Modified National Institute of Standards and Technology (MNIST) and a custom Human-Horse dataset demonstrate that our bio-inspired model outperforms conventional GI and convolutional neural network-based methods. Specifically, it achieves Peak Signal-to-Noise Ratio (PSNR) values between 24.5–25.5$\mathrm{dB}/\mathrm{m}$ and Structural Similarity Index Measure (SSIM) values of approximately 0.8 under high scattering conditions ($\beta \ge 3.0 \mathrm{dB}/\mathrm{m}$) and moderate sampling ratios ($N \ge 50\%$). A comparative analysis confirms the critical role of the self-attention module, providing high-quality image reconstruction over baseline techniques. The model also maintains computational efficiency, with inference times under $0.12 \mathrm{s}$, supporting real-time applications. This work establishes a new benchmark for bio-inspired computational imaging, with significant potential for environmental monitoring, autonomous navigation and defense systems operating in adverse weather. Full article

This work presents a lightweight and interpretable framework for the early warning of voltage stability degradation in distribution networks, based on fractal and spectral features from flow measurements. We propose a Fast Voltage Stability Index (FVSI), which combines four independent indicators: the Detrended Fluctuation Analysis (DFA) exponent α (a proxy for long-term correlation), the width of the multifractal spectrum Δα, the slope of the spectral density β in the low-frequency range, and the $c_2$ curvature of multiscale structure functions. The indicators are calculated in sliding windows on per-node series of voltage in per unit Vpu and reactive power Q, standardized against an adaptive rolling/first-N baseline, and anomalies over time are accumulated using the Exponentially Weighted Moving Average (EWMA) and Cumulative SUM (CUSUM). A full online pipeline is implemented with robust preprocessing, automatic scaling, thresholding, and visualizations at the system level with an overview and heat maps and at the node level and panel graphs. Based on the standard IEEE 13-node scheme, we demonstrate that the Fractal Voltage Stability Index (FVSI_Fr) responds sensitively before reaching limit states by increasing α, widening Δα, a more negative $c_2$, and increasing β, locating the most vulnerable nodes and intervals. The approach is of low computational complexity, robust to noise and gaps, and compatible with real-time Phasor Measurement Unit (PMU)/Supervisory Control and Data Acquisition (SCADA) streams. The results suggest that FVSI_Fr is a useful operational signal for preventive actions (Q-support, load management/Photovoltaic System (PV)). Future work includes the calibration of weights and thresholds based on data and validation based on long field series. Full article

Coded Aperture Snapshot Spectral Imaging (CASSI) is a rapid hyperspectral imaging technique with broad application prospects. Due to limitations in three-dimensional compressed data acquisition modes and hardware constraints, the compressed measurements output by actual CASSI systems have a finite dynamic range, leading to degraded hyperspectral reconstruction quality. To address this issue, a high-quality hyperspectral reconstruction method based on multi-exposure fusion is proposed. A multi-exposure data acquisition strategy is established to capture low-, medium-, and high-exposure low-dynamic-range (LDR) measurements. A multi-exposure fusion-based high-dynamic-range (HDR) CASSI measurement reconstruction network (HCNet) is designed to reconstruct physically consistent HDR measurement images. Unlike traditional HDR networks for visual enhancement, HCNet employs a multiscale feature fusion architecture and combines local–global convolutional joint attention with residual enhancement mechanisms to efficiently fuse complementary information from multiple exposures. This makes it more suitable for CASSI systems, ensuring high-fidelity reconstruction of hyperspectral data in both spatial and spectral dimensions. A multi-exposure fusion CASSI mathematical model is constructed, and a CASSI experimental system is established. Simulation and real-world experimental results demonstrate that the proposed method significantly improves hyperspectral image reconstruction quality compared to traditional single-exposure strategies, exhibiting high robustness against multi-exposure interval jitters and shot noise in practical systems. Leveraging the higher-dynamic-range target information acquired through multiple exposures, especially in HDR scenes, the method enables reconstruction with enhanced contrast in both bright and dark details and also demonstrates higher spectral correlation, validating the enhancement of CASSI reconstruction and effective measurement capability in HDR scenarios. Full article

Exogenous fires in underground coal mines are characterized by low illumination, smoke occlusion, heavy dust loading and pseudo fire sources, which jointly degrade image quality and cause missed and false alarms in visual detection. To achieve accurate and real-time early warning under such conditions, this paper proposes a mine exogenous fire detection algorithm based on an improved YOLOv9m, termed PPL-YOLO-F-C. First, a lightweight PP-LCNet backbone is embedded into YOLOv9m to reduce the number of parameters and GFLOPs while maintaining multi-scale feature representation suitable for deployment on resource-constrained edge devices. Second, a Fully Connected Attention (FCAttention) module is introduced to perform fine-grained frequency–channel calibration, enhancing discriminative flame and smoke features and suppressing low-frequency background clutter and non-flame textures. Third, the original upsampling operators in the neck are replaced by the CARAFE content-aware dynamic upsampler to recover blurred flame contours and tenuous smoke edges and to strengthen small-object perception. In addition, an MPDIoU-based bounding-box regression loss is adopted to improve geometric sensitivity and localization accuracy for small fire spots. Experiments on a self-constructed mine fire image dataset comprising 3000 samples show that the proposed PPL-YOLO-F-C model achieves a precision of 97.36%, a recall of 84.91%, mAP@50 of 96.49% and mAP@50:95 of 76.6%, outperforming Faster R-CNN, YOLOv5m, YOLOv7 and YOLOv8m while using fewer parameters and lower computational cost. The results demonstrate that the proposed algorithm provides a robust and efficient solution for real-time exogenous fire detection and edge deployment in complex underground mine environments. Full article

The rapid development of intelligent transportation systems and autonomous driving technologies has made visual perception a key component in ensuring safety and improving efficiency in complex traffic environments. As a core task in visual perception, object detection directly affects the reliability of downstream modules such as path planning and decision control. However, adverse weather conditions (e.g., fog, rain, and snow) significantly degrade image quality—causing texture blurring, reduced contrast, and increased noise—which in turn weakens the robustness of traditional detection models and raises potential traffic safety risks. To address this challenge, this paper proposes an enhanced object detection framework, ECL-YOLOv11 (Edge-enhanced, Context-guided, and Lightweight YOLOv11), designed to improve detection accuracy and real-time performance under adverse weather conditions, thereby providing a reliable solution for in-vehicle perception systems. The ECL-YOLOv11 architecture integrates three key modules: (1) a Convolutional Edge-enhancement (CE) module that fuses edge features extracted by Sobel operators with convolutional features to explicitly retain boundary and contour information, thereby alleviating feature degradation and improving localization accuracy under low-visibility conditions; (2) a Context-guided Multi-scale Fusion Network (AENet) that enhances perception of small and distant objects through multi-scale feature integration and context modeling, improving semantic consistency and detection stability in complex scenes; and (3) a Lightweight Shared Convolutional Detection Head (LDHead) that adopts shared convolutions and GroupNorm normalization to optimize computational efficiency, reduce inference latency, and satisfy the real-time requirements of on-board systems. Experimental results show that ECL-YOLOv11 achieves mAP@50 and mAP@50–95 values of 62.7% and 40.5%, respectively, representing improvements of 1.3% and 0.8% over the baseline YOLOv11, while the Precision reaches 73.1%. The model achieves a balanced trade-off between accuracy and inference speed, operating at 237.8 FPS on standard hardware. Ablation studies confirm the independent effectiveness of each proposed module in feature enhancement, multi-scale fusion, and lightweight detection, while their integration further improves overall performance. Qualitative visualizations demonstrate that ECL-YOLOv11 maintains high-confidence detections across varying motion states and adverse weather conditions, avoiding category confusion and missed detections. These results indicate that the proposed framework provides a reliable and adaptable foundation for all-weather perception in autonomous driving systems, ensuring both operational safety and real-time responsiveness. Full article

Infrared remote sensing images are often degraded by blur and stripe noise caused by satellite attitude variations, optical distortions, and electronic interference, which significantly compromise image quality and target detection performance. Existing joint deblurring and destriping methods tend to over-smooth image edges and textures, failing to effectively preserve high-frequency details and sometimes misclassifying ringing artifacts as stripes. This paper proposes a variational framework for simultaneous deblurring and destriping of infrared remote sensing images. By leveraging an adaptive structure tensor model, the method exploits the sparsity and directionality of stripe noise, thereby enhancing edge and detail preservation. During blur kernel estimation, a fidelity term orthogonal to the stripe direction is introduced to suppress noise and residual stripes. In the image restoration stage, a WCOB (Non-blind restoration based on Wiener-Cosine composite filtering) model is proposed to effectively mitigate ringing artifacts and visual distortions. The overall optimization problem is efficiently solved using the alternating direction method of multipliers (ADMM). Extensive experiments on real infrared remote sensing datasets demonstrate that the proposed method achieves superior denoising and restoration performance, exhibiting strong robustness and practical applicability. Full article