Search Results (275)

The current study presents a multi-class, image-based classification of eight morphologically similar macroscopic Earthstar fungal species (Astraeus hygrometricus, Geastrum coronatum, G. elegans, G. fimbriatum, G. quadrifidum, G. rufescens, G. triplex, and Myriostoma coliforme) using deep learning and explainable artificial intelligence (XAI) techniques. For the first time in the literature, these species are evaluated together, providing a highly challenging dataset due to significant visual overlap. Eight different convolutional neural network (CNN) and transformer-based architectures were employed, including EfficientNetV2-M, DenseNet121, MaxViT-S, DeiT, RegNetY-8GF, MobileNetV3, EfficientNet-B3, and MnasNet. The accuracy scores of these models ranged from 86.16% to 96.23%, with EfficientNet-B3 achieving the best individual performance. To enhance interpretability, Grad-CAM and Score-CAM methods were utilised to visualise the rationale behind each classification decision. A key novelty of this study is the design of two hybrid ensemble models: EfficientNet-B3 + DeiT and DenseNet121 + MaxViT-S. These ensembles further improved classification stability, reaching 93.71% and 93.08% accuracy, respectively. Based on metric-based evaluation, the EfficientNet-B3 + DeiT model delivered the most balanced performance, with 93.83% precision, 93.72% recall, 93.73% F1-score, 99.10% specificity, a log loss of 0.2292, and an MCC of 0.9282. Moreover, this modeling approach holds potential for monitoring symbiotic fungal species in agricultural ecosystems and supporting sustainable production strategies. This research contributes to the literature by introducing a novel framework that simultaneously emphasises classification accuracy and model interpretability in fungal taxonomy. The proposed method successfully classified morphologically similar puffball species with high accuracy, while explainable AI techniques revealed biologically meaningful insights. All evaluation metrics were computed exclusively on a 10% independent test set that was entirely separate from the training and validation phases. Future work will focus on expanding the dataset with samples from diverse ecological regions and testing the method under field conditions. Full article

Non-Intrusive Load Monitoring (NILM) aims to disaggregate a household’s total aggregated power consumption into appliance-level usage, enabling intelligent energy management without the need for intrusive metering. While deep learning has improved NILM significantly, existing NILM models struggle to capture load patterns across both longer time intervals and subtle timings for appliances involving brief or overlapping usage patterns. In this paper, we propose a novel GRU+BERT hybrid architecture, exploring both unidirectional (GRU+BERT) and bidirectional (Bi-GRU+BERT) variants. Our model combines Gated Recurrent Units (GRUs) to capture sequential temporal dependencies with Bidirectional Encoder Representations from Transformers (BERT), which is a transformer-based model that captures rich contextual information across the sequence. The bidirectional variant (Bi-GRU+BERT) processes input sequences in both forward (past to future) and backward (future to past) directions, enabling the model to learn relationships between power consumption values at different time steps more effectively. The unidirectional variant (GRU+BERT) provides an alternative suited for appliances with structured, sequential multi-phase usage patterns, such as dishwashers. By placing the Bi-GRU or GRU layer before BERT, our models first capture local time-based load patterns and then use BERT’s self-attention to understand the broader contextual relationships. This design addresses key limitations of both standalone recurrent and transformer-based models, offering improved performance on transient and irregular appliance loads. Evaluated on the UK-DALE and REDD datasets, the proposed Bi-GRU+BERT and GRU+BERT models show competitive performance compared to several state-of-the-art NILM models while maintaining a comparable model size and training time, demonstrating their practical applicability for real-time energy disaggregation, including potential edge and cloud deployment scenarios. Full article

Gas-phase chemistry has been identified as a major computational bottleneck in both the forward and adjoint modes of chemical transport models (CTMs). Although previous studies have demonstrated the potential of deep learning models to simulate and accelerate this process, few studies have examined the applicability and performance of these models in adjoint sensitivity analysis. In this study, a deep learning emulator for gas-phase chemistry is developed and trained on a diverse set of forward-mode simulations from the Community Multiscale Air Quality (CMAQ) model. The emulator employs a residual neural network (ResNet) architecture referred to as FiLM-ResNet, which integrates Feature-wise Linear Modulation (FiLM) layers to explicitly account for photochemical and non-photochemical conditions. Validation within a single timestep indicates that the emulator accurately predicts concentration changes for 74% of gas-phase species with coefficient of determination (R²) exceeding 0.999. After embedding the emulator into the CTM, multi-timestep simulation over one week shows close agreement with the numerical model. For the adjoint mode, we compute the sensitivities of ozone (O₃) with respect to O₃, nitric oxide (NO), nitrogen dioxide (NO₂), hydroxyl radical (OH) and isoprene (ISOP) using automatic differentiation, with the emulator-based adjoint results achieving a maximum R² of 0.995 in single timestep evaluations compared to the numerical adjoint sensitivities. A 24 h adjoint simulation reveals that the emulator maintains spatially consistent adjoint sensitivity distributions compared to the numerical model across most grid cells. In terms of computational efficiency, the emulator achieves speed-ups of 80×–130× in the forward mode and 45×–102× in the adjoint mode, depending on whether inference is executed on Central Processing Unit (CPU) or Graphics Processing Unit (GPU). These findings demonstrate that, once the emulator is accurately trained to reproduce forward-mode gas-phase chemistry, it can be effectively applied in adjoint sensitivity analysis. This approach offers a promising alternative approach to numerical adjoint frameworks in CTMs. Full article

Single-phase ground (SPG) faults pose significant challenges in three-phase four-wire distribution systems due to their complex transient characteristics and the presence of multiple influencing factors. To solve the aforementioned issues, a comprehensive fault identification framework is proposed, which uses the TabNet deep learning architecture with hyperparameters optimized by Optuna. Firstly, a 10 kV simulation model is developed in Simulink to generate a diverse fault dataset. For each simulated fault, voltage and current signals from eight channels (L1–L4 voltage and current) are collected. Secondly, multi-domain features are extracted from each channel across time, frequency, waveform, and wavelet perspectives. Then, an attention-based fusion mechanism is employed to capture cross-channel dependencies, followed by L2-norm-based feature selection to enhance generalization. Finally, the optimized TabNet model effectively classifies 24 fault categories, achieving an accuracy of 97.33%, and outperforms baseline methods including Temporal Convolutional Network (TCN), Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM), Capsule Network with Sparse Filtering (CNSF), and Dual-Branch CNN in terms of accuracy, macro-F1 score, and kappa coefficient. It also exhibits strong stability and fast convergence during training. These results demonstrate the robustness and interpretability of the proposed method for SPG fault detection. Full article

Induction motors are a common component of electromechanical equipment in mining operations, yet they are susceptible to failures resulting from frequent start–stops, overloading, wear and tear, and component failure. It is evident that such failures can result in severe ramifications, encompassing industrial accidents and economic losses. The present paper proposes a detailed study of engine fault diagnosis technology. It has been demonstrated that prevailing intelligent engine diagnosis algorithms exhibit a limited diagnostic efficacy under variable operating conditions, and the reliability of diagnostic outcomes based on individual signals is questionable. The present paper puts forward the proposition of an investigation into a fault diagnosis algorithm for induction motors. This investigation utilized a range of analytical methods, including signal analysis, deep learning, transfer learning, and information fusion. Currently, the methods employed for fault diagnosis based on traditional machine learning are reliant on the selection of statistical features by those with expertise in the field, resulting in outcomes that are significantly influenced by human factors. This paper is the first to integrate a multi-branch ResNet strategy combining three-phase and single-phase currents. A range of three-phase current input strategies were developed, and a deep learning-based motor fault diagnosis model with adaptive feature extraction was established. This enables the deep residual network to extract fault depth features from the motor current signal more effectively. The experimental findings demonstrate that deep learning possesses the capacity to automatically extract depth features, thereby exceeding the capabilities of conventional machine learning algorithms with regard to the accuracy of motor fault diagnosis. Full article

Accurate estimation of winter wheat yield is essential for ensuring food security. Recent studies on winter wheat yield estimation based on deep learning methods rarely explore the interpretability of the model from the perspective of crop growth mechanism. In this study, a multiscale winter wheat yield estimation framework (called MultiScaleWheatNet model) was proposed, which was based on time series remote sensing data and further takes into account meteorological and soil factors that affect wheat growth. The model integrated multimodal data from different temporal and spatial scales, extracting growth characteristics specific to particular growth stage based on the growth pattern of wheat phenological phase. It focuses on enhancing model accuracy and interpretability from the perspective of crop growth mechanisms. The results showed that, compared to mainstream deep learning architectures, the MultiScaleWheatNet model had good estimation accuracy in both rain-fed and irrigated farmlands, with higher accuracy in rain-fed farmlands (R² = 0.86, RMSE = 0.15 t·ha⁻¹). At the county scale, the accuracy of the model in estimating winter wheat yield was stable across three years (from 2021 to 2023, R² ≥ 0.35, RMSE ≤ 0.73 t·ha⁻¹, nRMSE ≤ 20.4%). Model interpretability results showed that, taking all growth stages together, the remotely sensed indices had relatively high contribution to wheat yield, with roughly equal contributions from meteorological and soil variables. From the perspective of the growth stages, the contribution of LAI in remote sensing factors demonstrated greater stability throughout the growth stages, particularly during the jointing, heading-filling and milky maturity stage; the combined impact of meteorological factors exhibited a discernible temporal sequence, initially dominated by water availability and subsequently transitioning to temperature and sunlight in the middle and late stages; soil factors demonstrated a close correlation with soil pH and cation exchange capacity in the early and late stages, and with organic carbon content in the middle stage. By deeply combining remote sensing, meteorological and soil data, the framework not only achieves high accuracy in winter wheat yield estimation, but also effectively interprets the dynamic influence mechanism of remote sensing data on yield from the perspective of crop growth, providing a scientific basis for precise field water and fertiliser management and agricultural decision-making. Full article

Road extraction from high-resolution remote sensing images (HRSIs) provides critical support for downstream tasks such as autonomous driving path planning and urban planning. Although deep learning-based pixel-level segmentation methods have achieved significant progress, they still face challenges in handling occlusions caused by vegetation and shadows, and often exhibit limited model robustness and generalization capability. To address these limitations, this paper proposes the SAM2MS model, which leverages the powerful generalization capabilities of the foundational vision model, segment anything model 2 (SAM2). Firstly, an adapter-based fine-tuning strategy is employed to effectively transfer the capabilities of SAM2 to the HRSI road extraction task. Secondly, we subsequently designed a subtraction block (Sub) to process adjacent feature maps, effectively eliminating redundancy during the decoding phase. Multiple Subs are cascaded to form the multi-scale subtraction module (MSSM), which progressively refines local feature representations, thereby enhancing segmentation accuracy. During training, a training-free lossnet module is introduced, establishing a multi-level supervision strategy that encompasses background suppression, contour refinement, and region-of-interest consistency. Extensive experiments on three large-scale and challenging HRSI road datasets, including DeepGlobe, SpaceNet, and Massachusetts, demonstrate that SAM2MS significantly outperforms baseline methods across nearly all evaluation metrics. Furthermore, cross-dataset transfer experiments (from DeepGlobe to SpaceNet and Massachusetts) conducted without any additional training validate the model’s exceptional generalization capability and robustness. Full article

Partial discharge (PD) is a critical indicator of insulation degradation in high-voltage DC systems, necessitating accurate diagnosis to ensure long-term reliability. Conventional AC-based diagnostic methods, such as phase-resolved partial discharge analysis (PRPDA), are ineffective under DC conditions, emphasizing the need for waveform-based analysis. This study presents a novel clustering framework for DC PD pulses, leveraging multi-level wavelet decomposition and statistical feature extraction. Each signal is decomposed into multiple frequency bands, and 70 distinctive waveform features are extracted from each pulse. To mitigate feature redundancy and enhance clustering performance, principal component analysis (PCA) is employed for dimensionality reduction. Experimental data were obtained from multiple defect types and measurement distances using a 22.9 kV cross-linked polyethylene (XLPE) cable system. The proposed method significantly outperformed conventional time-frequency (T-F) mapping techniques, particularly in scenarios involving signal attenuation and mixed noise. Propagation-induced distortion was effectively addressed through multi-resolution analysis. In addition, field noise sources such as HVDC converter switching transients and fluorescent lamp emissions were included to assess robustness. The results confirmed the framework’s capability to distinguish between multiple PD types and noise sources, even in challenging environments. Furthermore, optimal mother wavelet selection and correlation-based feature analysis contributed to improved clustering resolution. This framework supports robust PD classification in practical HVDC diagnostics. The framework can contribute to the development of real-time autonomous monitoring systems for HVDC infrastructure. Future research will explore incorporating temporal deep learning architectures for automated PD-type recognition based on clustered data. Full article

In response to the technical challenges of traditional reservoir inversion techniques in determining inter-well wavelets and estimating geological statistical parameters, this study proposes an artificial intelligence high-resolution reservoir inversion technique based on seismic waveform constraints. This technology integrates multi-source heterogeneous data such as lithology characteristics, logging curves, and seismic waveforms through a deep learning neural network framework, and constructs an intelligent reservoir prediction model with geological and physical constraints. Results demonstrate that the proposed technique significantly enhances prediction accuracy for thin sand layers by effectively extracting high-frequency seismic information and establishing robust nonlinear mapping relationships. Inversion errors of reservoir parameters were reduced by more than 25%, while a vertical resolution of 0.5 m was achieved. Predictions agreed with actual drilling data with an accuracy of 86%, representing an 18% improvement over traditional methods. In practical applications, the technique successfully supported new well placement, contributing to a 22% increase in initial oil production in the pilot area. Furthermore, this study establishes a standardized technical procedure: “Time–Depth Modeling-Phase-Controlled Interpolation-Intelligent Inversion”. This workflow provides an innovative solution for high-precision reservoir characterization in regions with limited well control and complex terrestrial depositional systems, offering both theoretical significance and practical value for advancing reservoir prediction technology. Full article

Over Over the past years, the field of Machine Learning (ML) and Deep Learning (DL) has seen strong developments both in terms of software and hardware, with the increase of specialized devices. One of the biggest challenges in this field is the inference phase, where the trained model makes predictions of unseen data. Although computationally powerful, traditional computing architectures face limitations in efficiently managing requests, especially from an energy point of view. For this reason, the need arose to find alternative hardware solutions, and among these, there are Field Programmable Gate Arrays (FPGAs): their key feature of being reconfigurable, combined with parallel processing capability, low latency and low power consumption, makes those devices uniquely suited to accelerating inference tasks. In this paper, we present a novel approach to accelerate the inference phase of a multi-layer perceptron (MLP) using BondMachine framework, an OpenSource framework for the design of hardware accelerators for FPGAs. Analysis of the latency, energy consumption, and resource usage, as well as comparisons with respect to standard architectures and other FPGA approaches, is presented, highlighting the strengths and critical points of the proposed solution. The present work represents an exploratory study to validate the proposed methodology on MLP architectures, establishing a crucial foundation for future work on scalability and the acceleration of more complex neural network models. Full article

Precise segmentation of cell nuclei in pathological images is the foundation of cancer diagnosis and quantitative analysis, but blurred boundaries, scale variability, and staining differences have long constrained its reliability. To address this, this paper proposes AL-Net—an adaptive learning network that breaks through these bottlenecks through three innovative mechanisms: First, it integrates dilated convolutions with attention-guided skip connections to dynamically integrate multi-scale contextual information, adapting to variations in cell nucleus morphology and size. Second, it employs self-scheduling loss optimization: during the initial training phase, it focuses on region segmentation (Dice loss) and later switches to a boundary refinement stage, introducing gradient manifold constraints to sharpen edge localization. Finally, it designs an adaptive optimizer strategy, leveraging symbolic exploration (Lion) to accelerate convergence, and switches to gradient fine-tuning after reaching a dynamic threshold to stabilize parameters. On the 2018 Data Science Bowl dataset, AL-Net achieved state-of-the-art performance (Dice coefficient 92.96%, IoU 86.86%), reducing boundary error by 15% compared to U-Net/DeepLab; in cross-domain testing (ETIS/ColonDB polyp segmentation), it demonstrated over 80% improvement in generalization performance. AL-Net establishes a new adaptive learning paradigm for computational pathology, significantly enhancing diagnostic reliability. Full article

With the expansion of vehicle-to-everything (V2X) networks and the rising demand for intelligent services, vehicle edge computing encounters heightened requirements for more efficient task offloading. This study proposes a task offloading technique that utilizes federated collaboration and multi-agent deep reinforcement learning to reduce system latency and energy consumption. The task offloading issue is formulated as a Markov decision process (MDP), and a framework utilizing the Multi-Agent Dueling Double Deep Q-Network (MAD3QN) is developed to facilitate agents in making optimal offloading decisions inside intricate environments. Secondly, Federated Learning (FL) is implemented during the training phase, leveraging local training outcomes from many vehicles to enhance the global model, thus augmenting the learning efficiency of the agents. Experimental results indicate that, compared to conventional baseline algorithms, the proposed method decreases latency and energy consumption by at least 10% and 9%, respectively, while enhancing the average reward by at least 21%. Full article

In many scientific and engineering fields, such as hydrogeology, petroleum engineering, geotechnical research, and developing renewable energy solutions, fluid flow modeling in porous media is essential. In these areas, optimizing extraction techniques, forecasting environmental effects, and guaranteeing structural safety all depend on an understanding of the behavior of single-phase flows—fluids passing through connected pore spaces in rocks or soils. Darcy’s law, which results in an elliptic partial differential equation controlling the pressure field, is usually the mathematical basis for such modeling. Analytical solutions to these partial differential equations are seldom accessible due to the complexity and variability in natural porous formations, which makes the employment of numerical techniques necessary. To approximate subsurface flow solutions, traditional methods like the finite difference method, two-point flux approximation, and multi-point flux approximation have been employed extensively. Accuracy, stability, and computing economy are trade-offs for each, though. Deep learning techniques, in particular convolutional neural networks, physics-informed neural networks, and neural operators such as the Fourier neural operator, have become strong substitutes or enhancers of conventional solvers in recent years. These models have the potential to generalize across various permeability configurations and greatly speed up simulations. The purpose of this study is to examine and contrast the mentioned deep learning and numerical approaches to the problem of pressure distribution in single-phase Darcy flow, considering a 2D domain with mixed boundary conditions, localized sources, and sinks, and both homogeneous and heterogeneous permeability fields. The result of this study shows that the two-point flux approximation method is one of the best regarding computational speed and accuracy and the Fourier neural operator has potential to speed up more accurate methods like multi-point flux approximation. Different permeability field types only impacted each methods’ accuracy while computational time remained unchanged. This work aims to illustrate the advantages and disadvantages of each method and support the continuous development of effective solutions for porous medium flow problems by assessing solution accuracy and computing performance over a range of permeability situations. Full article

Pipelines play a vital role in material transportation within industrial settings. This review synthesizes detection technologies for early-stage small gas leaks from pipelines in the industrial sector, with a focus on acoustic-based methods, optical gas imaging (OGI), and multimodal fusion approaches. It encompasses detection principles, inherent challenges, mitigation strategies, and the state of the art (SOTA). Small leaks refer to low flow leakage originating from defects with apertures at millimeter or submillimeter scales, posing significant detection difficulties. Acoustic detection leverages the acoustic wave signals generated by gas leaks for non-contact monitoring, offering advantages such as rapid response and broad coverage. However, its susceptibility to environmental noise interference often triggers false alarms. This limitation can be mitigated through time-frequency analysis, multi-sensor fusion, and deep-learning algorithms—effectively enhancing leak signals, suppressing background noise, and thereby improving the system’s detection robustness and accuracy. OGI utilizes infrared imaging technology to visualize leakage gas and is applicable to the detection of various polar gases. Its primary limitations include low image resolution, low contrast, and interference from complex backgrounds. Mitigation techniques involve background subtraction, optical flow estimation, fully convolutional neural networks (FCNNs), and vision transformers (ViTs), which enhance image contrast and extract multi-scale features to boost detection precision. Multimodal fusion technology integrates data from diverse sensors, such as acoustic and optical devices. Key challenges lie in achieving spatiotemporal synchronization across multiple sensors and effectively fusing heterogeneous data streams. Current methodologies primarily utilize decision-level fusion and feature-level fusion techniques. Decision-level fusion offers high flexibility and ease of implementation but lacks inter-feature interaction; it is less effective than feature-level fusion when correlations exist between heterogeneous features. Feature-level fusion amalgamates data from different modalities during the feature extraction phase, generating a unified cross-modal representation that effectively resolves inter-modal heterogeneity. In conclusion, we posit that multimodal fusion holds significant potential for further enhancing detection accuracy beyond the capabilities of existing single-modality technologies and is poised to become a major focus of future research in this domain. Full article

Penetration depth is a critical factor determining joint strength in butt welding; however, it is difficult to monitor in keyhole-mode laser welding due to the dynamic nature of the keyhole. Recently, optical coherence tomography (OCT) has been introduced for real-time keyhole depth measurement, though accurate results require meticulous calibration. In this study, deep learning-based models were developed to estimate penetration depth in laser welding of 780 dual-phase (DP) steel. The models utilized coaxial weld pool images and spectrometer signals as inputs, with OCT signals serving as the output reference. Both uni-sensor models (based on coaxial pool images) and multi-sensor models (incorporating spectrometer data) were developed using transfer learning techniques based on pre-trained convolutional neural network (CNN) architectures including MobileNetV2, ResNet50V2, EfficientNetB3, and Xception. The coefficients of determination values (R²) of the uni-sensor CNN transfer learning models without fine-tuning ranged from 0.502 to 0.681, and the mean absolute errors (MAEs) ranged from 0.152 mm to 0.196 mm. In the fine-tuning models, R² decreased by more than 17%, and MAE increased by more than 11% compared to the previous models without fine-tuning. In addition, in the multi-sensor model, R² ranged from 0.900 to 0.956, and MAE ranged from 0.058 mm to 0.086 mm, showing better performance than uni-sensor CNN transfer learning models. This study demonstrated the potential of using CNN transfer learning models for predicting penetration depth in laser welding of 780DP steel. Full article