Search Results (6,414)

Real-time, accurate and detailed monitoring of urban green space is of great significance for constructing the urban ecological environment and maximizing ecological benefits. Although high-resolution remote sensing technology provides rich ground object information, it also makes the surface information of urban green spaces more complex. Existing classification methods often struggle to meet the requirements of classification accuracy and the automation demands of high-resolution images. This study utilized GF-7 remote sensing imagery to construct an urban green space classification method for Beijing. The study used the YOLO v8 model as the framework to conduct a fine classification of urban green spaces within the Fifth Ring Road of Beijing, distinguishing between evergreen trees, deciduous trees, shrubs and grasslands. The aims were to address the limitations of insufficient model fit and coarse-grained classifications in existing studies, and to improve vegetation extraction accuracy for green spaces in northern temperate cities (with Beijing as a typical example). The results show that the overall classification accuracy of the trained YOLO v8 model is 89.60%, which is 25.3% and 28.8% higher than that of traditional machine learning methods such as Maximum Likelihood and Support Vector Machine, respectively. The model achieved extraction accuracies of 92.92%, 93.40%, 87.67%, and 93.34% for evergreen trees, deciduous trees, shrubs, and grasslands, respectively. This result confirms that the combination of deep learning and high-resolution remote sensing images can effectively enhance the classification extraction of urban green space vegetation, providing technical support and data guarantees for the refined management of green spaces and “garden cities” in megacities such as Beijing. Full article

“Hiding in Plain Sight” (HIPS) strategies for Personal Health Information (PHI) replace PHI with surrogate values to hinder re-identification attempts. We evaluate three different HIPS strategies for PHI replacement, a standard Consistent replacement strategy, a Random replacement strategy, and a novel Markov model strategy. We evaluate the privacy-preserving benefits and relative utility for information extraction of these strategies on both a simulated PHI distribution and real clinical corpora from two different institutions using a range of false negative error rates (FNER). The Markov strategy consistently outperformed the Consistent and Random substitution strategies on both real data and in statistical simulations. Using FNER ranging from 0.1% to 5%, PHI leakage at the document level could be reduced from 27.1% to 0.1% and from 94.2% to 57.7% with the Markov strategy versus the standard Consistent substitution strategy, at 0.1% and 0.5% FNER, respectively. Additionally, we assessed the generated corpora containing synthetic PHI for reuse using a variety of information extraction methods. Results indicate that modern deep learning methods have similar performance on all strategies, but older machine learning techniques can suffer from the change in context. Overall, a Markov surrogate generation strategy substantially reduces the chance of inadvertent PHI release. Full article

Recent developments in next-generation sequencing technology have led to the creation of extensive, open-source protein databases consisting of hundreds of millions of sequences. To render these sequences applicable in biomedical applications, they must be meticulously annotated by wet lab testing or extracting them from existing literature. Over the last few years, researchers have developed numerous automatic annotation systems, particularly deep learning models based on machine learning and artificial intelligence, to address this issue. In this work, we propose a transformer-based fusion model capable of predicting Gene Ontology (GO) terms from full-scale protein sequences, achieving state-of-the-art accuracy compared to other contemporary machine learning annotation systems. The approach performs particularly well on clustered split datasets, which comprise training and testing samples originating from distinct distributions that are structurally diverse. This demonstrates that the model is able to understand both short and long term dependencies within the protein’s structure and can capture sequence features that are predictive of the various GO terms. Furthermore, the technique is lightweight and less computationally expensive compared to the benchmark methods, while at the same time unaffected by sequence length, rendering it appropriate for diverse applications with varying sequence lengths. Full article

Visual inspection remains a cornerstone of quality control in welded structures, yet manual evaluations are inherently constrained by subjectivity, inconsistency, and limited scalability. This study presents WeldVGG, a deep learning-based visual inspection model designed to automate weld defect classification using radiographic imagery. The proposed model is trained on the RIAWELC dataset, a publicly available collection of X-ray weld images acquired in real manufacturing environments and annotated across four defect conditions: cracking, porosity, lack of penetration, and no defect. RIAWELC offers high-resolution imagery and standardized class labels, making it a valuable benchmark for defect classification under realistic conditions. To improve trust and explainability, Grad-CAM++ is employed to generate class-discriminative saliency maps, enabling visual validation of predictions. The model is rigorously evaluated through stratified cross-validation and benchmarked against traditional machine learning baselines, including SVC, Random Forest, and a state-of-the-art architecture, MobileNetV3. The proposed model achieves high classification accuracy and interpretability, offering a practical and scalable solution for intelligent weld inspection. Furthermore, to prove the model’s ability to generalize, a test on the GDXray was performed, yielding positive results. Additionally, a Wilcoxon signed-rank test was conducted separately to assess statistical significance between model performances. Full article

Plants vary in number of chromosomes (ploidy levels), which can influence morphological traits, including the size and density of stomata cells. Although biologists can detect these differences under a microscope, the process is often time-consuming and tedious. This study aims to automate the classification of blackcurrant (Ribes nigrum L.) ploidy levels—diploid, triploid, and tetraploid—by leveraging deep learning techniques. Convolutional Neural Networks and Vision Transformers are employed to perform microscopic image classification across two distinct blackcurrant datasets. Initial experiments demonstrate that these models can effectively classify ploidy levels when trained and tested on subsets derived from the same dataset. However, the primary challenge lies in proposing a model capable of yielding satisfactory classification results across different datasets ensuring robustness and generalization, which is a critical step toward developing a universal ploidy classification system. In this research, a variety of experiments is performed including application of augmentation technique. Model efficacy is evaluated with standard metrics and its interpretability is ensured through Gradient-weighted Class Activation Mapping visualizations. Finally, future research directions are outlined with application of other advanced state-of-the-art machine learning methods to further refine ploidy level prediction in botanical studies. Full article

This paper presents an automatic system for the classification of musical instruments from audio recordings. The project leverages deep learning (DL) techniques to achieve its objective, exploring three different classification approaches based on distinct input representations. The first method involves the extraction of Mel-Frequency Cepstral Coefficients (MFCCs) from the audio files, which are then fed into a two-dimensional convolutional neural network (Conv2D). The second approach makes use of mel-spectrogram images as input to a similar Conv2D architecture. The third approach employs conventional machine learning (ML) classifiers, including Logistic Regression, K-Nearest Neighbors, and Random Forest, trained on MFCC-derived feature vectors. To gain insight into the behavior of the DL model, explainability techniques were applied to the Conv2D model using mel-spectrograms, allowing for a better understanding of how the network interprets relevant features for classification. Additionally, t-distributed stochastic neighbor embedding (t-SNE) was employed on the MFCC vectors to visualize how instrument classes are organized in the feature space. One of the main challenges encountered was the class imbalance within the dataset, which was addressed by assigning class-specific weights during training. The results, in terms of classification accuracy, were very satisfactory across all approaches, with the convolutional models and Random Forest achieving around 97–98%, and Logistic Regression yielding slightly lower performance. In conclusion, the proposed methods proved effective for the selected dataset, and future work may focus on further improving class balance techniques. Full article

Land cover mapping is essential for territorial management due to its links with ecological, hydrological, climatic, and socioeconomic processes. Traditional methods use discrete classes per pixel, but this study proposes estimating cover fractions with Sentinel-2 imagery (20 m) and AI. We employed the French Land cover from Aerospace ImageRy (FLAIR) dataset (810 km² in France, 19 classes), with labels co-registered with Sentinel-2 to derive precise fractional proportions per pixel. From these references, we generated training sets combining spectral bands, derived indices, and auxiliary data (climatic and temporal variables). Various machine learning models—including XGBoost three deep neural network (DNN) architectures with different depths, and convolutional neural networks (CNNs)—were trained and evaluated to identify the optimal configuration for fractional cover estimation. Model validation on the test set employed RMSE, MAE, and R² metrics at both pixel level (20 m Sentinel-2) and scene level (100 m FLAIR). The training set integrating spectral bands, vegetation indices, and auxiliary variables yielded the best MAE and RMSE results. Among all models, DNN2 achieved the highest performance, with a pixel-level RMSE of 13.83 and MAE of 5.42, and a scene-level RMSE of 4.94 and MAE of 2.36. This fractional approach paves the way for advanced remote sensing applications, including continuous cover-change monitoring, carbon footprint estimation, and sustainability-oriented territorial planning. Full article

Dermatopathology has rapidly advanced due to the implementation of deep learning models and artificial intelligence (AI). From convolutional neural networks (CNNs) to transformer-based foundation models, these systems are now capable of accurate whole-slide analysis and multimodal integration. This review synthesizes the most recent advents of deep-learning architecture and synthesizes its evolution from first-generation CNNs to hybrid CNN-transformer systems to large-scale foundational models such as Paige’s PanDerm AI and Virchow. Herein, we examine performance benchmarks from real-world deployments of major dermatopathology deep learning models (DermAI, PathAssist Derm), as well as emerging next-generation models still under research and development. We assess barriers to clinical workflow adoption such as dataset bias, AI interpretability, and government regulation. Further, we discuss potential future research directions and emphasize the need for diverse, prospectively curated datasets, explainability frameworks for trust in AI, and rigorous compliance to Good Machine-Learning-Practice (GMLP) to achieve safe and scalable deep learning dermatopathology models that can fully integrate into clinical workflows. Full article

Lung cancer remains one of the most lethal cancers globally. Its early detection is vital to improving survival rates. In this work, we propose a hybrid computer-aided diagnosis (CAD) pipeline for lung cancer classification using Computed Tomography (CT) scan images. The proposed CAD pipeline integrates ten image preprocessing techniques and ten pretrained deep learning models for feature extraction including convolutional neural networks and transformer-based architectures, and four classical machine learning classifiers. Unlike traditional end-to-end deep learning systems, our approach decouples feature extraction from classification, enhancing interpretability and reducing the risk of overfitting. A total of 400 model configurations were evaluated to identify the optimal combination. The proposed approach was evaluated on the publicly available Lung Image Database Consortium and Image Database Resource Initiative dataset, which comprises 1018 thoracic CT scans annotated by four thoracic radiologists. For the classification task, the dataset included a total of 6568 images labeled as malignant and 4849 images labeled as benign. Experimental results show that the best performing pipeline, combining Contrast Limited Adaptive Histogram Equalization, Swin Transformer feature extraction, and eXtreme Gradient Boosting, achieved an accuracy of 95.8%. Full article

Bitcoin’s extreme price volatility presents significant challenges for investors and traders, necessitating accurate predictive models to guide decision-making in cryptocurrency markets. This study compares the performance of machine learning approaches for Bitcoin price prediction, specifically examining XGBoost gradient boosting, Long Short-Term Memory (LSTM), and GARCH-DL neural networks using comprehensive market data spanning December 2013 to May 2025. We employed extensive feature engineering incorporating technical indicators, applied multiple machine and deep learning models configurations including standalone and ensemble approaches, and utilized cross-validation techniques to assess model robustness. Based on the empirical results, the most significant practical implication is that traders and financial institutions should adopt a dual-model approach, deploying XGBoost for directional trading strategies and utilizing LSTM models for applications requiring precise magnitude predictions, due to their superior continuous forecasting performance. This research demonstrates that traditional technical indicators, particularly market capitalization and price extremes, remain highly predictive in algorithmic trading contexts, validating their continued integration into modern cryptocurrency prediction systems. For risk management applications, the attention-based LSTM’s superior risk-adjusted returns, combined with enhanced interpretability, make it particularly valuable for institutional portfolio optimization and regulatory compliance requirements. The findings suggest that ensemble methods offer balanced performance across multiple evaluation criteria, providing a robust foundation for production trading systems where consistent performance is more valuable than optimization for single metrics. These results enable practitioners to make evidence-based decisions about model selection based on their specific trading objectives, whether focused on directional accuracy for signal generation or precision of magnitude for risk assessment and portfolio management. Full article

Effective production control requires aligning strategic planning with real-time execution under dynamic and stochastic conditions. This study proposes a scalable dual-agent Deep Reinforcement Learning (DRL) framework for the joint optimisation of Work-In-Process (WIP) control and job sequencing in flow-shop environments. A strategic DQN agent regulates global WIP to meet throughput targets, while a tactical DQN agent adaptively selects dispatching rules at the machine level on an event-driven basis. Parameter sharing in the tactical agent ensures inherent scalability, overcoming the combinatorial complexity of multi-machine scheduling. The agents coordinate indirectly via a shared simulation environment, learning to balance global stability with local responsiveness. The framework is validated through a discrete-event simulation integrating agent-based modelling, demonstrating consistent performance across multiple production scales (5–15 machines) and process time variabilities. Results show that the approach matches or surpasses analytical benchmarks and outperforms static rule-based strategies, highlighting its robustness, adaptability, and potential as a foundation for future Hierarchical Reinforcement Learning applications in manufacturing. Full article

Soil organic matter (SOM) is essential for ecosystem health and agricultural productivity. Accurate prediction of SOM content is critical for modern agricultural management and sustainable soil use. Existing digital soil mapping (DSM) models, when processing temporal data, primarily focus on modeling the changes in input data across successive time steps. However, they do not adequately model the relationships among different input variables, which hinders the capture of complex data patterns and limits the accuracy of predictions. To address this problem, this paper proposes a novel deep learning model, 2-Channel Network (2C-Net), leveraging sequential multi-temporal remote sensing images to improve SOM prediction. The network separates input data into temporal and spatial data, processing them through independent temporal and spatial channels. Temporal data includes multi-temporal Sentinel-2 spectral reflectance, while spatial data consists of environmental covariates including climate and topography. The Multi-sequence Feature Fusion Module (MFFM) is proposed to globally model spectral data across multiple bands and time steps, and the Diverse Convolutional Architecture (DCA) extracts spatial features from environmental data. Experimental results show that 2C-Net outperforms the baseline model (CNN-LSTM) and mainstream machine learning model for DSM, with R² = 0.524, RMSE = 0.884 (%), MAE = 0.581 (%), and MSE = 0.781 (%)². Furthermore, this study demonstrates the significant importance of sequential spectral data for the inversion of SOM content and concludes the following: for the SOM inversion task, the bare soil period after tilling is a more important time window than other bare soil periods. 2C-Net model effectively captures spatiotemporal features, offering high-accuracy SOM predictions and supporting future DSM and soil management. Full article

This paper introduces the Gesture-Based Physical Stability Classification and Rehabilitation System (GPSCRS), a low-cost, non-invasive solution for evaluating physical stability using an Arduino microcontroller and the DFRobot Gesture and Touch sensor. The system quantifies movement smoothness, consistency, and speed by analyzing “up” and “down” hand gestures over a fixed period, generating a Physical Stability Index (PSI) as a single metric to represent an individual’s stability. The system focuses on a temporal analysis of gesture patterns while incorporating placeholders for speed scores to demonstrate its potential for a comprehensive stability assessment. The performance of various machine learning and deep learning models for gesture-based classification is evaluated, with neural network architectures such as Transformer, CNN, and KAN achieving perfect scores in recall, accuracy, precision, and F1-score. Traditional machine learning models such as XGBoost show strong results, offering a balance between computational efficiency and accuracy. The choice of model depends on specific application requirements, including real-time constraints and available resources. The preliminary experimental results indicate that the proposed GPSCRS can effectively detect changes in stability under real-time conditions, highlighting its potential for use in remote health monitoring, fall prevention, and rehabilitation scenarios. By providing a quantitative measure of stability, the system enables early risk identification and supports tailored interventions for improved mobility and quality of life. Full article

In the evaluation of vehicle noise, vibration and harshness (NVH) performance, interior noise control is the core consideration. In the early stage of automobile research and development, accurate prediction of interior noise caused by road surface is very important for optimizing NVH performance and shortening the development cycle. Although the data-driven machine learning method has been widely used in automobile noise research due to its advantages of no need for accurate physical modeling, data learning and generalization ability, it still faces the challenge of insufficient accuracy in capturing key local features, such as peaks, in practical NVH engineering. Aiming at this challenge, this paper introduces a forecast approach that utilizes an empirical-informed neural network, which aims to integrate a physical mechanism and a data-driven method. By deeply analyzing the transmission path of interior noise, this method embeds the acoustic mechanism features such as local peak and noise correlation into the deep neural network as physical constraints; therefore, this approach significantly enhances the model’s predictive performance. Experimental findings indicate that, in contrast to conventional deep learning techniques, this method is able to develop better generalization capabilities with limited samples, while still maintaining prediction accuracy. In the verification of specific models, this method shows obvious advantages in prediction accuracy and computational efficiency, which verifies its application value in practical engineering. The main contributions of this study are the proposal of an empirical-informed neural network that embeds vibro-acoustic mechanisms into the loss function and the introduction of an adaptive weight strategy to enhance model robustness. Full article

This work aims towards the automatic detection of faults in industrial air-cooling equipment used in a production line for staple fibers and ultimately provides maintenance scheduling recommendations to ensure seamless operation. In this context, various deep learning models are tested to ultimately define the most effective one for the intended scope. In the examined system, four vibration and temperature sensors are used, each positioned radially on the motor body near the rolling bearing of the motor shaft—a typical setup in many industrial environments. Thus, by collecting and using data from the latter sources, this work exhaustively investigates the feasibility of accurately diagnosing faults in staple fiber cooling fans. The dataset is acquired and constructed under real production conditions, including variations in rotational speed, motor load, and three fault priorities, depending on the model detection accuracy, product specification, and maintenance requirements. Fault identification for training purposes involves analyzing and evaluating daily maintenance logs for this equipment. Experimental evaluation on real production data demonstrated that the proposed ResNet50-1D model achieved the highest overall classification accuracy of 97.77%, while effectively resolving the persistent misclassification of the faulty impeller observed in all the other models. Complementary evaluation confirmed its robustness, cross-machine generalization, and suitability for practical deployment, while the integration of predictions with maintenance logs enables a severity-based prioritization strategy that supports actionable maintenance planning.deep learning; fault classification; industrial air-cooling; industrial automation; maintenance scheduling; vibration analysis Full article