Search Results (2,132)

 Background/Objectives: Identifying and classifying abnormal lung sounds is essential for diagnosing patients with respiratory disorders. In particular, the simultaneous recording of auscultation signals from multiple clinically relevant positions offers greater diagnostic potential compared to traditional single-channel measurements. This study aims to improve the accuracy of respiratory sound classification by leveraging multichannel signals and capturing positional characteristics from multiple sites in the same patient. Methods: We evaluated the performance of respiratory sound classification using multichannel lung sound data with a deep learning model that combines a convolutional neural network (CNN) and long short-term memory (LSTM), based on mel-frequency cepstral coefficients (MFCCs). We analyzed the impact of the number and placement of channels on classification performance. Results: The results demonstrated that using four-channel recordings improved accuracy, sensitivity, specificity, precision, and F1-score by approximately 1.11, 1.15, 1.05, 1.08, and 1.13 times, respectively, compared to using three, two, or single-channel recordings. Conclusion: This study confirms that multichannel data capture a richer set of features corresponding to various respiratory sound characteristics, leading to significantly improved classification performance. The proposed method holds promise for enhancing sound classification accuracy not only in clinical applications but also in broader domains such as speech and audio processing.  Full article

Nutrition plays a fundamental role in promoting health and preventing chronic diseases, with bioactive food components offering a therapeutic potential in biomedical applications. Among these, edible oils are recognised for their functional properties, which contribute to disease prevention and metabolic regulation. The proposed study aims to evaluate the quality of four bioactive oils (olive oil, sunflower oil, tomato seed oil, and pumpkin seed oil) by analysing their thermal behaviour through infrared (IR) imaging. The study designed a customised electronic system to acquire thermographic signals under controlled temperature and humidity conditions. The acquisition system was used to extract thermal data. Analysis of the acquired thermal signals revealed characteristic heat absorption profiles used to infer differences in oil properties related to stability and degradation potential. A hybrid deep learning model that integrates Convolutional Neural Networks (CNNs) with Long Short-Term Memory (LSTM) units was used to classify and differentiate the oils based on stability, thermal reactivity, and potential health benefits. A signal analysis showed that the AI-based method improves both the accuracy (achieving an F1-score of 93.66%) and the repeatability of quality assessments, providing a non-invasive and intelligent framework for the validation and traceability of nutritional compounds. Full article

Traffic accidents pose a significant threat to public safety, resulting in numerous fatalities, injuries, and a substantial economic burden each year. The development of predictive models capable of the real-time forecasting of post-accident impact using readily available data can play a crucial role in preventing adverse outcomes and enhancing overall safety. However, existing accident predictive models encounter two main challenges: first, a reliance on either costly or non-real-time data, and second, the absence of a comprehensive metric to measure post-accident impact accurately. To address these limitations, this study proposes a deep neural network model known as the cascade model. It leverages readily available real-world data from Los Angeles County to predict post-accident impacts. The model consists of two components: Long Short-Term Memory (LSTM) and a Convolutional Neural Network (CNN). The LSTM model captures temporal patterns, while the CNN extracts patterns from the sparse accident dataset. Furthermore, an external traffic congestion dataset is incorporated to derive a new feature called the “accident impact” factor, which quantifies the influence of an accident on surrounding traffic flow. Extensive experiments were conducted to demonstrate the effectiveness of the proposed hybrid machine learning method in predicting the post-accident impact compared to state-of-the-art baselines. The results reveal a higher precision in predicting minimal impacts (i.e., cases with no reported accidents) and a higher recall in predicting more significant impacts (i.e., cases with reported accidents). Full article

Recently, deep learning algorithms have been increasingly applied in construction for activity recognition, particularly for excavators, to automate processes and enhance safety and productivity through continuous monitoring of earthmoving activities. These deep learning algorithms analyze construction videos to classify excavator activities for earthmoving purposes. However, previous studies have solely focused on single-source external videos, which limits the activity recognition capabilities of the deep learning algorithm. This paper introduces a novel multi-modal deep learning-based methodology for recognizing excavator activities, utilizing multi-stream input data. It processes point clouds and RGB images using the two-stream long short-term memory convolutional neural network (CNN-LSTM) method to extract spatiotemporal features, enabling the recognition of excavator activities. A comprehensive dataset comprising 495,000 video frames of synchronized RGB and point cloud data was collected across multiple construction sites under varying conditions. The dataset encompasses five key excavator activities: Approach, Digging, Dumping, Idle, and Leveling. To assess the effectiveness of the proposed method, the performance of the two-stream CNN-LSTM architecture is compared with that of single-stream CNN-LSTM models on the same RGB and point cloud datasets, separately. The results demonstrate that the proposed multi-stream approach achieved an accuracy of 94.67%, outperforming existing state-of-the-art single-stream models, which achieved 90.67% accuracy for the RGB-based model and 92.00% for the point cloud-based model. These findings underscore the potential of the proposed activity recognition method, making it highly effective for automatic real-time monitoring of excavator activities, thereby laying the groundwork for future integration into digital twin systems for proactive maintenance and intelligent equipment management. Full article

In the context of intelligent coal mine safety monitoring, an integrated prediction and early-warning method named MTGNN-Bayesian-IF-DBSCAN (Multi-Task Graph Neural Network–Bayesian Optimization–Isolation Forest–Density-Based Spatial Clustering of Applications with Noise) is proposed to address the challenges of gas concentration prediction and anomaly detection in coal mining faces. The MTGNN (Multi-Task Graph Neural Network) is first employed to model the spatiotemporal coupling characteristics of gas concentration and wind speed data. By constructing a graph structure based on sensor spatial dependencies and utilizing temporal convolutional layers to capture long short-term time-series features, the high-precision dynamic prediction of gas concentrations is achieved via the MTGNN. Experimental results indicate that the MTGNN outperforms comparative algorithms, such as CrossGNN and FourierGNN, in prediction accuracy, with the mean absolute error (MAE) being as low as 0.00237 and the root mean square error (RMSE) maintained below 0.0203 across different sensor locations (T0, T1, T2). For anomaly detection, a Bayesian optimization framework is introduced to adaptively optimize the fusion weights of IF (Isolation Forest) and DBSCAN (Density-Based Spatial Clustering of Applications with Noise). Through defining the objective function as the F1 score and employing Gaussian process surrogate models, the optimal weight combination (w_if = 0.43, w_dbscan = 0.52) is determined, achieving an F1 score of 1.0. By integrating original concentration data and residual features, gas anomalies are effectively identified by the proposed method, with the detection rate reaching a range of 93–96% and the false alarm rate controlled below 5%. Multidimensional analysis diagrams (e.g., residual distribution, 45° diagonal error plot, and boxplots) further validate the model’s robustness in different spatial locations, particularly in capturing abrupt changes and low-concentration anomalies. This study provides a new technical pathway for intelligent gas warning in coal mines, integrating spatiotemporal modeling, multi-algorithm fusion, and statistical optimization. The proposed framework not only enhances the accuracy and reliability of gas prediction and anomaly detection but also demonstrates potential for generalization to other industrial sensor networks. Full article

The converter valve is the core component of the ultra-high voltage direct current (UHVDC) transmission system, and its fault detection is very important to ensure the safe and stable operation of the transmission system. However, the voiceprint signals collected by converter stations under complex operating conditions are often affected by background noise, spikes, and nonlinear interference. Traditional methods make it difficult to achieve high-precision detection due to the lack of feature extraction ability and poor noise robustness. This paper proposes a fault-aware variational self-encoder model (FAVAE-AS) based on a weak correlation between attention and self-supervised learning. It extracts probability features via a conditional variational autoencoder, strengthens feature representation using multi-layer convolution and residual connections, and introduces a weak correlation attention mechanism to capture global time point relationships. A self-supervised learning module with six signal transformations improves generalization, while KL divergence-based correlation inconsistency quantization with dynamic thresholds enables accurate anomaly detection. Experiments show that FAVAE-AS achieves 0.925 accuracy in fault detection, which is 5% higher than previous methods, and has strong robustness. This research provides critical technical support for UHVDC system safety by addressing converter valve acoustic anomaly detection. It proposes an extensible framework for industrial intelligent maintenance. Full article

Accurate estimation of upper-limb joint angles is essential in biomechanics, rehabilitation, and wearable robotics. While inertial measurement units (IMUs) offer portability and flexibility, systems requiring multiple inertial sensors can be intrusive and complex to deploy. In contrast, optical motion capture (MoCap) systems provide precise tracking but are constrained to controlled laboratory environments. This study presents a deep learning-based approach for estimating shoulder and elbow joint angles using only three IMU sensors positioned on the chest and both wrists, validated against reference angles obtained from a MoCap system. The input data includes Euler angles, accelerometer, and gyroscope data, synchronized and segmented into sliding windows. Two recurrent neural network architectures, Convolutional Neural Network with Long-short Term Memory (CNN-LSTM) and Bidirectional LSTM (BLSTM), were trained and evaluated using identical conditions. The CNN component enabled the LSTM to extract spatial features that enhance sequential pattern learning, improving angle reconstruction. Both models achieved accurate estimation performance: CNN-LSTM yielded lower Mean Absolute Error (MAE) in smooth trajectories, while BLSTM provided smoother predictions but underestimated some peak movements, especially in the primary axes of rotation. These findings support the development of scalable, deep learning-based wearable systems and contribute to future applications in clinical assessment, sports performance analysis, and human motion research. Full article

A systems-oriented review of machine learning (ML) over the entire flood management spectrum, encompassing fluvial flood control, pluvial flood management, and resiliency-risk characterization was undertaken. Deep learners like long short-term memory (LSTM) networks perform well in predicting reservoir inflows and outflows. Convolution neural networks (CNNs) and other object identification algorithms are being explored in assessing levee and flood wall failures. The use of ML methods in pump station operations is limited due to lack of public-domain datasets. Reinforcement learning (RL) has shown promise in controlling low-impact development (LID) systems for pluvial flood management. Resiliency is defined in terms of the vulnerability of a community to floods. Multi-criteria decision making (MCDM) and unsupervised ML methods are used to capture vulnerability. Supervised learning is used to model flooding hazards. Conventional approaches perform better than deep learners and ensemble methods for modeling flood hazards due to paucity of data and large inter-model predictive variability. Advances in satellite-based, drone-facilitated data collection and Internet of Things (IoT)-based low-cost sensors offer new research avenues to explore. Transfer learning at ungauged basins holds promise but is largely unexplored. Explainable artificial intelligence (XAI) is seeing increased use and helps the transition of ML models from black-box forecasters to knowledge-enhancing predictors. Full article

This study evaluates machine learning-based methods for retrieving thin Arctic sea ice thickness (SIT) from L-band radiometry, using data from the European Space Agency’s (ESA) Soil Moisture and Ocean Salinity (SMOS) satellite. In addition to the operational ESA product, three alternative approaches are assessed: a Random Forest (RF) algorithm, a Convolutional Neural Network (CNN) that incorporates spatial coherence, and a Long Short-Term Memory (LSTM) neural network designed to capture temporal coherence. Validation against in situ data from the Beaufort Gyre Exploration Project (BGEP) moorings and the ESA SMOSice campaign demonstrates that the RF algorithm achieves robust performance comparable to the ESA product, despite its simplicity and lack of explicit spatial or temporal modeling. The CNN exhibits a tendency to overestimate SIT and shows higher dispersion, suggesting limited added value when spatial coherence is already present in the input data. The LSTM approach does not improve retrieval accuracy, likely due to the mismatch between satellite resolution and the temporal variability of sea ice conditions. These results highlight the importance of L-band sea ice emission modeling over increasing algorithm complexity and suggest that simpler, adaptable methods such as RF offer a promising foundation for future SIT retrieval efforts. The findings are relevant for refining current methods used with SMOS and for developing upcoming satellite missions, such as ESA’s Copernicus Imaging Microwave Radiometer (CIMR). Full article

The estimation of leakage faults in evaporation tubes of supercharged boilers is crucial for ensuring the safe and stable operation of the central steam system. However, leakage faults of evaporation tubes feature high time dependency, strong coupling among monitoring parameters, and interference from noise. Additionally, the large number of monitoring parameters (approximately 140) poses a challenge for spatiotemporal feature extraction, feature decoupling, and establishing a mapping relationship between high-dimensional monitoring parameters and leakage, rendering the precise quantitative estimation of evaporation tube leakage extremely difficult. To address these issues, this study proposes a novel deep learning framework (LSTM-CNN–attention), combining a Long Short-Term Memory (LSTM) network with a dual-pathway spatial feature extraction structure (ACNN) that includes an attention mechanism(attention) and a 1D convolutional neural network (1D-CNN) parallel pathway. This framework processes temporal embeddings (LSTM-generated) via a dual-branch ACNN—where the 1D-CNN captures local spatial features and the attention models’ global significance—yielding decoupled representations that prevent cross-modal interference. This architecture is implemented in a simulated supercharged boiler, validated with datasets encompassing three operational conditions and 15 statuses in the supercharged boiler. The framework achieves an average diagnostic accuracy (ADA) of over 99%, an average estimation accuracy (AEA) exceeding 90%, and a maximum relative estimation error (MREE) of less than 20%. Even with a signal-to-noise ratio (SNR) of −4 dB, the ADA remains above 90%, while the AEA stays over 80%. This framework establishes a strong correlation between leakage and multifaceted characteristic parameters, moving beyond traditional threshold-based diagnostics to enable the early quantitative assessment of evaporator tube leakage. Full article

Accurate dengue forecasting is essential for sustainable public health planning, especially in tropical regions where the disease remains a persistent threat. This study evaluates the predictive performance of seven modeling approaches—Seasonal Autoregressive Integrated Moving Average (SARIMA), Extreme Gradient Boosting (XGBoost), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), Convolutional LSTM (CNN–LSTM), and a Bayesian spatiotemporal model—using monthly dengue incidence data from 2009 to 2023 in Bandung City, Indonesia. Model performance was assessed using MAE, sMAPE, RMSE, and Pearson’s correlation (R). Among all models, the Bayesian spatiotemporal model achieved the best performance, with the lowest MAE (5.543), sMAPE (62.137), and RMSE (7.482), and the highest R (0.723). While SARIMA and XGBoost showed signs of overfitting, the Bayesian model not only delivered more accurate forecasts but also produced spatial risk estimates and identified high-risk hotspots via exceedance probabilities. These features make it particularly valuable for developing early warning systems and guiding targeted public health interventions, supporting the broader goals of sustainable disease management. Full article

Estimating heterogeneous treatment effects (HTEs) across multiple correlated outcomes poses significant challenges due to complex dependency structures and diverse data types. In this study, we propose a novel deep learning framework integrating empirical copula transformations with a CNN-LSTM (Convolutional Neural Networks and Long Short-Term Memory networks) architecture to capture nonlinear dependencies and temporal dynamics in multivariate treatment effect estimation. The empirical copula transformation, a rank-based nonparametric approach, preprocesses input covariates to better represent the underlying joint distributions before modeling. We compare this method with a baseline CNN-LSTM model lacking copula preprocessing and a nonparametric tree-based approach, the Causal Forest, grounded in generalized random forests for HTE estimation. Our framework accommodates continuous, count, and censored survival outcomes simultaneously through a multitask learning setup with customized loss functions, including Cox partial likelihood for survival data. We evaluate model performance under varying treatment perturbation rates via extensive simulation studies, demonstrating that the Empirical Copula CNN-LSTM achieves superior accuracy and robustness in average treatment effect (ATE) and conditional average treatment effect (CATE) estimation. These results highlight the potential of copula-based deep learning models for causal inference in complex multivariate settings, offering valuable insights for personalized treatment strategies. Full article

Underwater acoustic (UWA) communication supports many critical applications but still faces several physical-layer signal processing challenges. In response, recent advances in machine learning (ML) and deep learning (DL) offer promising solutions to improve signal detection, modulation adaptability, and classification accuracy. These developments highlight the need for a systematic evaluation to compare various ML/DL models and assess their performance across diverse underwater conditions. However, most existing reviews on ML/DL-based UWA communication focus on isolated approaches rather than integrated system-level perspectives, which limits cross-domain insights and reduces their relevance to practical underwater deployments. Consequently, this systematic literature review (SLR) synthesizes 43 studies (2020–2025) on ML and DL approaches for UWA communication, covering channel estimation, adaptive modulation, and modulation recognition across both single- and multi-carrier systems. The findings reveal that models such as convolutional neural networks (CNNs), long short-term memory networks (LSTMs), and generative adversarial networks (GANs) enhance channel estimation performance, achieving error reductions and bit error rate (BER) gains ranging from ${10}^{-3}$ to ${10}^{-6}$. Adaptive modulation techniques incorporating support vector machines (SVMs), CNNs, and reinforcement learning (RL) attain classification accuracies exceeding 98% and throughput improvements of up to 25%. For modulation recognition, architectures like sequence CNNs, residual networks, and hybrid convolutional–recurrent models achieve up to 99.38% accuracy with latency below 10 ms. These performance metrics underscore the viability of ML/DL-based solutions in optimizing physical-layer tasks for real-world UWA deployments. Finally, the SLR identifies key challenges in UWA communication, including high complexity, limited data, fragmented performance metrics, deployment realities, energy constraints and poor scalability. It also outlines future directions like lightweight models, physics-informed learning, advanced RL strategies, intelligent resource allocation, and robust feature fusion to build reliable and intelligent underwater systems. Full article

This study addresses the performance of deep learning models for predicting human DNA sequence classification through an exploration of ideal feature representation, model architecture, and hyperparameter tuning. It contrasts traditional machine learning with advanced deep learning approaches to ascertain performance with respect to genomic data complexity. A hybrid network combining long short-term memory (LSTM) and convolutional neural networks (CNN) was developed to extract long-distance dependencies as well as local patterns from DNA sequences. The hybrid LSTM + CNN model achieved a classification accuracy of 100%, which is significantly higher than traditional approaches such as logistic regression (45.31%), naïve Bayes (17.80%), and random forest (69.89%), as well as other machine learning models such as XGBoost (81.50%) and k-nearest neighbor (70.77%). Among deep learning techniques, the DeepSea model also accounted for good performance (76.59%), while others like DeepVariant (67.00%) and graph neural networks (30.71%) were relatively lower. Preprocessing techniques, one-hot encoding, and DNA embeddings were mainly at the forefront of transforming sequence data to a compatible form for deep learning. The findings underscore the robustness of hybrid structures in genomic classification tasks and warrant future research on encoding strategy, model and parameter tuning, and hyperparameter tuning to further improve accuracy and generalization in DNA sequence analysis. Full article

Speaker profiling systems are often evaluated on a single corpus, which complicates reliable comparison. We present a fully reproducible evaluation pipeline that trains Convolutional Neural Networks (CNNs) and Long-Short Term Memory (LSTM) models independently on three speech corpora representing distinct recording conditions—studio-quality TIMIT, crowdsourced Mozilla Common Voice, and in-the-wild VoxCeleb1. All models share the same architecture, optimizer, and data preprocessing; no corpus-specific hyperparameter tuning is applied. We perform a detailed preprocessing and feature extraction procedure, evaluating multiple configurations and validating their applicability and effectiveness in improving the obtained results. A feature analysis shows that Mel spectrograms benefit CNNs, whereas Mel Frequency Cepstral Coefficients (MFCCs) suit LSTMs, and that the optimal Mel-bin count grows with corpus Signal Noise Rate (SNR). With this fixed recipe, EfficientNet achieves 99.82% gender accuracy on Common Voice (+1.25 pp over the previous best) and 98.86% on VoxCeleb1 (+0.57 pp). MobileNet attains 99.86% age-group accuracy on Common Voice (+2.86 pp) and a 5.35-year MAE for age estimation on TIMIT using a lightweight configuration. The consistent, near-state-of-the-art results across three acoustically diverse datasets substantiate the robustness and versatility of the proposed pipeline. Code and pre-trained weights are released to facilitate downstream research. Full article