Search Results (57,133)

Marine mammal sound classification plays an important role in understanding species behavior, communication, and ecology. Automated classification methods have received increasing attention due to their ability to efficiently process and analyze large volumes of acoustic data. Traditional classification approaches often rely on frequency-domain representations, such as spectrograms, and image-based classifiers, which can be highly influenced by user-defined parameters. In this study, we investigate a classification method for marine mammal vocalizations using a one-dimensional convolutional neural network (1D CNN) that directly processes raw audio signals. The approach can handle signals of varying durations through a random cropping technique, minimizing signal distortion that is commonly introduced by conventional methods. The model was evaluated using marine mammal vocalization recordings obtained from the Watkins Marine Mammal Sound Database under three experimental scenarios. The results demonstrate the feasibility of using raw audio inputs with a 1D CNN for classifying marine mammal vocalizations with variable signal durations. Full article

Artificial levees along major rivers are critical for flood-risk mitigation, yet many aging structures have poorly constrained internal composition and material heterogeneity, limiting the reliability of conventional safety assessments. This study develops a quantitative, non-destructive framework for characterizing levee internal structure by integrating electrical resistivity tomography (ERT) with borehole (BH) observations. ERT profiles were combined with borehole measurements of grain size (D50) and water content to investigate subsurface compositional variability and to evaluate relationships between sedimentological and geophysical parameters. Grain-size data from borehole samples were modeled using four predictive approaches—random forest regression (RFR), artificial neural networks (ANN), linear regression (LR), and support vector regression (SVR)—based on ERT-derived resistivity and moisture information. The results reveal pronounced internal heterogeneity within the investigated levees and demonstrate consistent relationships between sediment composition, water content, and electrical resistivity. Among the tested models, the ensemble-based RFR provided the highest predictive performance (R² = 0.81). These findings indicate that D50 characteristics of levee materials can be reliably inferred from ERT data using machine learning, reducing the need for destructive sampling. The proposed approach offers a transferable methodology for levee assessment and supports future applications in non-destructive monitoring, spatially explicit flood-risk analysis, and climate-resilient flood-protection management. Full article

Tooth development or odontogenesis is a complex morphogenetic process that requires tightly regulated interactions between the oral epithelium and mesenchyme of neural crest origin. In this narrative review, we compile existing knowledge regarding gene regulatory networks and epigenetic factors throughout tooth development from initiation to eruption. Signaling between the epithelium and mesenchyme is mediated by four conserved pathways—Wnt/β-catenin, bone morphogenetic protein (BMP), fibroblast growth factor (FGF), and Sonic hedgehog (Shh)—which operate iteratively and interact through extensive crosstalk at each developmental stage. Transcription factors, such as PAX9, MSX1, PITX2, and LEF1, interpret these signals to control cell fate decisions and differentiation. Epigenetic modifications, including DNA methylation, histone modifications, and microRNA-mediated regulation, provide additional layers of control that fine-tune gene expression programs. Unlike existing reviews that address these regulatory mechanisms separately, here we integrate signaling pathways, transcription factor networks, epigenetic regulation, human genetic disorders, dental stem cell biology, and recent single-cell transcriptomic insights into a unified framework. We discuss opportunities to apply developmental biology knowledge towards regenerative dentistry goals, including iPSC-derived dental models and spatially resolved multi-omics approaches, while acknowledging the considerable gap between preclinical findings and clinical applications. Full article

A hybrid modelling framework is proposed to predict endpoint sulphur content in the ladle furnace (LF) refining process by embedding metallurgical expert knowledge into interpretable machine learning (ML). Industrial process data were extracted from the Level-2 (L2) system of a steel plant, and a desulphurisation dataset comprising 5169 heats with 29 process variables was constructed using a knowledge-guided time window from the joint satisfaction of refining conditions to the final argon-blowing stage. After data cleaning, normalisation and correlation-based feature selection, four algorithms—Random Forest (RF), Extreme Gradient Boosting (XGBoost), Support Vector Machine (SVM) and Artificial Neural Network (ANN)—were trained and compared on a representative cluster of steel grades identified by K-means. The ANN model achieved a coefficient of determination (R²) of 0.7752, a root mean square error (RMSE) of 0.0027 wt%, a mean absolute error (MAE) of 0.0017 wt% and a hit rate (HR, ±0.0025 wt% for S) of 76.40% on the test set. SHapley Additive exPlanations (SHAP) indicate that limestone addition, slag basicity, argon flow rate, refining time and initial sulphur content dominantly govern sulphur removal. The expert-knowledge-guided, interpretable framework provides quantitative support for specification-conforming endpoint sulphur control while mitigating over-desulphurisation and reagent consumption. Full article

Biodiversity monitoring in tropical ecosystems is constrained by limited infrastructure, insufficient localized datasets, and reliance on cloud-based tools. We introduce BEMAX, a lightweight convolutional neural network for offline classification of endangered tree species in the Bohol Biodiversity Complex, Philippines. A curated leaf-image dataset from five species and an unknown class was collected using a Raspberry Pi camera. The MobileNetV2-based model achieved a 93.89% validation accuracy and an 88.33% field accuracy. Deployed on a Raspberry Pi 4 with touchscreen and camera integration, BEMAX demonstrates embedded AI as a replicable framework for conservation in data-scarce environments. Full article

In recent decades, forest disturbances have increased in both frequency and intensity, driven by global warming and urbanization. Remote sensing, together with forest disturbance algorithms, offers broad opportunities for forest disturbance monitoring due to its high temporal and spatial resolution. However, operational methods capable of predicting and classifying disturbances while providing official area estimates suitable for national statistics remain scarce. The Three Indices Three Dimensions (3I3D) algorithm has proven effective in identifying forest changes and providing area estimates in Mediterranean ecosystems using Sentinel-2 imagery. Yet, while suitable for change detection, it does not distinguish among disturbance types. Here, we propose a two-step framework for forest disturbance detection and classification, tested in inland Spain for 2018. First, a binary forest change map is produced through an enhanced version of the 3I3D approach. This step incorporates Receiver Operating Characteristic (ROC) analysis to calibrate the algorithm through data-driven threshold selection, allowing adaptation to specific regional conditions. Second, detected changes are classified into four disturbance types: wildfire, clear-cut, thinning, and non-stand replacing disturbance, using Sentinel-2 spectral bands, 3I3D-derived metrics, and geometric descriptors of disturbance patches. Three machine-learning classifiers were compared: Support Vector Machine, Random Forest, and Neural Network. The detection step reached an overall accuracy of 82%, estimating that 1.43% of Spanish forests (264,900 ha) were disturbed in 2018. In the classification step, Random Forest achieved the best performance, with an overall accuracy of 72%. Of the detected disturbed area, 69% corresponded to non-stand replacing disturbances, while the remaining area was classified as thinnings (19%), wildfires (26%), and clear-cuts (55%). By integrating freely available Sentinel-2 imagery, remote sensing algorithms, and photo-interpreted reference datasets, this study provides a scalable and operational approach capable of producing annual disturbance maps that combine both detection and classification of high- and low-intensity disturbances, supporting official national-scale estimates of forest disturbance areas. Full article

Background: The recognition of autism spectrum disorder (ASD) has been a challenge due to the heterogeneity in symptoms and complex variations in brain function. Resting-state functional magnetic resonance imaging (rs-fMRI) has become instrumental in studying these disorders by accessing underlying abnormal neural activity and connectivity. Recently, deep learning approaches have shifted the analysis of brain networks by capturing spatiotemporal information from fMRI sequences. Nonetheless, most existing studies are limited by relying on a single representational scale, typically restricting analysis to either voxel-level spatiotemporal patterns or static connectivity matrices. Additionally, the dynamic reconfiguration of functional coupling and its variations across different anatomical parcellations are often ignored, which obscures neurobiologically meaningful dynamics. Methods: In this regard, we propose a multi-atlas dynamic connectivity transformer fused with 4D spatiotemporal modeling for ASD recognition (MADCT-4D). Specifically, the framework comprises two complementary branches. The 4D spatiotemporal branch encodes raw rs-fMRI volumes to learn hierarchical representations of evolving neural activity, while the dynamic-connectivity branch models time-resolved functional connectivity sequences constructed from multiple atlases, enabling the network to capture dynamic reconfiguration at the connectome level under different parcellation granularities. Moreover, we perform late fusion by combining the branch-specific decision scores with a learnable gate, allowing the model to adaptively weight voxel-level dynamics and multi-atlas connectivity evidence for each subject. Results: Extensive experiments on the publicly available ABIDE dataset demonstrate that the proposed method achieves 90.2% accuracy for ASD recognition, outperforming multiple competitive baselines. Conclusions: The proposed framework yields interpretable biomarkers based on learned dynamic connectivity patterns that are consistent with altered functional coupling in ASD. Full article

COVID-19 still poses a global public health challenge, exerting pressure on radiology services. Chest X-ray (CXR) imaging is widely used for respiratory assessment due to its accessibility and cost-effectiveness. However, its interpretation is often challenging because of subtle radiographic features and inter-observer variability. Although recent deep learning (DL) approaches have shown strong performance in automated CXR classification, their black-box nature limits interpretability. This study proposes an explainable deep learning framework for COVID-19 detection from chest X-ray images. The framework incorporates anatomically guided preprocessing, including lung-region isolation, contrast-limited adaptive histogram equalization (CLAHE), bone suppression, and feature enhancement. A novel four-channel input representation was constructed by combining lung-isolated soft-tissue images with frequency-domain opacity maps, vessel enhancement maps, and texture-based features. Classification was performed using a modified Xception-based convolutional neural network, while Gradient-weighted Class Activation Mapping (Grad-CAM) was employed to provide visual explanations and enhance interpretability. The framework was evaluated on the publicly available COVID-19 Radiography Database, achieving an accuracy of 95.3%, an AUC of 0.983, and a Matthews Correlation Coefficient of approximately 0.83. Threshold optimisation improved sensitivity, reducing missed COVID-19 cases while maintaining high overall performance. Explainability analysis showed that model attention was primarily focused on clinically relevant lung regions. Full article

Convolutional neural networks (CNNs) excel in tasks such as image, speech, and video recognition, as well as pattern analysis. However, their reliance on large training datasets, often sourced from third-party providers, exposes them to security risks, particularly poisoning attacks. Targeted poisoning attacks, also known as backdoor attacks, enable a CNN model to correctly classify normal data while misclassifying inputs containing specific triggers. In contrast, untargeted poisoning attacks aim to degrade the overall performance of the model. This research introduces an invisible targeted poisoning attack characterized by low implementation complexity and high computational efficiency due to its computationally inexpensive LSB-based embedding mechanism, without requiring complex optimization procedures against a basic CNN model and a residual network (ResNet-18) model. By embedding trigger images within poisoned samples, the attack remains covert, evading detection. The model is then trained on a dataset comprising both original and poisoned samples. The expected outcome is that the model will classify regular images correctly, but will misclassify those containing the embedded trigger as belonging to a target class. Experimental results on the CIFAR-10 dataset demonstrate the effectiveness of this approach, achieving a 99.32% Adversarial Success Rate (ASR) against ResNet-18 with only a 0.02% reduction in accuracy on benign test samples. Full article

Various digital-signal-processing algorithms are used to determine the refractive index based on the spectra of Tilted Fibre Bragg Gratings (TFBGs). Identifying new features of the optical spectrum improves estimations of the refractive index. New or modified demodulation algorithms influence measurement accuracy and resolution. In this study, we used signal-processing methods to determine the local and global features of TFBG spectra containing the so-called cladding mode comb. Based on these features, a demodulation method using artificial neural networks was created. The main novelty of this study is the simultaneous use of both local and global spectral features for refractive-index estimation. Currently, these two types of features are used separately. Here, the neural network is used for feature fusion obtained in the first step, consisting of signal-processing methods. Full article

Geometric problem-solving (GPS) has been a long-standing challenge in the fields of formal mathematics and artificial intelligence. To address the limitations of unidirectional approaches, we developed a neuro-symbolic system that integrates forward and backward reasoning. The neural component employs a gating-enhanced attention network to select candidate theorems, guiding the heuristic search and pruning irrelevant branches. The symbolic component is a bidirectional solver built on FormalGeo, which performs rigorous geometric relational reasoning and algebraic computation. The neural component predicts the theorems based on the current problem state, while the symbolic component applies these theorems and updates the problem state. These two parts interact iteratively until the problem is solved. The solving process is organized as a graph structure where facts and goals serve as nodes and theorems as edges, thereby generating a human-readable solution. The proposed neuro-symbolic system achieved an 89.63% problem-solving success rate (PSSR) on the FormalGeo7K dataset, surpassing the previous best result. Full article

Traditional sensory evaluation of traditional Chinese medicine (TCM) and medicinal and food homologous products has long relied on human observation of appearance, color, aroma, and taste. However, this approach is highly subjective, difficult to quantify, and often lacks reproducibility across evaluators. Intelligent sensory systems, including the electronic nose, electronic tongue, and machine vision, provide objective and digitized sensory information for TCM quality evaluation. Nevertheless, these platforms generate high-dimensional and heterogeneous datasets, creating a strong demand for efficient artificial intelligence (AI)-based analytical tools. This review summarizes recent advances in the application of machine learning and deep learning methods, such as support vector machine, random forest, convolutional neural network, and long short-term memory networks, for intelligent sensory evaluation of TCM. Particular emphasis is placed on how AI supports feature extraction, pattern recognition, classification, regression, and multisource data fusion across electronic nose, electronic tongue, and machine vision systems. Representative applications in raw material authentication, geographical origin discrimination, processing monitoring, and quality grading are also discussed. In addition, the current challenges related to data standardization, sensor drift, model robustness, and interpretability are highlighted. Overall, this review provides an integrated overview of AI-enabled intelligent sensory technologies and clarifies their potential to advance TCM quality evaluation toward a more objective, efficient, and holistic framework. Full article

A neural network simulation–regression method was developed to assess the impact of sanctions on the level of administrative offences under fragmented, noisy, and short administrative time series. The study addresses the problem of quantifying and predicting changes at the offence level as a sanction size function, using detection probability, prior violation level, compliance costs, and auxiliary contextual factors. The proposed framework combines a hybrid MLP–LSTM neural network, double machine learning-based orthogonal causal estimation, the simulation-based generation of counterfactual scenarios through domain randomization, multiple imputation for missing data, debiasing procedures, and ensemble uncertainty estimation. The contribution to administrative law consists of a quantitative tool creation for substantiating and optimising sanction policy, assessing heterogeneous effects, and supporting evidence-based rulemaking and law enforcement decisions. In comparative experiments, the developed method achieved an RMSE of 8…12%, a prediction accuracy of 93…96%, an overall accuracy of 95%, a precision of 94%, a recall of 93%, and an F1-score of 93.5%, thereby outperforming contemporary econometric, simulation, causal machine learning, and predictive machine learning approaches used for sanction effect modelling. Additional verification through nonparametric statistical testing cponfirmed that the proposed model’s superiority over the compared algorithms is statistically significant across the main quality metrics, which strengthens the evidence for its robustness and practical value in sanction policy analysis under fragmented administrative data conditions. Full article

The Landslide susceptibility assessment in complex mountainous terrain is frequently limited by static modelling frameworks that inadequately capture nonlinear deformation characteristics and temporally evolving hazard processes. To bridge this gap, a continuous-time dynamic assessment framework is proposed for Shazhou Town, Sichuan Province, integrating slowly moving scatterogram interferometric radar (S(BAS-InSAR))-derived deformation time series with Liquid Neural Networks (LNN). By incorporating a liquid time-constant architecture, the model accommodates irregular temporal sampling and captures non-stationary environmental responses through adaptive multimodal feature fusion. Analysis of long-term SBAS-InSAR observations (January 2021–May 2025) reveals distinctive deformation patterns, identifying eight active zones with maximum annual displacement rates of 107 mm yr⁻¹ and cumulative subsidence of 535.7 mm, which serve as critical dynamic inputs for the susceptibility model. Comparative experiments demonstrate that the LNN framework outperforms benchmark models (including LSTM, GRU, Random Forest, and SVM), achieving a coefficient of determination (R²) of 0.95 and an RMSE of 0.50. Furthermore, multi-temporal validation against 189 historical landslide records (2008–2025) confirms the model’s robustness, yielding a 91.5% capture rate within high-susceptibility zones. Interpretability analyses via SHAP and Layer-wise relevance propagation identify rainfall and vegetation cover as dominant dynamic controls, while characterising a distinct slope threshold effect at approximately 20°. These findings demonstrate that explicit continuous-time neural modelling enables physically consistent representation of irregular satellite acquisition intervals and delayed hydro-mechanical responses, thereby advancing landslide susceptibility assessment from static spatial classification toward dynamic state evolution inference under asynchronous Earth observation data streams. Full article

Accurate prediction of vibration responses in hydraulic structures during flood discharge is essential for ensuring safe and stable operation. This study develops a hybrid deep learning model that combines Convolutional Neural Networks (CNN), Bidirectional Long Short-Term Memory (BiLSTM), and a Channel Attention (CA) mechanism, optimized through Bayesian Optimization (BO), to predict dam gantry crane beam displacements. Time-lagged Pearson correlation and Maximum Information Coefficient (MIC) are applied to select the informative input features. The CNN-BiLSTM-CA model captures both spatial patterns and temporal dependencies in vibration signals. BO tunes model hyperparameters, while Partial Dependence (PD) analysis provides insight into how these parameters affect prediction accuracy. The model is validated using vibration data from an arch dam in Southwest China during flood discharge. Results show that CNN parameters have a greater impact on prediction accuracy than BiLSTM parameters, underscoring the importance of spatial feature extraction. Ablation studies confirm each component’s contribution. Compared with existing methods, the proposed model achieves superior accuracy with a Root Mean Square Error (RMSE) of 5.49, Mean Absolute Error (MAE) of 4.34, and correlation coefficient (R) of 99.42%. This framework provides a reliable and interpretable tool for predicting structural vibrations in hydraulic engineering under complex discharge conditions. Full article