The implementation and application of artificial intelligence are propelling various advanced affective computing frameworks. Automatic recognition of emotions using multimodal physiological signals enhances the efficiency of systems such as health-care applications, pilot cognitive state monitoring, and passive brain-computer interfaces. However, challenges remain in capturing topological-frequency patterns in diverse electroencephalogram (EEG) electrode layouts, uncovering coupling dynamics across adjacent peripheral modalities, and integrating complementary affective patterns from the feature to modality levels. To address these issues, we propose a central-to-peripheral complementary integration network that employs hybrid encoders to extract and integrate affective patterns from EEG and peripheral signals. For EEG, the model unifies features from different channels into a single map to extract local-to-global representations, while for peripheral signals, adjacent cross-modal information is embedded into global affective patterns. These abstractions are systematically aggregated for emotion recognition by aligning affective relevance across domains and modalities within the central and peripheral nervous systems. The proposed model was evaluated on four publicly available multimodal databases using a leave-one-subject-out cross-validation approach. On the Database for Emotion Analysis using Physiological signal (DEAP), the binary recognition accuracy for valence and arousal scales was 75.00% and 77.33%, respectively. On the Human-Computer Interaction (HCI) database, the corresponding binary accuracies were 78.78% and 75.38%. For the SJTU Emotion EEG Datasets IV and V (SEED-IV and SEED-V), the four-class and five-class accuracies were 71.94% and 84.83%, respectively. These results validate the robustness and remarkable generalization capability of the proposed method.

Research on cognitive workload recognition via electroencephalogram and functional near-infrared spectroscopy has gained attention in brain-computer interfaces. However, complex acquisition environments cause inter-channel interference, limiting model accuracy. This paper proposes an improved dynamic graph attention channel selection method that uses graph attention network-derived scores to reduce noise and enhance robustness, alongside a hierarchical fusion module to address modality heterogeneity and a collaborative attention loss function to mitigate cross-modal differences. Experiments on three public datasets with two training strategies validate effective recognition in complex environments, with the channel selection method aiding cost reduction and exclusion of irrelevant channels. Copyright (c) 2025 The Authors. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)

We present a lightweight 3D Convolutional Neural Network (3D-CNN) to estimate physical parameters from fluid dynamics simulations modeled with Cellular Nonlinear Networks (CeNNs) exploiting the inherent spatio-temporal structure of the data. By training on small spatio-temporal patches structured as 4D tensors, the Convolutional Neural Network captures both spatial and temporal patterns using artificial intelligence models based on 3D convolution operations. Despite its simplicity, the network accurately infers the viscosity and density parameters for unseen data. The method generalizes well and, thanks to its relatively small size, could be suitable for real-time or experimental use. Limitations, such as handling chaotic behavior or long-range temporal dependencies, are discussed as areas for future improvement. Copyright (c) 2025 The Authors. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0/)

The rapid rise in inverter-based renewable resources has heightened concerns over subsynchronous resonance and oscillations, thereby challenging grid stability. This paper reviews approaches to identify and mitigate these issues, focusing on frequency scanning methods for stability assessment. It categorizes white-, black-, and gray-box modeling techniques, compares positive-sequence, dq-frame, and alpha-beta domain scanning, and examines perturbation shapes like step, ramp, and chirp. A comparative study highlights their strengths, limitations, and suitability for specific scenarios. By summarizing past events and surveying available tools, this work guides operators and researchers toward more effective, reliable stability analysis methods in grids with high renewable penetration.

In the field of affective computing, traditional methods for generating emotions predominantly rely on deep learning techniques and large-scale emotion datasets. However, deep learning techniques are often complex and difficult to interpret, and standardizing large-scale emotional datasets are difficult and costly to establish. To tackle these challenges, we introduce a novel framework named Audio-Visual Fusion for Brain-like Emotion Learning(AVF-BEL). In contrast to conventional brain-inspired emotion learning methods, this approach improves the audio-visual emotion fusion and generation model through the integration of modular components, thereby enabling more lightweight and interpretable emotion learning and generation processes. The framework simulates the integration of the visual, auditory, and emotional pathways of the brain, optimizes the fusion of emotional features across visual and auditory modalities, and improves upon the traditional Brain Emotional Learning (BEL) model. The experimental results indicate a significant improvement in the similarity of the audio-visual fusion emotion learning generation model compared to single-modality visual and auditory emotion learning and generation model. Ultimately, this aligns with the fundamental phenomenon of heightened emotion generation facilitated by the integrated impact of visual and auditory stimuli. This contribution not only enhances the interpretability and efficiency of affective intelligence but also provides new insights and pathways for advancing affective computing technology. Our source code can be accessed here: https://github.com/OpenHUTB/emotion}{https://github.com/OpenHUTB/emotion.

Emotional responses to stimuli produce distinct brain activity patterns that are often sparse in time and spatial distribution across the cortex. These neural signals also contain individual-specific features, complicating emotion recognition across diverse populations. Current approaches rarely address the dual challenge of capturing sparse emotional patterns while minimizing identity-related biases in individual-generic emotion analysis. To bridge this gap, we propose a graph-based emotion-enhancing network framework that isolates emotion-specific neural signatures by amplifying sparse temporal-spatial features and suppressing person-specific biomarkers. Evaluated on two benchmark databases for binary emotion classification, our model achieved state-of-the-art performance in individual-dependent scenarios with accuracies of 65.76 % and 65.39 % for the arousal scale, and 57.75 % and 66.74 % for the valence scale. In the individual-generic condition, the accuracies were 56.11 % and 61.02 % for arousal, and 55.21 % and 66.17 % for valence. Notably, the model's temporal and spatial enhancement modules provide interpretable insights into emotion-related neural sparsity through learned feature weights. This framework advances emotion recognition systems by reliably identifying universal emotional patterns across individuals while improving computational generalizability.

As power systems become more complex with the continuous integration of intelligent distributed energy resources (DERs), new risks and uncertainties arise. Consequently, to enhance system resiliency, it is essential to account for various uncertain events when implementing the optimization problem for the energy management system (EMS). This paper presents a preventive EMS considering the probability of failure (PoF) of each system component across different scenarios. A conditional-value-at-risk (CVaR)-based framework is proposed to integrate the uncertainties of the distribution network. Loads are classified into critical, semi-critical, and non-critical categories to prioritize essential loads during generation resource shortages. A proximal policy optimization (PPO)-based reinforcement learning (RL) agent is used to solve the formulated problem and generate the control decisions. The proposed framework is evaluated on a notional MVDC ship system and a modified IEEE 30-bus system, where the results demonstrate that the PPO agent can successfully optimize the objective function while maintaining the network and operational constraints. For validation, the RL-based method is benchmarked against a traditional optimization approach, further highlighting its effectiveness and robustness. This comparison shows that RL agents can offer more resiliency against future uncertain events compared to the traditional solution methods due to their adaptability and learning capacity.

Accurate disease recognition through computer vision is crucial for the intelligent management of potato production. Popular data-driven classification methods face challenges including limited labelled data and poor model portability. Unsupervised Domain Adaptation (UDA) addresses these challenges with a novel learning strategy. However, the complex field environment introduces a significant domain shift problem due to varying conditions. Existing UDA methods usually concentrate on aligning global data distribution and employ a single structure for disease feature extraction, thereby limiting their efficacy in true field environment. To tackle this challenge of potato disease recognition, the Multi-Representation Adaptive Network (MRSAN) based on subdomain alignment is presented. MRSAN effectively aligns feature distributions across diverse data by minimising distribution differences among relevant subdomains. Simultaneously, the multi-representation extraction method captures finer details from various perspectives in the disease images. The combination of these two approaches efficiently mitigates the adverse effects caused by various interference factors in field environment. Based on the acquisition conditions of light variation and disease progression, two field potato disease image datasets are created, containing five and six kinds of potato leaf disease, respectively. Extensive transfer experiments are conducted on the two datasets. MRSAN achieves average classification accuracies of 87.03% and 80.06% on the datasets for the corresponding transfer tasks, outperforming the other compared methods. This not only validates the effectiveness of MRSAN but also demonstrates its robust ability to generalise across changes in regard to light variation and disease progression.

Diseases cause crop yield reduction and quality decline, which has a great impact on agricultural production. Plant disease recognition based on computer vision can help farmers quickly and accurately recognize diseases. However, the occurrence of diseases is random and the collection cost is very high. In many cases, the number of disease samples that can be used to train the disease classifier is small. To address this problem, we propose a few-shot disease recognition algorithm that uses supervised contrastive learning. Our algorithm is divided into two phases: supervised contrastive learning and meta-learning. In the first phase, we use a supervised contrastive learning algorithm to train an encoder with strong generalization capabilities using a large number of samples. In the second phase, we treat this encoder as an extractor of plant disease features and adopt the meta-learning training mechanism to accomplish the few-shot disease recognition tasks by training a nearest-centroid classifier based on distance metrics. The experimental results indicate that the proposed method outperforms the other nine popular few-shot learning algorithms as a comparison in the disease recognition accuracy over the public plant disease dataset PlantVillage. In few-shot potato leaf disease recognition tasks in natural scenarios, the accuracy of the model reaches the accuracy of 79.51% with only 30 training images. The experiment also revealed that, in the contrastive learning phase, the combination of different image augmentation operations has a greater impact on model. Furthermore, the introduction of label information in supervised contrastive learning enables our algorithm to still obtain high accuracy in few-shot disease recognition tasks with smaller batch size, thus allowing us to complete the training with less GPU resource compared to traditional contrastive learning.

In this paper, the adaptive practical finite-time tracking control problem for a class of strictly feedback nonlinear systems with multiple actuator constraints is investigated using backstepping techniques and practical finite-time stability theory. The effects of deadband and saturated nonlinear constraints on the controller design of nonlinear systems are addressed by the equivalent transformation method. The problem of complexity explosion due to the derivatives of virtual control signals is solved by using the virtual control signals as inputs to the command filters and using the outputs of the command filters to perform the corresponding control tasks. An adaptive neural network tracking backstepping control strategy based on the command filter technique and the backstepping design algorithm is proposed by approximating an unknown nonlinear function using a neural network. The control strategy ensures the boundedness of all variables in the closed-loop system, and the output tracking error fluctuates in a small region near the origin. Finally, simulations verify the effectiveness of the control strategy designed in this paper.

The lack of complementary affective responses from both the central and peripheral nervous systems could limit the performance of emotion recognition with the single-modal physiological signal. However, when integrating multimodalities, a direct fusion may ignore the heterogeneous nature of multiple feature domains from one modality to another. Besides, there is a risk that the distribution of the multimodal physiological responses may vary across different affective scenarios for stimulating an identical emotional category. The inter-individual variation may also increase due to the superposition of the biometric information from the multimodal features. To tackle these issues, we present a hierarchical multimodal network for robust heterogeneous physiological representations (RHPRNet). First, we applied a spatial-frequency pattern extractor to identify the electroencephalogram (EEG) representations in both the spatial and frequency domains. Next, inter-domain and inter-modality affective encoders are separately applied to the statistic-complexity EEG features and multimodal peripheral features, respectively. All the learned representations are integrated via a hierarchical fusion module. To model the multi-peak patterns stimulated by different affective scenarios, we designed a scenario-adapting pretraining stage. A random contrastive training loss was also applied to mitigate the inter-individual variance. In the end, we performed adequate experiments to examine the performance of the RHPRNet based on three publicly available multimodal databases combined with two validation approaches.

Pulmonary fat embolism (PFE) as a cause of death often occurs in trauma cases such as fractures and soft tissue contusions. Traditional PFE diagnosis relies on subjective methods and special stains like oil red O. This study utilizes computational pathology, combining digital pathology and deep learning algorithms, to precisely quantify fat emboli in whole slide images using conventional hematoxylin-eosin (H&E) staining. The results demonstrate deep learning’s ability to identify fat droplet morphology in lung microvessels, achieving an area under the receiver operating characteristic (ROC) curve (AUC) of 0.98. The AI-quantified fat globules generally matched the Falzi scoring system with oil red O staining. The relative quantity of fat emboli against lung area was calculated by the algorithm, determining a diagnostic threshold of 8.275% for fatal PFE. A diagnostic strategy based on this threshold achieved a high AUC of 0.984, similar to manual identification with special stains but surpassing H&E staining. This demonstrates computational pathology’s potential as an affordable, rapid, and precise method for fatal PFE diagnosis in forensic practice.

Background and Objective: Diatom testing is supportive for drowning diagnosis in forensic medicine. How-ever, it is very time-consuming and labor-intensive for technicians to identify microscopically a handful of diatoms in sample smears, especially under complex observable backgrounds. Recently, we success-fully developed a software, named DiatomNet v].0 intended to automatically identify diatom frustules in a whole slide under a clear background. Here, we introduced this new software and performed a val-idation study to elucidate how DiatomNet v].0 improved its performance with the influence of visible impurities.Methods: DiatomNet v] .0 has an intuitive, user-friendly and easy-to-learn graphical user interface (GUI) built in the Drupal and its core architecture for slide analysis including a convolutional neural network (CNN) is written in Python language. The build-in CNN model was evaluated for diatom identification under very complex observable backgrounds with mixtures of common impurities, including carbon pig-ments and sand sediments. Compared to the original model, the enhanced model following optimization with limited new datasets was evaluated systematically by independent testing and random control trials (RCTs).Results: In independent testing, the original DiatomNet v].0 was moderately affected, especially when higher densities of impurities existed, and achieved a low recall of 0.817 and F1 score of 0.858 but good precision of 0.905. Following transfer learning with limited new datasets, the enhanced version had better results, with recall and F1 score values of 0.968. A comparative study on real slides showed that the upgraded DiatomNet v].0 obtained F1 scores of 0.86 and 0.84 for carbon pigment and sand sediment, respectively, slightly worse than manual identification (carbon pigment: 0.91; sand sediment: 0.86), but much less time was needed.Conclusions: The study verified that forensic diatom testing with aid of DiatomNet v].0 is much more efficient than traditionally manual identification even under complex observable backgrounds. In terms of forensic diatom testing, we proposed a suggested standard on build-in model optimization and evaluation to strengthen the software's generalization in potentially complex conditions.(c) 2023 Elsevier B.V. All rights reserved.

Previous studies in affective computing often use a fixed emotional label to train an emotion classifier with electroencephalography (EEG) from individuals experiencing an affective stimulus. However, EEGs encode emotional dynamics that include varying intensities within a given emotional category. To investigate these variations in emotional intensity, we propose a framework that obtains momentary affective labels for fine-grained segments of EEGs with human feedback. We then model these labeled segments using a novel spatiotemporal emotional intensity regression network (STEIR-Net). It integrates temporal EEG patterns from nine predefined cortical regions to provide a continuous estimation of emotional intensity. We demonstrate that the STEIR-Net outperforms classical regression models by reducing the root mean square error (RMSE) by an average of 4 similar to 9 % and 2 similar to 4 % for the SEED and SEED-IV databases, respectively. We find that the frontal and temporal cortical regions contribute significantly to the affective intensity's variation. Higher absolute values of the Spearman correlation coefficient between the model estimation and momentary affective labels under happiness (0.2114) and fear (0.2072) compared to neutral (0.1694) and sad (0.1895) emotions were observed. Besides, increasing the input length of the EEG segments from 4 to 20 s further reduces the RMSE from 1.3548 to 1.3188.

Accurate energy consumption prediction can provide insights to make better informed decisions on energy purchase and generation. It also can prevent overloading and make it possible to store energy more efficiently. In this work, we propose a new deep learning model to predict the household energy consumption. In the new model, we employ differential evolution (DE) algorithm to automatically determine the optimal architecture of the deep neural network. The energy prediction results are presented and analyzed to show the effectiveness of the deep neuroevolution model constructed.

Diabetic ketoacidosis (DKA) is a fatal complication that occurs in patients with diabetes mellitus (DM) when there is a severe insulin deficiency. In forensic autopsies, deaths related to DKA usually lack morphological evidence. Therefore, accurate postmortem identification of DKA remains a major challenge in forensic practice. The main aim of this study was to identify DKA in renal tubules by the combination of Fourier transform infrared (FTIR) microspectroscopy and chemometrics. The results of principal component analysis (PCA) showed that the biochemical differences between the renal tubules from the DKA and control groups were mainly due to differences in the structure and content of proteins. Subsequently, partial least squares discriminant analysis (PLS-DA) was used to develop a binary classification model, with diagnostic accuracies of 97.73% and 90.77% for the training set and test set, respectively. The results of PCA and PLS-DA showed that differences in protein structure and content were key to distinguishing DKA from controls. This study demonstrates that the combination of FTIR microspectroscopy and chemometrics can be used to accurately perform postmortem diagnosis of DKA.

This study presents a novel approach for using sliding mode control (SMC) for uncertain nonlinear systems with external unknown disturbances. A nonsingular adaptive barrier function-based finite time control is developed to deal with such systems. In technical implementation, first a novel sliding surface is designed based on the Hurwitz condition, then, using the finite-time stability method, a novel nonsingular terminal sliding manifold is established and the barrier function is obtained, which can guarantee the sliding surface convergence to the origin in finite time. For functional and numerical validation, two examples of uncertain nonlinear systems are selected to illustrate the effectiveness of the proposed method in comparison with the other representative approaches.

Facial Expression Recognition (FER) system analyzes a person's emotions using facial expressions. Nonverbal communication cues, such as facial expressions and gestures, can be seen in FER. FER has applications in various areas, such as medical diagnosis and treatments, human resources, police investigation, education, customer service, journalism, and more. Nowadays, Convolutional Neural Networks (CNNs) play an important role due to their inherent feature extraction mechanism from images. This work utilizes a CNN to classify facial expressions into seven classes: anger, disgust, fear, happiness, sadness, surprise, and neutral. Experiments are conducted with different CNN models and pre-trained models on the FER-2013 dataset. We explore how the model accuracy changes with the model parameters and compare our classification results with existing CNNs. The experimental results demonstrate that the proposed CNN model, trained on a balanced dataset with data augmentation and batch normalization, achieves a training accuracy of 80% and a test accuracy of 72%. Furthermore, the experimental results show that EfficientNet-B1 can achieve a training accuracy of 93 % and a test accuracy of 86 % on a smaller subset of the dataset containing three emotion classes, which is more accurate than the state-of-the-art models.

BACKGROUND:In the present case, we applied postmortem computed tomography angiography (PMCTA) in a medical dispute involving sudden death after cardiovascular surgery.CASE PRESENTATION:A 39-year-old man underwent aortic arch replacement combined with stented elephant trunk implantation surgery under extracorporeal circulation. All vital signs were stable and he was arranged for discharge seven days after surgery. Several days later, the patient was sent back to the hospital for chest pain and poor appetite. Unfortunately, his condition worsened and he ultimately died. PMCT scanning detect pericardial effusion. Family members suspected that the surgical sutures were not dense enough, causing the patient's postoperative bleeding and resulting in cardiac tamponade and death. PMCTA was performed before autopsy, which showed pericardial effusion. However, postmortem angiography with simulated blood pressure showed no leakage of contrast agent, which guided the subsequent autopsy and histological examinations.CONCLUSIONS:While many previous postmortem imaging case reports have shown positive results that provided evidence of medical malpractice, the current case excludes the possibility of physician negligence and reasonably settles the medical dispute from another perspective. In short, the PMCTA approach we describe here was an effective tool that can be applied to certain medical-related forensic cases.

OBJECTIVES:To select four algorithms with relatively balanced complexity and accuracy among deep learning image classification algorithms for automatic diatom recognition, and to explore the most suitable classification algorithm for diatom recognition to provide data reference for automatic diatom testing research in forensic medicine.METHODS:The "diatom" and "background" small sample size data set (20 000 images) of digestive fluid smear of corpse lung tissue in water were built to train, validate and test four convolutional neural network (CNN) models, including VGG16, ResNet50, InceptionV3 and Inception-ResNet-V2. The receiver operating characteristic curve (ROC) of subjects and confusion matrixes were drawn, recall rate, precision rate, specificity, accuracy rate and F1 score were calculated, and the performance of each model was systematically evaluated.RESULTS:The InceptionV3 model achieved much better results than the other three models with a balanced recall rate of 89.80%, a precision rate of 92.58%. The VGG16 and Inception-ResNet-V2 had similar diatom recognition performance. Although the performance of diatom recall and precision detection could not be balanced, the recognition ability was acceptable. ResNet50 had the lowest diatom recognition performance, with a recall rate of 55.35%. In terms of feature extraction, the four models all extracted the features of diatom and background and mainly focused on diatom region as the main identification basis.CONCLUSIONS:Including the Inception-dependent model, which has stronger directivity and targeting in feature extraction of diatom. The InceptionV3 achieved the best performance on diatom identification and feature extraction compared to the other three models. The InceptionV3 is more suitable for daily forensic diatom examination.