Search Results (195)

In our previous work, we developed a neuromorphic decoder of intended movements of tetraplegic patients using ECoG recordings from the brain motor cortex, called Motor Control Decoder (MCD). Even though the training data are labeled based on the desired movement, there is no guarantee that the patient is satisfied by the action of the effectors. Hence, the need for the classification of brain signals as satisfactory/unsatisfactory is obvious. Based on previous work, we upgrade our neuromorphic MCD with a Neural Response Decoder (NRD) that is intended to predict whether ECoG data are satisfactory or not in order to improve MCD accuracy. The main aim is to design an actor–critic structure able to adapt via reinforcement learning the MCD (actor) based on NRD (critic) predictions. For this aim, NRD was trained using not only an ECoG signal but also the MCD prediction or prescribed intended movement of the patient. The achieved accuracy of the trained NRD is satisfactory and contributes to improved MCD performance. However, further work has to be carried out to fully utilize the NRD for MCD performance optimization in an on-line manner. Possibility to include feedback from the patient would allow for further improvement of MCD-NRD accuracy. Full article

Robotic assembly of electrical connectors enables the automation of high-efficiency production of electronic products. A rigid gripper is adopted as the end-effector by the majority of existing works with a force–torque sensor installed at the wrist, which suffers from very limited perception capability of the manipulated objects. Moreover, the grasping and movement actions, as well as the inconsistency between the robot base and the end-effector frame, tend to result in angular misalignment, usually leading to assembly failure. Bio-inspired by the human finger, we designed a tactile finger in this paper with three characteristics: (1) Compliance: A soft ‘skin’ layer provides passive compliance for plenty of manipulation actions, thus increasing the tolerance for alignment errors. (2) Tactile Perception: Two types of sensing elements are embedded into the soft skin to tactilely sense the involved contact status. (3) Enhanced manipulation force: A rigid fingernail is designed to enhance the manipulation force and enable potential delicate operations. Moreover, a tactile-based alignment algorithm is proposed to search for the optimal orientation angle about the z axis. In the application of U-disk insertion, the three characteristics are validated and a success rate of 100% is achieved, whose generalization capability is also validated through the assembly of three types of electrical connectors. Full article

A stroke is a critical medical condition and one of the leading causes of death among humans. Segmentation of the lesions of the brain in which the blood flow is impeded because of blood coagulation plays a vital role in drug prescription and medical diagnosis. Computed tomography (CT) scans play a crucial role in detecting abnormal tissue. There are several methods for segmenting medical images that utilize the main images without considering the patient’s privacy information. In this paper, a deep network is proposed that utilizes compressive sensing and ensemble learning to protect patient privacy and segment the dataset efficiently. The compressed version of the input CT images from the ISLES challenge 2018 dataset is applied to the ensemble part of the proposed network, which consists of two multi-resolution modified U-shaped networks. The evaluation metrics of accuracy, specificity, and dice coefficient are 92.43%, 91.3%, and 91.83%, respectively. The comparison to the state-of-the-art methods confirms the efficiency of the proposed compressive sensing-based ensemble net (CS-Ensemble Net). The compressive sensing part provides information privacy, and the parallel ensemble learning produces better results. Full article

Automated movement intention is crucial for brain–computer interface (BCI) applications. The automatic identification of movement intention can assist patients with movement problems in regaining their mobility. This study introduces a novel approach for the automatic identification of movement intention through finger tapping. This work has compiled a database of EEG signals derived from left finger taps, right finger taps, and a resting condition. Following the requisite pre-processing, the captured signals are input into the proposed model, which is constructed based on graph theory and deep convolutional networks. In this study, we introduce a novel architecture based on six deep convolutional graph layers, specifically designed to effectively capture and extract essential features from EEG signals. The proposed model demonstrates a remarkable performance, achieving an accuracy of 98% in a binary classification task when distinguishing between left and right finger tapping. Furthermore, in a more complex three-class classification scenario, which includes left finger tapping, right finger tapping, and an additional class, the model attains an accuracy of 92%. These results highlight the effectiveness of the architecture in decoding motor-related brain activity from EEG data. Furthermore, relative to recent studies, the suggested model exhibits significant resilience in noisy situations, making it suitable for online BCI applications. Full article

Facial expression plays an important role in human–computer interaction and affective computing. However, existing expression recognition methods cannot effectively capture multi-scale structural details contained in facial expressions, leading to a decline in recognition accuracy. Inspired by the multi-scale processing mechanism of the biological visual system, this paper proposes a wavelet-guided global–local feature aggregation network (WGGLFA) for facial expression recognition (FER). Our WGGLFA network consists of three main modules: the scale-aware expansion (SAE) module, which combines dilated convolution and wavelet transform to capture multi-scale contextual features; the structured local feature aggregation (SLFA) module based on facial keypoints to extract structured local features; and the expression-guided region refinement (ExGR) module, which enhances features from high-response expression areas to improve the collaborative modeling between local details and key expression regions. All three modules utilize the spatial frequency locality of the wavelet transform to achieve high-/low-frequency feature separation, thereby enhancing fine-grained expression representation under frequency domain guidance. Experimental results show that our WGGLFA achieves accuracies of 90.32%, 91.24%, and 71.90% on the RAF-DB, FERPlus, and FED-RO datasets, respectively, demonstrating that our WGGLFA is effective and has more capability of robustness and generalization than state-of-the-art (SOTA) expression recognition methods. Full article

Accurate and rapid prediction of structural responses to seismic loading is critical for ensuring structural safety. Recently, there has been active research focusing on the application of deep learning techniques, including Physics-Informed Neural Networks (PINNs) and Long Short-Term Memory (LSTM) networks, to predict the dynamic behavior of structures. While these methods have shown promise, each comes with distinct limitations. PINNs offer physical consistency but struggle with capturing long-term temporal dependencies in nonlinear systems, while LSTMs excel in learning sequential data but lack physical interpretability. To address these complementary limitations, this study proposes a hybrid LSTM-PINN model, combining the temporal learning ability of LSTMs with the physics-based constraints of PINNs. This hybrid approach allows the model to capture both nonlinear, time-dependent behaviors and maintain physical consistency. The proposed model is evaluated on both single-degree-of-freedom (SDOF) and multi-degree-of-freedom (MDOF) structural systems subjected to the El-Centro ground motion. For validation, the 1940 El-Centro NS earthquake record was used, and the ground acceleration data were normalized and discretized for numerical simulation. The proposed LSTM-PINN is trained under the same conditions as the conventional PINN models (e.g., same optimizer, learning rate, and loss structure), but with fewer training epochs, to evaluate learning efficiency. Prediction accuracy is quantitatively assessed using mean error and mean squared error (MSE) for displacement, velocity, and acceleration, and results are compared with PINN-only models (PINN-1, PINN-2). The results show that LSTM-PINN consistently achieves the most stable and precise predictions across the entire time domain. Notably, it outperforms the baseline PINNs even with fewer training epochs. Specifically, it achieved up to 50% lower MSE with only 10,000 epochs, compared to the PINN’s 50,000 epochs, demonstrating improved generalization through temporal sequence learning. This study empirically validates the potential of physics-guided time-series AI models for dynamic structural response prediction. The proposed approach is expected to contribute to future applications such as real-time response estimation, structural health monitoring, and seismic performance evaluation. Full article

In recent years, the rapid advancement of deep learning techniques has significantly propelled the development of face forgery methods, drawing considerable attention to face forgery detection. However, existing detection methods still struggle with generalization across different datasets and forgery techniques. In this work, we address this challenge by leveraging both local texture cues and global frequency domain information in a complementary manner to enhance the robustness of face forgery detection. Specifically, we introduce a local texture mining and enhancement module. The input image is segmented into patches and a subset is strategically masked, then texture enhanced. This joint masking and enhancement strategy forces the model to focus on generalizable localized texture traces, mitigates overfitting to specific identity features and enabling the model to capture more meaningful subtle traces of forgery. Additionally, we extract multi-scale frequency domain features from the face image using wavelet transform, thereby preserving various frequency domain characteristics of the image. And we propose an innovative frequency-domain processing strategy to adjust the contributions of different frequency-domain components through frequency-domain selection and dynamic weighting. This Facilitates the model’s ability to uncover frequency-domain inconsistencies across various global frequency layers. Furthermore, we propose an integrated framework that combines these two feature modalities, enhanced with spatial attention and channel attention mechanisms, to foster a synergistic effect. Extensive experiments conducted on several benchmark datasets demonstrate that the proposed technique demonstrates superior performance and generalization capabilities compared to existing methods. Full article

In this work, we present a new partially coherent adjustable anomalous vortex laser beam (PCAAVLB) and introduce it into turbulent biological tissue. The equation of such PCAAVLB in turbulent biological tissue is obtained. By numerical analysis, the evolution of the intensity of such PCAAVLB in turbulent biological tissue is analyzed. It is found that the PCAAVLB in biological tissue can lose its ring shape and become a Gaussian beam, and a PCAAVLB with smaller topological charge $M$ or coherence length $\sigma$ will evolve into a Gaussian profile faster. The PCAAVLB in turbulent biological tissue with a smaller small-length-scale factor $l_0$ or larger fractal dimension $D$ will evolve into a Gaussian profile faster and have a larger intensity as z increases. The results may have potential applications in sensing under biological tissue environments and laser imaging in biology. Full article

Despite significant advances in understanding neuronal development, a fully quantitative framework that integrates intracellular mechanisms with environmental cues during axonal growth remains incomplete. Here, we present a unified biophysical model that captures key mechanochemical processes governing axonal extension on micropatterned substrates. In these environments, axons preferentially align with the pattern direction, form bundles, and advance at constant speed. The model integrates four core components: (i) actin–adhesion traction coupling, (ii) lateral inhibition between neighboring axons, (iii) tubulin transport from soma to growth cone, and (iv) orientation dynamics guided by substrate anisotropy. Dynamical systems analysis reveals that a saddle–node bifurcation in the actin adhesion subsystem drives a transition to a high-traction motile state, while traction feedback shifts a pitchfork bifurcation in the signaling loop, promoting symmetry breaking and robust alignment. An exact linear solution in the tubulin transport subsystem functions as a built-in speed regulator, ensuring stable elongation rates. Simulations using experimentally inferred parameters accurately reproduce elongation speed, alignment variance, and bundle spacing. The model provides explicit design rules for enhancing axonal alignment through modulation of substrate stiffness and adhesion dynamics. By identifying key control parameters, this work enables rational design of biomaterials for neural repair and engineered tissue systems. Full article

Brain–Computer Interfaces (BCIs) enable direct communication between the brain and external devices by decoding motor intentions from EEG signals. However, the existing multi-class classification methods for motor imagery EEG (MI-EEG) signals are hindered by low signal quality and limited accuracy, restricting their practical application. This study focuses on rehabilitation training scenarios, aiming to capture the motor intentions of patients with partial or complete motor impairments (such as stroke survivors) and provide feedforward control commands for exoskeletons. This study developed an EEG acquisition protocol specifically for use with lower-limb rehabilitation motor imagery (MI). It systematically explored preprocessing techniques, feature extraction strategies, and multi-classification algorithms for multi-task MI-EEG signals. A novel 3D EEG convolutional neural network (3D EEG-CNN) that integrates time/frequency features is proposed. Evaluations on a self-collected dataset demonstrated that the proposed model achieved a peak classification accuracy of 66.32%, substantially outperforming conventional approaches and demonstrating notable progress in the multi-class classification of lower-limb motor imagery tasks. Full article

To address challenges of background interference and limited multi-scale feature extraction in infrared small target detection, this paper proposes a YOLO-HVS detection algorithm inspired by the human visual system. Based on YOLOv8, we design a multi-scale spatially enhanced attention module (MultiSEAM) using multi-branch depth-separable convolution to suppress background noise and enhance occluded targets, integrating local details and global context. Meanwhile, the C2f_DWR (dilation-wise residual) module with regional-semantic dual residual structure is designed to significantly improve the efficiency of capturing multi-scale contextual information by expanding convolution and two-step feature extraction mechanism. We construct the DroneRoadVehicles dataset containing 1028 infrared images captured at 70–300 m, covering complex occlusion and multi-scale targets. Experiments show that YOLO-HVS achieves mAP50 of 83.4% and 97.8% on the public dataset DroneVehicle and the self-built dataset, respectively, which is an improvement of 1.1% and 0.7% over the baseline YOLOv8, and the number of model parameters only increases by 2.3 M, and the increase of GFLOPs is controlled at 0.1 G. The experimental results demonstrate that the proposed approach exhibits enhanced robustness in detecting targets under severe occlusion and low SNR conditions, while enabling efficient real-time infrared small target detection. Full article

Diagnosing schizophrenia using Electroencephalograph (EEG) signals is a challenging task due to the subtle and overlapping differences between patients and healthy individuals. To overcome this difficulty, deep learning has shown strong potential, especially given its success in image recognition tasks. In many studies, one-dimensional EEG signals are transformed into two-dimensional representations to allow for image-based analysis. In this work, we have used the Markov Transition Field for converting EEG signals into two-dimensional images, capturing both the temporal patterns and statistical dynamics of the data. EEG signals are continuous time-series recordings from the brain, where the current state is often influenced by the immediately preceding state. This characteristic makes MTF particularly suitable for representing such data. After the transformation, a pre-trained VGG-16 model is employed to extract meaningful features from the images. The extracted features are then passed through two separate classification pipelines. The first uses a traditional machine learning model, Support Vector Machine, while the second follows a deep learning approach involving an autoencoder for feature selection and a neural network for final classification. The experiments were conducted using EEG data from the open-access Schizophrenia EEG database provided by MV Lomonosov Moscow State University. The proposed method achieved a highest classification accuracy of 98.51 percent and a recall of 100 percent across all folds using the deep learning pipeline. The Support Vector Machine pipeline also showed strong performance with a best accuracy of 96.28 percent and a recall of 97.89 percent. The proposed deep learning model represents a biomimetic approach to pattern recognition and decision-making. Full article

The results of radiotherapy in patients with primary malignant brain tumors are extremely dissatisfactory: the overall survival after a diagnosis of glioblastoma is typically less than three years. The development of spatially fractionated radiotherapy techniques could help to improve this bleak prognosis. In order to develop technical equipment and organ-specific therapy plans, dosimetry studies as well as radiobiology studies are conducted. Although perfect spheres are considered optimal phantoms by physicists, this does not reflect the wide variety of head sizes and shapes in our patient community. Depth from surface and X-ray dose absorption by tissue between dose entry point and target, two key parameters in medical physics planning, are largely determined by the shape and thickness of the skull bone. We have, therefore, designed and produced a biomimetic tool to correlate measured technical dose and biological response in human cancer cells: a brain phantom, produced from tissue-equivalent materials. In a first pilot study, utilizing our phantom to correlate technical dose measurements and metabolic response to radiation in human cancer cell lines, we demonstrate why an anthropomorphic phantom is preferable over a simple spheroid phantom. Full article

In this study, we propose a novel improved Dung Beetle Optimizer called Environment-aware Chaotic Force-field Dung Beetle Optimizer (ECFDBO). To address DBO’s existing tendency toward premature convergence and insufficient precision in high-dimensional, complex search spaces, ECFDBO integrates three key improvements: a chaotic perturbation-based nonlinear contraction strategy, an intelligent boundary-handling mechanism, and a dynamic attraction–repulsion force-field mutation. These improvements reinforce both the algorithm’s global exploration capability and its local exploitation accuracy. We conducted 30 independent runs of ECFDBO on the CEC2017 benchmark suite. Compared with seven classical and novel metaheuristic algorithms, ECFDBO achieved statistically significant improvements in multiple performance metrics. Moreover, by varying problem dimensionality, we demonstrated its robust global optimization capability for increasingly challenging tasks. We further conducted the Wilcoxon and Friedman tests to assess the significance of performance differences of the algorithms and to establish an overall ranking. Finally, ECFDBO was applied to a 3D path planning simulation in UAVs for safe path planning in complex environments. Against both the Dung Beetle Optimizer and a multi-strategy DBO (GODBO) algorithm, ECFDBO met the global optimality requirements for cooperative UAV planning and showed strong potential for high-dimensional global optimization applications. Full article

Robust emotion recognition is a prerequisite for natural, fluid human–computer interaction, yet conversational settings remain challenging because emotions are shaped simultaneously by global topic flow and local speaker-to-speaker dependencies. Here, we introduce GASMER—Graph-Adaptive Structure for Multimodal Emotion Recognition—a unified architecture that tackles both issues. It uses the correlation structure based on graph neural networks (GNNs) to model the complex dependencies in the conversation, while adaptively learning the graph structure for GNNs. The experiments indicate that our model has strong performance that outperforms all existing graph-based approaches, and remains competitive when compared to recent multimodal fusion models, underscoring the importance of combining fine-grained multimodal fusion with adaptive graph learning for conversational emotion recognition. On the IEMOCAP dataset, GASMER improves accuracy by 2.7% and the weighted F1-score by 3.6% compared to the best baseline. On the MOSEI dataset, it achieves a 1.2% gain in binary classification accuracy (ACC-2). Full article