As an important direction in computer vision, human pose estimation has received extensive attention in recent years. A High-Resolution Network (HRNet) can achieve effective estimation results as a classical human pose estimation method. However, the complex structure of the model is not conducive to deployment under limited computer resources. Therefore, an improved Efficient and Lightweight HRNet (EL-HRNet) model is proposed. In detail, point-wise and grouped convolutions were used to construct a lightweight residual module, replacing the original 3 × 3 module to reduce the parameters. To compensate for the information loss caused by the network’s lightweight nature, the Convolutional Block Attention Module (CBAM) is introduced after the new lightweight residual module to construct the Lightweight Attention Basicblock (LA-Basicblock) module to achieve high-precision human pose estimation. To verify the effectiveness of the proposed EL-HRNet, experiments were carried out using the COCO2017 and MPII datasets. The experimental results show that the EL-HRNet model requires only 5 million parameters and 2.0 GFlops calculations and achieves an AP score of 67.1% on the COCO2017 validation set. In addition, PCKh@0.5mean is 87.7% on the MPII validation set, and EL-HRNet shows a good balance between model complexity and human pose estimation accuracy.

Multiple types of brain-control systems have been applied in the field of rehabilitation. As an alternative scheme for balancing user fatigue and the classification accuracy of brain–computer interface (BCI) systems, facial-expression-based brain control technologies have been proposed in the form of novel BCI systems. Unfortunately, existing machine learning algorithms fail to identify the most relevant features of electroencephalogram signals, which further limits the performance of the classifiers. To address this problem, an improved classification method is proposed for facial-expression-based BCI (FE-BCI) systems, using a convolutional neural network (CNN) combined with a genetic algorithm (GA). The CNN was applied to extract features and classify them. The GA was used for hyperparameter selection to extract the most relevant parameters for classification. To validate the superiority of the proposed algorithm used in this study, various experimental performance results were systematically evaluated, and a trained CNN-GA model was constructed to control an intelligent car in real time. The average accuracy across all subjects was 89.21 ± 3.79%, and the highest accuracy was 97.71 ± 2.07%. The superior performance of the proposed algorithm was demonstrated through offline and online experiments. The experimental results demonstrate that our improved FE-BCI system outperforms the traditional methods.

Functional brain network (FBN) provides an effective biomarker for understanding brain activation patterns and a diagnostic criterion for neurodegenerative diseases detections. Unfortunately, it remains challenges to estimate the biologically meaningful or discriminative FBNs accurately, because of the poor quality of functional magnetic resonance imaging data or our limited understanding of human brain. In this study, a novel FBN estimation model based on group similarity prior was proposed. In particular, we extended the FBN estimation model to tensor form and incorporated the tensor trace-norm regularizer to formulate the group similarity constraint. To verify the proposed method, we conducted experiments on identifying mild cognitive impairments (MCIs) from normal controls (NCs) based on the estimated FBNs. Experimental results illustrated that our method is effective in modeling FBNs. Consequently, we achieved 91.97% classification accuracy, outperforming the state-of-the-art methods. The post hoc analysis further demonstrated that more biologically meaningful functional brain connections were obtained using our proposed method.

Individuals with severe tetraplegia frequently require to control their complex assistive devices using body movement with the remaining activity above the neck. Electromyography (EMG) signals from the contractions of facial muscles enable people to produce multiple command signals by conveying information about attempted movements. In this study, a novel EMG-controlled system based on facial actions was developed. The mechanism of different facial actions was processed using an EMG control model. Four asymmetric and symmetry actions were defined to control a two-degree-of-freedom (2-DOF) prosthesis. Both indoor and outdoor experiments were conducted to validate the feasibility of EMG-controlled prostheses based on facial action. The experimental results indicated that the new paradigm presented in this paper yields high performance and efficient control for prosthesis applications. Individuals with severe tetraplegia frequently require to control their complex assistive devices using body movement with the remaining activity above the neck. Electromyography (EMG) signals from the contractions of facial muscles enable people to produce multiple command signals by conveying information about attempted movements. In this study, a novel EMG-controlled system based on facial actions was developed. The mechanism of different facial actions was processed using an EMG control model. Four asymmetric and symmetry actions were defined to control a two-degree-of-freedom (2-DOF) prosthesis. Both indoor and outdoor experiments were conducted to validate the feasibility of EMG-controlled prostheses based on facial action. The experimental results indicated that the new paradigm presented in this paper yields high performance and efficient control for prosthesis applications.

In recent years, the steady-state visual evoked potentials (SSVEP) based brain control method has been employed to help people with disabilities because of its advantages of high information transmission rate and low training time. However, the existing SSVEP brain control methods cannot adapt to dynamic or unstructured environments. Moreover, the recognition accuracy from the conventional decoding algorithm still needs to improve. To address the above problems, this study proposed a steady-state hybrid visual evoked potentials (SSHVEP) paradigm using the grasping targets in their environment to improve the connection between the subjects’ and their dynamic environments. Moreover, a novel EEG decoding method, using the multivariate variational mode decomposition (MVMD) algorithm for adaptive sub-band division and convolutional neural network (CNN) for target recognition, was applied to improve the decoding accuracy of the SSHVEPs. 18 subjects participated in the offline and online experiments. The offline accuracy across 18 subjects by the 9-target SSHVEP paradigm was up to 95.41 ± 2.70%, which is a 5.80% improvement compared to the conventional algorithm. To further validate the performance of the proposed method, the brain-controlled grasping robot system using the SSHVEP paradigm was built. The average accuracy reached 93.21 ± 10.18% for the online experiment. All the experimental results demonstrated the effectiveness of the brain computer interaction method based on the SSHVEP paradigm and the MVMD combined CNN algorithm studied in this paper.