Facial expression methods play a vital role in human–computer interaction and other fields, but there are factors such as occlusion, illumination, and pose changes in wild facial recognition, as well as category imbalances between different datasets, that result in large variations in recognition rates and low accuracy rates for different categories of facial expression datasets. This study introduces RCL-Net, a method of recognizing wild facial expressions that is based on an attention mechanism and LBP feature fusion. The structure consists of two main branches, namely the ResNet-CBAM residual attention branch and the local binary feature (LBP) extraction branch (RCL-Net). First, by merging the residual network and hybrid attention mechanism, the residual attention network is presented to emphasize the local detail feature information of facial expressions; the significant characteristics of facial expressions are retrieved from both channel and spatial dimensions to build the residual attention classification model. Second, we present a locally improved residual network attention model. LBP features are introduced into the facial expression feature extraction stage in order to extract texture information on expression photographs in order to emphasize facial feature information and enhance the recognition accuracy of the model. Lastly, experimental validation is performed using the FER2013, FERPLUS, CK+, and RAF-DB datasets, and the experimental results demonstrate that the proposed method has superior generalization capability and robustness in the laboratory-controlled environment and field environment compared to the most recent experimental methods.

Response surface methodology (RSM) and central composite design (CCD) were used to improve the preparation of carbon nanotube and graphene (CNT-GN)-sensing unit composite materials in this study. Four independent variable factors (CNT content, GN content, mixing time, and curing temperature) were controlled at five levels, and 30 samples were generated using the multivariate control analysis technique. On the basis of the experimental design, semi-empirical equations were developed and utilized to predict the sensitivity and compression modulus of the generated samples. The results reveal a strong correlation between the experimental and expected values of sensitivity and the compression modulus for the CNT-GN/RTV (room-temperature-vulcanized silicone rubber) polymer nanocomposites fabricated using different design strategies. The correlation coefficients for the sensitivity and compression modulus are $R^{2 }=0.9634$ and $R^2=0.9115,$ respectively. The ideal preparation parameters of the composite in the experimental range include a CNT content of 1.1 g, a GN content of 1.0 g, a mixing time of 15 min, and a curing temperature of 68.6 °C, according to theoretical predictions and experimental findings. At 0~30 kPa, the CNT-GN/RTV-sensing unit composite materials may reach a sensitivity of 0.385 kPa⁻¹ and a compressive modulus of 601.567 kPa. This provides a new idea for the preparation of flexible sensor cells and reduces the time and economic cost of experiments.

The rapid advancement of humanoid robotics has spurred researchers’ interest in flexible sensors for wide linear range detection. In response, we report a capacitive flexible pressure sensor based on a multi-walled carbon nanotubes/titanium dioxide/polydimethylsiloxane (MWCNTs/TiO₂/PDMS) composite. A micro-hemispherical structure array formed on the composite surface via a templating method reduces the initial capacitance value. Modified carbon nanotubes (F-MWCNTs) were prepared using 2 wt%, 5 wt% and 10 wt% γ-aminopropyltriethoxysilane (APTES), significantly enhancing dispersion and interfacial bonding strength. The synergistic effect of microstructures and MWCNTs surface functionalization further enhances sensing performance. The F-MWCNTs/TiO₂/PDMS pressure sensor modified with 2 wt% APTES exhibits outstanding sensing capabilities: it demonstrates dual-stage sensitivity across a broad linear range of 0–95 kPa (0–13 kPa segment: 1.89 ± 0.49 kPa⁻¹; 13–95 kPa segment: 7.08 ± 0.63 kPa⁻¹), with a response time of 200 milliseconds, maintaining stability over 2500 cyclic loadings. In practical application exploration, this sensor has demonstrated strong adaptability, confirming its significant potential in micro-pressure detection, wearable electronics, and array sensing applications.

Polymer creep can significantly reduce the safety and dependability of composite applications, restricting their development and use in additional fields. In this study, single-factor and multi-factor analysis techniques were employed to systematically explore the impacts of nickel powder and graphene on the resistive creep of sensing units. The creep model between the rate of resistance changes and the pressure was established, and the material ratio was optimized to obtain a high creep resistance. The results demonstrated that the creep resistance was best when the filling particle was 10 wt.% and the ratio of nickel powder to graphene was 4:21, which was approximately 60% and 45% lower than the filling alone and the composite filling before optimization, respectively; the R² of the theoretical value of the resistance creep model and the experimental value of the creep before and after optimization was 0.9736 and 0.9812, indicating that the resistance creep model was highly accurate. Consequently, the addition of filler particles with acceptable proportions, varied shapes, and different characteristics to polymers can effectively reduce polymer creep and has significant potential for the manufacture of sensing units for tactile sensors.

Vision-based classroom behavior detection is challenged by dense student distributions, frequent occlusions, background clutter, and large-scale variations. To address these issues, this study proposes DBRA-SCSA-YOLO, an enhanced YOLOv8-based detector integrating a Dual-Branch Receptive-field Attention (DBRA) module and a Spatial-Channel Separable Attention (SCSA) mechanism. DBRA improves receptive-field adaptability and multi-scale feature representation, while SCSA performs staged spatial localization refinement and channel-wise semantic recalibration after feature fusion. Experiments were conducted using a three-category subset of SCB-Dataset5, with YOLOv8 variants and transformer-based detectors included as comparison baselines. The model was evaluated using standard object-detection metrics, including Precision, Recall, and mAP. Across five random seeds, DBRA-SCSA-YOLO achieved mean Precision, Recall, mAP@50, and mAP@50–95 values of 0.695, 0.685, 0.720, and 0.536, respectively. Compared with YOLOv8n, the proposed model improved mean mAP@50 and mAP@50–95 by 9.5 and 9.0% points. Ablation and architectural sensitivity analyses further supported the complementary roles of DBRA and SCSA under the reported experimental conditions. The framework operates on static classroom images and does not incorporate temporal information. Therefore, the findings should be interpreted as within-dataset evidence, and independent classroom datasets are needed to establish cross-environment generalization and practical deployment capability.