Explore the vast range of titles available.

Current deep learning-based change detection approaches mostly produce convincing results by introducing attention mechanisms to traditional convolutional networks. However, given the limitation of the receptive field, convolution-based methods fall short of fully modelling global context and capturing long-range dependencies, thus insufficient in discriminating pseudo changes. Transformers have an efficient global spatio-temporal modelling capability, which is beneficial for the feature representation of changes of interest. However, the lack of detailed information may cause the transformer to locate the boundaries of changed regions inaccurately. Therefore, in this article, a hybrid CNN-transformer architecture named CTCANet, combining the strengths of convolutional networks, transformer, and attention mechanisms, is proposed for high-resolution bi-temporal remote sensing image change detection. To obtain high-level feature representations that reveal changes of interest, CTCANet utilizes tokenizer to embed the features of each image extracted by convolutional network into a sequence of tokens, and the transformer module to model global spatio-temporal context in token space. The optimal bi-temporal information fusion approach is explored here. Subsequently, the reconstructed features carrying deep abstract information are fed to the cascaded decoder to aggregate with features containing shallow fine-grained information, through skip connections. Such an aggregation empowers our model to maintain the completeness of changes and accurately locate small targets. Moreover, the integration of the convolutional block attention module enables the smoothing of semantic gaps between heterogeneous features and the accentuation of relevant changes in both the channel and spatial domains, resulting in more impressive outcomes. The performance of the proposed CTCANet surpasses that of recent certain state-of-the-art methods, as evidenced by experimental results on two publicly accessible datasets, LEVIR-CD and SYSU-CD.

The interactions between chemical, hydraulic, and mechanical processes in clay soils has garnered significant attention in various fields. It is important to study the transport of chemical substances and the deformation of the clay layer under the combined action of chemical solution and mechanical loading. Clay is a typical porous medium, primarily composed of soil skeleton and pores, with the shape, size, and distribution of clay pores being random, making it difficult to accurately describe using traditional geometry. A fractal model can be used to simulate clay soils. This study leverages the fractal theory of clay, incorporating the generalized effective stress principle in chemical solutions and a fractal model for the diffusion of chemical substances. A chemo-hydro-mechanical coupling model based on fractal theory is developed. Simulation results show that an increase in fractal dimension corresponds to a heightened roughness of the pore surfaces and an increased tortuosity of the pore channels, which significantly amplify the resistance to solute diffusion, thereby retarding the rate of solute transport. As a result, the magnitude of the maximum negative pore pressure within the clay layer, leading to a longer time required for complete pore pressure dissipation. Additionally, the deformation of the clay layer is larger.

This paper considers a laser-powered unmanned aerial vehicle (UAV)-enabled wireless power transfer (WPT) system. In the system, a UAV is dispatched as an energy transmitter to replenish energy for battery-limited sensors in a wireless rechargeable sensor network (WRSN) by transferring radio frequency (RF) signals, and a mobile unmanned vehicle (MUV)-loaded laser transmitter travels on a fixed path to charge the on-board energy-limited UAV when it arrives just below the UAV. Based on the system, we investigate the trajectory optimization of laser-charged UAVs for charging WRSNs (TOLC problem), which aims to optimize the flight trajectories of a UAV and the travel plans of an MUV cooperatively to minimize the total working time of the UAV so that the energy of every sensor is greater than or equal to the threshold. Then, we prove that the problem is NP-hard. To solve the TOLC problem, we first propose the weighted centered minimum coverage (WCMC) algorithm to cluster the sensors and compute the weighted center of each cluster. Based on the WCMC algorithm, we propose the TOLC algorithm (TOLCA) to design the detailed flight trajectory of a UAV and the travel plans of an MUV, which consists of the flight trajectory of a UAV, the hovering points of a UAV with the corresponding hovering times used for the charging sensors, the hovering points of a UAV with the corresponding hovering times used for replenishing energy itself, and the hovering times of a UAV waiting for an MUV. Numerical results are provided to verify that the suggested strategy provides an effective method for supplying wireless rechargeable sensor networks with sustainable energy.

Classifying birds accurately is essential for ecological monitoring. In recent years, bird image classification has become an emerging method for bird recognition. However, the bird image classification task needs to face the challenges of high intraclass variance and low inter-class variance among birds, as well as low model efficiency. In this paper, we propose a fine-grained bird classification method based on attention and decoupled knowledge distillation. First of all, we propose an attention-guided data augmentation method. Specifically, the method obtains images of the object’s key part regions through attention. It enables the model to learn and distinguish fine features. At the same time, based on the localization–recognition method, the bird category is predicted using the object image with finer features, which reduces the influence of background noise. In addition, we propose a model compression method of decoupled knowledge distillation. We distill the target and nontarget class knowledge separately to eliminate the influence of the target class prediction results on the transfer of the nontarget class knowledge. This approach achieves efficient model compression. With 67% fewer parameters and only 1.2 G of computation, the model proposed in this paper still has a 87.6% success rate, while improving the model inference speed.

Wearable devices have provided noninvasive and continuous monitoring of physiological parameters in healthcare applications. However, for the comfortable applications of wearable devices on human body, two key requirements are to replace conventional bulky devices into soft and deformable ones and to have wireless wearable communication. In this paper we present a simple, low-cost and highly efficient all-elastomeric conductor that can be used in a soft radio-frequency (RF) transmission line and antenna. We show a stretchable transmission line and two stretchable antennas fabricated with conventional screen printing. The stretchable conductor used in this fabrication method, which is a mixture of Ag and Polydimethylsiloxane (PDMS), can be stretched at high strains while maintaining a high conductivity, low attenuation and feasible radiation performance. The measured conductivity of the stretchable conductor reaches 1000 S/cm. Additionally, the highly conductive printed Ag-PDMS is utilized to construct transmission lines and antennas. The performance of these stretchable components, especially under different conditions of bending, stretching and twisting, are experimentally examined in common wireless-communication frequency bands. Our results demonstrate that printed Ag-PDMS enabled RF passive components have the desired property and quality for wireless wearable communication applications, which would provide new opportunities for wearable healthcare electronics.

Medium–low density UAV point clouds often suffer from incomplete lower canopy structures and sparse distributions due to self-occlusion. While existing point cloud completion models achieve high metric accuracy, they inadequately address missing regions in trunks and lower canopy areas. To resolve these issues, this paper proposes a hierarchical random sampling strategy and a spatially constrained loss function. First, we dynamically stratify point clouds based on density distribution characteristics, employing hierarchical random sampling to preserve proportional representation of lower-level points, thereby effectively retaining basal tree structure information. Second, we introduce a distance constraint term for mid-lower point clouds into the symmetrical Chamfer distance (CD) loss, compelling models to prioritize completion quality in trunk base regions. Experiments on the FOR-instance-created completion dataset and Xiong’an dataset demonstrate that our method significantly enhances structural recovery capability at tree trunk bases, with visual results outperforming the baseline SeedFormer model. Additionally, we refer to existing point cloud-based diameter at breast height (DBH) calculation methods to measure the completed trees and compare the computed results with the measured values to evaluate the accuracy of the completion effect. Experimental results show that, after integrating our proposed strategies with existing completion methods, the accuracy of DBH measurement from point clouds is significantly improved. This study provides novel insights for addressing structural bias in tree point cloud completion and offers valuable references for digital forestry resource management.

Impact ionization effect has been demonstrated in transistors to enable sub-60 mV dec −1 subthreshold swing. However, traditionally, impact ionization in silicon devices requires a high operation voltage due to limited electrical field near the device drain, contradicting the low energy operation purpose. Here, we report a vertical subthreshold swing device composed of a graphene/silicon heterojunction drain and a silicon channel. This structure creates a low voltage avalanche impact ionization phenomenon and leads to steep switching of the silicon-based device. Experimental measurements reveal a small average subthreshold swing of 16 µV dec −1 over 6 decades of drain current and nearly hysteresis-free, and the operating voltage at which a vertical subthreshold swing occurs can be as low as 0.4 V at room temperature. Furthermore, a complementary silicon-based logic inverter is experimentally demonstrated to reach a voltage gain of 311 at a supply voltage of 2 V. Yuan et al. report a nearly vertical subthreshold swing field-effect transistor consists of a graphene/silicon heterojunction drain and a silicon channel. The device enables nearly hysteresis-free transistors with subthreshold swing of 16 µV dec −1 , and a complementary logic inverter with gain of 311.

Acoustic detection technology is a new method for early monitoring of wood-boring pests, and the effective denoising methods are the premise of acoustic detection in forests. This paper used sensors to record Semanotus bifasciatus larval feeding sounds and various environmental noises, and two kinds of sounds were mixed to obtain the noisy feeding sounds with controllable noise intensity. Then, the time domain denoising models and frequency domain denoising models were designed, and the denoising effects were compared using the metrics of a signal-to-noise ratio (SNR), a segment signal-noise ratio (SegSNR), and log spectral distance (LSD). In the experiments, the average SNR increment could achieve 17.53 dB and 11.10 dB using the in the test data using the time domain features and frequency domain features, respectively. The average SegSNR increment achieved 18.59 dB and 12.04 dB, respectively, and the average LSD between pure feeding sounds and denoised feeding sounds were 0.85 dB and 0.84 dB, respectively. The experimental results demonstrated that the denoising models based on artificial intelligence were effective methods for S. bifasciatus larval feeding sounds, and the overall denoising effect was more significant, especially at low SNRs. In view of that, the denoising models using time domain features were more suitable for the forest area and quarantine environment with complex noise types and large noise interference.

Recently, the deep neural network (DNN) has become one of the most advanced and powerful methods used in classification tasks. However, the cost of DNN models is sometimes considerable due to the huge sets of parameters. Therefore, it is necessary to compress these models in order to reduce the parameters in weight matrices and decrease computational consumption, while maintaining the same level of accuracy. In this paper, in order to deal with the compression problem, we first combine the loss function and the compression cost function into a joint function, and optimize it as an optimization framework. Then we combine the CUR decomposition method with this joint optimization framework to obtain the low-rank approximation matrices. Finally, we narrow the gap between the weight matrices and the low-rank approximations to compress the DNN models on the image classification task. In this algorithm, we not only solve the optimal ranks by enumeration, but also obtain the compression result with low-rank characteristics iteratively. Experiments were carried out on three public datasets under classification tasks. Comparisons with baselines and current state-of-the-art results can conclude that our proposed low-rank joint optimization compression algorithm can achieve higher accuracy and compression ratios.

Ethylene oxide (EO) is a reactive epoxide. However, the association between EO exposure and the risk of developing asthma in humans is unknown. The aim of this study was to investigate the relationship between EO exposure and the risk of developing asthma in the general US population. In this cross-sectional study, data of 2542 patients from the National Health and Nutrition Examination Survey (NHANES) between 2013 and 2016 were obtained and analyzed. Hemoglobin adducts of EO (HbEO) level be used as the main factor for predicting EO exposure. The association between the level of EO exposure and the risk of developing asthma was evaluated with logistic regression models and dose–response analysis curves of restricted cubic spline function. Mediation analysis and linear regression analysis were utilized to evaluate the association between EO exposure and inflammation indicators. According to the quartiles of HbEO level, the patients were divided into four groups. The results indicated that an increased HbEO level was associated with a higher risk of asthma onset. Compared with the lowest quartile, the odds ratio (OR) with the 95% confidence interval (CI) for the highest quartile was 1.960 (95% CI: 1.348–2.849, P  = 0.003). After being adjusted for numerous potential confounders, the OR of quartile 4 relative to quartile 1 was 1.991 (95% CI: 1.359–2.916, P  = 0.001). Consistent results were also obtained in most subgroup analyses and dose–response analysis curves. In addition, EO levels were positively correlated with the inflammatory indicators ( P  = 0.006 for WBC, P  = 0.015 for lymphocyte, and P  = 0.015 for neutrophil). This study revealed a positive correlation between the level of EO exposure and the risk of asthma in a representative US population. In addition, inflammatory response may prove to be a potential biological mechanism underlying EO-induced asthma.