Search

In the charity sector, fundraising and transparency have long been key issues. Charity NFT (Non-Fungible Token) auctions, an emerging charity fundraising model integrating blockchain and NFT concepts, bring opportunities and challenges. Blockchain provides distributed data integrity and transparency via cryptography-linked data blocks, while NFTs enable unique digital ownership representation. This study designs a charity NFT auction platform on the Fisco Bcos blockchain, using multi-signature algorithms to ensure NFT authenticity, ECDSA (Elliptic Curve Digital Signature Algorithm) signatures for transaction integrity and traceability, NFTs and virtual currencies to reduce costs, and IPFS (InterPlanetary File System) for storage. The implemented system achieves 2104 TPS throughput with 492-millisecond latency, increasing transaction processing with low latency. Overall, the platform aims to address charity issues like opaque fund flows, high costs, and fake initiatives through the strategic application of blockchain and NFT functionality.

Public Key Encryption with Equality Test (PKEET) is a cryptographic primitive that allows an authorized entity to test whether two given ciphertexts are the encryption of the same message without decrypting them. The security of cryptographic schemes is analyzed using security model, and thus in order to derive reasonable security against the real attackers, the security model should reflect the real attack as closely as possible. However, security model widely used by PKEET fails to capture corruption attack, since it does not cover the real attacker who can adaptively corrupt users. On the other hand, many PKEET schemes suffer from a security loss that is linear in the number of users when using security model with adaptive corruption attack, which causes that the actual security guarantees of the schemes linearly degrade in that. Therefore, the goal of this paper is to resolve these two problems. We present a PKEET scheme in setting with adaptive corruptions in which the security loss is a constant, and in particular, the comparison shows that our scheme is efficient.

The deficit in social interaction skills among individuals with autism spectrum disorder (ASD) is strongly influenced by personal experiences and social environments. Neuroimaging studies have previously highlighted the link between social impairment and brain activity in ASD. This study aims to develop a method for assessing and identifying ASD using a social cognitive game-based paradigm combined with electroencephalo-graphy (EEG) signaling features. Typically developing (TD) participants and autistic preadolescents and teenagers were recruited to participate in a social game while 12-channel EEG signals were recorded. The EEG signals underwent preprocessing to analyze local brain activities, including event-related potentials (ERPs) and time-frequency features. Additionally, the global brain network’s functional connectivity between brain regions was evaluated using phase-lag indices (PLIs). Subsequently, machine learning models were employed to assess the neurophysiological features. Results indicated pronounced ERP components, particularly the late positive potential (LPP), in parietal regions during social training. Autistic preadolescents and teenagers exhibited lower LPP amplitudes and larger P200 amplitudes compared to TD participants. Reduced theta synchronization was also observed in the ASD group. Aberrant functional connectivity within certain time intervals was noted in the ASD group. Machine learning analysis revealed that support-vector machines achieved a sensitivity of 100%, specificity of 91.7%, and accuracy of 95.8% as part of the performance evaluation when utilizing ERP and brain oscillation features for ASD characterization. These findings suggest that social interaction difficulties in autism are linked to specific brain activation patterns. Traditional behavioral assessments face challenges of subjectivity and accuracy, indicating the potential use of social training interfaces and EEG features for cognitive assessment in ASD.

Existing in-vehicle gesture recognition using channel state information (CSI) of Wi-Fi signals requires the use of 3 or more receivers. Due to the lack of research on gesture recognition using fewer receivers, we here study the case of only two receivers. CSI data of 4 gestures are collected by two in-vehicle receivers and are then processed to generate body velocity profile (BVP) data. We propose a Transformer and long short-term memory (LSTM) (Transformer-LSTM) fusion model with a feature extraction and dimensionality reduction module (FEDRT-LSTM) to recognize gestures using BVP data from two-receiver CSI. The performance is compared with those of deep learning neural network (DNN), gated recurrent unit (GRU) capsules network (GRU-CapNet), LeNet-5, and GRU network models with hyperparameters optimized. Results show that the FEDRT-LSTM model can achieve the best recognition accuracy of 95.54%. This approach can reduce the number of receivers to two and achieve good performance in the recognition of 4 gestures.

Melanocytic nevi (MN), which are the most prevalent benign skin tumors, can be classified into three subtypes based on the depth of the nevus cell nests: junctional, compound, and intradermal nevi. Among these subtypes, intradermal nevi pose a lower risk of malignancy, whereas junctional nevi carry a higher risk of malignancy. To facilitate early stage diagnosis, mitigate unnecessary biopsies, and reduce medical costs, this paper presents the first study on the classification of benign melanocytic nevi. We propose an algorithm that utilizes depth-relevant, channel-specific, and channel-shared information from dermoscopy images. This approach draws inspiration from the physical principles underlying dermoscopy imaging and the physiological essence of nevi. We employed CNN and transformer blocks to capture local and long-range contextual information. In addition, our physiology-inspired module focuses on learning depth-related features from the color channels. To assess the performance of our proposed method, we curated a dataset of dermoscopic images of MN and compared it with two human experts as well as classic CNN and Transformer baseline models. Our results showed that the proposed automated method achieved an average accuracy of 70.23%, surpassing the best-performing baseline by 3.6%, and outperforming human experts by nearly 20%. Furthermore, we conducted an ablation study to confirm the effectiveness of our algorithm and to analyze its clinical significance.

The facility layout of the production workshop primarily focuses on the layout planning of production equipment to reduce material handling costs within the workshop. With the development of the low-carbon economy, a scientifically rational facility layout planning contributes to improving overall operational efficiency and reducing carbon emissions in the production process. However, the design of a production workshop facility layout is a complex optimization problem involving multiple objectives, such as minimizing handling costs, maximizing total non-logistics relationships, and optimizing the utilization of workers’ working hours. Additionally, it must adhere to various constraints, including area utilization rates and equipment utilization constraints. Existing layout methods often fall short of meeting the practical requirements of engineering practice. Therefore, to address the optimization challenges related to the facility layout of the production workshop, this paper establishes a comprehensive optimization model, targeting low-carbon logistics within the production workshop and optimizing the overall non-logistical relationships between operational units as the optimization objectives. Subsequently, this paper proposes an improved Particle Swarm Optimization (PSO) method, considering task collaboration, to solve the integrated optimization model of the facility layout for the valve component production workshop at Company A. Finally, the validity of the model and algorithm is confirmed through example calculations and result analysis. The analysis results demonstrate that, under the same conditions, the improved PSO algorithm outperforms PSO and SGA (Simple Genetic Algorithm) algorithms in terms of optimization results, iteration counts, and runtime. In conclusion, this study introduces task collaboration to enhance the traditional PSO algorithm. Simultaneously, we consider both logistics and non-logistics relationships in optimizing facility layout design. This provides theoretical references and new solving algorithms for low-carbon logistics.

The Fengyun 4B (FY-4B) as an operational satellite was launched on 3 June 2021. The advanced geostationary radiation imager (AGRI) aboard on the FY-4B has ability to monitor aerosols information. The objective of this study is to introduce the operational algorithm for retrieving land aerosol based on FY-4B/AGRI data, which is called as an adaptive dark-target algorithm. First, on the basis of FY-4B/AGRI cloud mask and snow cover products, a novelty identification scheme was proposed to classify the pixels into cloud, haze and clear categories. Then, aiming at the characteristics of FY-4B/AGRI instrument, the adaptive relationships of surface reflectance varying with the normalized vegetation index or view zenith angle were established by atmospheric correction. By validating the AGRI retrievals every 15 min with AERONET AOD, AGRI performance (R: 0.95, bias: –0.014, RMSE: 0.091, within expected error (EE:): 81.99%) was comparable to that of Aqua/MODIS deep blue AOD (R: 0.95, bias: 0.024, RMSE: 0.095, EE: 78.57%), while have the largest difference with those of MYDDT. At daily mean scale, the statistics of AGRI AOD against AERONET performed well (R: 0.93, bias: 0.062, RMSE: 0.18, EE: 64%). Additionally, the agreement of AGRI retrievals with AERONET AOD varied with observation moments and stations. The FY-4B/AGRI operational aerosol product based on adaptive dark target algorithm were proved to be robust.

With the explosive growth of user data, AI applications are increasingly affecting people’s lives. To tackle the problems of data privacy and network congestion, people raise their interest in the Federated Learning (FL) framework, which enables Edge Nodes (ENs) to learn a global model without data sharing. However, FL also brings some challenges, including energy consumption restrictions, EN heterogeneity, data imbalance, and so on. These problems may lead to poor convergence accuracy, slow convergence speed, and high energy consumption during FL model training. In this paper, we focus on the performance of the FL model under energy-limited training devices, heterogeneous hardware and unbalanced data, while taking into account the social relationships between the devices. We utilize the Lyapunov optimization technique to convert the original problem into an online optimization problem, and introduce two algorithms to address this online problem. Through our analysis, we demonstrate that the optimal solution to the online problem can approximate the optimal solution to the original problem. Our simulation results validate that our proposed algorithms can achieve great performance while satisfying the energy constraints and outperforms the benchmark algorithms.

The incompleteness of the normalized difference vegetation index (NDVI) time series (TS) restricts its expanded applications in key domains. Although spatio-temporal hybrid methods show promise in TS reconstruction, reliance on auxiliary data in most existing approaches introduces errors and increases workload. Furthermore, NDVI values marked as contaminated in the quality assessment (QA) data are underutilized. Ultimately, when utilizing spatial information, most methods are ineffective for the representation of land-use changes. Considering these issues, we propose a local peak Savitzky–Golay (LPSG) method for spatio-temporal reconstruction of Landsat NDVI TS. First, we construct a local peak neighborhood weighted interpolation (LPNWI) method that fully utilizes all original values to fill gaps. Second, we design a slope change decision tree (SC-DT) method for identifying residual noise, thereby mitigating the impact of QA errors on reconstruction results. Third, multidimensional calibration with weighted spatial reference (MDC-WSR) method is proposed to enhance utilization of spatial information by improving traditional correlation coefficient calculations and generating a multiyear spatial reference, which effectively reflects land-use changes. Experiments on Landsat NDVI TS data in the Qinghai–Tibet Plateau (2013–2022) show that: 1) LPSG outperforms other methods in mitigating the impact of QA errors, preserving TS peaks and details, and maintaining spatial continuity; 2) LPSG exhibits superior performance, with average RMSE reductions ranging from 0.00018 to 0.00750 compared to other methods under both correct and incorrect QA; and 3) LPSG demonstrates good robustness under various gap conditions and effectively restores TS of pixels affected by land-use changes.

Accurate and timely monitoring of functional urban impervious surfaces (FUISs), such as ports, roads, and buildings, is essential yet challenging for complex coastal cities due to their cloudy weather and diverse land surfaces. Synthetic aperture radar (SAR) provides unique all-weather observation capabilities for prompt and regular urban mapping. However, SAR scattering information is limited to distinguish impervious surfaces with similar scattering responses but different functions. This study develops a scattering–compactness fusion network (SCFN), which integrates SAR polarimetric scattering and object compactness characteristics for enhanced FUIS recognition. Central to our approach is the scattering object compactness index, which is specifically designed to capture the distinct spatial patterns and compactness of scattering objects and complement their intrinsic scattering signatures. The dual-branch SCFN concurrently extracts and fuses object-scale scattering and compactness features using tailored network architectures. Experiments on L-band and C-band fully polarimetric ALOS-2 and GF-3 data in Hong Kong, as well as L-band dual-polarized ALOS-2 data, are undertaken to verify SCFN's effectiveness, achieving up to 8% improvement in the overall FUIS classification accuracy over baselines. The transferability of SCFN is further validated using fully polarimetric ALOS-2 data in Shenzhen, where consistent performance improvements are observed. The successful application of SCFN in both coastal cities highlights the potential of joint scattering–compactness modeling for advanced SAR-based urban mapping and its robustness across different urban landscapes.

The identification of archaeological ceramics is a relevant topic in the field of cultural heritage, and the history of archaeological ceramics can be traced back to prehistoric times. At present, there are two main methods for identifying archaeological ceramics, the empirical method and the technical one. In practice, these methods are costly or time-consuming. A systematic literature review of thirty-three studies on the identification of archaeological ceramics using machine learning is presented in this paper, including the collection process to build the dataset, the image processing of archaeological ceramic images, and the machine learning algorithms used for the classification. The main findings show the efficacy of deep learning for the automatic classification of archaeological ceramics compared to other approaches and highlight the need for more comprehensive and standardised datasets to further improve the automatic classification process.

Assessing plantar pressure is crucial for fabricating diabetic insoles and preventing diabetic foot ulcers (DFUs), which are caused by increased plantar pressure. However, the commonly used methods for assessing plantar pressure distribution involve professional sensor-based equipment and expertise, which are costly and time-consuming. Given the qualitative association between ink footprint images and plantar pressure, this study proposes using the footprint images to predict the quantitative values of dynamic plantar pressure in barefoot and 4 different insole conditions (including Nora Lunalastik EVA, Nora Lunalight A fresh, Pe-Lite, and PORON® Medical 4708) based on a multilayer perceptron (MLP) neural network model. To provide more precise insole material recommendations for specific foot regions for better plantar pressure distribution, the plantar of the foot is divided into 5 regions: the toes, metatarsal heads, medial midfoot, lateral midfoot, and heel. Patch-based MLP with localization embedding is introduced to learn the correspondence between ink density and plantar pressure information. Ground-truth data collected from 52 diabetes patients is constructed as a dataset named diabetes-footprint-to-pressure and used to train and validate the model. The mean absolute error (MAE) of the models for the barefoot and 4 insole conditions is 5.51% (33.06 kPa), 3.99% (23.94 kPa), 4.85% (29.10 kPa), 4.25% (25.50 kPa), and 3.57% (21.42 kPa) of the sensing range, respectively. Compared to traditional methods for plantar pressure assessment, this approach streamlines the process of acquiring the overall and regional dynamic plantar pressure with barefoot and 4 different insole materials. Clinicians can quickly provide recommendations on the type of insole material for individual patients.

In the context of implementing carbon quota trading policies and the rapid development of e-commerce, blockchain technology, with its traceable and tamper-proof characteristics, reduces product costs and corporate financing costs by promoting information sharing. It effectively solves the trust problem caused by information asymmetry. This study considers a supply chain consisting of a manufacturer with financial constraints and a retailer with dual-channel sales. We construct decision models for bank financing, zero-interest early payment financing, and in-house factoring financing under both benchmark conditions and blockchain technology conditions. The study explores the impact of blockchain technology on the optimal emission reduction and financing decisions of the supply chain under these three methods. The results show that regardless of whether carbon emissions exceed standards, there is a linear relationship between the manufacturer’s profit and carbon emission reduction efficiency. Additionally, there is a U-shaped relationship between the retailer’s profit and online market share. There exists a reasonable range for carbon reduction efficiency and online market share such that introducing blockchain technology can effectively increase profits for all parties involved in the green supply chain. Zero-interest early payment financing and in-house factoring financing have critical points related to production cost, while bank financing and in-house factoring financing have critical points related to carbon efficiency. As the impact of blockchain technology on carbon price increases, decision-makers will shift from bank financing to in-house factoring financing.

Unmanned Aerial Vehicle (UAV) communications have received considerable attention in recent years as it can be applied to a wide range of applications. To expedite this development, it is essential to be able to model and accurately predict the wireless channels in UAV communication. Thus, in this paper, we compare various state-of-the-art empirical large-scale path loss models and provide quantitative insights into these models. Based on the operating frequency, environment, horizontal distance, and altitude, we recommend the suitable choice of empirical models for UAV communication. These insights not only facilitate more practical theoretical studies but also motivate measurement campaigns in the future.

Voltage sag is one of the reasons that hinders the high-quality and efficient development of high-tech manufacturing users. Custom power device (CPD) is one of the effective ways to solve voltage sag. The reason why most high-tech manufacturing users are hesitant to invest CPD is that the economic benefits of CPD are vague. To solve this problem, a new premium power service (PPS) for CPD based on trial-purchase model is proposed. By trialing CPD, users can choose to purchase or cancel the devices after trialing, so as to improve the enthusiasm of users to invest premium power and ensure the smooth progress of PPS. According to the operation process of the proposed service, an optimization model of service scheme is established with the goal of maximizing the interests of power utility and user, to optimize the return ratio of users’ mitigation benefits and the length of trial service. The revenue distribution model based on Game theory (Nash negotiation) is established to obtain the revenue distribution strategies of power utility and devices company respectively. The feasibility of proposed service is verified by a semiconductor user in China. The trial-purchase premium power service could be replicable and applicable, which provides a new choice both for power utility and users.

The composition and dynamics of wetland plant communities play a critical role in maintaining the functionality of wetland ecosystems and serve as important indicators of wetland degradation and restoration. Accurately identifying wetland plant communities using remote sensing techniques remains challenging due to the complex environment and cloud contamination. Here, we applied a sample migration method based on change vector analysis and a random forest (RF) classifier incorporating SHapley Additive exPlanations (SHAP) to explore the spatiotemporal changes of wetland plant communities in the western Songnen Plain of China between 2016 and 2022, and to better understand the decision logic of the RF model. Our work achieved accurate annual wetland classification at the community scale, with an average overall accuracy of 89.5% and an average kappa coefficient of 0.87. Our analysis revealed different spatiotemporal change characteristics of wetland plant communities in the western Songnen Plain and three national nature reserves. The SHAP model showed that MOS_IRECI is the most important feature determining the prediction results of the RF model, and the importance of the features differs at global and local levels. This study confirms the feasibility of annual dynamic monitoring of wetland plant communities at a regional scale. The results are expected to provide a reference for the fine and sustainable management of wetland resources in the western Songnen Plain, as well as valuable data support for related wetland ecology research.

Microarray data is of great significance for cancer identification at the gene level. In the microarray dataset, only a small number of characteristic genomes have significant classification and identification rates for cancer. How to extract a small number of characteristic genes from a large number of microarray data is a classic NP-hard problem. This paper proposes a practical hybrid approach to implement the feature selection of gene expression from the microarray by combining the F-score algorithm and an improved artificial fish swarm algorithm with population variation (FSA-PV). Firstly, the F-score algorithm eliminates a large number of useless and redundant features in the dataset. Then, FSA-PV is discussed to obtain the ability to jump out of the local optimum while retaining the excellent feature of the subset as much as possible, and the adaptive step and visual are used to adjust the search space and to move the range of the algorithm in different environments to improve the local optimization and global optimization abilities. In addition, a naive Bayesian classifier is used to test the classification accuracy of subsets. Eight classical datasets are used to verify the performance of the proposed mechanism in the experiment part. The results reveal that the classification accuracy using the FSA-PV is significant superior to other algorithms in Breast dataset, and the classification accuracy is more than 90% in 8 cases. It further indicates the robustness and feasibility of the FSA-PV in the gene selection process.

Physical therapists play a crucial role in guiding patients through effective and safe rehabilitation processes according to medical guidelines. However, due to the therapist-patient imbalance, it is neither economical nor feasible for therapists to provide guidance to every patient during recovery sessions. Automated assessment of physical rehabilitation can help with this problem, but accurately quantifying patients’ training movements and providing meaningful feedback poses a challenge. In this paper, an Expert-knowledge-based Graph Convolutional approach is proposed to automate the assessment of the quality of physical rehabilitation exercises. This approach utilizes experts’ knowledge to improve the spatial feature extraction ability of the Graph Convolutional module and a Gated pooling module for feature aggregation. Additionally, a Transformer module is employed to capture long-range temporal dependencies in the movements. The attention scores and weight matrix obtained through this approach can serve as interpretability tools to help therapists understand the assessment model and assist patients in improving their exercises. The effectiveness of the proposed method is verified on the KIMORE dataset, achieving state-of-the-art performance compared to existing models. Experimental results also illustrate the interpretability of the method in both spatial and temporal dimensions.

Capacitorless DRAM architectures based on Back-End-of-Line (BEOL)-transistors are promising for long-retention, high-density and low-power 3D DRAM solutions due to its low leakage, operational flexibility, and monolithic integration capability. Different from classical silicon-based devices, in-depth studies on the performances of nanoscale multi-gate transistors (e.g., a-InGaZnO-FET) are still barely conducted for physical description, due to the complicated multi-gating principle, finite-size effects on transport, increased variation sources and enlarged parasitic effect. Hence, high-performance multi-nanoscale (down to $\sim ~50$ nm) dual-gate a-IGZO transistors are fabricated, and a physical compact model is developed based on the surface potential for dual-gated coupling and the disordered transport with finite-size-correction. The short channel behaviors on sub-threshold swing, mobility and threshold voltage are investigated, and contact effects are validated by the transfer-line method (TLM). Regarding the specific challenge of dual-gate alignment, possible misalignment and parasitic effects on multi-device fluctuations are important of large-scale circuit design and analyzed by TCAD simulations. Besides, the bias-temperature instability (BTI) has been comprehensively investigated. In awareness of the above effects, this model bridges fabrication-based material properties and structural parameters, assisting in a threshold fluctuation-resistant operation scheme for capacitorless multi-bit memory, showing a great potential in future monolithic integration circuit design using BEOL-transistor.

Few-shot segmentation aims to segment a large number of unlabeled samples in the target domain, by leveraging the images and labels from the source domain as well as a few labeled samples from the target domain. This is pivotal in tackling the scarcity of labeled samples in remote sensing image segmentation tasks. However, prevalent few-shot segmentation methods overlook inter-domain discrepancies, do not model and leverage the relationship between samples, and often only implement binary classification but not multi-class classification directly. To address these problems, we propose a cross-domain few-shot segmentation method based on task augmentation and feature disentanglement for practical remote sensing segmentation tasks. On one hand, task augmentation, which involves increasing the diversity of the training set and generating more challenging training data, can improve the model's generalization. On the other hand, feature disentanglement, involving the extraction of domain-irrelevant features for segmentation, improves the transferability of the model. Furthermore, to flexibly capture the relationships between the segmented regions, a graph with regions as nodes and relationships between nodes as edges is constructed. Then, labels are propagated from the labeled nodes to the unlabeled nodes in the graph by label propagation algorithm to implement multi-class classification directly. We conducted experiments on two public datasets as well as a Tibetan Plateaudataset collected by our group.And the experimental results show that the proposed method leads to a significant improvement in accuracy compared to existing methods, demonstrating its effectiveness.

Managing sloping terrains worldwide presents a significant challenge due to the lack of structured management practices, particularly when integrating various fruit trees in irregular arrangements. This study addresses the complexity arising from mixed cultivation by proposing a solution utilizing Unmanned Aerial Vehicle (UAV) imagery for tree species recognition and fruit tree classification. Our approach involves equipping UAVs with multispectral and optical cameras to capture imagery over experimental sloping terrain. The collected data undergoes processing for classifying different types of fruit trees, roads, and buildings through Orthophoto images. Convolutional Neural Networks (CNN) are employed for image recognition in challenging hillside terrains, with deep neural network methods, specifically VGG-16, VGG-19, and ResNet-50, being applied and compared. VGG-16 achieved significant accuracy in multispectral imagery analysis. Subsequently, various image fusion techniques, including Brovey, Hue- Saturation-Value, Principal Components Analysis (PCA), and Gram-Schmidt, were explored, with PCA demonstrating superior performance. The study revealed that image fusion, particularly with near-infrared or red-edge bands, significantly enhanced prediction accuracy compared to standalone multispectral imagery. The combination of visible-band fused imagery with additional spectral bands yielded the highest accuracy, improving overall prediction accuracy from 0.76 to 0.92. This research provides valuable insights applicable to diverse regions grappling with challenges in managing sloping terrains and mixed fruit tree cultivation.

Stroke is a leading cause of global population mortality and disability, imposing burdens on patients and caregivers, and significantly affecting the quality of life of patients. Therefore, in this study, we aimed to explore the application of virtual reality technology in physical therapy by using immersive interactive training and designing rehabilitation modes for individual and group settings. We also aimed to provide patients with stroke with a comprehensive home-based rehabilitation and treatment plan, ultimately enhancing training effectiveness. Patients can engage in home-based rehabilitation through this system and undergo functional, mirror, and constraint-induced therapies tailored to different task contents. Simultaneously, using brain-computer interface technology, an emotion analysis mechanism was designed to map the patients’ brainwave signal data onto a two-dimensional space of positive-negative valence arousal; this approach can enable remote physical therapists to discern the patients’ emotional states during the rehabilitation process through virtual spaces, facilitating timely adjustments to rehabilitation tasks. Moreover, to prevent compromised effectiveness owing to improper training postures leading to compensation, the system offers real-time identification and recording, promptly issuing alerts when compensation occurs. The system provides a multiuser virtual rehabilitation space, enabling timely corrections and data observations, offering patients with stroke a home-based rehabilitation program, thereby realizing a localized aging care model.

The Community Land Model version 4 with carbon and nitrogen components is coupled with data assimilation research testbed to assimilate remotely sensed leaf area index (LAI), to analyze the improvement in model performance for simulating land surface variables and land–atmospheric exchange fluxes. The results demonstrate that assimilation effectively addresses the issue of significant overestimation of LAI values, particularly noticeable in regions characterized by low latitudes and dense vegetation coverage. At a global scale, the disparities between simulated and assimilated LAI relative to observational data, are measured at 0.90 and −0.07, representing 54.1% and 3.9% of the observed values, respectively. The root mean square difference (RMSD) for assimilated LAI is 1.61 comparing with the simulated LAI of 1.85. Assimilating LAI globally leads to a noteworthy 1% reduction in the mean relative difference of the global average 2-m air temperature (T2m) and a concurrent decrease of 0.15 °C in RMSD. However, at the global level, the assimilation of LAI does not yield a significant enhancement in the modeling capability of heat fluxes, although modeling capability of sensible heat (HS) slightly outperforms latent heat. Improvements in land surface variables after assimilation show significant variations at regional scales due to factors such as vegetation coverage and climatic conditions. Overall, in regions characterized by periodic changes in vegetation, such as forested areas in Western Eurasian Continent (region 5), the enhancements in T2m and HS after assimilating LAI are particularly notable, with mean relative difference reduced by 7% and 20%, respectively.

In this paper, a photonic skin sensor based on a balloon-shaped optical fiber structure cascading a Fiber Bragg Grating (FBG) is proposed, which can simultaneously sense pressure and temperature. When the photonic skin feels pressure, the built-in polymer micro-cap is squeezed, causing a change in the radius of the balloon-shaped structure. At the same time, the capillary-enclosed FBG can be used for temperature compensation. The pressure sensitivity of the photon skin sensor can reach 6.556 nm/mN in the range of 0-0.5 mN and the temperature sensitivity can reach -0.192 nm/°C in the range of 30-50 °C. The demodulation of the dual parameter matrix shows that the error rates of pressure and temperature measurements are less than 0.5% and 2.01%, respectively. On this basis, the sensor is applied to robotic arm-assisted puncture experiment detection, and the five states of the puncture process are analyzed. Based on the backpropagation (BP) neural network for repeated training, the results show that its accuracy remains at 98.7%, indicating that the sensor has good robustness. The development of this sensor shows application potential in the field of robotic intelligent skin in the future.

The paper introduces two decoding techniques based on grouping for symbol flipping non-binary LDPC codes. Firstly, a new technique of adaptive grouping of variable nodes in each iteration using bit reliability and majority voting of each received symbol is presented. Grouping of variable nodes and the subsequent decoding are based on individual symbol reliability and majority voting. To form groups adaptively, the cumulative density of variable nodes is used where a high priority group is considered to contain the most unreliable variable nodes and is decoded while the second priority group contains variable nodes having a second level of reliability and so on. Secondly, a fixed grouping technique is applied to symbol flipping decoding of non-binary LDPC codes. In the fixed grouping decoding, each group contains an equal number of variable nodes, and selected symbols in each group are flipped according to pre-defined flipping criteria. Numerical results and analysis show that the proposed group-based hard decision symbol flipping decoding algorithms have the advantage of reduced computational complexity and show better performance. Therefore, the proposed algorithms can be considered for applications like data storage and mobile communication.

In this paper, the computation of graph Fourier transform centrality (GFTC) of complex network using graph filter is presented. For conventional computation method, it needs to use the non-sparse transform matrix of graph Fourier transform (GFT) to compute GFTC scores. To reduce the computational complexity of GFTC, a linear algebra method based on Frobenius norm of error matrix is applied to convert the spectral-domain GFTC computation task to vertex-domain one such that GFTC can be computed by using polynomial graph filtering method. There are two kinds of designs of graph filters to be studied. One is the graph-aware method; the other is the graph-unaware method. The computational complexity comparison and experimental results show that the proposed graph filter method is more computationally efficient than conventional GFT method because the sparsity of Laplacian matrix is used in the implementation structure. Finally, the centrality computations of social network, metro network and sensor network are used to demonstrate the effectiveness of the proposed GFTC computation method using graph filter.

Media convergence is considered an effective way to alleviate the decline of traditional media audiences. Although media convergence has nearly swept through the world’s foremost traditional media, studies reflecting its effects on individuals at the micro level are lacking. Therefore, this study established a multidimensional perceived media convergence scale to measure the actual perception of media convergence from the audience’s perspective. To achieve this, a comprehensive field survey was conducted in a county in southwestern China, yielding a sample of 966 participants. Study 1 used exploratory factor analysis and a multidimensional graded response model (MGRM) to optimize the item pool. Study 2 then tested the reliability and validity of the scale using confirmatory factor analysis and a MGRM. Moreover, the entire sample was tested and analyzed for the measurement accuracy of this scale. The final instrument contains 24 items in six dimensions: communication power, influence power, guiding power, credibility power, political propaganda power, and service power. This study demonstrated the robustness and scalability of the scale. Based on this scale, future research can provide a deep understanding of which groups of people perceive media convergence in what way and conduct causal inference or mechanism analysis in conjunction with the audience’s behaviors.

Phase unwrapping is an important research direction in fringe projection profilometry. Improving the accuracy of phase unwrapping from a single wrapped phase map has been a research focus. Existing the deep learning mathods for phase unwrapping from a single wrapped phase map suffer from accuracy issues when dealing with noise, the large variation range of phase surfaces, or isolated area. In this paper, we propose a novel approach to address these challenges. We treat the phase unwrapping problem as a semantic segmentation problem and introduce a new stage to the high resolution network. Additionally, we add an object contextual representation module. This approach allows us to predict the fringe order map from a single wrapped phase map without the need for any pre-process or post-process. Our method can accurately recover the phase information of objects under various challenging conditions. We validate the effectiveness and superiority of our approach by comparing it with Three deep learning methods for spatial phase unwrapping and one traditional spatial phase unwrapping method, qualitatively and quantitatively.

Space situation awareness (SSA) has gained much more attention in the last decades due to the increasing number of spacecraft in space, which provides powerful space surveillance, and hence, ensures the safe flight of spacecraft. Inverse synthetic aperture radar (ISAR) is an effective tool for uncooperative target recognition and has been widely used in ground-based SSA systems. Due to the increasing usage of small spacecraft in missions and the similar appearance of most spacecraft, the high-resolution imaging is more highly needed for reliable spacecraft recognition than before, which is also desired by the component-level recognition. Furthermore, the 3-D image reconstruction of spacecraft is very helpful in evaluating the structures due to the addition of the 3-D information, and hence, has become a very promising tool for deep target analysis. This article proposes a high-resolution spaceborne 3-D ISAR imaging method with fewer requirements on the antenna configuration and high computational efficiency. First, dedicated processing is designed to get high azimuthal resolution using large target-to-radar rotational angle accumulation. We address associated issues including the migration through resolution cell phenomenon and the high-order azimuthal phase modulation leading to defocusing of the resultant spacecraft images. Second, with the generated high-resolution 2-D images, we use dual-antenna interferometric processing to acquire the 3-D coordinates. To get precise coordinates of scatterers in both vertical and horizontal directions, the cross-range scaling technique is used. Finally, interferometric processing is used for locating the targets' height coordinates. Simulation results indicate that the presented method can provide 3-D images of spacecraft of higher resolution with a simple system structure than the traditional interferometric ISAR image methods.

Fine-grained fashion retrieval searches for items that share a similar attribute with the query image. Most existing methods use a pre-trained feature extractor (e.g., ResNet 50) to capture image representations. However, a pre-trained feature backbone is typically trained for image classification and object detection, which are fundamentally different tasks from fine-grained fashion retrieval. Therefore, existing methods suffer from a feature gap problem when directly using the pre-trained backbone for fine-tuning. To solve this problem, we introduce an attribute-guided multi-level attention network (AG-MAN). Specifically, we first enhance the pre-trained feature extractor to capture multi-level image embedding, thereby enriching the low-level features within these representations. Then, we propose a classification scheme where images with the same attribute, albeit with different values, are categorized into the same class. This can further alleviate the feature gap problem by perturbing object-centric feature learning. Moreover, we propose an improved attribute-guided attention module for extracting more accurate attribute-specific representations. Our model consistently outperforms existing attention based methods when assessed on the FashionAI (62.8788% in MAP), DeepFashion (8.9804% in MAP), and Zappos50k datasets (93.32% in Prediction accuracy). Especially, ours improves the most typical ASENet_V2 model by 2.12%, 0.31%, and 0.78% points in FashionAI, DeepFashion, and Zappos50k datasets, respectively. The source code is available in https://github.com/Dr-LingXiao/AG-MAN.

This study addresses $\mathcal {H}_{\infty }$ filtering against DoS attacks within the framework of singular Markov jump systems. To provide a more comprehensive depiction of the issue pertaining to cyber security, DoS attacks are perceived as occurring in an unpredictable manner and can be represented by a collection of a random variable that follows the Bernoulli distribution. Consequently, the corresponding filter error system is formulated. Utilizing a set of viable linear matrix inequalities, this research establishes criteria of stochastically admissible for filter error system, ensuring a specified $\mathcal {H}_{\infty }$ performance. A filter design method under DoS attacks is proposed. Finally, the effectiveness of this method is demonstrated through a practical example involving a tunnel diode model.

The adoption of electric vehicles (EVs) promises a reduction of carbon emissions and a crucial step towards a cleaner environment. While more EVs are expected to replace internal combustion engine vehicles to operate on the road worldwide, their adoption is inhibited by factors such as high power demand. Unregulated or poorly regulated charging of EVs can cause grid instability, especially in grids that were not initially designed to handle the charging of EVs. This calls for leveraging the available grid resources to control the charging of EVs in a manner that ensures optimal grid operation. This work proposes a distribution network-level dynamic pricing strategy for charging EVs to optimally utilize the distribution network and balance the load between residential and commercial/industrial distribution networks. Different EV charging probabilities that cause the EV load to differ from the optimal state with a mean average percentage error (MAPE) as high as 30% are explored. Simulation results show that with the dynamic pricing strategy as an incentive to the EVs users, EVs charging load will contribute to the optimal grid resources utilization.

To tackle the issues raised by detecting small targets and densely occluded targets in railroad track surface defect detection, we present an algorithm for detecting defects on railroad tracks based on the YOLOv8 model. Firstly, we enhance the model’s attention towards small and medium-sized targets by substituting replacing the original convolution with the SPD-Conv building block in the backbone network of YOLOv8n, while preserving the original network structure. Secondly, we integrate the EMA attention mechanism module into the neck component, allowing the model to leverage information from different layers of features and improve feature representation capabilities. Lastly, we substitute the original C-IOU with the Focal-SIoU loss function in YOLOv8, which adjusts the weights of positive and negative samples to penalize difficult-to-classify samples more heavily. This enhancement improves the model’s capability to accurately recognize challenging samples and ensures that the network allocates greater attention to each target instance, resulting in improved performance and effectiveness of the model. The experimental results reveal notable advancements in precision, recall, and average accuracy attained by our enhanced algorithm. Compared to the original YOLOv8n model, our enhanced algorithm demonstrates remarkable precision, recall, and average accuracy of 93.9%, 93.7%, and 94.1%, respectively. These improvements amount to 3.6%, 5.0%, and 5.7%, respectively. Notably, these enhancements are accomplished while maintaining the dimensions of the model and the parameter count. During the identification of defects on railroad track surfaces, our improved algorithm surpasses other widely used algorithms in terms of performance.

Depression is one of the most serious mental disorders affecting modern human life and is often caused by chronic stress. Dopamine system dysfunction is proposed to contribute to the pathophysiology of chronic stress, especially the ventral tegmental area (VTA) which mainly consists of dopaminergic neurons. Focused ultrasound stimulation (FUS) is a promising neuromodulation modality and multiple studies have demonstrated effective ultrasonic activation of cortical, subcortical, and related networks. However, the effects of FUS on the dopamine system and the potential link to chronic stress-induced depressive behaviors are relatively unknown. Here, we measured the effects of FUS targeting VTA on the improvement of depression-like behavior and evaluated the dopamine concentration in the downstream region - medial prefrontal cortex (mPFC). We found that targeting VTA FUS treatment alleviated chronic restraint stress (CRS) -induced anhedonia and despair behavior. Using an in vivo photometry approach, we analyzed the dopamine signal of mPFC and revealed a significant increase following the FUS, positively associated with the improvement of anhedonia behavior. FUS also protected the dopaminergic neurons in VTA from the damage caused by CRS exposure. Thus, these results demonstrated that targeting VTA FUS treatment significantly rescued the depressive-like behavior and declined dopamine level of mPFC induced by CRS. These beneficial effects of FUS might be due to protection in the DA neuron of VTA. Our findings suggest that FUS treatment could serve as a new therapeutic strategy for the treatment of stress-related disorders.

At present, most research on supply chain network design only considers the horizontal game of commodity greenness of the supply chain, and does not consider the vertical cooperative mode, and does not consider the impact of their own economic development on the environment. Therefore, the green supply chain network of transshipment manufacturing enterprises is designed. Firstly, the index of green supply chain is optimized and a more comprehensive index system is developed. Then, the K-nearest neighbor and decision tree algorithms are used to pre-process the partner data, and the partner is double-screened according to the index. To solve the large error in judging a single weight factor, the compound weight method is used to integrate the weight factors in two dimensions. Finally, the branch-and-bound algorithm mixed integer linear programming is used to solve the quantity of goods transported by the partners and the specific distribution route. Through the experimental analysis, compared with the traditional supply chain model, the supply chain network designed by the research reduces the cost expenditure by 34.26%, the pollution by 30.25%, the transportation loss by 37.12%, the transportation distance by 42.45%, and the transportation time by 28.47. It can effectively realize the sustainable development of enterprises and reduce the pollution generated by enterprises.

The traditional dynamic window approach (DWA) adopts the constant intervals for the sampling window, which limits the trajectory exploration possibility. This paper employs the twin delayed deep deterministic policy gradient (TD3) approach to generate a reinforcement-learning-based auxiliary candidate trajectory with variable sampling mechanism in the prediction domain for the automated guided vehicle (AGV). Subsequently, this auxiliary trajectory would compete with the traditional DWA sampling trajectories in the optimal evaluation. The proposed method significantly reduces computational costs while expanding the search space of the DWA scheme, which improving sampling utilization efficiency and planning effectiveness. In contrast to completed data-driven methods that directly generate planning solutions through the policy networks, the proposed method overall ensures planning effectiveness by the DWA mechanism unit and demonstrates superior generalization capabilities. Simulation results reveal that the proposed reinforcement-learning-based DWA generates the improvement with 11.97% in planning reward and 17.46% in calculation efficiency towards the traditional DWA approach, which demonstrating a significant performance improvement over the traditional DWA method in AGV local planning mission.

Video prediction is an essential vision task due to its wide applications in real-world scenarios. However, it is indeed challenging due to the inherent uncertainty and complex spatiotemporal dynamics of video content. Several state-of-the-art deep learning methods have achieved superior video prediction accuracy at the expense of huge computational cost. Hence, they are not suitable for devices with limitations in memory and computational resource. In the light of Green Artificial Intelligence (AI), more environment friendly deep learning solutions are desired to tackle the problem of large models and computational cost. In this work, we propose a novel video prediction network 3DTransLSTM, which adopts a hybrid transformer-long short-term memory (LSTM) structure to inherit the merits of both self-attention and recurrence. Three-dimensional (3D) depthwise separable convolutions are used in this hybrid structure to extract spatiotemporal features, meanwhile enhancing model efficiency. We conducted experimental studies on four popular video prediction datasets. Compared to existing methods, our proposed 3DTransLSTM achieved competitive frame prediction accuracy with significantly reduced model size, trainable parameters, and computational complexity. Moreover, we demonstrate the generalization ability of the proposed model by testing the model on dataset completely unseen in the training data.

When faced with complex scenes containing strong edge contours and noise, there are still more background residuals in the detection results of traditional algorithms, leading to a high false alarm rate. To solve the above problems, we propose an infrared dim and small target detection method that combines local feature prior and tensor train nuclear norm (TTNN). To suppress the strong edge contours, we first establish a background edge contour suppression function based on the structure tensor. Secondly, we propose a multi-frame density peak search algorithm to obtain local feature information by combining the features associated with multiple contiguous frames of the target. Then we use the local feature information and reweighting strategy to constrain the sparse components of the target signal, and describe the background low-rank components by the tensor train nuclear norm. Finally, we separate the target image from the background image using the alternating direction multiplier method. As compared with eight advanced algorithms, the method in this paper has a better background strong edge contour and a stronger noise suppression.

As the foundation of human survival, the development of agriculture has always played an important role in social development. The quality of agricultural development determines the speed of social progress. With the development of computer science, using computer technology to solve problems related to agricultural development has become an important research direction in current computer development. In recent years, remote sensing detection has received widespread attention, and the application of remote sensing detection technology in agriculture can provide great convenience for the development of agriculture. Benefit from the development of deep learning, remote sensing detection has made gigantic achievements. However, we have to face some challenges. First, deep learning depends on large scale of data with annotations, which expends inestimable human resources. In addition, as the depth of detection network increases, the amount of parameters explosively extends. In this work, we carry out the research based on the two problems. At the beginning, an active learning method considering classification and localization task is proposed. Our method can choose some few but valuable images. It uses 34.48% amount of training set, up to 97.4% baseline performance, and realize the compression of the scale of dataset, which reduces the trouble of manual labeling. Due to imaging and other factors, there exists many small objects in remote sensing images. So we add the mixed convolution, dilated convolution, and mosaic data augmentation modules into HRDNet network. Experiments on the agriculture dataset indicate that the improved algorithm can obtain about 2% higher than HRDNet. To reduce the number of parameters, we adjust the weighted sum ratio of importance scores dynamically. With the pruning ratio of 80%, the model volume has only 184 MB, degrading 70%. Model compression accelerates the detection speed for seven times on NVIDIA AGX Xavier, with a speed of 6 FPS. Our work will lay a foundation for remote sensing detection.

Brain-computer interface (BCI) based motor imagery (MI) can assist stroke patients in upper limb rehabilitation and help restore motor function to a certain extent. However, the classical MI paradigm distinguishes different limbs and cannot effectively meet the needs of upper limb rehabilitation training for patients. Therefore, this paper designed a new paradigm for three motor imagery actions targeting different joints of the unilateral upper limb, and electroencephalogram (EEG) data from 20 healthy participants were collected for research analysis. A deep neural network model combining an attention mechanism for multiple frequency bands and a deep convolutional network were proposed to adaptively assign weight to the EEG data in different frequency bands. Then feature extraction was performed for each frequency band to learn further and to classify features. This model can obtain an average accuracy of 69.2% for the subject-independent case with the triple classification in the designed fine motor imagery (FMI) dataset, which is better than other controlled methods. Furthermore, ablation experiments were conducted for each module, demonstrating the effectiveness of each module. These results manifest the feasibility of our proposed method and the potential of FMI paradigm for BCI, providing a new training tool for upper limb rehabilitation after stroke.

The application of long baseline aided strap-down inertial navigation system (SINS/LBL) has been demonstrated to be an effective solution for accumulative position errors of underwater vehicles (UVs). To address the inefficient positioning when available hydrophones are few in acoustic array, a hybrid tightly-coupled SINS/LBL is presented for underwater navigation system. The system is composed of SINS, LBL and pressure sensor (PS). The hybrid positioning model based on slant range, slant-range rate, slant-range difference and slant-range rate difference is established according to the location principle of time of arrival (TOA) and time difference of arrival (TDOA) when the number of available hydrophones is two. TDOA positioning model based on slant-range difference and slant-range rate difference is adopted when the number of available hydrophones is more than two. The two tightly-coupled SINS/LBL models could be switched with each other as the number of available hydrophones changes in acoustic array. This paper used experimental data obtained from an unmanned surface vehicle (USV) to validate the navigation performance. Compared with single tightly-coupled SINS/LBL model, the proposed method can provide faster error convergence and more accurate submerged position fixes.

We propose a method to estimate information loss when conducting histogram binning and principal component analysis (PCA) sequentially, as usually done in practice for fleet analytics. Coarser-grained histogram binning results in less data volume, fewer dimensions, but more information loss. Considering fewer principal components (PCs) results in fewer data dimensions but increased information loss. Although information loss with each step is well understood, little guidance exists on the overall information loss when conducting both steps sequentially. We use Monte Carlo simulations to regress information loss on the number of bins and PCs, given few parameters of a dataset related to its scale and correlation structure. A sensitivity study shows that information loss can be approximated well given sufficiently large datasets. Using the number of bins, PCs, and two correlation measures, we derive an empirical loss model with high accuracy. Furthermore, we demonstrate the benefits of estimating information losses and the representativeness of total loss in evaluating the accuracy of k-means clustering for a real-world customer fleet dataset. For preprocessing sensor data which are aggregated from sufficient number of samples, continuously distributed, and can be represented by Beta-distributions, we recommend not to coarsen the histogram binning before PCA.

The landslide is a widespread and devastating natural disaster, posing serious threats to human life, security, and natural assets. Investigating efficient methods for accurate landslide detection with remote sensing images has important academic and practical implications. In this article, we proposed an LS-YOLO, a novel and effective model for landslide detection with remote sensing images. We first built a multiscale landslide dataset (MSLD) and introduced random seeds in the data augmentation to increase data robustness. Considering the multiscale characteristic of landslides in remote sensing images, a multiscale feature extraction module is designed based on efficient channel attention, average pooling, and spatial separable convolution. To increase the receptive field of the model, dilated convolution is employed to the decoupled head. Specifically, the context enhancement module consisting of dilation convolutions is added to the decoupled head regression task branch, and then the improved decoupled head is to replace the coupled head in YOLOv5s. Extensive experiments show that our proposed model has high performance for multiscale landslide detection and outperforms other object detection models (faster RCNN, SSD, EfficientDet-D0, YOLOv5s, YOLOv7, and YOLOX). Compared with the baseline model YOLOv5s, the AP of the LS-YOLO for detecting landslides has increased by 2.18%–97.06%.

Synthetic data has been widely applied in the real world recently. One typical example is the creation of synthetic data for privacy concerned datasets. In this scenario, synthetic data substitute the real data which contains the privacy information, and is used to public testing for machine learning models. Another typical example is the unbalance data over-sampling which the synthetic data is generated in the region of minority samples to balance the positive and negative ratio when training the machine learning models. In this study, we concentrate on the first example, and introduce (a) the Howso engine, and (b) our proposed random projection based synthetic data generation framework. We evaluate these two algorithms on the aspects of privacy preservation and accuracy, and compare them to the two state-of-the-art synthetic data generation algorithms DataSynthesizer and Synthetic Data Vault. We show that the synthetic data generated by Howso engine has good privacy and accuracy, which results in the best overall score. On the other hand, our proposed random projection based framework can generate synthetic data with highest accuracy score, and has the fastest scalability.

Deep learning (DL)-based methods have been successfully employed as asynchronous classification algorithms in the steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) system. However, these methods often suffer from the limited amount of electroencephalography (EEG) data, leading to overfitting. This study proposes an effective data augmentation approach called EEG mask encoding (EEG-ME) to mitigate overfitting. EEG-ME forces models to learn more robust features by masking partial EEG data, leading to enhanced generalization capabilities of models. Three different network architectures, including an architecture integrating convolutional neural networks (CNN) with Transformer (CNN-Former), time domain-based CNN (tCNN), and a lightweight architecture (EEGNet) are utilized to validate the effectiveness of EEG-ME on publicly available benchmark and BETA datasets. The results demonstrate that EEG-ME significantly enhances the average classification accuracy of various DL-based methods with different data lengths of time windows on two public datasets. Specifically, CNN-Former, tCNN, and EEGNet achieve respective improvements of 3.18%, 1.42%, and 3.06% on the benchmark dataset as well as 11.09%, 3.12%, and 2.81% on the BETA dataset, with the 1-second time window as an example. The enhanced performance of SSVEP classification with EEG-ME promotes the implementation of the asynchronous SSVEP-BCI system, leading to improved robustness and flexibility in human-machine interaction.

This paper presents a stepped electromagnetic bandgap(EBG) hosted on the T-shaped stepped stub with an inverted H-shaped slot etched on the ground to enhance bandwidth and improve isolation for a compact UWB-MIMO antenna developed from our previous design. The characteristic mode analysis is used to offer physical insight into the operation phenomena taking place in the evolution of antenna. The proposed antenna has a compact size of $27\times 22\times 0.8$ mm and realize bandwidth from 3.07 GHz to 11.1 GHz, keeping the isolation more than 20 dB. Key parameters evaluating the antenna performance like radiation pattern, gain, radiation efficiency above 75% and envelope correlation coefficient below 0.05 are investigated. The simulated and measured results agree well, testifying the proposed MIMO antenna is a suitable candidate for UWB applications.

Accurate and reliable mangrove datasets are essential for the protection and management of mangrove ecosystems. Therefore, the evaluation of the current mangrove datasets and understanding the differences among them are critical. This study takes the Guangxi Beibu Gulf (GBG) and Guangdong-Hong Kong-Macao Greater Bay Area (GBA) as the study areas and analyzes the agreement and accuracy of eight mangrove datasets from 2000 to 2020 using area comparison, spatial agreement analysis, and absolute accuracy evaluation. The results show that; 1) significant differences exist in mangrove area and spatial distribution among the different mangrove datasets, with the percentage of high agreement areas ranging from 10% to 42%. 2) The overall accuracy of the evaluated mangrove datasets ranges from 56.2% to 95.6%, and the classification accuracy of mangrove datasets in inland areas is lower than the overall level. 3) There are regional differences in the quality of mangrove datasets, with the agreement and accuracy of mangrove datasets in the GBG being greater than those in the GBA. 4) Fine-scale mangrove mapping based on high-resolution remote sensing images, such as Sentinel, and global mangrove mapping based on the Google Earth Engine (GEE) cloud platform should be emphasized in the future. The findings of this study can provide guidance for data users to select appropriate mangrove datasets and a reference for future mangrove mapping research.

In this paper, an integrated device which realized the dual-mode integration of power amplifier (PA) and radio frequency (RF) switch based on GaN dual-gate (DG) structure is designed and fabricated. The integrated device provides two working modes and meets the performance requirements of PA and RF switch. In the transmit (Tx) mode, the integrated device is used as PA. At the frequency of 3.6 GHz, compared with the conventional GaN HEMT (high electron mobility transistor) as PA, due to the field modulation effect of the dual-gate structure, the current collapse is effectively suppressed, and the output power density $(P_{out})$ of the integrated device is increased from 6.90 W/mm to 7.85 W/mm, and the power added efficiency (PAE) is increased from 44.3% to 51.1%. In the receive (Rx) mode, the integrated device is used as RF switch. Compared with conventional RF switch, the isolation of integrated devices is significantly improved. At the frequency of 3.6GHz, the isolation is improved from 19.4 dB to 41.0 dB; At the frequency of 40 GHz, the isolation is improved from 3.0 dB to 29.7 dB. The integrated device provides a novel function integration scheme for RF frontends, with great potential for improving the integration and dimension of future communication systems.

Remote sensing images are the product of information obtained by various sensors, and the higher the resolution of the image, the more information it contains. Therefore, improving the resolution of the remote sensing image is conducive to identify Earth resources from the remote sensing image. In this article, we present a multiple-branch panchromatic image resolution restoration network based on the convolutional neural network to improve the spatial and spectral resolution of the panchromatic image simultaneously, named MBPRR-Net. Specifically, we adopt a multibranch structure to extract abundant features and utilize a feature channel mixing block to enhance the interaction of adjacent channels between features. Feature aggregation in our method is used to learn more effective features from each branch, and then a cubic filter is utilized to enhance the aggregated features. After feature extraction, we use a recovery architecture to generate the final image. Moreover, we utilize image super-resolution to restore spatial resolution and image colorization to restore the spectral resolution so that we can compare it with some image colorization and super-resolution methods to verify the proposed method. Experiments show that the performance of our method is outstanding in terms of visual effects and objective evaluation metrics compared with some existing excellent image super-resolution and colorization methods.

Content caching has emerged as an effective approach to combat the increasing strains on our current network infrastructure. This method is further improved when combining caching with source coding. However, additional complexity is incurred by creating this hybrid method, as the source coding component comes with associated feasibility constraints and decoding costs. This paper presents an approach to balance this complexity with the coding gains by selecting the best-performing subset of files to compress, while the others are left uncoded. This problem is shown to be NP-hard in general and difficult to solve in an iteration-free manner. To this end, two novel approaches are outlined: an iterative-based solution, which uses the features of the entropy function to derive the most suitable files to compress jointly, and a meta-heuristic version, which is based on the Genetic Algorithm. When compared to an exhaustive search, the proposed solutions are found to be sub-optimal but falling above the 90th percentile of all possible solutions on average. Significantly, the iterative method produces results within one percentile of the meta-heuristic approach yet it finds a solution 2.31 times faster. The iterative approach has an additional benefit, in that it is able to predict the relative gains when adding more files to a compression group. It is thus able to terminate prematurely if the estimated gains are less than a chosen threshold.