最新刊期
卷 23 ， 期 5 ， 2024
- “在农业识别领域，研究者提出了一种基于改进Resnet34的小麦开花期品种识别模型，通过优化网络结构和引入注意力机制，显著提升了识别准确率，为小麦品种识别提供了有效解决方案。”摘要：To solve the problems of low efficiency， low accuracy， and insufficient related research in traditional recognition methods， a wheat flowering period variety recognition model based on improved Resnet34 is proposed. Firstly， to address the problem that existing agricultural recognition models have a large number of parameters that are not conducive to deployment on mobile devices， an improved Inceptionv1 module is used to replace the second convolutional block of the basic residual block of the Resnet34 network， reducing the model parameter count by about half； Secondly， in response to the problem of decreased recognition accuracy after the reduction of model parameters， ECA and simAM attention mechanisms are added to the model to improve the accuracy of wheat flowering stage variety recognition through effective extraction of wheat features. The experimental results show that the proposed model has an average recognition accuracy of 95.7% on the wheat flowering stage dataset， which is 2.1% higher than the original Resnet34 model. Compared with the efficientnetv2_s， MobileNet-v2， and GoogLeNet models， the accuracy has been improved by 2.4%， 3.2%， and 5.0%， respectively. The proposed model has better feature extraction ability and provides an effective method for identifying wheat varieties during the flowering period.关键词：Resnet34;wheat flowering;variety identification;ECA;simAM
- 29
- |
- 1
- |
- 0
发布时间：2024-06-17 -
Lightweight Network for Point Cloud Classification Based on Gridding and Surface Features Encoder AI导读
“在深度学习领域，针对点云分类问题，提出了轻量级网络GridPoint，通过栅格化模块和表面特征编码，实现高精度分类，同时显著降低参数量和计算量，具有实用价值。”摘要：Point clouds carry rich three-dimensional features， and their classification problem has always been a hot topic in the field of deep learning. The accuracy of existing point cloud classification networks is already relatively ideal， but the parameter and computational complexity are too large， which is not conducive to deployment in practical scenarios. A lightweight point cloud classification network， GridPoint， is proposed to address this issue. Firstly， design a point cloud gridding module， which divides the grid area based on the coordinate position of the points； Then expand the higher-order term function of the coordinates， encode the surface features of the original point cloud， and enhance the expression of contour features； Finally， two rounds of global pooling are used to extract local features and aggregate global features. Perform classification and ablation experiments using the classic point cloud dataset ModelNet40， ShapeNetCore， and the real dataset ScanObject NN. The experimental results show that the classification accuracy of GridPoint is close to mainstream networks such as PointNet++， with a difference between 0.3% and 2.3%； The network parameters and computational complexity are 0.11 M and 0.05 G， respectively， which are reduced by more than 81.7% and 88.9% compared to mainstream networks. They have significant advantages in lightweight and have good practical value.关键词：deep learning;point cloud classification;lightweight network;point cloud gridding;surface feature encoder- 31
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：A multi task emotion recognition model was designed to provide a more comprehensive and detailed emotion measurement tool for intelligent emotion interaction by combining Valence， Arousal， Dominance （VAD） three-dimensional continuous emotion analysis and discrete emotion classification. Firstly， utilize the relevant constraints between two recognition tasks （labeled as points in the VAD three-dimensional emotional space） to improve the recognition accuracy of the model； Then， a method and dataset for identifying multi-dimensional continuous emotions in VAD three-dimensional space were provided， utilizing their correlation for multi task joint learning， and establishing constraints between emotion categories and VAD multidimensional emotion space. Compared with traditional fixed emotion category labels， it can more comprehensively and meticulously describe emotional states， especially in dimension D， which is currently less studied； Finally， use the sentiment labels available in the sentiment category dataset FER2013 and manually added VAD annotations to measure VAD sentiment. The experiment shows that multi task learning with V and category， A and category， and D and category can significantly improve the recognition performance of the model.关键词：multi-dimensional emotion recognition;multi-task learning;VAD facial expression recognition dataset;discrete categorial emotion recognition;intelligent emotion recognition
- 7
- |
- 0
- |
- 0
发布时间：2024-06-17 - “在睡眠分期研究领域取得新进展，专家提出基于自注意力机制与双向长短时记忆网络的模型，有效提升分期准确性，为睡眠研究提供技术参考。”摘要：A sleep staging model based on self attention mechanism and bidirectional long short-term memory network is proposed to address the problem that existing models cannot fully capture transient and random waveforms in samples， as well as cannot focus on typical and important waveforms， which affects staging results. Firstly， a single stream time-frequency information learning module is constructed to automatically express the low-level representations of PSG signals， and to mine the time-invariant information and frequency features of EEG data. Then， design an adaptive feature recalibration learning module to calibrate and train the instantaneous and key waveform features that appear in the 30 second sample， giving more attention to these features and assigning greater weights. Finally， the features are sent to the sequence dependency learning module between the associated samples to learn the contextual relationships between each sleep sample， fully utilizing adjacent samples before and after to determine the current sample category. The results show that this method performs better than other mainstream models， with accuracy rates of 85.5% and 84.3% on the Sleep-edf-2013 and Sleep-edf-2018 public sleep datasets， and MF1 values of 82.1% and 79.6%， respectively，which can provide technical reference for sleep staging tasks.关键词：sleep staging;Transformer;bi-directional LSTM;electroencephalogram;single channel
- 12
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：To address the issue of high collection and training costs for training high-performance automatic speech recognition models， a method based on word coverage is proposed to minimize the data size required for the training set. This method introduces the concept of vector space models， mapping all corpus texts to a high-dimensional space， and selecting the text data with the lowest similarity by calculating the cosine distance between vectors. Then， collect audio based on the selected text data to achieve the best recognition effect using as little audio data as possible. Finally， using Hamming overlap to calculate the amount of newly added vocabulary to evaluate contribution， in order to optimize the selection method of cosine distance. The experiment shows that compared to the random speech training set filtering method， the proposed method can achieve the same word coverage while saving 21.31% of training data， and there is a strong positive correlation between the word coverage of the training set and the inference performance of the model obtained from the training set. This proves that while maintaining similar inference performance， it can effectively save the collection and training costs of the speech training set， thereby promoting the further development of automatic speech recognition technology.关键词：automatic speech recognition;vector space model;cosine distance;Hamming weight;training set minimization
- 6
- |
- 0
- |
- 0
发布时间：2024-06-17 -
Triple Matching and Contrast Regularization Algorithm for Multiple Choices in Reading Comprehension AI导读
“在多项选择研究领域，专家提出了三重匹配策略，通过对比正则方法强化正确与错误答案的差异，为提高识别准确率提供新思路。”摘要：Multiple choice consists of three parts： article， question， and option. The main task is to find the correct option among multiple options based on the article and question. Some algorithms have conducted some research on matching strategies between articles， questions， and options， but generally use paired processing or bidirectional matching methods， which cannot fully integrate articles， questions， and options. To this end， a triple matching strategy （TM） is proposed， which uses contrastive regularization （CR） method to distinguish answers and matches any element of the article， question， and answer with other elements to absorb semantic information from the other two. Experiments have shown that capturing and strengthening the differences between correct and incorrect answers through CR can enable the model to better identify correct and incorrect answers， in order to provide reference and inspiration for researchers in this field.关键词：triple matching;contrast regularization;attention;deep learning;feature fusion- 12
- |
- 0
- |
- 0
发布时间：2024-06-17 -
Multi-Scene and Multi-Obstacle UAV 3D Path Planning Based on Improved Seagull Optimization Algorithm AI导读
摘要：The goal of UAV 3D path planning is to plan an efficient and feasible flight path while avoiding obstacles and meeting constraint conditions. Therefore， an improved seagull optimization algorithm （TP-SOA） is proposed to solve the three-dimensional path planning problem of unmanned aerial vehicles （UAVs） in multiple scenarios and obstacles， taking into account the widespread application and computational complexity of UAV path planning. Firstly， a nonlinear convergence factor is introduced to adjust the iteration process of the seagull optimization algorithm， allowing individuals to maintain a high degree of randomness in the early stages of the algorithm and converge quickly in the later stages； Secondly， the Levi flight mechanism is adopted in the search method to expand the effective area of local search and improve the individual's ability to jump out of local optima； Finally， an individual optimal strategy is introduced to increase the learning process of individuals on the historical optimal individual positions and improve the optimization performance of the algorithm. The simulation experiment results show that TP-SOA can plan high-quality paths in complex multi obstacle scenarios， with higher convergence accuracy and stability compared to the control algorithm， demonstrating significant advantages.关键词：seagull optimization algorithm;Lévy flight mechanism;individual optimal strategy;3D path planning- 18
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：Aiming at the problem of slow convergence speed and poor reduction results when the social spider optimization algorithm solves the minimum attribute reduction. This paper proposed a minimum attribute reduction algorithm based on triple restraints social spider optimization （TRSSOAR）. Constrain the individuals in the population during the initialization stage， during the iteration process and at the end of the iteration respectively. First， a fitness voting strategy is proposed to optimize the initial state of the population so that most individuals in the population are in a good position； Then， in the iterative process， opposition-based learning is introduced， and a local opposition-based learning strategy is designed to improve the individual quality of the population and expand the search space； Thirdly， in order to obtain fewer reduction results， a redundancy detection strategy is used to remove redundant attributes in the reduction results； finally， experiments are conducted on nine UCI data sets and compared with four representative algorithms. The results show that the proposed algorithm performs well in terms of reduction capability， running time and convergence speed， and has certain advantages in solving the minimum attribute reduction problem.关键词：rough sets;minimum attribute reduction;social spider optimization;opposition-based learning;redundancy detection
- 5
- |
- 0
- |
- 0
发布时间：2024-06-17 - “最新研究改进人工大猩猩部队优化算法，有效提升移动机器人路径规划效率。通过Logistic混沌映射、控制参数W改进、鱼鹰优化策略融合及莱维飞行策略应用，显著优化路径长度，为机器人导航提供高效解决方案。”摘要：Aiming at the problems of weak global optimization ability in the early stage， weak convergence ability in the later stage， and easy to fall into local optima in the path planning problem of mobile robots using traditional artificial gorilla troop optimization algorithms， an improved artificial gorilla troop optimization algorithm is proposed. In the improved algorithm， to improve the quality of the initial population， logistic chaotic mapping is used to generate the population； Introduce new calculation formulas to improve the values of control parameters， making them linearly increase with the number of iterations； Integrating the position update strategy of the Osprey Optimization Algorithm to enhance information exchange between individuals in the algorithm； In the later stage of algorithm development， the Levi flight strategy is applied to update individual positions to ensure population diversity in the later stage of the algorithm. The simulation experiment results show that compared with SSA algorithm， GTO algorithm， and GWO algorithm， the improved algorithm reduces the average path obtained in M1 map environment by 9.72%， 6.07%， and 7.99%， respectively； The average path obtained in the M2 map environment has been shortened by 22.04%， 44.16%， and 50.3%， respectively， showing significant advantages.关键词：artificial gorilla troops optimizer algorithm;path planning;Logistic chaotic mapping;osprey optimization algorithm;Levy flight strategy
- 16
- |
- 0
- |
- 0
发布时间：2024-06-17 - “在直播行业视频流量激增背景下，专家设计了一种移动边缘计算任务卸载联合资源分配方案，有效降低视频上传运输成本，提高用户体验质量，为解决视频传输问题提供新方案。”摘要：With the rise of the live streaming industry， the problem of a surge in base station video traffic is becoming increasingly serious. Therefore， a joint resource allocation scheme of mobile edge computing task unloading for video transmission is designed. The original problem can be divided into real-time video task unloading decision problem and mobile edge computing resource allocation problem. This scheme utilizes the layered characteristics of scalable video encoding technology to convert video streams into base and enhancement layers through encoding， in order to adapt to the differences between user channels， reduce video upload and transportation costs， and improve user experience quality. Considering the characteristics of mobile edge computing serving users nearby， the scheme offloads video tasks to the edge for analysis， which is used to solve the problems of video layer selection， unloading base station selection and joint resource optimization. Simulation experiments show that compared with traditional schemes， the proposed scheme has significantly higher offloading feasibility， effectively reduces transmission delay and energy consumption， and improves the video transmission performance of mobile users.consumption， and improves the video transmission performance of mobile users.关键词：video traffic;mobile edge computing;scalable video encoding;unloading decision;resource allocation
- 15
- |
- 0
- |
- 0
发布时间：2024-06-17 - “西安交通大学通过“一张物联网”整合校园智能设备资源，实现智慧校园应用开发从“烟囱式”向“平台式”转变，提升建设速度，降低成本。”摘要：With the development of smart logistics and smart asset management business， universities have generally launched a large number of intelligent sensing devices. How to effectively manage and efficiently utilize these resources poses challenges for the construction of smart campuses. Xi'an Jiaotong University utilizes the "One Internet of Things" to unify a large number of IoT equipment resources with different access methods and protocols in the Innovation Port campus. Through protocol adaptation and conversion within the platform， differences in device data are eliminated， and standardized interfaces are used to support the implementation of upper level smart campus business application functions. The application of IoT platforms has gradually transformed the development of smart campus applications from traditional "chimney style" to "platform style"， achieving an increase in application construction speed and a reduction in costs. The design， construction， and online operation of the campus Internet of Things platform at Xi'an Jiaotong University have certain reference value.关键词：smart campus;Internet of Things （IoT）;protocol adaptation;standardized interfaces;platform-based
- 24
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：With the iterative upgrading and practical application of digital technology， scenario-driven has become the new engine of technological industrial innovation in the new economic era， and empowers industrial interconnection to innovate in the direction of digitization and intelligence. However， the increasing number of cooperative members in the scenario-driven industrial supply chain often affects the efficiency of inter-organizational collaboration and the quality of services due to the problems of resource and information blockage barriers， business intersection， and complicated collaborative relationships. Therefore， this paper proposes a scenario-driven business intelligence collaboration solution for industrial alliances. The solution first analyzes the new collaborative needs among multiple alliance members under scene-driven industrial interconnection， proposes a cloud platform-oriented industrial supply chain business collaboration model， and constructs a cloud platform service architecture system to support the interaction of heterogeneous business data from multiple sources. Finally， the feasibility of the cloud platform intelligent collaboration model is verified by taking the example of industrial inventory collaboration for the supply and demand of emergency epidemic prevention materials， with a view to providing ideas and insights for researchers in this field.关键词：scenario-driven;industry interconnection;intelligent collaboration;cloud platform;industrial supply chain
- 18
- |
- 0
- |
- 0
发布时间：2024-06-17 - “在轨迹数据分析领域，专家构建了基于真实车辆轨迹的知识图谱，采用自下而上和自上而下的方法实现数据融合，为轨迹数据提取与应用提供解决方案。”摘要：As time series data， trajectories contain both temporal and spatial information of trajectory points. In addition， the road network structure and weather conditions as auxiliary factors can also have an impact on the analysis of trajectory data. In order to organize and integrate these multi-source heterogeneous data to achieve analysis and mining of trajectory data， this article constructs the trajectory knowledge graphs. The bottom-up and top-down methods are adopted to integrate trajectory data and auxiliary data. The trajectory knowledge graph constructed by the two methods are saved in the graph database， facilitating subsequent data extraction and application. The knowledge graph constructed from bottom to top is a star shape； while we propose four spatial ontologies to construct the trajectory knowledge graph from top to bottom， which can effectively reflect trajectory spatial information and driving semantics.关键词：trajectory data;knowledge graph;trajectory knowledge graph;graph database
- 17
- |
- 0
- |
- 0
发布时间：2024-06-17 - “最新研究揭示了区块链和人工智能技术在基层医疗单位电子病历整合与共享中的应用潜力，为解决患者与医院利益协调、医疗数据安全和基层医疗数据科研价值挖掘等问题提供了新思路。”摘要：Exploring the application of blockchain and artificial intelligence technology in electronic medical record systems for the integration and sharing of electronic medical records in grassroots medical units， and proposing corresponding countermeasures. Taking primary healthcare units as the research object， a combination of literature search， questionnaire survey， and online survey methods was used to understand the current medical treatment process and information technology situation of primary healthcare units. Statistical analysis was conducted to identify the concerns and concerns of grassroots residents regarding the sharing of electronic medical records. The investigation results found that there are currently contradictions in the coordination of interests between patients and hospitals， difficulties in ensuring the security of medical data， and insufficient exploration of the scientific research value of grassroots medical data. It is believed that the implementation of county-level medical communities should be the premise， and blockchain and artificial intelligence related technologies should be applied in the electronic medical record system from the aspects of system permissions and functional design.关键词：primary medical units;electronic medical record;county medical community;shared;blockchain
- 12
- |
- 0
- |
- 0
发布时间：2024-06-17 - “最新研究揭示智慧教育发展新趋势，专家利用CiteSpace软件分析发现，未来研究将聚焦技术支撑与智慧学习环境，推动理论与实践紧密结合。”摘要：Through the visual analysis of the research hotspots and development trends of smart education， this paper aims to broaden the ideas for the future development and research of smart education. In this paper， 690 Chinese core journals and CSSCI journals with the theme of "smart education" are studied and analyzed. By using CiteSpace software， the research hotspots， hotspots migration and research frontiers of smart education are visualized and analyzed. It is found that the current research on smart education in China can be summarized into four aspects： theoretical research， technical support research， research on smart education intelligent learning environment research and practice research. In the future， the focus in the field of smart education will gradually turn to the research on technology support and smart learning environment， so that the theory of smart education can be closely linked with practice， and the research on learner centered learning resources and services of smart education will flourish.关键词：smart education;knowledge graph;visual analysis
- 8
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：Nowadays， with the rapid development of Internet technology， data privacy security is widely concerned. The medical industry is closely related to people's lives， and a large amount of valuable and private patient data is stored in medical information systems， and attribute-based encryption is a very good solution to the problem of fine-grained one-to-many access control encryption. However， there are some issues with current attribute-based encryption techniques， such as exposed ciphertext access policy and inefficient decryption. Therefore， this paper proposes an attribute-based encryption access control model （HO-CP-ABE） that supports access policy hiding and efficient decryption， which addresses the current problem of hiding access policy based on LSSS linear secret sharing to ensure ciphertext access policy privacy and security， and introduces outsourcing decryption technology to give most of the decryption tasks to cloud servers to relieve the computing pressure on the user side. In the performance analysis， compared with traditional solutions， this solution not only improves the security of ciphertext data storage， but also successfully reduces the user decryption overhead to a constant level， which effectively improving the overall system efficiency.关键词：attribute-based encryption;access control;privacy protection;policy hiding;outsourcing decryption
- 10
- |
- 1
- |
- 0
发布时间：2024-06-17 - “在金融安全领域，专家提出了一种基于投票机制的黑名单公有链解决方案，有效阻止异常资金在加密货币体系中流通，为金融安全体系建设提供了新思路。”摘要：Cryptocurrencies have become a financial tool used by many criminals in committing crimes. Currently， the research on preventing abnormal funds from circulating within the cryptocurrency system is not effective. Combining smart contract and editable blockchain technology， a blacklist public chain solution based on a voting mechanism is proposed. The solution first uses smart contract technology to design the voting process of whether to add a fund to the blacklist from a game theory perspective， setting measures of reward and punishment to encourage participants to choose an honest strategy. Then， for confirmed abnormal funds， the UTXO anomaly degree field is designed， and different processing measures of backtracking or assigning anomaly degrees are used according to the stages of the abnormal funds， utilizing the features of editable blockchain. Experimental results show that the average probability of transactions with the highest anomaly degree being selected in the block packaging process is only 19.5%， and the average probability of transactions originally selected being selected after being assigned the highest anomaly degree decreases to 68.4%. Meanwhile， the probability of transactions with different levels of anomaly degrees being selected shows a stepwise decreasing trend. The blacklist public chain solution based on a voting mechanism solves the problem of unreasonable basis for adding records and weak execution in the previous blacklist solutions， effectively preventing the circulation of abnormal funds.关键词：blacklist;smart contract;game theory;editable blockchain;voting mechanism
- 10
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：Federated learning， as a distributed learning framework， can jointly participate in training a global model while ensuring the local data security of each client. However， in the federation learning process， there exist malicious participants who submit wrong updates to prevent the convergence of the model or make the model fit deviate from the normal direction by poisoning attacks. The traditional subjective logic defense mechanism considers the frequency of interaction， the time of interaction and the influence on each other， ignoring the influence of multiple sources of data on the reputation evaluation results. To address this problem， this paper proposes a federal learning attack defense mechanism based on multi-weighted subjective logic. The mechanism calculates the client's contribution by Shapley value and evaluates the client's reputation in federation learning in three aspects： trustworthiness， contribution and freshness. Meanwhile， the security of the model is further improved by introducing blockchain technology to store the parameters. The experimental results show that the algorithm in this paper can accurately identify and defend against poisoning attacks under multi-source data， while retaining high model performance.关键词：federated learning;poisoning attacks;multi-weighted subjective logic models;Shapley values;blockchain
- 12
- |
- 1
- |
- 0
发布时间：2024-06-17 - “在区块链技术领域，专家提出了一种基于区块链的可信溯源方案，通过去中心化存储和隐私数据查询保护，确保了溯源数据的真实性与安全性。”摘要：Most existing traceability systems implement centralized storage and management of traceability data， and the acquisition of traceability data almost only relies on third-party trust institutions. This method has the possibility of single point of failure， information asymmetry， and intentional tampering of data， and cannot guarantee the authenticity of traceability data. To address the above issues， firstly， utilizing the decentralized and tamper proof technical features of blockchain technology， a trustworthy traceability scheme based on blockchain is proposed. By storing traceability data in a decentralized manner and designing a authenticity verification method for data uploading， the trustworthy uploading and verification of traceability data are achieved； Then， to address the issue of sharing some private data in traceability application scenarios， a authenticity verification method for privacy data queries is designed in the scheme， and proxy re encryption is introduced to implement authorization queries， effectively protecting private data and enabling secure queries； Finally， based on the Ethereum platform and utilizing smart contract technology， the design scheme was tested. The experimental results show that the proposed scheme achieves trusted storage and authorization queries of traceability data， ensuring its credibility and security.关键词：blockchain;traceability;decentralization;smart contract;ethereum
- 15
- |
- 0
- |
- 0
发布时间：2024-06-17 - “在网络安全领域，研究者提出了一种名为P4FO的抗网络指纹探测机制，利用可编程交换机技术，有效混淆网络指纹信息，为网络设备指纹保护提供了高效解决方案。”摘要：Network fingerprinting detection is a crucial intelligence-gathering step prior to conducting network attacks. However， existing network fingerprint obfuscation techniques， which are typical countermeasures against fingerprint detection activities， still face issues like high deployment complexity， non-transparency to end systems， and significant impact on network performance. Addressing these concerns， we propose a packet-based online obfuscation mechanism for resisting network fingerprint detection， named P4FO （P4-based fingerprint obfuscation mechanism）， leveraging programmable data plane technology. P4FO utilizes the flexible packet processing capabilities of programmable switches to obfuscate network fingerprint information online in a manner transparent to end systems. Building upon analyzing response rate characteristics of probing flows， the mechanism implements a two-phase fingerprint obfuscation scheme combining “recognition-reconstruction”， which integrates capabilities of active probing flow recognition， false fingerprint customization， and online fingerprint obfuscation， and it can alleviate resource constraints of programmable switches in high-speed network environments. Experiments based on real network traffic datasets show that P4FO outperforms current mainstream methods in combating network fingerprint detection， offering a more effective solution for the protection of network device fingerprints.关键词：network reconnaissance;network fingerprint obfuscation;network deception defense;programmable protocol-independent packet processing（P4）
- 15
- |
- 4
- |
- 0
发布时间：2024-06-17 - 摘要：Realizing medical data sharing can not only avoid a large waste of medical resources， but also promote medical technology innovation and improve the quality of medical services. In the face of explosive growth of medical data， cloud servers have become the first choice for medical institutions to store data due to their powerful performance. For security reasons， the data stored in the cloud server is usually encrypted， which ensures the confidentiality of the data， but is not conducive to subsequent data retrieval and sharing. In order to solve the above problems， a medical data sharing scheme with attribute-based searchable encryption is proposed. Combining blockchain technology and attribute-based searchable encryption technology， the scheme realizes fine-grained access control of user search permissions in one-to-many data sharing scenarios， and solves the problem that the fine-grained search permissions corresponding to changes in user attributes need to be revoked in time. In addition， the scheme also implements access policy hiding and multi-keyword searchable encryption， which further improves the security of the scheme and the efficiency of ciphertext retrieval. Through safety proof and performance analysis， the results show that the scheme is safe and efficient.关键词：medical data sharing;blockchain;attribute-based searchable encryption;policy hidden
- 14
- |
- 0
- |
- 0
发布时间：2024-06-17 - “在图像质量评价领域，研究者提出了一种结合颜色和纹理特征及视觉特性的改进算法，通过构建伪参考图像和多元高斯模型，有效提升了评价性能。”摘要：Concerning the problems in the traditional reference-free image quality assessment algorithms， such as only single features are used， subjective assessment results are needed in training， and poor performance for contrast distortion images， an improved algorithm is proposed based on color and texture features and visual characteristics. Firstly， a pseudo-reference image is constructed for an image to be evaluated， and both are divided into primary and secondary regions of interest according to the visual characteristics. Secondly， the regions are further divided into smaller blocks， and then different color and texture features are extracted for each block and multivariate Gaussian models are established. Finally， the model parameter distances between the main regions of interest and the secondary regions of interest are calculated as a quality score to evaluate the quality of image. Using the public datasets SPAQ and CSIQ and comparing with the traditional algorithms QAC， NIQE， IL-NIQE， etc. by calculating the values of evaluation indicators of linear Pearson coefficient， Spearman rank correlation coefficient， and root mean square error， the results show that the proposed algorithm outperforms the traditional ones.关键词：reference-free image quality assessment;visual characteristics;pseudo-reference images;MVG model;texture features;color features
- 12
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：In order to make the image transmission in the network safely， combined with the research of image encryption in the frequency domain， a new color image encryption algorithm is proposed. Firstly， the plain-text image is divided into three primary color components， and the two-dimensional discrete cosine transform （DCT） is performed to obtain the DC and AC scrambling sequences. Nextly， perform two-dimensional discrete wavelet transform （DWT） on the scrambling matrix to obtain four sub-bands matrices， then the binary matrix for diffusion is obtained by performing inverse two-dimensional wavelet transform on the scrambled sub-bands，convert the two-dimensional decimal matrix into a three-dimensional binary matrix， two binary matrices are converted into two-dimensional decimal matrix after XOR operation， and finally perform a diffusion operation to obtain a cipher-text image. The experimental results show that the designed algorithm can effectively resist plain-text attacks and unlawful attacks.关键词：color image encryption;discrete cosine transform;discrete wavelet transform;three-dimensional chaotic system
- 7
- |
- 0
- |
- 0
发布时间：2024-06-17 - “在地表水检测领域，研究人员提出了一种基于多尺度特征融合的遥感影像水体检测方法（MFWD），有效提升了不同地表环境中水体的识别精度，为动态捕捉地表水分布提供了解决方案。”摘要：Surface water plays an important role in the global ecological environment and human life. Dynamically capturing the distribution and extent of surface water on Earth is necessary. However， due to the high complexity of land surface environments， existing surface water body detection methods have limitations in applicability and accuracy， especially in highly heterogeneous regions such as urban areas， mountains， and cloud-covered areas. To improve the recognition accuracy of different types of water bodies in different land surface environments， this study proposes a water body detection method for remote sensing images based on multi-scale feature fusion （MFWD）. The proposed method first extracts multi-level features of water bodies and land surfaces based on a deep residual network model. Then， an Atrous Spatial Pyramid Pooling （ASPP） module and a Channel-Spatial Attention Mechanism （CSAM） module are designed to fully exploit advanced semantic information and capture advanced features of water bodies. Finally， cross-scale connections are utilized to fuse multi-scale low-level spatial details and high-level semantic information， obtaining comprehensive feature representations for effective water body recognition. Experiments on Sentinel-2 data demonstrate that the proposed MFWD method achieves an overall recognition accuracy of 95.6%， exhibiting improved accuracy in identifying different types of water bodies. Moreover， the detection of small-scale water bodies as well as water bodies in highly heterogeneous regions is enhanced.关键词：atrous spatial pyramid pooling;attention mechanism;multi-scale feature fusion;remote sensing images;water bodies detection
- 5
- |
- 0
- |
- 0
发布时间：2024-06-17 - “在数字化转型浪潮中，我国专家构建了公民数字素养与技能认证标准和评估体系，推出人才数字力（DCLC-DCI）认证，为提升公民数字素养提供了可借鉴的示范。”摘要：In the digital age， enhancing citizens' digital literacy and skills has become one of the important national strategies in China's digital transformation process. The research on digital literacy cultivation and evaluation in our country mainly focuses on the evaluation and countermeasures of digital literacy abilities for college students， strategies and paths to improve citizen digital literacy， etc. There is very little research and practice on practical， scalable， and easy to implement citizen digital literacy and skill evaluation systems and implementation methods. To this end， based on the concept of digital literacy， the framework of authoritative citizen digital literacy， and the practice of cultivating local digital talents in China， we will construct certification standards and evaluation systems for citizen digital literacy and skills， and achieve the implementation and transformation of research results. We will launch the digital capability level certification - talent digital power （DCLC-DCI） Digital Literacy Certification， forming a sustainable local citizen digital literacy and skills training and certification ecosystem. The local pilot practice has shown that the certification work has achieved good results， providing a typical example for China to improve its citizens' digital literacy and skills.关键词：digital literacy and skill;assessment system;digitalization capability level certification;digitalization capability for individuals
- 12
- |
- 0
- |
- 0
发布时间：2024-06-17 -
Experimentation and Exploration of A Large-scale Online Examination Program in Western Universities AI导读
“5G时代，线上教学成为教育改革热点。西部高校学生因设备限制难以实现线上考试。基于雨课堂和腾讯会议的大规模线上考试方案，要求不高，实施效果良好，为西部地区教育改革提供新思路。”摘要：With the advent of the 5G era， online teaching has gradually become a hot topic in the exploration of teaching reform and assessment models. During the epidemic， in order to ensure the smooth progress of exams， most universities have accumulated rich experience through online exams. However， for most students in western universities， due to economic conditions and other reasons， they are still unable to meet the equipment requirements for online exams. To this end， based on the principle of offline exam requirements， a large-scale online exam is carried out through the Rain Classroom exam platform based on template based propositions， combined with the Tencent conference invigilation platform. This plan does not require strict hardware equipment requirements， and divides the exam process into three stages： pre exam， exam， and post exam. Each stage is closely connected and the overall mode is mature and feasible. The large-scale examination test shows that the implementation effect of this plan is good， and it is of great significance for carrying out the large-scale online examination mode and diversified education reform in universities in the western region.关键词：online exam;epidemic prevention experience;rain classroom;Tencent Meeting- 7
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：Under the guidance of the "Guidelines for the Construction of Ideological and Political Education in Higher Education Courses"， addressing challenges such as the inefficient integration of ideological and professional education and students' tendency to superficially engage in ideological learning， this paper focuses on computer software testing laboratory projects and proposes a method for ideological and political education based on five-star teaching principles and CDIO （Conceive， Design， Implement， Operate）. The study designs various levels of testing cases， including basic normative experiments， comprehensive design experiments， and research exploration experiments. It elucidates the process of knowledge acquisition through experiments based on the principles of five-star teaching and ideological education， as well as the developmental trajectory of ideological abilities based on CDIO. Following the implementation of this approach， there is a noticeable improvement in the assessment scores of ideological abilities and student evaluations of ideological teaching.关键词：five-star teaching principles;CDIO;ideological and political education;ideological and political capabilities;computer experiment
- 17
- |
- 0
- |
- 0
发布时间：2024-06-17 - “跨模态检索领域取得新进展，学者们深入研究了不同模态间的语义关系，为实现高效检索提供新思路。”摘要：Cross modal retrieval is a key field in multimodal learning， whose main goal is to find semantic relationships between different modalities， so that it can retrieve samples with similar semantic features between different modalities. With the development of deep neural networks， cross modal retrieval has attracted the attention of many scholars. The consistency comparison of input-output queries remains a challenge due to their different modalities. To this end， first introduce the relevant concepts of cross modal retrieval， summarize the commonly used methods of cross modal retrieval based on real value representation and binary representation， and then focus on the application of deep learning models in cross modal retrieval， the main datasets and evaluation indicators of cross modal retrieval. Finally， propose the future development direction and existing main difficulties and challenges in this field， in order to provide reference and guidance for researchers in cross modal retrieval.关键词：cross-modal retrieval;deep learning;real value representation;binary representation
- 11
- |
- 0
- |
- 0
发布时间：2024-06-17 - 摘要：New technologies such as 5G， big data， and artificial intelligence have brought innovative vitality to the Internet of Things （IoT）. The deep integration of IoT with personal and family life， as well as industrial production， has brought profound changes to the entire society. With the rapid development and widespread application of the Internet of Things， the security of IoT devices has also been severely challenged， especially the security issues caused by firmware vulnerabilities of IoT devices， which involve a wide range of areas and cause serious economic losses. To this end， the current research status in this field was first summarized and organized； Then summarize the causes and vulnerability categories of firmware security issues in the Internet of Things； Finally， introduce the existing firmware vulnerability detection technologies， analyze the challenges faced by existing technologies， and provide prospects for the development of IoT firmware vulnerability security detection， in order to provide reference and inspiration for researchers in IoT firmware vulnerability detection technology.关键词：IoT firmware;vulnerability detection;static detection;symbolic execution;fuzzy testing;deep learning
- 5
- |
- 0
- |
- 0
发布时间：2024-06-17 - “在文本处理领域，深度学习技术助力生成式摘要模型发展，专家全面调查了基于DL的ATS，概述概念，总结模型及挑战，展望研究趋势，为文本处理任务提供新思路。”摘要：With the rapid development of the Internet， text data has shown exponential growth， bringing unprecedented challenges to text processing tasks such as document management， text classification， and information retrieval. Although researchers have developed various deep learning （DL） based generative summarization （ATS） models， most of the most advanced ATS models are based on the DL architecture， and there is still a lack of comprehensive literature review in the field of DL based generative text summarization. To this end， a comprehensive survey of DL based ATS was provided. Firstly， the concept of ATS is outlined， followed by a summary of typical models of DL based ATS and their main challenges and solutions. Finally， some open challenges in ATS tasks， as well as current hot and difficult issues and future research trends， are emphasized to help researchers better understand the latest developments in this field and provide guidance and inspiration for future research.关键词：automatic text summarization;deep learning;generative summarization;natural language processing;natural language generation
- 11
- |
- 2
- |
- 0
发布时间：2024-06-17
- Technical support is provided by Beijing Founder electronics co., LTD 鄂ICP备12016542号-70
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0