检索结果-内蒙古大学图书馆

SPECK 32/64: in search of better rotation constants for neural differential distinguishers

CRYPTOLOGIA 2025年

作者： Sarkar, Arpita Bhattacharyya, Malay Mishra, Girish Garain, Utpal Pal, Saibal K. Bandyopadhyay, Sanghamitra Pal, Nikhil R. Indian Stat Inst Kolkata ISI Ctr Artificial Intelligence & Machine Learning Kolkata India Indian Inst Technol MARS Lab Gauhati India Indian Stat Inst Machine Intelligence Unit Kolkata India Indian Stat Inst Technol Innovat Hub Data Sci Big Data Analyt Kolkata India Indian Stat Inst Kolkata India Lancet COVID 19 Commiss India Task Force New Delhi India Govt India Def RD Org DRDO Sci Anal Grp SAG Delhi India Def RD Org Sci Anal Grp New Delhi India Indian Stat Inst ISI Kolkata India Def Res & Dev Org Sci Anal Grp Delhi India DRDO Delhi India Delhi Univ New Delhi India Indian Natl Sci Acad INSA New Delhi India Natl Acad Sci India NASI Prayagraj Uttar Pradesh India Indian Natl Acad Engineers INAE Gurugram India Indian Acad Sci IASc Bengaluru Karnataka India Inst Elect & Elect Engineers IEEE Piscataway NJ USA World Acad Sci TWAS Trieste Italy Int Assoc Pattern Recognit IAPR Kolkata India West Bengal Acad Sci & Technol Kolkata West Bengal India Indian Stat Inst ISI Elect & Commun Sci Unit Kolkata India World Acad Sci Trieste Italy

The lightweight cipher, SPECK 32/64, is designed to be a secure cipher that can be used in a resource-constrained environment. SPECK 32/64 involves a pair of rotation constant parameters (alpha,beta). Different investigators, including the original designers, suggested different choices of (alpha,beta) primarily keeping the implementation issues in mind. The objective of this study is to check whether these choices compromise the security of SPECK-like ciphers, namely SPECK 32/64 and SPECK 64/128. For this, using both neural differential distinguisher (NDD) and pure differential distinguisher (PDD), we establish the following: (i) The pair of rotation constants suggested by the designers are by no means optimal while tested on both NDD and PDD. (ii) Based on our limited experiments, the rotation constant pair (6, 4) is found to be a better choice for differential cryptanalysis. (iii) With the increase in the number of rounds, different choices become equally effective. (iv) Choices of alpha and beta that are close to multiples of 8 can provide computational efficiency, however, they are not necessarily good for differential attacks (v) The NDD and PDD behave very consistently on round-reduced SPECK 32/64, and generally, NDD is found to perform better than PDD.

关键词： ARX cipher differential cryptanalysis lightweight block cipher neural distinguisher SPECK

来源：评论

学校读者我要写书评

暂无评论

Survivability of industrial internet of things using machine learning and smart contracts

引用

COMPUTERS & ELECTRICAL ENGINEERING 2023年第1期107卷

作者： Priyadarshini, Ishaani Kumar, Raghvendra Alkhayyat, Ahmed Sharma, Rohit Yadav, Kusum Alkwai, Lulwah M. Kumar, Sachin Univ Calif Berkeley Sch Informat Berkeley CA USA GIET Univ Dept Comp Sci & Engn Gunupur Odisha India Islamic Univ Coll Tech Engn Najaf Iraq SRM Inst Sci & Technol Fac Engn & Technol Dept Elect & Commun Engn NCR CampusDelhi NCR CampusDelhi Meerut Rd Ghaziabad UP India Univ Hail Coll Comp Sci & Engn Hail Saudi Arabia South Ural State Univ Big Data & Machine Learning Lab Chelyabinsk Russia

Due to data collection, there is a potential risk concerning security and privacy, so IoT reliability and survivability are of utmost concern. In this paper, we address the concern using two methods. The first method is device identification, which uses an extensive set of machine learning algo-rithms for identifying IoT devices. The algorithms include Logistic Regression, K-Nearest Neighbour, Support Vector Classifier, Random Forest, Gradient Boosting, AdaBoost, Light Gradient Boosting machine, Extreme Gradient Boosting Convolution Neural Networks, and Long Short Term Memory are used for device identification. The performance of these models has been evaluated using multiple statistical measures, and weobserve that LSTM outperforms all other baseline models. The second method proposed for ensuring the survivability of theIoT environ-ment is a blockchain-based architecture for smart homes to ensuretransparency and data protection.

关键词： Industrial IoT environment CNN LSTM Blockchain technology Smart contracts

来源：评论

学校读者我要写书评

暂无评论

MULTIPLE-SETS SPLIT QUASI-CONVEX FEASIBILITY PROBLEMS: ADAPTIVE SUBGRADIENT METHODS WITH CONVERGENCE GUARANTEE

JOURNAL OF NONLINEAR AND VARIATIONAL ANALYSIS

引用

JOURNAL OF NONLINEAR AND VARIATIONAL ANALYSIS 2022年第2期6卷 15-33页

作者： Hu, Yaohua Li, Gang Li, Minghua Yu, Carisa Kwok Wai Shenzhen Univ Guangdong Key Lab Intelligent Informat Proc Shenzhen Key Lab Adv Machine Learning & Applicat Shenzhen 518060 Peoples R China Zhejiang Sci Tech Univ Dept Math Hangzhou 310018 Peoples R China Chongqing Univ Arts & Sci Sch Math & Big Data Chongqing Peoples R China Hang Seng Univ Hong Kong Dept Math Stat & Insurance Shatin Hong Kong Peoples R China

In this paper, we consider a multiple-sets split quasi-convex feasibility problem (MSSQFP), which is to find a point such that itself and its image under a linear transformation fall within two families of sublevel sets of quasi-convex functions in the space and the image space, respectively. A unified framework of the adaptive subgradient methods with general control schemes is proposed to solve the MSSQFP. This paper is contributed to establish the quantitative convergence theory of adaptive subgradient methods with several general control schemes. An interesting finding is disclosed by the iteration complexity results that the stochastic control enjoys both advantages of low computational cost requirement and low iteration complexity. In addition, a notion of the Ho center dot lder-type bounded error bound property for the MSSQFP is introduced, and the linear/sublinear convergence rates for the adaptive subgradient methods to a feasible solution of the MSSQFP is established.

关键词： Convergence rate Iteration complexity Quasi-convex optimization Split feasibility prob- lem Subgradient method

来源：评论

学校读者我要写书评

暂无评论

Automated cyberattack detection using optimal ensemble deep learning model

引用

TRANSACTIONS ON EMERGING TELECOMMUNICATIONS TECHNOLOGIES 2024年第4期35卷

作者： Vaiyapuri, Thavavel Shankar, K. Rajendran, Surendran Kumar, Sachin Gaur, Vimal Gupta, Deepak Alharbi, Meshal Prince Sattam Bin Abdulaziz Univ Coll Comp Engn & Sci Al Kharj Saudi Arabia Saveetha Inst Med & Tech Sci Saveetha Sch Engn Dept Comp Sci & Engn Chennai India South Ural State Univ Big Data & Machine Learning Lab Chelyabinsk Russia Maharaja Surajmal Inst Technol Dept Comp Sci & Engn New Delhi India Maharaja Agrasen Inst Technol Dept Comp Sci & Engn New Delhi India Prince Sattam Bin Abdulaziz Univ Coll Comp Engn & Sci Dept Comp Sci Al Kharj Saudi Arabia

In recent times, the Industrial Internet of Things (IIoT) has developed significantly. In the application of automation, and intelligence, industrial digitalization introduced cyber risks, and the varied and complex industrial IoT platform presented a novel attack surface for network invaders. Several Intrusion Detection Systems (IDS) were advanced recently as many computer networks exposure to privacy and security threats. Availability, data confidentiality, and integrity, the damage will happen in case of IDS prevention failure. Traditional methods were ineffective in dealing with advanced attacks. Advanced deep learning (DL) methods were designed for automatic ID and abnormal behavior detection of networks. Therefore, this article focuses on the design of Improved Reptile Search Optimization with Ensemble Deep learning based Cybersecurity (IRSO-EDLCS) technique in the IIoT environment. The major aim of the IRSO-EDLCS technique lies in the accurate identification of cyberattacks in the IIoT environment. To accomplish this, the presented IRSO-EDLCS technique performs IRSO algorithm-based feature selection (IRSO-FS) technique. In addition, the IRSO-EDLCS technique performs an ensemble of three DL models namely deep belief network (DBN), bidirectional gated recurrent unit (bigRU), and autoencoder (AE). The hyperparameter tuning process is performed by a modified gray wolf optimizer (MGWO) to enhance detection process. To exhibit the improved performance of the IRSO-EDLCS algorithm, a wide range of simulations were performed on the benchmark database. The experimental outcomes depict the betterment of the IRSO-EDLCS technique over other existing models.

关键词： Cybersecurity

来源：评论

学校读者我要写书评

暂无评论

Microgrids 4.0: digitalization of microgrid with IoT and recent technology interventions

引用

IET COMMUNICATIONS 2024年第0期

作者： Negi, Gaurav Singh Singh, Rajesh Gehlot, Anita Malik, Praveen Sharma, Rohit Obaid, Ahmed J. Alferaidi, Ali Alharbi, Yasser Kumar, Sachin Uttaranchal Univ Uttaranchal Inst Technol Dehra Dun Uttarakhand India Uttaranchal Univ Uttaranchal Inst Technol Div Res & Innovat Dehra Dun Uttarakhand India Univ Int Iberoamer Dept Project Management Campeche Mexico Lovely Profess Univ Sch Elect & Elect Engn Jalandhar Punjab India ABES Engn Coll Dept Elect & Commun Engn Ghaziabad India Univ Kufa Fac Comp Sci & Math Kufa Iraq Univ Hail Coll Comp Sci & Engn Hail Saudi Arabia South Ural State Univ Big Data & Machine Learning Lab Chelyabinsk Russia

Following the fourth industrial revolution and subsequent developments in information and communication technology, applying intelligent techniques in microgrid is gaining popularity in academia and business worldwide. A significant amount of data is continuously generated by the widespread use of internet of things (IoT) technologies and sensor networks in microgrids. This data includes essential information to progress the performance of microgrids. This paper includes a comprehensive review of IoT, cloud computing, big data, artificial intelligence, machine learning, blockchain in microgrid and the concepts of digital twin and metaverse and their applications. By aiding in the design, operation management, and maintenance of microgrids, these methods offer a potent tool for managing the massive historical data and real-time data stream in a proficient and protected way. They also facilitate microgrids operation. Contextual awareness, security, and resilient operation are important for microgrids, therefore their possible improvement in light of these intelligent approaches is comprehensively examined. A theoretical implementation for managing robust operation of microgrids is then described. The discussion of recommendations in microgrids concludes.

关键词： big data blockchain internet of things

来源：评论

学校读者我要写书评

暂无评论

ABA-FWI: Augmented Boundary Attention for Full Waveform Inversion

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2024年 62卷

作者： Xu, Qiong Min, Fan Pan, Shulin Zhang, Xing-Yi Song, Guojie Wang, Ke Wu, Xindong Southwest Petr Univ Sch Comp Sci & Software Engn Chengdu 610500 Peoples R China Southwest Petr Univ Inst Artificial Intelligence Sch Comp Sci & Software Engn Lab Machine Learning Chengdu 610500 Peoples R China Southwest Petr Univ Sch Earth Sci & Technol Chengdu 610500 Peoples R China Southwest Petr Univ Sch Sci Chengdu 610500 Peoples R China Sichuan Egretech Technol Co Ltd Chengdu 610041 Peoples R China Hefei Univ Technol Key Lab Knowledge Engn Big Data Sch Comp Sci & Informat Engn Minist Educ China Hefei 230009 Anhui Peoples R China Hefei Univ Technol Intelligent Interconnected Syst Lab Anhui Prov Hefei 230009 Anhui Peoples R China

Deep learning full waveform inversion (DL-FWI) is an end-to-end and time-efficient high-resolution imaging technique for subsurface media. Popular methods are often plagued by location drift and significant velocity misfits at the stratigraphic boundaries. In this study, we propose an augmented boundary attention algorithm (ABA-FWI) to focus on the key boundary information. Regarding network composition, the wavelet convolution (WTconv) layer and the spatial attention module (SAM) are incorporated into the encoder and the decoder, respectively. The WTconv layer captures low frequencies by obtaining a large receptive field without suffering from overparameterization. SAM extracts distinctive information by utilizing the interspatial relationship of features for resolution enhancement. For loss function design, our reflection coefficient tuned boundary (RCTB) loss introduces the reflection coefficient to adjust the gradient map weight of low-contrast areas. It focuses on boundary regions characterized by speed transitions to minimize errors. Results on OpenFWI, the SEG simulation, and the Marmousi II slice datasets show that our method is superior to the state-of-the-art data-driven methods, especially on boundary details. The source code is available at https://***/FanSmale/ABA-FWI.

关键词： Accuracy Source coding Transfer learning Transforms Media Feature extraction Stability analysis Reflection coefficient Spatial resolution Synthetic data Full waveform inversion low-contrast boundary spatial attention mechanism wavelet transform (WT)

来源：评论

学校读者我要写书评

暂无评论

Attention-based unsupervised prompt learning for SAM in leaf disease segmentation

引用

KNOWLEDGE-BASED SYSTEMS 2025年 320卷

作者： Tian, Luda Yuan, Yingchun En, Qing Ma, Wei Zhang, Guanghui Liang, Fangfang Hebei Agr Univ Sch Informat Sci & Technol Baoding 071001 Peoples R China Hebei Key Lab Agr Big Data Baoding 071001 Peoples R China Hebei Engn Res Ctr Agr Remote Sensing Applicat Baoding 071001 Peoples R China Carleton Univ Sch Comp Sci Artificial Intelligence & Machine Learning AIML La Ottawa ON K1S 5B6 Canada Beijing Univ Technol Fac Informat Technol Beijing 100124 Peoples R China

In modern agriculture, leaf disease segmentation is crucial for crop disease management and yield improvement. Since most deep learning-based segmentation models require extensive annotations, pursuing unsupervised methods becomes a practical solution. Although advanced models like the Segment Anything Model (SAM) can produce precise class-agnostic masks without annotations, their reliance on human-interactive prompts limits their utility to unsupervised disease segmentation tasks. To address this limitation, we present UPLS, a novel three-stage framework that leverages unsupervised prompt learning to enable SAM for leaf disease segmentation. This method employs unsupervised contrastive learning to progressively segment leaf and disease areas and automatically generate disease prompts for SAM. Concretely, for fine and intricate lesions, we utilize an improved high-frequency attention mechanism to extract high-frequency features from the leaf area, and construct contrastive losses between global priors and foreground/background features. Furthermore, we devise a strategy to automatically generate disease-related prompts from the segmented leaf and initial lesion regions, enabling SAM to refine disease boundaries without human intervention. Experiments on three public plant disease segmentation datasets show that UPLS outperforms existing non-fully supervised segmentation methods in accuracy and robustness. Source code is available at https://***/Tianluda/ UPLS.

关键词： Unsupervised learning Contrastive learning Prompt learning SAM Leaf disease segmentation

来源：评论

学校读者我要写书评

暂无评论

DSGD-CECA: Decentralized SGD with Communication-Optimal Exact Consensus Algorithm

arXiv

引用

arXiv 2023年

作者： Ding, Lisang Jin, Kexin Ying, Bicheng Yuan, Kun Yin, Wotao Department of Mathematics University of California Los AngelesCA United States Department of Mathematics Princeton University PrincetonNJ United States Google Inc. Los AngelesCA United States Center for Machine Learning Research Peking University Beijing China AI for Science Institute Beijing China National Engineering Labratory for Big Data Analytics and Applications Beijing China Decision Intelligence Lab. Alibaba US BellevueWA United States

Decentralized Stochastic Gradient Descent (SGD) is an emerging neural network training approach that enables multiple agents to train a model collaboratively and simultaneously. Rather than using a central parameter server to collect gradients from all the agents, each agent keeps a copy of the model parameters and communicates with a small number of other agents to exchange model updates. Their communication, governed by the communication topology and gossip weight matrices, facilitates the exchange of model updates. The state-of-the-art approach uses the dynamic one-peer exponential-2 topology, achieving faster training times and improved scalability than the ring, grid, torus, and hypercube topologies. However, this approach requires a power-of-2 number of agents, which is impractical at scale. In this paper, we remove this restriction and propose Decentralized SGD with Communication-optimal Exact Consensus Algorithm (DSGD-CECA), which works for any number of agents while still achieving state-of-the-art properties. In particular, DSGD-CECA incurs a unit per-iteration communication overhead and an Õ(n3) transient iteration complexity. Our proof is based on newly discovered properties of gossip weight matrices and a novel approach to combine them with DSGD’s convergence analysis. Numerical experiments show the efficiency of DSGD-CECA. Copyright © 2023, The Authors. All rights reserved.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

TPSN: Transformer-based multi-Prototype Search Network for few-shot semantic segmentation

引用

COMPUTERS & ELECTRICAL ENGINEERING 2022年 103卷

作者： Wang, Wenjian Duan, Lijuan En, Qing Zhang, Baochang Liang, Fangfang Beijing Univ Technol Fac Informat Technol Beijing Peoples R China Beijing Key Lab Trusted Comp Beijing Peoples R China Natl Engn Lab Key Technol Informat Secur Level Pro Beijing Peoples R China Beihang Univ Inst Artificial Intelligence Beijing Peoples R China Carleton Univ Sch Comp Sci Artificialtemp Intelligence & Machine Learning AIM Ottawa ON K1S 5B6 Canada Hebei Agr Univ Hebei Key Lab Agr Big Data Baoding Peoples R China

Few-shot semantic segmentation aims to segment new objects in the image with limited annotations. Typically, in metric-based few-shot learning, the expression of categories is obtained by averaging global support object information. However, a single prototype cannot accurately describe a category. Meanwhile, simple foreground averaging operations also ignore the dependencies between objects and their surroundings. In this paper, we propose a novel Transformer-based Prototype Search Network (TPSN) for few-shot segmentation. We use the transformer encoder to integrate information between different image regions and then use the decoder to express a category in terms of multiple prototypes. The multi-prototype approach can effectively alleviate the feature fluctuation caused by limited annotation data. Moreover, we use adaptive prototype search during multi-prototype extraction instead of the ordinary averaging operation compared with the previous few-shot prototype framework. This helps the network integrate the different image regions' information and fuse object features with their dependent background information, obtaining more reasonable prototype expressions. In addition, to encourage the category's prototypes to focus on different parts and maintain consistency in high-level semantics, we use the diversity and consistency loss to constrain the multi-prototype training. Experiments show that our algorithm achieves state-of-the-art performance in few-shot segmentation on two datasets: PASCAL-5(i) and COCO-20(i).

关键词： Semantic segmentation Few-shot learning Vision transformer Multiple prototypes

来源：评论

学校读者我要写书评

暂无评论

Multistage Polymerization Network for Multiperson Pose Estimation

引用

JOURNAL OF SENSORS 2021年第1期2021卷

作者： Bai, Yu-Fei Zhang, Hong-Bo Lei, Qing Du, Ji-Xiang Huaqiao Univ Sch Comp Sci & Technol Xiamen 361000 Peoples R China Huaqiao Univ Fujian Key Lab Comp Vis & Machine Learning Xiamen 361000 Peoples R China Huaqiao Univ Xiamen Key Lab Comp Vis & Pattern Recognit Xiamen 361000 Peoples R China Huaqiao Univ Fujian Key Lab Big Data Intelligence & Secur Xiamen 361000 Peoples R China

Multiperson pose estimation is an important and complex problem in computer vision. It is regarded as the problem of human skeleton joint detection and solved by the joint heat map regression network in recent years. The key of achieving accurate pose estimation is to learn robust and discriminative feature maps. Although the current methods have made significant progress through interlayer fusion and intralevel fusion of feature maps, few works pay attention to the combination of the two methods. In this paper, we propose a multistage polymerization network (MPN) for multiperson pose estimation. The MPN continuously learns rich underlying spatial information by fusing features within the layers. The MPN also adds hierarchical connections between feature maps at the same resolution for interlayer fusion, so as to reuse low-level spatial information and refine high-level semantic information to obtain accurate keypoint representation. In addition, we observe a lack of connection between the output low-level information and the high-level information. To solve this problem, an effective shuffled attention mechanism (SAM) is proposed. The shuffle aims to promote the cross-channel information exchange between pyramid feature maps, while attention makes a trade-off between the low-level and high-level representations of the output features. As a result, the relationship between the space and the channel of the feature map is further enhanced. Evaluation of the proposed method is carried out on public datasets, and experimental results show that our method has better performance than current methods.

关键词： Polymerization

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：