检索结果-内蒙古大学图书馆

Scale‐wise interaction fusion and knowledge distillation network for aerial scene recognition

CAAI Transactions on intelligence technology 2023年第4期8卷 1178-1190页

作者： Hailong Ning Tao Lei Mengyuan An Hao Sun Zhanxuan Hu Asoke K.Nandi School of Computer Science and Technology Xi'an University of Posts and TelecommunicationsShaanxi Key Laboratory of Network Data Analysis and Intelligent ProcessingXi'anChina Xi'an Key Laboratory of Big Data and Intelligent Computing Xi'anChina School of Electronic Information and Artificial Intelligence Shaanxi University of Science and TechnologyXi'anChina School of Computer Central China Normal UniversityWuhanChina Department of Electronic and Electrical Engineering Brunel University LondonLondonUK Xi'an Jiaotong University Xi'anChina

Aerial scene recognition(ASR)has attracted great attention due to its increasingly essential *** of the ASR methods adopt the multi‐scale architecture because both global and local features play great roles in ***,the existing multi‐scale methods neglect the effective interactions among different scales and various spatial locations when fusing global and local features,leading to a limited ability to deal with challenges of large‐scale variation and complex background in aerial scene *** addition,existing methods may suffer from poor generalisations due to millions of to‐belearnt parameters and inconsistent predictions between global and local *** tackle these problems,this study proposes a scale‐wise interaction fusion and knowledge distillation(SIF‐KD)network for learning robust and discriminative features with scaleinvariance and background‐independent *** main highlights of this study include two *** the one hand,a global‐local features collaborative learning scheme is devised for extracting scale‐invariance features so as to tackle the large‐scale variation problem in aerial scene ***,a plug‐and‐play multi‐scale context attention fusion module is proposed for collaboratively fusing the context information between global and local *** the other hand,a scale‐wise knowledge distillation scheme is proposed to produce more consistent predictions by distilling the predictive distribution between different scales during *** experimental results show the proposed SIF‐KD network achieves the best overall accuracy with 99.68%,98.74%and 95.47%on the UCM,AID and NWPU‐RESISC45 datasets,respectively,compared with state of the arts.

关键词： deep learning image analysis image classification information fusion

来源：评论

学校读者我要写书评

暂无评论

Enhanced Deep Smart intelligence Based Driver Drowsiness Detection Using Optimal Scalar Vector Feature Selection with Convolution Neural Network

Enhanced Deep Smart Intelligence Based Driver Drowsiness Det...

引用

artificial intelligence and Machine Learning Applications Theme: Healthcare and Internet of Things (AIMLA), International Conference on

作者： Nathiya N B.M. Brinda P.J. Raguraman C. Yeseshwini S. Dinesh K. Srinivasan CSE (Artificial Intelligence and Machine Learning) Paavai College of Engineering Namakkal Tamil Nadu CSE(Cyber Security) Paavai College of Engineering Namakkal Tamil Nadu Artificial Intelligence and Data Science Paavai College of Engineering Namakkal Tamil Nadu Artificial Intelligence and Data Science Vellore Institute of Technology Vellore Tamilnadu Computer Science and Engineering Paavai College of Engineering Namakkal Tamil Nadu

ISBN: (数字)9798331538965

ISBN: (纸本)9798331538972

Driver drowsiness detection is important aspect for avoiding accidental risk to protect safeguard for al type of transport. Through artificial technology take effect for identifying automated driver behavior detection using image screening and mentor the driver face recognitions to identifying the drowsiness impacts. Traditional models uses various aspects to find the structural points to identify the risk of evaluation, but increasing feature dimension. The pitfall of eye lids and drowsy behaviors are so tedious to recognize the changes due to segmentation failures which leads poor detection accuracy, precision ad recall rate provide higher false negatives. To resolve this problem. We propose a enhanced deep smart intelligence based drowsiness detection using optimal scalar vector feature selection (OSVFS) with Resnet 50-convolution neural network (RSN-CNN). Initially the median filer is applied to normalize the image scaling and marginalize the feature ideal margins of face lids presence to improve the contrasting. Then Eye Pit point aspect segmentation is applied to identify the face cover eyelid cover region to mark the closeness of lid space count. Further the Deep vectorized scalar feature selection is applied to extract the drowsiness lid closeness of behavior impact rate. The behavioral aspects of drowsiness of feature margins are validated with absolute scalar margin of closing eye lid counts to identify the drowsiness pattern. Finally the deep convolution neural network is applied to categories feature margins according to feature weight with the class by range of drowsiness presence. Along with the class weights, drivers are alert to avoid the risk to make safe driving system. The proposed system produce high performance by identifying behavioral properties by correcting the feature segregation to increase the false [positive as well in higher precision rate, really rate and higher detection accuracy compared to the other system.

关键词： Image segmentation Accuracy Convolution Face recognition Neural networks Vehicle safety Feature extraction Vectors Road safety Vehicles

来源：评论

学校读者我要写书评

暂无评论

A Novel Ensemble Deep Learning Framework for Breast Cancer Prediction

A Novel Ensemble Deep Learning Framework for Breast Cancer P...

引用

2023 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems, ICSES 2023

作者： Shaker Reddy, Pundru Chandra Nithyapriya, S. Sharma, Nipun Maheswari, S. Jayaram, B. Katale, Tejaswini SR University School of Computer Science and Artificial Intelligence Telangana Warangal506371 India Anna University Artificial Intelligence and Data Science Bannari Amman Institute of Technology Tamil Nadu Chennai India Presidency University Bangalore Electronics and Communication Engineering Karnataka India Sathyabama Institute of Science and Technology Department of Mathematics Tamil Nadu Chennai India Computer Science and Business System RMK Engineering College Tamil Nadu Kavaraipettai India Vishwakarma Institute of Technology Department of Computer Engineering Maharashtra Pune India

ISBN: (纸本)9798350319200

Because of its high incidence and high mortality rate, breast cancer requires reliable means of early detection and diagnosis to improve survival rates for its patients. With advantages including increased accuracy and efficiency, deep learning (DL) and machine learning (ML) methods are quickly becoming indispensable in the fight against breast cancer. However, there is a need for additional study because existing approaches have scalability and performance constraints. Utilizing a pre-trained ResNet50V2 and ensemble-based ML approaches, we offer a hybrid approach to reliable breast cancer diagnosis in this study. Learning and extracting latent trends from complex breast-cancer(BC) images is made possible by DL's incorporation, with ML strategies contributing interpretability and generalizability. We performed wide experiments utilizing a openly accessible Invasive-Ductal-Carcinoma(IDC) dataset based on breast histopathology images of varying sample sizes. The resilience and great performance of our hybrid model are supported by the results of our extensive experiments. When compared to state-of-the-art models, our 95% accuracy rate, 94.5 precision, 94.2 recall, and 94.6 F1 score are all improvements. We also found that the best ML model to use with the ResNet50V2 architecture is the Light Boosting Classifier (LGB). This study's findings provide substantial contributions to the field of breast cancer diagnosis thanks to its novel approach, in-depth performance analysis, and trustworthy evaluation. In addition, it may help doctors make better decisions, provide better care, and boost results for people with breast cancer. © 2023 IEEE.

关键词： breast histopathology-image Breast-cancer DL ML transfer-learning

来源：评论

学校读者我要写书评

暂无评论

Exploring Reinforcement Learning in Large-Scale data Processing 3

Exploring Reinforcement Learning in Large-Scale Data Process...

引用

3rd International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2023

作者： Bhambu, Pawan Kumar, Ritesh Sharmila, P. Patil, Vivek D. Khurana, Shivam Vivek, V. Vivekananda Global University Department of Computer Science & Engineering Jaipur India Maharishi School of Engineering and Technology Maharishi University of Information Technology Uttar Pradesh India Prince Shri Venkateshwara Padmavathy Engineering College Department of Science and Humanities Chennai127 India Vishwakarma Institute of Information Technology Department of Artificial Intelligence & Data Science Pune India Chitkara University Institute of Engineering and Technology Chitkara University Centre of Interdisciplinary Research in Business and Technology Punjab India Computer Science and Engineering Faculty of Engineering and Technology Karnataka Bangalore India

ISBN: (纸本)9798350319125

Reinforcement mastering (RL) is vital for big-scale fast processing. RL utilizes the availability of large datasets and dealers skilled in this information to remedy a wide variety of obligations, from automated diagnostics and self-reliant using structures to clinical information analysis. Using exploiting the traits of the modern-day significant information era, RL allows efficient and correct decision-making in massive-scale information processing and answers complicated troubles. This paper explores the use of reinforcement learning (RL) in large-scale data processing. RL is a subfield of machine learning that focuses on decision making in dynamic environments. With the increasing demand for processing large amounts of data, traditional techniques such as rule-based systems and batch processing are becoming inadequate. This has led to the emergence of RL as a potential solution due to its ability to learn and adapt to changing environments. Through a comprehensive review of existing literature, this paper discusses the applications of RL in large-scale data processing and its potential benefits. Furthermore, it examines various challenges and their potential solutions in implementing RL for data processing tasks. The paper concludes by highlighting the future research directions in this domain and the potential impact of RL on improving data processing efficiency and accuracy. It similarly examines some of the demanding situations and possibilities related to using RL for huge-scale records processing and the capability programs of RL in the future. © 2023 IEEE.

关键词： applications challenges opportunities potential processing

来源：评论

学校读者我要写书评

暂无评论

Utilizing Dynamic Spectrum Allocation to Optimize Capacity in High-Speed Fiber Optic Communications 3

Utilizing Dynamic Spectrum Allocation to Optimize Capacity i...

引用

3rd International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2023

作者： Narmadha, T. Sahoo, Laxman Yadav, Dhananjay Sharmila, P. Mahalle, Parikshit N. Parmar, Yuvraj Faculty of Engineering and Technology Department of Computer Science Engineering Karnataka Bangalore India Vivekananda Global University Department of Computer Science & Engineering Jaipur India Maharishi School of Engineering and Technology Maharishi University of Information Technology Uttar Pradesh India Prince Shri Venkateshwara Padmavathy Engineering College Department of Science and Humanities Chennai127 India Vishwakarma Institute of Information Technology Department of Artificial Intelligence & Data Science Pune India Centre of Interdisciplinary Research in Business and Technology Chitkara University Institute of Engineering and Technology Chitkara University Punjab India

ISBN: (纸本)9798350319125

Dynamic spectrum allocation can extensively improve the ability of high-speed fiber optic communications, permitting Fiber-to-the-home (FTTH) and different deployments to be optimized for max performance. This approach makes use of a decentralized approach to allocate spectrum resources on demand, taking into account scalability and efficient aid usage. With the aid of the use of a restrained set of hardware and software, dynamic spectrum allocation can quick and correctly update the communications community to meet changing demands. This generation takes gain of the increasingly more sophisticated optical additives available these days to alter the spectrum in real time rather than relying on static allocation algorithms. This capability reduces the want to frequently reassign spectrum allocations, in addition to the amount of time and sources it requires, permitting the communications network to be managed and controlled extra efficiently and efficiently. Moreover, dynamic spectrum allocation improves the overall performance of current fiber communications structure with the aid of better using current optical components, which may be applied retroactively. © 2023 IEEE.

关键词： Communications Decentralized Dynamic Retroactively Sophisticated

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Fuzzy Methodologies for Energy Efficient Routing Protocol for Wireless Sensor Networks

Hierarchical Fuzzy Methodologies for Energy Efficient Routin...

引用

2023 International Conference on Intelligent data Communication Technologies and Internet of Things, IDCIoT 2023

作者： Prabha, M. Anbarasan, M. Sunithamani, S. Saranya, K. St. Joseph College of Engineering Department of Electronics and Communication Engineering Sriperumbudur Chennai602 117 India Chennai Institute of Technology Department of Computer Science and Engineering Kundrathur Tamilnadu India Panimalar Institute of Technology Department of Artificial Intelligence and Data Science Poonamalle Chennai602 123 India

ISBN: (纸本)9781665474511

In recent years, the wireless sensor networks are widely used in numerous real-time applications such as WBAN monitoring and tracking. Recent developments in wireless networks have given rise to new and reliable methods for enhancing network lifetime, energy efficiency, and scalability. The power consumption of the entire wireless sensor network and the energy level of each individual sensor node are closely related by the commonly used clustering technique to manage the sensor networks. In order to optimize data transmission, the fuzzy C means algorithm is employed in this article to thoroughly analyze the cluster head while considering the energy that is available in each node and distance metrics from the base station. This study demonstrates how carefully choosing cluster heads and node clustering, which divides large networks into smaller clusters, can enhance the lifespan of a network. The proposed network uses a multi-hop routing approach, where each sensor node can independently collect and send data in order to address the data rate issue. The suggested cluster routing protocol was put to the test with 1000 data transmission rounds to demonstrate its strengths and weaknesses in terms of network lifetime and energy efficiency. The choice of the cluster head node, the distance between the nodes, and the amount of energy needed for subsequent data transmission are all considered to be random for each round. The simulation results show that the suggested methodology beats cutting-edge routing techniques and achieves a promising network performance. Furthermore, the effects of hierarchical cluster head selection point to the potential of our method for use in WSN in the future. The following tests were performed using computer simulation, including comparing the effect of network life on the increased number of rounds before and after the influence of an energy-efficient routing protocol, and examining performance metrics for network lifetime. © 2023 IEEE.

关键词： Energy efficiency

来源：评论

学校读者我要写书评

暂无评论

Blockchain-Powered Solar Synergy: Advancing Morocco’s Renewable Energy Vision

Blockchain-Powered Solar Synergy: Advancing Morocco’s Renew...

引用

International Conference on artificial intelligence and Smart Environment, ICAISE 2024

作者： Boutahir, Mohamed Khalifa Hessane, Abdelaaziz Youssefi, Ahmed el Alaoui, Ali Omari Farhaoui, Yousef Azrour, Mourade Engineering Science and Technology Laboratory IDMS Team Faculty of Sciences and Techniques Moulay Ismail University of Meknes Errachidia Morocco National School of Artificial Intelligence and Digital Berkane Morocco

ISBN: (纸本)9783031909207

Morocco is the first African country to adopt solar energy, but it faces challenges in optimizing solar resources due to prediction inaccuracies and inefficiencies. This paper has the objective to discuss the potential of blockchain technology to make a game change in solar energy prediction in Morocco. This paper presents the results of analyses of blockchain in enhancing data integrity for solar radiation forecasting and enabling secure peer-to-peer energy trading that can improve grid integration of renewable sources. The paper considers the technical, regulatory, and socio-economic reflections on the implementation of blockchain solutions in the solar sector of Morocco. Morocco can avail itself of blockchain’s transparency and security features to surmount the current barriers to its solar energy management in support of ambitious renewable energy goals. The study presents a conceptual framework of blockchain integration in solar energy systems and recommends further research and pilot projects. The proposed approach can significantly contribute to achieving Morocco’s vision of being a regional leader in sustainable energy. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Blockchain Morocco Renewable Energy Solar Energy Solar Radiation Prediction

来源：评论

学校读者我要写书评

暂无评论

Redefining Safety Nets: Text Detection solution for Cyberbullying Eradication

Redefining Safety Nets: Text Detection solution for Cyberbul...

引用

2024 International Conference on Distributed Systems, Computer Networks and Cybersecurity, ICDSCNC 2024

作者： Rao, K. Biksheswara Suvarnamukhi, B. Famila, S. Hariharan, Shanmugasundaram Thava Vinu, A. Vasantha, S.V. Hyderabad India Sreyas Institute of Engineering and Technology Department of Cse Hyderabad India Vel Tech Rangarajan Dr.Sagunthala R&d Institute of Science and Technology Department of Ece Chennai India Vardhaman College of Engineering Department of Artificial Intelligence and Data Science Hyderabad India Prathyusha Engineering College Department of Information Technology Thiruvallur India Vardhaman College of Engineering Department of Cse Hyderabad India

ISBN: (纸本)9798350375442

Cyberbullying, the malicious use of digital communication channels to intimidate, harass, or harm individuals, has emerged as a significant concern in the era of widespread internet usage and social media platforms. The anonymous nature of online interactions makes it easier for individuals to engage in abusive behavior, leaving victims vulnerable to emotional distress and potential psychological harm. To address this growing issue, the development of robust and efficient text detection methods for cyberbullying has become imperative. methods, contextual analysis, and user behavior analysis as potential techniques for identifying cyberbullying instances. Each method's strengths and limitations are discussed, emphasizing the importance of combining multiple techniques to enhance detection accuracy. The study underlines the significance of context-aware analysis in distinguishing between genuine cyberbullying and harm- less banter among friends. Additionally, ethical considerations, user privacy, and transparency in monitoring mechanisms are addressed to ensure a balanced approach to text detection. Finally, the abstract emphasizes the necessity of human review in the detection process to prevent false positives and negatives, fostering a more inclusive and supportive online environment. By leveraging the power of advanced natural language processing techniques, this research contributes to the ongoing efforts to combat cyberbullying effectively and safeguard the emotional well-being of internet users, particularly the most vulnerable among them © 2024 IEEE.

关键词： Anonymity

来源：评论

学校读者我要写书评

暂无评论

Estimating fish weight growth in aquaponic farming through machine learning techniques 3

Estimating fish weight growth in aquaponic farming through m...

引用

3rd IEEE International Conference on Intelligent Technologies, CONIT 2023

作者： Kumar, Purushottam Tiwari, Pranav Srinivasulu Reddy, U. Machine Learning & Data Analytics Lab CoE in Artificial Intelligence National Institute of Technology Trichy India Indian Institute of Information Technology Tiruchirappalli Computer Science and Engineering Trichy India National Institute of Technology Machine Learning & Data Analytics Lab Department of Computer Applications Trichy India

ISBN: (纸本)9798350338607

Due to the ever-growing population, rapid urbanization, unusual environmental change, and dwindling water supply, the food production from conventional farming techniques won't be able to keep up with increasing food demand for long, leading to severe food scarcity. Aquaponics is emerging as a viable alternative to food cultivation techniques but lacks efficiency, hindering its adoption on a large-scale production basis. The adoption of artificial intelligence (AI) techniques in smart aquaponic farming can further contribute to increasing the yield in aquaponics. In the study, both linear and non-linear regression algorithms such as linear regression, lasso regression, ridge regression, KNN regression, SVR, and decision tree regression were trained on aquaponic fish pond dataset, evaluated on the basis of multiple metrics like MAE, MSE, RMSE, and RMSLE, and a competitive analysis was done to determine the best algorithm for predicting the harvest weight of the fish in aquaponic ponds. The results showed that KNN regression and decision tree regression algorithms have lowest error and highest accuracy of 99.15% and 99.1% respectively, for predicting the harvest weight of the fish. The results obtained could be scaled up to commercial levels and used in regulating nutrient concentration and aquatic environmental conditions © 2023 IEEE.

关键词： Fish

来源：评论

学校读者我要写书评

暂无评论

Study on the Preparation of C4 Alkenes by Ethanol Coupling Based on Bootstrap Method 12th

Study on the Preparation of C4 Alkenes by Ethanol Coupling B...

引用

12th International Conference on Frontier Computing, FC 2022

作者： Yan, Zhendong Yan, Liqiang Ying, Jiajie Guo, Jiang Yu, Hangfeng School of Artificial Intelligence and Big Data Chongqing Metropolitan College of Science and Technology Chongqing Yongchuan China School of Construction Management Chongqing Metropolitan College of Science and Technology Chongqing Yongchuan China School of Architecture and Civil Engineering Chongqing Metropolitan College of Science and Technology Chongqing Yongchuan China

ISBN: (纸本)9789819914272

The manufacture of C4 olefins via ethanol coupling was investigated in this work [1]. The Bootstrap method of statistics was used to obtain the accurate relationship between the ethanol conversion and C4 olefins selectivity and temperature and time. First, analyze the relevant data [8], where the amount of data is too small and discrete individuals. In order to better fit ethanol conversion and C4 olefins selectivity at linear temperatures, Bootstrap method was used for statistics and self-iterative method system was used to make the fitting data closer to the real data. A visual analysis of the fitted large data set was performed to obtain the relationship between the ethanol conversion rate and temperature and the relationship between the selectivity of C4 olefins and temperature in each catalyst type. And verify the results to check the goodness of fit of the equation. When T = 350℃, the change of C4 olefins selectivity within min∈ [ 20, 273 ] was investigated [3]. By resampling, the resulting discrete data approximates the linear variation in the response. A large number of data are fitted into a continuous curve, and the relationship between C4 olefins selectivity and time can be known after analysis [4]. That is, the conversion rate of ethanol decreased significantly with time, from 43.5% in 20 min to 29.9% in 273 min. However, the selectivity of ethylene, C4 olefins, methyl benzaldehyde and methylbenzyl alcohol hardly changed over time [12]. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： data visualization

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：