检索结果-内蒙古大学图书馆

Semi-Supervised Deep Learning Based Method for Abnormality Detection in Videos 1

Semi-Supervised Deep Learning Based Method for Abnormality D...

1st International Conference on Advances in computer science, Electrical, Electronics, and Communication Technologies, CE2CT 2025

作者： Chaudhary, Deevesh Kumar, Sunil Dhaka, Vijaypal Singh Manipal University Jaipur Department of Data Science and Engineering Jaipur India Manipal University Jaipur Department of Computer Communication and Engineering Jaipur India

ISBN: (纸本)9798331518578

Anomaly detection refers to recognition of events different from normal ones for example road accident, fight, robbery, arsenal etc. Anomaly identification in real world surveillance videos is an important application of computer vision. The work proposed in the paper detects anomalous events in surveillance videos dataset and is based upon semi supervised deep learning model. We trained the model using UCF Crime dataset that consists of 950 normal videos and 950 anomalous videos. The anomaly videos in the dataset consists of 13 different types of anomalies such as Abuse, arrest, explosion, fight etc. that generally occur in real life. The anomaly in the dataset is labelled at video level and not at a specific frame in a video to define the semi supervised nature of learning paradigm. The extracted 3D features from dataset are fed into the multilayered deep learning model. Experimental results show that our approach has significant improvement over state-of-the-art approaches for anomaly detection in surveillance videos. The accuracy of model comes out to be 83.96 percent, that is improvement over other methods. © 2025 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

A Fairness Analysis of Human and AI-Generated Student Reflection Summaries 5

A Fairness Analysis of Human and AI-Generated Student Reflec...

引用

5th Workshop on Gender Bias in Natural Language Processing, GeBNLP 2024, held in conjunction with the 62nd Annual Meeting of the Association for Computational Linguistics, ACL 2024

作者： Baghel, Bhiman Kumar Narayanan, Arun Balajiee Lekshmi Yoder, Michael Miller Department of Computer Science United States Intelligent Systems Program University of Pittsburgh PA United States

ISBN: (纸本)9798891761377

This study examines the fairness of human- and AI-generated summaries of student reflections in university STEM classes, focusing on potential gender biases. Using topic modeling, we first identify topics that are more prevalent in reflections from female students and others that are more common among male students. We then analyze whether human and AI-generated summaries reflect the concerns of students of any particular gender over others. Our analysis reveals that though human-generated and extractive AI summarization techniques do not show a clear bias, abstractive AI-generated summaries exhibit a bias towards male students. Pedagogical themes are overrepresented from male reflections in these summaries, while concept-specific topics are underrepresented from female reflections. This research contributes to a deeper understanding of AI-generated bias in educational contexts, highlighting the need for future work on mitigating these biases. ©2024 Association for Computational Linguistics.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Uniform Memory Retrieval with Larger Capacity for Modern Hopfield Models 41

Uniform Memory Retrieval with Larger Capacity for Modern Hop...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Wu, Dennis Hu, Jerry Yao-Chieh Hsiao, Teng-Yun Liu, Han Department of Computer Science University of Northwestern Evanston United States Department of Physics National Taiwan University Taipei Taiwan Department of Statistics and Data Science University of Northwestern Evanston United States

We propose a two-stage memory retrieval dynamics for modern Hopfield models, termed U-Hop, with enhanced memory capacity. Our key contribution is a learnable feature map Φ which transforms the Hopfield energy function into kernel space. This transformation ensures convergence between the local minima of energy and the fixed points of retrieval dynamics within the kernel space. Consequently, the kernel norm induced by Φ serves as a novel similarity measure. It utilizes the stored memory patterns as learning data to enhance memory capacity across all modern Hopfield models. Specifically, we accomplish this by constructing a separation loss LΦ that separates the local minima of kernelized energy by separating stored memory patterns in kernel space. Methodologically, U-Hop memory retrieval process consists of: (Stage I) minimizing separation loss for a more uniformed memory (local minimum) distribution, followed by (Stage II) standard Hopfield energy minimization for memory retrieval. This results in a significant reduction of possible metastable states in the Hopfield energy function, thus enhancing memory capacity by preventing memory confusion. Empirically, with real-world datasets, we demonstrate that U-Hop outperforms all existing modern Hopfield models and SOTA similarity measures, achieving substantial improvements in both associative memory retrieval and deep learning tasks. Code is available at GitHub;future updates are on arXiv. Copyright 2024 by the author(s)

关键词： Associative storage

来源：评论

学校读者我要写书评

暂无评论

Secure Consumer-Centric Demand Response Management in Resilient Smart Grid as Industry 5.0 Application with Blockchain-Based Authentication

引用

IEEE Transactions on Consumer Electronics 2024年第1期70卷 1370-1379页

作者： Ayub, Muhammad Faizan Li, Xiong Mahmood, Khalid Shamshad, Salman Saleem, Muhammad Asad Omar, Marwan University of Electronic Science and Technology of China School of Computer Science and Engineering Sichuan Chengdu12599 China National Yunlin University of Science and Technology Graduate School of Intelligent Data Science Douliu64002 Taiwan The University of Lahore Department of Software Engineering Lahore54590 Pakistan Illinois Institute of Technology Information Technology and Management ChicagoIL60616 United States

The development of the industrial Internet of Things and smart grid networks has emphasized the importance of secure smart grid communication for the future of electric power transmission. However, the current deployment of smart meters and wireless communication methods presents several security concerns that must be addressed to ensure sustainable and resilient smart grid networks for Industry 5.0, with a focus on consumer-centric technologies. In this context, we propose a privacy-focused authentication solution for secure consumer-centric demand response management in smart grid networks, incorporating consumer electronics. Our proposed protocol utilizes blockchain-based authentication to enhance resistance against potential attacks and ensure the integrity of demand response data. Additionally, we have incorporated a PUF to prevent physical attacks. We have conducted informal and formal security analyses to verify the effectiveness of our proposed protocol against various threats and to confirm the session key security. Our comparison study has demonstrated that our protocol provides superior security and added functionalities, with lower computation and communication costs compared to similar protocols. Furthermore, we have measured the computational time for implementing the blockchain component of our protocol while considering the impact of varying numbers of block additions and transactions per block. © 1975-2011 IEEE.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

Bridging Optimal Transport and Jacobian Regularization by Optimal Trajectory for Enhanced Adversarial Defense 17th

Bridging Optimal Transport and Jacobian Regularization by ...

引用

17th Asian Conference on computer Vision, ACCV 2024

作者： M. Le, Binh Tariq, Shahroz S. Woo, Simon Department of Computer Science and Engineering Sungkyunkwan University Suwon Korea Republic of CSIRO’s Data61 Sydney Australia

ISBN: (纸本)9789819609628

Deep neural networks, particularly in vision tasks, are notably susceptible to adversarial perturbations. To overcome this challenge, developing a robust classifier is crucial. In light of the recent advancements in the robustness of classifiers, we delve deep into the intricacies of adversarial training and Jacobian regularization, two pivotal defenses. Our work is the first carefully analyzes and characterizes these two schools of approaches, both theoretically and empirically, to demonstrate how each approach impacts the robust learning of a classifier. Next, we propose our novel Optimal Transport with Jacobian regularization method, dubbed OTJR, bridging the input Jacobian regularization with the a output representation alignment by leveraging the optimal transport theory. In particular, we employ the Sliced Wasserstein distance that can efficiently push the adversarial samples’ representations closer to those of clean samples, regardless of the number of classes within the dataset. The SW distance provides the adversarial samples’ movement directions, which are much more informative and powerful for the Jacobian regularization. Our empirical evaluations set a new standard in the domain, with our method achieving commendable accuracies of 52.57% on CIFAR-10 and 28.36% on CIFAR-100 datasets under the AutoAttack. Further validating our model’s practicality, we conducted real-world tests by subjecting internet-sourced images to online adversarial attacks. These demonstrations highlight our model’s capability to counteract sophisticated adversarial perturbations, affirming its significance and applicability in real-world scenarios. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

SAND: semi-automated adaptive network defense via programmable rule generation and deployment

引用

science China(Information sciences) 2022年第7期65卷 121-138页

作者： Haoyu CHEN Deqing ZOU Hai JIN Shouhuai XU Bin YUAN National Engineering Research Center for Big Data Technology and System Services Computing Technology and System LabCluster and Grid Computing Lab School of Computer Science and TechnologyHuazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System Services Computing Technology and System LabHubei Engineering Research Center on Big Data Security School of Cyber Science and EngineeringHuazhong University of Science and Technology Department of Computer Science University of Colorado at Colorado Springs

Cyber security is dynamic as defenders often need to adapt their defense postures. The state-ofthe-art is that the adaptation of network defense is done manually(i.e., tedious and error-prone). The ideal solution is to automate adaptive network defense, which is however a difficult problem. As a first step towards automation, we propose investigating how to attain semi-automated adaptive network defense(SAND). We propose an approach extending the architecture of software-defined networking, which is centered on providing defenders with the capability to program the generation and deployment of dynamic defense rules enforced by network defense tools. We present the design and implementation of SAND, as well as the evaluation of the prototype implementation. Experimental results show that SAND can achieve agile and effective dynamic adaptations of defense rules(less than 15 ms on average for each operation), while only incurring a small performance overhead.

关键词： network defense adaptive defense automated defense programmable defense security services software-defined networking security management

来源：评论

学校读者我要写书评

暂无评论

Pairwise Learning for Autism Spectrum Disorder Imbalanced Classification 6

Pairwise Learning for Autism Spectrum Disorder Imbalanced Cl...

引用

6th IEEE Eurasia Conference on Biomedical Engineering, Healthcare and Sustainability, ECBIOS 2024

作者： Liu, Shu Wu, Qiang Yang, Xin Computational and Data Science PhD Program Middle Tennessee State University MurfreesboroTN United States University of Tennessee Department of Mathematics KnoxvilleTN United States Middle Tennessee State University Department of Computer Science MurfreesboroTN United States

ISBN: (纸本)9798350396133

This paper performs a classification task on data obtained from the Autism Brain Imaging data Exchange (ABIDE) repository. In real-world case analysis, the number of autism spectrum disorder (ASD) patients is much smaller than typically developed people. To address this issue, this paper proposes the utilization of pairwise robust support vector machine (PRSVM) algorithms to classify autism spectrum disorder (ASD) patients. In this project's experiments, the correlation matrix derived from functional magnetic resonance imaging (fMRI) data was employed as a classification feature. A comprehensive evaluation was conducted to compare the classification performance of PRSVM with various machine learning methods. The comparative analysis encompassed various aspects, including different data dimensions, imbalanced ratios, and sample sizes, providing valuable insights into the relative performance of the algorithms under different experimental conditions. The experimental results demonstrate that PRSVM can detect autistic patients more accurately when the data is imbalanced. Moreover, the results indicate that PRSVM outperforms or achieves comparable performance to other conventional classification methods in a variety of situations. Furthermore, our approach can be further improved by augmenting the training set with either exclusively normal person samples or by incorporating patient samples and normal people samples in a proportionate manner. This augmentation strategy holds promising application value, as it contributes to improving the performance and robustness of our method. © 2024 IEEE.

关键词： Diseases

来源：评论

学校读者我要写书评

暂无评论

Cardiovascular Disease Analysis Using Correlational Analysis and Association Rules Mining for In-depth Analysis to Identify Predominant Variables

Cardiovascular Disease Analysis Using Correlational Analysis...

引用

2023 International Conference on computer science, Information Technology and Engineering, ICCoSITE 2023

作者： Siswanto, Boby Soeparno, Haryono Sianipar, Nesti F. Budiharto, Widodo Bina Nusantara University BINUS Graduate Program- Computer Science Computer Science Department Jakarta Indonesia Bina Nusantara University Engineering Faculty Food Technology Department Jakarta Indonesia Bina Nusantara University School of Computer Science Computer Science Department Jakarta Indonesia

ISBN: (纸本)9798350320954

Cardiovascular disease is one of the dangerous non-communicable disorders or diseases that has become one of the causes of death worldwide. Various studies have been conducted to prevent cardiovascular disease in the world. This study analyzed cardiovascular disease medical record data from the Kaggle public dataset by implementing correlational analysis combined with association rule mining to identify variables that are the predominant cause of cardiovascular disease. Correlational analysis can analyze the interrelationships between variables in a dataset, but not in depth. Association rule mining can identify the interrelationships of variables in the form of frequent item sets, which can be calculated for their support and confidence values. The result of this study is a combination of correlation analysis with association rule mining that can identify predominant variables to cause cardiovascular disease. Found that the variable gender=woman, height=short ( © 2023 IEEE.

关键词： Association rules

来源：评论

学校读者我要写书评

暂无评论

Diversity-Optimized Group Extraction in Social Networks

引用

IEEE Transactions on Computational Social Systems 2024年第1期11卷 756-769页

作者： Hsu, Bay-Yuan Chen, Yi-Ling Ho, Ya-Chi Chang, Po-Yuan Chang, Chih-Chieh Shia, Ben-Chang Shen, Chih-Ya National Tsing Hua University Department of Industrial Engineering and Engineering Management Hsinchu30013 Taiwan National Taiwan University of Science and Technology Department of Computer Science and Information Engineering Taipei10607 Taiwan Hsinchu31040 Taiwan Fu Jen Catholic University Department of Business Administration New Taipei City24205 Taiwan National Taiwan University of Science and Technology Management Undergraduate Program Taipei10607 Taiwan National Tsing Hua University Department of Computer Science Hsinchu30013 Taiwan

In this article, we propose to study a novel research problem to boost group performance, that is, social-aware diversity-optimized group extraction (SDGE), which takes into consideration the two important factors: 1) group diversity and 2) social tightness. We prove the NP-hardness of SDGE and propose an effective algorithm, named group shrinking for diversity maximization (GSDM) with a performance guarantee, that is, GSDM is a three-approximation algorithm to the SDGE problem studied in this article. We further propose three effective pruning strategies that are able to boost the efficiency of GSDM but do not deteriorate its performance. We conduct extensive experiments on multiple large-scale real datasets to evaluate the performance of GSDM. The experimental results show that our proposed GSDM outperforms the other baseline approaches significantly, in terms of solution quality and efficiency. Moreover, the experimental results also confirm that our proposed pruning strategies indeed boost the efficiency of the algorithm. © 2014 IEEE.

关键词： Costs

来源：评论

学校读者我要写书评

暂无评论

AI and Leadership Moderator: Enhancing Work Engagement through Intelligent Experience and Sustainable Practices among Indonesian Employees 7

AI and Leadership Moderator: Enhancing Work Engagement throu...

引用

7th International Conference of computer and Informatics Engineering, IC2IE 2024

作者： Herlina, Maria Grace Iskandar, Karto Triana, Ika Bina Nusantara University Binus Business School Undergraduate Program Management Department Jakarta11480 Indonesia Bina Nusantara University School of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798331505530

Artificial intelligence (AI) integration in workplaces reshapes employee engagement and organisational performance. This study investigates the combined impact of AI experience and sustainable leadership on work engagement among Indonesian employees. Using a cross-sectional survey of 301 employees in Greater Jakarta, the research employs partial least squares structural equation modeling (PLS-SEM) to analyse the data. The study finds that while AI experience and sustainable leadership independently enhance work engagement, their interaction negatively affects it, highlighting the necessity for strategically aligning AI initiatives with sustainable leadership principles to optimise engagement. The methodology references several factors: validated AI experience items focusing on productivity and skill development, sustainable leadership items emphasising social and environmental responsibility, and work engagement items measuring enthusiasm and job satisfaction. The Rasch Model Analysis and person-measure analysis ensure the reliability and validity of these constructs. This study fills a significant gap in understanding the interaction between AI and leadership in enhancing work engagement, providing practical insights for managing technological and leadership dynamics in organisations. © 2024 IEEE.

关键词： artificial intelligence employees organizational learning sustainable leadership work engagement

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：