检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Alkhouri, Ismail R. Zhang, Xitong Wang, Rongrong Department of Computational Mathematics Science & Engineering Michigan State University Department of Electrical Engineering & Computer Science University of Michigan Ann Arbor United States Department of Computational Mathematics Science & Engineering Michigan State University United States Department of Computational Mathematics Science & Engineering Department of Mathematics Michigan State University United States

Deep Neural Networks (DNNs) have achieved remarkable success in addressing many previously unsolvable tasks. However, the storage and computational requirements associated with DNNs pose a challenge for deploying these trained models on resource-limited devices. Therefore, a plethora of compression and pruning techniques have been proposed in recent years. Low-rank decomposition techniques are among the approaches most utilized to address this problem. Compared to post-training compression, compression-promoted training is still under-explored. In this paper, we present a theoretically-justified technique termed Low-Rank Induced Training (LoRITa), that promotes low-rankness through the composition of linear layers and compresses by using singular value truncation. This is achieved without the need to change the structure at inference time or require constrained and/or additional optimization, other than the standard weight decay regularization. Moreover, LoRITa eliminates the need to (i) initialize with pre-trained models, (ii) specify rank selection prior to training, and (iii) compute SVD in each iteration. Our experimental results (i) demonstrate the effectiveness of our approach using MNIST on Fully Connected Networks, CIFAR10 on Vision Transformers, and CIFAR10/100 and ImageNet on Convolutional Neural Networks, and (ii) illustrate that we achieve either competitive or state-of-the-art results when compared to leading structured pruning and low-rank training methods in terms of FLOPs and parameters drop. Our code is available at https://***/XitongSystem/LoRITa/tree/main. © 2024, CC BY.

关键词： Constrained optimization

来源：评论

学校读者我要写书评

暂无评论

FedHZone: Federated Learning Based Hazardous Zone Detection Using Internet of Drone Things

FedHZone: Federated Learning Based Hazardous Zone Detection ...

引用

2024 IEEE International Conference on Communications, Computing, Cybersecurity and Informatics, CCCI 2024

作者： Biswas, Payel De, Debashis Obaidat, Mohammad S. Hazra, Arnab Sadoun, Balqies Jogesh Chandra Chaudhuri College Department of Computer Science Kolkata India Maulana Abul Kalam Azad University of Technology Centre of Mobile Cloud Computing Department of Computer Science and Engineering West Bengal WB Nadia India School of Computer and Communication Engineering University of Science and Technology Beijing Beijing100083 China School of Computing SRM University Department of Computational Intelligence SRM Nagar TN Kattankulathur603203 India School of Engineering The Amity University UP Noida201301 India Dr. Sudhir Chandra sur Institute of Technology & Sports Complex Department of Computer Science and Engineering Kolkata India College of Engineering Al Balqa ' Applied University Al-Salt19117 Jordan

ISBN: (纸本)9798350349832

Air pollution can affect human health, so it is necessary to predict the air quality index (AQI) in advance. In this work, air quality data collected by the Internet of Drone Things (IoDT) is predicted and analyzed to detect highly polluted zones using a distributed privacy-preserving federated learning (FL) approach. Air quality data for a particular zone is collected and then fed into a Long short-Term memory (LSTM) based model to predict the pollution level in local training. Global aggregation is done by an IoDT-mounted central server to aggregate the received models from the local IoDT level. Privacy-preserving FedHZone predicts and analyses the AQI values from the past time series dataset in a multi-drone environment. This work detects the most hazardous zone in a larger region of Jaipur, Rajasthan using real-Time data. The result shows FedHZone with stacked LSTM having ''2%-7% better accuracy than the vanilla LSTM-based model. © 2024 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Segmentation of Mammogram Images Using U-Net with Fusion of Channel and Spatial Attention Modules (U-Net CASAM) 1

引用

International Conference on Applied Soft Computing and Communication Networks, ACN 2023

作者： Robert Singh, A. Vidya, S. Hariharasitaraman, S. Athisayamani, Suganya Hsu, Fang Rong Department of Computational Intelligence SRM Institute of Science and Technology Tamil Nadu Kattankulathur India Department of Computer Science and Engineering Sri Sai Ram Institute of Technology Tamil Nadu Chennai India School of Computing Science and Artificial Intelligence VIT Bhopal University Sehore Madhya Pradesh Kothrikalan India School of Computing Sastra Deemed to Be University Tamil Nadu Thanjavur India Department of Information Engineering and Computer Science Feng Chia University Taichung Taiwan

ISBN: (数字)9789819720040

ISBN: (纸本)9789819720033

One of the most critical concerns that must be addressed worldwide is early breast cancer diagnosis because it can help the patients with a higher survival rate. Early detection of breast cancer can significantly lower treatment expenses;mammograms have been used to identify the disease. To plan treatments, segmentation assists doctors in measuring the amount of tissue in the breast. This work uses a U-Net model combining spatial attention and fused channel modules. The U-Net CASAM module is trained with the CBIS-DDSM dataset. The proposed segmentation algorithm’s performance is assessed using the combined INbreast and MIAS datasets regarding intersection over union (IoU) and Jaccard index. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Incomplete intuitionistic fuzzy behavioral group decision-making based on multigranulation probabilistic rough sets and MULTIMOORA for water quality inspection

引用

Journal of Intelligent and Fuzzy Systems 2023年第3期44卷 4537-4556页

作者： Bai, Wenhui Zhang, Chao Zhai, Yanhui Sangaiah, Arun Kumar School of Computer and Information Technology Key Lab. of Computational Intelligence and Chinese Information Processing of Ministry of Education Shanxi University Shanxi Taiyuan China International Graduate Institute of Artificial Intelligence National Yunlin University of Science and Technology Taiwan Department of Electrical and Computer Engineering Lebanese American University Byblos Lebanon

Water quality inspection (WQI) is one of the primary ways to ensure the safe utilization of water resources, and complicated data modeling, fusion and analysis play a significant role in seeking the resource with the best water quality. Nevertheless, the challenges of missing data, relatively large differences in decision results and bounded rationality owned by decision-makers (DMs) in terms of WQI still exist nowadays. Thus, from the aspect of stable and behavioral decision-making in multi-granularity incomplete intuitionistic fuzzy information systems (MG-IIFISs), the paper investigates a comprehensive multi-attribute group decision-making (MAGDM) approach for the application of WQI. First, the concept of MG-IIFISs is built by modeling MAGDM problems with intuitionistic fuzzy numbers (IFNs), then a new transformation scheme is constructed for transforming MG-IIFISs into multi-granularity intuitionistic fuzzy information systems (MG-IFISs) based on the similarity principle. Second, three types of multigranulation intuitionistic fuzzy probabilistic rough sets (MG IF PRSs) are developed by referring to the MULTIMOORA (Multi-Objective Optimization by Ratio Analysis plus the full MULTIplicative form) method. Afterwards, attribute weights are objectively calculated based on the best-worst method (BWM), and a new stable and behavioral MAGDM approach is constructed by means of the TODIM (an acronym in Portuguese for interactive multi-criteria decision-making) method. At last, a case study in the setting of WQI is conducted with the support of a UCI data set, and sensitivity analysis, comparative analysis and experimental analysis are investigated to display the validity of the proposed approach. In general, the proposed approach improves the stability of decision results via MULTIMOORA and BWM, and also fully considers the bounded rationality of DMs' psychological behaviors from the aspect of the TODIM method, which has certain advantages in the community of MAGDM studie

关键词： Water quality

来源：评论

学校读者我要写书评

暂无评论

Challenging the Anomaly Detection Paradigm: CNN-CRF Approach 2

Challenging the Anomaly Detection Paradigm: CNN-CRF Approach

引用

2nd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2024

作者： Kanagaraju, P. Nadganndi, Swati Kedar Sawant, Tukaram N. Ekambaram, Saranya Keshwani, Palak Kumar, Yadavalli Suresh SRM Institute of Science and Technology Department of Computational Intelligence Chennai India Bharati Vidyapeeth College of Engineering Department of Instrumentation Navi Mumbai India Bharati Vidyapeeth College of Engineering Department of Electronics and Telecommunication Navi Mumbai India Sri Eshwar College of Engineering Department of Computer Science and Business Systems Tamil Nadu Coimbatore India FST the ICFAI University Raipur India Aditya College of Engineering Department OfCSE Surampalem India

ISBN: (纸本)9798350327533

Since intrusion detection has gained so much popularity, scientists continue to dedicate a lot of time and energy to studying it. Even after all these years of research, the intrusion detection community still has problems. Recognizing previously unseen attack patterns is hampered by a high rate of false positives, which is an ongoing pro blem. However, new research has shown that there are exits from this impasse. Intrusion detection relies heavily on the ability to spot anomalies, or unexpected behavior, because they indicate the presence of malicious activity. Data preparation, feature selection, and model training are all parts of the suggested approach. Machine learning algorithms can analyze the data for the development of a predictive model thanks to data preparation methods. For feature selection, the proposed approach relied on principal component analysis and iterative grading. After collecting the features, CNN -CRF is utilized to train the models. The proposed method is superior to the popular algorithms CNN and CRF. There is a 98.63% success rate when using this strategy. © 2024 IEEE.

关键词： Anomaly Detection Conditional Random Fields (CRF) Principal Component Analysis (PCA)

来源：评论

学校读者我要写书评

暂无评论

Integrating Instruction-based Learning with Bidirectional LSTM for Autonomous Vehicle Pedestrian Detection

Integrating Instruction-based Learning with Bidirectional LS...

引用

2025 International Conference on Multi-Agent Systems for Collaborative Intelligence, ICMSCI 2025

作者： Shoba, B. Raghunathan, T. Peter, S.Sam Karthick, K. Akshya, J. Sundarrajan, M. Choudhry, Mani Deepak Psn College of Engineering and Technology Department of Electronics and Communication Engineering Tirunelveli India Dr.NG.p.Institute of Technology Department of Artificial Intelligence and Data Science Coimbatore India Sri Krishna College of Technology Department of Computer Science and Engineering Coimbatore India Sona College of Technology Department of Information Technology Salem India Srm Institute of Science and Technology Department of Computational Intelligence Chennai India Srm Institute of Science and Technology Department of Networking and Communications Chennai India Srm Institute of Science and Technology Department of Computing Technologies Chennai India

ISBN: (纸本)9798331509828

Pedestrian detection is a crucial aspect of autonomous vehicle (A V) safety, as it enables accurate identification and tracking of individuals in an A V environment. Effective pedestrian detection supports essential A V functions like braking, steering, and speed control to prevent accidents in urban settings. Traditional computer vision methods, such as Support Vector Machines and Random Forests, rely on pre-designed features like the Histogram of Oriented Gradients. While the existing techniques offer some performance, it face significant challenges in dynamic environments, especially under varying lighting conditions, occlusions, and diverse pedestrian poses. Traditional models also require extensive manual feature engineering and tuning, which limits scalability and adaptability. Recent advances in deep learning, particularly with Convolutional Neural Networks (CNNs), have allowed for automatic feature learning, yielding practical improvements over classical methods. However, CNNs can struggle in low light and when facing adversarial attacks, as they mainly capture spatial features. To address these limitations, this research proposes a new methodology combining CNNs for spatial feature extraction with Bidirectional Long Short-Term Memory (BiLSTM) networks to capture temporal dependencies. This hybrid approach enhances detection capabilities by learning both static and dynamic patterns. Our proposed CNN-BiLSTM model achieves a 92 % detection accuracy and an F1 score of 88.5% across diverse conditions, outperforming existing models by up to 15%. By integrating temporal features, the model reduces missed detections and false positives by up to 20%, providing a more reliable and adaptable solution for pedestrian detection in autonomous vehicles. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

AcciRadar: Real Time Automatic Accident Detection and Reporting Framework

AcciRadar: Real Time Automatic Accident Detection and Report...

引用

International Conference on computer and Communications Technologies (ICCCT)

作者： M Fathima Mubarakkaa S. Hemavathi V. Kavitha Aarthi M Varshini K Janani Sindhu P Department of Computer Science and Engineering Sri Sairam Engineering College Chennai India Department of Computational Intelligence SRM Institute of Science and Technology Kattankulathur India

ISBN: (数字)9798331537579

ISBN: (纸本)9798331537586

Every year, millions are injured or killed in road accidents, with a high economic and social cost to communities. Speeding, impaired driving (alcohol, drugs), distracted driving (texting, phone use), and adverse weather conditions significantly contribute to accidents. Accidents that occur in places with low population density, such as highways, can lead to loss of life due to the unavailability of emergency medical assistance and longer response times. Highways, where accidents often happen, now have more cameras. Therefore, utilizing these cameras can help automatically detect accidents. Thus, this proposal outlines a real-time automatic accident detection system using surveillance cameras and Convolutional Neural Networks (CNN) technology, combined with GPS for location tracking. The system would identify nearby hospitals and alert them via phone call and message, providing the accident location. This helps to save lives in danger and ensures that immediate medical care and treatment is provided to the injured person.

关键词： Road transportation Road accidents Hospitals Surveillance Cameras Real-time systems Convolutional neural networks Time factors Accidents Global Positioning System

来源：评论

学校读者我要写书评

暂无评论

An extension of the ADVISE Meta modeling framework and its application for an early-stage security analysis of a public transport supervision system

引用

Journal of Reliable Intelligent Environments 2023年第3期9卷 263-281页

作者： Mariotti, Francesco Bondavalli, Andrea Lollini, Paolo Montecchi, Leonardo Nardi, Simone Dipartimento di Matematica e Informatica ‘U. Dini’ University of Firenze Viale Morgagni 65 Firenze50134 Italy Department of Computer Science Norwegian University of Science and Technology Sem Sælands vei 7–9 Trondheim7034 Norway Computational Science Department MERMEC Engineering Srl Via Livornese 1019 San Piero a Grado Pisa56122 Italy

Early-stage security analysis can be used for a preliminary assessment of the security level of a system, thus providing useful insights to guide the whole system’s development. In this paper, we focus on a specific meta-level modeling framework for security analysis, ADVISE Meta, which allows representing a system using generic built-in blocks and relationships constituting the ontology of the framework, and to automatically derive complex low-level stochastic models representing attack steps and adversaries. In this paper, we extend the ADVISE Meta ontology to enlarge the variety of the possible attack paths and adversaries that can be represented in the framework, by modeling (i) attack patterns available in the CAPEC database, a comprehensive dictionary of known patterns of attack, and (ii) the adversaries’ profiles defined in the Threat Agent Library (TAL), a reference library which describes the characteristics of threat agents. The paper provides a detailed description of the whole process for extending the ADVISE Meta ontology, and the application of the extended modeling framework for an early-stage security analysis of a public transport supervision system. The framework enables a variety of security-oriented analyses, in particular to assess the probability that a given adversary can successfully reach a specific goal, to analyze the most probable attack path that adversaries can follow to reach a goal, to perform sensitivity analysis at varying of attack patterns and adversaries’ profiles, to compare different architectural solutions, and to identify the system’s components that can be more probably attacked by adversaries. © 2023, The Author(s).

关键词： Stochastic models

来源：评论

学校读者我要写书评

暂无评论

EyeHelp: Predicting an AOI based on Eye Gaze for Patient Assistance

引用

Procedia computer science 2025年 258卷 1486-1495页

作者： Akshay S Dhanush S Aswin G Nath Amudha J Department of Computer Science School of Computing Amrita Vishwa Vidyapeetham Mysuru Campus Karnataka India Computational Engineering Ruhr University Bochum Bochum Germany Department of Computer Science and Engineering School of Computing Amrita Vishwa Vidyapeetham Bengaluru Karnataka India

In this research, we delve into the predictive capabilities of deep learning models for classifying Areas of Interest (AOIs) based on gaze patterns in patient assistance systems. Utilizing eye-tracking technology to capture gaze data consisting of x and y coordinates along with duration, we aim to discern the specific regions of interest on the screen. With 9 distinct AOIs identified through a 9-point calibration process, we leveraged Python to preprocess the dataset. Applying FFNN, LSTM, and GRU, EyeHelp achieved promising accuracies: 80% for FFNN, 96% for LSTM, and 93% for GRU. Comprehensive observations in terms of accuracy, recall, and F1-score metrics across AOI classes are provided. EyeHelp employs an open source integrated webcam; hence the approach is open and adaptable to various healthcare environments. By integrating this technology, our system can be seamlessly deployed in clinics, hospitals, and even home care environments. The utilization of integrated webcams facilitates cost-effective and non-intrusive monitoring, offering valuable insights into patient behavior and attention patterns.

关键词： Eye Tracking Assistance Deep Learning Prediction Area Of Interest(AOI) Eye Gaze Gaze Pattern

来源：评论

学校读者我要写书评

暂无评论

Projective Exploration on Individual Stress Levels using Machine Learning 2

Projective Exploration on Individual Stress Levels using Mac...

引用

2nd International Conference on Artificial Intelligence and Smart Energy, ICAIS 2022

作者： Nagendra Prabhu, S. Singh, Amarjeet Bhavya, B. Sreeja, K. Bhargavi, P. Srm Institute of Science Technology School of Computing Department of Computational Intelligence Kattankulathur Chennai603203 India New Horizon College of Engineering Department of Computer Science and Engineering Department Bangalore India

ISBN: (纸本)9781665400527

Recently, Stress Prediction in every individual based on their profile and behaviour is a challenging task in the current sector. Current system is a manual process where it is difficult to identify the stress in the college students or employees. There is no automation for any stress prediction. System uses machine learning algorithms or AI algorithms to find out the stress levels, these technologies are used for application development. The purpose of this project is to reduce stress in students and employees. computer science has come a long way in the last few years. It is massive and multifaceted. It has been used in a variety of applications to meet the basic needs of human society. In the field of healthcare, machine learning has made significant progress. Stress is a deadly disease that kills a large number of people worldwide. How machine learning techniques can assist in reducing the risk of stress prediction, is examined in this study. © 2022 IEEE.

关键词： Personnel

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：