检索结果-内蒙古大学图书馆

3rd International Conference on Advancement in Electrical and Electronic engineering, ICAEEE 2024

作者： Rab, Raqeebir Ahmed, Towkir Alam, Arfayet Rahman, Ashikur Leshob, Abderrahmane Ahsanullah University of Science and Technology Department of Computer Science and Engineering Tejgaon Dhaka1208 Bangladesh Bangladesh University of Engineering and Technology Department of Computer Science and Engineering West Palashi Dhaka1000 Bangladesh University of Quebec in Montreal Department of Computer Science and Engineering Montreal Canada

ISBN: (纸本)9798350388282

Image enhancement refers to the process of manipulating an image to improve the visual information it contains. Images can degrade due to various factors, such as operator incompetence or low-quality picture- recording devices. The degraded images exhibit color imbalance, unwanted noise, and hue saturation disparity, especially in dark images. Furthermore, any photograph taken in low-light conditions typically falls short of attaining the desired level of visibility and essential details of the image. Most image enhancement systems depend on a predetermined set of instructions. Deep learning approaches instruct and execute certain actions during training without knowledge of the properties of individual images. Insufficient knowledge occasionally leads to the distortion of images. Conversely, an agent that utilizes Reinforcement Learning (RL) can make decisions about which actions to perform based on the input image. In order to address the aforementioned challenges, we present a novel approach in this study that use Reinforcement Learning to enhance dark images. This method accurately emulates the sequential process employed by humans during retouching. Our reinforcement learning-based agent, similar to a human expert, aims to evaluate the present state of the image prior to making any decision. Upon conducting a comprehensive analysis, we have determined that our agent, which utilizes reinforcement learning (RL), has demonstrated exceptional efficacy in maintaining the authenticity and color equilibrium of dark images. In general, the experimental results demonstrate the outstanding efficacy of the proposed approach. © 2024 IEEE.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

CR-SAM: Curvature Regularized Sharpness-Aware Minimization 38

CR-SAM: Curvature Regularized Sharpness-Aware Minimization

引用

38th AAAI Conference on Artificial Intelligence, AAAI 2024

作者： Wu, Tao Luo, Tie Wunsch, Donald C. Department of Computer Science Missouri University of Science and Technology United States Department of Electrical and Computer Engineering Missouri University of Science and Technology United States

ISBN: (纸本)1577358872

The capacity to generalize to future unseen data stands as one of the utmost crucial attributes of deep neural networks. Sharpness-Aware Minimization (SAM) aims to enhance the generalizability by minimizing worst-case loss using one-step gradient ascent as an approximation. However, as training progresses, the non-linearity of the loss landscape increases, rendering one-step gradient ascent less effective. On the other hand, multi-step gradient ascent will incur higher training cost. In this paper, we introduce a normalized Hessian trace to accurately measure the curvature of loss landscape on both training and test sets. In particular, to counter excessive non-linearity of loss landscape, we propose Curvature Regularized SAM (CR-SAM), integrating the normalized Hessian trace as a SAM regularizer. Additionally, we present an efficient way to compute the trace via finite differences with parallelism. Our theoretical analysis based on PAC-Bayes bounds establishes the regularizer’s efficacy in reducing generalization error. Empirical evaluation on CIFAR and ImageNet datasets shows that CR-SAM consistently enhances classification performance for ResNet and Vision Transformer (ViT) models across various datasets. Our code is available at https://***/TrustAIoT/CR-SAM. Copyright © 2024, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

NADR-Net: A Deep Learning Framework for Predicting Neurological Adverse Drug Reactions Using 17 Molecular Descriptors

NADR-Net: A Deep Learning Framework for Predicting Neurologi...

引用

2024 International Symposium on Computing and Intelligent Systems, SCI 2024

作者： Chaurasia, Anushka Kumar, Deepak Yogita Department of Computer Science and Engineering National Institute of Technology Meghalaya India Department of Computer and Engineering National Institute of Technology Kurukshetra India

Neurological adverse drug reactions (NADRs) pose a significant clinical challenge as they can have a profound impact on patient health and treatment outcomes. While diverse drug descriptors have been employed for neurological ADR prediction, the potential of using 17 molecular descriptors for this purpose has not been explored. To address this, a multilabel NADR-Net and MLSMOTE-based framework have been proposed for neurological adverse drug reaction prediction. The data for 17 MD and ADRs were extracted from PubChem and ADRECs databases and then mapped based on drug ID. The resulting dataset contained information on 2160 drugs, including their molecular properties, and 1030 ADRs. The methodology was then applied to this dataset, and it showed promising results in terms of hamming loss, precision, true positive rate, f1 score, and ROC-AUC. This study highlights the potential of using molecular descriptors for predicting neurological ADRs, which could improve patient outcomes and drug safety. © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Utilizing Deep Learning for Detecting Malaria in Blood Cell Images

Utilizing Deep Learning for Detecting Malaria in Blood Cell ...

引用

2024 IEEE International Conference on Computing, Applications and Systems, COMPAS 2024

作者： Wahiduzzaman, Md. Fokir, Md. Jahid Rimi, Sadia Afrin Parvin, Tunazzina Hossain, Md Jakir Ahnaf Morshed, Md. Reza, Ahmed Wasif Arefin, Mohammad Shamsul East West University Department of Computer Science and Engineering Dhaka Bangladesh Chittagong University of Engineering and Technology Department of Computer Science and Engineering Chittagong Bangladesh

ISBN: (纸本)9798331529765

Plasmodium malaria, a disease transmitted by mosquitoes, has claimed the lives of more than 7.6 million individuals. The conventional methods for detecting this disease are laborious and pose diagnostic challenges. In this study, the performance of a convolutional neural network (CNN) employing transfer learning, specifically a CNN called ShuffleNet, was investigated for classifying infected and uninfected malaria parasite cells in thick blood smears. The dataset utilized in this research comprises 26,161 images categorized into infected and uninfected cells. In this study, the dataset comprises 26,161 images categorized into infected and uninfected cells, providing a comprehensive basis for analysis. The proposed method achieved remarkable results, with a training accuracy of 99.57% and a testing accuracy of 99.77%, indicating its strong performance across various datasets. These findings underscore the potential of deep learning approaches in enhancing malaria detection from blood cell images, offering promising avenues for further research and application in clinical settings. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Constructing a Secure Charity NFT Auction Platform Using Fisco Bcos Blockchain for Enhancing Transparency and Traceability

引用

IEEE Access 2024年 12卷 36924-36941页

作者： Chen, Chin-Ling Zhan, Wan-Bing Tsaur, Woei-Jiunn Huang, Der-Chen Liu, Ling-Chun Changchun Sci-Tech University Changchun School of Information Engineering Jilin130600 China Chaoyang University of Technology Department of Computer Science and Information Engineering Taichung413310 Taiwan Xiamen University of Technology Xiamen School of Computer and Information Engineering Fujian361024 China National Taipei University Computer Center New Taipei City237303 Taiwan National Taipei University Department of Computer Science and Information Engineering New Taipei City237303 Taiwan National Chung Hsing University Department of Computer Science and Engineering Taichung402202 Taiwan

In the charity sector, fundraising and transparency have long been key issues. Charity NFT (Non-Fungible Token) auctions, an emerging charity fundraising model integrating blockchain and NFT concepts, bring opportunities and challenges. Blockchain provides distributed data integrity and transparency via cryptography-linked data blocks, while NFTs enable unique digital ownership representation. This study designs a charity NFT auction platform on the Fisco Bcos blockchain, using multi-signature algorithms to ensure NFT authenticity, ECDSA (Elliptic Curve Digital Signature Algorithm) signatures for transaction integrity and traceability, NFTs and virtual currencies to reduce costs, and IPFS (InterPlanetary File System) for storage. The implemented system achieves 2104 TPS throughput with 492-millisecond latency, increasing transaction processing with low latency. Overall, the platform aims to address charity issues like opaque fund flows, high costs, and fake initiatives through the strategic application of blockchain and NFT functionality. © 2013 IEEE.

关键词： Smart contract

来源：评论

学校读者我要写书评

暂无评论

TruChain: A Blockchain-based Access Control to Improve the Security of Smart Water Grid Systems 17

TruChain: A Blockchain-based Access Control to Improve the S...

引用

17th International Conference on COMmunication Systems and NETworkS, COMSNETS 2025

作者： Mondal, Susmita Chakraborty, Suchetana Indian Institute of Technology Department of Computer Science and Engineering Jodhpur India

ISBN: (纸本)9798331531195

Water is a significant resource in day-to-day life, and it usually requires technological association for comprehensive management. Smart Water Grids (SWG) typically use cyberphysical systems (CPS) to monitor several parameters associated with water quality and distribution, leveraging the possibility of performing remote operations. For robust architecture, remote operations rely on application-specific sensors, communication networks, edge control, and cloud computation. Needless to say, the centralized control of edge and cloud infrastructure is prone to attacks and failures, such as unauthorized access or single-point-of-failure. We propose TruChain, a Blockchain-based access control mechanism to eliminate the probability of access takeover and aggregate quorum-based validation on sensory data for transparent provenance reliability. We also provide access revocation to strengthen security by restricting access to sensitive data stored in a privately distributed cloud. We believe our solution has empirical decentralization and a verifiable authorization method by maintaining a tamper-evident, auditable trail of interactive events, building trust without third-party intervention. © 2025 IEEE.

关键词： Cloud security

来源：评论

学校读者我要写书评

暂无评论

Effective Data Sharing in an Edge-Cloud Model: Security Challenges and Solutions

引用

computer 2024年第7期57卷 53-65页

作者： Karati, Arijit Das, Sajal K. National Sun Yat-Sen University Department of Computer Science and Engineering Kaohsiung804 Taiwan Missouri University of Science and Technology Department of Computer Science RollaMO65409 United States

The proposed protocol offers privacy-preserving authentication across several cloud platforms, flexible key management for consumer data protection, and effective user revocation. Performance evaluation demonstrates that the proposed framework supports low latency, safe unified remote access, and data privacy in the contemporary edge-enabled environment. © 1970-2012 IEEE.

关键词： Consumer protection

来源：评论

学校读者我要写书评

暂无评论

Indian sign language recognition using TensorFlow and MediaPipe 1st

Indian sign language recognition using TensorFlow and MediaP...

引用

1st International Conference on Applications of AI in 5G and IOT, ICAAI5GI 2024

作者： Kanyal, Hoshiyar Singh Gupta, Shubham Singh, Smriti Kumari, Sarita Department of Computer Science and Engineering ABES Institute of Technology Ghaziabad India

ISBN: (纸本)9781032874036

With this work, a technical method for identifying Indian Sign Language is attempted to be developed. This work attempted to create a system for recognizing the hand movements from Indian Sign Language using machine learning techniques. An amalgamation of two- and three-dimensional illustrations of Indian Sign Language motions has been employed. The dataset was manually assembled. To identify landmarks in the pictures, the Media Pipe framework has been used. The findings show that the approach used in this study is successful in identifying the other Sign Language alphabets and gestures. The signs and gestures recognition for several more Indian regional languages could also be investigated and explored using this method. © 2025 the Author(s).

关键词： Gesture recognition

来源：评论

学校读者我要写书评

暂无评论

Usability Evaluation of Online Food Delivery Applications in Bangladesh: A Comparative Study

Usability Evaluation of Online Food Delivery Applications in...

引用

2024 IEEE International Women in engineering (WIE) Conference on Electrical and computer engineering, WIECON-ECE 2024

作者： Islam Arif, Md. Ariful Kabir, Shahriar Islam, Zohirul Hussain Khan, Md Faruk Islam, Iyolita Islam, Muhammad Nazrul Military Institute of Science and Technology Department of Computer Science and Engineering Dhaka Bangladesh Bangladesh University of Professionals Department of Computer Science and Engineering Dhaka Bangladesh

ISBN: (纸本)9798331535476

The online food order and delivery service is becoming a very popular web-based application in Bangladesh. These days, a growing number of individuals are interested in ordering a variety of foods online. In addition, people are looking for quality food that they can purchase whenever and wherever they want. Online food delivery services are required to be intuitive, user-friendly, and accessible to increase their acceptability in Bangladesh. Therefore, the objective of this study is to explore and assess the existing online food delivery apps in Bangladesh and identify the best application from a usability perspective. To attain these objectives, four online food delivery apps (Food Panda, Hungrynaki, Pathao Food, and efood) were evaluated through Heuristic Evaluation, Cognitive Walkthrough, and System Usability Scale (SUS). The study's findings showed that each app has a number of usability issues requiring to be resolved in order to improve acceptance. The results indicated that from heuristic evaluation, Food Panda had the most number of problems and Hungrynaki had the least number of usability issues;from the cognitive walkthrough perspective, Food Panda and eFood was the mostly usable application;and from SUS evaluation score, Food Panda achieved best score from the users. Overall, Food Panda is the best application from the usability perspective. © 2024 IEEE.

关键词： Food ingredients

来源：评论

学校读者我要写书评

暂无评论

Developing a Model for Predicting Multiple Sclerosis Progression Using MRI Scans

Developing a Model for Predicting Multiple Sclerosis Progres...

引用

2024 IEEE International Conference on Computing, Applications and Systems, COMPAS 2024

作者： Jahan, Rownak Sarker, Anabil Yasmin, Ayesha Siddiqa Mushfika Hasan, Kazi Hossain, Md Jakir Adnan Morshed, Md. Reza, Ahmed Wasif Shamsul Arefin, Mohammad East West University Department of Computer Science and Engineering Dhaka Bangladesh Chittagong University of Engineering and Technology Department of Computer Science and Engineering Chittagong Bangladesh

ISBN: (纸本)9798331529765

A significant number of people lose their lives to multiple sclerosis each year, and many of these victims come from rural areas. However, the technology for early sclerosis diagnosis is not as effective as anticipated. So, we set out to develop a solid method that can aid medical experts in diagnosing sclerosis. Despite a large body of research on the subject, we attempted to create a far more accurate and error-free classification method that is trained on a substantial amount of real datasets as opposed to augmented data. The network performs noticeably well on 15,600 MRI (Magnetic Resonance Imaging) pictures using a modified Inception-v3, Resnet50, and Efficient Net. We found the enhanced InceptionV3 excelled, achieving 92.44% training, and 93.26% validation. Additionally, it outperformed several alternative CNN architectures already in use. © 2024 IEEE.

关键词： Diagnosis

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：