检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Miró-Nicolau, Miquel Moyà-Alcover, Gabriel Jaume-I-Capó, Antoni González-Hidalgo, Manuel Campello, Maria Gemma Sempere Sancho, Juan Antonio Palmer UGiVIA Research Group University of the Balearic Islands Dpt. of Mathematics and Computer Science Palma07122 Spain University of the Balearic Islands Dpt. of Mathematics and Computer Science Palma07122 Spain Hospital Universitari Son Espases Palma07010 Spain SCOPIA Research Group Department of Mathematical Sciences and Computer Science University of the Balearic Islands Palma07122 Spain Palma07010 Spain

The increasing reliance on Deep Learning models, combined with their inherent lack of transparency, has spurred the development of a novel field of study known as eXplainable AI (XAI) methods. These methods seek to enhance the trust of end-users in automated systems by providing insights into the rationale behind their decisions. This paper presents a novel approach for measuring user trust in XAI systems, allowing their refinement. Our proposed metric combines both performance metrics and trust indicators from an objective perspective. To validate this novel methodology, we conducted a case study in a realistic medical scenario: the usage of XAI system for the detection of pneumonia from x-ray images. © 2024, CC BY.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Low-Power Electroencephalographic Data Encoding System for Implantable Brain Stimulation Systems 6

Low-Power Electroencephalographic Data Encoding System for I...

引用

6th South-East Europe Design Automation, computer engineering, computer Networks and Social Media Conference, SEEDA-CECNSM 2021

作者： Fragkou, Aikaterini A. Kakarountas, Athanasios P. Kokkinos, Vasileios Dpt. of Computer Science and Biomedical Informatics University of Thessaly Lamia Greece Dpt. of Neurosurgery Massachusetts General Hospital Harvard Medical School MA United States

ISBN: (纸本)9781665427425

This paper studies the Delta encoding scheme and its effect on power dissipation, for wireless transmission from implantable devices. The study was performed on data from electroencephalographic signals. For the implementation of the proposed system, the design approach followed three phases. The first design phase is related to the initial modelling. The second phase included the development of the hardware description code for the proposed system. The third and last phase is related to the evaluation of the transmitted signal and the measurement of the power dissipation. The results showed that up to 23% power savings are possible for a negligible power penalty due to the delta encoding process. © 2021 IEEE

关键词： Signal encoding

来源：评论

学校读者我要写书评

暂无评论

Cryptanalysis of TRUTH: Fixing a Protocol to Make it 5G-Compliant

Cryptanalysis of TRUTH: Fixing a Protocol to Make it 5G-Comp...

引用

International Symposium on Networks, computers and Communications (ISNCC)

作者： J. Munilla M. Burmester P. Caballero-Gil Dpt. of Telecommunication Engineering University of Málaga Málaga Spain Department of Computer Science Florida State Unviersity Tallahassee Florida USA Dept. of Computer Science and AI University of La Laguna La Laguna Tenerife Spain

ISBN: (数字)9798350364910

ISBN: (纸本)9798350364927

Trust and security are critical deployment require-ments for Industrial Internet of Things (IIoT) networks. A recent protocol, called TRUTH, integrates security mechanisms for authentication and privacy alongside a Dempster-Shafer based trust model, to assess the trustworthiness of collected data. We analyze this protocol and show that it does not comply with the implementation requirements of the 5G standard, and that flaws in the identification of involved parties hinder its adoption by 5G devices. We show how to fix this protocol and tailor it for the specific 5G implementation constraints, by enhancing its computational and communication performance. Finally, we show how to prove the security of this fixed protocol (TRUTH +) using the Tamarin automated verification tool.

关键词： Fault diagnosis Performance evaluation Data privacy Protocols 5G mobile communication Data models Synchronization Object recognition Standards Industrial Internet of Things

来源：评论

学校读者我要写书评

暂无评论

Analyzing the Cause of Students Reluctant to Participate in the Classroom: Machine Learning Approach 10

Analyzing the Cause of Students Reluctant to Participate in ...

引用

10th International Conference on System Modeling and Advancement in Research Trends, SMART 2021

作者： Islam, Majharul Islam, Md. Sanzidul Siddik, Md. Abu Bakkar Mousumi, Sadia Sultana Sharmin Daffodil International University Dpt. of Software Engineering Dhaka Bangladesh Daffodil International University Dept. Computer Science and Engineering Dhaka Bangladesh

ISBN: (纸本)9781665439701

in the twenty-first century, seems many students are reluctant to participate in classroom matters day by day. Some days ago, a course teacher is providing a lecture, on the most important topics that will be implemented in every step in life that class majority students were nonexistent. We got motivation why it's happened and find out a lot of reasons behind uninterested in class after interrogation. Who are belonging age range 20-25, most of the time they face an obstacle and suffer and other teenagers. The fundamental purpose of this research to take action based on behind reasons, so that students can move in classroom learning matters. In this article, solve will be described and how to motivate students. Then it will be much helpful and far informative for teacher, students, guardians and also education specialist [8]. In this research shown causes, and also recommended how to overcome and solve those reluctant and spread knowledge and enlighten the nation. The aim of the research to help and inform provide to students avoid his/her predicted absent in classroom using Machine learning, particularly we used Liner Algorithm. Then, students will participate in classroom lectures as well as social activity and world would be more peaceful. This research achieved more than the accuracy of 95.9%. © 2021 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Analysis and Detection of Melanoma through Collective Intelligence with AI

Analysis and Detection of Melanoma through Collective Intell...

引用

Annual IEEE Symposium on computer-Based Medical Systems

作者： Enrique Fernández-Morales Carlos L. Sanchez-Bocanegra Rafael Pastor-Vargas José-Juan Pereyra-Rodriguez Juan M. Haut José Alberto Benítez-Andrades Communications and Control Systems Dpt. Computer Science Faculty UNED Madrid Spain Communications and Control Systems Dpt. Computer Science Faculty UNED Malaga Spain Virgen Del Rocio Hospital Universitario Seville Spain School of Tecnology University of Extremadura Cáceres Spain Dept. of Electric Systems and Automatics Engineering Universidad de León León Spain

ISBN: (数字)9798350384727

ISBN: (纸本)9798350384734

Cancer is a widespread global health problem, claiming millions of lives each year, and skin cancer represents a significant threat as it is one of the most common types. Early tumor detection via medical imaging is critical for effective treatment. Leveraging artificial intelligence, particularly novel models like Transformers, presents promising avenues for improved diagnosis. This paper explores the efficacy of a Collective Intelligence approach using AI in classifying cancerous and non-cancerous tumors, aiming to reduce classification errors and support clinical decision-making. We created five different configurations using various datasets to compare the results. The results show solid performance for the CI in the evaluated tasks, reaching up to 75.89% accuracy. The lack of images in certain classes significantly contributes to overfitting. It is suggested to explore data expansion strategies and improve consistency in image capture for future work.

关键词： Accuracy Melanoma Collective intelligence Image capture Transformers Solids Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

PIONEERING WOMEN IN COMPUTING: A CULTURAL WEBSITE BASED ON INFORMATION ARCHITECTURE 20

PIONEERING WOMEN IN COMPUTING: A CULTURAL WEBSITE BASED ON I...

引用

20th International Conference Applied Computing, AC 2023 and the 22nd International Conference WWW/Internet, ICWI 2023

作者： García-Cabrera, Lina Serrano, José-María Pedrosa-García, Irene Dpt. Computer Science University of Jaén Edificio A3 - Campus Las Lagunillas Jaén23071 Spain El Valle High School Junta de Andalucía Carr. de Madrid 2 Jaén23009 Spain

In an effort to acknowledge the eminent role of a number of relevant women, pioneers in the field of computer science, an exposition was developed as part of a cultural project hosted by the University of Jaen (Spain). In order to establish a permanent, virtual, version of this exhibition, a website has been created, as a way to allow visitors to know more about these pioneers, about their life and history, and their many contributions to the field. In addition, the site serves as a record of all events related to the exhibition, as well as multimedia galleries in order to improve the experience. In this work, we provide an analysis about the development of this site, which features information architecture guidelines, as well as accessibility and usability issues. © ICWI *** rights reserved.

关键词： Websites

来源：评论

学校读者我要写书评

暂无评论

Chewing Detection from Commercial Smart-glasses

arXiv

引用

arXiv 2022年

作者： Papapanagiotou, Vasileios Liapi, Anastasia Delopoulos, Anastasios Multimedia Understanding Group Electrical and Computer Engineering Dpt. Aristotle University of Thessaloniki Thessaloniki Greece Electrical and Computer Engineering Dpt. Aristotle University of Thessaloniki Thessaloniki Greece

Automatic dietary monitoring has progressed significantly during the last years, offering a variety of solutions, both in terms of sensors and algorithms as well as in terms of what aspect or parameters of eating behavior are measured and monitored. Automatic detection of eating based on chewing sounds has been studied extensively, however, it requires a microphone to be mounted on the subject's head for capturing the relevant sounds. In this work, we evaluate the feasibility of using an off-the-shelf commercial device, the Razer Anzu smart-glasses, for automatic chewing detection. The smart-glasses are equipped with stereo speakers and microphones that communicate with smart-phones via Bluetooth. The microphone placement is not optimal for capturing chewing sounds, however, we find that it does not significantly affect the detection effectiveness. We apply an algorithm from literature with some adjustments on a challenging dataset that we have collected in house. Leave-one-subject-out experiments yield promising results, with an F1-score of 0.96 for the best case of duration-based evaluation of eating time. Copyright © 2022, The Authors. All rights reserved.

关键词： Microphones

来源：评论

学校读者我要写书评

暂无评论

Experimenting with Normalization Layers in Federated Learning on non-IID scenarios

arXiv

引用

arXiv 2023年

作者： Casella, Bruno Esposito, Roberto Sciarappa, Antonio Cavazzoni, Carlo Aldinucci, Marco Computer Science Dpt. University of Turin Corso Svizzera 185 Turin Italy Leonardo S.p.A. Italy

Training Deep Learning (DL) models require large, high-quality datasets, often assembled with data from different institutions. Federated Learning (FL) has been emerging as a method for privacy-preserving pooling of datasets employing collaborative training from different institutions by iteratively globally aggregating locally trained models. One critical performance challenge of FL is operating on datasets not independently and identically distributed (non-IID) among the federation participants. Even though this fragility cannot be eliminated, it can be debunked by a suitable optimization of two hyperparameters: layer normalization methods and collaboration frequency selection. In this work, we benchmark five different normalization layers for training Neural Networks (NNs), two families of non-IID data skew, and two datasets. Results show that Batch Normalization, widely employed for centralized DL, is not the best choice for FL, whereas Group and Layer Normalization consistently outperform Batch Normalization. Similarly, frequent model aggregation decreases convergence speed and mode *** Codes 68T07 © 2023, CC BY.

关键词： Large dataset

来源：评论

学校读者我要写书评

暂无评论

PaperMill Detection in Scientific Content

PaperMill Detection in Scientific Content

引用

International Workshop on Semantic Media Adaptation and Personalization (SMAP)

作者： Gerasimos Razis Konstantinos Anagnostopoulos Omiros Metaxas Stefanos-Dimitrios Stefanidis Hong Zhou Ioannis Anagnostopoulos Atypon Systems LLC California United States Computer Science and Biomedical Informatics Dpt. University of Thessaly Lamia Greece

Researchers often come under pressure when facing the ever-increasing demand to produce a progressive number of publications, resorting to hiring the services of paper mills. These are unofficial, and often illegitimate, organizations providing ready-made questionable research components and services, posing a threat to the research integrity, scientific ecosystem, and publishers. Identifying paper mill material is a challenging and laborious process, while the increasing number of Artificial Intelligence services generating human-like text obstructs this process. The purpose of this paper is to contribute to the research integrity domain by proposing the PaperMill Detection manuscript screening framework. By leveraging contextual signals, it measures the probability of a document being the result of a paper mill organization or generated by Artificial Intelligence. The combination of these signals can facilitate the detection of questionable scientific content. Our evaluation has revealed that the proposed approach outperforms other open-source and commercial solutions in all examined evaluation metrics, achieving an F1 score of 0.97.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Co-Training for Unsupervised Domain Adaptation of Semantic Segmentation Models

arXiv

引用

arXiv 2022年

作者： Gómez, Jose L. Villalonga, Gabriel López, Antonio M. Bellaterra Barcelona08193 Spain Dpt. of Computer Science UAB Spain

Semantic image segmentation is a central and challenging task in autonomous driving, addressed by training deep models. Since this training draws to a curse of human-based image labeling, using synthetic images with automatically generated labels together with unlabeled real-world images is a promising alternative. This implies to address an unsupervised domain adaptation (UDA) problem. In this paper, we propose a new co-training procedure for synth-to-real UDA of semantic segmentation models. It consists of a self-training stage, which provides two domain-adapted models, and a model collaboration loop for the mutual improvement of these two models. These models are then used to provide the final semantic segmentation labels (pseudo-labels) for the real-world images. The overall procedure treats the deep models as black boxes and drives their collaboration at the level of pseudo-labeled target images, i.e., neither modifying loss functions is required, nor explicit feature alignment. We test our proposal on standard synthetic and real-world datasets for on-board semantic segmentation. Our procedure shows improvements ranging from ∼13 to ∼31 mIoU points over baselines. © 2022, CC BY.

关键词： Semantic Segmentation

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：