检索结果-内蒙古大学图书馆

2nd International Conference on computer, Communication and Control, IC4 2024

作者： Abins, A Arockia Darwin Wilfred, S.M. Ashwin Ram, K. Hari Prasad, D. Easwari Engineering College Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350387933

The large amounts of debris present in our ocean water poses a major threat to the flora and fauna of our marine ecosystem. The continuous increase in dumping of solid waste, and the very slow rate of degradation of most items, are together leading to a gradual increase in marine litter found in the ocean. While a large number of human-based campaigns have been proposed to tackle this issue, the sheer amount of litter present in the ocean results in their efforts to be insufficient. But the increasing development in the field of computer vision combined with the help of Autonomous Underwater Vehicle (AUV), presents itself to be a better solution to reduce the human complexities and automate the tedious task of decontaminating our ocean. This paper contains a literature review of the different Machine learning models using various neural architectures currently used to detect and locate various debris present in marine water (i.e., plastic, metals, glass, styrofoam, cigarette butts etc.). © 2024 IEEE.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Constrained Bandit Learning with Switching Costs for Wireless Networks 42

Constrained Bandit Learning with Switching Costs for Wireles...

引用

42nd IEEE International Conference on computer Communications, INFOCOM 2023

作者： Steiger, Juaren Li, Bin Ji, Bo Lu, Ning Queen's University Department of Electrical and Computer Engineering Kingston Canada Pennsylvania State University State College School of Electrical Engineering and Computer Science PA United States Virginia Tech Department of Computer Science BlacksburgVA United States

ISBN: (纸本)9798350334142

Bandits with arm selection constraints and bandits with switching costs have both gained recent attention in wireless networking research. Pessimistic-optimistic algorithms, which combine bandit learning with virtual queues to track the constraints, are commonly employed in the former. Block-based algorithms, where switching is disallowed within a block, are commonly employed in the latter. While efficient algorithms have been developed for both problems, it remains challenging to guarantee low regret and constraint violation in a bandit problem that includes both arm selection constraints and switching costs due to the tight coupling between the two. Here, switching may be necessary to decrease the constraint violation but comes at the cost of increased switching regret. In this paper, we tackle the constrained bandits with switching costs problem, for which we design a block-based pessimistic-optimistic algorithm. We identify three timely wireless networking applications for this framework in edge computing, mobile crowdsensing, and wireless network selection. We also prove that our algorithm achieves sublinear regret and vanishing constraint violation and corroborate these results with synthetic simulations and extensive trace-based simulations in the wireless network selection setting. © 2023 IEEE.

关键词： Wireless networks

来源：评论

学校读者我要写书评

暂无评论

Thai Winograd Schemas: A Benchmark for Thai Commonsense Reasoning 2

Thai Winograd Schemas: A Benchmark for Thai Commonsense Reas...

引用

2nd Workshop in South East Asian Language Processing, SEALP 2025

作者： Artkaew, Phakphum Department of Electrical and Computer Engineering Tandon School of Engineering New York University United States

ISBN: (纸本)9798891762190

Commonsense reasoning is one of the important aspects of natural language understanding, with several benchmarks developed to evaluate it. However, only a few of these benchmarks are available in languages other than English. Developing parallel benchmarks facilitates cross-lingual evaluation, enabling a better understanding of different languages. This research introduces a collection of Winograd Schemas in Thai, a novel dataset designed to evaluate commonsense reasoning capabilities in the context of the Thai language. Through a methodology involving native speakers, professional translators, and thorough validation, the schemas aim to closely reflect Thai language nuances, idioms, and cultural references while maintaining ambiguity and commonsense challenges. We evaluate the performance of popular large language models on this benchmark, revealing their strengths, limitations, and providing insights into the current state-of-the-art. Results indicate that while models like GPT-4 and Claude-3-Opus achieve high accuracy in English, their performance significantly drops in Thai, highlighting the need for further advancements in multilingual commonsense reasoning. © 2025 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Multi-Modal Visual Features Perception Technology for Internet of Vehicles (IoV) 6

Multi-Modal Visual Features Perception Technology for Intern...

引用

6th IEEE International Conference on Emerging Smart Computing and Informatics, ESCI 2024

作者： Mahendran, Rakesh Kumar Thiyagarajan, Abirami Angelin Gracia, A. Kumar, P. Rajalakshmi Engineering College Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350306613

In the realm of the Internet of Vehicles (IoV), this study addresses a fundamental question that causes the emerging trends within the IoV landscape, Previous research in IoV has laid the groundwork but is limited by conventional single-modal methods. This study advances our understanding by incorporating innovative Multi-Modal Visual Features Perception Technology by inputting various images from LiDAR cameras, Radar and processing using ConvolutionalNeural Networks (CNN), alongside single-modal approaches. Through this research, we uncover hidden facets of IoV, promising transformative insights. The implementation of Multi-Modal Visual Features Perception Technology using deep learning, along with single-modal methods, holds the potential to revolutionize vehicular connectivity and shape a more connected and secure future. © 2024 IEEE.

关键词： Advanced neural network techniques Convolutional Neural Networks LiDAR Multimodal Scene perception Self-driving technology sensor fusion

来源：评论

学校读者我要写书评

暂无评论

Fingerprint inspired advanced end-effector and its applications

引用

Friction 2022年第10期10卷 1739-1750页

作者： Sung Ho LEE Hyo Sung KIM Han Jun PARK Moon Kyu KWAK Department of Electrical Engineering and Computer Science University of MichiganAnn Arbor 48109USA Department of Mechanical Engineering Kyungpook National UniversityDaegu 41566Republic of Korea

The human fingertip consists of a fingerprint with many micro-grating *** main roles of the fingerprint could be divided into two purposes,namely,the enhancement of the frictional force and the effective transmittance of the *** this study,we present the fingerprint-inspired end-effector that has not only admirable frictional force but also electric *** end-effector is composed of fluorocarbon rubber,one of the famous materials to achieve high frictional force and *** various experiments,the novel performance of micro structured fluorocarbon rubber end-effector(MSFE)is characterized by comparing with a macroscale patterned sample(MPS),which has been already used in real industrial *** results are analyzed ***,as feasible applications,we suggest two applications based on the role of the *** is the conductive astronaut glove with high frictional force,and the other one is a non-slip pad for the next-generation glass transfer *** these experiments,we successfully observe the enhanced system performance and confirm the possibility of using the MSFE as feasible *** believe that the MSFE could be a useful and powerful alternative as an end-effector,not only in the aerospace industry but also in display manufacturing processes.

关键词： fingerprint end-effector fluorocarbon rubber(FKM) glass transfer system spacesuit

来源：评论

学校读者我要写书评

暂无评论

A Federated Mining Framework for Complete Erasable Itemsets

A Federated Mining Framework for Complete Erasable Itemsets

引用

2024 IEEE International Conference on Big Data, BigData 2024

作者： Hong, Tzung-Pei Kuo, Meng-Jui Chen, Chun-Hao Li, Katherine Shu-Min National University of Kaohsiung Department of Computer Science and Information Engineering Kaohsiung Taiwan National Sun Yat-sen University Department of Computer Science and Engineering Kaohsiung Taiwan National Kaohsiung University of Science and Technology Department of Computer Science and Information Engineering Kaohsiung Taiwan

ISBN: (纸本)9798350362480

In this study, we develop an innovative federated framework for erasable itemset mining to address the challenges of horizontal federated learning in data mining and resolve the shortcomings of the previous algorithm. The framework is based on a client-server architecture, in which clients from multiple data sources collaborate to effectively integrate information. The proposed algorithm is divided into two parts, including the client-side mining and the server-side aggregation. During the client-side mining stage, the algorithm introduces quasi-erasable itemsets to collect additional useful information, facilitating the integration of results on the server side. In the server-side aggregation stage, the algorithm employs a boundary strategy, effectively utilizing the clients' quasi-erasable and erasable itemsets to improve the accuracy of the results. Experimental results demonstrate that the proposed method not only effectively mines complete knowledge but also ensures data-privacy protection. © 2024 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Machine Learning Powered Genre Prediction for Next Level Book Recommendations 5

Machine Learning Powered Genre Prediction for Next Level Boo...

引用

5th IEEE International Conference on Computing, Power, and Communication Technologies, IC2PCT 2024

作者： Julian, Anitha Pavithra, B. Saveetha Engineering College Department of Computer Science and Engineering Chennai India

ISBN: (纸本)9798350383522

The proposed work delves into how recommender systems, like those on YouTube and Amazon, shape our online experiences, particularly in book recommendations. It addresses the challenge of the 'cold start problem' with the help of a Popularity-based recommender initially, adapting to personalized suggestions through Collaborative Filtering as users engage. This dual approach aims for a comprehensive system that overcomes challenges in new user on boarding and adapts to evolving preferences. The importance of continuous evaluation is emphasized for ongoing effectiveness. In a separate study, the research explores the application of Natural Language Processing (NLP) in predicting book genres. It uses algorithms to analyze textual data, employing techniques like Exploratory Data Analysis (EDA) to uncover patterns. The modeling phase involves text classification algorithms, contributing to a deeper understanding of genre prediction through the application of NLP techniques. © 2024 IEEE.

关键词： collaborative filtering continuous evaluation Exploratory Data Analysis (EDA) genre prediction Naive Bayes Natural Language Processing (NLP) Popularity-based recommendation Recommendation system text classification algorithms

来源：评论

学校读者我要写书评

暂无评论

Flood Risk Assessment and Interpretation Using Explainable AI 2

Flood Risk Assessment and Interpretation Using Explainable A...

引用

2nd IEEE International Conference on IoT, Communication and Automation Technology, ICICAT 2024

作者： Sharma, Swati Tomar, Vivek Arora, Sangeeta Gupta, Nishant KIET Group of Institutions Delhi - NCR Computer Science & Engineering Department Ghaziabad India University Computer Science & Engineering Department Dehradun India KIET Group of Institutions Delhi-NCR Computer Applications Department Ghaziabad India SET Department of Computer Science & Engineering Sharda University Greater Noida India

ISBN: (纸本)9798350368109

Floods are the unforeseen situations to the society which causes high risk to the society. For proactive action, it is important to identify locations based on past results. It is necessary to take measures before the situation gets worst. For the benefit of society, a novel approach to map flood risk usinga balanced dataset and an Extra Tree Classifier. The Extra Tree Classifier performs better with the Shapely Additive explanation values. It shows improvement towards decision making processes to reduce the causality at the time of floods. The proposed method achieves the accuracy of 93.61% for the estimating flood danger effects on a particular location. In this paper quantitative study was done to handle the challenges to predict the flood danger using the artificial intelligence ethically. © 2024 IEEE.

关键词： Risk assessment

来源：评论

学校读者我要写书评

暂无评论

Load Balancing in SDN-IoT Network 8th

Load Balancing in SDN-IoT Network

引用

8th International Conference on Inventive Communication and Computational Technologies, ICICCT 2024

作者： Sarohe, Sanehi Harit, Sandeep Kumar, Manish Department of Computer Science and Engineering Punjab Engineering College Chandigarh India

ISBN: (纸本)9789819777099

The Internet of Things (IoT) extends the reach of the Internet by allowing intelligent devices to discover and connect with the actual surroundings, as well as interact with other physical entities. Software Defined Networking (SDN), which separates the data plane and control plane, connects large networks and gives a view of the whole network. Combining SDN with IoT offers the advantage of centralized control over all connected IoT devices from a single point. This paper reviews existing research on load balancing in SDN-IoT networks and thereafter discusses applicable techniques. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Out-of-Distribution Detection for Contrastive Models Using Angular Distance Measures 23

Out-of-Distribution Detection for Contrastive Models Using A...

引用

23rd IEEE International Conference on Machine Learning and Applications, ICMLA 2024

作者： Hossain, M Shifat Jha, Sumit Kumar Walker, Chase Ewetz, Rickard University of Central Florida Department of Electrical and Computer Engineering OrlandoFL32816 United States Knights Foundation School Florida International University Computer Science Department MiamiFL33199 United States University of Florida Department of Electrical and Computer Engineering GainsvilleFL32611 United States

ISBN: (纸本)9798350374889

Vision-language models have demonstrated extraordinary zero-shot image classification capabilities. Out-of-distribution (OOD) detection is the problem of determining if a model is operating within its knowledge limits. While distance-based detection algorithms have emerged as a promising approach to OOD detection, we observe that there is a disparity between the distance measures used for OOD detection and model training. Recent studies have attempted to mitigate this shortcoming by modifying the contrastive learning process, which is highly undesirable for foundation models. In this paper, we propose an Angular distance-based out-of-distribution detection method for Contrastive models (AEC), an OOD detection framework for foundational contrastive models based on an angular distance measure. The angular distance-based score is compliant with the standard training process and circumvents the need to modify the training process of the model. We also formulate a distance transformation and solve an optimization problem to determine an OOD score threshold value for in-distribution and out-of-distribution data. The experimental evaluation demonstrates that AEC outperforms state-of-the-art OOD detection models in terms of AUROC, FPR@95TPR, accuracy, and correct ID metrics. We obtained an overall AUROC and FPR@95TPR of 70.42 and 83.98 from the proposed algorithm, which is significantly better compared to the SOTA OOD detection algorithms. © 2024 IEEE.

关键词： Contrastive Learning

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：