检索结果-内蒙古大学图书馆

2022 International Conference on Electronics and Renewable Systems, ICEARS 2022

作者： Ramprasath, M. Dhanasekaran, K. Karthick, T. Velumani, R. Sudhakaran, P. Srm Institute of Science and Technology Department of Data Science and Business Systems School of Computing TamilNadu Chennai India Department of Computer Science and Engineering Andhra Pradesh Madhrawada India Srm Trp Enginerring College Department of Computer Science and Engineering Tamil Nadu Trichy India

ISBN: (纸本)9781665484251

In recent years, Pretraining Language Models based on Transformers -Natural Language Processing (PLMT-NLP) have been highly successful in nearly every NLP task. In the beginning, Generative Pre-trained model-based Transformer, BERT- Bidirectional Encoder model Representations using Transformers was used to develop these models. Models constructed on transformers, Self-supervise knowledge acquiring, and transfer learning establish the foundation of these designs. Transformed-based pre-trained models acquire common linguistic illustrations from vast amounts of textual information through self-supervised model and apply this information to downstream tasks. To eliminate the need to retrain downstream models, these models provide a solid foundation of knowledge. In this paper, the enhanced learning on PLMT-NLP has been discussed. Initially, a quick introduction to self-supervised learning is presented, then diverse core concepts used in PLMT-NLP are explained. Furthermore, a list of relevant libraries for working with PLMT-NLP has been provided. Lastly, the paper discusses about the upcoming research directions that will further improve these models. Because of its thoroughness and relevance to current PLMT-NLP developments, this survey study will positively serve as a valuable resource for those seeking to understand both basic ideas and new developments better. © 2022 IEEE.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Towards Robust Temporal Activity Localization Learning with Noisy Labels 30

Towards Robust Temporal Activity Localization Learning with ...

引用

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Liu, Daizong Qu, Xiaoye Fang, Xiang Dong, Jianfeng Zhou, Pan Nan, Guoshun Tang, Keke Fang, Wanlong Cheng, Yu Wangxuan Institute of Computer Technology Peking University China School of Cyber Science and Engineering Huazhong University of Science of Technology China Interdisciplinary Graduate Programme Nanyang Technological University Singapore College of Computer Science and Technology Zhejiang Gongshang University China School of Cyberspace Security Beijing University of Posts and Telecommunications China Cyberspace Institute of Advanced Technology Guangzhou University China Department of Computer Science and Engineering The Chinese University of Hong Kong Hong Kong

ISBN: (纸本)9782493814104

This paper addresses the task of temporal activity localization (TAL). Although recent works have made significant progress in TAL research, almost all of them implicitly assume that the dense frame-level correspondences in each video-query pair are correctly annotated. However, in reality, such an assumption is extremely expensive and even impossible to satisfy due to subjective labeling. To alleviate this issue, in this paper, we explore a new TAL setting termed Noisy Temporal activity localization (NTAL), where a TAL model should be robust to the mixed training data with noisy moment boundaries. Inspired by the memorization effect of neural networks, we propose a novel method called Co-Teaching Regularizer (CTR) for NTAL. Specifically, we first learn a Gaussian Mixture Model to divide the mixed training data into preliminary clean and noisy subsets. Subsequently, we refine the labels of the two subsets by an adaptive prediction function so that their true positive and false positive samples could be identified. To avoid single model being prone to its mistakes learned by the mixed data, we adopt a co-teaching paradigm, which utilizes two models sharing the same framework to teach each other for robust learning. A curriculum strategy is further introduced to gradually learn the moment confidence from easy to hard. Experiments on three datasets demonstrate that our CTR is significantly more robust to the noisy training data compared to the existing methods. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Gaussian distribution

来源：评论

学校读者我要写书评

暂无评论

eFuseNet: A deep ensemble fusion network for efficient detection of Arrhythmia and Myocardial Infarction using ECG signals

引用

Multimedia Tools and Applications 2024年第18期84卷 1-32页

作者： Dwivedi, Amitesh Kumar Srivastava, Gaurav Tripathi, Sakshi Pradhan, Nitesh Johns Hopkins Whiting School of Engineering Johns Hopkins University BaltimoreMD21218 United States Dell International Services India Pvt. Ltd. Dell Technologies Telangana Hyderabad500081 India Department of Computer Science and Engineering Manipal University Jaipur Rajasthan Jaipur303007 India Department of Computer Science and Engineering The LNM Institute of Information Technology Jaipur Rajasthan Jaipur302031 India

Myocardial infarction (MI) and cardiac arrhythmias remain leading causes of mortality within cardiovascular diseases. Accurate and timely diagnosis using electrocardiogram (ECG) signals is crucial, yet manual ECG interpretation is prone to error and time-consuming. Addressing these challenges, the authors propose a novel deep neural network architecture for efficient and accurate detection and classification of both MI and arrhythmia from ECG signals. Our model strategically fuses a modified AlexNet for robust feature extraction with Long Short-Term Memory and Gated Recurrent Units networks to exploit the temporal dependencies inherent in ECG data. Fusion techniques combine the insights from these diverse sub-models, aiming to reduce generalization errors and surpass the performance of single-model approaches. Extensive experimentation on the MIT-BIH and PTB datasets demonstrates superior accuracy (98.51% and 99.97%, respectively) and reduced training time compared to other pre-trained networks. Robustness is validated through 10-fold cross-validation, yielding average accuracies of 98.76% and 99.48%. Our findings highlight the potential of this computationally efficient ensemble model for practical clinical implementation, enabling more accurate and timely diagnosis of life-threatening cardiovascular conditions. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Electrocardiograms

来源：评论

学校读者我要写书评

暂无评论

Augmented reality for studying hands on the human body for elementary school students

引用

Procedia computer science 2023年 216卷 237-244页

作者： Richi Rusli Darryl Arkan Nalanda Antonius Doff Valma Tarmidi Kristien Margi Suryaningrum Rezki Yunanda Computer Science Department School of Computer Science Bina Nusantara University Jakarta 11480 Indonesia

Augmented Reality (AR) media has become a utilization or an intermediary tool in everyday life. Augmented Reality has become a medium to contribute to human life, such as games, virtual tourism objects, and learning media. However, the application of Augmented Reality media in the world of education as a learning medium is still rarely used. Therefore, the holding of this research is as a form of contribution to the development of Augmented Reality media as a learning medium. The form of contribution from this research is a simple Augmented Reality created by using qualitative method to collect and analyze the elementary students data and Marker Based Tracking method resulting an application that has a function as a hand-shaped prop teaching aid for learning the body parts of the hands in elementary school students. With this learning media, it is hoped that it would help teachers and elementary school students in studying the parts of human body. All elementary school teachers can use the Aplikasi AR Belajar Bagian Tangan as a teaching medium in elementary school laboratories on their elementary school students or as a prop while students listening to their teacher explanations while seeing which parts of the hand are pointed out by elementary school teachers through the Aplikasi AR Belajar Bagian Tangan. This Aplikasi AR Belajar Bagian Tangan may provide elementary school students with their study and experience in using AR as a learning medium that can improve students’ analytical skills.

关键词： education application research Augmented Reality learning students props

来源：评论

学校读者我要写书评

暂无评论

Differentially private locality sensitive hashing based federated recommender system

Differentially private locality sensitive hashing based fede...

引用

作者： Hu, Hongsheng Dobbie, Gillian Salcic, Zoran Liu, Meng Zhang, Jianbing Lyu, Lingjuan Zhang, Xuyun Department of Electrical Computer and Software Engineering The University of Auckland Auckland New Zealand School of Computer Science The University of Auckland Auckland New Zealand Department of Electrical and Computer Engineering The University of Auckland Auckland New Zealand School of Mechanical Electrical and Information Engineering Shandong University Weihai Shandong China Department of Computer Science and Technology Nanjing University Nanjing China Department of Computer Science National University of Singapore Singapore Department of Computing Macquarie University Sydney Australia

Recommender systems are important applications in big data analytics because accurate recommendation items or high-valued suggestions can bring high profit to both commercial companies and customers. To make precise recommendations, a recommender system often needs large and fine-grained data for training. In the current big data era, data often exists in the form of isolated islands, and it is difficult to integrate the data scattered due to privacy security concerns. Moreover, privacy laws and regulations make it harder to share data. Therefore, designing a privacy-preserving recommender system is of paramount importance. Existing privacy-preserving recommender system models mainly adapt cryptography approaches to achieve privacy preservation. However, cryptography approaches have heavy overhead when performing encryption and decryption operations and they lack a good level of flexibility. In this paper, we conduct privacy analysis on the existing locality sensitive hashing (LSH) approach based privacy-preserving recommender system and show how an attacker can retrieve user's information under such a recommender system. Given such privacy risks, we propose differentially private LSH approach to build recommender system that can offer differential privacy guarantees for users. Our proposed efficient and scalable federated recommender system can make full use of multiple source data from different data owners while guaranteeing privacy preservation of users' data in contributing parties. Extensive experiments on real-world benchmark datasets show that our approach can achieve both high time efficiency and accuracy under small privacy budgets. © 2021 John Wiley & Sons, Ltd.

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

Characterizing the Weed Species of Agricultural Landscapes Using Hyper Spectral Image Analysis 5th

Characterizing the Weed Species of Agricultural Landscapes U...

引用

5th International Conference on data science, Machine Learning and Applications, ICDSMLA 2023

作者： Nagaraja, S.R. Mishra, Awakash Prabhu, A. Sivaraman, Himani School of Computer Science and Engineering Presidency University Karnataka Bengaluru India Maharishi School of Engineering and Technology Maharishi University of Information Technology Uttar Pradesh Lucknow India Karnataka Bangalore India Department of Computer Science and Engineering School of Engineering and Computing Dev Bhoomi Uttarakhand University Manduwala Dehradun Navgaon India

ISBN: (纸本)9789819780426

This paper explores using hyper spectral photo evaluation to discover and characterize weed species inside agricultural landscapes. By using excessive-decision hyperspectral imaging, the specific spectral fingerprints of the different weed species attempting to penetrate the crop canopy can be diagnosed. This fact can then be compared to spectral libraries for the precise identity of the weed species gift. With these spectral facts, researchers can map the spread of weeds, understand the composition of weed species present, and verify weed responses to exceptional farming strategies or weather changes. This information can then assist in the development of extra effective and selective weed management strategies in agriculture. In conclusion, hyperspectral picture analysis may provide a treasured tool for characterizing and dealing with weed species within agricultural landscapes. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： characterizing hyperspectral weather

来源：评论

学校读者我要写书评

暂无评论

Autonomous Vehicle Navigation Systems: Machine Learning for Real-Time Traffic Prediction

Autonomous Vehicle Navigation Systems: Machine Learning for ...

引用

2025 IEEE International Conference on Computational, Communication and Information Technology, ICCCIT 2025

作者： Praveen, R.V.S. Hundekari, Sheela Parida, Prasanta Mittal, Tanusha Sehgal, Archana Bhavana, Munugapati LTIMindtree Limited Telangana Hyderabad India School of Computer Applications Pimpri Chinchwad University Pune India School of Rural Management KIIT University Odisha Bhubaneswar India Graphic Era Deemed to Be University Department of Computer Science & Engineering Dehradun India Lovely Professional University Phagwara India MLR Institute of Technology Department of Computer Science and Engineering Telangana Hyderabad India

ISBN: (纸本)9798331512965

By making transportation safer and more efficient, autonomous vehicle navigation systems are transforming the transportation industry. Autonomous cars are able to make route selections based on accurate, up-to-theminute traffic predictions, which help them avoid congestion and maximize their trip time. The use of machine learning algorithms to reliably forecast traffic conditions in real time for use in autonomous vehicle navigation is the focus of this research. Deep learning, support vector machines, and random forests are just a few examples of the machine learning models that can accurately forecast traffic conditions by mining massive datasets collected from GPS data, traffic sensors, and past trends. The suggested method incorporates these models into an autonomous vehicle navigation system that operates in real-time, enabling adaptive modifications according to the flow of traffic. We also look at how things like construction, road accidents, and weather affect the precision of traffic predictions. Based on the outcome of decreased trip time, increased safety and efficiency, results demonstrate how navigation systems supplemented with machine learning, far surpass conventional routing approaches. The study suggests that further research needs to be conducted on hybrid algorithms for better traffic prediction, and multi modal data integration to further optimize autonomous vehicle navigation and in general, machine learning presents great promise in this area. © 2025 IEEE.

关键词： Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

New Algorithms for Distributed Fair k-Center Clustering: Almost Accurate as Sequential Algorithms 23

New Algorithms for Distributed Fair k-Center Clustering: Alm...

引用

23rd International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2024

作者： Wu, Xiaoliang Feng, Qilong Huang, Ziyun Xu, Jinhui Wang, Jianxin School of Computer Science and Engineering Central South University Xiangjiang Laboratory Changsha China Department of Computer Science and Software Engineering Penn State Erie The Behrend College Erie United States Department of Computer Science and Engineering State University of New York at Buffalo Buffalo United States Hunan Provincial Key Lab on Bioinformatics Central South University Xiangjiang Laboratory Changsha China

Fair clustering problems have been paid lots of attention recently. In this paper, we study the k-Center problem under the group fairness and data summarization fairness constraints, denoted as Group Fair k-Center (GFkC) and data Summarization Fair k-Center (DSFkC), respectively, in the massively parallel computational (MPC) distributed model. The previous best results for the above two problems in the MPC model are a 9-approximation with violation 7 (WWW 2022) and a (17 + ∊)-approximation without fairness violation (ICML 2020), respectively. In this paper, we obtain a (3 + ∊)- approximation with violation 1 for the GFkC problem in the MPC model, which is almost as accurate as the best known approximation ratio 3 with violation 1 for the sequential algorithm of the GFkC problem. Moreover, for the DSFkC problem in the MPC model, we obtain a (4 + ∊)-approximation without fairness violation, which is very close to the best known approximation ratio 3 for the sequential algorithm of the DSFkC problem. Empirical experiments show that our distributed algorithms perform better than existing state-of-the-art distributed methods for the above two problems. © 2024 International Foundation for Autonomous Agents and Multiagent Systems.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Contextual Reinforcement Learning for Unsupervised Deformable Multimodal Medical Images Registration 18

Contextual Reinforcement Learning for Unsupervised Deformabl...

引用

18th IEEE International Joint Conference on Biometrics, IJCB 2024

作者： Zheng, Yang Xian, Hongjiang Shuai, Zhikun Hu, Jing Wang, Xin Hu, Shu Chengdu University of Information Technology Department of Computer Science China Department of Epidemiology and Biostatistics School of Public Health NY United States Purdue University Department of Computer and Information Technology IN United States

ISBN: (纸本)9798350364132

Multimodal deformable image registration refers to the process of finding the spatial correspondence between pairs of images with multimodal and mapping them onto the same coordinate system. Most of the deep learning-based registration methods are one-shot registration, which is difficult to handle images with significant deformations or displacements. Reinforcement learning can handle these challenges by viewing registration as a strategic decision-making process which is a step-by-step registration. However, it faces challenges with high-dimensional and continuous deformation fields. To overcome this, we introduce a planner network that maps high-dimensional input state to low-dimensional plan, guiding the actor to generate continuous actions. In order to handle complex multi-modal registration, we propose a multi-frame plan module which encourages artificial agent to explicitly utilize the redundant states in the registration process and learn more accurate registration actions from the generated state frames. To facilitate the training and convergence of the model, we define an unsupervised reward function and incorporate spectral normalization layers. The entire framework is a fully unsupervised registration framework and training in an end-to-end manner. We evaluated our method on publicly available T1w and T2w brain datasets, and the results indicate that our method has excellent deformable registration capability for multimodal images. © 2024 IEEE.

关键词： Image registration

来源：评论

学校读者我要写书评

暂无评论

Sentiment analysis of the Indonesian community toward face-to-face learning during the Covid-19 pandemic

引用

Procedia computer science 2023年 227卷 398-405页

作者： Andrew Giovanni Gozal Hady Pranoto Muhammad Fikri Hasani Computer Science Department School of Computer Science Bina Nusantara University Jakarta Indonesia 11480

This research aims to analyze how the opinions of the Indonesian people in the learning system in the middle of the covid-19 pandemic. The research method is carried out by performing the K-NN method to determine the accuracy level of the data used. The research data method is taken through public comments on Twitter social media using scrapping techniques with appropriate keywords. The data will then be processed through training and testing and classified using the K-NN method. After the data is classified, the accuracy, F1-Score, Recall, and Precision level will be tested using Confusion Matrix. The result showed that KNN performed well, with above 70% of the F1-score for each class. According to the confusion matrix, accuracy also showed promising results with 82%. Future research may include oversampling the class with fewer numbers. K-Fold cross-validation can also be used to see the general performance of the model. The same method may be used to find sentiment towards a political policy that is taken; whether a policy gets a good or bad response, if the response is bad, you can see the causal factors that cause the negative sentiment. In this way, you can find out the will of the public.

关键词： sentiment analysis face to face learning k-nn

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：