检索结果-内蒙古大学图书馆

iSSL-AL: a deep active learning framework based on self-supervised learning for image classification

Neural Computing and Applications 2024年第28期36卷 17699-17713页

作者： Agha, Rand Mustafa, Ahmad M. Abuein, Qusai Department of Computer Information Systems Jordan University of Science and Technology Irbid22110 Jordan

Deep neural networks have demonstrated exceptional performance across numerous applications. However, DNNs require large amounts of labeled data to avoid overfitting. Unfortunately, the labeled data may not be available;annotating large amounts of data is time-consuming, laborious, and requires human expertise, making it unfeasible to rely on manpower for annotation. One of the solutions to address this limitation is active learning (AL), a technique that utilizes unlabeled data while maintaining high performance. AL reduces the annotation cost of large datasets and enhances the training of models with fewer annotations. Uncertainty sampling has been proven to be one of the most effective strategies in AL;however, it lacks diversity. This research proposes iSSL-AL, a novel active learning framework that utilizes self-supervised learning (SSL) to ensure informative yet diverse samples. Three main aspects categorize the novelty of our work. The first is extending the margin uncertainty sampling by incorporating SSL to select informative and diverse points. The second is employing incremental learning for efficient training of the AL base classifier, where the model is trained incrementally in each AL cycle. The third is addressing the cold start problem, as our framework achieved high results in the early stages of training. Experiments show that iSSL-AL outperforms other state-of-the-art algorithms on the MNIST, FashionMNIST, and CIFAR-10 datasets, with accuracy scores of 99%, 98.9%, and 93.5%, respectively, effectively selecting diverse and informative samples. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

Contemporary Approaches to Analyze Non-stationary Time-series: Some Solutions and Challenges

引用

Recent Advances in computer science and communications 2023年第2期16卷 61-80页

作者： Dixit, Ankit Jain, Shikha Department of Computer Science & Engineering and Information Technology Jaypee Institute of Information Technology Noida India

Enhancement of technology yields more complex time-dependent outcomes for better understanding and analysis. These outcomes generate more complex, unstable, and high-dimensional data from non-stationary environments. Hence, more challenges are arising day by day to fulfill the increasing demand for future estimation. Thus, in this paper, an extensive study has been presented to comprehend the statistical complexity and randomness of Non-Stationary Time Series (NS-TS) data at the atomic level. This survey briefly explains the basic principles and terms related to Non-Stationary Time Series (NS-TS). After understanding the fundamentals of NS-TS, this survey categorized non-stationary time series into groups and subgroups based on a change in statistical behavior. It also provides a comprehensive discussion on contemporary approaches proposed by researchers in each category of non-stationarity. These algorithms include clustering, classification, and regression techniques to deal with different types of domains. Every category of non-stationarity consists of a separate table to draw some advantages and disadvantages of existing approaches. At the end of each non-stationarity type, a short discussion and critical analysis have been done. In the conclusion section, it is observed that this research sphere still has many open challenges that need to be addressed and demand more exploration. Furthermore, it discusses the possible solution of improvisation in future research. © 2023 Bentham science Publishers.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

A Local Differential Privacy Trajectory Protection Method Based on Temporal and Spatial Restrictions for Staying Detection

引用

Tsinghua science and technology 2024年第2期29卷 617-633页

作者： Weiqi Zhang Zhenzhen Xie Akshita Maradapu Vera Venkata Sai Qasim Zia Zaobo He Guisheng Yin College of Computer Science and Technology Harbin Engineering UniversityHarbin 150001China School of Computer Science and Technology Shandong UniversityQingdao 266237China Department of Computer Science Georgia State UniversityAtlantaGA 30303USA College of Information Science and Technology Jinan UniversityGuangzhou 510632China

The widespread availability of GPS has opened up a whole new market that provides a plethora of location-based ***-based social networks have become very popular as they provide end users like us with several such services utilizing GPS through our ***,when users utilize these services,they inevitably expose personal information such as their ID and sensitive location to the *** to untrustworthy servers and malicious attackers with colossal background knowledge,users'personal information is at risk on these ***,many privacy-preserving solutions for protecting trajectories have significantly decreased utility after *** have come up with a new trajectory privacy protection solution that contraposes the area of interest for ***,Staying Points Detection Method based on Temporal-Spatial Restrictions(SPDM-TSR)is an interest area mining method based on temporal-spatial restrictions,which can clearly distinguish between staying and moving ***,our privacy protection mechanism focuses on the user's areas of interest rather than the entire ***,our proposed mechanism does not rely on third-party service providers and the attackers'background knowledge *** test our models on real datasets,and the results indicate that our proposed algorithm can provide a high standard privacy guarantee as well as data availability.

关键词： local-differential privacy stay points detection trajectory data areas of interest temporal-spatial clustering

来源：评论

学校读者我要写书评

暂无评论

Video captioning – a survey

引用

Multimedia Tools and Applications 2025年第2期84卷 947-978页

作者： Vaishnavi, J. Narmatha, V. Department of Computer Applications SRM Institute of Science and Technology Trichy India Department of Computer and Information Science Annamalai University Annamalai Nagar Tamilnadu 608002 India

The explosion of the novel phenomenon of the combination of computer vision and Natural language processing is playing a vital role in converting the ordinary world into a more technological pool. Natural language processing and computer vision are the vanguard of Artificial intelligence with the enormous potential currently ruling more fields. The trending Accelerated research of the NLP and CV combo are cancer screening, surgical simulation, visual properties description, visual retrieval, visual description, etc. Video captioning is a task of localizing the event and various features are extracted from the particular event and then the description of the entire event is performed with Natural languages using various techniques. The task of video captioning has some evolution of various techniques from Traditional methods to Deep learning. Currently, Deep learning techniques are ruling the field. Several models are being proposed for the better enhancement of the video captioning task with the default Encoder-Decoder framework and some models with the variance of Attention mechanism and Transfer learning etc., The Attainment of every technique and the accuracy of the generated results are highly dependent on the nature of the problem, diversity, performance of chosen dataset and construction of the technique with various layers, and the amount of data split for train, validate, and test. This survey discusses the captioning types, real-time applications of video captioning, evolution of various methods in video captioning, advanced techniques, different datasets, metrics that are used to evaluate the results, and analysis of the results of the existing models for video captioning. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Multi-Scale Time Series Segmentation Network Based on Eddy Current Testing for Detecting Surface Metal Defects

引用

IEEE/CAA Journal of Automatica Sinica 2025年第3期12卷 528-538页

作者： Xiaorui Li Xiaojuan Ban Haoran Qiao Zhaolin Yuan Hong-Ning Dai Chao Yao Yu Guo Mohammad S.Obaidat George Q.Huang the School of Intelligence Science and Technology University of Science and Technology Beijing the Beijing Advanced Innovation Center for Materials Genome Engineering the Key Laboratory of Intelligent Bionic Unmanned Systems and the Institute of Materials Intelligent Technology Liaoning Academy of Materials IEEE the Department of Computer Science Hong Kong Baptist University the School of Computer and Communication Engineering Key Laboratory of Advanced Materials and Devices for Post-Moore Chips Ministry of Education University of Science and Technology Beijing the Beijing Advanced Innovation Center for Materials Genome Engineering University of Science and Technology Beijing the School of Computer and Communication Engineering University of Science and Technology Beijing the King Abdullah Ⅱ School of Information Technology The University of Jordan the Department of Computational Intelligence the School of Computing SRM University the School of Engineering The Amity University The Hong Kong Polytechnic University

In high-risk industrial environments like nuclear power plants, precise defect identification and localization are essential for maintaining production stability and safety. However, the complexity of such a harsh environment leads to significant variations in the shape and size of the defects. To address this challenge, we propose the multivariate time series segmentation network(MSSN), which adopts a multiscale convolutional network with multi-stage and depth-separable convolutions for efficient feature extraction through variable-length templates. To tackle the classification difficulty caused by structural signal variance, MSSN employs logarithmic normalization to adjust instance distributions. Furthermore, it integrates classification with smoothing loss functions to accurately identify defect segments amid similar structural and defect signal subsequences. Our algorithm evaluated on both the Mackey-Glass dataset and industrial dataset achieves over 95% localization and demonstrates the capture capability on the synthetic dataset. In a nuclear plant's heat transfer tube dataset, it captures 90% of defect instances with75% middle localization F1 score.

关键词： Eddy current testing nondestructive testing semantic segmentation time series analysis

来源：评论

学校读者我要写书评

暂无评论

Exploring the Potential of Deep Learning Algorithms in Medical Image Processing: A Comprehensive Analysis 5th

Exploring the Potential of Deep Learning Algorithms in Medic...

引用

5th International Conference on communication and Intelligent Systems, ICCIS 2023

作者： Pal, Ganesh Prasad Pal, Raju Department of Computer Science and Engineering Jaypee Institute of Information Technology Noida India Department of Computer Science and Engineering School of Information and Communication Technology Gautam Buddha University Greater Noida India

ISBN: (纸本)9789819720781

In this study, we comprehensively examine the potential of deep learning algorithms in the domain of medical image processing. Through a systematic analysis of existing literature, we explore the applications, methodologies, and outcomes of these algorithms across diverse medical specialties. Our analysis underscores the transformative impact of convolutional neural networks and transfer learning techniques in enhancing diagnostic accuracy, image segmentation, and disease detection. We discuss challenges such as data variability, ethics, and interpretability while emphasizing the importance of interdisciplinary collaboration. By synthesizing findings and identifying future directions, we highlight the promising role of deep learning in revolutionizing healthcare diagnostics and treatment planning. After conducting a thorough analysis of various literature and empirical studies, we have evaluated the capabilities and drawbacks of deep learning models in managing different medical imaging modalities such as X-rays, MRIs, and CT scans. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

SAR Images Change Detection Based on Attention Mechanism-Convolutional Wavelet Neural Network

引用

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 2025年 18卷 12133-12149页

作者： Jiahui, E. Wang, Lu Zhao, Chunhui Mathiopoulos, P. Takis Ohtsuki, Tomoaki Adachi, Fumiyuki Harbin Engineering University College of Information and Communication Engineering Harbin150001 China Ministry of Industry and Information Technology Key Laboratory of Advanced Marine Communication and Information Technology Harbin150001 China National and Kapodistrian University of Athens Department of Informatics and Telecommunications Greece Keio University Department of Information and Computer Science Kanagawa223852 Japan Tohoku University International Research Institute of Disaster Science Sendai Japan

As Synthetic Aperture Radar (SAR) image change detection can continuously acquire target information under all weather conditions, it has been used in the past for various applications. However, it is very challenging to use SAR images for small area change detection due to the presence of noise and the imbalance between classes. To deal with these problems this paper proposes a SAR images change detection scheme which is based upon an Attention Mechanism and Convolutional Wavelet Neural Network (AMCW-NN). Firstly, employing Multi-Scale Superpixel Reconstructed Difference Image (MSRDI) effectively enhances the edge information of the images. Secondly, a twostage Weighted Center Constrained Fuzzy C-Means (WCCFCM) clustering algorithm is proposed. To constrain incorrect migration in the subsequent clustering phase, the algorithm uses a two-stage clustering algorithm. During the first stage weights in the clustering process to improve the accuracy of the resulting clustering centers are introduced while during the second stage, these refined cluster centers can be constrained. Thirdly, a Convolutional Squeeze-and-Excitation (SE) Attention Mechanism Wavelet Neural Network (CSWNN) is proposed which more accurately distinguishes uncertain pixels and thus reduces redundant features. The final Denoising Diffusion Probabilistic Model (DDPM) is introduced into the CSWNN to expand pseudo-labeled samples. Comparing this algorithm with seven sets of contrast experiments on six real SAR datasets shows that the accuracy of detecting changes in small regions is 2%-9% better than the classical algorithm. © 2008-2012 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

IensNet: A novel and efficient approach for iris spoof detection via ensemble of deep models

引用

Multimedia Tools and Applications 2025年 1-30页

作者： Sharma, Deepika Selwal, Arvind School of Computer Science Engineering and Technology Bennett University Greater Noida201310 India Department of Computer Science and Information Technology Central University of Jammu Samba India

Iris biometrics allow contactless authentication, which makes it widely deployed human recognition mechanisms since the couple of years. Susceptibility of iris identification systems remains a challenging task due to diversity in spoof or presentation attacks (PAs) that fails to assure consistency while adopting them in real life scenarios. Hence, iris PAs are the growing concerns that gained significant attention in recent past decade. To alleviate these attacks or recognize presentation attack instruments (PAIs), iris presentation attacks detection (IPAD) algorithms are designed to distinguish a real and fabricated iris trait. Aiming at the efficient iris spoof detection mechanism, in this research work we expound a novel ensemble learning-enabled model (IensNet) that learns three pre-trained and fined-tuned deep models (i.e. DenseNet161, ResNet and VGGNet) for better accuracy and generalized performance. The novel IensNet approach offers several merits (i.e. consolidated strengths of multiple models, improved generalization ability, etc.) as compared to a simple transfer learning strategy where the knowledge is drawn from single pre-trained model. Finally, our approach learns a novel fully-connected dual layer classifier via outcome of three fine-tuned models to yield a final classification result as bonafide or spoof iris trait. Our approach is evaluated on Notre Dame LivDet iris 2017 and Notre Dame contact lenses 2015 anti-spoofing datasets. The experimental analysis of IensNet offers outstanding performance with a lower ACER of 0.2% and 1.4% for Iris-LivDet-2017 and Notre Dame contact lenses 2015 dataset respectively. Besides, IensNet exhibit promising results in cross-dataset environment with an ACA of 91.46%. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Randomization Strategies in Image Steganography Techniques:A Review

引用

computers, Materials & Continua 2024年第8期80卷 3139-3171页

作者： AFM Zainul Abadin Rossilawati Sulaiman Mohammad Kamrul Hasan Center for Cyber Security Faculty of Information Science and TechnologyUniversiti Kebangsaan Malaysia(UKM)BangiSelangor43600Malaysia Department of Information and Communication Engineering Pabna University of Science and TechnologyPabna6600Bangladesh

Image steganography is one of the prominent technologies in data hiding *** system performance mostly depends on the embedding *** goal is to embed strictly confidential information into images without causing perceptible changes in the original *** randomization strategies in data embedding techniques may utilize random domains,pixels,or region-of-interest for concealing secrets into a cover image,preventing information from being discovered by an *** implementation of an appropriate embedding technique can achieve a fair balance between embedding capability and stego image imperceptibility,but it is challenging.A systematic approach is used with a standard methodology to carry out this *** review concentrates on the critical examination of several embedding strategies,incorporating experimental results with state-of-the-art methods emphasizing the robustness,security,payload capacity,and visual quality metrics of the stego *** fundamental ideas of steganography are presented in this work,along with a unique viewpoint that sets it apart from previous works by highlighting research gaps,important problems,and ***,it offers a discussion of suggested directions for future study to advance and investigate uncharted territory in image steganography.

关键词： information hiding image steganography randomized embedding techniques payload capacity imperceptibility

来源：评论

学校读者我要写书评

暂无评论

An ensemble machine learning based bank loan approval predictions system with a smart application

引用

International Journal of Cognitive Computing in Engineering 2023年第1期4卷 327-339页

作者： Uddin, Nazim Uddin Ahamed, Md. Khabir Uddin, Md Ashraf Islam, Md. Manwarul Talukder, Md. Alamin Aryal, Sunil Department of Information and Communication Technology Chandpur Science and Technology University Chandpur Bangladesh Department of Computer Science and Engineering Jagannath University Dhaka Bangladesh Department of Computer Science and Engineering International University of Business Agriculture and Technology Dhaka Bangladesh School of Information Technology Deakin University Waurn Ponds Campus Geelong Australia

Banks rely heavily on loans as a primary source of revenue;however, distinguishing deserving applicants who will reliably repay loans presents an ongoing challenge. Conventional selection processes often struggle to identify the most suitable candidates from a pool of loan applicants. In response to this challenge, we present an innovative machine learning (ML) based loan prediction system designed to identify qualified loan applicants autonomously. This comprehensive study encompasses data preprocessing, effective data balancing using SMOTE, and the implementation of diverse ML models, including Logistic Regression, Decision Tree, Random Forest, Extra Trees, Support Vector Machine, K-Nearest Neighbors, Gaussian Naive Bayes, AdaBoost, Gradient Boosting, and advanced deep learning models such as deep neural networks, recurrent neural networks, and long short-term memory models. The model's performance is rigorously assessed in terms of accuracy, recall, and F1_score. Our experimental analysis reveals that the Extra Trees outperforms its counterparts. Furthermore, we successfully predict bank loan defaulters through an ensemble voting model, which includes the top three ML models, achieving a remarkable 0.62% increase in accuracy compared to the Extra Trees. To facilitate user interaction, we have developed a user-friendly desktop-based application. Notably, our findings demonstrate that the voting-based ensemble model surpasses both individual ML models, including Extra Trees, and existing state-of-the-art approaches, achieving an impressive accuracy of 87.26%. This innovative system has the potential to significantly streamline and enhance the efficiency of bank loan approval processes, ultimately benefiting both financial institutions and loan applicants alike. © 2023 The Author(s)

关键词： User interfaces

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：