检索结果-内蒙古大学图书馆

Image Inpainting Detection Based on High-Pass Filter Attention Network

computer Systems science & Engineering 2022年第12期43卷 1145-1154页

作者： Can Xiao Feng Li Dengyong Zhang Pu Huang Xiangling Ding Victor S.Sheng Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation Changsha University of Science and TechnologyChangsha410114China School of Computer and Communication Engineering Changsha University of Science and TechnologyChangsha410114China School of Computer Science and Engineering Hunan University of Science and TechnologyXiangtan411004China Department of Computer Science Texas Tech UniversityLubbock79409USA

Image inpainting based on deep learning has been greatly *** original purpose of image inpainting was to repair some broken photos, suchas inpainting artifacts. However, it may also be used for malicious operations,such as destroying evidence. Therefore, detection and localization of imageinpainting operations are essential. Recent research shows that high-pass filteringfull convolutional network (HPFCN) is applied to image inpainting detection andachieves good results. However, those methods did not consider the spatial location and channel information of the feature map. To solve these shortcomings, weintroduce the squeezed excitation blocks (SE) and propose a high-pass filter attention full convolutional network (HPACN). In feature extraction, we apply concurrent spatial and channel attention (scSE) to enhance feature extraction and obtainmore information. Channel attention (cSE) is introduced in upsampling toenhance detection and localization. The experimental results show that the proposed method can achieve improvement on ImageNet.

关键词： Image inpainting detection spatial attention channel attention full convolutional network high-pass filter

来源：评论

学校读者我要写书评

暂无评论

Tree Counting and Health Assessment in Central Jakarta, Indonesia using Aerial Imagery data with Deep Learning

Tree Counting and Health Assessment in Central Jakarta, Indo...

引用

2023 International Conference on Artificial Intelligence Robotics, Signal and Image Processing, AIRoSIP 2023

作者： Hananto Kusumo, Rafael Edwin Gravita, Adella Prasojo, Ricky Irwansyah, Edy Bina Nusantara University School of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350327472

Trees play a crucial role as one of the primary oxygen producers. They have particular significance in urban areas, where they contribute to residents physical and mental well-being by providing a healthy living environment. With the important role that trees play in the ecosystem, it becomes imperative to maintain and monitor tree populations. Counting and mapping existing trees is a fundamental step in this process, and deep learning methods offer a practical solution. This research employed the DeepForest algorithm, a Python package capable of detecting trees in aerial RGB imagery. Specifically, this study utilized aerial imagery data to study the Gelora Bung Karno (GBK) area and TPU Karet Bivak in Central Jakarta, Indonesia. The results using the DeepForest algorithm yielded F1 scores of 21% accuracy for the first GBK area, 60% accuracy for the second GBK area, and 70% accuracy for the TPU Karet Bivak area. The lower accuracy in GBK area is influenced by the dense canopy, making it more difficult to detect individual trees. Alongside the DeepForest algorithm, this study also includes tree health assessment by using Visible Atmospherically Resistant Index (VARI) calculation. The VARI analysis relies on the tree counting results from TPU Karet Bivak due to its higher accuracy. The analysis categorizes trees into different classes based on their different levels of health, enabling a more comprehensive review and maintenance process. Specifically, the study identifies 27 trees in needs inspection, 41 trees in declining health, 15 trees in moderate health, and 2 trees identified as healthy. © 2023 IEEE.

关键词： Aerial photography

来源：评论

学校读者我要写书评

暂无评论

Energy and Bandwidth Based Link Stability Routing Algorithm for IoT

引用

computers, Materials & Continua 2022年第2期70卷 3875-3890页

作者： D.Kothandaraman A.Balasundaram R.Dhanalakshmi Arun Kumar Sivaraman S.Ashokkumar Rajiv Vincent M.Rajesh School of Computer Science and Artificial Intelligence SR UniversityWarangalTSIndia School of Computer Science and Engineering Center for Cyber Physical SystemsVellore Institute of Technology(VIT)ChennaiIndia Department of Computer Science and Engineering KCG College of TechnologyChennaiIndia School of Computer Science and Engineering Vellore Institute of Technology(VIT)ChennaiIndia Department of Computer Science and Engineering Saveetha School of EngineeringSIMATSChennaiIndia

Internet of Things(IoT)is becoming popular nowadays for collecting and sharing the data from the nodes and among the nodes using internet ***,some of the nodes in IoT are mobile and dynamic in *** maintaining the link among the nodes,efficient bandwidth of the links among the mobile nodes with increased life time is a big challenge in IoT as it integrates mobile nodes with static nodes for data *** such networks,many routing-problems arise due to difficulties in energy and bandwidth based quality of *** to the mobility and finite nature of the nodes,transmission links between intermediary nodes may fail frequently,thus affecting the routing-performance of the network and the accessibility of the *** existing protocols do not focus on the transmission links and energy,bandwidth and link stability of the nodes,but node links are significant factors for enhancing the quality of the *** stability helps us to define whether the node is within or out of a coverage *** paper proposed an Optimal Energy and bandwidth based Link Stability Routing(OEBLS)algorithm,to improve the link stable route with minimized error rate and *** this paper,the optimal route from the source to the sink is determined based on the energy and bandwidth,link stability *** the existing routes,the sink node will choose the optimal route which is having less link stability *** stable link is determined by evaluating link stability value using distance and ***-energy of the node is estimated using the current energy and the consumed *** energy is estimated using transmitted power and the received *** bandwidth in the link is estimated using the idle time and channel capacity with the consideration of probability of collision.

关键词： Link stability internet of things optimal energy optimal bandwidth residual energy

来源：评论

学校读者我要写书评

暂无评论

Comparative Analysis of Binary and Interpolation Search Algorithms on Integer data Using C programming Language 8

Comparative Analysis of Binary and Interpolation Search Algo...

引用

8th International Conference on Information Management and Technology, ICIMTech 2023

作者： Saputra, Bima Andri Candra, Stevans Calvin Wijaya, Franis Berta Suryaningrum, Kristien Margi Saputri, Hanis Amalia School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350326093

To perform data search, there are various algorithms that have been developed, such as binary search and interpolation search. There are various views on these two algorithms, regarding which one is better than the other. From the various views that have emerged, this study aims to find out which algorithm is more efficient in terms of time complexity and memory complexity using Big-O notation, as well as its execution time in the C programming language. This research was conducted with 6 datasets consisting of 3 datasets with uniform data types and 3 datasets with non-uniform data types. Each dataset type consists of 1 million, 10 million, and 100 million data. The results show that the binary search algorithm is more efficient to use rather than interpolation search on datasets that are not uniformly distributed or the data type is unknown. While the interpolation search algorithm is more efficient to use rather than binary search on datasets that are known to be uniformly distributed. © 2023 IEEE.

关键词： Interpolation

来源：评论

学校读者我要写书评

暂无评论

LDREG: LOCAL DIMENSIONALITY REGULARIZED SELF-SUPERVISED LEARNING 12

LDREG: LOCAL DIMENSIONALITY REGULARIZED SELF-SUPERVISED LEAR...

引用

12th International Conference on Learning Representations, ICLR 2024

作者： Huang, Hanxun Campello, Ricardo J.G.B. Erfani, Sarah Ma, Xingjun Houle, Michael E. Bailey, James School of Computing and Information Systems The University of Melbourne Australia Department of Mathematics and Computer Science University of Southern Denmark Denmark School of Computer Science Fudan University China Department of Computer Science New Jersey Institute of Technology United States

Representations learned via self-supervised learning (SSL) can be susceptible to dimensional collapse, where the learned representation subspace is of extremely low dimensionality and thus fails to represent the full data distribution and modalities. Dimensional collapse - also known as the "underfilling" phenomenon -is one of the major causes of degraded performance on downstream tasks. Previous work has investigated the dimensional collapse problem of SSL at a global level. In this paper, we demonstrate that representations can span over high dimensional space globally, but collapse locally. To address this, we propose a method called local dimensionality regularization (LDReg). Our formulation is based on the derivation of the Fisher-Rao metric to compare and optimize local distance distributions at an asymptotically small radius for each data point. By increasing the local intrinsic dimensionality, we demonstrate through a range of experiments that LDReg improves the representation quality of SSL. The results also show that LDReg can regularize dimensionality at both local and global levels. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Federated Learning and Differential Privacy in AI-Based Surveillance Systems Model 14

Federated Learning and Differential Privacy in AI-Based Surv...

引用

14th International Conference on Information and Communication Technology and System, ICTS 2023

作者： Adiwijaya, Jason Tanaya, Venansius Reynardi Anderies Chowanda, Andry Bina Nusantara University School of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350312164

Surveillance systems, particularly AI-based ones, play a crucial role in enhancing public security. However, they have also raised significant concerns about privacy due to the sensitive nature of the collected and analyzed data. This research highlights the importance of data privacy without compromising the accuracy of AI-based surveillance systems. We propose the integration of two privacy-preserving methods, Federated Learning (FL) and Differential Privacy (DP), to enhance the security and privacy of data used in the training and testing processes. Federated Learning enables the training of AI models using distributed data without needing to centralize all the data on a single device. In contrast, Differential Privacy offers a statistical framework that safeguards individual data points throughout the training process. Both methods were implemented individually and combined on the chosen AI model. Our findings reveal a trade-off between privacy preservation and model accuracy when applying these privacy-preserving techniques to an AI surveillance system. While there was a reduction in overall accuracy, privacy preservation capabilities yielded satisfactory results. Notably, the optimal results were obtained when combining Federated Learning and Differential Privacy with a noise multiplier of 0.6 and 50 devices participating in the federated learning process, yielded an overall solid accuracy of 69.33, and maintained acceptable confidence levels across different categories. This configuration underscores the need for a careful balance between privacy protection and model performance in AI surveillance systems, opening avenues for future research on refining these techniques. Our code is available at: https://***/slimmyYer211/RMCS-PPCV_01 © 2023 IEEE.

关键词： Privacy-preserving techniques

来源：评论

学校读者我要写书评

暂无评论

Enhancing Hate Speech Detection in Social Media Using IndoBERT Model: A Study of Sentiment Analysis during the 2024 Indonesia Presidential Election 5

Enhancing Hate Speech Detection in Social Media Using IndoBE...

引用

5th International Conference on computer and Applications, ICCA 2023

作者： Yulfa, Ramadhan Ihsani Setiawan, Benediktus Hengki Lourensius, Gerry Gilbert Purwandari, Kartika Bina Nusantara University School of Computer Science Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350303254

Legislative elections, presidential elections, and village head elections are all on the schedule for Indonesia in 2024. Law No. 7 of 2017 governs these elections, emphasizing the ideals of direct, public, free, confidential, honest, and fair elections. With the growth of social media, political personalities are employing these channels for campaign activity, resulting in a variety of user responses, including hate speech. Detecting hate speech is critical for preventing and mitigating the bad impact it may have on society. The purpose of this article is to investigate the usage of IndoBERT, a fine-tuned BERT model, to improve hate-speech identification during the 2024 Indonesia Presidential Election. The research focuses on hate speech in the context of the presidential election and the challenges raised by social media. In this paper, we utilized Twitter as a case study to develop effective hate-speech identification techniques. The method comprises fine-tuning IndoBERT using a dataset of Indonesian tweets on the election that have been preprocessed to reduce noise and classified as hate speech or non-hate speech using a pre-trained model. The accuracy, precision, recall, and F1-score of the model are used to assess its performance. The results reveal that the IndoBERT model detects hate speech in Indonesian Twitter data with high accuracy. A future study might look towards fixing typos in the dataset and examining expert recognition of hate speech scenarios. © 2023 IEEE.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Super-resolved Estimation of White Matter Microstructure via 3D Conditional Latent Diffusion Model

Super-resolved Estimation of White Matter Microstructure via...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Ma, Jiquan Gao, Yihang Jiang, Haotian Gu, Jing Khairullina, Diliara Cui, Hui Chen, Geng Heilongjiang University Department of Computer Science and Technology Harbin China Northwestern Polytechnical University School of Computer Science and Engineering Xi'an China La Trobe University Department of Computer Science and Information Technology Melbourne Australia

ISBN: (纸本)9798350386226

As a powerful microstructural imaging technique, neurite orientation dispersion and density imaging (NODDI) provides detailed insights into brain microstructures. Its clinical application is often restricted by the necessity for high-quality scanning, which can be challenging to achieve in practical settings. To overcome this limitation, we propose an innovative 3D conditional latent diffusion model (3D-CLDM) to generate high-quality NODDI index maps from low-resolution diffusion magnetic resonance imaging data. The 3D-CLDM is a two-stage super-resolved microstructure estimation model that includes training a vector quantized generative adversarial network and a diffusion model. It leverages the sophisticated high-dimensional data modeling capabilities of the conditional latent diffusion model to effectively capture and represent intricate microstructural features that are difficult to detect with conventional techniques. We conducted comprehensive experiments using data from the human connectome project to rigorously assess our model's performance. The results reveal that our approach not only significantly improves the quality of super-resolved microstructural estimation but also surpasses current state-of-the-art models in both qualitative and quantitative evaluations. This highlights the potential of 3D-CLDM to advance brain microstructure imaging, making it more feasible and effective for clinical applications. © 2024 IEEE.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Seamless High Availability in MySQL database Systems: An Advanced Implementation of ProxySQL, Orchestrator, and Keepalived

Seamless High Availability in MySQL Database Systems: An Adv...

引用

2025 International Conference on Advancement in data science, E-learning and Information System, ICADEIS 2025

作者： Loa, Jeremy Soewito, Benfano Bina Nusantara University Binus Graduate Program Computer Science Department Jakarta Indonesia

ISBN: (纸本)9798331513320

High availability (HA) in MySQL databases remains a critical challenge, particularly due to risks such as single point of failure (SPOF) and limited scalability in existing setups. Previous studies have explored various approaches to address these issues. One study implemented ProxySQL and Orchestrator for query routing and failover but relied on a single server instance, creating a SPOF that could disrupt client access during failures. Another study implemented HAProxy to balance traffic and couple database servers with specific web applications. While effective for certain use cases, this approach reduced resource efficiency and scalability, because it relies on HTTP protocol and database server are coupled to each application, making it less suitable for dynamic, shared environments requiring high flexibility. These limitations highlight the need for a more robust and scalable HA solution. This paper proposes an improved HA architecture that eliminates SPOFs and enhances scalability by combining redundant ProxySQL instances synchronized with Keepalived managed Virtual IP (VIP) and automated failover using Orchestrator. The architecture allows shared database resources across multiple applications while decoupling them from specific application servers. Test results confirm the system's ability to maintain seamless failover, consistent query routing, and uninterrupted client access through dynamic Virtual IP reassignment to backup server. By addressing the issues of previous studies, this solution provides a practical and scalable framework to achieve high availability in the MySQL cluster by combing ProxySQL, Orchestrator, and Keepalived, making it suitable for modern enterprise environments. © 2025 IEEE.

关键词： HTTP

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Classifier to Recommend the Tourist Attraction in Smart Cities 2

Deep Learning Classifier to Recommend the Tourist Attraction...

引用

2nd International Conference on Disruptive Technologies, ICDT 2024

作者： Gupta, Keshav Kumar, Vikas Jain, Abhishek Singh, Pranita Jain, Amit Kumar Prasad, M.S.R. Grd Institute of Management and Technology Department of Computer Science and Engineering Dehradun India School of Computer Science & Engineering Dev Bhoomi Uttrakhand University Dehradun India Uttaranchal University Department of Computer Science and Engineering Dehradun India Roorkee Institute of Technology Department of Computer Science and Engineering Roorkee India Koneru Lakshmaiah Education Foundation Department of Computer Science and Engineering Vaddeswaram India

ISBN: (纸本)9798350371055

Provide an attraction recommendation system that uses deep learning and is powered by the Internet of Things (IoT) to develop the smart city visitor experience. Users of a smart city app or website will be able to record information about their trips, including whether they are going independently or with an accompanying companion, the sort of companion (e.g., a family with children or a lover), the purpose of their trip (business or pleasure), and personal characteristics (e.g., age, interests, etc.). Using this unique combination of input data, the suggested deep learning-based recommendations system would suggest tourist activities and sights that match the user's profile nicely. In addition, the suggested tourist attraction recommendations system uses actual data obtained from Internet of Things (IoT) devices to offer content-based information (such as previously looked-at sights) and context related data (such as weather conditions, location, time of day, etc.) to potential visitors in the smart city. Regarding the initial case [(a) planning and searching for activities earlier travelling], the suggested multi-label deep learning classifiers achieve a recall, accuracy, precision, and loss of 1.6%, 98.6%, 98.8%, and 98.7%, respectively, compared to other models such as extra tree, decision tree, k-nearest neighbour, and randomized forest. As for the second scenario, it can effectively suggest tourist spots while searching for smart city activities (b) with a recall, accuracy, precision, and loss of 2.6%, 98.4%, 98.6%, and 98.7%, respectively. © 2024 IEEE.

关键词： Deep Learning Internet of Things Personal Characteristics Recommendation System Smart City Activities

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：