检索结果-内蒙古大学图书馆

8th IFIP TC 12 International Conference on computer, Communication and Signal Processing, ICCCSP 2024

作者： Kowshik, Achyuta Siva Sai Sindhur, Ambati Sai Reddy, Annem Gnaneswara Ganesh, Manasa Sivan, Remya Pati, Peeta Basa Department of Computer Science and Engineering Amrita School of Computing Bengaluru Amrita Vishwa Vidyapeetham Bengaluru India

ISBN: (纸本)9783031736162

This experiment focuses on employing machine learning techniques to create an automated system for classifying palm leaf manuscripts into three distinct categories based on their degradation levels: good, bad, and medium. The study collects a wide range of palm leaf samples at different degradation levels. Using advanced machine learning algorithms and an efficient feature extraction method, a model is created that accurately classifies palm leaves by level of degradation. The research incorporates Gabor feature extraction and statistical classifiers to train the extracted features. The classifiers used are k-Nearest Neighbor (k-NN), Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), Logistic Regression (LR), Random Forest (RF), Decision Tree (DT), Naive Bayes (NB), XG Boost (XGB), and Ada Boost (AB). After employing SMOTE, normalization, and feature selection, RF obtained the highest mean accuracy of 87.25% and a standard deviation of 0.08 in comparison with the other classifiers. © IFIP International Federation for Information Processing 2025.

关键词： Random forests

来源：评论

学校读者我要写书评

暂无评论

Predictive Maintenance for Hard Disk Drives Using Random Forest 8th

Predictive Maintenance for Hard Disk Drives Using Random For...

引用

8th IFIP TC 12 International Conference on computer, Communication and Signal Processing, ICCCSP 2024

作者： Baniya, Achal Vinod, Abhay Chinta, Hitesh Surya Manathanathu Pushakara, Bhavith Pati, Peeta Basa Department of Computer Science and Engineering Amrita School of Computing Bengaluru Amrita Vishwa Vidyapeetham Bengaluru India

ISBN: (纸本)9783031736162

As digital data continues to grow exponentially, the reliability of Hard Disk Drives (HDDs) becomes a critical concern for ensuring data integrity and preventing unexpected system downtimes. The research dataset contains historical SMART attribute data collected from HDDs, encompassing parameters such as spin-up time, temperature reallocated sector count and other relevant metrics. The study explored usage of various exploratory data analysis (EDA) & machine learning (ML) techniques for forecasting HDD failures. EDA techniques include Synthetic Minority Over-sampling Technique (SMOTE) algorithm for balancing class representation in training set and Principal Component Analysis (PCA) for dimensionality reduction. Diverse machine learning techniques, such as Random Forest (RF), Decision Tree (DT), K-nearest Neighbors (KNN), Gaussian Naïve Bayes (GNB), Adaptive Boosting (AdaBoost), XGBoost, Gradient Boosting (GB), Linear Discriminant Analysis (LDA), Logistic Regression (LR), are employed to predict the hard drive failure. Local Interpretable Model-agnostic Explanations (LIME) is employed to identify which features have the greatest influence on a specific prediction, providing insights into the factors driving the model’s decisions. Random Forest exhibited superior performance of accuracy 99.9 and F1 score 0.98 compared to other models. © IFIP International Federation for Information Processing 2025.

关键词： Dimensionality reduction

来源：评论

学校读者我要写书评

暂无评论

Shift Equivariant Pose Network

Shift Equivariant Pose Network

引用

2025 IEEE/CVF Winter Conference on Applications of computer Vision, WACV 2025

作者： Wang, Pengxiao Lin, Tzu-Heng Wang, Chunyu Wang, Yizhou School of Computer Science Peking University China Peking University Center on Frontiers of Computing Studies China

ISBN: (纸本)9798331510831

Human pose estimation has been greatly advanced in recent years. However, even the best-performing models are not shift equivariant. In particular, a small change in input images often results in drastic alterations in output, which are problematic especially in video applications. The prevalence of top-down approaches, which typically rely on a (non-equivariant) object detector in the first stage, exac-erbates this issue. In this paper, we first demonstrate that the biased keypoint representation and the non-equivariant network components are the two main obstacles to shift equivariant pose estimation. To address the limitation, we propose an unbiased decoding method, and redesign the necessary network components (e.g., APS-ResBlock, SSP). Extensive experiments show that our method not only produces much more stable results with shifting input, but also achieves better metrics with the ability of tolerating in-accurate detector output from the first stage. To our knowledge, this is the first work to address the problem of shift equivariance in the field of pose estimation. Our method could be easily applied to existing CNN-based pose estimation networks. © 2025 IEEE.

关键词： Electric reactors

来源：评论

学校读者我要写书评

暂无评论

I/O Latency Management in Private Cloud Infrastructures 25th

I/O Latency Management in Private Cloud Infrastructures

引用

25th International Conference on Parallel and Distributed computing, Applications and Technologies, PDCAT 2024

作者： HoseinyFarahabady, MohammadReza Zomaya, Albert Y. School of Computer Science Center for Distributed and High Performance Computing The University of Sydney SydneyNSW Australia

ISBN: (纸本)9789819642069

This paper examines how co-locating multiple VMs on a single physical server with shared storage impacts I/O performance, specifically focusing on latency of I/O operations, and the overall throughput. We introduce a Proportional-Integral-Derivative (PID) controller to maintain I/O latency close to a predefined target by dynamically adjusting I/O requests based on real-time measurements. Through a series of experiments, we evaluated the effectiveness of our PID controller in comparison to VMware’s default QoS mechanism within a private cloud environment. The results indicate that the PID controller significantly enhances performance for latency-critical workloads, achieving substantially higher compliance with latency targets compared to the default mechanism. However, this prioritization of high-priority VMs comes at the expense of increased latency and reduced throughput for lower-priority workloads during periods of I/O contention. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Three term control systems

来源：评论

学校读者我要写书评

暂无评论

RIDynaQ: A DynaQ Based System for Reading Impairment Detection 8th

RIDynaQ: A DynaQ Based System for Reading Impairment Detecti...

引用

8th International Symposium on Intelligent Informatics, ISI 2023

作者： Surisetty, Hima Varshini Gottimukkala, Sarayu Varma Amudha, J. Department of Computer Science and Engineering Amrita School of Computing Amrita Nagar Choodasandra Karnataka Bengaluru India

ISBN: (纸本)9789819721467

Dyslexia or reading impairment is a reading disorder that can have an effect on both children and adults. Though reading impairment affects the reading abilities of an individual it does not have an effect on the intelligence quotient of the individual. It has been proven that eye-gaze behavior can be used as a good indicator of the reading abilities of a person. Therefore, using eye-gaze behavior and the eye-gaze patterns as a determinant of a good reader and a poor reader can help with the identification of reading impairment. An early detection of reading impairment can help the person in getting the required help as soon as possible. In this work, the detection of reading impairment has been modeled as a reinforcement learning problem using a model-based learning approach. The proposed model is named as RIDynaQ which stands for Reading Impairment detection using DynaQ wherein the eye-gaze behavior of both dyslexic and non-dyslexic children was considered and the optimal gaze targets and the optimal policy for reading for both good and poor readers were obtained. Also, a comparison was drawn amidst the pre-existing model-free approach and model-based approach. The main difference between the existing Q learning based approach and RIDynaQ is the use of simulated experiences and developing a model of the environment which does not exist in Q Learning based approach but is used in the RIDynaQ model. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Deepfake Image Detection Using Light-Weight Attention Integrated MobileNetV3 Model 2nd

Deepfake Image Detection Using Light-Weight Attention Integr...

引用

2nd International Conference on Applied Algorithms, ICAA 2025

作者： Harshitha, Talluri Simhadri, Tanya Jaswanthi, Thadakuluru Rayvanth, N. Singh, Rimjhim Padam Department of Computer Science and Engineering Amrita School of Computing Bengaluru Amrita Vishwa Vidyapeetham Bengaluru India

ISBN: (纸本)9783031845420

With sudden proliferation in evolving Deepfake technology to manipulate media content, identity theft etc., several modern deep learning based solutions are being continuously developed for Deepfake detection, distinguishing the manipulated media. Therefore, the work presented in this paper proposes a novel and real-time MobileNetV3 based Deepfake detection model. With the fine-tuning of MobileNetV3 using additional fully connected layers and self-attention mechanism placed optimally, the proposed model proved effective in concentrating on crucial image regions over relatively lesser important features. The integration of self-attention layers had a direct impact on the model’s capacity to identify the artifacts manipulated during the generation of Deepfakes, thereby, alleviating the detection efficiency. The work also implemented and trained five widely used and recent convolutional models including XceptionNet, InceptionV3, EfficientNetB0, MobileNetV3, and InceptionResNet for best baseline model selection and comparisons against the proposed approach. The proposed model outperformed all other state-of-art models by a margin of 3% with an F-score = 98.46%. The proposed model also has minimalistic memory and time requirements supporting the real-time applicability of the model even on mobile devices. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Image Captioning with Multiple Perspectives—A Visual Context-Based Approach 5th

Image Captioning with Multiple Perspectives—A Visual Contex...

引用

5th International Conference on computing and Network Communications, CoCoNet 2023

作者： Ashwin, G. Chaitanya, V. Kishan Rohith Nair, Priyanka C. Department of Computer Science and Engineering Amrita School of Computing Amrita Vishwa Vidyapeetham Bengaluru560035 India

ISBN: (纸本)9789819747108

This study tackles the difficult issues of image captioning while negotiating the complexity of visual data processing. The complexity of visual data and the associated processing requirements make image captioning a daunting task. This research describes an effective method for image captioning that makes advantage of attentional mechanisms. Several models, including Transformer, VGG16, VGG19, Inception, RNN decoders, and Bahdanau’s attention mechanism, are contrasted and analyzed to demonstrate the benefits of integrating attention processes. By allowing the model to focus on the appropriate visual component, attention improves the accuracy of the generated captions. Transformer models, in particular, outperform other methods by capturing complicated relationships and delivering accurate output, and they have the highest BLEU Score of 70. The findings emphasize the significance of attention mechanisms and the relevance of selecting a suitable model architecture to maximize picture captioning performance. The variety of potential applications emphasizes the benefits and potential impact of image captioning in a broad range of scenarios. Several use cases are found beneficial, including supporting the visually impaired, improving product descriptions in e-commerce, assisting medical diagnosis, improving image search ability, and allowing effective communication and comprehension of visual information. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive End-to-End Solution for Web Security with Cryptography, Multi-Factor Authentication, and Secure Communication 3

A Comprehensive End-to-End Solution for Web Security with Cr...

引用

3rd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2025

作者： Kumar, Kanderi Johith Sai, Kandlapalli Aravind Reddy, Adapala Dharshan Reddy, Chinthamreddy Pranay Daiwik Rajagopal, Shinu M. Dept. of Computer Science and Engineering Amrita School of Computing Amrita Vishwa Vidyapeetham Bengaluru India

ISBN: (纸本)9798331527549

In today’s digital landscape, securing sensitive information has become an essential priority in the modern digital era for both individuals and organizations. This project introduces a comprehensive web application that integrates advanced encryption and decryption functionalities alongside multifactor authentication (MFA) and secure payment processing. The application supports various cryptographic algorithms, including AES, DES, Triple DES, Blowfish, etc., allowing users to securely encrypt and decrypt data through dedicated routes accessed over secure HTTPS connections. To further enhance security, robust multifactor authentication mechanisms, such as OTP-based challenge-response and QR-code-enabled 2FA, have been implemented to prevent unauthorized access to sensitive operations. The platform also incorporates PGP encryption powered by Keybase, enabling users to securely exchange encrypted emails. Additionally, a secure payment gateway is integrated using Stripe, ensuring sensitive payment data is encrypted and handled safely. With a strong focus on data integrity, user privacy, and modern web security standards, this project delivers an all-encompassing solution for safeguarding sensitive information and facilitating secure communication. By addressing common vulnerabilities and security risks, the platform stands as a reliable and scalable tool for data protection. © 2025 IEEE.

关键词： Authentication Protocol

来源：评论

学校读者我要写书评

暂无评论

Out-of-Memory GPU Sorting Using Asynchronous CUDA Streams 25th

Out-of-Memory GPU Sorting Using Asynchronous CUDA Streams

引用

25th International Conference on Parallel and Distributed computing, Applications and Technologies, PDCAT 2024

作者： HoseinyFarahabady, MohammadReza Zomaya, Albert Y. School of Computer Science Center for Distributed and High Performance Computing The University of Sydney SydneyNSW Australia

ISBN: (纸本)9789819642069

Graphics Processing Units (GPUs) are widely used as powerful hardware accelerators for data-intensive tasks. However, their efficacy can be hindered by constraints in device memory and data transfer speeds via the PCI-E link. This paper explores the use of GPUs for executing scalable sorting operations that exceeds the GPU’s global memory by a factor of at least two orders of magnitude. To address this challenge, we leverage the power of NVIDIA’s asynchronous CUDA streams. Our solution enhances GPU performance by implementing a pipeline model to offload sorting operations from the Solid-State Drive (SSD) to GPU memory via the host’s main memory. This strategy maximizes the multiway concurrency level to reduce the overhead associated with data transfer within the GPU memory hierarchy. We perform extensive performance evaluations by sorting 128 billion data elements initially stored on an external NVMe SSD. The results confirm the effectiveness of our approach, showcasing a speed-up of 4.3× compared to the results for CPU-based out-of-memory sorting of skewed data based on Intel’s Threading Building Blocks (TBB) sorting algorithm. Our results include the cost of transfer times to and from the main memory, a factor often overlooked in similar studies. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Graph-Based Cooperation Multi-Agent Reinforcement Learning for Intelligent Traffic Signal Control

引用

IEEE Internet of Things Journal 2025年第10期12卷 14362-14374页

作者： Shang, Jing Meng, Shunmei Hou, Jun Zhao, Xiaoran Zhou, Xiaokang Jiang, Rong Qi, Lianyong Li, Qianmu Nanjing University of Science and Technology School of Computer Science and Engineering Nanjing210049 China Nanjing Vocational University of Industry Technology School of Social Science Nanjing210046 China Qufu Normal University School of Computer Science Qufu273165 China Kansai University Faculty of Business and Data Science Osaka Prefecture 564-8680 Japan Yunnan University of Finance and Economics Yunnan Key Laboratory of Service Computing Kunming650399 China College of Computer Science and Technology Qingdao266580 China

In the trend of continuously advancing urban intelligent transport construction, traditional traffic signal control (TSC) struggles to make effective decisions with complex traffic conditions. Although multi-agent deep reinforcement learning (MARL) shows promise in optimizing traffic flow, most existing studies ignore the complex relationships between signal lights and fail to communicate with neighbors effectively. Moreover, the deterministic strategies generated by Q-learning-based methods struggle to be extended to large-scale urban road networks. Therefore, this paper proposes a multi-agent graph-based soft actor-critic (MAGSAC) approach for TSC, which combines graph neural networks with the Soft Actor-Critic (SAC) algorithm and extends it to multi-agent environments to address the TSC problem. Specifically, we employ graph-based networks and attention mechanism to expand the receptive domain of agents, enable environmental information to be shared among agents, and utilize the attention mechanism to filter out unimportant information. The algorithm adheres to the Centralized Training Decentralized Execution (CTDE) paradigm to minimize the non-stationarity of MARL. Finally, a rigorous experimental evaluation was conducted using the CityFlow simulator on both synthetic traffic grids and real-world urban road networks. Experimental results show that MAGSAC outperforms other TSC methods in performance metrics, including average queue length and waiting time, and achieves excellent performance under complex urban traffic conditions. © 2025 IEEE.

关键词： Deep reinforcement 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：