检索结果-内蒙古大学图书馆

Enhancing Voice Phishing Detection Using Multilingual Back-Translation and SMOTE: An Empirical Study

IEEE Access 2025年 13卷 37946-37965页

作者： Boussougou, Milandu Keith Moussavou Hamandawana, Prince Park, Dong-Joo Soongsil University Department of Computer Science and Engineering Seoul06978 Korea Republic of Ajou University Department of Software and Computer Engineering Suwon16499 Korea Republic of Soongsil University School of Computer Science and Engineering Seoul06978 Korea Republic of

With the widespread global trend of voice phishing or vishing attacks, the development of effective detection models using artificial intelligence (AI) has been hindered by the lack of high-quality and large volumes of data. This lack of data reflecting a real vishing scenario often leads to imbalanced datasets and biased detection models. Therefore, we present in this paper a data augmentation (DA) method for expanding the imbalanced Korean call content vishing (KorCCVi) dataset to address the existing data asymmetry problem and enhance the performance of Korean vishing detection. The proposed approach for DA involves using the back-translation (BT) method with three different intermediate languages: English, Chinese, and Japanese. The proposed method offers several advantages over the traditional synthetic minority oversampling technique (SMOTE), which is the main technique used to compare with our multilingual BT approach. Using these two DA techniques, several machine learning (ML) and deep learning (DL) models were trained on the original imbalanced dataset, the dataset balanced with SMOTE and its variants, and the dataset augmented with our method. We analyzed the impact of these DA methods on the performance of the models, demonstrated the benefits of each approach, and suggested the most suitable approach. The performance of the trained models was evaluated using the accuracy, precision, recall, and F1-score metrics. The experimental results demonstrated that the proposed multilingual BT method effectively expands the dataset while preserving its contextual and linguistic characteristics. The average performance of the models revealed that those trained on the augmented dataset outperformed the other models. They achieved F1-scores of 98.91% for the back-translated data, 98.14% for the original data, and 97.23% for SMOTE. © 2025 The Authors. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecomm

关键词： Phishing

来源：评论

学校读者我要写书评

暂无评论

Time Series Forecasting for COVID-19 Confirmed Cases Using Transformer Based Stacked LSTM Model 8th

Time Series Forecasting for COVID-19 Confirmed Cases Using T...

引用

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

作者： Penumadi, Sri Naga Jathin Nambiar, Vaishakh Vignesh, G.S. Singh, Rimjhim Padam Kanchan, Sneha Department of Computer Science and Engineering Amrita School of Computing Bengaluru Amrita Vishwa Vidyapeetham Bengaluru India Department of Internet Engineering and Computer Science Universiti Tunku Abdul Rahman Perak Malaysia

ISBN: (纸本)9783031736162

The worldwide impact of COVID-19 necessitates accurate forecasting for informed decision-making by governments and health entities. This study proposes neural network architectures for time series forecasting of confirmed COVID-19 cases based on historical data. The models, including variants of LSTM networks, CNN-GRU hybrids, RBMs, DBNs, and self-attention mechanisms, are trained on comprehensive datasets of confirmed cases, deaths, recoveries, and other variables across various countries. Comparative analysis indicates that CNN-GRU and LSTM models exhibit superior performance in forecasting accuracy, surpassing RBM, DBN, and self-attention models across metrics like Mean Absolute Error and Root Mean Squared Error. This research emphasizes the viability of employing advanced neural networks for precise COVID-19 spread predictions, pivotal in shaping public health policies during global crises. © IFIP International Federation for Information Processing 2025.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

The impact of backbone selection in YOLOv8 models on brain tumor localization

引用

Iran Journal of computer science 2025年 1-23页

作者： Ranjbarzadeh, Ramin Crane, Martin Bendechache, Malika School of Computing Faculty of Engineering and Computing Dublin City University Dublin Ireland ADAPT Research Centre School of Computer Science University of Galway Galway Ireland

This study investigates the utilization of the You Only Look Once (YOLOv8) deep learning framework for accurately identifying the location of brain tumors in medical imaging. We investigate the effects of model size and pretraining on the accuracy and computational efficiency of tumor detection by utilizing different setups of the YOLOv8 model. These setups include various configurations, ranging from very small to large, and can be pretrained on the COCO dataset or not. The experimental results, carried out on Google Colaboratory using NVIDIA Tesla T4 GPUs, show that pretrained models often achieve better performance by utilizing the extensive feature representations learned from the COCO dataset, resulting in increased precision in tumor location. For instance, the YOLOv8-XS model pretrained on COCO achieves an IoU of 0.278 and a Dice coefficient of 0.435, whereas its non-pretrained counterpart attains only 0.241 IoU and 0.388 Dice, indicating a 15% improvement in tumor localization accuracy. Similarly, pretrained YOLOv8-L achieves 0.269 IoU, outperforming standard object detection models such as Mask R-CNN (IoU: 0.212) and Faster R-CNN (IoU: 0.228). These results highlight the impact of pretraining on model performance, particularly for lightweight architectures, while also revealing diminishing returns for larger models. The research uncovers a subtle connection between the complexity of the model, pretraining, and the time required for training. It emphasizes the possible advantages and constraints of pretraining for various sizes of models. © The Author(s) 2025.

关键词： Brain tumor localization COCO dataset Deep learning Object detection Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Segmenting the Brain Tumor using Deep Learning Techniques 3

Segmenting the Brain Tumor using Deep Learning Techniques

引用

3rd International Conference on Electronics and Renewable Systems, ICEARS 2025

作者： Pranitha, Ch. Lakshmi Sumathi, R. Harshitha, V. Kavya, N. Soni, A. Kalasalingam Academy of Research and Education Department of Computer Science and Engineering Tamil Nadu India Kalasalingam Academy of Research and Education School of Computing Department of Computer Science and Engineering Tamil Nadu India

ISBN: (纸本)9798331509675

This study presents a U-Net architecture with EfficientNet-B7 as the feature extraction backbone is used to propose a robust brain tumor segmentation method. Using cutting-edge deep learning techniques, the suggested model precisely segments MRI images with an emphasis on identifying and defining tumor anomalies. A thorough preprocessing pipeline that handles missing values, augments data with albumentations, and normalizes data guarantees the consistency and resilience of the incoming data. The segmentation procedure uses a hybrid loss function that combines Soft Dice Loss and Binary Cross-Entropy (BCE), improving the model's ability to strike a compromise between segmentation consistency and pixel-level precision. To ensure efficiency and generalizability, the model training is optimized using the Adam optimizer in conjunction with a learning rate scheduler. Additionally, an early halting mechanism is incorporated to minimize overfitting. Using a standardized dataset, performance evaluation is carried out with 70%, 15%, and 15% splits for training, validation, and testing, respectively. To evaluate the quality of segmentation, important metrics are employed, such as the Dice Coefficient and the Intersection over Union (IoU). The accuracy of the model in identifying tumor boundaries is illustrated by visualizations of the anticipated segmentation results. The potential of combining U-Net with cutting-edge feature extractors such as EfficientNet-B7 to produce dependable, effective, and clinically significant segmentation results in medical imaging is demonstrated by this approach. © 2025 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Comparing the Pre-Trained Models with LSTM Combination to Improve Image Captioning Technology 3

Comparing the Pre-Trained Models with LSTM Combination to Im...

引用

3rd International Conference on Integrated Circuits and Communication Systems, ICICACS 2025

作者： Mukil, L.D. Singh, Tripty Sharma, Mansi Amrita School of Computing Amrita Vishwa Vidyapeetham Department of Computer Science and Engineering Bengaluru India Graphic Era University Department of Computer Science and Engineering India

ISBN: (纸本)9798331508456

Image captioning, integrating computer vision and natural language processing, has become one of the critical areas of work for better accessibility and advancing technology. In this study, the performance of pre-trained convolutional neural networks-DenseNet, VGG-16, ResNet, and Xception-has been evaluated when coupled with LSTM units for generating captions on Flickr8k. These models have been evaluated on the basis of BLEU and ROUGE scores. The best scores were produced by ResNet-LSTM, which indicates the model's stronger ability to produce captions that are accurate and contextual. VGG-16 with LSTM ranked second;it was almost on par. Xception with LSTM gave results that were moderate. DenseNet with LSTM had the worst score, indicating that the feature extraction mechanism of this model is weak. The following results provide evidence for the practical use of ResNet-LSTM to facilitate access: from assistive technology for visually impaired users to automatically generated content. The comparative study would further indicate which one has more merits or drawbacks compared with others and thereby enable a more informed selection of the models suitable for each particular application. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Enhancing Blood Platelet Counting through Deep Learning Models for Advanced Diagnostics

引用

SN computer science 2025年第1期6卷 1-10页

作者： Dev, Utkarsh Singh, Tripty Babu, Tina Mandal, Ashish Kumar Sharma, Mansi Mandal, Adhirath Department of Computer Science and Engineering Amrita School of Computing Amrita Viswa Vidyapeetham Karnataka Bengaluru 560035 India Department of Computer Science and Engineering Alliance School of Advanced Computing Alliance University Karnataka Bengaluru 562106 India Amrita Hospital Faridabad India Graphic Era Deemed to be University Uttarakkand Dehradun India

Platelet counting is considered an essential factor for the diagnosis of blood clotting disorders as well as other illnesses as COVID-19 and Leukemia. Hemocytometer-based manual techniques are labor-intensive and prone to errors, especially in dynamic settings where platelet clumping develops over time. This work presents a novel method based on a collection of regression and object detection models- MobileNetv2, EfficientDetD0, and SSD MobileNet-that were trained on a large number of microscopic blood smear images. Multiple models have been trained and analysed to find out which models perform well for our dataset and problem. This study demonstrates how deep learning and object detection in blood platelet counting can revolutionize the field by increasing diagnostic precision and efficiency in healthcare systems. Out of all the models trained, we observed that EfficientDet D0 gave us the best performance, with a precision of 0.64, mAP of 0.96 and recall of 0.65. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2024.

关键词： Blood cells EfficientDet MobileNet Platelets SSD SSD MobileNet

来源：评论

学校读者我要写书评

暂无评论

Intelligent Energy Management Strategies for Renewable Energy Sources 1

Intelligent Energy Management Strategies for Renewable Energ...

引用

1st International Conference on AIML-Applications for engineering and Technology, ICAET 2025

作者： Patil, Sangita Deotale, Disha Mali, Ramesh MIT School of Computing MIT ADT University Department of Electrical & Electronics Engineering Pune412201 India MIT School of Computing MIT ADT University Department of Computer Science & Engineering Pune412201 India

ISBN: (纸本)9798350355611

There have been several developments in renewable resources, standby sources of energy, and storage technologies. Because renewable sources are inconsistent, the best method to ensure supply continuity is to combine them with traditional energy to serve mini-grids or the isolated loads that are not linked to the main grid. In hybrid power system, many different energy sources are blended. Every time, effective and efficient Energy Management Strategies (EMS) are necessary when more than a source of energy is employed to supply a given load. This strategy directs the energy flow via the supply chain. This is necessary not just for standalone hybrid systems, but for hybrid power system that are linked to the main network. The intelligent control of hybrid MG has been the subject of a large number of serious research projects. A comprehensive assessment of the approaches used by a number of writers in their research work on applying energy management approaches is conducted. Standalone hybrid energy systems and the hybrid systems those are linked to the electrical grids are examples of these solutions. The aim of this paper is to give summary of various energy management schemes utilized in sustainable hybrid systems. Brief study is conducted for a number of freestanding and grid-linked hybrid system designs in order to make sure and find that which energy management technique that works well and will be applied for one renewable source is not always the best for another. © 2025 IEEE.

关键词： Hybrid power

来源：评论

学校读者我要写书评

暂无评论

An Empirical Comparative Study on Pruning and Quantization Algorithms for Model Compression

An Empirical Comparative Study on Pruning and Quantization A...

引用

2nd International Conference on Machine Learning and Autonomous Systems, ICMLAS 2025

作者： Mukesh, Narravula Kumar, Mallisetty Rathna Vikas, Ambati Sai Jeyakumar, Gurusamy Amrita School of Computing Coimbatore Amrita Vishwa Vidyapeetham Department of Computer Science and Engineering India

ISBN: (纸本)9798331505745

In a world where Neural networks and Deep learning have become of major importance, the need for methods to reduce their memory storage and usage has become of major importance. Hence the need for quantization and pruning has become huge and important. Also, the hardware used also decides the necessity of memory reduction due to the hardware's compatibility and memory computation constraints. This study compares different pruning and quantization models and reports about which are the best methods among the different methods considered in this study. The proposed comparison is carried out using the metrics of prediction accuracy and the time taken for training. This study identifies that magnitude-based pruning is the best among all the CPU-based systems with an accuracy of 0.998, whereas structured pruning has the highest accuracy of 0.985. The focus is on the hardware used, which are mainly the GPU-based systems and the CPU-based systems. The study reveals that the pruning method is better in the faceoff between pruning and quantization methods in terms of accuracy and computational time. When compared between the CPU and the GPU, concerning accuracy CPU holds the edge, whereas concerning the computational time the GPU holds the edge. Overall, concerning the balance between accuracy and computational time, the GPU-based systems hold the upper hand, making them more suitable for the optimization task. © 2025 IEEE.

关键词： Central Processing Unit Graphical Processing Unit Neural Networks Pruning Quantization

来源：评论

学校读者我要写书评

暂无评论

Digital Guardians: Exploring Video Analytics in Autonomous Classroom Surveillance Systems 2

Digital Guardians: Exploring Video Analytics in Autonomous C...

引用

2nd International Conference on Machine Learning and Autonomous Systems, ICMLAS 2025

作者： Sreekar, K. Jeyakumar, G. Amrita School of Computing Amrita Vishwa Vidyapeetham Department of Computer Science and Engineering Coimbatore India

ISBN: (纸本)9798331505745

The growing prevalence of intelligent surveillance systems has expedited progress in video analytics, particularly in learning environments. This survey article analyzes various methods for assessing student behavior and extracting features from classroom videos, with a focus on identifying test malpractices. We look at many existing systems that use techniques like facial identification, tracking, gesture recognition, and engagement analysis to monitor student activities. Here we shed light on the research needs by showing the benefits and limitations of several strategies that involve coping with occlusions, variable lighting, and small face sections in video frames. Our goal is to develop a strong system that uses live video feeds to detect test malpractices in a better and accurate manner while being efficient. © 2025 IEEE.

关键词： Gesture recognition

来源：评论

学校读者我要写书评

暂无评论

Image Compression Algorithms - Comparative Analysis 3

Image Compression Algorithms - Comparative Analysis

引用

3rd International Conference on Integrated Circuits and Communication Systems, ICICACS 2025

作者： Niharika Reddy, G. Namrutha, M. Akshaya, Namratha Bhaskaran, Sreebha Amrita School of Computing Amrita Vishwa Vidyapeetham Department of Computer Science and Engineering Bengaluru India

ISBN: (纸本)9798331508456

Due to the abundance of the new digital media data, the issue of image quality and volume of data requiring compression has become a significant factor of concern, especially in media storage and transmitting. This work affords a comparative analysis of different image compression techniques with focus on the compression ratio, quality preservation, and complexity. A new hybrid model of Predictive Coding, Run Length Coding and Quantum Entropy Coding (QEC) is proposed and shown to exhibit negligible quality loss with substantial space saving. The experimental outcomes show that the proposed method reduces space 80 percent and works better than previous methods for areas requiring high speed and relative accuracy. These insights are timely, as practical computing-communication trade-offs are paramount in the new generation of social networks, medicine, and multimedia streaming. © 2025 IEEE.

关键词： Image compression

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：