检索结果-内蒙古大学图书馆

Enhancing Fraud detection in Banking With Deep Learning: Graph Neural Networks and Autoencoders for Real-Time Credit Card Fraud Prevention

引用

IEEE ACCESS 2025年 13卷 20633-20646页

作者： Alarfaj, Fawaz Khaled Shahzadi, Shabnam KFU Sch Business Dept Management Informat Syst MIS Al Hasa 31982 Saudi Arabia Anhui Univ Sci & Technol Dept Math & Big Data Huainan 232001 Peoples R China

Under the umbrella of artificial intelligence (AI), deep learning enables systems to cluster data and provide incredibly accurate results. This study explores deep learning for fraud detection, utilizing Graph Neural Networks (GNNs) and Autoencoders to enhance business practices and reduce fraudulent activities in large organizations. For real-time fraud detection, we propose Graph neural network with lambda architecture while for credit card fraud detection, we use an autoencoder, validated through case studies from two banks. The findings demonstrate that these methods effectively detect fraud with balance of precision and recall, improving the efficiency of banking systems. Python is employed for analysis, emphasizing the ability of deep learning to manage and prevent fraud in real-time on dynamic datasets. In the end, this study concludes that by using deep learning algorithms, we can control online credit card fraud detection in banks, improve the efficiency of the banking system. We can manage fraudulent activity in real-time and on dynamic datasets by utilizing deep learning algorithms, which allows for ongoing improvement of the fraud detection and prevention system.

关键词： Fraud Deep learning Banking Credit cards Business intelligence Real-time systems detection algorithms credit card fraud detection graph neural network autoencoders autoencoders

来源：评论

学校读者我要写书评

暂无评论

Community detection via Autoencoder-Like Nonnegative Tensor Decomposition

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2024年第3期35卷 4179-4191页

作者： Guan, Jiewen Chen, Bilian Huang, Xin Xiamen Univ Dept Automat Xiamen 361005 Peoples R China Xiamen Key Lab Big Data Intelligent Anal & Decis Xiamen 361005 Peoples R China Hong Kong Baptist Univ Dept Comp Sci Hong Kong Peoples R China

Community detection aims at partitioning a network into several densely connected subgraphs. Recently, nonnegative matrix factorization (NMF) has been widely adopted in many successful community detection applications. However, most existing NMF-based community detection algorithms neglect the multihop network topology and the extreme sparsity of adjacency matrices. To resolve them, we propose a novel conception of adjacency tensor, which extends adjacency matrix to multihop cases. Then, we develop a novel tensor Tucker decomposition-based community detection method-autoencoder-like nonnegative tensor decomposition (ANTD), leveraging the constructed adjacency tensor. Distinct from simply applying tensor decomposition on the constructed adjacency tensor, which only works as a decoder, ANTD also introduces an encoder component to constitute an autoencoder-like architecture, which can further enhance the quality of the detected communities. We also develop an efficient alternative updating algorithm with convergence guarantee to optimize ANTD, and theoretically analyze the algorithm complexity. Moreover, we also study a graph regularized variant of ANTD. Extensive experiments on real-world benchmark networks by comparing 27 state-of-the-art methods, validate the effectiveness, efficiency, and robustness of our proposed methods.

关键词： Tensors Optimization detection algorithms Matrix decomposition Decoding Spread spectrum communication Complex networks Community detection graph clustering nonnegative tensor decomposition optimization

来源：评论

学校读者我要写书评

暂无评论

Keypoint detection Based on Curvature Grouping and Adaptive Sampling

引用

IEEE ACCESS 2025年 13卷 42272-42284页

作者： Li, Bifu Cheng, Yu Li, Weitong Guangdong Univ Technol Sch Informat Engn Guangzhou 510006 Peoples R China

In the keypoint detection algorithm, the farthest point sampling methods and random sampling methods are usually used to select candidate points, then keypoints are screened out from the neighborhood of the candidate points. In the methods, reproducibility and interpretability of keypoints are relatively poor. To address the problems, this paper introduces point cloud geometric features to build a keypoint detection and description network that makes the keypoints more stable and their geometric properties easier to distinguish. The details of point cloud geometric features include: 1) Grouping the point cloud using curvature, Candidate points and their dilation clusters are found through adaptive random sampling and neighborhood expansion;2) Aggregating the dilation cluster features by using the channel-position attention mechanism to generate keypoints based on the scores;3) Jointly generating the descriptors using the dilation cluster features and the attention feature maps. Experiments on ModelNet40 and KITTI datasets show that the method in this paper improves the Repeatability by more than 8% compared to the existing methods, and the improvement is more obvious when the number of keypoints is reduced. Additionally, the inlier rate is enhanced by 8% and 3% on the two datasets, respectively.

关键词： Point cloud compression Feature extraction Three-dimensional displays Accuracy Sampling methods Real-time systems detection algorithms Adaptive systems Point cloud keypoint detector descriptor curvature grouping adaptive sampling

来源：评论

学校读者我要写书评

暂无评论

FCDNet: A Lightweight Network for Real-Time Wildfire Core detection in Drone Thermal Imaging

引用

IEEE ACCESS 2025年 13卷 14516-14530页

作者： Wang, Linfeng Doukhi, Oualid Lee, Deok Jin Jeonbuk Natl Univ Ctr Autonomous Intelligence & E Mobil Dept Mech Design Engn Jeonju Si 54896 South Korea

The increasing number of wildfires damage nature and human life, making the early detection of wildfires in complex outdoor environments critical. With the advancement of drones and remote sensing technology, infrared cameras have become essential for wildfire detection. However, as the demand for higher accuracy in detection algorithms grows, the detection model's size and computational costs increase, making it challenging to deploy high-precision detection algorithms on edge computing devices onboard drones for real-time fire detection. This paper introduces a novel infrared wildfire detection network named FCDNet to tackle this issue. It includes an Efficient Processing (EP) module based on the novel Partial Depthwise Convolution (PDWConv) and the lightweight feature-sharing decoupled detection head (Fast Head), achieving low-size and low-computation wildfire detection. An Adaptive Sample Attention (ASA) Loss is introduced to enhance the detection accuracy of wildfire cores in combination with Normalized Wasserstein Distance (NWD) Loss. The experiment shows that the model size of FCDNet is only 4.0MB, representing 63.5% of the baseline YOLOv8n network, with 63.3% of its parameters. It operates at just 5 Giga Floating Point Operations Per Second (GFLOPs), 38.3% lower, and achieves a 77.5% mAP (@50-95 IOU), a 1% increase, with a 460x460 input image size. Compared to the state-of-the-art YOLOv11n, FCDNet reduces parameters, computation, and model size by 26.9%, 20.6%, and 27.3%, respectively. The thermal dataset and training codes used in this study are made publicly available at: https://***/WangLF1996/FCDNet-Dataset-and-Algorithm

关键词： Wildfires Accuracy Real-time systems Drones Head Computational modeling detection algorithms Cameras Feature extraction Autonomous aerial vehicles FCDNet infrared wildfire detection lightweight network YOLOv8n-based

来源：评论

学校读者我要写书评

暂无评论

Autonomous Aerial Vehicle Object detection Based on Spatial Perception and Multiscale Semantic and Detail Feature Fusion

引用

IEEE ACCESS 2025年 13卷 42897-42909页

作者： Rao, Wei Chen, Siyuan Li, Dan NanChang Inst Technol Nanchang 330099 Peoples R China Fudan Univ Res Ctr Smart Networks & Syst Sch Informat Sci & Technol Key Lab Informat Sci Electromagnet Waves Shanghai 200433 Peoples R China

Owing to changes in the spatial position of Autonomous Aerial Vehicle (AAV) aerial images and limited platform resources, most existing AAV aerial image detection models have low accuracy, and it is difficult to achieve a good balance between detection performance and lightweight. To solve the above problems, an object detection model of AAV aerial images based on YOLOv8s, called BSDS-YOLOv8s, is proposed. In the proposed model the Dynamic Head (DyHead) is used to replace the detection head of YOLOv8s firstly, which can improve the spatial perception ability of the detection head of the model. Second, to improve the detection performance of DyHead and make the model lightweight, a new feature pyramid network (SDI-MBiFPN) in the Neck of YOLOv8s is proposed, which contains a Multiscale Bidirectional Feature Pyramid Network (BiFPN), called MBiFPN, and a feature fusion method based on the redesigned Semantic and Detail Fusion (SDI). Finally, a method that integrates Soft Non-Maximum Suppression (Soft-NMS) with Generalized Intersection over Union (GIoU), called GIoU-Soft-NMS, is proposed to enhance the model's post-processing capability and reduce the missed and false detection rates, thereby further improving the detection accuracy of the model. The experimental results showed that the mAP0.5 and mAP0.5:0.95 of BSDS-YOLOv8s reached 49.6% and 34.1%, respectively, on the VisDrone2019 dataset. This is an improvement of 9.2 and 9.8 percentage points over YOLOv8s, respectively. The number of parameters and the floating point operations (FLOPs) were 8.03M and 26.6G, 27.8% and 7.6% less than those of YOLOv8s, respectively.

关键词： Feature extraction Autonomous aerial vehicles Computational modeling Accuracy Head Neck detection algorithms Convolution YOLO Semantics YOLOv8s Autonomous Aerial Vehicles (AAVs) MBiFPN SDI DyHead GIoU Soft-NMS

来源：评论

学校读者我要写书评

暂无评论

Novel Chewing Cycle Approach for Peak detection Algorithm of Chew Count Estimation

引用

IEEE SENSORS JOURNAL 2025年第1期25卷 803-812页

作者： Asmiza Selamat, Nur Hamid Md Ali, Sawal Ghadafi Ismail, Ahmad Anom Ahmad, Siti Nisa' Minhad, Khairun Univ Kebangsaan Malaysia Dept Elect Elect & Syst Engn Bangi 43600 Selangor Malaysia Univ Tekn Malaysia Melaka Fak Teknol Kejuruteraan Elekt Durian Tunggal 76100 Melaka Malaysia Univ Kebangsaan Malaysia Inst Microengn & Nanoelect IMEN Bangi 43600 Selangor Malaysia Univ Putra Malaysia UPM Fac Engn Serdang 43400 Selangor Malaysia Xiamen Univ Malaysia Dept Elect & Elect Engn Sepang 43900 Selangor Malaysia

Chew count is a critical parameter in analyzing mastication signals, yet traditional methods of manual counting by trained clinicians are often labor-intensive and prone to errors. As a result, there has been a growing interest among researchers in developing automated methods for estimating chew count. This article reviews the existing approaches, evaluates their effectiveness, and proposes a new approach based on optimization technique. This work proposes a novel approach to chew count estimation using particle swarm optimization (PSO) combined with a peak detection algorithm. The chewing dataset comprises signals collected from 20 participants consuming eight different food types, with proximity sensors (PSs) detecting temporalis muscle activity. The peak detection algorithm identifies key signal features, while PSO optimizes the peak prominence and width parameters to minimize the mean absolute error (MAE) in chew count estimation. Two specific chewing cycle approaches were implemented: a participants-based (P) cycle and a participants-food type-based (PF) cycle. These approaches were compared to the traditional All (A) chewing cycle method, which evaluates chew count across the entire dataset in a single analysis. Results demonstrate that the PF method yields the lowest MAE at 1.25%, followed by the P method at 3.46%, and the A method at 4.26%. Moreover, the PF method required the least computational time at 8012.2 s, compared to 9392.0 s for the P method and 36621.4 s for the A.

关键词： Reviews Masticatory muscles Estimation Manuals Muscles Robustness Sensors detection algorithms Particle swarm optimization Optimization Chew count estimation chewing cycle chewing episode maximum peak width minimum peak prominence particle swarm optimization (PSO) peak detection

来源：评论

学校读者我要写书评

暂无评论

Distributed MIMO Radar Target detection in Multipath Clutter Environments With Time Reversal

引用

IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS 2025年第2期61卷 4907-4925页

作者： Lian, Hao Yang, Minglei Zhou, Dingsen Liu, Meng Xidian Univ Natl Key Lab Radar Signal Proc Xian 710071 Peoples R China

The distributed multiple-input-multiple-output (MIMO) radar, as an emerging active sensing technology, uniquely leverages spatial diversity to significantly enhance radar detection performance. The time reversal (TR) technology can be incorporated into radar systems to further improve detection performance by effectively utilizing multipath signals. However, the coexistence of multipath signals and clutter presents a formidable challenge to radar detection, potentially diminishing its performance. To tackle this challenge, we propose a distributed MIMO radar detection algorithm that integrates the TR technology with a power allocation (PA) approach tailored for multipath clutter environments. Specifically, the algorithm initially leverages TR technology to efficiently process and exploit multipath signals, while concurrently optimizing the distribution of radar transmit power through strategic allocation, thereby bolstering detection performance in complex environments. On this foundation, a distributed MIMO radar signal model is established, and the TR-PA-MIMO detector is derived, embodying the proposed solution. Furthermore, a conventional distributed MIMO (CD-MIMO) detector is developed as a benchmark for comparative analysis. Several other detection algorithms are also evaluated under identical conditions to assess their relative performance. The Monte Carlo simulations and theoretical analysis demonstrate that the proposed detection algorithm outperforms all other algorithms for distributed MIMO radar systems operating in multipath clutter environments.

关键词： MIMO radar Clutter Radar Detectors Transmitting antennas Object detection Radar detection Radar antennas detection algorithms Resource management distributed MIMO radar target detection time reversal (TR)

来源：评论

学校读者我要写书评

暂无评论

Microwave Technologies in Experiments for detection of Dark Matter Axions

引用

IEEE MICROWAVE MAGAZINE 2025年第3期26卷 78-87页

作者： Navarro-Madrid, Jose R. Garcia-Barcelo, Jose Maria Diaz-Morcillo, Alejandro Univ Politecn Cartagena Dept Informat Technol & Commun Cartagena 30202 Spain Max Planck Inst Phys & Astrophys Werner Heisenberg Inst D-85748 Garching Germany

One of the major challenges of modern physics is the detection of dark matter and the study of its composition. For this purpose, there are several experiments where microwave technology is present. In this work, we describe experiments for the detection of dark matter axions and the most important design features that microwave devices must have for this application.

关键词： Microwave technology Superconducting materials Three-dimensional displays Dark matter Superconducting magnets Superconducting filters Superconducting device noise Microwave filters Superconducting microwave devices detection algorithms

来源：评论

学校读者我要写书评

暂无评论

Segment Anything Model and Fully Convolutional Data Description for Plant Multi-Disease detection on Field Images

引用

IEEE ACCESS 2024年 12卷 102592-102605页

作者： Moupojou, Emmanuel Retraint, Florent Tapamo, Hyppolite Nkenlifack, Marcellin Kacfah, Cheikh Tagne, Appolinaire Inst Univ St Jean Cameroun Dept Informat & Syst Informat Yaounde Cameroon Univ Technol Troyes Comp Sci & Digital Soc Lab LIST3N Troyes France Univ Yaounde I Comp Sci Dept Yaounde Cameroon Univ Dschang Math & Comp Sci Dept Dschang Cameroon Inst Rech Agr Dev Yaounde Cameroon

Researchers have designed various models trained on public or private datasets for plant disease detection to help farmers remedy crop yield losses on their farms due to plant diseases. Plantvillage is the most widely used plant disease dataset with laboratory images captured under controlled conditions with a single leaf on each image and a uniform background. Models trained on such datasets have extremely low classification accuracies when running on field images captured directly from plantations with various interwoven leaves, complex backgrounds, and different lighting conditions. In this study, we propose a model ensemble solution for the accurate identification and classification of plant diseases using field images. The model uses Segment Anything Model to efficiently circumscribe all identifiable objects in the image. Image Processing techniques are then used to isolate the identified objects from the original image. Background objects are separated from actual leaf objects using Fully Convolutional Data Description, which is an explainable deep one-class classification model for anomaly detection. Finally, the selected leaves are submitted to a Plantvillage-trained classification model for inference. Our model can detect diseases appearing on individual leaves of the same image and improves classification accuracy by more than 10% on public field plant disease datasets such as PlantDoc, thus providing a reliable solution for farmers and practitioners.

关键词： Plant diseases Accuracy Convolutional neural networks Training Task analysis Object recognition Plantations Plants (biology) detection algorithms Field plant images fully convolutional data description laboratory images plant disease dataset plant disease detection and classification segment anything model

来源：评论

学校读者我要写书评

暂无评论

Ensembling Object detection Models for Robust and Reliable Malaria Parasite detection in Thin Blood Smear Microscopic Images

引用

IEEE ACCESS 2024年 12卷 60747-60764页

作者： Ozbilge, Emre Guler, Emrah Ozbilge, Ebru Cyprus Int Univ Fac Engn Dept Comp Engn TR-99258 Nicosia North Cyprus Turkiye European Univ Lefke Fac Arts & Sci Dept Mol Biol & Genet TR-10 Mersin North Cyprus Turkiye Amer Univ Middle East Coll Business Adm Egaila 54200 Kuwait

Malaria is a blood disease caused by the Plasmodium parasite that is transmitted through the bite of female Anopheles mosquitoes. These mosquitoes can cross borders without passports or visas, making malaria a global health concern. To effectively treat malaria, infectious disease specialists must monitor the efficacy of the treatment by counting the number of parasites in a patient's blood at various time intervals. However, this task is challenging because it involves examining thin or thick blood smear samples under a microscope, which can be tiring to the human eye, particularly when there are many infected patients or when there is a shortage of clinical experts. In such cases, rapid diagnosis is crucial. One approach is to capture microscopic images of blood smear samples using a camera and then employ deep learning-based object detection models to detect and count the infected red blood cells. In this study, state-of-the-art object detection models, including CenterNet, EfficientDet, Faster R-CNN, RetinaNet, and YOLOv8, were explored. The dataset was generated using thin blood smear images in the laboratory. The results revealed that YOLOv8s outperformed the other models, achieving an score of 0.9031 and an mAP@[0.50:0.05:0.95] score of 0.5957. This study also found that various model combinations and ensemble strategies could improve the detection of malaria parasites. Specifically, the weighted boxes fusion ensembling approach achieved an score of 0.9186 and an mAP@[0.50:0.05:0.95] score of 0.6196. In contrast, the non-maximum weighted method achieved an score of 0.9324 and an mAP@[0.50:0.05:0.95] score of 0.6214.

关键词： Malaria Blood Microscopy Feature extraction Diseases Training Task analysis detection algorithms Parasitic diseases CenterNet computer vision deep learning EfficientDet ensemble learning faster R-CNN object detection plasmodium malaria RetinaNet YOLOv8

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：