检索结果-内蒙古大学图书馆

Specificity-Guided Cross-Modal Feature Reconstruction for RGB-Infrared Object detection

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS 2025年第1期26卷 950-961页

作者： Sun, Xiaoyu Zhu, Yaohui Huang, Hua Beijing Inst Technol Sch Comp Beijing 100081 Peoples R China Beijing Univ Chem Technol Sch Informat Sci & Technol Beijing 100029 Peoples R China Beijing Normal Univ Sch Artificial Intelligence Beijing 100875 Peoples R China

RGB-Infrared object detection is an essential technology for the intelligent transportation system. Existing most works on RGB-Infrared object detection focus on how to fuse RGB and infrared features. However, these works overlook the inherent differences between RGB and infrared modalities, leading to insufficient modal feature fusion and limiting the performance of RGB-Infrared object detection. To address the above issues, a Specificity-guided Cross-modal Feature Reconstruction(SCFR) algorithm is proposed to establish modality-specific correlation for RGB-Infrared object detection. Specifically, the proposed SCFR involves the modality-specific cross-modal feature reconstruction network and two modality-specific losses. The modality-specific cross-modal feature reconstruction network performs cross-modal feature reconstruction on RGB and infrared modalities to establish modality-specific correlation. The modality-specific losses guide the direction of feature learning for reconstructing the expressive modality-specific features. These specific features can be used to achieve more efficient feature fusion, thus improving object detection performance. Comprehensive experimental results on three RGB-Infrared detection datasets demonstrate the effectiveness and the superiority of the proposed method. Our code will be available at https://***/SXYSUOSUO/***.

关键词： Feature extraction Object detection Image reconstruction Fuses Lighting Correlation Proposals detection algorithms Visualization Transportation RGB-infrared object detection feature reconstruction cross-modal learning multi-modal fusion

来源：评论

学校读者我要写书评

暂无评论

Multi-Hop Cooperative G-SFIM System With Low-Complexity detection Under Imperfect CSI

引用

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY 2025年第3期74卷 4177-4191页

作者： Pang, Ke Bai, Zhiquan Yuan, Dongfeng Zhang, Haijun Kwak, Kyung Sup Changzhou Univ Sch Microelect & Control Engn Changzhou 213164 Peoples R China Shandong Univ Sch Informat Sci & Engn Shandong Key Lab Intelligent Commun & Sensing Comp Qingdao 266237 Peoples R China Univ Sci & Technol Beijing Beijing Engn & Technol Res Ctr Convergence Network Beijing 100083 Peoples R China Inha Univ Dept Informat & Commun Engn Incheon 22212 South Korea

This paper proposes a multi-hop cooperative grouping space-frequency index modulation (G-SFIM) system with decode-and-forward (DF) protocol. By using group mapping and 3D constellation, the minimum Euclidean distance of the G-SFIM signals is increased. Meanwhile, the signals are conveyed through the multi-hop architecture to improve the communication reliability. Under the imperfect and perfect channel state information (CSI), the closed-form expressions of the average bit error probability (ABEP) are derived. To reduce the complexity of the optimal maximum-likelihood (ML) detection, the minimum mean-squared error (MMSE) based grouping ML (MMSE-GML), the MMSE based grouping greedy detection (MMSE-GGD), and the grouping ordered block based ML (GOB-ML) detection algorithms are proposed. Numerical results validate the correctness of the theoretical ABEP of the multi-hop cooperative G-SFIM system under different CSI conditions. Meanwhile, the proposed system achieves better ABEP performance and is less sensitive to the imperfect CSI. Additionally, the three presented low-complexity detection algorithms offer the flexible compromise between the ABEP performance and the complexity.

关键词： Complexity theory Three-dimensional displays Relays Symbols detection algorithms Transmitting antennas Protocols Modulation Spectral efficiency OFDM Index modulation 3D constellation imperfect channel state information low-complexity detection

来源：评论

学校读者我要写书评

暂无评论

Robust Overlapping Community detection in Complex Networks With Graph Convolutional Networks and Fuzzy C-Means

引用

IEEE ACCESS 2024年 12卷 70129-70145页

作者： Al-andoli, Mohammed Nasser Irianto AlSayaydeh, Jamil Abedalrahim Alwayle, Ibrahim M. Mohd, Che Ku Nuraini Che Ku Abuhoureyah, Fahd Univ Tekn Malaysia Melaka Fac Informat & Commun Technol Dept Media Interact Pervas Comp & Educ Technol Durian Tunggal 76100 Melaka Malaysia Rabdan Acad Fac Resilience Dept Gen Educ Abu Dhabi U Arab Emirates Univ Tekn Malaysia Melaka UTeM Fak Teknol & Kejuruteraan Elekt & Komputer FTKEK Dept Engn Technol Durian Tunggal 76100 Melaka Malaysia Najran Univ Coll Sci & Arts Sharurah Dept Comp Sci Najran 61441 Saudi Arabia Univ Tekn Malaysia Melaka UTeM Fak Kejuruteraan Elekt & Kejuruteraan Komputer FKE Durian Tunggal 76100 Melaka Malaysia

Community detection is an important task in complex network analysis. A community is a set of cohesive vertices that have more connections within the set than outside. In many real Complex Networks (CNs), these communities naturally overlap, meaning an individual node can belong to more than one community. This overlapping structure is crucial for many real applications, such as social influence detection, cyberattack detection, and recommendation systems. Existing methods often struggle to capture both network topology and node features, leading to suboptimal overlapping community detection. In this paper, we propose an efficient method called GCNFCM, which utilizes Graph Convolutional Networks (GCNs), Fuzzy C-means (FCM), and the modularity Q algorithm for overlapping community detection. The key idea is to achieve robust feature learning for nodes and then identify the best structure for overlapping community detection. GCNFCM extracts node embeddings from CNs, considering both topology and attributes through a dual-decoder design (inner product and GCN), while FCM is employed for optimal overlapping community detection. Furthermore, FCM is guided by the modularity Q algorithm for accurate community identification without requiring prior knowledge of the community count. Experimental results on ten real-world CNs of varying sizes demonstrate that our proposed method outperforms other state-of-the-art overlapping community detection methods in terms of producing cohesive communities and identifying ground-truth communities. Additionally, the results indicate that the developed method effectively identifies good overlapping communities in real-world networks.

关键词： Image edge detection Complex networks Feature extraction Representation learning Decoding Optimization Games Graph neural networks Convolutional neural networks Fuzzy systems detection algorithms Graph convolutional networks fuzzy c-means complex networks overlapping structure community detection

来源：评论

学校读者我要写书评

暂无评论

Forensic detection of Timestamp Manipulation for Digital Forensic Investigation

引用

IEEE ACCESS 2024年 12卷 72544-72565页

作者： Oh, Junghoon Lee, Sangjin Hwang, Hyunuk Korea Univ Sch Cybersecur Seoul 02841 South Korea Affiliated Inst ETRI Daejeon 34044 South Korea

File system forensics is one of the most important areas of digital forensic investigations. To date, various file system forensic methods have been studied, of which anti-forensic countermeasures include deleted file recovery, metadata recovery, and metadata manipulation detection. In particular, manipulation detection of timestamps, which are important file metadata, is one of the key techniques in digital forensic investigations. Existing detection methods for file timestamp manipulation in the New Technology File System (NTFS) have been studied based on various file system and operating system artifacts. This paper compares and analyzes the features and limitations of various existing detection methods and confirms that the NTFS journal-based detection method is the most effectively way to detect timestamp manipulation. However, previous NTFS journal-based detection methods have limitations such as incorrectly identifying normal events as manipulation or detecting manipulation only in limited cases. Therefore, we propose a new detection algorithm that can overcome these limitations. The proposed detection algorithm was implemented as a tool and verified through performance comparison experiments with existing detection methods. The results of experiment showed that the proposed detection algorithm has significantly improved performance by detecting timestamp manipulations that were not detected by previous detection methods and identifying normal events that were misidentified by existing detection methods. Finally, we introduce a case in which existing detection methods and the proposed detection algorithm are applied to malware that performs file timestamp manipulation in real-world advanced persistent threat attacks. The results of which confirm the superiority of the proposed detection algorithm.

关键词： File systems detection algorithms Metadata Digital forensics Operating systems Malware Tunneling Forensics File system forensics anti-forensic countermeasures timestamp manipulation forensic detection

来源：评论

学校读者我要写书评

暂无评论

A Novel Backdoor detection Approach Using Entropy-Based Measures

引用

IEEE ACCESS 2024年 12卷 114057-114072页

作者： Surendrababu, Hema Karnam Nagaraj, Nithin Univ Transdisciplinary Hlth Sci & Technol Bengaluru 560012 Karnataka India Natl Inst Adv Studies Indian Inst Sci Campus Sch Conflict & Secur Studies Bengaluru 560012 Karnataka India Natl Inst Adv Studies Indian Inst Sci Campus Sch Humanities Consciousness Studies Programme Bengaluru 560012 Karnataka India

Amidst the recent technological breakthroughs and increased integration of Artificial Intelligence (AI) technologies across various domains, it is imperative to consider the myriad security threats posed by AI. One of the significant attack vectors on AI models is the backdoor attack, which involves maliciously manipulating the model's behaviour by inserting hidden patterns or triggers into training datasets. In this paper our primary focus is on the defenses for the backdoor attacks mounted via poisoned training datasets. While many backdoor defense mechanisms have been proposed in the context of text, image, and audio domains, a majority of these defense mechanisms focus on training a specific model to detect backdoor triggers. Our current work proposes a novel model agnostic backdoor detection approach that utilizes complexity/entropy-based measures. In this study, we demonstrate the limitations of currently existing entropy measures - Sample Entropy and Approximate Entropy in detecting backdoor triggers in poisoned datasets. Consequently, we propose a novel modification of the Manhattan metric in the Entropy calculation and incorporate it in the complexity measures. This modified approach is shown to successfully detect backdoor triggers in datasets from not only the Natural Language Processing (NLP) domain, but also from the Financial and Geological domains. The effectiveness of the proposed approach was further substantiated with the high F1 scores in the range of 0.92 to 1.00 across the datasets, and with zero false negatives for the real-world datasets from the Financial and the Geological domains.

关键词： Entropy Complexity theory Training Data models Computational modeling Vectors Time series analysis Artificial intelligence detection algorithms Data integrity Data poisoning backdoor attacks backdoor defenses approximate entropy sample entropy

来源：评论

学校读者我要写书评

暂无评论

detection Approach for Station Area Topology Change Based on Multi-Time Scale

引用

IEEE ACCESS 2024年 12卷 125079-125094页

作者： Hao, Fangzhou Wang, Shaohua Guangdong Power Grid Co Ltd Guangzhou Power Supply Bur Guangzhou 510620 Guangdong Peoples R China

The detection of station area topology change is a critical part of guaranteeing the accuracy of a data-driven station area topology identification algorithm, for which a detection method for station area topology change based on the multi-time scale is proposed. The detection method comprehensively constructs the topology change detection indicator mechanism with multi-source data such as voltage and current and improves the robustness of the algorithm to the error of single voltage and current data. Firstly, the transient moment comparison indicator, the adjacent moment comparison indicator, and the adjacent sliding window comparison indicator with short to long-term scales are constructed. Subsequently, a three-level station area topology change tracking and detection strategy based on multi-time scale indicators is proposed for the identification of three types of topology changes: user migration into the area, user migration out of the area, and user phase adjustment. Finally, a simulation model is built using actual user data to verify the effectiveness of the proposed algorithm, explore the performance of the proposed algorithm in different scenarios, and compare it with existing recognition algorithms. The case study shows that the proposed algorithm is effective in recognizing topology change categories and change locations compared with existing algorithms, and is robust to data measurement errors.

关键词： Topology Network topology Transformers Correlation Low voltage Indexes detection algorithms Low-voltage station area low-voltage topology identification station area topology change identification multi-time scale indicator

来源：评论

学校读者我要写书评

暂无评论

An Adversarial Attack on Artificial Intelligence Malware detection in Consumer Internet of Things

引用

IEEE CONSUMER ELECTRONICS MAGAZINE 2025年第3期14卷 53-66页

作者： Yang, Chi-Hsin Mwangi, Maina Bernard Cheng, Shin-Ming Lee, Hahn-Ming Natl Taiwan Univ Sci & Technol Comp Sci & Informat Engn Taipei Taiwan Natl Taiwan Univ Sci & Technol Dept Comp Sci & Informat Engn Taipei Taiwan

The proliferation of the Internet of Things (IoT) and artificial intelligence (AI) has transformed traditional consumer electronics (CEs) into next-generation devices with enhanced intelligence and connectivity. However, this advancement has exposed CEs to cybersecurity threats, such as IoT botnets. Consequently, researchers are employing AI for proactive threat detection and prevention. Unfortunately, AI algorithms are vulnerable to adversarial attacks, necessitating robustness studies, as evaded malware can cause significant damage to already susceptible IoT CEs. This article presents a case study to evaluate the resilience of AI-based IoT malware detection systems against adversarial attacks. Specifically, our method involves inserting crafted binary code snippets (payloads) into the empty regions of malware executables. We leverage explainable AI techniques to guide payload generation, coupled with an optimization procedure to efficiently identify optimal payload sequences. Our method, tested on real-world IoT datasets, yields a robust hybrid detection system with a detection rate of up to $99.11\%.$ Our attack approach achieves evasion rates of up to 100% and generates transferable adversarial examples. The generated samples evade a prominent structural IoT malware detector with an evasion rate of 95.15% at a minimal attack cost. This study underscores the importance of enhancing the robustness of AI-based malware detection systems and implementing diverse strategies to safeguard consumer IoT devices.

关键词： Malware Feature extraction Internet of Things Payloads Consumer electronics Codes Semantics Explainable AI Training detection algorithms

来源：评论

学校读者我要写书评

暂无评论

Low-Complexity Sparse Compensation MRC detection Algorithm for OTSM Systems

引用

IEEE COMMUNICATIONS LETTERS 2025年第2期29卷 398-402页

作者： Hao, Shiya Li, Hua Li, Qianqian Feng, Jiaqi Dai, Xiaoming Univ Sci & Technol Beijing Sch Comp & Commun Engn Beijing 100083 Peoples R China

Orthogonal time sequency multiplexing (OTSM) has recently attracted significant attention due to its enhanced robustness in doubly selective fading channels. However, the computational complexity of conventional OTSM receivers increases significantly in scenarios with fractional Doppler shifts. In this work, we first analyze the effects of inter-Doppler interference on the performance of OTSM systems. We then propose a sparse compensation maximum ratio combining (SC-MRC) detection algorithm to mitigate computational complexity. Specifically, a channel sparsification procedure is implemented by eliminating amplitude values that fall below a predetermined threshold, thereby substantially reducing the computational complexity associated with matrix multiplication. A low-complexity adaptive compensation scheme is introduced to mitigate the associated performance degradation. Simulation results demonstrate that the proposed SC-MRC algorithm achieves superior performance compared to the conventional MRC algorithm while offering reduced computational complexity.

关键词： Symbols Vectors Delays detection algorithms Doppler shift Detectors Interference Computational complexity Time-domain analysis Standards OTSM MRC sparse compensation

来源：评论

学校读者我要写书评

暂无评论

An Edge Computing-Based Factor-Aware Novel Framework for Early detection and Classification of Melanoma Disease Through a Customized VGG16 Architecture With Privacy Preservation and Real-Time Analysis

引用

IEEE ACCESS 2024年 12卷 113580-113596页

作者： Almufareh, Maram Fahaad Jouf Univ Coll Comp & Informat Sci Dept Informat Syst Al Jouf 72388 Saudi Arabia

Melanoma is dangerous skin cancer disease with high malignancy potential, necessitates advanced detection methods for improved patient outcomes. This study proposes a novel factor-aware, edge computing-based framework, leveraging a customized VGG16 architecture. The framework prioritizes real-time analysis and privacy preservation, acknowledging the diverse risk factors influencing melanoma development. This work has led to the development of a factor-aware approach, a customized VGG16 architecture for the classification of melanoma, and an edge device privacy-centric system. Tests contrasting traditional and customized VGG16 architectures demonstrate the enhanced effectiveness of the customized architecture, attaining a balanced categorization of both benign and malignant instances. The proposed system works well for processing skin lesion images at the data source, cutting down on latency and protecting patient privacy. Experiments conducted in comparison with traditional VGG16 architecture show increased F1-scores, recall, and precision in the classification of benign and malignant cases. The customized VGG16 architecture achieves an accuracy boost from 0.84 to 0.88, outperforming the competition consistently. Confusion matrices show a decrease in false positives and negatives, which suggests improved diagnostic precision. Improved discriminative power is demonstrated by the Receiver Operating Characteristic (ROC) curves, which indicate a higher area under the curve (AUC) for the customized VGG16 architecture (0.95). In comparison to traditional architecture (AP =0.58), precision-recall curves show a higher Average Precision (AP) score of 0.94, demonstrating a stronger balance between recall and precision. The proposed framework offers measurable gains in the accuracy of melanoma diagnosis in addition to introducing novel approaches. These findings highlight the possibility of transforming early detection and treatment, which would eventually enhance patient outc

关键词： Melanoma Skin Cancer Computer architecture Lesions Edge computing Computational modeling Artificial intelligence detection algorithms Classification algorithms VGG16 architecture melanoma edge computing detection and classification

来源：评论

学校读者我要写书评

暂无评论

Histopathological Cancer detection Using Intra-Domain Transfer Learning and Ensemble Learning

引用

IEEE ACCESS 2024年 12卷 1434-1457页

作者： Yong, Ming Ping Hum, Yan Chai Lai, Khin Wee Lee, Ying Loong Goh, Choon-Hian Yap, Wun-She Tee, Yee Kai Univ Tunku Abdul Rahman Lee Kong Chian Fac Engn & Sci Kajang 43000 Malaysia Univ Malaya Fac Engn Dept Biomed Engn Kuala Lumpur 50603 Malaysia

Cancer is a leading global cause of death. Histopathology image analysis is widely recognized as the gold standard for cancer diagnosis, playing a crucial role in early detection and reducing mortality rates. However, this diagnostic task is performed manually by pathologists, leading to high human errors and variabilities due to the huge number of images to screen and tissue complexities. With the emergence of deep learning, specifically the Convolutional Neural Network (CNN), in computer-aided diagnosis, there is a growing interest from the medical community to automate the labor-intensive manual image screening. Despite its promising performance, deep learning models still encounter challenges when it comes to extracting comprehensive histopathological features for optimal results. To tackle this issue, our study introduces deep learning models based on intra-domain transfer learning and ensemble learning. We evaluated these models on public histopathology datasets, including Gastric Histopathology Sub-size Image Database (GasHisSDB), Chaoyang colorectal, and Clinical Proteomic Tumor Analysis Consortium Clear Cell Renal Cell Carcinoma (CPTAC-CCRCC). Our models achieved state-of-the-art accuracy: 99.78% on GasHisSDB, 85.69% on Chaoyang, and 99.17% on CPTAC-CCRCC. These results highlight our models' ability to extract rich features and perform well on low-resolution histopathology images. Thus, our models have the potential to assist pathologists, reduce their workload, and enhance patient survival rates.

关键词： Histopathology Cancer Transfer learning Feature extraction Deep learning Task analysis Ensemble learning detection algorithms Convolutional neural networks Cancer detection convolutional neural network deep learning ensemble model histopathology transfer 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：