检索结果-内蒙古大学图书馆

DBSSL: A Scheme to Detect Backdoor Attacks in Self-Supervised Learning Models

IEEE Transactions on Dependable and Secure Computing 2025年

作者： Huang, Yuxian Yang, Geng Yuan, Dong Yu, Shui Nanjing University of Posts and Telecommunication College of Computer Science and Software China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing China University of Sydney School of Electrical and Information Engineering Australia University of Technology Sydney School of Computer Science Australia

Recently, self-supervised learning has garnered significant attention for its ability to extract high-quality features from unlabeled data. However, existing research indicates that backdoor attacks can pose significant threats to self-supervised learning. Additionally, due to the substantial training overhead, traditional backdoor detection methods in supervised learning, such as meta-learning, are not well-suited for self-supervised learning. Current backdoor detection methods for self-supervised learning can only defend against backdoor attacks triggered by patch. To address this challenge, we proposes DBSSL, a scheme that can efficiently detect backdoor attacks in self-supervised learning models. In DBSSL, we connect backdoor attacks with adversarial attacks and use adversarial perturbations for detection. Specifically, we first conduct targeted adversarial attacks on the samples, classifying them into different classes. Then, we employ the Z-Score method to analyze these adversarial perturbations. If outliers are present, we can infer that the self-supervised learning model has been backdoored. Unlike previous works, our scheme is better suited for downstream users with constrained computation resources and exhibits excellent detection capabilities for backdoor attacks caused by patch -type or global triggers. Abundant theoretical analysis has demonstrated the feasibility of our scheme. Moreover, extensive experiments show that our method performs well in detecting prevalent backdoor attacks in self-supervised learning, achieving detection accuracy exceeding 95%. © 2004-2012 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

A Planar, Wideband and Compact Hadamard Matrix for 2-D Switched Beam Antenna Array

引用

IEEE Antennas and Wireless Propagation Letters 2025年

作者： Pan, Yufei Xu, Zizhen Yang, Ye Dai, Xin Yang, Wanjun Zheng, Shaoyong School of Electronics and Communication Engineering Guangzhou University Guangzhou China City University of Hong Kong Department of Electrical Engineering Hong Kong Sun Yat-sen University Guangdong Provincial Key Laboratory of Optoelectronic Information Processing Chips and Systems School of Electronics and Information Technology Guangzhou510006 China

This letter proposes a wideband two-dimensional (2-D) switched beam antenna array with single/dual/quad-lobed switched beams for direction of arrival (DOA) estimation. A planar and compact 4-order Hadamard matrix (HM) without any intersection structure works as its feeding network. The HM provides four groups of output signals with equal amplitudes and coded 0° /180° phase differences between adjacent outputs. It adopts four universal building blocks arranged planarly, each of them is realized by a 3-dB wideband quadrature coupler cascading with a 90 8 phase shifter. For demonstration, a HM prototype operating at 2.4 GHz is designed, fabricated, and measured. Its overall circuit size is 0.67 × 1.34 λg 2 and bandwidth is larger than 47.6%. Finally, a wideband 2-D switched beam antenna array is constructed by the proposed HM and magnetoelectric dipole antenna array. The 2-D single/dual/quadlobed switched beams verify its effectiveness. © 2002-2011 IEEE.

关键词： Hadamard matrices

来源：评论

学校读者我要写书评

暂无评论

Millimeter-Wave Co-Polarized Self-Decoupled Substrate-Integrated Dielectric Resonator Antennas for In-Band Full-Duplex and MIMO Systems

引用

IEEE Transactions on Antennas and Propagation 2025年

作者： Liu, Zheng Hu, Peng Fei Pan, Yong Mei Leung, Kwok Wa Zheng, Shao Yong Sun Yat-sen University School of Electronics and Information Technology School of Microelectronics Guangdong Provincial Key Laboratory of Optoelectronic Information Processing Chips and Systems Guangzhou510006 China South China University of Technology School of Electronic and Information Engineering Guangzhou510630 China Sun Yat-sen University School of Electronics and Information Technology Guangzhou510006 China City University of Hong Kong State Key Laboratory of Terahertz and Millimeter Waves Department of Electrical Engineering Kowloon Hong Kong CityU Shenzhen Research Institute Information and Communication Technology Centre Shenzhen518057 China

Millimeter-wave (mm-wave) co-polarized self-decoupled substrate-integrated dielectric resonator antennas (SIDRA) are investigated. A dual-port design is first studied which employs the loading effect of the feeding probe to modify the E-field distributions of the dielectric resonator (DR) (Formula presented) mode. The interaction between the probe-loaded and DR modes results in the formation of a quasi-fundamental mode (Formula presented) and a weak field region at the coupled port, leading to stable radiation patterns for each port with high port isolation. Additionally, metal vias is used to surround the DR to reduce cross-polarization. Also, the metal vias form electric walls which can reduce the DR size effectively. A hole is introduced at the center of the DR to further enhance the isolation. Based on the dual-port antenna element, a 2 × 3-port MIMO array with a center-to-center distance of 0.49λ0 is developed. To improve isolation between adjacent elements, a leaky wave decoupling structure and two strips are introduced, increasing the isolation by 11 dB. Both the antenna element and the MIMO array were fabricated to validate the designs. Experimental results indicate that the –10-dB impedance bandwidth and isolation level of the dual-port antenna element are 6% and 24.4 dB, respectively. For the 2 × 3-port MIMO array, the –10-dB bandwidth and isolation level are 5% and 20 dB, respectively. Stable radiation patterns with low cross-polarization levels are obtained for each port. Our designs feature a simple structure using a single substrate only, with the design methodology extendable to larger MIMO arrays. © 1963-2012 IEEE.

关键词： Millimeter waves

来源：评论

学校读者我要写书评

暂无评论

Partially Hidden Policy Attribute-Based Multi-keyword Searchable Encryption with Verification and Revocation

引用

IEEE Transactions on Mobile Computing 2025年

作者： Chen, Liqing Xu, Shiguo Jin, Chunhua Zhang, Hao Weng, Jian Huaiyin Institute of Technology Faculty of Computer and Software Engineering Jiangsu Huai'an223003 China Fujian Provincial Key Laboratory of Network Security and Cryptology Fujian Fuzhou350117 China Jinan University College of Information Science and Technology Guangdong Guangzhou510632 China

Data security is becoming increasingly critical as outsourced data services flourish. An effective solution for ensuring data confidentiality in the cloud is attribute-based searchable encryption (ABSE). However, most extant ABSE solutions have not been able to simultaneously achieve rich expressive queries, result validation, attribute revocation, and policy hiding, and there are problems such as inflexibility in query methods and access control capabilities, insufficient scalability and practicality, and privacy leakage. In particular, it should be noted that most ABSE schemes fail to promise the privacy security of access policies and are susceptible to offline keyword guessing attack. Hidden policy-ABSE (HP-ABSE) has been proposed to tackle the issue of explicit attribute values in access policies potentially compromising privacy. Nevertheless, in most existing HP-ABSE schemes, an adversary may initiate attribute values guessing attack to speculate on the values of attribute within access strategy. To this end, this paper proposes partially hidden policy attribute-based multi-keyword searchable encryption with verification and revocation (PHP-ABMKSE-VR). PHP-ABMKSE-VR scheme is practical, secure, as well as effective, making it appropriate for real-world scenarios like smart healthcare, as demonstrated by the experimental outcomes and safety analysis. © 2002-2012 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Brain-Conditional Multimodal Synthesis: A Survey and Taxonomy

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年第5期6卷 1080-1099页

作者： Mai, Weijian Zhang, Jian Fang, Pengfei Zhang, Zhijun South China University of Technology School of Automation Science and Engineering Guangzhou510640 China Southeast University School of Computer Science and Engineering Nanjing210096 China South China University of Technology Key Library of Autonomous Systems and Network Control Ministry of Education The School of Automation Science and Engineering Guangzhou510640 China Institute for Super Robotics Huangpu Guangzhou510555 China Nanchang University Jiangxi Thousand Talents Plan Nanchang330031 China Jishou University College of Computer Science and Engineering Jishou416000 China Guangdong Artificial Intelligence and Digital Economy Laboratory Pazhou Lab Guangzhou510335 China Shaanxi University of Technology Shaanxi Provincial Key Laboratory of Industrial Automation School of Mechanical Engineering Hanzhong723001 China Changsha Normal University School of Information Science and Engineering Changsha410100 China Guangdong University of Petrochemical Technology School of Automation Science and Engineering Institute of Artificial Intelligence and Automation Maoming525000 China

In the era of artificial intelligence generated content (AIGC), conditional multimodal synthesis technologies (e.g., text-to-image) are dynamically reshaping the natural content. Brain signals, serving as potential reflections of how the brain interprets external information, exhibit a distinctive one-to-many correspondence with various external modalities. This correspondence makes brain signals emerge as a promising guiding condition for multimodal synthesis (e.g., image, text, and audio), which is crucial for developing practical brain–computer interface systems and unraveling complex mechanisms underlying human perception. This survey comprehensively examines the emerging field of brain-conditional multimodal synthesis, termed AIGC-brain, to delineate the current landscape and future directions. To begin, related neuroimaging datasets and generative models are introduced as the foundation of AIGC-brain decoding and analysis. Next, we present a comprehensive taxonomy according to AIGC-brain methodologies, followed by task-specific representative work and implementation details to facilitate in-depth comparison and analysis. Quality assessments are then introduced for both qualitative and quantitative evaluation. Finally, this survey explores insights gained, outlining current challenges and prospects of AIGC-brain. As a pioneering survey, this article paves the way for future advances in AIGC-brain research. © 2020 IEEE.

关键词： Taxonomies

来源：评论

学校读者我要写书评

暂无评论

GITANet: Group Interactive Threshold-based Attention Network for Hyperspectral Image Classification

引用

IEEE Transactions on Multimedia 2025年

作者： Duan, Yule Chen, Chuang Fu, Maixia Gong, Xiuwen Niu, Yingying Luo, Fulin The University of Sydney Faculty of Engineering NSW2006 Australia Henan University of Technology Key Laboratory of Grain Information Processing and Control Ministry of Education Zhengzhou450001 China Henan University of Technology Key Laboratory of Grain Photoelectric Detection and Control Zhengzhou450001 China Chongqing University College of Computer Science Chongqing400044 China

Group convolution networks have shown great potential in hyperspectral image (HSI) classification because of their ability to divide total spectral bands into multiple groups and focus on fine discrimination within different spectral ranges. Most group convolution networks process parallel spectral groups independently;however, they neglect the important relevance of nearby spectral ranges. Moreover, the feature maps from different spectral groups are not considered for recalibration in the existing attention. To address these issues, we propose a novel group interactive threshold attention network (GITANet). In the network, a stratified-split-concatenation strategy, which not only splits all bands into multiple groups for intragroup convolution but also propagates the intergroup information via the stratified concatenation operation between different groups, is designed for bandwise group convolution. Relying on the high dependencies among nearby spectra, the cross-group interactive attention block is designed to encourage significant spectral features. Subsequently, from different spectral ranges, a learnable threshold generation block is built to estimate the information validity of each pixel. On the basis of this threshold, soft threshold spatial attention is developed in the bandwise encoderdecoder architecture, which emphasizes high-value spatial areas during the fusion of group convolutional features. Therefore, complementary and discriminative spectral-spatial features are obtained to improve the performance of HSI classification. The experimental results on three HSI datasets illustrate that GITANet is superior to several state-of-the-art networks. © 1999-2012 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Multi-scale CNN-Swin transformer network with boundary supervision for multiclass biomarker segmentation in retinal OCT images

引用

Biocybernetics and Biomedical Engineering 2025年第3期45卷 340-356页

作者： Zhanpeng Fan Xiaoming Liu Ying Zhang Jia Zhang School of Computer Science and Technology Wuhan University of Science and Technology Wuhan 430070 China Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System Wuhan 430070 China Wuhan Aier Eye Hospital of Wuhan University Wuhan 430064 China

Retinal biomarker morphology is closely associated with a variety of chronic ophthalmic diseases, in which biomarker localization and segmentation in optical coherence tomography (OCT) play a key role in the diagnosis of retina-related diseases. Although great progress has been made in deep learning based OCT biomarker segmentation, several challenges still exist. Due to issues such as image noise or class imbalance, retinal biomarkers affect the model’s recognition of other biomarkers. Moreover, small biomarkers are prone to lose accuracy during downsampling. And most existing methods rely on convolutional neural networks, which make it challenging to obtain the global context due to locality of convolution. Benefiting from the Swin Transformer with powerful modeling capabilities, we propose MSCS-Net (Multi-scale CNN-Swin Network), a network for OCT biomarker segmentation, which effectively combines CNN and Swin Transformer and integrates them in parallel into a dual-encoder structure. Specifically, an edge detection path is added alongside to enhance the localization of biomarkers at the edges. For the Swin Transformer branch, considering the irregular distribution of most OCT biomarkers, a new windowing partition is performed in the Swin Transformer to capture the features more efficiently. Meanwhile, we design a Feature Dimensionality Reduction Module to extensively collect the information of small-scale biomarkers. To effectively integrate information from two scales, we design a Transformer Cross Fusion Module to finely fuse the global and local feature information from the two-branch encoders. We validate the proposed approach on local and public datasets, and the experimental results demonstrate the effectiveness of the proposed framework.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Real-Time Student Behavior Analysis via YOLOv5 with Coordinate Attention 9th

Real-Time Student Behavior Analysis via YOLOv5 with Coordina...

引用

9th International Symposium on Artificial Intelligence and Robotics, ISAIR 2024

作者： Ma, Shuai Chang, Heyou Yang, Xinru Chen, Quan Han, Jian Key Laboratory of Intelligent Information Processing Nanjing XiaoZhuang University Nanjing China School of Information and Engineering Nanjing XiaoZhuang University Nanjing China School of Computer Science Nanjing University of Information Science and Technology Nanjing China Jiangsu Key Laboratory of Image and Video Understanding for Social Safety Nanjing University of Science and Technology Nanjing China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education Nanjing University of Science and Technology Nanjing China

ISBN: (纸本)9789819629138

The detection and recognition of student behavior play a pivotal role in the context of smart classrooms. However, conventional methods often encounter performance degradation due to challenges such as occlusion, data diversity, and complexity. To effectively addresses these issues, we propose a method that integrates YOLOv5 with coordinate attention mechanism and GhostConv for analyzing students' behavior. Specifically, the coordinate attention module enhances the localization ability of the proposed method by addressing problems related to target scale variations and uneven distribution. Additionally, we employ GhostConv to replace conventional convolutional layers in YOLOv5, reducing parameters while improving computational efficiency. To verify the validity of our method, a self-built challenging dataset for real teaching scenarios and Citypersons dataset are selected for the experiment. In comparison to the original YOLOv5, our method has achieved a 14.8% and 5.7% increase in mean average precision rates, with a 19.0% improvement in frames per second. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Corrigendum to "Hyperledger sawtooth based supplychain traceability system for counterfeit drugs" [Comput. Ind. Eng. 190 (2024) 110021] (computers & Industrial Engineering (2024) 190, (S0360835224001426), (10.1016/***.2024.110021))

引用

computers and Industrial Engineering 2025年 204卷

作者： Nawaz, Anum Wang, Liguan Irfan, Muhammad Westerlund, Tomi Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University Shanghai China Turku Intelligent and Embedded Robotic Systems Lab Faculty of Technology University of Turku Finland

关键词：

来源：评论

学校读者我要写书评

暂无评论

A spectrum-enhanced attention model for semantic segmentation of remote sensing images

A spectrum-enhanced attention model for semantic segmentatio...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal processing, ICASSP 2025

作者： Li, Xin Xu, Feng Tao, Feifei Tong, Yao Lyu, Xin Zhong, Jianyi Kaup, André College of Computer Science and Software Engineering Hohai University Nanjing China Key Laboratory of Water Big Data Technology Ministry of Water Resources Hohai University Nanjing China School of Computer Engineering Jiangsu Ocean University Lianyungang China School of Artificial Intelligence and Information Technology Nanjing University of Chinese Medicine Nanjing China Multimedia Communications and Signal Processing Friedrich-Alexander University Erlangen-Nuremberg Erlangen Germany

ISBN: (纸本)9798350368741

Semantic segmentation of remote sensing images (RSIs) is essential for applications such as environmental monitoring, urban planning, and disaster management. Convolutional Neural Networks (CNNs) and their variants struggle to capture comprehensive spectral context for learning discriminative representations. In this paper, we propose a Spectrum-Enhanced Network (SPENet) that leverages the Frequency Transformer Block (FTB) to capture rich spectral context. FTB integrates Spectrum-Enhanced Attention (SEA) with Multi-Head Frequency Self-Attention (MH-FSA), incorporating more informative contextual cues. Specifically, SEA aggregates spectral statistics through covariance matrix normalization before applying channel-wise attention. By projecting feature maps onto the frequency domain, MH-FSA provides the network with a broader context, extending beyond the low-frequency focus of standard self-attention mechanisms. Extensive experiments on the ISPRS Potsdam and LoveDA datasets show that SPENet significantly outperforms state-of-the-art methods. Besides, the proposed SEA module notably rises average F1-score/overall accuracy/mean insert over union wiht more than 2.5/2.6%/2.3%, as demonstrated by ablation study. © 2025 IEEE.

关键词： frequency transformer remote sensing images Semantic segmentation spectrum enhancement

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：