检索结果-内蒙古大学图书馆

Disentangled Noisy Correspondence Learning

IEEE Transactions on Image Processing 2025年 34卷 2602-2615页

作者： Dang, Zhuohang Luo, Minnan Wang, Jihong Jia, Chengyou Han, Haochen Wan, Herun Dai, Guang Chang, Xiaojun Wang, Jingdong Xi’an Jiaotong University School of Computer Science and Technology Ministry of Education Key Laboratory of Intelligent Networks and Network Security Shaanxi Province Key Laboratory of Big Data Knowledge Engineering Shaanxi Xi’an710049 China SGIT AI Laboratory Xi’an710048 China State Grid Corporation of China State Grid Shaanxi Electric Power Company Ltd. Xi’an710048 China University of Science and Technology of China School of Information Science and Technology Hefei230026 China Abu Dhabi United Arab Emirates Baidu Inc. Beijing100085 China

Cross-modal retrieval is crucial in understanding latent correspondences across modalities. However, existing methods implicitly assume well-matched training data, which is impractical as real-world data inevitably involves imperfect alignments, i.e., noisy correspondences. Although some works explore similarity-based strategies to address such noise, they suffer from sub-optimal similarity predictions influenced by modality-exclusive information (MEI), e.g., background noise in images and abstract definitions in texts. This issue arises as MEI is not shared across modalities, thus aligning it in training can markedly mislead similarity predictions. Moreover, although intuitive, directly applying previous cross-modal disentanglement methods suffers from limited noise tolerance and disentanglement efficacy. Inspired by the robustness of information bottlenecks against noise, we introduce DisNCL, a novel information-theoretic framework for feature Disentanglement in Noisy Correspondence Learning, to adaptively balance the extraction of modality-invariant information (MII) and MEI with certifiable optimal cross-modal disentanglement efficacy. DisNCL then enhances similarity predictions in modality-invariant subspace, thereby greatly boosting similarity-based alleviation strategy for noisy correspondences. Furthermore, DisNCL introduces soft matching targets to model noisy many-to-many relationships inherent in multi-modal inputs for noise-robust and accurate cross-modal alignment. Extensive experiments confirm DisNCL’s efficacy by 2% average recall improvement. Mutual information estimation and visualization results show that DisNCL learns meaningful MII/MEI subspaces, validating our theoretical analyses. © 1992-2012 IEEE.

关键词： information theory

来源：评论

学校读者我要写书评

暂无评论

Exploring Triple Knowledge Cues for Zero-Shot Human-Object Interaction Detection

Exploring Triple Knowledge Cues for Zero-Shot Human-Object I...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Fang Nan Ni Zhang Qidong Liu Wei Jing Guang Dai Yan Chen Feng Tian Faculty of Electronic and Information Engineering Xi’an Jiaotong University Xi’an China Ministry of Education Key Laboratory of Intelligent Networks and Network Security Xi’an Jiaotong University Xi’an China School of Automation Northwestern Polytechnical University Xi’an China Lenovo Research Bei Jing China SGIT AI Lab State Grid Corporation of China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Current zero-shot human-object interaction detection methods often follow a two-phase pipeline, which uses a pre-trained detector to detect instances and then adopts CLIP to perform interaction prediction. During the second phase, they either obtain pairwise representations by directly performing RoI-Align on CLIP features or designing additional queries and decoders to fuse CLIP features. However, CLIP visual features often lack fine-grained information, thus being detrimental to capturing complex HOI interactions. Besides, extra decoders might increase computation costs. Thus, we propose a triple knowledge cues exploration model without extra decoders to explore various knowledge guidance for improving CLIP representations. First, we incorporate position distribution and semantic priors to delineate a layout from the predicted boxes and inject semantics by using the CLIP text embeddings. Next, we explore object priors by leveraging predefined class names and the text encoder to obtain saliency maps for humans and objects. Then, we design three types of holistic tokens to capture diverse attribute cues for human, object, and interaction, respectively. The above cues are finally integrated into a vanilla two-stage CLIP-based baseline. The experimental results on HICO-DET demonstrate the effectiveness of our proposed model.

关键词： Visualization Costs Fuses Semantics Layout Pipelines Detectors Signal processing Decoding Speech processing

来源：评论

学校读者我要写书评

暂无评论

security Vulnerability Assessment of Power IoT based on Business security

Security Vulnerability Assessment of Power IoT based on Busi...

引用

2020 International Conference on Control, Robotics and Intelligent System, CCRIS 2020

作者： Fei, Jiaxuan Chen, Kai Yao, Qigui Guo, Qian Wang, Xiangqun State Grid Key Laboratory of Information and Network Security China

ISBN: (纸本)9781450388054

Power Internet of Things is the application of IoT in smart power grid. Once attacked, it will cause huge losses. Therefore, it is necessary to conduct a security assessment to take defensive measures. However, the traditional vulnerability assessment methods of the power Internet of things mostly focus on the security of the system itself, without considering the impact on business economy and efficiency. This paper proposes a security vulnerability assessment method of power Internet of Things integrating business security. This method first analyzes the security risks faced by the power Internet of Things, and establishes its attack tree model. Then, each leaf node is rated from the three safety features, which are weighted by evaluation and calculation, and the activation probability of each leaf node is calculated. After that, considering the blind attack factor, the activation probability of all nodes in the model is calculated. Finally, the vulnerability of the system and the vulnerability sensitivity of each leaf node are obtained. According to the vulnerability sensitivity, measures are taken to protect the weak links of the system. The effectiveness of the proposed method is verified by experiments on SCADA (supervisory control and data acquisition) system in the power Internet of things. © 2020 ACM.

关键词： SCADA systems

来源：评论

学校读者我要写书评

暂无评论

Dynamic Access Control and Authorization System based on Zero-trust architecture

Dynamic Access Control and Authorization System based on Zer...

引用

2020 International Conference on Control, Robotics and Intelligent System, CCRIS 2020

作者： Yao, Qigui Wang, Qi Zhang, Xiaojian Fei, Jiaxuan State Grid Key Laboratory of Information and Network Security China

ISBN: (纸本)9781450388054

With the development of cloud computing, artificial intelligence, big data and other technologies, network systems are facing more and more security risks and threats. The traditional security architecture based on border protection cannot meet the increasing security protection requirements. The zero-trust security architecture which has the characteristics of continuous identity authentication and minimized authority allocation can adapt to the security protection requirements of most current network systems. Based on the zero-trust security architecture, a dynamic access control and authorization system is proposed. User portraits and user trust are generated according to user behavior. Real-time hierarchical control in different scenarios is used in the system to achieve dynamic and fine-grained access control and authorization. © 2020 ACM.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

Density-aware and Depth-aware Visual Representation for Zero-Shot Object Counting

Density-aware and Depth-aware Visual Representation for Zero...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Fang Nan Feng Tian Ni Zhang Nian Liu Haonan Miao Guang Dai Mengmeng Wang Faculty of Electronic and Information Engineering Xi’an Jiaotong University Xi’an China Ministry of Education Key Laboratory of Intelligent Networks and Network Security Xi’an Jiaotong University Xi’an China School of Automation Northwestern Polytechnical University Xi’an China Computer Vision Department Mohamed bin Zayed University of Artificial Intelligence Abu Dhabi UAE SGIT AI Lab State Grid Corporation of China Zhejiang University of Technology Hang Zhou China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Previous methods often utilize CLIP semantic classifiers with class names for zero-shot object counting. However, they ignore crucial density and depth knowledge for counting tasks. Thus, we propose a density-aware and depth-aware prompt counting model, which captures density information via learning density-aware prompts based on density-aware contrastive loss and incorporates depth guidance with predefined depth-aware prompts. To facilitate the training process, we design two strategies for standard counting loss and the contrastive loss, where the former prioritizes larger and sparser objects initially, gradually focusing on smaller and denser objects, and the latter adopts coarse-to-fine density learning. Besides, we construct a dataset named LVIS-372 with more real-world scenarios and balanced instance distribution compared to existing ones. Finally, the experimental results demonstrate the effectiveness of our proposed method.

关键词： Training Visualization Semantics Focusing Signal processing Acoustics Power capacitors Speech processing Standards

来源：评论

学校读者我要写书评

暂无评论

An Attack Intention Recognition Method Based on Evaluation Index System of Electric Power information System 2

An Attack Intention Recognition Method Based on Evaluation I...

引用

2017 IEEE 2nd information Technology, networking, Electronic and Automation Control Conference (ITNEC 2017)

作者： Xiaojuan Guan Yuanyuan Ma Ye Hua State Grid Key Laboratory of Information & Network Security Global Energy Interconnection Research Institute

ISBN: (纸本)9781509064151;9781509064144

With the increasing scale of the network, the power information system has many characteristics, such as large number of nodes, complicated structure, diverse network protocols and abundant data, which make the network intrusion detection system difficult to detect real alarms. The current security technologies cannot meet the actual power system network security operation and protection requirements. Based on the attacker ability, the vulnerability information and the existing security protection configuration, we construct the attack sub-graphs by using the parallel distributed computing method and combine them into the whole network attack graph. The vulnerability exploit degree, attacker knowledge, attack proficiency, attacker willingness and the confidence level of the attack evidence are used to construct the security evaluation index system of the power information network system to calculate the attack probability value of each node of the attack graph. According to the probability of occurrence of each node attack, the pre-order attack path will be formed and then the most likely attack path and attack targets will be got to achieve the identification of attack intent.

关键词： parallel computing attack graph vulnerability exploit attacker knowledge attack proficiency attacker willingness attack confidence

来源：评论

学校读者我要写书评

暂无评论

Sensitive Data Comparison Algorithm Based Spatio-temporal Label Distribution Fusion 2

Sensitive Data Comparison Algorithm Based Spatio-temporal La...

引用

2nd International Conference on Algorithms, High Performance Computing and Artificial Intelligence, AHPCAI 2022

作者： Yu, Pengfei Shi, Congcong State Grid Smart Grid Research Institute Co. Ltd State Grid Key Laboratory of Information & Network Security Nanjing China

While the IoT cooperates with big data to deeply reconstruct all walks of life, it also poses more severe challenges to data security. Accurate identification of sensitive data is a prerequisite for data security. Compared with traditional machine learning algorithms, deep learning algorithms show great functionality and flexibility in large-scale data processing. However, the existing deep learning-based sensitive data identification methods focus on the mining of a single content feature, ignoring contextual information, and the identification accuracy of sensitive data with insignificant content features is not ***, this paper proposes a sensitive data comparison algorithm based on spatiotemporal label distribution fusion. The algorithm can simultaneously model the spatial and temporal patterns of the data flow, mine the spatial and temporal labels, and identify the type of data through a comprehensive judgment strategy. It solves the problem of identifying sensitive data with insignificant content characteristics. Finally, the algorithm is independently repeated experiments on multiple data sets and compared with multiple algorithms. The results show that the Best F-score and NAB score of this model are significantly better than other algorithms, which are 0.812 and 69.2, respectively. The algorithm proposed in this paper can more accurately identify sensitive data. © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).

关键词： Sensitive data

来源：评论

学校读者我要写书评

暂无评论

Research on "cloud-Edge-End" security Protection System of Internet of Things Based on National Secret Algorithm 4

Research on "cloud-Edge-End" Security Protection System of I...

引用

4th IEEE Advanced information Technology, Electronic and Automation Control Conference, IAEAC 2019

作者： Gao, Peng Yang, Ruxia Gao, Xianzhou Global Energy Interconnection Research Institute State Grid Key Laboratory of Information and Network Security China

ISBN: (纸本)9781728119076

With the rapid development of Internet of Things applications, Internet of Things technology and applications have become more and more popular, and the perceived risk of sensing layer terminals, transmission and application has become more and more important. This paper first analyzes the security risks faced by the Internet of Things from the perception layer, transport layer and application layer;then describes the cryptographic algorithm technology and its application;Finally, based on the cryptographic technology, the security protection measures such as security authentication, data transmission and storage between the sensing layer, the edge side, the network layer and the cloud platform are designed respectively, and the security system based on the key system is proposed. © 2019 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Research and Implementation of Electric Equipment Connectivity Data Analysis Model Based on Graph Database 1

引用

4th International Conference on 3D Imaging Technologies- Multidimensional Signal Processing and Deep Learning, 3D IT-MSP and DL 2022

作者： Qiao, Junfeng Peng, Lin Zhou, Aihua Zhu, Lipeng Yang, Pei Pan, Sen State Grid Key Laboratory of Information & Network Security State Grid Smart Grid Research Institute Co. Ltd Nanjing210014 China

ISBN: (数字)9789819911455

ISBN: (纸本)9789819911448

Effective description of electric data resources is an important basis for electric data modeling. With the access of large-scale distributed new energy, electric allocation data resources are more complex and changeable, and existing data description methods are difficult to accurately describe. Aiming at the above problems, the research on hierarchical description method of electric distribution and distributed new energy data resources based on complex network is carried out. Based on the complex network theory, the hierarchical description method of data resources is formed by comprehensively describing the external characteristics, internal elements, organizational structure, and relationships between data resources. The holographic modeling technology of electric distribution and distributed new energy data resources based on graphic computing is studied. Data modeling is an important basis for data fusion and sharing. The production, operation, control, measurement, and other business links in the field of electric distribution and distributed new energy maintain a certain degree of independence from each other, and there is an "information island" problem in the data resources of each link. In view of the above problems, research on holographic modeling technology of electric distribution and distributed new energy data resources based on graphic computing is carried out. Based on graphic computing technology, equipment is taken as the core, covering resources, assets, measurement, topology, graphics, and other all-round information, to build a holographic data model of electric distribution and distributed new energy data resources, and support the deepening integration and sharing of electric distribution data. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Complex networks

来源：评论

学校读者我要写书评

暂无评论

security situation awareness method of power mobile application based on big data architecture 5

Security situation awareness method of power mobile applicat...

引用

5th International Conference on Data Science and information Technology, DSIT 2022

作者： Yong, Li Mu, Chen Zaojian, Dai Lu, Chen State Grid Key Laboratory of Information & Network Security State Grid Smart Grid Research Enstitute Co. Ltd Nanjing210003 China

ISBN: (数字)9781665498685

ISBN: (纸本)9781665498685

According to the characteristics of security threats and massive users in power mobile applications, a mobile application security situational awareness method based on big data architecture is proposed. The method uses open-source big data technology frameworks such as Kafka, Flink, Elasticsearch, etc. to complete the collection, analysis, storage and visual display of massive power mobile application data, and improve the throughput of data processing. The security situation awareness method of power mobile application takes the mobile terminal threat index as the core, divides the risk level for the mobile terminal, and predicts the terminal threat index through support vector machine regression algorithm (SVR), so as to construct the security profile of the mobile application operation terminal. Finally, through visualization services, various data such as power mobile applications and terminal assets, security operation statistics, security strategies, and alarm analysis are displayed to guide security operation and maintenance personnel to carry out power mobile application security monitoring and early warning, banning disposal and traceability analysis and other decision-making work. The experimental analysis results show that the method can meet the requirements of security situation awareness for threat assessment accuracy and response speed, and the related results have been well applied in a power company. © 2022 IEEE.

关键词： Big data

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：