检索结果-内蒙古大学图书馆

2021 IEEE International Conference on Multimedia and Expo, ICME 2021

作者： Yi, Jinhan Liu, Xin Cheung, Yiu-Ming Xu, Xing Fan, Wentao He, Yi Department of Computer Science and Technology Huaqiao University Xiamen361021 China Xiamen Key Lab. of Computer Vision and Pattern Recognition Fujian Key Lab. of Big Data Intelligence and Security China Department of Computer Science Hong Kong Baptist University Kowloon Hong Kong School of Computer Science and Engineering University of Electronic Science and Technology of China China Provincial Key Laboratory for Computer Information Processing Technology Soochow University China

ISBN: (纸本)9781665438643

Existing cross-modal hashing still faces three challenges: (1) Most batch-based methods are unsuitable for processing large-scale and streaming data. (2) Current online methods often suffer from insufficient semantic association, while lacking flexibility to learn the hash functions for varying streaming data. (3) Existing supervised methods always require much computation time or accumulate large quantization loss to learn hash codes. To address above challenges, we present an efficient Online Label Consistent Hashing (OLCH) for cross-modal retrieval, which aims to incrementally learn hash codes for the current arriving data, while updating the hash functions at a streaming manner. To be specific, an online semantic representation learning framework is designed to adaptively preserve the semantic similarity across different modalities, and a mini-batch online gradient descent approach associated with forward-backward splitting is developed to optimize the hash functions. Accordingly, the hash codes are adaptively learned online with the high discriminative capability, while avoiding high computation complexity to process the streaming data. Experimental results show its outstanding performance in comparison with the-state-of-arts. © 2021 IEEE Computer Society. All rights reserved.

关键词： Hash functions

来源：评论

学校读者我要写书评

暂无评论

Model Discrepancy Learning: Synthetic Faces Detection Based on Multi-Reconstruction

arXiv

引用

arXiv 2025年

作者： Jiang, QingChao Xu, ZhiShuo Zhu, ZhiYing Chen, Ning Wang, HaoYue Ba, ZhongJie East China University of Science and Technology Shanghai China Fudan University Shanghai China The State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou China

Advances in image generation enable hyper-realistic synthetic faces but also pose risks, thus making synthetic face detection crucial. Previous research focuses on the general differences between generated images and real images, often overlooking the discrepancies among various generative techniques. In this paper, we explore the intrinsic relationship between synthetic images and their corresponding generation technologies. We find that specific images exhibit significant reconstruction discrepancies across different generative methods and that matching generation techniques provide more accurate reconstructions. Based on this insight, we propose a Multi-Reconstruction-based detector. By reversing and reconstructing images using multiple generative models, we analyze the reconstruction differences among real, GAN-generated, and DM-generated images to facilitate effective differentiation. Additionally, we introduce the Asian Synthetic Face dataset (ASFD), containing synthetic Asian faces generated with various GANs and DMs. This dataset complements existing synthetic face datasets. Experimental results demonstrate that our detector achieves exceptional performance, with strong generalization and robustness. Copyright © 2025, The Authors. All rights reserved.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Experimental Analysis of Selective Imitation for Multifactorial Differential Evolution 14th

Experimental Analysis of Selective Imitation for Multifactor...

引用

14th International Conference on Bio-inspired Computing: Theories and Applications, BIC-TA 2019

作者： Peng, Deming Cai, Yiqiao Fu, Shunkai Luo, Wei College of Computer Science and Technology Huaqiao University Quanzhou China Fujian Key Laboratory of Big Data Intelligence and Security Huaqiao University Xiamen361000 China

ISBN: (纸本)9789811534249

Recently, evolutionary multitasking optimization (EMTO) is proposed as a new emerging optimization paradigm to simultaneously solve multiple optimization tasks in a cooperative manner. In EMTO, the knowledge transfer between tasks is mainly carried out through the assortative mating and selective imitation operators. However, in the literature of EMTO, little study on the selective imitation operator has yet been done to provide a deeper insight in the knowledge transfer across different tasks. Based on this consideration, we firstly study the influence of the inheritance probability (IP) of the selective imitation on an EMTO algorithm, multifactorial differential evolution (MFDE), through the experimental analysis. Then, an adaptive inheritance mechanism (AIM) is introduced into the selective imitation operator of MFDE to automatically adjust the IP value for different tasks at different evolutionary stages. The experimental results on a suite of single-objective multitasking benchmark problems have demonstrated the effectiveness of AIM in enhancing the performance of MFDE. © 2020, Springer Nature Singapore Pte Ltd.

关键词： Multitasking

来源：评论

学校读者我要写书评

暂无评论

Public Auditing for Multi-Version Cloud data with Storage Efficiency and Version Dynamics

SSRN

引用

SSRN 2022年

作者： Ye, Weiping Nan, Fulin Tian, Hui Quan, Hanyu Chang, Chin-Chen College of Computer Science and Technology National Huaqiao University Xiamen361021 China Wuhan National Laboratory for Optoelectronics Wuhan430074 China Fujian Key Laboratory of Big Data Intelligence and Security National Huaqiao University Xiamen361021 China Department of Information and Computer Science Feng Chia University Taichung40724 Taiwan

Multi-version data storage has been an increasingly popular service in cloud computing, which is shown to support flexible rollback recovery of cloud data. However, how to verify the integrity of this kind of data efficiently still remains open. As an attempt at addressing this issue, this paper presents a novel public auditing scheme for multi-version data in the cloud to achieve secure and efficient verification for all the versions of cloud data, which employs a three-pronged approach. First, we propose an incremental storage table (IST) to store all data versions in an efficient and low-cost manner;second, we develop a dynamic version control table (DVCT) to support efficient update for the multi-version data;finally, we design a custom-built tag-generation algorithm that can achieve authentication for all versions with only one tag. We formally prove the security of the presented scheme and evaluate its performance through comprehensive experiments and comparisons with state-of-the-art ones. The experimental results demonstrate that our scheme can effectively achieve public auditing for multi-version data while providing excellent security and outperforming state-of-the-art schemes in terms of computation and communication costs. © 2022, The Authors. All rights reserved.

关键词： Signal encoding

来源：评论

学校读者我要写书评

暂无评论

Ppdrm: Privacy-Preserving Drm Training and Classification on the Cloud

SSRN

引用

SSRN 2023年

作者： Xu, Lida Tian, Chengliang Zhang, Guoyan Li, Leibo Tian, Weizhong Zhang, Yan College of Computer Science and Technology Qingdao University Qingdao266071 China School of Cyber Science and Technology Shandong University Shandong Ji’nan250101 China Key Laboratory of Cryptologic Technology and Information Security Ministry of Education Shandong Ji’nan250101 China Shandong Institute of Blockchain Shandong Ji’nan250101 China College of Big Data and Internet Shenzhen Technology University Guangdong Shenzhen518118 China College of Electromechanical Engineering Qingdao University of Science and Technology Qingdao266061 China

The prevalence of cloud computing greatly promotes the development of artificial intelligence(AI). Accurately and efficiently classify data on the cloud is a classic AI task. The recent Discriminative Ridge Machine(DRM) is a new type of supervised classifier which outperforms many other classic classifiers in accuracy and efficiency, especially for high-dimensional data or unbalanced ***, the DRM training and classification task is tedious for resource-constrained data owners. Motivated by this, in this paper, we study the outsourcing of DRM training and classification within a cloud environment and propose a quantum secure and efficient solution for this problem. Specifically, based on the two non-colluding servers model, we adopt the quantum-immune CKKS cryptosystem to design a series of homomorphic security protocols for matrix and argmin operations and realize the correct model training and data classification without leaking the privacy of training dataset, the DRM model, the query data and the classification results. We argue the $\mathcal{CPA}$ secure of our design and evaluate the accuracy and efficiency from theory and experiment perspectives. Our theoretical and experimental analysis demonstrates that our design can achieve the same classification accuracy as the origin plaintext-based DRM classifier and the query-response time is reasonable and acceptable. © 2023, The Authors. All rights reserved.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Practitioners’ Expectations on Log Anomaly Detection

arXiv

引用

arXiv 2024年

作者： Ma, Xiaoxue Li, Yishu Keung, Jacky Yu, Xiao Zou, Huiqi Yang, Zhen Sarro, Federica Barr, Earl T. Department of Electronic Engineering and Computer Science Hong Kong Metropolitan University Hong Kong Department of Computer Science City University of Hong Kong Hong Kong State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou China Department of Computer Science Johns Hopkins University Baltimore United States School of Computer Science and Technology Shandong University Shandong China Department of Computer Science University College London London United Kingdom

Log anomaly detection has become a common practice for software engineers to analyze software system behavior. Despite significant research efforts in log anomaly detection over the past decade, it remains unclear what are practitioners’ expectations on log anomaly detection and whether current research meets their needs. To fill this gap, we conduct an empirical study, surveying 312 practitioners from 36 countries about their expectations on log anomaly detection. In particular, we investigate various factors influencing practitioners’ willingness to adopt log anomaly detection tools. We then perform a literature review on log anomaly detection, focusing on publications in premier venues from 2014 to 2024, to compare practitioners’ needs with the current state of research. Based on this comparison, we highlight the directions for researchers to focus on to develop log anomaly detection techniques that better meet practitioners’ expectations. © 2024, CC BY-NC-SA.

关键词：

来源：评论

学校读者我要写书评

暂无评论

PON: Proposal Optimization Network for Temporal Action Proposal Generation 16th

PON: Proposal Optimization Network for Temporal Action Propo...

引用

16th International Conference on Intelligent Computing, ICIC 2020

作者： Peng, Xiaoxiao Du, Jixiang Zhang, Hongbo Department of Computer Science and Technology Huaqiao University Quanzhou China Fujian Key Laboratory of Big Data Intelligence and Security Huaqiao University Quanzhou China Xiamen Key Laboratory of Computer Vision and Pattern Recognition Huaqiao University Quanzhou China

ISBN: (纸本)9783030607951

Temporal action localization is a challenging task in video understanding. Although great progress has been made in temporal action localization, the most advanced methods still have the problem of sharp performance degradation when an action proposal generated. Most methods use sliding windows method or simply group frames according to frame-level scores. These methods are not enough to provide accurate action boundary and maintain reasonable temporal structure. In order to solve these problems, we propose a novel proposal optimization network to generate start score, end score, action score and regression score, and then remove the redundancy by NMS algorithm. In the proposed method, we introduce a metric loss function to maintain the temporal structure of action proposal in the training process. To verify the effectiveness of the proposed method, we have made comparative experiments on ActivityNet-1.3 dataset respectively, and the proposed method has surpassed some of the state-of-the-art methods on the dataset. © 2020, Springer Nature Switzerland AG.

关键词： Passive optical networks

来源：评论

学校读者我要写书评

暂无评论

Towards Sample-specific Backdoor Attack with Clean Labels via Attribute Trigger

arXiv

引用

arXiv 2023年

作者： Li, Yiming Zhu, Mingyan Guo, Junfeng Wei, Tao Xia, Shu-Tao Qin, Zhan Hangzhou311215 China School of Cyber Science and Technology Zhejiang University Hangzhou311200 China Tsinghua Shenzhen International Graduate School Tsinghua University Shenzhen518055 China Department of Computer Science University of Maryland College ParkMD20742 United States Ant Group Hangzhou310023 China The Research Center of Artificial Intelligence Peng Cheng Laboratory Shenzhen518000 China The State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou311200 China

Currently, sample-specific backdoor attacks (SSBAs) are the most advanced and malicious methods since they can easily circumvent most of the current backdoor defenses. In this paper, we reveal that SSBAs are not sufficiently stealthy due to their poisoned-label nature, where users can discover anomalies if they check the image-label relationship. In particular, we demonstrate that it is ineffective to directly generalize existing SSBAs to their clean-label variants by poisoning samples solely from the target class. We reveal that it is primarily due to two reasons, including (1) the 'antagonistic effects' of ground-truth features and (2) the learning difficulty of sample-specific features. Accordingly, trigger-related features of existing SSBAs cannot be effectively learned under the clean-label setting due to their mild trigger intensity required for ensuring stealthiness. We argue that the intensity constraint of existing SSBAs is mostly because their trigger patterns are 'content-irrelevant' and therefore act as 'noises' for both humans and DNNs. Motivated by this understanding, we propose to exploit content-relevant features, a.k.a. (human-relied) attributes, as the trigger patterns to design clean-label SSBAs. This new attack paradigm is dubbed backdoor attack with attribute trigger (BAAT). Extensive experiments are conducted on benchmark datasets, which verify the effectiveness of our BAAT and its resistance to existing defenses. © 2023, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

data Governance: Status, Technologies, Applications and Development Trends

Data Governance: Status, Technologies, Applications and Deve...

引用

IEEE International Conference on data Science in Cyberspace (DSC)

作者： Haiyang Zhu Xiaofeng Chen Suyan Li Yejin Wang Xinwei Fang Minfeng Zhu The state Key Lab of CAD&CG Zhejiang University Wuchan Zhongda Digital Technology Co. Ltd. Hangzhou China The state Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou Qulian Technology Co. Ltd. Hangzhou China Zhejiang Institute of Mechanical & Electrical Engineering Hangzhou China Wuchan Zhongda Digital Technology Co. Ltd. Hangzhou China Department of Internet and Artificial Intelligence Sir Run Run Shaw Hospital School of Medicine Zhejiang University Hangzhou China Zhejiang University Hangzhou China

data governance involves legal, policy, technologies, and standards measures to ensure security, accuracy, integrity, consistency, and accessibility of data, with the primary objective of unlocking its latent value. In this context, the study initiates by scrutinizing the present status of data governance, subsequently identifying pivotal challenges through the lens of computer sciences. Moreover, to effectively grapple with the intricacies of high-dimensional data, data incompletion, and data redundancy within the realm of data governance, we present a compendium of data governance technology methodologies, encompassing data dimensionality reduction, data imputation, data deduplication, and data visualization. Further enhancing our discourse, we shed light on the typical scenarios of large-scale supply chain enterprises, elucidating data analysis and data intelligence applications. The convergence of these insights seeks to provide a comprehensive overview both data governance technology research and real-world application.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Ablockchain-enabled Efficient and Fair Energy Trading Schemes in Industrial Internet of Things

ABlockchain-enabled Efficient and Fair Energy Trading Scheme...

引用

IEEE International Symposium on Dependable, Autonomic and Secure Computing (DASC)

作者： Jianwen Hu Yuling Chen Xiaoiun Ren Yixian Yang Xiaobin Qian State Key Laboratory of Public Big Data College of Computer Science and Technology Guizhou University Guiyang China Blockchain Laboratory of Agricultural Vegetables Weifang University of Science and Technology Weifang China School of Cyberspace Security Beijing University of Posts and Telecommnuications Beijing China Guizhou CoVision Science & Technology Co. Ltd Guiyang China

ISBN: (纸本)9781665421751

As the technical support of the Industrial Internet of Things, blockchain technology has been widely used in energy trading, data transactions, and Internet of Vehicles. However, among the existing energy trading models, most of them only address the transaction security and transaction privacy issues that arise in the energy trading process, ignoring the fairness of resource allocation and transaction equity in the trading process. In order to tackle those problems, an energy trading scheme called Ho-trade is proposed in this paper to improve the efficiency of model trading while ensuring the fairness of energy trading. We propose a new trading strategy in the Ho-trade energy trading scheme that guarantees fairness in the allocation of trading re-sources by introducing an entity's active-reputation value. Based on the identity verification foundation of the consortium chain, the scheme enhances the existing PBFT consensus algorithm and improves the efficiency of model transactions. The experimental simulation indicates that the scheme requires less transaction time and has higher transaction fairness and security.

关键词： Privacy data privacy Cloud computing Consensus algorithm data models Delays blockchains

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：