检索结果-内蒙古大学图书馆

Securely Fine-tuning Pre-trained Encoders Against Adversarial Examples

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhou, Ziqi Li, Minghui Liu, Wei Hu, Shengshan Zhang, Yechao Wan, Wei Xue, Lulu Zhang, Leo Yu Yao, Dezhong Jin, Hai National Engineering Research Center for Big Data Technology and System China Services Computing Technology and System Lab. China Cluster and Grid Computing Lab. China Hubei Engineering Research Center on Big Data Security China Hubei Key Laboratory of Distributed System Security China School of Computer Science and Technology Huazhong University of Science and Technology China School of Software Engineering Huazhong University of Science and Technology China School of Cyber Science and Engineering Huazhong University of Science and Technology China School of Information and Communication Technology Griffith University Australia

With the evolution of self-supervised learning, the pre-training paradigm has emerged as a predominant solution within the deep learning landscape. Model providers furnish pre-trained encoders designed to function as versatile feature extractors, enabling downstream users to harness the benefits of expansive models with minimal effort through fine-tuning. Nevertheless, recent works have exposed a vulnerability in pre-trained encoders, highlighting their susceptibility to downstream-agnostic adversarial examples (DAEs) meticulously crafted by attackers. The lingering question pertains to the feasibility of fortifying the robustness of downstream models against DAEs, particularly in scenarios where the pre-trained encoders are publicly accessible to the attackers. In this paper, we initially delve into existing defensive mechanisms against adversarial examples within the pre-training paradigm. Our findings reveal that the failure of current defenses stems from the domain shift between pre-training data and downstream tasks, as well as the sensitivity of encoder parameters. In response to these challenges, we propose Genetic Evolution-Nurtured Adversarial Fine-tuning (Gen-AF), a two-stage adversarial fine-tuning approach aimed at enhancing the robustness of downstream models. Gen-AF employs a genetic-directed dual-track adversarial fine-tuning strategy in its first stage to effectively inherit the pre-trained encoder. This involves optimizing the pre-trained encoder and classifier separately while incorporating genetic regularization to preserve the model’s topology. In the second stage, Gen-AF assesses the robust sensitivity of each layer and creates a dictionary, based on which the top-k robust redundant layers are selected with the remaining layers held fixed. Upon this foundation, we conduct evolutionary adaptability fine-tuning to further enhance the model’s generalizability. Our extensive experiments, conducted across ten self-supervised training methods and six

关键词： Topology

Detector Collapse: Physical-World Backdooring Object Detection to Catastrophic Overload or Blindness in Autonomous Driving

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhang, Hangtao Hu, Shengshan Wang, Yichen Zhang, Leo Yu Zhou, Ziqi Wang, Xianlong Zhang, Yanjun Chen, Chao School of Cyber Science and Engineering Huazhong University of Science and Technology China School of Computer Science and Technology Huazhong University of Science and Technology China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Engineering Research Center on Big Data Security China Hubei Key Laboratory of Distributed System Security China Cluster and Grid Computing Lab China School of Information and Communication Technology Griffith University Australia University of Technology Sydney Australia RMIT University Australia

Object detection tasks, crucial in safety-critical systems like autonomous driving, focus on pinpointing object locations. These detectors are known to be susceptible to backdoor attacks. However, existing backdoor techniques have primarily been adapted from classification tasks, overlooking deeper vulnerabilities specific to object detection. This paper is dedicated to bridging this gap by introducing Detector Collapse (DC), a brand-new backdoor attack paradigm tailored for object detection. DC is designed to instantly incapacitate detectors (i.e., severely impairing detector's performance and culminating in a denial-of-service). To this end, we develop two innovative attack schemes: SPONGE for triggering widespread misidentifications and BLINDING for rendering objects invisible. Remarkably, we introduce a novel poisoning strategy exploiting natural objects, enabling DC to act as a practical backdoor in real-world environments. Our experiments on different detectors across several benchmarks show a significant improvement (∼10%-60% absolute and ∼2-7× relative) in attack efficacy over state-of-the-art attacks. Copyright © 2024, The Authors. All rights reserved.

关键词： Denial-of-service attack

BadHash: Invisible Backdoor Attacks against Deep Hashing with Clean Label

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Hu, Shengshan Zhou, Ziqi Zhang, Yechao Zhang, Leo Yu Zheng, Yifeng He, Yuanyuan Jin, Hai School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China School of Information Technology Deakin University VIC3216 Australia School of Computer Science and Technology Harbin Institute of Technology Shenzhen China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Engineering Research Center on Big Data Security HUST Wuhan430074 China Cluster and Grid Computing Lab HUST Wuhan430074 China

Due to its powerful feature learning capability and high efficiency, deep hashing has achieved great success in large-scale image retrieval. Meanwhile, extensive works have demonstrated that deep neural networks (DNNs) are susceptible to adversarial examples, and exploring adversarial attack against deep hashing has attracted many research efforts. Nevertheless, backdoor attack, another famous threat to DNNs, has not been studied for deep hashing yet. Although various backdoor attacks have been proposed in the field of image classification, existing approaches failed to realize a truly imperceptive backdoor attack that enjoys invisible triggers and clean label setting simultaneously, and they also cannot meet the intrinsic demand of image retrieval backdoor. In this paper, we propose BadHash, the first generative-based imperceptible backdoor attack against deep hashing, which can effectively generate invisible and input-specific poisoned images with clean label. Specifically, we first propose a new conditional generative adversarial network (cGAN) pipeline to effectively generate poisoned samples. For any given benign image, it seeks to generate a natural-looking poisoned counterpart with a unique invisible trigger. In order to improve the attack effectiveness, we introduce a label-based contrastive learning network LabCLN to exploit the semantic characteristics of different labels, which are subsequently used for confusing and misleading the target model to learn the embedded trigger. We finally explore the mechanism of backdoor attacks on image retrieval in the hash space. Extensive experiments on multiple benchmark datasets verify that BadHash can generate imperceptible poisoned samples with strong attack ability and transferability over state-of-the-art deep hashing schemes. © 2022, CC BY.

关键词： Deep neural networks

Teleportation of a ququart system using hyperentangled photons assisted by atomic-ensemble memories

学校读者我要写书评

暂无评论

Physical Review A 2016年第1期93卷 012332-012332页

作者： Ming-Xing Luo Hui-Ran Li Hong Lai Xiaojun Wang Information Security and National Computing Grid Laboratory School of Information Science and Technology Southwest Jiaotong University Chengdu 610031 China State Key Laboratory of Information Security (Graduate University of Chinese Academy of Sciences) Beijing 100049 China School of Computer and Information Science Southwest University Chongqing 400715 China School of Electronic Engineering Dublin City University Dublin 9 Ireland

A single photon encoded in both the spin and the orbital angular momentum has recently been experimentally demonstrated [X.-L. Wang et al., Nature 518, 516 (2015)] with linear optics using the hyperentangled state, which can be viewed as a bipartite four-dimensional (ququart) entanglement. Here, we investigate this process from a general point of view. By exploring a controlled phase flip induced by atomic ensembles in one-side optical microcavities, we propose teleportations of general ququart systems including a two-atomic-ensemble system, a two-polarized-photon system, one photon with the polarization and spatial degrees of freedom (DOFs), and a hybrid photon-ensemble system using two hyperentangled photons. The output information may also be encoded by different physical systems up to the special requirements of a receiver. These schemes are also adapted to teleportation of a ququart system with only phases or real probability amplitudes, which is beyond previous superdense teleportation [Nature Commun. 6, 7185 (2015)]. With these restrictions, half of the classical communication cost may be saved and experimental complexities are also reduced. Our theoretical schemes are feasible in modern physics and show the possibilities of transferring complex quantum systems for scalable quantum applications.

关键词： Entanglement production Quantum information with atoms & light

An efficient privacy-preserving compressive data gathering scheme in WSNs 1

学校读者我要写书评

暂无评论

15th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2015

作者： Xie, Kun Ning, Xueping Wang, Xin Wen, Jigang Liu, Xiaoxiao He, Shiming Zhang, Daqiang College of Computer Science and Electronics Engineering Hunan University Changsha410082 China The State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommmunications Beijing100876 China Department of Electrical and Computer Engineering State University of New York at Stony Brook New York11790 United States Institute of Computing Technology Chinese Academy of Sciences Beijing100080 China State Grid HuNan Electric Power Company Research Insitute Changsha410000 China School of Computer and Communication Engineering Changsha University of Science and Technology Changsha410114 China School of Software Engineering Tongji University Shanghai201804 China

ISBN: (数字)9783319271194

ISBN: (纸本)9783319271187

Due to the strict energy limitation and the common vulnerability of WSNs, providing efficient and security data gathering in WSNs becomes an essential problem. Compressive data gathering, which is based on the recent breakthroughs in compressive sensing theory, has been proposed as a viable approach for data gathering in WSNs at low communication overhead. Nevertheless, compressive data gathering is susceptible to various attacks due to the open wireless medium. To thwart traffic analysis/flow tracing and realize privacy preservation, this paper proposes a novel Efficient Privacy-Preserving Compressive Data Gathering Scheme which exploits homomorphic encryption functions in compressive data gathering. With homomorphic encryption on the compressive sensing encoded sensory reading messages, the proposed scheme offers two significant privacy-preserving features, message flow untraceability and message content confidentiality, for efficiently thwarting the traffic analysis attacks. Extensive performance evaluations and security analysis demonstrate the validity and efficiency of the proposed scheme. © Springer International Publishing Switzerland 2015.

关键词： Compressed sensing

Deep Reinforcement Learning based Green Resource Allocation Mechanism in Edge computing driven Power Internet of Things 16

学校读者我要写书评

暂无评论

Deep Reinforcement Learning based Green Resource Allocation ...

16th IEEE International Wireless Communications and Mobile computing Conference, IWCMC 2020

作者： Yang, Mo Yu, Peng Wang, Ying Huang, Xiuli Miu, Weiwei Yu, Pengfei Li, Wei Yang, Ruxia Tao, Minxing Shi, Lei State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing100876 China Global Energy Interconnection Research Institute Co. Ltd Beijing China Communication Branch State Grid Jiangsu Electric Power Co. Ltd Jiangsu China School of Computer Science Beijing University of Posts and Telecommunications Beijing China Department of Computing Carlow Institute of Technology Carlow Ireland

ISBN: (纸本)9781728131290

Smart grid deploys a large number of smart terminals and sensing devices to form an edge network, as well as a virtual network of information space and the power Internet of Things. As a key component of 5G and future network, the latency of end-to-end and the traffic of backhaul link could be reduced by edge network. Nevertheless, the function of storage and computing are moved down to the edge nodes in mobile edge network which increases the complexity of resource management. So it is an important issue to find out a more effectively resources allocation mechanism as well as meeting the requirements of each user. Edge computing refers to the processing of large amounts of edge data in the edge space in the edge network, thereby reducing dependence on the data center, achieving limited self-governance of the edge network, and reducing off-line threats. Although Deep Reinforcement Learning (DRL) has been applied to many of the work related to edge networks, there lacks the applications for green resource allocation. A Deep Reinforcement Learning (DRL) based green resource allocation mechanism is proposed in this paper which aims at efficiently allocating the resources while satisfying the needs of mobile users. The value of energy efficiency can be obtained when the algorithm achieves convergence according to the simulation results. The efficiency of the DRL-based mechanism and its effectiveness in meeting user requirements and implementing green resource allocation are validated. © 2020 IEEE.

关键词： Internet of things

On the Effectiveness of Function-Level Vulnerability Detectors for Inter-Procedural Vulnerabilities

学校读者我要写书评

暂无评论

On the Effectiveness of Function-Level Vulnerability Detecto...

International Conference on Software Engineering (ICSE)

作者： Zhen Li Ning Wang Deqing Zou Yating Li Ruqian Zhang Shouhuai Xu Chao Zhang Hai Jin National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security Cluster and Grid Computing Lab School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China Jin YinHu Laboratory Wuhan China Department of Computer Science University of Colorado Colorado Springs Colorado Springs Colorado USA Institute for Network Sciences and Cyberspace Tsinghua University Beijing China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9798400702174

ISBN: (纸本)9798350382143

Software vulnerabilities are a major cyber threat and it is important to detect them. One important approach to detecting vulnerabilities is to use deep learning while treating a program function as a whole, known as function-level vulnerability detectors. However, the limitation of this approach is not understood. In this paper, we investigate its limitation in detecting one class of vulnerabilities known as inter-procedural vulnerabilities, where the to-be-patched statements and the vulnerability-triggering statements belong to different functions. For this purpose, we create the first Inter -Procedural Vulnerability Dataset (InterPVD) based on C/C++ open-source software, and we propose a tool dubbed VulTrigger for identifying vulnerability-triggering statements across functions. Experimental results show that VulTrigger can effectively identify vulnerability-triggering statements and inter-procedural vulnerabilities. Our findings include: (i) inter-procedural vulnerabilities are prevalent with an average of 2.8 inter-procedural layers; and (ii) function-level vulner-ability detectors are much less effective in detecting to-be-patched functions of inter-procedural vulnerabilities than detecting their counterparts of intra-procedural vulnerabilities.

关键词： Deep learning Detectors Open source software Software engineering

ECLIPSE: Expunging Clean-label Indiscriminate Poisons via Sparse Diffusion Purification

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Xianlong Hu, Shengshan Zhang, Yechao Zhou, Ziqi Zhang, Leo Yu Xu, Peng Wan, Wei Jin, Hai National Engineering Research Center for Big Data Technology and System China Services Computing Technology and System Lab China Cluster and Grid Computing Lab China Hubei Engineering Research Center on Big Data Security China Hubei Key Laboratory of Distributed System Security China School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan430074 China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China School of Information and Communication Technology Griffith University SouthportQLD4215 Australia

Clean-label indiscriminate poisoning attacks add invisible perturbations to correctly labeled training images, thus dramatically reducing the generalization capability of the victim models. Recently, defense mechanisms such as adversarial training, image transformation techniques, and image purification have been proposed. However, these schemes are either susceptible to adaptive attacks, built on unrealistic assumptions, or only effective against specific poison types, limiting their universal applicability. In this research, we propose a more universally effective, practical, and robust defense scheme called ECLIPSE. We first investigate the impact of Gaussian noise on the poisons and theoretically prove that any kind of poison will be largely assimilated when imposing sufficient random noise. In light of this, we assume the victim has access to an extremely limited number of clean images (a more practical scene) and subsequently enlarge this sparse set for training a denoising probabilistic model (a universal denoising tool). We then introduce Gaussian noise to absorb the poisons and apply the model for denoising, resulting in a roughly purified dataset. Finally, to address the trade-off of the inconsistency in the assimilation sensitivity of different poisons by Gaussian noise, we propose a lightweight corruption compensation module to effectively eliminate residual poisons, providing a more universal defense approach. Extensive experiments demonstrate that our defense approach outperforms 10 state-of-the-art defenses. We also propose an adaptive attack against ECLIPSE and verify the robustness of our defense scheme. Our code is available at https://***/CGCL-codes/ECLIPSE. Copyright © 2024, The Authors. All rights reserved.

关键词： Deep neural networks

Decentralized Task Offloading in Edge computing: A Multi-User Multi-Armed Bandit Approach

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Wang, Xiong Ye, Jiancheng Lui, John C.S. National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China Network Technology Lab Hong Kong Research Center Huawei Technologies Co. Ltd. Hong Kong Department of Computer Science and Engineering The Chinese University of Hong Kong Hong Kong

Mobile edge computing facilitates users to offload computation tasks to edge servers for meeting their stringent delay requirements. Previous works mainly explore task offloading when system-side information is given (e.g., server processing speed, cellular data rate), or centralized offloading under system uncertainty. But both generally fall short to handle task placement involving many coexisting users in a dynamic and uncertain environment. In this paper, we develop a multi-user offloading framework considering unknown yet stochastic system-side information to enable a decentralized user-initiated service placement. Specifically, we formulate the dynamic task placement as an online multi-user multi-armed bandit process, and propose a decentralized epoch based offloading (DEBO) to optimize user rewards which are subjected under network delay. We show that DEBO can deduce the optimal user-server assignment, thereby achieving a close-to-optimal service performance and tight O(log T) offloading regret. Moreover, we generalize DEBO to various common scenarios such as unknown reward gap, dynamic entering or leaving of clients, and fair reward distribution, while further exploring when users' offloaded tasks require heterogeneous computing resources. Particularly, we accomplish a sub-linear regret for each of these instances. Real measurements based evaluations corroborate the superiority of our offloading schemes over state-of-the-art approaches in optimizing delay-sensitive rewards. Copyright © 2021, The Authors. All rights reserved.

关键词： Stochastic systems