检索结果-内蒙古大学图书馆

Patch is enough: naturalistic adversarial patch against vision-language pre-training models

Visual Intelligence 2024年第1期2卷 1-10页

作者： Kong, Dehong Liang, Siyuan Zhu, Xiaopeng Zhong, Yuansheng Ren, Wenqi School of Cyber Science and Technology Shenzhen Campus of Sun Yat-sen University Shenzhen 518107 China The State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou 310058 China School of Computing National University of Singapore Singapore 119077 Singapore Guangdong Testing Institute of Product Quality Supervision Guangzhou 510670 China

Visual language pre-training (VLP) models have demonstrated significant success in various domains, but they remain vulnerable to adversarial attacks. Addressing these adversarial vulnerabilities is crucial for enhancing security in multi-modal learning. Traditionally, adversarial methods that target VLP models involve simultaneous perturbation of images and text. However, this approach faces significant challenges. First, adversarial perturbations often fail to translate effectively into real-world scenarios. Second, direct modifications to the text are conspicuously visible. To overcome these limitations, we propose a novel strategy that uses only image patches for attacks, thus preserving the integrity of the original text. Our method leverages prior knowledge from diffusion models to enhance the authenticity and naturalness of the perturbations. Moreover, to optimize patch placement and improve the effectiveness of our attacks, we utilize the cross-attention mechanism, which encapsulates inter-modal interactions by generating attention maps to guide strategic patch placement. Extensive experiments conducted in a white-box setting for image-to-text scenarios reveal that our proposed method significantly outperforms existing techniques, achieving a 100% attack success rate. © The Author(s) 2024.

关键词： Adversarial Patch Diffusion Model Naturalistic Physical Attack

来源：评论

学校读者我要写书评

暂无评论

TCS security Analysis in Intel SGX Enclave MultiThreading

TCS Security Analysis in Intel SGX Enclave MultiThreading

引用

2022 International Conference on Networking and Network Applications, NaNA 2022

作者： Zhang, Tong Cui, Xiangjie Wang, Yichuan Du, Yanning Gao, Wen School of Computer Science and Engineering Xi'an University of Technology Xi'an China Shaanxi Key Laboratory for Network Computing and Security Technology China

ISBN: (数字)9781665461313

ISBN: (纸本)9781665461313

With the rapid development of Internet Technology in recent years, the demand for security support for complex applications is becoming stronger and stronger. Intel Software Guard Extensions (Intel SGX) is created as an extension of Intel Systems to enhance software security. Intel SGX allows application developers to create so-called enclave. Sensitive application code and data are encapsulated in Trusted Execution Environment (TEE) by enclave. TEE is completely isolated from other applications, operating systems, and administrative programs. Enclave is the core structure of Intel SGX Technology. Enclave supports multi-threading. Thread Control Structure (TCS) stores special information for restoring enclave threads when entering or exiting enclave. Each execution thread in enclave is associated with a TCS. This paper analyzes and verifies the possible security risks of enclave under concurrent conditions. It is found that in the case of multithread concurrency, a single enclave cannot resist flooding attacks, and related threads also throw TCS exception codes. © 2022 IEEE.

关键词： Floods

来源：评论

学校读者我要写书评

暂无评论

GoPIM: GCN-Oriented Pipeline Optimization for PIM Accelerators

GoPIM: GCN-Oriented Pipeline Optimization for PIM Accelerato...

引用

IEEE Symposium on High-Performance Computer Architecture

作者： Siling Yang Shuibing He Wenjiong Wang Yanlong Yin Tong Wu Weijian Chen Xuechen Zhang Xian-He Sun Dan Feng The State Key Laboratory of Blockchain and Data Security Zhejiang University Zhejiang Lab Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security Zhejiang Key Laboratory of Big Data Intelligent Computing Washington State University Vancouver Illinois Institute of Technology Huazhong University of Science and Technology Wuhan National Laboratory for Optoelectronics

ISBN: (数字)9798331506476

ISBN: (纸本)9798331506483

Graph convolutional networks (GCNs) are popular for a variety of graph learning tasks. ReRAM-based processing-in-memory (PIM) accelerators are promising to expedite GCN training owing to their in-situ computing capability. However, existing accelerators can be severely underutilized even with pipelines, due to the oversight of the skewed execution times of various GCN stages and the ignorance of skewed degrees of graph vertices. In this work, we propose GOPIM, a GCN-oriented pipeline optimization for PIM accelerators to expedite GCN training. First, GOPIM proposes an ML-based scheme that allocates crossbar resources to the most needed stages to streamline the overall pipeline. Second, GOPIM utilizes a selective vertex updating technique that evenly distributes vertices on crossbars by interleaved mapping. These techniques collectively reduce the overall execution time without losing much accuracy. We also provide a practical architecture design for GOPIM. Our experimental results show that, GoPIM achieves up to 191 × speedup and 16.1 × energy saving, compared to the state-of-the-art work.

关键词： Training Accuracy Graph convolutional networks Pipelines Energy conservation Computer architecture Machine learning Parallel processing Optimization

来源：评论

学校读者我要写书评

暂无评论

Non-IID Federated Learning Based on Global Knowledge Sharing Without Out-of-Domain Data 8

Non-IID Federated Learning Based on Global Knowledge Sharing...

引用

8th IEEE International Conference on Network Intelligence and Digital Content, IC-NIDC 2023

作者： Zhang, Yufei Xu, Tianxing Dong, Jin Tang, Lin Andrzejak, Artur Yao, Zelei Zhu, Konglin Beijing Academy of Blockchain and Edge Computing China Heidelberg University Heidelberg69120 Germany School of Artificial Intelligence Beijing University of Posts and Telecommunications Beijing100876 China Guangdong Provincial Key Laboratory of Novel Security Inteligence Technologies China

ISBN: (纸本)9798350317923

Federated learning (FL) supports collaborative training of deep learning models on distributed data. The non-independent and identically distribution (non-IID) of user data brings great challenges to FL, which may cause the global model to converge slowly or even deviate from the optimal solution. As an effective method to solve the data non-IID problem, knowledge distillation often relies on public proxy datasets for model training, which violates the original intention of FL to protect privacy. In this paper, we propose a global knowledge sharing algorithm without out-of-domain data to solve FL with non-IID data. Firstly, to ensure that the data remains within the domain and to constrain the deviation of update directions between the local and global models, we propose a novel teacher-student network structure, which utilizes a personalized global knowledge transfer method to supervise local training. Secondly, we also design different knowledge transfer loss functions to maximize the use of global knowledge and prevent overfitting of local data. Finally, in order to continuously correct the update direction of the local model and improve the accuracy of the global model, the shared knowledge model on the client is updated using the exponential moving average (EMA) method. In order to evaluate, we manually constructed the non-IID dataset in the real scene. Experiments show that compared with the existing methods, the global model accuracy of the SKH-FL algorithm proposed in this paper is improved by 5.07%. © 2023 IEEE.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

TPIA: Towards Target-specific Prompt Injection Attack against Code-oriented Large Language Models

arXiv

引用

arXiv 2024年

作者： Yang, Yuchen Yao, Hongwei Yang, Bingrun He, Yiling Li, Yiming Zhang, Tianwei Qin, Zhan State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou310007 China Institute of Blockchain and Data Security Hangzhou310053 China College of Computing and Data Science Nanyang Technological University Singapore639798 Singapore

Recently, code-oriented large language models (Code LLMs) have been widely and successfully exploited to simplify and facilitate programming. Unfortunately, a few pioneering works revealed that these Code LLMs are vulnerable to backdoor and adversarial attacks. The former poisons the training data or model parameters, hijacking the LLMs to generate malicious code snippets when encountering the trigger. The latter crafts malicious adversarial input codes to reduce the quality of the generated codes. In this paper, we reveal that both attacks have some inherent limitations: backdoor attacks rely on the adversary’s capability of controlling the model training process, which may not be practical;adversarial attacks struggle with fulfilling specific malicious purposes. To alleviate these problems, this paper presents a novel attack paradigm against Code LLMs, namely target-specific prompt injection attack (TPIA). TPIA generates non-functional perturbations containing the information of malicious instructions and inserts them into the victim’s code context by spreading them into potentially used dependencies (e.g., packages or RAG’s knowledge base). It induces the Code LLMs to generate attacker-specified malicious code snippets at the target location. In general, we compress the attacker-specified malicious objective into the perturbation by adversarial optimization based on greedy token search. We collect 13 representative malicious objectives to design 31 threat cases for three popular programming languages. We show that our TPIA can successfully attack three representative open-source Code LLMs (with an attack success rate of up to 97.9%) and two mainstream commercial Code LLM-integrated applications (with an attack success rate of over 90%) in all threat cases, using only a 12-token non-functional perturbation. Copyright © 2024, The Authors. All rights reserved.

关键词： Open source software

来源：评论

学校读者我要写书评

暂无评论

A Multi-Exposure Generation and Fusion Method for Low-Light Image Enhancement

A Multi-Exposure Generation and Fusion Method for Low-Light ...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Haiyan Jin Long Li Haonan Su YuanLin Zhang ZhaoLin Xiao Bin Wang Faculty of Computer Science and Engineering Xi’an University of Technology Shaanxi Key Laboratory for Network Computing and Security Technology

ISBN: (数字)9798350359312

ISBN: (纸本)9798350359329

In the low light image enhancement, single exposure images contains a limited dynamic range, which hinders the restoration of contrast and texture. To address these problems, we propose a multi exposure generation and fusion method (MEGF) which simulates multi exposure images and perform feature fusion and enhancement on these images. First, we propose a Multi-Exposure Generation (MEG) method, which constructs the Gaussian Distribution for each exposure level based on multi exposure datasets. MEG can generate images with different exposure levels based on the constructed distribution. Then, the Perceptual Importance based Multi-Exposure Feature Enhancement (PIMEFE) block is developed to fuse the feature of generated multi exposure images using VGG-16 network. Before fusion, the generated images are input to Multi Scale Recursive Feature Enhancement (MSRFE) blocks and obtain the denoised and enhanced features. Finally, the fused feature are input to Curve Adjustment (CA) block for fine tuning and provide the color enhancement on fusion features. We propose the Multiple Exposure Recursive Fusion (MERF) block which estimates the adjusting factors for CA block. Experimental results demonstrate that our method outperforms other techniques in both subjective and objective evaluations on real and synthetic datasets.

关键词： Image color analysis Fuses Neural networks Gaussian distribution Dynamic range Image restoration Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Prediction of piRNA-mRNA interactions based on an interactive inference network

Prediction of piRNA-mRNA interactions based on an interactiv...

引用

2023 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2023

作者： Liu, Yajun Li, Ru Li, Aimin Fei, Rong Guo, Xie Wu, Fang-Xiang Xi'an University of Technology Shaanxi Key Laboratory for Network Computing and Security Technology Xi'an China Xi'an University of Technology Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing Xi'an China University of Saskatchewan Division of Biomedical Engineering Department of Computer Science Department of Mechanical Engineering Saskatoon Canada

ISBN: (纸本)9798350337488

As the largest class of small non-coding RNAs, piRNAs primarily present in the reproductive cells of mammals, which influence post-transcriptional processes of mRNAs in multiple ways. Effective methods for predicting piRNA and mRNA target relationships can help identify piRNA functions, investigate the possibility of piRNAs as biomarkers and therapeutic targets. In this study, we propose a computational approach for classifying the relationships of piRNA-mRNA pairs based on an interactive inference network (IIN). First, we gather piRNA-mRNA target data, collect sequence data by position alignment, and construct a benchmark dataset. Furthermore, a reliable negative set is constructed by positive-unlabeled learning. Finally, we view a piRNA and a mRNA sequence as a premise and hypothesis sentence, respectively, and IIN model is used to predict the relationship between them. The experiments demonstrate that our method effectively characterizes piRNA-mRNA interaction and could be beneficial for researchers to investigate piRNA functions. © 2023 IEEE.

关键词： interactive inference network noncoding RNA piRNA-mRNA interactions positive-unlabeled learning

来源：评论

学校读者我要写书评

暂无评论

Adaptive Compression-Encryption Scheme for Medical Image Based on Improved Compressive Sensing and Deoxyribonucleic Acid Coding-Compression

Adaptive Compression-Encryption Scheme for Medical Image Bas...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Xue, Xianglian Jin, Haiyan Zhou, Changjun Xi'an University of Technology School of Computer Science and Engineering Xi'an China Shaanxi University of Chinese Medicine Section of Computer Teaching and Research Xianyang China Shaanxi Key Laboratory for Network Computing and Security Technology Xi'an China Zhejiang Normal University School of Computer Science and Technology Jinhua China

ISBN: (纸本)9798350386226

Medical images often occupy large storage space and contain patient privacy or sensitive information, which makes them difficult and unsafe to be transmitted through the network. This paper proposed an adaptive compression-encryption scheme based on improved compressive sensing (CS) and deoxyribonucleic acid (DNA) coding-compression, which helps to solve the above problem. In the proposed scheme, first the original medical image was compressed to a floating point CS matrix using the discrete wavelet transform and the partial Hadamard matrix. The CS matrix was then normalized and rounded for matrix quantification. The result of quantification was fed to DNA fixed encoding and DNA run length coding for encryption and secondary compression. Finally the compressed-encrypted image was obtained after DNA dynamic encoding-decoding and regroups operation. The proposed scheme was tested against 9 images and proved to be effective in reducing the quantization errors and enhancing the compression performance. For instance, when the benchmark compression ratio (CR) is 0.5, the CR can be reduced by 5%(CR = 0.4453) ∼ 23%(CR = 0.2636), with the corresponding peak signal-to-noise ratio values consistently surpassing the benchmark. Furthermore, the execution of DNA compression and DNA dynamic encoding-decoding provided double guarantee for the algorithm's security. © 2024 IEEE.

关键词： Encoding (symbols)

来源：评论

学校读者我要写书评

暂无评论

FIT-Print: Towards False-claim-resistant Model Ownership Verification via Targeted Fingerprint

arXiv

引用

arXiv 2025年

作者： Shao, Shuo Zhu, Haozhe Yao, Hongwei Li, Yiming Zhang, Tianwei Qin, Zhan Ren, Kui State Key Laboratory of Blockchain and Data Security Zhejiang University China Institute of Blockchain and Data Security China Department of Computer Science City University of Hong Kong Hong Kong College of Computing and Data Science Nanyang Technological University Singapore

Model fingerprinting is a widely adopted approach to safeguard the intellectual property rights of open-source models by preventing their unauthorized reuse. It is promising and convenient since it does not necessitate modifying the protected model. In this paper, we revisit existing fingerprinting methods and reveal that they are vulnerable to false claim attacks where adversaries falsely assert ownership of any third-party model. We demonstrate that this vulnerability mostly stems from their untargeted nature, where they generally compare the outputs of given samples on different models instead of the similarities to specific references. Motivated by these findings, we propose a targeted fingerprinting paradigm (i.e., FIT-Print) to counteract false claim attacks. Specifically, FIT-Print transforms the fingerprint into a targeted signature via optimization. Building on the principles of FIT-Print, we develop bit-wise and list-wise black-box model fingerprinting methods, i.e., FIT-ModelDiff and FIT-LIME, which exploit the distance between model outputs and the feature attribution of specific samples as the fingerprint, respectively. Extensive experiments on benchmark models and datasets verify the effectiveness, conferrability, and resistance to false claim attacks of our FIT-Print. © 2025, CC BY.

关键词： Open systems

来源：评论

学校读者我要写书评

暂无评论

DCGSD: Low-Light Image Enhancement with Dual-Conditional Guidance Sparse Diffusion Model

引用

IEEE Transactions on Circuits and Systems for Video Technology 2025年

作者： Jin, Haiyan Wang, Jing Zuo, Fengyuan Su, Haonan Xiao, Zhaolin Wang, Bin Zhang, Yuanlin Xi'an University of Technology Shaanxi Key Laboratory for Network Computing and Security Technology Department of Computer Science and Engineering NO. 5 South Jinhua Road Shaanxi Xi'an710048 China

When restoring low-light images, most methods largely overlook the ambiguity due to dark noise and lack discrimination for region and shape representations, resulting in invalid feature enhancement. In this work, we propose a physically explainable and prior guidance model for low-light image enhancement, termed Dual-Conditional Guidance Sparse Diffusion (DCGSD). Specifically, we introduce an elaborately designed Luminance Structure Guidance Head, which can be easily plugged into the existing diffusion model to emphasize the value of the luminance and structural representation. Furthermore, for reliable noise analysis, we provide a novel Sparse Attention Enhancement Module that is adaptively empowered to exploit the most useful region-to-region dependencies. This dynamic selection makes the diffusion process from dense to sparse, thus improving the efficiency of the reasoning noise distributions. To avoid noise amplification, we further present a Skip Calibration Module, which can be used to refine the local neighborhood that contains noisy and structural information. Extensive experiments have been performed to verify the superiority of the proposed method. DCGSD shows that leveraging dual-conditional guidance can support the diffusion model to produce sharper and more realistic results. © 1991-2012 IEEE.

关键词： Feature extraction

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：