检索结果-南通市图书馆

FTG: Score-based black-box watermarking by fragile trigger generation for deep model integrity verification

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Information and Intelligence

Journal of Information and Intelligence 2024年第1期2卷 28-41页

作者： Heng Yin Zhaoxia Yin Zhenzhe Gao Hang Su xinpeng zhang Bin Luo Anhui Provincial Key Laboratory of Multimodal Cognitive Computation Anhui UniversityHefei 230000China Shanghai key Laboratory of Multidimensional Information Processing East China Normal UniversityShanghai 200241China School of Communication and Electronic Engineering East China Normal UniversityShanghai 200241China Department of Computer Science and Technology Tsinghua UniversityBeijing 100000China School of Computer Science and Technology Fudan UniversityShanghai 200241China

Deep neural networks(DNNs)are widely used in real-world applications,thanks to their exceptional performance in image recognition.However,their vulnerability to attacks,such as Trojan and data poison,can compromise the integrity and stability of DNN applications.Therefore,it is crucial to verify the integrity of DNN models to ensure their security.Previous research on model watermarking for integrity detection has encountered the issue of overexposure of model parameters during embedding and extraction of the watermark.To address this problem,we propose a novel score-based black-box DNN fragile watermarking framework called fragile trigger generation(FTG).The FTG framework only requires the prediction probability distribution of the final output of the classifier during the watermarking process.It generates different fragile samples as the trigger,based on the classification prediction probability of the target classifier and a specified prediction probability mask to watermark it.Different prediction probability masks can promote the generation of fragile samples in corresponding distribution types.The whole watermarking process does not affect the performance of the target classifier.When verifying the watermarking information,the FTG only needs to compare the prediction results of the model on the samples with the previous label.As a result,the required model parameter information is reduced,and the FTG only needs a few samples to detect slight modifications in the model.Experimental results demonstrate the effectiveness of our proposed method and show its superiority over related work.The FTG framework provides a robust solution for verifying the integrity of DNN models,and its effectiveness in detecting slight modifications makes it a valuable tool for ensuring the security and stability of DNN applications.

关键词： Neural network Sensitive sample Fragile watermarking Model integrity Fragile trigger

Influence of nano-mechanical evolution of Ti_(3)AlC_(2) ceramic on the arc erosion resistance of Ag-based composite electrical contact material

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Advanced Ceramics 2024年第2期13卷 176-188页

作者： Xuelian Wu Chengzhe Wu xinpeng Wei Wanjie Sun Chengjian Ma Yundeng zhang Gege Li Liming Chen Dandan Wang Peigen zhang Zhengming Sun Jianxiang Ding School of Materials Science and Engineering Anhui University of TechnologyMa’anshan 243032China Analytical and Testing Center Yancheng Institute of TechnologyYancheng 224051China School of Materials Science and Engineering Changzhou UniversityChangzhou 213164China Jiangsu Key Laboratory of Advanced Metallic Materials School of Materials Science and EngineeringSoutheast UniversityNanjing 211189China

Al-containing MAX phase ceramic has demonstrated great potential in the field of high-performance low-voltage electrical contact material.Elucidating the anti-arc erosion mechanism of the MAX phase is crucial for further improving performance,but it is not well-understood.In this study,Ag/Ti_(3)AlC_(2) electrical contact material was synthesized by powder metallurgy and examined by nanoindentation techniques such as constant loading rate indentation,creep testing,and continuous stiffness measurements.Our results indicated a gradual degradation in the nano-mechanical properties of the Ti_(3)AlC_(2) reinforcing phase with increasing arc erosion times,although the rate of this degradation appeared to decelerate over arc erosion times.Specifically,continuous stiffness measurements highlighted the uneven mechanical properties within Ti_(3)AlC_(2),attributing this heterogeneity to the phase’s decomposition.During the early(1-100 times)and intermediate(100-1000 times)stages of arc erosion,the decline in the nano-mechanical properties of Ti_(3)AlC_(2) was primarily ascribed to the decomposition of Ti_(3)AlC_(2) and limited surface oxidation.During the later stage of arc erosion(1000-6200 times),the inner region of Ti_(3)AlC_(2) also sustained arc damage,but a thick oxide layer formed on its surface,enhancing the mechanical properties and overall arc erosion resistance of the Ag/Ti_(3)AlC_(2).

关键词： MAX phase nano-mechanical properties decomposition surface oxidation anti-arc erosion mechanism

Adaptive Backdoor Attack against Deep Neural Networks

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computer Modeling in Engineering & Sciences 2023年第9期136卷 2617-2633页

作者： Honglu He Zhiying Zhu xinpeng zhang School of Computer Science Fudan UniversityShanghai200433China

In recent years,the number of parameters of deep neural networks(DNNs)has been increasing rapidly.The training of DNNs is typically computation-intensive.As a result,many users leverage cloud computing and outsource their training procedures.Outsourcing computation results in a potential risk called backdoor attack,in which a welltrained DNN would performabnormally on inputs with a certain trigger.Backdoor attacks can also be classified as attacks that exploit fake images.However,most backdoor attacks design a uniformtrigger for all images,which can be easilydetectedand removed.In this paper,we propose a novel adaptivebackdoor attack.We overcome this defect and design a generator to assign a unique trigger for each image depending on its texture.To achieve this goal,we use a texture complexitymetric to create a specialmask for eachimage,which forces the trigger tobe embedded into the rich texture regions.The trigger is distributed in texture regions,which makes it invisible to humans.Besides the stealthiness of triggers,we limit the range of modification of backdoor models to evade detection.Experiments show that our method is efficient in multiple datasets,and traditional detectors cannot reveal the existence of a backdoor.

关键词： Backdoor attack AI security DNN

Fine-Grained and Fair Identity Authentication Scheme for Mobile Networks Based on Blockchain

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

China Communications 2022年第6期19卷 35-49页

作者： Mengjuan Zhai Yanli Ren Guorui Feng xinpeng zhang School of Communication and Information Engineering Shanghai UniversityShanghai 200444China Zhengzhou Xinda Institute of Advanced Technology Zhengzhou 450001HenanChina

With the popularity of the internet,users hope to better protect their privacy while obtaining network services.However,in the traditional centralized authentication scheme,identity information such as the user's private key is generated,stored,and managed by the network operator.Users can't control their identity information,which will lead to a great threat to the privacy of users.Based on redactable blockchain,we propose a fine-grained and fair identity authentication scheme for mobile networks.In our proposed scheme,the user's identity information is generated and controlled by the users.We first propose a notion of score chameleon hash(SCH),which can delete or update the information of illegal users so as to dynamically update the status of users and provide users with more fine-grained and fair services.We propose another notion of self-updating secret sharing(SUSS),which allows users to update the trapdoor and the corresponding hash key after redacting the blockchain without requiring trusted authority to redistribute the trapdoor.Experimental results show that,compared with the immutable blockchain Bitcoin,the redactable blockchain in our identity authentication scheme provides users with fine-grained and fair redacting functions,and can be adopted with a small additional overhead.

关键词： identity authentication secret sharing chameleon hash redactable blockchain

维普期刊数据库

Robust Image Watermarking Based on Generative Adversarial Network

在线全文

学校读者我要写书评

暂无评论

China Communications 2020年第11期17卷 131-140页

作者： Kangli Hao Guorui Feng xinpeng zhang School of Communication and Information Engineering Shanghai UniversityShanghai 200444China

Digital watermark embeds information bits into digital cover such as images and videos to prove the creator’s ownership of his work.In this paper,we propose a robust image watermark algorithm based on a generative adversarial network.This model includes two modules,generator and adversary.Generator is mainly used to generate images embedded with watermark,and decode the image damaged by noise to obtain the watermark.Adversary is used to discriminate whether the image is embedded with watermark and damage the image by noise.Based on the model Hidden(hiding data with deep networks),we add a high-pass filter in front of the discriminator,making the watermark tend to be embedded in the mid-frequency region of the image.Since the human visual system pays more attention to the central area of the image,we give a higher weight to the image center region,and a lower weight to the edge region when calculating the loss between cover and embedded image.The watermarked image obtained by this scheme has a better visual performance.Experimental results show that the proposed architecture is more robust against noise interference compared with the state-of-art schemes.

关键词： robust image watermark deep learning generative adversarial network convolutional neural network

维普期刊数据库

Improved Lossless Data Hiding for JPEG Images Based on Histogram Modification

在线全文

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2018年第6期55卷 495-507页

作者： Yang Du Zhaoxia Yin xinpeng zhang Key Laboratory of Intelligent Computing&Signal Processing Ministry of EducationAnhui UniversityHefei230601P.R.China Department of Computer Science Purdue UniversityWest lafayette47906USA School of Communication and Information Engineering Shanghai UniversityShanghai 200072P.R.China

This paper proposes a lossless and high payload data hiding scheme for JPEG images by histogram modification.The most in JPEG bitstream consists of a sequence of VLCs(variable length codes)and the appended bits.Each VLC has a corresponding RLV(run/length value)to record the AC/DC coefficients.To achieve lossless data hiding with high payload,we shift the histogram of VLCs and modify the DHT segment to embed data.Since we sort the histogram of VLCs in descending order,the filesize expansion is limited.The paper’s key contribution includes:Lossless data hiding,less filesize expansion in identical pay-load and higher embedding efficiency.

关键词： Lossless data hiding histogram VLC JPEG.

Road map for, and technical challenges of, carbon-nanotube integrated circuit technology

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

National Science Review 2024年第3期11卷 14-16页

作者： Jia Si Panpan zhang Zhiyong zhang Key Laboratory for the Physics and Chemistry of Nanodevices and Center for Carbon-based Electronics School of Electronics Peking University State Key Laboratory of Information Photonics and Optical Communications Beijing University of Posts and Telecommunications

The past decade has witnessed an explosion of research on carbon-nanotube(CNT) electronics due to their outstanding properties, including an ultrathin body, a high carrier mobility and a high saturation velocity.

关键词： TRANSISTORS

New structure transistors for advanced technology node CMOS ICs

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

National Science Review 2024年第3期11卷 24-41页

作者： Qingzhu zhang Yongkui zhang Yanna Luo Huaxiang Yin Integrated Circuit Advanced Process R&D Center Institute of Microelectronics of Chinese Academy of Sciences(IMECAS) State key Lab of Fabrication Technologies for Integrated Circuits Institute of Microelectronics of Chinese Academy of Sciences School of Integrated Circuits University of Chinese Academy of Sciences

Over recent decades, advancements in complementary metal-oxide-semiconductor integrated circuits(ICs)have mainly relied on structural innovations in transistors. From planar transistors to the fin field-effect transistor(FinFET) and gate-all-around FET(GAAFET), more gate electrodes have been added to three-dimensional(3D) channels with enhanced control and carrier conductance to provide higher electrostatic integrity and higher operating currents within the same device footprint. Beyond the 1-nm node, Moore's law scaling is no longer expected to be applicable to geometrical shrinkage. Vertical transistor stacking, e.g. in complementary FETs(CFET), 3D stack(3DS) FETs and vertical-channel transistors(VFET), for enhanced density and variable circuit or system design represents a revolutionary scaling approach for sustained IC development. Herein, innovative works on specific structures, key process breakthroughs, shrinking cell sizes and design methodologies for transistor structure research and development are reviewed. Perspectives on future innovations in advanced transistors with new channel materials and operating theories are also discussed.

关键词： transistor CMOS GAAFET CFET 3DS FET

Advances in methane emissions from agricultural sources:Part I.Accounting and mitigation

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Journal of Environmental Sciences 2024年第6期140卷 279-291页

作者： Xiaokun Wu Ying zhang Yinghui Han Yagang zhang Yuhang zhang Xiaodan Cheng Pei Zhong Xue Yuan Yuanxun zhang Zhengqiang Li State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources North China Electric Power UniversityBeijing 102206China Hebei Key Laboratory of Physics and Energy Technology North China Electric Power UniversityBaoding 071003China College of Resources and Environment University of Chinese Academy of SciencesBeijing 101408China Interdisciplinary Mathematics Institute University of South CarolinaColumbiaSC 29208United States Aerospace Information Research Institute Chinese Academy of SciencesBeijing 100094China

Methane is one of the major greenhouse gases(GHGs)and agriculture is recognized as its primary emitter.Methane accounting is a prerequisite for developing effective agriculture mitigation strategies.In this review,methane accounting methods and research status for various agricultural emission source including rice fields,animal enteric fermentation and livestock and poultry manure management were overview,and the influencing factors of each emission source were analyzed and discussed.At the same time,it analyzes the different research efforts involving agricultural methane accounting and makes recommendations based on the actual situation.Finally,mitigation strategies based on accounting results and actual situation are proposed.This review aims to provide basic data and reference for agriculture-oriented countries and regions to actively participate in climate action and carry out effective methane emission mitigation.

关键词： Methane emission Agricultural sources Accounting methodology Mitigation Emission influencing factors

维普期刊数据库