检索结果-内蒙古大学图书馆

Layer-Wise Learning Rate Optimization for Task-Dependent Fine-Tuning of Pre-Trained Models: An Evolutionary Approach

引用

ACM Transactions on Evolutionary Learning and Optimization 2024年第4期4卷 1-23页

作者： Bu, Chenyang Liu, Yuxin Huang, Manzong Shao, Jianxuan Ji, Shengwei Luo, Wenjian Wu, Xindong Key Laboratory of Knowledge Engineering with Big Data Ministry of Education and School of Computer Science and Information Engineering Hefei University of Technology Hefei China School of Artificial Intelligence and Big Data Hefei University Hefei China Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

The superior performance of large-scale pre-Trained models, such as Bidirectional Encoder Representations from Transformers (BERT) and Generative Pre-Trained Transformer (GPT), has received increasing attention in both academic and industrial research and has become one of the current research hotspots. A pre-Trained model refers to a model trained on large-scale unlab.led data, whose purpose is to learn general language representation or features for fine-Tuning or transfer learning in subsequent tasks. After pre-Training is complete, a small amount of lab.led data can be used to fine-Tune the model for a specific task or domain. This two-stage method of "pre-Training+fine-Tuning"has achieved advanced results in natural language processing (NLP) tasks. Despite widespread adoption, existing fixed fine-Tuning schemes that adapt well to one NLP task may perform inconsistently on other NLP tasks given that different tasks have different latent semantic structures. In this article, we explore the effectiveness of automatic fine-Tuning pattern search for layer-wise learning rates from an evolutionary optimization perspective. Our goal is to use evolutionary algorithms to search for better task-dependent fine-Tuning patterns for specific NLP tasks than typical fixed fine-Tuning patterns. Experimental results on two real-world language benchmarks and three advanced pre-Training language models show the effectiveness and generality of the proposed framework. © 2024 held by the owner/author(s).

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Improve the Robustness of Deep Learning Models against Adversarial Attacks in Judicial Field 8

Improve the Robustness of Deep Learning Models against Adver...

引用

8th International Conference on Digital Home, ICDH 2020

作者： Liao, Yongxian Zhuang, Wenzi Huang, Lifeng Liu, Ning Sun Yat-sen University School of Data and Computer Science China Guangdong Key Laboratory of Information Security Technology China

ISBN: (纸本)9781728192345

With the rapid development of deep neural networks (DNNs), they have been applied in various domains including crowd counting, object detection and digital forensic. Consequently, DNN-based forensic process plays an essential role in judicial field, which will simplify the trial procedure in real world as well as cut down the workload of criminal case judgment. Apart from manipulating the visual forensic evidence (i.e., digital images) by existing editing softwares, malicious adversarial attacks can also pose a series threat for DNN-based trial system. In general, by adding some imperceptible perturbations to generate the adversarial forensic evidence, DNN-based trial system may output totally wrong results. To this end, we propose an adversarial fine-tuned training (AFT) method to improve the robustness of deep learning models for defensing against black-box adversarial attacks in judicial process (e.g. digital forensic). Concretely, we utilize both query-based attackers and transfer-based enemies to generate a set of adversarial examples at first, and then mix up both clean and perturbed data as training set to specific fine-tuned layers of models. Base on this training manner, the generated models not only can improve the robustness to resist adversarial examples, but also reduce the computational cost since only parts of parameters need to be retrained. Extensive experiments on both two pre-trained vanilla models and three defensive models demonstrate the efficacy of our proposed method to tackle the adversarial attacks and potential value in digital court system. © 2020 IEEE.

关键词： Digital forensics

来源：评论

学校读者我要写书评

暂无评论

New results on quantum boomerang attacks

引用

Quantum information Processing 2023年第4期22卷 1-27页

作者： Zou, Hongkai Zou, Jian Luo, Yiyuan College of Computer and Data Science Fuzhou University Fuzhou China Key Lab of Information Security of Network Systems Fuzhou University Fuzhou China School of Information Sciences and Technology Huizhou University Huizhou China

At SAC 2021, Frixons et al. proposed quantum boomerang attacks that can effectively recover the keys of block ciphers in the quantum setting. Based on their work, we further consider how to quantize the generic boomerang attacks proposed by Biham et al. at FSE 2002, so as to obtain more generic quantum boomerang attacks. Similar to Frixons et al.’s work, we only consider quantum key recovery attacks in the single-key setting. With the help of some famous quantum algorithms, this paper presents two methods to convert the attacks of Biham et al. into some new quantum key recovery attacks. In order to proof our methods, we apply our new ideas to attack Serpent-256 and ARIA-196. To sum up, for Serpent-256, we give valid 9-round and 10-round quantum key recovery attacks respectively. The quantum time complexity of 9-round and 10-round of Serpent-256 is 2 115.43 and 2 126.6 respectively. Furthermore, we show a valid quantum key attack on 6-round ARIA-196 which has a time complexity of 2 89.8 with negligible memory. The time complexity of the above quantum attacks are better than the corresponding classical attacks and quantum generic key recovery attack via Grover’s algorithm. © 2023, The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature.

关键词： Quantum theory

来源：评论

学校读者我要写书评

暂无评论

FakeScope: Large Multimodal Expert Model for Transparent AI-Generated Image Forensics

arXiv

引用

arXiv 2025年

作者： Li, Yixuan Tian, Yu Huang, Yipo Lu, Wei Wang, Shiqi Lin, Weisi Rocha, Anderson College of Computing City University of Hong Kong Hong Kong School of Computer Science and Engineering Ministry of Education Key Laboratory of Information Technology Guangdong Province Key Laboratory of Information Security Technology Sun Yat-Sen University Guangzhou510006 China College of Computing and Data Science Nanyang Technological University Singapore Artificial Intelligence Lab Recod.ai University of Campinas Campinas13084-851 Brazil

The rapid and unrestrained advancement of generative artificial intelligence (AI) presents a double-edged sword: while enabling unprecedented creativity, it also facilitates the generation of highly convincing deceptive content, undermining societal trust. As image generation techniques become increasingly sophisticated, detecting synthetic images is no longer just a binary task—it necessitates interpretable, context-aware methodologies that enhance trustworthiness and transparency. However, existing detection models primarily focus on classification, offering limited explanatory insights into image authenticity. In this work, we propose FakeScope, an expert multimodal model (LMM) tailored for AI-generated image forensics, which not only identifies AI-synthetic images with high accuracy but also provides rich, interpretable, and query-driven forensic insights. To this end, we first construct FakeChain dataset that contains linguistic authenticity reasoning based on visual trace evidence, developed through a novel human-machine collab.rative framework. Building upon it, we further present FakeInstruct, the largest multimodal instruction tuning dataset containing 2 million visual instructions tailored to enhance forensic awareness in LMMs. Leveraging the knowledge of FakeInstruct, FakeScope achieves state-of-the-art performance in both closed-ended and open-ended forensic scenarios. It can distinguish synthetic images with high accuracy while offering coherent and insightful explanations, free-form discussions on fine-grained forgery attributes, and actionable enhancement strategies. Notably, despite being trained exclusively on qualitative hard lab.ls, FakeScope demonstrates remarkable zero-shot quantitative capability on detection, enabled by our proposed token-based probability estimation strategy. Furthermore, FakeScope exhibits strong generalization and in-the-wild ability, ensuring its applicability in real-world scenarios. The data, model, and demo will be publ

关键词： computer forensics

来源：评论

学校读者我要写书评

暂无评论

Design a Proof of Stake Based Directed Acyclic Graph Chain 3rd

Design a Proof of Stake Based Directed Acyclic Graph Chain

引用

3rd International Conference on Frontiers in Cyber security, FCS 2020

作者： Tian, Haibo Lin, Huizhi Zhang, Fangguo Guangdong Province Key Laboratory of Information Security Technology School of Data and Computer Science Sun Yat-Sen University GuangzhouGuangdong510275 China

ISBN: (纸本)9789811597381

The concept of blockchain comes from the Bitcoin system where transactions are organized in blocks. However, blocks are not necessary. Researchers have found ways to use a directed acyclic graph (DAG) to build a chain without blocks. Currently, these chains still rely on some kinds of proof of work, which is not environmentally friendly. We here design a DAG chain with a proof of stake for independent nodes in the Internet. We split nodes into two categories. One is client nodes that produce only normal net transactions (NNTs). The other is chain nodes where active chain nodes are selected periodically based on their stakes and aspirations to participate in the chain management. Active chain nodes produce chain transactions (CTs) that may include hashes of NNTs. CTs form the DAG chain, which helps an ordering process of the NNTs. With the ordering ability of transactions from client nodes, the chain naturally supports global state transition. With a brief analysis, we found that the theoretical performance limitation of the chain depends only on the bandwidth and computation power of chain nodes. © 2020, Springer Nature Singapore Pte Ltd.

关键词： Directed graphs

来源：评论

学校读者我要写书评

暂无评论

A Robust Blind Watermarking Scheme for Color Images Using Quaternion Fourier Transform 6th

A Robust Blind Watermarking Scheme for Color Images Using Qu...

引用

6th International Conference on Artificial Intelligence and security, ICAIS 2020

作者： Liang, Renjie Zheng, Peijia Fang, Yanmei Song, Tingting School of Data and Computer Science Guangdong Key Laboratory of Information Security Technology Sun Yat-Sen University Guangzhou510006 China

ISBN: (纸本)9783030578800

With the advancement of image devices, massive image data are being generated exponentially, which indirectly leads to more copyright issues over the images. Image watermarking is an important method for copyright protection. Considering that color images dominate the real-world image data, designing copyright protection techniques specially for color images are very meaningful. In this paper, we propose a robust blind watermarking scheme for color images based on quaternion Fourier transform (QFT). After converting the color image data into QFT coefficients, we embed the watermark message into the four channels in QFT domain, to improve the embedding capacity and robustness. We also conduct experiment to investigate the performance of visual quality and robustness, by comparing with some other common signal transforms. The experimental results show that our scheme achieves a satisfactory performance. © 2020, Springer Nature Switzerland AG.

关键词： Image watermarking

来源：评论

学校读者我要写书评

暂无评论

Aggregation-based dual heterogeneous task allocation in spatial crowdsourcing

引用

Frontiers of computer science 2024年第6期18卷 215-227页

作者： Xiaochuan LIN Kaimin WEI Zhetao LI Jinpeng CHEN Tingrui PEI College of Information Science and Technology&Cyberspace Security Jinan UniversityGuangzhou 510632China National&Local Joint Engineering Research Center of Network Security Detection and Protection Technology Guangzhou 510632China Guangdong Provincial Key Laboratory of Data Security and Privacy Protection Guangzhou 510632China School of Computer Science Beijing University of Posts and TelecommunicationsBeijing 100876China

Spatial crowdsourcing(SC)is a popular data collection paradigm for numerous *** the increment of tasks and workers in SC,heterogeneity becomes an unavoidable difficulty in task *** researches only focus on the single-heterogeneous task ***,a variety of heterogeneous objects coexist in real-world SC *** dramatically expands the space for searching the optimal task allocation solution,affecting the quality and efficiency of data *** this paper,an aggregation-based dual heterogeneous task allocation algorithm is put *** investigates the impact of dual heterogeneous on the task allocation problem and seeks to maximize the quality of task completion and minimize the average travel *** problem is first proved to be ***,a task aggregation method based on locations and requirements is built to reduce task ***,a time-constrained shortest path planning is also developed to shorten the travel distance in a *** that,two evolutionary task allocation schemes are ***,extensive experiments are conducted based on real-world datasets in various *** with baseline algorithms,our proposed schemes enhance the quality of task completion by up to 25% and utilize 34% less average travel distance.

关键词： task allocation aggregation shortest path dual heterogeneous spatial crowdsourcing

来源：评论

学校读者我要写书评

暂无评论

Secure outsourced numerical solution of algebraic equations 6th

Secure outsourced numerical solution of algebraic equations

引用

6th International Conference on Artificial Intelligence and security,ICAIS 2020

作者： Zeng, Ke Zheng, Peijia Liu, Hongmei School of Data and Computer Science Guangdong Key Laboratory of Information Security Technology Sun Yat-Sen University Guangzhou510006 China

ISBN: (纸本)9789811581007

Numerical methods are designed to provide numerical solutions of algebraic equations, because there are not analytical solutions for algebraic equations whose degrees are larger than four. In cloud computing, outsourcing numerical solutions to the cloud raises privacy issues due to the privacy sensitive coefficients. In this paper, we propose an effective privacy-preserving computation outsourcing protocol for the common numerical solutions of algebraic equations. We first present a protocol that can securely evaluate Newton iterative method with the preservation of the private coefficients, by relying on somewhat homomorphic encryption. We then provide two implementations of our protocol based on the two popular somewhat homomorphic encryption schemes, i.e., the Brakerski-Gentry-Vaikuntanathan (BGV) scheme and the CKKS scheme. We conduct experiments to compare the two implementations and analyze the security and effectiveness. The experimental results show that the proposed scheme is effective in practice, and the CKKS-based implementation is general better than the BGV-based implementation. © Springer Nature Singapore Pte Ltd 2020.

关键词： Outsourcing

来源：评论

学校读者我要写书评

暂无评论

A distributed multi-exemplar affinity propagation clustering algorithm based on mapreduce 3

A distributed multi-exemplar affinity propagation clustering...

引用

3rd IEEE International Conference on Big data Computing Service and Applications, BigdataService 2017

作者： Yang, Yu-Bo Wang, Chang-Dong Lai, Jian-Huang School of Data and Computer Science Sun Yat-sen University Guangzhou China Guangdong Key Laboratory of Information Security Technology Guangzhou China

ISBN: (纸本)9781509063185

Clustering algorithm is one of the fundamental techniques in data mining, which plays a crucial role in various applications, such as pattern recognition, document retrieval, and computer vision. As so far, many effective algorithms have been proposed. Affinity Propagation is an algorithm requires no parameter indicating the number of clusters, which is the most distinguishing advantage compared to the k-means clustering algorithm. Multi-Exemplar Affinity Propagation (MEAP) extends the single-exemplar model to the multi-exemplar model, which could describe the dataset with more complex structure. With the amount of data increasing rapidly, the growing size of dataset makes the clustering problem become more and more challenging. To solve this problem, the parallel computing framework is widely used, such as MapReduce. However, for the MEAP algorithm, it is not a straightforward task to implement the updating of MEAP messages in MapReduce, which without proper design would be time-consuming. In this paper, we propose to utilize the stability of data distribution to apply the MEAP algorithm on the MapReduce platform and develop an efficient Distributed Multi-Exemplar Affinity Propagation (DisMEAP) clustering algorithm by using three MapReduce stages. The experiment results demonstrate that our algorithm can perform well in processing large-scale datasets and could achieve the same accuracy as the original MEAP algorithm. © 2017 IEEE.

关键词： MapReduce

来源：评论

学校读者我要写书评

暂无评论

Resampling Factor Estimation via Dual-Stream Convolutional Neural Network

引用

computers, Materials & Continua 2021年第1期66卷 647-657页

作者： Shangjun Luo Junwei Luo Wei Lu Yanmei Fang Jinhua Zeng Shaopei Shi Yue Zhang School of Data and Computer Science Guangdong Province Key Laboratory of Information Security TechnologyMinistry of Education Key Laboratory of Machine Intelligence and Advanced ComputingSun Yat-sen UniversityGuangzhou510006China Academy of Forensic Science Shanghai200063China College of Information Science and Technology Jinan UniversityGuangzhou510632China Department of Computer Science University of Massachusetts LowellLowellMA 01854USA

The estimation of image resampling factors is an important problem in image *** all the resampling factor estimation methods,spectrumbased methods are one of the most widely used methods and have attracted a lot of research ***,because of inherent ambiguity,spectrum-based methods fail to discriminate upscale and downscale operations without any prior *** general,the application of resampling leaves detectable traces in both spatial domain and frequency domain of a resampled ***,the resampling process will introduce correlations between neighboring *** this case,a set of periodic pixels that are correlated to their neighbors can be found in a resampled ***,the resampled image has distinct and strong peaks on spectrum while the spectrum of original image has no clear ***,in this paper,we propose a dual-stream convolutional neural network for image resampling factors *** of the two streams is gray stream whose purpose is to extract resampling traces features directly from the rescaled *** other is frequency stream that discovers the differences of spectrum between rescaled and original *** features from two streams are then fused to construct a feature representation including the resampling traces left in spatial and frequency domain,which is later fed into softmax layer for resampling factor *** results show that the proposed method is effective on resampling factor estimation and outperforms some CNN-based methods.

关键词： Image forensics image resampling detection parameter estimation convolutional neural network

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：