检索结果-内蒙古大学图书馆

32nd ACM International Conference on Multimedia, MM 2024

作者： Hai, Xuan Liu, Xin Tan, Yuan Liu, Gang Li, Song Niu, Weina Zhou, Rui Zhou, Xiaokang School of Information Science and Engineering Lanzhou University Lanzhou China The State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou China School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu China Faculty of Business Data Science Kansai University Osaka Japan

ISBN: (纸本)9798400706868

Voice is one of the most widely used media for information transmission in human society. While high-quality synthetic voices are extensively utilized in various applications, they pose significant risks to content security and trust building. Numerous studies have concentrated on AI-synthesized voice detection to mitigate these risks, with many claiming to achieve promising performance. However, recent research has demonstrated that fake voice detectors suffer from serious overfitting to speaker-irrelative features (SiFs) and cannot be used in real-world scenarios. In this paper, we analyze the limitations of existing fake voice detectors and propose a new design philosophy, guiding the detection model to prioritize learning human voice features rather than the difference between the human voice and the synthetic voice. Based on this philosophy, we propose a novel AI-synthesized voice detection framework named SiFSafer, which uses pre-trained speech representation models to enhance the learning of feature distribution in human voices and the adapter fine-tuning to optimize the performance. The evaluation shows that the average EERs of existing fake voice detectors in the ASVspoof datasets can exceed 20% if the SiFs like silence segments are removed, while SiFSafer achieves an EER of less than 8%, indicating that SiFSafer is robust to SiFs and strongly resistant to existing attacks. © 2024 ACM.

关键词： Fake detection

来源：评论

学校读者我要写书评

暂无评论

Suppress Content Shift: Better Diffusion Features via Off-the-Shelf Generation Techniques 38

Suppress Content Shift: Better Diffusion Features via Off-th...

引用

38th Conference on Neural information Processing Systems, NeurIPS 2024

作者： Meng, Benyuan Xu, Qianqian Wang, Zitai Yang, Zhiyong Cao, Xiaochun Huang, Qingming Institute of Information Engineering CAS China School of Cyber Security University of Chinese Academy of Sciences China Key Lab. of Intelligent Information Processing Institute of Computing Technology CAS China Peng Cheng Laboratory China School of Computer Science and Tech. University of Chinese Academy of Sciences China School of Cyber Science and Tech. Shenzhen Campus of Sun Yat-sen University China Key Laboratory of Big Data Mining and Knowledge Management CAS China

Diffusion models are powerful generative models, and this capability can also be applied to discrimination. The inner activations of a pre-trained diffusion model can serve as features for discriminative tasks, namely, diffusion feature. We discover that diffusion feature has been hindered by a hidden yet universal phenomenon that we call content shift. To be specific, there are content differences between features and the input image, such as the exact shape of a certain object. We locate the cause of content shift as one inherent characteristic of diffusion models, which suggests the broad existence of this phenomenon in diffusion feature. Further empirical study also indicates that its negative impact is not negligible even when content shift is not visually perceivable. Hence, we propose to suppress content shift to enhance the overall quality of diffusion features. Specifically, content shift is related to the information drift during the process of recovering an image from the noisy input, pointing out the possibility of turning off-the-shelf generation techniques into tools for content shift suppression. We further propose a practical guideline named GATE to efficiently evaluate the potential benefit of a technique and provide an implementation of our methodology. Despite the simplicity, the proposed approach has achieved superior results on various tasks and datasets, validating its potential as a generic booster for diffusion features. Our code is available at this url. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Zero Watermarking Algorithm for Medical Image Based on Resnet50-DCT

引用

Computers, Materials & Continua 2023年第4期75卷 293-309页

作者： Mingshuai Sheng Jingbing Li Uzair Aslam Bhatti Jing Liu Mengxing Huang Yen-Wei Chen School of Information and Communication Engineering Hainan UniversityHaikou570228China State Key Laboratory of Marine Resource Utilization in the South China Sea Hainan UniversityHaikou570228China School of Computer Science and Technology Hainan UniversityHaikou570228China Research Center for Healthcare Data Science Zhejiang LabHangzhou311121China Graduate School of Information Science and Engineering Ritsumeikan UniversityKyoto5258577Japan

Medical images are used as a diagnostic tool, so protecting theirconfidentiality has long been a topic of study. From this, we propose aResnet50-DCT-based zero watermarking algorithm for use with medicalimages. To begin, we use Resnet50, a pre-training network, to draw out thedeep features of medical images. Then the deep features are transformedby DCT transform and the perceptual hash function is used to generatethe feature vector. The original watermark is chaotic scrambled to get theencrypted watermark, and the watermark information is embedded into theoriginal medical image by XOR operation, and the logical key vector isobtained and saved at the same time. Similarly, the same feature extractionmethod is used to extract the deep features of the medical image to be testedand generate the feature vector. Later, the XOR operation is carried outbetween the feature vector and the logical key vector, and the encryptedwatermark is extracted and decrypted to get the restored watermark;thenormalized correlation coefficient (NC) of the original watermark and therestored watermark is calculated to determine the ownership and watermarkinformation of the medical image to be tested. After calculation, most ofthe NC values are greater than 0.50. The experimental results demonstratethe algorithm’s robustness, invisibility, and security, as well as its ability toaccurately extract watermark information. The algorithm also shows goodresistance to conventional attacks and geometric attacks.

关键词： Medical images deep residual network resnet50-DCT privacy protection robustness security

来源：评论

学校读者我要写书评

暂无评论

P2 Law: Scaling Law for Post-Training After Model Pruning

arXiv

引用

arXiv 2024年

作者： Chen, Xiaodong Hu, Yuxuan Zhang, Xiaokang Wang, Yanling Li, Cuiping Chen, Hong Zhang, Jing School of Information Renmin University of China Beijing China Key Laboratory of Data Engineering and Knowledge Engineering Beijing China Zhongguancun Laboratory China

Pruning has become a widely adopted technique for reducing the hardware requirements of large language models (LLMs). To recover model performance after pruning, post-training is commonly employed to mitigate the resulting performance degradation. While post-training benefits from larger datasets, once the dataset size is already substantial, increasing the training data provides only limited performance gains. To balance post-training cost and model performance, it is necessary to explore the optimal amount of post-training data. Through extensive experiments on the Llama-3 and Qwen-2.5 series models, pruned using various common pruning methods, we uncover the scaling Law for Post-training after model Pruning, referred to as the P2 Law. This law identifies four key factors for predicting the pruned model’s post-training loss: the model size before pruning, the number of post-training tokens, the pruning rate, and the model’s loss before pruning. Moreover, P2 Law can generalize to larger dataset sizes, larger model sizes, and higher pruning rates, offering valuable insights for the post-training of pruned LLMs. © 2024, CC BY.

关键词： Digital elevation model

来源：评论

学校读者我要写书评

暂无评论

Not All Diffusion Model Activations Have Been Evaluated as Discriminative Features 38

Not All Diffusion Model Activations Have Been Evaluated as D...

引用

38th Conference on Neural information Processing Systems, NeurIPS 2024

作者： Meng, Benyuan Xu, Qianqian Wang, Zitai Cao, Xiaochun Huang, Qingming Institute of Information Engineering CAS China School of Cyber Security University of Chinese Academy of Sciences China Key Lab. of Intelligent Information Processing Institute of Computing Technology CAS China Peng Cheng Laboratory China School of Cyber Science and Tech. Shenzhen Campus of Sun Yat-sen University China School of Computer Science and Tech. University of Chinese Academy of Sciences China Key Laboratory of Big Data Mining and Knowledge Management CAS China

Diffusion models are initially designed for image generation. Recent research shows that the internal signals within their backbones, named activations, can also serve as dense features for various discriminative tasks such as semantic segmentation. Given numerous activations, selecting a small yet effective subset poses a fundamental problem. To this end, the early study of this field performs a large-scale quantitative comparison of the discriminative ability of the activations. However, we find that many potential activations have not been evaluated, such as the queries and keys used to compute attention scores. Moreover, recent advancements in diffusion architectures bring many new activations, such as those within embedded ViT modules. Both combined, activation selection remains unresolved but overlooked. To tackle this issue, this paper takes a further step with a much broader range of activations evaluated. Considering the significant increase in activations, a full-scale quantitative comparison is no longer operational. Instead, we seek to understand the properties of these activations, such that the activations that are clearly inferior can be filtered out in advance via simple qualitative evaluation. After careful analysis, we discover three properties universal among diffusion models, enabling this study to go beyond specific models. On top of this, we present effective feature selection solutions for several popular diffusion models. Finally, the experiments across multiple discriminative tasks validate the superiority of our method over the SOTA competitors. Our code is available at this url. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Actor-centric Causality Graph for Asynchronous Temporal Inference in Group Activity

An Actor-centric Causality Graph for Asynchronous Temporal I...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Zhao Xie Tian Gao Kewei Wu Jiao Chang Key Laboratory of Knowledge Engineering with Big Data Hefei University of Technology School of Computer Science and Information Engineering Hefei University of Technology

The causality relation modeling remains a challenging task for group activity recognition. The causality relations describe the influence on the centric actor (effect actor) from its correlative actors (cause actors). Most existing graph models focus on learning the actor relation with synchronous temporal features, which is insufficient to deal with the causality relation with asynchronous temporal features. In this paper, we propose an Actor-Centric Causality Graph Model, which learns the asynchronous temporal causality relation with three modules, i.e., an asynchronous temporal causality relation detection module, a causality feature fusion module, and a causality relation graph inference module. First, given a centric actor and its correlative actor, we analyze their influences to detect causality relation. We estimate the self influence of the centric actor with self regression. We estimate the correlative influence from the correlative actor to the centric actor with correlative regression, which uses asynchronous features at different timestamps. Second, we synchronize the two action features by estimating the temporal delay between the cause action and the effect action. The synchronized features are used to enhance the feature of the effect action with a channel-wise fusion. Third, we describe the nodes (actors) with causality features and learn the edges by fusing the causality relation with the appearance relation and distance relation. The causality relation graph inference provides crucial features of effect action, which are complementary to the base model using synchronous relation inference. Experiments show that our method achieves state-of-the-art performance on the Volleyball dataset and Collective Activity dataset.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Residential Energy Consumption Forecasting Based on Federated Reinforcement Learning with data Privacy Protection

引用

Computer Modeling in engineering & Sciences 2023年第10期137卷 717-732页

作者： You Lu Linqian Cui YunzheWang Jiacheng Sun Lanhui Liu School of Electronic and Information Engineering Suzhou University of Science and TechnologySuzhou215009China Jiangsu Province Key Laboratory of Intelligent Building Energy Efficiency Suzhou University of Science and TechnologySuzhou215009China Chongqing Industrial Big Data Innovation Center Co. Ltd.Chongqing400707China

Most studies have conducted experiments on predicting energy consumption by integrating data formodel ***, the process of centralizing data can cause problems of data ***,many laws and regulationson data security and privacy have been enacted, making it difficult to centralize data, which can lead to a datasilo problem. Thus, to train the model while maintaining user privacy, we adopt a federated learning ***, in all classical federated learning frameworks secure aggregation, the Federated Averaging (FedAvg)method is used to directly weight the model parameters on average, which may have an adverse effect on te ***, we propose the Federated Reinforcement Learning (FedRL) model, which consists of multiple userscollaboratively training the model. Each household trains a local model on local data. These local data neverleave the local area, and only the encrypted parameters are uploaded to the central server to participate in thesecure aggregation of the global model. We improve FedAvg by incorporating a Q-learning algorithm to assignweights to each locally uploaded local model. And the model has improved predictive performance. We validatethe performance of the FedRL model by testing it on a real-world dataset and compare the experimental results withother models. The performance of our proposed method in most of the evaluation metrics is improved comparedto both the centralized and distributed models.

关键词： Energy consumption forecasting federated learning data privacy protection Q-learning

来源：评论

学校读者我要写书评

暂无评论

Quantifying the agreement and accuracy characteristics of four satellite-based LULC products for cropland classification in China

引用

Journal of Integrative Agriculture 2024年第1期23卷 283-297页

作者： Jie Xue Xianglin Zhang Songchao Chen Bifeng Hu Nan Wang Zhou Shi Department of Land Management Zhejiang UniversityHangzhou 310058China Institute of Applied Remote Sensing and Information Technology College of Environmental and Resource SciencesZhejiang UniversityHangzhou 310058China ZJU-Hangzhou Global Scientific and Technological Innovation Center Zhejiang UniversityHangzhou 311215China Department of Land Resource Management School of Public Finance and Public AdministrationJiangxi University of Finance and EconomicsNanchang 330013China Key Laboratory of Data Science in Finance and Economics Jiangxi University of Finance and EconomicsNanchang 330013China Key Laboratory of Spectroscopy Sensing Ministry of Agriculture and Rural AffairsHangzhou 310058China

Various land use and land cover(LULC)products have been produced over the past decade with the development of remote sensing *** the differences in LULC classification schemes,there is a lack of research on assessing the accuracy of their application to croplands in a unified ***,this study evaluated the spatial and area accuracies of cropland classification for four commonly used global LULC products(i.e.,MCD12Q1V6,GlobCover2009,FROM-GLC and GlobeLand30)based on the harmonised FAO criterion,and quantified the relationships between four factors(i.e.,slope,elevation,field size and crop system)and cropland classification *** validation results indicated that MCD12Q1 and GlobeLand30 performed well in cropland classification regarding spatial consistency,with overall accuracies of 94.90 and 93.52%,*** FROMGLC showed the worst performance,with an overall accuracy of 83.17%.Overlaying the cropland generated by the four global LULC products,we found the proportions of complete agreement and disagreement were 15.51 and 44.72% for the cropland classification,*** consistency was mainly observed in the Northeast China Plain,the Huang-Huai-Hai Plain and the northern part of the Middle-lower Yangtze Plain,*** contrast,low consistency was detected primarily on the eastern edge of the northern and semiarid region,the Yunnan-Guizhou Plateau and southern *** size was the most important factor for mapping *** area accuracy,compared with China Statistical Yearbook data at the provincial scale,the accuracies of different products in descending order were:GlobeLand30,FROM-GLC,MCD12Q1,and *** cropland classification schemes mainly caused large area deviations among the four products,and they also resulted in the different ranks of spatial accuracy and area accuracy among the four *** results can provide valuable suggestions for selecting cropland products at the national or provinc

关键词： global LULC products cropland mapping accuracy evaluation food security China

来源：评论

学校读者我要写书评

暂无评论

A secure quantum homomorphic encryption ciphertext retrieval scheme

引用

Soft Computing 2025年第3期29卷 1497-1509页

作者： Cheng, Zhen-Wen Chen, Xiu-Bo Xu, Gang Chang, Yan Miao, Li-Hua Yang, Yi-Xian Wang, Ya-Lan School of Data and Computer Science Shandong Women’s University Jinan250300 China Information Security Center State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing100876 China School of Information Science and Technology North China University of Technology Beijing100144 China School of Cybersecurity Chengdu University of Information Technology Chengdu610225 China Huawei Technologies Co. Ltd Shenzhen518129 China Department of Computer Science Faculty of Engineering and Physical Sciences University of Surrey Surrey GuildfordGU2 7XH United Kingdom

In a recent paper (Gong et al. Quantum Inf Process 19:3, 2020), a novel ciphertext retrieval scheme based on the Grover algorithm and quantum homomorphic encryption was presented. In this scheme, when the server performs the operation of marking the solution on the user’s encrypted state in the Grover iteration, it needs to remove many gate-errors generated in the homomorphic evaluation of the T gate. And the server could judge this specific solution from the quantum circuit of marking the solution. It makes this scheme unable to achieve the low-cost and secure ciphertext retrieval. Therefore, we improve the Gong et al.’s scheme and propose a secure quantum homomorphic encryption ciphertext retrieval scheme. In our scheme, the trusted third party is introduced to cooperate with the server to execute the Grover algorithm. In each Grover iteration, the trusted third party can quickly mark the solution on the plaintext state, encrypt the marked state, and transmit it to the server. Then the server performs the remaining operations of this Grover iteration on the encrypted state. The trusted third party finally decrypts the iterated state. This cooperative approach ensures that the number of auxiliary qubits required and extra quantum gates executed in our scheme are lower than the Gong et al.’s scheme. By analyzing the security of our scheme, we confirm that the server and the trusted third party will not be informed of this solution. Thus, our scheme realizes the secure ciphertext retrieval with low computational overhead. We utilize IBM’s Qiskit framework to simulate our scheme, and the experimental result shows that our scheme is correct. It is worth noting that the low-cost and secure ciphertext retrieval will play a crucial role in modern information security and privacy protection. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Ciphertext

来源：评论

学校读者我要写书评

暂无评论

Research on Metaverse Security and Forensics

引用

Computers, Materials & Continua 2023年第10期77卷 799-825页

作者： Guangjun Liang Jianfang Xin Qun Wang Xueli Ni Xiangmin Guo Pu Chen Department of Computer Information and Cyber Security Jiangsu Police InstituteNanjingChina Engineering Research Center of Electronic Data Forensics Analysis NanjingChina Key Laboratory of Digital Forensics Department of Public Security of Jiangsu ProvinceNanjingChina School of Intelligent Engineering Nanjing Institute of Railway TechnologyNanjingChina

As a subversive concept,the metaverse has recently attracted widespread attention around the world and has set off a wave of enthusiasm in academic,industrial,and investment ***,while the metaverse brings unprecedented opportunities for transformation to human society,it also contains related *** is a digital living space with information infrastructure,interoperability system,content production system,and value settlement system as the underlying structure in which the inner core is to connect real residents through applications and *** social incentives and governance rules,the metaverse reflects the digital migration of human *** article will conduct an in-depth analysis of the metaverse from the perspective of electronic data ***,from the perspective of Internet development,the background and development process of the metaverse is *** systematically elaborating on the concept and connotation of the metaverse,this paper summarizes the different views of current practitioners,experts,and scholars on the ***,from the perspective of metaverse security,the social risk and crime risks of the metaverse are *** the importance of metaverse forensics is ***,from the perspective of blockchain,smart wearable devices,and virtual reality devices,the objects and characteristics of metaverse forensics have been studied in *** smart wearable devices as an example,this paper gives the relevant experimental process of smart bracelet ***,many challenges faced by metaverse forensics are summarized by us which provide readers with some exploratory guidance.

关键词： Metaverse forensics blockchain smart wear virtual reality

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：