检索结果-内蒙古大学图书馆

WiViPose: A Video-aided Wi-Fi Framework for Environment-Independent 3D Human Pose Estimation

IEEE Transactions on Multimedia 2025年

作者： Zhang, Lei Ning, Haoran Tang, Jiaxin Chen, Zhenxiang Zhong, Yaping Han, Yahong Tianjin University College of Intelligence and Computing the Tianjin Key Laboratory of Advanced Network Technology and Application Tianjin300050 China Key Laboratory of Computing Power Network and Information Security Ministry of Education China University of Jinan Shandong Provincial Key Laboratory of Ubiquitous Intelligent Computing the School of Information Science and Engineering Jinan250022 China Wuhan Sports University Sports Big-data Research Center Wuhan430079 China Tianjin University College of Intelligence and Computing the Tianjin Key Laboratory of Machine Learning Tianjin300350 China

The inherent complexity of Wi-Fi signals makes video-aided Wi-Fi 3D pose estimation difficult. The challenges include the limited generalizability of the task across diverse environments, its significant signal heterogeneity, and its inadequate ability to analyze local and geometric information. To overcome these challenges, we introduce WiViPose, a video-aided Wi-Fi framework for 3D pose estimation, which attains enhanced cross-environment generalization through cross-layer optimization. Bilinear temporal-spectral fusion (BTSF) is initially used to fuse the time-domain and frequency-domain features derived from Wi-Fi. Video features are derived from a multiresolution convolutional pose machine and enhanced by local self-attention. Cross-modality data fusion is facilitated through an attention-based transformer, with the process further refined under a supervisory mechanism. WiViPose demonstrates effectiveness by achieving an average percentage of correct keypoints (PCK)@50 of 91.01% across three typical indoor environments. © 2025 IEEE.

关键词： Frequency domain analysis

来源：评论

学校读者我要写书评

暂无评论

A PERSONALIZED MODEL BASED ON SIMPLE RECURRENT UNIT FOR ACCURATE PREDICTION OF STOCK PRICE TREND

引用

International Journal of Innovative computing, Information and Control 2025年第3期21卷 791-812页

作者： Bai, Wenjiang Yuan, Shuai Lu, Yan Xu, Mingyan Chen, Shuoru Zhao, Feng Kang, Feng Intelligence and Information Engineering Department Taiyuan University No. 7 Fendong Street Tanghuai Industrial Park Taiyuan030032 China Information Engineering College Yantai Institute of Technology No. 100 Gangcheng East Street Laishan Yantai264005 China School of Statistics Computing Shandong Technology and Business University No. 191 Haibin Middle Road Laishan Yantai264005 China School of Computer Science and Technology Computing Shandong Technology and Business University No. 191 Haibin Middle Road Laishan Yantai264005 China Yantai Key Laboratory of Big Data Modeling and Intelligent Computing Shandong Technology and Business University No. 191 Haibin Middle Road Laishan Yantai264005 China

Stock is a vital component of the financial market, and how to accurately predict stock price trend is a popular topic and open problem. In order to achieve higher prediction accuracy, we propose a personalized model based on simple recurrent unit (PSRU), which can effectively eliminate the impact of noise and diversity of stock price series. Specifically, we firstly introduce variational mode decomposition (VMD) in terms of parameter determination for denoising parameters and perform denoising on the time series, effectively reducing the noise in the data and improving the prediction accuracy. Then, to better capture the characteristics of diverse time series, we cluster the time series into multiple classes, and each class fits an optimal prediction model. Especially, to improve the clustering, we introduce the dynamic multi-perspective personalized similarity measurement (DMPSM), which can help avoid singularities, time shifts and warping in time series. Finally, SRU is adopted as the prediction model which can achieve high prediction accuracy with excellent mapping and parallel processing capabilities. The experimental results show that PSRU outperforms other commonly used time series forecasting methods on multiple performance metrics, demonstrating the effectiveness and superiority of PSRU in stock price prediction, especially in dealing with nonlinear and non-stationary time series data. © 2025, ICIC International. All rights reserved.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Removing Regional Steering Vectors to Achieve Knowledge Domain Forgetting in Large Language Models 6th

Removing Regional Steering Vectors to Achieve Knowledge Dom...

引用

6th International Conference on Machine Learning for Cyber Security, ML4CS 2024

作者： Wu, Wei Wang, Chen Xu, Qiuhao Kong, Wei Zhejiang Sci-Tech University Hangzhou310018 China Zhejiang Key Laboratory of Digital Fashion and Data Governance Zhejiang Sci-Tech University Hangzhou310018 China Zhejiang Provincial International Cooperation Base for Science and Technology on Cloud Computing Security and Data Aggregation Zhejiang Sci-Tech University Hangzhou310018 China

ISBN: (纸本)9789819645657

Large language models (LLMs), as a significant direction in the development of artificial intelligence in recent years, are becoming increasingly popular. These models demonstrate significant potential in applications such as personalized services, simulated dialogues, role-playing, and specific compliance requirements. However, with the continuous expansion of LLMs’ practical applications, selective knowledge management of the models—particularly the ability to make the models "forget" certain domain-specific or topic-specific knowledge without impacting their overall performance or knowledge in other areas—has emerged as a critical issue that demands resolution. We proposes an innovative knowledge domain forgetting method designed to enable models to selectively forget specified knowledge without the need for fine-tuning. The method identifies and removes steering vectors associated with specific knowledge domains, thereby achieving effective forgetting of specific knowledge. The proposed approach has been evaluated in several popular open-source LLMs. Experimental results show that this method achieves good knowledge forgetting effects across diverse scenarios and exhibits notable practical value. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Domain Knowledge

来源：评论

学校读者我要写书评

暂无评论

VIPNet: Combining Viewpoint Information and Shape Priors for Instant Multi-view 3D Reconstruction 17th

VIPNet: Combining Viewpoint Information and Shape Priors fo...

引用

17th Asian Conference on Computer Vision, ACCV 2024

作者： Ye, Weining Li, Zhixuan Jiang, Tingting School of Computer Science Peking University Beijing100871 China College of Computing and Data Science Nanyang Technological University Singapore Singapore National Engineering Research Center of Visual Technology National Key Laboratory for Multimedia Information Processing School of Computer Science National Biomedical Imaging Center Peking University Beijing100871 China

ISBN: (纸本)9789819609680

While the multi-view 3D reconstruction task has made significant progress, existing methods simply fuse multi-view image features without effectively leveraging available auxiliary information, especially the viewpoint information for guiding and associating features of different views. To this end, we propose to enhance multi-view 3D reconstruction with the power of viewpoint information. Specifically, a simple-yet-effective viewpoint estimator is designed to learn and provide comprehensive viewpoint knowledge for locating and associating learned features from different views. Moreover, to improve the 3D reconstruction quality when 2D images of only very few viewpoints are available, we propose to learn the shape prior knowledge to provide sufficient shape information for compensating the limited 2D observations. Overall, we present VIPNet, benefiting from Viewpoint Information and Shape Prior learning for high-quality multi-view 3D reconstruction. Extensive experiments validate the effectiveness of the proposed VIPNet, which achieves state-of-the-art performance on challenging datasets and shows well generalization ability in real-world scenarios. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： 3D modeling

来源：评论

学校读者我要写书评

暂无评论

Fuzzy Information Quantity Measurement and Feature Selection by Macrogranular Entropy

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年

作者： Zhu, Zhilin Zhang, Chucai Dai, Jianhua Hunan Normal University Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing College of Information Science and Engineering Changsha410081 China

Feature selection is an important data preprocessing process in artificial intelligence, which aims to eliminate redundant features while retaining essential features. Measuring feature significance and relevance between features is a significant challenge. Fuzzy information entropy is an extension of Shannon entropy. It is widely used for quantifying the information of fuzzy divisions. However, it has significant limitations, notably the lack of monotonicity in fuzzy conditional entropy measure of decision uncertainty in the feature selection process. We introduce a novel measurement Macrogranular Entropy (ME) and construct some generalized forms, such as conditional macrogranular entropy, mutual macrogranular information, and joint macrogranular entropy. The conditional macrogranular entropy exhibits monotonicity when measuring decision uncertainty. Additionally, we propose two feature selection algorithms: one based on Monotonic Conditional Macrogranular Entropy (MCME), and the other based on the Degree of Symmetric Association (ADSA). The ADSA algorithm and the MCME algorithm are compared against eight other feature selection algorithms through a series of experiments. The comparison was conducted based on classification performance using SVM and NB classifiers, as well as evaluation metrics including F1-score and Recall. In terms of all four evaluation metrics, ADSA and MCME achieved the top two rankings, respectively. Specifically, on the NB and SVM classifiers, the ADSA algorithm improves the average accuracy by 12.22% and 2.88% compared to the original feature set, while MCME improves the accuracy by 10.07% and 1.01%, respectively. Experimental comparisons demonstrate that ADSA algorithm effectively removes redundant information from the dataset during feature selection. © 2025 IEEE.

关键词： Feature Selection

来源：评论

学校读者我要写书评

暂无评论

Big-Moe: Bypassing Isolated Gating For Generalized Multimodal Face Anti-Spoofing

Big-Moe: Bypassing Isolated Gating For Generalized Multimoda...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Yingjie Ma Zitong Yu Xun Lin Weicheng Xie Linlin Shen College of Computer Science and Software Engineering Shenzhen University Great Bay University National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Guangdong Provincial Key Laboratory of Intelligent Information Processing

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

In the domain of facial recognition security, multimodal Face Anti-Spoofing (FAS) is essential for countering presentation attacks. However, existing technologies encounter challenges due to modality biases and imbalances, as well as domain shifts. Our research introduces a Mixture of Experts (MoE) model to address these issues effectively. We identified three limitations in traditional MoE approaches to multimodal FAS: (1) Coarse-grained experts’ inability to capture nuanced spoofing indicators; (2) Gated networks’ susceptibility to input noise affecting decision-making; (3) MoE’s sensitivity to prompt tokens leading to overfitting with conventional learning methods. To mitigate these, we propose the Bypass Isolated Gating MoE (BIG-MoE) framework, featuring: (1) Fine-grained experts for enhanced detection of subtle spoofing cues; (2) An isolation gating mechanism to counteract input noise; (3) A novel differential convolutional prompt bypass enriching the gating network with critical local features, thereby improving perceptual capabilities. Extensive experiments on four benchmark datasets demonstrate significant generalization performance improvement in multimodal FAS task. The code is released at https://***/murInJ/BIG-MoE.

关键词： Convolutional codes Sensitivity Convolution Face recognition Noise Logic gates Feature extraction Security Speech processing Overfitting

来源：评论

学校读者我要写书评

暂无评论

Learning with Open-world Noisy data via Class-independent Margin in Dual Representation Space

arXiv

引用

arXiv 2025年

作者： Pan, Linchao Gao, Can Zhou, Jie Wang, Jinbao College of Computer Science and Software Engineering Shenzhen University China Guangdong Provincial Key Laboratory of Intelligent Information Processing China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University China

Learning with Noisy Labels (LNL) aims to improve the model generalization when facing data with noisy labels, and existing methods generally assume that noisy labels come from known classes, called closed-set noise. However, in real-world scenarios, noisy labels from similar unknown classes, i.e., open-set noise, may occur during the training and inference stage. Such open-world noisy labels may significantly impact the performance of LNL methods. In this study, we propose a novel dual-space joint learning method to robustly handle the open-world noise. To mitigate model overfitting on closed-set and open-set noises, a dual representation space is constructed by two networks. One is a projection network that learns shared representations in the prototype space, while the other is a One-Vs-All (OVA) network that makes predictions using unique semantic representations in the class-independent space. Then, bi-level contrastive learning and consistency regularization are introduced in two spaces to enhance the detection capability for data with unknown classes. To benefit from the memorization effects across different types of samples, class-independent margin criteria are designed for sample identification, which selects clean samples, weights closed-set noise, and filters open-set noise effectively. Extensive experiments demonstrate that our method outperforms the state-of-the-art methods and achieves an average accuracy improvement of 4.55% and an AUROC improvement of 6.17% on CIFAR80N. Copyright © 2025, The Authors. All rights reserved.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Online Multi-label Streaming Feature Selection with Missing Features by Dual-Space Consistency Information Measurement

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年

作者： Dai, Jianhua Wang, Jie Hunan Normal University Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing College of Information Science and Engineering Changsha410081 China

Multi-label feature selection is critical to address the challenges of high dimensionality and computational complexity in multi-label learning. However, in some practical applications, a more complex challenge is the dynamic generation of features, known as streaming features. Although some scholars have studied multi-label streaming feature selection methods, the problem of missing features during dynamic feature generation in multi-label data is rarely considered. Hence, we propose an online multi-label streaming feature selection method for multilabel data containing missing features by dual-space consistency information measurement. Firstly, we construct the dual-space consistency information measurement, which can effectively capture the relationships between features and labels. Specifically, we consider it from the perspective of instance correlation, and propose a fuzzy α-tolerance relation in the feature space to directly deal with missing features, and introduce an improved neighbour similarity relation in the label space to capture the instance similarity between the label combinations. Secondly, based on the dual-space consistency information, we further introduce the consistency gain index to help identify those features with important information during the streaming feature selection process. Thirdly, by combining online significance analysis, online relevance analysis, and online redundancy analysis, a new streaming feature selection framework is constructed, which can comprehensively evaluate the newly arrived features in terms of strong consistency, weak consistency and irrelevance, so as to identify the most representative features in the continuously arriving feature streams. Finally, a series of comparative experiments on multiple benchmark datasets demonstrate the effectiveness and superiority of our proposed method. © 2025 IEEE.

关键词： data consistency

来源：评论

学校读者我要写书评

暂无评论

Online Multi-Label Streaming Feature Selection by Label Enhancement and Fuzzy Synergistic Discrimination Information

引用

IEEE Transactions on Fuzzy Systems 2025年

作者： Zou, Ligeng Zhou, Tong Dai, Jianhua Hunan Normal University Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing College of Information Science and Engineering Changsha410081 China

Online streaming feature selection is an effective approach for handling large-scale streaming data in real-world applications. However, many existing online streaming feature selection studies do not effectively leverage the correlation and hidden information in the label space. The measures used in streaming feature selection often face challenges such as computational complexity and insufficient information exploitation. To address these issues, we propose an online streaming feature selection method based on label enhancement and fuzzy synergistic discrimination information. Firstly, we define a novel label enhancement method based on label correlation, which does not require any prior knowledge of feature space. Secondly, we propose a new measure to evaluate the discriminative ability of multiple feature subsets for labels in fuzzy environment, called fuzzy synergistic discrimination information. It is bounded and has many interesting properties. The fuzzy synergistic discrimination information is computed directly based on fuzzy similarity relations, so the calculation is relatively simple. Thirdly, based on label enhancement and proposed fuzzy synergistic discrimination information, we construct a three-stage online multi-label streaming feature selection method, including online significance analysis, online relevance analysis and online redundancy analysis. Finally, comparative experiments on 12 datasets demonstrate the effectiveness and superiority of the proposed method. The influences of different feature sequences and these components on the algorithm's performance are further analyzed. © 1993-2012 IEEE.

关键词： Information hiding

来源：评论

学校读者我要写书评

暂无评论

An Unsupervised Malicious Web Request Detection based on Transformer and Contrastive Learning

引用

IEEE Transactions on Network and Service Management 2025年

作者： He, Shiming Zhang, Ying Liang, Diqing Sharma, Pradip Kumar Changsha University of Science and Technology School of Computer Science and Technology Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation Changsha410114 China University of Aberdeen Department of Computing Science AberdeenAB24 3UE United Kingdom

The World Wide Web (Web) is a crucial part of the Internet. Web attacks are becoming more and more serious and complex. Malicious Web request detection aims to rapidly and accurately identify abnormal attacks on the network. Deep learning is being applied to malicious Web request detection, resulting in high detection performance. However, most deep learning-based methods are supervised and ignore special characters, which are hard to detect unknown malicious Web requests. The labels of Web request are fewer and Web request data is insufficient. Therefore, we propose an unsupervised malicious Web request detection based on transformer and contrastive learning (UTCDetector). UTCDetector exploits preprocessing and 2-gram word segmentationto preserve special characters, extracts semantic feature by Transformer, and leverages hypersphere loss function and contrastive learning to handle insufficient Web data without abnormal label. Since the public Web request datasets (CSIC 2010, CSIC TORPEDA 2012, and ECML/PKDD 2007) were created before 2012, we collected Web requests from a university Web application server in 2023 to build a private dataset named School 2023. This dataset contains more modern and complex attacks. The experimental results on the four datasets demonstrate that our method achieves a higher F1-score than other existing methods and ablation variants. © 2004-2012 IEEE.

关键词： Contrastive Learning

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：