检索结果-内蒙古大学图书馆

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Dong, Wenlong Zhu, Qing Mao, Qirong School of Computer Science and Communication Engineering Jiangsu University China Jiangsu Engineering Research Center of Big Data Ubiquitous Perception and Intelligent Agriculture Applications Provincial Key Laboratory of Computational Intelligence and New Technologies in Low-Altitude Digital Agriculture Zhenjiang China

ISBN: (纸本)9798350368741

Video Character Social Relationship Recognition (VCSRR) requires a comprehensive consideration about spatio-temporal and multi-modal clues in videos. Most existing methods mainly focus on integrating multi-modal clues and modeling interactions among characters. However, they fail to discover key clues in the complex video data or fully understand the clues related to social relationships. In this article, we propose a novel Large Language Model Enhanced Key Clues Selection (LEKCS) framework to address the aforementioned issues. The core of LE-KCS is to mine multi-scale key clues from the perspectives of time, space and multi-modality, then transfer the knowledge about social relationships of the Large Language Model to VCSRR for understanding the selected clues. We evaluated LEKCS on the MovieGraphs dataset and the experimental results indicate that our proposed LE-KCS achieves state-of-the-art performance. © 2025 IEEE.

关键词： large language model Social relationship recognition video understanding

来源：评论

学校读者我要写书评

暂无评论

Kolmogorov-Arnold Network for Solving 2-D Magnetostatic Problems

引用

IEEE Transactions on Magnetics 2025年

作者： Zhu, Yachao Xu, Kai Wan, Bingkuan Lei, Gang Zhu, Jianguo University of Technology Sydney School of Electrical and Data Engineering UltimoNSW2007 Australia The University of Sydney School of Electrical and Computer Engineering CamperdownNSW2006 Australia

The data-driven machine-learning approach has significantly advanced the development of computational electromagnetics. This study introduces the Kolmogorov-Arnold Network (KAN) as a novel method to overcome the limitations of traditional multilayer perceptron-based physics-informed neural networks (MLP-PINNs), which often struggle with fixed activation functions and high computational costs. In terms of accuracy, KAN outperforms traditional PINNs with a 13.47% improvement in estimated accuracy. For efficient convergence, KAN achieves stable training with only 175 steps, significantly reducing computational overhead compared to PINNs, which require over 15,000 steps. Additionally, KAN demonstrates superior generalizability and flexibility, achieving an average 5.03% accuracy improvement over PINN in transfer learning scenarios. These results highlight KAN's potential in computational electromagnetics. © 1965-2012 IEEE.

关键词： Magnetostatics

来源：评论

学校读者我要写书评

暂无评论

Representation with Minimized Max-Error in Optimal Piecewise Linear Approximation of Time Series data 25th

Representation with Minimized Max-Error in Optimal Piecewi...

引用

25th International Conference on Web Information Systems engineering, WISE 2024

作者： Zhao, Huanyu Li, Tongliang Wen, Shiting Shu, Zhenyu Deng, Ke Yang, Jian Pang, Chaoyi School of Computer Science and Technology Donghua University Shanghai China Institute of Applied Mathematics Hebei Academy of Sciences Shijiazhuang China Hebei Authentication Technology Engineering Research Center Shijiazhuang China Institute of Biology Hebei Academy of Sciences Shijiazhuang China School of Computer and Data Engineering NingboTech University Ningbo China RMIT University Melbourne Australia Macquarie University Sydney Australia

ISBN: (纸本)9789819605781

In the past two decades, Piecewise Linear Approximation under maximum error (max-error) bound (PLA∞) has been intensively studied for effective qualified representation and analysis of time series data. It divides a time series into fragments and then represents each fragment with a straight line to approximate the data points of that time slot. In this paper, to elevate the representation quality in the optimal PLA∞ results, we present a linear-time algorithm FindMin to construct the unique line representative of minimized maximum error (min-max error) for each fragment. Through extended experimental tests, we demonstrate that the proposed algorithm is very efficient in execution and achieves better performances than the state-of-the-art solutions. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Time series analysis

来源：评论

学校读者我要写书评

暂无评论

引用

8th International Congress on Edge Computing, EDGE 2024, Held as Part of the Services Conference Federation, SCF 2024

作者： Tian, Yun Wu, Bin Shi, Jiaoli Zhang, Caicai Xu, Du School of Computer and Big Data Science Jiujiang University Jiujiang332005 China Jiujiang Key Laboratory of Network and Information Security Jiujiang332005 China School of Modern Information Technology Zhejiang Polytechnic University of Mechanical and Electrical Engineering Hangzhou310053 China

ISBN: (纸本)9783031770685

The development of cloud computing and the widespread application of cloud services have made outsourcing services more convenient. The need for individuals and businesses to store and manipulate the graph data they generate is growing rapidly. The unreliability and insecurity of cloud servers make outsourcing graph data a great risk of information leakage. To effectively protect data security, encrypting outsourced data is a useful method. The adjacent vertex query is a very commonly used and fundamental operation, and similarity search is a widely used and powerful tool to improve the scope and functionality of queries. After outsourcing encrypted sparse graph data to cloud servers, it becomes very inconvenient to use and manipulate the data. In this work, we present a scheme to realize the adjacent vertex query supporting similarity search on sparse graph data in cloud environment (SSAQ), which also protects the security of the information. This work uses edit distance and the searchable encryption principle to construct query index, and next implement the similar adjacent vertex query on cloud server. This work provides a formal security analysis, and also gives the experimental comparison and analysis. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Information leakage

来源：评论

学校读者我要写书评

暂无评论

Graph decision transformer for offline reinforcement learning

引用

Science China(Information Sciences) 2025年第06期 395-396页

作者： Shengchao HU Li SHEN Ya ZHANG Dacheng TAO School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University Shanghai Artificial Intelligence Laboratory Shanghai AI Laboratory School of Cyber Science and Technology Shenzhen Campus of Sun Yat-sen University School of Computer and Data Science Nanyang Technological University

Recent advances [1, 2] in offline reinforcement learning(RL)have taken a new perspective on the problem, departing from conventional methods that concentrate on learning value functions or policy gradients. Instead, the problem is viewed as a generic sequence modeling task, where past experiences consisting of state-action-reward triplets are input to the Transformer.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Efficient Direct Downlink Sensing Method Using 5G NR SSB Signals in Perceptive Mobile Networks

引用

IEEE Internet of Things Journal 2025年第11期12卷 15360-15369页

作者： Li, Hang Xiang, Yang Guo, Qinghua Liu, Lizhe Huang, Xiaojing Cheng, Zhiqun Pang, Yashan Hangzhou Dianzi University School of Electronics and Information Hangzhou310018 China Academy of Network and Communications of CETC National Key Laboratory of Advanced Communication Networks Hebei Shijiazhuang050081 China University of Wollongong School of Electrical Computer and Telecommunications Engineering WollongongNSW2522 Australia University of Technology Sydney Global Big Data Technologies Centre UltimoNSW2007 Australia

In perceptive mobile networks (PMNs), using 5G New Radio (NR) signals for direct sensing poses a significant challenge to practical implementation due to the high computational complexity involved in estimating sensing parameters. In this paper, an efficient sensing method is proposed to incorporate both downlink active sensing and passive sensing to estimate multiple sensing parameters, including delays, angle of arrival (AoA), angle of departure (AoD) and Doppler. In particular, it exploits the synchronization signal blocks (SSBs) to facilitate sensing with multiple remote radio units (RRUs). To reduce the computational complexity of direct sensing, a sparse model is developed to decouple multiple sensing parameter estimation, enabling efficient sensing method design. Then, leveraging unitary approximate message passing (UAMP) and sparse Bayesian learning (SBL), we propose an efficient method to achieve parameter estimation and association with corresponding RRUs. This method is further extended to general scenarios involving multiple path components with the same delay. Extensive simulations demonstrate the effectiveness of the proposed method, showing that it outperforms existing ones in terms of sensing accuracy and complexity. © 2014 IEEE.

关键词： 5G mobile communication systems

来源：评论

学校读者我要写书评

暂无评论

Prediction of AIDS/HIV Infection using Custom Artificial Neural Networks Model 1

Prediction of AIDS/HIV Infection using Custom Artificial Neu...

引用

1st International Conference on Intelligent Systems and Computational Networks, ICISCN 2025

作者： Sneha, B.J. Shriyashreshta Kodipalli, Ashwini Martis, Roshan Joy Ushasree, A. Devamane, Shridhar B. Global Academy of Technology Department of Ai & Data Science Bangalore India Manipal Institute of Technology Department of Electronics & Communication Engineering Bangalore India Gokaraju Lailavathi Engineering College Department of Computer Science & Engineering Hyderabad India Amrita Vishwavidyapeetam School of Computing Department of Cs Mysuru India

ISBN: (纸本)9798331529246

HIV is a serious disease that impairs immunity. Without treatment, the infection may progress through three stages, which might drastically shorten a person's life. Artificial neural networks (ANNs) are utilized to detect the HIV infection and even to find treatments for it. The AIDS viral infection is predicted in this review. The aim of ANN is to precisely classify patients according to whether or not they have HIV/AIDS. We have formed a custom ANN model and applied it to the Kaggle dataset that is accessible to the public, compared which optimizers offer the best accuracy and whether employing dropouts yields improved accuracy. With Gradient Descent, we achieved the maximum accuracy of 86.92% it is also one of the most commonly used optimizers due to its simplicity and effectiveness, followed by SGD with 86.45% it is a variation of gradient descent in which a random subset of data as opposed to the complete dataset is used to update the parameter and RMSprop with 86.21% uses the average of recent gradient magnitudes to modify the learning rate for each parameter. This aids with training stabilization, particularly when dealing with sparse or noisy data. The accuracy increased to 87.62% after we introduced dropouts to the hidden layer with optimizer as gradient descent, which produced the best results. Loss vs. Accuracy graphs and ROC curve have been plotted. With dropouts, the custom ANN model performed even better on the Gradient Descent optimizer. © 2025 IEEE.

关键词： ANN dropouts HIV/AIDS optimizers

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

MCITEBENCH: A Benchmark for Multimodal Citation Text Generation in MLLMs

arXiv

引用

arXiv 2025年

作者： Hu, Caiyu Zhang, Yikai Zhu, Tinghui Ye, Yiwei Xiao, Yanghua Shanghai Key Laboratory of Data Science School of Computer Science Fudan University China School of Computer Engineering and Science Shanghai University China

Multimodal Large Language Models (MLLMs) have advanced in integrating diverse modalities but frequently suffer from hallucination. A promising solution to mitigate this issue is to generate text with citations, providing a transparent chain for verification. However, existing work primarily focuses on generating citations for text-only content, overlooking the challenges and opportunities of multimodal contexts. To address this gap, we introduce MCITEBENCH, the first benchmark designed to evaluate and analyze the multimodal citation text generation ability of MLLMs. Our benchmark comprises data derived from academic papers and review-rebuttal interactions, featuring diverse information sources and multimodal content. We comprehensively evaluate models from multiple dimensions, including citation quality, source reliability, and answer accuracy. Through extensive experiments, we observe that MLLMs struggle with multimodal citation text generation. We also conduct deep analyses of models’ performance, revealing that the bottleneck lies in attributing the correct sources rather than understanding the multimodal content. Copyright © 2025, The Authors. All rights reserved.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

AugGPT: Leveraging ChatGPT for Text data Augmentation

引用

IEEE Transactions on Big data 2025年第3期11卷 907-918页

作者： Dai, Haixing Liu, Zhengliang Liao, Wenxiong Huang, Xiaoke Cao, Yihan Wu, Zihao Zhao, Lin Xu, Shaochen Zeng, Fang Liu, Wei Liu, Ninghao Li, Sheng Zhu, Dajiang Cai, Hongmin Sun, Lichao Li, Quanzheng Shen, Dinggang Liu, Tianming Li, Xiang University of Georgia School of Computing AthensGA30602 United States South China University of Technology School of Computer Science and Engineering Guangzhou510641 China Lehigh University Department of Computer Science and Engineering BethlehemPA18015 United States Carnegie Mellon University Heinz College of Information Systems and Public Policy PittsburghPA15213 United States Harvard Medical School Department of Radiology Massachusetts General Hospital BostonMA02115 United States Mayo Clinic Department of Radiation Oncology PhoenixAZ85054 United States University of Virginia School of Data Science CharlottesvilleVA22903 United States The University of Texas at Arlington Department of Computer Science and Engineering ArlingtonTX76019 United States ShanghaiTech University School of Biomedical Engineering Shanghai201210 China Shanghai United Imaging Intelligence Company Ltd. Shanghai200230 China Shanghai Clinical Research and Trial Center Shanghai201210 China

Text data augmentation is an effective strategy for overcoming the challenge of limited sample sizes in many natural language processing (NLP) tasks. This challenge is especially prominent in the few-shot learning (FSL) scenario, where the data in the target domain is generally much scarcer and of lowered quality. A natural and widely used strategy to mitigate such challenges is to perform data augmentation to better capture data invariance and increase the sample size. However, current text data augmentation methods either can’t ensure the correct labeling of the generated data (lacking faithfulness), or can’t ensure sufficient diversity in the generated data (lacking compactness), or both. Inspired by the recent success of large language models (LLM), especially the development of ChatGPT, we propose a text data augmentation approach based on ChatGPT (named "AugGPT"). AugGPT rephrases each sentence in the training samples into multiple conceptually similar but semantically different samples. The augmented samples can then be used in downstream model training. Experiment results on multiple few-shot learning text classification tasks show the superior performance of the proposed AugGPT approach over state-of-the-art text data augmentation methods in terms of testing accuracy and distribution of the augmented samples. © 2015 IEEE.

关键词： Zero-shot 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：