检索结果-内蒙古大学图书馆

Frequency-enhanced and decomposed transformer for multivariate time series anomaly detection

Applied Intelligence 2025年第6期55卷 1-18页

作者： Li, Shijiang Wang, Zhihai Wang, Xiaokang Yin, Zihao Yao, Muyun School of Computer Science and Technology Beijing Jiaotong University Beijing China School of Automation and Intelligence Beijing Jiaotong University Beijing China Engineering Center on Computer Network Management in Hi-Speed Railway Ministry of Education Beijing Jiaotong University Beijing China

With the widespread adoption of the Internet of Things (IoT), vast amounts of multivariate time series data are generated, which reflect the operational status of systems. Accurate and efficient anomaly detection in these data is crucial for maintaining system stability. However, data from unstable environments often exhibit high volatility, data drift, and complex patterns of anomalies. Unsupervised anomaly detection models are typically designed for stable data and lack generalizability, leading to a high rate of false positives when applied to unstable data. This paper introduces the frequency-enhanced and decomposed transformer for anomaly detection (FDTAD), which is a novel anomaly detection model based on a transformer that is enhanced with frequency and time series decomposition. FDTAD addresses data drift by decomposing time series and leverages both time-domain and frequency-domain information to improve the generalization ability of the model. The model preserves major amplitudes in the frequency domain to extract primary periodic patterns, uses spectral residuals to capture detailed variations, and incorporates a frequency-domain correlation attention mechanism to extract dependencies in frequency-domain data in a sparse representation. Additionally, a spatiotemporal module is designed to extract the temporal correlations in the data and spatial correlations among the data with different attributes. FDTAD combines a data periodic pattern reconstructor and a data detailed pattern reconstructor through an adversarial mechanism to achieve maximum accuracy in reconstructing normal data. Extensive experiments on 10 public datasets demonstrate that FDTAD outperforms state-of-the-art baseline methods, with a 4.1% improvement in the F1 score and a 4.7% improvement in precision. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

SCASAN: Spatial Central Attention-Spectral Attention Network Based Feature Extractor for Hyperspectral Image Classification

SCASAN: Spatial Central Attention-Spectral Attention Network...

引用

2023 China Automation Congress, CAC 2023

作者： Zhao, Wei Zhang, Hui Cao, Yihong Liu, Lizhu Su, Yingjian School of Robotics Hunan University Changsha China College of Computer Science and Electronic Engineering Hunan University Changsha China

ISBN: (纸本)9798350303759

Due to the influence of space, time, environmental changes, and types of ground objects on the lighting conditions of high-light falling images, Hyperspectral image (HSI) classification is challenging. Recently, attention-based HSI classification has received increasing attention. It uses both spatial attention and spectral attention to capture the required spatial-spectral features, thereby achieving better classification accuracy. However, it is challenging to connect the two attention mechanisms without compromising the spatial-spectral features. In addition, how to obtain the attention weight matrix which determines the importance of each channel or spatial location is also an important problem in the use of attention mechanisms. In response to the above issues, the Spatial Central Attention-Spectral Attention network (SCASAN) is designed to efficiently model the spatial-spectral information for HSI classification. In this network, we propose the novel spectral weight spatial attention (SWSA) module to solve the connection problem between spatial attention and spectral attention. Furthermore, we utilize a spatial central attention network (spatial CAN) to obtain a spatial attention weight matrix to avoid the impact of non-homogeneous adjacent pixels on the classified pixel. Our proposed approach demonstrates superior performance in hyperspectral image classification compared to state-of-the-art methods, as evidenced by comprehensive experiments conducted on three publicly available datasets. © 2023 IEEE.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Pre-training transformer with dual-branch context content module for table detection in document images

引用

虚拟现实与智能硬件（中英文） 2024年第5期6卷 408-420页

作者： Yongzhi LI Pengle ZHANG Meng SUN Jin HUANG Ruhan HE School of Computer Science and Artificial Intelligence Wuhan Textile UniversityWuhan 430064China School of Computer Science South-Central Minzu UniversityWuhan 430064China Hubei Provincial Engineering Research Center for Intelligent Textile and Fashion Wuhan Textile UniversityWuhan 430064China

Background Document images such as statistical reports and scientific journals are widely used in information *** detection of table areas in document images is an essential prerequisite for tasks such as information ***,because of the diversity in the shapes and sizes of tables,existing table detection methods adapted from general object detection algorithms,have not yet achieved satisfactory *** detection results might lead to the loss of critical *** Therefore,we propose a novel end-to-end trainable deep network combined with a self-supervised pretraining transformer for feature extraction to minimize incorrect *** better deal with table areas of different shapes and sizes,we added a dualbranch context content attention module(DCCAM)to high-dimensional features to extract context content information,thereby enhancing the network's ability to learn shape *** feature fusion at different scales,we replaced the original 3×3 convolution with a multilayer residual module,which contains enhanced gradient flow information to improve the feature representation and extraction *** We evaluated our method on public document datasets and compared it with previous methods,which achieved state-of-the-art results in terms of evaluation metrics such as recall and ***://***/Yong Z-Lee/TD-DCCAM.

关键词： Table detection Document image analysis Transformer Dilated convolution Deformable convolution Feature fusion

来源：评论

学校读者我要写书评

暂无评论

Neighborhood Combination Search for Single-Machine Scheduling with Sequence-Dependent Setup Time

引用

Journal of computer science & Technology 2024年第3期39卷 737-752页

作者：刘晓路徐宏云陈嘉铭苏宙行吕志鹏丁俊文 College of System Engineering National University of Defense TechnologyChangsha 410015China School of Artificial Intelligence Jianghan UniversityWuhan 430056China Xi’an Satellite Control Center Xi’an 710043China School of Computer Science and Technology Huazhong University of Science and TechnologyWuhan 430074China CCF

In a local search algorithm,one of its most important features is the definition of its neighborhood which is crucial to the algorithm's *** this paper,we present an analysis of neighborhood combination search for solv-ing the single-machine scheduling problem with sequence-dependent setup time with the objective of minimizing total weighted tardiness(SMSWT).First,We propose a new neighborhood structure named Block Swap(B1)which can be con-sidered as an extension of the previously widely used Block Move(B2)neighborhood,and a fast incremental evaluation technique to enhance its evaluation ***,based on the Block Swap and Block Move neighborhoods,we present two kinds of neighborhood structures:neighborhood union(denoted by B1UB2)and token-ring search(denoted by B1→B2),both of which are combinations of B1 and ***,we incorporate the neighborhood union and token-ring search into two representative metaheuristic algorithms:the Iterated Local Search Algorithm(ILSnew)and the Hybrid Evolutionary Algorithm(HEA_(new))to investigate the performance of the neighborhood union and token-ring ***-sive experiments show the competitiveness of the token-ring search combination mechanism of the two *** on the 120 public benchmark instances,our HEA_(new)has a highly competitive performance in solution quality and computational time compared with both the exact algorithms and recent *** have also tested the HEA,new algorithm with the selected neighborhood combination search to deal with the 64 public benchmark instances of the single-machine scheduling problem with sequence-dependent setup *** is able to match the optimal or the best known results for all the 64 *** particular,the computational time for reaching the best well-known results for five chal-lenging instances is reduced by at least 61.25%.

关键词： single-machine scheduling sequence-dependent setup time neighborhood combination search token-ring search hybrid evolutionary algorithm

来源：评论

学校读者我要写书评

暂无评论

iDrone: IoT-Enabled Unmanned Aerial Vehicles for Detecting Wildfires Using Convolutional Neural Networks

引用

SN computer science 2022年第3期3卷 242页

作者： Sethuraman, Sibi Chakkaravarthy Tadkapally, Gaurav Reddy Mohanty, Saraju P. Subramanian, Anitha School of Computer Science and Engineering and Center of Excellence in Artificial Intelligence and Robotics (AIR) VIT-AP University Amaravati India Department of Computer Science and Engineering University of North Texas Denton 76207 TX United States School of Electronics Engineering and Center of Excellence in Artificial Intelligence and Robotics (AIR) VIT-AP University Amaravati India

The rise of global temperatures, over the past few decades, has disrupted the usual balance of nature. As a result of increasing temperatures, wildfires have destroyed millions of acres of land, thousands of structures, and homes. The pollution and toxic gases produced by the wildfires are carried out to thousands of miles, thus threatening the lives all around the world. Most wildfires occur due to anthropogenic factors, which cannot be predicted solely based on climate conditions. Henceforth, to detect wildfires before escalating, we propose iDrone, which is a wildfire detection system equipped with an end-to-end CNN image classification model: XtinguishNet, trained on a wildfire imagery dataset to detect the possible flames or smokes in an image. In addition, our approach also acquires the weather data and the intensity of the fire. Contrasting with existing wildfire detection systems, our proposed solution is a fusion of the Internet of Things (IoT) and Deep Learning, aiming to provide a one-stop solution for all the needs required to minimize the damage caused by wildfires. When validated and tested using various benchmark datasets, video surveillance, iDrone acquired a high accuracy of 98.36% with the least computational power. © 2022, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.

关键词： CNN computer vision Deep learning Forest fires iDrone Wildfires

来源：评论

学校读者我要写书评

暂无评论

Low-shot Video Object Segmentation

引用

IEEE Transactions on Pattern Analysis and Machine Intelligence 2025年第7期47卷 5538-5555页

作者： Yan, Kun Wei, Fangyun Dai, Shuyu Wu, Minghui Wang, Ping Xu, Chang Peking University School of Computer Science Beijing100871 China University of Sydney School of Computer Science DarlingtonNSW2008 Australia Peking University School of Software and Microelectronics Beijing100871 China Shanghai Artificial Intelligence Laboratory and Mininglamp Technology China Peking University National Engineering Research Center for Software Engineering Beijing100871 China Peking University School of Software and Microelectronics China China

Prior research in video object segmentation (VOS) predominantly relies on videos with dense annotations. However, obtaining pixel-level annotations is both costly and time-intensive. In this work, we highlight the potential of effectively training a VOS model using remarkably sparse video annotations - specifically, as few as one or two labeled frames per training video, yet maintaining near equivalent performance levels. We introduce this innovative training methodology as low-shot video object segmentation, abbreviated as low-shot VOS. Central to this method is the generation of reliable pseudo labels for unlabeled frames during the training phase, which are then used in tandem with labeled frames to optimize the model. Notably, our strategy is extremely simple and can be incorporated into the vast majority of current VOS models. For the first time, we propose a universal method for training VOS models on one-shot and two-shot VOS datasets. In the two-shot configuration, utilizing just 7.3% and 2.9% of labeled data from the YouTube-VOS and DAVIS benchmarks respectively, our model delivers results on par with those trained on completely labeled datasets. It is also worth noting that in the one-shot setting, a minor performance decrement is observed in comparison to models trained on fully annotated datasets. Code and models are available at https://***/yk-pku/Low-shot-VOS. © 1979-2012 IEEE.

关键词： Video analysis

来源：评论

学校读者我要写书评

暂无评论

Study on the Influence of Embodied Avatars on Gait Parameters in Virtual Environments and Real World

Study on the Influence of Embodied Avatars on Gait Parameter...

引用

2024 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2024

作者： Zhou, Tianyi Ding, Ding Wang, Shengyu Shi, Chuhan Xu, Xiangyu School of Computer Science and Engineering Southeast University Nanjing China Southeast University Computer Experimental Teaching Center Nanjing China

ISBN: (纸本)9781665410205

In this study, we compare the virtual and real gait parameters to investigate the effect of appearances of embodied avatars and virtual reality experience on gait in physical and virtual environments. We developed a virtual environment simulation and gait detection system for analyzing gait. The system transfers real-life scenarios into a realistic presentation in the virtual environment and provides look-alike same-age and old-age avatars for participants. We conducted an empirical study and used subjective questionnaires to evaluate participants' feelings about the virtual reality experience. Also, the paired sample t-test and neural network were implemented to analyze gait differences. The results suggest that there are disparities in gait between virtual and real environments. Also, the appearance of embodied avatars could influence the gait parameters in the virtual environment. Moreover, the experience of embodying old-age avatars affects the gait in the real world. © 2024 IEEE.

关键词： Virtual environments

来源：评论

学校读者我要写书评

暂无评论

A Deep Understanding Video Q&A System for Film Education in Acting Department

A Deep Understanding Video Q&A System for Film Education in ...

引用

2023 International Conference on Intelligent Education and Intelligent Research, IEIR 2023

作者： Wu, Zhengqian Li, Ruizhe Guo, Jiahao Wang, Zhongyuan Liang, Chao Hubei Key Laboratory of Multimedia and Network Communication Engineering School of Computer Science China National Engineering Research Center for Multimedia Software School of Computer Science China Wuhan University School of Computer Science China

ISBN: (纸本)9798350342895

Recently, advancements in artificial intelligence technology have greatly influenced the field of education, particularly in the area of intelligent homework assistance. However, current approaches are primarily designed for procedural and logical tasks and often lack comprehension abilities. This limitation is particularly evident when it comes to multi-hop and continuous tasks. To address this challenge, the integration of Large Language Model (LLM) has significantly enhanced the capability of AI systems to handle multi-hop and highly interconnected inputs. In this study, we focus on the learning needs of students in Acting Department, specifically their study of movies and the significance of classic movie videos in their learning process. However, assessing deep comprehension of classic movies poses its own challenges. To overcome these challenges, we develop a quiz system utilizing Knowledge Graphs (KG) and LLM to facilitate a deeper understanding of classic films. The generation of video quiz pairs is achieved through the use of Automatic Speech Recognition (ASR) technology, which leverages movie subtitles for question generation. For answering these questions, we employ techniques KG and LLM to process questions and retrieve corresponding answers. The proposed method achieves good performance in Deep Video Understanding (DVU) task of NIST TRECVID, demonstrating its effectiveness. © 2023 IEEE.

关键词： Motion pictures

来源：评论

学校读者我要写书评

暂无评论

A study on the noise reduction method in turbid water environments using Spatial frequency characteristics 9

A study on the noise reduction method in turbid water enviro...

引用

9th International Conference on Intelligent Informatics and BioMedical sciences, ICIIBMS 2024

作者： Koura, Ayumu Kim, Hyun-Woo Cho, Myungjin Lee, Min-Chul Kyushu Institute of Technology Graduate School of Computer Science and Systems Engineering Iizuka city Japan Hankyong National University School of ICT Robotics and Mechanical Engineering Iitc Anseong-si Korea Republic of

ISBN: (纸本)9798350363043

Scattering noise reduction is a challenging project to remove noise under scattering media conditions such as fog or turbid water. In previous study, blurring caused by scattering medium particles in fog or turbid water was removed by low-pass filtering. However, in this method, the accurate color information of the object may be lost due to rough low-frequency filtering. In addition, it is difficult to the obtain object shape information in heavy turbid environments. To solve this problem, we propose a new de-scattering algorithm that combines Fourier-domain filtering and integral imaging by creating a filter tailored to the frequency to be removed. To prove the proposed method, we perform a numerical analysis of experimental results using conventional and proposed methods. © 2024 IEEE.

关键词： Noise abatement

来源：评论

学校读者我要写书评

暂无评论

A study on Image Processing Method for Noise Reduction in Digital Holographic Microscopy 9

A study on Image Processing Method for Noise Reduction in Di...

引用

9th International Conference on Intelligent Informatics and BioMedical sciences, ICIIBMS 2024

作者： Nagashima, Masaharu Kim, Hyun-Woo Cho, Myungjin Lee, Min-Chul Kyushu Institute of Technology Graduate School of Computer Science and Systems Engineering Iizuka city Japan Iitc Hankyong National University School of Ict Robotics and Mechanical Engineering Anseong-si Korea Republic of

ISBN: (纸本)9798350363043

Digital Holographic Microscopy (DHM) can obtain three-dimensional (3D) information about the fine structure of an object by utilizing the phase information of coherent light. In DHM, during the process of reconstructing the shape of an object, phase information is extracted by windowing the sideband containing the phase information. The 3D information of the object is then obtained from the acquired phase information. When windowing the sideband in a narrow range, it can remove noise caused by the DC spectrum, but detailed shape information of the object is lost. On the other hand, when windowing in a wide range, detailed shape information of the object can be obtained, but noise from the DC spectrum increases. To solve this trade-off relationship, we propose a method of determining the range for windowing the sideband using a function and then averaging the windowed sidebands. This proposed method can reduce the impact of noise from the DC spectrum and enables the acquisition of the detailed shape information. In this paper, we explain the reconstruction algorithm of this method and compare it with conventional methods and the previous study, phase differences averaging (PDA). © 2024 IEEE.

关键词： Noise abatement

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：