检索结果-内蒙古大学图书馆

9th CCF Conference on big data, bigdata 2021

作者： Luo, Zhenxiao Wang, Xinglong Zhang, Yuxiao Wang, Zelong Hu, Miao Wu, Di School of Computer Science and Engineering Sun Yat-Sen University Guangzhou510006 China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou510006 China

ISBN: (纸本)9789811697081

Real-time video analytics is a killer application for edge computing, however, it has not been fully understood how much edge resource is required to support compute-intensive video analytics tasks at the edge. In this paper, we conduct an in-depth measurement study to unveil the resource requirement of deep learning (DL) inference for edge video analytics (EVA). We measure the DL inference time under various resource combinations, with different physical resources (e.g., processor types, core numbers, memory), video resolutions, and DL inference models. It is observed that the relationship between DL inference time and resource configuration is complicated, which cannot be captured by a simple model. Considering the coupling effects, we model the resource requirements of the DL inference by utilizing tensor factorization (TF) method. Our TF-based DL inference model can match well with the observations from field measurements. Different from previous models, our model is completely explainable and all the components have their physical meanings. Especially, it is possible to match the top three components extracted from the TF-based model with the practical DL functional layers, namely, convolutional layer, pooling layer, and fully-connected layer. The measurement results show that the top three components contribute approximately 50%, 25%, and 10% time for DL inference execution, which provides clear instruction on fine-grained DL task scheduling for edge video analytics applications. © 2022, Springer Nature Singapore Pte Ltd.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

3d-Usanet：A Spatio-Temporal Coupled Based Transformer for Video Anomaly Detection Method

SSRN

引用

SSRN 2024年

作者： Mu, Huiyu Wang, Luhui Yin, Hongjian Li, Yonggan Dang, Lanxue Liu, Yang Zuo, Xianyu School of Computer and Information Engineering Henan University Kaifeng475004 China Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng475004 China Fair Competition Review Transaction Center of Henan Province Zhengzhou450008 China

Video anomaly detection aims to detect incidental and sudden patterns of events in complex scenes, which plays an essential role for maintaining public security and averting potential risks. Unsupervised learning is particularly effective in solving the problem of insufficient annotation of video abnormal behaviour in low resource scenarios. However，most relate methods decompose it into two sub-tasks: spatio anomaly detection (S-VAD) and temporal anomaly detection (T-VAD), which barely consider the coupled correlations and the difficulty of coupling global spatio-temporal information. To address this, we provide a new perspective on constructing a novel dynamic spatio-temporal coupling relationship by designing a 3D U-shape Attention Network (3D-UsANet). On the one hand, we construct a Spatio-Temporal Feature Modelling Transformer (STFMT) module to handle different spatio positions at the same time step. On the other hand, a Channel-wise Scale-aware Fusion Modelling Transformer (CSFMT) module is integrated into 3D-UsANet to overcome the multi-scale heterogeneous divides and disregard irrelevant characteristics, which adaptively captures global semantic dependencies and bridges the spatio-temporal semantic gap. Essentially, we evaluate the performance of 3D-UsANet on UCSD Ped1、UCSD Ped2 and CUHK Avenue datasets with the AUC score 83.50%、96.30%、86.70%，respectively. The extensive experimental results demonstrate the effectiveness of 3D-UsANet in detecting anomaly behaviors. © 2024, The Authors. All rights reserved.

关键词： Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

A Model robustness optimization method based on adversarial sample detection 5

A Model robustness optimization method based on adversarial ...

引用

5th International Conference on Artificial Intelligence and Pattern Recognition, AIPR 2022

作者： Sun, Jiaze Long, Siyuan Ma, Xianyan Tang, Yanmei Xi'an University of Posts and Telecommunications Shaanxi Provincial Key Laboratory of Network Data Analysis and Intelligent Processing Xi'an Key Laboratory of Big Data and Intelligent Computing Xi'an710121 China Xi'an University of Posts and Telecommunications Xi'an710121 China

ISBN: (纸本)9781450396899

Deep neural networks are extremely vulnerable due to the existence of adversarial samples. It is a challenging problem to optimize the robustness of the model to protect deep neural networks from the threat of adversarial samples. To improve the model robustness, an integrated detection model of adversarial samples is designed to detect the existence of adversarial samples, which consists of a multi-classification detector and five single-classification detectors to perform double-layer detection of adversarial samples and intercept the adversarial samples finally sent to the image classification model. The detection experiments were conducted on the CIFAR-10 dataset for the adversarial samples generated by five attack algorithms: FGSM, BIM, DeepFool, JSMA, and C&W, and the detection success rate for all types of adversarial samples reached over 98.96%. After that, secondary attack experiments were conducted on the model, and the detection success rate of the model for the second attack adversarial samples reached over 92.36%. It provides an efficient and robust optimization method for deep neural network models in adversarial environments. © 2022 ACM.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Few-Shot Node Classification on Attributed Networks Based on Prototypical Network

Few-Shot Node Classification on Attributed Networks Based on...

引用

Symposia and Workshops on Ubiquitous, Autonomic and Trusted Computing, UIC-ATC

作者： Yaliang Zhao Guangming Zhang Jinke Wang Henan Key Laboratory of Big Data Analysis and Processing School of Computer and Information Engineering Henan University Kaifeng China

In today’s society, the interactions between various entities form complicated attributed networks. The node classification task as the core task of analyzing attribute networks has gained a considerable amount of attention from researchers. The existing graph neural network-based node classification approaches all demand lots of labeled data, but a high percentage of node classes in practical scenarios contain only restricted labeled data. So their classification effect is not good. In response to this problem, we put forward an improved graph prototypical network (IGPN) to enhance the few-shot node classification’s accuracy. In particular, we start by obtaining the embedding representations of nodes on the attributed network through a graph attention network. Then combined with the node’s degree information and neighbor information, the importance coefficient of all nodes in the support set is learned. After that, we can obtain the prototype representation of every class. Finally, the node classification can be completed by the distance of the query nodes’ representation from all class prototypes. Our method is preferable to the baseline method on three datasets according to a large number of experiments.

关键词：

来源：评论

学校读者我要写书评

暂无评论

GraphSparseNet: a Novel Method for Large Scale Trafffic Flow Prediction

arXiv

引用

arXiv 2025年

作者： Kong, Weiyang Wu, Kaiqi Zhang, Sen Liu, Yubao Sun Yat-Sen University Guangzhou China Sun Yat-Sen University Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China

Traffic flow forecasting is a critical spatio-temporal data mining task with wide-ranging applications in intelligent route planning and dynamic traffic management. Recent advancements in deep learning, particularly through Graph Neural Networks (GNNs), have significantly enhanced the accuracy of these forecasts by capturing complex spatio-temporal dynamics. However, the scalability of GNNs remains a challenge due to their exponential growth in model complexity with increasing nodes in the graph. Existing methods to address this issue, including sparsification, decomposition, and kernel-based approaches, either do not fully resolve the complexity issue or risk compromising predictive accuracy. This paper introduces GraphSparseNet (GSNet), a novel framework designed to improve both the scalability and accuracy of GNN-based traffic forecasting models. GraphSparseNet is comprised of two core modules: the Feature Extractor and the Relational Compressor. These modules operate with linear time and space complexity, thereby reducing the overall computational complexity of the model to a linear scale. Our extensive experiments on multiple real-world datasets demonstrate that GraphSparseNet not only significantly reduces training time by 3.51x compared to state-of-the-art linear models but also maintains high predictive performance. Copyright © 2025, The Authors. All rights reserved.

关键词： Scalability

来源：评论

学校读者我要写书评

暂无评论

Research progress on an intelligent production and distribution of the remote sensing products

Research progress on an intelligent production and distribut...

引用

International Conference on Applications and Techniques in Cyber Intelligence, ATCI 2020

作者： Kong, Dezhen Zhao, Jinhuan Tian, Meng Liu, Yang Henan Province Space Information Processing Engineering Laboratory KaifengHenan475004 China Henan Key Laboratory of Big Data Analysis and Processing KaifengHenan475004 China School of Computer and Information Engineering Henan University KaifengHenan475004 China

ISBN: (纸本)9783030539795

Since the amount of remote sensing data that can be obtained every day has reached ten terabytes, the storage and sharing of massive remote sensing data and the construction of product production application architecture have become imminent research topics. The purpose of this study is to build an efficient system that can meet the needs of massive remote sensing data storage, calculation, and distribution. The article elaborates on the latest research progress of cloud service remote sensing data distribution and intelligent production from the four aspects of remote sensing cloud service development, data distribution technology, cloud service technology, and operation mechanism of remote sensing cloud computing processing model, and proposes cloud service remote sensing key issues facing data distribution, and summarize and prospect related research, with a view to promoting the further development of research on cloud service remote sensing data distribution and intelligent production. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021.

关键词： Remote sensing

来源：评论

学校读者我要写书评

暂无评论

Contamination-Resilient Anomaly Detection via Adversarial Learning on Partially-Observed Normal and Anomalous data 41

Contamination-Resilient Anomaly Detection via Adversarial Le...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Lv, Wenxi Su, Qinliang Wan, Hai Xu, Hongteng Xu, Wenchao School of Computer Science and Engineering Sun Yat-sen University Guangzhou China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China Gaoling School of Artifical Intelligence Renmin University of China Beijing China Department of Computing The Hong Kong Polytechnic University Hong Kong

Many existing anomaly detection methods assume the availability of a large-scale normal dataset. But for many applications, limited by resources, removing all anomalous samples from a large unlabeled dataset is unrealistic, resulting in contaminated datasets. To detect anomalies accurately under such scenarios, from the probabilistic perspective, the key question becomes how to learn the normal-data distribution from a contaminated dataset. To this end, we propose to collect two additional small datasets that are comprised of partially-observed normal and anomaly samples, and then use them to help learn the distribution under an adversarial learning scheme. We prove that under some mild conditions, the proposed method is able to learn the correct normal-data distribution. Then, we consider the overfitting issue caused by the small size of the two additional datasets, and a correctness-guaranteed flipping mechanism is further developed to alleviate it. Theoretical results under incomplete observed anomaly types are also presented. Extensive experimental results demonstrate that our method outperforms representative baselines when detecting anomalies under contaminated datasets. Copyright 2024 by the author(s)

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

DreamFuse: Adaptive Image Fusion with Diffusion Transformer

arXiv

引用

arXiv 2025年

作者： Huang, Junjia Yan, Pengxiang Liu, Jiyang Wu, Jie Wang, Zhao Wang, Yitong Lin, Liang Li, Guanbin Sun Yat-sen University China ByteDance Intelligent Creation China Peng Cheng Laboratory China Guangdong Key Laboratory of Big Data Analysis and Processing China

Image fusion seeks to seamlessly integrate foreground objects with background scenes, producing realistic and harmonious fused images. Unlike existing methods that directly insert objects into the background, adaptive and interactive fusion remains a challenging yet appealing task. It requires the foreground to adjust or interact with the background context, enabling more coherent integration. To address this, we propose an iterative human-in-the-loop data generation pipeline, which leverages limited initial data with diverse textual prompts to generate fusion datasets across various scenarios and interactions, including placement, holding, wearing, and style transfer. Building on this, we introduce DreamFuse, a novel approach based on the Diffusion Transformer (DiT) model, to generate consistent and harmonious fused images with both foreground and background information. DreamFuse employs a Positional Affine mechanism to inject the size and position of the foreground into the background, enabling effective foreground-background interaction through shared attention. Furthermore, we apply Localized Direct Preference Optimization guided by human feedback to refine DreamFuse, enhancing background consistency and foreground harmony. DreamFuse achieves harmonious fusion while generalizing to text-driven attribute editing of the fused results. Experimental results demonstrate that our method outperforms state-of-the-art approaches across multiple metrics. Copyright © 2025, The Authors. All rights reserved.

关键词： Image fusion

来源：评论

学校读者我要写书评

暂无评论

Competitive and exclusion and oscillation phenomena of an eco-epidemiological model: An invasive approach

引用

International Journal of Biomathematics 2023年第5期16卷 23-47页

作者： Ke-Lu Li Jun-Yuan Yang Xue-Zhi Li School of Mathematics and Information Science Henan Normal University 453007XinriangHenanP.R.China Comple Systems Research Center Shanri University 030006TaiyuanShanciP.R.China Shanci Key Laboratory of Mathematical Techniques and Big Data Analysis on Disease Control and Prevention Shanri University 030006TaiyuanShanriP.R.China School of Statistics and Mathematics Henan Finance University 450046ZhengzhouHenanP.R.China

Eco-epidemiological model plays a crucial role in maintaining ecological balance and biological *** this paper,we propose a predator-prey model where the predator population is infected by the *** non-negativity and boundedness of the solution are *** existence and stability of equilibrium are discussed by an invasive *** simulations are used to demonstrate the competitive and exclusion,and oscillation behaviors of the ***,the predation rate of the infected predator is shown to have far-reaching implications for the structure and stability of ecological communities.

关键词： Eco-epidemiological model Hopf bifurcation competitive and exclusion global stability

来源：评论

学校读者我要写书评

暂无评论

A Fine-Grained Anomaly Detection Method Fusing Isolation Forest and Knowledge Graph Reasoning 19th

A Fine-Grained Anomaly Detection Method Fusing Isolation For...

引用

19th International Conference on Web Information Systems and Applications, WISA 2022

作者： Xu, Jie Zhou, Jiantao College of Computer Science Engineering Research Center of Ecological Big Data Ministry of Education National and Local Joint Engineering Research Center of Mongolian Intelligent Information Processing Technology Inner Mongolia Cloud Computing and Service Software Engineering Laboratory Inner Mongolia Social Computing and Data Processing Key Laboratory Inner Mongolia Discipline Inspection and Supervision Big Data Key Laboratory Inner Mongolia Big Data Analysis Technology Engineering Laboratory Inner Mongolia University Hohhot China

ISBN: (纸本)9783031203084

Anomaly detection aims to find outliers data that do not conform to expected behaviors in a specific scenario, which is indispensable and critical in current safety environments related studies. However, when performing outlier detection on a large-scale multidimensional dataset, most of the traditional methods make no distinctions between classes of outliers and lack the reasoning ability and the explainability of the analysis results, which leads to the low accuracy of global outlier detection and the inability to effectively identify local outliers. In this paper, we propose a fine-grained anomaly detection method which well utilizes isolation forest and knowledge graph reasoning tactics. First, a better extracting data feature method and a more reasonable weighted strategy are used to find global outliers based on the traditional isolation forest algorithm, and then we construct a custom rule base according to the in-depth research and analysis of global abnormal data, and at last the ontology knowledge is reasoned based on such rule base to achieve the detection of local outliers. Through analysis of extensive experiment results, our method could effectively discover more abnormal data without big loss of time costs, has strong generalization, and correspondingly improves the performance for abnormal detection. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Anomaly detection

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：