检索结果-内蒙古大学图书馆

Global video object segmentation with spatial constraint module

Computational Visual Media 2023年第2期9卷 385-400页

作者： Yadang Chen Duolin Wang Zhiguo Chen Zhi-Xin Yang Enhua Wu Engineering Research Center of Digital Forensics Ministry of EducationSchool of Computer and SoftwareNanjing University of Information Science and TechnologyNanjing 210044China State Key Laboratory of Internet of Things for Smart City Department of Electromechanical EngineeringUniversity of MacaoMacao 999078China State Key Laboratory of Computer Science Institute of SoftwareUniversity of Chinese Academy of SciencesBeijing 100190China Faculty of Science and Technology University of MacaoMacao 999078China

We present a lightweight and efficient semisupervised video object segmentation network based on the space-time memory *** some extent,our method solves the two difficulties encountered in traditional video object segmentation:one is that the single frame calculation time is too long,and the other is that the current frame’s segmentation should use more information from past *** algorithm uses a global context(GC)module to achieve highperformance,real-time *** GC module can effectively integrate multi-frame image information without increased memory and can process each frame in real ***,the prediction mask of the previous frame is helpful for the segmentation of the current frame,so we input it into a spatial constraint module(SCM),which constrains the areas of segments in the current *** SCM effectively alleviates mismatching of similar targets yet consumes few additional *** added a refinement module to the decoder to improve boundary *** model achieves state-of-the-art results on various datasets,scoring 80.1%on YouTube-VOS 2018 and a J&F score of 78.0%on DAVIS 2017,while taking 0.05 s per frame on the DAVIS 2016 validation dataset.

关键词： video object segmentation semantic segmentation global context(GC)module spatial constraint

来源：评论

学校读者我要写书评

暂无评论

LigCDnet:Remote Sensing Image Cloud Detection Based on Lightweight Framework 30th

LigCDnet:Remote Sensing Image Cloud Detection Based on Ligh...

引用

30th International Conference on MultiMedia Modeling, MMM 2024

作者： Su, Baotong Zheng, Wenguang School of Computer Science and Engineering Tianjin University of Technology TianJin300384 China Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology Tianjin University of Technology TianJin300384 China

ISBN: (纸本)9783031533075

Cloud contamination is inevitable in remote sensing images, resulting in a large number of images that cannot be applied in various fields. Therefore, cloud detection is one of the important tasks in remote sensing image preprocessing, aimed at removing images obstructed by clouds. Most existing methods are mostly based on CNN and feature a complex network structure, requiring a significant amount of computational resources, making it challenging to deploy them in practical applications. To tackle this problem, we propose a lightweight cloud detection framework (LigCDnet) with a lightweight feature extraction module (LFEM), a channel attention module (CAM), and a lightweight feature pyramid module (LFPM). The LFEM serves as the backbone of the network to capture rich spatial and contextual information;the CAM adaptively adjusts the channel weights of the feature maps;and the LFPM extracts cloud features at multiple scales. The effectiveness of our approach is evaluated on two public datasets, GF-1 and LandSat8. Extensive experiments have demonstrated that the proposed LigCDnet achieves state-of-the-art detection accuracy while significantly reducing computational burden and having a smaller model size. © 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Adversarial image detection based on the maximum channel of saliency maps

引用

Optoelectronics Letters 2022年第5期18卷 307-312页

作者： FU Haoran WANG Chundong LIN Hao HAO Qingbo Key Laboratory of Computer Vision and System Tianjin Key Laboratory of Intelligence Computing and Novel Software TechnologyTianjin University of TechnologyTianjin 300384China

Studies have shown that deep neural networks(DNNs) are vulnerable to adversarial examples(AEs) that induce incorrect behaviors. To defend these AEs, various detection techniques have been developed. However, most of them only appear to be effective against specific AEs and cannot generalize well to different AEs. We propose a new detection method against AEs based on the maximum channel of saliency maps(MCSM). The proposed method can alter the structure of adversarial perturbations and preserve the statistical properties of images at the same time. We conduct a complete evaluation on AEs generated by 6 prominent adversarial attacks on the Image Net large scale visual recognition challenge(ILSVRC) 2012 validation sets. The experimental results show that our method performs well on detecting various AEs.

关键词： networks neural maps

来源：评论

学校读者我要写书评

暂无评论

GridDroid--An Effective and Efficient Approach for Android Repackaging Detection Based on Runtime Graphical User Interface

引用

Journal of computer science & technology 2022年第1期37卷 147-181页

作者： Jun Ma Qing-Wei Sun Chang Xu Xian-Ping Tao State Key Laboratory for Novel Software Technology Nanjing UniversityNanjing 210023China Department of Computer Science and Technology Nanjing UniversityNanjing 210023China Huatai Securities Co. Ltd.Nanjing 210019China

Repackaging brings serious threats to Android *** birthmark techniques are typically applied to detect repackaged *** based on apps'runtime graphical user interfaces(GUI)are effective,especially for obfuscated or encrypted ***,existing studies are time-consuming and not suitable for handling apps in large *** this paper,we propose an effective yet efficient dynamic GUI birthmark for Android ***,we run an app with automatically generated GUI events and dump its layout after each *** divide each dumped layout into a grid,count in each grid cell the vertices of boundary rectangles corresponding to widgets within the layout,and generate a feature vector to encode the *** layouts are merged at runtime,and finally we obtain a graph as the birthmark of the *** a pair of apps to be compared,we build a weighted bipartite graph from their birthmarks and apply a modified version of the maximum-weight-bipartite-matching algorithm to determine whether they form a repackaging pair(RP)or *** implement the proposed technique in a prototype,GridDroid,and apply it to detect RPs in three datasets involving 527 *** reports only six false negatives and seven false positives,and it takes GridDroid merely 20 microseconds on average to compare a pair of birthmarks.

关键词： Android apps dynamic software birthmark graphical user interface(GUI) repackaging detection

来源：评论

学校读者我要写书评

暂无评论

SMART: Speedup Job Completion Time by Scheduling Reduce Tasks

引用

Journal of computer science & technology 2022年第4期37卷 763-778页

作者： Jia-Qing Dong Ze-Hao He Yuan-Yuan Gong Pei-Wen Yu Chen Tian Wan-Chun Dou Gui-Hai Chen Nai Xia Hao-Ran Guan State Key Laboratory of Media Convergence and Communication Communication University of China Beijing 100024China State Key Laboratory for Novel Software Technology Nanjing UniversityNanjing 210023China School of Computer Science The University of SydneySydney NSW 2006Australia

Distributed computing systems have been widely used as the amount of data grows exponentially in the era of information explosion. Job completion time (JCT) is a major metric for assessing their effectiveness. How to reduce the JCT for these systems through reasonable scheduling has become a hot issue in both industry and academia. Data skew is a common phenomenon that can compromise the performance of such distributed computing systems. This paper proposes SMART, which can effectively reduce the JCT through handling the data skew during the reducing phase. SMART predicts the size of reduce tasks based on part of the completed map tasks and then enforces largest-first scheduling in the reducing phase according to the predicted reduce task size. SMART makes minimal modifications to the original Hadoop with only 20 additional lines of code and is readily deployable. The robustness and the effectiveness of SMART have been evaluated with a real-world cluster against a large number of datasets. Experiments show that SMART reduces JCT by up to 6.47%, 9.26%, and 13.66% for Terasort, WordCount and InvertedIndex respectively with the Purdue MapReduce benchmarks suite (PUMA) dataset.

关键词： job scheduling job completion time MapReduce Hadoop

来源：评论

学校读者我要写书评

暂无评论

A survey for light field super-resolution

引用

High-Confidence Computing 2024年第1期4卷 118-129页

作者： Mingyuan Zhao Hao Sheng Da Yang Sizhe Wang Ruixuan Cong Zhenglong Cui Rongshan Chen Tun Wang Shuai Wang Yang Huang Jiahao Shen State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and EngineeringBeihang UniversityBeijing 100191China Key Laboratory of Data Science and Intelligent Computing International Innovation InstituteBeihang UniversityHangzhou 311115China State Key Laboratory of Software Development Environment School of Computer Science and EngineeringBeihang UniversityBeijing 100191China

Compared to 2D imaging data,the 4D light field(LF)data retains richer scene’s structure information,which can significantly improve the computer’s perception capability,including depth estimation,semantic segmentation,and LF ***,there is a contradiction between spatial and angular resolution during the LF image acquisition *** overcome the above problem,researchers have gradually focused on the light field super-resolution(LFSR).In the traditional solutions,researchers achieved the LFSR based on various optimization frameworks,such as Bayesian and Gaussian *** learning-based methods are more popular than conventional methods because they have better performance and more robust generalization *** this paper,the present approach can mainly divided into conventional methods and deep learning-based *** discuss these two branches in light field spatial super-resolution(LFSSR),light field angular super-resolution(LFASR),and light field spatial and angular super-resolution(LFSASR),***,this paper also introduces the primary public datasets and analyzes the performance of the prevalent approaches on these ***,we discuss the potential innovations of the LFSR to propose the progress of our research field.

关键词： Light field super-resolution Convolutional neural network Transformer Sub-aperture image Epipolar-plane image

来源：评论

学校读者我要写书评

暂无评论

Multiuser Computation Offloading for Long-Term Sequential Tasks in Mobile Edge Computing Environments

引用

Tsinghua science and technology 2023年第1期28卷 93-104页

作者： Huanhuan Xu Jingya Zhou Wenqi Wei Baolei Cheng School of Computer Science and Technology Soochow UniversitySuzhou 215006China State Key Laboratory of Mathematical Engineering and Advanced Computing Wuxi 214125China Provincial Key Laboratory for Computer Information Processing Technology Soochow UniversitySuzhou 215006China School of Computer Science Georgia Institute of TechnologyAtlantaGA 30332USA

Mobile edge computing has shown its potential in serving emerging latency-sensitive mobile applications in ultra-dense 5G networks via offloading computation workloads from the remote cloud data center to the nearby network ***,current computation offloading studies in the heterogeneous edge environment face multifaceted challenges:Dependencies among computational tasks,resource competition among multiple users,and diverse long-term *** applications typically consist of several functionalities,and one huge category of the applications can be viewed as a series of sequential *** this study,we first proposed a novel multiuser computation offloading framework for long-term sequential ***,we presented a comprehensive analysis of the task offloading process in the framework and formally defined the multiuser sequential task offloading ***,we decoupled the long-term offloading problem into multiple single time slot offloading problems and proposed a novel adaptive method to solve *** further showed the substantial performance advantage of our proposed method on the basis of extensive experiments.

关键词： mobile edge computing sequential tasks computation offloading dependency

来源：评论

学校读者我要写书评

暂无评论

FDCPNet:feature discrimination and context propagation network for 3D shape representation

引用

虚拟现实与智能硬件(中英文) 2025年第1期7卷 83-94页

作者： Weimin SHI Yuan XIONG Qianwen WANG Han JIANG Zhong ZHOU State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and EngineeringBeihang UniversityBeijing 100191China

Background Three-dimensional(3D)shape representation using mesh data is essential in various applications,such as virtual reality and simulation *** methods for extracting features from mesh edges or faces struggle with complex 3D models because edge-based approaches miss global contexts and face-based methods overlook variations in adjacent areas,which affects the overall *** address these issues,we propose the Feature Discrimination and Context Propagation Network(FDCPNet),which is a novel approach that synergistically integrates local and global features in mesh *** FDCPNet is composed of two modules:(1)the Feature Discrimination Module,which employs an attention mechanism to enhance the identification of key local features,and(2)the Context Propagation Module,which enriches key local features by integrating global contextual information,thereby facilitating a more detailed and comprehensive representation of crucial areas within the mesh *** Experiments on popular datasets validated the effectiveness of FDCPNet,showing an improvement in the classification accuracy over the baseline ***,even with reduced mesh face numbers and limited training data,FDCPNet achieved promising results,demonstrating its robustness in scenarios of variable complexity.

关键词： 3D shape representation Mesh model MeshNet Feature discrimination Context propagation

来源：评论

学校读者我要写书评

暂无评论

Multi-Task Visual Semantic Embedding Network for Image-Text Retrieval

引用

Journal of computer science & technology 2024年第4期39卷 811-826页

作者： Xue-Yang Qin Li-Shuang Li Jing-Yao Tang Fei Hao Mei-Ling Ge Guang-Yao Pang School of Computer Science and Technology Dalian University of TechnologyDalian 116024China School of Computer Science Shaanxi Normal UniversityXi’an 710119China School of Computer Engineering Weifang UniversityWeifang 261061China Guangxi Colleges and Universities Key Laboratory of Intelligent Industry Software Wuzhou UniversityWuzhou 543002 China

Image-text retrieval aims to capture the semantic correspondence between images and texts,which serves as a foundation and crucial component in multi-modal recommendations,search systems,and online *** mainstream methods primarily focus on modeling the association of image-text pairs while neglecting the advantageous impact of multi-task learning on image-text *** this end,a multi-task visual semantic embedding network(MVSEN)is proposed for image-text ***,we design two auxiliary tasks,including text-text matching and multi-label classification,for semantic constraints to improve the generalization and robustness of visual semantic embedding from a training ***,we present an intra-and inter-modality interaction scheme to learn discriminative visual and textual feature representations by facilitating information flow within and between ***,we utilize multi-layer graph convolutional networks in a cascading manner to infer the correlation of image-text *** results show that MVSEN outperforms state-of-the-art methods on two publicly available datasets,Flickr30K and MSCOCO,with rSum improvements of 8.2%and 3.0%,respectively.

关键词： image-text retrieval cross-modal retrieval multi-task learning graph convolutional network

来源：评论

学校读者我要写书评

暂无评论

Enhancing OOD Generalization in Offline Reinforcement Learning with Energy-Based Policy Optimization 26

Enhancing OOD Generalization in Offline Reinforcement Learni...

引用

26th European Conference on Artificial Intelligence, ECAI 2023

作者： Cao, Hongye Yang, Shangdong Huo, Jing Chen, Xingguo Gao, Yang State Key Laboratory for Novel Software Technology Nanjing University China School of Computer Science Nanjing University of Posts and Telecommunications Nanjing China

ISBN: (纸本)9781643684369

Offline Reinforcement Learning (RL) is an important research domain for real-world applications because it can avert expensive and dangerous online exploration. Offline RL is prone to extrapolation errors caused by the distribution shift between offline datasets and states visited by behavior policy. Existing offline RL methods constrain the policy to offline behavior to prevent extrapolation errors. But these methods limit the generalization potential of agents in Out-Of-Distribution (OOD) regions and cannot effectively evaluate OOD generalization behavior. To improve the generalization of the policy in OOD regions while avoiding extrapolation errors, we propose an Energy-Based Policy Optimization (EBPO) method for OOD generalization. An energy function based on the distribution of offline data is proposed for the evaluation of OOD generalization behavior, instead of relying on model discrepancies to constrain the policy. The way of quantifying exploration behavior in terms of energy values can balance the return and risk. To improve the stability of generalization and solve the problem of sparse reward in complex environment, episodic memory is applied to store successful experiences that can improve sample efficiency. Extensive experiments on the D4RL datasets demonstrate that EBPO outperforms the state-of-the-art methods and achieves robust performance on challenging tasks that require OOD generalization. © 2023 The Authors.

关键词： Extrapolation

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：