检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Zhang, Wei Ma, Mengting Jiang, Yizhen Lian, Rongrong Wu, Zhenkai Cui, Kangning Ma, Xiaowen School of Software Technology Zhejiang University Hangzhou310027 China Innovation Center of Yangtze River Delta Zhejiang University Zhejiang Jiaxing314103 China School of Computer Science and Technology Zhejiang University Hangzhou310027 China Department of Mathematics City University of Hong Kong Hong Kong Noah’s Ark Lab Huawei Shanghai201206 China

Compared with natural images, remote sensing images (RSIs) have the unique characteristic. i.e., larger intraclass variance, which makes semantic segmentation for remote sensing images more challenging. Moreover, existing semantic segmentation models for remote sensing images usually employ a vanilla softmax classifier, which has three drawbacks: (1) non-direct supervision for the pixel representations during training;(2) inadequate modeling ability of parametric softmax classifiers under large intraclass variance;and (3) opaque process of classification decision. In this paper, we propose a novel classifier (called CenterSeg) customized for RSI semantic segmentation, which solves the abovementioned problems with multiple prototypes, direct supervision under Grassmann manifold, and interpretability strategy. Specifically, for each class, our CenterSeg obtains local class centers by aggregating corresponding pixel features based on ground-truth masks, and generates multiple prototypes through hard attention assignment and momentum updating. In addition, we introduce the Grassmann manifold and constrain the joint embedding space of pixel features and prototypes based on two additional regularization terms. Especially, during the inference, CenterSeg can further provide interpretability to the model by restricting the prototype as a sample of the training set. Experimental results on three remote sensing segmentation datasets validate the effectiveness of the model. Besides the superior performance, CenterSeg has the advantages of simplicity, lightweight, compatibility, and interpretability. Code is available at https://***/xwmaxwma/rssegmentation. © 2025, CC BY.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Composite Fault Diagnosis Based on Deep Convolutional Generative Adversarial Network

Composite Fault Diagnosis Based on Deep Convolutional Genera...

引用

2020 Asia-Pacific International Symposium on Advanced Reliability and Maintenance Modeling, APARM 2020

作者： Yonghong, Zhang Zhongyang, Zhang Fan, Shao Yifei, Wang Xiaoping, Zhao Kaiyang, Lv Nanjing University of Information Science and Technology School of Automation Nanjing China Nanjing University of Information Science and Technology School of Computer and Software Nanjing China

ISBN: (纸本)9781728171029

In the field of fault diagnosis, composite fault are difficult to diagnose accurately because of the coupling effect between various faults and dynamic working conditions (load, rotating speed, etc.). In addition, fault samples under specific operating conditions of practical manufacturing processes suffer from inadequacy and unbalanced distribution. Aiming at the problem of sample imbalance in the process of model building, this paper takes the bearing and gear of gearbox as the research object and presents a fault diagnosis method of the gearbox with less sample information. A Deep Convolutional Generative Adversarial Network model (DCGAN) is constructed for a more accurate fault diagnosis, where the counter-learning mechanism is used to train the generator to learn the distribution characteristic of the original fault samples and to generate additional fault samples. Fault samples under various working conditions were collected through the designed experimental platform for the training of the fault diagnosis model. The experiments show that the DCGAN fault diagnosis model can effectively improve fault diagnosis accuracy with insufficient diagnosis samples. © 2020 IEEE.

关键词： Failure analysis

来源：评论

学校读者我要写书评

暂无评论

Learning Pixel-wise Continuous Depth Representation via Clustering for Depth Completion

arXiv

引用

arXiv 2024年

作者： Chen, Shenglun Zhang, Hong Ma, Xinzhu Wang, Zhihui Li, Haojie Dalian University of Technology Dalian116620 China College of Computer Science and Engineering Shandong University of Science and Technology Qingdao266590 China Shanghai Artificial Intelligence Laboratory Shanghai AI Lab Shanghai200030 China DUT-RU International School of Information Science and Engineering Dalian University of Technology Dalian116620 China Key Laboratory for Ubiquitous Network and Service Software of Liaoning Province Dalian University of Technology Dalian116620 China

Depth completion is a long-standing challenge in computer vision, where classification-based methods have made tremendous progress in recent years. However, most existing classification-based methods rely on pre-defined pixel-shared and discrete depth values as depth categories. This representation fails to capture the continuous depth values that conform to the real depth distribution, leading to depth smearing in boundary regions. To address this issue, we revisit depth completion from the clustering perspective and propose a novel clustering-based framework called CluDe which focuses on learning the pixel-wise and continuous depth representation. The key idea of CluDe is to iteratively update the pixel-shared and discrete depth representation to its corresponding pixel-wise and continuous counterpart, driven by the real depth distribution. Specifically, CluDe first utilizes depth value clustering to learn a set of depth centers as the depth representation. While these depth centers are pixel-shared and discrete, they are more in line with the real depth distribution compared to pre-defined depth categories. Then, CluDe estimates offsets for these depth centers, enabling their dynamic adjustment along the depth axis of the depth distribution to generate the pixel-wise and continuous depth representation. Extensive experiments demonstrate that CluDe successfully reduces depth smearing around object boundaries by utilizing pixel-wise and continuous depth representation. Furthermore, CluDe achieves state-of-the-art performance on the VOID datasets and outperforms classification-based methods on the KITTI dataset. Copyright © 2024, The Authors. All rights reserved.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

Improving Variational Autoencoders with Density Gap-based Regularization

arXiv

引用

arXiv 2022年

作者： Zhang, Jianfei Bai, Jun Lin, Chenghua Wang, Yanmeng Rong, Wenge State Key Laboratory of Software Development Environment Beihang University China School of Computer Science and Engineering Beihang University China Department of Computer Science University of Sheffield United Kingdom Ping An Technology China

Variational autoencoders (VAEs) are one of the most powerful unsupervised learning frameworks in NLP for latent representation learning and latent-directed generation. The classic optimization goal of VAEs is to maximize the Evidence Lower Bound (ELBo), which consists of a conditional likelihood for generation and a negative Kullback-Leibler (KL) divergence for regularization. In practice, optimizing ELBo often leads the posterior distribution of all samples converging to the same degenerated local optimum, namely posterior collapse or KL vanishing. There are effective ways proposed to prevent posterior collapse in VAEs, but we observe that they in essence make trade-offs between posterior collapse and the hole problem, i.e., the mismatch between the aggregated posterior distribution and the prior distribution. To this end, we introduce new training objectives to tackle both problems through a novel regularization based on the probabilistic density gap between the aggregated posterior distribution and the prior distribution. Through experiments on language modeling, latent space visualization, and interpolation, we show that our proposed method can solve both problems effectively and thus outperforms the existing methods in latent-directed generation. To the best of our knowledge, we are the first to jointly solve the hole problem and posterior collapse. Copyright © 2022, The Authors. All rights reserved.

关键词： Economic and social effects

来源：评论

学校读者我要写书评

暂无评论

Application of Explicit Group Sparse Projection in Optimizing 5G Network Resource Allocation and Topology Design

Application of Explicit Group Sparse Projection in Optimizin...

引用

Innovation in technology (ASIANCON), Asian Conference on

作者： Kodanda Rami Reddy Manukonda Hemanth Swamy R. Sankar Chidambaranathan C. M. Haribabu.k Saranya. Department of Consulting IBM Senior Software Engineer Motorola Solutions Department of EEE Chennai Institute of Technology Chennai Department of Computer Science and Engineering School of Computing Vel Tech Rangarajan Dr.Sagunthala R&D Institute of Science and Technology Chennai India Department of ECE MLR Institute of Technology Hyderabad Department of AI & DS Dhanalakshmi College of Engineering Chennai

ISBN: (数字)9798350354218

ISBN: (纸本)9798350354225

The growing adoption of 5G networks in diverse domains like IoT, smart cities, healthcare, and industrial automation necessitates advanced optimization techniques. While existing approaches have shown promise, they are often limited by computational intensity, lack of adaptability, and specific focus on geometric constraints. This paper presents EGSP-5G, a novel methodology employing Explicit Group Sparse Projection (EGSP) for 5G network resource allocation and topology design. Its capability to achieve targeted sparsity levels allows for efficient utilization of network resources, thereby making it a versatile and powerful tool for various optimization tasks. Remarkably, EGSP-5G achieved a resource utilization rate of 90.02% and a throughput of 110.32 in networks comprising 1000 nodes. It stands as a robust and versatile solution for optimizing 5G networks across diverse application domains.

关键词： Technological innovation Automation 5G mobile communication Network topology Smart cities Medical services Throughput Topology Resource management Optimization

来源：评论

学校读者我要写书评

暂无评论

Temporal Feature Matters: A Framework for Diffusion Model Quantization

arXiv

引用

arXiv 2024年

作者： Huang, Yushi Gong, Ruihao Liu, Xianglong Liu, Jing Li, Yuhang Lu, Jiwen Tao, Dacheng Electrical and Computer Engineering Department Hong Kong University of Science and Technology Hong Kong State Key Laboratory of Software Development Environment Beihang University China Department of Data Science and AI Faculty of IT Monash University Australia Electrical Engineering Department Yale University United States Department of Automation Tsinghua University China School of Computer Science and Engineering Nanyang Technological University Singapore

The Diffusion models, widely used for image generation, face significant challenges related to their broad applicability due to prolonged inference times and high memory demands. Efficient Post-Training Quantization (PTQ) is crucial to address these issues. However, unlike traditional models, diffusion models critically rely on the time-step for the multi-round denoising. Typically, each time-step is encoded into a hypersensitive temporal feature by several modules. Despite this, existing PTQ methods do not optimize these modules individually. Instead, they employ unsuitable reconstruction objectives and complex calibration methods, leading to significant disturbances in the temporal feature and denoising trajectory, as well as reduced compression efficiency. To address these challenges, we introduce a novel quantization framework that includes three strategies: 1) TIB-based Maintenance: Based on our innovative Temporal Information Block (TIB) definition, Temporal Information-aware Reconstruction (TIAR) and Finite Set Calibration (FSC) are developed to efficiently align original temporal features. 2) Cache-based Maintenance: Instead of indirect and complex optimization for the related modules, pre-computing and caching quantized counterparts of temporal features are developed to minimize errors. 3) Disturbance-aware Selection: Employ temporal feature errors to guide a fine-grained selection between the two maintenance strategies for further disturbance reduction. This framework preserves most of the temporal information and ensures high-quality end-to-end generation. Extensive testing on various datasets, diffusion models and hardware confirms our superior performance and acceleration. © 2024, CC BY.

关键词： Spatio-temporal data

来源：评论

学校读者我要写书评

暂无评论

Out-of-Distribution Detection by Principal Component Correspondence

Out-of-Distribution Detection by Principal Component Corresp...

引用

IEEE International Conference on Multimedia and Expo (ICME)

作者： Xiaoyuan Guan Zhiyong Gan Ling Deng Wei Shi Jiankang Chen Shenshen Bu Chunliang Zhao Jianfang Hu Yuren Zhou Wei-Shi Zheng Ruixuan Wang School of Computer Science and Engineering Sun Yat-Sen University Guangzhou China Key Laboratory of Machine Intelligence and Advanced Computing MOE Guangzhou China Network Business Group China Unicom Guangzhou China Department of Network Intelligence Pengcheng Laboratory Shenzhen China School of Data and Science Qingdao University of Science and Technology Qingdao China School of Software Engineering Sun Yat-Sen University Zhuhai China

ISBN: (数字)9798350390155

ISBN: (纸本)9798350390162

Out-of-distribution (OOD) detection is vital for the safe application of intelligent systems in real-world scenarios. This paper proposes an enhancement to OOD detection by leveraging the consistency in cognition between two models, both pretrained on in-distribution (ID) data. Specifically, for a given test sample, we first apply Principal Component Analysis (PCA)-based projection on the feature vectors from each model. These obtained feature vectors (with correlation between dimensions decoupled by PCA projection) are then aligned using a multiple linear mapping, which is fitted using the least squares method on the training data. We hypothesize that the regression error for OOD data will be larger than that for ID data, making it a useful metric for OOD detection. Our experimental results demonstrate the effectiveness of this method. When combined with existing robust baselines, our approach achieves state-of-the-art performance in OOD detection.

关键词： Training Measurement Analytical models Training data Vectors Data models Intelligent systems Facial features Principal component analysis Least mean squares methods

来源：评论

学校读者我要写书评

暂无评论

MetaSlicing: A Novel Resource Allocation Framework for Metaverse

arXiv

引用

arXiv 2022年

作者： Chu, Nam H. Hoang, Dinh Thai Nguyen, Diep N. Phan, Khoa T. Dutkiewicz, Eryk Niyato, Dusit Shu, Tao The School of Electrical and Data Engineering University of Technology Sydney Australia School of Engineering and Mathematical Sciences Department of Computer Science and Information Technology La Trobe University Melbourne Australia The School of Computer Science and Engineering Nanyang Technological University Singapore639798 Singapore The Department of Computer Science and Software Engineering Auburn University AuburnAL36849 United States

Creating and maintaining the Metaverse requires enormous resources that have never been seen before, especially computing resources for intensive data processing to support the Extended Reality, enormous storage resources, and massive networking resources for maintaining ultra high-speed and low-latency connections. Therefore, this work aims to propose a novel framework, namely MetaSlicing, that can provide a highly effective and comprehensive solution in managing and allocating different types of resources for Metaverse applications. In particular, by observing that Metaverse applications may have common functions, we first propose grouping applications into clusters, called MetaInstances. In a MetaInstance, common functions can be shared among applications. As such, the same resources can be used by multiple applications simultaneously, thereby enhancing resource utilization dramatically. To address the real-time characteristic and resource demand's dynamic and uncertainty in the Metaverse, we develop an effective framework based on the semi-Markov decision process and propose an intelligent admission control algorithm that can maximize resource utilization and enhance the Quality-of-Service for end-users. Extensive simulation results show that our proposed solution outperforms the Greedy-based policies by up to 80% and 47% in terms of long-term revenue for Metaverse providers and request acceptance probability, respectively. Copyright © 2022, The Authors. All rights reserved.

关键词： Resource allocation

来源：评论

学校读者我要写书评

暂无评论

An SRN-Based Model for Assessing Co-Resident Attack Mitigation in Cloud with VM Migration and Allocation Policies*

An SRN-Based Model for Assessing Co-Resident Attack Mitigati...

引用

IEEE Conference on Global Communications (GLOBECOM)

作者： Xin Yang Abla Smahi Hui Li Huayu Zhang Shuo-Yen Robert Li Qingdao Institute of Software College of Computer Science and Technology China University of Petroleum (East China) Qingdao Shandong China Shenzhen Graduate School Peking University Shenzhen Guangdong China Purple Mountain Laboratories Jiangsu China University of Electronic Science & Technology of China Chengdu China

Cloud computing provides users with cost-effective on-demand resource sharing, but the shared resources also creates additional security risks due to co-location with malicious tenants. In addition to static defensive technologies, Virtual Machine (VM) migration, a type of Moving Target Defense (MTD) technique, provides an alternative solution to mitigate co-resident attacks by dynamically migrating services/tasks among various servers. Although significant progress has been made in this area, critical gaps remain regarding the quantification of these techniques' synergistic effectiveness in cloud computing, as previous research focused on either evaluating the defensive capability of MTD or on examining the characteristics of static strategies. In this paper, we aim to use analytical modeling techniques to quantitatively evaluate the synergistic effectiveness of integrating MTD with VM allocation policies. We designed a synergistically defensive architecture and proposed a Stochastic Reward Net (SRN) model to describe the execution and attacking processes of tasks migrating among multiple servers under three commonly used allocation policies. Moreover, experiments were implemented using SimPy to verify the analytical results obtained from the SRN model and to broaden the evaluation scope. Our comprehensive assessment, using both SRN and experiments, evaluates the defensive capacity and corresponding features for different scenarios. The obtained simulation results were approximate, with an error rate of less than 3.40 %, reflecting the reliability of our methods. Several defensive suggestions were further summarized based on the evaluation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Double-Flow-based Steganography without Embedding for Image-to-Image Hiding

arXiv

引用

arXiv 2023年

作者： Song, Bingbing Wang, Derui Zhang, Tianwei Liu, Renyang Lin, Yu Zhou, Wei The School of Information Science & Engineering Yunnan University Yunnan China Swinburne University of Technology Melbourne Australia School of Computer Science and Engineering Nanyang Technological University Singapore Kunming Institute of Physics Yunnan China The National Pilot School of Software and Engineering Research Center of Cyberspace Yunnan University Yunnan China

As an emerging concept, steganography without embedding (SWE) hides a secret message without directly embedding it into a cover. Thus, SWE has the unique advantage of being immune to typical steganalysis methods and can better protect the secret message from being exposed. However, existing SWE methods are generally criticized for their poor payload capacity and low fidelity of recovered secret messages. In this paper, we propose a novel steganography-without-embedding technique, named DF-SWE, which addresses the aforementioned drawbacks and produces diverse and natural stego images. Specifically, DF-SWE employs a reversible circulation of double flow to build a reversible bijective transformation between the secret image and the generated stego image. Hence, it provides a way to directly generate stego images from secret images without a cover image. Besides leveraging the invertible property, DF-SWE can invert a secret image from a generated stego image in a nearly lossless manner and increases the fidelity of extracted secret images. To the best of our knowledge, DF-SWE is the first SWE method that can hide large images and multiple images into one image with the same size, significantly enhancing the payload capacity. According to the experimental results, the payload capacity of DF-SWE achieves 24 − 72BPP is 8000 ∼ 16000 times compared to its competitors while producing diverse images to minimize the exposure risk. Importantly, DF-SWE can be applied in the steganography of secret images in various domains without requiring training data from the corresponding domains. This domain-agnostic property suggests that DF-SWE can 1) be applied to hiding private data and 2) be deployed in resource-limited systems. Copyright © 2023, The Authors. All rights reserved.

关键词： Steganography

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：