检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Tian, Lei Wang, Peng Liang, Guoqiang Shen, Chunhua School of Computer Science Northwestern Polytechnical University Xi’an China National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology China School of Computer Science The University of Adelaide Australia

Because of the invisible human keypoints in images caused by illumination, occlusion and overlap, it is likely to produce unreasonable human pose prediction for most of the current human pose estimation methods. In this paper, we design a novel generative adversarial network (GAN) to improve the localization accuracy of visible joints when some joints are invisible. The network consists of two simple but efficient modules, i.e., Cascade Feature Network (CFN) and Graph Structure Network (GSN). First, the CFN utilizes the prediction maps from the previous stages to guide the prediction maps in the next stage to produce accurate human pose. Second, the GSN is designed to contribute to the localization of invisible joints by passing message among different joints. According to GAN, if the prediction pose produced by the generator G cannot be distinguished by the discriminator D, the generator network G has successfully obtained the underlying dependence of human joints. We conduct experiments on three widely used human pose estimation benchmark datasets, i.e., LSP, MPII and COCO, whose results show the effectiveness of our proposed framework. © 2021, CC BY.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

A Survey of Deep Graph Learning under Distribution Shifts: from Graph Out-of-Distribution Generalization to Adaptation

arXiv

引用

arXiv 2024年

作者： Zhang, Kexin Liu, Shuhan Wang, Song Shi, Weili Chen, Chen Li, Pan Li, Sheng Li, Jundong Ding, Kaize Department of Statistics and Data Science Northwestern University EvanstonIL United States Department of Electrical and Computer Engineering University of Virginia CharlottesvilleVA United States Department of Computer Science School of Data Science University of Virginia CharlottesvilleVA United States School of Data Science University of Virginia CharlottesvilleVA United States Department of Computer Science University of Virginia CharlottesvilleVA United States School of Electrical and Computer Engineering Georgia Institute of Technology AtlantaGA United States Department of Computer Science Purdue University West LafayetteIN United States

Distribution shifts on graphs – the discrepancies in data distribution between training and employing a graph machine learning model – are ubiquitous and often unavoidable in real-world scenarios. These shifts may severely deteriorate model performance, posing significant challenges for reliable graph machine learning. Consequently, there has been a surge in research on graph machine learning under distribution shifts, aiming to train models to achieve satisfactory performance on out-of-distribution (OOD) test data. In our survey, we provide an up-to-date and forward-looking review of deep graph learning under distribution shifts. Specifically, we cover three primary scenarios: graph OOD generalization, training-time graph OOD adaptation, and test-time graph OOD adaptation. We begin by formally formulating the problems and discussing various types of distribution shifts that can affect graph learning, such as covariate shifts and concept shifts. To provide a better understanding of the literature, we introduce a systematic taxonomy that classifies existing methods into model-centric and data-centric approaches, investigating the techniques used in each category. We also summarize commonly used datasets in this research area to facilitate further investigation. Finally, we point out promising research directions and the corresponding challenges to encourage further study in this vital domain. We also provide a continuously updated reading list at https://***/kaize0409/Awesome-Graph-OOD. Copyright © 2024, The Authors. All rights reserved.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

High-Spatial-Resolution Remote Sensing Scene Method Based on Spatial Adaptive Convolution

High-Spatial-Resolution Remote Sensing Scene Method Based on...

引用

Evolutionary Algorithms and Soft Computing Techniques (EASCT), International Conference on

作者： Ayub Baig Bura Vijay Kumar K. Arun Maitrey Bhookya Ramesh Karunakara Rai B IT Department Malla Reddy Engineering College for Women Hyderabad India School of Computer Science and Artificial Intelligence SR University Warangal India Computer Science and Engineering (Data Science) Gokaraju Rangaraju Institute of Engineering and Technology Hyderabad India Department of Computer Science and Engineering (Data Science) CMR Technical Campus Hyderabad India Department of Electronics and Communication Engineering Nitte Meenakshi Institute of Technology Bengaluru India

The High-Spatial-Resolution (HSR) remote sensing is categorized through its ability to generate fine-grained spatial data. The image resolution is typically calculated through Ground Sample Distance (GSD) that presents the smallest ground object size that can be solved in an images. This paper presents a remote sensing image semantic segmentation for resolving issues through Hierarchical Atrous Pyramid (HASP) and Spatial Adaptive Network (SPANet) based framework. The HASP is utilized to resolve the issues of difficult background and high-scale variances through expanding the receptive domain of network. The SPANet has a capability to adjust convolutional operations adaptively that reduces the computational load. It can lead to quicker inference time and make it is feasible for real-time applications. The loss function is calculated which assists network to enhance the segmentation result. The parameters like accuracy, Intersection over Union (IoU), mean IoU (mIoU) and mean F1 (mF1) are utilized to estimating model performance on iSAID dataset. The SPANet obtains high accuracy of 98.74%, IoU of 84.68%, mIoU of 72.26% and mF1 of 81.59% which is comparatively higher than existing methods like Foreground-Aware Relation Network (FarSeg++) and Cascade Class-Aware Enhanced Network (CCENet).

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multimodal Imaging Feature Extraction with Reference Canonical Correlation Analysis Underlying Intelligence

Multimodal Imaging Feature Extraction with Reference Canonic...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Ram Sapkota Bishal Thapaliya Pranav Suresh Bhaskar Ray Vince D. Calhoun Jingyu Liu Translational Research in Neuroimaging and Data Science (TReNDS) Department of Computer Science Georgia State University Atlanta USA School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta USA

With neuroimaging data scientists have gained substantial information of the neuronal underpinning of intelligence. Yet how to integrate multimodal neuronal features effectively in relation to intelligence remains elusive. In this paper, we have developed a reference Canonical Correlation Analysis (RCCA) model that extracts latent, correlated multimodal features while enhancing correlation to a reference of interest. We applied RCCA to gray matter and white matter images from 7874 participants, and compared the derived features with those from Principle Components Analysis (PCA) and sparse CCA (SCCA), in terms of association with intelligence and prediction effectiveness using LASSO regression models. Eight RCCA features explained 10%, 16% and 17% variance of fluid intelligence, crystallized intelligence, and total composite score, respectively, which are similar to the percentage of variance explained by over 100 principle components. SCCA features presented the least variance of intelligence. Our results indicate RCCA model can successfully extract features of interest. The top brain regions that contribute to intelligence include the frontal regions and cingulate gyrus.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Kathakali Mudras: An AI-Enhanced Predictive Approach

Kathakali Mudras: An AI-Enhanced Predictive Approach

引用

International Conference on Soft Computing and Machine Intelligence (ISCMI)

作者： Jaison Paul Mulerikkal K Janani P.S Sreelakshmi Divya James K.S Lakshmi T.D Sajanraj A Binu Department of Information Technology Rajagiri School of Engineering & Technology Kochi India Department of Computer Science and Engineering Department of Computer Science and Business Systems Department of Artificial Intelligence and Data Science Department of Computer Science Loughborough University Leicester United Kingdom

ISBN: (数字)9798331518127

ISBN: (纸本)9798331518134

The goal of this research is to integrate an artificial intelligence framework for predicting Kathakali mudras, a crucial component of the traditional Indian dance style that is renowned for its complex hand and facial motions. Our goal is to protect and improve the accessibility of this traditional art form by utilising deep learning technology. To be more precise, we used the object identification model YOLOv8 together with convolutional neural networks (CNNs) like VGG 19, Inception V3, and ResNet 152 to predict and analyse Kathakali mudras. Furthermore, Few-shot learning (FSL) and support vector machines (SVMs) were used to offer a comparative viewpoint on the efficacy of machine learning method in this situation. The study entailed a thorough analysis of how well-performing different deep learning techniques performed. The predictive accuracy of VGG 19, Inception V3, and ResNet 152 for intricate mudras of Kathakali was subjected to a thorough evaluation process. In addition, an evaluation of YOLOv8’s real-time object identification capabilities was conducted in order to investigate its potential in live performance circumstances. The results of the experiment showed that VGG 19 performed better than the other models and had the highest accuracy in mudra prediction. As a result, VGG 19 was chosen as the best algorithm for this particular application. The results of this study demonstrate how artificial intelligence (AI) technologies may support and preserve traditional artistic forms. This AI-integrated framework is a useful tool for scholars, educators, and dancers, promoting a deeper knowledge and enjoyment of this cultural legacy through its accurate analysis and prediction of Kathakali mudras.

关键词： Support vector machines Hands Accuracy Predictive models Prediction algorithms Real-time systems Object recognition Cultural differences Few shot learning Machine intelligence

来源：评论

学校读者我要写书评

暂无评论

Rankboost-Based Result Merging

Rankboost-Based Result Merging

引用

IEEE International Conference on Ubiquitous Computing and Communications

作者： Benjamin Ghansah Shengli Wu Nathaniel Ghansah School of Computer Science and Telecommunication Engineering Jiangsu University 301 Xuefu Road 212013 Zhenjiang Jiangsu China School of Computer Science Data Link Institute P. O Box 2481 Tema Ghana

ISBN: (纸本)9781509001552

The explosion of searchable text content especially on the web has rendered information to be distributed among many disjoint text information sources (Federated Search). How to merge the results returned by selected sources is a major problem of the Federated Search task. We study the problem of learning to rank a set of objects by combining various sources of ranking. The problem of merging search results arises in several domains, for example combining the results of different verticals and also Meta search applications. This paper presents a supervised learning solution to the result merging problem. Our approach combines multiple sources of evidence to inform the merging decision. We use the Rankboost Method, a boosting approach to machine learning which learns a function that merges results based on information that is readily available: i.e. the ranks, titles, summaries, URLs and click- through data, which are found in the results pages. We combine these evidences by treating result merging as a multiclass machine learning problem. By not downloading additional information such as the full document, we decrease processing cost in terms of bandwidth usage and latency. We compare our results against existing result merging methods which rely on evidence found only in ranked lists;Semi-Supervised Learning (SSL), Sample-Agglomerate Fitting Estimate (SAFE) and CORI. An extensive set of experiments demonstrates that our method is more effective than the baseline result-merging algorithm under a variety of conditions.

关键词： Machine learning rankboost Distributed Information retrieval Result merging Information Retrieval

来源：评论

学校读者我要写书评

暂无评论

Resilient to Byzantine Attacks Finite-Sum Optimization Over Networks

Resilient to Byzantine Attacks Finite-Sum Optimization Over ...

引用

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Zhaoxian Wu Qing Ling Tianyi Chen Georgios B. Giannakis School of Data and Computer Science and Guangdong Province Key Laboratory of Computational Science Sun Yat-Sen University Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute Department of Electrical and Computer Engineering and Digital Technology Center University of Minnesota

ISBN: (数字)9781509066315

ISBN: (纸本)9781509066322

This contribution deals with distributed finite-sum optimization for learning over networks in the presence of malicious Byzantine attacks. To cope with such attacks, resilient approaches so far combine stochastic gradient descent (SGD) with different robust aggregation rules. However, the sizeable SGD-induced gradient noise makes it challenging to distinguish malicious messages sent by the Byzantine attackers from noisy stochastic gradients sent by the friendly workers. This motivates gradient noise reduction as a means of robustifying SGD in the presence of Byzantine attacks. To this end, the present work puts forth a Byzantine attack resilient distributed (Byrd-) SAGA approach for learning tasks involving finite-sum optimization over networks. Rather than the mean employed by distributed SAGA, the novel Byrd-SAGA relies on the geometric median to aggregate the corrected stochastic gradients sent by the workers. When less than half of the workers are Byzantine attackers, the robustness of geometric median to outliers enables Byrd-SAGA to achieve provable linear convergence to a neighborhood of the optimal solution, where the size of neighborhood is determined by the number of Byzantine workers. Numerical tests demonstrate the robustness of Byrd-SAGA to various Byzantine attacks, as well as the merits of Byrd-SAGA over Byzantine-resilient SGD.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Exploring Deep Convolutional Neural Networks: A Grad-CAM Enhanced Comparative Study for Automated COVID-19 Diagnosis from Chest X-Ray Images

Exploring Deep Convolutional Neural Networks: A Grad-CAM Enh...

引用

Smart Generation Computing, Communication and Networking (SMART GENCON), International Conference on

作者： Md. Sazid Reza Mir Md. Jahangir Kabir Md. Abdur Rakib Mollah Nazmul Islam Nahin Department of Computer Science & Engineering Data Science and Innovation TD School University of Technology Sydney Sydney Australia Department of Electronics & Telecommunication Engineering Rajshahi University of Engineering & Technology Rajshahi Bangladesh

The emergence of COVID-19 in late 2019 sparked a global health crisis. Identifying infections quickly is key to providing proper care and limiting transmission. The Reverse Transcription-Polymerase Chain Reaction (RT-PCR) test, a widely recognized primary diagnostic method for detecting COVID-19, faces challenges associated with cost and compatibility in resource-constrained environments. This issue is resolved by chest X-ray imaging since it is non-invasive and inexpensive. In this domain, research often utilizes pre-trained models on relatively small datasets of chest X-ray images, limiting the potential for optimal generalization and accuracy achievable with larger training sets. Also, these research works limit the exploration of crucial information on salient regions within CXR images, constraining the depth of analysis and interpretation. This study addresses current research challenges through a detailed comparative analysis of convolutional neural network (CNN) architectures, including pre-trained models like Inception V3, Xception, DenseNet-121, VGG-19, and a Custom CNN developed from scratch. The evaluation focuses on their effectiveness in detecting COVID-19 using a relatively vast and comprehensive dataset. The evaluation focuses on assessing the efficacy of these models in detecting COVID-19 across binary-class and multi-class classifications within a large dataset, offering valuable insights into their overall effectiveness. Grad-CAM is employed in this study, which visualizes picture salient areas affecting deep neural network predictions and improves model interpretability. The heatmaps reveal the model's decision-making processes, making their behavior more understandable to medical practitioners. With resulting accuracies of 99.13% and 97.66% in binary-class and multi-class classification, Inception V3 demonstrated the most optimal performance. This proposed approach surpasses existing benchmark methods, underscoring its superiority.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Adaptive radius species based particle swarm optimization for multimodal optimization problems

Adaptive radius species based particle swarm optimization fo...

引用

Congress on Evolutionary Computation

作者： Zhi-Hui Zhan Zi-Jia Wang Ying Lin Jun Zhang School of Computer Science and Engineering South China University of Technology Guangzhou School of Data and Computer Science Sun Yat-Sen University Guangzhou China Department of Psychology Sun Yat-Sen University Guangzhou China

ISBN: (纸本)9781509006243

Multimodal optimization problem always has several peaks that are all optima of the problem. A promising approach to deal with such kind of problem should locate the peaks as many as possible (e.g., all the peaks) and should obtain high accuracy in each peak. The species-based particle swarm optimization (SPSO) divides the population into several subpopulations. Each subpopulation is gathered around a neighborhood best called species seed within the radius r, trying to locate different peaks. It does well in some low-dimensional multimodal optimization problems. However, the parameter r, which is associated with the efficiency and the accuracy of the algorithm, must be specified by the users. This makes SPSO very difficult for users to determine how much the parameter r should be. In this paper, a method of adaptively choosing radius r in SPSO is proposed, termed as adaptive SPSO (ASPSO). The experimental results show that the performance of ASPSO is more effective and accurate than standard SPSO in dealing with low-dimensional multimodal optimization problems.

关键词： Optimization Sociology Statistics Particle swarm optimization Algorithm design and analysis Search problems Convergence

来源：评论

学校读者我要写书评

暂无评论

Minimum Separator Reconfiguration 18

Minimum Separator Reconfiguration

引用

18th International Symposium on Parameterized and Exact Computation, IPEC 2023

作者： Gomes, Guilherme C. M. Legrand-Duchesne, Clément Mahmoud, Reem Mouawad, Amer E. Okamoto, Yoshio Dos Santos, Vinicius F. Van Der Zanden, Tom C. Department of Computer Science Federal University of Minas Gerais Belo Horizonte Brazil LaBRI CNRS Université de Bordeaux France Virginia Commonwealth University RichmondVA United States Department of Computer Science American University of Beirut Beirut Lebanon Graduate School of Informatics and Engineering The University of Electro-Communications Chofu Japan Department of Data Analytics and Digitalisation Maastricht University Netherlands

ISBN: (纸本)9783959773058

We study the problem of reconfiguring one minimum s-t-separator A into another minimum s-tseparator B in some n-vertex graph G containing two non-adjacent vertices s and t. We consider several variants of the problem as we focus on both the token sliding and token jumping models. Our first contribution is a polynomial-time algorithm that computes (if one exists) a minimum-length sequence of slides transforming A into B. We additionally establish that the existence of a sequence of jumps (which need not be of minimum length) can be decided in polynomial time (by an algorithm that also outputs a witnessing sequence when one exists). In contrast, and somewhat surprisingly, we show that deciding if a sequence of at most jumps can transform A into B is an NP-complete problem. To complement this negative result, we investigate the parameterized complexity of what we believe to be the two most natural parameterized counterparts of the latter problem;in particular, we study the problem of computing a minimum-length sequence of jumps when parameterized by the size k of the minimum s-t-separators and when parameterized by the number of jumps. For the first parameterization, we show that the problem is fixed-parameter tractable, but does not admit a polynomial kernel unless NP ⊆ coNP/poly. We complete the picture by designing a kernel with O(2) vertices and edges for the length of the sequence as a parameter. © 2023 Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing. All rights reserved.

关键词： Separators

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：