检索结果-内蒙古大学图书馆

SRFNet: selective receptive field network for human pose estimation

JOURNAL OF SUPERcompUTING 2022年第1期78卷 691-711页

作者： Ou, Zhilong Luo, YanMin Chen, Jin Chen, Geng Huaqiao Univ Coll Comp Sci & Technol Xiamen 361021 Peoples R China Huaqiao Univ Xiamen Key Lab Comp Vis & Pattern Recognit Xiamen 361021 Peoples R China Fujian Normal Univ Fuzhou Fujian Peoples R China

Human pose estimation is fundamental to many computer vision tasks and has made significant progress in recent years. However, the problem of unbalanced performance among joints has not been paid enough attention. Basing on simple baseline Xiao et al. (Proceedings of the European conference on computer vision, 2018), we propose a weighted summation method of local keypoint, selective receptive field (SRF) unit and use the feature fuse method to tackle this problem. Initially, the weighted summation method of local keypoint is designed to make the network explicitly address keypoints with large loss value. This method calculation weights according to the loss value of each joint. Subsequently, the SRF unit was proposed to adaptively select receptive field size for keypoints. Firstly, multiple branches with different kernel sizes are compared using softmax attention. Secondly, the Select operator chooses one of these branches to yield effective receptive fields. Then, the features coming from the encoder are merged in the decoder using concatenation to solve the occlusion joint. This method enhances communication between spatial information and semantic information. The experimental results show that as a model-agnostic approach, our method promotes SimpleBaseline-50 - 256 x 192 by 4.3 AP on COCO validation set. Extensive experiments demonstrate that the proposed approach is superior to several state-of-the-art methods in terms of accuracy and robustness.

关键词： Human pose estimation Selective receptive field network Selective receptive field unit Weighted summation method of local keypoint

来源：评论

学校读者我要写书评

暂无评论

Identity-Aware Contrastive Knowledge Distillation for Facial Attribute recognition

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 2023年第10期33卷 5692-5706页

作者： Chen, Si Zhu, Xueyan Yan, Yan Zhu, Shunzhi Li, Shao-Zi Wang, Da-Han Xiamen Univ Technol Sch Comp & Informat Engn Fujian Key Lab Pattern Recognit & Image Understand Xiamen 361024 Peoples R China Xiamen Univ Sch Informat Xiamen 361005 Peoples R China

Facial attribute recognition (FAR) is an important and yet challenging multi-label learning task in computer vision. Existing FAR methods have achieved promising performance with the development of deep learning. However, they usually suffer from prohibitive computational and memory costs. In this paper, we propose an identity-aware contrastive knowledge distillation method, termed ICKD, to compress the FAR model. A nonlinear weight-sharing mapping (NWSM) mechanism is firstly designed to avoid the difficulty of directly matching features of the teacher and student networks due to the lower representation ability of the student network. Furthermore, an identity-aware contrastive distillation (ICD) loss is employed to guide the student network to effectively learn the mutual relations between samples with multiple attributes. In addition, an adjustable ladder distillation (ALD) loss is developed to automatically adjust the importance of different distillation points with the progress of training. Extensive experiments demonstrate that our method can significantly improve the performance of student networks and outperforms the existing FAR methods on the public challenging datasets.

关键词： Facial attribute recognition knowledge distillation contrastive learning contrastive knowledge distillation adjustable ladder distillation loss

来源：评论

学校读者我要写书评

暂无评论

A Shallow Graph Neural Network with Innovative Node Updating for Online Handwritten Stroke Classification 17th

A Shallow Graph Neural Network with Innovative Node Updating...

引用

17th International Conference on Document Analysis and recognition (ICDAR)

作者： Wang, Yan-Rong Wang, Da-Han Yun, Xiao-Long Zhang, Yan-Ming Yin, Fei Zhu, Shunzhi Xiamen Univ Technol Sch Comp & Informat Engn Fujian Key Lab Pattern Recognit & Image Understan Xiamen 361024 Peoples R China Chinese Acad Sci Natl Lab Pattern Recognit NLPR Inst Automat Beijing 100086 Peoples R China

ISBN: (纸本)9783031416842;9783031416859

Stroke classification is important to the layout analysis of online handwritten documents. Due to the diversity of writing styles and the complexity of layout structure, stroke classification is challenging. Graph neural networks (GNNs) is one of the most effective frameworks for stroke classification. However, GNNs has the problem of node over-compression caused by the deep structure of GNNs, which will lead to loss of node information and hence may deteriorate the performance of stroke classification. In this paper, we propose a shallow graph neural network model that is capable of retaining long-term receptive field by constructing a more reasonable graph through edge classification before the node classification step. Moreover, a novel node learning method is used to integrate edge features into nodes, where edge features not only participate in the calculation of node attention weight as in previous GNN based methods, but also participate in the final node integration. Experiments on the IAMonDo dataset show that our proposed method achieves an accuracy of 97.71% that is superior to existing state-of-the-art methods, demonstrating the effectiveness of the proposed method.

关键词： Stroke classification Online handwritten documents Graph Neural Network Node updating Edge classification

来源：评论

学校读者我要写书评

暂无评论

computer vision Approaches for Automated Bee Counting Application 18

Computer Vision Approaches for Automated Bee Counting Applic...

引用

18th IFAC Conference on Programmable Devices and Embedded Systems (PDES)

作者： Bilik, Simon Janakova, Ilona Ligocki, Adam Horak, Dominik Ficek Karel Brno Univ Technol Fac Elect Engn & Commun Dept Control & Instrumentat Brno Czech Republic Lappeenranta Lahti Univ Technol LUT Dept Computat Engn Comp Vis & Pattern Recognit Lab Lappeenranta Finland

Many application from the bee colony health state monitoring could be efficiently solved using a computer vision techniques. One of such challenges is an efficient way for counting the number of incoming and outcoming bees, which could be used to further analyse many trends, such as the bee colony health state, blooming periods, or for investigating the effects of agricultural spraying. In this paper, we compare three methods for the automated bee counting over two own datasets. The best performing method is based on the ResNet-50 convolutional neural network classifier, which achieved accuracy of 87% over the BUT1 dataset and the accuracy of 93% over the BUT2 dataset. Copyright (c) 2024 The Authors.

关键词： Signal Processing Biomedical systems bee counting bee traffic monitoring computer vision deep classification object detection

来源：评论

学校读者我要写书评

暂无评论

AE-UNet: a composite lung CT image segmentation framework using attention mechanism and edge detection

引用

JOURNAL OF SUPERcompUTING 2025年第1期81卷 1-21页

作者： Li, Hongzhi Ren, Zhanghao Zhu, Guoqing Wang, Jiaxi Chengdu Univ Sch Comp Sci Chengdu Peoples R China Chengdu Univ Inst Higher Educ Sichuan Prov Key Lab Pattern Recognit & Intelligent Informat Pr Chengdu Peoples R China Chengdu Univ Sichuan Prov Dept Culture & Tourism Key Lab Digital Innovat Tianfu Culture Chengdu Peoples R China

The primary impediments in lung CT image segmentation stem from the ambiguity in edge definition and the inadequate segmentation accuracy. Addressing these issues, this paper introduces a novel composite lung CT image segmentation framework that integrates an attention mechanism with an edge detection operator. We utilize residual dynamic convolutions as the encoder to augment the network's capability for extracting and representing nuanced lesion features. Sobel edge detection is integrated into the skip connections to facilitate the transmission and utilization of edge information. In particular, we introduce an information fusion attention module for deeper layers, optimizing feature reorganization and utilization by attention mechanisms and dilated convolution. Experimental evaluations on two lung CT datasets reveal that our proposed AE-UNet achieves outstanding segmentation performance, surpassing the best baseline network by an average of 0.93%.

关键词： CT image segmentation UNet Sobel operator Attention mechanism

来源：评论

学校读者我要写书评

暂无评论

Deep Metric Learning on the SPD Manifold for Image Set Classification

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 2024年第2期34卷 663-680页

作者： Wang, Rui Wu, Xiao-Jun Xu, Tianyang Hu, Cong Kittler, Josef Jiangnan Univ Sch Artificial Intelligence & Comp Sci Wuxi 214122 Peoples R China Jiangnan Univ Jiangsu Prov Engn Lab Pattern Recognit & Computat Wuxi 214122 Peoples R China Univ Surrey Ctr Vis Speech & Signal Proc Guildford GU2 7XH England

Thanks to the efficacy of Symmetric Positive Definite (SPD) manifold in characterizing video sequences (image sets), image set-based visual classification has made remarkable progress. However, the issue of large intra-class diversity and inter-class similarity is still an open challenge for the research community. Although several recent studies have alleviated the above issue by constructing Riemannian neural networks for SPD matrix nonlinear processing, the degradation of structural information during multi-stage feature transformation impedes them from going deeper. Besides, a single cross-entropy loss is insufficient for discriminative learning as it neglects the peculiarities of data distribution. To this end, this paper develops a novel framework for image set classification. Specifically, we first choose a mainstream neural network built on the SPD manifold (SPDNet) [25] as the backbone with a stacked SPD manifold autoencoder (SSMAE) built on the tail to enrich the structured representations. Due to the associated reconstruction error terms, the embedding mechanism of both SSMAE and each SPD manifold autoencoder (SMAE) forms an approximate identity mapping, simplifying the training of the suggested deeper network. Then, the ReCov layer is introduced with a nonlinear function for the constructed architecture to narrow the discrepancy of the intra-class distributions from the perspective of regularizing the local statistical information of the SPD data. Afterward, two progressive metric learning stages are coupled with the proposed SSMAE to explicitly capture, encode, and analyze the geometric distributions of the generated deep representations during training. In consequence, not only a more powerful Riemannian network embedding but also effective classifiers can be obtained. Finally, a simple maximum voting strategy is applied to the outputs of the learned multiple classifiers for classification. The proposed model is evaluated on three typical visual classi

关键词： SPD manifold image set classification Riemannian neural network stacked SPD manifold autoencoder (SSMAE) metric learning

来源：评论

学校读者我要写书评

暂无评论

UAM-Net: An Attention-Based Multi-level Feature Fusion UNet for Remote Sensing Image Segmentation 6th

UAM-Net: An Attention-Based Multi-level Feature Fusion UNet ...

引用

6th Chinese Conference on pattern recognition and computer vision (PRCV)

作者： Cao, Yiwen Jiang, Nanfeng Wang, Da-Han Wu, Yun Zhu, Shunzhi Xiamen Univ Technol Sch Comp & Informat Engn Xiamen 361024 Peoples R China Fujian Key Lab Pattern Recognit & Image Understan Xiamen 361024 Peoples R China

ISBN: (纸本)9789819984619;9789819984626

Semantic segmentation of Remote Sensing Images (RSIs) is an essential application for precision agriculture, environmental protection, and economic assessment. While UNet-based networks have made significant progress, they still face challenges in capturing long-range dependencies and preserving fine-grained details. To address these limitations and improve segmentation accuracy, we propose an effective method, namely UAM-Net (UNet with Attention-based Multi-level feature fusion), to enhance global contextual understanding and maintain fine-grained information. To be specific, UAM-Net incorporates three key modules. Firstly, the Global Context Guidance Module (GCGM) integrates semantic information from the Pyramid Pooling Module (PPM) into each decoder stage. Secondly, the Triple Attention Module (TAM) effectively addresses feature discrepancies between the encoder and decoder. Finally, the computation-effective Linear Attention Module (LAM) seamlessly fuses coarse-level feature maps with multiple decoder stages. With the corporations of these modules, UAM-Net significantly outperforms the most state-of-the-art methods on two popular benchmarks.

关键词： Semantic segmentation U-shape architecture Attention mechanism Feature fusion Remote sensing images

来源：评论

学校读者我要写书评

暂无评论

Multi-label feature selection via joint label enhancement and pairwise label correlations

引用

INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS 2023年第11期14卷 3943-3964页

作者： Liu, Jinghua Yang, Songwei Lin, Yaojin Wang, Chenxi Wang, Cheng Du, Jixiang Huaqiao Univ Dept Comp Sci & Technol Xiamen 361021 Peoples R China Huaqiao Univ Xiamen Key Lab Comp Vis & Pattern Recognit Xiamen 361021 Peoples R China Huaqiao Univ Fujian Key Lab Big Data Intelligence & Secur Xiamen 361021 Peoples R China Minnan Normal Univ Sch Comp Sci Zhangzhou 363000 Peoples R China Wuyi Univ Fujian Key Lab Big Data Applicat & Intellectualiza Wuyi 354300 Peoples R China

Multi-label feature selection(MFS) has gained in importance, and it is today confronted with the current need to process multi-semantic high-dimensional data. Recent studies usually figure out the MFS problems either simply assume that all associated labels are equally important for each instance;or that the labels are independent of each other. In many real-world applications, however, both cases may occur that the significance of each relevant label is generally different and label correlations are ubiquitous. Based on this observation, we propose a new algorithm, called FSEP, to perform MFS by considering label significance and pairwise label correlations. In FSEP, we first construct a label enhancement method that is able to obtain label distribution and further earn the information of label significance. Then, FSEP explores the influence mechanism of label correlations to features by using neighborhood mutual information and incorporates this influence into the process of feature evaluation. After that, a novel multi-label feature selection strategy, namely, Max-Relevance, Max-Contribution, and Min-Redundancy, is proposed, which achieves a favorable trade-off among feature relevance, the contribution of label correlations to features, and feature redundancy, simultaneously. Extensive experiments on both public and real-world datasets show that the proposed method achieves encouraging results compared with state-of-the-art MFS algorithms.

关键词： Multi-label feature selection Neighborhood mutual information label distribution label correlations

来源：评论

学校读者我要写书评

暂无评论

Semantic Camera Self-Aware Contrastive Learning for Unsupervised Vehicle Re-Identification

引用

IEEE SIGNAL PROCESSING LETTERS 2024年 31卷 2175-2179页

作者： Tao, Xuefeng Kong, Jun Jiang, Min Luo, Xi Jiangnan Univ Minist Educ Key Lab Adv Proc Control Light Ind Wuxi 214122 Peoples R China Univ Western Australia Dept Comp Sci Crawley WA 6009 Australia Jiangnan Univ Jiangsu Prov Engn Lab Pattern Recognit & Comp Inte Wuxi 214122 Peoples R China

Unsupervised vehicle re-identification (ReID) aims to retrieve vehicle images from different cameras without using identity labels. Patch features, which capture fine-grained semantic information of vehicles, are crucial for ReID. However, existing methods often fail to preserve the discriminative semantic structure of vehicles due to the non-uniformity of feature attributes across patches. Moreover, domain discrepancy among cameras also requires attention, as it can cause large intra-class variance and noisy clustering results. To tackle these problems, in this letter, we propose a novel Semantic Camera Self-Aware Contrastive Learning (SCSCL) framework for unsupervised vehicle ReID. Firstly, we design the Semantic Self-Aware Contrastive (SSC) loss to perceive the semantic attributes of vehicle images from spatial transformer parameters, thereby enhancing the semantic representation of patch features. Secondly, we design the Camera Self-Aware Contrastive (CSC) loss to perceive the cross-camera distance distributions to facilitate the exploration of instance constraints, thereby enabling cross-camera clustering-friendly representations. Finally, extensive experimental results on VeRi-776 and VehicleID datasets attest to the efficacy of our method over the state-of-the-art performance.

关键词： Cameras Semantics Feature extraction Self-aware Transformers Contrastive learning Training Unsupervised vehicle re-identification semantic self-aware contrastive loss camera self-aware contrastive loss

来源：评论

学校读者我要写书评

暂无评论

Unsupervised remote sensing image thin cloud removal method based on contrastive learning

引用

IET IMAGE PROCESSING 2024年第7期18卷 1844-1861页

作者： Tan, Zhan Cong Du, Xiao Feng Man, Wang Xie, Xiao Zhu Wang, Gui Song Nie, Qin Xiamen Univ Technol Sch Comp & Informat Engn Xiamen 361024 Peoples R China Xiamen Univ Technol Inst Spatial Informat Technol Xiamen Peoples R China Fujian Key Lab Pattern Recognit & Image Understand Xiamen Peoples R China

Cloud removal algorithm is a crucial step of remote sensing image preprocessing. The current mainstream remote sensing image cloud removal algorithms are implemented based on deep learning, and most of them are supervised. A large number of data pairs are required for training to achieve cloud removal. However, real with/without cloud image pairs datasets are difficult to obtain in the real world, and the models obtained by training on synthetic datasets often need to generalize better to natural scenes. And the existing unsupervised thin cloud removal methods based on Cycle-GAN framework with considerable model complexity and unstable training are not an excellent solution to the problem of lack of paired datasets. Based on this, in this paper, the authors propose an unsupervised remote sensing image thin cloud removal method based on contrastive learning-GAN-UD. It is a network consisting of a frequency-spatial attention generator and a discriminator. In addition, the authors introduce local contrastive loss and global content loss to constrain the content of the generated images to ensure that the generated cloud-free images are consistent with the input cloud images in terms of image content. Experimental results show that the proposed method in this paper can still effectively remove thin clouds from remote sensing images without paired training datasets, outperforms current unsupervised cloud removal methods, and achieves comparable performance to supervised methods. The authors propose an unsupervised remote sensing image thin cloud removal method based on contrastive learning-Generative Adversarial Network of Unpair Data. It is a network consisting of a frequency-spatial attention generator and a discriminator. In addition, the authors introduce local contrastive loss and global content loss to constrain the content of the generated images to ensure that the generated cloud-free images are consistent with the input cloud images in terms of image content. ima

关键词： image reconstruction remote sensing unsupervised learning

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：