检索结果-内蒙古大学图书馆

IEEE 2nd International Conference on image Information Processing (ICIIP)

作者： Prakash, Om Khare, Manish Srivastava, Rajneesh Kumar Khare, Ashish Univ Allahabad Dept Elect & Commun Allahabad 211002 Uttar Pradesh India

ISBN: (纸本)9781467360999;9781467361002

image classification is an important problem because of its applications in many fields like shape analysis, object tracking, image retrieval etc. Many techniques have been proposed in literature for classification of objects into two classes. Multi class image classification with high accuracy is a challenging task. In this paper we propose a new algorithm for multi class image classification that uses biorthogonal wavelet transform as image feature. Original images are decomposed into subbands LL, LH, HL and HH using Biorthogonal wavelet transform at multiple scales. The coefficients of LH, HL, HH subbands are used as features for classification. The approximate shift invariance and linear phase properties of Biorthogonal wavelet transform are useful for classification of images. Also, the lifting-scheme of Biorthogonal wavelet yields reduced computational cost. Quantitative evaluation of classification accuracy demonstrates the strength of the proposed method.

关键词： multiclass image classification Biorthogonal wavelet transform feature selection machine learning

来源：评论

学校读者我要写书评

暂无评论

Quantum convolutional neural networks for multiclass image classification

引用

QUANTUM INFORMATION PROCESSING 2024年第5期23卷 1-16页

作者： Shi, Shangshang Wang, Zhimin Li, Jiaxin Li, Yanan Shang, Ruimin Zhong, Guoqiang Gu, Yongjian Ocean Univ China Fac Informat Sci & Engn Qingdao 266100 Peoples R China Ocean Univ China Engn Res Ctr Adv Marine Phys Instruments & Equipme Minist Educ Qingdao 266100 Peoples R China

The quantum convolutional neural networks (QCNNs) are emerging as a promising solution for image classification problems on near-term quantum devices. While QCNNs have shown encouraging results in binary classification tasks, their effectiveness in more complex multiclass classification tasks remains to be fully understood. In this study, we propose three distinct QCNN architectures, each inspired by the configuration of two-qubit entangling blocks available in quantum hardware. These QCNNs are designed based on different sliding modes of quantum filters. We investigate the impact of quantum filter structure, filter arrangement and parameter sharing among filters on the performance of QCNNs in multiclass classification. Our findings indicate that the specific structure of the quantum filter significantly affects the models' performance. Moreover, we observed that unsharring parameters and more complex filter arrangements can significantly enhance the performance of QCNNs. These results contribute to the development of powerful quantum classifiers for multiclass image classification.

关键词： Quantum convolutional neural networks multiclass image classification Two-qubit entangling blocks Quantum filter

来源：评论

学校读者我要写书评

暂无评论

Fisher Regularized e-Dragging for image classification

引用

IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS 2023年第2期15卷 639-650页

作者： Chen, Zhe Wu, Xiao-Jun Kittler, Josef Jiangnan Univ Sch AI & CS Wuxi 214122 Peoples R China Univ Surrey Ctr Vis Speech & Signal Proc Guildford GU2 7XH England

Discriminative least-squares regression (DLSR) has been shown to achieve promising performance in multiclass image classification tasks. Its key idea is to force the regression labels of different classes to move in opposite directions by means of the e-dragging technique, yielding a discriminative regression model exhibiting wider margins. However, the e-dragging technique ignores an important problem: its relaxation matrix is dynamically updated in optimization, which means the dragging values can also cause the labels from the same class to be uncorrelated. In order to learn a more powerful projection, as well as regression labels, we propose a Fisher regularized e-dragging framework (Fisher -e) for image classification by constraining the relaxed labels using the Fisher criterion. On the one hand, the Fisher criterion improves the intraclass compactness of the relaxed labels during relaxation learning. On the other hand, it is expected further to enhance the interclass separability of e-dragging. Fisher -e for the first time ever attempts to integrate the Fisher criterion and e-dragging technique into a unified model because they are complementary in learning discriminative projection. Extensive experiments on various data sets demonstrate that the proposed Fisher -e method achieves performance that is superior to other state-of-the-art classification methods. The MATLAB codes are available at https://***/chenzhe207/Fisher-epsilon.

关键词： e-dragging technique discriminative least-squares regression (DLSR) Fisher discrimination criterion multiclass image classification

来源：评论

学校读者我要写书评

暂无评论

Active learning combining uncertainty and diversity for multi-class image classification

引用

IET COMPUTER VISION 2015年第3期9卷 400-407页

作者： Gu, Yingjie Jin, Zhong Chiu, Steve C. Nanjing Univ Sci & Technol Dept Comp Sci & Engn Nanjing 210094 Jiangsu Peoples R China Idaho State Univ Dept Elect Engn Pocatello ID 83209 USA

In computer vision and pattern recognition applications, there are usually a vast number of unlabelled data whereas the labelled data are very limited. Active learning is a kind of method that selects the most representative or informative examples for labelling and training;thus, the best prediction accuracy can be achieved. A novel active learning algorithm is proposed here based on one-versus-one strategy support vector machine (SVM) to solve multi-class image classification. A new uncertainty measure is proposed based on some binary SVM classifiers and some of the most uncertain examples are selected from SVM output. To ensure that the selected examples are diverse from each other, Gaussian kernel is adopted to measure the similarity between any two examples. From the previous selected examples, a batch of diverse and uncertain examples are selected by the dynamic programming method for labelling. The experimental results on two datasets demonstrate the effectiveness of the proposed algorithm.

关键词： Knowledge engineering techniques Optimisation techniques learning (artificial intelligence) image classification Computer vision and image processing techniques support vector machine pattern recognition applications image recognition support vector machines binary SVM classiflers multiclass image classification unlabelled data computer vision active learning algorithm dynamic programming method Gaussian kernel dynamic programming

来源：评论

学校读者我要写书评

暂无评论

Discovering image Semantics in Codebook Derivative Space

引用

IEEE TRANSACTIONS ON MULTIMEDIA 2012年第4期14卷 986-994页

作者： Wang, Jinjun Gong, Yihong Epson Res & Dev Inc San Jose CA 95131 USA NEC Labs China Beijing 100084 Peoples R China

The sparse coding based approaches for image recognition have recently shown improved performance than traditional bag-of-features technique. Due to high dimensionality of the image descriptor space, existing systems usually require very large codebook size to minimize coding error in order to get satisfactory accuracy. While most research efforts try to address the problem by constructing a relatively smaller codebook with stronger discriminative power, in this paper, we introduce an alternative solution by enhancing the quality of coding. Particularly, we apply the idea similar to Fisher kernel to the coding framework, where we use the image-dependent codebook derivative to represent the image. The proposed idea is generic across multiple coding criteria, and in this paper, it is applied to enhance the locality-constraint linear coding (LLC). Experiments show that, the extracted new feature, called "LLC+," achieved significantly improved accuracy on several challenging datasets even with a small codebook of 1/20 the reported size used by LLC. This obviously adds to LLC+ the modeling accuracy, processing speed and codebook training advantages.

关键词： Codebook derivative multiclass image classification sparse-coding

来源：评论

学校读者我要写书评

暂无评论

multiclass Object classification for Computer Vision using Linear Genetic Programming

Multiclass Object Classification for Computer Vision using L...

引用

24th International Conference image and Vision Computing New Zealand

作者： Downey, Carlton Zhang, Mengjie Victoria Univ Wellington Sch Engn & Comp Sci Wellington New Zealand

ISBN: (纸本)9781424446971

multiclass classification problems arise naturally in many tasks in computer vision;typical examples include image segmentation and letter recognition. These are among some of the most challenging and important tasks in the area and solutions to them are eagerly sought after. Genetic Programming (GP) is a powerful and flexible machine learning technique that has been successfully applied to many binary classification tasks. However, the traditional form of GP performs poorly on multiclass classification problems. Linear GP (LGP) is an alternative form of GP where programs are represented as sequences of instructions like Java and C++. This paper discusses results which demonstrate the superiority of LGP as a technique for multiclass classification. It also discusses a new extension to LGP which results in a further improvement in the performance on multiclass classification problems.

关键词： AI approaches to computer vision Genetic Programming Linear Genetic Programming multiclass Object classification multiclass image classification

来源：评论

学校读者我要写书评

暂无评论

Mono- and Dual-Regulated Contractive-Expansive-Contractive Deep Convolutional Networks for classification of Multispectral Remote Sensing images

引用

IEEE GEOSCIENCE AND REMOTE SENSING LETTERS 2022年 19卷

作者： Singh, Abhishek Bruzzone, Lorenzo Univ Trento Dept Informat Engn & Comp Sci I-38123 Trento Italy

Deep convolutional neural networks (CNNs) are the state-of-the-art methods in the domain of classification of remote sensing (RS) data. However, traditional CNN models suffer from huge computational costs in learning land-use and land-cover features, particularly in large-scale RS problems. To address this issue, we propose a reliable mono- and dual-regulated contractive-expansive-contractive (MRCEC/DRCEC) CNN for scene based multispectral (MS) image classification. The proposed technique increases the accuracy of learning and minimizes the loss in the feature maps by incorporating the CEC approach in the classification. Extensive experiments conducted on the Sentinel-2 EuroSATallbands dataset pointed out that the proposed model outperforms the state-of-the-art models such as EfficientNet-B0, RESNet-50, and EfficientNet-B7.

关键词： Convolutional neural networks Convolution Remote sensing Computer architecture Training Neural networks Deep learning Convolutional neural networks (CNNs) deep learning multiclass image classification remote sensing (RS) scene classification Sentinel-2

来源：评论

学校读者我要写书评

暂无评论

Maximum classification Optimization-based Active Learning for image classification

Maximum Classification Optimization-based Active Learning fo...

引用

International Congress on image and Signal Processing

作者： Zhengwei Cui Xiaoming Chen Jian Wu Victor S. Sheng Yujie Shi Inst. of Intell. Inf. Process. & Applic. Soochow Univ. Suzhou China

ISBN: (纸本)9781479958368

Traditional multi-class image classification needs a large number of training samples for building a classifier model. However, it is very time-consuming and costly to obtain labels for a large number of training samples from human experts. Active learning is a feasible solution. This paper proposes a maximum classification optimization method (MCO) for actively selecting unlabeled images to acquire labels. It integrated the information of an unlabeled sample from different perspectives with two steps. It first chooses a subset of candidates, and then selects the best from these candidates. Our experimental results show that the maximum classification optimization method outperforms two popular exiting methods (entropy-based uncertainty and BvSB).

关键词： image classification image sampling learning (artificial intelligence) optimisation MCO method maximum classification optimization-based active learning multiclass image classification training sample unlabeled images selecting Equations Feature extraction Mathematical model Optimization Training Uncertainty BvSB active learning maximum classification optimization

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：