检索结果-内蒙古大学图书馆

Electronic Information Engineering and Computer Technology (EIECT), International Conference on

作者： Chenyu Zhou Bo Chen Ying Zhang Wensheng Chen Binbin Pan Jing Ji Shenzhen Key Laboratory of Advanced Machine Learning and Applications College of Mathematics and Statistics Shenzhen University Shenzhen China Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University Shenzhen China

Underwater image enhancement, as an important branch of image processing, has attracted the attention of many scholars in recent years. Due to selective scattering and degradation of light in water, images captured underwater are generally of low quality. To improve the accuracy of underwater image details and color, we propose a novel underwater image enhancement model, which is based on a Multi-color space fusion (MSF) strategy and assigns corresponding weights to each color channel for fusion. Qualitative and quantitative comparisons show that the proposed method can effectively remove image blur, achieve color correction, and significantly improve image sharpness.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learnable Irrelevant Modality Dropout for Multimodal Action Recognition on Modality-Specific Annotated Videos

arXiv

引用

arXiv 2022年

作者： Alfasly, Saghir Lu, Jian Xu, Chen Zou, Yuru Shenzhen Key Laboratory of Advanced Machine Learning and Applications Shenzhen University China Guangdong Key Laboratory of Intelligent Information Processing Shenzhen China Pazhou Lab Guangzhou China

With the assumption that a video dataset is multimodality annotated in which auditory and visual modalities both are labeled or class-relevant, current multimodal methods apply modality fusion or cross-modality attention. However, effectively leveraging the audio modality in vision-specific annotated videos for action recognition is of particular challenge. To tackle this challenge, we propose a novel audio-visual framework that effectively leverages the audio modality in any solely vision-specific annotated dataset. We adopt the language models (e.g., BERT) to build a semantic audio-video label dictionary (SAVLD) that maps each video label to its most K-relevant audio labels in which SAVLD serves as a bridge between audio and video datasets. Then, SAVLD along with a pretrained audio multi-label model are used to estimate the audio-visual modality relevance during the training phase. Accordingly, a novel learnable irrelevant modality dropout (IMD) is proposed to completely drop out the irrelevant audio modality and fuse only the relevant modalities. Moreover, we present a new two-stream video Transformer for efficiently modeling the visual modalities. Results on several vision-specific annotated datasets including Kinetics400 and UCF-101 validated our framework as it outperforms most relevant action recognition methods. © 2022, CC BY.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Non-Negative Feature Extraction using Conjugate Gradient Method

Non-Negative Feature Extraction using Conjugate Gradient Met...

引用

International Conference on Computational Intelligence and Security

作者： Jiawen Zhang Wen-Sheng Chen Binbin Pan College of Mathematics and Statistics Shenzhen University Guangdong Key Laboratory of Intelligent Information Processing Shenzhen Key Laboratory of Advanced Machine Learning and Applications Shenzhen P.R. China

ISBN: (数字)9781728160924

ISBN: (纸本)9781728160931

Non-negative matrix factorization (NMF) is an efficient approach for non-negative feature extraction and parts-based representation. Nevertheless, the optimization problem of NMF is usually resolved using gradient descent (GD) method, which leads to slow convergence. In this paper, we utilize conjugate gradient method (CGM) to develop a novel fast CGM-based NMF (CGM-NMF) algorithm for face recogition. The update rules of our CGM-NMF are acquired by searching conjugate directions under non-negativity constraint. The proposed CGM-NMF algorithm is theoretically proven to be convergent. Experimental results show that our CGM-NMF method has superior performance.

关键词： Convergence Databases Gradient methods Linear programming Face recognition Training

来源：评论

学校读者我要写书评

暂无评论

Deep Grouped Non-Negative Matrix Factorization Method for Image Data Representation

Deep Grouped Non-Negative Matrix Factorization Method for Im...

引用

International Conference on machine learning and Cybernetics (ICMLC)

作者： Zihao Zhan Wen-Sheng Chen Binbin Pan Bo Chen College of Mathematics and Statistics Shenzhen University China Guangdong Key Laboratory of Media Security Shenzhen University China Shenzhen Key Laboratory of Advanced Machine Learning and Applications Shenzhen University China

ISBN: (纸本)9781665466097

Non-negative matrix factorization (NMF) is an unsupervised learning method that can be exploited for parts-based image representation due to non-negativity constraints. However, singer-layer NMF cannot capture the latent hierarchical structure features from images, while deep features play more important roles in the image representation and recognition tasks. To overcome the limitation of NMF, this paper proposes a novel deep grouped NMF (DGNMF) approach to learn different level attributes of the data. It is interesting that DGNMF approach automatically makes the data from distinct classes share different basis images and the feature vectors among different classes are mutually orthogonal at the same layer. Meanwhile, to preserve the local information, our DGNMF model establishes the objective function with graph regularization, and its optimization problem is solved using gradient descent method. The developed DGNMF algorithm is proved to be convergent and is finally evaluated on face datasets for classification. Compared with some state-of-the-art deep NMF variants, the results demonstrate the proposed DGNMF algorithm achieves surpassing performances using different layer features.

关键词： machine learning algorithms Face recognition machine learning Image representation Classification algorithms Matrix decomposition Task analysis

来源：评论

学校读者我要写书评

暂无评论

PR product: A substitute for inner product in neural networks

arXiv

引用

arXiv 2019年

作者： Wang, Zhennan Zou, Wenbin Xu, Chen College of Electronic and Information Engineering Shenzhen Key Laboratory of Advanced Machine Learning and Applications Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University College of Mathematics and Statistics Shenzhen University

In this paper, we analyze the inner product of weight vector w and data vector x in neural networks from the perspective of vector orthogonal decomposition and prove that the direction gradient of w decreases with the angle between them close to 0 or π. We propose the Projection and Rejection Product (PR Product) to make the direction gradient of w independent of the angle and consistently larger than the one in standard inner product while keeping the forward propagation identical. As a reliable substitute for standard inner product, the PR Product can be applied into many existing deep learning modules, so we develop the PR Product version of fully connected layer, convolutional layer and LSTM layer. In static image classification, the experiments on CIFAR10 and CIFAR100 datasets demonstrate that the PR Product can robustly enhance the ability of various state-of-the-art classification networks. On the task of image captioning, even without any bells and whistles, our PR Product version of captioning model can compete or outperform the state-of-the-art models on MS COCO dataset. Code has been made available at:https: //***/wzn0828/PR_Product. Copyright © 2019, The Authors. All rights reserved.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

PR Product: A Substitute for Inner Product in Neural Networks

PR Product: A Substitute for Inner Product in Neural Network...

引用

International Conference on Computer Vision (ICCV)

作者： Zhennan Wang Wenbin Zou Chen Xu College of Electronic and Information Engineering Shenzhen Key Laboratory of Advanced Machine Learning and Applications the Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University College of Mathematics and Statistics Shenzhen University

ISBN: (数字)9781728148038

ISBN: (纸本)9781728148045

关键词： Neural networks Standards Optimization Backpropagation Task analysis Reliability Computational modeling

来源：评论

学校读者我要写书评

暂无评论

Stochastic Variance Reduced Gradient for affine rank minimization problem

arXiv

引用

arXiv 2022年

作者： Han, Ningning Nie, Juan Lu, Jian Ng, Michael K. School of Mathematical Sciences Tiangong University Tianjin300387 China Shenzhen Key Laboratory of Advanced Machine Learning and Applications College of Mathematics and Statistics Shenzhen University Shenzhen518060 China Shenzhen Key Laboratory of Advanced Machine Learning and Applications College of Mathematics and Statistics Shenzhen University Shenzhen518060 China Department of Mathematics the University of Hong Kong Pokfulam Hong Kong

We develop an efficient stochastic variance reduced gradient descent algorithm to solve the affine rank minimization problem consists of finding a matrix of minimum rank from linear measurements. The proposed algorithm as a stochastic gradient descent strategy enjoys a more favorable complexity than full gradients. It also reduces the variance of the stochastic gradient at each iteration and accelerate the rate of convergence. We prove that the proposed algorithm converges linearly in expectation to the solution under a restricted isometry condition. The numerical experiments show that the proposed algorithm has a clearly advantageous balance of efficiency, adaptivity, and accuracy compared with other state-of-the-art greedy algorithms. Copyright © 2022, The Authors. All rights reserved.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Deep learning for Spectrum Sensing of Sub-Nyquist Sampled Signals

Deep Learning for Spectrum Sensing of Sub-Nyquist Sampled Si...

引用

International Conference on Fuzzy Systems and Knowledge Discovery (FSKD)

作者： Shiyao Li Hongbing Ji Lin Li Jian Lu School of Electronic Engineering Xidian University Xi'an China Shenzhen Key Laboratory of Advanced Machine Learning and Applications College of Mathematics and Statistics Shenzhen University Shenzhen China

ISBN: (数字)9798350356328

ISBN: (纸本)9798350356335

Traditional Nyquist rate sampling faces significant challenges as communication technology progresses towards 6G. Multi-coset sampling emerges as a viable solution by reducing the sampling rate. However, this method necessitates signal restoration before processing. In response, an intelligent spectrum management system has been developed that utilizes deep learning to directly extract semantic information from multi-coset sampled signals, eliminating the need for complex data preprocessing. The system integrates multi-band spectrum sensing and modulation recognition, achieving an impressive 99% accuracy in both modulation recognition of single signals and sub-band positioning of multiple signals.

关键词： Deep learning Accuracy Target recognition Signal restoration Semantics Modulation Radar Feature extraction Sensors Radio spectrum management

来源：评论

学校读者我要写书评

暂无评论

A Zeroing Neurodynamic Approach for Solving Time-Varying Optimization Problem with Interval-Valued Cost Function

A Zeroing Neurodynamic Approach for Solving Time-Varying Opt...

引用

International Conference on Information Science and Technology (ICIST)

作者： Haojin Li Sitian Qin Hongyu Jia Jiqiang Feng Department of Mathematics Harbin Institute of Technology Weihai China Department of Mathematics and Statistics Shenzhen Key Laboratory of Advanced Machine Learning and Applications Shenzhen University Shenzhen China

Most of the research on uncertainty and imprecision that is widespread in the real world can be abstracted into interval-valued optimization problems whose cost function or constraint function is a closed interval. In this paper, a zeroing neurodynamic (ZND) approach is designed for solving a class of constrained time-varying optimization problems with closed interval-valued cost functions. By virtue of the projection operator and the optimality conditions of the interval-valued optimization problem, the time-varying interval-valued optimization problem is transformed into a nonlinear equation system to be solved. For the reformulated nonlinear equation system, the proposed ZND solver has the ability to track its theoretical solution within pre-defined time. This significantly improves the convergence rate of the ZND solver, making the ZND solver more promising for complex engineering tasks. To the best of our knowledge, it is the first time applying the ZND solver to solve the time-varying interval value optimization problem. Furthermore, the introduction of the projection operator greatly reduces the computational complexity of this ZND solver, saving costs for its hardware implementation. Finally, a numerical example is given to demonstrate the effectiveness of the proposed approach.

关键词：

来源：评论

学校读者我要写书评

暂无评论

0-minimization methods for image restoration problems based on wavelet frames

引用

Inverse Problems 2019年第6期35卷

作者： Lu, Jian Qiao, Ke Li, Xiaorui Lu, Zhaosong Zou, Yuru Shenzhen Key Laboratory of Advanced Machine Learning and Applications College of Mathematics and Statistics Shenzhen University Shenzhen China College of Mathematics and Statistics Shenzhen University Shenzhen China Department of Mathematics Simon Fraser University Burnaby Canada

In this paper we consider a class of 0-minimization and wavelet frame-based models for image deblurring and denoising. Mathematically, they can be formulated as minimizing the sum of a data fidelity term and the 0-'norm' of the framelet coefficients of the underlying image, and we are particularly interested in three different types of data fidelity forms for image restoration problems. We first study the first-order optimality conditions for these models. We then propose a penalty decomposition (PD) method for solving these problems in which a sequence of penalty subproblems are solved by a block coordinate descent (BCD) method. Under some suitable assumptions, we establish that any accumulation point of the sequence generated by the PD method satisfies the first-order optimality conditions of these problems. Moreover, for the problems in which the data fidelity term is convex, we show that such an accumulation point is a local minimizer of the problems. In addition, we show that any accumulation point of the sequence generated by the BCD method is a block coordinate minimizer of the penalty subproblem. Furthermore, under some convexity assumptions on the data fidelity term, we prove that such an accumulation point is a local minimizer of the penalty subproblem. Numerical simulations show that the proposed 0-minimization methods enjoy great potential for image deblurring and denoising in terms of solution quality and/or speed. © 2019 IOP Publishing Ltd.

关键词： Image reconstruction

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：