检索结果-内蒙古大学图书馆

IEEE Transactions on Artificial Intelligence 2024年第4期5卷 1728-1742页

作者： Zhao, Changming Peng, Ruimin Wu, Dongrui Key Laboratory of the Ministry of Education for Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and TechnologyWuhan 430074 China and alsowith the ShenzhenHuazhong University of Science and Technology Research Institute Shenzhen 518029 China

Ensemble learning aggregates outputs from multiple base learners for better performance. Bootstrap aggregating (bagging) and boosting are two popular such approaches. They are suitable for integrating unstable base learners with large variance and weak base learners with large bias, respectively, but not base learners with small variance and/or bias, e.g., support vector machine, regularized logistic regression, and ridge regression. This article proposes two novel ensemble-learning-based fine-tuning approaches, boosting fine-tuning (BF) and bagging and boosting fine-tuning (BBF), to fine-tune learners with small variance and/or bias for better performance. BF embeds boosting in a single hidden layer neural network. In each iteration, BF first uses the Newton's method to generate a temporary training set, and then trains a boosting learner on it. BBF combines BF and bagging. It first uses bootstrap to obtain multiple replicas of the training set, and then trains a BF learner on each replica. Extensive experiments on 46 real-world datasets demonstrated that BBF is flexible, robust, and effective, and can fine-tune many popular classifiers to achieve better generalization performance. © 2020 IEEE.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Semisupervised Transfer Boosting (SS-TrBoosting)

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第7期5卷 3431-3444页

作者： Deng, Lingfei Zhao, Changming Du, Zhenbang Xia, Kun Wu, Dongrui Key Laboratory of Image Processing and Intelligent Control Ministry of Education School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan 430074 China and also with Shenzhen Huazhong University of Science and Technology Research Institute Shenzhen 518063 China

Semisupervised domain adaptation (SSDA) aims at training a high-performance model for a target domain using few labeled target data, many unlabeled target data, and plenty of auxiliary data from a source domain. Previous works in SSDA mainly focused on learning transferable representations across domains. However, it is difficult to find a feature space where the source and target domains share the same conditional probability distribution. Additionally, there is no flexible and effective strategy extending existing unsupervised domain adaptation (UDA) approaches to SSDA settings. In order to solve the above two challenges, we propose a novel fine-tuning framework, semisupervised transfer boosting (SS-TrBoosting). Given a well-trained deep learning-based UDA or SSDA model, we use it as the initial model, generate additional base learners by boosting, and then use all of them as an ensemble. More specifically, half of the base learners are generated by supervised domain adaptation, and half by semisupervised learning. Furthermore, for more efficient data transmission and better data privacy protection, we propose a source data generation approach to extend SS-TrBoosting to semisupervised source-free domain adaptation (SS-SFDA). Extensive experiments showed that SS-TrBoosting can be applied to a variety of existing UDA, SSDA, and SFDA approaches to further improve their performance. © 2020 IEEE.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

controllability of neighborhood Corona product networks

引用

Science China(Information Sciences) 2024年第12期67卷 327-328页

作者： Bo LIU Xuan LI Qiang ZHANG Junjie HUANG Housheng SU Ministry of Education Key Laboratory for Intelligent Analysis and Security Governance of Ethnic Languages School of Information Engineering Minzu University of China School of Mathematical Sciences Inner Mongolia University Key Laboratory of Image Processing and Intelligent Control of Education Ministry of China School of Artificial Intelligence and Automation Huazhong University of Science and Technology

controlling networks aims to study the models, structures,and related dynamics of complex networks. The primary problem of controlling networks is to determine whether they are controllable. Nowadays, controllability has been widely studied and applied to system engineering and control theory, power systems, aerospace, and quantum systems. Various classical criteria include the Gram matrix criterion, Kalman rank criterion, and PBH test.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Prescribed-Time Stabilization of Singularly Perturbed Systems

引用

IEEE/CAA Journal of Automatica Sinica 2023年第2期10卷 569-571页

作者： Yan Lei Yan-Wu Wang Xiao-Kang Liu Wu Yang the School of Artificial Intelligence and Automation Huazhong University of Science and Technology(HUST)Wuhan 430074 Key Laboratory of Image Processing and Intelligent Control Ministry of EducationWuhan 430074China the School of Electrical Engineering Guangxi UniversityNanning 530004China

Dear Editor, This letter investigates the prescribed-time stabilization of linear singularly perturbed systems. Due to the numerical issues caused by the small perturbation parameter, the off-the-shelf control design techniques for the prescribed-time stabilization of regular linear systems are typically not suitable here. To solve the problem, the decoupling transformation techniques for time-varying singularly perturbed systems are combined with linear time-varying high gain feedback design techniques.

关键词： prescribed stabilization perturbed

来源：评论

学校读者我要写书评

暂无评论

Portrait Matting via Semantic and Detail Guidance 1

引用

7th Asian Conference on Pattern Recognition, ACPR 2023

作者： Wang, Deming Gao, Changxin Sang, Nong Key Laboratory of Image Processing and Intelligent Control School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9783031476341

ISBN: (纸本)9783031476334

Portrait matting is a challenging computer vision task that aims to estimate the per-pixel opacity of the foreground human regions. To produce high-quality alpha mattes, the majority of available methods employ a user-supplied trimap as an auxiliary input. However, obtaining trimap is difficult, which restricts the application of these methods. Recently, trimap-free approaches have been developed, but the alpha quality is significantly lower than that of trimap-based methods due to the lack of guidance. In this paper, we propose a portrait matting framework via semantic and detail guidance. We use the mask obtained from portrait segmentation as semantic guidance to improve semantic understanding of portrait matting, and we present a detail encoder block to capture the low-level feature around the subject as detail guidance, which helps to refine alpha details, such as hair. The proposed approach achieves remarkable results on both composition and real-world benchmarks. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

HTRNN: A New Protein Nuclear Localization Prediction Tool 7

HTRNN: A New Protein Nuclear Localization Prediction Tool

引用

7th IEEE Advanced Information Technology, Electronic and Automation control Conference, IAEAC 2024

作者： Miao, Xinyu Li, Wei Huazhong University of Science and Technology School of Artificial Intelligence and Automation Wuhan China Image Processing and Intelligent Control Key Laboratory of the Education Ministry of China Wuhan China

ISBN: (纸本)9798350339161

Current protein nuclear localization assays encounter multiple challenges that underscore the constraints of conventional biochemical assays and sequence-based procedures. This paper highlights the emerging interest in utilizing artificial intelligence to surmount these limitations. Specifically, a supervised deep learning algorithm, employing the HTRNN model, is presented to identify mechanisms responsible for the nuclear location of proteins. This methodology embraces an entire data-driven end-to-end structure, making expert experience or previous biological knowledge unnecessary. By utilizing amino acid sequence features from both the head and tail regions, the algorithm displays remarkable detection and generalization capabilities, which have been validated by experimental results on present tagged nuclear location protein datasets. Overall, the data-driven algorithm plays a crucial role in enhancing the identification of protein nuclear localization properties. © 2024 IEEE.

关键词： Proteins

来源：评论

学校读者我要写书评

暂无评论

Fusion-CAM: Fine-Grained and High-Faithfulness Explanations for Deep Convolutional Network via Hierarchical Fusion 7

Fusion-CAM: Fine-Grained and High-Faithfulness Explanations ...

引用

7th IEEE Advanced Information Technology, Electronic and Automation control Conference, IAEAC 2024

作者： Zheng, Chen Li, Wei Huazhong University of Science and Technology School of Artificial Intelligence and Automation Wuhan China Image Processing and Intelligent Control Key Laboratory of the Education Ministry of China Wuhan China

ISBN: (纸本)9798350339161

The interpretability of the convolutional neural networks(CNNs) has become a research hotspot. A popular explanation method is based on Class Activation Mapping (CAM), which visualizes the salient regions most relevant to neural network decisions. However, many CAM methods use the feature maps produced by the final convolution layer to generate the class activation maps, which usually have a low spatial resolution and can only generate coarse-grained visual explanations that provide a rough spatial location of the target object. In this paper, we propose a novel CAM method named Fusion-CAM. It improves traditional CAM methods by combining final class activation map containing semantic information with intermediate layer class activation maps containing fine-grained details, to generate fine-grained visual explanations with high faithfulness. In order to obtain high-quality intermediate layer class activation maps, we utilize Layer-wise Relevance Propagation (LRP) to obtain the weighting components of each channel of the intermediate layer feature maps, and the intermediate layer class activation maps generated by weighted summation are less noisy and have clear fine-grained details, which help to improve the quality of the final class activation map. Qualitative and quantitative experiments show that Fusion-CAM can be easily attached to different CAM methods to improve their performance. © 2024 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Cross-user Gesture Recognition Based on Deep Domain Adaptation Networks 43

Cross-user Gesture Recognition Based on Deep Domain Adaptati...

引用

43rd Chinese control Conference, CCC 2024

作者： Fu, Zhongzheng He, Xinrun Wang, Haoyuan Chen, Xingjian Li, Yixuan Huang, Jian Huazhong University of Science and Technology Ministry of Education Key Laboratory of Image Processing and Intelligent Control Hubei Key Laboratory of Brain-inspired Intelligent Systems School of Artificial Intelligence and Automation Wuhan430074 China

ISBN: (纸本)9789887581581

This study introduces an innovative approach for gesture recognition in smart wearable devices using a deep domain adaptation model, focusing on the challenges posed by heterogeneous user environments and the need for precise, adaptable, and personalized gesture recognition systems. Traditional methods based on machine learning and deep learning techniques, while effective, struggle with the variance in data distribution among different users, leading to accuracy challenges. To address this, the study proposes the Domain Kernel-Alignment Adaptation Network (DKAN) model, utilizing a new Kernel-Aligned Multi-kernel Maximum Mean Discrepancy (KAMMD) method within deep networks. By calculating kernel alignments between source and target domains and adjusting weights of Gaussian kernels accordingly, DKAN emphasizes features conducive to transfer learning, enabling more accurate cross-user gesture recognition. The model achieves a significant average accuracy of 97.3% across 21 gestures, highlighting its potential in practical applications. This advancement addresses challenges in traditional gesture recognition methods and sets a new direction for future research in smart wearable technology. © 2024 Technical Committee on control Theory, Chinese Association of Automation.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

A Dual-branch Model for Early Detection of Alzheimer's Disease Using Resting-State fMRI 7

A Dual-branch Model for Early Detection of Alzheimer's Disea...

引用

7th IEEE Advanced Information Technology, Electronic and Automation control Conference, IAEAC 2024

作者： Wang, Yixuan Li, Wei Huazhong University of Science and Technology School of Artificial Intelligence and Automation Wuhan China Image Processing and Intelligent Control Key Laboratory of the Education Ministry of China Wuhan China

ISBN: (纸本)9798350339161

Alzheimer's disease (AD) is the most prevalent form of dementia, and early diagnosis is crucial for delaying and treating AD. Resting-state functional magnetic resonance imaging (rs-fMRI), a widely used medical imaging technique, offers rich temporal and spatial data, which has led researchers to explore various feature extraction methods based on rs-fMRI images for AD identification. However, the related work still suffers from insufficient utilization of temporal and spatial information which leads to unsatisfactory early diagnosis. In this study, we propose a dual-branch fusion model to extract spatial-temporal features from rs-fMRI images. Our proposed model can extract temporal features at different levels. We developed a Class Activation Sequence (CAS) branch, which is a structure that emphasizes the function of each temporal node throughout the whole time series. Additionally, we created a time-domain local branch for local feature extraction. Further, we designed a fusion module for the model to describe temporal contextual relationships and fuse features at various levels. We tested the performance of the model on the ADNI dataset, and the experimental results show that compared with other algorithms, the dual-branch fusion model achieves higher classification accuracy on several classification tasks including early diagnosis, which proves the advantage of the dual-branch fusion model in temporal and spatial feature extraction for rs-fMRI images, and our work also provides a foundation for the temporal domain characterization of rs-fMRI images. © 2024 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

image Sample Generation of Stator Surface Defects Based on Layer Mask Blending Generative Adversarial Network 23

Image Sample Generation of Stator Surface Defects Based on L...

引用

9th International Conference on Computing and Artificial Intelligence, ICCAI 2023

作者： Li, Wenzheng Tian, Lihua Sun, Zhigang Xiao, Li School of Artificial Intelligence and Automation Key Laboratory of Image Processing and Intelligent Control Huazhong University of Science and Technology Wuhan430074 China

ISBN: (纸本)9781450399029

In industrial production processes, defect inspection plays an important role in reducing the occurrence of failures and improving production efficiency. Data-driven algorithms represented by deep learning have made great progress in recent years, but need to face the problems of small quantity and poor quality of datasets when applied to industrial defect inspection. This paper proposes a layer mask blending-based generative adversarial network (LMBGAN) and optimizes the training process to generate high-quality surface defect samples. LMBGAN generates defect images and layer masks using the defect image decoder and layer mask decoder with the Pixel Shuffle operation. Inspired by the layer mask in computer painting, LMBGAN adopts the input image as the base layer and blends the defect foreground through the layer mask, giving it the ability to focus more on generating upper-layer defect images and reducing unnecessary background changes. LMBGAN additionally introduces adaptive discriminator augmentation and non-saturating logistic loss to promote model convergence under small datasets, effectively alleviating the problem of GAN training difficulties with limited data. The experiment results show that the proposed method can generate high-quality and diverse defect image samples through easily accessible normal samples, thus reducing the difficulty of obtaining rare defect image samples. © 2023 Copyright held by the owner/author(s).

关键词： Generative adversarial networks

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：