检索结果-内蒙古大学图书馆

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

FIFAWC:a dataset with detailed annotation and rich semantics for group activity recognition

引用

Frontiers of Computer Science 2024年第6期18卷 271-272页

作者： Duoxuan PEI Di HUANG Yunhong WANG State Key Laboratory of Software Development Environment School of Computer Science and EngineeringBeihang UniversityBeijing 100191China Intelligent Recognition and Image Processing Lab. School of Computer Science and EngineeringBeihang UniversityBeijing 100191China

1 *** Activity Recognition(GAR),which aims to identify activities performed collectively in videos,has gained significant attention *** conventional action recognition centered on single individuals,GAR explores the c... 详细信息

关键词： has collective gained

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

Short-Term Residential Load Forecasting With Baseline-Refinement Profiles and Bi-Attention Mechanism

引用

IEEE TRANSACTIONS ON SMART GRID 2024年第1期15卷 1052-1062页

作者： Xiao, Jiang-Wen Liu, Peng Fang, Hongliang Liu, Xiao-Kang Wang, Yan-Wu Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol Key Lab Image Proc & Intelligent Control Minist Educ Wuhan Peoples R China

With the development of smart grid and renewable energy technologies, residential load forecasting has become an increasingly important task. Short-term residential load forecasting is not only conducive to power dispatching and peak-shaving and valley filling of the grid, but also good for residents to obtain higher economic benefit from renewable energy. This paper proposes a new baseline-refinement forecasting framework consisting of two main steps, baseline profile construction and refinement predictions. Firstly, a baseline profile construction method is proposed to forecast the baseline load profile based on the similarity and cyclic patterns of daily load profiles. Secondly, for the refinement predictions, a bi-attention mechanism is proposed by combining self-attention mechanism and external attention mechanism to fulfill the feature transformation and is included in the temporal convolutional network to refine the baseline profile. The final load forecasting results are obtained by aggregating the baseline profile and the refinement predictions. The simulation results demonstrate that the proposed framework has smaller forecasting errors and higher forecasting stability than the commonly used models on two load forecasting metrics.

关键词： Short term residential load forecasting temporal convolutional network bi-attention mechanism

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

Finite-Time Output Synchronization of Multiple Weighted Reaction-Diffusion Neural Networks With Adaptive Output Couplings

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2024年第1期35卷 169-181页

作者： Qiu, Qian Su, Housheng Huazhong Univ Sci & Technol Key Lab Image Proc & Intelligent Control Educ Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Wuhan 430074 Peoples R China

This article mainly considers the output synchronization (OS) problem of multiple weighted and adaptive output coupled reaction-diffusion neural networks (RDNNs) without and with coupling delays in finite time. Without coupling delays, an adaptive control law and an output feedback controller are, respectively, proposed to ensure that the multiple weighted and output coupled RDNNs are output synchronized and $H_{infinity}$ output synchronized in finite time. With coupling delays, an adaptive coupling weights control scheme and a novel feedback controller are put forward to make the multiple weighted RDNNs with output couplings achieve OS in finite time. Moreover, the finite-time $H_{infinity}$ OS is considered in the presence of external disturbances. By the Lyapunov approach, several finite-time OS and $H_{infinity}$ OS criteria are given. Finally, two simulation examples are presented to justify the effectiveness of the proposed adaptive control laws and controllers.

关键词： Couplings Synchronization Artificial neural networks Delays Adaptive control Adaptation models Mathematical models Adaptive output couplings coupling delays finite time multiple weighted reaction-diffusion neural networks RDNNs output synchronization OS and H∞ OS

来源：评论

学校读者我要写书评

暂无评论

Passive Lower Limb Exoskeleton for Kneeling and Postural Transition Assistance With Expanded Support Polygon

引用

IEEE-ASME TRANSACTIONS ON MECHATRONICS 2024年第2期29卷 1193-1204页

作者： Sugiura, Sojiro Zhu, Yaonan Huang, Jian Hasegawa, Yasuhisa Nagoya Univ Dept Micronano Mech Sci & Engn Nagoya 4648601 Japan Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Key Lab Image Proc & Intelligent Control Wuhan 430074 Peoples R China

Robotic exoskeletons, which assist in stand-to-kneel and kneel-to-stand (STK-KTS) movements and static kneeling postures are in great demand in the nursing field. This movement involves continuous adjustment of the center of gravity without a sufficient support polygon, which enhances the required joint effort of the ankle and knee. This study proposes a novel passive lower limb exoskeleton to support the movement. The exoskeleton was attached to the right leg and comprised a gas spring. The design followed an assistive strategy of the expanded support polygon. During the STK-KTS, the gas spring provided extra contact with the ground, thereby expanding the support polygon to increase motion stability and propping the knee to provide torque to the leg. The effectiveness of the gas spring was analyzed using a Lagrange dynamics-based simulation. Moreover, it was confirmed that the support polygon was expanded due to the proposed exoskeleton in real-world experiments. Further, experiments with seven healthy subjects showed that the exoskeleton reduced the time-integrated myoelectric potentials of the legs during STK-KTS (13.6%) and static posture (37.9%). These results imply that the proposed exoskeleton has the potential to reduce physical loads and provide a comfortable working environment for nursing workers.

关键词： Exoskeleton lower limb support stand-to-kneel and kneel-to-stand (STK-KTS) movement work assistance

来源：评论

学校读者我要写书评

暂无评论

Adaptive Position Constrained Assist-as-Needed control for Rehabilitation Robots

引用

IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS 2024年第4期71卷 4059-4068页

作者： Cao, Yu Chen, Xinkai Zhang, Mengshi Huang, Jian Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Key Lab Image Proc & Intelligent Control Wuhan 430074 Peoples R China Shibaura Inst Technol Dept Elect Informat Syst Saitama 3378570 Japan

In rehabilitation practice, motivating patients with neurological injuries to actively increase muscle activity and ensure their safety are important. Therefore, this study proposed a position-constrained assist-as-needed (AAN) control method for rehabilitation robots. A human-robot interaction system with position constraints was first established based on prescribed performance. Aiming at implementing the AAN strategy, the robot assistance level metric (RALM), a constructed global continuous differentiable function incorporating dead zone and saturation characteristics, was introduced to quantify the robotic assistance and facilitate seamless operation. To bridge the gap between the position constraints and the AAN strategy, a sliding manifold was constructed for the constrained human-robot dynamic system, where RALM was regarded as a weight factor to achieve a human-dominated mode, a robot-dominated mode, and their smooth transition, regarded as a human-robot shared mode. The stability of the closed-loop system was guaranteed by using the Lyapunov theory, and the proposed controller was verified by several physical experiments on a knee exoskeleton driven by pneumatic muscles.

关键词： Robots Robot sensing systems Robot kinematics Muscles Human-robot interaction Force Training Assist-as-needed (AAN) strategy position constraints robot assistance level metric (RALM)

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Task-Incremental Learning with Feature-Space Initialization Inspired by Neural Collapse

引用

NEURAL processing LETTERS 2023年第8期55卷 10811-10827页

作者： Zhou, Qinhao Xiang, Xiang Ma, Jing Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat MoE Key Lab Image Proc & Intelligent Control 1037 Luoyu Rd Wuhan 430074 Hubei Peoples R China

Incremental learning models need to update the categories and their conceptual understanding over time. The current research has placed more emphasis on learning new categories, while another common but under-explored incremental scenario is the updating and refinement of category labels. In this paper, we present the Hierarchical Task-Incremental Learning (HTIL) problem, which emulates the human cognitive process of progressing from coarse to fine. While the model learns the fine categories, it gains a better understanding of the perception of coarse categories, thereby enhancing its ability to differentiate between previously encountered classes. Inspired by neural collapse, we propose to initialize the coarse class prototypes and update the new fine class using hierarchical relations. We conduct experiments on diverse hierarchical data benchmarks, and the experiment results show our method achieves excellent results.

关键词： Hierarchical classification Incremental learning Neural collapse

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：